From 58a53148d46e447f44ea54d7df1fecc56d6b82c9 Mon Sep 17 00:00:00 2001 From: Saturneric Date: Tue, 1 Sep 2020 01:13:29 +0800 Subject: [PATCH] Add --- .gitignore | 116 + .idea/.gitignore | 5 + .idea/bert.iml | 11 + .idea/dataSources.xml | 11 + .idea/dictionaries/Administrator.xml | 22 + .../inspectionProfiles/profiles_settings.xml | 6 + .idea/misc.xml | 13 + .idea/modules.xml | 8 + .idea/other.xml | 7 + .idea/sqldialects.xml | 7 + .idea/vcs.xml | 6 + __init__.py | 15 + bptdata.db | Bin 0 -> 40960 bytes .../bert_config.json | 19 + .../bert_model.ckpt.index | Bin 0 -> 3541 bytes .../bert_model.ckpt.meta | Bin 0 -> 3186087 bytes chinese_wwm_ext_L-12_H-768_A-12/vocab.txt | 21128 + create_pretraining_data.py | 469 + dealing_dataset.py | 49 + extract_features.py | 419 + modeling.py | 986 + modeling_test.py | 277 + optimization.py | 174 + optimization_test.py | 48 + ...ng_movie_reviews_with_bert_on_tf_hub.ipynb | 1231 + run_classifier.py | 1056 + run_classifier_with_tfhub.py | 314 + run_pretraining.py | 493 + run_squad.py | 1283 + server.py | 417 + tmp/epout/checkpoint | 6 + tmp/epout/eval.tf_record | Bin 0 -> 8382055 bytes ...fevents.1586543049.iZ8vbescrakld4m4drzcktZ | Bin 0 -> 2277221 bytes tmp/epout/eval_results.txt | 4 + ...fevents.1586536204.iZ8vbescrakld4m4drzcktZ | Bin 0 -> 15951445 bytes tmp/epout/graph.pbtxt | 592992 +++++++++++++++ tmp/epout/model.ckpt-14062.index | Bin 0 -> 22717 bytes tmp/epout/model.ckpt-14062.meta | Bin 0 -> 4075388 bytes tmp/epout/train.tf_record | Bin 0 -> 83795018 bytes tmp/eppredict/predict.tf_record | Bin 0 -> 48757 bytes tmp/eppredict/test_results.tsv | 134 + tokenization.py | 399 + tokenization_test.py | 137 + 43 files changed, 622262 insertions(+) create mode 100644 .gitignore create mode 100644 .idea/.gitignore create mode 100644 .idea/bert.iml create mode 100644 .idea/dataSources.xml create mode 100644 .idea/dictionaries/Administrator.xml create mode 100644 .idea/inspectionProfiles/profiles_settings.xml create mode 100644 .idea/misc.xml create mode 100644 .idea/modules.xml create mode 100644 .idea/other.xml create mode 100644 .idea/sqldialects.xml create mode 100644 .idea/vcs.xml create mode 100644 __init__.py create mode 100644 bptdata.db create mode 100644 chinese_wwm_ext_L-12_H-768_A-12/bert_config.json create mode 100644 chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.index create mode 100644 chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.meta create mode 100644 chinese_wwm_ext_L-12_H-768_A-12/vocab.txt create mode 100644 create_pretraining_data.py create mode 100644 dealing_dataset.py create mode 100644 extract_features.py create mode 100644 modeling.py create mode 100644 modeling_test.py create mode 100644 optimization.py create mode 100644 optimization_test.py create mode 100644 predicting_movie_reviews_with_bert_on_tf_hub.ipynb create mode 100644 run_classifier.py create mode 100644 run_classifier_with_tfhub.py create mode 100644 run_pretraining.py create mode 100644 run_squad.py create mode 100644 server.py create mode 100644 tmp/epout/checkpoint create mode 100644 tmp/epout/eval.tf_record create mode 100644 tmp/epout/eval/events.out.tfevents.1586543049.iZ8vbescrakld4m4drzcktZ create mode 100644 tmp/epout/eval_results.txt create mode 100644 tmp/epout/events.out.tfevents.1586536204.iZ8vbescrakld4m4drzcktZ create mode 100644 tmp/epout/graph.pbtxt create mode 100644 tmp/epout/model.ckpt-14062.index create mode 100644 tmp/epout/model.ckpt-14062.meta create mode 100644 tmp/epout/train.tf_record create mode 100644 tmp/eppredict/predict.tf_record create mode 100644 tmp/eppredict/test_results.tsv create mode 100644 tokenization.py create mode 100644 tokenization_test.py diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..df9efad --- /dev/null +++ b/.gitignore @@ -0,0 +1,116 @@ +# Initially taken from Github's Python gitignore file + +# Byte-compiled / optimized / DLL files +__pycache__/ +*.py[cod] +*$py.class + +# C extensions +*.so + +# Distribution / packaging +.Python +build/ +develop-eggs/ +dist/ +downloads/ +eggs/ +.eggs/ +lib/ +lib64/ +parts/ +sdist/ +var/ +wheels/ +*.egg-info/ +.installed.cfg +*.egg +MANIFEST + +# PyInstaller +# Usually these files are written by a python script from a template +# before PyInstaller builds the exe, so as to inject date/other infos into it. +*.manifest +*.spec + +# Installer logs +pip-log.txt +pip-delete-this-directory.txt + +# Unit test / coverage reports +htmlcov/ +.tox/ +.nox/ +.coverage +.coverage.* +.cache +nosetests.xml +coverage.xml +*.cover +.hypothesis/ +.pytest_cache/ + +# Translations +*.mo +*.pot + +# Django stuff: +*.log +local_settings.py +db.sqlite3 + +# Flask stuff: +instance/ +.webassets-cache + +# Scrapy stuff: +.scrapy + +# Sphinx documentation +docs/_build/ + +# PyBuilder +target/ + +# Jupyter Notebook +.ipynb_checkpoints + +# IPython +profile_default/ +ipython_config.py + +# pyenv +.python-version + +# celery beat schedule file +celerybeat-schedule + +# SageMath parsed files +*.sage.py + +# Environments +.env +.venv +env/ +venv/ +ENV/ +env.bak/ +venv.bak/ + +# Spyder project settings +.spyderproject +.spyproject + +# Rope project settings +.ropeproject + +# mkdocs documentation +/site + +# mypy +.mypy_cache/ +.dmypy.json +dmypy.json + +# Pyre type checker +.pyre/ diff --git a/.idea/.gitignore b/.idea/.gitignore new file mode 100644 index 0000000..cf30ded --- /dev/null +++ b/.idea/.gitignore @@ -0,0 +1,5 @@ + +# Default ignored files +/workspace.xml +# Datasource local storage ignored files +/dataSources.local.xml \ No newline at end of file diff --git a/.idea/bert.iml b/.idea/bert.iml new file mode 100644 index 0000000..6a3f7ec --- /dev/null +++ b/.idea/bert.iml @@ -0,0 +1,11 @@ + + + + + + + + + + \ No newline at end of file diff --git a/.idea/dataSources.xml b/.idea/dataSources.xml new file mode 100644 index 0000000..eec2c94 --- /dev/null +++ b/.idea/dataSources.xml @@ -0,0 +1,11 @@ + + + + + sqlite.xerial + true + org.sqlite.JDBC + jdbc:sqlite:C:\Users\Administrator\Documents\GitHub\bert\bptdata.db + + + \ No newline at end of file diff --git a/.idea/dictionaries/Administrator.xml b/.idea/dictionaries/Administrator.xml new file mode 100644 index 0000000..b033f19 --- /dev/null +++ b/.idea/dictionaries/Administrator.xml @@ -0,0 +1,22 @@ + + + + amki + asctime + badrequest + bptdata + codedream + epaper + epout + eppdt + eppdtout + eppredict + idcode + levelname + nlpdata + sckstn + stnid + stns + + + \ No newline at end of file diff --git a/.idea/inspectionProfiles/profiles_settings.xml b/.idea/inspectionProfiles/profiles_settings.xml new file mode 100644 index 0000000..105ce2d --- /dev/null +++ b/.idea/inspectionProfiles/profiles_settings.xml @@ -0,0 +1,6 @@ + + + + \ No newline at end of file diff --git a/.idea/misc.xml b/.idea/misc.xml new file mode 100644 index 0000000..fb94267 --- /dev/null +++ b/.idea/misc.xml @@ -0,0 +1,13 @@ + + + + + + + + + + \ No newline at end of file diff --git a/.idea/modules.xml b/.idea/modules.xml new file mode 100644 index 0000000..e84c31f --- /dev/null +++ b/.idea/modules.xml @@ -0,0 +1,8 @@ + + + + + + + + \ No newline at end of file diff --git a/.idea/other.xml b/.idea/other.xml new file mode 100644 index 0000000..640fd80 --- /dev/null +++ b/.idea/other.xml @@ -0,0 +1,7 @@ + + + + + \ No newline at end of file diff --git a/.idea/sqldialects.xml b/.idea/sqldialects.xml new file mode 100644 index 0000000..5b66d9a --- /dev/null +++ b/.idea/sqldialects.xml @@ -0,0 +1,7 @@ + + + + + + + \ No newline at end of file diff --git a/.idea/vcs.xml b/.idea/vcs.xml new file mode 100644 index 0000000..94a25f7 --- /dev/null +++ b/.idea/vcs.xml @@ -0,0 +1,6 @@ + + + + + + \ No newline at end of file diff --git a/__init__.py b/__init__.py new file mode 100644 index 0000000..effb57b --- /dev/null +++ b/__init__.py @@ -0,0 +1,15 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + diff --git a/bptdata.db b/bptdata.db new file mode 100644 index 0000000000000000000000000000000000000000..7e4b91dbe078a0f35a26c992d833b591e67ce20f GIT binary patch literal 40960 zcmeHw349#Yk^V^2BOjW>1dK6YEJ&O{0OKPB2mxcjG3K&k8ysMXQO2^ZU~D4WfJ02C zb>D}r!#Zq9wk{uqG#ll^ZtB%8nd z13hWx^*gHS)vH(az3O%Ey+3zXPS)lfc~52U%3A7}w>#aHL)2a5!e-e+Bgy>AOi^Hp0#g*2qQDddrYJB) zfhh`1QDBM!Qxuq@!2eYWJerm5Cm_9d|f3In=8!**#ZDSD#cqrUeh{ zrDu)fp;$N??@?osu2^KBR^A>j7}0{2T4UjB!<9yTAJu87OdLF{)rPd1I&HWn9x99t z54<+qs<%~X<(+6DK2~c4oAka;?NEhY+pk3ax6Uf!+M zm8Y~28)?(_t6FKR-Z-EKdbQGG{ZQ@1np$x)YGRu4kx(ppCKl<%Q$h`s4mIUkXh@6H z8!d&>soprGcb4i6FKWRSP{5+0v>=3v2S?+6)P=~|sXM@i3Lri{kI0lle5 zE3VQ{bSBgit#lwh)~t7&#q1JI!NlPM7=SiVY_y)ztGi<(W8l6qykF}-5DWJc1sVqQ zsvwaoP#FtXCE5ZI8);1o_h?1kb}@*LwZ=Ou6D3tzSCxE@2NTUzv5{)6szIhngFn?C z9~s$vF+}nKWkfHni$zZ)4i{=&CDOC>sYUh+Lu7#QU<7hu^j2x%aJ*N=(9jh=Nz^yT zLl8#l*(e**Mv1PVu2x$ut*xJpey&l5!PHo!8XxOz?TO5SG@Xw()P(f#_NHb@V8(TdLKZDo2xDafl87g^&^SpaAdJ&Fiw1A|%(7Gby6 z)r9B~CJ}Ph3RZOaiILRu>8kKyWXc^TaJ-KfJd)FyLXcEGq7Zhp5r-h%X%W9xF zUUEz;@6*}_5(Rz6$cr>;$Q;^EE-M+GHg-G_XqQgpD;bWOb^&0&8&Km%iioSN9eRIP$|KZlS9`qrtTx(&!Hf@Ms3OsHN`@gr1uJ9>D`sK3 znaWv_IqM0nBw|z#=nX}Q7Se5cOFo#SH+ASI8}ypm*l1F1gZ)gQ{FOc^W}k9K+OZRl z9=CKnDF02)em+jX=q;g{CH&|kgtvG$J9t8B=^R6=&U}NkEe|O zGHtX7je{GQon1ho#gHZ)qpvPlXZWa}p3;hmq? zRtwls2HzK31=_g8QUmFmf$=CO+0~)`!ZdDTjDywLQj(9<8=i zsIl42TgpkSG^pZEP>5T; z>FrD2u@#%CR?}on9$QWPH0`-E^W+T2Ehym)8C1*wj9ui(Y5^7WPz^S5X~90Z1iIz^ zF{hKocf&u-MYK&ILd-G%AdeKOj?nLI% zcGZ6bMuO1TZgN6p>1BcV*Z|S84CZ8+7#OqzL7YN0WL?Nyu=ayw>BK{)^aI%QsztSH z;h)uw9@GXuR(e!;<;~&1*GPiJnrk`Ea#tJUVQw#sT@zWSH8m74>bL zhL)BoFc~7Tv4ioDYHoKU`)LIxwHTROL1Hwj-S_|ojA5Jm6%<<0S!`vPQGB$3tSw$F z7zb1eY+-d)2B>~Bim$4!4`qxfL-_JZr}YJEiN+t=qN#RNDj+xUeWk$ zvy$@*sWZ{4B`39kI&`^o$x76`3q$V+Yhy4BX`8Xey##;T32)Vd*=nEEw4*%q^j*gBTVpmzGGT*l0EDfR0S=*ibvRc6&{AsS^y|qKs=> zB04XFRGOZBF+O&PR%7V^j(%~qo;Fw7M9Z|Q{X$)=6N_B(iD?1UhxAIEm8zvp*e9e@ z>iej2(h4@We3Cp^f+sAJEZ@Ek)@|DP6M(0 z3e@}fSJFP2vV7u|KZ!1_FpQ7s{oRRUU6y*&`fE)+5go?n1I-TWfW{WxPkT{ii541` zL!%t4tntVmC^JI+agD*JFRd2G`HP2ch36-}OMMsm z+&-8058hvSf9g$mU-5p&`!(+u-D`Zm_5B=v0nPV)-#2|<@qNPgd0*6b%GdAf@wNL7 z`09M+z5?Gze7k+yeOr7__#W}C_1)*Y+jpDqoxU4f;*t^<$k9WEEUEZbM>%0rRS9s@mXL)CM(>=fQ=$@B7-}1cV z`GV(@o?*{vPtdd1)9dN*G<)`WDm;arXFNMRTRqvHM?LF2_j^`&ZueZ{xyiH0bCu^Z z&qW@OC&T?m_pjYQbN|Twb@!LtpLKuCJ>-7X{ayEq?&Iz*_d$1~yT)DW4!EClKkeS? ze!~4e_dV`A+&8-yxvzF-xi4|ga!+?VGJl=<)69704>G@%`IXGiWqvF(oH>x$m)Sw& zoBB6Jfhh`1QDBM!Qxuq@z!U|hC@@8V|1AnkbEc%MFfJE+2NZaJk;mOl8ii zjssl2;%MUXhmJ-rUv@Nb`8`KHm)~*h=knW*eN=9_)KSOfTt_XJmpE#;yx39Aih7E-!QBb9t#l z<#Mhgkl{>Eb6xULF6O+z#l?HMnEeqhE_$Ae3qQ=ooAz+wdyWh5vs`$d;ljO}i_BeI zT(FaiS$SN{{16v2p5|iu4lbr`=OW`NE?nEVP@d$%`9UtybE#1NxRw7(+rouoGk^Gp z94>ypiHraAbMdS;@r@-^0ZZR&eq1-CTVCE-t=zCl}vc&c%1`;NoAFaq;cjx%k#?TzvE0 zTzunJF24RQF1~gP7hk=Zi+{d}iYFWkVz z=NEGEx$CKLwp_3`vdv``S<8s#l z*L39xCC&Nf^ykx+wCfyCp!OvH)Kyn1=>-@15p1LA0C{xAn&D@HBO_jN3O}`QDdKH# zIf_REu3!Yiwf-K&yk4obR`jU4I(egk(o<-O4Iegvvi=7lD<@gI`YP>j#iNvDyMzmOO@7BvksSf-{ zC2)Z?8-r270!WN|_k41f`ODx&gy#zWCEzn?XzU_{L(H5(Q473bsnccR>GcS<9w7JV z0g7%>MDA?2_+R6vhTy^y#}v>Q`k9`1*^BUV8^=4;1y{gbcAvjgCPB`}0KA7KXhFQi z7z{kIhdRutZi<*5 z?AC@GAUh@r5wf*$Mgla#H1)B0VEaveINvzX&-|oLLlPAvA)=_1?Qt3dGn|AS{_MlW zT6;T!Y!n+h7(n0={&%nufjb%=2TqCL2s^OE$+Hj#;lCM!EdzL>NjL@~#6wjiDV==? z+*u!LHQjI_a%ihiKYKPdG)Uy)52Yq`+FKZ}EI_}4(s=xGF#lG65u{we9{2``8pOT8 zMq{)K!7v&~>xfp%4)Bgn8gknSqqh!XjphY8gWo0EG0p^D9_oV2n8^iNPkwBq!Hj^j zAVMOI2!hu~zzvefbLl6`AUSFZmszM1tOr^1w4|9EX4uBIhV}Z<6rRX32ASc&zOZ0?Mfb~}uFKELrY8rr5`A&ZT@gy{h?nsm(t;TR9HdG@L zWc6#x8p0>6F9^;MXu`^q;iV`As6wsyq#kNOyxbgixL*b-OP@&1*cWv!Kvwxn{T~&T zfvywHXDK*(E~i9Qk^GjGYr_CA6ic&cCl&^@qy@nRa4Ht=jt}(VcOu#g2To~2v|I!*rlt`L$?1Q=;!Dbm%%ziDbMaspu_ZjNPgq9!V2K3b zKzzjP%@s2YA|pad_9~(P}g3dlho(^ z&u7uU6vbdYLmLp|>1gAFW-L9ZpDD9TjR-877R@Mzfu$dz_h|{%5ZcRHMzZ?z-wF6p z;Zy!Sf|S87hNZc3$Br%Aa%kog;*T;6>4mmeKn2TyVF<*A32}AjinZ${LWz0DM~h-Z z6=1Z{Oz@wHZON7?$USt1=CH9$tTx^mHi5_?P)^{I?#$V~E1SN^+MKuJsjT&R z+1q#09o3w?oruM%*AiZ8ga28Jn#S=8Y^IQfiT#4x2#ymWcn*+FilCirAeO@dg_VS- zSSQNd?EFY>;w4631&L3RJ;E6MNC4mhZmL=Py#zVz~T4DzCuv5eA1Onr4 z1-hnn9fR1BBI7Ws))vv5fn>yvUX3NCA0*=uLX$XfMh~7wzaj^sinM_iz3rG?&cqs^ z=>W zP?_4Gi8aI4q76qdHMC%M0n}Z2Yc7yd>N@{U0_ZYtF?HfO?EWBuhv{>Z`BgXo=x1p5 z=vJ@B(-?IzmXgrG+`;o}OZs*cD&)DIH2=^;O0cy-zkfwv}SNEfW&YO!|Y zK_NEJ)n;PX7LtVvQ?9cQ;{>;3LuL30@ubCA6qXP5#%6#t0U8rc5VqO(Hd(D09B$_*6F1L^FRx-PDuNP z1NQ*|rOxqhvz9h(9l)f>nS$E@NMrdbsDT?M_J*&Z*YKTa#!F^8)0g^lmv7sawKjLl z)?GWHU3TW=J(IKPf*H>A1^z9IZ&|!#>Eeyq+n(NgifGaGI59u&|3KEdr?T^QWv$!|?X`VN){f0t%l+A#a-PcdXRXP88VYPfUhXb@ zlc7*EPo&RxI@6c?HzrC-jL1oWlf5?Fa@PY7tjfB3#lu+-FTRJ+3K#ev2Si5;6~jKn zT1Smn8auV36Ix9epJd?=Afysaeey?#X{yUNdUFulIKEk&=1kA=KZ^YW*g!&o*PKN| zz%1*Bd%ywhVEOgyHr%yj(UL{i(~m`qZ(c(9|G5D3`|kIC-8F_u4NbYAXF;w z2)H+`FrUOMFbZvK#AxU?{l0Rqp1O_T>X~>}tw-p3RB#9_$HqE!HKZyF;Z&ZXe4b_? z-pFt{1vogp1C8D0RG~piJ^?r%ZLtpwzyxz7ujxqFa3vBD1lI_QD1azEYHACG%NP*? z2R70Kz|d6JUJvi?0_AfEFG(=Zkc&8Sr7hSJ8>!+eVifv;ml3X1(`rwCB@@J5dA(y8 z*M_yuI(s$5Vi075JCl6vZ70YR%bs%MXavz!x)ltDFw@}CH%@iL0|kOvKrRD(F1Kyv zN?yDyB=H@yh2w`}5nMY2=7a7W%I!peBecM9{`1!<=|#^vTQ+P(dXRNb?RYYWILafY zXfBAw&SJe{7(Og^7Lx-Sqd_%^1XHA=y&LfefshcVLqif<1eQRKoBg#3s!86p*XJeJ-u$05WBeW1+A{4^XNBiQAFG0jM7&6KFj}?q`Jg!-06|2}>fu zFiJm*+)|m)07^lyyM{W?s1hVCF<8p>js5=quZ3NzL{j`N*|cS$mw# zB2h+Lg@g>Ws*^OFa6Sij%eAvFBBsySgW_>E2`~c9iu7YtW{8S~Mg%1mt-_s4ivJx_ z&BzExOeoIT9NO@9^nQbrug4?kBtZc|rdYV8@P0(M;^l)TP5|j2jvqiG5fUC1w=rpc z0^6Z0wFeJDb3obLevLEzqt2$K3$uiPQiG^nhjat&;CP~8jL2BkfvrdCh=)|V$-n(- z1akK}8y7Fk0=R@i3ZNbw*(w4R2aPC66xKN&q0dE@Sd|cG8S5nDK*kV&4B!|C#3|7X zXc(u(IB_BtE(3jpM~LhahJ&yBvE-D+qJh{@se~PHqn9rA(w*-*fnw0ItR-ZY7K#%7 zgYV`d#Ij@|BHl!flqt(%M$ip}qbZjIlS2tEw5kI7li>geRZz^oxLwbuWhFDRz*M5f zL2mkrRRduW5RsSbd(~0Q&TIh`W?;>YZ*!(U?h&jS}%_<&m1TB+JS*K#S;PgZjt<%w(x^ zKN2eKTK9AwLN2mz5e>^5jJ`1uMJ#`)TC|7`NIBL)QACiEg;jZ_$P#cl6t1hn3OY&z z>uySP)q>+-5H8zOn2mNj$|Ql{EGoz`6uCSQ54Gkm<3lK!(gP*cpwyABjoy7$*Fw1X zSR4LB$PMjFgiEAUUGWyoEz^nmlugNGAmwbSmI{eLy+$%duu2SIoCN5Jpb_EyTJQ|6 zRGUDd;l}uAoxKB@+d2A|AG9t_o6)-DnlZH@-?G=shOjBYXvngbl$nG?BzL$)8>rDn zOR<;T_+}E)x)fDU389KeU7L0W181;yh;iJ6y6h@g10QwPj?W9i%!n>Tbpc3$AnlEd7A>NCwEYyQh3=1^qKG%Ee|Vrvv?7AS;VUd5t>zdp ziuFCFAaagx31iqXj5-AIIn+k&ift1r(T0jFRVURFfqmNHL;14+do8@uSz|3I*d8oI zlx!$Pq^V^FvY0_m#z>-HvMzY@q|ru?AVw_VQVd|6FckXbhas8BEMYZ`XE5A#yY_6$ zc@_c5oIJEPzUHx&uguGS3X#lhd$QI&or3`BU0ZV~y$o&p{bd-EtV+-w;YgrkWi5~Q zix$S4qw2s_*tT=gyrP0+5_bYTtGblW0tEqDQ(hp>kOd(nqX$0+>cBJ$>m=$*E9oHo zWHO#~!+d9Yw(=RV3S-d*mRN;)S%p?yX=bb11dM6HD&^Dih&^co>^Reg zc->TCDRsm+H31XJY+)&N0Q7yGgkko|R?JX<8rX-mNDX59lE(|i1RxBctqk6tg)3$7 zxj?8QSJ9`ed*6E5QJ^Nwj1bBEYvYTXzmz?}5l9%5$tvtahzHpOFcY1PUr)E!d7X4;>L!3l!NpHZ&P8mJ0DHVX z7tgITITz2Z^C!4ytNdqi(bi5LAgCbi$5TivsZmHtBuPta%}S>OL%f=Fn;grM-tt!a zfyvB+!2H9rlB!JjO6UJs&P@*AfG^z}^}gVp>$%(gd-sBwZ<+p$X&;|PV5~~O1PRvX$r$as+|0{xvsG53N**2H;z?= z*QWmX{($&PaORH(2|9$YZSvSjls4_sE%9 zx`RDk;8T&-rg*+<9$DYD)zpy9`6J_lT?;%T0gL;QN&dI^-=i`zb&j}kU2=W#=) zn6hax0cd_MSELI%QgOC2`$S|4;?JPWR(y>n<#B2pjD*h(iOnQ~C^D1Px04yp2xOe0 zk)bqyj*2uug*Uh^H)SD-gExVDHgLau0+Nk`V{HxLbF+0E_E=~qLijA7qwoo`N|!?z z$SPkk8)|7J$O>i?C2-VYl}LcS1Q$OKKgu!~ScUeX7N zXiMXA_<2Dk^xj_AWiac`C#v{C*eSmEFe$u3`3nmB;` z3shtsDtxo+Qn-O+1dtc8gkOia=Aj%j4^>$RZd&*_2J?p!9+0V`4o)6y;S;fl%A5oj z;}db#S`Q}<;EqEwMQ$>=FCCbH69=Do@YL4~Fv*fN5f! z5F{CcEH;reVHO649T)mcIE7r<@gOBt7V?T&Vddr{2PEXd%(Q$G5daZLPNCRjCRh`@ zgQ6sq9ZeckA9?}juWMa%EXG;n!Y}9JorsWKW~_xK0z<46y^s{(|IYHXp0ZT9XaUHB z>~2t1@(IRf^=RMD@VV?lEA~QnXSyyXQL<%cgYTG}4tzU;u5rP?I!L!3<#W>|m(0`# zkDj{rBOuQL*KEm82$KXk-5}}2wB(V^(#hhpqoMIl1zFj&8L(wVHXoZ$U@vm0qgu)i zM-^aY;d8z^AQL5qv+}pP6PI^V3Xnr^0Av%EQdEtnmeMlXq&%^~7HQM}JZQ7hb&(Z> zvm0UDKP7M{S%mpodz>9=h~ON9XBCcg{j>^a2F8Vj0gDQs%O_los4GOe^~28eJ+2Eu zEBJjtA|f^2nB^L#Wn7M!jg=SAR7>pnS+R$?_a=o*@&%p}b;POd`3QB5x{a}5vu_*U0Oe-yn%Ze@9$YF{Wv5KKRGuaWIwLd}|N%AYC zhBVFAq``{Hq8E4t9*GTiP=iRpn>^ghbj8~&+s}i~V$$Yy@9E6;pl6TWU3nHO({=dYr%Hh2NdGs!Fzb7*`^TRXRX85LePIpcF z+_dX5T3yx3OUez-{PZuU7o>aBUP>!Xy9^Jc{!>@qgGDf3c@?o6G9{Zwalz3ZD@`hN ztWt98+ojJMw{?L?&M7pvA|yHP*Q%q5=B_x0A4n&Lsz{u<9TM9JT~Vg?nOPFx7M7u9 za))!Yc+H+&TX$@irrM46+U+Rk|Pdd6y zG?`qwxYkllSOhrITx%&N^KYVyO}}EdS@|Ax=2L!Xy%&Mn!^&b&p;dMep$Ce1PS&05 zS!5<3s@jFT9<*jB^TOOgF3LR3sn>5?{5&=o5N+?$B|GoND7Gjs%QxbG zl;ftfghyo*Vjcdev0RIJ%GhuO!`(Q{lRiqB`n`9d)7i@RQ`mUhW5QIspj?;r`qK5d zy68?U=4|DAIN^!2&{|6*DE$5b88t3c$xe9Qx$-dxPu;t2^@9=|o^<-OdAu&M&v{-b zZzrGn_T?DzL&|rpAzSMi{Ry-pi9#T|+vErHPG&HPfLc2mI~P(V32IiB%0$UHXRs4_ zt@qBCqd5FlILz%3g8Zr(sX|yji zgI5;HTb8(KU?5{{Yn^fMNStq%s_$Bc5$;z0g+|C(i6t;1CuM5Qb|;b)Fq=bi^qwei zkk)l~(Ookcm?fezHhgP+A_S$uR|0NBX&Cehgih$NBqOIr{?y>}XWovp+-&9BWKUx2 z;Fq4Dm-pmu&&gW9n_`@3#=1&q&X2Vobu}2BI6qprhbn|NGFLk2hDs*Rm7bkU&(bY4 z+T+Q;k}_TUgg{FgHO^~%$bc<`1t^TdHF_A5w2$*3Vd@zDqnMUf2>?#97@-sywVD5z zrs};7oSdh8%Pwe>$T|pRsb?Y0|Er`C?}Df8lbKw!=Dr&PTB>}L07y9Hk);AS!00)_ zQ4aiK+DHlmdc}r%$g-(y#C}C-lR-!R9R?qKGUyjz87pKb#)}qg#rZ`GA@fCqn#J(PjR!h6_f=ZxAa3(~oopxb9mdMp$xu$>4%#+7CGFowyBY7l{u7#9V&K8Rz zRE_sQkdhEl$%uIhIkw$`-Qf}CYeI93#mv{jt_m?NP5LJR1NLss9$^0@gQ`Mg`~72B zp&=xJ z1RfWS0wf$A)02Oq5z{Ae?no$=5?;}Anm{eFNW~L3VhJr*zRaiV3YnJVLji~|UX2KQ zRbr%;T4mR>3{dPQ5)menJv5I6ixZnYG_T(G4oK=8Y#U1VsY3!0QljWJKox-GKBEbqt4yi|r4%B{2Y!u4z-3R1U$Lm$h^u&KXz&@-`Zw zdx{pp@XdGCuq9wQplJJs`<>rAB16r3$3k*?!=yQW;EAP7uP>1{B!6LrsS}Zz1F4!M zA3;(xlA(noO0-d~=$P-JV3`#|4Kp)=LEM{QA=Cv+pl0qDGK<8s&Mjs7kl*>-#hn zW=vh;;6D8Tqk&EV^g!E{&w#@kA^t$QurX8F%H)F5sH?GZ!p)7O$0I5A6O?x#h z6OXB@{8;!GD*tH-GAX40tLgndfVo_&{LbXs-%WbIY^bv*mEWdh2$5O-CDZ#o!8-0Y zVzAMk^f#K`ZzEe-zc$(acb4Ao{b2sB%CDdi#2-RB=zlZ4-{Z{WUyAnmn@#Vx_a8up z=atv~N@>f~S&xBCS1SKO{_i(1yC9`ou#zRFCK!AaLvkv=untyYO5(`#1|}GM1Y=mL z{M>%pB74H-2q1aoT&T32zM%;Q-^cX%_rGR>!3~hwy~@vE*Zp-83_gthbCjQ&i~9{p zFqnVKLvVrZRQ`?p_OMm{ybOPk9`YE86ATjc_nzc$J+`vh%1{12lKid10^Fedczo;n zn@;k#7Nno6{OjLKlE3$Y_*W`FvLbRg*8G)|{H=j9TBjr^ncrVL$=_-)^C2bv|3#9& z2U*D*cI*!JU^>$N<&yk8fIjCay0x_boFsoY-|tM{>1g^N11cDPhoKBdC#5U~FC3ue5+NR(XOd5Cv@TWvX@QXbb!h0PhO_%6&?kX1b zSV~vti!7M|O%8Qi;Z8Jf1%O{C$W7<}3mrQgzQ??`dRDtDGYe<#oUwWOr>1=_W4-GK z%GaGe&a2Wlrq4)Q?HEI~Klk5*&tPWrTpnwfaLLL222ezHX}R-AezhcU!U`*hQ(HVC zIIV%!s%JO!)MYL=d6FV@70&WenP89PYgf(ddmIgu9>KZP7kQ2~3tGP@>ao*k}C_H6NM0oZRx+LJv$0QEqnVzSZ^Ih>Mj zyCj)lREZ#|(WyKlkdEK%E|q-m~M_BzBmtePZs1w=#q=XE5~ zY%suEnzB%>cuE@>!NB1OGuk;)%r88kHxmOmL+>`A^gzjYs~L~9e`q013`WS9xSOc$ zI!#R-f^K*}v_&U1?L%Ox(={{2e`4t_#0_!X%S<86?mE2FLh_l*1#Z(N&Y^9(vWS-f zw+`rK1f!Qr+Z5g7oZRsCnUNZPbp+9bGX~^e_cV;+M_e=Dl19IljMKN21GXHII`ee| z@Czni;BdYXQhdjj&mYw$CiADRZCS&4_osa2GCX`Mu_QDlk%=Ub^_)cgGP{(iS?^)M z5-}AMheRo+Qgs#`A!LBJgCl)CdI2I087IWI4w1)8J49^aJxHaD+pP!5<1Dv^ZBC8+ z={rEK3tZDJXqZJY!Ouxk$rVYjIiN+uNw(rhj1g!TSmrq65SSoLFGyW&pDAA9j8_p$ zl#pVwk~?>_2@Y$EPQT;H95kmB3Z#OShs8NyA*qKFqs zagsM!Ct7f!q!B+c!MyFXeYvJd=5ur+UM@LsOBm3S9iRj${Dc)OuLxl5e@Y}!F1g}8 zAGk-$ydrlqoli%{h+P~n&HvC-i0t`Y84^9ANlzwXSj}db-6m)R%jLxQCWUz3rbC=x zZeV4Jhla4;qCsddjB@+R?H>QT#L)B7Dy7a)&>Tx*e`+?&}y_C`HG6nI;CH` zbSK-v?G3_9B?W;%GTP2Kb98MsydCb5<(>; zHzDB&hG$V*o0LpcUPH;(IVAZer{>l!S?=;AGJFU7i0VH1sl;a?adk-yNc; ziPqH~8$E~{Hk1Q~@$bS`!KP7ow}J4n^0IA`>Tq|WgswbbnqVmCYytQ{e99TaLjv+T z7AY0l(sWY1R%YfHGW7Cieh><2w##Xa5G>G2DM8ipRbnO*DyO$9B`H7Xf&JJz@nC3{ zCj+08?rW|Ez7YF|R#Hq6bZjKJZUZ>@1Py{us5Ki<9y7KJvJiu`m`#vvAD;8-3;U{K zEQ<}cJlYC^A=pIHfvp-M0J*cF9rL}pAjUjbx~L+-ub)G_B;}xNuw=oXoiLN znuH+iC?jW=ynq1{KpqN=%)&-N!bI%xq5ZHojxjyNSRa6%@j@Qvfk~#36If+r5Zewq zfN!nOd2UzMUE8vE?#$hs>*v=s5XcQeYPwvV_R0Vi~d|Y-2BS|SE z=E&wo{u#!n*j|BwLn4l>z;c7O=36sBcM*O&rU7FE>?%HXh|JuAXn#+r}j$}myuo*&eb)W?v_@h}2| z8V?}HJ}nQDRlW-*f6}O!Z*mhCX@TOs^il?+sR-@U?TArT4qoTG>Sh--^{P$SmR2ag x$1udIz(}n78+J`Rks9Y)pC)-HTr!f&o08Um_?!eGfot}TFbnHOjJ$j6zW{Fe09610 literal 0 HcmV?d00001 diff --git a/chinese_wwm_ext_L-12_H-768_A-12/bert_config.json b/chinese_wwm_ext_L-12_H-768_A-12/bert_config.json new file mode 100644 index 0000000..adb75ff --- /dev/null +++ b/chinese_wwm_ext_L-12_H-768_A-12/bert_config.json @@ -0,0 +1,19 @@ +{ + "attention_probs_dropout_prob": 0.1, + "directionality": "bidi", + "hidden_act": "gelu", + "hidden_dropout_prob": 0.1, + "hidden_size": 768, + "initializer_range": 0.02, + "intermediate_size": 3072, + "max_position_embeddings": 512, + "num_attention_heads": 12, + "num_hidden_layers": 12, + "pooler_fc_size": 768, + "pooler_num_attention_heads": 12, + "pooler_num_fc_layers": 3, + "pooler_size_per_head": 128, + "pooler_type": "first_token_transform", + "type_vocab_size": 2, + "vocab_size": 21128 +} diff --git a/chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.index b/chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.index new file mode 100644 index 0000000000000000000000000000000000000000..8c80a011817568d562cde3f6f35ed943ac033e16 GIT binary patch literal 3541 zcmaKvYg`l87RAp%0uwxu@CpH92oe-j42Z#JQ2|i`m-YQ*jBsUT9d7OxfVKDqqv$IAzpnSEAf{ny@Sh7^YO zhaqG`Q|?wYg+vZYb*5gbPF1Q^Dot9FP8ywoj^zdRn~OA~C4!h>AtARW@QKz?WeX zWV2fCzvj8P>C;!K)8h3x8S3Ckh{tiXqOG`rvl|k8nE+X(U&jcZ{iy8pOjYoBBpBhy zq6?}iuK%(wD4XAiWOqOQ=`)*&{InJ6Ds`q*OWhL22S^k3dUYCQ6iCyv^ch+DxIV-a zGsb(K)E%L{tRy0U%}<*C5+M^uH1pG(x&veE_$qapPAyex5~*H4DPXj6yaUwFAGA-f z8MR8CnWokTjzBitBdD8_%x4Ug?7IC|&7NGoPOV)jU8T+miAHqyQ!%xb_o>y0hrV&q z2s1rmUv{)_tnTsxi5-7+mO3*u64A_CFDjDv(Xk&#jEjpzsGO2eWxiXlwp@cXYZA3t zp&f`W-{MV0;y<>AyDMQy11yOY#LU=r4^7}lrcB>X1!&Ur>daKNN|UHpOCutw$pSIc z@P)(|B9-1PouPP#=ED_3C0t?5EZJKAT%(H%e_aPf8`^!TbG*S`zn{iLbWo_4BjU~e z`RXfSES#7*-n>h?(YHp2L{doL-+|hQIbm*|89RCL1M70F=(b9KstUV5)`rS+tx!Kl zgb!8y*%X&+#VXj#o`V&~^Pef9V5B5oRrvm{wgo95G9H;oZN)C7(pJTK3PbZv@VU{r zr1J7xP|~LFq(D1sNy0N_uoY@0wAiz-Ymn4huLZ!F;E6n)BmvpOS!VR*1O?nAum~1( z{~9KqoPsx-uJ7x=xm`|2t9WHi9>#=pGu zI|g=X!~S4SnPB`+R!;w|M~51;eOD6)K4gR?q>IwKCw{kADRh`cZ!-kb38Pj zE@0mKY98-;A3bQP78pVRt$0OFJ^_XNsO`>^9YxcTf6O3@6&O&&K#s;)a~SOUvaeDx zXklflX8HB-C>76*ziEg?GY`TjGw-ULJ~1|;lS1X-TP7_&G%6NO8o)U}d&7C}5II#k zgqkTy-Vm?1sinuc36#xM^P#Q3cosCvng7P*xhXt64vtAo&x$X4DB=Ew1M~fQut$(; zbqYEACK9?j;mxJX!(FX~H2SdNCbJNPbnOK>09ETv|4+PZu`B>`to7wWXOyO93*OlYd^GYfI9>|J204 ziGyPZ2k$uL+^K|PNXF{`9wph)k^-9MR|*@rG^)_U5yGT><`i_7lAf&tUmw|O} z|MAn6!M13?0&?Zf^b%4V*tr~PJ>oS@07)ClmTTg0?c!ZL1>9s(5rl52j)l?PdOoim zMi;~A!ocLE;8IEV;UduYKaPXZRdpY>1-q<*Cu|Y64ZD&~{2IaUwh!0OBYw$VDica{ zn=61bs1BytRI>s&Z>ap2l9?y@D9u5(WaddeS=(35Bs0%^xx%jse_m9dm_p7@s(^D3 zZXX zoytbCMT3f9LG$lPz!?K2q7P@ zD$MNea7zJ^>CI~3G;y}6fpetR-qOt`*Sn_)No%1FydP>$-cKmIEhOT_AMPZz?g~BB zDn58pPP)nL(Q7`&*_$6Z63WR%uz;z{0?IEBPFpwJO_&s~r+CuCm_$!sX8>AFqBvyggF6xVWy%bD+zQY-V2+1kLwO z9>aDGNpM09Tu(vyZ989~ooj@^k?g5vu7p?2zx!fzNCgsY#Hb5yRw11G&yqz>UarA) zg^=E}E$nJu%Z65cKtwe{16)rYH++HrnAmE>GW2o6c%A2dvz-u9jYJ%KRLl19su##s zy&vYub&aE)>Ds|!b}262y!tmqvE}_^h8x7-&dI(T;!5CYv~3%Z9^7=;MHvpM5>gzV zRKX>+wtXc~oA@B~DM_K+_xLi#TNByUTJAb#p!O+x4G#z=yM#eoh$vTzo7_359YCbn3vR*8n}G4us_m z73^vpbe@_^gb*9vc*>LvLIi=~wjN17(Lq>V^*BB%GA0EyUFJQ2dib9H>tWOZ3u>Y^ zNJVf4TE74s)n8tt0F`C9K@nC}qcVjQ4G$`z_;X}v1!090p|*YGuLi<8r3e<({<#-e zqi#DS53?4$fNsCWpKOkKLFxtH901aX2}@FmVIvs+=V{9cV%P+RKMJ0gP7JqNKs(<~ zZ3ob)b)ai~Q3If_Dz-!r9ZBeRW!ECmv6z>T(3!EPj}IpVq&vP}^DqTOZK z7Quq9Ofdaur4ND{i{)-sj|NkZ(=kK=|D0?=x~z)%WkS}&ad z(D!BaKMsR-TR<<|y33JED0u_(MQ zG>%V5rxn41^4BIHEh%Xj4W#591oHwNJ?nNwL&^h9-vKBq%HePD%QBqgc;aI@w{*mC zEf{v5TaZi)ll@c?u;cmdGPZ+jMtZtdohh#oj$FfG{e7NY#;ZB|L!D#X9~`o`yq3*Y zZts1EJls}jb<&JX_#bw~$TAVqx$hF8c=sp~vf`;yQzcC;~nE60;}G<$xryW)%;%UB6dl!`HQfH?Lh zVi_C3-p#2;4u6!z61wQrtK&1(+C=sZI$4u3{TGzXBHuT}#>TzhQ@*sU&A#Wd(AFbc ztw~DOj|}k~H{c0-L867PA})s(FGC|7lCKS6d}X~4R!nBhkis%kzeK)RSk%; iPDVrS|GQT5s%=WX2VLr6PcHrcpAX(LTO?WY?*9iRGnDuM literal 0 HcmV?d00001 diff --git a/chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.meta b/chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt.meta new file mode 100644 index 0000000000000000000000000000000000000000..7b374b58eaa23d5cb523bb33c1670401a227a347 GIT binary patch literal 3186087 zcmeFa3zS{gRUoMMs#Gd%Nw#z)+m)MyI3geC#rBh`WGgD$u_aml5KDF>*;zg5!au>0t;kdS}FK4IMH>2f)aKqyM8xFVHvva#=_Ra@;;4LtV-@Opr0CU6j&8_JK zztL%32 zM9uWCt`~m=UkQ884EqxyNsyt2L_;E8!N11mhtkV#EnapL99}$B6NIC8z&=6#-s;A9 ztv?OsCe!h7{c1QD%wBlpEoYZ+fc?pz@c(BYc?-UOU;6%s$NfQY2==2$^se+*p4l94 zY;OPiFTzb{#vA>?N`EpH5DcUB>FM#+2`Wi@atz-e|KYd*1pj6%{wDr|_!|`3OvfbA zya2Oj@o`8*RL$8-Q-UqA&unNT)^s1%9@FZI~&k^!Tz>XdgwSGQep){U2Iz3%+V~cy((WJ&39}9$wj+ zMuW>wJQ<>DO*b|-MjKbJFYGz`C2%lWyAllsC@($KX#7eBZ`clY9X$$1&{*`ep9$fH z;iPwUv~i_BLTfXe7GxB<`3BgHuuP`mjkyM;dVFIl{6R)pS+U^~+<11pF_~P9Hb=vi z{-v!op)awKLnZ`A;NaEq#@1%9A6t+$l%jjm`2xY=I2SMiCMhU(pvPcaj-VfDX)@8?R>ON*qPv zNJ`9bMJVBQFnjJA@+URd8Ytm*eC)N&+G7jtTLze4oD;G>AK=*r`04)o;6ZVel$o>- zX7G5y!w3H_^QJ@+A<6jfiWBGUF#FID54QlBf{#85$g1dCPaFef+etq8OQ8IXm>!bZ zxE*#sG}_o0*AP`HHQW#L@uwdgK9_w#{Lb(J4#iFYoJt>XeuH)*swAG-D~Js5ZNdD* z{pr)u7|q8am=i*Q1le;r`9p9x7vrPzms5epV>&S_yDtmt65`jA!$1Gp(HkLuLQEYz z{fneQS?}riuMnO?a3FCtvcK9R9`esy0NGQX zS`;2<_5~E7efW0tO=aVEIQAx|{oy*A(A0bJgU}+(J+}B5o*{w>)0)AnYWd+CP}7!S z_AxxVGq&X#z8k%K^qQ+qe&UYpz2OkLr* zuL-9E59(ojSHSZV@P|qAPbwQ9Lyz>v2jFozy09?~@RipVAjIFUF95s}<{#VG*u>*_ zfM-BtPp<#6MMFkij?_$oS8LW1FP{%1Uf;fip-dlS4PEbI+#hG2bbt(TvDh|p+) ze;=SCqLHTbv*jg$1dzi{J1^I4ODo&k{<^V1o1bYT8usZ%?-SP zPl8+E@QSzq#6R{n$IT3S8>^q3=c3fIqT^?4R}aGemFp{`s5gq%uOjXDqBpev%4lQdnc{EIeojUk zo11t>1R>mfWova6?R4rcpiIZs8o&8cxJ8J2B+PVT-iv>HVT|zI$c~M7WP&Yzmz?pM zfVU1O0xqoK6KJC)Mry)lWurvJ&6-%luY>~^(W}~6dvbj!qBsJs*iRGZiQ?isJ3$9J zy9Dk!dJsR2f3aBn%`H-lY&i~02>$>B6IlQu^Weon-iO`zezeAT1;-occ$$O@#gC)m z+OYH^+Bm!%C&5J2sYFPUpV0(XBZqro&*E-;Eh^#zaKpu@Pe%DdbCmCmum_(oa=v~2 z6C`r98TF?cZ9*HhvayAcguRkLH}Bcxh*AIWb zguZTnoH;V{FgxvE6|_SmlpkRzCMAcadj3myl1JdiOL*xTL<0<^L~+y!ua#HOrx$<5 z=c3FbPsRg0m(dFsUU&TC1fyyA!PWFj3wz+-wcv(JPvhw@(j7?Xv!y3U>dGSq!aCsr zBHkqIR7Ja91q*n&Odi~e`r>rSB3-$lwLG}&FNK>C{OMrh9qV#bG&O;jggi<|0ve?& z(beJl(cO;<2Ndn$tUxOLkmKc1AFx69XlSo><0`^amk0*(!y>Unf zeel>LhrZ|xg!#uvF@|O#6AR3K^n-dKF${)vnS&!b)6P|D7UtTIqC?> zMm)B7D8;eS%}}gLwSfSIrIg&Dl25Ca_ZOBD7XSny)~|jUUg%J9rR!i3@~GSf@m9Wf z5{ak-9%orBN~A(W?)TEx6;dOrCb)bB6L%u~vYftn0lm%hXdT`WCqIfzK!8>n&+P$* zQ<8+B2raJmN0Ugnpa)@JdL5XKF%gR%L&1H(BoSto&SGS8IEpXD=nQO)row?r<4Swr zGZ-nt52dM`X>>lC_C;(6Wn{lNamGoGC%NU_)ZKalBD3v;k} zBCE_N%zW$-jtYiGOJ~PXIFyEOi)2wCTN(`U9um>jG-@l(Kuj`JKJ#ig^z_DLig1Y! zMlpQB{v-$;3tMco(49V9Mt1Oc9I2HoSC|L9k(wQe1}KBl&oRz%9=$|#JPz_p{hFN@!DN*ByH z`Vvfyh<{T&QAG8#{PN&}NYIZ23W+`5nakaOK>j@`QrDOQz7$;*w|QZVO1h8E{UrH^ z-f%D}C6Hd0NMibrlcW1DqgG}=$y7F9A?Vc7tsp&ELB?lcAJ%px$%w}{gAuSLQ zHXAdia{eGrU%y0|Z;1a%&NUy+j>sV4jBp4~L4dD(bn+;sw1@o>x_yIcX-bIEXgr@W zte1Ua0mN6&!S1(Us9O?NtVh6zyNF`$J^JeW5{?-lSXS5fKxgjwoyQmNJhph#bIBj> zICjU?{@sH&tgiN>W2b^of9$7P@Kx~gH1$52Mw{IN*m3tOBFvr#N%p-HWXbluXZFnR z4R&2@?dp69%=K1AXc6&R76b?JZnykE>0zCHSrDOpY;WggxFHJ(#lnl>O=vIgg7>#z zU+GIhnNQ*gi5CALm+}J-Acrz5`mF`;jzacGVqV!OUgj_r0_ej zs-^sFRQW+S^@- zg)-A)Zv%RKbqj8;aV$w1Cb>gW$$G~X|CEukAPwsDg`mCArqv$H#SWE0z~oN%dUm98 z#7=eY!WFo&^3Xuk6^8`2VI7152kJcfzgiF`IhxJkW;DXmiXNu%pE=gW>_65?CSJkO zS9HI8(KaQ)8Hr^^bwr-Su`6-NWJ7t8UR;cF?Z;-EbntCwb21!oS%&Jbi(JYQ&)fuXer zPup9SdJB}PbBbM{Zb_`a;=>kAc=)2145b>cvqWX zO;(X8(y+y!%??a8B0w6omsSdBWwZB`~tk`N!3Q+Bd?oZ{fz+X_(p$Nn(g0rE<_cPOs@F_btRZ)&3KuFO_e0Mj@EVlM^ zu=q9&FWm;O30@V4my|+Q@DskNP1s1eazifU2-S*IClx`K*m*RQg+Si{O;QLk8&VFV zWBsn6Bz(3_sAnY`U&+pg>f=TtWoUF`L)ynA=ouSwGd;gVllE`l*0LlLC7llJf%nY9 zok~fiEA!~0y-;mgW!8meSC;rl+06`fK456$pM_SwlbP2Zb>|JD!h7lBaS)eEkwffZ zz8;IMvDaYLD5Zen-W4!biW)?HxrO^!Gia+;N?nqvY>A% zt-%J|3|pEMqQit3gKoXPaJ8L!YqVa&wvrWNqn7Tjf~fI_)~YWm$wj2*;_ z)bu6r(?Vv!89Ag&N8sAgikL_N8pz0g9pu{;B#6R3~m1f=-?Vlu`Q zI}#}&&GMmfq^>+j0|y$x&**WMGRj8cggRKIrb#n}a&ciKqJM~z?1V~j$9mkiP6Tw4 zkHxqW*Pr4+i5QsSrw*75FVP>azG_w96{I4kzHt@GYlbO%w+PwFg3#*hq(q7_2Ze4e-u3SX0$ttkMxKPc541nrgt27u;C=n;9Zmu9Qr4 zR6NM$>$S3(QZdy9Gb#W6is*l|&-Tq?nr$~q z2G4en1?iaE@+8{OBK$*CU@1+|@^PzAFIs(i$?DT5tUi60)u$OUcm!^!%<09ceVr%s zfx%uQOE`x(nN^IVnQ#k0RxEs99?o2X<3ZOH&a8y};J?@l+i|6^sZ02j0)!27>SP%4 zX%slS&o5eje#!FlCoDgIm*wXfqIZAJ_PbyyIBv=I-tmq~c3V3uyDq}F*jV`Ngk#ZC z_!gIKpKm98i_3Pxx43L4e2dEr;Zw;@duL^*(@+5xU4(C`vGCb3d&yGxmMn#DY1tMY zJK!czE7SPI_>OW`|VY57iA3f~D=;hX*96222I z!gp6=;j=r2?y?lVyDWw8E=%FN%ToC6vJ}3%IKtQP$h%AF^X=YQ>2(plyBiDN-Bu3x z-Il_4x25piZ7F{fO88D%3g1af;X7$5d?zi1 z@1&*ho#Y6g>cX*SXJx19h2tc=e^y%!Tb}k49@kbms3L&&Upa37l|}onEZKkMg#B0U zvj586_Fp+^4oh?DqUDh$aY1zp{^u5aiKcuGr6U>Jq^lmxiVd|@xv`v?5y|T%i9$m$ zijiHN9LO79LIb8Apu-`^>hj(s#u#rVjQ0}~jn2u(l5$KB;9y7c%4iFRW{$g?!{=~Y zO|rh`&xt|BfpUCtc*EPJhhVWGF}GNhz2iQdY&5axZx1SCe%lN4om*5xh(${>{zHB8 zaQYAR(Z=E*UW#RvqE98AoQ&&D_Xc}XXgfEQrYDOzng1l+(wR>92yc3Od4O>?t(nyx zrhl=bV-0Xd?jE@Vg7JQOK7bZ+9914KQ3&>*f%z8B@tA%7y$f%G?`%O*f}Rz3CR2Di zFPAzMZVV2@Qpj&dAWP0cPo`S+`Y!lxq*C0UA^Jq(KZF_8vJzuWiM#dt$W{wINnkBo z?3K>wOF@;*O$P9c_=8~vpO#QC!L?dqkC-4zQkly#uLAe2rsI0V z!f@O*iPNBC83i1~IiH-nsAbvvGqno~2+~KjaD-#_sDp3=&d1Lmg-S@yKh=T>|G2bf zw&>tX!NM7^IU0m>q<{KQ{920{a39Y}Y5w~xhVznG~#)A`a`2hbt7 zarr<#OifIk)hqD-g01*fuyL&Ipjcf$>6T~X)8M;YVfA00APU4X> zK}yc{(fM~@+_h`h%&uL^NhyYx%L+}8xNXBvwBWSiVR}r~#Fp0-k-E2wC^SZZG=YE% z3C=bGeli{4ZH5LPe&vJrHl;re{)gM}SR?SOn?9eS{QHDY#9UiLJpJpyrM3u z$yyD9^GkX{OL>nG3yv!h{Tew6Wx7vnYNgn!; zzl_8!Cs8vK68^_F+}nhBN#Ln5{ln^^iI~X=HkV+aggD!MBqNO6dlCV==;xT!` z%qs)ayPYJYI#vd7Sh1Oz7mPSg!}W&xx&{JesnrxHKx%77h&yL-YpKec`?`Q)+)doxH{qKjcGSqSg!ER zRc_Yv_siTT6ithRn5zCt5}s^hIujuD`$}vlogh6xIMt%YlZ-&|a(4wVrMHrkakOv6 zM0M(q?>n;>{qbGPGw71Q*S71@mBJFFE9gT?{OXb=VG$aoAT;xZ^r(cWC2o}-?L3|` zA8)r8(t9{vNWX8P*jORF?pIf(OvM!2ETfOQ!~c#OZq%3ua|r|M0-VZe$m4hhV@sfo`&QN` zn;WRk`1A+&xv`bvR?3Gxbvt=?;O11L?U<+)+8?fZi1Rt%nMG) z=h3TbObFoM@(o4oL)^Ro2T49sg?zg4Otjvc;=WCvQI2*-@;Br`+&kg+pc4miSs-wk z)rOyM!NaUa8g7#2h=&(s$TJiEV+$VjfCwx4vCXLbcpENyK*bmd@g~k?aciMZT6OOZ zz`l4HLpw2Pr5Hx0k~27vE=yQA$a)@6f&bKohdj^@_t++pk!@f^=ub{F@`-tVuH-YZ z3;un(9@w=vd`Y~-6K_IaHfu)+_-6{|P)R^Jic$?x2p>E-za?!doy(+k?A$E}RdB=` zOG(_tN>oB|ne7r)VZu+im4Ga{okamU!oq`~9&T20UVFN_b3ro9BA|4NhcqYh!o%Ma z?8ZX{-ly?<7fHX@6hA2w&^=58KcMk?7YncVCPhkj`EJ_jP4SAIUMir%@`z_Tl1D6p zp+~EdlHPjo5}^sh=&mNX{G(doQisBxJ}#$l*|N;_5>xq{hdgoA!KLE7qz z8*=%-BXOBr?6m=!CGw!!u>j4FYmTa=;;14yQ@13Ukx7y3u)-}`<;aOyIkqtNb+{@q zc?DKYQZl%3Q+4%A**`7$1XiRHY8J9Tnc%AC-Uv%sdsxHOA9b%O;`iF)>2P+e!kPIv zG~a$+T3$>>{Fwk{+qTQ`Ld830WnYQ#ytRCVpH{@7>l6}|wC4n^g1z}4>t|2>d)GK0 zx=s~%%7Pp#Ir{@HI1A|8R`08(l@Q+Hi0!u2I7a@iS4h*~99*i3o(lx8Poks_NF? zemYh$Y3q7fl=-yW6YM%8#liPbIKh8z!7)3?MD=`3r!s6$XRPdsX- ztk=8s+O7BAckA_A2gAXw>l@QspWYY^ZjDCKTC_gB;~m3Ew0!N>+sq(bTR!%>TkpHI zth*#uTCszlvqyPRfpSR$+yLp^ml5s6nZ=lBw_oQ_;(9^}z|~)Oo9UJcRr*`J1M!#w z8%Y=W@}sBpU%Eo55;3j#zqsTNKJJS6l7cl_#GA#`^(S=_y6&&-~g4xSm4t+loOc=#0&7EKzmlmCYz zXo=a{*x>ri@a+;QK9N+tI}h964!DFznc-6aP!=#BOzz@O2-WOUPZtvE`D~z6+YUACmQp@nANLA-R&tQ3fIaa*3#7}%xaQ) zj+7GbE>e2`Ya1SHOtw6s=bZ!fT$DY*MIm(auOXtA;CDD_TD8i z#(W~ee^JOO4Qv*^=P-#1-q%JJ(5}sP*H!*x3tn0!euDg!=;|=*jtlKM^z!U>FEs{= ziaGILnMYd3Sj%bA0a&YvPoSTOdqf(j@df4V)d4k1hNLLK|+ z9Rme*io+OBwMW+nQ4+k<_$FyUM`(79lqTo8aAE;XPqvw}g(nbU(osLHy~S(OvTTbk#B;D;1( z$H)5Dqwxi7E5-oJv@adgUlENrCaDHv8+KZeaOlj9IEp52L7D5!3Z{3C=U4OB0{U9Y ztGWE1vi>H`(nL|l9Q@yk{&V?ZvLo4*eZO0ST9dc@Z5QZk^=1b%XEklknUfKu{4?y} zxxZ?EWa!)$vYSw*;K8V2E%{M_hjbMU?+j}}{>p>P4ezn!$t1LwG58NHSTsb*OQkAEYNi^)uc?~V9s

wLX;lm2hQwEJD6I_tv39>S|KU}8_b@^E|yU3=Y-A}Hy^DMPi!HQ;H})3f*(|5 z!wc!kFugS0*nD{09}JO6!T$13p+?!VZ;MPo{>J8*!LMsIB75s4EcQT2dT_Y98jYj% zl_)q=`zwlXwE_V>NwoLE0&|w%Dw?lrZT9uoKPMtnh!jg%2r(}oW9~`>|FT^VtNnAXwY@n`uwC3cl#oh72F{Itn*R}gCc)t+lS2YzPrq$^<~yEFmI&ECrctT?u9`ZCwc+#DI5V3r$|s8ifp@7Ee`$ z7s9e3rARUB^Rdn0=^>KMD@s_SUmGhk7;kLiWw;v`A8(E~u4DUGf%Rmwfvm4;xo8Z6!oL^oj_}JZI8nV*5H!0Olkmxeg24gl$NKoB>>;sK$kzy- zzo4AUfON=9vGgRz0kB+Q$c-KM3v2yrA^Wq#Qiy^w-8vP4-{NC}rYi9+oOsH{Ntn5O z78Q@1S`7bBMN}-4A6apU>!bR^T_gt^C$tD&al+2ETFyEH6HlN75wmfxhrzOk$b{jY zky@3>(HFsY(#k(H+SnKeyW`*LS@{e&#Lz25KdqtWVECL!R`*8_4xi)s9s!2{ZeGEN|muZdcf5|iBzNzk6{ zPjIu4_+!h{R?>kY0b% zn?~ysq3N|Yu#zw0&PSYum}W?nGv0*Iz&rD4p=AJLc=TYod1;i3ge4s^590*z>_bCr zX-N%u8NRrpwjf|Ou&?})7zHhp;sAtc$bBZZt9uxj-UWRlEP6l_C%8*hLGN9oe1Z`2{kWM_~KV6S2qj$(Rili}u|EC3) zJ+o5UmQXU(BoGRt7k3r;paT4Q8{Xy@!R*8vh;lk?xv%Y0*bdz&3#_)9d`{DRD_$3awA7LhOUH-Ii3Z}8jrD1NxE_t;SsGYo;~+3DT~hWK z6VWdCx9zNph)+Ix@}V_cpkyjS)NM2wRxQ~c-cY&r}VDcO46FI%8vL0U@2eOku@ zwpHtc_s+C|)UGBU2T(rFKRFWyyO=Cn5>W>0U$o&aOL)mjU`v>?&QC6zrRz^pzU^?bBBneS?6 zqdFP_m=#Z0pFLDLMe4=t>Kycu6-8sJTge4{wUtwvh($gul88Np_w!ASQqYR9ptP-H zyHTIjS{ zRb@mRcBYlu4DExjZ^LoE5T^_b_N0HvdSN*9O%%Z9kW`BchQap5W79-+NPim41+y0( zdCOUyvuYE=h2Ljh)1Ha^$825C{{oh>o2yItv_&=Y6CBPL&z2mPT5OsTul0)_87LL= z>RE9VE^7WJKsBcTe6kI1^~pD5oR$@|V3X2nM3gQ{+nkhgjP@_uyqgXe(>y}Eo!sKZ z3%MXCa$Yzk9!P9-ZQ*y@@FtJ)oJ3|>cUCT0ME=mbsoY^IbiOE-9#IQ?sAI`uvxc>hY+GQX1V2GrcrP=sQ9O!{!&*v0=uxq;;k}abeB26 zzK?Xbl9`j@(GC{sJ~=ddXJGZhPGe%!tWzy43m3{sIV0JowBZoq_TbTG>*zX0@XsY-BDC-?dx;>(N^@LseLU&Q+8bSkyLK&Yx(Jo zW2vq+N@XC*DALhXZtZLN$$Iqs#Z!GyfnEE`N6;8kW!1j&-xoJB2yW&jr!s5JPLQ&N z(+hREwXb|^^T;Zz_LcwK3fFkGul&bXqpLostIK*N!E<))D;p7u+E?~7Xuoy!HKht%R3M6{J#`&xduTVR7#yHb7)IYp9N z`&xb&9pJEri&y(9y}Ud$wwj;y!*40>!L5CjVC2L;tZC&RIkVwl*S;p`UvZAWx`v92 z_sY(I?g9wh+SlUYvh{LH0ToYkP7jnGQ8P-lAgfFe;=agE5f!g?Gg+*wU-1&sU~y|- zi$~Gs$`U4CSx@Q3WmDCx_O*DZ#8u$dz82q)E8R@xZkB}cA zuL;EJpjj45>W`SznQ;Q>2%q#Wbcn5=qMj@r{ix?#T9&Zyj2M=j9@Q-5bHatTbQXB`2VdU9 z6?eDAhEx1-9e*g+a&&I1oe(#}k>wyhA#yt;jFDanid6&#D&W+4)hdGefLavSH6_9wthMPgYvZT8*FsB{I?z?iGr)b1ua_i)sDVL&!6SfaZytEr z&zFVX(ca+vWELS~&=QNovQH>KJf}3p**nc~2nP$$e3mWi)8|p8`SkK`ML)t6q)QvC zQylUf>`DKSO@G%bix_&#G%2aWZ$4xQ^8iT~AZtxi|a_NNQ4iDH5eS zgY;MTj@lg-mlzQvcs~6FI7pUN?v?>9&s*N<3Gz*$RZf+!3xt>n@4+g5O&cEJ7;J*J z7j={7S<#a0!HVhg`m?^Qn9A}cl+mIdN_RsPH-#nwmQST`@(?kpzR~w_DmszL)E?Qn z9fa>_Mg=_Hl2g;0JT;<+n0(I{Af0#=6#Ca~Saw&Iq;RSckMT@y^nH{O-!#GSE9>@~ z>3Qyju;w&n-^vHQhYc8zGtv_g;3t{!@(c@UBEK!wyhH!TmYkEO_p`kG%qMeWwReR7 z&=TR?qFJ#@%>TDK`6@dB{3xh(sCkhLHE*}CC;YrDCnSQFtk_!?cw8((%`Boz={37x zb4sB?EZySC+SAw}*t|#uo3&?}+Vw_GIOcThgUyT1!RAHVU~@tu;VoMxZ1}dzMkX;$bx_iwk+s zzqeEtU{*6*#(Z1(O(?`;Mw*^XlE-X`hwrz{OfHefg?b^xu30oKm?kb?S#^;E!$OEZ zw>&16ieti-%XWuEN*ArH6^xWm5KS#yGQtc#+kz(z7m#EyW_N%)Vf;7Noy(^|+ykeA zlbmS~3V7Hu0KaYCyP$Z!zWaCPzk(*agod(nExEWV0W}{-MXXOz> zCjwL6Tk+bCC$iK!U5NHOtYC5y9+!b3(zi>?quR>O+}&3uV>8AXN(zy!Z(5}ZT?_m(_|*6rGxWj{j1YLbui;%5Fy zuotpT4Pr>Mso1>sc-gX@mW`HqDNgNqE-jYzhtv#~V{0CxpSbpNx8^)+!czwae3N;X zP90{_lv%?Sde99&LQqy~Q44OGcT}f}k0#>UQ~e_s2LY1zD75->EjVp+wj{x$N=Hbo zCF$YDkF573l^JHX!`je?IMRKAm~|#OoKRGEQtG$*oRCJ@1H(8f_%=(2HW81!ZOx;eZ%TcG%NE%-7!sjbjgTcT*{Hmk5sQKV>0+K4^| z5v2{<3wb_uMhe~P%|4D~X(nKKd1AB2hyuliPr&jstz73}LrQpY1ZhJMvVgWnnQBNC zQK*P)?c=A}6T0Uz+NIDmtDQD7qe&VIvlJ!~Dn0 z8T|_8QhB01)x1IeXVyHEN)x#@J8AZbgI=jDl(-xCFjTpFbwR;5I- znR3F%%$-=NWahePiJ~ZxB9=DhrlBF{02?ToG55#%*Q4(fbBRF`(qx!78n4mZ|&!}Y;%B{zNH-nT!Bg}Kw=^g5o+X;|^VnSE(kv9;Xt zHQfVGcx$0AqY%uVM?{PR017uf^h~kND{8Lk0ob>^R~#1eI0HvN-6WnJG92J|9h@D4 z_ACx$<1CShG-3yE5(hK+T$qN8N|&Y^n-7osgCX)W*kAsMd;hOLQwUE(#Ir~ee2arN z=eRw)B^=}jhpVg67`NF(!J$jlzp{}HFGTisar;^Gdg$7DpHb$y{&?6&g1iV+{NwCc zXkE}UG*UIUHv1Vl>QZsh)o^AY)hVlw8Dn()vQg8GN+q&Rrir#Sp|~OjhOLWxJ;o`R z!)!kBQ(yOb`A{SL7A@9*`+QAn()8e3MS5ug3Y!wG(CS78Uu;8d0LsP5v*T&7JN_+f zMp&b$X&IZ71K!z&$82?}5J(JwnJR^>HC6ZYZ7X{CXdAWx(5(Ey%_;pNs2;-DT>~-B zBg;)yD)cJCxpFiO-TrkWV}aDc{@t1Li~(6C|F{`DyrGUuA{sDNpuU)(-spPK#Ks~F z-l&01zxE-%{L~lM-lv2^cx4nlzF8~9w7wA)23CG@-MH$Kz_NFtjqrb@ZGX|#1DP3C zr|>3aS6@UkgmCtDMrVe*cMTX(JZ-{ z(-VHb1#eb5aKrT}4q=Z5L#&_NhCXc^rNZAjL&8$FBHBKE_|+CX;S*0re>Qt177IPH z*AIS;iWHE~m6E&F)~~L-__SUm3BU5edp+}xM(&esc&kso8T+yQYl9HdyaBo_vF66? z1$9|obIF3gp!$Cug$B)Z;&E*!w?g$L%Ms-2s?+9F)mi0u(dh#GZX4d@QJ#~-NZtME zNEYN|=-pK22aXTJp*j0fD&sM)o-s|*M9z|yGfq@}LWj9UkbI>NQdF~`Sd_mzZp{q# zFGHNN7XMqIw*->VxmAEzQk|kLyqC&$)9{jcP)nFXpZe*8$7NxJzoa^fXu@#vo|ZJS zJ;j=vR^jDb;2O>=$~C6RET~?8(wkyY<;FOx^tdI7}0EMqnTxY1P7?g|c*#%P7I7V-}wStG3zD&KeJO?D%a_iMjz-Or5GYTNifn`MyS=4n()>vQ#w(8iM0$X0Q z{8_HJD<4-6%a;8K^NmHD2ozO`KETTYvX*WL6^czw^oya*7wh`iNL8BSn+7V05F=L( ziz^#Owu^b9eJoDer4yTY2;_ooqdk0@TA!hSSKXOhOKq^vd-d}Snxi!%;vc7*TfK4C zf*$M*_N<{z;TBbEtwvf`9m%!o8{07LQ-gn4#i^Qm)F#(63grA~jCM#Zu!Ej7@JSXPx_D2t53!1GUhF0__Fw?}e zQiG~vqF6qp-gThB@!WVdn8E+?aZ9P<4chlnbCDI$ilQ~u;ycw9*UPRU9MHxlB?6a3 zVAX?cD0GNTY>JN~n_dvAV-T24^<6$TJlRk^&o!{A9;3sjI{#0!w#*_ewDvLw-An>eny!wPrX@vw=X;Cw8)7*7J$YR2ICO(M=(wB$it@Rpi^;X& z6Vyl##Z#qtTrHu>$QAoSi45e3($mM37fEY$Soz`YN9bd1V#>TVLdKl1PZ=)@pf&-# z5T|d{3E-yG&C}8Fn^XgaH5GHK{G=Z-t8j%V(zFIM&H8Hmt(5S2oVN+F0s9}$d7($* zXQ+E1tcifU3e~e^>>njf(r!X4-FcJn1h5kFHZC>g@b@V>j@=2^&uMntRO8pR1Z^s9 zQc#Y#=D-J9MDL@LYc6qbK2<0faT1o=cTvGe=4{z;gnqP3w#rSU7=EusG&QOSzzKD; zY0*qlB?&Z|NBEf*yuqxBA`L1zN75(}>6OBrPDoUr@hO`~g9$_EEWr(gPGYaPL8*6z*0JT-3MS&dz>hDIl+uc! z9O{8p5^l<@RV|jU&Z(InWC-FL?VR?i4Bto<+H(QRLilVOPlfL(Gx2hO)iF*GO2!D? z7#w)V)5B5J6DBCSHqGw(J~L(T?l#9st0cSBDzAu$q3hZ^P7FdbQ(aC=@p4&n@|Ui} z2{2yp=Qmh~q&l-uQvr2RjfAuMNd_Iz9=tj{0XI!I1S*q%rrn~XU(tJG5PV8snx>A$ zD96{CU5b=ZMcRsC01}PioLOoyz^lhTF$KG~ip}8tX)qVeUU=j!XLZ-EWTBtd24*It z1!Pn|6H&&z=6Y10J(5`qC19pX&z2X%l;#h%;BKw#q(D|?2je7C#pr9vxu(Qjr{3z! z;4W~5QaO=%d!T&&e!w-_>ItzGX{^vvJBOJsL77%qj8K;RLWWY71~!YLBNxHRJN;a~ zEG8{GeXsF+c-ncEN6r^WnYJ?_%fVwRkk8Wr${=d1g^G zNvOv;q!Loh|8XB+d|w_*T!Q04*BDEb-zgvAf~5&)^?)XPz6Bi}pX3M2(Z_~B`iel_ zBS#=_4=fOZ4;d;4Wc=Nl(JbI0qmh404Lny19T}0dfQ6XMr+#6*c)y`^lpS(09YA}* zuW*SIp|l_bU!AZ`B@4*A=q-GlZCcdBPP4u%-O%7S<{+K`Ty1G#TH<%|i2={Ii3+xw z2*pp9jZ2yTlvZ@^l~o_JE<9V7VA7NlW&;M_Yp8Z91UF@fRm z3u6Nf*Tkg6SszY^{m~F}cU>_eyVI0`Wf>EN|7sUQEv>9|z&5n-icRdwB!DXp{U4>; zJZq?Cqbo(}Sbf%{BHFXPgL>aZt$EY}p3dzw`J+T%ruNZwnlzkf$QsZIKWE6+70Q}1 zusm|NGJuz<_Suu-vAcPUpxGSCzhUcK_V(HgRn@RvBd4{gPCfCwUH5*1nx;ky96j_^ zp@`axTvIKDA86|XeRUdUz^BT{G_@q#!$){0efEIt^)HaZV%T3jJ+sE_VqzR#k#}}X zE^dicw+ZWkWQ6l*aR}YW!n>*)Z84w{S_2@BzjYy4VM3Z&PjaiThF=&-F^#%qYE~hr z^=EEdruRFkEi+aqC*Wl{UmGWlO8hoW*ZON;6cHV3q0B>yZpuLH^R?nhp%lgtyjJ-| zql+~bh=11bV)Pwci0du(_*qBf(lN9St{$=cPS0B{hb%ZvcVlbbFQPx1L9^ClvR5Xm9yYY!CLY1v@$n#NX#m>_fdpBS+tOIt-b)B z^VSv(oyTlMto3(eYiO_cJ`>A-(1J?}2mg4tKgFS0VkAI!5{*{7&u&HI>$rJ+ydI6D z`tDluTQZFs^p+rZcDI*n;kHP9aj1A9wUN#6nYMDynLGWHaDi&PvHLT- z$W!G|-X@Qf8VYxa+TYxUW45+L&YRVeQp!oJ%;uica()fdg|@(Td=L(PP(2;qz+IPI znmvNlxrZakR9z=}n8~b~HC*9&TUKeUrWZa+M~rBqRb6=W1mdLjt!d>SIo*XnIX~eS z)rG+X`d73KjAXu_zuiFdUWsk`ODvAeQVqV+JY2S3EN6~2@LKaq?l5HCBWh;0);>wi zy4F0^t-Q#R;8ZaM;IbptRZ#~oX|60`3RR`hx);~WjhLgIH4l}ozSKE$^_ z>L=E9M>=11g~9rWN#(tUZhYT(G3Al;tOJ~}U=v~i@?)jcR+DIdk!pfR=M%NfsdI9F0&%I4R* zSCIX z#@QV#UMGl_$6;j(8Xc>0beP)lif?5XVOb3JLxr{q4AKWqW zfZFlpN}(ih`#9%J=@Ko_{%0$8^NVkHYu^?1z%M=k@$}@gpbT_mM(@#t!%Od&nkdaoGMfR zX-6Zg`XJl4SvZz{g08CJhJ1b$UiFTITt97bNUP&rwSX%t9W7)?{XB!d=uT+8BoK?5 zl&sN#mg$O>1*9}(J3_67WCh_DP>Y&$^eFkiYgIu0R^=@JxfQPQ zDrfnRH?P`AOPcepb}#D7t(;{eVo^EEekQ+i*5Z*~Drd{@Glz76TRB^PtQ_X3MJd$r zO_piHSlm9e4Pz0U!>QK+vvRf!oEHPauAD8u*)8tDN?4a)Lmms<%GvV6==g^E-44U)87Z!beHy*p;&qjGWM{HLd(3$53@{a4#J36CX{oxvOE+79VP6_*#u&BOsy8kWfiCDo@me= z_fRgka@Gu7rZ<*$k1ba4Tme8WjOu3qz7>8ms$6dK;-Tu5nv!SQvgt zJ^i;9;Jz5e0jX~Mf5lL^S7$}CFFO;aMY6L;56)w8==1nX>@|J>f0fTYU8U=+9%vld zp&G430WEgj$8d-CWz)Q`NYp9J+k zGn~-IqJ9bPvZC2>Jai(F)(pDoSk7)yjiwq-RTHMwyhK61JV0|oC@?^>x-0Vg(ISDXy>>>VBi zZ!Dh#WT%4`OZcRE4~Yt}VCFnla9c1_xr#*fSofMj@JlDGn}jQ-L7@YDzu1AR+D20o&wQ?& zYOjv!)7@*KB}?@Ta3K1(GM#Z07(6AJ_U3_?{d`%j-M9h2%rt~CXoOP~bJCtuR>(c`$=q1&{bv4$mI&t-&5BiG6DljU zAW4P(j;)Bty--AaU6jad!#DS_xL{EsFpz7NKSq(WUg7U9dT&5R+3o7HnSp zA`Lb#ItQBwtOAk&ZkZ#H;45H&zg2&&Hdb%NJ4H1E+$MoJAB0c(|@Y8-Clo)Isrlsra;7 zColEhslqB)=HMSqGL{6t(}JhGvosz9m4@vUX$bh058k_@NPemfJ>N)Hy1D8-c-x~X zW#Mh<)Lgf}!k>WUZF}&p8RIS+dtJ$5I@`^?aGP|1!JzzZ<%cQoxi&oRS=N)tk*TND zg}cTcNc)Fcjz;rD_4lb)HXBYxxwoa{xsdi6JCUT0yAy#a@2#MqFWXydgu(MHDKXr? z3?dbt?f}3FCMV%>*=*rL4k{x3J#9GUph{%Zivw(-Z(a7q`?9bCW^qil=n64pjWKed z*><34Hsm6~ZyKd95n(4f_8dCAYcPsJbyKlLl0;dzN>{hpHDcz~@?U+U9UeB8QcZkXlR9!;K$l=t*kW&xc34FA%fN zB!?4fC%cmzhfU#FHY>!^D(0-K&nk(s6gHQ{B4o8AW7&wM^rS&!cJ!HQ7bIJxex^Zx zw1!0&@rDnXOLf@bWx8W+B`M9(LGEwsd?sW^Ixm64oBeTrG>S&O;WQfe#j@T>&|Uy| zPg}Xe*(4f8XP<-K*`2Ox!O;iA^Du>L>z^o0H*E^cy?;!1A-) zj1e&(Eg1nPlr{t*3ut?ksfJV$g^F5j&q-#K8xbkE*G{5kWk~|fjY-^)B4`o4Q3UUo zrn;btdCm(IT@TB!5evLw{$u8heg$)>Jkg$Ng>D-KRf{D2nKjR((nPLJk2B<{D;Byk z+Z)n2XWEkFd3mDc_e6mrmqzK48cT*ZR59gk>MTknGuK5+6h(=Ys_4|b5mGCe78DTu zmCTU*W1@!R0`^Z}Wy-WK6)&)td_0^NX50t&1g9(tGvv3I{oYngi-Z&PAQ-&AZETBA zxhTIvj!d7EBhz;Wrap4zGdte!_Z6Z3adxGSOdYok{iS~!o9(=lR=E(6bnRDM~NbwRPZ_erfldu8J5mNvcWtVxB=&pK-dMttlp^* z$j#O{;!7DjOh_>O^(yqFwR3~e-rTA6$RB4V!wFd7{;2J#kU+>Wcy2V?CS3K7&ri4D zf-ND|HrAr`=_D>b?c%uZtrZck?hV%m!zgc^T&eq`n$@@u6pm6`uGjYNyiy*tfh_Bx>hz98TiIPB*uD<7jg<6pIXly}_Q9 z(bfdB2BJ6k@^FBYc5ucBTIjgV5ND2HBH^}{7}UYcO&6viqtd16#^%H0{$Pmw4EC3Q z;)Vz6&lFojI?eFSVoC5V4%(dK_Ux7zIsf2rbu}7C>nl-k=u-8sY-Ga=kqe1v7=8HJ z7F={y@8|mCVIK+dB2e*T`G^Gx9_+t*tBG=MrjFLt3WU#GVj(}u-NdTk z3%#L^D<&E+RiM6@px)?u@nWF{_5|C9`0`U3Z^Fj@8Lf6Fqij~=$VL^|yfnh-hTG-Un7)WUF zl-M#iu`SX}sfLHl@3-L1N(XMZj)94_XfVW3Mc02DH%`{$?_eTfDc@1_^>M%2f+u|9 zSzev0DjD;R9@z>AzeYt0NU}-E-CAclYi%lf-w+8znRhgDpKQZhee%uNuI*nNWJERk z8n;OVc+9IuP?I!~v!qh;mq^)lm^%o`SNagRS~^$Rdc8pJ z7$l){%Zgf3oeC?$d#P+U4KJAowS+13sh>W0Toy+7ORBq=CJZO5)p^~s*ms1%;rKVjtQ|3t%qv1#mrXq#mW@m>`_OPS zB9gW2r&ReJsuO(t=Owh&)FCYt$J{{x%&JG%deiGTxm@IRn@W?Y^iq#^BQTJUw5G2Y zdFhpn^=W^&9*u)`u&YyTnqMf(J1HKW`{c@@I|D1E;KMCAZ)XQw#)nwJM4n>EF?EGA z!7+NRw-r3ZaAxYI<#pc4j@-twA{|+up<2@@fH()16+s&7o-Elq3jm0%I`*c(me)&v zmMiYc#~H%1MqgT+3ldd{KETTYvX*WL6^h-C6qjTFJl6HGk*f60HwZDtfl4C8$d$w5 zEOY3jUy#@7Uco}9U3y9SG__qr0k67oxt7{spZDtL88k;{1@x$5`=A%lGOe+)g*&M; z)XiD1OQO{jdr-#HE*5od45D#&l#UmrMWN`Lx70TNS*aG1LQ5PQ;urL?@_uSRnt~RW zuslS$LrjP&S$n9AUdP)+2dVE$`f`BFhSXQJSufmZoK$e)Gm+8yk)Q z@7$Be<&Hx)xPy+%DWE9dHI|+}rgW;T(P8C> zw;!R8wTU70)(9DM!aikOo!)Iqt~{|@mT&D`@P&AWAp9oPfMHF=+$ulmN6ac*A&NAu z!K`L|8&O^pJxS?{{SW87&?E6P)IAW^L_l7J>d`ayj}qNLc@BM>sfm@4w{fW{hri1a zN%B5%N)AnT0`|+Aok_LOdNqri3eyK#MDL@LYc6qbKElM^UJS}o`z|UN$($`4j?j;G zHRE4dxZw9%L{p=R0Gv=an->0%(D{X_EdERj-e6Wmkp`8VBlakfQ@NAZg*$@RGo3u~ zKKTgLgz4tsPh0RB9ul%OIVl^8+XCiOIt{R@8HsBCZB9m&~@z{C+4J?sa|-h4oR&!>!j9iuntLeW}&76>Y^G6XZ4fnI-r%7vVGDk zI4+`Y2vjEjOuI!%KXI0H3{uXJX084!S|I@^J*xZdRE#RpRty7>Xbk7fQj0+$`;Muq zIV~}OjOu5YzmVgcF;k^yO9~t{#M#rFW<3Qz*n+#Y?x=dS@D7h|c7<|h2R7rDZS?@z ziZsfTXS>=l&3p;c@>%xNH7-bf40EQ#^3Z2dbmSs9d8ePtm&G7vr|&f$zx)L^Z1VtR zW4Um`+h^Av-dok%0kp!MgbBKq{cn<0TD0R^I_AhIu1C7T>c-ncEN6 zr^R?;C8U`DqaR?*zTeQ3JuwM=Uml-a zg5yEg7@w4(gU`31qvMnOaXI?vDH4d`Vrhry?STbEP4WMbp>jaR->n(V1|8xC34cq? zLsttO8IiPrg_z8zeqp_MznNf4zA4~z0PO|8!X-|G(t;2TWlTaPcE}SFDk7C1H{@&z zJk`Tav%V`2rUlE?-k6(~6M#E1Elf)c9X#JAD%ffw6n|beE)oiaR&?%_EeGajc(yEl zUsL?dok(-=y@qO+8be(&9acyaq!js5t>;;Vl(aB5&~QzRTH0mBYAUT*_^);`)Y8h@ z7HmTc85>=?u&Lm}k5X-(HB_^ClcIF2K5H^HH|0USCAHO+wHfBn-ki0<9yqfvZ-s5O z;9b-fNG;&$+)k4}O7vxFA6=(O!- zF9z)bE|FEY3G0Dmg!5=|Q0^r<-qpXcdp$RWMjNB0giu0jzDVP5T?pn2cyj5lhI$xD zF^#%qYE~hr^=EEdruRFkEi+aqC*Wl{UmGWlO8jm^2A8aT&?l2cM8{ev^N^yOG7#&2 zt$3o<(?Vm6QGC(pVvPmjpEbM~eFqofTEMTR1ANvIxpWMzgR4jE#?$jwi-ZL+fb7QB zykA6rG=pZu^Rt&4#8h6M*GIVcna+te7>z546W5Yy(P$<-R3GW*2#?9Ln3|k)ol{NFTrafNyTYF}8?G3cr5b#tdAMx7 zJOEp>U2gq}IIB^rN;(!;B1X-+*3iPOyvPzR(qQ4TBh^(=2QO)^EMW>&rO>(;*UODO zNRDKku(bIM--5>K$+1nAgu%lzoBeze?h2MHuc87h3+??Abvl+{1g~LI)faE`z6T>W_|Whu{XCkS>}%V+^Pqdbg@9I(om@~DW?1$HddolgR>@0UnT)BSe`Ix-tNY?=jg}Ab?T`A2b={HFLwAyEBYE(FHh2Oq z(;G{>NNJ7MHtcNs@z=NExHSWdMoepsvpGZ6;)3J)_YL~fes2}G^!BI0Trhj#k++=H znNdblNv?^aikEfme6Rp08*as3+}+RhM_W;FrWaIbz3`Gv6vv+jYy+??omLu zjd73SC1l)#*zu)dAiW=j<;E5;DR+G7UR;iQOvhX6E4USP&|4S&+QSKRYnX+-^g@Rrc6)mpN)>6ZvS_La{a_kP-q#C?WzJfGL#L6b}L9|J-f=9WAf_`b)9K~|qoY|5E!g(UB z$gr>$u_AKpj}<}uA{0kel2SuO_a|bj8ZW0LKQH*I+KI$}oFRN)MU@6C1aW2_4DfbqEi5i ztXC)k2~tPJaD&4aEYtayE{EzGy0E%g~e+LmSFvSI(B7F8S)!E0h#! zEqfxQh}K)_Fpj*-xuB#BDZ6sE{3ShltZ=JqrB4}*G7rh925>89%TLxLSPA0QL$0jK zS^oRts`Bhin(@MRK{_vCS8nAjAKN_e%Bq~@Kexg)Uga$R@%F)2A1VOc%2_rd7L~K? zXYwm&EgmVY;_yLBM$Cp=Ia@|U4s)>Vy<0JB`K5l0L@g#l1Y^0CvtjHC(*PS?T5FI7ro}@Wb!2aI!0BB^WuO5o=ocM~ z`V}uB4Hma@ws;h6t}J2VRsED+T+;*}j4x}ESZ2ejoGk&xu}zkQ@r{<&H~_bDHo=(} zyk%9+Y93e*N0r%RwzJwu1#l~8jR0mrDeJR`3QTQB2-mF|mt8qqCQ>k$2Aa0Ba<=@! z$~>6tHv1y0oJ}BB$JAf z%2|mupfUt06(mrrdq>&ifcb6n;Vm;yBUqw z+{gIoSOO!v0ok6PE_cYYO-pd&R%B|+LZ)Pl34I-+2SSC)v6CD=> z17_x*7o2!N$noWvu7Y>ETpWBuoR_o_exwB_IK`v693k6Y3~GiG+E~;t!Ch7~JC27=Bq~V04a$bK zO{(EkHDO9l=V0F%F_$Sls#(bAgbQuyEKEj!@a4U0ai3WF!*%?jRZ!%%+97Z=99a(H zLm)Rl!Uk!D8s4kkwNj@7PHeocI2m;I3_E$!CYSwZV7?XX>dc~je`EPzr?YaORPP~C z0T#@h#|mx>W-3>asGf`?KK;-+=No)C^GrT>ZL_~VcyPEj=?;dfGg(V{Vsv$x^g5ra zC9=LOm%acqfa1^LLCGSD0(;h3TRcG^t(9ES+i_V?Mx3g7!GfM!t>jP^TEP~4oF~@3 zrVxCUY~!{9DY!xOFKO|jHapeCGoLG`+N-1bboW|l$x?&xm&mC_DQ6FSE7KWAfx%N! zWp5N@f8b?5?}6;U%rt~CXo9vbZKLCimUm8J?S4pE-5?#nj;Rx?Clz12aLU-jg5@rWL{ley&t$Y z{0zuTrE!Mk8-58u-7({IUlefy5hHj${RQ90?32iVuI#u}?+YzqK>8?pRaICjgt#%Z zJG(dZYt)mjNvW;S_9D748*5Ll4@EC`a6tNz7cI${7E4;WpEJ`edH1KDB@$Ehu0qKY ziM|z9BTfcug$VgU-Q*!+zi6ZHH{LV)9nGkK=Rk65ss_6|$x|bG@W;gRuVsB{!24ohc<>~B)pJc|% zGu)+prnXe`4*efna!%TF$_lw>KA9VVsBXlc(DjIvxqLG*X)ALDTSC8REi<91Jn+|=0z&l zthMyKR89HXt>wMo8%s;_m}zg(IoP~t8*ENURH)cC1e?FM4Xk#jB)*I@^QCYzi*%X2 zK(k~Ua9$LQR0f=PIZ#C>SPlZbL<#VA8XmMca#9k<#Im?h2=G!_fUTHsE58Yac$`YJ zl5u`maGjZ%{C>;K$AsPkr-GB5F`){0xTi`Re%rj%LGgTj z?-pasnSV6N5EJ}P3!d`M(&=CjJ(q1EDVb3c3`M*Ue&vJrcKlERKGlYvZzL<-T=gEj z?a`F7@U|G$t(e6UZFbnOoFBC_p2#Xd$y@l~T{FgAHuk!b#dNltd*L?eKw*ddNV1vX z@VU0JtH72IN#w}XQ)>IvxF^rFJ;eT@Qdi!NQT=`DmCc5eQSNOyE01D75tvfi3gY-u z?OKluW*d?M!Trl1vfLJ{F@#UDQ>-c)`Ta!_rA&8Zd`|%G-cLsDVOJx$D>*| zWX1L(+gJ;38i!P;3a5|Hnm~R;RkH3PP?kWi7{m6t7M!*@Taw6Or6Z))lJs!nM;dyP zG!cx{CDDcl@ksXtV%C}Da6(btNs7_^gm1zf8r5-Zcmf>DW`$T<#hi8ZStU`H!se2= z!(-WqrSzmhV|MhJY8ND1q<*GBe{?oM8xn;NnM-xp;AOgFZ6ztq(JB2l%wp516}+dd z+~I5zjiR&9LGSENSGC}5_g)MvOgGIt$>fJoqFAqf69tB{hd|j*ov;8S6Hx)PVUc39 zOau;*F!gm-z^zn|sml{}BvJ%xMLLT`doeb>A-u{0MQeDOrS$N3E%-8~Pm@Y*g}zcv zVzoqZ$?nlfk)q_|Y5buyqoa%fmqWLDvyZzDn=vdePi*!WQJ~oH30QvSViNKoF5raH zh9G1CZI3e5kShE_MWd*RNWr~!5-lrB5@>Er;)WDKi>QmUGZLkR#NhqXR2Nh+&v}8O z>tPu-VnGgiLlVG`nKSwo%%$=~d#ZVZ{Lid;CY2^~ZF-y`PhGLlmD%2q{Ntlp^*$j#O{(v329n2=!l>s9DUYv%@`y}48Ckw*zBn*qQI_eX6{g#<#5!E>XT zY}&QP!kbp zBe+(kG=WD9>R{%k3)7HM>C$v#^WkxSFhsTo`^!IZ!vpnaipu4Nh-VF+@GTD7oa6TF zRv%^f!Qtv^G>+C+qTtY_>R;K&h8H3?4435hKKNPl?&;bcB%{o8{qeAm1YxOo$-3%# zs|=z{7z+3}J5*X1vZ?Se;e2CV zi3G;(8i;Wocy6jvp;r<1<^lY5BV&Qo!S>#GVj(}u-NdTk3k$1a%n|Aj4qP$OfT;rY z#RT<6*NdjqXhwvbm|_8reTXkV^~JUKDd7-a8AXq8)+#lvZ$yQGm7lzA#`Pm@`-`p~ z$jq=hg*Pd?x_x$1t1F7CHgT(JU{~`(4R1o%#Q%zwSx_SsRkcsE$!gl? zY}fX$4Kkt{ryy=zSWpii^h58WP5UpX_FzY$K{K6rT-(X5P<_dA1UX%lnp;v2BsMY^ zezy&8@+i;AVWe&jbtDUNGW2dLi}PwnsEo(FdIU8|6FEz&>~da^6`6{NF;^3kkoY(j zt2tFsZfz|jE^XJARHy22;k{J0nW(EKaqT#xPUDg71ircdQ+^q+!$w-9=Bu#Z8x0RpRS2&Ew`Uv z=-i(dZ`=h-!Ew5Hg8-pyFLd5%3x0u0@exl}Aa{0SeS*Q^_&2IE!zV&omrXq#mW@m> z`_OPSB9gW2r&ReJsuO(t=Owhw{e^=7m^hEF^`_T1qd14#n7*3Qfdx_Nr5^7_U?3l9 zlDbi6*L`wz(4B$R3!5uU_1vnJJubk9TX3H1KqYp-|S;0dL zXQp0S9upE9%Cg;RHD)S&hHCYq0OAztSP`T#4l+fwT8h}JV{ZyP3}n>6;s4i zDNohVfG=hTv5~6u&NmHI5+O#e9Lk;aSFk2&m!4?(!HDAvit(iV_=_4ziuK$2iWg!q98(_8i7~zX(Sz8bW-ExH6+H^fj0RH6 zAF=I6y$V5rP(9Bzu&ExS!>2m3)f?3@Ze;^zRMV4WFPUekh(Q z#p7xTRbRQ{CU}d|)5nxfl{Gr7{BRRk%RHjPj2~+gL+Gs$GUkMR%D9Z2=Gb2lO(b*b zY;aTR7J1;`q#7`+shC^kC;f<7g)2mnrZt$=tZySy8rdiMAI^E9N8)FwdmyZdfV>LT zqi5_NB@No%W@=(3Ghc$Tb7n?(T6wfh3$g5zX&jUKJZDj2!Q7(NJ5BCJi~aQ)A!1}HjA{D z>v&UoEeB^qPPU@{@yw;o@y3| zJ3jXfRm3u6Nf*Tkg6Ddq3Ph-|N-EDHbCE{0lKS?#PgwBRdNQ@r6vsW#6Vs@dpD zQ94$iHJO^5@}S<5+Um;M9H)4`i`oLI1w5VGY4S&jzD(_->oh5@1KT%~7koA}aTJuF3f9pc9{PoSS5d1pB(}F^%Fp^>#b<5N&-7~i>)BBy&mKlw})KN~r z%W}RpP8yZ?b?Y$91i@#6u!!hb3uPWsbW;Xm-LDl-^f{^A-!LA!VnB;E7Kne=@M82G zT!@=7y45fU37>UDE*(Sb;OY^(@$|g)hA-KTt$DwQ{%8ixs#~7f&)YWBhC`^mfR>px zivSm8)m{RJH~ZuMXcUclL!8Rc7h76zE64)qwH0fchLhfkm;}_veIH_r2oZKL)No~T zbVS?~tQ9X$D|2Iq#H?a_A4S-c#SHkyN{{EfwM9ecF&hzU{oU9a+UvbYEE~ENZGWbqIzDc1fOp!2SaOBN^c2rXLozK z7H*5w7l(=$QXAO}pJ^-SoVn9K2^Xlw8@oTVi#$~hPW|#)FeQC#w_|-CK)o#ed*1!*{r^6e#>yk^eCj{Zyibbiat`j}XWLC`@uJF7q zOSM+h55G%(S4*rDQC)cS1mdLjt!d>SIkVw_UsM+c6X+*>aHRw&yMd88B3JWX*)!0l z{BdNKYVeij;j;B|Y-nW7y4C>E@j#@iEu&^#Yrt?TFR~;!brlC(cBHy0>fj~Kl_gA} zsuWuH;(ED}2M$Ws2}_$*!F3%_9NT0`7$1UJMoV9Q^e5_cEP*qxVN%r>Z}Y%GimvxkbkrJ)d?sR2s!4M8}-q>BYwm4-@{NipU3u(48=AGYC2 z;g|17gXJGt9p>u3xLTv-LwpZT{lvQNNa~?G$+dAh_&^&x0hj5GrCp@7hH%CW5}pZ* zPi2FzZ^LnG1{RH&)*5GXhN{Iy%kzvzE$8-4YR6Yit)Gcrz>u!1A5j2J?f7!WvmDN? z*LR3sNM|>pfLeC)2|c2Ad^w}W3gtFu2@g49$Cop5m8fk!jLQmw6)e(T(333_GNMNR zM?KZe550>JF0tdwnXY*NSGD8I`Kc9d@xqC0=1d2J=s9y#ps=#dnLzFMa-~obxP6>+ zrg6z4!nRL%eV{Z2t~m|O922WNSN>g;o?Qd~eCRDo6EVZafN1r0l%DzStXC)ttx>a*v~G!?f(m z*|`7ji?4FF3^u!Rw)`c@XQ^JHx{Oe(YvEHF2+7MKeL!yIZ28H0^yF|W%df8ooZ(kK zc&~HCOc_>ob@=3 z&f}`zd)qm6#!j1jP0L}cQl(1J=#;W!J56fm=GbwP(l{x$QyviiqivsU*zNH-&N;{V z+iw=4hzBGjE(Nt9C}}|ms)}AfTpkd&2ihk@0dWzOi&Q9=KNJuM75;0!>&!Q6tyycn z*(b?^)1>mAZ{26+H|t^v>0noB=B%9C7I+n9&dS&AX-$?nD__1IwGmfLQ+|*|8RE=Y zF(Zo1S@D|k%vr^f9tfKyUPg5%&YX4G@L`UeASi%V9vmXfuyu*cV}#PoS@+9nxQFP$ z+WicODS+N zU`aD)9eRhoAax7LvVOY1A9_Oa`%1T{XBI}9IothS)JsdNtPBnz4h}||Ia~2818+r{ zv&@1KI4XLi;&BFP=4_O}29#3oJrpoSg{4T3jwK}AiZf?jAr&h%(6r3VS@(mzelXhM zBFzOO%V>g?e4>nhk6_S*y+fp z;g4!?||R3$r&??1TqcJuo6su@hvpZ87Xji>G@ z#)r#*HG+RR3V%+sA~K)ByQv1e@4?uQ-Rs-C^Xc{Z&b58k7%1t$82*glkL_q-U}g~3 z`<|w{xbsrO06Nt?_+_JFtPxT0mwn-M3gc8{wFC2tt(s2TW9g+*u2!*JfHR-P+sgzb z)%33)dv&-NbOjeFtR=g1GHf_%z^?PwiS~mDe$alXPDgIz;xe4wO_iQ%@^@GBp6}E!Sj&8txzU9iUeC3T6|_w zMH7O|$Ax?HxtrH#TN_Uvi+L%O(qbq_}K6|bxRWg-iT zitNm4q2}Wx)Xe$VM05~2tE2jjbd*crw_0Tnjc=BC%!GKRt z28_p+zUhf$3o8Sg>?+bi3;M60a1B_ce69Uf5@K~wKSow>#(G0Hzo2}XJV9<3#4H2* zG61i?Qf-ZSZT*{an>f*J6Iw0nHi=p;T?FyM`_Kdpn9K8m2fT02Zq0Yr-D7h-VP>No zHLspnK`5^oL~*ViMDegWSv({eM1jyFwgK~(@u3dL`d;#BO6^^KHtA=uzZ#kMXIE+2 z2C9D3KpyFgEv?&6CciW`>$#ciRdasQ=8P{iK75ea58wXl)~UIMBVL$9BjJ(DLHPfEMA6mbx;Ha{~q zClg$WEP6?xZIauUbLrkSR{M$*%SA`bP*e|-AT#%TC)vgF1PKEV{IOR0o)GwoJ+2Nf zp}I2ovM(fyu~mPiNe8~uyzbN;Ht%J{ws?*un+MA$L^gFTaq!WUGR<(2pg=m9)*%_x z6Rew7yQdaBS-tl|@j*Qy&15K3(}fPtCAUZ1H}dG$i(HTk9&3jbN+m~)ajZC~R9&t`- zu7q6kC>G8_3ZHCSHWg!a7IRb8r+P#=F&S=&No9Hdk1m&K%&fIFmo;iAJsHr16@9VX z1utsgYYy=r_89Pqc&d(=yvBF*+sK>#&!zXhaAAIZb77kTpP%o|Y#~}~^2L67&X0`I z8O|}&=qmdX{7$#K;)atv@q&*nt;;}SdTdf)V?TlvMb=HjjI^s)uRiV{c?pw{yhQZK zOiEHDkZ$cWg>51ctLyBM+shuq(-T7p+IKIJ1X**)Krfo0h|w#|LFDQEpON_%;hCnk z+May9oWy>L;tAc-Rgz-uQ3tIyI_Q>l2|QQ7^S@l>e(*XI^TeG^Up?W(lB(=% zOhcb(rO4a#MCI>nf}&JL^_E&%MmARA$vakAR9j~2i%J%4iWHV~8lMQsK^TH~)5^%! zR|+BdXKfD0b5K74nJIfSFL}Z1A;%%oPq&E~kD7;zlZwO)|Lgt%lf909bQ~>Z0@$L3+#DWXR^iB>1k zH_7Bb`HaIo#oMAGAzIXvk_5d@64-JWEl=c?*C8nOUnGRN@0?w`=Eu41t@&<~u0)18 zxK2LPhPlVgBgLttFjs!vcVafqL7P}?D?#9-PS52x{6LqF^h_aY?I}O>K2)T!X{%s_ z**iS}rNz25D#>R#c#sgD*Zb6y(JqZbeYnH@Xh}|DGu+>!-4!YdB>^u@W=rCg4|%^a zGS6uRac%qBd~0vF9Q-idnQz>8!EUvVn&n>Wawi>Q@Bhnv+Xdz-?hgTcD z%R+JJJfwL=`vY{R_(^7ou`rQ^&n_)Cpeqg(zCg4riwB`C$gbBJXYH*)&c$8k$f-MR z&c$J@>KNGBPq=!R6XBe1sxu6A!(;_!K0DS2>o}oJc0}5tlu9o8 zdh^EXvz^)I=6rK~u{YnD*&`sFNmdQZBa$36G~2=eb~nI`Ti|)9OHHzxa2E~+o&f%H zP_O{$KzVQ3u~3e34~Z)9!ECIWmg0sQ3 z`Njg$E2oKN+|KEWLyyLAkfs7&|AriK}HW!d(AG55u>`SzT8i3vcg6aNuug{pxk$ZuY33 zo%79wiiF`V)Xy?9KQ}Vxb#Wl2eTa@r;wWG>hG#ev8l%U0tLPz`GgBXR$9b2rWVqJ6 zE$b^(X&Ob4WW!PkBrikdGYRodv}%A`9D75!mDSRpm6|)cTm+VlUq{)x7RXA>L0$vM zDvu$r0Vo#o`ov82YUhUnl?22nwZrbpc8@vIwHuX9x?R!msA7ggIP#xyharZDWBoo9DDmE_Pp4eaPvm7Mn6N_Qd6iw+1Z;yQrGsz zd}q2@w-=>Np{tr#dYgQ&#Eqo$lEhiKtGumzl$wu*$l@ZThoCdWSlxyIFWeQ%X`)Tr zZu+PX!-mAhXQ?X=*?$t_!AOoMlz^37s~Gbos--gWK>)+y$>Sy4rgygaBvhzr4RNB9 zhl~U6c3c*Nt=ic{_$UN~$8$SZi-YhFId8R8T9L*HVi_|Zq1q-R)Y_!AeVvO3{oBN% z1mY?L7VmRIu|o`ML;gCb=@pSWfWWBK-GhCaTcY9SaoaUdEcN!SK`ix>9l6v&tpiQG zRhoUB(~#NZ?5N|8MtUuE1G%)R>F-5%>N>46@!gns4z^=FZ147mTU_p#nHKM)aXEwv zDJC~IU!XdEAgg+KT=1akz7>lc2=$R+d>cqTeKwP-ByGKO>4-ji{MMLgLa)|O^i4RY zjpNU|L&eoT?7H^NcCA>gntJ{@Re+(EVjGoT$YW+btq@6?YBJ-mzLjLU7&$~Y2I?Qq zWsyg9&cJIR)ItDVP4Lz;`lY;Tpdqs2Hlx~FArtaGsWg4~yZ@qFD_IndPx=`6S|O%k z*j5!_-ya(hwU4OQQgClMXT<&AK%*@D?1IrqzS)ZDnB?)UVTy9C$o$QSC~Cw6Ac4BY zvdG7Tz!iUUWZoTjMf(iu9gf7UL{7Gcys3F-@ebi3&wkI}0vVWY#r(_2e4~sBKbj06 zuJ6=q{fD^~pMJv6)*kYM(tU-H5%UuxqV=t3pOxZLLWqzK>h@?KQ1A;=NFNzKgfvW& zEqHB!1$fZ{Zcz9r(U4n7tPnEpn;7RJ0~j^XlmqW){-zdRCwF`k0&;f-doyMDzcfy|)Pdwd^xams+u^XT69jU|V)-eVW;@(3aZ zN7waZM4z;{)ct2E*<03%|07L%=sy!}lK9I)-~xD3jmI;-Qs@JonX~TCd5axHJ&>rz z{)KUivVQt8i8a+FyldG`k9fTurl{`ON)P}lkCE(IxEXXe+Qg@-6;_xus^4%fXy+Uv zU*e9onFhRrZ#8*d=nvqr!3 z?38WOEVm&4fW;)))Xf~BRg8u3pbEPkcB|;?Y|qRohSOI1rTo}7Vov&<+g6vl-LTuv z|E6)<0+fU8B8hKXwHG%8j`22t+j}cM9e_3N3PMcM8Q_IUzXM(8;u&sAZpR<*CkX;=6*7=X10huuTfElDQfUk%u4U5vflvcp0tL^P*Zdg215gTU=^DtN zO}-b2vW*|2zirza^SpaajdOVk)VaQu^ODv(L`>Lsf}v@M?2$*IM5`yfokMLQUH_l- z2F8!JsN)H9te8eoNB^7tuFu`8%-2R{jnk9Au#Y|_B8Ct9h~YOC3If&!3Z{>GnomSB z2Ymk1iqSOu5eE_T-&8+z+~`O|Qb7yRnGgS9n7ltwQ!X@My92Nn{GOI12&F<0a(5bN zp62u7(1JHRZR)PLrooT3q2&Q{ytrQxTFP(Ag<)=wiCn6{gvyn(1eZtr-~ul;vglUv zHdHo2c=_2#wo7NYwL()~I#euD0S4Yn*^_6VQZmA5sF8-4S~y#ay~S*E0dIFx+ah~y zsJgru1Bw6m5d*hY(X?QVEV6P_9(F3ikuK%GM3s34tTz27ZR%LqYqAbEZlhkJ#jYDQ z*b-Fslx99eO@U+yulsh!&Z9(KCcAVflkOxL>%z#g$~NLOzY}?_>zuVwpxbh{*ML{( z_RUH;+wT6C*yozB2c31XX1h^MvrnD8@OZ7YMY+*?l=)Svn;L0wbkkSF5uMnU?r`up zUN=O#>&%h1CVezApB(c*Uns+d_}UqTp_Wwp&0t-i`P~2ga|UAE zuSzEa(bM+WWeZ)@Tmb(~!Atb)kRsfcx`AApe@tYqZbO@J)FWo&@wy$zW-Bhb8k_mP zMSqw^vv|p~c)g{CHfTbfg|sHDDF7}is?R&XeY`7_?9j|>>mSTrQuVVcf z#m1B44PR{ei(rVN-RIR24Gxb@i&*t<%V?oOdBF zuWq4Vn8{SO;L1U{3+j+*o_6x*rM&Ek8pG}%gMT_U-;tNWUddVUhTi`%S%z=;+GOyH zvbW_5UO$*i>TMmN`Pvw549&U}9|_vS?sBy@S&G$nn@Sc@N7Zb;GDiEHrOUs93$XUa z>1$3wr)X2&AxnA%h5ID!A0C_gwXwv1H+#Z=0D~7k?Rr^uwV?H@?FAY3vXvDXnP0@a z!$-86@|LET7WJS8Ey=K|6P{-BRZUIT6&B#jOz5k3q#}bRn3MR3-c+G+lye)cwWnp*o(|B79J%J7C zrrvuf>@6~5i8ysM*8>hqVB~@CQ ze^71acwJmjX!(>p{@^P{e23e~StBtY8=FkTHGZRviv{tFxK-bD zi5XwXa%}-zY{pmesy%JV;6$-<%3g(WRG8O_+0&$Av9Rs?vM%U> zw!>u@*BCY9i;=PqdAOJubvK8>5k~+#-I;H!Z_TdFi~9@Fd8t!F@;%a7qF>xVAi84Q zqg#53xJUO9O16fW@x>^Ro{yqb3p(SA`>>09?Cspxx&l*C!6vqMuFW?6usq!MsTp4! zec2!oM{mP9@g)<53~Y~DEM){`yhMInwTOXP;glxwt3FrUWsVxzL@w<|s6+r&G_r{# zK{_jkUZ5YZsF6+N3Ta^pYrIc>f(%N;$R=_jIwYy+LB~+g9}Vn&xYpe5(DoA%A;UIW zl#VEgKPrLbWvE0|s#HUyEDN~B+RLHhmw~U?OeFbo5%?$?M|lj17}=yHPX!=#yNUup zqM+6YurxF!-2AxkEzL5O2Xq}~ecjy!P&n`SxF8*_c;Yy@UaO_Oq|anH3p&??QGbyw%{ zNCDKEn?)LIm1fS$*%QI7`^TU(b5_n^55x;0S5fAy{Cj0an`^KFJQT-hb4k5LX~mhd za&BATRg^g^U$>_~*K`S_@fI;FX(tW3@ab7>N1&bmKhqmv?YR=lP>b5^mW2f}7C z?C4#?NHb?$Hhh@FJC9MPo3!*=_oG>iE6j4k(`&_UC>(MBIM%1#(W6Umd+(B5GIa{$`;zF*1!~kt7@Z5dZ zKN+wPtu%ACTRb6pr9C5cpG=lxC`oKm3MMyOJVGPQoUM44fw!W}S!Tfyj&ju$kFz{{E6tpZ z64-!J>b-{orjQ&!>NK}(2v?dp>k6q@sez_tX3n}F?Dd0bp`=$==4?eVj;RGet|&Q` zTQTB0+)mEBjWlyMPF&+R%D6~*&Curx1ec}u%gE#(aptTiqYltj1d`g*L~o4V*ZMxg z0{tzlkP7{pPvhNx1AzO&yac4CRiS0o?7;2L_zg`vx3`>`)a~33z3uM9kRy8={@A(3 zr{Rx#YUKdfL%gX&%MPry5=FF~bzjmw(Wgzze&R`I?d|6E>;6;ho6H+e-BXMYmw5XK z{^cnAIn6TJ1=Xh=RZD3x`#h)z!hnx2@7nTl2kCiW4*; z)1LV(-d-jksiuGZ80lip87tj)2!-_>)pi9JDXb;Cb24l=YQV1Z)`|9m34YLisFDw> zvsz#2PWHKxc|ejp+S(DQ?bYCB(4kEveTQz>(R3Q0B#9`&hge;^{%Nz(u;x^OF;%;B zapa8c%T#aGBIaAcO|)`eyD!)k-fZqYRg}9z%Y?ODKCnO(m~DJu{U!`REVonpAGgYB zh|OWZd>Eg#5=sF&Hs0cQ2J#7!ClYN9^qBb~K8FM&tT=ccGPo7WL`RWeD~nrviWO!) zF5HvP-Ml{A+IVttZFjn{!1iRKDX#R`#G8)omn1sUcVrf(`}fJ+UPQDB_M*METfs6N zaf4B{*T&rd#))0BGPE19+VG*D*z_i1_?X$=e%zIQnm-s~vQuE5ZC^QTt`7FMr#B}` zmA2B#qzUt5!oxTi4D6CvykW06h)zzw<#$CoF@Gowge7TJ#KDFJVS<+(N^738%N#m< zhzMn8Aq8xm2bUHZUddL(e(F=}H{glv*{}*NJSx4gOqeb^xh5x zyCIkh>YVP=f{ZPiDUpQY=oxc(R2*14WG9P?Z}<(77j3m45*+cHp1nBO+x_kOkL#G} zw-od9!a9kD=-!OWeV!LOl-m8B8~WXNr)$z`d$MKchl&7}Mroy`xdt;*#FT;aMB%0& zD#U4M?FpgXsD~_K=Zg;dJ75m! z9RQAv^bj&vrmy7#=M1Dk6Jby387i(zp|F$Qh)VDoVbY{pds zTJc_E+#}b3GrAsZKAsFVAJ>A-m55jkwV-R_%g*kz~pJIla$BA@ zQ$g)rel|%I&-~TMyg$22%QjFoY==OB=8s-}WNkm0{LqT@(kbRg2&n+1uKZ+mRdmhnJXcT zih9JWzdABcXg0SBIqX%01SLsN4_-1-lN3Dw^S}*k>A6U3Dw7g+D5^S%IEX|liJD7B z)@c~fmQBT2oyFW#^{F0FPE3YdqF=^$xy(prA>{TUI;D*pHI$wVXu^uVSnh%si|{og z{=@C`yz6`-o~k1zukjsql%zc8Ra6c8BV%-ibIdfl%Dx1@)9tRf;Vipl8ziR3CIvS3 zBS=xi6AZo0ojMBfW-fS%=#fb!DH33I-j|3^#(T&1$n9m1;pvHN=nQ3VUv)*Yw@oi% zb*K?~(F8>`y~0#_^FJf=Ey6QRZM8l5vMGuubW2xBiuKNQ&PH|a zpV2_amSGBO(-Vg+V-pkyt^nz0)Cr5VpXR7wphM|E5*k1|qfHHzikGMuHw{iuwxsUY?S}2T4!ErqBN$LFXZ)6Y?7+%Y)pSy&2*cds8p3rP?XB3-cqX+MY-Ft z?rWdCW0gg@{BzHlH$d^v9pzKdnmzX(SbMrn;dsKh53p%uIw$gx7HV18;dJ#=yO}BIa>{vCe1hb z0JNp90>A6bk-8Dv=+tOrcX+kYyDSuk&O@44v_C+7=z>%HB(ucYbN1R>gV=qSIdbYw zd*nR~?Q|jnvYkT~8_+)oI)^|!EXxd`lgKVSUyW>|4&mE$7aAHVUEJHg{`AglV*zR} zj=KLPg$KelZRT<&#?6l=cyr6!i#P0y&?gsHug-VoTUX}AJs11`m&P`-2w8S5i(l`| zZ^vg(2bYOPGB3?`7Bdit$i+D%kMdkM;IVPRd|nKdhKAO#QM^I^Ft>Qvz;A}_= zCCe+Gb8xvY+47jjwWMFR{lx6d)F}i&wAYOj536alfe8+K%-3>6~7I+@&Qj@GE9N*hz&S(B} zP_O{$KzVQ3u~3e34~Z)9K^bFl0T~!EF+i!7_2U)mgRU1tr7`RkwT;4d0FEBwYj=Hn z?R~#J5MJJ#KYu;Q)KuSyx&Z61ytLr@sj>c|%M;lI)+BiIH4sNNmZOs`0lY^R4IsZn3ooJ=*6Ri`m|Mn!k)2 zTC&Jz){wMF@1bM!$0PH;oO}w(x5WdsMC_3_GxH}DQb4ZjJ?{1=xJ^Z4Q}>IxFOJQ( z=hR!YTsyxvG#mNBOa|ZO`gs?3=Co1PU#_~7Sk1pv<-v(Y13|4iaV=9@6a12uNOCxgpt3Bo5S5<`Z%Zu5rcg86b*s4}h>$BbUJ;=G--f1#DzOH$l zeEpfD?d#ZJNis%!%R{o#sK1;JPONO7eTu6CbfM!ra;Itz7_z%l|=o?UYG( zIT@m34IJ6=vkVD-fUK44)Pj6R(TBBPbiju$|4W4kk=<>s#iNd0AehGN&;e( z+M!=bKPxNg8EYwPfmZz=Da!B7m#OI*2)%gWa`3bv-tY1CBA&xV0p4k#d@u`Xjn}AR z;cLKWpwtv3w_jwVMAi$v(y!b>_3U|U?fMZggOOKr?4+kEtQcE0vHZY z9xvH8y|c|Hp+Ze-h!d4OxT|v=P+8qIcM&_A2p@%j@OWMZ$l_KSGMk(ob==WNuZx*c`dK;H)b#hFJ9VAbnfPw(+_pH} zYzz&t>cWD1&lELuH-6jUa^8n+^8RoO;2ks5;+-^rhmdQ&Ky~~;{<4S11rMr%x8fm` z%C>4qJ$*Kls=T(|j;_da0)j`Dy}yTWZGLM^G@)1PC;BFw)5d*mG5)t&l}kH4SQ@uI zILNRq0qqmA`E#lOLoLNND!-7&%z9cOk~Gz1_FsL6*dVtxSL(C%59hMTqdI5cH4thc zfUYKZ>lyu0UNz8?r3smk_erJc!`~-;WP7a;uQSZq0z|w2id(z6B#fGmjfmPuRBI`? zw_GaJZ3`M@;b#|&MzVHQOvfaTcMVgrt32~JBciAg6MzKj7Ry4rwG44q^P3~{?zk&f zFECq+y~S*E0k5#rCA1Pb*`D&I=AFemgr_`9nfaHI`9>KNel&3vPC>tG5BVYIzC!q^ zwRDDp1|VTEz4?g|(fZc2&q{GAAwb+YdE@oj&TMmYzPY}D%&J*bX>qA1zfbu;l)Z+nm_HM3lK9I)-~xD3jmI;-Qp5*1 zA7+im4x%1NRAc|bxb+&ae(3aw*V|!=>Yl9x0ig02$)1IqK?5I*meb+ckf?scxuBgh z7i|#~>FU+1LVOuss(7{3#8FM0on4QIBv}p!ys~ZuZC)Ok2ifB&-dY%M-Rw|I$8?rj z>=os}fn+N8w_!6(*~S~l_N>wGJUeCEG|Mfh@wI(w&J1Vf6vJsN{Zf8x8!;#S4%+HA z@0R3QgR@KXziHgI0Oer2NaEX8?Qv{%LSYW2bvt3Ww;tCAiV8wZ(iz}|NxuVK=i(V| zN^h6&9C?%NxqsPe>H5z0PiGSCTYX1V0q@Ym9uJO zk(HbBuu~C^Z2G@Mm3aoNHmxgd>R8xovJN+Hqh6xLt}AK|x8nQLnh#M^AX&oezMZl2 zC{dTmE*;9GyACX|74PpvUh6t%Z4~IX-0d~s6}o-1QqH!!|0VXh=IcQZRjk==RMYHJ zCojBg*N9)Gx~Y){M>l=lcSro>m7l=j^Z-eTwYn90R%fW;noE1>{V8h7jEd!oyyneUb4Ys)Jp2Npqc+MG zQcQCOV%)DvC)xH@D0ESC0sJ=wFVV9@ig4qoqKtj>kBQ9HZDfj zv6=5%^oMCQigC^8jNNd8H0^p*eTK=T0wZ!}|12+Y=4=?CXKTk}o6_zj*MpgU6Cm&(CMIW&SS?1Q4d1Gcd5Y>SYdG%&)9LP%M%ZF1g) zyu1nu{c=z)x^mdbpO^BoCu$74e+>TV*nCG`274uE#T$D6$7C74VYS)TA*XC+=Fpqb9ZHzXCW?hPp1npsWxmuen#p)aBja@QQ_v~ely}IIwxBR>`OR%Si$2T`kIntsSmM8%Wr=&i<6FO4uy%WAAEhy;C;S&U zcuCZ9Xnqmz4j<8O%3GS6HY#<{wj{%D+nNO#s7byi!1}qqMgk2RmddYC5q-lbsLUSJu(~UBlxh=^( zBafh^drD%HQZRDY%V7EDpYi8dLJ_ipNvtkjvtR^oi^f?NV=+ICx5dyC*r0Cez2`N6 zy}_f4f~{UbB^FX*?|V>KiM_&{)1{W$Lr#PLd&P*@dsZq;yc_< z&b!UW#wHVSjo&EaBIPwr@oy^5Y_*@K4%z&`*c?+Uuq(t=dz@wu#m%L|77}}iuJBev z%=il2`h)WXNS7#SY>?X2<3UG@yA`vbF*V~WnNJta72UYQ43To|!f+W-Grp2p6G6Gd z>2`HT0u_ds@s-S65482n>a>`zXjdsM-rx<49&?GhEi&>PFdj;G+FWJ+iVzIF8`?4k3Rog*v+JoolmAf0G+-R@96yj=pRVNGGFYV8dC$J2k~!8hX@XDI+N3CGz8{ z#RF!&53~0z4#E|Vz9H$OMmCW@@FSFH8WXvebm$GeKoH&%0a;>X6Zzq^u!J?_VqEpV4@bB}O)B$x{JH-IhQBNZIh8 z=B$5jDimdYTye0>1G)~gz9i(-|B(WxrBZ`Y`@!+90B{vOM_U`h*9-l{4Q3@aR;3w- zC7C>MMht3Fuv(GU_=+k5z6Lyn7}UhcrvqBWm$%BtzYU=PF#Ay-l6to(arR=I1iqc|>?H(A!D)e_Pownmyc>#h*-jvJfB!SjIGi85#9 z>-MxJ%bb-j-ygM+!*=6p7H7_i8Bt`;ir189&MKDNr_5QGB5~%d`y(Ib5QW3EWLEd1 z$;pP^A^|3XL%!n7S$FX?+(YzW?S6*DKBSqm?qW3lp{7fgIqQAg#sB(eiah#ujZ>UC z>(MBIM%1#(&zvBzN;79G_IZGogI!Z0^zOU<$$$laq?xnb;tA0!?QN&~WHJ`-ED=Pm z0L>M@Aa$#zg)Gt4ulooYu%wx@-BNV8Qqm;L`f)z&hN=p(&1~l4%vpyhNo-OICWooI zL6?PZ#F?`d&ob~MZvb##n3sT5l?#Y8h5LqPPVrmL zOzJtshu(JgVaN`>4S$ZppVRQiJ+*QG>>=LNp=AfwT8Sds&blwDpXk%3Wk2zx(+YR( zs9CvGb%z+=e{k*XLAUHTnKz!grx+hDJL!+$Uyj0`(>(8=!Mmvjyzjx-kKOCryYuPw z`OY=5#z092#_(qZe{60-{AG2M68WR7aPf{PT^lHEBOHXJoz z*Lmwi`@sZ1Xg}nqBafEiWS<+E2PDa(tsQ~d2sdM`1Xl*QU78Qtp8F!{J9N8_rqlQ& zNkm6qLqjZMr`acAI4{`gi?Txjkl1U!3i{ENdrp(^0Wsg{h0Y8 zK8FM&ta=SnC=(q;g01Yf_{=2mj|=zYb2qQgwl_vNRw}NFlV&kJ~uZ`Nced)YpWoS3z z#(%_y5B*#l~yKAm>&}!#=&4_g1Mlfor_0*vU{=khTjl*(P?Ms z&R@XKs`iF&Z}+$BKdxhPTUg|V|MS8+iH7Ljj7xl8XeeSJ+PmSBp@bLlP-@Y+p%}F= zI>$L8BvSj`c&BU9YJ0N107sbZYcFgqu0m~4ahLa>GPI;oT4`xML-GR|8~|)W63qWT z(D=1t6sf%_hzfBUT6;ohH|ima*!iM^em{85=uZxl0_+3Tu8B2vPqL>$@{qTQ6`PqI z0FI6H5HeTgRSnWH?8$?EFP={-mkGeHqT6rdJMAe1dv_uw+088-pnxVcj3TF9@ge2c zgm`(wyEo5NE455Sw?Z)gq|}@@=Tro_H)pamRkPb{%Z6Mm=I@mZw-HUHDmw{P^t9kv zg`5qUSkzGS@mB}b44|b5dutHj;~ms261w-irURR63CZN2h(sGb*nFG@Rp zO-BzlA5R9Gk88o^N<>}ASOVDmy<;OPcdGDZ=$Y@q&8h$cVMz-(AD6#svG@xGTeqj~ zQGx-VpbXf#-K5_hi8rpk8`xx5krrCefBl4Oz$)cy?YEMUy&NeY99gy351HjNv-t(( z%j5}iyC6!p4uQn&ECZoZ$84B2|EAm~PITLZR?E6gqLxb+Yy4+_CL6AN_m8wTgSk8} zc)Dd_GahWK@<<0lf^@lK@&)CVSOfVGZ7LJjFHM zbi@orv*iVeuvU#w4LS%cwk8JS)I>AR%$~418f~Xc& z?g<9<1nZ{N?x_V&R`2~#d{9qFGa1U%bfLp@$?XyMjXe7GA{XR>$J!xQ3;JsHr16@9VX1uquiYexKs+v$1N z`9wTbM@(MhJL)J&c~0WMWu%&w+CbNYg|+5K#^?;^m}zvCeF=W2+g)+P={#sDWqt+{ z(_@nY8~b@BQxpX+5j`@gBt_?-!Mz{xPYqypbprjpXNo$3e*Ct>2kQ3>+e9K(*V!Yt zmp!)6>50SXMH3VqS(8{d8{SYX9l7H6ba6AX~6Ofs*H}jGgydfW454oq? z#EeJH!^KHOVut^91uFf~akMPH)}sFjADNDWy*^3Q`Qe-oUO!mG>d{f!)^f?w{E9eJ z$CpkTN}a|mAxrb}*gPydMf7MV(ds1nCYk&vn^HYXv_(Tgw5TT~33{C*Fps`6kX1$= zg3sARV*Vl_%zfwV+BHAUZEwwYn{*{I%t5`|Gi{iA%sf(@N(yu3*X5!z&?Xk!N)R}y z({uR^KhTLT$WQrE?4cs<%NgYqkIx3GV1(H_JprY~x^<=b5qfj62 za6ihZeAi|TWVpXayDL-_N&;S*%$CF}8#~Q!jLdUdL0sFuHs9LYEeAhLcjg;6uIw$g zx7HV18;dJ#=yMCKiMAbX+~P1zzA1>^;nha(vQQj44{2V}{s7%6ev(;YYi|t_Chjsv zPTgsfo(@Bc9G?HOVaQ?wy6-^45QuPPbs;nm+4VZ`*hU?~x9Ki4G*Y^_w|)KTo!Q0$ z)LtBQ|4RxFglpQVA;SAfOg!dC6TG?Q?Zq22u;C{cSFg@@=37_h#XT4M|Ch!#vItoy zm^g;^+aq%!)xBSu?JQ;>5Rr>J(KX$b$^Jf{7el3?p*3vOZmtgdq@hHo9S66vW~qjx zP_i8?5i@$;L~IiwDkj+Sn8&q7aoP40volkt0ArxNZo;k(U-o<8XHQV`3TtE05fiZ=b^F(@Fg5?gGlMtkOr?j!7eWd3$!z7;*dEw*6F^|krN0_I*%^Otc$e?Ixl8j=?Gj>27D z$L5bm=6yN&%r!byJdTKzf%y{(DIkC1J?{1=xJ~`Bso~|YEZZJVx2rqyerdiqHs798 zZ_#q?{NCU*%J^$pCb^0aOwA+ju1x#isq)}Nqk%A8ow$~%EdsYmB)J}x8dpdgYlkbr z{PozpH={jQn-N|dnkW{~o)c}^NQFE3`h-WjJ%V5?e1tFaxfKk&)dt7O}r2JSWn9?s`C*J~K#;BfgeRe7s6(+h9!MsVP7Y5nT; zRo6O6Wv)HyrviO*p(0_pMEV!rXgj=dq=%4+G)O3fWzZU`(Jy&Bo9#2n-` zfUNQuLP4=C@Ql2F5&kPJ^<`%0+TKB8rh2vWLxD;HVwBpUwidp-;w4nxuiC;B@u>kX zTn?T##QQzIUg$YqT;!qvkGVqmU>4FEuTjN9%4hb~EqM7rsVUT;?Ceb;scU;t`-O3A9CJmsjy;hX^Hb>K0>ukMyRz(YuH>owM`4>6xji7 zVo?Hd6#|R*xuMu02DKr79n=(;r%W#@sRIa{N9qS;)1g#SZ{HfkQZLz&OC8iIlWANx z_H|A}W|Om{jyqzT1zBWc4pP%`w6sp^Onf(X&RHC8Hl~`Q*x_%s$@{}CfOpJHi+9oh z9zw490@d*Y`O6+27d)uCZ^h_tJB(74A&!UGaH9MmB;US`zG4xwwq6xj4q3D}% zP8;``V{X{1W#?S{HF&6WyLaF}rwTCCQf#C03wg|}rxhYeQ%z?4)wdRWi}MZwF`k0&>(gp-%|N~mxaRxDl0#bWu?u&31QCOy>-sUG zPg-2+9ZH$-NVTGU5;K1$+9dIpg}?>y=P{3Ge5HsFa62T66xl)41Bq(vUl_M2>wgus zeTGhtc)cB_sP5TH5CAHVk?dKx8FaRv_*4~{3P_^*4d;S(&Rn!bvf8d*y(+|);iZaK zOHCZr#F-_G@6GI{@ZLu}t62J<|{BIh!EkHTg zE|U1pS+xiER(v`DYuputn4~kn3zL55>Y6!gOKQU5ru246aE!dk_9RsHT)MuqeR+}~ z(9Vtr@&+MQ6kEL3$Wm!!qw)`g8t@V*c)q;m$M_w9Qdl;&Ao*S-$~Jz8{58uGp#a~zeV~xWnA8ql;6XsYkjpCC& zb>?d$v&QMk|LZ>b7+`w-(Sh1iANKLcZz>c7Kpnu!`jySu?rxF7uI+AjWqWIHw%D5Q z6yxH6`9vghz~?`$7)`?;AvE*fR6lgw=tx9TK?~8D5C348ygyJ=u6Nk(0PF?7rzHtO zsSw0aqjOnd9o>9B@@CWVi8niqK4?Jl*<-B-Jz!v{r6ORI-}HT)!)(m$F_FO*m{7TL zmf-S;A6($YMi$*F-iFF12roYy$#w}8n)1@2Vv!0k@LtNEJo}WA5k^CeG{oT6p4F6X z;y-@Gz^zp@Em$LqtlaQn=9j24&w$mY-=s|)3wur0;l^#$%i>c*yZD+9QDe{BN%yiUIZ@7ljgbyFh^j&AyjIO0%W&5dD=ea;Hp&)KOmhZe+^!p+9q zKPEC)x1miq>JhW?c-?YIpMp1*R$O#7HuHUp{xFSZ@selpdSPhetWklMKojaLq%~np z0dVDtYE;;YL_0ywn1P#u+VR?FWhr(@^eWb$QEWUpp3Mp1T&9j_aCmH5#HxQ=Mnk>d z+lS?0ccRYk|88V1A_x2jr?Wk1nq`LtO?T&;SEn!DnD5-0esI3CHQ(fm->M>ux-S)x zC6Wxd|n79u@qVCWbHk+Rgdp3i67Kyu!Nyu1nu{c^KN7OB|DpO^BoCo25gKL-DF zY`!BegT0co;tjq3W3mk2a1E9{gq6K5SMY+sTvBgSmCV=1Xk%#B_xMQA9(I?jwaHSf zz73IUSwzjm>q^b$D`T|JS-SiyxBzQ!oWAB1bc#0R9kQfXP`FRh{^7B?UmHvOce5w_ z2QYZy)2^4tvieV9L7`(EM!d&S;s9xW5$_Hk(Qe9HnqFEI&Crqzt2*InCSTRmbWP4{ z_0C4kJVxOv+ZNGc7am?flBa#OtnxF*9fa$^`ulidFu{IR2e-A@Tg)~W&_;GDo`o)K zbj^I%KN;BTZgpEmO0A$}7S{NRS=^6au7OZZnW2l~5jf4m$i?-o*|m9bf1&)Ov}-U! zb&yE379>May{?&6)6$Ey+*1Jyh8;L;P&Jil=g}-FPA@J(cPwg z#QcM5Gso-VfOw%m`>r`DX)Wn;#gP zV`>FuHxFQg&vq1i92r!5(rC|1ti#>V`m z4!q4E;^@hjn(>t?p$fQtUb3f2#bROG_hnts18w(3%EXK>M#}pJd30(>zDGJs2oYT| z?$Is1MBJnM2n8omGrkxF((_S7$JmZ0bjBC=VHfw<+qtoI1*W0`3}(-3n>lps?!eTH zFOI%!5J)GZ+-z~j*&3RK9<^A?2+DYg{J3fn1GB;@Smamzy`H_V5hI((rTqvMIDUY2 z`lL|DQqc@2k|yc!9C86_WE1&=w6KIV-X}jn1|?!-6S)u_l2r7dV<_m426jJKNuDK9 zBb#V~B=JWjkh~0)&>Q)90^L)v_Hrnoc*Er9gn!nZWbR36ZEnDzOPs{%dse<;LisnlT9esH`i09-}S(N^5}dZE9t zxR~cm?DkCzYErOTk=FQ%DgwR+JcSt4#L1@vTE$29%E#n8ByJ>ZQ0xGLR4CR+GiP6# zZQhu_YBFbCYTP3L541`%XWi9#5G@zfnwl==QYy}zb$`UO=fkZbOk5(VIcRrvNT&u% zGiT)t_CUN4ausFH%DC84tAAh&dRxMfmc!HtbE;`)?}Hp^5x^BHtJl&=A|k* zAIzmmoH^_Mh;?5@=B#*4dFHHQNejY8;puL#%VGLPnmOyT;lmu>d5k)|r={1rAI)N1 zVg6Zvjz|q=l( zMhP^cmQ@~O_B<~dIE*xNwqidWpiO-~>b~o#jR$|EnX}#E3DGMpbGG|rG8XSF5l557 zFG$_0X(3B=_3J)D1}tgjY_}90u9P&%vVNQoyP>Lls>q>J*MY^Evkp;`*rXJU-1RaD zz(_M^E1qTGttfMrSulj7Tr~*}XdHSa&YX=B*nm>%y@vv(8US_e6;B2fXU@7pDpqQs zX_=X`?gxAQV0<)o=`gM|bGD)w$J7EKSCpK}tr+ng4y|PYJ|oSXjT6`SjWRA$UV}8L zq3%G!2aHtz6KBqPGU@P`4?|Cc+wjNEGCmD|+k}3-bM({63;m>LF8N8cn z!22GI{n)*}y*r;?pYL4DqcKp@fie6U!5`bvLas5G8HDw|r>QRPywos&PBjmH*{B$6 zL=^mGUpSpkI4xKO96Ws6YT~qI{8A}bt2@s2&1doUG66|7{p-g_7t8ELJlW6{T%@p; z?9R!s;iv(-&RZwi4<`6Q`=LrcEcKm26P@gHBlCbHd9<}7P#fW9&dHb!YCJ=@?!g5; zf3zLiMACQYb{$Qp@kx@1I^8V6*&w$GYfcpyQ?)x6N6y&3O!ZbRV!jpJL@Q@rzhEV^ zjSsBfEDOWx|F{MJG3vy1U$8^G+1z`oD0hf9HrW7WK8(*=38erX8*d>ygJZJf3meZ) z`2Gi$`k475K8FM&tT=ccGPo7WL`RWeD?^Jio1>ZexNuKCck}veYvakqwcY8) zLcJ#s?YnDVeMLw5j?BVz|310fi-ZKGkC`pZH5lv ze#`HgKNJSSlC&z~NYx}el-4}6R-2#@`%WPul%0hX*wcA%X_4WTY-M4pK2n$4)~AGq zbaDIY9*pKIURVFiL@J6%m~m*4@n(qL+ks#=Udg$j(i?t5(0-!7}~4Ykpo>C(#hyn{m0%^FoJGi_Q&Q+Gh0Kc&BU9YJ0N107sbZYcFgqY%O*OKc{Nykxz&=pznpk7^Bzqbp58EbIc%+oZdOH9d8|firuF9(#q+{5V2mM}bBVzBe z@T=(d+xSj<3c=o;NJ(~cO9v<*`L$h|p}rKwsN z&g7;FGqP1eZzG!OS9TJr=xM>TitM(jg_@6(P&2!$&<8<#2-Lsk0B9+~-Wmk>cn39$ zgzi1B>A>b%LPE3LH?aBmYZPogo(wi0*MiNJh~)1yU8t49(&l@|MpW)p;mgo7--DY~ z0S3a77H~e!J}I_0*gEjd9MFiQ#^cp&C@k09b3s7-biX`u!E z*H5?xtWv(#ek%#JGG>)oze<+x!*Yu^zo2}XJdw6t5EBs&f!g2=S*oqE`8VY@aiZHM zv|83}617}9UHSU}lXWZ@SYu!=&kG*#zB#)!-&uE$&Gm$tjdIj{?uixW+KCkpo0G*u zl8F@vJ#sbwWqhauvOZhgGIeY3Fjxi5Fh860t^2P==Ka}KTDF0zVLMmaBpJ4CdX<%P z^o%(?Dh{k2GJo{)Bbg*&zBD%LxtZ)$bA>f{%kdN|&bIi?mo4Mmb>$ZdMdh(J^Pz*$ zSvGoKX)v8-Yj3h79}uJ;w;)j_J$!X+p3iLS6>=1-DNWICv<6bYsIn?Alhi+gk8FsJP3RveV$bY|!t9wm$#_yGTO`m&7QFn-*qlsoC9++70&SDr zzMN@G^{mIhD^4sI9kC11Z21J4ErZ2)Ah!?vu~zz?5crBc@3G0=M62J1!zk7!&C%(e zGjD*S(;YTJWo6A?X%3{+Ee15&MDuc)Wk>n1Ru@9ZWupv8lfmqV;)8lZn#oY6rVAaOOKy+2Z{*RhYm5eR!DH=^La9jV_|yW*&s>S6 zRn&)J!+dpQp3rP=6>`|C2nkA(o*ukpq$VkT$>Cr-16z775}V4TgdK{iPO9$$+~8{- zAk9|ZWS}jZim^J2xvA`wb(Gvs&xsTO66YB6BV%-iGr8B*YV0ce68uiLyW)n!N0t_*NG<)diRrOP zfsOqLQWV`?oI3OH9EFhM)vH$@m%K#u$fS}KGd(u2T-YWOvAWJ4xxMVMeNImtPA{6E z$i}!((~u)7OQkpeGcw;IJk!)x+mo-Clh{vDJfT~cPppE(cJFFGjb3u|;JGABF( z9b1MOSeu?WY#E!NIB*3>KWoDn5xse5gASzwNoWA=j5ak;DmGD(Gk{KHQB$)>!6Q10 z_MN2yng@$`phO@dDwt-4XYZG>yr5oqZW9z!H_OQ&6=dW5m*X$`J{f-9jh#=EwlASC5tvi z3QIbTPlU|IB-B?5A^B%*4##s)KLMF3dowS2K|JK+={7OrQS)$dQjwV9e_e4M{pdJa z7GG=8|3m^Zea=UwA1q?U=IDplvI$AhTkcoHnL567(opI&c@Rf1^YYj{EIUQ?XeZI? zB>Em-3KhtcwURXG%bV*f=#nETGzwQGKy+uoY*Ht9;_ za~j+jpJ~I~W9E_KR8p8LzpjXL&?Xk!N)R}y({uTagh2O{ALt$`(!QKgPJ&PeRWQQr zot}WwVqKz0s9*;V62kL(pL#OdrBSF4ceo#ARK6GQAjADV+FhZdP!jOcWVR$;`H=S; zBlDbA5ZAV^&A0Y;%jDDP&V1v>mA%FG*7{;=V{xSoeUm%?&wC(mcW<$Gi@nqW+*&7V z-gV|k-9l}2YBX|#8dq*TvOB!m=v@|yL+2sQE7~8RJH<~jORPJlzconBxXT5(sr=nISYJ*|ptvY@-h0+jJKi8Yx}e+rIwv&TL}=`YevR z|0RV7!ZmH?GH)0yx)zg)`OyS#Zh3q0#tdxu$;H*H^PTzDm3eW`#s2@Lv5hQ3<`RSi zsr9Wvncp6n3#sn?(rjlj1A&NKe5e<*reHoVhDt+2YuKpWTpjjFLy1m14lZuZQVmI= zWY=Q0&dQ060G&}ldCcQlqquDQiP@Q{Q-CqhUN_;p$QOS7``_V)aXMk2gNzLf3yuw0 z3ncB4s{(>)1DXsK+Cb$t@zB7i)d8ej*nM$luehW9pUE)6wk9=C2?H=-_X`719ixPEG^zv%KrHi0z>-lXCxcM`H$iHqI! zE8APUdo#F+*(R!dY$S7;YNO(EL+!0mR%mY2B4L$ZmIv}vX0oumY zCpI>Qa&&A+Z1H+Z4cuh@c4WR4J-{ut_Mk`md}A@&n@{tXaYIWM`6L>W7WaIxFOJQ(=hR!YTsyxvxJEkS zB3W6Ty=tKVo0C_JSKEFj2Q&Ffo9k!(ohlDbG#Uug)ro7F+9Gh9M3U=4sqHB(gGn>N zxKgh9>#=!nMtiO{BfK~?Q7oXzN%m0L?!Y+=IWA?`=J6KPq)hx*(%`ufAhO-E2kp^Q zrYx+q|5)`so4N!G#}JaFWe^!XW03r6+N-Ei>eMOMe3*Lers*Z$L6tO3a{6Zm-FZf4 z{(~wwq6H&)_f*O#H(T*t(VlXilLpSqi`lMs#wions#a0!v)%PQ$hq9!X)-;&u6f3M z{h6cfGv*k)`|&o-<6Y)Nag0v$KxA)8a5cxBPDkeVs3$&RWrVr2_HNm*FaxfKk&)dD@EMxWOo|_59jln>wCALIk|nh z9V)sKKO=ISR&Ayi-rkMiz~9my^>cc@xlpY&oV8#}?U&Qbc2o0nBXeFC2U6OH=!8S; z{8nRlhBKitdQPs29-=uj_0e!f?qEB1HlaCq`~$+qd8Z9WMVYFa~_sN^AtqcR?0+7CVo0pao7&eh@|{6o%LEtSrk znNvd@{yDPEN2s>R2(>n8ZC~f&LH{NfyMjWQ0x$c+K|5vYRYCK8mR*aj7ptT zD-I|K`zvUq-o7=6rCzclmpZ65o2>(Gr6IG)*-^(GvCV=kvM~p#={QHTU_p#nHKM)aXEwvDJC~IU!XdEAgg+KT=1YOcq`7u61k7t)YE6tLzS8x(fjc7 z8Tzd;(S%;jQ1neWr;RthcW|&7w~x85O|uWotL4*``@n4eoGQRjORqt? zO*NVQSKmQ~%AA8LS=;)Db6Mn3oip$n2(=JER};MTjD9Iw4b+^qhuL&onve;3pH!MY z{M~=RtzE&x!YGL6qz|316=E8O$)gXMn~#l%+DBAtDY&KL7?_euF{e+*bJ>&s!x0 zE5)UR5Fs5DsAHdt@C#E&9~nM`G)$5$*w<0IIoRS}e3WR&tt3_m8TZNOTx4wkZ_tzj z?`P%lhQ3+_SU&!y7GEd=;zy4hJpTEERHD6>{+-O06u$UW41S2zY4|Rv%O-aIP}~z-*z^NFR!pM zu=OY!)o(Z#v~%Xn4Z)Xfr)GaIBaTb7SwpfL2I2IP{*3Ldy5n}rBfzi&&(-?(^mSW z{Ma^PPWm0R)n(<-f=QGw|8E+%EkHTgE|U1RReQ{FQ`H;>?ydNA0M@uG2r)@#fEOnH z4s@N1XSgZ7T@oB4Z?ZiJm93Vp?`&V5BnY(eau#$KDAqto6~z{>HL_G1*{J*jp$5DJ z3Z5^o`7wS6pcEzyB2n|R$@d~rw(&#sw{3f4p7*6RHHJFZw{l+6dWVP!`%W-44KZ5) z06pRD9BK>c`v0UiFn+W}9Z#5J#WadK`rq`AqIQ$3NJEgvuZ_$arzd}5AAKB1WG)2q zVIP6~rb5Ax`9vghz~?`$7)?PBiAMf6)ejvvIuemo&_Z&N_Y`WNI{24FxjiN_*a8zO z|Gpnwaz%$n{OABrII?h8k~CNZKO4z*=?u44Xv#~6iX-YW1Mj8m$+J%>8DTWkNJC65 zoUO&)Vz#+}x4Wrrkv&v9o?qL@6Iu-f#S zw5elZugN;xxQ%+9u+5g5@~38^7pVCVH3gC-yzbi>JC71|ne5V`OnAlsEV)EA*zN8| zu@;`@cOtKKowGIybX)HB8t@9;zFDbkyBnjBmqLTAnEw*{T=Vsyvo6+bH>zp&sgoC8 zr={qPjunFWRjQjBX>fGYSHuws7jmOax`Ucej(MOjlwm`B?To@uOR9bHjBFck7HB?? zIP5T67J0OV{qgRZfwH@fad1UO+vGYSmQ1r{O%Hj%u#I*N43Wjlgn1$ba2{mEzq&@3|YQ1>vZ&L)8i+FPAFC#rbQev%cg`U+JYPjaoUV49unlhtexgxK5 z^VJ;EsE1$nAz^=;E&`&XHp&)KOmhZe+^? zMCR%?vF#{<>h#4M^POAM56*YC z=9^xAcToIRwavSu>Yd~qp}9-Y7`kfIqlNB0Q`9}V#&0`ptN+Rhz_TqvdeXpjG72GO z$u^LjcOfsYfl)HCH@+M+n^5>8h_K(3o9h>jS%V4kMtawB3|ClVpH+*d} z=ajO8=!#pt*F?I5Yq_M}rYf1QjnT%?tV{8cpgrs^S8J1{SbZbvQ65o8)oi{pM*Ezl z%fEsPu=d93YfeF@Xj9%HOL_%`SqwWa?H?YS`?ayee>ZcsHZmr!J>fr!!Q-Cxqj}6P zc;xGsM^>@M{36~RKBC=}w=}&Vs0TG@NrqLO@HCUJYHGSB=d~(jqh=m`yT)n9iC`BV zUOfoFQP;mjHu$?&bU0*z0I)Qo-7Ow6?KJrwrFA}I*KMjS&TeL^}gThMeeORLvZHZ(fh>`z8S(nij^AD=c9IuND3N4?K#~*yfi0^Pa zIqxZp zZUm`KJsxzlxTjziG^S>J1>vh(o4tBX1#oU&_DJA>n(>v)nh46>FXo6DU&+k%P}>m3 z&F>A33yTMJG0a(mZ*f@H`q?tG5FMswd@-Mn_yafND|ywPwq$Uk)P?)q-ZGeU^yEv; z_)3*f1>8O_+0&$Av9Rs?vM%U>w)-MwV#XIEWgqg$Kx7|!unB(Qm@r{Or!*|3)3_M- z=zg+9+@t#liF*(;z8D44^HD^{*p4Q2#uxWt7x&oPxv_NxrlM}FZ`no~o5@%~M~@ff zsTp4!ec2!oho%e;;z1zszzIESv6K;%@e=uQ)#3rO-iO)y)}+ZaxaZ6p4v&gEsurn- zG^4Oc`lyjj8rektAT2CmjrYk;5PL zEdhRjHc=o<4=JT9Pwc4~A}*!k%vtwGJbNOzH3#kfsfQ4rk!H@y8SH_0A>=B`oRxpC zL|J@h_~*K`9AoX#lWL#7H7_i z8Bt`;ir189&MKDlK-hhfwls6rWy6O#JgkNYS&5mm?njd|DZNDkOa_O1#hJ72;%T^t z=)v0k42gY6GiTk!X#7J>mn?JE`?!mP1h8xt2XKv3oH^^!D1k=QvdUx3E^}OQSe*?n z&77^+FL5DPK^{QM!LF&mbN5~UWWZej#hJ6+;(qi>d$#FTOMX3YmI#7U8c-~LLFzu8 z7P3TFzwRSsz>;Rpc1zLWN=cI}>&N+U;0ZYt?>exNX3jc9Nn(>yFml(+AOLaZY{jz- zycK26G7CoFsOSNfMQ_BJvrz&YP)fb`>?bfq9S8a03`Uwc>k6q@sez_tX3n}F?Dd0b zuB7_}<4QAUD~fSUEdX*w$*J6mq3>{5v^>@rY36L4xW;dkagp+xjJ+PxoTah5dyO=6 z){{{O=qdt9?P;PnM(=BVpIHi)+el<&K8<((4FK*7^AeDny3G#U?u_5iBpJNr%%pB- zaOiD!9|p+sHvF-3jZecL*POXb*F*dc0WCYQ)=Cu7cGi7K_e7sIE&GWlomRMON6pHm zN?_yr53apE=$8E^^Tt#66yw8X1N{;F%Tf4qn&|D?{a|A9S+0|P zZe$*iB#*Xs1ZpGPEVC!ER%f6?n@IW&-L9kQG(JfZQGySN{nKWnVa=%mW2$!N;>a1> zm#N;WMa;K?n`q_CD$WXJw()`Wn`LrB{U5i=hwCPgtlbxEA8$7Io+`@iqYY9vQkf6q zvsOYWK*z>g$j*RGtUBC5nBZlH(wgT;C~-`f<)?KC+u=h* zC_4)&VCy`%w8-#EwpwZ;+||ct>!2)m=2JpLy10FH4@UD9udDxMA{9j@LR-eE-rIp- zHw1G*!|);={mF|}dc$vsy!c(S>vJSH;s;rK!?(Bl+w~vUYX{u$e_mK8(GcC6af#0h zm4s?mmPAgL8A>fWH*`7F-hSMCH{R)*wA!9*FTfFI``Qaz3p-@DxXb%b8CudPt+X`P zU`7fLMN^$XW^I-dt4VuP5EbGywDyG1Zq!2-vGYX-{T^mdwi`ADjdEui^OM7*0Q*3- zYhsPvlk91bJmhU+<##C<0FI6H5HeTgRSnWH?8$?EFP=p!mkGeHqT6rdJMAe1dv_uw z+088-pa3=qbj~B^{nvzedBeLm&r~ZlML6W#yyesUlTvfuoKq3x-kizORGnR2f3IY? zjc6)W*-5CPrv=X{TD4?Svl?nXPD0Jj*(3cBw?>~B?P;-j6SAPC2zzUw=HngIEE2l+ zyru)2YYBClTE>fx9&A4T8U>q=Cxgw$wP154qE2z#<6DcKs-Aev_l}LI+^NEsp=Z7a zH>&~+ge5KDd|Zs0^FbFv0g6l?Xk}~k27H1t;AI9Lh&kTtrA>AfX`u!E*H5?xtWv(# zek%#JGG>)oKTlSV#~?x36Q%hD<;&y=a=RcpWZ4!2@J#I9@VMP%%o?KkH{~{QqT434 zTGnk6wOl$~`I`ij4OhPVukdWcT%H#^;C*v;YreDY9-He4GaKcodG*8!LV3*~igRuM z+K0`_;vq@@T7(|44Vb@-4|PD+k5m_Bs=do23-eba^Zx8A-P_ohzf{9^2u@}G=;cS& z^3d<UmBbB+)VbWxxyN}<#>u^xPz6PElSz0tcef@GbzG+=wNh~jow!pOlR5J zn=Hu(8Yz!5Nm)=f^3}0m|a{F>F z195jA`-&6GMMrE^G+RDFX3Jnv5P0E%Kh{d$6GD2$UP7s5@MYg~En-{!6{KH2B&x@f z&4cCpk*%3KAs!$XtQbt5SW7Fr)7Lf!q)5t*gCC{U?x_V&rv3a-d{9qFGa1U%bfLp@ z$?XyM4c}&Sc!A9YkF`Szr6Q^0Qwu1i#yB9&mk{gKUmckzG@Dz69QGF6lg0^fb#_BBQrm9c%h;m{w+!D2o)w-J* zH0bI;Rp}%QVkkWs(1aCzvD^jE;qf&?|6!kFJ?CNr#wX&bI%4t~-%&?N%5zvAv$bGr zjD~Yi^CM$)hI7m`y2`!;ztiomxZyO+jR3JbvN7A6t=puk*>aJ*GhkCpk_1I?-{mi z35x$SGT$OR)6`bmldqSP*iTVBp<5bBQe>d1+9?MnCqmcBVFuQwCk|W2CMXVEL7aX@ z0YcEBbRY>0pqFi(CI`dK!r7^vHA+L{OlT>AAWBSW# zrrY#HrK)U#qEtrpmRedyYgz+M)INE~DvN5%Y<*G5qD_&)l1}3jA+zZd>MMni{IfQP z<2g9$LuSg}%u8Mn_mF$KP0V=IJY1YqBxd+uSKwD49Y@RJYc2Yp@R8|kga;31iRkn> z{}J%PB39UrGHHHAoT=kWCk>@eqLDWv7T9?Ic>AMBgNn|Ey)RG3P=#vPDBe zw5TT~33{C*FuC()6N&kYgfRD=vuoGyF zarNqaXTEi1UfgrB|9@$0Ba4uQf{7zWzdbS+Qr-Kd+0J4H0uj0RP%maVHlG(mrJ9s z{rlhH@9xy028IR4hO7mW_Q+KM!L$KQh6-&h7RY$*H}TNGsMP_aT-beaXRo-U{GZ7% z!FC|pAkNJ9jLoy@1_}6BERjeZ13UW(R}XU{#K$dG!RE7LeXx!b+GIzh9ZIPbiC0l_ z5ny)%%(w-fhq}}xs|k0m72V`{qx#Q5!2+ZM<-KLcLOIGkB&xs%S$-69F)(6cfKn~% z$1BzcT`z`8W0?M1dD)s^J;c}U`u5uUetRIiyg7gVdXTB9z7chC%U^kE!Sz#P{Y952 zvI(q7@aAhEuEMQqKwNDTHL@Aqkn>mR%-I49oJa=*7C1oLnEJ%V#!!xq4T&u_RU=ny z`P-5CR`dY3*xG|0?emSrY;QizU&alcR^*dtNLt)$2%B4H{&-~Gmy=Hmf;3MB;T6fY zb;!&fUDf<26jDHn&3oMKPjH)x#-@s=g^@(Kx}%x<;@Et9PQ69Twex#JvymUnWD$vF z->Vjcu)i(Fs|U=#Q{};=zIda7FkPLvmZ>cQw@D8y^_u@PW9LVVlQWP?Iw8UrB>34z%ofRTRS3gtH1s(lUsQo-s&% zHSJYYDRnaUh))i)?=bZr^dHe!Eu{%<#(AA z#W6a01ChNY!IctsIvttcqn`MPl@aF7Zg1@ZI9&dZ`(g5Eg*4|Ca6Z4ezIO|nliNq$p`t6VxkR4Ps?GGm z+q)4Q_*+^(4^g;zJ?f_deRH8AVYo#4JF)q>kvXr611ar8bX-#JU!XBO!9$zowIZSfiD(^Hy(YPxfw#8UcI^HFL( z8X}8}kRF205c75~N-4@LAh;`>w!?{vXp)Z4cPvD8a;vZHcqfg^Ayl}~7>CJ?%@?SSAIPd69v3{Q3f_uyu|%SMHudz`OsbN!^ z{;TgGLuEGIN@KSE;anDZRObx520|?a(A5NQJ)>XBYqY&&X+kFCeNt)q@VDp+bkYaA z*9!5+2039^-P(U{u8rwqBck>Z)mjSfEoYdxiWz8>g`Zt88p$_XF&&dU-sK((MuU`T z{$@lJHDUsgK;2?l45HhYKs>Ph=E%G|?uzyq)H@uBTZx=(4|!Ab&f*=yL!SMf&j|z5 zt(boqnQxRakq}KD@}tRph452r16c1rmxax-%}00Ivv_=PE? zj|?9|8YamW?Autn!t4zSA0--cD~T0C#)p`gx0ABo1x-2depViD=&MzL<>PN^@r5EF zel*VUsKpObiS|;$8VFw;<0<(5>&+Xl&vs^;oAb@}1!Qi|Y+*lCny(s`1N%&~*tN*} z?%oVa#K-rMH^>NT-`q#;HQzUu9MXD^UAW65h!`AQ*N+i>(&AFrK1(%Rjr<=<+e81E zXp_WW76KQ*lWIJk@s%Pz!1*w1JO&0)4O`M5|gn7qFzC1Dyvd2-pwJ_eg zS^u8rO~oYl0h3B4+ORoP_<;k^H_JO@5GO@GHo=r_yn$@b8vV|*Q?^aB+=6hV&V&9~ z{doK{hm@+fvrVx|F`Ty2FXhL!5p&Y-psj9-M!P`sziHgI0Oer2NaEX8?X|h1QI+z> z4fj@jIsj|j6@-|iGr$Xzeh0eF#WUQL-YyA_kvG|%gvzE&*LSurPZ9*0jW-CXqS)fK zMwUt=8j!Sm%cKgRC>l)_jKQSfY&??s|);P+xEsh?@MWF40Wz=<-DZz z4iOXfonUAhqCyp{O0Ayob`G_Lbp3zQ8yG*@qK+rbv0@rU9sO_m#}@8hWxh5tYn-0^ zg?;p~pUS`Df0x=^ANKLcZz>c7tPK=QAD1+rh-41<{HGP8Y4{_AX8xP%hmIQ^iAXAF zAv*Kn9}JWC2WrYac^c>rz+UisT9P1?3PB7tI+v9)v{WJtTlkvKi$e?E?6j%7;#vVe z)`pe`%<H9bvTW*huT&lo?%9Z`#;t7rLinT}lPy+8WvRGDeGE@@%V1hhy z{A?uKr8C@Gp(!sNDy9ih+V@iSpj&emdYG22|g+uhW*$X*+&K-b1> z;y-@Gz^zp@Em$LqtlaQn=9j24&w$mY-=s|)3wur0;l^#$>xh&~Ya6jgEV!57@zQMb z#=N-e%#pSfb~G{{qNYHygx7sLW9Lz#E|Xn4lu38Y9MsW}?bK<0C-Pd?IcuXpx8-iH z0k6>Qo0ZD8yX|Fq5R>M=#6H)2J?N~9HQSA9ntkfzg~uCiE0-v+Vt$qCrbZeZ-Sic4 zL>AT~{L9_jS?@KU9P>b5D8q*M+8KqRmQ?%X8QD(XEYN%&aadMTnHuH-?2mWP43yn< zjDssO+9uZtu_QRoTWJV(+g`c`hREV&!aR`zIFB|5bS}|DQ~rz7TbZJ)bwagXy!N*# z0^^m_hLBijwfZ_E^McN(07;3px)pjBx8~AbdVh+VGQ${5$Z|zq^X98Lr0v7QFJF{M zK3UXebFW2o)JEAtifPV3jQdsTBw7NpB6Aztphe9E@ZS`?M9&T>!tKIU^jyGlY5p;h zxw;K)!cmWyjmPUYn=;92Z07qG{b3r-;w8`G^}^7`o*%r;nng>X33V3Iny{t-xTvVs z#FVvHJgdvz6x5E_J}XPHL!wu){)}Sd$?@EsgSMi8+o|x-efEqkYcOom` zFbdm=GvD>kdF&+?T4rI5ub9RC=q2h=&8#Nv@XIX`1i!RAC#u&q^XatoBBj8|fCa;j zu&W|RFBz_sG)=HnXzs&dx>05`wxz?Pch8IPrlTQFJ@6z z?iFzRykt+4ip9dV@5{QN2igvg2?8k-Grkxp`;dosouS^Z(r}SZ4av7dXNdq8!9ig$ z?$Is1MBJnM2#I?TGrkxF((_S7$JmZ0bjBC=VHfw<+qtoI1*W2ItZ!}aT$^p?(6MWb z)Qm5VzHAUkC!^eK30xX_)M6&mk{URQZWVd>WzS+WI;! z2ql`mM6NZM`mil!A|OkQY$7R73rkqzeex4zP$EV)kqglwNktDjhJya68;;5ciQ&q2 zJ%$?DL=z;5KPrK^x4x=qiS*#XUJBh)u=a8&w=(b*n~5Y}E`*O*P}0?$AjlFUo3!Mq z0HkhLp#Y?PMOInB#r(MNEzL5O2Xq}~eftc#ep#_e<9Js9xQd>mt+?^^hWHDI#TsHz z6Y44)X^pR_BH(MlQ;0!LoP0WY zE;a6!%LA>_%vpDJ7UGaVt+`pG!B%PJteia`ZVk%-2PMV2WVkw{8I)$ux-0Yg>w$P7 zd;;8>hF>r#-0 zdx##a-OrHNhct86U5v&*)O5)*XT6WRI7k4?diUZiwl2+_^=OnpBWhXYv3O5H(ZFG( znX?u9my<_-q&^>Y-}QZzz2iwUXS>Dy=;g_fhz5Yp9f1?ej(sw)6L0S?M3`8fRIg!bme`qXagflzQ)Z4FFTp%vo1R#Yzn{ zEi-f0{a~*j%s#`o(#+Y4VjNQofLu{>Dz{?jJMyOwcAJ-G&c=yr{6-lUDX(GqLtaIe zmuV}`ob_bX0lJDnQhS=H`o#NM-)B37Z1Jd?!(aP;5PiRLyu3xA9qva(Ah(_We3(;i6Yw0x-aRT=+mZUKk=lq z_IC67b^k5*P3DcK?kUEH%V8EH_?M&b=d?^kYe&t>rAqAM`wyDW;N4UM-uGba$L{s* z-TCzTeCL|BjUTQtP||@h{29R?oBk}<7|aa9df(Gj7k6H27(l0*2fu7oj5Q(({<1He z)*D#OS*5{y8<__r$)l|uf!YW+Tc*~) z1wDVX9oj_Fcj$Hzn7qV--$W%*e0aP2tC#dHP; zcFIYfRq|M#WO)Y5PAtXgZuk^Wswey5n*8d?k@)NP5`jSYq~SMHAVX&&gme?#a>sx{@f1A7nb?ATV%C+U(6t$bRUmkBR@mG=v3crN!a0iRYA7I47q$ zC=$vA;I@?>OytVCswf`AogDS^?fs;QXaeC^*6Ua4PVI%*@=hovo4utG3}EU9 zn=s-d%y_xNUE62sOSR+<%U^hMPTOczm=@r!8AZP%yB~6=^6TKHV29sgdk-bD2GOv(XfLTU{DDNr`Eq zURq$wSw<^fGw(*ID; zuDmpS5pS@sY8ySVa0q`r?)VRYIFy0$|Hi>$2_@q>&Z@SsYg_x=>Fbx=Gp}MQs-1s5gvbme#bg@Jf68iGA#yDlLBritUIYn7+;ZaKtJsE z6elOp23__-Uy|*6)f3~j(aMR?p=yTOBXtv0Hrl~)CSAvYJztdf+!v;_3j<=P>G-pR!_vqO0_xG(koL!>P zboK@4o!{+h7aS6+M!iCN0A*piYg3?-A4Z8{L=_^5s9DH2{%l^bdoC&I)QSK}}(V0C$7+%d91G4h1y5=EF$8VQ05XeX4ZNUF#}MM`5E z%kPvM87a8UPoi~YDFe;XB#xvASVRw`N#Y53U%KjoI@(hgD285^c@zthVg6I@jD81m ztvu16YRMq~D{r1@rHNddUS!BCS1b=@b~2=4&a@}V>+(d;pOXcOTpCqZY9Se^Y8qt9 z+tpcANoK8!o+!!^DO=I0dm|)iO+1(_`zx6t`G;f;$HUM+0hKA!p;o-0ZC}=0EGo1_ zF8R2>uFSYy+?w6uQJJB`Dc;z3w#8Rm)PHa`k?DK&$n=ews~_D8+hQ{6e3)ISqf;j= zN*#tQ;--st^u?__FQNlG8m$h3FHfXD$(X9WzGTx7Gg{Q45+`~cL@*7$!jP3hZpm3C zlk?vXFmsO&*VlDBx4ALeR^>|QhN81~EaHaD!$Gqm4Gp^wyVSvPHOmE zd_y;M=|E4F%)3uFnKxx&Xv!`aX7x^oKw-8n3`#W44hadSzutu&wRT|;`kOnu9?eKe z1XQ>`=zA-a5ONG&7|r-vo5 z)7NgWp6&1txZ%BpzKue*>j;#2Ip+g3UVNUIVziqM;j?4oRhWtclS9B-7R`e&Nn~4_ zjI;P$qo`w_xM=x8SzNUT2J!Udpblnkx-kukN=K)gTlY_ft7Bk$wzvH!Zg{{vQ!Z$3 zo**vpNfL?I2WWGR+f%3HaOZo*Cr^$hqm2`z?6RZX|KuYZNr>DK#u+A~Df5f&-P6X~ zK~d)E;bc4ng0NJa4En|DzxWk)sMHiRg+`_3-ezBs!<33gLk*__$)s$<_$ZAi#tmUc zLnfu}^acI6oF^uCrVb#CVV!kDN`RRG`uIm*cddToVJ|pX7VKKGX2`Gxu2&!!7oa?r zXo;9BVAW~_D35MGHJN6+i+|;DgmvCJq;aarA-=UQ9tsyoh>y7niM%znkstH)nbkym zurFo+sH%<#b4oo*BoO^|)eOa!DaqJfgc#R>=UA1>lZtS%`1JpZG8RA`Xzwj17V@Lq zF;)dWSlfk28pstB5loe+FO{fAJuhOZLEfwNP#CA5D4-|UKg3tx`tsTPm~;q_uZ}p-8kxb}f z{I5cplNlE9NFxjjIMC9j-nF_Ki_w7?+mff`TWLV*^a?1`79#8U=ARI0#yA{9k)?3Fwy)Vx>Lr?5x?dajn-H9FH-Yg2a$xli=P z*QVr~vt2vAHb^c8wo~u7w|U1CU;H)J9vmn%V5UotYnI%~?T@TLkh_agwqEb& z-4Rf3?JXoBZP%VuN7JD{fMvT;lH)5;Jz>gI>Sqc;Xup)i-(cOvIAJ(>Pfr@@LPxw{ zORJZF3Ie>Ghq;FHxOE9?IVx&pxV7MvyEX9YMtT}K+b>86a&5dmc4j2EjHZg!7of08* zhK~T3%|0ELjZC)dzVX@`lB{K)(&cwl&uoo~ z3~(&*Wu=#Wyc>dnex*6fb$B<);_Z65&Py{dn4(2cyt^lk_}PJwd(00hB$OjLX0LEY zIELS0Uhp8pnX!+y*Ljyaa$`lcZTNcPPqEfC3?R;h*nTt%EsjEYh;@gUH9%b~3|+0< zQ{28$GB04+5L@^pblEQXPh>q9iV>5>m>2eGY^1W~gAfddl*iL@Oz&{*9_Uc>6~x?& z9&Q2P;VyR**;gT0aJ+YNGMk5g@NuiCk__56U}KXl(8{7U+g?1IOhapi=G-6>LDEz% zXk(KSfGZ=g_Bl5gI>@3n%U=gJDHGd^B){;XIs}2)RG-^vs@J&&Hq{Gs_*4hBMiRSb znv7fdfT?PByzLHKW`Pzpbdgxf(sX;Yv?}Wi%o`&UQ{>tKtaXAYV zZ398Y{*u)RZsaibFNT@P$#i>w2?{Qgr`jFMB@xeX{^O>zT>tiMv7nP zBSYxD5pw2)Q_8r~z4<8Yr5O9?l=l}Cd)ql>@w-?9hBp;!tNe;IV%FgbQl#k(W;g2_ zrFlf=qolR$f4DaZJxV{r-UH!H1mIO}A3fuLlw;6#C3AJTV+_4zCFC_+YC6K-V3M>! z))kyAb9Rd!17FP8Kt(CNcv}zYeYA2dB<{^e1o5rqpe(!ZVuKOwU1h_Or_rt`zO;Rc zp7?AJX==0)fD`Iw(;^)blwVjGwx92bo877?)1aDjq>U0md+fo26J|glfnt9!ned_Jg9$?(EWt4xX^V<{LD^S{n%Ht|1rza9;M3X#Lpkj8 ztTf!LSt}!!Z_lZ@4>E)JZSuSA933nA9<0!w3z!Pwvu*q{AbkJYKF3MxB)jY?Z;6P3 z>$-9PnUm(Gx_W)}^!(HiQhaliFjNCEwM^LY;)lri`0ubTNqc6&Pyu^UjfS)RBvS_{ zpgpcnSVSEOl#+j@-NK~blCM2S`7Al<(cW)oV^ov2Viow~T?|5#P|O{Hczju5 zelEv3L#B#mO9dP@#JP2AkYPjezJOwZ-NokL?25x#RchFO!JfHlDW!NvPuxhlqxR9l z_R-Cxj++5F_FU&+D6IpVZGa?6e_G<^ z&igFLqN5POsXP5bepn7-4*EG5tBd&lwDPDBlK7`EZ0i7Jv|I$?ExY!Jc%jcaR{3sS z@;4M^2r-Cf5HAe+Ip8%Z=ZdFp%E*_})f}81IoXc-7ekr0CY#6e0D>ybW~eAh25C}R zwnZC_%%ukTX%7^{Z!kHa6;RN4X~U20Zvc#89*~n1kj3{bQr7l^^|$4CV;XmDs4-s= zXk?(z^|_Rm)Y`#9Lgoo#%MeL;J(wu@WFO~HNk}vQXD0?2-(JTj4~hfXz!{&^p%eem z6N?m|^uMl0ALBl4HkJiOGjG+SnJ>;f5EAcnR1WC)hmGj;fCuwP#NT1_(CtEpMkGC8 zK_>IrKWG;3M@mQA^As>0fPCP0xWtK2dJuviiGZ#Y)v6VwjlaS+E%srjs_zO-Xuy-^ zjvB=41#GPJFc`%*`NR-s`$z?wAwtE;x^d~|3l;;d!)>~CfD{hT+?B{i5=*?xQSH(Q zZpF}$Ootan$dgCrOG(eO3n^-0e4yc)n2IA;onX@J6W&Z(~(GgNKEcAcCiQ=K&N+z|p= z0@MRb{4_RA4HYa%S-4XtFU{)9Sm8=C`uRqMA#E)W1MeU(wiicO4?VMZg{~ zEJR8|(+=m5l{x~7^Bzr$YrD~}9Y@Eq3)q~l6m*cAKaZtFPjCD)pF>4YUkcdUJj6R1 zf4eXKx3ma$TF$adI{zV`g>LA`5)~rFEWHJLWl<`s*FA;eANtn8(4_9smY{ZaXUnyG zCQ{#6*d(NOvRV9T-#X_koc?9Fz&74E{LBIH)H;+m=150FVWkH3jhBKS;`{pII$v9& z=gnC5{w_T!25CUwk1nP_*DL`5W3F{WmP8gGuuq3~a5p5E=1&OZoi9g{*}6_>n9;17 zH(dDgIj-JOg5HkaY1ttJ6?vGbyh89yC_k^20t{*^$1= z`Y|T{%&O+ zAg2hjV%1#KSfSi_xZAwL1G&hKuVB7)qPAujH@!9}6`l`v92?F_qSbL&oq|rssvRB1 zc6zhE0jWI;Pw&DoyH}L4BNge>4Ba> z+ju@;JH7}i>md(&$Wzxh%&iHFdoBei=ETiJyEG^;7giy-lWsc8=xRe|X# zw()XSYAAias1OOr^Y3A)BfpB{2` zPo-0D)vy_OvR{*ZN!#OHM!=Poj+W%MpXbmQO8041K^8SxRfPgg>57*H#583V#le_! zK^F~Ui<&6CJa*-5G5^iKa<+8?kU>yv#9VBNZ5wQM zYhT(xXkHHL19B^8+b7fL>EYHEFVUc=r8GFIytWOi;8xDI ze~3n~6T~w^uB^&g{`ZPvHKo$?!F5FK4@%ic5_aWQ&bB{JQ@0MhvMOi!=XSWptDNN@ z-wnPd(O{?d%dVVdBjQmx%RZA|IqPwx48lg&#kRl4h}m!}XWPf>VGbjL5EGdVtFUbY zm&BN`D`(qZ4vTxR64vd{p!XiPa<+XK9{=!$i&r_TeY_n9F~YJ*%uV~QW>?N?Fmgh( z-n7yk8K+H5Bh%c<*%JNZ!K*()lS9pSb@EUTK>3DYy+oSBb?cP?Si=DBXflpd%dRHrCQ!O}9?1Oh#L4W0`5pF3c`Cv36>EI+_S><*XAx6_oP6 zd$7QiWeG!ATuJ$`!Ts-oTRGb%QnpY9nr2x!+x}o@9*iEwozX@TMcN*_a<+un9#b;{ zxvb(;>O_a`P&YXp%H>whx`9)A<7pQ$tvQdvx!lTGjYdsC*Taxhn#Srr(Z1HznVH0L z8zuR1`}AKU!2Rf`2uKau!w%Zh8DFUOeq41pul9cIdcmc8Ag*~9e#yPYhu~M+Yo&s` zyJ%5|ksY?tN*K^`*L?wZ?q4>|`w@*gHE|dBiiKk(!TQ(DFJ5i5w>?W-zI<8M-&1Tx z=)rIH!mmR#?muClrfT3mg|#2sTbtXX!PaQ9ZW-fe=NK?)t`EO@@JmM26LAc72H9rc z<4|W89#aTdx0)M2saR|qL}dJBoj47{C>BZX%p%!xb7_4+1^!DITf76uo~<}$!yC}4Rl+%PU}1v;|J;wmGo(NHVY;D$)32L zQ#@MB5wHzWGfFrRW`Np->7bm@#-e@;?zW=&aXb$qQNy?yC>zu^*@jaY!c^YQ*+qxt zT&C)(W+7h_F88IgG#SIyZ(KQD?9{9Na0dR+YbbJ!aYI}wu2{~B8=^Kp(gx8c=LhV& zR!l0u#Ktq0o1tZB@a!Ea*VY8*9pV%AJtQ{3viT!W!R^6J>nal4$uQ!x&LAWr_zyGh z!tG`AX$Q?K^x?2zP8@a6r`h@C#SHQFT0@nP0l4na)^=H z+Xi7*&*ffF?h9W5dGv|l*2rQUTVBCTo*yzsaJKp--oq@xR6uujT-x`AwlD}YD3+2X z0C+AMTI=4>nO#P&wokf7rM5%cN5O^JTz|#J_$2fOW&5;0@}eb*X@#WeW}femfvX^E znX>-=Ii=Sg>qlkZ3aiFf!P+50eNo38BKM0%{XFcRtS@XV5kZNOcuO1=;2tP%P1|7i zAbE~EG^q%{hoq124K**FAE9OlTC!qqDgwOJgqm4I*V1c#!RCrWB-tA_hDD-< zf3SJ!85(R}3Jx|e`39Ry5=rNhm#1mC$PG)2Z|e(IyHgoohMDf^`A3Bn`r?b8*Tli*n(*ba-z8DerLiFTJwQ~$ z6?uP0*B$KTd4(2uUmKnoO;*}&bG2b+wOBQudSk`C^~Q=@#Vy%QoQ)M0c=&Ao1NTw~ z%lTUIX}3;Z8*^p}HiINCrbiPEF^SLg#AC@>I$d2IJza(EEODx$wQeVo{B&Qeq(-vS z&1LrB%|=s{X3r;-k|ce2ubFr5vhmjy71NnD_wpI(fZ|p!6}JcjUQPA+LSH#!dA%*jF|?ZbrSgg_RPxTi!OEdn=*z z$%BL^WV1yGIgoI*_>sQ2B|w$Xrk4QN^3=MViuY|{6_CYo)uLO(kTu3AfM&BmkDP9L8) zfzn8L+bkE6;u3m`R)3)eW=Ef^c0sd6_A?#&Bg3PZ zO`uiho$gW{Hh4;RysadrIhJnNFu_+Z_V8}l@9$f8IJ-on>Ff*8JHOl2E;tJF-i-EQ zKw-LTQ=pQcb6KLu$V7C&j4V>@lWONkS;!$0uD;F=xSi@TdwC+6MM1*Wn(u+VspH|b z7AV?4Viv@cO{|B$TVco_O8pfY7T)QT6_+n4XJ3o~vPw`R9^6lUmO z*CTw?TUETV?`(^&xTt?Zk4$Us66O=!tB0>|%v{ap)(YxjTS-sOCP?)U?Bnf8hAR)o@*SpZ8)-DV}e{*NoBmD|brw35s z{-E!zP(sKtcwsb4DClgnBYw6g9`+@~`sVs*W4c`wpAIIY)l(;?*nJheciSC1&;VFw03`2-?`$?kYNp6uRt&^ zKzS_D8m*4vvVc{q5uiM}{nTWd?JoY6ySCTSgiVFexAw(D;Q|TqF;^jxx287oW1c>< znurhf#S8#d)e&J%DOMsuWo~5k%y@g{#OB8KbO;`UoCG+#`Ux4kixA^F@EogBc~TK} z2W3g4S!<^FUs1*ar~~c2#l%8>lsm?%zz1vCBWWOD#Y6;CCF)Bh>QT>&^MxAF6YL-2 zt8abz?0rl+gvZxLk8By0n%*y>#=z<)&&;^KzwiIhrGZq2H7LA^+0|<5QEAC3lU^bbW2ASy5NCZSNFvpXiIPP02TByLNhQ(78y2g^JGI{W26_P7zE)@4CZV z{594d94Itkrb~}&mfXtikE}qDyNgm2kl4vw@rQl!H3{Xpyo~J4p@C!pPUgu?WqkuD zu;r99P^yr;=ItY>QJUyk(mVA8=yjRrX(_NYQl8eX>ixQPK^#VW#~@0Ou&k&roiKQR z0Lymc@S=H8PnhzQ`k6uyT3%F7{0-J!j1z{F_w=MuYP1+C4PMT}T*Go|HdOTaDE-pU2w_b42Kf8bN&boNx25~SufERCAXfHu%t0tLR5xhw| z2d=2j8{92pb_ylaX#-DsE-6CzC?Rm!?9*Y{$Yi_j8?UV)$y)X)U4G~G1Rwvrg2c^D zcBv4MI3Jx}nVx~+{D_0yV!$zbg)_o2e66<^Jjifn?4#{<-sO(mXgSudtUtwC(=dQI7nT=6 z;-$BZP!YZ_(`!a0;(h_x+GB4PYwK}0$qHGOm;yW%ko9y!uu!Ze zi-oADE4vjN)Rlt<*hqDH=VJqvNQjXuhk7SHZ!_+G{f9i-?`7i&+iS2lE*q&0_I++uv=1f$O=*pnEqn&(478d;56WaZfTFI=)zM_IR*e_6h;GHYrfc3)+xT~- zT}VQkh|QxD*xX_1x}x$Vck3c-Toy_1vK zJp6-?TSb*fAZujB8?dp-7HDPB+K%qUgZXW2QUY*g1lB(1215r~)MokXz@~FTbqE5d zq59mN?I;LtNqneY=Ni~lFVNvr9oSk6Yb5z)cFqS(RkP!5ci1uuwBQ*|H|DZ5-5xEi zst*J6#$T=?a#tMAs)$_w1$&B(?Q%T(t#w@PfEZ*i!Q*lkDB99U#pL?p-a>c1o!W9fkPam>6Ro>{Z^24)7=ok9P5PENfoH^l?GOi5Ee6$$X$=tR~_b9F1 zFPlmh+6iXyyI2E;Hx+BE{E9SU*5L|Lr0ET2H|rZEs6_Onq_ymSxHkzsN!3r)(tdcYC{_OiC%>m(heVieLHuP;d=(D~ z-I^E*+f8OI=s+OKvUrnjaIVRi&_qy2yt#*rzSZQjWL$^|;i7=hL?%$|4<-{n)O;{u z$b%&~f&rRj$QP7-m8gj=$5t>APX&JQkSrlJQeg`MhH}{FS!uXgvsOkd-=0%*A7lpc z$^NIdC1)7H@4*V~xqztOPQQo5Rhvm+ZpHH zzxAcYm!Mf>o6X_ahnHejf;!4nRKeJ6z&KC_Mp1ZFaB(`^!DYl z1^dt>uS^{pl^go2Pc^asy(Ur1v%$zk@nVa^Y-j>;>%Gv^_c)kYP0!ad1&Fysg zqex#Sd9+EB+E1L}AWrd1j%;0{tcro{k-MD%yiB!MClzCNqsC9xBobd@>zw+!F+K-1hYyuAt(5eLB!*(og|kDI?R=qHNC} z;N9!$0QGr`!p6xmYq++6u)lqJrorrHVjMh?cXmuJERnUh3DZC_!g;tjSoad$GN8XW zIKvFOxL7Nd(3>yX_}dVIH73LbABmN)Gbr98bLMBG&8+a*s*PsO2?o? zz~Z^Q727i7g>ng=%K7>@X*-$trH3nG&1eE5I^IHAhZMt<0a^F!#gj%Uh+<~O-aOgp z;*AC1uNq#QzC#FclOgx#fyh;3Xaih5WH%no+hq8n-PpwUW%P#_G;42pW}j!8w$_t} zwqCR~hauESKvQPTBf#|xuxjv9@K(G!t;~%bBD0G9eH3X=mXav}-1FWR4V*_cBG&t7 z3v1}F_cE~@XS-{P$>Q5%76<>pb%Ws)re?`)L4)nl+R4FFr$&=AgC|FmjnNw2{kHwY z5XK|zE{7)}Wi)cotL*H%r;d>>i<`3>u^1^0PTQ@ao?3mm9P?;mY#F_N$&tjH{{Fze z{*k7igPqgW)zQ<6pJtI!F&%aXQT)BW`1-^=>~x-GmvsI^J`)vDN;scR3;a6-i}{&) zhu}BQTX$Q;$WZO}&X{xgD-ugaw|Ov)oD;^o`)6s=BWzmw+2B59(Dtc5;6@Ikj;DX7R7C$X#B;wKUvp z`$|0lg_X8;N(YpjtiM0KW`LjDiU=riyURnd$}YK9dX&8>{c!E+-Yj5|xUZ)xKdkYN zPCUNOK2Zx}tUO(}WruY01Eo=(R00FGuySH;xV@c)BWzgIuwn)Ae)}>AiX34E%$rJ> z`TA~q-iL%3HD10@;f!swJkC=N)}(yFX?yDCm^-_?k#ecak+-s?bUB?ll38+_oHz zgGx>NroZBOU8;7j%(|iTbKIBSlcV&Z;9!5T0jBLB(}7^fUA~;9!p-4yl#~$$PK+QP z1coMbUQC9I9BTmFj-mI-zWBPNyz4}J&G0WTYR0Cf%#XB1m28GH1GLKc&4+uU4X6^1 zI%+Jw)VD6pGp%BCfI}_Rmav&}j2VLD0^JSzrjTr>M=U;P->e->`!e8bTkI8lu$&Wc<{WsLv?)f zMsa<%#1H67>d4RPcu@~N@ojzUh%_7}-Cl)7O`%WbB1m6NWCmimKFVP7GTZGc@QbHV z*U-wb_qL^QI3dD}lqaHC5ck{r!nSdofEEqm`axvv;U+%Y&O4PGq1vI>#b`$IqcBQ2 zSY2O!M2SbkLt~+S)K&~<=YciEYQ&suMT;KiIRZ{jyGlo$pk6!)e10LP=AywuPUq2} zFj$?%cQeuiYa>Olam*odH-*7T+gqF4qZMZ8Rgc%Yr|8KEPPT3(yNuf-WQNl4(%Vvw z><2ouhPt(Bg?Wp8PPh-UC2UG#{EQ){FJZk9N>GC1v6gv!0&>{bu^oatPssnlg3T7+(SSs{-zwN)THMR9he zZW|FjAkvL1oQq1`Hgw^&kL^?VmAdVp3X72N^55<6pn+U=rEdE)Je2DX8M9JX`+PfM zVpQsC!ILCv!LQWSpyY&f{psa-O(ha;cBO7ff#bn}-Qf6p^I@Gr)FV;8a(GXb=5*bb z1qOEm(@qT(;i3K38C^rn&qFnbgc&ejqNC=EU8&oG7!C9KLd7g9YJIwGxGbUG1j5g+ z)NO&qu~fdq;cHJ05GpompanaAbFurw@$50Mq6uR z*`t+Rm|bvU?bP;kG~pW^!LHPGL#MRI*F<8vLt?n2WxV`KU5!djAlP5=ye?I{7&fZ& zEOF85a5`K$DVzAF*+RDK;RjxMH`STOnY{x&nY{AlM@K~jYf$!~mO~i=#z^iWuYtuE zs=aNnO|{bs79oKaD6g!wZX zfflbeMtDCcKBLeWPuss8IYa@&EP2OIcj2!Oku`rjmuZ7sIWyP+atpm>eqz z1w>{thT^kiF|19Nygnp;vd1jpsN`tn0IF&MOzUb8pJ^~h_nzJwZmixjUf&+9j&0ZUY)1n_2ASvX zH*jX6`F(z=hNLX*4?$bzMvqDK<>gyDJ~Xv^<^AeZ5x)GxNVLIei~P4NcbK5d%uw;L z$o3u+!3GPidDVViHF${&~6W+aY3&fWw#=KfGbZ& zH&0Gs(O`B#^_!fBACSXAWiF^S>e$*#^BP(`mwQgxkpKnd(bjmzb|k>=p2`=9e7~Qq zezjePa9>!wms{$opzdt#v~NHqUsyInu0Mm(8zLTD>ncZR}~^^$j# zx}*}BOQ=J&GM@+Tre_XGdWQQnd2)?x&9kly8!TO-zUA=#FL z=Y<@@1Y8~?>V(0{!L}Upy6%(&c#CmKqyXt8CBMpU7MX1EKHm)I!LZu{*#rUM)M67v z>%B+C75d_DeKVT}n_I)1)t+}oMH9alY%Jgj2wsy4m zio#*Zbco-4$8C#=Xdyn^7vGQ^%}!UDIl65|3ZrwDP+e+#TO!IzSn8mf_>Ot^LLYyX zQ4gek!;I%D=hq&f!~- z(K~Cp%_&h`g7@gdly1nJ$#x5T$&B$&G1h)rz8}_V6ND)?Q|aIJg>9*{7Fe`Fj<(!{ zK$pKP^-ZQ?Iauv?*l!IDU&- z&y4AeavFc}qxKc#K`^TRsE|tw>ybh{6o`$8lM@)2Sx(KoB-{6@C&p`|l@psA)6wav zdWR$(l?B7Z{4guonJtw6X4XurG|WNk8gfs*>_|_cB6d&wm9Gb2@9Q$!3!N1anpv_R zYN}D&_>snej5LaOG3tQW@F^AYx0sj$k-3JQZ8H2p#-vsorJ6Rw5dHJKZ|%}k%Q!qw zXVGd1Uc885=AS9jH8Zxe>w|} zI%-sxK?Yx*c^n^-NyLX41p*!9wIav>$~8VsB;(-i?E87G0FrIvi+p?Kpyk0VPlorO z`r};}K?V@aT_#}@5T-GAihzI&y0f{d0-#n)Uo2$-maU=umO*v$L$b%>VQ82b9GBh5 za%uMfzP558nIb~NCpVhk0%kNFYV{E8jnxm-MI3jC+p=Y!A`boY9?@)=k#8fiEw-S4 zLyu_RrzhKP%EAs^Oh4MC{5@AK-IpD{1X{Xx z%Wm|Avafsb!P@BEDQ?eh4KsSh=XJ>-b9irKtTHohnc^e5+0&?s=-}f0x^28UbGB{R zKb?|PZtEx;ak!;HjY3p%u-y#3Tpc_%oIKe6507kRnfc%HP^utg7f`pr_2;|rqqZ*) z#sK4ISE!UC(3V|UTL0DF_e2RJ$NUA-ZC2Ivo5L)h;ACW8a9b4C!y1+47!SL+(Pr%uUTg!hb3o*Yd^8z)BDWkXOFI2jLtCM;cVnOdsV!Xd|qQ4qhzkFlDRRd`hTo={rc zJE8-KOGhv$P3NnWY|wFz@-<-BVi4B(Jfy`pM^k+Kqp!PGcWE(C2i7Y)_s-Gq`h}2j zDaxbiMl-n1H@I#D0Xu19V0m==smV0kUHmJLBrad_NaW=U^@z7$(ugI*(_9rwN0S~u zvnq=Z^~Ee8RrMBu&MH=P;iyhRlv)W*(yAp0_Ql9@9k7nptURqClX{~l*>gJ%;_u?D z1`r6Ej*H=o9^vsQ`vRX#hW&T-qim+ceyPOX|KY9tD(U@36U3Ro;E_zgOy7W*V4aMR zFJlOFjEBe9MiBJ4SpRdtd#N$z`cX5pv>)gPJcVf}m6Z*Ph+;;zatu;hOIKMbVz}j- zE5^vypj>DKik3BpnFzrVQ^}OwiPxZvhlVU1w3zcLrUr=%q zOKJqsZf-fC7V%0^%4uoh?ozvfzFoEzLqK7`I?DMwrchADC;Q^-lJd^kNj?LmXp1W8 z4;EeI`|WFJ6Hrsa(R7f13GG@&*+EMV*tgOTn<>Yz08b#$-F>nt$V?dDlEdP2eeqCA zu`Vwsq+2zRGQiF}z1?iyMLR=jxcZSP+bCJ|WUG<01dG32%+}qzEQ*w}{KtM>#}O}W zk-Lo%g)PO!LM&fIzZ+Io;!An7WfHtwBtC*ejHXk4p~{o0Y6=c&!Y00i_B`Rl;iOM} z>69EZ%CoNwDY!q6unp;P>#o>x%+<N7z84e&iF1kh5v6UJ%e;7*!8Gc165IlEelZfVvmk73 z%r}C&_pk-udY&Vq>wMRc!7165l>|N&%k}k7b~1{KJpkymoOrj-Yo*iN9}}7^BAi%g zk}HjRtN(E*bebatqIn8J6L>RsmA7|zzxWe$w}=H`dr!5YKYTC1exO5vs0*AVm4L|) zX<$>z<7XE!)p?G%d}}xvuC0yMR>lxD9?J4VC~?~_NOYv6jJM^ejP2^S2K6@=0uY4WrjLp)r&2bKo#f3&hDIbSO! zHK_M-(IigGNKG4AdWNikun_j%$;oUU{^5aMMHy0JYiQh9i8rFNp)K&j-DZ_-TjnOo zS|(Ur$EPMhR|ap96$uzZ*aADt-v?fK5SL_TGXiGRrcdZ~&?(`KUOvOCB| zFWy!M^OnSQ*-MFFo&}7yR8oPwzWB8sHlxIHtQN#G6145?bNc~}dQdh)7mK;9bUl)a zUD^zD6WU@tc``nso~HO%ADg@H%_C=)IjN|(#gNP|?=SLI_WW`*ft$?or_ufoZ$j2a z`msd9*5M3R66y_Tl}9E*Bo*=N2yqm_#jp3peQ8Kix*tdjhBqC6VL380$Nwy?Cn<_v z)a_I4yQ085dRX6H@QXyZcG2udyhl_F)r`qx_OGc61 zt0u-A%fUrH83v>V%Sj^0U@i?kYXZzj2p|bM_eqXSfZI=`f+TW@aU^6a55PD3SPr0a zXp=TUAj;}!ju7A3XSr~l#2KRgmPlF=R>8);xn)96kn!?d3;GXvWYxs)@ovs4_T0%& z7@B}2JT$EBM>$l1NQy0xu`%(zXgLuZDy0To8}R>Rf)2Z7Y5*kUR?qqA? z;_RK`&(Qhb9f|A=WF}`)dq@#!LfMf+bLOJzw&4e}Tgg;ubo+JXiZjr?ZBpPMtF_e5^Xv=xg z6fPb#kuRxhSeDY2O2XL`alS5hE@lsg0)nbj$s~ZjAfGdd8ESBio(M~W*egG z9&7L-(Ujd9bWpY#VRSrrT=N}u5b}DllwIczLh2vYZ+vM8ONN`FPyu^2t|ghJpR9*1 zZ_^`^FUfq4T8u@~9hubWfZY7zU9O@8odjVaOfSk{kV5tf8FgMMbO%y2l<5NvHgC}W zNwb3>md$OaUHmax4{QgA6Wa74Mzqkju_P4Q7(!$?c>Eq$zR%V1q`v~FrNx*__pc)MDzHSNNR+8Zx;*aoK}(8hQ=9WZ^^(PJl-6n2RtEeHm-?Xg;{QKiU`(Kf%R`Z{`qsdz14Slf_S!%3YEE7m*eNtVV9 zb&Z*%pBL0PQt_W%xxYql6)DrSy@n^8#`Q$OsI> z&NQR&;5ls_3TzhHvgiJi$citZOT=g>IgvnICa^6I^fw1*NL)O4^U)CAXv2Py1!Ob= zdm_h z$0GghRXO?kZsx^C$xUP|uU0MDcjJ!-kU;g{xc!MB3fBNeA}ao8D1Oz@-4GuT+apM8 z9%;lp=s|Rj69C9?5iwSm`!s4~RzGlNcJ9q+=QtZ#-NyOE#4rkA+hij)s*%gPgYRFjs1$$&?V$q#=PaT6@7O=q$jlq&|%R<1TiLq@rA)1egVy0LU z*xL~xymRjldg2v{5!A=vnkW`ci93km@AbvkCq}W;3YT5d`49O_R75G^{9-cd)zwG3 zv=GLkm>#&$ZiE!yJa34mt1O*_zk2e8KFSY{AQGIIEx6NWs*DOoA|JPsgrdCFayjw$cs2# zw`D27H7r?o4(fuA5%w15%?c+_L+5-GKND8LeGfw zB$s#DH*5`uQebf`l`nCUV^&1-$`F*L_V6!}>@I>fubI=@$M18ZBS`0Tx5x;l_+ezE zhd+cW9pQf+VF9~YGG?+G2oMq=HVVkFSlyl!wGH{V3E$d@8H>p1!?Im==qsNz})p zRL!D3x2j!c8r7Li(r)>UZ1$96czj6pV7I~Y{VVRT0i6eD^m7=f-r3nfw+@qHP~QQ=ZX#(Q6$_Czh( zo5F@TbxB~W&9Czv=-Cc?6&Q)-ZKOhhVlRwCPe6BCp+6B8$F;5>K)-=T zaj~6N=t;PN@`K2}(+dA=+c^2{v_j9+q?;dwk*1^2Nscx0BTBsZX*kgui|w?s(u)_h zweF1KpJxV~{Ol?nv&D8=`NHQHa%wIbEaWtccrD_QIEpsq>LR|IktkYg9gfK7N%O+l zkytIZ$;t|_Jw#?iYIzB6DMY-G>srDPP>CBvTJquD0U)$W~_E5Dk3>rM@e<|I^U!1;##H55jDSU zQt@kb+doCiP#v-5)#~yO?r@G-tII#XT0By>j;E1xCwuv|x@=T@YIWI%derKAoT)`` zcZ%ZdTHQ7xdPGEf`e6QiVD)VKtKn6R+9}L{Grv~1eR^1kgqHwse+LbduxoYOr{SRz zf5@1%y4vU4Arm8>n?@7bLdvhz)u7}=QT*wZNUGcTxwCEjTHTTYMHH#cr5P+j365?z zAj*Bcyr)X@GXb_NFw4|@GdxGpI>WGIMj)DBtJ@qKX25)jj+!rat!@kA$XeazBxbE{ z^8=dqSH^l1NU`P<^J{fmU~w#!H_K9EM#p^FwYnwRyr?g)R+l)@5x{afxO5bYU90Pa zP=&GluOlp!WvM6(i$tksTX4U|VK^fK%&Uf_j&!I{cnp|btLuhNsgJLX#FPhn6gf0#2Kg3#t*%C;CKBvp zQ7Vn&)#_^B>#7b}>|94FLJN82Q_ni**a(h4Iw~eIlx2rit6F~H*FOkrq^INQnfA?1 zMPzr8_rT%{)%=F54(HYUhFveXbPo(bI19hzlIcV6OAZ;aj)uaZ&Mv6H&w0p}V=5}c z!2B7FKs1&$+xCiuW96WW{&n+ihQ;$frqa87>h6Bx41^J zfUPt>);&r#;xql#3he;WlB^P+;F?tk@sx>T9%e^CyLkdwV3z8NQ>IybfGbZ&H&0Gs zjbU~{^_!fBACMzC%-E=LLu^*hXJ&0s!>V0I%eyYBNE@y7rdsc zGB6FR0<<`_pREBC+B7#l4tAJD2?yXE-U(U_An076;(w~vBA80x-IRxQZ_Z@41(|@>ntSvm z8#iPj-%TcyMvIMoHl2Vi=3ycfGW0EoB1l$V{JXwb4!K6qo|_Qp@|T5x&~8WTdL{aW zl1Nd+YZ%j@Eox8{Gj}SQjl%j6x+RnI)=PtN{mN=3=lTGR4s`oquF9P+JABEWUbb77 zwk#~l2g+PfbENTUT^i(@y!z+95b={QlMwk75ApT58XOz=TBjZ5gNBbffjH*8>7;x#`3`q zyq)?qdl4jQJ0dD1Qvhr&gVgL9roY};oxM}m7kmxZ$*ClwLt(A*v0Dh~YUc885=AS9jH8Zxe>w|}I%-sxK?Yx*c^n@azaSF% z4>Jk`I>>89kO7oS20Of+eLt@iK(cLok#DaYv^<#Q$+Q@z1^eP%7eV^|pe%!A)2X^= z3jraoE)Nk9f5hkrt0c5mOJ6Ky0hYDE)jf$P8CRj>(>czj!t)_n8uBo-Z$K;Ibf^_X z%wQ7U19f4@9pbiZ*{3i>2UPsqzPoMTilQFU*1RmtRk%-&SKpL5Tj8x=UQCAlqB{~V zKEf{o(n%W@-VP)QLUI8SiQAaoMF)B`%pC?{o=TD0ufbkAQIO3`+}KiwQk<}N7}0cG zsnAnGLVP;J=siAMU)QbP=Ei7Sl|!NVJ7}lW?i%z9&4f?ACTm7pi2g4?F)gu+Y*bmDMpgF4uRS-w!xVemuY0@tDM!jIa% zKo|pzpIxD5r7g0y{;R$3i4sPR`3t1mtk5kS;ZnaX`M;m*iATI?vA($u6W+FqV%NcB zw0i2qbiBE-GTvAnpQsICz#ursOdDk4iTIHa;cw%R?K%RbWbPRTm6ZfHKTl1x=0);c zqab6SxM=x8Sq8EPrT{r;BnO;~S7GcA3?T#KSTsJuz>{s6zYxF8&vZ9dLJ{idbaU(e z$#8WHtk3qg|HO_Zn1{-8=|rf@t)2dAU1RsusUb-6J>!!nN0ZUUiBWdh(e8ip5)f`l zfk~4#X&c%^##w%;CyoZ|`P0M6cnCCM={gw+L{zwaNmE&mp7=F>kkzED!lTmngwo<( z6K!}{O7av?CUG0)&(J17X6ul*U>uht(d2^J0ocM61A0FG(brwu76QDX0X8f&e%N7F44g9WUmje+IS?WZQwY~BJw#)SxoVvet_Y2L#hfc0-e>Jth#X2H(^gA2=>Luaviab)~r0OASVlK|6QEb z00KeRaWQ?-IsB{TI7jlNu`iJh(RcNuY^KD1sl?v@;jONI>HS6-T!123{2zw;2E+tw zM?k)eACrFK@wL$-TSh^@|9h#iVfxWCv$P-R2RwyoD3z5Bi-=-IHd$qwQl{V{Ir-*_ zQRuqG_ZF;v?1o0!cIq7nUobEpX-T7?S9>O zV+!-=N2}xEbTmlcVh$4mJ@!nYz_cGJ)JwH5KGhR1Pl_p{wVc0{h=m~j3i};2oPZvz zrHQ*8Y6I4O*&RdglYQ}ZNqOgNC7*#(v_+M)WknnLe)}B7L?o{(3dNVuw&g&{0sB_^ zVKe0z7T^g4y1P*}1(^wlxa6?-TwgqtQmo6%329diqzte#Pj5FVFS!zbRiv2WmXKFJ z)W%Gv0|+H=u8t;yH8o8IC5xVH)uIQ;6L*ojc@h0)3S7;)5^G!deq#M13SU@q)tB;U z&m?%aNPGl`7)_`8LX{^~)g+v53#IrH+Vq4Ihm$__rBiau+_pjv=zoN*-p9woZO!>9 zYUIi}S1ZHql_`{WZcbGBNQ=glf7x4nw(1kV`0(CpEmm*2e|A4ug@u(9Ys2mB>_%~Y zwnP+}SQsUN6(2|Qgvd`=(7RhUPAR02&X8zGbr%OQlj66~3B_!5vR(I$*VeGCFZ-A- z({m(C$Um?kGnL}s-#lk86#`Q9qth$Zyz==3q5^SxWJ(=^L%=n6hhBIz$o#$-kEMCV zb%z-27w_qbhdoU~$hZ?v)G?L#cC;lCVH>^^-47yc1~vBC_E?j0rEprKK`aKvK;m3t zend$ip{;A9tPm+f?ncGWemTGx3E5c?=5_nu!xnt&d1geH%82PU-~H%;wyY%ZsaUSB ze}YAEXVIV>m@0d>&ugXA-5(R0EFzp(Xp$?9daplkAMk$thdc%Zh(AG>i&y}*H&q+S z5xy5-KhRVlwmBfPiE{Kq8rYQb_}Rs0fa*XGEHd{M7l)>=pKnV_t@F4&?UK?sgMt10 z`)PrB6S_9eLXM+QcEVn)Hdky|ja;?!!#xjgRf3o8lK;e3rlB}BWJVk?nu;*Al`m8f zN{6J$({c{+aP1ywm*f9vB_lasDRHqh+L~jeycKWCw(Wu=h?*X7lh55A-U^ zkhHF$abqRkhz^Igz$**mc64bVY_8)|6QC=Dx5#iF3?XcTo#pQXuRMrLGRcqf5J9$5 zUb_=R^wKiv7>4rQDR&0)DPN~6_>?aY@}N90HyP!B*GJ|7P*oD-Z-l%^m<&HCiGRrc zdZ}ub0yEJqVWYx$eQY4TTxbUa=@)OSgLzBhy6mMyFwX+7__ZE3q{Q;KwIH66pk-vA zqhQsab%=V(O4lQ)*u@!Qz5^(Itd9-e_vVo^%bZlyl@89w38`d!ogUVufOy>Rm;Epc z?VPpvG}`~+O~~3vKbAwR%w8j_Um2hxJ!O$T6DjttH5 zKTGRLnl(jXN6d@Z6xD<0%X(b=+Khitq{PMps>+G2qo7oArC+kLe!XWp~%BUIf^fu?SPsIWcjLzF~@Rn zkxzy}5uzwzHPUJe0w#$dleskXtO+n9A%G<4+$WitfF!B3$|%AX<4DL<9)NH5cQOZn zC@Z|t#mUKVeYCOx|Hv-M_Py$f@!Ckv2A+;iPsMlkSuR{BafYbBC6ZQzRj|2lZkbT; zw4R>y;$_Mf^dE9(%Ea&SZq6$9?CR1)c3x{(+mDJ-1@cb)UbLKu4V6*@t_{PK;5p** zt>I+2wl-Q@8AJQTP|o3iSr7XK8BdiwO*7kC$<`HeiY3wpXroG#0^?YMD*45^#q6>` zQ=_i@4a|ux5br3%iEz+F)V6pvUxlY zp{fT4Bqapwgy0z(Z4fk@BIL(CkPyGgWrJ3o;Xq9r^^80bFt#Pb>{q#%)p-&g_2n>U z2maa!=Q#6F^HR<|dmgDUwW9FwklBXly2lzkDjLaYGly;sIw;$WFrvpbQ{l=x>Tu=t zVkx`M9j??rs(Uoj5SBu038cVcjcZ9}=_l)9%iHw$<4ZE1qZVV4?8YG%!=k9-U9O@8 zodjWBa$b}{A|nE_SIDgM%1Ct}MMIfB&|vch?VmI|2sA%4lIP-&(RyGzIGoU?4>6*J zR-*=~nQ$#|4UK)Gt1~YGPoFz$Lvw>Tm>uvjH05+*K0e>l$Lh8W))hzVrbhb?q92R5 z>-Go|JxtS9;u1{^w?z`~c9qmLDqL|jBw^x*H$qPN>Y-iJF081%@q&kKa4G_AjHlz_ z+8Fwk268fZTA3EGegknn>|$GlSY2^eFM6a!sBl|L&@Ju&EE7js5DaYFW3|j_nw9a{ zxv|8rpKrV48u{;CK{Jzz3i?VbxrP_zef}WR z0Z=C>08E2cb7h7qEOQ6Q2n>Va%_uy0PVekc;4h$C#Aqlvkw9D~>?@lN4CrqT&Lm=- z_-F`kv|+!GTQXMLc&ccdPnKW*v+4w^!xl>${njnUb4i7cKPN_~QX2Q2jSSD@K=gp-I}<;SB=b5&jBTXRGO|sK#T#*dE}yNZYdHA?~Ew2yoc^%^SFt zY9Oxv4;QvEzyip8HF-7iliA{jdg3bWl6iEr{lsu~v|Tb>D3oik6OAv^1?w8JJl%L_1b+Zb#2hSHqGxe=+oV&Pcvonm>> zR~k8@OkBErF>F98a$>!*|BK=bPVR%}F}V-Y2csn*p`}lZfHceKSq&8$-!m z;Y>m`Bk8{7JSO)oG`VjIC3l7MV!X$PFh#tfC%)Y3+a9*P{|A-vrH10xH$CHdi3oXx z_6)5!ZdULI=6OK!)(g{0a>f%AA^Ggzm~ZNd=N86tYjblA%4@c#lkw_kb!B@EHc!K! zh9{o11W^KrE`uxTe{4TS{M(**z5_gnVC^o3cDD;r0s-t*5T*ri$pWC=*^}Q_zt}L4 z3Mu}mFP?)^P}YGI{5~q#og>%g!Q3SxxtmGuzp^ATeo)R+FXB~o7?9-UmEzLvCnn>K zCs&52p#yV$xcww_YtAiRF7}Pq*2Y`g?f(e#+K-POA8)9?wp;$SjnS(*KYOv*Gul{f z{?b<%k+aLhCCmGY$QkSpF9K)c^*wQUo6bkQz^F6sP- zyezNsKDP*1OW>OxM~!_%nzYzEq>V6BDm3rMM##05tNo;eS^9tMc%_=zM@!Jr?^JmO~sXFA2(RiB~;b~)9lMW|wLNafTnQcZ3zs_**eiZ9Q; z47uy&`8uto%U6~)SPIyGP9;_iAOp1RYF*p*XJyXTIC{|o6+_dEt~K$0RL$nbcfwlF zFrYF0g55ghi_`o*s$Ix8^0LFzJjiSl^B`ecZe>GF+2m?|griCb;65B}gHt%$sBr!# zH6Kocl{ciSoEq=`K}U^|+e4yWR066<{Qyx#lc>qUs)u50er!{B!pe^XbmCdS?+lo- zz3=jJ;*gtZjr##bQ&Sicvc0vrJz6P;MmZmN+Rrd%MnNUBMKA5Me2igfWtY5$lKm%5 zL8K(h!^Hok8s^6aLJ1+F?3f6iM+Iyj>*;}-(&woN-;8%c#nfhq?QDa1!$qsZ>2T$w zbaSTJLbmJS2VQwM+V#K+Gt}}e&RXy1W5bOnvR(3@dL%?cGZhG@#6_TCj~4+<*g0ax z2QyU#cH!X*pbBhmxz{QCLT*kv(o%nRZM;6_eRfYYUSg!)4P9{7u0-)@gsSl%V0X(V;!X!2v@)gm}3 zw#06DYFm3^E?2A-hxZh$?Vsf#M1EcTOEF(R?n}J$5`=Jss{3zVd|OZKt4adVWFE;w z`y{Y`g8#{Avd&JN{(WGnDot#%ND}Q_2T$qx{GwiB_|FX1?iFHrj|!C11fGc9t$NgF zJ-BglG+KT6snMy?V08r1(Nm-4gSi5)fq?$z74$d%&z`tMm5fH`QOh-f)2&l0lhGE$ z)8QqVU6@^PV(rv6OcDgsw)^Sf+Nn`?<>4#p^|C_psQ6zG-UJucur%Z^iIR80YzUyU zK#GaoIejY)pLlJH7I=oU8az3gY>d|E&~h1-DlctJR8%(MmO8>P5HOskh%oSA;9JH20>iI) zXVQ4JQxZtFAq_fsy#JSBH9!hfSm}ycBl9zTaWn-ooy>)q%19^GZK0!-@hp|gh@)$r zR}r7+i~Cd1ul9>(!HqnjysAsHw-`Y};u5LSBfroWcPF50nJh`gp*)!hBkm9-w)OoR zPG|8qJ@HjMD(Ye1@K2gs$?Yi;U5T+9{U$xKd`-s0M!dPlSIdxe3=A07JZCFl;2RuRx@rONe zEekd@-^BWU9w@KDMAsxiKpF1MEG#+pfQ4`xz>C+pJxl;v1k+t_aAC#EXx$ogjG0zw zn_;&dBV!NZyZRyqER|?rT5l}zFpJz5qKmF=E5FZhv^trvCI%5EDEbZ**Wwfvnea@4 z)!x4NB;#JQr%H_6*ljL}wx45)G#Ss;OvoKMU*QsU z*cqSA!#~_OR9Gj24%q)f7j$y0-Dyh{-6oSsmbkzJ$4N2(s*Jqwz|Wlr0{K1IN-{uj zs45eqHmXvJ0hX!CPVGG`prv7MQRZTJdvoNo(Mg0lM(Ct!0lu56z@}tW{dtD00$6Gk z7~39jsFDl`s2?Gs6Ri=Bj2gD$IiQ@sUe^I|sNSEesp>_$k9Qv{P zLtb;6u>0W_{gu9$u|KvLMF4k{LA;Gjpy`;0J?i1 zi`eJt2E<$d&ps#;J-2nxEhwg*a`!tM<}e$=CocAIkST=g_)wxDLOvusGh&zRNY1te zS9}IGe((|J7!p1W65>HCAT+-m<{+sia?831a~ikhU2vsxQuiIcs4k&`M%fmLsiGuD z0o1qmc|zF+yP9IoacW_*Spupw?jlw*KEDTo2GEvFJ|x=3lrrsEY^p1IkNem31a#OC zVeRI&QJZQ?Rb-i0l<9dQs-cDK@!DC-h*P?X8Coj?E@vzpHDCW9|LE(k zZQE6}q>PrsKk11B-XNCYmmnd?LkJSdEM=3#hrIcr?GErK6iNG9PMqio*lm2sk4rL0 zBfS%1k~T||8qqXQG;=1cd%nEhu-6Pp81be)&p5h)Q`r-~6NaTw&ZB%F#9ur`5qOu) zAAvRjPwgw$1VoEoBk7U1=I?kbFkr5tf=u7QpR`26o0A@@sUtoYgkH5Dz}u0_&mpFt z<%Vm9(-Tj~W`CD4V|nh4z0ghL%naG<`@Dx14VW@w?2r^@!@_!Rm3&@7EHEMMxup$Q zMkzhTuU-Yfi}N9Z<7)w>{a8w-;NWd}T3~Pp)dm31pDZk&uZ}Z!)pN|6fU#DFuV48C zvTD#wag?ofE5q%TDYT+&PE^dYGae5H6n_B?B5 zqy?M!Q46D5|Gx$HV#1w**mTz+SvThk1@`b;m zUaV)OU|0%qzBi|h=<;Q1_D_S;w_EPa7EWQYViF51h7B_`UO-@vXO*b0juY6Wknt7(O$KS%hS3p*E6c0)tTqa*A0Mr*jyE3P9$0_O;SD|C zP`pdTzU7Nb#bfL!ffDd5yTa3{Ui7@_iaRpr(WEg!TCrkT;fSGD#DogNqSr;rIUjTw zR|=X@IAt$FGb$Nh+@+HJ=S+5ybtE2w_C=VHPHNB+{ z6bv0JTM&UVm5(P-J=ZM2bQ_lwV z+?@Chi_$6rt&X;)Pgv3vEnvb%IO8q!RO!5zLwLN$^4H&JpeJAhkxY8{{wV2ok}Z4 z4G}4mLu+Accv3gtKPM(8+z*c)&*tGDCfoob{i)hqQqYR;?ulnvL`_B0abs^n1Yjm0 zk$?xV0Z}{b(%vs%uSDt0dI^;@0Vc_S1jsIFBSp#h|H@dL@t@0}NRZK{#HFUSwBd$# zQ+yO4HSf0B((FA5%^-atX-3Esh5}Z-MoT}Ba<8^PizVk<)(ejxc{&d7{iUmKl4HjV zs{}xl>32h#qyR3NvR6PIMk-|}#Oo|_WOU(J0Xkq>0|2O&&>*F?>6aXv7m@<-)97x7>E;+|-#)vj!E2IR$ zS64?*D}wie3651pDzb_X^~-9X7--8oA_+(GG?r-x2nS_SMMsx|aC-K~NWH)lv>b4E z(ei~*r(4N7>yp1-PZGT_0q7?ulmKSDEV}n2uwK1 zxKervBx*p#ul2-L)|;tfVv{|mWH|Q`1d=0-1S3gyhZGCAYKz3A!O>Fqt zpCxgU!$*tSD+8wUB%%J}m=N%?>Nn&J>=Y@xsdd~hq&HX8Q(Fw8s#GD75!Q?cX5UFI zwIyKB-dRtL!ht#$X<-Nj+c~62h2@6L$@`ka8NXj%3L7M zjH+g!tN_CQn&KI$b4W#jf+0;<3YWB*vAIeC_f(N`G{89mJ?>SjvYi|<=}ur0l+q&k z*3j@@l`D84WqQ}@s#()frR2(7sui*Ns30iCR9sM|kC!azW0GIy9rx^?`foS$A|BrY z6~iyvTsyO|xju%0f1?w_GlRQ_+xBsZo=_Bh8JsIzyasH54GJi|vkh}I5uTf4)`HQt z@&iz3V})bM)gyN;>&MLJ5lPY(?JmTGsu(uy@p1Jiy6Ew>&Evzh6d|NdS_%vf|=7T|qrE-WacgA7Y!!K&pnt z1RT2RKAz2kk+IK40Ld<>Aj+*)5o$ni6iDoymxC;ty}S#RHi;F383Fa6^OGdN;*1GS z-6pDTD*LAgr!iwiezAsEe5-{)@a>(x4PFa%N?nc=U`V(0cy92c0x`F-u|_V4G_u&t zIQ3FkX`4wL)79JXHZ%grz|ISMxsX12B zBvAIYD?w16R`sb6XOd=JBe4RMY&RHLtA}avlF~esUMc3`o8>)Ne!UX8gOTj{hLZW! z+!gteyldxcUIu+i{EPAn2g|;WQc(^T>x@?pL>~Rm%Mbz_Ekg)$Lozsw#lK>mv2PNW zJU$t&jz=3%dV5qBCt3`u_MX^pw8AZjUCVtKZcZvL^v0L%S$hsIK&TG>zFcX3dGm4k zyOlHBE7w^7n3=dieCy%{1UX`%8nn=*N^XOc#<;eE=G+7%*eQ#{Ufh5qW5D{>*J6dj ziu@ZWuRNKc--{@3YAWN{()oT_%E1R5 z??rC;_h>GQHdMBU-rQZXKY$OP5xeqZqz4NtO3)~AxFbt`mUMAzKm5QAUOB)sM|Q2? zTO1M>sTIWaXoA_8qK>||v8(u{;g<(cqq*HRYZGj}iUKDVQ!o?tY!&mk`0(C(v}GSA z2DKC=VhqwBz)Bt#d@cLr#XjoA(PDOiJ}Eeq(UKP^Hd;yi2$Y!XC>r(=9ucMJ)UOF| zQ$fSdQedb|x<&`IE#gJ6yM*$Ha&`HvGo5oll?gLRU3ei8uB#3-xO&Afau2Za=@13* z7}I%}Nzrf_b}6HhDPh5~FIU=o#o?^d-u?^r?1ADVIbwjT!+p6YxubQ43^#y|zsVMW zwUyx>syp&-6xU}-tXhtWkpr5@q<>qqeg9s30%`!l?^j zelealLF9m`zAkH3lVg2zHQRIF+UDlu$mVL!{xP7UCs9H$Nk~Zn&;uY&WQKL2WO zB^Q0|VfQt}p<-jg&GchD*AZJm3Z8h~ytsj@X7`-g7_N^`9G#9~mq@lx`(y5>to+Lr z?qvSiN2L$NS7yonh9~hHfLF=G(9PT@p0|CP9lr~$uUvKBXxi}N`7C+kO zJ02B&P^DWUnuiH$?oZHQ{v{M9@m6Tk=ZeBf>9flE;nI7B?6I%Tu+3k^3{s2}l7V7{ z_$v=L*65mbx%cfDMxf}ih}7#OcEhAz+`lOO|BVMq)6iBTek+Ly^I7HP1I_#PC4QX>ncS zZA?`7$qyH@Pc}}5R8&yBjg0)&@BPAEu)K-vQKq*-gz2@9dPJE>^_N2)uEwRhfsr45 zy0wlz9T0=;C6?$@TW^{#BnmwJJ*lSk=jxafZ^(7dC6dyC`Nb$Y;BO~KcJkGI%1YJf z>oF$6uGK#w!%J);kO7NIKW0ZE@mpl{x<;u`JZnq|%auj(RP%;|84M$Xh>Tf{8y!&6 zKnZ!VSse4ZD2bP&kob7{gptzmEQC{b+54RI+sWwa=Y zcDRT;oUm3+lAOlaaR#hFlQE)MNP90%+TYk`UAwKw5aCYNLR1;F?yJi|G15V3?lUu3 zG$ZNgiN7!&omLdI&`&oKMC4p+dKIFcb8bHw|KXfihJHBrWD)h*fqRk@4U)xV9rxKv za4Obgi_}C3NVL|9@9R@Z>{<$p2v6tDX(fdwd)-){^(SiLQNaYJF2)enlA5)`Yl-IY zHGpHS{9-sUC{q*{pXE7=M&yPt8|`2~w`j~tYfYMz$?b+@uz{Mz{{t zT@b-koeyKP{!QPTxBz)MP^;s_87S)!ojeJJXk*&)GP@g^dX0$Ek*uEhoWH;Us%62H zucQJoo%z15=zw5zk)geE@fm+i>waBN?)wr3D4Hl__aDgUQCmn$7IHkjB6eylJ*r0X zHOz@LhT><)Xk#066)|N4Pi&$~2S50rH#zKLHF&nP+tnZLhk$l=QR@%hDgsj4V1bO5 z4SmbX?%0m-x9>)aww8AhMRCrffwP)W&dj@0r#J;0_Yxw(qsB7wYL33ysK{{uUEV*0 zx7JpTJeDAhBt^MvS}9JaXS64N+e@ebztUKF2%@}dJpQ_N%fuQoI^@$;nfi|5a@Yec zpj^YxShQZ-cNSHY)80$-R{p-80PlMx$|)B1a$4)4{{~GMz&C}s$oAIe_GqP?4v$H4 zf6tzBag_Emay&de@j?8JyE4)w1>9O{9DWEfkP-UkXnPuVf>7=aTjEy?AF6G0oPfb8R#ktd2IeM+3Q{pcH2LyW*nW zqkf03G;r0)n1Vs<%&Fv)yprm4E#2X3ezz7!KaQA7X#F+vc?Q!i7JHWkXyaRZVJi8Z zeJcH8BHz+RPxKF(!%M*2oP@o4gNKG^MiaSVZ*Y7F#eEy{caV2%56RI?jWkcVr&CC9 z7&U$sR>d@)ktx|cO>L!d%p*+YXiN*@gi0~C7l)i(pfweCS&+b({H=wX{5cbe{Av?b zfBE*maU7=V&$TvD48-=$+Y<20+kgM)Q>R9wZyL2zY3=_TvwTVK8C?C(E))VUZ`k`9 z0rt~4=?_;urAL(ONMr|0*~yFbE$y4F@F-9fmu?o0lCo$#BmQSS@jx2H)hpQ#b)%f5 zLA?}>g1U{W$KfqL9KmPs4oZQ?3Xtqgs$grh)C~SKcIYH52?}PF4m+^QY?{H#r-~iG zOdK3RvES~C$3n%e#=$z!0~!MLgqz!KVM>eeA>KSiDqTNSlR z-3YqPso21r`0;)Mtum5bN!XAllHG=edOLr(fArJ+@awYCt7hzvpQ!!wXc@O!5Euy* zce{{+Rx?3HqvFK_Ql@3ubOITXd?5Yd8g@5<#7Fw#tE2f-=@Z;j^~km)B+HXq@ov&N zPJ+ZcwWLJbQLg-G+5>sMYP8>f_{yr$e*Oh}Wutu$e)-(v^Q1tCKh}x>P&f^{NWP9N zp7Aw}D4`!I?&rm?FdloCiA!MEqYS!1vcG)SvV3B_vqQwVUDA17qFI*2uYv`d)l67z z!opiC#*1I*g)3zj2dncqYb>Pbicj}Nu<37MVKx4k`xtC6S|2PO$Y873J*f%r%)o_& z8nECSr3L5_^$-rmx^@%lyWm_tlYc!8S!&-p5smYi=|gkn<6a_+Um^`EQ86+n>%4YhGMkr7@zHQQs}8N3zAAXFx<~irJ)N*fTT+*K`6+a|I5ai@sBW8ZiMynf|m=QL^P>Mo&ANrcqjo5 zwMB2d`y|>Xc@{2|Ev+an?B<$H?Bl-&O)hLtC*#%8>dN-o_{4~I8*xwnEI54$J^f#u z7q>J#Z#&^OXiEf+O8D=Wg+`Yi?fe%k4`4g27z^=S^pkS@&xt4V=R5ua$b|E+EVbwJ zwa?RDU6tI*aPET}tD|y2d98TlRd}Vibo+_Pc;m?xn2NByvOe5?5=KtU$%@?LqsPY^ z>c8xUeR2CnYir}J?e<@T|F|)FRp)0f7JEh;tIbb-rSSsF#HGs@7cZdsJ!i`&kocW` zdr!k%tE&%ZyW~IPF1#lH0pw0K;*V!;2-<<-VN40v+=9iaIM_6`(!0qeyx|;=e{DQffTw z$6gYW<-rM-8X&7!7lg3qRbt+EH7mr2B0cg#w2fz8JWZuaO+D~D z3yB`xerhuP|Ji%@Fw3s0UbNQUl}h#K?wD*162!!_|A|N2=Du@q2d~)#W{d^;u zJLX(#&biiJbIs?NyE^AwzK>9yk6tzBc+KA!<2T35Cww&0v6NNB7F3FWr?RwoO>#-a z6;$}5v|ff8T6-PE!bWmzE4?SOpi=cSG#S}=iI*%-rU}nC3At+}k#GUS>m%}7?j{9n z<3Ao@Y0n<{V;ps^C&tW)2=l_PN5IrAMVMe)4@IY$tbv|zQU(&wH!cgEL^8TKYTu-#iZ7!5D~381~;1I(zNDS z;-c~P>DAF-8_S0M@u0V-cmB!s(>vqA$PN0mz2pVI9s!Bux;546U`QlC8VQNicbGk5 zj5HoUE#}%Xo(Vph!FBAsE;4hMoLqE;?(UC<{Y_ixy8T4=DO+2rSoeBooh_Tt5Av(= z9a;owR!VdJOK0IWS>ZeeH};MOR5)8U!K?)C>B12Ty`u3I&re(vVjIoD48EB`%O2UY z6)tYjO)F-@?34`t3y})y;|Qs=`oEAOUfY3ZQTRgHDfCTh+U6xD4FL6=!*C4&UfqEk zD5PT^s%Ui9W_X-Q?6)odRE=duGbvcs>Bk>Q+zj3k+rMgwbczg9&Fj?3m20pPfy6ZE^$_VwFW z!fw=7?b-DFycrq5gB`eobU)f&TU*CJfnnrH)FXml0w%x@ci=dw26j&N*ZU)Z_OHJ^ z?Bn%@S{bqybg^~jP}LR#W9vscaAVj)TcQL~Pm58SDTJF-o&&tcxF+Mcj?L)q2PHyeQUqs!L z8Rp+p8zyoXiW`y%)RRx)JT#hJweCGR?taEYCx8q-+QYDs1kxHY<4Msflot<`$m*nB zD|9t#3I2{cIB;T4IbW9f%D6 zkwD~ib{T~RC*LRX88cG4v^%|TcLe{+PAI!_!fUi;s5Zh_y1hCU@+@ulp9rqjDyf)R zWXq1(KqlsLhrK-s#6-=>o$_y^p zSH=dx%e`-1Gy#y2A?8UO&2~8+z)y7J5n+WH7Dx?Jwp1oAzN3;%S~6i0N4(Uek|ncq z_}SS|T91;VkqJf<(DG>P#Ta1GYbn@Y(b$=p@HF6YJnQSnJE2@=WgaHR*F!CU{Ydb& zm4-Ze=a+&BrDdrv)*LXD=cqm>6*dhU4LxYtC9XiCg4(Hu*+mp1qK_szMpwGSxv&z|!f`b%DcvJ>b^DLsm-o#Yx_YzdeDjcq8!xOS2bFPHkEqsSG+Fz&6 z+jZr8tZ%0HbR8& zL_47%xES$~elMjyt;gtD%mv1u6^7U>sLHQX{)_PU7=`5JrpE_ zhO+bj=Z2*+YR+a=XeqAziISwvCT5OQ=%uZ0837@=84w z$ql=a9)t*qF?oW$O2y?mUni)-|BbNNjJ_Za7?)=v7ti>Hf~wQlk9xB`2v`Dd?f3$M zB9^JiO(9C-vU02Fy7&amTqY~I3$SbY{Q*(3^bfQSIi@m1Xq@cjYcz$nBIYjHrQ(NW zFqoTZFSVv+>#RYuc{d>so2cj!+v=;-Vwu`HLn~y4r?4340Y5AwSFJrc2(XO&1x1lP zLZBjIC%7H)2CpMtZAlP>qg>qQpqIZ|t1cm-u-g$vMzH5WFV`7!(94f_9`x$(u^?u4 zdvU7Ot9NaP3Ets#_Nj%avoFq6oT{kB`6qw5)>Nt_T5G&Ei_ioVJDydK~9hk4_cmpL^)@Ah<8}Z(cc`TqhQriX6|w*rnv9>V2@i{MEF# zO8+jaz71rn_{c>RB=Fu`@xQsXR8*D)^QSxAqMLAUDK8cM;~W}39Hw8#;PCqr?3EljwcezkR9M(5863OjLi7nmH8x0 zrRR902dQB9TQYctd!pA|w>DvDCMDIN-dzaGb>IVz&jA#XrnXism}pBAJi>qNmlUZ& zq&S5|CP)|l-*8@1hgcKYKoyOn=_jwii>gp6} zQm96Stej?#xEgKjS!05|QtI&2o#3$0NS!=#=SlhJQLY#(DtYkLO09rx{?IK+tYO7g zq77&4BgqBYPv(Ef;80+kFU`0p>J{D}vHvOi|G&oXdr3vTXc10%_b&;Leli&f$@q#$ zA1kdT8IK}8nbrP-BZ;bkgjzkrr%sbTUC)4O6(pqbV=kHDY{hWR?GyzFRsWBYdmei9 zawvxHF9b|9JQacMZuj3Ep6N$|?_emSo*r30L%Gp`U%`G+Ies_*Ts{2`y@v{a(R#jtm+Kkl!d+8S;2$Gy4U?9pp) zzKR1x{cU^83vYmhBoF&t>5o@1lVuxTEPn3I7602@bpAve3@-lWk4Co*_c+wxCrUAH;D45QLj{he`c-lcIH z#w9G^3)d$`r#JbdKup}2QyktIpWt(*jer&ZkH<$R+5d45yflMHwC&@#{+_+@+utdK z3(NEK-QLW>4nCY!YhAV+25@f%|3}+OutwD1#3_3lJBw?>4KX)t&+XS?+#C+=bU=@V#x@sJ$#-ycw?% zuTkY~x36njZ?BwX@cIm%;ITe!u`6JO&hrWVSGc5+VXT;p<_oXt!uL4Iu($8{cr;ua ztR2Vq#0I%7h~++(ll!UB@BhBUocuaAPQPZ|MzJ9Bym03Qqv7UVtNk;>oz;#0&Rxh{ zo>}-#*gsfbA8zkd|0UkBykl_ZaI^M*+Z_C#6Mq-u?qG9G{GT$Dv%-&O@E_aMa&n8s z?nSmRjk##zQGApGEIvNcyq?}XcENC62oW4&U+!jfxr-3YnygmL%G7{|xB?-!hnEig zvEit0l*yuiW^R$K2+0#;o9z(LC>nTM2LGW=sRFT7xL|wazjM*BSaWxRR7WARta2CGQ>7S$bb>aK` z)bD$s23c3Nv+ASjz`sX$)@JOgU-MXPp1yb_+s@jVwn`Gbt^<#?s%<>j+}RrSde3J6 zGe!h7xN`&Z2SyNUg+ZmTb9;czJ?+Nm-I$K-JF$6sW2ac+4u?D};luZYFys_}mKC!0 z#=|}HJUfTaX7Fg6d{nj2Rg~59Zo(>g92aND(!*<- zWx&hIGGK*4a>rmaUfkZ=S|5z+F}tiYRFLoytQX~08L1szFUteVaH#i;iAta6FJ_-D zfDD0XkRSGVrE9(toJ_y9hB6RJI&fJj09GdsKW#0A7X1Y)UT9$_?dP&;s5P#$T(`2& z$@1f6x!0D>z9;>vR(Up{G~s3cUVAp8wxjj_*}-V_z=6dAGc8`v{(KZGq>F9WEDI7` zC8*TCH>t_;&-vxG|Ci%$wQtQ$VDQbX`us%5QdXv`m0@6!%a4JTadG5NJMg@gc;{mO zQVU+}Bs#rUr=LaZz-X_+?t@58fHNs7SJ)#~DhNHkp)H+~%~GpY9=-$i+%*_%uVRfA zF&Gxz$jm4BT1Ml(k~~GH5;FMl@!ETo;dU&+l>(e3j)tVd2}#?aduCbyvlIPDwpnjv8fi|~yO#KfP%U2u`-&yT>{75!{rjB3p>r`DOH zIBu8ygcESAccZSKAgcH-RV}8L>b98NZDe9m#{OtrS(E*0X7{br;~cA(!Bs()L$Jw% zv1Xw*p~e1=+pzS^+iJWJS=N+V(8j>FOP8@YL5VCHC#X+hNGYK}0=)m%adinC{oi+%+xH=E_35S=+_NCL4x$=mIY?ck;7#?6S!jVo8; zusSA1?p#DgRmdjMD;HXH?#!2_^}j0O4(|?_`Y?cxbI0b+RS!fPlC$$jL|g!EN<^Xz zukMGz0){uoB`n-~3gpLsJi;!G?U6s`q5eoBv@)_fy0R^nK_o<$^NM)Nq#@w6*Odti zBYAj?_AL3IKW%5!# zs%&A_Xs7V|ah(XYM1`iv4K$`7As47MP{dR%Hz|!$Gj<@P(Zl=W;FMgtrg9krr{s6z z;gtGRIjph8MniCiGo*bBqXzd7ru=D&=A$e`fKJ=G~YelpW zS!*utb{1A;G41dHVy5md&zv zrDg3cwNyv>}i+C<CcLpr2E|M1sF%u<_-{_GOr!az7?|ADjYoSofHA0dV9hJ$ zw2_ADM5KGuJhkfJcjyw@M%CUQvnaNZEM=Botv+BJbw1V3P%H1+$-^c_LBMC}DvXr} z=42-LqXVD?J0s1S#pKtj`axV`Q_>Epe0rDyjSgJxz99j+EV!rl;LT?Z=7l{4%a zi=!Bp?(blF74kRHsX{5ugF}M1i$h+X9gKmB=Z0{`=BZPI5i*U>h4#beGq^d)hP0)` zTD8h>Oq-pzy)VOQ$v@JC0ri%&{lPXGLmCUoQ|T!oz2QeY@EmGY?40bc_eUaexZ(ET z@7J4vWi*Qqu?rUqHqE|en9q7M3vBH1A^21Wjt8%`C3N5sR$QG*B<1Qf|0tLvq-#v> zWP8|7^F)d@xiIlI7`tY5M{b>NlRwQ<7{|w3+egYiO-^pI?s7B6RR$EvLl;^oo&@}Q zAGyNt+d271(F#PWYMjYlOnoy8^xqO#Dbg~E%YbQAl~3pBG|J75Kw}@lz(|bbhh`(K zB)33D3?o-AuU(p@F2hKjL1que-%b!WIQuB6q2Eb;7;7;ia>$HlUbW{HH2G*A{wicR zm&m+YfJJHL7`i1#pWY-HDKH(XZ87IKLgcqIdfXw${rKO`a@8VHVF^Bx!L`9|voYvz zR=6et!B?J`FnUUR*5}2>gO`U;u7#j0d(e~ZnhlE1!mW}jcN5&uTZyB}RaU@xzqOP) zsGW=9`x*kJN|N2xp5U16NJFsmzn7<15<%BCr2^r_T@vxvNp(aD&B>HdM&oW;XJ{`e znXibmXF$Vwh!ZZUt$eldoQDjC1`Hcd-G2$D=iVD&j1rDqe-jdo0qRmLM+bcT}tZ!v7CZbz2 zRm{?QPkfNWC-5j>a)aO&PDN9quK8vJ=Tm)AD$iWP+wIh*!>-DOb1&3>HH0?02tsH# z=ejyRO_NMEIt~7ZxFWEzK3q95S}?B2old}l2?APN@sW+Dr#O8o)HMoI6hfr*JBFr$ zS!qV6fOpXpZrkW;#=@^XH7tD(exnN)s&X#g zdOo`P#Bzvd%v=E8$S=zGE7*5W)mJ6SdIZ~~gn?aqtJqDp3%AKi{xP_*cQmY$Uk}P^ zf+&14gQqFY8LQ}PqpfWmkG;sW!v4x4>ZD*@NTlL;0Pz*t>CXe2O3>(&EN`89n=@DEo3)mMb*p9VI!#E!|qijVf%8i z+QxeuIWta|d8|Yek}V&M(057UR$T6wK$L%Pe2Yc)xR;l4GOfpRXZ#+lrdxv$R6;s65cB(qaf8mlDWe zNkztZd3sAis58VOs4B(KCwvl(so`a6i9zPRuaI9Gtr`+Nr-dEYX)*ts*PvKL^{c&%uH+%Hjo3G-3x#b0>Ie4cB zUf?@k>5nl*EjC>I+?y-@H+q}-Q_OI)^7~sS{$B)p$Jl$@+1}b2toHe{qqgJc5*(Xb z`Ac`;U$QhI-|FE(l|@&Dh8`ru)&jDJo8!S~W3V>tj|You{4*>9ABYyS3cR&_GhrDo zN?g+G&f|N^wZFuKHDfeSz--1EZrb)*N=%$sS9tkm(^B)t>kUl*>y@1B7G$cM!b>`E zm2KtvCDfj!DNgNZ8C>ct16K@tiO~S07~yTE%~&L9p9yCAfY!2=kXpc@Imj7{GWf`1sMfiI zG#bcJmdLK9#rtd6L~WS>*Qz?~gp67dL!{;9R4+8OcqTaPRNXPsucYY#zG@mJhstHU zsPAxj+OOn#J6SccjRRHY6!#q+dmd+6eCcr2G&f3~Tra_4bHZGzcyfn$?%9g=S_@RfmqD$So9_$>nPgu(m!~h~diL#EUFJhFkxHqu~a< zSqS%s!4kDXaTDKK`*DgHLEuC#MyuPf!`6gJ)R_|qPUM0~z=`}?C^%6c;Fl@&I3q}m z@LOSrvRaXby9)GPUUj628IW-^^gb$z{=PH4`#V|qJvp}FZ^IupI`=u@vX(= z;=~H?2&9D&HlAXu}V~w-OlS6z(Dn;5!P^Kdn zQxN*3W+Qcp7uJ^?+`W)j(&+MQtTVPWzrHEc*ia$l314s^ zx@cA;1tL501Y%YG8>!{bMQ-ZqGuyo`{=;79w}^&JO|J+$0HvDHrA2lg8yez-f>Jdf zV!|@*qGNBSheW8-70p?bv7k+qp};#G9`n{%SaSo~JSq{p@JkL8ifoJ40-(0+iAKlv zYs$7o`*gd*vZ$9iT+gtrvcWzKNj~dwBY%Q(zv?)?NKNM7p9`zm*4;xSmeyMej)Jw1 z6SI9tYy4DapSdV^ABll+(utxdhSb$3b(GUMxsqb7xNSd6j5Z4&4{Jl#HYtIUJ7%PA z{@&~G0;$b-HAKc}{iA4nLrfe%lx70>Vx;vVkE=*3yhgHb9`F^6JA5*m{2K^jugoI} zjvz9LE=2|dWjoj@ijP{VFcs}aC~o^zGMh1L4<{^5k#}U39PwTvfp+E~97-6LT7f#5FfX{?$dw6&5aP&xsH8&-KkwuV259uY)ACCu{SFgl7v;6-&yl~wW%a0mPncf3m~x zcWJ|q4+Yb5rWwuhA2R#(1HbLaA||!%uri2pQ!7p}Ml$#lho?&X_AR*#s_ISbkgNu3 zeHFrQ(c+{q#x=dy*`XF4_Fw?HoTkQCtf8>BV-NO79TE0$dA-j`0h1frJvim^dY|u> zv=OZhP|;ZQq(JRePTx%lS4jt|kRVGFGb_Sa5j-vH05K%t^fw@(u5wfOr^QkAgpy} zCwoYke7P7L7|O-}$)E0Bu&fR&CyKVs=pwDmjy8;9hac+Fiw{jmBnfF$lix{F=&1OW zHbP7**Z79R!)Z>ZP-R2%98##8To+u=L?&KYx&Rbuu=h)vV@aP<*6-;ARJ~o9Tc1-4 z)-8(yZjCxiDlCKV3uHg>cDSu~i&H<*dHGeTz-Oom$;@(CIB1=(FqI4l2aVF1kJ^ql z|A{Vsn~wytLbBA1HoCUuJ#>ghiIEV|sE?|uOAuq#tOT6bKvN}=N>~vY&txI_Dfx;w z$WdKc{QhK{T1Jz|%;j&%WgznwdfDbQ)p)5ct)uzaYsZxUacC;x$jrKw}sDvyPkXz7(!L+G> z{N3Cdh*E+2N0X)k|4apg)tW{5o~AURi(5o+ST0#>H!AQL4$pG}4$FNcg2VFrbZ}UG z(lnCfQ4Lnme%O0(4`T(*Gg6*8;{Ov>Z%Qpzp~8g5Jzzk{GzY<{CMFlHHSV+*N4`u2 z3*{y>6=xb)D1QSbEL0yY=byA(7ctteia)h95O9XVb0RE6bdc`ukB0rtlY`Lm<_sOV z6h}tKV8@mn4PP|SFjkuHk#-h_vM#*}&+pwDTbFKG4-a%TC>3HAa9zcA`Mh+pq$Wlgdy@|nmlO6c@0FJmvTT}mUJGI- zqyM!I+(s@p{(AW0&bU7sYh>4q?=!n5nxOceP-bg`^>N=xX1ixjK{ES=*#wY?+gb0t zWBtuL2YFI6P84I4Et=cWLc8Oe?-JDrZqtKa?3LYH$%D}~!}ayWyZh^>by%|&CqoQ- zp&8Al>6!4~;8?8Zdan`QsZlO3vgN=+NFg2>vs?44@_z=JcZE_>FP2q-!^69iMKk)dyEH+^<8_MB&Y)xXn-ryi zMjt1pJgV4qvMgPGKee+9hnyUq6xY`IW{PU@iQ)R7H_QH+0F(-Gr}K9PHcueCZ#>*y zAM9|jNyU@=Sc-Oo`{ebRnL~%G=VRdHb?ivd2dr|8Bx5QSFGt6?sXm}{i=>Cop0nG{ z6fyWx2D~gE*s~NnFGjbo9Uh)KH5d)7t*KgeZ3xlk^Q4`>&l+6tf4K`=={BeBJGYT>Qrl0SUW?+iOTgEET?c+3ouYP5_SgF(kup7xJ_9J@q}dUh zk8a=ufkRHAAO5lf&rc^GJXeeBn#pyL0DB18Q^A36B^5Q^MP*mN@5nu-kL6bf_pFbP zx3-UzWScyp;m2K75rHKGguV#_CEB?l`=|A$JmKfjf~346?-#&pyD7s85{`;72dPz> zPy3LkQb4U!^c}G0uEAh?b#1tT3C`nY*hspBhk7FgPiV7@R+-QUc=cK7q{b$Tb&;Rx zri@6OlQ$(Nd0{#Lp+&-uw{v0$#xIQ<+GxqkN)M^t)JbW?jCwl_6mRSkNEB6fB3nTV z?oEoCI@Zd!FsZZ-rloyG9{|U)hbG zzG~b|vy0VP7)WmGt?<0w%~agh%0iuat?eT=Hy}tDW}shDR^_-9k?clFj7sM&@VfeLW$XTKGuyIN2C{3 zNin6k6r+r>yt5HCmQKIif<&y~Lz6u^9lH%iKJblM5=NgAc8x_zo$sbm#rr=tZ!R5f3gVb^^rSVc+(8j5I&C`Ehj`cXr7U#IX_)MuS1Z9alx7^3hY>SBEx0km?|hA}hCsPM)gL!pIwqkCBTYArpZN(^Fe_c`x+px=7-Pqp#A7=Jv zv>TQ6wb9l#4nSOFYht~gMbw@CZOy0VmpgWC#5eh*3jy7X&#BJaAC3dup7jM@e8e87wXj`#U(^EOHN0GVuBi zJjCp;szOkEwBeG!()&Y_&de*41S4JLQ5)2Q_4L;PDbqDkw#&gkU-~n z02LrR{9Cm%j8Mo+WKQ?#4hf=FGhD-kCF2oFE#L{JMkLUJy|go0-%aG8{4lGMfGN%) zPWC0h9c8~(2JU+0a3P54@xvXs!9Vs_kD(&9D$k@Nm?Zc1R{lVN!OKFU+V{nzx!&y2Yj3`a$MT@(3gGiWScpqur9WPg)dBu&-U14}-O|5SY zT?Bi_*dyNA-r5znI$4}zdzbq-{LcR<)5}0AD8JOx&XyJ)*KnkV!FL_( zaYvF{XZtsL5)*oV?FU!S@EBbF;Vk|k6E>Uy)6xDH!$neVDDIP*@;C=G-PP34p1t-CM6bBO=8ic z_#@jSa`CfAOb?Yc5J@HKR4XYSjj&2_sx1|#noCSJ9(fRTs+BvGK{Mu5(|HD7KJ{qu zPehnD@hc-Sg^co?Dv*!!IS3DU6X|u>Yyls#kKU|o0Y2Wz+m3IY8gKN^7(+^c4Z+X* z=im*IYZ`zQ;t7sJE*LIzPc8v(Y51mPtum^#0Li|2AXrk;O^6%=pFoS&JSIP38IO>8 zHGA^Tr^QEPJQ2pdMDBW@2u~2@>MEm!CC0iQQ+`?{{u`kR>~xhuS%1L5KW>nFCUw}zS@~seoSkD z8@|lY#k~{xFM?YORmnQ^<%}LB1A7hsz>jy}3eVP32|S7_os7-VR$VF}-(B>Btv26s zNKGCXew$NwG+trU)&C~Ca=qmXu`4%o5FQl&;icI%`SMPSufqe8YLy(xSmW?6F4?8& zT86gt01QV=r48_gNI}ZMuVnrar4iJtIzEXY-b*`hWn>5uCGDuap{PBBhm46tlK8g_ zRy-xauXG6-XAx2>l$knzhUSD*`X}ekd@zHjd+uL8-O&agjDeE4T+8m3_?lpS1KLV# z;1<|GeD8`ge7NIngK`VnI#gR-Hcr!dteh8?8m-_*!y^yw@9muzXXS(0#2<|qQusv| zKGK1w1)H&`F2;w%+&7k(^(gk|QRgtr%W#0sz#IL~O}YFA6pztH9%rXYYBk*KUMn)n zkeOQe`V`@pz}P&fpFG`>+R_FM%mH&7ItM+=Z6j2Y}uvjjG`>5Kx4@xPi$##_TVt-gH*T;S+8Is@Y zgeZz5*gNV>Vn`(tB2UAYq9PyrT%&`sazRU-N!=Y25$=xP4hZijmqwOjLH3pW6X_vEZb;!u*=-Y zL@Sl|0Bjd(1k1f~4Pb_e%eQ;pOpwq6X$9tX;Y|vxy1!UTztJn<4>F&tqZ92?o>Dq7 zxG!^x=W_oAe>?QLnlQF5eB3{0mTM@oJo(2ed)v2Y(KImc;4IfsEUO`OvFFU#W8>8u zabwM$w;e_FFE1oDR!l+kKR`wF|CE14Kxw4hq5KzBQ=y1O-kn&vlGQfhvj9VRd97+# z@V?ButS8QJ9#Pur1IBI)5m|*XkbHGkTYUKkh4p>g$4Yrz81j#~Z`7$RtHAz5qg3iM z3t|{ZrG2((oJgtWBQS!2R4%`vbKmNsRI+eb`7eSQNL9%i_peTlI#{*Z3jC{o)BJ^} z2cxr&O>+u2#M!H~QzoRe*zP@M(s1N-*T8k}p@sRnK_=Wi(v*AyTcP_V7Lo;Y3WYoN zBC2bd>_yauJtUf}q09(JZ1_ft>f#GHSPE4Q*a+=KP^h5Zfbf`>QU8mMvpE>X_w^R$ zhXy&}-y!zt%cYuR->^mS;pbYd%{Q9#|ClH1`2x^NLgY(spTZfEhzwY8(G z2QDqgPdDfA`pUYk|LA=?vMqRcw;}b~cZ3>K9y~p3=U~IvtfPI6z1$5}ye|Dfw2PYd zpdT=8`>>->?TGDjR?K2sN@b6tR{P{^%0gjwGgX-xIhRA-zRLzmlejuR_N(p5y2HfTQS9l?AZ7I;q2_dNFOaf32(BQ- z^57ee?TN95$!%z1O8Z7;x?~&qMh9ZzPh}%sKqbafE5DQ5s*OWs^qxKBgP== z@Fa%oT319q6=Q`G>C2cDn2X7?H8MEhIHs7$U)806b41@mdFLtp!~_KAjYj1N1aTTe zjHy&a+kk>ZB4+e1%CKOTnl#z^>Gr4w9NZ6k5AIxjd z!+3TSg#VbjlKdKq&-VV0HHdM4)tbAopQ5kPQcR2h@$jK|NX1v*K<*^MWw<>y1yI4z zXkJRvo~ZrnM28Lp#Qy|hj$I3>SZlrCeCf%;T}`_G@q;ChDD3SO;a{x6ybclSrb{^rR+Xc>*L z`TSEMiEYb1hA$du$N)wg?g=8}p5^oTWZJB=uqw-+&xYf@o5ISUElcFFss_E<{NrU^ zcqWm8(GYvxE7#WY>jLehIRQ8z&uEbdD$N{b%DPlHyNHep7uCO`aFJFp|^i|uVhSTS#qOhgp*@L zGQ!tK?;I@djQgXpCeAiP=}hw7)u~)9q131tdiksk*2jG(`J~rAfqee9O9nK!{q)W| z#>^;BV3efsk9dLHD$@HydwjAKaw=dRkWhaUQOLMvxW2wv1U}D87|n2Uj<^f_N;jbM zt4BXfJiS@=&z$%mvrWg)R7KmDSZL4~M@7QdUAPcnxVbhsvnn5TNDipmtv-VXX5j>p zF&{p=+20tRJU$-c>(RaawLj)m*%%L!WT+6aCfUkDD~s)Ror5#RK!-Da%6iMbWth);GYhjbGlvedt^-wA8IMj6SolL0M}M6~ zKYs_)tB~6l6`bS}HjB!g_y?cO;Mu_#==cjy_eX=Z!^2aj2BX2b(SG=R1~(_!P)sS& zlB!!c7eCU40ri%&{lPXGLmCUQJ=|(YXibKfl-}^89e574Dt1ox*ZU(8GjF&(`1|!H zU>VJ{adYNSwT>7TkO60j9uHn?iw5z4DqbK+BxSNwUPw2N-KmaXh46Y4!n~gz#(@eOlR6V-k z)ynqtkkR(wI*M%GYssMx7H!!fC+n3V9- zod~jNWvd0zSrjMs+eqJs2B0mhlfPYG|k@Y#%z3|`rZV0|ltF%jJ!fy*&piCJS0k6u1b1&3>HH0?02tsk12_oB6 zq#CSr_%uy2+2}OHTo-0F?{a9gU|f+qoe1u<^tF%D)axqLH40M{LZtLNc5|SGa~9bz zgLlysZrkW;da@Y-X#$_!q|jD_sz!llI8BWrKo%6WGN@&-s(z2=!KtNkj%VyBD>cQE zo{c0ticsH1h^L~}ZL&Up3~uZl4Xe-B2g@o!v0uZclXmTTh2P5HX^xfnwb9l#j@ese zah?5@#T7vXY7AFf$`_0kdGRTJBU?$u+V?d|m+j^05qlA&5zTHM=d>GSrrITL+=3JZ zo8siV7yR|+UYC)yV{TzptcYf?2B%M(Dd_aVt7^Oeo5~&&f$SZ)bb7MMt^%jgEH+l}03~Rlyn-On6tM z?qz#aVV~mc-h5mm}6wUh3ns6x4OBsD9xc2jdoVLcOd(5ndnW?dsgZWR;*h}(<>LGrLc=$;kO%38A>Avt zLj+zEyP`Zr?y5L{c&%uH+%Hjo3G-I(lV}n&gez(P7l1mcWiiC zJy?~ZXz&y>+^qcm$*KPr!QL@;-FCLOb_T2LHE?q$Q3zu!8(53b^T+EAO#kbZk=OvtC1xLcNe8a7XddK_mcgaYGAKis-64$I zmtussgzTy4B7#`xK26q@1TQfRvkngydT<32;7S$(-$XjY>T z2h+IC01pD8nOwg10HcIv@;7n41{rSs6OM)()J%i>!(fS8p?G%XYUXLCrwE*4Mi4lW zi_z*<9Wdt%IFSn`0Vncnq2NS)fZwMK5|ci9!f%Bg%4$Wb#!O^j)?o!2JC)1P8W8(!vc zL&LVp2Kz9iYEq9I`4gP`RmbsFYBK-+Tv*Mv?j9nsv|dzj6s&!snC(Ma;}=$CZ2!)) z_OH3qX)em$M`B<*b&6t0U42qVIgOJmDf*3GGs_e{9@d7eZBhaycg!f7zXH8RIu_jP z@Cd2Rcr`@EX#JyTd_xSCNbu1}>s20CkyLn%WZyinTY`B5LF|=zB*76xCebD15v&4; zHRA;A6vam^RhWu)BNVs&GMUYowTDw$G9+G2B+xEDl&ak#8r{&rM#_39b}t(Dxh^ZO zCWbg0PY;WPXa%i^04s5IC~rC;Zw5XSuI=FkxswME1=L~c>tvs2Hv&7c2jg#M_G|wd z$0^U}LT(kZX`Ol6RFo2nKbO>kRPcxov3mofw7LBJ8P4fI@m(WR)Q#||;XFFhOWByE z5PNQl1cQse%-DzTFc z{g%wlJwnv7wo|L&!FwIv0P$R?O1?D)IiR(DZ2e$lBhddaLA+{S?&ZLf& zOefNzB9D6b6Nh(8{A@;c;1v^`Ub4G+R(Rj`r$ZkMB$uPrcns0pGjfDPmSKYJ+|Xm~7of`e=tRIl4(>vtBnxT$~C^>@QRwRoa8yA zP&c`oa6J>5ctsZ+rzk+-4X@1x|6epSN}pI(@#!Q~y=a+RA6E-jE(-&0{X9!9Eb+t( zWZUs}xUF}KQ`^yb`BkaFXQ*n)%yd{hSZqsyj7X|IUtpPyd%o~V$_9Yijy3^`u78_d zgshee%|H&r)Xq3G;*Sqjys#>yC z!iq?F8l}4@<}2PXN40117fkl6WjKjMCx1&W1ev+e)S#-u&1nkq@{?Mp5e8E9@)Kk& ze?*C=xvw;G7Glw}j3q_Ph)YbAqH5U={1f@Hhkh-xl^B-Au^nWhH(3@>R2zwOX3|1L z5rtJdg4h7KObH7Se7RbO#w5dtPfT43r5ZyCV-CJTS%|2*G@9M{rFWBiAX9CsQ`;mh zvo_?xR(I+Ls4_V2gaw-AKOSLY8TZH^^LPiR%WF zcC<-B^Ws@Oc?rAEMn^1>ThNBVw8?<{-9-=URoY2KB zB0MaYthE~+c<5oUJXwJZ56gWd!o%|Wba+^OlFuixMQ_JLpni1z6IFXkEmonzgvL!n zc9DzL8h5n-9dsGO1RmkbRIpHPLQ{06!R7Kd6eZ_G@jFPwEr<{YDm;e)B-3CfL)}y~ zk?!t~hW*WxgCqq`KSBh~kB*Uh%Z`RG8rVoHL27BfaN1cI%G&lSJim8qY;C(`M=Vsr zeO)*~u731_UiZqiwfwkuwQh2)i=)BX>67f@#p-Z#ZFq996P0UCe!etURIWHkyHk|D z;rUGS6&QG53bo;L$h~>~!r2NwjGV+QQYo*Ut!@_NZfKni`h70B&y~$cO^kY&xP-8O zd9S>LkY&@fb4PBUSy%9D9k`8LZu~Xz#hr0~G}g$j8J%Z#O>QB_OG=ro4c5nfjm&Dw z4q5ak$a0_`fUg@71Zk*m&Jo`T*$oWh+K@w6*tapC- zL!LM?(9v>V#SuvaP(Ve{4(E*7=3eakSP;Yi)tnVCa}Sx1UKV3ji^8B?iv zMmok#^#Pq*Bt3j~8|3cD?1wL9z{~QH9jRiD``Y2*sZ)c|z}lLs*h)9}7+z>6>9Zad zd^v+Y0ZcV*O-s2k68GBI?02!?m%Ff)Zgbkca~l~awf*GjwV05x|GEzRKsrV3ob0dn zM_7JkCf&<@5Dr&rr%5Hbw zk$a3F%dZaZSsx#7Z67JA6nR3!kAq+$A<6)uZ^A%{b}rcdX}u{=_<6J-r;3<;w)-5s zwwp4n$ooR6Rhm!xke3DbT=!2XA0GGe7p4OcS|ptOcFsJ!G;VUE$Vgj~m(oM3H+2#!si0>jboi2vlcr{M zZ&K7G=fA~V6Lr1Xr z%u;~2Qmd&>QzbV6`vb!cY=(-p*SN(~a;{)yy}d(12d`|OIlF&NvW>GkCHd8YlR=p9 z7MezYl9xXnbvHy36t5G>s*Q$fta&#NzHC~7v}Ri^f`pgP62qAq{gk42r8U_Ev6G~T zIl0(^ayli?X^$UfPDiy~0dJ1G1J^`^MylyU+IK$IB?iwGiYdjV7-fuQk{C6a&ZJpG zAJYul6uN9uOQ+*C@Qqm#MxPRPjYB#~DW%OF3pyz+GG$a!+Eo@bQrgsnxe7BLD`V3s z@1It318?e(-1tz&Hht>ol-%ICFqqM#@ z+SPQ9kP;a$5rXWo&=e*Jz#gCuma_sqPjIQ#rbN z;7R$C=MM4m4%z;2Mqv2D>DMyR*E7rzIF$h>U20b(Z%w)CCUfd)=*juZFw)DV|P$ z*I=-{x;@&Wqb~lg!`tJndyxhxDT;Km4k(?v`9*BhtWKqve6vv)at7}svb!DDWmO}N z!|krBtDGtZ#L7})%o?2#J{5x}x?h$gLV)bLEkU-~n02MGh{9Cm%j8Mo+WKQ?# z4hf=FV_d_9C0!V$7Vr*}+o4875^m!xyqm~D`C(RsF#aRSImD;*6W|WsG*ctAC*S;P zk&QVq{^1VX;2-;|$54@4m1oisOp|+iD}SKC;AKriLKnP!w|A{S?ysKO8g2B)y}91( z(Q9wMif^=K%C-6aFMK`-3vnr|^v5f*KEQwO%@zNSaO?T9>#1B>Q|o&}7s1{!wuyJP zw{`}r{a)vBGY(O$_LuI!zhv&>!v}*rvar1dmwG z%h8^*kGXIj--Ejc|GN+WdkFsvAMk<%x^0xYR4GP4C~dK;oI=9NE^0Z#!eblHo?Q&z zdGMlMXMUbpM20`?!~Y&KiW$C=`7nyO{FHNFRv5x{n|BXJ;~NK~jnm^kn`hj+sPV7z zCptM+<=czf_(8FG)7%2HA;D+7g0Fsua#_7SCq!t=JrI$>?tFNK`o%S5ErgaCZG@Re zX(#+<`3WX(J`lxJuY~t@yyHfBhKZLai&Jdxav$g4`5$F^8At`?mwMXS61!F3U;Dw; zGdyONe>jVO$dnu(*YLH6!FL_(abJ^iWWqE%8`ea8jYcOs1dui9xouhCXokW)LYzh3 z#tm0Rn~6U#(atE*n!=bAErv08rN7}-wp!ZnOXUfiix?obBJ zm{U#X8Te|PM~{CZ!nBE>#SgbLN=VRr{#7v&Jm6Jut@*VqZ}9wC42Ris`*mxBO`P$1 z)_`OgioK`&FWcUl9C8~9o8g^aolL_D=jKJ54F9juDnw>VTI}={Gzu+CiTpmM#zA# z*+*|ywg4aRwO|TL6objOc$aMybrU*<8Vvc z0>npA2wbUX8&sa`jz^-r{TGcpCXgse^Y8us2bgHAlu008l_n1{=Fe1&N^JW2YG-2k zF|7q|_%cHmS5M@>2yQJ@CF{_aGfG+p_8R_yAMd~wo~@-4cobDS8JnZ6x>P{!hko4l z?0lF3w%UBlAvJkm_-zi}(R#&s#RE&dg7KWsaDC6j5QAL z;*wpO>ezLxN1H4C)%gYs6CBA5-ViBBIs28&KcX~(dR5|c_6lQy2`}xym60Jtl(eJv zhNAWi9x^5pN#fr!Sn-qu4+4V5S%lOIWv0%bp*i8mDKpZb;DZ@F-E;r)^^P|9KvHUT ztm^{aE%7zM`UbR>*mimZn`}>9RQ_SBCI%za~tS&u?s*)$4^@J9c0Q!c*&#bdOQ$JwcpS`9Z7$q>-g!q=w=zXZnSLH*?E zj?|VmXwc;M-`${c^x(pS7bw~k>k9&wn0)sy7cY{)N|qm|kNCLemeKfFV^(m7xUfFy zU*|6%O4&PRadENuS@JMxa7lj1sq9@{%_Ob0#;Z)kW|aW}KIC6zsxH0&d!i*Tg{kA- zJ$13*GA~}=KRXy5MXK|nx+cwR`G9wqL>PWqVkrM8H&>^qrsA~*;Kda|ux`VP?O-cG z2mh`9c(J|#EnZC2xeq{waSYJshTtsf{Nwx zOYN)fdax8x58kL)E`$51l9~?=DLKq`?(K{HQL$Vf`=KOAeyIK)p9QY{vh+YIXclTX`xx)Rr@Nxf? zS+1c-^5h?@${oso zQ8g8cSmfP^l`C0o6TWZ}ckaHsfx-JS@2Z|S!+At$s}DH#lgtBbZ4WFX0m)ZqwZ)fz zP*~r$d#selg(3f#`$nDGvW(b2Zg2=h5RgjyY}2^09B>X0kjmvZbnZ)GlS&p2EB{3> z0jVlkR+ zrQ{pf3f(uckgT6mDBQ6XQC*9%6=9!|p-e2KhruQ*_(qHB;tM!f3RTV52y8`AXkaS+ z2Pt$0@qL|zxuL-feLpA8*}CuL&u~h4)LJIZP%@lC%&M*gyAd-p^z~=8bXp^%k$+PH z=1DEmqUwLqv3~|1)J&g#ayflkGlkbz)@|KK@4Jz0vBSFysn@=e;*@7g&qk80*n`b>f{xjqe&exf`r>UAlp27B$rVqcnz;(?VZuxrbx6X<;ZoT+COxV0m9TUnx7-xn~YO7Nw2)>QNibOXNi9k-0ei zHQVczM(dWl&~_Z!dejkaDt{qYdLH2p^)DsuT>nf5t_k;BwNOEC)cCCVImZYca_b42 z=4b52$`}Ig>BIyzDV#O%XenN0JOhyPI&~ zn-?^~!6UX{1krpBWt~Ty%N4C@{XpY+UkSTIjO%ZQN20HD9J8lv{%5BM$`ovFv379qUOiCW0_!bDu~sV;m8kSgHUYh$=kR_n?LNZ6#0FB*k1&g}9QVT)#HWEnh<1o|q9$DweghZ?kpuhE<{XhQ;;dlN#A?YwX_&YO;F zJRsG6>`oV-9T!?q0GmRoTx7A?F<+S0@n~+*7@T&FJ=qfpS*xyx_H%yeg)~mX)D@c@ zEB7lAbxV=K{9cqY*1GfIGl3B}m-|ft^Z1WP*jrkA&Hj8BA}C8xuzf) z0Zrv+i*jTH5Kv!Aa7`JfubQtzXh4o=1l-*p4f~rX2j=RSG216%yTPjh64#b(3|}?UE_PFkd!e1dk-^K!RI#zO)|=<=rLAnOS%&NT z=$BL>22q?F$^v}*$2{)QS#}&9Ytx8_37_A;yjRkQvn;tG)}mFYDp;@iCYH1NaO?Fx zim!s+Iau5o_eWz*Y;8uyndwR>B%YH?XKk=P?rWq|(`Cp?7n<@-Zh`PWx>UG=+fMJi zW6XpsSL&TTwqlQzQIB?c+#t-svdXtfj3o?*TZi$MU-f>T@Z*7h6jkGdVJv_XdHog$$ZlwVQOKfU7h0l`ZtBnFYMR8Jen^SnA z5b`9$GOix~X|kH(iQ)R7H_QG>h0BCAy(|1?XHQ1rVZzm@8Y*^WvJ#hR(t*lGS!29PM2n#cHc8fHQMz5(_B;!9Uw3 z&pK&T^5G8qp#EEAe+>n^+BUV? zNnYh?BkB%&Zoh5~r>_pjXR#_5U!_f89aqnhX@j3w}jfSV-E zli0{%GnW3&MPGndcl|Ha>4j2q*US$m_#8q0W2p2CD;S5yOh7@aCZR2%NgCQh8N7Km zoNS~#v_?*w>;hvGj4T8f< zaL%S7(O`YUCrNU}M#~`LrY@^rm-ChdKZ@Gdh=y#`hrPpxNGfm@mKlK&3JDUrn}A3T ztwHm#*<_d9M^c7uW0n~)xsC>L=&%zF0^UyILk(6L0e0bxGJ>E;%O<-|^BmJt%Ym2? zcky}StEj}UV9T7k`yyP4z$^U-BDcyy`%Q2|ZzZVE-Zp-dIWxO-(Y5zT;Wsll=+;$S z8*Odl)Tc$J5%xD0Q6B|2qA{}Usri30mDa_V_(f={$isH>{oy+iJP*xY9_z9fq>hs9 z1Vg(1LEt{j5>Z=W#@Wfu3atE?q=Oz&8{KU$(2psBrktooE3psaO~-%iR(K za^ypmhlsKOlAa%^BGEina7$z!@zzc_d{YDBJY^xn8w$ns(EvAOgM7dcvX1c}kg|rK zjPV|RB2+1V!Ev3>$PGQ3N>YU}OvoKioC__Yg}=_=uwz)QhV>!~SRVFwaNySw%+h*$ zD}M;alVzH$HgPfRUF(ngtEaX`8~t%_t~Y!1+MBQ95(t_M48Q7*4ETNxuZf4MQPcvS zY=o4R-#$4UKj7K$OX_c5+n)`WDu6e+droD|W$NA^WA=`oGgrv5<F;YTS_z>A1DZux<<&JVkm2~p zK!tP1D3Oq=wZxbpYQUPDnS{sDbyx}_Q~cV^F?JxvgblO?jII1usJ{BB-_7Eg&`?u# zLbFlG4clgfT*998j1dCA3{~89i}?(frS&4LUZ7>JGurB}o$T+7kKj4Y#+_a{x^o9} zR;vWek~F!O^j!Frjx|d&t-bWKY6cr4Os$q zty;s*R<{5d21Mj~DS(LlSRfElf1lrC3<7KNT&@F=2ZGLEwIEe3rZFwsOaP66%Ef3+ z>)bwv<=Sb^Vfne>b69X1mwLe^|@)CM!X?knQIu zg3ddsB6MVtc*lqZKNBSpl&Cp(6eej`KYJQHAS#s}Xikf!QtJIR`?J*=<7%#7n?RSDZ78`Qvn)UqBf z^2fWj+>Ybx%XW`-^L+SBPyx2?7b0!6T^Vo`s&zV-?K@hdkCm8P8u@D!rnOV&4+a#- zC-sw)xVAc8YuRC>WZ|PhJ;d4;B_?vlhSY7_AGbfeX){|5%P>~;D3@J|D>&7|cA3Ui z9u?9d)h7?^l2l$7A8lol7=PkWNz;RoXkmb0uO~WEsajL4_b~izyT@fSV(nQ(Gr6#( zz{f5G#%q?JLDhZ_iC)OSVawj#Ku?|72*?7NyiB!pX?YS7baZXh89xh^s=K zy~pK+yzAgd?`oF|c^V_FuabSec>O4@DaRjsy@Oy?7U*n zC5sZ9#`CA3YaZ)@uFE)HuKGQDrQ1(n4L*jA=k51-#?VM$=qfbV3v7%y79@W74g2dD z-aE(0hL5*wG||@q>(dr~ze^i;9GDt&#j!q#ah7$L6{`X6Ixv zll$2{{N(VMo$rpck*p0t*?PXdzr4UuwnApmS26M|HO{6-!y^yw@9mv`Q2ZyG_@g;I zi@q-)o0TL~RxE2pFsw+VApyS(aBks&VQku?Ir-_eT>PyFK8PoNVT+)&Z8*5A}2 zX~0wSX5%zh(eGaa=XZJLv~ z;7wH^mU_`f)wcA848bTN(jgf2O;uD0ssS6r7TiNrAeI_f5s6KsK)jS+?~7_wn-#x5 z*>x6hA~JHrUsH=fCMYEJrK)6ek`lZ8qSjluy|mPLBXSaXdccdrf4HEOOe-lB!QT%0O56TGOABYgITCym~lN29x;feWh3Jh%ID0nSNHs&G7d`ohG$bg#go;^%{dchG_&& zbu+nStyv=u47#LX>M%)HfiDt4K)DA^A(;dM%3nYT0o6Cl=_5gcmMFc38{xAVJUiSv zL__HA{%F|WJUKAeE!#l{;bS*%j!OJnHZ*+EKvRcpJ}ul?c!8{1-wwC+Zi%c~w=9Rp zw;B)(F$wr@T{upydoTq*ycyh24Q}>+wdTFVs zSi^IU7K*imM^jTeygPx)@D1SJJb&|R1z!VBVG;ko;^5aia5J^9l}$xWWU9r6cCMSn zz8@y85bR&xE3Xh_+4N!@dwt+mYN_$Jvln;9{n1z>Ib%N54EQtC6GolD*b?v)N^EVg zKJIHIR@dpvDeMgBOq>Lz4MOsH}4$eDJ?inWKFeXZbDP^nD)qLDUVgM z5_Y@g7U~$pp1EJVyT5+g4OULk9h{W^Hj%Zo>0pK7!x0G zMi@LevpPPDukCd%?M}~+;M=qCd}%JLFV6R$8}#he@_h82fbPLPkQtzWSz{-7Z5p~KsMfZ zxV=8u;mA?;0^;cCY?L86Qfr%Hez7Plwo;=&_lgJT!u%@yApU&{}4e;|FaIA1&sGPt0Hyfk<4O$=iA&)t~0h!%uc`bPB%O+x!+9N8(Lb$R~QJtT@q z3^-XiUEGjE)*`NKx@uYlkT`*SDubKioo!>#-#iD{rZ^Ef;vT3FePuK5Ow%>VG`$6$ z+j|ZX)3mY_B9`eJx^M%fz9qx-Xeicad(^N?$rUL)rn@j5UA6ag%|t)=WnvZ88Ku

8CEE9fomKMF#n4$!ID6O^9M@)*+ zKz$F%HNb|P6uXmveyP4*b9M@@q58B`u({t*R`zy35$~`| zw*+J$B>@eaGeK-3J!S=q^W4#j(R$#|iL>rTV~J!uJL}q(FC+*~VLC4sM++@?^N)zD zsTJNPsR0Z5CUNA*j+}w+2_Gaby|-~SjVy-`*nz7NBm1CnxrlD{mNztWQ!Flt^TlEwdywu4m$WlgYmOaCPY{%RR&+FYB*^a5r zq^&%O$kfX)cGn&y;14qE>(Y%%`Pyh}8^^LPGR?83%_8cp|3r{%lK#`ITNWlx4nVxENo%JK~w+CKKnEFr`qkyEU#3kyA`&l z+FL(a4RZOD{ha--!C-rJd$dJIGyH=63GUXNNYj&a!gzBKS694~6E}abA7DmXs+Ie{ z1NLxztPMBtg45$>gb8QtK}ZTo?65wmnid>vaaBF#M4m48)D zs4&>v87wXj`#X3BtBqNqtkG29XFKprb-Mfcx!g+W4h7?_xYQMI1}LB14?Mp>DV`-2 zmkDLQ`$EFXbiBIpz$w6V@TY1=7M*~XBBOw466>m z51M>cnKQI7d5E!*aTFTnXIGV|OK|RRLT>!Q%KKjG<(bK1Vu(oLhdOZ7JJ?r`og#NC z&!8ii1NZh;{y>1U%j%)=%#sd;cjWf2^~e3yQ(L2r{~8%%~$ci-12_P9Q<|& z>fth2>5o@r>d$}f%@zNSZsYltjz@51MXk^BTm*Z^*y7#U-r5}JA?I8xbeb)FQZeyLT&#{XqhSg&-J?V_srG*5);~t zn+3ziGFW!@$oihmb&T&39K}a0=jD9L*~eTsk9SD#!T;`KcS-TTu4KYl<96VEJ^(NIY3n8$>MZ=0Pr%jhPp+!B z=Y$AoxjXtXSeys@PrtN=oEJ=rXF4yH&#!WyU^3=oN1O`j@a~RpxF`=T@u*{QZ0%j{ z<6t@Ovr8`lshoV#&%&0>#)TY|{@M?&p5f87{KHxNL#90LaSb1P7<|{k9``XRY9?=1 z>+rM}o!}5D)b(s zSgcclwei3OCzdeLR&ll$qh&IYcp8&fPc=nFW8yDOQwYV&5Hb5x=0hZsm_w{2W;7xx z!6CL(9Ac78(@vnaAe@*(tlXcBlM#oQ&Li;pr+a69BCNEDUfG8!yi{Qf3mR^x8GOR0 zdRlX5S-0THvF2tDE@Dm2fDKjlLOXV5xcQ#)zwm25eMrMTFc6%_$qRYqj*&51hxhP) z*JlEpe$8GTyudhFSF}erK)+;BOId2*g-6395AN^noqtgLC!6@A@gxv#(q(0I_tKGm zk&WJ+=fR*mWAZ1uKr-<+HGMmx`#4zU#}(-rDP%1@^g5=yYkXlrtIK-uZx!*-#Rtk=%1mo5V0-z zIq&?su@O!GF4hcK_bxE_5bGV@NGG652hTH=YnS=CA||}&8mn|nZ-R0fA)IP@g3bu`}O-Md(bA@HV% z(G%>Gpx)j6-shF*j+99rUz8@(aLk!3H;mgStg`nDS-;j(H503k=n!y|m)W#MI{TUTwb z?XpSmqP9!Ki}*mt*Z$-VwAH1ydlFG<&Lf3vG(utCK^G?AgB>^+YQdsh7##<5&k$l( zp%6$$m6aN4l0WZ#PRi9c0C$8Y@+bqPl9W7M>V-fwvf$+@E^fS;xid3)LL(KWjTBY( z?RPi29zD45;020_#QJzRKi_rw!KRjEObL8E_ibq{Rw_asba&&)QLS?IkH? zl5wi5RzGF2;1Mrg-#mw`tK*ry;cvd+wQ40GWZwM^6FKb1f$W49#fA9sr^6zSG zms&ceBGC-^gI~G&Qt_%Q9x6iAfiL)#i{PGVbPgrY8l<(rm6C$Yd9hfL^aNaTdF+Rh z!noi99}3{&8)+t?qN6;Vy${~z4gaMCP`OM3F*@c#%srvMT!7Zes?a~Eer*K%Kj{tj zKNjEc`3-r zLOgGe9iX~B_M6_RzEVfIKlv}JRzeVfe0!*J9b#r#L+~uDxRL1nZ3^C-`IhO#fz6Mo zw#0ub^bsTrc`B^7_VN!3&U^NKmFBo8^;QKW5nlCTsHEYJ#dP=&jedv2Lu`Oo!b`kZOXCh|< zkY^owuf}4JH2modZicVe=HVLqup6upU3!4%_%u|0xppG6?Yqi|Bf;5k+W?AfC6yhD zs=H>7kJ1D@`DHAnlkEo+4D)d*f2*!@j>OPHUu|i|!v%%{k?hB8|@`)1eSc;S6m9@r>nXbAuIo(-c~4Wa~r{x7;9f)EvKxEDywb#?B7W zAi2UX+R_cj)u|4o#ygTE0fjBdd)g$?i;3K06Glk^CQfE{ zkeeOI{KvJ#>GoN6iJQX1`7w`DvufNZEG6`s7qS_O5sPH#v3iwHbDBGX4s%s=p&b*j zn+I$vPJ|^)?qc^)$zOCJBJNc7;e|^G4ZuQqGI~yB`NHPY&785m7~#6q6_GVX7@&mo zG8FA@!i9s5BBFg&j{>^Ud=6!uC-e^wH$@sIHs+p5o=} zIPt?rqDqd;W;uDIt>kD-XwBH&wJOLhV5d@e>zL<4TTF2in+|pyVijAy;}V zE~turAswv$W>nEbZh1w3`ety-mrZQ~UBMo|I@uhJ^JW=+;B!$GR=HhSVT~vl$&aDz zEYIRtLm)R6t)0uQC&KFR{%HKf7vBKr6q04|JQC9L&`j>dV@amgTde0E%DL6x5=l*&aGn;o}1URkwn$&TkJvSXYj5D8hUu7~z>e(8lY zPQ%nidJ5_xu_>^b#SHkATm;VLepA3a{^Jq$Le(Dmqju7!F_IUKNCDYOJ*touRal4w z!}3yH8yKES#rs`RV7R^qEf`L8qdN`+%XMqPK>`pge_=ETu73&pIq5*~7o$+2TBs=A z5(kwSbSM|6H2{KB$P|9(BhjH;JS94mAB{wZ>N}>UeO%!Ey{Nu|TA@Nt)TSBHU69Mu znkG%xzwH7o&a#hyPIBd?T@pbj`Prfz836>;m&(h>EpjmXzaDSF$1`{q1CBmSFwqFO zyFVKCH%|_dRL9utU-#3k+Je`2V1>pvI6Bs*5f2lAVgK@8 zNk-1H4#6giz2=)}B!n+--Z@y@8TUtHO>Aw3&zb2;s3m+F4^*I*&e~vo+}B9Q z=t)h{P1(3wLV&4@7%Tvr#H()l+zwvmU$Eb+FNM1D>G11 zQ0_v2;pW=l%&L5rF)^lw56{AJngj3f+0FjO@Z|CF5MN8~?XUeYCrY?@V5Gu-2r8-8 z6<926H+g0f1rMa%s4T&1T8I;Vx&t@Rn9xf9%y0u=9yzNCgbz(j@P;lG;x1*=JHPxP zPvN#XwIWq$OB~1y!Po33@S3iF6HV{9r`ET&M)+u1nW`Qh-c1`{hz`KgfF1&>`7Bw! z+9<$N6er7%Ch4GJnuobLdsjf3^^W{x25gn0?DIpY>)IW@ly&9R}l=MlN15wc~Vx6QOwL_zO?>M}xJ)!&9dQqk*-N zP|<%k_=GdurQ@#+elCOOq}Y$PJeYb%{C40=y0AvPBW;hbjen5DJS@+a8WL8M(Iw?2 zytD()Cf38w$^LqOBtqrU+e1HJZ{(FhY(W<-XAV_+a$#W@a4_ew(51Gh56`q>UVj=9 zQyuFjWMjZ*3nbJIy(@tx^dTF(GxlLes?H z-$jOMPOU5aH2F7C|72X@Z_%5lBUdAEa@C-<{NpAo_;3bCLLFvf(BC` zSGJ(%#WfqMorPN@_3cJD+PglA`c_$hGwsc&TKMMX0(lFAe~9-#I@REOEsNTDsB8WmxY)O zt0}rHrRd;`T?$N4ZB|FRWl3nxl-Q5H)#e38imnbY<}7*>JLvl4XsY zHpx{ZNPY^Q8UapY95sUV6#O&-#LfJCI+zY^FOC2Q|Fsj&$5y6d;-pRa6t4E6xO(FqaC}j7EGB?yVY>d$l?aU;Uzd{ z!H{Jts`w;HuGnZ9MBLP6_3LupvfxKi`dZ2He3C#_4OSTecHxXNf}mKi$&8haGFmw>4B#lf&t_FpV!)PR&aRye0Kb{RLAR3n+GuMVJ9>*O zl(WCFxFU!h@^s4x{}FX*r#C9pA92OkGT;3RSKks5`=O zH~aP2yjh5qyTuFN@&A$c?Ln4ZRr!7I?R0wfOdz=$m^xZOSz2Ht)tLz)Fa;)QCJxc` zgk%On63BQn_fELdFK*vS&G=W;Sc(tC2L(Q&D9S^8MGym#w+T@MEK5l#P(YPHS;Ak4 zB}kFq+WVaSIQQ(c_G|6igFmYhdQ#5mz1M4hYprju&Gre=567EuW8oPK{EA)tu6KFw zO3y|Nv>H?XMUq!)QIQgo04K+8NpUjpV|TQ*JCYAS<4RO8#eM94+x5VFd(f{8x#aAW zmo5?{5E=^n_lU}h-PdLKgYa$v9M-Wg^;3uTN({h}k5wKZ%93Oo_yH4%7O8?ejA~i? zyxSxGxEl}O^nkcPS*Y*^Lh*ex!0%)LK443a8J8G#5GZl7*Qa=oI1x-$*z)VEsO79+9rP0udq=blh+3Bj@fOxxrs@Pdv>mB3y;^HErN-EbZek(T zqyzA%CA>4`8YyMbxI%M5^_57+)?z9M1f6jtHl$ZBpTlGcQND!Mk-OAIWr+IT(F+*X zes1Y@Z}yjNV1NlgT{=sYqI1j^wg{_5>k!_80dWyx@R|=ej8wRAj0y?qT1$%X_{Fef zFV({Gg3qYv_<@+Rpb|NWTcP&qqumr%&&7rs(+Ra8oCiNH1bEUjMF_@Sx0KHaSvoJm z+67wWI+N|e@bqAJdaSyjeBI>kUC3D#7*CUv$uVu#GvCp5W@*^kD?c+c*c4%w7t~T5 zUuLq2GE=gzIcp>agTtiMbQ=b%VVZ-DdxuMG@kkd2#-NSvCXQoq7_7cpD+a^SQ=J?a z3(+Dmm|DH{#A@qtMcAzRMM9}DfwBIDtm8i(jDjF~fwJLqlpn6hw0^Hp5d%cjQmks# zq4GijB5J)9Ktz2k5{T%(FK#i~V_Q^-lsyu42I~dsYB7yzxn=@b3{)+~YFZceIjq)B za}KM|MW4g^3*|LE?U}WM)SWe+%6|WIL}D-NXL1%-tzmZ!?+ZhK?ALGLYqp`h$5Wk z=_pyfHYyS!WY6a6ix>DJ{_?YwEKUnI(J@yK#ui1Ga#J7>QzV!38teWC(LMmOKm}61kWW-@I1)gWC#LsQG5A(bbv_`_-Zt=qw0Dd=q#W6=;g8A>tbNGdHg|aLKa7 zX7T(v=o;NEN<5Q7E`77@#tr+FyV4ydum&H)#q$pPJZETRFmxg3HX-3kP#HiiA!2{+ zejOut=MdT8@s`ad`g*SOX^XJmWsN(KJENTetURl&(RsWxz%8^kqiU4IhQsR}0UJZ5 z$3EwN&?8KTijJT;+T>!*dp^ug;&Pe}E@3LP!SQf4s7xG^mODS~XoEl6R_{W__Fb9B zV|HE^GrJ!veyTBOW0#DziUEZ&kH_pncVvxZZ3v14jLwSFS;!3Ck}x_Qc`~3VV6&2h zN<$Rul40d4>v~B;0e%(aQq#}0*rHAW_BDFZ!HHiU`x$gXgBuJzTm4OmM+E2=n6->X zZxrjY&l4S<#zq=AmRkB8$YlvQTG};4n)YT6e@2&jwGb0HqJ{q{o*!C}G$(!2MA&9b z(Sl5i4*X>lYT1GH$ZUZ6(?TuIhs3<%wBb+Iz+krH(qxRshcCNd6ARssI(am_rdEW| zF^zE>ikN{9#31`;EHgfzSJmaoPx0%3z8DM z`l8iaGUX%+JhFcyM1R6d8mS4b+<9V>Ayg=V1kTq-k74$8>XQ!qStcb3EQwoofhyjt z(?3rOA<>B`>kMV+)d>Zy@`XU@>I`DBR!6S-6D{QhoHl@_4@zn5GlX>p)0M&EF0QN>LtsPJW27v7OrTJuUfFxWOPhHLUu}7o(lP@ZRA0|>homCSAS6q1aUnfF z<$U&EM4cwRR*eAjne~vv9%R6cS7}Tw*=p8iQJMaZ$74hgQ0<|zX>tgtenDAACV?#a zo7MCY^vF~t(wgD6T@_)U=&+Z>TZe22-7}bs2V18{v2{zr$8O%1l=!!72%e$05KfxU zb9Pr=sS49~vX}R7Pb^HgEJwt*8W4V66S#b;DtCOK2HxcCV|G(eUhP zelz0KcxyO5J=#ryK(k-3EEN{e=DtD1_I$b?GfN%CgEC2RZD&45m4j^P7BTxS#vE} zm{9z4*(TS0Za9m5uOMZNA77Yca!)*^cawF?|Qzctt#pFS}iPfx$He@X3+1$jy; z0x~_e1?Wt_*5q<;y9@DLr=&Brqpq%}vCc07!-S99sLgMEaBjSLc60Uo|3O6aXpcVE zsN~hZxcZ^U8xV?o3XjJhRcURGQJHhqa>To&GDsSHVt8}=U=>c1z)8>9Xr|2IAQl(Np|=7|$c zI4C^{WWMQmXJZ7fMFS|6Ee1ZykQ}MabBg)G!nH>k^p?ouR5h)U^NY&&lN0>Y9>j$) z<`KTqZ=&U}$2!b4iX7=Y%6*q((1%CIXU>c!qYI}!Ny=BGozOg=%L;quEHa{R(w%e z1UsHDR_B;r$+ey&M9ZO z8_KU>-&Obz`znsccAZa5w|95+u5E3`aich-g>rWP=M<$ZOIidGKkux{<;IzOR=u7tw(bE8j5r zlle4Maw+pUdc~uO@(@?Cios;%d~rh!SxdOG>8oi^@i+zY_Z@a~y0dML23r>Z+Y}}u zC)@*-qBl0<&omt>rs?hMW&M{DF-?u75V1_()MGbM>RT~PkH=ztzDEnYlw6Ts9@ATy zk0RR}z8&@!VinaHrPSJYzO?I6R5Z+-Q-Cr2FR{^upeT6PoPeKeY`(lvAS*rl?z~~h zhtW^ZL)oP&&CHNQ=VPqC1W{CR?5!k~eJw4VIa*tD3=?t1#yy>jyfaBEgfTjoeVUcU zUR-7IHH)<9rN$H`=tEhpl|EuplqP4FTCzjYP;4?l=VLML$7YwkCFxdL6Z{3ULgyob z2A#d{o(P&5nCV|Gr2;!P6dJqkGP>KS|TWlKDxB zK{@M_bd4qBlQjR}E~4#iKGp@zjWlv8`=c%ep+uWmD9dv`a!L=V50YF1Y{*HoI|=BQ z`s+1kr{EguPb&q_+t%A7DbAnfmu?BjKw1J8HfM&|M0(6h7H6ys=d5e?CE~2R(O4oG z&(8Z8)e8xdQ<%?-odc23%v>sPhWW`pOI%Ggc$=gKEajWUk#WhjvoYNrSN4A5(t8_M z6V2glq@Qb#nyGm!>4$xc*b^H(O;QXdIGUuvSO{EukF!0C+7kOrb8PgLS5#x8$;J`t zY04S-cG4k9z1$#6aah}PI~Xc%t+tWs+}y!#>))E#xv7n>HC|6-pYAwcmu^(bhm-9c z=tZyc!-O|&R@Wuj?K0iNf$<*4PhELMaiw1aMt)O_n9x#4G+x-9>w#9U-ASgCWq*gl z_6JvL$aV(<_@Z|Q>%cpswfrL*gte2A`7|H)&W`h0?pFQOcVMMLM^6SseUyDVqNFG7 zTdms4_-G?J6REjyy@Bp6?3Kd}KFZVBQ;uBH-@p90{J+l3A5o!7r|7ZYazEAXUS@Tr zLf@@$J=Na&Ni)b5SBE+Kt45=pQ#+IGyS%pU!9isI7h3m4&e1>9QZy=q1e0)mAEZC>d|%rLJ@{0QtlR9;Mnu zB=rZB;yJA2GNH`(UPw@xPFFV^I3=w<`ct(di%!6Y$$ah6E(LJaLs6q8r5qPT7XDn3 z-x*d7zn|`$JvABaY>ZD2rlbD8{>7&^&hAb}6F(Ar8xMhxjHA%7IJ-=uF2T7Y=G&(q ztde-EB3b>AiXqXkpX##X!NJ}j*9PMz7F zY!0UVrT*ga7v1_o@t0dj<4f%0F{nq#V0|!MSE;}FxxZBYH@b}%*Lfbn#)?{><++UQ zpF)dwcV~NdbZWpcpGTRd_LuI0zvSqCKNRDM#_Fq5N-5hzN}|tl$YFQ1aRxSCI`EZr z3Rr3Ef5FojYk!eT8%6I+{oeAuOY9>Zw&v}T^*xi+F&6R7{eWaTe(zdQjl*30zDt+k zChEQL&jEfT75-uG4FLeQT~vzHSwwjJY%!^VK)kq&S_1Ub8@eetlwQ0_v$+E z!vXl`TK3tH*X4!qv{PSM8o>=)_lzdf7mp^JXQu-`6}W#{<3AP8_VFm=)2lo1J&ApK z$QT#ghS=YQ1YG<6)T;XXPRf8*_@f`h;sV%z_N6uCEG5=i!7P9)9|8ck0&%DNBp0?A zAcFOW*t@%-;i5XUWT&A#w)U?LV6a^9*=3i2QckfNW?=@fen!Jtp2WWUNMAUM5*3#h z?ut;n9Sncs)Hw*g;!S)0hx0rV^IAusz1Rc?NUk`f$ZhcARftlFQ7lL?2I%_HD7HUG~1Ojv0Xy|xeR5UUO_ z0b_X3@I1sG4ym5j+*#EvL~?AonS;i-b}-9fk-o$IDJ^pP6nB&gr9-Xd}NHoL@IKV(H(dngQ?L zMFt-lxhOfaKco|o(ZLH$<=SN)D<=#&0myHQK{lErhEkKBD7@W7Nu{y!%@OV7>ga{W$gYm zDXQtIRF09N4NxgLLb$lcbQApOly-rkx z_)5~&?rS&BLz+F{{0PFMxR}JxO=sBndBK6e3CAxV%f~)dxr&fN-B?&pTZsvuGg1%q-y)}P`FzDtay?n zAf&j)r+{s)Wp6pBaIOZ{j?A@Jq?B34sjXUl^=io@UcGT}el$4_G~-o$rI=gsq0nB3 zwD_ueQ2kM#3Nb9wP|hj~M~3T?_}qpJJ@3upT)~xg0A{x6+1rDWVSW8pWLVq5!A9g; zVv9$*@C*wIAe&u#7LEbcTCI_t2m`8LA`}`LH|u}<)S-;`V!zc1#eiyY@I%>-h()-S zcRC+g5eG8-w$=06k!f;S_FymsR@+y#qCn)PJ^(+2f?xG_t+q=o9kfjz0e;o$E5+-s zc&rFP2ch6sEkbxE9&hZB)&W;4=CQzTj|bNX)bjWb6@_ud2R@X*B{b5^LPcs5Vt*10 z|CIzlxm*G%Iu=4K0-?WJfYr$W`o~nJQD0J0XRi+i`)aik`ox2M^(i6wP*c)>yPWJD zu@P|9m$`~@zn zz@~RI0RqP(m6q3iGbJrak>@uJ>U?=(DKdCXGcY=+MV5Jv7p?lRPXJdRrXp2;C^#Kf zODOlI{%BIF18WgY(+Um8Y85R4_cIAmPvPDbb|WZ-cwb!#^0JV?+hd2QY~-WCslHZ6 zwLkSAOe-M>K%qTUwGKImtReWUmhzrXXqiqP*y4z4OZ?|T9|5vZpu%cvul}Ind|=;K zYfgwl{jms)HMK>R@LnCdb&_i$aU`BBRQb%FYCE1RRIR?D@4#xIQn2(?^&eu%LZ(=? zmy_!J8^O))SDzhC&U-ey3EWTbJEEJ7Voe5n_a6r$sx_|~`R+Wj!d^GUjC%)0QD8VL z_1*+Ns;NMrY`>O*xsutILZcF%VS0s93xhEhfkDDtJ8esWK>VRC0VRm~U-TVg0Z2a3 zO}PId78v+7d9c=vQy>eRnBN2YxkoX`ONAILv+CGC+ z-fNlHe5jh&Y#Co0DCxHLVd%}owwMvxMYL-kJtMV;&u7Fk|9yNk^rCy_5c`~K9xmF4 z-QGL+KYy703hej5Ql zF4b?fbo}sm%CM;U{D0SCoAsi_d1$p>XB4 zIe56AW28bU?GEovQ)sb~trJPyaf8rPbK)v8Z*;^-QlXk!uFu}=N;f=Lr#p}q%ep;R zbbuCP+Ii^NYy@0DVGHV>Hc9m2B2V4Ev+CR*kz5}>u^*l|weEB+2&L2uD zIdYp7CGC*#N(IlV}op4_1heqRqB?sBeitvFc!m87DF+VZ*t_08Z^tYK{eTfrW_I@=sYElovQr_ZxbCRJF~c2$Knp+_xWuDZi|Z|+ zqC+d!e3n2WWHnt6?C0Xj3u&B&t84D3;H(2;qG#}wTmsG&epA9c@#8W6Le)O?qjl0I z8p%sXq=0Od9#>r92RA-xzbp+5t4nolV7R@;8&uRrH^cra2@KcwU(wRlQ& zs6LvA4)u4i`Qb?sn0+d#ub@|`krTCPCUh6n@~oz5;-{8tu&7hx&M#`%JaY8aA1GiVL!%!V+a#WHUjP$OvZz)(<8h(hGzTiXn}Wx z>t}8|nXk)!BO-Ba*~a)MN0vr@UlIcMtz9f;yLP%OcdF{=Np?&B#ZlGImPLptRBO7& zD4qR$kKI6{T`)A)u!$JR6gjeSp&A?1>_{fcAU>14OkT3EqM7JAJ$ z(?|$j-n@IXx;q_Are3jiVZK6$imlYr8ICrl1B-M}8)_a)bGDxr(Xs#RQQ-<fI-$=12iNPDE{ay}V8VBbB8!FXCNB$H`2jrGUP(AXf*l%Q z*l%{(O*AI7J~%htJiEDi-VzA+AdJ;>@tEMt-qfQ)+?8zl7gs+NIo!6OR^*&X-H#`g9E9xba<)uZEkXyXgf0azIj>eLs6-q`|4$RYDTv)jFsCh1RqQ;lG(0*!b7nLdIU5N?y#h4Uvrl%|OLOeUS{|(B zv$1D3a;r`dhkv%mhQvFv_W0WP2U*O+@oXuQuwq6ByoAFIcGuI`Q;uBH-@mNi@|*c1 zYEaVAdhF-A>?Oo{*gZYi7))fSJbq{F=j)BUDu}_uVHcH!7$e>;TuTsdVKC<{u}f`H zADl(yy#72QraIQ8XwjOk>eM=!nj8~SOA>Bp@q15q6|R9W#gVX1;ECz>&atY~Tu|Gp zzsyi?Wq=8(;}w`D1pgi|R10cd!KcZ;3Hp~YD-|WOZ_=BmBI{>SQ!$s5l1jcfD+Vsy zP$QW};z(`KoW(`0SW%8dyI6@)^3|ZHD4d)B zV22%xb(qc3VCw?l#{E(RAzF&r*n*uGha%Z#o1(jNyCTiKm>uumm_(X0R_9HaWB=1- z*QaY#@!^h!p>lRWg{NwB_8dXAuo&9k5?dr~zxq;5@g$Sh=wG zv650MZTz<8A<^0Mjicy+J=UYZ1T|h;bpOF>dsNXZzU(!-tUxYX&Uk^;+1al}+$3q9 z#3+aDJI~zT&X%|YxU5loyYpbdfp71eL_Jx&Ck==S}GkD#fA5WV&Ai*at`|a0bhazzC%T3Ed490>@N! z%9xh5yV3nFk}_-?vkX1DY%~Z@+8Kk#Cka&5V3iSI7r`ha2#O_}3>Bb%Z=Su&@)XmR z)j-U|F;RN3Bsq6Tx)PnLxohx?2))Ds_u9qoP-XX<*-icRsIq%~jH?zX`)G$9@hhnh zC)+#F3tZ))od1p0bxGW5j4Zpd_(80+F25u$LhF4xLkSwB_)eZ5zZ1#xu(mT$BBI_Ue^`DS~IN^0#NZRALo44XtZ-`XR>{lmq0wYQ|w*c(6FgZ zT~4QpRjwqcd%|%)`^6<=URHi0+b85X9B;z)0eIgr03^T_yZBx2^4^u6jTmS(ru>T} zuhbh3G1T zzEGTNEi_TbWkiV`*q{3&1F>J^74cX#%38qHD5Nxg`|NQ1+JoWz?Q6UTs8wJ4i~RL< z_oMd**<1ZR$5?Zfx(}w1y@Tg0m2zx3G{#%g(PR^@MGmH;RUirheBd$nP_i90HjHBK zgXOK(+FuHITAb%9uG67yr!$ofU(;$4c=1fDY1x0d^B6ja;z?y)?5DeI%~2S!ufX#9 zbgBUC4;EW~eHFFTE)u1|-VtpBqShrsyoEKFsX74kK`!DR2pHOlS{{FA z$~98TqH%@hg6b=gj;+O15C~dYX;di%mnEcEE}z3>2~oa;){(o^m12nc-q8yf)>0&B zU>D;G@G!>y(hUqS0jNu7iBfcq*}@iKwdgpw=v9f~PYp&YTsTIBgmkSXMR@#TShAPK z2}`2Ondu>LAf_y+L{8#XsJ;4VH-*)6v7yFvLNdRCCp}YyVBB>}`HYaI^CGNWpjECj z*&Yl}4|b==std~3P43=>oE5EWczKZT_R7!93^qlW*-?t_Yh|zep%GCNS2&ko8%T2csZ} zUZ8CF9OZ{AGOgb$RKx%gwG^vb^^rnf-irewYP}RdM13q0i0HpBE-JP}zZlW6N21PP zy&zpJrZFwoOaP04s>N7M%d8!{`Ym`KfLD>@&SAB7nsZovF8UnSUnsBX;UK2P5}k5p ze;RcTUq3lhp2KZaJcDBrH_gp5w&yvCFOtKT_0U%1H|SrSg9Z@6uewX)Lb4GyIn~<>_hTZma&ss7i^FhU%fGi&P(| z&Q+tk=M(x;4FCFbJN+K~%UQ0s#C20HWc&G*sPj%O1soXwwMe{E!~&m5k_bqYOd{|! zctljHJP=Qdh%yP>LLYEHTWv9}=K8}ErRd5#e+Jhz8pgF|E7)mEk4O~kycg_Ra!6_) zRmWzpabJ~iEwVukj7Tl(;iC8<-YIsUOQpX>q(AOy4IBHJsf|#rs%nh*m@S3)uxzw9C}>w3xVmH6=#sy?;+6( z88}o9SwD$lgvmP)jmuQet}X;dIGITP%8X|n3@9BG300_*X61ysWp@KR*1z88mYu~& z>#O8nFCQgxMs_^FJ<@UBt+yVaiQ`KunQb&TFjuA55ow(%bgdYRS}Hy>W1UG&v4L-c@}$l$YlY z6-}9HrPoN9oT`cve3m$K^WaMDFS(hU>`j)?aa9#M8jHI!28A~TsJEE&zO8>mz=_lB zYwpReUK+0ltg^V)5b;$lKIQuxv}|2*(6Q?;xSuTXU6X~}YT38_LRI5;fZsUNHGFGI zMC@TA&B8Qzyg+F5YkJt|N&Gl?vYZ?FgtY_eDdg_GOFqxbG~{TDN%ehL3x%FLNxR+r zc$eMfXn2I}$BJRwiiob(Y}l_B#Xx64DB_!0pJDkkH?KHw$+E;|@%%aHTKT;2u(%5} z7?X{C%3bLW6Ig?f;o^CReV#KkG8j4@z*=%1Ua@IY3HxjJ>lnd1hsXwxw`?}i*HzA^ zEy8}6HSUrip}o&i$ZAm<|;kL36ap#hU4SNE4-2 zNv%XdhBi1Jt_GEfL(+2Rr@gVtA8j)q*}`BI4n}+)v-7f;+5K4YQ~QqiTokQwz=h&(}Svk4Q?>>Z1p$XR55(2V>tr0mf7ct4o_nv4IE1?eGcTZ1RO21 z+WD9E%rbvQmwL4j6F8!U|0$jyT97m+eLGu4UJJ{u~ zEHgfzQ`P0ke=}==_+Fv{ zvC_c0Ol%qj;^!1MhN2qXW`*z1cAcqGl&?~YKqV+7^(9lX1xblrebMSII+9~427VY| zDGd8JLi8uRq>-A?%AF@B8A3%`{-L#0gDdRo)F&PIvrI}7SQ5AF0#&?Or+<##Mk3Cl ztTU9MS0@y-$`=Bqt22njS{<1tnLvWT68j8coxyZvu(*pWucmfAuGUPaw90^ME-4~q zTU!hbqwFI@X`1l961a*VkMSvo`_zv`vSZW2(Mt2EY;@poJyt>n{?$dyHt>%ElJLtsP zJW27v7OrTJuUfFxWQ5RKBK0yVQM7@tLKPYERolpeeAVa4kgxtCHW0)W$x#)s|03!% z>9uMEn9r<-9QGgsZoEokYROi!HhC&B*Ah_UW{(j;K(&X;rpY0o`UPbfIcv+}e(Th4_J`nIIRzhy)4484VLfDQWE z?rnD=ZHM!Vt%^Z~1X=V<;1fP>qc*?w!MX9~+0E7S7DN-6*I#e}_|YDHu2IRWe{uCg zkvAX|`4qkvJ`ipCAMCM&vamf6oY~mkp1`|=RnmWSd{3?zMt)RR#)pQB3rxT4D;YBX zMoWga&|G~!vyV$f&DOEgIyC>0q5gYvyfNx8@_*B#PbuqcZmz|g#(v_^^(3(RrsJKB z5xg*cMURg%Bu8p(Q_Rsr`Eps(ISE2Nz4D1?sz(usM;=1|{-`YJr*ceP?X7n-wJ=7b- zRe1D@lUZ#Mn{@1cEcG=6b*)W`e<^9Rr3ZMrYQt*mIng$8jf9I4l+zsVdu!t%9aBN3>=OhaYZ<*7Zv%w4z{C&l{f*8 zy|I@un5b(!nbld$`Anyheyt(&M7Hl$qtVW(;dm1gkY_A#mT$g8ypd$*Ggr=m%x3_i zpq%IhTA#Ih2kZ^KjDbjy=*lr1`J^MFU9po5E?6P2NtfiAEj=Na-PcQ*i|9azm2a5* z$$T0rxs>@Fz2ec}NK-+kn*BTP*YU~9`QnBevX*dV(^u1;;&D~)*xz^9&FRjzIT~!? zY}58+nQ(X(Ohit&2P#ExY{t$sZE|e7D?>#ly`8`gs(6Q#Zt z!}NG8*5`Y)uuI7m8QpM~hiz{@N@i~%R#BZ%O09k8OS>LLMZ?T#D8O^^M*glj0YBH+ zJoj;HsSz!=w=qcPA_TK)gbi}+u(y&__O-Nd=BN+l9*vYM%_zSxrvQ*gQf-XU9FpqH zNm8#_q(v_^rYJ!l%4)6j5tE`cIlI)79Xg*F5=+Q1%Mt7?Nw?CP;16YmekVK42YR$9 z*$zhPR{u&&7h-D|UV%+UJxMG6X@_V^mzbC&%{njV(SPf`rZS)Zh9EE%7q z`3H9qZg1uAy6A9MRXygFi2PBPf>5H(jFja$9~bqB^Xx&AYk&sS`*Nl<8zsoP(5|Dwk1T1XM46%vyn3XKf3r8!c^{_7yXWfm)63KXW-p8n3 zNRXVud|oV}3@vx_v&7X@gSSa)z*4?h99fGtku1>UYrdbj^xnqROrT@pp7=3hPi*is zNimqd^p+OfnM|~KTLSjW_4YX-7eET92oC${M40K z7gzcVze8dBgDW*^d5PQP1NfqM2kW>1I-~FG zIG^Qi)lYo~*8W5-O*+_MP%N>zUhqg1>rlT4Dlluwo)}1KRQ*^>~ zbC6e8f|C#ie@003j!O<3H>Zu^|{8T|I1XCk6u}2fo zgW9IzC;38!(bn#0b!|M@1&pDzn7BjsJ6-mp8oh-1x!OwQ4khEQywsI$1|Xlk4?KTB zDW1bRE)&Xp?}Y@F>2!6&fm71zqd!$UvgibS7zau!L^vxF?9na-aMeRmqa~#r7ep5R zT#?@y1}xr~kyD;$7mSRf(6BhW%CIGVq&RoPeEal+RrI}J2z8VYl~LBut)J?$K@QyCZ~TD(XII4veY{=H_74Zs!KpLblg+`jztmqm{-Rr7C|t6I7ruo0 zEBkm1>Jc(nA57O(>MwroFO~m|ZsWyuo=32;qSli;EZ~cGoT2T$MT>QREm*EEJz3|TgeiIb_@#PCH z;y(4>5CCA?MWslcMTE!C7LzIn#JlqJMh)e1_FYFV>vxyMBfA~=;Q;(|Eh;5#S@od& z*^t-eh48d@Us)Q#4O{n&Ces&>CYxuc13neFe_7){70>qZDC5(sJMcZP9brtxB&K_eQ6CjbC){FMJEQ>p*1#jN@MoLO$$N`x3O#4ySt&` zqB^u>r=dKy_OA_Kuw3xjWtV_bPO%zhVTZY>2E(5?bxy>|>JR7P4;}3*&uF;blh}73 z=?m9WqT%vEEtS_g0`0{nI6#Uu(RpoH!ex%N{ll6=)>=Tmwu8#Y)I>-Afoo!cy)M|~ zj15=Hl9V>|a7Az}eRv)WM64lXXQ0RAuZ-mbDZW2M#w-#naeRj$G2; zzx=rTzs}4b@%R8`bpMuxeUXjcUEslBJ7bEgd?10)*GLc}}Bf*uI z#xhk0@0mVcc4GU?baQYHjUy4}!+tk7ziw>A(!WbJ1Kz!h3_eQlG%m#ckWN5G2QM&{ zYnOSfm}sP_vbUk)S67dnow4lHHLi$rFpR zWE!41Qx{#{3O-SWKk*f)nwixnbO?mWtL$6e15y7WwnJcwRmFytI#zLV0#Ko{U+A(Q z4(kq8Afv3isc@WX&z10tJ;gcDDhsT0)Pw;Mw>j6w;tt;<8)yBsOQCVLaD>T>@NrQY z)%AraH-;DGQ&g%rj42-hy9g~AxkZUK$yBAjyNnm=G5Uw@3b6arq^PE+QaMJ7HbA9l z6wtB78}YIGy6oDtz#*&CDLVr(hlLF94f+X;{cnfe7$yes?wBN5mQmCoxYXG$BnO(- zE;-ZXeI54vu-(J>uyW6BSi!F{TCY^bmSOolk3j3Iuisi%ZQnu1sO=K%;CQPxNSx%>AwdK`Ur>5(T0=7i|S zIaw1J7+X+U;f15X-K(|BB+tm;lXQ_B%T(<##MC+^LB}ai_Hb~$zQiL{+n0yJ)$(V> zlOzEl#Wg+!Y;!Fr=QMymS|?0;l2sX;tAVv6bL|x=WtMSjt5!c}wIms@-Z(fvnj8lX z@~Xal%&quPXs<(Bd{sTD{@4^l2C}L4*l=XHE{V?<1TuVkFfy#K---+~b=<=-pjxXn zvJ+uI^-F|8L*r)sZ^JExd>HUsolp#@76(6+?TA=}TY0DRkris&^+Ox}a*Y7LYW0=kbyqxAgrI{^@T(RfJX0G~ z2qLK{!)`?KcyNtCEsy_DQ5aWz;6n*qLL<#ARMheyg8zRK4F8n`K)GB3DLNKHECQjw zT7cEb@?N}qeK6QptCi3v9_*`63CV|=lK$J}WN$ngwPXS6>Vm-;-p0n)sh)3*AHN90Wfj=;ZYDtBc%;(ux^JeWg`18c$U2!`2QOdN zK7-dZ1EYglWSQr9(W(#o1aS3XDpK`_g41EOgmQ1{k0zx$uolrYtGop0>fFU_a^XB zO$7pF`?VCzmCUvjRPbvIV#qMhf{(EX3=-zrXjIRxp zbX)r{^k!mP%n0ow+BNTa$jo;d^P+qB5c`~K9xmF4-Q8QzN_rZ zy$MwK?Bgzg;#!HZLs@mV?D46VfTz9;rF63Wz-B8_O&zlS#K6MF)ivg<#Z8UFhqw6Qiq1YF?>}5P39mD1KU0b!+Zx_}k zLR*)aOz8!o0;E;vnaYeG6@D^?enr_QviNK^9ikD*v)jrIxoyrK?@d!^v5~D4N!)RR z&{K2RRb)&$sPx&JUFnAB>U0OvVp+H6iVie41D*8_jteMkLEY0PiC$dfiEuy$tPW3{ zaC)(y9jN@rx5Vl9S$>I|!o$Te4^mSpJ!#=9?KJ1;(Y(rMInCP-Y<_3J7k1zmDEU;J z1WTCRg%y=};v;;auA81;?y`iqW9%acmn2xJ@hgR$Iw()5=gfX9D>KgdVubHf*Jaj} zV1NqJt5CGJ2^S1HN{IHR9wl_6`5ef)$mkyeb+iDsWfYS50YV0|4-%}zkZS|=(a4vR zu?T|!F0z*M?5PdVd#j$nzEc^8dI2yquLkB8NU>a52b0BdDaX;MVpnU6EXSw$Ah(H9?#U8rcU zfhUcqONJj!sFCn9QKCvSHT259*kz5}>v>h>IZ+n6stFfOh!uwVptizSsHgi*u$%ig zxu5QHT8_CX{(H(-xShITA4)1Ya+?+8ji{0%Q3k-^;!UDS=sj2=esXOTHH*ZOqtbyj z_GKtJs$WMLv49og8F7~D%P+zfvsQD&21VNf&*= z8Bj`>kT5-=U}Vb9@hpztqO5j~dKh|aOQS1CsI;Mc_BYA+Nl3H-&?zL#;dvs^*AWnx zq_V-XXTbj?G!MR6ddu#sC)1+oxY_`meuc`0j{zppf-XN(d`^fgH9HpO zVRbwcTt`>#Q3vI-1QH>u>3U#47gt_L<1}1dsoAl1KPYa+i_z|#44#rpz`4S2N|+~p zJjP$B+NXZBPTE8xdFhB0xS`Uc_Fz{W!W}pb4693ZZD5!P@8Wy-t0XX7--8tlqYx$f zx_2T7R_nILs0<)j{la7rT>nzm+V*2fs8BCdqC(=Jl7bG^;;aUMx(s^w=+J^h<>f?l zs1{F&4%J5!(V_kh*0fIxyq`)!PI`qJIZ>NtLU%ze&uW?`ermY}8*jY~X4xk|C$;j* zE{UL%`fOQ_OaKDeysr>Wx{>QJ>0i;v6v3q>8{+VN~9;*E&Ufql}KB*!EJ^8e2?8g zqg^nBuy~PQ0cdtr%>laM*~IKOE4u@Q$pfsHTcRuNBPsFewCuHHs@U8f_Ls%; zWb50*f=<8xpDE<`x@?`sHv~GirV&q)K7YyDe#JNLbh2A%Em}pJEcBXhrjZc5ym|L% zb$2?LOy?9^7jny(cj8H>qCYs%G&unaOPvFtADpfr? zzK1rx5FLP(0ilL_LFkQrf-GNc6yPa}lhsGFc%oFm>}zOg_4tpG)eKLLH%9$M{%#WZ@y>}? zpB+p_!=vLfXGW9Jh0#d(WQV;p$9}BkLC-QB4mpuqb^MMY``I2F67R^`<7?v|WHArN zv!zJFiW!~0XPSYW`*U6P5@J2ja*dZtonc zI?V;Ot)|IQ;T#5-kUCz0X+rSt0YkN*))jo3{F|VEC5}Xtz`jXup1`IkujOS?Q!$s5 z67^|MqL$O(IoFUH$uts2YJ=u1E^5V!awINt6^`qCf47XkK_4yzT_~rCpCLYatzfVk z^U`+F1XqXASz37_+xM!`Xy?>$ya_4kGxUQ3>`M_dtYmi8f+8yWHW^^#)+mVd)Q(v( z5S%wuaJk#vfPJ$=4l>kOo!|*pz8chLSXzdsNXZzU(!-tdh%?Q*-T}?9UT#LfMDF2I@pl0;tGuXf}4*xFP=IcXVkW0Ja0;WP$|wcj^$Jwa#HgPYtfVmy;}?Sj4W;t9A1)hh6rNs z)x;hp$rT$dgM^#9ypvYVTbBGNYG0G#oR(?!0g?)w!7?K-LMcH)chkaic202nZV=Jf zyGY8gZOk&XXk(*6c+$@J6~zx?rFHowaS=Lk14`TscJh3^5TD;~?nLrDEPHvn&$oa&O12Xi zkrcQS`0aSNs#YzC1}}J9i@m+W@aOAQEnZqcece`~B^u!4$2q;vyIj;*tF>$1Mxjhn zWRWON{HRtfjl+$}{l zmq5^`FPHf|_a*kOZfMxlrY>nvvC5SMbx%0%XTK_rrbaQNCfi_u@GExlyWZu!D?J-A z&}vNi7fD{}X+)ufB*4kBTT+}1{Ma3B?T+Nb&z_rbW&`Zot_SAZgMMYmC1p}i6VaO7i^2RN(*l7SyE@mrB9xWiNtko|Et z9=_=Tae=Z>;SGf1`)GjQ$&hyd-0M@kN1O;I=g+F#_4X1m|I zKgizd?>WYrtJHlkT^Dy7mP$Fc92(=T>1eVE*CGeg(JByy06y@Tdx*}C8aqe9_dx-^ z>u-#^U%BVGitBV}+v!ZD!`HN01YSJTYFhSRuFb;*P!8g*_0wIp<|vHVS73R4I#mGn z2a7GgzKUAt&H^OF-VtpBTy(1<#9LT%(FF)}V+DXFK1%aSZWlAfJx|hn+E0@nWVW&eedW63~RY-G@grM ziX(dcr5hMv0#KLE5~b)IvxP0fYSCIkEf^5DDzwf50}dkG08){4^)PhJv1@NS2iV%#uZYiG;vUFaAwF|V$ zbtc<`;pxHd^cbAed=&O|le>4#oHb2O=$%gXj;=FH!`5E;nVG?+2vg`J*2wgUs7EQx zQYq3z8HvH*Fex?t9f85>tNl*HU~2W&6RWMq6=AdL7YU`t*lhg^+Ge9*7Jgnl7zIJ} z0%gPJDEBdBTEADQhyfyMDORA`%d_UMMV-IdFuND2*r+ ziOFTuQcB7lN@-D{!s$5d7rX4b@FPVnrkteEUn>8%@Gi|&oW|n%JHvl@Q=W;&tXzcs zV^pO?NkjF}(?zP*l(;tr6Z%pN|N3(~{T}?wS+2LlbyF^6`}viq^G+=V92o$$NW4?T z0-s5e2uPGnBJebLL{zFg5KoJ;Q~sb}%pPz*TWv9}=K8IP5*jC?EARXnT-RtA*P5+h zr!74qQLyt~uxrV(NgYA<^TM`0_8Rw93D+VU)WC?;vK}sqAM$Ovoq*Suy>{wR(q{h< zmG{*BLZ*$bD+8WFbx!AUeaC9_xWup`bwCVd4^WuaZk<0Ek(bWuryy}{EeKMS|KX?} zVr`2G6M17p*|zmZv8VyDUv@vd=`vdj%P>{-AeTLgD=^i=b(zLj9tP=<>XU~8r9x#T z_oqi&l_aL0I8@R+jJk%78tz09dOgXJO4piFy@%s(*AzXM5nIoqYI5Op=j^G;XlG-5 z8fZ)Xef^71Z=Br)B9mV&scqU0zYv(NS#buL{T>p%kby%KaVu}FA5$#_Q3foK#$~EQ z)rG(aClkqEnenWH0i}Z?p$bnIJE5-f-N26ZulKpiXR+Y=7PteKlOfy={7A=nx8Cym zs>8Sxnn44tb3&U+@)Sxq2lOC&ZB!&e$ezvB7ccNd{N-o&lEt~i#}-AHa#J7>QzV!3 z8teWC(OZ`2_tMSal+=oV8Z5`ETQ9Y z6*?M=yCMdKHwEZv^z*pX+jZ6rm)O_blU=bg1YEnxjoF)=cL^OBo?i!c=I3M;`yNkNpsS-vsL63VVf~U3oPEPfs-z;L%}{JnzR(Z9{7@#noHbE7;O`85NV*vE(p#N2Fn z9B3VG&^|1Hrzj?|zlXfPfWP6v$99sLj^2$m&1oxD`|bpS3dC0yH-@4b-DZXF&vuE7ZOQvKC zk`lZ6qSaeOm~W|%C*&jwJhFcyM1R6d8mS4b+<9V>AynjfzVz$VCmr~+OiB`161VIE zRlHfJzfxI6V!xYF))~sss}l-ZxS3U5~3Z(<$xK#+^+skOI&W`v_5*CcLi%uHwgI0Ac(4)Q?57W7ERXO7qIh(jp}p z_*WM(+qxWc0skK$0{`{hSiyhveK9msBm@0w2iEA974)lLU3N4`%)kE4tf2qvM2J@} zS(fBUiVwDMMT30Rf~_X=9fN$;Hu4}}^?5SntG_4)g2Wt@u4+B|FQQJ9UaLlc`OJFA zVGlCk#;Y`@mTWa^v#8v1^8gLlC>CUo5kWw;hsvhOA)xvNWf?h7%#N|_7M-w9bl6Mc ztwT11?iozRgRRpeyl#oEyB*go0L#<26xd~NOG^A(HU!VmTL>rUewr_ncUNAis@8Y1 zm-lZ^tXj9MRRq5p7p6`3f6-$nsC5qx(C=M092U2Pmf>c%vpN|K&z|NtBTkLChU3$t z-IR=K_Uo0Uf?|;pIbK+nsY7pmdS&mIy-pGe>iW*9Y#PmYR4wdv$j@m`p8_ zYliKG`R%uU%C&@T6h`s)8Zt?2INF#FEE22PTFjbPnlhiKjorUUkAQ@Y^)I?*uyyyS z$fOTp!FeKUt|bc-Lc%28lCGr)bAwX+cFXP5F^E5Nzk1JLQK=28EsC;XYy@J{Wj-KBdY>rQ# zn2x8XU)jH;_Q!%eB^3dg9@_$RreAAvxwqYgv>nbZe)3;gGN{^dC?5Cugpb>(&2N2h zZoGMRbM?Fh(V%(7g&&>Sqdoduqmozu;_8PYZ$K#WDLfv3%s-`MAMCM&vamf6oY~mk zp1`|=RnmWSd{3?zMt)RR##m|Hx4P zJvrVO^%wcS>CvZ@bv8Fg#kB%>^dwOArsJKB(XPOw{)|TkqYTNBTH6$J^iXb&IA2(} z_NbZ5w2nCEy(RKERZVN;{G#&xum-$F{QV&DuL|W6@+bo+7&nQ6rLW z7R9&c54eaHR$N!#;aj_>2OEQl%#2<}poe;cxC)QBcx*9!;ab0#{>3hPc{Uq{N;XmxFI2zk^J~7?iIaaY`ihLHIcB0CV(P4mc z3B0o900IMt<40T(&gw-)zORGrpq8s7>An)c-q_0+Ow=`=%<3%We5T{l_nV>RbQaF? z&3A}5lI(ou$~lnv3_ui=wZ+#56gY*9YLdO7moX3t5?wilBcF6cv@3S9!38Vir5WqR zbK;hpH_Pn4UdmiV2STiT!{krq(@@E!%;)G8k0$K~NQ-~*$;$cSh8nV#aAnh1(?sH7 zf8Sv@r#su`Xs~qwuuWkia>6}ODSBfwcBW~w>C{~rDl+Nq>}CCz5|K%brFaub@e|@r zJ$4hNz7@mtcr4cEd$h1i$rXuonjFkK)oGfGlG$5`Ra9q`QfuG&(ym8Q(J*sP0Y;0< z^Hm0BH2ghP0)DQsdG6!ZQX?vnHe%>ptktXlVe?Ho?5!k~eJw4VIf_cG8Ne4Vpa7mp zQuAeu&c*oHYZhtION}W?(1)^GD}BVIC{4~TwPc6RMU6<1#koi(2bH}g=~h}3{GqJS z!0^(!kiyFcdbB9nV1wp34pXMTVOmA3&KD0h>7#S;7xwRq6cjw|@in^VEc=r*JuI1@ zq!^U5K1tVDGCoQ35AGteB+;8wEC8UK$n^Z9E(M`Pn;9v~b1t4%N4G84Y1o4#*8m%G z((FzG`lbGQ&DklqhWgV=!SlBDXgVNVx+Ne3X$e@^oEc&h=`q`2an8rHqE+@K;;g&T zSRxtE&ifeE3ki}_n9qx)yU}ttKTBLqHF%q(1}x>9#gVmW(^_*)YA1U?ap}E{t7)U| zc+xS^QXKmju_rconxq&^a5PDSu@E?wfw0e5#zsH88XHYEj!;ii&d9fu4oT|ek}QR0 z-ik>zQtg;K*lqn=6WcMhnY6}}h+Ov^WA{9GE1&K-UzcuF%7>Hf9q1^p^23BTZC2MM z+3hmj!-4T0$4@=`*~OK94H)@NF=B$ucwu+02U@*$Cz(!`{T&M1A6%(%YXyyON;nJ# z@I~(q)`53MYd7{|QPD~}8JQ34VejlXpXF}VPkjgB{fX$Ag6z`~B|T~1YSmW8M;pnR zNX>=o4Rmi|OCyZpC8Oree#`w-yL*|{l?r{g!u3>p>nF`1S6m(D?5`S)c24a~w(lZy z`G0ah!QHwOWqOKEm~Ia8>Pm2O;^z;ka`U1P@&_R)B(cl;qScUj?08H-+LiJWjbBmaNv}*`sh#9jx0I>A13p)N4pfjRS!jt zmXvZ_5Lx(hMSf>k-l3;3GLAyS;_NEJmiUq4+!6Eb(+^hB_tHbCXN!p~-47gN{#2J8 z4-WR`u~X(w)fsdQa^U`c;|~NlyDC;_m%E(p9}cF2Q)jj(n}cb8slRyqMYq0CY_o;k zjlZkFJ|2U5gbdaP({+{li=X>T<$t5wcyXQQ5p1kz_F10G*#0SWc6WEScSoo0>UW>v z*Ff*A)c)cC{^EigU9g8@JkeNvRZ1ykdq_$2Sq?eujyBG~#!Cmjl1>3lA)pXzsoz_^ zcd7oDxX2e!M^xs`KGI=p-X2-sGhWB|0Xd)Yd)JC;KIP)~UAhc!pxq1q9DskWg@5WD z)%J47BDx*+-Vgv_+eM{FokfJl&lZy^2*kVc^u}G1%h`7wxvbw^UgrAmzz+xDpKDPm zDW$K}lZnrUye=<b5!Ysguv8M7L0FqP$qYuhKeu*HZGtfJ1| z-3<*F)uAOj4dt=5e{BGR<$}*Hy9AVSiq$X+JG}CYhF?92efN>R@GB)DE+*Uzp=LW6 z{=}(sBJNgyI1hj5C|&ScN1(mf1P4g5COWSTOSsIjwtrZ2$Xe|TjZI31&e&v#u)_}p zE>AAP1@^jNlQTA49qux<3yj4!6%?+0*y(o$4pVdn8Y=5 z3ZVoUB4mG5o+ryJEaecZh#8GYN^poBE)TJ4ndmu%hM1&7tlFQ7lL?2I%_HD7HUG~1 zOjv0Xz4i+15UW8S2pXP;*ux>!)0#W0x`jxNwfBA9wf1Q~>;wB{n3ooPlw=C?7Z1nB|4+h&AQ(WZ( z$>iVEeLLzC^Pi(Zgti3v!WogA1E#~<^=0(~++4i+a{gyCdk=-xi4uy=yd$m7)Kh3` zChbpuLtL-+Gs0Rj9U!|$f-5hLWvUL|Gkv`5#P*r#=HMJ|AreppY~*)?^XtY&Ed9Gw zGvM933ncjbA)SDX4qjj?*Dmu5MNDubI92J8-bCdzQaEOM6q+l`Z>S(5jruUT$9*_D z|5J-IruPsCJ>XJaz*&~Jg9n4_(_N@6{~K>rC*w`gy-S4{5^u&d>2&H)?)DiPo7gXO z*$;!}>gW4E& zvM#XS4?ZW=>KlMNK@&xifohS}G9n@YGBC2}J^~S;Z z(d0OAkXQBXV{XNVLVF$3;;ZUG^+z4z8XVjV^ivj&4A&*``GP=(Zx2R>_4QkkVWy6I zI0jT}wMKR#45)sIP-tk}tp9C&OCbXW{8lFv1FFTr4`n+d7U5Rj>3n2G9LVt7Rvio! z8NmFZ3wtma0;}z-T2UZ!Qy+jILcy>4yH?wcN3uFVde;adz^__;rFh*Hj};;4AQb$n zMF`KtMWzfY5*f3{gKGq8dHjcp!nooCA4=d78fj*sqQC3-Nsj$VF#J~%0OfKCr07@( zu?U3zY5`U!1Lz-9nMM;y<$mn-!C+slRzjb6u&+KPBp+%@`frz$z161b7c!!x$KD<; zmlA-QpuR>f4?MKg?&P86#R*njFnxx1e9Q2T>n;_Gyur#};IayAdN&gwa6D3JdEGZt z($c1b@nb3I_wbr#V02K6Eb|;MTJ>R{0IohQa!Gg#_LnJ48ji|7dWkuhmiQPyGkeN(cf_Xb)Ab zL(VK~2tKQ&yr&adrjrM@IHKATf3#vo#4TY!77A2YZSB<`G@K9Y`)bVzQK&x_fw88x zs1n|{Ao#?wCGljT%4gcNO{7p1ns>|k!TH|($AO4y&8tSfJCCfe*G(~Ld#K=&6;|1ME%TbMs^&FY#@7Z)x~+W} zdNZ*tW`uST?V3l=2_>qP`A%bAbi*C?IoCW~v=6(<3(=Ja$c|4#LPgpxp zjW+5W>B1DIS_ikiJk{Y-mvFckuzJu&^Y->!(xfkR*~@r7I)=;dyS8et-!808gtjg- znbHeF1+3(|*9}vd@uR{|#?Y@Q`$QI>&89=`?A)EArsbhcA>Ny&&|)K7Cz8112BD|s z#8qU4j%z8mi2YG=Wp8$+8=kAv9Y~91-JUBtnC!x|g9JA)zy%bxpzdjtL@zFK?OJKu z=ZO4$FL6_NxH#rbsny*mQHE)!IY*D?RX)pU?uje#E0D;W5d7u= zgQ++PmN2`Ey+b8m?y`iqW9%acmn2xJ@hgR$Iw()5=gfW!Y#xQO)wvQyn{&Sy;k(pz znKdOCpn~)&6vZW6KH@?P31TAW5+8Jw5baGpO6W%OIgoXc(LeOTgkZj9+$d49Kg>Qz zuo6SA4b(>?Urxp%3)jm zV7|ee66u=$u1zoTJpVkEbOcMR%)4kzLWsL~c|5E(RA)Wj<^`-_CNIt^vRID3_+acp zMSBfAX+&Kz{Afargr9zAOCUC7U+l6*?)99i@|-9OUDbq(Cd3NE&j6M1zCu0ScY@vA zzsdb{pVM;8P4SpfK8x*)8SF#JMPfnTh$=a@*Yh}b_bohj`Nf+=lhAvxLVV9u#c^!4 zRajYC$&tqmv3_gp%TRJuzpw=Mgrb*+T@-RhU5|Ywspz4$ye>g~GdLA%Sew9Bu!pbC zHb+&2aC$Dj-A^V}Sk-n_g*BmIq&^0+b3BXVjB@REj*0`Vhe@L=y|<rtYoQA6_H9OYs*Iv+#7BhpV zxm#(t$SV|SpC9e5M2KfxPQLB`(sI{P%l&#Z;69S3OZDavl;+) zL?zU1RoqELhidVZ=umw$5gqF9h^Bo~gjE8YDcyc5sjr|{sF4%3X(n_R)bgyRnfGEb z%(72_PHN?qT@pbj_1UrU4*M~V%R-o7vJr64U@{(TogO6+ z^4hCo3%!u)*l$E6t}WXb|K!Nh$VUM|U(yQqtz9ezzjnGScdF{=Np?&B#ZlGImKBN+ zRV&xESOoj|9=m}?yI_R<-gU!a@glzh(CpHi19Zc)iP>*fb_WVW5wDELz3e){h@-D1 zQ^n@?u)i#xCtKeh7IgY0RY(lFje9raf_gT z$=ZI!TJCgmn=|h<-%KMRczN^g(dzDWFqv9nYcqT<D^2<4x1Im&QQ-<N~qJsA5a?GQM(4Gcemohfq8jjAL zQg0IGhQ{qCjW@5^2N&51ngj3X`K`g``1A>QTJrQO`xwKE zwwpY+rIPB=u31O0nim{pzu9Fs(U{Qs;M{la+t5VgY<9lf13()~s8IV4%_U>@PLm2x6S-#pRz*7__ ztB+>!#5`2$kCD|3PmVW6{YCz7DqJR|8C>DVol%;($2ABv`CS1}cM>>H)A7#62wq~k zqQ@R2DE_UymR8!X9ttf@aXu=nbBl)VKfpx*lKU{4CcHgcBw7ugR`id*PlnkRL8m$t$%<;wZJhUwItzo z7Qgp&SK%56QydBF1fG~~?;NW-%>}isQVA3=A$7b0(}dvP1BPlrtti|UGfgFajc zx=?xZlv)xZ5*25Nk6tSnG;gDFl-qXW1Qjpp3hYY}GpuBG)q)}_`!*S1)H*2HVAPIT zF%WQRsxz|g0xqs2`(}q6WT>$^!4s@}HRve{=jK1yVaH+}W^**yx&XLwzZ5};mSQ%x zVCTiP5}xoeRQ83HjKla7t4ghUye~RKcFP%v-nJM|Hqx@Jk<(_mY6Qto$x|c1X`G`* zu%42iMu51vpU($zgMFnD;MlKrXhx3o-V(9$BS~O)s@7BUSBa0gZhnM7>rPBd3x)6JmBv)*-3=(eY z^7?f(Z&~uAsC_LD)9eEz6*z-sMqq?ef`snYWQEC~k;*q??;CtCBM6Emn=Gx3rmTd$%kmV{)zv`E#4%BNb|g7>NV*c8s<}is;On{j zym%>1^?dG774@6hP5t$#ih6yFs}?BxXonr~E2$4B+dI$;T;-vh|BcmkN!)3SEW5h+ zL9Dbcza%a~dkVp6rhF&QkKc*pd06)Hbf3Kdb(Bis3`Qgc?gW1O1<83 zSB*wHr*p-?fP36#R1P9tH!k&jg#Aj*ST)L8z||I=@?@L>Jr zZ1=n4?5+NuW30JK-3L?1-obO0N;$S18sn|$XtD{{A_vpaDiDPLKJb`(h|Z20l0s4B zLH+>KQd3g+EB6d?71!y|w$qtPhp%b12)uZv)wJxt+{0wN&TjV8UAE>ZjM!ITd3`!n z0QLuqEx*2sTAnf01y2sLcSPHOsC9`DZ(+@$8i+1ku!lOaD$}dAR#R#emgpu{`~B91 z1MsLNyfft*DP_^PLUTd&l}N|dVk#niLw_Ob)l*2XTt0`%5~6$wts{4-E5#7?y`vW} ztjVI+U%G(-CIEHmEK!QiFndm`H>N?gP7CQNjA#sVL37^!e! z))f-cwU!j&@rz-}UaE!V1)ovT@dGhsK_zk$w?ggJN4qJko{J4NrW2w&(^7IDi14Im ziV%#uZYiG;vUFaAwF|V$btc<`;pxHd^jLL4`MSy7yJpVn6a_9))X(10b!KVU+ABXZ zGuRYimKW4g9A9R#s_s+wwlq;jVlX&NN=>(6uo|XmeZT|PaucdWVzByZztb?7TD|qe zYU^=D*sS_RLa8x_T%c!N4lsgqd(tLycVwYVPex#_yl#>+tOXdF- z-le&U(^y=8XZSB~%2V5TqBQ<7s#2n)p?c`)A|=}rOz2B7{Oix{^n36xXSv=I*G;*g zhgiU7*{?*McWNo%$N;EC;+-NE_)L;SK%$8eSjmh(4IUAdDh~`ui*l8EA8nq7z*^+0~DsUTjvi( z)X8S`Q;@iJqOEBS)QWZN!%;oN+7=Zi^2UbLZQEaVKfLKOTMNrDRrMg3z_^lzZIz1o zR1eo>8ee%Bq(iDt9txBS+CBHrlGy#}(N-mi=_d}AG!K*EZs_$SM=D)wO7$L&zg<)G zTt;j?i_ET1Og#=igWy77x@N^0WcGVV^g;#>{T#A(&a}<#=PZ!MWvb^^7Xl-kOeBA0 z#A$vAgU%bE<@t2>aWN})!iH^B?Ft#Ydl$!#9m?F8H*I4&I zm`?emx9zRflAybKskc6^S$y;8#KJLPe6=UbCNR&iK6QimEP8CLi;*W8r#aELyQVH}mVhIc!vAEO)CO;Val3{o8zc zgx;$+Wx_s2R3PSN!wZ_LZHZ3qw1B5ed~7F~>FC{9)11r&Z>j>Z(u*~!wxu^@2u2H$ z4#DVeLTvIW07TJK*}X&sVx@s~nb^!0h_AY$xG@yf=r${Sf41xFBNVjNx-4enu&+{! zKqV+7^(9lX1xblrebMSIssxuQOJV;;i2j6^G*S~RB}6e8KbU)F&PI zvrI}7SQ5AF0#&?Gr@!h7ZET4+i?Ys8hF+ae&?;XDl&;Po7Hf56liSB*WmMk=EjX~x z5Y`z?R|bo_xbkXh*W+s4!&7?Jp)tj1K)>37HTq=*{pwej9SsumuYWTu z=>Iwq;?+x*C3%wKgDqUqAYZj$tI1>nyb9xxAz!tPJjhplo(%cwFN%R6XrNTpp?W{t z8Rq|ssMDm^su5s5vmSETgABOws&}a+Tg@79Aoz3{4gcA{b>8&{1fxRs$@o(7>JVS3GoHSqH z?5@00Rh#c*FYn);SZ!`uj)-ryDsYQ>QFP1xqQ_29>mD4S-@9%&EN%%c!_Dq>but>B zJG>o!9BoVoUJ~0gJByOo{fqPnNZ45aqFV-AcaMsk7Mv%t=325a zArzC;D#7!^Qtbh%%iwm)?bI=dKXbo&&tT)MAFTY&a;Q)CQkVmFpvDE(U*!KTavh^Z zMlEZ^aVcYfW=H2vP0z!ldEG0%BT(nJ7TGJPUEt{Xt-_X`PvJTiUl8~L^)~jw9!n?-+XKOwjqU9Tyh~Um{YS_5iInlz3>*_mvYxneE zV=$4K(aQ+*P;U@d;nDK01l&Bi)+=n|()nVSy*!%@!(mw7pUkZQdaxk{j|LKXyt~6c z`|q4`hP$Es3ie%v|FEy(Xl&Q{#B_V-SjCbl@>#sUY;gkzluO{1Ee8-7I2=FXif~ph zD)N0DY)1xPkJ~x8tl!wn7);bPp3Lej=6t48Nx#+*YP_ckEZ=;Gcq7TqXRe$Bna==3 zL0K!LPI*kc67z;$#y})UbmbV1e9{rouGq;27p#!i^V&?poDOzhFJ&&G10hzvVe%*Q zX{h8<=5zFlM}wXMZSLbLR`JQo`QnBevX*dV(^u1405Q9X^l-4!*W?^@|wCUL&a6SoxQC8QX;OZu^G-&WpC=Sn<(|I7^cT# zu|8j+gmtcQ0EhQ-bzy0*U}1Sj?QJDX1%Nzbqj~(kyIOF6!o^|LDiX0 z6!a1!d(9#(dZ{r*3Hnf0Yo(8v6s5`8r3SJ?=i^)Lp|8u{l5{Jr3I0%4Xe}TULBKhA z7W+Vt79|^O&>Y8M%Jer(tJu-*2|=3l(Yd$*`}ai(3ZC}(8r^f2{Yjc0mdsC549Z!b zq-!i0pQQN*cM+aOP_aGdV`hKUr681OGs9S(-^nQCL6U2L4LNCcCjtFZf4%1H6kJ37 zX{BJ#_&FCdM&f*zU%DkA18E6Z*qj++6X`J<#^T(2m3@gg>uxlbNXE1CK1TIIg5(tD z^I|V|3ng~4&k|Qt4c;cH0ZaL2ab#Sy?QBdBy2ks7OYd!5%>+6ox_kC9Voz-FG)Xa- z;AoNtV-5zb*Q+~ zuK^>!DMn2AnWlT|e0Qz~TD^8BnNF7d9SYkYT&YpM<=y})2tI%>dUvo6yfa!WB_Rba zbcrkNd}Hyf&^tTMXSrMTQ{RELKapu{;HM)>deXkts;!KVHj*=unhVz(=-$F!IoufI z`ZV^GBbW5|FF!8-uQT&U+)}Cl+GD@veyZKQ%<4*ozFXmXs=f7-W{@kc4s-TbjYd1C zb|%|*C3?Dw80??iPjI*HM46tV6Q-Mkyt)#cocQ@es@(kEAb|%VDI~GW`=okWaJ0pl zdMb!KUs(|Cn<9@Upa->0#ZU5u3Zt#v(dycGunQQ&E2wy|d)e=F*^g@U66WV>E0sHx zjJNVqSGpO1eD*%@`~jtS4(qr~DD!wBDe#`ibA`%uy1L=ODQWf5pQ;^MbOJt1=4+33 zDS)dUiW)5`<+vcS@aKyB&M=e38`YA<4lzD5jzYuY>?*^S_>tn=5%cZS4_49lYJU|g zMx^Ygy6kvxus4sLGIy%Zpkt5&_xBrrAi&vGu>x+t?H>-NgHva=C!2$5f2qHC{6)9E zPZgu4{wtosO-rb$; z-O;H*zx#{@4`5mQOLxIv!s(dop%_mzR$rA;O4%M#5`C6K4!fg`GqCZ}fv==ffJTfp zL2&7zaB4|BWmFeAm)-(;da#dl*qXOT*7uCpF@8W!x%}R>qMCBK_yf*{@*mhAVQfCq2@w3IG3Ig$Jo2Xw8dU~T6a5?*~BbW8N%i>|H z4*YNc{<#(x-e*Hzmlwj*PJLx*1UGEmGn!0aJeq8toeua^;QnQe|5QBN$D@o-ukOJ2 zhC)#16tvaeh`ZbVE@^d){yfGW)<8Zddv56%Me$&PjVTH zxgjXyLPm7=zr*x{9DG#uzj?7NTjg##&3 zae1L0#Dmyi_!FnjiEvx};XM4IBdxc?pwxNco2e`R`hrDPa61Z8ypmW6$hjow}0 z!C*UMimQAenf#l&Z$|{oe~39h^}mZ>^;&-C%K6WeE|n}c&yhM|1xcZ2im z#zrjtyHqpa-Mh%(BX>q5DfWkS0x~*yfvH@(%wxrbCU)U6W2Q%; zxw8C*3L?^|4^ziv_TlLKPc6=v-a{btfJ=D+XIb739t^HeccHTUZ@g8Vj5kI1E)`nw*c^`1)Yq?V3(O$eL{yon7qoq!;A7MDpefDl#hU2gqDolqSU9XX9L7h)JNza z9d>`36xH-pD#u9C2B;LTTSZ{G`?~Dfw7?;&(n@F-#2N-7!hBETgDF zaH+Fh00X(5N6Zpr%B zMcD5LpOb3!4ZxkCi6Y5Br47xdQ8e{IZe-ER(_GwiGxKL=iiAcf${HykcSjQrKSTTe z20x?EEjaYI11doTDv?GjSN0X7s;_qVXays=J2>q$?0(tRoKJ9_4*Qz zRBc}#3RlaY6;F}`gcR5K6tK;;%yLc}WXTLvQxyjooU4JgBXjK)DP@*%YO7X1d$nX6 zuiiK~Kbjl|GV-dvn#`^EP-w41T6|SKsQy^zjbl6PK9s;EG}6pMMJ*2^`2Q!t@LwxP!~a0ksy-ovSOh|U^|!4~Mg#p@ zirehdSzrj%F;WKzmGcY=+MV5Jv7p?lRPXJdRE=kp>xm0;HdToe(C^#KfODOlI{%BIF z18WgY(+Um8Y85R4_cInzPvLG7bTuf6cwb!#^0JV?+hd2QY~-WCslHZ6wLkSAOe-M> zK%qTUwGKJZtReWUmhzrXXqiqP*y4z4OZ@RsSH5g|oL$0zEEK4)+S;o>XgD9(_tlyc zqELS<0%J{WQ7F7moX8PR7OH$^Pqhuh9sVDC-yUSyRn^`1-cC=?o=Hfq22$s<#Im%& zMyiz(l7&pMYA>%r$V3hi`+9J*`@Zv|$wkj*H-Y zzB?x|MXht{$rcP4MS7sHOsevi({L=1OK;3aS+|kKO_wV-Xl6%(c_D6bQtx z?X4lb9Ahj1$p^X#_dmn}1HUE@)_VLep68R+=#}ib1qS-DUOBD(kqCdz7Wgj9yynBz zyk^Vz+CWLSwGTsYCbq?l&@Q4~^LPbE#nbtYYF>2X8unS&JY2L7yU7dDl?TX``mV9%~4xhS&!^MErTYJfKNs2$;Ww-HsbPSi@cWu>Pzg<|H z2yIU0OvVp+H6iVo(@z_dyge!m@X0fjB7d)g$? zi;JA>9EHRQX9xM&fy#e;OPqe6<(IfAJX{>}t`1|1k2&ot@igb?(Y(rMxf0fXn)5pY zzOchl;)ba>36?Osi@if7U+S`ixMS=i2$v*Ssqrg?ojNE_sOQXnD=Rb3`C^3cQrBhH zlwg1g(yLGumvE^l;PiX^fQ3Ov3DMrvqr6B$^Er@pkmCpRG8yJF zz>~=P1jCOe)JXW5W|!K&>(2azE^Fjo*H)G1L|N#nCa@ie6{fdymwLMI6uYH=v-{~j zr*W9;=D(+WXgf{(?EOh4M{cu%yb)D$B+397T**!}3B3m^#IIdqX&>g|IJVjKB&4o>25Sf~t=FxRT?SlZqZ{%j*)Phe(f=kHKZ(UR06K+aITlL<`Z@yU z0q38mIcLCzBs34cS$cB!zR9#`I<7Vyr(dA~ilt}Ar51GgnL?>TWU1M4yW^SQ8gacP zRCIWD46_6hA*<XftD0J7aU%~McG#=PRI#}|>@SPw z$=0`r1)Y9L6>=%a9JLDedtJ6p;~N4UThoXqN}s=c?SSGNcRIQKzMiYZ*7A9;`4$=p z!ONTXj#hW4gUQqqTha5Z88#Q@E54RsPcEI|Xk$9CNT(@)IonT*0C4a6&mI-75Vq65 z>?Aj#g5Dk}qaN+@gh6Dt_4eq1vkpy2&bjXRfPG7;%zm2x$arI8wH$=keAzBAP+S^* z!rs&i@7%Tvr@zQWRL~w>j(HRj+C!s*tW@+?u4HVyH5^?yqn;7jON{o&B0EKM;2pcT zHP{@VJq1roo_$&W^4cE@qJ&EXMk@RVppt4`k;TGxllQcx+?E57wq&CqA(Q=Pm)%Tb zLhFMI<4t&ZK_M7v4m(JMOuS z?d=IXT2`g1$Hw>5#uuUkuri=aB*b8m<>O@eYNG&8QJkzkn#B_(UYH%@?<%>YT4f(4 zs~MgiZ;bkj{NGf#OiDAj!jC&6$6Kd?t27<&Y>amM-DmW)XcKuM50e#=#e>$CL(eF| zcH#zbUMxI6aq^+sM5 z#9-!H^&){0Ef{)mNUt_P=~I+Bq{EZvt9>s_~GU6RmTuAvKa|B#zVu z%~@R3iWTKZs;L|Gr!?5t>BEJf3zaudsU;yIQE`s==(U1Dlkmlhl`WL+6gJ`)BW761 z?5YJtRQ7E$z^HXlvcaewvtl6Rxz+NG4mrqBV|9WjSovyDpJ8c5qDMOHc&x*0js{ye zH*PK|?Jq?TqNSLPE!cT+O?EJvZF{|OG(%&L^(Zhwjn@|4f3VsfRWyq)d(AGZ=q~%U zh?^wMlbDvnmQxl!Xo@}93m>Iv7fQuli>9n=tEv*7@G(^Og_Vp$eFisaX*spfZjtyw z$=vr9uke#H+z&MS8TKl5^n19`gJvLS@NT(ea$PJhtkm4`$#Hq2Fr}V2&DuG z-Ax{Mscr2hON~85Qig3~mf4sNM~|5)awO9M9-kyoRfAPVfL#Qmj36k|vdJE@JjL|P zY9MCfm?%A1lAJpvU5QTBo?++?RcOD3-P~V~Dzw*UvuY`EXXe)(dhXUK`$&f!@hhnh zC)+#F3tZ))od1p0bxGW5j4XR*@dH?CU4BVig!Yt|yJ-$Pd2akpB+tXLm#6sb1*oHB zJAn~NfjfcUj(4lt>54(a8qh7Ww{#f(e7&m0OADy4+e);|G|uTg-sPgkTCH94HVS2u zB8x2_n6z)-glVi7}I2rh{ zJKEYE$%mia&;VVnk^%dc>w)?9pkEns$=N9{T_i{#G!&p6jVizQ@CNpZF2f&$cMIUK zj)ke8I*rTcHXyIZ~5v6`;w6*WQ(s8`4t)q$= z*2^qlZ9Le8fnUcVOY0vn{t%5Pt27zzl^hrjrh_x*wkMl|X@9A|c;bb(zCdieg(iyM zf>8Fs{@foKi2Wk3h{vi?)&j0WA*J!#XNTh}6*pdAUvP$vYgTH1=^mhrJ35GO^7kBL z%~k3?m_qgrp0iZSvE|ShZ%s#&O}G|0n2uI~CCm>*nM#MRX|)Kvc$(F;?7tk{lX4ulVX>d?vNcCx#J&Q{>(i+Mus>LA`Sn%Q zA``+}qisOcxuE}Q5mAXck}{=HEJPJUTClvmng4cf9VDW zm;ltJvqUL6$82GXuv)YhLJJ1OMTo&}KHxA?;leR0B&2ICDZ=9y!;-yJ3(E^WqoU&n zV#tDz^{^Q{&2%;A#8$QSQ;fhS__X-s;KtwIY zs@AsJ9D08E-F1%yB5J)9Ktz2k5{T%(FK#ia(ggt!`Eb-3tQVxK#Wbennh9VrP_-DV zX_UenWHuWoEbq?P!Iai*;{#3koZ@v2Y z1XMie5kw>)YQ0ccD0AQlCs7(vCK8j&sHK#YyEOGFqhP<-W!HxvDQYq0B!&J``TvD? zX|Ccl7B}1-{>z*4RNHY^GSL9R{xPajqNJgE=;}~F6 zt1ZUWT)#C@imsod4#=Oub&ZB`t=S58+R`Hu1v~EryOu0d`wG8iC{@4GeO1D>$Obhq zBDJiCi{b};TW+V|^<^@@{)eanY~3$p+UUA6;3-t+bS~F-tVT~IF*KiKZ=*1+-8z3T zqE0repMu1-(P*XR8<+yH4@C74Yg<&9$Qv8Vwk;LJ{bl#Vn=Z4punbdG4{`~ND`=ls zl@y#(<5N9cmuYj`ME4<@Z&GaVa!| z23qHYHkG6?M>yZp(RF@RR3t*kp3T)4&+|q6iDKF0*Dw;CY7E$9~4g^aTG2bOl*tAev9fZx- zTSCXPtI*L{+%scPcvFC$kzgM8cDv51(h~csd$OyS#_KVuEUq;~{OlH=^8F3kd=Ojl z8Jhoj_md^QYqF4AE&H}#sA~KU@Efq|_PoPW|Iuc8{as5w;&IhHWb%x>~bgzgiRnoduzY zZvqb^UxD$@+`QtzCCd_<#q;N&YbmPoqDR~X8k7e5Nq40?OkfQ@hKuJN_Ib|G$YAJN z&h;7^4;xWZXMgQ}9V2+>5ZU1Imdz&mdbaavi?H8ijXP%UoOVd#m0Z9rv^Jw^l*ER^ z>m6ig`e)q_dW7jv(GfI9n_R4U&xhF>Lz>a?R=KOG=Otl|hpRzl;*boz^V5ztg5HJm zdCV?qKUVzIzG%YfFdmQDh3?22$=VQ97V~fxGDEi{j4n5ul_XRep0X|(R<5$HmoyaM zS3&MlNk6|zbnF|VgA>0z_A}^&1~(XswfdV#Yg=Gc!XrTnw6V_-9iGNU8aS3(`W(n* z2{?J)34&*IsaFdzfg@V@zv8)}1xa(#H%)|X#uQ~=i6D=^j6y9tupXHWP=8ve#rcq! zSDZHd$r>2ul~R1k{hCgM;Wf1)gpO&9+famr4{2yx`$jEcWe=}==_yM8dVyUuBmn|*~^1S&xxsV|w5El5i2>WfxyL1PbEsniDqQYKzz z|3--ZgqJi@6I!|R#3Vzg$di_Qjrybmf0ju}0!!lLE>Ojrb^7bH$s{^4Wu2i6y*iY8B%He+XW0CCGv~aZ2ytcu#M~$h#zq*LIr{Mp6MBu-^8!PzFbeW6_ z^s60MqhD6guYR>K1)6YD|7O;H#@C1tuU@h&$&(ZxY~hLq`Kkq5O$H6dmlE<-+sK1_ z)#u5Oul}MK2$B@f^|4T|wEY)Rr%A6>BfxxSJ>;+l8F1rOH&aWtnl<3SnA-?`2G1TN zf`Dodl}(dFK=ljCGBOEd(ci?UkHpBIeZ0e79B&=6A$0#>G9GN59mUoy3F!=PPD=b+ zHU!VmTL>q8Nz~oHwhxH93r9NLm6xfi_1)~w{_TlX>z3t+_*Sc$M(O@9dh8Un?!j35 zz3Yd=;$&zUZg#J$lhN?}S$;F(%y?@!K0De?fk3lguPhZ5D?@#6j=GB`Z@(jh%HR#) z{<3)UYaLz#&S4S%Uvcd1U3M$AuT@P&OJsVYj7*oW9Z**YI-NOMcr7|bn>)Nda0j*2 z#M{}cyVJpBYLQ$srZ4Q3)S_Z~K8X!S8`FVBVrcoVDb77J!WMC|*DcZ`AYo(u%T5lq z?j035EjUkP&9!7=V5d z=71fjae?(0`M--?$7qpJ%gB@A_-is6oL4O$L06*HJ&owG}_4idj6nO(ekx${M?DoPz5WirL^jJb!*d7SZ zZESB(;9bHh=|48UKUWMRKdLL^yTr7FS`_M!zt)1yx*>uhem0?6?+u<@qjosH41K#tWY0QWOb4#qQ#G9*W8ZBxwAGQ-uf-gR(| zEqD40D~k&YM~|5UGdgYUDc)Nmk5kpOM$RuP-%n2PPkRs-#+XO=GP)+4$73Dl8bywD z9_7BvG3di%<8$XmlaaH@gxX0vb3Uj8#s0O!emotKVmq^PWg_<$bSr?pp~tqe?akUd zZe!78Hl8B83Z*Ay80ri!$%z(LTvy-WTf1im8-t0=jBX>)L%l&hZ2!wgqn$Iu@g^iPPhH?w`|=&)jU+ptxpEF{U4-k{Ys>aAnh1(?sPu_V*ojOS-ddjs{zo0NWHMA}8Dfm7+H`Q5^L$Dim`*KJPe9fM1^1Y{sB0SlWm zLu?{FW+jWWy(HlPVRKg57m2g(Mq`O&JUj1WR4*h*PGLSTmPgC%7^<{d`Dci$sRnP8 z)PSXYvp8~MM-HKTviA~~-rKmEHeQb>9TUaK?4!h<*x+fBVlctcBn`$w;HZeSV>)LY zvQJyaMn9(-8%;KjP>)*9$hVUYN$TYdveZa*T<%~m?cbW%ajA`+HC{dB8szh+dx?R7 zeX8SpUAj>zA5OM+;LV6tewgs4&FZ=&yIrMwI56Jh#F=M3r?}Ry0VBUDMoc7bN|`s` zo$G;CuiZ(elVyL0!uAJOYRFs@2Jl7i4%TrAbVlFaaX!o4s-OA}!l}^FGX>eFB1(GF zzSXL&jE^>wGm)AL*Bj{G!j>+UfwcTv?x)(_%dD9QZy=q1e0)mAEZC>d|% zrLJ@{0QtlRo~jCxd`;^QD8+MF$7Mp9@4b+qGM%n&IB-f@ee|blM;4uc50m-Yqg@K% zs)wRROG-H|h%EfMBEK`td`T;c$ABGTd}JJjhQ--s5_JjA9WmcN{a}^E+bG$o@K-z8 z*iUuYiQr&w9y?|3RGmS`AqVasF#bS*v#VkS+9HRW>gd#_wxSXg)f{y7N$ z9EE?(>!dxTO}L2f;$0yCz_yD@kvfYAkDo0jRS<}GvpNePtc$D$! z)gAbr*N#xjVSg7AaP9k3tLpDREhAgukA4t~3t<1*m)4N8Rx@Tb++Zrp5qG*zb7715 zBN*M9y`vi%E~-OIb{fiK>%iIo2FnGXU3Lj5{ z;u<-HP=X8*vOjGkG_qkShge0-Xhc$iL+o&Qh_#bxa9|_F)JrJ}n74!xPv4v9G&exUlt8xpg(h4dkj zeLx_%Opq4}Bzpw zM(-~0V6dGr#j|`Mnf#l&ZzqpBB80~NIT}Q0OOP*|5y?4VI=o$9RzEUMh7o2m1~!Ite9xbN0}qP=bEZ?NN=KY8Yvtz zJqpd0s4V{*Z&fGb zP0_teg%~qz(nX=>B2|hiR_iF=HGqd`H z4uLRvm3_;5AnHHFb_h(dy8C2kM9~qdj!}N0%YHblJ5+&;vhJqBajHF6!Y_hw4)ys2 zt+K#6M@<+Iahr2(Ebj1sWaF&AcBOc@l07c}Ckn0FSEAe)UX)K!sp2rEd<5(wv}EKK zrM6+uzK5fzkI+9l>~(2URMS(b93w>=pi^%zGO${#FTL;FI8ki@^8}-qQ`WKeYpEb*b&1M3S0|L?N#nZgg?Khkf^v%likG<-z%QXXcNn z*rZGJ*n7L|NUR0Ra$#~DECNFau0pxF4pB0S))d+A2cMH_^$ozCpot>MK&d39irRIM z4MrBdJk7;TH#2``rbuX%qO6g^%-z4Q;h%kW<&i5id&$kPK)~1nFB}E#Uaegwc}51G zq>JQOX15WC1$Ig*k5i27L&5d>5|31EUmglq%byibk_3bl&-W={n`=Qir=HZ&CRx&v z5=iARI9CH}N9NiqQpzmj)K;y&cD1Dat==@aIGUUQGV-dvn#`^EP-w41T6|SKsQ%a# z1%mfuK!)p*_gZr1{= zQ6O?tAAlc1!LRzeR@+tEq7oB|OTxjgT79K>-4%}&A?P3!{HjF=&!m!+JUdtF;P>(1 z8i86K|DmEVuK2)*61ap$npvnwZ9?o%g5kfC04SGBAVtSQh(#duR|~K@8O`$D^3#!0||><#pdoNy`*RUiv+}rWqI=)FR6~$BR~d z*e8Ii50|9s>O?^=|DZO--XEL}t0k0sQ-3rm)q%B$rfG$SW3`GFf$Jj+>x@N`sZ)se z)ukXW3kkeEc8JPGJ`$YjYjsrnQ~$xV5`q8}+Cx?AkOR#cg6Cnyk3=7CQ|z6c&@!Do zu*DJ8miW(wJ_2N+K!w%TUj0GC`M|!f)|?Q9`eP9oYif%s;XQij+CjUUY?440s(hwR z+c?d~_wEGFB}yO*RjY64JFr@)6f8Ye{fAhxkSSL5P}GiHOVm7c{`KHy_kHI_lZ&3s zZUXnqIMf05?mrGhRBK*7^4)o4g}rWy8TSs1qQG!g>b(hkR8xUK*?uhrb0ug?VIE^F z0)vFPcG{K#f%tV8iD?jowj(%_4|Egme~1MJeoY>%b*C$y=abgxmF&3%2KupHIj#MX z2tU+@(kzy}%QCO|hpKtamhrWLl5T4shTcqUiy5I^M7!p>2>Ep6MK@aT#31vT>dZdt znum+_VK;dpy7BR{Pptx3I>`+$SEqi>jCE%$qLn&Q5nR7BE zY9Tzqz>iDyTWy_lAcmIuYD+VoBuUs;tR1YT7j=$wVZM@Qwuuu7_|zpFE(WX~w9&l1 z?YT@m4_d*KE%x~?yN&0gV|bQw*H-QI+l94>(AH%pQ+h$D0BO}3y3a6`89yrgWDNa^ zvQK33*=#yQJs9N;N`aPcRG)mcpY1zx87Dj~@!m9r78}_*k;EN02t73?t|GS-VtDbU zO**Lb*&AGCP0!Wo4y47hZqF4RXj+WUBZ%Vy3R_V3v`L~D7daU&fW!%>7yH?P%71)I zoPM9>m$)fBTpaV7B{b7JO%!J0Y0lB3d6my{nxi$^_CuOqDDVX=T?SKe5-eeM7gkgf zxsQcj$UeT*WeIV|*hdg9Nw8AmR|-3IP@YiFnf;cw`8?PpA)pycU6)x?f&nT>uR>8= z!iCzXI)>dx2Mju*5srdxG@k=m7a9EnMIGswZ<)_NNU##a=o+YxM!uYkMHmcLk{<2T zIJ(x%VZWUOoDCs@HOXp#jlyAxRpuMaDUq(}@3`p&xbEkuq$5~jW!^<&5<=X?o$;{R zP@VO7n-{Q#nY=iw$YMGA;)AgZ740?fq!G2yU5bGki#?)B_v zRF&sMS?H=JTr?q87>`Z7Slo4Q(bKc$OR^z69Qf-XN(C{>6oH9Ky1yb4@L73}>|XFjkZF*`P05A0`O<%Kj(!_}3V z9c%YHToGbjMgn8n}kw^xWVu>4YONfi$}VB$3^?*t?)z;tk!J>2N^)H`i03L zxc((@|9q2w9!o-ndZDuKP8?KH(4ktK)c}Ocw1mfH)So<$s5Eq_7Eg%|)khQ2q5clm zw0Ff;RLJa;Nqq&qLXDiLO*5gppq6JfO~6m;NN>3wjLHr*P(I5(0XnIbS9VDRoz!Q` za%2J!&|j)9AIB6UTcpnZw!?mmf- zH1hkBR=9s{pP231>8{+Zs+OnO$^MI?s+KK_5K*XBk!w*P>Nf1>d+bIU?Sc{Zd)E(# z#S8rkK(qU54$uwHCT71`*&Qehm4XaCz3d8WAA#LruOd^$=Jv3^ES@J@-yRlp`XyCJ z74G*`=BQP$-|Mn<8s8A;*qTP5k@|@;q+GssKrxj&o$OXx3oDq(La+H28VSM6oA-`Z zcc+8N)GM|w%vT6ev6Wgn!_mfc;3b_on@E&&{JGbq^=`V5- z6|_f}V;)6>_B80alCkmDaCG5}dMPBcH6kv9H?P?vi|iE5fp_fU)?jmd_SAF?uO;^{ zul=zgO1MN|q{4pyDyh~LSuAWfc@)QKL(6Se0BK9sOFL%jvfu2on`umFeQ;sC2``Ua zv;@K$go`1kB#rAjW0w8U}Zqx2CDfuS-#pRz*7__tB+>!M5&$G+dpab_>Yp+3{Q_YM*T(p zZz^0Sr5Rk|$DJ{iunAc1G*F$UX29<;^}y-wT!&Z}j`<=`4y z?m+)`abe-;G2TBQiYC*^`4NwKz`Yo3@Q@emS$6#i(;|Y6LcHiHpp@!17e1ZfbS&OE zb>I2HWHdZBK6h?3895sX?aYFpYi8)y*(W;eB{}wEEf2QuDChzpzN4S*u_5t}tUbOq z{y`S=a6DUzB&?XxWxaFvb6xggVm<7h9c&CHGE|#eF11B{a441Y`tyjG>R9&@t?HCJdD09pA+;pob{4<)bWhja*fZtongI?V;Ot@_IhHERZ#kUCz0X+rSt0YkN*))jo3{F|Wvfb$=h<~QigQ<3$v zsHvFCN$G}Fs8~@ljM5$F8d4*fM&d|q(456ZtyocxRAQ+=Vv(o`_I3JjA?QNo%~NVg zh)7hNBR+bqV9+FdA*}5LS|a;m#0)E$UA3Ty%Dzno7_|;cHW;;IRt$tTD^IJFe4|4S zGSpa|;0ac~8pI>0f(tImutz%Vc&x*0js{yeH?BPaBur_4DS{9!#cXWB&Wj6d$#z$6 zSERWYu@n89l1OvL>bwba?0>rKhIEZ8KHRY|6wH@w;lc%9kh3dD27q2xF}HUY9R1aA zk%+#|f1}Z*V;U<`PcUqd6A`P(a)~P7!EQVs)YxdrLzZ=i{deCuggKJq=|waB#~ydV z%>-(@Rc&F07?&}0uwPH80c*+NJhvQO+l8E-PAZiUep^OO*KIVWcg`N`QDA}^uPwU& zV6{D}Xck}gnq5}OWkbP&$ydv%|Mt_X%YH55CQ0)oMmcObfN^$eCSLBSv%@i)IMN=m9ZY|t1vbaHTcr`d@Io%{$6_1kSij9^*!cARX zzpmyjOMVo!ujLuq`96{goWU|9FhVIoLU)s~rIXlJS!vMB*h3^`*ts*y;PFWURW(>; z1lUC|$_RpD$tH6$70{@!?1nwme2VGWhpK^?iDRPlKuB`#kaQ(FRV#_C+@Z?lx3HW0 z>rv(M`uJ5X2>x3*tn4NFEBi=?9q}uv4=39@Frs9YhjRWmR@WtQr!lhZQ1Js;Xyp2Mcq{t#soYEpQ%q#t7ysj&PwPso; z1fb&SKF;~_(P-z)&Saa89`#T+G;C^9m(!_Yl`9GAo^agHex<1-13QY9)UeRUvgYa$v9M-Wg^;3uTN({h}k5wKZ%93Oo z_yH4%7O8?e?1q?rKEwXF8xP;~fVeJ>o<#DSy#% zUChXhJ({YHmY;dduvpvV@|BL`ZEYP@%&=Z&0c+#IE)4uS4q00NfboZDJXxj5D(I0v zY|sAO9~p@KBCm+Ys!`Sgu0tWE@!MyI<2N2%UvS2hYnl&cB9@P0RkvJ;R9$#$3aGy35uag%SG-EU!$Q>6OdpFj+#BFQIkhE_KBjqP}-M|17fVy;+C`IR( zEo>21i;kOvfR$I2Sou?fkqQ@%Q6V8+Ye^9vzZjP6rJm+gFUYiEYd8>77E~f9aVykb zeYBgx>V?=)V>+RyAl#?`p7cx+f^pX^0; zyu;=c1ujy|;n-Wd&MXaEd*x?l2Ad+x@`755xt^i^MxcEby5miGV~=wFQWaKMfucl`0R! z(;}iw0^KTmoBP>ni*YsAZ%veH*MwTBeCPt9X;fWEUtU{+ms;TEM+VQ}x<^;3|zwiaQ%eC>ohs6g2VqI!t6 zEhq(AOy4IBHJsf|#rs%nh*m@S_Wh%mjz;w-u zGsx`skm!XB94eoypF~l`P3S;0E>j(@E(At6nMnT1jAtDTC><0DRVW`>Ii+sv-N=sj zZ}7RTXEDzvT0lC)5+W(S6Y>EO({?aAg~ z+F$A~o_OJ{FA!0u<+XA=B(YaTMIwak*<5|`JYU3Les-3{seGV8QgqWfmzpF+oX*L| z7DbqHQy>siB$xB~*8LBrQ$Fc!duz4S-0Dq(i=)X2Q0%I{WXj8Phl-|5wM8_9f-*Ok zICIm&amAUN?Ddw=acvbk8jCv=gTk8v^!L>zzf(8s{a^M~_heTujn@NKSzK#~_}Uhq z^8F2-0&@k)CH8sulO?`uvXEOX`?g=GYWxoH8|@;_ZJ^>{A0pB$OoPV@gin&Cu`AXN zsz;2w4=kA-oA^JGT}Wns=H?X#E?Jh?ES^6HUE`v{3Sv6gaA~(du_g9N zccnW_U=2Qoi{~BodCt(tVCY&}^|S_)9jY7l*Y4Lbf_Dy)4IXdVY@)AgoljeY{Vr?V zp~;<{M<2&6v^Jw^l*ER^>mAx~oEyXK<+JVwJ;HRT=m?slO)l0<=Yz_x>du%bVUNe` zyd=!=a5bn*9FmqhKee`G?2NXRc38q6kn?%WE^0qk{M5b{!nX`O9@&v`~xlAu%tR8vbMr3?Y}W95q!e z7iM2_za|#CA9eC*culPcp<^23HWVR1iDwaJ_DOc<;mCf*GUM}^Rb8G;KIU!5!u7~l zv|_dU->Vvbcd_wa5OcHP1mLPg+J^8q5`qfz`9Iq8U^BK z7B_{W8r^1v@6UFfQQK6o&K;t^LM;N7ppev;Ovx4`C3f{itGDPdj}&`To-+M6Li8uR zq>-A?%AF@B8A64Yb6Dnq`}5bRPde~tnUo~3Bu?%ERlHfJf2NZ?jvwqwhK#b#P=;Qe zP|zx02$ZhQAQo$NWSag25(JjmrwQu}rYnQR?W??++V!|vGo8|Y2U6YJQSp9|C`}XI zR{~e@<8eOaaKHMoNOo*mI9h35Id|GKC1l`VUBui|@c%v{@L%6eIQVbg607q93(1V3dmP& zBM)M?Ud)d(=3Sr0kvK?dA-)y>qBt!7Pw16wXMu*ZlX zpxQ%a)8r6P{erTLoF`_-K)^Q1e!Rn89B&=6A$0#>G9GN59mUoy3F!=PPD=b+HU!Vm zTL>qp9-41=c2{1eO4@g`JNvgMmb6>eDuQ2)3)80izv!`3)Vc=;==ZK44vUkaWw_bx ztWHM5^Jn?Zh%@7@;rQ%mHzlK*{d#4opjdez_vWacc6O5ETFN^zs0`i!?k|fszt-V3 z;2ai_NpYzERkhhmM62xWU3M$AuT@P&OJsVYyh3pK+5vTipwpS7h1UX5w7J9U19wnM zO}w4Gx;q_ArWVOH!}h{nNv$Pp;{Y|E#D=4d>A*{3duC@*5_{buJpvLo*1zoJVC&vd zk<)_nMAlqO7A7=DcWRG2UDS(Ab{xOmayxYl;?LZ#-aptlZ-bS!&>3UWE6pG_N0(%u z=%p|R>_CkRtiQ(!Ny=BGo#xG^iXdQSK$#}E3SrsA6#S0o&LhgB3!XQN)Xe(&}DaKvtc+c%lnhL z6+jO*q~OsEHz30QcTPFO-B5l7`<}vo*jI5Zw(ERqy1jF}V#yTwEZ%uh#P9IsD3`!1 zTMi&Fa5#R%72&L2ROI_Q*iHbDMTO2@+shbC)HR;W>MZ7brc+7ZY{jfRf$e|!XtZ-? zINpR@<*5st<(uyiZzS3I%$0K>^BI6BC{ta6sXsl*9_(cdM1n+Dj^W5B9TDw{oosNy z3VCVhlKdEN9nbNKUdmiV2STiT!{krq(@@E!%;)G8j|M#jbUdqHt#Q_)V6t+)xS@uu zC0yC`)wGs?YtN`4+242AE$Pm-IT~zT0&G*5h@5Z_REpl%3=z|GsFu z>-1-tzP`t9rqs7$n4XBm`g|J}b}6|crN{JE=A(}GMv{Y7`bJ_E)fuJK+IPOR>rqrR zEV)JjzQh#;@2L{-bB)b&H@21<+Zd$t5m1c+6>ZjUBB|_aY2nP#+L~Kh^XSQFp=~Y9 zCvvbnk_usr?txv!=!m^?krutwn4$!ID66&7M@)*+W6Ag=%|Ez{%sc-83oZxgwObO-yJvsYr681O zGs9S(mv+oVV+DJd6U;Dq$OZsb7qK5 zq{nO+i*v6Jzb^YCan{{vERl?7=Y5Rog#^hd%;&{uc#^&t%-#G9aW&Q8ZIT+Wly4SC zPHoh+GdkIOiA(Qu=W2eG*b^H(O;QXdIGUuvSO{FZR~nmyecCcM`dQW3XtHsHdem}8 zzMXVPQZEl9ON~^=%o0XY}nI=d;|c`l;{0+MhtnVew-FKNV5ZllHAvZDoA4k(`OtT)5sq z_ZGJF&p>nF`1S6mn7?1G23Gufu28UB;|3GUXNDAQAP z!gOypJn|?14UoBUgj@cFYyN9QZy=q1e0)mAEZC>d|%rLJ@{0QtlR zo;38Yyomh)rFahOxJ)SXy%!Qxrqk682Tn<=kN#Bc$f6VQVLTcK+z5HJO95QI^y#IdK1g@dpB&T@@=lUG8djU^ti#&YatxY!0UVrT*fH7vA~;@t0eu-%IRc zF{nq#V0|!MSE;}FxxZBYH@b}%*Lxnp#)@X2+WqU|T^yv;c?2b0h!NyAmzLrh_D~9HRW>gd#_xEJERZ5KL`0;QuxQbPTE^(@7pb+GGOlt z0RXmLREpGDM0oscF{y$;yb2E1(}0zyG_IIj&A#KvRsHU=$gn!_!$J7xC@#FuguE^< zgr}YQ+R_Ma+`4}>nZ9T=**rfT@TtK4s~Z2Qc&3j>8J}L=f$v4zgp^s2{ar}FweL@@ zs=xoV3}}Tv`avu%fc;Iqpv0i~Q`HO#^e^XM}e{=}IJ5G+L~;(xd(BB$K(%2OLY_C)rbNBY9Y zlr*_WaVLXXN1(mf1P4g5COWSTOSsIjwtrZ2$XdEr$KTvg3o|s)k!sQtD5@<|vJ!Xmu#w4zhQwSx<5Fz_RokiQ0x4c-& zAyyGH8j+OX5IbBRVpTLrKsYIfShYVDCld}an@7NFYW|)1nXu9(dTj#U7zZMV-)Y7^ z6jD8{xwEQUh~!w#VA+57jl;nS9WrDm=7(0n> z)1j`-{v~7roPEtf9lXdmSyyyGRmQ$#SxY(G02ANCzWd1K{R7L7%l~y|{)mcGx=D|h z(fwN%_C+>&cYz0k?Tjg&Zm?fSC%0TEtB zCwoY%>#%oHSe+=L=*&CP>P$U_re@Oq^f$x}YGr5jsRG`h1a)2Z;o!^e8k}mfui8L>l#Be_!(RP44jv4y zPj{iR{BOKfos2g{_bwG;NW9t7hx0SDda;?kJNUek-H|HE6N|EB8tsEq311$Ber=#? zW>%ljArL07vTu10ME!@@4uL6F6-Dbx^JuqWNSpmam;G>9cc=mxW!+7M<5YXDgkJ>V z9P0B4T4jNCj+!ta;x^~nSlr?N$i`WJ?Mi5zEgWGoHGEuDMu+Ngni%h9@S=Q*N)?AO z za?fp8p*GPXmM-8~rRDcL0>caVMsH%-+)twLi53YjvsZoKlMNK@&xifohRuDFlL% zMK4ctansGrpP4BV8l@;}q=4LAl~2E~(e>!FE00_uC=`N>;-pVtU~GXGjskbD)-IDg zBZE)UMRF{|u|=mO@3`fqWM8vb_MzZf5gIRP}|Rek%ITk)aLUWc^!s(MiU zQEjbGKlNmh*8dXc?BbkN7LE+pCGoip8MZ^G44wD41S7-x`mM;YhHbBD#LOkOc%%!@ zu%H04*|m?rfNHJQ$WDX-)h`hW4UL=izYVt(s(_aP1Aem;iUHN);D@pu5sPpu?{q%0 zA`WEuZL1fhAgJISdpH;ZtL>{=Q6O?tAAlc1!LRzeR@=p$2Sc(Ea*Y7LYW0=kbyqxA zgrI{^@T(RfJX3GH1|^8H2ljYyjX*7r|4>mFSA5_@30y)W%`8+T+PD%c`;%b!uOtA< zsOptWIV^|Ckar4)$LY4EEJ(CG?30`|49d@}Z`r|8_aqdnHObEztfg z;c_Vfs0r$8pHB1p;OJwG_;i%(fI%qBW1+0)sIYfkDDtJ8esWK>XU?8q)I^OA*{q z03;viCfxrJ3k>|4JXq`Tzj&TcTBBF8=N1^~$9m2QDurA{zEt*F0Rb54*_=(Uk|tj!#47&$dov zc74|mez>>KKIQ@_u9X-&lvQ`j9-nLpcm(Y(Dqmo(|~U3MGKN5^paeb-j)_1lHD ziO|+%CR2Jrr~n!FGIXC|Dl>jm_{kXh6=k2u;ptHusaRG%bsC(KZ(Tj^5_a2qG zx_1bP6HYJovjdg?_?9^RKFcq0Q+T*I<~2)b_U$95m{4&ip5`1qnpgRPPIG=Oz!!En zp7(>PI0=?8yNkU;C12{Ygt%kuBM6rySgG+Vg`GMmPpIe2e#bMU1anR35Cncl`Yt08I7kn{|9CtBv^?d*9PjN zkuN7>5e5S-&swu(<;e}?cR4$7WIw+junc{9{BA4z?IhrA2obDFRs(Dl4od(q-(XIO zbWMN9jVb^WeU3^xf+bewT{I>k#9iDO533E;S&ui9yO)NFa4&{&hB&LpVmbQagRu)0 z?KSYE5p~J%qX{(urG93Blo(lsyrvkLRU56q6x9Wcw{SIq^1(fUFwy; zQ|y-h&F)wJoJMBupS?e+&j_oBoLXZy^ZxT&H@4*W3GZ#+NN{&hg*4US! z$Up0OWrIE z1gmvh!9fNPtbSoK2(EugyMGkUMNcaBSQ0AK3zfxN;-Hd(4%OnU2A~535=2u9xM}E6 zEuIn`s*fh3L;W4BX`dE&KbeG_^a?d{qBhNh?t)sL)ieP=MctV7Ba^@^`vmBuR$kd9 z5p+_YEz6M!KtO+~x_qos9T$wt8agUNWXb#{bT$IxuQ{hSY) zAk2OvB5`fm#`q^kmPURJ)WtWgFa9*w6Ra zjWpT?BkcFC9}bHb`W1j?SJfP#8=g(fezUSWP?$XWdbuTvJZk7@+`WoS6`R|`{<3(U zY<+uJ(CL>{A??T?m($!d$$qcP)@gi0pkr$qfyTZk%7AeB+5tsI?sT%7a4nP}dg8t2 zTWBN%FK^yETHT!vCR0mnMWxgXn+x-mP)o2Um(FmsF&(5!XU_K1CY}H6QQ-<RPIx*YQ;BDAN^@hce{Zw*Hm&ZyT9_Rdmq?7<#cWT$8j zyki%)2Akuvr{HPHvoGskUi)J~lyHf_NQM6ZR8p-gvRK$|^0-S^*>Ha7uUIebn5oNt zv&(L#F`@Orh4Ci5JaX~>nLzmZ9u?xQWYgbQ{ZQm^+k#q=IwtO21jPF;?6tk{CYs%G z&uwgPPvFtADpfr;zMnR}5FLP(0b%Io(g;87<7D}2qX17)oUA^Y#S^_WFncw~Rjg+p zC94^p9&e2Li~QeIxJ*hjxWbP+BQ8ilYNvt7G#&43jCT9oXY^#iLg#)gbLRRIdzh?{ zEFQGB9C}yFdda~xw%mbv+KUSdM~|6(*-l&y9Qp@D(PTO~KjJYDxEF&B9`b@c%dS6R zT12o>h!-~m)nIUQ@L)Q@=~%pT>b~=X$!K_NeD2(6GP*Px37_b&m*m)wwLI9~{U9U` z|7?#9iFah}@wM>}vY3bC*-|87#f&cNox7jwvKJHUVfXA{V=$4S^2FV-pRYIasvwpg zSYcq;Z2 zAWU&2tP^-@y1jF}>NFSBw(2i4^yK51kUCz0X+rSt0YkN*))jo3{F|Wv0N961^BeT$ zsmS_S)Ktvnq@t^EH4w*Td$(axFScoVSuQx`bbkQ&J}5=Ux-<}5C1#foyI7I~dB z;{>?4fq$JoTnM^QdGnN75+V{6=ZKG9D;V@7%%`vszZfyYN@iCrD5A1&lL1DpgOUwK z?U)q(dg1KjTNaU7zSD+BIb00_68)bW%ghXHe%>9q*7_) zw>72Gxx}UNSdRh|)Oc;t{RgYp0Heyckb5(z2|P(`LD9 z1j$dyQzO7>oTEmto|2zNfVjDzbDJq=A+>`tM*ym7*Lk7$uXf}4*xFPuPTIF$V0(4K zm7Io5G#4$JHm1K>bIEw#l>VSnoQ`pO6VY&A;1|}SDHD3P7Va5Y+#oo-Bk{?CwYtlVJ$z*s1?tLT`ID=(IV1!bFgzmPILLkE>_z+1M zwvAbaw&QIi22a`E;%O-os@)Xl^s)3k^W1{pxNOJCwbR{}f z>o9bOs-@q;Ztkx~)za(ZSG6D{cI_n&2V)=Uup@pY_2FcD2L{-!@=(tI#_GBx?leZ0 zJ*W5qth6q_BrZau!9xqv+3n>!d2akpB+tXLm#6sb1*oF|i$yohMB9Gp!Q>Q1Nsh=M?Okoyj&GJ?f!uXxP-IE~it)DpwNJJ>j^Y z{U)9xW+`@Mmd_-uro*q;#qWBT_pbD8#6YVtsXljsY9V+ z@ZTwVGBDuC$0`pHWl6FP{D6r>i&ViKCi93t?#9D6Js>Vn7Am}fP<$T^@H-iR4~V*o z*Pyqy+gS(EBZe}obRQvub9mm_+I;xmqz03mE#)DlL z_;nnzwEh9(57BtCN|W^tuVx2^gX!SRx$Vj3VA@~mFP?betuGLU(lX?H3AG~j=l;k* z>=$`OJXVde7H}O3DUIJgI~>3E==wdtk|TAH=L$REvKa0a7m*qi)4$5?Zfx(}w1 zy@Tg0m2zx3G{#%g(PR^@MGmH;RUirheBd$nP_i90HjG|Vy+&M9tv@Z!#k+u-loBt` zRa~b-+fHXH9loa3BJkpAR@1WoaKiy?(j>3q21(w&RQw3muu-NkJtEiQ@ zcO^si)@U0LwJs6jU97pd$uE}#TlZ$ck}{=H5!tX zTe{tw{iPciU;xtFYuV<_~D97>-P#3F+fBu#i~~HoIvBBK8#u*Oe7Fd>!ko9 z>SK{WME`wpixKr+RZOn|jyCYZJ{)xh>jmj*F^y@tW&&6YR4vA8T9@`YtkzC*4y(^a zpTqhK$ObhqBDJiCi{b};TW+V|^<}@f?Dl^44^ai!x?jk& z(RF3OQ>f1AT(0j}jh^akK=VxYHVV_)t@8&X>SVL}DM(yfi;xuMe;}%dSlgn)MBdm? zwrwSZak~oYa_pDg4{y56*1|GORXxZhFs`IwTV)}9s)y?`jjucm(jnC+4+Tnv%1U0B z9&J^Un113=Nz=_p#9KVtYM|GX9I14zDb;&8{&r2#a~ZMqEXvDNgbRV`niXe|+3z9I z3mG_64p~2mTB^Hv=mr)@<1*C`RTlyyoJ=HtWyZ4(29yqpgetg?lrtcRBJ|2Bb?fg& zcD#Rs&#ga;vDbIYKHPEMt+#xnI*d!988py3C$y;~6*yoDcqq*i4@vA*QIQBCdp1{J zJkJ;Lm!I8B7UvQlTNGi+O@Tm6kzCH_TlYVhPWhy_?XA_4pu2k0;Noa<0{FYD`jROx z&mAh7GSwCl#@t-u%*|^;nVanO&Cqe(Gpf+hSln}BPPe&g1epM|jx z5os2t!Q%zOCrRGv6>A68BgWkamdvJ1{3u9Xympn@8gjJ7r20Org+ip7*;`of-H&6w z&Es8mkE7ucwjV2oZ7U+WTC-ulS`-7F1)+#mAx~oEt+AclKHLgC1e_qUZ>kqfIW>Oy@&OjUiFORA__a z;c8HsI3z81e%jFn;nQy(kJ))y%m3~i&*)OG7GeTNwD5n$b3+S~=A>_$ z2-}P)T2_h3171dCKkFMb@FI< zO|1x_V;bW&6d~b59+3Sr%^9E99jfZ`Wb!d@I~J}-#-bIg-Tz+Iu)B-h(SNB=!_ND+ zntS2D(XZ8rI|KVDQGuA74KFOL?rEX}V>`)ANAJd(=438-Qx%AnUaV0yYk^q(YGD#| z2*y;5^t>MVjXm}NQGr;AVO=IRjRNtZ;-*kkquZ?T{n@T_8VHdx?Xa&Yk54O&Z@VaI8A*aB5U~f)J{985z&(K>4C$;@VDW&Iy-p zGe1N1<@Vzq#$O+}gIa3h?d;Xv>0mOQQ%ql&pWHJ_Y&hDO4lELDo?&}tXEA3md)*>E z0unaXzwG2->)uh3(}MFv)?7;#CNw{1Ku&i2PLY|FFu2`vJ9P}=&)l!xKiD{LgO%Uz z4)uv%3Uj~?)VRR)rThjlo1_ zMz;~@q23^_!lT(FAdKk?N7)y;?9Oa93};|@e=@fM=)r~*ydndnU2kCjom0+mHDZtongco0QCi>DGipj-m4Y;k`ua5#R%72&L2ROI_Q*baJv%RuS@ zirH&>8H0(s#*6YbHZ{B;%Ggg}`(Hj9?VK5oHzBKl$HV=7gaOMp-yz;e zvh$fM=RoE&08voZVyzF`N<0A`>}3o@f<#x2;m9W)5$%edY;eH}d9z=eS!S>3rOZW? zS!c@;QTc|+pUkJBl1rJ-(JLNJn}?(^hGDXDzPO==tR-C8^wl&`xsLsPhuxCyY@4IO z)+N9;g^9=s_duoSjm^xHX*yI))7#l?{g)6iO^wY6mg(zz>}E=RD~9QbSgg-iXknL< zD-vl__;&s0qoOwwtEkQ>rPjXlrCpDrqG6<*0*uzZRnKTH%D{|DhnxjYB?+etHoub5zCh83yTmWZ+ro#Jz~UiKMcxrL`B#(YeSwlcYizqjTA(S#az{ z-B)6-T%<)WHKr&*AIfU2^bwPyG&#G}k{ybMVzX{AAB#bl8|;lqx6+#6FPIfN9}(1o zaTLY+eLY%~Y_LId9ET~>-!QGp1k*<5oR2Eot=Vyu{`(>Y1y6f?jqW+i{v=HgOXepj z2IZ_z(lwThPtyE@yNI^8`B)bKAj7s zX=hrR*>@b>hkcaT6B|5DQVb?Inxw&42prdXK==I9ma)-?tFh5!;|TSr<&1ng>5!ye z-XKejRKMg7_R{{XiT#q=*jeM%L+<|^M)wlOR5I-_szIG^Qi)lYo~Rw^`FCXAjb$UYTO(v$YBR&8Z`w2_>N)LgjUK=&55mxo5p z3uM3LeyZKQ%<4*ozFXmXs=f7-W{@kc3v+fsS=*Ux-$Ul||KxsxyLBhZ^c0;i-5liA zmEh#W&mXAdX3lC@ndK7l2O%jWvCI3UdRlO_#hH34h&C+m+rnt&eEHWfd~ z7b=Xlc1Npg$prP^LQco%K`Nx29@b_b;E&E((0o>RXeii1bmpx*BmX<0X8AqXEadw$RU4nB*%(qWJSRKUM7Qx_-`cqwYA~@Ka$4;3$RcFv~$btI@ zj6V?I?5bFSZbSRy)$G7{Y$O#4gy#S<^Q^#$TDw><8+#6A{-dV~zt2h(+x z`ir0YOXYu~+jw!k=MikIX!cni71pm}2d2>C-QC&V9i177`y)+ceHU1HeNdLwR8%o@1F||tu-(8d&>_j)&CL~+Jc`2vk!LI znzu*R_l(yuzF%?_ziX|ird%$5@0H7NiTMHe=OFxZ6#g;G&e*N#MRYstT_FI#wu?%U zI*SO8pDiX;5QrDuypBfoyF*W5SDdK_p{v<<9J#9BT^92sI`G3m_~$75Ovvl#;yBDlj(~_lg;zf0iO!ozpC+{if8(Gl=11+9rzx?xW#QkilxW?E+pXE_or6X z-+x*Lw89_#AQl(E{d=y%hVs}t zur`3fa=~YpT>?ru#cG&^9p<7M41eMb+;xP2S^eQ6{GlVQr?c|Zh8sPRedm$Ba3due zE+5oMd95SRUTlH`q*xQ3*M=or=2+W5tT|*YrNu?pk$>Q#Gp>mGS`{w2R|lJ%vEk}) zm#JM~EVikja5Wx;;N%j@whE06DVt;oHSQFeY)0oI)r;h6vf8+6$F`r5s`v zF{2Sl2@bKt&xl~xVe~DbY|OeI^=rGoV}C6>O=`eXWo%kXX+_5HIw$IzrkCIM9@UbwU(c= z4+mFX8p~82yr=ni*{SVw)6KyJ+(K0C&ff@rH#om;Y{b&POEm-Dy^9P!-dpz*CHq4< z0T~^4jv4yPj{iR{BOKfos2g{_bwG;NW7_Bw088wU9Nt2@OdS>BUO?o z7Ujw`+LIyt9>_}j;8Z?ZmiKD|RWq~tgbsl)d6j+3dm!pR#C8Zwu_n@y+c3&6blDGw zb%!dDQP$m5I8L?aO8CW|;v8s|1=cxg!hnd|oNME9hnOxqoU+w;kp9}0(0W@q!eny! zc!$A@@+m4+9LAK7fL(-^jNGEsHjI^wRLQ=(j92V2>SBkzE=`JRdMcG;q-X)5L?4~d=h%$+C>12EnDyb|E>?wCW?sEI6QlPlr7> zZ1*retlV=ORzS;*#Kl*B&m++K>g%`GRofTI2fPzkh~LufLri;yX;7;1%8XEb)b4+-}B!RMq}eFJbOXrf3mP+Vj! zIhMbSz{sMPr@6T4X6Db#6bX$|lr>U7?*4s^W=Ef0dE^SgoDiKmCu_pb9C_g=aQABM zGRZSC_#|CW#j@EeA-FYl`hN0c9}2G5mw2RV`|?n@TK=qfk|ZFcc)m{o+g!^m=QNNt zD$k5gojK{rQDtzh2G)+uwO6E+S;ncYT777>r0}iYG`KjLoB*2fs=k`ct@u!AuR~gV zRXwQwh{z)Y{gj0x!*xk~ZbOFcxGF>E{Vl=Bu)cmPGR)L*566INt=7m+gaOqr5ef~B zoAtka=8&}C3>fg6olp#@76(6+?TA=}TY0DRkri4yH?x9ksNh*JRJP0)mMtwUGZ2Ef(}B#uUdrgOl?s09s-U?D$1}E;W4O4 z1_Ca%JpMyPVO;Tn4<&F3jWn}RQOko!0)G+=|CIzlxm*G%Iu=4K0-?WJfYr$W`m3BM zs(y|6A$v_Q*jKBS&?g@3t4|5ZhnkZ9+vQ|$JQ^t(*8Q$UqUR)^1wq& z?M@zAUYuan1%opjE$#t2vQqVpvV>>d|0tJLVAH#q0Dcdo|>h}ky!)giT-qasWN_Ai@qG?*8;aIJrMd1F_XdSpQ z`x3R@ZNnP z8$FioD22t7g({!fQ*FnSg{svz^c`3&R0@`!s{TVPS;!PCjtT4_ggB!pFr1ZoZvr3HR3K2c zUrWJU$=NJ6-WSpR`7= zWYz*zC~nAb!5xq5oh`pXDuI5iS59kxB*L%nKa=4pGiUFz%xk{3n%8U@UmGatw)SD@ z&BV5t5!ywxYo1GkPe)$#uueZzIiIP{?6ahI zru2eP0n$q|be~}=Gk#R~$r$<-WuM66v)Obg#Zo#$eHPVzw(rPgoVK)thXk}21ifhr zEjF@sB8fY05PE7(Tt()mWt;#NhOD)b*c)8whUe;Z2hw6$x95ru&|-8R5%G4V?3UvK z3R_V3v`L~D7r9zuq8xbQgwu=t>_Fu|z9mk-&+<#$6do>)c~fdNXGoP{+G)n44m)s;WBMOf4`jsoDCs@HOXp#jly9G0OlLa zDUq(}?}^8_f>i8tRMHVFu`=(XF$p2=;?8(jZK%$AyqVm+j0l_0}Mji^h8A5Ex{@Ds&9c2{*9HTy!BHFB?KR+Z;OS?H=JTr?q8827jGMe0ODQ-Zro zJpp)%-O|6={RE)X$jtq-_a~Jcxy=glMqA0zv$BYzcw`I;PZjvZn?#e)d$2-$&t2g4 z(UrFHD@!XmDjis3Uxt#S`h_L1CltNcu^$KPznoO`P+MM?puQQLiZ!fFU@O?eS6iE- zY2kx?BB{cvwyP?v2?ZndF_4|(SsZ8R5`Jp6bKG>Wf~$f^;uCEspZ!fTeiDms0CWn; za(JEy^mPQ(@~GF7p4_O`Ud|32+0V<}%RrjNgCF}m%{c=$B%yim&C-**_f4in({c5N z6P+$SJ1(`L%g+=_6(UQ`j@um%z%_~hlU;9#!7PD9$ZEPC*w4Po3u&B&t1C4-*6!z7 z+?u^hkM&acQ*sG7SNKf{^TdzG`3qJ1)sNOmn`k629gzYzRC?4NoED_gz_7Yh*9L~$ zi@}5y@4rd{!}UE_!Ekf_dA*DsHn1gmvh!9fNPtbSoK2(Ew0)BWQ>@MB4+P%l&# zZ;69S3OZDavl@U75J;R%sN^Q1L$!EHbf`X>hz|94RMURCbNTh%>Jvy=T7Mb$l9w!v+M{d|wzNTXda z!hY}i;jnn2Ujb-#Rm}mq;n~FOH!HgXg&}VZ;wiff=DqBycomr{Hn)fUW$`@O`u4D( z(=Vw)&~&*9VA+L^nFag3E?cMZ4S|lWX#^^TC(3|u`Pu=+RPJ0|gExCVr?T-af!X*MD75)QINwu!XVqv?-q0Q?t~#Fp5o$bPfSZl*Dz z^}&VlCcHdy(GmzR)ngy^Z1(yd72>XB)8AM9P~>pif?AQNbmnK|h;6gi_QIQJcE>%p zvAsQkN6V^I_1O4++W10r09FQs99Iq)G_PEieVi;`Z4}@sij&nxvv{KPjoE&1S|jwM zWHrOn`SNL&f#3K*@uG2tdnvQoiM!Ws)GkRL?PSOZ_n5>X29<;U` zdT1hiwJfF_Tw}`}c;0$(Vd3a8v)0jRYeDe-0Z}xWPR@^b%mePlV1tLeV9&DaPnZ@F zY!u?f4M8;++-isE1gB&1&Z+y(4<@7GvGKWcqsi#fXe4~1!(NhOKi2YKduxP{EAg{E zHYDDWwa3@SKgePpj%Q1egcUP7ea|$CdVo3=`?)TAF|i(Y&ki;Q6B#N`+#UP*dLyq2 zVlYvxdW_HrsS8KFLMko^7|eMxcBw7ugJY|l*PlnkRL8m$t$%>UB?*|2T9R-(i{E>? zr*I8~DUO770#8l1caB$`=7QQ*sRRm`kUCz0X+rSt0YkN*))jo3{F|VE_su2LDivyE z-=H^7Mb^)vreZE9r4mr8pp6pz$`jcBmybp}XNKcV!0u07;9Ns$B-2P7sSTR5xTqB? z%8^>+b&gHbzX#XxAY^0Ye1H#+1XLygr5o?zvxLDX-xoLrJ&k964aSclmh4YqJ@99q(7 z&klo#bL^KQ2+>l^#unzui@RO%;$Fl~^lwVy#Ti==oVfq#vK!Jhs`zln!cZ_@jfD#r zd_m5xATO)$tz;B0Qb_NxZ;^<;&VQrPrDGZ^Qcp0F7IBlLc@m=>HuQk7 z@j5c@*DFG>2Ycb8H0?sExNFgrb!}Bu!V^A*%D%9YafsWXI=q{_w45qx_n8iR^I|;N zNXxQDPMhVb5hOn)PmKVlagG|ndP;s80pjL z$@UIRpIhajod1p0bxGW5j4XRr@dH?CU4BVighstk3)9){Ww4Xy#_vS(JS=;8iqBqv zIvTK8bkht*Bn9pSemmZ+B5FwC>wZgz;m_BrTD-J?`ns(|%UZ4wH@gWQrK{+0%WF$; zUAxYi}i&wdlnky@E|md_+Dt;4U_#qWBT_pbD8#6YVt ztWZ=i{Xlr*QAAVLkNmM-hmg|A}_Ml%Ga>>~#FI^-^AT$(E z0pC}28U7%=TL6c3EKL2>p-?e6ft8=RbB=th@&HkmB-_9bm`Jop72M%OM^FyM{mM-w5RE6RG&$y_dOxYKKleulV!y~M;<0L! zwSenTNNN1`+2Q!L2gBWOf2)Ymd^Fqr()cy(P5z!^thq|v2UE!2!E=^MIkp@c<1HAP zxe3=IVT|S~5QP9f@R)lj*^boNTN9b{x$tYsw@Yh(iOYEbH8P&7xK4+*oz7G`d`+uG z;KkFdre*)-=$^FgxX$lu*iU!anximcUxDTI=~MyOA1t=~`YLLX3E{2LHXv$UBE-8` zb9B`ScxXFPWqP&NY6?;ZIvDWYBdQku&XjAUlttqT%>~t0A{|?csUQ%v^v0-C3NA}X zuUtNd$r7S`39Tb{sVmM9^}VAPFs!9W+{Kh@Z~^?K8yH{$P?ydUrRW^9g)PEr(QPvg zP;sDOTFIObIE+-daEuBG=~_#Q@c6~BWG{^qmVy~I%*_wPlm(T@N!$vxS0C-BuzDdj z)R<1FbnHx59exE*dZq}$xa*ek86ivOMOeE)t6XQYJs6%H>`sqY7nHA`+`DV$tWJOw z?{Ms`U1yett-bOyGlNYLW_dv^#qnh(Yj>PflZ(S(aF~>uZo^>B?oOQ?8wbl+90see z){4PU8Zi+-i^O1R_0|)st;ZE%v+5TKrN#ut`WLc}|9ChGg6IXxhR-p6xFXZ~y+TC{ z5K&99s#PB;EybvW02v2F)Osm^i27J05Yd01_n)+|YM3GlpRf-{oxyrRx>`(QTCSM@ z76Vm_v6>dmU2^9M9(fL{wbPu#>T}WOu>L}MO%Hdx*h8BAX;dxbhRM0|9B!lH8BC4V zfo=GXnq%)%}wU}~} zLVv0J|H8X8S8*DP8}1JOAouAT4aM77?E1m!$t9fzAd*?@cJ_9tN~8X{voOWTlWi@HoC41 zcnZ}yoy+wdtI@loJ}y&i4w@1;dmDvm?bi8&5p}Xz{S+jwt+qKs4N1g9ABgH9*0!iH zkvBG!ZQHc9D~YgQc0au7GFuDFFje&+m%z9}B@dW&;<`-ZD-VNoNcG7>dnA?DrAJ$p zB&MG@RMI?*M^2UgJ8g?WuO~TD=~`2&_i+5}nxf}2V(VFCc75nenQ$R6U9;j0GW$Iw zdLaXc%BBd4o4^8TT&8+n)3^4&zd21`V{%32iDVg64@=s+5%3tD+(iLiTK~zIdK5 z;x9kDmn_aD24jmNOt~o#h$)iG`F!jC2h%B^^tQdV3YR!{;2!5qgNvidiS5bes=j2( z%X5c{rc9Lrn^aYl;IqUDo7dF;k_Tb)^_I|aT@^YSi+ffK3U3O~-&dDVU)7?@dHD2QFEb*esqu2VLus!Hcrxv3y3r z(%t@kw(rPg<#h+NVvE(!*eBhU?l6Hh_!uspci87SLnDKsi}N~8y!M2M{k8jbjNqL^ zWP`_BHk;_{I_J|CVZX~7cg);5?T|((${e@Q+Kj4E5*rS$cX$Flr@Yu_-4A+%=}^%T zG)J3UteMUSG@RM}&DmsFB10P-4_AZA#35<9^V5zt2%moQc+AerVrKWVclgQUF}u(m zStD5+f+7Kr&)(n?x1X>4 z8C~ktLQLR@7XGhzZfHT$ob*i-VVf~U*_Rt!pd#fjqfpBZtVd=8)Snh=aXuvG6{ih< zvId6vkiVDMm)x(3h3-e4JQ`k8D?;d)#<&edNccc>;?)@`>}M=9KA%(7<;mn@-gYcp zkBmhtR=fYbs_}OhyQBY7pT?i}Z#DPAf1_VB&`|{cC{cl!n+-2$uC^sQXr@M#1@W<+ zWTvBcV@-3={Ulez8C8K;>BSmV+tM2{1fzvWhhX$KJ&P)F4AcjR3dBkS>oT!v6o{Ww z+!Tsxbek2vKihRC3fSx`)FMy`3Q2v*lx#s#Vpm_ZdJ78O&`PB~Ab8klsU9!0eZEmsGW6<% zf>!xLpmcQxu~@4koBavIA#v!0Iz0O{VV%KrWw5w?l~+@{9#?C|Q@SPUIfQKYcJ@J{ zG);J430%dG$N7}Q{p!ae*|BNiXr+05oFo-+7!~+e7ctwwf9~sJyjQ~BM+E-sy9o#X z(KpS=K)>37HTq=*{pwc>Q=kba^>1bc{a+(Oyn4y9Bu`R&u!SoctgtenDAACV?#ao7MCYJCKVTb@TBKdvUyV$cE7UgUNWXb#{c; zE!#_z3v_hQHzy_jEgOPo=q-ek=F`I6m6xfi_1)~w{_TlX>z3t+_*MggF)qjcqQ_2A z>mH1?-@AS|EKY`&;b!-`IvEYmpXE0r&WyK)CJBL8=h>liIEYFT4Uc+VNo?C8Ro=|y-nultNk1fQTS zGWN|yb{Dk^9J{zR*c_idH62gSzN~+F?T-a{N-6>}J+=ktOuyFTa&Nl}p=#DfI=>_q zPW3n4 zKq&Gl9M{|>l{R+SBR!T-7Pbe1a~s>+6L^=fO8SqD@6Q#($dBsE`1T@FApr`tDE8$H znSY}tLtALBKA+jgojfB{uF?iNGhFEZ$WZ@1J>D4g7x};G(WjJkHaFLTauS?yfdh0s z4XnQDcxPj@%b_cSQHJD5ZJtxi7Z#2lW6)b7k5kpOM$RuP-%n2PPkRs-#+XO=i5G8_ zy|KqS%r%M}={(APmt)X}$HwQ*jV2>!lc@&YVuLm|UKv_d>|ZAos_w(VHMZR8FRUyAGwrBZ(`|jtf_AV>Bj$efD54V;~YFx^fIhKIw>PSL|ei3s%TWa|a)}aI?%_(My?&=s<{-ZSgbl=&RJ;?cBuND`5lPgc$sH`I`|ge#lAnkFjOvA^%IThg6vb2Qk(*`{+T1i(b( zgnOV;^u}iFOw-z4h&oc;m7(IQ-p+38zl4aZYHa5JviIgumL5l)Xy2+x* zU;e(#OsA^ewX9y9a=eYg^in9+&nyJgrO5|!1*f0rGkU8|w7j4p`i{gyR9BRy9)0gk zy9`p%NzzOL{B#*8eXn@}eXiqt@kT+#^qhJ0fvxN?`oo*3_H!XU!zFpBJM!o%QtwT| z?AvH1R*oJ)D3Ph7rAT3l(KcqY&VpLy7oALtUTR9w1p826wx;ixDX7VLQwwz{-irB} z%?0JccSOCF)&zf~Dio3w9iI+DQu*p9m(!wTM-AG>bQp8`8#StAkSykFN<$T3Cav;HEJePLX*4@CvzhW{^~Ub&BR5lO zXVN-XB36)o*`EaJw>#e3r5jB7*8biNeK^%>Hko9bHmesavD+PX4@V#GacOh1(HrnK zV5;xTkC-so=XZc^1=>5`1+6UaDy0+P-*=$%nAp>d6|=#U=u zgB|a+-01k!JpiT9!6yTHvMPYmtMc1e4Jtk`k#nM9?)mOOX9s(w-#AzK1?uC@-`+iY z`Xkl{v_2DK`RW%UjItJ?i$makOUcPo5XwWD9@Ho4wJR_t$|&2DVo*xyShGyG-$ z72K#NCiRS*Fg_isS69S~6J9-3SrIGGqy>WdLdX=6ILP)%?P z=3QE-4@IB{AG&eULOW9Y5T9!YO3+!CzIIveE^(pRCqCw3Gt4+Cx!;BV`I# zhfIE55pNDlqvFLx8T!b$6eFC^U9M0!!Mu~7Zy!IgdW5$~V8{tiefE;NByRTZv{S5| z$~owMy$0@{b^eh6XP0CJ>4urS^lNZux2A{F&1-x6*QbZwQ{9u7Uh(>S@?S) zhM=B21{bG?7tPW?|9kgT@mV^@dv`OZV8^2QYkBTaXAkwx?!k?{gW2Yl?$QfsDVPms z?H8x|7f<+GpZc+oN_4C*sgxqNYbK&E_Qc^}wsTE8UKH>k-2f2LAkp}#;$^8?2(#Fq zAX8cWY)4&SS7hC1n#ZW=$Z++-Nz%NPPpsZE{g?~A@|4SypYW{H`eoCn^q+#VKk#Pd`A8cA9&Q(x4D^z-9nX!EOsVqhE*R~(Yf}J0EM-hv8tPs* z-TKVUC-Zh&{^O?pN5_nX(p-lyU!cHgq*x0b%i#(xXIPux);zK1Cl{4Z*{)+1xlZ+u zEOIAontW3_B*J8>>~WW>ky1RL4dyAw_f&>Yhn zaUfasn|9xh%)Wr>;a7q|gw2HNhOE~ilJ$~!RY+;NgF&8qyf-;rP9!%M>YlAMoNeooB1G|V(RcrW5q+2y@!hu5c1 zrpl123e`7@*Vhe4fc?AJG{|=E0-I0eS)xd(Z%-E>X9rKXl*40wT*9Q^2#y_ey}Sup z(^SE6mq*gIa`hdiiBw75w-A-x9cbY9Gr{XWW1Mq(PlBKeUdj{9W$o_Z>4WRzOQ@j# zox7@|@n*hz7Yi{eycsfsAsc(@$HenhwL4NS@^Yhqm4;sGZJ7%<@SOtBn?kfSGP~kS zI5czCN9+*fDQ{Na)q5c3gU}rUH>_m8u>GfwRs1@mFJ4lgF54X{K}Ny5nRXnT&t>p? zBAmlZes)XFw#tM&M=cn-<2KK=0m|Wd)s|QH!n5?2*NOA05_Wzhr$}wBAeBnyLUyBk z%#a!lW6YQ6ypU2x{!o(qk0JAyPJ6g~V&UV^`R|O2qB%V^Rbj-8h8|KfWo-N4*UIND zsk`DDN5xLZc&5WFYZ_l{K}%c6@q0++hlCCP*qqG92si~2xEy;SDC zK0d7Q8N&)>Pftv~D*c{^F#GEEgS;ve?F8#0vcu*po>_fhNt%Dg1CU+9-jj$L=3Zow zS=kswxvHO7Qs+Z6xS|WA`(RJlLg*n(L~4-AelugJZxzpz#`*@}ju?qvRD!Y+Y+l6A zpDT5e1)FEJxbbPmS7zo58Z%H}r_gJ6GnxL2hRUL=o^}2ixymBnnDZ+v2vl|=J0MNy zy|KMoVsmwzQ7~pwr|!7SBt+Jon#Ua(t8<)z;B$#Jzss77#Y^>2azF(ZNqomBH7|=Wb(^ z&6Lv6Y_3u77eipFy*8bf_V%n=y#vcHrP#uW~HD1eK!)67FfzHhw9sqYcPe-i;cxLE{Z z#yAgRPYC^u0br52&_9HtCYP4#Ys6sRSX)7#aIkN_Dv!K0mF#a9CwtqTQm>wnpsDYY z$E5(EE~u}G%Y~Dc@vfS*Sejt73#R9AM+|OQA|n=dXz)#cn1A)x(f@p9OhQ21v4 z=}M}FTtqXqr0v*Pfdsr&F?Gz_W93UhtPBypJ+_uAM}A(s)VIS?`e#0Hvm`+SlJ21z zJE}9w8iD7vm7Ighc2s5Wj{QhSO4C&nc0NU63xBe&(W(%x4u45$v5;`UYPOgEuy9_u z?`uac4~6+>PuOc}MiubhD+1xc*tv_|!^J{n&0wRQXh2D)BE&+a^^NfsvV~@Vt4cK= zgo=gSu#(Q(7ElMJ-aJu%E8gsW{E6BAP3C5Ig7UMj1)SSJ1}*cQe;i28*1Ubj^*oSa zucu;*zH3F1u$>j--UL2oQz5}%^R^UROXjx}8caCjGz#fXf*F?V7z<&OaBWY!rH~-} z%|OrxQJ6r~f04NXp)`^gmPuBBLApGUSQDErFkl}bFPJ<#du-M402KV+)APSj0?AwhZ1$pFFkdNyPWJ~QgRwD3arH3McnfSA#>!+tH@!Igm-AwyM4*Jlwsv zgu`lQ=)mbM&5;}49-5uWORI{2%&{*%I9`};Z-Hl;P!|fnG_9tBp9YuLYuOtvvR$e_ zTvCm-*ILOuN5;akn)*eP!~;e;W!v6c&DDLE)oZ%1_FvuS%^cUM>^a3P+}=S@KOIFm zW;vUSH!#W(2gf2DPlh#nB|9=DtPdc>kLE^c`{O7_GX`MqYk_h!4=jLv!qB()!pomT zksik7ixsGEHmBYNYMTHe?De~pWM&`?O>^#@vc)MI) zr(m8N*Yg0i0MUOu8b9U6HvoEy$Y$_766o6wXd_vl(4XIU;rVuT_WYUb<=xX-G)bR8 zqkcclJfm}{g685kOCLUXeE+c5EZk%eJAQRsjDn?~Qw(Yzve?xzUU((AMk?+U-LH5% zBKj!{(*BBL&0%J_-!ieE`95@Jrj|C`InO8@oYp z&;kfH4~z!E^>1R9f7rhJuScOmYiLE@CT=P*=+GDkCV<5>M4yZyaO2RSF+L?aG+!Et z4(<2koA#!g*5@BzZ>5X>*P`|ntc4airB2O=-34PFn40arSgg;okAP0b@-ic2851*!&?A4sP-1c~>FHbfU*Y~!%r}OK{F79ph z(v5G@0V!wD*5evH^6Fcb)WtNlk)va98S$K|G2DLPtkD~HI&CK#S2{%3hSGb@uSpXj zee>pHv(HcASZ#}Wt@L|0*b#%67JBL#sI?c&t!TEy(fCjt&?{X?!k>^hLmWQ(x z>ZQ-elwB{ad3i#RQLgE2+6El>6uXq^$>;<20jdn1&0oH~v$LANGX<0a2lLPYJAz_T z?_HMf+~UFMp3FkjOFMOW&Z7tzz2JG-BC=;GwY|GFdveoUG1791tDiclE~lCB?!CD? zy}o_*^5M3=ExCJp?Vr6w3$DB|row+3RHm{kkSxSgK1+#0>Ou4%y6zVRN7dIasaK~- zp^MWex3BA)M{fR;X@u`uPKCJ3*mQ3#|I%xtY(z&oafXj<@65UxD_gGAlraQ}%}*7LOXV!D&HJaJ{g z{5ChC_u>UC}S1JZ-F z^=vS<#p9C6kJaZdtF6QyfqQ(I{-Blg@LXF;ELf?eYxSMGFIZBqO*{?l=Ax69c>DiMw2GIjjvUeK6<4p{63kr{}cdy#97VoN8LHLQC$68u5X5M$Rx{ z>PYIhv$FS|uJn8d!t~C|b^SihBF>MdlZq~pty3;X5Vnbt5hl4&d)Qzvw$;z~WRIUPxd z+;oq++{@p8|2BQQ$mK%Co2RKGk$0lrwZu~|2nK1ve1r(mv>)|H0cTiMt1ARWl=o9@!uIDF|Q zlE0DTuD&!t5M71YaUqc|?oCD)_mH~OeK1NF=UB(r#Qn{Zx-Wi2jUMh^84Au!2DEVE z7v$gpBG_eG>r<7>@=A;2qy8~T=Cv%hu+w2Gi%xdRzT(0dJBt zXJS&r)@%ZuW8MV5c3FOuCLEMfyA~{2w^>y#c;v@W!G;xzLpymlrDsQzx}HhvJ5Pp- zjchDy;`pTtnJ+@61N|W~V zYiwT;@k&l3CRU3UEE}`mSu4qK-IV=8GjPP(VTq&>7Lw@M3v0oW3F|GOJxkU%5=>r| z<}4VpEMk$VK9ppw*w`|NcvCmyugmk6D}7YzT4R-fSZVf?Nfu zm-L*@&QkK6nA1?HsUJ#0!(z%ZSQqB5p97T+`tr#HL)B1~CBQDZqAWpEEYxJF^b9`) zT*cHc55$Z-Cd!^2nOr+ml~TH)&MGMbe9QiQUS`L7v!UHzqh8&;7(~0*k8!mDRX^WR z=XsR+*8biNeS*_!*2=TrSiM*YcN#~Q^?RR8QR~$=O7ezFAevk(s=;Bd=keO zEgenfi9RAp=!t&%G1#s8u8yLwuQyzLl>v3TL7^qtpMuYNQXgQ`MGafQU2{xA88wi- z$Z%E}8FU&G*`3rsgpYM8u(nF;hz8VqF{e1+KAYXxys^J`B~?#WKeQxmn>N*})2TuS zw-Tr`;W#gUt5k^wfDR|S@=SFm5B=8mb^XS|3r}RP*oAw2knLUBvk`?^$W)FxH z#=^8W9Te`R0lb+Z&FvJ(Zms&7nCi)A1XuDG+}HCnazm%4a!1SNNPS*FTqpBQ{b|Q@ zx3=vnZd)%_0T;HX2l~LT`}HcVd)E0!FrF-z$)w5F{b#GQThqho=C!^3>(j&TsqV>3 zuXz1E`7fhK0DnIp8RY&Fw}^)hW5oqbNJ!~Ceg1HKGou7tVt;4-3U>d!(?TuhqF~J6w>e^>$z*ud4xxboR04-2*_G^vq;gm z+-GYQ-|T>SI>&TyQws@r??qs0!4J3f~ncHod%YvpVTA!)u%a-yfcXUcb^6foij&58Jxz{0j+ z>Lmy?S$|au2azyJFRy$xN5xWi`2t!8dTE(r-SyesODn8dAt9T6zT0%C`geX{Kn1|O zbQT#!?=f4H5nx7Fc@efhRXabf!L#TC4g&`~FN_iiS*|4}c-f0#(N!9xfwd8?0rG4@ zj4`M{PWfA*F#8Cn!s?Tut;Q`v>r*5u&{ulK1i^XNE#_I~(`RoKh#xFX;nggiuX|VX(TjAO@@18!RNE!CH3LX8;B>)`K$^w#OBav*tnh zpoTQY`Zu(C{Kxx)Ac!@v+VHtIn_P+0`dO_a1c(?@z+t5;gWAurlM%vkW#>=_^dn<$ z3Ls*>ED(s;-_KuEB-@DLHEMtZ125|5g63drkY$V0q?YeY03bn)F<@%xc{py)Ag5oG zekyDpHnyjkht1ao&%^c`ir4gvgDdrXii@fG-XJdIzWr;(Jj|Ek{n+kTyexqfPwx=~ zEI^FCnph~-zyU>KM#Q8@$Xdpjl2USar}~m5b+>#62w0|S)$w6z-PWWqtc{ub_2Ly)LR{qwkdg zW>CGSbNRjpOnNd5JoL_JlkZEBS_gIgV8FU`etddKO55*09b_}z)D{WqXM*++Yc3j0 z#Kwlzwk>a!?1e|_%lsE_`qCCq8O9F1j*GCbcs=C}adxVQ?`0ZpJRH$ss;^uVM#W@q zTzJ$w&BnIaqch?6YBC z2#g=ud=7H=dy03C>|8$LqWP= z(coqC3f=?i{_cI8SMY!guDg(ZdvP*^xBID|>v-R-2cB0>#zoK^8E9P-+8iW(e7ID= zNlVa7Ak~)#bs|FS*|fg*XIW(i^)_TpDngOi)H`Urpa;bDG6Kn#~R(dy9y$fg^!?JZW zjO&M>@MZw5i*3Gd%P)HRJ$v<+{*zrt6h85eFdHykhN10I+H1WK26_phfbVQohw*)WO^eXv ziedvQe*t4{TSi8gV6Vo|5S3iA{F8|V76COGQqfA`;x zA@*E%Hhp=^^(Oindat&SJqzqR45bD=H}kmO47a4s8N*Q~PMnPIpi9pzmc5`x{U`s0 z9kELHYK4r4KJKl#j>^`qoxalvcAQX_Zc9U91#VNupr@ziD!@ z%THd*@i&>jOdmQu;}_F@3uBT7H#p-A)>}SFBH%U2Z9%`6X!3L%Y2et@(a(TfD-8$3 zvc9W;NGOwCqnp}V-4ilItN*9>QYj(np7ibADq?lmW=|mlVx#PD3i@W0)UngqQ>_M4 ze|0`A-iyRoF?H$*Yz%ryK{j4$=_(n3Kla}yCiTb6JQiN39*BI5(a#Bc?tVIOB6v^76^z6^$#pb)B5yPB zl=^TYf;dZ?jL=-g7CLRys`(w)`q<87sbjqXmpLsH-l+&;GcI6P#g;e85R46CIs{|C zQ}?F!6sS)nB8bfhT&xy04T5;odr*pM>^7_ZestG4vPJ)C>JXR(N)mp_jchLo#csY4 ztQOnJ39>h0m7;ztA^OX!q_H$f8~1Es(n71Sajp}0am$jwPCe7K-2LIxv^j&yV^dYlq2>i)bAwZ8QfCV8Fy=0)u~G_ z8`hR)X|uagFX-UdtDjA@Ov|gU0IvC8@6S#-JY)Xai!SW6aI_rxxVwx^drb!Z&5M{A z_$Qa{_=bUOm*XcBf&aQUAoy?TWwKPD-*^Cazg9uNd33R(ktF@=-)RYnphM8ec~$*& zBE)M=t|<8=p2ZA6IWgPPT zt3*3Z)>exE+e_=M4E7KQ+_0!IV=|c5E)Gopj>oSjf`G<{GN+S6K=XitMsBBP2gMgI z2la~`_1f_4s5XS&G2P#u?p~c8EpM5LBGh+AE&LIO^fmN=;G}sy=hDjC3^DzvdQ10B zk;F7&PTt>YEO1Cls2^BXms5|uE`aXx-CJ9|hov;!ysxYKv#lqtW^YDpZtrewU!5Jq ztf=O{z06cSuoZ@1Uf|f_r^|W7+dtgGkm);syQlLzzb@+AfZNE3|9{}r4=$?bf zQxWJ)J^xc$Oe^9n(yv%n0mb*JET*S8P6Id#;o!B*WM2Psgp=)GsZ^CW#hm-29sY~GAo4eEN+gC3iZXaHKTleRInk+ddWG8&QF?IS~oIbgI{fXfT!Z zrPppl=(RWCM^OX8ssB^UDuNb<8^N`my}f;XFJZapzjyl`ZS}C^Q+1hq!RKNg!Pu)m zX<_Bx*pf+0G?!oBa*bDmF)xOW+oAbyE%3jOZ12pvC$s;iN1s)!TPt%LX0O1BoQ794 zbbUmt`VO~m?9BAF(dU-4qYRTHHU25)Cr;dTulwR=*o93>?=2`i5l`!hpJIZ$yIkEFt`CS}?zIE^RwQIBenRk;3m%t=^XZANc>NDaIDdicc zl_8k{+;N%q=iSR{w>96uy<<#8({l3kS|7p|CH49OmwXE=Y**h?TL)LCJJbDYW%Pyw z<51rqF54qL-8{S6!?^8&I^EG1MV-_y>fgm``05Xr)LUB1VLb;H?@zWp0O`R7EAU1s z;Kd8-f3z8A{cdRX3ig#A|FEy#y`j6#mk;-D++V6>dPkJ;g1~E(o1iL@2}mcno__LI zgj=nmUVGoJb0-;$`MTv6n~8air{&@7EqG4%9-bcYMt~XF?CiJpCh^81yS;SHG?48j zKu##rldWkz^|uQAGr-p_x7dhs5nXW%M|;u9yIt=_Yj6RCysli5cy%M@4Ysmyo8s%?ghYTo5jo;L&?@wfGkm4#mQk9%Nxh-_x_*CgRjNUmMO?^kyTWL-3N2)?0;pbvvRhtTe zB-THO!dVzMh<>G__e6oL|)W>S^sV&W9vP~V^AHNZxkthzIS{L+4V zt=Lm&4eeK(0Zt$ZT*%;OFWn{(1K9`wYR(pVQ_^X+QgM=UmU2hp=(&?6^+$y(&&b2G^N(^}C7LR7ba&G=Z!5wo=HkJEskqW(5N}>L(IY?+e#z zemHR@HguXvF<7qAOxld|fHQY;{d>2+0~{MYE{}~)b{wHSO}RzAT~&}tt6V0QnwjmG zH>)>xUmv+0Q#+H^xe_s}Dp@ zs9s$WFHU&%5ZNJlCnaVVLZ*nsLAFn7PYVvtICq?Si7G!fkg)VmeJBDw7&nbRi3=5G zy9cw?3)|BJjWJR=roL%Oz0#r=S)ChK%5o?aZ^cVp@oAvr6W(~d0H1^yRh#dqZ%Y}T zzU|8iV;&wP66=`M_l8nwW;%Yj^@LN%>VvrGKC^?B`N5-WX;e75Y+E)J86!T7gzJ2_}n(uqrL#XFp1P=+R`s^ij zN!;w+X{T5_m2=SjdJWt?>--}D&MwIcUUy!3wmQ2tJ)CY{+uOfBJ?x(9p1kym*WZ)> zGI9j;DfJ5>s3(uX#p&Tiv-HpZ-aS=(md^3s-OMT2u?W4E=MHuDP;c=b+}JyqZBDyO zFFe8MV6|VmqJPP=`~71fmFQSsQYl4j*GxoT?1{s{Z0DMGyeQy7x&c^e_iiwurPHUgq3`Iw z+@}BBrG8g(yL=uzThxOh2_D#e$87)bq1pcRCl06CsleSk8XxtBoQiV3di93>z6$#+ z6MM)f$@r-MAZfU;f5xirnMbOI*5jieBI5+=KiITJoU4d+R;Ucf2EWXL{I%^zvS8;& z-cg`j{qT~sU6hknwbM{cTW2p!^}%xDxeJbf8Jym#EW`RMFKpDvbJRzl@8&gPB9w)R z%&@FaO}9RC^GV%v^QM*kjO=6`mqtQtdOk(|F5+iG%%novfi<-nr|4g5Zn8cup(6_1a zcjo8X3Kn{@y}0NlYZ|<{%cPw8poBfGm9yb3^2Kp=2Fsa8AK04h>a$XBX4Y((!($1= zO3x=aJkFr;Q{^A^KFK~L;fHPnx8&l5p5sYNdw9PpIROr~S+s)}*eB})EnsC}lPfNT z*9QXb`psx`{+0!7WMl76I59wTOmD=2WYurleLGH|I{djq{Yo&1u$fTZaF!^}g45yQ z@dE3B9}~I-Ij_>JJX4*?M`mmLx;|Ra#b!T}BI|Mr#jd;qv(6nS>C&vqKl>f|>(yz{ znwfXz^prm*W?mX*njO3sajNX{-nGN)(d@G*xii<&ku)Tz!XWB2`i+qZz8aR6i5E z{xilor}rcXy5OZe!Ccnv4xT=^KE8ws`ro;$IvQ{0yLYh=LxMM%8++=<#Pe0PJ5nz4 za-)Ei26N5idH0u=MrQl9!l9Y7K4OO;PkFQYuHFMNAB64@xM4*=2DJf{?*h;Vy{Ru= zQlBo{9V$Ua!MmAu9GlN&@OvVhQ?<)PkT}~a6Y?CjVCat9JlDoo4k5ej8jO?&dvq^6 zOYg3oIIliZ{a+`$Q9fo!jfOGiOLSgHDI$I-cg980oSvGhFk(hS z52=k^nl+F5yd`y4T;r(N=@`#+n2)w){C!6~C<{Y=B2Oih^=YD*o*3B>0NkO+zHB4=hRZ&v*c` zOW1o7QN!Ge404MHx&OqHIv<+B64TY^xB=|hX{0K_x@M@MMIy_RnI#A z469+vwgR)GWE+p9&Mvx1Kri|Ae!luey(z9s4vs;Cc&`T7Enl-s^pD_kx9Zgi4K zKuGUp90J?5WnOdI5I?pcH8l+|fbw>2FGVTyiqq%RVqh{;x|vb^Vo48_|6H)FR1W0hKn!`fuCU5L$Tg z^&dJ?kY|k3e<}7GLLqEfo!(ni@)=h@ZS`g%Hp}+1_lq&Eao>tOGZq4t4 zxog}?6($O&1fW}EeL3)U+dDLbdJLrK))`{DK&~KwYhO; zdULkFd2ZFda$VF%KBG!I%{&&g@nA!EIs)G##`h)ydT_G{#Efws#GVk_8w0>1YjtFOLG9+v`Px~RRT9v4np#=C0LVu^mu zHkY2mnZ*2EWIbgd4piIQbz>GQ;jVUbg`*2EX|z}`l$mIyj}@k`*VknVyMr;ZSjez+ z(|t|@H(#9U)t{dhuXl|Jg>UAcu4YrMWr*MOgacNyz5IuT^TLf?J92p_%s+d=UQ;uQg!kT4HONUgLJd^b%#n_ExEiRm zzA@fHDVZ7Is#47dp=uyEtfnOElH*u4(BF!8vmbwAwtthko1LKiLe66#)7%77Y2W$h zedMdrx6inq2U6(uROI;)5}`~oBy4BJxHo~1*)>Qo*t|Uh*AjBiz<<(&NrpO?&OFy5 zdqYRqBwX8tdj<*OTlX>xPpbz}Xs9l%gRH=a6cc`1HFMUhs`qJ}b&Iv+$Wu&MsMc~} zf+dMAQBjPuOQZ>^eiS%Tc~Ty!44g143?;V1Cw+Mkt9t1^AUw1-?ZVbTMSf)R`~SGwH`y;@UeeJf}RmX899ttB$V8ZZB28+)*#)nc9Gn zjWnF}Bwn)w6SU^rCPjBEWnVH5uaI*y4O=qN@`Qp5z^#VdwWmgNk=D#Bw zyT;4q!e-L@c&v!*OXEg6m7tWqdWqBS>witL70QiE7U>pS+tvfEK56$zzX2S z7(ExQ^+JNt=j9?VuhqX=;s7#JV2(I~;j8&I?R1l4m3CY$fe6{e=jrR!Uq~}M@3I;h z-C8avWPLM#UZ@D3Mx-70OLQmT6jO>!foZk zyK{c&sZ-pw2>rTewxJ~)kh;a-x@2K1DM$lMr<1W>v}$R*9Kn1@KB!GOmE^Z|zaf`- zW%>p!b=1275iHlZFzBDR?56mY`A~4s==e~$-t?vXE9$2PpHa7-za@LY~i>$yiq?2F50X{Ti(lxI!O?>BP zBAgEcntNu*%(^PL7iBR|;cAsgnGU?6(_ng2)`g4RLMtO%f`yi+2hqjnMZo-;;XC(? z695yC69PT?W^Y?_p~JG$?=uNUc;5P>#Rqx|U)U=g`@zWu<28jD=sc#t2+?6Cdv!I@ zg`tGM6Qf$ky^GZvMd+bv+NM?X*ejN*LnY{T>feiJImdZDYa_2Q(0$VR3LU?uZkGMq zU=E`fWiOR?B+$9ULR8reUcwP~oNmA-;iTt$RR)|Qi<1tANwzw!`sWQ?b%%Q9`P=O~ z+)JlBN*|OGZkx`GYB~Hb=a;q?23_B~o^KsrpG|iu%H95kWoMf;n4qGnUer-PCC1hD zam~Ds8{RyGO}TMgc`fb?fK%WbA8hMQXSD=ccsA;R&?lmYmwW}q;xU&=u(}L1I|ZPj zGgvQmFadsB4o>^s5T3~gAu~I>Mtdn`jjg=KTmgQ+y03c=cm=rMR@c`|?;g;OvnAVV zKF9!|O^8{^W}TfKA28I`b`Z`o6*e>zw}KIU;AuwwAMh$An9P5U)D%v1~ zN{{((h!@@2G?N>r=Z9ifso6MsV|>oFGpPQid2M9E2B*GX%-Rx~EdQHBmy9+HsN8{WK2a2dqd@@hzZL(s}CHyJ>bp|LCQ zWN58cXcAeRg_GW&54f%m7oXU4<8P$PKge#zPC^co4L(zn$e(Vinkrcht)= z0Oh{r1I48xbKf!D-=6MXozc=EdMeD`*uE8&QahasF!c=?5yT*vIA+`tk0v%Od5y%PbRqy!+#b>_sdL@X*JMV2w>-dg^p zm#UO{DF%jU3Z6Nyq?cM*QmOyS+=ze+>)OuV-oAb}uv{M9yZw&%I**t%%7G_mw|4sg zQU2=tqO!X|hO(#7s)H5nkS%8Q-BCE?k?oyX_hk0p7#vbXi^Z_gMY5v7 z$0J(1aJYSAXQr=$Ket@vVv3UovKQl)g=@Pzv5rRI^(0b|}+_M}M2Ob?P`uySI`hP3|XR(xHn-%L}^xlkc>ioQuQ zuS(#MmX4U2iD&z~`Gzmby$sZ&)y2VCwPo}ap1n$v@lWLEwfR13H*WFbyQ zS^ZHb35YQV6EyB?ju{owbem@^sn+{D>Lrq-&t8=*l@Xajq(Y=;9(`a-ze>4%c#}4c zUcQKAH|npL@2*q}WXE*As^*eW)!eHtbngRX1XIhTPO+L0j>xy{ zt#uItXd^(YWSSf8Cz97u%}GT5LALj3?=f{u)BDab^!^){L+8?(DkeUEV!DcqB$-d& zLJ30(+}s6L>N3SEb6M$=T3B!3H)k)g0w%Rv@rC)v%vU{N=S$^FTx;i@e*ha$=#MOd zXewS(haAr&52pCNWtrr`R!~jaGOu(A1nGs!B@mxv3YWm=7V4HjaFf-W9qZI18?S27 zC!ULT*6a{V6)mZKk8dQuClsAGe78y#kpRVhPpwR3$_)D@W+2+jGrWV)(wOX(3DL%{ z^$Jj~MAdI#!#fa3AZt3=QC-Wt%{cYmj#|O# z%uL`{es>Nj0lvA0X^7@eCy<|*&W=X-6uzY(mjyp1Q|&yZ|A^>s(w*$pHKI_RsEWip zf`RIZLRHOFEeLS*T`4c*jas$i-8gn}dkPs1lvtH>&P_E^%#oCqE=Vd~nsDk~jT0R+eMxZrKV#B&=56 z%)(U;!8s&p|#rieXCk_$rzo|(#GhB+9+pi+>l^>krKh86W|lUlBL=ZWmY;?(5e=>G~g`i(km z+;E&5t#9dmY2H$u_Rl@ zd6Hk>i9=v>H&m4doS3?tA1~cvjzbMnL^S>%Wmk;8_$@!b{z{MSeve8xn?%8t@P{Na zbz)oLF-JT}AaUmxvl|80m_w5H6zf&O()b1vIXUKD-P^r7JzUM3P4oKg(6&lq^6cV_ z7qKSN!np2f&9rJGU)N`L({k)6NasYrj5Kw(rt|Yo#O%2|cLAE;PRn0?F zeY%$Z4`qJkW4Yw20wfuQ%Gb_{*3&<7L$7bPcONTW6(un5Ne1TqN3xAyoE}~@hn44l z@180?<2YPj8}vSLUNTvH)BFMEcc`<6de8mf#@@kfbDF&~pbuT0zeMht>h|+zGxtt+ zPoDn5_jFcjzja0b7Cn8LfU5saHl=0&bXToH@Z5Qa_I78RZ`(VR$Htgb_apPZgL?tYOn*c$?Q%4mR- zNuLQ#<_Ca#$iaV?i{S% z(60_}Up?H`Z-B3s5y&Qv>hEh*i-&$bbF{$HSUo!#&Ev=JHajm9HR=xLc3wD6s2}L4 z?qfT9SEf6gyV>RO5B4A2J2>c`D*m@|fc(Zeva;-8#+!z|BLDw%)JHc>&Kgr`G?9Z< zD%TU+nthLn=!TkfM{^OKPkZFsko6uJIrX!!SAF9Zdq%e2yu^63rdGQz_ITsq3_Yb| z;itXO%nUR(1LFm2L*DKTW{5Vgy0Up#-{P;{X?%Ckxcw#U@z8i2AoFlN*6)_{n!#l= z*bk`tyZ3np8@iAzQ<%&>8VN~ukRdeABjfcevM+KE+}*vpw>8^e-I?B;?Qfo2ogN-$ zZ~N@+t{%*GuB|@)#BBd&)mn2Ks@k$6{$}IgYt*Z|7eNPupMp3_#$J619EGuo!h<1E zxUfAvxLfa9=V2)SalekHjl`yj#B0^%?rVZ0fj%iH8v4}@C?_T+8jW{U>Ct-o%$E^w zG;w&fy4ZbHKpfC#1VupanL`m+HxYQHEFUuK&@s|*;t_pXPG{wyGiwgD?RJk^0gl^=t^KoHJn%Hi*Iobf|cti0>XAGq7-c zArcOY-1zlTT+y?tUl~1^!FiFUKK#!?yrpN3uRua!;oc>7tEib6nHy8nOJ1*D+r2F7 zB|Jlrcx88ofyH96<0I0LKp-ax>a$bvV_d#RCGdLFN-nBbbsylg5|&wL0a)beD}eGjoThC32`B)`b3%fRq0c@a zeG(=DJ==7H0DvibQcyTP51t+R7Jbrdk~INm5n`ZCCh=NDU>ceg7J1TZ669Ge^$Bq* z3C&XoX6RegNn?h_^H4^?l!tiw=RYGP@35BHx5$!inV@-SKA13wTc+$eL7~8Ughi0T z9zl=>yv2HaFCa`nVubbxiz02(DHs{0m6@(ZM#=m_2-#6&xH zL{*j7IEC;QeTyz>-omZn+SAZ}!7CFnZy`4X85fMV=v%}|^A><5EbPe(jF_JYH3th1 z!dnb13Z;1qracrM74sJ1&qSg?_ZI|%3@kGBVGv+34X#&1%(J~E0Jty>!!XFeqEcUr z1)dbyQvzD4#0X?8Fy3Nd5h=}EtgY4>E0_mn!WHutoNq@wql*CHEd~~a(!9l-Vd&T! z*TobB_X$DnBN$^~k*BW!%28*{9VQ3?IWBolP$-aHvjHD{5++jfL;(7fJt-(0$eLte z(I>qo=?i!ZlNaJ;hvX?iVZgG&B2Rivf*6SQi4Y_0R4Gm+p?Mf@F|eqU#te<;wZL2O zpAiyl3^NQZvZPxkXdapmJh(0Hr!pguI4Pt@3@w5b_6UMBNZ2EYry!ov7y{`LLyIDX zJ;K8#m{7!*yXQ$k;Xp9O&>~5Gfs_*+!CRPMMAd-*IYFU7c#EM$mo#tTmTApIv|ogn zY3C0ypCL97<1L03anig6@I)E*6jKwSPYf9fOm8u?D3sLL%lZ`m0RCPz+*GsjtNXPeOZ2YZydg1Tq#FZ!xrpl;$n^4Bmnl z05Pe-c{JuW!dr|i3Z;3Axo4weZ>Coah42=Vrv!xoS(A(`@}$=!h=Kk-A~)MR>ZCD4!$_332rih-#|QOU|E4-3X3H91yW9Q1aE;@LBbxJ z7g2@dH4f8Tj4isPc?&lSYfnS_g+Uc!3PS#@kkDhi#n>WFnzsPVWbqcfq7joCq2?g5 z!+49aMWHlr!L$d!qliV!LK5>9!G>bI#n>WK9|i%|%OG!IiYWk2aGwzpb_|1zEh_c3 zSin4gPYLLjlIH}40(nnlY!NBVTMQVyMTlW`juG=243iYXTZ}CVrFo0FXZXin%v*4u z5E5Lp04(zK6#%h>K5E{=M2<7DVd!(rN1sF}fO*0+#Q=aQds0wlAiTxcqEC8FGF%96 zA$dwr7!clKVv#4kCP57J_X%-UNS_oG4urRuSky^lhK7+SZ(+(qj2ZaP2nqqRESp$l zNw-YUJTxCdOte#`7-)!$!+48{MUcWCfs+H)Cou)l@QHXzNVu_?V`5RHut$Uu2Fw*k zd}E=_!!X6fB1wLM)E4XX+9E2~{m%)?3`B1+vFMWKE!@BGth5T7TAwkZ! zm{`O~^A>=a9C&B@hnSiOePU2}koQC;7KPHh1=CmxkBTXX@Mi{v2*V&2nffpYu$TsW z3yn1*JBMV|DU1bG(PCwOy&c=V-a zAR~h9b*x*|Nn?hF9Vl;sSU`4jTOML8!8QzSnHE{nEfaiFWIl-dsmusuEYKcd5u~t3 z5R8JEJL1cocnabvnRQr?&?muHM~Dl11WgRsERSyko+kx`10nV{ERy6GNI7Q_dJB^m zQ8nOyPEaTi-eSX|OPaTE%d}=9+AmBv#Jq+4SwSH|c#92-IBDJjFq6ewgcxfpO-yzK znu+NxHY^IIc?;%q0X%9lN6cG^{Z?R!fD9u~U85TMAVt#`A zcqBF$2;Z>C(^mk*YB(WpVf<`uj#6K?q@Dww3eSh1L@cO@A^?8Mo`j4(T^XN`#lzxH zdQ~zOFc)Tq5U)xkPYH?xnim#)(#sOWLbO$wScvmN`lO(Ed^VOB7J1URpb#0+vRVXG&g&kBrFf8BxCaBCKu z(k#ZfRThKu?IH11{|ocm;!&E$n6s><(HGMjOtbJBuo%==1H_8D!DEDsQB(~Gj6->h zH9ig{9!Jd<0XS6lq>#8PY<;q3aVWh$nJk3IkUS+Q4%qr+&0jReDBYD3Ip zNS+cD2dq!@EuIwi2@ktqVj-SCkXWGk7ZeY)Ml7b}*GM_v5ju zN<{M*=;P2a)0oqsM z;t&8}1jeB}hI4xn&0|a$JO-~i#IKle9*qPa{gnuYQh&YeTP z8X#8G4IYD+0Rg!#F#^j442L?G5BYE?VMg$55dczUW?=Ea)+f#lL^O}FF5oeEnIT@h zNKC-ufaZneI_bPXjA*G>i1S5g9*P~DJBn!R(0CxqV<1D)7It78hWAa2G3mYuJ_^kU zgS^G9QTCjWc%ytL`(~nsjjiw)lBWd40qqkOPYU})Ydi)L3pD?N;(_67=WZf?jg(Uw z!DFPU5&v_7Vu9%~oJ)#m9>X26nvLj~@n%#pk0Cb%2^h*_IJXtiJO*G2i^t$aMoe;q znuEm$(_=W-7|}ci(;xtunpzT572(fB!a&aTC>C+`6iK zz<3PjMk1QW*f@F~Lt+9FBvhy4Tv0@?PY@%}Uctjz;-rz9fyD#c199#wqOn865R}I- z{X&c#_|HJxr12BhH=WCh=)MWshh~I{gScTSW61a}PmBpH~H9l7strB=mRTSX1vL`^^m&MeK5o=^dTVuS6_lf!6MC9 z1w;Xhx*gz0O(Y&s4|g95jRg9jplG1|!eUBkzaR*Wl>LHuOi(1Se&JkVB<&X-qJal5 z67&U+9Z0y)N^$Nn5>!fUArE4HUOeK;$ZrJ}jjuxbhI6IStNDGSW(Qi*yog*ejpYsM zb=^mVjfH0k7L{+oqGGY=YsFDv!BSpq-ezL+sCrBHO<}QN*@H#s-B^SyQ~LTO5dy#} zFGinfV)Pm6(e7K)#R#wnixP%SocoZx*aTQRNA#Hx;KH>4ivosCoU4$$9T%_x?Js6N z4r!{A2O$pAYrFr6oFC!bf^@g9a%$ZT%mx#KM@$*KQoW-4a#k7mzdp$Gk6@l#jC!BQ zbAJZB!FYd3J=lGv+lB7fovVzz*pzr4zfUvc z5rSDo_9<-L;@oKDS+}e$hUbtxCa6--+^}!PdvXIYF=;{DT9N05&_0ysShtwegC!bI zL@5X}d&uDe{~A!bC!jC1ei@iU!&*?ECQ2enAirJlriOO&ftC zY(T<<&n0YL6tM!$48)Sw<`SWOD9_>CdnBloavmeZ8;L5#|Dce-D~#t@x7gFebGY%Z z8Hr|yiHMxakUuOaDj3hPZtecBnflyO0c-pTM5Lb~whb72U2FcJ{;x3JIsbM|cb3w)o^ z@!)0eYmMKpQV(?R7y6Ap9P=G{FQk9Yp84*)KZb{$wInc;8U2b!AXkY73!a} zY^0;lm03u&Z(5re{hXj!{3ccpmWS-)^?;_J8U6cAioZl{bSIt^N53z~x8KHmv%A`Z zx^D-w$2a$9JJY0u(Y71+Ymz#*ZRGsgx0})|aq5cczoke96k9TOedF|_!#*@3$ zr*~h<@dSM$GQjGW|E0Ray<6_!aC-Ia-IEXL|Mt4>|Kq{_gL?-Dy^n5wwz&p=+6O-! zfS(S*Pe>aN*vmfbWgqsk4|~~%z3jtY_F*pvu$Kea%K_}=0QPbKdpUr;9Kc=@ zLz+O&ZFDI~<6WGgj*voa;%XQexb=b>w*voa;%XQexb=b>w*voa;%MIAe4cN;K z*vk#r%MIAe4cN;K*vk#r%MIAejg201$u9Ew8uED``FwzUK14nrA)k+t&nL*|>&WMb zzia2HZ3e;)ByqbK<15q~v)f`1J9p ze;)Byqbl&{HLQX^NBq^m3jTS-Uk$C`pGW-F;0peE#9s}s;Gc*6)f)rg&uh{g`5gAQ ze-8Gye-8Gye{MvQ*ml2Dsgu(uw-35Ycb-@|tvZ*w%Xj`_C;#1I>u>YmZPOdGgYFr1 z>-k%{%cp;_qdvD&C)MuJ`#Fc*Tk`+uoq6;Dy)AdReRxx!;(1Fh-s|3e;q2+CXAiiIubJmaW5^H)h?XbIU6|^^DE>lTNEs=kuQ>Jy2iUQO|UFuFFqc?=I{A^qPcU z{iEE@pd&_Cbg&-T-g+`O{F&;^+h?;In_Ju0b@X0%!h}>k)lnbiT9!X>a}9{lv#SW* zc0rx)bWg0DJaOX0U2cFnL58Z&s>eoeD|DAHKC#oAk5~^f)OZs1>|OO-*L&yh-IJFe z-#@Gyp-l^Vj4@LER!8+-H{0L7cJt-iyW5Ae)qCDHyZZJ^Ph8iZT|MaDUi)VcvyDu) zl@)dC_1VF}^s!m@`Q49w`Q8(|TQ7KE_u32g_Fr&mdg%p)k1yEO+nJwg4AZ|N*gDG2VJy(`n5&Fm*voycGzP#^B7=hvLOy0?3{zqhlwHM=p})s=SjX7_Yeaj0{> z=RI2cl3GZxXVVceTQmYM)facv$6nppzOi}b=BDGl-VA<1?+m|V^+{%pJoT~!HS+Bp z^*q9u?yV2craSj;zoSQ+-1Ci(O$cO{f(*3!R~_}wAT#r>0>q2le1RGcGnnd2bt*!J zl!Ldqwt91s=6oBSOS<2h0mS3i^LZ3Gc<(}T?4s^e*Y$*TYx$QRm5F2mZmz}EFPdMv zF)y|rzq+HI3+0!Yyue}1&4j~O#+EEE@A#_UKQtssT7e=CeYSbdnP;$( z(8)-)A2Vqz#y_n(MJ??YLpQH1n_a&$+uGXRee9sxykFbQM%Vl+c4QfnGp`LR&u@(F z?dt6LGa6to1&r((XJq@{hcbsgMF7-)B}R5ojO+_-Q5I)F*^#}xvwd||4C_PFEqp{x zy1vAe91dTL1VC

8{`vVl68j}{p?^=pzM?Q@25Kt9lR#CE;Tl(1-Z z4vR5#^n>$3=BUWc#zM<>$T>L3ZW~R`sgF{F)qj(W|By5O1q*Z>ef>@% z1I9%LIBL?OoNA}|2@dhYGr@FtT4oyecxnbX#h3y9ha@dVoU|B6nq=#s&A6%WCURs_ z!%^#V8hw@>VI3rL9;ROa zh&5gPr6fVtISI0nqJOmP^>;0)9>jO#^v<~;AcCJ*fj5wZKm>ohze~`yr1=#7471^6 z2oYi1L*k)Cd!By}l_ez$D1;(~vGPV@G9`XurA-Jg-t2Yn7ePe@eWs7&W@QSOT0GLQ zsPCl`rQ{s>^Gz{f5N}8_8cJ0E^coP`QvObqDK1lq_mqC5QMM(u8$0y_RI-#bCVh>N z3lM8#713A3_X~-Qi0CZ(1zbUAap_PzUkW}n5}dUkHTC0Ex|ED5eZvwkQV^Fmf20DV z%ZRUD3N$4mb@WLbg4A*8RXlYZe)VY(fuEuhsH9!#E2x;GwqP<=e5+Nk8`(wMf+TV| zT0DukQoY5wNmiM950yqGck5$ZfWw2JU2MGb5W8wQr+R1Z!{Lo5^c&eX^nTy;aMnH3 zz4hwO69!@0~Gb3#UIpVUmM1;vlGV7wQlw;U8#{n=`#&|no2+}>>EDsD-Gj3YD79Ms zxhN-Gsus@)hpTPFW@^R72&nodDjiCimcH=rT>X5wuwD_l1`uD9$23~~nJ6P%ZWhmo z4V%Sn*-iB!T4tfBDWqT0vc4&#><@`nH~|e&U_eBuWK#%>mhGQwZ>6e@=tu=8^&^?&#Tw8to ziP`?mN)wt?A1dAj5!gn}(uEL3{f%U%(N)IDlD^H7qxFtX6SAL_Y%IOUgA7F&7HOfM z2<=PRPgqQBe_%63+)q+EM<)u!fAmWpmgYYiw#pSJW*bmUV6Y$VZx?i&MEpmeMY#4I z!u9w%PJ-8xXQja_I7VT8q<|3fuko^F&SKarGF`@h=;GqcH%UaVE&EMk5|;k85ocWe zfU~GPf-HX4#%=RA86hTN@u>Z=ooSBNJGzCEn1m&PMbbpxOU#55KQkW>-ASLtruHW` zQ;xCclQ4G4 z+e+(B+^$j^D|uhxO+jLLh1kM;@G!8H9sJJUr_lw+!6^;{d&KUMiA(BE1|=6u>rQ~d zh+}25q`}6DbjDVpSg#T2M0Nv=Aaz*$?BLwu{UGaujJcw8CxeoyrFAFF*h@Lv$P^L_ zGJ@TZ=}vS}GWyS;P}DwsVnFdOgObPf@h-r|y=>>0H(dpA6HLpzz~@ex?nKugCs!N} z_b7d%`wYdq3`z#q2ZDiN2*Wf;U2)|(0U%Lex{PrUqGN|SXDLbb+_N!UqWtl?l-Ez_OodgA27km^@Q?=q!38I(*dtvi_uY|>sqYb0zo zi1iyxqh-1iT|S(QNW#vx<}o>il)S8O3Mo4OYWxbvzXcf3X9d<#efE@a3SrT*1BaG9 zLf2@NL2Fah$BkezFGWM4|S01|=s;>rVQs)5F8; zG~K=3)nMI;VPwS{9fHSG(>&*l6OryjR~cufaWt>T*>$p?lx!@mJ3+=N42$?D9dXML z+DDe0u%ED)*nz{u9;tJ5KcV=KVadbN{6}LT7O6X7KRmXVh;%0d7U4QP7ip9Q>tgzZ z9XC=mJuG=v8oYwf3hN`qEW0?}iO6&rc-6(lm~V!T8deNmMb?uIlI|fCCK7nT#0liy3#nA;%H){=AG7^r_MVW zmTW7nJ3+@PtgrajI$~jl*g~@SWV#bwaGZSMo;PsRzEkI&3`;JS)|~)@5zop-0$Z`} zB+v<&?nIX#Cv_Ztc2Rzi^+7jVN_R3WnOa(R!i>Fyvo&;_SUnPKOhgDp=bh-H9 z-aaj2K=LjnkL%-IfW7%-JIC#wchX;F+Ahgqsuo(=BGaAd z@?p#*BZr+mh8|P8lM(COSTibdD;$#+0j;zo5I6=zB!}p{6J1Q4%7-StQV;}VL#m8F6e?1=}vToaWZ5~KGLR-nEyx^6_WoLvCex1tqJQRSiiHe z5+gj3b!4H*_zzuNoE&lZ)FX6_OjPm=wh>DU+ZY5~+re1$#!t-n<74}XNDHg$i<2Yk zkw)3-X)|&(CShq|ku+(mm&8xow6uioCbBgSi%lIlZ0eCZNB0v-cQR(359=mMxiSmU zo$yvUQGMBKoxiqax|0!$V;wsj>rwkhCkv%J8JBD;tvf-JC9JO$BO0uY6$>-O=F4;^ zy5Jad(AZ&M+xLKMoOHTSx|4Cq#nQSHAl@8Ol!05ss*6BVBEly+??jg$Cv_Zt_Gu6S zN_R3YnOa(R!W=aTXUm(Y#Da`qOUVqAop+*(l9NQmJzxIy;cS zx5DAU(|}om*ma-3=ad{N;S|E6WykB}Bgtg7$&pe;rAkz*#? zC)Op#=^-1;OgLK4Dc#AWqN|LNB@^zYUE&>_w60S+wn@px z(z+AG9yt~{3f+m&cv(NuWyZ-Ahlv%p459sGvn1OZdC%~Kb>3^k)BMVIiUUhMCa-`r>Bloh}4GO6{ znXt65jX}VLAB;tB{M?H_KDLjDbSJvLI5|R||5{)nvbot+PMeV<*EkcF78Xepc`q^i zNc_wkAuS;XMTkjQZ0aON3o8sh*(zzzDc#A0bv~?XYq{(U(Vh7AW#~7oY*|J1U$6Z# z-HEO=PETFFSW{52$i%+II(FMoJmx6)yis+9GC%A%b;~0GTucO7biy?_V(ENP3cb7OYWA|oeX1i zCp@q#7GW@KqSg$BC`ESOiLNJ3PB?t+(Rf8;GwQsP^^&QjbtiM7$4V8a;1cZhh*cO& zqh;rv=97D2Pqu8NjepCCz z>r76TFh;tDDZVQBot*$SS7f}5lhq~0Ns~BER%Abgv1l%n+~ZjaKbk7vDPV`<$9VyqmCn01S|D+o=9s4CIB#k$NmnZjja zk2EZ#?qt2>Vd?W;8>6&H-3j~Q{&qnZlt_1?D~yvN>CPz-GbHS|(fkMNyjSp95q(6A zPeleqbP>^EeY&_fIa2Vck$9}#h>@6NgQbOSj1VsTpclRVjj#Aw8@J8hWMud3th0F3 z4d-lvqxFt%q0|~@gQbN<(j=mn#81A*LwBOfjL}Ut95(eBdrs+2HdyDwx`|S*%tCZ0 zE%7BH>li&nbkB~iG)|^C9Bb6<;f=oEKf0?>x|5BPZKZW5=x{{z6|uZRY(xZ3`P+_q z(bb*p8=F^dZa%iNcV)V>xx2T2eHHp_Qx_a3U!2ookG1cV?qs9nVrkt8FjDcXY$U)F z>oo$MklnMR%a4;f&hZ0_@*|>nDBa0M$<)%i6K3oe%-MWW91i#RI!PsH0Dg2z(`gz++O_BNm-QCwP>V0Is zlte5oNkQ`^qPvJ?8Db;C5rX=AS<1NVERiw}J$tOZ{{WRSYb8-jt5Sf$;*g>^hDm^P z#cGZ~Co(~S1o86N+cig4yC_d0D(I)l94dKSTA0ENe-NW9mR1BC6A{DD73GmGOb)lZ z=7?^e7EvJim^Ic}wBw`Hrm!FG#X0mof2QRZw4F5`!@JgxG|S?g0LBK8A3&-envYpy z9Y+felN{d=^AtUNHckO(8i*5Sx{#021;*+sPLQ3CBIEYgc%#W$uKrK?I!hU#cQ3-hw9v$0n?X=j0aiMvUKI^zx(42^FB1Uu~10s5f49n^Q z<4in#hiyH!o|E{b&(hI01_>96Fjm6xQ#Ahg*lr@zspuMGrAeP$ZP>-$(ME>GCw-QV z7ReNOM4hI$EIz5&e|Zsq|Ur%(}*w3)K*vN~(2FR0H?g zuVgwEU2UvnaTwTk?T>A&q)w$@a@6nw!|PwRy8J10(MnJCPY+~jIq^a#>$j9#&%J@llwq}lChbgY;cS(xi2FEU^qFALfl(M`m-Ok_Yr zFA-u{9@{!_&OSJ{p40e*rK4>O5-t?Mbt*h|K7RhjZS%Jpp^jG97^|B&EbGzwN3Zv2 ze8SSvBAFuZD7sE;Bl>k;<7OR8!SibJE>C{vW}&7bERCYJ*{*q zBKw6pl>v`|9iH>|G_vL+_9bNQlRA}Q$;Hw-74(5v_0n@uk#fNdlvt@DHeaYy(M89Z zjD~B>glOMEy1SA(m0`)$(mE9&# zF0E5xj*W!T#b^i!>iIWSwuj9 zBSXAs<4&PYMb{uJU55TAljCP`PJ6+j_ZEk&^KHQ#3Bx!36+{4K0ptm09o1Ft6rQJ| z3yhT}&KU_u=p31ev`%GMGP|@+W&LP$DvZZcWtv{6qHBtk7UM{>Z1)f=Sr{>+bt=P> z$E9^DbC1Dl5?>LS{UEVsgZXHx9e6YwGM$PpAy!gwm9j??5YJpvvbOYjwPs)zvE~VA zx5Ng$NW>zYOL(O1&>?M)&^dZ8AsO2tld&bnL9(hb9VCF56qyjwNn|>eA&;>gI_lVC z>^r#+G%6WeTBov+LZ_1Iy+ZL$iokl|K9H_7&XhB94x??;dJ64uL^@H)k1je^e@SvmZToRkq)ug2GPSf$1sJSFaJFI@MxdQ! zY6;I%(Iv>r9*4a>7H-n0pFB@xRPwmAPK7x(5=J+}Hbuh?R>_IA9l>rWnNCF)CM%B| zZg+9MkOf0$63NGmS?Aks#OqYdycIx701|^RDGD7ZAs@qIc*hRIdwg9d`Is^5d|NO_ ztr(z%33+Kz`L&ao-bu5)DB(C4X) znT#zFYRRg`VAB9%Qe*-d5^~UlJIcDuSe>Lr#x`AG)b~)g%Z*FM_JPOtr(4^z-NS>` zE3^H>)!FXVy{*~)>dy4$Y=3j@TneEI4}iQyo!Y#*Gd(!yKCB+dZhbscurt8M6?#!&Y`Y1 zR=zj~AslDtzmrUtlB>O6m@W}=l{cA0c}QLAJ{XcH*wg*3gRU%@RC3ItWykB0W?4AX zV(8R&lW9`Yv9Ah96Re)`^$u?mder36BkJMqL!o(unMbx8;TvCL9`QPMh}Wa^jxG`s zt4vC^Rf<(`)d-#@2Gk~)t?}na7Oxbm@Hp2=x-$@34W1&g%A}-SrC5d5M&Jp<>njQ7 za@s(;v&gWDt~U-=nK;7QqxFyOEk8g_LX(nz6-+|q5-mi!l1j24*T}sNsQ)FpI7*is zD_G+nU(LrYwTNTjAhyR`G1%n|>UG^mgxG~;BHd+ViYHxptgJ~w z+b$qB>LWx-vxR~w`7v1 znUox`m!|>t`5`3<-G%%D=MW}>X_^-h+$r2DnDF@D33*~d&OCGV87)SK=4mD+`|E|s zz*w|}bwco?z&xs`q$l&}Dr4ashxt89@94GEH<6Q5$@|{zo1AJ`vyD#jh?!|#sb0~2 zIcuiz8b$SxQqA(0vUL-W>z%BVV`7=yqw@M&$fPJ)-upyJ(N^DhgGr1_>cQ?S<%xkA zMsi*laaD_Sr4T+d$7 z+Huyk2#ljJNsz8Zmm4cr`u-=A{m!Bm`CgGAKAzrEGP$0;BIJ1(6u^5+TYE)91IflC zJh4aD8!KO&V`}A%HKl)KRw|M&SuZ(U4`0%-7E_ksiDK}Ie9_N-dSn3$`I2=WS-Z{^ z;T~yVJkYk}Xg$yt94L8DF{vVla3TxHra>2#>5@;x$Kr&s6u0u`&ZE(5WV-f^X*RoNvydGT(b95wpZL$VvTLB{9M!{W_#ZI08T1ZwB(y)JwV>-hsp(6;a;5FB@bYk-Fg!smH*_ ze?y3&`{zow^waGaU)?!ay|F*r+P-?Yy|;U?dbmH`J-C+DeMn9F2nia`SxW0QCRy%R z_jT`?OO_Cu)t}1-pGAV|fadAwf43;TLyV^O^<*xTl;;J0E?_0e*!F*|8PHd$2fFtQ z2NZj7L@)R=NjhYO!)W;ZbKLV_k02cCJ!CeNoTYPbAYNs0i-rV@H+`c|oZ;%PWtqVu zEc@pi!m>^4BPxPAO}c-sBrBbpYGq-TtQ8Vu2z^Lqtm(27K=t7Bx`!iDZP)&YEYV9O z6hrpUl}u$HpHr|_j?W>yt=9X@1oINL(VciE7|e>5k7;bjDRuh%sqV?sU-+Jm`aM}| zVDXSTXcqS@!V40ysE?52Tau20`uHBq9^c%b?M!n7`L^vFtFm1zLk+y#jNwbw`R+O} zhO8&8JijrBx2v<~&vchgFLjqsKShI3^#`&+W|4~hbBx zvE4pFls>)@m}jcn&z~*ktGwB%r2(8)mE3>Vw?cK)?s_9~FGwU|y zUkNinhv1Gz2x4+zt;Yzd9q1(Ch?~{)K1jple#=?#64M3N)N1#|b6ub~@7rdi4)J`(M%$EbjKWKT_X|8L2*g0fE!esbj zpcM>al3~zHhGu|mGPDq!;N`OV3Lra%EoTSk&KuhlpzLT;2?_J#OM(0t#^lGanIFv% zGXX;@v7^#QeI1Y?BTj}SgKse~lxEt13nv8kO(@N-H>QuXdq;>!O5Do8cqCsPQ3jsnY%z_ zvk{Xs8_k?)hMFvLn^3KW$=ivPzB}`uH>jj@oVB$q?tZV>M`wJ&eX8DuUD__UItm*+O3T)0IsFa0^p8r z&&~D&kiCNT(+5(h8S@nFr~OE0BCuu#HTlZ48T6RRptq_wci$M4K@bsm@Whp3zu@87 z!P}-cW(ix-yP-S+cw#{w4S+mqM&7ofg+(nKL%$cwrY5@yWK&xpdfTMhrh04l&B;W~WKc8KHiO#p&dWD@UL6N`srcTSTA{ve-bG{O4I87cD$ed%*a5y`HbKtxN$-!3h)lj|wT(clwCP2P4qipk~ z4Zo-+!#RcEYoJtVa-Tq|L^PS#8$feWy{h{F2hh|8cmeI@tDtlNIA%e*tOM!NjI>Rc z)}ll-7LSY=oRXTXXWDEDXtQOuEO>ndPDudAEXbJ+AZMEKwmH*stim$~qL}-~kYXk> zCvv$&vtSU#+&>pX6ZQd$8HjiajwgU(7R5Q+ylJeU`c29c=sQ`)qHnTBOkv_38~1p+ zgW1lt)yJQh?cc0~!Qmp|OqS9B!u=?MxepM`K(xt(wk@gENO`$uaBr*Mq&k7riO3Fd zK_J<&0H6alo+xs;50J}1#F^>SdMK$mEfc6T8&8wqOq)GoN}`cHNz^&y0D;s^KZ;E5 z17vc)8Eb0A!%)`P#?)Sz!7b5#lg9+ormgu(3=F}XO45(A*}OSpMe)XcfHwx>ZO6Xl za8NykE-Mh}H2F&)bB+ad6F=GGdLJ0+M6tzvfGzGfqfHK(O{zAX_jf%Q=`{IEAa&Ym zI`MONLWvKk;^{{*#(jV>?l&Wzr%$oqS)7We(ZA_X!c$h0&0JviMD^%$eVIt&1FCrX zQS5OaV2}IFc>C*~*__&qW_GS?6DT*?O(1M%6{O-B zw9Pk*)hNttYL$70R6K(y=6L`x&p_-`^rgi`9HfE_u( z^-z*)a%gXB(4;(p>^T;FN4&k$U!5KvX6H`s?XG$xKj7ZhAc|lf00c7-?PLUVtKk6~ zPm}5dQl~8}EG{J^8!+r5TAzb^TZ1Tac>s{hgJ#6TClfo}wH->Bx?xevZDo&`uV`dX z5{VC};u%Dd$pe5)9yDW}tnIWHNpMSa$l{RON*ggS1k)x>%Yjrp!zkW(2=K;0ylvxX zIUE$2ApsYtS01*_6~|A&xb{Xg8_}_W^vc61ws;7z#Xz)e-n1Gif~{f56G0Wv@ae-B z$Is<{ol0XHeHg_U4*|v)i1_h9I{xT1WEevhXB@FWk7~m;8AdcgOt&8BES_N$dprc# zV<6t7S+*~=97u0x0#!Ui7LVK(;t>l^T*On$tu&|yQt=F<808_rC<8IKS+anTgd-p5 zES}-hhg%klSAcZm=z$cZ;u%J<%tL@>1|o0sXi>g~Pn@8NXUO84+v-PR!3i@#B57&x zES_N$^E?EYXCU^>scnIAqnZ~;1r1pobXyq}Kwtn0i)iEoQbEHg26_lE&_JwhM`|$? zKI{{q;u$vCP=J44EUpGJBMI&ys~p@i9Y(RwLx6n-;!T=m^C|#WO9DwW99YFOZ1SOL zGe;~#>95wXa&$K#2L@E}45L`*A;3BVQKndD#%Vb=quK~CsNxy1_~wWvBi5(PrOE=@ z3#8&1MKR4IfN2IIO=hu8hSowzG?r7~?!<`2F}Hp**3qbP!T1Q5(Xv?*Z2*mbhG)n+W=#>1kQ7uk4_>{tNM z0UJ*gxjX{MWq!Yy$j{wGa)Qq^z zkk$hdj&yW%8Y-Rcj{%k$h`en@3yWI#bOxQpGj5x277I?8+0-iY44uU@j$)q20P_sQ-saN+ z;})J#(6P*876*MSX1D;d1kfWuoAe+RG>&4R#{dHj#F}iBHoe+&-HV#;1gW5LlMMy5 z7{?-Q0#GNAHYg~9Ug#P}vCm_GeFoxfv!h)PDa_;1e?e#Qj9J`sTkCnT0%ed`;sdI9 z#!;;E7+{@&C^K74D-j~7AL1;Yag+Z9noL`%BG#wOr3%(DL_k_WDxPr^(>w;4W+2j3 z+0!TG*v8aW2;nKI$$A3W60y<>XtRVCBTA@4)-t#|F^*!H#{kPbZpPb=e9N&4$H35$ z7~>}431rT(05bvfE0KfERasV9O^zpuVjcq&GZ1l`J8k+JBT`3PN=WJ?!7_CEK#gY-MJ`VO zav6xYO`q1I6+Uf2)XjuNEw=@a#gs%NwJ>3h3?M_tC{LouKsbn^q$g*>s>K*5v8K7ssOjTJi%fu_jTB@dRLufry8<1Dkk@h)zSrGhuPYPyZ>a zCc^+Xs-UxYCQ_*9kC$5g8=bqW~Fn3{*UmD3*Bwu*^W@Z5}Nw zYT?-gsdy$VzPYW1yI88iOtmJNENWWdSv->{=6M1z&p_;LJ}odb;TZ)|K`ajXSOBB| zvILN0fZzzEf}$Ac3BW*4nz6PS)Sh=<0OrwRKvdBB|Hs|C23e9F=V9oX#XbrIK{iEd zJVBYJM3LZ-U{Bvy-#d8EVzGD;yIL%k00?fQXXfs11G6)O?jC|m(IWT)K~pkiQDlgc ze~=2k^q?fl`s-f~+4^Bc$l-8=C0h>Ba@Y)oY>TpFDNNJ$>Aqc6=Tv@~m3i{io!cfO zNbF$t)X6;g)t8y)%eq7NpEHs@aM)`|ijazD?&?0zLH8N>yA7RzeI>kbdLjp@c;>3! zvm+rg%TOkTg*tK)jeb#U?&>uJXE%0mY{UvsK~nq zs(9wEPV*dent`X=h#1Tf!iz=j2|yLkeCQr?#%oqkP^?+SGk0~F=b+0B{9U3=IINf9 zW%^vzVjk)}lhGFy!pd7qq098StHnGAEoR{H60tM1E1`;Ku4*w46+PzboasOkANY$} zb610T4jRnB+bvTK>I@-=$yM>pRqf@CD89|mV&aU;is0SW+|^v3gXS{ucpE-L+nE1~ zgWuwrt6Iw&5k2PS28kX{@&gC3=B{S)95j=GuY21nat_;u-GI)4ts7Mjc_YHcoD0IR zVP840b>r$A&q3dKPW_$#B50##c;ArA^f~Qcg1fAZh#Y?=!0UlM}1tml=4zjia@BEuYSiif5tfH#@FC&1w{7G&!5ga$1DG#j|jAo)@6=41Ayeu4p51 zeIAFcpoOXj?TDz19tDtNxbq)5f{==5;p#vyKnEK5x{ZSY{nVEj^=L6rD`=tWKW|W8 z4PuJRb8LAvY6UG^-RA}9J_CP?5#7dBM$&EZmB4TK7qu3u-t$nAW7eol3rl?9w|Ew= zuJZzPoq?C77SOQ2sHN~B{6(z=?SF1+*U+qV>G*_u7eN)z!qsVBfKD^;bo&Mw%C`LM z395J&7pKRZ@lFQ<$c&a}DFQ1{C%I1^~ z9{3twJ;6O7d_X7J5`;fdu`{HJ{qGk!d~oLTNE^w1A}^@@IZX}0>j(6bEg>{RAw-=n9$gz>gF8%mJ2CY2 zNE^zomvth$a9?p%zXIb2^p-90vxnk`y1WgMfj#0ZzkV5JK&nQwqx>?^_T7)2MS0d1DQfOyo1hE5#?VhQx3EwQvf zu|%EU#?rdXJqwx@T#dh!Tm2@AsNJFa(q{3AG7bjqPzP7hnA!z0XhTfxq7F51fBp%# z(YQXR<5tlwZJloFwQ5GQ1c(<mLM96WKgH22nZivmwFdHrGclI_%1+?pHMDLPaCRI=_g=FCf|vEgL>&JTc+3hQtTn zh3)#9(7UJ!y-S^*|5|9{YiL(9tsUA01rf9n6lQK+kXW>-fOdUN=Uvov-lgttBVafK zd)r4n+y!&zk*e3cF|8dK4MpBdxV1C#^_NGezYJX7hL1USws|B!cZQkINYzt*q?r$| z-6>diK^4%**Hs>&t}<}=+A<%(lbM;%=#lo7{aih`3TWi(ERRrU88}@TLW4U@dOKki z(1^A?LFVIl0lD7CiG%}PM-_YpK^4%**JU1|E;Dd>A^PffSMPg9jN}TcfJU?h3PNb8 z49Ki^iTOTpSeYycs(?nm&hrR$o`KuN!)s$~4K5R3L{J4Zq7BhCMv+@mHs zunK78>qd`IHySwqES6UIFlutzKwFnru?lFU>PrvxUNWmylo3Un&bF9>Dxi_CLp?$r zYT*77F||Iw;#SdU=w5Y3{{yHs22Pfs3TWi(RF6=n8o0ZCzYOTAzC=mZi!ix1ZJDl- z;xZC(5LacABzVm_@^z_4s7pPfE|*?H@~O6gH>@wQ3TQ-|rXYG8m6}=LviIm!KqFsw zdW5>uz|qRYdJqo^Z=&;G@`5U$@gwa^o0Tt}SSs_HfhwS}uOmH19ckd~_GK}cFQg(% zj95%h6jMZV%J4SX8{1(ty z)q-|Jkhxz$9KvaLOg%u^HL$fa_BEZysOb#c-H&~7?``M}=V>N)ex$u-bEXL6s>plk z)IA_~_Vt&?sJ}d>E^p6&V2|X#c%TYsJakXl&)0%>rC3JBWIn#G@)&iMfy0ZvdWQ5h zlljmlC+G~G#_SN&rRGY4ynF^%0gZi~%J$TR@Yc`_g7XiZY@G?NCD%(8SlFo}dmjaDN+9YjjBkQvNT!V4y6o_V6`MQ=nM zpYi5CuiiBk979jOZuA6oqk+rYxEa>S{9XZWMNeo~6ufVSqNj}JsSps(0t{cmPkgQD z32H?Hr?>GlxKHxmKkzPWLi?g1f;Qs)W4`HGcvV0XUn6>g8qvVvBA>RQHKhIh*Kzm- zG*z{s9SM&aPt2jOA@PAS#Hp_dJw;7u;B;jO4em|p?Su!srm7b7M!bN`oevU=oSq1` zcBa0j^At6mfxFu`%y0(wzhA@%#;u*Hs@FVJ_G8WzVKfwZFX7yI>gz90QGXe@ybYg$ zJtDnZM8?=x3M#%iQbnF9Ke`9(!R2v$Ol&e zO?{o^De5c(r?(+AxKHw5K=4y`(?{A{cD<|4fy zER|XGONOutXy)rq&ro+7I9eGqL)b07iKK4<%~XA9NA_bDw{#AfdpQ)$pP>q9=Icn$ zP)8a#TSB<3(P{Z@C>P_KjG3wz?TDC+SL%7qz2>?L)03|oJwx4S;PUqL2X>c~`ik#6 zdNXG`2Ov? z2gk?TFI3NOfBf$K+c)0#>e0cS!{xn`5BuM<5(ooL|k_VqKYw*=v{W*jyAcF4CqeRTi$ z_=V+zyN9R(#?a7h=+87U^nCU4?T_We5GUX%m^otx^L3tQsPjCd{%)gZNMAW$K;er?;02`W zIB!IQ=jA30U#wo(em*x0xfdC35YBvU<{4@;Ga02^B!t41FwraYa^^c zdeq}iDLp!UNmCNrQ}vlo*B#=^Qoc9#dBI=kt3lOYMpJ})zLxVIYB>X+w?Q(rOPrZd z&`p6L8oK*@tI1zJRlT(RqA`C-;)Kg#3Z~OgTejzGMDL+S^d9wlev;CL;2OML{RWJq zp?lJ=HF5NE_1W#uX1I*1=aGs(elBJ=C<`qyBH-H0$)8lpBc&Tm@T4d#e8RP>EAcuLMXoklM#Y zgm?sc&)2=)L)~lOZ}HUHARE$Gec99%ChJ!9u7?VqoD>rTPmxRtw~qFFo$Ec+x!$8b zR|e3~{u16ySE`?fq(=LwpmpS^{9kNH?IYEPwm+yMwIt1UzgfhyITdRH`to(G_fW@r zk9xX&vkc=U;hm!SN>KQI5~$E<-?VJNe5whUk5?bt{)jzbio8wapn+eO+Vgd(_fVI5 zk9zxTXDfsl2PoQb8_dTHNzra;*_g?Bub#hT)?92WaX*&RnH3_DR&IFYdG2dZ&ry3CcwKxewy&4LoghXsz9XBf+S86i$A0Ux zyE!VMxvx1rN6l&Aab^4r>q;RjT3w(qFCp3@1td09rDcZ{6J15JBjzRKYfR5kV|q@# z?s?Bxogx%evKSk;dgin-y2co?w=amhib#lXt7q;7I|F-~$$(US=8edr3nTzmmC)SR zX`Z7_^PGCT*z0Fl7vVCD_C3K1Xd?osK)NIgWS9)d*KMAoZu6XaUHO(7+zH~VkRdVJ z1qCtm$VhCRooA>Ln)^D>bJTgBQ=b=m4f&X?*WDFV3C(F&6a>*myol@ym^i0O{KMr{ zr#_-r7;|4YdXBo$!0*ae(Hh(!zl@+tXij^iAdZGA>FshB5E{kX*Qa_a5KX=g^&EAm zf%lcsv_8*C^*>e#EmXZ~2f^8uF3PZ>y=IeuLp-}vU(u*q_`22$)U^iw7kAbcW$SdT z7*|jwv>3X7ozu?%T8@FQ>*0Xe!q>fCpzbyBcW=@pf83{yv;kd>v+f1$q5`7Zh=eI8 zp#)*bCQI-FcH!$>FHq+i_`HpPfvuEsAvs8~N@(#&`_^`G%x?d8flS7sn92I%bRKYFeRTA8?{kG00*sgx} z?YoB$_HVth|LWcQw+`;^-#d+H^k@6wU%yHO;>gu)xD4cp{LBOWqSxs9(EVNcN2>T2 zOYs{fxik#U0Wsrh@CM(an4u0YK|e#f8Ge6vbY0clbx@N0L0Mbm3|jP{@9zfp_}(|5 zw~Jg}B6fy#boG;vrBtn5NAx_?5YaoU#r9qg5fvdG`3>_^@D1bY<&Lg{UT*lqomY?U zKYXwsJ{^1b>b=*)2Vq-QD&&@L4(Y?H$50OgR@IX?9^ZcK$*=y}R<9oblMdau{a_Og z@2YNW&p|jGFTc2dw7h$8mf^m3_+aSI=Kc(f2UnYRbRD#5!ygLqP|7dgLIUEg)mv^n zwSDQyOWT*9{IXm{Rlye^CRB~tp%U4TGy(Iz>OI@<(gaMAmq_(h7#FT4?C3gZ!iGPj zaUpv1iE8V`?Mv6Uu9fu7RT4N)R*&C!>@;wy-`T1@^6dE+uRpLnz8&6VI}ab+IXGGF zuReeM`mN=iJBRmPJ>L1!{i8ek%l4mNSe|vr5AM8laCCTZ>+bTo2isqJ`iJd5U;VD` zA}bL=XspM7ZReSLC(EO&Prdx%(9k+LJb5E7Y`3m`E7||ss<+;F%j!+22>pMuRlPUU z#(G2BPuBnGYC(6koZ&U`A6%+Fk!fPM?Oc)Z*XuZAy#7~H<}_BXXYLpfDv|Bg_u{+F%ldF65^_rI{bw}0}+ zgQXQ8@4)!j-|a<+_!^1Y*jv~7c%%AfTh%izOKWUIC@Jo`UNhuCs$Zm5V;Oe|&hF4O($? zt&5vagm3;YdqI=nLg)-WreKxuet6npCBRH!r^!)L146p1S-GM5hnK4Fb6DyCjOCa) zw2VX&K|hP@vX#I=Jcd)2`LObNQM+z=a&Q*woq#9l*SrgpXpMuD+pnEpGOYOM{Zo-t3^Qr)tR$r+=icSGwi`sA%hz?oyz(^7h-Fue)5>OZ_(-6$cE zBIxcl)z6*k*7oQ6s~a>@X$>ITk23Y5QQO*3kVf?fK1fEMC=AJHW%Rz)^fWZ~f7OS}WS7~v zED|6bDEHP0%ETWeSCv5-Qa8ExB(uZmqq(;MNn$^sg`}1t6HH1ZDEr`FZB-w0&_v?{ z0LGzwn@@@juU0>Gx%ylQvmHH_G7-UnaNgK@4Na(*vE7C@@LKTJlB=&R8JJO3BE%m7 z>*Yp$rrKL*R?b>7@df~-r8vFqKLBWYii2on5t#W&!k_l#tjPwm5m_oS&LO!?ZoaRJnG#^H7Q=-r z&BS2@!PU=Qs(v7kz|ZGOSEGFf+I+Srdrj7T1WnlRj*>6Q*i1gzIp2mI*57R)_2vj- zj8?yJxq78!WMqT2SirDobqG}+G<~60syo}a9DN}aJ>VUDDv-@gzLF=XS@c5|ysnu? z|L;ESn+~gIz5q-*q@P=y=5y(v%f2f?p^y#GauLEJR2IjSOM8)D2x-di&F~{<`>yw9 zdvAJ6tDpI}4x`b^>sHl-$)6WXdIg|$Xs?$*>@Q!gUMuMaN6)Bq%&^!Ui+B-p;iHCK zU#ni(zT;=t2muVXPCvAj+0HwEd9=^aOM70(2k;sNKjFg+pDYx;E7t%N5AORCF#h$+ z)nQRbIB+~f=&)FxwB9>zuz6U$y1gukO%^xSH;MVK6nw|gP}<_FFtFwjWo^$_6|Hxr zVAp&}Yic?1D}!>UjkmsXw#WXTG4`FLU1n#>MFNBc=hO<1*16uMljJRuKcC)Bm@f3d z>gw_pUyXqkOP8)a@o2p)1v}=-)}efR7GA3)ryligWVN!jO`ug@F@ z^(V4DuV%XU^dqe_$Mt;6;di&H5Blr|1!8$mO1vgp*o0p7uP#-e@tNv?miOdYY|X8- z(m3SWod%1asXn#+k`Id<2CDye%hr)NM|j(!QMXue)R$wnW-SX3aI1gErMGv*$BE0J=0> zK}ahjWH#W-SL)bQukr@7Az3OiSWJ%c%SGP>gzPYiNZSY{SzR6!BZx$|K5$JXV~Xvk zWR^?l7_R8lI6?_`&GV%Pld_N7K0~jGVoX-HYeC6=M?G02Kv+~ppUWwXvYs;%;}o<> zZP$YEEYvKCN_e>9nmN*TEeLnaF{I&IkY5+^qj=6L@-$uxO54k3$1)MYqR{aT9^cTV zqPbS^~Ijg6hOx8xIY=GxvNM}9S*=II9 z3k3>`PltG>a6Q@a7{#cwo?NZ3W=d7nuv&A?TFadUDrjWrWP^GP22fD`!2K zY%m*=r6PmHnrlnmLX zD8>D+Tmw*Sn4(lXU?3RVrYKeW%*OE$p~GQ$e7(h{Z;Ddo69P`_o5VOpxw`}AR`}GN zpcvnVfi9%%7Ix~uS>y>xw6%CUCOPCuqeKdGzst(^K~h| zy=)Mci3k>j4k?lMx|DYo#h`79a(5>wehr+6hUYq~DcMjXP#i#hPu3va_z5>Ch})XRXp4CpRi9+Z1JI zuN=D>rYO;h2w@uP>M_+eMTs|-4K44*ihnSJx+ok0&EuZ?gi6mSaDo4&kpin4E? z*$73dSLR_-ps?`V6X^*+rSZ@;pCW)Uf^Uf!e}|G;UtIXq3uR~8zs zAf%NMG7FT4Hr1=V!E8vDiVPN$4wa7Y0z!5eVOis~AX#0W9wS!}=++0WDZL1zYe8)z zR6!fYYeBeco-aL^lzrU3%zdHB86i_Nzk#i2 zy0e~4Hki$rr6PmHq(e$mIIT?_Mp)MPTa@uTGFMiEhT_^vO_gF9P}!y^ z*x7R!#W1No*d zRX!x(H(!^6ITS-n7~h70HI-$p9=3h8o^^_HB+Qe7MA_7G0{-BA+;T^+iQ=2yd|fK- zG8?N!0)z$U-k{f|l0yja7SbQZ=US5}eK7+;P7V@j8&QRt8od9POuXKDN`%2ClN%Fb#^HcawVfep$}QFivrw{{Iv zlz2^)ulh`FaA};P};OTVT`{;*;$u+8xy8Ovvi%c zN^@M#XrXP2va{DSy8*iyrYO;h2w}~Yi9lUFrrM?`@y4>D0TT{pHarvewv4cpO zY5Xn9X4Q-qE6y6uTf)(L>c(n9XZe9=o1)}A0B~lQqU2UZSj)DwB{JI-CBM0Byp@Rs z7Ih9mk8i}1cNImRZHf}CtZ$Cu`y~N-8#_hGx&h$NFh$8O%K`A@!NdKd<%7G2w+~L1 z+fQvjar^GW?b~NILXql~d6*O^EIb{W9iJkA zF+wQ+Y`-1CtgkKuk+zXpbIs+TP4&v`nJWtoR}j+595OjB6yMGbof>~TMBZRFBuhmG zi%Ew{$ESLg9Y!eCcr8d)mq*1&V+P&&z%`{8;XsV}2V2$0^}nSKcg^#q2a~doul4ev z5}3A+t4j7eZX}8X2#d-{e^VG`J!f=U>a&#wP~Osq!SrV1-aK$xqq-}~4?wDgp z!?hs4F2|4TIjefB#wki^d)e$*CL&lA&ct%kck?ItgOH#2V)WLNt7#KH7be`ODic<# zJ_2@h){}`hK#RvyPbO<4RF+~$XFb{3XErGEOgf}Ag)`m6VFa$mDa!ab z&Tlb_{1zpJZ}i zrYKo6_&n73Ta@e*)GUfFtk_l8%%ip`N_N+0z5q;`ndK&LM$XS?#f-BzYn!5!&380` z%0&o=&{^~)pK=MU{+!qz0q6zUev1+uN+ikgOi{`!uUqV6o1&EW?w_LUUteU_Y$cHK zwQc6b!B%CPq7xg9T`ERu zqQdQ~^{i8r_{%Om4Q zXJN!}Wfj3Bz8nMgX-ii}HsvL@OeM35+EVPd1@c*`_G*#0*`_Et4*;AQrYN~p5!Mr|jSPK54F#Op|mEo3pGRPDqv#ONMb37=NIGSJ9vZ-E~ee>nrke+T* zN#}rw@wX^tA2_-!NcGA*ObQehp7Jkj`9uQDKKTkyEJ2OG9l|_ZT?Qg;BeUk3%R!s! zmDw{_78jQpC^?HVqqCk| zO`GtEGT}y5nXp>*5wN4Po=m&}T0EY5GFcm;(pO411naFQJNwLrXQ4o0@fnF{3Kx$Z z4-m|Wgn;BTwkb-2|J8ofOsT3GR%@=AE1mUZr$|5zN9s^ocNUyTKk&8ve|7pYrQU&cA1UUA_2mJQx+C^ zZ$u@B5a7wNX#42w1WojEWfe({@#PpWX04^Gb&8Vr0VrEd*QLM{oNpU2ot-XoZz|)_ z_|aM3Y)83TCL%Z#&K*)B?{%r+ER9o?lcH0Uoz;|VD3)JIhAVK!YTU5;y{+m)`cJTT z_RI62VTuy3i5O6{yD2;SZMIEOa{J5XJ~Tzyd59DUEZ!UR~bXY+iXXq=+#tjoQP z3Dcojy3Sf!j!V#V8h^~oUY;>DPEmIDdS*9ZH^USqS`i_v0a3oGwkb-yv2190PoBk= zL;B;JqQnj&X{PbFD4SI?TC6y0IBy9@>!};7;%i+EK;sl8=K+8-!xSaAD#Ds*whD-q zZHkiLTsGdy!~%;t$Ioo}rYLz=QRLaCD8b75<|w{j5}>D~l&l*7{tQ!; z?6L@eDvZfS+omXm-DM-NTu87ebbJlRH$^GL82-zou807@Ub}dLA zEl-1y#tgdkfon=H!iX_NrbzPIezve(3&LIVeCffY?BfPmPPuDkyB3t}cic!62@n>Q z4*gAG$jWDGoT7whp=L=`!owBU%#pS!O1NW=Ar04p{JI12qiu>(+Fmw0mWc=!h4c26 z<-6vSjJMGsng?;#ldJI@J{Kn3s45dyt18*iMu2jj;H)PTZ-5q$r=CpKMyPE1JiSSP z>;v0xQIhEF>@yplg#v}e=c2`D;rg)S0Y65a_2g_%Vt@l8d_yJN5-zCpoT#W+PN?ny5cLZfiv(idQgQrv%uDN4ly5R9|Y z+YnA=o1#?hGaJW4gbs`4=<7hfsYsPiNV#eTv-P?Z%%Sk9IB5yv+c2=ERHa1Mr3v@; zm2*AgOV;aBuxq}gHMN}hl{vJszZs#fb+4GOOQl_AW3@r${|u52}3mvZaoV^QGOE5<3xsW3pB4Z<=J!J<%>5_xY> zOC6VSigH?Xin6nsk_}Jeie&;VRc1|5cJ|Bjpkay_Bl*(=9xhAB$4B0`w>W@+Fj@Y>_X z-=f4D%Z8TsG{6?8zjQ+vJqG=B+gOz4No84KR$k8`QYy1?SoG| ze0_iS`gIWYEx-9DB?X_@pjp{uud_kMD4A8IbUq?NG5gZib>}05HrXq)Z@#`W2m<7_ zMPT-IdV8|&KYO{lqeo}oKC>Z;WUtJ_q(I^E=}_(X1Oto}LIG&gy)x^o3qhoAWY*Lc zgnEohnQUd#y)t{|3PZysgtW25i8nx>$5T@#Ya>`T$aDI@SyOiQnaz@g0)+=?^m|%`Ys8KV zY_teo8cDRC#ZQ8U0kuNYME3lzLHc}A|2yzQnI@~^9A72 z%tSW<0O#k16VIBktAAzx__c!v%j4~*swZx|dHeE{U+%d#`T6X76ST!_gqDjC9-?!Q zP_`~Cm-!;c7&-OlFIn!IK~P589A-p})T3xLPF{K4svp}prM!3lIA#C(9Jgj6wTaZ~ zS1wntxoDLHM>cFr#|#hNg?P~}XhbV1(2Dq%( zreN26S!?P$@hc;^s;kx>`>ofe(k`>XS|mVtbUGwP-nNwFE|NkUCn{&cSRGlLWXJfD z4D5HU<*W5~DS01&($#cr3OvD&0=VPS*_EPm(OWvCFLz+A& z8+T}=vf~6P5O}~je#gr7^0S5_sXn;ruNbt+(#>>c7~kK9Ui6(Ts$r{_4jbzGo7!rQ^5A^GaI5v0n0p03KSlndyUE0d^!MTSA0n)#vr)X zG%8@3{cr^g$yJfgky&%iC813L%j}se4GlLD(#i;$n*L^wX2&*FDQ_?vlcgeq2c@jn z@m)2@PBHSPZK{$yTAmyukVLmXa82n|7;&aPd+lYF%62^ncg-`V2b8jp<(F`S*nazM z*MpM%j+(Mafbg(%C~*p7Rz6hY^&mV9HBX{C9CaEDT4ohQ@b2C^^K)jkYiKkO!{@Q~30mu%oI}Sgoq$#|^3} z6K{Y%kEfj>`ywy*1@(eKl9ADu>mYTDD9u zDlvnY=^4M&!@R6GzVxK8vWMH|MZW3Inljm7HYQ6&1`o=S7yzK(()zY=@|t*zLh|P> zRfZ|QI25tmf;@pRCmNDt*Hv<-n93!f?dJ{KR3*D}zP=pR-)$iEnkvSqajLS<4~{ys zSitabjebSTHznEc1_BWA6u_IpI918+NX?_@;)-2$%}i?hjY@Xc9G7M$y2-1N^K%Q- zd^WXBRm$c&5Lzxmc!)+nj#Vyg)g6?`F5v5Twy8>RDjlf{sFDb)q6s;9<#o$`Y*Ur; z-WdUq;fB9|ec<&HkhM)!YPOl32A7T*9=ub)W34lNgOeH;*+|zoRVi*jFBw9kapKap zsY>xBmZZoM9aUesR2gJ;;{HoaRVp4Z5RPqAm8yMa19^zh;lVs>k?R|wR5`^4UGSbU zPF3!q@hwAR7~hJ4F&JfOn?SR#r0ba~v|gKnUGrtFsqe(EjNs}sy9w9IdTlE0G8?Q# z0)$6rBr)=Kk|gu}u(W-Q7Q&J%tw?r^FUf#?+VZssn_8zTc^`l~8q>8Y@C4`E22E#u zi&X(dJ_gkD9*1t@nfQ_qQs;4B%eTsmji3h54}jRApzs zJQEJ#QMpUls1*dJ%$lmiHal8HXsWXF5GfFNz&YeXzNyO2?)g;FI91tMmwO)*u0ykR zowc+qm*k#OStv$lWt*z(?DfoUz;A}BO0*(MGqWvh-u1?*O1!abPV=5T3$F3?#zN;$ zPhyMxJZAf;%4QcCjaHmEoHvrA_0;XvBB|u##-FO>JOF@ZYWGk-FrNjKvPW5#-h`rY zs**F=L1txQfrs60W9v2Flq7GvA9}W_O0W}sixl58(U@znrGXk30o%9FY=|NSEb}laPC7|thMVR%~r6AHd zGHb56B(zOcGJEDqL&Hskv@(KbHdQv6obh)>lwO4cai;Bh5bm01Ob;k!A2-mt=HyG;^`Kp^~9#E=E5O563Iw7qP8EE5qt3};rF zYF~1`>qN<0giF0Ocf8y&Fee(4W7k!3rx=*F zX6@-q>IMB5P}!aH^`&8|vTt<+(k89-5vg&ivVW)9Xe|~nJX{@;9N!dXzuPEef96t^ znSdpbRr4sixMKIMWl}w2Pygm><3zmqxbe3t*ZXtsTfc0RyVC0ujurR^2t$7tcM`Lfp3cLI|_d<&t?x>>mu&Vwzh$>@wY0aqt8ZRnTX(FI1#G=v0U|Dn;H(&_*<3pqEnTf)uh*5RE;Lc z6ZoL~RApzsJQEtGD)E{KcAB=X#%9}ZRdV~w#vPif>^wvY1RijX-}UlMRd#mIGo*2< zva_z5>(FdnXRXpKCyzlHrz$&p<@n7oRf$$a5EGb9wyov`p(@0xGd)4+^m7E8F*vBwc$*qdO*6>+2 z>WovB{N}O&S0)yC*g2#jl+j0}-6{A^9z5JX zT0XdYc>CaFx&74k6Swa^JPsF<*Akctb>C{_@>G=ny#5M@**9O}8H54y+6b7v)`@}8 zCZo2g%D#POLlh}snTJV%!sBz%nC$$mN@iDlNhco!XnNYFDw+MLOF^V_WY%1BNoZ5R zGJEDqL&Hskv@(KbfvnZ`yCU)ivoTpJGFVVf9eN$#y@l*CLaN4(u9DT|$uXkSLAO6} zP3ct_ai)zH#rKu*dJyiKXG{+$Wgk0-DI4-Iesoo`-=U=|5+FP*9ZH?TH4sBDF()P0Xu}nnpFmz~WeAk1LwLcJsXHA)S1N3=3HD$6kg5`$Pl%0KM z%(YZx@St>vY6>T>@{v}ysmk~~k_i(7bD|+RcHKw7owlh;cISM3X_%_)TOILdgF_#Y z8mB7zcbd(h#R7(ht3#6Go2u-08%=d>Qb|Y;oYxaw(0lRUkQnSr$Vl5prJa`@QAm3D_#($Jy zZBv!vGxU-nG#V!^ZTqdtLMXOPRf_vBF;%H}0K##TBWDA;vQ1U0_L&XjAwq`-bM!M@ zz9~tSQwTc6j0km|Y*Uqs0j4V9nnRSfJwsf!kI}+i^JT56?*t})_!feE;tiJj>fI~m zYg1{L*6;vbAjvY`r$+x0j8=G7-VTaOMydd9O`*ZxJpvPE{_7PE~eRlU{cj zG9t}{;qPr#AF@bqlk%TuLc>&L#|OgeRbV%TsV(+wQe9RvsaGa3{#b8MZ|1PCMi%{ zkEyn)O1!abYYK8b>mbezqxF{m5Bu&c8(w6@=aCpzM{~xO;v)G^(|6-&m>@PW2Y)vHvj+{ zrYhNG5dsU)oo%X8*j+XR%Y_6F!{}$Xd?S>C0|>6k54l5>uXoPKwCSuiTHaat-^c$7|CpWUw>-@kqL z;P`lZS>4_~xO@NhjrSFFf*-5lO5!vA4XH~)q@dygURxVVm2Nr$EoFMu|Fu;;SI859 zItO%e8*P$(;#>UmhktGtGTx!yDHa_(HfOE3UT?5@yE@q3FN94F3Znl($EXV8I#8&>te*k2< z>djs2v08M^ z-07`1BToQ~4(Q|tVrPfH>UmqoMZEQ9WW0kxi$w>E&AscOv}+w$I%!SZVgqaSA6=?m z&P-{L8U?)C36G+dd6L&A3qJxT4R0%hgofWb-@6^y<<6#U6RFiNU9Mg&8XehuS}tgK z)H=TS^Tj6feYIL{-}Yp_N%ZqlpVO~c=B8xHBh;+=;akh2lbz+g+xPD*k9O`JysQ^sUhdz>xPH`0= z*)&@`Zg>nw-wpDOaSd^DK;Q5eFJ(9X&R;C;G3g~opKt(KqdqEhjr!F~l|d#XpBOur zncI8SyT1gS|LW!H^QGWi$4NGx2MHe@(~i#$oirKr=d0JYUn!75E8u3oZUw_Fe2z{^ zU*;waEUGMf_3-U$uwJRm*R5dNe93ETP4P=35^i|5*GLiIrymEiUAIcx%*Je)AmKqe z1zBpFPrI8_l22^ewEemjJX>AfG8b!LNo{GX2d8i)d8IO6w}LI7c>yHIw(C}IVG)y( zZ899q*RA~ivcXs=DtI(T(j@owDQ~|Gc;E$b`pNSbO=(S1O&d58O%a8JFH1V^K)c6XR5PG(tU*>@U@+(%H-(J-L^I{$S{qpFtKINuhv z=nCY;!D-3_pzfMJZdKLxZ8RIHP*ux3O$rtsrt)WeoMHoJr+ft{9-viK%RHUSjR|`~ z-6XT9wp`RhRbm+H6eqLkGfx2ev|L?COLGWaF`7|K+iVVMt?)N*E_deeBV2)E2JrsZ0ZUw9zCG+rx8`^#p_ zLQ%n^(V@g~UPDU$BFt)?;v|c&W?%RWnsBAAgjg;62-wnFZzlc#xvZ<+OcqDz)R&ZE zO69IMI~&c0X{li0aVm>(O4pkmzYsXer(&GqWU#7cPF+Q@GHK1+X`JGeJOME3%4t{r z;GBGRz>m1zOm>(J&0^8PW7F}IU!^nJ#AgJy)+x?79?#q9cb4~#mpcX?g#&r)!jFJS zjZ>WL*7@4gGR4`q{E%#FoZ{@?YBp%g1r3i{hp@*v#o6yTid^FqC;JHQg-`Ndc-h4+ zyJlWBPI0o^KJx}(*4*qjc}a4vZaT90BzyX!x0cUk$_{WeiwZ^wk5q>|s9<8OI$P@$ zCpZ|Q93d|Qf>k(;CojEj@sM$fQ{FuDvziutK>q_orj>4)?8 zaQ)G~2%UqS8YdAH$;V`z;uLpK3s1-v2$10zp1AfS`EgC-6sKsw$~l08EmNF|Cm=Xy zqqvFO$~eWT+GsYO2MHe*(+h`G$vMTTawCD8b&7M>`4gMWO&XZK8KUf!f8@?O#VPp$ zfY;WV0v_Qz3=Q^r?_GO5V7_jZwwaCDGC{(Fbhq`FyV)c;fS^s}SC9?ZuUo-(xSESF zEpxF3mMmS`_C%%mx)p4hD{F1nt=z(hNt+8Dfl_-6H($5%`^yGnp{U@|7)g`d*QdPw zek>ZNICs;rOaA3ea0mzYI?EMi#{s4|I~(R(!ImjbyeLB7Dd<=sa|2C5#JqBeb&8YQ zU^e``!NB8g3<#$sM(3a==LZ`Mty7$xr>dC`&hd2?E6sUjr#L&C<=D128@Uu>FHp^zTWSLCjyonqwxb8-*af*}k0jT*b zQ=Htg17XfM#mVn38+V1`fk)reT5q-V=bYl?y+sjdoZzk>}-IM@9%@n7YYpEr- zm2rxbbp*hnWr~wscOW1dr#OY}Wh1d*Snz0cNP<4!e9_=@)yvzTwSLh+a2bW;f45aV zElsySxPSj{_=dAt^z)9hFNGxkcHtsU^Q&0@)mAmPS_kk{!b9@QxydSp6777>^zU7& zK4!JaaldaoNFERlwarTnAU;-oWcx$CfJnlQZ$(u9%~th(wJX)-)%VcQ*H%-7sO9f% zRWqAiAn5&`Y4eNxTIugzsy<>f#S!b)ljDPZE_NGUY~uMN)rYn}Xp3hq4#gMDX={P| zjp5sOk9Qs%E$qsK5S5NX347VGfvM} z|2zLnp2}@y7b?CJ=ioHDe!ca}tp-k?s$SZD(E}&;5X84R{kT;1`&;al`y`(OALudNuynH{(92S?-aD8UHu_gT0zPv-(Bus#m`#cfeUp z)?)4_F0~dvMm4cFF#)W&wQU9?(p2zwW$CYg+1540Q+nakGc9C z{wu8-Lmj;b`%&|FnSgL@J#^4O==-WqZhxW=LL~e;hl^fUQW{tN&Q|r&4BV*cV;7!O z7KdCVz<9EH{KjLaz>wo8Tui+$*G7PrlK#4uYg7|K)qgK%-CV?+pI|nfWE?Dy5r^L3 z%^psRoID_hk!;po|Z?Bn;JdvNs=@<{&cx0ZMA9Nv5Nc!!i+yz{TV zXkWChpM+IErCs&G{o}(^uH0X?+n3&_S3jd&`Q-i=miP8g-gvN7udVjQm+l|kQLlO^ z?>PCz;p@>ChgWZ0(_wzT)T0zfy!1V5bvxhta=7|xiN`eesY1veWIpJ4dH0p@X^}T} z`fa)GeT2pi*F(PeaCxM$B=_*ooV9w$;I<@Rl@Si!Wy|Uz8e44f#g9n{me`ISJVIl~ zzA1<^USgl~JwWm5`u?0JXkS=AkiWRb{T4P62Tsp_+C>N_3!8|KQXGi|gy31)!?Cal z`xwnJSj*d8Lqt48bPMLk72H2sR#*Bm|Qpt`0V~ygoMoTL`YxKNgj51I|^b?EGW8fK@ zC^NnIn_u8Z z9;Q7Y6T7g_(HxZt_wkd4Vi))fjU}eH3wg>Q?7}={cxwh>7h=so>;h|w!t4T{p*a^5 zyEKXUVC({(pm>d$*rj=d_T^<_m(J>%Bd~PBMp8M^&-4}|6Cb5G6iX)zJWG2zmQEP^ z7|k(RvZVMy24b-^so*0tcDNp*rIp;nH`XjJTePI&vPI7+w)g_YL_ek2+yx$?i7!hh zY(#q^{Vy>KWq}7MUR{<>Sn~w!3(L|8>nyHuzlBZ4>thq~QHmq6fDk-OdpH(0VIQM8 zCJWl*2Mxp~@DUn2EUy&ukRjNFdC1_l48bPEmSNarjJ~v6mQCOzG-qOAlkw1O0v@1v zby?V?d4l$ZWnq)f;u`l`*krOkHW44CI1&q+fM;nB$HFG;V>HKPVUzel1F;EwgvJgF zn~;YL!6wW@2DfDhHX*hQ!zPpA*#tg9b0!uxnGDS)-~ozPmxWE5Cum<-7B=ZDu5rJG zO{VK(6Y)`sBeAdvc$W5XENsF)MsrLSHi;iJ5Sze9XzZ}C33CkKf9-w%2S=gj`g7$@FVUy0{8uwe+WVSvw5g(;E5(}GvXK4?| z!Y1rvG{4P`tVs;SRb2+k5U|og-yV- zw1;D16ZSEhW3sSG{Gfr@1U^DzhlNeZLxx}z<{^XIG6b6tTZUni#qewbAE7xD3!5y4 zW)tuL#jDH0Ce0JHFDwh2bQagR|D)oxY9tv$yBD_ ze}k+uyDPRoy<(YeVuyn?%PRuy(^`kL+t-E}8^iq8S4(RS@{`iOvp=zTmHi&y>7Kb| z>kqj~-{X@%i20vxRe!AeNkj1EQ1S)b?|T0J-8b&tfBo>_?gy5)58l{$`r!Cvf7JPK z?)j&n|1je+ackG29~61&+BcCeem+)Ry78v!r?;vr-Ti|Pb!u#scXR%wXMQJtcdPow z?oOT#zkGDE{pPd(cQwxo+tr)Tcy#-zn@>FX)b^WR*xI@_Yd&=O-PN`2x5E!zf*rS%4XX2<6G6& ztwJVpJNXk^1g*B6#<1j~>Q}=i(v#oc^_~6KzuYw{-464Tl4Ib-s(SV4{=*0Rca{&1 z&%Y!4b=?={)vtfKYl^(;YgJXh2KPtXPuBnW>X{cuSKsn-`^@^Rx44(A{G@~Gtsh_h z(8-6xH~-JRslJ{5XWNg_jgt6vW8TP>g=<_=RxzeFWbUP zFFbQ2`lY9=kbk{;VwL4y-~NWM{tL?ocMoqLoGkA=_weMw!;`D+X-2;=uw8x4y@#)d zPxqhw<>T#3SDw7~Z6qt$uDD6z1 zIo{LvhF>YVb8vDHp7n?K!|wKPx6AAwpB$V#Jl=k4`-yYcbMak$Z}oNOiT(A{p9%{f zC%ztey>`ddH@|%P9qBrv_>+H*EVmmi_td#}Z|v3!a>rEvenmwWjNcl~f}?uD`d?%X zrVGYl)+{)7EI2_XPaF$Q(SlRQf-|(>%(37eT5zvku=kYzv#sin+<@`>#?-F)fQ;~P6W z$6tJSaOWK-%j1)Gymol9bNjW!d&}eHJC2vHp8Z(ut;1JOk+XB_;CQ+Jr7yj{fA`?^ zTl>@9(f$vP!XJKN=k@z{mUq7=?B?Vhlk2Dd6VvGvY-s&bs96{*!UyS&?;q{oJ9vG0 z%-AxnZMk;y?ZKAQ9}U>L`R|@>IQqhQS}CXd;y={CRsCr79n97O>cfZ55ANOBKMrp@ zc6(70_k{4109)6-v9{xDZdQ#QZ>s)&_(nb6!S++(wRdv=X!+9UsvHGZHSWEtHMAQ{ zr%BuSyYS2>;r&@XapTQrzi4vpZPnvtG2soU0`aOt@gx7#9EW?E}}!-szRaoE%=1?cP|nQ|KqLdkF$0tfHVEyl8x3?xRhMxkz!tEi3V_!U6)j9gA-r#P8+CX_L0%c5^ ztg+q4DPVaYz;Xm&*%vVpE~njXXmL5ZNVxoE9+#sX1uo|`E_ZIeqrv4L>*5mpEJzV- zj&_vT#QdkbfU+(hV{cekNM=ky|>^S5~x zoodke;hII?ck?|Bi%wQ7YWvly9X5#XNcMBEk5#co^?#VEjyt&Uf9s(V+9X4xOzmum}t9Z2*GzGzhK( zghg=bc^1Lc;BLrDfTdS934-H$X%4}plvE%%yn^Yq%^V=u97HjAce|d@f_HwA!236O z@Xj^heYOVgdv3m~0q@8H-j(pS2z=*S_(n%5XAZbPu7n2iTm#=aL|FKyzGvY}4ekc5 z&_I5iDSHzP$5Idu!=&I;VE88xh8FY540L7pkcfF8+_CMdM{ypwy{kn|WXdPB#v^K?4_1CwmZ|4nBT-PS%2-2i$ zPuKXqadY0_`+KzbUNLkPl-<|(?JBQs)mvtm_kY-`{vCD`F?4%F9>dh=E~LCv980NKAd_lS0pxI&AfHBRs=js>38)y#q*WdrzYOJCM#F^BWc zSR&&@ip65M4iX;2srPvdQ=_{WCfCBf>|7;Tb*#o>flSIz1&}crBSDsmjN>?op-ejF zQOwF@T+45mFpAdMfCkrJ74qs-Cd8Oy=u3kic0K)$j-CajoSKT2m<@YjEKzYH&thR* z2M7=2)bl)ysnJ~+3siiC-VG-RjwMNT{mw&}8r_Xq0%3uYA5YO6 ztFu`2lA2S2-XB5zEF{k|1W^oJS(qg?rEA#@6FQs4zd_T%-cY9UAjX{c)xh1mIc1*`EM$SvUC0Wp*f8P0S!n)G1T?2oAjX`? zEix$7Bf;V^tww?+I8$MgcjuVBr93)>q=uC1Na5j}9sv($YIGOQj=q&fh&3XGrh<>d z?+H*LR!G}vB8?`4`{N|WGwF&)fh*&=mfA33v{`r-v>c;U5yY6ImS_O?7cPmzhBttj zReG$j)jnoUz0X6L8r_AmgPApYtPv?e70evo$NVF=`)aWjb zg*us=6i(`7jV5bEs!|0nhp+AcFe6PCnicK|sTj_rWgfv$Cgoam!vxZ1k+Pu58l~DG z#vHXOEB1Znw6K(0Q-Kn*VK0z{9`#72$_jq%BjnWcJdml;T_8IMS)nSuNj!$Fm}ANgExuyQX|v;$f94 zD{Qq7zNznd_)??0@OAR=Sfk1sk@{1CVW`R)X;oQ5lteHdXRESCwd{t8`-=nPR2{^a zvj!NqBcURUsYr?0@{la7Qf4(F!l%=z?|B$gqq{K9?{JMZ%B&HoJ+l#vq*+K;Nf6uy zNQ_|8NsmHR&Z2Ai4HG|`g2WG2V$6x$A~(1U zZEP@5b~;=v)x}}6Yl5*r#UTncFs=iHhjHq89>&z@E{p{#?wJ`LYlK{5Qhq8BjJnff zq~k&}!@E07bj08#J@Y6^#iDpzt8$oV-z@M7nRGBj!NZ138hE|nEpgbc629I{S}2J_ z3~Jz8hX@bf)b~7msnK2d_K|KMhBfM~F{wKh7>3>HkiAnD#_2R0PBDy0+guIfT9I=h zU@XiGk5l0gV~%=ecz0*Gnc=Y2H9=TnV3lAiY_;ctspEGZ!qn(4garnET%KVXYXny{ih)bI>fvH=IK8Ur zu-6Shu1ca6w%Ui>)bTuWsnG?It5Ii-Nr|aIE}9t*nL0%-ni=MTBZe;Nk4G^or_QzL zhKcpfg03(#JWjPij5+F=;l}BO1=3-g8-QAs9xH6M54EY!dDK#)3!+w|#~PE`Qh{1D zGdxCmEHpC=0w@M7X@;u-Tgz;i2;VHg3NypwR1L(Kqn;V=_`774OXWz+W{0eDRT`|| z(>}DSWdkJpvxa)aWjZ^DEtnMuRmbRiy&KaLRCuG+1!T zus5J$Jd<9ynu%+b4Z{Y}IvX&b^mq9Aq;V<+V$AtyJ)d;*=KGqjcg(=oJ9b6)m-}c| z*tQbT$w^2*0fh-%8XFElsp&q~p~3??Jp&%l)aWjt$sMqS@Z|b;qS0cFNv){>^jFb% z(imy6G9xNRH0g{-=Pwt-wcLh*K(x*VAUgaiDfd;6ajFVp%=uuA=y%_IXM^ZnC8A5$ z#Z;ff>^eDw!eCa3vVwE_lI+y|JeaA`T`)`7mHW^WjVNnO>QDuk!_Sh^S;LUd(*)Zd zOfim0k32eh5yz7rUDjsR$iL0wI2fYHv?I63{pb(XfPe4JcQ=3^n*iTGvK02bCR!FQ zg+nwdXt{|KmXBIeTIW&Z%7CuL zHyTd(rs`|=R~M%HXMuRU{aAJB#+%mnlcj%60CcJmV$8W&<8*Ozufgd%`fyr)cz8mU zO3aRN;;ff?3YWv7%G5x)ju0Nosq1+tQ=_|3c4XKZdDeuKqzY30QN+neo@I-p7`~*3 zu7+u*AhGf!3ss z5gx#)^LYSMqq_k1awWDxfltwtYH!ZnKad-h?RM`mf$c^ba8;JQfNhP z84Sqne5fRpo>0XSa|B)>OH8biX@#x!l*i$3#8nLsWNLI5$d2+@Bh#9Y5;PmbNT;R4 za2p{phDlpJ3RW4zwHSv9plF@VZgNdh&5*_%sa=%U4I`tWB8sVc$-8R|_e!E;m1HY8 zw-3ds`*{>oqq`{f(y>OjH6g{PqBO>xE+pLQ?R1eqiNQ=->JbQKFxMI#CXOx+A*WIz z#vDh&+>U023-{V)l#?s^5e}$SNII*lHh?Q|I$grbc(6>?n;jI<5(+L9+=N z3AuEH+=fz&VbV!gW4IRWFafk#7#41wCaJcFF(-11oK!mqIUWxc37M*#m|c5OED^Fw z+7+DJhvL-zJc_B&T@*VBd8(0iO-L20fHEdzr0voXavUfTn5WqyugQfHc}bn>RAI!x zLJek1OB+m?^$ijSw<|B5yg-)dSS9XiB7|>rrLO0JOpWdW*+R$Z50_3g;;sp)LKPT> zoob}*Vmj4Kc*O7}4fH5Z<$Ssp>M-%XSzIi2s#84?V@~82nTsunH#i?fr<&@Tm_2&| zEHSZ4%r&WFga>fyd>+8m=q`Z!n7C(@c&br!O-T8vKr!xABSjaJRI_0c1DW*DqllG( zT#Iy=DB3I_3nPh>R7=E|qn|XqKHxBM*sGF-qhpj`JQ(egB6Fe=afPk+rPHb7c@R^h zyCAl(aQXwpQ;mpgLP}3XYYUAehP<6>xZp@46Gbs}N$Wg1a}m1J9`#nV&IUB7g-`JW zL*jc52D;R{iQ&We{K6>ni090r_ngC%5xZ`;iyTIr!cJptTrW(Z-tuz(fi$28@ z@^y+`_$eMPjAHPTW_biU8N9U)hl%#h;@w~4r5@8%F2tDgu3Et{zPZ~dIKBfbIL<%f zQCK(~_PGJLRcW-sR(m+N<8&Uk)aWj59h|GtXiZ6psle?UK%*5hbc$PT1jUdgopCi} zYvm0SbaR-yR9oZiqY$+1Am) zlWppD9<5Nxohj3vK66BWiG5M+%SQ^aywyQ=_{$b|lRjS=N-)mWqNG z&J>3now6^SDeeuZ7|*0FuEujMyI}wjt+N5OuP{?QP1Qk+IUlXLxOb*lbbrArn2ME{ zttTfT>B=zjr4CjIN)4dvP~icco&gVNYIH$BYs6VoQg13a8O{_>kvI#^6!%9|jA+s# zkIr8%h->u?6GxlH(1N;Snu>%NbM&GO;Q!L4F_kMZyT%yRXW0@vt2A1{xqYQ^>V6)~ z)aWjl9i_2GqctU!sDhoLup{L2l%1ilLky-E$D~!R#&NB`VL~WcX9Jj7m?@s7G9kts z{irbO`jYiBl`Jv)_99uTmsL8g;NCtYrw70znHt@TWTAjw4H+lj` z3%b1|7gNy^v*#hXSf$houI=OE)cHJssnJ~k_foI?xt^IusWl~Kr~<`MsWnwAwS;(z zflRvUY9QB2943k`4vAxlwu5%wt95%A{ozmBwL$>VWG(t3{m9L zky~UuOh>~8kNjg>LS{WvNw^$@+zx^kA>^>%H9=YESBFT{K)H?) z9?Gfnc_>q(yHJ)Y;hv;gBjuWr8dQN|m{f-xK9`W&P>L~3I_YW**Qy-`!xx8;Q*jYv zj)Q083D7bqrZOjH*8|qaArdtxuA_uUaq4~^#nk97ioJxa(RR&96{>(TCS)Y;(h+hT zQ8AcFA6*URTDrr;(Pn{Jm?@s63M0mx$SrbqZE0zPX|ukW;;`EdpktM|D{Qr=(baK1 z4`ga|7s!_SIGrh;X~bPKQiUooj5^gbr0s$;#hPCyo{S&$a32%*%oNWwimn+cKNTp(Ni|Y* zFn;dI#N z2H;ku(Q0CYZ(*fQ=W$Do?&3DT5xu9;Xwec=fm<|F95Qr@TR2k;A}EF|>5QxSwpQLS z1hrXwE6fznQh^X-j((=NQ0xoD<1NKFzt$2`bPD0Dc#6?Xn&Q#vi)h}vP|_^n3jx6pMe?lvLV$?o zawTxs^`nP#2uTGiH<7}!a_0zGI0K`*a3-uQ;X5T{jWUZCn~I?&_=SLwxl=f!F9h@l zRE%fR9#`YJR^Bjlwpr}_H+Xj5O9euVInUJW{H~if8g{;}W9PzEaoF$%Kv^Zv3R~@A z=8pGyC{v@mP?nG@%d_@0@~k~lgesUh$lmEs6o*uvLfITkF^Wl(T#e#ddBcQ#w9W=x zCH*Y{#i>9@V~+H(jAv`Ozvt$=8n}-va9>r_l45#~YL&d(win7F7B!SyhY3%~ssDK# zQ=_{$7V2Wn4;1fd1X_EfDpi2_Cn1=TKr16sVmOn&c?3f_$*%P`OdxF*DGS$2dn$#N z`g$qOiOP=>hs`Pp8RJubKP{BXdn$oe*lHgkr=I75OpWdW*+Iw}fz}?WMHLuE*Gqdy zpoOlN*f5I0o2}250=$1+;Nnyw#F+DRO|ds_&KrvTo*s&Az+13k9sIr~_zIIt zAqF)(T!#qH!>R9i_)??0@Fk^jFEOX}u|}S?N9s=nhQADX7|FAY;S|G|bkfxvT&r)G zxZf-emSCKUgcx)3VcgDs>IqO7Bi^H0C1%S*vam{^6}H;P!m00h7*nIWFm|x8MxeDv zYR_y0BZZag#WjIEkP<&z5ZMT`0+x zM8>Ilh(U!K%K5C_4v30CrXnV0%U&Q$C9z7m)r1HSw-%|;>_UZuPVVruNhcPv}3uA$b zdj=2Z8X?ynDL)ldjJneyd8fCH?sRmLS@0EDIK-f$JKcl`3*V0KS@;5@yYMxa#OWsh<{I@DtveMMhUs+3-YJN2 zI?aYt3}ezZj}BjkajnQSur((&ry_^`5un!!89N29 zAvR**l5Tkvp)zo53C@K8t}ygJPnAN9IqISR5&a$Ua;oXD*9}0fN}?6E+Hd6f!ZHJ ze2es0%D9LDOPb+PWXfDy%WRmi*(|{Rx&YWz4aAu9?iz=ao1+GYZ|}!p$tsu1k(kX! zXDGj2Ft(+k;ZPH5Qsg=~c*v$+=OIgt?m|}RHEVu|c&^c4%}Ft-sB(W8;TB1-)NvFe zmvq0QQ{qnwJ%*xPk=`;HM)yp zN1>~cV9iNUsUTiBGdxEUEHpFR8&5HsNhe&5=2~OJg#TutS(q7~r&1usoX9OQwW)`T z#bY`vlf^um3X_x?#!*e9WLNmkt zaT4Kqk*&s>_o%TF06g+9&*^^Xt_6 zJd~-?T_}@VB#Ek%@rH#)k2NPHr~=ARQyDUO+Q~vP4lzVx6qD|F6s~fPU5jm)2-+-G z7G{R$sV0aq=h?bC?i&SWJzug;rt&0a+g?hRZi7{ttl-@~N>2UHem^6*76s$4HgM zMn?=@(m9VnCvx$kN0k+=vjN~0R9V3gMMkYwWkIhO>#?Xl*T2@~p%hDT~f(ZIM|}sY@L;ya80K(rAUP z_GHo>@ADX@Mt3nxT3fx`oc6agLaha-WQmr=O}AdFy5t^;e|@b6}H-!Q>UKiVN8wg!dPJ5o|)l= zM#!}w<);F{use-(T(~>Ugi#D$(ld_^UCyLyRSpyFn+0B}t(A(07;_AnH1K-K?ljdf zFku`NFO+agm>OQB(jf*F>S5q99X>4fgTrB~N)GmdSn5ry zBwJyteH@(nod+>Bx*M|s2ZO2M6u%nH)`FCs3KB-chYKXxLSw^B6eZ9l9dk8wYZ(p` z>zl>F!rbs8RSPlZL~fDuQT2c^IGtYQbl9sByXyU7kc;`VN~RUI+K1iL@jP~^(Iv5~ zk!USQovFYsOrViM3k?o)(UE|cbjhPrmlNn(dc#EfW`S3j9A2afA;ui_Kyc&rR{gBc zS*816Do|p!8l9m0p2@(iN|hBn+K1cJ=R9ty(Is)Kk!3ANiK)Qt4_ANa{l~|Luiks< z;OOw+*4^cE56)6_dXI!=hd~G>m?d5D=)||G|6;59*4wWwZ-3#z{g8U^AKp7WIXt*~ z_`}Pi(=D8z3==w=1={eN!`~rTq7Y@N$A>#Ew-&rVu~mJSuu~=BVm#{K9{=N3b(P;m ztmNnqkgHnHU$q$C1+UP1=4OarNvhshlqJ~SeXat+KLR0)R9R?@7{*fqV$u=UgSeL4 zTnG?{(L+9zQdtmVk9v}LQf`u11T+;XF+2AH`X5VxPNZ2ppzB}}0G*zK0BCA>7g(f? zC6&Zgise|I#bcRNn+hy{bqj(S3AE5eac^8Dz$T4yJ+NyH4iiqB#m;|60(M!YmU^gI zJgHS_EFzmqmzbj%6xoSji$``HEdsLBV-S!{4eugb0y4cJPLUn!wRmKcQdEKLpTWpR ziY+u;+#fFqv`PD15A9l&!^G2OLHln?&`#Auj6G`I2RMP8!dXN$)iE)<9u(DydW%PO z9Vh~-(-RO-O%3m&+QZkef{RBrDOMGz4wH$fenuKDB>fO$DgiXv4(!|fx&vsUUQgjI~_UHKbb;`LQ6=3%1g z0wFpT9WnOgBf1?SEkw>#{KV|pi)fLY6Ui5k=sHXUM5q225KRs5BHBUDyRqtv2Qw*C z6|zkR%t-i!bejasZA2wtChd4VJ=f|UCXk|qHlTtQTVkmSiLocEB}T?b3&EM{p_pBJ z;VjnAi3*H|a~&uGoYNB!;7kqg!kM(hdKo*t$J&k6U_7QtfvO;ArZ0vBVSRlu5?BeU zNh>}AV^7juYk!z{x6R4F~KU1xgch4C7l}OQvLX5|B9VP;% zQ~wK?riOPh?V;#cB*w#;l&RSi9TI#h6H`%i8z>2wNe4cPSsBc`7fLQBQFO3GAs5T1 z=yZ-8tp1Y`68AQsL|pzv_+oB$7JV0ahKy^P#*A-$`TjnP6io1#OWPcf~c z=s2(vRFnR@9@VuZh@pqgqG+kXy_+hR7<&v2ZZNB68(fwMI_y@7=;$crq#}c$Lnx{c zT?dJP=+yfHqN(9sM3bajVC*ZMd|Ra(+a-mn0?fF<9TI#BW;DS(*5jTe2pNvf8JH8#db-Fs=zbuiR~h_7*4p^Xh{%FdhZbsdk|f#e;5R8 z7Bz>7NLJdtn~Ibedm_ikM5i8mp5U=(8O>BD<=roay;wPiFjS$r4iN#(sqY0eQ^UJx z7I+wpG^apT3B-0ud8z>NpF$T^A*-jnjE0(-7)dZpTJBM(_F%Y{`!F%ISr`tV1>!NB zs*)IcBFD%WR*yA<^Xb=a?^0zHvtf))Jta%+uMmSO5Z5sxfH-x&0Agx*Nf4{lV7sIm zRe<tB;loa&Gmd(`vIjo(Ylw0Eft zirH)8d#Mh|Jo&0sG@%jr<_CjPFWaDIfFn- zuuD4X(Xsbnw^sBp@v~VBEKND@rotn}9`%%S$K@p@Uty=4P!a1SUtzPo=uJH@pqCon zMXyj1tEQYsD#_O_DLWM$45yrTk>U$YIm390K|IRWdhOOC9wz9cx4{Mwur%cyEK%G& zsi&N$>zQ&6yH)}^If>GAtW$h7!6M|-=_v?+riOO`?I?;zD#h0>DL)l-jHjG;k>U$Y zIrqj@0&LPwkD{Q5oohJ{1MJOWXKBiLH&q=m_C$`6>tpqlvv^X9N||b)yj#ynv~}XI z;NxDZPLDxAHZ{B?vQ@gTT~d-NSQ<|`?;_n7nsV-smjv3RvmQlo543Al4--$D1#M9; zl8TQQd(?Una0124XR3l?b{!q3{6fk>wNC!ETL+53*XaoesHTROM72u(wM&Xs1z%$| zXGr#G{fwj{VoW7~CLQ)D0%SmsE|dUFnsN@7D3b7JwNuWl^Gnv$Vb5zy!KGW~5RjUd zb`vT*q&sK8LmC*~g|tvptEQYsDhU{^UlnYPr<_CLPjQW=oO?nmK{jc->ycgSdl-_~ zEUK17V53xq#Mq-2fyM40V9Gh{c1@U;uBk#WsxVy#iGbaK(5t2OyF-8kR=h=C>0?w_C$`6vsbkU44#le1QzzIgfdQ~ z^*7EUJ?lhZVY7X8bjSY!l&RreD7)!dB?23fQdNOxECL&8MPNd-B#0*6_$Y9DlI~jJ z!yw=SAvzTzG4|vGwVg@TV@XnIkEkAs*|QhXB01}XU}3Yph)(@4AetK9MYM;URYI^4 zDO2;ojFezV(MiDEMpOc3(vH`IxmNlxffOyY0oPbkOKg-1k{ElkT4H1vwGf=CGK$%? z7tSI<>m*^py}fWwPe6b(HM|RFGH}t$*y+^8NTmxKkpfkLX{IlRgkgPsF%nn_s!1z8 z0%H$F*YY1Go-PnYrz$1Jo_vaKXHpAcnu?{EJx7Nr7YZ2sqDpL}Q;Ci0FcIqK)c*pe zso`Brdnj6^5*v{+Rng{V@@*s&Q&DsqQ3;qy2R=Ig9x&HxASRGD3(O)#r=lgso_vaq zhfxc`nX0FlT}Q{61!s|>b+WNh9Vh~v(-RQjObzeC*-O!5m27N83e|jyM!GQ-MaO{^ zqk5dL9~)gL{g~L`PL)dxJ`4?RFvV7G$~pLbOAXkp^B#Jx6+lejZx%g^4enH^#MooVwE>We z-ak{l6ticXLhDlmk(_lxu~8i+0+dt#3s9zpccI)@F8572k5y{15h+m>chM~HaNm^kSS1h}k@8dlWH{vw~Rh=JiHL99}P(P~tYUBfBokj+yN zqbX-Dni3R~Cc7TRwZMmAf@q-)xSf)woX4pSiLpmL<=ptav`jnfa}xlp(|?7{_OWos z?*f3S;UxjA(tpt^RKdb<$~mO(6u@Z883a;-UD8pHASc(w<66M7@r%S%?oVW(?CuT&9-;8W3W9Uubzrk)qjOAYU$SEz_pQ_f?RooydDvCoWssLRFivZ~K z6a+w1!@Gbc_fJxloT@{{D#h2Bl%EPZ##7E?r1(Np&b@J!0GqVa^}w#>JPZ(`g*G6w zmZqG?sp?2$kMt1$^^~)CQvFlTVcVO4Y@PTkY_>PiPLDxAHZ{D9YzfFcpUxhubYEjq zk}8mmr<}(~_l2gM`{N~nHtDQKi0nzcYgG>eh0RL5CB4Wv6(2G7sP!V?1d5l>R0YNC zddMbrh)ES+*MTDNb$S8jKz?vWtXgc-2fN5%Y7t@ZidZH46jY-w2fHM|>g>0X~8HvEycuIhr zF9I9a>K=v~k}&$F>TCD!KRkKx@MQli3XiuRt1jJm)3tS>y^=&=!4d_K6MLj`TCEm? zC#0Bi4*OL?c~E-RiNKmL5ulviLbN1^Cf)cb zaC@rgTH(VWdR<^mbs5a7Lq>EeL}Kj82WmT$n(kG2Er>sNF{XMbX3t(ki{z{mf`!fY zCfuq21w>QByNGsl#wIEu*qD^53UbC^4#_E(;kXesl+CvOwG=>CrBoyqUbiF5-^hvydKQ88t6iRS)}Mxw8Yqx&%E(4 zY9TmN^%S#fZ^kXPxI-|i;9Lia0O#}s1UOT}yKr_>v`RKMA%&^}(+ou;-I$7^=zO#MooVwSnVH z-am)^t_jK_IqQUCVY9tZPW>-HnHt`OavwSO;#s8@n~)M!foG;ChSXv>;by}sK{V;T z>k(b6e;5R87NVso=SeD3V(f_=Bj4nz$DwC<99r&%QJoaCUvH`{R?a$sSa57Fnp58k zXr_jD(JT;i-<0!2B@mmC@>Bt2IORM+`Y<%*%*057VbXGsLbay@R?B^u7>X9!fWEdg z7eGu6?}FGN zADODuU=vb}DnJaUoI^HGo7rf}nG2E_#nXHh*yKW~z@#bXV2PrStp{^*mdjT!du}<+7Hros6^b`a@Q^UJ}b`-@^mEvnc%1;HB@sx8&;3=5Vlyh%f zCBP=_bTzQ2J-V;W0=6{e94t{}*6QJ9@c;^b4=ikZ1CSl!Qd226(ZVO%onzpU4Gixh z+fym4bYHY2RV3PY%6W=(Uuep?Kd=&Llg_#x+O?{O;keC$wxk!Crs5;Up2#tB`&un5 z0Y{}QpTmxos8;V96T8^|3NfidbsZ=Is?!q?P)!Z*qT0jPD*4xx6srnUW2s0;_9oiFXv(=KViIJN zwtE!7J;<*0Jq)rpi>f6N*tDz&EO!3@Q_f+xN=!#bDZh&__+^^biNL~U`^xH$_XSK- z!@HRF(6mYfHYHW70?t?jHr0y2*mz2SOxo|!>GuG+R`*;8AWI^!X(~cu>`{xrpyNxn zyu*H#P{s%?M9(@ASlDbYlvDo;P^N}=p-k%NUcOF01~^rTz^0^BRkX>l2yChqfeB%i zAewaJ^@y$&J`4gji<(75r$Qvgo_s{NGpTwkNk4ZnrFtl4&t61JJ?{{Ts)V}^6CvSF z{VyPz8s0^;hn!VHuqi21^TCXiU`WwPz}!Yu0%p>V*MqrM`Y?gCSzs1hVyPgBu_vq3 zMTSue!I>(fm|c6}EY{CDN!YXw6amiZ2?%hehIirI%f~(6^qi@5VN+6|DlpCT#gH(p zuP;UdD@OG!UmZ5>QHMpZhYiqpiWHrylo)*EQ*=9%S_sorEXC~Ei)oRfbtiK!^Mji?07qyryeu&08q)j&)jMGI{J zm_>?CMN5o5`4k-wqZWcQRZlUyPEH|1wJlP#PBu2J14V#ydIAERso`BXyD3^F8=I0s zRe@=SqLFS)MbU9!C8#F-_b3c{P+d!cH1YJvudc)fcdA@s>`_lS?{UAqTL{rqKE>>K zP=eO!#-??U2#8L-FCdy4-bJ*)$9+?gGnH;^N(xm)zKt8)A;G8jRcOkYji&_2r1P!@ z@~lTM7A>>^^epAt!4kzCv?14K$5G2bIqX;2;7-n04M;af=<2~@DN~{( zQlj3LWLj_PX!5|gjK_$S?)aofxa{kSJgkl!tiY6#I#$^#EHS8&6k%)`110GteMwI3G(0O z!b~tUS@Q}aiZwT7uUIo@Jhf(JVb0xh9-0cohHQIkFl2bkd5G%6&@E>HObr`mDR;XK zugZM{3_V*mtle@RHmi~Ze$1Dbf!H@#yG_Izm*lM)v8^PGh|QfjC&E{Zm@~dfBQ{lo z4cTtgV8rm2Gez@!F-EtXB{Vgxm?hcmR=g_k5m5hZS+RD@d1$HvGhbc~a<8keH?#`# zGAKdzTQy)?{TB%}YrsvpD+bINPYsyG)Y`I~ks4jqn_9Y>z-Nk~R{&9B z=%(Zq>*b86)~jrQO}CtfrjoBA+jbh(V0g=!0(Wl0=$13YQ^SZ^3O&+(W-GiZ;t@dq zY#Fh3%X!$WH4^wS-*O&5omnz8nimxHt zej02U-*O(JiZ67_Ig6`?VY76)-LO~XJc1!a13d#-Ywecvuvv8^@MFH^tSVrWJ7r!2 zC6K<=K-*UQMKaC?+D$bmmdzPYEnADLGj(4hQ{C5)ZAlG*HooONM0H>2mUA9hv1yO| z#b3ir@z=ApW^45##EBt&ZLSvq8rb;unO8ymPFS_A{3`%OS+AQKP^_9Wo?7)9>-ES~ z{xxKqRl|CXOGPNy=fO8B6;UzOFld%xkCxa>5WcGG5ug+e^bD-k+AZf{v;4@yk9J!7 zW$P_xQA>tf&P1~@1YdWdR`=Al5-=h*yHGcEpqMmgJT+-$Pi?y8JTjGl4cY$HVAuGT z^AMGQp zBVw~Q-ITs!)12|trq@}fO+{cswre#6-?#{jqJ6$hqarX7PYpw6>36dsk1|DI&sNN> z6@d{ahL}4NjGjI=SBpUn8J1_F_bHjOwmcIMEg7-^6FK5e}SVJZ6#vgl?ZlE9DOQ0qIXn>m6e$v-8F z))r@iqRFCHFi|YJDSySHIpe8Cud_It3c*Hfo9Z`aR0(D><{D8AV`k}iyD_gSeFX47 zTgI%V#73rSut+fa^6bi;v0JO^yZJ=%Q!;0Dfwq-|5wW>gJJEn*&YbbooY!8UO?6=- zwm~)6G|r2SP+?duFUIiFuxgfykHDK*qF3cVf>lQYJp*|$HRx_uDOvc@@?!qAyuQ4u z+cYnh5@=pGC#EjZwkomF3MR@D-ITv#)12|trq?;VGF6F<*f!O$e-`3xR3>IxqHCZu zjG1NNBOA;Z^Qsz10Md8R65T9X68Q12yzyn!4Mp6%dP<;rU2|5KXj|FXXay9-oSPa@ z%$YNun)BLAw5e=t#5St_B^uR@nU?4nSPiRY`FFcjuS$XhJUv@ht=)1SH7l0{ej>r> zfp?2cIJoXM|LC*6K=bk`f#P*7T3w)Rbz>yjY|!15zGBgw@zkP~1v+=jd1R^^8?lY5 zVSyGB+^B90C%6TeGz^(#-XpwbmgiLkkN~4+%aGMryIH6t@Y5oM27=#YP4|e`O9?cu zYs%{4Y%3HaVzZ{)l)qxiobl9@bBlA%p2wzYu@T#%8Wv|E;6~MAIN%mx6cVz)NaaNv1ZPAYR$^RoV(>bHWi4C z*!I+5$nci)2-Sz7Th0QQ8aB*Q?)FuARqi9K2+!84tle@RHLH>Ye$2O=f!H@#yG_Iz zjX2Vzz8}DAxB5svVbF9!wZe#E#7*HVM$8%Cq!F8{!A5L1YA|AW%bB8ie%g#~IZJ42 zSTReoN4n3%)~f;^0rk(86>GPgN6k7UfuBe)x>1;~Lk9)b_1yaQn3q8bv|_uor`^<5 z*jE2VLe0k1O}Q%u%o*RL0h{W-Mr;>qFkpDgnZkE&!047Uz(>P&Sw=mAsBXLC@2J8r z?UplfVu+&6x11A^Z*oUWlp2@cb*)$15fk_<%WnXPy!_9eR?x z4y`W7MD zzDdhA)qRcGmejCH<6F*SRQH8$Ip=}ZFm0Aux108=s^>exw6&7XakKcy!jD$`WiBiM zsxoe$iQ;F$s%_<8L~Pcon;KB8nlrvht2UK?joD_^uwLV8P73yU@Qq4ERH!rznq}DS z2ED55u^7Pb_*$*qavnF!k1YIX2K{C0EoV_nhHF$r^HVZubx&<80V86wCf(G5V$z)P z)TEU?wdt1g*i-^GX8TuzUE^EMV^ji$ZaHTV)39uocDGyhs=i0CglB7etrdZdn`KA> zKjzx5So~YKnC)5(=8TKL#?~S* z5l;<6X6bjkA+M@?1mHhghO8BVjhjVC0zc*=Fevzqro4GAlt6Q&Og*=UbK`}&Jll%E zNVr*3ZpvRVWzKkN%IhxA6H^h`m~B-J_Kb_b#?~S*1y->|Py9t-~?3kw?#DyH}My0{EY;Aju@?s2F4Xb9Uc)L}v%6|krMFTzK zx17h#DkXs*{}Nr_Nwr&(xH@@ zo9bSoDZuAxZqpK7BdTG{ECV0epuR*;zN5;;v?ZE2F_eq>m+1I1>SoNDD1J&2x3)wR z7%f#M0VwhkooGNdXTW%B&TB8xrm``UtZpjMf370f#Pj0&;&)3MXx}jSaehRibZqAQ;Tks;7-2% zIx*FaP1r`&ut4Jk_XO3ApD%B9LuQ$GyCJVCfCLyt13lxh_QX^#)*{vhhP=rH zH_`hnEY7wYbWwoOqnyDnsRP&&fRjJn5xAlY>R5JXPn@kplUH3aEq{NSTxJM zM~0XQxL4Ibf<-@DOS5*%dD1LW68JISa-MASmNU`%l&o3XITH*`3-byh%EH`~y<*Lr z@zk2v**TjE#3pQeYA|GY%Xxz8!_X~f0Za`WW+`{O4X?_51k68MHmu!po;0hH1b)o7 zoPpRkSi4QcJ_|-{D+wcFvrBSQ_=*v8#y4rirfRSW+l?BG7~XQGXrAws(Jf~QO${q% zNp`yxuL^tw)IVERtle^+H0zKAej>r>g&5{r&Ot%OGHs&sSukK*{TC6NHQ=V)6$9ps zZ_wyWK8o-_-O1b!mH z=(cOV4xQ-6P?1fPJ|*i_)77?;FCsQ;y-mq0*2@`BtykF*TW&d*2d0v*3EOrWOc>sB zo}h{^bjumSBr;;6Je5U+)r=8;N7Z`$#!m6shwn@uzI%B=IfA3p^V93ov~zuW{r*vTV5;ewunnfcT7R+l=$B@*(?{oD zJH9$SesDJZnM;BfD(Mn{)38mJ5x3jssv<|g%d@q>{uRwOn}tIHKR>Xt%}*bGGT7!l zZkroTG4o0#fx30AQ(aJPC0dgeAe5w3Q__lca>i5ZR2Edr6{qsRRH8LuyGMhGo}+;$^fmDmWddA5x7uWLrytOOGHIanF#XAi$G80p6!p3UY9$I0o#>(h(# z*)olMw82wdUVS7G7->!q)3HGp|I1FXE5;bMva87o6N;TSg{{~rXFRpj-&MSPbn)c+ z^2zm4wKK`CTy2zu+a-(qw^q}?^`waG|ZGGu_I-x zaW2if3~U6{JX>a}Y*y!-7;3EW9yB%6>#O7Q**7k(9`C<#{QmUn&H0}mOs|hO68X+4 zeM*~E<-li-IKzQY4G__0Q^|`qix^LB)_>qr9$3n%h=z1ZgUOO9isCi^)nlp6NS z^4E6zT@}iFN7(N_l$Ti&r^SB1xLRfhhrO`Oe%>PV%^mj4NqrU!*i=e2CpNQnO+~L5 zFlRiqrgaS1Qc6X{rc)XW_?O|Diz=z?SZY`>OJLhAcvT-GSa3AZGf+nL_caSPl~GMr z7JP8{dxHgkTDIT~=bdQx^-LJAsd#EmY}R<2Qdf+ZGoD(~I>u|Mn<9SFDGf!<--gDE z%BD;(YS=ByT1V!ViJVuJFy9e&s~pkNUdOWX;*-~$-EJ@^HK+3_ExSriYQ8fx#my^t zC~pXb=j15f<17?YAy8*ANU<3nxwzj^1 zTU&UW#X15%gVn(Qe-KNA9)8&k(CL}*+b|aUhvQ$SA*XfLgHTN2>xu`lSi=c+NvZQvT@639; zs(BIM{cM@*-_y*sS(+p8GhUhN{^1VCF5_=jPpC$$WpQJh0S65l`u~hPC!R zXtt;(%7UbZ<+4|!?Czdn96TN=<_VlmGh-kLXwnHG2Y#al(yxbxCm50B4|uikp{`0?@8`@8idyPU`X z9D&+`-ZB2jTthjvy;>bQPai*+o}8SXKbq~ocR}i^|DI9h&ff24mh-{lrTZ@|j$|s^ zDSn|L{rIR9k!TpP*2Ul4DPDSQUMBPAlgHDm(}%N8^RF@9yXD?tap&=LHamVa?Tk8u zFJB#>oKls=C+J+@6B#+&KlsGy>=SPuzZvE>&{_P2o#JD2EgW@odicI{{_Bf(r{_o4 z?_W;aXd~yx5?#bcVwljH-R#h`Kz^{Hi=(cljow~v#J!)W)%8Q2FR1E*;^Imx{`^WK zWlkfXN;PtMF{2`^i}N=6P&RS57RK@W8^#eIcq|l>(pf7cKla~c8Ywl6^jnO*_v5v0 zzR;Pdx&eXzBFYBqaR9NwI4K>7RYfN1&Ds@;Y6-sQPr)glY zr;WDvrCR30PG6OIQ#L`|f$ZYfR&wv>)HJH|45Z$&{^^CtyDG4sR~{16(*UN z3T@J6;*`NcM=7!|0C+GSzWPiw4RsA;ues>N!sB`@=nlv$i``>3`(j#t7SpY7 zL5(a;8Sm!>a*ZZLwL~9EHjoFJhDzJG>n=dG*n*CV-wdkB?(ETrCt}yONI)$w<7{jd z`UCNPri}ObOC6*D*00O~*Dc#f4OAM)&wC#TxZU40}anJAcnBan*3R;$y+-CCt=!=TL zpz4j-yv{cp*l8>`^U`IU9T12%s)&}PR7F~Tpv>e4e!jM$H#$G*u%V{Rf_4Ho1?5|p z9H;woW<}`pc&eHRr#nBjmMSW_a9!vW z)Ag4n%jt%lDGJ99^2=1~MtS#&=He9R7zCPYHRbkS_X_$-LaHk*%TSJ5#|NTlfRgg` z{^g*lblc7}g?CASoOKtadx^)3Qw&QLrNnA*ychIW{D)jST9To(B^mCu-c1QY3mTlK zewLi38+N7^bd&vH^}g7m*meE5ZgH_cF`YQOAW$1buf<@?!6^ZhnvRxVC?|VARlB(# z>3lKo<{A;eLMTDxZ7BV7K|ASoooNR>6_RLYU30nJG?!kSR9LDdC9`H$4Ltr1*NB!@ z=pER`-Fv;(+ZQ`O*ivtEaRsdve=(>c-JCO3ptsY4Dw?s!2II9}oJLrxBPCVyUbllx zi+{k?q2&;I-OyE^;HPVK9Cm)htd6G0YL!slZNX=mO&p+?uz*T79h>!d)_~D!lifV= zmN+d+>d93Ay>jo(TH3F5zT}X$DRLUy80U7u z<7dk4DLS$w4BBl&Rq+MO2|U%gw=$b9K=HFM-KG~Bo-APG29{J0RPfvzCH52z^;<^G zia@yZbyCCDrp%@aOsc7mFtUqI9_dXp;+X0WOb#&y^(9{ z@eBbY*9LdA5@{KNo-IRgd)@V_7F(^Fa0*mra|GxVC+`GvBig7UehO4-S%tnetI+N^ zC@Z{LO<^nB%jOr*DUdqTrZWwFo@roI8}-Cv2g>xx;dkF!pV02CC@;NQQDLjp%jO(# z>;Tf_>!-4-okwyC)aMxoMsuyE+~R~zfhY~A=&ILZ@YbP(D8*V+VXM^3<{R)<3CywX zqLiEh^?9~|5nD@9@uc^_DG((D747s}Z-NAI1c4g~4W~f8Y_0*F0(GCpDNvth85p(N z7HRopbl4*CQy@wQD#}r_49mgWofAQ^an%*!6sVWYFQ8MPWSVu&B{>D^^UMMx(}cp? z1wRGqX_sDsyYn4E z7;RM&PJw#a!~r@50#s}~Hse_XMym?DpfGR>)YGyCeQVaB-C2<&v!J><_twj%4A2I& ze43{~eV#C2R0j=cE6nn05s06)gh7AOveZx&PEbJ`ibL+LmrWO-_*o!m!wb!TCkq(4 zHPcF5b{pj0D6uCm{euJDI@_HsDAMA}E9$LHHdSC!O?{A&O%mjGAGe!TIs=|5=%qB* z`pFN@Pl0+`x?o_r$grlSYT*G*g{@6Ln=L?F8wQK{WScf2o+Mzz>QE2SGd~5QWS(L| zWy`INC=M=pmEw9O>SuEUrj;l@z1X-3@$3Mjx{V3(m59=Lijp>8kHFLzi;t_au1Dx+ z^8+S#S086&b)~KtjhyI!X9yU%wVKp%c|8Ip@)QkOFE6a==xVVA9o2V%`q>--It7Z6 z&`6DMDAAU#Mm$@<=&p5?Uyb-F5T)`IMGaaE-trVEBgtB0;S{K!%^T!gUqfS?UX^$T zfe~CwQ#B`k3PjmFMN66Y5ZWF7h3+NO`opRHeh5^qsFsgNVh5!%54qB#RV9hkN zI|a&1FKDdZD)qBD2OK+qWcr2{r{ol9z%vev=4_(jUNAUt7pSjg90n~)6z;V=1uFMN zDb|{*w@TT31KuiuIo4g2l2f1|&o=Z^Vl_*Ov;($ETDDPgGYgE&z=~!!=oE-De~L!RmZv~K5M0_iiQyDzkj*KeQ=mY>*5b7xPb4r>RThJ; zz+E6p^eO6yDh#wxz2#YvAhMvi`5QcoyFi0%;sBijf$=vTn++vgw6}Ckq(4wVE^!1Wth{v8Q;pVarfh5l9fak<@SsG{~k3Osc65 zGO}sWd}^#weLduv0!DMqbbfGt3N+Bt1;e(_955x-q5(?E(;K!ngKV|{ZEYYqn>Hby zBw)nWDpE~|Z*2xzl3+M)c?u+o1L#I2G$e5b+1!9>C5n$RHf};ZJHV)}^^=Dr&*V@# zPw{h>=M>jmfMW4+RYogOy5(jYQIor?k2A8mQc=`61sd@T!5}5KR+Bm|KLw&ho~)sf zEkn@uE)X-eT5Lf_^(oLGnFPW%*zvU!S@tQQ#kPLBz$R!_JK zG|c7_&?yjr!Pc@0Bc4@YL>pDaPk|`0r>Ll>?_k@V0%c_uv=g|gx3a@*egT~VsZ(w` z(=g(h21a$@OcgA6(SkC4ik|u{Uxdrch3i76IBb=M*_;E89l+Vbh8Cyf6llaV4vgk( zq9HnG;9Dh111h>2wBFB(q5(?Et#|k8XA_3md;{Jp0dm$|l#)}R5zjU-VojEW?gCLl zP|?n?^;Ss`2hfeOdfux~fri;!13CrjJ&jYK5zjI(YHPhHP+XkY%s&Z`Jg!9PKt(wt z%`RQ1Ku~O4b@g4KVK%>jPJxnX)-{*p6llaV3ye(Fdk39?yFf!NvoIRBJOu)R;L=uU ztzU#2W^)SY6ev)zy?Bi$5*Vo}i?urVT_8&IDFzZ%7%0x>cFR*BL1aO3^(oLW-HEfr z0XhW&<8L}P8}qCIqt&Pr{vzB^%NmRiS|0RB(g@-X@K6bkLg#GC0Bu0~PxBOL%o7H~ zlFb6RwlP9X%pf}0!D1DBGrWWDG(*|6zycn5!!t*NR(C3 zM&P1eiALGnfN3R)k1;lGLOeUbs1BTI+YxakqI903q%GDXFqOq}3~Gy3qHKP^jr^fhduuXvq4oV!OLQf^z2Gr&djP5pI;t5uj6`c#cMD zd_#%0

PkvjvRqT1VLhqu{4Nl*&^S)#`D@b}zzZBw1@LoC1xqd4rtmYiMlKs}j#3 zFoJ7UspiB_fhe1&XlcAgV9i7l1aBlZoC1xqxde0y1YoeWtiqUQ6&TS*uK6htCH52* zS+ff5j)Sto3(AZFh-`iVodT&dZ93C1;hBa}O0~M*2ucK}K$Pi|^)$)6m(cDCP+n+3 zKQ+bmRw>=RvqTV%9Y8XDLyJ>#3N+yvhfzv%;42&Fp%>vsTE=10ZN2{yMZ@)Apf?oE zjkEa%yj23^th*>Br$7^)ZD7RKDpD3DbP7ZXLB*16r3r2C0`<`)t!9K5UW6NGa}DSe zsQ)xhfhIi5z^JYDqQrILDG;Rt73K6=eisM|rJd0%DGhgl#@YM=It5AyS=U^WQ=kdY zEHE;S3lD!0jxv9WMlzX&cAor+BupZk_E?Zi=+H^6F;}#@SSXNj3HPMK&#(M|$HF zXu>lEjOIKRg9gDf2b9@Ulw`Tcu%@PJ;Q>w6Tbp#>&DJ5bwPCQ>{3%fHfF}vYDX~ET zD(1ms*JC9~Kr`FMZCM5<3N7fTPU4KyeKu2zX(ftJFE(sKy#t;dU{u#~%-biw3qBK1 zE3du_G|A=&&?!(nStEhIp+sBr6sUK=vjvRqT1Rr$Atq zZDkdD2Ry65h&HN-p8`>0Pf<}OtI+Q7Co3~xMPVyD$>tZ(DUdqlhBFPl1D0nautVrvyCixN5onkd-@nz{8R zNDxO5xRKCs3N*>)8qg`w;3=E}^$vKJfl*tttC)^>3PkBZ#dIbuPl2FNfMW8LhEt$P zx@Bh@S#%1N5VEehG*5wgU7lHBWa?X&@DwRO1)6A?g|0QT(C#=0h%0EVJ_VX&a|-Ab zD2%>$0tYKgKXvi9RmR} zHV~ZseDUJzC)4xm)9d$-dfhl}AQ5Y1mx{1I{MtLyhwr}mw7QC{7xzG?bEEYe@0g%cG$G45&#<8+3auGjOVSfUfc+ z|NFJVvRMNZMhg_Jzuq)$Db5+JG?mg@YfANS|1nUKmNn>Fvj*)>8x+9>Mb2N5UvZmF z8kn?GA8BkLncXy|IAc}}6Ex6znc^J6N=+%@wVLt<^^bw$ z)3ONNNz3QoG6D*kXhZa0t(BHdDCB%#Lungcp~^V5u+mmau`##skAR}ID0LP7jr~D-O3?J}j4)Ur<`Kh3ckTdzMJW zp#(^`Z)l;q*IcMdkdI(QXKU)BN8`X2ic*D&y7FZVZ4Zcsh6ADizWMWHOJ~weclS(L zcnbxTweA9SuV}hHkc(i{29`8eb#MMOS?c;IZKx;38Z!;6#Y)&AM>`_ro0?vv*Un?k^eLx3B$vo>C&F#c! zK;EI7(ixc1jOKRncY#v0yhEAEJG46~0&)V{pj#>{b4j=E%#oopqcF?%A~=v{=%%!) ziU{hOf3k*|iI!m~M=ei{1aWY|s|3$q?pbLi-KI0Gpi?72#>S&HkW=WU#0JUh?1EFH zu9j0MtvQ8u_lP8^1?@%oxo)~IXL><<)6vsBH7Y?mfstKnD4$M1LA91nC?_qGZmM{K z7TS^nfQ=)4XjYd!7)Tj?~5M@ek z;7P;d@{$oaHKL@UyfpQ@EpHh?v2ZVDAhkX<>SePJ=+r2oWL=|4PL28?@4)D+HKVS+ z;M9ongNjlzZ!fevH3D+t`kH^?W@#?HY{mhd8iiT57QyvFnt{=3)Cf2=qI965k?6w@ znyOo#EeS#k>WNN`dfBuBIyC|lY&u%&gPZ~*)~FV6YShzm3jKb|yGD{cg2au4*70;N zn@~V|)A7?hHR^+O0wcR-JQtqh2h3JG9~Y8iUj%zSr?z=+p?HW@|Zz0mwNp zs)Hg_b zrK7RofGH}x>u0kOcnbxTweA9yoEi;4E`m{8Ye{kW@U9W14HX61auIEhkDzYmVhhF= zof`GCnFw@hG=3VVMgx$8VC2@SQ4B~vHKL@Uq90ohVtJ*h-K8Ta!CFD}sZl?heL$y1 z$?)qMO>$~90C@*SXRR5~Rwz{fMjRxt~o|QP@Py!r8Y-piMPK`z&AHj&u)(b^fSOjLup!>go&b8c}{wQA)q%sSywb7r18Fx1+m8!)(R@of-ua zwidySK$?NkYSak0YeeZlMI%v$K~r_hvn4@jL3hzzqhU6!fKH9T1RIamKu&=Xt70$c z44oQL{!dYhttR1i%TpssYPDW|Gu+G@W)lis@QNorxO{wlJ^wV+#rghhdiIU|Z$6n` zy}!SqFfGX^j6gnt(H*pqjn{4$^f~18bZ@BT6Gpb$gyj<{Z`Ti^zTWwA?m<*tl(ov} zbI8lfl*4Q~VZVjSHofkEYyu;=)|KXr3i=*0Q-tD$DRRwomN z*<^xEKlRy1HifDLR?Cyp#)YF1NG33we~i&5?4@3;G>9Er)&nn zwmQX!9-B6%F-Rf|Q@R6l&ecgvB8n=I5e7HgkG}o)sD4R#HdB3e3AdNvPfe~yB z_Q2{ynLpX*jcuuf+k5M}yIOKVl?8p4b{Rd4ve|@Q&fPUmBeoOrjX`38QQpR={&OBn zQ1(yJR&-M$yZ-Hct#*b#)p@;@GX%n{)%MqFwPkY*Sx?xQX~SXP7^E9UDap0I)RpSD zfU2}~!`NEpu!fgpva5AepG}OinFn+>(bxbi=h<2kVhoZHjBKNxz}W=l0~M1R=$^IR zJ7{^a1r-H;qR3*D%|{H4Zm`zXrc)DRkeXn0>oN_`b%HG~r3e*e4flSbc4}{S-taiJ zL^NC>D)G^+zEL(yAp{lRbYepbR&tay2002wbpT>EqPm9;X(@xKC~TA;w0&qH%8%3QV`|LX{jQjX{2bkz4CYS*Y$6U`sX9@)KiQe&SyHy_F!hprokkc$Cdf zd>%W##uB)`5iJjV8hME^$V)JKYvm|$*>ZGHG2thXzEvR2*?cYdg?@|p>~ z`o_yB-OaN)26T2o0m`hnq-z_v<=KTKwO~?F$-+3BYCt6m zXogKkZWEAaVAR^=hd!!Dc|k=n);vSIs&7$jwPxzP-8h?MfVzt0fD{9xyVjAi z*ep}*AjGEJprWW&DTbDtS5=g?%Idt`INhx?CkI1pVV+GdJ|Mlo2(ERdEZ5ff#gF$0S*0shDrvqL_fR0wcGUq&hIz0#PbZQPHsN=0=oAkh+o7Xm!eF z6Ktzfd}@-_Paf`#tP zVJhyM1_g2ogt}g<*8CuY{85IKhdeuB%6_dr%LLi8_rGi4?u2$(XGo=J5{1Q zp`xs^?Ws~;G+ZGn@zJT$B%7rWf(npsUq2yUWf7XEO8odB<2Bk(@s!29OA#AC@EbA^*%~PfR0mx42wSF@evHKO44z^M`?6BP~F@)FBw zS{^__5e7vT^{LV%n~{L0O3C=^8cy?6seb@+5sXk18sVwZM9D?a-0DKH*YYq42r8%{ za1h;ynWVdU7JI`}r7+F*qB)RyV8jNFR2iZhF_b1$Ol8#eyhxA-7rjns{RN6ix^riG zfu~Ahd2w+zz1rW@ zngVW6^bbHLfst3MBA=(A*Fw^}E|daP^fB4{LVehM(D}Xmup5X0$UaZnA6}IXPqNv9 zSGfE)ykK-eN`TR&ntISm@fU&2wUj{jU~f{({P%W#Pe$gReSckbFv;%wKlAWxwtsmw zJvn_izlb;6zrH#?pMB%v>hV^w-vzsUMr@F>7FaDU3zQyAv@L%(Dm}=;SG$UAlV32J z;?J&JQF)Nv=zlsl_@)hmZ|gagwN{iZ$nE*J?GtB2Ln+RvoB6opmusu}!Os1L)mWqy z6cETS{@q#=*-iZCR%AaG|T@&YkJvhDmvl z-KT%vLd2${r!L>4b5aAsbY}knP>Hrj?^>@YD4RdC%!1~MU#)eL-JgFv9elDS^`?ou{?*2MGRq*?Y7hk7f?E3Wj{iA-D@5al6l)73C`p9%Z z$Q+fn6YoYladqUqFV-%r*XgP*D-u*K`mpaV53>92Pt8R?n||}?YI=5Dt>o{VUM@2u zNAFEfAH8!u%Vz6eWh!{7c;Wu@of~&=bZ*}LmD|N1XlQ}&yE&n?7P5o$icT6%TQQ=S z4|0R={dld5FLWkM7a7UbI{A~eIb`?VpU`w-vWZ=$kh{h0`?ux_DZbXw1>bL%2PuKI zF0^s~ISwGyya0@eU2J-}KIu+{^t_0V{LwH{!t2UzO?)_Q=o9$*bj+Xt-m0c(B0S|701 zM{@wy`hc}QV66{W>jT!n6a&E80I)UytPKEb1Hjq<%?nr?0M-V8wEy)jSv$Q<|?Kra(G;&gS^h9fE8c>xz{!^d-1co2-RwMq? znE%A($Z4P@G^LbQQ1%n&ht^Nlit|IOCu_y|p|z8>;{4FcnZ~VrLRvS|p=jA;KXHC& z(PXW_@SGo7F53QH173YVKPK#yQkJd`|6X%CkO4f?=L+d1K#rdIClC|Re&>ETQ zhYv+7WI7ZrknAVU4=s6tTfgwu+duZx_YWxj@X_nbCr8uA z52hz4r{|B39#&IyZgig8`)Kj<>Fnsy*~Nq7v!m1V)9cQS&po$ykL6Wv7q{-;xVN+S zO7S5yPVuwFD{sB?3)&prHeAOAeS zKM(QGBmDCi|2%=bV{RxJfF$< z(3CKNKMR9k-YIhe2f@5k{sgwhyi*nhw#K|;`Y0AeV^BN@f5yC1TnJlZ-YGtWtugNu zC&Jd4cZwI0-Z3E*H=>~_c7#7;-YJHJt%WHu?-WnMK``$WSHjkqcRVA-m}m@&GvUve zcZxS*Ys@>vov=0Lo#Ic}8uLzZDAGF~n&MG3G{vOwXUsdrrm!{UonlnjS|AYfPVp)n z1oMu^px70SLGdg68S_qYENqQ=r+5~&#=KKp3tMB}DZWK|$Bik@MMG1p3xCGEQ_Krn zW8NwDg{?8~6a&N70y~)Z2^@oBVl*_x#qej$JH^MaHRhe-WY`+>PVq8qjd`bh5eW85 z1O~ji&l5*rzIpBl{tN{&&mh6pP$2U>5^Oz$WW&gnl1u>alxIS}_TjGs_$#Ji1lx|` zuTUe9Fp4<lU0m_iKzK}Z4x8emw8GtjS)VTv+P zTRaId#Cyr^*SrMSJTUjtLvjD=kxBT$7iR%G(9=` z)-okE+o$Bz>$BJE|3rzd4*zZUg(&Mq8Mw3#@K$l-{&U42+bMdjrz4Kggs(2%K3pXX zXZw_0JgF0iGm(g2cM(xSBEByr;`-v<>G{$1`W zA9WGZhlG4@N=UQ1TGWdV90N!L14m6mRZlDuLH8~k2pvKq(ARv`4bV{%$O%Oz4o_X+zBk9Pn`G65=nLP;d+ zFZsyAp4fz_`o$%YD7r@OkrnpL==o2h{N7gTprNYTc6!zpMq71>5&n1=Eha;E7 zB(H6_gsCk?QzdrZCiQJCnkupTM$uHPzDCfJS?ZjQ0Xy=X5*vi6Ht36{df26- zXbQ;7=-hqLR1Z6M6iun5>2BWVx_a2nW9);#B?B>^>*`@Ak6jlaN`;p%n(AQ}kD{sY z&dL^}sUCLjD4I&hQDx(crh3?|qi8A+V>CElG}Xfn9Ys@sAl=M-PP2!-IdYn%_RWP2 zM^ioQ%aQd9WT|lSxuib!T3ionWfI>82Z?e=akqWOtnE@G}Xs09Ys?>UPkBci>CV6xua-GB~5qpPS=Iq zJjOlzMN@t3YKQ8Wbz(#_20GzZw5Bd6JWiqX^n`*LLc0$D1Yd@gB#J-LaftwmD< z?8A{uVv^cYG&R7k+k~ktMpFaqyiMxcS~NAp?i)o@vHBW8OJ=F_Iffy2oU&`~r62-3~W=QKyynWtm{Q_AkoO~{6ggv>5sI5g)BkaSGOJb7RQZzNfuG@sEEk;u#?7U6t+gdb5 zpQuM=wJ4g3)z=7GGE1G)F62kVeEszB?B>EG)3QIMnqxP1&C7N<%_1q*u|r0s{a(CDf+-Jk`Rif5^_}8 z_@XKLlq@0!MN@$oqrv&2sWEowD4GHU>1O70n)Go{BqHQA2Tw7YqOV{gvXJ!)WT|j+ zx+Lt$O+;-inxc;zB1s{a#3Z$)Xo^1Jhe$GEYKzeneW(tRWm4bPqNxdX-zb`j)z=7G zGE1G$G0+#ukhF42Y!If}pf8%DFLEKWP&5VPWpwVoXo@~eg-AouluDZJ=6$Yeu!~30)Zi&ZQ}h`dBq0<{CFH2Gq2OiUxm<+( zItr%(HAagg0J$i_E*-^FfFj+{e#1!;cIe1&4qX8i-86KG3j1^90R*~KNO`Hk?%YI` zA*7=Fk1knZPmUZDlT|jR^5P16a1*XfL6uj7*om7oXo#xl3oR}UVn2?;s#t@K=q0n} zdG&~0dQOiG%2Z3F;Kd^``Y;|^YEWPWBxZE;^nDQ*ZP?AD$V#P6clb^N=wgSD4FE(j z8JIa;7=1ttsS~>}K$QxyKj%~9M=9lqCvVTX@h7>HvsFmpy%C3g4Ng#oHmh&iJx;YKNnu0~HI zx)QFDqUb82N0p8@6l{0rGTqs4+RZLYQ zhRG~@PX8d>D$V(a24$+moY7SuJ9iXa0f`yiyfeBI?w6wIN~KMA_)Zrl+%v^448$=R zm^q^>;gTtKVSp+XV$SGFxM7N-tMSu_u7qo)D7s4MQKjRIuKGAiN6}TF#%OlV=t{Un zilQq(k#1~G?Xy!;2I8g6P@t z`_rqVZu|(a{RL2}qjeF6-THnD!cLN^IlYy>0?K_pB-#q^b7kjJAX;sxaCDGOGg`^=mfHXiK=L3ZpHRINkL-UD*V? ze!&VrSd)n(oY9tWUsZ5rfGibe&S*=xtqP;9@U*w-9d>85CEQVk(N;o`Dj#RGC0t8| z(N-YGXnf9StHq_%E+9!axnFjL$ik)6dsT>hd2w+@nJp7&{*Z5a?8IN;)2k5iUDBZM zm&1!JpLuvT+rPY;o}50Mf2(q~e|>d)KKsVS)nl6|{*-ICLZ3~C)O&XRmWBSjSneT3E}5}X$K-Pe~-j;6Eg z)5pix7gzhI*VC(oWR5N=Wq)>YaVfr{{ZD?O_;ltv`|SGo;k%t%U!i~Qw*PSPuH)mN zT~DXd)2Zj_)c156csdO|okpHcV^62auH*Z6-2+#zuB%t+>eX}g>brUkT)l>_UL#kp zv8$I08>I^yr3)LS3mc^i8>I^yr3)LS3mc^i8>I^yr3)K97dCn>Z1h~%=((`bb77+yJHpVV&j9u6myRb2FVPoRL#>9nxmvy9Wop zZe3ru($}r$>(=*m8~C~neceXBZew3JA4(`i>H$~Rhmx)jC0!p%x;~V2eJJVrP}23G zr0YXT*M|~HmU>V^*-}?GA4(`+>YB}m63UpmX7izha;C1?d?=x;sb@Ayo4UIBP(q1Q z*K9tNQ0mk*n-3+FJax_HLkXo%U9Ob$A3Z7F^ zRa$=Yr?XYmuhB(|>&^$Nf2Cvpmv0mwsO~s+{?QwU|L~2wcRC9m4_|n6cJbi&>}Ym9 zy*zw%ajpCDgXzi1>G`AC{(Bc!Cr8up_rvc?8(d$!J3T+Te*bbBH~UoD?DArEIxnP; z8+~uusCoZ^6I1{9>OR6DdU{KJJLV9*?IiS=*VfzG*Y|%9(JM`C{ME%=8~yQ(SsOj| zRn|lFY+6iD<09wc;$N1kMxRPb=&3OLBH6ingyXB^hi{V(69EA(cj$%Xjae<+gGe0^oL87S@b6@s}T|U zL%T#I2>qa~jQ&!4mKCC3=uwf591UxqK|gF?jgHXIT4&KO^r&V)=#Sf5(J%I?Wl{y$AqF?M&&4SPm zdL*hw=vRBm=ud?o{fBx zmeq&|{h?hV5`=!xRz`pPQDGbUg&q~@2>q;m2K}&oH9A5+Yn?^E(4(3Gp+9bKMZegm zngyXB^hi{V(69EA(LZd7e%QWZ1))EDFgJ_-q-8ZCLVswNhyA+FB$#gmgtA=D^?Kt!w09b=ucW!BO>&Nc8N$3`axS6{qaXr zZ0HwyRHP&Hv-TPE!}ita2>q;e7X3nxY6gVn1GF>MVD$0_^fbD5;B4;VU14JJP6PE4<8)DM{)sqHB=Vr7vvxxd+0V>n>S>)&M(H>_BS>N3nTp zNSan-LTd;O6Tv`h0F4*o5MSCo6du7nG(JyzXnHvI(EOz9q5ZMdL+u&VL(2oDhuNn? z4-JoczEFJNWM17xC3JW6VmW$1IQMktgZX2~$LB|nk7w`hzcnwn6ROp|XH~5B^5Khj zZ+32VZrt6ue}l^PB64o-?XudvyT$GMRpWv_jr7Br(!B;Xv+JwVlj+IP?2JkW_ZRPx z;!3(#RhbYs^P}0il=O@!|lZD@0RmHL0ld2bfuzb` zR!7hEoz_9$DsJ3=ZeBJ0_UXw293L#6e|I{)JUTgjOl8!cd#<3D(Q{8Pb2Acs*ZNL> zE$4OA-LBjJn>)p8Z7|bVkygZs!ox1J#j3O!@+8-f$!oDL{hv3BFRUTG#QA$ryN2ss zX>WHE0NBJJY!Q&$J#_Br8nClxa`*R`i@9Cgx__hdT)pbVYmw8tNC^C$_AaHvyR@*6 z^~`rk+_UOkR^&;#{T=o$<@(;GT*tdC#j}oMSV_3Mi9wimiOsq;cFE_Y?&Z0;dvS%B zJA0zfDgK+Cb_>eGi}Q!a*ZWmIYDR{#_*g~jC{PT*;_GP?mAy%^^^Us3rB9xW+N_96 z2C6%7I4th0wq4I3-l3iCbNAC-tq>#@|M!ieY)?w(F5OfknSQO@U1w=;*EvyRyUxv! ztlag=AAn%4Z#ARQh{RW6d+|TDjKIG-t!_DX?kxY1y{JeknqBWYvG5dPwX`IKY_kl{`UK& z?VuW_T8Ki>M$twa|F~s>@8{o{UQvRt`d9W2FqV@@_Geiv`LW_7_dncu;qFR}zr8b0 z8W#VorGdQl%_r08FHJjlmVaoMC#rZ&iT)I$^~1#n@4wWbb?;-vi?hYli|zFXikrJT zJFCU=`QlISv`c+``S_Z||24W)Gr#k^VSoS6ZWOt%@-+0hBBpb9^{4EofExW*&Z^zE zZ6VHf36~XpmDpU0q*&+e#XWQKZj|wQdzFg1aq#@!OU0eHFORQ|=lPR^JH@wzE8SPq z*~OD9%3tkQ?pILXhgW{=?%{2^&pqEr(d2cuUXFc!x_Du^4o2wMb1(lCO5I+J7cJ>7 zfC$S`B9Gme`jv^?ng3zFir&~W8C;auD$a9hELu^%c6{~v?DccH#eRHt`b*Q3pSgse z(~GauOG&ljSF`H|oRFf<4ku$bcWGZFgM3QnZ468X{;;9#X#kO!{1^N%G5D*7mInw# zQvSG8F8F(fnP(0}D!d@UpEP73sZU(1LsXHB-*5^8f6fpUfyLz?tZqT1>n}U$!k;%p zrEdaV^V95&Q-!|~ZlbELz!0g*e`^=p7XGpclYyGomG!XW$F2;JrNYb`ZS}C*N6}WOG>=Q& z#qh=(ZS}CjN6}V7kSZT&QhhX>Bdq>SK3q!q!%!tv+_>CM9k$+UjGM zj-sttiH(pZvjlqGL?1i%oGKfnsrKlNw))u3qi74r%;@mF(N-Tjd=zb|#Obcz?aHw0 z$5sHsnheyu(N-Tje(cHsSt`uD(bfRFeH3klS1z|2Z4I!)N6}V7kSZT&QhhX>Bdq8en&B!q!%!tpRrECM9k$+8Sb)j-sttiH(pZvjlqG#1K38oGKfnsrKlN zwuacvqi74r%;@mF(bf<&k}M^vH`*FvSC68tK#tM)ywTPWJ9!js0g`l+^SaRy_VCD! z#xFK+KiV2$|BiftK$r?EuZtRC?`|S%i_z8y`*q}^n6$PQZH=%yH(_h5(bfn%bdwUd z7;TNPOGnXGti(o0lUV}YZUQ^^oGKfnsrKlNwno^^qi74r%;@mF(bfn%d=zb|#Obcz z>&nL1^vH`*Fw zSC68tK#tM)ywTPeJ9!js0g`l+^SaS7_VCD!#$O27ezY~l{vG)OfiM+TZWo2UyNRqV zMq6X-*O7~2(%M?IHNo!OgsrVcTNCWiO-kHiv^Bvl9YtHQ5*r~+W(oAVi3xV@IaM}D zQ|-|kZB4M7N6{9LnbF~Uqpb;c_$b;^iPK%b*Og7M>&I3A!kP@!ywTPKJAUlS09h)` zywTPKyL}XGg|EhJHQJhBhmWGIgdkNu6xw9olaMY%VPB8Jtw4{_`Upr)ny{Nk@fM&- zH@d&#CJH-wLz@df-b)vv12#s(GYe0HMbtIZ%1KQtj9)dlUWP>+Qe=?r_Ba!s#Q{G<3qo2 z1r`NfKx#(UPY=pDsl%=xMP4d(`Xb;qg)UwMuql8TCj&RPOY7n#0J}6mmkPDN;8rGH z08sE1K3Zl9zUb*#C%mxhN6}Y8kt!i~^wq^qA4Ok*9-|4mqpvP@_bB=TH0g%t_NQI! z>ybZ=KZ^DgqOUIY^vER$l&R2iyDH)StBI~{MqefN@W@p$d2KKHDzRfX;cL6mSBrbF zOkUfJzDn%eQS=ql)rfF1%c0v_2)AK#+HBCKTBSSs60XFe=nF{A==$B!mvAo@MPDj) z`XbqGzJ$xN*rfrw zRH(V5FX5&vioU`}o3|T%3D;v$^p#MgO2{33^>H+hqOU-Y(FEPmmv9>vMPGm>-SFK0 zRJaX`{Av7A=%)~U3D;tgOAshiq2+c}1MKHbbZs;G5-!OiSHS-j7obTuJbD4!cVsNwj)mWOkDpAohhJ3N>m&`k z`uzvftLy#g`NNBo>DB((@%z)Oqpm=j3NAlkIJCDBW(dHja@Hv^oT-OjxMvc}MqqA? z30G%r7&C=nZdD4GXKku9oIg_ixKEWM?BaVl@VD{yeud(CoUB3{#HqIFj&p>ov?$JL zIE^J~#x;OGzUEXbuFx48)moFMF9&{mz!f`!J%E5G6JWSq+XycPf@=eWsc@s)d`^Ah zrQq{epAGzHBt=&OoFw5z;IjsjR2jMB9^v{dihIZ;+Zm!euy1jB77?YJpWCsHvDb%= zHGWe0DMVw!JzD4)1kzM^xm}iUqt-^)Hls1&UM+N4Ok&%M#)PZ2HjHgI8WS$p+ElsC zXl#t#JdDO-Rkm^Xey4#QepaCk;#AvoM`OZOTNsT2xz*v0`n4c;G$vfQh0&Nwp1vHo zUE2gN2ZB9-peGY!xT7)Q-mT!;0AVWJ+|ihD>lQ|1;Zbzcv7|d16Yko=Xe=Q~m61Cd z6Ry|7Xeka*5Yz)7Z9bJ-yd-X(!%B1dsQHPd2w+@=_eC-e%VU^cI>Zk0#vAY zle*;rzC8Pxhi9|>%d6?h>BH+&`ULv^_0{qD>>C$Xk8R@kGtQk&d4Ox{Kb@mTm5w(F zUmjrR{5hE#qZN7GUwMGjGM~3lq?^zmbE^!O(SJN8X%UMYshXls^161!a_ zZ2HZktLfQsh2(cmFPHnzqxYt#kKVbKKW6$T-NfJw%O^BqRI&V>KUjPw`?1j3_3^`Z zJGZ_<|IB?}9NwhQeWMpocqyHDDV=yJop>pocqyHD z={fPzbK<4v#7obKm!1riI=_;FMTIo`cAy`op|Xx z@zQtVrSHT`--(xj6E6cNUItFQ44ilwIPo%Y;$`5(%fN}3ffFwSCtik5ybPUq89MPY zbmC>`#LLi$m!T6cLnmH_PP`1Aco{kIGIHW&apGm- z#LL8qmx&WElgX|tG2T5m@OJEaJC@##J#WXpx8uOuap>(h@^&11J9?2rsa_xCx?bdT zy~yc$k<;}er|U&d*NdF47dc%oa=KpRP}0|j9LoASJ9?2rd0*#@UgS{b*EypXIh6Z# z&gexBWq*A$Qu^1~(Tf~P06S;&B8O7I&KbSPp(L<#MlW(G4eXrJiyTS>`(~s}u(P8V zIg|@_&gexBWrLkFdXYo4N8NJA%$__g_dXYoAP8NJA%WU+HbFLEed?3~ey97-7bW~7Xe zi5#kr_s-}<4pp9eXB>X=CAv-Vt<#6o`>%cF=k6cu-6?k7?%X&i_KM=x@wZQBog4R_ z+q+xryxF;V@1O7NeY|-2Xjw!*d*}FaI_o@N+`0dO&dt03e5arrAm5)Y5r5<2(donE zvo|lUJ4OBHU1M!|O0G)Jzj?z@vi=%fExGP|p!!!j`U}MePOjg-oRUE9sI2e(r8kNX zR5wFAzy8MIuf1{iPG=#L!xtW%T|78GJDOcjFArZ`T#kSIV0vy;}AV|#-4Ne=9B5w z`}>OvfQRTEEuqK!u7Rz6eRuj0y@$odUtMXo(I4M5w$W2x06s)d)y4EQuD&j1l#^TYER^R*<>mj=Qs`gqa9jg#rtz5E$RRY>qte~gE!dndSCAUNLq*v0i8WBCU z71|{tL6;9fTN(ZFG@8s7ObShPBV2AGWVXN9bp*v*;IkR5Kv-$L+1?7yDGRAoPPCiK-F$ z)m}3C%a-Ve?JHIg`peZjJX!Q7Evpd``a`=!BnbVWt&IL&YxD~}D$)`9S^Et7Vf$)y zgnrgKi+-U;H3LF_+}?_Qu}?J%LOkwf6t8mq-8ZCLVswN zhysKWtx(j?m9qXVEY8sAfRukK0?(FZQWsLFflP5>+Gg ztG#6O_gkVLwy#)0=n!?(9@Pv8{c(FM`o%ugEC~IeN1|$kezlj3{y|Ih!}b*`2>k;y`jeK`hzR|mT_O^M ze$ZA%fBcDPTih@7s7OcXXYDiShwZD;5&Bu{Ec%5W)eH#zaeFKJ#Xi+62>qZ(qH2VG zwU>J zqa*aQ)>-rmJ*pWH`s4Oi^oxC}SrGa`k3`i7{c0~6{iBxXhwUp?5cC$s2JT2><> z^oMqdND%r#TN(ZFC(>-_7kX5rBlNTO8T7;U)#wQQtaTRsLXT<&g#Ng_75!qLY8Hfk z&?8YbLciKeM*p}a`eFNu6@>oq$*(N>la|$p2>qd5A`*mt&{jr&{D~SH`h^}9=?ML- zeFpuoeKk5lKWm*uztE$a0ii!`Z$-b@rhDiGJ9=Vg;c;d~zj= z{-k9!B0_&?mxu(RAGBQv0>0uK9~W}?!{=dzE*0^%4v_)CcV5{H^A6ScHXa%#U?XYK zrJ4gi#T=7_j37%`qf<2x!gKyZ5{K|fVL)CDm4*5RImicJf@(mAiXHegZa8{vm;!vV zHfdUoiSUiy(63Iz(J?h@tXM;Qd}y==p-V+R6glgVvj*6q8Xtm=HP2gv(50FKT0`8S z0qo_OX_l}?r)nN(4Pc%c)|JwsHK^SbYkF%~Y~C7@rq!6x8bZTF zFwh!6s@x zTF7oQX(lcFHO(RPA_M1Gy3s-`d-ra<>K!2{G=XBvD{Tv z?#;P5y?%Z&{Wevk-Yn>Y6hGQd>0i0JIN7Hs)t^*Psvn)6)8pz?!;4MY>R0x#aXnZ3 z=^I7ao<4d$nyB1e2hO`25ladJs)&C3^rRB~-Uo~4-Q;GEFo~t$V3Uk1> ziW~QzE9e3EK|5hoi-G9T%-WIGjbcrgMsIO%bwT*eo#MyZ39PYNt;kznfD*hnMc~#J zh5zSf@k47#Fj*FS_+85(ue7+k87ORK81wP~7lU-o*x55J5BHc~xn114f1^`{SYb^F ze#q@_Bs5BHr@LY)V*H~wi@t}L&Rx39 zMw0znxx3!--?n?A(6-&1;aj=u*8~8>T=!~rWPaf>g!d~175}Pr)c)0JbuqGYXZZ)i z%|=q!?0VOwYryIJ`sSUz7mM3(F9@NXqI!3+_}i^5cgfm3I8zcaor;6j0@Q)QQ&KBb? zHt8QIZtm{vtm5P6_ddn=`Ec>U`!6;4DgNwE+qBx3kFQA$U!&VL^UKp4UJ?F>8%6$$ zNe#WO$m!f&{mJmDrp5>>H`aFZw~&3?G|r03N_?qGDP#8bDrI%!;Q76miaT##9$y{L zQ!wQ6EE=8fR^DFRYo`R-?=Nnjo}XSHm7QB}Ui{4E?&rULefi{QI=eo7e0+UzwSRg& zy;^*-?dWoPb#!)dak>Ac+3fVu`6pl9|HX?3_v!vN9RclMQ||u$LE|EQ@jb-{tFe#I z&W^5+ADmH|t(v5Gc{yos`S*VD;YU{&Pbil(y__BO_UDu=>ACauYKY$MhZcXR$0~lT z_z3GYkD)3)mi@{+m0v7gjE7nF?|eP}v7%`Aqx`p{Zt+vaM}K;H^=L|)v+Cb(J$d~2 z`0D-6i}BxfZ=GGtW^>b9L|{_D^Od5Q9e-=OU#*8r+ljZAcPP`CK0i<;*P0o~JDuka zKk)eCWO`Qo>P~U58gicAK7aI+7pLb`-EsNkdLC~sjC1b?SIboIuuJ#$pF8~iqwDG8 z%cJA&=(ykSjmpWxVSjWmIqn^w^vdqZ$>72G!Ex`wK0)u=L|*Zv_+Uasr>H5~jkw+w z$z6%>hA0HeNzSHodm!j`C^1m{>~h-kCy$S4p}smX@BMM$W~54K*8Zfd`{XkBrTPE1H zq*@?M4`{uV5UP0#^*~8s1FC`mZ5MS7)+Z)V8DAK*QLTZ{uhyH0=vR9!$VYkt-7@GG zdno9S`{dCNTPx^i?K9{X+P9e@gMKzc7X5PV4?;g|t)L%N>Qq2N=m%|O^oPa~Xb}37 zc1u!Wl~(NgmW5r3&B7`zwcCOLSVso>MZCazHL=T*Em+(YcWRmdtm}i#m*d+yL`INh zoUZ}4IU+N#&C##}K~BK2R|}ay5@^1Z6GV6L!wIO+k|fuAnrlvB)&TnDtU>IeSwq~(U=6UjW(}-E&KiUct>(yC1Dhjn4baic)&QGp))4%%#Tr24 z1(cvQgw_#&pfw~7m*k?zNPUq;jEvM?3kFbR1iD3>pnZYZW62V7F>#-!A)s|1wqA~I zY9E`eADY{j7Zk{X+XTGi1=uX2_ynejf;-AGTi7 zkI)ZZ^byey+REq;jU&(?^e63>q$1z1zDOgYU+pEMA9RZtLFgBIELlRHE$)*?KWwdF zleN#FUufTEh79`I3|aKc?*k$9!`4gs5&FT4J|g--TN(YKaReHK{-oWKRD^!@MH&(P zYA+f6pj*TULciE!$r3_;+$WEI*jm9RYo9^C(7w$K8T7LmvgnuJ2SVtFt(Wv8^n(|D zMD&BUGWtW~2s8-&NxLPf2>t4dG$Q)dUNZVYw}=shezC`rC4~OCPagfSwSrC7K7)Rt zeVZ9F=w~xz(J#LbgwPLLFX>0<2QT`F=m%|O^oPa~Xb}37c1uzb`qdX{MD(k@Wb}h> z5hDowVvi+D2>o%NJo;g41)HpW2K_?&HZx?<&t}M?Uw$75p&zzh(vQ#&Ui1;s58BG; z4~-+xAoM5gmZT!|t1r@s=vRBm=m*^*MiBbN9!r)G`r|%%^uyK)Hd*@&`i1swX2_tQ z&5%XE{5}vuKWx3EAE6(-=p&*Zw3X2x8b_c(=ug@$Nk!;aU!)PyulADB54uH+AoPnp zmMkIk$9?kXhpiQCvi2GD3+>y?kU>A2A&Y+beISH>*m_AnLO*!XM?^npE2BR&jzELZ zpR`+&D!mLLza^nQ$gq_sAZ)KsSl_b9J_liYRU-agQ1KhLir=*;fLm0YRn_)K$8>AF zdXPb?I{Pnfu1XsJlbeVC(ari}M9&?5_y3+*3B*O2da+>fZ_bMsB3yKCzg6L57gkHI zN=kXr;;ss=-Q^PTIDDp`^^dO|U!Be?oT-9(Sz&?tRIUbg8l_*Vy5yg; zf*mJmNBFrc*p(5Qe2NQp`iaGS=MSm4_-g;`_&Sb=MI?41An&1$B#Q%~91N8a z4Duf0D2j*>M&ogO-UE)HP=RpH#&CEKVT^Stad;0#iOYMy`lMFg1CF4`19^{7aAXKi zkoN%nWphE^BMgu+668ISz6$Hm=6%7(PXaflKQ1Z@LS_`ciJ7ybF2F zJcL4rZIFx;$qLsNgV<=vxDgvISwAA^2LutKBN5mELn!)zo*-268e9ons=xrU>3jj> zY#9b*4CaRnofY<>H;GpTdoTjMi8_=l5(qLdRz^6`n~1|GVuEOm2XcB7ID|q6gghI> z<4uG?)~3bdO&BdcZvumgcHRUIp~wV!ld!gXy$KjVHXZ0q!U!3If!-wPtgw%EHgR#Y z-I*Z2q>dyD1nq3VP#MXfosBq(A|kY0#N*^03EGvw5fmy+V`SV&POi3?#6(HPiI^zK z`Vg!c5KV|K1oHw$P&9*fHldi@;6RWE0{vxkK|7l;K*mVW&L-)ru#UV(Tw1-J_fSWY z#euK|hRO&Ac@J?EMMMar@i;#30Y^}%KsaY(IJ}22#=4X^ya%Ji8r4gyhmJ`zMl6`N0P;XyayO6BN*g8#8DIxA@327 zUom707$AF&y4Q7-L;Z9NvRb;_@Ew-c2j-0Y^~e zfxJh!Oyl$(pucP`$a{nVGDd>DN77ee9eI!VYRP)uLmf#L2l5_ZsElBc_Yg->M1;IY zJdV$Mz!4NGkoRC?IJ}22#=4X^ya%Jibpj-bc`d5>@z$>}{nf7x7+_Xq=I zj0Aa)q_4s{@*eTkob|kiI+837uL6HaY9^o>V(|dsavbiAd5eCQ@3GyCEUxjt#J>siP>v<1# zBv~BDdw`)bfN&wQsVF)j1rgkfcNHFc@H>( zA`j#}!euz8_W=E6b3xuC43IGrIes5{Li6w8Qn)@%ii<7gvw>4LbNCOK({NW4U`qD))0e6&dyC zeP`2e9$ih(7GI}$=kziwsFBv|z3J(rcdqfRk5?a^T|78GJDOcjFVVG-^v(PKb*H#} za{c}#J&ON7?iBxWr~KzT|KIxG^TprUDZZz=e@9=uy1u$Nn}524zU4bTfB1gq?(6jL z>(lG^h1+;9em%Omm&$JW-Kjf!pDk`5Jv=*}%{u>f@$t^Tb$0Ra{*NsBHNSiyeY~Og zt)1fgs}9wNIKEcC_RjR-yO;F+3)1q#@aH2kwQrn$d;V>buN*&oS0MS`f@IFDNPg=N zuhFL-782#;b~^mG-8Q_%-`7Yf{#fxr`cBC7dU|wmc0wNtx|)B@cc=JMJH<;3&@EyFpXGz^oM~X&>-|D?Utk>FRou`t8DhWU!Is=UGG!W8a-#w5L@KZD+t-X z3ei{tgwH%2U(?sKu1_z{X(My?jm0;Bt3BkQZH|T=eD})}3#-wrA@(GAF@V;fc3YS; zR2b+NSvLF*i`Zq!7SwXwsktCP;KAn0@ogQVB_YdrWInCth!zFg9C>Sij=r*Kfo(Y| zg1$cZlz01gSImIL*ocE2>nUBC8-Ggagn|q{c0~6{h(XK2tvQu zW62Uif7~aJe%M;UCTpKTztFzT3>oyZ8M5eyHzGvz!`4gs5&FY*0|WX&TN(YKaReHK z{-oWKRD}Mx$lQ*8wU>;3&@EyFpt5`+tIJ~lF<*kMT{Wyi#?VsA@s+6 z^5}=H6>PHh8T1S7+su$bKbs+oeprz%q93+i(vQ#|%DoNf2W@5ahsF_T5c-pLOHvW~ z<04->`qf@C`a!pd5rlrR$C4$4{ZE2>5G!`4gs z5&A>9vH|^|t&INAI06kqf6{JADnfr;q-saM+Dk@1=oT@8&@c8_vV_nd_sOFlwpOso z+Go%&v~M#*2K{V?Ec#(Zu!w%xdPzS*e<-&#pdYlA(H|N|ph4(Q+AT>%=#Psm?dVr~ z$>;~&B1RDU#U4wR5c=ajdGy293N~5$4ElxkZDz=zpUsd(KdfjK(GOcM=||`f<)Q}k zgSImIL*ocE2>nUBC8^*2^2GTpsRwm@dE&3$Dqc;B&c6HciD&lH5$k`}qs7N3e(U+- z3=|&v79K1{ZOHPV4?nxNE#Rb>d`(rE~tn5=#-^lhNY3h z($K=vudHs)FsmK{o=|0n$`iEpQHH6p!qiAMkic&7ha^3X08gmw1n^YvCQ}`aGwe(h zcE*8dnXlrPC3ePu9h4st*;#CQQ-&r@hJLnKEElWGOh?@Vg`VwstMXtbcA99laAsPCu1pz7C1|BxWHI4#S8C?utNa@ z_H&?)be@A{>?Dx_XD1F77(1p|L53WP6)@xgJAw@*Q37Wu4igwdra0j}Q9me9Kw0c? z?XH={X4*+&1kOwxBrs-7QNp)HW>A=b3X1_Vi})>Ff|3A%aZ|P>5i> z5t*u`Jjkr2r4}4eZ*iYryR|10BP1jq9H4`Wpq51lip{0Wg$1P+7Eo_P?&6Q`h`gY< z0BwNc8#H(WlwhsQMF*u89Z+urUw=&I3xx-06QuG5ji}7U2&EPyP;UccV%S4M zD&Q;H2&KBpMG2)AC6uNF!XFVKh@u3v84_6%3@aBWlv5DAGhg7!lRBZ9f)!h)U_7W7PEfxK@Z zIdNX#z9G*1nl?neTzJsa!UO7U(9s_kH$)^g&KP|9L1he^Q7%g8Yf(aPVlksXENvf1 zYMe3nd;(xhwy1tCQ0Qxc0`)d%i;nPw>^2FEGX-DI5ShZ()z1YBeJxm^-UhaQN#qIz z3-Ec5m@9Qx)Xzl=eJxs`-UfdD9g!atEx>m*AU}eJa^XT>3m5vPa6#q>g$p*0; z7c=y=n1Om5bR@b5k`}k5tb53AjQY9Ap|3>_)Z4(5tR)mV*t8_?jQY6{qOXMz)Z4(3 ztR)me*tDc=jrzG*qOZjg)Z4(5?Cnq}A$U8mnLvA^elCRQYas;nHt-{(4n+_G>Ht52 zx66eN11)sun-UD-euAXMeFNN2giOAES~SQ74FfG`pxy@kh+A=z6z2wR#Q`@kt{dcn zg@G0zOxiDd%g$dN#z{$5nPEeTe`9+yu`Uc-}C$ZoJ zo6H~=APlqsfqENwkvB*rD6ST`L5fqsrVY{{7bgs~IDvW_tViA;k<>V2aD${Wh7Bnf zEey11VPJ|D#I=>A#u@D^goK(gZez>op*BWNfWG7PnlVPFawWR6hC zVB<*MAPsY2!%zzwsJFrKiSB`<#dVZ*57`aUFc&)vwb+4r8+a16MDpT1*|a2YkcPP! zVyMLs)Z4(3tR)mf*tDc=kcPQfVyMLu)Z4(5=lh@lojP;Uc2 zGU`wSA)pTM^Z&N@t)Y@#S9+?}vyUW8Ed}Ft4_V|c>Ito%ks$MK~n5w zAStTvJ>6ZSdvBGis`YRZ;sO5%{$VCz7!upegdv1X2tycx6Y_#UAP_>x&p`4k5SRq! z|ByhERkhEpI_K=~TYK#z%|M1bFt+9b~j2f71d< zyyedjERc3v`+9*i2$6q8ipvtP@@bo`U3?Wq70@v;$_h=#2khibCA&ngZLaF=3ts5ttuLZ2*gMu5Hi|e z63=spKj1tEiU?E`jY7O(r11tBZ7_$X1#y)%&jE9Ao*RYu!bsx_GTM+9=!rUUmTwR` zQMYT%ua`ul5K|awOhHB)W``Dd;wax3RNw()Ae9@1D8fji2r}AW4LwmOp7O0hC+c?A zZ~`|9k%N&&4rH{!5PG6coa7sVPSjNlA&*Ol7K}7nFf!2sUJr?%d|OaGRD8FAsU8}I z*uh9+2Qu0)MZ6vor}=iFdPubkKSUvpFw!`}$ixxok}6K~?Ltc`uuFNILM&mVv4oN7 zX9!vkiMxE0P(4IVA`e!GCyX_oF#2FTK^*2Ag?K{ID779Mhj_wR;|Vg_@H|nTAg=Pw zLDUqQgXd=);t6AoC&*|6Pv8b2o?tSF*F)nFR~TztK}H)4qKQGA<yhls;iBMvg!@D%YmLd3x|M_N5J4iSj4Mj&Lg!6aG_iM#x; z!+8!A5U3s+hj_zS;|(&}U=9yCL>q9(0dq*63-N`q#uvt>GYhmH5@-4IgX*E(tD5wB zXdGe+V~r`uXv6H#dPp4Q8-wa0U<{;k;|fK1GjUxN3N4v5jv&Jg`HlWWB(Cz!K?Tvy zoC1UMftGtNyycrx@@%W{1Ylc{DEPLSb`^f8*oL?QDvng!+!N&)S78jmxFTinjWcaB z{B=HWh%}%z7TBgeUKQQ|%qkKG-z-zFCaKeDJRj}Ae|)ZB*GJP_m8vJttRU0qs7#|{Vj4XEbr8`&6%>5`<+g(IObQZ= zj!G~(#;pjJD#T$vvBlxXt0>Q?AhGDE#G+#&7Vpz%N?hg}g(|7tYdG z(J^gCe4iVH7zL`PfI*o1b%Io)qf!YOZ^#gLxkZRdpn_^=k;3>oK@!nXNra3y7{d!H zag`r$sGzDCL+C>%NFO>XeUR}6bCiNgyye@23aa9pRBZ)SC&(f?DvOZu2HTW^N?hmL zh6*axHu6vfxkRZgq{k&XrcDZE*+pCeExW)r1s?8&TMJkn$@rHGm7gVr+Br2$O z7AcIc+By;#qml?2Z!m@zRIqy_DyS;PkjJOmI1-qn?nlUYgE>k;1)E2rf@*L$U9YqA z^=y8AXe+3yeI$WxDvOZu2HTW^3igsj1(j+Wd8mS1qSRN?PZOqH3cl>Z&XQ=^1-5CA zS8%_g)Me5)%d}si6jZR$Br2$=UF=B<5{gpON#8Ql)&(!9VAn}RD2j&Z1y!}%BzVA7 zLSgGS$@whoH;D)ZG>OeqkWiF5P5Sd>+Nn^CLWII(lu}StJ5BmCRcbZq8D;A=$+HFQ zH5tkl;ex8#X%d*FG7DR;Nj_KNG@tyL=1M83s_iC$Q7Xfb@rGqfnJctSVVWzoppveU zfr6^kbkYw&(?$i)f6}DWpDUdIK=lC&s%q0oFjFeQknx6@Qo;{yQ{eCeCXxJC?K%mJ zQa34VeJ6QAC5<}$c|rx%z(@nDO(%gt>IQ|3H_Q?*sH97$ZxAY|fI*o1Ra;I1i`2ad zThmEiP)V0g-y&2{?JQClU$x&PFh(U2GTtyfyr7cyoW3!rpsEC#~9}5?H2gUC4NYWt>pJE|rK- z6b;i0s%m#iV3JBGZ2c)Yp@97<5ut!4v3UyaRFpbX`dP!&UrI3w5ek!0N z)SA*a%Ct|RWD7(pLfIl*P)TpekXb6Tknx6RN|`IfEKGBy6jasrl;EjS8HS8E7^Tb= zVi=~mQVS~SG8vewQj!$N*;c&PbCgN zU=qoH)h?C5D3w^)`c(3Q3O1@l1=Y}4P*s~$0)tdaA>$2I8mr7Jn?JQClU$s9aFh(U2whooNpptf;{`8=Ns$vX#e1i0$)Sc4L z2`2iW6jZQ5B`T;!(FIktRVA1zb!)=btWqhcV7p3GP^q?&hbqV=N}VhHi8FBtzU;!@ zm1x-owrP)7@R(+)jiqmviA^X473^e*3My(Bd(wi0qSVyVx6E{uf)fhZ)e;ekqG5VL zRc&YKKVYSPmYzwrhL)UAz=oEHP(YK&JXKp+0;5zyVQpr~&hfx*mL{W=f~wlg5*VeD z3Trz{Y7`1AE=-bclidPf(jUfxnH%rrEgKG zy`?`zCMv-TD(MI6TZ9U#oka@ctM;}8#;8|DWxQc}ctHi*TcU!hVhnkFs+}!?IVydS z@rLa%rJ#b{Em1)=jxMOGjV^(0DvOZuhN)5tD%k826;!Hi_@S!yy#&UoT*B7*k}tck z^(9(%fo>pKLSgHO$@whoh=~XVG>OeqkWiGGVfyoAA{2^Ih)|e}QVJ^RXBl9#rFNLU zQ6^HMWD7(pLfIl*P)SG2fLW!6n7&!2Up4O1ux#IQ|aLni-)O8SZV2BCrq7=*cB zwLK=VNIjJ);|&>t7gVrACMu|Q7AcIc+8h%Yqml?)drV$XNh?u*dQd@CF@`-pLHbbY zjp=7^Q*+FWHVlE?F`wSP`*$&)pJvNLZ7zZ|q0|f02XFd0<2DD6Xu>rgJo9~s4y5g3 z?@J%PX`|uSIDAA0K5=}yI$SKXliBg+y4otwS`)OQ2)+QJN!!})!e+c7`O=4B;~(%a zR4rI9r{}9%^X1v0etlKI2F=md{LOfSIq$Iv**MEL=Q9>_VB$^Np)OQuTU+oq;|+;- z#qB}AOZ+)*4}a7JXmDz@Wqvc>Fg0`y**MCdnoosH!4H1g4(%GGZEcC)j5o~7oX-nd z;;(OvS!LgwEVI+;_R+sHTj)=JXp10iYwNlgX_=EHHv|3V@VsWmyV>mao%O0!X#^%X zgVCie>YMR-bY0~$ll&^T1TE?h-#(q+oSsfr>ulj8F0I>FlmGa+#H|m@Y_*;*v&rht zba~Re@0aJ_UbLS6%dll|>l*(1q|^F1`=s4!(T}vxI?A|I`K+sa)>A&~E1wON&xXop zBjvL(cbQ^(M{!w48>hIeqqwZ2xU8eNtfRQBqqwZ2xU8eNOffH2T$U;>OEsB_%TmQ< zsp7I!aapRkELB{VDlSt@=qfJjDlY3PF6(Og6qj`smvt4FbrqL&6_<4tmnkrNipzS6 z%X*5-dWy?>8WzQ6J;h}`#brIkWj)1Z3Y5O$vcBT7zT&dJ;aamt+SzmEk zUvZhDZlJhqptx+HxNM-fY@oPoptx+H*`T;=ptx+HxJ(f_R9rSxTsBl(HdI_TR9rSx zTsBl(Hq`7>TsBl(rVJY?E*mK>8!0XuDJ~l+E*mK>8!0XuDJ~mnwks|hb2>LxTsBr* zHdb6VR$Mk#TsBr*Hdb6VR$Mk#Ts9ulx*(+_S7BwcNkbhiJ08ks+?i62E91B`rQ}q` z4Y+x{A`)K7ilwy6KkD(1`urmW;gF9S@sIdibL>JVnZ?2;=||j8!X@b_?k8cAbQJfK z@JKp}`$3gmtg?np;*KXDX79GN`fjr1e# zCt;0r6!(*GMmmc7Nf;v?#r-6Fk;M~NC2WzY60%4?;(ijUNJnu$2~nh@6!6?n!V;-K z?k6r!C?b<543U1s{UrR5j^ch2c1TBYKM6Oaqqv`h8M1idgM=3{RYD8tN8C?B3h5~B zC!vIN6!()5LOMz@nfr+oMInPsm2g4&5%-fYK{|^2Nq8U~#r-5KkdESh8jq9DCy#G8 zjW70-r>Cd+Zt@du1NhI9e9^nc#@liuKe=zZyxqR`V5^yY$o$O7Tx#1EI)Ft51u{zwfW7Xcjrs_$>va$zkSq}U!LEY z9cM=`|H_{~YTtM^xqo|JsQL6uPd;>S^Sx)V-x(X z^|3O72;EO5k2;{69~l<>;OpPbZJQmH&Oy{Cw5CKe;z;JlOirjpQR=7uSDl zy12ko`FHCRKIWr4GNGrL#TNw6C~4ekcHijOLh`+Dy*&8pVlY^}{uggNl{^9rx$$_C zyd|M}_e?suu9{bpdk#KRjo9YL%9ie|NdDDR$q%pCpq##S_TKT$@vT?im!C9uPsgLl z)%9G?MRn6j@7les!aqCZe|F3NOy!@SNnTh?m#b{^^qgkdd(*SUX*Tiybgj{PIC(Hv zQx?;|cJGJtFt%C9_!cB5;^;W4^7J--^<;LkY9{V~ zp2^DIe01}lLQ781wjVvQ|L0=9nr)9hG(WokV}59$`Q&-_-g>f<7yq^45>Fmg|MTKw zNjwkdLbryA%+`+^$%AjrU)wAO&HG-SuGXy{h!X!#M-IuB$gGa~Vefw=dFZXBu(#=2 zoc!WFHu~%>qsAn~&zT3YcB>a&X^6pLqO{))!@Yo4%9f45i(CZ2Pa*WKmhq z))FKwZ_D6%@?xWZ&sU!L`sm8Re`zF-oz9QtMB6*ljljuF2y>7K`?%k^H2Qe7Xu~J%2lM87l;<*ZKw1;An8&Ex6xmBrjCuUohejv~Fu` z7#a;M-2(ejBe`CcUOvhXXR*`DFXr!Ou{?^xxrQfk3kJp1UhMWDK17rHF;y8x;jozM zYytg2BS9Cw+l2-M-qln=_=Fera&q+R&duk?^ApK!r_=W(`Kp2_=1*W3PC9N%f4`BS zJDNn-=JE-KoVLDZhrAIJ57>Fx3Rr5iLwD>S1Mb)W@5RIeZg0cl_ZkVhTj_atqlx>s z-FnI5-wPoWEDParR0tPu>|*!!Z6RC~L^0D4cDdXd5l$C(anBYJdol5VU4H*3jRbWn zDMp0b_RjCz?ZVh!Ajhq51i}JHqPSUbjJCn?kHh(`q%8Ne4|)`g4Np!qBP$wue+&JeH z6%AxrL_AQJ#Qs4e`Mgcb#q$r4j$2=ikP#2nEfItQccLk<+^|6uSQQivI(48)vLd>8 z#ExLHQpCgqb^&1pyyvcvqsi?j#MwQ-dol5V)$9s2asT4%N`=IOb=lP#!$s-}{VRuX zO)*>)6b-uTvaK~F_T&{qZreg)FD4pr>ay?8F$XTrCRmIhXu17|NN_+@JaAWS7{a3j zoX2x77#^OWc+jrm5Hg8)FmO?M$-@3lArTG=iwEy2{(p&?u&R%*gmM`GQQ@$(c=+lL zOXm=>$xNt=W_vHX`VK6)%pj-0@7v5LJOjvu6~u*^jCiQ( zLXk#uVL6J9c!Vk_8uU;t+=(t4v2yDYB`ZZtJYW~Og!p)x_9ParX>v)x1b8nd9J2AXJe%LOmQg`w3OA@SaI!9loi zYUaXn1q^XvC?Fc-$gO{b3s-b-U-lpu2E*e4x;2k*;hhFUR*;SjXkVC!hzIHt8R5b} z3(o*@RSLtI!D6b*W;7VbnBjacbLfs&OXCLXX02*QOuO?wgx zxtd(2Gy&d=i3hA^R|uH<7iU*0BpU32ihWFcA?CR6QmEwD2S#J+}}YW-{WTy0tLN~r82XS>Zc8!?K0DFlR68DWexHVBZvdG` zc?X}KfPm{oc*lhZp~*cvO_sLvh$ce`HN5K9#J`Jaaz%;aPke#`jIROFt%I@~i;ffT zR9G^EycPfvWg?>nRF~G|M-6nK7^qRQfxM;wQDr8i23U3J=*M$qIVXttg(@i?cIsf0 z)J2r>$d%U~Abz39sR7xAg%$Fi9C+tj+_i}1Lb_)c_ww`x3suUmNY)hrofa|g? zp+TOaBEuy<#%F?5Ni|@*>}w5-5LcF0ahL*QFQ*1%myHQmF8ZfLICtB~fcMC48N`(X zvT8tgm5CupO3;#N;rb&DuU8#z;_!3ge$8GwbgM9PgaD* z)i6Naoau9pOESxtJfwiY2KJNbeo}t%hGuV|N{R9j+);*xZ=}R~ICUXHxN_glmE{aK z;>u7$Jk0#~HsQ(@CETBKavR;e1LJEzbn77D$~zU73?bd`5mzQM;-TjER0vlFYLsjs zH-QjWWz(lLnV%F-Agab;CfJnWvj7e|!w$dyL_h+imj;vsk4UL2uBo~k{^ z#o1O~Xkdc8ms10>TOdfbg^;_d6=323(hIt+5>D%JRAfQ!E!H#l!BYTUMTb z_8b=nhP=Sc6c~FsH6XieOt^9pa)~BtTv=|@Ag&yc6%W1ZDicGHl%OkdJqQMgC#eQ( zSDg^947_Ao344$$2c^}3?=}huS5_5jt79K><*>LK2B@1egIKQI!?#8hIq{Ht4l>Q& zQCwMGZ-}!ukx~P$3lYMV2X?M3Z$Cs_8A_-D)2)ewD_4~8tRb)9L0lP(uL04mgM=&Z zR9G^Eyk7`$Wg?>nRF~EWR|aa7Y#?vTL|mB(sR349I)?FFSzbMa_=PGd9(G^dizCW- zrvnvNi|@*>V$A*pvXsFOt@s(ACsO3pfb6m{;mSo-CBmt3;R@QH9`2$@H@^m{^FnM#TG@X&<_;mRXBSC-x(h$}-0 z@i6nYA%rVelyHAa#}Kry493Sp9J;E5aOIr}ONJQYi+w~!4X7@y5v~l>DA_<>&5gJ+ z6H)`Lx^#@=xw70}M@xq)DIRuyeVq_bqD+Zgyceg)sR7xAg%$FiPQ)XkZ3ELWC}Q1~JJ5V>;R5Q=7Rpo&Mi z^n*fN85&+M!pAN|2v;84xw7;YL0lP1h!2Fktq9@D6(x#4@!dW!z6M0M4ic`sQ(?&v zV|=%d$fyC;r8UBpff^+nNQWxKm6?zlV0TML7rd3EQ+yMa?Ui|Q66@7yNfbA0iIShi z65`p3P?m?T-_l)9W4#h7Cmym33p3lcmV;~~0rP`EN` zE+J2M?#cZlmcMD!f5POH7UG{wMl{s?Y?g3k^k!b9hGzpgk%h`RDkK_geh!N*9mN+H z*|@Tt#DZK|l@tv-Ka0hPCs9TtS59L=uB^z3hTPxX?St9YQ?=)~kP`JtwpHZBL)Nn` zR7HKW%XPC(gWxhN?b_F%VVWebcw$HkyiV=V9G#6x!3m~rLe zl$uw#A?C^fS<%q*n+}#BDM9CPJqQPhCn+Abt4e$CzIV>*TfZahU@{j_`sPH*CQ%e>cDkmCpezKNvW%{x_oK8nk}m@ng>0er17&W1;7Z3g7&u=y9oS2Be~|o z$O8_67`0pr!4ing_iI3K5w`hp<@j`UxL9TM)Fn(HDJ1(MQPLiS&Nd3 zhZEsUTVFPdr~%Y5(jPUFPgb=!@-tMz=DO{S-(*OM0;{d>Oh%-2U(4z4E;oX%IPY}I@^c`|9{5z?Nl z?o5{_6N&kQXOo8{lg-XIIG#v8nmqREa(Xfo>0irM$MRnXY4Y*gr}LZB)5&U`Eo6SD zC-U9N@l@W)(0nS9F@=v0rza;O^sgol7yj5hYrdV#l1B@F5<{AWe;&H`Mzo$Q{_PFX zv~}-yrpwv%=4tlr#ZmjkM*3{=ORt^J*0bs9?EA7l$^XwK$OQ4#**aTJ#UZQa3(51z zbF=fs-SuQKTV$uR^K7!7uHJ4RzLtOIJCdOeUQCkJ-Ltdlayucdmqe@M|%S+M`DD<3{ov2Xm3Xd^Bk*|5XT~*$$&{S{Tk`S3{#N;Kd)Q^Xk#y zGkm5#C@UJeOCj=pAN0eGs*ZaYpQjHCiw5sl9C`Cd%tmqtA5bPDC{A`Bprg`wpyk51 z{+4$KA5Ug7;-NZ9N?!P+mg97rKE;RLg2JM~r|v;yTPua5)!u{1JuYpE;m?pkY4PwK zlO*vBiXZCn%a^R@OXj5Oc6&%Xt8cKS-o1R8Z<(*Kw3olkm&7x8cPRDv<%?1;;(6q9 z1c_($Ep*(wm}~@vMJKqCy7jH<uWXgj2eaeq7!uc@qc->YWe4#>T<0CcNnESFSQQUncBZw) zbzaYqxCY%Xxz{hdeHP-n=eQk-Yqf70_wZ#Gxf5$#=cgA*T!T(k-0PQ}xe9TeA4wx| zt@b!kJd6v*ptN{;I&c*u;n%7{g?3^ zJpJcRitbOgZ0UD2UzS6Vo5MW^|Kq1uk|(yuwl?RY3I~z?(@(E7|ATypPXD|+mm^=< z?pV*}2$pvisqdfusSKl-+36qmWc{R(e7brZX+3{CJD;rIUt}Rzz1A<721kQ)XJ>y) z&M#KwFXk)Z9rN=Lv~Fu`7#a=C{qAsiqU=s~{PtolcSR<%LRxV@EB!$uNuBRY9Sn*j z&~}xxgWGAVg6NDg>eO*l`umNfZ3mqnP6Y>@a5HTJc_Stsu=BDNaNrte2e;Dp0Pn@b z19oS)e-GbBhMKs4aU*Rigo0&h(Or_YhH&xTCFFV?+(p~NTNFew(-3yK+!_)4dzTn8 z?x9T}*9&2n-{lTue#Dk0iu}Y|<<6DAK;j14Ma)1Fbu^o(kn#A3kinjT)2{?d?AUhkWTHNpD%=M-umNZS!t|)Y&wL zxQc*WgoB!RlExiesgQWEE;N2uE-%3~yY%=ew_sD;Eqiw?xtMmML8flm`)#?S76wUu znc`mA6^LqTpi}PFI&#o)r@_kzfc0D+M0a;xCiwvl!XJwJMK%hY;x5^$p!i62>l#T` zL>G@(+#_2N6A##}YF5B|?h3POinFT+crPX%u$o<=ChlLHU8#_Gur9k=W4K6Np?|x$ zJGKY7C@30qe!|BZ5&;e@ho?~6T>`mQ_Dx;({W<2qMNQ__40py3h>G@f>Z%Pxc$9$i zc*cFPDeQXQb`=LXfDUvSR=vv$LbeMPsk*}g8+a(3t&2NiD`KJ}yz2lHE(|r%=?;qnsO z#U+XS&?i|aikH}ITv*;4h2FBE3W^Sd{KZj(M-W{!V&%qxOgKf`*!3dY)zpLw@2+kK zxv;!r3cVQy3W)cpTmJ|buIS*t#P1si!{Y(EHIH!Nod!czxV!*=!G-gK$7G=>-ukw2 zVRt(D})9YYaPG+{gHA1cAH}6A##BU&4ip zno5*X;llE!Rm6n@qN0K4uUa*PM+rFf^#~u2ps;Isd*mt)!i9k@%c_SDNCbt&d)$@& zgWtb4MBg>^43#5s*~_y4&bNjI-Q*O zNYC%n(CiH$^C<7&(-RPIy$J8PEh9pccb(+IG`WM%Pe2JZyz17(%?mF8o3AKQ{E1Ic zfblgTx^nQ0}w?44@m%K6nB`j7(RsYd&lWN#{`Mx?v_B$~Z}DkaKC@N`8V zk^6cPp1Qpx!j*U3X~SGu&Tymi6;MLFSM%fBgezB+aDU3l?JokZ4aV1i=+;5Pm3JyE z8RD|fZ-Fc4w}8mfQS9?;!4~L$*7e|!w$nD|1I7Lo82v^>9 zybp8b-WYOaD4_;Sw!4~M?G7e|!w$nE33I7Lnk$ZoYG+167P;>vxTZ525+AgkFHs#1K6vn`cU z1Fp-q)>tkwT+EgG_&kvZxhSayZ1p@5QKtAApC?k})PU@=G2zNZRVBiyv4=jss~3<} z1G=kB3_((YuJ{|@)}z2Tl4`(q)d}IsK$&GFynM$?wFNaj^q0Tmg=X(at~|iEN`@aM zS01?iB*K-UDjwwne6NH`sS)9BFNtvF-CK0RTzP=+?Li6gFb8f;BwV?oMDZuSxd+DA zfaumi!j*R_EE(eReMAMWoWGBVEFHx+@EEzW+)qbmJXJ~Y;V@M9;)pUHxkJ1cr^u-R z*{yaY+j^=(TzQDIts;swxv?y;SOB3wZ?Lh;bN{VJwr^fT$B_Kd+4&S zH87kq#n1RWk>}^VoEnf_HYQxTsH#LbHLg6scliRc;-L>*Wnu`D5_InG0lv*gfo~+$ zfbFUi!j*wC(@Kc%^97~VfbTX6{=a{39DPV5xw3S8LMP}(A0}7M8$J=P3{~+cmwr#^ zS3xQzK17BtLjy_qk}m8EwGI)ecv#KX+nh7hh?QNsNx9YfF=3@|<(;?S*wge&h< zSTe-rI}Qt6Ie*6?SvrdEI5cu)xxbG1g(@jN9P;byBy|yGO620bI7Lnk$ZoZ>LJo0d zx%rKP36RgOx_TRW^bU1N4fNaLR@)_$bG#CAG^II z!j*S#`weqt=`Dh|GL%rmt8PssT)Cn|@h8692gcWc=+;5Pm3JyE8RGK2$px;Qcc+r2 zqYI9jm)@I<$gAY2)~ z2R!)hI~d=W@3)VAzdg_1fO%@sK4uRol@lE@z1?jSlD&Z{9_1sJy@`}~xNa{=xbkOl z9wywp^?Kss z4cHx&f(0#n^PIxxp@evtZcSudxuS&obA-7v7#|POt%Hm!?^GBv#9)BAGLaDv)ulDYm4OF<7h#(pSB_6thl^!)GCN++=I5)!^>TW?x;0;(h5eGY<EzArf6_=^*vMUF-y@y0+Jk`(^i^7egLzTx;BTLHYLU^ZDZC`T5CFJ3Uao_N{DrJ3DG0X`)_}Wnn#A zzkk$z={ZHzEAP$LN9}L^aVx>!x)3vw_-0>OXUpk&zWk1)lthF5iui3w#W;UCJc-S+ zBvmYXEr<1Lw*F<{69eXL=Op&-5tHpXpJYKhvW)f2K!q{!EYJ{3$Fg%Aegx ze|97N*^TsPH`1ToNPl)C{VA(xls|is{_I8ivlr>lUZg*Jk^byO`cueKlt25C{_IEk zvmfctexyJ9k^byQ`cq2iD1Qzj{W*yA=OEIbgGhf4BK`+M`i50Ik}VrA z0_}O_yKlZbTOZ%CeW{}&o1UB;N>P(sSG37@q^NzJ)hl^rwBe>r>)bTI%cK>mpJt7E zk0&2V9=W}o-(5^jvc+nW9(*_X*fN`*yg!j$ugU5BczQay z`0^K%qhz$q7W3tLa`$}mSNE>AlgT^zo6T2~&f(3?KH~Izawc`_;j7d2oorjn9Xyvj zI$1C7PGn1Mc8+v+4*u2y$s4jteP(*&yxgF-1qo@aCAGA literal 0 HcmV?d00001 diff --git a/chinese_wwm_ext_L-12_H-768_A-12/vocab.txt b/chinese_wwm_ext_L-12_H-768_A-12/vocab.txt new file mode 100644 index 0000000..ca4f978 --- /dev/null +++ b/chinese_wwm_ext_L-12_H-768_A-12/vocab.txt @@ -0,0 +1,21128 @@ +[PAD] +[unused1] +[unused2] +[unused3] +[unused4] +[unused5] +[unused6] +[unused7] +[unused8] +[unused9] +[unused10] +[unused11] +[unused12] +[unused13] +[unused14] +[unused15] +[unused16] +[unused17] +[unused18] +[unused19] +[unused20] +[unused21] +[unused22] +[unused23] +[unused24] +[unused25] +[unused26] +[unused27] +[unused28] +[unused29] +[unused30] +[unused31] +[unused32] +[unused33] +[unused34] +[unused35] +[unused36] +[unused37] +[unused38] +[unused39] +[unused40] +[unused41] +[unused42] +[unused43] +[unused44] +[unused45] +[unused46] +[unused47] +[unused48] +[unused49] +[unused50] +[unused51] +[unused52] +[unused53] +[unused54] +[unused55] +[unused56] +[unused57] +[unused58] +[unused59] +[unused60] +[unused61] +[unused62] +[unused63] +[unused64] +[unused65] +[unused66] +[unused67] +[unused68] +[unused69] +[unused70] +[unused71] +[unused72] +[unused73] +[unused74] +[unused75] +[unused76] +[unused77] +[unused78] +[unused79] +[unused80] +[unused81] +[unused82] +[unused83] +[unused84] +[unused85] +[unused86] +[unused87] +[unused88] +[unused89] +[unused90] +[unused91] +[unused92] +[unused93] +[unused94] +[unused95] +[unused96] +[unused97] +[unused98] +[unused99] +[UNK] +[CLS] +[SEP] +[MASK] + + +! +" +# +$ +% +& +' +( +) +* ++ +, +- +. +/ +0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +: +; +< += +> +? +@ +[ +\ +] +^ +_ +a +b +c +d +e +f +g +h +i +j +k +l +m +n +o +p +q +r +s +t +u +v +w +x +y +z +{ +| +} +~ +£ +¤ +¥ +§ +© +« +® +° +± +² +³ +µ +· +¹ +º +» +¼ +× +ß +æ +÷ +ø +đ +ŋ +ɔ +ə +ɡ +ʰ +ˇ +ˈ +ˊ +ˋ +ˍ +ː +˙ +˚ +ˢ +α +β +γ +δ +ε +η +θ +ι +κ +λ +μ +ν +ο +π +ρ +ς +σ +τ +υ +φ +χ +ψ +ω +а +б +в +г +д +е +ж +з +и +к +л +м +н +о +п +р +с +т +у +ф +х +ц +ч +ш +ы +ь +я +і +ا +ب +ة +ت +د +ر +س +ع +ل +م +ن +ه +و +ي +۩ +ก +ง +น +ม +ย +ร +อ +า +เ +๑ +་ +ღ +ᄀ +ᄁ +ᄂ +ᄃ +ᄅ +ᄆ +ᄇ +ᄈ +ᄉ +ᄋ +ᄌ +ᄎ +ᄏ +ᄐ +ᄑ +ᄒ +ᅡ +ᅢ +ᅣ +ᅥ +ᅦ +ᅧ +ᅨ +ᅩ +ᅪ +ᅬ +ᅭ +ᅮ +ᅯ +ᅲ +ᅳ +ᅴ +ᅵ +ᆨ +ᆫ +ᆯ +ᆷ +ᆸ +ᆺ +ᆻ +ᆼ +ᗜ +ᵃ +ᵉ +ᵍ +ᵏ +ᵐ +ᵒ +ᵘ +‖ +„ +† +• +‥ +‧ +
 +‰ +′ +″ +‹ +› +※ +‿ +⁄ +ⁱ +⁺ +ⁿ +₁ +₂ +₃ +₄ +€ +℃ +№ +™ +ⅰ +ⅱ +ⅲ +ⅳ +ⅴ +← +↑ +→ +↓ +↔ +↗ +↘ +⇒ +∀ +− +∕ +∙ +√ +∞ +∟ +∠ +∣ +∥ +∩ +∮ +∶ +∼ +∽ +≈ +≒ +≡ +≤ +≥ +≦ +≧ +≪ +≫ +⊙ +⋅ +⋈ +⋯ +⌒ +① +② +③ +④ +⑤ +⑥ +⑦ +⑧ +⑨ +⑩ +⑴ +⑵ +⑶ +⑷ +⑸ +⒈ +⒉ +⒊ +⒋ +ⓒ +ⓔ +ⓘ +─ +━ +│ +┃ +┅ +┆ +┊ +┌ +└ +├ +┣ +═ +║ +╚ +╞ +╠ +╭ +╮ +╯ +╰ +╱ +╳ +▂ +▃ +▅ +▇ +█ +▉ +▋ +▌ +▍ +▎ +■ +□ +▪ +▫ +▬ +▲ +△ +▶ +► +▼ +▽ +◆ +◇ +○ +◎ +● +◕ +◠ +◢ +◤ +☀ +★ +☆ +☕ +☞ +☺ +☼ +♀ +♂ +♠ +♡ +♣ +♥ +♦ +♪ +♫ +♬ +✈ +✔ +✕ +✖ +✦ +✨ +✪ +✰ +✿ +❀ +❤ +➜ +➤ +⦿ +、 +。 +〃 +々 +〇 +〈 +〉 +《 +》 +「 +」 +『 +』 +【 +】 +〓 +〔 +〕 +〖 +〗 +〜 +〝 +〞 +ぁ +あ +ぃ +い +う +ぇ +え +お +か +き +く +け +こ +さ +し +す +せ +そ +た +ち +っ +つ +て +と +な +に +ぬ +ね +の +は +ひ +ふ +へ +ほ +ま +み +む +め +も +ゃ +や +ゅ +ゆ +ょ +よ +ら +り +る +れ +ろ +わ +を +ん +゜ +ゝ +ァ +ア +ィ +イ +ゥ +ウ +ェ +エ +ォ +オ +カ +キ +ク +ケ +コ +サ +シ +ス +セ +ソ +タ +チ +ッ +ツ +テ +ト +ナ +ニ +ヌ +ネ +ノ +ハ +ヒ +フ +ヘ +ホ +マ +ミ +ム +メ +モ +ャ +ヤ +ュ +ユ +ョ +ヨ +ラ +リ +ル +レ +ロ +ワ +ヲ +ン +ヶ +・ +ー +ヽ +ㄅ +ㄆ +ㄇ +ㄉ +ㄋ +ㄌ +ㄍ +ㄎ +ㄏ +ㄒ +ㄚ +ㄛ +ㄞ +ㄟ +ㄢ +ㄤ +ㄥ +ㄧ +ㄨ +ㆍ +㈦ +㊣ +㎡ +㗎 +一 +丁 +七 +万 +丈 +三 +上 +下 +不 +与 +丐 +丑 +专 +且 +丕 +世 +丘 +丙 +业 +丛 +东 +丝 +丞 +丟 +両 +丢 +两 +严 +並 +丧 +丨 +个 +丫 +中 +丰 +串 +临 +丶 +丸 +丹 +为 +主 +丼 +丽 +举 +丿 +乂 +乃 +久 +么 +义 +之 +乌 +乍 +乎 +乏 +乐 +乒 +乓 +乔 +乖 +乗 +乘 +乙 +乜 +九 +乞 +也 +习 +乡 +书 +乩 +买 +乱 +乳 +乾 +亀 +亂 +了 +予 +争 +事 +二 +于 +亏 +云 +互 +五 +井 +亘 +亙 +亚 +些 +亜 +亞 +亟 +亡 +亢 +交 +亥 +亦 +产 +亨 +亩 +享 +京 +亭 +亮 +亲 +亳 +亵 +人 +亿 +什 +仁 +仃 +仄 +仅 +仆 +仇 +今 +介 +仍 +从 +仏 +仑 +仓 +仔 +仕 +他 +仗 +付 +仙 +仝 +仞 +仟 +代 +令 +以 +仨 +仪 +们 +仮 +仰 +仲 +件 +价 +任 +份 +仿 +企 +伉 +伊 +伍 +伎 +伏 +伐 +休 +伕 +众 +优 +伙 +会 +伝 +伞 +伟 +传 +伢 +伤 +伦 +伪 +伫 +伯 +估 +伴 +伶 +伸 +伺 +似 +伽 +佃 +但 +佇 +佈 +位 +低 +住 +佐 +佑 +体 +佔 +何 +佗 +佘 +余 +佚 +佛 +作 +佝 +佞 +佟 +你 +佢 +佣 +佤 +佥 +佩 +佬 +佯 +佰 +佳 +併 +佶 +佻 +佼 +使 +侃 +侄 +來 +侈 +例 +侍 +侏 +侑 +侖 +侗 +供 +依 +侠 +価 +侣 +侥 +侦 +侧 +侨 +侬 +侮 +侯 +侵 +侶 +侷 +便 +係 +促 +俄 +俊 +俎 +俏 +俐 +俑 +俗 +俘 +俚 +保 +俞 +俟 +俠 +信 +俨 +俩 +俪 +俬 +俭 +修 +俯 +俱 +俳 +俸 +俺 +俾 +倆 +倉 +個 +倌 +倍 +倏 +們 +倒 +倔 +倖 +倘 +候 +倚 +倜 +借 +倡 +値 +倦 +倩 +倪 +倫 +倬 +倭 +倶 +债 +值 +倾 +偃 +假 +偈 +偉 +偌 +偎 +偏 +偕 +做 +停 +健 +側 +偵 +偶 +偷 +偻 +偽 +偿 +傀 +傅 +傍 +傑 +傘 +備 +傚 +傢 +傣 +傥 +储 +傩 +催 +傭 +傲 +傳 +債 +傷 +傻 +傾 +僅 +働 +像 +僑 +僕 +僖 +僚 +僥 +僧 +僭 +僮 +僱 +僵 +價 +僻 +儀 +儂 +億 +儆 +儉 +儋 +儒 +儕 +儘 +償 +儡 +優 +儲 +儷 +儼 +儿 +兀 +允 +元 +兄 +充 +兆 +兇 +先 +光 +克 +兌 +免 +児 +兑 +兒 +兔 +兖 +党 +兜 +兢 +入 +內 +全 +兩 +八 +公 +六 +兮 +兰 +共 +兲 +关 +兴 +兵 +其 +具 +典 +兹 +养 +兼 +兽 +冀 +内 +円 +冇 +冈 +冉 +冊 +册 +再 +冏 +冒 +冕 +冗 +写 +军 +农 +冠 +冢 +冤 +冥 +冨 +冪 +冬 +冯 +冰 +冲 +决 +况 +冶 +冷 +冻 +冼 +冽 +冾 +净 +凄 +准 +凇 +凈 +凉 +凋 +凌 +凍 +减 +凑 +凛 +凜 +凝 +几 +凡 +凤 +処 +凪 +凭 +凯 +凰 +凱 +凳 +凶 +凸 +凹 +出 +击 +函 +凿 +刀 +刁 +刃 +分 +切 +刈 +刊 +刍 +刎 +刑 +划 +列 +刘 +则 +刚 +创 +初 +删 +判 +別 +刨 +利 +刪 +别 +刮 +到 +制 +刷 +券 +刹 +刺 +刻 +刽 +剁 +剂 +剃 +則 +剉 +削 +剋 +剌 +前 +剎 +剐 +剑 +剔 +剖 +剛 +剜 +剝 +剣 +剤 +剥 +剧 +剩 +剪 +副 +割 +創 +剷 +剽 +剿 +劃 +劇 +劈 +劉 +劊 +劍 +劏 +劑 +力 +劝 +办 +功 +加 +务 +劣 +动 +助 +努 +劫 +劭 +励 +劲 +劳 +労 +劵 +効 +劾 +势 +勁 +勃 +勇 +勉 +勋 +勐 +勒 +動 +勖 +勘 +務 +勛 +勝 +勞 +募 +勢 +勤 +勧 +勳 +勵 +勸 +勺 +勻 +勾 +勿 +匀 +包 +匆 +匈 +匍 +匐 +匕 +化 +北 +匙 +匝 +匠 +匡 +匣 +匪 +匮 +匯 +匱 +匹 +区 +医 +匾 +匿 +區 +十 +千 +卅 +升 +午 +卉 +半 +卍 +华 +协 +卑 +卒 +卓 +協 +单 +卖 +南 +単 +博 +卜 +卞 +卟 +占 +卡 +卢 +卤 +卦 +卧 +卫 +卮 +卯 +印 +危 +即 +却 +卵 +卷 +卸 +卻 +卿 +厂 +厄 +厅 +历 +厉 +压 +厌 +厕 +厘 +厚 +厝 +原 +厢 +厥 +厦 +厨 +厩 +厭 +厮 +厲 +厳 +去 +县 +叁 +参 +參 +又 +叉 +及 +友 +双 +反 +収 +发 +叔 +取 +受 +变 +叙 +叛 +叟 +叠 +叡 +叢 +口 +古 +句 +另 +叨 +叩 +只 +叫 +召 +叭 +叮 +可 +台 +叱 +史 +右 +叵 +叶 +号 +司 +叹 +叻 +叼 +叽 +吁 +吃 +各 +吆 +合 +吉 +吊 +吋 +同 +名 +后 +吏 +吐 +向 +吒 +吓 +吕 +吖 +吗 +君 +吝 +吞 +吟 +吠 +吡 +否 +吧 +吨 +吩 +含 +听 +吭 +吮 +启 +吱 +吳 +吴 +吵 +吶 +吸 +吹 +吻 +吼 +吽 +吾 +呀 +呂 +呃 +呆 +呈 +告 +呋 +呎 +呐 +呓 +呕 +呗 +员 +呛 +呜 +呢 +呤 +呦 +周 +呱 +呲 +味 +呵 +呷 +呸 +呻 +呼 +命 +咀 +咁 +咂 +咄 +咆 +咋 +和 +咎 +咏 +咐 +咒 +咔 +咕 +咖 +咗 +咘 +咙 +咚 +咛 +咣 +咤 +咦 +咧 +咨 +咩 +咪 +咫 +咬 +咭 +咯 +咱 +咲 +咳 +咸 +咻 +咽 +咿 +哀 +品 +哂 +哄 +哆 +哇 +哈 +哉 +哋 +哌 +响 +哎 +哏 +哐 +哑 +哒 +哔 +哗 +哟 +員 +哥 +哦 +哧 +哨 +哩 +哪 +哭 +哮 +哲 +哺 +哼 +哽 +唁 +唄 +唆 +唇 +唉 +唏 +唐 +唑 +唔 +唠 +唤 +唧 +唬 +售 +唯 +唰 +唱 +唳 +唷 +唸 +唾 +啃 +啄 +商 +啉 +啊 +問 +啓 +啕 +啖 +啜 +啞 +啟 +啡 +啤 +啥 +啦 +啧 +啪 +啫 +啬 +啮 +啰 +啱 +啲 +啵 +啶 +啷 +啸 +啻 +啼 +啾 +喀 +喂 +喃 +善 +喆 +喇 +喉 +喊 +喋 +喎 +喏 +喔 +喘 +喙 +喚 +喜 +喝 +喟 +喧 +喪 +喫 +喬 +單 +喰 +喱 +喲 +喳 +喵 +営 +喷 +喹 +喺 +喻 +喽 +嗅 +嗆 +嗇 +嗎 +嗑 +嗒 +嗓 +嗔 +嗖 +嗚 +嗜 +嗝 +嗟 +嗡 +嗣 +嗤 +嗦 +嗨 +嗪 +嗬 +嗯 +嗰 +嗲 +嗳 +嗶 +嗷 +嗽 +嘀 +嘅 +嘆 +嘈 +嘉 +嘌 +嘍 +嘎 +嘔 +嘖 +嘗 +嘘 +嘚 +嘛 +嘜 +嘞 +嘟 +嘢 +嘣 +嘤 +嘧 +嘩 +嘭 +嘮 +嘯 +嘰 +嘱 +嘲 +嘴 +嘶 +嘸 +嘹 +嘻 +嘿 +噁 +噌 +噎 +噓 +噔 +噗 +噙 +噜 +噠 +噢 +噤 +器 +噩 +噪 +噬 +噱 +噴 +噶 +噸 +噹 +噻 +噼 +嚀 +嚇 +嚎 +嚏 +嚐 +嚓 +嚕 +嚟 +嚣 +嚥 +嚨 +嚮 +嚴 +嚷 +嚼 +囂 +囉 +囊 +囍 +囑 +囔 +囗 +囚 +四 +囝 +回 +囟 +因 +囡 +团 +団 +囤 +囧 +囪 +囫 +园 +困 +囱 +囲 +図 +围 +囹 +固 +国 +图 +囿 +圃 +圄 +圆 +圈 +國 +圍 +圏 +園 +圓 +圖 +團 +圜 +土 +圣 +圧 +在 +圩 +圭 +地 +圳 +场 +圻 +圾 +址 +坂 +均 +坊 +坍 +坎 +坏 +坐 +坑 +块 +坚 +坛 +坝 +坞 +坟 +坠 +坡 +坤 +坦 +坨 +坪 +坯 +坳 +坵 +坷 +垂 +垃 +垄 +型 +垒 +垚 +垛 +垠 +垢 +垣 +垦 +垩 +垫 +垭 +垮 +垵 +埂 +埃 +埋 +城 +埔 +埕 +埗 +域 +埠 +埤 +埵 +執 +埸 +培 +基 +埼 +堀 +堂 +堃 +堅 +堆 +堇 +堑 +堕 +堙 +堡 +堤 +堪 +堯 +堰 +報 +場 +堵 +堺 +堿 +塊 +塌 +塑 +塔 +塗 +塘 +塚 +塞 +塢 +塩 +填 +塬 +塭 +塵 +塾 +墀 +境 +墅 +墉 +墊 +墒 +墓 +増 +墘 +墙 +墜 +增 +墟 +墨 +墩 +墮 +墳 +墻 +墾 +壁 +壅 +壆 +壇 +壊 +壑 +壓 +壕 +壘 +壞 +壟 +壢 +壤 +壩 +士 +壬 +壮 +壯 +声 +売 +壳 +壶 +壹 +壺 +壽 +处 +备 +変 +复 +夏 +夔 +夕 +外 +夙 +多 +夜 +够 +夠 +夢 +夥 +大 +天 +太 +夫 +夭 +央 +夯 +失 +头 +夷 +夸 +夹 +夺 +夾 +奂 +奄 +奇 +奈 +奉 +奋 +奎 +奏 +奐 +契 +奔 +奕 +奖 +套 +奘 +奚 +奠 +奢 +奥 +奧 +奪 +奬 +奮 +女 +奴 +奶 +奸 +她 +好 +如 +妃 +妄 +妆 +妇 +妈 +妊 +妍 +妒 +妓 +妖 +妘 +妙 +妝 +妞 +妣 +妤 +妥 +妨 +妩 +妪 +妮 +妲 +妳 +妹 +妻 +妾 +姆 +姉 +姊 +始 +姍 +姐 +姑 +姒 +姓 +委 +姗 +姚 +姜 +姝 +姣 +姥 +姦 +姨 +姪 +姫 +姬 +姹 +姻 +姿 +威 +娃 +娄 +娅 +娆 +娇 +娉 +娑 +娓 +娘 +娛 +娜 +娟 +娠 +娣 +娥 +娩 +娱 +娲 +娴 +娶 +娼 +婀 +婁 +婆 +婉 +婊 +婕 +婚 +婢 +婦 +婧 +婪 +婭 +婴 +婵 +婶 +婷 +婺 +婿 +媒 +媚 +媛 +媞 +媧 +媲 +媳 +媽 +媾 +嫁 +嫂 +嫉 +嫌 +嫑 +嫔 +嫖 +嫘 +嫚 +嫡 +嫣 +嫦 +嫩 +嫲 +嫵 +嫻 +嬅 +嬉 +嬌 +嬗 +嬛 +嬢 +嬤 +嬪 +嬰 +嬴 +嬷 +嬸 +嬿 +孀 +孃 +子 +孑 +孔 +孕 +孖 +字 +存 +孙 +孚 +孛 +孜 +孝 +孟 +孢 +季 +孤 +学 +孩 +孪 +孫 +孬 +孰 +孱 +孳 +孵 +學 +孺 +孽 +孿 +宁 +它 +宅 +宇 +守 +安 +宋 +完 +宏 +宓 +宕 +宗 +官 +宙 +定 +宛 +宜 +宝 +实 +実 +宠 +审 +客 +宣 +室 +宥 +宦 +宪 +宫 +宮 +宰 +害 +宴 +宵 +家 +宸 +容 +宽 +宾 +宿 +寂 +寄 +寅 +密 +寇 +富 +寐 +寒 +寓 +寛 +寝 +寞 +察 +寡 +寢 +寥 +實 +寧 +寨 +審 +寫 +寬 +寮 +寰 +寵 +寶 +寸 +对 +寺 +寻 +导 +対 +寿 +封 +専 +射 +将 +將 +專 +尉 +尊 +尋 +對 +導 +小 +少 +尔 +尕 +尖 +尘 +尚 +尝 +尤 +尧 +尬 +就 +尴 +尷 +尸 +尹 +尺 +尻 +尼 +尽 +尾 +尿 +局 +屁 +层 +屄 +居 +屆 +屈 +屉 +届 +屋 +屌 +屍 +屎 +屏 +屐 +屑 +展 +屜 +属 +屠 +屡 +屢 +層 +履 +屬 +屯 +山 +屹 +屿 +岀 +岁 +岂 +岌 +岐 +岑 +岔 +岖 +岗 +岘 +岙 +岚 +岛 +岡 +岩 +岫 +岬 +岭 +岱 +岳 +岷 +岸 +峇 +峋 +峒 +峙 +峡 +峤 +峥 +峦 +峨 +峪 +峭 +峯 +峰 +峴 +島 +峻 +峽 +崁 +崂 +崆 +崇 +崎 +崑 +崔 +崖 +崗 +崙 +崛 +崧 +崩 +崭 +崴 +崽 +嵇 +嵊 +嵋 +嵌 +嵐 +嵘 +嵩 +嵬 +嵯 +嶂 +嶄 +嶇 +嶋 +嶙 +嶺 +嶼 +嶽 +巅 +巍 +巒 +巔 +巖 +川 +州 +巡 +巢 +工 +左 +巧 +巨 +巩 +巫 +差 +己 +已 +巳 +巴 +巷 +巻 +巽 +巾 +巿 +币 +市 +布 +帅 +帆 +师 +希 +帐 +帑 +帕 +帖 +帘 +帚 +帛 +帜 +帝 +帥 +带 +帧 +師 +席 +帮 +帯 +帰 +帳 +帶 +帷 +常 +帼 +帽 +幀 +幂 +幄 +幅 +幌 +幔 +幕 +幟 +幡 +幢 +幣 +幫 +干 +平 +年 +并 +幸 +幹 +幺 +幻 +幼 +幽 +幾 +广 +庁 +広 +庄 +庆 +庇 +床 +序 +庐 +库 +应 +底 +庖 +店 +庙 +庚 +府 +庞 +废 +庠 +度 +座 +庫 +庭 +庵 +庶 +康 +庸 +庹 +庾 +廁 +廂 +廃 +廈 +廉 +廊 +廓 +廖 +廚 +廝 +廟 +廠 +廢 +廣 +廬 +廳 +延 +廷 +建 +廿 +开 +弁 +异 +弃 +弄 +弈 +弊 +弋 +式 +弑 +弒 +弓 +弔 +引 +弗 +弘 +弛 +弟 +张 +弥 +弦 +弧 +弩 +弭 +弯 +弱 +張 +強 +弹 +强 +弼 +弾 +彅 +彆 +彈 +彌 +彎 +归 +当 +录 +彗 +彙 +彝 +形 +彤 +彥 +彦 +彧 +彩 +彪 +彫 +彬 +彭 +彰 +影 +彷 +役 +彻 +彼 +彿 +往 +征 +径 +待 +徇 +很 +徉 +徊 +律 +後 +徐 +徑 +徒 +従 +徕 +得 +徘 +徙 +徜 +從 +徠 +御 +徨 +復 +循 +徬 +微 +徳 +徴 +徵 +德 +徹 +徼 +徽 +心 +必 +忆 +忌 +忍 +忏 +忐 +忑 +忒 +忖 +志 +忘 +忙 +応 +忠 +忡 +忤 +忧 +忪 +快 +忱 +念 +忻 +忽 +忿 +怀 +态 +怂 +怅 +怆 +怎 +怏 +怒 +怔 +怕 +怖 +怙 +怜 +思 +怠 +怡 +急 +怦 +性 +怨 +怪 +怯 +怵 +总 +怼 +恁 +恃 +恆 +恋 +恍 +恐 +恒 +恕 +恙 +恚 +恢 +恣 +恤 +恥 +恨 +恩 +恪 +恫 +恬 +恭 +息 +恰 +恳 +恵 +恶 +恸 +恺 +恻 +恼 +恿 +悄 +悅 +悉 +悌 +悍 +悔 +悖 +悚 +悟 +悠 +患 +悦 +您 +悩 +悪 +悬 +悯 +悱 +悲 +悴 +悵 +悶 +悸 +悻 +悼 +悽 +情 +惆 +惇 +惊 +惋 +惑 +惕 +惘 +惚 +惜 +惟 +惠 +惡 +惦 +惧 +惨 +惩 +惫 +惬 +惭 +惮 +惯 +惰 +惱 +想 +惴 +惶 +惹 +惺 +愁 +愆 +愈 +愉 +愍 +意 +愕 +愚 +愛 +愜 +感 +愣 +愤 +愧 +愫 +愷 +愿 +慄 +慈 +態 +慌 +慎 +慑 +慕 +慘 +慚 +慟 +慢 +慣 +慧 +慨 +慫 +慮 +慰 +慳 +慵 +慶 +慷 +慾 +憂 +憊 +憋 +憎 +憐 +憑 +憔 +憚 +憤 +憧 +憨 +憩 +憫 +憬 +憲 +憶 +憾 +懂 +懇 +懈 +應 +懊 +懋 +懑 +懒 +懦 +懲 +懵 +懶 +懷 +懸 +懺 +懼 +懾 +懿 +戀 +戈 +戊 +戌 +戍 +戎 +戏 +成 +我 +戒 +戕 +或 +战 +戚 +戛 +戟 +戡 +戦 +截 +戬 +戮 +戰 +戲 +戳 +戴 +戶 +户 +戸 +戻 +戾 +房 +所 +扁 +扇 +扈 +扉 +手 +才 +扎 +扑 +扒 +打 +扔 +払 +托 +扛 +扣 +扦 +执 +扩 +扪 +扫 +扬 +扭 +扮 +扯 +扰 +扱 +扳 +扶 +批 +扼 +找 +承 +技 +抄 +抉 +把 +抑 +抒 +抓 +投 +抖 +抗 +折 +抚 +抛 +抜 +択 +抟 +抠 +抡 +抢 +护 +报 +抨 +披 +抬 +抱 +抵 +抹 +押 +抽 +抿 +拂 +拄 +担 +拆 +拇 +拈 +拉 +拋 +拌 +拍 +拎 +拐 +拒 +拓 +拔 +拖 +拗 +拘 +拙 +拚 +招 +拜 +拟 +拡 +拢 +拣 +拥 +拦 +拧 +拨 +择 +括 +拭 +拮 +拯 +拱 +拳 +拴 +拷 +拼 +拽 +拾 +拿 +持 +挂 +指 +挈 +按 +挎 +挑 +挖 +挙 +挚 +挛 +挝 +挞 +挟 +挠 +挡 +挣 +挤 +挥 +挨 +挪 +挫 +振 +挲 +挹 +挺 +挽 +挾 +捂 +捅 +捆 +捉 +捋 +捌 +捍 +捎 +捏 +捐 +捕 +捞 +损 +捡 +换 +捣 +捧 +捨 +捩 +据 +捱 +捲 +捶 +捷 +捺 +捻 +掀 +掂 +掃 +掇 +授 +掉 +掌 +掏 +掐 +排 +掖 +掘 +掙 +掛 +掠 +採 +探 +掣 +接 +控 +推 +掩 +措 +掬 +掰 +掲 +掳 +掴 +掷 +掸 +掺 +揀 +揃 +揄 +揆 +揉 +揍 +描 +提 +插 +揖 +揚 +換 +握 +揣 +揩 +揪 +揭 +揮 +援 +揶 +揸 +揹 +揽 +搀 +搁 +搂 +搅 +損 +搏 +搐 +搓 +搔 +搖 +搗 +搜 +搞 +搡 +搪 +搬 +搭 +搵 +搶 +携 +搽 +摀 +摁 +摄 +摆 +摇 +摈 +摊 +摒 +摔 +摘 +摞 +摟 +摧 +摩 +摯 +摳 +摸 +摹 +摺 +摻 +撂 +撃 +撅 +撇 +撈 +撐 +撑 +撒 +撓 +撕 +撚 +撞 +撤 +撥 +撩 +撫 +撬 +播 +撮 +撰 +撲 +撵 +撷 +撸 +撻 +撼 +撿 +擀 +擁 +擂 +擄 +擅 +擇 +擊 +擋 +操 +擎 +擒 +擔 +擘 +據 +擞 +擠 +擡 +擢 +擦 +擬 +擰 +擱 +擲 +擴 +擷 +擺 +擼 +擾 +攀 +攏 +攒 +攔 +攘 +攙 +攜 +攝 +攞 +攢 +攣 +攤 +攥 +攪 +攫 +攬 +支 +收 +攸 +改 +攻 +放 +政 +故 +效 +敌 +敍 +敎 +敏 +救 +敕 +敖 +敗 +敘 +教 +敛 +敝 +敞 +敢 +散 +敦 +敬 +数 +敲 +整 +敵 +敷 +數 +斂 +斃 +文 +斋 +斌 +斎 +斐 +斑 +斓 +斗 +料 +斛 +斜 +斟 +斡 +斤 +斥 +斧 +斩 +斫 +斬 +断 +斯 +新 +斷 +方 +於 +施 +旁 +旃 +旅 +旋 +旌 +旎 +族 +旖 +旗 +无 +既 +日 +旦 +旧 +旨 +早 +旬 +旭 +旮 +旱 +时 +旷 +旺 +旻 +昀 +昂 +昆 +昇 +昉 +昊 +昌 +明 +昏 +易 +昔 +昕 +昙 +星 +映 +春 +昧 +昨 +昭 +是 +昱 +昴 +昵 +昶 +昼 +显 +晁 +時 +晃 +晉 +晋 +晌 +晏 +晒 +晓 +晔 +晕 +晖 +晗 +晚 +晝 +晞 +晟 +晤 +晦 +晨 +晩 +普 +景 +晰 +晴 +晶 +晷 +智 +晾 +暂 +暄 +暇 +暈 +暉 +暌 +暐 +暑 +暖 +暗 +暝 +暢 +暧 +暨 +暫 +暮 +暱 +暴 +暸 +暹 +曄 +曆 +曇 +曉 +曖 +曙 +曜 +曝 +曠 +曦 +曬 +曰 +曲 +曳 +更 +書 +曹 +曼 +曾 +替 +最 +會 +月 +有 +朋 +服 +朐 +朔 +朕 +朗 +望 +朝 +期 +朦 +朧 +木 +未 +末 +本 +札 +朮 +术 +朱 +朴 +朵 +机 +朽 +杀 +杂 +权 +杆 +杈 +杉 +李 +杏 +材 +村 +杓 +杖 +杜 +杞 +束 +杠 +条 +来 +杨 +杭 +杯 +杰 +東 +杳 +杵 +杷 +杼 +松 +板 +极 +构 +枇 +枉 +枋 +析 +枕 +林 +枚 +果 +枝 +枢 +枣 +枪 +枫 +枭 +枯 +枰 +枱 +枳 +架 +枷 +枸 +柄 +柏 +某 +柑 +柒 +染 +柔 +柘 +柚 +柜 +柞 +柠 +柢 +查 +柩 +柬 +柯 +柱 +柳 +柴 +柵 +査 +柿 +栀 +栃 +栄 +栅 +标 +栈 +栉 +栋 +栎 +栏 +树 +栓 +栖 +栗 +校 +栩 +株 +样 +核 +根 +格 +栽 +栾 +桀 +桁 +桂 +桃 +桅 +框 +案 +桉 +桌 +桎 +桐 +桑 +桓 +桔 +桜 +桠 +桡 +桢 +档 +桥 +桦 +桧 +桨 +桩 +桶 +桿 +梁 +梅 +梆 +梏 +梓 +梗 +條 +梟 +梢 +梦 +梧 +梨 +梭 +梯 +械 +梳 +梵 +梶 +检 +棂 +棄 +棉 +棋 +棍 +棒 +棕 +棗 +棘 +棚 +棟 +棠 +棣 +棧 +森 +棱 +棲 +棵 +棹 +棺 +椁 +椅 +椋 +植 +椎 +椒 +検 +椪 +椭 +椰 +椹 +椽 +椿 +楂 +楊 +楓 +楔 +楚 +楝 +楞 +楠 +楣 +楨 +楫 +業 +楮 +極 +楷 +楸 +楹 +楼 +楽 +概 +榄 +榆 +榈 +榉 +榔 +榕 +榖 +榛 +榜 +榨 +榫 +榭 +榮 +榱 +榴 +榷 +榻 +槁 +槃 +構 +槌 +槍 +槎 +槐 +槓 +様 +槛 +槟 +槤 +槭 +槲 +槳 +槻 +槽 +槿 +樁 +樂 +樊 +樑 +樓 +標 +樞 +樟 +模 +樣 +権 +横 +樫 +樯 +樱 +樵 +樸 +樹 +樺 +樽 +樾 +橄 +橇 +橋 +橐 +橘 +橙 +機 +橡 +橢 +橫 +橱 +橹 +橼 +檀 +檄 +檎 +檐 +檔 +檗 +檜 +檢 +檬 +檯 +檳 +檸 +檻 +櫃 +櫚 +櫛 +櫥 +櫸 +櫻 +欄 +權 +欒 +欖 +欠 +次 +欢 +欣 +欧 +欲 +欸 +欺 +欽 +款 +歆 +歇 +歉 +歌 +歎 +歐 +歓 +歙 +歛 +歡 +止 +正 +此 +步 +武 +歧 +歩 +歪 +歯 +歲 +歳 +歴 +歷 +歸 +歹 +死 +歼 +殁 +殃 +殆 +殇 +殉 +殊 +残 +殒 +殓 +殖 +殘 +殞 +殡 +殤 +殭 +殯 +殲 +殴 +段 +殷 +殺 +殼 +殿 +毀 +毁 +毂 +毅 +毆 +毋 +母 +毎 +每 +毒 +毓 +比 +毕 +毗 +毘 +毙 +毛 +毡 +毫 +毯 +毽 +氈 +氏 +氐 +民 +氓 +气 +氖 +気 +氙 +氛 +氟 +氡 +氢 +氣 +氤 +氦 +氧 +氨 +氪 +氫 +氮 +氯 +氰 +氲 +水 +氷 +永 +氹 +氾 +汀 +汁 +求 +汆 +汇 +汉 +汎 +汐 +汕 +汗 +汙 +汛 +汝 +汞 +江 +池 +污 +汤 +汨 +汩 +汪 +汰 +汲 +汴 +汶 +汹 +決 +汽 +汾 +沁 +沂 +沃 +沅 +沈 +沉 +沌 +沏 +沐 +沒 +沓 +沖 +沙 +沛 +沟 +没 +沢 +沣 +沥 +沦 +沧 +沪 +沫 +沭 +沮 +沱 +河 +沸 +油 +治 +沼 +沽 +沾 +沿 +況 +泄 +泉 +泊 +泌 +泓 +法 +泗 +泛 +泞 +泠 +泡 +波 +泣 +泥 +注 +泪 +泫 +泮 +泯 +泰 +泱 +泳 +泵 +泷 +泸 +泻 +泼 +泽 +泾 +洁 +洄 +洋 +洒 +洗 +洙 +洛 +洞 +津 +洩 +洪 +洮 +洱 +洲 +洵 +洶 +洸 +洹 +活 +洼 +洽 +派 +流 +浃 +浄 +浅 +浆 +浇 +浊 +测 +济 +浏 +浑 +浒 +浓 +浔 +浙 +浚 +浜 +浣 +浦 +浩 +浪 +浬 +浮 +浯 +浴 +海 +浸 +涂 +涅 +涇 +消 +涉 +涌 +涎 +涓 +涔 +涕 +涙 +涛 +涝 +涞 +涟 +涠 +涡 +涣 +涤 +润 +涧 +涨 +涩 +涪 +涮 +涯 +液 +涵 +涸 +涼 +涿 +淀 +淄 +淅 +淆 +淇 +淋 +淌 +淑 +淒 +淖 +淘 +淙 +淚 +淞 +淡 +淤 +淦 +淨 +淩 +淪 +淫 +淬 +淮 +深 +淳 +淵 +混 +淹 +淺 +添 +淼 +清 +済 +渉 +渊 +渋 +渍 +渎 +渐 +渔 +渗 +渙 +渚 +減 +渝 +渠 +渡 +渣 +渤 +渥 +渦 +温 +測 +渭 +港 +渲 +渴 +游 +渺 +渾 +湃 +湄 +湊 +湍 +湖 +湘 +湛 +湟 +湧 +湫 +湮 +湯 +湳 +湾 +湿 +満 +溃 +溅 +溉 +溏 +源 +準 +溜 +溝 +溟 +溢 +溥 +溧 +溪 +溫 +溯 +溱 +溴 +溶 +溺 +溼 +滁 +滂 +滄 +滅 +滇 +滋 +滌 +滑 +滓 +滔 +滕 +滙 +滚 +滝 +滞 +滟 +满 +滢 +滤 +滥 +滦 +滨 +滩 +滬 +滯 +滲 +滴 +滷 +滸 +滾 +滿 +漁 +漂 +漆 +漉 +漏 +漓 +演 +漕 +漠 +漢 +漣 +漩 +漪 +漫 +漬 +漯 +漱 +漲 +漳 +漸 +漾 +漿 +潆 +潇 +潋 +潍 +潑 +潔 +潘 +潛 +潜 +潞 +潟 +潢 +潤 +潦 +潧 +潭 +潮 +潰 +潴 +潸 +潺 +潼 +澀 +澄 +澆 +澈 +澍 +澎 +澗 +澜 +澡 +澤 +澧 +澱 +澳 +澹 +激 +濁 +濂 +濃 +濑 +濒 +濕 +濘 +濛 +濟 +濠 +濡 +濤 +濫 +濬 +濮 +濯 +濱 +濺 +濾 +瀅 +瀆 +瀉 +瀋 +瀏 +瀑 +瀕 +瀘 +瀚 +瀛 +瀝 +瀞 +瀟 +瀧 +瀨 +瀬 +瀰 +瀾 +灌 +灏 +灑 +灘 +灝 +灞 +灣 +火 +灬 +灭 +灯 +灰 +灵 +灶 +灸 +灼 +災 +灾 +灿 +炀 +炁 +炅 +炉 +炊 +炎 +炒 +炔 +炕 +炖 +炙 +炜 +炫 +炬 +炭 +炮 +炯 +炳 +炷 +炸 +点 +為 +炼 +炽 +烁 +烂 +烃 +烈 +烊 +烏 +烘 +烙 +烛 +烟 +烤 +烦 +烧 +烨 +烩 +烫 +烬 +热 +烯 +烷 +烹 +烽 +焉 +焊 +焕 +焖 +焗 +焘 +焙 +焚 +焜 +無 +焦 +焯 +焰 +焱 +然 +焼 +煅 +煉 +煊 +煌 +煎 +煒 +煖 +煙 +煜 +煞 +煤 +煥 +煦 +照 +煨 +煩 +煮 +煲 +煸 +煽 +熄 +熊 +熏 +熒 +熔 +熙 +熟 +熠 +熨 +熬 +熱 +熵 +熹 +熾 +燁 +燃 +燄 +燈 +燉 +燊 +燎 +燒 +燔 +燕 +燙 +燜 +營 +燥 +燦 +燧 +燭 +燮 +燴 +燻 +燼 +燿 +爆 +爍 +爐 +爛 +爪 +爬 +爭 +爰 +爱 +爲 +爵 +父 +爷 +爸 +爹 +爺 +爻 +爽 +爾 +牆 +片 +版 +牌 +牍 +牒 +牙 +牛 +牝 +牟 +牠 +牡 +牢 +牦 +牧 +物 +牯 +牲 +牴 +牵 +特 +牺 +牽 +犀 +犁 +犄 +犊 +犍 +犒 +犢 +犧 +犬 +犯 +状 +犷 +犸 +犹 +狀 +狂 +狄 +狈 +狎 +狐 +狒 +狗 +狙 +狞 +狠 +狡 +狩 +独 +狭 +狮 +狰 +狱 +狸 +狹 +狼 +狽 +猎 +猕 +猖 +猗 +猙 +猛 +猜 +猝 +猥 +猩 +猪 +猫 +猬 +献 +猴 +猶 +猷 +猾 +猿 +獄 +獅 +獎 +獐 +獒 +獗 +獠 +獣 +獨 +獭 +獰 +獲 +獵 +獷 +獸 +獺 +獻 +獼 +獾 +玄 +率 +玉 +王 +玑 +玖 +玛 +玟 +玠 +玥 +玩 +玫 +玮 +环 +现 +玲 +玳 +玷 +玺 +玻 +珀 +珂 +珅 +珈 +珉 +珊 +珍 +珏 +珐 +珑 +珙 +珞 +珠 +珣 +珥 +珩 +珪 +班 +珮 +珲 +珺 +現 +球 +琅 +理 +琇 +琉 +琊 +琍 +琏 +琐 +琛 +琢 +琥 +琦 +琨 +琪 +琬 +琮 +琰 +琲 +琳 +琴 +琵 +琶 +琺 +琼 +瑀 +瑁 +瑄 +瑋 +瑕 +瑗 +瑙 +瑚 +瑛 +瑜 +瑞 +瑟 +瑠 +瑣 +瑤 +瑩 +瑪 +瑯 +瑰 +瑶 +瑾 +璀 +璁 +璃 +璇 +璉 +璋 +璎 +璐 +璜 +璞 +璟 +璧 +璨 +環 +璽 +璿 +瓊 +瓏 +瓒 +瓜 +瓢 +瓣 +瓤 +瓦 +瓮 +瓯 +瓴 +瓶 +瓷 +甄 +甌 +甕 +甘 +甙 +甚 +甜 +生 +產 +産 +甥 +甦 +用 +甩 +甫 +甬 +甭 +甯 +田 +由 +甲 +申 +电 +男 +甸 +町 +画 +甾 +畀 +畅 +界 +畏 +畑 +畔 +留 +畜 +畝 +畢 +略 +畦 +番 +畫 +異 +畲 +畳 +畴 +當 +畸 +畹 +畿 +疆 +疇 +疊 +疏 +疑 +疔 +疖 +疗 +疙 +疚 +疝 +疟 +疡 +疣 +疤 +疥 +疫 +疮 +疯 +疱 +疲 +疳 +疵 +疸 +疹 +疼 +疽 +疾 +痂 +病 +症 +痈 +痉 +痊 +痍 +痒 +痔 +痕 +痘 +痙 +痛 +痞 +痠 +痢 +痣 +痤 +痧 +痨 +痪 +痫 +痰 +痱 +痴 +痹 +痺 +痼 +痿 +瘀 +瘁 +瘋 +瘍 +瘓 +瘘 +瘙 +瘟 +瘠 +瘡 +瘢 +瘤 +瘦 +瘧 +瘩 +瘪 +瘫 +瘴 +瘸 +瘾 +療 +癇 +癌 +癒 +癖 +癜 +癞 +癡 +癢 +癣 +癥 +癫 +癬 +癮 +癱 +癲 +癸 +発 +登 +發 +白 +百 +皂 +的 +皆 +皇 +皈 +皋 +皎 +皑 +皓 +皖 +皙 +皚 +皮 +皰 +皱 +皴 +皺 +皿 +盂 +盃 +盅 +盆 +盈 +益 +盎 +盏 +盐 +监 +盒 +盔 +盖 +盗 +盘 +盛 +盜 +盞 +盟 +盡 +監 +盤 +盥 +盧 +盪 +目 +盯 +盱 +盲 +直 +相 +盹 +盼 +盾 +省 +眈 +眉 +看 +県 +眙 +眞 +真 +眠 +眦 +眨 +眩 +眯 +眶 +眷 +眸 +眺 +眼 +眾 +着 +睁 +睇 +睏 +睐 +睑 +睛 +睜 +睞 +睡 +睢 +督 +睥 +睦 +睨 +睪 +睫 +睬 +睹 +睽 +睾 +睿 +瞄 +瞅 +瞇 +瞋 +瞌 +瞎 +瞑 +瞒 +瞓 +瞞 +瞟 +瞠 +瞥 +瞧 +瞩 +瞪 +瞬 +瞭 +瞰 +瞳 +瞻 +瞼 +瞿 +矇 +矍 +矗 +矚 +矛 +矜 +矢 +矣 +知 +矩 +矫 +短 +矮 +矯 +石 +矶 +矽 +矾 +矿 +码 +砂 +砌 +砍 +砒 +研 +砖 +砗 +砚 +砝 +砣 +砥 +砧 +砭 +砰 +砲 +破 +砷 +砸 +砺 +砼 +砾 +础 +硅 +硐 +硒 +硕 +硝 +硫 +硬 +确 +硯 +硼 +碁 +碇 +碉 +碌 +碍 +碎 +碑 +碓 +碗 +碘 +碚 +碛 +碟 +碣 +碧 +碩 +碰 +碱 +碳 +碴 +確 +碼 +碾 +磁 +磅 +磊 +磋 +磐 +磕 +磚 +磡 +磨 +磬 +磯 +磲 +磷 +磺 +礁 +礎 +礙 +礡 +礦 +礪 +礫 +礴 +示 +礼 +社 +祀 +祁 +祂 +祇 +祈 +祉 +祎 +祐 +祕 +祖 +祗 +祚 +祛 +祜 +祝 +神 +祟 +祠 +祢 +祥 +票 +祭 +祯 +祷 +祸 +祺 +祿 +禀 +禁 +禄 +禅 +禍 +禎 +福 +禛 +禦 +禧 +禪 +禮 +禱 +禹 +禺 +离 +禽 +禾 +禿 +秀 +私 +秃 +秆 +秉 +秋 +种 +科 +秒 +秘 +租 +秣 +秤 +秦 +秧 +秩 +秭 +积 +称 +秸 +移 +秽 +稀 +稅 +程 +稍 +税 +稔 +稗 +稚 +稜 +稞 +稟 +稠 +稣 +種 +稱 +稲 +稳 +稷 +稹 +稻 +稼 +稽 +稿 +穀 +穂 +穆 +穌 +積 +穎 +穗 +穢 +穩 +穫 +穴 +究 +穷 +穹 +空 +穿 +突 +窃 +窄 +窈 +窍 +窑 +窒 +窓 +窕 +窖 +窗 +窘 +窜 +窝 +窟 +窠 +窥 +窦 +窨 +窩 +窪 +窮 +窯 +窺 +窿 +竄 +竅 +竇 +竊 +立 +竖 +站 +竜 +竞 +竟 +章 +竣 +童 +竭 +端 +競 +竹 +竺 +竽 +竿 +笃 +笆 +笈 +笋 +笏 +笑 +笔 +笙 +笛 +笞 +笠 +符 +笨 +第 +笹 +笺 +笼 +筆 +等 +筊 +筋 +筍 +筏 +筐 +筑 +筒 +答 +策 +筛 +筝 +筠 +筱 +筲 +筵 +筷 +筹 +签 +简 +箇 +箋 +箍 +箏 +箐 +箔 +箕 +算 +箝 +管 +箩 +箫 +箭 +箱 +箴 +箸 +節 +篁 +範 +篆 +篇 +築 +篑 +篓 +篙 +篝 +篠 +篡 +篤 +篩 +篪 +篮 +篱 +篷 +簇 +簌 +簍 +簡 +簦 +簧 +簪 +簫 +簷 +簸 +簽 +簾 +簿 +籁 +籃 +籌 +籍 +籐 +籟 +籠 +籤 +籬 +籮 +籲 +米 +类 +籼 +籽 +粄 +粉 +粑 +粒 +粕 +粗 +粘 +粟 +粤 +粥 +粧 +粪 +粮 +粱 +粲 +粳 +粵 +粹 +粼 +粽 +精 +粿 +糅 +糊 +糍 +糕 +糖 +糗 +糙 +糜 +糞 +糟 +糠 +糧 +糬 +糯 +糰 +糸 +系 +糾 +紀 +紂 +約 +紅 +紉 +紊 +紋 +納 +紐 +紓 +純 +紗 +紘 +紙 +級 +紛 +紜 +素 +紡 +索 +紧 +紫 +紮 +累 +細 +紳 +紹 +紺 +終 +絃 +組 +絆 +経 +結 +絕 +絞 +絡 +絢 +給 +絨 +絮 +統 +絲 +絳 +絵 +絶 +絹 +綁 +綏 +綑 +經 +継 +続 +綜 +綠 +綢 +綦 +綫 +綬 +維 +綱 +網 +綴 +綵 +綸 +綺 +綻 +綽 +綾 +綿 +緊 +緋 +総 +緑 +緒 +緘 +線 +緝 +緞 +締 +緣 +編 +緩 +緬 +緯 +練 +緹 +緻 +縁 +縄 +縈 +縛 +縝 +縣 +縫 +縮 +縱 +縴 +縷 +總 +績 +繁 +繃 +繆 +繇 +繋 +織 +繕 +繚 +繞 +繡 +繩 +繪 +繫 +繭 +繳 +繹 +繼 +繽 +纂 +續 +纍 +纏 +纓 +纔 +纖 +纜 +纠 +红 +纣 +纤 +约 +级 +纨 +纪 +纫 +纬 +纭 +纯 +纰 +纱 +纲 +纳 +纵 +纶 +纷 +纸 +纹 +纺 +纽 +纾 +线 +绀 +练 +组 +绅 +细 +织 +终 +绊 +绍 +绎 +经 +绑 +绒 +结 +绔 +绕 +绘 +给 +绚 +绛 +络 +绝 +绞 +统 +绡 +绢 +绣 +绥 +绦 +继 +绩 +绪 +绫 +续 +绮 +绯 +绰 +绳 +维 +绵 +绶 +绷 +绸 +绻 +综 +绽 +绾 +绿 +缀 +缄 +缅 +缆 +缇 +缈 +缉 +缎 +缓 +缔 +缕 +编 +缘 +缙 +缚 +缜 +缝 +缠 +缢 +缤 +缥 +缨 +缩 +缪 +缭 +缮 +缰 +缱 +缴 +缸 +缺 +缽 +罂 +罄 +罌 +罐 +网 +罔 +罕 +罗 +罚 +罡 +罢 +罩 +罪 +置 +罰 +署 +罵 +罷 +罹 +羁 +羅 +羈 +羊 +羌 +美 +羔 +羚 +羞 +羟 +羡 +羣 +群 +羥 +羧 +羨 +義 +羯 +羲 +羸 +羹 +羽 +羿 +翁 +翅 +翊 +翌 +翎 +習 +翔 +翘 +翟 +翠 +翡 +翦 +翩 +翰 +翱 +翳 +翹 +翻 +翼 +耀 +老 +考 +耄 +者 +耆 +耋 +而 +耍 +耐 +耒 +耕 +耗 +耘 +耙 +耦 +耨 +耳 +耶 +耷 +耸 +耻 +耽 +耿 +聂 +聆 +聊 +聋 +职 +聒 +联 +聖 +聘 +聚 +聞 +聪 +聯 +聰 +聲 +聳 +聴 +聶 +職 +聽 +聾 +聿 +肃 +肄 +肅 +肆 +肇 +肉 +肋 +肌 +肏 +肓 +肖 +肘 +肚 +肛 +肝 +肠 +股 +肢 +肤 +肥 +肩 +肪 +肮 +肯 +肱 +育 +肴 +肺 +肽 +肾 +肿 +胀 +胁 +胃 +胄 +胆 +背 +胍 +胎 +胖 +胚 +胛 +胜 +胝 +胞 +胡 +胤 +胥 +胧 +胫 +胭 +胯 +胰 +胱 +胳 +胴 +胶 +胸 +胺 +能 +脂 +脅 +脆 +脇 +脈 +脉 +脊 +脍 +脏 +脐 +脑 +脓 +脖 +脘 +脚 +脛 +脣 +脩 +脫 +脯 +脱 +脲 +脳 +脸 +脹 +脾 +腆 +腈 +腊 +腋 +腌 +腎 +腐 +腑 +腓 +腔 +腕 +腥 +腦 +腩 +腫 +腭 +腮 +腰 +腱 +腳 +腴 +腸 +腹 +腺 +腻 +腼 +腾 +腿 +膀 +膈 +膊 +膏 +膑 +膘 +膚 +膛 +膜 +膝 +膠 +膦 +膨 +膩 +膳 +膺 +膻 +膽 +膾 +膿 +臀 +臂 +臃 +臆 +臉 +臊 +臍 +臓 +臘 +臟 +臣 +臥 +臧 +臨 +自 +臬 +臭 +至 +致 +臺 +臻 +臼 +臾 +舀 +舂 +舅 +舆 +與 +興 +舉 +舊 +舌 +舍 +舎 +舐 +舒 +舔 +舖 +舗 +舛 +舜 +舞 +舟 +航 +舫 +般 +舰 +舱 +舵 +舶 +舷 +舸 +船 +舺 +舾 +艇 +艋 +艘 +艙 +艦 +艮 +良 +艰 +艱 +色 +艳 +艷 +艹 +艺 +艾 +节 +芃 +芈 +芊 +芋 +芍 +芎 +芒 +芙 +芜 +芝 +芡 +芥 +芦 +芩 +芪 +芫 +芬 +芭 +芮 +芯 +花 +芳 +芷 +芸 +芹 +芻 +芽 +芾 +苁 +苄 +苇 +苋 +苍 +苏 +苑 +苒 +苓 +苔 +苕 +苗 +苛 +苜 +苞 +苟 +苡 +苣 +若 +苦 +苫 +苯 +英 +苷 +苹 +苻 +茁 +茂 +范 +茄 +茅 +茉 +茎 +茏 +茗 +茜 +茧 +茨 +茫 +茬 +茭 +茯 +茱 +茲 +茴 +茵 +茶 +茸 +茹 +茼 +荀 +荃 +荆 +草 +荊 +荏 +荐 +荒 +荔 +荖 +荘 +荚 +荞 +荟 +荠 +荡 +荣 +荤 +荥 +荧 +荨 +荪 +荫 +药 +荳 +荷 +荸 +荻 +荼 +荽 +莅 +莆 +莉 +莊 +莎 +莒 +莓 +莖 +莘 +莞 +莠 +莢 +莧 +莪 +莫 +莱 +莲 +莴 +获 +莹 +莺 +莽 +莿 +菀 +菁 +菅 +菇 +菈 +菊 +菌 +菏 +菓 +菖 +菘 +菜 +菟 +菠 +菡 +菩 +華 +菱 +菲 +菸 +菽 +萁 +萃 +萄 +萊 +萋 +萌 +萍 +萎 +萘 +萝 +萤 +营 +萦 +萧 +萨 +萩 +萬 +萱 +萵 +萸 +萼 +落 +葆 +葉 +著 +葚 +葛 +葡 +董 +葦 +葩 +葫 +葬 +葭 +葯 +葱 +葳 +葵 +葷 +葺 +蒂 +蒋 +蒐 +蒔 +蒙 +蒜 +蒞 +蒟 +蒡 +蒨 +蒲 +蒸 +蒹 +蒻 +蒼 +蒿 +蓁 +蓄 +蓆 +蓉 +蓋 +蓑 +蓓 +蓖 +蓝 +蓟 +蓦 +蓬 +蓮 +蓼 +蓿 +蔑 +蔓 +蔔 +蔗 +蔘 +蔚 +蔡 +蔣 +蔥 +蔫 +蔬 +蔭 +蔵 +蔷 +蔺 +蔻 +蔼 +蔽 +蕁 +蕃 +蕈 +蕉 +蕊 +蕎 +蕙 +蕤 +蕨 +蕩 +蕪 +蕭 +蕲 +蕴 +蕻 +蕾 +薄 +薅 +薇 +薈 +薊 +薏 +薑 +薔 +薙 +薛 +薦 +薨 +薩 +薪 +薬 +薯 +薰 +薹 +藉 +藍 +藏 +藐 +藓 +藕 +藜 +藝 +藤 +藥 +藩 +藹 +藻 +藿 +蘆 +蘇 +蘊 +蘋 +蘑 +蘚 +蘭 +蘸 +蘼 +蘿 +虎 +虏 +虐 +虑 +虔 +處 +虚 +虛 +虜 +虞 +號 +虢 +虧 +虫 +虬 +虱 +虹 +虻 +虽 +虾 +蚀 +蚁 +蚂 +蚊 +蚌 +蚓 +蚕 +蚜 +蚝 +蚣 +蚤 +蚩 +蚪 +蚯 +蚱 +蚵 +蛀 +蛆 +蛇 +蛊 +蛋 +蛎 +蛐 +蛔 +蛙 +蛛 +蛟 +蛤 +蛭 +蛮 +蛰 +蛳 +蛹 +蛻 +蛾 +蜀 +蜂 +蜃 +蜆 +蜇 +蜈 +蜊 +蜍 +蜒 +蜓 +蜕 +蜗 +蜘 +蜚 +蜜 +蜡 +蜢 +蜥 +蜱 +蜴 +蜷 +蜻 +蜿 +蝇 +蝈 +蝉 +蝌 +蝎 +蝕 +蝗 +蝙 +蝟 +蝠 +蝦 +蝨 +蝴 +蝶 +蝸 +蝼 +螂 +螃 +融 +螞 +螢 +螨 +螯 +螳 +螺 +蟀 +蟄 +蟆 +蟋 +蟎 +蟑 +蟒 +蟠 +蟬 +蟲 +蟹 +蟻 +蟾 +蠅 +蠍 +蠔 +蠕 +蠛 +蠟 +蠡 +蠢 +蠣 +蠱 +蠶 +蠹 +蠻 +血 +衄 +衅 +衆 +行 +衍 +術 +衔 +街 +衙 +衛 +衝 +衞 +衡 +衢 +衣 +补 +表 +衩 +衫 +衬 +衮 +衰 +衲 +衷 +衹 +衾 +衿 +袁 +袂 +袄 +袅 +袈 +袋 +袍 +袒 +袖 +袜 +袞 +袤 +袪 +被 +袭 +袱 +裁 +裂 +装 +裆 +裊 +裏 +裔 +裕 +裘 +裙 +補 +裝 +裟 +裡 +裤 +裨 +裱 +裳 +裴 +裸 +裹 +製 +裾 +褂 +複 +褐 +褒 +褓 +褔 +褚 +褥 +褪 +褫 +褲 +褶 +褻 +襁 +襄 +襟 +襠 +襪 +襬 +襯 +襲 +西 +要 +覃 +覆 +覇 +見 +規 +覓 +視 +覚 +覦 +覧 +親 +覬 +観 +覷 +覺 +覽 +觀 +见 +观 +规 +觅 +视 +览 +觉 +觊 +觎 +觐 +觑 +角 +觞 +解 +觥 +触 +觸 +言 +訂 +計 +訊 +討 +訓 +訕 +訖 +託 +記 +訛 +訝 +訟 +訣 +訥 +訪 +設 +許 +訳 +訴 +訶 +診 +註 +証 +詆 +詐 +詔 +評 +詛 +詞 +詠 +詡 +詢 +詣 +試 +詩 +詫 +詬 +詭 +詮 +詰 +話 +該 +詳 +詹 +詼 +誅 +誇 +誉 +誌 +認 +誓 +誕 +誘 +語 +誠 +誡 +誣 +誤 +誥 +誦 +誨 +說 +説 +読 +誰 +課 +誹 +誼 +調 +諄 +談 +請 +諏 +諒 +論 +諗 +諜 +諡 +諦 +諧 +諫 +諭 +諮 +諱 +諳 +諷 +諸 +諺 +諾 +謀 +謁 +謂 +謄 +謊 +謎 +謐 +謔 +謗 +謙 +講 +謝 +謠 +謨 +謬 +謹 +謾 +譁 +證 +譎 +譏 +識 +譙 +譚 +譜 +警 +譬 +譯 +議 +譲 +譴 +護 +譽 +讀 +變 +讓 +讚 +讞 +计 +订 +认 +讥 +讧 +讨 +让 +讪 +讫 +训 +议 +讯 +记 +讲 +讳 +讴 +讶 +讷 +许 +讹 +论 +讼 +讽 +设 +访 +诀 +证 +诃 +评 +诅 +识 +诈 +诉 +诊 +诋 +词 +诏 +译 +试 +诗 +诘 +诙 +诚 +诛 +话 +诞 +诟 +诠 +诡 +询 +诣 +诤 +该 +详 +诧 +诩 +诫 +诬 +语 +误 +诰 +诱 +诲 +说 +诵 +诶 +请 +诸 +诺 +读 +诽 +课 +诿 +谀 +谁 +调 +谄 +谅 +谆 +谈 +谊 +谋 +谌 +谍 +谎 +谏 +谐 +谑 +谒 +谓 +谔 +谕 +谗 +谘 +谙 +谚 +谛 +谜 +谟 +谢 +谣 +谤 +谥 +谦 +谧 +谨 +谩 +谪 +谬 +谭 +谯 +谱 +谲 +谴 +谶 +谷 +豁 +豆 +豇 +豈 +豉 +豊 +豌 +豎 +豐 +豔 +豚 +象 +豢 +豪 +豫 +豬 +豹 +豺 +貂 +貅 +貌 +貓 +貔 +貘 +貝 +貞 +負 +財 +貢 +貧 +貨 +販 +貪 +貫 +責 +貯 +貰 +貳 +貴 +貶 +買 +貸 +費 +貼 +貽 +貿 +賀 +賁 +賂 +賃 +賄 +資 +賈 +賊 +賑 +賓 +賜 +賞 +賠 +賡 +賢 +賣 +賤 +賦 +質 +賬 +賭 +賴 +賺 +購 +賽 +贅 +贈 +贊 +贍 +贏 +贓 +贖 +贛 +贝 +贞 +负 +贡 +财 +责 +贤 +败 +账 +货 +质 +贩 +贪 +贫 +贬 +购 +贮 +贯 +贰 +贱 +贲 +贴 +贵 +贷 +贸 +费 +贺 +贻 +贼 +贾 +贿 +赁 +赂 +赃 +资 +赅 +赈 +赊 +赋 +赌 +赎 +赏 +赐 +赓 +赔 +赖 +赘 +赚 +赛 +赝 +赞 +赠 +赡 +赢 +赣 +赤 +赦 +赧 +赫 +赭 +走 +赳 +赴 +赵 +赶 +起 +趁 +超 +越 +趋 +趕 +趙 +趟 +趣 +趨 +足 +趴 +趵 +趸 +趺 +趾 +跃 +跄 +跆 +跋 +跌 +跎 +跑 +跖 +跚 +跛 +距 +跟 +跡 +跤 +跨 +跩 +跪 +路 +跳 +践 +跷 +跹 +跺 +跻 +踉 +踊 +踌 +踏 +踐 +踝 +踞 +踟 +踢 +踩 +踪 +踮 +踱 +踴 +踵 +踹 +蹂 +蹄 +蹇 +蹈 +蹉 +蹊 +蹋 +蹑 +蹒 +蹙 +蹟 +蹣 +蹤 +蹦 +蹩 +蹬 +蹭 +蹲 +蹴 +蹶 +蹺 +蹼 +蹿 +躁 +躇 +躉 +躊 +躋 +躍 +躏 +躪 +身 +躬 +躯 +躲 +躺 +軀 +車 +軋 +軌 +軍 +軒 +軟 +転 +軸 +軼 +軽 +軾 +較 +載 +輒 +輓 +輔 +輕 +輛 +輝 +輟 +輩 +輪 +輯 +輸 +輻 +輾 +輿 +轄 +轅 +轆 +轉 +轍 +轎 +轟 +车 +轧 +轨 +轩 +转 +轭 +轮 +软 +轰 +轲 +轴 +轶 +轻 +轼 +载 +轿 +较 +辄 +辅 +辆 +辇 +辈 +辉 +辊 +辍 +辐 +辑 +输 +辕 +辖 +辗 +辘 +辙 +辛 +辜 +辞 +辟 +辣 +辦 +辨 +辩 +辫 +辭 +辮 +辯 +辰 +辱 +農 +边 +辺 +辻 +込 +辽 +达 +迁 +迂 +迄 +迅 +过 +迈 +迎 +运 +近 +返 +还 +这 +进 +远 +违 +连 +迟 +迢 +迤 +迥 +迦 +迩 +迪 +迫 +迭 +述 +迴 +迷 +迸 +迹 +迺 +追 +退 +送 +适 +逃 +逅 +逆 +选 +逊 +逍 +透 +逐 +递 +途 +逕 +逗 +這 +通 +逛 +逝 +逞 +速 +造 +逢 +連 +逮 +週 +進 +逵 +逶 +逸 +逻 +逼 +逾 +遁 +遂 +遅 +遇 +遊 +運 +遍 +過 +遏 +遐 +遑 +遒 +道 +達 +違 +遗 +遙 +遛 +遜 +遞 +遠 +遢 +遣 +遥 +遨 +適 +遭 +遮 +遲 +遴 +遵 +遶 +遷 +選 +遺 +遼 +遽 +避 +邀 +邁 +邂 +邃 +還 +邇 +邈 +邊 +邋 +邏 +邑 +邓 +邕 +邛 +邝 +邢 +那 +邦 +邨 +邪 +邬 +邮 +邯 +邰 +邱 +邳 +邵 +邸 +邹 +邺 +邻 +郁 +郅 +郊 +郎 +郑 +郜 +郝 +郡 +郢 +郤 +郦 +郧 +部 +郫 +郭 +郴 +郵 +郷 +郸 +都 +鄂 +鄉 +鄒 +鄔 +鄙 +鄞 +鄢 +鄧 +鄭 +鄰 +鄱 +鄲 +鄺 +酉 +酊 +酋 +酌 +配 +酐 +酒 +酗 +酚 +酝 +酢 +酣 +酥 +酩 +酪 +酬 +酮 +酯 +酰 +酱 +酵 +酶 +酷 +酸 +酿 +醃 +醇 +醉 +醋 +醍 +醐 +醒 +醚 +醛 +醜 +醞 +醣 +醪 +醫 +醬 +醮 +醯 +醴 +醺 +釀 +釁 +采 +釉 +释 +釋 +里 +重 +野 +量 +釐 +金 +釗 +釘 +釜 +針 +釣 +釦 +釧 +釵 +鈀 +鈉 +鈍 +鈎 +鈔 +鈕 +鈞 +鈣 +鈦 +鈪 +鈴 +鈺 +鈾 +鉀 +鉄 +鉅 +鉉 +鉑 +鉗 +鉚 +鉛 +鉤 +鉴 +鉻 +銀 +銃 +銅 +銑 +銓 +銖 +銘 +銜 +銬 +銭 +銮 +銳 +銷 +銹 +鋁 +鋅 +鋒 +鋤 +鋪 +鋰 +鋸 +鋼 +錄 +錐 +錘 +錚 +錠 +錢 +錦 +錨 +錫 +錮 +錯 +録 +錳 +錶 +鍊 +鍋 +鍍 +鍛 +鍥 +鍰 +鍵 +鍺 +鍾 +鎂 +鎊 +鎌 +鎏 +鎔 +鎖 +鎗 +鎚 +鎧 +鎬 +鎮 +鎳 +鏈 +鏖 +鏗 +鏘 +鏞 +鏟 +鏡 +鏢 +鏤 +鏽 +鐘 +鐮 +鐲 +鐳 +鐵 +鐸 +鐺 +鑄 +鑊 +鑑 +鑒 +鑣 +鑫 +鑰 +鑲 +鑼 +鑽 +鑾 +鑿 +针 +钉 +钊 +钎 +钏 +钒 +钓 +钗 +钙 +钛 +钜 +钝 +钞 +钟 +钠 +钡 +钢 +钣 +钤 +钥 +钦 +钧 +钨 +钩 +钮 +钯 +钰 +钱 +钳 +钴 +钵 +钺 +钻 +钼 +钾 +钿 +铀 +铁 +铂 +铃 +铄 +铅 +铆 +铉 +铎 +铐 +铛 +铜 +铝 +铠 +铡 +铢 +铣 +铤 +铨 +铩 +铬 +铭 +铮 +铰 +铲 +铵 +银 +铸 +铺 +链 +铿 +销 +锁 +锂 +锄 +锅 +锆 +锈 +锉 +锋 +锌 +锏 +锐 +锑 +错 +锚 +锟 +锡 +锢 +锣 +锤 +锥 +锦 +锭 +键 +锯 +锰 +锲 +锵 +锹 +锺 +锻 +镀 +镁 +镂 +镇 +镉 +镌 +镍 +镐 +镑 +镕 +镖 +镗 +镛 +镜 +镣 +镭 +镯 +镰 +镳 +镶 +長 +长 +門 +閃 +閉 +開 +閎 +閏 +閑 +閒 +間 +閔 +閘 +閡 +関 +閣 +閥 +閨 +閩 +閱 +閲 +閹 +閻 +閾 +闆 +闇 +闊 +闌 +闍 +闔 +闕 +闖 +闘 +關 +闡 +闢 +门 +闪 +闫 +闭 +问 +闯 +闰 +闲 +间 +闵 +闷 +闸 +闹 +闺 +闻 +闽 +闾 +阀 +阁 +阂 +阅 +阆 +阇 +阈 +阉 +阎 +阐 +阑 +阔 +阕 +阖 +阙 +阚 +阜 +队 +阡 +阪 +阮 +阱 +防 +阳 +阴 +阵 +阶 +阻 +阿 +陀 +陂 +附 +际 +陆 +陇 +陈 +陋 +陌 +降 +限 +陕 +陛 +陝 +陞 +陟 +陡 +院 +陣 +除 +陨 +险 +陪 +陰 +陲 +陳 +陵 +陶 +陷 +陸 +険 +陽 +隅 +隆 +隈 +隊 +隋 +隍 +階 +随 +隐 +隔 +隕 +隘 +隙 +際 +障 +隠 +隣 +隧 +隨 +險 +隱 +隴 +隶 +隸 +隻 +隼 +隽 +难 +雀 +雁 +雄 +雅 +集 +雇 +雉 +雋 +雌 +雍 +雎 +雏 +雑 +雒 +雕 +雖 +雙 +雛 +雜 +雞 +離 +難 +雨 +雪 +雯 +雰 +雲 +雳 +零 +雷 +雹 +電 +雾 +需 +霁 +霄 +霆 +震 +霈 +霉 +霊 +霍 +霎 +霏 +霑 +霓 +霖 +霜 +霞 +霧 +霭 +霰 +露 +霸 +霹 +霽 +霾 +靂 +靄 +靈 +青 +靓 +靖 +静 +靚 +靛 +靜 +非 +靠 +靡 +面 +靥 +靦 +革 +靳 +靴 +靶 +靼 +鞅 +鞋 +鞍 +鞏 +鞑 +鞘 +鞠 +鞣 +鞦 +鞭 +韆 +韋 +韌 +韓 +韜 +韦 +韧 +韩 +韬 +韭 +音 +韵 +韶 +韻 +響 +頁 +頂 +頃 +項 +順 +須 +頌 +預 +頑 +頒 +頓 +頗 +領 +頜 +頡 +頤 +頫 +頭 +頰 +頷 +頸 +頹 +頻 +頼 +顆 +題 +額 +顎 +顏 +顔 +願 +顛 +類 +顧 +顫 +顯 +顱 +顴 +页 +顶 +顷 +项 +顺 +须 +顼 +顽 +顾 +顿 +颁 +颂 +预 +颅 +领 +颇 +颈 +颉 +颊 +颌 +颍 +颐 +频 +颓 +颔 +颖 +颗 +题 +颚 +颛 +颜 +额 +颞 +颠 +颡 +颢 +颤 +颦 +颧 +風 +颯 +颱 +颳 +颶 +颼 +飄 +飆 +风 +飒 +飓 +飕 +飘 +飙 +飚 +飛 +飞 +食 +飢 +飨 +飩 +飪 +飯 +飲 +飼 +飽 +飾 +餃 +餅 +餉 +養 +餌 +餐 +餒 +餓 +餘 +餚 +餛 +餞 +餡 +館 +餮 +餵 +餾 +饅 +饈 +饋 +饌 +饍 +饑 +饒 +饕 +饗 +饞 +饥 +饨 +饪 +饬 +饭 +饮 +饯 +饰 +饱 +饲 +饴 +饵 +饶 +饷 +饺 +饼 +饽 +饿 +馀 +馁 +馄 +馅 +馆 +馈 +馋 +馍 +馏 +馒 +馔 +首 +馗 +香 +馥 +馨 +馬 +馭 +馮 +馳 +馴 +駁 +駄 +駅 +駆 +駐 +駒 +駕 +駛 +駝 +駭 +駱 +駿 +騁 +騎 +騏 +験 +騙 +騨 +騰 +騷 +驀 +驅 +驊 +驍 +驒 +驕 +驗 +驚 +驛 +驟 +驢 +驥 +马 +驭 +驮 +驯 +驰 +驱 +驳 +驴 +驶 +驷 +驸 +驹 +驻 +驼 +驾 +驿 +骁 +骂 +骄 +骅 +骆 +骇 +骈 +骊 +骋 +验 +骏 +骐 +骑 +骗 +骚 +骛 +骜 +骞 +骠 +骡 +骤 +骥 +骧 +骨 +骯 +骰 +骶 +骷 +骸 +骼 +髂 +髅 +髋 +髏 +髒 +髓 +體 +髖 +高 +髦 +髪 +髮 +髯 +髻 +鬃 +鬆 +鬍 +鬓 +鬚 +鬟 +鬢 +鬣 +鬥 +鬧 +鬱 +鬼 +魁 +魂 +魄 +魅 +魇 +魍 +魏 +魔 +魘 +魚 +魯 +魷 +鮑 +鮨 +鮪 +鮭 +鮮 +鯉 +鯊 +鯖 +鯛 +鯨 +鯰 +鯽 +鰍 +鰓 +鰭 +鰲 +鰻 +鰾 +鱈 +鱉 +鱔 +鱗 +鱷 +鱸 +鱼 +鱿 +鲁 +鲈 +鲍 +鲑 +鲛 +鲜 +鲟 +鲢 +鲤 +鲨 +鲫 +鲱 +鲲 +鲶 +鲷 +鲸 +鳃 +鳄 +鳅 +鳌 +鳍 +鳕 +鳖 +鳗 +鳝 +鳞 +鳥 +鳩 +鳳 +鳴 +鳶 +鴉 +鴕 +鴛 +鴦 +鴨 +鴻 +鴿 +鵑 +鵜 +鵝 +鵡 +鵬 +鵰 +鵲 +鶘 +鶩 +鶯 +鶴 +鷗 +鷲 +鷹 +鷺 +鸚 +鸞 +鸟 +鸠 +鸡 +鸢 +鸣 +鸥 +鸦 +鸨 +鸪 +鸭 +鸯 +鸳 +鸵 +鸽 +鸾 +鸿 +鹂 +鹃 +鹄 +鹅 +鹈 +鹉 +鹊 +鹌 +鹏 +鹑 +鹕 +鹘 +鹜 +鹞 +鹤 +鹦 +鹧 +鹫 +鹭 +鹰 +鹳 +鹵 +鹹 +鹼 +鹽 +鹿 +麂 +麋 +麒 +麓 +麗 +麝 +麟 +麥 +麦 +麩 +麴 +麵 +麸 +麺 +麻 +麼 +麽 +麾 +黃 +黄 +黍 +黎 +黏 +黑 +黒 +黔 +默 +黛 +黜 +黝 +點 +黠 +黨 +黯 +黴 +鼋 +鼎 +鼐 +鼓 +鼠 +鼬 +鼹 +鼻 +鼾 +齁 +齊 +齋 +齐 +齒 +齡 +齢 +齣 +齦 +齿 +龄 +龅 +龈 +龊 +龋 +龌 +龍 +龐 +龔 +龕 +龙 +龚 +龛 +龜 +龟 +︰ +︱ +︶ +︿ +﹁ +﹂ +﹍ +﹏ +﹐ +﹑ +﹒ +﹔ +﹕ +﹖ +﹗ +﹙ +﹚ +﹝ +﹞ +﹡ +﹣ +！ +＂ +＃ +＄ +％ +＆ +＇ +（ +） +＊ +＋ +， +－ +． +／ +０ +１ +２ +３ +４ +５ +６ +７ +８ +９ +： +； +＜ +＝ +＞ +？ +＠ +［ +＼ +］ +＾ +＿ +｀ +ａ +ｂ +ｃ +ｄ +ｅ +ｆ +ｇ +ｈ +ｉ +ｊ +ｋ +ｌ +ｍ +ｎ +ｏ +ｐ +ｑ +ｒ +ｓ +ｔ +ｕ +ｖ +ｗ +ｘ +ｙ +ｚ +｛ +｜ +｝ +～ +｡ +｢ +｣ +､ +･ +ｯ +ｰ +ｲ +ｸ +ｼ +ｽ +ﾄ +ﾉ +ﾌ +ﾗ +ﾙ +ﾝ +ﾞ +ﾟ +￣ +￥ +👍 +🔥 +😂 +😎 +... +yam +10 +2017 +12 +11 +2016 +20 +30 +15 +06 +lofter +##s +2015 +by +16 +14 +18 +13 +24 +17 +2014 +21 +##0 +22 +19 +25 +23 +com +100 +00 +05 +2013 +##a +03 +09 +08 +28 +##2 +50 +01 +04 +##1 +27 +02 +2012 +##3 +26 +##e +07 +##8 +##5 +##6 +##4 +##9 +##7 +29 +2011 +40 +##t +2010 +##o +##d +##i +2009 +##n +app +www +the +##m +31 +##c +##l +##y +##r +##g +2008 +60 +http +200 +qq +##p +80 +##f +google +pixnet +90 +cookies +tripadvisor +500 +##er +##k +35 +##h +facebook +2007 +2000 +70 +##b +of +##x +##u +45 +300 +iphone +32 +1000 +2006 +48 +ip +36 +in +38 +3d +##w +##ing +55 +ctrip +##on +##v +33 +##の +to +34 +400 +id +2005 +it +37 +windows +llc +top +99 +42 +39 +000 +led +at +##an +41 +51 +52 +46 +49 +43 +53 +44 +##z +android +58 +and +59 +2004 +56 +vr +##か +5000 +2003 +47 +blogthis +twitter +54 +##le +150 +ok +2018 +57 +75 +cn +no +ios +##in +##mm +##00 +800 +on +te +3000 +65 +2001 +360 +95 +ig +lv +120 +##ng +##を +##us +##に +pc +てす +── +600 +##te +85 +2002 +88 +##ed +html +ncc +wifi +email +64 +blog +is +##10 +##て +mail +online +##al +dvd +##ic +studio +##は +##℃ +##ia +##と +line +vip +72 +##q +98 +##ce +##en +for +##is +##ra +##es +##j +usb +net +cp +1999 +asia +4g +##cm +diy +new +3c +##お +ta +66 +language +vs +apple +tw +86 +web +##ne +ipad +62 +you +##re +101 +68 +##tion +ps +de +bt +pony +atm +##2017 +1998 +67 +##ch +ceo +##or +go +##na +av +pro +cafe +96 +pinterest +97 +63 +pixstyleme3c +##ta +more +said +##2016 +1997 +mp3 +700 +##ll +nba +jun +##20 +92 +tv +1995 +pm +61 +76 +nbsp +250 +##ie +linux +##ma +cd +110 +hd +##17 +78 +##ion +77 +6000 +am +##th +##st +94 +##se +##et +69 +180 +gdp +my +105 +81 +abc +89 +flash +79 +one +93 +1990 +1996 +##ck +gps +##も +##ly +web885 +106 +2020 +91 +##ge +4000 +1500 +xd +boss +isbn +1994 +org +##ry +me +love +##11 +0fork +73 +##12 +3g +##ter +##ar +71 +82 +##la +hotel +130 +1970 +pk +83 +87 +140 +ie +##os +##30 +##el +74 +##50 +seo +cpu +##ml +p2p +84 +may +##る +sun +tue +internet +cc +posted +youtube +##at +##ン +##man +ii +##ル +##15 +abs +nt +pdf +yahoo +ago +1980 +##it +news +mac +104 +##てす +##me +##り +java +1992 +spa +##de +##nt +hk +all +plus +la +1993 +##mb +##16 +##ve +west +##da +160 +air +##い +##ps +から +##to +1989 +logo +htc +php +https +fi +momo +##son +sat +##ke +##80 +ebd +suv +wi +day +apk +##88 +##um +mv +galaxy +wiki +or +brake +##ス +1200 +する +this +1991 +mon +##こ +❤2017 +po +##ない +javascript +life +home +june +##ss +system +900 +##ー +##０ +pp +1988 +world +fb +4k +br +##as +ic +ai +leonardo +safari +##60 +live +free +xx +wed +win7 +kiehl +##co +lg +o2o +##go +us +235 +1949 +mm +しい +vfm +kanye +##90 +##2015 +##id +jr +##ey +123 +rss +##sa +##ro +##am +##no +thu +fri +350 +##sh +##ki +103 +comments +name +##のて +##pe +##ine +max +1987 +8000 +uber +##mi +##ton +wordpress +office +1986 +1985 +##ment +107 +bd +win10 +##ld +##li +gmail +bb +dior +##rs +##ri +##rd +##ます +up +cad +##® +dr +して +read +##21 +をお +##io +##99 +url +1984 +pvc +paypal +show +policy +##40 +##ty +##18 +with +##★ +##01 +txt +102 +##ba +dna +from +post +mini +ar +taiwan +john +##ga +privacy +agoda +##13 +##ny +word +##24 +##22 +##by +##ur +##hz +1982 +##ang +265 +cookie +netscape +108 +##ka +##～ +##ad +house +share +note +ibm +code +hello +nike +sim +survey +##016 +1979 +1950 +wikia +##32 +##017 +5g +cbc +##tor +##kg +1983 +##rt +##14 +campaign +store +2500 +os +##ct +##ts +##° +170 +api +##ns +365 +excel +##な +##ao +##ら +##し +～～ +##nd +university +163 +には +518 +##70 +##ya +##il +##25 +pierre +ipo +0020 +897 +##23 +hotels +##ian +のお +125 +years +6606 +##ers +##26 +high +##day +time +##ay +bug +##line +##く +##す +##be +xp +talk2yam +yamservice +10000 +coco +##dy +sony +##ies +1978 +microsoft +david +people +##ha +1960 +instagram +intel +その +##ot +iso +1981 +##va +115 +##mo +##land +xxx +man +co +ltxsw +##ation +baby +220 +##pa +##ol +1945 +7000 +tag +450 +##ue +msn +##31 +oppo +##ト +##ca +control +##om +st +chrome +##ure +##ん +be +##き +lol +##19 +した +##bo +240 +lady +##100 +##way +##から +4600 +##ko +##do +##un +4s +corporation +168 +##ni +herme +##28 +ｃｐ +978 +##up +##06 +ui +##ds +ppt +admin +three +します +bbc +re +128 +##48 +ca +##015 +##35 +hp +##ee +tpp +##た +##ive +×× +root +##cc +##ました +##ble +##ity +adobe +park +114 +et +oled +city +##ex +##ler +##ap +china +##book +20000 +view +##ice +global +##km +your +hong +##mg +out +##ms +ng +ebay +##29 +menu +ubuntu +##cy +rom +##view +open +ktv +do +server +##lo +if +english +##ね +##５ +##oo +1600 +##02 +step1 +kong +club +135 +july +inc +1976 +mr +hi +##net +touch +##ls +##ii +michael +lcd +##05 +##33 +phone +james +step2 +1300 +ios9 +##box +dc +##２ +##ley +samsung +111 +280 +pokemon +css +##ent +##les +いいえ +##１ +s8 +atom +play +bmw +##said +sa +etf +ctrl +♥yoyo♥ +##55 +2025 +##2014 +##66 +adidas +amazon +1958 +##ber +##ner +visa +##77 +##der +1800 +connectivity +##hi +firefox +109 +118 +hr +so +style +mark +pop +ol +skip +1975 +as +##27 +##ir +##61 +190 +mba +##う +##ai +le +##ver +1900 +cafe2017 +lte +super +113 +129 +##ron +amd +like +##☆ +are +##ster +we +##sk +paul +data +international +##ft +longchamp +ssd +good +##ート +##ti +reply +##my +↓↓↓ +apr +star +##ker +source +136 +js +112 +get +force +photo +##one +126 +##2013 +##ow +link +bbs +1972 +goods +##lin +python +119 +##ip +game +##ics +##ません +blue +##● +520 +##45 +page +itunes +##03 +1955 +260 +1968 +gt +gif +618 +##ff +##47 +group +くたさい +about +bar +ganji +##nce +music +lee +not +1977 +1971 +1973 +##per +an +faq +comment +##って +days +##ock +116 +##bs +1974 +1969 +v1 +player +1956 +xbox +sql +fm +f1 +139 +##ah +210 +##lv +##mp +##000 +melody +1957 +##３ +550 +17life +199 +1966 +xml +market +##au +##71 +999 +##04 +what +gl +##95 +##age +tips +##68 +book +##ting +mysql +can +1959 +230 +##ung +wonderland +watch +10℃ +##ction +9000 +mar +mobile +1946 +1962 +article +##db +part +▲top +party +って +1967 +1964 +1948 +##07 +##ore +##op +この +dj +##78 +##38 +010 +main +225 +1965 +##ong +art +320 +ad +134 +020 +##73 +117 +pm2 +japan +228 +##08 +ts +1963 +##ica +der +sm +##36 +2019 +##wa +ct +##７ +##や +##64 +1937 +homemesh +search +##85 +##れは +##tv +##di +macbook +##９ +##くたさい +service +##♥ +type +った +750 +##ier +##si +##75 +##います +##ok +best +##ット +goris +lock +##った +cf +3m +big +##ut +ftp +carol +##vi +１０ +1961 +happy +sd +##ac +122 +anti +pe +cnn +iii +1920 +138 +##ラ +1940 +esp +jan +tags +##98 +##51 +august +vol +##86 +154 +##™ +##fs +##れ +##sion +design +ac +##ム +press +jordan +ppp +that +key +check +##６ +##tt +##㎡ +1080p +##lt +power +##42 +1952 +##bc +vivi +##ック +he +133 +121 +jpg +##rry +201 +175 +3500 +1947 +nb +##ted +##rn +しています +1954 +usd +##t00 +master +##ンク +001 +model +##58 +al +##09 +1953 +##34 +ram +goo +ても +##ui +127 +1930 +red +##ary +rpg +item +##pm +##41 +270 +##za +project +##2012 +hot +td +blogabstract +##ger +##62 +650 +##44 +gr2 +##します +##ｍ +black +electronic +nfc +year +asus +また +html5 +cindy +##hd +m3 +132 +esc +##od +booking +##53 +fed +tvb +##81 +##ina +mit +165 +##いる +chan +192 +distribution +next +になる +peter +bios +steam +cm +1941 +にも +pk10 +##ix +##65 +##91 +dec +nasa +##ana +icecat +00z +b1 +will +##46 +li +se +##ji +##み +##ard +oct +##ain +jp +##ze +##bi +cio +##56 +smart +h5 +##39 +##port +curve +vpn +##nm +##dia +utc +##あり +12345678910 +##52 +rmvb +chanel +a4 +miss +##and +##im +media +who +##63 +she +girl +5s +124 +vera +##して +class +vivo +king +##フ +##ei +national +ab +1951 +5cm +888 +145 +ipod +ap +1100 +5mm +211 +ms +2756 +##69 +mp4 +msci +##po +##89 +131 +mg +index +380 +##bit +##out +##zz +##97 +##67 +158 +apec +##８ +photoshop +opec +￥799 +ては +##96 +##tes +##ast +2g +○○ +##ール +￥2899 +##ling +##よ +##ory +1938 +##ical +kitty +content +##43 +step3 +##cn +win8 +155 +vc +1400 +iphone7 +robert +##した +tcl +137 +beauty +##87 +en +dollars +##ys +##oc +step +pay +yy +a1 +##2011 +##lly +##ks +##♪ +1939 +188 +download +1944 +sep +exe +ph +います +school +gb +center +pr +street +##board +uv +##37 +##lan +winrar +##que +##ua +##com +1942 +1936 +480 +gpu +##４ +ettoday +fu +tom +##54 +##ren +##via +149 +##72 +b2b +144 +##79 +##tch +rose +arm +mb +##49 +##ial +##nn +nvidia +step4 +mvp +00㎡ +york +156 +##イ +how +cpi +591 +2765 +gov +kg +joe +##xx +mandy +pa +##ser +copyright +fashion +1935 +don +##け +ecu +##ist +##art +erp +wap +have +##lm +talk +##ek +##ning +##if +ch +##ite +video +1943 +cs +san +iot +look +##84 +##2010 +##ku +october +##ux +trump +##hs +##ide +box +141 +first +##ins +april +##ight +##83 +185 +angel +protected +aa +151 +162 +x1 +m2 +##fe +##× +##ho +size +143 +min +ofo +fun +gomaji +ex +hdmi +food +dns +march +chris +kevin +##のか +##lla +##pp +##ec +ag +ems +6s +720p +##rm +##ham +off +##92 +asp +team +fandom +ed +299 +▌♥ +##ell +info +されています +##82 +sina +4066 +161 +##able +##ctor +330 +399 +315 +dll +rights +ltd +idc +jul +3kg +1927 +142 +ma +surface +##76 +##ク +～～～ +304 +mall +eps +146 +green +##59 +map +space +donald +v2 +sodu +##light +1931 +148 +1700 +まて +310 +reserved +htm +##han +##57 +2d +178 +mod +##ise +##tions +152 +ti +##shi +doc +1933 +icp +055 +wang +##ram +shopping +aug +##pi +##well +now +wam +b2 +からお +##hu +236 +1928 +##gb +266 +f2 +##93 +153 +mix +##ef +##uan +bwl +##plus +##res +core +##ess +tea +5℃ +hktvmall +nhk +##ate +list +##ese +301 +feb +4m +inn +ての +nov +159 +12345 +daniel +##ci +pass +##bet +##nk +coffee +202 +ssl +airbnb +##ute +fbi +woshipm +skype +ea +cg +sp +##fc +##www +yes +edge +alt +007 +##94 +fpga +##ght +##gs +iso9001 +さい +##ile +##wood +##uo +image +lin +icon +american +##em +1932 +set +says +##king +##tive +blogger +##74 +なと +256 +147 +##ox +##zy +##red +##ium +##lf +nokia +claire +##リ +##ding +november +lohas +##500 +##tic +##マ +##cs +##ある +##che +##ire +##gy +##ult +db +january +win +##カ +166 +road +ptt +##ま +##つ +198 +##fa +##mer +anna +pchome +はい +udn +ef +420 +##time +##tte +2030 +##ア +g20 +white +かかります +1929 +308 +garden +eleven +di +##おります +chen +309b +777 +172 +young +cosplay +ちてない +4500 +bat +##123 +##tra +##ては +kindle +npc +steve +etc +##ern +##｜ +call +xperia +ces +travel +sk +s7 +##ous +1934 +##int +みいたたけます +183 +edu +file +cho +qr +##car +##our +186 +##ant +##ｄ +eric +1914 +rends +##jo +##する +mastercard +##2000 +kb +##min +290 +##ino +vista +##ris +##ud +jack +2400 +##set +169 +pos +1912 +##her +##ou +taipei +しく +205 +beta +##ませんか +232 +##fi +express +255 +body +##ill +aphojoy +user +december +meiki +##ick +tweet +richard +##av +##ᆫ +iphone6 +##dd +ちてすか +views +##mark +321 +pd +##００ +times +##▲ +level +##ash +10g +point +5l +##ome +208 +koreanmall +##ak +george +q2 +206 +wma +tcp +##200 +スタッフ +full +mlb +##lle +##watch +tm +run +179 +911 +smith +business +##und +1919 +color +##tal +222 +171 +##less +moon +4399 +##rl +update +pcb +shop +499 +157 +little +なし +end +##mhz +van +dsp +easy +660 +##house +##key +history +##ｏ +oh +##001 +##hy +##web +oem +let +was +##2009 +##gg +review +##wan +182 +##°c +203 +uc +title +##val +united +233 +2021 +##ons +doi +trivago +overdope +sbs +##ance +##ち +grand +special +573032185 +imf +216 +wx17house +##so +##ーム +audi +##he +london +william +##rp +##ake +science +beach +cfa +amp +ps4 +880 +##800 +##link +##hp +crm +ferragamo +bell +make +##eng +195 +under +zh +photos +2300 +##style +##ント +via +176 +da +##gi +company +i7 +##ray +thomas +370 +ufo +i5 +##max +plc +ben +back +research +8g +173 +mike +##pc +##ッフ +september +189 +##ace +vps +february +167 +pantos +wp +lisa +1921 +★★ +jquery +night +long +offer +##berg +##news +1911 +##いて +ray +fks +wto +せます +over +164 +340 +##all +##rus +1924 +##888 +##works +blogtitle +loftpermalink +##→ +187 +martin +test +ling +km +##め +15000 +fda +v3 +##ja +##ロ +ｗedding +かある +outlet +family +##ea +をこ +##top +story +##ness +salvatore +##lu +204 +swift +215 +room +している +oracle +##ul +1925 +sam +b2c +week +pi +rock +##のは +##ａ +##けと +##ean +##300 +##gle +cctv +after +chinese +##back +powered +x2 +##tan +1918 +##nes +##イン +canon +only +181 +##zi +##las +say +##oe +184 +##sd +221 +##bot +##world +##zo +sky +made +top100 +just +1926 +pmi +802 +234 +gap +##vr +177 +les +174 +▲topoct +ball +vogue +vi +ing +ofweek +cos +##list +##ort +▲topmay +##なら +##lon +として +last +##tc +##of +##bus +##gen +real +eva +##コ +a3 +nas +##lie +##ria +##coin +##bt +▲topapr +his +212 +cat +nata +vive +health +⋯⋯ +drive +sir +▲topmar +du +cup +##カー +##ook +##よう +##sy +alex +msg +tour +しました +3ce +##word +193 +ebooks +r8 +block +318 +##より +2200 +nice +pvp +207 +months +1905 +rewards +##ther +1917 +0800 +##xi +##チ +##sc +micro +850 +gg +blogfp +op +1922 +daily +m1 +264 +true +##bb +ml +##tar +##のお +##ky +anthony +196 +253 +##yo +state +218 +##ara +##aa +##rc +##tz +##ston +より +gear +##eo +##ade +ge +see +1923 +##win +##ura +ss +heart +##den +##ita +down +##sm +el +png +2100 +610 +rakuten +whatsapp +bay +dream +add +##use +680 +311 +pad +gucci +mpv +##ode +##fo +island +▲topjun +##▼ +223 +jason +214 +chicago +##❤ +しの +##hone +io +##れる +##ことか +sogo +be2 +##ology +990 +cloud +vcd +##con +2～3 +##ford +##joy +##kb +##こさいます +##rade +but +##ach +docker +##ful +rfid +ul +##ase +hit +ford +##star +580 +##○ +１１ +a2 +sdk +reading +edited +##are +cmos +##mc +238 +siri +light +##ella +##ため +bloomberg +##read +pizza +##ison +jimmy +##vm +college +node +journal +ba +18k +##play +245 +##cer +２０ +magic +##yu +191 +jump +288 +tt +##ings +asr +##lia +3200 +step5 +network +##cd +mc +いします +1234 +pixstyleme +273 +##600 +2800 +money +★★★★★ +1280 +１２ +430 +bl +みの +act +##tus +tokyo +##rial +##life +emba +##ae +saas +tcs +##rk +##wang +summer +##sp +ko +##ving +390 +premium +##その +netflix +##ヒ +uk +mt +##lton +right +frank +two +209 +える +##ple +##cal +021 +##んな +##sen +##ville +hold +nexus +dd +##ius +てお +##mah +##なく +tila +zero +820 +ce +##tin +resort +##ws +charles +old +p10 +5d +report +##360 +##ru +##には +bus +vans +lt +##est +pv +##レ +links +rebecca +##ツ +##dm +azure +##365 +きな +limited +bit +4gb +##mon +1910 +moto +##eam +213 +1913 +var +eos +なとの +226 +blogspot +された +699 +e3 +dos +dm +fc +##ments +##ik +##kw +boy +##bin +##ata +960 +er +##せ +219 +##vin +##tu +##ula +194 +##∥ +station +##ろ +##ature +835 +files +zara +hdr +top10 +nature +950 +magazine +s6 +marriott +##シ +avira +case +##っと +tab +##ran +tony +##home +oculus +im +##ral +jean +saint +cry +307 +rosie +##force +##ini +ice +##bert +のある +##nder +##mber +pet +2600 +##◆ +plurk +▲topdec +##sis +00kg +▲topnov +720 +##ence +tim +##ω +##nc +##ても +##name +log +ips +great +ikea +malaysia +unix +##イト +3600 +##ncy +##nie +12000 +akb48 +##ye +##oid +404 +##chi +##いた +oa +xuehai +##1000 +##orm +##rf +275 +さん +##ware +##リー +980 +ho +##pro +text +##era +560 +bob +227 +##ub +##2008 +8891 +scp +avi +##zen +2022 +mi +wu +museum +qvod +apache +lake +jcb +▲topaug +★★★ +ni +##hr +hill +302 +ne +weibo +490 +ruby +##ーシ +##ヶ +##row +4d +▲topjul +iv +##ish +github +306 +mate +312 +##スト +##lot +##ane +andrew +のハイト +##tina +t1 +rf +ed2k +##vel +##900 +way +final +りの +ns +5a +705 +197 +##メ +sweet +bytes +##ene +▲topjan +231 +##cker +##2007 +##px +100g +topapp +229 +helpapp +rs +low +14k +g4g +care +630 +ldquo +あり +##fork +leave +rm +edition +##gan +##zon +##qq +▲topsep +##google +##ism +gold +224 +explorer +##zer +toyota +category +select +visual +##labels +restaurant +##md +posts +s1 +##ico +もっと +angelababy +123456 +217 +sports +s3 +mbc +1915 +してくたさい +shell +x86 +candy +##new +kbs +face +xl +470 +##here +4a +swissinfo +v8 +▲topfeb +dram +##ual +##vice +3a +##wer +sport +q1 +ios10 +public +int +card +##ｃ +ep +au +rt +##れた +1080 +bill +##mll +kim +３０ +460 +wan +##uk +##ミ +x3 +298 +0t +scott +##ming +239 +e5 +##3d +h7n9 +worldcat +brown +##あります +##vo +##led +##580 +##ax +249 +410 +##ert +paris +##～6 +polo +925 +##lr +599 +##ナ +capital +##hing +bank +cv +1g +##chat +##ｓ +##たい +adc +##ule +2m +##ｅ +digital +hotmail +268 +##pad +870 +bbq +quot +##ring +before +wali +##まて +mcu +2k +2b +という +costco +316 +north +333 +switch +##city +##ｐ +philips +##mann +management +panasonic +##cl +##vd +##ping +##rge +alice +##lk +##ましょう +css3 +##ney +vision +alpha +##ular +##400 +##tter +lz +にお +##ありません +mode +gre +1916 +pci +##tm +237 +1～2 +##yan +##そ +について +##let +##キ +work +war +coach +ah +mary +##ᅵ +huang +##pt +a8 +pt +follow +##berry +1895 +##ew +a5 +ghost +##ション +##wn +##og +south +##code +girls +##rid +action +villa +git +r11 +table +games +##cket +error +##anonymoussaid +##ag +here +##ame +##gc +qa +##■ +##lis +gmp +##gin +vmalife +##cher +yu +wedding +##tis +demo +dragon +530 +soho +social +bye +##rant +river +orz +acer +325 +##↑ +##ース +##ats +261 +del +##ven +440 +ups +##ように +##ター +305 +value +macd +yougou +##dn +661 +##ano +ll +##urt +##rent +continue +script +##wen +##ect +paper +263 +319 +shift +##chel +##フト +##cat +258 +x5 +fox +243 +##さん +car +aaa +##blog +loading +##yn +##tp +kuso +799 +si +sns +イカせるテンマ +ヒンクテンマ3 +rmb +vdc +forest +central +prime +help +ultra +##rmb +##ような +241 +square +688 +##しい +のないフロクに +##field +##reen +##ors +##ju +c1 +start +510 +##air +##map +cdn +##wo +cba +stephen +m8 +100km +##get +opera +##base +##ood +vsa +com™ +##aw +##ail +251 +なのて +count +t2 +##ᅡ +##een +2700 +hop +##gp +vsc +tree +##eg +##ose +816 +285 +##ories +##shop +alphago +v4 +1909 +simon +##ᆼ +fluke62max +zip +スホンサー +##sta +louis +cr +bas +##～10 +bc +##yer +hadoop +##ube +##wi +1906 +0755 +hola +##low +place +centre +5v +d3 +##fer +252 +##750 +##media +281 +540 +0l +exchange +262 +series +##ハー +##san +eb +##bank +##ｋ +q3 +##nge +##mail +take +##lp +259 +1888 +client +east +cache +event +vincent +##ールを +きを +##nse +sui +855 +adchoice +##и +##stry +##なたの +246 +##zone +ga +apps +sea +##ab +248 +cisco +##タ +##rner +kymco +##care +dha +##pu +##yi +minkoff +royal +p1 +への +annie +269 +collection +kpi +playstation +257 +になります +866 +bh +##bar +queen +505 +radio +1904 +andy +armani +##xy +manager +iherb +##ery +##share +spring +raid +johnson +1908 +##ob +volvo +hall +##ball +v6 +our +taylor +##hk +bi +242 +##cp +kate +bo +water +technology +##rie +サイトは +277 +##ona +##sl +hpv +303 +gtx +hip +rdquo +jayz +stone +##lex +##rum +namespace +##やり +620 +##ale +##atic +des +##erson +##ql +##ves +##type +enter +##この +##てきます +d2 +##168 +##mix +##bian +との +a9 +jj +ky +##lc +access +movie +##hc +リストに +tower +##ration +##mit +ます +##nch +ua +tel +prefix +##o2 +1907 +##point +1901 +ott +～10 +##http +##ury +baidu +##ink +member +##logy +bigbang +nownews +##js +##shot +##tb +##こと +247 +eba +##tics +##lus +ける +v5 +spark +##ama +there +##ions +god +##lls +##down +hiv +##ress +burberry +day2 +##kv +◆◆ +jeff +related +film +edit +joseph +283 +##ark +cx +32gb +order +g9 +30000 +##ans +##tty +s5 +##bee +かあります +thread +xr +buy +sh +005 +land +spotify +mx +##ari +276 +##verse +×email +sf +why +##ことて +244 +7headlines +nego +sunny +dom +exo +401 +666 +positioning +fit +rgb +##tton +278 +kiss +alexa +adam +lp +みリストを +##ｇ +mp +##ties +##llow +amy +##du +np +002 +institute +271 +##rth +##lar +2345 +590 +##des +sidebar +１５ +imax +site +##cky +##kit +##ime +##009 +season +323 +##fun +##ンター +##ひ +gogoro +a7 +pu +lily +fire +twd600 +##ッセーシを +いて +##vis +30ml +##cture +##をお +information +##オ +close +friday +##くれる +yi +nick +てすか +##tta +##tel +6500 +##lock +cbd +economy +254 +かお +267 +tinker +double +375 +8gb +voice +##app +oops +channel +today +985 +##right +raw +xyz +##＋ +jim +edm +##cent +7500 +supreme +814 +ds +##its +##asia +dropbox +##てすか +##tti +books +272 +100ml +##tle +##ller +##ken +##more +##boy +sex +309 +##dom +t3 +##ider +##なります +##unch +1903 +810 +feel +5500 +##かった +##put +により +s2 +mo +##gh +men +ka +amoled +div +##tr +##n1 +port +howard +##tags +ken +dnf +##nus +adsense +##а +ide +##へ +buff +thunder +##town +##ique +has +##body +auto +pin +##erry +tee +てした +295 +number +##the +##013 +object +psp +cool +udnbkk +16gb +##mic +miui +##tro +most +r2 +##alk +##nity +1880 +±0 +##いました +428 +s4 +law +version +##oa +n1 +sgs +docomo +##tf +##ack +henry +fc2 +##ded +##sco +##014 +##rite +286 +0mm +linkedin +##ada +##now +wii +##ndy +ucbug +##◎ +sputniknews +legalminer +##ika +##xp +2gb +##bu +q10 +oo +b6 +come +##rman +cheese +ming +maker +##gm +nikon +##fig +ppi +kelly +##ります +jchere +てきます +ted +md +003 +fgo +tech +##tto +dan +soc +##gl +##len +hair +earth +640 +521 +img +##pper +##a1 +##てきる +##ロク +acca +##ition +##ference +suite +##ig +outlook +##mond +##cation +398 +##pr +279 +101vip +358 +##999 +282 +64gb +3800 +345 +airport +##over +284 +##おり +jones +##ith +lab +##su +##いるのて +co2 +town +piece +##llo +no1 +vmware +24h +##qi +focus +reader +##admin +##ora +tb +false +##log +1898 +know +lan +838 +##ces +f4 +##ume +motel +stop +##oper +na +flickr +netcomponents +##af +##─ +pose +williams +local +##ound +##cg +##site +##iko +いお +274 +5m +gsm +con +##ath +1902 +friends +##hip +cell +317 +##rey +780 +cream +##cks +012 +##dp +facebooktwitterpinterestgoogle +sso +324 +shtml +song +swiss +##mw +##キンク +lumia +xdd +string +tiffany +522 +marc +られた +insee +russell +sc +dell +##ations +ｏｋ +camera +289 +##vs +##flow +##late +classic +287 +##nter +stay +g1 +mtv +512 +##ever +##lab +##nger +qe +sata +ryan +d1 +50ml +cms +##cing +su +292 +3300 +editor +296 +##nap +security +sunday +association +##ens +##700 +##bra +acg +##かり +sofascore +とは +mkv +##ign +jonathan +gary +build +labels +##oto +tesla +moba +qi +gohappy +general +ajax +1024 +##かる +サイト +society +##test +##urs +wps +fedora +##ich +mozilla +328 +##480 +##dr +usa +urn +##lina +##ｒ +grace +##die +##try +##ader +1250 +##なり +elle +570 +##chen +##ᆯ +price +##ten +uhz +##ough +eq +##hen +states +push +session +balance +wow +506 +##cus +##py +when +##ward +##ep +34e +wong +library +prada +##サイト +##cle +running +##ree +313 +ck +date +q4 +##ctive +##ool +##＞ +mk +##ira +##163 +388 +die +secret +rq +dota +buffet +は１ヶ +e6 +##ez +pan +368 +ha +##card +##cha +2a +##さ +alan +day3 +eye +f3 +##end +france +keep +adi +rna +tvbs +##ala +solo +nova +##え +##tail +##ょう +support +##ries +##なる +##ved +base +copy +iis +fps +##ways +hero +hgih +profile +fish +mu +ssh +entertainment +chang +##wd +click +cake +##ond +pre +##tom +kic +pixel +##ov +##fl +product +6a +##pd +dear +##gate +es +yumi +audio +##² +##sky +echo +bin +where +##ture +329 +##ape +find +sap +isis +##なと +nand +##101 +##load +##ream +band +a6 +525 +never +##post +festival +50cm +##we +555 +guide +314 +zenfone +##ike +335 +gd +forum +jessica +strong +alexander +##ould +software +allen +##ious +program +360° +else +lohasthree +##gar +することかてきます +please +##れます +rc +##ggle +##ric +bim +50000 +##own +eclipse +355 +brian +3ds +##side +061 +361 +##other +##ける +##tech +##ator +485 +engine +##ged +##ｔ +plaza +##fit +cia +ngo +westbrook +shi +tbs +50mm +##みませんか +sci +291 +reuters +##ily +contextlink +##hn +af +##cil +bridge +very +##cel +1890 +cambridge +##ize +15g +##aid +##data +790 +frm +##head +award +butler +##sun +meta +##mar +america +ps3 +puma +pmid +##すか +lc +670 +kitchen +##lic +オーフン5 +きなしソフトサーヒス +そして +day1 +future +★★★★ +##text +##page +##rris +pm1 +##ket +fans +##っています +1001 +christian +bot +kids +trackback +##hai +c3 +display +##hl +n2 +1896 +idea +さんも +##sent +airmail +##ug +##men +pwm +けます +028 +##lution +369 +852 +awards +schemas +354 +asics +wikipedia +font +##tional +##vy +c2 +293 +##れている +##dget +##ein +っている +contact +pepper +スキル +339 +##～5 +294 +##uel +##ument +730 +##hang +みてす +q5 +##sue +rain +##ndi +wei +swatch +##cept +わせ +331 +popular +##ste +##tag +p2 +501 +trc +1899 +##west +##live +justin +honda +ping +messenger +##rap +v9 +543 +##とは +unity +appqq +はすへて +025 +leo +##tone +##テ +##ass +uniqlo +##010 +502 +her +jane +memory +moneydj +##tical +human +12306 +していると +##m2 +coc +miacare +##mn +tmt +##core +vim +kk +##may +fan +target +use +too +338 +435 +2050 +867 +737 +fast +##2c +services +##ope +omega +energy +##わ +pinkoi +1a +##なから +##rain +jackson +##ement +##シャンルの +374 +366 +そんな +p9 +rd +##ᆨ +1111 +##tier +##vic +zone +##│ +385 +690 +dl +isofix +cpa +m4 +322 +kimi +めて +davis +##lay +lulu +##uck +050 +weeks +qs +##hop +920 +##ｎ +ae +##ear +～5 +eia +405 +##fly +korea +jpeg +boost +##ship +small +##リア +1860 +eur +297 +425 +valley +##iel +simple +##ude +rn +k2 +##ena +されます +non +patrick +しているから +##ナー +feed +5757 +30g +process +well +qqmei +##thing +they +aws +lu +pink +##ters +##kin +または +board +##vertisement +wine +##ien +unicode +##dge +r1 +359 +##tant +いを +##twitter +##3c +cool1 +される +##れて +##ｌ +isp +##012 +standard +45㎡2 +402 +##150 +matt +##fu +326 +##iner +googlemsn +pixnetfacebookyahoo +##ラン +x7 +886 +##uce +メーカー +sao +##ev +##きました +##file +9678 +403 +xddd +shirt +6l +##rio +##hat +3mm +givenchy +ya +bang +##lio +monday +crystal +ロクイン +##abc +336 +head +890 +ubuntuforumwikilinuxpastechat +##vc +##～20 +##rity +cnc +7866 +ipv6 +null +1897 +##ost +yang +imsean +tiger +##fet +##ンス +352 +##＝ +dji +327 +ji +maria +##come +##んて +foundation +3100 +##beth +##なった +1m +601 +active +##aft +##don +3p +sr +349 +emma +##khz +living +415 +353 +1889 +341 +709 +457 +sas +x6 +##face +pptv +x4 +##mate +han +sophie +##jing +337 +fifa +##mand +other +sale +inwedding +##gn +てきちゃいます +##mmy +##pmlast +bad +nana +nbc +してみてくたさいね +なとはお +##wu +##かあります +##あ +note7 +single +##340 +せからこ +してくたさい♪この +しにはとんとんワークケートを +するとあなたにもっとマッチした +ならワークケートへ +もみつかっちゃうかも +ワークケートの +##bel +window +##dio +##ht +union +age +382 +１４ +##ivity +##ｙ +コメント +domain +neo +##isa +##lter +5k +f5 +steven +##cts +powerpoint +tft +self +g2 +ft +##テル +zol +##act +mwc +381 +343 +もう +nbapop +408 +てある +eds +ace +##room +previous +author +tomtom +il +##ets +hu +financial +☆☆☆ +っています +bp +5t +chi +1gb +##hg +fairmont +cross +008 +gay +h2 +function +##けて +356 +also +1b +625 +##ータ +##raph +1894 +3～5 +##ils +i3 +334 +avenue +##host +による +##bon +##tsu +message +navigation +50g +fintech +h6 +##ことを +8cm +##ject +##vas +##firm +credit +##wf +xxxx +form +##nor +##space +huawei +plan +json +sbl +##dc +machine +921 +392 +wish +##120 +##sol +windows7 +edward +##ために +development +washington +##nsis +lo +818 +##sio +##ym +##bor +planet +##～8 +##wt +ieee +gpa +##めて +camp +ann +gm +##tw +##oka +connect +##rss +##work +##atus +wall +chicken +soul +2mm +##times +fa +##ather +##cord +009 +##eep +hitachi +gui +harry +##pan +e1 +disney +##press +##ーション +wind +386 +frigidaire +##tl +liu +hsu +332 +basic +von +ev +いた +てきる +スホンサーサイト +learning +##ull +expedia +archives +change +##wei +santa +cut +ins +6gb +turbo +brand +cf1 +508 +004 +return +747 +##rip +h1 +##nis +##をこ +128gb +##にお +3t +application +しており +emc +rx +##oon +384 +quick +412 +15058 +wilson +wing +chapter +##bug +beyond +##cms +##dar +##oh +zoom +e2 +trip +sb +##nba +rcep +342 +aspx +ci +080 +gc +gnu +める +##count +advanced +dance +dv +##url +##ging +367 +8591 +am09 +shadow +battle +346 +##ｉ +##cia +##という +emily +##のてす +##tation +host +ff +techorz +sars +##mini +##mporary +##ering +nc +4200 +798 +##next +cma +##mbps +##gas +##ift +##dot +##ィ +455 +##～17 +amana +##りの +426 +##ros +ir +00㎡1 +##eet +##ible +##↓ +710 +ˋ▽ˊ +##aka +dcs +iq +##ｖ +l1 +##lor +maggie +##011 +##iu +588 +##～1 +830 +##gt +1tb +articles +create +##burg +##iki +database +fantasy +##rex +##cam +dlc +dean +##you +hard +path +gaming +victoria +maps +cb +##lee +##itor +overchicstoretvhome +systems +##xt +416 +p3 +sarah +760 +##nan +407 +486 +x9 +install +second +626 +##ann +##ph +##rcle +##nic +860 +##nar +ec +##とう +768 +metro +chocolate +##rian +～4 +##table +##しています +skin +##sn +395 +mountain +##0mm +inparadise +6m +7x24 +ib +4800 +##jia +eeworld +creative +g5 +g3 +357 +parker +ecfa +village +からの +18000 +sylvia +サーヒス +hbl +##ques +##onsored +##x2 +##きます +##v4 +##tein +ie6 +383 +##stack +389 +ver +##ads +##baby +sound +bbe +##110 +##lone +##uid +ads +022 +gundam +351 +thinkpad +006 +scrum +match +##ave +mems +##470 +##oy +##なりました +##talk +glass +lamigo +span +##eme +job +##a5 +jay +wade +kde +498 +##lace +ocean +tvg +##covery +##r3 +##ners +##rea +junior +think +##aine +cover +##ision +##sia +↓↓ +##bow +msi +413 +458 +406 +##love +711 +801 +soft +z2 +##pl +456 +1840 +mobil +mind +##uy +427 +nginx +##oi +めた +##rr +6221 +##mple +##sson +##ーシてす +371 +##nts +91tv +comhd +crv3000 +##uard +1868 +397 +deep +lost +field +gallery +##bia +rate +spf +redis +traction +930 +icloud +011 +なら +fe +jose +372 +##tory +into +sohu +fx +899 +379 +kicstart2 +##hia +すく +##～3 +##sit +ra +２４ +##walk +##xure +500g +##pact +pacific +xa +natural +carlo +##250 +##walker +1850 +##can +cto +gigi +516 +##サー +pen +##hoo +ob +matlab +##ｂ +##yy +13913459 +##iti +mango +##bbs +sense +c5 +oxford +##ニア +walker +jennifer +##ola +course +##bre +701 +##pus +##rder +lucky +075 +##ぁ +ivy +なお +##nia +sotheby +side +##ugh +joy +##orage +##ush +##bat +##dt +364 +r9 +##2d +##gio +511 +country +wear +##lax +##～7 +##moon +393 +seven +study +411 +348 +lonzo +8k +##ェ +evolution +##イフ +##kk +gs +kd +##レス +arduino +344 +b12 +##lux +arpg +##rdon +cook +##x5 +dark +five +##als +##ida +とても +sign +362 +##ちの +something +20mm +##nda +387 +##posted +fresh +tf +1870 +422 +cam +##mine +##skip +##form +##ssion +education +394 +##tee +dyson +stage +##jie +want +##night +epson +pack +あります +##ppy +テリヘル +##█ +wd +##eh +##rence +left +##lvin +golden +mhz +discovery +##trix +##n2 +loft +##uch +##dra +##sse +speed +～1 +1mdb +sorry +welcome +##urn +wave +gaga +##lmer +teddy +##160 +トラックハック +せよ +611 +##f2016 +378 +rp +##sha +rar +##あなたに +##きた +840 +holiday +##ュー +373 +074 +##vg +##nos +##rail +gartner +gi +6p +##dium +kit +488 +b3 +eco +##ろう +20g +sean +##stone +autocad +nu +##np +f16 +write +029 +m5 +##ias +images +atp +##dk +fsm +504 +1350 +ve +52kb +##xxx +##のに +##cake +414 +unit +lim +ru +1v +##ification +published +angela +16g +analytics +ak +##ｑ +##nel +gmt +##icon +again +##₂ +##bby +ios11 +445 +かこさいます +waze +いてす +##ハ +9985 +##ust +##ティー +framework +##007 +iptv +delete +52sykb +cl +wwdc +027 +30cm +##fw +##ての +1389 +##xon +brandt +##ses +##dragon +tc +vetements +anne +monte +modern +official +##へて +##ere +##nne +##oud +もちろん +５０ +etnews +##a2 +##graphy +421 +863 +##ちゃん +444 +##rtex +##てお +l2 +##gma +mount +ccd +たと +archive +morning +tan +ddos +e7 +##ホ +day4 +##ウ +gis +453 +its +495 +factory +bruce +pg +##ito +ってくたさい +guest +cdma +##lling +536 +n3 +しかし +3～4 +mega +eyes +ro +１３ +women +dac +church +##jun +singapore +##facebook +6991 +starbucks +##tos +##stin +##shine +zen +##mu +tina +20℃ +1893 +##たけて +503 +465 +request +##gence +qt +##っ +1886 +347 +363 +q7 +##zzi +diary +##tore +409 +##ead +468 +cst +##osa +canada +agent +va +##jiang +##ちは +##ーク +##lam +sg +##nix +##sday +##よって +g6 +##master +bing +##zl +charlie +１６ +8mm +nb40 +##ーン +thai +##ルフ +ln284ct +##itz +##2f +bonnie +##food +##lent +originals +##stro +##lts +418 +∟∣ +##bscribe +children +ntd +yesstyle +##かも +hmv +##tment +d5 +2cm +arts +sms +##pn +##я +##いい +topios9 +539 +lifestyle +virtual +##ague +xz +##deo +muji +024 +unt +##nnis +##ᅩ +faq1 +1884 +396 +##ette +fly +64㎡ +はしめまして +441 +curry +##pop +のこ +release +##← +##◆◆ +##cast +073 +ありな +500ml +##ews +5c +##stle +ios7 +##ima +787 +dog +lenovo +##r4 +roger +013 +cbs +vornado +100m +417 +##desk +##クok +##ald +1867 +9595 +2900 +##van +oil +##ｘ +some +break +common +##jy +##lines +g7 +twice +419 +ella +nano +belle +にこ +##mes +##self +##note +jb +##ことかてきます +benz +##との +##ova +451 +save +##wing +##ますのて +kai +りは +##hua +##rect +rainer +##unge +448 +##0m +adsl +##かな +guestname +##uma +##kins +##zu +tokichoi +##price +county +##med +##mus +rmk +391 +address +vm +えて +openload +##group +##hin +##iginal +amg +urban +##oz +jobs +emi +##public +beautiful +##sch +album +##dden +##bell +jerry +works +hostel +miller +##drive +##rmin +##１０ +376 +boot +828 +##370 +##fx +##cm～ +1885 +##nome +##ctionary +##oman +##lish +##cr +##hm +433 +##how +432 +francis +xi +c919 +b5 +evernote +##uc +vga +##3000 +coupe +##urg +##cca +##uality +019 +6g +れる +multi +##また +##ett +em +hey +##ani +##tax +##rma +inside +than +740 +leonnhurt +##jin +ict +れた +bird +notes +200mm +くの +##dical +##lli +result +442 +iu +ee +438 +smap +gopro +##last +yin +pure +998 +32g +けた +5kg +##dan +##rame +mama +##oot +bean +marketing +##hur +2l +bella +sync +xuite +##ground +515 +discuz +##getrelax +##ince +##bay +##5s +cj +##イス +gmat +apt +##pass +jing +##rix +c4 +rich +##とても +niusnews +##ello +bag +770 +##eting +##mobile +１８ +culture +015 +##のてすか +377 +1020 +area +##ience +616 +details +gp +universal +silver +dit +はお +private +ddd +u11 +kanshu +##ified +fung +##nny +dx +##520 +tai +475 +023 +##fr +##lean +3s +##pin +429 +##rin +25000 +ly +rick +##bility +usb3 +banner +##baru +##gion +metal +dt +vdf +1871 +karl +qualcomm +bear +1010 +oldid +ian +jo +##tors +population +##ernel +1882 +mmorpg +##mv +##bike +603 +##© +ww +friend +##ager +exhibition +##del +##pods +fpx +structure +##free +##tings +kl +##rley +##copyright +##mma +california +3400 +orange +yoga +4l +canmake +honey +##anda +##コメント +595 +nikkie +##ルハイト +dhl +publishing +##mall +##gnet +20cm +513 +##クセス +##┅ +e88 +970 +##dog +fishbase +##! +##" +### +##$ +##% +##& +##' +##( +##) +##* +##+ +##, +##- +##. +##/ +##: +##; +##< +##= +##> +##? +##@ +##[ +##\ +##] +##^ +##_ +##{ +##| +##} +##~ +##£ +##¤ +##¥ +##§ +##« +##± +##³ +##µ +##· +##¹ +##º +##» +##¼ +##ß +##æ +##÷ +##ø +##đ +##ŋ +##ɔ +##ə +##ɡ +##ʰ +##ˇ +##ˈ +##ˊ +##ˋ +##ˍ +##ː +##˙ +##˚ +##ˢ +##α +##β +##γ +##δ +##ε +##η +##θ +##ι +##κ +##λ +##μ +##ν +##ο +##π +##ρ +##ς +##σ +##τ +##υ +##φ +##χ +##ψ +##б +##в +##г +##д +##е +##ж +##з +##к +##л +##м +##н +##о +##п +##р +##с +##т +##у +##ф +##х +##ц +##ч +##ш +##ы +##ь +##і +##ا +##ب +##ة +##ت +##د +##ر +##س +##ع +##ل +##م +##ن +##ه +##و +##ي +##۩ +##ก +##ง +##น +##ม +##ย +##ร +##อ +##า +##เ +##๑ +##་ +##ღ +##ᄀ +##ᄁ +##ᄂ +##ᄃ +##ᄅ +##ᄆ +##ᄇ +##ᄈ +##ᄉ +##ᄋ +##ᄌ +##ᄎ +##ᄏ +##ᄐ +##ᄑ +##ᄒ +##ᅢ +##ᅣ +##ᅥ +##ᅦ +##ᅧ +##ᅨ +##ᅪ +##ᅬ +##ᅭ +##ᅮ +##ᅯ +##ᅲ +##ᅳ +##ᅴ +##ᆷ +##ᆸ +##ᆺ +##ᆻ +##ᗜ +##ᵃ +##ᵉ +##ᵍ +##ᵏ +##ᵐ +##ᵒ +##ᵘ +##‖ +##„ +##† +##• +##‥ +##‧ +##
 +##‰ +##′ +##″ +##‹ +##› +##※ +##‿ +##⁄ +##ⁱ +##⁺ +##ⁿ +##₁ +##₃ +##₄ +##€ +##№ +##ⅰ +##ⅱ +##ⅲ +##ⅳ +##ⅴ +##↔ +##↗ +##↘ +##⇒ +##∀ +##− +##∕ +##∙ +##√ +##∞ +##∟ +##∠ +##∣ +##∩ +##∮ +##∶ +##∼ +##∽ +##≈ +##≒ +##≡ +##≤ +##≥ +##≦ +##≧ +##≪ +##≫ +##⊙ +##⋅ +##⋈ +##⋯ +##⌒ +##① +##② +##③ +##④ +##⑤ +##⑥ +##⑦ +##⑧ +##⑨ +##⑩ +##⑴ +##⑵ +##⑶ +##⑷ +##⑸ +##⒈ +##⒉ +##⒊ +##⒋ +##ⓒ +##ⓔ +##ⓘ +##━ +##┃ +##┆ +##┊ +##┌ +##└ +##├ +##┣ +##═ +##║ +##╚ +##╞ +##╠ +##╭ +##╮ +##╯ +##╰ +##╱ +##╳ +##▂ +##▃ +##▅ +##▇ +##▉ +##▋ +##▌ +##▍ +##▎ +##□ +##▪ +##▫ +##▬ +##△ +##▶ +##► +##▽ +##◇ +##◕ +##◠ +##◢ +##◤ +##☀ +##☕ +##☞ +##☺ +##☼ +##♀ +##♂ +##♠ +##♡ +##♣ +##♦ +##♫ +##♬ +##✈ +##✔ +##✕ +##✖ +##✦ +##✨ +##✪ +##✰ +##✿ +##❀ +##➜ +##➤ +##⦿ +##、 +##。 +##〃 +##々 +##〇 +##〈 +##〉 +##《 +##》 +##「 +##」 +##『 +##』 +##【 +##】 +##〓 +##〔 +##〕 +##〖 +##〗 +##〜 +##〝 +##〞 +##ぃ +##ぇ +##ぬ +##ふ +##ほ +##む +##ゃ +##ゅ +##ゆ +##ょ +##゜ +##ゝ +##ァ +##ゥ +##エ +##ォ +##ケ +##サ +##セ +##ソ +##ッ +##ニ +##ヌ +##ネ +##ノ +##ヘ +##モ +##ャ +##ヤ +##ュ +##ユ +##ョ +##ヨ +##ワ +##ヲ +##・ +##ヽ +##ㄅ +##ㄆ +##ㄇ +##ㄉ +##ㄋ +##ㄌ +##ㄍ +##ㄎ +##ㄏ +##ㄒ +##ㄚ +##ㄛ +##ㄞ +##ㄟ +##ㄢ +##ㄤ +##ㄥ +##ㄧ +##ㄨ +##ㆍ +##㈦ +##㊣ +##㗎 +##一 +##丁 +##七 +##万 +##丈 +##三 +##上 +##下 +##不 +##与 +##丐 +##丑 +##专 +##且 +##丕 +##世 +##丘 +##丙 +##业 +##丛 +##东 +##丝 +##丞 +##丟 +##両 +##丢 +##两 +##严 +##並 +##丧 +##丨 +##个 +##丫 +##中 +##丰 +##串 +##临 +##丶 +##丸 +##丹 +##为 +##主 +##丼 +##丽 +##举 +##丿 +##乂 +##乃 +##久 +##么 +##义 +##之 +##乌 +##乍 +##乎 +##乏 +##乐 +##乒 +##乓 +##乔 +##乖 +##乗 +##乘 +##乙 +##乜 +##九 +##乞 +##也 +##习 +##乡 +##书 +##乩 +##买 +##乱 +##乳 +##乾 +##亀 +##亂 +##了 +##予 +##争 +##事 +##二 +##于 +##亏 +##云 +##互 +##五 +##井 +##亘 +##亙 +##亚 +##些 +##亜 +##亞 +##亟 +##亡 +##亢 +##交 +##亥 +##亦 +##产 +##亨 +##亩 +##享 +##京 +##亭 +##亮 +##亲 +##亳 +##亵 +##人 +##亿 +##什 +##仁 +##仃 +##仄 +##仅 +##仆 +##仇 +##今 +##介 +##仍 +##从 +##仏 +##仑 +##仓 +##仔 +##仕 +##他 +##仗 +##付 +##仙 +##仝 +##仞 +##仟 +##代 +##令 +##以 +##仨 +##仪 +##们 +##仮 +##仰 +##仲 +##件 +##价 +##任 +##份 +##仿 +##企 +##伉 +##伊 +##伍 +##伎 +##伏 +##伐 +##休 +##伕 +##众 +##优 +##伙 +##会 +##伝 +##伞 +##伟 +##传 +##伢 +##伤 +##伦 +##伪 +##伫 +##伯 +##估 +##伴 +##伶 +##伸 +##伺 +##似 +##伽 +##佃 +##但 +##佇 +##佈 +##位 +##低 +##住 +##佐 +##佑 +##体 +##佔 +##何 +##佗 +##佘 +##余 +##佚 +##佛 +##作 +##佝 +##佞 +##佟 +##你 +##佢 +##佣 +##佤 +##佥 +##佩 +##佬 +##佯 +##佰 +##佳 +##併 +##佶 +##佻 +##佼 +##使 +##侃 +##侄 +##來 +##侈 +##例 +##侍 +##侏 +##侑 +##侖 +##侗 +##供 +##依 +##侠 +##価 +##侣 +##侥 +##侦 +##侧 +##侨 +##侬 +##侮 +##侯 +##侵 +##侶 +##侷 +##便 +##係 +##促 +##俄 +##俊 +##俎 +##俏 +##俐 +##俑 +##俗 +##俘 +##俚 +##保 +##俞 +##俟 +##俠 +##信 +##俨 +##俩 +##俪 +##俬 +##俭 +##修 +##俯 +##俱 +##俳 +##俸 +##俺 +##俾 +##倆 +##倉 +##個 +##倌 +##倍 +##倏 +##們 +##倒 +##倔 +##倖 +##倘 +##候 +##倚 +##倜 +##借 +##倡 +##値 +##倦 +##倩 +##倪 +##倫 +##倬 +##倭 +##倶 +##债 +##值 +##倾 +##偃 +##假 +##偈 +##偉 +##偌 +##偎 +##偏 +##偕 +##做 +##停 +##健 +##側 +##偵 +##偶 +##偷 +##偻 +##偽 +##偿 +##傀 +##傅 +##傍 +##傑 +##傘 +##備 +##傚 +##傢 +##傣 +##傥 +##储 +##傩 +##催 +##傭 +##傲 +##傳 +##債 +##傷 +##傻 +##傾 +##僅 +##働 +##像 +##僑 +##僕 +##僖 +##僚 +##僥 +##僧 +##僭 +##僮 +##僱 +##僵 +##價 +##僻 +##儀 +##儂 +##億 +##儆 +##儉 +##儋 +##儒 +##儕 +##儘 +##償 +##儡 +##優 +##儲 +##儷 +##儼 +##儿 +##兀 +##允 +##元 +##兄 +##充 +##兆 +##兇 +##先 +##光 +##克 +##兌 +##免 +##児 +##兑 +##兒 +##兔 +##兖 +##党 +##兜 +##兢 +##入 +##內 +##全 +##兩 +##八 +##公 +##六 +##兮 +##兰 +##共 +##兲 +##关 +##兴 +##兵 +##其 +##具 +##典 +##兹 +##养 +##兼 +##兽 +##冀 +##内 +##円 +##冇 +##冈 +##冉 +##冊 +##册 +##再 +##冏 +##冒 +##冕 +##冗 +##写 +##军 +##农 +##冠 +##冢 +##冤 +##冥 +##冨 +##冪 +##冬 +##冯 +##冰 +##冲 +##决 +##况 +##冶 +##冷 +##冻 +##冼 +##冽 +##冾 +##净 +##凄 +##准 +##凇 +##凈 +##凉 +##凋 +##凌 +##凍 +##减 +##凑 +##凛 +##凜 +##凝 +##几 +##凡 +##凤 +##処 +##凪 +##凭 +##凯 +##凰 +##凱 +##凳 +##凶 +##凸 +##凹 +##出 +##击 +##函 +##凿 +##刀 +##刁 +##刃 +##分 +##切 +##刈 +##刊 +##刍 +##刎 +##刑 +##划 +##列 +##刘 +##则 +##刚 +##创 +##初 +##删 +##判 +##別 +##刨 +##利 +##刪 +##别 +##刮 +##到 +##制 +##刷 +##券 +##刹 +##刺 +##刻 +##刽 +##剁 +##剂 +##剃 +##則 +##剉 +##削 +##剋 +##剌 +##前 +##剎 +##剐 +##剑 +##剔 +##剖 +##剛 +##剜 +##剝 +##剣 +##剤 +##剥 +##剧 +##剩 +##剪 +##副 +##割 +##創 +##剷 +##剽 +##剿 +##劃 +##劇 +##劈 +##劉 +##劊 +##劍 +##劏 +##劑 +##力 +##劝 +##办 +##功 +##加 +##务 +##劣 +##动 +##助 +##努 +##劫 +##劭 +##励 +##劲 +##劳 +##労 +##劵 +##効 +##劾 +##势 +##勁 +##勃 +##勇 +##勉 +##勋 +##勐 +##勒 +##動 +##勖 +##勘 +##務 +##勛 +##勝 +##勞 +##募 +##勢 +##勤 +##勧 +##勳 +##勵 +##勸 +##勺 +##勻 +##勾 +##勿 +##匀 +##包 +##匆 +##匈 +##匍 +##匐 +##匕 +##化 +##北 +##匙 +##匝 +##匠 +##匡 +##匣 +##匪 +##匮 +##匯 +##匱 +##匹 +##区 +##医 +##匾 +##匿 +##區 +##十 +##千 +##卅 +##升 +##午 +##卉 +##半 +##卍 +##华 +##协 +##卑 +##卒 +##卓 +##協 +##单 +##卖 +##南 +##単 +##博 +##卜 +##卞 +##卟 +##占 +##卡 +##卢 +##卤 +##卦 +##卧 +##卫 +##卮 +##卯 +##印 +##危 +##即 +##却 +##卵 +##卷 +##卸 +##卻 +##卿 +##厂 +##厄 +##厅 +##历 +##厉 +##压 +##厌 +##厕 +##厘 +##厚 +##厝 +##原 +##厢 +##厥 +##厦 +##厨 +##厩 +##厭 +##厮 +##厲 +##厳 +##去 +##县 +##叁 +##参 +##參 +##又 +##叉 +##及 +##友 +##双 +##反 +##収 +##发 +##叔 +##取 +##受 +##变 +##叙 +##叛 +##叟 +##叠 +##叡 +##叢 +##口 +##古 +##句 +##另 +##叨 +##叩 +##只 +##叫 +##召 +##叭 +##叮 +##可 +##台 +##叱 +##史 +##右 +##叵 +##叶 +##号 +##司 +##叹 +##叻 +##叼 +##叽 +##吁 +##吃 +##各 +##吆 +##合 +##吉 +##吊 +##吋 +##同 +##名 +##后 +##吏 +##吐 +##向 +##吒 +##吓 +##吕 +##吖 +##吗 +##君 +##吝 +##吞 +##吟 +##吠 +##吡 +##否 +##吧 +##吨 +##吩 +##含 +##听 +##吭 +##吮 +##启 +##吱 +##吳 +##吴 +##吵 +##吶 +##吸 +##吹 +##吻 +##吼 +##吽 +##吾 +##呀 +##呂 +##呃 +##呆 +##呈 +##告 +##呋 +##呎 +##呐 +##呓 +##呕 +##呗 +##员 +##呛 +##呜 +##呢 +##呤 +##呦 +##周 +##呱 +##呲 +##味 +##呵 +##呷 +##呸 +##呻 +##呼 +##命 +##咀 +##咁 +##咂 +##咄 +##咆 +##咋 +##和 +##咎 +##咏 +##咐 +##咒 +##咔 +##咕 +##咖 +##咗 +##咘 +##咙 +##咚 +##咛 +##咣 +##咤 +##咦 +##咧 +##咨 +##咩 +##咪 +##咫 +##咬 +##咭 +##咯 +##咱 +##咲 +##咳 +##咸 +##咻 +##咽 +##咿 +##哀 +##品 +##哂 +##哄 +##哆 +##哇 +##哈 +##哉 +##哋 +##哌 +##响 +##哎 +##哏 +##哐 +##哑 +##哒 +##哔 +##哗 +##哟 +##員 +##哥 +##哦 +##哧 +##哨 +##哩 +##哪 +##哭 +##哮 +##哲 +##哺 +##哼 +##哽 +##唁 +##唄 +##唆 +##唇 +##唉 +##唏 +##唐 +##唑 +##唔 +##唠 +##唤 +##唧 +##唬 +##售 +##唯 +##唰 +##唱 +##唳 +##唷 +##唸 +##唾 +##啃 +##啄 +##商 +##啉 +##啊 +##問 +##啓 +##啕 +##啖 +##啜 +##啞 +##啟 +##啡 +##啤 +##啥 +##啦 +##啧 +##啪 +##啫 +##啬 +##啮 +##啰 +##啱 +##啲 +##啵 +##啶 +##啷 +##啸 +##啻 +##啼 +##啾 +##喀 +##喂 +##喃 +##善 +##喆 +##喇 +##喉 +##喊 +##喋 +##喎 +##喏 +##喔 +##喘 +##喙 +##喚 +##喜 +##喝 +##喟 +##喧 +##喪 +##喫 +##喬 +##單 +##喰 +##喱 +##喲 +##喳 +##喵 +##営 +##喷 +##喹 +##喺 +##喻 +##喽 +##嗅 +##嗆 +##嗇 +##嗎 +##嗑 +##嗒 +##嗓 +##嗔 +##嗖 +##嗚 +##嗜 +##嗝 +##嗟 +##嗡 +##嗣 +##嗤 +##嗦 +##嗨 +##嗪 +##嗬 +##嗯 +##嗰 +##嗲 +##嗳 +##嗶 +##嗷 +##嗽 +##嘀 +##嘅 +##嘆 +##嘈 +##嘉 +##嘌 +##嘍 +##嘎 +##嘔 +##嘖 +##嘗 +##嘘 +##嘚 +##嘛 +##嘜 +##嘞 +##嘟 +##嘢 +##嘣 +##嘤 +##嘧 +##嘩 +##嘭 +##嘮 +##嘯 +##嘰 +##嘱 +##嘲 +##嘴 +##嘶 +##嘸 +##嘹 +##嘻 +##嘿 +##噁 +##噌 +##噎 +##噓 +##噔 +##噗 +##噙 +##噜 +##噠 +##噢 +##噤 +##器 +##噩 +##噪 +##噬 +##噱 +##噴 +##噶 +##噸 +##噹 +##噻 +##噼 +##嚀 +##嚇 +##嚎 +##嚏 +##嚐 +##嚓 +##嚕 +##嚟 +##嚣 +##嚥 +##嚨 +##嚮 +##嚴 +##嚷 +##嚼 +##囂 +##囉 +##囊 +##囍 +##囑 +##囔 +##囗 +##囚 +##四 +##囝 +##回 +##囟 +##因 +##囡 +##团 +##団 +##囤 +##囧 +##囪 +##囫 +##园 +##困 +##囱 +##囲 +##図 +##围 +##囹 +##固 +##国 +##图 +##囿 +##圃 +##圄 +##圆 +##圈 +##國 +##圍 +##圏 +##園 +##圓 +##圖 +##團 +##圜 +##土 +##圣 +##圧 +##在 +##圩 +##圭 +##地 +##圳 +##场 +##圻 +##圾 +##址 +##坂 +##均 +##坊 +##坍 +##坎 +##坏 +##坐 +##坑 +##块 +##坚 +##坛 +##坝 +##坞 +##坟 +##坠 +##坡 +##坤 +##坦 +##坨 +##坪 +##坯 +##坳 +##坵 +##坷 +##垂 +##垃 +##垄 +##型 +##垒 +##垚 +##垛 +##垠 +##垢 +##垣 +##垦 +##垩 +##垫 +##垭 +##垮 +##垵 +##埂 +##埃 +##埋 +##城 +##埔 +##埕 +##埗 +##域 +##埠 +##埤 +##埵 +##執 +##埸 +##培 +##基 +##埼 +##堀 +##堂 +##堃 +##堅 +##堆 +##堇 +##堑 +##堕 +##堙 +##堡 +##堤 +##堪 +##堯 +##堰 +##報 +##場 +##堵 +##堺 +##堿 +##塊 +##塌 +##塑 +##塔 +##塗 +##塘 +##塚 +##塞 +##塢 +##塩 +##填 +##塬 +##塭 +##塵 +##塾 +##墀 +##境 +##墅 +##墉 +##墊 +##墒 +##墓 +##増 +##墘 +##墙 +##墜 +##增 +##墟 +##墨 +##墩 +##墮 +##墳 +##墻 +##墾 +##壁 +##壅 +##壆 +##壇 +##壊 +##壑 +##壓 +##壕 +##壘 +##壞 +##壟 +##壢 +##壤 +##壩 +##士 +##壬 +##壮 +##壯 +##声 +##売 +##壳 +##壶 +##壹 +##壺 +##壽 +##处 +##备 +##変 +##复 +##夏 +##夔 +##夕 +##外 +##夙 +##多 +##夜 +##够 +##夠 +##夢 +##夥 +##大 +##天 +##太 +##夫 +##夭 +##央 +##夯 +##失 +##头 +##夷 +##夸 +##夹 +##夺 +##夾 +##奂 +##奄 +##奇 +##奈 +##奉 +##奋 +##奎 +##奏 +##奐 +##契 +##奔 +##奕 +##奖 +##套 +##奘 +##奚 +##奠 +##奢 +##奥 +##奧 +##奪 +##奬 +##奮 +##女 +##奴 +##奶 +##奸 +##她 +##好 +##如 +##妃 +##妄 +##妆 +##妇 +##妈 +##妊 +##妍 +##妒 +##妓 +##妖 +##妘 +##妙 +##妝 +##妞 +##妣 +##妤 +##妥 +##妨 +##妩 +##妪 +##妮 +##妲 +##妳 +##妹 +##妻 +##妾 +##姆 +##姉 +##姊 +##始 +##姍 +##姐 +##姑 +##姒 +##姓 +##委 +##姗 +##姚 +##姜 +##姝 +##姣 +##姥 +##姦 +##姨 +##姪 +##姫 +##姬 +##姹 +##姻 +##姿 +##威 +##娃 +##娄 +##娅 +##娆 +##娇 +##娉 +##娑 +##娓 +##娘 +##娛 +##娜 +##娟 +##娠 +##娣 +##娥 +##娩 +##娱 +##娲 +##娴 +##娶 +##娼 +##婀 +##婁 +##婆 +##婉 +##婊 +##婕 +##婚 +##婢 +##婦 +##婧 +##婪 +##婭 +##婴 +##婵 +##婶 +##婷 +##婺 +##婿 +##媒 +##媚 +##媛 +##媞 +##媧 +##媲 +##媳 +##媽 +##媾 +##嫁 +##嫂 +##嫉 +##嫌 +##嫑 +##嫔 +##嫖 +##嫘 +##嫚 +##嫡 +##嫣 +##嫦 +##嫩 +##嫲 +##嫵 +##嫻 +##嬅 +##嬉 +##嬌 +##嬗 +##嬛 +##嬢 +##嬤 +##嬪 +##嬰 +##嬴 +##嬷 +##嬸 +##嬿 +##孀 +##孃 +##子 +##孑 +##孔 +##孕 +##孖 +##字 +##存 +##孙 +##孚 +##孛 +##孜 +##孝 +##孟 +##孢 +##季 +##孤 +##学 +##孩 +##孪 +##孫 +##孬 +##孰 +##孱 +##孳 +##孵 +##學 +##孺 +##孽 +##孿 +##宁 +##它 +##宅 +##宇 +##守 +##安 +##宋 +##完 +##宏 +##宓 +##宕 +##宗 +##官 +##宙 +##定 +##宛 +##宜 +##宝 +##实 +##実 +##宠 +##审 +##客 +##宣 +##室 +##宥 +##宦 +##宪 +##宫 +##宮 +##宰 +##害 +##宴 +##宵 +##家 +##宸 +##容 +##宽 +##宾 +##宿 +##寂 +##寄 +##寅 +##密 +##寇 +##富 +##寐 +##寒 +##寓 +##寛 +##寝 +##寞 +##察 +##寡 +##寢 +##寥 +##實 +##寧 +##寨 +##審 +##寫 +##寬 +##寮 +##寰 +##寵 +##寶 +##寸 +##对 +##寺 +##寻 +##导 +##対 +##寿 +##封 +##専 +##射 +##将 +##將 +##專 +##尉 +##尊 +##尋 +##對 +##導 +##小 +##少 +##尔 +##尕 +##尖 +##尘 +##尚 +##尝 +##尤 +##尧 +##尬 +##就 +##尴 +##尷 +##尸 +##尹 +##尺 +##尻 +##尼 +##尽 +##尾 +##尿 +##局 +##屁 +##层 +##屄 +##居 +##屆 +##屈 +##屉 +##届 +##屋 +##屌 +##屍 +##屎 +##屏 +##屐 +##屑 +##展 +##屜 +##属 +##屠 +##屡 +##屢 +##層 +##履 +##屬 +##屯 +##山 +##屹 +##屿 +##岀 +##岁 +##岂 +##岌 +##岐 +##岑 +##岔 +##岖 +##岗 +##岘 +##岙 +##岚 +##岛 +##岡 +##岩 +##岫 +##岬 +##岭 +##岱 +##岳 +##岷 +##岸 +##峇 +##峋 +##峒 +##峙 +##峡 +##峤 +##峥 +##峦 +##峨 +##峪 +##峭 +##峯 +##峰 +##峴 +##島 +##峻 +##峽 +##崁 +##崂 +##崆 +##崇 +##崎 +##崑 +##崔 +##崖 +##崗 +##崙 +##崛 +##崧 +##崩 +##崭 +##崴 +##崽 +##嵇 +##嵊 +##嵋 +##嵌 +##嵐 +##嵘 +##嵩 +##嵬 +##嵯 +##嶂 +##嶄 +##嶇 +##嶋 +##嶙 +##嶺 +##嶼 +##嶽 +##巅 +##巍 +##巒 +##巔 +##巖 +##川 +##州 +##巡 +##巢 +##工 +##左 +##巧 +##巨 +##巩 +##巫 +##差 +##己 +##已 +##巳 +##巴 +##巷 +##巻 +##巽 +##巾 +##巿 +##币 +##市 +##布 +##帅 +##帆 +##师 +##希 +##帐 +##帑 +##帕 +##帖 +##帘 +##帚 +##帛 +##帜 +##帝 +##帥 +##带 +##帧 +##師 +##席 +##帮 +##帯 +##帰 +##帳 +##帶 +##帷 +##常 +##帼 +##帽 +##幀 +##幂 +##幄 +##幅 +##幌 +##幔 +##幕 +##幟 +##幡 +##幢 +##幣 +##幫 +##干 +##平 +##年 +##并 +##幸 +##幹 +##幺 +##幻 +##幼 +##幽 +##幾 +##广 +##庁 +##広 +##庄 +##庆 +##庇 +##床 +##序 +##庐 +##库 +##应 +##底 +##庖 +##店 +##庙 +##庚 +##府 +##庞 +##废 +##庠 +##度 +##座 +##庫 +##庭 +##庵 +##庶 +##康 +##庸 +##庹 +##庾 +##廁 +##廂 +##廃 +##廈 +##廉 +##廊 +##廓 +##廖 +##廚 +##廝 +##廟 +##廠 +##廢 +##廣 +##廬 +##廳 +##延 +##廷 +##建 +##廿 +##开 +##弁 +##异 +##弃 +##弄 +##弈 +##弊 +##弋 +##式 +##弑 +##弒 +##弓 +##弔 +##引 +##弗 +##弘 +##弛 +##弟 +##张 +##弥 +##弦 +##弧 +##弩 +##弭 +##弯 +##弱 +##張 +##強 +##弹 +##强 +##弼 +##弾 +##彅 +##彆 +##彈 +##彌 +##彎 +##归 +##当 +##录 +##彗 +##彙 +##彝 +##形 +##彤 +##彥 +##彦 +##彧 +##彩 +##彪 +##彫 +##彬 +##彭 +##彰 +##影 +##彷 +##役 +##彻 +##彼 +##彿 +##往 +##征 +##径 +##待 +##徇 +##很 +##徉 +##徊 +##律 +##後 +##徐 +##徑 +##徒 +##従 +##徕 +##得 +##徘 +##徙 +##徜 +##從 +##徠 +##御 +##徨 +##復 +##循 +##徬 +##微 +##徳 +##徴 +##徵 +##德 +##徹 +##徼 +##徽 +##心 +##必 +##忆 +##忌 +##忍 +##忏 +##忐 +##忑 +##忒 +##忖 +##志 +##忘 +##忙 +##応 +##忠 +##忡 +##忤 +##忧 +##忪 +##快 +##忱 +##念 +##忻 +##忽 +##忿 +##怀 +##态 +##怂 +##怅 +##怆 +##怎 +##怏 +##怒 +##怔 +##怕 +##怖 +##怙 +##怜 +##思 +##怠 +##怡 +##急 +##怦 +##性 +##怨 +##怪 +##怯 +##怵 +##总 +##怼 +##恁 +##恃 +##恆 +##恋 +##恍 +##恐 +##恒 +##恕 +##恙 +##恚 +##恢 +##恣 +##恤 +##恥 +##恨 +##恩 +##恪 +##恫 +##恬 +##恭 +##息 +##恰 +##恳 +##恵 +##恶 +##恸 +##恺 +##恻 +##恼 +##恿 +##悄 +##悅 +##悉 +##悌 +##悍 +##悔 +##悖 +##悚 +##悟 +##悠 +##患 +##悦 +##您 +##悩 +##悪 +##悬 +##悯 +##悱 +##悲 +##悴 +##悵 +##悶 +##悸 +##悻 +##悼 +##悽 +##情 +##惆 +##惇 +##惊 +##惋 +##惑 +##惕 +##惘 +##惚 +##惜 +##惟 +##惠 +##惡 +##惦 +##惧 +##惨 +##惩 +##惫 +##惬 +##惭 +##惮 +##惯 +##惰 +##惱 +##想 +##惴 +##惶 +##惹 +##惺 +##愁 +##愆 +##愈 +##愉 +##愍 +##意 +##愕 +##愚 +##愛 +##愜 +##感 +##愣 +##愤 +##愧 +##愫 +##愷 +##愿 +##慄 +##慈 +##態 +##慌 +##慎 +##慑 +##慕 +##慘 +##慚 +##慟 +##慢 +##慣 +##慧 +##慨 +##慫 +##慮 +##慰 +##慳 +##慵 +##慶 +##慷 +##慾 +##憂 +##憊 +##憋 +##憎 +##憐 +##憑 +##憔 +##憚 +##憤 +##憧 +##憨 +##憩 +##憫 +##憬 +##憲 +##憶 +##憾 +##懂 +##懇 +##懈 +##應 +##懊 +##懋 +##懑 +##懒 +##懦 +##懲 +##懵 +##懶 +##懷 +##懸 +##懺 +##懼 +##懾 +##懿 +##戀 +##戈 +##戊 +##戌 +##戍 +##戎 +##戏 +##成 +##我 +##戒 +##戕 +##或 +##战 +##戚 +##戛 +##戟 +##戡 +##戦 +##截 +##戬 +##戮 +##戰 +##戲 +##戳 +##戴 +##戶 +##户 +##戸 +##戻 +##戾 +##房 +##所 +##扁 +##扇 +##扈 +##扉 +##手 +##才 +##扎 +##扑 +##扒 +##打 +##扔 +##払 +##托 +##扛 +##扣 +##扦 +##执 +##扩 +##扪 +##扫 +##扬 +##扭 +##扮 +##扯 +##扰 +##扱 +##扳 +##扶 +##批 +##扼 +##找 +##承 +##技 +##抄 +##抉 +##把 +##抑 +##抒 +##抓 +##投 +##抖 +##抗 +##折 +##抚 +##抛 +##抜 +##択 +##抟 +##抠 +##抡 +##抢 +##护 +##报 +##抨 +##披 +##抬 +##抱 +##抵 +##抹 +##押 +##抽 +##抿 +##拂 +##拄 +##担 +##拆 +##拇 +##拈 +##拉 +##拋 +##拌 +##拍 +##拎 +##拐 +##拒 +##拓 +##拔 +##拖 +##拗 +##拘 +##拙 +##拚 +##招 +##拜 +##拟 +##拡 +##拢 +##拣 +##拥 +##拦 +##拧 +##拨 +##择 +##括 +##拭 +##拮 +##拯 +##拱 +##拳 +##拴 +##拷 +##拼 +##拽 +##拾 +##拿 +##持 +##挂 +##指 +##挈 +##按 +##挎 +##挑 +##挖 +##挙 +##挚 +##挛 +##挝 +##挞 +##挟 +##挠 +##挡 +##挣 +##挤 +##挥 +##挨 +##挪 +##挫 +##振 +##挲 +##挹 +##挺 +##挽 +##挾 +##捂 +##捅 +##捆 +##捉 +##捋 +##捌 +##捍 +##捎 +##捏 +##捐 +##捕 +##捞 +##损 +##捡 +##换 +##捣 +##捧 +##捨 +##捩 +##据 +##捱 +##捲 +##捶 +##捷 +##捺 +##捻 +##掀 +##掂 +##掃 +##掇 +##授 +##掉 +##掌 +##掏 +##掐 +##排 +##掖 +##掘 +##掙 +##掛 +##掠 +##採 +##探 +##掣 +##接 +##控 +##推 +##掩 +##措 +##掬 +##掰 +##掲 +##掳 +##掴 +##掷 +##掸 +##掺 +##揀 +##揃 +##揄 +##揆 +##揉 +##揍 +##描 +##提 +##插 +##揖 +##揚 +##換 +##握 +##揣 +##揩 +##揪 +##揭 +##揮 +##援 +##揶 +##揸 +##揹 +##揽 +##搀 +##搁 +##搂 +##搅 +##損 +##搏 +##搐 +##搓 +##搔 +##搖 +##搗 +##搜 +##搞 +##搡 +##搪 +##搬 +##搭 +##搵 +##搶 +##携 +##搽 +##摀 +##摁 +##摄 +##摆 +##摇 +##摈 +##摊 +##摒 +##摔 +##摘 +##摞 +##摟 +##摧 +##摩 +##摯 +##摳 +##摸 +##摹 +##摺 +##摻 +##撂 +##撃 +##撅 +##撇 +##撈 +##撐 +##撑 +##撒 +##撓 +##撕 +##撚 +##撞 +##撤 +##撥 +##撩 +##撫 +##撬 +##播 +##撮 +##撰 +##撲 +##撵 +##撷 +##撸 +##撻 +##撼 +##撿 +##擀 +##擁 +##擂 +##擄 +##擅 +##擇 +##擊 +##擋 +##操 +##擎 +##擒 +##擔 +##擘 +##據 +##擞 +##擠 +##擡 +##擢 +##擦 +##擬 +##擰 +##擱 +##擲 +##擴 +##擷 +##擺 +##擼 +##擾 +##攀 +##攏 +##攒 +##攔 +##攘 +##攙 +##攜 +##攝 +##攞 +##攢 +##攣 +##攤 +##攥 +##攪 +##攫 +##攬 +##支 +##收 +##攸 +##改 +##攻 +##放 +##政 +##故 +##效 +##敌 +##敍 +##敎 +##敏 +##救 +##敕 +##敖 +##敗 +##敘 +##教 +##敛 +##敝 +##敞 +##敢 +##散 +##敦 +##敬 +##数 +##敲 +##整 +##敵 +##敷 +##數 +##斂 +##斃 +##文 +##斋 +##斌 +##斎 +##斐 +##斑 +##斓 +##斗 +##料 +##斛 +##斜 +##斟 +##斡 +##斤 +##斥 +##斧 +##斩 +##斫 +##斬 +##断 +##斯 +##新 +##斷 +##方 +##於 +##施 +##旁 +##旃 +##旅 +##旋 +##旌 +##旎 +##族 +##旖 +##旗 +##无 +##既 +##日 +##旦 +##旧 +##旨 +##早 +##旬 +##旭 +##旮 +##旱 +##时 +##旷 +##旺 +##旻 +##昀 +##昂 +##昆 +##昇 +##昉 +##昊 +##昌 +##明 +##昏 +##易 +##昔 +##昕 +##昙 +##星 +##映 +##春 +##昧 +##昨 +##昭 +##是 +##昱 +##昴 +##昵 +##昶 +##昼 +##显 +##晁 +##時 +##晃 +##晉 +##晋 +##晌 +##晏 +##晒 +##晓 +##晔 +##晕 +##晖 +##晗 +##晚 +##晝 +##晞 +##晟 +##晤 +##晦 +##晨 +##晩 +##普 +##景 +##晰 +##晴 +##晶 +##晷 +##智 +##晾 +##暂 +##暄 +##暇 +##暈 +##暉 +##暌 +##暐 +##暑 +##暖 +##暗 +##暝 +##暢 +##暧 +##暨 +##暫 +##暮 +##暱 +##暴 +##暸 +##暹 +##曄 +##曆 +##曇 +##曉 +##曖 +##曙 +##曜 +##曝 +##曠 +##曦 +##曬 +##曰 +##曲 +##曳 +##更 +##書 +##曹 +##曼 +##曾 +##替 +##最 +##會 +##月 +##有 +##朋 +##服 +##朐 +##朔 +##朕 +##朗 +##望 +##朝 +##期 +##朦 +##朧 +##木 +##未 +##末 +##本 +##札 +##朮 +##术 +##朱 +##朴 +##朵 +##机 +##朽 +##杀 +##杂 +##权 +##杆 +##杈 +##杉 +##李 +##杏 +##材 +##村 +##杓 +##杖 +##杜 +##杞 +##束 +##杠 +##条 +##来 +##杨 +##杭 +##杯 +##杰 +##東 +##杳 +##杵 +##杷 +##杼 +##松 +##板 +##极 +##构 +##枇 +##枉 +##枋 +##析 +##枕 +##林 +##枚 +##果 +##枝 +##枢 +##枣 +##枪 +##枫 +##枭 +##枯 +##枰 +##枱 +##枳 +##架 +##枷 +##枸 +##柄 +##柏 +##某 +##柑 +##柒 +##染 +##柔 +##柘 +##柚 +##柜 +##柞 +##柠 +##柢 +##查 +##柩 +##柬 +##柯 +##柱 +##柳 +##柴 +##柵 +##査 +##柿 +##栀 +##栃 +##栄 +##栅 +##标 +##栈 +##栉 +##栋 +##栎 +##栏 +##树 +##栓 +##栖 +##栗 +##校 +##栩 +##株 +##样 +##核 +##根 +##格 +##栽 +##栾 +##桀 +##桁 +##桂 +##桃 +##桅 +##框 +##案 +##桉 +##桌 +##桎 +##桐 +##桑 +##桓 +##桔 +##桜 +##桠 +##桡 +##桢 +##档 +##桥 +##桦 +##桧 +##桨 +##桩 +##桶 +##桿 +##梁 +##梅 +##梆 +##梏 +##梓 +##梗 +##條 +##梟 +##梢 +##梦 +##梧 +##梨 +##梭 +##梯 +##械 +##梳 +##梵 +##梶 +##检 +##棂 +##棄 +##棉 +##棋 +##棍 +##棒 +##棕 +##棗 +##棘 +##棚 +##棟 +##棠 +##棣 +##棧 +##森 +##棱 +##棲 +##棵 +##棹 +##棺 +##椁 +##椅 +##椋 +##植 +##椎 +##椒 +##検 +##椪 +##椭 +##椰 +##椹 +##椽 +##椿 +##楂 +##楊 +##楓 +##楔 +##楚 +##楝 +##楞 +##楠 +##楣 +##楨 +##楫 +##業 +##楮 +##極 +##楷 +##楸 +##楹 +##楼 +##楽 +##概 +##榄 +##榆 +##榈 +##榉 +##榔 +##榕 +##榖 +##榛 +##榜 +##榨 +##榫 +##榭 +##榮 +##榱 +##榴 +##榷 +##榻 +##槁 +##槃 +##構 +##槌 +##槍 +##槎 +##槐 +##槓 +##様 +##槛 +##槟 +##槤 +##槭 +##槲 +##槳 +##槻 +##槽 +##槿 +##樁 +##樂 +##樊 +##樑 +##樓 +##標 +##樞 +##樟 +##模 +##樣 +##権 +##横 +##樫 +##樯 +##樱 +##樵 +##樸 +##樹 +##樺 +##樽 +##樾 +##橄 +##橇 +##橋 +##橐 +##橘 +##橙 +##機 +##橡 +##橢 +##橫 +##橱 +##橹 +##橼 +##檀 +##檄 +##檎 +##檐 +##檔 +##檗 +##檜 +##檢 +##檬 +##檯 +##檳 +##檸 +##檻 +##櫃 +##櫚 +##櫛 +##櫥 +##櫸 +##櫻 +##欄 +##權 +##欒 +##欖 +##欠 +##次 +##欢 +##欣 +##欧 +##欲 +##欸 +##欺 +##欽 +##款 +##歆 +##歇 +##歉 +##歌 +##歎 +##歐 +##歓 +##歙 +##歛 +##歡 +##止 +##正 +##此 +##步 +##武 +##歧 +##歩 +##歪 +##歯 +##歲 +##歳 +##歴 +##歷 +##歸 +##歹 +##死 +##歼 +##殁 +##殃 +##殆 +##殇 +##殉 +##殊 +##残 +##殒 +##殓 +##殖 +##殘 +##殞 +##殡 +##殤 +##殭 +##殯 +##殲 +##殴 +##段 +##殷 +##殺 +##殼 +##殿 +##毀 +##毁 +##毂 +##毅 +##毆 +##毋 +##母 +##毎 +##每 +##毒 +##毓 +##比 +##毕 +##毗 +##毘 +##毙 +##毛 +##毡 +##毫 +##毯 +##毽 +##氈 +##氏 +##氐 +##民 +##氓 +##气 +##氖 +##気 +##氙 +##氛 +##氟 +##氡 +##氢 +##氣 +##氤 +##氦 +##氧 +##氨 +##氪 +##氫 +##氮 +##氯 +##氰 +##氲 +##水 +##氷 +##永 +##氹 +##氾 +##汀 +##汁 +##求 +##汆 +##汇 +##汉 +##汎 +##汐 +##汕 +##汗 +##汙 +##汛 +##汝 +##汞 +##江 +##池 +##污 +##汤 +##汨 +##汩 +##汪 +##汰 +##汲 +##汴 +##汶 +##汹 +##決 +##汽 +##汾 +##沁 +##沂 +##沃 +##沅 +##沈 +##沉 +##沌 +##沏 +##沐 +##沒 +##沓 +##沖 +##沙 +##沛 +##沟 +##没 +##沢 +##沣 +##沥 +##沦 +##沧 +##沪 +##沫 +##沭 +##沮 +##沱 +##河 +##沸 +##油 +##治 +##沼 +##沽 +##沾 +##沿 +##況 +##泄 +##泉 +##泊 +##泌 +##泓 +##法 +##泗 +##泛 +##泞 +##泠 +##泡 +##波 +##泣 +##泥 +##注 +##泪 +##泫 +##泮 +##泯 +##泰 +##泱 +##泳 +##泵 +##泷 +##泸 +##泻 +##泼 +##泽 +##泾 +##洁 +##洄 +##洋 +##洒 +##洗 +##洙 +##洛 +##洞 +##津 +##洩 +##洪 +##洮 +##洱 +##洲 +##洵 +##洶 +##洸 +##洹 +##活 +##洼 +##洽 +##派 +##流 +##浃 +##浄 +##浅 +##浆 +##浇 +##浊 +##测 +##济 +##浏 +##浑 +##浒 +##浓 +##浔 +##浙 +##浚 +##浜 +##浣 +##浦 +##浩 +##浪 +##浬 +##浮 +##浯 +##浴 +##海 +##浸 +##涂 +##涅 +##涇 +##消 +##涉 +##涌 +##涎 +##涓 +##涔 +##涕 +##涙 +##涛 +##涝 +##涞 +##涟 +##涠 +##涡 +##涣 +##涤 +##润 +##涧 +##涨 +##涩 +##涪 +##涮 +##涯 +##液 +##涵 +##涸 +##涼 +##涿 +##淀 +##淄 +##淅 +##淆 +##淇 +##淋 +##淌 +##淑 +##淒 +##淖 +##淘 +##淙 +##淚 +##淞 +##淡 +##淤 +##淦 +##淨 +##淩 +##淪 +##淫 +##淬 +##淮 +##深 +##淳 +##淵 +##混 +##淹 +##淺 +##添 +##淼 +##清 +##済 +##渉 +##渊 +##渋 +##渍 +##渎 +##渐 +##渔 +##渗 +##渙 +##渚 +##減 +##渝 +##渠 +##渡 +##渣 +##渤 +##渥 +##渦 +##温 +##測 +##渭 +##港 +##渲 +##渴 +##游 +##渺 +##渾 +##湃 +##湄 +##湊 +##湍 +##湖 +##湘 +##湛 +##湟 +##湧 +##湫 +##湮 +##湯 +##湳 +##湾 +##湿 +##満 +##溃 +##溅 +##溉 +##溏 +##源 +##準 +##溜 +##溝 +##溟 +##溢 +##溥 +##溧 +##溪 +##溫 +##溯 +##溱 +##溴 +##溶 +##溺 +##溼 +##滁 +##滂 +##滄 +##滅 +##滇 +##滋 +##滌 +##滑 +##滓 +##滔 +##滕 +##滙 +##滚 +##滝 +##滞 +##滟 +##满 +##滢 +##滤 +##滥 +##滦 +##滨 +##滩 +##滬 +##滯 +##滲 +##滴 +##滷 +##滸 +##滾 +##滿 +##漁 +##漂 +##漆 +##漉 +##漏 +##漓 +##演 +##漕 +##漠 +##漢 +##漣 +##漩 +##漪 +##漫 +##漬 +##漯 +##漱 +##漲 +##漳 +##漸 +##漾 +##漿 +##潆 +##潇 +##潋 +##潍 +##潑 +##潔 +##潘 +##潛 +##潜 +##潞 +##潟 +##潢 +##潤 +##潦 +##潧 +##潭 +##潮 +##潰 +##潴 +##潸 +##潺 +##潼 +##澀 +##澄 +##澆 +##澈 +##澍 +##澎 +##澗 +##澜 +##澡 +##澤 +##澧 +##澱 +##澳 +##澹 +##激 +##濁 +##濂 +##濃 +##濑 +##濒 +##濕 +##濘 +##濛 +##濟 +##濠 +##濡 +##濤 +##濫 +##濬 +##濮 +##濯 +##濱 +##濺 +##濾 +##瀅 +##瀆 +##瀉 +##瀋 +##瀏 +##瀑 +##瀕 +##瀘 +##瀚 +##瀛 +##瀝 +##瀞 +##瀟 +##瀧 +##瀨 +##瀬 +##瀰 +##瀾 +##灌 +##灏 +##灑 +##灘 +##灝 +##灞 +##灣 +##火 +##灬 +##灭 +##灯 +##灰 +##灵 +##灶 +##灸 +##灼 +##災 +##灾 +##灿 +##炀 +##炁 +##炅 +##炉 +##炊 +##炎 +##炒 +##炔 +##炕 +##炖 +##炙 +##炜 +##炫 +##炬 +##炭 +##炮 +##炯 +##炳 +##炷 +##炸 +##点 +##為 +##炼 +##炽 +##烁 +##烂 +##烃 +##烈 +##烊 +##烏 +##烘 +##烙 +##烛 +##烟 +##烤 +##烦 +##烧 +##烨 +##烩 +##烫 +##烬 +##热 +##烯 +##烷 +##烹 +##烽 +##焉 +##焊 +##焕 +##焖 +##焗 +##焘 +##焙 +##焚 +##焜 +##無 +##焦 +##焯 +##焰 +##焱 +##然 +##焼 +##煅 +##煉 +##煊 +##煌 +##煎 +##煒 +##煖 +##煙 +##煜 +##煞 +##煤 +##煥 +##煦 +##照 +##煨 +##煩 +##煮 +##煲 +##煸 +##煽 +##熄 +##熊 +##熏 +##熒 +##熔 +##熙 +##熟 +##熠 +##熨 +##熬 +##熱 +##熵 +##熹 +##熾 +##燁 +##燃 +##燄 +##燈 +##燉 +##燊 +##燎 +##燒 +##燔 +##燕 +##燙 +##燜 +##營 +##燥 +##燦 +##燧 +##燭 +##燮 +##燴 +##燻 +##燼 +##燿 +##爆 +##爍 +##爐 +##爛 +##爪 +##爬 +##爭 +##爰 +##爱 +##爲 +##爵 +##父 +##爷 +##爸 +##爹 +##爺 +##爻 +##爽 +##爾 +##牆 +##片 +##版 +##牌 +##牍 +##牒 +##牙 +##牛 +##牝 +##牟 +##牠 +##牡 +##牢 +##牦 +##牧 +##物 +##牯 +##牲 +##牴 +##牵 +##特 +##牺 +##牽 +##犀 +##犁 +##犄 +##犊 +##犍 +##犒 +##犢 +##犧 +##犬 +##犯 +##状 +##犷 +##犸 +##犹 +##狀 +##狂 +##狄 +##狈 +##狎 +##狐 +##狒 +##狗 +##狙 +##狞 +##狠 +##狡 +##狩 +##独 +##狭 +##狮 +##狰 +##狱 +##狸 +##狹 +##狼 +##狽 +##猎 +##猕 +##猖 +##猗 +##猙 +##猛 +##猜 +##猝 +##猥 +##猩 +##猪 +##猫 +##猬 +##献 +##猴 +##猶 +##猷 +##猾 +##猿 +##獄 +##獅 +##獎 +##獐 +##獒 +##獗 +##獠 +##獣 +##獨 +##獭 +##獰 +##獲 +##獵 +##獷 +##獸 +##獺 +##獻 +##獼 +##獾 +##玄 +##率 +##玉 +##王 +##玑 +##玖 +##玛 +##玟 +##玠 +##玥 +##玩 +##玫 +##玮 +##环 +##现 +##玲 +##玳 +##玷 +##玺 +##玻 +##珀 +##珂 +##珅 +##珈 +##珉 +##珊 +##珍 +##珏 +##珐 +##珑 +##珙 +##珞 +##珠 +##珣 +##珥 +##珩 +##珪 +##班 +##珮 +##珲 +##珺 +##現 +##球 +##琅 +##理 +##琇 +##琉 +##琊 +##琍 +##琏 +##琐 +##琛 +##琢 +##琥 +##琦 +##琨 +##琪 +##琬 +##琮 +##琰 +##琲 +##琳 +##琴 +##琵 +##琶 +##琺 +##琼 +##瑀 +##瑁 +##瑄 +##瑋 +##瑕 +##瑗 +##瑙 +##瑚 +##瑛 +##瑜 +##瑞 +##瑟 +##瑠 +##瑣 +##瑤 +##瑩 +##瑪 +##瑯 +##瑰 +##瑶 +##瑾 +##璀 +##璁 +##璃 +##璇 +##璉 +##璋 +##璎 +##璐 +##璜 +##璞 +##璟 +##璧 +##璨 +##環 +##璽 +##璿 +##瓊 +##瓏 +##瓒 +##瓜 +##瓢 +##瓣 +##瓤 +##瓦 +##瓮 +##瓯 +##瓴 +##瓶 +##瓷 +##甄 +##甌 +##甕 +##甘 +##甙 +##甚 +##甜 +##生 +##產 +##産 +##甥 +##甦 +##用 +##甩 +##甫 +##甬 +##甭 +##甯 +##田 +##由 +##甲 +##申 +##电 +##男 +##甸 +##町 +##画 +##甾 +##畀 +##畅 +##界 +##畏 +##畑 +##畔 +##留 +##畜 +##畝 +##畢 +##略 +##畦 +##番 +##畫 +##異 +##畲 +##畳 +##畴 +##當 +##畸 +##畹 +##畿 +##疆 +##疇 +##疊 +##疏 +##疑 +##疔 +##疖 +##疗 +##疙 +##疚 +##疝 +##疟 +##疡 +##疣 +##疤 +##疥 +##疫 +##疮 +##疯 +##疱 +##疲 +##疳 +##疵 +##疸 +##疹 +##疼 +##疽 +##疾 +##痂 +##病 +##症 +##痈 +##痉 +##痊 +##痍 +##痒 +##痔 +##痕 +##痘 +##痙 +##痛 +##痞 +##痠 +##痢 +##痣 +##痤 +##痧 +##痨 +##痪 +##痫 +##痰 +##痱 +##痴 +##痹 +##痺 +##痼 +##痿 +##瘀 +##瘁 +##瘋 +##瘍 +##瘓 +##瘘 +##瘙 +##瘟 +##瘠 +##瘡 +##瘢 +##瘤 +##瘦 +##瘧 +##瘩 +##瘪 +##瘫 +##瘴 +##瘸 +##瘾 +##療 +##癇 +##癌 +##癒 +##癖 +##癜 +##癞 +##癡 +##癢 +##癣 +##癥 +##癫 +##癬 +##癮 +##癱 +##癲 +##癸 +##発 +##登 +##發 +##白 +##百 +##皂 +##的 +##皆 +##皇 +##皈 +##皋 +##皎 +##皑 +##皓 +##皖 +##皙 +##皚 +##皮 +##皰 +##皱 +##皴 +##皺 +##皿 +##盂 +##盃 +##盅 +##盆 +##盈 +##益 +##盎 +##盏 +##盐 +##监 +##盒 +##盔 +##盖 +##盗 +##盘 +##盛 +##盜 +##盞 +##盟 +##盡 +##監 +##盤 +##盥 +##盧 +##盪 +##目 +##盯 +##盱 +##盲 +##直 +##相 +##盹 +##盼 +##盾 +##省 +##眈 +##眉 +##看 +##県 +##眙 +##眞 +##真 +##眠 +##眦 +##眨 +##眩 +##眯 +##眶 +##眷 +##眸 +##眺 +##眼 +##眾 +##着 +##睁 +##睇 +##睏 +##睐 +##睑 +##睛 +##睜 +##睞 +##睡 +##睢 +##督 +##睥 +##睦 +##睨 +##睪 +##睫 +##睬 +##睹 +##睽 +##睾 +##睿 +##瞄 +##瞅 +##瞇 +##瞋 +##瞌 +##瞎 +##瞑 +##瞒 +##瞓 +##瞞 +##瞟 +##瞠 +##瞥 +##瞧 +##瞩 +##瞪 +##瞬 +##瞭 +##瞰 +##瞳 +##瞻 +##瞼 +##瞿 +##矇 +##矍 +##矗 +##矚 +##矛 +##矜 +##矢 +##矣 +##知 +##矩 +##矫 +##短 +##矮 +##矯 +##石 +##矶 +##矽 +##矾 +##矿 +##码 +##砂 +##砌 +##砍 +##砒 +##研 +##砖 +##砗 +##砚 +##砝 +##砣 +##砥 +##砧 +##砭 +##砰 +##砲 +##破 +##砷 +##砸 +##砺 +##砼 +##砾 +##础 +##硅 +##硐 +##硒 +##硕 +##硝 +##硫 +##硬 +##确 +##硯 +##硼 +##碁 +##碇 +##碉 +##碌 +##碍 +##碎 +##碑 +##碓 +##碗 +##碘 +##碚 +##碛 +##碟 +##碣 +##碧 +##碩 +##碰 +##碱 +##碳 +##碴 +##確 +##碼 +##碾 +##磁 +##磅 +##磊 +##磋 +##磐 +##磕 +##磚 +##磡 +##磨 +##磬 +##磯 +##磲 +##磷 +##磺 +##礁 +##礎 +##礙 +##礡 +##礦 +##礪 +##礫 +##礴 +##示 +##礼 +##社 +##祀 +##祁 +##祂 +##祇 +##祈 +##祉 +##祎 +##祐 +##祕 +##祖 +##祗 +##祚 +##祛 +##祜 +##祝 +##神 +##祟 +##祠 +##祢 +##祥 +##票 +##祭 +##祯 +##祷 +##祸 +##祺 +##祿 +##禀 +##禁 +##禄 +##禅 +##禍 +##禎 +##福 +##禛 +##禦 +##禧 +##禪 +##禮 +##禱 +##禹 +##禺 +##离 +##禽 +##禾 +##禿 +##秀 +##私 +##秃 +##秆 +##秉 +##秋 +##种 +##科 +##秒 +##秘 +##租 +##秣 +##秤 +##秦 +##秧 +##秩 +##秭 +##积 +##称 +##秸 +##移 +##秽 +##稀 +##稅 +##程 +##稍 +##税 +##稔 +##稗 +##稚 +##稜 +##稞 +##稟 +##稠 +##稣 +##種 +##稱 +##稲 +##稳 +##稷 +##稹 +##稻 +##稼 +##稽 +##稿 +##穀 +##穂 +##穆 +##穌 +##積 +##穎 +##穗 +##穢 +##穩 +##穫 +##穴 +##究 +##穷 +##穹 +##空 +##穿 +##突 +##窃 +##窄 +##窈 +##窍 +##窑 +##窒 +##窓 +##窕 +##窖 +##窗 +##窘 +##窜 +##窝 +##窟 +##窠 +##窥 +##窦 +##窨 +##窩 +##窪 +##窮 +##窯 +##窺 +##窿 +##竄 +##竅 +##竇 +##竊 +##立 +##竖 +##站 +##竜 +##竞 +##竟 +##章 +##竣 +##童 +##竭 +##端 +##競 +##竹 +##竺 +##竽 +##竿 +##笃 +##笆 +##笈 +##笋 +##笏 +##笑 +##笔 +##笙 +##笛 +##笞 +##笠 +##符 +##笨 +##第 +##笹 +##笺 +##笼 +##筆 +##等 +##筊 +##筋 +##筍 +##筏 +##筐 +##筑 +##筒 +##答 +##策 +##筛 +##筝 +##筠 +##筱 +##筲 +##筵 +##筷 +##筹 +##签 +##简 +##箇 +##箋 +##箍 +##箏 +##箐 +##箔 +##箕 +##算 +##箝 +##管 +##箩 +##箫 +##箭 +##箱 +##箴 +##箸 +##節 +##篁 +##範 +##篆 +##篇 +##築 +##篑 +##篓 +##篙 +##篝 +##篠 +##篡 +##篤 +##篩 +##篪 +##篮 +##篱 +##篷 +##簇 +##簌 +##簍 +##簡 +##簦 +##簧 +##簪 +##簫 +##簷 +##簸 +##簽 +##簾 +##簿 +##籁 +##籃 +##籌 +##籍 +##籐 +##籟 +##籠 +##籤 +##籬 +##籮 +##籲 +##米 +##类 +##籼 +##籽 +##粄 +##粉 +##粑 +##粒 +##粕 +##粗 +##粘 +##粟 +##粤 +##粥 +##粧 +##粪 +##粮 +##粱 +##粲 +##粳 +##粵 +##粹 +##粼 +##粽 +##精 +##粿 +##糅 +##糊 +##糍 +##糕 +##糖 +##糗 +##糙 +##糜 +##糞 +##糟 +##糠 +##糧 +##糬 +##糯 +##糰 +##糸 +##系 +##糾 +##紀 +##紂 +##約 +##紅 +##紉 +##紊 +##紋 +##納 +##紐 +##紓 +##純 +##紗 +##紘 +##紙 +##級 +##紛 +##紜 +##素 +##紡 +##索 +##紧 +##紫 +##紮 +##累 +##細 +##紳 +##紹 +##紺 +##終 +##絃 +##組 +##絆 +##経 +##結 +##絕 +##絞 +##絡 +##絢 +##給 +##絨 +##絮 +##統 +##絲 +##絳 +##絵 +##絶 +##絹 +##綁 +##綏 +##綑 +##經 +##継 +##続 +##綜 +##綠 +##綢 +##綦 +##綫 +##綬 +##維 +##綱 +##網 +##綴 +##綵 +##綸 +##綺 +##綻 +##綽 +##綾 +##綿 +##緊 +##緋 +##総 +##緑 +##緒 +##緘 +##線 +##緝 +##緞 +##締 +##緣 +##編 +##緩 +##緬 +##緯 +##練 +##緹 +##緻 +##縁 +##縄 +##縈 +##縛 +##縝 +##縣 +##縫 +##縮 +##縱 +##縴 +##縷 +##總 +##績 +##繁 +##繃 +##繆 +##繇 +##繋 +##織 +##繕 +##繚 +##繞 +##繡 +##繩 +##繪 +##繫 +##繭 +##繳 +##繹 +##繼 +##繽 +##纂 +##續 +##纍 +##纏 +##纓 +##纔 +##纖 +##纜 +##纠 +##红 +##纣 +##纤 +##约 +##级 +##纨 +##纪 +##纫 +##纬 +##纭 +##纯 +##纰 +##纱 +##纲 +##纳 +##纵 +##纶 +##纷 +##纸 +##纹 +##纺 +##纽 +##纾 +##线 +##绀 +##练 +##组 +##绅 +##细 +##织 +##终 +##绊 +##绍 +##绎 +##经 +##绑 +##绒 +##结 +##绔 +##绕 +##绘 +##给 +##绚 +##绛 +##络 +##绝 +##绞 +##统 +##绡 +##绢 +##绣 +##绥 +##绦 +##继 +##绩 +##绪 +##绫 +##续 +##绮 +##绯 +##绰 +##绳 +##维 +##绵 +##绶 +##绷 +##绸 +##绻 +##综 +##绽 +##绾 +##绿 +##缀 +##缄 +##缅 +##缆 +##缇 +##缈 +##缉 +##缎 +##缓 +##缔 +##缕 +##编 +##缘 +##缙 +##缚 +##缜 +##缝 +##缠 +##缢 +##缤 +##缥 +##缨 +##缩 +##缪 +##缭 +##缮 +##缰 +##缱 +##缴 +##缸 +##缺 +##缽 +##罂 +##罄 +##罌 +##罐 +##网 +##罔 +##罕 +##罗 +##罚 +##罡 +##罢 +##罩 +##罪 +##置 +##罰 +##署 +##罵 +##罷 +##罹 +##羁 +##羅 +##羈 +##羊 +##羌 +##美 +##羔 +##羚 +##羞 +##羟 +##羡 +##羣 +##群 +##羥 +##羧 +##羨 +##義 +##羯 +##羲 +##羸 +##羹 +##羽 +##羿 +##翁 +##翅 +##翊 +##翌 +##翎 +##習 +##翔 +##翘 +##翟 +##翠 +##翡 +##翦 +##翩 +##翰 +##翱 +##翳 +##翹 +##翻 +##翼 +##耀 +##老 +##考 +##耄 +##者 +##耆 +##耋 +##而 +##耍 +##耐 +##耒 +##耕 +##耗 +##耘 +##耙 +##耦 +##耨 +##耳 +##耶 +##耷 +##耸 +##耻 +##耽 +##耿 +##聂 +##聆 +##聊 +##聋 +##职 +##聒 +##联 +##聖 +##聘 +##聚 +##聞 +##聪 +##聯 +##聰 +##聲 +##聳 +##聴 +##聶 +##職 +##聽 +##聾 +##聿 +##肃 +##肄 +##肅 +##肆 +##肇 +##肉 +##肋 +##肌 +##肏 +##肓 +##肖 +##肘 +##肚 +##肛 +##肝 +##肠 +##股 +##肢 +##肤 +##肥 +##肩 +##肪 +##肮 +##肯 +##肱 +##育 +##肴 +##肺 +##肽 +##肾 +##肿 +##胀 +##胁 +##胃 +##胄 +##胆 +##背 +##胍 +##胎 +##胖 +##胚 +##胛 +##胜 +##胝 +##胞 +##胡 +##胤 +##胥 +##胧 +##胫 +##胭 +##胯 +##胰 +##胱 +##胳 +##胴 +##胶 +##胸 +##胺 +##能 +##脂 +##脅 +##脆 +##脇 +##脈 +##脉 +##脊 +##脍 +##脏 +##脐 +##脑 +##脓 +##脖 +##脘 +##脚 +##脛 +##脣 +##脩 +##脫 +##脯 +##脱 +##脲 +##脳 +##脸 +##脹 +##脾 +##腆 +##腈 +##腊 +##腋 +##腌 +##腎 +##腐 +##腑 +##腓 +##腔 +##腕 +##腥 +##腦 +##腩 +##腫 +##腭 +##腮 +##腰 +##腱 +##腳 +##腴 +##腸 +##腹 +##腺 +##腻 +##腼 +##腾 +##腿 +##膀 +##膈 +##膊 +##膏 +##膑 +##膘 +##膚 +##膛 +##膜 +##膝 +##膠 +##膦 +##膨 +##膩 +##膳 +##膺 +##膻 +##膽 +##膾 +##膿 +##臀 +##臂 +##臃 +##臆 +##臉 +##臊 +##臍 +##臓 +##臘 +##臟 +##臣 +##臥 +##臧 +##臨 +##自 +##臬 +##臭 +##至 +##致 +##臺 +##臻 +##臼 +##臾 +##舀 +##舂 +##舅 +##舆 +##與 +##興 +##舉 +##舊 +##舌 +##舍 +##舎 +##舐 +##舒 +##舔 +##舖 +##舗 +##舛 +##舜 +##舞 +##舟 +##航 +##舫 +##般 +##舰 +##舱 +##舵 +##舶 +##舷 +##舸 +##船 +##舺 +##舾 +##艇 +##艋 +##艘 +##艙 +##艦 +##艮 +##良 +##艰 +##艱 +##色 +##艳 +##艷 +##艹 +##艺 +##艾 +##节 +##芃 +##芈 +##芊 +##芋 +##芍 +##芎 +##芒 +##芙 +##芜 +##芝 +##芡 +##芥 +##芦 +##芩 +##芪 +##芫 +##芬 +##芭 +##芮 +##芯 +##花 +##芳 +##芷 +##芸 +##芹 +##芻 +##芽 +##芾 +##苁 +##苄 +##苇 +##苋 +##苍 +##苏 +##苑 +##苒 +##苓 +##苔 +##苕 +##苗 +##苛 +##苜 +##苞 +##苟 +##苡 +##苣 +##若 +##苦 +##苫 +##苯 +##英 +##苷 +##苹 +##苻 +##茁 +##茂 +##范 +##茄 +##茅 +##茉 +##茎 +##茏 +##茗 +##茜 +##茧 +##茨 +##茫 +##茬 +##茭 +##茯 +##茱 +##茲 +##茴 +##茵 +##茶 +##茸 +##茹 +##茼 +##荀 +##荃 +##荆 +##草 +##荊 +##荏 +##荐 +##荒 +##荔 +##荖 +##荘 +##荚 +##荞 +##荟 +##荠 +##荡 +##荣 +##荤 +##荥 +##荧 +##荨 +##荪 +##荫 +##药 +##荳 +##荷 +##荸 +##荻 +##荼 +##荽 +##莅 +##莆 +##莉 +##莊 +##莎 +##莒 +##莓 +##莖 +##莘 +##莞 +##莠 +##莢 +##莧 +##莪 +##莫 +##莱 +##莲 +##莴 +##获 +##莹 +##莺 +##莽 +##莿 +##菀 +##菁 +##菅 +##菇 +##菈 +##菊 +##菌 +##菏 +##菓 +##菖 +##菘 +##菜 +##菟 +##菠 +##菡 +##菩 +##華 +##菱 +##菲 +##菸 +##菽 +##萁 +##萃 +##萄 +##萊 +##萋 +##萌 +##萍 +##萎 +##萘 +##萝 +##萤 +##营 +##萦 +##萧 +##萨 +##萩 +##萬 +##萱 +##萵 +##萸 +##萼 +##落 +##葆 +##葉 +##著 +##葚 +##葛 +##葡 +##董 +##葦 +##葩 +##葫 +##葬 +##葭 +##葯 +##葱 +##葳 +##葵 +##葷 +##葺 +##蒂 +##蒋 +##蒐 +##蒔 +##蒙 +##蒜 +##蒞 +##蒟 +##蒡 +##蒨 +##蒲 +##蒸 +##蒹 +##蒻 +##蒼 +##蒿 +##蓁 +##蓄 +##蓆 +##蓉 +##蓋 +##蓑 +##蓓 +##蓖 +##蓝 +##蓟 +##蓦 +##蓬 +##蓮 +##蓼 +##蓿 +##蔑 +##蔓 +##蔔 +##蔗 +##蔘 +##蔚 +##蔡 +##蔣 +##蔥 +##蔫 +##蔬 +##蔭 +##蔵 +##蔷 +##蔺 +##蔻 +##蔼 +##蔽 +##蕁 +##蕃 +##蕈 +##蕉 +##蕊 +##蕎 +##蕙 +##蕤 +##蕨 +##蕩 +##蕪 +##蕭 +##蕲 +##蕴 +##蕻 +##蕾 +##薄 +##薅 +##薇 +##薈 +##薊 +##薏 +##薑 +##薔 +##薙 +##薛 +##薦 +##薨 +##薩 +##薪 +##薬 +##薯 +##薰 +##薹 +##藉 +##藍 +##藏 +##藐 +##藓 +##藕 +##藜 +##藝 +##藤 +##藥 +##藩 +##藹 +##藻 +##藿 +##蘆 +##蘇 +##蘊 +##蘋 +##蘑 +##蘚 +##蘭 +##蘸 +##蘼 +##蘿 +##虎 +##虏 +##虐 +##虑 +##虔 +##處 +##虚 +##虛 +##虜 +##虞 +##號 +##虢 +##虧 +##虫 +##虬 +##虱 +##虹 +##虻 +##虽 +##虾 +##蚀 +##蚁 +##蚂 +##蚊 +##蚌 +##蚓 +##蚕 +##蚜 +##蚝 +##蚣 +##蚤 +##蚩 +##蚪 +##蚯 +##蚱 +##蚵 +##蛀 +##蛆 +##蛇 +##蛊 +##蛋 +##蛎 +##蛐 +##蛔 +##蛙 +##蛛 +##蛟 +##蛤 +##蛭 +##蛮 +##蛰 +##蛳 +##蛹 +##蛻 +##蛾 +##蜀 +##蜂 +##蜃 +##蜆 +##蜇 +##蜈 +##蜊 +##蜍 +##蜒 +##蜓 +##蜕 +##蜗 +##蜘 +##蜚 +##蜜 +##蜡 +##蜢 +##蜥 +##蜱 +##蜴 +##蜷 +##蜻 +##蜿 +##蝇 +##蝈 +##蝉 +##蝌 +##蝎 +##蝕 +##蝗 +##蝙 +##蝟 +##蝠 +##蝦 +##蝨 +##蝴 +##蝶 +##蝸 +##蝼 +##螂 +##螃 +##融 +##螞 +##螢 +##螨 +##螯 +##螳 +##螺 +##蟀 +##蟄 +##蟆 +##蟋 +##蟎 +##蟑 +##蟒 +##蟠 +##蟬 +##蟲 +##蟹 +##蟻 +##蟾 +##蠅 +##蠍 +##蠔 +##蠕 +##蠛 +##蠟 +##蠡 +##蠢 +##蠣 +##蠱 +##蠶 +##蠹 +##蠻 +##血 +##衄 +##衅 +##衆 +##行 +##衍 +##術 +##衔 +##街 +##衙 +##衛 +##衝 +##衞 +##衡 +##衢 +##衣 +##补 +##表 +##衩 +##衫 +##衬 +##衮 +##衰 +##衲 +##衷 +##衹 +##衾 +##衿 +##袁 +##袂 +##袄 +##袅 +##袈 +##袋 +##袍 +##袒 +##袖 +##袜 +##袞 +##袤 +##袪 +##被 +##袭 +##袱 +##裁 +##裂 +##装 +##裆 +##裊 +##裏 +##裔 +##裕 +##裘 +##裙 +##補 +##裝 +##裟 +##裡 +##裤 +##裨 +##裱 +##裳 +##裴 +##裸 +##裹 +##製 +##裾 +##褂 +##複 +##褐 +##褒 +##褓 +##褔 +##褚 +##褥 +##褪 +##褫 +##褲 +##褶 +##褻 +##襁 +##襄 +##襟 +##襠 +##襪 +##襬 +##襯 +##襲 +##西 +##要 +##覃 +##覆 +##覇 +##見 +##規 +##覓 +##視 +##覚 +##覦 +##覧 +##親 +##覬 +##観 +##覷 +##覺 +##覽 +##觀 +##见 +##观 +##规 +##觅 +##视 +##览 +##觉 +##觊 +##觎 +##觐 +##觑 +##角 +##觞 +##解 +##觥 +##触 +##觸 +##言 +##訂 +##計 +##訊 +##討 +##訓 +##訕 +##訖 +##託 +##記 +##訛 +##訝 +##訟 +##訣 +##訥 +##訪 +##設 +##許 +##訳 +##訴 +##訶 +##診 +##註 +##証 +##詆 +##詐 +##詔 +##評 +##詛 +##詞 +##詠 +##詡 +##詢 +##詣 +##試 +##詩 +##詫 +##詬 +##詭 +##詮 +##詰 +##話 +##該 +##詳 +##詹 +##詼 +##誅 +##誇 +##誉 +##誌 +##認 +##誓 +##誕 +##誘 +##語 +##誠 +##誡 +##誣 +##誤 +##誥 +##誦 +##誨 +##說 +##説 +##読 +##誰 +##課 +##誹 +##誼 +##調 +##諄 +##談 +##請 +##諏 +##諒 +##論 +##諗 +##諜 +##諡 +##諦 +##諧 +##諫 +##諭 +##諮 +##諱 +##諳 +##諷 +##諸 +##諺 +##諾 +##謀 +##謁 +##謂 +##謄 +##謊 +##謎 +##謐 +##謔 +##謗 +##謙 +##講 +##謝 +##謠 +##謨 +##謬 +##謹 +##謾 +##譁 +##證 +##譎 +##譏 +##識 +##譙 +##譚 +##譜 +##警 +##譬 +##譯 +##議 +##譲 +##譴 +##護 +##譽 +##讀 +##變 +##讓 +##讚 +##讞 +##计 +##订 +##认 +##讥 +##讧 +##讨 +##让 +##讪 +##讫 +##训 +##议 +##讯 +##记 +##讲 +##讳 +##讴 +##讶 +##讷 +##许 +##讹 +##论 +##讼 +##讽 +##设 +##访 +##诀 +##证 +##诃 +##评 +##诅 +##识 +##诈 +##诉 +##诊 +##诋 +##词 +##诏 +##译 +##试 +##诗 +##诘 +##诙 +##诚 +##诛 +##话 +##诞 +##诟 +##诠 +##诡 +##询 +##诣 +##诤 +##该 +##详 +##诧 +##诩 +##诫 +##诬 +##语 +##误 +##诰 +##诱 +##诲 +##说 +##诵 +##诶 +##请 +##诸 +##诺 +##读 +##诽 +##课 +##诿 +##谀 +##谁 +##调 +##谄 +##谅 +##谆 +##谈 +##谊 +##谋 +##谌 +##谍 +##谎 +##谏 +##谐 +##谑 +##谒 +##谓 +##谔 +##谕 +##谗 +##谘 +##谙 +##谚 +##谛 +##谜 +##谟 +##谢 +##谣 +##谤 +##谥 +##谦 +##谧 +##谨 +##谩 +##谪 +##谬 +##谭 +##谯 +##谱 +##谲 +##谴 +##谶 +##谷 +##豁 +##豆 +##豇 +##豈 +##豉 +##豊 +##豌 +##豎 +##豐 +##豔 +##豚 +##象 +##豢 +##豪 +##豫 +##豬 +##豹 +##豺 +##貂 +##貅 +##貌 +##貓 +##貔 +##貘 +##貝 +##貞 +##負 +##財 +##貢 +##貧 +##貨 +##販 +##貪 +##貫 +##責 +##貯 +##貰 +##貳 +##貴 +##貶 +##買 +##貸 +##費 +##貼 +##貽 +##貿 +##賀 +##賁 +##賂 +##賃 +##賄 +##資 +##賈 +##賊 +##賑 +##賓 +##賜 +##賞 +##賠 +##賡 +##賢 +##賣 +##賤 +##賦 +##質 +##賬 +##賭 +##賴 +##賺 +##購 +##賽 +##贅 +##贈 +##贊 +##贍 +##贏 +##贓 +##贖 +##贛 +##贝 +##贞 +##负 +##贡 +##财 +##责 +##贤 +##败 +##账 +##货 +##质 +##贩 +##贪 +##贫 +##贬 +##购 +##贮 +##贯 +##贰 +##贱 +##贲 +##贴 +##贵 +##贷 +##贸 +##费 +##贺 +##贻 +##贼 +##贾 +##贿 +##赁 +##赂 +##赃 +##资 +##赅 +##赈 +##赊 +##赋 +##赌 +##赎 +##赏 +##赐 +##赓 +##赔 +##赖 +##赘 +##赚 +##赛 +##赝 +##赞 +##赠 +##赡 +##赢 +##赣 +##赤 +##赦 +##赧 +##赫 +##赭 +##走 +##赳 +##赴 +##赵 +##赶 +##起 +##趁 +##超 +##越 +##趋 +##趕 +##趙 +##趟 +##趣 +##趨 +##足 +##趴 +##趵 +##趸 +##趺 +##趾 +##跃 +##跄 +##跆 +##跋 +##跌 +##跎 +##跑 +##跖 +##跚 +##跛 +##距 +##跟 +##跡 +##跤 +##跨 +##跩 +##跪 +##路 +##跳 +##践 +##跷 +##跹 +##跺 +##跻 +##踉 +##踊 +##踌 +##踏 +##踐 +##踝 +##踞 +##踟 +##踢 +##踩 +##踪 +##踮 +##踱 +##踴 +##踵 +##踹 +##蹂 +##蹄 +##蹇 +##蹈 +##蹉 +##蹊 +##蹋 +##蹑 +##蹒 +##蹙 +##蹟 +##蹣 +##蹤 +##蹦 +##蹩 +##蹬 +##蹭 +##蹲 +##蹴 +##蹶 +##蹺 +##蹼 +##蹿 +##躁 +##躇 +##躉 +##躊 +##躋 +##躍 +##躏 +##躪 +##身 +##躬 +##躯 +##躲 +##躺 +##軀 +##車 +##軋 +##軌 +##軍 +##軒 +##軟 +##転 +##軸 +##軼 +##軽 +##軾 +##較 +##載 +##輒 +##輓 +##輔 +##輕 +##輛 +##輝 +##輟 +##輩 +##輪 +##輯 +##輸 +##輻 +##輾 +##輿 +##轄 +##轅 +##轆 +##轉 +##轍 +##轎 +##轟 +##车 +##轧 +##轨 +##轩 +##转 +##轭 +##轮 +##软 +##轰 +##轲 +##轴 +##轶 +##轻 +##轼 +##载 +##轿 +##较 +##辄 +##辅 +##辆 +##辇 +##辈 +##辉 +##辊 +##辍 +##辐 +##辑 +##输 +##辕 +##辖 +##辗 +##辘 +##辙 +##辛 +##辜 +##辞 +##辟 +##辣 +##辦 +##辨 +##辩 +##辫 +##辭 +##辮 +##辯 +##辰 +##辱 +##農 +##边 +##辺 +##辻 +##込 +##辽 +##达 +##迁 +##迂 +##迄 +##迅 +##过 +##迈 +##迎 +##运 +##近 +##返 +##还 +##这 +##进 +##远 +##违 +##连 +##迟 +##迢 +##迤 +##迥 +##迦 +##迩 +##迪 +##迫 +##迭 +##述 +##迴 +##迷 +##迸 +##迹 +##迺 +##追 +##退 +##送 +##适 +##逃 +##逅 +##逆 +##选 +##逊 +##逍 +##透 +##逐 +##递 +##途 +##逕 +##逗 +##這 +##通 +##逛 +##逝 +##逞 +##速 +##造 +##逢 +##連 +##逮 +##週 +##進 +##逵 +##逶 +##逸 +##逻 +##逼 +##逾 +##遁 +##遂 +##遅 +##遇 +##遊 +##運 +##遍 +##過 +##遏 +##遐 +##遑 +##遒 +##道 +##達 +##違 +##遗 +##遙 +##遛 +##遜 +##遞 +##遠 +##遢 +##遣 +##遥 +##遨 +##適 +##遭 +##遮 +##遲 +##遴 +##遵 +##遶 +##遷 +##選 +##遺 +##遼 +##遽 +##避 +##邀 +##邁 +##邂 +##邃 +##還 +##邇 +##邈 +##邊 +##邋 +##邏 +##邑 +##邓 +##邕 +##邛 +##邝 +##邢 +##那 +##邦 +##邨 +##邪 +##邬 +##邮 +##邯 +##邰 +##邱 +##邳 +##邵 +##邸 +##邹 +##邺 +##邻 +##郁 +##郅 +##郊 +##郎 +##郑 +##郜 +##郝 +##郡 +##郢 +##郤 +##郦 +##郧 +##部 +##郫 +##郭 +##郴 +##郵 +##郷 +##郸 +##都 +##鄂 +##鄉 +##鄒 +##鄔 +##鄙 +##鄞 +##鄢 +##鄧 +##鄭 +##鄰 +##鄱 +##鄲 +##鄺 +##酉 +##酊 +##酋 +##酌 +##配 +##酐 +##酒 +##酗 +##酚 +##酝 +##酢 +##酣 +##酥 +##酩 +##酪 +##酬 +##酮 +##酯 +##酰 +##酱 +##酵 +##酶 +##酷 +##酸 +##酿 +##醃 +##醇 +##醉 +##醋 +##醍 +##醐 +##醒 +##醚 +##醛 +##醜 +##醞 +##醣 +##醪 +##醫 +##醬 +##醮 +##醯 +##醴 +##醺 +##釀 +##釁 +##采 +##釉 +##释 +##釋 +##里 +##重 +##野 +##量 +##釐 +##金 +##釗 +##釘 +##釜 +##針 +##釣 +##釦 +##釧 +##釵 +##鈀 +##鈉 +##鈍 +##鈎 +##鈔 +##鈕 +##鈞 +##鈣 +##鈦 +##鈪 +##鈴 +##鈺 +##鈾 +##鉀 +##鉄 +##鉅 +##鉉 +##鉑 +##鉗 +##鉚 +##鉛 +##鉤 +##鉴 +##鉻 +##銀 +##銃 +##銅 +##銑 +##銓 +##銖 +##銘 +##銜 +##銬 +##銭 +##銮 +##銳 +##銷 +##銹 +##鋁 +##鋅 +##鋒 +##鋤 +##鋪 +##鋰 +##鋸 +##鋼 +##錄 +##錐 +##錘 +##錚 +##錠 +##錢 +##錦 +##錨 +##錫 +##錮 +##錯 +##録 +##錳 +##錶 +##鍊 +##鍋 +##鍍 +##鍛 +##鍥 +##鍰 +##鍵 +##鍺 +##鍾 +##鎂 +##鎊 +##鎌 +##鎏 +##鎔 +##鎖 +##鎗 +##鎚 +##鎧 +##鎬 +##鎮 +##鎳 +##鏈 +##鏖 +##鏗 +##鏘 +##鏞 +##鏟 +##鏡 +##鏢 +##鏤 +##鏽 +##鐘 +##鐮 +##鐲 +##鐳 +##鐵 +##鐸 +##鐺 +##鑄 +##鑊 +##鑑 +##鑒 +##鑣 +##鑫 +##鑰 +##鑲 +##鑼 +##鑽 +##鑾 +##鑿 +##针 +##钉 +##钊 +##钎 +##钏 +##钒 +##钓 +##钗 +##钙 +##钛 +##钜 +##钝 +##钞 +##钟 +##钠 +##钡 +##钢 +##钣 +##钤 +##钥 +##钦 +##钧 +##钨 +##钩 +##钮 +##钯 +##钰 +##钱 +##钳 +##钴 +##钵 +##钺 +##钻 +##钼 +##钾 +##钿 +##铀 +##铁 +##铂 +##铃 +##铄 +##铅 +##铆 +##铉 +##铎 +##铐 +##铛 +##铜 +##铝 +##铠 +##铡 +##铢 +##铣 +##铤 +##铨 +##铩 +##铬 +##铭 +##铮 +##铰 +##铲 +##铵 +##银 +##铸 +##铺 +##链 +##铿 +##销 +##锁 +##锂 +##锄 +##锅 +##锆 +##锈 +##锉 +##锋 +##锌 +##锏 +##锐 +##锑 +##错 +##锚 +##锟 +##锡 +##锢 +##锣 +##锤 +##锥 +##锦 +##锭 +##键 +##锯 +##锰 +##锲 +##锵 +##锹 +##锺 +##锻 +##镀 +##镁 +##镂 +##镇 +##镉 +##镌 +##镍 +##镐 +##镑 +##镕 +##镖 +##镗 +##镛 +##镜 +##镣 +##镭 +##镯 +##镰 +##镳 +##镶 +##長 +##长 +##門 +##閃 +##閉 +##開 +##閎 +##閏 +##閑 +##閒 +##間 +##閔 +##閘 +##閡 +##関 +##閣 +##閥 +##閨 +##閩 +##閱 +##閲 +##閹 +##閻 +##閾 +##闆 +##闇 +##闊 +##闌 +##闍 +##闔 +##闕 +##闖 +##闘 +##關 +##闡 +##闢 +##门 +##闪 +##闫 +##闭 +##问 +##闯 +##闰 +##闲 +##间 +##闵 +##闷 +##闸 +##闹 +##闺 +##闻 +##闽 +##闾 +##阀 +##阁 +##阂 +##阅 +##阆 +##阇 +##阈 +##阉 +##阎 +##阐 +##阑 +##阔 +##阕 +##阖 +##阙 +##阚 +##阜 +##队 +##阡 +##阪 +##阮 +##阱 +##防 +##阳 +##阴 +##阵 +##阶 +##阻 +##阿 +##陀 +##陂 +##附 +##际 +##陆 +##陇 +##陈 +##陋 +##陌 +##降 +##限 +##陕 +##陛 +##陝 +##陞 +##陟 +##陡 +##院 +##陣 +##除 +##陨 +##险 +##陪 +##陰 +##陲 +##陳 +##陵 +##陶 +##陷 +##陸 +##険 +##陽 +##隅 +##隆 +##隈 +##隊 +##隋 +##隍 +##階 +##随 +##隐 +##隔 +##隕 +##隘 +##隙 +##際 +##障 +##隠 +##隣 +##隧 +##隨 +##險 +##隱 +##隴 +##隶 +##隸 +##隻 +##隼 +##隽 +##难 +##雀 +##雁 +##雄 +##雅 +##集 +##雇 +##雉 +##雋 +##雌 +##雍 +##雎 +##雏 +##雑 +##雒 +##雕 +##雖 +##雙 +##雛 +##雜 +##雞 +##離 +##難 +##雨 +##雪 +##雯 +##雰 +##雲 +##雳 +##零 +##雷 +##雹 +##電 +##雾 +##需 +##霁 +##霄 +##霆 +##震 +##霈 +##霉 +##霊 +##霍 +##霎 +##霏 +##霑 +##霓 +##霖 +##霜 +##霞 +##霧 +##霭 +##霰 +##露 +##霸 +##霹 +##霽 +##霾 +##靂 +##靄 +##靈 +##青 +##靓 +##靖 +##静 +##靚 +##靛 +##靜 +##非 +##靠 +##靡 +##面 +##靥 +##靦 +##革 +##靳 +##靴 +##靶 +##靼 +##鞅 +##鞋 +##鞍 +##鞏 +##鞑 +##鞘 +##鞠 +##鞣 +##鞦 +##鞭 +##韆 +##韋 +##韌 +##韓 +##韜 +##韦 +##韧 +##韩 +##韬 +##韭 +##音 +##韵 +##韶 +##韻 +##響 +##頁 +##頂 +##頃 +##項 +##順 +##須 +##頌 +##預 +##頑 +##頒 +##頓 +##頗 +##領 +##頜 +##頡 +##頤 +##頫 +##頭 +##頰 +##頷 +##頸 +##頹 +##頻 +##頼 +##顆 +##題 +##額 +##顎 +##顏 +##顔 +##願 +##顛 +##類 +##顧 +##顫 +##顯 +##顱 +##顴 +##页 +##顶 +##顷 +##项 +##顺 +##须 +##顼 +##顽 +##顾 +##顿 +##颁 +##颂 +##预 +##颅 +##领 +##颇 +##颈 +##颉 +##颊 +##颌 +##颍 +##颐 +##频 +##颓 +##颔 +##颖 +##颗 +##题 +##颚 +##颛 +##颜 +##额 +##颞 +##颠 +##颡 +##颢 +##颤 +##颦 +##颧 +##風 +##颯 +##颱 +##颳 +##颶 +##颼 +##飄 +##飆 +##风 +##飒 +##飓 +##飕 +##飘 +##飙 +##飚 +##飛 +##飞 +##食 +##飢 +##飨 +##飩 +##飪 +##飯 +##飲 +##飼 +##飽 +##飾 +##餃 +##餅 +##餉 +##養 +##餌 +##餐 +##餒 +##餓 +##餘 +##餚 +##餛 +##餞 +##餡 +##館 +##餮 +##餵 +##餾 +##饅 +##饈 +##饋 +##饌 +##饍 +##饑 +##饒 +##饕 +##饗 +##饞 +##饥 +##饨 +##饪 +##饬 +##饭 +##饮 +##饯 +##饰 +##饱 +##饲 +##饴 +##饵 +##饶 +##饷 +##饺 +##饼 +##饽 +##饿 +##馀 +##馁 +##馄 +##馅 +##馆 +##馈 +##馋 +##馍 +##馏 +##馒 +##馔 +##首 +##馗 +##香 +##馥 +##馨 +##馬 +##馭 +##馮 +##馳 +##馴 +##駁 +##駄 +##駅 +##駆 +##駐 +##駒 +##駕 +##駛 +##駝 +##駭 +##駱 +##駿 +##騁 +##騎 +##騏 +##験 +##騙 +##騨 +##騰 +##騷 +##驀 +##驅 +##驊 +##驍 +##驒 +##驕 +##驗 +##驚 +##驛 +##驟 +##驢 +##驥 +##马 +##驭 +##驮 +##驯 +##驰 +##驱 +##驳 +##驴 +##驶 +##驷 +##驸 +##驹 +##驻 +##驼 +##驾 +##驿 +##骁 +##骂 +##骄 +##骅 +##骆 +##骇 +##骈 +##骊 +##骋 +##验 +##骏 +##骐 +##骑 +##骗 +##骚 +##骛 +##骜 +##骞 +##骠 +##骡 +##骤 +##骥 +##骧 +##骨 +##骯 +##骰 +##骶 +##骷 +##骸 +##骼 +##髂 +##髅 +##髋 +##髏 +##髒 +##髓 +##體 +##髖 +##高 +##髦 +##髪 +##髮 +##髯 +##髻 +##鬃 +##鬆 +##鬍 +##鬓 +##鬚 +##鬟 +##鬢 +##鬣 +##鬥 +##鬧 +##鬱 +##鬼 +##魁 +##魂 +##魄 +##魅 +##魇 +##魍 +##魏 +##魔 +##魘 +##魚 +##魯 +##魷 +##鮑 +##鮨 +##鮪 +##鮭 +##鮮 +##鯉 +##鯊 +##鯖 +##鯛 +##鯨 +##鯰 +##鯽 +##鰍 +##鰓 +##鰭 +##鰲 +##鰻 +##鰾 +##鱈 +##鱉 +##鱔 +##鱗 +##鱷 +##鱸 +##鱼 +##鱿 +##鲁 +##鲈 +##鲍 +##鲑 +##鲛 +##鲜 +##鲟 +##鲢 +##鲤 +##鲨 +##鲫 +##鲱 +##鲲 +##鲶 +##鲷 +##鲸 +##鳃 +##鳄 +##鳅 +##鳌 +##鳍 +##鳕 +##鳖 +##鳗 +##鳝 +##鳞 +##鳥 +##鳩 +##鳳 +##鳴 +##鳶 +##鴉 +##鴕 +##鴛 +##鴦 +##鴨 +##鴻 +##鴿 +##鵑 +##鵜 +##鵝 +##鵡 +##鵬 +##鵰 +##鵲 +##鶘 +##鶩 +##鶯 +##鶴 +##鷗 +##鷲 +##鷹 +##鷺 +##鸚 +##鸞 +##鸟 +##鸠 +##鸡 +##鸢 +##鸣 +##鸥 +##鸦 +##鸨 +##鸪 +##鸭 +##鸯 +##鸳 +##鸵 +##鸽 +##鸾 +##鸿 +##鹂 +##鹃 +##鹄 +##鹅 +##鹈 +##鹉 +##鹊 +##鹌 +##鹏 +##鹑 +##鹕 +##鹘 +##鹜 +##鹞 +##鹤 +##鹦 +##鹧 +##鹫 +##鹭 +##鹰 +##鹳 +##鹵 +##鹹 +##鹼 +##鹽 +##鹿 +##麂 +##麋 +##麒 +##麓 +##麗 +##麝 +##麟 +##麥 +##麦 +##麩 +##麴 +##麵 +##麸 +##麺 +##麻 +##麼 +##麽 +##麾 +##黃 +##黄 +##黍 +##黎 +##黏 +##黑 +##黒 +##黔 +##默 +##黛 +##黜 +##黝 +##點 +##黠 +##黨 +##黯 +##黴 +##鼋 +##鼎 +##鼐 +##鼓 +##鼠 +##鼬 +##鼹 +##鼻 +##鼾 +##齁 +##齊 +##齋 +##齐 +##齒 +##齡 +##齢 +##齣 +##齦 +##齿 +##龄 +##龅 +##龈 +##龊 +##龋 +##龌 +##龍 +##龐 +##龔 +##龕 +##龙 +##龚 +##龛 +##龜 +##龟 +##︰ +##︱ +##︶ +##︿ +##﹁ +##﹂ +##﹍ +##﹏ +##﹐ +##﹑ +##﹒ +##﹔ +##﹕ +##﹖ +##﹗ +##﹙ +##﹚ +##﹝ +##﹞ +##﹡ +##﹣ +##！ +##＂ +##＃ +##＄ +##％ +##＆ +##＇ +##（ +##） +##＊ +##， +##－ +##． +##／ +##： +##； +##＜ +##？ +##＠ +##［ +##＼ +##］ +##＾ +##＿ +##｀ +##ｆ +##ｈ +##ｊ +##ｕ +##ｗ +##ｚ +##｛ +##｝ +##｡ +##｢ +##｣ +##､ +##･ +##ｯ +##ｰ +##ｲ +##ｸ +##ｼ +##ｽ +##ﾄ +##ﾉ +##ﾌ +##ﾗ +##ﾙ +##ﾝ +##ﾞ +##ﾟ +##￣ +##￥ +##👍 +##🔥 +##😂 +##😎 diff --git a/create_pretraining_data.py b/create_pretraining_data.py new file mode 100644 index 0000000..5340d96 --- /dev/null +++ b/create_pretraining_data.py @@ -0,0 +1,469 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Create masked LM/next sentence masked_lm TF examples for BERT.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import collections +import random +import tokenization +import tensorflow as tf + +flags = tf.flags + +FLAGS = flags.FLAGS + +flags.DEFINE_string("input_file", None, + "Input raw text file (or comma-separated list of files).") + +flags.DEFINE_string( + "output_file", None, + "Output TF example file (or comma-separated list of files).") + +flags.DEFINE_string("vocab_file", None, + "The vocabulary file that the BERT model was trained on.") + +flags.DEFINE_bool( + "do_lower_case", True, + "Whether to lower case the input text. Should be True for uncased " + "models and False for cased models.") + +flags.DEFINE_bool( + "do_whole_word_mask", False, + "Whether to use whole word masking rather than per-WordPiece masking.") + +flags.DEFINE_integer("max_seq_length", 128, "Maximum sequence length.") + +flags.DEFINE_integer("max_predictions_per_seq", 20, + "Maximum number of masked LM predictions per sequence.") + +flags.DEFINE_integer("random_seed", 12345, "Random seed for data generation.") + +flags.DEFINE_integer( + "dupe_factor", 10, + "Number of times to duplicate the input data (with different masks).") + +flags.DEFINE_float("masked_lm_prob", 0.15, "Masked LM probability.") + +flags.DEFINE_float( + "short_seq_prob", 0.1, + "Probability of creating sequences which are shorter than the " + "maximum length.") + + +class TrainingInstance(object): + """A single training instance (sentence pair).""" + + def __init__(self, tokens, segment_ids, masked_lm_positions, masked_lm_labels, + is_random_next): + self.tokens = tokens + self.segment_ids = segment_ids + self.is_random_next = is_random_next + self.masked_lm_positions = masked_lm_positions + self.masked_lm_labels = masked_lm_labels + + def __str__(self): + s = "" + s += "tokens: %s\n" % (" ".join( + [tokenization.printable_text(x) for x in self.tokens])) + s += "segment_ids: %s\n" % (" ".join([str(x) for x in self.segment_ids])) + s += "is_random_next: %s\n" % self.is_random_next + s += "masked_lm_positions: %s\n" % (" ".join( + [str(x) for x in self.masked_lm_positions])) + s += "masked_lm_labels: %s\n" % (" ".join( + [tokenization.printable_text(x) for x in self.masked_lm_labels])) + s += "\n" + return s + + def __repr__(self): + return self.__str__() + + +def write_instance_to_example_files(instances, tokenizer, max_seq_length, + max_predictions_per_seq, output_files): + """Create TF example files from `TrainingInstance`s.""" + writers = [] + for output_file in output_files: + writers.append(tf.python_io.TFRecordWriter(output_file)) + + writer_index = 0 + + total_written = 0 + for (inst_index, instance) in enumerate(instances): + input_ids = tokenizer.convert_tokens_to_ids(instance.tokens) + input_mask = [1] * len(input_ids) + segment_ids = list(instance.segment_ids) + assert len(input_ids) <= max_seq_length + + while len(input_ids) < max_seq_length: + input_ids.append(0) + input_mask.append(0) + segment_ids.append(0) + + assert len(input_ids) == max_seq_length + assert len(input_mask) == max_seq_length + assert len(segment_ids) == max_seq_length + + masked_lm_positions = list(instance.masked_lm_positions) + masked_lm_ids = tokenizer.convert_tokens_to_ids(instance.masked_lm_labels) + masked_lm_weights = [1.0] * len(masked_lm_ids) + + while len(masked_lm_positions) < max_predictions_per_seq: + masked_lm_positions.append(0) + masked_lm_ids.append(0) + masked_lm_weights.append(0.0) + + next_sentence_label = 1 if instance.is_random_next else 0 + + features = collections.OrderedDict() + features["input_ids"] = create_int_feature(input_ids) + features["input_mask"] = create_int_feature(input_mask) + features["segment_ids"] = create_int_feature(segment_ids) + features["masked_lm_positions"] = create_int_feature(masked_lm_positions) + features["masked_lm_ids"] = create_int_feature(masked_lm_ids) + features["masked_lm_weights"] = create_float_feature(masked_lm_weights) + features["next_sentence_labels"] = create_int_feature([next_sentence_label]) + + tf_example = tf.train.Example(features=tf.train.Features(feature=features)) + + writers[writer_index].write(tf_example.SerializeToString()) + writer_index = (writer_index + 1) % len(writers) + + total_written += 1 + + if inst_index < 20: + tf.logging.info("*** Example ***") + tf.logging.info("tokens: %s" % " ".join( + [tokenization.printable_text(x) for x in instance.tokens])) + + for feature_name in features.keys(): + feature = features[feature_name] + values = [] + if feature.int64_list.value: + values = feature.int64_list.value + elif feature.float_list.value: + values = feature.float_list.value + tf.logging.info( + "%s: %s" % (feature_name, " ".join([str(x) for x in values]))) + + for writer in writers: + writer.close() + + tf.logging.info("Wrote %d total instances", total_written) + + +def create_int_feature(values): + feature = tf.train.Feature(int64_list=tf.train.Int64List(value=list(values))) + return feature + + +def create_float_feature(values): + feature = tf.train.Feature(float_list=tf.train.FloatList(value=list(values))) + return feature + + +def create_training_instances(input_files, tokenizer, max_seq_length, + dupe_factor, short_seq_prob, masked_lm_prob, + max_predictions_per_seq, rng): + """Create `TrainingInstance`s from raw text.""" + all_documents = [[]] + + # Input file format: + # (1) One sentence per line. These should ideally be actual sentences, not + # entire paragraphs or arbitrary spans of text. (Because we use the + # sentence boundaries for the "next sentence prediction" task). + # (2) Blank lines between documents. Document boundaries are needed so + # that the "next sentence prediction" task doesn't span between documents. + for input_file in input_files: + with tf.gfile.GFile(input_file, "r") as reader: + while True: + line = tokenization.convert_to_unicode(reader.readline()) + if not line: + break + line = line.strip() + + # Empty lines are used as document delimiters + if not line: + all_documents.append([]) + tokens = tokenizer.tokenize(line) + if tokens: + all_documents[-1].append(tokens) + + # Remove empty documents + all_documents = [x for x in all_documents if x] + rng.shuffle(all_documents) + + vocab_words = list(tokenizer.vocab.keys()) + instances = [] + for _ in range(dupe_factor): + for document_index in range(len(all_documents)): + instances.extend( + create_instances_from_document( + all_documents, document_index, max_seq_length, short_seq_prob, + masked_lm_prob, max_predictions_per_seq, vocab_words, rng)) + + rng.shuffle(instances) + return instances + + +def create_instances_from_document( + all_documents, document_index, max_seq_length, short_seq_prob, + masked_lm_prob, max_predictions_per_seq, vocab_words, rng): + """Creates `TrainingInstance`s for a single document.""" + document = all_documents[document_index] + + # Account for [CLS], [SEP], [SEP] + max_num_tokens = max_seq_length - 3 + + # We *usually* want to fill up the entire sequence since we are padding + # to `max_seq_length` anyways, so short sequences are generally wasted + # computation. However, we *sometimes* + # (i.e., short_seq_prob == 0.1 == 10% of the time) want to use shorter + # sequences to minimize the mismatch between pre-training and fine-tuning. + # The `target_seq_length` is just a rough target however, whereas + # `max_seq_length` is a hard limit. + target_seq_length = max_num_tokens + if rng.random() < short_seq_prob: + target_seq_length = rng.randint(2, max_num_tokens) + + # We DON'T just concatenate all of the tokens from a document into a long + # sequence and choose an arbitrary split point because this would make the + # next sentence prediction task too easy. Instead, we split the input into + # segments "A" and "B" based on the actual "sentences" provided by the user + # input. + instances = [] + current_chunk = [] + current_length = 0 + i = 0 + while i < len(document): + segment = document[i] + current_chunk.append(segment) + current_length += len(segment) + if i == len(document) - 1 or current_length >= target_seq_length: + if current_chunk: + # `a_end` is how many segments from `current_chunk` go into the `A` + # (first) sentence. + a_end = 1 + if len(current_chunk) >= 2: + a_end = rng.randint(1, len(current_chunk) - 1) + + tokens_a = [] + for j in range(a_end): + tokens_a.extend(current_chunk[j]) + + tokens_b = [] + # Random next + is_random_next = False + if len(current_chunk) == 1 or rng.random() < 0.5: + is_random_next = True + target_b_length = target_seq_length - len(tokens_a) + + # This should rarely go for more than one iteration for large + # corpora. However, just to be careful, we try to make sure that + # the random document is not the same as the document + # we're processing. + for _ in range(10): + random_document_index = rng.randint(0, len(all_documents) - 1) + if random_document_index != document_index: + break + + random_document = all_documents[random_document_index] + random_start = rng.randint(0, len(random_document) - 1) + for j in range(random_start, len(random_document)): + tokens_b.extend(random_document[j]) + if len(tokens_b) >= target_b_length: + break + # We didn't actually use these segments so we "put them back" so + # they don't go to waste. + num_unused_segments = len(current_chunk) - a_end + i -= num_unused_segments + # Actual next + else: + is_random_next = False + for j in range(a_end, len(current_chunk)): + tokens_b.extend(current_chunk[j]) + truncate_seq_pair(tokens_a, tokens_b, max_num_tokens, rng) + + assert len(tokens_a) >= 1 + assert len(tokens_b) >= 1 + + tokens = [] + segment_ids = [] + tokens.append("[CLS]") + segment_ids.append(0) + for token in tokens_a: + tokens.append(token) + segment_ids.append(0) + + tokens.append("[SEP]") + segment_ids.append(0) + + for token in tokens_b: + tokens.append(token) + segment_ids.append(1) + tokens.append("[SEP]") + segment_ids.append(1) + + (tokens, masked_lm_positions, + masked_lm_labels) = create_masked_lm_predictions( + tokens, masked_lm_prob, max_predictions_per_seq, vocab_words, rng) + instance = TrainingInstance( + tokens=tokens, + segment_ids=segment_ids, + is_random_next=is_random_next, + masked_lm_positions=masked_lm_positions, + masked_lm_labels=masked_lm_labels) + instances.append(instance) + current_chunk = [] + current_length = 0 + i += 1 + + return instances + + +MaskedLmInstance = collections.namedtuple("MaskedLmInstance", + ["index", "label"]) + + +def create_masked_lm_predictions(tokens, masked_lm_prob, + max_predictions_per_seq, vocab_words, rng): + """Creates the predictions for the masked LM objective.""" + + cand_indexes = [] + for (i, token) in enumerate(tokens): + if token == "[CLS]" or token == "[SEP]": + continue + # Whole Word Masking means that if we mask all of the wordpieces + # corresponding to an original word. When a word has been split into + # WordPieces, the first token does not have any marker and any subsequence + # tokens are prefixed with ##. So whenever we see the ## token, we + # append it to the previous set of word indexes. + # + # Note that Whole Word Masking does *not* change the training code + # at all -- we still predict each WordPiece independently, softmaxed + # over the entire vocabulary. + if (FLAGS.do_whole_word_mask and len(cand_indexes) >= 1 and + token.startswith("##")): + cand_indexes[-1].append(i) + else: + cand_indexes.append([i]) + + rng.shuffle(cand_indexes) + + output_tokens = list(tokens) + + num_to_predict = min(max_predictions_per_seq, + max(1, int(round(len(tokens) * masked_lm_prob)))) + + masked_lms = [] + covered_indexes = set() + for index_set in cand_indexes: + if len(masked_lms) >= num_to_predict: + break + # If adding a whole-word mask would exceed the maximum number of + # predictions, then just skip this candidate. + if len(masked_lms) + len(index_set) > num_to_predict: + continue + is_any_index_covered = False + for index in index_set: + if index in covered_indexes: + is_any_index_covered = True + break + if is_any_index_covered: + continue + for index in index_set: + covered_indexes.add(index) + + masked_token = None + # 80% of the time, replace with [MASK] + if rng.random() < 0.8: + masked_token = "[MASK]" + else: + # 10% of the time, keep original + if rng.random() < 0.5: + masked_token = tokens[index] + # 10% of the time, replace with random word + else: + masked_token = vocab_words[rng.randint(0, len(vocab_words) - 1)] + + output_tokens[index] = masked_token + + masked_lms.append(MaskedLmInstance(index=index, label=tokens[index])) + assert len(masked_lms) <= num_to_predict + masked_lms = sorted(masked_lms, key=lambda x: x.index) + + masked_lm_positions = [] + masked_lm_labels = [] + for p in masked_lms: + masked_lm_positions.append(p.index) + masked_lm_labels.append(p.label) + + return (output_tokens, masked_lm_positions, masked_lm_labels) + + +def truncate_seq_pair(tokens_a, tokens_b, max_num_tokens, rng): + """Truncates a pair of sequences to a maximum sequence length.""" + while True: + total_length = len(tokens_a) + len(tokens_b) + if total_length <= max_num_tokens: + break + + trunc_tokens = tokens_a if len(tokens_a) > len(tokens_b) else tokens_b + assert len(trunc_tokens) >= 1 + + # We want to sometimes truncate from the front and sometimes from the + # back to add more randomness and avoid biases. + if rng.random() < 0.5: + del trunc_tokens[0] + else: + trunc_tokens.pop() + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + tokenizer = tokenization.FullTokenizer( + vocab_file=FLAGS.vocab_file, do_lower_case=FLAGS.do_lower_case) + + input_files = [] + for input_pattern in FLAGS.input_file.split(","): + input_files.extend(tf.gfile.Glob(input_pattern)) + + tf.logging.info("*** Reading from input files ***") + for input_file in input_files: + tf.logging.info(" %s", input_file) + + rng = random.Random(FLAGS.random_seed) + instances = create_training_instances( + input_files, tokenizer, FLAGS.max_seq_length, FLAGS.dupe_factor, + FLAGS.short_seq_prob, FLAGS.masked_lm_prob, FLAGS.max_predictions_per_seq, + rng) + + output_files = FLAGS.output_file.split(",") + tf.logging.info("*** Writing to output files ***") + for output_file in output_files: + tf.logging.info(" %s", output_file) + + write_instance_to_example_files(instances, tokenizer, FLAGS.max_seq_length, + FLAGS.max_predictions_per_seq, output_files) + + +if __name__ == "__main__": + flags.mark_flag_as_required("input_file") + flags.mark_flag_as_required("output_file") + flags.mark_flag_as_required("vocab_file") + tf.app.run() diff --git a/dealing_dataset.py b/dealing_dataset.py new file mode 100644 index 0000000..d8db627 --- /dev/null +++ b/dealing_dataset.py @@ -0,0 +1,49 @@ +import sqlite3 + +conn = sqlite3.connect(r"nlpdata.db")\ + + +def create_dataset_ep(table): + cursor = conn.cursor() + sql = "select * from " + table + " LIMIT 20" + cursor.execute(sql) + conn.commit() + + dataset = [] + + for row in cursor: + eid = row[0] + tag = row[1] + content = row[2] + if tag == "5" or tag == "4": + dataset.append([eid, 2, content]) + print(eid, 2, content) + elif tag == "1" or tag == "2": + dataset.append([eid, 0, content]) + print(eid, 0, content) + else: + dataset.append([eid, 1, content]) + print(eid, 1, content) + return dataset + + +def create_dataset_pdt(): + conn_pdt = sqlite3.connect(r".\bptdata.db") + cursor = conn_pdt.cursor() + sql = "select * from " + "predict_data" + cursor.execute(sql) + conn_pdt.commit() + + dataset = [] + + for row in cursor: + stnid = row[0] + text = row[1] + dataset.append([stnid, 0, text]) + print(stnid, 0, text) + + return dataset + + +if __name__ == '__main__': + print(create_dataset_ep("amki_test")) \ No newline at end of file diff --git a/extract_features.py b/extract_features.py new file mode 100644 index 0000000..60e3830 --- /dev/null +++ b/extract_features.py @@ -0,0 +1,419 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Extract pre-computed feature vectors from BERT.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import codecs +import collections +import json +import re + +import modeling +import tokenization +import tensorflow as tf + +flags = tf.flags + +FLAGS = flags.FLAGS + +flags.DEFINE_string("input_file", None, "") + +flags.DEFINE_string("output_file", None, "") + +flags.DEFINE_string("layers", "-1,-2,-3,-4", "") + +flags.DEFINE_string( + "bert_config_file", None, + "The config json file corresponding to the pre-trained BERT model. " + "This specifies the model architecture.") + +flags.DEFINE_integer( + "max_seq_length", 128, + "The maximum total input sequence length after WordPiece tokenization. " + "Sequences longer than this will be truncated, and sequences shorter " + "than this will be padded.") + +flags.DEFINE_string( + "init_checkpoint", None, + "Initial checkpoint (usually from a pre-trained BERT model).") + +flags.DEFINE_string("vocab_file", None, + "The vocabulary file that the BERT model was trained on.") + +flags.DEFINE_bool( + "do_lower_case", True, + "Whether to lower case the input text. Should be True for uncased " + "models and False for cased models.") + +flags.DEFINE_integer("batch_size", 32, "Batch size for predictions.") + +flags.DEFINE_bool("use_tpu", False, "Whether to use TPU or GPU/CPU.") + +flags.DEFINE_string("master", None, + "If using a TPU, the address of the master.") + +flags.DEFINE_integer( + "num_tpu_cores", 8, + "Only used if `use_tpu` is True. Total number of TPU cores to use.") + +flags.DEFINE_bool( + "use_one_hot_embeddings", False, + "If True, tf.one_hot will be used for embedding lookups, otherwise " + "tf.nn.embedding_lookup will be used. On TPUs, this should be True " + "since it is much faster.") + + +class InputExample(object): + + def __init__(self, unique_id, text_a, text_b): + self.unique_id = unique_id + self.text_a = text_a + self.text_b = text_b + + +class InputFeatures(object): + """A single set of features of data.""" + + def __init__(self, unique_id, tokens, input_ids, input_mask, input_type_ids): + self.unique_id = unique_id + self.tokens = tokens + self.input_ids = input_ids + self.input_mask = input_mask + self.input_type_ids = input_type_ids + + +def input_fn_builder(features, seq_length): + """Creates an `input_fn` closure to be passed to TPUEstimator.""" + + all_unique_ids = [] + all_input_ids = [] + all_input_mask = [] + all_input_type_ids = [] + + for feature in features: + all_unique_ids.append(feature.unique_id) + all_input_ids.append(feature.input_ids) + all_input_mask.append(feature.input_mask) + all_input_type_ids.append(feature.input_type_ids) + + def input_fn(params): + """The actual input function.""" + batch_size = params["batch_size"] + + num_examples = len(features) + + # This is for demo purposes and does NOT scale to large data sets. We do + # not use Dataset.from_generator() because that uses tf.py_func which is + # not TPU compatible. The right way to load data is with TFRecordReader. + d = tf.data.Dataset.from_tensor_slices({ + "unique_ids": + tf.constant(all_unique_ids, shape=[num_examples], dtype=tf.int32), + "input_ids": + tf.constant( + all_input_ids, shape=[num_examples, seq_length], + dtype=tf.int32), + "input_mask": + tf.constant( + all_input_mask, + shape=[num_examples, seq_length], + dtype=tf.int32), + "input_type_ids": + tf.constant( + all_input_type_ids, + shape=[num_examples, seq_length], + dtype=tf.int32), + }) + + d = d.batch(batch_size=batch_size, drop_remainder=False) + return d + + return input_fn + + +def model_fn_builder(bert_config, init_checkpoint, layer_indexes, use_tpu, + use_one_hot_embeddings): + """Returns `model_fn` closure for TPUEstimator.""" + + def model_fn(features, labels, mode, params): # pylint: disable=unused-argument + """The `model_fn` for TPUEstimator.""" + + unique_ids = features["unique_ids"] + input_ids = features["input_ids"] + input_mask = features["input_mask"] + input_type_ids = features["input_type_ids"] + + model = modeling.BertModel( + config=bert_config, + is_training=False, + input_ids=input_ids, + input_mask=input_mask, + token_type_ids=input_type_ids, + use_one_hot_embeddings=use_one_hot_embeddings) + + if mode != tf.estimator.ModeKeys.PREDICT: + raise ValueError("Only PREDICT modes are supported: %s" % (mode)) + + tvars = tf.trainable_variables() + scaffold_fn = None + (assignment_map, + initialized_variable_names) = modeling.get_assignment_map_from_checkpoint( + tvars, init_checkpoint) + if use_tpu: + + def tpu_scaffold(): + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + return tf.train.Scaffold() + + scaffold_fn = tpu_scaffold + else: + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + + tf.logging.info("**** Trainable Variables ****") + for var in tvars: + init_string = "" + if var.name in initialized_variable_names: + init_string = ", *INIT_FROM_CKPT*" + tf.logging.info(" name = %s, shape = %s%s", var.name, var.shape, + init_string) + + all_layers = model.get_all_encoder_layers() + + predictions = { + "unique_id": unique_ids, + } + + for (i, layer_index) in enumerate(layer_indexes): + predictions["layer_output_%d" % i] = all_layers[layer_index] + + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, predictions=predictions, scaffold_fn=scaffold_fn) + return output_spec + + return model_fn + + +def convert_examples_to_features(examples, seq_length, tokenizer): + """Loads a data file into a list of `InputBatch`s.""" + + features = [] + for (ex_index, example) in enumerate(examples): + tokens_a = tokenizer.tokenize(example.text_a) + + tokens_b = None + if example.text_b: + tokens_b = tokenizer.tokenize(example.text_b) + + if tokens_b: + # Modifies `tokens_a` and `tokens_b` in place so that the total + # length is less than the specified length. + # Account for [CLS], [SEP], [SEP] with "- 3" + _truncate_seq_pair(tokens_a, tokens_b, seq_length - 3) + else: + # Account for [CLS] and [SEP] with "- 2" + if len(tokens_a) > seq_length - 2: + tokens_a = tokens_a[0:(seq_length - 2)] + + # The convention in BERT is: + # (a) For sequence pairs: + # tokens: [CLS] is this jack ##son ##ville ? [SEP] no it is not . [SEP] + # type_ids: 0 0 0 0 0 0 0 0 1 1 1 1 1 1 + # (b) For single sequences: + # tokens: [CLS] the dog is hairy . [SEP] + # type_ids: 0 0 0 0 0 0 0 + # + # Where "type_ids" are used to indicate whether this is the first + # sequence or the second sequence. The embedding vectors for `type=0` and + # `type=1` were learned during pre-training and are added to the wordpiece + # embedding vector (and position vector). This is not *strictly* necessary + # since the [SEP] token unambiguously separates the sequences, but it makes + # it easier for the model to learn the concept of sequences. + # + # For classification tasks, the first vector (corresponding to [CLS]) is + # used as as the "sentence vector". Note that this only makes sense because + # the entire model is fine-tuned. + tokens = [] + input_type_ids = [] + tokens.append("[CLS]") + input_type_ids.append(0) + for token in tokens_a: + tokens.append(token) + input_type_ids.append(0) + tokens.append("[SEP]") + input_type_ids.append(0) + + if tokens_b: + for token in tokens_b: + tokens.append(token) + input_type_ids.append(1) + tokens.append("[SEP]") + input_type_ids.append(1) + + input_ids = tokenizer.convert_tokens_to_ids(tokens) + + # The mask has 1 for real tokens and 0 for padding tokens. Only real + # tokens are attended to. + input_mask = [1] * len(input_ids) + + # Zero-pad up to the sequence length. + while len(input_ids) < seq_length: + input_ids.append(0) + input_mask.append(0) + input_type_ids.append(0) + + assert len(input_ids) == seq_length + assert len(input_mask) == seq_length + assert len(input_type_ids) == seq_length + + if ex_index < 5: + tf.logging.info("*** Example ***") + tf.logging.info("unique_id: %s" % (example.unique_id)) + tf.logging.info("tokens: %s" % " ".join( + [tokenization.printable_text(x) for x in tokens])) + tf.logging.info("input_ids: %s" % " ".join([str(x) for x in input_ids])) + tf.logging.info("input_mask: %s" % " ".join([str(x) for x in input_mask])) + tf.logging.info( + "input_type_ids: %s" % " ".join([str(x) for x in input_type_ids])) + + features.append( + InputFeatures( + unique_id=example.unique_id, + tokens=tokens, + input_ids=input_ids, + input_mask=input_mask, + input_type_ids=input_type_ids)) + return features + + +def _truncate_seq_pair(tokens_a, tokens_b, max_length): + """Truncates a sequence pair in place to the maximum length.""" + + # This is a simple heuristic which will always truncate the longer sequence + # one token at a time. This makes more sense than truncating an equal percent + # of tokens from each, since if one sequence is very short then each token + # that's truncated likely contains more information than a longer sequence. + while True: + total_length = len(tokens_a) + len(tokens_b) + if total_length <= max_length: + break + if len(tokens_a) > len(tokens_b): + tokens_a.pop() + else: + tokens_b.pop() + + +def read_examples(input_file): + """Read a list of `InputExample`s from an input file.""" + examples = [] + unique_id = 0 + with tf.gfile.GFile(input_file, "r") as reader: + while True: + line = tokenization.convert_to_unicode(reader.readline()) + if not line: + break + line = line.strip() + text_a = None + text_b = None + m = re.match(r"^(.*) \|\|\| (.*)$", line) + if m is None: + text_a = line + else: + text_a = m.group(1) + text_b = m.group(2) + examples.append( + InputExample(unique_id=unique_id, text_a=text_a, text_b=text_b)) + unique_id += 1 + return examples + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + layer_indexes = [int(x) for x in FLAGS.layers.split(",")] + + bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file) + + tokenizer = tokenization.FullTokenizer( + vocab_file=FLAGS.vocab_file, do_lower_case=FLAGS.do_lower_case) + + is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2 + run_config = tf.contrib.tpu.RunConfig( + master=FLAGS.master, + tpu_config=tf.contrib.tpu.TPUConfig( + num_shards=FLAGS.num_tpu_cores, + per_host_input_for_training=is_per_host)) + + examples = read_examples(FLAGS.input_file) + + features = convert_examples_to_features( + examples=examples, seq_length=FLAGS.max_seq_length, tokenizer=tokenizer) + + unique_id_to_feature = {} + for feature in features: + unique_id_to_feature[feature.unique_id] = feature + + model_fn = model_fn_builder( + bert_config=bert_config, + init_checkpoint=FLAGS.init_checkpoint, + layer_indexes=layer_indexes, + use_tpu=FLAGS.use_tpu, + use_one_hot_embeddings=FLAGS.use_one_hot_embeddings) + + # If TPU is not available, this will fall back to normal Estimator on CPU + # or GPU. + estimator = tf.contrib.tpu.TPUEstimator( + use_tpu=FLAGS.use_tpu, + model_fn=model_fn, + config=run_config, + predict_batch_size=FLAGS.batch_size) + + input_fn = input_fn_builder( + features=features, seq_length=FLAGS.max_seq_length) + + with codecs.getwriter("utf-8")(tf.gfile.Open(FLAGS.output_file, + "w")) as writer: + for result in estimator.predict(input_fn, yield_single_examples=True): + unique_id = int(result["unique_id"]) + feature = unique_id_to_feature[unique_id] + output_json = collections.OrderedDict() + output_json["linex_index"] = unique_id + all_features = [] + for (i, token) in enumerate(feature.tokens): + all_layers = [] + for (j, layer_index) in enumerate(layer_indexes): + layer_output = result["layer_output_%d" % j] + layers = collections.OrderedDict() + layers["index"] = layer_index + layers["values"] = [ + round(float(x), 6) for x in layer_output[i:(i + 1)].flat + ] + all_layers.append(layers) + features = collections.OrderedDict() + features["token"] = token + features["layers"] = all_layers + all_features.append(features) + output_json["features"] = all_features + writer.write(json.dumps(output_json) + "\n") + + +if __name__ == "__main__": + flags.mark_flag_as_required("input_file") + flags.mark_flag_as_required("vocab_file") + flags.mark_flag_as_required("bert_config_file") + flags.mark_flag_as_required("init_checkpoint") + flags.mark_flag_as_required("output_file") + tf.app.run() diff --git a/modeling.py b/modeling.py new file mode 100644 index 0000000..fed5259 --- /dev/null +++ b/modeling.py @@ -0,0 +1,986 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""The main BERT model and related functions.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import collections +import copy +import json +import math +import re +import numpy as np +import six +import tensorflow as tf + + +class BertConfig(object): + """Configuration for `BertModel`.""" + + def __init__(self, + vocab_size, + hidden_size=768, + num_hidden_layers=12, + num_attention_heads=12, + intermediate_size=3072, + hidden_act="gelu", + hidden_dropout_prob=0.1, + attention_probs_dropout_prob=0.1, + max_position_embeddings=512, + type_vocab_size=16, + initializer_range=0.02): + """Constructs BertConfig. + + Args: + vocab_size: Vocabulary size of `inputs_ids` in `BertModel`. + hidden_size: Size of the encoder layers and the pooler layer. + num_hidden_layers: Number of hidden layers in the Transformer encoder. + num_attention_heads: Number of attention heads for each attention layer in + the Transformer encoder. + intermediate_size: The size of the "intermediate" (i.e., feed-forward) + layer in the Transformer encoder. + hidden_act: The non-linear activation function (function or string) in the + encoder and pooler. + hidden_dropout_prob: The dropout probability for all fully connected + layers in the embeddings, encoder, and pooler. + attention_probs_dropout_prob: The dropout ratio for the attention + probabilities. + max_position_embeddings: The maximum sequence length that this model might + ever be used with. Typically set this to something large just in case + (e.g., 512 or 1024 or 2048). + type_vocab_size: The vocabulary size of the `token_type_ids` passed into + `BertModel`. + initializer_range: The stdev of the truncated_normal_initializer for + initializing all weight matrices. + """ + self.vocab_size = vocab_size + self.hidden_size = hidden_size + self.num_hidden_layers = num_hidden_layers + self.num_attention_heads = num_attention_heads + self.hidden_act = hidden_act + self.intermediate_size = intermediate_size + self.hidden_dropout_prob = hidden_dropout_prob + self.attention_probs_dropout_prob = attention_probs_dropout_prob + self.max_position_embeddings = max_position_embeddings + self.type_vocab_size = type_vocab_size + self.initializer_range = initializer_range + + @classmethod + def from_dict(cls, json_object): + """Constructs a `BertConfig` from a Python dictionary of parameters.""" + config = BertConfig(vocab_size=None) + for (key, value) in six.iteritems(json_object): + config.__dict__[key] = value + return config + + @classmethod + def from_json_file(cls, json_file): + """Constructs a `BertConfig` from a json file of parameters.""" + with tf.gfile.GFile(json_file, "r") as reader: + text = reader.read() + return cls.from_dict(json.loads(text)) + + def to_dict(self): + """Serializes this instance to a Python dictionary.""" + output = copy.deepcopy(self.__dict__) + return output + + def to_json_string(self): + """Serializes this instance to a JSON string.""" + return json.dumps(self.to_dict(), indent=2, sort_keys=True) + "\n" + + +class BertModel(object): + """BERT model ("Bidirectional Encoder Representations from Transformers"). + + Example usage: + + ```python + # Already been converted into WordPiece token ids + input_ids = tf.constant([[31, 51, 99], [15, 5, 0]]) + input_mask = tf.constant([[1, 1, 1], [1, 1, 0]]) + token_type_ids = tf.constant([[0, 0, 1], [0, 2, 0]]) + + config = modeling.BertConfig(vocab_size=32000, hidden_size=512, + num_hidden_layers=8, num_attention_heads=6, intermediate_size=1024) + + model = modeling.BertModel(config=config, is_training=True, + input_ids=input_ids, input_mask=input_mask, token_type_ids=token_type_ids) + + label_embeddings = tf.get_variable(...) + pooled_output = model.get_pooled_output() + logits = tf.matmul(pooled_output, label_embeddings) + ... + ``` + """ + + def __init__(self, + config, + is_training, + input_ids, + input_mask=None, + token_type_ids=None, + use_one_hot_embeddings=False, + scope=None): + """Constructor for BertModel. + + Args: + config: `BertConfig` instance. + is_training: bool. true for training model, false for eval model. Controls + whether dropout will be applied. + input_ids: int32 Tensor of shape [batch_size, seq_length]. + input_mask: (optional) int32 Tensor of shape [batch_size, seq_length]. + token_type_ids: (optional) int32 Tensor of shape [batch_size, seq_length]. + use_one_hot_embeddings: (optional) bool. Whether to use one-hot word + embeddings or tf.embedding_lookup() for the word embeddings. + scope: (optional) variable scope. Defaults to "bert". + + Raises: + ValueError: The config is invalid or one of the input tensor shapes + is invalid. + """ + config = copy.deepcopy(config) + if not is_training: + config.hidden_dropout_prob = 0.0 + config.attention_probs_dropout_prob = 0.0 + + input_shape = get_shape_list(input_ids, expected_rank=2) + batch_size = input_shape[0] + seq_length = input_shape[1] + + if input_mask is None: + input_mask = tf.ones(shape=[batch_size, seq_length], dtype=tf.int32) + + if token_type_ids is None: + token_type_ids = tf.zeros(shape=[batch_size, seq_length], dtype=tf.int32) + + with tf.variable_scope(scope, default_name="bert"): + with tf.variable_scope("embeddings"): + # Perform embedding lookup on the word ids. + (self.embedding_output, self.embedding_table) = embedding_lookup( + input_ids=input_ids, + vocab_size=config.vocab_size, + embedding_size=config.hidden_size, + initializer_range=config.initializer_range, + word_embedding_name="word_embeddings", + use_one_hot_embeddings=use_one_hot_embeddings) + + # Add positional embeddings and token type embeddings, then layer + # normalize and perform dropout. + self.embedding_output = embedding_postprocessor( + input_tensor=self.embedding_output, + use_token_type=True, + token_type_ids=token_type_ids, + token_type_vocab_size=config.type_vocab_size, + token_type_embedding_name="token_type_embeddings", + use_position_embeddings=True, + position_embedding_name="position_embeddings", + initializer_range=config.initializer_range, + max_position_embeddings=config.max_position_embeddings, + dropout_prob=config.hidden_dropout_prob) + + with tf.variable_scope("encoder"): + # This converts a 2D mask of shape [batch_size, seq_length] to a 3D + # mask of shape [batch_size, seq_length, seq_length] which is used + # for the attention scores. + attention_mask = create_attention_mask_from_input_mask( + input_ids, input_mask) + + # Run the stacked transformer. + # `sequence_output` shape = [batch_size, seq_length, hidden_size]. + self.all_encoder_layers = transformer_model( + input_tensor=self.embedding_output, + attention_mask=attention_mask, + hidden_size=config.hidden_size, + num_hidden_layers=config.num_hidden_layers, + num_attention_heads=config.num_attention_heads, + intermediate_size=config.intermediate_size, + intermediate_act_fn=get_activation(config.hidden_act), + hidden_dropout_prob=config.hidden_dropout_prob, + attention_probs_dropout_prob=config.attention_probs_dropout_prob, + initializer_range=config.initializer_range, + do_return_all_layers=True) + + self.sequence_output = self.all_encoder_layers[-1] + # The "pooler" converts the encoded sequence tensor of shape + # [batch_size, seq_length, hidden_size] to a tensor of shape + # [batch_size, hidden_size]. This is necessary for segment-level + # (or segment-pair-level) classification tasks where we need a fixed + # dimensional representation of the segment. + with tf.variable_scope("pooler"): + # We "pool" the model by simply taking the hidden state corresponding + # to the first token. We assume that this has been pre-trained + first_token_tensor = tf.squeeze(self.sequence_output[:, 0:1, :], axis=1) + self.pooled_output = tf.layers.dense( + first_token_tensor, + config.hidden_size, + activation=tf.tanh, + kernel_initializer=create_initializer(config.initializer_range)) + + def get_pooled_output(self): + return self.pooled_output + + def get_sequence_output(self): + """Gets final hidden layer of encoder. + + Returns: + float Tensor of shape [batch_size, seq_length, hidden_size] corresponding + to the final hidden of the transformer encoder. + """ + return self.sequence_output + + def get_all_encoder_layers(self): + return self.all_encoder_layers + + def get_embedding_output(self): + """Gets output of the embedding lookup (i.e., input to the transformer). + + Returns: + float Tensor of shape [batch_size, seq_length, hidden_size] corresponding + to the output of the embedding layer, after summing the word + embeddings with the positional embeddings and the token type embeddings, + then performing layer normalization. This is the input to the transformer. + """ + return self.embedding_output + + def get_embedding_table(self): + return self.embedding_table + + +def gelu(x): + """Gaussian Error Linear Unit. + + This is a smoother version of the RELU. + Original paper: https://arxiv.org/abs/1606.08415 + Args: + x: float Tensor to perform activation. + + Returns: + `x` with the GELU activation applied. + """ + cdf = 0.5 * (1.0 + tf.tanh( + (np.sqrt(2 / np.pi) * (x + 0.044715 * tf.pow(x, 3))))) + return x * cdf + + +def get_activation(activation_string): + """Maps a string to a Python function, e.g., "relu" => `tf.nn.relu`. + + Args: + activation_string: String name of the activation function. + + Returns: + A Python function corresponding to the activation function. If + `activation_string` is None, empty, or "linear", this will return None. + If `activation_string` is not a string, it will return `activation_string`. + + Raises: + ValueError: The `activation_string` does not correspond to a known + activation. + """ + + # We assume that anything that"s not a string is already an activation + # function, so we just return it. + if not isinstance(activation_string, six.string_types): + return activation_string + + if not activation_string: + return None + + act = activation_string.lower() + if act == "linear": + return None + elif act == "relu": + return tf.nn.relu + elif act == "gelu": + return gelu + elif act == "tanh": + return tf.tanh + else: + raise ValueError("Unsupported activation: %s" % act) + + +def get_assignment_map_from_checkpoint(tvars, init_checkpoint): + """Compute the union of the current variables and checkpoint variables.""" + assignment_map = {} + initialized_variable_names = {} + + name_to_variable = collections.OrderedDict() + for var in tvars: + name = var.name + m = re.match("^(.*):\\d+$", name) + if m is not None: + name = m.group(1) + name_to_variable[name] = var + + init_vars = tf.train.list_variables(init_checkpoint) + + assignment_map = collections.OrderedDict() + for x in init_vars: + (name, var) = (x[0], x[1]) + if name not in name_to_variable: + continue + assignment_map[name] = name + initialized_variable_names[name] = 1 + initialized_variable_names[name + ":0"] = 1 + + return (assignment_map, initialized_variable_names) + + +def dropout(input_tensor, dropout_prob): + """Perform dropout. + + Args: + input_tensor: float Tensor. + dropout_prob: Python float. The probability of dropping out a value (NOT of + *keeping* a dimension as in `tf.nn.dropout`). + + Returns: + A version of `input_tensor` with dropout applied. + """ + if dropout_prob is None or dropout_prob == 0.0: + return input_tensor + + output = tf.nn.dropout(input_tensor, 1.0 - dropout_prob) + return output + + +def layer_norm(input_tensor, name=None): + """Run layer normalization on the last dimension of the tensor.""" + return tf.contrib.layers.layer_norm( + inputs=input_tensor, begin_norm_axis=-1, begin_params_axis=-1, scope=name) + + +def layer_norm_and_dropout(input_tensor, dropout_prob, name=None): + """Runs layer normalization followed by dropout.""" + output_tensor = layer_norm(input_tensor, name) + output_tensor = dropout(output_tensor, dropout_prob) + return output_tensor + + +def create_initializer(initializer_range=0.02): + """Creates a `truncated_normal_initializer` with the given range.""" + return tf.truncated_normal_initializer(stddev=initializer_range) + + +def embedding_lookup(input_ids, + vocab_size, + embedding_size=128, + initializer_range=0.02, + word_embedding_name="word_embeddings", + use_one_hot_embeddings=False): + """Looks up words embeddings for id tensor. + + Args: + input_ids: int32 Tensor of shape [batch_size, seq_length] containing word + ids. + vocab_size: int. Size of the embedding vocabulary. + embedding_size: int. Width of the word embeddings. + initializer_range: float. Embedding initialization range. + word_embedding_name: string. Name of the embedding table. + use_one_hot_embeddings: bool. If True, use one-hot method for word + embeddings. If False, use `tf.gather()`. + + Returns: + float Tensor of shape [batch_size, seq_length, embedding_size]. + """ + # This function assumes that the input is of shape [batch_size, seq_length, + # num_inputs]. + # + # If the input is a 2D tensor of shape [batch_size, seq_length], we + # reshape to [batch_size, seq_length, 1]. + if input_ids.shape.ndims == 2: + input_ids = tf.expand_dims(input_ids, axis=[-1]) + + embedding_table = tf.get_variable( + name=word_embedding_name, + shape=[vocab_size, embedding_size], + initializer=create_initializer(initializer_range)) + + flat_input_ids = tf.reshape(input_ids, [-1]) + if use_one_hot_embeddings: + one_hot_input_ids = tf.one_hot(flat_input_ids, depth=vocab_size) + output = tf.matmul(one_hot_input_ids, embedding_table) + else: + output = tf.gather(embedding_table, flat_input_ids) + + input_shape = get_shape_list(input_ids) + + output = tf.reshape(output, + input_shape[0:-1] + [input_shape[-1] * embedding_size]) + return (output, embedding_table) + + +def embedding_postprocessor(input_tensor, + use_token_type=False, + token_type_ids=None, + token_type_vocab_size=16, + token_type_embedding_name="token_type_embeddings", + use_position_embeddings=True, + position_embedding_name="position_embeddings", + initializer_range=0.02, + max_position_embeddings=512, + dropout_prob=0.1): + """Performs various post-processing on a word embedding tensor. + + Args: + input_tensor: float Tensor of shape [batch_size, seq_length, + embedding_size]. + use_token_type: bool. Whether to add embeddings for `token_type_ids`. + token_type_ids: (optional) int32 Tensor of shape [batch_size, seq_length]. + Must be specified if `use_token_type` is True. + token_type_vocab_size: int. The vocabulary size of `token_type_ids`. + token_type_embedding_name: string. The name of the embedding table variable + for token type ids. + use_position_embeddings: bool. Whether to add position embeddings for the + position of each token in the sequence. + position_embedding_name: string. The name of the embedding table variable + for positional embeddings. + initializer_range: float. Range of the weight initialization. + max_position_embeddings: int. Maximum sequence length that might ever be + used with this model. This can be longer than the sequence length of + input_tensor, but cannot be shorter. + dropout_prob: float. Dropout probability applied to the final output tensor. + + Returns: + float tensor with same shape as `input_tensor`. + + Raises: + ValueError: One of the tensor shapes or input values is invalid. + """ + input_shape = get_shape_list(input_tensor, expected_rank=3) + batch_size = input_shape[0] + seq_length = input_shape[1] + width = input_shape[2] + + output = input_tensor + + if use_token_type: + if token_type_ids is None: + raise ValueError("`token_type_ids` must be specified if" + "`use_token_type` is True.") + token_type_table = tf.get_variable( + name=token_type_embedding_name, + shape=[token_type_vocab_size, width], + initializer=create_initializer(initializer_range)) + # This vocab will be small so we always do one-hot here, since it is always + # faster for a small vocabulary. + flat_token_type_ids = tf.reshape(token_type_ids, [-1]) + one_hot_ids = tf.one_hot(flat_token_type_ids, depth=token_type_vocab_size) + token_type_embeddings = tf.matmul(one_hot_ids, token_type_table) + token_type_embeddings = tf.reshape(token_type_embeddings, + [batch_size, seq_length, width]) + output += token_type_embeddings + + if use_position_embeddings: + assert_op = tf.assert_less_equal(seq_length, max_position_embeddings) + with tf.control_dependencies([assert_op]): + full_position_embeddings = tf.get_variable( + name=position_embedding_name, + shape=[max_position_embeddings, width], + initializer=create_initializer(initializer_range)) + # Since the position embedding table is a learned variable, we create it + # using a (long) sequence length `max_position_embeddings`. The actual + # sequence length might be shorter than this, for faster training of + # tasks that do not have long sequences. + # + # So `full_position_embeddings` is effectively an embedding table + # for position [0, 1, 2, ..., max_position_embeddings-1], and the current + # sequence has positions [0, 1, 2, ... seq_length-1], so we can just + # perform a slice. + position_embeddings = tf.slice(full_position_embeddings, [0, 0], + [seq_length, -1]) + num_dims = len(output.shape.as_list()) + + # Only the last two dimensions are relevant (`seq_length` and `width`), so + # we broadcast among the first dimensions, which is typically just + # the batch size. + position_broadcast_shape = [] + for _ in range(num_dims - 2): + position_broadcast_shape.append(1) + position_broadcast_shape.extend([seq_length, width]) + position_embeddings = tf.reshape(position_embeddings, + position_broadcast_shape) + output += position_embeddings + + output = layer_norm_and_dropout(output, dropout_prob) + return output + + +def create_attention_mask_from_input_mask(from_tensor, to_mask): + """Create 3D attention mask from a 2D tensor mask. + + Args: + from_tensor: 2D or 3D Tensor of shape [batch_size, from_seq_length, ...]. + to_mask: int32 Tensor of shape [batch_size, to_seq_length]. + + Returns: + float Tensor of shape [batch_size, from_seq_length, to_seq_length]. + """ + from_shape = get_shape_list(from_tensor, expected_rank=[2, 3]) + batch_size = from_shape[0] + from_seq_length = from_shape[1] + + to_shape = get_shape_list(to_mask, expected_rank=2) + to_seq_length = to_shape[1] + + to_mask = tf.cast( + tf.reshape(to_mask, [batch_size, 1, to_seq_length]), tf.float32) + + # We don't assume that `from_tensor` is a mask (although it could be). We + # don't actually care if we attend *from* padding tokens (only *to* padding) + # tokens so we create a tensor of all ones. + # + # `broadcast_ones` = [batch_size, from_seq_length, 1] + broadcast_ones = tf.ones( + shape=[batch_size, from_seq_length, 1], dtype=tf.float32) + + # Here we broadcast along two dimensions to create the mask. + mask = broadcast_ones * to_mask + + return mask + + +def attention_layer(from_tensor, + to_tensor, + attention_mask=None, + num_attention_heads=1, + size_per_head=512, + query_act=None, + key_act=None, + value_act=None, + attention_probs_dropout_prob=0.0, + initializer_range=0.02, + do_return_2d_tensor=False, + batch_size=None, + from_seq_length=None, + to_seq_length=None): + """Performs multi-headed attention from `from_tensor` to `to_tensor`. + + This is an implementation of multi-headed attention based on "Attention + is all you Need". If `from_tensor` and `to_tensor` are the same, then + this is self-attention. Each timestep in `from_tensor` attends to the + corresponding sequence in `to_tensor`, and returns a fixed-with vector. + + This function first projects `from_tensor` into a "query" tensor and + `to_tensor` into "key" and "value" tensors. These are (effectively) a list + of tensors of length `num_attention_heads`, where each tensor is of shape + [batch_size, seq_length, size_per_head]. + + Then, the query and key tensors are dot-producted and scaled. These are + softmaxed to obtain attention probabilities. The value tensors are then + interpolated by these probabilities, then concatenated back to a single + tensor and returned. + + In practice, the multi-headed attention are done with transposes and + reshapes rather than actual separate tensors. + + Args: + from_tensor: float Tensor of shape [batch_size, from_seq_length, + from_width]. + to_tensor: float Tensor of shape [batch_size, to_seq_length, to_width]. + attention_mask: (optional) int32 Tensor of shape [batch_size, + from_seq_length, to_seq_length]. The values should be 1 or 0. The + attention scores will effectively be set to -infinity for any positions in + the mask that are 0, and will be unchanged for positions that are 1. + num_attention_heads: int. Number of attention heads. + size_per_head: int. Size of each attention head. + query_act: (optional) Activation function for the query transform. + key_act: (optional) Activation function for the key transform. + value_act: (optional) Activation function for the value transform. + attention_probs_dropout_prob: (optional) float. Dropout probability of the + attention probabilities. + initializer_range: float. Range of the weight initializer. + do_return_2d_tensor: bool. If True, the output will be of shape [batch_size + * from_seq_length, num_attention_heads * size_per_head]. If False, the + output will be of shape [batch_size, from_seq_length, num_attention_heads + * size_per_head]. + batch_size: (Optional) int. If the input is 2D, this might be the batch size + of the 3D version of the `from_tensor` and `to_tensor`. + from_seq_length: (Optional) If the input is 2D, this might be the seq length + of the 3D version of the `from_tensor`. + to_seq_length: (Optional) If the input is 2D, this might be the seq length + of the 3D version of the `to_tensor`. + + Returns: + float Tensor of shape [batch_size, from_seq_length, + num_attention_heads * size_per_head]. (If `do_return_2d_tensor` is + true, this will be of shape [batch_size * from_seq_length, + num_attention_heads * size_per_head]). + + Raises: + ValueError: Any of the arguments or tensor shapes are invalid. + """ + + def transpose_for_scores(input_tensor, batch_size, num_attention_heads, + seq_length, width): + output_tensor = tf.reshape( + input_tensor, [batch_size, seq_length, num_attention_heads, width]) + + output_tensor = tf.transpose(output_tensor, [0, 2, 1, 3]) + return output_tensor + + from_shape = get_shape_list(from_tensor, expected_rank=[2, 3]) + to_shape = get_shape_list(to_tensor, expected_rank=[2, 3]) + + if len(from_shape) != len(to_shape): + raise ValueError( + "The rank of `from_tensor` must match the rank of `to_tensor`.") + + if len(from_shape) == 3: + batch_size = from_shape[0] + from_seq_length = from_shape[1] + to_seq_length = to_shape[1] + elif len(from_shape) == 2: + if (batch_size is None or from_seq_length is None or to_seq_length is None): + raise ValueError( + "When passing in rank 2 tensors to attention_layer, the values " + "for `batch_size`, `from_seq_length`, and `to_seq_length` " + "must all be specified.") + + # Scalar dimensions referenced here: + # B = batch size (number of sequences) + # F = `from_tensor` sequence length + # T = `to_tensor` sequence length + # N = `num_attention_heads` + # H = `size_per_head` + + from_tensor_2d = reshape_to_matrix(from_tensor) + to_tensor_2d = reshape_to_matrix(to_tensor) + + # `query_layer` = [B*F, N*H] + query_layer = tf.layers.dense( + from_tensor_2d, + num_attention_heads * size_per_head, + activation=query_act, + name="query", + kernel_initializer=create_initializer(initializer_range)) + + # `key_layer` = [B*T, N*H] + key_layer = tf.layers.dense( + to_tensor_2d, + num_attention_heads * size_per_head, + activation=key_act, + name="key", + kernel_initializer=create_initializer(initializer_range)) + + # `value_layer` = [B*T, N*H] + value_layer = tf.layers.dense( + to_tensor_2d, + num_attention_heads * size_per_head, + activation=value_act, + name="value", + kernel_initializer=create_initializer(initializer_range)) + + # `query_layer` = [B, N, F, H] + query_layer = transpose_for_scores(query_layer, batch_size, + num_attention_heads, from_seq_length, + size_per_head) + + # `key_layer` = [B, N, T, H] + key_layer = transpose_for_scores(key_layer, batch_size, num_attention_heads, + to_seq_length, size_per_head) + + # Take the dot product between "query" and "key" to get the raw + # attention scores. + # `attention_scores` = [B, N, F, T] + attention_scores = tf.matmul(query_layer, key_layer, transpose_b=True) + attention_scores = tf.multiply(attention_scores, + 1.0 / math.sqrt(float(size_per_head))) + + if attention_mask is not None: + # `attention_mask` = [B, 1, F, T] + attention_mask = tf.expand_dims(attention_mask, axis=[1]) + + # Since attention_mask is 1.0 for positions we want to attend and 0.0 for + # masked positions, this operation will create a tensor which is 0.0 for + # positions we want to attend and -10000.0 for masked positions. + adder = (1.0 - tf.cast(attention_mask, tf.float32)) * -10000.0 + + # Since we are adding it to the raw scores before the softmax, this is + # effectively the same as removing these entirely. + attention_scores += adder + + # Normalize the attention scores to probabilities. + # `attention_probs` = [B, N, F, T] + attention_probs = tf.nn.softmax(attention_scores) + + # This is actually dropping out entire tokens to attend to, which might + # seem a bit unusual, but is taken from the original Transformer paper. + attention_probs = dropout(attention_probs, attention_probs_dropout_prob) + + # `value_layer` = [B, T, N, H] + value_layer = tf.reshape( + value_layer, + [batch_size, to_seq_length, num_attention_heads, size_per_head]) + + # `value_layer` = [B, N, T, H] + value_layer = tf.transpose(value_layer, [0, 2, 1, 3]) + + # `context_layer` = [B, N, F, H] + context_layer = tf.matmul(attention_probs, value_layer) + + # `context_layer` = [B, F, N, H] + context_layer = tf.transpose(context_layer, [0, 2, 1, 3]) + + if do_return_2d_tensor: + # `context_layer` = [B*F, N*H] + context_layer = tf.reshape( + context_layer, + [batch_size * from_seq_length, num_attention_heads * size_per_head]) + else: + # `context_layer` = [B, F, N*H] + context_layer = tf.reshape( + context_layer, + [batch_size, from_seq_length, num_attention_heads * size_per_head]) + + return context_layer + + +def transformer_model(input_tensor, + attention_mask=None, + hidden_size=768, + num_hidden_layers=12, + num_attention_heads=12, + intermediate_size=3072, + intermediate_act_fn=gelu, + hidden_dropout_prob=0.1, + attention_probs_dropout_prob=0.1, + initializer_range=0.02, + do_return_all_layers=False): + """Multi-headed, multi-layer Transformer from "Attention is All You Need". + + This is almost an exact implementation of the original Transformer encoder. + + See the original paper: + https://arxiv.org/abs/1706.03762 + + Also see: + https://github.com/tensorflow/tensor2tensor/blob/master/tensor2tensor/models/transformer.py + + Args: + input_tensor: float Tensor of shape [batch_size, seq_length, hidden_size]. + attention_mask: (optional) int32 Tensor of shape [batch_size, seq_length, + seq_length], with 1 for positions that can be attended to and 0 in + positions that should not be. + hidden_size: int. Hidden size of the Transformer. + num_hidden_layers: int. Number of layers (blocks) in the Transformer. + num_attention_heads: int. Number of attention heads in the Transformer. + intermediate_size: int. The size of the "intermediate" (a.k.a., feed + forward) layer. + intermediate_act_fn: function. The non-linear activation function to apply + to the output of the intermediate/feed-forward layer. + hidden_dropout_prob: float. Dropout probability for the hidden layers. + attention_probs_dropout_prob: float. Dropout probability of the attention + probabilities. + initializer_range: float. Range of the initializer (stddev of truncated + normal). + do_return_all_layers: Whether to also return all layers or just the final + layer. + + Returns: + float Tensor of shape [batch_size, seq_length, hidden_size], the final + hidden layer of the Transformer. + + Raises: + ValueError: A Tensor shape or parameter is invalid. + """ + if hidden_size % num_attention_heads != 0: + raise ValueError( + "The hidden size (%d) is not a multiple of the number of attention " + "heads (%d)" % (hidden_size, num_attention_heads)) + + attention_head_size = int(hidden_size / num_attention_heads) + input_shape = get_shape_list(input_tensor, expected_rank=3) + batch_size = input_shape[0] + seq_length = input_shape[1] + input_width = input_shape[2] + + # The Transformer performs sum residuals on all layers so the input needs + # to be the same as the hidden size. + if input_width != hidden_size: + raise ValueError("The width of the input tensor (%d) != hidden size (%d)" % + (input_width, hidden_size)) + + # We keep the representation as a 2D tensor to avoid re-shaping it back and + # forth from a 3D tensor to a 2D tensor. Re-shapes are normally free on + # the GPU/CPU but may not be free on the TPU, so we want to minimize them to + # help the optimizer. + prev_output = reshape_to_matrix(input_tensor) + + all_layer_outputs = [] + for layer_idx in range(num_hidden_layers): + with tf.variable_scope("layer_%d" % layer_idx): + layer_input = prev_output + + with tf.variable_scope("attention"): + attention_heads = [] + with tf.variable_scope("self"): + attention_head = attention_layer( + from_tensor=layer_input, + to_tensor=layer_input, + attention_mask=attention_mask, + num_attention_heads=num_attention_heads, + size_per_head=attention_head_size, + attention_probs_dropout_prob=attention_probs_dropout_prob, + initializer_range=initializer_range, + do_return_2d_tensor=True, + batch_size=batch_size, + from_seq_length=seq_length, + to_seq_length=seq_length) + attention_heads.append(attention_head) + + attention_output = None + if len(attention_heads) == 1: + attention_output = attention_heads[0] + else: + # In the case where we have other sequences, we just concatenate + # them to the self-attention head before the projection. + attention_output = tf.concat(attention_heads, axis=-1) + + # Run a linear projection of `hidden_size` then add a residual + # with `layer_input`. + with tf.variable_scope("output"): + attention_output = tf.layers.dense( + attention_output, + hidden_size, + kernel_initializer=create_initializer(initializer_range)) + attention_output = dropout(attention_output, hidden_dropout_prob) + attention_output = layer_norm(attention_output + layer_input) + + # The activation is only applied to the "intermediate" hidden layer. + with tf.variable_scope("intermediate"): + intermediate_output = tf.layers.dense( + attention_output, + intermediate_size, + activation=intermediate_act_fn, + kernel_initializer=create_initializer(initializer_range)) + + # Down-project back to `hidden_size` then add the residual. + with tf.variable_scope("output"): + layer_output = tf.layers.dense( + intermediate_output, + hidden_size, + kernel_initializer=create_initializer(initializer_range)) + layer_output = dropout(layer_output, hidden_dropout_prob) + layer_output = layer_norm(layer_output + attention_output) + prev_output = layer_output + all_layer_outputs.append(layer_output) + + if do_return_all_layers: + final_outputs = [] + for layer_output in all_layer_outputs: + final_output = reshape_from_matrix(layer_output, input_shape) + final_outputs.append(final_output) + return final_outputs + else: + final_output = reshape_from_matrix(prev_output, input_shape) + return final_output + + +def get_shape_list(tensor, expected_rank=None, name=None): + """Returns a list of the shape of tensor, preferring static dimensions. + + Args: + tensor: A tf.Tensor object to find the shape of. + expected_rank: (optional) int. The expected rank of `tensor`. If this is + specified and the `tensor` has a different rank, and exception will be + thrown. + name: Optional name of the tensor for the error message. + + Returns: + A list of dimensions of the shape of tensor. All static dimensions will + be returned as python integers, and dynamic dimensions will be returned + as tf.Tensor scalars. + """ + if name is None: + name = tensor.name + + if expected_rank is not None: + assert_rank(tensor, expected_rank, name) + + shape = tensor.shape.as_list() + + non_static_indexes = [] + for (index, dim) in enumerate(shape): + if dim is None: + non_static_indexes.append(index) + + if not non_static_indexes: + return shape + + dyn_shape = tf.shape(tensor) + for index in non_static_indexes: + shape[index] = dyn_shape[index] + return shape + + +def reshape_to_matrix(input_tensor): + """Reshapes a >= rank 2 tensor to a rank 2 tensor (i.e., a matrix).""" + ndims = input_tensor.shape.ndims + if ndims < 2: + raise ValueError("Input tensor must have at least rank 2. Shape = %s" % + (input_tensor.shape)) + if ndims == 2: + return input_tensor + + width = input_tensor.shape[-1] + output_tensor = tf.reshape(input_tensor, [-1, width]) + return output_tensor + + +def reshape_from_matrix(output_tensor, orig_shape_list): + """Reshapes a rank 2 tensor back to its original rank >= 2 tensor.""" + if len(orig_shape_list) == 2: + return output_tensor + + output_shape = get_shape_list(output_tensor) + + orig_dims = orig_shape_list[0:-1] + width = output_shape[-1] + + return tf.reshape(output_tensor, orig_dims + [width]) + + +def assert_rank(tensor, expected_rank, name=None): + """Raises an exception if the tensor rank is not of the expected rank. + + Args: + tensor: A tf.Tensor to check the rank of. + expected_rank: Python integer or list of integers, expected rank. + name: Optional name of the tensor for the error message. + + Raises: + ValueError: If the expected shape doesn't match the actual shape. + """ + if name is None: + name = tensor.name + + expected_rank_dict = {} + if isinstance(expected_rank, six.integer_types): + expected_rank_dict[expected_rank] = True + else: + for x in expected_rank: + expected_rank_dict[x] = True + + actual_rank = tensor.shape.ndims + if actual_rank not in expected_rank_dict: + scope_name = tf.get_variable_scope().name + raise ValueError( + "For the tensor `%s` in scope `%s`, the actual rank " + "`%d` (shape = %s) is not equal to the expected rank `%s`" % + (name, scope_name, actual_rank, str(tensor.shape), str(expected_rank))) diff --git a/modeling_test.py b/modeling_test.py new file mode 100644 index 0000000..817ad2d --- /dev/null +++ b/modeling_test.py @@ -0,0 +1,277 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import collections +import json +import random +import re + +import modeling +import six +import tensorflow as tf + + +class BertModelTest(tf.test.TestCase): + + class BertModelTester(object): + + def __init__(self, + parent, + batch_size=13, + seq_length=7, + is_training=True, + use_input_mask=True, + use_token_type_ids=True, + vocab_size=99, + hidden_size=32, + num_hidden_layers=5, + num_attention_heads=4, + intermediate_size=37, + hidden_act="gelu", + hidden_dropout_prob=0.1, + attention_probs_dropout_prob=0.1, + max_position_embeddings=512, + type_vocab_size=16, + initializer_range=0.02, + scope=None): + self.parent = parent + self.batch_size = batch_size + self.seq_length = seq_length + self.is_training = is_training + self.use_input_mask = use_input_mask + self.use_token_type_ids = use_token_type_ids + self.vocab_size = vocab_size + self.hidden_size = hidden_size + self.num_hidden_layers = num_hidden_layers + self.num_attention_heads = num_attention_heads + self.intermediate_size = intermediate_size + self.hidden_act = hidden_act + self.hidden_dropout_prob = hidden_dropout_prob + self.attention_probs_dropout_prob = attention_probs_dropout_prob + self.max_position_embeddings = max_position_embeddings + self.type_vocab_size = type_vocab_size + self.initializer_range = initializer_range + self.scope = scope + + def create_model(self): + input_ids = BertModelTest.ids_tensor([self.batch_size, self.seq_length], + self.vocab_size) + + input_mask = None + if self.use_input_mask: + input_mask = BertModelTest.ids_tensor( + [self.batch_size, self.seq_length], vocab_size=2) + + token_type_ids = None + if self.use_token_type_ids: + token_type_ids = BertModelTest.ids_tensor( + [self.batch_size, self.seq_length], self.type_vocab_size) + + config = modeling.BertConfig( + vocab_size=self.vocab_size, + hidden_size=self.hidden_size, + num_hidden_layers=self.num_hidden_layers, + num_attention_heads=self.num_attention_heads, + intermediate_size=self.intermediate_size, + hidden_act=self.hidden_act, + hidden_dropout_prob=self.hidden_dropout_prob, + attention_probs_dropout_prob=self.attention_probs_dropout_prob, + max_position_embeddings=self.max_position_embeddings, + type_vocab_size=self.type_vocab_size, + initializer_range=self.initializer_range) + + model = modeling.BertModel( + config=config, + is_training=self.is_training, + input_ids=input_ids, + input_mask=input_mask, + token_type_ids=token_type_ids, + scope=self.scope) + + outputs = { + "embedding_output": model.get_embedding_output(), + "sequence_output": model.get_sequence_output(), + "pooled_output": model.get_pooled_output(), + "all_encoder_layers": model.get_all_encoder_layers(), + } + return outputs + + def check_output(self, result): + self.parent.assertAllEqual( + result["embedding_output"].shape, + [self.batch_size, self.seq_length, self.hidden_size]) + + self.parent.assertAllEqual( + result["sequence_output"].shape, + [self.batch_size, self.seq_length, self.hidden_size]) + + self.parent.assertAllEqual(result["pooled_output"].shape, + [self.batch_size, self.hidden_size]) + + def test_default(self): + self.run_tester(BertModelTest.BertModelTester(self)) + + def test_config_to_json_string(self): + config = modeling.BertConfig(vocab_size=99, hidden_size=37) + obj = json.loads(config.to_json_string()) + self.assertEqual(obj["vocab_size"], 99) + self.assertEqual(obj["hidden_size"], 37) + + def run_tester(self, tester): + with self.test_session() as sess: + ops = tester.create_model() + init_op = tf.group(tf.global_variables_initializer(), + tf.local_variables_initializer()) + sess.run(init_op) + output_result = sess.run(ops) + tester.check_output(output_result) + + self.assert_all_tensors_reachable(sess, [init_op, ops]) + + @classmethod + def ids_tensor(cls, shape, vocab_size, rng=None, name=None): + """Creates a random int32 tensor of the shape within the vocab size.""" + if rng is None: + rng = random.Random() + + total_dims = 1 + for dim in shape: + total_dims *= dim + + values = [] + for _ in range(total_dims): + values.append(rng.randint(0, vocab_size - 1)) + + return tf.constant(value=values, dtype=tf.int32, shape=shape, name=name) + + def assert_all_tensors_reachable(self, sess, outputs): + """Checks that all the tensors in the graph are reachable from outputs.""" + graph = sess.graph + + ignore_strings = [ + "^.*/assert_less_equal/.*$", + "^.*/dilation_rate$", + "^.*/Tensordot/concat$", + "^.*/Tensordot/concat/axis$", + "^testing/.*$", + ] + + ignore_regexes = [re.compile(x) for x in ignore_strings] + + unreachable = self.get_unreachable_ops(graph, outputs) + filtered_unreachable = [] + for x in unreachable: + do_ignore = False + for r in ignore_regexes: + m = r.match(x.name) + if m is not None: + do_ignore = True + if do_ignore: + continue + filtered_unreachable.append(x) + unreachable = filtered_unreachable + + self.assertEqual( + len(unreachable), 0, "The following ops are unreachable: %s" % + (" ".join([x.name for x in unreachable]))) + + @classmethod + def get_unreachable_ops(cls, graph, outputs): + """Finds all of the tensors in graph that are unreachable from outputs.""" + outputs = cls.flatten_recursive(outputs) + output_to_op = collections.defaultdict(list) + op_to_all = collections.defaultdict(list) + assign_out_to_in = collections.defaultdict(list) + + for op in graph.get_operations(): + for x in op.inputs: + op_to_all[op.name].append(x.name) + for y in op.outputs: + output_to_op[y.name].append(op.name) + op_to_all[op.name].append(y.name) + if str(op.type) == "Assign": + for y in op.outputs: + for x in op.inputs: + assign_out_to_in[y.name].append(x.name) + + assign_groups = collections.defaultdict(list) + for out_name in assign_out_to_in.keys(): + name_group = assign_out_to_in[out_name] + for n1 in name_group: + assign_groups[n1].append(out_name) + for n2 in name_group: + if n1 != n2: + assign_groups[n1].append(n2) + + seen_tensors = {} + stack = [x.name for x in outputs] + while stack: + name = stack.pop() + if name in seen_tensors: + continue + seen_tensors[name] = True + + if name in output_to_op: + for op_name in output_to_op[name]: + if op_name in op_to_all: + for input_name in op_to_all[op_name]: + if input_name not in stack: + stack.append(input_name) + + expanded_names = [] + if name in assign_groups: + for assign_name in assign_groups[name]: + expanded_names.append(assign_name) + + for expanded_name in expanded_names: + if expanded_name not in stack: + stack.append(expanded_name) + + unreachable_ops = [] + for op in graph.get_operations(): + is_unreachable = False + all_names = [x.name for x in op.inputs] + [x.name for x in op.outputs] + for name in all_names: + if name not in seen_tensors: + is_unreachable = True + if is_unreachable: + unreachable_ops.append(op) + return unreachable_ops + + @classmethod + def flatten_recursive(cls, item): + """Flattens (potentially nested) a tuple/dictionary/list to a list.""" + output = [] + if isinstance(item, list): + output.extend(item) + elif isinstance(item, tuple): + output.extend(list(item)) + elif isinstance(item, dict): + for (_, v) in six.iteritems(item): + output.append(v) + else: + return [item] + + flat_output = [] + for x in output: + flat_output.extend(cls.flatten_recursive(x)) + return flat_output + + +if __name__ == "__main__": + tf.test.main() diff --git a/optimization.py b/optimization.py new file mode 100644 index 0000000..d33dabd --- /dev/null +++ b/optimization.py @@ -0,0 +1,174 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Functions and classes related to optimization (weight updates).""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import re +import tensorflow as tf + + +def create_optimizer(loss, init_lr, num_train_steps, num_warmup_steps, use_tpu): + """Creates an optimizer training op.""" + global_step = tf.train.get_or_create_global_step() + + learning_rate = tf.constant(value=init_lr, shape=[], dtype=tf.float32) + + # Implements linear decay of the learning rate. + learning_rate = tf.train.polynomial_decay( + learning_rate, + global_step, + num_train_steps, + end_learning_rate=0.0, + power=1.0, + cycle=False) + + # Implements linear warmup. I.e., if global_step < num_warmup_steps, the + # learning rate will be `global_step/num_warmup_steps * init_lr`. + if num_warmup_steps: + global_steps_int = tf.cast(global_step, tf.int32) + warmup_steps_int = tf.constant(num_warmup_steps, dtype=tf.int32) + + global_steps_float = tf.cast(global_steps_int, tf.float32) + warmup_steps_float = tf.cast(warmup_steps_int, tf.float32) + + warmup_percent_done = global_steps_float / warmup_steps_float + warmup_learning_rate = init_lr * warmup_percent_done + + is_warmup = tf.cast(global_steps_int < warmup_steps_int, tf.float32) + learning_rate = ( + (1.0 - is_warmup) * learning_rate + is_warmup * warmup_learning_rate) + + # It is recommended that you use this optimizer for fine tuning, since this + # is how the model was trained (note that the Adam m/v variables are NOT + # loaded from init_checkpoint.) + optimizer = AdamWeightDecayOptimizer( + learning_rate=learning_rate, + weight_decay_rate=0.01, + beta_1=0.9, + beta_2=0.999, + epsilon=1e-6, + exclude_from_weight_decay=["LayerNorm", "layer_norm", "bias"]) + + if use_tpu: + optimizer = tf.contrib.tpu.CrossShardOptimizer(optimizer) + + tvars = tf.trainable_variables() + grads = tf.gradients(loss, tvars) + + # This is how the model was pre-trained. + (grads, _) = tf.clip_by_global_norm(grads, clip_norm=1.0) + + train_op = optimizer.apply_gradients( + zip(grads, tvars), global_step=global_step) + + # Normally the global step update is done inside of `apply_gradients`. + # However, `AdamWeightDecayOptimizer` doesn't do this. But if you use + # a different optimizer, you should probably take this line out. + new_global_step = global_step + 1 + train_op = tf.group(train_op, [global_step.assign(new_global_step)]) + return train_op + + +class AdamWeightDecayOptimizer(tf.train.Optimizer): + """A basic Adam optimizer that includes "correct" L2 weight decay.""" + + def __init__(self, + learning_rate, + weight_decay_rate=0.0, + beta_1=0.9, + beta_2=0.999, + epsilon=1e-6, + exclude_from_weight_decay=None, + name="AdamWeightDecayOptimizer"): + """Constructs a AdamWeightDecayOptimizer.""" + super(AdamWeightDecayOptimizer, self).__init__(False, name) + + self.learning_rate = learning_rate + self.weight_decay_rate = weight_decay_rate + self.beta_1 = beta_1 + self.beta_2 = beta_2 + self.epsilon = epsilon + self.exclude_from_weight_decay = exclude_from_weight_decay + + def apply_gradients(self, grads_and_vars, global_step=None, name=None): + """See base class.""" + assignments = [] + for (grad, param) in grads_and_vars: + if grad is None or param is None: + continue + + param_name = self._get_variable_name(param.name) + + m = tf.get_variable( + name=param_name + "/adam_m", + shape=param.shape.as_list(), + dtype=tf.float32, + trainable=False, + initializer=tf.zeros_initializer()) + v = tf.get_variable( + name=param_name + "/adam_v", + shape=param.shape.as_list(), + dtype=tf.float32, + trainable=False, + initializer=tf.zeros_initializer()) + + # Standard Adam update. + next_m = ( + tf.multiply(self.beta_1, m) + tf.multiply(1.0 - self.beta_1, grad)) + next_v = ( + tf.multiply(self.beta_2, v) + tf.multiply(1.0 - self.beta_2, + tf.square(grad))) + + update = next_m / (tf.sqrt(next_v) + self.epsilon) + + # Just adding the square of the weights to the loss function is *not* + # the correct way of using L2 regularization/weight decay with Adam, + # since that will interact with the m and v parameters in strange ways. + # + # Instead we want ot decay the weights in a manner that doesn't interact + # with the m/v parameters. This is equivalent to adding the square + # of the weights to the loss with plain (non-momentum) SGD. + if self._do_use_weight_decay(param_name): + update += self.weight_decay_rate * param + + update_with_lr = self.learning_rate * update + + next_param = param - update_with_lr + + assignments.extend( + [param.assign(next_param), + m.assign(next_m), + v.assign(next_v)]) + return tf.group(*assignments, name=name) + + def _do_use_weight_decay(self, param_name): + """Whether to use L2 weight decay for `param_name`.""" + if not self.weight_decay_rate: + return False + if self.exclude_from_weight_decay: + for r in self.exclude_from_weight_decay: + if re.search(r, param_name) is not None: + return False + return True + + def _get_variable_name(self, param_name): + """Get the variable name from the tensor name.""" + m = re.match("^(.*):\\d+$", param_name) + if m is not None: + param_name = m.group(1) + return param_name diff --git a/optimization_test.py b/optimization_test.py new file mode 100644 index 0000000..4f2dcf1 --- /dev/null +++ b/optimization_test.py @@ -0,0 +1,48 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import optimization +import tensorflow as tf + + +class OptimizationTest(tf.test.TestCase): + + def test_adam(self): + with self.test_session() as sess: + w = tf.get_variable( + "w", + shape=[3], + initializer=tf.constant_initializer([0.1, -0.2, -0.1])) + x = tf.constant([0.4, 0.2, -0.5]) + loss = tf.reduce_mean(tf.square(x - w)) + tvars = tf.trainable_variables() + grads = tf.gradients(loss, tvars) + global_step = tf.train.get_or_create_global_step() + optimizer = optimization.AdamWeightDecayOptimizer(learning_rate=0.2) + train_op = optimizer.apply_gradients(zip(grads, tvars), global_step) + init_op = tf.group(tf.global_variables_initializer(), + tf.local_variables_initializer()) + sess.run(init_op) + for _ in range(100): + sess.run(train_op) + w_np = sess.run(w) + self.assertAllClose(w_np.flat, [0.4, 0.2, -0.5], rtol=1e-2, atol=1e-2) + + +if __name__ == "__main__": + tf.test.main() diff --git a/predicting_movie_reviews_with_bert_on_tf_hub.ipynb b/predicting_movie_reviews_with_bert_on_tf_hub.ipynb new file mode 100644 index 0000000..466857f --- /dev/null +++ b/predicting_movie_reviews_with_bert_on_tf_hub.ipynb @@ -0,0 +1,1231 @@ +{ + "nbformat": 4, + "nbformat_minor": 0, + "metadata": { + "colab": { + "name": "Predicting Movie Reviews with BERT on TF Hub.ipynb", + "version": "0.3.2", + "provenance": [], + "collapsed_sections": [] + }, + "kernelspec": { + "name": "python3", + "display_name": "Python 3" + }, + "accelerator": "GPU" + }, + "cells": [ + { + "metadata": { + "id": "j0a4mTk9o1Qg", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Copyright 2019 Google Inc.\n", + "\n", + "# Licensed under the Apache License, Version 2.0 (the \"License\");\n", + "# you may not use this file except in compliance with the License.\n", + "# You may obtain a copy of the License at\n", + "\n", + "# http://www.apache.org/licenses/LICENSE-2.0\n", + "\n", + "# Unless required by applicable law or agreed to in writing, software\n", + "# distributed under the License is distributed on an \"AS IS\" BASIS,\n", + "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n", + "# See the License for the specific language governing permissions and\n", + "# limitations under the License." + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "dCpvgG0vwXAZ", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "#Predicting Movie Review Sentiment with BERT on TF Hub" + ] + }, + { + "metadata": { + "id": "xiYrZKaHwV81", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "If you’ve been following Natural Language Processing over the past year, you’ve probably heard of BERT: Bidirectional Encoder Representations from Transformers. It’s a neural network architecture designed by Google researchers that’s totally transformed what’s state-of-the-art for NLP tasks, like text classification, translation, summarization, and question answering.\n", + "\n", + "Now that BERT's been added to [TF Hub](https://www.tensorflow.org/hub) as a loadable module, it's easy(ish) to add into existing Tensorflow text pipelines. In an existing pipeline, BERT can replace text embedding layers like ELMO and GloVE. Alternatively, [finetuning](http://wiki.fast.ai/index.php/Fine_tuning) BERT can provide both an accuracy boost and faster training time in many cases.\n", + "\n", + "Here, we'll train a model to predict whether an IMDB movie review is positive or negative using BERT in Tensorflow with tf hub. Some code was adapted from [this colab notebook](https://colab.sandbox.google.com/github/tensorflow/tpu/blob/master/tools/colab/bert_finetuning_with_cloud_tpus.ipynb). Let's get started!" + ] + }, + { + "metadata": { + "id": "hsZvic2YxnTz", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "from sklearn.model_selection import train_test_split\n", + "import pandas as pd\n", + "import tensorflow as tf\n", + "import tensorflow_hub as hub\n", + "from datetime import datetime" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "cp5wfXDx5SPH", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "In addition to the standard libraries we imported above, we'll need to install BERT's python package." + ] + }, + { + "metadata": { + "id": "jviywGyWyKsA", + "colab_type": "code", + "outputId": "166f3005-d219-404f-b201-2a0b75480360", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 51 + } + }, + "cell_type": "code", + "source": [ + "!pip install bert-tensorflow" + ], + "execution_count": 38, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Requirement already satisfied: bert-tensorflow in /usr/local/lib/python3.6/dist-packages (1.0.1)\n", + "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from bert-tensorflow) (1.11.0)\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "hhbGEfwgdEtw", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "import bert\n", + "from bert import run_classifier\n", + "from bert import optimization\n", + "from bert import tokenization" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "KVB3eOcjxxm1", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Below, we'll set an output directory location to store our model output and checkpoints. This can be a local directory, in which case you'd set OUTPUT_DIR to the name of the directory you'd like to create. If you're running this code in Google's hosted Colab, the directory won't persist after the Colab session ends.\n", + "\n", + "Alternatively, if you're a GCP user, you can store output in a GCP bucket. To do that, set a directory name in OUTPUT_DIR and the name of the GCP bucket in the BUCKET field.\n", + "\n", + "Set DO_DELETE to rewrite the OUTPUT_DIR if it exists. Otherwise, Tensorflow will load existing model checkpoints from that directory (if they exist)." + ] + }, + { + "metadata": { + "id": "US_EAnICvP7f", + "colab_type": "code", + "outputId": "7780a032-31d4-4794-e6aa-664a5d2ae7dd", + "cellView": "form", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 34 + } + }, + "cell_type": "code", + "source": [ + "# Set the output directory for saving model file\n", + "# Optionally, set a GCP bucket location\n", + "\n", + "OUTPUT_DIR = 'OUTPUT_DIR_NAME'#@param {type:\"string\"}\n", + "#@markdown Whether or not to clear/delete the directory and create a new one\n", + "DO_DELETE = False #@param {type:\"boolean\"}\n", + "#@markdown Set USE_BUCKET and BUCKET if you want to (optionally) store model output on GCP bucket.\n", + "USE_BUCKET = True #@param {type:\"boolean\"}\n", + "BUCKET = 'BUCKET_NAME' #@param {type:\"string\"}\n", + "\n", + "if USE_BUCKET:\n", + " OUTPUT_DIR = 'gs://{}/{}'.format(BUCKET, OUTPUT_DIR)\n", + " from google.colab import auth\n", + " auth.authenticate_user()\n", + "\n", + "if DO_DELETE:\n", + " try:\n", + " tf.gfile.DeleteRecursively(OUTPUT_DIR)\n", + " except:\n", + " # Doesn't matter if the directory didn't exist\n", + " pass\n", + "tf.gfile.MakeDirs(OUTPUT_DIR)\n", + "print('***** Model output directory: {} *****'.format(OUTPUT_DIR))\n" + ], + "execution_count": 40, + "outputs": [ + { + "output_type": "stream", + "text": [ + "***** Model output directory: gs://bert-tfhub/aclImdb_v1 *****\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "pmFYvkylMwXn", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "#Data" + ] + }, + { + "metadata": { + "id": "MC_w8SRqN0fr", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "First, let's download the dataset, hosted by Stanford. The code below, which downloads, extracts, and imports the IMDB Large Movie Review Dataset, is borrowed from [this Tensorflow tutorial](https://www.tensorflow.org/hub/tutorials/text_classification_with_tf_hub)." + ] + }, + { + "metadata": { + "id": "fom_ff20gyy6", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "from tensorflow import keras\n", + "import os\n", + "import re\n", + "\n", + "# Load all files from a directory in a DataFrame.\n", + "def load_directory_data(directory):\n", + " data = {}\n", + " data[\"sentence\"] = []\n", + " data[\"sentiment\"] = []\n", + " for file_path in os.listdir(directory):\n", + " with tf.gfile.GFile(os.path.join(directory, file_path), \"r\") as f:\n", + " data[\"sentence\"].append(f.read())\n", + " data[\"sentiment\"].append(re.match(\"\\d+_(\\d+)\\.txt\", file_path).group(1))\n", + " return pd.DataFrame.from_dict(data)\n", + "\n", + "# Merge positive and negative examples, add a polarity column and shuffle.\n", + "def load_dataset(directory):\n", + " pos_df = load_directory_data(os.path.join(directory, \"pos\"))\n", + " neg_df = load_directory_data(os.path.join(directory, \"neg\"))\n", + " pos_df[\"polarity\"] = 1\n", + " neg_df[\"polarity\"] = 0\n", + " return pd.concat([pos_df, neg_df]).sample(frac=1).reset_index(drop=True)\n", + "\n", + "# Download and process the dataset files.\n", + "def download_and_load_datasets(force_download=False):\n", + " dataset = tf.keras.utils.get_file(\n", + " fname=\"aclImdb.tar.gz\", \n", + " origin=\"http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz\", \n", + " extract=True)\n", + " \n", + " train_df = load_dataset(os.path.join(os.path.dirname(dataset), \n", + " \"aclImdb\", \"train\"))\n", + " test_df = load_dataset(os.path.join(os.path.dirname(dataset), \n", + " \"aclImdb\", \"test\"))\n", + " \n", + " return train_df, test_df\n" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "2abfwdn-g135", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "train, test = download_and_load_datasets()" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "XA8WHJgzhIZf", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "To keep training fast, we'll take a sample of 5000 train and test examples, respectively." + ] + }, + { + "metadata": { + "id": "lw_F488eixTV", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "train = train.sample(5000)\n", + "test = test.sample(5000)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "prRQM8pDi8xI", + "colab_type": "code", + "outputId": "34445cb8-2be0-4379-fdbc-7794091f6049", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 34 + } + }, + "cell_type": "code", + "source": [ + "train.columns" + ], + "execution_count": 44, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "Index(['sentence', 'sentiment', 'polarity'], dtype='object')" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 44 + } + ] + }, + { + "metadata": { + "id": "sfRnHSz3iSXz", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "For us, our input data is the 'sentence' column and our label is the 'polarity' column (0, 1 for negative and positive, respecitvely)" + ] + }, + { + "metadata": { + "id": "IuMOGwFui4it", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "DATA_COLUMN = 'sentence'\n", + "LABEL_COLUMN = 'polarity'\n", + "# label_list is the list of labels, i.e. True, False or 0, 1 or 'dog', 'cat'\n", + "label_list = [0, 1]" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "V399W0rqNJ-Z", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "#Data Preprocessing\n", + "We'll need to transform our data into a format BERT understands. This involves two steps. First, we create `InputExample`'s using the constructor provided in the BERT library.\n", + "\n", + "- `text_a` is the text we want to classify, which in this case, is the `Request` field in our Dataframe. \n", + "- `text_b` is used if we're training a model to understand the relationship between sentences (i.e. is `text_b` a translation of `text_a`? Is `text_b` an answer to the question asked by `text_a`?). This doesn't apply to our task, so we can leave `text_b` blank.\n", + "- `label` is the label for our example, i.e. True, False" + ] + }, + { + "metadata": { + "id": "p9gEt5SmM6i6", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Use the InputExample class from BERT's run_classifier code to create examples from the data\n", + "train_InputExamples = train.apply(lambda x: bert.run_classifier.InputExample(guid=None, # Globally unique ID for bookkeeping, unused in this example\n", + " text_a = x[DATA_COLUMN], \n", + " text_b = None, \n", + " label = x[LABEL_COLUMN]), axis = 1)\n", + "\n", + "test_InputExamples = test.apply(lambda x: bert.run_classifier.InputExample(guid=None, \n", + " text_a = x[DATA_COLUMN], \n", + " text_b = None, \n", + " label = x[LABEL_COLUMN]), axis = 1)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "SCZWZtKxObjh", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Next, we need to preprocess our data so that it matches the data BERT was trained on. For this, we'll need to do a couple of things (but don't worry--this is also included in the Python library):\n", + "\n", + "\n", + "1. Lowercase our text (if we're using a BERT lowercase model)\n", + "2. Tokenize it (i.e. \"sally says hi\" -> [\"sally\", \"says\", \"hi\"])\n", + "3. Break words into WordPieces (i.e. \"calling\" -> [\"call\", \"##ing\"])\n", + "4. Map our words to indexes using a vocab file that BERT provides\n", + "5. Add special \"CLS\" and \"SEP\" tokens (see the [readme](https://github.com/google-research/bert))\n", + "6. Append \"index\" and \"segment\" tokens to each input (see the [BERT paper](https://arxiv.org/pdf/1810.04805.pdf))\n", + "\n", + "Happily, we don't have to worry about most of these details.\n", + "\n", + "\n" + ] + }, + { + "metadata": { + "id": "qMWiDtpyQSoU", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "To start, we'll need to load a vocabulary file and lowercasing information directly from the BERT tf hub module:" + ] + }, + { + "metadata": { + "id": "IhJSe0QHNG7U", + "colab_type": "code", + "outputId": "20b28cc7-3cb3-4ce6-bfff-a7847ce3bbaa", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 34 + } + }, + "cell_type": "code", + "source": [ + "# This is a path to an uncased (all lowercase) version of BERT\n", + "BERT_MODEL_HUB = \"https://tfhub.dev/google/bert_uncased_L-12_H-768_A-12/1\"\n", + "\n", + "def create_tokenizer_from_hub_module():\n", + " \"\"\"Get the vocab file and casing info from the Hub module.\"\"\"\n", + " with tf.Graph().as_default():\n", + " bert_module = hub.Module(BERT_MODEL_HUB)\n", + " tokenization_info = bert_module(signature=\"tokenization_info\", as_dict=True)\n", + " with tf.Session() as sess:\n", + " vocab_file, do_lower_case = sess.run([tokenization_info[\"vocab_file\"],\n", + " tokenization_info[\"do_lower_case\"]])\n", + " \n", + " return bert.tokenization.FullTokenizer(\n", + " vocab_file=vocab_file, do_lower_case=do_lower_case)\n", + "\n", + "tokenizer = create_tokenizer_from_hub_module()" + ], + "execution_count": 47, + "outputs": [ + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "z4oFkhpZBDKm", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Great--we just learned that the BERT model we're using expects lowercase data (that's what stored in tokenization_info[\"do_lower_case\"]) and we also loaded BERT's vocab file. We also created a tokenizer, which breaks words into word pieces:" + ] + }, + { + "metadata": { + "id": "dsBo6RCtQmwx", + "colab_type": "code", + "outputId": "9af8c917-90ec-4fe9-897b-79dc89ca88e1", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 221 + } + }, + "cell_type": "code", + "source": [ + "tokenizer.tokenize(\"This here's an example of using the BERT tokenizer\")" + ], + "execution_count": 48, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "['this',\n", + " 'here',\n", + " \"'\",\n", + " 's',\n", + " 'an',\n", + " 'example',\n", + " 'of',\n", + " 'using',\n", + " 'the',\n", + " 'bert',\n", + " 'token',\n", + " '##izer']" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 48 + } + ] + }, + { + "metadata": { + "id": "0OEzfFIt6GIc", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Using our tokenizer, we'll call `run_classifier.convert_examples_to_features` on our InputExamples to convert them into features BERT understands." + ] + }, + { + "metadata": { + "id": "LL5W8gEGRTAf", + "colab_type": "code", + "outputId": "65001dda-155b-48fc-b5fc-1e4cabc8dfbf", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1261 + } + }, + "cell_type": "code", + "source": [ + "# We'll set sequences to be at most 128 tokens long.\n", + "MAX_SEQ_LENGTH = 128\n", + "# Convert our train and test features to InputFeatures that BERT understands.\n", + "train_features = bert.run_classifier.convert_examples_to_features(train_InputExamples, label_list, MAX_SEQ_LENGTH, tokenizer)\n", + "test_features = bert.run_classifier.convert_examples_to_features(test_InputExamples, label_list, MAX_SEQ_LENGTH, tokenizer)" + ], + "execution_count": 49, + "outputs": [ + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Writing example 0 of 5000\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] i ' m watching this on the sci - fi channel right now . it ' s so horrible i can ' t stop watching it ! i ' m a video ##grapher and this movie makes me sad . i feel bad for anyone associated with this movie . some of the camera work is good . most is very questionable . there are a few decent actors in the flick . too bad they ' re surrounded by what must have been the director ' s relatives . that ' s the only way they could have been qualified to be in a movie ! music was a little better than the acting . if you get around to watching this i hope it [SEP]\n", + "INFO:tensorflow:input_ids: 101 1045 1005 1049 3666 2023 2006 1996 16596 1011 10882 3149 2157 2085 1012 2009 1005 1055 2061 9202 1045 2064 1005 1056 2644 3666 2009 999 1045 1005 1049 1037 2678 18657 1998 2023 3185 3084 2033 6517 1012 1045 2514 2919 2005 3087 3378 2007 2023 3185 1012 2070 1997 1996 4950 2147 2003 2204 1012 2087 2003 2200 21068 1012 2045 2024 1037 2261 11519 5889 1999 1996 17312 1012 2205 2919 2027 1005 2128 5129 2011 2054 2442 2031 2042 1996 2472 1005 1055 9064 1012 2008 1005 1055 1996 2069 2126 2027 2071 2031 2042 4591 2000 2022 1999 1037 3185 999 2189 2001 1037 2210 2488 2084 1996 3772 1012 2065 2017 2131 2105 2000 3666 2023 1045 3246 2009 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] i have been a fan of pushing dai ##sies since the very beginning . it is wonderful ##ly thought up , and bryan fuller has the most remarkable ideas for this show . < br / > < br / > it is unbelievable on how much tv has been needing a creative , original show like pushing dai ##sies . it is a huge relief to see a show , that is unlike the rest , where as , if you compared it to some of the newer shows , such as scrub ##s and house , you would see the similarities , and it does get ted ##ious at moments to see shows so close in identity . < br / > < br [SEP]\n", + "INFO:tensorflow:input_ids: 101 1045 2031 2042 1037 5470 1997 6183 18765 14625 2144 1996 2200 2927 1012 2009 2003 6919 2135 2245 2039 1010 1998 8527 12548 2038 1996 2087 9487 4784 2005 2023 2265 1012 1026 7987 1013 1028 1026 7987 1013 1028 2009 2003 23653 2006 2129 2172 2694 2038 2042 11303 1037 5541 1010 2434 2265 2066 6183 18765 14625 1012 2009 2003 1037 4121 4335 2000 2156 1037 2265 1010 2008 2003 4406 1996 2717 1010 2073 2004 1010 2065 2017 4102 2009 2000 2070 1997 1996 10947 3065 1010 2107 2004 18157 2015 1998 2160 1010 2017 2052 2156 1996 12319 1010 1998 2009 2515 2131 6945 6313 2012 5312 2000 2156 3065 2061 2485 1999 4767 1012 1026 7987 1013 1028 1026 7987 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 1 (id = 1)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] this movie starts out promising ##ly , with an early scene in which frank morgan advises against gary cooper ' s marriage to his daughter , anita louise . frank morgan , playing an una ##bas ##hed gold - digger , loudly complain ##s to cooper about his perceived pen ##ury at the hands of his family - including his daughter , anita louise . i am a fan of all 3 actors . frank morgan is ( to my mind ) a hollywood treasure , cooper a legend , and louise a very lovely , versatile and under - appreciated actress seldom seen in the leading role . i also have nothing against teresa wright , and while not blessed with great range , she [SEP]\n", + "INFO:tensorflow:input_ids: 101 2023 3185 4627 2041 10015 2135 1010 2007 2019 2220 3496 1999 2029 3581 5253 25453 2114 5639 6201 1005 1055 3510 2000 2010 2684 1010 12918 8227 1012 3581 5253 1010 2652 2019 14477 22083 9072 2751 1011 28661 1010 9928 17612 2015 2000 6201 2055 2010 8690 7279 13098 2012 1996 2398 1997 2010 2155 1011 2164 2010 2684 1010 12918 8227 1012 1045 2572 1037 5470 1997 2035 1017 5889 1012 3581 5253 2003 1006 2000 2026 2568 1007 1037 5365 8813 1010 6201 1037 5722 1010 1998 8227 1037 2200 8403 1010 22979 1998 2104 1011 12315 3883 15839 2464 1999 1996 2877 2535 1012 1045 2036 2031 2498 2114 12409 6119 1010 1998 2096 2025 10190 2007 2307 2846 1010 2016 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] i was over ##taken by the emotion . un ##for ##get ##table rendering of a wartime story which is unknown to most people . the performances were fault ##less and outstanding . [SEP]\n", + "INFO:tensorflow:input_ids: 101 1045 2001 2058 25310 2011 1996 7603 1012 4895 29278 18150 10880 14259 1997 1037 12498 2466 2029 2003 4242 2000 2087 2111 1012 1996 4616 2020 6346 3238 1998 5151 1012 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 1 (id = 1)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] soldier blue is a movie with pre ##tension ##s : pre ##tension ##s to be some sort of profound statement on man ' s inhuman ##ity to man , on the white man ' s exploitation of and brutality towards indigenous peoples ; a biting , un ##fl ##in ##ching and sar ##don ##ic commentary on the horrors of vietnam . well , sorry , but it fails mis ##era ##bly to be any of those things . what soldier blue actually is is per ##nic ##ious , tri ##te , badly made , dish ##ones ##t rubbish . < br / > < br / > another reviewer here hit the nail on the head in saying that it appears to be a hybrid of [SEP]\n", + "INFO:tensorflow:input_ids: 101 5268 2630 2003 1037 3185 2007 3653 29048 2015 1024 3653 29048 2015 2000 2022 2070 4066 1997 13769 4861 2006 2158 1005 1055 29582 3012 2000 2158 1010 2006 1996 2317 2158 1005 1055 14427 1997 1998 24083 2875 6284 7243 1025 1037 12344 1010 4895 10258 2378 8450 1998 18906 5280 2594 8570 2006 1996 22812 1997 5148 1012 2092 1010 3374 1010 2021 2009 11896 28616 6906 6321 2000 2022 2151 1997 2216 2477 1012 2054 5268 2630 2941 2003 2003 2566 8713 6313 1010 13012 2618 1010 6649 2081 1010 9841 21821 2102 29132 1012 1026 7987 1013 1028 1026 7987 1013 1028 2178 12027 2182 2718 1996 13774 2006 1996 2132 1999 3038 2008 2009 3544 2000 2022 1037 8893 1997 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:Writing example 0 of 5000\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] i just watched this today on tv . it was on abc ' s sunday afternoon movie . < br / > < br / > this wasn ' t a very good movie , but for a low budget independent film like this , it was okay . there is some suspense in it , but there are so many bad qualities that really bring the movie down . the script is pretty lame , and the plot elements aren ' t very realistic , such as the way a 911 operator would laugh and hang up when someone is reporting a murder . i don ' t know what the writer was thinking when they came up with that idea , but it isn [SEP]\n", + "INFO:tensorflow:input_ids: 101 1045 2074 3427 2023 2651 2006 2694 1012 2009 2001 2006 5925 1005 1055 4465 5027 3185 1012 1026 7987 1013 1028 1026 7987 1013 1028 2023 2347 1005 1056 1037 2200 2204 3185 1010 2021 2005 1037 2659 5166 2981 2143 2066 2023 1010 2009 2001 3100 1012 2045 2003 2070 23873 1999 2009 1010 2021 2045 2024 2061 2116 2919 11647 2008 2428 3288 1996 3185 2091 1012 1996 5896 2003 3492 20342 1010 1998 1996 5436 3787 4995 1005 1056 2200 12689 1010 2107 2004 1996 2126 1037 19989 6872 2052 4756 1998 6865 2039 2043 2619 2003 7316 1037 4028 1012 1045 2123 1005 1056 2113 2054 1996 3213 2001 3241 2043 2027 2234 2039 2007 2008 2801 1010 2021 2009 3475 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] from hardly alien sounding lasers , to an elementary school style shuttle crash , \" night ##be ##ast \" is better classified as a far ##cic ##al mix of fake blood and bare chest . the almost pornographic style of the film seems to be a failed attempt to recover from a lack of co ##hesive or effective story . the acting however is not nearly as beast ##ly , many of the young , aspiring , actors ad ##mir ##ably showcase a hidden talent . particularly don lei ##fer ##t and jamie ze ##mare ##l , who shed a well needed sha ##rd of light on this otherwise terrible film . night ##be ##ast would have never shown up on set had he known the [SEP]\n", + "INFO:tensorflow:input_ids: 101 2013 6684 7344 9391 23965 1010 2000 2019 4732 2082 2806 10382 5823 1010 1000 2305 4783 14083 1000 2003 2488 6219 2004 1037 2521 19053 2389 4666 1997 8275 2668 1998 6436 3108 1012 1996 2471 26932 2806 1997 1996 2143 3849 2000 2022 1037 3478 3535 2000 8980 2013 1037 3768 1997 2522 21579 2030 4621 2466 1012 1996 3772 2174 2003 2025 3053 2004 6841 2135 1010 2116 1997 1996 2402 1010 22344 1010 5889 4748 14503 8231 13398 1037 5023 5848 1012 3391 2123 26947 7512 2102 1998 6175 27838 24376 2140 1010 2040 8328 1037 2092 2734 21146 4103 1997 2422 2006 2023 4728 6659 2143 1012 2305 4783 14083 2052 2031 2196 3491 2039 2006 2275 2018 2002 2124 1996 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] here we have the in ##imi ##table charlie chaplin for ##sa ##king his slap ##stick past to tackle the serious subject of anti - semi ##tism , and into ##ler ##ance in general . he portrays two characters - the sweet , innocent jewish barber - a war veteran , and the ravi ##ng and ruthless dictator , aden ##oid h ##yn ##kel . the jewish ghetto in this country is not safe for long , due to the w ##him ##s of h ##yn ##kel and his armed thugs , who routinely rough up its residents , or leave them alone , dependent upon his mood that day or week . the barber is among them , but is befriended by his former commanding officer [SEP]\n", + "INFO:tensorflow:input_ids: 101 2182 2057 2031 1996 1999 27605 10880 4918 23331 2005 3736 6834 2010 14308 21354 2627 2000 11147 1996 3809 3395 1997 3424 1011 4100 17456 1010 1998 2046 3917 6651 1999 2236 1012 2002 17509 2048 3494 1011 1996 4086 1010 7036 3644 13362 1011 1037 2162 8003 1010 1998 1996 16806 3070 1998 18101 21237 1010 16298 9314 1044 6038 11705 1012 1996 3644 17276 1999 2023 2406 2003 2025 3647 2005 2146 1010 2349 2000 1996 1059 14341 2015 1997 1044 6038 11705 1998 2010 4273 24106 1010 2040 19974 5931 2039 2049 3901 1010 2030 2681 2068 2894 1010 7790 2588 2010 6888 2008 2154 2030 2733 1012 1996 13362 2003 2426 2068 1010 2021 2003 23386 2011 2010 2280 7991 2961 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 1 (id = 1)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] i really hated this movie and it ' s the first movie written by stephen king that i didn ' t finish . i was truly disappointed , it was the worst crap i ' ve ever seen . what were you thinking making three hours out of it ? it may have a quite good story , but actors ? no . suspense ? no . romance ? no . horror ? no . it didn ' t have anything . < br / > < br / > it ' s got this strange , crazy science man with einstein - hair , the classic thing . not real at all . and a man keep getting younger all the time . it seems [SEP]\n", + "INFO:tensorflow:input_ids: 101 1045 2428 6283 2023 3185 1998 2009 1005 1055 1996 2034 3185 2517 2011 4459 2332 2008 1045 2134 1005 1056 3926 1012 1045 2001 5621 9364 1010 2009 2001 1996 5409 10231 1045 1005 2310 2412 2464 1012 2054 2020 2017 3241 2437 2093 2847 2041 1997 2009 1029 2009 2089 2031 1037 3243 2204 2466 1010 2021 5889 1029 2053 1012 23873 1029 2053 1012 7472 1029 2053 1012 5469 1029 2053 1012 2009 2134 1005 1056 2031 2505 1012 1026 7987 1013 1028 1026 7987 1013 1028 2009 1005 1055 2288 2023 4326 1010 4689 2671 2158 2007 15313 1011 2606 1010 1996 4438 2518 1012 2025 2613 2012 2035 1012 1998 1037 2158 2562 2893 3920 2035 1996 2051 1012 2009 3849 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: None\n", + "INFO:tensorflow:tokens: [CLS] story chinese tall story tells the story of righteous monk trip ##ita ##ka , who , along with his guardians monkey , sandy and pigs ##y make their journey west on a quest to recover ancient sutra ##s , finally , they reach the final leg of their journey in sha ##che city but all is not as it seems when the city is attacked by evil tree demons . monkey tries his best to battle them but is overwhelmed , knowing his master is in grave danger , he uses his trust ##y golden staff to thrust trip ##ita ##ka to safety . < br / > < br / > the monk ends up being knocked out when he land and when he wakes [SEP]\n", + "INFO:tensorflow:input_ids: 101 2466 2822 4206 2466 4136 1996 2466 1997 19556 8284 4440 6590 2912 1010 2040 1010 2247 2007 2010 14240 10608 1010 7525 1998 14695 2100 2191 2037 4990 2225 2006 1037 8795 2000 8980 3418 26567 2015 1010 2633 1010 2027 3362 1996 2345 4190 1997 2037 4990 1999 21146 5403 2103 2021 2035 2003 2025 2004 2009 3849 2043 1996 2103 2003 4457 2011 4763 3392 7942 1012 10608 5363 2010 2190 2000 2645 2068 2021 2003 13394 1010 4209 2010 3040 2003 1999 6542 5473 1010 2002 3594 2010 3404 2100 3585 3095 2000 7400 4440 6590 2912 2000 3808 1012 1026 7987 1013 1028 1026 7987 1013 1028 1996 8284 4515 2039 2108 6573 2041 2043 2002 2455 1998 2043 2002 17507 102\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 1 (id = 1)\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "ccp5trMwRtmr", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "#Creating a model\n", + "\n", + "Now that we've prepared our data, let's focus on building a model. `create_model` does just this below. First, it loads the BERT tf hub module again (this time to extract the computation graph). Next, it creates a single new layer that will be trained to adapt BERT to our sentiment task (i.e. classifying whether a movie review is positive or negative). This strategy of using a mostly trained model is called [fine-tuning](http://wiki.fast.ai/index.php/Fine_tuning)." + ] + }, + { + "metadata": { + "id": "6o2a5ZIvRcJq", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "def create_model(is_predicting, input_ids, input_mask, segment_ids, labels,\n", + " num_labels):\n", + " \"\"\"Creates a classification model.\"\"\"\n", + "\n", + " bert_module = hub.Module(\n", + " BERT_MODEL_HUB,\n", + " trainable=True)\n", + " bert_inputs = dict(\n", + " input_ids=input_ids,\n", + " input_mask=input_mask,\n", + " segment_ids=segment_ids)\n", + " bert_outputs = bert_module(\n", + " inputs=bert_inputs,\n", + " signature=\"tokens\",\n", + " as_dict=True)\n", + "\n", + " # Use \"pooled_output\" for classification tasks on an entire sentence.\n", + " # Use \"sequence_outputs\" for token-level output.\n", + " output_layer = bert_outputs[\"pooled_output\"]\n", + "\n", + " hidden_size = output_layer.shape[-1].value\n", + "\n", + " # Create our own layer to tune for politeness data.\n", + " output_weights = tf.get_variable(\n", + " \"output_weights\", [num_labels, hidden_size],\n", + " initializer=tf.truncated_normal_initializer(stddev=0.02))\n", + "\n", + " output_bias = tf.get_variable(\n", + " \"output_bias\", [num_labels], initializer=tf.zeros_initializer())\n", + "\n", + " with tf.variable_scope(\"loss\"):\n", + "\n", + " # Dropout helps prevent overfitting\n", + " output_layer = tf.nn.dropout(output_layer, keep_prob=0.9)\n", + "\n", + " logits = tf.matmul(output_layer, output_weights, transpose_b=True)\n", + " logits = tf.nn.bias_add(logits, output_bias)\n", + " log_probs = tf.nn.log_softmax(logits, axis=-1)\n", + "\n", + " # Convert labels into one-hot encoding\n", + " one_hot_labels = tf.one_hot(labels, depth=num_labels, dtype=tf.float32)\n", + "\n", + " predicted_labels = tf.squeeze(tf.argmax(log_probs, axis=-1, output_type=tf.int32))\n", + " # If we're predicting, we want predicted labels and the probabiltiies.\n", + " if is_predicting:\n", + " return (predicted_labels, log_probs)\n", + "\n", + " # If we're train/eval, compute loss between predicted and actual label\n", + " per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1)\n", + " loss = tf.reduce_mean(per_example_loss)\n", + " return (loss, predicted_labels, log_probs)\n" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "qpE0ZIDOCQzE", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Next we'll wrap our model function in a `model_fn_builder` function that adapts our model to work for training, evaluation, and prediction." + ] + }, + { + "metadata": { + "id": "FnH-AnOQ9KKW", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# model_fn_builder actually creates our model function\n", + "# using the passed parameters for num_labels, learning_rate, etc.\n", + "def model_fn_builder(num_labels, learning_rate, num_train_steps,\n", + " num_warmup_steps):\n", + " \"\"\"Returns `model_fn` closure for TPUEstimator.\"\"\"\n", + " def model_fn(features, labels, mode, params): # pylint: disable=unused-argument\n", + " \"\"\"The `model_fn` for TPUEstimator.\"\"\"\n", + "\n", + " input_ids = features[\"input_ids\"]\n", + " input_mask = features[\"input_mask\"]\n", + " segment_ids = features[\"segment_ids\"]\n", + " label_ids = features[\"label_ids\"]\n", + "\n", + " is_predicting = (mode == tf.estimator.ModeKeys.PREDICT)\n", + " \n", + " # TRAIN and EVAL\n", + " if not is_predicting:\n", + "\n", + " (loss, predicted_labels, log_probs) = create_model(\n", + " is_predicting, input_ids, input_mask, segment_ids, label_ids, num_labels)\n", + "\n", + " train_op = bert.optimization.create_optimizer(\n", + " loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu=False)\n", + "\n", + " # Calculate evaluation metrics. \n", + " def metric_fn(label_ids, predicted_labels):\n", + " accuracy = tf.metrics.accuracy(label_ids, predicted_labels)\n", + " f1_score = tf.contrib.metrics.f1_score(\n", + " label_ids,\n", + " predicted_labels)\n", + " auc = tf.metrics.auc(\n", + " label_ids,\n", + " predicted_labels)\n", + " recall = tf.metrics.recall(\n", + " label_ids,\n", + " predicted_labels)\n", + " precision = tf.metrics.precision(\n", + " label_ids,\n", + " predicted_labels) \n", + " true_pos = tf.metrics.true_positives(\n", + " label_ids,\n", + " predicted_labels)\n", + " true_neg = tf.metrics.true_negatives(\n", + " label_ids,\n", + " predicted_labels) \n", + " false_pos = tf.metrics.false_positives(\n", + " label_ids,\n", + " predicted_labels) \n", + " false_neg = tf.metrics.false_negatives(\n", + " label_ids,\n", + " predicted_labels)\n", + " return {\n", + " \"eval_accuracy\": accuracy,\n", + " \"f1_score\": f1_score,\n", + " \"auc\": auc,\n", + " \"precision\": precision,\n", + " \"recall\": recall,\n", + " \"true_positives\": true_pos,\n", + " \"true_negatives\": true_neg,\n", + " \"false_positives\": false_pos,\n", + " \"false_negatives\": false_neg\n", + " }\n", + "\n", + " eval_metrics = metric_fn(label_ids, predicted_labels)\n", + "\n", + " if mode == tf.estimator.ModeKeys.TRAIN:\n", + " return tf.estimator.EstimatorSpec(mode=mode,\n", + " loss=loss,\n", + " train_op=train_op)\n", + " else:\n", + " return tf.estimator.EstimatorSpec(mode=mode,\n", + " loss=loss,\n", + " eval_metric_ops=eval_metrics)\n", + " else:\n", + " (predicted_labels, log_probs) = create_model(\n", + " is_predicting, input_ids, input_mask, segment_ids, label_ids, num_labels)\n", + "\n", + " predictions = {\n", + " 'probabilities': log_probs,\n", + " 'labels': predicted_labels\n", + " }\n", + " return tf.estimator.EstimatorSpec(mode, predictions=predictions)\n", + "\n", + " # Return the actual model function in the closure\n", + " return model_fn\n" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "OjwJ4bTeWXD8", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Compute train and warmup steps from batch size\n", + "# These hyperparameters are copied from this colab notebook (https://colab.sandbox.google.com/github/tensorflow/tpu/blob/master/tools/colab/bert_finetuning_with_cloud_tpus.ipynb)\n", + "BATCH_SIZE = 32\n", + "LEARNING_RATE = 2e-5\n", + "NUM_TRAIN_EPOCHS = 3.0\n", + "# Warmup is a period of time where hte learning rate \n", + "# is small and gradually increases--usually helps training.\n", + "WARMUP_PROPORTION = 0.1\n", + "# Model configs\n", + "SAVE_CHECKPOINTS_STEPS = 500\n", + "SAVE_SUMMARY_STEPS = 100" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "emHf9GhfWBZ_", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Compute # train and warmup steps from batch size\n", + "num_train_steps = int(len(train_features) / BATCH_SIZE * NUM_TRAIN_EPOCHS)\n", + "num_warmup_steps = int(num_train_steps * WARMUP_PROPORTION)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "oEJldMr3WYZa", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Specify outpit directory and number of checkpoint steps to save\n", + "run_config = tf.estimator.RunConfig(\n", + " model_dir=OUTPUT_DIR,\n", + " save_summary_steps=SAVE_SUMMARY_STEPS,\n", + " save_checkpoints_steps=SAVE_CHECKPOINTS_STEPS)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "q_WebpS1X97v", + "colab_type": "code", + "outputId": "1648932a-7391-49d3-8af7-52d514e226e8", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 156 + } + }, + "cell_type": "code", + "source": [ + "model_fn = model_fn_builder(\n", + " num_labels=len(label_list),\n", + " learning_rate=LEARNING_RATE,\n", + " num_train_steps=num_train_steps,\n", + " num_warmup_steps=num_warmup_steps)\n", + "\n", + "estimator = tf.estimator.Estimator(\n", + " model_fn=model_fn,\n", + " config=run_config,\n", + " params={\"batch_size\": BATCH_SIZE})\n" + ], + "execution_count": 55, + "outputs": [ + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Using config: {'_model_dir': 'gs://bert-tfhub/aclImdb_v1', '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_steps': 500, '_save_checkpoints_secs': None, '_session_config': allow_soft_placement: true\n", + "graph_options {\n", + " rewrite_options {\n", + " meta_optimizer_iterations: ONE\n", + " }\n", + "}\n", + ", '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_log_step_count_steps': 100, '_train_distribute': None, '_device_fn': None, '_protocol': None, '_eval_distribute': None, '_experimental_distribute': None, '_service': None, '_cluster_spec': , '_task_type': 'worker', '_task_id': 0, '_global_id_in_cluster': 0, '_master': '', '_evaluation_master': '', '_is_chief': True, '_num_ps_replicas': 0, '_num_worker_replicas': 1}\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "NOO3RfG1DYLo", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Next we create an input builder function that takes our training feature set (`train_features`) and produces a generator. This is a pretty standard design pattern for working with Tensorflow [Estimators](https://www.tensorflow.org/guide/estimators)." + ] + }, + { + "metadata": { + "id": "1Pv2bAlOX_-K", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "# Create an input function for training. drop_remainder = True for using TPUs.\n", + "train_input_fn = bert.run_classifier.input_fn_builder(\n", + " features=train_features,\n", + " seq_length=MAX_SEQ_LENGTH,\n", + " is_training=True,\n", + " drop_remainder=False)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "t6Nukby2EB6-", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Now we train our model! For me, using a Colab notebook running on Google's GPUs, my training time was about 14 minutes." + ] + }, + { + "metadata": { + "id": "nucD4gluYJmK", + "colab_type": "code", + "outputId": "5d728e72-4631-42bf-c48d-3f51d4b968ce", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 68 + } + }, + "cell_type": "code", + "source": [ + "print(f'Beginning Training!')\n", + "current_time = datetime.now()\n", + "estimator.train(input_fn=train_input_fn, max_steps=num_train_steps)\n", + "print(\"Training took time \", datetime.now() - current_time)" + ], + "execution_count": 57, + "outputs": [ + { + "output_type": "stream", + "text": [ + "Beginning Training!\n", + "INFO:tensorflow:Skipping training since max_steps has already saved.\n", + "Training took time 0:00:00.759709\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "CmbLTVniARy3", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Now let's use our test data to see how well our model did:" + ] + }, + { + "metadata": { + "id": "JIhejfpyJ8Bx", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "test_input_fn = run_classifier.input_fn_builder(\n", + " features=test_features,\n", + " seq_length=MAX_SEQ_LENGTH,\n", + " is_training=False,\n", + " drop_remainder=False)" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "PPVEXhNjYXC-", + "colab_type": "code", + "outputId": "dd5482cd-c558-465f-c854-ec11a0175316", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 445 + } + }, + "cell_type": "code", + "source": [ + "estimator.evaluate(input_fn=test_input_fn, steps=None)" + ], + "execution_count": 59, + "outputs": [ + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Calling model_fn.\n", + "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n" + ], + "name": "stdout" + }, + { + "output_type": "stream", + "text": [ + "/usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/gradients_impl.py:110: UserWarning: Converting sparse IndexedSlices to a dense Tensor of unknown shape. This may consume a large amount of memory.\n", + " \"Converting sparse IndexedSlices to a dense Tensor of unknown shape. \"\n" + ], + "name": "stderr" + }, + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Done calling model_fn.\n", + "INFO:tensorflow:Starting evaluation at 2019-02-12T21:04:20Z\n", + "INFO:tensorflow:Graph was finalized.\n", + "INFO:tensorflow:Restoring parameters from gs://bert-tfhub/aclImdb_v1/model.ckpt-468\n", + "INFO:tensorflow:Running local_init_op.\n", + "INFO:tensorflow:Done running local_init_op.\n", + "INFO:tensorflow:Finished evaluation at 2019-02-12-21:06:05\n", + "INFO:tensorflow:Saving dict for global step 468: auc = 0.86659324, eval_accuracy = 0.8664, f1_score = 0.8659711, false_negatives = 375.0, false_positives = 293.0, global_step = 468, loss = 0.51870537, precision = 0.880457, recall = 0.8519542, true_negatives = 2174.0, true_positives = 2158.0\n", + "INFO:tensorflow:Saving 'checkpoint_path' summary for global step 468: gs://bert-tfhub/aclImdb_v1/model.ckpt-468\n" + ], + "name": "stdout" + }, + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "{'auc': 0.86659324,\n", + " 'eval_accuracy': 0.8664,\n", + " 'f1_score': 0.8659711,\n", + " 'false_negatives': 375.0,\n", + " 'false_positives': 293.0,\n", + " 'global_step': 468,\n", + " 'loss': 0.51870537,\n", + " 'precision': 0.880457,\n", + " 'recall': 0.8519542,\n", + " 'true_negatives': 2174.0,\n", + " 'true_positives': 2158.0}" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 59 + } + ] + }, + { + "metadata": { + "id": "ueKsULteiz1B", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Now let's write code to make predictions on new sentences:" + ] + }, + { + "metadata": { + "id": "OsrbTD2EJTVl", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "def getPrediction(in_sentences):\n", + " labels = [\"Negative\", \"Positive\"]\n", + " input_examples = [run_classifier.InputExample(guid=\"\", text_a = x, text_b = None, label = 0) for x in in_sentences] # here, \"\" is just a dummy label\n", + " input_features = run_classifier.convert_examples_to_features(input_examples, label_list, MAX_SEQ_LENGTH, tokenizer)\n", + " predict_input_fn = run_classifier.input_fn_builder(features=input_features, seq_length=MAX_SEQ_LENGTH, is_training=False, drop_remainder=False)\n", + " predictions = estimator.predict(predict_input_fn)\n", + " return [(sentence, prediction['probabilities'], labels[prediction['labels']]) for sentence, prediction in zip(in_sentences, predictions)]" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "-thbodgih_VJ", + "colab_type": "code", + "colab": {} + }, + "cell_type": "code", + "source": [ + "pred_sentences = [\n", + " \"That movie was absolutely awful\",\n", + " \"The acting was a bit lacking\",\n", + " \"The film was creative and surprising\",\n", + " \"Absolutely fantastic!\"\n", + "]" + ], + "execution_count": 0, + "outputs": [] + }, + { + "metadata": { + "id": "QrZmvZySKQTm", + "colab_type": "code", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 649 + }, + "outputId": "3891fafb-a460-4eb8-fa6c-335a5bbc10e5" + }, + "cell_type": "code", + "source": [ + "predictions = getPrediction(pred_sentences)" + ], + "execution_count": 72, + "outputs": [ + { + "output_type": "stream", + "text": [ + "INFO:tensorflow:Writing example 0 of 4\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: \n", + "INFO:tensorflow:tokens: [CLS] that movie was absolutely awful [SEP]\n", + "INFO:tensorflow:input_ids: 101 2008 3185 2001 7078 9643 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: \n", + "INFO:tensorflow:tokens: [CLS] the acting was a bit lacking [SEP]\n", + "INFO:tensorflow:input_ids: 101 1996 3772 2001 1037 2978 11158 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: \n", + "INFO:tensorflow:tokens: [CLS] the film was creative and surprising [SEP]\n", + "INFO:tensorflow:input_ids: 101 1996 2143 2001 5541 1998 11341 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:*** Example ***\n", + "INFO:tensorflow:guid: \n", + "INFO:tensorflow:tokens: [CLS] absolutely fantastic ! [SEP]\n", + "INFO:tensorflow:input_ids: 101 7078 10392 999 102 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:input_mask: 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:segment_ids: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0\n", + "INFO:tensorflow:label: 0 (id = 0)\n", + "INFO:tensorflow:Calling model_fn.\n", + "INFO:tensorflow:Saver not created because there are no variables in the graph to restore\n", + "INFO:tensorflow:Done calling model_fn.\n", + "INFO:tensorflow:Graph was finalized.\n", + "INFO:tensorflow:Restoring parameters from gs://bert-tfhub/aclImdb_v1/model.ckpt-468\n", + "INFO:tensorflow:Running local_init_op.\n", + "INFO:tensorflow:Done running local_init_op.\n" + ], + "name": "stdout" + } + ] + }, + { + "metadata": { + "id": "MXkRiEBUqN3n", + "colab_type": "text" + }, + "cell_type": "markdown", + "source": [ + "Voila! We have a sentiment classifier!" + ] + }, + { + "metadata": { + "id": "ERkTE8-7oQLZ", + "colab_type": "code", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 221 + }, + "outputId": "26c33224-dc2c-4b3d-f7b4-ac3ef0a58b27" + }, + "cell_type": "code", + "source": [ + "predictions" + ], + "execution_count": 73, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "text/plain": [ + "[('That movie was absolutely awful',\n", + " array([-4.9142293e-03, -5.3180690e+00], dtype=float32),\n", + " 'Negative'),\n", + " ('The acting was a bit lacking',\n", + " array([-0.03325794, -3.4200459 ], dtype=float32),\n", + " 'Negative'),\n", + " ('The film was creative and surprising',\n", + " array([-5.3589125e+00, -4.7171740e-03], dtype=float32),\n", + " 'Positive'),\n", + " ('Absolutely fantastic!',\n", + " array([-5.0434084 , -0.00647258], dtype=float32),\n", + " 'Positive')]" + ] + }, + "metadata": { + "tags": [] + }, + "execution_count": 73 + } + ] + } + ] +} \ No newline at end of file diff --git a/run_classifier.py b/run_classifier.py new file mode 100644 index 0000000..1ef4f06 --- /dev/null +++ b/run_classifier.py @@ -0,0 +1,1056 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""BERT finetuning runner.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import collections +import csv +import os +import modeling +import optimization +import tokenization +import tensorflow as tf + +flags = tf.flags + +FLAGS = flags.FLAGS + +## Required parameters +flags.DEFINE_string( + "data_dir", None, + "The input data dir. Should contain the .tsv files (or other data files) " + "for the task.") + +flags.DEFINE_string( + "bert_config_file", None, + "The config json file corresponding to the pre-trained BERT model. " + "This specifies the model architecture.") + +flags.DEFINE_string("task_name", None, "The name of the task to train.") + +flags.DEFINE_string("vocab_file", None, + "The vocabulary file that the BERT model was trained on.") + +flags.DEFINE_string( + "output_dir", None, + "The output directory where the model checkpoints will be written.") + +## Other parameters + +flags.DEFINE_string( + "init_checkpoint", None, + "Initial checkpoint (usually from a pre-trained BERT model).") + +flags.DEFINE_bool( + "do_lower_case", True, + "Whether to lower case the input text. Should be True for uncased " + "models and False for cased models.") + +flags.DEFINE_integer( + "max_seq_length", 128, + "The maximum total input sequence length after WordPiece tokenization. " + "Sequences longer than this will be truncated, and sequences shorter " + "than this will be padded.") + +flags.DEFINE_bool("do_train", False, "Whether to run training.") + +flags.DEFINE_bool("do_eval", False, "Whether to run eval on the dev set.") + +flags.DEFINE_bool( + "do_predict", False, + "Whether to run the model in inference mode on the test set.") + +flags.DEFINE_integer("train_batch_size", 32, "Total batch size for training.") + +flags.DEFINE_integer("eval_batch_size", 8, "Total batch size for eval.") + +flags.DEFINE_integer("predict_batch_size", 8, "Total batch size for predict.") + +flags.DEFINE_float("learning_rate", 5e-5, "The initial learning rate for Adam.") + +flags.DEFINE_float("num_train_epochs", 3.0, + "Total number of training epochs to perform.") + +flags.DEFINE_float( + "warmup_proportion", 0.1, + "Proportion of training to perform linear learning rate warmup for. " + "E.g., 0.1 = 10% of training.") + +flags.DEFINE_integer("save_checkpoints_steps", 1000, + "How often to save the model checkpoint.") + +flags.DEFINE_integer("iterations_per_loop", 1000, + "How many steps to make in each estimator call.") + +flags.DEFINE_bool("use_tpu", False, "Whether to use TPU or GPU/CPU.") + +tf.flags.DEFINE_string( + "tpu_name", None, + "The Cloud TPU to use for training. This should be either the name " + "used when creating the Cloud TPU, or a grpc://ip.address.of.tpu:8470 " + "url.") + +tf.flags.DEFINE_string( + "tpu_zone", None, + "[Optional] GCE zone where the Cloud TPU is located in. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string( + "gcp_project", None, + "[Optional] Project name for the Cloud TPU-enabled project. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string("master", None, "[Optional] TensorFlow master URL.") + +flags.DEFINE_integer( + "num_tpu_cores", 8, + "Only used if `use_tpu` is True. Total number of TPU cores to use.") + + +class InputExample(object): + """A single training/test example for simple sequence classification.""" + + def __init__(self, guid, text_a, text_b=None, label=None): + """Constructs a InputExample. + + Args: + guid: Unique id for the example. + text_a: string. The untokenized text of the first sequence. For single + sequence tasks, only this sequence must be specified. + text_b: (Optional) string. The untokenized text of the second sequence. + Only must be specified for sequence pair tasks. + label: (Optional) string. The label of the example. This should be + specified for train and dev examples, but not for test examples. + """ + self.guid = guid + self.text_a = text_a + self.text_b = text_b + self.label = label + + +class PaddingInputExample(object): + """Fake example so the num input examples is a multiple of the batch size. + + When running eval/predict on the TPU, we need to pad the number of examples + to be a multiple of the batch size, because the TPU requires a fixed batch + size. The alternative is to drop the last batch, which is bad because it means + the entire output data won't be generated. + + We use this class instead of `None` because treating `None` as padding + battches could cause silent errors. + """ + + +class InputFeatures(object): + """A single set of features of data.""" + + def __init__(self, + input_ids, + input_mask, + segment_ids, + label_id, + is_real_example=True): + self.input_ids = input_ids + self.input_mask = input_mask + self.segment_ids = segment_ids + self.label_id = label_id + self.is_real_example = is_real_example + + +class DataProcessor(object): + """Base class for data converters for sequence classification data sets.""" + + def get_train_examples(self, data_dir): + """Gets a collection of `InputExample`s for the train set.""" + raise NotImplementedError() + + def get_dev_examples(self, data_dir): + """Gets a collection of `InputExample`s for the dev set.""" + raise NotImplementedError() + + def get_test_examples(self, data_dir): + """Gets a collection of `InputExample`s for prediction.""" + raise NotImplementedError() + + def get_labels(self): + """Gets the list of labels for this data set.""" + raise NotImplementedError() + + @classmethod + def _read_tsv(cls, input_file, quotechar=None): + """Reads a tab separated value file.""" + with tf.gfile.Open(input_file, "r") as f: + reader = csv.reader(f, delimiter="\t", quotechar=quotechar) + lines = [] + for line in reader: + lines.append(line) + return lines + + +class XnliProcessor(DataProcessor): + """Processor for the XNLI data set.""" + + def __init__(self): + self.language = "zh" + + def get_train_examples(self, data_dir): + """See base class.""" + lines = self._read_tsv( + os.path.join(data_dir, "multinli", + "multinli.train.%s.tsv" % self.language)) + examples = [] + for (i, line) in enumerate(lines): + if i == 0: + continue + guid = "train-%d" % (i) + text_a = tokenization.convert_to_unicode(line[0]) + text_b = tokenization.convert_to_unicode(line[1]) + label = tokenization.convert_to_unicode(line[2]) + if label == tokenization.convert_to_unicode("contradictory"): + label = tokenization.convert_to_unicode("contradiction") + examples.append( + InputExample(guid=guid, text_a=text_a, text_b=text_b, label=label)) + return examples + + def get_dev_examples(self, data_dir): + """See base class.""" + lines = self._read_tsv(os.path.join(data_dir, "xnli.dev.tsv")) + examples = [] + for (i, line) in enumerate(lines): + if i == 0: + continue + guid = "dev-%d" % (i) + language = tokenization.convert_to_unicode(line[0]) + if language != tokenization.convert_to_unicode(self.language): + continue + text_a = tokenization.convert_to_unicode(line[6]) + text_b = tokenization.convert_to_unicode(line[7]) + label = tokenization.convert_to_unicode(line[1]) + examples.append( + InputExample(guid=guid, text_a=text_a, text_b=text_b, label=label)) + return examples + + def get_labels(self): + """See base class.""" + return ["contradiction", "entailment", "neutral"] + + +class MnliProcessor(DataProcessor): + """Processor for the MultiNLI data set (GLUE version).""" + + def get_train_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "train.tsv")), "train") + + def get_dev_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "dev_matched.tsv")), + "dev_matched") + + def get_test_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "test_matched.tsv")), "test") + + def get_labels(self): + """See base class.""" + return ["contradiction", "entailment", "neutral"] + + def _create_examples(self, lines, set_type): + """Creates examples for the training and dev sets.""" + examples = [] + for (i, line) in enumerate(lines): + if i == 0: + continue + guid = "%s-%s" % (set_type, tokenization.convert_to_unicode(line[0])) + text_a = tokenization.convert_to_unicode(line[8]) + text_b = tokenization.convert_to_unicode(line[9]) + if set_type == "test": + label = "contradiction" + else: + label = tokenization.convert_to_unicode(line[-1]) + examples.append( + InputExample(guid=guid, text_a=text_a, text_b=text_b, label=label)) + return examples + + +class MrpcProcessor(DataProcessor): + """Processor for the MRPC data set (GLUE version).""" + + def get_train_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "train.tsv")), "train") + + def get_dev_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "dev.tsv")), "dev") + + def get_test_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "test.tsv")), "test") + + def get_labels(self): + """See base class.""" + return ["0", "1"] + + def _create_examples(self, lines, set_type): + """Creates examples for the training and dev sets.""" + examples = [] + for (i, line) in enumerate(lines): + if i == 0: + continue + guid = "%s-%s" % (set_type, i) + text_a = tokenization.convert_to_unicode(line[3]) + text_b = tokenization.convert_to_unicode(line[4]) + if set_type == "test": + label = "0" + else: + label = tokenization.convert_to_unicode(line[0]) + examples.append( + InputExample(guid=guid, text_a=text_a, text_b=text_b, label=label)) + return examples + + +class ColaProcessor(DataProcessor): + """Processor for the CoLA data set (GLUE version).""" + + def get_train_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "train.tsv")), "train") + + def get_dev_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "dev.tsv")), "dev") + + def get_test_examples(self, data_dir): + """See base class.""" + return self._create_examples( + self._read_tsv(os.path.join(data_dir, "test.tsv")), "test") + + def get_labels(self): + """See base class.""" + return ["0", "1"] + + def _create_examples(self, lines, set_type): + """Creates examples for the training and dev sets.""" + examples = [] + for (i, line) in enumerate(lines): + # Only the test set has a header + if set_type == "test" and i == 0: + continue + guid = "%s-%s" % (set_type, i) + if set_type == "test": + text_a = tokenization.convert_to_unicode(line[1]) + label = "0" + else: + text_a = tokenization.convert_to_unicode(line[3]) + label = tokenization.convert_to_unicode(line[1]) + examples.append( + InputExample(guid=guid, text_a=text_a, text_b=None, label=label)) + return examples + + +import dealing_dataset + + +class EPProcessor(DataProcessor): + """Processor for the Emotion data set .""" + + def get_train_examples(self, data_dir): + """定义开发集的数据是什么，data_dir会作为参数传进去， 这里就是加上你的文件名即可 """ + return self._create_examples("amki_train") + + def get_dev_examples(self, data_dir): + """定义开发集的数据是什么，data_dir会作为参数传进去，模型训练的时候会用到，这里就是加上你的文件名即可 """ + return self._create_examples("amki_dev") + + def get_test_examples(self, data_dir): + """定义测试集的数据是什么， 用于预测数据 ，在训练时没有用到这个函数， 这里写预测的数据集""" + return self._create_examples("amki_test") + + def get_labels(self): + """ 这里是显示你一共有几个分类标签， 在此任务中我有3个标签，如实写上 标签值和 csv里面存的值相同 """ + return [0, 1, 2] + + def _create_examples(self, data_table): + """这个函数是用来把数据处理， 把每一个例子分成3个部分，填入到InputExample的3个参数 + text_a 是 第一个句子的文本 + text_b 是 第二个句子的文本 但是由于此任务是单句分类， 所以 这里传入为None + guid 是一个二元组 第一个表示此数据是什么数据集类型（train dev test） 第二个表示数据标号 + label 表示句子类别 + """ + examples = [] + for column in dealing_dataset.create_dataset_ep(data_table): + # 加入样本 + examples.append( + InputExample(guid=column[0], text_a=column[2], text_b=None, label=column[1])) + + return examples + + +class EPBPTProcessor(DataProcessor): + """Processor for the Emotion data set .""" + + def get_train_examples(self, data_dir): + """定义开发集的数据是什么，data_dir会作为参数传进去， 这里就是加上你的文件名即可 """ + return self._create_examples("amki_train") + + def get_dev_examples(self, data_dir): + """定义开发集的数据是什么，data_dir会作为参数传进去，模型训练的时候会用到，这里就是加上你的文件名即可 """ + return self._create_examples("amki_dev") + + def get_test_examples(self, data_dir): + """定义测试集的数据是什么， 用于预测数据 ，在训练时没有用到这个函数， 这里写预测的数据集""" + return self._create_examples("amki_test") + + def get_labels(self): + """ 这里是显示你一共有几个分类标签， 在此任务中我有3个标签，如实写上 标签值和 csv里面存的值相同 """ + return [0, 1, 2] + + def _create_examples(self, data_table): + """这个函数是用来把数据处理， 把每一个例子分成3个部分，填入到InputExample的3个参数 + text_a 是 第一个句子的文本 + text_b 是 第二个句子的文本 但是由于此任务是单句分类， 所以 这里传入为None + guid 是一个二元组 第一个表示此数据是什么数据集类型（train dev test） 第二个表示数据标号 + label 表示句子类别 + """ + examples = [] + for column in dealing_dataset.create_dataset_pdt(): + # 加入样本 + examples.append( + InputExample(guid=column[0], text_a=column[2], text_b=None, label=column[1])) + + return examples + + +def convert_single_example(ex_index, example, label_list, max_seq_length, + tokenizer): + """Converts a single `InputExample` into a single `InputFeatures`.""" + + if isinstance(example, PaddingInputExample): + return InputFeatures( + input_ids=[0] * max_seq_length, + input_mask=[0] * max_seq_length, + segment_ids=[0] * max_seq_length, + label_id=0, + is_real_example=False) + + label_map = {} + for (i, label) in enumerate(label_list): + label_map[label] = i + + tokens_a = tokenizer.tokenize(example.text_a) + tokens_b = None + if example.text_b: + tokens_b = tokenizer.tokenize(example.text_b) + + if tokens_b: + # Modifies `tokens_a` and `tokens_b` in place so that the total + # length is less than the specified length. + # Account for [CLS], [SEP], [SEP] with "- 3" + _truncate_seq_pair(tokens_a, tokens_b, max_seq_length - 3) + else: + # Account for [CLS] and [SEP] with "- 2" + if len(tokens_a) > max_seq_length - 2: + tokens_a = tokens_a[0:(max_seq_length - 2)] + + # The convention in BERT is: + # (a) For sequence pairs: + # tokens: [CLS] is this jack ##son ##ville ? [SEP] no it is not . [SEP] + # type_ids: 0 0 0 0 0 0 0 0 1 1 1 1 1 1 + # (b) For single sequences: + # tokens: [CLS] the dog is hairy . [SEP] + # type_ids: 0 0 0 0 0 0 0 + # + # Where "type_ids" are used to indicate whether this is the first + # sequence or the second sequence. The embedding vectors for `type=0` and + # `type=1` were learned during pre-training and are added to the wordpiece + # embedding vector (and position vector). This is not *strictly* necessary + # since the [SEP] token unambiguously separates the sequences, but it makes + # it easier for the model to learn the concept of sequences. + # + # For classification tasks, the first vector (corresponding to [CLS]) is + # used as the "sentence vector". Note that this only makes sense because + # the entire model is fine-tuned. + tokens = [] + segment_ids = [] + tokens.append("[CLS]") + segment_ids.append(0) + for token in tokens_a: + tokens.append(token) + segment_ids.append(0) + tokens.append("[SEP]") + segment_ids.append(0) + + if tokens_b: + for token in tokens_b: + tokens.append(token) + segment_ids.append(1) + tokens.append("[SEP]") + segment_ids.append(1) + + input_ids = tokenizer.convert_tokens_to_ids(tokens) + + # The mask has 1 for real tokens and 0 for padding tokens. Only real + # tokens are attended to. + input_mask = [1] * len(input_ids) + + # Zero-pad up to the sequence length. + while len(input_ids) < max_seq_length: + input_ids.append(0) + input_mask.append(0) + segment_ids.append(0) + + assert len(input_ids) == max_seq_length + assert len(input_mask) == max_seq_length + assert len(segment_ids) == max_seq_length + + label_id = label_map[example.label] + if ex_index < 5: + tf.logging.info("*** Example ***") + tf.logging.info("guid: %s" % (example.guid)) + tf.logging.info("tokens: %s" % " ".join( + [tokenization.printable_text(x) for x in tokens])) + tf.logging.info("input_ids: %s" % " ".join([str(x) for x in input_ids])) + tf.logging.info("input_mask: %s" % " ".join([str(x) for x in input_mask])) + tf.logging.info("segment_ids: %s" % " ".join([str(x) for x in segment_ids])) + tf.logging.info("label: %s (id = %d)" % (example.label, label_id)) + + feature = InputFeatures( + input_ids=input_ids, + input_mask=input_mask, + segment_ids=segment_ids, + label_id=label_id, + is_real_example=True) + return feature + + +def file_based_convert_examples_to_features( + examples, label_list, max_seq_length, tokenizer, output_file): + """Convert a set of `InputExample`s to a TFRecord file.""" + + writer = tf.python_io.TFRecordWriter(output_file) + + for (ex_index, example) in enumerate(examples): + if ex_index % 10000 == 0: + tf.logging.info("Writing example %d of %d" % (ex_index, len(examples))) + + feature = convert_single_example(ex_index, example, label_list, + max_seq_length, tokenizer) + + def create_int_feature(values): + f = tf.train.Feature(int64_list=tf.train.Int64List(value=list(values))) + return f + + features = collections.OrderedDict() + features["input_ids"] = create_int_feature(feature.input_ids) + features["input_mask"] = create_int_feature(feature.input_mask) + features["segment_ids"] = create_int_feature(feature.segment_ids) + features["label_ids"] = create_int_feature([feature.label_id]) + features["is_real_example"] = create_int_feature( + [int(feature.is_real_example)]) + + tf_example = tf.train.Example(features=tf.train.Features(feature=features)) + writer.write(tf_example.SerializeToString()) + writer.close() + + +def file_based_input_fn_builder(input_file, seq_length, is_training, + drop_remainder): + """Creates an `input_fn` closure to be passed to TPUEstimator.""" + + name_to_features = { + "input_ids": tf.FixedLenFeature([seq_length], tf.int64), + "input_mask": tf.FixedLenFeature([seq_length], tf.int64), + "segment_ids": tf.FixedLenFeature([seq_length], tf.int64), + "label_ids": tf.FixedLenFeature([], tf.int64), + "is_real_example": tf.FixedLenFeature([], tf.int64), + } + + def _decode_record(record, name_to_features): + """Decodes a record to a TensorFlow example.""" + example = tf.parse_single_example(record, name_to_features) + + # tf.Example only supports tf.int64, but the TPU only supports tf.int32. + # So cast all int64 to int32. + for name in list(example.keys()): + t = example[name] + if t.dtype == tf.int64: + t = tf.to_int32(t) + example[name] = t + + return example + + def input_fn(params): + """The actual input function.""" + batch_size = params["batch_size"] + + # For training, we want a lot of parallel reading and shuffling. + # For eval, we want no shuffling and parallel reading doesn't matter. + d = tf.data.TFRecordDataset(input_file) + if is_training: + d = d.repeat() + d = d.shuffle(buffer_size=100) + + d = d.apply( + tf.contrib.data.map_and_batch( + lambda record: _decode_record(record, name_to_features), + batch_size=batch_size, + drop_remainder=drop_remainder)) + + return d + + return input_fn + + +def _truncate_seq_pair(tokens_a, tokens_b, max_length): + """Truncates a sequence pair in place to the maximum length.""" + + # This is a simple heuristic which will always truncate the longer sequence + # one token at a time. This makes more sense than truncating an equal percent + # of tokens from each, since if one sequence is very short then each token + # that's truncated likely contains more information than a longer sequence. + while True: + total_length = len(tokens_a) + len(tokens_b) + if total_length <= max_length: + break + if len(tokens_a) > len(tokens_b): + tokens_a.pop() + else: + tokens_b.pop() + + +def create_model(bert_config, is_training, input_ids, input_mask, segment_ids, + labels, num_labels, use_one_hot_embeddings): + """Creates a classification model.""" + model = modeling.BertModel( + config=bert_config, + is_training=is_training, + input_ids=input_ids, + input_mask=input_mask, + token_type_ids=segment_ids, + use_one_hot_embeddings=use_one_hot_embeddings) + + # In the demo, we are doing a simple classification task on the entire + # segment. + # + # If you want to use the token-level output, use model.get_sequence_output() + # instead. + output_layer = model.get_pooled_output() + + hidden_size = output_layer.shape[-1].value + + output_weights = tf.get_variable( + "output_weights", [num_labels, hidden_size], + initializer=tf.truncated_normal_initializer(stddev=0.02)) + + output_bias = tf.get_variable( + "output_bias", [num_labels], initializer=tf.zeros_initializer()) + + with tf.variable_scope("loss"): + if is_training: + # I.e., 0.1 dropout + output_layer = tf.nn.dropout(output_layer, keep_prob=0.9) + + logits = tf.matmul(output_layer, output_weights, transpose_b=True) + logits = tf.nn.bias_add(logits, output_bias) + probabilities = tf.nn.softmax(logits, axis=-1) + log_probs = tf.nn.log_softmax(logits, axis=-1) + + one_hot_labels = tf.one_hot(labels, depth=num_labels, dtype=tf.float32) + + per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1) + loss = tf.reduce_mean(per_example_loss) + + return (loss, per_example_loss, logits, probabilities) + + +def model_fn_builder(bert_config, num_labels, init_checkpoint, learning_rate, + num_train_steps, num_warmup_steps, use_tpu, + use_one_hot_embeddings): + """Returns `model_fn` closure for TPUEstimator.""" + + def model_fn(features, labels, mode, params): # pylint: disable=unused-argument + """The `model_fn` for TPUEstimator.""" + + tf.logging.info("*** Features ***") + for name in sorted(features.keys()): + tf.logging.info(" name = %s, shape = %s" % (name, features[name].shape)) + + input_ids = features["input_ids"] + input_mask = features["input_mask"] + segment_ids = features["segment_ids"] + label_ids = features["label_ids"] + is_real_example = None + if "is_real_example" in features: + is_real_example = tf.cast(features["is_real_example"], dtype=tf.float32) + else: + is_real_example = tf.ones(tf.shape(label_ids), dtype=tf.float32) + + is_training = (mode == tf.estimator.ModeKeys.TRAIN) + + (total_loss, per_example_loss, logits, probabilities) = create_model( + bert_config, is_training, input_ids, input_mask, segment_ids, label_ids, + num_labels, use_one_hot_embeddings) + + tvars = tf.trainable_variables() + initialized_variable_names = {} + scaffold_fn = None + if init_checkpoint: + (assignment_map, initialized_variable_names + ) = modeling.get_assignment_map_from_checkpoint(tvars, init_checkpoint) + if use_tpu: + + def tpu_scaffold(): + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + return tf.train.Scaffold() + + scaffold_fn = tpu_scaffold + else: + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + + tf.logging.info("**** Trainable Variables ****") + for var in tvars: + init_string = "" + if var.name in initialized_variable_names: + init_string = ", *INIT_FROM_CKPT*" + tf.logging.info(" name = %s, shape = %s%s", var.name, var.shape, + init_string) + + output_spec = None + if mode == tf.estimator.ModeKeys.TRAIN: + + train_op = optimization.create_optimizer( + total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu) + + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + train_op=train_op, + scaffold_fn=scaffold_fn) + elif mode == tf.estimator.ModeKeys.EVAL: + + def metric_fn(per_example_loss, label_ids, logits, is_real_example): + predictions = tf.argmax(logits, axis=-1, output_type=tf.int32) + accuracy = tf.metrics.accuracy( + labels=label_ids, predictions=predictions, weights=is_real_example) + loss = tf.metrics.mean(values=per_example_loss, weights=is_real_example) + return { + "eval_accuracy": accuracy, + "eval_loss": loss, + } + + eval_metrics = (metric_fn, + [per_example_loss, label_ids, logits, is_real_example]) + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + eval_metrics=eval_metrics, + scaffold_fn=scaffold_fn) + else: + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + predictions={"probabilities": probabilities}, + scaffold_fn=scaffold_fn) + return output_spec + + return model_fn + + +# This function is not used by this file but is still used by the Colab and +# people who depend on it. +def input_fn_builder(features, seq_length, is_training, drop_remainder): + """Creates an `input_fn` closure to be passed to TPUEstimator.""" + + all_input_ids = [] + all_input_mask = [] + all_segment_ids = [] + all_label_ids = [] + + for feature in features: + all_input_ids.append(feature.input_ids) + all_input_mask.append(feature.input_mask) + all_segment_ids.append(feature.segment_ids) + all_label_ids.append(feature.label_id) + + def input_fn(params): + """The actual input function.""" + batch_size = params["batch_size"] + + num_examples = len(features) + + # This is for demo purposes and does NOT scale to large data sets. We do + # not use Dataset.from_generator() because that uses tf.py_func which is + # not TPU compatible. The right way to load data is with TFRecordReader. + d = tf.data.Dataset.from_tensor_slices({ + "input_ids": + tf.constant( + all_input_ids, shape=[num_examples, seq_length], + dtype=tf.int32), + "input_mask": + tf.constant( + all_input_mask, + shape=[num_examples, seq_length], + dtype=tf.int32), + "segment_ids": + tf.constant( + all_segment_ids, + shape=[num_examples, seq_length], + dtype=tf.int32), + "label_ids": + tf.constant(all_label_ids, shape=[num_examples], dtype=tf.int32), + }) + + if is_training: + d = d.repeat() + d = d.shuffle(buffer_size=100) + + d = d.batch(batch_size=batch_size, drop_remainder=drop_remainder) + return d + + return input_fn + + +# This function is not used by this file but is still used by the Colab and +# people who depend on it. +def convert_examples_to_features(examples, label_list, max_seq_length, + tokenizer): + """Convert a set of `InputExample`s to a list of `InputFeatures`.""" + + features = [] + for (ex_index, example) in enumerate(examples): + if ex_index % 10000 == 0: + tf.logging.info("Writing example %d of %d" % (ex_index, len(examples))) + + feature = convert_single_example(ex_index, example, label_list, + max_seq_length, tokenizer) + + features.append(feature) + return features + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + processors = { + "cola": ColaProcessor, + "mnli": MnliProcessor, + "mrpc": MrpcProcessor, + "xnli": XnliProcessor, + "ep": EPProcessor, + "eppdt": EPBPTProcessor, + } + + tokenization.validate_case_matches_checkpoint(FLAGS.do_lower_case, + FLAGS.init_checkpoint) + + if not FLAGS.do_train and not FLAGS.do_eval and not FLAGS.do_predict: + raise ValueError( + "At least one of `do_train`, `do_eval` or `do_predict' must be True.") + + bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file) + + if FLAGS.max_seq_length > bert_config.max_position_embeddings: + raise ValueError( + "Cannot use sequence length %d because the BERT model " + "was only trained up to sequence length %d" % + (FLAGS.max_seq_length, bert_config.max_position_embeddings)) + + tf.gfile.MakeDirs(FLAGS.output_dir) + + task_name = FLAGS.task_name.lower() + + if task_name not in processors: + raise ValueError("Task not found: %s" % (task_name)) + + processor = processors[task_name]() + + label_list = processor.get_labels() + + tokenizer = tokenization.FullTokenizer( + vocab_file=FLAGS.vocab_file, do_lower_case=FLAGS.do_lower_case) + + tpu_cluster_resolver = None + if FLAGS.use_tpu and FLAGS.tpu_name: + tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver( + FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project) + + is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2 + run_config = tf.contrib.tpu.RunConfig( + cluster=tpu_cluster_resolver, + master=FLAGS.master, + model_dir=FLAGS.output_dir, + save_checkpoints_steps=FLAGS.save_checkpoints_steps, + tpu_config=tf.contrib.tpu.TPUConfig( + iterations_per_loop=FLAGS.iterations_per_loop, + num_shards=FLAGS.num_tpu_cores, + per_host_input_for_training=is_per_host)) + + train_examples = None + num_train_steps = None + num_warmup_steps = None + if FLAGS.do_train: + train_examples = processor.get_train_examples(FLAGS.data_dir) + num_train_steps = int( + len(train_examples) / FLAGS.train_batch_size * FLAGS.num_train_epochs) + num_warmup_steps = int(num_train_steps * FLAGS.warmup_proportion) + + model_fn = model_fn_builder( + bert_config=bert_config, + num_labels=len(label_list), + init_checkpoint=FLAGS.init_checkpoint, + learning_rate=FLAGS.learning_rate, + num_train_steps=num_train_steps, + num_warmup_steps=num_warmup_steps, + use_tpu=FLAGS.use_tpu, + use_one_hot_embeddings=FLAGS.use_tpu) + + # If TPU is not available, this will fall back to normal Estimator on CPU + # or GPU. + estimator = tf.contrib.tpu.TPUEstimator( + use_tpu=FLAGS.use_tpu, + model_fn=model_fn, + config=run_config, + train_batch_size=FLAGS.train_batch_size, + eval_batch_size=FLAGS.eval_batch_size, + predict_batch_size=FLAGS.predict_batch_size) + + if FLAGS.do_train: + train_file = os.path.join(FLAGS.output_dir, "train.tf_record") + file_based_convert_examples_to_features( + train_examples, label_list, FLAGS.max_seq_length, tokenizer, train_file) + tf.logging.info("***** Running training *****") + tf.logging.info(" Num examples = %d", len(train_examples)) + tf.logging.info(" Batch size = %d", FLAGS.train_batch_size) + tf.logging.info(" Num steps = %d", num_train_steps) + train_input_fn = file_based_input_fn_builder( + input_file=train_file, + seq_length=FLAGS.max_seq_length, + is_training=True, + drop_remainder=True) + estimator.train(input_fn=train_input_fn, max_steps=num_train_steps) + + if FLAGS.do_eval: + eval_examples = processor.get_dev_examples(FLAGS.data_dir) + num_actual_eval_examples = len(eval_examples) + if FLAGS.use_tpu: + # TPU requires a fixed batch size for all batches, therefore the number + # of examples must be a multiple of the batch size, or else examples + # will get dropped. So we pad with fake examples which are ignored + # later on. These do NOT count towards the metric (all tf.metrics + # support a per-instance weight, and these get a weight of 0.0). + while len(eval_examples) % FLAGS.eval_batch_size != 0: + eval_examples.append(PaddingInputExample()) + + eval_file = os.path.join(FLAGS.output_dir, "eval.tf_record") + file_based_convert_examples_to_features( + eval_examples, label_list, FLAGS.max_seq_length, tokenizer, eval_file) + + tf.logging.info("***** Running evaluation *****") + tf.logging.info(" Num examples = %d (%d actual, %d padding)", + len(eval_examples), num_actual_eval_examples, + len(eval_examples) - num_actual_eval_examples) + tf.logging.info(" Batch size = %d", FLAGS.eval_batch_size) + + # This tells the estimator to run through the entire set. + eval_steps = None + # However, if running eval on the TPU, you will need to specify the + # number of steps. + if FLAGS.use_tpu: + assert len(eval_examples) % FLAGS.eval_batch_size == 0 + eval_steps = int(len(eval_examples) // FLAGS.eval_batch_size) + + eval_drop_remainder = True if FLAGS.use_tpu else False + eval_input_fn = file_based_input_fn_builder( + input_file=eval_file, + seq_length=FLAGS.max_seq_length, + is_training=False, + drop_remainder=eval_drop_remainder) + + result = estimator.evaluate(input_fn=eval_input_fn, steps=eval_steps) + + output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt") + with tf.gfile.GFile(output_eval_file, "w") as writer: + tf.logging.info("***** Eval results *****") + for key in sorted(result.keys()): + tf.logging.info(" %s = %s", key, str(result[key])) + writer.write("%s = %s\n" % (key, str(result[key]))) + + if FLAGS.do_predict: + predict_examples = processor.get_test_examples(FLAGS.data_dir) + num_actual_predict_examples = len(predict_examples) + if FLAGS.use_tpu: + # TPU requires a fixed batch size for all batches, therefore the number + # of examples must be a multiple of the batch size, or else examples + # will get dropped. So we pad with fake examples which are ignored + # later on. + while len(predict_examples) % FLAGS.predict_batch_size != 0: + predict_examples.append(PaddingInputExample()) + + predict_file = os.path.join(FLAGS.output_dir, "predict.tf_record") + file_based_convert_examples_to_features(predict_examples, label_list, + FLAGS.max_seq_length, tokenizer, + predict_file) + + tf.logging.info("***** Running prediction*****") + tf.logging.info(" Num examples = %d (%d actual, %d padding)", + len(predict_examples), num_actual_predict_examples, + len(predict_examples) - num_actual_predict_examples) + tf.logging.info(" Batch size = %d", FLAGS.predict_batch_size) + + predict_drop_remainder = True if FLAGS.use_tpu else False + predict_input_fn = file_based_input_fn_builder( + input_file=predict_file, + seq_length=FLAGS.max_seq_length, + is_training=False, + drop_remainder=predict_drop_remainder) + + result = estimator.predict(input_fn=predict_input_fn) + + output_predict_file = os.path.join(FLAGS.output_dir, "test_results.tsv") + with tf.gfile.GFile(output_predict_file, "w") as writer: + num_written_lines = 0 + tf.logging.info("***** Predict results *****") + for (i, prediction) in enumerate(result): + probabilities = prediction["probabilities"] + if i >= num_actual_predict_examples: + break + output_line = "\t".join( + str(class_probability) + for class_probability in probabilities) + "\n" + writer.write(output_line) + num_written_lines += 1 + assert num_written_lines == num_actual_predict_examples + + +if __name__ == "__main__": + flags.mark_flag_as_required("data_dir") + flags.mark_flag_as_required("task_name") + flags.mark_flag_as_required("vocab_file") + flags.mark_flag_as_required("bert_config_file") + flags.mark_flag_as_required("output_dir") + tf.app.run() diff --git a/run_classifier_with_tfhub.py b/run_classifier_with_tfhub.py new file mode 100644 index 0000000..9d2f80f --- /dev/null +++ b/run_classifier_with_tfhub.py @@ -0,0 +1,314 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""BERT finetuning runner with TF-Hub.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import os +import optimization +import run_classifier +import tokenization +import tensorflow as tf +import tensorflow_hub as hub + +flags = tf.flags + +FLAGS = flags.FLAGS + +flags.DEFINE_string( + "bert_hub_module_handle", None, + "Handle for the BERT TF-Hub module.") + + +def create_model(is_training, input_ids, input_mask, segment_ids, labels, + num_labels, bert_hub_module_handle): + """Creates a classification model.""" + tags = set() + if is_training: + tags.add("train") + bert_module = hub.Module(bert_hub_module_handle, tags=tags, trainable=True) + bert_inputs = dict( + input_ids=input_ids, + input_mask=input_mask, + segment_ids=segment_ids) + bert_outputs = bert_module( + inputs=bert_inputs, + signature="tokens", + as_dict=True) + + # In the demo, we are doing a simple classification task on the entire + # segment. + # + # If you want to use the token-level output, use + # bert_outputs["sequence_output"] instead. + output_layer = bert_outputs["pooled_output"] + + hidden_size = output_layer.shape[-1].value + + output_weights = tf.get_variable( + "output_weights", [num_labels, hidden_size], + initializer=tf.truncated_normal_initializer(stddev=0.02)) + + output_bias = tf.get_variable( + "output_bias", [num_labels], initializer=tf.zeros_initializer()) + + with tf.variable_scope("loss"): + if is_training: + # I.e., 0.1 dropout + output_layer = tf.nn.dropout(output_layer, keep_prob=0.9) + + logits = tf.matmul(output_layer, output_weights, transpose_b=True) + logits = tf.nn.bias_add(logits, output_bias) + probabilities = tf.nn.softmax(logits, axis=-1) + log_probs = tf.nn.log_softmax(logits, axis=-1) + + one_hot_labels = tf.one_hot(labels, depth=num_labels, dtype=tf.float32) + + per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1) + loss = tf.reduce_mean(per_example_loss) + + return (loss, per_example_loss, logits, probabilities) + + +def model_fn_builder(num_labels, learning_rate, num_train_steps, + num_warmup_steps, use_tpu, bert_hub_module_handle): + """Returns `model_fn` closure for TPUEstimator.""" + + def model_fn(features, labels, mode, params): # pylint: disable=unused-argument + """The `model_fn` for TPUEstimator.""" + + tf.logging.info("*** Features ***") + for name in sorted(features.keys()): + tf.logging.info(" name = %s, shape = %s" % (name, features[name].shape)) + + input_ids = features["input_ids"] + input_mask = features["input_mask"] + segment_ids = features["segment_ids"] + label_ids = features["label_ids"] + + is_training = (mode == tf.estimator.ModeKeys.TRAIN) + + (total_loss, per_example_loss, logits, probabilities) = create_model( + is_training, input_ids, input_mask, segment_ids, label_ids, num_labels, + bert_hub_module_handle) + + output_spec = None + if mode == tf.estimator.ModeKeys.TRAIN: + train_op = optimization.create_optimizer( + total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu) + + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + train_op=train_op) + elif mode == tf.estimator.ModeKeys.EVAL: + + def metric_fn(per_example_loss, label_ids, logits): + predictions = tf.argmax(logits, axis=-1, output_type=tf.int32) + accuracy = tf.metrics.accuracy(label_ids, predictions) + loss = tf.metrics.mean(per_example_loss) + return { + "eval_accuracy": accuracy, + "eval_loss": loss, + } + + eval_metrics = (metric_fn, [per_example_loss, label_ids, logits]) + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + eval_metrics=eval_metrics) + elif mode == tf.estimator.ModeKeys.PREDICT: + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, predictions={"probabilities": probabilities}) + else: + raise ValueError( + "Only TRAIN, EVAL and PREDICT modes are supported: %s" % (mode)) + + return output_spec + + return model_fn + + +def create_tokenizer_from_hub_module(bert_hub_module_handle): + """Get the vocab file and casing info from the Hub module.""" + with tf.Graph().as_default(): + bert_module = hub.Module(bert_hub_module_handle) + tokenization_info = bert_module(signature="tokenization_info", as_dict=True) + with tf.Session() as sess: + vocab_file, do_lower_case = sess.run([tokenization_info["vocab_file"], + tokenization_info["do_lower_case"]]) + return tokenization.FullTokenizer( + vocab_file=vocab_file, do_lower_case=do_lower_case) + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + processors = { + "cola": run_classifier.ColaProcessor, + "mnli": run_classifier.MnliProcessor, + "mrpc": run_classifier.MrpcProcessor, + } + + if not FLAGS.do_train and not FLAGS.do_eval: + raise ValueError("At least one of `do_train` or `do_eval` must be True.") + + tf.gfile.MakeDirs(FLAGS.output_dir) + + task_name = FLAGS.task_name.lower() + + if task_name not in processors: + raise ValueError("Task not found: %s" % (task_name)) + + processor = processors[task_name]() + + label_list = processor.get_labels() + + tokenizer = create_tokenizer_from_hub_module(FLAGS.bert_hub_module_handle) + + tpu_cluster_resolver = None + if FLAGS.use_tpu and FLAGS.tpu_name: + tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver( + FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project) + + is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2 + run_config = tf.contrib.tpu.RunConfig( + cluster=tpu_cluster_resolver, + master=FLAGS.master, + model_dir=FLAGS.output_dir, + save_checkpoints_steps=FLAGS.save_checkpoints_steps, + tpu_config=tf.contrib.tpu.TPUConfig( + iterations_per_loop=FLAGS.iterations_per_loop, + num_shards=FLAGS.num_tpu_cores, + per_host_input_for_training=is_per_host)) + + train_examples = None + num_train_steps = None + num_warmup_steps = None + if FLAGS.do_train: + train_examples = processor.get_train_examples(FLAGS.data_dir) + num_train_steps = int( + len(train_examples) / FLAGS.train_batch_size * FLAGS.num_train_epochs) + num_warmup_steps = int(num_train_steps * FLAGS.warmup_proportion) + + model_fn = model_fn_builder( + num_labels=len(label_list), + learning_rate=FLAGS.learning_rate, + num_train_steps=num_train_steps, + num_warmup_steps=num_warmup_steps, + use_tpu=FLAGS.use_tpu, + bert_hub_module_handle=FLAGS.bert_hub_module_handle) + + # If TPU is not available, this will fall back to normal Estimator on CPU + # or GPU. + estimator = tf.contrib.tpu.TPUEstimator( + use_tpu=FLAGS.use_tpu, + model_fn=model_fn, + config=run_config, + train_batch_size=FLAGS.train_batch_size, + eval_batch_size=FLAGS.eval_batch_size, + predict_batch_size=FLAGS.predict_batch_size) + + if FLAGS.do_train: + train_features = run_classifier.convert_examples_to_features( + train_examples, label_list, FLAGS.max_seq_length, tokenizer) + tf.logging.info("***** Running training *****") + tf.logging.info(" Num examples = %d", len(train_examples)) + tf.logging.info(" Batch size = %d", FLAGS.train_batch_size) + tf.logging.info(" Num steps = %d", num_train_steps) + train_input_fn = run_classifier.input_fn_builder( + features=train_features, + seq_length=FLAGS.max_seq_length, + is_training=True, + drop_remainder=True) + estimator.train(input_fn=train_input_fn, max_steps=num_train_steps) + + if FLAGS.do_eval: + eval_examples = processor.get_dev_examples(FLAGS.data_dir) + eval_features = run_classifier.convert_examples_to_features( + eval_examples, label_list, FLAGS.max_seq_length, tokenizer) + + tf.logging.info("***** Running evaluation *****") + tf.logging.info(" Num examples = %d", len(eval_examples)) + tf.logging.info(" Batch size = %d", FLAGS.eval_batch_size) + + # This tells the estimator to run through the entire set. + eval_steps = None + # However, if running eval on the TPU, you will need to specify the + # number of steps. + if FLAGS.use_tpu: + # Eval will be slightly WRONG on the TPU because it will truncate + # the last batch. + eval_steps = int(len(eval_examples) / FLAGS.eval_batch_size) + + eval_drop_remainder = True if FLAGS.use_tpu else False + eval_input_fn = run_classifier.input_fn_builder( + features=eval_features, + seq_length=FLAGS.max_seq_length, + is_training=False, + drop_remainder=eval_drop_remainder) + + result = estimator.evaluate(input_fn=eval_input_fn, steps=eval_steps) + + output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt") + with tf.gfile.GFile(output_eval_file, "w") as writer: + tf.logging.info("***** Eval results *****") + for key in sorted(result.keys()): + tf.logging.info(" %s = %s", key, str(result[key])) + writer.write("%s = %s\n" % (key, str(result[key]))) + + if FLAGS.do_predict: + predict_examples = processor.get_test_examples(FLAGS.data_dir) + if FLAGS.use_tpu: + # Discard batch remainder if running on TPU + n = len(predict_examples) + predict_examples = predict_examples[:(n - n % FLAGS.predict_batch_size)] + + predict_file = os.path.join(FLAGS.output_dir, "predict.tf_record") + run_classifier.file_based_convert_examples_to_features( + predict_examples, label_list, FLAGS.max_seq_length, tokenizer, + predict_file) + + tf.logging.info("***** Running prediction*****") + tf.logging.info(" Num examples = %d", len(predict_examples)) + tf.logging.info(" Batch size = %d", FLAGS.predict_batch_size) + + predict_input_fn = run_classifier.file_based_input_fn_builder( + input_file=predict_file, + seq_length=FLAGS.max_seq_length, + is_training=False, + drop_remainder=FLAGS.use_tpu) + + result = estimator.predict(input_fn=predict_input_fn) + + output_predict_file = os.path.join(FLAGS.output_dir, "test_results.tsv") + with tf.gfile.GFile(output_predict_file, "w") as writer: + tf.logging.info("***** Predict results *****") + for prediction in result: + probabilities = prediction["probabilities"] + output_line = "\t".join( + str(class_probability) + for class_probability in probabilities) + "\n" + writer.write(output_line) + + +if __name__ == "__main__": + flags.mark_flag_as_required("data_dir") + flags.mark_flag_as_required("task_name") + flags.mark_flag_as_required("bert_hub_module_handle") + flags.mark_flag_as_required("output_dir") + tf.app.run() diff --git a/run_pretraining.py b/run_pretraining.py new file mode 100644 index 0000000..b118f62 --- /dev/null +++ b/run_pretraining.py @@ -0,0 +1,493 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Run masked LM/next sentence masked_lm pre-training for BERT.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import os +import modeling +import optimization +import tensorflow as tf + +flags = tf.flags + +FLAGS = flags.FLAGS + +## Required parameters +flags.DEFINE_string( + "bert_config_file", None, + "The config json file corresponding to the pre-trained BERT model. " + "This specifies the model architecture.") + +flags.DEFINE_string( + "input_file", None, + "Input TF example files (can be a glob or comma separated).") + +flags.DEFINE_string( + "output_dir", None, + "The output directory where the model checkpoints will be written.") + +## Other parameters +flags.DEFINE_string( + "init_checkpoint", None, + "Initial checkpoint (usually from a pre-trained BERT model).") + +flags.DEFINE_integer( + "max_seq_length", 128, + "The maximum total input sequence length after WordPiece tokenization. " + "Sequences longer than this will be truncated, and sequences shorter " + "than this will be padded. Must match data generation.") + +flags.DEFINE_integer( + "max_predictions_per_seq", 20, + "Maximum number of masked LM predictions per sequence. " + "Must match data generation.") + +flags.DEFINE_bool("do_train", False, "Whether to run training.") + +flags.DEFINE_bool("do_eval", False, "Whether to run eval on the dev set.") + +flags.DEFINE_integer("train_batch_size", 32, "Total batch size for training.") + +flags.DEFINE_integer("eval_batch_size", 8, "Total batch size for eval.") + +flags.DEFINE_float("learning_rate", 5e-5, "The initial learning rate for Adam.") + +flags.DEFINE_integer("num_train_steps", 100000, "Number of training steps.") + +flags.DEFINE_integer("num_warmup_steps", 10000, "Number of warmup steps.") + +flags.DEFINE_integer("save_checkpoints_steps", 1000, + "How often to save the model checkpoint.") + +flags.DEFINE_integer("iterations_per_loop", 1000, + "How many steps to make in each estimator call.") + +flags.DEFINE_integer("max_eval_steps", 100, "Maximum number of eval steps.") + +flags.DEFINE_bool("use_tpu", False, "Whether to use TPU or GPU/CPU.") + +tf.flags.DEFINE_string( + "tpu_name", None, + "The Cloud TPU to use for training. This should be either the name " + "used when creating the Cloud TPU, or a grpc://ip.address.of.tpu:8470 " + "url.") + +tf.flags.DEFINE_string( + "tpu_zone", None, + "[Optional] GCE zone where the Cloud TPU is located in. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string( + "gcp_project", None, + "[Optional] Project name for the Cloud TPU-enabled project. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string("master", None, "[Optional] TensorFlow master URL.") + +flags.DEFINE_integer( + "num_tpu_cores", 8, + "Only used if `use_tpu` is True. Total number of TPU cores to use.") + + +def model_fn_builder(bert_config, init_checkpoint, learning_rate, + num_train_steps, num_warmup_steps, use_tpu, + use_one_hot_embeddings): + """Returns `model_fn` closure for TPUEstimator.""" + + def model_fn(features, labels, mode, params): # pylint: disable=unused-argument + """The `model_fn` for TPUEstimator.""" + + tf.logging.info("*** Features ***") + for name in sorted(features.keys()): + tf.logging.info(" name = %s, shape = %s" % (name, features[name].shape)) + + input_ids = features["input_ids"] + input_mask = features["input_mask"] + segment_ids = features["segment_ids"] + masked_lm_positions = features["masked_lm_positions"] + masked_lm_ids = features["masked_lm_ids"] + masked_lm_weights = features["masked_lm_weights"] + next_sentence_labels = features["next_sentence_labels"] + + is_training = (mode == tf.estimator.ModeKeys.TRAIN) + + model = modeling.BertModel( + config=bert_config, + is_training=is_training, + input_ids=input_ids, + input_mask=input_mask, + token_type_ids=segment_ids, + use_one_hot_embeddings=use_one_hot_embeddings) + + (masked_lm_loss, + masked_lm_example_loss, masked_lm_log_probs) = get_masked_lm_output( + bert_config, model.get_sequence_output(), model.get_embedding_table(), + masked_lm_positions, masked_lm_ids, masked_lm_weights) + + (next_sentence_loss, next_sentence_example_loss, + next_sentence_log_probs) = get_next_sentence_output( + bert_config, model.get_pooled_output(), next_sentence_labels) + + total_loss = masked_lm_loss + next_sentence_loss + + tvars = tf.trainable_variables() + + initialized_variable_names = {} + scaffold_fn = None + if init_checkpoint: + (assignment_map, initialized_variable_names + ) = modeling.get_assignment_map_from_checkpoint(tvars, init_checkpoint) + if use_tpu: + + def tpu_scaffold(): + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + return tf.train.Scaffold() + + scaffold_fn = tpu_scaffold + else: + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + + tf.logging.info("**** Trainable Variables ****") + for var in tvars: + init_string = "" + if var.name in initialized_variable_names: + init_string = ", *INIT_FROM_CKPT*" + tf.logging.info(" name = %s, shape = %s%s", var.name, var.shape, + init_string) + + output_spec = None + if mode == tf.estimator.ModeKeys.TRAIN: + train_op = optimization.create_optimizer( + total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu) + + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + train_op=train_op, + scaffold_fn=scaffold_fn) + elif mode == tf.estimator.ModeKeys.EVAL: + + def metric_fn(masked_lm_example_loss, masked_lm_log_probs, masked_lm_ids, + masked_lm_weights, next_sentence_example_loss, + next_sentence_log_probs, next_sentence_labels): + """Computes the loss and accuracy of the model.""" + masked_lm_log_probs = tf.reshape(masked_lm_log_probs, + [-1, masked_lm_log_probs.shape[-1]]) + masked_lm_predictions = tf.argmax( + masked_lm_log_probs, axis=-1, output_type=tf.int32) + masked_lm_example_loss = tf.reshape(masked_lm_example_loss, [-1]) + masked_lm_ids = tf.reshape(masked_lm_ids, [-1]) + masked_lm_weights = tf.reshape(masked_lm_weights, [-1]) + masked_lm_accuracy = tf.metrics.accuracy( + labels=masked_lm_ids, + predictions=masked_lm_predictions, + weights=masked_lm_weights) + masked_lm_mean_loss = tf.metrics.mean( + values=masked_lm_example_loss, weights=masked_lm_weights) + + next_sentence_log_probs = tf.reshape( + next_sentence_log_probs, [-1, next_sentence_log_probs.shape[-1]]) + next_sentence_predictions = tf.argmax( + next_sentence_log_probs, axis=-1, output_type=tf.int32) + next_sentence_labels = tf.reshape(next_sentence_labels, [-1]) + next_sentence_accuracy = tf.metrics.accuracy( + labels=next_sentence_labels, predictions=next_sentence_predictions) + next_sentence_mean_loss = tf.metrics.mean( + values=next_sentence_example_loss) + + return { + "masked_lm_accuracy": masked_lm_accuracy, + "masked_lm_loss": masked_lm_mean_loss, + "next_sentence_accuracy": next_sentence_accuracy, + "next_sentence_loss": next_sentence_mean_loss, + } + + eval_metrics = (metric_fn, [ + masked_lm_example_loss, masked_lm_log_probs, masked_lm_ids, + masked_lm_weights, next_sentence_example_loss, + next_sentence_log_probs, next_sentence_labels + ]) + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + eval_metrics=eval_metrics, + scaffold_fn=scaffold_fn) + else: + raise ValueError("Only TRAIN and EVAL modes are supported: %s" % (mode)) + + return output_spec + + return model_fn + + +def get_masked_lm_output(bert_config, input_tensor, output_weights, positions, + label_ids, label_weights): + """Get loss and log probs for the masked LM.""" + input_tensor = gather_indexes(input_tensor, positions) + + with tf.variable_scope("cls/predictions"): + # We apply one more non-linear transformation before the output layer. + # This matrix is not used after pre-training. + with tf.variable_scope("transform"): + input_tensor = tf.layers.dense( + input_tensor, + units=bert_config.hidden_size, + activation=modeling.get_activation(bert_config.hidden_act), + kernel_initializer=modeling.create_initializer( + bert_config.initializer_range)) + input_tensor = modeling.layer_norm(input_tensor) + + # The output weights are the same as the input embeddings, but there is + # an output-only bias for each token. + output_bias = tf.get_variable( + "output_bias", + shape=[bert_config.vocab_size], + initializer=tf.zeros_initializer()) + logits = tf.matmul(input_tensor, output_weights, transpose_b=True) + logits = tf.nn.bias_add(logits, output_bias) + log_probs = tf.nn.log_softmax(logits, axis=-1) + + label_ids = tf.reshape(label_ids, [-1]) + label_weights = tf.reshape(label_weights, [-1]) + + one_hot_labels = tf.one_hot( + label_ids, depth=bert_config.vocab_size, dtype=tf.float32) + + # The `positions` tensor might be zero-padded (if the sequence is too + # short to have the maximum number of predictions). The `label_weights` + # tensor has a value of 1.0 for every real prediction and 0.0 for the + # padding predictions. + per_example_loss = -tf.reduce_sum(log_probs * one_hot_labels, axis=[-1]) + numerator = tf.reduce_sum(label_weights * per_example_loss) + denominator = tf.reduce_sum(label_weights) + 1e-5 + loss = numerator / denominator + + return (loss, per_example_loss, log_probs) + + +def get_next_sentence_output(bert_config, input_tensor, labels): + """Get loss and log probs for the next sentence prediction.""" + + # Simple binary classification. Note that 0 is "next sentence" and 1 is + # "random sentence". This weight matrix is not used after pre-training. + with tf.variable_scope("cls/seq_relationship"): + output_weights = tf.get_variable( + "output_weights", + shape=[2, bert_config.hidden_size], + initializer=modeling.create_initializer(bert_config.initializer_range)) + output_bias = tf.get_variable( + "output_bias", shape=[2], initializer=tf.zeros_initializer()) + + logits = tf.matmul(input_tensor, output_weights, transpose_b=True) + logits = tf.nn.bias_add(logits, output_bias) + log_probs = tf.nn.log_softmax(logits, axis=-1) + labels = tf.reshape(labels, [-1]) + one_hot_labels = tf.one_hot(labels, depth=2, dtype=tf.float32) + per_example_loss = -tf.reduce_sum(one_hot_labels * log_probs, axis=-1) + loss = tf.reduce_mean(per_example_loss) + return (loss, per_example_loss, log_probs) + + +def gather_indexes(sequence_tensor, positions): + """Gathers the vectors at the specific positions over a minibatch.""" + sequence_shape = modeling.get_shape_list(sequence_tensor, expected_rank=3) + batch_size = sequence_shape[0] + seq_length = sequence_shape[1] + width = sequence_shape[2] + + flat_offsets = tf.reshape( + tf.range(0, batch_size, dtype=tf.int32) * seq_length, [-1, 1]) + flat_positions = tf.reshape(positions + flat_offsets, [-1]) + flat_sequence_tensor = tf.reshape(sequence_tensor, + [batch_size * seq_length, width]) + output_tensor = tf.gather(flat_sequence_tensor, flat_positions) + return output_tensor + + +def input_fn_builder(input_files, + max_seq_length, + max_predictions_per_seq, + is_training, + num_cpu_threads=4): + """Creates an `input_fn` closure to be passed to TPUEstimator.""" + + def input_fn(params): + """The actual input function.""" + batch_size = params["batch_size"] + + name_to_features = { + "input_ids": + tf.FixedLenFeature([max_seq_length], tf.int64), + "input_mask": + tf.FixedLenFeature([max_seq_length], tf.int64), + "segment_ids": + tf.FixedLenFeature([max_seq_length], tf.int64), + "masked_lm_positions": + tf.FixedLenFeature([max_predictions_per_seq], tf.int64), + "masked_lm_ids": + tf.FixedLenFeature([max_predictions_per_seq], tf.int64), + "masked_lm_weights": + tf.FixedLenFeature([max_predictions_per_seq], tf.float32), + "next_sentence_labels": + tf.FixedLenFeature([1], tf.int64), + } + + # For training, we want a lot of parallel reading and shuffling. + # For eval, we want no shuffling and parallel reading doesn't matter. + if is_training: + d = tf.data.Dataset.from_tensor_slices(tf.constant(input_files)) + d = d.repeat() + d = d.shuffle(buffer_size=len(input_files)) + + # `cycle_length` is the number of parallel files that get read. + cycle_length = min(num_cpu_threads, len(input_files)) + + # `sloppy` mode means that the interleaving is not exact. This adds + # even more randomness to the training pipeline. + d = d.apply( + tf.contrib.data.parallel_interleave( + tf.data.TFRecordDataset, + sloppy=is_training, + cycle_length=cycle_length)) + d = d.shuffle(buffer_size=100) + else: + d = tf.data.TFRecordDataset(input_files) + # Since we evaluate for a fixed number of steps we don't want to encounter + # out-of-range exceptions. + d = d.repeat() + + # We must `drop_remainder` on training because the TPU requires fixed + # size dimensions. For eval, we assume we are evaluating on the CPU or GPU + # and we *don't* want to drop the remainder, otherwise we wont cover + # every sample. + d = d.apply( + tf.contrib.data.map_and_batch( + lambda record: _decode_record(record, name_to_features), + batch_size=batch_size, + num_parallel_batches=num_cpu_threads, + drop_remainder=True)) + return d + + return input_fn + + +def _decode_record(record, name_to_features): + """Decodes a record to a TensorFlow example.""" + example = tf.parse_single_example(record, name_to_features) + + # tf.Example only supports tf.int64, but the TPU only supports tf.int32. + # So cast all int64 to int32. + for name in list(example.keys()): + t = example[name] + if t.dtype == tf.int64: + t = tf.to_int32(t) + example[name] = t + + return example + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + if not FLAGS.do_train and not FLAGS.do_eval: + raise ValueError("At least one of `do_train` or `do_eval` must be True.") + + bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file) + + tf.gfile.MakeDirs(FLAGS.output_dir) + + input_files = [] + for input_pattern in FLAGS.input_file.split(","): + input_files.extend(tf.gfile.Glob(input_pattern)) + + tf.logging.info("*** Input Files ***") + for input_file in input_files: + tf.logging.info(" %s" % input_file) + + tpu_cluster_resolver = None + if FLAGS.use_tpu and FLAGS.tpu_name: + tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver( + FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project) + + is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2 + run_config = tf.contrib.tpu.RunConfig( + cluster=tpu_cluster_resolver, + master=FLAGS.master, + model_dir=FLAGS.output_dir, + save_checkpoints_steps=FLAGS.save_checkpoints_steps, + tpu_config=tf.contrib.tpu.TPUConfig( + iterations_per_loop=FLAGS.iterations_per_loop, + num_shards=FLAGS.num_tpu_cores, + per_host_input_for_training=is_per_host)) + + model_fn = model_fn_builder( + bert_config=bert_config, + init_checkpoint=FLAGS.init_checkpoint, + learning_rate=FLAGS.learning_rate, + num_train_steps=FLAGS.num_train_steps, + num_warmup_steps=FLAGS.num_warmup_steps, + use_tpu=FLAGS.use_tpu, + use_one_hot_embeddings=FLAGS.use_tpu) + + # If TPU is not available, this will fall back to normal Estimator on CPU + # or GPU. + estimator = tf.contrib.tpu.TPUEstimator( + use_tpu=FLAGS.use_tpu, + model_fn=model_fn, + config=run_config, + train_batch_size=FLAGS.train_batch_size, + eval_batch_size=FLAGS.eval_batch_size) + + if FLAGS.do_train: + tf.logging.info("***** Running training *****") + tf.logging.info(" Batch size = %d", FLAGS.train_batch_size) + train_input_fn = input_fn_builder( + input_files=input_files, + max_seq_length=FLAGS.max_seq_length, + max_predictions_per_seq=FLAGS.max_predictions_per_seq, + is_training=True) + estimator.train(input_fn=train_input_fn, max_steps=FLAGS.num_train_steps) + + if FLAGS.do_eval: + tf.logging.info("***** Running evaluation *****") + tf.logging.info(" Batch size = %d", FLAGS.eval_batch_size) + + eval_input_fn = input_fn_builder( + input_files=input_files, + max_seq_length=FLAGS.max_seq_length, + max_predictions_per_seq=FLAGS.max_predictions_per_seq, + is_training=False) + + result = estimator.evaluate( + input_fn=eval_input_fn, steps=FLAGS.max_eval_steps) + + output_eval_file = os.path.join(FLAGS.output_dir, "eval_results.txt") + with tf.gfile.GFile(output_eval_file, "w") as writer: + tf.logging.info("***** Eval results *****") + for key in sorted(result.keys()): + tf.logging.info(" %s = %s", key, str(result[key])) + writer.write("%s = %s\n" % (key, str(result[key]))) + + +if __name__ == "__main__": + flags.mark_flag_as_required("input_file") + flags.mark_flag_as_required("bert_config_file") + flags.mark_flag_as_required("output_dir") + tf.app.run() diff --git a/run_squad.py b/run_squad.py new file mode 100644 index 0000000..edd4c3e --- /dev/null +++ b/run_squad.py @@ -0,0 +1,1283 @@ +# coding=utf-8 +# Copyright 2018 The Google AI Language Team Authors. +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +"""Run BERT on SQuAD 1.1 and SQuAD 2.0.""" + +from __future__ import absolute_import +from __future__ import division +from __future__ import print_function + +import collections +import json +import math +import os +import random +import modeling +import optimization +import tokenization +import six +import tensorflow as tf + +flags = tf.flags + +FLAGS = flags.FLAGS + +## Required parameters +flags.DEFINE_string( + "bert_config_file", None, + "The config json file corresponding to the pre-trained BERT model. " + "This specifies the model architecture.") + +flags.DEFINE_string("vocab_file", None, + "The vocabulary file that the BERT model was trained on.") + +flags.DEFINE_string( + "output_dir", None, + "The output directory where the model checkpoints will be written.") + +## Other parameters +flags.DEFINE_string("train_file", None, + "SQuAD json for training. E.g., train-v1.1.json") + +flags.DEFINE_string( + "predict_file", None, + "SQuAD json for predictions. E.g., dev-v1.1.json or test-v1.1.json") + +flags.DEFINE_string( + "init_checkpoint", None, + "Initial checkpoint (usually from a pre-trained BERT model).") + +flags.DEFINE_bool( + "do_lower_case", True, + "Whether to lower case the input text. Should be True for uncased " + "models and False for cased models.") + +flags.DEFINE_integer( + "max_seq_length", 384, + "The maximum total input sequence length after WordPiece tokenization. " + "Sequences longer than this will be truncated, and sequences shorter " + "than this will be padded.") + +flags.DEFINE_integer( + "doc_stride", 128, + "When splitting up a long document into chunks, how much stride to " + "take between chunks.") + +flags.DEFINE_integer( + "max_query_length", 64, + "The maximum number of tokens for the question. Questions longer than " + "this will be truncated to this length.") + +flags.DEFINE_bool("do_train", False, "Whether to run training.") + +flags.DEFINE_bool("do_predict", False, "Whether to run eval on the dev set.") + +flags.DEFINE_integer("train_batch_size", 32, "Total batch size for training.") + +flags.DEFINE_integer("predict_batch_size", 8, + "Total batch size for predictions.") + +flags.DEFINE_float("learning_rate", 5e-5, "The initial learning rate for Adam.") + +flags.DEFINE_float("num_train_epochs", 3.0, + "Total number of training epochs to perform.") + +flags.DEFINE_float( + "warmup_proportion", 0.1, + "Proportion of training to perform linear learning rate warmup for. " + "E.g., 0.1 = 10% of training.") + +flags.DEFINE_integer("save_checkpoints_steps", 1000, + "How often to save the model checkpoint.") + +flags.DEFINE_integer("iterations_per_loop", 1000, + "How many steps to make in each estimator call.") + +flags.DEFINE_integer( + "n_best_size", 20, + "The total number of n-best predictions to generate in the " + "nbest_predictions.json output file.") + +flags.DEFINE_integer( + "max_answer_length", 30, + "The maximum length of an answer that can be generated. This is needed " + "because the start and end predictions are not conditioned on one another.") + +flags.DEFINE_bool("use_tpu", False, "Whether to use TPU or GPU/CPU.") + +tf.flags.DEFINE_string( + "tpu_name", None, + "The Cloud TPU to use for training. This should be either the name " + "used when creating the Cloud TPU, or a grpc://ip.address.of.tpu:8470 " + "url.") + +tf.flags.DEFINE_string( + "tpu_zone", None, + "[Optional] GCE zone where the Cloud TPU is located in. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string( + "gcp_project", None, + "[Optional] Project name for the Cloud TPU-enabled project. If not " + "specified, we will attempt to automatically detect the GCE project from " + "metadata.") + +tf.flags.DEFINE_string("master", None, "[Optional] TensorFlow master URL.") + +flags.DEFINE_integer( + "num_tpu_cores", 8, + "Only used if `use_tpu` is True. Total number of TPU cores to use.") + +flags.DEFINE_bool( + "verbose_logging", False, + "If true, all of the warnings related to data processing will be printed. " + "A number of warnings are expected for a normal SQuAD evaluation.") + +flags.DEFINE_bool( + "version_2_with_negative", False, + "If true, the SQuAD examples contain some that do not have an answer.") + +flags.DEFINE_float( + "null_score_diff_threshold", 0.0, + "If null_score - best_non_null is greater than the threshold predict null.") + + +class SquadExample(object): + """A single training/test example for simple sequence classification. + + For examples without an answer, the start and end position are -1. + """ + + def __init__(self, + qas_id, + question_text, + doc_tokens, + orig_answer_text=None, + start_position=None, + end_position=None, + is_impossible=False): + self.qas_id = qas_id + self.question_text = question_text + self.doc_tokens = doc_tokens + self.orig_answer_text = orig_answer_text + self.start_position = start_position + self.end_position = end_position + self.is_impossible = is_impossible + + def __str__(self): + return self.__repr__() + + def __repr__(self): + s = "" + s += "qas_id: %s" % (tokenization.printable_text(self.qas_id)) + s += ", question_text: %s" % ( + tokenization.printable_text(self.question_text)) + s += ", doc_tokens: [%s]" % (" ".join(self.doc_tokens)) + if self.start_position: + s += ", start_position: %d" % (self.start_position) + if self.start_position: + s += ", end_position: %d" % (self.end_position) + if self.start_position: + s += ", is_impossible: %r" % (self.is_impossible) + return s + + +class InputFeatures(object): + """A single set of features of data.""" + + def __init__(self, + unique_id, + example_index, + doc_span_index, + tokens, + token_to_orig_map, + token_is_max_context, + input_ids, + input_mask, + segment_ids, + start_position=None, + end_position=None, + is_impossible=None): + self.unique_id = unique_id + self.example_index = example_index + self.doc_span_index = doc_span_index + self.tokens = tokens + self.token_to_orig_map = token_to_orig_map + self.token_is_max_context = token_is_max_context + self.input_ids = input_ids + self.input_mask = input_mask + self.segment_ids = segment_ids + self.start_position = start_position + self.end_position = end_position + self.is_impossible = is_impossible + + +def read_squad_examples(input_file, is_training): + """Read a SQuAD json file into a list of SquadExample.""" + with tf.gfile.Open(input_file, "r") as reader: + input_data = json.load(reader)["data"] + + def is_whitespace(c): + if c == " " or c == "\t" or c == "\r" or c == "\n" or ord(c) == 0x202F: + return True + return False + + examples = [] + for entry in input_data: + for paragraph in entry["paragraphs"]: + paragraph_text = paragraph["context"] + doc_tokens = [] + char_to_word_offset = [] + prev_is_whitespace = True + for c in paragraph_text: + if is_whitespace(c): + prev_is_whitespace = True + else: + if prev_is_whitespace: + doc_tokens.append(c) + else: + doc_tokens[-1] += c + prev_is_whitespace = False + char_to_word_offset.append(len(doc_tokens) - 1) + + for qa in paragraph["qas"]: + qas_id = qa["id"] + question_text = qa["question"] + start_position = None + end_position = None + orig_answer_text = None + is_impossible = False + if is_training: + + if FLAGS.version_2_with_negative: + is_impossible = qa["is_impossible"] + if (len(qa["answers"]) != 1) and (not is_impossible): + raise ValueError( + "For training, each question should have exactly 1 answer.") + if not is_impossible: + answer = qa["answers"][0] + orig_answer_text = answer["text"] + answer_offset = answer["answer_start"] + answer_length = len(orig_answer_text) + start_position = char_to_word_offset[answer_offset] + end_position = char_to_word_offset[answer_offset + answer_length - + 1] + # Only add answers where the text can be exactly recovered from the + # document. If this CAN'T happen it's likely due to weird Unicode + # stuff so we will just skip the example. + # + # Note that this means for training mode, every example is NOT + # guaranteed to be preserved. + actual_text = " ".join( + doc_tokens[start_position:(end_position + 1)]) + cleaned_answer_text = " ".join( + tokenization.whitespace_tokenize(orig_answer_text)) + if actual_text.find(cleaned_answer_text) == -1: + tf.logging.warning("Could not find answer: '%s' vs. '%s'", + actual_text, cleaned_answer_text) + continue + else: + start_position = -1 + end_position = -1 + orig_answer_text = "" + + example = SquadExample( + qas_id=qas_id, + question_text=question_text, + doc_tokens=doc_tokens, + orig_answer_text=orig_answer_text, + start_position=start_position, + end_position=end_position, + is_impossible=is_impossible) + examples.append(example) + + return examples + + +def convert_examples_to_features(examples, tokenizer, max_seq_length, + doc_stride, max_query_length, is_training, + output_fn): + """Loads a data file into a list of `InputBatch`s.""" + + unique_id = 1000000000 + + for (example_index, example) in enumerate(examples): + query_tokens = tokenizer.tokenize(example.question_text) + + if len(query_tokens) > max_query_length: + query_tokens = query_tokens[0:max_query_length] + + tok_to_orig_index = [] + orig_to_tok_index = [] + all_doc_tokens = [] + for (i, token) in enumerate(example.doc_tokens): + orig_to_tok_index.append(len(all_doc_tokens)) + sub_tokens = tokenizer.tokenize(token) + for sub_token in sub_tokens: + tok_to_orig_index.append(i) + all_doc_tokens.append(sub_token) + + tok_start_position = None + tok_end_position = None + if is_training and example.is_impossible: + tok_start_position = -1 + tok_end_position = -1 + if is_training and not example.is_impossible: + tok_start_position = orig_to_tok_index[example.start_position] + if example.end_position < len(example.doc_tokens) - 1: + tok_end_position = orig_to_tok_index[example.end_position + 1] - 1 + else: + tok_end_position = len(all_doc_tokens) - 1 + (tok_start_position, tok_end_position) = _improve_answer_span( + all_doc_tokens, tok_start_position, tok_end_position, tokenizer, + example.orig_answer_text) + + # The -3 accounts for [CLS], [SEP] and [SEP] + max_tokens_for_doc = max_seq_length - len(query_tokens) - 3 + + # We can have documents that are longer than the maximum sequence length. + # To deal with this we do a sliding window approach, where we take chunks + # of the up to our max length with a stride of `doc_stride`. + _DocSpan = collections.namedtuple( # pylint: disable=invalid-name + "DocSpan", ["start", "length"]) + doc_spans = [] + start_offset = 0 + while start_offset < len(all_doc_tokens): + length = len(all_doc_tokens) - start_offset + if length > max_tokens_for_doc: + length = max_tokens_for_doc + doc_spans.append(_DocSpan(start=start_offset, length=length)) + if start_offset + length == len(all_doc_tokens): + break + start_offset += min(length, doc_stride) + + for (doc_span_index, doc_span) in enumerate(doc_spans): + tokens = [] + token_to_orig_map = {} + token_is_max_context = {} + segment_ids = [] + tokens.append("[CLS]") + segment_ids.append(0) + for token in query_tokens: + tokens.append(token) + segment_ids.append(0) + tokens.append("[SEP]") + segment_ids.append(0) + + for i in range(doc_span.length): + split_token_index = doc_span.start + i + token_to_orig_map[len(tokens)] = tok_to_orig_index[split_token_index] + + is_max_context = _check_is_max_context(doc_spans, doc_span_index, + split_token_index) + token_is_max_context[len(tokens)] = is_max_context + tokens.append(all_doc_tokens[split_token_index]) + segment_ids.append(1) + tokens.append("[SEP]") + segment_ids.append(1) + + input_ids = tokenizer.convert_tokens_to_ids(tokens) + + # The mask has 1 for real tokens and 0 for padding tokens. Only real + # tokens are attended to. + input_mask = [1] * len(input_ids) + + # Zero-pad up to the sequence length. + while len(input_ids) < max_seq_length: + input_ids.append(0) + input_mask.append(0) + segment_ids.append(0) + + assert len(input_ids) == max_seq_length + assert len(input_mask) == max_seq_length + assert len(segment_ids) == max_seq_length + + start_position = None + end_position = None + if is_training and not example.is_impossible: + # For training, if our document chunk does not contain an annotation + # we throw it out, since there is nothing to predict. + doc_start = doc_span.start + doc_end = doc_span.start + doc_span.length - 1 + out_of_span = False + if not (tok_start_position >= doc_start and + tok_end_position <= doc_end): + out_of_span = True + if out_of_span: + start_position = 0 + end_position = 0 + else: + doc_offset = len(query_tokens) + 2 + start_position = tok_start_position - doc_start + doc_offset + end_position = tok_end_position - doc_start + doc_offset + + if is_training and example.is_impossible: + start_position = 0 + end_position = 0 + + if example_index < 20: + tf.logging.info("*** Example ***") + tf.logging.info("unique_id: %s" % (unique_id)) + tf.logging.info("example_index: %s" % (example_index)) + tf.logging.info("doc_span_index: %s" % (doc_span_index)) + tf.logging.info("tokens: %s" % " ".join( + [tokenization.printable_text(x) for x in tokens])) + tf.logging.info("token_to_orig_map: %s" % " ".join( + ["%d:%d" % (x, y) for (x, y) in six.iteritems(token_to_orig_map)])) + tf.logging.info("token_is_max_context: %s" % " ".join([ + "%d:%s" % (x, y) for (x, y) in six.iteritems(token_is_max_context) + ])) + tf.logging.info("input_ids: %s" % " ".join([str(x) for x in input_ids])) + tf.logging.info( + "input_mask: %s" % " ".join([str(x) for x in input_mask])) + tf.logging.info( + "segment_ids: %s" % " ".join([str(x) for x in segment_ids])) + if is_training and example.is_impossible: + tf.logging.info("impossible example") + if is_training and not example.is_impossible: + answer_text = " ".join(tokens[start_position:(end_position + 1)]) + tf.logging.info("start_position: %d" % (start_position)) + tf.logging.info("end_position: %d" % (end_position)) + tf.logging.info( + "answer: %s" % (tokenization.printable_text(answer_text))) + + feature = InputFeatures( + unique_id=unique_id, + example_index=example_index, + doc_span_index=doc_span_index, + tokens=tokens, + token_to_orig_map=token_to_orig_map, + token_is_max_context=token_is_max_context, + input_ids=input_ids, + input_mask=input_mask, + segment_ids=segment_ids, + start_position=start_position, + end_position=end_position, + is_impossible=example.is_impossible) + + # Run callback + output_fn(feature) + + unique_id += 1 + + +def _improve_answer_span(doc_tokens, input_start, input_end, tokenizer, + orig_answer_text): + """Returns tokenized answer spans that better match the annotated answer.""" + + # The SQuAD annotations are character based. We first project them to + # whitespace-tokenized words. But then after WordPiece tokenization, we can + # often find a "better match". For example: + # + # Question: What year was John Smith born? + # Context: The leader was John Smith (1895-1943). + # Answer: 1895 + # + # The original whitespace-tokenized answer will be "(1895-1943).". However + # after tokenization, our tokens will be "( 1895 - 1943 ) .". So we can match + # the exact answer, 1895. + # + # However, this is not always possible. Consider the following: + # + # Question: What country is the top exporter of electornics? + # Context: The Japanese electronics industry is the lagest in the world. + # Answer: Japan + # + # In this case, the annotator chose "Japan" as a character sub-span of + # the word "Japanese". Since our WordPiece tokenizer does not split + # "Japanese", we just use "Japanese" as the annotation. This is fairly rare + # in SQuAD, but does happen. + tok_answer_text = " ".join(tokenizer.tokenize(orig_answer_text)) + + for new_start in range(input_start, input_end + 1): + for new_end in range(input_end, new_start - 1, -1): + text_span = " ".join(doc_tokens[new_start:(new_end + 1)]) + if text_span == tok_answer_text: + return (new_start, new_end) + + return (input_start, input_end) + + +def _check_is_max_context(doc_spans, cur_span_index, position): + """Check if this is the 'max context' doc span for the token.""" + + # Because of the sliding window approach taken to scoring documents, a single + # token can appear in multiple documents. E.g. + # Doc: the man went to the store and bought a gallon of milk + # Span A: the man went to the + # Span B: to the store and bought + # Span C: and bought a gallon of + # ... + # + # Now the word 'bought' will have two scores from spans B and C. We only + # want to consider the score with "maximum context", which we define as + # the *minimum* of its left and right context (the *sum* of left and + # right context will always be the same, of course). + # + # In the example the maximum context for 'bought' would be span C since + # it has 1 left context and 3 right context, while span B has 4 left context + # and 0 right context. + best_score = None + best_span_index = None + for (span_index, doc_span) in enumerate(doc_spans): + end = doc_span.start + doc_span.length - 1 + if position < doc_span.start: + continue + if position > end: + continue + num_left_context = position - doc_span.start + num_right_context = end - position + score = min(num_left_context, num_right_context) + 0.01 * doc_span.length + if best_score is None or score > best_score: + best_score = score + best_span_index = span_index + + return cur_span_index == best_span_index + + +def create_model(bert_config, is_training, input_ids, input_mask, segment_ids, + use_one_hot_embeddings): + """Creates a classification model.""" + model = modeling.BertModel( + config=bert_config, + is_training=is_training, + input_ids=input_ids, + input_mask=input_mask, + token_type_ids=segment_ids, + use_one_hot_embeddings=use_one_hot_embeddings) + + final_hidden = model.get_sequence_output() + + final_hidden_shape = modeling.get_shape_list(final_hidden, expected_rank=3) + batch_size = final_hidden_shape[0] + seq_length = final_hidden_shape[1] + hidden_size = final_hidden_shape[2] + + output_weights = tf.get_variable( + "cls/squad/output_weights", [2, hidden_size], + initializer=tf.truncated_normal_initializer(stddev=0.02)) + + output_bias = tf.get_variable( + "cls/squad/output_bias", [2], initializer=tf.zeros_initializer()) + + final_hidden_matrix = tf.reshape(final_hidden, + [batch_size * seq_length, hidden_size]) + logits = tf.matmul(final_hidden_matrix, output_weights, transpose_b=True) + logits = tf.nn.bias_add(logits, output_bias) + + logits = tf.reshape(logits, [batch_size, seq_length, 2]) + logits = tf.transpose(logits, [2, 0, 1]) + + unstacked_logits = tf.unstack(logits, axis=0) + + (start_logits, end_logits) = (unstacked_logits[0], unstacked_logits[1]) + + return (start_logits, end_logits) + + +def model_fn_builder(bert_config, init_checkpoint, learning_rate, + num_train_steps, num_warmup_steps, use_tpu, + use_one_hot_embeddings): + """Returns `model_fn` closure for TPUEstimator.""" + + def model_fn(features, labels, mode, params): # pylint: disable=unused-argument + """The `model_fn` for TPUEstimator.""" + + tf.logging.info("*** Features ***") + for name in sorted(features.keys()): + tf.logging.info(" name = %s, shape = %s" % (name, features[name].shape)) + + unique_ids = features["unique_ids"] + input_ids = features["input_ids"] + input_mask = features["input_mask"] + segment_ids = features["segment_ids"] + + is_training = (mode == tf.estimator.ModeKeys.TRAIN) + + (start_logits, end_logits) = create_model( + bert_config=bert_config, + is_training=is_training, + input_ids=input_ids, + input_mask=input_mask, + segment_ids=segment_ids, + use_one_hot_embeddings=use_one_hot_embeddings) + + tvars = tf.trainable_variables() + + initialized_variable_names = {} + scaffold_fn = None + if init_checkpoint: + (assignment_map, initialized_variable_names + ) = modeling.get_assignment_map_from_checkpoint(tvars, init_checkpoint) + if use_tpu: + + def tpu_scaffold(): + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + return tf.train.Scaffold() + + scaffold_fn = tpu_scaffold + else: + tf.train.init_from_checkpoint(init_checkpoint, assignment_map) + + tf.logging.info("**** Trainable Variables ****") + for var in tvars: + init_string = "" + if var.name in initialized_variable_names: + init_string = ", *INIT_FROM_CKPT*" + tf.logging.info(" name = %s, shape = %s%s", var.name, var.shape, + init_string) + + output_spec = None + if mode == tf.estimator.ModeKeys.TRAIN: + seq_length = modeling.get_shape_list(input_ids)[1] + + def compute_loss(logits, positions): + one_hot_positions = tf.one_hot( + positions, depth=seq_length, dtype=tf.float32) + log_probs = tf.nn.log_softmax(logits, axis=-1) + loss = -tf.reduce_mean( + tf.reduce_sum(one_hot_positions * log_probs, axis=-1)) + return loss + + start_positions = features["start_positions"] + end_positions = features["end_positions"] + + start_loss = compute_loss(start_logits, start_positions) + end_loss = compute_loss(end_logits, end_positions) + + total_loss = (start_loss + end_loss) / 2.0 + + train_op = optimization.create_optimizer( + total_loss, learning_rate, num_train_steps, num_warmup_steps, use_tpu) + + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, + loss=total_loss, + train_op=train_op, + scaffold_fn=scaffold_fn) + elif mode == tf.estimator.ModeKeys.PREDICT: + predictions = { + "unique_ids": unique_ids, + "start_logits": start_logits, + "end_logits": end_logits, + } + output_spec = tf.contrib.tpu.TPUEstimatorSpec( + mode=mode, predictions=predictions, scaffold_fn=scaffold_fn) + else: + raise ValueError( + "Only TRAIN and PREDICT modes are supported: %s" % (mode)) + + return output_spec + + return model_fn + + +def input_fn_builder(input_file, seq_length, is_training, drop_remainder): + """Creates an `input_fn` closure to be passed to TPUEstimator.""" + + name_to_features = { + "unique_ids": tf.FixedLenFeature([], tf.int64), + "input_ids": tf.FixedLenFeature([seq_length], tf.int64), + "input_mask": tf.FixedLenFeature([seq_length], tf.int64), + "segment_ids": tf.FixedLenFeature([seq_length], tf.int64), + } + + if is_training: + name_to_features["start_positions"] = tf.FixedLenFeature([], tf.int64) + name_to_features["end_positions"] = tf.FixedLenFeature([], tf.int64) + + def _decode_record(record, name_to_features): + """Decodes a record to a TensorFlow example.""" + example = tf.parse_single_example(record, name_to_features) + + # tf.Example only supports tf.int64, but the TPU only supports tf.int32. + # So cast all int64 to int32. + for name in list(example.keys()): + t = example[name] + if t.dtype == tf.int64: + t = tf.to_int32(t) + example[name] = t + + return example + + def input_fn(params): + """The actual input function.""" + batch_size = params["batch_size"] + + # For training, we want a lot of parallel reading and shuffling. + # For eval, we want no shuffling and parallel reading doesn't matter. + d = tf.data.TFRecordDataset(input_file) + if is_training: + d = d.repeat() + d = d.shuffle(buffer_size=100) + + d = d.apply( + tf.contrib.data.map_and_batch( + lambda record: _decode_record(record, name_to_features), + batch_size=batch_size, + drop_remainder=drop_remainder)) + + return d + + return input_fn + + +RawResult = collections.namedtuple("RawResult", + ["unique_id", "start_logits", "end_logits"]) + + +def write_predictions(all_examples, all_features, all_results, n_best_size, + max_answer_length, do_lower_case, output_prediction_file, + output_nbest_file, output_null_log_odds_file): + """Write final predictions to the json file and log-odds of null if needed.""" + tf.logging.info("Writing predictions to: %s" % (output_prediction_file)) + tf.logging.info("Writing nbest to: %s" % (output_nbest_file)) + + example_index_to_features = collections.defaultdict(list) + for feature in all_features: + example_index_to_features[feature.example_index].append(feature) + + unique_id_to_result = {} + for result in all_results: + unique_id_to_result[result.unique_id] = result + + _PrelimPrediction = collections.namedtuple( # pylint: disable=invalid-name + "PrelimPrediction", + ["feature_index", "start_index", "end_index", "start_logit", "end_logit"]) + + all_predictions = collections.OrderedDict() + all_nbest_json = collections.OrderedDict() + scores_diff_json = collections.OrderedDict() + + for (example_index, example) in enumerate(all_examples): + features = example_index_to_features[example_index] + + prelim_predictions = [] + # keep track of the minimum score of null start+end of position 0 + score_null = 1000000 # large and positive + min_null_feature_index = 0 # the paragraph slice with min mull score + null_start_logit = 0 # the start logit at the slice with min null score + null_end_logit = 0 # the end logit at the slice with min null score + for (feature_index, feature) in enumerate(features): + result = unique_id_to_result[feature.unique_id] + start_indexes = _get_best_indexes(result.start_logits, n_best_size) + end_indexes = _get_best_indexes(result.end_logits, n_best_size) + # if we could have irrelevant answers, get the min score of irrelevant + if FLAGS.version_2_with_negative: + feature_null_score = result.start_logits[0] + result.end_logits[0] + if feature_null_score < score_null: + score_null = feature_null_score + min_null_feature_index = feature_index + null_start_logit = result.start_logits[0] + null_end_logit = result.end_logits[0] + for start_index in start_indexes: + for end_index in end_indexes: + # We could hypothetically create invalid predictions, e.g., predict + # that the start of the span is in the question. We throw out all + # invalid predictions. + if start_index >= len(feature.tokens): + continue + if end_index >= len(feature.tokens): + continue + if start_index not in feature.token_to_orig_map: + continue + if end_index not in feature.token_to_orig_map: + continue + if not feature.token_is_max_context.get(start_index, False): + continue + if end_index < start_index: + continue + length = end_index - start_index + 1 + if length > max_answer_length: + continue + prelim_predictions.append( + _PrelimPrediction( + feature_index=feature_index, + start_index=start_index, + end_index=end_index, + start_logit=result.start_logits[start_index], + end_logit=result.end_logits[end_index])) + + if FLAGS.version_2_with_negative: + prelim_predictions.append( + _PrelimPrediction( + feature_index=min_null_feature_index, + start_index=0, + end_index=0, + start_logit=null_start_logit, + end_logit=null_end_logit)) + prelim_predictions = sorted( + prelim_predictions, + key=lambda x: (x.start_logit + x.end_logit), + reverse=True) + + _NbestPrediction = collections.namedtuple( # pylint: disable=invalid-name + "NbestPrediction", ["text", "start_logit", "end_logit"]) + + seen_predictions = {} + nbest = [] + for pred in prelim_predictions: + if len(nbest) >= n_best_size: + break + feature = features[pred.feature_index] + if pred.start_index > 0: # this is a non-null prediction + tok_tokens = feature.tokens[pred.start_index:(pred.end_index + 1)] + orig_doc_start = feature.token_to_orig_map[pred.start_index] + orig_doc_end = feature.token_to_orig_map[pred.end_index] + orig_tokens = example.doc_tokens[orig_doc_start:(orig_doc_end + 1)] + tok_text = " ".join(tok_tokens) + + # De-tokenize WordPieces that have been split off. + tok_text = tok_text.replace(" ##", "") + tok_text = tok_text.replace("##", "") + + # Clean whitespace + tok_text = tok_text.strip() + tok_text = " ".join(tok_text.split()) + orig_text = " ".join(orig_tokens) + + final_text = get_final_text(tok_text, orig_text, do_lower_case) + if final_text in seen_predictions: + continue + + seen_predictions[final_text] = True + else: + final_text = "" + seen_predictions[final_text] = True + + nbest.append( + _NbestPrediction( + text=final_text, + start_logit=pred.start_logit, + end_logit=pred.end_logit)) + + # if we didn't inlude the empty option in the n-best, inlcude it + if FLAGS.version_2_with_negative: + if "" not in seen_predictions: + nbest.append( + _NbestPrediction( + text="", start_logit=null_start_logit, + end_logit=null_end_logit)) + # In very rare edge cases we could have no valid predictions. So we + # just create a nonce prediction in this case to avoid failure. + if not nbest: + nbest.append( + _NbestPrediction(text="empty", start_logit=0.0, end_logit=0.0)) + + assert len(nbest) >= 1 + + total_scores = [] + best_non_null_entry = None + for entry in nbest: + total_scores.append(entry.start_logit + entry.end_logit) + if not best_non_null_entry: + if entry.text: + best_non_null_entry = entry + + probs = _compute_softmax(total_scores) + + nbest_json = [] + for (i, entry) in enumerate(nbest): + output = collections.OrderedDict() + output["text"] = entry.text + output["probability"] = probs[i] + output["start_logit"] = entry.start_logit + output["end_logit"] = entry.end_logit + nbest_json.append(output) + + assert len(nbest_json) >= 1 + + if not FLAGS.version_2_with_negative: + all_predictions[example.qas_id] = nbest_json[0]["text"] + else: + # predict "" iff the null score - the score of best non-null > threshold + score_diff = score_null - best_non_null_entry.start_logit - ( + best_non_null_entry.end_logit) + scores_diff_json[example.qas_id] = score_diff + if score_diff > FLAGS.null_score_diff_threshold: + all_predictions[example.qas_id] = "" + else: + all_predictions[example.qas_id] = best_non_null_entry.text + + all_nbest_json[example.qas_id] = nbest_json + + with tf.gfile.GFile(output_prediction_file, "w") as writer: + writer.write(json.dumps(all_predictions, indent=4) + "\n") + + with tf.gfile.GFile(output_nbest_file, "w") as writer: + writer.write(json.dumps(all_nbest_json, indent=4) + "\n") + + if FLAGS.version_2_with_negative: + with tf.gfile.GFile(output_null_log_odds_file, "w") as writer: + writer.write(json.dumps(scores_diff_json, indent=4) + "\n") + + +def get_final_text(pred_text, orig_text, do_lower_case): + """Project the tokenized prediction back to the original text.""" + + # When we created the data, we kept track of the alignment between original + # (whitespace tokenized) tokens and our WordPiece tokenized tokens. So + # now `orig_text` contains the span of our original text corresponding to the + # span that we predicted. + # + # However, `orig_text` may contain extra characters that we don't want in + # our prediction. + # + # For example, let's say: + # pred_text = steve smith + # orig_text = Steve Smith's + # + # We don't want to return `orig_text` because it contains the extra "'s". + # + # We don't want to return `pred_text` because it's already been normalized + # (the SQuAD eval script also does punctuation stripping/lower casing but + # our tokenizer does additional normalization like stripping accent + # characters). + # + # What we really want to return is "Steve Smith". + # + # Therefore, we have to apply a semi-complicated alignment heruistic between + # `pred_text` and `orig_text` to get a character-to-charcter alignment. This + # can fail in certain cases in which case we just return `orig_text`. + + def _strip_spaces(text): + ns_chars = [] + ns_to_s_map = collections.OrderedDict() + for (i, c) in enumerate(text): + if c == " ": + continue + ns_to_s_map[len(ns_chars)] = i + ns_chars.append(c) + ns_text = "".join(ns_chars) + return (ns_text, ns_to_s_map) + + # We first tokenize `orig_text`, strip whitespace from the result + # and `pred_text`, and check if they are the same length. If they are + # NOT the same length, the heuristic has failed. If they are the same + # length, we assume the characters are one-to-one aligned. + tokenizer = tokenization.BasicTokenizer(do_lower_case=do_lower_case) + + tok_text = " ".join(tokenizer.tokenize(orig_text)) + + start_position = tok_text.find(pred_text) + if start_position == -1: + if FLAGS.verbose_logging: + tf.logging.info( + "Unable to find text: '%s' in '%s'" % (pred_text, orig_text)) + return orig_text + end_position = start_position + len(pred_text) - 1 + + (orig_ns_text, orig_ns_to_s_map) = _strip_spaces(orig_text) + (tok_ns_text, tok_ns_to_s_map) = _strip_spaces(tok_text) + + if len(orig_ns_text) != len(tok_ns_text): + if FLAGS.verbose_logging: + tf.logging.info("Length not equal after stripping spaces: '%s' vs '%s'", + orig_ns_text, tok_ns_text) + return orig_text + + # We then project the characters in `pred_text` back to `orig_text` using + # the character-to-character alignment. + tok_s_to_ns_map = {} + for (i, tok_index) in six.iteritems(tok_ns_to_s_map): + tok_s_to_ns_map[tok_index] = i + + orig_start_position = None + if start_position in tok_s_to_ns_map: + ns_start_position = tok_s_to_ns_map[start_position] + if ns_start_position in orig_ns_to_s_map: + orig_start_position = orig_ns_to_s_map[ns_start_position] + + if orig_start_position is None: + if FLAGS.verbose_logging: + tf.logging.info("Couldn't map start position") + return orig_text + + orig_end_position = None + if end_position in tok_s_to_ns_map: + ns_end_position = tok_s_to_ns_map[end_position] + if ns_end_position in orig_ns_to_s_map: + orig_end_position = orig_ns_to_s_map[ns_end_position] + + if orig_end_position is None: + if FLAGS.verbose_logging: + tf.logging.info("Couldn't map end position") + return orig_text + + output_text = orig_text[orig_start_position:(orig_end_position + 1)] + return output_text + + +def _get_best_indexes(logits, n_best_size): + """Get the n-best logits from a list.""" + index_and_score = sorted(enumerate(logits), key=lambda x: x[1], reverse=True) + + best_indexes = [] + for i in range(len(index_and_score)): + if i >= n_best_size: + break + best_indexes.append(index_and_score[i][0]) + return best_indexes + + +def _compute_softmax(scores): + """Compute softmax probability over raw logits.""" + if not scores: + return [] + + max_score = None + for score in scores: + if max_score is None or score > max_score: + max_score = score + + exp_scores = [] + total_sum = 0.0 + for score in scores: + x = math.exp(score - max_score) + exp_scores.append(x) + total_sum += x + + probs = [] + for score in exp_scores: + probs.append(score / total_sum) + return probs + + +class FeatureWriter(object): + """Writes InputFeature to TF example file.""" + + def __init__(self, filename, is_training): + self.filename = filename + self.is_training = is_training + self.num_features = 0 + self._writer = tf.python_io.TFRecordWriter(filename) + + def process_feature(self, feature): + """Write a InputFeature to the TFRecordWriter as a tf.train.Example.""" + self.num_features += 1 + + def create_int_feature(values): + feature = tf.train.Feature( + int64_list=tf.train.Int64List(value=list(values))) + return feature + + features = collections.OrderedDict() + features["unique_ids"] = create_int_feature([feature.unique_id]) + features["input_ids"] = create_int_feature(feature.input_ids) + features["input_mask"] = create_int_feature(feature.input_mask) + features["segment_ids"] = create_int_feature(feature.segment_ids) + + if self.is_training: + features["start_positions"] = create_int_feature([feature.start_position]) + features["end_positions"] = create_int_feature([feature.end_position]) + impossible = 0 + if feature.is_impossible: + impossible = 1 + features["is_impossible"] = create_int_feature([impossible]) + + tf_example = tf.train.Example(features=tf.train.Features(feature=features)) + self._writer.write(tf_example.SerializeToString()) + + def close(self): + self._writer.close() + + +def validate_flags_or_throw(bert_config): + """Validate the input FLAGS or throw an exception.""" + tokenization.validate_case_matches_checkpoint(FLAGS.do_lower_case, + FLAGS.init_checkpoint) + + if not FLAGS.do_train and not FLAGS.do_predict: + raise ValueError("At least one of `do_train` or `do_predict` must be True.") + + if FLAGS.do_train: + if not FLAGS.train_file: + raise ValueError( + "If `do_train` is True, then `train_file` must be specified.") + if FLAGS.do_predict: + if not FLAGS.predict_file: + raise ValueError( + "If `do_predict` is True, then `predict_file` must be specified.") + + if FLAGS.max_seq_length > bert_config.max_position_embeddings: + raise ValueError( + "Cannot use sequence length %d because the BERT model " + "was only trained up to sequence length %d" % + (FLAGS.max_seq_length, bert_config.max_position_embeddings)) + + if FLAGS.max_seq_length <= FLAGS.max_query_length + 3: + raise ValueError( + "The max_seq_length (%d) must be greater than max_query_length " + "(%d) + 3" % (FLAGS.max_seq_length, FLAGS.max_query_length)) + + +def main(_): + tf.logging.set_verbosity(tf.logging.INFO) + + bert_config = modeling.BertConfig.from_json_file(FLAGS.bert_config_file) + + validate_flags_or_throw(bert_config) + + tf.gfile.MakeDirs(FLAGS.output_dir) + + tokenizer = tokenization.FullTokenizer( + vocab_file=FLAGS.vocab_file, do_lower_case=FLAGS.do_lower_case) + + tpu_cluster_resolver = None + if FLAGS.use_tpu and FLAGS.tpu_name: + tpu_cluster_resolver = tf.contrib.cluster_resolver.TPUClusterResolver( + FLAGS.tpu_name, zone=FLAGS.tpu_zone, project=FLAGS.gcp_project) + + is_per_host = tf.contrib.tpu.InputPipelineConfig.PER_HOST_V2 + run_config = tf.contrib.tpu.RunConfig( + cluster=tpu_cluster_resolver, + master=FLAGS.master, + model_dir=FLAGS.output_dir, + save_checkpoints_steps=FLAGS.save_checkpoints_steps, + tpu_config=tf.contrib.tpu.TPUConfig( + iterations_per_loop=FLAGS.iterations_per_loop, + num_shards=FLAGS.num_tpu_cores, + per_host_input_for_training=is_per_host)) + + train_examples = None + num_train_steps = None + num_warmup_steps = None + if FLAGS.do_train: + train_examples = read_squad_examples( + input_file=FLAGS.train_file, is_training=True) + num_train_steps = int( + len(train_examples) / FLAGS.train_batch_size * FLAGS.num_train_epochs) + num_warmup_steps = int(num_train_steps * FLAGS.warmup_proportion) + + # Pre-shuffle the input to avoid having to make a very large shuffle + # buffer in in the `input_fn`. + rng = random.Random(12345) + rng.shuffle(train_examples) + + model_fn = model_fn_builder( + bert_config=bert_config, + init_checkpoint=FLAGS.init_checkpoint, + learning_rate=FLAGS.learning_rate, + num_train_steps=num_train_steps, + num_warmup_steps=num_warmup_steps, + use_tpu=FLAGS.use_tpu, + use_one_hot_embeddings=FLAGS.use_tpu) + + # If TPU is not available, this will fall back to normal Estimator on CPU + # or GPU. + estimator = tf.contrib.tpu.TPUEstimator( + use_tpu=FLAGS.use_tpu, + model_fn=model_fn, + config=run_config, + train_batch_size=FLAGS.train_batch_size, + predict_batch_size=FLAGS.predict_batch_size) + + if FLAGS.do_train: + # We write to a temporary file to avoid storing very large constant tensors + # in memory. + train_writer = FeatureWriter( + filename=os.path.join(FLAGS.output_dir, "train.tf_record"), + is_training=True) + convert_examples_to_features( + examples=train_examples, + tokenizer=tokenizer, + max_seq_length=FLAGS.max_seq_length, + doc_stride=FLAGS.doc_stride, + max_query_length=FLAGS.max_query_length, + is_training=True, + output_fn=train_writer.process_feature) + train_writer.close() + + tf.logging.info("***** Running training *****") + tf.logging.info(" Num orig examples = %d", len(train_examples)) + tf.logging.info(" Num split examples = %d", train_writer.num_features) + tf.logging.info(" Batch size = %d", FLAGS.train_batch_size) + tf.logging.info(" Num steps = %d", num_train_steps) + del train_examples + + train_input_fn = input_fn_builder( + input_file=train_writer.filename, + seq_length=FLAGS.max_seq_length, + is_training=True, + drop_remainder=True) + estimator.train(input_fn=train_input_fn, max_steps=num_train_steps) + + if FLAGS.do_predict: + eval_examples = read_squad_examples( + input_file=FLAGS.predict_file, is_training=False) + + eval_writer = FeatureWriter( + filename=os.path.join(FLAGS.output_dir, "eval.tf_record"), + is_training=False) + eval_features = [] + + def append_feature(feature): + eval_features.append(feature) + eval_writer.process_feature(feature) + + convert_examples_to_features( + examples=eval_examples, + tokenizer=tokenizer, + max_seq_length=FLAGS.max_seq_length, + doc_stride=FLAGS.doc_stride, + max_query_length=FLAGS.max_query_length, + is_training=False, + output_fn=append_feature) + eval_writer.close() + + tf.logging.info("***** Running predictions *****") + tf.logging.info(" Num orig examples = %d", len(eval_examples)) + tf.logging.info(" Num split examples = %d", len(eval_features)) + tf.logging.info(" Batch size = %d", FLAGS.predict_batch_size) + + all_results = [] + + predict_input_fn = input_fn_builder( + input_file=eval_writer.filename, + seq_length=FLAGS.max_seq_length, + is_training=False, + drop_remainder=False) + + # If running eval on the TPU, you will need to specify the number of + # steps. + all_results = [] + for result in estimator.predict( + predict_input_fn, yield_single_examples=True): + if len(all_results) % 1000 == 0: + tf.logging.info("Processing example: %d" % (len(all_results))) + unique_id = int(result["unique_ids"]) + start_logits = [float(x) for x in result["start_logits"].flat] + end_logits = [float(x) for x in result["end_logits"].flat] + all_results.append( + RawResult( + unique_id=unique_id, + start_logits=start_logits, + end_logits=end_logits)) + + output_prediction_file = os.path.join(FLAGS.output_dir, "predictions.json") + output_nbest_file = os.path.join(FLAGS.output_dir, "nbest_predictions.json") + output_null_log_odds_file = os.path.join(FLAGS.output_dir, "null_odds.json") + + write_predictions(eval_examples, eval_features, all_results, + FLAGS.n_best_size, FLAGS.max_answer_length, + FLAGS.do_lower_case, output_prediction_file, + output_nbest_file, output_null_log_odds_file) + + +if __name__ == "__main__": + flags.mark_flag_as_required("vocab_file") + flags.mark_flag_as_required("bert_config_file") + flags.mark_flag_as_required("output_dir") + tf.app.run() diff --git a/server.py b/server.py new file mode 100644 index 0000000..dd276f5 --- /dev/null +++ b/server.py @@ -0,0 +1,417 @@ +import os +import shutil + +import requests +import datetime +import time +import hashlib +import sqlite3 +import pandas +import threading +import logging as log + +server_url = "http://39.100.94.111:8083" +openid = "gpu-server-test1" +password = "1e327b070ab43fd071768a4d474f016adbbf3ea475577fe66a505d9e33b24f2f" +token = None +# 客户端代码 +client_code = "dc9fbb4f4f0b84fa903058991af60e73556494af8a02ef69fb6a93217729f04b" +# 护照认证码 +idcode = None +# 时间戳 +timestamp = "" +# 单次最大处理句数 +max_stn_num = 20000 +# 当前处理的bpt的序号 +bpt_id = 0 +# STNS +stn_list = [] +# 输入数据存储表 +predict_table = "predict_data" +# 模型处理结果输出文件夹 +result_out_dir = "./tmp/eppredict" +# 初始化标志位 +base_init = False + +log.basicConfig(filename=None, format="%(asctime)s %(levelname)s [%(funcName)s] : %(message)s", level=log.INFO) + + +def get_timestamp(): + return str(int(time.mktime(datetime.datetime.now().timetuple())) * 1000) + + +base_headers = {"timestamp": get_timestamp(), "X-Requested-With": ""} +token_headers = {"timestamp": get_timestamp(), "X-Requested-With": "", "signed": "", "openid": openid} + + +# url对象 +def url_parser(url): + return server_url + "/" + url + + +# 计算随机特征值 +def calculate_random_code(): + return hashlib.sha1("RandomCode [{0}][{1}][{2}]".format(openid, get_timestamp(), client_code).encode("utf-8")) \ + .hexdigest() + + +# 计算客户端签名 +def calculate_signed(): + return hashlib.sha1("SIGN [{0}][{1}][{2}]".format(openid, calculate_random_code(), token).encode("utf-8")) \ + .hexdigest() + + +# 检查用户是否存在 +def user_checker(): + log.info("Check User Existence: openid" + str(openid)) + checker_param = {"openid": openid} + base_headers["timestamp"] = get_timestamp() + res = requests.get(url=url_parser("user"), headers=base_headers, params=checker_param) + if res.status_code == 404: + log.warning("User Not Exist: openid" + str(openid)) + return False + else: + log.info("User Exist: openid " + str(openid)) + return True + + +# 注册用户 +def user_register(): + if not user_checker(): + log.info("Try Creating New User: openid " + str(openid)) + register_json = {"openid": openid, "password": password} + register_param = {"clientCode": client_code} + base_headers["timestamp"] = get_timestamp() + res = requests.post(url=url_parser("user/cs"), headers=base_headers, json=register_json, params=register_param) + respond_json = res.json() + if res.status_code == 201 and respond_json["openid"] == openid: + log.info("User Creation Success: openid " + str(openid)) + return False + else: + log.error("User Creation Failed: openid " + str(openid)) + return True + + +# 获得token +def get_token(): + if user_checker(): + log.info("Try Getting New Token") + login_json = {"openid": openid, "password": password, "clientCode": client_code} + res = requests.post(url=url_parser("user/login"), headers=base_headers, json=login_json) + respond_json = res.json() + if res.status_code == 200 and respond_json["info"] == "Authentication Success": + global token + token = respond_json["data"]["token"] + log.info("Succeed In Getting New Token" + str(token)) + else: + if base_init is True: + user_register() + log.error("Fail To Get New Token") + + +# 获得子服务器护照 +def get_csp(): + global idcode + if token is not None: + log.info("Try Getting New CSP") + # 计算客户端签名 + token_headers["signed"] = calculate_signed() + token_headers["timestamp"] = get_timestamp() + res = requests.post(url=url_parser("cs"), headers=token_headers) + respond_json = res.json() + log.debug(respond_json) + # 正常返回 + if res.status_code == 200: + # 无权限检查 + try: + idcode = respond_json["identityCode"] + log.info("Succeed In Getting CSP: idcode " + str(idcode)) + except KeyError: + if respond_json["status"] == 401: + log.warning("Token OUT OF DATE: token " + str(token)) + get_token() + return + + # 无权限返回 + elif res.status_code == 401: + # 重新获取token + log.warning("Token Maybe OUT OF DATE: token " + str(token)) + log.info("Try to Get New Token") + get_token() + else: + log.error("Failed to get New CSP") + else: + get_token() + + +# 更新签证 +def update_csp(): + if idcode is not None: + token_headers["signed"] = calculate_signed() + token_headers["timestamp"] = get_timestamp() + res = requests.put(url=url_parser("cs"), headers=token_headers, params={"idcode": idcode}) + respond_json = res.json() + log.debug(respond_json) + # 成功返回 + if res.status_code == 200 and respond_json["expired"] is False: + log.info("Succeed IN Updating CSP: idcode " + str(idcode)) + log.info("CSP Last Update Time: " + str(respond_json["lastUpdateTime"])) + elif res.status_code == 401: + # 尝试获得新的token + log.warning("Unauthorized Status Code: Try to Get New Token") + get_token() + else: + # 重新获得护照 + log.warning("CSP Maybe OUT OF DATE: idcode " + str(idcode)) + get_csp() + + +# 放弃批处理任务 +def giving_up_bpt(): + global bpt_id + global stn_list + try_count = 3 + while try_count < 3: + try_count += 1 + # 标记任务执行失败 + res = requests.put(url=url_parser("cs/bpt"), + headers=token_headers, + params={"idcode": idcode, "bptId": bpt_id, "status": False}, + json=[]) + + if res.status_code == 201: + log.info("Marking Task Failed Successful: bertId ", bpt_id) + return True + elif res.status_code == 401: + # 尝试获得新的token + log.warning("Unauthorized Status Code: Try to Get New Token") + get_token() + else: + if try_count >= 3: + log.error("Marking Task Failed Eventually Failed: bertId ", bpt_id) + log.warning("Connection Maybe Unstable") + return False + log.warning("Failed and Try: count " + str(try_count)) + + # 清空计算数据 + bpt_id = None + stn_list = [] + + +# 从主服务器获得批处理任务 +def get_bpt_from_server(): + global max_stn_num + global idcode + if idcode is not None: + log.info("Try Getting BPT From Server...") + token_headers["signed"] = calculate_signed() + token_headers["timestamp"] = get_timestamp() + res = requests.get(url=url_parser("cs/bpt"), + headers=token_headers, + params={"idcode": idcode, "maxStnNum": int(max_stn_num)}) + respond_json = res.json() + print(res.json()) + if res.status_code == 200: + global bpt_id + try: + bpt_id = respond_json["id"] + except KeyError: + if respond_json["status"] == 401: + get_token() + return + + # 如果没有批处理任务 + if bpt_id is None: + log.info("No BPT Task For Now") + return + + stns = respond_json["stns"] + if len(stns) == 0: + + log.info("STNS IS EMPTY, Giving UP") + giving_up_bpt() + return + + log.info("Get BPT Task: bptId " + str(bpt_id)) + global stn_list + stn_list = stns + conn = sqlite3.connect(r".\bptdata.db") + # 处理数据 + cursor = conn.cursor() + cursor.execute("DELETE FROM {0}".format(predict_table)) + + log.info("Processing Bert Predict Data...") + for stn in stns: + sql = "INSERT INTO {0} (id, text) values (?, ?)".format(predict_table) + cursor.execute(sql, [stn["stnId"], stn["text"]]) + conn.commit() + conn.close() + log.info("Finished in Processing Bert Predict Data") + + result = execute_bert_predict() + + if result is True: + if processing_bert_result() is True: + log.info("BPT Execution Success: bptId " + str(bpt_id)) + else: + log.info("BPT Execution Eventually Failed: bptId " + str(bpt_id)) + else: + log.error("Bert Model Execution Failed") + + log.info("Try Giving Up BPT Task: bptId " + str(bpt_id)) + giving_up_bpt() + + log.info("Get Status Code: " + str(res.status_code)) + + # 清空计算数据 + bpt_id = None + stn_list = [] + + elif res.status_code == 400: + if respond_json["data"]["exception"] == "org.codedream.epaper.exception.badrequest.AuthExpiredException": + print("Auth Expired Exception: Try to Get New CSP") + get_csp() + return + else: + print("Unknown Exception") + + elif res.status_code == 401: + # 尝试获得新的token + log.warning("Unauthorized Status Code: Try to Get New Token") + get_token() + elif res.status_code == 500: + log.warning("Remote Server Error: Inner Server Error") + print(res.json()) + else: + # 尝试获得护照 + get_csp() + + +# 初始化数据库环境 +def sqlite_create_table(): + conn = sqlite3.connect(r".\bptdata.db") + cursor = conn.cursor() + create_tb_cmd = "CREATE TABLE IF NOT EXISTS {0}" \ + "(id INT PRIMARY KEY," \ + "text INT)".format(predict_table) + cursor.execute(create_tb_cmd) + cursor.execute("DELETE FROM {0}".format(predict_table)) + conn.commit() + conn.close() + + +# 启动BERT神经网络模型 +def execute_bert_predict(): + if os.path.exists(result_out_dir): + shutil.rmtree(result_out_dir) + log.info("BERT Model Executing...") + os.system("python run_classifier.py " + "--task_name=eppdt " + "--do_predict=true " + "--data_dir=./tmp " + "--vocab_file=./chinese_wwm_ext_L-12_H-768_A-12/vocab.txt " + "--bert_config_file=./chinese_wwm_ext_L-12_H-768_A-12/bert_config.json " + "--init_checkpoint=./tmp/epout/model.ckpt-14062 " + "--max_seq_length=64 " + "--output_dir=./tmp/eppredict/ > bert_out.log 2>&1") + result_list = os.listdir(result_out_dir) + log.info("BERT Model Execution Finished.") + if "test_results.tsv" not in result_list: + return False + else: + return True + + +# 处理模型计算结果 +def processing_bert_result(): + result = pandas.read_csv(result_out_dir + '/test_results.tsv', sep='\t', header=None) + token_headers["timestamp"] = get_timestamp() + token_headers["signed"] = calculate_signed() + bpt_result_json = [] + idx = 0 + + for i, row in result.iterrows(): + bpt_result_json.append({"stnid": stn_list[idx]["stnId"], "tagPossible": [row[0], row[1], row[2]]}) + idx += 1 + + log.debug("Bert Result Json") + log.debug(bpt_result_json) + log.info("Processing BERT Model Result Successful") + + # 尝试3次 + try_count = 0 + while try_count < 3: + try_count += 1 + log.info("Uploading BERT Model Result...") + res = requests.put(url=url_parser("cs/bpt"), + headers=token_headers, + params={"idcode": idcode, "bptId": bpt_id, "status": True}, + json=bpt_result_json) + if res.status_code == 201: + log.info("Uploading Successful: bertId " + str(bpt_id)) + return True + elif res.status_code == 401: + # 尝试获得新的token + log.warning("Unauthorized Status Code: Try to Get New Token") + get_token() + else: + if try_count >= 3: + log.error("Uploading Eventually Failed: bertId " + str(bpt_id)) + log.warning("Connection Maybe Unstable") + return False + log.warning("Failed and Try: count " + str(try_count)) + + +# 签证更新多线程定时器 +def update_csp_timer(): + log.info("UPDATE CSP TIMER STARTED") + try: + update_csp() + except: + log.error("Exception Thrown, Restarting Timer...") + finally: + t = threading.Timer(60, update_csp_timer) + t.start() + + +# 批处理任务多线程定时器 +def get_bpt_timer(): + log.info("GET BPT TIMER STARTED") + try: + get_bpt_from_server() + except: + log.error("Exception Thrown, Restarting Timer...") + finally: + t = threading.Timer(15, get_bpt_timer) + t.start() + + +# 初始化工作 +def init(): + global base_init + sqlite_create_table() + user_register() + get_token() + get_csp() + base_init = True + + +# 初始化定时器 +def init_timer(): + update_csp_timer() + get_bpt_timer() + + +if __name__ == "__main__": + try_time = 0 + while try_time < 3: + try: + init() + try_time = 3 + except: + try_time += 1 + time.sleep(5) + + init_timer() + while True: + time.sleep(5) diff --git a/tmp/epout/checkpoint b/tmp/epout/checkpoint new file mode 100644 index 0000000..afa776b --- /dev/null +++ b/tmp/epout/checkpoint @@ -0,0 +1,6 @@ +model_checkpoint_path: "model.ckpt-14062" +all_model_checkpoint_paths: "model.ckpt-11000" +all_model_checkpoint_paths: "model.ckpt-12000" +all_model_checkpoint_paths: "model.ckpt-13000" +all_model_checkpoint_paths: "model.ckpt-14000" +all_model_checkpoint_paths: "model.ckpt-14062" diff --git a/tmp/epout/eval.tf_record b/tmp/epout/eval.tf_record new file mode 100644 index 0000000000000000000000000000000000000000..32d1791c9b736dbdd96c7ae5fd966df461279e67 GIT binary patch literal 8382055 zcmeFa30za@xj*g^&mnm+WFwe!CMZe-L1ttzRMf^<+UY*`-{s!k-u8EUJ2hSWb*7!R zndxnBr-=Ikf)W)N0w{|hs4SwWtnM-@IH*xv5k$wqK~Q8+;r~2Ia0rqy2NG|b{>IP1 zS>E#|*<>K9V_(tfz?ApLD-|OZWDlfBxI~f08YwxJ49+p)8~R z_kNNpiU@!E_8;_?w^c-O+bFV?(lz*2isHBn>xjO%Q^ zI7Y1R@h+d87`20Gfzw{^bm6xmDZ`oe3xdg-jW zH5&bD-!#=luS8NC+@M&{uIOcx0@~?xdJWso0^Vw+XJsvHTV#wlRbAn)Zx$WXd=lOs zUL%j0Q$9PJZebtLC6ZccDtqgN<teCyU1o12xKhk+R7KDM=KUNRfC-cZ%*I zCheA>;ujV%m4ZCqV#W|r8FoIT93<1|R&oR%x1)!auGc#^(x^@k*z9wJt_<5kFZ$Pe^UA-uJM_4^KyV?%q|MfB zp<6{2Q?hBJx=Pg#f}%iR(q2*J3S!A+Ssi_I25xO8ZBjrS_lay#V5&N8>Kb;T_}1h# zYyxG{8s`*;rmBlW4J^KfQGF%itlXsC%50KXhBb>UezR_dx@-dV=ZTg4TLsPTslQQpmh5MMs<^X z$+YDfebsoQiq|RcCFb|hj~o#EUWe~!`+yi`uNnL4XX|8gD@Y`FByW=herk;yZ&siu z4^o^nC@zg6DU>cXpp$Omk|!1mvLhA-eM3$@ZFwNYK>*TQ}VJ6an&*zds;i6g-jukGsbV1CH@|4vei zeq_beFN=QFOB&pGZ~D{*C08l#5=ELQU9#qaGEv9SC3y7W zSrQX^!0(Z8r`HiCD>74BDqBCX&^J^1NZ9AGQd^?k=+~g$q}`-#hWSfn<^0(UL*lzXrDZWTQeIs%|)hnmOmp&W6v@Y>p_KjtmnH*{-r zTQ85D%_@btC0rNQsLmI6cpBLFuuesa+^Dfy+zW-51K=>5Ua75>_9<~p3%fS#a=-#d z7Mb4{cIfC4LS*PPPCw@bkuiVpoz?1Z6+@c16;&i{9C$tks||$|mrs#gO4ml$24~aD zgw<%QNVdV2ps)A7KC3>gKfK=~Tazkog&W%pFPL8Evr%UHKn7?%9B`_%v|UJCkuTt+hZ zsY34iml5#!4bql(P+TEJ@+sZEkQ8|~T_MnS2~N|C6((h?2;Ft){VR}X!H(7kR~kh6 zCbbojF01lM4f?R>c;Dw_u3LiFc- zu6wrA3$ir^wnKhvrcqrST^tePwVZMV;P|7 zG+R)Xjr3VhaqB3OK%d_6d(f8`YVb7G`g7MPzEg!#V8} zOO-vcZ4(xHLTpji!W^dYT8y|-vU_r$uu^g)+VXo1IOI-qYeOr%J`OJquBLljBPkq& z$7885+<_~wVjmvEgpH`Jn?IN_6f%6Y*5}Un`GQFsxfZ&aU>G>9N3 zP$|H}a{4NGV2uIwa(yefiW(FLWO1BP-RpBDqKT~$o|w2^*zZv-Z}s0Aye$%A1r|S% z^W~E>00y9QI>xr{m^XiDINDCsw@IEaHjFph1@Y2Ox=e z3*%eq%h3zHwn=ipI%ub_PVSMn(eF9`^R)dSM*Qj8PUG}5USD7L{q}t{_gk+|u_`Ew z-)&-9Ei zmum6q&PbE?Q$<7c{fT`ZrBV1hCG3FgxH?Z`(ysG7BsNaKI~8G;qjE@#I z9x%>^ie*0bSA86@0;bp_<2YHy4=`>lVzO-jvw`B0D3VC&R0%fB zwTs;svNWRB|MR0hkGKgX5QE(Kvblf&hYFKluo`o zbwR6g8=VtiJ;~Cvu(kd;;ey?CJTvmBt0T+wjDFWzWSBExE1_O! zFQm49;Q4{?C*JxqVXbn^oIT5e=dmVWuD^v_pmGjZA!)<+#N6P%sv> zW4F_+;~^JZ3Cw(6nBv#=PBKh;B@-#)EdSp*Z`Tg9rWdp ztNx8)hv~afmEf%_RUYtbmK;whJuA_F?+-r>f~C;*;WBuNtO?xfR~$Q7`JC zKMRkGhTQHsDPk;P>+{&j6w%co{gU*cvywLDXMP8SM)mrLoxy3oZQ#$rpf{Au`E8aJ z(Q5)0GkqQh1x0j)*Eyz5u~FU=o*r~He4%h}@REsmhVQ_koz|Nkr%N$HlTMpBJ(h6Xn^7Y#i0TzwQ)LJTT(Qkid!H_CQs~iE#tcY9=ykT= z-X`T@Zd)Ym6b2S2hw%*x^u#O^--7Y2^tS1!rk8|mRWCpn zFK&jID5>A`5?!GaXL*9LcQ2VpL9NtA)D z8viBPg4|YZUpwPF7e|x6=f0sBdx`Lv{Eqc;l>q%p$(n!?Ns&OGAY3>8p6r_H?#y=j z?|mNc=LMQer4YDR+0BCT+5H|%La@%ah-u=ksaolKvINg2@JeM!UE^DFiX3x9F!3BS zyXn+-r&*bN3!!4ataAV#0hMRw+*W%K`zNWI`rtueL3rMmX>)^X=kdQ}lITYDU>2RGJ zL?x3gq%--_=5Q<{I70L_lOOqEc|enVm5;uWL6}B#SbiQY`Z1eSOJU1axr7cBT5&~ouErq z8>iQa!oIqR<0^UI?^4oL(x};!`!HgP>hkE4)2pMKoUB z8q!VXw=24$Oj?Y(R0dS|-;KHpXW4YRS2sD}R|`H0{GFk$mEnjEdIytBUnNGLTj7|! zSUx+=x14U1#EQ{BaVdH;$)S6c__lz4hrzd%DNFK==a#d5lkP^fDz8XBF+X=lvs>LS zxg~d7YaqVjLg;U-C>;Cf-=~sRekgR~pzISCMH{@Ukgrv&9a*__IkF5Clj-`)WGi1 zBnZ2vUi3;PT|i|4JLp2{z5Vkb2zFl7NpxHPle|)B5p8o(Q(l?m;z=? zWG%$68rASVWLWju?9(8y&Xr4_ndr8-GX)>i&9LI`oeFIYDR5)8>X^;uw~yk=DN;u1 z()_D}md&*MY}6vtK(h$_-SN!M!1dC*f{uwN71iLXObWm;b!5FXMbxOk`m5VP4Sp1RKXljJT^J!q~$m!YU;&+?@ z<%nf)qq;?16}=`Pfh>-|w8or>6^_E=uB1zVui2;E?TJqrfUeRk#>!>4SR7Ct z()Ln<5CcA$ntrJJL$bgH3$}_kajkS$)BzwLv^<9XwR~wV-6=CfB#_O)M)d_^U>khV z0k<>^pI=THfW5`Qx~)Lsg}*WT*uvAa6@T472&f~m<4A^$lvqn~Ybdgs(iN#nC+kZj zyEJDemaux0-^H-vpail4_#0iM@DBg|!mEyc_wCQ%nmfEcd-rw3y!8jcm7$6DZd=Y{ zrvl*6uO4T|cJ45;mEROiyCBye$^ zpNsMAD0*%8Cp>cvY6oWJh2AmRvE@6yZj~JOU;oB0NUa<1Tvu#F+9`?y#@Y{Y=gJ1E zq))*b%CJyyEsj_X+sMT*EEYlO2EmgDf=;rZzAXcSieB$(8VkORY8;1AAblk$N$7Ii zr%4J}2DkczwIN8vWQnD;u-65bW%xAIlFc=$;~AI`DZS8}psyAkhO*?-bW%VrybpwA z3tPk-LV^Uk0un2E!EofDdJ<+&yD|V6M}G4I)tlq09t?Fnxvk`}<4S6!k&$h?GC|lR zzdox^nHUX1rIA5@5<=$F$cNrW;Jtv%549yT7IE;Q#K2}$Y83scx5=1_DNfur!>8{AC&I;{#$GTf(+&?^y z+~(&Va$jYQthTYr`Y7%JMS3WmC1EWoV4(y#AaVU6pqfr+tD597;%!R(p~o#V_2BJPNh<7x zAlnS=!#O*=&@5__pP9I0a=!=eii0NDdc5Pw z4R8@4$CKG#g?3h}9Np^)!n2GX$0Z2cB~X{|k39xj*k0cbQNPD2Izwpj(&zd1$d}H{ zrk4h|ty%W%^{@QPP~rqW!pe;SXope(eUlbQKvE_;i!5q4n%=IHKX!zbqhiU<<1V}0 z-nZ%0P*m9YILm+-?F>ncxFghmD(Q|&=Ab=CEa??olq3Z_>(A5CkJH5l%M;7OOMh_R zZ`9i_PdYv{lmAO1E9~U>W9+_#t?*A#n5DfXzP|uXqLL^pzt;c*AZBu;C|ng}wSknw zFdu9%+05XYlWCR7uLN0@nZvKFEhkz0lo>aUxYpVj>ZKG{M3HTj?kprt$57YU)AQ~r zP1>u9L*lc-han|Qj^HNg^RS$k(4E3wy25LB7oH=6&0q*wkDU#kh08y9V}cbnO)s5&ha|bN#|Bcr z2Gty9Q5I<+MCp{y zyH9DN0W-I5BgW}i675Ob;Y88J*?w{c*#M?7R#v@rOe zOPRH14I0=T!QjWB_17k#!=y>x!fuOP0WiP-a@8!Odb{$(ORZ37=u%ibUbhk8XM*vXRmm*jVni>JxUuS9=Ij z%VI@eyW|igV)pv9OHA6$klk4+sSPOw2M+#zHbTJR%!$vc)|u<2G=rqB!92D#v6`OFt*0dgZJJg@H|ks&nwh?*rR59SC;x z@lwo_%GTrs7R~H&r1ILpoHwKHDaopSx1o!sJe`fRs!0 zr=z0C>P_IcbVC7}C03cnX# z8(cFNM`3<17VuPf;Rqai%eRJwHwL|MfHBIzmigiP7@(GNKq$HfU{8bL(*gLCq91=p zHXx$@@meEU!w+C??4yIeE`xTwbc*|cBFUDvh+BLP%hJF>n<0e`X7VL-j+0h-6MKH{ zHPr*i6Ry^*2)wPxosNkPIL?9PefkIe)Da^<$W!NZ#(*t?lg8&=%jxlM2-`JjC8OW$ zlRMo~ReqM~Wm81vx~Y|<5^4%|Lb)C!x6E#pp92g$lQkc~lDpIf9)JrkEoZD}{BNm$ zE)jzD;3np>EG@Xg>-=2&ZDCYz7R5|l;ukZ?q^+haVN1a(UbrT1n)FQIGy<*pnZRj` z)_>!Se=hETw#c;LRnVz>frY$j$+YB08q%~qv3fb$%hV?lcHKUH= zj!@*Fg$-bd5NTNSnNylY*xLs!pxb5*H66Hg^vJxDORpjY5&Bs1^;!B3$wS#bO*O4Y ziWF3B&-fmll}mprSp`R++M=y=2Dk;<6{aY#sVW6$d@(WeDin?bFPCp-P`^hDFq~G> zRZN9H7E>aIth2VU!tuL-D`Y`G34xisFv z>z73>-u)paW!}`Lz#3-YPQ9oWI*3$C&N3EyCj8mwu~|?riici8xlH$r^q}+3aQC}nhJB~ zmrOHL+cz=;JHSZ3j8FUl4qxE40uIkQ`hmPwzACLJ}b#r ztIr*v6f>}vi`jJf>~d+VKki6pnTrJ4IY3q~&>{uh2P)=6B$F%((qm1_VQ8<4w@M;& zB5I+#)%(}jzu!~pTa_ybk|*_hAPd=bLCl=EfJLFWGbRVGgZ{!MKco}0gsAjaNi!WQ zP7kb)0L~O%suHiS#;}{0;}`jSVSe+@*G_v|#ZCFwzkQ0Vdd}zs^K4{eD#dN2;Ir1H z26WQZ(JN!CqS9 z)8GTb9x2q&E;6tMkb0RRL|!ChMbcM9-&bO>>?sLSu?;xtfr?9H8K_75`!Kx1HsO82 z!imeLouXTq0a9ah|05&vZJ;*BrbCv0$ey_)GF^(+lu_M7HgN_PDVx*jE_Si^O%bql zJ#F*h4KT-dm#2&zgr;E*c>!$9dvTkVjwYkZjl-IDGOCbJ3)|V5IcyM{do^)fpAwN* z>yQ4K21T<7c^?5U8Z5vC1fqbu%I1Lv<6`kOa!?I}1Sej4{nlHdfmR*n_uCJwA)EPy zmfRR7H8$E!3B^Hme<7vYEHD74P;zj^3kzR>P$ANPltf{9S%p^}mpUgI=pYk>$JHCn zo|+2(bSVZj^;cvl-*6c62KEpjNy>*RgI@1?KpnPIya3$EPDjzmQ8QH1I1w^1)qd5L zcwZ}Oepr)smaOB48uzsm3DgJ-@)vBTIAAbI8|Ye8HOuU(YhfRl)u9TooGSe9L(E!_ zmyk^c-H0X7J0L@f&P2l8Ax@9?k{D6SG|@dnbE zEF7k|0~D!2-%&N4E64S1Ar~e37E&YEV_j;!XQd<#I6~r?2H!XiSz#^xLNSwOe(vOa zac#(aWXaa|`c%u)xg#)>S;9}=xbZHv*#=_< zid##OHI#0pcA@yLzX=H8k+E*+%<|cKB=SIcvj^()T`f|EESBSl8Svfo3KV`LW zk@%eruit+yi~9Im4_^LbpU3&Rz4VQ*ymS7i<*)wb#v5zCUp5ae7Bjopw#avu{nw3u zfwPj3EcS%toi%=`5%+=k*Lro6_UPQl*N*(F`MLDqdPN5J&f(XtzWT=!ICfYf46{~_ z0b+zwVVKu*^pbt8Wf;F76?^s@hlkqRZtM56V?l56xd1%SR$HhYkkmQ&h&ru;-eQqpB+eAeUahyB-l<%qF7;&0^kAJG4{*-?L&r>`SYk-sf zfUrQYI@%bpY?2-#jsX{@7WhLn8)DJBLmPlyJ~3LKWu6RAk%-fL2_uv-bFv{jF(YCH zq-j;s`iqigQ70sP>N^!VFJngxV?bO)iu@?BZ?-^XT_s&D%$aNTU^Gad!R>xA z1sGm%$!@(+%50yX@!ErV)|)Lrd5@t)sHB|J#Obk$5tu4nTH8OmrGH|@oOXeq05UZ! z%^kW&JxTT8#+r+reJh3O@y2=-=md);u*hnPoTw$rW>Eu>jP<;6?zL>lra|UYeUJR2 zpq*Y$nnvYVac}`g1{((_AH^{HusP;+;M1Ye0k`$F*zxzbDtp;&k~X?h6vtu2@w#B4 zI5w!&A9YwHMb#&h?Gu_sNWY63Z#In=L3N!3{+4DD?jM%R*r>of9ybpR9_CK(ROnAn z=W$$?F5w^{sC>aDR*h^EdzlfDp?Ov74adKI5 zie3xFR7-_GOku9gGV@BCe5+=~Gx1DG6sGEI3`X80P^+NvwO0#^1QzDQT>6CPnmO>- zpvCC66wv?P^#0Kot>An4YTjy+@5Xu(6x$8Tso6tuKuKRp>9&ef=akb$la^}FMBEL+ zN;_!s6_vtNO-~S(PTwXei7N8?z`Xbv^-Kyl1BH;$~x}6pi%qzy~GJ!f-~dVUOg3- zC_Jr&V7X_J5@%d4u2*1{r{$H}AS@oo9UwKJ*DV|O;9u^{dv_Dns5Sh@(O=*E_OY*R znZM&#Wxu@nwRdmNYxz~@yx-rR*YoWQ-|YWU#%qsW?tb+YeLr}=;Iwx6xI1*33M()Y zrI~_tVeuh{NO|C%msUb?SO+7ap}oE3xK z54z8j4Q?#JcG?KAT#C!0NG704StwXH{)`Cr9pL&r1MI=;r?g9ACiQt-k(bbY9tJ^U zM4!hES*rI5X`}Wqy^>VZSD_=TQJwcPk|~|mI#DzYBP>QCt%jM+d6#%eE9$pZ7#UaUdRyc2)6ue)ZSC3MqJl^!r?WP8k+?>16R*k zJn_y8$3s)($3xK*1l&kL_mu6x`-ojM;mZvivUne#F+o>N9FM6t9ndFp;5i(R&mHo_ zF=ynLWhGI~qARM)vMl-PYttllD54X0%3rezi!o zN!~2kDQocDp(FQ2bzcbM@li{>}=+o3qzUBRBbNnr`fhSY`v-&nT{&B3+cu5(+V@ zap8eD9|Nxj*4()fJrY+$x4|FL3bEK)e@mGs-U71|Z(;&0Dr=RJYjiRx^E>GW?utCk zso-T0)x+A0G!<@%(8`lN>X}T;k0{Xe#98qwI>O8?_0W&srZK6HdO}PF4T)x8Ye92& zNl1TS4y5X1bf$xxWs=BYvWRQ~gCKW$8$Iyu%h2+JKr zuUQh0LSPPOZ#v}~iru%z39h_t#oq3p`zXk*FAerq*kJDg#r06+9@qwf4_U0?`s*SB zO{i1Z&Z)IvkO4W=RvGqw?e{43$`v5_98h;`p8z#y0hTdnwygx~E^97ci;Gw>9%?nP z7c|!&2_ePI`4E#9vfBig*(3B7Np5iLw7c>Hg3Z)Y@5V`ap?3uZe&wvX-8*Ehl{S;SCkqSz;_La-B zzg=TvDeM_yp=DX(d)Ir(xKh;`zf#$H?@}42#hnUUC~VYryz=gS7=K#3beu`Mf~}pp z4a~&j6H+v29(?58%eIM3+Kuv_@JBP(OVKE36+MC-Es2~C>+|S|IxwS}MjA_;?1;LK z8!hTag{&{y!=AtjzYF$}^$I;EJIsor6^70or0`3#-__aRtAgV8P-GXS+dl8syaudf z*8C5+?CAmu|E(O~GTFj~inkJkSIs5N*cyF-F`#(*-ry6^2`_HKj$lZwtPH!ROp!N> z?nbpL^E4?F^yv4m@VX@1$W?_H*liO!0fn7YmxLS)&ZUnAo|6<%qs5iuy|cdvM8n+U z2cuy|Ps+L7HrAgf^Eue%8$Jv^A;0cmU&$T z>XwJT>2x#6_ge0RH6G^U9eQab0%qtOt~j?ZU`}6o|La4^Irt>hy72~PhsvW_R1DOG zJp>zIwy;N(X{si+2==losx(ze6vRL1Zb-JBUpl#qoj-cP;Mi*N#*Jgsj(XI8zvh=b znl>YDZ0g(TnP+LFni_F64ATtI0LrCzDNfHm3w5ZC>Nu_o5K`ff+5wg^cz0?rC7!;6QPjTxgl0fN}K#aXzp~s|HbA3;WB1O^T zk5T;^rU|OjpHYVF!1_7{42%9a@XWLL%?lV~O^>+4LwgeOQ8nC{AHhzyB1^|rbN>%? z;pwq{XO++YlPr(1x&J%gdN5SKaNGUEj>fP|V|po8ypkNF_3I|CB$onj`+=Xe9muiE zXD{;UAwXDw4p&qr@&dbq;$7-Hd;H1r5g7x9=i`T6c)&|!jQQ>5E1_cRUh+=K_`i_V zZoHR3!Oft(4uQFL;bynLhgyVueSsN5*lo%#GoZZ3B-D6qiJiL`aI&_=Q;9ScpJ(_d!Br9mOS3Bo4T3NUy-mSbQ+5lT<4gb1|Wnl0%c@g$XY_BlY9+ zm39zA&!+Rw=WlCF@xPy&H8gPIww8&VAPJ@f5roX(;bikEv8P#{3e z-ILAI!;;H{nLy8E`5j&I0kZ^u4_Xa1aWlGZRZ#peD9Fjm)BTWT_T&>~Iro0m5v3ImBz?rKkqgdty&a zyjkjLU~l=Bm?={}NAyTwHyDCWK7I^iUa%YU%JFZ!G1Q6nTm^LPxXLmjE=uAe%pO-u z>2Ern9{Ty>J^r*@e&7&HT>dVed;$9~seQL)^IWSWYW(ryY*OyVlIX0BN^+Fq4pC%3 zX4oE6EN3h~4eUqsdO;r4D`7H`Nqd&*qzw`EqE2B!R7%j9kUGZ1v_jqx%9-9dB~h8) z-7`P>T8^MA%A`#Vz_SclPGnx>zMzFr`EgXdQQk>k{8l}rb(^&1vr7Z(Uw-fXOR_el z<%w%>J)Sv3istF}zkFrhh#}_?Zbl(Poc)|q zHcQq-aVCnKrgZDrh2jP%N`t)9M1e8D{G-vgb83b^<`$b%RI$HywkltWr9>$Ik)Hv& zAs~-9bVFFMpw|ZrW?_PY$OMvY73!72u*(c&T$d@6ra=2+25KBXROc~qT)cGPR=zls zTmj+-WLrBpae!jFojoVdon9Yn+b(H3=mNhMb)Uyag5udG?Na6SSt;_IP?L6VaK+pjx(B$3 zo9Sw0EpX+uOS-~)y|Itbh_@Ra3&sNvpUWp6^N)0>eLM`eGrb+X`aW&lnYw=vP)B0j zSj&M{34ByVAp0g-A_{FIIvVvmf!qykb#o^ri?zK438i)JSB{OXk3Gle{t*nlCh}8$>0O%lu4!ixv70A`+#@4B2K@ys}k# zOMy^3H1w?_nNuwB=w6@7Fryl7V!~{*=$=@}P_H1L6c#tmmbK$U?0o~xrULnHpCx|9Oik$hiA%=i zs5kmq(&au0?we`SVqw>QK?|roMH8QqAma}m+a-3@S zyP{BFl}IYQHZtYG=wdW+z25yEG2?;Y%~H^ckJd0n43g~E)4kp+XE{q>=8uu5iWjHA z#=dwtQrDI?G#UT7vXAVf+@tNkVOlH@L*|?P%?5TMg}o4ATf&%m{O-dLzaetY(b5*NnT>%GZ^W|K9c7g zh0sXm9?dgvXHSZJ-id~`_y7awi}%TP!7frRtC>36>; zhT*hWH#TNUY<8b)ii32zbV}Du8pAs2_ph-X(8fIvI6@2NRrvSL>6rkoIrA-S@xC4I z^+1{WDM@AU;UMHBDiG!iiss#qb}e*9-o4Vr*%9`!)Beprd^S`gv9s}GYHJINv>;~o zn}UdkzF2yx$Er)@$iWNf3WI)*2I#4V!v}ao%$JeOCS}42*GY3nrS_n2tj*z5zk(wi zVE@w}ZsUi9bK>9Is$4vHq6eL-m=kY*0ebwcAR!2ANxx zVjc^^7&Q}py|P4t2`$)v8~XNuCs>aROlSd_5AeyS#3!HrIX@mTT0Gl;8%`ciJjw%J zcpiHrDEjtLi}*{T$l9@$m-*#_F4Gns&tRC`s4feQXWoPVk^*|<^`2OW3@2BS5FLHA zKQ;V^i+3el?vY_>lt6Ld@s2-V?;m3oDw@r^{z4vkKo_OK9`!hiTS<{+*aanD3{I=W zi4~D6XSLJM;6oek{XhMK9PTOCU*q+qbl-5w&A+|$Bfebx&4XXJed1+-<#q+L}l*<-PPa-Skq%i;Z zNSZIjU`#C#EGLI-l7AXmG3A=7G18<(wznfpn`V)?ny!@K^{vr)zQqjC?m^?yt1NI( z1}qA#R~&{+JIh_j)Q!H*qmPbb%Q<@3SYhPY%Q-xX7i_2*kG>r}&njoWcko|#kbQ0} zXU^NmnGY!rLOh2k-G0cgOy=@rdUTfLOPfU)p-HD>#JS!^^%_BnXa{p>@@5|()XJr| zimOPYI-jf?Z&V*qE}YP=s1WpfY!z?fN`gDdbwMo23$#3fck`tfM(z?|wl}2oV!BI- zIkJqGFi#H~M;(OD8`-oa+q(#8Do3^I3?2)Lrk9ecMO1NREMGR4*~2hk1cwO!n^BcZZVe zd`awKJ5)RuVCQM4vEUW<9(9e}Tp|LgM-V>7(xq#vXNx7rU}WAtMg)hEKD7T_)z`ji zRY`t!^HeI?Y0U zMIM70ZLOd~#EQ=mi;D!R&$2ZbPu>@@Gy-TgvNd1?i6PC>2mQ)jg3`%Qga%WT(=l_f zw5SonK{=3J2Pxag$bwJZV9ZTWZ61R1ga!P)aXjGhxr#A@Bfc=9*PkHLCvc16_B9K7Xehu3*+<-QK=+G~TTI}bNFZ--N+d5KT}2M_RqI_dpDZ{7ujktNDb8r!dBgl>h7H94X8p+GS`=$t%9b8|+F zILFhdJ`~vl1&Y=3v%xu@n`O|8o56cfG_CFxoTBg0vEYcoG5ON1vJ9`-kWnuE&;@cH z6o$?`x^rGT6ZNBnx1xud%6#-+H_lnGV^~)!v%FhHIW%-RRo(SV3A%$Wn&7&*cm5ZQ zC4y4y#{;b5q@e%DWh8^&p5(^qZk0B|X9vXS%_FTx=~8oVQY0X(Bd7eEM2yYqH<+C?(`O+r& zDQI5gWSIj?=^$jDw(pDp6kq(vo3l#45jvpazHzIVm(y;$B2X4vZV2L1a!`hzmFq zhExmv9&L0ShkdQe{7`Md!nDMwbozFXGd*}tLc;)BzIcgw+~u_ijY<6=@V_TovC}#? z{4z;(W9*dMU}r1EZJ|gur9;A&4bzSUZ!WKn8g06Q4 zcJNcZY|5ycve;jyQKH5E#G|~!3oXn&e8?{H;dL!%%(GCY0wF>g)L?-!Mukwqu-@QiL3$nGxYZ;aRs}jU0pVPmpXGO z6Yi6xL2hhZ+hNCr?Aj3c8dI*O8PYQ1N^?7U)rcCJb*e*mj`wIM>Z3%E%O$@0%jmV>%p044dK3a%5Ckqlg!- z_RX4`DP1pUQ5w}tBMz%vkuS#sF+&xK!RLIy;LuoJh#C9J2Y+gR-VNqvQObmULF=gG z*Vez9{zeZtT%k)cvz9D;p-ocF9A`F2E%$CuiV?3C_R__(R*dgscLgqFYJ)4hTF83A zC2^*7S7?zLF`$sb{Z-^vIP_qi{?XKdr<}QoIte=#965vv#|H7j4t3`J(`OV`X>;s5 z-@QcE^7GHRv0A&sMmBAtIOsC<0hrgK!pUXPdSvQ8Gf`hIZ3Sl4T>60Q(yZlJqPirc zBr<9Cv4|uE4C?hh9-2-Ud9IuVm5RWG;7t4svzjA^)G(7d@zS*K)+LRm)WeNIVyDyt zYZ9Ov{HKyt?ERo-QPuQpOl7E~FM92S_VLK2QxerAU+jHT`4zGOiRgyayQdrPaJF@%} z-k&?~^>+GJGW=WSwZV*CUH^k$|HQh%tpEG#uaPrutj^xG*)pzE+%<|^v1C6XV^%vZv?4|7bX9-_;50yTz>>QfR6$Cn;b z{=u7{PM4_g|0|J3O&kZ9+Z$X6#Mm8lvrn%#BrfDjTiAobP8z$~wTezjO2)^8=F1>6 z;i#xHx_owq5bDwB5@niPk9^z)pxkb!ud+KrJ0Z$_f8Ihd#s~`J$hEpcaT%&3PuOFC z8X?dIYrvyCb_UQYr_J6s)X(C!V#tmIC?jfJG}=x`+lEZ)H)o)UZD9AQODDTn_jqy% z9EHV`PrCLBFZnZu?4A05hbp*sG{YhlrmmJmHH$JLs>d(!1L{2S9dCUWxx?axUhfR) z67b7y6<=cT_H9MHbl=PiOupFlOXKKT4EY|qb|GMPo@?Q*-^Xi$9i_h3P3ng=X=lm0 z=ZsYYs^ABOp*K?;Wc{a6x`QFrbiL?+U~6>zoJ47D$YFY?Bu{Wx+UL;|JpU13IRjce zB(i-Ti-bD@uF!oRJ0+jYT_Z?;7OfCR2x*tQ?(9Q)VL{HD{#KM|;#O3VG&fdd_Szt# zkm4Z9lS}E2MYq#;C*}xh{Ick#u)84R) zJ>Dkmz?h4YJ<-|pRv-Yo>s<;RLR=;IMv9Hc;?G5b&ptMWi#;q!_=;*MUY-mLx^IYr z*ul_)Z(Nq4mSWPzLr(1{qHEz;n|cXSoz{|M?yg^qcrU1+pnOL43QXE!RS&V~y*OqI zjbm^G(#RcG=Q$x{Jp5ume;BG0ZTjBx-U1JsJ`4~7VTqYh-4Isc?@ zG=Yom@sc59eiHU|=}=l@ch#bZ9in11nGXO%E!s3t$FV{Y&*;(o?Sf>KcqU%DCZNo# zg>8j6Pow%I>_{-`Spdc%fIQw&tOqUv#o%dq0c7kuZzgUYO7vwnRb`l6%8h-qWi~3z zW1`Xqs5OHP2LNPgwFa%GU*zTaXY3=lTor*m{$es`0wnhf;TqU=B!10FViAl<;>H5mxO>kfd z0AL{zlDz2m`<;LvggcDZnOc)A&T2ik!nh3h`3MIOWPIMbV>B~d2OQOq*Y!HS(2H8 ziYX@TE#V>G7UrsC7xS54j^K!>QQQCR*zab3{j@fZf!2&uw!k*l>hm!;Z<9lIiE}mU zr5mL0=27u)ynE?C%jPYjcFIa$TrcQi7vrVsSMfrHFg^5gw$Lh9vjh)KWHYguNI|zGqVnD#aCQi6H)h^9PpieNe8C9tIq4r9Xol??-}Ub+c%`Wzt-xd&XzsJ0g(zfBsgwo~x#Th$>JLyL+-cziNdL(@G;-Sg8Q$AJPmG6}%d1p#1!m6Fkr|jTLIPf(a z*(YxADf=ri%--O>qWxWR^UpG?W~}_jYmH&d&{=Ujl77H`!37uwdlVhbAqVxAO} zPQ{94vR?AL(tj?Qg8n&^wp4WwsCAoxV5*yR(6@Z|%{&;=CFoO{w3UK~vMuyVZJVZ! ztO+n`c0Z#=^dmrs3-IvdlsTV#^AGHRKPEBc99jRR*>y^7cAXrGgAUIbl&(GKfgn>_ zP8G@yO+G8MJ!{mOKnnxSicSux7ae}(LP!m0T<1fI=u<%zQ!+INgt@M6=Q;wQSVhT` zi|foQ%D<|P4UH{AKK!74J(1$#DYA;vVWZMJQF?;_3qAWi@TbvhqYI14?f#vPeVBG1 zci9bIDy#bjSg!u)yA4yV!lV0Viltk+;Csq0Ha&Rh zxc2Z)P(DBAN>vNR8SvG=q*}$CpL;K=RkA>Br9mLzHU>}#sK zGcng+e}F)UQCcOjrh21hf^i*;_dIT%iiy`i=5upKe`GTa2UyYA!fv0?0A(TAUc~nhFm1;(CBUoQ>(dA{FU@pU_)aBiHkcRAVB$7XD`b+jLUZ#~9EEJ$ymejYK+G0! z!=3R3p?!!=xVZi7&^F<28waq{68s2hQTZJnqEmdRQ00M+IZ=`uOxi6nZybkq|3(P zjLo73MW092g!qt8>09A_9u;1fWcTQHpFWR#K{vRT>&WtP=#<(Sc~^8!g7e!0-Lt~h zaCkFcvjaN&*c#4%wDfIq)5E&w6@2#FzmXVz3bPvvHlxiBnLu%I6j_O(&Be?`FC%2y zrA*injN+|ks=YO~^AS7evLgU;&MWboWns^$f0J7wA-N-YnDsBgi9`Ecwz%4X~;Q8G#0hTT78Oly0+Z2ebr%qBUUT z)&L)OBZF;x?}|?OVjY_+xEQ`@k$Ir#i>7heI35f^ZQP^Z9mQG&tAFo&y|shcy3cZAQkFVaje)?AqImC1JoLKPa>CMlDfw5b6=lggz}-iMvQr*}tL>WkvIn9xr1m~-60 zkh0l0+N8BSgHXFqN+ei4Qsq-{u0%jmlDJ9U`wSVldE(~j3+|*5 zJO4xOJG7?NPMIj$4cPr4LHIvNc>h<(fL@_E6E@I}1q3vf^0YvcgG&)Fm4C;^%_k0WAYD%YOrqWf3u@nTZ%p`w6^sM zNJHT=B9M69RocWCK*ce%e!&DTy2T5vV`G`0P4Tj-C4beQ`4jjE`Ay)qx;e?QQCK!o zoPi>1DIGfcTj|~4-v?r_D&a{#eWjocd?{V5nbopp@*}@hKF_j_(J>%BmUb?J&6CrP z-S9@x1S>R}UOM{@Ng7ULbzgZ$ifsUpMRCyF!$|4=+UK#vZ(mT8yh*<2CH=O@?kE$p zUue$cPg5<6=w-7~43{_pqR&k^)OC$ux5n(7~r9`EBo2;wq%!xIdTQ894v z!%lm7Rt^FTa*u>lMuFQD2wiqtn%>$EUxbHu11HCosVz-PGMqXCe8 z*bJjOQ4KMB6Gc`iy?l1VezI~LG-DrZcBW9=28tw6I?Q|T5@b$E6y#1a24JK%1=wa& z%@K;5GMX2J>12_Z8>U8`f~ISi$j7kohxG$xl;p z-@lApu)+8VimRi@5lXj}?wi@5z(j-9UiyI^bw>3q;kq#7F*UHU;#z-DWAde|eR1ri z=mgK*bhBt{aJjk*I+*2);U<#VwL@dG3iat%^jAr?1`mzu9`8@Qk@3HD>anT7N0&pV zsj@WKBzS8?Nn{)S`KuuTrXI=4s#N+~fBeVy8^!Z`F`Ni;>>AH6$$Rmqjmat`Jg_rr5E+bttTV@*7n+2)jMBz!z3g2$X zj?EC(hQx=JdEF<6!J+yo(5=*syjH~3>%jBle>|L*d~)B~iMPMGm-e&Dr=Oi1bDE^M zu^t4z+dmbRrqp zE;bIDmKj*|(&0mTj0kj+swwUCi2C+EhQt>fI0MCBF#U*c@j~&~Z3~OPHr)!Azy7Ud zAKBu@U^!$1mfaM$lLFgH*TU3>U~#HZjhQC2kD{AcNdhlD|>h%9JOpaZ*{J{*K9f4&scEGwZRA6too&;f=9f6TpHp*SD}3Do^k zt3T3o;O33}fRSqz(i>2xGO(3l8BV~KlQA^_uhGEC{H%RUUA;pc8k&arB?%?jN!r99 zG)kae?`y37pz@S|g;%EL#>^9%hccvgH#aG1P;?U$*DUI#v*y-lP-Vs^PQHn_s|~H0 zTA?*3>YXs(?x{W68A|wtVlGpFS@cr5TpG1aOBG{6%qli1VA;5ysLmNHXEkceUOhi=6P5m2*SynOOcqWL z>ZC1q%jxCg{`!}Kc_!_RnGbyrLnrDSO`;I*|Id5LukQN#dLUT5rYd5OX;#kK7SJ)% zX7LAq;JlBWj2zytZ|{Mx6n))ivK2XxxVja@=*CLYJ{v4-qqwaU_^x#4z+ntN#s70- zUJG*=WJ+G(Qf>aU)iYAnk5qSrS(=^X4D=(3fC4_pjCfs3y4!o^tb*A+k2S_MgK#H19m z*Np|qIU7MzM{!3e*l2Whq%aaSmMje-0epT2cHMa7qHCowgw+gr8hNsI8hWY?=*SC! z#wTBz0>nOv(QCs}R&@|?|ZyszI%G3HKQK0Rx zx?Yf6kRU|v&>u;Az0t_hLwH4;6OpZXKrfFls{53?J@t8F)TXP(8&$4O6tb@HNZ`aD z9Q>as4Sj3s@2ukH=Ik}o$j#@>LSHs^uwMHa#dTAp3n?~}r?!FRj0SZzFg7$OF3a+? zvEp(hTA4$RD*{A6QCMt8*`(|GYnPv@ha6*?LRejdaEPv4YMpjp3&P!7z1sGVo zU8Twc!yW<@dEjttr>_eZ0-AtjFUrK}dq^BuW=le{=?ZRb1SI^3uS6JFgV%N?<{v+P zQGJ`2#b=PJM*GcjTtMYQTS^04q`K&J$9#XUrVuJPo8+r}^xNm%AUWU@NFe$WILw!B zS0)JIU0{R-hx}?J2(uMNt~s+Wu>RHOv+r?M?0G-vK2J8d@m2?Z#6k8`F2zCf*-U8L z2mL{zE_b@684kE?S|GzXX3pX{pF%ofywp-}i$2FZnZ8|dZ9)zWzw)Igq52j}X)XU; zq$r~_7#%A3%M(YRu#bvil;try-uSn*GptB?r$Spp3fvee$83UQQj zi>^KWYTqT()`g)u1P(xm{u||6L7*(aWB|P1EXveCh(fa$xD-u(i=mArZ#x1x=>aKIU9SSXV-{ow1;1~(ZOdpfinE-@V&|pK6zvJ zcMa8o+}75zqt{rnNiF>JjlPLeWO>D+j8S8aSA>wSTRQh*h}7JP#Cgp~I< zC3v`zVIr5dQ~^ItAs-3ih7&|1xb}sBw2!!d{Cn>TjTK$TbnmYt#m^bK>TS?fOL0{c z**6dc!$_)8(@7Txfoh684t)>+s}5NfkwdaU4B_qKpbw#7H+T9qmA;AXRK>%dWngcI zUkuY-z#QgKz!U@G(G57zo_a<;nrDxl?M?4{UFk~Z`a|Rrd2!hPWVyK7>I_*q5 zovzdOZNDyGr_;*!{%5+*+c%wdrtKi^3kV7>pdl!TARvq4%I2ux4vNZ%xF8W62Nh6M zRQTU_5|u=b=0M;V?fdhmoIQA+EBA9h_i|lx!IxSyVHUusKzI(Wf?8Jggmi-t=}UB* zJWhb0yBB)0umIXPDB8v$1<^BM^6Zn`AZki^{{vfwh!=|BhJy#fJg=Ekryy-?UJnhS zL(KX9?Scg7-9vYR8xp=nyLZ?&+BvOT!_hid8k`$`j3kKmc^RM*c^ff`_bc{+A6uii z;#RLcFaJusa_lzGTT{+XA3jXjg@h;5$IuPE&6By2(9fEQ+Tl{C{GDai`E=RB*GX%C zvyKyIEazKAcJ5N_ZHjbI(M{q^%{s3$0UKQ_K;F#|RH#}3LMfL(sA;~tUY+h<5D-h% zX_J1j>zlFEDeYFDPUZw%FKt%Us81=+s8&s^4%kE=8?!$gV#eA7-}>uk$jpem@fhRT z9CS$4LE1&8}YotgFR!*5i~SpD|w?{1!X@jJKvQ@4sT ze4`bTquJBTCS_L~Fp!+KnB$!%&Ugc2@OQCU2sQ;lJTg2jqTpEZGd z#AS4`wATd-51W}UH6>t^;zD(a(5Zp+lZH^}UW(j6s@0VWUACske=Ypt7=&t~OT2Zc zJJqR55v-Yz!Qgb*<8q2UD{57y`u4hHYMLNMRjW*+U%-S#fAD)PRP&B)665K0nG?HN z9Ept+(`v-GCR7UQf=|#@fX0)7z;)u+ESPmieng48vrt;3Ub{-RlwBL}Wk6ZL3Hr7Q zSyX1>+1o19`JpgdW-~KLnhetdHAPK+TA`ag83&j3N~#qn35 zuxty6DxA*Rnz4Fo)WZXdGEyY7ZJ zM3%X1(wZMtlld>0;|xjBK92KBid{jG1S%RE`)~?M5nP-Z6N$y}$xO-jVj|mU1LWf` z%ovtIe==(J5C4;UhGXj)w&Uxs->{Y3a9XFBLzt(0>e|r7ugo2LNOMfyB}nqcLt&~|!D+Lq_CMJi9#+40fodWeeQuf74(mjADr$?0-p%-imfVEDCGDUHh zpaM50%ey8cTUf!Q3%)v_cMNEE?{;r@KQ*CBplkQYq8})_)XAeiY;pS#Yb+d#F`h7b zY-jr0a3sdKVbrAAP-;tW;j|MD2k(ZtUFuP~@I!2VKP&8TM+&9ds7`2g>7XlMef_+I zvqN8aGDXM-fBduVf8P7U`+xbJd_Kj_p-4=B+3nQKNi%>TtP+xV{2 zf`cQLo9de!+M=011?XXeJ4Ba=L3CPCO%{Q~WrwIC_@cbPs{@)iYeXx@=E2(KJ}IV` z6Z2JF5a-?I18W%2LpbDPzOsoVO}(vJPMW9m`WZwIMJ<{N`dvLFJ$6o&l6dzAfJqxF@hqy>O^yzn+Gby%>7>&Y?Xr^SvE1TNj06)=!cd>1-$<+2LLfl_n5d zFCAAh?U>wL%x|POL)y*UWrCGQ*skBq9DChJuOX0W+2wypp03fYA!|sl3)1H;_jmw> zx_G7I7aR7zYcA_=IqvWm&+Gf@yllJ@!=le@wd452)13EzMmQw!kX8We{jrrBkdG+G z*uCSC`&`J`2kqxR#^D#-{qP@!FM6eKj-*L^(Q7Gt+n z&r@szMe3>O^1w#v^^hX3N*^q+#ae`0in+>I54{?B)}e|8Dpyhjx5B%^k-Pe`*A}5c z@M+gi_V_h3<$@Y4+16|Bx@SWg`i_74I34l_U}V2Zd}S&IVL+oN^eCxT%^uUEnj^@f zD+Mt@`$U&jdS8>%))EpA_2Wj5_7GhslP@g}&jq@C-FClyBIFjqs+hT>=Z-dtTc^O~ zyZmpvEsDsO-lU%|$H5C=Rt9U3JKWeZoeBKWr}u3WfKIz?=1A#fYVug1EY+N!+C=Xq zn`ophy)t!;rdkzCF33;N$05#kT$AH-oGfF|KRH3~bpUhGb6zZ-9>}2E@z_CM!fSe> ze*LZOt8JtBPgfH-@e&xg`T8t@*HA0~dgLLe4BS7`W0E+JsQg(iJuExHg~tu}hx zRp`|>=|<@=PDA|gGH4js{Qz#fjG21zSGMaNr!^lOD;OhqA0|!+bVq;=4P06zsT#&{ ze>4#GjsK&ohwmr%`1coyedC9j`VB3bgrP=-5_gbN}#g-qO@0fm$abRwC5IK zPT1bS9N%`}af%nzYcB~7d6X+o4AZFCHxO1c^RSEtuYqvh+MHX0HA34Ng-?{kdH-jG zW3GvR>w-Lny3Fh)m7)#uW&rf@H!sNlaha(E`C}}s%n{BN=!cPPou_XyEa*0Uh-gOO zN~LARbLgG#o+ry+niWrhl^?R6VqrpBOGOuX!3+dBzsnd7P8OD;j%X-$rRuQx)YN8A z<2A72fKF;A8#@0Iv+0cI&m6 zcHiTZ3-y2K=r#L11oqu35R-UB68?E$VtBE3Sf>7gyxXCDtLCS_By&uiq2#WuS{I)= z#);Q69NJ!+JTOA?I0to6o=e}5E@9%v3=5pCplK(zt?n6)otV1jwPP)li*b6$1(NE- zeHlQ}(r4ManPQ=&B$J9ZHz%8_S2aL$a6*c;M&{@k2D+{YPeH#!Gqa7%hYkzVG=fA> z&~hO>0=iw9^kUD?g&XNOkcAw&D7~HIZ{+VI=^Rf4^-6QtCnh zG6Pmbw9(m+^zU}R7X@GHW~dJ+9hXHH_~@FwV*US_C0O_UCb|=2(V3cm;NHsDjN{6J zDj8qFz!)`%esjf?T9C`s`CY_&zzjMJvV9=}+;1yYu9w?51WqL`Ps+?A+aY2=u z9&Z8O+>gdCC223simB9U3fw}mKtz>IMaRoJXrxF?@@;g3JWad+9Ul|qpx)Nh(Ru`` zaLXrULLcjK;Am|GVpm!B}YW0en*hT|IJkrZQ`l>nJO&atxXr2=tJ{kW0 z+JPvk1u=_KFa8&a8%2O_yN|cAj$%_NvI^8b>HBnzGRv*V3l!o&iSL%8MqC=aoaDk+ zzcjdk&6sXr&$27sbKQ3`1;fUN!TZHJ{1x{9KwkcZ^G31Kw7>c;vEZY~#c-4?cxfz3 zrWMRrQEU=L5~*l3AdotKk^#V83n3B8!tD?=mpM)d}{2VplRV&mWS4^$>hFP9uX#!o$vmsK0_3)Q)F+kR1w7R*abyqr;c)y<+yDNbdPPHefYgd-?~*mZc5~c2 zlb#m{BsmeOk;rmm)f7fS8Rr;MB|%Mx7)0F~G|+Tap@@Aj*J;>W+deC6-v~ zn#7I3s(V|$z;m`wya0Qjv*_C)$qZgQXGSOJu9_vD)6NoC3;wFUR`D+~*NGDd z>#eL#62&G`Bp!$_#Kk^~1&8QFQL_M|gRU6#b-Y-pogR_B)0CZ$bJ%6v(^P*YPn`#I z5(6x-n3VQeBH8G~u-In>ixP@0rbq!5-JnQjuwM7P!C+Ks|^#rv?RvqZ<^>(AMnm zY?Hmf=`%lZZaNG#@F0G`VdwpHS&#)dt>3>^M6ySbDy#9?NwFmqDW;;U$+A~F=)|!` z)jW1JDA{$99#y@zN!;{S*8e#hRqA&AXWQSccx(2rSA1Kq-sW~4gjujfc&qBJe4XzG z7?CvkTF`u`wyV?r{o@@o)_!NBFgEg(RO{8P5=shrMre^$$BGa7`W%NDDk%pCuEwLY>4gt0sP*8~)SZmcdwQvM1m zuC>viW`nNC>0qM*^8~d@OmpoIyabgI80m)&XY5lv7H}dUS6J_jVmFR=s2zlShavva zgYJLv{e#{XbMwpNBTkW&QG`Rv9?kiBI*;rT$C54((XG=ieCJZs1^Fp$7k!yIMv`P% zf_yq3igjm8mbs;yu4w#@`HFV9Vp%X=v2rN+k_PvdAH5UD|9na83-gyBe$AFqgO7c2 z;M15NT}ufyzFVVW#c zQv+FhU8`47&_04YAgTrJ|KmH9MsStT`9UeJN$>Mu15#P!;CQJFMFlFU@W>bYz3am9jB zAnFLm;v-Ysn+2W1OU$kCvS2JSty3N*i$HcQUa)YKgNC}lff?NJ_xD7{9OGvKE$7BwJE{HV+y*sE)mbAnb0 z+CsV&r+thns9Z6KW=pb}Ou8u~FHD!>-%ih#faLWqQAal1!8;yPQSo>QX zjvh@DGGh?aFE8z2Cjvj(=LWBfoAD=!SxWmZybc0 zzH~n|tqpWUlifCI$_a)E(G+26#Up*|hs>#Iv66200n3683{f2RhU4%+5O(Cww|4(~ zB{7@C2CrM*dUX*CTA$*FqI&HCpLkI+Q%oDY>a|I}7+-7_RME?%`QSN&IHSAD@TUhEVZr$UPFnNL0J->BhTk@m!$To zwrsaAN%)0>8@JXIyOM6wOI_#t=1X^*a!4sKv3&U9hpQqmnu2~ztRy+;3Nw^|=sQTf z!O{2TA>0-mS^SjprqiyNq1dHQX?X(0#!+N3ijSitP^T(2bf?TX?$DUU+SFm07M>aw zdj|Qbn}*>4ujTRx-||JBwhgaN3ki-ER;-}w@{gMcrP!uII!xFs^QiTR7Yx@I{_(Kd z0}AvTAKx%c8~i}g@LqfPU|bF7&3+ONSEKwtF;}&S+2@U!t9Hmsozmt;U1sXF9ps`M z>11+A75v*y=aPDDa_BXATSS!%`h*-vip>QDdog9R9^rbN!;ZgydFej`Ee>f>TVX88 z;^&Y$F-rDZ%{|*Gb}I$Qi8gVuMBP?l2?a>3M#Kj{@IDQ$#!C0C^3vcWjb5EGeVcq? zV6Lb-0_cbWzW`#e%@K1&*F;CfZFGmoz|NTzBWZ!_?)^u7R4a5j?iVCT@ahiHLf5v* zEgHuftN|j&21uq|hoJ@^G%>gS+$0NfKKZnH5833zg--{pu&{$-L1wN9s&Ku^y=s-v zA!t^FGBd0ot?>yUCWXUhvG)?ydNvD6o8YPpO}6J;`6^tVt~)d9rg*j(B)_dtkRP4UQo*Mys-!$rD|I$?@Gt<9EDKUJt`@eEP{DXk)eWb@ykl`yO-XtHy`% z`pP=*G++APSto5#!^f65@BfSo{Ih0lmaJ!sRqGVkz=$!{O>{Ng18#1DYcf+oFD5s= zZh{oW9CC)l1tNoDVmMGvxF)!sBcMIuX+FEm7kL+#de;~q>D)}R(}|szGge0MAjN_=^EnmWY$7tl<}%%)(RZPJb#K5%S6w%q zF}+fQI}beb!k|h!G>z_Y*$Leuc%}Kcu4I}~@sBe4x)QJIb-`W|sCR-S4W^2*X9Q{~ z-9GFPmwUC*AI=kD$H+fsOQ10}IpVHDw-FLaSZ8VO4QZhlx(%?D(qZW6Ybc-E=5IN9 znZ^;9{Zkw4?0Pk@SEM}?%>SZZaj~?*4}^*52$}^Ie%a*I@X*gaejz;G|KodxV}sZG zcU}X&A@(29w(1{zrY0wLbAgPdPh3||v8fbUO+`PFfYEL6C=^`wY6&|#U58y@s4g%;W&~pThL>}E9)pVk4BNQbRs`gBCv>Wn_vE+x1XFSNkr{=T2X`gEGP|}!R z<&bSJO3s>%&fM;;;l(vG#t2>!?Q3t8?!GdfcCv#6_u0pwH z{+P4V@v38*q(L_;@Was)GxOZ7LE@@$y|l)bwuR3m;UIZgJ8joE{0l_evSNDP z*yZG|e9N?z?iD6B`O{;+Vz#RGdB+PXK{Rut=8Ai#aL<%sF*pN($RJ1GuG_!JI(x>& zVrp{leefxXdBOCUIkL>yF@_#Vg8~{evgn(xg~j0rKN(Gx z{N4hItJ9WEBv)TBNG!0TRJcd6T@>j=?bG$_Ig-p^y`x^eK(aOzD;sN-2Lj?{m4dwi z{{r>{q;P2!Btk@FGu<;KTeH@6t*3s>RTUQ0*0=)MORE4T5jty9lNc|+D#&Z1Gi7Fey|QnrjXVMPp%I&QH!O4Ht*(u97zqW+aH)q!!4$Y$`}cL`guQH8gPH4N1pfB zB{PLy$K>$>ARe29To09kn_}o`#43p#8GceKlzNwes3;J$;^$x;XF(^JHE&7v znm8bA)zv7jxH&DI0yp;Z-u}~6+r+igRx@yh}jjlF<93W83Nmr}bYr>J`oa-NUDsnSb7f zBB^YzFR-^+T>VM{N%8;Jonfe!vwHRRkF(WnL%zdp8~X8AsLz*t@rA849-laj6Z+ZS)mkr+D#HU8CDJA2Vlr`Bd;?)_LuQlBsK+OF-#&7%YhW zcpB{N!4HDI&8GuwtGHhhgmBb+BT)}5d1yd^xjPV&MYpCjO4~e!0S4BCV=pMI?;85O z|M_g`?K!*1O;c%v!$l7lJR3*(=jwVz^P7FUR(N zEoo~0)Z@UOTp59@)&7s7-ea5eR``7P4f5T7cwN2rjGzO)rI*Ct;1C_CQ=3C(K=O6dwN9NT!{WGl zw$}w?V&&r-gB?Eh&m1ZqzGV6vb|NTUV^&ih>luEk+{wLIRc>3inrx5xE7UN+r%D|_lcfB)wa%VhIW@#w#ir7z87 zlVdg6tfkmBI5S39fdq3b)k+Ei^y-UY_sK@rLWqh0-Hx!-Eyi?wIa%r6>z5-;VgKz> z9`*N--%Nf4k5X+V$sI)wS-sai6bllRB~)|{gpljCrJ@$iaaox-L3CTy2{r5&S&DR_`PxiP6QnX4?{Qw%#d#gRen_L&H2vam)Y8$# zKIq*wiOV5)`PpS=KQwuEK*Q#mz{BL~4Y!8>qJ5 z0v!SD56QRJ5s2Z7279i5Nf7h%q#b`6WidEf<3~C&+lf0l(yT0DBE`m2WC;j;_b>cGVQ)(Joh ze2B6eOJ*nO8K`vZsY{9X1Kcd=_(XU5FJvJG|LU$H>L?+>N3$ZCGja9)2K)?s}O z$0mw};*a%I^maifvy)0y*NPjxo5Wj$`SeELuT)njSI`%j1koBzCt0IG1)43e5^D>o z0Ka;b%qYG?J{L9%X04{~efQuS3ueq3LZiYHDo++%rmg29aljWu{_eL|Y`KSDD5Xmv zSm+aRUr({%v#!MovrcA#rVwU}OuAGUGv%_^M)+A4oFnOV!7Xg8sSsDu%?xt8>eW}K zR{5VBQ>hq22s|32fkENX-9z`2+c}9!@?C8^)i`Yqk7M#ca_KtnQg7Wg?-S71nK$ul z02DB%(P=^VRGD;=Z>A>GYZ#-x_9bken1P`icH1X&!=Rrsp^-oQ;8!J~)!+Y# zB)>FM4QS2vu`HPs3nGPSRCKQ_Wzu;drs!d6pxJMqFeRc%y47o+a7WMsd6(c0vq3oD z!yt-duS~50#UbSP zA`wNd=yMI0r7u#R4#q;Un`4hF_3B(vlLr6vuiJcaX{LL|XuKQK-3A$O( z%q)~{HjYQuAcA~}PTlT$gsJFR256}H`e_+dt#X%@_R9yjc zw_Y6+R3d&v_SshrLb~4Q6TU z#NGwR(iS7F$R43PJNZ&d4+A7S&GBMF>mg7;@AM*M?_AERmkN@jV z+vPEzrH}KPUBa=L-W#w*XaL^^n|*aLK{@gg?>eAm*MZbzC~}Qp6~Q*Y7EQ0q-heW? zf<8OFTZnwYL+=3%5F^hTUWc5c+io7Q#mLic?~8|o44^YApZeipk{9+!Nff4)OHBD_ zy%#p#T=$<3d136i(It{LDH5}-bu~s9!F_YoIhaW#DybS8OGnjj9rsRsdE?ZJgjC2xI{x?bY3+9PBq42<{GXAe0KVrm>f+OFu zCG>ou><0&JClc3hlL8gqX2~rT3NoSLuy9-<|6K<3c^cS4uX*6r6#CyGx*SQ0;_fgM zm<-T-IL1ai<0;(cBcM87VcWpxw9z(>&cYZ;7c}Lz`BeJM5u`F56KVw;M4PnDuVbN6 zGE<{81h)C?h#VHekVj*{4+xL$8v38yKPyKDhmOtBapgAo#PFt;Q9`l{00d3vl zsrYYR*h05l>MYx7T3Ku(D}wdDYrJ*H@fkA(cdtxEj**+L`vW)2b+w+bCW7;KLDQ{t zoZc5HJ)0R_qEPRfB5tA+g;_MpmiM~cC2{PbF>R0-Iu0rsdIhR{-E`}k+%dr{Wj_8FnpaXt0ABnCP3;rEb{*0|=hz)!A}W z${5*TF{=2vXNqFCI$vs{T}7Gm6YkqJx;(d|;tj%N1}3AMt__NUNstvlrSlw#qZc?1 zL(G7&(f^IS5Hs@KE6Xnb$P$JzP~Gp4V=qk@<|`{xb)I4yC{j;Fr@MDDb?Pj7yHba! zNoJCTrD}*%qlMb@YBI#86T`0vyCach&7?PofxKqsdPrk1tf)JP~gdQ;B;)eWEuS{58_K6b5mVeD4eLrEx5Dbg?EL zwHwkU%y5#L@P>y2hTfh>@Pfz4{eL_{q?Q?KQR>D2B5_XKPrTV`&RIvXDHK^nMb|=& z0_x8qJ+xl^&~Gjhi$icOeSDJ`30lzSDGpyC*)M3Lhp(CR<-mu}mduba3qGW-Q5#5s z6XWBU6+S8`7M$Q+RCI-+M|DmxZ&DMf(=PtO${(N7J`5?=WI=g?f!;|qdY=*9e0875 z5PC&BtIlJUY?*SdGK;Qt#lJ>zMR1+EPK$-I>B7?S#Xcv!PHAV=Ym2;aw@G|+<^_3t zaD}D{2)0_nE-~%ktdx#NJ!J=Y#JnNI;X+s+w#SG(`x;&*hWggK^Rq%M5K8yWSV{7o z7(z#^AXG-NyD74Rir(Xs3t~+d#(u8J4c!rx34LE_bhaPL7+ld}A6TpM1_)@Zf2Go= znYlIP?DQpW7|uzYTCc4WoR@zkE*2IGGRUnd23frppH&uogES~&ppv6l&=r1NnKcQM zhgpJsp?Y=Fv@%CAb2%PB&jyCQXZZo>_?Xa2TVcVcL%>d4ImIC#_}~W@@ zg>2>66{8F3X4l29=S{sj8JkaY!wbFEgZTPg_a>>qtw#CK&jBXIju9FFXzY0E;4k3@ z&AjvPb=po+eDX(53>}UsEtAft(*hm<^G_D6mX>=jl*EPRLZ%>*nJqz^lOrjD2;mj5 zKQS}#ntJV{Yc7VR3+n}Sq4nC!VJBd9{yDv#93v%eox&mkK4Y(VuNd#UD1Yb|>v1c* z)$`Gv4nf}lz-2vq^|Lp4XpI=r{Kh{=z32+bv7-&{o#BgJA4E31vU0rjUY+*1cPe-| zTYh@`Tb18v`ITV?N@FB5$GvMnnG)|nmhyV-e9>wUIN1E%a{#S+ZME7UYKy3nb^vj=@tT8ED)@d)zxDmAgp4g!p9E;CTpyM%M1_2Xp)6-wd zA)icivE@%FvGOY(6A!f0TO&?-<&T*y*yVrRTdzhdpYGm4XA3sc2VL)mT+!lIyv+K# zZXw=)m4^TW{ z+~4g(c8nXw3FAL;w=IEqp_WJ-Wf5oD*<-GND}pLr@q*K&kSS%sZ3*2f*Q;?KR^Bt~ zqv8d_hz}m%7)q1&6v(*2@m6QqSHH1LGV|8OjwctKH~{m|YWBN9v5;uIiiA96bOJde zI5g(V*AKd4)urwhME~__Oe~__C|C(U=n%l9QJWHqzM%_1S{7XIJy)s2V^|*EC(ju#dB7fW@QZI6ToXesk?qqU19_hq1i69)&)K1E^rKJJnXk(7?~vzC ztO+X@m5<*AgfF@#dZ+vvV>vmgQnVeG>salnSMTvJAHPF=L#@jeB)DcxY6(dlTN<`E zuvm3hQ5@Dn>ssAoB$uR@p|0ngK-UMo{LuB-w;iI5Cm`cSS3e%`2xijiGl3Sw6}1({ zk}M}i9B2ym@wc{9>{f~tQqiX38nub*T>~Py2!#d(CYjcTE>d>Lb;zWUMX#A~Y69ja zZ@YDd>)L|WvpKTMA(+6<6=L}WFf198Ai4~5> zEc6{a#{sABQwP7z3pkYUA@z&zTY$6UXMVpVhn*NWSFM24K(X}{IYmWxGC;xrER{_m z3E&Uwwy5&L3?gJ5GzjJe_WISS6F_+TKK(G{c8JapdWv1^|5)$6g>IwQuvOqlANA;U zNfF>2b{Lr9^=eF$*Qs@TCp1c1fFH)F=ymA;`&8_?Mb+zK5M2?@mgt(qM%Ai_PT^Lc zO&$q>dNqC;5Z|#0qRZfX>D6b5(f^RDd^}!pAZ)*APGq&Z*}lMoE{N9#G^;hVm_= zKaf;m6D=;^%gOMt&F}<+*$=U2GzWPNHyHX^bC1;h_&Y~l^g_?G)3I{Aibm$4WG2(M zXety2Hz?8|y^%&^c^|m34T_G4IyLsK43oLw5062PyPdc5g2#yI`!^9^iwB$A`_mF) zaN_k*nUzmkK(WAnn1h9U74#*>0ECq>)Iv$6cgFO?(-%rIY2qCDuzOT(_AO%U9FfPD@rTUQi<`#} zbKnLq$WUXy{O~EF{&TK z`hOHRa$G-JTrRYr#KqqwmdtZv!?M8&5y=$0oFdCWn@`p)T%t)DUk?@H_kpVAF4+z7 zjnZ+eKrGdfG1~s$t?f~}F<^V+5btwipdYg^vc9JM=eAWaPp@U1*m`i3?936xN^&JN z!6$u?ZaF(F)1zKn;+7#dk``&a;J)yx$mnq~T<>uj!X5jdur6M36uN6~OII=2LrdK-R1UN}xMOCY4R z17wSEw)V)lUYB`+ne;AAUf2TfEX@MR<}h?zXT^gqfH7>A?hJahJ|0QzVkOz~;b_8rY=w9t{PBH5y)k{k$24OxOREwCzaT4L zn#>T?LG&q%*hsP9JsYU#Wauc!jk*lt)va<&THg%1L$=BXj+^uUOy*C4A*26QG2=h!|TC3^<>@x{Zq!OIOA`S!ej zyknc{c}Xk(two z!vZgR;M8aSFl`P8s%Fhtu*Q}zlY?k7oh%oXx@``~RTPIKPm^w&>LzVgfHAShW0Y%G z*l@m;a)Dh@vy1e@E1F_9UXEoAG2Hu{P%mc&Hr%rbgS z;&{>uMfpY=Yq>KuO&+zvp;5v!8yb)!Cm;Or&$|D4?+@?)<#&&J5p8hr%w5NUni-NKDfq&HvG zEWj9Is_#eV0@PFIMFyyimw8I^Ie{Z_hWgIIte^td+=&3wtUfV-UJ zyHB*jzg}DDwOR7$cJtAqNe+?|cdT(Tf$|At2C?-~s+^f(!Ow3_PR%2QqsVcq7hg%S zPzSJ^ijEI9g1*?^fXeCcEzStdm#$T}NsGe`p~(#H!9kz{6BhVE`?P9~KkOtjoBU2G zYs7fn4f^QUu}&4-%S?@t&ep_^u9YqmE&|~PRD%FIR4g{xIswdvF)UMahd^J@2CvHL z8$vS8Hb^!9++ zf`{Gx-VBx9yZs(GM0I*dz6yi+4p{EJ z@#}H z&1ChMBN3m!y51KoT%EcND$8~GzRk>i8hJELy^%{JFM)0>rng{y*aR6YGwFDm?3^@1 zlO}^@xj_)e8bWcc*b(sn;{J|so}CpCyxgFXzw+5KMqq)@pAL3alR5m{Am`0_k#sAt zt)SQhio^lKuckEYve#XohyI{39j5C5Nt0GVBIy*C6Nkxa+55e=A4B#&(J|LmeQ)IG zi5672$TZtYrV|^L&#lIJ8^sn;Ab%8H;MSsPls1WznOI4JB+~=jtXgGD2oh?>3syzw zHRyriMGXo(4ORTf%mYv;z$ACC%Uajjua>CRvnA7xL5+7>kS;&WC@v;2b` zNpYx_h5bl)>_BmEI0gi!r|_~SBMz<}_x4l^Zqk@v<&bUstceplG$*Z4wx41_{;wQX z?4d?+BZP7`c_5v`=QA$=nM1vHi?D)D^WEUlHFL?^xSuK6;BkFs(wiW>rTzP?&rZtl z;?(ce^|?6to99e>nUbS3kP)Iq+Z$3WO&YJC@$pwP|M=vQRrEQ_5qxMBy=TV9%~5|G z;DqK-)b)$Ufk82|*>m<_R3n|ZUV%e3(%f2t(Qc%4FJumBbm!$DA_*=q&}SJHxV9^! z3k3}<*fAu8z*S$BS8qVC%XZ-nsi_ai(wWpEOqXQ=`8@K`bjXuKGc>4{K0t)@JAMv@ zO~0Qou!H|M>&lO9ec_j6J;mWU8w3TOn71^@3|_aqp~5{l7n&=ti4J&f3p^>@3XPa~ zP~O?((Gij8SFbH)%g46^^J2YruHWu}7|C9Fn!nz+Msa}Xed|njwkknS17Z#Mq+aDw zHBNuW#-zqj2QW&Q#=ZAa+UpWCWg}IgFe|7y(oQ`6iU%?DPe06o zM|in8BR`#x`$yYudZz^nM}PbYzb3al&j$9qtO*?7Jh_3b5$}rVW?FzwvsDR%3SuA$ zm+kSdP#EFJVyYzK!fSb+OQ{-hrQljfmRniyQtwr=Ey7#jMpcb89wN(F&bd9Zn=yt} zMCjFDx}Tah+p`r`_V|RIvYTFxLDnD;)USi$DEq++LDaP3yf=L;j`BnH@FJpj;sq8E zD)cFj%BNV+G1x>!8zd`*%@CPRmp$34)26-q#T)Z}Q1ugJT0qvb#iUHVDAFKr^Elvh zo~eM0LmFL1uXL{ij@~l@VDbp9AjNctC^y_O_;|`F3?48~J<=fu_yK0d@_!tE(PiNB zBd&WTk3T0(hTO$s=v%9m9+y>srok$~>9OsiT6Lv&1zkZ-X>rAbMyXZ4oZvn2k&A-! zBdQ?Zktta!tk+i28R3gxSq5s>NaC?_d`9>xAk{++xC#Ys9}6fFWQ12kq_AF_Dtq*) zRW$a!3<_j}hZ2_w8k`fu4@%=3Vn)ue*q$jLpKKzf{k26p@e&G@y876oqZA9yaTOJP zD(d6Uu-`$4x+!^M3kA7!EVaTDR!f(e1tN5e_6C%SbSO`7iKIh@d^hwgTobJb-r-)a znjehIsFk6OQXS?;_XN%lT^YJ9piYosQbA7;oe?w!bbwQJ37YzK4?_wCJ*s$7n-oLM z4GcO{MFHy~HiBq6ws|)wiU|aY@bmZrZYVKkg+87}j2mPBc(~)wwhfdVeTvOYkE#h+ zeRCu^lK9{nRhbkQsSn4Uh)AOc$Gpw+e`C7b?pHmKhY#W&?pNJUcGG#kryU9ZnfJEE z&IHbT?I+~GOA`!iwz5E{DHfuGHB|H+Qt4ABy~Ny8B{N7m|7FN5Xqr=_Wx)Vm_Jjt= z@mB->tLf`gj?s%oWeetwN|$v-%pX;+oi{2?_GG(Njz_ELYPui*)q3u_?<39329Nx( zqw4h^KS+5ARnFz9Cm*{ ztnA9bcKCc<#>Z?Zm}CLgC!aR&A)B1oc|KqTl^qnjjUq)0gF8c-bYTU|r=i|R`_331?1veGjgsy?vHz>PqRSB{dRRS+1{V; z~W=T5$F)=LHl-( zrh}M3;RVL%G89f^QRzxyo+On-D`bid$4e4Q4+0eKJMB)}YY&{G0jV_i#q1hwa?L zCy(j8UjV`}6Cqax=Bafkhk$6pz6Y$=+d2UzcJeFM7Z2<^U|VGEIBCUN;q96?$#;8EV_m&&b#}$oU`O<|DE!iSGBi$O2%yg<6 z-HITpGQ{Aq4=7eMie+bLk-`fo)Tk3P#9E8{VGMmWh3tN5B643?IV*=LwwfaQspvc) zAddx!+Rtu-=3J$y(R*QFp{$#}7E&pI*xSrSey77~lszt6q89nBQ}jT)K#b(F7bww9 zt`Q<>(JI+t`s!p1oNW{)cmnF8ZmafaYJhaD{z1WZn!gGlRwnN1)ykp_jD&~{L--lt0ZU%Fl*`QR__{de^Hzke_KZ}0!% z&%ckB)2s`jsQ#9~_0!M2JHPZ$FyhgL9+Q;JHBW9AL@WRc-eU8b^U*ERQ4)mLBvCkhsQ{di_?-2khNA&ji=Zp z6j?|`V}uM>C7rY`RA z1$Cgf*e1ogY5a@P2wk^3212w@1K^2CRYN~NUXez}NV44jml4q#T8`fPLqyM+7rn#j zxFpN1Qh?o!di4f*vjD475+*L5cmkBPK>M7oP+- z*DOYB#-;3~B!}M;%8C0WVa3!Z-n)xpw^L**6|D!w?!<`B@D;&XV6=2)!7GBzl5S{0 zdtGXzD}p!3Z;4NMmIWvI8X%>Os-}0ttDtsRHHQvOO= z@WvU-paiQHU>6hA;jUNXrH+D}!=UyEehz}%&u*Ojwrxd>)2^5}YFwH^^1|+t3dL+m zHVAm-$gqeG6f7bz5nu>i1i{-Z`t;a>S9Gfcvn4p-ZS)>-8J#BUQRxyyIAbH*o*4<4 zIM~VVq!$9;@pbP4X%@6PZS&K0D37@<4@JUY-CAJWT%)Nbdjz*tS#&17W4xJe*qm3u zr|8uk6WW97f>%c3I$|%m1%$*raO7~PJc3NWKkdg6FRL`-l}Wx!-mzGvPqTMjCDr|v zH8`>DxngB|>M0iL98OTt$keM>qh`($j}-rQdbT7-hWV#9D8n@82~a>1xqKTGH(mFr zxB4W>vO)Qv+p|iL0|e9t&_G7dE7spQ4pcHs1WL)w9seeA8jw?Ct~-Mlq`t7L*#ACIhfY zdt44d#4s~7BP?Bp<*6wWqoSJ5@;O9rrFT+`A{GLt&e{o8{&@AB2>Zv~?k~P$0AlyY z*?l25K>CUPP~R1MekHYlW>Mu~icO)&Dk?f<%+=Sh2chnD zqv||pCtp?R)yd3i#odr}hy=Hd-9#sXgaZ0fcSAbdhX6SJjwEk1_4^dZAAE4_mG{q&(6kwxB0}Am}-)-{F@Hxy*D&2j}E2VDR z1vS#0)P>jLv}Xd+Wp_g^g&Zb}ARDnVa&g3pU~H~S^_{QPPpwmK4oDQ$DC@OH#T5Y< zs|K?2ajjGKiDu2JP%Mq8h6qWo6hEg2auP?NJ0V+eA^0%4?q3V7R{GZ}eL#TBt4^8B z)VR)mrB!Kh))nRB*D8+jqDkV7tbY9=I}h>z;Qco){J@r_+-Y6x^;TGU%sY<#BTx!) zARu0bdGTJC!z3>ZE5i_l;CawuI$&wZdkG;-oeZ*_D?8XY$_nz0M*?%!VtmKZlJadEvE1C`4 z|3)44dtWtUidR-I16~K2IaYUu)2EqcKG2TXW#84)`AE3 z`(0eaJqQjV`2mzs! z@Vy?(d@Es@g#4kG5T3B?I}?Z94I?r^E_ZnbAcVCMqC35W98#-(v^^i*y+Ahv}1gtC^0xz z)HNB|Ilpw@=4TLe3YYmdiBHnYzg{F<;#&($!kM83n)6}v9M+-K?{GUSux=I1`1@cK zxtutb2i{5_6SSIQlPR(s-4~<(YIeP%LD5RN8y2(NB}YAmZarZ8f8q05O%uND1BPgy*DUY6mUo>; zdihT*=y>-i^%gnl#Expa6>Kk4>_v(+Qqi|noy^Lym7dMe9ss2`P?mYX=ZvCWyIFE< z>;rjAm|oK`qs<>8V`GazqG09NuE}4*@$(UTq&tF+lXg+Ie61oCXvFKa_q=aIhR5X9 zLg>F<+YMC_M)9tx=1-U{>K0a+XovJ_^O=XBGqjxKhU5N;ar@*Cfk+Rx# zUiQ!%yRqxE7)^_T$?K~3iE-Q2hVc)*_sQ)M!sQMBXE?n)PCfTa!wde@M-e^eY}LJ- zHp0fC*VP2%W(Gm70L6?kw-+zF=v7896Xr}T4zKmx2pQ{j0o_nPc1)NWRH11b9%NJ(Z>s#@>LKNc4;0(fFy5IgqsjMLyJ{ z*D6uEz1!UonoB{kpieGkSDP?@c=50@oR>ZyfxO@~@|Sauow02@;Hd7uAzcD&ghkWz zzGk7vlIKo6}s_p5Xd8yN~li_Q)?U1WvNeAMy#|IdA_3gzNe)6dz|B zPX`+ns2`W1sfX2J`Bdn~P4ZnCUJB0h7oNkq>V_e;8$@;wzhf@^mFUDA+lBW_0t=4# z!RJDw0<%ZRf@pF&u$G`%P(^PM8U$D{RS~!#^p^W&0L!p}#BOw0jEG~#{`pUF-}^?F zEd!#Qh<^Y6Z=!$y-UqY(@~BVFX_rYHbd5z!qjwj5L0+fLfBOQ|#UZEfGT&DDEyYG* z_0T4S`g!SgkA6S*4&@>43Fg@QD}=VVc+oI`)!K2a5o%9>4OOi}_Cly00-eGhs4XzK z4%>I%Jc4#(U`VrqPos$&1O1doj%@gK&6$u2@}}5 zu$wfd2rqjr6?M}I(@G-JWt(U{sTAV*^US%x_~0(mEo>66po)PR@e;E=pzpQWf;3ri zz&fuxKr3FOJRZ;~M7B%=vo0W(WC^!BLZQii#MyeT&)R<)L41$Y|21sdU>cE}I7`IQ zh*UHqVN$ZNXzVs7mFbv(iX~4UsnZtC_-p$2_k6<`b>_|HpRD}V(%;~umCSTO3h6LO z7Pdna_?!SQI|Y=8RWv?w!K-?8-^V)6ra2fqd@;g#ef7b$KN)9PCTrprm6LT&yiA5w zai0WN9>wNRKqD311SFQ7&{uY!u2EJ%&*ExDraxBBo-pyg-~amE@Y~7+sCsSr$ptya zc{~3r=Z%UPYpBMbSG;pJ3a`FN?-jKwQIH_Z&2eLH{n7`G0PDv&Y>(H3MV&ji>t|sW zyhvT6Hjsi*d4c-_TEf-=%hN$WXxPpGfuhZlHsGa&s@B~B z$K;2h4E{1)XX@5Aj=M{;H7R4Ty-U~ORqh2WUkg09E0GEU79LC$0a^$RvLWK-^-wsn z6;%B`g!re&1J5ja&4d=sN{<@4oHx_D&AS}*NO9VGn-Ai-%WI^ZH!}Av|8M50ED%b2 z_4j`w$xdu?imkAnNwE++Nds&{E16z>NKq@k=817PojEF32LU*Y=dTv*gr#z^SM~H* zNtb`VFZ9YLGq`>(9hW1+u>MdP88e~=0brhxTH_QIl{TZi-zFIbEyC_nhFr^S5yxN;rn(~Tm9R^QSM z6lV<j8DueO5cG zC^m^AiBM<_IzzZ#J>{AX{6VQ>+ci3r8`|KJMQ>N=un&Fs%J^_xco{NB4Wu2<;b+Lo zcE-+s(aChNhNM%t)#uiPSjlPE4f0Du#}TR6@xq^6Vh<#qTs?eS{2(#@-i&P9nEgwV z>Ep=q80iEEpCH8s))C=Ef;9xV9Vd=N#yAqpA$O} z921Mt#L$mTqB^W+!&)~SezUANMsCWd4l4xg9)AGAa6*9FiiEn4SrKe;FBTP44=Jf?Y2OBfiw;7XMA#B&(1T{~eX zgF!q!koy2x0g%;Up*5Z{iqAk%en6dOm8#Za|vH!s|v#*r-E;M@l^ab59<0#yjS?Qg>*6~$sQxYFWcV0B;~E465Jld z5LgUszfTyNgZzfTX#-#L9?-XJi{f9BIfSFOzf*NPB$-+7yK@@k7n2zjY=(49G7}5s zB6@YFuqenxc!J$qj&dNLhd}$xOdluWc@K2Z`Q<-5mNUvSn`n(6>Bww;4yqF`YSOIc zo`(zC}rE2Gf^KOjQaRwlnz(Gdcn2; z-Og!WMV^r6NIF&Bp1Mk(hu#<&(IFG*a*qdc?9b{2z1?+QO=7H?Z=s9nQsB5XeGdN+ zto`ulV>Sk|X(ucmUB+u#9J%LT>0h%mIfvY4OFWP^7+G9(7;UmN9q7S|-jd6A zT^;nv)hMpGVT`-iW$l=_*V^6e>T!GY9~j8bfe*$&KF$7n_zi~B`Y7)&eAjoX#pI+h zzse!oUNDkk5P9#j(AiJ1dnr;*MIRAd3dxZrGbNK&c&E56)&hL*sWO3IIF79p9EZx( z21Vrr-B#C3dY*T&Dvd_oY`mZexVjVyT}}2*gDxH8xDCQ>Y$jdcHkd`#z(H@98L}VcHdk_kUi9$TUu+}v zPe)>%xW1Jm4&NbP;N2F|7Sb`LjV|(9=e1AN3N;L-{u>}L7-ql4{sf=RxDE9#H!d#w z>3;Zqi+$PgGrwPw!!MY9xoTx!8YmY0o>NqGrUvVYKy@Lc#|8OZdR#UKC9f zS?>=Y&Y1IEM(bN^zL_7r{973_mc6xOR!LOBn>lZ9{ASvl88hm?x%nGs zzL7EG_&Xg@N4{S?qu|X8?_PUbcjKF1{h}o5dQ|bR&wS(5jO$USX14#f>Wze7WXvde zBmR3OQD44s^4kt*JbpZ|js?}@#}0a<_TZESFS^&})IQZ zd1u~=AO4h-^P6{^7!!?Fm^eaF>(`kr*#iW2Bn}djm_|YM~{le0(DaJz@eSTOL2a&Jz0&)t{?k_pt$iYub!& zzVogHQOpwm?~vn89Qtds!pH@RJx7r*sOT1`Su*>@IQ|Wa6vbVIDQuL?V93Z)`VCCz z%}lmmgCdJA0B&Ohk`CQv^PP9e&S_V`osEga`uvL#JL#Cn9NFH0T4fwtMxXafg;pb7 z6Da+_(jN3D(e+Lcffg`gv)co>p%TMW$1bo~EVCX^i8sWwK)^8)lWw`fRlvIlK!hc~ zCYGT3P%iZxe`gpl!v(HqVu{y0Hsa0ARl7Va2>VSVnMD%$*(@hsyMv%mpBPX&#X^Zg zD&SFrt);Q#qKPeG?hh`S*b-t%JAIj{*CvMFWNtBMM8)INWpUaAKAZdv>>1x}!=mLd zAVfVA3ywK?`TgRLzh%M5f#06{itKY@E7W8)>U9)bOF{o5w4fHTUxnz3*kb5i+&Qf{ z9GXrvCY8f9x?8w-Dw4x(0gc2=%{s5W0d3OG@bbWTK?|+>LeM1M_S(^)%w!_vGoPZ3WWx-oiO*4Qb8>ID6=B_jhNuElfp)5y&KP@dQ`D12-cHYJrD3*6r zT>*w}GhGNud|5Qc%t4y$IrgT{SPlp%T*kzX<9$B*nVbAf&(|y1jA<5Fx! zkSYpi!Oar${=C5AF#CaKGulIrJDbn5cUS-O0HdD;Nq^e1ZyDLZ4@piuiB(!5shDC5 zD3V7-pO-hgw+fe&G+CLn?>0*1oqhKi;qS4atMK z6G*-VizY&Ys9X#a?LymiThG?@f4a-*&d$zyy2;t*)}6EMpSreZ8{*vxf`SSv7r6-{ zS8u4?yn)&(Skx$@0*Po9MNmY%@cVp{U`ZsJF9c5P&U~%pdwT+T-uXPA=W}_#Z~Om7 z@n06y{CC4Ir;_XJP~*b(XN8qE`vZ!(N0FX^@bpQLfIxQf9J-1?y+xk@xA9nyb36Pt zsQc>DLy=AuuAvi-`>cFnr+c?NHnfGHNGDRL_Pbr?A`g5D2klt&ob=FmUW=j^>VhG_ z)I%ERONs|xtu*qGB~9peixYRz>C9beHK=r^(W{hTwDtv_6k!w__lO;Ib1PkU_}Mgs+&c+4WTW;7bF!PjU0SRhYFuLq5A`G z$e{NMl-qNswgjic-r_vRW!2fYfA8nxlot3E^uN24q_cyc3mdIUE2C9RG21AT4}{@@ zG?3Cu4jlMt*7fXG+$BkqI~91PQQIf%B`KVR!s8%!t_NeL*WQ&M7sicWH6Ccfp$KGW z&|R`hksR0^9tUSujUScplQS%qlZ=$rFfl1-cH3%ox@^O}9rev3?)s^9!Vb>iN!z^2 zsdV8Pwb3_2+(T;V+ftCH;dhAg0{-%X6@8Fsj3oA6N6fdAT+J6;OW}K4!aNAnPoL& ztnmT6DpqeW%$r%^V_fp&k9ac6l5_O?Z(Sg(pR=?(IKP8JI-4m5aupk?sL!D_2^|RJ zlhP${Q1!%KpO9!B0VCU+bQev!2i#^LSOdKA(zoZn9tz= zXJAJ@Fve*7v=5wvMc(Hn7LdsQ{;fu`ik)BL!X?Pyh75vF8pUj+NHP`GE4VH&C@TC6 zQ~JD8RTq`JG^U({_i3PL$mAM99=hKRujkS+Lab3`UpqRy85{<7hWAD+pQ@YwSD%+H z2yqi@c9Cp$2ytQ0^?(&Zc2GMDn^j9fYv}~f za!-SNg>O6ft+=_qOOgZY`Kii!>26g5NYbER<{U0iUs z=02Sqcog2KSI{$Kho;?AR|=7DR3qOiEQK0X zDvQUbiTe~w#zV%zy)|U}B>2dn?GraPo=big-O>n0)o-|2K>7pzRMr+puo48-0dU#ZKMMF%7gA?s!Q!vCY zVsna>u^9K=|M_VBILk7#Xlc+P((lG<02ekEo2(Y4RTQ(5BFm|$PZg__yJzVR_#X9x zzL!rGcLR2gEenlhz8u1j&D@T~VRrXlf^5l%z9#M7c4&&Q=AmbLnivYInxvI&d1n9?acG*5HcQbe8vJ3=J)V?xB88dln}dcpv6uJ@mXnuECv4M zK!(xOPip379_ytBK7PkAF?LuOc&B$RRgZh#-HI$tJCyrEV}U`{B|AnQ$rB=$c|2|( z9?NgO!xL@#_7BX#nEYhr6^vW1X~(64^ygT-a1VfNfArHj83lQ<6$W z;Uay%9d2(SaZqTC#R;QrP#m)! zkC&9MUW_ud{K5>;A6Q`W+O|Uz$(QV4;=C zSso`}`~bo^{ceTcm%_1i!et|D-+KS;5+4ir8sFbJk0iTrgl(6V)ybon9ExO8Q42gy zLZs8&K&Q*&r1P-Qvdz2TWrM2TedUZ5FE15t1zt4WX|LOUC@Xy^yj@=Jj-oa=MF&F< zhT;gLX@G`pddx;>)C>@-_3VYX_kX|pKP*_8xxhG$G=FUj)qN{ubx;h*mbYW+9X6+2 zk(AL1;fcVdnMHTYO;#cg)N;~8mqXilSvckc`rY=z<#fMB-&Wvv?Uv_*P%xwp^zvi| ztXL~a_bc*_=WSyOMcMQ^W*w6tNDu%qg+YOV`xE@Lq7M1eS^aMHvTnuhpltVeULG!; z%3^gkyg@GXrei9gL~&G;A}!$R@IUyAr9(CxFW|#Evv$h4fU-#qGGxqvC?GVJck*>x zJWtZ`?oH6Yo-f=6Rqju$vm-?3R}19`U)jR^jkHxOp7+9-8^E)gq`3M$W903BL5Za{ zW*dWQO4xy?i+LAnDw7nI!Ih#b96UQ#Qm@09aofomm1B1H=w~0j{p&wj?2#f{AtX21 z*&`RuXvbPPPW=?~fFk#(sB_E$?{r=@oj0W}!az2JUXc}nz}n?mF-*H)ohp%B7HpyK z$h#F-XIm-i5MAV_C`-9{Q=0ts+KS+9UOf>TLLa$BLp{WX(8DC_m1NSc*$|o*u$mNv zUy!U5;ByOu(!~bFK7OA6soAGz?|TUecJd?}L7x=b?EF)hMWLAK#50$*=A*C1?wQdw zW$E}^;4m$N9M(bJqR?K_qOPSsEPeB8^cQLFePh zeIVJ#>d`}i=HD?%@nOb~8?`R$JqfA~#QxTTz`)r64q=#ZXIaZhF0EbEv6pHUx<1JX>^FQ<3VIz_4gX5tzOL^Gk zsb|8PJzz8(0GC*7IE5ngUV@{!ER@)pzPrB)?6A#P*a`LbkWOeZg3lvKQQ`v|oTM55 z8yP_dgLk8{syuyARv*0UhGe{c9R89mL9NSHncJb9Z2=7)@ScI@2*Zyt!ppN#RUPUb za^3bx-9U7LtW0TiTF?+K?9-=s1Xo#lN+WnDc8{_`?F(mLDHGXtt=MVz*df_I?MuZi zNrw2VSJ?8Aq&VSQ&(BjOlazpH;X_`cAXnW0@fF;EY?n07(ras>Uuc`xLd8SgElEXi zf#Ab*pM}Ei@U`4(^2C;d-5)#T-7IHj=)K91BaiGPtC?|K7xaI2Z~1rI6^2a))Pdkx7H3!N`@svOp51oKb=OXE z!z%o9iDG~`=NuJfb}D;-VxkO41;B+$RAL1l?BW$SfWzcr$Q{_-n;1y&Dn_WW)g?*M z``X1wlM-e%w09!|N;lus3HM8A&VsKmg6< z3xm4oRAsVn-vr$*MJ`CRDWmc2s_zlTBmU z7qW*vL!HnOhz)HpM}GEcczy8E?%MBIAohn-DL*6!TzEg#Vg;HnD5jnwpHoq#;(Y-q z&0$bL<)>*E1u;CJu`cyl698Cg)YvSP8Y2v+0%0x+u zPH7^+A}>Xsz#N=Ds!4ZmmxIg)c1tyQ;q`j=RQ@s{8%r_#`EnE=gI1p5zZr`sr!Q6h zagt>%ddM7GPEyB{eO4yFkYeDI-aUo9D{I zTKSk!NL3wUN&@g&C%=Q<;)x_Sn03I2b)wR2L9tk;!Y|LOS9+38=fPP`R@g0$0TKsz zHzrrlK!&^A9}9c3GzQo&?Gi&N5dH(+93&}4x>2k-8yK-n(8{lx`V}!dj?)V5Go9`O zBO4S+yjFOB^A&XJAAB!)kZ_Xwk%sZJ$ zNOf58gH;wlm+fI{k8FgU*GjLWkR}BhTh&RC?grGmW=zh8JrXKrUXUyVrc)zqv0D_2 zf_6eCsb5{;2g^H*(-2(Ymkyk_xex_d6r^kO*(yxumCsHOjTWus=9-w3Zjw)tloYv7 zQ&rKzg77}>?m6|LS&)nZHaK2|-v$-jfP*AWsaNT+)U1Gm{hAjgoiht26${N!O$e_9 z_#Iilr3lSIUSn7%{m5<2ghS$cz6W`F z?IIqwW2Gt!MI{0ELo&FB1=aKl(&nRRB+JT}tfiPW6p5#zuxJ7WV9esoI&>qx z!Uj@nkHW_9DO2SrFwk9j$}mp9#10tmz2ko*$O4QjZ(l1VdF)aXF6?H5662sXshnbV zQltz@=)Dh0Pz$789`A0-KqPT1R1Gq-^I$F|Tt+p>dZ3Y~9q4Iqek03oulkVsoc|fl zaSe!I$tt+$AEBRI;g>Dx2x;@d<9y+LT9*jX<4PbPT_mfwIJeH2UN%T`z&UotuVfnI z-4{5w|FQT@8f1xF1O>8yc0(qs9bVmnSCBtQ4{a6!Ns0LiDgdW&up0yO2Us6g7LLmD zCcK#+G1qq?gBR|EIzcka$Up17Eo|Oa;bsOma3Hm5(H{nOSN7Q6Mw z3T+K37*CE`ec|^}OgTk%Qc;)ujef2CUTL=kY0)}3mlcO59h#)O0$YGSE^MYjb~O$5 z5J?LB@>h{Mxj}VDUMNb1vUK?5aRY%n;B=7vY~}BpiN)-AxjPery;i2CxhMivH#N-r#c-m6|bu^>F&>u}g5(Y25h!d#CUPNO#7 z>k`)x){d?aCN0(0#so>RcqJ>>_ zf_InCBexn(4=0{vLce=T#PSI2(Ce0;@WrPxH4lCD+G>8O&*C|~A#s7`Q8sud@bJzm zO|ef(M5DIYr;?u%aZ1Y`l@6lHfj!kzZ((h1?4#;8pL}=jNQeIc3 z^S2E`Y9(-P^Ff`=?6` z|EwMe_oe+kf#iH`qB@`~G$^XGlVZv!QjE%sNs6P==sB=wD~p`}_ijFJ$C{*vL4pc4 z-Kh8hoGiH0HVa8Fk~Pq#8o(PN9jdcXjE_s4E`@W3EDTx)f*frj{UN4m<``kL5G$G? zKi@1$R5~jFJ#66EH#Dq4GSmZpZ~T0%ZE2UwW)JKXYhlCMCGqW-5;$qdpZ3?9p~yk!|Dyklz+za+zx}#deIstn0OKHM*H;E%l!5Yic>9Q1@qSLrEDJi)6u-zPJDxLid|`!-Llgsf?J6oN zDb&bse)Xbci)S3?G}-Y_`p7trUR%Virqh&}o=M-Yc@vMbg_)kIZ(jQr$S7(*n)~-J z#CY@WS5JKBtDB4e<$~$aY~d2n;lvyJL&`*nlWSg`UmE#`uZFCp7r!>Y5YOPCwRHTe z^Y2FfVVI_hwdMH=QVza)3=LKzmI;iqW2+#8$CIDE?%D~O6P&4q*82p_5?hBn$E!g<@ z6T?2Tg&j6rI0ARr3LCpAW(P$|si=5iwx_uw3k(6&&(&}eh*6plUQciG1J=O~NFkY( z#C3h1mqIH*cp=rZIxtmrM*Tpd!v>fp=`JZ!$YFu$eszKG4h^WoMx=+gde*C}p*9PP zp`7K|A_7=J=h-I4X`u2Rw}pP&0-fFe>i=_c>^WPWu37=4iDL8=IfXp=`!rCP#}spN zXi(^QPybFF^>GW?psdoIAP>A6waKsMyqic}d8_%mo2ia>c1G^_)hX=;P*d0}-9P25 z$o#-3m$gmq-Qn$|N^@3JHM!8cn`^m%BUuK;+x3?FkYE4O59+~P!nX5N)zxVR&f@VW z{rb46Ox5H&Axp+%^D}bvZ}Q2NW)Ef;?(ge_zkLO`IC-!ADs^WRQg7VqdASKWw-2?k$U9YsyvrCc` zYTjfIDErPwJ`t>2#zZ(~yt4?FjT)iaxJg<(RaX}NdC*BwhtqUnoQ@qQSs?q^`AKM_ z_e&Pb^nykKD5TPWnA^yFIgJie76_ ze4%VrE(ej9UO_|fjerKugBSBujx8Bad`#?5u$^%B?=ZY4>|2hZf;4aa^tDRb0*{^_ zDvyyE7luc=6+Bi`%qohkq@u71*LrQnFLuMOk50fjz<|Y1M>OpGsLA)PQ0wXeD%5RtD$L<-~cL7KbK-4Ky4Y&`CG_ zWap}DFIu4C|3SrEvVN$@iVN?3p*3I-0JcyJX#H%aqHv2lu-)yI=8K!T$YRtao8L3t z=+`7QDEhcP;q#F*W~;g|46C-W%6q=Ks2caU=9239qh%@__D#mvRJ>q+aXb|N<51g9 zBsSzBE^I~Y^esV42zQlot6GOu2-qZzDCvsOWeB=~qY6s=WupdA*UDaOqZrub=2B7B(nseT8hvBx)>(y_%&{ah-z@Af1MdfKyyiF~!NTzq0yFhr&>xlnZP$!7>E>=U8 z6Yi5yF6ef6D~&(9BGyHK4h(j^;N3`?*ChSa6YlIKu}r@B1WAG77!w5Bp;{_&)=}tf ziH+DW2WzLYHN6nj$D#!g^sru%ee1t5l45=r_Gj!A^CLgKshTZeVnrA- z1s@Ju`qC&_nWs;QW0-jQ7^87@-oG|Yu-)XiYyq5|Z3_yjB!sW@N)s3I3PqTOL1Ig2 z2>c9)4)G9Io)LemS$z7AO}6DL&SKn=-pX#lS!#HbylYu-&i$Y5&=Gq0q9Ppxyumy`2t^n>gT;!fX z%*hIRLoG5RI%glC_M!JJYD%qSz_0oHw;IVR7xsg zsoA+OU%E|x6+GPifT~lrri60_imFXIZ(xtm%jlah1{P^}Xs|zv&bneB4MTXopMq86I%Jb~p?}7wf z2~b6{C+t43)@6{@CI!_PA*@lCe6X1ag;P*S(4bhMEe&dv?GIc6KT}meQF~i%^n*7< zjrYpn)j>#0mJnXyhfk~kLBaxV7j5#T%x@ecZ16Sxa(I94)Jjo24}Bb6ta4!xI$y|9 zS4V?{!clo8SiVa5YaS|Upez&FI52vI)M(o$^^$t-Szydv>QNc&vfLc@0r``Mf2ww~ ztXu{6e)|cD9uMNjgG2A}6cbC46;u>PM?n-TA`U89&5Tc@Z{>&l!aDM`AM(LbK573` z9^z~7x}ggFy27cpkYJM!aNWPm9kDi z=D%UF5|QQnc2eWQJD`hJ=Bt5Xj#K0q6@|1J=CpA(y>{XmVhq44>DyC~9~F6{P{ccp zMs+UO4OQ}c$>*LxZH{XetSN{PSc0~y_e=*8*62BnsuQ43k~_6czB*WkJ8)!vYNeZ` zHC}CACY9O-uU3ACEZek?Z&c*~jZvy9!{?ep*95e?h2BN-Ia8@QcU)_Fs9B~Vd^uAIYN=cu&aIoNkGW5 zNN|;}gB+743tC)t)ljfvqzk8_>doq4B)0-bW|L1@cx5n>hQa-x`SrW$Reheh5W4|C z2%Qk?anV7k;XLrFq|qPBC(9_aOnV#9>u1x*%4(!{0J|>o{0_rvjSZZJM9fo9vV+sg z>u2J=VS!WAn^$wl_UCL1e$onJ2Pvk8B9&B>xxU2=3*>++3&*nSYz<%vY!VRIVHrO< z(wJ*%f|U!D8cHHs$su_M?87n3)CxA}ki3c@%1}}1pnSi&N>nI9^42z=9x`AC5AsT< zmWP#xZC0ACU6Zs=Fbp^ip3~8T*5G$x|JpvZI#+Gl@Vw`|%}m`u-U7rexim7ef)H0V zzZOD$SStmjc-&(3W$p%V9i|mmdc<)$M5)S5$!_%t;vWEf4?stNxh2;!^KX$3kG0o=w1@GR;0Nxe$ku`*}lMify+jN zCD%biVii++@_-E(9J{xr>O#!(&K!3;{M2kyP0_?MvP*iD*XDEon~CJuJ7>RL6S?fg zbFZv>Ifi%cm5uIt)hF{mJ~g{~)LOu-KJwwrxYYoobYS1n7-H)>?je=^`q!~&1+>I~ zAj2Hcuc!M27bW1Gnh6Kvg4fSp{PMhP4O$bBy@GHmP-tLX)K=j>)%Hm&g7Cwd!r=hN zY$}HJb`C#eA2>Va{vy(r_{$DsmZ^IpQJG7}^UUetSms6uICy=6I8GY95cI%vr``xT z=_EM@NB%aJAGaezI^`bwANLU7Q`cR&4@e6Iw*4M<8YLQmvA$K+4U#6?RVm80kXm}} ztWL!x$2hK z{o&`-cgPp)I)+^LyFqSQ#ai1a<`PAksi;CaN4Z?3R~APc^gSQgMQS*@|xLUr_dkL=!I^+k2&Xnx+?{{ldHL8&JYdxo2tpgozh}HlN zbG8QQ4J+vca>PA#8boo*#YVp@O*hvFX>n7Qpv327usKm=e~lc4FUyl(pQoI$_syz^ zB>iFw$};$CTS(k^QeZVbsT8w;BI|+KkJkZ7+dJ}VdI>3vtndQ@&^bpHTP0U!-X|5o zUA#+@UD8{;df7?8gA%+lMjKT2pRB#l!nx;JEo821hCVtKuKCC6fW|0o8Vxm**!zq` zK9BW;vot9CI!ZY|d+1|>Z+l~m#`&G0yO#4UmgChk>kg0=?A9k2j?ZqfGJq)*vyLK( zR1`{{f3E5S!gdreUm9BMg^jHRQr%KsnU9$WE?<~IE`^Sf<%3ye0}fanMgs>!j(NSj z@x?z_z_I&1PZ_!9!l~co;|ACC-lv#eirk^1aM{Q9c%;<8(0mTv0Y0~>SfELIPo6Js z{SO@IuNRVuViZ@g64Q&Y)O&jGnlAb9P^}CtbB>M!|MphOc z7i^e<-iE1}%zR+(FvDLV!!Lt?8aOI=#s#nOYJ$oG7&;hoN~kH0DuB5(zRx`DL4o2a z*xje9*7|3NFL@i}>%9$3CiiZHt`%$|C)`Q=~z**=U!o zPq34D&I3<8H+=2ncG<3J$HP9%qw>jC>ccWoLAdi~+rgP|6ci2)=d|0{fkJfu0Oxg! zNm;z+pWh`_<6+A>_?vrzVm_nD5h`kdaDzt5afc0+b`Mnd?C2?j+KX;iJAT`#TV-s3En)q9%XVYrDM zMoO#xmFz6Sca*{P~8r&o9#ne{)F;IM4=>*T|=dj7VsER9~33mtB8oCZ~| zbab~@BLelX;Of{z_Cf9Sc_;55s}9DbJ^mGwp$?5{mL`@N^QFtt?UiFw?8LhQ4*s;) zwhiscf;rcXn?Zl`}EplWa&#?wD}r7b+>r)2=9oUNaKnfhrUg!qROYzKy zR71Wv74(N-&$QM50W5NByi8oJ?O})A?WdjD&&d!^i~v4EjOM&k=cBF{*ftV~QJ@I4Y*k4Tk_~%exTcO{tiJh<`0UR3< zLn`Sw*hC-pYLK<^cX4&uT%ghjU+IymDfNMx-$ssJRpOHqnyN`vWszJc1@3oirMpA$ zVvMkxdl$qabXR0~oSnX9@>YH?@JjW&EetZVL^a5ecA-yT5;>!F=SiPudT74TNMD?o z!9DiMnO8<;mK?Dzj{34s-1itS>|Ydgo$Xg!BDPc0{+>W`TsW&%=AK(b)D2lZN!gaOAM+58Wh8RtV`(BLC@Y2Vlku@f2>9`(f*PYZB<*-Yk=gz==v zYJM^(2AIV*VD!9_AIq48aWV@=D z7dqeP-Bf^>AMyBe)UQ8?`osVK-f!prS+ay;7E&a7>`G(o{7w6` zUlzS$!ARq~i?T`icyh)HBS$FaFhvf4g6>-_pku4o-jS{4-jVI_XyC8@W=qHd-a0Yb zduS&hcgc|rtP*#TOOjgfBDx^$A5RRE(&&d0Jg?VYl2it7<5odu zLM5F)C7;v++Y>w|UnQ=Kcm%6LEuGFm3jk#$lY1ozyc3%IDKW~CIG)4C(i7%eq+SdC|a zn4{=+1d-0X=ZoupHMaX@S2c&k&JGw^2vcObGP)wz)KdjyJ+nq)iH7_6IyA?_9ef1O zv9e&}lIq_UPqr*Ni~cxqB}sE(kG8^UH7ueS=u^m}qAENR!t?wSUhNaCg|NGcW&6RL zVvhFvp$~^0pV$$yKM))G76ze~&EIq-@YH^? z1)f^SF(^=MpwX+FYkr3c4;-f~T+jK8(?JismtKqSY}TGIC^o7WP%y|{=;6C8=$djt z$O@=VL7){J?H|}f+fe6!wr|S1=e-%fl^_55MM*BF!lTUe6WTBZ1%kcm{`&@5I`1;~BK8`ndc!XZmd`o)|{K4D*YU}*k@iCjE7O}&v!amPV2zx=0Yos6I zaQ;?@KKS7bJi=$#eefcE(l*&N7GB}W6z|-Zpp?Z|hg3oa#R+mq3f)EK zm!(0_pq)l%`_u*|Dj&I3s+%EJX?~v(hJ+pZE?AovhmS_`b@*_3ZRGQ=f0h4Zj3&17 zH!7EqQ?i@Dnhmi<`9W@}d_B1!Tsu*(?c!baT|4n0H>7uVl%IF4_`HmX;ggA#=2lWolwKv1X!hn$(^wX(uU}s;x zKkJ7FA}q*Rmn%;vJ6#w#jaJAxOfd&2QVqFJ{#lYnx68|C>$Q2l8)s*TyA^SqIB^%f zM0B2W(#rrf;LD-yc^Ut1$ZF3A;Mb&cFan2HOMF`SC`e{dHOLC!nNGTxld4+IYm@7> zH^c9RRMQx*J5Q1nwa~+cgbv8_uyRf!x6d;lIE2zemyUEDb0UxiVG3W43|H)c)c!_g zxh*G*%O;Y*gBjGOxRzqpP$Zs;LN0~@2|Y_0_BF5EmylMi-|b-N!O&4IsFfU0miB|qmfD3o*6fgkbjbP?$O{9Zx?^Np7{=Y2q)?!1P$Y_1GP*tm zYRPoakZ79RL2slM2+hr|RkM;5*g>#j0vI!Bd@Y^?Q6}j3#*|b%c_c@}6C2SYx(xl{ zOFayVRAsX7UX2b<05O@OlAjweH%T$K0h+(@0eE*amq&UinyyOFUrhEzY7jk$N~VV* z>wG>WAo?aW$V|U1{L+trzyvq#AT?RZ-Q=^>3&&0mT@K9+)qGu^Fg+B1hrODUB20@N z3M`=U1~jQIOA;dPUb(=w`j-v*TsWX(r!E*9Q69OK_-vnKYA4C2n`vF9X9=gx=iDpK z^?;sjWE{3&&mPLDcg%{2jCsl8_55J#uYOB5vfEX;aK>_v)zVf#F?kfpG4HDCXt*ffQi;_4X5gLQ-^PrEERg3K4=SKTAJLl$vVCx;B1sQ)V zoA4L1lAR%OVdo~_3P&3$CK;_>)Bz}{LB)Egp6Q^W+XMnIx%4iX&U|o?oQ6~Bp{c4E zVLB*e%{TpT)C|S|7=~*&2A&_SpI0{YMB4Inxa_7V&1%9EC?<{~pPIQ_FwbLFUd^3~ zmST);g=Kmin&r`&%BPIW%4EFo>j|niY+ImR787>bosG0*f78fmfUZO2MK&vL8fg=Q zPlK$_Gc)uGr+`x@X^=Gs=K>dE4y-rOcLXiS z9;b}%;&n>9cylX)SCdZZYH}&;KG~q`Rv7sYd3doP{1&&HTM*s{s~8?7YI2li^fLDh z?qNYMxiWJ#Sxw3yh`d*Ae&bK!jd!Ul{Fcl-O?pB3_3$tiUY!oVCn3rPb&v1u`JDgb zcPxNPr~1B6j(=?cb=?Y37bvEQB6=z+nPiBw!v1)1+FEY%iyzjDT0JtvhbN^={$#$8 zq_}P32U#mB!$-V&Ct%Mjm|Zl89JDT~Yj z{&`4y&fe~S^`9?D+U4fswG(%EoqOehq&2c5K(B3srX9#=5h&r@9(gHJuf>BN(F4gP zpMgh8BOMWE7z2Pk2{{KM%*P{stg`L1b=lREB}I{KkVB}}95dBI)ZaU7mISn`^n{)NyGd5P}zx?e3YLYp1T{RZFWid74rX?WmX4f@lXm z*UAS%B2o((icX(8I+c4nyb-#%%_D&halk(QCZ9abNi{w{Qs;Tpknz}v4DSW2m5-W3 zb9^RQ4An#C*m9EU!dvitR;H(rVzyFb3&vnzA)>L`b8YAjb*c(AB|1zLLp?KEJz+!D zLz;pgK;pZ6b^$2A)Tk49(CLYL^afeBM$e?Gmhvh^`vVWq2JcEy8~9Tr#m0~`=M2CM zIq(QivI5Mw-|qg)+tdJHJRcq>kKEV|;IcJ}WTO?l;wa`*iY%j|%stmx?<9h6vsqa* zXH>xP>^Xcke4ahVXxx!o|G`>c3m|@e;)PQr#f2>h@O=+TQ07w%XozKDru+7HurBX} zZ#{pzyjQYb`p_TuOGfD5y*gvN9IGHMgD?D$*Qj0f?W4awrG-vjAGi+g?*`>%pnGUg zm8t7TQ<(bM(?66SF>r1)Cf+dN77J7i|5o+t*zVaD2SjkEE{W`77i)3hT+t~jgw#?DFb&mU>fodZ$%PgO z^TpK>9dt=#pP-w+Vit6;)k%6tr3l+jE^|S%Iw_(TYIx0!)<(LQl+R9)W&y8ZS-7bL z7<=4+uhn;vtilfjXwyS;>9e9#&3zad>a?a}1GjF)tnKc|kqC|5jnGJ&=N}hb^9q#m z$xa6k7k9%z(ql8_(2HS$Y4S_LIUW`yeXKkEXR>T4nE)3qUCFV+%X*4QqR46xdkpId zH?dA(*%PulV7tCln=phj<)gIe9eO6%0buAcM&qpgrRj%1Pl&hGZhVsM(1lG2 z5I79l`8*byG|~eSd^_cvnO3@fpbZv8BItY%9TH<1po#&z&WvI4yfK0aJAjyN!e{`( zc1-&K;pld~#kJTDw*a>ove<>U!I@TINTQh46j?<@eI`m$)Bw|Vy{Or@L=i143$Nj9 zhqzf9jhUU{T6^2i(9vUuG2rNcPP)%N28Q6HqyFm;uV1#UWO>qXxbT`{r(Ol?xsjw0 z*Ohb*@+rqMv7mW&7<7?xL8fn|M|vpI=Hc}*Lyp6cv3Lsglqe1pBW(;ABLi*vlUaV;&-|F2b%Xtndx&qS z>#o`B-u>N0fn`Z4|MvGzl24yAcN9uw2l*TuC}uqckO3B{vji)jwomGYlKUQl85jh& z?z%d`+bBf`H5XgXpq~E!EQj4#$1jpTQ-)j6@uOW^(@B{NI~)yGNH{<-)fB0qqPj!+ z1ZR|bWmz}|%6d3)ob%8tmOk}@IRE7n#3;Q2bSqcFHhZo1(rfQ?whEWac0h&72IUU* z`l;3YLeYw;kCdJKDj$4(l3LG65TDXMa?^*m_~?NWAysn@HmXIWg7?TR$>XTsVUpr~ zG58ip%O3R`Ij~q7KZo|MaMOrwAS>fXT@WAorY#?oUC0e#Jzdywf#mQY>$8bsQYf;H zit3qe^fM^#gJ>(7nHb^eupXkTgz}KCh=S0*3G*sN*MjQhSsILaAP4-|^rj65O(TYi z0|R7)eSeKDmCx4%A9fgtcE~&F7LpU1>DeURC57sNs##+L0h`%y2oW~VjK!hkwW8@E zmUSmXnYES_K4)IZXI6Hhl4AB!WH(5Vd!nJy5OgcDA>xK|@y(({WeEp_hvqrc3^`<7?Bc`rf)ns!8Nxnn zj!{*w+iLiN&cdK2{;@3X&DSnU7Eo=9^2j?tH?S+{(qt?kOz=EBshmm_WQYy&!;@P1 z+q^1*w|OlP*2`PTF_Hm2!=*k8gy%Ufm6PY zyJqC?Nm78gIkd0qu%hISU;z+iT<|XlO`m*N&;vnwU6MzU99!a@v9Jt@5u1~&p=|T~ zXdT*za@$`|{pbVB;`7?JLleoDUz=p{vT=irVIRfZrO0he7RNHc6b8B`ONG}I^HQK? zr~>?t>$2!MxpaLntXJv07HBSvo>M2sE}~4XE`^g45JN6=qvza^0c$~>szz#N9qf0@ zGCkZkr%a6Z=hD|_qx+;W&gIleloIWCC{ zJ|iuXpL-<@IGJlA5_tHvuJ9i%M6h1CM$hB5S_ADbTtoR_(|-KX7xOK6IrrnmGs$gs zi<#>>ab&gCVzy`;vtS%~1YPpr+3U4w^j&hC#Pbdc3q{dE4-{Rp+pwZlL>w2UDGNo& z%9ATKhV=?^lm%S9FVE%8iu@0oFR#*!Cnc?M?V zL<6%w5XtT;B61{#>5bYRa$ML$^0XLKJ}+9O+&1g#jOO5MAHz)Z6G-P<4!?`}$@?zyw+L`~`GQR)&_oDvt{s;g4`zQ&` zxDkpP%3OI|GH-98NZW|MHSrcL>{$xM_S8Ti=L zTlaFE+Ub4}J0RG4KVKa?7X}aqANZrEuW!i{i~G*63eF+OQ??%*v?(_yl*OU}~HVuK1PWMaddMVz8x1azis7ecTo$fom4)`un z&O7uUJKoes;`JEe=Rs@Tp&PnS5XZr+11hvt5xw9GFI~6fq2wUQRUh%!rFa*si@YC7 zpeR@#?>@{tgri9M%DAw8^(@c~dEGve)}Btj`#sD4s?7Vrd!&}#{>p_hX0+P9pP?Aw zfNntPB&_5_-D7C>NK)X&11oddQQvS z;)XslaFcCjj8I&WqyXBXfOc*i2M4K|jI?H^aWR&?_~oO}=7ifLU3*A_EW_v1z+5_a zDm=X4#XjIi%$VFP(zPj8fXr7eH{SiS0GNzwIMtd%UUxhy=_{ONj)*Od0p^AP>5wq9 zI;mI@_GjCs>F3Hu*=eLkMOcFZ%et2E7RhR-HG7yO&kc%>u!96MnI`oCy((3OF+Z?j zW(nEaX;{?ysiIP(E1(_cIQZS@N;dPbJYEUVPf@4t0{tajn3ak169vzcOm9dRQSRVk{ASWk@nJ?||6LZ9{0 z>;U9L%;u?b0?GP!r;3A|SRDW(#FaUakBWHa3YC8MKcCkCd7HAo&W2m>RG z&QDe#p(aX;{cZIq*le|2cn@Vqy~12j11)UnKFNVabDB_ntr;4eZiE;VD`v%UusLO= z*NRyUpb8ty)Vn9h$|it&a-LKM8x-aPBk;D@aJ~p@QuGN{%sNTm2mt3Aof<3!tA>b9 z4d+G(raMY!Er(}2X|co5P3QWBfhs#jr8?fXDyINc8SqIOkwT z?>@yl2U6(Ruuw61pQx2S>2*?_quD3wC5>UjpWk6lfTOS&=J3OQOkMh4ZMHrN8~ftI zTO~UlO`JH!bCv8M_mnIJWJ`B?WN0cp?r1J(uE-8>P2KX%kcrx$>Y3arJvVu*aN;OD z9-D??PdK{=&v*WNkZqvOWnEZ1LAhMoOc|G|>Q(@?0!B-A`ld}tRF2i;e)W+%IL%)@ zFh;Lf`6Mssk0|PG%N@QZHp5PF#J2CuGw*xnVL&83RA(+)+6%|T-Y)4uKNRajuETkKf`Jk$IB+)p97NY+8)El`lTo+z&A0yTc{d`+ zyn!tx5OU2I-XuCy_-v(-XvNvH_{Ku*4al@S@2^5@j3KUsFaifo0D+#FA0 zt%^tbDdqu1?omY5Z(<6Q{}U>;dODu5zaBDBLpc^(&%LbzN3Xxv*|320Rp4l{BC}#Dr>5)MbQi8 zlQAS3ikYzd5!>WZpjr1BrzK=xV4vVJ$Ux9TbrtR~uZH=y!=`~5Foupm&Ax$7_Fp&Qv?73PSgm=zRRN=3y%KanvEqu(W*6|?4# z`B&cjO?~zk-^t%-2Zv*Y2iM&&{P^aneYQ+6Psa9L7th2FJ4~;jAsAS}piWnpFUGii znz)HMKHZ>76K|iAry650KO4jYZi$2FbJA`0LEMr3AIwCHoiM%=9wMP-Gb(cEKNMLYaKJrQ+d)5PhadM(OjWdWrS7IH2g|9McG;-YfNc)j)( z93`tOUR>}Z9%aE%78DZid#Qn5GJfu8!OL+R4dBVGy8=hEQGfOe0h< z8x#dlL7gv72(JgXHckA6k1msIq`&m(hE~Z6Ka<~T;Oq}fQalKJ2pX!8jn4_izMX81 z5k$*6-B)^<->eZ*yLij)pgPrlHFC=x1u0dRu?aR(E>C74=U!8 z^)8%I+-ZgXEffPnjhm^clE^|jOyX^vYX&rRZgn1ZJ z$kv>J#9)(uk|H7e3`Y-W(v^j`hopmU8wx~&RA>P(<~6F)Ccpz&43ngAKGa5nuR)g5 z35c>Dg7>p;7uc41KUdo$JLTU!q|?Nu3R9gPis9XE*&@MJeog?cb1><-bRq}qdZDWK zlB9qx632?m!VL)B!;dY7=2s>E*+&{&cul%#Wjb0Y<^n~U284<*i)5zuKo_cwz6)(cai9{Q z+poq@AXFv*8~3qq8fWM_Upm4wpUMgC6IA#?EjR+J9rPTpaQ4kig!;=|8toAd1QAR{ zEs$?ELfHYE>6=b0~`jyC&v8p^$n=e@Ar(}J;IHj}v|!G$qaWQDN|ib==bM!`waq)s;z#t24HR=@OMjAku4+V)3@F!TX@3m9`uhrkoVrJS1+4BdU z`tFniw&M0*lk7I^um-jMucMemiX>1`do}UAUV^JlzE^dCZcXr-VC*5u7h@Oz=&OJt z8q-lDg@VJ+v&4+U zTj6y|UNu9nt)<~)zW)ja$F3FUfW^Y*lsqS}MzDm=GrbgNIm!(m>pX24WbE*qZcuj7 z2YgGQeFn>^7cqKhC(Wj}P1I`>#EXKGLKn`dgPmu+44LK@ld<|%vi@L4`ek$Qc~(w< z>wW=qLInofaE!~=+1UxOESY&*dPZF?DW|r8;`3S%^-bqsc}6vTS)1=)De44DL^uvx z`qCEq7&#=?YYV-*xg9iKOPC#{5o;AD3jJ1E~X zw3a=)Th@icbauLBKb38g&5I7I6m`(i)Rs_Pv#6Ed6_Lj|u0E;ORRuPSp!%-Qv&tt` z)hop~%)D+z9;X&4AdU2O_5Q#D4w7}=mTm~Gmap{IEeuK*W5Yl(w@c6{+aFjdiszxA zc^v1u8n)h=z5dt^u!lYx`p7MQCgL-V-VtWt3|}qQ3FvwfZ-dXU1J&cZ?W60rm($;T z-enxOc-~Z1(?>y5twG*69SB?}9^^&~*NK5ZPmsw?QtXW|ACJ{?2_HK8{3G2aQuMU}#^+YRsG=CquP>*f(&#gs zJMwM?=F>6g|Il|OC@EbFf@J*7@OvTYOwsJj+5K+jp941-`FKMm|F$4cs5AQY63j8- z2nWeJF<#mixFob*w#pYqF`CLrlc%(6FyD``>v&fDcJPeYd(F-e4PI)V_S)XNIewn^ zzTbr~3xY)Mk(s2}g`>8|tuRqdG2jI6p`xG}fQO=;=Se-V#ustZIn5qBfatf5B$GOF zR@x!!18R;KVZJa?+3$vTViD=d*CDfr`Z-CEJxW!kyH|MZ@;?Y*f`~5)fG$8b^3%aa zDHfZXIunl(tOYqndS;~2InL00934=MTn2^J>NIY}&+i?4-ixiwDW!UOnaGqR+OLN4 z%r~$cXKdFQ{N6eU3dg_VPP&s7C|-!bO$WBJ>UR1j zDsg|@rqJ!0y;-x(s~EIyjQnOJAdZ!vi6KbHyY6m&P>7vVg zQ9G$|Ur7JFQo3llqij*vJA6d~m*zF#eD%(sRmyX&-ty@J>}N9-qmrm|Qw_?%Whkv@mVb zKEHTgwW;z?uZioa0;#`!Z13Psqf0c5l zN8g-8h%=&3b#=xn2#DkTCqXYANk?$}hyH7MaiOcgs~YiYI&8c5q$RXEW6z#Bdj4OU z0-m6xa@o+a9r`I`Ni@ncdAFD~6ZZ3PTZANcOL?21doDM?z*KmYs@p)pz$9y37LL2} zbz&@_0ph9Xna7w7b3Rv9kyvG}X2jEM?_(e7^xGSEr03Z!dx_)RKl+hn&n5Q#;003e z!h5b8R(q~X6w^$Rb5v9{G?!GLJT)+kT-Vl!u*DxHVCdj>^Za2>=09`PzgIm)(>7O+2! z92DiK49@WEqL+G{b)-#rq!6<*D-It#8-T1JMtyK9@a2(23|+Vc$4)1+$C!fBli1L?tx6>7G)mUb zKqbOkJgCD+QY1})g5-Sxt^A_d35xD;tVr(@)YA>5O7qBVhvuqhi((fvm}45O0$|zA z{X*OZj8;8#qt^*BzP=f_NS3N<<>Ng`+;-n;{|)d<=BXY*`+UAKhu-UZ6>Q=~OBd>8y*HNTeDWkWo(+$2ub8@@z&kLeYZ&u)U> z2#7YCC3>zdOH&q}sKnnL2z%I=His8LC){X%g=`L7{=qy8;>so+my>oEHhPP#jNV;} zxlNHTjp69@(9c!XCUr7hJTD>qkvv^i*q}&}W=+kQd?{3i8e`~V<zw}rOVYG5!0)Y*LYod%L$Nme;&d!`W z(}KKTelcqsDRE)sov^ZNH55}xk-bz@5jTli`P!;iudA<+TKab})Z$kcz1|%8;U~Y- zYrA+G)w!H2WDze(QTFPG`yzjL>)Ua!wtjERn;(Ai$BR&ZWRPDWAAa(u-_1`7O%U&t zqPlBuNJ~g8Q|P_GJx>ENb+EfM%+k4A677YE@k)ArTaH#dtg+2BRD_pIin0SiBB5iFOy&)8Z z%}lFig$IiFrqRZ*N=No-!}{RuKIAA;><%zW7wjYDLw?VH*b=39IavI1r6}JYlsg1%qun+$csf{sV@&bE3LMn5zxM5! zU|CA8{HFP@WchfIogBQINv0UEB5SCqo8JJwE%`QX+?*5vXuC~0M(}8tVxtU;f}4Zu zx$SZTGis6E!Job5HLEEY;`z}!#cm2*mbc=Q_k%ukvtXm(-fur4(a)Ki0WF+^w&n2@ z6HAd5=AO#xeupvcNFT9}-^n@m3Z8du%2lt`N6!B! zhh8*sX{hc1_fX``$SjGG00U3yz)Osmpzu2|vn4663kpTM1=#PS!|u#@vS<1zQZ6}! ziKoOyMmqQQp0|&Qp?r?eEpB?-#;UJL0^Lr3Sd4TfuUpa1Z}VvkLm8I8ZSpz9E0ycD zW@dYQy~VRmo(yFaIltcIGe*lKdo3r#+pYEupHekI+Gx<`WyXBcnKakGN37tENHe=A-u%Ga;aZXv8j~Cwh z=WW+)Jr6eK!-Z1?c6<_~7lG42msu8NGw_(d7~|!5}K+ z=(Wg>ovNx-Z)UK^?wn@ioaV{kC^kW*c zc|7fo0O;|}WB4NP_4FdKWueL7uWcc5;|Wj|4hkfsQVbB%ugBK2v>>Rs1`g(EVV|H* ze$wxtWb=zPlbSv9phz~)zfe>QEX^CG2KOZbJQ}UN_Yhy%3@}QE?7o^V z8y5S;Z}QeW@7+zMK1*g6tGnn=A+BB`zpk#OyLg7kV)YmybTGgiL+jut#^yCUw65NJ z=U;3Wk|!Hip1CK9#Ifp+u>pqlyV!@x!n>-g-j;==Y5JKz zlEm>q-ZXe2$)Xs@gr`zbr9QooQd#!Siq}7^jy%ELGb@K#B1l!mP+$6V1S}Dth)oCm z1zpNDUrSZBkbJLPvRS%QdUw>leaF6_)(eX1$ohkF`jP!leu#*q>u&$1&XaXcvB0F~ z=dvYavkMmlRa@b=lwyE3vw(`i#xuS4|FQQia82b`{$BBhJ=4zV5@NN_p*J91 zM{o$ZgqBY#4R4p{hu2G%`(F{Sm^A3LeA3O?OPSs+5$t7Qi>zG48)f>M4bsdiw{TrmPo1I0iM+6gLB@AB%R zv0x;dcN@9@w<|j74q$zTNDKB`$P!_(z5X(`JBi^`dvE6LfVz*R083~*71;(w z7JqM(M^X16bbErV_o-5KNiU1<({;f7c~5o5L+@>^_RXJpeP#yTF?$UNn2gZUh;Di> z_+AS=8egsA=X>ZR6)r1PmE^AP^|{IPg#gHxf~x#3Y1y0s?&qHOLVM_CvQ{WZTtcsk6$)Z+(iQPT@iqA!4#uR{OnL^Y7U5`AMP%EP zrg&r}iKw4`+wYtIT&V`}At%Cnpw_lu(k3sBxWhle)B;RSDw-wb3l=B7p26*fUT~+q zwkiL-_>(vA(=#h?`owMV*%~z2;(K2-N79aYZ~b!>zM>H{T>6SV%BsNZc>pcQpb}O` zmkW;h9+zF@j0+$hFAN)CJiccJ^DWT?y|9e?vB~GhrOpr9k!=nJ_67vz;P| zRAeHj_{&$s9|!J%pO{UZ&$$p#OLuv;%ZkHu+zV#4%j)@E!b`&P`E6Ky>f7R{a~T(? zuz|>9-7_8#IlH{K#>D`UKWol@NLEcE8Ai``E5*c9B#w$ix`CswXNPHh>LtKXiP;<& zWQ(0r=w0rDJCHP0mOcC#W*-BqXKHp9+h1U%-5u9qUjFJichi0pW*Rv_KRocBV=3+( z7Y&wM7KG{h4y|LLX&DsSt_m2Jx9ss+vO&h;rkPpYvIX#kG#K?ZyaJCW9vdaB&j zU~ZJY^^G%R?Pvuz4s3U38reok6tk6rm5D45TIIY+)up`7ZKBTzkIOPuXOzwAnt%dY ze|{OSO%@llKB7kUe~2-Un4HoXznWk`hiYfZ|By#c<_C`w0(D@QE2LA6$gf&QF>5HY z5*s{LKtKmuO`uPIk-a$QCnq+V}KOO=H%V4S2}$<=i1v?Dnq?jEBocc&Mis z;C?@j5kuTaA?rx4Yy(j7wz@=7*eyL?UdV~%U2<7;BeYR*BUDG@sXUKDAaT;{R%Syq zXHf&cow+r2)hmmZd!|CRG!~ebyWo%xXw3;V#CR6Ym5Ms1>vJOYTsUa|6psV?|{NG%OSlw!;O?WZwE%e&yZwUoogACy}azq_a~^4lG9w z8Of0%ih-2ATrhE&xDeDims~VhSBwS3*fCJ+lRvZAb0yS+{$=r>AW#^RKoa?yJo$0e zM~nZ21G7XINxM8FWX0=;xQF$1$Vgq<%0HzT0FJknAyA{-B|ksR;AXpcvkejT7t2?R~PgYIc&c)l+oAA za$t;pzrD_G3;4m-v80xF~2=pnYxK&<*4ybZ>yz5E2h}(kP3{ zfcUz@O3qfxpZ$}OY{Bv5MmrybYf33o5(_iz4=I9F(W6;MydVE=RAcw#>*^*o_EE0q z7tUHbE0g=jJb9gGs%QZC!;v1Kf-V#{&K=WxWyup@`*J5(>8{<_1MN^zC}}S@Mv0*vRRXW4mu)Oad0Qj$!wFOc1!% zfSKvNuXT3knKuok1-rH@nykpLW(Q=xY%iQ^5*V#9G9q*R6V4ma-{z6*VJ z2b8D9(eC(d=-MywDDk);yC55KYLH^sNJk%pp5mF}N5R9143L%uG~-Bz-tCN1X?PaB zwi@nS#P1%i4M~*rgJBM2)E-r;g3!@*m9?+ItpZIA+(YTSb@ZLEE-9Ea?TUkJD^t?e8g0~+c13@zYi(zvqVLtM$+z~i@>Tf$lY2V5H(MOb6)l9 zn4SfM8aO9h>*XS|eKdprhh|319wAgD4a(9~v9BH!X&R)M0airzgrq}Nz<~U$`?0xL zu#_y)K;4>))@L20Q{bQfiEC2?Ug&n#TnRcT!cnndFJZ%lQ8s83`l`S8p4cGgWPks1 zBU$gjaxTY6&LvaKHi{%*h&q}6hlf^?L)yt9F17~7^)FH`3xrM?J>ls2O;_Ebo{aN*21~D>cNyjX5 z$$`bl10ylgMKPaHq!YymGR>xNnj+4Dxq|*b^vq~dKq1wHHAQGnPpJtzsEpy@SsaO` zek;Fg_9{UviH4dn4c>_KZ&86+;^R|#x$8jbXbOE3)bOKW2K?wHDIEBPaPk(8-=NGtz=>n^uy4xSzEYAidb zj17B0H**r4yF&Lu-E0h}l0>^`QbKg}c_C&8qEpGlZn8!w*~Ca1Ws;GgV}-W~4=#Cp zYMLbBuzqJVX%A>Fp;@6-?FWlIC1jig&d7(ZKehxC&)#MKGwdKS`G;4lOr1Us3kfrB zoI3J}*I`~Ywm=LcrJQl#8Ea~ba<(antI)v55DR)OcoK`%XCUQ597^woZRm;HaU z0tXd$?rT%Oh~0WHVd|E9mH?CoRytwTA@*!}M(W#tJuJ6m#^Y}p`5oqwa<9bs&!(Fn zHhyBqbzc!PTcumwhMkld zmNz4p*lx#LRlobe%g!@+JRB&YlzT$=znaPEch%%_Q=kJX3XSnu+L3Ex;WCfBWn7NU zi-=KtCKC@;{#5p|g90@LVxUvD!MS0|VR64}L(uY9mIaPKCM@hJIW9Ob`Vw}an9;ue zexP9oeB)cU@<}GU;I#wml3Jtvaw)|WQ>1{3yc~XW7DfcLifvO${F>E=p+96Q5Y3hG zD`*rl1M)=Xx;jsu!nDg)&$}=`*&nOfkA@FAZByJ(Y?UpX%!b2Q$&(3Jf9*YJ8i4$Q*rsN}m8T`u(B&Q3 zFEY4C`Np5(NSG0=a@KjMW@9CI& z=_9`$`m?u@Z=(oUH|nLe^m?B-x3m0h&fDA{L1){oq=WyFrs4HV3oeU0WqR(OdtvK+ za4X&}@AB>l8Faety$9q}al~EEtMIAXGi^Q)XnK}<4?1;|OuwXnEs$ap`}zQIjQ55g zd(9l|EyL#X#LgJ6nZdK?2S+||HfXQfFID}GEOTK0?G7V-w25M3DH4O7-nEKm)!-a` z_(y}`Ym7=CaIPecv{#T@xOpGu`t!(1#8Je;e1Mm49@?ZW(aXncCa+uY9NDT+*wEJoE+ zW5t>y)31HuXY=xB%5OgCG-Z{1sbq~A?#LbLN+=M~6p^)_MFbN*hMYDA)rgnBk`SQj zC#y(y*Z?4kDJ#J(f}CA#&@>F4S16oE^AC#7vcQu_qP~ zL9=aa-4NxKq|huCps((<*F*7;(@Mdu@vUoR1DUZvfbG4dA8bjO{Me8Gi%Wkct4GsF z4s56Bj5JaL#enU-k&47E%q@4^#Gu%?FuutA4M7yC>)WciF z8FbpC+y?w``^1Tyb-`ODo16!oRt2Prb}At{L)LSX{qGYTpUbUFgbNRG>mkE(p_bW0Z~4-~Mu;92`7K-` z=%5!isXO7*Zb2po2Ws8ULFvdL?ybOj*#@!LWBSysrh##XWH zo~KYZ9GOSfC_5&q;P&r-ZkprBCIaBVGRjOwWSR;qGBhZRR`M!@2LmdEZT}k)B+Ek5 z&n&O~Pq=@_?~ZBhA{$6JaFE$d31%M|geD*;rZAw><8d_vdoGl_TnvDOXp%dxF?hQ> zo~ht&_t+xJ{!}%IhR|^}III!R_t^2*)o=b>WtwF0?55a(_cdk`4+?}A zA)hT?UIK=#W~<~@u#T<>0xCHBzyGG|W#P)3-e&jD=FJY@yV?8wABTP4h$|&&?2?uo zSW+A`+6EO;3@DgANR%+ZT?6f)2Dmw);4?X(*up@$YUeN`M(Z$NM!dW%phGqwAI4H8 zxdiaKCoTx>q&2BCTKTO|;iIGXhidkRRsi9~Qqs#`GZ(9*p)r}FA8K}N2|Eh`Wu-N0 zJ?<(~e!7xtfQ^&C(^9g064`0=yyGb*jw0*95jsp>k|ltUD^~FqgspXJlhsR0>3xCN z0MM3C%YuK84!b`#(fI@ry^*#;4-H| zxrV_GZ#P1BPP-8r1yvUD;n{ZY^=ynQ%ON|SJ`yl<*jSX*`>zJSHx?c+$HhD`!vm&o z+Kw_0ova5J)cUt`>Se>!8pr{*iQExmlU?K&h=(=}=-Nmi1<00N69D-KVFh7$-vQMI z`{9iRv_@4i#z3{*r|`Y|j>rt^C3)JfeoNvVI2o5i_se&Ay6p3@dNDWgcZF zRzhW>+qvSoQw%t1dG*`}WYZVMdIdt`5zQFVC|o-+y6379OwLeD9Ys!3ksIc0Ws-p-pfqBsV24U`8I(u|zi)PFgpR%}&Jy)P zvxMz479W>Y&^WS{UMi^Z+W+celvUe3w$I2Cb@PEe2YJ*poV(}@=V8+wuFXhO#mhev zu9bK57e{mV2vb7#3ib+a0GXv$fqg=>)3!40bCwEpbQTb$Y7&@o9{O@K1&^tu^aB>N zfTFda8dYJv@TBFqyUTbmrkSy9EFuTqIcFJxY8%BQpxbyjy?SF%Yw&O$ZLhp+jy_8g znJ_AZRrDYzu2k;o@H2FkYW((ftl?q#3Xf^z9G6*tFjY+syAv{_Yzo3Jx?hu@7GnqQ zR(|o!IH(Cm-jQbYA^NsBoyJ(rCZDx#ry-_S`Bv4JtG;pY`xo8-(vy%AgvA?f-E(HW z@P>Eax^J`mE>;!{RrR;GQvb!Enm)?hf0G<_U`zM9kz8n|m?nyxL5~^|y&-{GmIyQI zHF)?)`iW~Izt|J0(}oqvO83>1`$-D)XSfDz7Dbb<5Y#al&J{t)pmo;EyX44Pi+r^e zLFmBEm6eAALtC6Mg|kF`aynA#;Md#}_l9LS4{I0n#^G1pnzNUara~f>b>T_#syR`C zhk19y*1KpV%~?xyz?EF+u^#$GSqKBm z7UlV)*g@&}bJ_Chep%9+rU{Y`tDejxL*^~mLnkXAF1RAraaKYRUp#uLa}3aIDf)n@fc=(fJoZBJ;t{JOf` z`;i}>9dt_eNf%^su|hq1Ue@qA9C?n@L*t#*ylCELKde+w5qEhmVN!f;BU`LNkXhwo zLx8bEkTR=u{#a6M9e9gmCdIacj+ZZgWn&OHagkqRc&CH`uvTb1en6SW9lw3iunb@W zBXdH-9_RN@|LI?R45B4_=-pzHGKo|ey}^4a1_+F@aMOt{2Yqi0L(l8eWWXILfaB$B z-D*_WuLXN&BuWyd!u4L+4e9<6a6RZ?cM|7b&dU;2Nj7@jrNR2^0H;Kw;bJEtOs%n$Zj;YASxZdab+#}vQ-3DAcWu+0&7K*E7fLR)I&1@}3 z%=~NFKUXso+4%GqM7jiit@9Pry|Tk@hs^AL+rM57TcIZD=9x$62d_1&TZ13@EuVB# zcEmf!Jyo=YxhsigwkcOk8Yjde(Ac5Y#?XAubgWQI{b<$?;!KmB92OjAvasqw#y_An z*d760bkK52l{8PQ(D}DBm2=0Pq{(tzJZBN@aETQzCQ7^O{7iH4pUtOrV8g>q#{CBO zBcZpU9XoDVG-t^P-_Fo1!H^T~V-EUW61I8Vos-FI_SRGc?X2(@JG@|BY^J?`;iz4i!KO&WAsISH?3dTgGtYErU) zlk$eWgF#k0J^ zzQ^V8ouoZ<#heu0S;@s&2W0K!>Wt0YG^io8x9DJLjXaefBj0m-9J7N-)gRZMb~DJ6 zpI#yhNvs3QlUyTtl1ed9_m@OPHV8kJXL72-fq2ieJg9^Jh<8A_+^b%8hnMWH-(j5~ zHKGK_TIr%M`=vl)!G-w`$E6FmRieZmpfH(}B}V!?CeJrur19Nl>7?`v<19I6gs5W_ zbA%#?sL11z20rY_^~Cq+6RZ$cbCSiivmopP%~UmguXaqwblL_!X4vKjc6fr%1RIPc z1ndQdi~+ZLXbIG*!V)t5T^;gTf-aYpAuB^P*zo0mDkn7AUxOVsi-eHV3yt#Dx@i@Q zfKr=u0(V?N*g8lRkG9&yYZdaI3MosDOfY5H$79p;LZ%w9G@u+vZgy*gr@^ZK9+dceIK z#tpcxB00Qd-_x!Moa3BFem7^m$l@9sDlL&nX2!e#s$Vbq<1Z%|aHiT>@;~H}6FZF-y!Xq6ABtB{%*Paovh#l61^+C#493yOSim}pSwoSPR3rux8x zmqKwTKb?+&UC<->QnyozHV@72@o(K?={j4!Lbe&nPV#5YcK*vtzk8NubI?HOCmKED z5*880hR4gxU3QAwIX%)PoKBTyCFeB3kNA7N&*z@Dits*f)hxw`iQ0eOls37>4kK~> zzQ?B-%!_w+|Lm7!+ZV=+-EX8$vMDB$A{kUzMq`3b8YJ5XKclSKfcwjzgbC>39dUX?lG-pW@DT>@TS;I<@3jrsy&x` zE}N7=`sj5&9|z`4DNuFM=SV7f2=&l=-KsqYo$$EYv)pCBu$q_S9yLv?SngRb>)=<; zu=h^J3RLMuh+Q!?VM29k>$L!b*2(Y6TS|27w2lK8sn!^2of3+HFk>DSxm1wqy5D!u zX{i9+u9$BOQtGOQX@k9*52}h z#SVAG^1`7FS9tWPSm_}u^Gi|hc^TwO$NbBkLht*c(eJZK3ZCc!}_WZeEpSR zUxPsD`;l}7*}=|Wap1(bDkD*oPce|kkWED%@s8oO%aKI?35QQB|GIdcIu24^s=y?{ ze1%(qNgHfpe1Rv|gDHYN$A+9r!w-uOzcD-j9r{{D-<%p?Ic~FA_We@;T7!LK z^=0_I`_WoP?SMGf|`YR_EG^xv7mQ7t1(5R08Ve}i1&-HRK zv+^8WAA&O+1*#Lh@}N!`j)O`tAJ9ZW^&w@`;;oDz z*-WvFs%?f(T7@fCh@!r_GTPTv4Y6^`Ik4?wCX|S-6wHnrE}Y%u(-FFksh9T3p%IXk zK(K0+^fy_Gp)q`gRktR_-3SmG*N9T1Ad~g74{B$Zyc7KJ;eqw^zO%|GfuaZ&IHoRkGzX$4NC=Y`Q!V6jrS% z`<;sA zIr)u=KQi^WIjmpSjCV~>@fR;g%Gzvh4=9lYZxv%)Nqz8!;YeeomLW~|7tm#5Y|GDh+9O!(Djzx+4T@*gw2e2MB4 z(BdXakf_#RF&^eAM+v&1nU>~mSYKG12a>{S71^Fh4=_v(px5^}HcJpcdC?+!w#FFS zYK&Z0+a9x84ac?b`py(-O=0q^4szVTpP0dPxy!B)jO*PF)+){jcXD-*iQmj<&bmK6 zjZXbTDy=GY#?_EUMX!8MD0p7l<+-ww7iKb#5lYr$dW>KEg4@kwdg+HgYsS*o#et*A z2aSx}Cw*NuIJXFUWmy8~>JlfY;5G2^gjSI&>yoa53Lhk!ECq67jh^Fph4VFWhVxxl z2>m4b-+}X4qGHe3uzH{fN`+(v%|9f);_dDOV$8G#63Vx*?5EmxoI%zD%EYQMe0%{; zEWZ3+=r`A!5;Hlh$zp~vs*|qZuVb=ANUxtvw=47%%UTog;bN${F`S0+mRj1XzNoIpPd#IW#)&ZNDwD5eO_^TuTR2-#aX1%%sy_pL~Ja<66P(1ii+ppCJj2-+zW z2v@l7q6@ql=+7m~=OuE^ky7#Dnf0=DOsOh%8ZxLft1mE1raVsU$1cZ5LYrk;MM?+~ zy!H5ASAYJ)<*#4)zCH~uUY_Lh{^vh@e;56J0u}v6|HAis$W?KHvR>M(?wDQ^aG2E6 zmpM1)+~L(raW4G+YEB)gm+EK9a$zOoSb%VJLxsgzHb}kY9`kOv0ajadvJ_H0n&;Mm zjoL;d@EoC-Llmi^BKrx_&E-Mws5`KDS)w#mmSA{C9~)L{I9U+3RB#z}#U$-d`bec;fhO^nNp58FCLy|#Pl2z2AYvh(2Zl=%->X5YZ_bOM0WP}WN zH^fv0oS|XSz&uT13maYF-zQMYC&9DDx7H;Mo=}(;v!lm|0RevMP*y$_> zcE;>55<8nHCYB;GRAiQ@69dWg9vM*TVu@*oq(!3T6e&R*WQ}*c&@vFAVzCYk*536W ztK$Zg23X8Y{(CIRUIrLdvb=pA;lm7-T-ZJ;W!04>}nO|E_h~>s5D~OxrY~OArk^G15Zm>uiU0a zwdKIE#452CCePk?ao+4(eg@$DHuvy)vXdQf9M}e~HUdro#q6a>4$$!PGw6F^_55`I zCePCF9Z=a(CrYQ!b5O=yBpV`%fE?_y__P?0KJhAMx}E#IF{!0Wg8T9edUeDgG#$dL zS)yuAk9WuPI-oSZOIO)phi)U(JiF{;JX173`Iv_RHH|+hUPKaJG84DN2pgFclR-hx z=ty+W4Ko<%{dQ=ReuUkCvP7SLqjK)H5Y6&{g0M`Lj!uU(*$QcgWIsrlA*ZK(-t#@u zJ(h7U+0VWiDogIgkl{7B_*zD8o$X&nbPSF(*HiHHH=O916ZrZ(XXcfoh<-7rImAg%CkLMuIGMpQkB+r@6 zc9%2*_0Fa=*I$sFNMNxY;hKx5m^g~8ry{ZH9V5uavLPq@uY+!-w>cMr*W{M_xNmbT z&xfWzF;)(^4>pb%~g%Q9B;qMJIB)>6);n2xz(O+N@4fWpXg5O{*xIGr&Da zs(D(U#-Kr`q`+2am0UKh_y20te)-{hzmNRMuilINpP&5tcfX1h(~J|Ls8N(2^X7mBMIlHV0$GIWtbbTJCZ=d{9;vusvMQot_2tsP<4QRwxi7 zZDROf(aX|YPqJDC$Mt9Xe{1qvFMHa-aELD*xJ{7GKpzIaAw`gC z6%dvSDD<|qWOzQ9SRnTK6WRDGE5uItfVb!u-!>R1zw7+nb#l^y4U|ue43vu$bDkn+ zsmS8+DEWE$Jz2r5r$^Yww1^})-zTdAZZdcIXG5FSb)Lw>@OW&q`t++wA$Yw=oG|q> zDb)T;b&hm{OSMvRhqu}F3V)sYhRjF#+?O8!EVvVeT$B4V_cgPAJ6FUCX@fFj$LWhNinHs_~5jx03IW%pDPcmH; z3S~XOM-7D}JA88IwMfo}E?R`mUXY-=U0y-&1cHo|5WAL{=-1W?d<=^W`ys>dcd>Gu zeXV{*hbe&I8`jEaCRUUc8*w7iM?!$5Ao$RPV0l(Gm>BEnE$SIK%?Rf#%^$z$|(Fc^L#h0PWj#jbX zzs_@Ac+xygr9`LPL3o>< zH*^5%iu*`~a2=Bp(hfYLxO|1~H6px6t2ipI;X&>QlGhJ8#fVOe>tR29m>1`ECTIyy zvp1w)qN!0;O0fG+oo6fml5ju#Ek_kqVD$EgH~C;u&}pw!Zj@Iq^jASwej`8!X^;0g zerbN$%}DhJM7FjE<@=JhMkRwJKN z9eCfXP~0J_4t&IW#JfJ%?o#Mk4P_a&pS{oCXI3lwH@X_+Qt4aYI78O53oJS?D1aS% zMBFHeVzyGS3X#>mJDCJ!KV;5gmHb(+#~omje5&Yr{${_e5}hPNmFc%*>iEILGITr< zTlNOUlleuD_WtP)26X(a_TLteJ1<$<%6cOeyJR9WNRbCr<;I*6zYV-})f#mI z=_eVgRBjTr^7ZBg_ZHw@u29&`XeyLD7@Sb6*yPhiH!Rrgr|&AzF7Jl@;3Aw4TV!P~ zcnNjv_Y@1|K?V%2j_=t^a@a{I2TpxBVI-l-DF!G$O0Zv$PPq}nPza>f;4V51yos7B zIyO8uEJ;u*>X;rAlxtvIt_2b*O}eUnn7T}#;fdt}x|!`j`cOVSmt675g8JE-fNVOC zd-~NvXrG5ETdhELhU40yG_uo1xmPS@*b`y)oP#HFXeqhVFw-XFAU28HMisp&K~m z0jIoM`R6!;GQ8si@WE6{O2c)WdgwP+BSZfc zUBT6oX7ypRi;nZY!L5=-Pr*4C9#D1w&^r9bq&n?@q$eO73z7o=-`Zcin7zS)@nVL( zA<3tJ&J*ciQ`w_RngaO~{H5d$eS zvpU)L9PdJ4p`rzd0(*GbVU>~|`Uw5twPtm_w3bet(aOh9T16bWC%Y@j=`m-WY*)nf?^IIxGsOke<$J~70WCFl(+jllRN2EWI@ zY5*(-%dkDdv>2zujJckzrjmqBw8()$VMZV#M@u@yC5vS{7<8Utes8;6gT6ELs-6^H z=PnVXsgSN?*UTj|@A_->wb6Pi4*e+zgZG4Ys?c+{$w%LE@pIWK%GNT`GSHaFztKM0 z7)*cO+q9FXQ*hHMKM*fnOvC zl?~Dp;Eu&0w(VEiWPz+-c$0~2{E?L=obX=7U5V*l*vytQZGH=rpemfDQ#L57_>Vln zMo~OHSM6C7&>^``Z2f*c@9D5wg6B$pZK(JiF2Ib3crJ& zVgF(_R+t{4M{Eq?XQnfc>CIauS51?4Un-cGN#xxHyRue(qJJ8&+E&p8G;VmS=!=}x z8NJ*t{}$$t6$>!7tW^!l#?RBn8XnIqq463Qjt13!50lr%s;>o@38Tr^=Tzb1)mAm0?+ma-X-x8cg-iRO~Bt%4H=sF z6NrFX`E9}yZWY}Gjh(moHhCg7VXjO^Z{W2O$Pu=6t~hHfjSx4+hqlGkKKI)b`3A(i zes0Slvc`e^^UyqXL~2ta#cZL-CZu?Zm-mK!EZ^mPj?NBiC+T#5Sn|AU@(c5|ieB#O z$>lS2F1fN)`|FdZOT%utpWbKhBlDtT6bC)^;kU}ZW9lI><66MXR($`EgaCGfd7Kso zvn?wB7m^MweD?N!)X&V%>e)MPKCRoj+J!C#iScL6*$>I8(URjF_wOe_;WR>uY^9iZ zio^kJscSeoloeebx21Zc;^voYU)QvA4hEnoLqAzEok82)HhA7bu>-;LX0q)y^B@>S zkxZz1)z_04WJbPI$|R$1}lR) z>HVSTd)F$~OirdNIpxze+nlTApHRz1LBy6=uXxZEoe{$X;NIV%meh0@doh~v> zi#L-`4)Qh(y5bD;rC||YHid~?-Ok&5yX05>3Vr*+wQh&F?a*T9#Pl<-Wpl6$7)N2~ z1J@yY!4eSPv6By}^cm*JAgrU83Q}iW5@u63zS_H>|0~%P5-A}QrL7zW7CrdKHlIbm zZL>J66u#r?>*60QY5wOTz#iV$h+6cl*Q=#xgSwsXLn;xHSr0iCd!mO-tIFqG=CqPw zFIk$ZMbbU_5zty!@oQl#g!Bp;q{e>AW1CE$R20Zs`LPfx#^P98>M5Jz-!{-6o@L_B zqctz~ygM%Ni5bs3#`s$#bwCsc@qgUn4ijj!@*$ERfdpV!uBLem@0n1E?-Eteql_NVk1*!7sbGqX9pE|f@oFiAXyFwpXM}3 z&kHla1Bfj7`X-o*Fn^?7j_o{#oc617LK~!qNt|$r`s8$AY>t;}6*oB9Kxk`A(P6U2 z)_~}#LN%Gph9B25vA?2pidlt0GvlrZx z#RM(ktd(!z9Sp4Ib-N^s9!}pK-mJdvQXpD0dv|!rtPaUms_tKoe}Czl$9}Sl-bG#g z!P0kI-?Af2U}eXE6dU5=*{K5O$P;E7kn-n`+RMl;2M*30F;vH&#dBl^mBYd7;}~S8SP3MAkzN*^a5z-dDxA1^O)T zhUvJK^p8JRw!@EXwGh6vRbPVUVKiKZj@zSQ~7y}rdN)N-)vrB6I8S}bl6ptIV z_}fzx{=)zr|I}-Jq?w(NabQEH&j@HYD5irVtyE+IDS+y6Oh?n;p6`}BqDymBQ4QO= zSP`@etDLO$NLK;dfiQN;kQ46m4uAsdktRy|iRJ`Joq^gbNs$k2$n}*yY4inVje3n* ztLTz8x~CEiR?nlt%MzulpwDxoqFvE1>4lgVM4tn%gH9WALiVwmu$<5sj?M*t1E4*Z zNzrdSti9J8k_f%yzlEj%$A zeBb!)$FIC^fZ9jvSG`JZu!EWddudh}@hJ3D%sqYF?J;67W7}gy;ZWmB6HM z$MueDynMaS97wvGC3V?1%YSi)l$tmpN|7OgJU>2%X*(4_@a8} zcFxCv39uc!CWrq>9EjC;IkfES0Z7@+4QQk9T&cF=QRQwn-1 zj)+ngQ4Hkm=At^y^c&#Tc^;)Jg0SaijpQHRxBVM9$DlOj#Qc0kBdFGR`LWk-`-Ad~ zIQCl4%%jjBrJv|HP3l3XJu`PsJpow}Maq4FgHD;8BP3qF(xuwHU7o_*r9Ls=riv6~ zh0R3PXmV!^=RSGouO4y(Znne*oF|({i(xykU@A01M+U_}w}=!f64zcIX3}S4V2>i; zfZwF*Qr_@b_lgEK`=E<fL?XAbW~s#km=P1^h?TJ za=2Zgx|t7M&q=P!4LSv>bkZP?szA0~0*KL{Jp?K&HtZ6?;^!OmqLh#>dKFnBNa17+ zccZVO`z2^m~b+xC1V@Fn8(Q|CisM$ANu}SYQWC%fBEKR5;qAN z5{{JWJ17P^Tx~;QrsX_z#HOjxp$ZCq18;@T9^q9ex&gXa_@}Kv)pg2;HvE zLcO`BpPX{De=@AqYhjy5Zgv)SYvHg(h*Lur|0vi1l2o6x&18=Q3z(Be=&hib0~Fau zMb=9SysPMR!BM&;C@rj8QZ_Am^72Uu%6qaEleCI|cwC>Qlbj&y=MA`KaZkbKKT}mi zq9-q9Hgm7>5~rLdpM`xMJSgj%-Q#;kxRYD#IpBu8Ujy?h12R2utKSw{_G*{3H!J}< zgyY7`Mk>l`WREIg@T6I=vbw+u+0-`&>bRz+jKkKRnK4@+Lwj0IXxHpQaVy;)+(Se0 z2e9rmaw>w(N_MI?PVNL|re!>gM;uUg&F*(?l;XL5a$MD^O5vTA*a#5jR>^vxn48fy z=h*?}+lRh!$FvtGn@k%A-Yc2u(mB9Q3EUmt%5U;q6`)mY_Btc1mtw)P2#K56qi`RQ~4M5FNaZ{>G+ z0SSwIn=%b{KW$)q@A9vje>q^A(yq9+5W=IHAa=tqOw0{MPXVfs=F+KCSTbG6isQWYb29%Cj1Py)oqOo zz4~*R6-_C_sBxiK;6wTr^BWP>|!)Y42gc%^WUutM0xVB8QJrj_xd1oz3! zS(;98%pO$YW$XZL*F+f;6tIMn$=&w*IJHa}|0RQ~q~&WjGf3eia>huh9Hy9Rid0~f zq=Lp4it+L`(D%8JdJ+qT=vIDtPy@fxEq%CmPPz)~bx)8YG1S7-m_`F_jMCi`1)5Y<4-`5+bOnu8 z>{$g&={OYLmMlR9e;v~~O!qV7)aud-^y)S@gl&a<>+o&j-l;=QQF9Wgp|@g0Z5~~8 zCa0g^cciQ80tV#y{wJZY1Oz4dH@t_Ox;WRQC}ELldw~1!OYj#-jtOw|TEQxTrY;~_ zvfk$(Ae@K8lF4+VqAuWbc>y?nvHfrz=e}|;@E{c_ABr)70dmPCIy%;^!lxRJt2wRw zTp3=+jAcx?X?1UtIc(+vJ7Fpt_+g#Q0KnvFzxpkScVGY(7)haYirGPtWM~CCOgyO1 z(5evj$!`TFg<$zjm24TKi4ozcd%^Xx6Qtku5yTy_FW2LqT-b@r=EBDKb+=yh7|&zt zPpoh<@!wVQzeo%?5#AC0hHRKbpi}&aeRm4QY^O*f6`2xNM<;~m(T{&4n`sa4Q)p&a zF|CJJ%ydlO=vfaLOZ{X^P_w#+w?iFcd56@oJOyi3Vl3Zh>AkF0!f~DD5vPK_Z`zTM zO&Gv|%^x$31(!o+!Ui5@ZN>zh_B!naUTPPe5Ee%T`0G`VN(jlQw_7C&my4K}cy5IcfA}ugl^hQo~6J*}&UFXZWCQjS_Tu zoh2Gf(u)!4=suD|+9CB_2kNPhT=t6%2PY!j*e=fo0f%!Jgw;D^7T#fA&XfYxkW=o= zg6T1wv$!{bIkbxNAPjNZBi>LET&C(FIMEexGH5wW;2Uy6a)R^n{;=G@%7A_1czHd) zgq!Kx;)j2{!|J|+TjsUC?5$;^<~%M6%@WiBWyGRICCVNe^LI;>wf^|Id)i^ZRIe

={Ry$+o9h$(8rN!*2%YJFqcV6YP*!tyWw;rs)!ESo1N1F%7oA3Bf^fE}9j`_h?NYW%y zYV_jnrWja&bSg5JTQ;XldW4k7^Ftt^4HD`5pgg)bB7=S?c_`_FXz}*xEvobC){yPf z%V%K!xm}zDl0eF#QKoLDj$Y;77Mu&b>IHV(Q?XtE<9E>N2mGQN%`dqRB7N7?7v?Og7H-3@ng@2SpvBu^h7>0t$QUxYfFTxSQ}&Vl2mxvL$%FSzR_~fcqf$LO>o}&#aawDY}At=A=MC?&&5eHsG>|8?%oznWIMIc(yDncB76 zKe{5m6^Ks7%YrqMQfkfIyOMLfQfgJe3CKSMdtW|vCW%GPP+aX~L%Rb7j>}1fs#{a2tHprmb zFID}GEMq5w9N66lVTBPoZ4<@BQX~d>NpeD4B#@kg!r_w3a1M?p12)%bs%lQ>IQQ&i z0iscs+J2|Fzl%LK-2e;Ky496r`)C_#2VQ)rJQ=b0ITQoQS~_GJsSqL?8+NBLRJOJ9 zW5|urbh?vn^JwLF(D%gP4hBxE7Ri7saB^0Go3%*RO{4P{{lYOE-~!Mmcp-|$qD#y! z9}85NRVQXAAHxOnqUHC0|F=V?NfQpMe$1pn3_x*w0lmh(Jm|Fh8M@kAg9Rn%4$+rJ z7SQEEn)0Ad*|HEYRY;Uz`75WrH0V}OQ{l(~*CV`go+jS|BMI^HwQjZZc1SWps=aLv z5nI3maWe)ajC!kic#N`fo8VQwqxd%l5wq_-Hz~QrPV+dh6K1thip2wp>8Hp&@aD-X zgkADPNrkYM-Zx{r5`_^~o3sXZ(FF^7

>3SFl!vrJnr+DMva$>L44irq3VBvoV(e z+H653#gG%Gl{Q1^e!LtAW9j5T%z)50UOS*X4H@ew)H-RHutjo8k>Wg@!-icd(@2X1 z=fTm)Vu|DMe6nbx|54ddSzj3DO4Ob>BlPJW`)DXm;i!`D@vd{bS*$cVJJA8GiaLzFV0>vHmCaB#)OLWxC~6 zkOGtqq{+#2;W(G?S%eI|F>LF_f9`zC-nzTev=Wj{x{L!O!%RVEEB~A_)peU9%MZ6k z4>)^-w|Ga~F~l+c62n4NJa+}{ak2BA+l;Ao@dZJJ8C#?@qR_8qfhO+FD+?}*J7pQ% zUg&?0sY-v=toXKO;D?Rs_}8z9o7Cfm5euR5WdCdr7>qV&yzsA|nudkUL|-r?7OCC_ zW;O9Pao783u5+8fP04*r`~AVQC64A+D|h%Li*|*ixsJPNu@FQy_p!bRMfsPm?=Y1VY%I^0&v{Yp-5GjA zdc-+?a{Od;h@#RNcR1j(SOSR`EuTHEZ~5j&^ImoaFfeaLO%YZ8dYeK|0apObhuF6H zmU|SqtE%YAfGgq{&SCI7VR0j1KIv;7{jdr&4bJ zhxT0ytHa{FKc0Jdb|z<)poMhsFUZZ$ z2lT~C3K^G|(wdcc`n>FS#A+oR*D3UN;+Dv1hM?U$yMOjevdw{$8uuIJ0Ay24CI!uX zBe%P!xEv;XrflF{R}aVwl%GqM&jThws8r5_1OUu<$7<(6rx;N#l%Mu;2e?|MpY+fP z%r;-ly^jqq;ckIzTa|epcEhwaI36ij?1f}@IR4KyyMOVf0WQloe&=0M`Gtw`UNQn# z9mN3o=`kv@j@u5UcikTK5bN#b7JA?H#g-7)xx*A;s!GyDV@Z2PNV+OX&;d>yOhV}m zD~(v{-!Z*ip1)wgHELeG{BGD?SFAZ~kzXS{P=g;Y?HNGzH;whE#%@)unAodHp+R>wq;t}hIt z)*59EJffHZirk03B@o6(H(!Sg)75Uy&IwHqNp&5dPk%MZXMj$dU#-G?Bn=YH6hKGH zED^S0!@f8v&Nm=~;A)5l`}D=jcXI3b7X!9AXNfAw>dF12Q2{kY`l;|EHY6!$;zIgK zic5m;fIN}ETO1Q~PMP71z5R6bHYh~ZbWHCj7X!AuqW|7&!t3%lw};{lJoN7XBS<5V zh3NSla8SAmf8GO*96WnZs@W|*pwz-Uu!`W@0I_S}6Hf_gg}-XcD;i9%DfYw{WzAg3 z!W#$tnStf=PamJ^Eip}uW3#_Q8I$fE>jf^JQPMkCPQY0E1|l|AIDy~2iL$BR|s z=!G|Ba1?J(GRP}+#|Ij+zDLD@N ztkE8~Wh+10o%TIv-cp*~{p$PQGGJogzx(}&oN!bZA!t+|f1?@r)tc04An&(I9{Kf?rIU;N4nkE& z->V(dufo~QGg88K(TSv9mZ7>#wg`^Uab%tIy6J;X#a^(1G<-)Z5CXM9Q|9$S`LaUa ze$vkrgl%?>67)08>ihDAQGpL(x7p7O0V}xvmc%J#{x~dVUWeuFu`NT=^YsIptCr%= zJd#GqtE8ON_i9Zgj>CqJ%t)6_KJ5yR6SH6eEW(5Z^VoU_D@;uM z>Um!>mV5-q4VJ=&ex!cdN-^;iiNnhILtLc)LaS8K0nM$kH~U_12tAd94SUKB!|3CR zD2ssfM0(g7F6QT39a>coWKc${<9qg!oY4#v2R2(z7y+Q1V!-WFLPef?qv&6*EztjL zRv#w!N$HpWu=n*tuQ#jXf2{ji@z)2PR(`bsKgpBn#Cdz&)`JRZ3|cQw_OGS?zIX>* zImhXtbLJjU-u~+1eef}7ZmRN+RSOm;yosGg2flXcyNmyHhkt^BCg`CDpvbT#GITWj@$U@8&Ah-EgaYKo4D)uxi9-ZXT8toz#3R0-zL>y4Mss&dCPa2w%)Jx(#u$MrB3Sss+*-WOgEv|YLL^5u^p26KzI*5%IR zL8mxUKxc>L(hcE9NS&xd)WJV2UP}6)O;J1DIQJ1Gj$^~tD8UL>ynDM+e?PFKK&ZQ3 zn$6uKY!eMS{e!L(tr846wZFPbu+4SQ>59BwHt3WlZk(I!k8^c<W^ZpPS`w4@jcV63>%f>xCR!qh8QV$woyz1 zx`86o>8`J1-d&3%S)_^K#c-lsu;VZi{r9^z1T`}|B>FZ-6&<#a$buB&h2u$$ic=P zxq(@N@zE|8z{5zjz=v_c2s?O$e4Q#?V2~H@)O>F*DP!kObzn#R1taNll46cggPCOL!GQ(VozBlT-w?^0)i!4AGxW;g!-?#MSQY;B8A0_0T82wuA2SDi9wD zUB+ovUxjq*k&;jUx{zlYEkW4|zUxALOm~LAbT@ z>jWj-t21)kaXe)9!g;F!V19vY0}b=@TYJU4T>7n3&A-nwfa3LQ znb9Q6fs>Yw8ZGmFih&OBg;Zqk{9E(uMEakNY6$QG+3bB%zCz)z{=Jc}0G; z^oiH{1a%@LJwdgB?vh91I?-K`J|ujVE>IuxJm9*E?vQ2pWQXL-+eFAR+N|E}Rpegr zPM5z8rHEy4c&_dkanW`-F)uwv*(6gR&dWUiYXf-R$f@xnpExjhRvG1C4^Yf~irl3l z+hq^=rPLX}GEQZ{9&WO7om-UP4m0{$WX!=_xZ{xhR_t=0r0`;b5Lg|uK4E!Kc~F~2 zwdWm1KPthw$ZsJ|bB9@|KNBxc2&m^H=suovBlPJVT15tT#heSW9QQkr_m-(joYE(} z5ZD`*0`8ov{1tQ7%G+gy;$3tJCz_WYQo_-n-^?wXl`5(to$4O?y1u*l$AS2VMNw_? z9PU;Lbnl`o;qymr@}<;0*>P3-90P!D3%(IJwE9pFV0Pg1_-*1pLJZ)O{YkT(-Y_R_>YU@!^f` zjqc^s%coS^ncJQuLC=A^%~7q8f@hNRdkD>F`S0{Ak_){fP51 zx4<(|ZTh)UT_LmYVS-xP%e-= zWjaT#=z}M>zyJH%;oc*=#RRV;x6*!c)0EDEO~A>4 z&6z?Ygk@07PKu$Tb&qPPAk{SnnmFgnHoHD`pc+Y?=u}{{`m?Z2K05CdL7oV%*~RZ~llRrC$NBu5%)hUr?S~yU z1bBIHqswdqWoD^mWhC4#+7VKc6U9_sl z10dmBUQ4i)Sw~j{bq-Gr3EFYTP#XxfL9jc#&aV43OA z*}!R{yPc7DYM5caKkT~e>Urs^bdQ}>x#9-+7XPyNPnrtl4yF(GMwk0W9>tF8&)^*x3bQrjWzWEWzgS%cR%c4RYxJRc}8>HcldW zMz8ixiUDW-cC7TmzNcF`rAm11)-gzy5Cc2}=Y?JLrJy0FYVSv$!|i^my%PfRIBO)( zIze{E9Ur3o$?>gzyeuCV+u`Ql;8C=dx zMO=^ot;GdgC@TE#I|-FUB6%UPN_j5n@a$V4F)Bz|*ZW&y5_V^bI zYd|4y*ku_HgwUi|j&UfkU3kv-urHMKK_biVw6K6PKsI_b(dWWD6{mo9W;d*K(3aFL zyy1y0NgdFxpB~r^%hrHj255EQ0(PB#;PIg_Ple+dK6{6UMu7=2+dKOw4EV8ZhHb}e z!ah5|KMIiTvn(Tb+8P}zh2hze{ah4e8Ibe?VSkI_6j;HjD#bCM5}Pn#e#?}+V}}VQ zt8T;RO}1uNAmiU{F3(WO(hP^Hu1a91DhKw?aW0(@hsI zT*p120-s1r>;0OgdnNULSlok)E|OIZyX2}esX_M#WF509u8N05!0HV%O6d~6PLBGJ z#{j6Tv92<-31+sLlyR9^rC)tBq`}(+n3f+`yi3xj zL#zAfc`2aeK!}h-#iMf9b?>`#QOQu98h~jW6CXPs&2&wiW$+X#e3}gY!4`!<+^ae~ zHK0v$ioPk;hITSY;_kQ_m`P6w%SUbN<*cT}d41!O6*t6|!ZcPCT^{rHa`*VOyYEom zoY@+u=ih(zvFHNqsV+N;tm_$5VIwM@F@j@Gu|kFOhS-ZXo%~_6$%4GIcV{-yy)d;9 z&1Q!114<61qg7OVk9Uv95x()?d!Syl$NLKRF7QM>@a^LrQ*DHlLTg+t-2`=`x4ku7 z4Oi=i?^hF`@Z{HXk0_Hkb^b#xBS*4Do1iuBI;U1vDe7{^1lAtmQ2W3*jVOo;evgG$#)8S&>6CY6WissrqF+RNhfv zB54e6*8f+yglyxQADQ<_PdAGJcC{UBecroU4k$~x8U+gbs*Nx8G%~gzA8(agidVV6 z4kS`*=G}BFq8uMV9C`MvO~J^G4*tt(U7XfY^S0D7u)-!k;Joe@u@Z*KQ}qx{Y*|>T zccI{*s*f~F)unEw|F5w82`vB5pWhy`R2^{IU>l^%N9}^MDLM35XHfB||6S#FNthdz zMeCym-SZ>6xYtx_Jb+uNOPo4RGH21Q|!QE1o6@oe@XY?C*FmWpln% zAj)|WyCeq)k_r|Cwgo+mSjAM*Nt{pkolLi?W_D}badI{~2V(xWCF=s}{8L0r!yro( zcinrZ;6ijGWQ_2s5C{r%bj9so@TbMppH?kwiMuOl;x`bj|29zp_iSACXUB86 zRkP2=9r)}$wgR1vyBWOPy({hlxfy)kD`!^Myhq}$xFZV(xsP5uvtVZs)Dru5#U1$O z#Rco$Uh+!aesO5m#Y_* z|DT$-KYr_%fBMG~>g0mIzW2)|Z>BG7ee1yI?|e)9UBl7i3;*`sFT1~b@r~bYqtdB0 zU*Ebgm8yHY{0+m&;st+y@0T|h?EC)Og-$EhV14q&h^{+t&@MB7|H&C9r1ku~{qJPe zbfPswS~?{Mmw5w({bi>JW_{Jw;TfF0+*-aq3LAaY_g}4?U&~t}sPM^}fd?G$3%2rm zV)?_iGK&2Vv%-e+I*`Xp*F_0U*jTmY>d(kpc3zM3{+B^r^60(RR!W{hkxf*5BaMYJ zC^wPDU>#?w2$hl47v=p77J{Wg>QU#PXQ;!(Q&WU=mN1{1T^9Um{%g*RHI_x-PP@Cd zQk{+rAc;_Yn;{vH==sf_SiuXD=PevaK1Wx^)Igq^b&9FRY%zW$(%S*@THFmrvO%c9{5~%739{3#`V>Kqpi3wG>%x ztSCZ`K{Zx;bnrjot@23Xyf;O5c=Y45>B&z;v}^R2g4gVxywmoIWPM@X2_KW|@q<$@ zoh2FUTo31UxXC^dHI!-R#%ak&D-$w6ioIzmHt@C^wlogm5Tkd_DE*5454n+67)<~lQ z*4^mMK$M4t1=l%E{MNW$Sq`}eT^ud*N*vO$I5}jPU}8r^j0|Pn8lF0=`|V7VfeHF` z)grQ)oq=)Uy-kIgf!RUHq3&Qi6_1IbPVZ{p{m>G0Sk@#{qfAlwNX0 z{7Bq6w^v$B=Lg^!2bz!1=5|6D`Rs1Ty-57q?M%!5XEvE;C&q}C&S+zM*>$D^azPa` z%q(0ZNuoY+(SiqaKT+x+N!Y?kj7^=1XHq%ZP&}7Njhwq2^U<7RqpD_ib{o z0}=ter-MExDT>@9?v_0yb-b>5c0<;L;A7_Ln!m9ZpziFuAb8KT3|;)UWpl}0C+_H2 zZ^m=CbecSI8X2PESH~R^cgjxFN4e)^SmJt0nCYqK_xNiRN4fa{xpcmGr2q?G?|%L~ z$9PRG4W>i_^kMXQ&RHqmV_@s`>x;>O2Ey{d+x#lt{b(qw1rjn$PJrSKc)o{2Dqin# zZH+q|a*sZtIP9OoTN;;7--h%>Cu1B}r)-kea}R-UoXsf$3bKnK1yNY>fZr%1IFH-o zZyasK^ep-`eU6_Nc_k)$_R8s?>^J?6;YgG4f`4A*F3*vpxrU>I?kht+l@HQ*v^DPF z+$%w=rXS_reRcKp7bKIFfBxfdya)ljX6c~EZP!`|=B7k9fs`lfo-n)ou5B_lcKIZB zgIM1St>NEyhv(f{HBkX3qZU4`->CroT1kZVk zR(q~3`A(Q=x$61GjWSa3!t9zrBWcvGsgjbzj$22?8!Pyc1O~VbH32Q*y68?OlYe$T zFu|OH?Qo=$m;N7uSD0?Gne{q^~`pI+c1bVU_7gC7XU2bL`%lo?l|MNk?BZEac{ah%mc zGmZAxR5ls9T_=v(S)to)fv|sz0##qJ*BMpVw!|I*c8;Bq$gput)fjw<+>Fo$cglxd zvgmHrmRMc1vGI1;rJq?VFBA=WAh$=WVoU5zDH7NWyC986Cxb^Yc!Q-4811`83cwGx zmm9>jkuGtF+;Nng#*vizXx=kK_haMAc$?LWv;C&tBg zGhC!n@)U|BQSmx@nfznb$9#=q7wIIJJ8Go2$Mp)EWOsN|y4c{{*$RtE&oj#bSiyq& z-74nZ?@h4Ce$Y6b++^p)IPK+}El6l7A`=jwaewAK&*u{9wlm8k#ZlI+C!ivPW5B;1rI3s`S3j>y0 z>*uXr3Lmy{wOTAl~GKf=l=P1_#kdEt5YVGgM7lWL(x z@!orwNIIZ8O=t58MaRju*-%*rY>V$rX;GI=xEOI$9gc=4AHWJ1FHJku`I2R;5t{^) z6Yq{{%{-nmN)B1cB4~br>r2c4>{Ok1h*NCI&{+K5<&)K~=4! zj(5_xFXoY0jcrH;oD+&Z$Q5LXP-XhCWV@uvwMm*9a9x@oP&psp+udJc1+fvRjq&#h zIPCK(cBek>d3Ssds8B@)!)6seEs7FuHs^7~y$CkklJKXshwoZZl!V1s$GNVqh zngZvw#`VXW|87a3{DQ1AR(PO|fm|p%FAJ;oeJy4=*Ji0hvCj(Z77NNKY})P+7TlFU2IIig&`?FBeIh`GdBDjklfa;pis zA#U)09MtLm(W@FoHFQrel`i%G*a$oQF-6)h91^Mzhomtmwsw}Uon^1M>}S(F8Q7Sk z=(8d1@}hV4ia1Md!5751SfNPN)1R4_#lXv7NhidnF_(k3C=y~FYBMU7WoHh(%g?D9TNE$i$O6S}wT%D(u&fru48(`|}5^gnTv*4U0o8KwJ(x(DW zf?!ARy7>*VLH9l4JIs=>DFA`3)9%P0{#fu~v!Gr2wL1D2gP;%KSWU zLrWaCCZe#1x?=uGG4`~o(?g1YTCSeAJ)o1(^Sfo`LY*vohTSkVG31!zX!N_WP%x2W ztdKLUJYhw>r7C&0*#vzi=bb{OQ01`OY*lnJOQjkQBiAJ1M_ zvU>K;yUkhk<;B0EOy=aV$K#V^$i->qLIG%x;9i3 zP%O;gEEbf;Aa9CBu{y3;SW6!#8%6!xwIGF;%u844`Bk$ANK&|-e<`q=ua!I`>GH== z_m0}hOF8QT(&cG(?xYxhf<_zZXTD}P1LIuHYJ!}%L87I&OY^+f?%&`wcouT+6KwK1 zwD5{JFETf3o9GU&T9^UNgGlsME$kIm#OUHGsfXMSx4g&#pDMSCm?eT;o(+OZ>T*;v zrw}w)N5=RV^nuOC*uz}n7{^FgRZ^KsEwmT4#^L8(f9l$$@>UsM+dx-R>%$uSFEibr zjok3o$jDW0m&N$)%4PLFy~0oBTIJ;^{4TjsD}ldmd#ti{w8&4)^*GknmeYZS zS--I)g>u@RxD}cw)K4@B%Ish8Ug{gy2#}bSdY=>OW) z!mXvNzW9JNf+lCetR&79KxgzU+YY1AcRBv`*Sfz~5pG(1R;BgrCc9pk#pk%$;!{P* z_fw#q6c6pJuqfnstP42i4Z@+}#lr0~>%gN~v?;VW3Jc{Pxo=gSRILkWl0qafpn!8k ze3;iH?V5*OX_vkCC^v^zxgF#WaxVd@S~z*CZL_aN0*tnm_2f{OL^koA>ITM)e4-2|q}x4&|lywA=UId9a86qvzm3nkx7!TiLxC^{8w z5`*d@lD~KGJGeSf)`J{%gMSjQ!si%vtr@i!CyWEOv&0@ZY(HUQ*T;M%dfoG%ZR!)J zc&rS;e;ZY(k~p`$+rxFTdhRAsQY`9DHSwPu%K*-3RG_lcD)Ouc&wkp`96W18`<%0G zF2-njrrxD%U;Bm0goONZ=1kHx)_UZ`?(9;t27vpNyq_YsK^}tX^*9H0Ea{4E?!fa; zp8`R%9C9~WFRNAU@zes>>3JavnW3ATAh>UM*TFBH-xqV5J}s%GYrKc)ZeagEE4>ZN zkS=DJ9&*unHYj({>-{&*e!y&+iObO8kc6;DOp8=Y9}pc58HUfm@o=-=A6OKfC(_eZ zF^gK`is9q%&Gxu$bZCz^m-dO(s9oCP#^OP~o-qsR!(rORW^co%RANw*X^ zhn5EChzBD2+><@Gakr^HdbL`)THX@ZFS{3!##GEJ^1Cj?t`8jRI@IIf)eWSCi+3~$ zH-zIIs7r9TM7HU zR!)iGYKYYbCsIA6$}J;c!ocik{mLAf!?5wOvxg}te+W}>V0`q&@KX*W3k2$yg)XmSZ!iM>+4t-}G z8`D=SX9tMMQ~)!Y!hmh{@!;D$(jx_y`vLD zC6v5~f-Q3V9@%aB2(YWHnBF4MOEO`;HHtlCxjXQ8LE;^g^=ZsmA821H`c!NZKHYvL@mmj1x!VyOC2 z>?80^;9bqUA_`-H?c$9!WXTphU?W!8GmEE4p ziFp7Hn4xMHB`>5v0m_iP$PrhBE#qzDu8Kj9`AwqgaC{A^49;44b|iYTZIX7Op}E@y zm2qxHKzu3sI`q_UjNGFfa%ls_l$v?=Bb70qL3}qUbm0Q~^pJKrvOIH}`qw7Bys_)B zFZtMs?aE5CUH2d*e?XDDRD4y`F3-zR56Rky20@qmW=^HDQd}Rp8&vD}hE?%8WtG%M zZhtiVg+;R^?%L<~e&y6#A1+L%2EJDF#=xSpaq4W&xv0ayQD?{-6+;%&{*&}X=+xwlmw<{L3zIEjrhu`d8&>EM- z>mged=RSu)re0}2B<1j3GrtoQ4D&*7z4n=#H)%LM%V(Rwxe+X~GFDXR%<#{POyJwb z-w2$$?2MHYZ`#3m8`XKGq2yaAvYCocV_JNwe41QgeQTEHs*o-#joIhk3RL&dMO8#E z7Gyy&=6i?y)c;t=#g5bl*QrQyOb|{Zv+l6Jf2>hnnPmdV*LVK(7bNpJ3tdBgeH09e zC^;m(@~QYrTBpSL-~$5}Og6JsxhA$Hsx|HooeRhMqyz5hq|P_X2alZ~7`?0EYz(c0 zbFFb%bRqX3e`n0dU7y(md(XlILGhXOu=n>F8PB?nx!dsvfxihZkLAZ#5zTaR&};^Z zDLIr{?Z7N!2R{)ym{65eJ;FEFO;_{R%DW@hfM!XPYcKuJZq-Hkb_sN;VA;aokhBi8 z%Ynj97oEnSzo7m^`2lya0J4!$1yS}QyBzuN#J`iBgU=?z&i;e-IT&Lf?!0DtbAoEo5~^}%YLtW%+0;?WCC8c6rp44&^=`N5Eb zP}O7X=rIm$Pqf+3ID(n9Bf7kL|MQ}^$*Q#cxZ+)sK9*H+UaJTtfupTT0VU6;NDdW` zsUg&MQzOk@8gmrASmXfMLkU5-+#O^i!otkQ%cAMveAk6-_ZFxYPHFy+Ha=8vQrvM}+~Pv|=8}z|!E!4scddqN_m@Wu==MjZUE`HQ<44 zS5(@Jl@$KtQ2R5Psh`Z$j(0CBbMw+WKGlgTlesw-|Bnr%jGei0;(bz+nYlSc$w43a zAX4Le2yBua{M3LI-|z6~RpzPkyt4wj+#kg*^D$V7Ys_|43c)Ik}8E&w_CSLy-Ip695-HiiMmpJgE6Y!B7UY3-i#0xBY-WwefXZA;*2;OUP%BDS+Ia-XCBvYG|wHP=_Z`FQ#_ zO$*WQx_{S2K6K(x%g1I*%Oy&FfgDBQSmB=9eLnF||60o0k79d$u_(Do zwT|g!?#XHwETN7p_|qZ7p8>CaZg*Udw3FEvc0|!g55^97wZF^Dth0ZDec zKd{a}LvWOr%~5Y)5(Rnm0p&@tE~*afkagI!sjOLQAB2LwHBTXB^cB zLwe~hZbGa&n}>f2MQgp_w~eu6?X0DarMfqw&qm~0_$y9^+?U+v|4-j!#Cn_;`ox-8 z4|L&XaJEZYbVl$UdYLDn>LRz#{ebF_ zQqMFJbsAGm8%roO3S8+5-PCpd=^;SaC9(Hza$*p(K%DWdU2*qztjnJ#Lyqsn&6HNi z@wa+kfr9Sy-@GEmyk@DajC0QS4sVCYttoAwWMz2lI^(tsw*5C>9kA46dAgC-c_RqG zdp^1VB8QR#_x4sQzW$9m$=Uf@`ms-502aiglM9>-@SX;|bTZV&K!Ha*23ZCmHZK&V zlTL+3RYRsiTX2FF!j|cr)HT-Kr~mnSz^X;2P11sz*hq41tnw)*wh+T+D7iz)Z&Kt2 z6@S1LBt9AKtO1gvs^(Y3h819b+vnUB;E z$5iJ7)H-^#XI1!W&m^BFNEJcVJ2UK3zS80 zdxU?Qz8S0mJ}#{Ly8|s!i^`ZPzeU3?iNSh5jP#u+<#e0!G`)PTx{bjDJ6?Z4fd0f? zqiB$Q?DsAV^b${l=Be4TXZ5@7C>rByJax!Aj>f9odiC9*JWGlI zr`=&$;V#e#OSx^5-Jy#GyFn2b3E{B?qAsvlh*b~$(UtSD$+8d_thID{V7F?+j9%bu z+vw3mXS$XP^-#m3QB;cbez~gSBu~`}?QFZEFxO>oe;B9APS)qCS)c3ycBuLPzO=O4 zlEeQ6p{KPXMUW7-e?c*wr7VD^-PX8Im`$HsNiK_y_-ypW=-^2ZD7z=?XAZrd=-C=~ z`YTsq?OFjDjZ9_pf@_?eVR$Y-?318vT)Za znk=-;@Bi9=r2K2~*DmT}4nd`P2CxM712z92chZfbWg&-^f4obt2u=``g)Yja4~6uB zYttHcpMw_zf5vB*pCntn`sk5MTjSb=+h#Aqr4a8Na>=5LBskcROQvEwFlQNWTx8P% zWZZ14qeU=zs`J<)U=jJvS}v1s@nH zo(nkDAwz*_T>O**&XTY`0A3O}TDN7{AuHvBUN|09Kf^2ZuF5!f^68gdORH>?R*9xCxP7@7nR%uLJf(|p$9kR-x_P^h5+j(taLkDfI#woSFc+&^m=yPspV zb(;2Fdea-0Db1(D%uej4SV?p0y|bcweN*RU1+I!24ytm)`kPI@bHP9W>u>zH)}A&QE(g_fPYau*y-;_x)q8L3ImvpE=DX_yg~rnD@+kraPAofc zI0_0+k82$qZ*BV0smavwIdQ(b#0)muD7l6rTd4Sv#RpsMP~rVBr#HgT{&bXwQt!t0 z{mU^!vhz^~AkVOh$#&KGPs!564m~5=kf&$c-bZTNR#BGCPcMi+YNf|}mlt*^H2CL; ztK3%492TAzB0JSix|D{F?*zd|uT}=Bk*c)AV_#SYXu9-9fm}x^r<-0Y2LU06(^yRm zD2`u_W`C(o^0sNEdGLiV{+Jwc;=IyTGxKnol7m*^hg5ufcy?q>NHNgRV@9b>GU%?> zMK?eXQWl*b2?gIqKJolWpqLLIASXNwIxX0^`;blut8jxNazlSh5=V_}wU}$dootI@ zuM|fea=9VKvEei-M?D1lgUxtIZ=1&43`pW+M<&ix@0Bc#Tmu5N!!Bjsdb&_B^6@OK z(4=58VqR?g%f{wdUya5}C|!$qzizq-RB`&>t4X30gDT4mR4J4^i6ZYK5nib`FmGLg zNzPSO_+&_q3b5;T3O$pN`LPWXBUew|Usjly*3`MNai$3qLBFnAL^e++6=siq2PMy= z$ad(4j@c&KC9a6ME9s9Oq(7aLPM$tuWM_cjgET>~hS%=1%TqU}8H6k_EM6{dndi`X zvTa7j{uw%Qept&O zMLzN@iTxO;0tT5{;O=ad54$9h9zz+!BhOt?!!B9DYvsLA&IRof)!Y@4DWq*ad4^@>GZB7Yj>7xq%(DM%4$&U?hQf zw`8G4aax$ey%N;VJqWL-q(OV)GtXuo*>>U)3qSsXEWD`hU(CqU_DjIFX0$6{eGHp;QxM9EVrl0wB} z$*!@T4a-f`S;3ctmmt3Okgnz)Qf6~*M%dS<7TH9+3^DFOV4ZP!^^x($Y^`z z{O?IBJ3rQWH*TcN3^Tcu9D0#7RQ&3BOJna!jQ?(p(+Rs(Nl<2lS`sNtI%tlr2bTHX zn3Li}e;U<9lyzRs5!A|vC^Stq;m#d zXQW?Ejyb>#$uguF(ssA}@r%gDe1wxYXe%lv@evcc)$0p?`5(&;H8y^S6QjaP@7e)n zDHoA}e=&hG4>gJEz)`>R1%kl55 zdl_Rmo79ItYD=*!D{-aCS9m6iPyl5r ze*GiUYb!WRR$$~dhyP^-j%isH#ecSBFnmE)7Ld0cwUyaU$sy;L39@W%nt&!rfjEhW z2#r@&^i)51%r)Kt;O}mzoLyhY;D*n1m6EFZR5Zcg_%F{|r>MozV#V8w6 z9@rkP3CMQEcA7%&3Xb}rdjW%2X^_1UJX=hbmqq6)&Ud!v5Z4tkecGyOPX$R%5@mz@MomjQXIYjxm|%Ua2) zsLdcPuqL`Y>fD?mm*sJ#@=6cYZ!ZMppvgi1^h`IhkCvn;8(*?_}|gYs4w#y9&^g?G{`A~S-o(u>J`gLLZvK}V&I zgGcYsnTmt_#pDj0nX;A0cH3ceCd_`|4t&#l?#r2z3A;KmJgkIWF~8p=ZKRuo6=6;M zCS)!kS>l`CP+i3B6=nrCg&W`3N{-O@3d6G+g${&i9S;_^C~@@8Xasm~e=T(RuxVBM z?zQi{L>{@Y@?D&`LM_8=&$EV-ucF9uD*hT-E$igucmOe&Zi7ZZTO=yzCi| zvnxF8xXH1;mTcO5Jw;&(w#6O)(Lu6}U9ioG7amZo8Wn6SrR2pF*-6E>D0=CX=ziut zy^hHWs1d5yfag&izQ(6nx=DnI;7-n=5RB&QqJht!zgDiUR&JN0ij`i{CQ)OBEDEV* zN7l-g0~H`p<|C=eQmMwno`wc(gWb?%M?-@{xm^Bnge7H((?-v%kdQ3pw7Ycz-=kiZ z9t+)hIUc2)>)u-W;_LRF#POfKjnh2-Kz6>uy3rUTLp<&E)BHN2$&jqta`k6qEjvTv z#QU6`W`<-dB?p&w6BVD$!BBYdg00>S+a-zNj@-75 z^NvpyGl7k08s((e88%ji3+`F{cBBa(vj0(UB_%HmK2Dh7qlS`$k6Q_XR(uSF9272_ zzFe9POtVRxo59PcCxw^A^uNAwc4f>Z4iXwKl@7R<2ObMN98%3s<{eyc21Ht0<;6gm z*1{R$tqZ8|xfS&Y)|qqUQSAFK7keeT?vmVl70(s`C24QOpd_7hT39IPmkoN_$9Azt z&?qpC{n831*d6xzQpB?7ec8b#^FjfJ-9hP@cbKOCfsG;ook`-8}A0)oBcQS=aPQ2N%Qm9lJwNtrRV31_(Wd2$tQ5zlI z#e!B2q>6#%M6fg{8P*ZJuF9>RR~A|sRT;HCAd$B}*gmU?)#;u%q^u5UmlNMqa^N=7GHGlq=h+^@0cvX0@@Q!<~s*^$0=6B!4e|NyTWsnEL0{c>MQqVv_ z9rfq$zy;@`@E>5}uL^I2EixvQ-^Cm<@}*-*+z7?vDUtyJx|?q5eV62}Vfo##ix$(nfm7{IBZ3X{!>{hw}ZbdBoE{> z0RuNNHh=*PB zR9FRmRjh^UE(i~*QUg%i)xg@c!=nyJpt9&ndX4-px0ddK$-gnTFJ!Vs;tsTFwwNxf z*I2h{AB$i5@_Q!xb@AVp%_Vo)*)J!K@2oeoUrVRS6Q_|OsP+XJ2CT|Vr0zk*%W~MO zpClcmB=%GJiLfR~mm5@TE~xS=TQJDQ#c-qVGM`i2yWH-G1VMpnAUa!|AvsPS^EL+E z0Txrxm!Xmt=7xHS$XUGe@-*VV5q4ao{h3cmo^^!4P7H;#a`W6cYCgY&>z!D1hk71SD6 zD;bjE_t5j<8~ApMN@A;c7hg|TB~n|Jt#RF=RMke`6kj}23U|VhF#P`It{eTJ?$e{0 z>$JA)$G1M(Jl6!npPZSui#E*RLs2V|SxHuqHb%>WBcodq=t04zum7o-|GVZQqebA6g zaZEmUyQ+guon6e$;GBRGjRA2#_c%cNfU;%YadJLRWo8d;SeJ?&S%BO#>l~CkJ>-`5@z2v(0-MN>8Zu+QEzp7O=G}l;zp}rek1uAV-ps?f_z_ zqk1cW6B+4^*Sz1J-ZkstUj-PGg#- z%Xk^E_VvD+M?VCgyAZaD|Jup+-8*)D@LLmXe%JhyQ1XcrhZa_wc}0&X`KJ`QPsJlE z6q*nq#cknK_!MzBy{zsBXE_Pr&@HP3snDAdZ9(dzq}01Qygmr2QZ9pRE zb$5NN6z!2SU;5}Z6a3n~^ie)3e_?#5(`NRpo|4y5q#F7(!z(=ogvX?vvL~;Q|1%B5 z9e0V*FBMV+Y_UI@BF*ivv=ko8QRi_D8qGwSW0W*j+c& zLav-&9`xxum&M&Oog^!y;pj{oSy@rG3>9AuidK5ML78Ntyc>3Tupl8enU}CI zAy%VEh|LYeqOoS*9=KKA0#35gwa!(c|C|tu{|9CQ&taEpSpCu%BrEA>4$jpm(%iHX zkk4vWw1E03u1>=)_)i9qbZ+(TrjvO_#Zhp_P7>_?V~NU10d2e(y4iVlz_SjtA^wI> zEL*Xhw)V_Qb5#dPnAPCE7yO=!ub%`p>oeY{k9RY;nS%|#SfILHa>u=9Hp|Im_lV7} zW(NfGdmZ`CIuOQiaHuqo>49G{IXF76M?WHs(}~{9GdNGlTPV^@#h(MOZ>{`h1iFh@ z@Of9J_v-{^)F!%)9)Nx90p%{zCr*p(gigpocT8=e-uC8F4cr z!FT^F8U>29Hid7B0p?uMb??)3HV=Z*3Upn2IMCxD#q(-oq`4JN6p9uLbP{wfQ9%MZ zhqfq&p)$1@NO+Stx4m@`mu55n>?Y9cr^)^eiy3D%fzG=YEdRl4K5iyt{Y8E54`ihi z@7(guAitTCr%_}*6_1&-#R8DHSoelDv{-i9bN81oF31YT!m&2_b|8XAjZik2V*!IP zFJ*<&uTS0PX5pei`iP zQ|E8zQ~^Cj2i+ZkS79^YW^PqfmiLLMeEKre;kGI^QLsN|HwX_Wgl!4PgCdmMt~=$MND>mUYfff?R z=oY014oHp*jtds4le~0Zz4SE(r3MsIQ&2kP!wR$f~vaNbYfjZX2X4lvef8O7Fa zMeCy0%CUhn)ldar<%U`Tx0z1nF)xXSxyftH0MtJ$CJC_^Jh>dSMFIIE`=*edJhn+b zf5Xf?8O?sLtUv!ToCN3HKM8~0Nw);c3)0YT1@lKgciSyDs9V}N$GDdCkrljZ;h_-l zgp7n2unsQcZR1u83q@<(_KB0ZMVw5pwR4Oi0=TMB&?ean?2_!_0Bn#qY0R_vFc#O| z`NEIiTwwBlzJB<-yUD@nz>+t*IOao2ev~2&RQ!^#3=Y<5b(2*wM@hMCK-?F28RYi# zl8PAM37{%s(#T5rF3(Ep${dgmj!FtIRprwi1Us)rj-X{a#OV1u==9Jk_j2zI;W}l7 z61nm6!EWrnLtraO)?HfG=*`k~H-Kr%}#Lx-*{DnENp zRrttzjp7Jh9exQ|Kf584J7Eae_Uo89P;482qF22jKneQU`(L!|{9v<4Iqy$^Sm_hl zE2;N8?N=`@i9N=o&CcZhz8!L#|804+1v1@;_fM2Pq${rMi`N#k#$5`5Y78ibZHd_G zSxXN{x}X>~b0L(-&bchP1F9~4-g^lFN*+4cbXp$Wau9>|@)D@@wQOAE>UwAgI zk;ac&EVwNG$Op$cD9i}n2D_`I@CvsF5Lm}=#PZiP@n6R{_VcOJ?0zQ&Za@2zCE2JI zPETVC4c6KLZ%wINshe7TKv68zl3u!vDHPb3%VwPa3U-fkoWbmPnw7C}-i>Mm@6BOf zlar%=J>m-4!mb#{dH>6x_kFalwv&=W9yE`N&r!B=)tgl(eYGl$V)gtC_lvLZA?hl( zr9nkO8pX+1nq+r)Rc@XB4Z!Jy-29jZRo{O#9m>k!#Yx|jkQ&8&c&DOS>L{m}9q6Aj z40ynfP=7MihuudqxhAx@1iLII%U&3?);V#gc$e8sXHoJ@ zilkHVD5Zc6T*w=V73H|~1yw-gu?8Kise8rl!4qdK zOOjRG|GLCvGlX}9zah!&))FV)<^aRns1~YhN)AZRpyEsBACi}b82{Z8_us8?Sd5bu zaB!9un)pEMcy1rvhi!7I*0>C+XTgevJ<`T0QHrt!=}%0H>D3efUVQ7lw_HsSX?(5r zA7rT$FDu*4COwssr%)t`ipK;YE+pNuWLM1Hqz33A;m|?f@GKOt+*wZkiQ78Slm8@k zBd{}RwI2k0Bgh1c-ihh^XvqxdtZO3>gpNkdXIy_+*pw80 zJ)#kE0omY@*32m4rE<1<_tED`DR)Swg=3g;=#<@)9(5cIlf~4q#)rlIj(W=uAJg8L z-s)|#DlI>*c$cKVF#gD1Gc*-YawyHrL1w^w?mDjajVqA3IxD?Jp@VBCS<`+W1) zCak=%>##5Rm>pJ}IBK@itdMGul0TrxU7&J;R*x+L*p|_`uh&7~4EDnBdz1hn#jr~) zzY%De2Hl~s965|T=mXHe*h_bD`=Lp@Q=!i0{Ns#h*ah>?2JWV--UZC3lKjYF7p#+N zme!MQ8k#sq#=b7w71c}Qsf^%Ux?2XyFF+lYr>gS9;yvxGrBZb^FPk?2$x6I3AO6?M4~Z{DCvU>b%J6Yw#yDnff9kuI9r{v)2Z=m82e*22} za?EA#j;OYo^)s7AJLp3H_c@2ae6+>^Enz^VsBPv(cQ(J~i9c!Movd&$*$-oJzjYjp zQSe8_{_!U{ma#mi-CJ3S&K;3e(?*H*Y*;#w@wQl*smvBvxn1Wxp3~r;E&hbR#`FD` zwekzXMjChm#m!Sgrp^i?CSQZaOi%byb`bf)ulE0kWxM_hQq`g}Gd6jY972IvRDAUp zaASoV^_saz3fSOZKBrWaEvR&_a_d%IW-8q~Wl5rJ4w|4Uw`>m5tL>ERgv8<@<`LVf z6V|gh5mXt##rl)j-LN9)Piy|Qgz6FmRWr|X`sh_mz0WaKmD?#%uXMNaG;C-6@2twQemuju`d=%MQM zhY=c8N^~`y8wk5*V>%&=uJXf%AYC+GqDBS??B?#Kp-%41<2+7V z=6IYT9CDhKGxE}gnqHkHah??}y7Z8GZX<@X=AELCDC>b-^&C)__R)o3ydq8jTGrYH z_hkC0l%U<-df5ZnJ=s~9na5zqTjQ|&Kbbnq$G+nZKHi7Nq2>(Y5fEC9(%HX{%a++0 z{d2SS+TK&dYIdAAvH#<<+mBdEzp(LPowqTHSkWh|c&AOW)%$?*q^W*2i$fnV%?SZ(G(?JMHRXrQjPS=#V`I3Cf`%4B6~4WQHBgQ@+aB zI?s-n%V&+r%7#t*^(pUqo+(-un*Qx0WI4MGkrSJWT(j-&MoPYcBFR*|fnVVolP$vZ z3-%5?bzn-NWGjqt34GSH+xy%He+c=UC4~>0HN<&WM`DHWr})d4O>qz)XN|fG@r_ly zCT>!AqTmQ!E31Qc(}N-Z4;p|o=4-#X_@>EbEZgv{?~|Gr#%8pe*^E<^{6mTyrQ&r! zZzm7!v?QSih_ z-;u;wEikn#xA68t9d|F?2-JrLncNZ1LL7FdV7nY*Zr%{+l~n_q?{)ft2at9|s*zGW z3HHc%5qmJQF*eqGZX1rXwncBwOZxfqPE$DG-W(2Gl~6R=L>wRb}d z-kvn@W6aqktil+h+3yfLn7HdVx?0x7St%&XRPGC@q#k~5BX{wP6oL6W?A}Gj>n!>_ z=SEnYyntIStM^Ib)cG4r88uuzke%mN>{pdGNs@b_Qe(=97p}}UK?8GIuR)B=roifg4PFdJ7*dyQc z-6L1$D3?bz$;$oT>tuSPP+Pewdd#c)C*2d95A+m7`JBX*rY} zqVHRAN$dqiK&+e2jVu;sDtd*dLD)!Lz$x?A(~We2_}#_9dOu{L+ZR>~i(jiEjVTZ( z^LoL%b+9Wt2?;C*SdvjgKEtTIy6Vf(CST;8y*smsj$Jgxi46#-E04m@2b8>)B2`p; zv$R*3CDJpE;8tacs)fsV`Y4?@=r#;`WN_}uPKrA`%9tBrTOw;AV+m?Wdb-rRP_RkR z2fQvFe7qo$vp=Sm#!J*4eEfrV)PfvRdLX`oP1WqR9_mtWRMkRi)jfbFK@kxw5uB3@qo^1BCo zt^;#tC#T&Fe1UKj$3V)~er~z{-M~Z23*O+}Dv81_hFtfqr25HTZYcvq zh@uOe4*o^D-e-``2EyglU&{J!6cqa39(50f4A$3r*R-md@cE^p}Tg^}A z?VH^+{|uc$9{2*;n_;YEQ3>}$AxeOrlk`Eebg}HQkG-HZG5kDZb+QwL4!xf9t;tBp zIa>>gYpZd!h9ZqPr>4<$+sgFO{D6!c@VBiY_-pJh;MV(ESvM z!RAM5!Mm=4ypTcU5n}I=4IY6)~u6ex(W9k|PoxJ5h9s}G5x409 z@$IRql41|1@ql2T^U+`bc)^!V00~((=X<1q9YCBoR?}$)kaLv$3`Lr#cs;bd>xg=z zM-vTx?R@NiUlP18tUxs+ONkkB0oGYRU9^@yE+2NmYf#Y*#2sK0SozjU3gxL$kr+$v zs-R{5s&}d=+x5wvZIX?Vdz8a2?ZOUzl1K}B*Lc4U_-Ksn1y)cDZ%p9ca7?{9ekrd*^Qxi*iCpO^AI{{kf)BojSV^s z*4H6eczVzm1Rb0GJ}eWM(D8fSj6ah#>{b~k-bO>Q#VC+xQu1^REXMDkaADcvsf$wQ zc$9LkdmDL++aIUhBU{WtG5hgNG-V&lzie$2i7*%E7#iF^;_t zu`_JVoK1X7T6Cuch!vvq4Kv!^&dx86G9aQ^y3OyTxO5g4OB(63rqmf?Q(~QgT!=*!WOzXC-v-!{ zXzA18cfvG^Azobomhrd7wF~#hncq>uuK#k5$qlAmPGd z{Q?f&SHw#Qs}won@N9L6YnIP+1+-Kv5#o41PQm%xVz|-@M8)c+`9eA8L zrqXBzo=QqyK@lAlUkMsBZIUdJLF;lucouyka4;CQy~quCUziFKs#@^5)%#~HgC3AJ z$qK;@24wC@Pt*bEs$ML}kiz~qD;R5NPT>obW5qZ>MqNgz13}X{wh{^!>}1aDj2*kJ zZNEFY=X;hOixtP9Y5s9e2mBcEDcy>yn!N`)*R$b)SIyoTqn9=K7jW_D4%hQEuqMJ; zJ3YDaehBQJ?06&E=RB(gz&>QR`on(*fHm%Mf*TBHHhvQ*9sIG?to4vXss=~WvNuX3XazdkM3U3d8 z4{w5YjF*o|93Y#*9x-DYar?kz)F`YF&yPb1Q~%~!?(!!7X~u-%plWE$svcZ~rUmqR0mvJhX;;>d6p#S$ZspqK8-csjqwQ6fDZ{s0q0b93WYsNP(2cg#vuF z2$;XI8=;88f1dUR?S`~5z{ldyQ#hUX_gUG$Ga+r|s-*d(mmSia*X)w_%_!R+Q}RKI zJfPw+%X~=r_aT>mtb<}?qomo=pkz?;K1xcV2R4iD1+U+jr^bSA)HbW7>o}coHHHF* zT~fTt{d?$bbRIYPWi=kw^SfnPbVBSU(h0Y>E?C!owj&0Ih2JuAj~yEATJ}ZRDYY z?bwr>j>!^YwX?Ldz!j9~gwW6p8-88%k~zaJjgZSs6;%TN_JyzqbPZ?F1CyOe9CZ&g z_G#&z0;uhwKj5~)w$0f1j|uhcNErPS*=^s)Ob3J*(ikbVgWC>Wg;)i1m#&VgmF%E*LKDdal0`ot8-dFdCG5*&cloVxpsUQe7O_)&g6JiA z3#;jE+^X5FaRsXTBw28m9*Djq+{V=?4llr?OL={ykc(F~nXY_#t(o||ut3B3i_XNq z##48)+6ho`pC5khdDlln9W!23<0r}YdQ`t`O zDSA_kIu~rqhYBdhVCp!?{(|h;Clp0sJM^?Zc-W;yaaxFMK#=0R0x8Y}aCx#hDbWQ% ziz-!@W16J{l0_QDak`$TKE*8$yc>NtdQZ&JKs!K<1%AdiE1z94d1SG}&+(pJgD;z0 zWfzfZFUfOaVwexi@Uw@K15->172nUSl~;#%gWLxYCDR3}8$77y&k_v)Cvq?S&l|j9 z7xZAO>A&6Jp#_4y6)srNqX)|v*D3kwBzChdv+UG&+Lg#kS3X(*5<8U4bYJ^Nmyx*fr@#0sNtix{$cz(5hc#wv zP%S8z2|AsZO2_|FDq(hHdhslNe(yL+Q^NGw57VI!dt`9y4}FXuo%3d}<-hpzFFi~M`AIujL{g@M-!~c|+bB8w zFtQ13M$NC^$vEsI4&J7_19wUbP?`6rWPvdT>2OgiZ#6(`-D9^7{X6&KWEUqzI~ z1(j76TU!Mi6vYj!5|_qSP()PtpHo%XR7q*Qf?`TDncsB1<=*@7-S3@y?peMgnL8KH z3+SwX2G=%;<3ClKKFlZZcycf}l@9xTNig!myQglx=o?`pt1TMZjagx^A;H?H2KrNX zFp^+)>(v>Z>^{L=_bW82uVrrWQ&OS2Du&Nt1@S(QZIkhS{FHdI7&u%so8>pdvT58M zX$-Y6A;2S9s#PrxkMq&nMSuG=4($Z?6NC?Y@bqJNU*QGpSMElg^Lf#!{qOsJ>~qik zkmRWP#&kW9C43CF+IsqgVCi(ni#N;Zu@M!P-5h?-3l-FXNqj_c4=B=u#AIu@JSlQcoU`eAP{Z){ovFt~nyZ3?!8_g7dupmdh06}Q>kh8O z(AcSYvg1>aLuf<=!Jc@2COFF#qCKH+rh2>gU>WKqA;l3e05Mc>H3ozlX~FeM3}{DZ zCT2F^HhwuZ4|#v_5Cf00>D#V}Y^r)-dY|WQktRKSO_XK@yJhlV2d8x;j{95BiLaXA z_fB)t_sQYmc1zByxIqxaP%Z5fiff?AaVlozoJ+ET*;m+F-z7c`$~(eT_uj}V<|0Xw zCk5T1%cRT358kO_ZiX!%-x7UIwjdI(VNtwF<&r7=F4*rj$qGmgI_1|ZN8P`U7<|@D zNpLmEp|6JB3u}q4VqmK-xEXc`q+XoQ0V9ZwFKz2*Xn zgR0v(RLsCzmt+f|8d@z>L(3BEk=4^lK|S;m&*DiNeM@8w$}0bDuG>80$r%Rfq5+@A zF8I(7Lfa*f%27{WlNHmA5Et@!EB>FHP4CIJIe69=XwzeOH?lqlFVvpAH8kWiVggK& zOUiMwY&^*{1I#9h+dz>7Hmb_C}`{9DBg4eo9*iU$sHJJY2|bw4n9Y<<7HkimO? z?eD+~8O|%76FnBQvrNbcT(oKj+0RdBc4Bqrv>CvUP+ToVsv-I=^mOQ6Wg3^`v0v3k zw@oUgZul}4RHZ4lZv#fH&2N6yzTyo41kwkHuD6)ZyDO4_ycN6q% ztoGU;lENK{`V_XuEz$VjCF2X|(~=94D)~}7b}n{Lhb`dvToTy*7~UNZfAQM!R|i*~ z-Cp75Xqe z81vDvbx2j>ugRhNJT8apSa2A~1|l@Oyc!K<(^CDaS&Z68z*N!+QAi@zCs-GnB-iNa zRUYM`4T@3-?$-iYs$)n%MhF)JYHWo7pLYD<@`EqEpJ=IwJk3!4;;9?y0=@Er+Y%uZ zmQc)x#Cf0qazMK7^XJ7Tlstc!QEy=#CCo1-y3R4JHPh#R(oV|8lU6g4ew^ZtQly@W zIYHuu8-3T2i_(KpdQpS?s#{C+M&EO9>XiEeYlPRP*VEU7lN2q{7~61ddaew}Mx~E^ zROK zkPPg}I6^%{MxOGkk+m>g-n)ps<`$c#=qYi}wD!{_)YwoWX2 ztOVO-stcJj3^G%7MAW&)3$y9#(*|6SiLX8g3A!JC@z|v9$yNSPRtvMhbS!2%V)_?P zzEgW}WiQz6c^02OZp_77mRTS#)H;-ACL%XbTmnVnu%I|bCk%4lqG52EaC>ld2=0OH z6WC zr=^Dzn@Fr|gz};q6f2^P|80rhL-qt-lr0vnhyrplwoAC2+2~#tx&eZ?G3jHYJ6=zn zv_lpbn&)}eyCr%Hqhso6z34u{h$@Fc!+ua0Vd5xgSeGg_fB40@Uz=3iCEFKHBA56X z@|{?!JT_Z~KBKr#DRP~PIR&oD8gNzKQef(kR)v7brsIY0m(y3+YUvqSo-|4Rk@x_8 zH}b&5M1iJUpjCI$A1)AU@Fu1^Z&g)GZ;}C*t!&q1ttu^`bSBoiKjB(5)q(;CV_lM% zJklv?WgmHV(t0Hh(MzBh8y;lHvglR_wmHMZh15U>zTt64mE|-Zw2Myo76Hv34xdS9 zOY3ReEEu@caM-JCwQ!J`1b1{#+k-}m#is?A-B-+DT{G>Ur~GnoQ?)?foY1)ggxK(;CiqNWAN9t zYw+2z;-&65LLDiQ82dIyTN9rD>v%Im+2<#lMM)1|^z27H-N-c4aiQxywdzJtp*0z@ zG9YhDh%yQflngrd4qZe#gRX*WXha$)kvb&!=$|j;ybO!&niWa_jcS7FNnNS^+?@8 zo0hVk?*|+s3!T^(1vQC>1k|mgxV02nO~s^yUy!T`*D=*F_2^wF6JnK`L`QF*hMPj$ zKrviT$6n(Fig7=GRC(k@2L&2xP+}M(68goqSxx|t-! zdz#aKCd>ILy-vK%$}*EHn<*}lBI}@ftSrm5Q=$c`4=iA9D0h^?q=fJFNuE$mDnkmH zItc$CleNYKHQRt>+T+*^&yhDga`;PcTXH))t)2lC!b293C*}3h0`|*tnRRaUWQVw( z?wi#Tohj9^y^&|8J(|_;k}qHAvBt08B`Kgl(l9AGD2^GUdFM6dhM6DvblNfkuLXs= zp7N#B|5!|OXxRngcu-d{wJ}!Q2^vY0s!6p=kOG^ZF#^w%iL@C-PkM~a!QJ*CFH1$^ zX-lsY`wXpUq{hvOrqH`xgsvf=MBNJ|ZcYJz<-qkV8`tioF-SJ*HTT< zWueW{{VwrvR7|6xtxpgqJi~O+%bDe#=iWS_&VszzV!A98;-g$UqKhquNHBEl65=51 z48!3u#AS+bkW=EDvZd44k>#!<0j8}Ov&;JWIb?Vry&raQ!2}bMuKo7H-^mIm-rQ!J zVPgx$fd;i6Z2feKuu*{mq{U~kA|6=6ddV6tnQl_W&N`yZrsJfVq@ed=$B+_==WO$y zHG4Z)mng&8#(w%Qr~dMylbl}V3LuJhGyE56e890X@P$KAo#wF5{)r34xiYLauSlk^ zc;RIY@?~Rz`pv#AO06o@4~QDNqD~kRnaf3MI6ctN^+wi9_fISrZ1zRF{609orO2K# zn(juP*m=Iduoo;A{rF+WrG*t)mhOB>;ux${AVqbyMcoh96a&8)#&jdyN8;oLEZhgK z90>hJMYlMj!$ASM783@i{IT@(7>>B|t*f(KENOt8b~kK=pQurB(-SN08L3`CvNS4I zLsf8e;bhY*6gvf-5l}4vFO|}tNS6~h$)FbkvDiq|2%9E@|EwE@h)si!pbV3}fn8z7^74@4dMR>`ia~ailvfKR ziR@9Yb*@D-`jrMN?3K5nLQR(}F+4+kASg@Fp&oESJ{aU=DGpwb+%I=Tm~)EcxHml$ zr(OjKv2pAAajOdbT|_BH;%O14P8#LV$DW39H5+D%XYFD-168Dsz&VfsSuv@ z(VPe7_|={jA^0!Ss<3mOXBPb3>qDU!RnHqK;rFow!6>260*oU+O^@uG^Zb}zC*FKp ziRn#YGSrvYA|zmef^xULkzl1na4w`$x!TL&tC{78ZtLV*_H*>P_0=uWZ_O=+0LAkB5fRU}pF%txih7FAG`LmK^FC&w#9^y?5^dOi6@&E*P zncx3{br)W+O%p!`6K(kj>muW`Z)_^Gof)@x60W@ODD1wU+oQ$!24cb{_2C*4+u_6 z9?J^78|gB)i?WmAY~a^hIcv4o(5s7r#^@c}`er{D!DH}R;Stuq#9<7h&buD#?{~bb zGHDvo$9{2$r1H}=oOt_FVJ0>TDGq`mcTh1#-pI?|+r1xo0^3R?rXFhA*j{p;WKTH& zA^o|K5z;MK!*#-rL$6%STyWED5OmOOaJvfjD25^;<-uv3aVL|@qzG~U1XM7MOyrRIflFk4wrl7hBT+r81+3Zc=3-Z1ZlOe@z#4#8G)+K&S>BI?rKzC;ZgFI1ST=o@sg(CCn}DR`A&qIB11`vm4tAd| z#UW9jU~E23Uaxjg_Kx;@9A-T0pW`rEi1VuWM;9HPY>5!3_3l`yPlDB%FG#?oiNQ&j zc~hu7Njhn?pJKJ;H6Dh@rQFF`;ZS#KmvZ0hSf^|ZmKD38I|kx`A9O>n_`6vKJ31{w ztQ5*UsyxpiOZgKE2l8L+6Re$*D@z99HU?N2DnV?=#n4`kKjw#vL;HTYImonssefx- zDam+g_Ak|D`_aV{(O210 zz?jj-^+tAx^8)dHJ=6+Hd$r0RnKaOkr&ZPY>sY)pY5*Aaa~^@r7vj65LVw-Cacaz z)P`i!tHWEOTLU3?OIRsviLMQ~O*R0V%|hYo5v5J=di8ukz-#akG`7CfoR0f=rMC$l zO~2YZk8E*b1!SKY6mlsJa)Q&Tm@AO)yZqJWnVJisho+qLK~Ao1s^h9Obj0K^w@Ep( z7MPdn=>l-W6woJquoU|>MGb(YTX{{99dO$tb}DvtoEpLx7iKcTfwooaxEM|a`OP10 zUjN5ZLG%WjyFvA6m8T0}2L#0Fun&%%RWcc(07vje+R@c^5ZN*K2)cZj{q1VY5@0Wh z3Sy-=7?MS{%CF9V4Q;Dj8{Xm3s7Qo}(=o;aPMK|(7{TP)+4aa9-^w69cw%`#$@fZp*+_r!~V{2{la-md;!hv>zz14uvcY zub8ndd?A#L*EITGqvz?^GQTDun$ChicdZKT^Ny;GVEMM~nQ_Z89C6M|;3Q{K(0E=TWoz7%Wz5*2- zx~H^6V>nv7Nax=@MXUP6f1O7bouY1r%;RmIEzwO3{%&E6PrV>16Jp_)xX{5T+>Klc zyQMs7u38tdhC8S%^#b<%na3ve(fI5lZ>&nu>{B;fJqVisHn`F_?DuK3SqIo~Tc5Mr z{u*T&jPyeJcdUdQp}(d#tSt@ghR0d_GRo)Y{<~2=&XfdSqjjdYC3oTDis>7~W zN6QCa%YTD4iK9&kA!o@(ewwcH2GNibGfg*(;vfzog^EF4*68!il^JU-q>^<)V;$Be zCL#2pAQ4z%%lwk1=O#4-j0a#@c**Iwz-KaobVjM*e!7M0*uwq>z zdt6m%;3e*~yCAClGR92cJnO_7yq)EEctR?5>s5g8JbEdZZAcOshFctopkeC zz3z~{-RBVxr3y-;&O~(5`B2JiyXSyQ^IP4Kr8AB2!Tbu)aI$IOHdh>1rMNB-Z;Gl5 zj+187RZK@%9J5S((~yHc;4YIuo(T#YZ?1Xm(!W~P z{c+m$$x0=k21$2hGE}>`4Ps!U?`d+0ErE>x>#yGj334c67265dKKAMK#iA|sbV6j# z%uZje>SLd+?#soaGP%!s*Y-~Qv&I~ad%PAb>gPqf_q^zVvwC_*aO^Cj8iOp5d154r zC}dX7$foPY&RcAnJezSb8k5aq3|_bxC#cB(u_Z2ElBLH=__ZO#WQS~*5_FQpX{(^t zGAt1lllnX=K|?+-xi7#a!}PyhvmGs7Y(|l#(!m1}_B8GNTlcTNX;N^Dy&nC7)bmqt zoj7JzZ>Ex*p|~cBG(tS23<%hO)TSwP0dtQ|lLOlgQp@i3!HgqJ));U>rL~6? zd#_aLF#9NxZ8FppOP3uCOog&KS{3Gmf?A_vfs`1As)PvI!-8xkWAcDYrMlMB$ojR1 zbhFs?ZlsRIR`4L~hx>0-XF%K0`?bpZ+z$p90y~@@vQBdu?9IchV>hxak!It$ttg{< zR|LQ4v9sG;tAwkCeLx@DOY#J5VY}v}(Ff*S`dS=QO`4&MZ#^?We~?CB{--#mC3+W> z;z3c4 zp#(d8#)HO~JnWdum!Km9y<@IuNfKfTfO zSA}V{I{NmvE|PUG&1$vVY_&?GI0$apM#bpqZWfIzdEm3Y8+q3q77$}*1~@0nL;F1P z#ON1*tH>OhO_zsuKwtyZ)ZL{#@rHwS8v7x|10QxY{V=d!Z|Rigqv|@bN~1LcNj$}^ zp~y<$J_mPoQBZjlmY0fQmO9#&N;LyBQmT`w; zmlc1_w6fIw=KQCmX1F+fCl(FuW}=~q;uGcJ;0XVsVEbaxnW749dse6 z&-x&Z&U;}{JB@kHxHZ;jJs?ZfqkA;g7*m^KnRDi(m(BJwT^|vH}_LRBn{){rivy_UT8_B{zAo1Ss_e6 zD~Tu7lh4BLDUt2}*528D>^k8VqS+Y{FYKftc&_W$sa4#Pf<(ArsIW!Z1$Q``aCI)#OySXBe}ApAcpd(5#Z7$kPKxiK*|ExxKH!7^V->0`&4YrpkY>Pba

FJ~g9WaR#9RKq9z4IsH!NG5wC#9|P?tJsVaVbu1T856se7K$zIkB4CV75M1 zQyj3?9-?9z6=iOXihQPr+@^66ETs#SnmYeDAqGFdzGJg*HZb~B5KV1JKI~9R1Lt)q zv;k|vtApdnwFnLBF*4;VE&u~yT1bK5@nc`aokt${vppYf#4;v8Cyv$SYn_t zUgNdIr#rF}s#`Ok_VonUf!r^r*l1f<=B3}r42$M z&P`2`WvVx*Ya@5Ucz7Kp5QJ(?T&x&y(SxHCU7(|4e&z#<=ZKOaw?~c|UgG7Ie@fc= z6HER%E0k_%hi!@88qh^ogk134IH^ITgXobg2stTH*7y`E56x_ePGRzytAb{_TY1vw zlq6lS)hj``L7f$#7d?hNuoUSLu!7u%0+qNeYNs(qW#8+1`GwF1!DxTdqc%MwpAoYW zcukM7p;Oy%k8t%#)ezjDxW7hHl7@s zuDJzE(#A<0YQsUH_o-0Lns7rbXIcQRP7a5};7sy`#$Z23|MG&yE1lnv|Jc&9>$K3Y zV#VDT+(nmz=X)N8VP2D^xTQiqGp5F7(i+SkDe+H}x6^B99iQ6mrYRAeqKhT5;(R)L zqNZ4q5Y+=@bBEQtX2HH+r)rh&eoYfA-Ve&n`p{K@=o=bUK=B355S`?y!_fp1FVKu* zrhF-Ql4(o5_+8)CB>4q%hL@Rbvv*M(Sd4P1m;(_hvTpDVWvaJ=maDn1Eb-T>+C|xP ziuAq`-HN~gGo=aq;hJPRd203)EIAM_%oeQo)GP_fgsqW|ZDo!8%L6VL?ujFeDt?S} zJ~}8GP5hYtE~{JO>ZqcxSni*lR?Jx05o3;7ul#%@Dk0r0;GgOmf$gJ?yd^JL@(S7f;a^4)vgy1|3t}Z?6ZgItk_iYx#-Vno>D?!Y7oK4> zHQ_bk#t^RTDXUEx0yVPgV5D2l3a<`vfG9DaZa$D2x*nO2GWutI{p8=~niibd6|=%g zn-kk`2h0|iI}~?|A~(^7n<-u`*u-3t{m)J}Lp}F)Q7#x;uez>t&4Og-K0)#9?cNzO zJ;@3`CF!J%2_l9J5@hbrrXTy9^_kxa{>r^RS>ams4snrpyGM&^vErcslo@)JTM2!` z`!QQ!FyVk*7ay0QPL@K%U-lF<15`}95UQs!y{X+}g#seFFx90@{L$oG@oKkIlA}t4 zRMRo_^bHoCpn|H(jfxgP>IvTyWCbv8&I1B@RU+hWcf`qI2LXI8tB@0%5d& zUpacO?5ig6^iFfq_sL-=7Ef(vlm7|DfzRnUiYJVjgJSC6xx-{n*+j+7(a+I->*P0& zC^IJC^*gTWV>K8zjHC?Z^qsFklnmN^4Yl!_7)+wv8PN-|8%?TDT>%c*enS9`cmAL6 zd`+u@=1AtSg=eC@e6MUpN z;U#={fQgp?8kb*}{`j8;7rjzZ2dNrvm*u?1Npi^yX^j+jf+EMLm_4#SIx~D>K;P8Q zj}qB3_f+Yb;6A|MM)xLQN81`u$i|BcrI1`F%ZkX2I25!%-4cCYwtk|*BiULw=6S3NI^R}0Sw z+uhbgr%GF*JLg=I#WANz!EDXJU)!NvjSbFt0L;|XI?mRp)@dy*fqWJGajK%NI>!S+> zOrd7~()RsvLKBM3+voqq{QG7~MUDQ*Qt zmQpd=Sw&L~w&3iEtKS~eSsn84dhP@dX*-sO)}_L5+G|ePmMMXz)#nepYSxl;C)R1I z%-~Q=aj^B?NyV%cEc3zdzK6a7KGU@US@c#mfmy|DoYWMZPVOmtfmfz+PB!yMdRM+c zoT5Ux@X;4@L-065T@TKv^AVc`5P}!kCs-45ecIjWSgPcdzaw(CpNksXl#)>g6fe|J z(QjXSU`ZzLv_b?jXout!Bv4!&MOL9EjIl46Ad)XP_zx1% zwGA&!jQjcWKMOvwt&L};-TcBfenJl`_ zzi8@JU#K+)-y6u`Qtqa;3a9nkd&Q&J? zM{l#}+Vqxa9ow(00qp>^R5%jHoFTVe~>Hfq))k|Q_>QBi7jGIN{b<) zvs{8n5u08+LKi{%ME3Km<&q;&TGdkFW_fi;nVVLe>K@0O0?B6!7HpW7Crw5TH%~?! zor8TBKYY{AB%9Q0R#cGfZ@JZ29h1s7H_vUG||p~!YB<`~%-frecT z7JbD~PQ4r*SccOsS)w8IX1SZ@BT@)~pcsI$Az%$xA}RC3_KraK5Lnp{_925f;7~(C zC%$e%$(**_)g;e}_fA0TJ7nGuP#lzLE1_Z_8iuJ3+2{uHZ6TLTcc`z`ND2CWx{`joN%-}2KK|jswOfj;NT_u#`S+WeVL_0+=?d|P4W0K zerCvwyeio4-AQ9k6JEnqr_aa=HGYO#RfS|naB(D3N-Yat7VhBB5?`8qb1^Ef`H26iiJfuW>K*b|Agz&h0kvXY*O5yMUU0BjN@&mZ3I7W@GG-OXP~ z`;uuTTC)DTKOz<5$ptfkf0E)prpQNB%*U!lQ~THsQK@@7K@yB!(h|Kxh8x8?vSo4= zbMCv>$o@%vY#*EL_Sal0?mLU9yYBmDErz4NUi|sq*=DEup{J+j47r#57Mw(y`c-~u z$epB^At%{C9z5AW{gs!z8qvgAT?mH1!DkkEO`P-EJnsB;a)TvR7N4EA^U74P7!TEi zk|}N*MYd2em}jYD7b_aUdIK?A+0tE5A*G96=2Pfh5wagD9wbEWpQ>XTlsVz4BQNR&YnC1+}&uY%hu7 z&0M);ZNNp@r=s{NSrC$iX9puPy_07>0urL*s(e9$*L_k6b&cED6e%wWXMOJMg2eEm zZB$n>ezNak*P$aXdR0k`kt<}6s*zDG&F|P3xB5UNAF@YZkZcGjWa?ZYKzvp!SLU|F z=b|iU)a`9nAi^IXRt6g7%e?S0ZmN%0v*ixyX*Lk&T|%t|1|Ju6sPD?t$tG!+qQV<= ztSwVIJub*V0m8z9J#;G9E$CE~3onGWL>t?lCwjS#S@n+_7;GIlcP_R5kT$JCJ>L&F zMi!2T>SsfR{W^+UOOe%7jE;RsABZq;CtxltQsZh>_>GQ|ROa!q*!Y3+7;$u(T7S>O z#11$w7s!VdpG~za;%22*8U{K;BA6_3f(Bz2_xhkcze%+Z64=Ka7B-?{%)@ecp~86; z|JSaOA5+>#DW5^ZozR?4}`)j zRy06;YfXCiHQ*1xfbV1)$mhkeZUzt6YR^PLnF_N9n%^pO%lU54v=!_*K{ZlU8B-N% z!1xcb;EGdpFS(^a7g!?O{1(>9$ocNPBnAfs*Yt!7Zkw6XV2l^8@;~hx4=ilOk%=t2 zwR64;G8n^zu?V+Tb&PBANu30gZK@-xDtVj1kA55khvtmiA^%Q%tS^i#X(hrtq@eD z0LwGfN70;>_ZftHy8$rskk?Z?@t{fv>>$SIi_avh;L1F#Qd%4dTxX8f*X+g!|7qvd zgC9nY?!5P@rPXDJ*;H(&IEcI6M8(uHZJ~Q)C9iK3bhD+hn;@z;%Rm0*-q}s6c2XO% zN&1sRbw z8WATQrNXh##1((+8nlULjyW1Pd0}^4^6}SPEw|S%iEqe?m7pW+8k5T)X?1R3Yv?0c zlj>mb5?|z1bM&I~^z7P*p{F}K_{xd5Gg2+9T03p5f|WYe$O*j440&lBgTdb2k=P*5O@lvapv_+KzB~P5)M8A&Ph`_Q z3Yjb3m^qDk*5}?dB-oCYLTe)$o@`2NI2aY2@RBI6{N*Qqok5uvoyQ)JPmq2mPD%rJ z@ep5P9L24o$Z{$MO&bP`43bKZy=cIhvjGsJGQoTY;RT3sZyo>Lr3ofLT>I^XzmpaG zOeM~%UqOudQ21=2I1rNSshA?~fj}q@8=e}lJ2;2FIwRe^!%lw}~Pfs2D>e z_DONJ+tnFs0vo@v!QAtnc7ufAHG_6F=C!&|v-eEbiA@sZ-Qo%VO;$Vc9v9M*hNv6cC~gZb zZZUCC8V7lLwg<;CYq*qfpns}>az$9DyeKH1Xm)#c3)XOZm1*24QE=NbFZ|(Q%cvu7 z!wVj-u&wJ;UUU-(paHxjhSDchK-`2C^d8Mx@|p&@{Y$|esHud6_lNZ5fXi6UEoHl6epG-WT@>YTK?l56J+-P!b3rB4#)H4 z#6bot%z_LaQd}=Z?m>_N5ENoHne`JoB@2V}exG=53Rx~rAQ%1bglLWtEcgjUIw3s? zYugwKcEa(XOCMPV%SE^8d6=_=#D&+GSTWY6=@z6;S|;8W4i*$+pn?vHZ(a$j0HP-> zNDI|$bnK3BEJuqE=uqc)Yt;~-&`uu+>ZEr<0B*las=6$q-v#UAltmn$+DWIW(`Ze+ za1*mYTu<)?gUnu^1L{UaGTkRwIb+^ES*$qSyG^2521NARsB#+XBg)0(l0E(>uiJ z!E0Vc7wr7l=}@>;QW}+}PJkGazjupP!c<)IE|l%{NuKan7$=;M^$_P{EsE`4x>;J4 z;j~@QMK%O1cbgw4&7mLtx??t05`hpD=7Jl#|Bbam<~K1|>lLRI-|l>4J!`bfFBE^| zI)9$wfB2{ik}7}A-eA2P?Q5I3`HiAt|4bS=FPl`^H0nF^u~|FNWVXup1fQESKh~nK z*Jy~6p<-nO$7ais^@)3-0`F!l{nJ{t31x#txcGPmoLKq+ec})`{7K2KZsqRaI~=lj zR0^}FVCl(3C>+|Y+zQ00pprMbuJPI(tZ9`aed>TqrMzF6$fA|GXhu)u8V<5VTo#HKyI&V!+2Ii`wvPK2 z%%IKtWXJn6e`%>(Sy5r&Z3DO8dso=%+4q!-rvCj3yO6r4%v68kjbULe(Hmx<-KquVa7nfQ}x%_uh7*lxjnQs&mio{%*_ zm~M7(S?Ih<;a+9qYzHiow)ife{_riYH~>4m@bXG4{XxICX)$X0)!unz%Xk80D?_As zF2waeoK46q9rDJCe1Va=8)|yEr>x8(aStv$JE_#;}Jjw$S z*=25tY@v6F|A-aAXT32#pRnajIMV6EYf7AVc~>ptyewtQOR{0IB8kvBaR`jfheAx5 zvXIpQC5@rzfcH&SgV8iMUrmz(zm(usSTb0sJJc1EH06S0^g~9Y_vBZIx^LwMGAW#_$jmR!-l=)|i%LAU`aI2+sG z00>wak{>d@p7y=!-qfkR~eE;u-7&3^+78=tK{q4ioA=N+_&49VhP>@ zPnc|pXN%9=47$fqaUkVuSZ@Fa%?p4(1qR*{(!=UV8(RTfY*}P(OSEZZQIU6L$Z>+V z^8%BClIi;x<{{5zu;WRe2GI%Mlj^e24ki98o=8Q0fP8{XhaC*!(>fs639lziG=cBi zJAU&!vdxKIqzBCO^L&cSrAQWXgtv>Z)YTFFHUmJc$4he+S)k4ermOQz(|sp=XbO0cJ|G$z;C9=DdwW>6#^LzJZKm1AZ53f05XFhU>p&+sYL(l0_R-M>wS?9)SO*2+1*XY3y zire^ffgW&uR;MpiGA3XQ)7X(=&VGVo;HvECkPaBcpJe-(-)%18p3FpFDZ> zU+;bGA1g_zRgR=n?cn;}7`_L6j*z_4D%Y_`WVK@!7CbnWHe+b8!x;U3x$@eze>N#4 zf9&|;uQeQG?YkR% zkeC#g=a~dC5qVOK62e#pqhE4Ia6Zt^=gJHQfZMK^VTZZ?$iLaCNChJyZlxL-z1+a! zmEcw44-rK4)i6-QRT^xMJH8cBr^!Lr?NIPyZnl5B$J}ITO89^a6haATQfZ1Kk+G%< z!WS!M#0iTfNHJ{HurT%jv_P^IJYIQE_T}%JV8*Tr`X)K)#9OOQGvHjJxbqY_OT}c+ z=R>+7We%uTlNj`Z#Dl}KR#qH&Se-qyTd=}+g)f#SZcygERxHS&(jKVmEVa5tno?Nyrw1(S_J*1i*e9?rkpItw-l^l3!1X+t29*~z_LV%^X*7|G(syke8)&@LsUF%Z}7UwqhfwEmxJ!jF>90+2} zkm*UgsGY=4ToST$`rXMXA;>gfT-17|wR>z1x(0S)3|hPbl+=mr2@)^7r|f1iYII3R zZE&V(0c`qVlcFlPL0K%hrdaECY(zy{Je2B({$ERo{YQDiL@Q$4c)iVA|GtW~~m;!399rB$9d^(u?MXFyXLE8$|AP1_MO zg6X%T>jS?V2j`e@F@63g?WAlxRB#*`lXINnj#8waipc}w|HA?uv%)h|*6*@Md7Ezc z*gv(*?V6%Iw3ev_TA?McEe_l#DE7q&DI@@|bJfcq2c{E@#_V?ihf;W%n~_}@lv}Ko zg$w*62skb`@DXW>z}nXh1+;bqqwb(%52;H0PtofnzgA@^ln*n70Gwv9T?zxNbuPqiYuka zUL&i3R&|wwN?*mW5G6!*N4Cm83apev<*sCDJKaS z6@%e?7`}?T=vKLb9jS(7)39@{q*p?f<2}-w0<4yCR#FIrzzsmpZ`BsZ@8vzawOv#G z>>)?xE-xW6j=lZPRm-AZe1wSeM#WmIU6vW15s@lQ67PFm4~2}iP`dJhn&j4H$_NfsHy-{a|{5q)DX(-A`k@-NyF0qwUhZG(5ZT_e@o6*qgjD zIeYT9cP(kyp02p!#JI6S%~m5brc6R<<`x4C>{6+DzfQ{zHcY!^moYiFE_U8aTK_kH z(;D=cJGO#o$CE>5qP~FQfIT<|sopj&~Bvrb3gUdbU#femvo?6+ig(E+n!eGXoz zebJhqZhbi01ekBv{>yH1$cdGfb7n~XnBqR7NF5@1r(22;J(tFU7aAmwD4aciA!SI` zE|u#2Dx;tR27Pe$`~?({6Lh$3b%PZLss(yiKy_@OiO|zndZS5oN`ean8v1dW>GQZ8 zjt;4-vNh31z19WI!w}~kVhl3G$mb2D4Hp^UVJVjA*v^P~u@f{W1vSD%w)@SSZzfKF zC)qByMg`|)pnv9x&0hpBoR51barw55)t%5soDyyf8Xs^OaKS1jC{S#FWj6mCp4ZT} zcmKRmM*lvqckjHG(4#+D_u+R;GG_n32L6m3<7eY_V*Ta1nM7)(IB-@rgRdO2opfx9 zuyp3*uzO+Pl!WB(c2Qx_E?8$VKBCTlv0~+`Wt=gqrk5O487uo9R2@_SKMD{D>R2f7 zBTlAQb9aJURL1;4^mfJx7&))V zBvvYXp#LEcxVyklfzqWksz%mH|4+WWG-5}vrb1F6***!g93YSsTVk?zFyEn<=#nHh!V-VWsXts~n5l8qE7UBX)>)33hrFt1*_YV#Zhi!E9!1 zMTJ?XF*vb~ihI9}`@3b?EzC%`-B64*NJ7Qnn&!h*U8i zIv7RM{v6TFnG{<0hdt-Wrk6&c-D8H6Y>LaENGcU$+*l6)B}tyR&mDL!b}4s4@pRN> zjTKmd9iY(@gNAWKjF)d2l+a8dNduOn7S(5@+mXH@yOCnRi{s#;kiPwauL&voH^VNG zt>XzebBBD2J17oF%MA2 z%R9|U-zSIp$rUGFe<1l_24LdA^xQ2WaCMv4BPm{8D%61|d+x`o)1i1T&2NPQ?jld=-N>d;W4|pS zD-=42kFaf>8-vEicYloV&%xgsPvN9M?>e#*_VK(~(DUP~;>7)2EI3 z8+lUPIy9*==xg4OeG+^-!wyeO3(#l-8X&5EhjSIVyV)hL6$)Ti zvP-$|^|iCA=y`K(duf(`>~qgY{+6Hn2)i1*BqB)h8yCfk9>kFRqm^HJI6IY!o3rTa z4YT(N;zIh|mqO-tuUD}$NmNJ9y?G(@wkz_WH~o9H^o-0wNVS9!f8<(@;Fxn>$fbP# z^jhGnCTSA*`?9%Y^LU6&8!Fs$C@zyCX%Oi;`GUl#IA(*|sKK}ry}X9>1l5q0(HbM8 zIJ#NV0(QG;bnH5HogkMfptAz1g3|&H1{VrWL9*hg1ff}Zar|e*OdeYqb>x?M&4}|_ z$Jwt6a%T-fWznh~Wd946Cve&f*GDL>mLk=FYl!|wc85X(yTifYGm;MFWyx9Oq!BF= z-ts;}N=TKb4x)%NgX;Y2ywg2eqU#~gs6>(&j@Pu{W<09Is|#FnC$|Qkf=tI8S}*EQ zHiSM5+evifhI%p661_*6B}j4GBpn=cp|GCDyEw*(WY!3AZ}t&bo{R+j)^VSE@7oRT zCMcDD<(sF-DkpYR0~O5>U3V+RZKmLj$6&pMt8BH?zQUYGK}w7k-f8D54_MeDMd#6QN-HjN8$Y z7f7hZ7w#8WW*R+h@^{|V6zFG%ij^n%L`Dtk0;I)bAp^{t?p7v0%u#DFHDzEz2OkBu0NLMU&A! zvf49K{lNuU8xzMIRG*gb6X->tjxcFV9M>fSLr6_zyGNhkp!zmxR2`21JEQo~lrxfS zx8$IDItzG8I;M8fb?QpGOnqodOLU&JjBE7lW^XH3G4;}9`l9?1TQQ2R#}P=NCx9Dr zFcKh{^_{i%(XTAyZ~0)yi5<{ZB6IVk36Xk8EW-M9=O#7SXc{wnX`^xSd_auGQ}~bZ zqb>hdXc>R&w81D=0)x+zUU$qKL#~au&^ta$d`=23N3=vEIS1;e?V|Uy1l^In?igj% z5{;>37(BT?q${c;8ard7wVr30Mh*mwti)FKAf9jg`4AZL#SeeT^NF@Kx4zO0qqysg!hDDz~eEp2aL82%H7~8S^;Dox#AwlLqWcv#Ro;sqFF~&aE)mb z+@8Km5GTD7oDvZ4Sr>6q-2i-yH{27#G9i|`&+~?4F*WoVm%XpB1;2POgE9h+hs}aBkHnuIemssf>Jw&rXQwv3F;tCEM-!iyS3S~^78DuY!?#jdJ?}*)|arjQ0+rY zh1Y5I8!IZTc2?0;bZ^eb$`*NITiv_-Ug7gw+c5dhAH)GItWW+h`pYXB|Nft)fu=?3 z54&pCl5{6F%vPB#JH-^Yn<6`@m`Y%Hhj2Il$AJ%gYJ{DVMmpYeMW9yIMNaypdlYze z3qGT7h7E0TjOZbT{v#46O(liW8o_OPGl`p37P?cmN7e_6%sS7TVaO3Z`a;A8)Y!5( zj5=<3p@zEugQlB86KbYx`Q8z-l%Hzj#QSE5mm6aI-$-%mDH2b`AT3R+93_r{;E5Nz9kh;Pi1n1&0vNoWLM$)k5OmSqd z%-E*{bl_xKACWH4^VdO%cck`Cacu%y_tRT-LihDqj9#DDc22mW8*SQc{tgI2_czmp?H>fcy zUmcJfiNYclDMC@23(dOo}?pi8l(F zI@kZ(N*QHT{pOz@`jrV5k8Tyn$whv!aN<4oBQsdsqPUwBxrY3hnCA+cBWalclB1B8 zCDwN!J{vn@@uBMytV3H*wC~2f{)<2P|YIrwC7zv zJv7&3%14tqpkK^NIgDsC)24eHmPOY+1B;Y9e+uAN+d;e-CUvJuAcnxT4O&g(RCF;G9>& z5{Ua9B8T=-TnR-AVOKwXA860WBpmNq3FUg$2CP<|mwY1cciG22nvxu3EY*Xdz@7B& z0BzvAo0w~|V!BcFr+$~SQy0E!I4=t|oZn|-WpKXHcO69B{3$sI<8flA?g`BCI5e_g z-@)mD-%&@-kg-Q{_+i2HMG+?CNL**9lif~iL_T4Lg-VJmqsRd&=BnV5EJcXaP>muz z9V^~X8bY<|vxbOXWRKRW_K*^P$dJkt-*!zPnSzC+&mE+JkzzWVZel*?-GpR`40XI~ zs|2HLH2VafNeUu**%gX7Mw1}Op4bV+Rn`XVblJTk`Fv*V$?w~C`@?%9hS{` zX=kq-DeO9ExrMi~_w7(-srAYnx>It4xybB`I7JsoQt0Msu~YT3V;v0B*%<}ldcw5m!v6-ovy z580+jn-nL-zFPyfd#nw}hP=j9rjfqPrWpEQ|CZ=ZHX*81(mX9*m^Nt@b6RmQ_&k*6 zdJNm5d$QB6ZGkP(aiRBQN9a4A*I$R+MHmzNZ}-x{GkZv!VYj>Av)2;>@(ll~)ADVu zcOr4b92#3c0f)n8d2qIQ{tnJIuUU596=L?@_5W#^Hpj={=)}sF6{g@qW~IE!e+!Vh z=7~Y7>DZ-^mTF!sPgv6u zw<#gyEZOM9OB`r7L)4utii6Pe6e{LZc9Xn_+3S|^W?n=+eMVjux(SSiEzuh$)xc_& zMxT)Gesir)At*v%X%4zCYH`n?x2f*9?~b}D-y9M*az})0N#z zZdFs>m);uOwh2;yM#}kV(N4UZKW_%B6BKuhB1e!cPsgG`Fi)Hku2n6kdO?fQdXzvp z935-iN+B7r2LFTmD*PRAxyEFu6WIpWUb0)f+-;sdB4cuca(z%0lP5-{2^$<1#EPp$ z&7$()JyOkT&!gZ*)6-C{4t8QY-PW=>Z>9lPK()sf*NO@G!U2~9L7nt+H!S;YU*T{2 z;m2d?v;&K(8~v9jm;iO{w-^3SR`3Iq6E96bS};UTZJ{^_MOshAbPG}i=R!I}Slkkd zTf{*^!b4gQx*ArIY?a4~vB2)wEJvpK;)e}8CfUv|VJlPK4L4y!`8UmWvTHoBatwW` z6%+?y;-yqf|0KOA5$G^Mlo2%a1@z$b$S^llGe?2$5{~NlP z&0jq>`%hZcmXHL{q-#S4`<2L`2v&HP_kPz?bfI9W`$qQ?i1kc|UDHL`r=l+J8@}6u zfg*EuK2S@-)L+hGU>P zoBUvULSzfKA)t+|WsU@wxfOb!0NZVcGEENbB5G{2ZE`8pv{)O^#N>)I$cBJSPQ>Dj%iYVPg(1B&w11{z3v(6N@W#;)pK`< zFESMnU-ZD7#g8BIY!sJvHXDj(UdVeTV9$deTju&YtyHp-*IOlAFfmEAO?1OO8_bCY z2gVYg^)ttm*!A-n;DL+JJB`5+KV0mZQu)@une27A;ftq)rPOAsl-lZ3$ZmNRiib&dtKm5y<660Xmoy3s zf%m)YpOnC@U`wezvXa-a#8{ye|F%ayB_SIH_&0V(>XgYr!`TdYAkj{_V^g=Mux%ZQ zzxmHU`A;^XuIFcpr6i5t4$OJYo}|)jvrt5Fkb{#?#l(IM6S)>Dj)K3 z#lfFNlubGx(gdkfU34jZNw!abb)oQ19fWiByIeFKq5UYIUP3Zy?71!A^PZ!zA+VUZ z(Sei2prdfQ_nqbS!~~immz3jV*?3?H9lAB!L~$D^l0d~6>zkur=_7DpYZfzlpn^b0 zBb?~iGQV8UUbzF_oYj9MuUBeyl+nLo{R#oRUBTaWom?1ykXIC zuM*~X?3#5bsE%j@y1aJ>Ly7Ll^C8G#ZM>Td5pWj+yS#Ui`@ww>9(|l_^<6#dl;7SN z`F_3bEz$VAU4j%QKeA8osVIkTlccJRe+6y=j;ng?qE_McY1p(54b#YuFXziS`{@}b zOeM2F&m()Bc&qk_8A@s?u9_khRLrh9OWlvp$r2Pn2+ie)?S2(O`~8Ny))d_m{jq9k zcp|$#bX$OK`Z2Omnj~t5t5C)?c;Rbz1I?I8p)YFl5+)hM`?}2~BFtuYUZhmkCF$Z=89TYS^Z-CO?96eIp44tc|jz7gv;;cGMNs_1+fM6PstGT3-;prnf=~+xOnbJl0}{OJAk;< zp>i;v;((eq3#+;8@h^~c$r8g0l}HjfcY#}hWUdDPdx~Db)cWoaFY$rc)J3mjB>Y|< zjLul}I*@e7NwXpGI8zP5Uo$kDoKkJc53t6;Zo`7m5!6o5lS3^BWcCNg`CeEe|cHrOHm$-A}1Yc8aY6uz%-8l zT!TxIO<>{M=)7ydZA(U!7m9q#3hzm#x-+awj%DKq-nt~arHBjd5?+*M(CLxa6kUpv z$&Q(2t8d@->9;zBLvC1?Ew5hQ{dG%VIPG#{1q4{sY~RF{OusS{G9-|sqEpewYEFjq zvF&t6#Oa8!MS<-gaDe*o4ER_F!7%=bSDN0<_~7|;eNL>LTA}O1^#JpE!FHzVjohl* zIirGfKxD2V0%9RlAFL*oA&V7_G~`I-iIK7EpjuCNI%v}}bh>!pW^k>*-=RY|i^H)yTCP-p84Y*@r2}!Avc0a29Qet1E>gwip<2h^t#=4 zyPa+?yY05!k#_5Dw>zE9PJivRMF&T`fT*B?8bE=_#S0hl%5^}zj0y}Yil`unGY*P? z41)^)=ShN+NHhl${>RylpOKT3bKc;2zj@y0eJ7km?^hRb6j-$ti={j3IMG!N`8uJ?Xp&b7%9Q?q#Cu|Z6or7_U? z!rNYl7aN4Fl4h=MKfg=U4gurz&=Hs0V0msUoV6>lf{^(oBZQAXgj3GWzO%XD8|H}n z@wcvgMry{ZqGK@t+AYFjO_Z{MB4??XWdCj;nT{8pCRjC76`>%-!=Z2-vF+X$aUsZZ6XG%bPqpJnqZ4 z%!!Xqh^hhV*cxt$d{C03`DFe;*hX!YA1S?ZAgsgxg7Xz!(I=w^?9>u z$3?PfJnO^Wg%yZH#s=kgQOXPoc+wuT*Q65NS)1X2UvE=F<)a zp^B4AR!6^^Cr(imLP;%}sv{BaIu|+WePKQX9P4?>HUD8z9Zf zc0v5B+4R8t-hi9hRWqa-G}f^=_A!pJ1RVP0z{YB-*KpKWlAXI6YpO#k(7&1U7bJ+By|ig&D}O8Juv&MH+o3Yu2sLk^c|z2&EaQ*f z+R8uRt3%JI$TQpX{NlAgI}q=aD8&00^R3PzSU^L1syq4J}Ogak|6eP9J25 zxU2+GOvn-3UbtMiU#%Pc*uKcai`xhWgEGP|c<+CyfWl^|ee&~*hZLpqoys$+Urio- z6ZUtWc_Sn6O)DSxS<&-jc`PnVuo25^7pzv4@~b87f{Pq1rfk;Uir7f@Mt!D0E(u`I zce|@s}8WGzCrmz2U z;pLY*1;r$m-@V`x=N1H0%0Wvz(MS)|wE$bBZ!RkH8I;EJO|J%}2Nm_Q<-W+lnIXO& zfow5F(hn9o5@PmY=b8S)lQ$g*BSzSnK0Em4GgsieinnC=|Sjf%%6QBV$mvY&87q2QHse@KphY5EO2EY_?GloHxr>#-NQ zk{*O@(dQ&Lpe4}2zozJ=v6B!Q$@;i;l5PHLs5MlN=Lz8QDC4c8)=^i22METCk#NoE z;7AP|6z##=+ZHV6ou(@Qo2vsy0t>>+lzLTez|93!P+@c3*kpEZ<~`A!1=S$fe3}f* z@20m!*U_tH80cQf7V+pk*wuEA7e50eRA=8=zzn?;f#E0|8UADh!?;<3E6P##=E1;a zvW3O&%>fIj7$_ywVP;Y>gZy&QuuEQGIYfTj1*k93DQFkm@a$jMBiaHp-Rn63axte# zt+-j+PM?ar6#2>A3*Z13%8xAu5nNu1kl9=^dcbh;B<7y!M=`pdkqM$s{Oj6y>$EkK zxEzZ!l{UIZPLn-Q$B#)$jLzt=fT39hbV0DT?eNp}DH1?zxjgKsCXtgCq$`fx7z#~8 zQYeb(irArUBL-im)`v|3R8B3`eBZLwo~9r!HuF- z;d;&ib(5;exbwl?&tb)|JWEsRH7wr{)uhVf<2lpL_cm`6gm$rNYS?%V&*OgksAf+D zniVt@NxX|3#~TWJaqv9X!qK2$wB$@**mmsSUv##Xl{4alvtC}O+^+s$#w{6&Z0l94 zW*ig53S&dJ`Nz)yZeWtgG4O|7FdIDVay()%ayN~6XT1uqcm#$2=%luTcz34Tk$`N_ zwZl{k+3!^%=@OvRJ6QwdC>%Y7vq=Jr5ss+h)t3T1&1+KA zFH4t^RAy@ui(Qw47Hd*2rG&uEE-J=Aw`+7ws*{>N#TCUZ&dSIXMJIHpVPV1s?V6eO z3-qc?x7z`!0rd;+$Ud1r;D4Mj=xQ_`d-h4rpbRiM{2sj}$`a@zxIpq%+-AAe5lp6Ky!ciLG- zi_3K7@y4H6YuGX|7%blK9I{wk@+c(`%V%L|4*T4X1%D8}((9^Z1qOOCyPObS9GMet zpigOxQf}=uR=gc_i=VUA8~Wg}jB}lbsjsR}hT11b}5Jd^Wp!}*Ehdj#{c-WH0gT&%I}o;#(A&$TV9!A zvtse})*22Qv~-)ovY`UH544I5P(!Ud0#Oj>tpMXZ@yDm^g2dyi zr|y6eB&Jcfj%7|J(#qmpt&K>lvHvhpoXA=4uZIBKIj(HV=q1)JD|>m#pBWE_|4;}%=plr{`oPI z$>M}>g~f~)Qp&v)$){q@eGNnv=2mLzAiaMi>YUdJ`2Qlz^*(MET_!#cRRWFB^3$!@ zuGzv#<)9whiKsTk%|)2rzy10($#vu1CEOy;KHiC_n~UBLMbai`AcHXq9>ZbG!Q;7e z+ySGNhKiU`*Z4CtW`1^b``5{37Gq}60yCdd$}Wm@P%%$b$6>qLN#_8Y$B4RU$&g!Y zG=xG@Kt(Ev2fNcqbWoa91|a%KP#PQK+e9~^ElUrL@kcZdJ@!Okng{MMggk@_M-Wik z;n_v6AUC;vikqMgi`ttIcui0iN0x`J_t-NBFHfgqqertw7++0MjHuybv5hTFy2EcH zD01pXo7bT0Lt5seYg2$sD6ovAgMvt(Le~I!=lEy?oefWKNUc#sZy!E8QQE6M1VqTlwSGsb0W zvOT+H1M{&gvO|#&wJq96H*28F!?164s$;SY5bGc;+BXa{#O^D2^$lz01SSz#7Q4VU zI2E$PGu?*h-k<3I@ySt&BLCIPUt3OHm9#^@=|P1d1jz`SwHLuXLJhUQ+?Qw5B~I~r zEGCr!rYsg5i$R?5q97sX=sdGib2#J=t4RKM{2DBd&Y!e!bdFL=2-hE=Vsbr^QX@~e zj&H13Ymk|`KhD3FGE=u9G7p%o4Iu`)GPr9VmZyN{fxU=$sOy!bNwYKwvoQac6{0h_ zCV7153o#9H{#p0f4*q!?Zoy3eD z`Cs2?BI{Y~67R7;F~Dw^s1#09ZTJ+g(?N+zG1TRr!M27)W&X(#6o`epT{`S3Vi`8-j+j4COsxi!`3C)HfouVg`@{& z`n7`Mva#c@V_v*)&HQ5Vh)aj!0e#lBXSNQKv(U`P=t5C7jPX-b*d&ev7aYe*-3K#bk1)#@@H z_&;srwtI@=P-rDc-8l=)fQwOIeV#sVL|p2rab>QbRJrv$Y>|JXAtj$1^^`Uz_+%D&(5pm+~8Eo_mB=y_66BapBtRmg{gi6QT6iGigM6@ zz{)zkYFK{Ab*=jl0gGH8rq}ezKL!DL<8MM^zOG8=cMTQ;SNfh zwP(CHk{WK2yb+q@bEbg5-{J||HPuEh)#5g%zhLBBQA-ZVzB=8!nTlH#S_|Y9j0Rxw zLbbzU>$RRzuA|6mDrPgkQ`#c!;Tmfx|9@kLvafm7>0lY>1L%a5&}n8wXm%a^19|kN zLBuu-L?lqkH56He3cPyAD`tc*yA{!-x;4dGGLL8A`S#@TrBizF=@0+@9lucv@wWFm zlFqE|g~b-4%mNSlC?zm(8AQ|yx`@%BWlb+6SR$!JD6)>)&dOkvPIc{km} z$>DxNGI;HzMOqzjNmMCaEcc@ARb=3F&s{n+t65tg z*1NEmUac^28lx_V@;MpdOEDb0U3Cvi8gfXpb_*w8GD6>VC~MrflQ~h4j60YEPyXx1 z8K++|gCy{`WlPCc7DsVQEv9`prG$>Jom9*a$lCNn+o&m;pCd87KBYM;!5)p}!g_g? zZ-rmFX0y0SyZ-Cvew_KnnIAuRb@R9P|6S>l^IBYVRym7P+cZ-XoqV(u4xHHoVRzkgiNn_P8EXmH2iF3aA3gIcvgGkJ{>l9f+{HV`^S9Ld3L@ z-@G}^-)uYzMt)REGG>r+i|^=uN(sriJfP?UDgbPXG82X;`=cFRF3c162p~X|=Yi#D zO)9U)lrTqE6x0^h2!d4DY*kMGsZW55ImWMwBMtO%SBJ;!?U(UQ~RQq=iV`Pm-i`(q3gUwilha4C||67L$6vR@4y!04#gmp z7;N`GPLg;n(hShbyhyf6jfp9I1MjaPD}-mgA?DgD0$m^^7R1P{14<1nVCOI}J9rsl z2IhTtZ=bg|Gpt>BY}mgv=}{GsVWWD&)kX2bY`OyqV1dSs-lo2&{@B;3qJTP>I+RFE z_V4xF%-_tv4*f$TE-0jFpzEbK`ROx9Tz0!(0XG&(6+W3C4^%7;U%ng#84Dab-~}Vd zOqafVqi2qJWpWW~4w7tUdmk3Z-aoWhU=L7A$cXQwVvec{xs~#xfyQ{+=I~aL?#_bN zumkG2P@P_LH1L9Vu>uO6ure6=z4WT(i#PMiMK|c<5zX=umw|bmKDf`uYVMnh9-t;` zisHCyr!+n2l61sn`>cey4#k?IP-a?p9RM%$5V6;;oO#iUT8@A9nxtUK4Jd_t?~q>) zT}ZFsbu7%|4v>{(hx+}{UEB=MweB4Y?*t_R@y-R0exEPohx~?JD(ORhW&B*lAwRte z>3N2@iM+(HQvT7#FYfEU#V=?Zk6`Z5593t(yToJ?38u%0G%F$*Kdw2k$#9KL_E$k!{m)G5Z-H;Je}eL~&*@i&T~7O?s!z_tpM%ly!eU>jhr@EE5tmXg+??;@ zW^m%8+eEiy&=ae|l?t2I9BI~`!r6o&>|_)(g79>YiuNB`bNhTrs@-g8g>=yg%4GkH znK~2=Iq!AB%R^ZZubgkH9}xH6*`t!Vg@A3d!6~JV|Vy@wK*6QlK$5XBxeQz z;hr&E)1{QMm?HbBm{D=$JR$NAV-|n=tP0qW;^H#ml1|5yx}YWrmaS%luX5cw>m=!S zYn0VR>Qy~-yZeX>iVU{99|~O|EL5C=mUg_S%qLFPK;u{r4PoFDWA?;ttVG7sj(z@& zzPjwK47vOL-=4KjPCwm=#Nt?;jf8bBCxZuFi<)$#KFHSW@T?DeZ<9w2!4$xs;M9dM ztc^RS=_R|v2iGL$YQD*==EC(kSH?iX>Aps5p$_b0BTR2M!|0+}yy%uokFUhp;DVdp(Ta z&qto~DYM;7pSNS5s_oHiKkxAbz*y_W`c%sQ$y%0~iEqN*O&mo2$F8H9loCqww;_e; zrZ&(bsQJ(pT3v3?7z$gNYcd_ z;U=Yq(Z8=oCAr2chh||NMOsi6^serubNpk4=VdyK#3hP1dO)L)8@|;ABx6@4z4L*J zCqY?Gm(Q=5-{8H+3pvPi*mDEnBBM~u(kgmTv~L0Is``*{qy>{h|^sioCavBmwdjERq0rII^ zjq^D?MlAwU<1~7ovwzNu9uZh`aI-e*oa(%R55bpK&7;x$FoE}DA_0PcPsH+?=^_U(c%)%ctsh1y;osWv3jc%UzI@C zZat6_Ano8ss^uCd6YNZ8Mw~Yvu<`$LxQ`y+A1Sb|#(R2$&Emzy27zVG90tmhtf$P{FH|uMFu1s?D;ndJ!rB zl#5P8A%*5~@&T`c=Xhf8$pGBK{5TlxwgEo)jaR2@%&y9*n7?cyMa*0k7Kdx^=)2u7-r56xi!o8DBln5rOz+k8?lnNl85b07%0Brz5k_LMK)b7s37b3 z>-go|3`fWdI1FA6##+QTzY+00Gl=4UckDhnGlSf=c>1lB@-js(QZe`~L=BEpJgAS2 zUD!@yXP5W_$G-T6C&bSB6v$(M&lU?OFL`#;CJByZ;AqJAD)xHoZim*(w}qec(yOld zAJnW?Y*Xky2(J!C9#ov9a_$E}Y+%ZPkGPccVujtJ4#hgP5S_*PXPpk@NO)7Frq;vxQU#O3y9GiW`=X@&(?DyUJ0?J3 z@oHv+6%WNZkCS0JkVJO|VP4J1)`1ojL*GXtbhjjBVTV9sT(9b+iz88H`p8r?V^0hr zjzzG55Rc{{5&mhIj{2dm1&kYtU?uV4Jk|JePOrv{sFq)Fx^i@eAXP8&z zf6l8zSsu2|t66(X)}crZt{3ERdpwIHzmTqs#xftgX1nW0QO(*_{5akgF}7VlQX4?F zU{lcQ;B2V(Jrz~DxKeXqIn1-(&vSYj&$wZVyD=fk&hC%?C!7>aBy zm5zdR5%;u6Pc?1(pQHZ=p^_diWm zj7Qna;w2};!k}!Vloe`Q3tMe+?5m7;8e$?bwX8Zopj32Ch z!%q4t+uzS|_QmY|%6-HC<^PqNSD5tK|M@$T!s0A9lmw2c5zMBPAo`n5#k467sEJ96R1D#MrP6HWryj_Ascv2vH>{GG9Eh{$czCz zY8fUC(Pt0948S=j3zyF~E53ENIr|iF3kxL6PfGAVveBcUyE#k?gl+PJUE1n9Wr);HGAKF1!m$^QVEO|8j64 zz1m|yfvLjr&ti-0E!jQ6&AGjFC0G!=s$lkJ?@sw6o?f-hA75iY1V=HAc`T^H@0`%U zui~7bE1)+IuPBjVsuGj*mzB$B)eEXY+i%^hVZ}<1%iMa|qd=fsh9>TMLA=L}1?8eW zAc~Vu=LH@V?*I;Ecn1=j1(CgO9Z;mTi+f+bfrl0HmtRf~Di@u%SyaIkZ&vnX@-J-w zlo8&jrLnWxtfg>RYmRKl-k`o85Fd$<$l;+ROVdVjIl8QnO=0bHxw;jbBW5DwwjrcY zo}GzXGm_!^rg z9AwWNj~*LvV0fAldZy=Xyz`^U$P2JImjwKIV?69Flrotj8>kp$TF>Qd;OhN{U21_1 z9#gBE!u0;BkVGAJS+4!a$3XzGbKV%>$j%WRbJ;Kclz-H^e1^4l#YU-2wk9_)5mMP- zc%WCC$QgEt6&{UBUhUY@w(ZL{LfP&|&JC2$C(NP##lW9w0huf{K32QN9!#*Ur zwunl-@`QULfEvhSIJlLMbV2F#5O>WY{Cm$-(zj)D8 zL`xfFAmZ!3A+nrKy9{yZ_m|brK3UeUDsZ@z&|Sg?rIODLW|AhD7gII7sxq z%N>;(p*JK}m@e)m$27Vfem&mjWT08o6t+3KIP$g!q>HdJVTij{lkA5^CRYOFWGG3~ zq-qdA_kKE^COs+4(%{mB)hx;WSAtK0`ceGrsYJJr+X~eUDGJ@ez#4F#+l_9GalrW3 zR|eOrp+LrXG5EtNP_0++>&iFa-9Gs(*^%&W8U(=%(4@I$W;$)4_e1(BMX?gvHoNJ9 zB}iPFFsqPbya<(-EfSm)Mw%t=9P?@nSeYU1TV5eG*2zdFd8ZdIO}Y0g#}>vZo8`si zeX%Z1nF2Bdozm(sz=1wuig}6{KxoVaJO3~y_1~^Wo6S$)`)l*bAr_mT3l`?*1f{H{ zNHrC6@gGbD;Wft2qm3j7sBAE$*yC9*KuVi3alN35b9;e-pD#((oOyM~Es=vAO1)qP z?na#Rx+F^EoRDK{+EtP!E%e+e#jBgONv;@|IUuT}&jI7>-dVR7)N?P1+6A{H84|n_ zM@3p=yn0fL9RqMMwjq;&5F`6SQ4tkyPDUP!y-7A=LmsRD>(?YLiqA$Iuz2E+SXi6_N|{TM9O!lYaqjmX zE$Ne=^QxnBsG=Y2rq{kX{OSg%JAXg=gsNV?AMDFMZfQs#L}brG-}7Iq_=R588}-n6y^!{R;#uZoG86x-du>~l-=16d=4)oSJmhBnl$3vIl9X30 zfOM8po~FnNDrVJ;L(yyIAJoRrxGsUkh^(34to@uE2yCRc0l`(Xwq5XryjyW!UgI?^ zH*j{5d`TZS(+}8IfkJvW-A=YaZ*~DBB-^m=8!E)H0RkOwTmb9k#nAWEB&!sdJ|jxpA+_wX4%XHU%3;6O+?LFTvE0vI(-tuZg}R<6^<)|dgk;@01eL^&rxsMM#DJ9g^Ij+<|`g% zP6Z=aO|C7W2mj_^Qk>or@v70XP z9Dx4FcDh=P@8j+V`@C}^4lhO#o@VV0&wSxQZb2j#X*GD?Lgm4T!7;B+gyctL`dtq{ zLib17a+oLjMR@K~W$y(xzBi1%2vq;0mQrh~874Ub_I{I~BzWwG>9N*iwx(M)q&P`> zB`bt!GT=HlS=MSFUA6ck=uM(i(;`JS2VIr#BQNZPOOqKHf{_yqHS6+(M&2?T&wtXV zCkw1e{{l7!m@e@-_M(e;osbNW9{Q39xD>rW%4<)=ZI89UN@sM4k>=x$EZM&si0t<8 z&ANt|G{6`OQbp&9LFi2noW$5{k2*`b9+De~$-|VHkYciBC~Q2=XP=%d2*w(bXMYm2 zcNtliX@?dEZJBBId2aoC>z~PL7AKnw7WN{QQiA`!iHf;LU!_0xX^4bgYPuy7?M6FY z1<|v58Jv&2N;kv*jW$;aW>43?s~KU#z9AiQ6(ej=pFIrvy>+n+la+_X+i)A**DcZx zMVw3z_NJ58K`54UYvE~7Js)x_pgx?v(;rtF(*;1s3Hxklz1Io2aqN;~dmfmh!}eH?e5W!&|Io*b4o$-93bKP4I#`^CJZgcC zJ(Lo9oD5V<2B!d)o;Xe)wD@)c#qT{{FRAiv);?I!D(a?F!cizvuj){gLS-**hESd@ z8(8YGFC|NJPmu$x>36@9DBk71ebz0fWg?m5<%vl#9yt;uW5)XFUR9>GJ2DQ518Z5( z0Rdxe#$y)+wkuGqeJ*%Np+g3`;~jgk@e7A;Z_pZ4g16lG16%_gq8W|M${KT2YL;B>m1g4 zWAB5o|54jJqGPVJ{iw%@j4y5Zvyxsn=rWY-=XN@-nT#Q>>;N{y5gl{+#b2CWB{d^LbWikOBx$@XA$t)`_FBw) z2Bicbq){(2PB^up4~*w7l@4Y}SQ zA|FonL+t>9mH+VgrDJB?#wfP!!0vzXd+q-g<4;y(y`a_0^mns1E_%6e5X#{8a4SiZ zb_=JM7tbk?oY!I*?rKF`-~~}u_DHTqoL_tqB43Yu8mF$TnkhUS34v#g#RwjhcjK(x)&f~iZ*5r| z@US6+g#nLtL9DQh|CwTo6l-H!MPD)ECm#2$Ke%>=d0V6Hd|OABGYh7&7#-;r8=ei6 zGLa%{sTf4Z3Sl-dspko|L}98wSEjnH}FcTKIN&DsN!OmPzDqIxrTtMpM|gJ6aK6J4RHLk2^? z!fO7U`H~qnRvv$6t996vwb-x`2u1F5tOmo*DZMI9R^X`ec81erhqYihqGK*+gotVH zdTiQbU1`W#M3h>143BFEFVgu@*l1NRGw}7QK5kLyYLC)T&>G_#Ntnc@?Q3gTPI#Am zr?gpHE3JnLL%a);vcAWCkuAlkw!{voJXuz(jvNDu@ke0<6{@B0m5;3>VoXYdSiH@( z5kD&ecKtqRxL7XS;ZsMioO3%=Px7PkqYO~2j6FHag{R=%x3^L`KVSXTyGwqrSDheR zLpFvaar3241r$u@gfTTf6dZQ8t)k%=$GlNfb$-VsgS8 zMR#O+O@b0JflL=WJU@gC-J{8|QA}t?>_o+c#-7YgL)+g;oJ>s+i?^{hYJxhUO#5*-x)3h#cZ#&BqF1hPWH#!0{Zu+W;zfEz({fuBhT9@lu0V%k-*SvRsaVULnN3 zB9r=^v#g1oc(GcO#=M*iD4ua|k2LL$HAxa{-4Yw>ELhvKtXh3Vlm$vo#_G~s5H`d) zwP3Y+##3fzffzX;)1&@+__Ve7pbgc&Q4)qtGh4wW+$T;D>#$bfFeq-SaW#oA&t7zDGNus~F7( zdu{Umto=cgp+I7Bb&w4Tq!#IxIVaQ*P6yV%Ot&KencUS;$^On-HJ*Vj`;h&N0i5(! zama%jYYHURE+{tWj&c-GZC*n#Q){3P$~J`^l=aFwRM-3(fl;Jaev5n@v1Sn{_;Pg3 zz^T|LPYqa>>DMm62DwB|aU^n2_kmxUGIPYGja+tj&MIUd03I8&aSxn<0cNIc*R&6R z;%(l=biTac8rjC;7AX)~8Pl_nPbq<+B%6w9hRQHgm@JWF2uYiD7L*H&rN^lJtXCoH z=WP#^az^4<0}Tq!syaG-X0`_4GWQHz(FwK69l)#R#Ix~T&m&)K(;jzLTb+4pDQ6%nA=&;o27Sxula33IDnqE>JTp3&r8N7PgIrUU6 zLyW=XEc957`};%LzkI`NO!fqD?vYAnArTe_&aPURo_b1oh9ak^m`8!B;|)gE%-`l+7V&7o?FE(eptO=_2#KGO!(9iWEZ8#BtvDdQ zxoERr1%FUFFuykdxdrxfdT0=ihc=o@Ua+=al>kr-O6Pt(Y?*6f%a?xa--oPg3{rDhu@oKdn5Mx z?KD!)V&~UhT7N*_Y|~S4<)HY}?_ucNsx!+DXnkYU^AO zYqwK2vO8#YnzcyWnaH^l^o18v*{9PFxJcyJtTpYIIyiN7k7ufDxu`l=ugVQ*2}~4U z4|l#jnjlJ?W`ITqW`q)IsLE%~i=JFQ6nZH1tam0?XOJdx-iN~xmm&8d_t?-pNg00+ z|BO+zr!DN0`NmCm2k6Xg08*q55ID@^n;9RQV1MO2$V?C_qn2ZIW*z>M5mctnzjCu= zGIfqD-a^@^b3`HuW4sQRH*|dlf)eI*)6N5m%^9{2K{ju9+8b~7c^L=JR5@jcuStfX%HCPvunz&FCS#nRPQm~q8$xYJEDN`;i)9jqVevlOMip)R zj4C?C@3L?g$>8Vt#sxMjZfZd=8R+eg0Mg=Qw-uS5S(7!66sYyi^bHz!91JxqQXg&Z*$mH!>7Nn z4&6T8p~c=*C$SNrHF+0#z^8dvbT#mf7Z#!H^xxrOlB>l$co|TTq$senC+*Oj)tn}^Vkesl z#+aBQaM;Ggcuf~eqU*jBYR1W}@7^vVxh%#>odr$~Q_50`6a&9??0Y#kKCjy7rn~7uRF^DV=9AsP=VQ1P~pZ?@MUo&^4bHW}9RnF|2vx;=iTfMVE&;Tz&tAY-#lb0iy zlL><4$!t35h!IGp<^FkH)-1DCX_}d2<4C+4i%sx%p-RsqQ z)d5l|Xx2i6o{kGVs>X=?IuCqjh+9W5J*qC`;)|ta{G}%YH@X{-or;wS!o;}XI0+%f zVHv@M%5hU?TJL{YYfEhGb&N^m1`rd#+G^cZ$>zwvqL6SB==@os!PTK>k3XiFGYbR)0adIbVhI^y#cy~PmulUUG7PW&EjTlrduz4 z!j<;g?N<}UUD_vCU}ajnU_g`=P!80cr%9$?HytY+h=fvYNwPmwx`F&7 zj&W>ZF_dpt@diYb{6a85_$bot{5d)NGeS7^Qs1F(T6caiN#d|LRBNM8tm@@N0D$hY zvJZGYaW{lpofK}FYLa-Ez^vH`p$UvUxmyBWZap%`Y+YQ$nu8>p#nuIk$(VBd1C$bW zH~T>D(E}aQ_65a23JyvW244exFS1*v1J4xlGrD~B-iQ|IAP}iRcZR~$`I#GdfS`(F zYH+=vlkN-Wj4eDlNy7bBIULmeACkuW6&WGr?V8I$R!6tAav znsL+g%hF{emBqL@Xn~ttN|{3e*Lh4oc&>U?nV*5b-oMbhG^E(GO>;n!;o0w%FU**m z%N=&m9Qn2KPqG!o{d9${X&f`|nD}`~P zdjDd&mj_k&+NJeeaDQZa5ar8{I!^HPMvp7O_^Qxzr?gmIFT-<4$ppmHUNs_pBdsVzsdPC7p{&;xtpY}tB5Var?47;I!I@7(FzVp(4bra+bkCS9>1{BVA zd*1_j)cxx7+K*JHG?^f1GDr&Lb@b>se>%MQ&-)$Q3FJs1I0OOaL717HIiJM5;W`SN zuT=hp#IYDQkWe0DNRlb#28tw7F=;{f=+C3DHK!d&m@wUUA)=PIX=Wo7ojB%8&tQJ+ zg$08l9dao%SnwXj@6ecGaVq97n@AC})riI0pn3}sRa43eiX5Y2+69T^7BFRW(p>b?jO^l*MqP8Xu%M7#HCf#F2LQ=7t->uE*j;O?gYVmfmb*v zjuolf|9q z#@QbPm_an~Q`rY(=XjbJEG~Pjw17$xr360e0xBk(-t2uyb<4Mme;UM!^P|$Fwc@2$ zCFSlX$+5+y{8Nfs^wK&SOP};AJrwJg$ZNbRahJy8!ISLxq<4*RqZEN)CJ?(yLAr;nUglveWWNXU7PbmvC$;O%# zQ_`eaA>A}q7i4L+C~J7fLT=9Orq2aGcm=7M9AE|PnUaaX!Jcs^dyQ>4j8pA2O%_;G zvdnB(iha+@Nf(QKmX*`T`YfMQ%6^L60~-PmNl_$*RlN*Z!cwnhuQ*w&=tz{2Dkx9f zr0P)RLhn}!qz+r8=}@Sz8zc$Jqd@Y8kz8QmfJRkJWYmzGd}GPVeSTkLiem4aRCN&) z?3YIA_L0j`-BWZl17M zUJA8VdE!EOyP!=}KoG|xE@iwfSiLZHVpK-n7mW>)@Ztc!U}kNwUZdL#Tt^MEijd;S zRtes)ol5^IETKpj@+IjzeCfFVM}B7H#l7vlj-)eli&(tYm07HH`zU1rMRKW_rm*|+ zJ|o#}k7qf=iFbNz^f(&$P|_88724&iR5jqp)Nw}Wk2q}$+DNMA1UODR!4t>3YPda~ z6`pmRI9?kugq)BUy7tmVPLg4mh-I_#d?ste#=ZaB-?}_(-9*GBpTJ_o*l0OwQ+zV_ zb13g^m26XeHuvMXb&*Sx78?UFahzVdUy44@^;w->nPStsahylIW^D@e*^gfL_WG}_ zeXa3ZzsL74MV`~>osf!f0A3J2cRvDW>ukx{}TTXQ}M$rS1Zslkby z55Z74+`WD7leTl3pF57DFJR{I^#AiWx~xIL#E7uCcgO~WN*AA%O~$ey3j(YSCgtf3{3-=33*#sy*y=R4Z(m4 zl7wL(1^-N?*|2-IonrlM&p0E4G2LDE(AApxfQi9iarmy(!cTon!jDXY*g0~9{tTEh zP_oCwpm9p0*Yqsh0@SNWzF*5p)l{n!A!7u49#h*0DwV$bZnA%Z@}}!0=rpNVH2j93oglvlYSfB(CGH*b5VKl*V1 z>11)h<$a5Q%Y91OOOd-&3=ofEk0EWW0mV@NCXbE);3{p9J@n|NO+kJTfJ;%7dVQw8 zMRXW183GlshaS2%GC+FZpqu{F4$nNG8^jC1fM7_b1E}Eis&@AVF0`UG$ZE(A|6SqT zka&WZ*_s`my}-<0L1GuSiEhHj=LwsGfg~F(uzYD z*VzW&t_M7xBjDtR&B)Va@9H1B_~Gxx<~1u*u(^#SerZCr5SblQ2BD{v+bFUX${_yX z{_7`xkRG(dqYJXm>2z-3-LE8yPmuNgdqqecr&k?>*u7pA=e;U?wa0+d*nF(;$sFfa z132kr)oVXJV$IV3H1!6HLnk)4_|s%*vL+yQLhTkTN$7+SinDFmxKm`$#E*NUlTX?< za5G|N)=JDaL;lw{n#g(<`?-58tUx-Yg#NQsDyEimLY-}7DZV5{RXQy0zc24zFbp-8 z#gSh~S4N)`eKL2|3_RDR>6gaOm>O`{g9fYB&dE414R0F_ZA)OO!gK9B|&1RfZ9GQVeTp(wm?q}cAfyiZWKMaWlD~6 z-t*YTH(QQB>Mp!P;=eTB2DAZ;@iw+n$`p!hq+&i26@X<)jM9n%Z6zj)o65YPsb9+nasr=Ud&6j=n-m2{o*ZrK;oUbnwl zVAC8kTF2SD3@>q&x1L2pu3;H)fK*)t!kt z1NoX6Lupwafh=sAi`s*D8j%tzpgD7O8i`4kp;rTYfd=f4@5WMv z4lTEKC>AC+{T6pKvWwLFyq~>g-9`2#+27dcIl~qYv<)qw|FV`p7>NuCSa{GNGv$Pu zROsQNP138708EG7Mms!n!gY6Kz4TQ{E$ElklVT0zZ<?V7Duy-#-1iM%G24$1|>wuf&J>xw)R zIjPDzdQ;dXjjn)>4K_gb7aNNmvaAs1+eqGC}Z zuUG_0*$(KsDu$Z%F8K|mfpc8}+0fJgEJnxj!a`5vC|HKy&8c7$p72~adDF>kE?7(C z)Bm+U7b?xJi`#>N%VZ0)Knjcf-vbtoj)78UQ6v)!6L5R?ZYJ5tOZLxK7aLVM-bG31 z7G;v~NW{BEo_8PumoGGg9FEGSPsn4VJAuEnN!BepsCjooq={7SUB@moW9-7j0dI10 z#|-7rLywH&w_fx{vA%CzlQd~pNc*LKy(T%Y{WH>J_tMAxmL`T3aZ34hbd6WNs30;{ z_z;>0SCXSqHQa1^i?cbgVVWj5=?4`SlYvbYS=#;z$*TVsxtyOEguRxU}K#pjOE8ujIV*yik$ADp{BddSWB+nF)*We*OX zeA&m`=;TA&I2gwaOegoFKeH~kXKjkdMge`ExL#Jts|{|02pd)jbbyvqyC7SW8`$n% zE6st5g5^PXVY`(W)+^fw!Hs6^260?yqi;g+QQvhC8(HagL+(^l!VD@;K!ZolL<=LR zP$ird^VQ~sB_#c?8%WN0oER)#SZXX5mQqRy#h3f3n1VUwVW-LJz${YgRjp2w?&s<@ z&1~hP&rlt3V}WU_-mKaa5j%64`9EyeK$ZkE2mgl1aylb?HSFo1?CqOXtKOAh88BHwYs@PLpv*RyesnAMvURx{Yi*2;(zJB?VQ{hhb9px;jX(6- z&!6h%uE!H(&-6dOcnS|}Pycw!j^{T^(yaq(tc8S)NMCUz)?%lUgjuN8(h{c2RTW2; z1Yp%WmKv0Ek&yyAG3cz9t2D@FXO{$g-pox0keEYA}s`)6ym%vk~61`z6ZC{Vx!4)Xp+^9PGw=Z!idssDI_ ze84Qw%3>D<;;LgR);3ehO%zF@Vt^?!9Cd#hWH%Pn(f#vpDt7vJLM}N^xG89buqwdW z_Qc98*l%8}Zgld?;g2taa@^I4|Re7M^${0HW zk$7Vxdi(ZcoEpFfq9)i80I$s>r$SHMCtv|3?t{dY1~Y|>){jhZ|z6Z zlXZ&e(L{h3@M+@rt#zQjBq7)$3p71ejmijz=&5Iumw`_9U(eI&p^oV$ce${N0~E!) zZQjtri9}!7AVY^uI!F$u!^>B=el9NoPD$7tInfYJ4ki-;fPMFHetJ$}?WHh*35(Ny zHcD!-7A;Ft>Xo4ElpZHdK!8y$I4d;d$#+S4RK(;2P#fWN~_Pe*Ni&$72n6Xiof}U#=Cq;p4R0nWjm{uvMywYnn%qsCsnZIXF zJ5*om>RmSlbuZ{poD}p%0WT=_Cp38Fgd=0u1YKGCrfqUavF|e`aup+_O#k!$1ZDl& zyxBVbojd-dV+L7aL15BPDeqC_4i$4wTukqdI0}rb<)Yi6+rsndLK^GdSA_$at7bcO z4LDbINK*1%i7!r^A*e%Hv1og^fxaQvVUdxkx)rtXAbZ2nVXM}eSqV@Oe}L4>cSPlc z=ZRB*+N3%-HZ)Iy0^Ga92g$Z*-6c^Wr-o!}>gmfb7f=OMtEdjj@XsrE(|z0$-~NE5 zOVjCI&#ITpMeF_7fZa%@*1h({-+wZ{4!Q;lsL@{)6y4z?D&BED!pp-&crYb4%LZ!=}$GaoNT&(;o(3mKdqRV_3|E%fv6#% zRd@o^$GBj4;3Rl6{4JiooDsaKIosd)SCJV-tG8YMHA!GGia;oSOi*wKrA(v97LY|V z)<*5(-Uim}23Zc>yU?^Nza6^XD=xf{3-RSHc@oHio8)~bKlM*P^l{EG$HzE>IN=l{ zd`$o02P^kltM@YTqgcGDx1sghAVXVY;@;Q>WMfkV7wegRK76e2n3~sDX8)wpy7<9H zJwd4#%22jK(SIT5xU238>}8>F@oW2{TfkvS^xEwXp6GGcPH8bnw!p>~Sz=QZpvV%? zAlT^xE{*SAk?yeMwgO*bA`!^6!pgZHXmTQQ!mFWE5Z`m)>^%R8Gx0YZHyez;u9ki4x2{9+ z2xMu7UDhu?70^hp1{rd!1V+9yy{aj!C2ZIQNE>{LBXN7s;d|T{Mo;7H;&y^!Nn2PF zuTO$$4J^;O&woH8y=o0<6z$^TC&vrx)u=DgNn`U{@yuRO8{RalgEY&t;A>Yr^OU^8 z4`1oMNjO3jKz|3#VKVaKqIlaf- z;9IU90$MP==4hY}x9{t?D}!>vvCGKQ_o+E|CsJuJoK&tniRz*u_BC@K!4}`|YOJY3|6D3)2IZFLr`f z)w3py2{fKHj-%gW1P$u1bL!l_F&gc8>&j=OhS@U4Vu!fhVtds@DPh-smWt`3tJUe^ zb+E`_PxVE0p?KJ3)xuQ^JLV0$q>C?vULw}t7p2c(Nv&iTk4)MtX4BszmMk(0^Y90H!7yoYXgKNPAf zj;xWGg!j^+UwemVnsiXASFPjY1&#D)a%1HXqg*Q!5HfN1#=VvSQg`-#W6_JgIw7e8 z)I@YBF8kHPO0(;=VtEb_Cu56Bd*FymvL;!xBB;|RVHR%B`vkezuYkb&J?lC#fXKABz0$OKLM#|s%hc+q1vgVK-vb*qH86_9a6mW2wx#IS5gpI)1F z$@4?cA~Bq9k>UpL7N?!A^5}5nM$h)#OavdcM|0de%0cF&DHh_HJw2VOYc&Fv$Vc?YCD{qtiali3c4#fAiA z#K$C<3MnOs1LspQro2(|D_4b?p!8u1!Wi{aA1)bk>l3s?E$|T(@z@YqPal~Fq*6Dz zvC%0rM_g`r)(ZyRxiNeH@kQi^QAAfhw@2ox7r}_L@3dcEp~n^;Rqr82-W1G{RndbYXGWXG`>h$Yq>g{P)6Uxd{*BkS z%v$ofqu0YO3(c^2yLjdw$T}7mH04>~bvva@#kDOaU%fY?MGBszDS5YAu??!%uk(x%iVOo3eN83=4guKOAm&h6*mEY&=$^BNnCV; z3U6%}ta97x)z2*nC<{9${)BW##cMxscO0H31D?kydUC)RB(`o~1Rkm*&9Lc3?;UB8 zri-&BI(>9eXgl2^Jve&}M>n`2F&Mc#dZUo3cf{o|6rbq*i$coy=}?B$EgMn{kYRbU ze>)95k+4)b4i`_)%cPL<^l%P4!w4zUd*jai=0*2b5~GuNqiI}hUBrf#G-yY}u12~# z;F8Eh9EFwsDKni3iLrCY7#?G9cH%KJc>JD2ILnBN*XUXp{fGab?k0M%E3As8xGFX{Ew9Wk zKPpzZIpnYZ`~PZ#%LUE_(Zuj$v_Fo3iOn&tZHxb{wbS~vDPge>V#75>qLosw6mFR+ z5Be%hN62Q`DjAxxZO!9{ZtvtjzF~?EY#WcC{XTiqi@w!u7aXRWBxeLW!y4%gQH7qB z^gXv`ZDm-8VsAvV{ETVL=GC z)$ZoC<=B+QQ;Gvwaq8OD3fRYUOQMRkTZu4F1<71`fnZP!Y2O8f1~!zl@LU zO-L^W_4F6QI;BYG0YyYXZIVol^MS<_QD$U06-_w!yX!ZlOom~O#r}s4hPhR|O1i_d zkRBw>+6-l(=LThEaE@lHv>x_9*PkCY4v}wafy&p~{zRqG_)JA+AQgyx`iJ&mO)N9S6 zE*cJXQAtA%|bFNEa*oz#pjo{5vBOlt_^EfCfUIibYdb5^2wLk97>D$`mkK+NSeod{#4= zT?va<5F5LaKE)1C%%pVF~Whi|rKn=p&67p!n#IgnjnZ2M7U*KiKIXwj)X zC1SJD$P{dDBZ)IefyH-IPbs%iWGfX@#_x(Kk)T>I?)i-bt;i*vO|Oo^UH?(_N^c_* z*4~JLNJB`T@PTtn_(mUr{ZIDEw{-MgXZ~dN{bX&a@;ipXt&7Y6Nmu-99y!3`Rpny~ ztBoTz(8)aquY%yMOm^NL}ti8ekvTL}uk#6w51%idSnpHC%DiS%LlVM)` zjH}Q&)8x@kuTgf(`sBNVChMe31mI7g{*0qZqMs4KsjI*F(Wd`f&2?B8U&TACNfRV5 zE)R?MI7?EMu~0XJ6jx3Rv%1z(<5vgFYhtg^>*XJnZM-(|p#cWXa zOLM*YrK3Nw4ycYc_-2Z8xhKfhkc}bf{>AEENU&`4->yCse3)uc=5mhNz!g{pcZ?pMFN@$=i?y!{V%|L(@mOedP*LyFb) zvSuwbX_-H4C>n(_3zZql34x1&^>~J3^6a-JVriK9Dt9(7`4{U*3X@a~ix(jqahF(O zEB}Bmkk05qZ)X{FQb3dcvJ(B=D% zFE=)*I4h->eEaE6X==a)+1b$bqH6U{xc}Eou4Z?oHO5cTYou#i2AC+68A+xe}m)(X0@ri=kzvA~=WJ6=AA&hcJ*~b{w!g zBTIU=X3e5|PUS`%J6}xj@&A$cC2&oqcmBTO8x&Vi> zTq6L}tGyIkuMPu4>tM1_pEj=WE%|}hT}v(nn!r=u#d@<;2-2-_O$cg{RS0W>9zHK?gcCCQ1{P1+hbm1U?n+^pGOqd z!p|VkG~*#01iSORS(> zSP|Oby;#~#FCts%WL`F1B^hwRx{n%iLWbKBN8rgezzk!&?Ff{?p;!UNc|AzUB*7=< zELcvJ;Y$q zvI9f2!Y}phPkR*$jppU*cbCi}51lyTvf4zUZSi;}ZanFyRQChE@G2#RLacw*C=S1= zr5m9V@}U1#3B+2sRZxFY|I>zVB~xqO2D#(FPTp?$XY#e&v*Hr~D#Mkna7E)+DBMot zI9EW}5bwh}t-(98;f|ijLg8xAxw+{5i2LH%rEn!T_{E*;;dQ5Y6X!CgmUky+=+Do9 zD-DVp{u;jr!nCP_BUj8zm*Y=;0QA9z-=+-Cq>#%Tjh}YLi*ZZgY5W~C{Re-?frw0h zrxj{r2gBRbeJ=PLVfdR}M^=$-P7K3pljXOVVt}<^C#5>(yVG-#V7F*3_d!JY)a|a9 zflbTOx%!x2U3h!dSIXYi&5L{Q?k|eIy%l7-x4a$q-jcU}bulJge$jh*4Awie3tx<5 zZ?|dW#>;FWwzDl`hnj=G+xcDEh?<^n`<*87PK+9$A{Y`DSWPjjD3S!CC&0DUD1io1 z1L4+5LDj@;hzyTi`Syz$v=J099vGt|R-hQKh$#4Bm=P3P{Ib@P!U+Uw2!_1(6%=!j zBKx6$BqW2=%wH0eO(O#U%8bC)B_G1%pDLR9$)pWRcdeSB_{|xS_rpsc=e3iLaexh;5uFvrp=) zU)p0TSUaMgX<_*Y*D+iB4=)a!ruEfnvbNUqh)ds95W|O}u3Kt`N|o zP4mOfkO3FNP7dMQ3>t4^YGc@e)*f)QZ46eR9sh?*ABLK9=U5?~iGQPwNoEXR>tiZ} zkI5Y6%L`b&2QSI9Rbxc5m#n&i~`L@XMjJ+58m){6>lkwbgygZPd z&#;;d=dB?5#W!cX^}1&nbHX*sVp<#BODecI;b2X;kVr-ni%+mus5?@lz;lL7Mg^VB z!8Na2j)WJ6E2v4GF3;m!0Wr4<8b{J7E|GO^ARv$32D@a{vUa*UxXXU3Od|yKb99>@ z*bCdNKs|2eoek1AjH{3Tf0xfE>z#P@*=w@;Y^NBg)ZI#{w7d?t6P!|UzpPyj{08uK zL(oI-i!piA8vPf|IiWZy=%NpQ?Ivt^PtX;ic=j^M<={PgLwe-f>8ha1#GyuTG*Gb@ zEG&(~3M$lRHz&RoXoN~|=3i4ujuS)WhzV5oQVh`I7h&g7y1W+37uHROBFZDA&jSnl zGwB@Qg{%;&n@Ap~UQurp>l$!54D!Xf^nrkM`Ps=@ff~_sC+rajsblB|hnkPnk%kfH zgYE*Hz;V*;raL+s;F#MN{cB4o7{NJKz!`t`$hSTRAif_Dm6;J)|h{-5*&WiaWkEKYaCfX_3Cz zo;TM#Gx;Tfij_Pj7Fb{j#u-Jjs6<>AjhdrckyfN0+|QTM_r^5Q&C2GCz;=xEky%ss zUo|gfc&&@U=c>-`AG@BAZ&6 z`_oHruYWh=D+{sLtRZGYP%Z!NjK`$cJ&tpEY?uK@0w;&!$T;`|?Uk=a8h1C}*|$SW ziYJgdlTE~7im9SVIi*Vf(cQ0Je5XDp``wlgFUBl^O34!VEcVOxyEeq^o1v4;KL@e4 zLfLuG)|h$-_#gRd3bp6GOyASw;yXoAXFT)ibx~N&2DkOh?3%W0LYep;w3@H*$QR`X zqwIKY@EO>j;^O>gt{M!+22j z543st)R)A}Sg8Wj_|-~Mc`cmt!dfY!daZwpZ2hE%K3bnz$!_{gNO?rg+gs@kL94jM z5jqLp-ONX`bIg4o?5i@puClvsZA$s?KJmTWr3ZaZlsJnaYnQR3a~ z@f#PmZZI$G{*vIvN_~1W{}jykpKr^{pxUZmwwBj3GskteS3eZV=wulH9gU(jVsl`o zXp3JP!CYw)3-nwJ4E@E6@hnW7R(*3Wi*=+`vJ!Cz@t1X zyZ*_K-!m>(kKg~=&qxhB`UNMBsoXMIb8{~m^IIzmLS+F!Jh#62!*w+| zHTzj13k&f}r))>snrR(vuX|ZkD_s*Cs|m5G&P^8^9~FGDI|c2W$87Uxo`W5ZYb=Zv#3@GSNYY-I(eC*`q}C7Ye8M4 zmVY3sLU-Go8(vrQu6rj0?H2XXHzE>(as%t2bq32CO2qYc?`UENgqNI)7biwF z?Jqyf&e+ZTk!-WM5}g%sw9ShY1?oeltbMRf6iY!xzzLO+^6E> z^s;ZZ2h_)GBj?_!pLZ{0muU2nmzV#il`gN7B5@QdbDy2u4wN7EcbkTRovq*=W*EC){@D^455=qS=msdB=x8VmCR+jJ;LIJRc@z=#S1!FiooK9 zD%p}?E$14^7c>i>N!3~ojz51x(0t^?+z^y6+9Bv6t7hjw*Cwu0MZw#-+abDF8I%!o z-@9LyOZS70_r+-`ymFsRQATvl3e`Sn(XDhALtlJ8n7t4|Tk1QdN zm1P{~Q?fWK)Cjf2^sXYZbApBFwiAbDPMg55oMH}8WFMtU7C~^1homq~BsaL4^T7Lo z_hn#bPH+Qph;;c5!B)mlcL$YI;n=%SN$U7e)j8*&YrV842!g~ux|!W_xc!z_CcQtT zII=~az-!{F4+SlSI7~C_&(k48lES;pQ3LN4)RNmnX~Hp>vV;I|n5dvy^$t7KM6Gz^4jeJ86`l z2$RE?w~?ZU9TwQ%{ZT;zBRATZb*VvD$XX{}&Pq(|Ne;z;c7&Ev#Yi)g=OJ4Yva`qt^GN5p;aBm;SrmUm!w8u#wZ6y649g1k8Lie#^&6}Lt@ z^^}99NUR$U3#$zE`|9sqkr;s@>=gbFNpWJJ6q*15b2`FOu9UQxgmhY%fMt& zC3M&Co3S>yA*M*VmRla$;dO%6aFaRBG!}1Q-FR!*;5lI53`+q$0ZcBfCavK4Vl7C% zs9Sc<`wHa1w{dIv+dMFS-4K((IU1Bnw|aJX^~;|Ejb8ceQFf?|6nM6{%j}?LdN=dE zf4CbpxGd$$v}ErDQg1Q`$0!C$45}y_wj+`SMbP;T;mbpP!o1}}r9Js}OlOo^0 z3Mxc1fdrzD+YqxDI&M+7B+c)(w3+{TVE?RsS(i9H94Y=DMlPFBz-{n;Dz5j;l_X4P zjmcJ?Ql1qrmKM3)osqriZ|OAFk5D1Q*~)V+XiBKx2rZ?jiL=t^eUl$8ux zMEpq_#ZumSiK(;h*bk5`hVJ=2CN?bL>a?%F?roMjdJ9>tXPVLPqIU!qM&t*p4||_? zFLG;DCQg1BIpAX0bmKmIk4WbO`>8(nPWQ^$o5IzHf{_iTR#r-oVY<&Fm&TAyp8pc> z8(?+uxn`K@TInb)sL8u-cQlxcKJpi=%;~t_$;bWF+i10Zd~V!Dl3^+G`2(nn+!+$&1ZlsxmkJ2O^L`9_QUQz+tTRaQ=g=pZ(JqlZ_o zNU-l7e{>jmzFg=JhJVj&zr#94EbQpV{ceNXbaPEn2&fE6QKwVPYKp9)RB3*>&{5Lo zevZx;?e!jf+o4t}a!pjeecxLf{#u*gejCQJ=Pm5sey4SrdmFiz%p*N)@}5q-ELn*Q zX?cfcf#AX0H^sZ4puAT4%pV!*Ptn(eYbMt4nuMP+2@@J(65)EI{4q!*Y?=Dlu?zr% zsI(dBLu|_6vyS|}N3dw#45NSYc2iyw$$!oKlM^O(?f}K?qeuy*0ua#UZo9avJ-cQ$ z#5B8=id$p_?rW!h4n$gu{crITg0kjn6*c1RlF#IQA)Ry~-7G8>V^$N_n&de;K17%W zI(`zbMqJCkC5{a#^&dQ=Q5c@egr`h*Fr%8Okw9b!TFeed^{bdXY~2{%_>b@X@Owt^ zoSql@Avyn=fv3X+Jk1nylOoqBRUfw$+ESj)xJ>$8%0o|cl6ZYI@~|O&Qmo)$lt$4e zE)VT@ImFKt_q&`E71RAL)!tpaewP(~p9l53)CwO<`dzw-mg#qSNUw%b4n%0-PON0_ zcX>P+&(uWW28vy*KB40;pRosiZiZi)@)$WM>f^8V*)u#G?F%w9%)R2J38E~#unw}UGZ*H189`S6 zv%L$*1}9EefbkjPb>&eEH2-g-R9NCr_fG5_bpq1S>@pvy&qE;GqKpKpXU{v_yl*L{P;6FPqRrF;00sVurwIM?Nl|_(ziD#C8NaY=*2kn>9hqFFa)P-skx5Ea)23(*W5iEV9q(u&Nz8-eB(B^;T;A^(CFjm1v zR&N!&ayEXVD0*H`!AL2+afcS%5TWE;I(rcpu>R*U{R z_GPEbtEUYpJ`e2Xt^}P6sJ}X;+yKI%NIwa$cW2gQ?_o(4Fqvxczd-R8LJ74L%pgxg_bJ z8|}tyWG_rT$C{nMWgSylKm73bu0|L3_*+$fA&c3S%Q$h#_*N5#H;rOaDUwX7aPONe z>V&BOO(;r0jt3-ZJt02I8&z4?$Y#n`z>IA4kw10*&G$H?sqA!OSXim-#GTG&uP$kZ z1SL=mbX}*(mZ{z1XX0|VLRtHy0eQOoHc-!K6pP+Y{+^-ZR_lY^2l0Xi4))9?xGf6N ziprtWWwqx{&vQ_Nl}kSbR*(d@Y-IziZg}s$IZ3=#lbiV^;#O!XE#_p-spepGuYnn5 zUe#78J-00^4n_{7apP`uMOA&rXkkR&|2>Y>I`Jm?zKPAaMKRYYa+OlulphSd9jNs~ zQmVzACa)%WA57~VZUv2e6}7x_Ub_5F4D!t!^L@tK${dkF#2o;IukLYHM`l9dTek9v z{_b7e>q3xN08O|f+>9uEwoRJkdEOJ`wm9*wI-jS(cjQ2J1wXBdULrWcMHY;P7<8Co zp+*%+F#6?55an;Ck3%`@p;>XRT>#Ks5viof>$+Rvq&Qd5<91E5jV!>Cps|K8Tfk&} z!Lks~rIrV0J$c<@{|ySzQ&WIzToH=v(F1WjuZ^Tn9B{!iwZb*}vj_bjh*4>Mhxi;c zK4Cl$t7ZliYdo(77f|@m?r!pn*%}>IUW{Y!BUV@$mpAX^BA(G41#kG)39^)(!{fxR zbe4(nT1zo&D3U^{PR+bI_m=20?AD&0d$;xEX$mh!G~I0jsqcMX{v39k)@{{ zcW#Pm$#V1RC^q33Cw6(P6nXUo9D?SO99o?)c-CchBjK?03;kPE(e2-WXx|@_Gb*5%Jc{H{ss`xG zNS9w1HuK{J`GN-ox~fq`={Sv^ODQRG!=C(fd5gTU&l)>k5xLFJwWV)fMJIDf$zuuqsa?7u_Iz7FN7(fTsmHW zxga$*8r}26z3$lhANpCb4HLsmpaU+iYkl&8c`BKWbLzxvjFpUdw>+71*n6ADEx%iS z-9i09z{rX+xM_Yj%1q(xd`$9=PF}fgo zT(oD%ve&G_7KHkSI2!9HCXFJgl&VeIBELGfibfL5hclDsfBUS4w=heNJ7qltGI@LIqIZqRShD8BACO8Xj=n;$bqEU1Q49!WpQKbB~CWyLDF;^*anNk5E4DgFR5_d<|D>O$S`4{HTEc?Rv+j(7U=26HhV#jGYIpktAKN=MUE^~EBzLQ)We-AJgjFh z^e*;!98f;>+!=vC?g+t-$GDhJeGJzQ1i)4*2S5q*G-2C3V9o2}E}xBGcSa1Kog4#v zS;pcpAY+BEaTCK+myS2$Yw@y>qh!Fv0t}tl$pz}6Ay#Y^#UxQAky3TIrpvJkCMUew z7iot6CUs9RQ?{UFmzgmA32KPBDX;U?D7SG- z#CXpmxalhd4n$u5azT_Y(o5t>4L92tMxz3bDWdEO zBX+>~`OzEw=1~`?t-Q1nN2zt)CeEcx1JxOl0`6r_a!@YaMS#@KzVmI%B-w}wTZXdh zJ*+S>&Q0Z+X3omuw3x8Mz=AAueI6aIRZuXH8d*zUox9sLZR)~nK*WpsHf3~sz$OUi zFH9Dp{!K28i%piY0i4zi(^`RP4c*+6t{TNCtCLg~IuKI>ixP6j@dV5-O}32?dg(bwpSb?^JVdFJu#EQBOk{bf#^3fg31Ly{?G1w|4l zRgy!@&Cr@Y#gxS3K{iOX~vLJ~5ZAaaza2iV%$<1bax? zMKJVU?z6@Hz%r@vBiUf`F3`wt~)6L=4&lcyzMyQNASgb?k>o<08A25dNzSs2a zapWnx)x(KB&X#sW5shBT_CBOhAQ#}l63h)Ni83;{a1+LWwv+AtQ~ZkADl|_>!OoIx4kY#7ed-T zTZzwIkpt}obb0DI5rAMP2ruS|ksV>cr3z>y>!8*NfAqz7)cEt#<#*r7;oc?Ppp!S? z5-Vs2(u`h7`b3)|ZMLCn&_vqlyA1%fj;;^3Y|Ve&DcKAGbXXs+39kwDwX7Qf2#Bf- z&eJ*VlTlz3#1N<7mhFrHD})6m&HV1jD!nS=;vhBBVPq-wD{L1;-7C1Qz)qS;FAlP4 z&TV-62XC<-JO)R!&v)zau(W#CPI#+8Y&1x+zka2jta9Q)Bp~e@QWTO&F%aw6K&kSS z8IwwZxU-6v8w{ylc$H1>WXd6JyOU{%*~QJKd&64ft+FR%=kzPWc4-kaHv8cbAjCcp z7-ke!2%(mI{F@KWBO`3W?au3WBVz@a(aqtaG8ih7HPc;m zF%%j7`L_H|k7UWAB_0nQS}tTZ2ewYO!PhW**nuzE8H{C9d;ZVG2#ta-e)$QBwdCVE zv02cVz$JxZR#Id+r7H35mRE3qPb=OPdCM$_;1%64J%)clqj1IZa^(PNU6!kHYXmZg4JW zz8xctpo>{SuZ-F_$KJaZD|5&SA7i-uoV$MPe|(JiP_9g@B%3FYgC=k6PKtr%w_HlK zCN!IFdGFHO8pSqo3kS^2Llf%adYX2UBU?7=)sTWi|r8r|d&2bflhx~{p9s7d-4m(4@)3S3jw}dY6mlC#&>A>toU3y&U$4e-@FMa) zTqC11KO>*27k0+Vkqr}K{T6mp<+O5NWpDnfhQx1}m>q31%E{7^uLF577DG z1BUz5>m~QXFeA`PBL!T&B41!=DNK|@?f-u|@Z)3UQ)S!fX)P%#bD24(ty8ih99Jhz zm#>k>3G(FG^j2uG$@Wt3^~N62b--wnNyiGlkmZX|h-ZbzQRP7;Qd)y3`RL4{b%QkI z%E%^t$N&yHWMwnQ|L%Jq*D42nAk|;ikUi{7lk-+>l3J5hx`tvZDRPKXt@8WAyNW(5 zt_8~37U&7QJWZqMA^FocNsdacaWbc!ckg%UjaUIVoUIr=4#jm$ z#`{JP{m;dWZ+^`k&uCh6mQC z-$E#1yxTXXHkqdvofZ;S3a`5(G0K$a*#qM1+r&j&+%amD{W8adkX}8Nwx9w&;;$aW z0asa}VqDn&`*7y#9>*zzBvG0lc5UgmMTeDLUg|YbSU=lKs+0%B^}uvq5gIE{cfz+5 zBtw!2Vy9}n&EW>cMvz!o3LJopAw#kQiR4dberry0;sMG3oHN4X^uRk|S3ir*Cjpd#;Z@AKDp`rA+g?{x+6QCODFp z-twf{fC{~hu>7>+&ay*AZ2kMI%u(^>u3$nWH8y=W7vVs7J>MGv@O zpWu$knkk39OF6s2RmpX$5ZDa>`Pnt7eE@a@ezJ(QO z0_^;3B3K`tK6#I#I}&*VHcw5TsQ#NU1KU2+j-6Ta(OIn}X?U zXkNEReN$sz$l|p2#7cz?wr1!cHC-ny;N<(?7VhGmp!aY#Ot zunR+9`j#xpo7U*R$}bC8kOsXFxciu^PEs7v>WC5>`ygO6b8Y!qc6R1JChT21!)RyT zZpup{`A%HOa>B%N9H5we6e*!ph=d;U7$T(iLJD{&Lwd{?q~4%%0=mICjZC3mhvWd% zS(O8ED@NE+*$|$#o*ImthO>bA6VPuc_3w4!CmFIBk;w8%LMo2FJA4F9PwH0c`s}M{8uLh zkvY$}a6ImouVhd}4qFfn3JeA=j{8!58CRCF&jU**Rx-O}XWz_sQ{%5Q_)BjM$0MgL zIADdmaW@L~q%cP0xj*T-O4fa8+@n$xBbQ4tQ2D)uQZ>_`gI1*aP*6LKl+{RBoiD1H zb~_ACV#Q0os2K797rgpqsZr{Rz^zQZqF;7jig`pWuPhq1PFv)EbzIPIlopM>sj~42 z&lT4uh8dBv#V>0uDP)HfC*FPo;r@{AcLl|OKeZn!zx8C085}G}>X+?z)kWlsu)#GO z{HF6L(4NOx6w(Dr<7!SNkZquBgBGe|kgQA%N~b8U|U44Z#&)ZV)$ndSd9`uFV@1LM>c}wRkwZq!yie_SvS~3^_;lO z#R_>yhgU8C0Fu!~?C{UAlCk2q2Xi>Cr+qp^Ha<@FeA%2J3 zi!qM|>6IPeM@0|0Porp)rUd7JKYui+SJLdK&J)B7+CnVk#%+bvP($K~V{?RBcY)3{ zhW~n@5o}k#yJQx5$ZipHVjHyD#4TSuo{1Yz`T>S~AR!H3?0;`=h2KLE;I0B{bac}D z0G4HRnleXucV*xR4$7J;J*?ql&KO zZ>N#g;XL?GK-Lb+SwjpGKBNRaCU<9caBG5&)2YmcH_E7cyfTl5m<>U-d~C{ICu$1I zhyU?zbo=pY+?-o_mSw}8zzQL2lVWHJgcLpMxH{`NVrc^d9L8ie1Q`A+Mje=X`t@Xw7fT(UWSD!J*z z`@Lr-_U!@1d`^+Olqw@2i;fdu)i0E4--zlV1_7aNIc6m4Jy#R;;DI4bkS{2J&Wxg{ zlbqAc?f|IJyUJb6%Oc3st;RdBm>dEJjS^&F)+@@OD$zkDmT#4bbUu5%Tf^$4y8}`P zw9~Z6@A0xjfdr2>Z`7N+SpdttdaRJ~_{!b~@aB;iJK z%hX%p>cUC4BJs{8LF)AIe9-hfNi-grP-kdj>YT;M84TwOQ)k`CS*Y3?SD1SKp1Jeo zwBbxE?o^9>zk7oWl~e%%7_2OyGNc*OGZCAksvHrReDPjeII`kZ2b3F{7}TI;5hpd!Fxe|0l|JJt9>QD8PWP&wht zrXlT64$!03N#g~(qd>uTUC{2i@a>sbC-3oF7)(Dj>o}b`@iM(5>I9uM@dT+5-f^#( zcE=qzeKkQTylye>6~QP$BdhX$K&Gff+yVMn-Li(D!kB`Xjp0q9`bJ{3XD#w>Zi(;1 zX^VpK`+$o!+z^7w7l1mKBAK%&_=@m`xL#xjCjtSmxF)t@qsPoeVl$SLbGn zHcswTS`MY!3{FRQV}oN_K>3C#U;F)NqD@Zhep!h&b;>XDA5YHnyE7v%D2}=1byNH) z7gT^bEzrHJ6>SJ?B@HpP(&G?{!cE~SKMW~!OiqR{!g-&>s69}Z>e!9qD3D@de(iV4 z?H6&|f{cb~Py6n8k~@JMHTgvEqnHwk?50!)=!cOT6o;bE`<$ndT?}6vBD?v6eTscCqu&C;B+!JZi)aBJ ziXPIyl=>DzkIOcyT%m4}Lma{{1(XwxDf2~0SfTIt#03rS%7=J%xz7^s0T=YEG0%re zkTBduSDjBYe^qqi#7E+KMQ!MP>GpTJ`O9akE4b*ax6&P+hXVJ^PMl~EC2ohMa~NY~ zN9YfJ1Z@2N>Q4$Q=#RVc!#8*`Bf>JL{Q5T}{Y&FC6`SBOn_|Ef&om^4v3WjU(C4lV zHxPt^V}xF~)>Vy3G%QoYMtj^(>9Zy}=(1{;Vh$NIC&S?4)w46))zP48xUolb{a35^ zZhhTL0&s4TJmIz%io|LZXdaOH`aWpM8Y+ac>E^Iv2o~yl?$j6$!0um-0&}&4Pvwbo zf$np_#jpvl3B+n5JyjoOaeIMEcC#H@C`JPZi`HZ`7#Q*_J0tUF`qC+L32Qdl5+~l7 zSrNR(SauUdx*a}q=()DjMJOK^ThowIQ}v6dEnCU3_cd6 zY?7c4h*dM__c01I5a^sFOKu)@Xol1Bb-eT`1?Gv;(6 z(O63BI`JxX$ix*cq!{1_+z$CPMN#0M$r{CKUV`8(>GM=?6DIZ0rX1z{n*K3JMx zY{(HP#qOB3LSE$A&CM5JoM^#NbMuWqV`4GZs;0-S((G zH)l^a!b~?e>?Yaxn!&8h1ZF!ZW;;c)DOF))6?il{df~!_hq*evj1};!O@x&r8~xC`+-)^!k?r(c1Qblqc;(7We~n_7AY0%#S9+8<8YNn+ zKR!9Fb5FR@w#fchy_xJ{7cF+;#qErV)v2Ty@R`e?6U1*NlRzry+nAHs>y3raM>*PX zRBovBg0x0}o^Um3&YIABvx+!c(H1B{d*+?(r#_`z8IC-xXntY^mnIj>HGUhsuSoh} zQA&-Vx*%E1 z@ZYG2hXm$;%h8|$dUI4xIQFVS9$S{;iXDBpx7q>vLbZHMbs>RIEVYW;=Z*nEL+>9p zL6`csDv_hOj$iECqSzCn-@er=vgt~)7bMJZ(K6i30;#OQ_KKxu_)E3}&^p+b#NPR_ zw-H!BJ~!?n$#CK&X}<}G@+l^lB3W1&UnS|6X%y|e{k}=OTK+!nEio2SoDaPsY>~Bl zALOj%CIc4-w0r6AsP=4;AM-utrQ=`b)JtK54Rz}0yl+TYOtkH9z2VN$0fX&Nv;BVS zz_74NQU7wTsNpRmSQ`KFZVoANVz8Vy0m~7JsisH;rAihhGfmP(f?cqmJ`}Vx9P4P( z{ID7V^N@q~$dF9@JTP?v-{!HKbK468G!JK8ojdx7;OJoVN{eDUW-Q#v3P$7KKlfMn ze>1{JAXxJuxjcb9GMTMbih%{AiBe@LAu@&x1t3g)z8G-9P!|Z$KrutPct;>i5bvg+x(rs298jgf!m3OW_7Z4d(ZEpufJ+mU zFAOG&aZy12^M^tt7*;3B6Xn=p-UHPQ6}0|>Oy^S(S|^R;;Hz@b$jwqBE56PL==p^e z{7NhuuAs5Yp;6N1wL)sWoXUX{`pVgUe$m2Y;wg3$>aWB~6=3xx%eYz8R0@;HZr4h5;(>Fc2h;mI5a zI+d2+3tLTv<)&oI4v_IB_n?iun8XxAk((4Kde( zx=1a*z+Xq-nVs){c19UhD_l(qCv|h%=~Yy@TMC5?2(7Y~u~C0#fso_Va*`?RL z!5eo)*!GIm*~k$AL6H|Rma+DTY{shvOgZutE4+>i-rn)AzcN~nPgW(oN!r+DpPhH} zMV6Wnp7c@-1cSOL)j_CILIt`$d7&^l6pO+&N>BoYQ=HPka$s{qNx`y+E3}xUf{Xh zGWB5KSuxgI6>$6H>Pwy)zhzTS&c>8h24t?)Y0(v-rEqeC_kAcHA8^4E_T64B5!rM+ zgoa`TCtdaNsFjRPa*R}x%Yiwr1O94^G35*J1!{|8m1qa8Z?r-Z#lgSe2oHW~rzJUg zJo97u>#p{JJ7X*hM}iVV{C2F(kinGLV2m7TbF7Zd${3CB?`-r5G#aDe%)h3RoC)NJ zi7T_0Vj$2~gdK+%i#QWFu%@UCxPq@pUtXL zo~E@Nog{^KaNfc=st#g^Rmw#FewW82o|h5O3H+iNoTEWG$}QYaZp;5-c(Bv>pd&1d zjd*#r+wB0mQne?p(%gnv`|$l>g-18H0$9LM)>@-j~Td3wNhmt*{{` zPEhW4h<{UDA9MKaTVOFR#_aPy;=PXZF!FP8i8w=|6+;V9Xcx~BpMF!^jL;UC4&s1U ztl&$nD*Qa#+;WHs@7F*3x$4(H{pI}MiD|}#P|mwu2g<0SU9Zpm&b`rR#RjRD9tL9j z>m~PqImjWJQo~K5Z8#Zj6#HMdj+1{_{eWt&`Bi)1OVE+C)ZCUhZH<~0yK&DmUx1D3K+QYph`3Q= z=-9B!w6T1PZJK44W7^|(>vqFJLH_t({&UFO?;E|4Z=a#wCl{R93+XgjOKwxlEs9*H zRGIXrP+^NQU72)25}j2ODcF@N1=C^a;EYyf&fC z+*sJ@^{ddx4rNnapUgGS|GidT%w`kht)iGDiX>91T)M;c5h$#*aM}SwyXG{GO?}c! z6JrBPUV3IMPFP3DqeG{I%v<4|b}h2f++IfSbVHH2K^ol}(Pgn_&YqBS?KLw%^uotythTZ z$#om|JdG*8Ci%WV!$=stujYLw#5)>d>Ny5QlxO~Vg5$JS(Mw*r)Yv|U_6LydKl@d8 zvH}owIG^u5)#&;x{&;c{$^6nJd&*6gt6daRK#@F3b^D`S8tabmUn*DQm%&e$cL1>^ z@Tr0{4pg-u9XQD2bON;o6i1N6iM`~hCvs~T?#BMcYe1w~6WGS73CyJXpgh-3kTRK8 zyWq=YMEksGnlC!@Ya?RrMz5Mq?mDrvyv(E~?hA_Pp-4BjIvW&`bWkvZYP*}g%HV%! zOhft}B#0^o$92DZnn-P^o-2m>8psRY@fs1k<(EUh zxIfEy|GF8DShWnZXA4?h;a#(X0E@e0y*5hBX*$^04<}ABSS9tq_z#M#>WNt%3R8G_ zoL++BsVEMpR~p0@DbtrvS9iD?r7U!Go+3-Ra?+lV0hic`>UdDBzX#-HjseUl!Gbtq z0Rh`@{UG;b@as-97AxrTx<0LQni?6#t|%K6>Wk5L+}20O%GA}qE%G+$F`)W!%(Pm3 zquI^1#UUJWl9d}W{#{YgtkKA}Ik6jJMYavQkI)sVSE$hq$)t0_E4Vq~11?C@0v#a5 zA~YgM{c)VrD;aRXuR6(gPO_fs~Lka^f&}jhCvT4ofYvf#?8`iTYr0p zd}@hs%!%>vnaL*l8pT|p$R*rBAtxB77hvP8^8tlN{f;Ud0(F>{!Y-im(Y5@1|K0NY z0d;UQ5{p&NRwJv_cKSShMseE<$Uj7h`DEejlUDaB*xbQkDq|T9?cfchxKt~@`g3gjqeLt)wP|Hbj zQ)>;%;1aw;qj+Tr(WpRVhAGxBUjahv#wo1(x31=0^iCW8v(gt|JZVipr#n(^r^|PG zme2!l)$6CMMIHxqLyj~WmfmukF<|Lo2cENz0ZU=w2A>a)niH|SoN993=3QcioaH=? z;hFpoPS4vR?xN4VcT>EbKK@P@xg{QJoY=M_c9d)SDxg_M%C}nAH@)sH(K^UwADFU* z6h!uM)#)G_YlwvbL0FVFe5^t6VbBvbCea zxjKyl;zt6|k^p+kS2~^}pSF?1=sB)N{=;AKKy%@Aoe?77A&s&sjBs1~&N zbo678gh`l)slxKmavxMxY?T!Xk&hX()9X29frpeR8iC!|_ueS7AKSYf0P!3dhJ#nk zGK)*r?TRJ;b<)`f@4xP$w?VmIobV*K?co~;#8Iz$+;K%z82&kUjWL}bHpno2(y_1D zA!Eu6uXkq|4a?6j%-%_Me`)F!&Y75@YKp0#$U#bllI5t2wSfWO1^rzN=xd>jvN!U$ zXx;R99?GACh1$m5@7@4iY3ie#fmvsjPxYMzYJ56l(k0UAo~1lZ&ikmb&KsY+9JvaX zwPreB6gQ{O-Ei(6FK#l#g`at&QL70o=Z$XFVuLs{qciNw(O@!TQSGnh7%}9(XvGe) z|4W0ROC}gPK`}=uQbVa8M5IsNqkwWL$sIQ}7PQ=v*YI-bebKey>Mp6qZ)LbfSxIot z@uC{53@SqLaF;BL-tD@Z10wfb{GGmgqd-*#@wZX2BXD`lj=;m95)0T%=dKG%5TueD zzMVk}5PMi7j6BcUa9AcQfKFQHQzxwmP4Z6hj0=Y`Z2RwpeRL$0hQ->;H#|q*_1x=m z9)srSI+;&W*+I#PLomBcptOZzG!)rLsUDMy-YHPaiC^nuF2*#(VD0bKpxyL?i1Mkq zZVfS79-gV?w+r#ij;LB`uG@>5eE!zi{a{2GvHNdfVZh+CfAlkAq+s9&dx`c-gArJv zhNw9fQ4DZg=2NPpA<4Wp&z;;H?!#%#^d>!nB^La*z;@~;e=+BXxPl)W(k|Q+utI*( zrvTQn`77M|+&6l!^4sZI7Cj#e8E}jekiThFo~MiF?utm9=l~3j2r@630}E3=%2UbP zpADGTIXP`y-%8~Yc7e3WwIc9HTI8rT6FbKNGdTPkZOa@FKZ^bDVCVOK_vVK`nqo8} z-`w%hFUh7ajS)FuVnhlkCXXUHsG8RW&4~lzCElIxl~c>1v3eU98yumQr?wj4fz{j8yMU<@bQrsCM=ANmo`9s37aEivQY8+UT!SB%)$IgQgv zs=hQ!%ncJH)lv*Jvz?+;jm%EZ8-gt&>^*9ByCR%_XSPm~FY0j3q%U)}dz=UaPc)CS zD5MJ{Itw_-B8=@K>CYB&D}0CeU|@sy6=6G&{}ga;MD>zn5svYYk9^JhK;E}dcHVRT zq7W^pHSXf>lb7=b0xv|>OLK$EVe9=M0?S#B`6lx=%C}5K^(4$~r^|N;_KaY$VrOoi zLy_$PYXpW@zV)luooaBSd;>Y;-yvy$hQ)Fph`?!pb?~V>^a!1xv1sYG*BJ$3budn0*(oyj>J1a5D-|pl$;JLN%nDRa?E`_k zd_gX~m$^4c3G3-{Vc)7vp@@!x`OBG$?9G|E4ItH4GH^f~t zPmtq*Mc~>zcx!L)ABzH$MSa|(ks3v*xF`^n`Wj;TxW{Q6?~iAGq95`PtEi&D2jX%! zland>lMOLN+-sbSN$0%RhS!GX&^`h}!N^Id!L7na4_IcR)&z@f-&t@WN8@_0XfTQc@mz_DvnXRaPSF z6*t6e^uyPv_WX=j6wu{p=TR5uRvnb-pCdt`!@#ONRX+ zI(j?Z%r6Q=g?U5A2&M*dp)783i28wWg?uq*mEWfDYK~)DgkH?#2%z)gVFw<)_3$4x zZbqx~C-tS@k%S2l2p{@EUQaRU6j@EFG#)*X72GTu2m{jPs4R%>EeUR@yuDGNqcKEQ zK1MrH!yCftJ$2xFXg;*{kE12Wz-f7Mi$80dZeF_2#&$UIW@)Pl!qX@wl_JTMsw}$1 z|4$3gy>mm{D$C(^M>fQq6<>jV!b{}O3;((m0&-*F%?zT+S<(HA9~5(yMvHMu^_Mkd z54(`96UR?#O<++&F_jcKgi7{%y$u8@5a@@V!Xn=VSa%DDzF1|eCvY+JA|mb64bVQz zRs!=OtsY#9&?mmfTNA2I6O~0bN{WIzTs!G@DYORSaJf?!OCeKKCr$Gkcq31dCTph? zNSwexA+=Ywbu`p=tUzRgYhrrKAXM3Zzx%t$AH432sK^$l-Zg0pcTGThShsv2EE!_# zjv_30EoC5&Qt{`!G$E}C-!Dr2!&wMw=m7-MV=F}<>8{wY{Hj*7abWh)`6 z?0-UW3wlj&iaTZ{bIRQ+gsc1>(MN!Guw%+e9;E&_8JrHds{+XX6Fqxi>xk^> zm(Ch`oGwSbx;Ak!T?aH9nRF5}cvnMArW=x;-hbl>X^25=E}eD3>$s>3l~MRuVmo-} zlAtPZe>&+l@!%P}(;jHEHRx*PUH#5Pf%$&fX|0fzjXcElISoX8&-2pc zV&F(YF;8UA#RiNXJdF~(W4GSGw6HTSc9}<;24Q7f#$Py~`@?UH{>yLcejZFd|I+v` z%S|XPpHj>h6zQQxjA&R<6x~xb->fc@YE= zbHb~A4YyUmPC6&NMShas6{T(jwdYvD1>i|(CW*#Ai~-{|tbeR%=J{ zFfv1|P6oxSqevQ99eJs2bx0C#yQCpzt#@WXCQwjnUC)v_>BjJem}4X>{5GdeI;ONx ztia(I+FFb7`taekO}{sSokuz?6|^|9aad)t>xmoB^i$*sr9v8ijpCqkX9RYg zw}HZ4TSSHwi2U22C_gtC#T-DpLAb)>C`4WElJtqV&A0+vmn)>#9bJx348Rhb@W4_M!6Cc zx{%s3n=Xrncr!c-+pKAQo&zqJw8A_TN)+S^oObJDeK}M9`otSSMhxz0-yKhKzcdC4 zSQdsDsC^VuLXq8+YP$z!v76^DTrwn$z1;(o*pI%^FfSR(tafp$=uE$D9xY$5)xXXb zZ1Y(Cje0n)kGT!}?S=Dh%R8W43mfh6xgOzD@v?W4Ie#3?i`#SQV`sfy;VR)Ji*{&a zCmS~JTVqZ==Cm%d6(V*+ttOJ-E~ruH`<+ z3I)ziJDnQ&L?6z1{vum|@ev1gGTLGKM>fC1JZ0UaSn!a?|8I-i8~+nYb8)ww_Kq5B6A)bXGzYJq7M5sv^ z#0f6C>U@x-Tm3BTQ5fEy99r(SFC6a5rOShk2JM@JccudAA>s=Iyt%+JrBUEe#gWf% z#*l5668C0j6k1oj2+o-6y0;cf>*!dw1d1gZSwd=<1?W|m+2PGPq<*{q-`+KEU~}4) z$x2IFowQ!jP8;eU@~v*6<2B1d^=+(U~kUsnV!hdo2K5*t1aK=3siwl3XrOeE^)HqJ6nl>9E>cF}Fp z`(edOJij;`k~Zg{PxM_+M1tmo8(*EyJ5$HKY73z zFSth1<=wJFoFe#@()>>3M${IqU${kH`ROY z55W)Io!J8}z3z4JQA7fTQGo0O5+X!%~XsM+t@5L3#|@-gRXgoGbE6cLotvX*8))p@W!+K9~C$cb9Hp}+~u zJ8WEjG|kJLL(^&ZSrsNm0iBnfZjgGwM8<8e zK95^aFtx&C$t2jyLj@Hk8E}UHS%Sz`@aB==N~1tkEyndKLJpX&M zia6as8FNB>lvgY300#M7B?k8L1;}w$L03$>p54yIiOVbYnygpbDF!kSThZz5b5Dk7FS6rl!GEcetm72Y zIetx0BMc<09veOSxP{R0qUAKx$)TnG$7wuwg06#I?|dln^-rFY7&_l!u(EN-V*;h2 zFGdQR;h$UZA3x-erdivGb5mBDwc`ba^1{iGJ-}M;W(ZXFy4UiP!KiEuKj-a01X!F4 zo9&IoryY0lUrwvOwa{omu6}pPEb`EaH^Qq;qH>GJGjZcdKczzU%OvLh8>JDkf`d_M zvzN|#`AmJxg|Ba=+P~HE!^(FHzEb(_lebU*Jm;&KUwQl$EMh1nrT+h)y)S`l>dyA} ziZ>)b8nO{gZiz*RAc)FhsECbiXX(6Yr!({BO=sSl`M;TYZRT}4Z%d9{nu~5cl%Ae zPqKU?sMX+agLl~pYG``?O(0cV6E6HfYxm2v+tM5cA&{1+Z zxQkfsE97nfDp{^-)8u_28G=Rr8K6JENYd<=DcU10^nN~~)>L2WJ7Lyj3LBgS&{9@m zd>?q0rVjFHs-?=?(s-tw-a~dz)wcMUj+PMUBfngVD_-dFJH*D)Lq+XMFK7`BMFh@S zsKC2w_sc%q^jmB@=Y?nGT)3{m|HYxdW?2_3*r)+$mG^m6_(9?H3nh_tN|3;n7JH1B z8mi?O+Jy55>d7^_e zaG@Q8JMTc|CJQ$8^uZY&uoFaq@uH(sbHz>~W*O0G7pg5R+wc$01GsHe>tsK^Xq^ml z*$9k{^iB(%8r|Z9vt62UD^#lJXx6M2$2@v_hTk#Y#O^M(hkt~W`9v>q@4s0G<=B+3 zxbQ04MhFjy=8bBQqzQnsTW%irmZC$tCh91;>JuYu5q{3^3+tl)e1o*e^<LLybEk9u=zGUdR zw?C>S?0#6bqIJ%8_XZG*UP-DT3bQHj&*d~;<;hd%vN{@4YE5STP0dd^&WLMND8=(ia|HzWg<^xSKtPYHdSu$Jwswb zkh-SL7el}K;HBz9g87=O937qH-@v;-Zb(a`@ND*8s9i(%)eJQHEtccVZs=TAp*7U) zk4IiMW2)&}H*?5tcJq_0+7RGG(eSme_K_Wv;{;8T^^6vKJs#$z(>YMG6A!KHF^uR;C|D07D`;3> z?8MK@Z@(wB?xe;B8ZPXL*yy{4CIZRY=Rk=8`>b@zTje+-jVBt+18N^AoH+V6Ze6?Y z|D}(`Ubt{R$A-OluhzTTt94c@=lF|_OrshIxWam7JN=1xwf{;lY?!5|R|bzKLhOTv zc^a`XjLz6yJp0G{p7q|p=o?fACGZkZKjDTHq?f{LrKr?`O-LXb+Y5!MP}{SEdqr3) z#eL&4@T^9B90#72Pdc%AhwsGM=78CUW$G$Y@Z1E~?8VG! zRc`gyVrg=gN(a5Z+Tl*!ay-=IYdmCHK{*DCqXF9pKI`yZ4zV^xXJ=j-%aX>0XC51Y zvJB2O`GFukKcBuivy#3GVh)EE6w&EHxtw?gx_pXIbhT$%MsS+4khDU(fLzW|g8NgU znC#$7=ZItQU~}+B4-}iXbjX3Jy88S7yZqK|Tmw*cA(Gmt?1P?GK*`Mk9$S{GeY{1x z?T?;1WTaW#q&3%Xt~DrZtO8o+3yb?;nSlSa5EN8 z1Y-9XCr=5F&}AoZzj6{DUFqBncEBxL&i{$^V%KGZ>o%6JLv6#2)52!=9QTbxdx|U% z&~}j~=6)E)*h?Z?$E8)cX=qtO!?}1Ec6V0LnD|fY{`$5x$LKi;j%~46%&exERTNoH zX)Z))&w$>x5l7f61RD@Llm0$zXe>7E!;i%!H>IgGgg4bS=yfp-AKSW+zxZ>~VR~ zhQ7r1!yX|c>bl8@YX_&jY>f<;buw%~7I|RpkZ08<`J<62r(5J&WP(AUi*J;eT>aV0*Zm@B!|+RjOw62 z^Q-_q>^5;L2ZW8FHU@etkSgBt7gtY7_3wZdaw~sw_*=^+)xW;^jqdk0zYzxwg04fK zg%kf&@q<$fAd)qwg!851ZbYT?f*CB~@(4FW8_vQFE4WZ^Zl6&;)eM)1%&Fxhja_`* zb=T?82Y$FpUlGORQ)D}(S@PXxh|)Bwt9X6FOwm$a1?YmN(g|AfvOvcrlk-uWmVAUwi}2}$)w`I0h^PM=O$H88rJ;=+s-UI;U1 zKDU>oxOLuZ zpcAFGf{arDbu<8R1|Y2OS+f7z`>pGF*(`Eh*rM2|{@w0npgY1qYJ~b|GxgElZ$12% z^sje(-ziqX`e9pswht1P*Nn%fZ_SUqYn_F2S#Q-wUhVX2P*+E_fS%rsh_z%ptX*sP z8zc*%InN4cYuMtGDJrEZnG9xmbdB^F-5=e+`7#W>K(a-(&dC zukPWDEq30>nG>(btrIRT3ltlfpjs(*OGaiXs^Omr+2kKbmU5sMmopyklkaICD4u)| zXB=e(iU~hEUMU=l$Kk?kQyU(~gXo#MI#58c{X(8- zeH1orFs_vI1mm$zwhvCjZ3}KY13;Gd9Juf2Z>?4JoOm2IB)71Qz$Q*}*#5v=4ibNK zQ24Yvpr4CfKibr*cuD^K+=58lb57ByZYBmvo4S&}%co6^^t48GtGrG9>-%A?KF%{r zqXWw$@K}zp13%S#8Xqv02B8tk-5|LYrlZ?w;DTXcLw6+BJ1i3P2F9QkV+nL` z>YZIWFHJd)A!4@RaNzhq@m6oYzxLR(E+D?jy8zk*F7--N_V|`e%?PiNE@CbS*C01F zL6pSJ15HL9eM|;r<;K&RpDtbQ*i^gbxOR zp4n-bflyky4oIg7925e^0?hvCB(ApJy?R;;9K?$3G$pprEu37UDv0doqVH1dQ$20a zt&Y_5AfB zJSisR;GFt-y^#gsi$hRrSX%(FT|fPJ5bo5|k9KE6Nk+Z^+aBhqZqDAtxyq@Zx9p|9 z$t5!F5?&Rjo15up`i?=jzi9EB2i@?$m4b7@Um6Jx4uR=#$U0&hkj?z%^l#>vf%k8p z&MhRx?11OOah7uyk=rVYfnEs*D9vZmy6|jzA#>XYbR%Vbq>Be4v(RKLN3~0KB>0n% zPeSq|_N!3hHAb*Z@$ofroqErk*C5ziuil_&k*0W-d0ZE5<)nCSnw+azCO8MG)O!9x zL5=M1cC+dAZX*{bv5(eh^`{F+%qR+sMvI1Qc(EJ=jW$DWk}KN^^%Ud?{9O zN1ROq<$zc0qlf6&Io0AhHzM9ke5DF3`Lao{5hJ+6V6LtK`~M}ra#+$SYo)i-SEnqG zs++rzljDB}Qhj><0q;}HfZGO1Us!rzC&>+H=Xa6M`Mn&>6vKVuCY1~N{MUitPv7val;CxVKfJnNi`@|bXT2si-zGcXpZ5A2y>FWF z^_SfJH_34q<~y`mAf}OG>M77o0>o?qH|Cb2be>LN>J0`;Zdj3#I{Aa$R50~vZ)ilyYffsXFY9|t-6od2+0M>E#6V@-LgJ$gHY!&;D*f& z4hB~ERR^0&z@V=$v}Z~8*3B#PsFNpy5Nf5Q6Rvc!2Q+%1*fZRtKZg}isG|jIIwqTK zP}8i24@uG_vdiLC&!QMmGD@Q~{a*3RD(NAAZ7eUAS023ynm_*HH^2W6jhJTK2t|!R+H~CjZSmHC&#fy% z*sQi(_b&s|-ta}i28v0h$XZIHqxT0U%ku(S#c>`@@=o_;`K|eTWSum23|}Njou>Ed za2`Q#q_>}qy&36xXYD)j=kG=+%@#y`>SqT@COa?Lg`rVl0gYmcDWXU|rRkDxf+=ZT zXr2hWYp)Bfod%w)vIqiyyQlSgYYQVXp=w<>13TMq;Fd*bvsC4wwbJzHKF_V3Ocm_F zYSofROo5^#WEIq{VzBW8$1A|G!Hi9B77sWWJp7;hO6z0BOx-W`en2)pH<;OHftg&2 z0sFFz(qyTyMSHURuB=N3K|De++k6Y$Z!{0G4WlF8GEcBFcH)y7xRF(5P zW!K>UF(HQ)m7tA~EYF1^c~k$gJM?j~GbkI@&6nmU%^eJARGZ4Zix#xVFPiJUS9>J+ zH>y+pcg#CG3$HRc*$Y*pcH71brzhH@(OaZq58r6hXrt6qPNDIzD5-t=MOg*Qefl#oBE_i>5o z7gh)F^SKg=j6OTp!9%Yf`=6s9dG_DM<_cB|qY2+iKk>2DjFXdZzIBzXbzz+Bw7|($ zib%CEMH~GF-0r{9BPmg*2d;U3+&tUyAoo#ykvx(1fbvdD!q=r*cj^F&Aw%?~Py(9vrmvCliU zqyZJda=MtaVlK8sOb{6)hKO-o9kIpI=#atws>Q$W3imdfkRMcSX&~#^c`h!zGAgz( zINK-&7{yyC%`TrS9($&)nf1w3BN@WiASn;s>jAQ~3A{9AoQKxLW@r{y(whXyp#Np+ zLzkw+vq$Y|*f}t4QS!+Du=D<-J#Clr4?62!9nXmkxn_ZqM=eLL)7QjXz#mNwuYp?o z?#MVuTO&VsX?V9BlLWcqlORu94y&a-zU9HUrMelbyiW_U^Ky-}fz-%OklpUf1QpOY z1nZivim!l@c8#n=xG@+fJiOABiNGt6VERyz?SGYnl2P6AA|VQ$<_U7bOqF?;kdR}^5JKu%vi;dX_o!mtNlm!3@YwR?WA%f}WsQV() z#_i^7F--(f756Us5ZyZ2*#^ZtF6=QdiUA$PCAK}8b3c65<5}0x&jLnVoQH|<{SZQBqSA^9+IUPRj;-!;1{g&uRJlj|hmZUb_n$_+q7~zgjysOW7BW!hl&$}fzAoXXk4SR)BTD^s&bPgR-sMv>W8LSXQyNc zP}&sdp1q25ij45(CTUXxjTr;HahmeD44S&A_6LF(KL&*>B{i@Ef5AFnt(#J2_{lgc z`Bw+6iM&sPY5-hXDQS*A1Ciwf4&vO}nArR#rj^ROO8tB2>fk8I)LxNm;)omjzx`qRU#Us)HtxNJ?OjS?0E7#_4A$bG5fm>ify zQVH~-NUMU(QUCt~2b*Qjgroca`W=9VN(Qaj0WbHKW+C4SS;{%*vYNy>Blr}^ndqJx8$tC%mig-F z>jB+P4e+xXd28R(i2z_{)~ZkF*Myoq)8)xsJ4vAnd#0x>Jkv6Y0dcZ@l%`pp5xEZd z+eUIpMtHJ3&41U7d}uMJEf3ntB>C&;Z2E&X?t!2}k9FSdz}dSoXA?AuO7hQ!ZR0tp z6ohT(OCqu3d?F9^?xZJV@!>gvx23IeV@Gaq*XFLGd1g$^*!cb9WT^|=lq`!;PNA4}6iK8s zy|WCy$vm}|`Y;dFzPB^{mORYRjh^0>IpkH*Kop*C89N7szU20LT1^`|1mKbnCd zUm5sqa?XW;0fDw*um3fQxlECZlm>eA@Z*_2f|N+CO2)aNNfIOIlBW5`hgSzR3)|#c z{0Ao@EZBJ94W-e^atw-XcE2C?z#Dte;gWdhYLKVv_v)6%Gbjh#BtcIWI=@6{`v~qT zml`B^&7d2yDcgbE6d#UFl#z*^2TX=CUm)4t_1NT~{XmyM*@m|$wX6L|*uJl#j~MA$GpHvlcq)5h;9 zk8xCo?ZU*U?mV`OSYcwKXU&^^A!e&{Ei7F)zgk%7Fv*iTF+kURkHk z^gBlDp(1~`_+db$EOrJSuUDJy)+)aoe8iywcy?jOUh^?FxL91x3Om&My+xO;)$X2? z;ED}xJQPxevPR_nBgrXGFzBXxX^Hgd6NtrAcE|@iA>@$nVsn#^k1U(@tWzfUagImn zp_~x&H@zUL(Z!47WC${$hV&>o=93|);$+S5m1F&Kn)0Hs%y)oyJnB4%FptADWaqG3 z*$y@XIJ$G$S&o-iU%f9j&pNWdf31$J9;t1J3zPN>EoP#0irGw&jnFoPn>Y2UPqMsN zlqqVJl~Qrc8bKz^I&~6*BrdpElqjfEmwF}6E2VDAcF#OIwO&0Akbwam1!_hI1D03W zM$IdJ*Mr7Fn{;99*q}`wd>uWN8$O4*r-+`PO{Yv*$T_J(KV;ubgCvQxd6o+sK>N_( z+Xu&!!|Rz{oJ!f^koC^4aULCh&;qeU1N+k(SKOQan%Ry#;AZ}ml)JF^3QOc+Yv!{Q z0}VK*YMFO_IBGVHV~m6y2FAfnHG{^sEAr*~ zNv}QY9LHTgTYNg{Q)CP0j`)~SIPMF@b-$D;HT?ZiNV#j3S4Ck#xszRn;V&pVAPm3K zx!bmZFp^K&rTImeb$bApP21QI{BL_fyNQGKe^{7>M3LCw!GI2sbvH;lP=8sKtLmDg zz2%MJRn+UoY+nMw{xj%j!h22KLh{Uiei z0Vk?H&v?)~EuDOX*DucnT@(z2%M7s68Gl?x@+T1xvl$ljIzTZnLF_@rP!tHn<)rp9)LO0dI-=6iNXh@G&_f5J zoe*{?gcd?y%67;Td9_kKKTY|NXDrJ~r%yu-*8vr9KumlR6St&~+?FQt3L`#3u|Z^* zU@cXXq*ac!fc*}cHJma?b_~MPPGL1ju1hI=;oQMv9}bzPd&ee`+mpyDiw+J06my>< zeUzqFiX=0=$0}aE;wY5p8ADWc$`1Nr*xfJ?lSTRw%5LwT+6`|z?ktIHl@Ge1b-+)_ z@-JRa5Zwx?;jiMx2BLfp7Wx_Cm90tzQ3906G44VXT@|l`9uR|WN&dN<3?4c$U){Gl z2rsB{$HuB4-*t$qqf1D;3JVNPq$SfL$aD)lwJBgxV7}lgN1N%V=N|^>=L>fE47y=M zB$PV6CPvyMdOq0Krjmb*Hg1D3T-DTHIt75EG#JS!YtB3w2iYK)ljR3;HW^Byl63>3?9D?HXn0CzBro0W148aeA7&*m~vtxZH~C z`d_v*kg;b8h!GIv*`hmyv~6&fT74D;B9$7w?EMkri5mOhVs=m-8&v1*unm`=?*A2S zt+fIoj>8s4Nffh&A_M-O!Y?lXxlh_d+-@$$X^u#a1fz;ARvJ2Mwz3=d!&gC; z+auz3tUH5WiEIiq1LoLkYfDL%3rFy(Eg-O)Vu1HnL}|VRYDwpmb5dw~=U)oVI$|d+ zopeT73&gfMXbBuUseAr8#qy|4-)&Pb6l8gxr?+#KPQtdpjcPpqi^wGZ+kv%k-bJ5u z-+F1A=h8`O$}GAesEE$-ces6Y01%9@80)}WU(d<_QoZN*Hxe^!cDZGoAZH9jtzPqb0BgK2^Uj!=>E%2T>w_jb&)y9!^-n-^5e*xDo+I>M?^m2KsQhUtEZ^5)pHe5S& zm4UuY%0mwVM55Z{#mTtDF@19~q%hHf z6@mRqQ?@Hop<6kmnv%H&Xi8f(ttbL-l}XmZul7VN+KZc^*~O7RhcGhIPgbU9!hbeL zRa-OV*q9#IO*6pycz8r;6UD5jND`&_T(%_oy-vCLM;%mb+#_x3di627i|GWxNjP>) zl&0MJB4h|hdcIEB#~Mmb_#H;N|@*8#VW??q`* zWuuQIaSsM>3WU$i(4jY-j)~Uh$hC*bUd18qd7(Cu*Db!v8IWNMhd4nT1&zr#h(he& z8|M*6#fSHM9+{Fa0HtK?z-!W4E)Q*;6(?9RyF{c1J>!Gkccu9M4_L_(jun1ZXne{? z`+>Qm=G{}jG&?}QZQ8CNEtAMX3j#|w#dK1ngVG?8Bwu_W2p>82O5GILOB?wn?ewO= zOxQ&JYrj{WvQAm!Sp_pY227EIg2L2Bo=O8gK4@)F9UQB;D$Z85${%>+QIBu8TrbQEFi1AexB;COPG$1k{$=qCa>aOH(Cvmam!rMG*$1x5N5=m> zAFLJE&EbFHd-Wab+>*;K&}}5WS|O>Zn~^N90A>}`8G=)+1NkN$y$7^4oHmQuda>;_ zFSf3C)_zvA+k`*sa85y-eCiJaZdXCoVNGNyRWcLq0(eL{qKN! z8744d1ocuKH0_l0S1B6R*TvJc(O4+602B!-(*lW5Tb9tO)v&{FkZMwE-T4z}&AY69#vXN3S zaoO7=wgiE=zp};00Ia$6kThkbY>BabFM4FTOv+TIwm5plb*tu18O}sZM6YA=UMg4TXxK*Aizc~|%E@e3mD%!UfTaQgiws$xLdvhm# z^MJKXos9;%Ep%#hZzOh{(>BZ7!ghl11&XzSt)M@v=ch*Zkbf@$ z7)OAP2Io!zw$)8JUcI}mOG8}NeYH{S(c@bJuIVJbLy=RJ>Hk*KM*(kxFfJpeFfyk5;zyg z$=9wfXjJ1?ukeBRy3f%04L|-UhDu<&h_(k+wtWpYN7%uU6{@I1b-#VXT3X76Cexn> z+;)ZJiqZ8mw!FktLalUHNQp#??HTHnSXp{K=qYRgc7M#1&t^LWPrjSujZegExQVfJ(@+i#}VHpqYP|w#kMR(6%EngU2=8J?G z;DK{aa2sT8o^76u>MS~ev(2lC=_H#a5NL&PX&Dc6B4E49yH&39?TXsw1-0DMI{9V1 z-bkmcy4fOsBy;09ynl74+3)szXkEX^Cim^SF9WgBHv+3o`n|4)R{~5x!w?JE#=lXG zaRBVY#8+cAUR%?rY>%_XgvA=)qcvb*fq*e zxrZ|jvzqfL|Ce^&vu>~IvY4>Z_Z35|AmSm2<$+Adx=@%WpaEVwy=2O=@GM|cWQx9& zY*c}m1Fh}ySqZJd#|JHT;lpw|a0VP6yC*ArOqiSh^Iu!Lr7nvP8~$>J&$g+bOf6Ak zS=mL>=ZQUqKAUxFKJIB$Z&jI&jt?qK!@&w0kBy-7cCf+*^-&J(JA43f(G4C+zj1E+alUziy=zFA5-AAfP z6|ahed(-L7%4|@`&V(c501VR*vxSHAfnol=*epD$PZV1}wifAPa|0J93fRyZ!&>I5 zY2drI@!|wndv-k>IYrvj%0|h@9~Jz=IHh3BaA1X^@q!S1kG$#{PiZzA>Qg^ENHSg6 zY*bj7jbe%^qDVfa!Hhfv#RHB`?cwhUhzZFPW1#AbhzFi&%1)o-bQ(8}Tg^v-z&P*h zaOf2TTgG18@!B5WMn*T|qEn=KSfSwYImfie3Fu%o)lh?< zyd!wlr<&v9UQ#Z^H7?T42Hgz6$-hG{6kr7hR0CjLAbKn4x{jZrU=0r^Kw+d?vVzCN z%U4&>N!-55;&{qq=Z< zdgR`yQqQ>HR#_FZ5|W-fxP_#Zpd(PKTJzH4kb(t8bb3%9>5x0EQxvC!ePTQz1E@od>P{a6#Ox0%VkfPX9|&4HsZo7RoCP76 zPeqmVR>_?gis#*dPuD?a{hs)=G%sLNU>3dEA3f&H@a`-k+1z_1N7d#z<~PFe?}B62 zOuvm&NW#n@%Jk1lAw|zv_@4WGFLA%=W%wzOwMTzhb_nZO^>+zs2au}FyIUyM#{oLDu z*wY&g=W)6`^yt(E$uaseXLmr3I7TpLfS8Pz<6&a2QGD`wwoT@(4|dd9Q=(nfOSeUG zHqfYUP!hS$8X)c!v}YG6Sqe$u5|% zx}#%Gwha?%g0^!k)mknbv$s*Lh5ZiEax_UmT^}nxF&x(+MX{bnknd`eq>Hk}+H`sc z=zQgg&?h(c&%^-kF3wdB7HDa6<=J$tH}q7MFAHB5ZftRLT()rX&LBN>*jyf32yI8= zJhXbURnjjAF|<+jbp2(!(^kGuTe zdDiR4j|A)av6C7|C*2XYbQ0D@EEHrY^SG_@>%eF3BL#xh{;l#wpum*wzlikG*^cr) zpE4}05aO8S$y0tK(_vN!p*rJt?s?WV(_<%9`0bN#r}ZHBwsaCwnEO2&IO+b|ypGdN zp(T;nu~5(7LvGH*!ryGra_RXND?O`>Z`y@NK38G9t9j7+597?@V z0<6sOLEBNlX2fa4E-bJ(i5+Z~WPkSqKl6;Ub1b?IhCa4%ev8*;!i!jqB zi164F&T8mKvV2Mj6hjwsJ0d=E^wi}j6gfIoWE(}f6|vEi&4_85)$k!na$z53mjybq zC?=C4X_V$5<^(P|!s# z^F8YhVI1Vr40WrjRON)H)A#AcfuDqc{i>DXfhNgWf(qU&SrKCc;fNER&7Ev;`nC6) z=m}<6skZI=BYEiNve|Q>(GHLI$5YG-iY&vq2^T>hAl2&N7%;I z&fCq(%uM{J)75U)Y3S#hp`hoV_-4oVSG~1&K{f>KyFqYv-%JB&+%A3nx49gw4$Tuh z3>!~4*$Jd^TsfJ*dBo6@Y^|u}vLSpM>RB1#7)OH$TA6Q>e@P?;&Y&;DxYMlgf!aIF z&a(_7-Lj}Wna!b0TO(zoGTu%*2xnj3bY4< zJLonph*)0;HAq0A$<(c3uVRTb%?qP%pt#P`7Z~}u+uCa{PDXYIY~6M2``O{-CvQyh zc+c!-EU1VMC(X}|`#NahmE572+Z4G)X}*ZO2~4bHd5qu=Js|tUUC&37R+_Sw4{dX! z0d3$y;=;M!JrktA3=!JPISlE`z!AZcRyJX_@#=IsRi z?g81Fs2dUI=o=BO@}9^Xp!I(-<(6EV1(CkfN@Kygyo!Sq@Ig21E&^iglFy|v(P>Kj zHMHmG?Wzu#*7nFc=`xR2xh(3Q z4BC|YlzUmthGRqM$haNdUQ1M~=KsGzPc!C9zwzz!WTgwQtaB|;lu9uhD0r?K?9PBn zEXA^e^UzKK$$%8W>P83Df`rghOlov>(8HNk^E>GAcsta3O!TMcCmr=vEDC)epnQZ?Z8C z6$`99X=ojO%r;h3|Gnd%$I?63g~PQrdgo@V*1d4W0~LcyBB8_`otA3>XFYLkt)mZm z6u;!uRnp@_%K|fwPWC+h%M9D|Li2Za%{I?3AC{@BNP!E_FDEVLmxB}o6V6^rlfn5i zEFN+`h2pZ%0k^$TmzedE0k=YNp?52Pz-^y=6>q>T$8Qy{A*i1oa4QH{`^tb@J{j=q z_skTPd31Ul=d}3rL$uA{+eZ$IZ_aFwNK>WLTcKBCOo+j^m7HR_<>mZD-Z@a7c9>tn z1`s39IhK2ejp>*+Z5xph@<7z3-Fd)<|yc5+{d42Lnl)hmi~WRfAkT zouxvGtkD%opl~5p5xNKvv588Au z|Hh2G*9wnJA)mRh@rtv+Qa{Dqqeu^>StMBEd6N%Ah(X;U&VDiHD8%GdOFgggp~K4L zwkg_MFCLwm2VuGL&081Ukha zg%tY(xGbDlnY{@|dkY^hX7G95@4ZCUvom|HTO3UGSj@9Jih(MGOiGjLe_GiQwJ9<~ zTrMmZ7W-fsY)ZfeiAkLdOJnbV2H7c4%7&6!&?dvu=yp~7~&bceYs)J@> z^?3Q>?1%ixL*=NWtV{`ISUc&Q_44fTc(DunNj8>j6>~R-qi@~snf!vW%n=jZm~TdZ zBptY$V|Srh4Hz4=u)6UWPqvMg>3`S$@g?&l^PT*6e?>O4n`B%VGy5$jnF5M|WOojw z!8YZm!CNrUrSpso++wjIc~=*D7%2actwN1 zDHV14Ks{7+dFXk%dJNS}4#9^_;EX;qVcXD*6jlD#z^+(p+ONw-4Q`9pcY)0(-23|^ACs~%((4?wM!SgcR=m@QvY=j zFM;;!bxJHc>5lA@uJg_lVGO90Uh1zMivOG7L|YEI$b|;b8C<@<59MgSLrfO5BCh|^^PyGkojs%;2|1>537Wn1()di8Q^5xLE zHGj%L@IH~tuGNEoll>Fx$|sk#O*U$onj{ck2Djy0aKBuO9hw?yJmZC zv+D>86UfT0P2@yX1Y0wATozU~SUM=t3-0le$ekP=gyFPpbJLX9qxOg~T2~!ZJ09vo z!^VdNE{0v@{7qK4n85#n|MQ9F`EJRwkRxQ!jnxfYn7OmX!VO!DoYJhATM_(; zc$q>6E39TY7I=>{y~_*%d+_6Yv^@R|_y5}Q`(QI7c6Ag(e3P9Ay{O9A}z^LyIZYAi7Z1ei^jpc9P=WM|?ucCi2 zd_93vr*477YSDrgdFO(=vU>GbJy{cfl)(Ap&<)wlIFFBs==Dl)-69Pxb5S}@qEfdXE)qWxtRkGvzjNUpqrdBYYn2uL%}X=JfNxSFu$5^!x--NmK*$dA=jw_+ptyhM7ZOx=3C}B zKG0!@8l$;zAU15@$~J0pe|o<<$ZS4lr2l;l$$4%jp2HT{*-J4{o4FI0-u=5Oogq*Mdr;Lyx}>MT?5~OX z08ed|XMO+BH?+7P7qPcOQh{z#0TRmvoU>Chx#hwZ$7cl{^{hwnIG=cl?bmmNK&$I6 z4!-tJKPq9)ylz);(AVO@R6WhbC7;q(p?Lxf4PzvC zO;n6W9ED{LNbu_-t-#=0>2*Yv!@Wr`mW<58LAUK*=P}{61w5#dh^nc561^9aGByO2 z2R#wfb_39s_cDh4PeQCHcj`y(W|$p&x$riYs4Uc=7~T|Xx%_kaZrNc) z0;g&!6m>(Bt8VVWInJG%?7`I|qhYx*D!kZxkz&TjI%@?)8$yHSP#S@3oqjrNewC^Q z3JOi(1}AGBkMx(=|1?Ls+Ijm}y``?ZVC3IT7rtm-qXhh>>^-u9-5SM(6Fhq@)+pO4 z27KAAl%~O)-i@K|LLTtp%m&gdzcl}x;&{})pzFfryc*~vzMOZ*v()QqM4`B!&J=+h zyj(Ad;gv!MJsr24ZgEua$dS=$KjJNo2Ywp)j*i(ou;jtNKkMt3TZ;3-iWzx=D`H3t z03Sc3P5wbEu+;LyW9A@TvXO~nMJ5~TlVvs5WBoD~*R!&26Z%%Vgv5=eiO&VyqzYl|^7N*rPj@6+|Y^I9KuaXPu%yB;I0_WvJs|unWtOu?!x{#+v1M zf&>tIaXum*jTbv;Ji5&38?2x);g!i#4p~$5A7_%dt_wwM67f5Piv-0{>!)`|HiM9E znzCJL;4k4AgWmiNlI*|%H;lR`L~nwuP4%?h0XwIo1?u+%F+->cs+4^ero|vCCZq@5 zK8;ub{McSdV?(uBp9Gqx^LwIurd0vE!1Q^;w57<3z(9J+&k0zHT;@bx0uir9Ef5AL`-FW}X z$7K0PNkSJ6uIVhMw2c%4A)$4Y2HP(}2@z5r27@~2;syPlsH1EWOdWI^`tSb?)&$$w zxbj~;U7==dEKly*NeZ8v_0cH{)Ra*S6uR!CG^@C=fqH%-S;=7gh%J8S6sWUO9@_6| z0Mpc{Zt=+%B>AU{uFTl&{;9V~1@B<+!8sVGUG2S+X;jxi`3Iy2rQI{iJU*RKBhilKj-l**27f#h`@s}#+q8JhS>~`2Pha<%%goYn@I-G3Vpb=6< z6ZQ;dHb3i&@!|zH$WCu)r5BON%kWd zD!VGL7Ncvsl57sDQ?LC-<(m!PkAHploAIySUhvsF$9{SJmz8hkyoM=*HIeseJWJ0< z!5`F?S`4C%*NrMu=jk<3`ml!qjzJ#_R1G&6Pug|_sGi{SOq%e?KTb2xS-n4%E+t!C zc+RS{n6q|K%npha06~toE;MdQgrV z9mfXeS&feCTKk77s~=c*xp__~FE)DOtdlPk7{CT;bLH7k=DdnqB)lEgLH`BhpK%*H zbV;$du(sCQ07{*XnS+MFZ0qMJKyK?+4&Be4Nc69oXQDYDeX1vAlSq?=Q8_~~pz?l< z(zKGRf?|;#YPSZywfq;GC*Pdi9R^EQ*>2hPz(zHe*0iasc>5%%+H+2ERifvgcV7gu z?=>K5G2ph%t9t%IZ(s;;JADS^Wghzi@Mn5pyR39xniB8Us4kj((EXg^B1jV)79X3{ zAsiH+a363R;8po!d!sUsqRE%$H_ID2+r&pBMmuby7>Fl13XdNfB+X3IOs%RiF%gJt9rns78IE>#%yqb3tkowot5WU&wLJdD`H( z-WS8ieAECNW($Yk&r#bK=GV5knWw6PFMstH62mS~?851VG>aKGkz(R0vVziJ(S|n3 z-xRy)BWJuSe4vlt|A6?0ose+UNZWm%A8gEPvo3&kS+cH;lIA^Py+@(b* zb?PM)F4xjwIaW#c(>1VCZd9lG7r%K0_*JX@cMBJX^uf#n@)^lz&XQ7}p(tw#PPvYoX3p6j9YQ5(0P z``~5mX^?-zWk~jn3nWX0jL#ufl;b(s4jUS~9=QD{u^JlJwbDn{R=jOp0qwHJ#)cqR z30)Sdqw65jvq(?^ZGEx&FB>WpO`3uZHWg1rh~0pF>YW^W0z1HV{5B?WuGyBP%l|cx z>~Y~>?xz+;?g+&|1hInB98jgH(4wIvV1B?CFE0u-_{L9$ATEjxw#rS-Z!%SbbJA1? z0+59AD9Na&Ps(-kI_M?P5*?ez_JuWv!Ac#3=Ppc|$#F@KoiRYT_q>Y%ap z4QVc?I_S^++&Ji0jsJ}adUoT;)>QO_{bNBEI~@J;z}q*zZpP7qQhqzBdT#vNYZj=h zr5I@6bP9`E_i>RggS5XoWj0+wI^E6k-6(K*T#Rf`)Kt~t;t2&$GdKktY?J<(ED2;! z@mvFB7#ab&h`2$wY*iU(V>U@v`*-;O39C`r?^QtzvK1iodsmzwYJr9+Ikd5BP9w;+ z;`Q1!kqP8RM4o7opqYFc43Mu2>zbDQ0y{6;nFzE7CmTfm_s<^`|L-c0-tKi7v@R>= zYO90F>A$U(Yqy2p6?X{F2AvNAx}Nu#>7HeS&n8+5|}SHD++Xag{}`#tXiX9!w{JrK7Gjs_)&klB4) ze1vy{gT@*qGmxK@4t?hzDqOh$%2j-Cux_~nMx0Pv;ZIPU}FU`Z^eYLjAqyB8qd(3JSU3UfbAiehF z1T(x;+xGpDJY*-G>$-m#Nws)O;wfeYMV3*T{g7ZR^EJ}e9a728<8|SRg{u~LsG%6(T%Dve7|c6G4yg2eBwN7*=6hPuE-3Rz zRYHFgzaq$d?vZuUz@%yC>!#|FQd7c75OvY`Ax)X#Uc}Mr`Dg{pJTm>R2utT>@U+SD z9{zdnED&ctC|F6ft(@xMHbI-99Ts$2tRg^lALB;btUg%i>%Df#wOOQ!(;jw%qniq$ zDOEX*-^$A4eLAE@Z4e*7kOe63{$4uxta~;)xL1VjbUafMxmSTpW~h)mPM5ci?ytW4fTvHkj#Mf18QpN=@y@!IV-)!`H`ERjy(_F z0nd^3E!aJH;;JFiyWe`}H`ZbekF$7P*yXUHEYZ#Ff*@O>pn}vXwS}aW9F1s{9RVKS z_=DIU1cS9_jli*9U>m`2z3cbBbwZI1Ey0E1Vk7;C9ztnE+YD_VIiSQ8SCeF+`n0b$ z&3~if3t=y`04no^M$`3^JKf_1JH2~+nSoYG1xwsn7 z6I}4tmWpr5`#t+Ti)A?~2bLMw38}|@2pfCG3Z!n6VvbSdFr}%0yz(Z=2|=SeSFm0b6Ot#^^HIG&O}X0tOT||I8tFlP6LaXLt^V2c z5zgX}g2=X+O;C-N?4gsCbNW4FgO8D=Q|d!&L+h07>g!S0#95%^(Wq_;O&8ybs^)je zkYb$r57*yYFWSMyd)F2JPi~m9bJ%hJUkRR|s zp2&lUevP!#9b@)rF0?zN)_7xr2sZNtd*|()hewPlrdnx5@RzbY(SB72RA`SePc+iF zISMf&-FqziSRrP@-hJnPGz_2t083y4#E7x5 zzDbhHK=VTMTcDGt2<00NxUYZNRCS-Dx(O({2~p^5h=)Lz5lHQ|(*{Y;oMZGdK|W9_ zG1LtGf;UKRg<)?yRzZ)akDC3cv7A{(`1JHIV{ry6)J!<@(eaA!m{IfEq3(YsC)vg2 zT-Xb~VS%wr6w^SFI!bd+wbaunsjP!0v8n!D;MQ*AUJG3=1#^U4h4Rn?L1)E{oH)TU z?+e~1=4zYd<)M2$(jbYfO%Zj2=J)lWWNx=So`F;a*XS`H2`-@bNgjac$BBT8LM<}K zVg#FmN){k5410U#iO|_PM@l5x^3Y~(h6qndiB9snE!EAagF=dip!42?ZaP(qj}8jy zHo_OADe+$R&4Lbu;t1%pvC9rfPVC_Qh-U`ZjE&zvPL{F*mkS4dvn;@sLNV(ok_eTO z&`lMLm{|KD`T?*VpwVzLsP6X6`9io-eoN8FwQoG_c(nGN!Uh=j-O({;ofnAiTLZ&Y zgGa>1P@7Vk$#N`##R8H$@SdN6$h`qnhn5L$$sIi1Jng}=80x3p%Av>EO-I=sM><7j zi;>AsX(kCHsRy}mMF^~jhP4<;qZr5*ZlE+96fH2nta~8=NKM!+??_}9X%P2`^m3Ns!QAwv6{OJipjDC18#bc7EX)+_?7b-==IYpWveE) zMC1!{xeFt=g*U2`xP2spcR{i-7&F&H=MK2lE80AlN8@z^ZfT@dR^f-+x5LVXV_ve4 z1#@4a@A0j%!tS#<@7}YXkzCf`*qD7Tb9M(nU>}yk#K3QtX;F)#bxt;9o0}xqUS!ZM zCw#Z?tm*-D3vZIF^cp8Gbu>dh*5S10dRAs;;!n4Ey=7fi@3H`~QSTooNT;{>Y~q+C z>f33o9noS1lQ}_NHT5A>KdtoA7IT`I7N4pqmxWoXME544_LjHqr6p1=UXUl~sYOCZ@XL51S8`YV9ss1URh2qOIE8tVIykp8$(L-K;^eSfP9MCxv zw|Op|vookseGO6%55zUnWArv|$COX}w)iy5Q=)spug<4e@rKT02c<`2H{ymvfi~hd zjBe-E)b)$3IlV4xifoXLtGrubJW~aIk&8i4q=A&Ew{h=9CCgWF2V}V%EL>XnVn57K zsBwql;z?Y*q611p@R}IGV(72Y3I$5{qKpMi$#R1%$9?xyhy<5`=3)jEpTKeW252_z ztD}1f5H|D02ZA7({n8h};2d--=B${DN+o(4d)floT+iPvE}fSRl89+aw0ZrWcp3(y zH&2eAY1cB}p@6Z=K*33NV2GV%>ic(nx^@3gmxWiRg{^ti>l3-`M$Xrxp2i&!J3Y2} z?URg?(=%*T>_)|~YsO=f9V*uSZJoQdL?Rm=jSB~PZHOnHr!#nZP`tY<24kW!#R0XL zeku<|`;boG2m4Uw(JH@BZ29K3oRn;s zyVc*=MjNfuMPa7r{%K_%C|(8eA*i5m!2D%5-t4t7#`DIG$O}9EK&&OhUDmGHkO#+6 z1v(XPe;`z`l=I`6y^2!kyk}G)!}+C7d0vRqh69Gt&IitNwC%j>*iU0;U6wSzyvdrn z$;P_4aDLJTsdJwMdjX{RKU93?i4+SlAGZzNw&pl%3XBaE zgr(sb%nHXB0>5Kc~D*qiMeihx4a}WPhgk(=CD7^z6WYDzYhCSJ8s18opf2A zcUaqBIg`yx=cZW=ACe??OJ^507P~Cm*({0y|2A!?75NHsBp9hP+I}xRf0;M(UQER# z9nx*$8u!>2iy&T`Vf?p^!OEjf`8Kh^zB)-e&|~WqHfXnSIc(j~LHoN3K9&ZXVX~{E z7&NP0m??F{V&L~t%x;Pl15*leL49Ef!6yVYpV1CVgR*#If_u zsIg!H3480<1EM_06CRMyD4(>%M@LqZmxdTTWi*LZ_LO+^q z#>aMBy;!j=W3euq&N4pK}hMfOsfJ)~Qn3zS?_seYaJ8qpzcD+en6ftk6@3%n@o zH&UnUpocn`U`QoK5Jw6k`?;4%vOF$$5op5WDi}3gu%z`6w??)&Fim+o3|f!C76jFL z=I9{4mgMgs#IOd$<7UNbk7FRh%DhbQ%V_((wNk1LiBYuRZJ@i5E-IOd4<4leGr6C6 z7K9&l-%js}x+<=dR*=s;&nUO58r3_vcBmbOUHA8g_7cOb&?CYWCaCkHb_eux zkH45rH>&TyQs{l0vmW~6ol_KguL8zhKQ~u=FDP9E#Yw=kbqaPLM;d!HKYkBq?6HmJ z!_3}OQuCB_^3AudlC|ty78hOY*umxZp`=GUi8|(T5E?rZbe); zQoq##3P}{Rh9U`+rbu?)yMdqRuifT#-@U~LYR)Qox*4srGa#u5-8G%~L`S%_{ewTk z#ZKI1+jdlj{NhEe*>;?Y^5c;%7j|b8Cl0p;izYII6nOyT0(k}~_b-pG1Xbm7Vb8Qv zWP9G$B(`|ljPvoo_Z(;bFyle*-X3Bb_Nf`3j`Gv&;TMt5K#_6R8GNL0gIZ7 zs2mYnON)R=QQ-giBtc0`Y#vD1==^40uln54=lgy>-_Q4Ryx;7DOKu8rA+3_{^wN6u zdZu|?gynIHbhFn!UX3JfKJc;ww_x0?OjePk5vZr|?iDNa*EX=$&dQ*XDhqp;G~h62 zyA^u$vxi+*M-_%w`D&uk_dV=?hE+qy@vv|%MT*Iv8qp}!W8srxdB-AA9wSrU>v@lZ zzt=83MIWWHmI0{{HPQNCN`df>Ms_SvrO>$7C~_59>>R}%@ye*(f%x~tE0ItBITVjI zSH}6rp9qVV8bH|cFKcp0g#&{ST8t)4NoOb-l)Rh(Jr!veR6Hff+NF6QHZk%8oSu}0 z;aVha@2Yr{|}T?2(I$jFc<1ikdSz(f|->-9QgD`ak1 zhMG^$2CTe?8B}I_3;vfk9?B`1{Vv(`oOPn) z8=0bAlq?ljfibBhm6R;d|F*X6LdMDw zpYmmvA>;K8*IlLz&K9VT@okQHq1REpcvo0+NF|UZpnmfOd7|KcXnzD+W-N%kvj{bM zPth&W{SrK;l3$Ki&mV+UIr-8!o1?gpjjR!Gn0Vh z#zICLR`pMXBhi7E>J~W?&&$y-Y87QV7qhU03M8tb&Jxz>nF3uAdN0ZI#}5A-&t2|4 z;yp{cLy_|v`zyzsu$;4lyPvL*>2JcTl>2BrppsvZcX~ptRgP!-{Ohb%QH4KvQ5Xc! zpIz@=LX$WG9sk}UU5R@r=PW=4*vE@)n7_If7qP3v!A{$3~jQ;cq8 zWN{ofZgS`63+8{v;N-lQcl2X&+JV9K1}LK8!4H$(`zqqSPytrhI9+sgx!kq9!p$$!(DVUESmEaX&!fknZi=x>fn-j z+r7aJif-c{jXVWQ{YF8l@NzV|N1J5hzXln1uagQ~ig&W`{1&p^dxSIUv?gjmG{UKL z(ci`BYhjN0NzJpmo0x!}5#}kE)DsoYd$G8_`)SPSmgh>|3A#s)-2)YL2Z)+58c-%L z+2XZUqQc_+_Bz1`6*Hac%K~HugX8*e@EX~{EF0mtrj$T0(}d8Envwx;L>3h@LD>ZB zfb{}nAU&dQsjCaD?-LRuHq&d2)Dh%!L`VydcjI3WLUkS7KK_>A@g7_YLp-(Nu`p_k4x1lhW}m&oR*Z^XPe}1J zpseN1vbRX813N(FM)1$0WVsZ{Mroxy2n6;*qdr81yQC!{`{h6;pU*}$4K)x|LCFq6 zTT5eew+{LG7L><#b%!Rx3Mn&UtG)Qfrw(HCAe`z^Mj)Q?@$H5zV!%qVQ|1K{&us2; z+`t3LF+$c3O16z6$yCe@x@9iLnL4;1vRl|gp&v_XXk=?#6Z|1Nm4h5+7Od7(2gbtv z3~*%O3HCW$|GS^xFzrWh*ajvGjS@Qqs4{`m7RD*BgtSXD1ot?g21p}!+4PPxXc?tY zg#7mRe^ma@FMjvtpZ-(0mXfWeNbJ)Smwg7*;!_z}42pG(|0}7%V(dr``iN|I;MoQ0 z$|v}*xs(iml1at1@gK5Q@rGSt4ml5LXlzTw-o@MOtpW8yl>)pVY?JPi;_39T_K-9Y zh@J!&5j@%n3f~ z9nX`z2c(NNZLoa;GwT5SaW9+T|r8&HUS@GYD{Z{wa zhTh75eekP!Uzv6cScd?sI5lhv4$BBIm)RVxIoIfCuoQzo6R#zE9N6Cg%DM?wp_q~# zq@Wl)CKFa0=g8o~@&#$U_2TxBQOV%KKEK=1D^P0^HAs>puCne1e+Wk7(vl`_d+0I$ zu04veJqLVTPz_xYuBF={__s>D5g2)q z+f13}fsT@1_FP#39kN<_o1kI|#zYRy*KwU;?NqC%5z@h^MBNMpcwn*s<-?ubD1xEUY2!agiAv>hLe?)IKukj9}&Z*+T@`zM%f`qW5iXjslgv zV|gVJr9xDs$GGYEC6#k+JOCyyZA#-|@+@2KG8+$vjj`_IbpK|7!N4qU&)Y=uXOYuJ z-|ajDrU%g3u)pWjm+T{@^|_5NU=!T+%bLOS}_z;4|z9nclluvn@UklGUz*E zWX;90>NWn!ybNItX?aEKRSJcWd;LfJ$De6K|6)qunK%vD_;Y3i9>)z${-N~j51;p5 zeJw~!qGEXseV4N}qzZ^Km8cleDuNm$RGh#bYAoJG>FbluZ&k5dU_I0yp+w2^@v`kP zCu}O%9tI>27;mYeuPwbP+zst-Rs7AqyCZYMY-l2F3u)Uq&5R&D^Op-}zht^>wy+Y_ zf?HYSi8L{9waB~VcimHhZ$sA3uB~9^CkiWHz4@E%c$C?zci5>Y^HSr>&-=1DYn~ch z)1lDzP+Z;Avbv=2EC1Rej-{Z9ah?q)HfzviC0t=@v*u~GyM_^5sm!XfGp0S%7TS_y zgOejhCE9Q-OYP?Nk&=kPg{h!dcYu>iUw-2O$>DAGRnIS9a1W}tH;eC1SKh?>aTxbF zZ1=98^(1EF;IL-nQ|EPu-3=b>4{CQ`C0iYMiCkplT4z%-;F;S^#b6Ex_-k2B^xdT& z@o$KfIMWTg^~kS;Kn_PIyN1;RC@EhZ)WN;wiERsbVJ0Ut%+6+oIee!F#*E-Q9C-lm!33ikBgnWN+cKEQnR-tiX>|&$0DI58k~YnrpBdiiGvm zWH+;l5eLrc95b>q2PheI6{vxGA)<;t>4Ta^H-&1J66}etln!HXnL&woK&8OiIy~^9 z=X!C68|I4jY~G_zI|K=`5=dg!vZ~m~>21R`jnU9rftPUuW9d($=ho(>_a=* zAn9UdyMa-8NFRw>16^QhdZSK{&t8T@KVsJ6#H{VGcF-qrou{Df+Vj4;t@TL_f_iI_ z@}v(AHPo~s7O)ZHXtrNrYDK4L zSwJ%@-FL|KqU%QK1|Pgr4P|pG`L(6xf}NtAfL{8r1Y5+hGYz{&khyQzRcq!8uo|Og zjgid`J_Yy>{x~*Z%Czw`qm%<s6v`6?I7uxpXV0KT>b)Yqs4BwC+ikU$dElA;(=`E-wG( z0aNqguod+dOa&7407cJ2EXoCn9&Am+EFT`mF8bZ9EzZiblIyG-$OJ-4FhzPdxJj~~ z{&+#zl8ZuYB-41ch1k=RwHCt`W(n&6If<*7VYssM_yM`WzllEgi{m8gh4F8yjpm>t zN>)gb{Z!08w~8gD{_nlBQ&66F2C8wUB0gmBrr;}5d&%rr%WO#t2`VDyYz-2eBF_f5^q z3leCtV2r9lkh^0YNsHPH;T5ETQ#MJq2Hf!f%u87+srM@Ktn*Xu=M9oJ;f4ju67D6s zUvdbVH{!hN+5M7oi4r#o$tQe!O|?^p$}`4$3a31yN>}~C8%$HF2wVHCzizWTy^x+HTLr!jfv$I z(OL9O;6SX{KT>zjjmjL8r73ypD18UQKj@?20Jczc+Z4d!3)E*cx$? z#PV80%3ZDo_ISt6(L^JqPIJgH*HU2(Jp?+i6+Y|W{HT*wvig;4f>EbibV>Af*HNcf zX=_ME0E(JpxCD=)-f#6ApD#NUIR0yyoD?aZDHV=zu;O@Bk{sDCy$!q&lhOvZ5{DCf zm&tE3L1aPS{1#I`__>O>SnzghxR?jXWEHSJ6L(3|y$?n1kJLmriwju8>_p&yI~8$X zSm{#uaN<>382-`#%_rf zBVZ*etay@bFR_e@m;d%}-lil?4jX*2K)ZDL)v{&S3DCi<4C&((Ksnxam`uJ{u;m|{ zBwNTypTj_QdlF(Sx+*8u1^x(Fi8eC&t=O-A@_O9Qs(!e^&OL}$qQczNVOtoOJk2sH z%y~24W@|RO8oZf`uYdaz*~rYBabU}mXJlE@DcMd6#w-RkL{g*}or9Eg9&3}!J(ssI zs<&4GAv_H6?N#6ugQ+BYhVG0;jX9Q#ZnxXcl2E^Q{SD~&!@>H^B%2vJ9C%S(YlMyx zN(Ndq2S6>4UcdaN@E`=)D#6`N6J3SAqqF~(xjc!g{z2(i%D&UNeEavRzSHuf{lW`G z&B8O+mS_HC^YWapBk}Xzr3tdikYVUtMD@m>1ua<<@`fQ!u&9(i!#T+%}jFIgN zyDPrK*7*qg@($lcc{elMU!{PGCydPsxB9 zC{ryk9WLRBc1J5A?Woy8Quy-w~4IV zuA1mip>gFhjYqHwCKc*q?Dj!e1A4R5is7Km(QbI)tcR}$1RKzD=X-aHNggw_IPP~q zP8p%4f|5Z-s08~zHo8I5CoSL}y^@zNNCJsgT^DMdUmNI96?p=oDe|a}y_6@v!p)(( zq$uEebWT4zL8eoD#cM8tJ2U2ly{ni^!s4V2FI6tx@X}e|6!tlQR{@3pw-uPCik+u$ zWg{@{xe@u(SH>4T^WTjp`>BV&ne^~rkTRWts_ zw9)YCc&r0Mz(VsOP8r+1cY38keeGSh{o&i@sTG}`Idr8!10^bJ#6zxH?mek`ey8U{ z)`065UYh6+Mg79t3u+X$noKN@7c(q8o>VM9&_-u0`{kQ|bvxX&V3LUmabV}g0zpGP zLGJW}ZY9z6FjrLZ?*^yalJJ7Th*^ae2G6tArGGs0ZjmV_p0*s%-^iDFX+@b~=x-nb z#@SHpZxx;%7e^g=%XY%~`>T#U@8tHt00r@}thsNMc_oBphP{OhI;GASde=FjD3OXP z4=K`N2%pv}_S3nF-LRGcDeqiheN#jJCr7bKM}=>%xSCbCcsy;G+jtslwm4!L7iKa! zGdc@$I_4M_HQ(L;tKX7c%oa5cY*~&PEourV8Kib{shF+OmAq}funNW+?7fPk0qQwx z#A8k!Q7F!VOW{^g?;_pOw^dXX2)s*cWQF1Fqt8R8o#b@2&#dEPpf!@p>E()FP6aphS*^6-Q$FK{Ygp&thF^% z48%(%R|B(LcggepF*&%|tAwwmPe4orw^F2=y?P}rV(chX$@AHRq)Af3M`BBiv&Qm{ zMUFakkk~m`16j)%=8igzr;A6OFw{Tl^x2&Ckt+Fy*%|azezyegZV``fU4j(k6lbVO z!fqUX&-im{NE9x*bj_3p6LOrAK@PBtirMaeLonw5QM6WaS2*T>iB5(OzERL1(n{Kd zhXU6w`p~nFu7-qQEY&GJ0S<{aToXMM+9i#jg$(=Wzw+r@`Lo8BkNsrbEL8Pf#pzm< z1sOn1^ronf__sW73JHy;K?ihk}WHd@+D8P{b|cc`PbinclYydR7Sim29Npc|C=JkqFMAf z9(eb0n&qXQn}Ixgr`LVg5fEJ1J-2Jon7<|(uSsFIgvIi@6{AkqmTD#XEj*S%uU%B> zvV}bG#$$ct0=FTkNvx8OaME1M**3$Jl~6HfKg=Gp6=*C&#Z1(Ns-Cy#`?O(B8vMTR zIkL)u=cG)dIcW=kc&#zo0-b6BHkxHX?e~^*^KW!cra{?Bo7Zs0$Ai@U1!P+0}6>5nx zH|zo=;_>iag-QWxeneN8oU$-OkOKuAkgQLUrYyv8I_zW`DAzu8U$wZOlS3Z?(NLRx zq>taGl>=k;%balV@$+o8!E9h0mY+6VPQ-izV%DeM`VW%8Y%T4;6)F3TfWMoPrBP%D z6@z5>C)iq1t7vz?=|JTfQsjv_qVCW;tSUYdNM^z!dcC`Pe!d_PXi2BZ7EJ=?wnFkV zpL6w}ep5S@;JyP#Yb^x#Q^K~p_R*QIHo;=(46jStBpLSU3GG{i@syM>kd2aFCDlRt zrXY-sXrgZkS8=k~LlK$tK9%%(uU&K~Fb5J(*feB&b7&Z>Nx)@-F|t< zY5y0K`T_?wClHJER=Huzo&en0Q9qMlO<3Cx@LR$u6U*B$1MBq{uodW|#9N=Oa;V zE^oE7$AbDLWfHrMqPcJF@i8lm!Q*SEbAyqKG-GM|>pyzlgR$DV4M7Llm&JQzDbl_0 z{onH695O8JhBov8AXa)9-mO5ZpozXt{+vhcr~cd|y6sXH*~G194@TDW{O8^s1kCC6Z;rR)wr={#+1pA&}DN>EI=V|q3j zfnw%`Uw%!=Hq0hY0Zyw(oC8NgKplNTk}rjlZK23!DyGY`2-HSvy)byz$WyT|%R0E% zg7W-Y{Wtp_u?y?OvqsWNT-eW*wD6Z>gbV7_C%ei`{ZWU_@LBLM^{P#4{0sb%9JVe< z6TP3e#=l%z1u5PR`tt~NAo`wFbTe=UYod{PKq$1U(sV##nlCt6jIs@4b7$lp%@9y4Hbf94&As^@xY17vwddn^3*DPHBxGvN`61G5K4|# z$+ojA-D*I(ZjjyueZj*%eKa21=BtUWXAiQP_!WYSMD26e?I0VkN1akN{h^!PYzSkc zF>&Bw+YV!|dnB_7C`GgOhYEuWdp72;TS>74yRc10E^Hko11oYI#ne$8T_wLl(!-7M#$`Vw%ZfAE%)Dd>GDi(*$N~Bo4-(Dvy zBV}%_(^;Ouw7gw1`(3ihflUjj!A)2l@1kU>2-lb#=mNb;`Z&lgP(@b=K4Ky78nlXs z4!c4RXaMq4AVN^TecCW#B@`YT1A7g<>2obZ!CWbc-{G!ZrV%v@6&rWmMLowivy8$6Xiv!K>^IQ?dbv^Sp*J{Fkn zTSFJ}huq5-s1(Qi8$?a8=GeBlRXPxX_jO5IrKoFNJ+Cd~5yJ z$?RQA%2`>04sQN8A;x^^Eh7St=}R zhxYLoZ;uT#C}1?Sj=Nrn=FV9Eh5>V#)bMx6MF;k5KQ;OoZcwsTifE{qR?!!dJXR(r zUoh;}O#?Sq-?6A0?Q>1`}S}$!bMIgPal5d^e=JJe*^Zyq4W*~1k#0dE_zwx~{f9q=iSi#u8 zl#xsa23VC5zz$Ke0~E=pVm=F1$#Ha>foxbS0_zFve_hZ;sH^Q&A_XyuisiWFxPgO& zB}uDzcwsw1ol87D=A^$BJIwaE={bz;k*l^5WGs*UB+oE|%xnKfeqt)c?Xb>~1#xX4 zl}!_!;dOB1XC3`n#kY?Aa2<%Do+F2aUBH_0jE{bjFZkKFSq+P4KgQ-~8Nq@&e&^Jy z0R~HQ>eVe3WG}NyYsa00Nu811DWzo48ed4oq{}xfxkK`n)+;iD-`?)BVaZz8|Bj!9 zN7v07Ki(338R}|dc^I~-6E<MJ4a0)$HAY(3IRCQL$uIfOc+*^f*^{0Nn677yYqB7CVBr z&hFuNhHB(U5NeNE#%O#dKkP=2vH1=o+cM*O+qc~@?fY@qDaS%X(m8Sq_33Q@LnmR4&|PV=DpUJbAMPc=ej>PE{eS?*!IJCGBR1R}I_lQa& z)Y$hSOPRgS9<7Rw3a}r){j(f_!GX!*Zf_@v%p4d8h77bsPRKc_C|Np1c2Y5Fdb?W_ z-Nr{jmNKtZykS=~F(u(Rn<0h9n3I}5>71i@NFM-j*u&VEwGzw$V)ijMKX}hG@t!HO z*VD;o$4z+<3;bQ!ucHgJ;93|r#+;Dktp(_*%f(NMhkP=HUGw62cK5JJb%~wOFpqh+1J-Kj` zJKoSs^8EANhuLew+Js7dDc=qOO2u4|55nm-V1QT?c2?3yuqIC%u0(|9^OEM<#_^dd z+{~ILYw+|X*RX%^X@-*lKm}j?_B|47PQ~4UVWKjENfISXpvVR)29rdmg$O(EVOK=r)3AS?ks z6*MGjk?Sqir2HKnQmg@)Sqoy_!AYZW-?4wIF>O#|5{GeMyJMl9?F5u@WAv_y7tdFo z@!7;tX3cAZ#dC@@hu6jG;nx7L+I8zGmHc2-t4!JJ+`)ao{(_xD$1h5Um`o0DyFI%h zr-qbC1K~?=+%nqwEjNSF`S&)mf+Rce3hJPd8Ofq#FbSn&Kn2afL%PrNI_qp?*^&?0 zd%>AF$UZ^m`hMd5NSMdU1wNH{mo*UE?Uyuq6?@-d zx^h4CVd$FRRKE^xtE|>12OQrIJ3&`=-Q_dlkMr17 z5!4NWrP}5@e&zYdJpWr#HOux^%SNcN1yGEh;qTjez;tGML3|hs^H?k|LvUPCL+hH_ zu6W<|Qs&O&_8^Uf$F1tTM6TQzZ zk#&V!e&fcnL=dmT?Lqz{Vb`mPtaryRvc1_b9m`3Y5!1)md_N`b&w-}-MTae7u$VE# zL=Z6ScS$i%kQs(q0*&lcP$DZyaL8rMsR>+-+%UKxWA+-cYPv!$lNoZGkFS{rGV*R` zl*q41pZA9P;ixRZsMEuUMnR>^I#z{mH{=e{4rthEiY&I4i%pFUK|8$;6XeJKRG1^! z0$T^!=lQ+x=P6pl#;em5{p=Br-Au(~;4;xWH8avSIm8GqlwZisGp6?B>2$OMPb?N} zOiI|f$UbfjJtXRsRxG*4JtqTLM4y*yM77J`+4L3tOmh)d)4KxSduO9=-)=~mFszdT z&4l~ywP_hN2YxtmGRWYQ6!#RZB5DVYs6ZTc!uq$Il9f`V2qP*f(q^&79p{-GPvtqF z?*@HsX_9j*#A-hxL#`P#9IoM>qVwH*Xsnh)*3qLt-l4==OMNa8^$z>QcOle*+zaaY z+t@1B;lN5+1|*7VeQa9rJt>q-v@J&4w!@5Nl$eRL%!oPiQQTC#SqF~VTJUCj7xgZ} zayUcbTuSh5sS=sAW4+U*8>LVNqi-0(q*7hrCUzme1t@Q)BcU`&K-msC7M}F-CnFC{ z7sU?i&01LIuHq$;3}iEg1kU^})+t>jib{?w*UellMBborb|2j%1r-NfL8KBZRn~!* zYxntPlA~@ns-JNXFf$|n8Sb*w%P;}`tZ~L=lIFk@&=I2vD4&u+wcK9N@SyJqd+A2f z!975Z1-@Se6;Yj1P4sEfE$L_1bFM8-RXkd@hc1ZNADJk)D;Of0=yqv&pqf4h(8`4= zrcBfjl<86)Xg|zM7A%ITW_)aI{wE`_P;a(x;k{-6%kM9zeV?2#SBK!hspWPf0DVZw zfJEv7_JUW@S}5y34vIhVoGSiRx_Taj;rWGpWpO|f>5}$@D{-%b?vfs0!%hn;MViSe z^hlBR@NX~Hin^p(u1EcJ`DU%CmQ^Og>N#D>OQN6;YNpD?`bXl`l`iwTS%-OM0?miM|7QRR zi-TkrtXTXt>rYt5iWyIM=C1zkY*YOhhsDo+BMbCc5e6xxc6);MDGTclhD6$h39{>R zG|_vYtNu2-MYzXTFI1HnGQv)smTWYR7Gr!^Kg3hIk`X?rkU#tUhpASt1(Dl+X@U&7 zdZ6^MiazX9$j7?+K4D6jp1S89BrE8kF|^tOGzhuj6-$sLK{p*VyQ>rlzzLWl?d8HM z1?#NUV9rYy&ZXt}Gn?Euxfg)JpXA5Q zc)Y&cz?)n#xnW~8UXJTNBuM?cpLc2S+^|c0_53W?eyD(J3%SKjWNn{U6uBNMz#CXe zybO9XmdZgqSbztIL|AV3S!i2GCdg1t5|q|MX0q_L{ud^9B_sIGY^aL+{5J+CdFIHE z{fO3qS5&KKP7LYYr(}H;`J9SDJD4IxE|>(_LCM8ez@U}zGk^r9axR8f9*J}mG5Tf{ zHGMzyXn3!rYzh8Xsh)EM0)Xkx>0wal;<^`te^tPMl;@+Y5NPLShV`>okp(6|EsUFV`a z!A|0QqKN0>Gm9im;vNtuZ-@F%=#OJ>2|IIQrz*ZVJ`z13C zV`vRd%o>p?V8EJNKbiBW-INx_0-p+Y|NTWDV1t+}E^GCjhGR~b=u8ee#`Qv#VsN2; zjXLIpOiviyRDw?O;KF)gAG`gfd#<1JH%d$Wv*zv<*!uIG^!t8#5UhEP zR$udinB^ag{$#+kQqy7A-L*zjNC_o7M3Dnj%(`#i6vk2*EiPZMBEzTD?{4t@$Q3&l zB!?dMzXNgNL2^@AunffrI` z;QgRpz}@AX?_0ch{p|IT6-#Wk>arPp4A#*)aD4ey%;0k`J@=NWMv}uCojfC>lTOKY zQZNcJD{5Ie^j>HPSQC5}G|piKr1Ctb*t@ieoMvSNbcC;CWruARued}PdUVNbHvZ2Z z3Fhc{_Va9YmKi$cI{kZwDN`8}kJoWyPXy+x32UG|lq`b+P1qPM{b^vbhZZ!FkS7MH zT2uYTLM!L0|0Io{}sMfZz@3Wf|wHvas^mVmCw5shAX+uFs%sr zWx%Ku2b`B<`VsktrrHK={-d;flp#1QKgh;smM>l?p9-so1J5-USUoVz28Eliz{($T zS@SUkb^`ntXujtwkJ1AX$&kF(^T;TXoT`Y`D#IdBU3v_eW7P9&*vaf7&r+Xs7bOsw zDJmf6iL^7Bv!MV2Ay*_RnXN11Q`5t49o!rR^5Fsd%#_RzLja9WLRS4b6-*`mZRFkO zoiTI$@|(g{uzo(c>^`i5_j_$~ub@kXrNY~bivp5@dg!=WwVaJ^yF79gcy5qvcfH5i zNbZKNou$88e`Kdh%)R}V{M8?sGG{go-aTm_Er&I zg0)Xm`V`H)2OErx{!2CU?5P}LG%}8B%YXY5$!~%US)n`MyIV~1UYMrGQ${PI3QAT+ zkrFCqZS;z5~!allLKH{&HtPi-hw1!4`b4~O)NCoJH&YQR+{v9NC4tDO} z>{6crKU*6a!@q0l zFWUG8*=p|vu{HI9QEw6hUW%PEFOYaF5UNQ@gub9b zD_J9cYlvKc!U-gpLM2)|LD@X?m;l`T6YO)e=GWXW-ZUWN(PxDsa)TK%9603i$Osvq zQL?)fxr1zF{gNb53e`lXu@V<(xs4>3eHxMgsX>F%{;1Q?WUG>23R%e+bvi52az~wZ zigE&mLfb?4(H~2)g_`IG(tV5jLC1U}D>k@}?1O1*%d0ox_$gn@(v{|dU)X`-uemwq2p5}7PL4@3ri2ue-cTr|Zdo zWt}8t!8MY^s5DpoReMZ93c14! za1LAyv(~6&YlxEFr$`?avl3QeAKU>ds~&L@X9sH~4^#DjUCGlW>9KBAPXzFgKIoPX zB1;L8=}~&2xqf!myfLR7Ixc9`Y4^Nd??&;kbR7??DC+#qkxO%x+t?$}ClkL&IYeN- zmYyQWJ)Wz$LWZTg=jlbi&XFr2B@uZ-Shl;Z`iY`_dlKih{5t&WcX) zkn4(k!6tY7YwLR=e*J$BJ@Z&CTe<3tX*&axprHdp&qDJ9qAg!=5bRwKt@3RN8I~$* zeMY?WT9$9s39bSzK;{zV3_3mc>UgUU2VR?xq|{gnlD-@S-T2tIr`!--9GpuRcy&T4 zdk_WbM#}=9G1goI4xwQ zPWr47qy6~|ihr;|9_vSepi>8f*hQ!`h&BW{&hmL}E`{Mrz0O~utpR2Z)cj=YEW`X3t$kZbRyuHzF=Q$ytQnIjSt3O?V%Ht|$2vJ3G!M8j zy%GkG2hBQ6JidI|Hke@|AW8LH%`-_Xb!#iO=7i+y=o&(UV_!H_o+|P4`epvSF@w^o^= zgo*KR+Tj#HbSXCXgm>a&Gk6FFY35!7T*-=w70Q z-p_hYt@rQlcdTnR$WjB5MXt1@h~2GXO~C9 zfwnGqD;XKP8Nbf*-TNX^u+AHkTY7qhh7%6YXu)$?||e3~w{WnS@FL&Y34&w85e z&RGVH>j&JQt}vLBWz|t(r2T~n>5ds0pL>+-Gm6}$Vzx?G@-V7~cIO@VpB7>!+XU&n zOHiVxk|PJr)(|}QscRwstRK+C;QtcI9asnHS@tSHw0194dh~g2cGXLy>{UDr@AK@T z->cGHjZeoXAm?7QxJ9mi-jM63uK2uWu~sxB%67rWqHMw@cP-uQj?^}LX8Pye9q#){<;=`((LFzF~1b?`U{N6o!Bu zHkPaSPrqo>7AFgxL^X6APvu+5Y2>M0YFU@)q0kY3WOT2hi#+T6z*As<1m8P`NNZll zL=L9K@%oj{eL@JnlOFnmB43+wuVEJd2AP_h>kYnoiokao7sT?pnMztLpzc*l01 zz5Yp-KmO;+JyvI5U1I7UJ8T5kg2P+viFTnHR4lUH_UZIzv)z!uq+FyU?F249DRy&a zyS0kSMZgge(9F`(+5q%@Yz|jbm`xVGrq8kUJ|p#1avyo6NJ}O4UX?Cb@iXGz&AlK`6y(qYi%!wKb8?}W7|%ZP zN`y={ayFbHi2`(+N+mIwlDjb&an++t76hojj(NqW48gt(U&(kC7O1(tz`%DIVx zTVa}Le9DM_qZg=N^0$Q^jJoWF^nv<+3I0vn_?4WK1QQpJgt5F9VJ=+*0TBFK;IX=( zDsHt1f8AlHxbv^o6qu$fE#&z&xn#0-vMK_VpNTPj0~}?0WR)Nr+bW6$f8;C&j$o`f@Y;xBJ|BpFmvXr}=ABJYKM*Oo$ zQ%IR)%t-_E!DU`8@^W#CEu=*!Yb_qna^^RtY&@7*3*x8~NL@qPJX4a{=PK)9fp`{^ z+E|>Rk{=5!Vqc=WSipI6*A4Rtt9W@r;F-Pbm*8>R8v+6vuwYt|HatgxoL>KL`g`-g zei8HTo4@O?kHr7%ai8-FKB{CF%&G^^etB?aruMU zOh8G#px0T;{m3JSP6k#aTrxp{OO9f2ovW#U!0f%2ZOiX|^(&FkfEmf(zS=@Io5PF)udfam zVJ3r;?V?C36;tF{BkJ+qw)iB$HPUSme@ge=9~SR^T5^d#3agT9OSdgPDm*RUFU)Ys zn%5F~huGa$(T&U`5Mp*jZGF%(LaGydPn%XpII06qEY>?zL&mCvE&;D1LvYaJBWb?i zXh08LKowBaGwu_{%UTFbxMf;4EknRe#Bt_+<^TSiXD}RZm&|^bY;xd?CNySEDBIXY z$x_jx>8kE(1N(vDwU5>cHv#WBXdy@HincQZ)qE85L1}E9UZ$@s##(sTZ?-UxHfHcx zk|z!Rm+^?+Pf8w=7G@DW2d=3CTG$C->Y`+zq;O+GhRQb;l#P`s((5F5G( zT`6n81yODd$)T~zTRo>$rt*cgPpLD8ce+^f*Y)xfMa zvH%`=hgHC8k+=SGBH%@}iq=ch!4y_wbMs-69)?o7X6+$kP(DVd?KaH8@~`o;hgk-g z{o%~uX|kFb%pBOU$uff3R!Wvckpyg}Ity_6pi`<1*F%wyi@VJ=OK@1IiF_Nl z(}Ma!o8|xg(Y?qgNAm{G^`spj6nhSQCAf)kmj3JGY*wW|8AYZ@Z9^fn>!#DkhVKwVQ|F zJG}3DoS(_dls)2XXJ;v(+KrXws}Xg2wuGNrmdidUSRHj*j=JTWqFTc9SO=HAa~;zC z$sP|#TEO4#N!9Z+rMm;0*ey0^lFP5^l)141fuWTky~1 zqyu}vT}DX0Ldia)NE0+|yQh0(f?EMikUBpt` zu=MyM%sE0UzFtgMztd;`fY0PH#S9}%uTGdv-}$=1);xIa8*h?22evi0jBL#%O4dk` z^AOnk=1t*W(bja3K29mroZk(tW!(+Ez4(C3cF|VBNuqkC-|s3t>Qo;V?^WxS5dchB zL>rF0_VKfj1*=)!D9{ROS?hqC@0Kvl<9NV*C~z8;^hotqXD_MYUKW2&Hv~01ua8p6 z@gDshn&@KBY+(DTp*N6T&n7M&-AM3BnVjQ`m|LRB!(^hN`O$2BL9_}7_^>d&iuAo9c5-uo%+>`?97Y+ z?LQZoR^T~oveH7e9hTL_F2WvvtdPQ%{whdCC2=(JerYeA4Mj0n>#zqtbSz2V=6lao zO~>+VK?Cz;qh&2djAxrKd+OBBGsDQ)hOQeN!(22!^^MbH?JQ7EpP19xPRX`XBnc#O zW_56TJsZN3=3}Mv8V)wGlmu$s5@B(1T4HzA0Mj>P{rfk))4uzR-oN7p^e*tea?OmG zLtn{%&h_4cN*AsFC&1I!E&haiUy;N?zVlMQUM_Rz);d^P4|`+7^LWp?pJ9B}OTStD z1JgELhn;K+j4YgrlBH8*Cl!-HpO>$8>j@o!3MueDkntLe>X@`-en|X52-Cr&<%rxI^_|YM9++1;;1D85nGXl+dN_K`K zC#jekk(zZJ)V?+>(9%8pOHj$ZiibrqJv5ek=kqX**G=a@b#5z=ZVri_^x=YF_>zml z?$AWmdT~7mmnX5jnt(!}v#VHwH0{Z3?9avSIh9=Pv&+35`d}bND9K|D5#&WX7I;zE z!Nupr1$EO{b3SRmXEG3bY*DP)m?Q)(Lu~kySKbIRz^S;WXcbXAaAc;z2s-7Itds&m zkr*x6rf3aG4C(bK^?N5(79Rj2D&cwI#jJNq{kDYMbM2DurHe^V!0pA*W_^v_=laMU znz>Pee(UTWes)+auZOSkS|ff3v9np4BCQM5HxD4oB4ny)>{K}rZZmcXo1y( zE{oTQn{DUG*cv{jFqrHYq_W5PG_Ij$=*dUu6`N*qOqGD^J)exF{NcIX_#1Dy=l@ znSGFr54PSMUioVSM1J48&x>?2gNOq=LC~;00V4gBtd}A^R1Ah>a+hz04(YA4+gy}Q z{H-BtWEx4npem$?9{BD%f8EShW{0)N*Gmrvsum{-O6m8jXys+Eg23YeUCUBl9d#;R z(M#UDCd+4ka80Hx7JcEB=7E=#vrB=g01smP22km&dSv=-c<~8ivNGr#$#lsHUuzh!C^-dQ9>i(j< z@+Y#68R{H3qnT@Dn08XK6pC!2V(z*l-%W;&b4I5QRuwV~8Z4c3bNsH!3O0<$Y;RT#JE+)38({MZQ;K}S_~5ID7z*JSL?j>Wgn+vR*?sS7WepBaa7%Rw!9fn z4boac9QCe#XD^*C$e@#;mI27#T(RsXcFr!j7U~f`BCE)Vuu;&&xwbTx*AR3*@`0c( z=yr6Y;B$Vq|5=#GhPjV`&FJGmO?0gH`hfksRSSSqNRlhe@GGV>WpzPQw@C~kHvxUm z+?+Br)w2pRrn0PdVDHt!vU=DJS4~*`dkTWX*wv3=W0gFS)#RMS$x-C_AM<}C9CLcy zc{#7`?>pbH@D)v(X^M$SO45NLVj&@gY@N8sQSJ{c43AxOiPr0*XhFe1Cv~t71)F4< zVOU3w?c3ATz&#loGGzl{#RLB5%%Rsz0rI@%Ae*3)qE&RoTPuOs$LFwlL)bx-=UMHc zYDg>{SM)*j12X_Og?Z7Xkso@lbD8>_V8yF-)FEf&sm#og{qh&jdqtF*PW9`C8cLiN zFb$m+eGU@R>iJiFl}M|nKMn2D&_kWadgR_Bt_#X_?wO8$*vVjoqO1Vl6fa?fkr^4c zIX`;dDLyit??6PPa;{FZyGN`BcIGZ9%U07@ybrxJE%su2rP!6COe%rhY^nK3mv1UD;4Kn`*KD)2hROjV5CFh!3?@W zUZL2r2t#bxRff_x4H3r_P2w6(pZJ&pkE`S=--NJUlIPz}wm?bKG3P6!UDyw9%z^ME z&K_2dXR{dcZSL_rWPq^;Vi+OgDJx?*U^J$-*|BV4ev%xgF(qMk*jdK{J2YzdB2lH9 zKJKFB>X=V2%`IEp9}&knI2T1DKxSfIAs>iyGodlDgS(29@YgQ7Lhg|o(9|Jyfmv>! zay|vIi?xdqS#RNgs<=J;A@Ui>|7k@z3cMMeVZAbJU*zXx(q(WHhSMfkc?v@leqfTD zma$?kG`srue|TWp7W9IIOfrl@vya6?J6zss)&XO$;^T!iBJEsUDV5vXcQH4Z)*;B; zo%TJp;eYxYP35Zur5i3>vp!j z!8K7(M`{%N=y;bV=YD949wI0LWeW&=*=S7)B474U8=U`#+n**y8LW)|>W%x!5oT7# zf#;+vMpouDC4**wI^^+&rVyRbxn5;Wd7FJ1oY-u35vwE|bR~0^zw@3>NF{@Ak;8m+ zm?X$rMMq!J!Y;124!A(o0%iy>&Nm`ec1f|3q($B($Ks|8D99~|!YpDxH`Q+}bku45 zNOFY!Nab8)&D_GTnk?SN){ccIb8gDc; z$Adsh0#L!b#)d26L6;Gg1e(|hk?Fwp(k*H98UyYanBqW`qSEElz|lpvSb&?1kBtyx z^3M( z$vhc32FQ4F(==@TWz*L`dfsDq@nJo5F0IVx>F5d@B2YH=NHEfCE6*&hlwBm#2K zq5`(EJG@(x0*JW4B$6no3@M0cV8_V@{GiQQ(!fdPJQChs+#}8NYxT#K)oJKqY!dI0 zRRtcFjyYk=2i7^Hv3luTc0T)fK=nM-DoByy{aE~uRds#Bl(5{e3_3T=1`vA;1sh}S z@pGrTk(ex9ZK`GSf<*Z&sND2R6J%IRsb5v4NIwfb#J&!`E{dvymH?!d+>$ugiC%lT z`lb7!wuHemhHEkidB)>xb}E1Uv9r&6MFZBujky#DH+ZY$V@@b3dJ(n~!eF$!q^+WK zX>stFQ$tVz=P-GaZb zH4zHa9|*@D)A+D$zOdL4cKHoUK_LXapfM+;O0D%c=$<5751A%CM=Atkx};^2>#R}` zcmGQq(J!uX%ITmVx#C0Y0JSOLYuu%QKg}UzYR584Yn@Eh0~|JMZ9&!`jinFJXFak|5Ecp0EH(!oU~=R2$l;a%w+$Xnw&s9UjiQ4TDb&&zLxgK8td z;P&EV>A|Sa?WO)YbqGvts4d;i%*4F)dNogMaCOgq=k*(8i@E+(2M+QcGID75P%_9) z@1kO`556DbBWWTqADy0u$<~lvei?MDSc!8|iuB>DpLjO$3#ekxWYWV=ghDCMi-oAl zG}Q4=7C4p$+*Zde1Lvjs(lcHLko>H1#$}SmY(3<_@PXyeg!NE9C4&yey;RIum`@H% zdTE?yT1Cy?cR-jb5SHKa10E;X5Zz_pUJQ=T7ow_=XDy-}+17Z&sPAq&-)HA{MmJa4D1~P7}Ae9CA^9A=)8` z<#kDKNy-;Y(=vDBz*#%?6YrUhZOfC3zq`iUbv)Si^>1Gy8y&d69mp>xOc&{tY$pY8 zJVpydC_}EzA#pFS3C88Ea$Ij7>;SG=FfVxY70l>7w8v`1>;PJe1bfd1rhhrhNH9}I zIrG=wu8*C{Y~sMKhlSY$i)k}wD?jINl&*6=I!EWP4!M>?Vi*s_y4TH}UbDgoN2Y^q z$Nl*`4}3}+?8e~teb13qvq+}V=edQFZKlX3DkhefPnw*wh04P4!3b!rS)we3@{Me_ za*^G2PQdEl8F2LK zfd+5#$t<$Yh^cd!l0Bfv02PB=Gld@4mUbVgd?)g;~*J!9|7{bl&}#_nrk_`YthOm;11^QlvTQ{bG`_Xr+tAB z$rks6q+Ik!IIu_q(u5iGpyV9c?0ea-&!d+sF^-3Y~ZxBGX?lJt)UAd+WDRQs=zd;Q^b3)U%Ag?fSo94qucq7 z&}pcWpJIK&uZaXs9cjkI8YH`p8j}&$UHfF%7}*NP{nY>UVQT+x4B(qN@?$@ub>OA* zYNM~>J|*j;$mh6p)=9Bua{jtHU?;00Fw6C*AJ$DPhe*Ye)d5F>@gUYbVoAZoy2sBW zw)vL1>|>31onSY5DdRY4pg-0E)sh%@{hU;Clq-GCx?wv&z9656RZ#k?Q>0k5a_6e*l@X8;@MOW}SRVv@6^3tu(FZXzYE9KlwgGoe z0Zyw(9J9&OfxYZ(qscRcl5L^LW<)729Wf0J{cA;@r9Qf4IZzqX_k~DdJiY5kCMbGh z^z6U&W=>7_-xn@gMLO#>5K$F!+&vdcHaAnfE-E<$fLYVaXQ>-p>v3U!EKK}`%R}G% zh0I_tTptczBRiPc3&)*-p>}ejv#q9NAmNrp#e5RH-TQ$1dg(=BQwY#v2RQ5y|S|&DNlJ>69?Aj+p1Oj zxQ8MUP-A~2zN2I>AjkL;6ut7)@zWF-8kN8QPej%iY!qNgzOz5!x zL6G=ZQ5Ko{(hk`T!3t22fvyka9ee=n9LVmH7=nCJpqCb08FFOa5g+UYnaa}I{6{bjxzF8sbBo^XAhtIdN>^}`q@3OlGjGUx@41AtMrlkldcX>-vCuslw1b) z6#0JCE_ev55XPJaTyw*QrRcrxRbX?-I+rJ_ktqzr)3@3JFi$^q>PHv>W~OAY`Cm*+ zZXMRhSZFMaqsG1xPpx&`;&z;SJ38I>rnHH>-SrAt$IEm{^-GUBE!Tjsce-zeAVY9L zgp~Vgm%)X{=V+qqNK1HKkUAhyppwINl5BE{9%kE%f6fMonL0eoJb@8nsO7IbXkKJ+ zZvXwm#Ro_cGw0TU7sHK4lUOY!tD-={HKtf_JOC#mm7rk2 z^vi?;f-xr~qf3<)2Y~8NBCFP?mR8SEVolvg(yd^;GUsK^yC8ocR%Ur_1YHpI{4u9X z5E50>z0Mun7I}ss6Pkq@g0SK7Sl|Jo-=FFr8f(3Kqu1lh^`(;~5h>EXfGmQibox=v+*Z-BEBZjl3WO0ld_3~0s_8?(39XztK?_MeaT+PPGu?TIa3&Y<9BXxZ;~Lf_&i&kU<4BC z!0Y5w(<)3ROKb<8uPoGOHbbZdqy9TZIY531(-Bs6>%C$7gT`c#t;LYZZFb*v%lNRu zRCvl^O@{?3sV(G)P)~PRy|`6$k!*H3N2Yy%aipvnWV?^3nIB+eC}zI#t({-v7@U&% zsc)PnYaQ4rfw1d@qKEC2Y%4{QsF*&wmecGW%L6SRX#pqQw~sD|5{AEYN*jP#Za8Yl zV|wvn4KNH&rJ3Qg-yuf8p#Bo`v;XiI_f}uI`Zn3dY{lciLE2&?811EGFr}z4{nsiw z^zx-}$P#0*zeZlFSP}V^ZxbZPpnI~U3N*~D8@Es_e?}h-9Xwl z;`Z!66KX;D%(X3Zoui-`v(Eck5dOV;#t?+Z62+#J;5+XCsrEMQkXHr- zR(1rQeH}J1C^od-tBI}?!d?64c?AQ&E=M2D&o#BpO&9iP1;HbnepE|7U@;WA7JPi}LUGg;HMs3Z+$U|EFui5l9 zTfH+8r`%uj7#)7DTidfTYfYiGrXim|nG0JL4$;dlIt_>ev*l|eb!Fgwtc%20{hiQr z0$@RI1G=4j?Z}*zNUR-)=76AmT1R*oHUQEloLjjYHVkX`Bf(8?ePpd^$Oj-U?8ff2 zaAWf+7D8^DsQ4;Lg`h;7Al~EwzNkUn3boc4OC52`go?Flb-CoAkxvoAuX4oq~*8d;QXc=#ku1>H4RiQ4awN;Of{+LEiKd z>1rs91wx(yW)OdU-ITo`C z+FEZyQ`vd?9;R=K=xvZx2Ff>Bn^saz^J_XY8HK6vF+Y!96wEruGn7BzJJc$P@P3c-G-j;Wd6 z15HMoeUfLHgdgn3j3W%lmr(iQ$maRydXSgrJ!#{&-hS7^jH3V6UHmOs$9Uj|B*y|isT7+;kyTXu5$JI@rhg17EF{?KH4)pGF=opaM2zv8 z(=YKt#N@4-e@Oj>84+RcPMbz<@mp(Lc(1(FqOo~^V*4m^pNik6MGqBoJbG<`wm5j5 zcWJ1uC#p-HBp3$yIXuN;$2NIoNE*abAc9*aFAZJi{Smn(D5OiISS^YEPp-5uLYEt2 zP#4mx=9Nf0MY+;te!Zf4IwugsVKP-XT%U54SD&&Qk^|RdxuGL&rmp2S@RCZz@Yq?w z{nxGubh*;2uh-4erTUe5-WBRlNsy>(zCVa!KY?OhVWKQ*JBetUHq?{ny~PcUebv zU#biMM;N->Lw6tqH%uF(CDM$bl^(jhn03&*?x=;xcKc!D%-cSogP!CzP?B0@gms_t zvzTz*$TZ>TQePRik1;4~B^SVOJc7;BhidE;OB0@c>zb^HnP_C#h>Hmb!dU`}8yEj` zKB&x|fUXN8h9jVx6ju>jPYUV9+I(1QTIME_0#MF00NqdZyxqb5-)>oWRo1OsLV7}4 z<5HN5#ISG+@cc{!l59rL=%vGQqMdT~ab@#w>}mIL;Y~9~yU$)fv?1tHQl5BNcvJYgL3*GAiYo_C$tEbhlnVKHOkg5@M?7Eb|!~T)H zPu8d0q%k&j;GH7+K(sE6bV7KhC>+)5oBazQgMdAXrhn{1+PE+@8c*EEc44dde||S; zk5Hxy+Zhg_Oq5+t7w?Q1Ru(atf^=rFI1i$}OT;P6u-nJ%ZeP=RYS2~RD+?aRbSXE4 zWU4l+m-+AZJuO)uI7l``w#I$=+#u22Og{h%S|5?SZMp_vMcz4ca#1s2O3tf#%}ZQNUL(;k%>RJirY6x677 z*!gGLyPT)>&fmMZjKy8G_X?n9T(*&B|^y5)3k=uWRqUXE!YuS_gvuZ6wYPO(tey#?wPAhm3?H$z^vkv&k%e~w16Of25QEi|f^ zp^dreTj^Wi3vKb>>zt2l1HQv%MFZ2T7@)J~r>dM*Y6;WH4=UDfiJf?O&g+4%{cfY! z43%G3P5nJda^c0M#A5O{Q*0JGO7Ue2Z>lasyH9Ieljy$hp?P~fjtW3CieB%Z5!)+# zD%|3|*88Ed!+%#qZ>%#H7K^W$?}S^t?8FmZs|@w7_N}S*l*qZTBg;`D*ZKDiko7?t z1HJk*i%hpb)zc+k9oG-MsL5@4=O>WM3(xLf@3lvAo${2Hyzl739EM|;&gB5-$uuo z*OV^JYuV)BE8~7#wpf8qQ!H$Uk5TdHdW^W;(KG{#MTKt%L^rTa@HD$8AXD{`cLAss zLjDKInK~k~fyF63rZ^b(KA9>!$e33fSuVT@86!L!aVuu>gwTqn#c^~?K;0Wub*Hw1 ztI|*Q1dO;fE2?9+X!YucvOb_I9r8@2c0{8B4Q8NBPmB|aHlfMdG2wMSKJ3kT!}FAN zm-4f*d>2k1a&#^ayq>2q<=8gOU+#I1K-FXmgA^zRD7D2-#Bt8!_`YR64(d2RVSKM~ z`zyniOulkuU z%f1WTKi{!}S}t2dH1CktdTaKSmueB>U{pb6OWo-9cl2Z&_4DV09GmdVXFvx%%Z*?0 zH~wwtWtTPFrp5WjU_A;OVmX}NSk!bWy2AsPlFZrIq^wsz@R&GE*o1^H{Z0qX^NWY! zM#5N@tntR*23u38xGXmY2PsO1w0d4H-7CU>VgFJhz6GTNU7~K~q43HG-BwLAl{1ZI=jY#+=u>j+4a`CSL*uk6t|6<9>n{OeU>-OL@*ZQOuEH zMP8%D)J@evW2AF$>>gRCN2Q>~V=c)O*LW_Q)+ze@+MwrRrYz>x+)JUK2@J|5;%X{Q z+!wQ6utWAZDmiv2BG)%DC`pLkUPVY+SP#8XP!aN3)F;Y2v)7QjQ4aj2+cig)Gv?4q z<26UF8^UB4ZW38nGr6pd;iykaVKZLCe_P|$2XB-&tJeqT$p&Q`p=K!sbgnM?f1=v0 zJnDz^vcqlzq)>ay6RQCnh*8cT`nbT$Ihx~c@B{DPN7BLmpU2`BL|r4BTsXf2Z5^XH z-byKUJ4Fhp_!f0nbTWI>w|`D!SVj;;Np1kK%CIs$HkZCNS9dP#vFf8B-GzC3ffpkY zG%`orwtIn+6tL9l(1$thj@#XIyBpG;xlD$f%Kcp=I(}ZOwo+zw9d# z&F5gFO%jw#Zm3XW70C6?VbtpA(oopU-&Pbe*L`z*yXh{vDyE3WSnT1@Jf<63m;0O` z$S`WwjECcdl=UzUKcD%ZyDw|aCgoK8pVyJfS0?V(WMO~~Q*0eY4pH$a_HJrihu!ii zWjlj?9jD}#5qLvqs%Jt+7|3{_`~X~KQyb!X&7M%BjMxc~o-1Z@rh~L2b4uF?{bxFS zv>Qq*>qw`(%J(p-4>M*6aEumpGjjpteIIxh(FyjFe2o)CmX?X*%pPRPd)_>3t>f&n zz6*ykGp0v-O4ogW}l-nhzgoALC#rYp7VX%I?c}+!YQ}-EkD1$-1;cgY+1r{ z|B_A$`B@ehP68dZuq=Bi7I?pQQSk-fpx|nf#qJVRFvp=it9@4G+(RLCfn9W0hz?uY zDa%NQ^~F@rGq9gA2SY0ay5sI$@?5%5Q!4;f@aw+0bdGOTjKR#U0jgxm zeX|^!MzR%V90tSkvI7D4Lysx$Hj4DFo|_>^h5T9X)9A-io`# z^wN2+HTW6mwIEVmOb-i^vyJ{Xl_#N zK2pGMMRVQY6FFkBqV1trh$8Q#;xWsaAU+to#($$K|BysDocqRnSNdR)bG6iUcc733}y#(FkTd06kQPI1s2Wgg9`AQ zs)5-+(l6UG-87b?;ot})7;TD8^K1J<{FbN#>%Q`qwOo+P;)X*2s6dNlIk@-M8MWJy zbiI>C-FC&d`8!KGXLEJ5%2lgz5utPbe z)dvB^b*~~DXhq8eAF8u`is-E}1M|)xNoMg9-qIZ?f{(=D82+WZ_G1cB~3di3eVG^E!t*0Hf5rm zZISFiu*r$CS?lKMIRA#$svh=jF_05 zM7hgOkMlxOEkAL!-d6D{d>Cb;bHcuf3EEcIU ziY=x{Ar;>N-+%WUm7KS;#futf+);E`w~yOV=HRvK@o+Ez^&Z zejqAXr8rBliF8>=s~Y0_YrHODNR8kBm90;Yb5=$zH@1Efzvp<(itBDY^S6G~^FuRS zeqQnKL!{Y-Yau?h!0`==y-JZbDjv)FyC5yNLVR0sR|VDReX2xpuC#|XX)Bv@koO_N zoF3E48ds;ZaA5iY$r*HyOhTC`bd@VbYlQO4wUG&89A>>&C-h5KFI@J9UJYbp=%by1 zlF-$2i|FI-5OIb_a-|6oUDI)Yj)_koZH5HcO>g6;?DT0CAih0{=NY$By}r9V+T{H)x~k}I zdP&3vKutZJDa?t}tMOJYy)*(>w;O^gxs$BosQDQuMn;X`oI`#XS-8+^;hSbV^hA*V z6H+^cv|Ie#&rSgd)d*tP(?1S*9E0;SkZ^IPM-ypkYJ4c{og z3k0|)-Mb?)1sD7>RrjV{fUroe$GvHvk&W}#1H0`~A>P3*#34aPWFrWW)zW9w59uYI zAHwErNH7r78rP!ei$a>FCU<-;McN^$Sx_6gg*_bF2k$(H8Rry_bKZFTipMvvyyney zWAXIL1?#?N2Gq*;4?QGjUYTg|U5npE2gP2Y$R#QsE75bI`93wscodWxbX&LuA^=G2 zfpkN)K#$l#cFs$OEGqW$=P_jr#0Y#Z!QIRmP3jf~uONGBm#UHA@m8-kwg^(OH=uDB z&#SdbewdiWaq0qxydZo593uq~hfWaJN}3f|qk|P;=q=*?9@@Yd)aX5;*NAVUd)Xyl zJZ;F!)?bbDewt@H~;;IikH3BWl!ub-@?#a(3ppnS_SlCwnx$H5V z;>v;^$)a=8X8N?m*6FhG+v9$W?|gHQ%;KD1%&fCU2cM!**R`tz_Ku@+RhuYw zBL#C7|G;mh_nx3K&6=Q3!nS*r3sTqzQ;%pMrzv^piRZP^hS#5j_X+w169y%|J9@_g z<~gGsd4Co?-MltwlULP|oGIj>#n-xnVu9GFh>FL)SuBicfqt_x3ja9)tO!_m*$L78 z60z}$3=OtPKebP#s{#%L%p96t=uusUw%j(b>R3DDr2ZqmKUyo3Ydn3<_ciWYyneK< zyGa*l%e4#4cxn0Z(gIRFWsLqh7q%IfEWmw|Vvkc$D-mDmTP9or9><+ok4U=o&}(;~ z@OQJKbNbNy9N$K|fLbmzF#R)%y)QF(H&Ia4FM=ERGpp`5T9XsGY!Eru0ypUtn?jK_R6KTQL(yQB|3P3GLc*7lH;)JG zRoXL=<1{h0&Hp$Dx8pVL`5(h@Og^5O_{-PL_N6-f_bW&#zrB*{=GMqb3j`gc*!>jQ zL&dkM4|@&DL4KlUdWQt5WpJG}>5GFJn`fU26R0T3rB^=*TNRA)9h2x5MhuI>le{r& z(WJnz0I<0#`vkiKpt`hG-NW2c8JNRib>YRB+$B$Bjth+2hBZ!W^6U*e@HP6?lHje1 zU^9B&-+45ZY#nQLa^2OHG+1Dyiee#mwS$UBDS;B{kXLbV&GZLh55l%gYY5QQD!SxH zg*^%#c0eHkia~WLx(k#`b}Au>Um98-jlP+&l(|_{6<~7GT8!{oO-|9L=%b`V(kbc` zIhO6jk=PhLxAq=$+o7L(^*wPnqo(mI^?xQy$3l$@Tb3;rsL7z%bc&=<@rh!rmDd%3 zL{_4>F6^2JEym~nO!Nw6Gc4?#XP)nK!=mb&;>oX~N{PX@g6@S?WSRK3Y|uv^)B+i% zTFKprO`1}=!Z}5m&zX~#4;Q=NUwQZ|W^>W@w|5H3?kVKF#V_zE#eyNP1@90vypL#C zsWPNVY)WK_H&jXO6D$vEhhXO_)h1Pjve6rLymf;h2(>QK0BOJjbbkzPqzkkd&+Jnr zvnA3ZxBDq z%sjiCzv#+y@bKbW+`rf{97@-<+SPM^6J<|a&UOFKxZ-jc*~KjqI=DK_M&qfeu)LPW zt?`H(-bxhbN$o@Ki!t=Y6MRAb#fwfp2oYTP12G$!O1Hc-WW_794+2W~QBm6s6q`Yj zbSl2a1F2P^A{2V}Zu<5qi)CQ(OcKIPZik9Axolwt`EpKKZrcy-TckdLH($y?h#5h}O+~PatH~rSHI`y1;^}Bzj`G>68 zWO#lyaD{Ae;rgjP7Dk|mVgVKTRQx&rgn&d)&buaSiCg~mHQ>~`5QpbCRjqNIqUxA> z(gVsvA8N}%A+1r2s+XtGzq-(HiI0CQWD zs7diO*tvK@P>d5(PQ?=r{_)o4cO%U9!ff_BWPc_BvrAQSOzxs`9 zvevl0v02lKLhpJPYJ2G46;mnS`Q6q3wRhnXD$lpyuRW|moyIQ3i}inFy7X!E8iJQf zso&|R7w$J*+GMiQi0BgZ#q02N7pPn|&Wlhytg@RMRe?%OKAl36E#$jLDE0|O9#iqAT4?AQgz|wyVaqkS($kP= zYf`QY0fxC^I9#sTIloGY;SUu0=!@x&N@4nx=lq_?&jQ~%GPh)reCego0?kT@rkqmN z!gjMfrct>{^-e;>F{YiV@ko&-YucFu(Ve2JvJUVonj%}(OK^PDYp4a)%;lN~+NGLv z+E${g4EYdNh)pv}ea>kO;GSl&OCx~qMfnh#*={OZ3AT&W!^2fEN3;fYE)9a?k%tyO zh)5CuGwXdIxkI)P95ErHM26b-=;Pu5=lqfhQpa4D-=5Pa`2X2JoD15KV!z?H%?Nz& zLiYE`5f^rnZ&)l-A5kn2pq)WE6TLPUc1b72=lq&PyOo2o>&6x)W1)A{V&O*l(_loI zUVRy)J+novaacuHExW5K3V-6coz4nt5|v0z_fVj^fG*VDncYhaKpyFlMsO|qW9Bi#v@GSql3p+#{@(P_I?C{mAFG}j6b9ckcdT^DC#RFugAP=lp zm*}eR(%=?Fm*{o?{2g~tJwy+?T@S=RuSRE)1g~MYeN0WHL3tsr*?)i_O{jC+X$PQ) zH$rTAfg44JKgi>(xzt`N#BeaCHT}&s*&lC$$O-EG<*}R9xxU5TmCO*`?}wSDo>}+l zJ<=0&qim40#${^uM}XFFYz1T}_j)`cjr6i9MS_~?+i1MHd`fGa@vgFwJ_!DI)xtp^ zSRifr-fe!y0l@j(9NGMtwmirUoWGeN|MqP20`}{gxCT<@!VB0biv{cu#R7%!UMfCy zUcTpn8To=$v5)1wp1LaE6AMs<T*eAAM7Ih3y0O#S(D_U?SaT zH^AIl`#tV&-QKVD1@Ciz z>tpO2T-WXQniuxHmtD#2=$nO_3-dBnf2^fT=j;u~6Qe+>=$h>QoVu|6GT5G6m^bXU z$^2&-WEr55Ql212stL!I?>~c@apTviI&P3Q=TosTX`c5Vw^_4*KO1jzU6?}y?&zwy zEQ$sHY8@57e2!i_;?|;WWy>X9G$x^tW)m|@n8j*=rnIL~4;?8Q_Y6L)FYv;P^~+8? z=f;a7JU+-;go{J#YK!)oEE!rjuf`gHnx#E#Sy(5~Sehuc_3Ed7cV@T7EmkD6ixoqj zdUc-knk-4MECf?1NKU50`3VIW5IDWffYDyWSW!)dTa2iNb8q$n1_hVy)^NU%OpT+d#!1xwegK?^C>xJL&x& z=YPw(h4rOcYB(BIf&LN}6TP~U{fth5#ihgJ5|b!yWqORu4#rkXq*>vggx;plX;)Is z-hGhCDUPmW27wy_qpoHpt}J$^09k*H(O7J5`CKMPm+oH^jeI?*9SBakBS6O=QO_F0 zXRg=_CwQM5QI9_QQRQUwPI>8yh+|O8!u{|VhSP&Ws>0s&u>r%Jk9SnJ~H(Cp_iSfJD;3Tw#oCPunsKtd8FJO&?L`?W@Y*kK?gaX zEc3f6GX~w!3BeOg5vR>uJyVCgrYL;xR6AiKP@Y*FOcxx02#z6m0cBEQ>+jQFcH4pT zZpzuBm|9g#o~FdfT^?LteI`Z7FW&o=?!SKe%b)%BUDeS1;ATjy! zg7@CD7NX%3N^xNagG08)WK+6mq{HY_9hzgL#6TiRv@EwF?SDHuZKkeBx>Eqr6M3Pw zlk5(t@u&;i9nc>GCJCP^ptIN&Aw_gyG|M!A}sDK-lO3h{~TN5VA!^K?l-Bb_JwM780o zr-es_dO=m_-q05BbZJ>kS!XGJi*GKWzH)`A)b^gF4_LUS@^*r-a)6zpj1D zpl9)n{uvOc4_l$ZI<_u5`4+}IWj5evyhGdd-mFJeA=W)Wd=?tl4FM94j-h2!_JwA# zM}Pv%`wdB^?{CNP$%9d%h1+^Uzepz zmqz5oonDX|ur|`UOOee$dB(SJoO3Z2>Y5sF{*yf-n=YKp;}F?gH?_m#pms-ep0LDQ zuSH5ApbILC!Glg3FO9gJCaZ$8pC;-Zg!*dkL`U9}lPm<(G9K zWtt5$PXL4Lb=jw1Ns(@*R()Nsz7~1Y=dcv}xq5(k_NdUgrUPedoLI1LqHTDa+kO4< zPb9y7*~QB~nz40Ww`8&VS}`7_3Kn|+4t$%m#{@e_h3`$(3L!MocoYe4s)_{IigTK- z6As84CZch9e1NcFM(uor8xUipQK+C_?M<*&3gY9YxNw+cr^V(tpJE|szX{3e?g~L~ z3WPiE3bA;^#Px3CcgOfWE<^bDh+DEKS%gxa_=NKf%s5SqooKP;_3&SP{Rc{ONF?{Q z|9qch@@uAX;f1Km0wVLIn`IoP?kkpr~ z`WZCfM}^aJDRwhOvZ(kQ%A63~8D$4=Sge3;^>KjRObcpIjbqa<4!;19DjF)y8FWyd3Mtpy$AB6nA`$!bWE^;KzEObxMs7Te;gL-G z(Y0T^Kf`QA27aPgPPV*cR;1p-id0f80I7_M&-=zT*;YZ5=p=clsFkD!A;*$ly8-^9h?Px1r5S{ z0jk0}+T4s2NTc?==0huh+Iz|kq`a?BubyQF%^Nq0lSm0a7sZ9$*b^3TIzX}eD6*T1 zH>h?hlRO@JR?(drtew}Z@5?iYu0@^4)QXBiQiLf?Iq(Qz3ty)f21{xs2o)rz)pboj z7KS`2sIFKmX<*t^L!RAH$fZ(5*9y*(Mh4qRQEIc&*MZ@Fhr!Cx=<>N5ZtdT`RN)j3 zFLuD_nbgbcwb>!aE?FZ*`?FiQhMbih7M>Fq!DhOM8JK;Y-bU9-`e(Gp4GHoE`0N9^ zD%8;MuF@sW>zZCg zs^F&zqH@jNouVFvK5{!bGV^2KB6^kLOk{t|h}-7K&d3@Mtk%G8KN#x1J60U-4j2JS zke$9|lp}w*e7^8!mw&4(4x9$=AL$q_rRy&4#c96Zv(Ejv?0$$Nn{!5;BGs!~;|!`Z z>U!`lOcZG?P#1qol@U}OutIo;zCkLKRa0~6&mgFO8A#MN3Qh$bhQi3!xZ?pS5RI#3 z?}aRfH`jW%GbfY-;}v**KEND-<>!CG`0nz;GUcKD&$nOp`lP+F`)6SXU#4mg@UC@= zvdAgz`oO*z;FWi8}cmF+!EZ->W~aF zTb>FmsHs79(E7DcwHdO-xPE>a^VgvQO|RlW2<&)5)A}Mm=8pvZ7u-Hej zP@Ymj#ka{(hZp&1_3C%93btOJ3&jQ5@)UAATCc_|PcGf9y%u>w-37vgdUdMsTF?=# zUYjAUo`=Ms*&z?;g!%7a1^(xPV$BNUjyW>P?}>MUhhwE09Et@y=HB{MUgl>~wf9et z44cyo9p9PqP4bx=6r>W5-^TyPFXP|)ub=*6(Qm(~C%68ji9UzW{%C`;hGJJzWEnE4 zVH6kHv|uIrf6%*w--8WknCP#|_<8)$u+%-c8dz2Y=E04*g`=xlkVOBGJNc4!{I5mM7d+8JzR2Z4lt7`&VqU-4v#RF9V z-3>&>Es9Rj1Nx|-HLhM%>5Fe{kT$~GDa>bNBfEN9pHJo7U4qT-d>0=ZVD+38g~u(P z;N%4>YGLjVI<2!v9LXBwByR`ff0KUX)k7C(+exnUYJgt78x%oPq`jah+N7+KYze}x z^m?x*x>K}N;rMv^Sl`CDredst9DI$}RJiWWIzPDZ2XSW8)1`SWn;hil2fJ`pa{C1#lHQ@BKlxxg+RAlvn~Ym zm!az{m(GpQotJDC;M2W|qmVj4iMq0wb&80XV25prbW3Y|vcxVc!f|wX#x+RyCWx{&-dAd)T1QZ69VY{Z8D0omv z0?YqVYB^JtC@z+rf^5?}@4S;8g3(M9kN=|bnd!mnJr}{>d1Cmx+5a}+rIy~LIu^D@ z)dc$~Tn?KRNSBbOL9cs|RL{F5=u;sISBuN!x`Ux-Wd>D$Og}`6T7j4!gfZws`Xk>y zL4Kf-4ef^N2H5vZ)eY64DUQ0RjNa+#0>PF{!mqxGd zpOfIP+sIUhZiR{ozRt{ee#qu`_}LjdK8WY5{1)FIf9>xwt@D;HyV>H%QR4O-<1of% z)tf#hcIwQCTPkBes!7a!#Lbl8z;x64m}AS~cpe`i(>319G_7&BqZ@>u$@hEV=&f<(F@vh)pd}Gi${Ue+8obsi zTBZR7Jk4=UhTt%$dY01{B^Tmai*h4 z|K&9^yuMxf^Iwq7E)1^&7Vs*g*kXzlnsgP=hpnYC6@~T#mn)$8!fL8jiZmgnvAUF? zt$=l)cQnl)dVe+eW-O@KXR0uexYw_j?3f9s2!xf+Aw+ZpTx^8^pXYeNWwK)KA1bU} z-B+X+g~Ks^I;%-}N}B-{AX#i9z1impWYTp<{kpaFN&_i^v^h|5XeM%{;sqo=NU-zS z5PgUb_J+J~o%{74& z(&zTY%(*Hy)Fo>4a(iOhj=(fmX zkeJM!@e~?N2AC9P$g3;*0ewL+tSkymocSoWo^GGj7jr%!h3R2#KsY8Ld`XD&tOc+6 zf7bc<0!NNM;I+t5-BtJd8M9CF{>gz8WC_1eo(o%&d<$%?qu4ZxB%_1;IdwA@`0kPQ z(vSRmp^KnPi>*=NvNnWco=uT=Q>X;o2&`~1X*)YRN;1Z* z*f`JQ3dWD|aW-G#Wh*9W|2$jvzs$D6@V#$;o3xG9isZtnqQ@2%;V#AArbrhwG%7BI zW=+$p2R&**4*Dnr)N?_|IX|$7$&(sXz)bAcs;+=iDZSSk?XsCmgCWGP=!>fGg(4bA zN8{ES->eIK7=>!oxC!cwEoSNhJ0dX|P@ugIeIzyFlQZy1d@3dKxS*Ih$|QiE^`77Z zpaQ+9o?cDv``%aXfC9i>0)uM3e^Dr?Xv$KAdUZi~y4NCTth-NlND45&R0iYV!>hj2 zx3E)mjwArdGCZ>llDU{97}Jc|iojIE;xCgyXr#?!I_UX8J6}o~FnNDt`6X*1d5Fl$B3}>N4KkC)h2~rM}s;@Pj{O z`&=ZK-n{e=Ti#sizd}=}IU*PmqzD`7%7vdUtnsLUKXo0F1F=V-FDXIXq};_AWB;Oh zxUo&vMRa&tAiY6!jsNk%AF5Tu$_FYOXv8gJZne5xV&oQTCurYl={lxHjX}LU%_`_z zsqxtCo&~JqV;T>xIkg!g4zM9y{|#eZ=Qb*1wL|>;-@<3SVFuKk4?gN7H7*ROHVZ(V zp;!>xZKUE+|EiuIjH&Y5@fu2W*G%6Ow5TR>UvQdmqu-T?ed?u9e09}-Dbqj?Lc_s& z|5dT|q$O?xvleR2t3dq}gR|JAuxv`&SI>vEOA5*IDUg3=SG<;~>J;G^+vzp_8~vb6 z3-Z&wGx`LWsm^mRXG&$Aq7#1W{kud@g{eVTSk#E;g`zKK(D``!^0W9H@xl{T&^g=V zRivZpkNHHF5YaU~gER#6&A$<~ShxYma}7)*Ag@VO6Im8h5?m`e5&KB_v8X0CQP?*> zMLGzz=2gmZ!hy#WjT1RMK4E;1aw8}0v!6T=nXS$nmo^+BtNFP%F6?GQ3};klLN>*2 zphyM;b%f2{Y2eTHD27zV>_D}&Yq~*ID%%5g2fXW`WX)}|Z3fRZXngZxev2n;?D{M| zFHhC}X!dk7IJC*D>PQYhaJX<(=AZ>Qc2F#kHWvXnrX+}WD|17aPuUZED{j#Wze_as zU0;)7r}jHcK?GKdO8=MhNUY4Zi0k=fCMME7)F?~E?d$D-@g&-opTIjDW7SFv{9>Zmj! zcDCG*Kz+18I{Fhc zRR&clR-qLh`t#$ zKnk?Q!FOgiI-GyTYzFK%J~rDEUJr63V~pL_q=u<)U9;u};FG#^VZVoi;-JQ3$m=>( z!xYO#qD#aOtZVh_h%AA~;uh`Wg$eU@1Fv^YH#Yw=?O#ogDGtt+_Ce4Qc@RoNp~M%h z1MmQY+32AY#Xwh+0@Bi_0F2m~w@2FzjU0~jukA_j-LN;a=u}m^{&cgQxY76 zzE@-K%=+j}4Eq3wmEzW1keH9y6qGGr0ef~N`|YIhi6+kmNN``5og-cHHdTYL-}9-; zkuyB@_dL#dAA2wdUgc$7sDHgSc!V}PLId9qI!zM!InFLz+nHzKIIpGHREi{FV=ok6 zEt{DYHtg0RL1MiMrj6|-b%D-@jP)$r2qEj&oPNIW;DfKcm_pBm7Zwf*y`!)-%JE$x zE`nr>><*}aR+8f2kGwa8trVw|s{s$fz{Y@PXKiGY`;g~3dL^|L zls%h8RRLJQS1i+^(_>I=ThJ1{-toBx8=*03=@|Ra(OBWOE?3W;qpXQVo@GIEVSB+L zhSW{hijvuTiYs2XVDodD+$}4l7XwobW|a&Mbw-~* zl{``N{J|V}_h#A0(*n#+$W!+8Dx!a7oRC8nPDnY$mQrLp6^|9+Ilg=SfTB&Z*AKPC zcSP&fcpcLgdDIGWd{Nz`fL;{~%&wKbNELu>wyP9JAu)Lwa*hW>yXaQG2QlF8%GYc5 zgu+tf9EN$GfSH&j2!OZo;}1i9%z$Zo{nD>V#w!DuN(;bjqgbfc(o^y0-M2!>KLuP> z-C6&oF?#hqzY&RU5c*}Ij}?k!bhl*5ajP}^peb^zSEm;;k9_a}#90FBd@=7>2Tl#N z(>lq;wI3!9G$x~la$lnOsy&+<7xr*DHa8hTxpXzXSbQk#yaZyed;Nwy6U5kb;-r}$ z^<`Th+6W5k*PMR-`jI>T#~!^s7xrMcSlEsXicP0T3Kd@$HXtt47}O;mx?*vj##Gd4 z+yX_n><7KKBYv;=4x(qn4ri{2;a8dNpYH?W{Spx0tc+HD20a}KGhVhXg-3v2LA z48c{Ug1H2;zD41+z-X2zzOn#YGTHXpP2D;Vb+yNeB8ZDKH}C*7m0gAwrLM)AtMBF`FkPF zP?w8M`KHQq?2km;HARoOwM*J1oiqmdk>ptK+X769$jFfZWLJn$U}1tBc8ePO{qdbW z+Lbd$+?o~HUMW)JyXueCMer#jZrH7TRCzcwkI4x{1Y%ixjR%HsOd?IsTNs^$H4X?` zU49PY48_OSS$7+}RG}7*4ume4@w&M>oE(Hqo_lk&4!1%`)!;1OmFsJ^eeStNaL#q^ z2+OqFy9%e6El^z7uXSXx3zuc)TG*6yicO)&8Y&*SpD?&q4}t=zLAqvA2g}S>ujl

bB5)f|FM$CZ+*RgCu_!p=Vt>~$c9&DRoY{LpdyOhMv;7QcIR%^ zVAdCHJ@%s1Yma$vnqL%RV8OegbRCg;5WX>;rconq$HK}W?%xX6@I?&iJGq?BcDrJ= z9yl3oGwx+f{=OM0sx=|sB98lfh+71R9C!N~G6h_d#!Zzq?-jP}viu zYY-OD`GK7xjL&{X56U}%CJ5{PcM00TU%|q0pm5GqnQpXz22%=ih3x}6ax7;uFo(l1 zt<*zft5jtwqyp_{u=e2qx0y2LJe?kk zpQBe7Y7U2WdR=9TnR3sS6#nyKQ)myup2eKaSDs_DOt-@79{IVM5mxs27k?xPF1$X0 z*5@dnH;rPGDYBZ1Po0s#35CMlL}l8QT?Z zEZyFwHK;xVKJZT1(ircjd({SgLH&XyOGy86cn0tv0^`dOL6yYu>yRy&vFYdTVBGM+ z%%pck-`FHKV@C2o@*YX!w@|ro{0;y$3SoH^n?sRoM3^~bhmjQ1et5*qL_?3eqtSGB z1GAqha)Hp&@!%P2&c_9)u?BMRwbG_G#V^bd343?iG;(VSS!z)*JwUO26uD2u_%j3+)hEKC|6o5*(bng*Z#nq(5jfhsHed_(5+J%i)v>LP(Ht%#yh6762SmD6jB$6 zS6but2i}@nkHrXy5#G#zyU(<&b2hFl4|1fPVTj{jfx={AX# zyD+X9v&km#eaLdkd6@v$ZPp09|AX{-X5e`~JdsKE@v}{?`+r6*SlFgUiakn^1}eT@ zz8ENH(#2WRIz?M$CFHpK{=i{n4R9ttB8R<>2nWc0dS%!?rkctkS<||rKcyclkd?lX zPMVL6{|424Xx=_eQiG0#r9@U}H%Rg6&7eG4N0788kI81P$~s8_o$PaK)*wlV>?h0T z?uot@yi!~ib1baR*9zg5xO(B46Mqmd{OspF@Y5*)=dDA0E?d3J5zb2#=LR+G5ACbEgP@Ze`<=Q&WRb2+TLDD9NS{R}3Je?=ROFeX7)Q~4gHDQJTX%m|S zmnF54-883TZ*!otZ7a}hUN0R4!Bf9Ds)#W| z=Lh>r^GFpx=(uox3WQchxy?r?7RrBXsQBA-K3yP4n+d#T-aUYY1aT3l9oIwxjIcNY z(@Wb_UGmzwqpeux=s1m|3Pv+)~E)&;*V>)mpc@K!Ka7U)o>M0bL& zjm%ViDp{lIj?xEJhS&KHK{>>2MKRL{VoHZhE1Tdd$VWl9byajX zeM5zVRMAz=3EJ?3z^G-y;?VIh;5Hg#8 zn<#c81#=dE;Ek&bugU6s2VooBE9_lxHo6hmK6-$wp*5~lUMgR&ND{7>ZzY$QiL)m> zvE+EraEw?xS6OUb4(_t%grmT_(igexOf2sy(qjQda(&QI;gIT%^3LpSbWUjdjBZhm zuu9$n z6n;j;g}sq73nQYZ*i97Kh}yRmx6Lov=`F z6VeJ;b~7<)0~>2=Mx$tADs2R(4BxlzdwoTy7C1U$8v>sCVg9HUQkL1k0KV5RS(B&H z<%aaoe_kxU2LajEM6E=kI|gcTh<{Yh1q)=!8aHPw}bpF?|jY5tAl0QZyzBGlD+9 z)*-184k^mypMR2(AJQl<=P(RfBGSGySV~uZbxE_eqOHq1KpaxLCR)`3sD3c07QgmH zrdKcZABb5t^N9@NwCkpJc%aT-0lf{_SX;bLv${O@haR=imG4~72mO?B&bnh?<5zS;@lyHBf?m)1 zu#C6$(A;7=?iC`%ea73mee-TJFY1?a45F+6X6xy5(8i6ZF(NyY4@}*svKArb<8!;P z`@|txdV>BWw2|)g*x{Wo?Uz?myQbw!tEo;+)-)_aS^}kj0}%Ns&}@s&m!<~Q1s*5E z!dgi-0HaHO!s*qf<3`nU5aEbhnOkO_x0agZP;|U4tAd)$s=)RBMoNX+$ioX$1wcCZ zKy+n6#@u^~_BTFgh3+t{w$4;#1Yx%Cpca3?BW@ouH=_2xovGS454X-AMWaypNur0l zuq(pkVLg;)7eV`(kyLZ1;6uss$Z|<{bWSK-(g$VvRmD^RA7&w{PE^I@1Z~wEfC@bb z0@sCYWU&>kF0ffK;&uS`SGqG1tLFl_3|+7MIO@eq&ba5r)?YN&b1#@YUcYwNO(LJ{ z+PLUtZ=O0OuZ*aYv@?+4#)^j`de=-H#(kXcAzNd_MzC7H=Ja!3V4Zw*m)~y|njN5T zAN$cRa>#`ppeq&*(1#Rzk|M`Jel&I;b3C9~)Gt3CF#0%CwKq0jdYmK)hagwE-m?X` z6)!Qpbf%_B)GI8DDfe0C{}DYrqZxL`n}QO9w!|$FKNVtS>7kIt+Sa)0dAoy`zj0O8 ztt=Gu#XzMVELkbcMY5REt^23eI6NGXJds`Z{rr*7ev~rqOofeb;<8{bzCKRi@q!ce zo3HIrTN85f@pN2xYgSq`Wdrn zzOiUauda@5kfY~)*V_O~GA_kVwp!!I)HuOn<7?%!uFka2PS5s*xCLsqi3AgN;zg>EYz1!Q#ur6PqXcnKzw#ftUT5ylcw8-Tmt- z#zV9q!;G#pv^?6l`<|=Y?2|kzLEIz3(yNKZkY_UuWwkhzslq6HH+~!#(NP>i6h0*}6Lm5F6JHy7W)V68<;@h9F17600Nd7ZlNF#rJ*T&W*^2QBDDqZE&&h9iBOH28Q^-<>-|w?|Ybi z)Bo08{4H7e%J`;*78YkC#b#1uEfsH~1s^15Ku$z=`zy<#4d4X5$>+m}ssPWV>@l2i_1)uZ{&BJy8cSD19D^b=Uf6Qsum`l~k6MgUDK?2Bt4uv> z89|v-p#uQ;ok6^>V1hjj^EBFm2xrXri^g2kx9hiWn-THeh3xN>BQ6|QyD1v5b@I2S4_K^~%1yoF8jHc=b)K_T5C_3nT!`m&Ev z{VLDhph{ti!jrr|l+?>Be0K-5f~F&ODIN;zp>GM6gy=q!_sRN{n>4zup@}oQMA^(v zK_1fuWCXYA)!w<#k~ZRYH5TODniVC|ByUJo!)Gk@ydxT5Uz9kuo*!Fq^30ku-V4?r z{+x4OW@1u9M|Q7uBKf7NZ{bK6quLenO&xgs!0TUj!a|l95ohdYMXD%#Dut<(=0drG zQ30z>ZcwFqoF+~t&$9v0+oFu)nh5%GzJSw{p zGwfDB<3S_c?B0j8VHZ>3!`#6Kg6em)#_D zkV{obpmOE1q=Gq0is)7b>ROP`yGhYV56N=muoW(u(I;92dz|fF=b?qMl_YDR&JF2# zot(h210sB9o!9gHAo5ks%x}KzmDwiBF&%oTBW~E^Q8YV!Hr7uzDVKzdxOIw_g%yAs ztk>!av^l=rbTYeJVsL&27iMjo7~wVF<9V1DMkfEJ_?MHdE3^3&2Dq@V!cntbHWO-3 z0@g$2$a+D`+^51{mwR5DWl-kH^4KNe_2f6!_gmsF0O>+jSd+-|!NuA_*m~gzPVs5R z2``J)YCQ0H(K`|Qck8Xy%lQDtg)_k%y5^O>PopY8YXXykrY$wplw=6SELG^ejAh;1w8nyyL4vw{xJ~c)X2Sez`APA*$j+z(bbm~rAU_b9pLI#z zs4ZN~pqw?6QlA{ahKHxjx+oG?F2Te%R_bmr?&?c#|y=yl;A(7ohY7l;J)#){B1 z;Ra^0xKws9G&`h7uoyI#bc@A{#gCN*^j%>=Wkf1htny?u}rIlmLJ zONE^x)7A8tJlSGRPe`#KOKV`!9@WzWWGR#;;MC!>(66Lu0v^fhvf}|M!hP~G^V8R5 z=ue-Zj|veuwUGwZRr!90?ZF&jqMTu`xL3fx`@vnqUl&JWrSwq94b=+a&F^M~^aE$% zV;OdN9GX)*X8?kF`+|>x{!WeXEJ+iVN7n|m#SeYF%Kpzzx7PWXAye~=i1V4Q1W8a>Tv(xs#X@j!7Zs0v zwRfWmAgj6AeP!71mN7}fYRaJ8AzLpvrd(Ad*aqM!J3#OxY1D^IrpENdD(?Z+W51)2zADrnliCLq>)CQ3aIA;6>s%A^FPC5TqIM|LsX;uSC0b-|xv&H{!BEVW0^b}ravpyKa( z;~c{l8#XSoV?rlKdbwu8h8gC1)v4EctsavH=cx8tmsj$!5-uF9;V8yzRS(LsAm7Z| zjN9KMw82NOo>=lr#DEPDGH*tlikN2xUo>+2A7`13#2Yt?lSs)cvvoOPVNwoI>^=(m zJK_rk#R1)tUtd){R3rxd#`Krz8U|AkELEY{fY=N?%}GLRK|I0ikY$G?K-;0|X}!8z za>;k2ydb1C4krSP>D@WIn`+xH5pAOvi%|k3#}}1D7|iA6_^uG&1yTADw_@g^yo;`iX<$+!54?UW{H_p+ zp)fJG#_Jdelq2B_yn}C;E+n&qK--kyZ7%rjS9VQDxAv|OUxxZ$ECVi)BBM-pNG`o3 zWPe0TBvwSg7V2Q=Sy?xYD;PF0e)*t5bttThE(yX6_O+mWP|iC@lG&vZ`k=$1Hzc=I zy0unhRUZ(U(t8+z~fMI@tW=V6!=Tf9KIuvX$Qs$%VJg4Hi43 zDvI4jksVY#YB}O>9(o71QZ;H|!4zH#(wPhf@>ChE39fQj%Za@QU|dq9Xrrp48v;$Fo$0k#CPGEo>?M2co>Y|OL2z$U^Kq{Rc)n&8=d;Q?2FYJP9?waYjQov6Y zXbI;^N8E~Ny%xE3YQZIO+K$;atW1m*7!5#PSef*#?>21oH)Ex1LDV&}iQlT@!aJyH z3uKj2>~@M2fJ6WgdEc7b1luF@hf$3_fBIP&R{835S!|6E`4S#@jJS=sRm<+GkU3$< zbI7w?dEWc;H}PJPbf+NGX{wI)Vv8CL2nQa^e|X}F(2SPwtbaH`mcKG+*=B*3^%T2~ zB5727LPR!$3KBY#e*q=mSl)QiCq1TKe$i(atXfVpFk{UdA7G62zO%0Ve0Is>zpl(M z8jND7YCGnC%)d#wE(Cf++C7TsM#jO8r!6yPD=KUm%aPB_{PerO`?c9z7 zQ{#~$-4nVWTBvvURB3gEK2KGb2D5a-x6%mE*3W_PbjM}1_+$pbzz&;K<=ULn!V0%^;{4=f%l7Nv;`jy z>l8J)8?`{26pO{Hy&pz3DHCArtAr3zuPjs3DLMzj6~k_i{4=z}Zf7DYp&L6Vwh$Vk z!rN3>Dbvn0Fk6CLHX`%MTi;5OnE|N$%hy^+stbF zjA@~S`Njw^G_FZ*;^iUF6qpF~T+pilQ&czC>C2L>M#tI%cBFwBYb;*=@8qk$J}~{o z#r-bq<#H7F=LdF*uona)v>!@Pkzz+7gN?4 z8+|?&M)+Ri##GbOJ3s#$vxQlY6Hx@9v9HQS(zw%juT zwAM!4>gDy?CV4*TQ?)26eA}7f;B=opm9Cgc^2_$wqTNI8Xmo{|TDn#J8T48gYKCF+ zu3I5KPk-!NNpxr66PFNZKY9X8A<01-XYlo;oPnOvNE&s_3qaiH87q2|k*x2qruE=c zKIOti{v3oKr@#nZB74Z{SqHT#!evt|&zDchR3%O=VpeKWm`&{JfLW>aJX75_Qw z5Hb$w66Torg-n#VPlYK$tiY>a+E`Sbxw3$-@fi1%@W;ovKZ@6jUW4E1uqO6+MeGO% z4aj+VkQ^or!mLp20J)~BWU@kwpc|xi;eOe$TUO{cXd+uo?U(JB-I|*lawV)aZjjWA z`ePcI&gd%xT}t6D4 z6#EfHnlNRM?o}I<4r+>6H*lKt(2c%1z8xOAb74cOq%hbSL0wFf`&OtiKsIgIc|$w0 zwvLz*jYnRq@;w}8A~j1C1Mde=lO<~kLrkPokfJ=LP1X!6H;XRJYm+C4Q>5L>s+c5* z_+0egHLXCKq0Q6i)wq=&aq9r-$2K|s$7BNQf2PVbbOrpU@Tn`e4=ST!W6QDariGU` zGpTv)q;~84azrc-?TkgA-SeNs)*F6CW8|jtq8BmLOHFy)W@4!|ov=E6K<;s+^^GgU}Dm*jEDXX|TF422hVg<>hx<{1-V*i7?lGd4^X&iSA)e*!l) zzMbXPYmE(;bv`+8U?L}|r&EJkfOW8PVKGyHzC%!I5HtqS8E};^BeuO0!wU!Yei)DU zxp6RtL74R2x7uT@c?n$>8yrl7CYgp@I>|#<=iBRB=2r$SV^CCGOZSqbsfR_KqV$+P z<)xtA9!ov+>LHII58Y9+c?PhmXM43v+9kDuB#%P6jtt7{0?(0EUU@MOJe*)pcmc-B zJ^J#L2NG<2lpkQiqyG1|U-lNF6ZDX(Dh4+mI&3n|R2_@i1pAUxvqvC!Wr~9pXbXHZ zRr@_#fqKW(wbKacDsuj z_=e=ckc~j{7E~lcL97Ub5wStrGy8JpOwYfaIa_Bb=bQ1&^i02*)Sho?HX=J-y%giK( zFZkb?-URh(8%d|^OVaO*wg33~&h*REx~4S)i9#ERoPIB)8kBZ3l}Xc%kUZs0kIU0a z=&ot!hz(XPqX}X;^R`^IjDIgSCV=|K`BUC^GwhXG-Z=jeNqEWjN~K17rEH3UEwzS< zK&tK*$v)o_E>t&R&?lGf3Eslh25oTC@`fB>t>Gbks zgc_=)9g@9D8#i3A{equbb=!)VvT_XVH=UX!`prQvgBd9tdar_{zhtXTwUM#dMKMsd zU4Z;?1yBL0PT_S(+Grgm$3;>daDx-$R_0b9XbE^EhLU+at0(NhZ4?IVRx(9FMP7T~ zKptD|47I+gY*T>7D;YNJg(2rTRyQo;!M823ZXVz7F3Mjw%K(|af03>xnXk;sQ)2|C zQi>_2NFfxmfVS_)N%OnqZTu7;ZMarcMn`elgAa>xXjE%0b#D*Oac@@D@wB{XX3g|^ z0M9ON)uOxMOXI1fab(cHiW==uQD_OZ|QSrAwNwQ9v_S410T5z(q z@%-h-50$2Et#-TnG1I7eefDy}-Zylr4!BBo^U+E`lqSKeU6Bp(lQjP*Cv~wXTGT`K zE1@0|7t^cY$YKs4Upy3~PEi(!+7$_2a9N2{i7lddEa(PD8V{4#PvK;>SLV&hXmQ`^ zKbbVslq`zPCdz)-TVjSr3bB$N__CL-{MjP8;8JKC?HT>tkI$f34L{EeH0__hwcIq({E9%tObWU;q-uelgmBm+KS*l) z&?~pPQ}WEWZF;~$RG8jo^Y=H^I||0LE4JUl3UF#irl6nfiW5XtQ2)9POK4W}Rzf%) zixVIfEj-}2N32nl(b|9^2MohcNDrc6f-Qu{(`Cozz|em{^p{%#O^aXc)(bIH=$axa zE=*qW#Vmsg}%lFP3?$}VdZN~mImB7X6kFaR255RJPK^i!Bk9( z1Zw3QJyxr9J){`qK^eskIx(;v8n-Z2*UV(O*Dg$*rVTpiHvt5utc}C4nHVv}ws>hC zHoNw&`GILO1{+_)j#m{ktsA-Yo|*lm!lRpRa8zSiQxEiS49em|)hol|9Q8$;MQ#@s zo+GJ{-0Psr9kH|G0HlU=bx(DRw?nULuN>`E3{=46U~BWBY>#6tXHBq1iT&{~(AQ!V z{zXd`0~}FJobAx;9L)fmE>cAqM#GvrR004}xdWtu-w(0g6|=vD109kaXmhsE>Wvw4 z!_Z^>(=vcATIcs~-x|*KeD7v1DYxS;pEE`dNFBuhi(oYskwO7Cu+zetkB_*-ZhPk;iu}yD6|+ft8ynQ9HhtrDgt9;xX7tjd+iU-RxSfs z$s2SAox_cLeaU{Ga|CGvfxk%voJ>f?lPE{p;$x0W(&&yLZ1@^-YMqVh$%*pvSp%*H zl4JBou%r=nKk8%&UQp|%zrCKu%s6sVmW8AT>gb@fL6-v9v=tjYyOx>@*8b@54Mv}4 z`5SgPn&ba})&&`GBy|ZqkbZ)%wSFz~n1I5NT>6Iek-S-TLz*zN zFl0w)4SzpyIvu4e#E1P>Oy8?)R;}QDIjsjg;9Pn;r`N5BYZb#~3kX>b6!TZIf)KUx z{P(t*CYhg&%GhxclbKX<(zH76byccYzi>-H4Cgw)XrJp9-YSnGAjr5XhU1}G;_}c+ z(Vg(mzu(I(wV{ z9UvKxX^nU0ft0%^BzZh?%yyhMG$Y5{DE(51S?IzLUEC#&Ug><3ukMl#$n{lSYo+G` z3PgA5E)e2YpHvLUvy};6c}O+4`Sqi|ZFGwynZQz(;I)xAARkUJ#rWaLvJfoNjN%=) zo}8J55#!;t>V-e8fhOG;QJq=K{>iXl2?U#dO3qIq_l?$tPKtq0St}KhK{oR8=>1L| ze06$YhXnabbc7Qc1$rH;1IpmUB+xQTM0mmjLp!0{(Zf5(@tn<{raO$3W z!D#bp{;3}&B4WP}@-@Ybcs~QuO1nyylRP_ijSd?<0+kd~PLUGOfTw%s0n{KK2c2sn z>dm4Ceisni9;F9?1a||kTyaL2#oebo=!^UBJ5o$Jtd|xq0C+=K*Rz2S9g;quD`Ms> zpSt9hd#?;PZCWSX!s)nD4+Py3Ht=IOWA>q+L65Z|{mci9@c=tWCx5*DGt)qh-R{`U zgl0PE$gl#zuGxuGm*8T9?IXC$8gfA07)U+o{@B2k8FhkK!(!A?tbdRlEN<=exc4eL z%Ehu;Nw<5x`+-Fb;R9iV4y$|)2PXayUmXlg4ZQi&mApX*Y&Kf!lf|qIEYw}ITh>h< z4zCAw)-MU(znmHH?4B9v|F!rCX$;ajeJyUaMR15UkVaknCK5=$$o_7`c*%ymmRYZSCFc^b0qziV(D%i?A+ZbAPSq&-=;b69I-1d+j)n3E z?aW=SI`Fv3>M2dCj4Av6uINWui)!9E`m>=WXI0HA=sJQ0ZUqNIJ^_1X z?(s|w9CT=eRMZ9@zB=uZ8kpvTN&#*B0q1HOsdmd``^MEs%?5d9mnd^5o{O#LpIq16 z{X0|Pc{b&FcI?fXk)%)dYvAW{yWJ!GQ@t+w)hc2{SWq?i{T1(M6x&=omDrWmMc;B> zYcoN-XJ>>ZI-VWCMn|S~e&heLi*FXo;vkQW1(fKgVofAaXrk}B)p4t<#_gsIi{aEw ziX7mH(hGJ&_&HmNneGFm2buzU#HT?Php`be#JZIHxnuWYR}%Ozhi=%>Ewb=6DvLak$kYLTSMZ_cqQe#VQLksz`r ze%N8=gxN<2V%VoJBU))7Qi)v43 zhonhSDNVIWqv~@e%yL{jXAm2lVug!IKmI6qr)g2H-3C$2R0}V2tAr@oQQy{i!0i-x zn7fLp5_P+4!CQ^toFn!0#IX1vLJZBthpJIg#>dHljJ5qzqIR3*Q zT_zjt*kT3ABk&k0v&upAg;1KwEgVh0TF-Q$Z*cQnt|#@TNT z1gSIvOFqTqQX~t#klh4`Bbzv&FGmKr<+C2hbYvSy%v0}PJ`2P4pkCNIs~o)80>O1| zHQf@Rj&g+FBPf+YBa}xk7ZeDrAD|zeSIhBZVpWFU4i9AIpU=V1M|Ev83@eQ?dhG$S zZ3@|E^!4wg7|5CBQ4!U`v%mxPK(Yezrze=@F3l<|uP^o2I`xt$?^@5YkVJVZ=y9E8 z`oJkz;eQMM1#%PUwutpY-o99`(}FywT277fOQ;61>P}(A*Yy%k!L!e%{YIR4pZ)C5 z9~$to=Y1C`xy>#vYsXe+t&u|bLyGCApvVtY%2IsN=xy!?eUWGdTL6$V0XrGh$aRNR zbzqMKo%0%{SfzsdXPrO{T$m=>#eqiZsp>RfgwjHx6nM=phfqtb1u;G;qQKfQy^F=)wo1DMNxf z(amrtq?@t9x;I3lNc9JOc)&#EbfEjt^K-|Cq(N*)-8Z+*1%V1-z^NK{ECwCB`5m`` z9bW53|2e{xvwQiRHSSphJIQ_Ot%rX!ASr5X%mUK&%Gj~>Mt1Bm#SBp70Tpp?Mw#cZ zldTU2a~fU2kVy4J@0zeu{{+?HY3k_LcEhr*->F?B$XW&*pDhx!ZD1nW_?xD7@&Skd z%|3F;X}4!}Kt8kt4>%7wAnj**V4}Q7+DC741wXA3lAs;X&yBpG`c9}-e%OYmb;3aq zvVHIYI0$~6)+xt56B4Ot9tWHYz(}FPrk9Atcf;eUk;@)GYlG|4+sYmHLb3Bkf5@CE z?}Alpa6Ih+5(#!|$$qv+{Kd!zfYS52-TXHmb-Y%2fdHA^kGmar8@ruh z_w?h{tSl4z;m3RjE6X%_*OA5lYPx%SMY5G<_H}9Wedt#DMmQ+DtX%0G14IvJ=zLDO zdxbx6W|9sG9?J?`CRpVbD=(Cso3_F23|$$PHtUG*TGwV(hvWu1J9G1_4{QiER>*p0 zT^loSUU*}A;^1}*cW(s(B%cJyjYcVUdFV4c=sUF0=*G73p&=v$oerSRh6WQb$-;gz ziUp^{_{7`j%O2ksjo)9D}Lwdw9 z)k(cYa(mu@OQT{P(Bj2#u$#KZuk(#uIu~Ml)dA_izlbrvAz#d=N9(0z@uOO$r?El% zwBI+tpQoMig_o|5Cr2467;ru+!z=XaaO{FSR)yQM05m?r0(SYK?S=ng0M%!)QE!kN zb{z6sZ4`Rxrx<7w?}3I`e*UZ$?`^K>PED@8+^=Tu2}|`o@2i{e&}L5T?9=>~kbVNx z|4_Dx|LqC88H(AMRR(1O0l{NHw zr*?TX6X~DD-RSx7%_bH8JM<e#x%c=OMsY9QL^p3&4be z$rxJCz8e8H^R{agH<${_a9m^Bu=0vHM`qevu7Hq5QTR1Tz@KGOl-2xl_gIe!c8K(| zWeI>r%PoI-LtR%uV3N zdfl9Zc?T31g(Q`9lh4gLiE`vS!UX$6`lxJ~+mWDV8QQcMP97u^(gQCEqXRqW8oEr_ z8*+1Qd}uuPvaDT^K@U3QaF>%S;yj@J{9ylr3a1Ym1GF=S9G-qw8$X?Q)T0d&SsJCf z0~8^PxyQ&o$XS)h^fa$XUac=ct%hCR7D;PBlVYP2{+jDK_>0ykAA}-*`89WpTXaC; zA5?OkM`Z`F1Xa)44=W!T1*_)G(y^0*E2aS+yRGptQ#a7*-zzJ1-@u>*&SBYZaVyE2 zUCTkjqBj0@rfNYi7a2-YBxSPQGrBNK51#>%FjinpyifuF0t)xb_N4&>>9|%nE0XJH`5gK?q zB#BdPWl=^<+St%AVgMVxG>?W+jLT&Iz#kR8>>6EIcC1lg76YF9CQfc(7k!-@83y$8 z+8M1-aDf>TJ(oB*BsyiTO*79To9A-DLv9bLp^qqrp88_I5f&0+?}hL({AgT;rj zxRe!ICN27Q?QT=z88f7ZIm%REVQ7_O#H2uQ*EvyM>zNG7Me8JLq*Ft~uSS!d=7YgD zn79_m0+z{_rbwdTjz~e75DBBO3>4LYdlsIU1ysM>yZ(`ay{|2`?&CfFxp^{AEWY^{ zU&?B3>^IOBd2POjso09$`nzU?Rfelx0H?{M%o7v=8~J6P^>nGvCZ9|U)-{9qKyiZ{ z!3HiG7OctIHq*K|9St#K#O`QAS@jqz>`wa5CwYraD`(h5W$YLH#09tL-$UEa&huns-6`B`E$?N3%Yd|`e}Yn)+c zKHh)*-CHkvMCB&QURbWEqV{;kdgTPJ03RrWtKn|pRQbk0Z%m=2O1xrzld40u*Q<{E zn0Jtr240_kRs|e+%6#rxj{Z!eqL|mHz~ibL zQVg)x$CXq)?e-;_wlEF`EB{5S-q0y{p zC?=I6Tc`*uqG(q1OS(a^_ZUbCsgY4`2dKSZK%z6KT#Ug8UHGY%Q_aouoFMOREc0NM zFZu<8u$%{bjb#^qYwM%&G@aS+j?5gH2#69xg<}Eaov?GikkciN66|~(6m>xQJca{p z)U$H~bLjXGl>Zx`vlrGv!)p!i1P{9TvjVHRHKE!WB^+p}vVqMpKyECMA7Of40Nk;k zUijaIn>hw!@}I~0j*=C2y!K=mS-DLV6GxF~D&m6d3X|*K#_u56{#TeaKIdJ&a7|U@ z`uE6eH=pLdaSP|!-0e0y&dRh=zxeZa7QgJ7JM0L`;9l^x&pj(>*vwWu zJ1{F~O!~_gUpL7N&`6v9`~M_~b_|U&BWUDM45(S9Q4#4=pZ-;!bu17Syx!>Au1MkN z-3KIZ(TnbN&fDYH?R3z$#HVEX1MtK3e{H<*u)@g%8WL7GnLPRXH_A+fiR{&&5CiP9nF&9hA<{exO3$7Opm^p1vWD6r}rAb54*lsG9D8ZbTac!(hq5ypi91 z_(%1h-~Yp}{`4Q>)fBUWB9TTwFn#mvLvQ^f?BJ1-_RGhw8O(_HAG9BkEw4-mVug{> z*-kM~ERsn@tbqc?TDM}?8eZOPb%n=qw@N8i)_2N){-(mCL0ZF8r}?zYu`#ilyCx9I zVblD(ghS%5gC04nkwgb~I#I5th^RKx*ckq|3Ju}%vZFg&{H0#T!DG}EL2$<}OTWN@SO&OGdX(I_yKt&S1p z3vRo%%a4i=^K2f?9{cxeqc7GlEV~ESZW%M)zBKTOsjB8HlAJN4!HMizrM}ky&gY2w z8_B9LbX>re9N?YyusNDdF=;uvP4Ber`ycaG{q9c&>mic-b13O%x6s&eye!73hdFXG z^Ozz7R0Qffbj&!!9oCJ8&OZRkr2wpmujID5995+A9{H6+8$4>~odzPOgMR6O>73iJ zr*3owcFRmUmV1B{&v@jgMpB^$X&1eMWP*I!4nMp`cV5qOdEm8Z7bJW;si=|k(0Fkp zv`v?U;|RIR3@&~Qen^(M{*Cg`O3@DJl-HBSt(%Yc9rnXt=D2eG>|&zj?GTi@p8?$? zHQrW7)(B9#@3gXCh)d@15S0tCcqfYL3GGx?E{G4&&d3rU3BopKyJe6oyS#YWBm)GM zS(TrV#}0ODMG52MdvpkBJ0?hBs(rJ*=~fM6pGnQkpwCNsS4WVe^akmw9S1vuRS;d)LQ^0qg~W91K8+5$lIT6F|B%imd?kHAz-Gr zIE8amRxQk-qk(v|S%rBVwApR^ruUowA%^;kMsY*hF2{87Npi%ohK>`hamAJ6h+`k8 zL(<6Um~o$bRCbcY0cT~T;E1D*aKaW9Mi9}xu?;ND8<@YHxb|b(U`+ac=5>NBx8rzR zx{-OV{x-8Hz+@-#EX@Jrn8-G2Ud}YV>Wu4KE=n=(y zLy{%`H0cWU@8gN0TlUmSp#fqq;z*3P$ z(d=7WydDv1yQt2%t>}16<#X<3hn7pf|M)di9x%H_iy3CGH2QNz8y|%}3j{jF5gU>? zKj(|Lu$1AsBiQC7D@aWK#ZTT^=xICHhMKL7if3jdgxN_kX=}4J$HND=ET@Jvl@_CVTY5N#Xas?CkwimW5<1zwv0t#9h0* z0~#%3y(*kGi>d?Sc+reyP_;4)jX^p3JBy(D;LFEBrn#zTgFJSedN7lbIt1+idi5gg z5lZt1Zl@W#sw=FvvTVvM`u;4s$l}SjXf*3xWZnjOz4ebPmm2KNg@0TXOzyF>H}<=W zBpZ!(W-BK%%O;aSD&liR953nzSHvkkcY*z@h|cwj6r7dT&_|S)MM=}ng5pu6;J&O^ zeo)-NPnuT8-47eJF0dedph&dM|FZXG_jHgR=>zH7HsD$4R3iI#JnypipuCO08Mb&D z1&)$KZx%%gA_aN$2H@hY;g`B^7J(p}Xsy$E&b?r?cxR>QppJMiBtv}DBi%n20@Ee3 zO##iS<3JMK;9Sew6?)mb+HcM+{&^^;jOQ!EZqD86wS^gUs0uxa)F9K_q2guz^$U|V1IPPGs^`7H<$!Vly|%LOYQ~di<>x`0(zVW0X64Gb=5_C_yr`_Y;4>>DsRz(U<(B-qb`3duLxvPWq zdYY}m#(*3;)3tOyxUc;CGW{nX;52~y3&JM6>+|tQkMDUfVu$_kbFK#s$U7u)ytC3G zuMNC%%TjU{L+k~ngUxA6@Wu{elJ6Y-m1(mRo5E2$-lm&rjlx}02VJ!w$*DsUAJRat z^j?BtnJn&c2}WaXyCO6H#B?+0r;{C!EP;S^uJQeci(ku!-J5v7FG-ReTc2_x>yt|{ zuuP?+@)7DpqEIiy@bm)d_;zm)rY*CTc)FTKu7(0&Y}Y9EE6Sm-y2EE_kNYmC0zo2g zo6TtFbaTXR>WwXhbtZZAQCQiKN!$4UoHf}1l$C4z4}%^jy8-OCAat7%tYRr<14Y&X ztbjK!kG>?tL={GhJ0?87r=K?q@H6@Z+xX~Dw4Cqz7!Xmvcw-gGe#!J54jRpSImMJv zq=JusIP2;JuN}T6Q01K#HiRs5X~fb z6^G^eBspCY$8$kc#_QgUW~Rjblo;FQ)Cd)1mh#qEi=*8R#ZO z9^C-ijoG65xmLi6@eGajZY*dTZ5(SK^XU9z;F|{X@-H{fq>{aMY+jm;%*%0#fx5jq zDk7H~%T4EPU)%)cdhzmRRXu%K{H3r_*uZZjHz4Ah2{GR`eh+U2Q_HCZTGfln20re7 z4w71k%a*#Ipu4AUap|HnRkxU2E|xW15#M)S2Bmy>XB~G(XsIl12Hpi_e+%oB_^N@w z%;~E5kw+1?jgRklEj&FQs^+o4V{`*E+R?FthLy3QnsV+Ek?T}Mx_e1zAE^ef75fI%(1$H;QbaQ-w1q67h_VLR*HAPAOI! zfmm)}9*CzOfHCrwb>emf>MjAl2|eI~#qwq1I(O9{6Tvko@e;0MC@e{6!b-64VR^sXJ z8J`utt53`7~8An_Hs|y zj6Tm~iY-Oc(^bP{02{sC{D+iIo?&k^JNX}ulGS#+(atvVGm=+n(jDS%< zF?kfprXsXXIn?I2n#1+_M2ACS{9w0>d|~O1Hwb!)`qRg}zZ4cY4LLMJ^bNa4_P?uj zO8sHG9It%Li?JycBnz-SH4RT+jXu*>ILZzzzfp%Tdf63^^w|JpA*zuyfQCxBdkwG9 zJJtu95DNskjt6E$IW1{(YIB-6W5N;}R+x3mhG2({;7M++K?Xx2bPUfTyVynF?0DaM z%E;{0P>hx$d#Q+wnS;=yw#&B_h&V8swv2iZyi?KTLYlye)3cs&vi= zyB#Yf9}GHt&^0fK^8wae7t&aN08v?Pzhs<~XzojV<~Zxwer5#Q9AIT$sP&IRU3`Zf z*tEaJlUzIYQ|gQiMg_%yBEc>y0$a@rLpngGYt8f?sBCJG3^}AhNCaP{`F96zW}uTk zj+aL7bSe;`zLBoAexL73X?*A~K~U$s8=OYRTp-lbxYYPTlG^hN;Jj{;Y;}vCVl7|Jww7Y|W}z=o~vW=d$E3h_}{5 z8C2x-n{ytJ%CG|ql0^5o8Ok~y@K#ADz{HFVB%@8Mb&s(E3H76W&ef(tH8w$eJ1&DV zBe11+gY|^?xOY}3&+3q*(M=OC?`nCii**t5;?d&r-ibApgtpu6M39>p!Qi}^ViG74 zPemN`tro_5wao)@JlRS{jU6AlI=2$ey7`!T-5X@(RpSL;D=T~Pg0ZZ5nw7nn{Nq1* z{JV?68~ODm@&Sowx7o4daBHysSQPL?Zce&@hD{w6wN%cY+lnz-bXV=0U zI%(Q=mzc0_`d&yPw{OmN=<)1!FPHU;`<<6@3PT3ydmcB*P1q#vbShINat9qwi^}L_ zoUPOHK-T%Xb39Y%-6zB@()DDMb!KJEU;W5wFnK-7H_7@CtB2Ws0}u0scP4#oK+w;w zohGE$j`uwAMpQp5CNq(f$zv+wqF+-`{rnRQ21`=Cwt8*z#Un?^Wm^BLL7L~Jbvi|s zg=9fVZWgzJzn;``@EYyR9ic4&bCTzRed2 zjcx_w`843Px5eXO^0bWQ?Id=Ohvb~6z}0}lcWSqtC!0oVX|vy_h(HJ12+t^sVt|i) z8x^r?%6d+paP^e!^r|V>L0kZ#jlCM@zw1G-CWDr%tBs4{zo?vkq0 z9jjgXLUhd|=q!H?GDo&{g8^M*L+&_XX~FHRkUP2bW6oj5u--U7?7KiVv)fzQFZ9W7 zBRJ(z3{*N~P!V4eBp6H#>=U1YQtmR}3g6Yv8(r$@?Li0mr#)gg@vswE=Uu|R5B|zZ zUZVV}`xWtF?oM)s&Y-W!lOH3#F~k(a6S!+r-!(CbT7QO`9d>Rb1?P?-LgAbEi;qe75--C z%U%(8g2Zr6O>0!BQE6?!Svv!Bf?B6zq!ENgP*T|DSTLFy7QisN&(?2P0b{cGtH-tf zWbi)(f=xdq=h^umcAN{kZ{&Y;QVj6Hw^9+k@@NKSZgRLfYL_-TlXFIx;||?Ds8-T0 z&(?Ls4*>m3yCQ)L>rET~1lbvs&fBZhU4`5o`jWaF_Z4An{QFMogTAf8>+%91VnA0| zHx3?Xk(7HULzCG>zs;{Vf}0$;R*t=~I9fl1eK3r&S6R(3l4Wu(3e{J|JwTm&9cpAW zUeSS(!JC6YA;kq_G;J=8T!=o$@vxjCZ~989)(KkA0t=~8S3nlH97{aKvU^4W`7i#I z%%KcOd+hS~6!h4QvX!%AJZ&|?Qw+tdr$`hw+;`B&mC26lIuSd(9|08< zcs4bXLD><{mC(c5MPuX*I!N_u#b79qa7>rDED${{`W~RO7uEuW!{Ly9N~ADMoQo8O zY83uLgQ7m?nj$IWf;RRs(j!){aYcL6LFdrv5Na-6ls-e<4Ju%XQ@LHbrhw5jB3DZ^ZE`y)$3S za?ywHv88CAuLhR2dm$aVPC;b)Kg4ZP?GQ(~=IBJ)zF4evs(1_K+%PGCqSQsQGp?xR z23*d**|R!jnUZGJZQc>r%EhINZ-fszBvDJ-{~qZ67vBy4xa6G;iw7MxQn$ln7Bzow z&|%NwFMc@au>X7M)X{hFc-PMc9Ud)C`qlZxjc_Nvy7>e6xoFVg?oVSEeZ21{TBpIc znpI7le8{XOaC*7eH=W7R2H`JMGpZkNiAl3;IU2`RqgXlfCT)}jjrz81!23f3;P$-l zA|QQnVG!7mgYeun(N2E*kA-woU}2!U;KSF1yD94@Ejh10}=VGTaBmgdL>|Ih#cV zqIAwB@z!ZrP>ix0-Sj~afr|EfKrpv;Rfsuf{m5&1C9=~Hm%15vRs7fk$=XW7i|908 zkc{GMXFlCm*aae=g82zH%_H)W@U@4Fhs_-~<6lViE3?R6G%}8-DCPu3jv^Oq2Mwzs zE{cP)CPljfJ8^;QYKL#3M5B085`>WkTzV@!vgu`l4mvWdKosqUrKYeGND$>K+UWHz z(Qfbt4Hc$osspmPhh-@eSQAh%D#IwSqYC(~3NsK6l z+z8UkZnE_XCk|>)P-M}IcV|=oF~tBVRmVqavdoUx&om>5#8XTRMb=XhUxi%~SDMS+6*+x4#@%X^f|Zao3O@3md&y?K8f8J<_*dyOWfqyTf8=V88A^e^j-x?XJ=&W zIEhhhgrHp%vy&nPU}XF+`Sy@zaMQ8-rP=EX&qR4uXuskMuQp(DKn;w|qEdJLHM_m* z!Qbwfo8^L66wPQ-UHZj0i?1wB{&x90$!{mWef$0PUsr^mwG9QeWdM7+q>T289ds^8UV5fBer;_DduyE9HSzTd>mOIV50|VQ{BR&-_^B} z0#{ulg8~;99qArW=s{Ma12$3WR7EKW$qhMl(0y-o%xIC6ffNGnw(f-tIiSdLqP)bt z+%w8acOS;Wk=Q&@zIsuzUo3aXVY|y2`iNtPM8nzbn?EZzFx}%M(bM{ATn1dmQuAsJ z+>C}LlVg^FWyDqH4b!~Ol7!VpEjen1h;6Bw2s==K0`y80r zD#!2u(%hecdaen`lo)1-HF8GGt}S0K)ZX1^${_r#yq6s>MP_({S53LTxOMTde{Wmd z_=9ynT0Ny6_Ep12R!`9=YD4=%j(EoK$~~J^yGgHnLs;v)?ZDc%DWKf5kCeIR(B&ZB z)WpHM!-NGJJdVnsXkOVF)I)Cw>w=xt4Jn8`x%KMu@kx>b6E;eC=TRW=OQTPKIP+-;S>gCcWy{=(Ven`s&G(YWx$-+?X}+}( z6-bn$zEF-=>HIeS(bqJJGc!u(;}xr>BzBa==jiTI);}ThWXJ( zMUfZK-D_v)6_)|&SH*=uL)`+2+K152WlNx2j_Oe!eCvVunUTz+x^Brzo@omon-VcQ zc7Dt>{G~H1X157fjt%_1KoxL%UJt!} z(G~IEuE|Tn??dcf3;(w9Yk7^Xk%9){H*=0lOGI&;6F``TWF@I#iSiHbg=kz(ki(+u zsv@@w3$viLCD!X34>Uxej4SWpV|$EsfWrzAqZpFW4p{>z=B>ymYa{jh6Sv-c*@?42 zaz~2EfqYIh(<)!a+v|h;&%j>4-w%1!a%u3QbOD=TVS^=+)Y8xLQZA9OKS+f#YEchgLm#_lMWWU*$ru@)gY(C51dy^cU0{q=0pXp|bIZKf< zSP@Yju)*VzEAZQ44FR?fL<%6}CB+)3gTCpULJn3^sqsa6AOItE|cNay&(xoU2WDQ*J?hJe`68F zZi27yo*{=ENE88=_%v-$qf}o}sC7yT(FR32MNyV6B37Eg(VqIzk6Ad8tY*-DUCDoP z|IVqGJ^KX2IKT>ziNbD2RH8&ok0nHF;mkkhGK;|aocnEciXB)t?)nd3zQNoqI=}f4 zS@+7sWVRbw%@m5+OpydCBGvb?Ul~;JB+AqM(nPDBvpsLlt>e`~fNy{B_4#=a1kfly z&E*!0EVWwWpGb?K^wffl{{t-?;eTi$dUaGyb8AcoI!Aqml{L}owYZ>&5oTl!a z7sElRoyv(ILNfaHEuMR$yKMi)yczl4p9^YC71!BplbVzs&(pXunY7m(C<`fqF}^*9Vp13PalX4YK_UYQX)C^gqmj^dq+5fG-Bg zCOpw4j|bX;eZD)%c3oMPM$xReBwQ;!1r0k#2y{I|sg_1@jA$SQ*&titqvdVzK-qSk z{v~h>0D_xIFGy<~Qsyn_Z7*7r}Q#HW#_CgD0G5DSuAFI525qy_t{`Y}- zhE*xUD?5o4+3}7Wa@Qkv+|?Abk0N`hh+V#`gFAw>GwSK}(=G%h%wHYcq{^7N-(#g5 zU!PSCa9e~!-p#6qWFuJ-d?RRCK&?kVX;$s?-OQX84LYRFI_5v<5GhFH;=2Z6lJ}rP ze@KegphFpnoj&NW#-l@q1%b)RI9`KnEs#i9<&X6tVU=HtB@@N1#lLbk?U%6I)^Rhv z5e?GE-aYiL-{_J?#SoUU(hRel4+{XnazyJ~ZhnCpEe7*ru}bx_?|n30$sQLNlpM{Q zo#3@xps$Vme}M`MFkxegi_tN}3KNqS#lH9Mdxll!!atQgBu#d_%JdkmGVK)8N|DP{ zMDonzbQP6A(xITKciw<=@~j8Y>3JQx67rOTkXGIyxHYX^wl=JaDyDl$2CO6+YH1?1 zbPwmAq%E*r)}+$E&!G~2+W6zMDri{X9H;Z=q$*(4H``?&m(N)tZ&I!L>CNy%-#h$o zW$zyU=Zf%;>;6yrBE0spAOq;al4cePTBQxbn}R`yt-eXzTyDPS3PB7ff6h+VCRN_A z)Y`w#`qBNjq8DZT;H(Or8!eL|YL`_}$;u2-!*k4)!?t52l|QioO9APehyu=V1r+~LY+oG?pGivv^f4}_6vXRiE*Ihx#N!aBrB9mDm@+I=wL8Lh2Q+{GZOid)pP>+ z;|Mo6j$)!IvW|+#o4vsadOO?YWy15~0#OST)=gYF-N-4l2qq(Mn4s6@VKViXZ`7L> zWxiD57c-@o`}y0TISA{!Q^U$aN^mbsL!AOtTLxv%!h+u@II-FpJ1DW2bLqr*I2Pf< zc%7IKgqX+2qRT(@HLX}Mlj_HSc_*!Ba?aCLJHuCwA&f}yB5z1<%Pj~NSo8mu&6jP> z&wbHY=D+MwHWl`p>6!b}i$T*3i^-37{N{IL>u8IK9apSC(ZYzuq>y6rDUwS?^oNvz z*Li*ZCh%jSM1jV*$d|%;QDWd}(azaOvw7G5B#2*EEhvY+5$)ou;Yj3}%k2$mR&60! z+#0go1R`}AterOxR!Uo&M_THKU$@0KS%?>l$FtUjS)VrT z`h~%WB>ney-X`a!kb6d7=XHv?Mv<#X<55j-e*F&Agg_@}GcXE4caA4iElZC$V#8;P zBwcjdKQ(YSxj6?Z5UQaF5firwqN)XIEMip0yOn#Eg&CZ8^#sm~2O=r`oh9USMFCdv`(F1Q@JKe*AYPCVpLEyPd6 zzUFG-8DSO|$3}LAA%`QrF+hXX3vu}&2c(n1Ue)xe&szX3oS`R=ua#ezolx^n^eC&9 z{ia?1K!4q9N`KDA{Mhk2W`@N*F)+hxt8YEMnR8#3Nv|W1c^gS#Na6Gj*?n>B^fRC_ zm;yA9#S5;8bGb+PP0)H$>RZI^u@Rax4p2&li6Vl zjwd(T_ys#~F`Ov00Gpcc~?2NFA=Z-wQ7Fi6cwB4!|Jn630t%L+QzCI(Bj1`x_RWLB^nJZf+p7pyBn) zem=l`%v%xzbWvH9R-tvO^oIJ%Sz5(PULLGfOIpb`mo6Ge@j@_lbe+3ZQB5aO8byO_ z7Y7LOP{5&4x=Oi9dBgvPe;V)$EfcKxsh<5|Nuv~Zmo1WdPQPTS?vFVeWg%v)yj2@9 zoEfXbY9*TVqmx(v?ll9#y#Jv6fNZg2gjEvIhhet2BS(U6zo|Q?T0oK-L$Pu5B_R!t#C|G$sG;hW&p}AQ$ z^Ch!8JRYd9`y_T(vw{k>^Rv5k3IkMBC*I#rGT2pe+p*JfzzA3+6jMZz9aO{_I-NJn z@VN&5>vy=o2`3OBk6HY@EVZc zb}fJw8f(LCJO~?MQmnqj{l~zs{Y~lb**Lm(9C_Jggqci=0WW7ORR04}$BxCTf6^=8 zA6)BsVCE6eESF;c0ztE?LE0;C=Bxw4gjUkvSWh>+y~jCwHc&1&-{mF$Pu65f`D@X? zni7=R?N-?g<=AeIb7YmrUGeZwc=ACLah%M4AY-#z4UdL=+SUghj$uk~e!Nw@a6% zu>RWd?EH)oBhL0VwLX*iDwsBn?W`=F`Cx#% zTRhCEFzAr%*T64>)XWib##tNC4Z?VxZdpYbxT?r9ILVNe7L* zsF7iq1-$LQIxrUEC8=SDC7F@}(SiA}?JWz}C4#nn0|eOb=|A zSJC<2>Py1Cp{O|lFYA4p#AQO@kPFeL7}Z$XovmzAXp}k1DxjX%DDdVeDqD$7vAU=) zNV7*#TLWV_MZz7x@{uTCvgEU+{iJF^6?8Bj2jV&Xuhpzb4bzMJ;Fu|#`%Y_Iksv~? zbvh7^)fjha{LLZ-xxT}nzT_`Ukzl^UW4Eu}ie9k7-K5e}gKv*#OKrz)jhQVqW(iJm zRykHFfZGM%^gCBAh!3e0?Qu>Jp;-ev_EZ2Rh8M%x3kD9WZ)~};>CBOZrfhgYd-BuP zwrtkSz710i#_7gyFa0-JYsbbZ&nS|TOfj&uZlWSEW!Nr1HSLjDohXkY4YJk%c+{%{ z+T{(hYGH=+UZ7IGBfUA-=9BznvRKacllyG_f|UW8ysj+G!Oeh;mN(9SL=x;68>L3r z$fg*mx71J(+rN87oX<%IiI6IB-RuT_5~x6Y2{I;Cp_!n;^wnFj->&+8MR*La-g!{g ztg7Xlb<7a;aT2^bh3DOEju)0NnJ}Zk3MN!#*B^pSOT5`^pY1r@Z>IPgz2PNQ;@fhp zY&t4F%!}h)6lMo4$x&`|FAM4bTCqGDiOJOUbR4Gv>O_%-p_<+_wM!U9-4ZrRPx^P! z;2$~1a*>z{Dx-9Ntc$m?g6j#&p1!cES-S1>t$#J8LABd0k{Qa=W{P?2S>RVQJWzw30IM?y?0@$%#G1Cix+D7TTss*J?WEfT?=p`poJX)sN6fUFAv4t2K z^`w-A_Ys%1uZEMCCXC+TLb#$CW8}$9SX|@ zIj&a&Ck!PPg5${wVoN}b9Q*KZgV&nwX3T6FkR%qrS!m*tALMOxc_h}q%W^5}EbHoB9daV8^?e0;Ao`R*Ff+0*43$ zC3H1)%HkZ==xFBA30?i0>T;yJ65 z9q*WmHrTCOYDP^$M|+G4Qw5-esH>f4HxZzqP z?ea?bR$$srqk(dNLJ9>&nYtthrGUF2 z&!KM%ooGDZ6j+3b5py+BFVB8(B;DJvLHW<(x>%BB#~YMdqYX+K#Xw4SCkjC%%8R_} z#Cv6HVMS=1)IAQ?ir&yv=qtu7`&P$g(^qmXIbrf0G`-hNZTHMk=D6q4 z`;}le61+e#es&Ihf+YK%AZI{FwI}qH^>;HC;lRjgF!#o~2w~-*PCCJ_-S~S0YHluw zeVyF2W7MoMsxJA4V)`i3OGO-aLoaerhCLHikSKvPeVk*9P`&cIm^M-8O^;uct$gg= z%6$~BKB<5P3Htb3ozPtj=ep%ZL0!UKT%!?K;skOTgU z-D&qipa`Lw`-)VUgnp9U{VN0Z9^EODkjqoZBO}9bhhlD1rW8ZG8MH zG>YY9!1;n{ReK`-ULh1aqQwL@7p{L(2q`DWMRc#g(b5Z4~L0Yey z^l!wGf_u)7eZ|W(gjv}>GM3>J1z1*wTWB4a_XQjJ?`nJbo zk7m_lkA1%FvM!H(oFp#R!)u*Zxg|lzaJ{I9zUsEq74L+djkrjLSJ!%UDzmuP=YO?W zPy6@SBL&J4%e^0kUmigQSnHARmE_zszg@2X%<(<;mQTTp){$qhG@YJqE*(eKsY?C#DWjQcy6E*%f9sa5 zS{XRw0?WvoSpmm>laC$rK+hixz*)TW@HBGEj(2HMM!U2Dig`c*-6D!Sf&+~4ygV9< z@qzLy1F9)=f#@kxu+ZOA;F6SXolqt~$+||o;_RdR`w?PL|r%dD21x@p(Y4s_BZyO}n zonpgIO+)29NI*bXsf^YJXp}v)#%T+qzUq#jRTlE4FbjA-fW>)14D>0e%cdT5S?#i2 zp!-emH%6CB>lDMua>0NEXh()L(8xjbifq%Q^Doae1)rHYE&_kz(`)iCLE{B24%R8I z{BA5)jRDz3`3ieG@+P{T***aPd6; zH{{U7(M$1a6bEPN|6USa0?N32vHHKS$=C1-MBRMYkK<_rBZIPb`7LL4g06IT$l;>j=GO=4 z?1i<;wNv{EjIP%lMRJ)PVvTaI5>-mOgUJ3c}y@kOQYzTS1UOZ)GVuke*K+phq${*F9a^| zn*NZBoYl@PARmfmRHNt*iF*^IAHA>7*C_S}UxZCznP{iD+wmLuN&jXQLVi7`51yW? zpqq2z-rOJDtU^$KBZ`~`J||r!YE~gz669cYqYeKS?I6bT5DG>5$QV3JD|-dc_>}qs zBiDP&Ugt*-NofP9`hMnhf-E16Ros5v0g`S6osAR|OOXvAQGwBLBnAP66Bl)bN8aos zVBa3kyFYIXlV{miY3?dZr{3KCwmNR!A^fk86BREzJIH?5Qg1 z$5ut8u1*Tv?b{;B;#Mutv0Eq48k(n_kp)UOH39n+Iz@9Nn8w)ZW$~b_SXMa)%NS9J z+xqCS6dqM~QiOj7CJA{NarshzxclgP+t&V(i#02XfvK8%+(xq*7!H)&%OT z2725Df+ht}Phsw4$f1tg4e5*Yz(ZWrxq;3&EFNr9luWk;9G;!WF#+V+VXS_H9Y8$4 z{$=e913;9~YY&iZuZ;Hs+Mpxgw3A|><}{Cr`0Sdz3cP=y!2{BVzVCL;JLs3rJ0~fl z_lVDvRZO$0jlQqEtla8$!mU}=4%>oguND3e33}PR(zIEV**N)y>4mV*0(%5 zM2rbt!$wE<)9S#|4Y=&)QrX`d@bbsbKVB!FkLH-zarMS6BUD_Zm@g@Ej*94!YP_NY z)18(DHcI=Q`{rWL!LHfrMrnFrg$HCyPmpLPjc!tasS(z2fGYfp*(ht#E(bNGeD7UA zUbfHsk@zA>lz*VV^@Au1N35Z70K6P42lRRfwVZ0~-_<%b^2GZ-y^SA4Mu@Wf%?q$+-b4u= zM!C&3!0RKes+JV8Ge>sp5g#`K-hPUyqDTc5(f3Bjj6`6=TjP_WI>G2F$`Tgc{KZWm zn!qa5gqhh&T>e0q+A&MqM&F#PKI~Trq8=>_;2~(p0q@87La%HWWN>x&$1VomR{1gS zFNL~mv4%e1wO@jordC=SZ2_zoV8)`;m(P6IA~C>6cvtu#i5ty2*>N5W*fvM3S?LtB zjUp*j#9fH;Wx6ISdw7?{C|r%Kz-4p~539{IUfW!oRe5yhyxqd{zD2-vh5UMj(`~dc z8PCI5fM7g=ZQhx;Fz@|;&ObNhx?)q9Xvg5GFapmLQs7qk%{dRCYUCI>K3mIMqf7*% zn)96VB#(1Hbc3=>*sSUcErF!!m!4PMv;id|ty7d=_Ck$98=!TL@;gEL=4uo+1=CnS zG`tqs20r>x8&m(7%owbR^TWOiWHYGRb2K&qQ~1|J%d^q~V`f^%gA`kJMBuL~ zn_rUpx&bK;B4s7XnL-X3eZ{*e1_(TgsfZnd)G*BE3^~M*8$mZX%LSeD8l`%D@gav& zpU)MLG=Q4BVp8MN3k*>^gVb2E2UH5YD_+@3*pLJilWvHnUh%3_-r(TP>48YXP$00x zB{cKJnmatMFMRkJ)%`U031d5M-oB3#1+m}faMOEdonCgT2xLs#=HA9fIOouvz%{Q8 z&~btzzqeYu@FCC4Y8<-Xxs_Qny%8E8SNTK- z#?3B;4k&n%L7yHsrw_RqTB1N=mMtu;TcG1WP4G{yV2uE|psLl;ZgumdPtdD*{5*$1> z_|400`KdjF8Us-oGA^6T`%^^FD=pg6nzUW6v*a>hL4&!R!3eqVtNjkCGv|CDZuiz9 zcSXD8)T~ZFoja9;v--fsq!~GJqe5V%4?NfT)n9Qb)Yq!0s=xNuDQdL`hoPYV^(Y@q zHX+|g06gMM9~Lw1s<$qQPiqGvQ^MB= z4T#SQAArj&UsD)fx)7jg7iC zNt=UCFKQuIPAjAexXXEYQP{Gpo|9ruhe~`YJL)X8GU{$5vhcYLtF~BUk6hd5-F3ze zxv&1{#(z$v_S|#VQKp`v_ak;fyiOv>iJ4q`%jKG2 zf&B)WVV`R8QnX9VBviX*HO)(ky%K9FEGNX=IeIY&f`?SLJ z54izWyVUe*A%`-@0wS(-@m2aZI z#4bE!g%;2C*2;hXvk?2@TsEOY&;1@y45hs2xvdO@loS&kP{gzb_PQLk8+4NO>= zomnq2Lc%7xfF?~dvc)GZ&r6BW`xqrq;x%2DAjm{YZlvBQ7IpG1?z2K57U_&kqja4|xq zfrP!#**?JvB%bR-d+YCRov`PMW@E~F?q5bhmj9^z+Y@H#PI6z^3GpFK0r`;I2kA5S zXB~ZQ&HN#+n+q<@+wR{J`Y85cc&*QlMeR{%{*Pw9L*M)(d&9+JjadeZDoJ`K0{7=9 z)Q`mZYDh)F%4_UTk9!$h19kgncL<+U`eRmghNybcI{!icGU^i{r2b>x)c|$ePuHbL zfb9Dc;GDiE?$cnIxQo-=#;5HZ+f{5e+SBf5rR}D?^J>jI8ml+u2UR6`RD}me`C4q$ z!~sIyKv1-;YTL#;@KrY<6#+VXlc zAN!skrEbh$H@!`zrxOILv=ePY>`D~bVe4ridFC~CDEi6Ue<<=FK~ZDQS!x}-tXt1D z?5T1a%xxv)P$iv1#BHY5LXHDysCLPjfC^e?_PHPsad6lRqnt=4qATJIP>q}uVW+}6 z`334e4K_${u_78H%L!m=%v`P<_~n0RU8Afh+2xgYoO=I-K}mrPN;VMk41!9xw#G;i z6^9hfMoMz*%4X(-p=%ak6*QK27S?o4|GI8ASK?yLE44MWGNp9XxxXC zQ8P1xW@>Q3<7d|sdaf^sVyGwF%Q+!D$Ze2a=e)lF>4gy%7~$T^)m@;vXiVuJS^zk) zl=@hjd_KF}c+^Ka-otS~#I5>dPF8kv9(X{R8!p2XA_=+blQd3He!q zY9-=Eq6h`#I(nI41rOcIm~7qgtqHe@5=_7v~x|nNy zE`$#PBVM6*sRr*Y^nM3_1?;!73`*kDfJLn?0tPMrVO~oc?8iRhJ#~2-ACSf_RpS{M z0ga8~Lm7l#W!^TbL!J3J|^1)K}7U`g$L6xp)lVf9RK!O@87xt-BI6Y*6 zy4DAsIJh~_1%}2ez_?HnRu-K^XVbWANBlh%7>Un$SR*yA% z87O6IWAf=w1i$U#oCDgG|FWF5YH!YMmffeWX%Bv@^7X^tYSW(9wrTsgslg^7hwG%` zXSS#=&ux`AK#uPY?)mU?vWYBG^gxs1x{`o^z>)&(pn`|>ZTWOh4g57sdyHp^WOALj;I)k_|RDn>X#nn85`PSh>;G*OMZHXbyMdfaV?e||mb$M!uiJXWzVH0Lk| z4$NIY!=Qc?+Y?a4SxX^@MiFO=`ZQUtsEt66T2;tFbCZoL+zLq(StU6*>(s1+;vKOc zL2wjP|Mo)aUlDx;cv`opKbx5pGXw-u!_m1Ry^70GozynJQq5YR07m2a)aOHvo?zpE&v@>$sM8 z28|c~@HOP}Z;e2b8DD z)41Luxdp~(Pli9Bkwfj{VDN84PiQ@~hRbwcK#BCRb zN_Bmj>P202M)1ak>p~B34Qg;4HG=%_kUi2wpHc#UT&{%}6a{PL#4wMmhtnFqI_r5) zs3;aB`j`n3v3$QpiqXd*ubh~|C`+@mBq_Eq9E{E2!c8VgqW{%Mtnoi5tmT<`4Ux-x zmm-hXs_Fqkfx=*OM`dnhAT~^MIl&Zb959+au0;YXOK3{^sXzY4zPGN&n%)fEaxwkJ zT-c8un|4WuxQ!bx=$1jO8)Mp&r5CJHbRHJ275lShV+D(;uUBP0vQH3xL6#bZJZIdd zZ{$>mA>l}|paM9C%oYYvf=kiyj4@H{scPnyDTaPTC4zKKQAoTX9&Hh{91e>6xb@U^ z4%URDHiBXFG+yVRiULv-=-g|$XaI)ScpsN=O*juWM&&d1!n2U%#Pyi|(g6EP6*lnk z;CLuQT?=N=q`jm|5&bnM41=B{`4f|@eLyx&V!{b2y7rcu1n(UC^4J#y6AV?puoMZQ z_1JM$o_C7q0Jn>S(JM<*&7{M^NmRHoZ^AK}VyB1J13==ZU!aV0GJ>I(FBj=vlI4da zx1b=#FapgXK%&>RWRm1U=N5vUlWP+h8Dini|RC9 zfx1m@P<8sXaEHAtmHH#4Ex2DQ7U5Iip2LTb7bYFtEP5L28BAm%fx9BC0$uOdChv%7 z<|Yy&{1uinC)YRU;AOi)P3+7$(7pX%wLh}w6k`J~4-O(RP^0w(WGvh+zUG@rT>AP> z*eQP|$qU*U)~3x4hLkSlasF_~8Oa^x3gsO+MnBPC-zUphR7iF~pS|5w*MhC;I{wI= zP72>SQH;K8qHq~Kmn)A~^(KETwrV5UU%%2yrFw7{-*y}Ql1s=p5mY7-cTD(MXbYzo zT7xFbYa<51NqaN$NDRh|GdPVgy<|G?oNybNC+?z4q4ema@}k6TVzv>@IAm_~yt__V z*)T-ak4qQYr^$G%Vat&Ca-KY>Tpv^h3`QwYslm&oJAJyKv8Is&8CLxDL1&?~DA7NK zgB@o(`DF|_2CkO|Ykau=S8zBZJA4#;RsYz&mVnKIo;4IbX888;btjk;dsFgt-!2=DBwGcBfCIC%FeNMwe|c^Q0&F zM=?Dh<09DL-ds?G-3&}50tXr=URC6h*w(;A)uHxvhB&v(vOg$3kif+Nq8G9adl3jGuJj|DB_CBf9T~^eQP4g-=*tWP$ znMC)HmlZ`JrBQ7%NWOz+C;b|zRe%I^gL+@cJI>qME&>&kCbJ)R8hkh9n(qZ!718qF z2CSiSUl>JHYXhnsgd9wsqR;E1R)kkX8&ptoPn*arP9r#H(s-F)jt5d$j64r}9i`A| zgE?+R00OAP?T$96cP(Bf&>=Mhs-Q>y+o4zEHMr^yAF^u@<1`SQzVVMgl&qd(Ma;W@ z>Z+l(O`{ImeAFKha%e@ii-@xj7h|$Wf?zeI-0O-U>##G@pn4$N0WlhL^Ka}MZ&C1i z8hu-<{dCu*MI1CW8`OvVaf(ErrMNg)3gXS(FR_;d){L$7!#tE?fvY}$*L>!qzl=_Q ztWUj>m2u&@dS37AoyxO9XHMqd9-&rFqo7CKsBq1CLcWfm(ug?B=YT-n9wj<3>p9Ru z22;&2pa%>r0l7+844hZKPyKA2y2uI`PR;0=n^*zEbG_@=at`Q!WCh0ej}vcFA3JLC z>bbzC?%1ICDj^4pro+N7y^OdHZrQ3g-}zH2Bv&J+XMwsvku6vkVJ3FWB|lc+B7<4e zH~q017oc4qUI^a2`Z<{MjhWlGgsbU3?g6D)`CVIhIrwrgJUl?*L+uiC?=>9jB=;!m zxYrh%)59^`J?MLFAx>`o^62Tppq`+W!RxS3b%mx^`q6Aq$yg9|f!fY1B+X{H`RYw& z^&+DHmbIlzuTVGr^wZ&B%(sozS7NkEbn&tM7m4{R@f%Hf)fcPv=gQzSRD~u>k|6>~ z?@=Pf2;@GPQPi)w=hLSF7tT+Ces*PWnXrLOkGUAq0~T~5rP~-#Kc@g9QyT+}P!>K$ z_V%Coc+V$_q4B=&v!1*Am)8p?f)o!9G%+-Z%?{oZb6T0H%nBY3$&&Pe_DSINaWT-e zOEDZ`ica%}RH`=~iLWns2!W>0UNWfrsC$}RXk4AgJ0tm+GaTYJkeG!v0i--j*b}WEG3Z#D~(DU^QMrH zZzZTaA}&E*60nMDAnSxfvSLwTR4b78(AXWEyGsj4dmX1WugSX zDv6#xEiZ{FiD{zT44?6W$K9a=X7J=Moc6Uqs~!AUel&?PcyPqJ#s(B+gd75t+bm5< zN7@Ht>|lQwrsH>$9l}Q7+%z{5#+(;3?~I-{$-p~@9)g?xtAQUxS1rCt*KxCE)Nw!L z0g+5%>=hvvzPcVM;|GV0{py6SV0Q~MI2>q}sZIabT4_%Z!ob5(qwb|v@K*4$W?YK6 z6m$x@zCs6^R_T7&Qm2S;pbP3Sc7>)W+xwy?eY5ma)JI+$!FO44ZDYQO6?3X%LTn z)U6}TAU=MW?)ze8^o`~y)Aab+!l~~vLUVyr`)bcf z3IqlrtRA@s%37dv?hVu(pS41{1NsM|UmDCApmSag4l?Kz99BCE(!iZV7tcq~ zAz)oFZ>Fy*aNAuOSQoMP6*sCIR!AAMr)KR8>y;YG zkJP%G3wDBnFwR*)<%CVHdaww_NzqS4isO2gd0|m{qj-O;r{B{6$qAa)0Ag_mf{_)hgz#R~P1*Hy(J*(F;=-SaJ{H;de;1*~#%5)t-l zU|dpW^AIy49C*=)r5kR|w=Wc8V}5wB$;wbxRKxEUAP@5@e@t=vz&l@!tnh{2ReSd9DLQ-P1aIgxyA4|b$6zfc5m6yK$lc~wDLbev)f^#raSrk7gWLcS#|r{m;5eSe^`qI@pZ?6A&}pnj z@nCC&fwl=#6fZ{`$@5&@Nzn=scuM={AB3Fo70Suqjy^PV)I zk1~-*P)SS2uVVAu;Ay5cpkO`((rI$l8px-JG% z8@uS#*!5An06R!2=q}he4UVw}#>mK1g*@Ji%;0da-I?-_k=Jk7drCZZQ_SEMX_;|Q z+$~Gng#Suj2YNiU+y z?rg9)p6%_3-f+tnWdtv*phA58(%+k8Bk|yw|Nb_m_uwqG3LB^t5OSb2&L!fq<|5^N z3THraU5Xja%LL1%*QNEamF<(ARBoDKgs}@io|DFdQc&M=#TH4X3Rr2mN1#;xtZO?C zjR;}X>$umMQFFNCkR48f_}c%r&oS~?mBNruWMvdZHhhC>kj|1H_BKiOQw`y@KKr~k zd}-3QBo=8pj|+E5pLtGJxR`qM@Han>wd#+M_g0rsAG|O%$7gJ`%@IO=h@kd^wt>K0 zjsodN(b=jy6L2P=gWCaCY_Z_BHy$5QV&6O6E>TVBd3hfso%C^Ql@JcjSEHK3@ZPW& zP(B5x@dltHkq&E<(MgN6*vMlFJRCQbr-WA(M)IaIU!9~)e0eF(b5m&~k(_;Lf}lGx zg@g17V+*;M#rpXb=YjY6dd2qW^LKOp=%2F`i(a)tsrB!d7f{vgpya`RS?J<9D#7tX zLJmdN2Z*>OSTKEl?qgm9xk9iTvIq`9nmryHRF9N>TvSnQTJWKO&txO_lyECII~cDv zbI%Fev{+gF`Lnwe20$cq6@r>t%)7?>9w}2e$?^sXP-{fCLPM(|*>#0USwUuUOtT;v zic=qiS0?1hbuX5j$BzA-|Mkrv-&d?ys-E-5_o)&OHfuh#LB~Eq{sBSlCgS23bVpZ0 zvb{mI)2CaFi2uJNm4jp{?1eQKR$4ug7->$(;&Zoz}SaK~}ZYEhf=`?KHP_!}#F z-q?O9fck`;Ch}mfU7}5=szE}2pP=p%ai2?k2AwVrLlkf~^u2LAveUSBAY%sqG6pAztH;EvYJ? zDC8!uPhrg)i+rxeMSa{J5@Hl?fhyZ|(!u>le*M8K_I;{6*3xF^Jca0Jk#_R0aqCD!@Ve=k z0cB?E1{DUBWRM1>xk9ITx4VO;ok=^el{eEot{!Klnx=Lvx~#HixqPm|1`N!XEz+A@ zOo-bJE7%$8I)y}b$o^!mG25d|5gikz`_%;-)JxMf%LK=Sle1Z4iy8PjyBZ!2=VQi9 zUsUiMdw&j_^cN2{qM5uu6edvZRAV>J_1s2vmfd-)wi}WaFhln&*vswmBJRLT4KLLO zZB`%T8GYQ1qj4a}vL-l`yXywODcZ4cmes^jCnwcYIqXav4^DO1XJbxpC*;KhWhCOz zP^cz>@~chNA;i)Uv$fDHOrbWsd?c*aXUBIgEVh(-AXe(goUjz0LAA6XYP0Vk*x>7Z zE5%#L<8uo`69l`>o$klC`+kN9$BmcKmc%pOV?3A3rvI_Se%)bE<*l-vYUs609OqUO zTeRi!Jo(s>qnx6^R%jDHq0J|&zk~G$%WC%V-^{Ga9fbQlS3A$V@(o3}RiEtYDNmq` z(*+ z8aYSECULFL$o<>=j)b*p@!xdbY3-2L=i~K3t=gNBjT{_ztx{)=oY7xJ+vD-W#KDs< z2y%p3)f?=qRG%x0mZ7H|rsSgY2Fp5533GzBa$Eh^x|=4$0Zg1s#C75AxB&2JAjl~b62?>I|KwDZO*uc>l=X8^4ynYU5Pn>^sI~VR7={8f0O`wr{ z7k$VzbFR%RCDN%Lq3$~8TyzgvN?eC%#Kc6C(cx1T&9X(b z3st7A0(E%iJW4+(*y?4UbjN0Qp=?WpLnGeM7-QzD~AD?ub zg~cTKR-XJ3fB&V;efF*tk9EW{I7c#xp6@?gZ27lUTlV@pf4To=n|8^DZ*INFn4*j?K2 z#Zn(l;GAt$wMMLu?esxce`R1F1Y0^ublSRh@8yJSoduB-Q@Qk(pzlc=1FSZz>D9TXbz;c3IP|>YY!%-UyAMRpqQ!y2!6BC- zWb|aaesA#hv7+T&-I?D}iR{qg!6D~D8?eC zH9oaG6Lj&w@UvyqeOMU%`NJ_=XxKgOw$pe2`xnP1(ksS;p~28Arjw7YA}XT0Nvs&j zQMB+OQHeA&q2vCzi>{qz?jBS?u7p4#&>It}+(BQkU{_MbqImNO+>X27MDOZFNUV;J z^oJ(Dtb52U3EQK5$OXIC(j&_pkiG!LDc88Al2zJlAbV=z>x^UqWb>6PI(&4N$-9B+ z9lML*1vY)s*n)~d?>IZ`Gzi}mG@r0HZrSX4JQ!3Artl}eNwlsHYF85ln80~}TO@uQ zsk4xwWZy~o+3DjlslDJiS^_#6w(|Q^LRr|sc1SiBr)TxlFxj@Y!E!FG~yG=Hb>w_*& z&kWx*BY9q%w#a`MXFIo~um&JFWGtMqfoAzSN0CkW`i71<~+rNn{(Uixu zYBz97G!4{gZ9kPcyG@%5?}u%F>$o`z%U$>$MxO!Str~p(`htzh+K3y}R!svG<6ohx z!j>H6B~weBTpcF@FK2a+?S0N6ab>=@(aUP~ls@?NpQ!k0l)=VcNF(IQ1htBYGl6wr z=0?K)K__0<;aBpG+vE2R{tV|nO9v0Sd5G1n)pI?Ww{CFXooa=|^7rQ+g0xkSW!|N7 zY_OV2$kz~55)p^1Nxu5Vd|j5LlxrlhxY5Zks%tUg3{+kFj?T@-4yr9FC!6M35%Kz! zqBT_UH0p@W2fvq)R}oYtY$}MGKW|=GKdXoQ3#No(D^wk(%a53E!(JCC9Q+9od-r6Tvg@VkaYy{fY`dyc}vz!{M=BT*y=!l`*l2kY}R(6t{u1Ot6BN5nL=n zB5V_|hM8$5Zz>JyVxOgD(fyh>RgrjUzWS#BUg(49mc*nnco`3F$4uge0am!3awlCs zpFg5o&TcqBt@7YP5pc7Of=3o1-#}0qU>V7>g7tK>>@$Tybt$}BWa1-<)1%m)fCsW3 z&{y$-+cG3^n#3iD9XQ;80){DBfrI$vUxSy}GlhAq?qJ~lGDWvS72)1c=*5X;L>ppE z(aCajiQJawDDEz3<3P3H0EHDF5C??vLW2q~n#05xD((<3#pZL+$XS#U(IWkT?BwT? zpU8Sh*DZh}qGE@cr!0aa6T0Y?6=tUV_~ku0-?S=}KNao0K^=Kv9J$wQ)K42BKTS}l zh`2P~_GswpZ0`Be%E8hgbdT?+opMuemOKdP2^@DlIU$sJOaW%i2YJ6d! zb;$-=Erk38K^?{ZLv>_H7?huti!y^C!|8z9pw0zgVDVTXtZ4DEKqe%tQbYNp`HmK8 zA%w1Xez~6Nm2Rdw$*YQy8KAE){D&FcP?rQ=+(Ey*DDd=sNR=x}1Y6|BIaxDKP*}f) zZwG~AbQ1;XQKg(>Q8$_B1HNKQ@_4bRM7%GokjhYF-V}?hsL8~X1zXM`>$7CVg&9^% zwZC%qw^W7)+dsQ(P?Ar`A?e8gC?Ojps1o{soDyHx{$qk57rSKQz)|RdH0<>rwFQJ; z7l9G#5#3>VWu+OM`%G=_8n@tWW^w%qyH#y^?CpQDPw)0vONc4&Jmh+Gyx@Mo^_Xro zx>_s9MDITCX6_lPj%!fY!wA%+Lg{F;utRxAKIDb3cyF8bFqI%VOX3*5TE>p4-kV!N z-VeZ$p802}CUU33wOATAaz@sc&tF{%A6CdA9xnRMYhSGF7)-!L=d^BrSaIM*1qLYh zg{=s{BsQ~OWs=(ZA*QyDJf^k~xpX&%XayXSs2U+QkslSNN zKQY~EoLv7^$DgUBX;8^G+6R(J$YB{;N5mm_V7mla3e4O9E&PhWb_quOhrNs>@;qTS z!K8lu*onZUpYl^DVPhBWuP_eL%>eEz2(4Mc{^X zj}mG;xyOVi*f*G>*J})r>kK&}eHz@IW>UIh!8LAttO@oM1!Sq_ekeLROwr(I?SgX9 z_*mS9E^SqH@{zUxb{(7o@)1~_x(XrmKb?6dpaL9iu5!GxgO1HtnBnB9h;aJi3j&?Z z$H+$e4pkm&%rf+x>f~?bCWf!2&^$dUI;(8*zN)-5zW&iKpIwXSxv_I#!yK&v(+PPBL9Hg@dd2!EJzcJ-1S+3?NM|YrhOyk( zNw7o4$ki#FjCsoa1I#!vzH&O&K7hz3Q_h3Y!VqpO1kadd&D6C>lcZ@pTmg$hYUbpI zV27 zq=!khm7mDRDJeg>d-4vL)q{hIps8DKRP;!!wyE%r@IRYweOSeY4^A7!-UTj>j6VB5Lpk}CC62H-WZy*w0h9>$?ipBL5IcTjxj_DLk1H+{wjw=wv0 zcED+zGQ-UfA6sot)#I@%5(9@&G55xTlhh;sHU0%N2dLA^lT@>`J35bdTG^&u<8L{_ zG0RowrRns?>+nX_Sa3Lt>j#;!;9zwk4DsK(WuJ5Xf*7=nspGii(PaQdX(#xBl4G)C zCuOM`BWb4qH0CMy5BzdT)vF4t!P5HAdzMm}9&E5crtc^(UlAdP0>iCDTx~?p{GR#U z>fh$k$kp~=Bj@IZhCnoRQdAq!qOx3E@B0|?2bQTVrS4s1>%yF{8}ln=X_PzrK{zFS z=7qv-Fkq)XzO`^x%df3i3IBgDy+mD`MlH8t749eGeFSxzh|7yIx3ivZ04{qyed|jp zqAt#U=rjmKzj$J)y+Y2L90ex*9Ho-wm|@)^L~80z{(YbhhA(aPqORG4{(H&08ZbbU z=xiF>4PsefO?XZWWIP5}slktH3cWxbS2aj_$z{Yle?fzx4VuLuQ+}RiUCb785PDFZ zqXv9CBh#n?$OKrU%#!Tn9;K2(bokz=7qTS7UR|6fFcf-(@q(4UE5Z+jnaK6N)*>;3 zswT8P2)DiHPI<`f5Pl>pADsy|LHROEe%x0qkYOcX&-M5(nHrU0Z`V9mCNqPrbBuE( zQcvSI17W{MxoTk#rW}1F!|y_Yx($qD_#U)NM!G&sjhp%;7$t6xoNA z+1NFni+pOQ&62l;kb@&<6A?F}Kyg!qt@08CyF$|-Z%v%%1Le^BsUxyH5tK4xJ80Z~ znYolpq7KTsN&R#;qi7uPv1+-|!gl>}&aZY}v(KUQ*v%0`R%EWFwtb{)Vthb0twYCM z0yXISG1B*lFnzHw3bGc{c-gT>RK+5M0N$@R6Dwd2+#OQa6xu4z6+r~ET>?1SqeP70 zP0*~7gr)Xkh=r%WRds%DefS4rgPHNA*5|e)gMtM&U#+K8fZ-4)vna0R(iWB>_VJ@+ z97`e#Ffl6`2OFv>Yja-x2Pe}M4$RzwZCYY)iLvS|O;F&6gaXBaR#lZbEftT~hPwH? zY4bs_5`oQ=?mb~A0*@^}`+43^zHHC4J=SRP;3|0rHtpo#_vT_F@K*n#=seC5(IdLl zyItNBngUhtx2P1(op9*BM{gG+;ld}ryF&^3oX05bYhX+6Ck_F5LtE^H86q?z(ZV(j>U zfk-`FqpA!9BN&tOi^R|uB1*R|qBGJ+Zqb5aY;J|NbSakYT5`NU=Ooh=0M!R(U}jS9 z2_VU*6ZWC~)Mue9eW3}B9@2hr)bg(RYFzGgAI0>L*SL3lH&bhQ^$Ryev2YYk7;TOa z_>?hc!FjFYex7~43qx87F6aHQ?SSRK8eX6nXOW&hJ~w#=$OIko?UIv|Sw_f?lgZi7 z!v*&-e}Nqs2%eh(pxAGUnFtGz=l*3B0}D_|ltHy)Q93Y*>{dUVtErh2nB_ z8V~ENmmc8Wll4TR^9Yk3dLkPo&C+WN4ZsZ0E!!e-Rf?Zu3{FIYIG)ipqr3R`h4xkq zn^o!gvv0OZ`ZVXsDhO4U@{Yu$i)v;z3-3y{`*+EziIHo7iNMvs?lVB`55Gj?J*^Ux zq?*WowH3OKt(aEryT*U*YmK~>)7)jyPXsUKg~c`S7&FTcj@@1TPgWms=;kZ^RNFMF z-^Pb=osgRds?)-4&;Z13D|lEjicD|T#;TUXd~qS!9#IJnVBLUZ5b8QN`(*^fGF~j& z8wye0&FbL=ht&=64Bpa7_Qq`FF2$9-lW+D8tM`R8kUE^Iid(AL=2y(^qL&d^HfW-7 z;wA-5+^bkEUD^wkJ{Xn8uyrT@mZk!N*X9PVO|T?>Fn3c}vlNy^*SY^>D2zPE5fT%M}g%D%jhW<0Eyqb~S=3W*9iK z$mtX9d4zR=dhGZ31U2NvY61_|C!1{aNeUrfO;CwM9BL7)N`xx-5&l#-oJ8PQ1m!Gz zM#MTfJe?cr} zD7=AemUY64gB|LPlFdL5)9Kg3bvtaFg^9fZ z$ScKA;ZEIB=89Lyb3@Na)>4ll->y)-nOo~qq&O*Rn5<-f6eBwbxLfd_&(FM0e0l1$ z>Go;gW5MCUJ`;vi@O@$B@V&oJ-I!;tt<}>jqw?gHf#7-VqR&!z%}g1E4W`;96~LlW z4;eo!{HC9Owk(g>LF9?9VEHcWZ)b)`k!J3@!B)GYbolT0PLI^t2l+z$hjHmUi)MMNO`unCveB7+JF5@h!rVs?>LxBZGB;o(rAN}3PKJU zQQL_)EG#JzTm&C%C%-H#lPZ3>A?AaiJRZ{fqyT$Yg$B7eYJ(a%?@w#h=D&VgTmQ9G zqVvt9uO&`9Om*^y!_JVUQ6I?)$wG)iRw(f4G~Ug*n}S`Bpb4O6w2qkoJ{&)KLi#_) z?5UmD?0h`fo?#$@+AX_Heimr}VN8ntG?Ng9)6Bm&raWXuoDG3RgYSk3JLjEg37}l1O^o*q+`aop3YVt7NjcI zbGu}^KJG}|B452nSu@iR+$@0{Uel}r%0w1JD;X$2nNuOz;{?KVqx69OzFbEc?&@$+Gvsy&wh)gLe+&%}~F1jo#0L zmn*ad5U3iluCZ06OZ7PlZkKrIUExMNJP`mMum*?C)9yPtyg4+(-jC1V06(ru5n;W= zE#Ydqfb5Z)6X(I!!=KW{$?`t_s)ahtHXQV|3>rz}94s755V)VikM-y4{_mYxXsnhS z&sEB6f4C#dzRA#Y<*H?9JcLG)p6-LT-X`*n&(cGhf=Nx+eDb|^2EI>jp2P!Y_#RGu zO)Itqc$_Sgd9S|nLvy94Ro^|jC#+zQN2O5QkxgPWmdjC ze(qh(r{aC;_vRLm>jT=fcjonycVD?9FXSHeot(wR8SIQ`JL~JDAZYmFIkmkb)??F2 z7+k9>6pvz0iuTOyr|X1AVyfn4hBxywsK&s)*ws`c=W67`$Tsbv@EzPEF`w}J=$ZxZ zaZ(}@cunB*=n>|GeWI>XyB7oFko^c0+=>Baz2RV&MXdNsk&k^Ps>dRQp|-S+OaVMJ zk=O(VjLyN>46dHu9^K(%rqb8}%sqPgD78KMg!+-VDi*rSZ1Xz>=BPQv#Jw0{Z+nhF z#4Yie&vC~JBvWVJe)+A5gaSQxFHbRq0!{oD{^7tB9@Lz+gY)JJW$?QRZdg5y7YwR? z;Al8S)&TKCxuSZJZaXk^K(-109F+@ID3J1ojlyBCgWVoZP>rub_HTkeH}4{aH;Csp#;tqKEy(r|PIxt0R& zton}c9sad>J<#{hipk&-+fAL5rv^6$ z4teG9_s_e)!A>H1oU-V?!o(4h8{yJfbA$3`b(zLJ#?|CPj*uIkvbW> zNa8*Qx22O_l*bMMZV)xYl*~HjLO}gr?YSt%V#9--=?ugfRiaicCOBUfRf+0^c-%nZ zu|c)jw~ux&uc#v-ut0{R$K5u@jEwES`{tv+v#vD%TK;c?RLeB#mW_(OLdY)?R67xu zBI+e;Bf6o;K84e*K0>xYiMO6YD@dDX(I=&+F|luRP=mzC!TYeJ+-CAFx=Yn3yQ;YPQsT=7)z!%1$YC$jtV7|t zt*}ZM$$qLR1iP%K$KZ{HVzfP)Wp{W*;!P6sFPF(N%|crAr!NF7#5PuyPi*9=B4-wx z1OK<@x)hGjKlTgzEF+J#cK6tT|4H6ZGq*Lmk<&|lviOqtFIVZ)i|@$_$eNjL+N+Az z=x)A=-%Bo+#tYW)62tRfHK_34JVBoP5pO;B0%ygXZvIj7#{8$F@TpLE=EVtN_?feh z{{~hVCT?%t^DTd?I{HCl&RJ@m2UB8~+dyS2AHSwj3FaawL@1>{;bkV7Gbpd2rSBnsRK5fI||Q8Io6IP)Ys zlyI~GyC&jn^I$`V!P(ZyPav*Bx1c(p;La2G&+nZLoNL@N*(v{Gs1j}FBf+1Wp<^pX zHbAm{(!FO32aisXXZ=_sK7+{xSItJ$>>QS0a~zos|pSG6&sX5$-@( zPD$RubM78uRv{aPQh4@hXC7;RFr=6js7=sVr+y(=wU$cXR@J0_-$$h9jPytDoy=Ee z={qu&llU*|pLhRjF4-?TC+`pfV}#^3q>VIk9*1nDQX;Y>AH`sGW+%UqQ^#qLG?1Ob z-IKOBID-$H9n&O%gcUvr!GXX2foC;C=Vbov5o)Ceo1sv?J1Ws|Jt1F5P-#RQHdt;| zof7H_ql{#gG!K~5u8EViwLVqciU3`?Pr1*eDiAinm?TV?--Z=1rhM%eC$~NC`x7YP z;(SV%$0Wt*=^}O3jP<@%Kqk>GT<=@!v)*^WH;Jwmwm^%QkJQ-a_@a0S0=h;rOHv_o zKVaCPWRx0lJ78F$gt%1D;rk=2`uP5F;!Wyf4^|&{Y}Ch9LVlT`I*7Q(k>#9IWFB;( zt(~Qx-K<)=v_OqHASoh~#1vhiF6RL8ICQ0j%8X91OpJ4|K4a<9JYMx8XxuJ9>$D^+ zTV;|Qr8dvf4Ny3oE5hM?>bAE@whBr#?}_^~z(XMihH1=iT*X> zhK#jVm;ir+TU0G?DUw51dYi4Fc1bQcdE1nCG(gH0h<)IfZu7gqNtOdg_2}QwG1t~O zQ8o&2E;w|6Sym{Ua_*I$*X%2@7)qJYWs3ct^2wYqWNK&t@(&}Ye}37FK5h@NQQ_i% zl63zU{K;Rn5(*+I^BpFHUNm}e~$C^DZorkymlkPp=^WMq7h^)BS%wK{AbPq-BN4fc}vWg9q^V#`>o7LWVr z&B3R&lcWQk2gt};<6dBl%>KI{=MUN|4v%$~FsK4F5tg<~27OKZZdnCzwHkq~RX0N0 z3U%C*BxB2G%0|*7QkaZaRx91qrEhK#J?}xh7XFrCEZ?v7K{6UBZn&zry8wH5<&qzZ z27S|bf5t{$7laptx_q^9PIqJ;uo=|}(|FMP1e$pbN9u0+7ek2IZSY|Ol@YZt`qf>q zVF#7PA69+GJ`;&eF0=>R9t=53x#TewGDl#Vt1+wyw$sa_%q|wlf5ow4AKebwiC6T~ zCr$IPMv}XrV>j>g3%7qtTJ4Yi?*|{H5h~sZ9rXT!bM)o!;#}llyOqfV5vzO=cOk~nyan6d+5WsQBUt0z zA%wmFb_{MCZ5VsTBY z|B}Pp0xLQ5eX2s3EWZpX1by^s}||G4ohw-#F;$)x%;hXtp| z;y}2mRlAiqmfH9-w&RL{?Vmaggai)uX!Nrp?8aSi+~`W9Q6ML7`*YiL4^#O7v`0dJ7k-=Yj`CAOO{kY zIR%QlM)*jaju@tCV%;CS}&kFr~*3JTk&g;_x}C-<-ERN)Jw zjt<&DY!4xaR>C`pxK92i??&Dl-XNJBg;@)&+T9^toOr=TVmb5`Xd+v+iPV~y0jNVi z%^&tj<(^)2THB^Qz38y_aaDt)m)sm=9=Fw4j@aT~$4{19`pTNfl!%^?9ta3#Nitt5B}xe+ zX;2%yBm{0PQE_z)n#^#9?2O3Qa2%+0@S3=4YJq$ags>HMpI|{#lno?#?&K2K`PdO|q3d zi~*ga27IwG?EBD3JAQ7l*r4iFltmw<4vAJS*cmqLb%nmGsf$3yJ-n*{8r$cReKhu} zX!T!9^^l!DBR?(xsWXscZ{1EDj*&M${@xdhFl~eYONYG!iUk?L@3=s`>;>TrAlT13 z@h83)9Psz{ne=0wwI1wPVW|EsP^XKQ&&?$(V~lfh!V10171y9mR==i9Sj(#u_G#L* z1<N!&1$vhww} z?PES3Yp*hdd=mt@A}FP3_TDPSrq8vr5(GIxJ0US&XC!NU+?a0WP#!EK?eKYb4KgbU z2Nj-EB!}PqJC4<8efz`yBh-8Bj8+d`csAP@t?LMR8bKvPV*zy=Y;t#tD`HYaJ7Z1? zyJhGB>7v*0+`q8c%iKu}*+1s?^RNDJ+DiMh0XA_S58lc!ucyjE2g@G`Ya_J`H- zbhtUjT*Uta#R%GdfAQ9TMOjsa?7ww6)Q%TMRUEfb74?L?hM=m6I3zK`=9z8U^+7$t z4_{65J|S-4@AlFAbb~Lf2C6Q!wY~oGNZ{NU)4&6C9J_YmBZYnirkrbpdVLRdCGL@tgU1yi4>Eb#ZWya+y4DHgvh6 zF2d7!i~Uu`n{FOI}%-Da4udqpy6in2dIC(qpoNbHjT4hP>`d-d7p} zFVRJuqS=Msslhc1Dmab2&67)O!v;t$L5P!+1b%hX!U*fKl%((7Np1Jw0P#^9rBX}C z_YzbU5vPwz6CM=ox8Cu(N@25&Q(Rb#KmnJT`8Ssed27cvyzVIZY1W_EvD=PfMfl#p8NY#3_(GR z?6$*pcsVq<)?s6coS4GNNs+VzkyH4!duB3DnelrrJL}uNR!vg4&-HscX$zW=T7R z8#k;v%)gQ4%IWvy%NJSE;d}3ho~oMWz#-?s{~ zX={T@Vm^-93{F{e#RK&WXE$t<`?#BfddOu&k-7x#DP8=@f((vH@~1&40Np4li75g? zn(b5)-!+To$``>iTMlSo!euBu_vt3@e4+XkPIs1FFCHd zswfN2rp*i{tsGpl466OJ+a)Vr{z$U)vf?bKi>?=z!C;T@+QJ7C;CCatIBqN#njnan zQ82+PKe|q3{<~G_+*p+QGIhg)H%ISJ9bM`7fROhSR38y{hfD=-@O0iv-bwJ38B};) z8Iu#XEvz5JY$OZ07sXrsF!^~0w?nu_lPkI?t`E8gy&K=3mci|gv{1mNa0VdBDOsL4 zt&c8?E*6!D_l4y}WvhC~0(DWy$SgR6o_-*yi#SE@hdCSgO=PiXZzvWP;`?5eo}tot zW+E?0iYyj2a`4^Tw3Xs+{=Tpk0R|P$wAD{fFB6z%)%rY;m}Y@e6b;GxsW$oD*^L~% z`8RImw=%GgPL@N% z;VX2T@~#GwcVVT`1M~eYAt0wLx?OU6fv#o7L9vC2$Kq*11ZB-QK>=$|F*vP8FF6yE zm@A=!2%FF^GwR`QZR92fSfOKjHS!Xb?P$y8!Ahmt1|TJb9C#}W&|PeT{FhuZ5qypD zvDJ$zGAhE3jhUjG2}C)S=ks=t^SaCCvAb;e!RKA#A)z|`#pAN6}7N2z$t z+Q~|m8zoaG;WFxp$vJ#GV;Awo+H0dPFVSb%t5`JbW%~MJf7t!E*3~AD z82l!6f}Nq_!TRG>P^P`srXA$&7WYDv+|2O1$~`gXIcL)OPW9|8z~G^06|@w@rJ2b8X=jDpQ?6yc16patFC>+K=URex3XZW$x@gaT!!G zZT5Ste9Y?;SBX}bA4v^_(HJ_z%~)`O#C*)t$i&Awn`>a*DKaTcZT|f=1$oaVD_LEf^jTTO^@jzv!q$l z$?Ehn4tO{v-ih9%Tz>GU&pVrbixeA@7JzMffcrS|UL*wA4Qga$$30Pjx{k!qdV*jZ zr$zd{;3f^J-xzuCj?M`Kbij6L*sGK5i-cNM=^kZ9@O{6Gh52NAL=P-Z_+X+B5bPiK z!@#?dED>~*Rk3S$sR2F0KkHf+qw#F{!HFza(ltdP_y&MMx``@69@Shq~P4k|rU&eu(XQ>URuRjf2%oec&uig~i%7#3X-tD6YxhcGTYAo)7W~C%9XTYvSS*uL-?Ids0UG%UQCRO0{dz82NkW!(iF{`^>a!a_H?&BuMWXBGBwMg^T z=YUcHS^Gg#Y*F{ncY&f7IO5K8^P#i|-M&a1r`yI!;S~7~Nc7XO&Ux&QC zw}i$pBzal(*2RS%sSWCPu1kwJn}a&}Ez%2=aZc^44s68wgKNELJk>i^$<#%{%CeDjj79(s`8iyDsy%NRVqo3?Un8g z-5{!t*`iJj-s*Rod(5vhaIN+f`9OYBq@TYrG@r~3-60$BKTK^9>7y<|8{aHRrGK0D zP-!$335afPlLcfBTx;d*QY@D?N2EkFkcZXvGg88zx;i=*sLiO` zP-pWl>`$t__Pu9M%gZLv<-s^&An_fHUE%-P%%qsp(9XH_yO+esuh*vC=9e4WL|!J} zqfSpMg)b{-BYW{q;EAgjzs#F%)iBq8)$wO4$&mxva~E8SfyW(-qyBV;It4^zx_)3A z!@zpGq+6Eetuu;taBoq$qMTWK7P`;7JVqvLb;Fnm+%&+vB7N!Yum7ukTPlw&hGFQJ zg_-E@E6@?=$jsMT6$K7 z`G2r`7gQ?NMQkJSzLYnPt01EWjEqkHAe}a+*>_z;4L>QS)&C)NUWNvB&EH&j>yPD& z2k8%HP1NofBTaS~W8RpJnRtQ&9p96^^~Fj+E`lyITY{kpjoCMNF~&sZtItOFkPk`R z+8AeI7n)qMF|5x?*Y5-cV1Aq>I$ZIVPR$|Co!4bwLsw|?W!y-}^#rw+h)d^naaO#X z8LsP&J`Je|(56M&CevZVpIV<5=|~3`mw&IA{|t7YDf2w8zQRhYd#;^#`RG>+U#wI; zvwFZj@s{eVPV%O|i|enPgql;xc5*IP-qvycwX_*l3#9#(v%jS>9Bo}ZICo~3jcUs$ zvn<)FVUmuyufgz6#5w+GZ6_?`9KT4+^;dB3{R@do$m zJg30m*q(5u?8nT-tmf)on=)h%6OY}+FhFA6^bVhvuN_^C4p(z90arXj@UA({!gbIW z6@pAfVx&0K9i;Yxr2*#`D4P-~k4hA>m*@Gez^4?_^G- zaexQw8HUCIxnz#wAe3R}C=9CYfjb39s9{oPnl(V}rcC_1z7Hm|hct={3_I%A|M8!? z-~aqK|NOiECtgX&-y^7aN4mt3wB0tg^IIQ<+E?$tAih(ETK`r~>B~nTlHACtR}2Fo z2$q=SaduOEw4TmY_s6tpx6JB!6$l?d)iiQ)XW>0qOIge7*W^?2b9X85T6VzuKudFy zLXy-v5@SJ~{FHh3G!0}$?D}~IRleU9#WnF3m+gI7;KvcB#*VoNQ3~;z-@jrtQ>y3u z@qMb~g>jTXyl_-k#eIbQ1A^L3#4WiDl}33{nL$YQzJE5>D>a87g8ZtBiq#zSi0Em& z;{dlALWKiVyW}JQD2-RhP38W%jLe1N$_7~@XaDT0kxTN)vz%P=0%r*z(0h#>DL02? zI9a0q)yO7tnc#?v$Y@-S9v9_^ztZ=Uqe?4!esXfkX)4Qu1Ii%&QFtgO}?o^#UEN|Lv|V_seH^$g7h@7 z9a}j!{gLXp)tsg8-fe`<9|iu$-19@Qt(@=7RkF0osT|aw+4Bz zal6K*BzK6AKP0FDA})ox4s zw3zWiu?VLqjrvGtp^mSJ?k2Z|mIV)c;eyvh;sf|w;qQTOQ?587!-hHcDO~}%U0&%u za+*lsZFmbZ2^dsqyzSBVpbt+La1h{b|2%j9Y=e4fu_!MJ(`i=4Y?-x;xJs8Vz9;k8 z^e(|1p-yN8U{dzwUsA~)yo{CDEMqx@Jd2<&5Jf=_1 z(wzw`5v0VXi%ev-yn)OYWk6_Xk`yQl&1p4p-KsF3OwLMkPKo)SynD}ksJTdTP?-;; zW^LLcj)~gAtI*t(bWxkd8UE+z=81>AHpn+Bb3;v&nRvnu98YBD_5)VnAYy;5@BH;bQ8SbN6leV#S2E z$gzaFL)->cZpOe)KdiW(q;}shy5uLY`}=QzvX} zj|M_sOHg}>xCY6PtedRk4+=l@ZdI8iy1K9}a`da=z$&Q3-=hQuA>l}U$aPKzr`ESo!t!uzK`__J$S)bJN- zIXgd`2Loe^jn>H^!)s$$aHArAVh|}Ea*V~hRH;PS%S{O!d2pScAjIk zvLJhX;jt+y3>u&}v@F_4?he!ShTfL8ONzw9UhNXBfYz0BI-mr3W9X!c+gVZLQsS(b zTyY<}h2%%|`inoa>Wi-Lef!%~y9e7H4{S8RO+tR1piD#@`oYn4x|^yDG-2^4IMh#( zc{~U+sT;^zne6`SQ@1L#)OF$gBV z?}Lq8jS^Edra={RkH3a4d${dzSc<6GJB^2q7OX|x71Jzfr%wc!_*Wx`Bhf9dYxLjj ze_4Dfej}8t5C)Z+}%$pl*kFKrvW8 zj+kTEpzUeK%fzmY#V0GYP2KlT#cRJ<{uH!_Z}M*PFs!mlyK8PU={C!DtZ+t;I0K;3 z_fOXG+eN2Crdk!y^7rQ+qK3WLVG1}>D4*ZO{qi?)|MB~u|9Z*qo}_h6)?dK%K2(m4 zx=JPFYX~Z7B+6q@Wyda;TQriRFFez%t{eAML|p%qton;ZH--xqE>^v~M2;c2HKG&9 zeNb_HF!zY+Gbwg%?IulN+UoG8QzRxdVHyRjNp&#ZF2O8r{d6~4Kh8jfd1Y}OWbJ1v zy?j}1-|}~C0}>C;Ehx275e7n@O;8(&xE3k0>gefDz0aw8$j{7MUjzWsS?*D;_iF~Y z0GF3^JGTc4y(Vp+!yX3q0&~wf^MWz;Ti^fKo;Y_bKsIjN7)gEf^o{wj<~{m8%Z1JLP-B83I|yERr0%h|be}id6}w_(1nbb-)-JJ- z&9RgYyU&d2Zf9i7^sGC_nbnD-pifP&VW$1+Q4p22pwtyCDvBFOKx?rGiiisT_f3M5NH7Ny?x}O@uX1wE zcfQE^p7Vb1`n&{W!>u{_kk(j^4BY`AN7zljwFNrXUigvE+XqnWptH~WFNu~hWHtdH z4)ZNpi8vR%aaCH#%keCTsNp{rpx&}iaoTQuJlU+FO$cH$p`FGSW#=B5KIBSy*J~X~ zVW+WhSPh08m0=O*y%Yn6Qx+AOrbfS7wtKP+`#=mjo#3^E7-@2=cv(`61vR+03OnUJ z;dx=JA!dNiU(8~8;uX%8Al+u8>Q<1^O>8%P*L+EBLd?j*<;hv|8SV4wTFg&pX-tUG zhfYl(f#ne%x7P)DzMD&4Q;!N-oK_ygEoe;+aC(7FAE(?lVbamn44?6uD zHD$r8oblY@*B8Iv>EEO&qK%(dELyyj>hwRW!Nk?d;6tioigJ3;58WK?b00xEtKPdd zuvLgH;Wq}?hNl7>>g7dOrFjdl1L@5PaSe$Hz8|p7_b9!ed;}Tuc+Z%$4tp%all8E2 zFqVa3CHlE5xz)G!W&U$@HmV~@q>kKrB@0*|@n?dvjzCyQWoR6mEy08sZBA&C?|bER ziC>K%iJ-NUIrD<7ANYGQOr1INpc;*$t*U-~*A~7L?^PBCS#o0nxOMQ^5Bk=> zV=Py^qW-j%XOcjF_^7vzEMvE4aoEZRu8m=P)<%knqsV$Hvdg)IpBw&|Uo{PAn2o`| zOz5Mx%gsT1&uFV+D=J2t!x7hZ79ag|JQeXA)?}ukl3mA!Jr)AqmFG$0*VhCH`9yvDVH-`acg$}Pl4%*7JM%Aj&`w`j{dO)#Cl%d4P6sR~5DS&nn zXn@&SYZ+}B*g<2one2P*obz8IUovSl>Rly&A&;H7o1y^W#$h_jdWu;?k(E@Wq2p#F zR4bR$NRcpst2%rwp1;DwM^D@_D|79o@b<;2p(Z>i{;J(h_OjD(IjqZ^G(%4X#T=$c zDOAP_VtCu8te;gKfZ21 zg!U2GRjPQI)6!{UH9)#Ed{n2#e-;J6*KfAst>qkvLXwp1< z>#8&gO89G_g#UsD9Wgo5gX%M~NAo+q&iI#2-=VG%WV#GE;pdGn$M78Z#*gnQkMBv2 zb~mfHn!8SzZ+RY%e-Tq+FehORsSSi^F{HrsyTu6-l-qo_`RY{IplZmH##G%hm(}vC z!45?5a1^me^YFVJcWfQa*S8iQw(Qiw*`TA9-Wx`)6HHmjoK`Qa4uoB%HZVcCfvn=K zA>9g{YJ!t&*`UC2YyRxa)=@A*IQpe`4>f%2^PZrPLZ1Le(!=5kz=WRG`)iXOxtBfr z#D~tCMUPp88jUBeXKU4`g%w)eH6I|1xApT zW~_1F#n;2`qWyb0eO;Lw)P-;cZ|JRSGg^?)~Pag*Y$jHQMVQIil{@ z>l90-Xk)%I=!99{c;h+e$DX~wbC^dy0a^ykmdOk24KH)4Q}hDc*mVz_F-@H%HEZZ5r4edWMpXYDoG-r9f-syLp|9Q`uY84ho zB&pLD+>qadqWI{^b&7Iv%=~0m9ke7|t?ZLkOfMI&nB1uOP4??O3kzRQ|MvPtNsBIh zFYEQBMYk6we(lJQ8ed=a?}fkk4cc+3kn3B*Z&0mg_Dm^K<4iiY^8pwEu%iZStKhQD zAhw@SV9$;QtabxVPnzhxK2z?Qe2imV7cAZ>KN0KuSYQW*&4 z@mMC%EK3Z9TG>`$J&g~qr%!pMinMJIrmj?E`ynSGwlBlR!uAd{TP*H4kYh2EeLq@9 z&i6&%Sn8H7RsT(4bT7>-7s4 z|22b-fjpDHFaE1ewTmiKWzI|jd!}^ZXA36^BsN0h>HF_^R2aU29W;L3nO18lFRZvR zuwT;|aki*O(eDoF9L1i=vX4ZYd~)rR$iwRC*!b*O&1sLpte!o0!>Z+sq;WhcksQ`E ztfWL9`QlaS-!a#_5aPPo{A9^(k4kbHh`u@}l+(-j0^8Um$R2=YGl$Qg0+Iqu`sd7) z58_D%hlzS>%rwekiYcN%%Zx<(6KOxeH7`tdZPK(UqC}|-wpoEHQw@@hc6oGxf2~$l zFURUII6JpfiFd^MJn(E4;)tjBm++f3ne=tV65cu4Z8_d!=gr4PIrnsrv3kU2p`SmN zm7Jp*g}+=q%LJT7Em`YGHao+GyMhvGmWLZI2Pvk6BKx2ljOwx&D%>}Lo3lZV_G~41 z#SA_XCEB&nh-9_ykI}5nK1Z#6xlRl{-uxqY}wk7O`ao%!DFQnVvguI15JSr&pabW zrj?6KZuoUYs}M7iv_;_wK;l&94G6-1LWgJG@Ypo3JXoh1P{c=Mi`wl+mtbXa#sL#6 zGur30`rUEAo7b56p6FcM<w8O zP3p+e*Ef`sG!E;r)n*FDK8k?|Odb_^-2+*=FsrrS72>>mrhLdhLiX}weQE@?vTLD0 z)0G0PlW@1!jwII(xg^Z32Q- z&DHFJ{ZIfiu@faJe^4flr@t776%i}_#d?8q2Q`r4yg4-_&2$zxbyR8LslX^|{9~lw zANpiu-DnBXsm{q-Ug-(nJ7oa+T}B1f2y|1+h3gr}I5^=Q1Gx+M2SbGC!^-J%0g#4` z6sCs*uy|-c=*Ez`O=~y#?B%EOmNR)_I#rf?9}s+i$1gs-oX!_R4^8i(?}UO@P_1{+ z?NIA;S$r^}guhR0=xv7RuaUN?QhB+&%P#vtHr6Zhd0T+u;~{Mrl+V8eL?1{2)F9_< z-odZw6_aO}6o?nBz&c;@$}h)beskEHXT|*9|K?R`Ivod@Ogf)>Sq*G%(UVt9&T=p0 z;c*Im5vU&UD8r|~W2dZ#-=S{ORFmCwJQ(7g%J_)A{Jry=nO6D~i4&agyBFR7jd=&0 zmit9BEz&%3ihm3=g07df+q2h*UBP3=^gY$)U^mOAbuWndz1vLaBvZ_0ifp1HQO3aT zbc75jplRi95IyL~i4zzoqHuF+g!EqAu`1@;+1qDTxnT_st25d6!e7IETV6KpNH_lL z@;}Jx7iLF-uFk_U5|SteoCzDLNF@5fI9Rr%O__$?Vxdm;4)%S=WQ5wec|emJPwdZ~ zG0L*B#KtHy+4s_WjSS^y!p0x=*2I!s?6ARMb-Bt68$}cY=_`ANShbPazfM{!XqH8Z zwAfu2+o@iI=2UAm)ldKrYK(WOE3jIg3)ai(o%2QcFvQrIHzPCy3Yljj2S4;KqAv#{ z$00KA_Q~vRk+JDTZDxNx`@z1b`3oJRR+><=IX>_LNo0o_4(qu4%utg-F}o;|ibTuZ z@(K^!Pq2`^M^U2aBPZQXh*7AlbI(-Y0dM~&u4m_+CMlAgKjbf5er-r4+Zb9gxNlN&IK!<`|X0Q+*cO)(5UPlX( zr(6v_?Rf<_eqwmrBQinfxN}uZ~SG(Iqc7TlmJa)&<*XMXCOkyTw z>Tmx*5;(jIK?ck)lO~;FAPkv8MPgEAf^xH@blx8KdijT*>zF7}ibSVc$9z1kNmCjI zr9yKd(Zw~LZqzi&@$=i|-~~yi&uZcYcF%aRT%obAyyZ;x{mg1RqQ2=G_*cuWQ=E-X z>&$QxM=|RuvWAM((|I%ZlIsfarRqUyte>)qME$Su^4J6k`>wW4v$2B2OK$i6^=IR$ z=HRe~Vx^iRU7hX`?b$3_?WN5JPUMQ2>CivsTyV4OoLBxdEhex)t{sh*(1)HEXKHm* z;g1Rw7%Y{x`;KD1qBcS58BGLFn9pPHPvnmazhp_J!dbatrIXP4e?KhUpAE* z{J7Cj!m}zcZGm362@+gTd2O3BXwRb2(Uyb#l8-i<{jRZ+BGk2R~Wyc`sNuHb2#=asu+g5|rpU>7@VBqr}L5i@{-<6Ob$0A8>JI zk)~UUwC;M?NBUh;L$OWWN93+6G_F%&a@}sXGuW?g>T-E5)%z-f1_L~4Gq;Mi+qIBo ztYEMNhvDpG^htmMn2_sidy+3S52TPn)2vpYU_L|3bk997$LyE#tnKe4v%0aQ+&md2BEw3)1 zUUA37h&||38@QA?=XKW`xsdCHdq^8erRMi$s4(alc;eDUz>v(--jHYUumg0f z5Wfrg6;PyA(np=wxfrL=srDTRcKyQ z$t3b)f;TFn39zfs$&x0`HE9;FR#1uAb90J#yVT{vPC8bZ8dNT>bHq;c3<}sG$|NnG znZ^Dm?l1VyGvkR1aafCMotBM+LtO1DK4V1@XjNUA3$3%KuHhNlotC*Y$njL0vdlY^J~ahOk;}X{L&rk= zgN;~>R!k@Dc=PT--R=&{>gZ>?%X8SvYo!Kyspudy=m9XNK$<3D1uk|wiFup_DD0JIgnmf&X0w1`f@tfHBWq7esea+jxDlVAsvnd_82j3kQ;V{ z@H10c$8_q9`^_PyJ!^+gS|Z6~w`Xx!?LA?(XO&Y7=)ni5$i4h3_cq{~)u}grX(#L~ z=SUL}#&ru4y*kt_$|zA$_&(`?Qx@+C!S-f&J=wn(8a?CrCe8hbO7JfmN3U4)d(4l* z32FnImBTM>3s^_7SDyftHnh=^2=e<;3YvR<0MA05mh{J9T8mNG?4N%b7GY zm?TonMvBB?&UQI{%d>P|CJAMvOo>W zh-W8o(3!P|I^>p|9Bh&*IczCeVel|UDXRlk3d+URfv9Pp4LBRnrfh+N!HJb5u%x!a z!a=KVb=p&+AI({4+DpDs`~7@!n8Qhv7tQvK6BP3iMUGIBtChEc&WU%C^AHUz@Td!j zrZ()?n08T@2fFYu^uRqdJEr0M1D<)4I}VvuFaJL0@o@<>|d^4Rm3x>NqKM=3;l zcDVLHZR?(}BI(7z%$Z{bmaPD>qF6llv{eK==YsX!W(2oC_153glPy!Xt>ip!RTTzl zu|-0&Y^|F%LHWUTWf8B3pABY(wpvo`iJbe?WG^3TLW2|M4=BdG7SBEs`>xKj^N;lmtJUGI2(Q%q zymy{S3c22^OCTka$yu|P{V2r%i(nNMxnJ7p*%O{BN_5S29e&)X+3+tne_Hb6ny(MO zapj9A-b703<#ME>z9(-}B$3R3DA8jG{ie{5rEPS41l~X3)FD3aQRZ?os7~0V!36y3 zx%GjW0gy%@C|J1nUvw&P&ht`5`(~B!O5Jw2_S2k2Cx5d>D**#F0P!UTqog zdDfTD;eEtPaH`l7iM0&6^^t%W2{tg(B8Ns!Kv8%r^ml0k_iJh>vOMWjne-hVkk_3N zm-ACY$EWq!unuEIf(=uTYlwA8;+lDD(p;0Sk)r&MY_fkc`PgjxuB8~Tek!QQV~`r3 zDyf5H!m4Sl!c4lC+>c07bqM=__PRw`M}6h;$7S&*ABW2Cq4gUR zM22p5;7{vs?+6g#t~ggLn$!JxuWiM$qpHElpod1)`_(VW9VRM|7xxaF@=Uxy|`u2o|eNL__c~BFdp? z!pR%H`N5wp*?OK;Z8>b_SYhx2E58ZG`vP#*II zqBzIeHAAa7CJ+qW?7*Ma_r)K0|29i((neNqz4rge`WGfWCC5yu+DgQ)r6XuG*;ib&6IR4;{OmKl}UI{5+q%)uCsszlkGgsN9vxKlu9f1>y zR@`dX8vVT`qUwMY$vXZab<2G1Hs?>gFeAm_nKgEZFYvhRqRkQQWZIPF^g8!WdA1~5 zWPetFON81C_Ghhf%c<=8dD8b~p_VN`tn|*fI6afzKIQ6kW2N{zm^hvSnLQJ;ZJYhl zwmwX=nrnFvBr_;;jeXbQGk8aYe6jhKemoAqW363d&I^%Bn zUH*EXC`X#Owu1Osfq-Wmhh|nFrkpPQ$JNh!^IXhV#B{|rz5oXLaD5A@P4LuJre-HEXAcLn>5DrTyD?zCkCSm$PMM*%VuuoPdSi-yOvjh@3 zD;3A)HE9y3-|_76{gB>CPLX1lCd~tR%#>p34m(xoJZH(q0+HuTX_w2tE6u-XS;+pZ z2bRNLm3&g+8n$jwYP|<4%nuso!%e84PFql=)^4WS zVNm+(+8dzRV0F+zf27gY-lns7Ltou2&-aKEytnwnh1&NP7c6|BNT-vCF=CVv`XL|w zRRmY`he4kH@_w}m2qD*gmYs7zX?=4=Bz*Up%BUe=m9txyuf99Nw!on4xa@BnDtOKt;AQ*v6%jjuxGpQlz*!YwwgcC_wEE z-|LU(@$U@(CQTfSg0yhoJZNdS&-un292*tBg)#T8Ms+4 zkT&4_dPwJn=DRdOW5pt95RKoHPA_q85y$YbmZU>-Qh0Ng@!KCu&+~TibAcRlqx)yj zwLT%}usU`65!Y=V8${qikm-RZpo-U~yyDmGksEp-s7Z5Rev3FqRH?3YPvEbb-#L2~ zkeft__D}0#nl$%Zmx$t=%Yz@uPLYkCk39!G<7PH#PVkqAmJ7;*ce>x4)i2&Vze#h2 zyD+S2TPFRyTlKsBZce>9yF;9yJm>YPGF4SAEe$OTZipCwlH|ML<-sZRgt6%wHziy4 zz#-d-qVf)a-w3ec(d zD(*)dQ&ddv=j)goa=lFZA=E@8Gc)>zj7p6C(ZmWfLsID3pU87pbe-yC(A(L5pb7CSAr&IauSwI--!H|^G)aMXJ2lEKKpu= z=dz%Si#GhK=@(U=@ew*ztyeT?Ajy!_e2LdSw}m9B|FQV}41PZ-IY%XT0uRjJ<<_K$ z^?BfVo!;e^Nv~#Cmv1LfJqxoD&aeU6g?sxiT4u~xNi{((0-b6NH1yf#yv@CR=HZ#` zvl>H8oX`kym{yWNF&iinOGRRDOk;8#bn-p{O?Q_G>Yy1SBv%@#7VW=kNL)sowZ4JJ6OpjcsjD)hi2DC`b``o;t1WKf~15Hcz51Ap=Z<{mS_;IIuC zs4+ZuE$uqMwLH`5*QQM*@Hel#LfSdJi7YpBzV%W}4@J7D$kb4*EHzf30x70YPv4WK zL&mm|eI1)HZ-UNuhU~d((p(@FOn_>{E{}RexwwbFZ`xf*+ct3VXG`|*JHvLl>4mK! zoitLAE;cgnC(Opi&PFo+Y?s9i@-+22px>UJP#6_w&+?C9Jl zYl4jM?uaa2#q`U;TZGya=&lGf5r)f9Mg{!L;M8sQTk78`OaU5P zZPGN$mj&VHyoJ}5T?_9inlw0Kx7(4h<-83Z6MgRl*Q>@@a?Rr@j^am4s z{=4B9Gs&lu$r`ggvvkT74WWm0Q*#RfL@Y44K18Or%K^bxY=u?_C!!V2i|g3VzXG>0_*Y|sE7dz7q+5>OW^mwbA}!*h3o9Wj5Y_TDHQM^|Q(A>o%Mx48>O)qFW>Hg3P(J>-PL<`c zU%fb8oh!?s2j=XiQ=uEcIY>y~JHHvKSfKYb-NY#+VF+z|7j8^bC&35k;{|GL$f!rlVU!hNCy?U@#V`d#hxoeb!uD`=)U`k zwnqsLVK-ng1Y>E?DO%9r9w(?%ESZuYkS&P`Uh7t?&UERPx6i^{luUXPFO$9;d?!3j z4Q+e{m?<13+UOD`N{9eY-34)ZaD#k(1UBl_?h4I-64^uQd>%CWg}`*m{65!5o+2Io7e%Hj-Ca@&z|zQ($&wr@o?jeG_Wtd^C#8(8bSyz{S^@aFfsvbV{W$)M2=cjWG- zm~@Klq#^-SYNG`XnO10O4?1B$uvyjybcs5k*c_r=z)qYE@+#zmFAExa*o#uB10jH$i3jyI$)^3WqJVGBXVCrJ!pH$pTA_ zUlg9U!1#BgM)y|5>zCfVwGb*Be$@HH#@Ekkpk9Hu-!E-JlKQ~(CQS`l5n42>)HQ9v zfK%FnqVRNxMsDU^4@mX7I%|jI6R$0v_G%bl38L{q*&xv@r?kswR#2wAUY~U49TQ@T zrkzrd4tClxci#mPZDwQMrx<9ubcc#OwD1PW08_9v1lpSvzz1Vp18jS(LUjA3(fI+1 zuKT2STySf?B}fk*P$VeNh>t0H6oUaDgB}e-L5T|NiHHgA_9zzCi+6#H#eo=CkM&s& zZJRok3Cd5vwea3`utTAA_+dz=vKRaV`7Rym7@rtu(4<8}G(lPG^=Q8KjQg1Yv_~;B zFg0|ksC}-X#Zw>2cfk&ontr#PU1n9egboA{!G5%lTNx zVf1-*(yPE_Y2-n%15Rz)epw9yq_?R~cYV=>yoII04pPNoS#{M+Ry9z}DT*A2gi~1J zT<{*%8oWnEv*N=uXM!JGbBA6^u1Zs(2g+&59q(Pfnck;KyK^kmlVm8H?REAt)Jxd}Algyw%g!S~i`u(ydVs889h9t6bhm z;5YbIf;?z~<}%00*?DCyHG-W!c!W%T+6-ld2gC@BebCpst%7y!q_fm`Jx+jm=|$ni zp2wz?$Q_c;VV7kY!()HmEv%M>yQ=k!!91-c^V$m%kS#KkSm_i4G%hJrWLbn>@@~ES zjvzzb?)pdk)Tlw7sz`iLjGm!Hzb&#o{5yg^X~wK%pGpz~vI^Ug?o-rwBERKWWXtow z^z`~T;Qn+H2OP~k_s5TTroHM9ANAIeWs}Jcvj@D9V&W*Wo{CJEjm=55*rL4489Nsg zctE12s#BcLckHtp;n%kHagK1SLr<}i9^BP1qwkmQv8?ptth3NcRi81z%Rt8%N0x)` znJSrxW|hyv%OCEsN&j}W7Hi5-)Cbm`3SqDqfOTYW%HPa6A)0=0ul7sB*Z0_n%q{(a} zJVr4`DN;j4RtHoB8q-R((SoafrC?YCpVqupqTKL)ezqi8mNpagYHXQ)BWyi$BkUj$ z$G!6oTE*xDzbi%_i?zIS@=h@pTXxb1fsLS3sqHcNt&74Bf`1e4|BQ6fJ+wB9*GDj; zsC{mkcb}|Pr7e-wc%KQ)3WM?6fu(Cb6HnNaLfQ^9R>toLgRLTJ#0wqYYgzLJ6PTL6 zd^3aWXy@YcvCc#ATumQ6UZVO{M##5zuDzwx_| zEL#S$v4}amPg!X&nB!Rh!BijZ%~|(JtmoFyt)VT!<>F59pwo%huS#Qmv`GI}G7GbR zuua1z((YO=KFoAy139#QMZbKvWIg z@brtae4G6$={_0$%yfYiBFKeh4^^9hHp0Hc@)T2h(e5_bV@CVg0qCl2%C zBDP0lJcnh^j)1d8W(?mk?D*BAPPHZXt#r#8OEyXghgDcB6`5)3V$VhuAOScb3=v-F z1{foV;cc6;1uUS6?gz*M7o)DDy)QQI>PnNEgofgSHycUpNUWe7R!R4miG~!4fqsxl zmuNMKUs8Axk3!9N6oO>{o*Y4xWk$ z_Fu#L4(ofMCQG~8GIx%RRMKs}~YA+<0@ zd+&zp71u)sq0SQdVwyBD^XuK>1l7Wvi0B#1{my_#wmjG&NEjyw3~*pACTzIDx+EFF zIHit6T%;_OR?gb1R@6{Et@p5Hoyer1_6z-H^=u+nNJ_2HgkYiQ=oGC^hF^3WJh4D+A| z_SlpQVu#3(QCHYT42?R6BQEZuTMk(I+&L>UtoYYE!mxG-tkx#aJ0L0&L-z-OrBX3> z!UV*Vg|Ts!Cz#EFZC}TjAAK(`z@&N1O!*+5WU%vnaM!B?ap2)j(qf8%w6uIGa`UVN z{-EExO)gjFZXg}%KN>${CRzgInr(iq%cG92qZ??fyw2vAxzr1XhT_OfsO-l~?M;yD z+BRnpXv~gEx@o;)4?jPk%%y$qXP)^1c9l4`!tAM*YW?U5cI|NR-F0jHESu}F*rcK|kF+?~MHGC6GC)S8Kn z5{lVJL8G?F9{vvgIpAYnIk{UJ4B0y^E=4zQ7Ad=kF9hv=~J@|O%AYD z8o8^mh#QGY&d_&U6CfK!2W8*LgVm;dzmRl7-vI+-QK|Hna;QKX%eXfN>uD(AmO#rD zZ+q0^vVfa~*ctWXxRLk6H?9aR>o2SnF7$#E6ZrFmyXoqPvf%hvu{hyw_zTftlnqts)Ike_?H1o z4j0Z^%~n`hdSBIhqf|gDM64mxwkUV1lVu-?u!Ds5IN32BQ|}uVM$J!?>IoRs$TnxB zLTXdS@#^V&vkxgc6gy?b4oiL*Gm3|##OPog^Ho+TeyQ!heyRvI?MkZ?y7S4N$xy2~ zTrDW0n1d84p(5A0q(b$0OE40a#_+I@wP9tD7^g?k8K9TdLrrfR?=s|NAp=#S{64`Y zG*K+96x~u9y4s*VVPN&g+%dz!dY@%M*aC8E=xNW2x!ZtZW2p#-+oiT*+{-kyTxLJV zygE;AWVcM5&FuQF=*K^{bkVV~syVFHSaIlKv6>+UfSLELw9)%-5540FG;BSv4cfNO zX1{Rm%)2X4_-L#_>sS%cptC*1e{0HRo z&YgnT>d+@#klD(_on6<7)0H>w_o+zTlRPYq31>p=(`sFE>WdCx{#>3N#I4>Ip=NP2hO4wb>Z~qfEyN9Mli8=cz1PEpk?1SZSV! zrNYRliVm|*s1nBVmlKE#cCEEWSxS3fVg-#!|M4yPK1+Rrv(T`j zL2L_XnZ8U^O05q$B5G5VQu*q7aXi!Mc@< z1oyet%3|Fbpu!8)*;skuqEzojPvmOs2#MjLX|@X7$~6Q%|M+F8-U-TMYMr`DP(%(3 z%jsCx;+e=xhf;2(`xR;TY@7-2OL&>B^*tsKKH1aQ9E^apQBHN8`AN|83wXF2mm*dQ zcv6A38#3*AXXPIYx~Hw2Ujo7>b7tGL7WWirRs(G}l_koZ-VdO?cn?2L6fa1eIT51B z_9E6{D?ZBV+3%8dVf4%Y^8CSfOoAts>i-%!#o^?WPt8Qh6^d!5fE_3Df`M@s*k%ji zAWon~rV@j52h01=orBs-nM=RhwNQNFt9(Co)!v$u1(^a`V+B}FK&~u7xdIxC!YRA~ zV@w4=Yok+bRTTzlTWO>I0|jA%@M)wH_bp-*er{CMT@PmCPo&SmXHmKMq71zU;FJh{_tA3N$`BP zBqx;=jZ|C8;aDH2F~c(9KBSmRij*Na4lXe`hx=XiZg8m5R`D={e+#cPAlt88ti{fG zra>w6#}X~(Eug%r;(=(uK5Jc5s=W=hg~)zuL8&%+f?vyvWH@c^(n@AH`@74wi z3yP8i<;4Zib}L{LFg!g+sEoRdY_RgI%y{mZQNMe%i`FI1r{7+E($$2b(tr8t8M2n0 z-ppZqWtrh?E5&S~V0{2z$bSm0gS$wFqI4eCl0k9Vwkhjpr9eX5pc59DO>}^nHEu>* zY&%|?mUPL-lCJs%@$_3^-7aLJe4+yf{nCd*S24kfpv5c^oea7`KM-Pa4ayOO3YI0K zp3^Ar3Bi_}NCTG%*DnW`1$QXw6?X(ietSEv$Iw))@$!sl8JfwyKg*{aU;1s6G+FhB z!}rKZcG85yhSV)HX>yrjfUWC16}jfEYr`5eMpJ`ttfX3n{n7^4Q(mXQ{_c_H0uyNp z9UrmJ_r3y*Wl6aN3!6Lra{{l0cY=Pso`H-|FwjvA#v zfu{l+=Jt{G%mH{$k=(HZm;~@`#`Cs592@_wR)&q?(OQ|wzNf9L2qPIKoXw4MUvl>= zFPkLLjsLp*53+hPfu@7Q6wM@x0q5~XOqH)RY4GiM|yyPPXj`=)^lC1b`u*&59n z4ICbwb9Bx`DLY0i*#=pYf;dL_jpa60kbUXKACAnm%%*)oY~g$}iMNwtAR%g-F~XQ1 zu}irc{v;@ie-Zsfgb$6%QF-u({Ni66c>{a6z4HzZh~afYv1(4h7Uo0##zl_SrCC90 zxalz#oUpj3bH*QkV+j%~Pxlf@P-6a2dBA0{VIRlVCR%7p0PfA(A^R0wbhc!vXqztv zY`4g6g=E4*FAD2`PmX@L?9xVtJzW+LYD`1fy|A3!W3KKA`7v!$I(q*VI=L<7aL6## zOg3zwm{^LeqarT`t(lj~Bq(po(?weRt4&!>?}7%t4hQEm$gxotJaeC;Pq0Gkr0_>S zIcRCPve{lZ>^iezf2XUH=ch<+&W;7Xzas%Ll2~Pm|55tlj~9RP`suGXytVeNf-e_; zCB=UN0c9hSCR|gwn(dlmxvQ|j26s(z;944PY&}su>ZI8CJB8N6hF6MAUU$VaOg;pbN=0nGqk2%<#@z&0qJJv zd*?7n;>;Q)FP+2;Qsfcpr`SHFMcHCtTe{(K-UqZ)V3ytFgTdpY{*BUPUXL;_ES)Zy z((PM3@1!b5axbJeJWpIEz^?LmZP2MU5a-DYOHl5n8$9pv`}sPi68!b~0lVA|`scy8 z9j-mfU2b_{duH^CF9mL%fiYrCO@MmC3)A+vC-S%W-ymf!PbW_FN_youuXpyfP?T$! z4|5d!%^iw3=duOLSi_q+CV$^VU8Xh#cZX>7UZaH6+Zm= znl}cVlBj~O-(9%sOZr8}7w)DLzZU(KBfr!OA47oUto*zpiPtK`#@G|IL9t(QM@pNg zbF)%&sIIGf^y7&rb2#6?N<=wDf)-ti5>rs>ol_*2#bpa15e3_Xb&A{PVgxypUMh-} zV`5yo8fqqYg<>C{SfBMwr?@A)-dURlvcs6EkSyzSOrI$B=z3OgJjI0lj&^+|ykfRV zpuAhAsUo@Tb~O&G#mCIV&tZxwrARSw$SI;dmwMh2yxmH-kSwSqf!2V6Zr2q0A^o;a zg^b62;YTHR1l_tp?yZ7JvV{JH2GkytX5BNL_sPXGdAI$5|^fJLmzaZo)+4{o3 z2hcKX_sXFd;LlE{BCkaBL0iIhDAe8nX||wseDkU_-Y*{jQZC--kt|%tY=dskn8)9w zNeDH5N})H;SSmOZyj-v>XfrQ~r}u7C7AwY&3Tt9uoM^E&?YPhPAIX_wV>Ae}5e>Y; zs4hm&ST&Iaf5U!H8=mB_(Gz#bN-5y(p5DH>^M5S6C3Cj&EX_W>o7`?=-5~R0%=tO2^vE7ND~Mu}?Pwi+dgUksV@RySby*GEB(E zBg0+D!yYnxUwJ}Mip-Fh5`!TN$d|>YKCLuNmqCGWmve%$3F~UG?R0ctI{ncUuobWZugnEM>QoKF zqG_3ij5TP*say{(_##E0tVD**i4Dbmx8*tX>WB`Y_5{{GOc>0hYZO?ItW9={krasc zf~$V7pbZqu@oeyH9{;hK<9GaTaCzA#SXcnkZbQFW-STh$T%BsG12U(@@PJb%3rKg+ zeKN6WD^FI=7FqXX@B|(NzV{7}e=bB|Nhwr@p(?A*3CIgW=h=96v9T_eEr=Mu^Md`q z?kBT8emI_%P8{Agth96*qLe%0tS$6Ep~{v3hq>^=lsypGnQ&m(gbGlB*stsO?ueUy z?BCTUC)lyb_cxNl7sd%zZ>CICQ%nU#4pWgayaDA|d42#^l4E&CIaHDBRBzXTX8nol zYUSJQnl_IDx|?owxz0Q7SuU=N0B4x;tY^FH>iNB7m)jn<>cD-98qZ6CRdCA|*<-)z zK&;0Ba4BvDH8WZKp>H@X%59|2>D~N1VrJoENAB0|d$;)lyH>Q|C z)corCcgZGpAmXrSDKrC78pVLD(NU3y)!hWjwIC4F2p~W=%HAnEee@8VtqshjcY?ju zEZgpZ{S$O*OvX=Br!7cZ(DrKEs}mr5#tM`1!r*!LSNz0j->*$D3H;3~uaI_jFyU~# zX1Q4wLNCSiP^62B1hYM?h~61~FFb=^CoJJZgvNCpbDl&)UEeLi5%(>!Zu-N8d2~uZ zHzYc>h&#p6O#7^K`fVJrV|tD>CU~jfD1F-Vg7`csQ6C`%iY>BwISlCXNOG=MR88xd zzr?xAt3_N*8iG4C9br0k!|PY2*fdqAiW0?pMKkXX&2op{9)K6N$kP4)IN)@I;OLW! zhDJBJB>Pw!oy1F*9P~dUKBj2W)CWHd!QTu2r8?DdQY(2P*TxFGCxXb%2Sb32n{q1D z((m$uSj^VkNAhR^7)Mx+oGMA7q3<@5tiD2bfSgRMOy2F*rW}6O zz_yYtIptpLg7u{rJn-*9C*-``MS6kM7YLpp^q5JvNOh_ho_56?EJuAHOIJ5B*T$_~ zIS@rdfEd%yZ)fCdE&F)9ASklZ^#gr~{jNsqTDwDibM`6qBWZm2Lnv?WcU=qQUDgPoY5ZWZI`z0N+z@mX zTjnNkHacV_+u*LOURc1SGFM~=q#67w-t~~9^xLQ~;|a`KJ0Gyk9Yw=w(LFgzly!oA zDkwmQ+G*pNwPfSVO`104fO1J_T4)BpoY$m5N)^0+K)DV`Q8R&*B}cFhis$TwkMRS- zViOt{44&T1%K1#qi26yBWp61d5&ZV&zli+p&wl;(|43=ZiBKHYo2@j$GNxtbh;~8% zVm-50aq87H5Gy7zvYy!rEB zJ7&|^{eyMvd?WeSM=fIkoE0-xy!*yx0Y)ndsCSI+TnDf8h)ZPC@{f3@V^3ON{*mYx zI*(qib2eF7N54?{K3UGr%HnYNd8e5%zKLSuDH20PVx$>3;V?>ydT7O5Xkv%~qefLW zkd#g^l57HzVT*ArXe?)6o&FPTgb5a#GZm?%m>n!QYDzcDQFYNL- zJg?ZLhQHhGCfUu)4etY9uNFbIAWD=IQAwX=@Mrs6AjR=dp$|y$?`olrp9%T-`}pYO z%MGui3qfnSH#?qRElh_ys)a-2YxsJ_rvb)0jrTNZE{3H_8$-HX@!Q4;Jlla}p|Cvn zqa`%heg4X4n}7As#R}IK(h?FqBPkHJrff*5)uu?&!5Y3T-{g}?pQg37yumqW7@;OJ zVd928I@qo9SVs@Ma`)Zw>^>Y;X|3!&xYJ2$V(bC3C#+kZEK32VNG%fO$G=(_v?4f9e2HLd-Ew-9?#^z9b`dyLPq=oSd$&ir zdk^2(cul)87_aJ7c=h6fN7BlOWkEM*FJ&tI_qgZJz2ln2%b?#gcJyvy^2FMW$|a&o zdIjlsE9Vy}+WGlBygx%-LA3afH{NRa6>!EJ>7js6X0DyNET|GnB5}ras7ZRCvvr89 zkoVG`_paYQ%Y?i|Em`YGHiwZ{XQl}sq?i(l>;pDC&ntipWFAGj$1Gl(vQ1SYtAX|q zYk3=;2b|XOD(L~Ia-UR>qs|XWZQ#vW`~8|UjlAe7iF|1CuDAp)|63tdylj4%OFUFK z;`!Zfbpghb+RNerz}CgU^W)WchYJ<^iml-@kMEcjc&OLz?G}hk;CZ*`<-d@1?2IlB zYeU&)P}@c^NoXNPUW;h{ajNqTk}au>fIe6t)k^7R1~x)0n>L!|r(CZ^B+GKCT*?mH zn@y|98ca6LI{rcLc};%(^DbKo4B=~(AE|;)Wxc#8yi{_57sES8u%!MYzYF3fO{Ea) zru*KyD$V3A3yOOKe`2fo{`H|n;dflSL76h%uTyiv!p&8i8mjAreC9=%$+-~+##!}2Wimaw0JLRk8x7VEI%y{{%_?dp6AXQFjA7y^x7GQq4KUpdMcF=Pu^ zCq7>j3_4ah2M^X#GMPyO5#JI2p2U65I)UVwiL_LT0U4K!of_JdoniSdWiBhAYtT^e zXwV7y6f#2306owJaTTv*);?9AAj)gd35!N1q?XY(*jVf$V}vN1+pNPTzE640vV|fW zjBz+%WToLEc7-W|J~1g0L*yA#ql`_h2Axa=1Q>K~g*N8MpR7f4o`xwk;PbVk5;F4I zjR`#frjci}-`W4S>fA!hn$}T@K{u(5C0p^qSESlwQ=m=3oQl9?P5jG7K7EHj#^Jv> zn@2kQK1Uy0e<4SR>T*}<_)34i$3ItA+$Dn9->t$n^tbx zDy$G>@UKG^QYYQcUm{AF4SIKjd%3VvtWBYph?Y!!Bu!FdXNgA(9rhlu!Mw#*WccVP z>@WD!(qAi0+Q{bkzzZai!_KsQW|~I^#q6R;Dz=8u`y1LqVUd|3+h8fPhrfnY1Rj8F z*xMnyLvM%ZWjNxU8iGB6QXw<0d{MIOIJqo7D?b!az;~n~GQOpE)WWk=TPL~do1-k* zRyZ3kveFa)nKY0UAzRe0tQG(q9QL}8x@tB6V$}N`b&=KnH)*N!yRW;OB*%BFwx1^( zIV?E}%p}JyiUFeJ?NlTxl~^UV89bwgeEc?LnRjv?tG|+^S!p?%FB=d#wx7V6ss)S--dn%wJYk(;p z%g=6ybbE9N6W!`*{6CL;BG&Gv&p0oYX}jnGTDv=}55kE3u2GO_oy*@m1MD7TE-m)V zn-V8zQ!bY`Al4v(4&av~Do|V|I9FvS(qISpE2WXWc1n1@L|NSMgHD|oC7~| zlapH3rhDdxZFXr$NbSfsV}BU$uvyA~D>ISrCa{J0v<)7tssP2V^*xdEcGW@2WlJvLUn+ zS|n<-czq3gMX7WX$m}c%92#+fWt@gJG~x;Y!ocvtySKDjdg2KD_n1->*Y|e z;W*x0F=s{kjd*%Ag-w}%e$N6E&>DZZJe`!X0~&{Yedo=9_7TM# zp~#0+B*vUE>&D>eYmm1Jv0prN##Ey;Lz#kK6Ih7?-~8kvP3idY{Hu@*oiQ6gZD6R2}p0g;{NQ z!yNYAqfN}p_~EXL_|!42;GauSprv2zsh2?)2<*oh<(4JYR?LkN6hTm_O*z3Jw*lQI zJz_L0+W*?IzisyUZ<8*vYIV#!()q&ZB5TbIw8s?FPm%i=wC;C};awuw&f8cCtHm-{ z>_CEL)G?B%pshZ2G}_{+&Uw5>NV~){kb)_06>geW4)#yKt4_5oU=5>~!N7iH zyg=JZ<5gpika)LppF$5%JAmc0T$mAx+DSXVMU83Y5f#%f2edzFiugCT7=u;W#OZHLoelVa}@oU3G( zUX~u37Pdo@4BiD)ht@IqJnhg|an?keRNj5`6%+paepmK3*)kb8_J><5yD0{OoIBCy zU*^&(%nw*M|7`F8=>KQLyZA<;RV`Z8%`$kun`H-I)pmq!55Ni#?Lqo(IIhHTf@J`M zPR2hDd-H97$44|7?B+(XloBt<4z^b+ZnBJ9SqTpndK`y$1B-&%L*fKUq(F&25S>S# z4D%q5LNq9s*AaqT2#qRqsT^?aa=xQH117D}3+yn67?)=3K(Sm}`+jBx6z;k^4*vaI zzhyTz&c>~*^jE_aJZ%SkM%+hEx`E|;$hl3`s7h03fyD2JUL&Jj{Xg{nEb=cu|J{H8 zHqyw@L5-Bl@zc}adjBoc#&hJCmp&ymFO1^TYPNqiQVj4@of_)KX7WpCO7`+&Bv9@I z-OtLrF=0=q$`|E}v=^BrQ}lFCxK5SiI~aluZY~Gc^Nj_L+HEsZx^!zmEQ~WyqYf$gmOwZxzOn8_*3F zBZvL2+dVLlZK%E0sS}hK@7eCrLErK$otFt-ek^Xq?izc-UHh z6+SZTT!R(;xKALBP!=S4BZKe8;MNeV;m74iSY(U3oSS7RZw}9lf-V%sd>u#cEMvyW zc!BVo``O_n^~*kamKnmID=s@L?3h?ZfIcux)QS?Ffof`CFGD$kPO&nOE+~xpWct~_ z_|Pni5>*Py1h?fzOCLs}=P$+MWEhg*ISNQ)2GEe8c@7Z9df7UF{!#oOc#g?7`o+g{ z_mBd1wh?z-1LU-sZB#`u?uhdxYh zP#mSJNnvP@G8ZajbHhIo_JwrQII=9bUc5S@guh}=x1dq8pPcqY9+5QH`-)G5O_~Ra ztJ1AL=e%&PBG?_b$ZGhfT>7L@qOw=cYU}{HaiePpWoAFeim9jf?fXT{$*-C)R5az$ z&&W{@=W^-IP;s7O8Y$8Mjiz}kfT|G0=o^2bS^=;)u)oK7veSS4BSnB zvhb?(IMIh)lNJROLp#qV%~qd!;2!UDtzK9TBRioWyJ2AqWWpr*-Y3R6&obDfpqlK4 z?J$LY?1|qv^lfz!XBkLf7xyZ=NgHoBeb~QZdcQK6VFw_`V8skLLm!?3!G=y2UevbI}p0IcDKjY={;Eb@W zFt{hn54Ghd{{~})$&{SJIuz)$YB@M)%{;0OQzqZZ2#gjlp=neeTmM>~@~9qlO)DVfz{TSFyU26>zC} zlbByw=3{ZzOt(T{QV;xR(KD6@Y0*8C@6w^3PyzOIacn`v)6o-m#0n9UHr@N{Ez5i? zE991&Bw3(d0Hg9Hs53#AdYkeg-4c9;?t$Jj6ZIYrJx5!fzCj5bx?zHTvU>X5RWd?~ z6{Wv3Z6vSnsr4eCurs~6` zCy%7rE~}_5ywZT4@HBN)(0;$Fuq)DT(A1&BIs~e;H+Yw&*^p>{L*51o8BS9aesT)Z zH5zF@F4OvqY@DpECj~oB?qw)BC zxT~L8^ZTrIt5q9BYo9BYd*aTvTDcYQ(Bga67dpp@x}kcjnYr$0tp;6ya|fc}16*-hh+K9DR>0!5;xcGKVl!rcRvSp*G%EH0?e~;_1B;fv5PMqDQ%(+$Y1o&^9H&fsnB#plpH?tF4>* zx37JW$2S4Y$=_)yiRLiCfURO!;avj7Y@kRi6}e1M=c)(xVQi>_0SZ*=ioz41SJz!g zEw*hnZF;me4YC;*HqBNRc=y@=8RU z^F5bNF;?yvoQQh5D7;0f4@>jymNzWg=njs?3B;1kFq&ZV;EuZdS1u-C{7rlAf5<8h z3yci2rQSj@2^85tMOFti$a8s};uN0>aR(iTiFDAhtH7fbY74Sm46SyMaBt#Po84kh zlv9Q+!_8ZX%iAqGcfBC2dsg~JSp1KzV5?1_rnNyM671(f$Ce0?FY?uE-#4x;hZ2Hc};}H@Z&W5leBx1gFRrPG0V&{dsNIr*Eic@APMJ2_QQRE;M z86_&9my$I4sh%_|NzZDUG)tNGIRFvI_t{*q7PnR9u;^M2=fpZB?ZztHp^!z}>D zy~t38HQ z0xKwf^HE<7`GCisYP!udZ=#q4imaoe4hJNv*SeR9QbHS5cjR&IdQ1`?d7xvzeg1+f(T<0(-4F(sSE zQ$sW{gKouh;WAd4EPovzp^pVV@Y)kLblusD%&0MA4#v#PpfhLiMXvDc;ztflOh^ z%4B7{cagWbQYcf4wL&ox>=8+uqQ_RTRJu)_jk&&D+2>F=)cD$ZVT!xHQIn{y1_ljN zSjs4=Os(@ZMiALs9ox@OFv0nV8;H&imj7V36+~__?Lm^sZPUZ!ddUxMAXiK=g%sIG zMPY|Mw*DZ_ux2ejIM*v#VbJPVLMks6+UP$8I#jc)RyV}haK+LO-Flonz8P=ArOkX} zdGLLL&HL9w%IUPwv%XHO8*nbva&z zf=+o6y^(w#uxG)Id0j8xhQ^Ll5dXJa%MjLy{7uh>sNnZM4yPMW2$9NU5`=~#)u&lucF%7XL5v8r?U*+1k^t*}%1!_YOO4(}*- z?Op1d)&qqDWMo(a@miJtOgB|wpdu*%(NR= z$r>X`lj)JbCEerjEIeV*g5(HpkC3I~46yj<4ccz{QBd?$*kX0lD@EnfV?pbfv``D% z3*36v_m<)_dGpjPZQ_)%L)nTfXCvA{z&`gjPKcgxs8@GUVg=4SMN|Gr)^hWB`D<25 zt_{q#Q%n;6vQfD*)TwXO9V1nuWs>#624$Vl2;DeG#TP*1z7o9HLRrz2LRk_pDmx7< z&zm)_V0zwo&blu6-iNF0MfGwH`vonH+DQgZzRl zR4me?=q?vQV{)5Mz0y*waQGEeCP8>uB(boyBk`*(i*qHXRF?%9pTp+V@$hO!iK$T= z9!DRXee|CO#a5H^+J&tjlGQvmIlFC4PBO)8rO0L~3TqufxeNqk4VrzCP0S;=cv-y? z_d7#-qRoDH#<%Z}+jKiRS!2JClL?v7@S~j@Jgwk(yO}H@8@O$ScpUuSX9J~lih)x7 z6e`NX42CJ|0riKzO$-94-ve`${o&QJeviL-KsVTFO|rM8L?%I$Gc7N0Ty)X6A!^8_ z89QR0H^P7Z;rHLQ-w*LNA$cJnwMax)sU zKMkGXVKo}1U;X+yvWA<{;PD1I$Hr)+P|P+8<}3;;ud&5Tj{?7IJY&6~)2mC0d?!^P z9IP)2$403&*_8X_o^YgJZWFm(C93MdaY1*v!Wma^gi?_6pcy8$6@sg5ngAN15 zq)=oV6@?zCezUA}VQN5%+N9_UST}nR9F!dqS)o0uXvs&j%yC^LNY0g6m~qAhBHRHo zK1G+GOi{h{11msk7YBcboa3=WdfR5oTPfx$MJ|J0y=q+qmcs#QKr7u0%gSe90xA|* z2u4xd;fnlr@qHoYIh&=2 zBLGjRfY{;UGTku&g?*mKe1#iOrk`1VZIaa*=}hnF$ue%s6_1^nR2$p1fnwq*vWAMP zk$))pP=d{6<~qy{ExIj@9Eq^UJWDu$V+eA}kK!~4bN=XuuM9c3bN;xJ{@UgihFtdw$=zvIhvb`i zKmRYXotu-&9f+b z$zT`}atI`>dBZw@5&n@*FBp{~UKH9zMwXLLMgJZD{wJu~~4Z&~d}!^@W@kc&Ky&;kek zFlLJz6my*-Z9vX77dhBaG;S{tM;`^fuw{ZZ>P}@hxdvg^WM;s#QCIZFRr!O+uBn%U zx{2j#ySgeOOL&7c`j!b%xUNyR*>AZrQT;K=7915<2~1Eh3xr%>zA6V^>IPtUdopmV zv{+sqxSF~AashC5{kcv3{#}*r-~~QA70HSMI$euY*w^IUbRW4Y&j%js6u(ERLP6Km z&otSxUPaD~>mD`qIj@7>7vwwH8NWEA$8PBv>qj=UadK^_zv(KrzUbvC^~#SmX_GN* zk9xaSTCP@TB|NRcay4`MCsT_n(zvk+a|Q?7Y&!WoCvZ%tIkvs?MF+?GX8( z0xm6D=WP;a(g#2QZ_sUp-_iNYnMPHN1SNaN$0y;kop+^|`ZTh*fkV*f;qkI{m3plu zXD!L)=H>EuQ&eNKrX8l3gA_SHMfE|kfkAu3zswy~Supe02xJ@e$~^gzz$S%3P#kCy z-}HsOw^5Lzc@*>!IT6$et@0oG>OptROq0JhqMKxRb_XPi2Hh;8;pKGCtfO8az9Php z)^QMUbSk|IXFK$qRq1KajPA^vX|KKcWh;30%o5xol{|KFuGqk#j$%Hh$SEqyB5c_~ zr)X+GT_C-y8KiL;wB%SW|09$wUs?yUGxDN^84t!lx^PdEM2;A*|cTT^f`hw zQL7l`Sxn*%73PI7iHvn|_$O9un@L=-YrYiZGH;3iV3r?;iK7bmIOx_Y1XWed8qYFF zL31ZjcQ!~^F~kuo-0V^H|NQsc_V$RkQ9Kq4bJf2ytWK5c7cWSFdA$H$uYMPOTyj)= zLZZ)HVn-UxA^_U~5 zhAt1@t9s~rQ*_d6qxPi+^5H9r1YN zRwF*;d)<6?N!L^tx=dpO)`*yJ%E@WJJ@Gql>(bTq%7u5xW^R5DfAb&&_*sVS<#$mG z@EIDYs3Pe>Z93T~IzP2YtuGg@_sa`J?$g`z6DI5PA^7_d$VyiWS~dDB5l83Odmoo$ znI@Lsf&SREin@x4yzZz;R4ItOyIUSKnpi<|9#Xi|0n9#|NE2wTJn4O3X1uFBGKcVZGHML zv*QP~(-vAW=5e&5Ac$U!poH=#RG7)%~^XBC_WpY z)A2JYW}+ZDg*>d5el#;>*-k2j+DPFjWAKGs4e_UE-LW&AAU3 z?+7T|X(4OaJ|VImJ`t#^q|2q{z~=$khoW$!23W_M6lt;&Up#beo8p*&;@F&z>Ml+| znfRyw*z{HVkRTUFna5EdmgYy?SS6}av~m}Hk3KNNq|lcvjA3eL_CQHDlEk3<1GsZl zmV1OrJ7cSOxw7A*E_k1?bm3=G4COTHb_*`gI4dj) zFZR!o^ufQ4x_H?>;X1!Wb&9MS+V*>9xp3F?cwV_;#Ozf)e%f^hU%mIA^X&O9`RZR0 z76!`VS+l!d- z7B?)iWL@fuib{Gnoj$W!`Vq;IV{eJGXWOAkas(0%4dmpTI6-1U;MlDbjYRb}C@okf<)zuJtWe-KJYT56v}b;stsPn${~(_vggSQ?m@3 z65r1i<-x#yL6^bqdcXr1jx0J4tdq4BSM^?- z`n|guSwiUg5;qOgZcexTn_z)z?m0x1;{soN%HU zH$w5)<7DAI#%3X8|FcqLv;?PwVm}bl`IrOsh!pHmOQj+FK3>en+#&O1T3va>4w+4n z-)^^GWmwE}nR_0+fnZ@`t_=P9vgxh6StxMUfa6I>_c z&H1;hVOGCXI(cyh+0Wy`fHO9}M%MRR(b(XB-G$ySYs#Lt{?DRDJtwl8q==CuywpW6ZPesS#%Uis3NVcpADo^e;SH3% zh2a7Z9`Afv>`}5Hiw3_nL9|Z=igehg@t0+iN2(1GgKqm&38Fn=P}+|bCzvLvC9BL8 z{Ko?IgI`V}$iu$RJ4;(Z^bOv(C&x)z#n6^?Xm0kLyQ)NC8fYu_PA?CFo%3aqJ;$g? z5UtkLL6rd*ywZjF;eGR3J;$;Rjsil@F-5~Ja{`U+&8PY{PkYn4NEvQQ9+CzgFH$`= zi*9BJ3lLkjK(DC_jUAVww9DpeeN6 zl*Cli>;3jlwTxJ)SSjj&>d!>AQBw)aTRFWXcJ`*=C9n>*Yc~6J)5wa+jxCoWXBdcB z4sGEw>=U<@?BVt$UbGcwec$t|B{4k4S(**bHd4%bimaugDgd$m!FixPDxM%yj-(6q@MXLHga!E0_$SsePJfGM==mfnrYO*)iWeN6mhEoPT=TgXQ8c42yjA=ceZgxiRz2Ik^a}X5I8p0*lwrtpcMKxqxrv3MCNFq1m!ei$cbQ6cQ2xU^tPKu;b zQHX}<`KhuB>5iHEg&2Q?3VS5x+~ITG9m?_1Xe^VI3!zXSxN-+P;S`qYAJ)QwMU;Nj zmcfA;bPy*aTL}loAMGtaX}u$~Pbj}g!AG*g1`X>eW-UclQBfGvH#2Wwn^-H2+VZZP zUe?#|$h@;Y?7|D|u1Z2>XIsq#fdTSy9AU-|c>_p{pYH$m4X^#CKI^ zgRVHUMAi|EjBFT2I_o^puw!I|`~>x*A60Miv937ApOq{nNj%=;0Gr0JH6({(vM7>4 zMOBKc=|Z0z|9z3Se6Ptd`mx>Xj7OucQe5bBT8eSroE_R67}SxAO*}!O!5AzW%@q zPr33d>>%m=@E7^ycx-R**-vx&4)Zs;vgh^r*+Q#vaSL#ZA+eK2Sg-i|`@=#sd~r&o zn2i)!PettsYt$47nx*=x8A(9}g4Ngrcm`6A6$=L4a-GGc@@#|*yUM{x413D;hwLCO z+5V#++jDGliIeacF)WPSm?2u@nF`S`*yCc$I@I6Kc^GmxB#y!^5_2v75Lt$Pf5Zl_ z@rj^tfX?_BJ}WA>d1N+!5%KbT=%N**$`tB8Tedzzy*%f?=c$)m__@H|5yJ_@}Q22Hc{m#4E2`E=m>E&|=-P-7sNMr+SSAX`Ucv*%g(CC;Ir7)!88tga%jwjY& z#3PqrgdH3sI29A7U${=&)1k1ijv%YZhr(6B2C`4}F*y-_L|YaHl=ef3`~;Fox6*jS zk;U)EQ0s*-Zz%0$zS(*vBXnH^mW)^wZS}Z&ZBkg6*vx`(*g}{q>wqF9u>UyiNF;zi z1=Du=j8No@ObcD59#tWU6DDlLS+qGhiYetbVZ3b)3|q6~O?z*Jx3&8${zs=zd?XeC zuk-Hox$57p!Q+Y`g9bzOsA;p?(<0E0g-+;{U8a5$(%N*XGx!l1-oX$ zc8&%*Q@VqGA_GzsZtXlpCpcGzOx&41CK~lKbL6K)kQ>4@QKA~hMKUIHvItLDKBBPZ zsaaT>nLTIqOJz(hq>+vV){=EhwENPfOY^CG$|6i+4B8%QR1|_ny`?IvNpaG1$MolU z)!gxDVS|(9jTz!DZ5JE^<7A>owH!Y4 zO+ND?hv|RjIL^MvZjN66OLCdLb_j3RCKf^OTv?fW2i-gGy1G2LgKl;&o?%WI8)5!i zX_LaJiIyCa?*NMJ=Y4yf_}V!j!ThCn;*WmjQ0zz;!3Umb{4({r(z-C6`o=$BAse|Z zOgxUWKrDY)+Gr=mfWYc@%ff`K57fhjriFs1eIcyb9cv}g;Te!NhQE86v;)OXIh_}{ zB6zQ~%pH$SbU}d2Jsz%*8d^-Ai3cZKeEZ`!6^pDOG5+iFOj61XBs|UlU9bVkM-+3C z0v4yJBvSNp(o5yinusSyMqQ)sUf_qq%r_d~2w9jqm6+eDf&!;#Nj4oLN{_tlUL$}! z23@bY5IEqmRx|)!P?qqae;%C)M1nO?7`H<7(7&5Df!Ga>VgFQ%IvU*T-dPVtc=uQU zH(D$WVd|NGjs;UR6=Scd$tAGMV^0`Ln}!*j%@nhNBJotze)kSwQn4uBZH>^ksC&b) z=08zw&>mKSM>>{S<~$KxV8MCtF+A|7V7kCw1CO^}3X7^;Z+L@vx9&mYlE()|UAgog zxvp!_Rnvp+CltBN1y%tUm9r$re;j6?%bY*?+K*u}9ek9Nr$X)SS@CmU>mu~l z*$E9Kc@hCp?_uA#T#A7<=Ur4(k2)UWXJ=(apu^p!)E^71SJsj)h?YV>12$FU(aR<; z(Ia;rcG{p!i#~m3tGK|YRlJ)<<;u%JTW5nxMo0qbbb7+x@^VHuCr@SvjvRc;+-~DE zC;YX=5&LKV&)-*>wNZ0UTBE5_4bcjYv{;lM`p=2o0KxK>u+;uvWX>>x$x#JchwO z!@halvl?}0eQP3?T=O?gN3G0c#WiW8u2qat*P)vmb(XF2J<=tO7aV-~lJHa2T4)bK z5vnXY3(B~$$tO#=EVM&q^h)uz{K#&lF=P;EMbHhtAF|B5QI|fmLu>%sb$fD5z#RzV zRw@SN{U8I{sOwY$*I-Dz-~$2v+#cbeJlXpU*&DVov{9D~ca_q2RsAGawnKXcR62X0 zli-?km8cG$#J|jd#}1za5&kvGm>m(fnRe!`w#aMAFStS0X3=%&hebc#A+oQTB0m4`xajT4H(oF0q$$Td6MyuEsNtpa}|&Kqz~k6-*`8giWmHG=q{rU z`34(wF1B)`hslr)a?PK=SP2!SR$KM=p%HR&JNkGT zq>#g~&(oA#oV3t91)$gEIkIM!?0|a)9!jS;I2!!I( zr|*(sSxNL<^gdg~mWq{3s9iz6r!x00<$&7E=wrFCC#(wE+%eOLqRyRRIat@-2&GfU zRVY871`*0S@!^O9AILsJLVM7yTb>BjpLniY9529gaqbS~vmHc{leaEUf##_b>=yEx zDe2BwJXjtZ9hTex$|Yi;8AX}Pfq1S>tshV)n|IUZGN?0JB&bWB3;F1sex(c1MmaS1 zPrY=GCq{=Ijz44+w@c>z?3KT-G~s(63Oju)JaS->EP)l247;>wH5i|9_F)}|#VF=> z43GR(`NM#*$jR|ILc$^^cS6t-oEDm8PF|fMpGuDfZ3TL+GxB{4`#n-ZFD*jJ%5v#? zzi|YJyWJi5`h zn_i>d?3dwL=3b|)rhj!#njx(6O7}{C-Jr4D_A3`6W|UB|Vum{e9DSnVr8O~gtk&ec z_uG$>-Q27RkD&rJZo^_|hbX3)0xkEboAYZICA^vyXw;>=n(*o)wGEitgZZ6SF+Nx|EAQ3Ph+6@bP?2kU$9+MbN;MpGN)!R zI*-Pri0+W4$TkSp2)4`|S1*4JSg*(P2VuJ zEDYP1x?I%h`;7PD(2EA!`ivo5a*6##kI(?Y-!IyfpZKKzeH8=5PKu*eOlpJmk`}A!56A~16M&HRn!KCtBUj~rZc|57ccFA3!POaGN6~KQ zRk$4gdHE+^wbxzXZBZMG?$ck})Q4fZ(Py|%Zj(yY1K(U>OURbj637KXue3*)sl87( zsSA86Rp)}oYfa%i;Vv;|PY=q;yiv;K3m17>cQJ1_lO<%s3zo^*XR}vHrx@@*QmCjr z{|D+0y6yG#P{C#vSJ(wT!Ps7$J?AVig<+lsOS+NPzEaVpH27r&VDNn$z{43gW{dIk zJUjagJ8nMz;khz<4ldrB5Ej0z4tgC^rO1J1ZvCeo4Uj)sDQVQ0tNIfn;)BLz;)hKb z2Mi2*%Jqlr7#PVl`rGN;SbIX9r?)CR&c(5i>@+D-Wtnn)5%?DsqAX#X^r8F$h_#e1 z)SvJ^C*374nWHZkp71u2k~!}tif&IUnbRZxh;+*ziu9TEp_$oqxAKvuOQVk$v_cC{ z?CjwhYJ(Do6n%*ZoIRnlEe<-p9s;sRNSU+QQr}aW|eaq+J+;~{&*Bf%S^#w>xOKn20188n`XlLS*>mYfM)3wmeuVpck35dE z&$3&Wk(Q|6y|gy!`7D6^-Q=?|;ulF{K)uIY%8C_4*>tXCG+*j5PmUuf8S@Vr#Q^Lm znU%kE(0wRvqd)(968D8!gMjvYSlVVA#UxT>BNesRJ6eSQ0oyb*8zZa0lD*z#)1yV5 zGqRvweu+U-7*OogyvuVY%n{H$Ss9))m=iDeEIRzoVkH>$o)6*35 z5k*cya}-@C-l@$N98ewiZ3l8cSc&0Go%ptoL1={I)#SEMo%n#NAG%OZ613FEzsybU0}Rwp(pE9qFt0o9|?hPJrP!;=|t{AhHO!~W$R!Nl`7jk?Px^XvAi zJ-ppEv*_?73p(X#@;mYlXplE(K&P@s-bc!%9Z*A3sanpMBP*6ZfI+v4phI$_*G1^u z+w5mi!gY0p;urutI%GQd6ej>rJf2$aWuNimk`3nZ2Ad@p7%w|I|FYn??=mIQ_aj-P ztC7MoagG3paV{pDGN025m zhi$Q|0wvhrMf0IwtxJI(v}@rtn#3ti53~(`_fL*Fq62~QP|aPk5V8y|`#bChR1QP#=2VDbn?v{+f=zP)@^+`DSjbId2>JzH;TQ8PkS# zH+et*FS30SIc&4L$)_0j{$){7IUp^4B=C&9UEMvSSZL6$V;XgsHOZCSp4=+_l$3jR z=#CSNJ$6Y8!g^F?qGauzmm76A=Cwc_`c>dnxTQQNty6yNbLVB3g(Vz^%P<=~>~PGu zU>SfDT==_CzeN4xw*stiseXN9DcQwiwFFQk9=1zAKrsar*$V}$qD{e@Wowyp(p<^X zg2kN#`S|tQXQfTPtLjoGs!JEc1|~s-iD3OHD3RLiw@;WRIITG=#2oR};5P3n!Ai-_ z>H6ro$3dH?O3?3t^hxb%ybas3oXO2MRy+@bwo{iHPI#U;@9o-$_N&hq#B*g?L|etj zz<0%fOeF+!1|oO(93xetE|7I=^D&0G%>1!gABg#p|N8ZB_5b^m|M}VPeWs=nb?@|Aa%0|>$jvkHj=QQIUQkx={!sYH zt%g|sY*RlJVhZu7^r0{{Al2zy!%nk4P6kjz#^8DTD#sV9#(;&`Xh`<0?-kSZM0Flr zKw~`3DA)xGFepumWxP#_0gqp~#NI4DjR(4QKGDWsRt-$M7fko{j&<-`Jszd|gZHjVPEjb}8 zmL)5YjbInOM;@&PLN>w2(uX3KgTVUsIb-3eS#!+^b}WpPB>2(yx93l?;zMV8M^Bb< zTYh*PG6B~5VQAVwG4T{xLq%l+qxg>LdUSL@SCvgSXqrGL?EjZAt78aq4N9IhYEkd` zf3sIa;o_?DxN3k!D+RYMF_LJ>a;3pjZ`2^;ag|`LFjv;2T<47?vEyY_EYso`FwCy& zv*yw1*S??gQ-gh7BX1LTEH#XMz}i_(XN4w_WHD@VchT$BXWlYs_WeAGET6hEG%NID z5-q8CDSz?1?<2j@3Q-+U&L{oSpt%(3a;{%Ulu`sMBRx!W)qDEd}z|BZ09~ z%cWh?B$6&GrT!*tSCzTP1=R{p&M2j-u*W)G&wi2*rR3LopTArRto2 z1XN8hj(Uh*DKTr9rU{amHR|<#hoA!BnmnJr#4Phh`-6$!jBiy4K)r8!XpS7SzZTMH zeKj3Tz1#GQU2+rAp9rc~VkrueA;H#F>6kTEQQsS_H*DxthCNAP}| z&wIVXys)*Rsu}C2HtN<&K88(kCS54#Cj*fmL{=}Tg*4R%f{S4$MM*@4AWJe3Sx#F< z!D~){gZ;!g2OjpLj{P$`JVx3Zf8#6D4%x3mT=qdc-W#*5T1bktn>P5>DX>o-MY6#d z8MNJWVZh$;B+4?+w8J18J1#hlK=M5n*;66$);%e)u_TYljWV^Fp!<4ScvSkiqT2&{ z|FC?dC_GVJ0e$0V=H8We(B@*2*r1}YCUu7@VKTH@pqO!ipozgSY?Hb|(BAU$uP!V-`-Y7#{8qIlT1AW3+ePw)I(iM~?t2)0B6 zp6#0Q;2vQG*#R6i`trb?4EAebXg1_b zeW7wWtfAH_(Jrb-*rn9(WK7d{)8&F}V4rZE)nJ`S$8f^>xXUiE}`TD&_fLJu)Y0eYTLA`55p;4@EXX(WBaI_`9D~i!667+JJT>qq zcV23fSvNE3;w>#Co?F(2ztvSF-^O7vP)rI%woy@gfZPT+Ybgs&fwc$xYBju0H6SA5QE(5;`8E`*_cQt1x@ z4$Vb^wrmA5qU?c&pBPW1jBOT z&&Egim4uZb~?x-nyI}k-DUX(qSVM{za!c`VXctlTjRoD z7|Yx^iVBv;U3QuiD)?*WX`dTs{=V9en-oc;#r>e5%-z!Q2Cl6uvO~`Yo+HS9{wU7B zLln1c%~<4@_-g_Qi~Q1o$%*P^qQe2W?@5^4BrWmHmFx*KDhE7_$}|}|71-|64?S?8 zH%vcNH!K{V@Qfn}vGYNk4~G#(Jjne1Sd`WIxEK5<6=~&hnmO8LE8R^opHbv*mhLc= z)H)oHs4fiH6nw{hvtPsOD}px#chOmNzenl9B2^a@(M3y2HG8EuMX7K$$7gwHkNViG zW+_s1VS)ZYWR4GRUGa)RvrF5qMrjrNl&Hq@&={nOm-RvcE()+9zH)pnLMiU;$@nw| z{Fco)<$2Z}?`=|?nYslksX9XHmF40Apdm!h*{C@#+UZvWVsYr7?S{ShIwma?QQ8bU zvpDx0NPriqjs>R5Dx_=O*Se2@%u&F|4pq+h_%x2VEumjNxA+ZTE9l-jJE4IjbBp%y zI2;HiGQ(>?@(>lEshn-`WtZLS{uaQz0-CB3dbCfexvK;97sCz=%1 zLH&vvdA+hZxK3C#k?V(kIE_%}xhyfY}F_TTxhn_nD z7xm_wor00Epqzt~_?1S$bXTmAkYhzDCVC zDdZI}@ZY|8z#|&WkP%dD>w!BgeI|N4sF!P5!dk_x!Xjx6_*a8&$%2N((UA$FZLdW~ z8Z;R2t9a?87kXR!gw0a)n4E?$%X~WzWsc&m_*Lne7mbC z25KgcP*Eij>9U6}_bQr#pByEsH$F&`s9qB@B-QAFUOH9z$tv! zBiBEATBGjf+&b|#@6!=^5v}6fuU(a|^-2uIXQ~3zg(h*mxKWor^N}jW564@}Y!AFF z*gE@YK*daF%!Bh@{;<`;vG@DTyEuKt`RhskPyF<^>{(v86m#V=Ud zELzf_><5C=<^H|1wka|L?rIZ67kzF6oQuh);!Fs7pVeZ|+ZiU-_k!;o`k<`rrB#AP z-91uG$4HWxRj$jXj}`ULgzgwFcCYHne#2^VVmE!~UrEImW=qv!=K%xQ{zL`9{` z?x}D2+*56pC6i{@!+!zsTNT*QnbSx>bf7 z{p>k7Fg7ckQCy&nqJxU;IkmKDbQ~dLMbt3s$$B^%VC`(G{D83{%Hw|DC9;(pqIm55 zfC}cYO?o!PK$~qk6@>*|1`YO1CaQP%G`s7wW_Hb7;dgX?yLy#61se9dg?EIX%qVlu z^fmhG<4I>&qB>Js2Q5-9ByF-mB`4 zr9wdbVHi>>DFza=hpDKoLA&Ra1!9!2#CNkFILn<-HgnUrmSiu~Z}+MtxwG-S#cRge zo7mZ_2~@{qcqJ;y|#zaA*R20Y89d6FxFy2-1*3T>-BY7^ZP zUb-*?N_3EYHDB0GADz}c%b-2tZ}}eSu_xus^;zAwAyQZa4U(PhFg)%8MyZTw0=MzJdB_Blg&3mZ40KFtF>NCXPlcqIja(ud!NILmwext5O zT_(#ACIw+vK(eIH1L;avPI`Q=-(G3T+>21-a)FMUgrm7+)*PJ~)+GimIAJ258iLcz z@HbpqcuZ88WNU1{mSoNmA}`fE;MUD=lN#)XL_`XyG3z=)7rh`r~_evh{Je& z3Q?{-!40nyj_jQOqPMJU7Vh`o6}ne?LRANv;LD*ny_?>k-tFI@tb;D2F6n3V#}N8> zq{?<(JpZXFbr>BMbN1Bpt~$XD9Y2wl8phIa$>W_ZOT%T8;^4H^kfbhvz%8bl%Y)Af zn-pj1TgvP?8-&@@9(@G`!%HE<1O#Wy-tlUJ;s_@fErm~ma&_3mH(zvW#8Rl~Y*?7g z)G0FpkBT428+E6p$XAyu?DT3_n5e$QG(achc||6TcVVzD+11MzkOuov`20RzlXjnr{yf_6Bzu8Bk*~H-)GWfak{o0Ad62r34Yuy&a)hI zyygtBO_)%3<*T3BZ+%`U(h!z?8rE{B(gVT*!D`4Q4+ue95i5P06o+5YV;~UOhxB#g zeqb#!(M92z+D67K_nSEH{L~^yD(d&C&b$JV!O+72&@*{O{z!#tSh($W5+H`A&w0QY z8qqm_vfH5}4b{=Vwc4DxRq+c+*GP@GJhm)rY%)a;DQ18o_blWP@3*O!QFp=Zt)x=} z&q^M-9nuU+^TTUNu{>pJxo0tG!pATV{p-ZowO=W26=P5LZJz_G{P6s6P!d;fh*)wo zz#u&CYY=ABSo(07?hRilxfHZgk`vab>r^h&CXfq)TC(16>+G7q49}W5d108>f#Qbv z;5!~Uf?Pj?a8PR0+ySMhJ01hVMqP@Z<+s@3l@VGeJ|?P(i1oSzJ-WA`g73z>1F9bH zOF@=9DS?K5~QXQqVC`hwfnDexH59N;+2`PD z<%k^tBeew!=Dw5uqNC!RG()%_($DE0H)k0%+r5v!vf1w=$X$(l76QM2jwv0D3lfI^ z5KgF=xY+PXxjiQu3(XiR@MS|c8q@_;fqgzV{avgDF=(7ALN@#_!hY=Ft9OzI$o_lS zU*Y!Zd7B)}t#7>$YV}`Kf70(D`@S%HrjKp*OcfLZ@&%<-)anISePbt`meq(0>2;wi zCUwuhu3OUOSp)t{oZyhWReHc@Rm4GIS#Sa=o>BJFuJAPi$w|Ywt|`q0)+=Q{VXKD`evr20{C65VVtGz*=sHUAH)P66VHtdw!+FD?SCU z_6qrLD@blETtA!K;06*NSL&{`*+6|xF?|&2p`v=!>zG8fQS;zS>)mUiPv#zplVI(e zzQlLH6PlKwyLK0p#%-Bd2&FZZszdTEa|#7zP~4j{tH~eeicsXy6jtODBgs+guL~QX zGA~85jj@*Rw9?o~jzOGc#mayDO6$_4Py9w$m@0wKv^nLvyEg^*g9ljZ! z1%fu8ON#~r%cf&0!syjKD{Xqcrh;Gxa4LOKio|&a4LmJxSXiPeP@mGe!hAUeGcGV# z8$bEC`SwIL2Ah|iKr!nmvYLuoI`j{2Sh~~~71$)y1NE)9;XhZrGq%4jr+3LVhzrht zK|JN~-n(@#I!FC-L9@GMBd7md5gUoA$OAL-i;_E&*{=$*gD z)67BMN@J<@|1FcCS#aSFoD+seSPonOU-OXMFnr#Jx&NYvU(y6f0&LJ;hsrK&cr0N$ zv?y&d{xZ{Fknjr?P#*DrQ>=bS=YKZ;nXKaGhwwOw01Pt2(vwLP1Jz%fsHpaTYytE% z>e2*TWlc;PlO33(%@M45>#T1^_y%$;u#2u0v&s)&C;xfVI~YpH7RoB`g)~Bz69yI_1L$1OWCbVR+wlb z2n1x&kF}AkA+bst0~XPdHL9a?T&OS2ZTg=8;mMI}Ai>FwOqfl5XTj@MJMuu7{uU|c zW=DAJi(a;|BWEb4mLfG&)HQL(+?JWUJRXWp0jD_f+FhfIbkWlcn)U7l^rrAfZtLBf zeVgV!ayvyAt3RGw1Lbe06upW~;i&=TvLY#WW#y@ETn?n=>3><}2R4N46KQedBClRDYMUvE;qD{fr z!hTMQDRHD7+eOl3NZ%SSYFN0^s|9EiA9(EvLsu7B%<+y(p7}8GiEti83NVP(3D6q z8!57$in35*=&NQV1wm1_VrfU%Nnm$cnhwetxazF&9M?iEPAGCFOrA5=r|loKcv*4r zFI77(kWJih!DBDCzy=o?6q80l1SYCJcu?9U-Ql%D^bklrO1^2t=>6h0^`EhfJ5h~+ ze3We~5`gD1>!{aVRV}$O4`}7Q@&a9MaX6u5m~Z-=NgnfM#hRBt9g8d|k5R%V3%ami zk#WI3^ZyYRS@beU`e8LmJ8fLBBM|H8u`u7|ka#k8+`#eVIaeQj=dRq%-h@2uobq@9 zVKEeWf%yt-3$h02fiQB~a$xB~)@0bHLQhaT6cVLG-U?|BhCs|N?SN;lUprJ4;T3J?eNtE7YQ0T2x7wOV|79PtI7bOb}&) zLf^4KJdrQh?!E5iEMVk5IU`<_s6Gv~NsYeE!Da52c&}xcu>oz=FdBv^mcubw9d;Nk zzWw1_K2{hRe^#=TB=LBkbI=AdITQl{{0z_n@V*L!Jh>9!TK6{i70&sHbc1@tj=*Fk za`UwL)GIfHm(v;Ff%eOOOR4 zv&df;J3A@J+~SP&t|%pm#2fmZATf$c(kBDsWtfN7?-(mrtz(nNS9}zoZBHUr6H@4w z_AyyGiDcS*_qI~ZW{PZ}qOe2|h+oxRKosnrg}{SC%m0xZBT;?Uw@$o=w3AHw zs`^Oa0m!~sHoZ6I)oU7c1CjlpACLE^0}+4v%xXGm>IaewqI!^hPNh+3&ouz&1m%(a zQKvxHeh^MjrWXJE)ShswE%}#&d1<7Gn>)he^iI7EzCNUwN(#iUqgv^n5Tv$7g7s=T zUXVNEijN)>dnmquLTT}`i;AHO22IwKN3eC?<6bAeBCk~B&&ZzvR2r@HSz=P)C9GIT z4M+fmG-!SYaYaj}5pL)SF+$SOqQ_VT$`L3aJLWC86EHV2635tv`;)iir|rv|o{qNg z7%MDAOzl9Tg52gsFp-JEoJBq1Re~nPb$9)eZ40ASWuhhEp#bTwYYdG27aKow1i(l` zxa1tW(fRSZgAd-d8kkGpkDX6?xfvK9M`<_ObPz6|$N+o)BP!~ux=?Ue)uTQkOal?2 z<$`F*gE=S3{s;rfC)XET027fo)8yN$0MUlQjhEpQ+k^!H2JaiHYWhCZtz>FHh{PxHnJ(JgItC%;v^teY+rToG&w=q7d0 z9krI(52MtHyQD?o-SWiHm4bFP{`~tP6@o5L{OK4p!r%WF90AH=jhuh_k0&>D+LLMX z)-z&Z`97sC6YcX%@zi4hieL2WBiOk~T88ly$<3Qgf8JJa8w2n&vAdH-WkZ zD|T_1Z2FXTIVdnw@aLctKG(DuZ@GyJ9`)h@j35lHRp~APtI`>o0^gySLJoOZa2UgV!<5@_{Mb z(dK%fnHR6;!0PGC>_DsiMgA81rHf^SsF3cJrqZWDsK-)h@)^+=2%4By8e2kft>*r| zp8ic8Kf~c&J3gor?qq-0BdieojrR0^xo@vMPxsTiE*?A3IX2cNg<`f*0GTM$^nQ{C zwU%)~SJauIS0$;Se|TKpuBj6BN&7>(J?@7kh^|NqWbLZ)gEI#xlL!hE;J-<*UV_(+3SbN$q-fn|fh`^%fu6pgNJRhp#-n*;9D*r~^#c!m| zESuRW@Ar-ks-L{yCtjB3)9Y*0ZJHbx)UG-e(eKfqd*s$FN($Od-}Sg6StT)vv7zHG zRPp0ejk@F7=BX)DTZHklT9Ty2t?o%;xd*R1#m+bgZBBb%XW*U_v?+Q`!4K`}G`J)u zdF=d_*jUHi6az`MomABMH*jyAsK)doFu+TeqBJ6^8Db-Rz4FY|EoOt+D_oihUDJ?? z2)IWMK+DO}91CXPp%LeMjKCUkOlO?bzFD%xUMYe_Vga4j0nZNFQnF<3 z8ax=-4f_BHwwpkyG*_ZW%9BAiWH@Q1*I1aVobgLObBs~F*=LUEoQs^kvix;m|MI2N zN4>2sO4BPB-XWWL9D6CWu@1W^2B;GZRMbKL=%lqr%F5vs^0PXVBoCJ*u<5*`(W9#Qbci8Wny=w|`~(6-GSI3BMD(uK%gmo=)Kl zvgELkFT_ZS2(--`G`FX(c9@VdJEG4_R!Gp&>>MR~m86HR4SlsS&k}dwb zJS_jW=n$Qi713K?u2aTCQrXh;eA}lvxJY$em`QJyITZ%%Q^y@3>|;6e%D2AVWt?uc zE82CdD##9Qc7?|P0bQzLk-i7s7upOh}dY7opVwfX~-HI|L6QtlCUg&}YC zIwma?N{&~8i*iTME*O_8snG$!nI>Z7vDo)JcPfwEwNLJGS#x*{9+n(sjV4W0P4D-q znpLJL@vSB4GaqvdS}1N^V_@8k@l?*2DF z{-*J3+y23{7{m3^5(DUdlm$P0IY)4kT%wQrmQry+8<=g1hyGo3(#$4iBdHe`&FG!w zM8u5Ff;_dHSWh?)8SEBhg!~hgxGke(EY$ft-T<*s=c6uSI)U&!)R!cxn-u4#?tjVB z4vpM$sF@f&75_uPG%0Sr6gL~276;wZ!Z$@EPQlKYp(3gh5Pe7s?ShSy$t6J-ry$IE zzaYK_Yb==Iv2)9U8H+KB12YmYYfyehw~%B{kXj%+nN`sAyltTgN?Xs6YC(hY zIw)+m`gP7lLxh?hC>ekiP3fLTRAuhz9zE*qUZ~cQKC{^+>d$<|%$JkiJY`RR%*A`* zF=kkZk?%*Gdo|TPdTMTPA;?JGo_s$d!}D5ToaEfAPF#c@eZHHA(~CP#Se)?T!6YilV6{M-{%RN>Uig5LS1{`v!x|RN8o83|!mGfOpiiX_Ozsh`V{&EZgH6D)5idI>((iZgfR2f5Iz|!?2ti_}9X_pM zi-<=Dz0Wss>P^@fZT4$+*EcCF1=v`ZT`sI39bpH6#Uqto;b-w*I^0JGDsK2-^KYMd z{&`Rs@j53zcf$X=1)Q2>MU&3-j-D*zF`80s(6oVK;wiF*ioy^|Hoe&o%msLzFyTl} zI^Mk;=DBbX6C=)>D^GI5#Khd?ztPzDBJ#Gxjiv7pjX|$(0)rdp7$#w1!oplxxy9oY zR+_8R@*+m}y&vT~IS382Js9PW7}aB(pg~=ZSa4@7#2P$yS6PTPwoKh3yrS-=JHB-l zqAv}<-207^*Y~{kaB=Ay&0i}I9Dvfv*g5U$UGzP&SDGmqFGz7g(|AJacXK8@w2!%Q z+0O9Tab=0nGqpQFlQK4_C=3|FfM`yS*HFwM zcJ}hQdSvy-7>hwu1iZ+t<}y7H%CvYl*kX@lEe6eo@Em{hvgRCkaR$#bofPN6Yhr-r z6MI$`-mW_=bTL?|cGGJOL_zc`!5K-V_o}R$8Wa7Yuu1XIcdckVn@qXQ^{7^mQ9Zgc&A1Xm(91(?tsM4?ybd~-PcRULcvoQ8DD&7596f8*HU zmwr8Ek#*-{{MY4~q?E@ymkTyKmyak0L>E4!qS6HUAP|=Ty`c|9$q*r3Cb{B&ef~jh zqs|CD#;xL#g$2SD9+wt9kQ^q`zX_4%p7DpP(cf&D|KYsxk z@s2UTofF8Z%U=&TW-r9R#ckp5w}h}rG?aWZ@%2Hsq&HT7WzcOi@Y}{nQlUh4(CtJ} zJ{?D{%CjbS(SP3I1B9lk4U$g5L14%o3rrlI$c7-@fQ=Uu(W{Rc`6_ zdMH{Z$`4-^ah^0PGNB+TUbduOvrKzHz9d7KA6}p;_FrNW?+Ap?U@xHbq?+DMV=q@9 zNu>|U(4$>qz7{XTGx4%(U{_FIy+@cnbLh$4^w5~a{)fk;^fDSujA&0rIynYRaI-Nw zdd45^t(+yoprPng&f|rSK z1si$}>axSG`WMmrr=-jF$nzu4%J$3w8Xs-4cM;tU814->F6fIWgVNoTGty-j$T|NQ zQ36z6o|D!o`$>lI`hvdiW(j-3M{B%Xiao^RDDme%;XAt5tlrQ3cRy(*W!wrWdAwpZ z+4xEyQ_LxfR8vuy5He^>&280|Dg}u6%Gbdr%M!;=0=ku*pfF|mRIhA=<`1l6z%6uL z5cI$qGzlR2&;z^bR=OP+S&&8}Ewoc$85nNCCz=#34$$`vUuI1I007fSJM58Zn_H^qERkn z@8Y{EgCyM}S&<=c)SdV0o!%0%3?%F_Fj1T*VQu|sL-p2P>s#j8hLJ7H*^amBkry~_~I@u_Skszc0xO06t zG|!`455|HAPJZ!(Ltpu5>fcvyD$2`TVtT+-Aiq~(K*DG#-^MaY@T>#)8RJhF-Qx=$ zh9rE$x^%psbNB|S;kI<}IEi%KX6b08mCf9o(x3s)t z(@&G4UA6#uX+8359I>k6#YnR0 zx3&6LbM>psA;!c7=!vvxSC*=K3atrD;^|A5~rX^0vr}C*D^*Uz2vs{RUGCgX2 zOn_Ky<0;0&LtMHX&s zJ<7@?kv#d4K!dgmgbMVizPi~jj(YO%^Jnb2ukN`wjB3uGJL=eP;IW2D<=Eh%@ieQ+_U6eI6b7u_t~?3dxGzfTSd!AE7zlXmqZ6<)$HnR5^@;!KPY79)=7jFarJ z7@>D@V)D;Fedk55qpDLC0aPjkpk{_EOBRIr(N)RuuLY)6i{CpQKJDYlW4g$mMJlIa1Yd>)U#)R+pwI%rI#QZJh#%AIv7-m^Xg--~tXnh-(osY;jNGb%GOI0t6vXm9)k&ic-9-tT$d z=Y8Jq^L^aff(`STET5e9WSbq&f{zAScaXE742U%!KKH6( z4$3M>#k5k$PR>;w^KFvl($ymEjLZ=IReju5vwOS;BU_{LeILlvB$aOa0yK&hf&@6y zEWIJkr%xy{o$%fa`m8Y3a78!$h0GDR`FH39__*klDD>eL^KolJ9y%p~EU zHpMNcRQYB3C)2k;fnyy1k-(u*Wpj^8Vj}k`q5er|7p)iHpR6~TbvI1l(XzRfRX5o0 zVu?@tOS@+o49o`~cJ3uxpR<4oaJ)wZOm#Dr=2)$_2*e4z;HO7Z%`o1vzh`bMx;0&2IB zx4j_pUkl|pv8gK06d$XlnXtyWSua%flfl0rY+O5`-j^joMliB*1w z98jKg$l>(7GxU(dc~0D{mZlr7B{P&&KdQatnq2^9&*5Zf+5hBubcVg8~qVKRLjRh zH9=i6{Azz1aya$FW#1lh80)**Z3#s4fEy(YLpl7W^ua_lbnTe%Yeqx2gMZ7b5mFkD z2!>>>QJ9)?6IPjv7o>L?n;;p5nj<6)xl5mUrZHqAKj-v zu2F0gR!JI3ll0c~RBn~^i$7S)s*Pcw#&xRt2s3-S!FGMW=&k=En_d|9!)~Jiv6y0@ zd{rJ5y~$~l+ZQfS2zWZIFXalf@tn}`aW5j;Y^i#55*2edA zpZ#OjES*Di`v;AkwRZvT$DY~g80UF_hmf~<>t4z>n~I{`YONvBh89sGRm zCaAld3qQBeSj&CF9XCEJ^K?g<9+^}tW508bH>56Zrllq9b}hn8u?Q?Nqb~|^K45bo zmY`{kx+KQRZ3UdQq%Ckz-o?jc+bkC?AGarrNz1Uq9=VpaJJGAm+mq3z+pAR0YZ-%$ zak<}rj%;A(2;1=h1#!a>E>}Lqz#?cC6^#|?m%3czqg>*LSL7(%d~Hq--7RmGA$@P} ztgEw;k5}ah{0XPJnJcFa5WF1!#}r%qZBvA?$JG^)uvq-f^{UnJ=- z3{rrSHv%a|6a%ZGxtL#|QS5|rMvwJX%RHK7-O^4uig{Orr^;7Fr9%P&OlDB^d)VQ! zyiiaZ)F~g3)XhA{g(^mvD;q0VF1R`yX4$p@%|y^*z|B~lKJIfTe(A97dFRYD!>Z&) z-!9c58mX1dQk*br6v&sbO<#`nur2K8v6`^SIaZ_1l+$04cfV_}A-mpim6Gd@WTjCc z|31YGP~$^q&?QH;+rO98%>+XJ4pPQJMce#Y=jQiu zi|A^mjDzW6gD$Dl)VO;?+C?8)0Oc{;B!ez-0mFXy-geRV77RJ8UYH}+67?l9E-7IT zmvP#dI!=$P8rct+2y`X5RdJSS7vhNlQVV?Ar5yM!iM&DAV+_hH7mD=1^{c9CoXw+9 z111k($ZgmG{|aoW8zhU+V?o~cFw`x=?+x|Q8o=?!b`m+b`u$==@a@+l_@9G+DRwiU z?LUG$Yva;P#AGzK&QMXR9z?BCPeWDpbA4iHNF{n zA)wvBfAE{G1sMIqqk}GoW%@HVH)IytttT#wEeE9$7SO!ix>9p6)0DpTf+)_IA$u*8 zC2|_PTcawuot#Edoad0kiNGVTwnnvj-U&ZKlH^!jVvWz2HY%bvqEnu1vV+Z&`)s}S z&RdmXrUhW_c0O&Uq^n*lS+%)$j-Wl3#Oj3G3 z$fWQDsTdR%7E^bD0~5>s9MQ8_!ncgEElqd5VRqLjm z`l_wG+VSAU$g{NpR z8s%5pB>UuQoXKx;C=b*qI>)OHHlA)`i;eMo#(3^g2mZL!6dPt9D$-vys#5N8 z;STR@kX#JurWnyEkb~H%PH3e`mqlo+w`-KYj@?5Yr2!iqVm0F1Z$M$&ueG9ou5qRX zQtj_~>3^c;eJy>4(@P_*_G8bQpk~Es&$5VmU@hJAI_}vTRXwlLZuGjt!vimh`^6uVH%A{|o6dvW49eko~Tc zlX|0ZrkrA6c2rD7SBuU98zNVQ+_Xf9;bMZ*W!^E6wnPznj3r~0kt_5)*B<(k_@>u& z`Et*Em6pGpa}qebbp-9_Mn=oe5Ve3rl|Bm_+pz&^tO6Pv5MKVZeD`0zX#mZ;r_+8+ zK5-1R*D^1B{2GF=^Rz6T(2!Q4#J&vy5q^~Y;KF~YtCxuvR}vW`#+Vm z%M7T>nEw0sNU9x2S}Kg-o=Y)XD3SrWDv}{tIp;xSk5^CdzJS%FfpeHXB-SX3CA(*A znc3n7QyqFcr_%QpAJ1>4)4gh46PV@G;R^1Cmu+A|jIcCr02{Sp{T)^Srmltk!SRO% zoUC|n?_F}jj&X9`2q*0nbB-cssOX(AIqnrN4)2Bd2`qLh^S#f|#iS*w|Ce_bZ2o^2 zzP0ja4R4=bIQ-SYZ&rRo^R@dx;X6jx@m4$D{Q8%#b^R*!x7Qae39nTRf2Db5r+2EL zffMhT?vp9d^5c0;{PV)i${4{easD^2|7z%8|9t*?=f8by!D>!SV49?hzsWseHgc`D z(wWRwzg@y!k}Ammk59fG|MfV>R4PVL9DQ{<`UF#}-%w&2i() zoF%m2xFu+0hcDs$_pX`p9L$K|;wYrAafWI>>h914raKgu>IT1I(F3#O%UgPlr4QfI z_N;XstA}sD6RW@Gzck%cONdR3*M7%c;6;rL$ZVjP6pAF{8gweKpI|v@uLAWGhaJ*I zcfvb4`vTOv!drs!shbND7fvS1Ed{vFXjluq7PRxPtKRy}MDo<^cnW1EPYshBHwI*j zAN>ZG-A&`OO^Ewje5&Umsd8gLzq~@(Bt>2S)+ivW!#R*Jn^g`}JIsu2@NQHsbIGvD zWLwv1tl%=rKx0AAkL!|^g73_H%izA0IETGVE{#T6v14;FWMnRGQA`g-KBuBfx$VL( zx|W~7;N0yP$(pH!C`_hPt!sbyB_Mf$o~9EP4Z)l>tmN6s#0bb(S ziU(q>kUs3N)%6ri2K8U3P&7bJz+@h{T&eO;xcij(^iC)uR|oUaL=MbN!Ll_=d!@Vm zC>5YsPG4!~pBkv^s&{WaaZYfk|)O(TA zn+t=D7ARB+3|6RqAg&0%?sg8u<#i&hSndR)iuw`-pgS{Tx!^if&(hK{k)Y84DJZz2 zRveC+^S}s(Okd2K#{7IOeOHX*SfKzFZ}7Oa&|wWg7>zxTyL4h3=B>*p6sswLXB{^! zFj$=LANc8Zve%J7m}CSm_Bh2Hp~yihx=phC6{NY<(ly*|LN!V(LwU~PaHM_LIByKt z71=2-h9I%Vr%Cc)Au7A?{A~ySFn!8Rof*;&gh4!Byjl_7z^N4`2kLtNxNxTwt4cJ{yI;|Ng+4ich4J*H)qBmT3k^Ht=suD@7G8BfhnTF5-aj)-EX)|BlEiK@n?z{e^x+nzmNm>*=kUXn*dt1PU z)HugGV^gh)GN=)n4+=)Dj2aKr2E{vT>BF)%d2LY3#45?!5=P@d&*NKI!N`6CK6K{B z<))?9*#vy-_b-Emiji+aGQ}iPWHl9y3uEdl9(TglgQ-*>bka&tO3-?v)nkM_bHhSt zKeHWcUodNUMYm~am<=ZEcvf#F!i@!&THKOF*{*w4y##IE=V5nTu7K9dct*a7^@qKp zLfO3Sx26f#&z0_nnY3!Xx2sQ>g_R))lqKMFK{RDv6L?RU zBD>UUd5A7u3CF{Hx=>)F6j?ovEX2cT-jPj7>8Lz|>l34TVe|F+OJ$qrD+nshusPgw*B_B}e`J5n1y`4>~oUo$eg z_4G3GLCuUf9wxIxVY3dxs!siLvSwN_L*AliKDyN&_2`5jwxI#@c6*el zh5f#BS67@KUSt?#D!fn1NS7V+W8#dY0`5`F9g5rr9u>^Mb5%F|wctgh%I}eDvy17N zFieE56<(qHWW`+FJiJf@PY2b7P>+-Nk(yXiy z-P)7>+cj{89WW$N8-z8|pM07|aTY`d5}5jcF8DwUF)epq%4?Pub2f+?B9eK=Q!8t~ z5~IE+Yz~g+m6-BY(RZnda3t*aoifLfXp`WyXgO#;11;jD`+@MdfCEYlBiGCJ_(3j9 zZ$z!ohGn{*vV81t^wjM(JJ4~gcgS=&V>5BKPQFP?te zr8VlXB9oUZx-C!lN#taU^Cbm-=b20wt2#{Vpfk2Du_lgU{(YIBsdH?`tK9C~1=Ilu zVigK*l0}{Je(-+&A#SB*&(fMt$+EUy5D({jmww@GTBplQIV_Zf%b-h~)pc_>g~SSg zHiF4MfA3PIN1}I(N3Hr6J)<8i1BT({dPZB;Jk1Ii_8Z0gEb2_ozZqQLx5|DyNSf`~ z^}T5nF}X}J7bwyOYy@dvFmehKMd-#}3u)$Lh(7buLV4mQ*%JTHJ^Ex=^e5g;eM*?k z#TCw;;06g+iB#Wmy5zA~aN0MCR}`Xda_Z4(oZX_;`yonK6fy{P1%L#l-#P8O0%YQW z5rs92@#aKG;>QKgPq=;Z0uFK;`e5a{bY3#l8{G$V@*YT^L3NQMyjWMDgtjPv<^`XR z-ubYDehaDPW5)Yffq(1)hW*(w-WGP?KjOY!bv^jplSr&kfkWIb-h`Wt)i{&LaJ4Om(K_iuiUe8vtd_G@Vg%+p82 zd#_T=6^dM*IP8FH=o$1yF$~E_ z7R^;*@To2!RUS(QxhRj1oQ?+06$nc~po(`xidW(E>K>_ps>rqcq}RTDZPi!RjYKPH zljz$_66CrLJ1llWd$o_$2Vhn9TAjilmY}N*N|D|P*H64r*b~_~Cttg`0NJRy9=PzS zZ)FUJU_wA+Z?sJAFb~j&g{LI{Tnh`R>EZIbmNJCsQ68KVQuKf;Ne zu@n{)`Hb=0E4lEi??3M;mB=DT5s+$_M5l6_ef7*b{8^~e;Z2>k+Vudiw)34ImB&FI zUYn#qd=#WzZ4CuWd*0+tw(0sKB88n#!#mZHWje>US9%vQGx^UVYk3=)QgBHj6Yo+8 zCM{_vM{~;L&9n#-CihInE;~rb=ei#0P{tIG4e9jzO z=5Cl5x4m-ak0j-VnHQHE&5Ls=28fj!V48~_%qS7BjI41k_AZ8O>K(FMbaJ>(t+yVE zF6Ge0k|JfL@1WamkX}$@gs)I^%jq`VBh}iTfjbVgPs;F_nEU?Fv-1}1I7nh9ZxIP_ zI>CtZ1Og<`{lm(b$XLNDK`}&VVP>Q3p*M)Os5%$KM5-}?s6L=(eluiGpAQ&x-Ab=u z2HoR4FNkBDS5P*yhU15X9vQZ>hUN|6`{|8;4>8yj+576vq?Daav0sM^bhJjs+Uh7~ zFGZ@U=*#lDY5CqQie(}D0&a@caWB_P%KhtRs5gjWc#WRh+~dd^pQRiW7SbsC-p*Xd^5Whv;K2dCqba6&*TlC>$T=e7h_hU_UHQA69?AQ>Qkq501VoqGD9NgI! zMW-SbXo0DLSLLuWl&w7K3(0fyij>J@)!Z?cb;fw2>`$@92K!GK-yNPpU$6f#SQpK8 z!9#M|k@Ok8mRgGGpvXlkx-YzXUXH4p?C?v~)u1VzcTxN>SWEZM**hJp%J&G><^IPN zsqzcnRU-BEmnu0)B+ch!AS}({4yKFUa#A00S{HOxoDh`(b)13JRx<+x4(OxqYyH*v zqEaB@XNG8L6i7%8zfWHZ`P^k`h#E7HzX;nqBXt^k9QG3w*z%UAV(|+I7 zn6yu`O~qu)sIXvA^ie<*=dQ7WeY)i0K zILjPcqY5|&NvW(s(!q~&>GtUQ zdeED7hmY~h&7B$-M$9#4#pCQSa`o<(LsJc=XX*0517z62ZY2Srq}oW+vxZ{gDY6nI z?&OC^n*^7P)%q~xKOiX$JJ1qDSh1rmZi*EmrWSf?r^yY7*pM7>maKnaX5f`Z_{pP~ zY>H%3(Rm?BylzfrNH+&*F6RQy1sn)Sgv7Q%mqZTALJY|ul($)`hEhu~+$3^vWxC2U zIlPHZ3@YZfdfL$OvJ_V4d~%C1GL8#>)wY3W>g2MSXxlM`&Wx|VL%bs3wm3`l=yyxh z>8Mno#x?N^6dNI}K1T46SN!rpK{EgZr(>1kmbtA_n7=egpH)2MWeT!H2YInR#V+;4 zW&M&@c1~8V9_lH4?6cuM2W+GXM5(9ICRR1q8>VyT;)OxxFujy$Spc`+T?{{ z9`}odn0S2sWNoBEY-II>x}%q54)|LMVR6n%9oeMi#7$5vb3==~N(h&77)Kfmy;zv7QS zYkXRQHj+PXSeWs(_}4amZ7cm~>)305VrO_PM9x?S#nN-EKxf{k5xf*sY083PGc4J$ zoiW3i!i6XZ6*=ciK=l+dr@#@mu}fm6t1JSGnR{({xT@6qKTKC3?RFw;X0@?i(#=64 zMK=^%8FazGa6Ww*%mPM_A?)7*70mnOkXGo5{K~;FbvGnC8&$52+ zWpG}!^MfyvbaoSEJ057Njog|dih%;#sW^T!k9I)gP#oxSx1~o zLe*WNo0a=!sBw*}GqhRSPilq#iz^ioB512m@VuKyemdy|w`seie#yK7cJQH||F7R09xE(Y)Qss$3pF^5rg)EO= z&)MbL8da}4;I`=CV zxPPxWHRQS*O1W6yW<2{vwhR-`ZpYeJC^m)1Ohl2xjxk|IkprjIS#B8&24Qi$f&4^o zb%m^he@hqx;8^O`Bh7p%m%ik&*(1xXQ-Rsg1CkEc$3(pje!@#90_!F9z>K<2-;`q2 z^9<20msS}}oiN&~r+Zj$T(S&0BaOwF+g3x++Cx7X_~#msjw_OQ6<@t5&gVd}*oaO= z6Te4zknG^xqRZ$cat3neZ_!&p6E$0$%zGd%lkJ^1PPfHE$ecGz!nYV^Dz-LGo*!WGF9(n-<-JFFDswK9SbURFWT?=pXPLt=1% zg|~(8k|cHxupJwpZAK1oCdI(gV;bmS&cWbHMO3f216G1`at&C}R9_rmZ-7RHfnKyw zFzHI2(_zCz@j=8YC>XPms}cNmQ&C;JolKh%(0$DFM){^3RR@1_L>^=m zsZq59BP>PASi$aBU@nx#Df6hFcU+O?ewh?NO`;^;5mG%5YTZIjTjCG$t+uRrfe&PB>?p*rO%Tf4H zFQHHn@BEL5T3F^6vVCUOxSrK>x8HffyT4d^(Nx0FZrxim+KOkOhPxtGe) zAW8Ak*C2wQtAb28*?!k02d9^DL$PF{2vnXAzw<} zEX`4&LMALQ;7ym|(;v9$+tT=K^z5L2`)n*d+Y`0L6|C&Yl;FQBe=0P%EGyG5{U=Fa zH(j#hMd~6WL$sMr-l4T+cyyj&V-4X7}u6O^ilB&9xCtRO~!azTn;7J zk^gk>o8968_rb``9#|O^=fDTK_sFim!#3BEd(uxV`}@g_Hr`@oJ*d~e)AJpv0Ubxb z`@IWftsSFdyAe9JPzqd}jrzR{FUx-L@>J9?}XA;O<-W6BRyW)|ns`S;;xe%KhbkRZ*(xMLMbIe9nr%Vl( zi_~l6CF1p5bpnGbTXg{GLQ!(~Rc^5azXMc^^6BPektPrvT0x;e0e+Q+`5^xAOs9O@ zXncapac+$b=daEE@X|L7_`3JaZ@o$Oy)gK?WQ4C%6w^qN<5YAr#BeitsGp`$wK17Y zoPR6=SMrBuXyVA}Mei3zuP!=)*+q9ksv^v(_LDlIuV9aQ z-FsoFXjRM*@HM5;#JYh(fbfvX+&AUb&@5JtQGgzRX zG;BUY*0HldcFf%@GqOP06q89onsxNK`J0?7mHMB+(S-U({d7JjmNfB)WcfgpY4C21 zT1zg7n`DFD>t~*upT^06S$;fsn~Y(U8e>d*e0tZ*by zdSS$p-p3S^N|Cixbf!n;?EVN41S%5Tk$oo7D7wX&&b?BttO^mb*sI~MWY$xRICq3^%Co?8lXq^IWO36n>K(QT4Ox_(+d%$eJ{aiGvr=3M5CxzV7ShZ(U= z{6vmEX&&>~u!ImM2xuVBq&6s@UhmpO_i;ZXh0encxGt47vmJBt#a;ZP?znnnZ96$m z`tE49(zNF%w(oP8yIEwwO%{LsWs+#e@!w)2BxoomogyDo(KXKca7qUs6BpwGa%rp` zh3V7%K-4(~0kl;47qdT|eanB7)8gsnUbeLmkNbnYsd#*=O*dGLjZ->RHx8RF5WFBm zkeTI#Z4v#m*ZAmvqW({76og!aP14q=Zj$EQC~A%R*z;Uqqo^#Rp0q@5;GFPk5}uAq zv}xfy10dK#3N46D(WGlUa*>Byn8Awt=F_<)q|A{(pmW5;xPfA7DYA!(J`iy2-;s`# zGyg(phTpjWEJ~O+GfPy?y&%66o~=3q6~gkBtx=c6P296W427m~2Ape!OQp>aF;8IH zgKmf8QH`R(yKZK!ux{qv-~^@@%5Wxv-Y(V}s|BIJ7EV6hNbe^|dOAtE+^q4+kMH4+ zX2M4tvqImndY|prBV72$+bUC4m#5>kcI>B^(QY}UEANVnkay(J8N)v2WUKVZ7il%yRG|bysjAEDGKVTP;VnK>n;g?<8J~pzbAos{0O7kBrC- zBD}pqa+{~)L*~y)_vzIwFL$`^3ha~b4MT=yCs6Co%GbGR6nIpP<3RB2+a57Lj?Mai)0}>9K*hCrYhEVT zUYID|aw8M=CB^hpPJ-jbh0x ztTo=j*B{f1Pj!HRN42i$r(L`jw?RZo(Bm0dh1iH$mO|qN1O7+7I zI9cxx-{zAkg4OTK!seiC*Blj6r*;ByN0f>rzI`G8)B)m;a?{K`iflK>wVpQKj5^`gfbC->Hp% zV1=)$d3m9SWuQC5oB_3n%^nAxTBD8&i#ZLvt*(dYmN!~|lt_Kl8nuLzO`3y-g$JE9 zidy*_ZXN755JH0BmdBlUxor>~jvtJqBG?bz^`p^1I!_TVc`PTAsI(Pg5pAnii+ z5x+a(Kq}{)A!Xd{kYN76#p-L3WAzF!%{wV>7emTyr##OIS^OzJlL`b2`v375Fln6+A7p-( z?VYvewkh-944EDuZlOSZQ`jGlo4sT&OfOOJtqbxXFsE$SFE@Wx0|i&l&4+1Yg)E=W zQ{9rB_AH*EegJugTYiFP<9N2z(`NJhA9uva4I6ce`R6}MwBCM4=Al&yEv5sD-4b`r z3>CoF=OuEG@}dS_s30@s^bLt_R@sx0Y6p992`=lrYVOt;75?qb6ct+bc z1ls*Y@1&bj7R;~!u*k1Y3j^tC7(nkA=KMow02Vz<8$L60O|yGP?Uu3Kd9&)>83wOZ zm9VmoY_{WN$2~?KXbHu@!gD?qjmmPE+pO+bm=K}fP2UOc;2xG?Er(iuzCUC+gPJld z@)gs2lqbdc{#En3rEpy))L6VMA0Us^S!%hNA+$Zg~0J{6Wav#ced+ogQUC7^G8*NvHsc__h$6az@S?}P)c_N1q&~8G3zJBc?E?_ z{ivvttHPW04W&+-kO{H|!p*^dB2Z_A@G1YINf!Q_0Z4C^{dSNv+cA)C8Ug7t#ay6B z8|I9mUyG>&4?VH6Du@IkldV~*Q9O{G1eL-X-UFAN5(xE`geLKF!ZXD=pmvy}+5>6O zNj$Z7#sFj#l!P{bOiJ}U1kxwm^8p)P&WFr|J*qw#Xse(gPpUk|P5tO2G(HI-z0wTt z9=Zc2(VbAhG0lC^9^jrpSsFN1DC(8w(??Z{K6V{)(EqFQYzoq-^b3cTzWZi8YH3V; zVOwI%9QGa`9|MH!FedxCPnu}~tmi5Rz)YE|PWc{HC0!;~LxoPLW`tUr8KOS^gsjRK z4}FYNmf>WK25fZ7JWk&F^`#q52D@|UEA=0drR*GGJDwqJF|txA6q8JmL@FAKqG5nd zjkXELBP=wtRH^~>ChJe5pVgn10R~LLvl_DU`Gol|zhg?UeL*J9W~jIMbCT!OIA@6N zd7q+D(|;vYHbe5XZAogQcbkQ%82x4&9u51(^xs7pjD#X6-U`k@c%umYoL(r#sZreWYLuo)juAXk6VxcG7Io|PVFeRD3eUDi_0n4xUKGbt zSL9_LTO?Vk^Rve_NLh%F#~#FZHpT-Z^TW(2i#q*dW!1Da(2`?VIbhTYI(zq3Nn znHd>1$eX&Q&sc&>!uu6yXBjhVXb~Jwfq%Qpe8Suw#4CZscw>Qp|CR9HF9%xvicb3(n8d_?(+xN~-3zNv_Vt$Snlm+9a4L znyV`C)A*Fns)U${I!4gzq24)t6$fNBk1I0#Z;@I$E@0SCqgd-yKKDz{{XF$fWiPB* z7_xkVo25m5AgN3*{T37xsfAj=`;^7r+XXF(0Rl>f!1u$CP*;R!hG306tU}TideA9# z8f$~a8h<08$ZpeO=;xF<}_8 zN4dAj8ljB3(-K4&-7rZzl3PcG0fqsnyr(Rd=+RTj0T{c6jMo&GAeopN#WM> zA+-v`_BB4+gq7YPoQ}+UEv*y2#pRjZ{#`mj^#rCx(V|#{g`|)juFD(7tYR&MR?_@4 z=q0f1h~$>7Fgv~tdw0U`xcmkF%M4!^(cugmR=}Di$YD32;MEg{*B!A_rG=FHHpu`v z;o3#(d8O)gK?lSIVqMYX4tUHNfx`}uM&`uXx1iYE7ZCQ=SKDq;sFyo0g{lc(N@|0; zY{)sC@Kc`n10A&mR_>(zzKnlT{b{G^Dxcj>fy}JvVW6xvs=%$Blg&k`(HKEKr+|Y0);X;J2+)@ThIe>dr)4; z9p47dCa2>Ojh;H0eJx#2BS13f`#M!6kg;{i^Br~=AXv?;I-&vCDJK;5l9oqjjX5ori136XGT#4|gBM)KO$F z72QvUWF7p>z+CzmeGTNhZquiwv4RWYYo4d2UHn?*Dv0}MC=*kDe`yzJ7Ger2J-OUa0uQick8&jpV56hSv@TCDk>GEk4!rN+a8t z)~J*22f_#}0~_i2WHTS=7N_Mbbk)` zoL7-kPJ095rvWR;@kHQ&WSQfEfIg`1Y|TD?#q?{CmG#I|<)skG&?s(W?UdS}SV1E_!E6V%e=nPd&9?1{TyOqik3to*-(Z{K z?+L|R!;s_P?+{DkUYH>#+i1v1rI@u8Swls4@GE5l;=|$tz-t1wnrs!RbiL!E@oADI z^ExMU;ZKJ?i{U*Htd8T}ec$=j7cv7bGN%9jJ(Bvu;G)6^7r7JzJo^m8OqfRPSgZ?G zK6mA`r2*@LE-9DHUha5(#**1999yDrQKw9Hh+giv$#JdY*`MRlcIO@5r=x5={8)?= zQzv5_eb43=$Dx1wJ7vJhL)V8V$&iEHiZ2i#v5v@vO{AFB6jXqM%wnty(+|p?=jP*_ z<$Y@`#j$ya`fDxeV)m_0{0>$l-Bd4+<`Sg=5VJD=H<{$IGaU9iR1#44908vyiUCWq z9gDTc@tPrms->Z{dYX>Xh=^);KA~uqwu10$KD}R7<9bGk<;PGa2XurNErN*btZMOf zc{it+lRxW9aH}FwbXIc7BZJ-yl{=3SV9jlE+T>O!(9&A}2?l44FklJHOpVAG`$la~ zWT*Mc`s2X{;AHvatS8%E7~lY#Z3J*?DF&2Pc2UuJbVqO!#9!4IgvFY4J#-Ge9oUgI zLAnIfEu37BMWwQmP`Cue=3)f({(Dp!RRW`v`zv&-_pg}`ILQPetwx3DQtwl)n!DQd zfGXel5~!Dha`{Jt+y`ID5o<}Bq~2eRe1x07Oq`hfmz4ozgsq7HXv>j#ZKix`yN$b= zVMwFlC{k_JzMrNQFvvx?p@zVx&5a(-&Sn zuU3du);7tYi_U(WB$bK*UW}I-SD8_V{5Fg-DLz-EaNyAr~=Lvp}bvfhrt1B|s1;K`#HnEz%{(Hnh-gRAGY zNwT@A)70m<+XdSNnNC=f1$GBrl3mLHH(2gA(qZ9*?{QXdc9>RjSKzIv? z)EBv>uu#~+Zq#Yq7q_UY=`G;l=?~CQCxN*vj}e>-y8eMIWGjhP?q>Js0 z;5gA@bJVax3l;k-FR`BiE$@}?UqiC(7%dG(XsMu>?G%V!M3>1|1jNWYWq+=U$P&d9 z*hqe*JaD~NmncT?kZ1lZvQ|>%jh;>X3VJOm4o3+u4A?#7wMMNa8R1$!OyRw(i=YXj z<`H`E_n0<=?MZg1xxMS^)K?6ssh;)M<)p}tQFFuyHG3!q>W1u~q7OmjRx3&3<*F(; znNE8`F>v1+m8E28HKlGs9JlgvM|$pp(8<){)OG%93x8CtFbs%7a2>17kJNSFUa)CWpD5{qfMI6!XdW)3a7C%a_lJM^x&cB0!MY!kGtXJ&!n;PW>Y|uVid;sS_AXiV zJdHvN`G%;MeFfxu`s8i$0`XC(@D5?ytK8G{C67b?72%z7i0S0OfV3U>FtxNgkGn;J z%ozyZ!17exT(}q@DsKU54TO0*Ase!l9C1nrSuE(4Ce7Q)Q}=OKMkEGxhV{DFd8iYE z5;-^B>O2;Gp~Us|L=FnLV;*E(z%H>mCU7YyCa`?2`m9GH2Zlf=b*}1+?o``k5rD$30%V7v#9wFujEqkM))we9MWY4#BXFnRG>^V(A8z75@A*h?XU(jH$}s2v zUz<}Y%i|^m6p0~K1Zpu3y0k_e2A>Aq=#x(QD%(2AM$er+7L0DU*^YTEjAn_pew3tg zHCUqGw3D|<@(Z&PRcd6TvM2@$v!zqfDLxvNPPqk3t719#rIG3@?zraEL7xcJOjqOa z4!TpG%GG+Rx4NF^pp~kdJDNSPZYQ!wl)+Y5_iH@Wnj7=e?1@yUvfl*~a}}y$1c@A! z`ne9|OrYdQyJ(n;A$20#wI?(mf`00!$A+gKoAnN`0fzNH_1P`ew7$0A&0su!(y;jq zS!c(daG8-g$fg+J=WW)nFWp1(Ws9Ov)BlHNyUb477(n+-&(8U%*XaaB;&#e#owp3ckR`@Km!x_ACw8X%_*Q%ABQr$5LIF}NbK_tl$8sUtaIG~(A$%wCFAQ_*_0U*x{- z7k9!F+X+;GO4)wT^1!1KUC12^l49%ti(mK2in%2(*9NU{*44YMg(w+{Lw3-YK->$< z%dM2B%hmcjG>SVeS7<}YZ0s?;(W;j%WvtN$)_wGk`((m>@Pe41-48B5_~%N8-5Yj@ z-lbUM14Vb8*KnI4fZI#QN@BeC@X*0VS&n+6 zjSsRi9rnwh|IXcQd!Bdc%8w=2gK?4qp$-%uyb+A@gI%h+nJ~B^y%KA6qoM*vKPC&q zmcaJ$IEEWGhU=r|S;2NnPx9gA&wIU(B4x3pO;P}P959nD6f8ojKo@;fQXU8*Juj~d z!s-kwoF@wx&=2gO@u=f$|C=2&{$6wSb<<*XW@^G=`gxBO(u7mx)sVO~?11T8=x@TD z7&mdg&>oZ?*(0st0v*J^-@iqX0DdPX4nn4SP@`v?1Tzfr#x7NQBuay(_!J7Ratr*r zWLqTdLHUq_u6Jay^6X)tTla~#33tDO%V|2N78lAou6&p1f^VNS*TQ0BZg$!=(6w z4?Fjgt&ZfN(WH16#q6L+Ij)M8O~-YvBwjVF#$Sf~MlRT3vB{dv}>7ILvX*}fx3C5WT3l#~{BU?DB^1H$JgK=4S_bU|=1omOs zV)t#bQ*QTz7rP_#a?s^KKryWWv6p5B&TeTYzySr@)yd(AEX*KyR28L%{tFV3vSzjm)#F}lwgEJZ zjp{go!p=s1=&*5t$S{g!@z-A_iR>sU_8T`K#YTg%hGNnw@-Y?N>HW~JZsyj|6_E)M zsq%E6>w!)DW8{G7KtKY><)+F{l7aC3b9O2A!g>ITc3W*w9lJl-!)fgHtaUc$yLZ~3 z_X##G6{ze|+mg|A(A@R;NJY1rPfwfNzzD!a_W5=-TY-@Gpv)f^1JEkt- z|JB{J&=4Cp#*V2=X37#(c_NE;8=axN=(*9eCAdqrftqAJUqdHagrH}4xYe(?(IlL1 zn)3XDV8Tp-^bJUgN7oc%w;%iWlEa+3nIM%xcenzv`aazxs}Sa>z}=X4Z2ltD+UN>K znsjR5Rr#V8XluDVzq33ZwT7-$mS5MwX>^EZX&MepCYa4u6gaA`JV7zY=Sg`U!sOfKaZ+v#) z20OMCW(qfGY1CA%bi!2By1B)ao|{)pwJ4H!sgSy(g&Ay&VCQ6#9@zoLCfFFYQS*Rl zoE!UnQ+n+4rfHfYs-Aa{d?AgA#6+#L!aaTu#3#w1n{AB3s1NwjLCXN}=<3P)Z5{xl zq~%QMaa`K(YZzhre<58)w!APgk9wo=sGMRTPp6oQZvT%?c`S($l+pWm#o;TV`rIHl zk$0bza$!vp)}61-DFY6|VyCO}eY~qt_c=v^)~Kp)TnWBR+W+G|IS%<@>0ULV`vUNq z&aiu7t6r*^kJs7&f|fzdcocei&-&0Cj()|-beOSQOJZh7Y2lQ6U@CE$eC?}OU5dne zl&I{~FDwr^tjwVANSgQ>1%ku|-D<8w*mbwLR;+)Vl|`X${92(fm7FxA)p+fDt-9LA zL$Y!X>K$#Lzt?lCYhvUrx>0afx>#_Su9to6xt7$+8aUhidm#n0kh|SqD>)cC5PnVt zqBhl#H?EyAKuY{ld>TDV{PKJTTuP>E6or0SqMXny(Xz0XsBE%@C#~0z!$q%8K<)M* zjBc5n@`&<~O5a;x*uM1K4`O0FWGVALff@isbU~;_v7dX39+ams*urwhBShcI5u$JB z2*H+y9N@U$9qtN8jiQ{xZp@pUSH$olv3OGy>rcF9np?;wN@vHjDKnXfjRAvlXSZ#r-jElvw-;vVXsfxI_~zRifIZb281xO(z7S&fxbMYlLJOxgK%>jQ%z!HCdY62&A? zWEB-1Bj^E*{w@^SQPt8pDkKg3Loi_hB&@>~VTmSokhm${|GkOG7uYd>Ajb$m>nUa( zMUtrK``@@I{tzpB7r2#k()=^PfmrrZwz!Ng4ljkt_0_qPizZ88#NI(*iJ4I@m(H8F z!?a|o-P#H>wOn;xz)95u@pZS=uIeiu@sU09Ep8iCL$c(EFJ(P26?nuWnT%h3AEKM2^{|XnGqVV&L>J||7i{*37aRm(+SVu( zXS)cMmv6pyQM@GZ7W_<(XcuP9%%Tfa2W?#tfE7@T=Ha9Ct`!fl0t!`_zUE)0h8ghD z?|$zBS!>5DfuMgq!rjq>^|hW$v=D?NEbC8(K4C*snEe30WcN3>)@? z%ygLHrpkM~Vy17AT#Jr;nF;`MzJeXI++#2&$|SqUP(h= zu}+Q2sCuCkUC6|+saNo-o2K!|rH>0IVQ^yU>`vkghZR1i{_LY4ebcn4$I~*`c03C< zQ|P0TZd0CuwWXt~T&Ocw4>}AQ#a@04s5`EdH}F=w_COKd1CZ2#vnNzOuI1PAcg+}b zsER0YJH|N-q{v6xC8W_a!@q+*z{vuqx`}Una9PRHJY^{+dcZuAUjNY#wwl(KdKxu$ z?9`a4HwC)HGw%j>%G-sfK{*Lia&{5x zRJTZh-))z=nXOS9M5WvUMYqS0!?F3z^IN2Qp&Vu!|BUeJ+(vqnQ>Cyq3Kz0VWhY-5 zayTIB^=*+B__o0UIn2bFc56x}j0Y&@9T!U=VA(b1fijBMLZyBC$_E^Sll&=L#$As1&ufBk2quZnZ+0yyBVvW zU7zeo|H*&~d9GYQZrbtGDPii!Y4b3}JfO&zR5Yer=+eEK`4!#<=k+p3?qL7-O80}p zP7hrd&@CQz*g^W_m`I$f(t@@h&a;mKFJM<>gLjQH7PrgrM_rB#sEvx(YZNx|C1B}dRA-yZ6^~7d2jWe@&jFQvCYy_j zAs=(9l{L;N`4RW>#sDaj7v9W=sXipP>XiN4=fvvWtf{M29?g}s_1ZkCkN}$H?h!(6*r*l`(^n_F;0H=&sjCE!FyZiB*EbEbH@#fFky7Z zI=6(B*|C?~Xk-N%D5jPod#LD^xes|e!DC2=I7z2`%eS$*Sk=4@ zL0Ie(kGziuHU`}B+(FKAZh06E4ALf7#0oBmuX$b&nH)ymY=Pe`nc=o(FYO%MT?%Rm zA0Ab@?zW3k{%S%Lf($)GoTsP#$%HjaNytTQI`yvMwmRcF}m7#{q2d@$Ew zX&n3=VoBT!6J^OZvP-EHvz8)jsA$Y+*DX>eIkgGZ8dV*6@N%P{ELUov6V#QNT3SK2sYU^ih(%7Rw^2U*Ck9EkcUo?n7}36GrVf5i<>3NS1tp& zL`)v-6laJsL*g9Qd-XwCg)MYn_%g?X zK~FZU*&9#VV%0D98#L%?`;ROS17KR-tXf1ivU5o7*VQIFjc{K;F?kfprlQZ2PI)t5 zKbgX%A$_@pN0X~&{uVGls18~vXi?y` ztHWKYn7~4@-u7gvdOe!mFwmw^qzT%cAsU22*I1KIOD_wman?d@I(;?NHp!syZ7VDS zju&uG#@5ldCWJP)*)`Ak*Ux*FM@&_M|qf>_DuHr)Vnom zg;S3*llS3Y|8}2b@)}78q<;*M5B~bMO-cyB&asg*rSF)Dbo)fRcC7nZ*%?X_v{5qM zU}wHx^wxioO?Di#-)&?$iYcamB6(DFrFi8u;G>1-gUrC7ODfct*+H&B6c~&`iVq$h zlqc()@g1a9hGdWpuU32K3vN3PJDi_g!9C>^D@pgks}tQT+-((PnFw6;_F=3}AOAU4 zaGCn@y0?X<>%w;HyqH<(jb-jdbkcuZKLfOY$e;9ks860pF9yvDOu*=*Tjp+7UYk=X zOCpotvFJK)9QYW2r?a?|6+Wn!{t|!6bj)GH8?fIg9x;=LhzYClZmA(#B>BLzKM+31 zy~{le+N62hEfUa^l3b0-1$C6)A*_Cv@w2W4$Vt;*W7E+Idw3j>(pef7=or_f;hcx9 zrX+m3H6msR_qfv7z_}m15|$b7a~CDK9fZgmCM|EJp^gR(r8nrivQsn=7I7AgyI7yd z=3PGCP9Il4vbu?t9hrJ!Z^#K#R+`;H#0<|2$)q~QY?(&~AD8OXWghL$eX`og%DouT zKNg#U5%)~SCOcG|e%N^ZpR1WKTCWsY~k;vDn1O~QBv_uAx} zoEn9o$VQTRiJYOyFfa%Ui{bQmW;+>Rg3tJ9`%PEFZ00xZd ztdvegchR-{1m=eH7|EJh<%zUJARn$06icv-!3qkCtLP6y;k#M8{c%4Aj{o1#@~7xG zfB#1GU*CM^55JEV(~JY5sL>X8Qq|v{YpS)y#znQ`6dg0lT(}gNqXGg+3)IKB1-j=r z4ow1Ho_gmkG8<3bW%~mkt=RM+$Y3^BruJ?pC3b8!jv1Ma8j69a$WAJHy`NTAJ+Ifl zkmSKthe=lEuWXM6kp9?uMc*Udr+q8baHR?A1>g-+`@2nJ;$#+IP-gAU=NL~!t2mj9*r7VcN?7v&|KhQe^<7X2l>@?0?9>WO}uPwj1*vFC% z_7(9+wwM{4SW2;qBFm|0)a}X!bwCYPI|6eT{h>omhqa-z1wPj9b>#7{Mfe6rn&y*# zD#}SE1+Pr{koaw9ODCToOVygDCJ#j6gEnLr-q-Mxcl`{q|R7feHP9=|uULZ`LcVp%yf6a}Vb@JOmi)EU-9(N%j_?i5ibOX^zH4m5v%#ono zK#GDzlA3|Qvvd)!U?P~I#HrV1_n^!G`@T_%vPF~^f#uf&E;@S4gf-I(B0B6$Xkb^^ zj2I-(-+1fKKkKv%vOQlf?Zk;1EAqrQ-fRs!5YWda0J~_l?|IQtC8T14uuU9UMGYJ& zJ}sfU1Gjs2$PfA#1dh>+#2p6D51Av5Sa)HLms}YCLz7)erTV^2YF)|aW-sI=O3_S_ zCMx=G{VsYacrBjO&R*li%*_cbh)g5vz2JT8^IV`lIT?~8Wiq|=fC{Yj2ERQ11Kw&c zyZ~GRL5Dn#k6m4vXviPYP+}OG$Ja|6lsY;NPW-*!rPrf)4r+Ln2%19+BG-F$0tLlo zzarH?`du*R)+R9W?RCImV`0%{V4DD!l>t(|Ruq2&)-_D{ZuS1C-=$5U7o4AvL01RH zigV|`c<{N8nKk3i`S=Tm+WMFc7YK7&ftJn>1a59}Mm{s2ia~1$C{ACya$Eg_4HZP3m(@7o8AZ4*{EXBsa22T?33Z8$1>X&+(r) zLP!qKW(OdT7PG_N&k4v>z}w^cES2ZDth=4q2eP75mo^1e>av6hY^oTFRCtepz4m69 z&b>+9L;8VGw$vSq7<6Nq9N-3}VMpB|2Y-HlV0WNt5(-ZFB%Wk)TRk~3I1Zc5rTZvF z0R`1f(G9>-jHLQa>TX^a9V>473eaY`>w+J9AO&C%TgqQNaalOdp=HQS8ni4tHnf?2 zKx*heJo-u7w;ulX$s23Gw*Q^QKe+z($?yH4LA^A*hAtEJuusBHGI(dB`o&%q$6@3K zoEJM<96$PnmtXx?=|3#{O`SHvYo)go9U+XXEEJxm!Eb5~#Zbz6FS|NJFZ>dpowRY7 zpWQG#>ChLQ8wGvmg~BWetkpVz*E}iUE~vD$3916iMR}s$u-Kq%VGYb8 znsmu>x7_dySh#7G$7W}XmkGB`#)VuNt=R{zw?1Nx0vkFW$`hwT2?L!Ygo+2C0*qxk zWqs@hzif9y$(&Y|LT^#_d2|a#k2)JM^kN$^+}TlLjT44mdo#57fn{-vl^T*2Z(Nqd zP*>Tq`S)SIZCfBcmHbr62gW1D1rsifkh|&G z>GXNU9@YGIc_R7Cg^<$FYF}MoOvqs$TuB+%RI$uivWG2IG>LPL5&Fh!&x3U zW(%xsz2=UC`H7CqKk%XLP}( zK$5JZZ%3S=G1G+~{_Hcf%G+Ox6>s-6HQU<^N-GA(QHQOAviOhJ-+Rvl!2Q4Ul#!cG zTv)Zj3~Y}mMK49}Q_)A2sr>7*cD6IDYgURli&rmR>sRRuaFh4>B76`%n22mrZvcY9 zHo<-QVV{1vj>!!#Ra}ugyQ>D0+D*al9Q(3Xxz_Kvw_b1td}?s+VAYP4{u!#H@`dEG zq?3)0YJ#4lN2WY_jPTf(TZu0Cp*%;DKl?L}7*eEaQsXbx8CEI9k1VD?RiyjgopCGR zGJ7x(NJ~_E#krCA@o=3_hxp>8_^88@YSKqvoOD=n25JO54F7~Cbu%Q>Fky=S6nxrK zE9-1X8#JNTkv5E7qjLRM{M0f&=(H<&D-l0fyhEjpW=P%x2dr>vpO;~7L?U;{b>zv< zcxBR0hi~xfa?DTe;Bd_EVfj=};21aM_3tgV4D)dDw4Hd?v=S6rsV?A1!9+QNir9 z;8np`Qw6)(Vw9`{-vt{{(9y06Tt2r|q`AgAvKZ91p||R)I1C4ztjB9*+0s8*7TIvI z9?lzqAXdt6isq)z>!PpCzbxsSyVd(B-681a-Gln_IJe?CAi*IjlCL8xq0=Ca|1hk{ zt5G&aR)ss99EKpX2d{JYoipC#`Ck8B%SU8|6MMcovxHJ2r2sE!y-^xyp>X~9J@j2Z zvNWTifN*cU47oPi1WST4=w(pqgHG`O^3mpIJW9WPvLo7rjR&FsEhlZRB*tv<(M2h4 zQ&5x@okrI2(f8XTt%certbBzg5n_vnED z3Y*mmVYQ?|SyrTa>V$TBttv;D@2$b#^{O|X2St}xM5xDUi?Tld-CK?BCeZy&bN0W; zawj%Oz)?7)wRSV5NTA3%Y&$U4n0K>A&lMpC>86};1QwPEW`TUw7?~o2wdXYZn9UE4 z-Qi}B_W6wKvE-_A+I6xO_PI8}Vp)uEH6JQrX6PpaHJnFsLGA1wc_z>aW&5Wp zATOa<4uR=PMV=^;(N8|(btVw`AnbxPP7t!{7Mq{22e>#vh+22D_mq#xy8P_)YmFqy zmFzcriL)sMxEbkGG^p^vH=D*g7^N3v(3kvs$yQaZs+Cs&bUR)20?2w+LDuu3JcF)? z+8S6VJ2UyPq>1VB-bI?!S0r6@g>->1dEyaAx*Ve$6MKC7!#{)Dm^kfJl>A`*C&7bL z(cjxk$PQOhZ3dS8lwvPM^3eq?6eW_|qE##;`S=C1GobKbuU9wEc)XWr%6PjbWP+SR zzdT!vmle!D&lpb*xSUe?{oCtD}sxU-v8w#%XrN5O;1it(`_Z{5;Ln_hCLB!9tWnDx-Vc>!7>l& zw{@61QfO-34$P=^OI8lJ)I)}<%x_K9*udI$G(WXyjs~*f`PR`q;dFbu#5Cu~KY6Qx ztaak$&rY++FojY;8+IZUy+zPCCkFh`ve13*ngn;`C+nAEMth5NHy$X3pgv@U&P>iD zpNXJHt3>5snJm+ojsOh^3x^}ZIvR$_7QW_oZhV*Agpia;fBX|kcx4L4p+0=bayo-j zY@2{@fGsbIYa9D-!y*vSM2YO$Av@IvvX-27TtyMOKfm5?o|-)Q6j> z58>~v)EOF^PD*C=@w#~(P@T|4cR~T{3DV8OB4dz^p!;X_l9o`%E!sv5mcjYSLB?pH z_!pm9(#3EAi}QNoU_gc%m3T@KOOaJnv|*uzA!~!=9cncj1HfbBm<%088!<8TmND5r zvg>T7rPz$qhTE*j$FxD{rV26=4PH8FCR8!qm|5kiRrb*>lA=K9Jqj(FfN%dAs6g7H ztQE$&ZSYH(x?I>LEB8)iO5IW1p;I(k#ncxMmW`u7`2L&!eDR}4b(WJ;Laq7BvOShq zd465v#8$!zeztq=VXk^NOE1jBv{i;TwvIr@cDB&e+?}W>UJjlXZ z0U@Age;j63Mui^-dbif`Q~JB|m>DJqwmA5&%Sg^})xJ(#d2q~ZMmb0+z#rL5MQ6d1 z=O|gl6#6GBs>lu+sU-@b$PyZ*L2Tw6xTZ&GEpNYDg(?>8%7TCj6_N*l%=472>1}ks zyjD`lPvaraJJdr;j({|0s@n$f6^Y)fQ>6F0D$@&&s1`6sm^flTFpU;H=H|$LOC{^* ziTPv8yZ>QAPxW_i`H^-fULY?t%a-4x6n80di;CV2)hQoeWg$=S_JVIVDw=|>ffCUZ z{;>&hVzeOLytFBpNQWE|WN)g1kfFB$3RX~#H;aiO4+vg$g6xHftxmou02^u3rtBr{ zOci9an|w-njfzdvn$-E;Yoe0p0>u@a)zJ-14fB}@tILf~y&RPeDmKk*M_4R_pMX!V zmBfaw4L=9v-^*N4SFlcwPgw5Sq#p3E0^(<#djh+HO`O`!t`n8{rIY1ztI3j}0skjH zJ&{L(bTl^bT7uKg5VJrWhh2;5&^pA1S#?vDjd5GUOt4GyOW#QHT**l@J5WX`4pNXz z0*WtdrhCZS~#?6+gy)-(+`?YSYn`Ol)Z7JmfJSbigc) zl1(Ud)EhC>X4&+MSX^}L+QJ8;S*$y;Yh`7ze)+4HCA;QdWtZ^_y;m^Xyb@J8K6}O} zhhYBY+k^@8TgGHB@BE);Sqe>a3E()fc>(dYAyMIEO0k)OQHcH+YW_eJ=OX#cqj65Q zxWEUS{fGZEqPKq}!)6mYhP?hG`HYd=^M{KK9iy=rPP}lpVlgsQW%PZI1;Pz}P#GKr zT?8osCwU8mV}^uhqhuQW*b!0GIj1i7keeE&^jz zuSgmr1d7l&KGuPQ*}yKweC85^nF3r3+qDSmC8O!A*tqCT9*;)Z#)*9mE6O$)l*3xx znt%pL4c!J5vqv3S*J&LWwm`!A-Hto+5+HGC(b=emrT+sbNLUVmbh;A+&Z~)=5pau z*ER^x=>^!}UQ4vzsfMzGbJ%{!i{%^79i_`5)VIVnDL}_76<-YgJZvqSz>5*?;VpGN zAEB2nbM2hi29yeqC0f6R&|Lm<*DR)*K;s(RahEO?pQ9T>>!b~#Wy)uFw6b@Z$8$R; z4&G7iy=l7bUW?6mdFptKYGYVz40n)ItdE+f`7-|-CUlMK{_W@F2sfkS#2)NbGb}bz ziZc{BO+_dBb8vW| zwa|mxC4xR4Qi0raZ-_{rS52yYpT|r38F7Q9m<_v)4RkC>#!IFF1CxAo$y( z3Dr6=uB?P=cYOV_vJvHeVw+3#wuSg znjK6WB269Y9eh6h)9L%6S=N}FYEmBztd-nX#qrlDVnVJ6wzw^aH01r6cR^wyDWDGO zXEKFnq`M;WM6u$NqFg3Zbz`PhSvd8MV7uqq(0ijA_hAOYeoz?Z;iEXjZEo9`x-!nP zfy8NnW2LiY`}ed;V~&&VQ7LQXb7)Sc4=v2r$mrPpW)d6~)sJMtu!1Ad_L4@X* zu#0~5mCKTDUdg0nzl~EmMOqRsE}3*)_D~iNii}3bvQE0hHHK{tO(ZL#mO=?YLU@61 z0&Bdw42oJB^?T@Je6{d=DwE4JDO*HZ<)^;&K;2nmXd&Jn zSSK$NwKJVin1sJ;Dubk2b#4nuA5TZGV2ju`I$pdu#Q5(-0mE0rgueE(m=8+)OqMEl z;9vHUR3{EWm6@5)-IM}4?6Rro_3Q%SE_ockEvx`Qw|;z!yXNeK0y(gE(w8L1{DFvg zp%7Z?R?2~>6-26vY(H?X)jS|e(HKR1&abaKf>O9~n;xi9CPwuu$On&spMy=vMH8W(n}rU79vScNUp zBt@lfllnYKp_ha0u`^W62gq`d3y>`Db>#jpzV+wBmT6I^oz$!(I`h3R2URi&?mB4- z{m39rSUa&uyky#9|A$jo`&W}S>ORk^ps^foE~9N{%kV<5a55Xzsp|gKLeot1e!=*^ zlGRQ;(`1{?G+QV|B3iBJGotOh1Oqvn1{E)h#d;tv&k3vpj&P_+M(@3nj}@KQ*(UkT z*T)DWoZw-YN?tgsc00)l9mr&FzTC9$Mjf z(Jn}6?FR=|FF41HWdf8VIP5m~FMALtaE$xl^bh`JqRDpX=Y?G+$=p&|PFzw^Y-VY4 zCrz%pRXk1UH6SP20 zK8s~>{0tT9v7+w4-yLR#IAZHLFZN{^8R_|jpZv1lvH<7#Zc8V|gq1QI>2p8-xY-g**Tb zy=9Xie!?4U201P6APwrA`F&qIF~3o*^@}6zvmo2e;PFDY>$7U;Wv+Pta@TgIR)`FI z$$qDl*F5wQ56Cjtw(#YyH8eiguC~G^m^^0{hCg5@gcu%U9VWxIwtX++_0%7l=AJ>F zb*F`n6#@8|kVlZh)-f2aLn@d-!k8jnZX}3nhrt{ZdUirdaIZ(5ppIV(LU}*rzAFUz-bchq(&O@7bJJ9ZA#)Kw;doRQvvJA_7CLyLkuoVD9iDt3O2_of zgp?Mr`x&3pq~0XH?ydJJz%q;}_rrG3d&vekmp*SJPwTmme?oMWu7@I?8oI#?@++Qw zw(3OL6JW!`&hn^jFyscXWfxA*`=aY*rz#!^PJ<(Yf4`sgy#^@o`b!o-?r&Gw3ZMb~ z0McCI%aVS1`FNf0X<06Sr&dx$D*5$LGuEUo7+*3eZE{gK9$({EG(B7RB+Ras$TVzr z!HaDW;B*x?y!>nLxrGx<2B~?*x%bJ2FBm~UzM0FLPASqTqNSn_sLn`xVP;E(_QY6T zcUX#llC+zb6|`|;Ze*DllBcp7I-b2RzaNSFXl^oi`3$lE=Lb3hEoMVFQs|gH!~r|j zp;PtCt1gpFF!^rIZ+=g<329sP$0evf(R5Xg-YL!(}4Gcx7F=(L1M^O~* zq{JXiuSd_!CDR^G1iiKc;UBfQXG0nCfXmgeCS?Qs0ksreaO~VXzhd7NFk=B*gb+v}(mXTfr;&TyQL`UzO)*}KU zfCKcsNTa~QU7kL|{zGnMthge<%46()*g6u1ae`k9{L+*E`i{w%+?Z8)n;hd7fN|oS zPluVAIZr9hQsh%Ax(6!GfrO;QAl-a1Xq`ccZ?8C&e^z>ILhdvjwfc?Q^U>Sw4eO%s ziXM2J_Sqxuqf;Tqx5G0x5|>zQg80bp$Yow}{1{3TUE2Zsw;%0)7+}Iy&713rNxBm!2CB@^Qb;NGP*B+uoeip?x%2bnIZ^wW zD$)=E-pZVIdA08)Ru_0lj73+~yu;8-xldlo@1hgJOQ+;SX_eck1z&0TZs*tE$zl%8 zzZa0HDxGqUDHbP$H$;?(Hi|oCM_C4%&hg}I|N7zjhZ_x?)D!1 z7Y+v(WW4Y}c0S1o8RPz@`PVlsGmWoEfr6F1WCkyfP6L{?2M ziYgOT1vRNlLu+T?nONqU$Cv?%3z%?nu`_0@KJJr~Wgl8DjGWffwXz604gr=tV0>vb zxF=~I#gZ*^772IyA9SnaS1RvGxA@@<(@x?hZsdJB9~xKuJ0VDyLSL3FhZXioNGcV< z6F#}h=EDOVg$S+yc+SJHI)tM;!O5CXnKd63DNOT>$D^JLWRnxmGa$Dzq#IX9DbgvD z29)y?tD(9bbA?z1u2mv)ng;8NZ+dO#;eq7v)=*G9cU#34$)LXN|Kc)oa* z=lz){h~{wk;cy3p2S*(beANty5;Bh@8s*^2o zJuP11lflbURf(6nj+t?JHiou=;@S0MwKZ4q%T<={>hsksPCWEhT-e(nM{$_9U9uQD zktb%BTNATc2~W*#Gt1si$$C-;W6hT#5lJ=E{$eq8#-Up{+t z=r!*+1~ImQVyJz1ZrdR^u|D@Cr|Q;QcE&yr7bl)gtaQtk&n^&Qk5IF8b6{B@Y7;LE z&`+)kte2twFiICY1_8qlA17Q4{|JX1wT_FqzxsZ=W!BhfSG`s;#JC*E2F^mw89J3; z0Fkpwrk!0b%u*f}Yuca`vXWUYZ1IEji&k|8$f7le$>8b3q*>bH=TL}nf`p}&a0nQT z548@7;SxRj7XALMmzxhK&Kg@WAF<+Pe2j&GW@Ipuf_FtdmT0g<7J2Rdf3v*_f7Y=Q zcw+OSw@k~Iol}{+q++;g0w?xDE}1P->L|r2iX4Z;jTjp}o7DH^-E807BxxR<>9>Jb z391wurflGSD(#j$l--!wNpI)X362J)(3^S3{gNhR`M+QBqwDjbtm%}pOQIJvsSiQ1 z)B79PQuic~%=+LAbO;@l_eAFMTX}K(n)$Iz7K8da^+42YxatsImBp-4EQY8|86=PL zg3rxMo17PsHyQ7(nz}HY<8(GE^J8g_UbxTAR#}JCG2O4e`$eCTE`m8pk|oKYvIMIo zFEW)(90)_8&_ofh*MKHXh4ebJeq!M)P$ZY{cWY)h%qa^smU26EJ{nb2joTP=yxKY> zhMAGZweHXBwk(h2lFe~qR~G~^hD2+3QVK}#=umqtf!z+3xmxA&@PmOG>}~P-96bE)k7SHLDB!s7ipPjh-bxn?Nn<8;)hst>SgD4aPK%dA#Jp%SHI2?gv z)8)JQ-<7FMGu83vkJpp@S0>O_XNH$5N>NUc5-NJ*-2Q3h(8oUD^4M!<1P~~wPfi|i zSroE~+2>Oa6(o8&>T(UZ91P5#y?%VD`&z%n%=+;dp4;nlT%~ChT>~#DOL80})KNJC zrBn(cL7f(EEuEht?w3QaJSr6foPt|I>v#h$mA+4Wnx(63t<$zXZuS7taFdnwPqgjK z>gtbueq;g>yDH#YSM zGDey%SrRlwCrLGDCa;Kk0-{X$P^XkauLW%3imZ*!56A;?q@@!ZLpuR4nY11%{)>3G zMTT6Hk$a(hN}uO|i;mW+jB_A}Vv}`zEQCNwA8cf|Hp`^+F}j`bWRi<}rW&|x-Y zrM;d~#8G4osuXMyG|p+G7Yd=F9`x{5dFp6g@MCGG-D|F2J~B3<;^hyr^A(&>@mlS% ztmr?OQ1SP*%V&`5PK=7BX3Vg?l;S=`x~OO@TG<%^%ibKx!NAziY_FSqO%ZQx@D93O zP~fYhPboJCXmAq`c1QMlASuvFb$*Bj?*zv0h)Q2HBc_9BK@!-Vv(qNS71)8)EIlwo z)8%n(cG2X0v+HHZSGd7XFEticK^(EqvxBY*+%4RVsjddb_|O8nG;{~uBB~tx1&~A$ z*YK&~D`2@a;L>MUr)iZd#`ltIY!x5Z?B%mHPyDxo1KdfU2frKT;Pmo+vyBheAiZKS z+b@4bs0<@+)+En1K`({qqy~hOrEuO9 z;xGPfQHhP~DSU%a&>qE!4Z8R67doFK9@>$9^|n^fj`;Z_z3~;n#0OI+#IPpYvME03 zJlW_<3e9XwCZ&MXb1D^GI`2`K7J~H}wAcwOa@0aEa)zqKJud=1jji5B<=tHqjI{6n z`0&4XO*l(3=tI0TUajh}NRaq!anI040Zm#5M-Z#9jk_sDtwWx%vM zbS{0$iCF61CfFUgEIbpaiI5-BrqCJQI{LODgj*#(CrP1KGHWJi{j$Yaj*~||6E)7s zi$Ir4H)*l#D*XW1?y(RwS(zkV;sdk|P$rSv$NZ)aJFi?|*Ygh-r4Z{=J*|B+H3sry8@_shCpiqeuZ2-2e;nCt(=k zM~=)`DDNwy3w^uUCt-Vl{U4V7K)=A^6gJ?3wGo}NET$Lo%J*g&Bv_EuIX+aYMCAYt z@~mPax*Z}meIDr0oC&=IN&zd?7nuT)-9R-8TQcm7WH0b=!%jw6_Yc14bPQNbg{%vO zzEB!$%nxC|td1^(eCb%rTf;DNL596X-InKYLdLk=4;BS}(Py3|(^}l`OR7oD++5&U zZB#GjpJTcV98*o|b{17;TgfhYgR;@5eu`Gv&a{fIdN+&-$rKliJT(`#4Du+i;e-(? zc*~*hNK7-2{FApD$XagIFiz~?f=11dRa6S4*g}y+D*7&TKK>mwMJxGfAfH|*h-L10 zw}ckRktQe()aMpbS3ym{crKf_n933B)W9toa?f0BKVx@}5{wk!j7TL0BbVQ+=t zR>)36(x!(jmNltk6`jEsRXgYy;U@7JVMkbxC_lW0f5xjKDiwfnhrY|!!BzQD`QgWX zbdZ(C>-$~s%Gl7y4pW>N2?Cy47TmD!z(0K&H;Z!chmQ&^`3$*~jXSYLvBG``sR+eR zuX+KJDOW(VNjG~D)X!q)CPCwDo1lv>pRyE~5wXF=Zm#Rov9Y$n5lhw{WB22ezOg;c z+ce)a{QJIlNum?4bc)PokSt08GF98D=vA)yQD^8y0jpiJpuglQ@cSI4n?sw_t6cY( zjxXkCF=yTDf{bSx)H=E(ph9)uJq7M6h-@WUOrzS7Sl~GNW=6zORs^BM&61p3cy*ek z-nrAxJZH_Ya+FdWp-2@Kjq6(Mbc8yF+t5Z08^%R#en@_Zq2d;^mSur^#hTCE@<1E0 z9O5TQ3^cugPunbw3DHl6LIhYo!&Su|H-sAOy-rtU3$Z@?p-;O~6E|)9tW+>JD`ah+ znp7s8q^sHlNcORpA6i z)GIP6S*ft}lskI1Me8b<8*eWh}K5M}mS84A78UY{bxzJH}@7L)qcK z&M-|z#le4FMsi-6#m_M_b99hWK=tlkD*Ckal=O-e|87t>!EzcXVCK{bvb;NG{Vq4= z#)nh|^}Dn)=h;qCHECxShj#*YPD9J=Kd!QeAqSAf;N>ZPd7>)8Mds#L6Tg@LvpSj1 zy_?seLXcGnig;OG`OJx5C)h8OZXA`HitM6MBw+`#YD&Iwy8zpjdav zVt5;+fJTKaR5X@7pgS_)vP-;1eAU}1-$OwnsUZZREIou|mAD@_?GSvi0$nzk`V-OXaPf>!x}?DcO_chRTeTx{qb z`R9SCKB4uilR5x#BZQcP?8!64^^|dFOJ`IKNS2of>I9$5TSXgYcc>cFNpImk2yQ7i z%w9UHeb#`#-Qi&$aM(X6W8MDz`JX?3V3|dHMVwqKImbU7RQAz-NMMuwdYBDL-H zsS)FI<0J1%?nAN5AO3KHY~;fC|LjG6e#>5{dS-*3|BR6w`OW*Ue`%U&o^lbZib%#Q zGfy2do2T|t3Lv@KMMWP8dJ?D$?2vDhE@Jl0*}xP>Z177Xr$SI7B89F7e|8bmN)T61 z0yCgD`><*^gLgE@Vtsaa;#QLy4U=BbAush$_PawWNTH-q-YYpL-DP)bXS?BMB;a_# zJ;HH!Z~n}Lo5txEUL)tZ;l_EZ(a8fd+}xlP*C=w8itb~RpfS2mcE=4%U^*h8>HYTCUQ22}l^EHp-#M)$j9@by-0i-FtPP{qLko;dHBnnQWp3^WSW7`yh~T8 zj(830$lDET&y3Uvww?#w-EzM_sZChZhfhi(2c3Aqb>0j;Cn&{HiX5S$V}zgiFP)t< zvqt=Et3h2RdO)tJ_kCscn`_@V^o{a2(NoU(a&FL3C0?8Dj~;RjD6y4}KS2_~GF=XB z4n0avDUL{+)VV>A0*yy^@sima)2n^YNo#-P-|wSDkw!MMGjEWy&lWm^mLXI+c57;0NR<)w=(h2E7Q#VXw=U0{zwKNA%fuXv3E6 z!*9dG?0AWF+wxE2ORj#=W5~Hc`EWd{6X>Zr1f}kn%S>Po^SAk>xTj6&g%XHdKGK?P z5-0mLf$OVvKn#8V?KtvQ9%UU6!+9#Lzy0eXOY6ZUuH(dOEGu?p zgI_0ojkndiOO{3qDzlh+=oAe=yZ-Hn#~veYJ+d+!FFVWuMlXAmUC&?r{8($GX8!EYT;`Bxxl)hV(QGKJ>qr<;FY=;M79aK*$x2~{-WSxC!NaVGTkg{ zImYo5!i{HY=-u90Wh;A^iJx$esgs?ygQLp<7enpK%eF_25pL$?<81X$0|%XxluzPG zrW2zG8105QBKs&s0Y!EruY?ws#Y-fwap4t8jq0%>rSx+tc&Xcbl|7zvGEg{W{BSCdCt#a^K$y>+1 za#U$oSH_;e88uplJb)W;e$*ievZSo%GA}vt!pI7ly&n2+Q|OeCyY8BDQ66Nf>Sd^g zU?4;sd-e(2Q8Z+T$6)j8p;yACgA3-nqW4G~w*|8kR{-aknV(cjv6Ui8K!R2;dj|=u z_IOuK)eDlCo|#2rh_!{(%W`=agW~wey=vTVQfH`;OtU)hro-yUhYYdp5E(it4S9fr zHV@pp&}13PeML-%mB8;F&wANOWoxK`pr~G!z?RXy9yexcm7h=lbb3uBQbSh-0bhwO z7>R&m#r3jN$_i;aTc{jxxi-5dAVzpU?AfK;CHq1&y&jePI|S0=h9^LbIACjFzsq9g z34Z|}*vivYmHZ4EOF^pwjVwqtkr)WRE=e7sxbeu)G!$7c-5wF%j;xFhbTb*8nlDv+ zOcrx9I8MAk+Gb{OHc*Oqio{aU=~Iu&Yu!@Bn*O(-)+VBAPPXuln^9839&t2oqhK>E zxIK2%S8w+F&0M7k7U~nfEFo#Gz(F?Df80YUz@q0+(bxmFBq)R4IC10L4%O}sUU&g7LbRpC_vY+~I2F^*Hpm9Ll1 z+v?S!va8Q%`Z{d>vS0M7jryy7{sG)vN2e8xzwqT_2<%WD3A!KU%o@uN!JEiik@fHY?w*iYZHP?7$bPu^hH-M6PFDZzOnN=?<3mtGN%!hp&msQ0ZwDwLs||RQ_^=Ei?j^JLJtwqa)C6c;+{@fA<4RVut6dIGq>kB*Sd3 z+(aohP$Zs;HY&MQ`&N5j2--uJx?dcR%}B9KwQnpid2aAqK0@zuqyxcW{73o`Bf8JJ z$*cVO#=Ea#eJ@hocE}fc7V*+m-Mq_!cD9G!HoM;?MtH)fhh9iZ-FHTu@JR{S8BqY@ z!{-f!DNDR!L$^*oFnghliDv>7Ws2CgK~*dG*hIBieAAcD)IjY_10gnFw{QX zlhz5k_=VFKPXEg#VY|HA_q5MZCKY&mH+mETa&%y83Z`Rz;Ks~C`4c|g{Fi=`9UbO!LTs=Ky9bj>bF246g&qnX? zES^*7zy6I&k`8$$uRF3yy~8t>IZGac7G#6E1$x=A>kW_eh#pJ+5zG8_@JgEo_35`b zf{l>^Vgw(>=@#pN7%m&V<&Qr}e%)k+{Qqz1J7n`0%nAWr*^m^{4oU%j&NeC<`xs+{ zeI7M*BB}PRn{q>VNTpT9GHsp@Mz@-|J|B`=qD$uprnpmh4p zkVe@#ITSfL*6|rBe9&HS!UM;Pt;6S?pMU3&iwP>Z4}br6663^~MXebu;wZ%$imarf zKLrjBfHkFv8-irT22Qm)hP?5 zF}`=foh=a{QEJTFEe4v@OLR7!?VqW-=Hc*(#C+Urw+rU?IeL#1FsQG5O7)K>Y3Ri3 zXe&v>7-4Q?0aF0-i;w6g<`C$Y9AnzCQkH$FI6JYOX>oryN0{&3%`5cX!Nf9o9xFh< zaV)T9JD`TRqK-xrJpZeT2S+VqvtOuyij_d}`p}916fNK9lPZP^>Vm-aq1i%gxK9;d z<6Tf_dc#^l?nDRHrA#z72!Jn}I|#-A0r-1a4Y$5o8CK(N@;3gaIs0E^xf6REz$7)q z3EfO75-75ciY^Q7^F(r=PP!!&D@-)T8v6tmMTs6tGsC-WP+qm?KyZSAJwNy*&$14K zVZ4rU0kaza)lyl;X}u0$<`@EvcuEmVkyVDUEA;N|i@M>fSrV`Ys1mS+dCbhg(BZHR z6GQJ9n{95G_~V~`@jJ1}T%-v$wvt#UHW#^On9x#+WQuI2qEqP0l2krwH)CN3R!!rP zHek9;lBNfP@fbX~)jM09G-m30wgF^_^WfM8>QnS_0?2FM-0dRyt_dJ-9_s!HIqt*& zxo!rK3zXsY;Kha>r7PsIpwS#Zp^8-UYv@iOcUk~B zoW#KV5M;!Cr%ryCe}?{C{!Wv+fk|bqhSkt|d7-?Y#kQ|>6>8Vr7IlZ66;;T!O62Iq z>k`;267*i1l&ioxHK{Lx@@qfaEILEChBc|1lzX85`Y7~+B>-V!D_ay;>i(&;Tar96 zWpV|`vf1p|jsRR{NaMQm6g@B5-~=w}((ex(8cpr26DN|b)XpMTw?PL0Mg7|7n_jCt zH9I^nGR;!dof+$ij01{ZGVb<0y-)`wqb8%{65z6sEEz7k`o$Z4J@6TT21G?r)>Ke# zWNS#Zn_k*MLmM(F6I~r^)mPg`-@c>!vZqnoCEzgnOOXqO*XWzbu_sse9@)CsOj zdM9M^ZjcT`d-!Rg&b<;yxsK8`l1hHu9Ia=gc=Ry?OXY6+doNZ&lT* zmUvy@#RGv?3MAKbya!QQWiJ0gl+BiW_;17Zv2m0oxAJ>gU3d<&;&9#)_H%a^-?S`> zblRnmm1<3F!9bs+Nxdm(iBC0I7rdIcH*j-srYenh$@?O+Ecli*bxMOe-y3UHdYHw` zE@7{9S@5HoO9PGs4bYA1Ua|=y`fL3%>4kiZ>K8_R>JWuIyMTFkJdWz9bwe{8nCHtQ zcZUrwmHg5-l00tp)z14ikdtOMw2V@Kg1~+%dWmaNz(-rX)2OAc#ZX4y3H|N2C8f&y zs!OUOd563~eNc&I361J>dXxAt?+ln(t#Wx#?eqrq#_+v?P#LE}a=KhzB0-~z6q{`V zaCw!OQ(Wen8m^ZvckKb!S_iAPMp?O?z-Je@nVZ#7!EEs6N59G5E;j)vWzrx2L=s#< zGHU4Tl|d<>)I9}MTt%nFTc<)b8L9MX2yd8;m1VaAbac{$T9P~MD9%fPhkbx-795JT zvc)n~>M$O^7*uKBF1^9ue#p-ef5AhZ;fPIc@3YfZU;oGZhkI!gPI`XicY-W*;w5*g z8GzSOinSD3O+{Y~jQ5NY?w4q4XKanSF}*;gSs~jh+eR-U?~c+l;tU*{(aQPBqq~6< zTE|tLPz`+1s|b*M&A|ST8fU=!k9__KULx@8OC^SNOsj+FV`r<7N{ zK`K+L#6o!;y-=6}bZeRfUV{>^+YZghWcrWNy* zh1uSaXW4VpIx>ctFUFm3cFlS@C8QHiD^@hUH;A*8g;O8ML4+t~u2y--=ZpuEbD_BI zcCX8jM8-dQIpS##)2z_kr8{M~vo$I+{3262PjmLGj+}-@1s;2%#o}4k!LxSNMNzGx1rc1E<(sK$&O!;Tzmi|^RDYbyeVz1e{Wts`JK^YQve zwT+KWrPe2dQ|^<8E}S0V#D-;)+0?XxQY@v&A}ShH-jm(mHU9UZ;lDA)j;+WUqY8tk zSllbvHPO-^by^Ey#rI4BO2=cuY+gH?_ss{8rdtbrQ`x+a%jo@6w%UO`?df-A(}v9h zZPU~2cM+#IoZ9)HJDV+gg1O|CoY+KI=@woedIYLMO5F8=oWMMoW6)X67PisZ;x-=8 z#VQ*d9;md+rkrHAN<*%z9&r%zh=<%!fsvc)y?AlWeffGmBxqW@fQkv0@h5^?DH|oV`2yv z+jE+whxmCBV}O#UW6H_WJ-y5E8`j}+A?!Q;mXQ)Jlcy7d#!6ge34fP7j*mflEK?eA zLFuMt!kc5 zy62e=Kz@>8yPAfZXaY$I1mLU^43^_rU0$$rKQGb6GhN&*%Du{Ba!P(-AM8vKnD_j+BJ_gGT;=9T$KWO5Z|JY>3vd81+eGtP@E z-h+%6OWusNz3zT+2yMia*@QKcox#tu^{40CZ=d^>2~)Ec=%;O*gL#h-L9A ztuoaupTRz%3Tg42^q74aYLEu|u)IN;NtYAUrAzWGgJcx$ArV}*_=p^jW~gxCe;@$v zL;^YdRZfsK3V017cQcyx<6GQI=UHkejuqZ)z^EA%|2FQEyM89TkK3qfxauBxdi@mHw%t*Ds#sz zfz9?^#b}NM6-tWdV4MKiRgTKB;r){LDk$oMMgY60KA8;*?$`%1&pme32g>XB@0(zP zO!JI$?~@HqypjTy)FCA*>68Ne9W51o6IS$Cg`i0z>-c(UvvkR{#2`&92%%!na^r+L zX#$I_9$IB}M3Wr!fS|iWtGpWm+>uxS<=7gy(G2{kNB0E}uqZKNP*xSm+mMrm%#4vZgJNv;n#bC;tOkB(Hjn7TVe{hN7-UrvFTv` zgBx}9As-#RWm3*W4Ss5avR<}ZxOrkHOq8c-oXOTq$PY1e&t{8t^o^Nt3)bLh=jMlG zi%)}^Y8qYSox$q|kr4PLR)dz+3c-N8rk3d7*Mf_)m|jxG!&rAElzn6MMhCsct3Rm4 z{VqRG*y0ZQMWaEg4KvkbjL8ck-~~`_Y3MjZ0N~KKgr>P4iuVHHrKnC>{e;63hpp^KJNPfKG^0Aj z%}h}j`__HYJBiMbcA~lCo#>`PKKJXAdLK=9Sf9t5@Z`DezOi0r+?Dfnno`ZLmy0?f9MB!Gm|H1&g zXE~LjdIYl|{z8itP4YbX883~~&R3k^_u7N)JHBDrUG229o0UHLYrM_?jb4yRZje|T?^WxVJ4B$x9h`-zpP&b6;ymMo;MvTgJ4 zgGgdL_2Fjf!)n0bSc;4wVn|D5^6gMbx z4WzTk^@tswt05$)Rbmxb0hohSe({_mK}ZJGDyj#!_ZaDtWYXADfQ^ut2SC?2g~o7A zrwoHE*br$5DJ~RdNi^N^(>^+>J^~vMw94BP!I$37!yC%zdI6>_+F48_tO|nFe5Y)M z4A=S)lEi$yVOie}l&S@ybQS&v&C;_IcfStGdN(~g;UcZ_#`JaulMy<4v-D<|&KK17 zm}?##G_s`?`VLR9$Q?kIfU?W}b11WJdxi-tz83g6{*a}5krmmWX4#^Uosd3^ftjI6 zeN>5sGCCT+$%OC{)se^~NIWic-$7sf{-bX-$gY0>Cg{-Jdc6p81DJ&DW0MqxP(|a& z*Lze$^4u3`zXP0x#Ca`zdGQ|pc#~Zmp_cuIY{}o}*QsAZI2Q!ZHeK2*r0j^nwog36dD7 zSraw*SSYbb4adA^EyR2VTsFGzn%l?rdK>`?QX=d?w)N_Gs;^xvgahcX$vk6j1 zDL@G;jf&3lHk8!tB^!7N>}g)R5_7}IXlAI_YGAOMW}9DLgdv89aHQVO8w(@7%MH=$tr2cf<@rMU(=#iuM?@2U;aAYHC%FCF!cc;$djTgq}<&CNG&~Od7O+ z%0vQ`JFJOHo@-bmVd+gDbkJP#Pgcgv)f|`ve;oh2de{I%Z{_R_qDvVad4hZEwgSdh zyxRIadtJu~uT;(3`d^J^G33Oqu9d|QR&ruC+u+%r4Mkd342)qr?<0EWghRZALfxDb z^B{QFENx=826jg5{$E{7#k8LIva87#QS0B=kOfZci&#_G0}_ZrUt9(cxb*wSPw1a^ zkr~_c;NaVBlhqi!cnp7XvKr%la#_(MGGSsx@|FJ}Yg|c=nH#v3QY2Ai6Y_I(vlh-@ zJm~}i)5jhKq(cRbr`;?@qz&iRkgB0Um@#~w6ho&IT%E@NsK1TL<=Y)6AIpMwJl&Aes@xtM> z9WpO`jNMOu-*@HDExi$^En2eTCLud_vlLl(4e_xI71DfQFLaY~pU-KZRxllvQ}xn2 zZkHwk5r_lq!LXxZ`-l&khG8FL_k-4r#BedOoyYS$|F6kgeRJoLiR3eG=^Q7XhL)RE zsPs{a2Nb!7PPb7%0N1)G;n>I17^AFRQCSR9ok34D7Erb`RY52yuxqX+TX@F}W2$v- z?TqFDDG;r8&j5~>LLmDvq=x!D`#iC>^15Or5a=yYtq?vC#0cA&EfY3lS%sl(zt0n@ z?hMgfv`zyq*f?txJ;4t%e&m1)lJKpcxKd~c;ofme5tobdM2C3oOdUuLV+OK~hw(=c zS2`Ptzg2+<{zPu%FP3faTS<uQ;QF=Hr=(zK?-4`_~ers7C^}MI! z#B-CCBB-Q*zi;&(q;2S+{LAw*R9XI5akJlJgSg*igSdFk zVQ8YgEn3CKhA!aO0NIS5)(4bN>5}NBar{#ttcTemRKq~Nt9y|Rp-|A2z#97sKvR#w zM=St-5*@7zMtp0sRNav=a|Vl< zVvxB8+Pm>LZ1b$=b%e#DWHG(lS5qvx$?Af!Pr>jD#EPA^2+F!8ESTi}-~YR$9ecic zS@Q8!wp-FgUz~nzc0NCSMfc@j8#<2}mT0cQmQ1$Sr!E*L``T>^6NhyaHH18f9l_ycT{+?c;xOpJWFo zTdP0qGU5LFUm{a1)!DfC+fF=fLSO9=2BbJjv4$cmsp!SfHj4{0!-V&lN8_Bw9x-E| zSb)K^2@-Z0WNr_s`g6cCOHbR1L)#|U&O7aMH{_H6-@@E!>mp->Wuk0erB^$9etsF< zsCdFJb^k{hy>a3Vp8C$S<02e5R&(CC z@b&@<7sc3%U?jZ5a4!mTY1Xp~geVEQiCG}5pqCnYI5e%IYtRmjc}x2N#%3s(O@mf< z?1mFgmTugZn`d@gmb1Jf2E~eI(;&8>kh?%$I$1mA`jk^ZT6o^OlRoLtcI@Fs*k%w7 zce4YJ?fqLbYuOX@e6+=R9VlxJOtsT5N$-k|ygvA|L0u+#Kv1NmhQ6W5_CNA^jaaMP z_IjnS{S3v+hRJ3O7;VSP9%r}n)(wgGqK{6iO|y+Yd{PoQI9#>1^LkO_yx9zNf>Io% z$Pp?U8}`tHMn<}2{0h2GewTmH|1q==u6m;vrWzzdg^obgwx}#swm$}Z7X=tApUP)n zmt`@xWOcmrq5|+xjA6D0=A>u`QxD`DNOpHdw4G<*FB!g1G z%##A-^D}a$8x>+x=#$EuuwKWQhz3DZ0TI&_`Vrkk@ZCaGdgE;Kzq z0+yd6k*Hu@$zZ&FF|*chOL$ADri_1woMSL~gb{$!h?TQXg|rDWR8^DpbOv2KXH>8n z1!k<6lc9%gLC5gdSjWsTQJ>fLZF#cKGPuIU<#pnD%}R_2XPtT(GO%iu@sV{E~U?lFnO<8YdW8-0R5Qb$;LS|7VkdIWa%<2jq+s8<;z02BwWtfUM3% zDjGAnx?t!`g+^7Q6QWhdGMK`HoS$C@EO9gy0w|*ke(Z79>)zzSOERIgC@wgTECw>M zSfEWf;n#1=TPdO ze#M$Vwkba7JlV*th{B22T+pvJ#1G1(6cEx&rJ{3$+Y~k87EzOW51S?~_9#%d5?$~n zh~QmhmU^_ZcK}4YJPSZXwT;Fzc>G>uD|;`}c>HC!c**d*?DWWC^|B|A?);B_Frj%g z&C^Z{6f4csJ*1QEg@nh|;egU1duQ6vr38b76k`+iiknuFD`MQOTSp zd*uKB8#RnHmxY4!THte9?;J|8gCZFqYaP1NT{}BTuvwZB zP%pUS)}+pnBu;H*AIekc-9V(CHU-aUm3R#>XYF9#X;Nb+tq|J^t&vS z!r$^vhqWYec>jj?1s(Jja`FtF7cEnE{m0|!CbYcKnzfo_zcMZ>#C?ajtOqCsko^`? z(F2Aqq=fLTlTV9x&{-hZep`UsWxU6d6j~o~YWxY(q}~atkZI!Tz!)J`24m4%Y-pck ztF%ddMY7JVhpAT<%6lc7#P!Nb=G25Fptj5QZzYXB4eFR5KKg3$HyhPgENWT9I%Oop!i1dLZK$1r(LFx1 z(7Q)cC+k(za`60vxxaPkm;db3Gy3Uxb}UyOVK*HfOXyLy7QN!g z-A)Uj?!)4(cVn&o?)UcX%%n=bF#ay^=njj}9VF-)f-(>Zh>C`wLlPsWC!mN*<@`-2 z9+dY%P!2_Z>jYh*;`pNY8nRECFWxBLIOSOQ5^8IBoeXa&ia)NnD#_!_-#qC!IlmEV zf6Mp};LaRXE5v!K!FO6DyezpMh$B>r(!(m}b_n)OH%aj-$VkK8P8u^8q3Bs$Ihkhs zez$VKo-OwaVtE+2b+XHG=#pgI>#o)l{u?{EKYpJ;rk+65T0q9tS!XSb@|r&yfN%?QbY((q$sM*ILT-KSGl0_Ma? zL%{aSc6|ETj_lm^%nr`s&rox-wF2w+Bb$25gp}+XVd?L0u;j^yG3y zhSDPXV5wlM<=~Qp(c6o2&eB>X?z3y;i|HY+0w{&s9gXUomXK~qcv<)k^0V-@{+bjj zGdd%>CG!4^%_`(DgRpN6i3ju$cUc@i=yifN1m(Y$8j>~XYV_r}Zl4ya4r+YibfXga zJG9DUFc1WSdH7pX%gN&$h2DcSMHNWe_xa%GgQh^l(rDD|L@iH~;wV+XK}Iw*QbpWM z4qk)*)+GK&rNNFPZiLmyb%&mTCo|$^F8ks4?S<9YEMgvP8%y`Pmzx+TL>?Cuv*zU+oKVJvJ1>P+Gfa!um<)M1!D8 zzm7T~%9J#!H*)TRa7exKf}(hKspPD>B=)v)el^tJXqBtDg;0ZgIbbC?y`Pe4Gs@zc z{!R-fSs|EE?4+OR7x7r2FgUW8Ce=;{p5h?e8YK}VK~ z$KJz}o5tV;D`32`WnE05Go1gPYa5va(2!t!k#L2uKni?uD9nl7vWdc-;Wc5%vs)He z9%HVMX;ij>fHXQLBRQ3>pGY=%dWm8+-Eza=xPKGgeRUBZ0`y;|)GZY-)J3Zn1EmPtVr zRX4#TUo3+cZ3?~g1a*3%))#y#P-@kfVzy6C7Wa_#GNe1)DcK*6s*C+TpF;P2GK6(A zqYIo(8H@@OPr&iK7Y!&}MAaSdd>cWQ`Hou@h^X(jJ-dU!doE3HC~NI%O} zmB1ZYOdjC5)5~>gN*KRT=TESj63?v-IcO~Q{!c4buFqaGg}OeTT4qCR`#C}1C#YT` zAzysQ8-txF5C$!IB52QpDv!0XMj9IOassZ;SVY%AQFpVbHuzcq3L|58`XUH!fZZ?V z7xT68d#0r+PJ17Kj?&@)R7QrHq(os)z*AQCl2{$NH9Y0jBjL&7;(%uU{nt%FMdH=+N6y!iw`1mOm1znnYid)J!#Zqu%uEpw_rDu*f?D*F1^O~=jOsdqUPDl;hy(ggM1r9vW1{GqsNe7 z$rWJG9@*cEL4xy+D2Yf_b%j3!-QUESmxI?%KvEbZ4XoRV6Oek_{T7B5n4Y@CqXvni zH_A;H{`!3@JbqdH@BLK$7X}`;ZQ#*D(3c6SiAZ?AlWGM!vnxh}#D#z9q%rxOXej5t3y?!R(T-2UItV$ z82R2Kz8`De)=|x(OPqU>4#*S%dx%zP^n=B)Dsp#Bj!GxSQZEL$juK-}(dP84kx}=_ z%&+Zx%Zjl_+^nBcRpWs`WBAVMG(n#rsAELJ$)7!(dx0(ybyCOYmb_E@&Zd9dKQAv- z7n;n=b&u>hb*0cy+CvK^*Sl&^hRT* zuOliq`alG7ieV%rFZ5GsFLVzjPOF!pyw#u=zA|#Vc{|Ac@{=)IrGa-p7DQ!Zibq%3 zKNd4F@^}6GqpX$=&kdS=Yu&Ch_C*j3;n+-00dS>sk-bs7*qb?Z>|Gz6CSL|FEyX z^Vkg&LrvXL^1$RR-VfjGi!GhrDk%Ty6$utkYPpXB3;nUoa!sHy))A zP^!idgTYBmMp^Dloc4+EqiOrXOJdU$8>Uu-t%<(~Ol1v0%j1%vy3Vz@a1sl{y^|Nt zFPjV48NX#Ke(q?ILysgaIt_Lp&3t$)v2J>6KuJt(ux63Wy${`pf4t-0{)o4`?I5df z-*bHu&5@vFfz`>KcYfnRY6ZKEl;75Y~fR z-XS@IR+%;F)8H-M4Sss^4CPuQ@?0D`j&U?5&mGF8Cv$QirZ|(o_F!-@Xx1+xt|@BF z6i1nyR)K+kf>Z9h3V?AbqA@0g1`e_KtpR!Sq}f@i)%U8Z9>gmdzTzo}g0*YBiB?Tz-l3q(luvofe9G{cDTIy5queZa=quZwv#odMG{D zK}}o9zh|E!WXQy2`!$IAD1-v)>mCe*@bl)$9Z+$E(QDk!{NICS#tJKBcy1z1lzgzv zUV_GBeUCgFLzY3%n+UMO2`9b`&7|OSel%_G*DSAE1t2RTNaoxLKpDqAewAqB6l}k7 zy*I&NH&_IF$Ij2he=9545{mA&Q2W)yW6{B|%=P)?O};x>t2Eb`fJBI%L=J^c*-d8~ zjG3nXImaFJ8lE$Pdrq-J#VgguymS9#^)Y_m@%wh_B)f2j=UPMRhK+@}M9`pNbPoCc zkv<=#g_0w$`4v%Ts4nub6s5q+$!?!(@^z8NrHS#`+=AH3sqNFTP_CJO3Y7CqBr?$8 z%9#u;gQb)2Kw^2j$_Vv#{l0p#mkV@biWE*AxtFRDoKe-uv`XxLJ|8})SS%Zg%9E4@ z_WNQzpt;{!Iw;4T5mJ$13^zxG5k(WZKYSO}5mhh4p0u8Tm5%#v4ubLVhR5mOjzYKN z7#(6S+smf6(1YQ{pgx?40TZ%cj^*AdJahO2-k~HTPI#twCzUTw4bj84x{F-kbjh}( z^S~Y-j*jfybzg2dc!aQ;p8?;2)6}3BJNxCayhYR&8}pM)&?^Zl2{k)AK#O(P#A4tM z(`b6-8PQ|s$PAk+M*uPG@-ch4gY04odCcd}vQ3}*=XJvTfE)tAGwyiW-Fv7j;sbPyNv zu7m&|ikZcUyeG+r?jp?bXiz79KTxXlQR_sH{LJjg`C>CSGUB71n;5?{96i{zk?B%E zidMD5_ma8o5NVU){$o(t3PDAn<$O+ME4vAFEqE-TV)ri?`RX;ruYc%FiGT-#hoM9O z8+UI(J`ZHmLOSf7jkq3OB#YSNT3^jp}0W8{f+$I;Iczj=F()mv$N zcX19??!n&5c^g~t2|*trfG#^B)hA2b@LCP^_;;FOCp9>!AC5v$gDT+|)#o`qWLk8c zU>#KuYgA{1l*L_~wr^U!61rn!D#I44jOw8nUFc%PX6nJ*Ty8I1Q$D*i^x`|0B^`3h z`*vUrOy-=UKGfX(g;9;K?u#Btt_D25;_%$}K1?26)iz8kJ9IFuH_tgfO6aijZYnbV za?#g{rC&5t^QpBXwFr8!IVrNiR~A8Q2`Ym~SS~vRZ5joHS;m30IRbUTOGE{PPF&=_ zEd&J|lA;&O3JBD2NSCF{w#MZ;-`(!ejExF64vp;GMa;O_6LRyn_64xd5(#>+pUY4b zd!GZYDA_Aam8K4$N+xL}IH3%7({{qh6KQp;-Ke>Cp_)#pm!Fm zdj7}C+e%Ekr1&il9V?Wuz=!L0D-HmQ6+VdDy^_8EW%WuFISL_ldnB*KgIAo@HVhv_ z1pSDhKF7%X5?+x!g@>%8gYu8P%@hX4z{Rpg<++IqES#l_!c0NiAz_#=#2qn2SF#{x zix=0+kdg;i8a!1+u2FOWfoMGhSWMDR=v~t)Pf)v{ArY^|c7m$N6|WlTTyp&cY%0Lg zEuaxG(zj%|oyK+$lm^I{Xoxug-qE096>rF^6`J2HLtz7gA!LnjE$2YE7x||9UIN|s z4l;#zM!ZD4D<)}7atEx}FI>eOSbv%15Vva>SnR*kk>h=P(!#dEp*Nb5hyh zhr!#ia$;;I%MnD_+%QIO*dgMBpPR-z({Aj+z7B(u<`Uq;uy9dcQ)qIdE%i%Kkg@=` z-CF~W$zg{aKR89xD(EM-_lD3%Ft0)@rn*2e|r|4Z>(7+p-vZu*DBV%;Jaou z*)JH&m8V%55n_Jl-ERb0ol@26<<-RVwAZ!Q;Tp zNU5lY%upu#=)ymDo7g=obd78{TnVZtVTPwN$})>bT7pT}ja81XYF| zx|xcjWJSy!IX2fB)tQ{*s=c8{$qhs~f&Z1owaHV3ZSqF-2I4*V|21D-`d0Znm)|<8 zHmX|%9pAbl$>i*to+>`U$qNQXQ^gg@B~>1$Jot9tSzcanqq_6o4!nK+7iZN+K|^D& ztD+$;_}=Yi3-$O@XadA&qj4w+%CTpoQ&vp^?bA% zkKZOm=YifQl06-kZj+)AYn!qmv`wB#pO>~Pt|$(3`utOPD9lnP@8;D=YuvGIb`6-F z!tIDRvjg+`sy8gnD-n2}Rr*Gl%*^P9em$%Hp=q z$r7&>*2zua$MnXQ&+dw<1*-K5(dp@Jasz*ti|6e|M*J?#B{ZU6Uc9&?SyxPM!qG7y6=-Bk*i8 zC)>B1cP(J2u+V=Yr!^{t>ZAI7s|0xUF)}f}e0Gx*YxQ(tx-d(!Ax)9xdqG+(&jlJw z6b^Xd2HZiI+4*A5zkLzMF$wKy{fn?T`XJV zdnsaFY(Ce(uY(-*U1^;F-`5Ft#~tG4aSV|kPx?3*w&>s4@coN(vwSKh=a79-Em2)$ z&6iI7JaO6~UP(YM?Svh26f(@V=6PpE4=-l3bN*jCe&e97%mMLvNt*(TGH&`ofd=GQ zR&WeK8f)4`)D9D5_>S2@jLnSxT)HH_BSx$Gu!_X1Zprjh4}@2!9{HJzM@}hgpfRY2 zs-CpK(%|u!e;i9BHJu!ETRWoeNxDSo(tZV&jzIAomaw!ba5OxP%^;BUfnt|fOfF*? zf`x4%UtAUW2sYLWE7OiaJl!t4T>0#T+0%^TqyCfC-&G`$7_m<9^kAKFN=(S=? z`3inGj6pq;v?wZG*D7DadXo{V(I2daEc3z9@zgC3_Ka59lwu7M^h1K` zBNBFp=a4IA-Q=|^4)EF(^9Ojng59%n$oY-x3-kU|`}L3h;jDTGX+7NM1Bd&+Zau73 z4)BtB?P|RI@d)cD9FrV|gy%Kc7VpZiTHbviMbav_hr;jJMYD`*Jfl@!3pf&7B5C31 z;_*Rf1^P?NhZ^X0(J2sgzZF&wybo6c&O?Uun(QLC#|IzUw%`HCXKe948CMh50iW9E zGxDceMn}yiqZ*&n9&(i*j&mfYFSbP9DnR!8FUFHW3?Bd858|u<+^CmlQe|J5_&}r0 z8heDGp-HobNZ7}{4gOV|qANzTGrTYc8yG-XI$5j}SBnZo2f6p;T~XL$xrTp&f_*BM zprzX^w3Z%|&-P}Xvu>+C^J8R&Z4cifx z6Lc9tm0+R|B?$}XUQ?97wk|!aXs$tALJou;d~I8ZNuJC}7p{m;7Z!7S!No!ctyQpi z+G6nKGtCw%le1B*1B@)=4a_jAJECmwJPjQY)x2D?+NDIkF2&C1U9o5GT&FGn`b;Z! ze(~w7T~x6LuYzZ6&`?XzRRp!4NZ2rOuitwm(^u1H#Mp1wCciD{i7M~`k2uk=^RaJLhl`r~PEDJq^g?~g|x{cmQEeFPVe`0dYP`(#!O9skE~ zT&C8&WLv`BHn`eG&^iKWX%Yr`#~_Y`+c;C$;?Vry9ppXvA?`xXmWjLgU7(EKqSzOX zCFotG<;voK(~>Mmn7IHLFPgg10p~@dI)9!SIB#7PowJW7d2BT)LwpD)+^D`3u`{lR zOrq+zjX+AaPkv9{>#pqRb0)+=1UzR1_ncxzz(}U!{KhTEzG^iczdf7&UFzU?3UW!q z*8EQix{jbu5eXP((JHYcW0`PgOc$7JWLGdj1h^PPvMx=}QsoOfVb{cQ$zPDf2166owWy(}I|0YYL(mLk>FCycLKO#K z?;O+y0b328WNpPp`~Aaxtf1k<=dit1I?(Vvf9rbG^!oWfP^s*;(w>_LhD7S{Wpz72 zgOj5r5>PxnO>sU3NzRA78UYW@B2bY8ROlw8hIIQtd)0yP+@GJ}p5bakwu5)8!9cZz z^2nKBibFHQ3MMv-q(j4x`l9*MzJAVLknsy*K^Sx;>t+4EnbD9wK~i7~LKX=R1OZ`$ zf!{`bKK&!%Seg%3_G8quJc=8Cc3q)<*|{RNc;A>_pw4=5G`82qp)e72D?wc$5*91;9An&hs!q@)%HDi~k#=C09rQ9qTh85!t%q}G&Lz>lo8O8_vL!&31F1Rlyw`Vr!ghn z5VY*Ib~O&s4PphYf;PpHz*6viA?e63k6Ry^98p7_pw@tV>`mTEew(5k?)il3k(9=L z8t?*?Jb$j>tlMDp!cTtw1?*r{={M(1l@&(E6aKoM+T+16f+*s!u;KxN235aGA_4Qq z*bk#s9**73G4OYizlhk^n8fC_$(_BjXS*Of9BZNuRlx7=fNR4 zhGP5f3B|JyNEY)}K}F4;fHqn$IU~pfiXx-BjoKT!Z2Tg6L+EBz8=W-XS!43z8F_pK zOT;|B&F#OL5tDYWNoC)P$7ZMP!2ubD#yxb5(-a%UJ+hS9x#V&%Os&3$rMvtOa=Xmx zIT}+?A64kz%hmeUL}puKLoE?*Q9Um{N;a8TReD8s<2s;3J4l(NiLC)5q%~wQJBpg^3sKEh3Zk0 z;wYs#5}qu^HYg-4gL>3DZ{&ilj93~CjAcky-%4Uy%d87Z8@ zwT@R3i)ETaUQJS@k9C3H$}lk_A7%lgkq2jqQTV~1KKv9| z@iWQO!b~wnX>Q3@P`%tLsK3}j;w6pJJpuaQ4US9sy>MPxAZKLn#P;Y5KhMz%DjtQc zwdd7ips&Nn4ATmwxtDmTXof&wWNHhMLSe zL6vBKcrjOlao{cftEXF%M{Z6ajXaZ%Vrb;M-FI{+=NIc<_D#-B-ug(hA!$?e1g!BH z@EtT0%3G^fZZ%OI^(v+g576~=4w_}%LT>m@Xnc8++?~u zY{*rVf6tpyPeqjB!aq%T+v=n2n#}2@YS>lAd2j{8WgCyCj-XEwz%ZGBP6c*vw{cKg z0HbZFC%GYdk64oyoe8c>Szu=LkXMnsR@%mC=6A@;XG1|HsfR8zjU_H;aene7X%iMK z0PgL!D0r<^W<(o7JOQ*6u&o@&G7BFh zPSEsGr+ncAI}c}ckg_#d4!(OHvY$EkZQc=q{YJ=RTS6FiXsft|A`FmVXrzF@09PTF zv-8Z?GO7a(&GD#4cJ3y23(UR|R)pGWRLlPGKIHMB8o*~(mKeiR{ zY--3x`8|&EHdbH1=dRdyP96D*y?V9BdZ!F})J6U|q&8&FR2|ulL$&`F-dg zIQF7A9=q5u$lNTKor=tX@K^a8pc`Eoc13YwO1g4QNUN&fCsTZf)U+s;#vObEQ+QR8 z`NDkRRx#QX%__fB;!Zif1TuvHz+a8wJ$q#ZZPhY+-ru^y-ou1(p5!1;lg%xi+y(2BekzJK zYARH_VyXl+k&CE)Zo2HfT+oQtleL^GZU^b&s#w-Awp(R}y>+QcVF%-^pPhWynV7ui zJ_i&-Ox~=J!As(pyh?PI#;99C>{ywYn} z*~`s?d`=BC3un{8tw_C5cFz- zT0tb3i^7_u=EO)gw|w>?h;w$09m(0#Ki{$M^6B+s^@`PU@41iJTdf!0`TL?vM}@Ba zDxDbH|Fp^>b2HAMSG{0c$e=bOu3fA`*1uNu@=qrX*i6t6nOzTxyz*T!ebg-(Dha)}QrIf6(1Bn#%QZQ06a%I7d%g0E zXeUfA%ddgu6tEn9^p~@m0?*w_{#x;!_w9`b!!vr#ua~8zaRiP^6Rip`#8PF zW^ou22>GEW$bQN3IBcKj_~8{vqdJXS7rqVTS0;W)Vef zaqBqpoV$`0ZVAJ{r4;)jlQNsN=_Z8vy6N_2=EqYDPSFsj2zdw)6J! z5;oB_&;84o!afLa&}l`far_T!Uw4O(nSmJwDTk0f?E0~K`Afc*e%YJK7YTuUFF11o zG<##+SvpiJ7Q~usow2SAYea^;dKIhbGzCU5GC3G!VCEO=HiLF&A|5dK)ARou;@%N}<BX#4M9Wpr-Dc@ako#^Di2I< z7GYCyK3~(S+{LfrwvvdU_E}ohj;RZ$-VM{F@r&dueKja#xaGB-6E!+=X$WS4QbTm) z6-hpS0ai?=hIGN!?urE5_Z8|^a|(5JJT++Z-QOYTE`qv2Bs}4 zW0UlOT$3s;pKajpGEaisg5av#$rpSL=2&Lq=J%h7@Q)&!cEYp#>xCPrrhQhb1$UM&0fVEr@w!}(cJlR?1ART0)@wkc0-i5usNl@YOC{A7Mv83)gG z9RB(b*&o@bs6BRb#gKGf4%?_AP`NAwl}n%@p7#0l-6{h=oBT+4b6QDYZ)__xe$|l4 zkqxRoKcl*u-srPLd?xT%+>zj9-kHE5NrrzWKTp^IRM?+T$(&nZt0SJm@5y|-;DrmI z?1F#AvqP|gGLcJtF8=!>={N8zm8Bp_p9eMi{XUvTWt%czjFkL+(vv|$UYLC@_12S+ z3(8X!L_hM=D*L3ZzLsUs*`?~)*|Iw_dQjO72H9 zN!w=)%GbsrgEngM)yukMX9D$Pve;Z-jyhJIGd0MjUGJ9*t9p+ti|Pj+7@asrqWO5* zmhj7=YvVK~dB0)_FJ0OXLefYXbJGvotuwS|siLPjAcE}*GHa`xH>ik8_Yucp`Cx2|;vG^=HJeaJ9p^LHMwHq@!Y0FO# zbdaDT$W2LY;Cxi?JU8((MXCHyU^1t63^@~6g2VY)cW?w|aC~pzjRgwpsr9=dlqCnvL+#u z&XCu}`L|;p`IvhdEb>F{ae&R!!x9vBn=}V!codg0gJPtNcHyu8awF6Vl;0H{T0>>O zWaTj+4>PO`qJ*HKfpHg+uuyo6tdW=d*2a8J>%({W*2o8VBOPm0gBXxHP0>i2q}b@1 z6tR7d<+>Vqz0z`@QJqUJROz`@Fya8u^^kH7Nl$<^3amW;7G@-c{P4hkzU=%O%f>GW zTppJ+UPmsD-a!^|&vTL^41AjYPuPz^<^*1Fvx&{|2=(SY=MXewcHOQzLJp*S+}&y6EwW zjZ=<|Stip#d^}lF#;Vma_L~d#{@m(?Ji1jVr7nB0yZXpx#%~eyO@itq5^4lTfp?{g z%;dGhS~3uIFUp*{#BM`iY{GbZqBag%wclR-N4hA5e%Cle^uGBxf`pg>^Hf$Op90tQywQ_3ER zNnlhXHA_49GN&p$oyOc%I_H!!S$r=l8*0EG0lZcP8u+=~ZPL~B7Jrj;ES#EAVZ_NL z#*y1uVT3q&@bDsM#9%yl(P9vTN&crR5@Z<>Xmo<=t)#f)!?-aB*ffzL7>+B zqD7Gp!FCMV;9HQLa+&g!<4ACLjd2N1SQ(U8HvZS|K6%+0kdH}le$5PE+aXY2J$I=} z$`iyqeql4WE`7raD6jl}-Tl4MR(C}HXU$fsXrvIn2d9ZafMHk&znY*yE4Q3TSU0ar z*00#@cQ>dc7;E`3Jb&WNo(QM{KPj>tz63?TIx;WxuymUg4x_X}wq!R~PhxMhxpuFU zQwBPz`=kZYcmf-$@ITmk=Zo7E%Y?W^MkH#L%XkAbFoop2p>hsYkJ0taZ>JH)j4JV= z9}L;2l09}AVn`b!-)M=*0LxBMd=m}%L3B;>#f?g!V3~?#-**KKetW{Z#*ST%ZF#QK z6YS^!%s?0+qW?;6(Alk%t>)$aPi0G~Z7-R5sj)FHdk7lhDFsA=fnO@=5af{P%UUbq zFnZD|(37Q;?+CC885Aseh5otZD!**PpsY7{Z4P}fdnSu2EH{(KHOoQve5fTO(+@-u~-ZEYnsG;7Sd_&MG zSTb=P)ib9{bW>6TTi&HJ*Fj%oAC<03i|(3HLiWcmo!P@P%S)&8V3x{?3 z0%nkO{pZ%NzU<=d3txLAX^Cp|yZuI*;tqsGhP59~PK#12UM0Tz0zQ{BNw*_$*4AlEWDRV( z?#a8NG`pkqWDdDV*bltq^B<2iA0@-Y93Tv1~WLO=`ZLl2Z4u`ys#x9w%o8Kf&9BchP{Hh(B<`+$h z>(8??isN4C3i-{;uFz_Ty970Hj{>h#+vnV%3q`B^&WY|r?qEe|y5M}we$|2SOilsx zo0)?8D17FxgyxWyWIM$Dvi#bVR|F=(V%Ui-4m}2*bVpPsr%RSg>OyoOIl>$v2u(=W z`Y+<$@v97PQ{w*YGsQ+}4yg;f;EL)y$3W!J<3H>o=Rw5^M8ucgKDyAJme*q~3j+o3 z9*FT7)w;9cL2UBXO_67fZJir4D&XgdFeDlo*jS{W{CS=nce4L^yU8RrXVb7Ym)8>+cMew z$Zm)(QQaYTbIe-p6)~-XWRSqz&1qBAM;d@x2PiF6*8=cJQp6(OJ}FX>^!vCm{Nn7K zJo|0B<_N1f@!aUkfA_t8{d?Bvi#YH%;E3xfl+F*UOS% zhqNZ-p1gN5v;`o=UTahebt~#tR0Whcr0_I4V7)_0*RsIXAT@%WZY6NymTWZ*vEf5h zFW1a!Zr~$(P%ig6wI%!lP%f6kH96!`YQM_d99JT*mqCK-%|a2})FXR*tu|zJH15f; z>O9?FgEXI2f-Xpfp@OKUME;3)<^%{8?^oUAFW9cS6{RO{MP+)gjapD1d_l!rMC$C5 z@{A>8d4TD;N(fm#!QSU(lce%sh%tD`ddLEp6IA9rK@$xXo%!NB-f0TtW&Bv8BaO6+ z6&snpd&_ZF$awxxEh2;ysQTuz^vKUUaqV>ZM7AEtSD_uJpFE=<>F*50N*8L1q?gKa~H&GK`B zpwAK1rztO==| zT&O6Vs43unGAkqc(}0rc52zyUJ|Qr{J_M1Z4!IusweNuj@?NS=z9`HT)ElfPd%0_4 zug_Q$SR!f_+={~OIl6C{xx_K|gq>8TOvwf=Iqcm|c~z^#kA;qgWRQaFA~C7vU?!a< zf>^F9BVBHfm66{wduWHE-yX2{kiQ^yhat_1ZPY!$43o)e=I;gGOl`;^UK(F7Oc$)C zA^BVzyg8znvjWnnsUg~s%|Pd{I$DEbOKE%quQ&jb>W!Q<2;4NjobFd7*J zP9J^m^`AD|2WGw?U||U3V8ek{i7eI{l-M$pXXZv9h|s^bSO$~|%pR!I-}SQtutLSN zx4Z2~&F*b^_7$985L7VKdm>MHmQTfGtuh(h6cnyNrc?*l2$#--Lzr;sCptFRcmB3| z=pU^v%LjQC*QsOUsdgJLz(~+%395lexIp#!U}Kk7Su8J_c9h)r-JZFBy7}eKxqqtp zj#k+$s)*~6q=o=9Hg!C%i`)y!uv%rI|0=($l1F}ZvUCng6RmjX?%bBBg*>ftf4E^z z4Y_Z+37Wt1l$)s)fyLsr;`W)1>YLxX_x-c;I^RC=i`*|`8V752>qP4#(GTrc?&ccR zLvLM?EF`W0ZTv^^_-|D6qC;zup#~CU9grZS zRAdKPLK-5?-pb+F{VHUQNtbDm`59<|_*x|rJhzfY=#m0bAydqDux7oJuSb`_BDYVb z>5*Y*raU->_-y89P=BW}DnW@wnBxo8d3Erws1sDgB?>XzgB;HJ;_T=y5{~wMHrBQ$+f}`(Sw^;_SyhKPtd>@ltm<C#VxnbBp?`)Cf!q9psS7|_4blba`ZK!q;ydDr_r06|m?BD>EMc6K>+9X4!7Vuh4( z=TiY|j>fDbP|L{|q7u{pLR*xh5VDyZO+^0Zcl+4M+UL~#Sai0=iu#8tFIuM~rQ}ilMDQnC(Yn8dG(l}(+V?AE?kH{)g3U2EmzOr$CDI7i7X;kDsQc>~-|;JDM1C@2b$>Gw_HB~zW?{0@1c zvtssA@Br>!(PEVaf*xtKCgs%ys9@S{9?$hw$we4&aAqC9jMcjJit69i8|`)1Ja*mM zW3z5;CuoTNYl(z=|M1?Cx#k~fpbTl4)S^I{sa&#W<}qchbV1JKqF3=q4#DkLssZaUnx3 zESk1CVxOXjdlGnB+aNpnAPy~<6Vk;FMOl~^ix-|5HJDg|=9Pc&U-C=7HOlh7=G-5t zq%TZiUapNL+CpDVD*cDQX~~z9eGl%*Lo8 zQ8({^;wZTsy3+dNw+7@w-`TS9D}enyY5eD)od}(0RN`wZIakA%iY{>yr)7i;%D4LM z4c!xPg1o2d^T`b6p&wV z;aBUxm%-xOcVINvUBc?y_uMqx!9!J-xK>zrg?TNamUysD0a3kSEURe*y^f&P5D6BR zpjCX$+1Z6L=FRO13MF5!1>Bv9qOW6GbIZyNaUC3AJQ6Fwz7mw`OW9XHy;Rbi3^h*8 zAW&{p@8PVJb_q%)b)u~S1H2;M0Iz-4*qSw*F9%KAi>KDbS1==G?ZJ)X;;hzUqh6j# zm3?8H(?%QHa)h8kb)tqyI7(u@L1r|{Lg-&RIyYNYB`BVYeXsjfNkmUvhpIu8&4n#= zK<%8Q_!dzjKZz(H(R)SK9;nmZ4!$d>WEg{FU~gy*Ki#dGX*ne+JhHbIXfm zeLiO45L}naRc%TXoqF%R$0e2TjoEUH;*Tlycr!$DxcA~8mYuzt)1-sUUzVni?1BC^%S zCafXoRRp!1NWfCYT`@*wF{fE%77PF1DYtgWKONf{w|pA&`k{L3WQTo;8bc*m(w7YU zBLCgtOCk=3mq5)kgsYpRSSF2`G2pK#k+%xYMD#0`MlOx)jlCRKIkVpXkVK;+4MAX@ zDQv!S#xZ;0_vZLRH~=DqTQW|a zu(2uzg03K_eMCa5U`@zkP62o0ln#0MY#`&YGD<5#DY_kpyd(7)0f zMQQX?4SYP)&BJ>Wffp=Wa!LtR0nkOC$H5V>5#cC#jJyQ(EO<>n_i$`MbcVmHmMG^Q zgApE$Q3K!h#&4F!FR!%k_MzJ{|lOUj!(4h zM7aAMJA3!5MOAg*w3?N-4)*?tI?m3lc<=_M)5fe^B@l)SpQtT%>yy*rSuCD3({w&WK;?qcQOJ0&fb|q8MqjH-_iUJsxF&)x0x- znpSQPl^T-72Z6LIvPY1?cVx;P`1VJ*#7UrxFrsS?#eXYd%^$5+NG$zH9Cc?rm2AVF zp7;tqKv4ZeLJ_wQ*oF5^AC&Lm+?>`#=W&zccg8gbuHuywcgcFcE#7T%qhF1xmY2nC z6nGUs1)!ZS_O?UcYq^3*ZU)3mQUqt zz1%j%PN9yhrB})q3XSTO)M1`s($2U?iY)F$F5Yca*YXDCebi=-DG2-#X`SHxKfT{0 ztCL-pbjfhQd*lZ3BC4C*KDmy&Vd^oeL*5_%*BdjlxdpKs#97eNn5)X-UJW$`RR;`_ z86lQWPUUV8Z>N$Y_se_X3gq)SMppCc1RmRV#cI9rTrc!H-_N}NLn{Qody4of^(niB z&~yJXs@uj^T_fnL1l5fGb&a4aydtialS@MRWoUlra=*eUcc7HcsQ$$NL0k^GjVzFs z@eNQsdwoVfS4$iCsC9z4Jsg3Zjpr$>o~oX;1?catN|MER_j&m-IkxDYr>Y~g%8IZu z{(Vp!y*?vrk{-PB=E&|?1HW>%2}oQE$W79L&@z7Yq|^L$(H~3J@*Y8^%fMR|xbewI zC#V5g8aGjRRkFpuQ^I$W_UX{Qhpm%ztrBJ=r>&$RS}VD;V#Sf}=id%4hTf`cJPYZbQxI^wHf zsJ-GWP|0>#j=~XCjcj=8G$U40Ff}_vq!og3#}CO`-W*j^qptK}GTT;>T=cWr47#Kgnr{f*zsR z^4TY-h9IpHlNhIhR{O&#BnG$a#gT|CQTQp00NB|8f=Hioi{drSKvV_=z}9PUACK*F z7RlqAyj*XfP^gN`^*%%`RG$Q@wC$?0KoH92eON_agZ)2(R`Ybb&usf#4<#mujd+|9 zcgzM#G~>#)m%V`iJ7sfJ7PdN&o1l>=JYU%DbCY+(uU%cly*twgq2s%;22d5)6_Yzn zPo{?SNVY++xhG(693H(YDIk-@YeLpRziln&jH*u7?{`PQJkN|@^ovfo1|To`6!zzt z0Ww0$g}AzB&JX>pmg#%7ThCML*{y1xo3o*cZERLHK|@i+Rw4l#Oth+t7@LN`HP=&h*rp-qvGB-lxbmFK{47Fq=r7oD;#()*MfU}Q7FqsElYD6W5V#qd9e z?29UZB4b#2?FNEgPf#hAARLC^+9u~r2D6Qia4UHv7^>L*+AF~5+YxVfe~ZKaepcV; z@EhIyn$=V4xvt~GluHl(zMNtUixl?3La(e0sDjD@DW@}#hO1V(g5o;~P_r(YZk-QDEkaSuR6&6|Zv1#6EWtMmg)x9n%#j*%`hW zU-=Ga&{`7>|40Hic`R=h4i9}!5j}w*mmiZ+^fTWg}4OKI^0|qKF znT4>>vh*AX&yx&9xv{VAMOb-uHLyA23RsLd20N^b+w$c*66@+B|J&P*)SB_sE}I8b zN6JjiY$JeCG>T&c+Ph7}=A?_AHNN^k^m1$>}V;81#S1Jim(p(nZO*?^0*z*`$Kb}KC4QEyjBp=Ee_b_U!rQ19f&wV zKI9;;?kYdL8AsgBy~IHt>=d5mp9EuXr;rSyNvZ<|Ec_=xXPqIaSJ9&=^-hMwIBFIQ zdDRP=q?uxk<)50S$mN0n5B{T`{!?~D=a40_$f}*o_1F@S@xNV^6lBGpYW4DJYO4pE zvHdn?Y!^Z26Ci?lBx(laU@ zxr$EbToH6jE`!M`i|Z4u4}^k0NtREW{1ISrXy(DdO1CvbIc=|;HB1cmv$I#}U;o7$ zKe6H^EUN_yxIGvz_igafNzf*OYDHd{ERR``HK|ol0l^(O-U;Gs*}!Lo zoZx%JD}8gw<#DH|D&P;_&4CM$oFA@g72q}Zvv^|#wQ(86Q;hZvHqW+PdGG?r&<%t{ zi^$;#A-^=m4ia~5t-w)}7@q?IOSocTz93JM%rmbi7$R&_WJI5xoyM)741Gg4RAsWX z32xj!IEkKTr`<(In9)<-lpXmst5G`g%S)e8huAq$9-M}1wJ~9h1YJ*1r!6L|G#DHE zamlNbC38w4rJWJdL*i?rYH%jB)9|hZtl};StBTww)c~TPG9Jh`;igu>)c|y~YQQ}~ zcd83Q$R_B^LFuM;ZZ9}2*tw&r5gdgRd*yc^mW(4Vj=wE2@Y^^Czy|`O$ier>R*uC$ z7;HEmwqp=(wv7$}H4FX1hN28vu@^ptO-%~Uri&}E5d>L=7~E?~XegxXWjY_ZB}0Nkh%3R%yAe;s=(iQaM`#ddO>{ zJZ{&-L9gr6bmE#|eAWBL(wSWojcQ0!@it8CP~lzaVSh;!;>WoJJFY>CttoL9DR@6K zT1Kj;xXatW<7Fo)uMSuoy2`JJTOCkM-H|7A8pv%yU2{GW^uR)Km-<|=4|*eN1bYHd zN=g@WM}8opQN5XH{Ce%bES`5(-9YXQ?Q{(!n5NGes4$J`isS5{@}+HEZJ+>*7 zfsNBr=4oN2*I1%8SjvaY;n}{$oNED1(u;FkVjdivA6C$Ca7~~`TE>L-$?^wJ~7w8 z?}ZMuF4<(@^lPVz#+c2tVTA$3Y^YgHUEWL?!my(+W==XL4);M zPb6HC=sAmoXGpyYs)0>{dI5HsS4Hj*KP_nj(VSbdWN`+6EJ`SzOaeP>Jh{#7FIZt? z+|git$}B53bc%n?r}mCigzmwubf4Pb62`p+y{+p zY0+0C&GE_8jcOhF=#8$hBi`G{6|;=$I{{U|NdJj2!*7HC$MKjA+bFJztdkuj6NR0m zQN2GdaoSb!c12}$bL3%p5m(D?1yi()ex&G+-!rvCeTcd&nco-HAC^z;moJ%E9CIhY zNjt<2TXv@Axo@3=A2YU2a)Z1=tT-wddbfK~o;mNgLHIifyV^sP^gQW=&f2)>^JM zWI)m=OH<&hBL5_4X24XLldu_ia$NzYkwj37a_%Y$8z};Ye^Y4~n5N1$oE2Qi}3~AS)txOXO-Ux-Z2h#7~o7-MqtSo=qN!;fxK5?;NP5vzd<&ZY76cM z8hc1YVH-k`{PgSYLxYo$c+Pyf;1n|?MhLUK@~$EEtM+Xx9=msDXhcD}608P>KrV6u z-+28K?@mc3A81PSKDl(4OslLC-1X6{=0I4H1 zN0-EOK;iC~&B%-Az;-|2-f#Zt=&a_zbF=-rilvc0R&bPm43d>`Vf!dD^~WxbCdIr#@EQ4iGz5qC4|cI z;k~_2;vi?BknVwEp|DSi{2jdt9ndMl&INbDz@S$ah~gLKhPPJI^2X#_GOdB^_8IbO zC3}JJ0jWZu7B0RH^hK+ndS%I!eQs%na{&3aOXbM#T>eZ%Z1Fl}?|*pgwuQklIst*! z##xr1SPzg(E>vL&N)_2gHqP1_etkx%JcT;X88f%YF*2d2GwkwNWgL)|p(Ff1oDyp9 zf3P9!^ju3xF?cIOQOUfUek*;=b(Fc$2e~LMjMhYREub&9-Df%UO(MN|(io~-U-1P3Bzh?DLzq#|5|3Pi>VDA(-#D{T16c983AeTtMUU%$FYKmFNX^O;J$H%X= z%2J^9F{obm$H&ll?_|C1XkTXzuI7Mk3z)8z~4W2 z4HrwKEj7*Nh$-^;BGp(YnMLMsQ>K8+y2`IvQ77+!{>OH1UZ@Gms7vIzz!DA>2Y4g$ z)MRsqyza_PL1~Hs#hD;jqM)lp((JOrHfOM7XK^fVT?8Xm*co^I^WWyZYh9Ht{@EYbwjq_lfV3TnyO=CQqUw5emeTEwrGGW$*j+QnPeV z%}jSsQPX43^fLX=*xfZ%jSGsFASk$ihD8=x1OYcx)>5IgRIpHSL1hVOSy%{&6e|4B zNrI9{G!GIc+Ef0Dym#MyH}LNF&OP@m-{F0 zlFL?6ZHi>33ep00lWk(CH3=LNPw(mFod<OhSWl3g?#Imp6;j=jvdf1Hq@%O%d<4 zL5<1#@=S@L;kmGS-qhkXd&2HT-~GJnGiz0>xfSmf*<)ESMrNsXAlI-Usiskd)SNBq z3C-;Dp6>ImS%Zk@-EXgJJHA#q&73up3nCoYidSW_-jz@+kXRK!4VHjdH!OBt=?B|v zpzWPi&F>F63|no?R!#Lu_0bg2Y5W0#tKnUr2H@M&Yfb#BhZEHX?_S*AIt4;Cqjmw_QYeu!lNvu9ZvK(0LfzrgI zHsc2Y$a=Khy(*PY=j4H zdQ4kM@DlUrCN9FpflYh>s!=G*q*y4auEjbRox%Ghm=$eE710!XoP1+1JaB+r@x${yz7Ix`8J~kw=~z|2bHA}ZPWfvNc$CY?kX-YabR`V0tMHg zE0hM}L+q?PN(+@78)Oic%CCY*)FJs{XXKD;Q8%ha+;@3m42Q>xSXuBIVjZ@4 zK8kgX)~g;CbVw|Rk}@r74Xy%E#nI!kqLtq0@}xOaJrEe@S1;?Lk#}S3qEErqXOo#3 z$Af=n$v1DX#k1!o?;V!=>@!J7-OC<7xn0>NgGzr)Q$)vt9umzC^6F)nd{{yV>P^){SkPeM4xIN@sxgg$o+PVl8de(}yfO*5*w5xU=N$VvxR zb5l(ekOYd2qsRwTBx)C_H1OHE8b~$|DjR|6!X`56xlj7pi&kOc+1u>i!U+>oAFGe8 zd&`K43k$AHA(y8CC){Y=?>5DDP^68Dyeq>(*=6)fA!vPPC0N_Run28((?MD6<9wVE zjI-2w?NG*s>BB1Vq-6e+lTFK#XIv!uQ;aETSFDi@xOUSApqlnh75a*S+`_=R9mk;I z{u9pzUXLt}+zLLT#{G!aD>g`z7PNl)0I6n>fxAIpVKhlmr%K}+h;CfV-6Es>S)X$A-$?8$lYN+P zB1U&nY%)c*QIV^r@AtSUs%ByoHze!$DPF~jG|6qUi>V>UdBsqO^)_jPil_G!CCnwS zk0;HzYZaPCi#ogE$8=!7U#(qcUaG`lz1bFOs6_LN!;e5-}>UQxCxG|!j%vvS6k z?=*tHPpeuttymEs-Yze2EqB@J8#Arb{}Xzzd%L_#d5WK7%yW ziR2&l+z9+oTjyQ^dflX;6t7knteI5c+9m3i$FgUAW87^EMC)-jTCR+F{9^EM=n>Q( z=^JcBj>viGZnDpTw?Su3uuw&@r24c965qTbK^zsyAA|v`;A{21>Eb;> znP8pW=&~nhi%(ZzQ^aS!7(5X@?G$jVrUbPFREXA21H-Ouw%IlW#S^4(n0t#|ISC)? zTNfX0eAxrP(ZYuX(|cq&bh1=a=`pAna@r74Bi7V1CG*pLfD9Mt;6zv})V6ZL;v8@f zGq60AO|NhQ#}rR~*swWi&nsf?w?Oez7`od#pZS1^@~@-MyKhon1XuE)=RjyH#HYY^ zX<>};2LFMLSC{8cFefqd{Gn`mbwk{#;V&E4p0{qFPa#L9ktP!ne~MxuDDpV?^#!G# zO%QjMCpvlk@6V&zpN6irg#Anl44 zkEH1_)Xo`eW{pn*aGHfD;+)gLxMqjj`|Pk*|3%wV{?R;a;S~`y7J?iKf=ZY*qBKck z#Bt>|Jxf%BY9Mq2*)O>ev2(^HQBy>WIv1iTw<`yXeYPvF_&2Do_@_zqeN-9rE@s2b zVd!~SwOx5;*6Ia^RJUZ+{tspy5RVV!mJw((_*?#pab4NGX!q&AF$$sIb^fk{eC)sq z@TVry@fyWmrpQGqvhr_lNLKro{on?$gWdAbDNE^Ax>YbRtDUYQrSv&*h3EoXA!-9p z(e(v7@g4spg`t;LRSjtdm41hx+*j^$eLy&DJwQPL09pGa*jORSnIYw%;$KsN5g942jxjRMDe z^#}RnFgII`FuS1I zH^ZyOua>?J^++I10ClbyzXQQlw6O_&UOM`5V3hC^4=j0vWdRHwnjy1p(6xop47wQx zOLMMp%Ot6E`5c2!4eBc;DM}YSAVs`G!ts`+um(z2N*N2?#)q-%!=L1v+k?62JPsUk zXTf@m)k&bbB(hIn_-+zt@#T;b1>u}W0ZCV-AGl4btl zEhLVcjB#LK6qv{uEybo#WIGj!{0pgKi~?&>R8G$UMCj=A5F9b=q>uP&kZpuC%0?CD zH>L9qs3$8*f(t;5qG5b9&hf5)5SFWE83CeO5_p58IIz}RW-{gZ6btMPSyUtn`SLk= zVysqlN2voHwOP`{)7(*>7Yz_(X|J3Q!BKjn!Npnopa2b4m?>tivRFZ+S=8k=?34|0 z5t-^v+gNkQnu2l6zsU+Q)|j?v+?v5S=Yj()#`z87yNw$xB32A6GtXSK zkn#t6MBkmhL-JkjYp0 zcv@AR?{R29;g_VqDt6emU%tnuSbh$6sRu|EFe@CIkrR&Xhj;~{pale!^Yew*8Qg=O zeI56mUnnQ26>yj@lh^mfC*zx^k-Ip}m*Ymx|9Rd|zH83^@oXt82j2Kt-~{PgxR)&q z?hk#Uvtqzz!O{x48;ZvMxk4WyXfQKxncgKjAYUUpO}e3sL@cYpis|X<;gFMRh;El+ zqrFrX8(@0`jyC&VR0mdMAh=;v)>#q7!qz7j)!vU) zS#-2tqw0R(eUJ+|^v)Tv3p({g4_eh`pV*)VStDq@rJ?cj0ZF`A7UCE5)nbu-N~=ni z4oIW@kj@VKASJ^D7e@=TW&?Cs7Ao3)w;dz!-@>) zWzi+3!Z+DAMHp=WDpQdIXA^^y;cm;ID%z>`GcOOp#YX42S{t!YLAV`QVKV77x=vNe z&*nu5%V+MPTNdA!q_K~^n<8#X+9B5WktChhDTrh0OU(W#V5C|hWhPnOC)qnf= zyY>-CBlA1~42=w7m(SzBD?Ib2QPliv@jHX0VJuPOz*_LgGruO6|r6#yjDgRNaF(Xp^jQC6#I>1D(B~dzKm6c4PT0B z3JtOXrc0!W62>tV0f{`Vx(=zbkkZva0+HzTz^5$ufln5#so>T4;&1(;Y^n($=@}`+ zx#2k&v@DIjzfKo!o?+my5#$zf26l$m%g+M!ZW{l%u$I=zGnX8e+)#X~E)d;Q4upQe zo~KuV1L#BHJ!QRMgUvY2p7Xf%mpEhLMFZOCx@9bkq3%%EgLhp0>vC6j2A_1+>~rs% zgXN-If#SJxlG5+Z=GPkhnB6idr;2yZY!ez+rQerM|0CJ>lC4U4Cacm;irs-KRpd>N z)jkHn-@&^84(GnmJ_!8QSHi#sG(+MZ>FlgW65H*VUo>OZfWTy-c+r41x^7vB{KLsN zzwxpc`duZ=m4Hm`D-AH01BcV zQ}>WmajLjdv}^83AS4SM4m{2W!yfXDpp)Yvjf(DR@+k#2~3j-FFYR3G&gaKosH(6YA6N<-6vHhsP0yZK+uq2t7wMj3FwhF@6{{aRUg4h_LHxZ~O} zKgwMnRg+5B%rSgXe$b#>C+wP1=`HRM@X3st`;lk;!p}(p$qRoE^A6ku*^zIrX3Z0v zxR`JpIGV;n9gL&uy}?+o00XKt2#0*(2_`KF4an%v!VphjHN#E4eN8o^=8Y?qM%`-H z!%u@1Q_XoCo^?h!Fa#`c%is!z|F?Udgf#`TV>GQUt@f?VIQB$YH}hlPYtL58LS+3Z z_x~}+1Q&tfz?zQ*v^1*PRr<^+Ni!s5Yoh$y0{1&7hfkVxcskFHikAP_-_&j%L-kV0 z2@7#mnd(*nY9duMFzy2JF02UvfMSpRo;q6RGsyt3Zt_w8d7FM+MzmZGULoe_jgSsu-}0f5DS6*iQ=+wtW1JLXPR=CR>79(_x04e7~p`7 zko;rv;VWZPfr2HYq~0eKTThWQP*-aP zFx~0tUDtzO`JfV0ze||w%m;2K)NxFXOOFg)etIsvsE}z0@cswQY8;e@s53)_cZb|Y^8>$*AZijD=j z5`Bdmtj3w7Fzitc?A$BX%Z>-Fn7pNu<79K1u!(f$v_ww*rX;@F92#7Il6KT~)y^oW7iuwSob~NbTE#UgzPh`5R*0l@*V`5*$ z>F$}AuMV1f2Dw1Pft3^sK0~Acst7<5pfp%9a%io3zcbonG1CAUU^N(Vk|##r;iE}o zc;P5#CT;8UW^>Q^iNM2wC5Hug7}gnFbxu0#q=u+mjN8W&GgqbP4XX0Xo!95u1}?@) zTZg#9gOdoDKAq3c|A&~IVHcwfR~WI zD9}7|$7%z(0F{zQ60oNch(Q7vB)gJ-!%LGV>~=1dW7O<9h)s$E$1oNzgJylPDkRgu__b;# zPiTj=Wnz-yshoJhi_ZBnr>)GCgv!4=XP!33MUFVIzG@+%EScHOD+|w}?*?k(#CMi7 z1{I1hg|Az9Ao3A-eU_Z@^gTRv}Rj~*1 zM**2iu}8BgCIsm5dPpK~6Og&Ak+s?^eco0$dxAL5h#LEbw*P4=Ft14TiXg{AVI|Z- zP;JG^(1~7YU=`D}0KMXJ{f}agDkjIv9c9^JEo;o~s!Rw($tL;w? z<)@KSZdO|d)?gY;j3^&cEEMJ}ry|q%n7NA4OIhkH4=j7q!D^}mwRD%LH&iF+l&6aC z$~plg!%k?A)yJ7Y@b&zQ5QyENH`OAn62gwE;~|0x>3%8#Fu`?80D75MMtnSH*l7>c znuul7g3!>1u1BQwutl1*olrF4jrZ#xZizDwLweSO?7*v(g7)n9naN^z;tqka7Yf- z6?IFo=|R}~@Urmz-nI1R@2zY7j}UcSW#Q{1W@_H40fgPx;MnN**|b`y!f>}VogWngx{$#{GZ)4UvwN6CKeVL zLtz;pe$vx^EAA+;U}&DOQKh*?a^|N6<g+&?`l8(7jlxshry!plSD5<SUJw_;`<60PgdhNGmPR2E+;ER8MmqfiXwiB(%>Jv|~ zu@u=rMPj}UGR7@O0fJ!$m5c0P)3*X&Rw3JpacteC1J|9p;C(oN42$)TPHHa{~Wl~Pv zx|r4Gaf~mOR%#(yv6e=b3Z0%E>47|kLO*d_0J@Fim}7ogwLWxyGGhj;0l;X9Fe$SS zUt(Du%9TI*Z*%XS1-BR!)d_3~tehL`2F&6S$Z^*#z03|s_p-@sEQ{Ar5~H8P(2l%! z$k!Mb5bgdru$De3&zYY?-(#90sstq-t%7DzJA)&w;p14RLQ^ecvKY6I&>1g{YODFfh0E$XY}lib1!EfJg3ogV!l@)N9=K`Bu>@fPqqYjpPYi zc|BxH0FIvstZon(bH*RYOTW=8(0n0BqU6h>BHqWM1D=KOt99^JkzT=uMG!0(+UAj^ z#@}bf2%CjIDrS(*TvDwr4Xy?jxfVr@8@^Ym!1#aEcQ09iAu~DiYkUVtEE~rZ&ZwYo zhQ4F-H)YJ%hkK*tH^qB-<*k>U&@GK=6?A_C&G@y;s?@t7l{DGs-A_Z3eKOuyv-lDd z7tkv?>3LoHg|ghW3h1yWNw49Kg^5W+YwxyJ|;NME97M^DGhCOZ+Gu!CI=$i5i&Vx$J`L2 zJM!kcJmd1^^9j=CsFKXifp1H4=T61)zghCNas;OSQBvP0Gkem zwcVBBA~QMQdGfkjkD@2HOv*vR->e+|xe*s3D|HJ<^Gl}IJT#g8E{cWfnVnRm)@z3{ zNpW7Zn}<~uGhnnfdfaSD*4Hz^Oh4$1SpvcXpvB`o#4;dMx@K`jQU zP~rkMxiz?T(aZZp6+jt@+is*tf-sfMA`DKt1zA9eybUwg%&OoW!F*HvVF;8%H$e+o zO@Q;559I7XsqM{~Rot}^E_awi)v|07PxpRgUXtqB9lGOg1Bit(Rfm`D`J2{dJ>N=F zTn1X&3VOZjfbaJCHINHbCto?^9+X}!;++up%(Q+YK4bBG0^m9nLU^76U)uoVO1^ngKLz3jbG zU*J>sDD{0WKg@pE7;`eM2h5na+pTf^n=^mwZd|l}QoZXE+2X+Lf5j#X)NYD}aD-h@ zF4a97;$|V5X()Q#3`ScpV_vb=zx0PPZ5Cmsg0ky;kmW zCnV1sk9x>WNvZF4DOUZkz1$fm#7yu8CpL%KzucS}Z%*oFfeo!oR7IZ%#l^l;9`CHd znv*A;|5tpt);&y{KdToG$nAY~Se~Ze_Rfp4n)b(&wVyvYyJd&J!yNcMMP?x}glp@vWwY;g9^>&7H8thwDI%b|6_(XXxX#wr@< zq^cKeof|`y_DIz zoK{E5Z-Hl-x#7fuEuukIwDcA!4aOwRg3xPzgKm3%p*aRkd1B18J19?_o3!xCk`^#j zoKkDOK67pb!;|JC(j!Tho$xE5`{cSgm<$S01;qB6pW%UVos?&U_4uFrbMNFdDUQ3a zXDqw*mERdxo|yG<;iQY(%Hz0y8Tr7(kNucp2PyKvK&=6C z9ajC_Wg-JMQS3&FtfwLouxP{8Gs_}{!I9ueeB+L?bdeQK!4RczCjSNpU<2S~FzPMYcIiDj&3`w!EXq;eA}2}XBK({XV66e(Gn%}W z4HJRscdP#^*E}wqiyzT}qhBmUjAQX0@G{H0A!=Mxw!E70-s35Aj8aA&yS|d_a$sK~RCOQa_t{IaV28<}BC*sQ z7MH`ia~WPbDaOY_#FM86ck>xu=aj89Zi_Sx?z-R%uPr{BuE1(3)_cR+bUBb+)J<1= zCMotY$x@7-?Y2+wB{u|&Efs7z^CM$%Tb1Vg^r7-CBVH~nxH5%Yeq~lA2nZjAuG0T!NpcMb{AT))62+^C^2bzBAB>wREq{t)y@o`PgKVDkv6Y!eR8-pjjc=2P5C(#D zx|{5KWrWpPllP#CVu8@%2sof1bf#w^uqe;Y0cxO6UHAFs3$HU}!9C<^K#6}-#9sGu z*OQ({ofEx`nYOC45}^Ed?^kq0Y!lbZR(cxlZ1+royDx+v=Vy4`<2OYdUzpB|=i?cU zqg4&K-VQ2vInHlqY?P{Chd03{(GfAbGZ(rUA@s{;@-|7}1|i3NIS2&Jj9Tx~DHc-o zQ>aL^BG=4o1}0`hJQfz^)qz;l^9s=eu1MYseBQm#XvmU!Ab%jAl=58T!-(}#Z@fKE zcZT(uQ+qh9?ZyJ%h$xlXqgL)xSZr^S)l9R7VR~~EOn!#PkMV{X(-ASylmjdtru-EmbU8gC=KoLsPw=n z(P5_|UK*1L&Y;qfy1XB_<%zchWP|SsoVR&mgZrq1ZJvK>;eJ3X(yt~d4q=%_%!8>H zm4?=fa%k+Yx#(`Vt(Cq6Dq%L%o{pV?--2GMW>Kr)pt6#GP6>lzer6e6ApIPo(?21Y zu~{W;o^SYthMiUk`$7%JxsviZ4(p5hQcO$AM51sUcm=Z%g)_J~k!PrBgjLgYknK`F zr!*j6rRi0~vPp^}`PtW@j7^1a28*TQlN4EMD93jh+>y8&Knkma%GBwHc%?QE)Osp@ zxXQL?Hd0TAvgy(JZ$Db{*F{7WM{Yq(pAMou@`Zi!i(voB6HeL^H+t@@fyC(klXbl2 zwSBji85giey!5w7#Y>hVcg;l4Jxj5`dr?P4Vo4f&;q@LXq@1% zODlHDnO*qCCf-pstU`%QBCLps;#^gxI!~BNm(PJQv3wHN+dAXFVG;Tz#^EBj1~$(~ zk8w0`g_~t9>FbNWcHv7#KwbUT?QF8&flbcmO~7-UVyh`qK}GH$CA=MTuQAc{tC`Wq zjS+eOQ2s{avb{gq^~UM1H$}93HR|skzpxRM z`?~O?r{U**h%&z7S4@8%Tt2rNA_&p!IMM_4C~Uvx*BE>?;P^z7ckKrmt{^e(Zy7RU z_xtB)MDIAB?Kf( zAvs~YZ-wvV0c6x1TZ5HR-6!qvk30A~CSu2PU>CRrJ6?<83wf(xhhMz_i`Zx|Y-|Tx zOGm%9!%veX+{iOjS{ru4l8x^TIlcdhSgWcO6^h28gK(ThYcMjVhQjfCxOQjie^&El zH@w2%OC%2J)WHsn3a{m8GFu!TE_T(aNn5j}r}52zut-^MA~rfbbbNJjTR~ zc_UX)Jne6Jxfx46nfm{xFEmfTwUD=(L2nVal7s&JiWX}>eR;jC4(G;g(Cs7{2rc%wFYlXI&D7BiP<1Kyp(3wJixsETnijgDc&CBS(!Q%T&7LE*x1wM+MRAhpdDqHJ@O)0#4+;%D*y zsW3{AZ3%vt$yNu}H1?awglvk16uLAB1Xmmj!BBv9RV9O7y?R;4qO`duT{Yz{SHZEP zRb>Z3Sf`?z(V%%1O0mLrc9jsL0j7Q@3V&jGZBbA6;WfuWFoEa zywP#TJg&fDL*y)!EXotY@*>2=M}r|vhVXE^3JeQR+^R{oaIZa^ImwZ!-~Pi_YsJO| zCta|$g~V}FSsmEcQDCw>X(={^BHO9ReBZ)gED#Sx@gDi@bLmuT_RxcFH-oC^Gpbxk zUq}uV=Y;YA0@QIX&MKX(gn6sr^i(+LkL(Gfk+B!g_g!vYy!WNz$}KU0K_&l&<((-5Ut|(FU)1K=~cj|-AbZn z8?HjKBDp0S>0@VT7~jirAY|l?w(F;5g!s;zF1E7v(M3^7c+uh^C;Z&7!0@@pTT>c} zK|5U{O$CG?Cn<8(X>&8Y9*{0@mKxF|FoJN{31bP-G8IpH6(x(-PuH}0MEl*Bbh{t( z)t?>&3jV+J!bm!Z_0nLZ8O9G!eA>n_C1&z`%+m=^m*puL!(-;SM#)~;x|`E9~o3{G9FRej>Q*)5yyXA-@dBB~i6xD*^vTw{*;9O4x~(O)d^HF1zyi6*XC zwsP1Bce9M;p^aa0d)duuL-K^MRl=R~4EYSby!DVLYe@LQ=^7U2)f^0m4ME z#7ENfXDf8r7-wB(zfg?93HOAa)W5L~#Ed5BB3sE^qiXDE37e5Xx zX3nSzAVgh9>yT`%S#%^Ui^k$~SRFjYD~`GAQz1$u9j^8l++63vsB2hwg0^esG#8Gm zef%9+@d6!^|EumEsA>p zdOHDqoai2gCqerrg?xOug?Q$bvQfL3iZ%NWL0=(vh1c{P7fV5cYk z!W@W=foj1og*9ZvN9xS2nH zua8mG)GpgxMl!kC@E!N%06xOe5_3PrLLs9<;8~WYi`(Ucq-xfE`RHTx@nK3A@+BPN zJzj8$MEM_AKIR{AsbeyO)5ZHG0{ogT*5ND=Uw%_A6PGc0Hs;ulzZHOR2k#0p-MDVNPF6?a_;@WL7o<9cecG&=erxC zvTp__%KN*s1@G_Vx@!J%8Njo_pt_Hr02|{dQFd9PwJ!#^EnrAwHcxlP+ zXBstK@Bb`+n{1y(4w@*Gdngu4YGhE6r{SR`DYm_>L0j%M0-MjJD-xG!dIWntm*WpG zuYWG;4^0!7&&d$6Dk~cY0v9AwVE~x zkp3TgohGZeSy&u+9fCs8qioWfDRvV@HUckG7~ol)N)ImXAseS(R%kNS!@g8?@^ir1*T1kS z;vA`m`eik4y-UhnhGdr&`+?JcXL!4OG7w{}{ClEM;K9VE->SP``%71z`R>@lwsNcY zDrvna-uWR(RMh#_NIrEv;5{Ti>0F_@51bD<^i`;ka!J}0v0Zv6(9k29f0}%t=m1Yq zF0-9?ANHK(0q4Y@DQXmt6!v=P<9g7epSo3#je93f3ytHxc|YB9!TGPN%JWs}plju# zok7c&qk*kceSy6xN>VgS2B&LPNb%pni&~7fTFfZxfU1bt7Z#%=1?pO*(nC-v0Go!L z4k~klOPJ-$Kl4TEP{T1+fKF2E42pwx+i2nw#{cMd!FbFa?a$uv)vm3Pent(Yq^lST zUOTWc9|}#4(m4)MEMyWEQ;}ELl{4Cw-jF;3623jbHT+%fX)YZRS7rM=u(fb$o1%N^ zeW)^(6ohpvi{$CN6H8YVFWbZsE!u`#t;SVTe@H z^cX7swcR0IQ1hoAa=7#kH&SXBsw*2!&A1-_^KRlaHIBP$c(d3;GZA-?l`h03b> z5}=o*gFurL8w8063-g4UCT5+_AQ0ASRTpHJ6t&V(Q(%goC3*kA^I3}I9b^e5H zs$3_`dSGxJd1Cz!*L``u$Ow$}DL4O(#5%Bo0!6qH>Qn@o{yR3sK%+3yUj(~w;R z0Y{aK;`pVaRr&;?<;Y`z`m4c3e9RAbzmp>nte^3xy642^(z?63t#uFgCw|A>h>xFC z@47^`I50jS3TYHRc2g`^(RNXhn2^5N<)l<|iSCl&y3iJ^B?t_i6tp6&c+e_9O7aWK z9)wjv6pE&Z*DPv~wJ7jLTWO^hCTBb-G2q3*fHrF8gp#SJr}37U*HCp>1_KLKQFED2 zdA#$3u&tpvo_WF=zmw!Md5b8Ge^H9$^W`pyURy)$3Lv>E!mp zgs@f`&mbo#N%7&*P7pg9B+}6Wn?Vje=xX~#hznkxE;-L0*&02Z@G@m{anv_ocAjCa zSIPWT8qM%Ml0>hg-k(c$^J?kxfKGWFGY~lF))aA#y5nCjZ3Wcwx)n-8{cR zma7`_TP6HZ3BeE;nveFMX30Y+k$`NYdBREC1aSw`_yNM@&XD^Dvt}EW+wbT9>%Wto z4y@cBF;Q*{C^nZO*;M5IZy9W`O%a>FoIke&N{-fp8?f7>IiOU2oOj*(IB%E#b?-u6 zvd=1Vk3K9(_Q|Bd49%+%6!B{W9g7+x@aM4aCFVlJ1mwl@A3Jvdn97akk3RmZ+@wi! z_m-vRq)ZO0%2=RTY84(qUTvf6WG2Ti|fuEk^VCl5RQt620;kNLV|-6w?&}9g7^cG9@SEO$lo%dWMC? zVMUV#)#c)>{qh0HBW0AZLe$}UlP&Tuclm-8gl36#^cnwBWsYw=f2HtqWjR|pkqR2z zL=Y4hs(prN59{@^(-Ib5LLL3pq6)kt0cwq^R*t6=Q9l zMis1C6OEPTG9%ou!DZx$_2q<(DgU{rv+4)NMeFcSz273I9eB~YZK4EUrC6x+bAgJC z=122!yS*La42I-~o%4j<&RF#LA$?Gu!NLlXq|ky#yo0w=XrS(Chh+^`j5u+WaEC7> zF?+34=?%iccwyM-mMoV+V-09aex0x5W-3YX$o-foLA=Hv3^=f@f&^B6wsW8B2|r9d z!<9}4@y3R`_qbhSVp#|~hK!L_!YHAk)lksVdVq36m#Mxyt}`n^#|>RGj%<%J=eB*8 z9?XIF*%tV28&zF25*tUqw#K^_j#BAc?vtY`n@y87FfqGjQcf+y>@?e$QhIOx{_jn7p07fjOW+9aYOQ6o8l-$7}e7IJenddy7$e&x@}t3Wl;m2$7=~} zelreAg0>1Ur?@v1qLkoGhs;0Tg{7LaOY4_b(7XI^FT5{rjHvzF{5Svm`d2PQjP$Oe zPsmF>@^~l<)-4<9XoG3AW>UDnqZ6Mu%abxj3aI0TV{Z3N(l-Y-4_IeZlkN)nTau%`rC*1WOf{59 zhK$n-;mBo-^+}--z0zRLofFHRkbW-G;L?Io{8{QU*tvkKF^$Po=ge>A^$>{0Qq&T( z(&0Vpg4PAWx3zn$@+oGnviK!A^pz!JkYd^gZQSu`ZrA06Hpg88{^;`;H}f<#hgF^r zn!E;2^5=%+Jxj_&_3|3YHt}QcGrYSVwRD;adHG|*lF4OIo7Twc1DYc6_#+Z0#$qwp zYX=h-@E8_2>}vzLXjy_y1jCj`o%j__7^X5FESvM!m5>BE(E<4?4XV1Sbb<7`ieNcQ*7eFHVLB3>1p>b2V~`WqjA3(vEJyWO&X*tj%~>5TZ(+kg24ih5oo z-9KxAhLO%Ue{fTOmL+Wc-Zg#4;lNT+;@l74SW&g~HzU1vyVcSkys_fi8_>|8(hPXG_JxWEW?BrbtuOND7#bNS|w~=&>wk_9kU2 z-54>H?OBUKurdU7yW33wYLjX4!lHZr@WS%7OQrYS( zqyp$*Td$LA?zor27JJwUche{+nnhi1!%i#tM|^A_i}$=KtJfhq}OOSg2Ztrr=_aJYfzv7`p{v^iB3j zr&kIa+~fHh$RTMDUmqQlENu_ThM@5E;hL397hSS+6R*L!J)}I?zLP)BiASkm>$x@= z&+>XRhBIOc8?`LQoB_&VVPb)r3PTw3u*TtCStnf~h-Dwh3V}f(c19O%UkI2ycf*Ob zs2Jl`yEbt{#lts)b<>R7-U!|AHDsj&2QNTv_EG6+2^1Sgkq@Xyq_o>0Z-=T4jqVU& zAZvhxk|=xbRxHk+RhY1NxBV_~!$f6M$`SJ{c!!k~7V_GoUppn$oOISnD|nf7v&%jg z2vuok>gBDlm1$ShNlx%5*Ti8JGVG;CJ_lHwkTLaZ5wrixe1q$-$gr@Py%{_t$Fho7 zco6=UGrvlDamGL(#w?u*N|Yr{pIlgQ;=g{*G}`_WCtyrzEdSojmz{IJRge%m1j%E2 zA%Gh5F=@S`dp^PDn=6(SZwt<(5-zI#KmdT}nqF%AP4(4mJSe&4!9H0bdm!sllpouoob|>YzJ0`c11i> zPv=VF?2r+}30_aslxGfX4!r!Iy;nZhX!BU}`&k=Fssr0R%1vw@MHCBpsku~SR7e^S zhXP$OBu%UhdDl?%9im;!VD%fBZWJ%bs(}d>nB~{a?(#tY-Md;9)B;PHn-5CA+5wpDOT zQXX*2wOekeVA<`F5cXJk0&Mr#mg3*x$^9jN;e&fl2M%-noKb9liPQQ-W&b?ozskp1U9;3 z{-)L|S#V|!CT5mN?kEy~45!Ao0?Lk^vr*A%YbMD|qkqwSBFDzg0RaX;0nwtfvW=)34n|7J-Gi3?~EwaM1`AX{ZY z5WXDud3E4U%wL)HZ8Roubl1Q1Dm`CQ)*`aH_zJ$px+i9-$ZshutBcc1QMkb zyN@D!smNo3A*WK`gTAc-G_|G8J?UCS?}t+DO%Vs=$9yX18t6zj%xnz6UJ%*p+A6?a zO%cd~Q4n-Wj_pWYdY-=Id)~cOa9z@~@DtC?ZnjryZG;2cO#Y}*ILSjQh}C^+?h>*< zO>JNsz(r}lXRjh@x~2ihE>gwdf@zQ;P0au~rPG_<;tkO7%QV%K7<{zL2w|mPzBc@Y2M+frGA-3>fRcg4*@u0R_LwGFT%1 zXY02oLOJ2U?qdrzaFMMEV-Ow(7BhOEF|v&g`Q1rfSAa9SqF)Ue8c$e{q>ZUZ~e}=Nr_n>7f!mSkq=BLfFD!rAVnTfkqxraP*f?k zs)v4s5CwR~ttcFH%_`|SK1QmYQKdO|(YKfa1|x!Q$&LU`RkBYiJtR*EgUps{rd5Ev zS%ovw_`tU{=!%Avjf*dfy69Gy-p~umlb%r&)GQFx%it%c6g*G%4b!rqF$U z>!+K$j2-oNB^I2((bF>K9OrjL+*wjDZ5CBa*G_AU_;t~Dx_*@L&^!GOCN&$k(dwxn==DsLhuy zTcLS;!p8DloX|`Sw)|#_dE&Xlf(E#cMx}S{r&vf5E~Fxn!yJTBk1Sb=F>O^o5YKjA zlqc+!LzG)O+04h#Hob=^2ZAe+{SV6)t`hczf}AqMlQg)us~S}~KzttW47TZ<`K9t! zK?hqXLUux%dJb#_8-rN13&$gERvpMSt7xl++EdyY^9SyQPiUqdpOe)gA z$PW$;Ko~kT^daRV?nh(!-+eAo!l;mADlCE609&p^aV{`s4}_NZmx{V%sr23uOhvW1 zBpR9Oi9%~+99w_>v})y9^ML(lch(NPO|p<(V9=A0Nfgrs^MrXotKUZ7@lOe|+Z0=z z8Y}1C;{EozU|FoxYAQ2%Ms?ZA$7vOb=H?V~V7%-$5g|zwyO|=JsK`t;dNgsFYE)@@ zBuPOzbT!i%*dSXcyrYP5nPkvdi5OUC?1hc7$8~)?^WMu|FBkO;U`hmcLgn21fvt3j zxSSsqx}7)3`#{}Ibb@;MJ^pz*Aq-0u?wkZ>62p_We#VV^uCXV`=B<`x#2EW8O8JxG zH9s@T4WV$$56PuzV{~(11>?Snbm*kmc8av1O(03pD(DQ}FTq_nY|f7pq`-#?mm;m| zap3(xjGO>#+@M=LKTBOtA9)>mBFB2zj^T0kDBxUI*BhBxo2sP`3M|><$ZE!}P9X9@BLDw?+%C07y zUu;C0_v#Pw$zccfI$tzFRUO5iphyh_B8bk3FOp&JkxxwYcr3ZZ^w2&07-m11XU>Tq z^UsGzv(?JRMTTC{>>iliD!MK#nqA3cd);$|x*H+Gkh_`=cE4kSR3AKU6~uq#h9p^f zdtp;VsyB`{?7j9Y*}UFRJU>9LONPDo{1VS6w6NI|He7@ZbB=^i#|LT)f@fe$ow8wZjsf;8Q`LNIw*M?MbLPMhbBp_Dofp= zsNr|YGl2O6FJWkKHB-T>5X3CT9(aod5oW}OJt6r*q4yko?yOo7_a8=(-45UM8oBMj z@@k!lW&8_@?WagD75R|FvS=Nzkm;mx;!YY@LLFW2+{@EDqZ-i)t!Rd@8PuE6j8ij5 zv(7yS`o%#{tpBhk2&_7o90YkrdQ&zmSU}Z~q^KmFko(pW(CU^dP7A6Ag!F|VGtlz$ zbQ&MmTuhfk5||o4TyVklmCFFpwc^8r(r6YKbgkwOLxsI~=VG5CaG+sIqZAtOsi__QJJj935T+O>P)F@S9(p{(B?FuKcX@ zA!&SN)T$m6M7C3G3q`I|k$U-LU^3C*2kZi~)DP)WNZn6TAnb;n3V0taN>U_>%fhqh zyMcMaOnRqkaIvONo=Vrr>r~ZJO(K&DZTFx`_2A-6iH^V!t@H!mXyqzlA^8;C$T2R% zPFN{ca|D>*�=TYZh9fvJ zH#rzT znkn`QMJ|m<4BQx_v_Rbj>gRr}I;h<0R!`$DUaKBQdcsXO#kf8)R6!6FZFvEeN+R^+e`HWMv$7!isDLfJO-mLSuvgSb*!P z8BmO3(N$?rST+=!(^SLuvO&;FZ}K|M$9D}SR=XlCXl+o9Uo}6S2c|RF92&}Qfi*3j zU#CjrV}(P*z5%5sR^++{+C92&@D&4w0Z1OaOa?$dZw*8S`$MxgFi=6RS)`+LVG*u^ zGT9yTVYJorFN<>K?{t1OwD+mM_+rS*+&p1nUymF2Y~F2xF3(f^f&998I1{pr1 z0#h<57Lw9;qDeSwEitAy2Eq#UD#pYOYEX+$y{bhKFb+7*6e9 zV|%D~C4%k;jcX~^3 z_v3FFrOMRacRnL0UKy#eeOYFi%z;;&?V9dD#5INhp$0hr&R#{ z96=WN>UMe52Y?4!)$F2q+UvrMdWerJ4NX$CD=O$dasZ-~VCQEzOj4W^Tn2e`ho6P% z*?Hn_XPq3KgnbJ;6s4in%*Z{dbQ7z&=}|%N3;T#U%?=0ViOay$kHcl*Ugr)U9fj7v zN;Y^sM!hL7JhWP3UA!~T*UgBlUpAAsNdh-qIdFhK@GXs6tGr4>>~X8}jS^NtY-D*@ zwDPuW51qN>j{i-FzbFgG$_f>{N0LL*f}o~|oAf>UhUC%Wu36R0J~&4+-MVS1utdd7 z)2i}#!+}i^xcOT*t(){OeA+g0o}{Q3bbDxdsq{h8>b6NX5jDjsbd0PxTtBVG!m}MY z$rtLoU!?bj8L_l2N1jFwIxvG3ysmS-YkdjwsXH;(jyBdnEq<~^QvY*@s;Z(%mFX>nm!)_5@mu#PLN6|%JX6hh> z`wVZp^b58r;#SBGaD?tR7N&gr?5u~pPW2v-w>u)CO0^Tv-hy}+;^!L(l63XOGJh<)#O78+l^AIf<94i#@+u7h)7Izx-s; zUQ+D9>(n_Dfl*Dd6%;v2MQRscoOPPbRO1WW6j9>Q&K{w!vgq|}ia6;Tr}ulVo3<}F zSs63!_QG1aDdL15Sm2b`r5FsGL$6oV%+spYP3vb8y>jR(Y0R{y2$UdM^vWP)18a)7 zP0lHE=4YyFeA6J9qk_%`Sz`;Z*+pJV3_kk1Ibn!OT(DIuHDX9~PxKoS@4y%;G{I0B z#X@*uG8Ne)*b4?)gD5#pj>6K?s}Jf)QtSt8=;#{X2AKw`*A1z>w5OICje}!ywV2B(rfk+O;0@qVI6`2b(PKJ~rj5P!|Vjj?4 zqtPzQ;{bb_?0^!M7tL-L7evoD^qaUo@71NGwJ$K8Vy zQl=*S;a{hkyCk{fx;wBZz=CHrL0IHpMIL((@$yfLCF0(m$vjp)9-S7Er0 zz`G^O<~`Qm(-cuJd+gmq-}su{DjrSkZOYN7c-Y09gZIB!{ zgo9+52%YE^NZBk z&#r#W2#cjKLnI7_ z7V86FGu3S#7?z6RFuVMKkuPwZ)$y~ zKH!Ea2VR?UO)!;0vD+z_g~+IH-jFm!j2M*n%LgQVp~=#HOUhgmLK8wa^4pj_ZgGpd z>E6ZMW|-4sH|xw19cFji@AOMEfBtvoB!v!3?qq>>&>%P36_=$|eh&i;Z~x@Y?5Mf1 zW|BvBpR_|x%IuWDC0~0eHwui@+5hodlJv?5j8YSUkwvkftfW$rRcyNWB-wLoUIv6ZX;I z55sPG+*3B@Jld!c;J{4~IBYV^Pv7~cviU~j{rTPYV`Ps58)m@8JxcLCOtE08DFK-y zO_qKn%vRrCnC^Bfqy%8!6mieD+;5w>nnC`jZ1oAhbLx6os{mUc^OGT&tH_Q#P>+{Jf8)y117xl1AiPp#;c)@T#vE_ka=f<0v@InQCrnd+_@{;c zpS>>uXzEPY_J}7WABJoMb521;A{3;R#biWnw4F|8>1=m;XFA)xbLStK#XC#q&Y$Vr z>C7G6MP*Y^K@Ffl0C9mJipp9Ms#QQyvABW3r53AzNU7rgzDaCJB$^WvZtVQy)a0y* z`JVHB@Aodxqk7%~9WRM$AIW>hsDU8MJq%8JDYld%yHQdJmrFV&m@_?n+Z%Pj(LSL} zg}j>N(&xc==~U`DM4t4K8H3u_|mNx+>& zq604$iRz~6l#3R*s#MFBSTb8j*C5J1mQ>r=mu)G>9w#XD;IPaNE8J|Q*i4FSq~bDW zbVOdKH%DvJq2N1-XHX-n6IM>(dihE~vv`?&hsdBtA&DK(_~ve3$1;rE@!^{Bf7)H_ zXL3Qtb7*p=(QzVLUF;K5T8(3K9eAsG}4E57wqTzxn;NoIRj`BjM*hqlHn<$bJ9 z3*PZ^>ko_&qgkym@-{x-7&b-_FbxGK96W9CTzK3*4n|^j`o^>^f46O&{DNc$>~w&h znfxK$7K;t){ZOLUs!98HEtL43hMd5e$%pw@qYm;eO3#VQWA7+)B)2B%!+Pjry#Cl6 zNgJGb3C?WO)R8kW@uK@PPKTWp4Em-`yBv1@g{!{1_#ZB~2dVAN!KdhTFWqLgX|e^0 zEMAGUQ70ss!B7?`&G+6UyB%IjZ}4r^Y~*$NTzX*x->A4RJVob7K9=W5@OvPlD|~Y# zJEfgIoj|wvKrlFWKc_zhfq(~YS zmmt3sxjSOe7o3Bv;3Qt9Po5}=hn3qW#oPJ#iqtl>L4_d>{2VTNs5t(rF8b7`--gqJ z?zwccT|X$SdBp;b6luTxNdKk<;N5P@kk}GYj=yY zfi<@(5^Fh;6$eXzQ+UnvZpFSA@NBFGO@}`I!wS^Uz$42-ibIh9zgv8g#)?iP_R3Qk z)Y$#`DUD~OPppbeXOLmQpgs_q4s8pVeMpQ>Rbw^+-6s8vccC!-$)s@h;l>gE&X^VZ zAO3_R_TiuY!Rzd^UYyrp8e7CiI=>Q_Qt(fCAihT?mbE_V6Jf>UvNg{WerC%dt|)l= z9bI#VeH4t~?TmY~zI>ByVb$X+B@fOj*(tW#^ITcLhG{DWdqXx%gZN}n1Kq)VFz>AR zgYY)Z0KsD=ib7riZ zHBseLED*hCQ*oPvOW167lPoWEzjRC3{_uU^kXHH>24dc!g4)1e_fk8Hxr8>&67PL; z^7&P~d;WO=ox-%&FG;~GbHp)y@?bpqs51~UPeM<=?`V%sT)u4Pvp&l-@y~?yG8ZLi zEh?uj69AVNovJRF(*>)Y59Te7>kb_gEAiP-a2V^)t{ICr_RYlzvkc{ZuzaFzJkDeD z)^=iVokA3@$Cay@o)&{{*hg=slw1i6kjUS5CI-`uV-|Yj-YPp5;DFEIgpct}|Nd~6 zt#k;O;$dO0&onodh?e=@5JJQ|i#|H3NK&ZEoP>2oK+xOE$1KbLe>`!7IsN>e8wY>o zuNl9#75wHBi}TzyKCz=)Uczf<%b_7ZI~sH#z}V0Y-_jnB=%x?<^zEwP7G9CRPK8}c zV~27M6VO<>swNZGzUJRPwM`9vL7-tLX^6C=NP&(M%yH!)10+G65GT18i<%<+YUo>3 z7V|M5pxp-{+I+ej*sr_kO_Sb_tqV_?0cWG6=WYe$9IpUn;-KGU5FxFKgi4JpdXV|s z#o4Pw+KTxu0~SZznElfy{B0U^7yR`@1UJLn2!xyvUrA85Ku@ zmPYK9R{E|APgQS?*gvgJgLgnl*}q|bEa=;@ojzy73i+RZ4BoPbt`%GeY>O?C3__zD zmOWIwc8^@3JH*ui4>@;8x*8~UF~ku(oTgAptk2rNzyhmTZ=dQUmE0D|o;wEvZTfIW z>^Q|9rAPx6w?njfuJ+8N1SXwHicI1i3OAFZZw&x}Rk5@ey59)1i_K-YcLQ zqB{JZ0{_z~vEjd6wkfR2xm4#^ppPT)Os7xnE#&>#f~W6YX(FVT8=gEkVwY-Fb+~jq z8$X^rz@SSHy>3DOH?GUENmi%a@(Pd>1$UCN5MA&QsEJ8tvVcRl0_0W_qS~3O}5DUz$hfl8?n}7@g|m)I#17 zQm)X?I6`#6l>+^YqaYO0KQm`q->gEcNf(wvP^S^H6G+3iLxkidLtfbd8R2&oM@T1$ zpMphUC>x-sljY0gnUhYKz9$x+>Xe%&SH{#qw6Q)sRkADsYZcE=!L4E%ip1an`OEcL z2@~gl?3qhNxH zn`SYu4r=w5h*IY4CH3m1kkCIETRLYoN%J@GUFLz~AajO}UGK@YS%`w~o}Tc}#f_F0 zYy`?hlN>n~Z44<=W0#1o&UWc&JN8&Rh6%^+<;t@@&HTw(TjF1j4H93ma*98r^34eE z3%nY1C8#@SNU=h3b$0x$jkCIgYJ)CI(|r%{N)>&4NI&{_2PH`^M-M$e=-Z}2_o5KU zcy+UGg&r2v2DNH#3Xb^qyfoyMJ8#IVTU-$|Q3_vQ*V$B^;_li}zg zjxgf3G^={xU2I$IX{V;LhTJ1rVTpd7!i?}{;XeLpxgqu;)Q;%GF1=79$&pl0WiKop z6NcPJzX+!xcy!SiymA`?kL^u~8}sc*QTXoR`vsrPsGgh^j|`;4s;ipRPxcC<~p zL)+OuuUOC#wOG(Y-+uX$ba3u{X+A$w)}|@snGSZ&U;A%B#p!!sdM(8BKnYHPEo_zH z?i+8*csM)k&JL%e*V#87BNah>^3}yNr&|n4|4-#B$TknoJk(g3h*F9Lje%k+u7tk= zIvX!UHuJA3ngY|A6@pE`2vbGx5`W4&2pZz=1NZY@X^!L=yTnhYycgR8{Ygg%7EYWB zTIbh9`qiDn;^?&ClhE>eHf)RkK7O$pkGKIs&Vt6;Za4}B$Is&inxBWB3!yBadEocp z1R3&R8D3zOA6BxFM6qiqvWkksQaPhV%d%cw?7AGI<3Bo{`E)>VyPvDh<^+gw3*QL3 zW-DXju~Us5VG^@|R(W8n5*0CvfITx~`gQrnS(xzIA=<|3XJ9#1i#&@yC$8l^2-zS& zRSGxQh{uNT=@_v#52mfh?&A6#oG>yzIqlEucow7ahxhs$$x?2vs|SbebF56vdWub< zNFo*Yi43_akT*(;B;m<{o1-(PKaBcB*5f;7Vfsg-?l6WPT`)$kFXq0pY?=iX%lRPb~-A01D6T1l^s&`kS*S2pc#k#^0_K(q1K}1 zY2lrw_4E=Eo{$#2Y4-ZiX7BTpP~jAL?+h`BF!vBZJB*kSjL&Dg+<>Ckk3Zb}k_9n8 zy>W6gIlv7u9$eJWW`&re6x%?N!&DqrLG{0w9sRf5={+P1>Ua?tmm({tnsU;h^t?%S zGs>*OjHjY_<<^{c-?mL6KOTYg zV2{L3s<>W#P_@QStD9RIant|&6d-o#q!Z+4B5`)f6D0?C$YIjK0?i^xS}@p-4!Kd; zNkhA8#8t*L9pTAM;e*VcFr>z;oj&AMD%19<+jyse(F`PWl^}V*Y!x|N48K@_x91#_F*h(i8!f zV|5C5K3^4C%!m0Me8ZI!?l31vis0jz`w;r-Q@2L8X_6!tNv7;U;K2|hgPQZJCN$8S zgAM*i!nVS~)zxG8^uZbN>3e*qONQk0n{XN&&n@;jSd;s^7c8h*&{4RS6mj!%J@;Qm z8m+*8fMOx@Q%1#|e(jQU3Asa`k_;(|g4O`J3X&n>YN?JC(AQ&5%j?6BL7`=pFhy`{ zN?%}d@F#&t)zU*B^(z57GThfJd=SzkFBX(X^fOiB#@JR(;n(ZmOk7a#M&l3mef6}a zNa9kSaF;y;rl;a_43G9*&8hk6GcDlxW2L5+6hC8J9Z<|YEM|L%Vk;<8PQ`sRqj%nZ zX)phRPlcpw>VYZC16NVzdhHbUAn<=;<*S}PFlERK=?|;;2bFo?$&@HyANVv6y>Ky; zr)s7#_N!B_qRPCNLs!#~V!wB;DwFr2B29(&1RomN!d(@jaVU6>i_B*iR_ZkF+l6#^TBqoLuZe8ib;Q3{n}H9Z*1>`%3M6 zZ3|3_-SPogP+tl7fbE46T8yx0b-@{dTX}l=zHhw9v|#}Fl3u)lZOt1PGHL>d_UVxT?O}vUQoeaghOsj9;$`fsgPL-Hz zu^)PYR3#4sIY^wijE5H>B0GQ~Mn5BdR)_eCyebj|4ruc?16gEAqD0#Y1T-ss;%8mw zb<0sp9S?eJ#kPH@jo>hi^RDz-JjDXPZ|(TSZ^#x8hTj1zOI1vzr#Lal}T9)-pNr8P1%;2`*GXYEDM51un%8-z+ED+&m=@jt7Fw+puYH z55yuWHIZUyz)U2Z7KrQmymw=~muhgh0_Fe4LNT&);$$F*BrAw!nrT)^R zH~5;*DxQCZNf8|U7P#ZkbeS5vWqJkE9or!30g1pNXl`4WE$R}k6`T)Wr_PoxY}4dH zycv&lDNu_ER->|PQ4zn=rx`lxa01RAE%l7+VDvel%IdA5=3xmHSys1 zp&g!!_1JcB*}vv7)VMEn`3koyO|=JsJLc@L0rwt2-o5i(&@K90wo|o*?(ImY+c>b}I^zI}E(5${XJD=Y>_q;eK zxICg+Q9}Dt_)T4~Cw5wGm2D8&WAl`DqTBWAr0@^sA(u=WZ@XU!54G5;XCq(c`bcC2 z*A|1XS{H_IkxKK({POvR+10=wZ$6yC?&iNI?1?oo&!AtCN3S5a(UeY4g8{l`P`8OVyY@@~@?#CgDPa2L8;u_@6A=j%bb zv2RvSES}>oC^?1~yOzUEGYLl;HtI&p<%Rf@w> z-fO0sI*5V*7q|92(a!)fGmN=Hab`M$eQ4!@x9OGsdH$}SOh&c=PQYX28@YHdH*<1I za7X{F7f|J?^s-|FmG2Ge0(w<=?mX-S(J4>MKXy?W3x^?c6c!wA>5@}9!Ge0@^*8?8 zmRr>hg()gVJR;an@b71;r`?UZ6?GvJQXG2dypUn!U&qm57-qgl zzLAUfavBED4Q>CR;BK3P?eFEW7XtT6_e%d<9($45D61iRNrU9k&KiO-#$xrLYZ*$O zJZRjo@#K5C@+c>4cy0`#@GsHX0*l8nBjaBh$%+Xi*J@(lK(Xs7l0wCqyUDYouWD9K zC?o}%K3RcxHLp!m^n9OR67S}ik|MxJIt~n14*S#Z^#uV&V!@rCuohtWe%k*5+4zi= zwSa8rF!xkXu>k8FD()Jc$kv1xPDg?{nCu4qFjKcQ6kE|iv{r&boKTSqLi2Zgt7mVC zh9UxVXOZ+PojIgLl}7jatd7Ha*lhUd?bFfe^cnVHO?560@wXu5joPhe$$D-HJ`c{! zlvrVCJH=*GWGfZdE;Eompxebf)a`BC$Z$V2QO`ttm>TN$FmP8GDyX3nZ9{) zmEZF4V*cmvS~W+$dUt-C=0CIE%=%8l&mMgL=7ND&|MN5i)(4_Gh5JGd3y!L?y^*kT zd3b$T0o|xrJPBzcD}9cUgQ}IG=G6}hVDI4#MD+k4R65fDl}^}Uzgtlmvt@d=H(sMt zc0^UiAiw>PR|CBdilTN+TI{`kVyouB>!&sN9ur179u?YNS(c~&h97>jd#&d!T@FPNaT9~h1MJ4fi0@gSXRsLUs-BYcZ{kc%QV{x}_Z$6hQ1lG)bnl~^z zi2TOwveG%H=Rh&Axq`*L9`onJ8?}=aWebz;^a>zQtP5cdi_ zN%^>!zdY~^6na05$qH$gZBw;rifKHCWebg@#M^w2g$ob+Y@LX{@KKdcbvdAlHedH3 zFb4{AtLPMBgtdJNyGz`wM!vHY?^J=2S3w(iDR2+GwiRZvbO!I303`A#N{!sAnUgHn zo33otT$dN}%y-7`Qx%Ed=(NWEe8&Zuu6LpB7trka>MMn|b#>2FSGJv6yW6CS##N0$ z4Z1grMTYs~Z2p8Qq{W0DHlUgfpN7ycq`pb$ariv;1&o`GJo5Q(|;)TO4U2 zZXS&X$Bu7Wc{CR&_5+HXrQ%?vOtJ!w#~4j@Fgm|MzacLp(oT@{wm7=w<0(^A4a;@tuaMcQ^oKSM9;ib;~-LGx~% zQs-9_uEkYZVl>wGy%)4r08hMFWGNbaHUVBDjuTQ7+c&msh5w}S~zXY%Qpv% zn=2lZCp|(VAG3KTM%etH_wUTIEy}c0`*>WL?cFQ{ZM<^DHnt7QAt0<@>679fvogY0 z{d|{$Kj!BbjL|zzHpg?HOuBerjjiCl$Hrgmh}&Z?be<|NKwA~Ija@&nTa4k|@<2c@ z7H4LJl}4Ddo7Dm!rpgL6DxueUqohdU{#Cbw*nRwYb~A=KVRxLN_`#YlTc)L*(|LDD zjRz;IE?P}A%@li_B1fsXE=4I1nummYVvE&RXYZ5g!!n*bGu}wznO2Sd$2rt{Uw!Zz zrm3pv>hRJqm?nZtXKa@q48hLmY+kuf0q@GU%IE8pMIlwGp9Dqwm>S2`S&8S*MlK-$_i5T6kA7;YAP;1A~W<3y)|GL zogW-8It`JIL>5|7>BZ4Tc60D;a)MM(!>uNf-lvw=s`+Ss!*_2kI5Pk84=R39KL504 z2VEF*SzHZG_~tWk^|U#-pT0-(=*HNlz_VgQShKuM(=6BfYz`ir+W=C1=KVFQbI!6M zem0P9!_jA-;M^N-IHIch%N_J+GL^m2Csg;PYy95+)VrNsuD&tT6yF%*&QbVP9R2B? z`vIN5iqlW;xgO1#qpDn64jC?4N6%d!5<84DxvFfCi7g4Z;Ip0IrFiI-MORKWsP#~| z_0X$}ZUri$YH}IQfGRIwX3!3hiaA)*6+c%`Uz7qt15$%z(eZQdg?7n+q%}x;Bf5oW zP_OYzoCKviOJ|%Do2$LRp~?)+0+ZJ%%u}HXZIJ{yjS69bv@51xZPL+(EH~ub4KZC1 zM0^+mC&>z>7_i8c4agpP<*JHAMG`Z2O)vA%3#ns|#+J{)+7ADM*F#>uf24R)hNqa2g%*$09-ujVcYTNMl*S<{7c<|KLXEn85rC88$ zyG+HEg$(lZ`P*cF{KtQ_Y5GEULAU*exoOing{?np`3NFir=AB)RE@D=Y3qGE9L zw8nhjEnmjmME}eTRRh#bU`ArK<~*-LiTMbfG96?jdjifw{sE69&q46zPb-c1vp%Rs zXjI%6;)g!LyK>Y3+jWpeb-=!X=461Wp8x*R2ew*W9=me0qsx`eYz;df-WroL&B&C_ z-=(M{tud9;c7R~$Aisfrr;5&!7~ID~G{Q)5g2o8fyK5IWXe^PWRDaK6gP1j;-ylsM zY=gS3Y|tf&Jx`Hy*lCH@9!savnO>%DDp>wrrjXYF%x6`?M7Bq|jn~c8U`E|sw_FfC zK(5f6=Yk;#HL}Rkhe^-Msg+ZAP3oCmAHG8h6v29WhbWV`JUkCb&G6)WsDCc>mB4fe zcY1=^F=K-h!bW9itS^2VGC3h^T)uJfN0TfFYk%?VA4wXwP_+jq-%G9Vm`kysv7w{l zuE(q+`4P7RdT7Hdm!x`7C_Wjs%|8RGnm?R;+h;XF2~uQOTgpQ&NaV+_p%c8#C8$pY z#nT4oX~&%f)zj`f+T$JX*RZza5nOz64~B^ys)<3mFYt%8+-w+Wux46Kf+u??go`UqLf zd*EES=!pa9ENZL`i~XvJCw}-74u*v-`pL5%9oiF;DLYK^z4Jn$N;sy6t_|!F^azU8 zLtcBpgltk7pboegeA%nBH$atpA+Jfi#lJ^zO_4PBl8JJ=G+>kLK?uaP>DFjHTy=4B zNi4p0ksiTCDZbmJs*E(h8fd+`kasu`;!TheMng1OgX1D(*Z>7KpD^l<9O5)dQR?sV zmeZC=s{ebz$H)>7c7vdqW>|;TI*LuE$XbXpin1g<KsQwVRg|MY;Xd__%kh+4hLC4@qj&+Ky)4qibGbQ*hR5Wfm}evCBU>%L&uA% z#L2<6JUuLwYk4SCIw(CFTgzJ-dR=a0*0Rlt%Tg>JNflfbgUq_3QwU-krt1cx9)xr$ zx*(hk`Eo@QaTAmr1Z)+IDzd{OBpD18P5r!n=}9o+Ng(*&8yw%6tH3Q-UC zXYF*fbbs@bG>2aa@*l~PD)B)`V;=I~>EAVVr?j1|jNJ3w{TU53CP`6duo*OPfyE?x zS+u)RfMe(s65KQRR(G{;U5>UJQ*C9jAIFI2W>$zDscaBk3NcjIO_TqM&I0cf?`>9rd7-_>kNQxHi z*z=mjKl%OX%&R+#f7oxy-f7QF-3$&0Cv=WEZ%27oIZbOv9~m39HkN@BG3f^=|VuBu6u zs%{3iN2mN)4Cy8yPXZbOWMR`OdxRf~&8i0RKC37!+&e&HQ~Iu$s^CS77NItR(NsW) zoy1rO2^?3&y=l;)S{%iPPs;KJYp03zA|&tQAR)K zp;w*?*+4PFnBv_ohIHdCd0MdLMwrz^egag)Fe`6hjdc%chkcmB)YDfKSG}>l_d(R% zC~b$dNR-CMFLEki2#Ee}$rD-A0nD>j@e^5-P%2R7qniYJ8j=B-7Bf~@VG40(34WPk z8Gd`rmYNk{v{z@>C_apW?}cgBtMI|dxsPI-@gk4SeA-9krIj0Bd)A8#jm-Yp`&1D9 zH!?i|dwg#C_rz|5<&zeBJ=2-}b1(R8g}Sl*zHN$TdKHDt)Tw@8d)z+5Ip+tBIubL2 zMY7pt9}XiKsO|q2+o-S@sH`c!{R2t&;AzibHSOh5Ea;SJE()ojDr34s>%-fbd!&J0h#M#wRx3E6$`oD@SI)a4Ug%z|jAA0Q8ClnE za>5CcTPCWACd#T8-vx>7A&i6Bdo9bE&u8PGH%>Q@^yI%U1= za&#|0XPRqa;|xCRILjjsw3B=7+nYCj_N8&QEN@?sY^5FkH&~zrR|I7NOPBc-O4wiT zn14xnPP6LuHh9I7_`8Z;=Cr(7IHb_?mPTv}yduYUD+S$5S-?h#YcXRxq@%!)?X6t8 zU3_X=k;Y#6C&wrjWKiqP+<#g0`%2`vy%L4&_F8P=$F+`r2CjhG zloz11rD9H!WQoZ99-HBSwao|%ZY;8?<+Vdq=c!<;9X+u+m3BA(f*3pM3~F5L$49g% zwMa6cr~6?&pTtWJ9#V8OS|mg_@4~w|gQQX5!;YyLVYd^=v2SBW@O{Teo^JS=uLVsN zU-`yKvStDSF~4EXz-EfwM8QASMA(?7Feit`OWX&(*EC{{qE;~`cLT!id{;P_z`}|N4uQ(5GO9-HVDAhPJI3p z_p9H<{r>;{>^BSl^w~ediBERvC5Hd|<_RwgI*LE}%{wICgKZ7aU=4$93dJT;WHl9s z!ImEAhkoCErL8a z#jw&U&i4t$_EV%6D$=Fbgn6{KhQ1k*BvG0&(Sr z(50dGh_*|S0d=G;LL;k1B{H4zqp;LT`O#f5iL8++49N*t6OM+~%z0{Fa+!mMkP0oM z&rH4mj(0)G?zym;!X`{kn5+rhmq@s22HssZM3q z^Q~oSVRkJD@z^L|EAvXeZK9KllkC9@a62i`^1!rUWUWP7zZ~CQ2ANM#d9n+#56G8O zI}a;^<3-)#ToW~NRU~H0$HKZfWX&E;x{qGRz@T0I@f*Lf#fYa)ulA@M2Yg#1Z~7OB zT9sLnJn1z>ig(MT!_scW^;xLlj%3H_GmAu9r(ptXAHPft24TwfZ~rroI=4BD?EQP@ z>y|m}ox%e*NTUb4HNd|%%&lpo*wYj_NyTlKCh@XptO9R@EI=0&Zf=(vc^S&?s0_$p zCPkJ9?pCyE`gz&BVqPy4M_^TWQsfrzE&iai&a4$w`fLbU3x&U>0RvD3ig(i~@eY_q zYKN40nk)45e-3+$or3ki5kHnbItbvSYrXMceaNll44!<5IrDS>W^T){@W|v}{ z>0U=-I3dz0gYw9*aD3%rhd4otBF}kUwyoImSo33tRlN%q%jwMe1)1Qt0ACHvK9XK$ zmstBS>*CNS!J2Y5o3;a&fW?{p2SsTyY0pL-fgv+3FK9>_p#{FG7e)hB8 zKZjh>*yn3@JW2x2h|?!l16ZyowoZj0fVnA{lLg@Xqg!9V2^!Xyx^;^iG?sm5-X2?} z!Y_!AYe(0x3);U8YE!>A66h=ufzsui=4gm2nYrdR_%+k@ycQv(y9fRD`1eh%3%kz4 zDu;(&ReWsI?i60T27 zND_*E05hnG6-;aSA2w{41e}m!A*l8sun=PPFqB@o1qjxI=o~Dp@8OQEmwu>tm2Yuz zy+XZ~kmVy~B0YHUw_6QoM|NHP^y%?C6GC9o|Nr#iGM5LebOenLnraZ z;OQFvPjvXV9&Mx7@DFw1z5U;MgnWzVrv9%yl)PsP43Ay7*b&Cdl4OJ(2g1EB8Vgqp zYST<{UVL6$KIeqAdZI31Y$5A7V#gM;oDe(ycLP4J|H?9x9QoF*U}E&(V964z5Pl!U z-l51XDsGwY4Ph5u18Y}&r^R^bQnHOs5?rH!T)dl3BOT&RyoxYwH*-#AA{tumUmkdk zZkE>wVDk>W+#dyf)-S+%T2nL906Ffb?V=C7s5OYs^MIrL7Rak$bk;r}7RrF<9 zmLvi6di1n`TvKUl=j{?B`3$7KdzstxTK`mntMb*e4eERTbzvSGz2^kGakGmOqSY4I zy|HgcHZgF^GkLD1B`sF2&ASv^Ly<} z^18)a!uE$(P^-fAVSTa+>hiQ4Nd;lmd**(6?V&?qe<2UBNfZDen^12gawPHDgV zgMH&HC{k_T_ZRZO%VU%6AmTK9y^~0>t0}S)RK@&L1YPt1F_-6$&9dh4pVW~Ne*7x; z@8M=uk_zWN_caS1#`pg6BT_#B)J=x}?58RABt@EVnRzBgn@{%>!6%EDLNUYOS7Sp=p*lH@h*fpOSzU8iK1IK?(mSa!3^8n+~E6!bXj) z0j6#an8r+<09o{A(FKRwryUtSdyk+~`yCv4PQ&N9%hmfoD)4^R3uPWDYM8m^vm_g) zJ(0=I!2x#SG)J)N#BE1?2{)`xnDLdjp0k*mkl$4$RWz1I z@JGx>+5Nyfs?L~F5Vc1##wIaP4)KP(77lqe zDh~;>LeD;lV$(TDjSAw!u_#Q$@y_dl8L6UK9NP+@1{;zf@O1O7&)QCdH0 z)+ze~VCRUVc(HT2+ji%CeOOkgJ14&3!{q3Qe`XAZU+l*B_ASIn0qB1hFJ59xO!fuI7uumQ zGwZ}ztOhnmX8|v{cI?B!00BoKfhG5f zS=vpr%4ZbIZbJ1^cC>a|@V!{5t*&@cyCtkPtdMtvJUVPp8<}h1KNs?{1$cd@5Kn9N zZlpg7Gu5$}*F!qx17-=3zoQ};y5mR(GUyAPMDHt81RHsmq?e>CP1j$aby|*}0m6jl zXhET>XO>R+=;IC(yt-ZX*|m=jG?8X`rx35qqwx}LlEe@L6b$fXia^5si^2lAs#+d`YOLm_VNbPVHfGob zWAv8OG&TO$e<|4?YFRxuyu7Z0)gF1_*$4h_5mz6V@X?3Y{0w43ciztd)UMdwF0IFLM%-y=)EM9*v5c+Ce&ZUahJS zVg!#a?4+^w*0tzz3^t$nHcx)zLP&9f&A9*SpBVV8v%MhOs-6b6)$+h5X#db20Fa#bY zC%Wp3^pmN4U9hRaSQp%{*3U=vIUoZeCYK9ztZ9k%s(dsiTwS!-ILoi zL({uLH%wkfE=rmetC$*x6Hrd#jWgUoo0y#ZKnE`~tXLO6&*Go7zPT)qRB$U#@Z4ZD zXd@5zX5ORNBNTZT($Ml=CXvMkLtSvTcLBXGTmONwnHK@=uxjyO4`S7D41U3lXOOJ48Q2IpeRi;jm#MAqrl9)DA)=dvW zdIL|#d2bP_cMXD?^5^$Jl11l?t@Bv=3{+2h;~;vhPh05|{q4yXEOpI`xt(oidH70!A)r}a zFKg4Z@(eU86nqL49%jY3g&$5@h}_UuUs&&^0H6!dVm-_*+6vN7Tx!b{9Sc?~j69Uz&Z94gbEfU)C(ku9AJaFXP#^YAGdwl>^~rhTN+nQvtVx7A1ALR zSzj0jr_u_IyD7Gq0{$%EjM>Mpp_hwL5D8-cfqO|OWHLL2NN86RewI{;SHG|VXzWgq z6hT`ccJP0w+6NAhfnFK9OkiX-E3fmr0Y+D6-=8`tO_uCaJq82YpWmr%7huV$&wbiO zbXQHIZ8yr}6)~P`Wr>{*m6Kqf_5`k0H3e2bA1_McmGUm~sweIaKsuk}k*o0i+k1hk}~746a&4TP=0Ja$)u&-6MIdS zM6QE?ZKu$nK1PyY8Eux?#ztsdCSR5o1LqTd5O$^Diu%6)PHwCmj}5Bhkl{STJoZ^! zcG!o^2vrc{cDxWh{aI%T(!FSycp}XFH_RO3W4sKQL9Ua2x=YfgSu=T4;3@iGNULTe zuR(Hea<;S#C{@R%Ytt!=a2aeT&*g*>svzxV(!W|v%`ZxRIY?SY@{~Pz4*J;2{B%(4 zMT)djap>6Ul!m}g@rvL)+Mvcn%ES5w@T#lmjPPUNv7Ba+ zL#RcLssl$S0ZD_r8w?WmM7-x+!$-x6^oizNWiKdC)P+^i7b5Sd%^C%$bWt8yB-$Ah z?*rWn>C9UH_q=z;wD>2o>%2>6Yg^=bs;m6PBCL(L%3n&7p$ZzU7;;_V9rW~`Anj9K zAg6H0*-Slt!PDSqA9tG4ug|nyqH{?%cy1zv*oic>%M74Kf-=`SRlDr=ES#CDfy3^p zgZs1}nfr0a;KKdBqrKwy4>o?)c5-^$`Ssu>x*amZg}LfxMLvJ6PbM#0Sr)Jh;!TOs zuy~GI$?Khd(EpsQMcgdU=eKFD0GCJabR%+AMH3l{d)U`<6q^u zhiw-~9vcO=vlv3%B4`E(C=WC>KMQV8*$%X%X!3JaNa-@hCc$wKJbmC?d4_!ujNsXh z5B%HCx(JKca{b#kOG)7b(qQEXR8VX=MGREjfclQ~U|99EHci{B^{*ZNX8FQ4O{egk zD@@J&yNZMJuWCx?5AsXr70=5CKED#l4Yqfdan=y93MNHn(f4Md-*DFNls~RvkFiT& z=vt$Uv(BknygP-J-`(Zl|zM-RHY#sh~Q1KXlj)`ZWcjKZ{=MVTCcG zpDd~m+9#_3A!D=qxkNPR3k-Xg{JRP1CC zgWX|pm| z`>*guLyhRcrp696VmiN4nHH>vAWC*LV2$Yr*dCo2iv*h0W3h~~9!f_+WPO$Uw-(2% z0&E#p?NF0$_|9ceqJrRR;JTMK@GGeD*rg&vV2RznbLEOW}3;^g-RyW$8vqHgCPZar&}|YPYeEIr>A}eI(uVn{#?vJ=ZIaKb*GQ z){J67T;3^zeB3?eNk_-Cj&8!Gz1woa>$r%WKYYnn z4b)>{VMpJyKD<*HPhEkMuM{XY&;>WkfFVo)X=zyWBR^Q1W;efoCZq)Di4Z;L{$_EQ{sN4rs*|_ z&9ulHrKkgrEl3Z&kex7{sStQLaHS$1y7~=Zn5&g_9$Rl;pK##aSve`RD zkf$mPIT!*uL(018$YO|!f5=!<1gq;7p&`heg32LSH0~m6qN&^Ocw+6gN`aox7nY=mju*wi{Bc=z? zvMi!a(G5XS-kDta#x2s0zCq)hfTHa89Ptp-~$xTpaok zCr;vKZ&qx0rC4i$%dyyL0@CBba7nRZYF#p(g}ToBR9t#A&O-g<9@$Gi^gS$1XELT8 zr3dK>>Lb4Xh0M@Aez6Q2Wv_^L&_lDZGoVJ;KwkzhWzoC+Hv1(<_VN44V$m^jV`hDL z>Fj>>9lv^LTdEHqexGFmyb9%Rc9TkhP8BZ#I={$$AqQ1Mg2vd)Nhiqi=a+??4L%%L z8{DR;;@1lDXjB}?2g1KIWNq~N(1W}JdT{PepL~7_(=2NQ0r`I55vYyL3LW(AQ`ZQK zp4TaxWmkRjVGdnBA(>hH+_}I@!7C=D@s5)&+5T^CgQPkpxf4RJ$(!X($Xq^J5I(X~ zu6>2?#U&w4t@!V>6Sm1qJNdwU{9<(`+obB71u`Rx_XP~UW&GIdf}kbXv`pc8~d zP155in}>1K+unziWx)IHejmiip>W~}=XMFFq43=B#aFL}ci9>$JN6)9!6oSf!Of^z zUXAJ)JWGp14?`<_Da7Y*2Q<<;zTt&+{7+tL(_B>L(VK&tM5Vl&Q4Lb$;CNqI1C3_J zpa&tFWT}Gt&{u!e_Zm6H?1g6PBP2dzbFhxzEZ6(2^Jx|)Fm3*)niB8yd+6CoPmtcg}mDVHL9CYdY?nH3_Scc*EClkPB8p?UL!qp zX}|-)qwk9rs8fjlhFz2Hk!}Wq!P$a)z6VsjeDl8qG7c}~lS`1HNQ8>=%hE3fxrgU2 z>u;=U{1;o+5iXX_gG+DhaFo>Z9?m+Ux-V^))rD0Gw8w#F{C|Z7$1NZ zncMepnnTCOckF)dS+BE!ijH^gLfJ+}c(bs;uQ4oJv{RZpuOu`jyc_7H*0NjB`y9Lf zIRp%g6FYX_imRV{=ZE?K&R9&v#?B}TNlT~P8VmB$G<4$Y z=`>yk|Msl>m_mOndg$P1OHd9}o6fYrx*eMiIv8Z&G+*Iv(&14ph};PrZWe?nZtmh) zfbxg;`Wwko4+ct(6;RewYzjpZaaCSN8Wks`X~Bd1+?ZC4{+p*YiR^u8znlB8M;b*3 z&BsWWx^*8XR8YY_EmNNP(B9WCNxNcTRYNxmi$bcz7zL>k7fx>vxh}`p3T6*dV|RRg zX++uAE=fCuA4HrFH!7A-+Q9D)Tq^qf+DUm!B z$BP;z_nCACg}1f_?4t97<3*?C$PaRiBnNj2vjh49(G@vB_o=Z}Da?(3E{29C2rU2o3Udxx9ojVONPldtAVY<84NZZKu?d1ZzKf}Q zK-v&LB`bKBxJGz6CO71-`TPVQopMWXj_-#Nd+7Wbt(pw)y&)&~9Ra5`)dBHtDY0`L zG1eBxYRe6%soathZ%bO{vE?szXvoUv6fVHDKqusz3whX$xRJ%RtyyCNrXj`RNm$~b z?E?;9RPNIGoejID*e0tF$D01_(mni2p9KGY7gOKA0%fij0--tK0K-9*8d#i1-{y9|UuuLPW=vjwT@vLLK?0lsG=vwPCE;D=s2!iqtP zAS+_*CLzvPav@%xbgOmycYQFLRE+0lI*FZB%spZdnmRR!_W)e zDQp&B;kC1C1xAG-uxd)H=K5mxSp)X&2`LyPIcn^6mLJy^N4FSb!BM+0m`_5yk2$psI`eQ1zkk z7|EH|L+|}E0;ond1dNueHKh?pCYOa=66Vu*w3}{L+!5>pa;aYaC1E$cmw#;Dkk@Ty z-_%Ni9s|qcmc)aLh3r70Zu$lBV!?q>$O|8jfWm&E zo_`&LE^wjT3NhCzdMnlOLy#j;+@`hW0@-wCFTwIm?a~?N#LeQfs`QDA7M+Q~bKQfB z(LurlkukO`hM&H4{JR#*vg!|q?vmr&EQ<%j=%$rrxj?ZWP@p6*Ea_7cU|#aJ%ZyC7 zJVE}@3;UOm(-1q+pr^znbKfolmfb0b1;`9zh<)hQ&X{8_t7lh-BYOh01Ber2wY`e9 z@*X-~v|qY61Dz>$u^|bE`GU9xHrpqgYs3Z=vGKV~>c+UNHaNs=4rT;wv}iqeq<1 zSta?SezHrJ@cBX+47ps^axh*%A_ALC9z1$jkW-fR|{^_cYMbH zv!=(HvsrM95qSKj+vCVDHyW>RD$6_+t%9p9iqsc!w0X2tG^CiAO5|(Z}L7 zOjTH~P`hbzH;s7dRN{&4vOcx;;_O|(fY>+nmK>vad%>F@{^02w`sQuX~x9%VFCRz;6 z^Lwz#=B3d2RvsR;PRrRl-YE1IMKVsC{(G&s<@|K33R{ z3Dmin74vN2O7mgwUTMGBpFYU~koFhP{*k0{1BmBBpOjhwB$r}gZKR{((DyLYLuoU4 z{qz!$GRuL#2Yut`W=o*Xtwr80%T}+C)gu2S;^x4M+VvB=K?fw;J2lG9wNB2!6DHS)!pwP#FZwqk2>6E#uA`xW%pfnA;y>^Hy z{?FUG;Ecd5x^ikKt*4Qg;_Z6H$*D_3@uIg;Bj$Zj1wuV=6dtc-&Wq2BvF{jFicpKl zpvJXM^VDv+>m04oL+6npaV?NIL1)~(obUhWS*NEw2h#7?Vzx|Q)WHk_2hU>BB9xnN z;$IGdq-_W3CqoL9q*}Dh_l6Kur49h1y6K%UoK~vN7?F}xP}`-N?!WrcWfC9 z><|TH1t)>$S2=Y+rt{P0fo`RFqa#X-++6$QU9zm;JW*;avXJYP*d?LGoz>GWL~4(C zcgapn*7>>7J9K(HOb;lBADH8*oW{d*R|t*%6IQ-unT1v*B+Vh+9y|-JvEmhcK(T`q zxkttIGKEk#3-ZsfTwnhDdg$iqqI;w@GF-$U5faulX^jC(LX%(V1^v4&x+-iNdo8L( z9xqxcx-P%TXz`5fXzW$KEiyq%-aRTQpsc1@Slk=d4l2Nz52`3{${qieiQMUfD9ai)uY`^%ld`AqfQ+Ls+{7mnC zFB}ZXn7$wAic5LrKG`snY?NSg&NiGg_hCT`fnZ6G|hWZ~}x z%S5(RZ(Mp*Jvf4AM}V;l%3ZPVuzb!5sd??K%~N%OpMpNSA?Dgx)bS0&BL{pByUP7< z4}a;;Z_l;(A0bQE>>&F+*#A9adYr`R@%oW^{pPT4IVWH2YN@eN&Ywl`))H&XAvetG`&`Jg(3 z?CA(PTtD6PM-ssz35tE2G|TjK#`J3{v-%m{A%p1YJUFCmRi>&xrT6=`DKbdMbEx2m zSHe;Z%pM!v~! zm5)QkG-H9d$D#Uk4BAIr^55QA_pA$!v#_g`FnL$Ty5{oV(}yqY>xy;FY!=an9}h zE=5TIh$)r?bPBOXxdyTe_-AZf`r(7;AcTfrGA8eT^IF|4+tLS*U2EGZW+)9nfZPGB(_B`U z%ox;qdeE=V{oL8yL1M;>yP?8qY8m(1)$aMDSzUW>s}wZQ4-evQpxE^kNulBfUaOtI zWZ_0h4~^m=O@W2{4g3uu##!}}8hSO)ZIuu9hTU-(*}H>#_WS)-@+RN1B=QRNT0)kO zv?TK2;Ri1DVRaAb6uXWh$yD6>N!^pNyKHzd?$~#mJsLBINB_|UWAtj@h+H8B#kRFC zUyyK|ol2QDC@8LqOz~eF4NH_hVKX1AUN8)Yi{@p#3qFhP`*+iagjK-3u?Q1GMy3pM zhPbZjR;-+zI0-IGh&8D3&7z}|uvq{o7?4g1IdX>-_rHX5)@7{pca4AJFsOz(Fi&~w z64)J(t!=Tb#VoaYNEG8SzoPw{*nen8SX6qhS!nB$*qO_(lfAGmI5cY? zzf7F=vUWHBg7`cH9WK40HOJ2NQ(W^z+5gS0j&K2n)rD@}-~s50~C9gB0W^x zH9Cn! zjJ8IImswq)gG5adxgMia9#Gx)-x<>l8oFBCksZC#KVjkk(c)jYaWJfO_WFt1uBijW ze7sS*%&(GdRqha7}i&*yn5dH$3|GUkm}xafSD{M=C7P%bN1rKag|}hMvL70Oe6Eh&p6J zD20^rvc0SPmhnm;T#M;NR2gcF-3@BZS7#4-ZGe5|{is89GIFtC_;GhnSxy@rmzkH5 zZ^X$PqbjET@XcRaa5DbE52uqZZqu6Qx^rZyl|j2pu{{*|n2I|HHM8xqHGbNBI)$Lv z%wmy#1{O(ZP1M}D;>NZW)5ejYB*Bm(pWf-85IYbBh3~Dr4k*qn5~VX}aCE^dr)0{C z`C#H&RTJQ^YYIHrP_lnV(PRsBX+`)8LeJSzO6Ndw9g= z==UZSKacCK)JgZqQqy#`TCsUjo2Hg$RQ9PiGdH&pr-7dh~vU^rgK56s5^9N#9y@y^YLAq#xVJyvKxAWnj)?LDDinKXZktK?Jn6P$;`qZfcXat^cN>CSmR#E3?(Cm0+#RQzZZhIg8Uitj8 zAY6zexxn_Y(g>aMwEV0hWkxB}rl}7vpL5&0pA^e7$YCj(rIi9b&wM|;re0mhGcusS z67cy|ju;V}ksceHV)LO6Y}yCVNU7g5Z*)!>jj)>s*Ar)1K`@nKlPI!=iYo!BE>pWV z^t%{=gMOU=Yc9r!8_R0^97K@SHDj@HEo#Q!Z8=~)c0q23)wLfA`r<|B!)ui_s_h}Q zyh2`PXf6~J-=RO8l@xh8@TzaZ+>Ef*ku|Ez&_#NODGA*I%kUaiJ)}hPqnFO8@Xhg! z7gdeMd41YYIXL{zdlygp0(OpbGa$>pH}4~$cWa)T7tH? zh`I*d-Un4#G@6fg*#&X8pBB0I%nB*xfw;7GViK`41h%_J2Nm}v-Z?>K{H9m+aWabo zBfca4JxSrV`u5CD9tof6X%rM!#0X2l-f7`3)I3n*J)x_bL3vy)$DEfDejwEqLL@e2bH;94675k191 zAzKdS?9B{~=}e8VfyO4_9nyngK<25ssA`V9!(>d?DY3gjdtF}08}u`u-lx8(`uvdl z><$dCvq-^d%hl)@e!p*(Bt~g0Na>26l1UD7Ly8B7@*rb13@OJbwvi(BRNPLVg*p5i zGqVL4z001~Ft?aAdtYVh1bQg$xx^d{OJ`P2NavSFqzVi{`TQGX)r3mH5D*q+Ns@!x zG>4}2^5aGQkcV%UwQ1g0E{!m}a!IP^trWCrvcW*%{@Zj3=m^$JKAm~>1)w&Wbz^3V zpe6F4@6Ooq5voUEDn@W)+Dy~YL99+v=3h|L+bhRK}w8Hn3z{7${pDO{aF-|}9&LYcEbUS;0 zXRm8N|1ZmCT43?VN=+>(_TWI=QLBMIM6neVDL1P=BIg~l>s<-xn+2Tstvn=e=oF$n zoI%|#J1IV@F!lK7`1V5lzx@BP_bqTuW!K&w@q}b!$csP{K}8}6;>g2LQ71Z`cKVn* z)7#sxy}j*x?QJXVt$o}(ZPRIIdV3KOpS%>5hXzm}@>E1cR9=pX;}aZEMnvTi&~cbi zP-GZ|Z>=OaByuDN5^l8L_!~KoeRlBdf7V`m?X~`Yua8wnW(_%EvqX;r-a1GZ1+KgX zr5=(aa4FO$>=Uk5uU3Pv3p#X=PuQmX7;+Dj13Tq53h`PC2(u|+jFa(#VB~Y`%-w#E z)b@Wh!RP9aO1~ft4h$d2!x~{HIw(aeMXnnYvo5F}iL+tj&;T4gZPY?Brqkex<_>Rb z5Tt4Cn_o>QhGVD=)fP0!I|DRnbhor9bjS(WKe}kxtCkG7?)228`C%E_u|TV3tBDk}0wUvmFY#$jXU0$QLvyKXz?gg69A_Lz%azUt(vLdh-!r zw;Hl|)}E*ApfSzwzyGkn1dSD~d7DZ8bGCgA3!)KL=?JALqsRde#GtEXiM(~QA_Ji^ z5RRb^&~W7uRLgX78|exc%j+5Rof)_%)TG`xJp+J+-k9VHK`-#QWF<+QM)sN^Ba2c%UO+n1=O(IgCe|eQw78Up9r3%c0^Ep<;B&(Fn#fKOmV#`N zJ`(4P9t9N!=Zp4v!8&^%=#ko$q}(u0tS~ZS+O})33w|?J7;)SsHP`LKx1V<=8ccAD z46GOGU{|r9w9$GA>@F(mq*#{J?in>*$2mx9=@R)Vm`v(q_jvc`)^P61G6QTlZ7`XX z@u8QgH``rjgEg>_624bs!YpK`-QH~KFVgaStLxGnP(JaXjWzfA4kialTzw0Ni_0l`; zRnxVQ3pC_}`@9AI#ZYo|*LmZD0)OLX-0 zCI)%k?6@)n)_Q?h zsfcud$VmY2le_WRIH0t*AYqq<2Asrqz^*VVv!<<-DU z@c{_t>irIeZkn0m^>AjU*oK&~9Oa2&$g(%voo|orzcbIYK6VnR%192oyadM$A(Hpa z)@=tV1x#jpsfZ*WE#%&Q2o#BlU}7|N(!0(E(g%aOfTB!(XT~lGGR$aI7wON0o8)K7 zrP<&ZYgJhJSi-pi`At~GhN*Q&{F)^wVhY_HQ}vJ>L6xvrnB-$i7)=Z}76!<6u$l7m zcW!_FkA<5rvn%!TT9U#k$IMElffBF<27Sm$FYA^fVG?dSBMHQSD`s+I`orBt+hGg3 zC8&;T3=GsuF*!3gM624aNb~7ZJqWk?+RApgu_0)nE3&n8o@KeK!+O6ArLAiX;yOSV z-Xl!|8Jq^Crq=%+WC`?&A45X+M)m%{GYb#U>jhDM`q_JEyqBOD;AZ%0jF+1w4`rGh zx_n6oI6nift9dtk47oAD*s%^Ms`3S$+?}4KVS8v}e`gk6CdeSO-KB@DTD2-M410lW z@9+fY!v=mCNa}T2Pep-rC8i^2FoUoZ{y+-EJy=JgRwQu1CPF4dy%cR8vvsl^W-RQE zb*GF#WA<}Ki55*+wd_h|kZDHSkleG66tSDp9N70cYc``*PzneRmq9kKdz8-=9zA>EIDVS}Lq1j=$cqx1x zBv*7xH%K*m+%BO?7v8_?TuJNsA9)nnC6U)!1Uz$Anb=^D`ZT*~=>LSP-r$?2p%quQ zzfU$gaLyd;XpPw7N~09pDUw7*e8Mev>yYVa+~C3;>8d#og1dou>SRD6cp!MCMYw)Z zw>>LcEk}YKI4qyG-#IIAP#41wzhqgy=&;6wp&&7tTj!4fx*OsKb;VDCV+_J8$4Lzj zukY}QA^E;nJlBOZsZ*7?s!rEzdWX+@4{hc=JZm7?LBwQ5p4Fj^?y-Z&n{R&h^@)V= z95^z}5W=gGG%i8MqkiTsNU8>!O(mpE!rwR-8FvnF3Nk39FTO<&I^7f(LWNo5l4f8W zKC8G2YinD%rlSw|xNv2J*XSME=<3mg4Qrlv$u^u(jFjOI!%xdf!-_plk(-{Y1?N}9 zabgwSGmrCczydpg9QFMo*uM4gW`n|h$ju(N+X*{U@?miE7E4m-rwO1Oc-_K46piB5 zuv?C?*}Kkqx;HE_3`;$Z1;m7`^|~&G8>lo}88qN}k^WSwo1Y4SIZd7zXu(zZf46-T zhS_81ONZ8Oz=4jLv2*T6(hkd=X@(78W1%@JqhKOawCioYMx#j$sN>_yFOg|gu~>7X zspFS%Gw6!hI{KIjL!@8+v8Sn?$eXw4CNT4xA1#jMun7$JeSh+GOAR-NbzT|t+Q3>E z_?iZtFi9^JR z%lYM*M>YZyEAUWN^0MeTrf^C0+cP(l40cg_2VU+}m<3n%QVNhQ&Z8o7RC-vty3&wtO)v{cz)5VOIW&&&|0_8sQOIY{u$n#iv4 z1%g;S8Q^~ESI^n$UV!PZm|TQ6@bZzDG)W@l%8phPhk_8}+#Dd%0q+vC13KoOH>%gS z!{9Vaj1}s*iUm^akbIGjPV#99x)+uSS~f7+C7?`g>bzGT$1$dh8-FQ-8@n?+Hv|K$ z`JzmCw3{v!mT)tt8#j%SLheBLho1Q&6tr(pHYo1|*2ywNnIW23MX|@#WyrLQelK9a zHe>coD>06n&+(WpYu%RS(IAHF1z&pf}Eo1@C}ty;VzWXS0Y z?mix@+2#zmX*%R*7VlBp!D878`y)(^O%6Zae=K>r2{gCs$&Bh>n5kgoEQ%Bu^Ve@LT<>CsNr-U-SdBJgFSftI%KMo*_x9hz zpZ5^n`5-+6V#~h*<{^)BEZ|rG%xKB-%luv3IF6=jo{sM2CUDlMQ$=Mnv`VbSwR8N% zG^VToGCZl-9uH6T$Pys_-%(o$K@Ddp!c)O2u=tffx2)lWwkfL>uGogqRRD$P|zuRIrY zvvTPwGUSA;M8+-dD{irF8W1jYexKVNTu16;Lr$$Q#~pV=>IMyZ*lpZW;c1x$sI2Dd z`8B*OZa3W+x)GdjDfgoBYEIzTuIHFi88WFbT!_Vu1|ip8 z8CH{^{;>wHXRA6vega~KkKECp#W-ZX=pw0;URiE`(6d4M2-{*yP>GUYaoiGW*2IFI>u?Ln!Ol*MpL+i7D{v&CT2U3j0 ziCSk5Io)+B6148qE=i>Kh1AG4@i)?V@ji9l^yDRF+k8>w^A)0`|=C z4lKR*hcJ^x`CeH;CMkVk_J8Wl4CDKh0ycyys0d@pqroFd;_6^CRf7^E^&q!*$g2T3 zYOrVs3y7iw$7gGdLPl6*GUSv4n@06gV<}=A-$+ZALw7-aVnvXNq`3){)OE9>r4Z%Y z>1iP3Z6&(-P~j-XZSijZgR=E56A9vs%Y2TQm7aY#e$)E6AkD+FgxX>AWf+R-yXZ=O ztip(bOG1eFLq29#aJ_VcJRKBT_t01nf6(iUDu>R7M9e0j-Nfi=mu#1$3rbowxS!OZ zyd>-=xM|Q892r<3)=Lt2%^=OHg}Q-K&xs8DQ$u4@Bs|so_%AS9n$fn!e-<5k%~D{5 zjrZ%gGckx5jST5-r4$JiiG>1JC^9)n(w9QO?yz`=N9+Pkb--s|ySMyvNsp^d0b<1v zT8|+sI6{~u^|e~yJEI0@*TgsW~mAN zg4hrSwdh)5JNmKwELYgf``k;byftI}tjRVWI6T&Y#e5(q*brJG>W~*g%>SxlwMUCr zw_ld1gRF;o#tul_YXP~m-}J3$`~L0azp-Nkz&famYE$5%{XjGO52^0YdnywyK#X&U zA*T(Ch%jHJUOM1b3CoVeusq+jUUpdRGLH(o(Pr+))_0g^k&-v?Sbi*i{V! zx%O$gG6*;D?ggIb-{GBd?^F-S(+s6egHGuojZ21t<9NoS9*D#bI$^YauCJ!4m`J%n9VJ_ zDMcnlc2W@+$Qeo1!i%apsaAPl?p=PJq!Xkyu|n>EtY3`#$Zf)d;vQc1vfi)(TIZh` za9vbO>t;XZ*$ptm17xKK@)Tm8Ja4Z9X5_qgb9?nf_5>XmKMZ?<$sv8>3&J99ilm3m z0tU+_buEo$M{T5sm+4$jKL(1q3{i~hNzQ5i6iJRBwzfZVVC2VRcI*Y1k^S;{l@Bd_ z5;pFu1IPCnlCih=#fD_i8G**!C7|hYyBSgqaoEbB3{k#dz;%*yn%H7v^ABv+m6gx1PM8?B2Y?CYq=0+~M($<$Pj@E7RpH*|1>SoJ&IO(u=~~{`*7f z`TL=SqH__X!J4k+(zm>u)OUgl#C0|+XMZL1^X(r`gj?2u7by(fvRFfpxvhEPdiOSdx3m{L(MZ97+pz$9 zZpe*3R#vZF86LOW`}X(9CB&9Y3Y7Ey|M4W5MYRKRY6|+d+qWQ z%OA*^)YbG_L8<>C|H82SoI@_fVWsp2d7k@zPNX1DRW)a~7>jEf6`hMG0}t$gFd37K z5ns$xjmOPquW6?M zEAk67AK6o;hv}Pa9VVEbw%>JTOsw6#H}ktDOk`36uagVxa#S7IpnPJ6mg|(FnIcV8 zM6WznaggLg?oVs*YF?V;Ea_Mj!SO}zQD2>86WFK*&dIPIk_}Zj z7rpBE#U87jwaRjnLNa6avI<_R>e2G^ zUmbKxRa{V|s?L0M@7Mpb?zOQEnsv}KU-Q`DY)s575p$TBxJA%ig@r40cuNlDf zl%kCyH>ij<{sqy7@W(u;2$lDdT2b`U`;uz`B_Vs25Z} zX4bnvNl%Gqi;E`72Ljinr6QXQYhdke9ayWObUR#eg zHBA&a&w9Qgurk(iZSAn3Ifm8pX<7DM)J?_RUd>(S{CQD=CJqesKKEWO5-s1KT?$&N zmuHsBC(V$supx#Kf7TG$@*XQ2GUex0AFrE8(tray$qY#Y)qG@8*y3`6Qv&nP+G#y6 z_sq!f-yR+_?E$c(UsCnEV^Dw6U<5ii35jjubG*X3zPyxN>5V}du0hcOq$=# z)yuFj{GL2bk|aWX+jSw;Jl*^f&kN$~i`NS}bd8Y zuxAK6()T#J`N!wB2oJk9$u>yKfRPK2cF+2F5sDHvOY+2bo$<=fIs3TA-jSOYMZM7- z?glEI&q5!$j+<(f71Tys+KhQ^d=N}O{^wUiO@I+I?o1qJ4fU2exKyPt` ziZ~h2#y6@m=;%J6Z|{{?(@koeoASLkaiT%~wn^Q_UpuXcbdVfyZH)r9G1ID{s3*xa zb>^Y)4q)co=~-;hFKpxQ0_3Dx$1(D*=O{*-3R zPc&0a5Q<(GcoK+FnGM(|0T>7Ne0Q2bD4tSmrpN{=0vY%p!q5U0f1a{5EY@S(X2JYT zvOTi~5p4UkMf(QJa*B$;_$nqXA&-+5h)J&d z?hDBjX$D+tL96sg_)-5Nq5);&`CY+roVcZRZbMEeJ!d2iu9mHxkAxLkr4~qshkLpl z+6RlIAt&Ujz>M2MPH7l&L9K_#S{scdD21FH)faTN1lb9|@XW6V_4Wgn(q+3{m%%LO!>0IdT%Jq%0h&F-JbZZFg-uW@PsBz+6B=3soT+rU3C znGUjHbgc!^_@FTUr+#zHZJ#Cer^5zh88}I?K(s^tXleJ%Z8LQ)k%CPDyTuhQ>pAgW zNsD6^8P6wmz^xr@Gz#M%I`&VpvKUlvb-w6z(**Qh-jUnnv;$8-on{kI6QyXR$R%_# zl6>A*ZII)l|l$p3OlcZ`6Ixu&@{Rn+O42wfi`sW&_ zVp*i1TL_BVO3aHp;EhvFi|`sp4+MnC^2(sSfyjO`+!mx3uoukTD z?F&iZ)X~^kz91iH@G)U`t9!>>bc*jfH!0XTwrrsF%sI!Jn=BB<3R+Wc7i`^NS=ZsP zR)?YBWb3QQeFXmN9rx29fuKPWO5LQ!>P^FA9VA$crHz`hg2dE$+rCJ*Oe}ao{1XP% z%TokP0nvi=avQ;M1w}JylMHZVq*R(@tpw{(WU?AILWnipWc^tF-tSGb(BYrDNy!~{ zu@whi7;l(5l3(o$O3_b|2UJAH^ioL6&xGI{a4>?%m%Jb3FL!zl^EH-+)_7|UyJ3i~ zTUr*X!9=^#ur~gFMJc^DV8}O@P9X3oGD4muS#y9SHw#VT@Co4LkD zLr#_cyA--viEiEUk0dZ7`oYviYfAj~$_8cS%3@(}*lm|;3I2ah+38{sZpjxN@~U=C z4>atOr%OY*5w+VfZ6=ydP?KE=C%+)BS6hL3R2V$&T z+{M$;C~ufA(9}y0l3sbTe88UF-iG-h_3jBgJ#bJ&F4EHv zNwee&ZV$XCX@y;?xNP#Kt@+}zY{P1{aNIBTzuo$A;nz%nSzgZXA~g=|G~X~oay_Mh zXzy7nq7ZT_2j|7Zu44tn|G#j(?wc%c<0o>fB?ivNQ{waza`ZH-Scfb{e zo%TT>x02h&Pnci6q*N9ckjz~-D;4-wAhmaXiRiIcFBgEn5CgkOKB@lu1G|H7k*$8a zd~!eos!6R6UB|mdIzU$yWPWEGl|62HX3!_amE04mmYH>~1Fqv)!Px}tW0?|`K*#2+ zz)t=3^Upqf-q+|@d3K8Ai6LJOiM@~pJ0IjJb##j4PH-vkLhkp})00U6$POCg7~^q1 z!U`HwZhO8n*K#4n#%p$5LrNHyU5FkKHVp0d1qXSj=T6Vku=3DeF`VFk<1rHB9jIiA zkYE4)zchdP>F<8_$KQx!D8(9zL^4>5ael}+9(!YZ`guzZFE)^H;6N_}yH}M&=idxG z)mJ#0FF5H-2c6pK^rbC9I=|8|EOy-%a?oF6B&JOcsg^}6kf>Oz>|2y5(%VaY$OTa5Zk%W;Nr<2dth1jwd-IKa$0Kb>AsJZve z;TX)n`MyQYzA!PE+h$9kW=hdSkwz+FHKalfxL#eRSqnjm0oMyNH5UU0(Ze739A0jg z-w=1pH6^0dIk>64ZBaLU-Ma?LLz3lZUq1Wt0a4XrJ-e&1S}z@aVJ57mpphnrJ}%zj+{~$$j%|vqM%Rd0$OebZM$>fX)xZ7jsAau08;8Y# zjSfR4^_Sx8n%!t8HJZbt4VmciVpKn7y&`oTePPLDeob%|yOyHWiS=qviWM za6aW_v-##Ur8r5E<5a|&uiX$wQtk4K2L6pTyuE>UX6&K2xKsvQ7H0T$ERW+{3wTJ* zlD>s&A&0UKbQ3;PR?FJ>dKsRjNcw}T=~JM92MXw_CP~b+TyB~q)455F6HYg6Y@JO< z1|D0Ymu0C=@EX)BBLh!(4LU^zri<|Jlbn?kN*lKtUgMfk#zYZg17>(7QtJPhh_CCo zUOcENjNDjCq7;b~iK8M4#7F&mxe2^9zico_u+f7X7PnwQcPn(Vi=M5Wc_!Z;t@pe= z5M;728y>{LS5OKd+%KadwodOB-jGKJYJKXZnslEY*f!kY zyJPN*QpI!G0!ifCF#48 z*l$>>`W!O*Rz8u#F|s_~7se`*Lfi4tHs^*X^Fxm`<0S&qn4^S%vU{-|BB8xe=)xjlV@rx%&#p)CHkRucD| zdAfyWxYkmNbc&=<5jEZ=2I+(xdfUtxb^Of7ye(usr<9Ia^wBFfm*0ZT;5g1*&s*Xy zekt7pGtVTglh~tqRPfpSA~TwA9{o<2rNrP15;kEFAHblHEWYP4UGo zm8mho<81gG9_ex5?6w56^xCMYiXn=*4ud zPo$txI&E@ZNs>4e?i@Bf3(l~bl>-ee40Vr|4;j2|u1Rb7p z*gJ0Nb>Aj+w6vbz#?KVwam&MZsjmB0ho2(*Who@>m6CtDzdYuZm~ZzkKSgq0S-axT z$=~_&rGIYt_JF(#cml8cCb$nmIB#d*?S-ev9$L>YUxH=R7rjn!FVQI3-OJ6I6C0ks z6v%n``#qZ_1Fpv3ywt05Rt{Yq-njs1FJIEnP7e7*R?8Wp@44L%Lz2u?j(#@&2Ia7G zT;!Y-tqF|co$#uY zT@+!TwR9=H&To&~kgrW#BldP|RBy=aqAg6@Pwt!v4<24z`DAY+#SU!8&Y4YIm6YNr zMGjLDAJ07HojL<`=Q2aE1^_7sb@MN~Ul3mksv=c$T7n=>aKtZJo*Qxy7#e`cgT5#K zT%wzQNw}MfxATDbzi(kF)GkF!wNQH9z&REW1M1zCbj{papio6~(?(;b@oo)4wwn9w z00~9g5tdBF-wVF2nzmWm0P}~L7g08Ji^ws^j{m*TIJPcd7znCAlRia-l|rebl*1f zy10BvG_RIE&Zx2+; zZl4zgTbf7?)5)!*&nw$Kp(F}RXy1o&0xXTn1BwT5*6U@lOh;0Z}vg>5Sg4e$m8X=7j%H%dv3wQmtzCB z&P)h~8dj`VyDnZMDw%f&B8zphQsL*YXgW4$r|8(M8`27wc%RR`kfEqLAU5#Pyd(35 zoIX(X%CE_<2iHqWotx!l()H?iAFc9+^bi=ecpn4Z5qjU5@JW3OkIgEi@XrL$IsU0= z6fiz@Y5b_Xb?i?+P?%g~*Dw06lI;$>#yV)`^y(-D@Umu65$M{s5~!`_L`$!5;sW-1 zK%TE`wTD)@Z*kwkJkBX{ntVaEewt1W`};Bk_IUo|(YFTJh>9F!omT zKK>@`j7~=Xv1yikoD4*k^-`pvY~w$cN4ezyFG~VXtI{Nf4ayEXUk=hsP<DA6UoX8#PS4MQ3^=1(7$?e6<@p*m3}72wt1RMnhZ^tU=0Pgwp{yN9RB`s`mFEle zhNqCzFj;|3Fnw%+EugcgP71(GBt4MR6RBPPCVs40M~QN`vzW(9WC|pM+#Ad;;rOjJ`7Z z&ErI;b;y|*hOPUI@w8%x;iPHjUa~Bpby%n{l*A&36x2V?KqK~9V!@onGyJ8nG^rjHJgV(gmzrM3| zMg8Pag~rZ-;Wkqr(F|8i z0eMm%sT5|0w0q>xNY9weYlAg3md(K}V`gl!`~y&YglvskPQRqc{hcTOjSFWJLy$S1 zn9jxnsPfJzwa|p14e2+3MPi?`EKuOa8c}nwlTxHnWIGj63w$q4>O8@E@`Y=|tSoM$ z0_763Nwl;ds2PyAEbW!!uWouptK7b*PBN|`9mnftje>D}#&{lKo}%8V`j6UgnDBAx z7uP-^Cter}(`JU821-#+kqgM^+AJxXwUMKVBAwiLF6M;df+sNyS=W$UZ7;V)h{e>n z0K(-u)=m#ErjR_QUOM1*GXzWrZki9cbl0OTsfU$f7qy3EUERy#!Fey)%b?nBZ3 z`S<6}v{cV|LBhZcN^C~7=xI%n$NOC4N*g99N!1)o3ey~gj!s&EyKR+ z|6R-A3PTK|11eC;d{6L@K_(lB4Kqc3oFpF&l1>hPy)x)rx>$&8$*~?9oXV>_yXj8u z4rMQ_UrOnUpycT+zE5obX)NEd#kY*@r&-^g+2D_swCk`nEcZMb%V(Kpr#A|I@oTcf zfwusl7;=QCQb;NCD3VJ>Ah~=Br`*3A*3IafptMniOX}Q*Jr&vn zm5+U7P=?z{AAxuKIQT)ZrVS<>qiNB)4M2{Sov=JH93!l)$0nPWtZ+QFJhK0sC3Bg> z#^M;b$goawgL{s+Sz?rP%%CHeo)BwQLr#c`9m)aMfu;NX+v$(whU#l{RE&hrcEiS( z<27ECeA$D2j@Y3jJ9cHNWyJOc@l6=wyVTz zs2Wc-I?XZp9?@^l+)OeY*!KWYy%E00UP=MQvw2iR@3KRZOZ2umd+0{#Rl(Y6=I1ff zFwY;>rPSl)b+UWn6bd=UwaUATFeMyQC!5q*?U*6HOgdeg)O$VZ`Svjy)-V0nP(MuD;vQYVVZ$3jh$a=^`4=$7Z+ zFTtbL>T|O-4ZJc*5_hkx1EL$q*PKYYg=a}o=%B1xg_1T%6Mo!({nAz;1VJi;HoIe* z1$L`dCW0_PFSiP|o9+vdJERaYbSDl%8vxE6e#W!WQN-fvAJ(kZS+c6K@l71KEP#O= zHjk4FrThE*(|tajQ8?>>cs1vY>YykKJeG-%GK&{)Xlf-Y*mgVN_R}`;qZgT`qMv`b zxQLWE@Kkis%v97+ib{$c1%k4`jVo@5b(|(OG6dy$J|u%qd7jZwd30fyXVobHoAcKf5ga2{+0Aj|(n zQYSq&>yRN8Q>(lQ*KL6zD^W3;Jz%~zVN9^q00>*Yz1)(m>}fr7$4$v%;v~aqf zZlD6IRbn6?SHp(AygTkj)lk#Mo-Lp-)~JmODr0qOgZs>WOo!{tInTS7o2|MhH^han za(i66=vc*3RUG6R7{bJnf^~vh(mV1Yrvw6p*TAOPpw!%wcF0eL?E_Iy%+8JsOacld zyrsbnVw9V0;@nl(9xvlU$I~Xn26!;yxSqK1|`aW8YB71(THV|#pikhJ4=X0+fI$as(n&F=(?*{fh%)GSyti*yz^$A`^IT&@! zCw8&-gEv?GYB)FLjZb|@hXXHrqRhx*?o*0;6#0~jI2m?SWtd8ED;hOq@C2z6>ZO=@ zk}v4w#`tKm32tW}_wSY#d&C53dZewwPPqpE!ing(zveE;-RIC6tTRKWrJRoO=?!cn zwP7d25HH5(VCL!qjYYS6ZmW9bahw~VxrR(L!@7?-pzc^23gs@H^5Z<@aL@tPniQ*Q zY19`^06v>KJ}^F91o6jJ*vX*NRku#(RUd-^tCwc0^dOLP&vl*q$*^>oBaNT93biDU z`WHcrW}o+}7lR&V%AW`I&;2gqUC%N}_rfmK@K3DwsN-)`m#R~|>XmISSt5+JbgBAX zAM(|a+Bbeijytf=bJHx;dx=s&?DrfM@fa9ZOJ%ngl!kT4v$=bn z+aZe&gofh5iLH0nhknk1abF{9<_POj)P+>3>;NEfD zrRox|7F_W;;0%vnBl&{uoYjI)rJr-wiLlL(6LM3_AME7Y3Duhk$ z9{S_uH^lF@%d0^>WliAU+T};#pp8G=rVW<$H$977vgtf|VtC>5L0P{?FBf+*Di=Hm zOAWZ}Tj6qDe1!XGZj|6OeQeeQjhX#G#|C0gw{=;tFiI8H9N&bd<9CdqeMb;19VKmI>K1led!KZ-4T8WrrWl>Dq| zQ=rKd-FoxRev;?F^INT%aVe)1kfvTrMI0pYK9G~{J^V9Sj%T?()pV=8iL=vx(5Y2^ zkbEG(pTkLQqPAnNj$+Y(MEN;QmeFMCI19*HBtC5K40_pp=mGTX7^3cHK0k| zA#V{@EUR!qm1<3wECD8ucA%QErNLl)vm z@)}?_XTTK;F-pU95Ud5V6wGVWV1@D}x}P9r5YFjH8`J<_HBAGjjXplNMTpb>UC?#? zEL2lqFA68C-~TAnxbc#GMuYkNkFqv#+$hH1+`?Nalc)07?ePW3S8~`oDYC=NQ;DM# zn<%oLiohL&IF8QcL-%%>)+aV(vQ2~uU}GcIK1dv^S5`=!a_Zgx_&d)8i214CJxyX9 z*mVW^xe*EB+bG3WiX>1G(MwNw>7YEdL0$dkO<40bs&iFIqFTCEb|z#KXWN`f2FTdJ z7`9zw4W(`GvjPV7%ImAYXIW6>uw`cqB|=8glH}>{T_$l*Vv9gJBGX0nZ!M#tZ2oorXda z<5{CumZ+Zp7${zF9fOVSjWeq-HEd}{{KYS zXnm~0?Y3-HE6L+%knmv02}^@^sxC}IO55~H zT_;;!m~?@CW!FBdFsx zO7ZA~AZA*Vx{ZHPTE-pV#_{&hx#DIigv;s+iBvz3bV&L;a zy&!u+oGq%&%!w^wY|$4ps<-SG%oLfXpe+8jRuad~jdfrM0lU+P-90U(NT)~&Ojy8O zjERbS=(M zvM+Uiw90!f_0k<4u?w((xr@H%63dU}BauFmVdJ`5tGu9kB+eIXa=Ql$r4soW)h4%- zDh$P}7UYXCKc<_u6}rZOA2aLo)QOES^|K?+)wBseegEQfmPAd1P}Rt9<`zm3Pm#@3 z1f~u{^3ZP3H`e5mVqqKo$h8jS2cXn))ue8Duv$>s=sj824YMI1Ek@z6?nyFFd40AC z6sp(_Rb=N2V^lyke#ER(L@B^4*Fp3iw779Ge89U`j%3FWspr?y$i0JClSJi9hMchS z8HC}v@jeCq@IAP@2U(IZC2OED;4+O>s z$wv?##JCESW(m86=^=woDdAh7RU3Un9v2kFtED^W8+>d5`^X%~N`NfLJt0SZai2LK z(3%Syi&$;Zq>c>4E$KAL2Lfmbf7kG`xO!QcFXks0pTEO9<=&}o@xo8ZA|)dHc`Fn@ zEyIs3RBeI6Cvbe$0ZVkg>+JR4G83*cX8ra*NisWJIdBGBsTrv+EWhl z&Op+0#Lq-Dm@5YvkB2fnjZ~A{-0EoE{Cq*QRO?pn-plQl9#!SLKmmJ)JTfp}obS^C zt3G?xWKA3=ECFC;^UL>o-&%WuuKM<;ePR=|zQ_nvs zJ4-s0dKpSBYnAVNU2@$wqC6PsC(+q~<|tuU4yw(WJdDJkG#O;l-SpMZ9&kgYF)TjH z6Aa(N0wtTcNE3(9==Q}*q%bzCnmzMp-!hq+fVB&LKu$epnN1yL2IVrP_>dxXRK!j3 zb`a#=Ej|=Dz^$cMX46rmYWB)Or;FN`_zY;!E5O?ian9)H=Fdc8Oaxd(f$F zagTJ+>DxRC6>!M#tuhSMQGuRNo=MBjsn@}~IyqDyqMyq3PtubF#f@n-86WA=j2=%BKG zKbzTK@_prj>Dq*n@s9EqwvjpQexgnFcEsb)NTJa|jlMc=w6sK~RYeBcC;yT)W=1n2wuOtKw|8GVTxwbU%zp!Tk!spG~2KVfVcn)G1Wvs}Fv7YkBuKTD}Ha z-O3)`5g)c|8}skA!ieB5o1)w(R!8I^h+%$|hP2Rt*SScv#h`4H~h7$G2X6{g$U`;%WHV zp;aW+fivjJ%%+k&N&y+$yQv7=BQ`i1&610dA_@E@szX2PhQxzwBw3F6+U=fsoJ?LD zUr#>}BfAL_o*a2;$O-ZofStb?7!+)>rePjDwpcNrL74lU=p9yAnfk}O-~Ii+nXt0$ zzh8TmTy|ir+&9BYC#8T0-c2eZIYbXzKpAwNZ&z>}2baMWp)J4}+9lg1De@>)#fAgh z$xU$~jnTC>I#Mv;nk>(8$CTS4C&W^+yg;x^fi)CWuGq=~Rw>w(-`Gla5CJcpa}&cpUa8@i#>=&WhRjf_v~w zAP);B20znTo*2-~VaNWA8IG)gHsySCd&)$1D;+rA$Dr$t2{5Qh)&*8e4}`Ti&}H1u zgvhUyL7!8$@$ZM-cb&`}4t7Xl9(h~esebo+PL|b)4(rb_R3fJNWz)Hk=Co->ynI9W zpi>8(9~$F3==6~2h4?c`P$x@=>*=DS{^{Wvekqb1*0KC?y?1tlZZr_%`XH7Mu`(aj z;(*o1r<&%e=yic7$&k}1p0NW5RG?mUL_j5;Qf#Kk2B;r`4FPmjpy0XP^Zp`?o!EaB zG|EU>gOX7mv->4xl+4_+EWwgghfRuu15YRn6g){jxPeqjw?i(aj=sGR-Qgjp-Bg7# zH$(?A44TPBjx~@m`5ShJ$;b~lQ{n1wxnu3HvkAi7my}+z=PTI{UvL$a3kkpwI!;NI=)rF6<3%g<#!<&`)7+X!wK9a`K?osNM5W_$lu* zbcZaLdm5Hj_~C_|X8F}+S|u=yx~w#O91F0qKni;Y^{FRDV&U1x%*JJuxrM5K_g`~} zX$so!lzE=4bKs>?j@fjxol+!GBoSntgW~~B1@s^ki}Xk{fLUEfw~Dvm`1MRD?@=eHCab^4aWTOKkV~-g!(kR7tiX=e=2#`!; zNzh25X&ppeFm3;1*Ty9s@`~A!f@dyNZ9OV%z+wG%-o$@PR^XU=CE}lFS#rFw@lzbw zr)6M-8{k#|qiKV?PEsv9>Wc*olM0(J2iO__ec5{Qj^6a^NSJ1_7i#_68nW7f?FH~0 zj99oNQi?c=Y@#Bv4dP^4C9-Cqaw;;{Cr@YL6xi%_TQkb`*MNHZpTo{f@Y}=PUNvE& zXfEd-sb;rZ?7+*J>t+b5qZH>Ua)yc+ptG0NsFFhtIB$lG%Np;;Ugh5HkOP_FTQ5Z> z@_f`KK0Vbzyk3)cPtt!^}M~XiqyNKLY`5;IZvF()%xu7 zuW(7@tzG(%o2N?iNQ8LXd1bXk&)=>%!NdEwJCH9(V9qhI1A6L z0^n0_D!U6$!)xp-yH4l+*Ylq2t*7-8eW>y82KDLhfBr`FiVfdP_#rA_X3&Y@8>T0c zyW~^9{@~W&L8nHKIte7^LTIhR%g%teI+JX-DCBi?E!6J-Moj6I;^Xu-+| zQDmd4L~?eU!J8R{+vg}z4_cY(W z>2$>l(1%u;8UTq9l8QN9{04Q@S0BDw{o2WYySltl-R{{C+N4HCm<;H+Y{9nq)=#K5 zKp-2FWHlBOVqUJYQ$LASwDFteN!(lzynMvp2GzL5z~bMe{(#_Fsb{Vt z!5NeJFA}|^Szg1_ss?4_oKCEMvxbJ_8+LY|F`m}!49~y&RBoGTGCcZa!8b^{1AD~f zW~QWoQoxcrhl+?4;LhHTxwU?ffn6sng;dDMnd$OP@-nHW#PfnUhdw-~8&tqF=L0d0 znXIPLvX;(G9o%tNGTw3KbMNgbiPHhRlppArO=h4V#w9s;lF9cb#b?! zfgx%`cp*23J|x?vz^%l50h9;f6ooY%hrA5=QoG$sIPIRF08_3pGQRywgxFH ze)(HB#N}}H1c?jkAveX?3j4BtJu}w8+5|mcB5JIUv-*C@v@7#0Cn5%W--KaurH)1! z;~}RGIi?%7lATZnS`bnhR1i|{UPabV>V_PvmumGZnRr81_TF*byPy8(`cHptvMd3= znK6TOva>7>JP}1t9XS#8QHoxQ+@&Hi=#0Rv(<_Ci!!ZUIC8(P7AXpPAC=45b62fZP zW)JjlT7*Z!b4BYQANQz##WKj2ggge#N_=c(vi$zs8cx1%r4S0GTZHQuRk)lbCq*!+ z3G;oS&h@mcYR)~Lj_!jm5Cn}O;@8WK48(@rLbyFB>);#@h!f$*rp_G-jw1NE-E>Qk z=9V;%(;GbCs#VqnwFIr5w_erZ4LLL&-ut+pdJPFR-JDMOkkfTB)H{Z){D8AHXeDl( zcMGE+eHf<8+>itwejs{j8r}Ryt_`5c`%6~epD&G%-H%TEX#S6$cS*V^K>{qC*NBpQ zcDpr8nmD$bN?6&av4@ydcQ$Lx&O~1A&-%M1{eqbAe*3eZNBs7uzh3!AF|BYSl;c`b zP*WahZ=R4W9U?ePK93X0q^^Bb&&ZK9heW-I^<|{OrjvByVyo+ zcWIfivBO{qY=)B2J7UJpY4O>=c$j9a=9jL#O_CfqFt^_f6WNp^iy~So0!JGwH}taY zFKd;Uf)6qe$?|&Ve4jQxy1eZ$H(YlzlOnqX;+~Ad^9PAha^t4x zYJRVqVS_Joo*a zvtF6vesLdxdLK2mJI0S98PIOz-PF1LcI8o?GH$4VQF($lI~*`uk5QmPaTmUJX(IVh z4m>9@U>RXy6AyOa_#8u^C?$N`oEYDOq?YcZ zb(|fD+P)+=?ED7=ykULBh<)@#WEbrj#_Jv*P z63;e%bwJlkl|fjTdc+R;W~^|*Y-+|c^v`^T*}`1?a9!vElk2*oHE%P?XBU%mU~6*P z%+WnUDS&L@02R^3@A`p}d;`gP^b)KI>f&#ZR&$%w*Tee>R7lWZs1gV}+W1!kn-zHwO8ba&OWL3=a&K02z}9raE0J%U`P!*h8r4;#*k&yL zFt9d)l?iaqL<*~$DQlOWw=6<@uFMTXso_n}LP>dOfq!XOWl)k&G4CO_n5Q`zPzXr5 z>yaj^2r^37$$2GFwqfZ`#duOMp^q#&HeED zyyrbkQZFs!R0dGh#B6H>ExowVUmx* ze%x_C0tzs84XarTmC?Nqwg9lcJu9e8Nm;#aqh)~p1zF@Ugzi5B3Dqc&P(5y71S<=? zIS0~gsYJI^-sM!1YnG@(cFMbrU*uj9ROG%I*fiP*uAIB+{hq4@*X+IzpQaG*rWP1D{i0m{xfoJQWne>s0H26gOFZPh1YH zl9{|~&~AV(A{RxGAfCBVuu*O2xog}&v1^QW%9P!e$N#UTtLm^pHU_t}$FEt^=NCH< z#Od+{Fry%OSC90P5J>t-vt*;%uJg*%Ur#ne{q)oJyTWd!IXwLr%m3Nrs;>XfV-Lvr z(de-qIP8AM%zbU46jv#7g^I{g?ey%TaWlAH+Qv6l8zNI+D{vm@W&z2*5*cbyN3Aq0 z6QZ}tP_~f;Hbb*5tRl1xPA|LPpKZK!95gdER~5a6jpH^xiilwBHZlCDDuFlTgsN3J zswPEsz`77@l`l&4(9t;%?C;_OBOyrI86&-V8uuL!%snvo`0V4eA*&JtImvQltc&;Q z@l2i$l5S&VTC&01=pN2^uRa^E%$R#O=;41(G2u(KyX-IIu@iISQM_gcMjnV~j#xOw zQi_cfSw}_aAkisb0EE~WqRR-h^Mu1RE@LBw=}G%uKRiqI14}9|HXe%uC&4k0d?5t_ z=I+-^2iz)wm~t!izXOS}5n*Q^gaKIp^7-$@&pV$nl5_nX6-@8)yXhvd>eW!)r32c> zB%f6|m~A3W^TYY(eeUDn2f>px{#FdQ#caSZLN>n4&i-c{iE(}2;gQLS;XU^4rjNO2 ziHh8dJsQ+|e^C2_c&KB~<=%!mQk1`p_lf4*^hhNuCkY)^2i2Zn#KtgS>W&pWsF&}m z!oOnjFPgu3JC___w}Iop^9cx&jqoo{P>LFgR8kRF=!f9Fwzw3BfL*T0- zs|HeUw}GdM%q~Mt?VdF=uL*N#^qfm&{gMeUvDn+BC-58NfC*04UGMopL5ig&^9$mZ zFi1<|Vx*2A4ZD3U!frQ1cosdCq)E_WupAOAL1For-Or2OmB)Y2WLrdDznenp*x42b z-W>ka%(mR56fG3FimPTMR=Fxpk@V1AZ{7g5z-+oqlri_7ctf}z2-SPTw33fF=Y=q3 zy^i^h)pS0$nqDioA#VfD$`r|C-)jLU{f$ykN!)%?CrtDhZZjY+&?;lzo_0@Rl z)=%}5zdvYr(^d?zH6vp4Hmv-asc+TOiO)M>%4(0b)8f7KbUAe>uuc}`7Bj79QIq<% z^v25_^4kkKIBTbsagX}POsjBNJFVXpkDAn<@Z0RZivvSz1xQBulG_{;Pj$!&Bvi$s z8_kwkOAZ^&XUIdsB{Rr3Bb_xanl+dp+soew0+vYVmM^N5>Lshy=Vl{A$~jJJFz8{b zdOR`oKjefmLHEP-vMz9Edu97DK8x8Xzx*9BJKnbLGhSKw>%YW{QYiFmS`omb<)D<)VVjaV_2QVL*nDxxCd zV5hJeRxlZ&gR;A{rj4IUY6G&c_LG#Va_K1Nhcb{|o%I1P4wgJe{YxOT4b2n=?fQUf zyER~kbjZmlDzr)u{UF;QuCb81f2@c-+X1y5#@K%Hv)j+RIsH9n6KZO|QvG)lJq^lc zM!xPON|8vBIHMEUEHS>v3{W`G)2E#ElH8CCIyb~d-^6GCt|RYS`#wiLGa1LMzO#<| zmHW=SCxcu}uy|K<`Hy71182H{xY!8ukU}YvDYAu%D1^M$_YG+$cNb-eZ+m41w(&C+ z`&7mR6A0K(&z;;|9BVO))5v6Cb@($_VV0V)rSr`59=OFVyf(LbZZ6%VjuZ8}9t+(B zE8QZ}L8|G!T+BMz?RHnv2*-CPkB^HDJSGz&?BLNne_4_x>EF}I(+(U$VITq=f<(4k zi(*`B>7)K@XSN8haq6Uxeecs-$VqOj>mD~_#Q!d|Dv>4fM&5fEew^PSuL;i!-Os^R z_M0o&BIt>~YCIUW2R8r7jG%A-;k0u8@I3Qg$1$?Uf#;c1X7kKpN^y`P`>BXEfju;e zw=@LR(gmU2AeRj^Z@wj>&AbBAA%kQ`MQ>P!Y5d znM6jt$t>;$VBBe3a)s7;=L=AK(tfrC2X+ELu_Z2^>WdXnrba#7R5sUSWBT5b#*keO zY-6g;Y|MU2v6mu+s3?aE-Tfff(n@smb&$u_D(~e!BoG#9;3UIR3~!?rWo$@k*o~kQ zJfsNhm+YE79ECw*M+V?wr zbVE(B2{c(g+1p651B2$A8E7gg#ZihJrXsdZZ*dvkIgSg8;e978eMy>HW6415B{w z(5T*_RX&srIrYHVL6;t>)+aXPkvqVv(jSDI{qW&RdcZYa-YqZl@1gNlowS-~xZU@X zel~V^$ZM;+!31KU7(eWxOb?OedTD7`uPhxBfHe)uSOq#%s0o5W=k8#=B$L<4tpuI4 zZaE@*(~1FC&461QAIqw{4F3j0+!?#lLaJy)wCBBisExmN+C|Z}nMeK0sckd+T}vTA6Em%v29H&SH9u`vl zyhs6hFSxsfKU-DE?A4$wlSTBIg2vmR4sKJTbF zu;s;u_+5Xu+zQAyd%Zg-MJmR6BUZjYYnM;A@G8BH+n{b#*V0;_I*>y^g|A|d)y~~? z)r?p-RQKBD(_!0a*E9aXtl#T1`m*IMcHdQay~O2@5x@Ff#P9!PV0n$96l*9FIjMU{ zqy6)DglfLB!em4iz4KuksbFVB9N0r`HbeP&N^yoFwN!+0;j=aXGpln{bu`i z(U7t-ur!IA8G?D(&5})STA%&%u^h1kRz`OTvXr8sStqyyc{5;!K$nUM#8}kIfj9-{sdc=hjz{{4^7~G?3zl-!*>B&U9T)s@WI^i8Yb3X3p1-@ z#mQ)3vEtg@sSTp9n((xC>p%Z1sbYsG2VPP`n(l}#?2D8FLOLH{$^}kDpF&+KP^Q2B z;HzD)zVlbD5@{zjr=bk~j4H_o@~S|f3U4;O^3Goyq(_&(^H)Oy4;gs7jlkd-8n9iklALP6Yk1fsGi1@M@rqrxv% zfV6wYK~hvEaPwCNiFB{vf!rAI)H{?vzq>K{{V?MU!mA9f? zO(v!M>)*aaHoY(=1yra;n3Qx%kwU>JM4S*;%tmEItrC5Fm^P|l3mi43fIn)w))V#{ zw5n2J2kc#COd6XhYjlh`uC{6L%af1tEz>O>)qQ8m15FOep1WRf1(+kN>5HNsUd*Ci zC@AV&oZ^*a2ky}^-n+E}Jw`V+zs+t5_o#3AcfCwTkZ%gNclzPLvtUXL{7?9LwP`!k z`vGZ#Y~(oxw=Ro6liry25nO|!4=$TuM(B_{aYUVts3#9peE7*%UiP3ZD!FJ-WogLc zEa~^g5|>RAlc&B5B0$W}$Rm<>SuHP~yqb}2IBG?VqdU@HGJ^Y9S&`S?D)@KKOtTfy z$^WB-RJn1@!p9bN;Sj~_r${~07@IoYD`LQrz-IVP`6kcLr)##xES3>PSUxt~_CGCIfentVU+qPT--+9hY` z#nR)-lgh`xut8N3(-?TcYgE8vg`+2?!qJTp0LXj`I~;Xqp4e0)xwvZ;TGiw_3EG65S%JpSdY|lcJZaSsHy5aPaw<<)nrWheMDUe#01T zvx3C9*JU3IUv}2?YX3X*#kf?@HjwtsSKXKfK@Dtv+%7x*#{2KT@4`ihjS+PS6((C@ z<8Id-Vuy;Y_o-vnaWc0x6m}wCSm21%OF#)(NB75G^Ii_>PsW`UaXUb(Wyqr<7MHdz z=0>*|QWoeKQ;e`e$7{#l&i@~4h015U$lN&nveROD^F+0BuPT{=?jXTRpK94=C|v{r zF<6q+#Pmz@p)Yk)NRB6_=dhd+h>$UNhulrTcnwToqgt3)K*Zh-Yv??7NbEh(NyqI& zXU*H667i$d7QncChtJa`^RI-{|6f zRn#37dt#i#93d0l75ZxYYG;58<69u6b@@g!sX>0+3}y&sab_kPZol3DFjxNnG_Mm(Yoa~OX(VDX%UD7@6Vf=h-*}zWJ-QMXUh`^W2kEk@bUq#kN6{=1{Gb>7eI1x&t1I}XrJ^kq{ zDD7Ibq;Q@YI1jnG?~pn-PSl*YfWmQ#IZBacY{oJXW|@eJp(QIrjy%J-X2JGk+=kw{ zOt24{Oirt`kWPhcOsj7y5Ar&q4Uwl{d1i|h zde}K!38k{PtqbwlOj>U2*VySNXqO-*0#+Ju9S~M`ubt6r_kuCG*RrTe;ma zp!H!CfI{yWBLWzR15@*uhaYdm2b!9}uo4M98Ouk*<)+h{gKK6XSpq)66wG6VSBJs& z%#_3a02`Rv2is<$;5A|f)=rO{V`S-gQeZJw>nSFkB572DDaeVGo4`1sRdh=Spp9P_ zS%Xdh-6Bnn%Zg|9ve^6sY+cLdW!t*nRmbeVY+C^^x9tgQnKJnc)&;OHRbbx^IckqM zBiab!6p>D(3$BOq#1`qbnVs@x@gd%r0vG`3Fos^#t^LSRVGs@0ByixhK=4CG_L4qXoAWRLPKrszcm1AucjN{_SAUk^i9eJL8(_m|Q zQwIj*{$nY)T$J1rT0bWx<~DT5xE6&Un5ISc>1FYA@x zA?v&=q2VJta(9qkUCbGN*MPhWoaTIHn-aGi0=Of9d~9%>_0thRF8t%yf2Ao=;gSNmhn`32z~AfLpQjuRuak4M5;D0uulXZ>Vlmd0KG&coN^ z%$QkUB+DVyuZ$PdW`VH76mx(g4Je9xdeV;9HAZ}p7Ef&Jm_Pr~^yR#x%sJ3&$K%T= z{QrF1T%pd7I5iLCkgNSydG&IWnN#zy3=@i+RN>d(~~N22U882!^#wqREWivjU1v?go$suH3 zO#5@0$PA=h{<KvmbJYMr_mvz<(jfxhDjI1T!y9mo<*3F%yn4NFp z#x2H8786hn#XzA;1(nc4_mV-6F5yGbr^-~&5=$DtP;gOvdQwWffnG9x5pW*&gj`cE zmTG-6C+QSU zG4uyvw8o)Dj)UXP8dVk$v)Xa+cRF&}Z{K=*#ebNw)A{Xhev`Dju{-ja1sZNr%yo(w z&>cyO?FD%+YzpnB`v|0Nz!BLuIYZD5v09yIJyS)}J>S#);L%%&-)Wn7_{ZzNe{|08 zAc`xw=JkFZKNW}px5o983grfkB60b?65b(H%akr3OA;(M{*abjrG=rl(wpIwb2S7vu5>F4($hr03QF$WRLv zpVb*?`fDQf|EIFOp>{y$RQy=M$+2DAXtHmWYy@|xK-Iy&e9eruiir<@Oq#~S01gkL zo}n06kF=VKP;fQQpVaQCu+B2LZ5@Q2yg_1QTI`_A5CbC3;b#aopk46l;3NJHiB3gz@eTY#qF%0J zaI-zwY>%=6E!&5W;(At~r6vV#TJ;;V)%k7fPb0{u?EEA5MLk(z!PozgVql8wqY`qM ztmu624n<+`LP19;RO#h#Qbg;5T7V%2auCYI=CwBHg2nhus}!52HKx13;Bt!S zL$P?7=@w0L@S?~fx+nsphxI`hqRvI$oiXH57qu}2sz4u2PvP93-bs>!o51LlGU_Vz|?x$lz+Q52`2vf5*AV3VMK#|;dt@Spr%IzU=UYdo=tTt}~;vXFN{SUX#<&Wl?v@1`?mYNs{^ zmI^L%Pzh;uaARQotfLGz=4Z)!LOv2=Q~rl!MZ{omwXD`w78{cnSlfj(x_co(zQzYBbs?jHEV*MxF~_3J!i074CZ_+BFbK|&9t;H~Jrx?gP6;laSq=AoWc>P32pW$?Poe%03 z-jSg|)A^vIoZ`Su%H82LROQr#P(X1{l22~~|8uo;v2?vC2OJ291@cRR_rrp|j$R;W zrW<6gN<=w`0b4G_<6|$x*+?#6w@PZNjVZRyaJ%ga+HTG|L!QTNXEsZgc`c)={nO`Z z6^rA$rT4tga55!j+-iSq@Y2b+RU$|a9`eAxw)v-hKb(mS{29QKogu$1Tq(c9AMz;X zu${kFCLYRB;#Qtp|2l&KBa5O$qh^}Fh)gI+^T+QiDhBKj^zgYnty zxI!Sv7~N}q^6zk9gXIu7@I>ca!fpuMwnkhQ5c;d1ny00t?1e$((ktVuJhGU?Zd1&s z6uI$O=b>G)=M9J-Z|7bxHj{iNsf$Vv#->MfWJ;A~pqHab6D^Kh7Jo%p7u5wFmBwpP zt{w#iS73r$F|!1W9K>#bT?ALzg>)+@9%>caxjE24slf$!i3rJb zE6AQV_J%$bCo>R8KgyJIE(k#oQLqPFda=<3V=+0=C0-hA{4~Yf>t{l#4wfpn@peh} zE9>ZUvj@1#ydVIsRke7gaO$EeNS7*O!qX3v<%eT!pPaq_)U`t{jCq1}~%JV4)p@j~} zO`n@xMqz_&Cx~nec(0$*tW1nNC`S2>Wjv5jomfU;K~qI6u42nZIXyi2H<>q__{yyx zJ*s_Hw#t21q={cU_ntKefZIAmcDnjo;S+8Gorl%_DWZD;8Ys-7Yi6zX&x$UDpEo13 zqK%5H8Wd4QKf4AWD#KeMF*QO_)~EDenf!E2Q;)YV?P{ zBNuKv3JVwh6>L756)ax!-0=P{tTkGJ$!A!AV>QLBqR4U#dKsgh$eW%BMGQzyIp!}N z|HY2Hc$;s=#Ru8Fc(?V2L;kzu4`F8Cq-oxo8j^3TN!yLnk3j4`48D~V0}7nmsDx5M zR{W5MiMag$_tEtG(=|F#PBglfMqIaX>Sw0#fQ|@UTO%P6X8BNOrh!JCZtQ9R_C%2J zHeF_9OHA{r3pUjXI@tg@f^kQ^Q)j~7KF0q;vfdDEhE3Vf52{HHyG4{61FFsfP&+7Q zD@97Egig->@a6I>d4{}$-|CAcROp&$LC`EOw1b}$eTRnC3+j&G+Q)RNL6#I@h$#XZ zS8Pw%AnBi;=L5kYX$vfCu|GK$P2?-VdO6Gi4{wTr&_t(U#!Y1f=f;sw-I`Je8~ z^lB6_OX1|tt^8^AoYL=PfA{XZ4n7_`_}!~--JFy5lO?ZJ16gzSK_pJhW;flk@zisl5Tpulk~ZLHz~G?^$fAUkKvom%-l2U50DkF zjG5VDVP-NZW<5pHse~oLr^$}E4>=Gc2{T26FLE#mQ5iDetLcd==B!ku@brogJ;%-& zV+o}vL+$E~+i|q_%g%Gw>RZPLF<#IDUm!>V0=oiwBhh3|?1s3$HaIz2Q||{IX%Nog z-HToAH9wt;OKHp?XzI9`6EUrHkl-al#y;9zp$B%z=YcuY( zxBk7=I@-df>cx$#r0m3I5O}%FzJO{;2fsmL;19~u;@Z^re|~A+DK*Y@Wpl2|%jVpd zmCad1eLP1OxGw+-LnVW<2XT#@{ZK5oKfK(pgTHEmPMPd;_sf^WhH1wlkaq7rq)1Q4 z6#KD+(TfJxc`$p?Z(?`;50e8z-!V^9r>348N6xaFrra1f4=kps8x(VmB3Ch`prd>I zx?z~N$9E_;D)z?Rmvw{OIC4C*=4UhtO7mvNWbaD?cTkH!e*_@LEM_qeBk?PQty4#ASX20mVm)99|LR=?Z$du_qg zd04Yy#T-3k1BU-^f3NQ(Gu8~Vr(YsjuME~|EU;EeG2kN=PzgKz3b}*wRlHIG;sLj7 zc{ir@aZad!z{~R!;dN;jB>DH(`AJaW;(j9R(C6D)PjkAqcJc2`` zM9T<%qhq&p{XBYOv>7e3KWehcb~moUIc72b^%S#@A~jS(2Des$*@1_1+vXe!M+fKO z+?F}px!Hahz@Dd7Ab(z|c(1Y|7E2^dDaw3RT@+Sv&c_Ruc_QWh4j|muq>GA;iuR?z zl!v4M21ya*9Sc<1ekmgCNRm&y1jmjo)b| zOW4^PH;(k=Ti|US#jK@BDwS|G7~(o_Xp;OdiZK+pM^yoQ(|YJpsN-T!SBBhmX2jO8 zv4D}S+g)~S{lTw%`Asu0+P`w6kW{h*h8sIQCoO<+kYXAsQU@VG*;esJRi&(7yvlcH zWI^oc>4oHySg%G#3)6OQEb>_lOgz9fwVvOY+BtfGI2JInV|i>J9xc8& z^!Kih#=f|3@*ae!ho>f<*p~vqDroaGledESAFU_bAStmi9QX%1<;b8krUr%$Fh)Dr zZ0^r^t!g@6K*{Gb|`6!=v?1?Sg35H3n80?Np4iCHjX_e3>HgU$CjtWUB}

-UzjH|&pf)xzxoZyU^nx)F=jvrV^}t`fMTFFLPsU6otp{ed6|-4x#rgEpObyz znGy|Z+hgyyCTYsrsK$_Dg$9$GSO|7pj2R~5IdoSR3NVT9LYu+{LC;PiY-|F%YrVgo zaLhWl{ffl%>}0*Mlv1x=Ng6r%D(t825jRLq``!zvj!fa~nrZ+HwMa`u$;^uQOo3j# zD6)q>8JHa2=9@L~Bz+4CY#R6_Q;Ljt<% zV1Ha~)S@@mc{f0RGagw%?)X90l8bw1go`2@==CDKx(D=6?#oU|Z-YQnT3kik5G+*`bRI06g$FY-UOdnaUSWSCvX!#Desq7ob?brKR^Hj^iO8M2 zE+{MRwzxNTnE=R0$kpgYVO6rC*b)&oQ?DV*?(u)@ZRALq0ySYcz_P}ks} ztce%dOh|6*xY(g-?3M2`%8u>~+W=F^kcUYF1F7tXJpNiCyiO{F=Z&&&?UIbhYpxY~ z8#d^y;AFBV!|rz7X0D>}pZ>8t7UZ8gHLD@)@lWKOHnCDw$Pxva{5C0`X%Xglo2D}4 zs4At)={|W@^byr@@y@WS&;{SptDB(mN(+HI(_Pq^TWpWVYUIN=0@yrtQ-*Dg{BL0U z&Wu0z%3gK_CXDN+l00QjbOQ{0Uo1GD%;y}8i2Sd{#nVw|H{X2o@cm12vnRVgE&ME5 z$8MqQ#<6>7mm0QEE~FR`V9ucupdl9OX-Y)cc~(U43&Xy$G_E!{Q!*g`Op+A8eIoF* zLI14=X+5!<7OQOPxbg#od6ji%G%~LCyr%)azP}<;Wmn^$-{A z)|O1`QJZavA$oEq*~`wBxUuhg*20z?rI=<4`ZE(gdESFybL+k zcq2Jjm2!8a2CKpw!rMbp3KH2rjuVtL#;$Rw1z!;LP;@q6&|@$pQLt8|3)6++ePDnT zcl_?iZbsgZgN7z(v)S)+02(w?<(Hw&93@I|DolyON3Zs2^#LU~#Yy1^7h5Ak(=b=% zg_|RT+E;%u{=xq>qiMzh!&Gu%Jo(&W7`iFu6N+?E38)lp>?&5F%uPQrS1lq{0&=&( z_li+3OyNIH=fD5{`)8m==rk0s?T6yELU`fO%v=kb+hwUd&3euP0iOThs=N-4X^=!S zky;?AfpW^La<~~Q+-}pIiu7R3!Jn??o||3H$>nxS;mpCGW_iJn5>a|EP^_BoyCpT< z2@k5`l+)Qj1`L9=NI-_WC;Uwk1qG@Ox6?L|Ff{#Cy= z=#R_{10EU>=vnIBBJ7bK5@kwmi_fsF=pET-ZM=)6%^AgSVt1>3D3|#j{aJ3o&iOgT^i$+Ml>i!HPlVJf10k(+V_Nrg%pEP49w&Om1)-)~mdq^U z)kEEtZUQQc4tbo1yi+%Qjhh$`EDigWSfid8d0Si}%H*8)su3TIFtJMyd3*%MHJ7Bt z*3p%Lo$@+hL(Yj_E+6#3s|P)Lxkm|L;wI0jH( zL2gYDe_1Z=Bl~6?Cm1`@0l`Fx2pP6bzpg1v^O&d~t8;R|YcoG|CrHI$akHM_Kk8j< z=W|;V8>j#7553Ij`;+GM@5!?9q|oAX-#{@L6j?(hEPv5-`VloA^Y1G z+sJBmhRKa5H0Y`vwt&}B%qEI#q!Mbquzzx6c%865=>BwMLf5N5BE9kzATWkUFbUN! zflqpY;Fk1VIQDw7&se~n<)mhEn@8|LvArc?)~dtS#)C~t!HuUbJ2nt=(F+B|VsNCB z{2v71oS-qaYFEVWq3^|>AcsIhK{FPhFbx(ga)3%W0xF^gD0|Y;P2d1FK~>9^=oXog`W+KT z$Sage7Dx9@)07G}Gv`67rA=`&vPG5xOLioD*b==r6zJb4R;UV8x5DyONUVh-fyOWc z27l_JdbusY>sbfg|9wCv-wG9*21yqU3NZ4;(JB7G-zC2wOb;Hh1!9eu7Y1o8FL*vK z?4Py2fSOtM(~UneV@V)b`)zWDouA^q9vCc^hWja96w^tOD=1Bgg>@A~lN!;fz_x%E z??xW#gcZgV`)Ug4o$}g%LD{}=9lb9cF_baQj3T73I}1^(wAjySQ|P%ajh3qvuunfe&`O;e)k`4V)V2cR>q>&#R+Tcz7HTT_9K& z)FlL#t-7cxWtFlG`dl=~l~vBIk{NkU3_u==cOO%vh<1jhl5X1-Jg&x|B_eI_*+Sw$ z@{Y(lkmI)HHFknFy)r$_NKo_J1=JwG5E~2~-dlZ4i`$bz@LgKDWe{D&HTt=+(pq6Z z-nK@^`N!S6f2anB(-`#9oC-G9X6hM-D*rGXuV%AV(HM%_K zImoP;cujsjeEk$nQCN>Se|7*EJr8uHi~_K^bbP7m$^aW-tJPvva{n*C%3 zllDU%cP9UW=>^n=Ln3WV+hSbi`eGFL?<60Jb42UCN|HehKhFwh7jdRC4EO5P79SvM4!1?Ghw-hFlLM4=)ApQtcO=5pVL`Z2Wks^DYg2b~20iM-FYz_4QOQ2nrb2-+ zChwMbL-2jJ%k9F`pQpy3GU*C4en=J+%nUQUmxJqrSMaC%}fo_&y$k5uYfdZ(HvoVizuSCJAo zj*dVd$T0tT55?@FNF|kUj#)4f3s4t_CP(YliGo_*;?Q-I)^TfjowBEFU}s|tP=|!n zq^7d$brGw0{iKlI0yGP^!df9-g2%Df=(68F{w49#JCD*grs?QxXGY&f2!|)Qu-tGS zCRkY)Dmm}uFF9uL{N}^HX0phQ!INtNp0yN{N|9tLp?-qa?;Kqv?gWP1jm*w4U@N7! z$F~E$;22Ie<^gmN70w%O2cN_W6|bdyRPv^EXWFw_Jva7L?ew2*p>K-s#TuXk`g6L> z{~mNi8|Vk*7?=$W&MN4^po(IK{LYN^j6rf|#yyTv95Dmvw|nK;zK@{LzHb@?Zo{{6 z>-Yme{&Rlfr%KH|8Af(HiC%`4CYlbYK$H7@{pv?kbpd7@qe@<0PqJT`713S`i?o$upaP+Y zO32~W%(@6unXx_F#cSAz2bN~|fOr;)^> z$&01N3dzR6NB*Wd1Q!9p#*neZhK(n<=Fry~3eHS4+miM-&b&v`-Pm8*X<;4mDFz%E zEtRmH+fUk~4yuw!JJSRKwmos+nkuo+u$gWWZwfiALMpjpPG(pE_W?O6Y?12KebaJ- zE{ZR@zUPqTz(1K-Tr=EHpK5=4u@OS1t`#N!&OueiBdU#_E9JYS8Z6x`qw}~mv&QZ- zd;0Hlh{$5tczWnhV=J0f8l412sV?7`~J&_9l6mCem$>{dle#Y1yB?Ver)~B z{{{B2&!(g0TNnJyQ%T!Ts^^o9wz@UkIA;T)lVP!xVv2#lY#x=68Mb}SgE&JBO1qTH zTBH|2EatM`0zv=8UM{ep^RRcPF)oc~kQ8$&WOsym^^B1fmgZ` zDkGg^jFZi<5p?#ZeN$KeY2Capnz6DqgmafPyfRCeixwzqp%~B%IYK3*#w9bovOAF1 z)`u?;oTNJ;9n(SYQ}vLiX9i`vJqx3u${tu&Qh05?H-R%m&)}lEDB@G!y|U}z)HcwY zCDb4T3kT1Dt-=-}EM380m%~A2P_3d% zcEW!oZP=p%+Y|TZMH}p)n6U8o-+b8%>ucjSaC9Lcvp}z#)J%V>Od4;HAqjh(Fxz*@ zc)fawIwu@O%|2BQOmw9|v3UDz2i)Rj7w*`%F19>a*{oZ?vgQwS+l-(cMzFogRNqX= z@=#Rk7?kbgZ1HJU4oq#J%l(QuTfW{lXL+braafrm8t`tGA(wWsYNxAW9hT$D79#ld zTRy=xd+Z<*LfuUp4HuysyR>$=2#X~;V1L>zuJvl9w+Rn|YTmTByZp`G%R48=og$fT z95ey8^3A%(PF-9Bgilx4oFK`(a^DdJ=X?@(645ugyB*+cqc zJ7?^h(LeplbQgOBBOeY|ev8n|#U||FzRSGzAax9F>8j4v> zkyTb@yKR&EB}PsQ`hOG0Prm?GXnl>Tc=tV_85GO2F8@E0{F22~N-cnrO);4iSx+T& z@HbJ*zkmOWSLbRKDe-9{G;*lrlF6x{%VG;vx!f#8xoBBf=>n+O$A_(z@E`v5Uw$2A#?0dPCaoemH^xk@1!lHW45)Y%Qwe8D zzcT4ezzbWSY1}WB-?XlXWaD(XaRGsy z`bm>Sa}jWe;tDAPc-0?oo8)1jCp4_bSeAt6a$?uUb%U2<06{9OdCKErxb?d9pcwB) zQJEo+ECo`jn7T^S$bh&M*h!u07dgc^JprEcP}j4y^89G3Pu$mIhsvbkOY4mk0|CXg zRKihj17E8s_sijR$$*=dek3mD9E&_Bx-2RTX_3OJO_}U-G{iLq$M(3f!iMcfxa6qp z`hRSnWS)u)v!`DoS?tn4Zj6l@3v859%od6im?tiN3g2I|!0*#}b?9)I*EXx&XbM*3niv+8!PX*ZX!&3N_K!tff%Qsvy)=!1T`||ChdnbUXs$zXy+N`R3J#V=yV82t z7Qej13kvID5ii-J!KM|=&k@v zhbSjnQ!Gdk6$N%c&JT%X^{`CI0l}A6X}{z)gk6R_P@5pzuY+%*( zgVW_t{Kh|y${F6%kOgiWVF6--Vfn!{ib z_&7BFqZ-MP-E4pSZA>_?#VvdkIoFJd8Sj7GL2Ad7c8gbkjAD*ZqzM%Qfl(Jq!9cwU z60b-Tjm=a>rab(#jDmr&E%crku;F$5Zec$`Ex`2t1PKXmMd&270;w$*)O_dizsb%wm$?}8U#zEz5FMZ)tDz>+;+tnY9H*w*)q#r}8S zInk^%TbKyl-_l5-8_zliElgiE#Xw2K4#*37b$Ds6$TzFd{Io%-Y^$saie@!^Zx{kH zd>D;lHbUG;J5=G8a35z0o=oAhJsnAuVWXfU?&@JX{4(Oi!UPH)@klzh@ z+zmg^d2g2jwvwhS4t+1lf9uTOX%&Yg=Q%pB+#tPrmuFXGKD|GDsi0?CiKvmYI`~X5 zCY*Qr)d%#)_Q|m&JyUX&l!a=iE*)PTmn&MwO^aPN{=F+-)c^G8oZl7EZJeSATuK`x zsl1b(3!wlaU$ukVDevRxM8}1C;TB-?!k-HSg~5GfP*%sk=M6-0URp&huP^pF_+$wv z>j`O;`}U0{E89r@bw-)UI)(VG33B5svz8(zr#F z!4n$rK0EI@<7@v7kL(OPyo~I3jN~r+czMV)uCnHkblX6`9ac!wLcbFR8{{({E0~su zis;qSI&Pby5XdYvSW|#i+P9?Jgm*b9kUqrVoy(!Y4iHc6#YhI3%{}%3VI%l5?rN|1 zBkS7QSEPW&P9<((bf+RaC_CuJG^4B-BIgR}qVrsjm}f@OK?pr_pQ{i2ch#DI_cgmN z-)qc1L)O}g^}F$`Qf}eS>_8ev=s`J zTKL;%T@v?#v*f%HdHBJ1$T1Dd(F&%a^H*Wgh zn>HZd7m?&&E!)f-4ou+{i#4@g)&Bb;I_P5HIv?$fZ;&-m64HuE%7xJHwV6+XcrAA#Bir#>CaAc>Mew zZOo5-%(Kf+FOYXhnyuNzjl(9uMLjGpm`gEQie!QA3XR>z1$3qOgrLZ)k&_}?qF%x~ zA=;w)kZk8Bjqjd%hhFAe5WQr4C#RQN$jy|Tj4qBD3plVvf^lkjJX|jPX@71R_d#9j zKUNXSOEPGS-VPkDTGigr23mtj z;vF-$344?sd}RGK#Auq7TE!jD1?ppwC}Xd=01c)fwej#xWZb+El_>#=68=h`Ly%v& z=DEfbRLnn8WKL{Q9rCU4-7W*|$Ds4VTVY)>T2=-ENGjt7e}+{)Ba%pQvDdRmZ@!Nt!yT~ssc0jWS0 zU<_Y&V@S%!5MZi7xiwuUx+-1btyN%0Iv4|#BQNIklWt)z&;zfHD)bzbH%L%Lx1jEnstv0+i%_lx{plsYTMj31%roIJ9_jWa(m zTMnB(8z=_)3inV6re>}juLaSq(lU;w#S`PIN11cn4*q?w-q>Xn9zN}RFQ5Pf%<*_V zY4IHNfLT_!P1xdzCD8DYGSkrIagpCzXhAWjKx7#*dDhe;cBpTk-nW;_;$%BlNzDK0P5gUI68WDaJytv zh^9n@X8*J=)&rRw%P!ew3BK&HnFK~K3@l#e$S-9z4DM@U9{u_2^*=J37=d8zx5=4T zrakPwg+c0~m`;jZp%M;-=e_=bJxB$o>f-O4ylO%tr-A-R*v|CB)Vg&_ z8Kj7jjHy^qBC3lj_uDM#;3o>Yg*CHsc>5;za=U>~p;*vC*NAV*iy|I+?^hQ4RjJTk zrII}5Dz9Q#gPESHS7R@AryT#((Z!)T(I@=%>iZFkCl#o&{Y=*&-A^}-EB{2n67O7b zo$+s!G`^V>anH>@KY7Xc>(UlaV8N8{oALY)Zv-C<+bP=K_0Rp9{(ED3?#oVtjS9D> zir!Ve2D$=tUJU#VOaZOIRd&C3-?Vngj)?hK(~A|vNbpn!e5L)8r0|LuC%Ihq-|@%? z-u^S4afX$p9H$NwfAw{V!TNtD*6a(3T%~Zk%SLF+-`!e$n=sn|Z z1zq+_;s(G}X4tm53j}TI)p3`^nPE+H76?wM^=e&c9`_QIT-GU9K#gXn{K(fAxKbN+ z>vl8Y4Qx=ADk{7mV;GRv)YThSpnxFLwRHY(-vs^urKhU}r_#I3>O4wc+cm`xMhd zk=q!iOlC6VJ<>g4dsVo^(sU|veKdbwAh;z>7rsq2T&Sj$glctCt5{CRH;H!J?pp_^)od%3D$@V0r{%^%DpNh&!t|GA-{%nhbZm6 zP=GF$2LG~yvI|jnV=-R3BUp1dum~1>i=|p0&^H1aMDWh)=%nzW&;q(5rZKQY)G1#U z|8q8wu~ECfp0v3d`m_2hoPDpx*h0ruy?5jNN6ExeB?XnU%fP^+8^sr z5(UNL6rQm#e@~2#F87>&%*R>GkuG@UBeYSLx3RKV?)zP@yzqmMtP3#Rw#d#-vE^~q zLvg!gr4P8nMKSqki!Yc#TL7~4d~Y^r4oT1#lTZjTjcU)%Vo?TU46 zo!er=PPNKr~T_0u44$>S$f$m|@|`XgUarC)bbF1}jjEJ27)>Gh?1= zyg%se(%-(8dJ&&yL-_l@ ztYC)}_jU6LFg^`SI%HDJdWxh|2|1h&ejl_Rr9?HT8fcVrXquPDT`pfO-|E#PH^}lr zQ$%Zg+s343#yqaDJ3X4Kw?wP4m{&kbE~T zb=hy>byiXgv zX;WbSq*snD%>zDdkT|{zdys2+H2wbc16+)dK#Lg)Qy%*#g9KFCv7-oUy4oe3@;YI1 z98mXk%3Go?L}|8jF9?mctJS~N|4{f2P|9=`phgTF8#c#y$QBAlaLhh%mT!t|u=Z5g ztc2V+%4VnB8`6Y8XJ-u00_7y|*p*Sv8h%zI;|Nw*J-yE)CN^dRjMvw z2fa{m&-)SBi7c~CLeBdgsQ5?d`Gp`(~h6xm26 zXoK6Lu8aF8cFG&$3b{8(2i+;p1gV9jh_cuM-?Hek*a~1ySNF>W`z(c&;OlZt{EX7VIB)fk{cr=&jKMC6tjjRtEq&KgdO}s?rQF)FE&C3 zW*|P>@4hOPhxJC4A$Ry=gN9-6KZ7FVKY#sO&42&+zkc%jUy7Gd%p!^;S|S1Wp58Jh zTkLRA^e^|HnM|HtzPPckX=j4LYB6Lr*O=N0hdeNEGu9Bn@pj03j0rF-e~EpV46N}H zSeptvU*u9=Tg(6#`D$vt_Vdqk7R5I!^y)jL*0*dT)eLy&h*~7qyfD7;Io&Nk zD}h!ENrw;W#c5So*xD_{z8OsXnluuSsu(+H(Yllf+7y}ts!qifkl0QSe)qiM-DKvF z6z_E(r0j^-*j=@V-fVCuK@oct!bTj5F%z>vQ*Fa$J%ZhCk981r{xSXHhcC#axnB8A(3W4 zT>sY1ouv4cDL!hlfJzO;K!Z;Om5?*Fn42dqr_uZzC8#RaOxH!>k+Rt3ah>wr@fW-r zd3oYN4_(LsP@TiAlidE;L5~6XlGoFCCsczT+Td-%L64y~(`OEPoKO`;4jMHlpvifr zE(Gt@g*g*`j-cvUkJ-)8-AzIUWc$mUnZ#(=ieu<2+g#QO5`(-*nHeH&%MxnA%A9qX#te52vwqjOO<|f&`E}`PucJhBTzOKt&$n=_#pWDb zLBo(be;}X#L{^sMwT5x+IWIdXvSWA3A4yGrwi&Aiu9J1#>L?d{+s}XXj)LMvPjUWv zR-m9(KYTC6x;5+B=$jiC57=qf>W!_LWg>-ID7YAU%O9!*u*KpEx&3Ae<*hK1+g(mBUIFl_j#uPtAxtbW;Z!jJhxjYT4;Sy>U&MUS3si~R@e7$n#~ceHn~ zgTx1WckBx=+YeRp@_Lf(#`a^ch5gt{F(ni!q7qt!i>0N|FM(0bRw+bj3EB%#ztt+z zCLNf3hLm$Vd`zOY`B0SBD%~4*5?E2YpuR*i;8P+f5nS-v;k7Ip2~ReLJFnWzawHwY zi?!*n*tdM3llIu*rT@aGQ>-~p+%_a)hm57gGY^8ORm#PY8srwm&}s)?gEZ$E^31Tm zWc#5>=^jI)!WJ4%8u=GKh8;Bir~G^QZ<`&Vuhl1g?_5I0UApR_PM2PvkJB6aAv zpo>yO=R{++rZTW*77(=2wE;sONYsXv^(OX7)T`X{22hgJ(UpO{@_ZE@fEM&Dxj}*j zvnxp%6uDLi&vR}}ON$$dK~YQN1+7w4?|m$4xmUF{23a<=ihLCc{*KPyVGBA7OZ5Dm z7Xs^f7kqO~LJZFgqlk^)X(mh9!N`r*LC{7%ESR>AV%Ab5l}cE~y8^A@03#8aBTb^9 z%Bz5G0A_9&{OB$^W5VbL`gt$g);Siyuywo3j{WuQ+;3RRE!a^oxIGIU~SDq;_cuSHb*SB4DuLKk5D z%&`FVFc>*n?l#7O*v{W>et({MQhCVDeTUS!@iOPU#WZo8Vt})~nMzpCSs=h-6HOKA zm18?{PIM#iOstl7!|D_f;2Uaw^| zCY7L^2_x}AMq~k>+CoRSo?_C$EmjurHSym2jj?7xeQ#H34yhPVS}n%y0L4J^tQG{% z`D-V1@!J#)xZZ5zplW%w{{WpoyHQoo>Gsj!KqNB-s>Vs@fpCAwqYvtY@N3p=@j1`I z7#DuVLmuZz2DpoDQS0I}C7p^}QY-@LQDsfcMUtrmtm`}- zWssmvC>3Di-UY9!K+RIYP5GFIUzRA81C!Bj)IGq8(56S-kvh z6ayssTR`0acs`(j4|*(-?0KO8-#S>C0#z!rO}NSPlSxQ**}>n(*NFyY`C5N*4^yJ#cj<=5YHi$hBSHz{|AYZb_mfnmQ|uQS3fU}A2N ztzv*VBTTQ(;cQfY$SW3n=3giL%s(Bp-tjj!$Qps!wn5ekOUv^leOBhg<0LKib2^1U z>$0>Oo=_G$z-8@mJLl_o?x$lsY3mDk-m%8r)e+X(9c&g;ZXAiRBSDfLjBYcCpsi3N zeQ3Ah-n2Y!888fA_q*v=0D-sbe)H!eYKoxdY0MH&Bf`h?2G&XER=o3@f3nUUvB8HM zn;kourry|1$_Bb$g1mlbg-!Gh=rFtQy-%1U$YCDx)^T%H2Z3hxFnN4@pLnb9@H1M) zRll~FE`D#kD~l^fgqPvw!%6tCeoXHPS+{j(3b zZkyUTdz%nd@{n>K8;eksGm~Qq8sIU^4*^NW+6mo4)1|DP4)Z|17#hP3!w9xV2pD#r z`RzA<^DFChG#ks}#xs$fRd*k`Bwj#Wl@EF(k)50tez&j%y1uGJ*Ty*Di^J<6hDL~l zk>B=x|6l#rWV3nsX6euVlWcm)^tnNrb6CExjAB5vy^u=C3HmrNF)}MS*Ly!7^TUt) z?($H>AzxJmU5Hhn-E9cyCn-_=-es|E(p})dY85LvXPKqZ8+mQgl^ksD!R=kj9eyLV z8!sFc2SH)iqP*}IjPL^chGkZ3^*HPN;y|grE&P(;I$;W@S$1zmj_8_t=o^>Biz55~`I7h~ofenNJ?wpJa-v`rZ@^gd55ZTs zqfWRwc#BU7cW_##eCM18#C5P@iz6!|;t(cAa#7Nw2|uw`)NtD>Tss;W*+CU^>S)YM zrE#^vnfwxwR&_E!b20!5xdULqtV-h|1E4n8b;E1@>RArA_3bX(w{JFVba{-s_Q@YR zttmsF!?l?-1?r%Utu~U>*Z)o-Do0+t%iivD9?a^+yKnvSjraU2qs?Bc?2npkvYnmR z>c0Ona?HYOt*01Bl-EE@h^&Dwl3+TgT{7UE6tRk1N_RuYd23W{K&@Au^8R$qo-m|6 zFN=K;mndif2A@vOGB3?dUsJJ-k=h4JZw$~IuqSSXtdoNkIT(qni>g!h5S)gzio<~r z+7k|m>y(*{bG{r$J_M^t=Eb|u0l1^5-X7Y%?n`C_J^1oJ|Ag#khafjzHD9(s&`F9p zL6MKBglmx6SqK5PKi!vs4m-1kpaS~=l~z^s1+2j;EI2##gNixv5WKN6S&f6SZ>(wG#6sA=)(`C`^ zlH(H(#X7a#%BiSfiHzr6{|Pa$p7zSO2~C}CIE5XGO$pCeA?aMXUuEF=uVTG$ zyX4&LJuzB^9%{p!GhJDalT$Eb{ZJ?DzjH2#SaYVcSzWuarP*nLnS6?Ygprm?IHKC< zxl+DMsyRL>mzy}LmupnU-VoM9?s&FJ%b@aG&tR=4mWLz5VSntwX&Xa^K+-aW;~H!X z8^)2tWY|MousQhV@7}R)Ze;@#H;(SwX@xaWDAWhtpMIC8xeR?#Z9q%eEAN-o2O%Mt zfo_F@H@5!9Xhwz=G)D7_Sl&C8TJ@&2KGZ7$4LfR51`vbND)OjW#rBE05TV%_ktOd5 zxk8E}F2x;x<4o`sKg~n=6~CJ^c5qg_zIaj*4L0VWZ;$kf@S1$v#08$fk{Wx>dwKkF z-#+;^*-2k*aB_5w$PkbrzZiO*%y*V7ylEU+!N!*BI{X+*Xgu%o#XhN(BD1rS%U^ee zq_}Z~OPR$~qNSKDifo_~&XRubUT)HqwNV|>E67RV5m{y6GA|fF_@8Od=Eg3`!2m4A zHD2db&(x_y{=Cr}(K&W8qbU=&6#8QV#fF_ar8|Bt{JZoeahtl~JC}eC z@-J8ADd9c5B{RBIW9e2H2g^}ZjAUd-a}O(2Q0m!pu6dfLl%_Wu{z?{)Cz~xScRIzS zQ6z;*K%qyIwxdpTRG`r-27HR>18;>b@zyHv9Fkd|n|+7Lc3p+z6O+$o+=s0*p4{)M4R)68^xTKP_<wX6R3fh#w*{eJ*(JKZ$_QstB9n(RNg_705gC3c|-NKu|qrb-MV{UEK zpvT6D(@I@f3e;ZqkwVXMNHQHH!=H0Yihqk2`A8##-Y^u7aPtM2wGTbc_m|BkW|-~t z$T>!qzB1wa0t*1Hr`h;UAM|)2Ef{ux49a$U-jWtYq(+<+zG#?V z@V42VW`h}aPafUXzx~$RU$d^icH8BSomy+`3vGjG>H#Sc8TlnErdH1?<}?OkJQWKN z_RXjO{q+_p#@~VZEn?TyOu@3T5V&PC>Q8OL=tk6$yX{*FTeVH1xcmEq%+_P^dy`fX zog2F{wHEeZJHQ^4s1CB0M^&ALJak% z%{NP|8Nr_I>Tut69kIjvRqbCi*|gRv5#8ZKwe@+3%r{6dUXNu@u7QNzv2q+Mb|2xA z6ZWwZI`P!F(a=n~aZ1<@%_J5Oqe9T5&<$#jxA;!hm32? zh&{Z`e|PPFTC-j}YcSk6Tw;gYVpB*4uj`+B-+279OX?kpBx+0 zWX{u(bqFj^FCVk5e_l|taGcpIRpssa19{|OA3wI-TDN7EB9H+u4F)Jvtlu8(g zPmDKGj5cxlfv9+_tcBHxz8DWyH;mElami19WUb=#QcWQca~(G7Yba(lMOIM>8zhx} zea2*72UJI&QkO|8{2Jm``~Mw=|HVU()5RE#_po1M|GWR@mvmpV&gQb2H{96aw3GWa zvBaWA3oKNG&=uSPl>u0eI+mpX=CN@MFJq6GqEo*p_{Ty8RwrvrLfP+k!(1X5>tY3maWTdQ z&$*}kJ7u}v{l0fmxSlqZfft8j z9q={vyWrppke;E|>pFxM4e}<^!AEL&y&6cwL9V`D(j;0AJxXny?c58(>%bn1od+eN zx~KuJF4QSsFN;@iS`a$$=dk&C#GCdfb{h`8`$h-E^4MQ~-Mm!!bMc-Vq}h!F=+`XV zP(8(*qR2@q!DJplJ{T(14Ya0Iu$e))o(n$%bodSQF44ZQ`XC^;r!`2`d7aayXj7Pk zc!oSo7bBN?x%`Lfjhqb!O*#q?}k`xi3qE#kj&koZ^9Lrf6-^*$cMlR z^RLZKUbfkK(sJAUot*da4%vC<566q~;QZg|$P-!NVVr7x z{iV^gXS#9r)=qn7tF$)|DH|`-t9&0o^=dk~B76|nLza1!QDqdS<}&0bxSa~q`C2cu zKUh_QEmQ7;qS7S;tEkS_CbvXko&aBSO?XO21(Q=k9t z_JLnMUI?#V_6E5=o-BE7c-o_%V(wF<=&IEF+1LhhW*KRUDFRmb8!xRX1&r zESAEuPkL?URAjks&aIIpmax2m3Lc4|7 zP?-|9ZmK3RvIH(cT@tNgKyY5LWClDYMYNK$I=C#bED$22pw)IozLb|sE=Oyw%}f;Z z1If%WMN5?CnBp!Fgg^a1*2Xt08YRV?OX3Px&K_mJIVt9JPalw=uyq;t!=gu6nX9jT;2UvL(?wc>kj+r#yS}56V z-QSZIR_KZJ@kWv>x)@XA{rG2w{3^#FX^+w>)_GS#TMn9^%lz)JBhp+^13-2lz8{XB z;q=g#`3C;(z&lfAzZZ>(N@*Sm*aH>t`u$^W9b zW)|ED2}jTwBb$^TXf}OulrE6#@I5*U#R}ilUxfc>p4@D~)~AJ^CF|HteeN5i1m5A{ zhPaSoAY7A!T*OVW_r(~D>y)R>t)si7X>qk)cYsDROV9;jX@Fc%Id{bb9X58%S^wn+ zU(BAn^{wQ&x8^+j0^T`F*t}?PM-HnOJ$ZEJ>Ew;g);#!ZvPo_X6gv$0S;0LpH>dI% zBnEyis5`d^%VMuhHTI72kc&dEz8TrVPl0f8ZB$uoU0_|HfuA)ojdzx_BW{OMul`7p z$~z^!P1h+`FfIel$l+p|LPrM*`*(l#KNAkT?0py&{+sE~rgq96|7=r#I`!kJy92YL zO$Coz#kEnF`I|rlJO}(QtfQ_Jw~dK(#)wAU{$M%n47<5!dwknn#yt0Uf8KYFtbb+Z zo+^vEr-)*poI00EXyX(q4dB`La&vynsqt-UtjjD(4lZtcUsBmlx-Dn zgrwm^|MgQAaZUpk3n!~tyb_MTJT>ZrV z#8-mNb4}UM52{JdD>K*BSJy8|M|oz*nL^H?~?IbK@lz=#F4l7e5xFr zp;r%ZQ+OHjCA>#m(~-OWEwWQkAaj@;SMG#(Ch!iE^;2uT&duH+SwA(MyV_;B?NQ^! z=}0j>FomDqW!+Z%imZ_BG$EJLgEKHHca${KJt6N4B|f9oh6e-7Lr)Vf82q zE*SV-vWnOt<<^)IFAbF5#U3RP9EMU+XC#-L0Eyv~&*SY;n-*67?`!vJ^F`LtTenRD z*onvL;%|kOQO)8$QYS<-lu^ZkHpNL{KdJTp|Lna9TvJ(=Kkh4hA-OSRBal1^iUbfu zWihm<747Ppo|)det9yE;r>A@RKc=UNzlx=1zc#lY< zY^!hLOnnq`-8ZR5{&bDWcsh`d1LbKKE;(Qalz=_EIM&h|HZ&{l@d|}V;(bkGpiy)J znMib3Bm?qXdU@Ci|K^BH`Vkz(%^#3sA!ns1nK$TB8@7wS=7&KHqqGh7Hj85nKx70& zHHI0O^1aF@S8_s~8ko#8-ISrKnvn~MtvJr8kCLl4(nutp8LnF=><{Q7=ZErb3xtpc zJN$ce*qaFZ>~T9jcEjIidtaY@raecj#?F0{3qDP~v*Z`%>Ff5qwbRM%iDae4?)PUD z^ME2flol&9Q253u2XntW+LZ7QwF5F3;G2pgyh~CLqQ*pN%!X@}?U5(*c1L&1(xr(^ zld5^%O+E+?aGK{W&ZW!dWvI5#D4VxaayYa-=u|j#&;tCF$;@7^nHMPy)-G68fEp-| ze0E18M^X~Uz(3)e3?Vf9EpUqJ;TAw$LILTCIpJ$!(t+O$NoeunP)^<@=>t)!7<$LP z(3-mNYZUdxl+nkyZGvKuo~q#E_EWR00~YMvAPiF~{wQFbH^>da zH9WgZn(EmmNZ{xt10LNXz2sAX=}m5jAXB_E8oxw6zQ=%l?i26!fTc`F_{X!*-X083 z;bD6y-+?TkW87}uVa^!mS=ZTOdn~T)do-C>eXLjcZtFsU%RQo`dN;)sP{crK8)eAW zvz=ZdG}HK`DKY|&@%j}*oIbfeN0le0_ABavRt(ccakIg68A?w;d(8X>{dcTUNK-h& zzaHg*I|BiubUWr#>}KQq?__*SXrA7gHxL0ke=tw`dmcpxa`aWQ{Q&kZ#E`9%j{? z_Ad&D*de_&Fw1N4-?`EF-#^!WFUCBlymerwo)o)r9I(;CKpdr*YKl}KL2XWGQRq3c zC8~q|P!u<3zp?|C^r#Qk!X!X(N*(Y&Rfnw{y4pq;h}*c`1pnWn8dR-@$;4ELo5rnB z?N|0sii6rXU|XiQFQ`#NP)MB;(?H|@5EX=S^Bx6M)GAgAcY+R`Ggu1kHYoOn15Edg z*}&K~D5I2%Pqf4J8 zz-Mig-H5D%rOv*Eg>+_Uxgbth3q=lR>1z7AB7Z2EuScQV;e8f8-ANgKgC1DGUkp*H zYUzNyK~Mnv?#)oN(Hm65!77bBT6b4g!!byW(YGXxkl7g=VPmL*x1J6b83+ z?GWSc6rme+*z-JTn6#0D{&Q%Rv=X9YrKF2)1x?NO!@KDc`E}7w;T295X^qH>(xpH+ zHBDh^T^WIaC?vYUaO(}f-Z}Uij+{Zqf>CxB&W^iBVD@#-*L1;V)ct)(^+JWFA_*{u zBc$6Ix`~1Mn)yZH8(%LDA0i02a^q#S4K@JOJX|X5j$I!C zUvCV)Z1{I4kLc*n)B3DE<8Iw%Jl@-$(T>ml{o21T*5$)M4uE?mI%MYFuD+*84b=5{ zHAWzfM5Z`b;-ZM1tzotYGQ(4Ytv+0FzIn5Z2+R}AqVpS%lGW_o8aG~M@38Qf(n zMN%nkwV((libwR4d0n39qO>cn2A>JjPfHAZKRivbMo~WloRrx&72{}G*nq=Dc{QVA ztNNs;{_|duVx$XUh%q_lp4WPxtAUVqb19+?Lb4T8Z}}xohy6z*ohrsC{uRjyId028 zAj^Hmf$hQ36su83%9$r^&myBOp+9}=&DGXR=;wV6H;&%gSXTEd3PqQshh~{QQBzzc zF8uP$Wt>)i9bHP+N9^-!v(0nOZ!0u4WY_0R>Fk8Xj8aZ9Wfa*@X)#XsfzM%i!?Y$~ zY&^i(Jh_2;LDUjry2@bcE%p$G{_8m<)=ap$)QKG5xNz-qz27-BY!0?Yj52zmB6g1Ma&Rh%Zkve#<=d#Q(bdJ~{QuEOYKy zOiHa3bCDwFA$kh}-~;l03DSSmiO&G(N24mAdj?2BS~PvcbcCA#N9i3(Jcks+2Ko#f z)rn2#j?$INcDhNE9`XRT2%0q=nnxhkX@Fv|2Em|#=DCLEC7&CF`)jPLurRYW|*mTovh^Fh=iP9 z+?c^Eg;n#G2(JeBaH|L~FPmMFw5VLaBl8MHho&r7w(;Mtrkhj-*ds~}bXCraow#^` z1#k^aY~x~-nR5cgo4HtK#>$DWe(efb_sUFAdo7TaO))^axRuh@h96}f`d9I~=zCsi z3S_7<5rW(eY2>tH%@b5Kv}ihGahfRBtdc$SM~ZO!r;0I*s$-BE!w1^g)VMC_yZ>bd z$!`9f>Es$akht;uGhhLdE{f@-$j6ix7d{xOgQd?0WS`#+Kg^)pBQJ+FbgN`g`H0p5 zYhO2g%kLyHoyE;vSTJRZGlL4DbnZjd9=U*#J8m2&@oo+d=(Wg?7$RPO|<_Bkhs0vh8B=i1+_Z9kf2VwKGq6Vzp&_PP*HuyV^ z@lv?ZS3K-x4#?fM_Ms>7#C0Dtqvz?T4tT7OXrG@NSRS&MQ^BtZ+B??558LD8I5bTDkL{kvcEvU{ zMsaqi{PxNf)~h8O3y>Z~`5dgkct~!~)Mb%%P$T^y29hC#|NjDHc*Iyed-t`}A0GNi zl)rhVF)oO_O18N1Y6$}WBdUsaQVc{Ib1CfsZcgxEu<2*hkiswewP)uw2v&HX^9R;d z^*R4iVu-;z)(rjZfr2qyT({2|3^s*CR(N;w_J+C+-eWW&!>DumD=QdN;V;r9C^(FfUGsm9a+JqzJx?GC=0`2LEk7Xyy#`QelB`8}K^}dh{p~ znF8@8Dc+q)?~>})`(RmZrno@dt|$=R%t&E3d#t)lX<#Ekf;OI>6_e28mqc5NjoBwI4C&m8lyBu zgXA*5H~D+i z52L>0g_{wBZXbk3G#IBtLlu9mX4r|1nvzZqf6l@s7c8cclN3`=ky@mz!nhW4&kP0s zQMMGMgrTH)(BlyY6Qwppm4@nCBksi@EeS5zmI{l3HDSP`l8f)KHIm>|@DIOn3N$e? zR9V1n1AU=^GHFC&t$mteqxkR}um$hk@2$Hf%jaM_yb~j9m!$jrPD7!IqnW$*VaBvh zcJQY|D6?&LzVYS1G*}1UY{av2BH9F-!%RPM<i)5#gG% z-)pU+6UZ!ps_GNLHNPdo-k7!g6?3)(meJe%hICwac&90P2N?`SLK=T6za@ z;xY$t_|db%iTh6L)YqoXu`ZdmQJJg6US#!$!Qk*KD-^kJCe_8HgFm&VY;{-52u1cI&EOu2 z*+r3jN{dt92LId2RXz_Cm7@CCZu$nu95u+WNIFB+6mc%1ma|#i7kNE$4@zAXru!`K8ar%9!<7@DJX*J7KF1Eq$GP91XuVK+zCyr#?ItK7jYQGY+$fBe zWB@guDGq0%W5^U6Y1d%kg@a}vP+r)DOD@_5il*=C!bxVQ<<`$y|3X$x1X`|<&v`S& zfL&P+!i#jTD3KYE8|i#lbe04`DC<;s*`f|e-R*(5Q1aWN(#NbHM?DDp9{1DP#3he* zv`%e5?ok|C_x-}<{~Nc@I+L4CWW{|aPhuk>eWms)r0MRGE|sQm3VDgXophqQP}HKi zCb`J%_Dbh;(O0AgfvIC0O$$4WkA_Cqoa>%&e6{sz=M@2mjYS(KIN@?_-@*^S4h(W% z-Z;G;q|yvC9(xt~F7dqeMkQS+>YmfAY|-S9I?&`h9=&o(kDy7TrxF%5Yg`Q(b~t%z z6J4FFGU4Z8=c!z6U&TgYe=4Trwd^FQ%&Xdz%7i z7;Q5`oNS+8Nx*;yvH`czYrF?MF#d*wgV2wv5teWjQ#XZHhT=YJWF=vvrY%oOKH_HxfFbsqyE59pl_eMYAYYF}eAq@6Hv zEmt-v$7pKogwb<`^AQ`+)xuZ3LMHU68FYZ*PryFT>KcbSgW3&-Z=kZ`%vIqL7n3KGeV_f-h= zh&*2S`#4avGH_7CzIonxdTromN<(yi07Q5q`@Qo0F2G*HX_CylC5aR2!zw{Lt!k1X z!1bob9!QNc#f%CAXP%t>{ohQrCS7q`A9cTl-FV7dit)QL%w#;kEfE>$Pos-tE|Ye} zp@r||d_$KnxyjwCTE|7+$8=vP(#V(O(pTR?cO+k8gi;M$#w2k2y_~*?c{WTA4*pX- zJnPB1N3870gp$JPix-%un5OS9%Oz!QJjI;1fXW9HbDSbaDeZaKox#=bP)$`lxhgSf zHVk^COFs~%DfWhL@lE%=8eyRKgzNrP38Ei;1oM=V$y#n3J$z-*qguL>*Q8oMsg=Kp zQx>X^!aOFZ6{`%*W-x}o4d{FZJ&sN(_B#@m0SY?RA=}l6KOOdk{_&v@hnbVdilwKZ zvU%Vb#@J!WE9-YL);l>iHb?ZlvUz#aI;S006^af+1@Hwvmd^D z{9f5Td$(OEsDjtjhM5gaQRlv;#NfsTrq;s19H1Cb*xN^GF~r&+=oBT*PZ1|^bRFtm zDHeNG4k=q-^Nzv z*3gQDSni8V?WSy+vr~^r_RK$_Xo}SHKM~vori8r~#MDk1mKSuw3P~^M$Jv*Js)|PI zWsuu?xHcFL4uqc%Eu#v3+vQo)@_mXpXOs;Bv@!?hxAKpZTJAo&RD5y4w$S{bOTh*> zgNd;h!;ZL?jo&m-tcLxmxp4L%Ki?lK148}$umAIHiFxXg|LsdnWUU)dJs>eZB2=GA zF%V$e3?d*9bY1TABoby6Q~<%ld?0=6_W~~c0`7k0lSqum7K?O1#LrwOhobwv9QHz= zf8Mswyb=hR5S?*O@}3zX7r(oF4(WAcgsijBVOaJW1Hudglvd9x7d#}@>Q0$nu`M(` zgSfx#jxLutv&|n za(sPAD){Lk_*-RKUAt$#wjEwg*Rb`0nHzvq`PpBv9eQfu;$M&+g2yBOhTb{-UPr>} zXwx~j<;7rUs3LCedGbp$48OR$)}MU*%G4dkTY$EoVjfbYm(r$+%jOL!)b0%3C2gZk z#UjA00((%R%Ru6vDTciBG)1HAl2k87{uRid6XPYI<2XaNsXm4chP~5um!zMH)+mh9 zY;{va8|+`}8fC~tRl`Yu^4I)e;OGcM!RriFA4!6!;GhTc1f(fYgt;`dG_;k!7i4fj zXv8q%GG{M8nU@liAG{^%I&6Y{x}btLRLe5_`{*t*J<#=9S)*q`pw3h{m!`Nm>w*+H zXEIdyJ0Zz0YDuJcx2wA5Y-EKr!1X(A?ANVZd+Cgd!J7JEvXI8j-=&%kM|FXwZ(ug`W}CO0%U^)9{QbO?{&n zv}vSS+=i(8Kw^eYmumyaj?JJ$&X(`9_!$-YH!<=jFNEt+NPTY9@cUauZVfsV}YV9ih<4YEtIx$ zNGq*dfr@D>c^A2QK^DXp6Fsv)KKw&)dOrBtHQ3r{<6oV*ft$qXjWNm-gq4CZ8wNYZ zkOiWinFu@j9qTP$J$dXmKQh~n_HTXt>*UgC&Wiiyu9MF!kZ^}$Zc)TYX>(!!Xn9yM zsBfZLUM6fd9S+@1bj8sfbRWr3UzPN5yZqaL24;hI1F+C#sDUN04AjYo7~ioTQ`afm z%hYhHxyzM3+(brK8@3As1F)k@ip=1KsO8FjZgEVj1l;QVbCLM4RdQ;o?z*Ih-WZ$A z+sp4z4;573m0?LW1|7QSQu#@;HgM1b>0D39E9w1ni(@kBbAmGjm1dDj1~*}Q1gJH- zf36O<%ka0uy|Vn#^phOI)H8HB?Z_H~wyo4Z1#eHXc8%FsBsa#P4W}6uk4^?beWoV} z4SP32A^tAz#_2 zU3I#GyIid>&sG~#!Y+^vZtUs8R_BNSbRNZk>#L`1-^)X=*82fqp^k3$?D9M!xD3l9)UP=Ug(&yQ`%P1q1aqnhZ$RXj$uaJ9CRD{y)G_T z8G^g)+3LaQ%wWvT(OpvX1E+490z+E{0H3&@G_(|f#s0`Hs2t-1(N0P+DyA6=7 z?f`>~;~;zvl`lEL4zz##)uOhkW>~2cS5=X1>|o`_LBGQmu-Z*Az;I@uw5ZZ}RZ$07 zT@4ZS@^;|3>GsP|RjNKj3pL9)6~(6a`LziO#804pvWc^q)9l+jyNA0fRL26Pj6>eAd5TD*2>JQ1exv)JAOG?v z@4oYtFvz9}uzMWGcBAd7>%zRZj*KNjfj}_wrzgyO z=eB4v#U0T4#M~VquXQRsQ{1At8rDr3WX*!(b3Y2W8g`U(D*RL62H2u13U89F7Hq$DqqnG$+qofwSIkYHzz$0ZA-o08=6yBmity?{8|JTtg;C+(} zdm0tIjsni6IdkB`Xqb&wNr}`9IMF@PuSl{RFR}_PAf%@lkg`vQ?QpYhqwX>(S9g;S zx#c3THW*Sls?=R41;Q@RVf)f1SU;&lfZO5?5u=XNXwS&ua}Ex=!=GnwL)&xID9ft} z)1`l&Vx0(PBj4(@`mqFcH9;936I_-6S>+yiG7qDjC@%n!PPkYgyvNrKsc%BKW0&*U zv`2Xg4h_U8{W$v+tATLe`MP4Lm?6BkYVOvkL61V;iwm$- zmKP?&pd8Mi$DPR)K3nEvDNKWBCNPIX8BT^8SJOBAT2x!+cgJE%B7A+IlLDZp#~V`v zO`H_KEPY9FaY2UaBy*8l8IhpGTpH7@`H)j%pf^eHL=N9ST4tqc$LH#92*lD;w$h=o-!*-VX0pNf+>q zm|k5!O*t3_eEG1A^$u3%k@~{jZz%rNY=tuJ2%nJV(bnJYYeh+qg{?7C3`GB~Qd)gf z54RJN{QAAJrnQQ6ebSzocHnF7_Zkw|G5}=&#ASGEK^CnTR4r;aRnuC5R>DLgk*n?v zs^JXq&U&W>HK;I3*d3eC*%ej6kEaGh&(Zg3y=OdC3Bse9^k;JlMFTKLSIZN8bszau z(zi$((357U<0#yw9Q0@t)PNQnsKxbrS4K3c&Ijre=5Gy+hq}J4p_oT~!^|Q*2s}Dz ziY)J?QJt~}3aHsh@%^Js6#w6toM#5e7DjDj@7`C6j(eJMR`-SKzmjF_W;-{I00EKl zh+T|Sib_mzP~f<+qg0L46_NFOl03UcWpPXGJ|^i@epY>P(EN#*vz^7=_I z_$))2c(-H~e9XwJru(FrJ8F=e5meI+@;G57gnIA_`ZX<@YH2l%qHk@2Qc%JE^9@Bk zsUPYHTBqp&0PM z-v5RN%i-XU!v}vFKG+ew1U5&oM>sY7w`8aehc5TW4)L!T@NnByWb%if|IAua#fEZ; zQQkw^Y12;=p?#Vnk6r@Q(#VAey5w`Wh2EZBB1f6nar=VX`^6r%-E6<0j^1y3CKzS4 z?7qgm;AHULnP&U6=z4wvDRARR4+vw7n1~Kh3@B)qP}&ww^SpHFSy3Os6%uZLq7dwP z(ekid`j%|BZ!sWv@uqok^2Ph**-V^prSO*QjzX6cwNYG8<7K4w!%m*aKK4o%7(f*R z7ikY>7pq}0wTZhmDn-1-zasibm|e>ydl1DI)x+qv{c6vrzWg3msGjiVq8|kD&EWZ? z?%ca1-i^Tn;_4&7vx#D0t7sjiji1<}DJ4CMl$a#WGF~Njn_s2C5VCS&P83$Tl+C-t zb3)qq%OCu+JJ>+uSu58cv4RGbwBf&gZawX=N#1hf;E#>j(-eKV4x3tNg_3wkwlkdg z*&r_#7009v>G3oQipOm`l=T|k^)NDh!tdUE6k$!v_lm&7CW+q;x>Iona=Y3UPk2e= zQE@)nV7}bCjMmEa=l&wt@|N{>!t;s+?(0W^xai2~Q}Ey^5|}!Q8^f#F5O4t zWRu9$0YPUvgl15Jb~pc&;MSZoq>f%cX&?P$oPhCi96aZ0xE=)X3ASgM0p|N_UYd%?s^H(3cY0o~^Gnt3lFd8}yx*6fmS8L^ zLlq}nGJa&MY@hhJd>vo-jr?3Kbftl7zi6Fj<+kiLHZrFw_-z8Cd^IuB_a#UzpGn`D zj06-=vtJ+_^e}m#N4ZAn5Dma-QljeP#rZpJ(fR_!u?4Sra(O|I&begUT#N$V*Z%d+ zU;j~TUe9I;He4r3ZoHlaU9A!8*$j%=LXl0Bb~CvG$=!y~;n(3n-ZEL7FoAc&uOV!= zPva1FD(MnQ)7){HO&qBya1H=o)(b0uQ2#Z%we9b#B&nbK{`{iQ8)i0;Ng3kWT*X8=b}h1a4XHJqo1`Iz8}k* zo&STQt(u&FVkAu8eP_ic>qQZpIEowRJ=j<-ncSNb>dmH{D3hoL@(E+5 z5h?>B9~Wd7tc+S2byv1IXnf{32kGp#LYhI971Rpk7L`=D)riuX_#$T_K zhKMw-u2EGnwGnuu#s?3>1Ihvy!%fEJH9sf&yfy8R+pdOe5DE2jOC{a(CyI@)4|wSP zdX)nn-Mr0Vrq})RaE~AN{o{m(W1%FG`();M(0(z>4+Rc*R79Mfl^J3bfWiUP4t%EU zrXN8c9|jM2q)qD#8SqG)40Qn^wcPYT?4@p|Q83_<&D0ByldYaxee}#R$(T{Z0vN1j zkg=kNoj+VV@cRd!cc%MXZl-uwOs{hIwMBD`lm_Q=tLa|Q_k)aEPPM$2qmRN(VxypH z+C#Do*dF`2NxV9`(D!j90?z0+e6%jSCJN8w(mOf$`as>Sh#vxFPDp&%vv=u1yL}!A z@eb4LSODcJEDZr~41i*XrL^BfoVQ+EJ--9s#tuuC#e#f2#UxWCk)26=K=H~6&s#zXwnfvf*v0J-7zLO!hZ$Shi#kJELBq5&1l%F3 zyKzcuQnkU9SQ1?ma6<&s9F4z=fq%}wYFe?f7uuW(|5%yHB9+rlbg`PxUU0MBu(XY? z-+AXmf8NPNjs&%WmJgJg?0|&Jt~mqpyr_(!ls$+Go0LfhJvwM3sG9cE+ZWWR_baRB zzMbq-tH!b{lq+>@e0WTtuzQHv-)I{YqeX^Oe)w_d=RGo9EhzHSbC9R(Mr0k0RY6JO z4czW&xpX>I=~wwx@b#QrVFFN^wJY+Yi^eO(lRZG#54IyMSOJ9k!f#{lsm$|~=H!pd zN!BY91FW)`!S+!M6x{Blv`w5;@h0hpX`jHpb%C%5qAyrn-RM&iw3X9M4}PN+B6pCp z{a8^c*e!{Nz+Z37Cg~AL158v;g14&f%k`>Hpk{l;oHdH2B$s=0X41^sunP599AO*7 z%pPH*%{n$ZAH(VOKjyrlG-Kw6XC^e0bT{5^I%t7{0*V2>mF<)kGgt{mr?LB`dy+XENymZeL_lSi9oSm2?j`i+c{7jz1NL?wQ}AuHeIgQBXE-&-_&` zvo~WN{xJ;ki<+`CF#T`5p|D;ay&{o38w;nMlEa}}eG_LM2`v!T5lp8XV&=rXZ5J4c z7mbwt+m|oy#raot@BYJa>$EwyjpNx!ku%UA1YZyhdZ6q;qVL|&L61&ZKZsu!2=!1M zoe@|xBQ>ylQLCs;uvFMB*bNG$<1_pY*bV0o@F+d9vKtfP%S(E!5%PRk#f=xxHc$gv z=i+7wPCs?@pUUR_X)g$-AYWpQPrKrZI9qz0#Cv8-Zz$RbhV9Dc;hp1w8wXG`E*iv-H^;FzxXvtdu3eLVhg7wmtvrlHIveQ#u(;ZCp7_EqJF;wYLRbx?VvZz zJxh1WuSkF2qCo~u&_m`HOvl_A+zN*M1I1R^3NN%vUH)feuEmi<2r_xBqgV^)o%;HJ zhlWL)aU$|um_zoyGB`PHfs<;Asi4RqN{h^f;L>vH=xlX%%n9{%-cnxsoV}qdfE05F zjp)GtaVwxrQ2*5me)ZR`O3sQJWEp|oe)o7yLsCmO=PdEe07k0I|9FUl=?_`d@6Uc~ z`45vl8)S7raCSGubhD|`6Y5f@7^yKpXtbq|)4?M?DYtc3Y`CR4TqJ9`FZo1} zd1gk7P&$m@~oBhF88tyviDi)`mOkJi+s=WI_lqXgHe)@5HhF_sC)jdT8OOhCwW zI3PdhsgFuilm+CgbbGk1B9jML*`<3237ed>UnSoR5K* z!=+HO6IG*H3RlAF&6*#5Z0*Xrt@W_sw#GA0g!sQHEq85TH@!)HQgn3Q8K5C72w203 zXI3ivfJXO<-x^M-=o*vnGQwQTk#bi8SXYl#V#xszjcP=mKRGsZu>#SA#hfdDnPCReAOF-|PIkC4h>lx;=pe;_ z2uTs8%@_8OV+2LfI_c#Td$_xU+vnUS{fe`6p=ia#Ce7W2>0i9?&EFsTX79q^WKr>7 zzO%4dGvIMc(&xL2w|rt|V6JdOOuhu~y!xerH>ww1{Zi)l3f^eeRPb+t`?EEuE(GPd zoHz?P7dm5u%gENOphJ27*!#^8Gj!^{xUP(BXSYOhA6Jd`7xWkS>h}Mnd#%c6g^LkZ-jkGKrIibcp1mfJW{RmEn!Z7FAwU9+){DR?C~1 z%Z_^*rIiAd4@Ip%2Q&jl4%}$R&I8$Ky*qyHf9`oM|GYakZGxPD>--i?-lA;!{qR)y znW0)a@g(mE??_nu#H5IQY9up?qi=$`p<3P2Y`EB@0}P6rPxFlC+r+r zbPa#7Zun;WtT?`|Nd_qrP?l1`ZRche!4>^RBf51)HhOjaO1O^eVPPW^R*8S1ny4JgiRU z=|CVsmaiV1)8~b1Y6rbq9RuMVb|tAPj9 zCeh(Ly!B9xQX1MP)g_TzAUumFFeexn>6fHw;Wrh!BcT@;^mE(zZG2V%9tUya(7eP3 zP^=8kYZ0@(<9yAix-{$jZ%HaUcgc+nPmu)%w^I!4aAr_i6q`lQr%_flDFc-Kc1I@z z-#biGn?W1C*HKUnw|CX+)<&l0H8V_> ztqiXvgC6W~=DvOs$R~{;mRUv^dmt=L1$zG*)7Qu8wJ;Vd&Rl*<*;=z$6#Y%XY5(JRa5~qYS_9X{GMX7`WBEX zmj<7u?|HRomhlWU&TtJ8$-G>3p&VOuXfBXGxTfb+OY;>w=%vg?@iEZ7MlIPk`cY)7 z2uoEmRQDq*rm{3Q_CUp<>F2pd>;T5fQc$r+zMS}ZFH%DVwzzPF7M3{bFidm9uXoNl z|6+wvkl<77dq%khl!KA`a7)w~rS6)f5@=YOzlhXB3H%D4i-Ue+0SddJXLmO%puFbY z9prC4C%Nrv+{P@^q%uY)GDGfgndHupLMY2GJ-a7t>u!4xVXLhx&$C0sM`wQXee0?; z8|7TBi>`u^MSrhA4BUch6PX~S9Nd9~>I2+je;^rLB|FC3qPiyD$n;68rmdKBIC#(l z|Dn8DV$>yRD@o#<_RsK!ThWy?b>Z~kvu9|YGti@)aP}~<8ydIu|H>mWH+;+N$Hf1- z{60D5#(oT_#g6b*TPfxuMb1;&Mp>KS{#UL?9tz1&|M9)ci&nq6NqI3mnfK7Yn|D0A zB05cxzGy8sCn_E#A-Xr_ zK}h0s+@iwc8$NqTD%U{kqqfsM3L{-BzXi+sVtJbY@92psn}<=G)If~k9GZ1=MsCCf zj*-p>VXpEpw08EU^EnInqkI=^owDmY!V35kzO}5BvM!~5MIw7PC^>O~Jv5^oRwKO7 z#->8-rd@%6SP_Ac(w&fB<|H!uD7??b&d%_dIs&KRwyu25F8Sj(zHc@_7k^Ovh%~W_ zz_{_QX19g8YNr?o3tpkLMSk7ncyvinskmLSJ*wZUlvG2N^c~(Bep}>u&c^BWq4yLy z+_HJ)VJo8QActZL5D=YFH^@GbUk6|%dv<_`7S<|!BELPmH)t~_P0=7Yz}*(UH}txq z9@K0yR7G=FO)3xRP-8vAc5+Q}Pm#|l^25poymCUlVorxa%i&W4%&45i6|wwM7|@$4tfwrU}no4u{r-Th%P<+Mxmb>W=%iY|2EnD z%D@a{kw(ml`4j`%H#wNAg$g8@UM2El%6{QxVT-1WPJ&wfIFL5W4DN^b%V%KJ_#r_W zh{cQd#~Nn#(`D2>(eZ#XD#5cUx>}IOt#@YHpVc6Z5i3@`Iqz*yVNzs}87mKdC|g0Y z-Pk9pwix{)iUBt9U6l3%k_`I_=`x*hY9B$WOx(0a+mI5IN$2|MK&4_1X4$mScx1pG zhDKRs#ABb`(bq*c6_|X}5vgnQMwX#nF{@(lc~wnA0d?HIcCwE@Mi`8+63&AF_QcN7 zNy+~1WNYqCw_T3EkWI4%NOQ65J`b|RYdEJv&xWRPO*A+Kk5`D_p#G zBKX6f%`zL4vKfC^Np`xiG5Ns4m>i}U-~lP6v<;&3bd_JB{H)KBu*0FJB~3n$NSCL6 zWhq!vl$onwt{GkYgfVo4` zSrcB5y6{dmEl(}2YaJ=8aS8;d^gQ_*GE0?542F^01Ss>k(g*h}^-)fq zi0mi+cmYs`E;$V?=8jl_X2NInx(C+c)HYO|Z}>Guw+Z$_y3!ILVLu8a?C*ykfuASo z+q0eibYJ#V?ET8VtRH7zXZ7^mcRd>8C7mB&o{63?Cs&aSH=c>gEoP!ZiUET79h4SP zx`etfPg8XB;yHR@mbwNw?soGrnw=S%=#3%*wLCJiOHxetN^khJXzJ+> zfnMD#dq5gh$=sb%6eVyG#_dJ2Z9C#@44e96hvaNg`hYbFmD{?sHtP3x&_#YmdKGMU zlmwho-SE@3E8>I?V&ZuR!(BA?wnxrhRM_spHCI@nV#1^5_{08YGh$p2d6jH=Wg;nM z7Di_$#q6L+E~V{@$@l4#=0lxZ4*kCHnr}N~=QW6OxGfsw>rL?48j3{mX+@wrp4&*8mH8?b?$48*3Z#ip2kISRC|V z!KpRTIfvd73geu0^pSS93cGN<7q^ zz~a#m_IAa>_dZ9Gy^bFFR9JEwnIo;uMy$7qbd)sZD`?5aYo+KBjj3dJN*WDTWFoYD%&qd$k>n@Y)q1v!vf^V6q87i)fkkm z79wfxU0IQMmCxniPCD-Yz_uWQ(kMcH{;S{U{^!TP{K>oTNLEnH`xJ?LhGaOh{x*!q zd*0@Ow*$nO zJ;1JhH?kONR1KV^lTLElXWipHB#T#*2GQa=I!}mqS4`az{1|p}oXqB8J(%_Y!U7!@ zZ>)QB-h11ouH^b}{}y3}%Br;Py=3=jP;p}$anb@R6%=!bA_pk#9(gj)NMDIf_AC!8 z57V6_S<^78j@$lm^V1XrN76B{EdY_L-Cyn`Ox*SP6Cy8SK_34vjcl23W50q2} z+Aw4FH#Js;!b1t_~Btk-*LgOWd*H#{x_v-i#ebw(+A`>-#!w z#?OQA1fC>IClbBIkgTJawG>I9w3|X8H(rPZ$Gy@9!7@%O5a1-t?{vmLbr8Ex?A7=nYBw|V=lkrQvlH=WK_~si3&xSPCZU(fmWks7bL9OjK|wo z1QJrU-&SI_6Ga~SQ)J~UgbdcR34W6UPvtphvtq&XgseyRKZWuRI&(n4;anDrD%rnHq%3xQ!Uq&{g^40;T+ zO|~o0B^~$0m!0?qW9Kkb9cuWub77xZXSK80zIR_c2qiKjfssZr>nO68(ssxnk@J!Q zF)|l)(nX8fIj91YADqP7qUa79Z}-J+TsSnkjyID3<^0zfv%&Cs^xz`d=*AngK$1RU zX=gp+V`PScjVs!vVCEX;;+5=5cS6_=)3YA=tv>C8O%J%dFr* z{p@dk3DyK`&u0R;@ibCo0hy;XYxzTv=>`I|8ypDti%t=v6e}?+MXBVbSEcAY2UK5T z`$(QT3%K`8_vHH&ajNABzAlFlTj&fM7^h*xMn|@x!}(@&=X|pj34VXgPIB;bwn2B! z0w(noQ%jK}loof)VE!3;NfPh#D*+NNd^6}_3Z`YKc82bfE)^b{1qGVuqwW-T&W;z} zk8G8vfKXw%swBWj{!>C@l6`|eyblrj}RDj=n9JRt#5+X&aC zmSRAdql(fZcW4U;4q{U24to27o|t?nuDde-jEpWmc?$vR$n8wXz zn0YxYUU-LX%#5<{NR@nN>o;G;99FECSSsxI+7fj=A}Qhk*C~`n0eOLl}ENQGfl9uaH`qPs1sLJk!@~lXAWD~ncWlv zp0I(^ZYIYAt^ng>CVigskhGEtuRe$>mr)fF*L+daE<;r&?(%CBbjS)tNfCSIZ{U}B z_K|XWEl8lIMPW;%8R_=|qA*hDvQ$g<*sz*)Tm%r?*ci>8bobw{^ETt<|7#_SNyVG zN#y-V*MC6Nw~Klkv)d}^vW4l8 zZcr2Si5DiS!RZ2VN=yTNC>joz3Ri17X*lSNyujHc(4iDVzgGvCIa4QZW0IVy*Gf9) z$9dmrkniOH|Gr^Hfloh5ni*z%Kt2gw0UW0056|>F6^*_-^ojp9e-V9k!;Dit+mT*7 zuxQ4hhp9jOAlxi7e-P@2>=b4*l@T5EPaXX7SQ#_tzYUwabC30wk=xdR+DN-=3@Q$2 zo>v%+jBD>@iu*W)<_+oh-h0pWX7QOJf+PEf*`k@t->|chu|Kx1M)Vy(04h( zdoxtWKGUtBeqsUbdsvSPU<#w6^ z&*jobKq9CO0(zNresI;a`ye=0C+?aOhe?6E}m9T*oj z$Z(*0hnx6`ex+Juo@?g3_fZ?E7;Oj1jW-%ES&aWFia9}%IuJCZd!zfLiSt*^N!N@# zZqihW9+G#8-um!MAN;iKr^ZElzLN08$8RPrGA??jSp(rY#o`^BdKES+2p;aBPMk-t zo&>D1U|5hasWPIUTMGLG)pUbkz~jlx^N^O_#=klfGejSQ0s7egn50-)IpsKsC#{kv za--mhW0M*i)Y$fEUVOz7=-8gAetWev-b-olT6|c!#`c^ znmjic_9Ec9hAuu~8v&yPabH{U)|MroclDlQQCAg7;(TEO&_=gtnmGAFWa>on=u@hD zUN`86-WB|Poa3I25&4oN-cn&4+*K{84ZEjkR~SONB`$`Jy+C>17`PTGtbjtz`{^%v z)_E^(Yd&n`xI7HaBDlTK=zC0p<-Z;BWKYaRXam)+d?*sw9erB;Sc1+;h6?j-@g6K~ z{8-T-cQHuVMASZDW9uPSA0WBx z_JrJc$#c}g!F(k36B`^DuPfE!m83nQ`-ZRet(-Rx_|Jf?_*_XP-z;0zE}$cn#;l?CzjCI-ipV zi7MHeZGQIzCz)1JJ+L-5gV4$zZZ+K}t(aOzzmrG3qkG3dz0;!kP}L%RM^C-eq{&rp zhtBHdsF90}nl(6?p)jVY%MSVB;B09t5I#PP?Is=Jx8^+dsiW~@@Jfqj7hN6JG3%DB zUC|4>?-{Drh!#x_B!4FJdcBYdz4_%k3(iH%2=Kh*@MYLyHFZ(njap1w=QFUI``p$a zvXPe2t~l(u@e9jjjkK9uxaUbw0o%t} zJ=d9&ioH+YzJa&5H|O1F+my6z`t?_>1>@Y-5ZI7_YZJWx=2eM-Q^7C&(pAa(;f3(C zQ;`x2H6&1AeLmD6?hsuC9_q6ZcP4`-#DWR^w zI`M$byyC_oUK?o&`vE??pklv8b4l@#)X_VY`(kUk`yj!uQsnY=wDrK*3nJ@wu71Du z2b;fbO&Rk}iSz?oJm<38*#e0GhztA}}%{%xC&H~u5< zu*AH?k^k*WO=RsWvozjqu{6%47~tgDOlh%Z8B9EXAm6Vv6{$9H@}iD%F&cMC zsuw5oHmmn7{B&VHCwH7;C-z}yE9to-8cgUbf zpnOCz4HP*wEcDo;xIu53|1ef(1VzW|1T}~s%-*a%=f8Kd4uy@IfastL2*+{tffXDN zNwH|q1FJe3WgTE!FpSnH>ro^!X^MVt%!TcFdheh|-NJsa9>MClTliR+)oZKmv=MTbx}WL%I^eD42YE0a9G@ql-8N{CJ9`p|Z=ySmDI6x9p%~+s2IIqD*M~ zPvzV1no%W|{(CIxVrO^Uc)^ook)|E@8uNrA{gn1PXnbu6>;}vvL6uO38ViU(Bturs zs}v1*pniN@crC9(UN$d7l^0bLFhm3Qf#*(Pf=^nMkvfY; zDh^jtCN{BwusHug58Sd$Q=A07N(>I;2PAQB!bf${2h_UjBqavpHcQ1z#W%TIRmGwL zae_}0rxO5qk{nPk;arfO6~S(2fjAxXm+?E2;L|AU2BG^5^`J*3$J8VC1p>HUc|qp_ zc0$~_E~J<~?E_kwFcrrMpD4~u)pbNBaoSX;!reA!`+xhhCs;Guz9L?-4X#_1e1bYi zp@B~2T!=PGJ5lY{uY$je+oEjIw9%)Dp7+_DJClzG>;jI52I)dk zL$zoAC8kA#;&PW1Mk9)jAAtqH-@K2HzUUmo}+i72%Zt(2<+?U<9 zNdC{QaW{YXX<9WJhqvt947Vi+{g3Zg=33KV^aJdM0^<(O#4YDT9$8<~VfczoZDBd4?jVTp1Yh3zme;#V~Ar@WX=i9pS3@gM^ zg%P>l*1)jw{EbjLj1C%3(M>8{wV;mP$hj)n4ollMf!?oIIlcp6|MMT;0bup~-8cG` z@ST=Vb!K1O9{xN%T7*;s!yNUo~AR z>ZZ%T`1F=5BwdQ7D^DV?iOvWO{2KVI!=cskCy_OrbZHIeCRCBPXl_Ja5S0==zg@9% z;_caWkUOz@GL-rI9F%mZ<0qb_)1shWHTwF3WAGg-j`XvYtr2NuaZSqzj5o$;|tHcdFG9lp#l_Sf>JK?Gp1EiFk^23h{PZp$cdl~ zm8pj@&ER7K#`93II5QU9u$qu3WS3<72Ag5@>pjQTk{maN5uj#7{JEH7_EKawrOo5E z&&j2YVC$Fh(iB?)tEB@l_v{RNKo$>pEM;0=Z<#gVk)b^Gdd?RzL*h9bye=*{FG33O z;?Vs-rC1O6NmG=@Uh@G-s|CBb3CvcIw1Sk0kT&NNmN*MFV?oGhZCRm)itf8yY%PW8 zw(Bb!8jbamlfsk24v4ZWnX*r!%LgJ1R0+U0#WAUYh>7jdYamiLjsXI;=ooDzU2%>T zI$jHy^TYL%%r@oL&szUNR!syhr;%UX%@hN0Sx;$`zKq1CW%RA+`p`_eMRO~99VwCb zNtSvVr}jlY3GS6@amj|Hx5+rE?h-NwJSx@n-3QsMUE{&_%! z3fnzEb0@BZokHYj*cO`?HLk#5gABHVJEr>|2F(l!G0!G-UtCv4w!5*1dc;DXH--2gUge)#AFD10Fft zB>@Sun>5Rn52wPb><;{F)(m)Lk;js3>G2tLGj~9t$SrYGOcQ)ee(;vShqIbucF?QT z+b5^+8e|Y!iD`l2B!?#{`|0YhDgT(X_uZwC@s?A zXtdPTudfI$CVc4+gVy;iZ2TBEUNGB8)9LeS1hw7+Q&w|y7rEVD4W66B z_VUj~E%A4Fs&N=_c4FJ%{vB<&;->y5pLgz`db&axa&BxvSE~Pdi`Imrw}%#!hNHLwqk3!>PO)6v@2?P_vL-0FXq_n5=n=kH^wN( z)X4F+7j~k^QAfv7V=>|UpLQ&?R&w5E;iImlm;{QfqO?u_w3<`~UH(z}8$15l_-50> z-9H-dhiLjM?Rb<9KaR09Wc7t~->LUkvVqW7Q935Yzam=K0ks*af%m*J#UT7`Dul;q z8rBzKl(l*&%DOhNY1}sU>;SX%f_5wzwY#ltF#kMsl7Mes)&5a;?p=@-w0-j^1xW6T ze3EBj_%=~Y8b#Jo+9RPmr8_~dL8s^RdxO+8)UZ`VR7^c6(V_CDo>Rj+s&sOk^~K}F z4m2r51)6 z?GQML6K1Gx%}Eb}>USv`kR_Za%9A7`uwOAQb0a$-m<)+C2i#U)whh8>{`TFCVOKOfl;1=o)?61f8;;7zo8rE#^Fc@=q)p(0!;%_dNuv z9#!CXi~|Df9dk=R!}X(X+enP!HB9)IN&j$REH!^_JnPt~`NPcDL`EOgqS+dB#jkl@ z4JRq$I=5I_NAI848`C`RU^r|R)60c;4GH*LG^o7Uu6RU0;`DjB8d#2@3NYqkV3n7A z#Sm_GPj7wWP7%px=bpOp1XE|>os?0`eu@-RTBJolNj((N!<0p+G*1Q%k3Mpl;AvFF zTuO%ak{*HfVWGdSHsF91_6WK>`@OFL`5P*ap+avv=SUc$r;}gFIYq8QPQfx+hPSSK zMkYvXt_}Rmc|NYg*s(NGBi=a&R0TJms3)7T6a34H#bndxjD{ZoF(L_i2gQIcLpG(| z!+oN>Jv&2HN~#4&{P%c%r~d(&QTkcPmZ;M{&GRtrB#DC@r7fD;Fx)`SP<=oiaSy;w zE>H=;hNE8WQhPEMKsaaUNAJb<6csz;j>bBVl}$X=jaR`o(s40bvLUK8RF@akM=nWE zlMtxieI^)pM&)Csb4ndLSa==3kPwE09)nd?Fxh$9!mxaQ+ZK*Q0GqF^V($ z+J8{4D^xEL+ahpI%L^7tlGpZ;iW@d?dIGB z1}Wqo`|OU6W6#g@+@RQV!d(F+k$v-f0<$l&=={c`WVIWoqV2HoJJKm;BSlgvZ8p<_ zG)J&5zQ@;{QLczutAy01r6Iz&7MTs7WOTCMcFkm~}Bip+2n4SGNcI3T> z)_eTV2U^|O$FQOKgRBZBP6f<&?v|v1r9gp>ct;b{#xP`dB4rGp;8;`JNNK!Q#~n*L zkQ>h`HqwEHii7pAhDBnv(~`kRNQ!J%tmafq9X~JBJc@S0#pUr2Yfw%8m-u7;=6S@p zAo42NG7)O(M=le0QViJd+~FitU8C$&cqbiCDq%BpK#t*q0&$aSh?=gA?ukUTiq6O! zZv(wlSRjPEQK5Mc=QGN-=g4-ZVQ~cRACf)AGJR&$?b!h#BobulGw{jH!&wamW zxdB(xttm&@IHqnKWwk+2T1OWFHA#0+A!oU=SedWx4eIjS8<@n&4A~rX!uzJCgWIl% z7cN)!1_4!M02GbyRGz0ZLwbWg^vHXA_+L%HUdW~|3%E=$@4jlXfxgJi3|b;gmwxDx zOmclHqB}I#B)vfgxTZVn1&doW??c38__h`D+p~MPA9`#edw>b$1Hr9u{4Lmig`Du^ zu==jpc+2uDb3V84wa%;wH%*ZlnB!fnY>?N}jr5T)NXLgffC~OO((Q*-*t(s< z>>-BQ1RoPU<2?w?ZjwF;ZW9zSnEU{kB3=Lw33G5(orOM}%I>DF!Cl{Wzy)1Fvyv;Px4%lJZc{Tmda^ z07{o{x)e{X2rQq2xOy1!5VWF(xR5iX#WBcA0*oX{AjyNCO=@71vPro!bVWpNgfTi1 ziu&?7ePo|JPaV&%jDT!^F0Ps}w~4LA7=6H;2A^lG6Pch<3I6?W5Z{t^DN^Ple~vJ!E6piivyVr)PE36%p~A&0%pu+{LetJ-~I& z!+!BIaA+uA+@({lvKk8a^$Mna{oF$9P_2#a{e54(3T$P4FvH&oZxA36MU79pqLVJ- z?B`=)+>ojkRN9>v8TcUkrl@dCY?$@vIDin=pW#S@R|Fx`-zu3BXr6?24Ss(=(Yx^^ zRB15@?WGvNasj37;)65?{~}W&L3KIHcTJkjizZ;d?Sv zNScMp*)5tpx*v+lix-;SyTtf8faPaXjQPkBtiG%bJ1l?capPZ9X2@tx{Hw_&n_z!@d@uh>ViQJ zs8BHA<7~5<;G~N%1QSv4_Q9(uGen<6Yh$MEt(aY6LfaOaq{T(BE=^X$d;2{ zlk{_wcn4al{%Q_BM!8{u9T#n(O)azx#%nY}IVl z(OBh`7&8IH&0zf zAiwP-bCIj-l&z92CH-8W0L_}#CP)o5&}rc}6>vE<5G#UK&mF1>>LaJTon}Vt#?5oS z%jhSZ1u5IO$@}iF?<>u?`Qe!f%_RMmNt!%p;mZ_I45T-0r?gk*w$1=3D6)V7r6^pt zS_3~|uW?d_-=N2SP(-TmJx>xWi9~Zr!kc!czk!>3&pp9={`Wc0`hV~cX|O@d z>64v+);v}Nj^RRE4kN?)to5n1>-XQAuLsND-|4I*JGp_yjuqcy24FcrvHK}f0{N3n z3jc!JQNea;Exp5alaCs;;0(G@bXR^s+9d=nIF7$tjCqm>AZT)S1QITDyrRQ2v((7J z(MM|LsIR$q2k(<4Ikt+2oPqC8N59g}wD3k8vVp&)n%`}`6e`Q1V=RfR1D-1%4V-*l z51oki{5Vp;4LWwL(E$0<82|45ya%;E2|TUEgzQ_b(mZ;v zCqQAp9QEH^$6|R8y`MY^ECxvADA8iV)NUlf-r-uzyXKy)#Haz>kJR=3@(zf#-sPW| z+e7ClJ0(L_+rO+19_wId%JRKyDrV`y^R2?)zD~B$utj7;ahEfsS}wny=pLIqbLEqfs^letk5fwS<3IUfAz9 z?1v`~%3~oldw`eBIYemAT37`D=6n&yF{bsGaboOK15cEd8pGuaVo{p_*Gkbkr^|_b3 zR5{g#UVnaAZYx}Am)lZsam0+-!~Ye*i03;LvyEhOjlR>tqHzUSM1v^A_1Q3;DEJIc+Rcl8XQP)@(f_l<~1OWE(dn){gC;Lk0laMX?1G zu=7WrpV8|WBSOc09(`eEDWh%{luPsGCwrluS`9RV#~gJXnu|P(JoD%}-a2_F@JV9n zSfhATK!FI9-1U&fu)`I&+oZNKn9W0%v0-EI`Dq}T@U>sQQFzff#>_<#v16mhM5r7| z%XLm$wXVlbuVbw8L^Fv(KUb-X_`Rt9QOZGsR}! zJ8hBzmldKFqAa&gg{}|{&Z8ZSwPrownX@TYCWaE795H*zCou58-t1M)(W4~#t(kEo zgBwchHw%YU8K7hj#THQ{pNh;B9QJD`$(;Ci|;Os61AuX$zeadhG%v%O19TFdkE^wf|!iv3WsNZXnLB4;13ldOo5UF#NaZXiG z9DsLV1t@Y2q%bkIA{n;)DL?y(O=JW&?a^-gBCmh#Vv9&GUfwF5{#O!b$CeK0rDL|0 ztrWWjrCX%d!qIu*M;x$+MzaF*Wzl5@e1BL|mN$PB5C{)BueKhW#j`NO9kT z;dZLq%vatq4lRE{v{(~C=Uu^Aj<`pT^-r~-#XK-#kWVE+kN~y~dg*~^C2xKyT}oq@ zNo^?_2BWL_+uU>LT4@jc2KID1qkJgo4&LUQCR3x+zc_paSVN-|Td_)wJIV|!AL9?! zff&rPvt`AwablB+jG!3NNpcIQNbo7PN`a^*2a-QJU`y!`f)*H}<+j0(kDD0az!c^^ zN60wCC4je^&YOQ)kqIYKrd#pZj(Hf?c+6v_v)Uo2m)U+T$|6s`;7$D|^iE>T3*-hj z9oK&UWn_&(tL;IGy+@HgUHhjDdXHdL1a`2+pdXk*aUSOv?+z{$UFCHJXU|35tAb9} zMg-3ONVTv_i34tnIL?%ktbU-^CtqZjs{@t?+_g1L92VzY79%N08k%*!KR} z4-c_==sABnaEWZ5Mj-q@M$O2hSZJ@51!{(8foGv8&vUo<29R0cV>n1hhown7!zv*o z37uePBxfWbwM21H6}G(6r&hi^mre*+?e-{e*x?9&-E&o}Z7dO;Jh4bfSxqE8IT0rz zWxwX?)+OivX`Fq@MHRB&f0?PgOr(WN;MIXlxhNiXK*`b=6wkirxGCUZaIf5!P<|3b z=7teKqZ#jJ8=gCG=Zu4KtcVLP>^LxBBC@m^XnWohH7klJRK;{`QIm}Z?Sh&k@oc#A zCZ`^L;X4vba^>t;RG7$>TRSbAikX(eToo0&bjqRRnU@s$NRUDA55}|tjbhXEVo1GB z2UgR1zuU87Lr2GH6xkF81hB>k=UDAg=Y`cuG1KyS#|WM|=UBN=cgCs^GY2}x){&{t zCUG!<3y%5AUpMZ03v`ZS^pVXJn@o`n7~sW-u3-{lGp|Y3tjMEtX-I}Hb}F7s1|4pX zbOO(2g12zeKkQc;Uw&{j%s5b5P}H|x&%lYhkrQe7f*lJjraL9QQu`mh{GchqbeA0e@ zK`=4fb;_V6l%#g7shXf9&1A25v@yD$I{y4p@yGn@z=56W{jN05@uIkvG^;lAQoS=J z`y`J9=T!+jt7@L6AZjMk4R#+>b6N9z#P_?u;i(rJkJyv3L}SMh!b$^qwwq!LDYBD_ z+~T)|-B0?NbgyRB#&5j)z0_}9`PNOKKIo%Qe7h0&J=O>wd1ER;LO_4$c1bS(CIL6z z*V-g|wJ#gzAK0y;Y9cWL z-RWt(nO=UvkwwfXR*&Rcw3h$6G~^!`4_0G9xRbf;w|QQZ zEQgK_Jq?j#^=iRtL6x?(2U<(F1UUgw3pX#+@+YS=eRGjYyVUKNuXb!TA8W{PW}J6o z>X{XAWGI8i`81LJh#QU`{B*Y}K#!x?l)k-W7q<|i9V;>?4NzA_u}}z9PDTFf8gnrG zo_rXXBh$Sq!|xMhDXen(`+brl+Tq$L%X2FAZG?KUG@lH5wOjv;etuC%VQ7YTnzT)L z)(>KZf(z1e-Y&OPK04)eN3~Gb3RMI0b8d(Iq6H`E8bJ@OanDp1^0r&42eS>(HW6?= zaZmEKo39&p#Qj`xD41y9yD6X}5Lq?UQ41>-`#hTQQ#10?h?;-GN(dv zyeelNnsttjodn($a}Y7kR$}w`o@qpkS8sYJ=T9Qz)RiaYh8^!NCK6OKpU>xGlo#rG z!eRr8cpv#z`}WW`pTAEu3OqkR>J-hYW8nqgFMhFoX?(={5t^6Fz@lnVMa|C*UQg-; zcil9k*yXM?pFhmR33k#^{^?%B4)2kTkaHa6KP)fdHMt#La*~%1#fVh_2j(2`Y!s%> zIWfOVaEIPauUWXB>6ACC9(+CbrQ=^IToxZuwd{OE&R24m9{W*?YWI)gBX)g9^I~Ge zc~zBDf?%&Q#t&y&&5IK>tJ;5}8!@u%lZfbuAA^adbU%?~ay zH0vh}$8j-g*~zi8&omqx{wzBks8@Niys}ft?k|kWbIL&Fsis&cQ9A&%3NN)uHj+;H zF1J-crhI`EGga((=Tex_QBnTI-)@s&c2k;96t&HL^sW{arWB_6+<{!f-E^*3l<2$+ z*>!Q;zdjpgovqRa&>D-72_HxOYU;bjxleX$sGCU2>JZj|{{~ZS&&vu!H_uxs(0DZm z6bbH!uAX*ucIQGod(1P-^)T@Ho}+7NOx9XEErs8z!n6Op0e#G^v&be^!IHH%F{M>b zn-(kM>~hq;y65YSk-{a%#g6ws6S%5ayw7_j_4W#C4X;I&?9wK@zNlB+OB%v^d|FhO zmK1A$R|2)fQDI?+oc6QL~y9wer>T>v%PDdZATW@627&yTYq}B>;=# z^vAw=(sy~iq*0jf{sDLeS5M1ZirdM3@&R){6!|O%$gq59#u5JNX^p~O_iFKZ_Wa_D z^vv{R^^+9W@jqO3!+KC^<<67IZ;nb&INW z<{?EKf3**O*RVq$iJOi`RZhKdXNHeCaI-`jO<1KdLtb%{Mho8v{-tp#zTNIyCW`Bk z;SwFk|Io5bI0(@>m`CT(YOJnK^y+hm^zR#>s%!}P zfWSS-!aFGKa6Rn%uS6!1v38P#74c#Q3d|qE2`^K&{`KQu7;^%dAk%4K)5H3~ky{ip z#K#^Cha=h~=jcp9ivrJU;OB9L`i$Edw+?;>AM;cSMHmyn%lDyJXx$=Yddq+|O8LAT zE45{K@~K$7Do;*nl{4H{#cuN?I-02Gf1uyOTs;0Tg*0-rtL(TM<+j0w_aVhzp~xjF zvKwShj(arlcn{0f7~{MP9St;4Tcv9q4aFsQLpsFuymxsOzO@hr?U(QM%~z()h;$nvO((iGxvEczG4z2v zBDI3N`Ot>^@i7)wd_Xd2-5)gKFywj=8lI?+k~G&9s#CLpf3rq#)bmLF{2M&@6VbB6;V{R-z=HvNdIIvE-)-8dj9n=9-P&jP#PIcOh6Usr@Auur_Gsqk} zh1&~grDx6UOU)R~>3z3ft)c9@q0UaHpWuQz&9J-&yaD&*{qS?w zj8diginv2~8A|E$=ze~V5~=x)@}flC%hjy}bu$Ew8TcMEgd5a-0@wY_I1`zRJJEjM z6;lb-jq*)C+Aw@i;3JRK9tR~@%l!=Hbw-IhEqIqaDTxJa(%Rn@Gvo z%v&MKq1TYr0^|?bJ+H(+!{xLBYWvZhgEf;_45vn3XRHO+m$@YK<5~3aRfC>#FL1=6 zLUN1I70nfhkomjOxm=n-pHg%Pb>#7@{Ys&iWTxYwwI~(MLxdS3W6TgU7yYTjxXF~* zZ~o{3Prpn3@x6iLWR)H7QdtJO)Fz5eqDVXyiBvJVm?XLe!4lW?8o(rtiRJb^_4z*i zqS>f;`ZQL*$_W)y*+(fKIO|bS@$xI@$a-$R96L6mfLeZxrjt&w?^3X4(0aC--sjm5 z%?FymO`Hj-4*01W=va{e3kf-pDm*4A+6Icasl_%MOF9W+Y8tDbAI|Hn735sg-h1sN4)-^ki9sq{{noQ@wG29(k&L5Pv&+dLfqtNmgN&D!ll9I(E4mxL# zt}Was6D$%-=g~vXP-US}qBEyedd#s4G^1w4SUq}Tz#F5vSoiQTfH?tg>fCRHYz)=| zPwW`JgY4zDZP{_Y#Tf%JR!gzK0(+2(EO8%%B#J1}K*-@mJLWdV;8UFPfKjhM zaL9QbNr25Po-GMV^n$25nB==&XbQ#_wX6I$^S1CHe!$0U_X?L@_XZkA>*}_$m&CJG z0Ylm|8U<2K+z)N@>k^&=52ID&pe+FR7&R6gT*>U6y<88W@S3n7(*A{k(1-zq?ocd* zwQf<7tNcscGZooON<(XDY|_2TKhq_Ox=%16W1DZ)oJ0nicnglX{Ux7D`1W6}{<185 z6?IaOGixQ|1wEJ7-l=fkfZ#<&{K*Jl#PCC7T%umK7TT?Vz<=?J*Tf+6& z_*VUo_L9o+1cn`l11=dL;1tEaN0DRbq(OrW8CbFN_2s!$0yRnyw2IUjE;W!rrje=( zTy{VXTKtS$U}ptyU?Z6Bm^HWqtgB?ND^BWl(_xsRhF3K^1q}ZT`cp8~)9v1lj5 zr6Rc0Q#)7Q{B3iz(ezdReRP+w2?{)uy^bl>*+gd)c7ny1HD|01Wu{2r1R4v>Dl`9_ z$HPpEn#vXb`krwl^9!P%n238`^iAU3k*pHclFyD(nDZX_z))LDO2J6F$m^q1f*OUj zkUCKmg0;liOD^$Ni8lH-DRP7JgDZVjC|guRVS|#vz&?6(mKGHz$ZuM9HTR`8jD}mM; zV|17G6dOa4)l}rsmx|^UyR7@kop2lC@EP!U227uEx9u)-g6Wjb->yym&oY+NfZBD4 z-X$IV-K@%+miAJ8MC!CtUmbQhuS#Kb88hWeb3`?U)w#A?G&9>%j?rIi__vImnypx) zru4e+BiBb~$gl%9fZ4BX*=itL5-4^(MPjJPDyLgZQk^n^h4QS&HTRgw|0%B!y7<*f2c4?LJ>n{-ee$CoJ>q2NyV6I`S5PN; zsr-|q9~zr<2=Ph|-N}4H-wMc_Do^Qq}x4J%f-- z#;nUV_wycEE~)$^(& zb^|aZQ|tzcBw%v}q^{4Sp+Q3jG)UFf9pusPN^b?026tO0^T7PYFdqWu&uNX3rXetn zUuVj1NB-NQj_P(CT{h8C9Si*-y6n^@dFWS4M~Uj7sJunBGkBkTn_H$6_MU_maiN_| z9Tc=2kl?w=5GWIBm0%6HJadls{+zMTxOs}*>cb|wx?l`Qm$O?#;%D?g#@yC#eH#9k zJ>hT?lJ9OkXDueqXFg5c0P|(XC}@1rDz8TQ8A zdU)Pt9&F`C5zmydmqKnP_5Q8P0iyU{Q%tDw{5j3%(rM^39?zsd9)?tg2#WYnnn z_lwOW(T+6^6Pv}uu%4MYbP_Wx-w@hBpM_u?Fa*~_MiOw{0gcfUmyW`v$=EKqLl*#S zBe;&8mJ@WT_P;HPdO?qniur$COA5J}y>^_V04WJ$B0YyF76O?2smNMkgEj{|Nu*1` zMrGu-4BIg$33_ChcYW}MgOIunYKRNwGGsv6cY#Z@;+z;NVxfx@9%1QxPv8eU%yPig z@ev2C?;i|o58UUN!n8_P&)ec>O&K~NteLia$HK*;<6jIY<{P)e=8|||$LKN9)f>5x z^5{!T!0JUZ7~D$1*Ri-vuEw-n?AU#oKp7HA2J84E=ktqWpx8Qt)+N#vxZru@hruiA zQ?qR(rYuK`MW8TwvroTom~s0^6MYv~PdoYi?it<6ZSHA4HzfB0*FX=(ov!C+T;m<` z&8E-1u#((?td8n{Ca@b;I{4+CA zv8ngNRC*)@-{Po?8R<6F~JdZ|NWWdZZ;x4EC$in_^-p5J3q(G72Rxg(( zBMRJfq+Wvrje=fK`!6lo2TX(Noph71)D!K77FLaiIdtai7*UL4C$k(`|4%Plj;+>8 z#c;VEsJ?au<1(sq=vt{pfpdXimmZysF7-{*d&Gx*tLe<3vmVx{Zl1^@ja{(uk1c;& zIW3a?+6>?K{hwOnq*=QaFeb8jvE1vd*12-lV?+W@dPyt0MTyEku1?30h*aR006fZAZoWT1)dR4jHW@5)u6kCqh@n_(KV(No}8I{I0 zV|EJ`6ZK&=(>3m6qrdG8(ggpnbYe0O6yqUekUVn!~ar4I6 zvD&IJ@Wv%kY&=EQL61QtazUxHoWa#L>yzN*#YnJgGtu#sl{O1NPr2V#ms%>LQj9HM zyOk3rY}qLPcRliXeS&B~CzCoS-gUpY0hlu4relM?6;92nm29UgHat29-ke(c7_=ui z;daIsKYb{{`T{(gbapP z3-3B(9>6WZddLT;rh6kA<$Iy3s}sUN9W;szyt2a=EuLPslO1+w681Va13of&w|&+) zqb;=8O8I8vE~{l=8Fxynoa6?Uu9Bps#^i*bt1fRQ2n(@qQudPmkWA4A!C7sCz#R8< zuM(g!%y#Z@-3F!`B;PVcA!$DU#KTfw{Q03|bVxh1R&bKsTmU|}pX{KnN;*gmgojp7 z`)t}Z)37BIUg@2AR%RuMWErN$lT22D)KhHeqK|@}u$S$2b24GbchTj(eKZnVLNW4c zL4l}4n8@@nD@1CH3}(<5NT2(*&|K(ooX4+*m>_ULYm{rS9d-(XQ9q=>T`4N@sSbtv z^Lbbf*&)nQ9)dLLL*CGs3*4_6P~y06?vR`MBW0P)8p)B?M+&0Ha#IXw!A_FhK-!!F6z@^~S>&N)n|qY#P$6NwSsbO{M5Op~nF344+p_C-2x6g^0!p-Vr&?D{?O^5^&X4>{lRsDZPC zQf==Nr~ygySnsS}@!XK>ZCT#@Mp=tZS|VEoGM0;)CuZOTnW^DPBZtgPoNG})_j+x3 zwSsFhBd?aF->o_Jt3Il*le>^9!~%0C;H8K>%V zi8tG^BgRBxzfSfXo_j^CO%KIL;!0?xjX|U$4-0o0&PX#G-JEB(?1qvTgC80@ZB6(_ zF;VfK*O-I=to{B3I!Ax#-8K74z+mWPKTEUI+AVuD6S$j`tgzoIhf`!$p_5*X{k!_Y z|3K?qyOnDs*FcSZmts>WvJpB^YZD-KG9ZojLHH>}8Yr7}5m8}a)EtDqX(PT-Vb%yi znhO)NC4koKHFs9a$8o~!lz@zSv9a^kgpagHP)kE?ZbkV0P^kAU^%!v64J&ZpGlzc; z%+ULuBeELz7FB^ufvDGU`AyG3&xdX+MZ3kzTU5=wv7bSE6o1TMbGk$?>x&w0%c5PI z@qSpo4n{L@S{C~iC7#P_-khfwB@Qmia*|`mq67@!G5OW|DYk?ndmz7BdVSF`WqH^T z^zyyyaWHg)OPMsoM|bY#0*#{1^D-2FoeNkYShH}i?=aZ752UNZYW=!jDDXy)U7Vl_ zD%BgM*PYOcxx_y_w@a8L*rGfR+}K@~+Pg-uF?(+u55$J&o|xIP6Wjnj^hQzxe-t-9 z4lBv(FHGEUhXKk`D0U-75~;`{ejmMybP{zIKNg&Gb)MSXH#EgMnNR67Yq@02Ttp6H z(aagGGwe%W>nt%&c{7n&7UiE3v@1AEnajUPZpgZV`=Jc%Sa6Fxo#~s~&da4gV0Vkl zq%oq+E(rly$_vtR-bak}_m^@0*&Lqj<2Q1gX{>&V+Z$@P8sWBYZLadwi;npj|K31y zxrq+@O~)cf4May7#gR`d(m?W7CX%p4kvA+E4a* zsyD!g(hz8q=ey4{UxKY7)c7Wp$+xqu(%sInu6y{~+|}JQ&hs8Ac1L@;0my$!qfeu2Y-|K6e$ zrbRitaEB|jP%H)aYcBL(#h>arK@N?K81bR$VcN8n9H=JigDohR#;&1AiC_aKptV3? zwbj>8#7PKFN&Gfj`V+mZ(cBU}BIoT`y@fV2W7OL&itV6C8(PZ9LR=&M$NMf|8pkqi zk^-0NfNJr`mzw6li6$93=rQ1jg&Q4owOFHUnxk&!m3idNzbicFhOEcX)94gbVc+mS z=9=o-!gdGhD06UH-Jm$Pra9GO9ilKg37yI@JgKwKpaD`yl zp+|gu(N4O??|=la)ra&EtmJN$rUvczRqOa$&X9eIK|$0^(*U&=&YqMzW}YztP)<0T zy8bs2hmG6J*lj$>MB|u^PIsJ3$*1IFX0tRiXn8q*fNt=~P;8yq8F(WQ6Q}OLXFSex ziHFWHIgmEB!f8DZ`~gQvGqf;Xo<28y?tSubes0L}vo0;FHLmRerxXYLbLd3=hS24Z zT!npTn9VMimaG&Fl56gbj{DrB1uI0BDs?u4kO^p*dzur3s2AM!yI9g-&whOjB;P== zJ#MhqCs`pHlwd!5Y}L3Dxclz|LI%lUQYWb6;n_SsmR;T_mLkO572|B3n?JFIBfBK6 zVN>)QnQ}+@U&$kG9Hw?09!fVrM?A%@qsUq+5>vhP+YiQvCJzDzEA?#IVlaHtKDfSo z@uFbVh}!bK<75ptBg&5PkZpj6REph1kt8bexMYCzLS4H?p~k-AWsA>9^5~jHYCPL2 zMYK#RM4k>yWXhUZtgzKS8+_#Xo4=Hu2EL%v$ z;v8++;ztn)5m*AJQJiE)0^e?ieBMm~70}dkK%5N9^V_N6wBxe8}S|u48h7e z7=q**M_ECRW;xinAm}qJJ@d$7u)Q!#^_nrAy4~usCc25WO4o{RG3wHg%Or1p zoLeTZR@w<2Ik&p?fH4Jmm2{;NtQ8M1wiUxx~&3MwQjpl zSxIw2^@MzR%AD4?^xaRp>n({n+c8W`#GD%-rLRR*6;R_}C)oFVElJ~z9u4vPtxG3m z#sZm*3p3`j=8)aN%5TW^;v-|$@Bc_r>{uazh~OB3kVCPM9h*T#o?pB>c%>*`0uFto zYCB2l=0bP)M+~sGWh?XOLeXl!90&&WEd>1{g-IYcf=h$zJlmOk;9J68hNd0t)_Brp zuly4t0Zv%s4=^zYr)_4+{9~(M{=a&h9Q)R7FVbnpI9WM$EKcrH>>Y~SrXu^90`Kk5 z)j^~C=u zKv=TZcJ>OWsW`4hRVYf}4NAAO&8jyqzLfjYy%*z_?xc3UeDgHt?k@SAtfsF{u5d<5EuWY6zq8eCxXo>$_Rtyx;}wRW6mbkxXm+4CQp))v8TJSbXt zZWM)_U`NFt-FwBjEwA0Gye1m$Ldhc7Hb^siLaOc-mO!hHT{B9RtBcgv{je=J_F1W(% z84|{7Xv=%6V1QT{Whg zZtBeaZkLuklpOp~4{TOfK&4Iz!Y{z4x_?)1dgbwU!+CM63=U)LaQ=HnZ{Q?@sK5Pu z&)q;hl(M|CQ^{^ShSDhmC{ZoXq7%YJ>KhPS zK%U$VA)09zqr5NgqTk46)c8AwN@3ES)Rlgfe$5JPVpOjqo(BW(Q@T)52ryJ9yQGWs zvsXGHw`Mxfc>gC!riPp6+J2)4(dun0=a9B((bzH^lv-pzzQkA8_@zL1pwcjCLWO2{51qf{s6ag!nBX@^ zP$uNjHy5Z6`JR#V^RwnOyM6Md(y%Tmh7U#@cDkOU)hLK?-tNFAS%M4xRoET{{~5sR z&d+pg3dX`$gB7=pbTHm!%=O<2yp3IBn;x41Y@2?+smW*VtnM=|Ua?zC-9%MQ2K^}?Md9Ju>s{-WhiB{B z=N+Dn_WR-4*hU1Ir((TP64R)+7cNy%T7bzAfg_KGPWNA`Vb-I9o ztTmv(WyBY`V4_6*LUpQZBlN22gq=2N-k?;w&*p${I)y=`X%zXvB|LQ(jd?U((u4Ft z`il69SUut!KLe7&0}y!~!drrzfc}uRZbsZZFC@OJRUvWCE=%973 za5KG06G7G!P~&#OEC6ytmP_Kig_aP2*zZP5Ob~!nD)xD{0#fdVfSPs85%r;BUMpKf z;eV68Gnow{vzFd^gI;ZYn9Aw}j`ODjm&j&rv2i&IW_P`>LsI z)KqXpMV7b^IzuR3xN&I_)ezA@uboyN)-?Blydf-RTE1HeZptT@`8A6%Yf{izcHzXS>91H)>tM$yG*RoYhd&4(`ba0Qh`)JWi>gkL#qW}r`KJP2_wYNRL(uAJxBTyg_A@2^Yo;~I zY6Wp_2?6`P5yP?G-LiIf4Jlxw^iLRQb)>$Fl7%cL`E=+s5x*$p9Uy!JYE zD+{1`C>FMiIH1S;THvRFSQ=C8lfoo=Mf3Ut+X*U9yMqTp8pS)njUC6^640CiqDTQv1Pb4b^H-Mm8S1N<9_CYAU$%K-1H{yTFDzAezO3HRaY1SD-VkkB+kTEf?0DqgrkLfkG_fbpO6nRy$egjN}&h` z=fL=U9(`0^<$BRK4V0!#Ns?frM~tY#rE7LIaHCYyNSlxT#%yJQD38`CTX^Re*ZL$Y zb7&07B?O$G+bis$k(w}T&V~@Yh9d2XxEJyPu<;$TU?c?8E$IQbXAiCMiVw;RItpSh zZ$1|JWJ5<~G&|7%)!mKX40)H~ZIE>v4I8W7YDU~-)vfi{_PXm~*!=6VH^`RHSpXl3 zy2d0`=Tj_@j_#l$G2R8hQes*xa3|?vzaNKPk$gfrgt5Z@kPm}8nNpA4V%)UU?XY8E zDm7M-LR*DFQq9Nh4INQjwl7Y?z*51)>B~3_HjSUOtLNvr=&_O|Ol>C#(@2rQHCTtVa3J#k`BgX!wAW1`3ei4gTd5 zc;*K8+v4}f8#1xZ@g z2Af}_@L=c)Q6t=O*{_e(@m9NC;-}3xL&kd-C-CN&*1RWrWt`V<`p_4k8tc@9- zTj04$P!Qa$(j8?vcB=|J*Yj2hc6@2PwHxQLS?!eBxUkwcnm18&Ly7Agy(o%)Yi1nD z;3kUf*aoXI5Jh_^h?)Y;5-8)zUPxH6F`_M8eZeg==vYMB_cV%f@1p{Z*VBeUUsC81qdNJkN|I&A1{q|jj?sFHDHc+K3aLnqSEF=4=?Ltju_MF^>KgNpdva_p z*P`m0UGKMbMuD;!LIEdDjC$=a z$7g-syD^|jzFQk}N3$9`4MGGJTU(Zf=p(SmlVJl0n0h(62?$fOSpPgHXiSO!%1<1O zOaHzgDW)b01yg5cxu_e3n72{3p+6yim487vC8}hCv4@%sSWXo!0 zLp0twef)rN=Y=l_Hca$sIOJQ*FQU{Xi(;qWTu=-SKV;HxcH7_rr4V^8d-%x7bYjLn zNw+Xfx(|}}6L?TA(H(e$w`Fd(^0T|j`4=USXQ)wB`&A1|*|8P9(?D%Wqu9+9Nv0xWnHZ*0*d}RKHGHK7JgjJ< zo)vd1PbtypZIfhyv0Ev8-(~V7#}sZ5Fnqmj`N{MqG+o~QLWCY13*UafgH&;|c0;O|d5^(g4k0AU~ogWZ0p=JJziz1UtD$iH>5pX)|j|DAwXVRnY_-D zq-6VbQ<_Rj)}|a5N)ADqO|q97gG9Id(}Ms-B1|%z*y>K`(c?|#!(zf;F`*l;GvC}| zOj}^L>4GMR2x_In@<#bS_pOSH;?j_Qe&xaepN^mwRlghw(9h9@kh--!bh|W%uJYOu zbePooC3<8s`{li`8$I&IyEX_OLR!{J_Fib5QyPpebHec(1RrRxSQ#qjL&3!R|M>KT zR@~%GruU+8q88NlJZHDx>-VRdixcOOo72czgSheoiXEUxKNb1FOtj17q)WW#1 z9&K(3n4_3n(!*<)<y6DdcO=mmM4PPg2!>BmtBv0f4PxLR4|< zKf7U)j<~_qs9TW?ErrW5<3vTZ%(rP(aEP*~g zjY8vfd2Z2ckcQQe9@7?Z4J!83I!^UB-bnW9gfzcaX}3Jd(dLC&W@2hwv1UtnC2BUO z7*_^!an{(eU1g$tuY?d?$w|mus!c*tlBs}bw>;=`u^nPGSiDw5EWX(x(%bro^ zxk8|U>d)&7%O|Ird?FCBB{W9o&iVOU55A@sBvboe`wkJ@dK!VB(Gxt6R zBBR773390R@_OlnfJ|OyP?r>^yX*Xjc?15JN0-aGq&hy!t%`>d;7z&Uc6T;Ppmh*w zQkVFgQed*(4WQTPT>21_>`J8SA?FVvmvG2=&B76f!k})tFerzvm6}r;JQ1XfD_J<5 z9L0m_-<&q6DS!CEl2S`Vj~xfGO^BW><;cQeIi{!A@mf5hWf+`_0y>;GGUx(+1>G-v zU(x3S`uT~um|Wg3e4|ge%X5d5Igl{DDCS(5D4@)l+mbOt+3R`6OeS_~j+@{nLD4gY zV-gjV$yaG%5zJX*OV(%Q|G++@S6ep%_ugdZ!g*Ph0$tH8f;xv6nl^&RE|4i9&qW?kNfgq>v>HjCkg>m2iXS1ZW$}S9ohL!uAA_VoeC%Bp>xUA-&jnnRfgD3d6jJ@sq8*5jI z^!RzJboyUO95?*fvA$bifS;`ty9NK)$PbtC zv}Awi4&LOU37aGcn~B>o8-;1CS@Q}v+-`blah0EbPbujsUPp`L#T7Jb%J>Hr^H8I-w|d6a_DJ>szZ){<1G%-L4M8h12}0oFK& z)a=MlTwvlu@cCAzxprnvuA>SnjTq`b>K3L_Hpv`8=Ub4$=iy!Ep zxk2Z|X!>9X>$a?p*CE^x)CA~>3Cah;iF8p8t*c#3llB8Ic9Q6{zjd>m3>ME)WEf0l z#Sf;DF;0we>bL(Z`OgwPTIBzJv6&>=Fnt5-ue`PBOrN^B@n#$RCbVKn~z zwmQ4;mHDsE(uf!Q;JsC?hUl=b*r2)8hQ!F&v-3d*&p>{O! zD`$VQj+~%7h{mfqph&P%FtDJX!Jp;)bHFJyIHOd4Pu|1COoI-GOtV||l6bZ%V2B=b zpJH^3ZO~J1@KP#9V#?GHiHj85KRzaoAglo&PvCsCt(mCuMn4xKvlBI)wc zC@zyB=YA+aEPehy0m^-s+#uWRKUR=2R`I18Ly4P=dGv6)=s&B32u)$_29_I?B*3yW zto7&tFFjDfig$~V3^Wx8&uky=d%SQQk@WaJTOS1cHgU!0-7ePo+6uti(aG$aJLIO` zS+>nq(V&s%EbgT>|2!7%GOsEPW() z(zYX%X(lxnk`p#;goE*><%W&r2_wHY)@1D#8zyv9%pSmy?~t=Lu8TQCxRu<745e1- zBev!_^?*+lxLPylqmp_*jS@0QbRAP`79|Ku#p*-epGZ2PLI_KqdmTHO_|SA)IQTgM z=1EbaKQrEvTXv2cV3MX(e0!Q+lT~%Sr6wz;K^x<-2I&ThO`ynn{oeq$dDyl_ZV0!RVEL?LfySV)+nvS4GGqP%`od7Z zNcO6b++wQv-E@bOjl366d%)%`^V4Rq%{5LUY$}ueZkCT;QZy{xSV6MCFp}c1fuz_+ zu~3Dvn~F^HiK6y-mh%g|tHpO0ruqDPpJ$0rw}-lc9u`!KbAhyJn7JzIqPu+H6gGKJ z^tvJY*jL>EPF$o(SmV3{?ggTRTEG0c`MeVM20GVQjRW&Ab_gt$Rs^GYATz0(7+tU) z;Qr6H*H;>IDw`l9MzgZY3A;MBN|OZ8`(v$Bv;bXKIy+(8>=6efr5qr)ebzYRyA8?f zlO(z)`gE=L9LJxD(<^MhA+Xy^R($_Wy=-|WG3Etw!;XEdYYb{%2PyU*Mf#}7UiX#r z9(w40Hmf!{rUzYDy?Ku+`$_ljx_)=%cb&f*{89Vw-b99sBjHE@*~b)Xf3Fp6a@3_M zKJ-WsoP)g87S&<0lRoS><&}%e zoc4vj?|zORB%2&#{5s`#=hi`8hv9c>6us`pX&jrY#L+FPG@on;aDOpMU#WkWe7EZJ zUKZa4`OOeKte{tj3S5!|yPz$tx=B{H_>81xy2f3d>69QyVUCk@$Wc+>m1EL$vtmEl z=c!TL@^6&yo^fdj@+Dw=S$&k~E+_DMz-x=#gOz@3;67`r#}gH%Mtjug$MGfO1kQ32 zrc?i^US9bnz07L;+Ra?DkDESZ$LZEkST@FTI!dv16sd+rUrxJ8Dloubx&9J zchZ}mj}_(eL8eIG6K!@m>ypWKc`Qew_I~JyS|e?sQ|DaMeubr~l9fR>3OFk$ej-(Js`tCnCdB~~8@f+|XCDDY<-5{8`tDcSB$lqD+3z*i1vh(Q z3kPzNGL&-5jb9lD08IGV50g3|jBO4$;D1~(EboA7PE-?yUOV^o_=tzUIlHVsyjgW# zHEGWu9uub}u<)^Rxt-e**sbsG>s_y=eBMb+`{nUL?Ibe@9Q)7?WPYA=hcJWg3RYv6 zk3L!Nmv)PD=$dhf&ABW%hq0V6JBhA2PMDqYSBwBu4ffQiQ2{2narSQUYAgR3L`&ffl^ZAwbv~V8h<o{9KCYb&sInnXLY!NW1Ipy@6`7-aBw4W-dlIL6fqg8*4xP}M0~{n3RJ#m`$ux=z->5W!bN>7XJxrGFs-Hp>aL$o|kcn@~6I>N^^DQ!_Bd?9_v_l)FSUWG5VY0@Uz zg>=)}GBo7Z!|svW-1BIpuRFqrFJ38X;3e<|rC9lneVsQ$;N>KsyQ7O#f+2An3kN6?k%Wge>Q@*o3$XO4*ikDwGN7mc1 zwgVaeW1K7L6#Fg(FEwOb^fTH**2_RKI9i@q9dPgk?b>J1L;Q7%@aL}J%;`3lGWwl| zv#_zzj%7S8ryXqSiiKZyv7~ayj#C^=R1WQME#_TwzqCZ%Ba4+mZS=c~j|$3xYJJqBkx^~m`9 zJDrteCpTo+vB)@PfQ$nayPqN@R3y@`r!c1!X)`JoC%6{QFRl*27S?I9G+Aztx|O*bQXPPutndU7aqH5nZp%`bcKszQ zTH>69%4d48$)hJhFtIc3-!0B%^w8P7!S4f-$_+YpEL4CxVT|UIOR+mBl1W9b5+w3> z&$tzce!Cj-326^&Q6W$AYC*2DJ?x&Oz~!nkNzg+#2nHbzP`NM%G|_d7I_NTwbhi&7 zRB2<8GjX6y3;^cdRh;^ZKbpzMbCT=|Uhk?#8lG(B%>|$lVx1>)Ut95oGV|{?`*E5% zpS6Z@(+Ta?r@!`4@l|7Yi{1L-OgK;)9dkqpyuMI%qhmE)Ai@u=Oke1T1OAsIisKJE z#EJKah8=4B2H@vj(O&p@$*)_8BW)}VxGAG!cMEF))x zs3o@ofzJf|4%K{2k8Yq3xqTqLA+zbCm@QGpu7>{D#KW}8X-+F*zuUIx_ra|}dZ@_% zrrt*Oa8pR^*fV$90P!^x3spuH$fj_Ty~N)Znj2K$q&~@3(?>)p(;-)M$T>aiIFt&d zF#q1>t4+b3UyD5_R*H}US)-_OS}D32hHdbW@Ci<@72RUMjn?QmvF;gzp zGJ_&b9J)1IrO@f4n2!m?l~5Po1+G$?O1m)&!Fj#M+-u82PJ(aBC;!^^=l|Brmzdau z#iVB%S#RJne?+lE6uFNH^%;;j+D%^|TYNL7YZTaE1<6_~*tJ1DGyvtU@%s6vfa*7jX z>{oLLX5aDryk|~}u&+P_5j#uPn(Mt>Fw1prOU84< z?6$s7UEKBCS@W;U-XL4-IF?v$paADnY%WE1Ag}rsps&0f4l3>JG^xg`WPWMLVW&4R zr*2aKY=%^O_#3!yZ6R;8CL9082NB9TnkHMnaMLJ1( zSTs8KIELjMES3Qbc#x0srWr!qbwa$M=>oT&MimzDXlw|=bSrN zpGb<$15e0trYlalL0Ip-h26u?qaS(Xc#SyV{JPt>%>URI{!%3x2iR2>>NG)_4BR#2 z2;Xk=9_k&61|I2CdzM$7{{|Uy-~?d%Rq6;NMU06xtfSbq6p5xH2YiyfP}hGO{x_+c z+h;Q4*`n*Sv6Ff9gOe_cKkruhCb#>tK2kGBjVw{xnzk`zZ27!FAabwc;me-Fj|sE7 z(mnoaJ$t>M*r3Lr$jWEAz3EOl-?`7~)f(qd2QHD#+`NGH>)#|0z#bzi@+cOvVzQ{n z!@`qaJ_~Kcp)_Bfz`N?wqH0{$CMkB`E^U-G&>x6XXI42Cdi026Ln|NyBGvg3Y$}Of z4?M7sTCH=6bGOj80%iiZ8U>n_kz_dCq2>Q*|IZ3wm$9YNHLBD8UBb1C8lE@&)U2u& z-zV3haH@eGajKUVc(3-jDMtUXQpwVc?l8x{@1^W z{Qa+9U;aN5nsp$Q9qSm|4FpFr#crTT0u`ytp2O;C?GE(M5r@BRo_E|gZx*VI8ikEl z+wf)NviL@K+V-RA-KW0eH~-H?ztdz1Q`<>`9UIU^1`yCtY&r$1XC#^kxDP=8%MRgc zKW&OP^h<4L@+IxeT2WzeJF_X^L1|EiPXz5HFv?qxoK*hwe(J2mpsucj(^>|+O^)V2DqLw6^A8pPMMA?L>~rK z1+@p%`|YIjMOS9tTySM(o^tu}R_~QQ$jMt6G~}A#moMt0Yuvl6ZaG^GNnBuJ##qxx z3VgM|&sYoQq$Sy{G$kf3j zT4jRAaY)?iWX><%7kVGK5G!Y;h~h!&v_O6fAK{+*(*J$MRkl|&`vi~ zL)d9W5p-RQ4oVMdRuw{G`Jb!`FEf!dp7!#@IZM$p_WtCS^uPP-wUis*zEwi;$*KKt|VO6*|?DL*ufN4e8{|-|OkUt#w5zqkEWxinvfEqm^ zL#~)OK3THBob?`+_}bq#T;s7*o3?MqY2rNPZzmpn$=K$#+uB7FR_uquZL`y6IzTPjSuHH*Ti-)Ax*(>1n+YM)6WHjAZ%PLJ zL%o{u+v3-TNRu6V=RYw}9NH=NDn(j>`=3tcH!F~sE1%au-*VFym$XB+S_*>=v~+cZ zBM#W@?3z3H=4Yz0fGES|x+GquM&9*2`qC1}Mg`Y=FHm=NNwEYMns)Se0Z(}-w@6sE|u*1RNOi6}Mlj0M_ zh!{ur(ye0KX*Cw2<*6%f$wf|Rp1RrTiPB~eehTud3uf5!9`h4a=1C(?AS~?Vz7xHJzYqw*ili4893y-9b&3X38q)06={CSiBl*%6)rd=@xOr9;*i_dS zHHuqQ13cY*TXlOc87nl24uSeYtYpoh)0sZVl#FNVm);AU+67EGAM>Qk~Gm(J0M5!}(&FZ5^mbOV&P}i8s z@O$!vfE9xO&)%27HIb%yTSA58#gL0&QcY}?2p!OF4nvE$(e9pmrsv$*ncZ)8b`IIu z>Df7YX7f4sAnmOnD0rZTLk>Aq6vdNEL2YdnY}B-PA^{v5MNp(s;Cr4Vv?Qi9MG`jc ze9dnpRrS^zrrv)&?{og2kHp96GXAkC$AXJPtHir_;A0b3PB!We%-$3Frv-^r0Y3>i z*>}Q--P3{HNp%4f-9%A4F}`|?R8QEaUNIvpY~7?4GlpA2B#L@t*G;;m>QU^7*!mI} z|X0kfyQMcN_^H*Oe#f^!L4StXrUKkPZU}) z{ee)!BS;bzNnK^C9mP|38Udr2+m&0`5#Y#K^3NY#S!>@*=;`goixXm5ItC#;N_lCL z{nE`7lV^Z=J8bVlTloxSu?XthT@?Xb7RhO_pnTvmv5RBb!E&N8R~2a8bbfM$%qMA% zHXSbp2Dn8=h>#qL0mWn!m4Fl&7~{F;Z_zQqC|X8 z{40{`OLo{i-fW5imwP6afQ^Y4ENGUMMJ!lgN%g9nya(D8TaFf>+J})cWWURcSQ2?# z(5hZgNFUKu3WkVBSy^3qww`>8ImvMEb-C|}`64yP3X!JoEy^bcxCs(3PCL42gY@?) z29n#3QVA`>$EqHP*FBaU34g3g6<{1cg@>x`A#FA!#~ajxidB9_-F;b>v{RfUdkp1b zAMrK>b^|d=h5t6*mdWMQO5z@ot)bgO4eHX^lDJIS7Vx{_@p5^pz!bPvZq#){(6|7o zRZ=90qNeB`az+C=2XQ80y(Ck%I`Tf>6;0T+S7M~_aq!s%c~t!GWpC=NSTaSwoJkJ( zl8xV-+>!SxGL$uRF?6TB2Zh{)yj~zX%Y*{VN(tt#!^LzhNhi8Si)o^3SFgX1z~?~8UR%H2Z%VEaP3 z2Q;>Gy5r~t0nLwh|2#%y6+!QoPxw7q^}=jt#Ws?3GsSGgcQ;`};0D!>R}Rk@{{4iz z7=k`CibL_*afv~l`;zMc9jAfD73EG(?Bc-hCjRJkd$Z|jPrmnRYlOu(+AggQ-!C<& z>wpv|Q+8HTDFB|!-I7fZ@Y^q~5mX9*sJ4uMnYUbih8!UckvEjBN~}UEk`6?9IOD*T zghm%v``Nd%OT7N>*P{Nh2$^>BYiQ)>0x}e73z&I@Z+@Wen>gsRY@X$JE_LJ0Rr8GB z8T6_6&d@jU-+?a-`ds|CqhA_O70OmB27R(WfBTIA)e`Y#UYqiWAXAtX)*nz5`s8@@ zYL=r_Pz7T8U zW435AkvFnLdPtKA#I)$*!_qQKhCVi{?hnY8^ama1K~IPo zUYgTMH}O&hJqje%Ef!TlhtLmWAo1H-+7}LR92PMy{&o(v*M9Q{g?%=y*M^Q*(tI&1 zKSOzs?vz860C^D}D#ljUSRb`BYwX;rxOWsMkuz?U;r*p?R*|#5P?1e4d`XkdyLyab zpf<0DO2Ar8ppyVkSfZ#?jBk!XjoojHMQMRmAw{B69yZW^Aa4|=3YH6p{PQ8)|1hjB z)N?Z(^zpDuTS6V(KDRzIzDOqZ~uLPRlZ#M=fyM0eJ@sa*Vr^CUo@UsIGzk* zbMkvo`*nh}D;oW_Ox~$z2-+^W%&YXz2)H-nyu2;Is5?jZkpt4Yh!p5qSunjK2-?C( zu1gF2^25sFZp*exyCBh_A}%MOC$?2_j3kQE11jQLq37j6X;WBVRErQt7}N*orC~;0 z5iivb@4Dc(I`R_Sb&j_@ZiU}5k}bL~EASh>YZ0_Zub~%5HW zAG9}rtU4F4ankmnY!QCr@V(ohRe7gsA1{Sh7QGz2oY}lm9tdAmwJcA&E#|WL7+E*z z1kb4JQ6@8%kzU*3k`oN6uUxJBm3__(msq+NdwWptpbk9M> ze!q3G_1Z%JirFSvc65gFru9Ay{`bCiU^eP?+0kf8>5s#=wefp|E1)8>N^XL>`)pof zkiJ-QMAJ!U^H$DV&0jIg%(dJ{y5*TN(MjVR`u;#uOtGXhbj##nL%>ip-4Ie=`Em&k6i*b=f~>crSrroPJ4m( zzO~ekn0X;qc~m;|y$X`eO&)o%`=`nV1O#)|vX>(|%EHg^cdhe;g(`UaVeM+kjvf#Xy=_36+o=a$t6a zB9}Qx-v`bXNYQImw(X?P{heWduR;9|lA%^bW+*$M-I)Q{Pm5uH)Srty7nwONbDBZjpy*=C#ri$dx`dtd zNyTO0MBNy^T<&(H;5xYlEshu5rL&ySGG43r^Y87W;a-atmdH0I`DXKaAcGG<0U`x@ zvmWJl$_wE*n}?jZ9|l#?4@o{<1JMQaZr&g!nNhV!?*5&`e(6Su68mY~e#{9a<9@u| z_}!PR`%vgFE9a99FO2F^VY6#(qZkN@=2|*Jbjq88u0`IQXi%fIt;h7}PwU^zAJCY= zMCp!Tld2q4_;Uf5#rlSTTETTOik%CQ9&{^?5?G%qE^bGZb;6D{$Ni(NP*MCrzlH4N z1{Lqs-^pnks8mzTVTv4}5{^ths)b#s2hvqPR@{0GnhmuALzqeU5I74lx3xh5Do+(% zAxRHeMqzs*OKij>{JT((L6RL47y21OlB0)w@Mt&hDs(Nv#GH?#s;t~pr~#J+nPknf zBa;ngv(X*Eauq?JLdo`n%fK>Je3E0Izvs1E5=%y2i+_v1zMAZueO0n$>apxmu zkR6bLqQ4>Sfo%K+S(`u7=$kVfj!e5E#-4R}a$|U7T_%Y9D%T$BfP$t}|IP+?iuwPaH>F4N2vkdY8Y}|kukC?ukFf;y( z|E>J8*eYU{Z@T_7lKjG$x+OLOX$!?r%AWpo-APi`5LPv>X;=lAVx8n3l=S<(Uu=*22a(oueh9^D)1A?YEt zJjg)8gx4N^Cw)-6J*HEMUe!Vx0(V${m>QQ7J_g+konCi)HN?ZdR8E`AxCNh|bIO%i zuNMmsmRN5)^oKvB#a_z`QY7t)6iL3eiZ(L1trWhzNUlFWd-wDfAu79lU>{?iW3gz# z$v6lPV=Ly&iFI#eg#qOPbeAvF7_|NTy^tym(cwpQmZT`CPqv^_eo_E-GYBVOR4t6S zD_*Hg48n+zi-KwNczX(fb|+jDM+LQ4bCg!~WPMurM`Rs0^~8&<;ypI%Ndd)xOD-E5 z2j31lAib@`s2?UPp$D#2S<3IGJHVgh1HemAub}} zIqT&DP@XfhYnLZaUG;;=VLdjZyNqo2VpI8)&C*v=4A{>5hTFn4%g)k8kOW;U=?^sL z{UbkBssL@9+MrgY-ckUau5D7MYkU2{h)xnMm#6SRPP9igLTX7-Xn$ZeX#%~aNMr?w zuw4OM|48goKx1e4-r&mj68G=4R|-Ft!n&R5s7)O2pTVT(y|De)QEzA zs-3p>r9|qz=?keGX3z)E+@86LNrE$_bMZ{KbUQSm!98O z$cy2?(pM;jcY&OEyJ_ybZF8?lcg(%1?1COLzxC`IJABEVlP3;u_2@}B265lem#GNGFzBR7K^<%!darQPxz z2)pNLQ(sx}%8*Ykb5nt)seT*XPv_04)n-Rm@f$puO#JD?@z#gS{&G4XTJm0_V9t~k zhwW=~UJyybQpuANx!8A5QOMiwKfqt&n-+DPUh2Do^hs|hA%!hCTMNl1nvG0fz(U?d z@iN~^|E0bg!yAM*l>1~KKyD8WcP^b`(kPMwRTr~1L(6Rw*fU9@{s28D%V+Q_)!lL=7xEkc950;HuyB04$IiCx zT$<&II^xA%Hx_l|LX3VtZ)5oOsd-wg1MQ@*L}KIydX=H~q*-=Ct4|R?b9X$~C@+Za zgjDLPNX)@s6Xt23zSAff78g%`dki5sap2e2kT(#ORk@OPK@R>1ZN4@?Sl=SN0yLfvq&=}_7e$S1P*E(&jz(|iwm@KoP?Twge2_%D zMQBk`^68JnoituIdpdDS?7xu#g)b@K+qfQWcDe&ajxryz+xz*CU(ceG!3jF#a|24i z&5n3lkEMCapw@4`wvSrcbJaag+-M0dGqdd~^N*o*fZ7Q@(}g=X&Q zKsJJ9S?FRtOYRA{;En?no;lPh{62FA&s=23@A!gux_@q!G2y>>=_S(k!g%i&*>HvR zQcMp;x~PP!(zTH$y36mdde0Qdnua)#eyb+w73@c}M|xnkiT-FJ%D2`0CVA=H)qLEZ z%ntl!S#MY}n06S3vWC?)c6XqU{O?uRgu`L;amrjTx^bRJX`(5WO@cN28( z`Kf|j-h9{_2ST@!0>52~KC(DGpU&mw2RF!T1bXKuNFSvc{q_`yGAK-GG7^jVcrhs;O$ z=6tw?qEb=2xWI1%bf^J_Ti-SQ+XBzXm-!lXEszZbxs-9Y6jf7qE4t+^BoVwY+xhnd z+X8p+*ZAL_?tbCrlxTYX*6DWn#4ef~xCN-k2k(rxk0HGv);CMUsYQ7A^%O~iJTqjE zbRYkKq$?O2oJd=FgFeNQD)EtUWcDtNu4J%HN{zNYsEAi1GYJxxN{TQp5W0w zh1u@5*~q9&idj#QbSeS21WfQniBm=6jsqKB8}w>=!uoa)YTrF=V~t!tj^G%1t4BZk ziyD1C>(xO?%;VRu##sJs(iwH<;eY35Hwd#r4!>OJ@2=O|*=2BgrJbG7P2;$2I;OPh zuTfT^F%UR#h79^}n!t+LRLFvW4L~6U||H7ZE*9{aC>0vqn%jdu_!)&>JQ}ff{#C` ztGsIO=KoAJq+)UA7t+{@*T^)4>#?9Xi>{np4h(=-#OWbpsW4bp$0?|w#`x4?diDf6 zRJ6a5v`1Mtj+~IZbz}!OZPSah4nDBiII1Y-5JmP=2_N$dXP_0*A}pZ?XFeo*p%MsF zX^S8sHZL@V_r7v(DA*hR4gP3<6!GdMIa6!sB2hE=u6rS)KZ`yv+o-!Pt&_GYR|OS` zGI`*9=53?bPS{1a2dxdNfw)1rqKBUhjL3`RhCny%OHM=Ui9zz5Pws)-pQLYw{bQxP z)xFkDK2g;PHp|WccS@P&ei*bOi!g*-k=F8R;*7w4e;tC)`!t6%mg|Yrw+LGS zWEUo#jB^tp*p_efKw+EFJ?Ay;>xtYdX-f_B+LXjvX z^0DdycrPlao+eeo7Ga_0`ovB-))Mw8YU3`8v!)q!O}spEep*9hCv8&P4_L<=6Ww%l z%kG|~eby}OTaP2>S&%L*+dePSYt>ekP!D?BR|cAB%)bIRDkQTFM0xm~cf0`3K*I5z zo;}M66600LHv{cm9$fsXUYv-)LZ4zXcP+s3^f85?C2CpHbelc#5bY4_z7vf zV2sTgr`Las%^G$oGEx7v@jR=h@$2_{PmuR~Nsi5izLsKADUwVjVC7uK1U-^{FO4pM zmPcl)?+R6u`og5Kmf21}XxH8w&UdpvXa_sE_r7mm`L*G9ldY)GCNHlhTYMe-{a)+{ z1@m=;PP(09fEBioO6XBk%^SMpYHzEgp zsx+BQrGEn{*Deprmu7|OdlXr8p)7@GRGsh~4XD%LW;I$lE?+@flNbU>cFsrc4lN7bBR(vIZ3ynqquaH3+zXynx5C6d0I{IQoiJgSLASo2F-dBLj^b}` zHIda`>^6gjTq9JJEQ*21-bN}RNzgj;kfwabS=F&A9dYg2YoQ%6iK0Vs{rttMn-Zh$ zgsM1nS@`y#{(xFlQ*>oyf!3{y&bm;eKm+Ss?!3T`1_v4|b@;YqPp}msO+T)9hivo( z9)OVxQA9BX6xm88oROZ9VnfBfab>|Dgczq_=}UU-5y->BW{#cWk94b5t%_yL`_nU~ zW0Tml=*_ZQuhj|a!t>xeQs-ABkJQurhau+aS9b zl~0$6-xKW+H_2-!t&$8xe?fL=U811hALJPLe|2`1VU zmL5`~F{nSS%E#gk%v4Da$=5c}MG#6Z(dbXUu>iUz>+gg$La=ya_-cN6j0yH@b8||R zHJ(Bn2V6OpDbE6j3)lYiFU4ws6)6!LzjcBv;f541HmLv@BapI=V%Ab56?V7K)01jd zgFfqDu2uE(Qv{j3HKbe6E=f`4@Y`l>PU-EU*$Dc1s;h`FE zoIXuH49p5i;nj!l5x0P{Fviak9)**S(?bu9eOgZP+k3_Dg0J2Xe%8BoVPZ3cq#XXj z)Pf{YCa~b>kMeW)#gYnD^S_P-ik-x*?cTr+H@98khFkw{%D!h050=(?S4FLgB-vtZ zqNqyT0>T6NLh$qD@NQBI;n;_BtLO%Kp}&V5nT{V2*Hv&li~G-VS_SXjlE+V7>;DU@ zvhk827E3y|^?8tvq1F^X;Ov1Fo))D3>f`~>vF-H=F< zMFSmmBiOkMQPS)R#ymzeR!zX`SOp}pP|+wa<#ofzMsVRLQ5~dQoTtr_ zR89s~j2`};Dfg#g=)B9Xklzdy%W1Jj2Hi`T0cnDixCVKTvIRMh(LRTYQ#wy-0tTw$ zkWFg*qW%E*n(M(?^m}2MvSpA+iw!(5V3ih^DSKa}#}C-}_G;)Rbz;2U@)xEGGL+>+ ze~Q_~tApBSuPq+t^iGWX>#8|Z>?xaAxQ-6Q9RfZhB&@0Q$K?gCPU%q0J3UGy`^-?9 z6rFPYN$}l08Id9>r1R+xqF+W9iVR@4?vq=1JD+kpxJxu}6;!tS1|GTVFR`76+oE`_ z7jJgxz>Uv(Xd0_FPLU%T6hkKHvax5{LxsMMG|S2&mQiD&(u_og(|8?u(-?f_hSv}K zinhIK-MjLlm>0B73QX$FHB2)DX+RO}cfldlb2{GePHPx9MuX zvf+!ZKfOKoyskxvLF*1_Wn@P{mm+a`ooYZ(PM5}>@#_{qHcZS0RX*J{sZX3G-Jlv! zWd=Y%TUiy8E!`vSkS_#Dn?+aAHEv{_bQ)_<3q?m~+=jvn@_5;JD|!|!onB9de7sfw zk}WoPSWPjjD6$-hQ|c$xhv#Y&r<*40F}KfSJKVD$yEAM2>=`_Bm6JRfpS)r4hu^d! zV%e_`KOm=l$!(hle}!T$QRE_(@QI>DxLkQpwI_Hu8Q_yIHkgwEvZiG-HDn*ys(n!d zp`CQ4|MiLGbBwwRlQQ@XnmjrQ5(qGPpx|FHb)a8*GOm2i5cKr}_L!(d(S=F)QZJdg zh7|Iu1s!sut}D1fcot@E()D~9@7Sd{PUD!p5J%qgrE5~lu^|MH%ig#qEe`7E509wy zH-u#S?UgvYy*RUsj^4>ppL256IW432+FQ+={)N~xU3hJHh=seNiPt{;(Lc2+3h6pQ zC&2RnWUL@9U=gp8G=PgSIs75H7f>DSEaaR-z_W&NL+!I>a_=>6sHIO%Sn;}5_%wg% zRsq@LC}qlfU6JIRjijuj7+`3sq7pFvlNeYe1s5JhfO5zhL5Zdinz(?)-U@N$c5pdi zZ zQw%Ndns}?_g%QP~%E`;D)3^jgZqVW;23-LgJG9>W!>@m6--g2O+&PcHn9Il|@S@&{6%J_TNr*zJ@Zb);h zP66YGL)O3?O40GL*4_` z1q)?6#Ah@YA!~P$phbw|@aP||;e@i0;zNF4(u_`F0oFZjZ0`hjYgYO+VJi+~sRfNEx z9DIK!W|lxcn+6mQAO}q)2VVs~&rM2`tXXz@CWINw$rbUC59X`sPeJj}AkfrGiqxqQ z2WBr6?3o5?NkvQc(3;PwM%XDr&@ut=-o-HNY)D@ ztSW8rzmsB0DN;lw^uB7EoCEZcAE+0Te&t4ODkOmJQ?3$R4Xpz)RT|sRuZmm+_7f{OkYs;E!;MeewOCNw#ta@JEI0=iF5BZvzHbcwD^q zT_7y2%P>NT0W|azLOw1YqBRB`h)^qn7^w0RAVR-fej&0D$e4Ep@0p?>=493vi!dF} zjRNth<#7=xPtE796ZOF#|LZTUqUDY4^^?foy*Ns~%qB|SPcaWEa-T}r#bh(NASQ-? zH|g$v?&#a+=NKf z`K;kPKlsKc-#D*B#;6SChy3%vX8JyoUGr<8(5;HTq5MRFa~Rb4txy5QPm^q!T&6j0 zp7U=vm6gyY?b7ti;d|pM1JY+LnOW(dD!{L(j4X{cs84F!wMN~0)7QLpP5RfH%AMB1 zE#ke>E0HJT@b56{8vV+Gu~x&Pd*S!o(o{a7({vF|qf=>~an~>$Ic>Pq&$eutXfLI_ z)?8zeI~INz{hGkX!0%zs>@D^;r)#vDoj2(rpp+$n=ez7B{kFi%kh%y=5ZKGLFye$X z75er<2Ui~t3LaO1GQ6=#};}=}Z@R?_oQZpBP*g`73SV}e9s5GZ3<|IXqQwal6 z529+I+ergt%GA&UQI-Bh(t#*Y?M-x^wlE@}?xj~Np>gZ`enUPx`B%hvVaLSVQD($6 z@jlix$W731Vyou3q+gmGT^LX>ZOErlenXiV*rd+aKB6BIi?6`!LM#@UXzUVK48^_K z(aF((!t{_1`N3#Nc7t)r%;t%sLY=F4dWH-cafv(HIN@pBUw5C>e%p#C<;t+Hkh9zr zYVYO6Ae}boyGAk4z~>{d{=-tq;qWGPCv=*C?5iW>{$xxDLmv|c#UIgUCf3Jx%Fn9I zt#~Y~HI^p#)w9szYzh1#t~VS}Pd(+w^IeUsW0lgt#- zcS)ra{GM2&t~2hqq==s_MGXSupV;?1D{Lrqj9kLppX_rUiJLBSdegl( zT-vE@efwKhfMiqsUnOT=7!mZb4bHDp4D5(TDgoWs7V?@_AW_WM7KJ8G-xP8+ZYvZl zZw+i_^aas)+U)Scw&feAwQEy&t3z6q{Q*UxH%MUwY+_sa8A>GKLt!)MQxguKXw2ju z^jR8I1tE^cSO@k zL!b01iPKXfZVNtE+=ALsjLkXION{$_GAu$|`ThKt=}6VY5c`^%7ldh#h5qYte3?I} zQmA2@OW{LW0dI|JbLx;YEBX9q>ih`jQQ`cY?il&Q`aL}R7JpuAa)IhHqNwiE-h&N( zo2Qj&^v9;82s)M9V=Ddo0xEHpTa};4sv--b3!vwq2^x%`m{~7DN_45tD zGes{K$lmbR6^-4HtBt6oAhtNt7kgmuO05MN+fl$l@;=@@Z1H z2+`LB72v#lx?BF2e*Us2wz;#*wtYT33!~W$r70$F8kRXgkPF|aO8Zd~KSShbWdPwtk?R$YkK2`jNAT+=sa-8Z1*1ib8C2RU! z@k-^UN%z1aN9wmDn)7}+l0I2WV9w+t*sT&tCwbSTUR%F4`&86b+G=g}emmq8N#bU0 zd9NKwvTc;4H59X&BCDu`8_I+9{n$Z8xvEauHmg0xpiTW8`W5~cb~fh(D|f-@>K(j~ z_n7>xp!lQy!tcm3ZlLhuASn=pjnJGoP)r6z)=&wTNvrZAxPR06P)sT}zxnC2EuOn# z5lWal@>{QsB|@BohE@M{b-g*<|HC)`@@0EoP!>W>6rI@65F>+2W^M|}6k?VQ_(87# zD+Sj5W7`-UpkFU3mb9wVLoj*QGO`@>o_4wZ1igh|<=oacuD_M!XPS(u!P$NtZcqN? z^t7!H-Qh0|Ptuv8oHQKoee3#T9tD5hDo^(RYv>QiDMv}}Uc8muvXL6i6l0{wc`D(Q zAU)u|>;P4yD%3RTI{x|ok2~gO{qw-*p&E}qt`SwW$1pWL|%q2+BK|+=` z9gat}WcWJpR4bN49$PONgpNu%pwwVvol=NMoTu@QZn{I>3fcsYhgdytS$N)@Bhm)p zIXYW_5qP5xsV1Qyo!&ubX)JkwIa3QEa~}B|@Gj>?m)l+ihwsNzC!Ki<%wyPJg}

G{9%41uE`Pu`p8=q$eAW!(#j5LU*W6O=zA!zkDFC$!D6K>#e`RkmX zV11-=z@CTAYn{L>%xb7^Wk;V?Wk3Q}TR=AR5CTJ+LUzPvP&_@OZb*e(TIix;(Z0bLqN7M*y!!4M3 z_9QziqQ7CEwD*GSEG#L0mzaeU*Fm;uW?0MI4sur2%0S_|{|R7H>QvQ`2KfLKxTlkf zxQl}F89<-(=@oQjW2(A`MbW5$V#ds~7uwaU-$?kw{urw;5&O;EO3GdsVRFVsnAA`V zlrbEn5)RH?usEzEXz9yF-L`qyF=2yjAoKv;E;*!GJgJ+s^7@r&f^KPv=qk??Qyz@@ z#XDl_c=shAgEh0*cVooySz9MDZM*}O{sF8G1em*ADW^q_Jef=v{;VG*5v+?-b1=~-PL z$W2#y??mrQQ>@q+U`{P3hR>L{<**H2c2EoyjBkU|Kgq4vRx5YC(etGzpPF=deE!Y) z|LFc=+gHz zSwWGdR08U&Ia6DNjq(QJ!?~VcOUEnm+!c0wo5#c zBwoK8zDVAqFov6ikT0N328xOyAILTEL=&jzLd@a5V}H{9h}%^i^+)?oDGr(sdM_a# zOZSwVz(@SE>f6e{7}Q7?tX~<3KhT$_SW*>vQ!x-Wrv0qBt-kwLnQcaP0rQJ%|Jq~U z0pdj@GqFhE3V#))E`r5C)y}~{zn-yRVB$OA69+uql%|WL~@hj=TZq+Qy;2U1n zS1^m>&4oeDvd8gbTo1$n79Lo9_9+kORwcO-@bh9=umt=p1%FjQ?24&LXz1)zoOxB> z6MJAbWFAc}pHdct6+*z}P(8sAT%xKDH>g{ASGbq=us?K`D;zK~awB4y&3&g&#MB1c ziw>{VT3Lj_A&o(sMVEyf6I2R}x_z2%x<-}{Oi^i)!!LVc8+hi^cXpwlIf(}@vTGa; zHUz5d)>}ti$uRO_(ZP~olnJyM*rcmfc}TGX{OtOks4U4f=}|~{L=KPtjV_|jLdhyH zJOCQUQ_fx(_5FCOiW>Uf$`Z2Qmt3&X?M_fkJw=XC35E13IWXV~JH;kp_0$ve#~SE{ z)=8H_q8}Iq*3$)~KYmlRzDMyO3W-s%mJ_6Oj-|@KK0D27^&}3j#4M!xCSF-x3ixe zKji0}<@7pxZ&vm4wKsm}B$LK9<>0j9DV1>_kOL4BJ&Q^YdaEZXA zD4J6zz-*L;fVNpCMWJk^0vR&a1Qz(+56GT+EMS+Sd_woMd(#I5^NqSZx>3Ccq;MJk zGOtEZ7=eF7o6ho=4S&gzX&;F@>G^p10sY`L%U^Lhybiwe&df*TVca$N%l1#NoV-KP zIdQ&0eKj;KZm)EOb-rcvj$1%uq)jki{Ao^Q|W>$NH}3*~S(upELPj<;8{;-yOF zj{FqKnTlQK%41U^caoC;$(4ci+Cu;GSgd%u4s{roP6ilhOd%Z*2}0s0OG|8{Iz!n3 zMWh`u6{@o;tPyd0WPrO+W!r4sBu%Z;a01S__uqQwV|!*_uLTYZS8u*HKiCkmJK|_L z2p3IS8jXIt8o-rDHfDDq#=Tgv%*F{Z%G|QVj-?aM6I8%-=yrG27NG< zkYssq-h;X3MyEqQnY^`Ool!l|kMzLo3$m?Lr5al)|7pR}&!I0z|L1=GNJ$S z|H=H>GCTvgK00Z1D?{-ZV6-Z-I@qALT+^3`a+wz4@t-aDgF%hJ&YXCG6wrvxjOYv? zqiTRZsB8*!J8coJexqS74zmn_FFMLcLeR{Kx!Qf9bpnJw03B;PaiWbxKYVsaNSAD7 zVCCc?pNiRMfH@S3KPj)>3{!qqeNk%FuoQoLtBI`kVhwA%jhUH6F`Fr}kxIzsU#gajv<&p+r0m!5}eiz*@FO?Z}z2pPsvT&sJ#dBx**T;xwZxm9=6GG^zIXOus zs{5r=7iCt2i0_GiMN%EbFTK~uBoJpB5x>l)m@O2^q!Q4tgAN~5i=fMLiKbb$Nv*#X z-3U!FvIR}RXlH5KgbYH++OY!)%{;Vy*{;wiU<3M(?HrysVu#JpmJhDk=j(c{s>70- zi|p|k%CnNilI<~iWQ)oc9H1A9iY0iVM^UC(9DYWNykwbjeP&o5T>1dEh6)uH^AGtX ziHf8rp^*V3T}HRlCD2dhbm0HjbQm8(Q4So;kZ89llJw?)~ zgq{3*(@zVJ)9ZN;6fOSOW~}44Djo|fBiE49*h>Fv!h6!IN<8Bc&^VR`+~kboIXrPT zV{cEBJ$K7zD#eb4p`|@4lUJdt5*xG+6q|soyHVZ(*_WqeZG0nnplDF&Tl`zbAQt&# z$v6Sv$!#A0#Jfpt4gGEbegII}C*RHBzs`!tk<(iVXBcvAMocY=< zVAUKEmc%`THW$a~rM`FO=6t>Kb07WlGT$+gA)LTt_0M=5Iq(J3)?K{$WvjXxzIevJ zkYiqK(6riY78fYyLy9z@YT6TfQoGS_rE(vbmq~(C5YBGn7lq~o^e9tS|@-3C6n zcIct5Kcc|{JPBz~f-3Qm@J!h$35DFb0&QT2>4OG}Xde1x0 z1!C-S#(~Z<{$g&K6YCh|QtRM25_;sq<5OCVkW&C8n9U)r-fi{80Y(U9fEGDoLGX z)o8PnKP)DDUznNzP|0K+n^aCO@!b-XCP4$UN)NMfiP3_~+Ko zdO*5?e>)0$K32|M#k4`+>czfU=ER!62keE_?~r#!bpvAp(#`kq6Q>`Yi)@nJO3bCX zFeOQp!b^!?9(hD~nKys_{3Ej(#BH;x!)tl-U0v~~7sgE(czR+N&#+4w2PO|S<9}-U ze`u9Ft>1k6ZPM(;w)7`9vg9_!bWp@ZB|Mz8h8RdwWDC7{Vh<#ro={ZLCdFORMSB1A z6Tnh>TW}2I#b)(ck|?V4-x|8ax6*&HZ$4n9L{t>K1bBQF`!+%rYf*5g9FOY5hi5dX zcPg49)2D)(99t#C{-ek=wjE>_s6UH+Q+QqAp?OFeK@9cB_f70nra@zLyajt6FO0tr zeZR8kQ}hx+0hsr>TFi<~W{?P|LAZ!#2x*Eeh5WD^%1d+bp3bOR87oZPgtlkNo@ZWT z10^@K$s1n!n>~rg(>jnByYN`3Ix6D&#A{VO{6##JL-{lkwkOW}l|g;>pUQzWC{ffO zkjx}bzZ#PxxvIP`D+ z$rdjbNr!Ai(sqg|rbr=`fWg7)U`QT8cg&)ITS6={$e8d>N@QO^?`s1yie80l8)Ubm zy2wRfZmEjQhmsb^Nx>$K{ee~dy>n7|4InctNqU~@F*{0Z*a`}+3-82uc5FD1giGcJuIA8}Kjy*Q@|I%JPfo^Mf18%1tX2|Y?Ak^^Tg)VO6R&xyD4?*(+z zDU!Y9)(p#8=*0Qn^lsWjo1kYWmQhstuZ^siW+;)=zD`jI%yfDz+fLUe#{XSg9Fi9K zh(8D&);ofeUx9`vr8KtfON_4(;st|xFR2llTWKNj_nv8MB8Pl#klV-|(I`&~tb)uc zK(}(Ke+INN=?X@6iTl&=U%b;&T!T5LnY@&c#i|pKx?@luR_~c&qR~0&Pz1UKnYK9P zeB=nIU%EhjFUpFwoRGYAWQP~arVnh;Q$;Z#toBn0*hUuHs#Te#3nmJ;Dxt|56!RaM zc14^%`^+Q*bdtfJ2Z}RNBwfbywY)=`Gg4fC%e6-NW$5Xc9o?g-6zo!*B`JawUaS=~ z$Vw`qSdyV^<8Klf)tllA$PQ@xRT$RF=s*3>O$+YC0&v8H6EnJJ2q#R8D?A+KXP+wb zv{vlJ>31v%F(~7axxLVTBe}%%C^xDP0UK35U2Q%pq>uSwqm>o2#tttWh`w>Va1R)q znS~uNe_3+zl^M8`eF$VFhR@E3fO0`WCf?I0i9+e za|O=eG$o>Tb7P-fb63NScpaxq6m^QX&^^?6e`bi+zfR#xR%sZ!h-dvY5&3bd?aP zi3+D6rVX(RI@LaD%zLbzSscjY)Wv-@`8tJ~78khM1?d+=M5N@=Y>uslXt82k#Tv;`pk_ zcKRdI7l0Xiy~M}>CzqmdN_KdMya(JQCg@<)qpXKke0StcYLg5o9{^pA@*3HoqMhy| z7}^AaFUs=6RU&c3j?EUz>3}1O*muwM)S=)AE2{qZr`E${8#kHby<06gW`no=6axnJ zZs=h{XUg{RZwuB&f|m__B$*SJiMG-B?>(S_-XD+`aX2C?6!_fEs?e&|V+690-Z%{! zyGEx$1J4KkJ+XbU$0b|&rLiZKg|l0PeF5p(ciutjnJpn(qdVz?+U+s!)1loCDvsb` z8~&~4FMRhqR_I7nzZ^#z90?>Z)>!V?0OUHwT%pJ%DgmqKTR{XpqA?EL9tBwo(6tvS zt=`Gy0i6uwY1YvC8{p)vfqGbs1=a*q&pH-kROqWTxlG6Otgyc<6zTKGJ!N4;iD$&J7^axFK#+R!E@HML2e+B300@+#$}MV39!Y zB#AO*eX@7nL6*>EWFh7oi`Pcpo>|5N_35S}pI#hpMrpU)5Rx3NFAKUqnN4cB9Y~x+ z77S)V61#MIbyL+s`%2muL|ta7vAzA$o+;Y`t2B_$J2_8_zT7JQ;#V&5v*^mn#gb-a zvB>RgRV=GwJSGQJ9zCO_^eVSjr_%2E3wsCo}T-t3Tm+~kkhav_lAz!;m zWTMMsEhP!>pqOfuw}Vk-xd7Rm{2HjT>`_AQORAt#-Umc`Ia4o_R$%|=QL+gIH?0a6 znAnO9H~j%D)4cRe`ev{dCQU!Cc!z9sv`Klf7PHR=3`G?p8cOfWMaFM4!BDpBtNRUb@apv}k$VXcqTS%W)ZwPar zlJfN9A3Y+>_;`9kH=SYMZL)t!FPt-smG3ReCkMPZI&{$n7w=OHsJln0gg*}Y1l_zvtgka~FoP=#vh_84qmu~c+hsZWW--3XbgElTqdP3@eK z>Ag!rkbDYlE;N!-B8!8PCz>T!hx{NAi5#0!GAVa9syIL>LYGX^ZIE5*{d)`<3!qgq7MMz}hh__PehR9)cKls}c$Dp1;1& zeb8J6I1d&sv${T(9V|hH+3N9DU@TfXy&kx|ytkJh5ZR3Y!D@(Z|2 z=A>-wcj8Gs{p^e`m$olL1+wW=FU9$s=k!}{;UioYQu@p!UGFSK=GA*pl*sB^ueDh zAXqGn$fp~j;3+w}QC=*`WzYrJ2zs+oZ7vm2VyGTdgKsEvwf%J7oLX&4WK&QY!E=^b zMvp-!vE)GeaOxBrRGxcY$b09}@2q>yaF>G@1WPPE5gO$^%4D#9`UBg+{nDc}(M^y} zg^gP>lxrvKg77vL2|<%<{zqa&SXNjKo$Y5i+X_xR6J5n`2Nl=MNeI%C)>wzEOo~=9ZbyOga+S5^_v) zSXc##$C#9khFkAz8@0&tyi&d)aDQx5WHa<|c*K7Mb5*EPKxtnjHw3l@Cxtc2F&(){ zZTbIGNE_g;W118Ra$aEVl^M4lhM-e8wI77|oO43_XW9zE-fxGTB1v9sZ)MwTTWct0 zHAPk-Ynr7)fjK4O5X6vLz;H(o&gdn^b6;JS0ZuM6%iW$k%}w(u3yQdGUtPdOE_tz! zxYY(ipAzpi%GdL*Nvkyd5GK%f`JDhJyaI?FKv4oF-T(1?fJC3d2n>3GkAUxNH!CM$XfF_4 zICc)Y@pCtS*Pe9#>6#`lw!T@2=v#zkZ(M^i$c`zEbAMakMt}_`myD_oLEJe*Pzt_`-yz3^t-Cm12@9 zvI2~D2sU)e4+DK?t)>Su&-cB&g6Hge?&31o-tH7wY;PHpqi_ErJM6PwQh*IaZU?Ol z)Hi`qv@RZ9wirv=2wtWkpB*v1v8a|-|Nojtu=_bF8a?ZF`rAjdy|&zhB|eSnYudy; zQ<`O0Lk(&yMbjhmZa$4{R?V_(iM|x7(okQ(RL@qMt2W85FA7VsGxiQ z6Y+nn%4t~64akz_rkr}Q@3+@RIqjer6Gd7qu9ZYlmZXqIQuuO3i_k=GXY_R>OHxO` z^IM_95)|B%ptc^E#m_-y{qe9SpcFd%@*UtJ!YvG1sx1!B2bXbSM3XvI095pGo%EB- zSjUw(y`HDJvm~>#D#t-l0HZ? z-|K&cY?_!Rz~+2Kykq3_x9`uzW9(IH)D`hk{kkDf{S-;z75Za-ufs)m^uRfsa|>oT ze4n#JZ-0MHkbTY`OB!5_?3#2_NEW>=bc1Z|gcRtU-4Rr->!t@uqxg_!?Sw&~-(3~7 zlV2yT65<#mT)V04oOB>A7t-LersYjTwn9%l+~NPTEYT_k{7V18jtnUVAez7Dy zK1C6w!A4HizCSh#@W=r%3;N;*QQs5tJF^`S{x7m;nCjV}qwc0MAI9h-` zC41a(VkBtU0p`F;AD8y^jKA6k8NJqp!xB(LFEDb|RnMwq4B8CkuHY;FdSuJZgM(Hj zMhrIlJpuz;UqEjOhLikd@gA2y_^M}6g#7$hztR8CfBf>tzx#!B3B|lek;G@O_sFI6 z&}U8v8UN$AZoc{1PkKRgn9O0Lk^hbb3hWlqIZ)W$GzN#0clv+i~)kIU-8P&%-4dfYRq?P zw_GG^y;zOeW#eAkN->Z%w}nbb4=DutIz1*gq9$_|@=|j6eNidliSY+!1EX2vG}I5O z1!qVyXta?6(yxZ->V?m zUW}3|8W zb(3-=`TQO94#l~k%E-#dp4cMgs!;SCX7g&|jJ!_JS{vlMr+Y+zj2<#i;NuDKaDoiA z>=*0)(_U45LA)(2iY=xwSPD!mm4}$7ogb75Nt3x$MIe+3fRn{ysANpAjPMA2o?7}* z;OVJ3JaOXv?)U-wIFQ#GGc1uEtcbP{0ijV_CG4VMzd^O5S+EP>QQ^IkYZfQm|cAJ1Y5*X-qXp8~{PsGWHPwqwnC#f>-hfMX` z_LyN=8cRKLzShEd(I_mRQt6M$N0ylHhOlm2pB7=KJRgKP8l_1P3L{ZbLA4?XiF7htT>897t>kMnsl~3KE$_Y3y`@9ad0q|4B_Ve+c z#VYI9%EI?$I*3F&y*s_ed9^3$J_ z&D`=6yx3YkXrmgJQVfuA79fofx^Elh)#RGAoxUTlibTi+&6|GY)^HRzA*+$()W~DNdX^^K#&imENav?(kIXUpis5?V9 z1sTbbfO7E(#c4I9iRl(aZw)-HPLXs=jk*%ix~Ug6%c7V1E?3@Eu4mHX?oT%AE=)S5 z%9@r>52%*;x=$p!0S}*UnDD_(GVoa;BL^ORiudcReIhG*-YuW-d$Nj~u<>3o3O4u1 zpwniG*@%B^!kw^2NMp#EdOfUBHk?g8#Rr84NFMK?trpc6%p!^`Qr^eXs02>JM3&^8a5#|k7%qf4V3 z{UC8lb1tZJMi;$TQp>v?RlvV0#d>HofpQt6u0ejBP7dv3Z*V`Fb#oOuBe$B-d_UVg z?9izQ`ufp-Ed4ekx`2O+ED>ywb<^ACU6byZdsF#G6h|3w0I~@_Bk{tn;9Oab_OgXS zh3&bz?ffW!GyfFIhDw~|UL5g`qkv=Fpnp_`l)UEuS z84G7&pR!OXbdh3yFS*QX6K@GB;O~{*Qrvp&9KF~#5!}y7f}Hqcq+3upB}Y;O33c5B zyFl$x^ht{)OMM%JUBSzIv*<<9ABh`*Kggu`d%$w#65n^%zwzKpC%$*-bNlDsnOppo z3t!v!R`R?%bKBl7_;Tj=ljo&=d+5!iZ=Ro5Gq3ME7rxg2qYZCv{PIU|z2uE$Z!i3O z(Hl7S(!9Q}{HEy36?1>{+ZJ)QKU!+Tf9F}%RoP{7Sep)e?4R$1Vd5zMAx0gh^cB*# z1;>Jh&o$}RP5RGK0C1$R^4RAebu61Ew*v6wSHE_Ztnp&+O_|L$mP;{N6xmEAw8H~W z6=1>ZO>>Z`Sb}k+PCA_rJ~`bHiD_~l!CJ9W=ve`LFXj}xB+>k_V2vN;vOh~?xaAuA zhULIXGcNJB`Cqjs(0oCxC>FX+OHpKQNSWrqY=ioY_9%aW$MoT6EvPg2dDi`&x|aB( zmS0m=u`v)haE1)}aD$ZBM$15;jMSA=C}t%^mQe{vwOlO1Y`PLn15~B{ufV|>gW6L| z{_J_re__Bl)e4A3?@n4pvb=a7skB+*ofK0_ks>PLIP4DFu+t4Rgay)WL5jer+cNd6 z%Am%^!mIhkue7S~$9B_Orgn5_LN$&G2pfgTk zM5TwQpqvp3&rukjy_loC%1HuHQ(mh7gP#>K$6v4cD_O)%_4Q&(k6askrBh5AMN+5) zOG7{Xov163$H=Dm6VP*_RSEePr|A9wlY0&2=dYmiAbS2>?i^wK{GDI=T4A9xy{jh+ zxxvDVqtBqNj_|^yQA`R&R-*cv%>1#6UPf*z^^b!(u^XSz5VFxz*%QxLEGIGXj2S$0 zY2Wu({MT$NCITLu$RPW;VZw`pgy(HAahzi6C{jx$T-0n2dLNn{Usc{!w<@q`tcz|5 zL zO^UhS26YPWP8j%Ol*ehy{O4%A{}*Z0zJI$rZ~xqjpL_hdin+gVMZ<9tNzd7(ST4H& zvOoU!kACu!6-lALtej6ae8&9#6*hR;Mltym$)ys|x0XoVgru#?fK&9X*IK6xgck9V znar@m!b-`uz!RDp?J4lyc7_!OgA;{!Xx@S)Z~h7rS7ggW)2D)GSBqm@S@GvAife#) z&g3p!X4hHP-7)-^y?(;7u{Zc_l@6M@{;+V%oZ~v&3CPn``eXB>$us>-> zE(fo*4|lAI@Ut(Ee?fd9EQRw($=)I?p${<7>qu+nUb6HY>Lcsp%KY24*e$Daa(-Zb zAk>s)(I2QQRCNmMT#Y{-lqm>Tr_uh@UgA-(kIR-$dj0v$h`iW_VQEHWg6i~U+5K2Z z^EWCiZ5Q>M{q~5FS))Sbk<9&PEfkmKf7U$izQ!)>CaCFO{>G|-EL)y3n{;}yqhzH` z3}=91`YG}dspGI`84@2rML-o0s$=;8mU#4%^{*sB|AgbQ4bj-52-~G$SQ!#&#pjtu zV3_NP2L?+mUcVzx27zKpy@l-A33^O^>lSXG(n(`|P*^I^QztPSkrd zka@0IX6|jC$t#i?)K~nALiNa4rLURr-VB2}Q&`5=rwQ<5JG85|j5qN1YH-p!ENpG`3k*~_F74rz~QvO-MC z@3`r1x}YoEJ+5mUejvoBh&MKIk{u&^NkO*$`=h4ri^rP7hW0JYx}^Mana# z@W>^0q=fIG@7mkQEXE~t$N@svuHa&Rr9Y(5$Ol5v$+?@CB`v4x_yhlsyf1-kDn0l1 zi06>J7_t#e&H#c$7!XGmLq%+yneNkezP4ZQ?Y+16ws%_T*K2#*+xAPRo#}!*DvJxK zpaxJDK|ontkX=O?H^4zfaYZ6Jild+mqQdt)Npwggl0yRDaXS8fa&ndz=e++s@AE#( z|1aDnyhU~Z;VT5TLf4VQf-+^JIM=O?!N14cQrr9H$NNia`^^vI@$onK{l8a4g;}#| z*dVXCquR=xl@w2X_VZ78P^&CawP??1TeZlQc1F=88c^mkHzP8Hc=fJVtoSohPT}zZ zxHDUj?%%34T`5srm+c&rjN51-V*#Nt__}^UAXEQY@Xoi)F3A%?=D(9_9=jwLEnJdQ z6bmgs4O9#=ye$^z$zW|Mj94aIBh)o3up+iJXniQqlE!1d@v~QZINA?uw6mNXqYx%9Iw$ilzM!mpKb-X#`i$8P0$h`2kLlp?;VC^>glK8&v+a zcTSVlJhnm5;67wqw3%WzQ7{YUj`tK*qA-!!G9_WAQC0>U3iG9%Pz9yCEnFXp45y{a zKKXdEM zLBisR@%g|F5GA}?X<(Um?L6{_b5LpeMbtfdbGa6(!nR`&|Rx2!a)g|~Z3CW!?ptw2v6Z)}e z&;_4c#60yzB4m6Tp2?N&f;Ns!XtZe;RcdOXur_1b5%-=*-FDIoTw?8_1vzw2WT7mP zS%9iQ`=ndFQdD`&33UQ{LR~|S(hW2)c7q1vjCi*m!GhwbwZOZ*YVHEP8dKLV2P`*) zJ%^fe4qk#gZU|$p|C_gUiw~EI1^xyQh>eDzo#G0pcRDXQ;&)!MOPS%C$UqH4aA{D6 z>wwE{Xm-wZ>sPH6o~D!LuAJPWXb;;Og$6A>aLfZrcE|dyAoLO}IO)MZ3qsVz*+q0&3`=e_)6# zs*}(EZ}Up^R$;vl`DD1-Ngmf<$5~V}^i%9-6nQ|!WC+WFD5E{B3k2?}Y28LJmhqpm*v2>%wMb+lGF$y|Df4!^S1n_KCOKDjOE(7?5)m zn7ZM2&=|6=2ucsaX8DWEvKja2Vo+~g8C4wE9<~5k6&K*uZn9-8t8&eY$0*_8q@~C9 zFt*_^T%Kd%_NNjXEr(^4DdObwXny1Nty~y<0$AGj;k-@nr^-X}W;a zuv+jm>|q$zKAjCU$g?%Y(A`$$dmuu0)hjkyuP&8jNC#ZX19YaxJ5(rwgVlV4S~lud@YEd_b8D) zZ_p(fGzk|0b=81!ZB)GH2?;)tPG8gLK+TO#57DbPN}Ck5P&{U!ugb8P4Edyy7)^)d zj9ABo*`17Oyft3^-~Ic$)=~mo@W*36+lKH!2D<@dVf&y1;Xc| z23fr@hpuv~P_F_(qce(D?Nu4h?gK#?un1=`WA8@s)u#Q5Quta2gSeoC$5zUQW;N!u zsv?dF(R2U@SsLmo7R7$b18!x0{N&-_tGTcpnw`zL?Nx+bTRUY@enndlu;!;amv>Tar z5=u^EFn@o~tAWnZlzQo*zT>cWB2(wrs?D5M0(apZx5JQNYAotKxZz-^r)qKR zbX;(Ae6|D#cU$K$ZKMhtWWA(X(n;ruO-bUrMyCMSF2wI+gtv0GsfHi69hvS2rsJk2 zAC_!OG&B6Q)d{a9AC58` znvJ>246>KUhNi{B&>W%ILlmjS9Oi;OO{Oc$&n||AVA9xR!3GYegkq^}*H5O+}$w>U4)zEaxHI z!WJ1nID&aj2&b<8yrhaX1Ihhy?*+1v$3WU+0i;}tg(jy=D#j#~hVnFR@=oO%SDoI! zNCGs?(5_PDhH0%d5puyJw{Bj*-9V%2H0BzzG#x5jwJry&jdBPIBTSb=NEu;lM=o|gDC%P>Fg>z~f={MNt^9?q}({;#^}%D1n`>f~#I zU8aLMO0S{1zOwZfpU#KYrNH|@z==<_iw?s*$H}c(|A@FcR=na}soADviZ)y$iLY3o z8R&$D6y4}4b~8mbA<>5^gbYf-9je`K5PN~(?IG{e?gsivv~Jbp4nHis$f9=#^a#rQ z9!wu6lBv;{7e~y7z0cYO50S7OY5GU^+-dcZM}9@DxgO6;NZjV2n0>?XXWe zA~C{%3tkRXS72*CXG+#XeYWUPd0w_l`y~KL z`k~w3B#Mov$Vw`vS+vc!M{vna_eg$Jy6p8!Ze#9;3_or5AY%9@$M$R}y*$Et_rk?^ z@E8#`_d0qtHWfTCUpXW&$g@C1Mu+tu@zMPr7^%NT(mZtCQJo-B)2v9EQ6vE|D|#Hr z#q7b!b5mh;FqTM3`ue&@|6<Tipmk~5P?kHuHcK(Xxj8njV*kHm&N@$M0%)908bMXno4_;mQiPj1!r!kL|Pmju7E ziOfdn1$NPt-dUN`4uXBZrrH;zS09OJV{tT)sDsZhj*sKrROB+fvp)ZQpG93{ax|F& z{Fh)+$Tfn`rG$5iusc<+z7SkYZ<(@*!Ldhji5!_AtNG&gafI3Bz~T z1I*j~j}2wt3dr9X-S4YWSZ%gQoG(RPbff(Us`5=yp&yR*a44In!`KUc=oaK?YDgEo z5OsY`T~Fw^kJWnR1RRfm$kAQqgyV^!sh|CEWRWx;FH$x{(hNYa+^OjVAVHlrP$t&~ zuL#GIZY)#;GRzd!*@)C&9d;66Y!(+T)CR)x2_#yEitu(CBea<6!2!0*b_)!^_+7=O z(~oIvjB50bXPpTxC%|c_IUF5!hIz_1oW61Jnm=t`(RzR2f1E7l#!1IxP-R#w%Jwv?oyqXo0?2UTJnZ@t&X0u4=#@W7VghSeCdCk4mN|Lq^o!_ z$>nh{zusb+_fc#aMT)7Ic<*@cn?Tc!`8=Rqcnb@t;yv@cmWJ!)O^OXZAio&d%Ieki z!39u#u-@}yU@3!Is`uOEpGsCk=(dZ74?D%%*lbNJ+pH*H>X>bwYlTCGoO{o8DKw%u zsr<@5EDW^{BOY^~_+g%~Z90Yv@2^)qDw%CY&d)xcQ%H(=jGRUb-9tQl}wKKm2p zfJ@TUJrH5ZU~_1s^0*_qW*n(gi#3h^IAMs||F$qG(2Sv@Z>=jQSvK zuu8fLYH2eyhlO=aHV6UWr`4p&EmmADSS)^toReujb* z=DYCv4v&5Tjy49Asm_88H&BdxH#gYSExEhSx<`_?dnOzGkhkY$%}Y|Ddy8D|ZE|do zh0G$*Hs?VtumSo#@}+nM7BX?Iyo0Hwo1lGhZ&aNyZ)O79qCPLzU6mEk#`dFUPw4Wc z)dFb3r(4wf1O`zeM2lCrZkmFrsV>POXc#d-UfaaQtjhwQx;iu?9Gq+H{L!_laSw+* zLb`U3{UOGK6|S&2Dn16ydpNsmPLjLvAl*M(+yHwxpdNtk-9`nTgvKZ442``97?jql zP5*bfKuP_v-#U*$7lX1#z}uE0PT-s1Iqinby26gP;ImOPcOj&W%@b^wRRrvpZ3)^R zTu#LW_W+AuRz#-sfjo^_2{sMM(m-$RYu%)SF$q#*)JpH)652PVLgV-f_A;a3fSG#P zunu42gcRz4=EE2ARJ=vXUW?W3dFH9ly*aeXO&?tn@@(&&E#B_Y%Ov}QAi*a-D8Q1+ z-Un(shVU?$v7YI5ehV&m>=rB-GrTwe6{A<}FeYpazzHf7zwzNm z^L@;YkYQff6|#8}d>w}-cy>}Obbe)1F`EKUg{*W|HoGDbI#+`cF=i z!rz$j`i=zAJ7%<$PI>ZE0+Qwi3$&c0Sm@_yreZRfYG`jSrw<0})u^pm0amnJFL7MwN&w$>&*zgItOHUyA@^ALBdQ>;d*OkXT)mo15(5TF8kzF!tJC`@F4Pm z`^R1bF0D)f)LbOE7Wr@YXq|j!?tn{4;2M(Uc@~a(0|%AZFQQlPlOOhNpf^XiYMT^C z=^{GEKb!9GgW#O=yE>w??j* zbV|}HiJJrvWmimvv>7SateqMgVxlhYQWZ)c0z3V!nH|isNrloP$?{1{CgqSEdg-L2 zfj6`(CS}t(bfu_M*{XtZv}fh#NCLZ&X;rn30V%nY;LHiC_T0=3&Z77w`~I;a%f2-38$FaW^7Nk~>D7Ke^Y7l6S#cG+Ha6RNLuc_9& zzUp<%Ulv4mK(YXeX|g>E7a9!AL6&$$qBJM`V6ZMvmQGjsp7%d|*(vTNn9OJwCHX({uBCxa>l(?O(d3FW<;cNxbf)f*;Ms@9fFg;m zMR72=Mpy$xTj}&};E&Yhi3@Px6xC%}Cxnj@nJ-8T=gmh~&-j|Lr%6~@L$-clQlS+V zR;!R=p>R5vim47wQ5|?Ai|z+1nRGf|ns2P`I;pn!Pu z0zCZPFcKS?(+6}mm>7L{r$?(6dDF7#HL7m?4Yo(P_mk9gh z%iy?yuAF)*{ElG2WiflnI}d&q$dk!{%N5lX@5_+`E-hZ2qzEYQPYO0irgPlgjWl77 zj{A%}VF13|ItkbdhHifyLIDfEX? zK?PeR+r!q7S{i*}b504%wE8{P2v10|z#-0&#X?x26GVKoB zYa21?o=<+NG8>(ZYlF{|4U-5IQVscT=TIzk$Y)S7<*-dy?}~}%ZT<$(ZNy?qT?>N= z;>v(Lna-4y#%p;p;}w)CMtVxTBVSzQJ7%7u@y&34_d~y?4!qlu9<}}Mhgk&pOJCLc zKi{Zyq*aE$|7XNTt4vO~Y2~U2Iwh-PaMQD$zE5ADl1q1}mi)kw?0zda@EcE2?C0;=X23YAfD|)ka{ur2Q&QmbdA$fjm z`&EwiitQ;F#v$Qv80*IiH!blpFElOh?OjMxzc35UK8uAWk78lv*-pje(mTa#NQYl1 zZES;TW!qGXga-n3C^wm+N)c44aji-AzwZtOop|%b>4#-)s%}!L(G@doY&+ei!maYy z`QZ*jezce#;kjbv=D9yMVu7fURAgq!o#es3i<7^k5TivyI*5$ClMS5 z)N?yJ>VsjSW!r{+{F6WKlbNBW{QFm0ND_~uc!d_QN~hQ@6iEd_w1{RJ^(Zid*sASg zliWHZ@}XjDgU@2JNZhJj2L&=uqPu)DglFXjWu35JGzQD(Fzelc4 z&ARi63k?a6_sll>q+(|`EB27}5T-r-_U@p~QyPQ&RhjNRsvc#Y%+#6L=i9G>_7|p) zAnyn^l9WpgzGqxBr3TU%+#`6ZIy1LfQA6%3P%|fvRMXd0`C^cgbH1pBT_Ce#eHjro z9B+s(9`O2^d8t{pb@4QEg~vH|>Ly8X)P;El)LH}|^LhjDOkR334$t^DejFX|FW}Z0X@~W3s z#U#^J90jY4KO=hrI<>uT7<@Za1Il%ze6C*IPhXZDm&B3F(!+4|D7|7*qjy|zTWF^e zuMT~_Ra+lyx`u5WPh}9ZaIXq&_T1oeM~eSC(-RrKQid|2UHB`)v$u}k& zxhT@KQ7Vr5Pz-9G>QzoG6fa&r|%R0~=lYk84WhgQ#j z|F1y9wQwQ2;n^Aky}_diTFCM5d#;sgAXS0(&wO#c&(0uZ!p}8cZ-ufP)1P&cMcNa- z`O;h(Rpw0RhOR}N=K>xWGnkC#A2Z|`+h(d>S?fR944|98y!2PHlE-URt_4_9DHehe z>rp!6Q)6}SpbJ)H-ScTww8<|?lKr8iT-qh6Wx)LSZJp-m+zxr2BOY?X!HASeS>^f*8ntwzMEu zH(W%EQwKConbS5C`k+hB z@=}NPq@}{xtG4*j#rh79I&oghG~ddS}G?KDa$X;jQZ`eN`h=xjeu>ZNFCK)r?8 z0VRm1X=Lr%<9$q%O*gsr2w-a;iP4=#QaN{AmaowuowJP?#RJ-J&KJ)yo0W9cPx8qg zZW|~5f}ebB0g-x&g*bE#71JM`EI|LY%B@*c>X%F7{#$p-^?qt6jJ<$6zo z-Es-ig{T>nFGYDE;1fi3_)ZP3XMJJUfCzfD%z>l9k?mD#xGSRo5j^tYxBl~ypBYC_ z+2bpT{uSFURa&5>h+?7JcLx=7BqGT_3rJ_tAwfng?2GGC-Bv9YUzi6?a>y5W$a}lT zaZ(hK?7tA5ncNu%r=N$4FLY~6=?&dI$b1Lg>Gq9W0azcrKD1u8SPW(JCL2>ZBQ|SX5SYny6h=3Hkzuf%m#T zvfJHsPOpZ-u-TciR?W_+HNsBFQy~KZj$Iv^9=PANV1B>g$(NRdR4BK2cFe{*Aa6P& zZA$*U&&P$%PM@lXOpgksUcFq~7dfEhuCdDvSHmqqcAn!1yY{VRv5r_A9^=Y}#pxDy zDzhRR=o=CJswdt>K(Ak-yeV5pnneB4rIAac?~iGB!i&?x0U0koJceg}b~&QtJ&nXJ7bccy-@z z%!|?fpL!_BZElMZkM~C_EhtPLQ*0ka9#S#JwpyS>zwe$qNDch>d*5pf3YoS|_JGCFS*1BdWDaFE4tjZ~;3*B^BuA^6*h}J>Y zL76YTu0)O#-DTN;5);+~N>iW-2xp~za>!<3S^^nlQ&g1!_vAo-4Ez&%$R-#Y4s&T; z3KRz6rwsU4nEuw`t_=r<%*AOsL)b1_7Mx8Z2}-`W-~DM=F_RIF|APDO`0qCb9Wg@& z?iX)KObo97ony_T^#d;55P`T{rs&vnwN&|2T@=<8eYd(ZV^2oybD%Xg?KGD_?*r>tImQ4kkt#j!urvVsSHk&qohM@bh9EK>i1TNZz^u9pcMzxp|?Ytlve_9u!{t> z!c$c7p6PV1doF!aeF?f4@mH1_b<=q3p>Q%y6aM(#pZ;d8m>H#Hkrep(c<90!5WQuCJwgh%6&R)ny1A`%m z`228`({1Z9PBvxYVpZA->pc>0-C-O1p_K4+>8)8wK(%)Xx*q;|U9}6E93 z$l1W_4U3!intAmJ`0uI(WYZ+F*TTp@7eVPzeI!O+-fKXxdmC7POyz5JeuFO98gw*r z+3UOA7KL9IEI`2lXg6c6o6}$HkQd_6J03EQvky6ai}@Qn zT|W>y=X++T#QnDNAvrl*wF8g6)!P=hzC^JXC~^*|8;yPk6bWv8eK8Ok8H`&SB#-#u zV`)FAeLaR<*G0)ii4IE!+vG-u0NlUeVW+pQB;=U9O}<&xt~P3sHH$icYyd!y8>AGs zTzQQIj8e0rOM(a3&v|A(_H4o_{JqGRgq;RBD#V6QsMA1K>R`}L`A&Hk?CjH}_?z)7 zWjaC{1GHW#lc7@mWnU(Tz6>^ukKLzV@(r^6W%$;F_bOb>_$qkxUmucK9y`K%3oNan z*aV8KqGIyJ8015dsC@BBb*=m~NPWL-mR|Cr9J^%C^gH$&M)$Ps=^rktCqKUR7uF_+ zi$}xb?6nPJw9LCnxs&da>=4;{mAsY}IZeHC}^smD>vO#@W!m z_Y z+A}Y9ydo~Q*f-5wMs&`-oQ%$d&>yGWv8E}3LhB)YWl0npPmz^W3>HTs^%Mp>kqGnu zld-p_QOr4SlH(}cUOeGz1S8|!f0_R)vuE;QQrsKl<`>Fw|js$oC8c;DXPC; zSCv2<$XwX4*L$;1wFmgc?V@f4KD5 z8654MW4Cj%megzC^qfkWjnY$(r>8)CgtrEi0KM9f^)r!TS5qVoT5Thaxaz9hioFUW z_ImdV5{0_)MS^+0>;Z&hQ#-n+IRRo~^>~zj6c>+)@_iV4%)4`;a&{&d%L3BSX~$JFdi0egB&>yv|GFzw&>^~ zJ3q*44~*>Eyo?SlHlMN$qbV!T{r;Vf275VGIFs+=dz2*2aL*ZbJ?(nw>c2WqQZcu{}-0g+_R5>p!Ls*1hVL zdX>S;p$NuU;i5@(U9H-uFh+ARnFNxBL1;+g-E_OAG>dL~Voe`M1il0WFS|4v;J$1; zXItG$OGqA%;d0ahF69&psnk*`rchSs*`fShHeIZW3yxKF zDECHXO5;g~vRVCm)3p}u8K}2w^lkzXu@`T`Tx&I5CTp2v;(|;GZw6J=0to1$^3!fj zs!!R>{nMU0v9ZYs*cl2Xi}`ot11IcIcYpPj54C3O7{aHfk%})&+~&Lm(vDMX0|oMy z&~AHgw$4Ow4mIV#VIs|zfoF^`sR_O!OJuqf{VLSy)}i@2Lh2)TGo{cMQZq+)*7Kt| z*oguM*>nv-duAk9!TlGjNg2kQxd7Ouqp;TZ|9( zC?AF$l%A7uwQf$v-cTG`9&-eu&+n5(3NxCHefv9?$vSRm;&Entw*{KEQEWN|TlN@C z9HGyHRQDjYoG9&ptye9S?tLnOfL!8hRce$xEcGh&Dhaxd(#jpGUCMnkQznnqlJ@zG zjT$7M4>2Z3e|tByI=~E)-|jk`M7DDS36C?Dbrv8grC2~sA;fULoaECo$MkoKYPHWM z?`HLJ$rkfcOP?)a>qwgaWob6us!gXaOVepQzrzE`i|@V# zGTX`=K~E%C!!h!Njv6%%8IFm@b@;MWqZ zF;LmFc_ojgpW$jNr7%OUWgP0dl7us=t(wuE0_P=PpUwMfV|ign!V@NKtK8V<|Yo zU3c5FI_MGp&;^58rI8uoZF1CavHQz18o*oH5pJLC{bq0i{KVq@Gk@@^ zS6y8R=uo2y9bo6)Y}j3oO?cOi_irC0>=?+&w{U~R*2lAd8fErjHs&fb$X*_Axmzr} zk|Pv*h$6LAOgo94+Nd}W+|D3IIpi2O+DAd_;AucT(-5>bq}O-{3M{mfZef}1hI{&d~=&11?RX6hVOykMQi7x##Gf=yY|C zX7{{qy2LT*W&41{&3-()+YxMW0uq&*@XopjGms>%^S6`TJO&b^IfkStYbmyhA_ssA zjV_jUD9fp?zy{_x=!iZ0oJ4(ot_4cxFn*Ixp95KqCeixIiJ-mFpw0lL#WUV#=At&o z4v)2=WrB+{Gc`BB)r}3=H~)$Zix@t5s5F4q)JEwtQUdLlj}*6N-I!zd8uXI691{q= zdWk2>}ufM>IREL9DyaHaF(* z;i7HmDDLjt_H8pvKFm9Ci!@9kH!R+KE5)9n$Z0AjCA>6fZxBkFREOq7cK z2b5Q3J<1)%>-mDEj9vf|$Wce71%i#QWA(0Hb4ArRrBqopVZ@)+ICA*?P7faQVq)5^c}cF!@aWeuJOh|JP;K`6@0o=sXVl*~o5T zFac{T_IexSklg}qlGg@ZFz30#6UmK`JD@uBU{Hf>sYi=vj{xG!)uDy2!qx-n^!x0+nF)yB5M;JQGt&RQmgMl*=Q(U)BKJ~kDMfZu zF=u^N2wSzM$bz#XgJ?mM{Dxbrwnu3Y-SfJxXae3e&=M0amdE;G@!l0#i?;CH*7?{z z5U08<%aHYj9hlz}*`nwTn*SZ(s3u0$%G=QUxsdB?w3v>0*a*s=rJ>=HR@hPVmI)(N^Xy`NJaoqXF z-QQ}=*5p{s-`10oSIqrwvam1*DHb}pE2)?)EDMj^HD@s3^CNVItLc+I>qzpP<+HP4 z5yGDxvj@!zh)t5u@_>rd>ME|>P>p8NX$4gSN1z@&Q zEEtciRLm-o2Ng~c>VS?-EH5rsdGMDzh?+G4gr{;{OW?W+7^AVI`U_;TUX zfGz4LoYT+TXJV9y;Qos^K4BXX!#Us;vH$e%)^&R?mJjncK>{K$LjyQlDK?EF8&NZ^ zOr!IZ7`636GIuBh}ihfL%ZI5bf4vxE|A>vPC)+T`hh)`JJZuSVyBP@xmVnp`NgQ zcZAQ*aq0HvD|m8?{`?`VbJVQV1~u|#zsB1^WQ1DNmYgK=wJtEu-)kHn@s)LpJ} zD+;O#%@aI;yjL5fMGXXQ9mVcyRFJZCVeI(C-6Adc(%0XfVTRNPA9hrd9j}-Je8d7c z`zdx0MM{v<(I7I#Ap4>k=-a|0p2?($t`gpGKL=g+jowSd)slv&Dz`Ormj+<^AU&{+ zZD8Vpt3o?~N3UfL=Cu+-@jjDccV|SI45gS$BkKI{+9tAS$|zu9dxD$^5Zk|n`xV;@ zR@}m%bZcrO-d0RNq41FG({s+Hb`e^azg-8Os%`Q#dEXS+;6SvkUG&F1D)HSvUjE76 z`Aeufs?&3qnm!uf<2p!Wo}Rmt`s3cYhXwcOr^1`EKIdX#$SgSwiy>n<^UyXdhVy+V z{HV##y35u^GiwpdN5A{#=zjMcauFm0%IVEwU6!U_a2J?Rb)6Bp!1yvQ9;Eqgahm{V zzT%gU#c2Zg>l=nn{@44~duJ|QFOQ@DHg?)OJoIW5&o=h+ZJAOttzV#LcQA=ehU*R0 zpi2itX|R-PsSqm>k=e7}Z=bw8s*V|}NdFL=aRdA34#|)q#^jJtr#E_;y_A;s_AVr; zub7vz&tmz>qu3mZY^P$%-S^1O2}`{TJwHD(wW`$8rNBUhl`@#Y#yeWIhKLnYyXiP> zp=`4Rs$)pJupf3zhsQJmGg?F#u`woCFb!(k`~}kjT4=_~jMVQpkmZv|mc{GeK(Xs7 zvIa`4-H@vr`!kSrqfOpNPI}-r6m=DEDjsl46FdGtYyZ1;{C9B~!S=fzX07M%QgiET zKl8N?!12~!wGnLF8+@EO8?l=7lA_2e@6M0`m-|5XersxaU?Q_SaFe{tSQ=4CI!Lvo zj&bS+X`++sCmc9_uFWBQ%OBX3mdtwcXUi~xxHS?~@9*7zx9m-g72HQY4 zkF{apJRi?ja%aB%7sui@AH40r|G_t7|ChTNDCJ*!=QLS8oHxQ_&lH-hhiry7Q|u-R zAQOWuYg@fg2?O_9+x)8py1m}m2Q%oBFUFd06Olx-2+2{$$+mEZ@{6z=i!-*Nto+Gq zt*?4Rbxyb`7T+(>9Vcf)+vNT3c}#{dHY8tW(wsJF7aO@>&|S)qUiP~TJl}nCkTB^` z-__z7jF3QZ7Mfvxyw2-9XEDL{KEa+DpYFW}L7Ko`*q0tpM9Q!@4h6A+>a5!Qk z9d-?^f%?vPL6HO~nl)&7@}+yds|EO}YphpD^Yo4aHO>OYOUJQ|n)<}S4`-SY^5(U? zc#_ZKS_9~rAF|FIpjg-jmr*guQ(xj`0DYnq)lt}PwF^FtsD#Sk)8Tm$$cLzV)@^we ztc{6G3g`<_JSj_a-7A-d_W9tWUiW;u!gLVk_brC50U$f_G$m+s?eu}bI=>FpO~nz{ zI5#IPQ=u zKk2r66f@VoF3daPYLM4?cBq_$1N$b13tH?O(K&b7MoZU|sz;N}X65EDFa4FQ z{YytA`*`dD+mP&AC$1L8in}C<&y^<4-5du!`e3Si1Z%_#@h`m^PjH=Qj<5gO?cCw< z?8&kHdH&GWxL3W$17s$|y99}hE|FQ`)1+Jie-FAGmUXC8gGxf$)tBc0uL4Nl7?t=8 zvrf*?tIMLc2AxxWN&^KravnA*t0k90bHbZ|m%B`P*xPuiRh^=GHZnwEoQD(cP)4Ce-2t^}NN&v3m-s@$3tD0Hn?u242EIs&TYuK#!Af0%Lf9~E1^PA+i65sx=YeHJ+S zgkn1>(m}0kY;|dc(3=3Iolx-&`D#^Z$P=<)ug@NF2aa#U)Z&1xRuP{@cC*SwaezG;h3W-P&G1A|tlhlk`Z7H@jP%?!2=bZ7qn+GM;w`r zIC;Sn|Lj**FxLzX_lFHBWZ!UEB>pl&5Fip6k~cU?vCv0fN5$+EFPpT?JI<$`Sw5); zdJ>x@o4f~H%DgtR>GWMFfN9m%%*duw;5%I^Y1LL})(A^jRPj3uv_Q3Vt-MIk7zrCvyHA<4Mp+nGbBCky+AfjBG6wkWO8yT7K-aKshB?bbxDt`)g8%S z>g7!ep!b)ojk+4u6I>A4PCq6ckR0A8xG^KiJ;@&#nsy1WduIpTsldBVsy`x*z1)%B zfCDdoq@f+UZu{jAv%%%>cedhq-3Q+>BjwiYx=+bR+ zC{~Px+Ng1U5F)(&vIGz|zwWhDd|B2#D<5KggR&~OnrS;C23$;M5}5;0iC^--Dz!xR^nX%nY|h;Wrr+G3cZpy5$Aj4P5T7S7*BSMe1rb_@g?sR+BHr z#-#zM%_*iEgYf4K)j@E-p=hT=b&af2b};B!+5fe2;43nceH!k%-CrAzfAaX2z>F=I zK$pd2*>LL_kG*Fo8XSV46pCF(kt8Z+nGl)eVZud7+y(>U**g#`RRjINAWX&Jb50ynA8`UQNOu573G>u#9F{+*629M#s z?_-m*v<;78*3SvQeEsX^ts6xk&PF4sE9!2@;+65+|k`h8hMZAZzGw zPCGuLT)p2qNy%IDVk7NX5_m%0quh)nbJ$ebpw12KU`r!`oqGNkm!efuuhFnRBSHJ2(q(OOda_0cJ=8or z3mxalAha9r)=NHdKQq^;a})1o0QCe+IdO49M>n&hfRt&88r`q6`>D6*x&Fku;*PgA z#71=;CJ#~MGl$NXB0&ZOS54xPO|E&<+XA4QHQfF^hTU(@_KjTT+V&{UzrxLYG_QaA zwsor7MvgZF)D#Y~yOOuh>G_2wRx|)Nb2GDtY;1=rJ=jBDheA?u;9+OkK{`~ z@#;~2M4(y@YT_;ky4`EYxu5}fe0z95oi(ox_D*-fVBL>&aN*&O@ZtA8!<}`+{%}J0 zglQ)yKKh9{eA4l~?|zrGO(KshA`-VL_9jIPSQJ|uh&eIcg?Sg|6$%Ok&7RdFs2NC; z*9PitOkOTC06ET}%h`y9X;XFw>6Ust0(H$ax!%8IhOShJ3JnWSxalh0nqaSregKrz zRWSx>xB69SQ?hA@h73L37~HRlcf-L} zv#VKfS|L0Am-kD)1*3j`WU<>6BXEiH+ zF_x1$+(g@Zj@zX2HrMwp_l;rJf&yH2*gSTNZAc9~(*=9r{!|4eSc{FVl7lYw!3A_i zIOd?Lfytp0;xkY?44d&TNex-y85`Z8Oki8oW4%_78tsnQCp(_K{k`T~Yx1ZUTT%GC zI+MK?7W6rx6jJPN_Wx9L#P2w>W^Sywfv!~^Cx!?h+a+~=J<-^UaxfT^WbKkY0o9?~ zJl6_S!k2lk0n>&RGK-kijupp@^zKIu2(V)#jqS*F+vacYgMVIP9f#p!QFxpfuo1P% zbnk&;EM#$5%C>@-#cJ_WRfjrO+#A`!91Mm+ERWjY9rX3E!7v9M94-^a6*yeRcI4i( z^>MGgV|H&!r#$&7Im#{b$YUF0u&^=bC>F%HnvvUZ04Ohezja0S;dRxn`45$?+SK5~ zg0#TJq*AmsD%5#zpf=p?PcKf%psrOTxg zD}9sP9)RQBDH>2FhCYN!i3fsOp!;4i`w&wbbz|mQAi-&rwrcwz0=!0e&*y<)XH>d$ z5pztF>$aFppRQ-p11psE!n=yx8C9Ad0gj+oe>A1flgqls6a3a>Bv$<}#-o zUbC%R?T3Y}5ys}c>I47d*6HXk$a-WWQGMGpcj{88=+Wg%Q$X(%t2D9VH3yinB?=smJ6>JOgFD5z5lLQPPXy5WVhC0 zg(#ueT@)$6if^plMy4R_fi=hOfqmUd)(IV{wzjeJaFvLKy4p?)Ab z?WtD-(WM6%kOP6$f{#SqQK)i-p&GrqT!suo&i7`-lJWEo)XX z-dYtKbZO0sPN*_SQT38$dTIDJ4gUR+@px%?u^?aEL_?_){D=wBlg5FCD{23RpjK*@ znz9EZBH&bkW*ylsD2O~NHD%w&5MrYN)^mSl4CA!Q>@U9{b;boMJkIXepli>iF9;4! z*I_z(57fF`WFAB!VYIO^7B@wCvcZTpd8}A>Qk^GDg`T6oQ-A$ilEUMS4oI>NDbCNPScr|KQ!(wLeS%daRnsUf zoUtUNP8cuT5P&ql88X<)FdJs11{*|8kQ>htG=PW|_BeH@_5`H^C2&UYjhSOQ?eOkL zZ697bZ;oH&wD3^!(qCS-X5{8#O?Vt&wZY$=ttqD$i7|lEN3i;~O}=I7P9NA7n`l)Z zx-SwZ``0nKbRBcZyMP*N7?~%WJDMCviajIzAbQIb>mKJ9lf*n;Q*87-UzB74_gso< zvpQv(E*9FIcS#eZ7!cVM*iCfU_LZVK@a8y8K28KT0N68i&bi07IobW7YmzmQiwzm% z4Z=D|n5~V%Xr-<+5TImWo*kV$QqfUkCbv4zfXTS)3`|7g8+U zAm~sY(&Wx<)uzdhlVY#??%nhPy<7Iw1*bGmfD{3`!{|h2%ao-)%O+h@?F;IV^?Nk_ z;>_2UPij#toAj~pVDLHRL61`HMk?{m<&&~Cc>(ChY*v?qMyCq3DxzsFnq1)7QNcrMe4j57g&2dbg+-F5D!)9kzaQmjvS< z2Q<|pV6Oz4Qqd=p%|2BT2c6$?;4BQ-Ikk4*#>wTfpj8?V7Y*}{yuMN4;-nT#Gk#RZX_V#(}r?e^OEwrv^#Xq$gY zV5fI#aC4}xNm(j+>YeScgMF?1PS7DJ0t16mBPo)YBKPCvy4Z``A?pSIa62*H^KYGA!&^zS&rMoaN}uwq3T+p>9~ZtfT$ksyH2fx=NoNbzybc6TKBuMt4Q?}o+74}U zxI>}aCAb8OOpkI&V59f@5M?C^;SVAoxF@jqcMZ`Y+1pCR9=1yOJ~HGYUrH+5%DnHm zZ#)X{48f7jLx%yz&FH!Q{x^x%P6`)qm&ef*8!ipHF^k0q-hj3x12{0r{#b*bJSSt? zSy6`?iq>+W5M>SN@M}>UEmE^5GUDdaSl0wpgaz~#b(iFnxXLXHXu~RkFt~qOa!lz2 zKv~Vav+!j#tivDu+fILMT{!S!l#R!`AREO5*_yQIbm^^GNl;6VWdup8^uVL^^|#)y z_;%Xc>4AtbdGihn zmkHD4ABnLtt!PU3^y<(i$@);;#@8EymXWPNYlIKJ-8vteO!sjX*n!1KVV4-!Vl^4w5<{igz}N?s@42hv)(2tyzbKjpAIlItIEZ z2wtlfY?s~hYM}Mt>1GHkXP?oI9YRbqG-{mKnqViL=Y$jLyPN(*tT)75yb~U;LN@mG zJ3LC}teo5%*%^T%cS%4ox`b?l;@LbVdz?h>hD@}*AR6+>xE%g?)rm-J7lgMd%Qb4cbb|Yt zOJiy4c*W1m7~=w2XT$Pk<8gxQgimV!sIVrl=Ix?lgH;>3CEDbd1MYh3FawiISBG}e zMuJ}?iM=E#ir5Ys%a}6;Hk4TLD$moQ_e0FLqL4SXagHT;xFBVyc{3lsY_c7?#tkXU zOA=X;d6{|h+{Qy>6^|2pJ1o|}G>Y9wkz^`nxqnY265@1#X3LY{{qC)hk=^aqsWFN| zWdT{V?wIl-6qeP@8Mm?*&fsup%I$xc6F4Sr`CjzRd1i35{Ag)5Dd*M&&f_i1ISVL# zM6pLGa)^q_a8344B^TyBkrk_s00Z%WG8eQJ?h%xb0*)qjQQ)V*5pWb%rjj|Pdt0^n zLHQ73Gbo{U(QV6=9zh*lN_wW(`L$|W6e+58!c@|r-0F3W?3-B@&@ZSKor`EuT#;2p z)G+7e>4B}<6t`S?B2&O1W0Pr2M_e<`$Q#Bs**s-G&|bLn_WTX*)|{cdU6O3DfEwtc zh+YVi*UZs9liAiGL-k2@t%>Qa*sEQHi~+UsT3VL_Dn417IwnOB7d_?_0uD_33sdFH z6Wk0=jJW4W#IsG0>-_CxH;-4NQx+zqmSQ1EbAXEJq^m^V!ITa?Jaldf%&6czs6 z1u_YPE-gTZ*)Ko{K5>6G5?*!0U_75~2Ou;S2IKR87ndm61kd()B|kJn=;y`1=qJtG z)+`<;0Y9|>+%<{?4d*uC#1S2(O?>(Ld=gEOgL7DwyBoP zI4?1XGVEZ^9ScO-gzD!Po+HoJ!Iy6h?0ePwN;fi##Hhgm#VWgi_y~4VlEqMt4>bvx zBW_h!MH~b2>iXdGP-l<@d?&47j+|mrLcgoh z){z2kR*1hrCvwcf3RO~UIR%V>F}sv>!^+R;0;y%{dx)fo}wC+mZkdGAqeVpCKn)NP?qkJ;~W8fsyl27Hk(N%7y%yhc6UsWJzlV2mbemb1*XLZ^aCD5Fz0T&o|Gkd*zXJt<75ghSL z1ev;;$@yZFj`F^swty5>pL>%M9A~_7yGJo|-7A632**xlWCd(R-RHZC8-mzqhbpMX z9!n#_X4o4zVQJ#x>$kRO&F0AvJ~fS0Od{tktiW-KZJ9< zh$FK$_LDZcjzeyf#@qIhcmMB}X+F=;G%xH5+01QU#bX}`woyZt#+?)k*{*CVrkH7C zQM~=SSDCC+RLq!yfsbkU7VHcviin+!bsq-$r24v7o-7xn5_QF1<#SCmQEP;2W|Ys> zt1*=d$*?1?W0e1L0_caeG>)Ep^4^LpYu3pZ%e(pO@k2xFP)s~$qBKOT5N2sQR8_ut zo_+FkX-2q-H{%jXWUy%tuu&PDCO-}2PIncl(>q~N%LvaRDXKQrh8gMf?!c7Es5jFj zZ&o0481xRbsqlcSw`0frH~|_)ugH#H#pxE?_V72~siDHmCMMHAYXd3bu|Ivn0wC2C zTS1ZiRLl;KiwsI^z~%|+E!*TBswB7L1nVU*h-;t?fx2DFeKV1~Y0w2pNKueT56vJq z=3voYI>-$Tx->)fGM)a6;Qp&7c)v{7E~!%GhuYncOO z7C<(*L*C_G=($1gRM?_Trk20g{f#r)LRqHtj}PQk?oD&M!)i&3_VU+r-umN5|NO`F z?_d1k<9D6akTA;cz)<63ee9I)gqPn}cUapOF5uy@4{W2rIU~G6qq`W~?{VonAH9WJ z>a@Yp4`5j$uX8R&;HM1o#z%0NS|sLJ<11eU!ctp_U1>k1L-#k?zeAm+ zfo7IZe4fZ)W6fj;Ye9w+cHIc74&eS7gRRm_=7ke7UNVleFa1y3*4JFjd!>R$|Mek> zJ*>zTF|OJ>+=cS8dVJXbPJZuITcaFz}c~NKcCGBxA`_s{If^L^e3m? zI(6a?k$9&wT0rTg$vQpq(P1nf_55=34QF> zEC7dfs~2vLy6CvkvL}xMNYBlPeFvk%8z&%5`1s7X*I9F+z8LoB@iy27KWe`CYB(_T z_!^twnicWUT@bUy#B}F)E-t1Sv==F(oA1#-Wg96`zxcyDuRN#AlqecH!p2&2Vr7)5Cr-bhmUuLRBxq*4`3ylvW z$2+DX@j~L1(kXaRuihnnA{%ruMQAX1lRD$vObnzv3*tayIQG+YNe%|#zZ2F+{9)Ea zN~h0~Zm)RH6A*JbEz$kLA?ytcNavj7hP|(+tkO&|W3TOvbH5|WJjPy$g_E2`v9NUM zshD%~ugLz2B+MmDEuAlTs>qO)$r{DGpxdA;Y`|rguuOJi&VWlUgFk;4OI-){ifwY} zn$zHk-;bq%;>!4M0f7BWE!&VM$Rz88*io4$OHm!uob%UR z4X;!qk<8cvV7N)L({>E^=$M}41d0jbAGP&bHzsk3)bcp5XhR-vBXf>v7qyeh`CI&V z2koHSN$k`M^OEP>nBA%Q1k~ZvBe!Yr^VN_Rb}-~Jf>MB zZ1-pu;rT@A8AW2~O5ss@g}8V~M4qpPXiCIH8BSxOT=I>$+_l1IEbkQI|M)GJl{)lDM2Z^2@ZxJh}MUh+UvCTR!a(x(Af%AZW?{Cc9t?Az_= zq&RIBtOnvXDZDL@`_)O!p)X-0PhqQ9yf_F>h}=ZNV-V};{>~DFT94X zO(P#WjMH&-LesCnO(+Yy^@sfFKbjh$DibB2LiTeVN<#wzs{v%k6DDwEx<5 zW^QUb({>QW9Rvjx)DV zKRr3i7bo9y-tYb1<#~E(z0-E(Ls`FrK5!#RP_5uWW3=;$N!bNiy)PbY1+1F0@!n*4 z#TjNFL{(%|yz!zFX7<90(h|@jiw!&}PvjowR`BmTm%4*wBJ7J+3HE@f9dLPb@@=sW zSUWlF41%?zj{iP(;3z4`eNAL?QBwJvn@Fr3Z<0Y3aZn&dOEKFhvXzR66-5TLdZo^| z9@Ou!!Z#+UlA{aKJ6$4~oEE5JsaNL$#bpX82lYCa(3sh?fwN%EG_eB+8WC$o9sfP- z08&F?}c?TXfsKOSYHW5e%ny zyOb!iMa7^)1qDx@-2mhLuc}INTE1_3I?_$bd|IaoFe)I0n-ii46 z?|v5Xm!JRsPrr*0(~JY5sNp=D)8Co@jkispS9OjBoJ#);c<18CfJ2FLC!^_*_R5c_w5k{; zTa+lj2YZnvtAb9eVuB9HcEXcwGNd%k7Oi(e(U1X$CIfLCv@G*7XBCEY$I1(B89x)98fBWh^+^?)Z4FCYpBy zT9!McT_I1uewQ-pwM2P5$oMAs+>oN7(mSn@HoFWs99H2!7OJ{o^MrhJz#+7Uen||0 z-G=@2+2>9LngF;au``eCWf$PFypmpq1PhCU2n2V9B=~4_j_sr{1pCr7*F&+7 zzC@HOET-F3=bS;?bryClBApOY2pQP#DnnxZ@%{ooT%&Cdk*9MxPN+N`#^!&sgUa9E zQ2l5q|EAr>1}!jZqUTfO-33x9daqL4j%CtXRfD9!Z;gB+0Wkc`Sq+QfAD^(Z>|hc2 zsb=p`SiB?|1q&eRm9KP8lqXTEzkWYF2e=lnth-vcD%4;m@O<@DJGF38>tASKLBaLX_~up8v`38`;{8xio#}r7I_J@ zp;SxHD-)+p+`94HFK_K+8-2{@9-pui>@3D|W$lOKu^4vTxn{v)V7p_3q}Mrm)_N#` zKxU05U~OLV}hPB zt@MJIii)uZr5YgTos}rRCe8%ewzG?w5hkUke0?66JWgMygdL%^a8;C5ZGq z{JN&vWtU5opq?`Uos+Cok^a$b$J*X24y>05vfFj9~u zj^%;x7X=Zs1fLuV1bpS6DC6B)C8z$MVxY{m-AszgZb{`?bd#tMZoq zi^-%!Z~D#;N$J#KBERg27q{Aj4EbSBNeFT~r1P1RqyJ7yeb>Z3@lc zp#~hxL(3VC`LYzKnnHoMWPYu*!vD0abOxSjV0KEz)N3#n_*=9uBV1tns#u2q{m3or zhL&d9?IzelA3ajTqpVR4rw%OO<%Ix_a_<8R+NCEZD`E8cd{~Qs&!3o_Bg+UF&Van~ z{n%fA$z*53i}`J&oSo}p$EBIq%q&n1#enSJ8H~9l`;{)L4)1`5f`b9wl45GVc;g#s z!JE0&QcN^QPc5EvWX^KOl~c8fPn0=4EX_=ooDyJbhfw6EVH z^=fQxDRtj08q1n(Edoa{FQZ=?187BWtu1mjVd{J3J1&t;cD#YkH$%xTih(w@9aO|W zZpinAw*v_`re`z>K5;x0Hqd-v%=G}V?>&Bu8?GFfsB8vfge%o*AP*dm1c8;m^?2sp z%C8vHZR{-1)OWJ`h6=6O?RsQEIt`P0D4z!F5N7!rBnA$5r1HU%8ccYf4FH9-9H@2K zquAsMUDAniBj06gU@WIrff3wZ=N(`rmi9Q`;cW`sCTOOS9B65c6!@VR_t)0FX0TiWr$Wr5lZ1&+r4C1ZFpGyLyA`Zv?^BVXtD4RW5H zOJm2y2pwjN)isK_LV+YtM4AfC$|_zN-7nKb&TEn*Sxgz{Ea&cw^-%HI8r;k;5!M0m zZHg#GR0DjwU13KYfo~TW$gmI)gk~lEGAx+U`fmmvqAcGok}t0fjuh1TwLk+(pGUkW zmh25Uq^Y7vFSpm_Q<6BN!|xMNX4*b|`}CVm;0t{x>pM%^AV64bU_sNMC)FKt+Y?9D zDnR~60%S25Npy*E!GD!iIV%F(JN`TPKQ@e~J;{zkITqTJOoUN#ROm|Z zL2@&+L+JcK6>++%1YDLou4~m00G`C`qE+x1Y!Pe@gy9D~ll1?d6Xs)+Yqn!|#e$3< z*2d+A-;kF;ZCt)z95kljnUhN3%5-gUZJN7`*D<#of^SG%*2}F7sZb{QSM#F;SH-tH zZjs%b93Uqvq>m_fFMcA<;^+glid6ChuJ!Sfogb4{S+%s!7xD=3%z*{n9yn%BNUgf` zNBUpKfA7=qI>24~yo0{uF;T3sG#FQ-9S;u^tIpp2q2nh*vj=RjWXBGUg(S$+PRMjR z32>C9(#_cEquHsd3I9k091ffg8g&67ydoQL_=ld`Nt)+&O-f((m=I#s9FOm&#;8|a zhvup6c5iJV^Ng{5EMh?At88D0{BrgyGlEW$Zs&}k@x0Yz`em)1=rIjGj)z$7nWyYJ ziiymqG#(s8&ykV>xZ;x*qVr zz}zI880$eap_qAY@OMKt4fvb3x&+?X(2V zgi+c7q%#wNyA(MqF+h96y;ZiBpb`;8D*_F5)=!m&mgGC-I{}$3Fu;gn7-MD`hPuMh zFT|RE$OT{-r42VZ)??wBNg3tfxoI8K33dj>J%m{>9)%e@-Vj?*m_aW!G2o_LQ=!!J zG!T&G{VxSctOL-PFfgu%0!}dB`sd;S>f~jq3h67e`yf5MksMrbUQ|feC=PihkRtIh z`jBTI?~%Njpk!hpy_-Hiy@OvQ-Xu!o7KxMHTAcChMzWjUw*Xm~&nlAL>ec$d;stuA z1hSV~Mq1rdLhznWUQWn^>Bs0c&&{4ivvuOg={8ustOL;~hUdA<O;pCcaiT`R^ zkq-T*?@!2?mnQgn%WVB>pqP4!)FH1!JEWn@y&IiS13(ii!bU?Z+Ao^9*)vIrYg&Wk zP)Kc{iO0Qdu72U(-~k6zCg}<}4f1ELG}d2kckTCYbjtJY_mB3#$00%@&EoXBpzd9Z z9Mg`cJu(+yA``!$XeJJvNw>{SmX!G=a7zN)Ir)Bt!YsjdD&R!6&SO-rBl`DMR&L?RJY~L9rFZjIN2}L#v&a2Skx{ZYQuQ z@21$u!mea78*PerZgTg&=Rb zYRhX;Al%tX*GLOPo`&fL99keBiFN@uRQ-lexr+>&E3`vrR2_4SRSEMwgLrs^w4Qj2r^+Ns!&!a;%HJXdIU$9RRYXY~Wb3Vf zu2%ee&!T$usqZ$^?a(US9;_3^a`5cfXL-0^c$jlz1;a2y+4PiUGj#Jy``%b&GCA{> ze$-6#cDz4tF!OKDQVf(;RbjIlmQF$B4jV*LM2Eq3L26YEicEk}@o1Ikf;9LC+82~i zss)C#2I#G?S0u`f=YZt{OQc%lNQMYy;U7YMIF^;8fvM*9K=DDcq)QG31fieYy0DEL zwqwHs`|?3>IZZKD6gf^s^pGQp8}8lnwlEMbj$x1j&#q#z}C5Bi%!ckr+3ges@3E{69yb}7!CR~I%x0? z%fK0C6?>)RmAiQ_x-3c}w^|uR?$g)Awdz$7*Ti+|dUXXU=i&K0`a@~9?G1!8$>216?N$aUvSWjD!OY+srxQ`(q)l?<`-8h)D+uV2>Sk!JyZ6hkyKe@^I@j%bNR8BJ zO;TOcTr|3%xX_p&G#1!ujFlJLm1`MDWpv3ZL3L-q0Sy|idw6jP(wmjLwv5G9HqP>x zU^K{ZjD9#aG_iuwl$=ct@xM30Xz|`kPjY)IS#8F>*h?{wDAJ9(p6TvS!lFH%`T&`9 zNg%HJG*V(F1~oeE_pX#&4;pZ&k)js;ldyO}s;J-Lim-t>P4_#r{ixp|McKwJh7=_Z zqe&oVIO`$!@^oCB%xs{Z1F0NvH%>Vk!okU$eOw(l7`2MNxew-|mM2zd)-#$9g{{G> z;ORW4Hq|}HvjLe4U`XU-JH0?rfhYVMx1p zl#fKKY2sF5@Qm;t9WN&+N&us@&HX>##j!!q+NXZ}-&Y>I=sms1#YUa=CPwtXx{R*o zmwR;bY82ZYKNc2=^?`W3UR^~SKUWJ|Wo1B6f_=R;iXF;oS-JOi$L+3hce^wn|3$ik}Up z-O34Ki4H4A9sh03iK;zAOVI5$O=_X2yiu6olg~Nf2=$^^|A(9E#=vUGKRu?;iL0UC zV7-%`SD?H>YtE3m1?clmj>-})mT7*Eht5@9&wKELt(V*W$r1m&t<=;R(<-b8y3~6 zV?$H9v0?Y1<+25yI^^2pIuX`mOqdzGVBqnn2w;B$E6h-Ts%!twP>Ex^wJH|mjM0o3 zlF}Ff?0~wP10Gj>YZcj(CeZcyBv}cH&mWkiBUYf8@?_zIYeJJFvS!=$Uy~Sij))z1 zI{}y8pfK_dib4MA5qgb&A2Gka}z{5&T&Dm@H5Z~xa(d;chYZdo{=M{ zR~t_x%CGw#1jXb@K?f^32opvo01U$n>A3M=+|W97yIoW)RF+@+(ls%1MI8w}&P}Jc zhz_eRNG54o&3{pAjDYzQlk&mM40LJiHDdqcrsg2j2|W(R3VF2EFQ|}Yk>;4 z(tjVfNuI+!1=WZxq#i44AzOPy*@|+$^h44wL!o&*Li+xCb*5+^{KDy?`|{YpGS$bz zD(G=<1yNre#NyYwPFT=$jKB1tAAe$gbarp9-6|kuFD22_O%bB8AbF|#Ifl+efs24E-7?M0Jc|Vaf-a5c4b~1xitrCEg@byZ>=OBjNcw*jwFV` z(z|-~16h~+>61Ajz1+yK;sr52%L5+z#k>~ByX#lOX_ajW-J?W3bbR8N#|7tdPL;=g zCXSa8RN>ztO}7ET^>`68$Se*%95;~IS;J%QukHVu2}M)7e)S2dWVbc8V?XAG8RqIJ z1}Y%VqwLxV?^b>(X$vlbj5boTX3!ceOTgOScB!Fn7c2DQczc6;hz57jdMCWOMhbc6(;)rp`8ZSvEu8DizN1 zGMOAz%$icNV=6gf_O9-w7$6MG#CJ6jxWP(*C>cszP{Yi~HVFI*Euas2cqaZGFFFJ- zWh`h#994B1OGvQvA73>nLc+W7Di+MN0lV%*I1n~j32gJLn~h?53Wh6qv|GRUo%|}_ zp>0fd8-20RvXl;W+US%NDz%CxaBNCB$o7pMZ?u1(iIJnSTN)uQC> z4R^h()|NezX7G6SVXXLlHoycOu);5O;qTNdLkpSML}ctZ$e(Iv9XC--97STNh+2-J ziE-9xH;@5^n##`56rU8IbV0gcBCOoVlWZl1Mjm7Wu6}Jx$cdq$eY-Ut7J~JsWmh=p z6Q-$d%`r5%VNE5d9%0`R@H&XDiE;0graBBYphs2X@BFZC8myf0!Jejd1CR4AR#s!m zhubc64CUdqTYOkx*6sTGHSynXKnt@zuo@cAGU!8`VrBo#$u|O4?ncHr!DD!w6(*>t zLn|_;m>kp=&E|Ly#q6d?Ix@WDZt1oxL50DLc!)D#r~}Vo z;vk9Si&90~+>UwFdgwh?I%^du4rf&2Gw!TVaxY8K?>{knDB~23VNIEo^@Y{z|H^Oo z$Hx!l@wPyK{j{GLJ8&}TJ%V1BdyZw$kps0ET?-!?G&@iNBwK)c;K_O=Sgmy3aLIq)32#B=EJPW{nmS#a(g@irQc-6r!W>H(oHv*rY*g z#y+y9T@HouL);&x>+GssnG>AWdx51X;Op{(L!_h(i0}{-uGMv|Prt z?l?bwc$sWrmmap`t?+)c6)BTq(kPNjMIZ}CiYQ&Y(<^U==3_~(99#CVT>9pG?5YO+ z?0K2qv7AC^n8&vG0S652qXU%ecORMn$D{;2+%IqG+<0t^r7>-BzWjf_|L{fMBG+;- zhPK_OUaeE4EzIH+0~60aqE-Dli;Df%KVSW?1L4alouWh7J+qv%0#B9wi}A$efEHPj zY6Z22`g6VdENPWg%|F9TjBhfY`8U}QTTJ(JV_05ehDkP3e<^9&@}|kIEI!utpXBs# z{;VCB2eg}6kINJT?3T4u1a5j-!9~HaiXK$TpYHFQU21?5LAeu7#qm&<W^by-jkX^=m0#gY}+U*QN?-LxA7 zgEs{hhTL%1!N?_n5IQi*gv7T=PBWY-C(6gh){{KPCvumW;hSiIY`v z!X66GPuoPCu?&S_Y|j({f5lIShWG6j9u@-j7(0O%9p8Tge;l%UPt1BV@_bo^h>?ew zkgMzv5moc^5B|A`3LZH3I`0R@z~!*t>2*#O>7C*h6hk)$YP^=w8IBiVB`Oh?2=gax z2V`CHXVCI7A-EZC(C){-{=v{vMmBq7JKh6XC~|z(IchAV!@4?z!v-I{(<)wxu)@DY zc!RdKLUp*Gl1G zW&8qn*zj;-MZ0+SLEt%_w^M@cMki$@!ZIjA+Yq(}nnUkQWKPLe2LU*9~m| zvs*vILT6fJK+oJ*{|od3S!-~;x^D3Xpk2=puXJt~>ws!5TZA07E9a%t_39JevAjfi z_snG+JcD(Npw?=jW4}PhhqjBW=t@qH+XVS2>;Uu3Pz*X4yA2rWNp^ra^^35)p^3ql zWT7cFvpx9~lS`33RK!7LC%wuwj)yA@))%Y@iuC}h<4=_FZVi&?*9w#xB+qH4OI_kX zmhrwbkRT&boa0Aa&EBx;#|))QcZ�oX~G19hcVQMX{# z^Z^5@`+)^#$XTx)N=#tKazG$py+M6^Qg{-{$ce@T}-hC$Nur_|-m&<%s}B?HyDQQB#v%M)>< z5WAUg%}JbQctRJHE$D#c|K8v@-cC+BG=pmP1SQSH`!ng~v(6Bh;&YBAfi=>dpw7H@ z{#~izRJP!rW3mhKlvXI$&aVMS1sF#Uc${`Cou4>uHO1O1owSKyH4P&*ehw4Uu-`!G zf3AL`LuT@jQ>Op!k0f!pK(8IA5ev;$u5^k4DcBUCP3GTs?g@$ZSUIoXp;~sIw{qTN z(#c!CsKamNJbhrjx`{N>_xTyzWt^6IW#Rn}`o;YYE8bi=FU2p8mo(wZuww=_OOUd? z-LgNl;LSg7AKJZPx6ZPKehi4$0~H3w?NbT#tM$4ZR%xOH%hVrwbxdLy#W*KcBY^D` zj`M!Y2pG<+=}#OzKFef1-h030DB1neq@Pci*_A^S1NrCzDx%!G&jXpd4tQu4AX$Wx z!M$N+Dx+jDCY!tDt#nt&R;cjrfTVP0{gmNuNi$kLHM~LRllt@y_uExJ(}D+nDgdYNp3_vohir zj$b#K7XRqkKOiUUICR%yW-KmJ%tsWdrXni+&oGy{aj)uUwDDI!^aW!zAVn2e!l@Ex z0NH50y2WtrK4150EwoZM&ux+4oWGU%l-op>Io=1g#5TTuMy6Al>OSyOSNb32l(_Wq zdO#&{Ij0BiZIE}*xgqR#*yoYx86Vp3aL1!V8XubC*E9DPjJ?bepUqIU@pVwEc^Llg zA(^u>ph*p%JI-riOB**Ps0^_m)_r6L8CGi+b>dK)-*-&FTK&hP56D?N2G$)jU^PSbgN=>Rd$WVaozz2Xk~^t96xUM^z|s?1G<>C*(`>-+p=fBK4PY5GH5^1qQ|FIX_9 z$!xj$h+?WKauzfegX(z8{V;WTUR5pal3!k!q|&OkD|Q0)`UWoWZ1C%(pMnZnCTMcR zgJ04Gt5iKxuUNq}3hQMM@r9>x4a0Uvt-?sjzDxXo*UU$HF}#{i8!4J%IpD2J1Q4wF;uzT7}^f z{sF?&;4$sy#B#LmQBuu-OMAfQr#eHmD$PEZdr*1v^sd{&WEUvlER`05UK*ZU!2{O{ z9`6Acd%)qkdkXkov7Bn@O&|+Jm}v?_!1I^v^y&%T4qZ`rYQO=#rt?sYuE7hPpYMT;LZ9yE6~MkB|Hg z%-57ax$)VVT_M4&ZaznYX`V!R z6Vx_?dU9~Nx8ZYWk3@r+9sF-O>Q_1g|J_A!V@&8aD1U~3l}=>0LT_|XN-Ng{JqJIB zC=zYQ8#-vK98~0!NHH5JvH?{dkCPU873Z1{$d|XsAxKm+Gf{pPOc7+y@HG2{YJ`_{ zv`cJ^%xGiT^19_UZn)sInmN|%Z^A`EM}8E^v}0UUn&ILg#Xv)DJ{6&zUG81we~#>O zIjuS;#)T|}{s`nf<3oGMT`6?Aa5g|UOf!G`OgwjxyUyjD3eRa38(p=tce&K7_xo1S zc-A;dtvY#e!Q!s){GWG)um0t_#dT`i0*eiDhF@czowX%!*l7z>{L=ron>C#W2xejBNz-=C4TbztGxh!xr z+E6yV9^ByWHrs@ZMNL`jN%l*FjFV=_I7~4IDN;m5Jn`z1myyI7k4R=n7ZoK040h$#tFOsBLaTu6?D~vrv$a0C!VK;$a)Oj_JAZd`a z$WM6dy)-GJE5fwk6dHuJM7Z$fJ8gGGUE_@bt&m1*qqCiID4QB3Tf)Y6pn2t$qouQc zI+Tx>P4v)?S0f9w!Wr}$zi5y9kRvnxZPa>g7Z-wt>plNZ#6XYb@+h>7#J~RJ3$^o4 zI3I)>5EcG)R(pSX-(>5QY;6eNQtcHDRwG?pz7qdYeK^c9ntQSCbT z7t+U$9?y>PvdzrQ#8Avyimaj{-dp-tq%`WJ6Bk_*pCS7vY|2NRCTs98;s6tL(Phdz z#gqvTeJ*|HNIyF~*zvNn-3$-06tj*ZtEq?+%2*MWd7;YQUbn&tFXyu`aiasm!h;ih zIyLJZhw)Tx*)bq2RBc5GP{E{wUV@Qf+}$+O)lhncd4_7})=z;r%O_!tVj%BrGf0YS z6{}wDlztMHEkdnN2-E&!gQ0r*4o`!mj7V$L%e??krm0%wz0OPOfnMz-x6~cJ!8WVS z=+Lq3uv)lc6HtA1DW~j3FP=a`7{q~AQ6^i#)JRW?56wuYi|I=ov>C_<@sN}V+vpl- zeCZ(&Yy~+Q=r0i6fay&EF_Lr@R!3lm_crKt>2*dnhr_Sq(Wbyet{%$yRtDw@F`|q5 zBiQkR?0bit>t#1*p|%Ltzr@TJtQ_n{%M$Dkg&0K4z2u2t-Q?&L|8eJ zZEs_Ti0?hB$Qc^lw-7{+4T~4)0}stOuZPv}A9!D>{DVeIid^252S_6#M8u)xQF zJsUe9e5rTBW)0w6SLtT#^wC5@VQixO5$Oi(Z6{XZ_i5>}=b$9fHq> zYUDjHw`s?|g$3QbD?&)psZvC>(hC0?sSc{#GU+GID0il*_gEp-^RfhmA%!xeZS993 z3q;cNPBqePt}TjBp$c-sKx3Sxu|wzCEKkURW$1V$hwhh}Y|80xe(NgPIF;m?y}COo zCWRu~sfY~tF`MVcK@|`ZjcWFj7$#kXEjxRHp2#s0nde&}uI3vXQZXThrP7*0*L|`H z{?%C%#XWmz-2#J#4suVF*79 z?m-3=jgc(t%)-44IP#wg*U2{m(Yk8AYmKx+n#Zfci~!z`@!3(W*x0SmvlnBG2ZNwL z%%SDCVM+Y--f1R)wg11yzmqj~?Ad@LGpL|8iDH12ViOfnH1iy{%nwz&jWk5)t7uBN z7?nS(&^a2hG^vqUb~qSj!Y1gHWgHA=74x@VyyT41=w**t0+CyKnO;vl>J;|K^bg+$7q3Y-p68O;pW} zVPPTJ(8aBzS4gotEk%U+KTQp%j?U*aFj0cb3n$gE;LKQ!i^&ouF~Mc~S7m#BOy(kI z;D-l@&W@Lo5;FskM=^UTl1)V(uAN`6 z#v0fTzhu%P>*Lj`cl@|Byyne1_3@zFP|;XMYSqjCz5K1xf6*_pb=AG;$F?2>mK_Sy zP&WOG)e1v-UjFmfhtfIQEz^_*g63609rPX7oAa^I6@w?G^GgD;ovTHz^;tC?t5{2U z6TS@_Jcn$cG5EQOK4t}tDc}6%+ut5f0ks|H{45kuH}kPtDN=AGv{s!7LLw(zpM-UI z>YeURUlUrVu2=Uuqqum9@S3=nTLo1~V2MIM0=noAu8C7bCBl03->$))W)-gts`TOT z^1>a;GD!W^sxL1rqit=5*nrHk1u|T*DefLkdBbIBq}D<}v{i;uQAt$Y&n|qaD?B}@ zithYM{@3=<`84tv9G%`dGnrp3YV}I+$%OdJk|j%`fX-lBkOp7rC$4*gF%EWEwKo`f zP$uXFAOEjx(;xTQV60wc`}Ox;dz7=~pR2B*nO^_|bPye(dn9kyN^i(Fd+Gx>lAMrt z66Gm0#zy*mYNg#FdR`l_P~DL0XKZ!5%&YJ}7Ck%hfj6(fr|yuRxwf7Ck!Ok(Hbx%GrdO}}Uzsw)vJkt`~r7<`dyVjT}xQw@^-H1my)%YK68uo$WBfKpn6NE+7TR!idK zzzm=6l1U>YCKlQ@sjwd`TT}t;tOE}C{|SH+E5tlA977Lm3Kqj)Tb_IrKD6Y(ZZ|;| z>W;3uKa$^dZ&T}dSKY5m@aUZKs{20ruKNT`H2YliwQ!%Xe4iu_?KoW z+GqAAq*9EQBHO44%mEs?bFd3JQH~A92|h-KL(DCL=qF~-k)2=Y zUW^@8V}$zHztGAjs~r5M4bAb`?c!r0YlB8ktJp*L&+Bz*4aVvS+~AY|MN=kQ+FL7s zwGnQ!cA`fZVnQylvT0Lx9$NR>i{6`KSo;RE5{!4b8sofJgNQK>J+8;tA!6wXk~!2+ zyd-;G3uZ+hsuSJwyd>@sHwIS9dPu$cbZBCLPMpYVA>;FV#xQNxfC&AfF$_JfM}kt8 zhkn)MW1RTql~2fVc0Pt3heDdoe2iL(siD9F2R`mc-WjSxK$Uk+^+dc5Xt(wVb~1Q2 zfs2HIhr*g(z3+^wGg+KZNgOYVDxu@ZC1^o?p|EoAdUXc< zkkqN$3?n@x{ZP@Aq^eirn0RNTX#{-Ucn^+&-}d=E9~i%cEiSEvl{GH+i1)Ar+08_O z)7vIk=^j;0C3o#Om%QGLG_s#!9#f=;ia@^OoS;JiYdKHE!0ZQNd4W&FSYZNk^j=Yd zwNP0$;GpwbEzmniWfn#oZEG9m63-sHt=Hb*uG>-Q6BuyFq_wJK zPQLuKs+s73`WgS*Y89GvmEBTDvRm#-fAT@lt0n;Y{$BqU**aX|r5zV>A273CyD0|P z;C5oH0BQUWIi{(yI8}5>V6XEs0n!V`@eX+Og0dsRY%sy85pviZ08s=fd;(v4>B#?M zYlF;9!vtC+^Dkog$ufMt9sS>BLpfLMb_->JMJ2%}hS>!2ep*H2FK+xQX`=F(qy6Ti zPyT2lOw4u5lRr$|D%&j}EI7XpockR%P3xC+$dS(PjE7F#%J1VufPRM@ zNj1OEx8I?SbDC}v^gApItd>Pd`W+t06Xy3jY@X2;&?8Cli{-?IeKO}HeRO)gI*W7u z%UA&uHSN48QqaoZx}b`Vn^gb+zB{*2g(311xA0g<{Q3ftBJ+&_2>2j}OY%G=r;KtGIymJ5Gj<5$usF zp_7;=;--Mj&ElqvP(}BA^_sX$vR`!3 zGuyR;ZWrsso1M=&U!Z$PrKB2sp^iC4a32=PtaiK}isS9@+)WpH6fd|cI;*$?mSvky zCw-B$iH<0vfr;`AsRS#A&z3FJzq(bD?x*vwhn#5ff~VuPsrLKgIB5%4&SU8(+X_UZ z8yI%jvJ9eOd?PB=H|q;S`RHCKTV#QCPREP&uXDTSxLRP8IRa(1LhnP<*Qzx=q<8_y z+8D0g3aa5?4z?hCC#})?BzfGK*9rooxb=?}fxHCDFeE`NNvo(9?&4kxz2IVa95-bd zs%ZaJK?yb!2e3Ju*rCSa-In3B_wL-E%rW^-Z|l|NBxfo)ZMMrjN-@P0IY31;$qm0D zzf;yp?}kOZncp}2g!lCs{l0y^7&F_hOqB2A>Nt^tYPd(MD5mor>(!mI0gx-I3eVu~ zgV*@x{2Ieidg%G^r5_m2b%$*EY6iCvB!@P8?ssii+-SIi<%?|TKi%dtVxwTNt_Kk#Y@cqXvqgv@j9G7UJLa*PisjY1-jSn5sIg)X+1-zUQg)W2-ya&N4|rgu z1>TBd;zgu~zXQHgChX>l7wlZn?~v!+?;q`PfxhcEZgklIMiv~SaiQpyUGF8v3?-Yg z+gOkV7AzyrVl}Ljw`K#~)oHhVj&}k|r|%GTdl>}J-nvhBNZ-<2<$NZ|wxzbi2*3Oo zrhA0ZYZ-n{EWTI@K>6=fFV>86_jyr&Xk>53KYqpvVH}&siYSMmM61uQ67cme6|y1@0A)@k3rH&Gpc_1@J?!64WaW z)7>GO26-$`7gV?461`t}C}7-HedHH&3`XFYjURdB@m*zS1nkyfUVM1v4q7dYiE7IZ4-mG02r&aIIIT ztD5;)f~2p3D($Q%&Rtw=Tj&kzlCKVW6dFCX_@|e@x#5eusjl$Y#jC&76<(_@5`G4b zPU(zoK6UDqQ_JWkQVm*dhi2GlBgmq|{~4Y~J{TXKpFPbA*wiWaKm3Yk!q2Rv@17(p z?HE64X874mF`Fn7M@69Dy4~@%=aL?0a9rtfkSo66)#{lM) zfR)7?VH{guVg-yTM|~E&|Drd%G)Q)Fix)Hs6MQa$ZdD2FiVzN{U^SU|FwUcusIc{< zk2)SZRA}{Q-~EHhTpaqDiP-{)=a!J`H+bJ;)kE4_bZ_rCHYnDnjf>Zu5MDK(e7kE?{a>o^&VsOMOl)!-| z>(SdBN(v_LwbT zJ18cZB3r14ikXc<1mHH<=>RP_mhvUaQKtm_`54B?Y7Nnva?UyB1toezw$FYJV<)V% zSmsX6Fyq+zlx2%`XGin(pPTUU^~aFK?zyDkVt#=(B<;#T?9 zNHgiAnQf{TPK7^+k_SGVlcd`0f{|z>Akez>x@fWm_Z-oXCHbS((K}&6AH>R`;IJ^H z*Cj6qw&H~$c|om`D1k1hOs1*y$AZ=@P70mni**=UMXeO*cHFRX1BXXL7a0J`JN!yc zrK&=i5me!?+3j+H*4%MjCAb4S0ds_p35w>xwArV@a1P+OhJ4cmxc9RT-y$dNcsaaj z2BCV2siVjRD#A$3S?=8+IgfO}_%V|{Paor=SX(}drDEpC!bTyeV$-OUG~j@pt(yWP=j{PW&+gIejd zO+Hu#L&Ea0WxF*@itd$nnz#ODD74tL^4PI0vjD_0zg!`w@i$+g*Xu z2lc$pk^*@$AKkJRSSTOUsa|O!gvgR62>n&KK*W$vkY&;_Zic8J+AE``SBld;#{q|4 zc`hulxxj$2jJg5VMT0rCdya|n`_82^c7V_qY#kA`P+skNk8c>_AQw-hfd_{>hX&{e zcuW_~>~)1btHy1xGUMhYO|kI0eR>>_v73xqS&x{bq^`$ zrE;NgnS*DGWIf_qwbmp0H7zNU^~-I|25NDB*~7`=P&U0Foxc3m0@Iq}{NQ9FIcUdg z3T&naIT%$GbDSdOR75p@^Ne14FTrk?X8t8khddtG_%?I9LOT3<=}o{}jWNAu{@F#y z+NXCqyXY#i_KE5|vbcJuKHm$WmxX7@aezq-6u%m;LiPHBvIW>8v6)*myEkAnw@!LN zQbxxHcK9_reMA>i4UCRyk)=9!bN2*~&t+kcjiH9e;^%Q;g_WtHs=rtHw-i1<4b?^ze!A%rocgWmaKZgwl~l>KFIoPp_l}U#6xBr!YL@7gWYf!1Ho_6s-WEt zJQciv0_X{%@tFsf)5iuaEQcP~?UvCZ4&HfHVnT-SzVHtuj$KN@jtkOr%`l^*m>m>J zrXt=-U!3^W6XDh0y&PWs{S99p_|C02Z-t-u^`$Rv|H_A7><&M)xaY@_Kgjq}%9qO) zefGuXZyfpJibc)ep9~bS0nf8VXI!wb?%^M_cLoggW9@dIW5F}sO-IgaQ6%{#`8A5m zIZbo3pl7q!S%bZrHJ~@C1&PU4x(s}bCeY+M4fbR*CY+6m%4AMB{JYrfQ(pP5ThVhx z6BcR{F3>t2P{t@D1G2?WYdJNE@+4(J07R{kHL=gNfA%GhG?l3Upc7`OLb=yFPtXkc z7GwXkX1$Y+lf%(L5*cIH@vq~ZYoW!um|iB(l=>G#Po&OiSrDp3B3pi+YZn(u z`Ed9|x$z-QOeoS#v>22!P+7FurG^Vd2f83!K(>LfTMrQV;g=mJSPz^+?{J4iH3+eB zQ)i$SDP|E3&**}hNan0W`4L5~@aQb#KWLBmh;oO|%Tb$PzoE~bdqC$1p&xvRs#f=Z(OL8roc^Eswf@0m4ZePG>^2Y0=5a@zHw zE;B2uXtyQR#DX)8teunf#?!x%4x(>f#v5^OWGJ)Vcffk5zx;`GdFX)*NvL1h>E?-xQ>K}Q{~WLVYJOdpqFr&xhs9KjBB z&}FUx`oIp(O}SBLK&$AK-r;3}pZ^57S5oLaPAfE+g1^)S#eG$DfnQ0e@wv=dt&%ui zzN~PXfxjKM0Hyx*Os;UiA&$97?mJ@{&3)%g`mi68K42jY)O;CxmEk*^>D?eSk~wRc zV42_yS?5xz+9$(5cK}T$mIbD(cDNS?mwV&5ZF6HecU+;qFyJOs^5B}w+F-NxipKQ7 z2)o2zeyzd~ zv4xG*3-vf}-0=3lguH1oPK%Fq{Uv`%jlC<`P9}sfeDCOZ0M~@y}Xmle9=! zLmv`fb+1+L{C@AZJHK1|)t!r1{ddJ%5B{ZIed}9|kUz)`zac*!{y_HD#xGtIe{e(o z!2=n}7Gk8}5(lLX4{}@i<(wxTd7wvF4JE>N0;>6~ewCau*-cmuvp6S6*W6y`9BwBN zvEiMyz&!YfZ1(JNPIdlN@sLx&e+L^M(rM!(m7I>b*w8fgaJM7G>{A$J7qRd(AU9sByxHy%Io?^30wX!S3 zur`+QmJ6^G154NU1QmuL8}xufwrFKg7aU>dy9Si`+`)#CUT0m924i$6Uw_dv?^Q#| zC>D_6&DX^k9g7m+N@sWs3s_H*ZP4-)17R>InI>ia`^l}@kpU}%H24IjWFSxm^c|T% za#sya+9iRX+rr6zCCuKB;XWAN8CI{vl=a_idh?ei7p_hUy9Yk!Aq96mbhf}}>%i{D9q+4l`BNsws7$WR~^FfuS z3<6WQ3(lNXYrw!^Kk(Kt$WW_@`mdu+6p%!=-*_ONVkOuyM2OcN0wt%)ppNF`)UT1?KR%Xq7&|zQhz! zsSuMoB_JzS9b}kpY@EmIv0>FN0}gmp=a?#LraPr6(1bDJuGiQBGyL)R)$AKAFaLytUBKA;5CGN8mj+HPfX89S%pG4JCioaq!m-e*3M^C*OB(MqTl}3ZT(BAz7M`D|(?50Gc8F^-EDnAS zQ6&1MF)X{x3`-)#Y^2BrD#FOPVB`Tu)|O0qr+ODx6DcTtwVb09y+66guo@TEG}YM< zBp0U@{l&?Ii}y4a-zBT{fzO+Y~K>$$xyt~%RL$T^hzghHRtJZI=$KR>2Y_+ zC*mTp@l%RluX71)JU-&MBUG!p@ZBumOQhBBL~uRSSI3Ih$U7nKgTc&ukof3#Kti&3 z-V-R&-xrqTc+b7r1tLVMbDqfUk~OozcNpWv4s^pf9yU4kJObH)Zf5M-8$+w-*|;xu zyp6Wd_GOf$h~*)5#%_A+qDVn^NH^{Up%0T98b!Yxfx@4#%96rUN;$>;#q%LMfTXS!8v7SaE?&S zA&Ts$B6hjN@@}}d$%+?%q?UVC_#=`o*yYkLjTLo*JXVU&E|-(CD*Eo?r5`*3QR6*= zE2L@eVZS5cABk$ES)6nTp%jbmx!0>#K)tx-=OY&_jieqC8$oMK=o!(<{4bz~`j0E# z@BQ9{p2d4BJ;`l5M$c+9AG()f9#N#5is%hX;G%9U7NHsyBaGzz8e|A;kd)6fFc>1C zK@ESW59BR(gl0hc5i?ZfoVK|=l4h60fIRO3hke{eA!L29v%okS-WbP=e(lmiBMlw0 z2tE}CsOR42^+(IyWcgX)i(eKKq&eX6 za+JhP`RbJYe;F!pVztJvHvE+g5IQJy2=Oz0c z<$nSyd@1yDsEn+kk8_t#*Q<{5D`8tz<(?C=OO-Icntxt&ihq9GAzzCIWW2V@;=tp) zz-pzVqJNe6>lZx~bOn5z;xA~9LdQjeWc7kRFfSc65cQ+ZRu{KbR^*Lg;nkBzwZaw} z$}c}_W#2N|a#S_ZV8gr7S;FQ?jW(A7b0};Dbh18{GzhC5 zJLp`u&Y8di-XcFKuH=~)N!jpSx5CVfxVM_9B!)_-e&+`!to++q>MP_UJKjHm{?VYN z)lVW8GiV+j4>39SNeXogOr zNKbTCfRc1cGylT3(xGw&yGL<*^eNHvG$uxQt>VJBio7w%hjC;yJzB+wie$I;c~4w{ z(_3SpwZ-gsnSa~saXciq#i|M1d0pY?7cNP z^j688xrM=PG~NIw&A0dC#?8p6Vsq8(ZD8k}6dYLe{)^74fQ6w2?yWS66*qY|dAIWW z-BvrnLe=kvl~iMzz!g54@4jP=gLa03R>@^RYeMNLVCGb?yvwEJf0Rg3IybHAU%QAHW=c` zoCWF=!HLt%?ZO02+gQxa1pQOYhINzfl?I!ZDWPNdE|O=*%hWluWvYx~^b|P^;;^dC z+*DC9l@`3(^~PLKDd+UMoK|U74>@%Ua)KJ<1)(_}G0aT}p`8ac^q!Dt-hLH?lf^mo zTX}*!LAgQaMC(&6L8a+ult@v`bs$ z*F|j$bLJF=B+4I5-$}~7vwUx=u~j;X%JM}H^iDd7*-82x{+{KFH}pIFEz5VEd?z_U z-plgEfG#q%H1i*X^f;$@olx2UAJ0vmO#t)UsN+Au4loy#)BZG`LQgy1fm9FC~=2A@+sfu<;=$(X#lh-b5GEr^~yFi8g&zcufdLvz|}3TW)Nosd|_fG_8+ z=T!@fscW+@DF+|ds@v3crf$<#nLaR!+w*GKtNIzKs^idRJ`sjx=oDLv5wq23=pYmD zxjbx1-%vgsG2#90=f8;f-OqmW*1KYwaUhf(8wd-mIs28DXK6u<=(gtpCXH0NUxH0f zwQQ4LIo$(Ez#4v`_hiia=M%}bg z{Ga=oj7Zhujm0F*j<+%uW(K2(V)7}HOGR9rU8h{nOYo@?)+;UwQ(Zdg^&WaBTwFG~ z)T?*XHNsEic|mJ{>8Y82oO_Q>H=H@@TOF8A?}LtLB%;iq>(xoBYUx`adtRW+74_;Q#M8rE7(wf{P2~Y*G#Z+5UCE5bar_mJ6@HJnSpFS#Xw8- zK2QW&^p_&&ThY%r3d@Y1xuk57C2?B466N|CdqTB}bK(nu`Wcx{mHz3fG4)gN~&1Hd7xrp~7Y;vvIO0Bk7arLkW)T zb{VojSrkQjxu~QI<#Fq|xx&(ET6a+SO_U?e$0!EeI%6aP{54ET3)m_}ZzSsDz{7gDucAh*5 z>+rlar%2YvOLori$y3&=G5;4mHPyLGe%@o{)MUS7;p?Hu4(~gtye>TNaWXVkbdj`) z`gont!JwVwZAE69Dj^~WD6O-s_Oln>%bh8=q{AObUJ>vMo& z3MrCDMPOeMW{p?Pdl0NC0rk07FH9BJNDF4=g0mY1I?xX}kWH?p4@2(QB;j#6UO>>4&_&`} zASta9b<+FAMYE09`gpbK!XMOsw=4Y5ng7wWxZuq?_2(DM$a5EM0U}$Q5`%HEfjwb` zA8NyIfA>FP6MkeLe6f~nVCM_laqMod8QM}PW;;cas0gGkN3ssgDPyk^q?$paSMSsm zR;*YVs8v*kY!jgSbKgm0q;$*{*a9x2&g@w5H0q!>xZ%j$bAK61BxJWDbs+^BU@0#>5)uYiiiD#+h8`W7=o1ndFB4E zUim8TsppdP-IHXc9WOv>W+q`X#cZNT92KE;Svl_tXUl6BJqzWUwbN4Q0j1t6RZ?>SQKUkuv>veZywXgUbBy=K@uCFzp<7-%$X)F?(XXpD zKk)N=ML1*jbJ+d^s~6OMqnHtYpqqxa6S4`V*sAX*12eq zxxSfhkU&!h?uL+Oo@z2H>ejb4WSJd%ASq_{E}mjyDYA}=FtNxZMKZ=rPLXcs0_6>Qq7B(F z^I_dA5A)yzpJHX(rhMcnJwCM1o{b5yU*kwDh#i(mT7lSp!}LDTxkEC0D1rn@8Lgs2 z*aCrB)I7ZHdCzGrc`NRX6T)=TZi|9gE2i8!IDKv*CVr7L{-dQw0zU~g8I3KOG95X< zF79E+0I4-IF()Vnl7eMaM3VAsK+4Si*;OL^GaKZD&IUXPNC$z$ep#2S4fr@NFWjyy z2wxu15^farISZ*jl(^XD^Y`Tek<@4M6w%p$LS^;;XYXC$n#!*IanIlh$;OZuft)jd zAQ1+{5y4PVCpfQm+S}Ki+k4yF+uQbbr1y5(-)*P&pHACpZxKaBk%xi`XaMCQ4?)Bi zDhP=0QGr2?;u90mac~e28C3YMMWT{OGzSuX(f&H0%6aV_IQyHm*IxU(zDurCuL`U5 z&tM-&OO*G5tKGmWBS%RAba5hdJH3&rd;jWNQ1v^sj!mW#zFrAsF|e}G*o@gIfMkZM zXPnpoW9yGZOFwvun_`6maws@SvOi?I{|*8*Y4TK6Q9xpbDr?!2HNG2QZXetGvHlxc zxnuhawbFruE^{;VPFryA?yS)FEV(IMW*{eaVy!S)*0KdK%hb@tq8>-y6ySlJyQr{~9)&oMH8N)*04` z0}g+>IzwTZPs+thabk%KE7_-pp**fs(jm@qO@-1An6s{9HO4aYs63~^d+4P=RYsVo zmd{=#&XH)A!aAvw6b0^|k`+)08;F^*<0K)_R4!|$&!xb1!Ck?{lA$U`xWhbDERrNa zp!%+~9bBT0F~@R75243c8u?(1u=-ZRK=oTDTNA(f`#&W6CxRHo@TmSNiv5%#N2wTt zYz?H%R4@(v+IeOClfliw_aXHmUX&v;YOEb64T5E^Jy1{fV92p)hx{A-x)WeNKzQ; zR-E(BhRU_Z`?2te@zO<;N9SueY$%=#(5ME5+a%$%OQ+Yr-``-V5BHK7B`bvmYGqAw zTpeMlThxh9aorcP0i=7b^A`nyZpKZ(-z`mu#WB{M%C(Q@%v4dTZmks z;lyD(D?6bNLkgs4NXi?JezvG<#P`T`Ajdr^8*Y5!2<(aClf!Fc80WLC|TeB#39R;HRRcML=+ZCsLk^>fl?#1EAW2%!F59X=TRZbgZv7%9cGeRLamKv1J zDx_<|+JSe1-f4-bV9dTI*Y|JpKy&SQjL&m&h$iGtNXY)S$!Of1Rre)1I#N!u6EAt% z%}m2NiakS-�KcUSleU>eF?|`j+gLR!`~>UkF+I+HtaeDrjW~ z9etw`9yq;l0c{j)ipYSev1N6>s<{|93p=<Sg;Y2(N#T?kj_N43h9Y~Zn6nU-D~D3+^+Pe1yYjqw zdPP!LpZi|d6v57@TuB9!F!w8YEyV3^g&yKpPpRWAS7-PDSD)@+7KI*z)_=an{2Ajj zZ9**B4u^mJ!=GbCYo@ix46@_Avw{nt7;Ac!b914uu}xYI#EY%a z81Xslco`wxneuDN{_sUlgMUP2!rcM<;zPl6*d8txr85=Y+s9co!#1YmkmVS0^c;7Q z6E0p+sMCIAXyXDB?|hXyyBjyHNJiB+JG-+o)Q(h)$#iINrE>#U+wc-JlzW?xKIn;AN=e^hf5#1 z%rv@cRefIisRPU@@qmkgY*cAsi(;*0F`XxUbetV@KBOK5Sr&yZ6n!Dhg+0X<^_4d- zgMWKkT`bwgt0%W+j?G_vdNdwic{$+nohzL`c+sUaa5;p_m{YJE%9L#a8?%F)6K{p9 zxW?|5u9G#&kE!&t90rBQn0S1^4IGb~IsAZ|E!my6AUMF}iyLF z+kSF3xWE-$+Kj2~s)R{>y9_-u7+S2f9j7zyb;nihHY%ve!A6z(V% z-5-}g@2C(m!id`Y5;uhS|K_U-)?`%N@AsZ3Yex$1I`RJC4l_ejNU@O2oC|`J-@5`c z&L7}^ZNj)o8LtkwlrLEJjheU8sjq%s@@D0~#DBBet?*~Nzok`TT^Ce}+avCg=@rM` zSnt&h`d#M*>%$!xco`LXo|qYXgXVF6*3lDKuz9oEgcL*g=c+uorQGq=x~Vdyx; z9-+uVDyCFc=)Z9a7QnpXjk-LQk&m(o2VD-i8X$XPNAN|rt!{%Z#&VjP8XB0HP+e1l zC1=u^QkhnHfZyW>iO)d!PV^c8-MH=kNj#8i;qQ`_hIa{e&e6dxf=<~9^IT(CDz7nY z7qbldLlcLeP{YroFS-?sq3-w?arOjp){Y({huc9%qu&QS6PkW|wD$;E9%}Z^v8e0OjAeYnp%IyHjlIQE*x^lx;+^ols_T* z%L5--qT(e9*;>Jno>w0dCt4QLHVJ}r82P^_-V?Q0-b9Z-gluLvIRq%r==9j!KYh=V zw(YcDtQ8_Qa%L|~tp-1K?E>&&+l1PHZPW5;tPX#H=$SpyAGiJzM0h$~+v&s8*H68u z$dOdhxct#UlWxV;unyY3nLg(4+h+4I;xu61mp#w8nUCLz)w{=HKAhOiwPHT5fs+e7 zhF#OQ@oMQ}{yLxPuqNdx(Mjp4@lod*KGwEkXn33PIR5#Y|N47NAp|RW1J~J$(`te< z*}`e9OpX63VI8lPITK;;*n%y8v`sJ5mX7TF45u%o^Sb=g_GJ3_K3dGon|YaJy)=uN zDl=!bgknMRIG>8C4{nm@ixP*j3^f>Y!}L?Aq>D?MViCx^fo6~ZRMdwa(gv&y$Nl3& z{z-y<_dyqJ01#H7BD1j!9S24dH}u&A`VnWKM;Ez$o~=)}E^P3&q|H8Euf>UH4=d!@ z%kwY5{Gn49xL$X^BJ8Jgc&U*Eu28c@TuE0*KKgx%AVak?s`NFz;*?~(j38$W*gA+~ ze9pSzt7q;cD^0dUef($pNbXBxOZJ=Dk_w6~qeuyqDx^PEVkKLx63LIZe2$aCIhv!P zCqt1_tyw?05&~$bTf9tkoFHBDvUJep0x6`m5U=f$?+L0@Y>*v_NDjzz2LrT&)CxCy z+>mc^%b03+ka0%SBST{GbBy45syX@f+iwRx3V)P0B$G`}9R8>?LqR3QZl%Z;DyHX~ zEko8LhlgRU4jQv`N+T=0H_C2K*Yn~h4t}>0?w^)_N!P;SXN~V^`I3n`wJ}uIM^>_p z^22mSNTYC#Zv($Zd?e~CS%K^3@Wrf7VRsnVytK6#cFWVJ#&N6{tXq{4!Xs4lN9360 zLip)<$ccGtRu;#ZL3bcGv<+q;Bty5;m2`F3km4K0QDN6ltIAg)JI&~VEOx&s)EwaF z@{ze#NT&!c|3l(@`$Oi*(2mK7Jn6>nN7lhJ|Ig(QKQ@_{^FN56MS7grysUa<_*U`a zSJ;KGkO3;DMSX`ptzQ0W>1&;Ioqsp~j(b)>7mfKHrLPsrPWtBZQv{8Qet)3Kin}9k z&Pa{YsgKj`{A~I@|0+rJX{MX#0hg3uV?VL!)na+ zAv*OJ!XjRspa^(VTK4AjI+;$bXTFpggm^96Eq6*~@J=Z8pz}#!Zphn&#@{@4Zr;=u zc_HrxWa-ulw|QfqdS<8Mmt%sQrPZ?Ixpl~mWUqGT2fgWMf}v_qw-buQ;FdHaXtF)co`DQHh3q6B~I$*uVn3qgJ%rfXdv>8?(BAJo?vUSCD(<^ z?xGVz#0m>0?H?{n4H1}Fwu&r+9fIYg)4Nd^E5ha0A<`)9^xm%Mlh!KYrUANr>_M9>83JYs6Tw2-w~u$ z?hUTs>v_!}#Wm=H3*KwqhKM3w4w&`S$Ql|0EIm_U-xzt1k4{R0OJy}-CzMH&J+Bvw zFidY}>psTOal|Y?G}4?#$9eal4jf!{aDmBViQYdqoLqI{)zP4t4|0cMZ&TzJ6|*g< z&>w5rA*M0@*Uj{~Hwv>QIaoXsL%_*y52shVEfs751xwU0)l`y7SP0>hLv^IGC7Kl1 zF1pF90b<1YDx=)#Z8_$_YL(ml;i-~f@OGs-{-8^fyp=(+Z#xYiqA<{RrKyocfp_Q* z_?&xiKJ@KKk~GOzvN_>#A_LIJHQT)V+$$r~VWp-6ZZL+S(;1C1&II|`P)neYZctQv zoDJ3(AV`@eDN^dEVzxi7?99GU_WLpn`(BLt@QKCa_GMW1PsQPuJ6SJ@3uI-FYr&S; z+d|i}P3(f=Z(NpU1#FS_$O`y3NvhzK3UlzQqCfNAK-PxpXYC|*wx3T1$>NAJmMq#k z>Xh|kdM(%APV4qqSy8tM^HraL>ysLVimAKZw}Mb!oUhKuz}G1ba zKE0cN>(xH@0he2^Uf?y#m$+j+R;}_OQ|Eui_gGXB?-1!Cw_eq%_6F9{SX?K0&Q?(E z=y5;G%MP+9QaGdGSysc2(BNcQs6lPcnipN68B5>o^yzm$1#v#4KeWo4{a_9`7*a}O znmb-N9HO79>77wEr!o>Jij`i-(KG&MPfUz3jH8Fi2%XvS5+|5U`1gpPxcG}7) z>%UIUa|_cr@#a>anTzx}#e(jTo{DLbW0qTs621CjNr`lCFzR^&teWHqD$Hnun$Cj5 zG_rr|fJSagCWs9!mL@O>fH`A|6DpZu?9T``W1mBNV_3xz>#G&<`bbXrzK|v1t+IMQ z2o82DbRm$l0H57F)CyOAT9A}6esI4BHqz`1K|RITxwyB6xeNHQh4f+9`=pCT_v&`g z60nNs(G9x9P8u2Gqlf0ROc|$pBLaG`CvNNSOki7^U$Tf3hpfuyPD+PAbM9n z^_)0gg*Va}+~#T%-gN^d%~`dyrdD{EZWAVv8xd6!gI|%V1TJfph%XeVD)UqqbW z_J<}5a{kAa_esl3Gp%%)*&98@UZuz-%*4Gj=?++z9g0p`Qy+{ev_*ldf@O7=JLY&EP0t#>q8)?!&{eHSBmfFI)1T zCYT-i&g}rA=LRz;UaG~J`Pw}cdxs*oshC^e+#P-Eo4M}YOg=;a(wQDVy>hui&(k0u z41C?NJK&=h(r0|PzN#sdl=J&Y)g08K&!@4?1z3Hdhn>i;buZ-QO)aDk@QtfF3_G_8 zuQGikn_d)D1~G%&H@^tp8#okMXp-YT9D37>f*{oo=pYC9TIE%yNQ&`l%pBk5jfJE3 z&BE(fU~vWtel0w_s8ucruMfc$AEss;Q{@L>W!!>@qKIw4|JdxIRbnI@Jaz1~kS;OH z3H!eeyWzMT2D=j<++oH->_5 zL8WnMuLZ!;%{bnz8YM)A4W?rb1Alk2e(Ixj)lcM0$VP6K$%)rhwPq9GHi|8$ptNmF z?~I|f6lx;f@;uMiBuNZ1Lj>*ySMXX{NF2rm{_3zQC3@vX1%iIDF4B0E4H^X%yydg; z=az_NCjg!G*QGopCfu&U??2OtX+ZA<<2b^fyyqDAJQ9#IkFc|?kk1uhD5U0f&wc^7LosGH0 z`HgN$3iHtOMYzX*vnsW{{wB5cOFg0V{F6e)jO{Zf4{JIhGm^5r`<)hQh6zT z209~;c}m7*&9*rObf0V`Yt%g|2p`Xw9yfC_Mj&!%&JDBL{|xuFjQv>&u~qPK-{rph zW{)~v`CORewcaZ{4u&-QH2^Q=^qjJYGrpCKrq8QV@z5tf!0xlSl~=%R7>3NSm7eT- zncFZpt%P^*cW3^Cz~rM$%lOA5WXVh8*n%wIFcSM3id{{S6x`k#TJ_+*m2s_%l@Lz8 z4mlBl%P9PB6t2Y>pv3{-WB3&AP5lDHBL2R1k}AAGuO8{LL;rk9G}pw3A(}vT*sT+1a~#>t>v$qkWn@5k4s636C>lO88Y@#>~4zez?nhM zoMZcBEuN_o%(2z13Ew6;7_u#>-A8kKra{&U`kRYGTV)AMmQr)gd(D(HVuK`3l))bN z+#B4=>$4ykg0h;uXN7$%1Gq(u()c$HKP-^)UxS;$O=MK~~O#X9VZDeL*>) z=kX%-nFri{0V_{)LyxvJF~Cx0z-bpSRwM=DCzb`T4<1l#@LVzx6WR7D<@_9g4Bj_Dv~;bcSfW>^jmL^>v>>q)7OZx|2_zF{>aNe3VFJlV)#^G@%56>K z#95ce%pkRoVj&f8Cy=0^zRusSF@MF9e!1j<6mu*a`RBx+LV2yUfL(L@nM;8wo^}3* zAT#rSx3G=kKdhdgs~Viu&twI_xn6oDeEIy-^0SJ5X1(Wz&;q(Zb#n$b0Z9o(cVrPi zgEa6&StuXSBRWQbqWbnWflG zv5>I64O95f!G2B+jfrlUwtqf!`}9jplB6+Y$*kL6pTT~Am88@AtgKvtjJhsJN?GQ4 z64K{-BC7*+>eR@Eq9ULwJZiV_)y6PAQw9B9_b(H#WWOXUy>jf!m9iZ=&$J6eH|>cb zZqV7e^Xp%}VS>(2ZZ>C;Jug_%3!NF9j!`TmLp4w_t6dxT$3QuxkiMo!4qNC6>?U`e zI!}e?5KHD?1Bz1ve~aKM_%nBB)CX?`L82?tQhEnTJ@81 z00~)4O`hAQu7@^9=xV_QvIWXzRL$2^%)jnl2DRbz(QtT=T#9Po@0^3PRuZq!ZPf0w zJpncn1V;-x>j3-V`=|MqF?}w4O(*87SlOCrp3^8y;vpMpNVtaHr|Kd|d#i;if?DPI zkal5n*kS?Xrl6MnxP@=HBWJW&;MOkn$FKkKMPDSPNvh_gGY{k#lLHlkeUO>0RhBC> zm}-J)Ayprn`Bn21P1y{OEHZc9U7)Hj12|wr(&+P)eKd`_7;VUpLK2 z|9v|1eR9Bwof+Wd4qHioMzM_)IYGsw&qpoc?#NHPo`1I!a=Wwojuyo~`^+kELnyub15Ro`z>pH?perBKXgTglc6 zcTG3IHo(J~2SacjRHW*HM5VvBvIiz_48SC)_=!s<;v>2IT1e==BJBbf8$T9*oxfaT zZjguk3?t#jvdbvWYMO7i&#} zCIpJ{UGp+TtYkY~D}Bnu$d2SX$QdEl2VEcoB-|(; zgE_L8x@!E456_W4_e@>^okAdDtVj_UXxzzckIPr%9^~{x`s-fRmIxen%*Vl%~jSB2{(4~ftpQFKEF_-_GI4d0c)hZLESt8BK`P!ip5Sls=6YO`d zq4D#w!dJ5R6Mu2|+%|d>@X(z$m;W0RshY8Hs+<^nR(-TVZ3)BZ1jL&=5nwBda8S2pYKV_-AbojGFjPwghC4MUS$etoN16(wZXeGqRt=k zTInTWd32d#bx>s_aPQ2ND(9GT5MX|uVslW-XhQ_q)obVLE=e~=eTmsg#ECL$MCVG+R zDk+?{nMMXhKHSkWJ6N3>xfMCQxS)lRo5E4t6z}NNIo^#^;zV(faD)y#woeYYtH=t4 z31oBwXQh=Lac1QcMtAvMs_wsD^a!9yxyqP8jryPYs{4>hwvt78(5pE)g)!AIF2veIieR`eKB5kYzbCMph9!WLhH)tvEqlic_Ts8kduS`})L z02@_k&Y|lT|;HDMYK#CcF^R?~xkW9obH6;zaeJ^oX}Kb@P&V=~jg9 zYm(BTvsMbo)IXx2Scn_XNUt zBzAIkz$OpltpS%C(mt=U4DT<2q?_o4S4K_4lW zob*K*@%!L>phiQS=={7vm+Sl*`WnP~+Jv}am8C2R&LxeoNHfp|P?5pzw901pT`JsI z$JN+DU%Sm#qr=p*d>8vb%Q~j=?(FV+(es;f!V{PdxLsqj92pFUjK_Ff9fTn_F7?E z;asT=$N;mlGAe8C(5AqZK!abg1PJ6%96{3|?`Mi7xR@S4=^`#z8Zs#30VMd0#6N8= zf6=#!>Ze|W5b6HeSe&WTwT9R4j=H9(p_&>AVuLL_e4vfSvP^qGo$}Gm6wh3KwI7HR z_Q~pam!kBr$?j{E1?hk82ijWoB{E8nz4Fd z7szlkEWtn zfyiYDZA1{4F75Jcg!U&Me_xaY+yjaeSHG(1Q8XzJ z_%x6vWuMHR;Ia)hRwo|w8+O5qv6EDVe2cdUH7)CryVSjWLgq1FpqyFZy? z!p7o{d=p7Fx7o*uw|A<|kW@~wr4%W`qkhkl2=R^cWOR9hi4!$FsijZ_Ko;8p;JLPRy6w zW@cUTC>HWWv{1?hQm6+6N1|GkD_+|!IU6wGk}z%Sv|^C+$Wx^S9P%Gv%DqZJu05Su z&n}yb%*nF>%}~i>fZ6W&U8~WW6gwhf#IBfc&+V9QnuO+GElMQCPVAr@F&p?j6bn{k zJJtd1C7s}hV8PE~NqzA0+2EBHF#Elmz$eXDHPVNtBM(56r>Y4%_hvF5=9QW-oIQ*M zKfAp5`Qeiq+#>=df8cnx(>OIj@rO$)Ok^tx?1Ykzz`=}6n{OXWGFmu1Wj~%=7;$*H zUgYM%~DRvb_lBpPss%_$TKvJ|J0xQTsaiD(p z%TP}UX(r?91LMfrjv}r&7~AuuRqy<_r7(vTkwB!C#`|1?YjppVi(ya=^;kq+zykYx5ufpg+WvN*d?#9|G3`86u_$K}Sn-`cD_Mi=5R_(iI0l{?FdISfR69ps-~bjRkUU&Udaqcd6(Wq~w@n^brMt>Jtbc6nIu`|Sg3s;pyXgdH7<{PV?F zUlWcD^JZQqS={0wP8^>H{qEmCF5`kBheOjxdW(95WN zRU4!c=Y=7sahcnP?eq;Ek|4KDD}o&!oQN^c7_*F5hdHnu)mX7LMhIDLUia<3LJt#K zesPg3AgL1x)TABunsX@@^wzSdm_$KZKt6~??4EK*Xb`5&u69Fp>JDWN_>%*Q)F_QH zZ9ALZ;#Mpv2yYg5&>t3yT9l3_fQ|O$T74^ay20s->%235V(cY<%Sf8j=7L%Y@j-zg zpe1I6Gzx2>$mQLcnz$f?U!v!7Qlm_mo9fjj#JIa=58V^FS(P$PPcDk_=Sk`QDIYo- zzoSL3ql0{$&^w{|-baIzO@`zgW>fj{jFhY@?@gc0iI z=B2;?`_kar^ZNB^yC6y!8*1L|cjl2Ukr^WL)dj+<#3#VauU4!r_Nnzq9Edz*kCjuU_*uMAV3TnWCw!fvpS*l&+psL6`gK?W2e?D&sRECZ1dzEf9vc<7Ry7CW;gN zO%Qs>9#0@zZow7j4NH@KW`HZBSSSy(nTjze)<%^E0@Hr_7SAf`Cfg&^D~!|G&Zy!* zy`su}L851e{1op#B==MRap_ag@yb%Ri!TBXGC82dvnWEVye#dUl`kq2@0*oRJBBl1 z4|olSg}siMz7N*C=-KC)K~3{x=cWi^;onVANeXpsKy^5sS>w?NDqNj(`XooMksjF^Jx{|D`v_$B z#JMi?N{6y|%SjTi)4Lc-jWr4zg&pF>zIDNyJ$6W{-LT*kCKp#I_RTpZDs-#koqBhu z^d|02YnAa6b0kL;o!*^_tZ;+y9!d5wc3&yJ_65$eU5#LY+IMIG{$n+I3loUSWp;wc*u|V$5ymb?lFDvNF`RKm2dGC3A%f z1K)Y0(!`4DUoS5$z{qFO4*AdfCl;Mi{? z=QsSsvvcToklS4Ev@H9r(UG+x(+u-b<)lB7L??Dri_PYY^%R?dL7SLfQo%b5$>%4O z{meD*o>?7G@TuP&ir};;jfabUAI{XECfo8K-5D=^6YjWmC|n6ZU)vE3nkb3#!*s0ZqT3D#(~sXI!y%8$ zX#5@zwu28&|LcQajU^kyi9@bdvN1M$B=OFW-9K;oMvM9)l({{vcI@nDZ2)bsF4mel z^!m=uzkcukGfgKS9R5Kxx#PssNwQg?xP`B<0~EPW#T?~zGmTJv5{pspjWP&t2@ZK) zS09UlTw%IWW$!o6y@>abO$HLMR5GmQ6p#cTM>LMkNqDLEmRBm*uPAzM6$ zpBGH&j(l=n9acKT(6s?oPrR$~Is&|MnzN`6j@{pStnYm= zvV$(-Gvw4UF)Renx|>%DmzUlJXod!`TaA_BLX(wh3SOYCVViSG{q_|FpbW zhJ~Hlg@zd?rfHR@3dLbMgSt0=@ZCS|ejD51$<}Yc3AWOyd&z*JLa{e^XV^900heT-Logj~_r3!E zF5}_9`+OEN11^`QuMOB5oD1nCqlD&|k@wsgZ*##=!216!s9F~+u<(u0@V-6a)q9^ z$wQ|ur7w^yrfX`Guxl!`qHFkgcAU)ur~HOo&zw;9LfwWgXOfUjn&J#$QAf4 zGG)mS6nJz8+iR7nUSC3KaiDTB6}D`9+77bkni&ocJRh6~{uK8|%YAAtyO2%{GAp~< zUDAc3KB=Bw3Z*@=>GjGxbR5Z}?-8B4k6;DVIv(D+19z^#ohxu>loQVLU%uHBqdpy zr~f}5D6kk%+dyH_XdF&Cf#ManZ8Oh~MTyjj*FRR2NDJwGsxGorupt-~Y3oCVhzrJQ zV(VnLW+p)e-Jy)_77zp4C5L?{4BMqMiv#oKq3)z(k!BmHM)W})3N8O#Th0hgZpi$z zYY&j++*U(QycG#r3B$apOp0Agku+m5222#ufW|+{{t-2jwVKNOrKsqo(J4@!5f zvJ}@=@g0YIPLBtcD>xoEbnwAXdj^v&m3y3aVzHvq)5n5pHE6R`%|ZFEJMK%=Ykbeo z+Z(uK{>O1tN04K`^-qnD?IZnEQ-@yQG`CRU$NM)~IviZQZ6`JwR@@UT>r?H9dYF|% zikKfBa>bQaLHIb(Xe|cbRtQ-%bNFfNcdt*etXFEKEa)M6ozEFiKI#=dIl|qQ8XA?6 zl6}_sY^KZkWr~Lov|i`4ke38cYn26br>|~)N6;PEk-g{JC1_F_TW^=HaeoMTfK9#z z{$(NdYUeMtt#5q9+1u=*eFuTtm)2>?a{rm|r|7ZJv7OkOSfgXR70Q~RN;>!&=}eL& ziC5>}N!k^t#j(Zf0Dq`VlmoVjt)q_{!mamX?>pgB3Oipjd8aNC)h?3n#Ke~aX8z_5 ziiLGbxiR{_pY+eom1x=_5RW9v*wCzin=^*g@j>36+2m2dyXM^r@plc*Ggldd0yJr{ z9x`a=mq;7==fstYHsKeFn=@*Ix+66W(!;z@#l}7%1=X#>j-DzvB$yeukxuLe2u`-{ zm7+Djk-g}pOo3QJrv=mom5H(R&XBP{MoO`$-@VH_*J<+PLd|Vu~H2?;Osd?82XG=Q7wY=Lii|v6TJHy8W zsj;Jo8+?|%+g)i{%G7DsJ=tbv=W%gT16?BB8;n&44KVU$5eaf!^xD}ovT0y!!h}sd z+4muCn0Qn3Lto3aFqbut6UWo6tem$*6w*nOb!;i5j@?rjXcP-WUcp{h18hT$HfjB!TB-miyC#$qs>cv)k`VCy!)cuhD+wDz@$(~619O<5Upcg39q$Rj<}rS*1+o7F++2%)erwmh zS52$j!2eUdfUJALN*nJqb3r##Y(7OcQZY%AE=aAY5hn+r^Z-f_X_2Q$(0fQCBg0{UL`%BZZlnv-4!Bim^Q;>vat;n211Q~0wEpGBKD zFCWXK^5S`+1{$L#C^>tI-s!V1M6;H+P;@2G;CGzlPRV=Uf#`T@Sm1)~h#9H1TgqYbB$3er?^WCq_c*BxMW`(YmPZtN?l{YCL{X{kv3Ks30V-&5bif&RG zMT+|8_DOe0yCeG@X{DdMn$f)5C#@ZKWZhnmSg*oay}fC#3O}7@@-fru96ePAX_>yUsU9O z-F?}_Wuh-;AURyC!u!KpXqESQWe~aCLi8jb^osU`9}4bPr1A8OUB9cNfsWZ1bl8bK zf%K9DU;CQL+_b)Tvw&>pMm=+4U$)r{VGR^pN0AyT2B`T^ zfN=TjY`UMnB|?*_KF+qvvL%|`?m7j==X)ZD^i)vDx`xJ!8>VdKYqA5915n`&)UAVH z@1|B}@WU;iYPWtmZ(beLWz{pK{I#zQx){$Py&j5eXVaf5i=Zf3I)kB@`rv9HNQ^B6 zV;|5BGo!-~$AHB>zS)<3OfdS%sR^e^rW5ZG?=}NQF~t^AB#(+|6Yh*EpOY;aQ0xYg z>Py}$c$>h2Bngsud4cOhkSWF6=8aF_)nfk@yhCL9q-+ShteaLME0J9molwTP9(HW5 z$JmB{4CDKprrajbd)v+?Th?26x)z}mTNW!d7;ukoby(X>WT+KF^)H~e?iZJdv1)UR zvRPd0rqKhxs+$3Vz2u-Gb}rbFR@vvOopJ}vPUZtPTHKiTV*fMixXJVVfk9$2D7nHl zS4k4LdC7?b_oZg$M@z9;6j?{b94A{MPO2VATV*ReGzR)H6&iQ^kgtLy0P6?3r2z@^8Jooo)&8m29IcIvt&K+fa$9Li$Vb z3P(tpJbBln8~sO4#gl&QePrEmjIh{$rG8PJ;SVOPynpQ!A>9*6s+mI4qF311SI9#u zCR27+TEXw0+9t#@gA1tv#V7sz3h#}wOxaQJ0m}KkGxkmXQd%M^ps%YfWyHH+8}|0( zgs4OQNrIa*wksYA9tye?1*&!-w(KME1arKXPONs@6^_s8)GIvB(hGwUJeN&8Odp{! zTf3j1B!D74{6<+n-QzRh(id4K-XXs4UlUQmA8=XkTBNR?vXFP(ZNTMpz)4l-&>TnbDh*`RS*qb#HQB6mz5aLEd-56LB0!a;~hePNpZ^}7*=$eL-1 zg8GnVy2v{_sNWs4(R*k7a%_YBXwdEdlha^3?^gc2c7D8{35BYa3H!+gC$?C7%`Day ziUonnLMjHCl4qhXgK+A0|87Qe0uFm3bAcB==u+X0o(-yW_46ymyHr@Ft~BzD@77l} zElR9Whbom=rEburkY4Y#ayo8GJI3>|1CZ@)u!j2M1mp=D9;{Zp=tVrRXd$Rmgjcgk zyh67V${g1s5hfYjBW)yha%$A}aV}ZM0T~WKfVECJA%nUR`p?aOU&Y8WepX3$3Az>c z$oY8&AqsyzeMP6f5Yj~&g$?{2feTuHzH)x$yNCa6+0Pf~VBh2mX6+lN&~~HnR8xmu;N+f6c>nwV z{3pwWj+OPq0l{wnO7&rhR*8}{C?DJb>*`;xqCRf_=8bnYFEGsi^?(bWmj6rpx7*(9 z`97Q$LF64T-TlUff89PGp3tc`ff6y+(Ae$2S$#*4J=LC5IlNfghh@Ys;INDK9VJe~ z;=BvvgvH5zb4?55z(p%c$ZjWI7@sj)7$2e7gA}Q!VvspeBI*T3M%A24v&+Pq9zTPu zK$Sb?qIl4yMTzTXi0dUV2~0Y3fNzl9i|%*7EX{!p!5&$oP}9OoXIhkcCH{YCm!K#R zni*ehpcKrL|C;3e9#}{SLxqjPB;aTm>#5@MdYRj(L&EmM_%OH+KO80WIbr;j zmOuU5|4cGL>BfIu{3}U#X|@eulX4gdD1%~Qs$EUR7|TX!um)<2Qd8~L1f}Z`R;UcA z^;j~i5JECg%5EG}g*YN`@ui|qUhE3?HeS%aT3`mQ+u71-zLl|D{OkvlhsrC7zxz|6i&OwRMA*-!$8Ip2~P&YCQLkeeB6#WVd9k^CU;%8 zq~mglq&e}_VuifBW=cs&{M>y075Xz^(dpDZ;wnj-@P_mjvpY1=wM@K7P|hnNpmYH$ z$;RXR#MlPnY5teIbHcBy7u=>nZw0#1sbfPc`PXMRJ7Sa2CWJknRjeJ@A9&Se3GZ79 zhQ3hLI4iQCS4rWt4(6OJP6YZoDZEB{*W5$?oeI$4QG!A!3huN)Vp*eNb&y5}%Ak3S zLAJ)DkuLYcjbp9Sq5g)gF7`+$us;HZ^=RGPT-^EUZOh0j7t`Xz+c;L@v`C-^Wm&p{ zIRrYA`=?|~9fS&ocyv-aIMcEC7&auPLz~UV(@ht7e(PAHVo)_7+m%#F4a9 zW&w>limjo@UMl9#AD0A|%35VbQcao!)z}uZIz@eO66oj`1-8m`yiRt5H&Sge;5F#- z2^tR|@YC6$U`f!l#J>i+eNgIh&;{xGSO%;ch?54WrCcn!57om~1}q9q^{5HRhr9cL z$FdLDQFwpMzK+eP7|D`6^V--zW8IF7K$e?ORrJ4YmQxTH+v3D-i?zv#*%cheE`a*b zkqfpdwoko2yHTuDFHwIwB`K_2(dU`Vzu~V_Uy<(j?xJt_qw_MLxGe333L1C7(Q&Zt z89k`R2Ab!!j_Cft)7NBb4D)7QCRt8wYO2gkO$o(6B@8Jk0!|2LwhcHr{IkDu0o*knVTyaD ze{4~nHFsZ{(1k zkDl2Yb(D9U)ub~O-dL`C(B%S2XR4wafOA_#m&(%l)k;l4_z7}U*cGMeR-E$bKGtCzz8QI8GIC=WP6&^n@)|4LPE znL1=8=cY{2?71_l_($E*A109-5jz!WQ*b`e6hI)Xoi3pVX630?vUtms<{35W{Z%f1 z+^F6;mk$`_v)14F2#%x^uQ98)yM;d5B>wKLZKUX>*(5$}W{j#R7B+z@shIt|Y9%H) z=S$;XTO6wSQjFFnR-{P_!1zk53{~QchNj#rm#+=jHm#2og4WM*((G48J{N-oa*{+F zP$pgs>#9QdVpP!JB4tcXXO_Ah27##x?~~HpDOWff7wb42)mPwIcE|chak{hcjgKa= zChWN1?>$e}jx=34F$of+Lxwprg%k?{letvPR=0LlPvqm0+KsN8paNu^s7MMfZ@FY8 zl-fqYFihRRf<4u4+kn5KRq7dBa6c3z2{g;qeU22J92*`u4Kb&Ot((NDiSO6`eUbFz z7A$%Xoe@3h6~o@LDC1SFGLu^3-U5koi$Y^XOWaSZ^Q50crS1Z@Tbk&JYwE3hJ(B^? z;5p(9v?{$4J&Sl)Dm6t=&i_=YleIFjAZWQ$woz(dXL@XyGBY-uF!Q{J2mE)>`nk!* z%vxxePA)p}()zxcJ?Ws>&neQ5?(bQ8)lZH8f^xC%DxxJVVR&8{(M4Z}ysPF=3y4}sU1LrIGLn3cBYJ3pT@t+#bi=VAItkI{V)y*zEq9hn(D+ND|Dl zs_#>5A4Tp`F*$-NSoF3DjjO192$Z0nWHw!<$XDffZ=c>9rB$~3RL!|bgQto7MO4ng zRckt9yuHJ}Onfk;l*X7`o3MngVo?@hNcI-#`svIg*`56%%jRAcgVU0)+TdL*i4Uoa z=%A4xiyvnCh_rw{qE+otXo{2rA;-zZ08OE!oS)7l2nXHLRQYtC$`H}wqgfQH4KQ|F z6PD<+NAJ*71CHO?L=t4YIV@^+3kSTM*2Qo61+?VE4dO;JADo02ku2R zx;DDnt>~h??A(*-&>mK9!i2ZlAJn>=ELYXL-)bf+CXynvc_oWt*HQ56$DnX>jrdk* zR(Klde=QEL;GaQ{%r7hE_uc{ND!TW|Mi!rC9p^|e~-$KKaDO&`u19{S$t!&fck zshpPQS#Ac~T#D6FB#Vl{nGqwtMLe7whjLFK*bHf}`$BdQ6r`)6w|Z&=8X^689j_|* z!<@*~QO)8te8-13xq|o6ci-M%wd{`*cqgowzu@vaCJPc>C2S{kPV6#YHp560#h#?d zaVnH7byWxA=P(z@ctYLv3dMDL5K%>7=hjm3;fM zw2#DzE=d=P8ihH64h9|Tr9Q{SWsysLF8XB3wn>(FoRa9&&7OMj)7Av8lU?#n;&sS( z@o|l%RmM;3i)KpPR-Z}9-Se7lFxW>0zx@_WTtmtSl^$1%w{j>y( z4d`C-!Kp?4HI2p+A40?w11f{5;#u6iZM@b8C2tpeM2>41H_I z8>C_5$O%Wo+By30-`dlDX&Fj%+8GJ*A%_K+cTwziid0ZBCBbD8dtTQ-@;vM~R7Ms~ zYf+{MQl=rnt5OV1uOcO~uv%rkK`UR=93>Urn>}iRizS%lsL5c<`JJxF)q)MpDxeeZ zpQQC(GbLMc+D8j)(V8%HXS3<;{#-XVpD}Qpuwa7$dB#U^y2Z(+O!&^%@)SJNJoVe7 zy+_ERi4g4{K0vD}HiaT9sTgAcOrYnY@~t*t>Ex}!dZyZM{nQTn!*Tr zV}6*^%XeN6_m#5QSuQ42lz#QAPe`m2doSS74znjI6uXil%c+<~MWq<)^bQd}k@kQG zqj4U_3JROBI@S*vMSD(Iogm!RBeASZ!NvJ-V(ONa0vk!}4bRWrANmZ^r+gZPeMG0u z;bnSl<85Rc6#M4vbZg+Zc{|`N+L+n|D2o}#T3`0Oo^dzn zCziG4oz@GnQuH26+b{K5PP!wt%3Sv^mDgwQ52;nu@u02%jq9O(!Z^|k>MzBTtblF? z3)SNi6@tQPJ+pEoE$R}H_Ps~f>mU|_%a}}A zKeI6a%k{MIp0rvlz$Nc#wF3dcW%!@A9?u@avE1cmMckI&`WH+24yT=dtmumDpMM2{ zX*zXlRJLC=$Uh$m&IU=!#&M7vd}^F*1I1I#9DactD5m*s-!c~Y4(DB?S(WdQVekXx zIEJG72l#qk`P^0(7wt`Q<1Nh%2(jFVz>_@shBwSQyW|F0a==#*Ex8(!z~qaPc#bnS zwgKyDJL3>XWf)xdHm}#b=m|?0<*K3Az7{9Iz^R_;XKu)`e83mMkiyg|8(=PpfIxxV z3ak@=&uNwa+s*7%HOcW|d=$`tiIbRY;1~>yar_Ym!$!@1hnx({E5GsAE}m=(k$k7* zgI|&LBk_El*nI3U3!9WuENpNWKp0lEkp=#XtVKDb_fp5J5vK^cXZE{S($$bni^0_* zULQy)_A_-nO>AhfbOkRqv;~Ce23-omH_`vFgH_->{pVIUCxw{wF4v9y z*y|LzM#Ws8z0T*3PdwB$+US!lH3-+u8E{FT(!}Bs6xGB$``A!}U!mXU;^QPaU_hZs z=kJ=cUC}3P@XYtFcH1AaUU^b<4~`*oUj_exMmM;=P4ZG+moz=RMtlue#Rb6dM#aK3 zet~LlaP>S*iRcm!i;wqvE)@;9q(~BdkRHC?)etq{(lFx?)MPqDGGzxPg+Qw8iCib! zB-QgY2W5?N)PZl3HwqhtSaAz~t1fAhq@PKJA~se4XJ_=7W67qokpaie{oNpWzu`sa z!1cK|h28Q5A7ztA;j}gyNwAp124KO6+lDEfzNh3vYay^l0?c&mjnu4vt5vqkk$G%G zgEMN3j6Ae1`x8qY)5%>9; z-WeH?UDmGHC_^7d&vejDLJcw{U`4zB_aXV;lUc1*lqtEA3)Y zmRG9a{~0H4#;Nem*$B&7%gW@?&)0%B!U5H1GDxn$csqVQQ7P_}&n8G*HjM8kE(cPa zV*;qw<`_=z-+2RF3G-i(Fea<9HZ|lNSu>FUb#>U6vw&hjoic}tDfem;>O$5`>!Q0A znjH`k#e_d_G+TK2sw~kqsNkE*FPwZ<+~eBitvN2fO8VVTgjV}u+EqC}hliWRj{Mqk z^dig`71l=C4%ayW#d%2|MN5DD4NIQB)B3$unD|1jAO+RE`(n|s~wO=um`Cp;_bJGP^!F-lv~m z%{ThZ>nUReq;4S3KnSOh)~fX0aY5+}?qUwQ94A{MPO2VAhcwi1v!RW~%>}D*7h6Yk z=(#ajNOODn=k_bOEYgf%Me2=q){R001c=_2cc zcZQ-#I!PPbV;fla%stCPuFNjJL6|SXC)(A$1lcs1UdCL;_=&0EvEn5YPkME-Ic$&Q z2)%EzPF>`=jfWb1`Jy9qYZ%vg*C9xGf|(~rBLl+fRc?mq>-|R`C``D@p8T8Nl5}o{ z$%)zBm1Zc+r`U}Y$);kiA=Ur|xvIfgh^k`#b*5erM}1s8Kbe=r+u*HJm-A~OSiFqj zS&RCAKfV}~%sZ_<>2;Ad2-6(9i#ZllS%rihZk!#s(Ds@MCoU4zE|UM!;N*ZAPIgc% zh(VWAG26Vg)9d`R0y2U#=M=w|Drf|H_Z>lvct_x2dcoiRuz~u+`hZ+_<6V#tLb1C7 zY0i{g{N~8pK{_=MaArY4Kb2Os(lb@CjlECr2(F)=DmbM|oROt0i^xzl%DI}HXN==m zaP!2}JUV*D2XVZ$`<&~nvFvm^ud@%!z2T0}DvC{}NFo(8ly+1|pY+Y;gX}!eW!!Tl z<%(U??c<+tdi}P7f>WnjO%5y+*n7$#cq;Jg1 zRrP{%lEN>MZFKD-+ro9~T1Bb^_a)ne?Vu7?MSrd+3fu@8^-Zi!-9Gc?j3nMZUX6zx z?ST2{*b0D0-+RlS=DpePjGK?L`4nkDEI5K>zK( zlv_?QR%|nJ(pqJy{EK4E19v1k!rN>JZwQA(W`}(Tt^DmazfCLc9d()0SJQdFSy$Ik z9hRXuD-o~F9#wO;2Of^T8huIHt|)*lGo5;~2gbqH2S1432pTBkiw&ERF|P40uBm-a zTi5kprkt*_Obl^a=VPbY{IHo~0mB=qn4EA_S1uM6yDg(W^IRLUOq5D4`QHf{QXU&n z6!JRB%^97f0#p(8z?93Tk4T$*J3;wLv&A!3)CC2g^{*E>vUVB=3|smfaPBDE#u394 z=HAP=`7b8No+xw%yom8b~HsDKWjJmld6P!#3u1JtNs zP*Hq=ASQ+c!5~3}Z>^a@nQ=~>12c5AU;8th+2@=+aQ6SKwf1`aS3O0JBHv#z^u*IK zm~hl4MtAp)h(kd;-5v_FK_4K+pSBkYdyovUo#{|5SL)T+{TU+GO|TNz<)w2AB*os^ zLi(mKmModK%^wJ{4Q;LT>ccN?^VeRHwu!O8Fg@^7Wd#soF7Z7}e=$1;5{>pNb(0sJ zQdcX6|6Q7Y+IP`t7=!GwXg0j8yw{D_qB*Z8>*{}H-}t@>O&`s!`HUQYZq!_-*-LSe zV$W0LLn`L8SFXvin5{YbYE-)#V@H^M(IG}!j|4%k+kMhaDrpRSZ*<=a?UhnB`vaPy z?+Xef7@rz+$)-2@4!YcuXF!Z;x3C^Il6Y+$eJCi-ca3PT8p#inWXnUkp(AN7-3Rn# zyXU32E|K&_ZUwhshOAzN`CB+QiopaK&S3l|jCY}CYf#*Db zw_h@0t8B(!SCWF~23yC?uyu%H4^U(u_-+EuOO4oC9aW0z#ii!nr^ zngp0Hlco7+mNs*0o47J`&-B)?wgu~ipDJ@_Y!+fO!U|}Fpsft;AqT-a%%LlV*pATP zcFvuVqp2b%H77M%BY{M|3>nRPA#r_F;Vd3#vcX>Cx`aQ#*1Y$J?QSM5^^W%Z+hi3# zE#qiAQ_Ky!w2VWmY&JLF2?Kh$FeZA~b`EmBcFU!0x;$y5(Q5$>wb@1WJ;r| z$vA=n7tMBHjCP4fci;5R@WYnCa9Tey$5Ll$@+GU>45f~!Ar_S}gQ`-WlO%I0`j>}- zDyEEUT+H^R2GKvdRU6e^y!QDM_m-{t!?R8&q+<%9#O+`N)ScZ0Ld+$S!boTbgMEw& z=wj9Gxu-&)+J&iuh7lRzTD`hyWA=Tr^q$wkjqE28dS}i!P&RK3yHbltEO-E z)h15gJ{!A#RYGj5)GciultKz>gL_q6VR5g3^@vwrdC@1Ik6?X)eL;=D0U4+;;EQ?2`=x| zfHbxC{mJQc)zmAzQ@O1_UY?`2poHg?cDbAzCCavkC6?VCIhrVLf~0(;uH6Z3FSR|& zR7E=71TFXyLi0WKP|>W#GGl1B30!iwMH@nIi;j6>-+8Q-AHD<}|Hv?w3vq@!be9qu z=4}tJ4T871G6Tw z#l(`O5`&{HPIN7xi_R5hLVMFJc?Z$f&fO)&oX#Z@OgBqaWX;fipicEF^*Q5%S8A^X z^a`uU0Z2E~P1d6O4(s@@$Ok58X(~gL2n;rZ#H|BCTs0M#YAQnyYIdUIZZ-5By{btR zRk$@O`$#iu>^s^Xv2!|Ju!>g{iN|J}hXCccln?C6UtM^>(*&uX)@;2%HaM}B41I@3 zzy56L zwd3Q?bz9u=KfKg_qj|wOjn1!;bnDhT-aLO1XJ%v;umE>Cy@&24H8jnvvhCv67eBils ztOVH(F-xXzc9N_QKvPJ;&7Xg(`qi?3>sfH{=Y^D`FvJ5bj}(&G!Pr2-dH36IeDpm( zlRff~JsD5*PHc~qo2~yYiY=hX4k~6nbX(o%9>)&GY?tz)nc(kekoWIt52ttW-x z{QYRwUe!a2Hp%tNDDZQ(nAc5B#W36;5jM}UI8q5`2oNkUp4jfuo}^Q z!7eGNCX7EF^GkqyJg9!#;e+o>3gH37ex8ru#W7Ir9NC5=9fKCsb6Z(=%qpT?qU zq(am}6^v*<#3+rnZN3bG5O91`*i4Sd^%@6>BUr&PVud#bY`x*N#j>tvz4dtE^!$J_ zcjJF`Ot)LT;D+}AC>mr5YK}(bqzi8lmjF*pA(ZK*`6rVBI)yZ)k9K7RRe&EH1feV&)*Kajtx zy|%Cm7#w!I49y*(3NV(^-g>2B0be|g5hWwV$(JA73J=qD{PfoJO>Z?>He%yr7CW)C z7t(b{xM}h!HkTq<&1=-r}vy3T={>@h-%Ctz&9{0B|V+oze7Ro~B*E66r} z%F2nsQf&s7Vv5~OkwPG21)8Zss9d{fs9cL><9ts;VL(OHjyYNErmyq>O;uUY!asba=5*2P&9RQv&f@W9-*^UMq zue@+-;>)2X(8TL{_K=;{`YAavG)|g9qk>|AF=Rg#gB=SRRC@%#AOt-J8pK`FZnhm- zSagNeD9Qt?rtXt%aV-ZmvoU;M5SC$e(FeZO5wX;3Ux4wBLs3VBh2a&fp&ljd3OIwi z={O`S&hnUOFpi{R*~%3xU$Zq$081S)s3eo1xKO(}WD(y_gl%A*ua_z_N=jP&vblH##W|j9t1KTnP zbd{>Mv6z#CDL`PLC&+Gi;{PMx=oc^FnCc+w5& z2qc=p($)_aNcH~TjjCsz!yW?$oj^vcS7#_z1{${Gy{a=l<-w^6Oryegl4hG@GH)ty5*p*GiC;%{?&h2zg#jMA$)|yMQm@|Wby#@buRu~wS4!8xZtk#XS46v@0j6kBZ3x7g)3ZQDuRScx zrF(@9u9#jDAJwVRwn2{{jEsIp3`5G(L>n8}F(pgWE5xuodM$@sPN*P*41U|1`3ZtW z2O^t7F^%l`;IN<2^sxDXWduV`M8E&#uVQ}x#%~wCEvH!*LOHR&h$Bv63|{5SN_}8+ zuL0?ViKE$?_^2J>*o@#w0b%i8+W8|7@nh$2{_^R#72(gy_y2%3DMI)AeV524eu~hE zO{{%piZGXA0YjNo%x%B*QwyS6nFeJwI11D0aw$eb?|!Ehw)s$i3t_Hn@(f6i$P;|> z%ElRqFBe2yjOb9HDRw!yMvc2TyVy*e3&sOOmdc28PJ5jAj}KqEW0?xdkzJWfcSAZt zkvB3iqaw(cIrfop{dqP#O0HWSI?nw%K2te;-~V{w1&)RM#4RjQ{ zo+9h0m_2TtP|mU`D%l@xRctS*3%SYE__aeTk_o@}Q1VfEOT>eGN#^wc+`)JSd>8nDXvotme(7 z2i&!vhFuBOt79cgd>8%i;#Q`|I|15V>X-pKYkmzx$S{j9R zLx_%nY&t%w(X&!q1r!rn;~lvF+8B9PY9*oCW=vXYBl~`S0+fq46nll6q>t)v+O1?a zKg-35Q3m;;Bb>}t6br%lGAagHb8rKlJ?*88NGT%^x)JuA8khDhA4Yv zX*dW$PauNtKyYiyfd~_Vq^=9LlRZwXzfPNBqMBkWC~}aB8GhlBj=t2#(0C0?_p!|d z^mKy;LLDG;bc~Vt58~xmt&Ie+b#%Q71qsyLO@U#QD~-OlXb7U~JqFOtgAM8msU?yw z=r@!g>cnbq+`gX*Ns6o&Zeib_V?|W(SpWxt0)-BrU*m~ud8xVynr|h3Z7ka-C-x(9 zY@c*Ym2{W1P*4f#rvuETa!n6eN_{B9balNN&$TnBJKN9H5u+D!mUv`;X8$0 zfXP%vzrxX5@-g7*2`t!iK)B26o3EIJ%>!ZP&q<{d3!BSk!sax^Lb#%yin%UMcJB&o zVfSbwb)p^mtPw z{ax)xZ!||^OubL~Kz`kO`1+kcTf7-sO^w1t9|J^q*fjjX!%Or~u9siw3NZod)(>wN zlRSR*j}wcYIy3l{Q7kapmOyu?X;;8kgYMd%s1C4c^Ch6U`p~4E@xZ-R+6Yy?Il?0E zL6>GVM!veBM6ip-4Ag7#TtPe09t}oSm?t=|NQ%r0NC@3HWoWiBWk}nktPRcxzd{xs zhYJ_lO-9)tY5Y-T?T*{CGm|XY&z~uGFb7q8zGT2Xm&ULG^ui5CN+aBIA`Q!1vda`U1_)ACbL^e=tB1P6xF}m<{*+;Y1F+KDi zAVpjX9Ilm+%YYOhd(dMu3P=&H?w9!sbrS`@cum(O-QmTknYrPVPS zIP$uO7j`Fvx9Qw0OLv_XA{@oIy-bCyT62<4^o$Qko_b1scXl>p>n)Kq$@MSZoOy7; z;{SWoz_}^H@Q_};!=pY3FE5*Wm_DWMBW)f>6=>)7&`Uzn{J{>L?p5i9D$-V(Z&JvZgryyoZ@;a}G=9ft1Vb8n^?4Gn#PMqSh`e!bb2QGg>8lC;HFUp}KfVkL?|6{$ z#+UK;;>%1Ov~d8U=QEe#7W4-#hE8U$Jaz>a<#nqfKcs zb63$cXQh9tc+h3D&w8@C$H!RBkp{>Uu0 z)q79SirJgld|9^`9P9be>Tio{j;r1L4qyJKk7gadd`A16<+TpZ8@}uGdvm2_Ou%Uc z35S=zl_?HP1b==Rofpul+NRJkn`BMOu*syg4^}y5;RRu8_h3z~C4B%#-F+{!mL2f88kOkUX-HVr zRt9Zjdzp51wr~rqQ@;WgxQWUGgL%lDh!#Mja==5cPKpHThR%qsLCEi6kC3j_KYt91 z;?T9>wJ6T(7XRb+>P2HQVV!ul#9_j&Cymn0id|E(45ZR!-@Fbfnn)`hN7k_!@;dr3 z-5~9Uq|G{68NJ%KB6P(#xNH#lqX4?q7slR$n-{x(dy;9B6W#HamMpba73I7LC+TKe znIwu$pvY>d?w*ZetTGgHQ_|_%FCY`mIN3@TOJvi^Tii1qAGt-`={FlreZ_={mT!NQ zL-smx-1)58`X8a#8j4g>F*k)>ftYOoh5Q7o`&YRESuiHHzz?k0xGBU64peu21T7HE zoVz;>|7$22S|Ul8b*jps_~<^If)49Favh|@VSciKGcQYH$hvD3Z}Ue6ETFH(DuUrJ z)xvGk<)iOvU=w7X=5G6v^P`0qH)O2DCnm=2eq+@SO)xs}vw)wIlh#Cw^BU#kBQt2V zP;4_r&S6WKS+`CU#iyXH3wDD@_A9CYqI}Y zx}V_Sr^+;69Lv(QGYRBYgkD|a_s|C^7jJ>n17O?Yv%&Ry%PXirdC|Ax_xPJgf?)RGFH(+g}xvNWgkSG$zG$BIf@#kYXpct{co$Q7y$* zQKX!TLI0i6dsInpnjR}!LZLOS#pb3+jMd9z0r@AE2!OSXoxcWMc6&8KfT2NEDa8}e zBrYF_0(u!7-i91oSeC6C|HR3Z~-e)kg*OK{!+ri4_TBL!NCQ;(FP? zbbRy(f;)MEXhOwrGWS_lRywU@;VAhW?)89Ha9r>SP0F;ft?XzYzYT~O?bKtsj~^n+ z|NHoyXI3iF9bdYG;Q5fBayd_IJY_cfo#us&3B6Sr{%R98qEEh2PBNVs8(^1?z(x_pLSAnH zVgn7+sySQ2x2SuC+8yEjq&VmX)SwNxqbZpi*v_J}md?|p&AUZPA`@h}bT)KqgW8VV zz|y(5lxw4gT-Hk}MOf8{#f;svS2HDG`$C&3!~g82ns{QyT(DUAWG~e4!%XL4WwT`? zb*EK09Ier}`lkow3ERXKq5FK=0V?gxRYMYPf?!|3!ma-Gbg^Ke-Im>kW#WYl8|Jgu zCA^S9{rUb79cGdynf_TD$u55WS|^SZd|)PfDk=65MGjCgA1YQ!K2(&*^y*~_*c@*U zs*^35y?frGN@#U}p8Yx+ORP~08NyDN3G|p$-Xs0wQ>9^GA9Ov%P!-bdAs6|g;-D^I z%{vpYOWH;^dah;hs(e{xXwCfhfdy~9v=TPbV=S=-Z194c6~V(bG*;M^zf<_?d=r40 zUSFO~%A8moT`+^haf$_>{$VO6cJhZJ{MQt{?_1a87oiaXxbMNmu_`=4FbF+JyQG6I zmD1Ya=ICDGI&k%`_TB8$ANf9Mjy^)}L1sgha?oXe$VTCyOMGC~{6Uv%LQKrOIj4(B z6?6j&=pE7BX|-Y`?Lc449Z|I^ZU%m;aKle|(^>*nh}uU}|N3b09EA=8Y94(H{{w!) zC^2UF+EwdY*T~uDMi_OQA+L>MuTfB00EQ4!mZ7VqB%;Qzm02k%0DoILz13Tb zvgo?^R&Ur-K^Aaz;4UdC@xgW!3nzz2d|Ik9Z7bj~j6$Esva zLg;Q13zf6L^^h{H6YMRc)anMm9a6+(YOrEtdf?B}f%n@W*QRE*Krk5rdKE?vw;sG!Cpj9qW@-v(6zLoUZ$I}}LQfX7c+DE@CX z#~r*^oMRq)5Aza26EFX^;xGSP$$1B*HKLuK=+QqL(5}h=%IWi@oq1>RftS|_nxi)p z0JZk<+Ze@KOFvLrQQ1hoPXtLaH^T}jz;=E3d%{lT1 zJqjzH_bp+xhk&=Pbna>SF*2w!<>8GJhH&}$;m84i;h1vp8*ea{;h^UP861fUJ3K0b zOd;nU_i8A2HPVqi)qLc5fMXf95fo$jic#Hrq~JfJpY_y(MrEF$4w?~Q1PGXYWEtUD z*0M7q9b6(%v8ZfR;symNgmnz266huy|9>>NklyBRB-F%`mP1e^o#!JMVfK&@*JRNUD?rq~FJFWquHd2&=pFuRGD&Y5sk zRCDkTrftnDJCFL3kDsx$-c=JvZgmDI_8vv<8h6M@)xFgliT`t7Tn@grhhd+GwTX>{ zi3qst@J#msy8b`36$@^WJSI`JGeTRf$YVByX4983g@L3hj!Qo9Z}zM59$>WBAtQW1 z_~0w~GW>{B54m(q=_eI##wjlcZ<=mmQ_XNonN~%N^T$u_hk5RY#WRNFV%%(-u4?kU zDD9&0e22JViXqY(e8Ts zrT(xU_vUE4sz<4V))ScHTPfTSS_*sdK~+-JDRs5tMDU=>4)V!NDBuUX2&^Y&9s2+; zfko{&p!*-o$b-|yq&Q*;$d^@YNX+f>iS_?ndD^R+-7m;=FQZDMD%^8two0IDu6RGB z`NseIxw#kU8byQRhQ_o)~I^F$Zj&a87gNwfzepxKG|tf!rXf{jnzaG7n8;4`CruH}YA zQN^lqcM~MazVY3&WVI7}{qxM0KaFBHQ}DzgScDze*D-PI(&)Rh^y+rO6}PiMi#jl` zP?4j_6z+Wa_$#n`A{WP)i_K>(Y(v5$UBPGSFY+8=Jz-q1u^1UsMNWv)p*wkzTajBU$(?~z9w{?+Nn<57(^k2S%Tt9v zlJ|qP@xhGsGF~sL^Qvr@obOorXuPdG;u`Y z8cZ|5a{O_U90)|dARjAuZM}c`WY43P1}PtOII(`=Fjb8O07&&<%tfg29BYM`tL37J+9OJm(I7iX)!} zpC>x-lLGCpI?rm}!=66q`+RCb-<3Z+Yr%fP;D0!$#&3~8(ZdH;*&TR`I?ZWX*IbWw#(}a z)s@gKK@Jxn2#3+&$$>C>H1+f=$NSw(n&utt`M1fc=SI`ym}#2L6sx1idMYMWu$C?L zX$aM8(&=pPDnW%BJshdZ1|^cbX&aR{8Dx=r^goBgV>Hk}J@aS{9dz(ZNyB?DnxOH6 zg8%w0*}@MRPAohQnn9zGV!=U{L&YS!@6qfR4um`uJw9rJd}u{y>NDA;o$D{ikeIPHL{&h>kghQhz3)>>H z>5ab4(fy>%(xCDg=>dhpP1DEkfx?5~^GqH697oR2YxCf|>ih5C z4*TNYO`1j`+3*^NL5EKSFVCSPFpg{i_VBgi z`EyZT$svhJ_PkXx`LARRKP%UHr7Foc6QNruHU)p#m@>K~s78qD-2f@aTZO|S3{eMFcbsMd@dRBVCANI5rN5Q#|H`QKs_JYK%A=`eZUiE|Wp zm@RlJ#crZVGUg~Y!|r7{WF|tBG+^Ej?4grsgQ5;Dy>I=mgTF zP85NVB5T977%Xa0Y@XZUq16la1tDDu(h~L)@Z(~2UzVnJ+7+peNs4TsE5+UP8sU&j zqq0zNRX(UfTCj9_kDFdy8M?*?zqG-%UEScC1pYmPXU0g&Qytv-5}w4qZfzchP9PYL z%%#tVc8K$3wLy6D48#<==sLm2QtcV~Xz&^OEVB0Kn9i^a!5QG&sL~uKhVD4+(~4zI zTOHW^|NP5E%W%>f8Q)-7*&vB0+(3>pIkZh`zz zkd8?WyhE>m#-I+DR{Z<~!?Mcpet-DAIQKX|{Y(5?&u7X?xu%Dt%POVo{Byup3#@`0 zCU;4ZQ>%@>JA1XqIB$daXfj7@uaWa}Bf~f9e&uwN)^d?)_LFRWTFZ&u|A)=&-=kjvNiScHdNYQsTrrrbaU)9H!W6ic}a04jPq4ii2TN1Mmh$fE2+Ld#Dl-B>QVy4fY1M_kg-D z=3PTy4-_zo=sApR_KAnMb33CGR;aa20v)qkg(L_DiUKriaNm_L>s1-fuW_po^+p0^ z;2NJRp?4z3BAo1tizo0hYItx<8f%FzzfAbUA1xhqPg*WcEMqv_ab>cODVL$(D4X6N za73h6-y-&u3AQheEhw>lHv3)4jgkk24Rwr3u53yQzDPDsB74nV-yDkFPLT{MX1k_R zyj;0NvPRSoyVo-JJmKfcTTsj&|J{qxr-4m2DeAMYB*@ytC%n78J3zAEWX>v%2BSx; z3z}(dRyf#FwB^}9@!lB*?Dw3l&BXoet-PVm>-W@5(5!vdy)uK-FAD|(_Cj||ZQn~B zQ*h@QH+@5>wqpurUKdD;y&qqJJZ~AYhET9FkYRMU>hw#F1@P$c>(~@0Uih8(?kB!) zTE?R}w@FrB04LrxbA;9R`1gb$#q}AVfqB)MPR$;n_M&u9)u~E!)fNd_*c@RekXbku zJ7#i+ACSy$b?67lilNItHmRBnYT$e1v=f`KpP1>us}$Qpk!C7p&Ajp#2URDcx@g_( zBJUFK9rT*XM}&Fa{Xmg~q{Q{~5z$u9d!&{?W`%mA??KHG;RU)NYS5*cF7jS8?;LxM zts^PXw}^4{pr$~uK5_@$MwYyIH?sW2^uSs%HuXiyoX+TKMPXz?)Cs6c|A_1mw2}Ly zLENLNQ}s?8R5eGhkX{xXReUC0IjP&bQm}*W_H704TY}fe?mGktf+M2Oq^oRgT*VJE z*0m_VQQMHlOZHI9D=(#wrS{Z$l`j+(j!dn4M3aJ~>KBZyRa1hqB?$%AHOJLqG6#UbOl z>Nc(Rusj&Q#r_h9tQAXR!q3iDs>4nCNcA`ERYpeGuQ`vyihfat}p*- z7wmB>p}({o{#;-}&>xTY)stn*AB5dO^FY#;3;TJa~xeecIKJ$2v8qeKUpog@-`ieZjj z?-6Yj?saU|gUwK}B2Mgr364ot7aV`v(mdg_TXtfoaP-Q|5bb~YI#{gD(HTsGcu7ch z;KIIXol^8DW`(bDO$V2MX+*!HeT5^JgC90V-s#xmoge-=FvKKAZvF6fG0Ed6Mx5A_ zQfDSc$|x4(N(mJ+FfUn%Bw~G$OC&%ojSZi447yORNV_1i{Ql&0x-+a%84pYex%3|J z;$ee!yk?#Ma_HOM1vc>)D$q^n*W0G)V)W{r5koFbf@}ybH7IVnH7bowdhHh*;D;S6 z!o_09UWl6iTMb2-u=CL4;b}7H0_*`JHF^TYuBON;Dh4aF45j5*X@iYx$6JjX|7IJW zq`e*vuX(tQKPy`b=Y8e=*ZqWYES9SC{$+&2QcV@KGR1+3kW`k+Hj1lN9TBG9EaU2n zwHe$7LBI={cDOkFtm{5Y{*-4b9L&Lz(l)PIl@NNvb2qF$h<;@=y9>G&XfwhecHLOc?7_NPp@Skcxp>k$x~CYPIk7@HZHA$0imjl?K`Lg61nVM8B6Um^ILU5!-<^%g zXnOT!uO%Uas$2mQ?`-w#)HJA;1!{Z42Hf$|uCN+K7o8G}EeDrKy6HWz<27866P^Rv zZdG%JT<)qG=2T7TqHl+!g+oJl`nFFOjYK}$+n)Indnl_h!|IU)GCLR(r0`~LSiRl# zwLkfrAXGT?dMU|pVv$p62A4e)3)$8ARLm#yuE=90C#Gq2%+82jIi%_z3qsmy{Fn)? zpVmgD&C@3PpO}Uh<_WfXCkcV{zD1$e3LwQi9ZVg`g}PiUGqMLD z%va8C@Gzft%%8a7VMXdpc=xqG+_$VKdrmwl9Q9{$lRjOL_>;kJZ~j)zt1Bh}-P?>c zqJeo^y-z|XCaBe*69nZiu9;jjre@nb7m**7cuqcs%df2b>O|HgPu%bKT_T&TsU#;( zd)#LxV{$1Lh*2}Cm<^M!xb5^R(WEgc1e%$C^Wwj1uSgF@;bHH$YZibs!31gTK@Hly zsIitte=tRhxlTr=z;d@Lk^KeA>^82;A!AT~hM0dQUB}<)H z^8h8}2xB*iViPE`8rg^QCAh}P{_W~gA1qHtUE|nojr9exY5lF=>d;Z{^|zv0CQkZm zplU3>6DO7{9KMr0K|i#`z8bJmlnI-qLDd1*WN2KOC>p;W4v&}LCJZ@HRy_Vq%#Vcs zcEH=R5g;Ff!--K*WhPgOD0UY`3aFSyCDL7>zj&huFaYY9m2T}!en6!()^({D)M7NK zN~G1IUU@G4z}-+l*zH;ylrF=%&%>}!(thmSt9oBhF{PEXh7P$Ddw0B2#m zh1bMA-V6VPxsZ4!=ash>eja(xuXx6a(1m|;s0JF#l8~3lHjHFqQkD8(@rT559V3e$*E2bvR$`3GJQ76-@ zkBDmg915i2J`>}pZ;1+Ja0ZD;P&1wn<1EX}>ZHgP&MkpvRGp0kb$-zdet2x{UljdHg`Q%bzvv+lxb+nuA(rqbK3*G zs{ zgp1{G`K}@9PV7p95Z?&E?xxs6isVr->jU=AT=8-bo%pqD@-2`&58_{n{qECc*^t2) zKeCgk z1l1&|YzM};h_**x!V4G^+FskV&&HOtjmmr(wx2C>)2<9(Q>Fdb{oLGi z8Rl_dlO$}8)Hf?-oTxZ$Uxr67=%YvF$OC@M05qq>?d_Of{~_khKfm$Yg>TDOQ0y{_ zEOA_*#>T(BC#QXpCEXSuw}}()i8*@5B?F>!=|qt+kxjdM!QzVxI%#b-D2=?3w`eSX;c-&a4f}17dVc9BHE+lQv>?!T7}b$U>6j#8&EVGcaUQER=<(Q8BkewhJE0ANs6g;^*qs__I0swB~_)wFj1L7Sfot z*d~7H^MR(CKBl<^WLfdd`{R-qG8#OLhc@BHgOvaGN|WcTn7t1ujde`9I6ZJEz%jd;Cxifh zI6X0iLk7No#QiF55+Z#+@jppoomhxumsXjL#u2sWI_>^Vgy+jZ$Ydpb4j$*$$b`jmkFhhRK~`WMtewZ;fc3 zFLFWWOOpK^nLK7~O>IO58Wz@Wci0#|WMuuhH*_odWT3q|`HF=L{mElarlp4Xt*@?5e>3yoB!7DYFg0WFmLg?Md4k+GA_HT}E z6XUgqg5rG9)0V0@Ngj%d1($;DEunaz?9q1J>Vci{_DW{@Z!FW|pA$QTBZF!(OvJ~MjrZnG^_rFfqD0!lzg`^ zL-2TP8PF3P5oXa_{IRCT-e@r_vOOSUXS5imQwDZsT2f*;trFrOxkAeEKKesho*)Ml z5T=ujx7snSj4g}oh(&zbocy$p_x2s`e>54Z|I_%Z8RQdd$?#6>5_#XOJH`Ww{emKW zR7^U3p7eM(sn24m=Rs-d+^XP=@Fb5){~AF%lOSkPX8;iZW|B0}kSqdSDHO=-*`(H{ zLTtZ7Q6lY){AgBE6m~goQul%Y(J_@^6JL$GsyQ0mfXsRV1K~8X3FHd;W@QL+>1;Yj zGw6bJTzZcIpeQijgvT{>zqF7(NVl-qHDgIgzT7YY8rRraIhS7JX&fw}D?;}xYDBd` z2_pE03;}wlwF!b`VeX6u*L(?HW$d_s<%+rVXVTR`Y|u?ZkBuyi)3(UqhUC$2{VgEM z1iwwWsti)<#CxSCv%S(WiUs?onu^JAZH!v#R^{DKuy5uKWk=u;UFxeHX$f-USR2+hgeP9TAQwN_WslNO?H)yPRW!^hCxN|HO zUrsFHI4Zud8_%_XeJ_K%ca2*{c($f??rGO0l6+Yb0ZRM@+RC79Y`V-5HG-V$#urMQ z6WRI1S3Umw_$-r@d8;D2h7>xnlsRE0Wy&eGj3T8}%#PRZFSx`m^?LA~hcBmkohQho z-XGQ))*PK6>6LLS?0)k%^nu240gQll}d^<}U{+C0`d*5`1a1|ms+8`zxvVAP!lZT zbv=8?&S%USbJ7ed6%>1rBKxrh9qo-fZi{YJ?yaOrH2t!45lL}kU9;azfaFmuwAtBC#e5pr z2u`mKXkt9zjvfu{V2l=N9=JdU+=pC@S-b_|hg>&(r67D2g@WU-s6>85vpb{�mDY zb!?cxHmq=n2RmZJaw=X}nK1cJGuK7Z@d&}N&oRG> zDSCCbq)H0=Vhk=}ahn#;bVM8!7!L4;e@;oX;?9S)mSjqNfa1g(Vh-x1B9mXQJB1;K#uder7Xhlsf^6-hSAz%QM{5hv$qmD-sm+vMfeyfq}BUs<{nM{VlO zESZB83#QFpOsvH4ZK~q9qz~I0XOi_Iq*^!XKP}SNBa^{L;f)Zcu9$N0B`t25OlJ*s zOu4jPv%}*u#G_GT!p;ZFnb8(%qmk1M(<4@jDy6l-+BUZ~yA+t&2p$gGhGWFu;DavmNqJr{q^c$vpa#@irA&frFW^$diKFQZDEN7$#L<^S5Y(=Q{`7byh<&8`H5UKbc_DZ{?Z<-liKE9)${t6&gvn9L<>{?93td!n>N_{&c_$g@P0r02H zWZ%;*pZ)ODNRv4E>HdNYQZk7&nu(Lc6btgDf{HmyXNb@wtfUiNTOdib@f)jta3*+w z(N+rAxM2s+d>N*?V%($|{ByZL4i54|@Bh#ThqU>UHgUiEQ3duIK>wY#OVFy+h1>+| z@!Z^95y)SmSLe$TMCp<>(3b!R>({bt+%Vj~#;szke6Tzo{}U4BX#-Xyv?@7mO$VJ@)rS zmI(^ai8kU$XF!>U5)m@;1JmiP-WZ+0fbu^iQ_fZpajv$5PMoOzFvODGjSnoGS3Gi& z)NK@j^QNx8JVbm1+M?8M5Y z*i5-(Q7rJ@>!}!29_L;$I@i$mip4_6YF^}R*q)*94b7`IaWRWS=U&<3dOz%Qd2b|U zt~eS8whDwN3?T;zg5ha-X_X1PKe!?DGr{B`dorHrpRxR=ax>`cqSyk8?4V*2WIeRe z>3%f0HPjUQYf)U0q7qqzq$_RWM&%yQvr??!LH$xmcPsS%4T^pOrg@VZ>F>e!@3vdj z!mI}7(L!6?ro(Gs^UqO(Xe6p^{x;h$;QxgR zDU-DM=io+<0~Tw{x6$T#Ut}uWVkfT(zYMKg-qZfyz&vf9pbog7uxz?#He7Ov?FC{u zg9iX>SXWKgD-r}t+-wTiJZ+t6(N^dfsDzAh#2(@Yc#ooiU(C9}RV0KBN6hxEb zI>0Pigg^6T_}^7-z!+2+iYYJHW-4EXH-QdZgqrZCaH#~si}-s!tuAD&Sa~FT_#b>J zpm<53i4Q#0=PZ4?POH8+NPiAN6t_jO+T&5Gf&pC^7QVwYVGQ`iUn^a3j3iOL3qD`)@JD~mq0;S|E5!>sPZJz=gcC35P<=m-TW*<^ zz{dyU#DOA??4Y}ACWUTN>96r>V0v(tXlDqiS=kwH>tbm#j)h3uvd&Z^yqVL0s5K6*3>9Zpgm1$jqbb^s8+Ez z3*RyUE#n?o%%AcDniHqLtTAI08l>3!6zQj8z~!sD5R~b@MV%keLw5mJUEzWoOvQrF zRYl(06b;IU0nO1Z%4H$(aAp^rISgk`Em%e!U-0p_?koTrm=X6E$ZI^?nff1S8(wZ* zP)V<0>($8p(hoIn{bB745|*|5Jq$q2s#l}0xkKEk!V20ZI-q>n!jt6rx7pgC*}aP;$Yi?>vR zPHS6pD7Jh_o*YtkFcjYl%L~!UhETmGThl=2fx}4Kqb!3Yi1B4_=e^ukWP>Ayc-cr3 zzApO3FTZ1wPQh_A|BW2uC!L&F4Yr%9!SfUg{m&Yx7}O~lqMJ}-0G!`ppN9dg20Svt zwV0~cO%hc~_>N2^!|;jVzmcQ`jWm2AROsszU*1rQ3z+^zsmG8;vS(m)9}y10?(J z{_gefF9(hyoOVOtdRqe}ykTfAAx81ZHqibq^?}fmApXe?0#2;lI7s6^kZx3*0Y^>w z+)}qQK0_dV$^(~b?u48YH%AxJHBs36E+MqZGtCQA64z+9c;?QiuyrwQd$b%^jS2}I zbJ=s4TOhr;O8xRrO~Oa!^M?tf!CI8diABt(W+LV~#a^MvB`Ris*{`U5nkD?_Wu`!6i_9da3RNvA(hm%twI zlsdx|OAbImW@~cje#I3iuNrd62O)-Rh@B8yI|UV|IBREEf}l~|udrrVkHEsHAT|mS z^W-ffehDjw+z|V0)(1baY(me`y}eP~&LW#aabO}rZQ!THntF$12wV9R%!h1#(B`)} z{1f+swh~uzUjNCjN`KKi#Uyw%YvZfP*5}6RIb#}3=*~$%zqGAF%{2j_RXgE%XZA@2|P}Y2DdOZA`}j_9{q}vh)z0r z?uqbZhq8F9=NOBkPG8gM>R-Y0Cu#;^?F{RN}EffqI=`Yg)( zfu(=>$^DP>sz}1&MLq$ISv#Qcsznhu9pJLfzijUQ7vC3XQM@DxuSm5zrcJz8UN$$I z&X9qVMty01oX;g8rf^-5_md79NS|Hs6Se`rDJ%%U-E!GnNIS2FS=*V7;l;qenh_q) z6jBMYYz9*5+x=f^j68_zmheR21e)Mc_saS*lI6rYs@4p*dnvYvBD<)VR%V+9D-Cjl zsfsS@262wy4WiM}|Q-asQV1u%b zu2y7-cF{Th*>o#&5QND#|8~Jyz`nip8h<=<$PPOrn-`v`zb^RvkmbIaPwd=@oi-fX zf7lm+Hx?7udqGKHS)X?IycVcxsTA59B2QgeTQT(1iR^sAX9d5_v~=n^?LLUZfs19| zH^M-KFh$;1+>lnLMRDG@1Om&L8gY^sleji8OQFaK9eMdO)RE%>0b5tyOkUV|g{@7; zOL9#7w&dd^%XqibVuT}tjm?tU#2qvit)NR5sLCLf6Pe+`{M<)+!DW(5$4U?|;9SLm zm)f8V0T9k;4!Ac1BnV#+Hty8zcZ`7|??~yO74Qdl( z62&^NE=a_vq_xJHyf!iZ!g0Yv0oim(M2(*znKV`coJaB@mn==bB%Qu5EQvhfbrHJ$ z47sfKII21HEOy1QVv0?8^LM{o`>b=j;?_Hx*3H5UU_jV)2Dd&)Xfz&j$Xo{4D3#1_ z8i!B(q|&StWk34o;?%n;+org|w2^np=Dxcp2)4cAUODzB{{Im)49=~y{;0OZgwtv& z4(3EmAjY%;s8?tcW9sKWtYe97a$@+|kKa#za>6awd%0!fM?U)=CwA!R%{JGG6uXup z?^7|zm53Ra=s3%vPYKgyapRJ%@B{*E0nigSjl)2myrx2G@}R`0UH>&%JBdIw>j($i zR*Frf$R;Xg$?OK8g*gIAA-UjLXlAviX6$jh8>wyb0jC6${9-N?6G0ik)a&8eyR(&jkpmxd|qN`)vfww^M8eMYd8g zn1zBK`di8p4eVu_1n45x+Qn)a^vj$;wcF`-s zfqxlU?biA4@-;YIOMQ|e8|Y16jc3p^VJs}sF-SS&aM@vm#Ax8;crZL@~f`{w^>|NO=ue);zAR zsFvTsOJ+=r@y{KgOiJaU$HUWP(1q6o&TCYWEoLexfnrxvWEB;IVe2gLn3npaPu?#q z@f`>CV!1^4`RVYnc9-MFvR}Qo?pc=+>GU2!9TX7u!&Ygz;EJ1e39zgGBi8xpi#$#n zB@5i}d&Pe`ai_m!Pu}OmuHfk4Yp8z4${pmbb0Y6nzG5NFaMwoyhbzQKX?BVC>a^h;g z(`GQLrq~LK9E4JK)k9I4tlPlogoWtUeh1`Ryw=VfWZt^OF7<+@1mbGd86QYH8*yQi z=S3-A*rY6GE9ql?7a`qp_(UN`81s40y~A$v3f zbo#WzJ~&$8c75ti_u4cZdu5=FnK(V4NzZSi8OgkvE`sKT63&Qab{rerXZnFKv zlr5f>$AEt7krfn=cIa2{{%qCjCed^0-%IY3CVsY$^S%(I$4r8@Q*0|mu2L}tl5H%u zNV%&>mv>2H!5gzSTw5c+e8?8XfP0(Afe!c zCh-Dp5N4)?JyBA~M>;vNiOo@fhy@#H5n;E_JMKO7Ipt?i3f3Hb)L=Vh%kD6>I$;I0{Y&=m zP-J6+)v7u8&*3nzSUH>7_b=vU!v+A0S9gfSgDklxcM~wO~z!?Zb}NiNH+4bX`R^E*lT9f=1}Z*ieykR zNJM!N=q8(^+r;~SlKG+M}%Ub}-Crht&vWV>omKm>$pRSrH-p=gLiLkXB1BOS2+60UVffaNe&#k_vrB zYMDFkdi7Oif7GhrQz6=YlC#32!NpM-k>gV?KN^5MUPi|q#obrmvNTQkY+Iaocg$h5 zqB{hs9`UCZEBDHSSA-kcw_y_tJKIb1p@T$$q>to;r^_lNdCdE$C8NP#9?QSM1)Q-PQgKee0)`2?dnpWrhh>F9z1+o5aH<6w>JNYKl=QU zHQ68T_xmo9O->vPf>sVAa^Z6+79!7?z@Mb9ou*@Md0vFP_l0`(=V6c%-=t0nUL(3I zei+aoJ_{bzA(z6)0bzpds;oi0SJe+;g(meGpS{v8uKE7Ul?|bev?^>_drM>D(PwLm zW~VWJmg>GU|8n_RPf0*l@)pIupi49A=@uYH+Ar)^?DRrH{0g^3(FJ?9C9`FXYzKxd zv)S(&e!%!?^@blm>j{tO(n_b(L$pwyPb#H-Oy0DXe?Xb8ZF<7tW`Y$Y@q)yJf4!i+ zHkNuUC*IX^)MModIyHcrfd0sjX5J&s(Iu0U{j(*-AE!~dpNJlw_34Jc0$t` z)D+#MObG4w%oN`8I~VOp&0}0h+r`Wyt%CRaGYa|)z7UKm&E?fP#haZGIidAk7&!8P8 zOdO7so(G@3?rt(s%f9j5vt+e3igG93zUG;MA&p`;Q}C>T3b36y5TcW4k4E;8rqGS^ zwCS=c6WO))P|#^PW(2I6jI=gu+#D|@4DU974A@sBaLvZ8evYR7+y5v`=#bu#zDW}K zIZ&Ngi0m@MO$Non9ygVWsiUz!#VzQut5?T+_A$FuXFcPBj|QEcvL~n{5_h_oUy>k5 z_0s#V4bO*+>kpHjv@QlzmTq|#f@i*mjKDDIhblR#t4&KmnuVB_3wm{V} z>%Y~s0Qy?ERTx9Ey~4eIcc;Zl`U3kQ#csJ0I+!|gl}9WH{{PZmXwSb*RynbkEXPc3Y^GSKvt3WcbSVdbGY<(D zwI@`U1=vz)yGN1WhBsywB7O6?Rd<`-d|uFSptP`@nin)Co?7{rnL-mZTmoHUNt_cy z10t{^9H=^qT~Cp9RLn^>ZJu6e;1AEGH+nQcoMgylv*LCbFt3JoepwEm<+HQLXWMaM zk8iVm4qmvJkQLPy@mmuvf`9kI3*_c=W0x*B>p9s+vAq=OreZQgh2bbiwAFNBct0r& z$3{rG^bu7B?(*-ImxEWufGKqBW@!wtMDEtdd5-079b)YOxmsQJz}I)J}ncEJ(1%8EX6<`e1BXuTE7$Pp_@soj@3S1*mF4 zi85HOf-LC(K4hdd@);ZZ78(#0D_JSI4WuehwY>gO;Ch-g^l0IwU;2S0?{?|Yv%(=@S{`?FrHoG#nX z=+(`@7vC9{A*d8Ske?zQ!1j7VardRX;9K6!(YJ&fr&dQDru)eepQT_T-%_^29har8 zWa$gd(bwfWXyY7t>J+k@M!xz4**Rr%bhf5bd`@{mwQ)vHL@piA*bgxL@bFZ-_G!y( z^Lbtpk9xE7+<;{*hO-JmNM#ziJKsRDi4<8&#q@_I1Y<+gB0&qA=-MJ}gD&gYn)^@| zXus9A_3_(?3R|bM+jYE9G4X%Thn=5pvSM9in*AjExv^pon_*}l#X=-{Hx<*UtkxX$ z#Jz2r+oDEgCFDNAc9=P#!CJ3Q)iwp}kCXkeV80?FDKakruF_4`E|q-lru|r2<(&ZC zwxJ5NUY;OIz#bdwG&jhs)iF)#b~QrsQ>a`{vtNwTMwC5O=nP+AFPQz0`>Ms3Db^gh z$yhyG8QSc#Tv_a$BzhqK|JZvMxF*vxZG2ww4atKcCxXd4fFcnL;K(Ud(Be#Id)OYg z+uiNAZMWTSrQObMXWH$TwcF`TQSl523La1cDiAr#!LytkJdX+riUJ;x2mlzHh=|llwQl=ptOdau#4`|GyC&4S zX=#+0zM$wQQ}Kni>EYVz2~IE$tFN&0I(*lDBKr8NM%(hU`lG*sffZ~Bhk%Hv$-68PVtk@bQYLKSobv1>=b4_l7~})9 zg29wRsJ>KZ(fj=kxlC05%2w?V6cR1xoNBx4@mcj?bC5#nu2(VlW8sUdu=^c9OQX2A z@Ui=dtt4|ceJ4h(i&Z}-0R9B0{^RY6#Xm5D>GvJKzehg)(tzoK37A?a<~l{LQn7|$ z*@?NW&|zFL`)IhH+5nk{b>I`916Rz4AY{Bt~v!uurNA&+%qREbB5g`b1$Ctc&}LCV!YcO zcI*WG$N#G|nw_AGzow8JC(dcsni!IU6jMf#{RX1^Q_3WfzB#NxQO~(4?%=9RgsXxM zMt5-Q_=jMP!B~D4K~loeG?n@xZ*xcs!Gu&Ly*dOWkuQ?d98479f2hqnAZgPd!}O6I z-8B|ac}fwSKPID04t^qqE6u|<(3;I!yy zy-QOs%~BzA)1Z49L5l&rOUDCoq0uLokeo@MT0$2H1uX3di1DV-c94uWn*ENk14(bY z{DS$aJid4GANP~v?3`UEuH?O8Vro977@)O0h0bmzeJ^Mg?+9?h zRT13BtmmA=x&^c)`qUP-1>EB<7pQx`c6MQfTY_6oKrZJLxfgVtUkH+6>5{X8V^Not z!1XofsH!yBC~WoIQ==E_R&VF%X6Fk^y+#+Xi^|K&=brb1)W}2E34Ds35H-g2uO8Ml z8&@=T;Cp$kvhE0bpeBCu^@dkH^cxqP>ZeaKV~`Ds5OB+L#cNoW5;gU~U#sVt{Zvir z4pv5H+NPYildrmm5*M5fLD()jPME+w?7N8vMNr7Ky@_|&w^`C5OLo;4AE0i7L5&j0 z_(xo7!)n7Ez9v}>60>jpq!3`m&Fnz(H^p7;!N!%S^ZO4=Nv;#Su@HnCQ`32bVh&QI zjEc?W9G;UNaGJg@?BQ({)pDC+4n{pz>=r$msa4|9R8^L^SbwIM+rwKO(Gio&xh|{- z!E4ueUxZ4yijdV2O)=$>p9<@Q!!Ebx*N5*Gou-RU;R`~6)YxAP5*fE2W z;jI;zu=WyG*r9Ix_=nemjM(|}`|ZcczUi>kkDa2!6az#CrI?gyb*yH{dELZ6uDt#CVMgf4{ztuw6g#om zxnP3nDvCKykt0;>4!<6{^%pw+w&(%qf-dH+pOHbg1P-`&s8&ZTU0O{)dOe4`I^r^& zO*e!Ua90NRNp`y*g*MphUW1-ZF*Uqg&SePU4!Ub+54!h60hx#Drm88X)C<`=T2x6P z%Xz!G#{*J$iGC}CQ$-0;L%ejm5YO2k+JNtgQESy%R?w#IeQ#IBtM2pa=sN)gH0F{r zMd}jH`i1%=z$sUqq=TL`vAb2z{KZAqqvFLqQ}K!wDqagaee1WsGTN0ww}^SBQEWY8V-0aEUpqq<746d#WmM1`_x6wy`#VsgYs-xzU0 zLNFZ!Y%eIPC3<>~%rGsTCrH339-e8IEQNy85tpUQ1v>sxtY}#JXvwZ`J^tmxCDm`H z{PR-v(tf%$*l=Rcua1B3;1W25*J|k$jyjjKhpzEGO%HQ#`Yy#^52>?pV5E}4c87t; zOQH=PtHVB4q|e0T>W?{hLi0rUvQC<nGT{16_HBJn2=W2MHC(08zqK^LaKWUs7tiG4y-0!GEte(E}CTve+eZ6&t(O`7`d+P^e z9lPMK^A-dV3#9vP{Jx0O^pN-c1scWWka%7;u->-O?NHc|2^tPMeuoT?Iy_=C2A9qX z@psd+mztZ5FG-n#1q(+Dyh`DTK$O`{_1mdzO)t=FW>3d zBg@|S=o>Pl$;g=XpYM@0b|%AljUmME#s-YCDF$$ofs8@Owr=bvJST};q?ZTMYpr6v zsX?YC*Z)xfw)AI;FiZUmIQoN{<%90HW;!-$U@cC@bx|D%74^^kHRjCOuaq7`fe9Kk z6tjaO+o;%0ZmHajepsrY#%urz0`)>~sk$_v!+RaR_*A$-o)dORs=i3d6$89DAG=np zSeh~0U|?x9`yIFJC_Vq{m2>9ni_;b)SXeQz162p*kJ;csJfrsqAsu$yqIUV9XE%Q; zeF*)0SqlOCO|Eq#S%$!Pyg)yDjp zfa2IRIiw7tbRac$D!fUiFG4`_{m%>JR(TI{R|tTF0VNt&2y(?EE<@fu^l)IDk4CXx z0D^X)>&!)lvUYi0h;EcVAy<{cHN;>H1D8cMOB#5GB%pNnj7Db{1G<&UbVf6;6M`a@ zz~O-1_|Is=?}ERXdN2J}CM<#&y?=kPB)q~xI!$IsT8@WDKzm;s7Bp<@3$>~hzuTzn)tn=a;NLg{Q7{R9g2 zb#w}^9o}J0!u!K67ht2)Lsxml`)d@d=AD@P9D+S~fx0+4YjK_Iu{c|$qw7GG`M7@# z6gBM+P`7)n2uO*+){#OUuisH@pIvuX~oebjI|LOVI@q_sTtpvPyDY*0kQ z=yL%SgQ9V(;h-zNNm%KZ#L1yFib3Dw5x_sd9(^_%8J4%PLtkb2Ti<@uh`xwrSK3I0 z6Qi%$WG%ZuG3O|9nu;yvwhAkScj*q$1}}>=5(i~+9zmx)&NTKuRs|jw<9qdOaj)z# z9Una6f>VozYkydwht@Z3#*M|*ftk?VR2fn@uRsDVky>a0K^+kQ4}y-~7lu8432tlM zumq=oQ$QV_bA~SEJdz?G%pqwfARRx)M*BAvc(TR9vBJ}9bN_Kk^?W0q-n^5WK=Rmm zQcfJ)J7a>rqZD(9BIQ)<3MO@7CS&-wA*L~=DdwrqCUS3KnRgPWDW(_pZ9VjfsEQDf zpr_Z9Bu=+i9=F2HAWeZSp?G|i04SiDPn0RVO9HJM7#v;>_pa~&L3lOAlt4;+m(q@| zy?vo(3#i7w=X?J;OZ>VKJpR9{SW31{Cs5)vrsioM#Xt&Y4|K{$-g95K=&WmDIQXY~ z801RDEL++PSZS7Ze)G0hJ}fe*NUmpAUU0>4#3cvVnxLVX>U_IFdLnRr)PB`k`}Hxu z^aoAy8NW21y$-PZIy)_mOz}@D-+0xDejkOlhUm+`vGH{StRVUZS(K%~H1{sAc21t} zO)rgNS5yuV?Ah`AKM|JL<94Fa?01e8Zl|p*ytCJwNS;j|!HGBY7AWTTg@KlQnmnH_ z0S_j@TYX19K&k`Rgw=Z{GO2zi1M|h|o8pV&vPdmGWx80-ilxp!{{D!$lj5{natm(9 za_X+U5wtbi_|M$0!A5G=%$=S&kXL*n>yRM>MNSy>a*C5U$*)(;$&czK2S|xm)r`ZO zdkaB}lbh_BHM2{4@%4j}jUj!Ytk(e=oyTWac~!U_pM9Fs8#?HY$7;FxQFb&t?BOuc z*606`X`yo8srn{mbW(;qhAxs}7k0MEY1dR}i5%-|B~i=42l!O2HoGOn4Qc! zP4D-D&GSW{-Q3oYY~DL@f(}*Hocyq|$nA4GRO$lm*@bCx*mrYVI1QdDBH$mgqmB7| z-mE~&MApS{?(@;?_u~8S^ndSPM*}X)b@NH969-(LnHYpFiUBH*4ye!sV&o%q1sAk7 zmkSycmxVynTn&Yi&n0CtAa_PS+%|rxSAzn%nQzIDDGiLvP$vmNwRBE#bdk?3@p^JQ z{28f}F85Wh5mb?Vv`&JGNRY!`xhHwUhyHy^)YYz#;0S_ZDgt zxBLzPXS9Jn?}mrQE5pBsjJTvkKBFt)cd3Hl^eu6wXjqoU1wN|+PJ^^J6r+ae9K&eU z^coz2iqlG>N=__Zq2V z7v6K?ebplqoZY6FR*Kxf&MhQ`&~rk|fJQ{p7_CoZdL#?f`#4FUsnHjKB+<*DS{*vH z@!B+b4@sC&!%JF}Fyo{I`1UH5dl*Ah8uFaJ2{Y6?9(N6-47#NK@|36ww{%!MFN!aR9P|GawsXk$`V0&kEMM}$Mb%)AY&SC#JnR>z zY=^^(F&y#AvMt(O{HuHBb;Pd}P%KmnpBFZI7m{{@3}u7vyS(BAmqQ-D-YD&+&y$UQ zZTzi%+q|z%34i41e6qsE=uK1b(=s-`QCs!(WyW>t#y_lzAbsrCDJM2Bn@m#h@za>) z)5tJx($nN^{Lj2A+z-ksz>h{&t3mgM@Z-YX$fg+V)!awt1B1;GI!@3q3-#poD}iJ< zx|umbt&3V42q`IHpQLv75vos8=s7gMPjZcu6xbJe1$sN%Nx!gN7AMf_s)y7GQ)agD zGenuZqYLqxUM|v8V^I&9s~%~F2v)YJPldy>EEQ^7-t|nJiIN+S=_K&F@UsBdi_R0g zw<#t?6d#PoU=5oUFWv1`Cp}Ha1@G{yn4@*;h{4)6!>5<%N5;>@Zc~kFkb6Xx;9X5? zRU_d}n~T3#wOR1TQFlr5(!F$CmBu`Mu?SPu+UBGvf9)QC_{!cvQ|CF3j9Pm$*;~3?>dd~}1 zR>-E5!*7KjhW+ws5Y2k{I;Qqv0}EIcq9$*6cO zP`FZ~*ciQ^qu!|)1WlzJ)$Z9@BD=dTtQMpRJW+eEQI;2^@v2UpHV~|Ov~<0B_RMMX zBNp;u2wi=_JR%oc%rP^z*W(a04Oh%oW0DmQBysN9P9@A5J*;8N`V(xsffa10h1_2G zJujn0Y53*AcgS`pj$lHK&Y0*?F2#UI%U&w>V_>g-$lKtCT{G?SPw9tH8`dgZ3nVzt zIQfFO$ZgzG<{5o>QH^g?OykVm`ghlq8~yU;VeQCmxt61u-|E2Xk=NL?0KaVDY=ePS zcd^0>^^V)yr>v~#X_jP&vcAE67Hyan9cGq%Y+mdCU{n#Mi=a6gyJ{s2FUh{de>iM$2h}DcZZztTze^LJ3?~P_< z?Yg9B(#6hKcHX~?Y%uYahbU%{B2THM z;1Lu#X9Q>z_0j>!o!Q`XZt{38HqgB3d5WQ;8kG)*LQ9O?#Z8cuOD20E=TgDMhoKh- zwXBAk=T0C$z8k883g`H+3Z{Z76P9qXGb0P)k?_GD8$e1C2<0kfW6|7bWD^uJF%tPP zkP_lxx5elySGXZvOGZEm$H<@FD}PL{ARB>c83#3Rs8`YkQo4(-O$Jesc6qPdX&vQN z-~GG%SDgvFPPQ_j8qly7wk(Jdz*3X<<46>=JRs|adS;xv#IQ^#CquUgiSrFL?)yA2 zdVu^3cSBFp`GT7em_&_3^+()ndg(Urr#{cz*G1zwd;0Q|wy0w*EPKOg@eWoN^|dD( z&;8e{&NJB!&0D=tKHtwt^~)C@1#uuSM*4^h8mDH-<=F!gjbCCIChVd2*b{@)c$GcT zU|~9J1Br#f?0A?RNCIZPB{OHqd8LBY7MODyW*z19(A$Hq3OA9{A}m1b=NR6d1=YH< zqWr0V$e3AWftE3MIsT<(wEQUIvfmt|-xL1_p9GS@&hK&JP<4gL=Bt=upuIg8$bKa2 zXYv}ow+9w_mWT%F zOmQ_`28C#b=yo;4fd>~2h;bl?*dZ2J8oLmgymkb#{vuMenUhDpY)hOtc5H!A5{Ta* z?Oh37Sv{ng**v4wO}#>ZO3L7smM#Ix1OEDj81&sCfZ*#R=?9g3bZhla>m87Z$NgY> zz+Wo)81aRZ_#WMALk}0+f?;NC?HhYhCyZvjy3hEtX%C#3fM)EQ@jue4qn2&X&;Rx- zKc7At6xRJgO_sB>IZnLS$}q7xDHM}LkquO=QRJfoYQqv4C5(l-`E$zh4tMRBVz)h|y}XD=_NBld$iKqxscZ6f%7Hp9-pd$uX9_%Ir z;=#S*{(!`JdOq4LRY&;I@DZ1;fMh{sNSeG!QRmvr#YRJDxbSLqLz2(WWaX|2&`dt! zf*Ybu!sRg^&sJA*!M;HC7Kk?MKWq|oOx9sz?2^ImH5;6cz1H&TGG4ChU%bEdBNro* z3O@Vo`y_5U(U?qFGQ}iPWIYvI;)Tgt(Br}gMjQOQ!DF22+_ArkgI-(pSxJ$&myE|mV>sem^&`zC^ZHKbnr8` z%6`1`yzzh|p6iP-S=kTg4KMunS90Dl&qK4ZAI|%i5er%CPVq&b)P<`f`rP*`_ll4C z=W+(QD;FK}JkOx6dyeY2erdHcBb#^O`oQ}ZA4AsT+U-s%`Qyh#38twCK#)u z7>LQ8!_Cs5M`l=;{Gq%v{KjJ3A9aSu2@XJ&1@^U|EUF=`sH3r|rCy3!3|hKWl*Eab z?r=kfCEQw-gL(sUlwyvtS&}A4#z1_3gXxtY^nMhY%&VB44WYZ^fpLqvpnx)$)5fot zHYgGqeV+}DRURn&ecTNhUEZ;!qQg!A{2YnnzhIBZ?hn58+gKx{o`(F7jI=qiA&N7p zTj-&fPbu;V6^kU1&61nqlRo-Br8a(7=#cj+?*iB;r;rL@;;AA!Sw%=Lw@12H{E6)T z0z+43W*8=W@stsy^fNdLpd(2tje>)MV)&y>m;tSOHXy72B7Zx>O*4*z%(eauL`j$AEw^q=`fAJ|JE+296ld$5?fFA5O)&_mVRpX0RC>Ta+AS`4^;&cfy z28^Y@AASBS;xG0XKHO|=oEjL74=&TArch&L$6gCvqDnKD;kBT4Tl=+J;&U?iU$r49 zbxxYR^H=XSK^6YUTerk%@>M^2*X;S2{#JH>$KvtZ{j(<;kk#XNUO!W}yCvG3M2U?D z?ZoR_p2_;QlVY~xG{o*#?wE;vRq2ufF4~n;KlN5oo+#B1s+^EleJFI7G6N(Nv7_?C zjdEL7o{4ZX{y7#X|J){wKazcqeuF#Bdory1&I^1?tov>JwSifSPl)UI8M8+~oUSP* zHG0r}80fZ3xVnI*n43}e=cCgM#ZkPCeqb0m$Z${w?YY?R;%M7w3r_Tau`euM8ZCD4 zq?R%IH15y(Z;-7{?9+fA=a_Jo=y;&UrtnQPgmH)8VDwR=3UE+3% z$^}R%jX}>riUIkK{Zwojz237TWW8q_R|iV#>g^1uxP~22>iPVz2phren2 zrg?_(OXAF0NR;YmY?Ddg=K1UO652U~wsKEf{rJbf&?>Nuf2%`}hWt(9AIvMW*+9aH z9asnnkFh6vCw&)cWS=eED@%}|SK`wYb3I%q z>3(bJaoEyldi8mK8oq%fJJ@@{S+)u!V^$U8ONRmk2k-pBli+_gE-hO=_{Q7h>MORi z^qD~69>v_H$ZaaNgHB|yg(HP@L17a{yH5GW3918%gpejxU_iH7Qbw?W?qlw9*a#o= zZsW&GH_Nf*D9t%fdRGy{C_ItL=`(~Hs8~Jr#WRs#@eqC~%X%$ZB@Yo2amjPzOk%=&Ad4Y7d@+v0W7LT(rIxT(*{S~#nH z8bc5uYMk+YkOczz0F3e0&;v*zm?qyo_X+qtDKpD`5kqP66bSd$g`~hvOO06qBqvax zraw^*g8;`VUInkhE!A&8qCTNK&L8AX2p97yn@aeYfA&Q0EyHJ=s;X(V%->edHQKKJ zpGjAfJx;vyt~Rk#r4&<4kpe3AxT4$_o!~p7(wQlwR@QM^j3MBc;#F=_H<#Gmy7Ia$L_=9xT;u=*psD|rckOGPoCMKE}{W&{= z<_kZI%>}GL^V;uyR)&~2oH(r=vd}_O6j>U*GI%Rkqy1jz3FdvW022B1a0K)?!agba zbv3q#XRFc!j`Oj^$`I5=rJs7Kq3pAOdZ$-b6ICR;?1`hnTl<2=j)-A{1?#S5g_vnK zzw)C4=9PL*ix>;Fceu_B$XkW!Gxw-A`hDiDJ_VXm8FWrq+6;S}i7%c4s}0E)4`tUY z*uf(IR?hlRW=yXYSK%OXZW*2zL37%cj zguo)9`u+lFf)9n>qa%tUVM=5jT|O6!8`|hHnNIQ~bT^&NO`fMtj6#R!JXzDRhcAhX74yHV_hL&>6uQpgd&h1 zfKI=)=?KYkVgQ{m0Z=)`lu)Dys@1ux-C-jHUhy7DwJgtH2So*!yw&I^W4t%dzXXJN zYkZ-RUZZ&6)hpZR*A=cOIYGMYM1smL+MonaybXyN>gZk>T+l{md-Oz2wlVFc`Lri& zUK)H72iU>p+i(AAzIi!=(=JCA>MYL5D|p2oJ3NehH<)ieq0A7WvRngHKRLn|nP7fc z;9`PN?0T=~XvjHjCwE~74Y40tmOI{ogy|Kt%jb5{S{ma!Y4Y8wk%&{OY8oY^ zVXuu{v%T^}CJ)%8n;2|pLhUq-q+<3?*PWi4nfGOrU&2^Gj?Ef2X$)pB{{3Tf33;bA zE*9kCG2{n~K_H~O+PxE$70#1F2*I=pk4E8pExj#3oyy(A$@Wq=DKN8W7nkIVpS&$o z{>7tC{t9+hCglf9l7D5iGKFpt^T=%{c3Or_>_#`mbW!91==ClGy-6tit`9+$CuE%h z7qwlXe*?Wp?=S0;4$4>YpgO8uo*7mH`ae78X1Z1^eioWF7sNC)iZ(q}Ju1S z;0;M3seT$&7aa$V@L!f4;+6|DM8LH;=&?EE=9_qdhTL<9GAnJ+j;~`mFw#k-`e`9h zqrMe#LUfk&h_h4}kp4jb`48M1qm4Xp`04HPD_+QakNNa13m@P>XuQzV29uTGdGk#z z>4~@uRXcg2JV6)z;j-%zrzM+W1-WU;k++Iw8X?#G#A!igB z5C)$7YV!p~!WPSftI`(?JL%J`U_#Zd{A&^42$MIjZT*O>pHB9f%tkuJY^6vl6$@mG zuo)e3fiEmP-)&Uqk`icV{>U5n<*EWfz(T5}+ZDTJPtn4le@-mtVA9Xl+5uS2f%8_N z{CU~xLh}k5Ho0);{mWn%J~kKrd2LM_zh&0)S*X~P1F6nWg&X8$A^DKN#Prl?rT|^! zJMvr*+%ObpIo#49>R2LO@BUhg*2!GAyYcLUD5L$@s+DDugHF86G?+~KDT?`sBGp(> zg*A5Iqg)NqNtz`X>Q|%3fNJkL8TdKaMUp|E4ruo(p`V1-$>7ohB$aCr`N4Z@IGJwC z!|J6-g1AjG^e>%22cJ@q?Y949FA#jYn{vxn2W-H*<#n)4*o zb`;qH4HnxYo6*6}2yVX_Cs||!lK;vL`Q)$@1L>*>kj_xdNs64HV#`HZdig9UQizCy zuBF57O|t!Nz!cI0GEaG;tx@3kwsH5UGU%;Q&7cl&)o1_fphUoF5w-|X@;#ZidrlkQ zKs<62@{!r9T>mzYdanAEuMYUBF4LJj!~PtbcDv{#(OoZfE+>!mH^JUIfd)tHj3gtP*derlnw z*YHYhfKa_OOO-3e#R(}RMmIbiv?Bt9_CbvG2(|V~c8F>9{rd{@%@P|&$BB#nENrdz zc^ru@_pe;C?C_g6#T~L7E-L;cajuf>Uu#;D#4)_?7d{skzq#zMNt{28Ueplt02;;{ zV{BipteC;cfXa&DZNB8siEoP+7}uyjRK(Pf0w-Rh&Y7%H$0^WkA_u8hjd!EBMu9pO z$-G|Z&qOC@E!E%w+NY{auNqiM)>60*L4Hb>F!LOcRNWG1D6wS;-=jWihO(HO$tjC` z?v89#mEh0hLzSp{xnQ{uUGun=#eyep}5{NYw_rb zRI-Iz7Bgg2@7m&C?ntT*{U2}hWSBfU`@dl|8P035?>x=?g?W9Rg)+Ef%4%6lU=8OP z>6+gSn~=+%mzB%IHcGJ}ZIf$1$8Z!C*g8VXyqaS21qn0vg$>Kj$c9}~M2E!{;lnN+ z;YG6NU#XZ~5njpfb3Z_?_+|t+z_{BC8k4W@{|_1qPhb;qcQ{7b8U~w+z6e;H6D+d6>^bfKX zCZ?m<|J|3ww*0f`;oITH6-em1WG~szF5B(AR+C&bSxBoX2C6`hQn8t?_4Edhi>eOQ z>WJ;GDEPf*y5SfKeCMjtXX3jd-ae4|?%}lro>cBv?x8ORFXs$-cc_|T_Hyqn{L8ZU z4gN9cVZM1w{NY{s(Iro0=Tz+x9WiY*j``wV9IZ&UUpeV-iml%S6MH$%o5_AR^3ML` zWAlW8)8@J@WHPcAZ}DI2Ul4hU>~y^@%;i)-U&02zTHs&Z&Z&}g&BwEcgHw2uo( zHoNJze%#4k$xJqaq2pIh{N=0mLJUc}S5?Dt2&D zDp#)>rSZ51LR7ipof7B|!hkuJL1KWjNi`gyt_aDXYhPb0D1%bTWN>W)L_5E+jO+T5U*9ng$Gi*`Cw6%)#Ey!&`)K6)jF;Ah=;(ejEE}Bp z5fIT7&=_0 zC=u66>ZMhr-cy|eV?*;dzc3s}(SRO!HZ+N+LGr5-*3yrKYgy^%+K4vW*4mdrW%05k(AQ$-_)0WdFvkLOSd`vfaS05c-_xV ztDk;K^4>jjW^xOZ-m7?6M7t$&KL;xd1_-#bl{6OgXSnU2)62z*r3$z5xld*4{sp_| zppYk4&Nj*#Wf10_4fWd^g`vvDVJNZ=E0(`!jsguUtT^wk>fDdN@naw3nxtDCdW-Cs zP9VECCXtj+F%bL624n7-2)Z+RS=M#(9dhJeE+)G;D+HIYg!&;b8(aiKsmjqe?&-Tc zu)biod$0T)^m;&j$p9I2ui-U$$4eb%bta1zN7dA_LW}dR^WXXGTS1el;&k4>%u*F6 zN=TpbP2vIbN3&$ryzPNTxh9l_zYsnkx#Dp~p*B#H_eAL=8bzK!=ji>B-lW-xbVr%P z&&Pd>pY%`WO|@T=`WFi=xJiprL?~q4=C)h3-n%IVGh+KZ&dAOvGI@F251&E?b%$S$ zIEm9OD^u17dqKUSeK%$n(i;$rSa)k)Rb=}^T*IB`ra*91K~C}tZ4vk<$2=?ba~DT-``rDkcIAX#u31bA9h7Zj&S z*$g`oqDcKgimUGzQ^;@I!#Tp%>H#+#IWpsGF_5B~4i&(R9Q=E54M?gwv%=WE{ zVvbX=rHe(f#5DPh#pRq79!OhbS}8NE&b5oiw%cx3G!<&B)RGy`6c z%OUA=H44-(ZKF>>Q};1t%lvGWUb7S|4*w2G_jo4*Y|$QQ@^^5T2X#;>Q|W z6Iq-w=Ozoi-}znH6lkdy3`iQ5P_fs66sSXXBpN-`9(uVT z6XK=Sl3^f0)bX!z;{9*)p=^cpb82}Sgk3?$}L!nPXGs83r`#c!df5jI+N!CN!1yxb10BWQY*4Nu19ls}^lR|11cY$xw z#AK;11b5N(QsmklbYH))hhOE@MceEz+i%M3JCUrW%z49HssB)}{of@Xip6x5!bYZ^ zucLbx0k~D8Y%{>Lh!y4MB=vmt8Cid1HEf}6#x1*Lfj!FEjc@DKAB2=@XBbUQ=fAgp zK-M{NbWUqx1GiHQtfrf(SVKPOFyw)N`{i;pnzS}(&8>62>bh5ahuZ{Uy6!2dHTZlI z*vkU;JDaxQ@2W%0TNf?#{FQ;Am6n60OHDCxf>oX%ITKPSt_VrupsdwiQ07F1 z&Zd|X(9GASf2~pA`5bPy*LGDh@6wVz^c@a%#vYKJ3`FJe6UryD9bx~FC@uofitbTFTFl6&^?E!oY~_$!S0QF}}ytC@1%K<7OF=tT1X9Gg9o^RBCwn(UBv zRpVD0^x?9FK(YU00!MmI!vxSTWZufzG%rdIc1GqMB0g=%sOjE@517p16t(da$d%XFGk? z;l$j+8Q-)0Fc#PU)$2E|5k|k{->xhyBqgs{Rl+3`G}KT`B}I-wc9~vJy1lTCvD@nm z{g?-NM0%B=Ts9!C<6al`&_!?G5~oYbpaM?AtyK2S9d;>G9#oznHN2-$HJomr`wNDF zzpk4um(_ZstTk-=f_L~;@pgGONPD^EzC+v^-vQDt@1mRKSXHG_Y^z5CsIxgg}1t%{4# z@vqM;3^!104)KzBX*1LxV5{#3m7on-?%Ow4jr|Pma|b<-(tVL&t>U0e50Z?ux))*v zSQP>CII4n39h6};DAbRn2-pr9mEDSE7vBK>@XC;*&}jtD8(k|K1mX2cZZ8)rwNRO0 z#3hMzhO0X{+8|^W$3Suc_=W?L9X=J|X*2YZV}~Hj67sA)4mQg&!p6y3zjiLI?6IE_ zbp<0oK1edZG+6{NVPjUf{S;G3kvwFn#xxXW1KK&LXwVJ~vuey5;EQHSlR{7JY6ur; z6uV}&FB);dYKcq^$_Q#zRRr1|kbQB_?5@QtT%XFaRBe1mv_9AgU@Vp~>n?p%`$4C9 zR?-5GM!jb;PY2}r*jaL(Tn#~G@^(3@Td7wDT^C}jV2M|`?^NcmSw2j~DZCuPQ;ZMC0c@de;6CDkJtFX}?<=!yD4RtNxQEzZtI}i3<2F09q zP^Q%=4v=a-4ucnszE|(_xC~#Gsz3~LYZQHveeUW?ItjR>ABJjaR2@X-U#lG_mR|yE zS09VlS^NF4{3WWkP8WF$Kl#WOdffCp3Y5jO|c z74;}j$e;%{BIShTfW9BJj(9zMi6q3@2WtIwHbcVtkFe=V%a9n)o%yRQ+1I?4%>tAE z8R*?XW^8qjI4$g;H!eyVMdl2BQ9gE47lH=0mi`pny%86627G(5M|Qcu>hG9Me=Xi* z<4oB6dDh;@Y9^ewT%i5y)4y#VC;pNI`7DIHw<#-?qwqbX?9uZjBzS%-JfUn*=*a`J zNE;0_NKriy)oBeuz0w1+ZhDBR0{Z!E)iWp|ZiDSq7yXRB8W!i5O&4%`=u)rMAkeN+ zRES}l*GBL2Ky3u{s_`&(klW#lV&gZ#;xc47VJ+S*U+1G%E-?@HIPGF&A+)oP#^4UL zFM8#1E8Ob%<+5SIrk^G2$=c}+F=t6N-5AqN?}vOxF}Iq2$UEhU z>{es%#(KVC`A%0|6pnyqvq4(S9dKXHspIe0AA>lvXCK!Nps>|GPK*l6lWp}GR`!wd zKfQF0%;+d*%=*vwNSYIedn^QtPt0u<-t|fXYUwUvS!A+nvk*4MM?|{W`{+@Msr_yp zDh$gsG4B|JD^ffv{S5sllc>V5YEC9$J**nv){9u- z**s|}CW|7wso00XK%L3&;cXABRkix85hOri)txzA-ks9Z&{~@n-Y41O-zRwvt#3&p z3@&RFn>?OFaaJ3R*LOx-BKEKwmgnnBffOp+=lru*y*L8B4V>>5&h9xq^iBUvuq~HE z9=?urHQoHZ9*xpZVcl{p8ccD*3LB=kI`kbYD?+8e{gu5TMk}%|t!FAt2;WT+=2x_MH&IPUXV=72w%1>j(AVzQ1n3QN0V*W{uq&N5G=Zf>7 z6;53&N)&;Tp-7E-pl$T#sLddmuBESdTnbcc7L^8r#JYP~@Hr^a>5`$5>gFX~7;@yUl}gY_`(QTM0xggz%pKbkAEimbCHz za#y|_@|^o<%yt3`;iBE54*pWZ+2>z-`pu6GDSz|J#;_4k~}3n0p$)}xU2+~ ziD59laVV%1zx__*yGG-2_!k~h^6*Qu_^dO@gFmB~0g61qY=K5q>IG#pHQ`xMyp9!u z>6}&b;{6A?8YYp+3_BWt%?Jt58`lLY*_<|hYlw!?`nGvgxFta^Y&Jcjyb98iuxSRa zYc<|q;kGNP5K1KqxFar@H$zLdbLMkVS{QuHY1e1Kx4Fq}ziQYeSEV!X)l|5l3{wU~ zZ4O2cNUG*k3Q^RlfrH~*H_0X^2GD*J0PUd|V71x_Lg}F8x13t`;iyWVLB22bUth@h?SFGwIq3Q|$BbUm z5VPqmhYR^l1|;@C!USWm0tr>Mg8Rc5BS>^%v(m{ic59sT=1a(R6G)w>m@^bPNyQqH z)gAm&;0#6EbVyPQbyMk_8qYFVJ9f=(k4WZ~&&`L{tkS5GsICQ-++<$9vL$jL@_-6* zpTPx0jW=r)*Eom#%jpMRTKcwsBCKvXko7*}-LA+Lob6l$*~a_NKW|r;zYkXiH#8fM;Itm%QXSOCODPSaQnMoEK=aaoAQGun0X;m0M!c_ zbVjg|BkiKk7KrcEg+Tjlp5S&^3qeU+XyFE_>`th+DgZ|7Y*iD`)81cjnucn;$`GWP z-Qnh#&TX8J$pT`6Ii1ug%f@I|%_j$4jE1M+v){f?;>IhuabokLF@Z`l#UxT>Jv95P z_6Jl#N9Pq!H5Q4rf@-y$yb_iZVHG4;j%km}EkmOA+CSYjFVJJNEq7wDSg12x#WPeI zs&|u3JZx1kIJdB~D&eeOSkGysZ$LY61*}gtE)?u1%qrN7Su5Cof^B#FqWcr-RnI4- zg|!H?!qqjt+XG>KCHh*t8qapvQ3K8Wn>*ix5O1nq7gT-^y5ADR-gTE(S$I+)_Jdc_ zr@yX0-z))1F-L)c?c9%nf#IcXQ~QN?SvDEt)KX1*>)SoM%&Q#Om=PzA?pdgbr~>73 zY+-1MDVwv!r(E1BtXN!4R|;>yzXg%kl$SkCj~{@ro|e7@U8Fkx1O9f-HBJvb5E(aX zh}U5@!Wtk(D~OEeHOrrp)3jlf+jO5dj&2VWP8xA9K@bjnA)dbYg5MPU=Rpsnck*wo zWGP8uw=HsF3>BOBExRcO9GM+dEX3#t#Ogfrm0h7d5nc3S?^WIz!EplU-{))T)gBwd zP6lcvI@lJileT%BlUMMtvc?`5_@ddc2_s)Ls2$G#vG*(G<~5s6TTNx5nDk&&Z{#)j zW?y`xTZY6jpUJvlXS;L-$%5?QrpTpOHr^Do(HjUhi7o&YVV7?cbdKJPY7x_!R; z6GDU4QbhfFb>kj$)@L?LkrVHfE$~5Cky~PYb@B$-A(zuD1pNW(r`$4M2;6@~y+Y6v zwSfnl=u|(A>K^D-Ivhs&xwHXG6Jo$#CrR6o)I2%;vQD29?`kY0tUI7t@n%%AYXk5d zsgt02>M1D`rnzY)j;@>=XJV|K`*Ch{=poB?tTWYESFOnyaG6L z&yzbqWzr@*03~$koXzrFIntVdn7RHu7Ub4)&q;H+pU93cc}niN7eR4e4X?~w$FGIb zJI&u-`Tn|pI==)2&)u>Xx62x(*JRanmDh;9v)9%_YCPA}^q74Sc1(1-?^VzD9E`pQ z)l496bR@ioWQuNy6S%7app|G&t8k;=XWpPrj44Hky-M#bY?idkt7WT#;-&Rd!8>H@ zwA*)*Umne}B^hUP{8~ZlxAvH8!&;DA#olp^BAeb9RxLX${zQp&Na}~42K`C{F|MKC zT%))rKf%M&SA#I-6hy1O6qDh+D|>RlvzL_BUiM6!*gLhb-NESETJTWwfO0__G!mM` zgG!KfGX}^SZ!L7L#(OkIt5=debP}&a&%1lTw@5ezXtp(f^Rh9p_Yup+<@@}Dm$}9W ziA#XX3bN`I%iruZiG!q3%qEI#q+*Y&kV~qKUlo*1Lo~LHe>@@^djBhCCva1ML}S_1 zY*|L%>(jr%f<$gkm87TSmp+78lTMK&9Io ziH;=}A>CWp1#yr(Q4OcUt#57(=XlWJ;HDU?OWQtoBy^pp78FKnIN0@ZRP~s*(xWT1 zDJEM5)Uf*4eTE-ateqz(p^+Lr94(yyyJLGIi3LtfUC{B~*%3jOaWal`LOod5anW2V z*l8nb7Bq~tZkbHExZ6w1>0eL-_4j|mkvTDWK#^DHJsVYZQy$nRe^# zH!e(uD$0oiDi)}sGUhh9ukwB@{=;)oNz{Lhz5{wOAR>S2)5EJ*G`XDzJAZHz7oPm^EuDTBNT)NNob7`}lZcLo%0RLstt zQ>|YS@BW_Sbt}I z(yYG9&b#64%v|6%8I=+zwlNk|N=Cx7>H7=XBS!oyLN4*QxR3Z3EGhnaQ_M(sj`4Vn z_f@iqjQGbX&zB^sZy^hG2}sEQj5R5xoW96U z#OT#-AjP~h^l3+|Booh=wWyeQ6#Jj`&HuPgY+Pq#e|@`wY@AMj?QzUHlR+^%DY6~u z_uHX*d{k{A8TOAUk-c)Hdu*2@S%tni7#nEXfba=pv<1|ZeC4ctv}4w5dlcw=;asp< zSe!Q!yQ#aw<-d&x`TM(<2_Srv zx}tczUV%zQae~!)hc)}nAvkx{^Rf6-VSnTb*Iizx=}$zLM0oeGONXdXf7P(dK=|mf z;^;onUE}$9&WT{dl^5NY8eTk+_02ZCI_5iiOK&L9e(QbRVVBcUqer^}MqjOueiCW8 zr$snA^kwdvS%zb`MOljtuhvCXyl!|^?49Ogc(p=M19emQ%NB_S;4_@IyO7n_gPQ-1 zX%A^53j6=w?>t#Con)HWsZA8Kks=8oC;x++P>l!^EX#7=BpCEPp}ftn6IM?F|J+L^ z!8+6Q(rpeu!0y>QEoJJ%iz`~D89@=hCio-aZCSU+}(9HFa$X<{EIg;&QPh9YMJg`461y_g<% z^hZyTtKnBgs2cJVie2n9KWv`-i|^({uk14~maX_Wi@5WKr+#?wC;uupnwG68f!E0v zcBaL7|1wf;Vp?)22G*cVDi#`dAQhc9r%zNTDFIE=Hhvoc;WlU{KPPK}G;^uE#|v~! zW+Cb|ehneL+}(4UVonBDD%UeD^LMN4glkJEv>or3Mzqgyzjt^2&|K}+X>-dKWN1sh zuFKDoWI0wKU`gctH}@&~IfL$Xw3goBHy}Y)jLMK_1kW{b)(VDQjte$GmsdNdm)k2l zBZKX<*R@EEVuR;NPlUu%Sv#1`63!{t`jC~Jq>$tpyI_oCuBBc{p65xDD2cPtPowCS zr3jWTJ?OR@2vM~`)vA-8wnNBvbWCIsKR;(PRM=U=GTk57nTPGZBz~@i=;IE*YM|iR z=vPhekv#S36t~lB!qNkd^PiIQpytpNQ^Q#khUd})wA>ZsG5vIkxan-blg$`mbK|v7 zn?-KMW$DlAtA7MSYLa7O&$m%b8bvlyu_(xaP%_s>Y>c-7fXy!mM{ z0n0^-0lMTfRBQ&FNuN=)%X@f3yvveK>97m_*BIR^+rU}nox?rDA9mR?r%0&dr@N!x zLy_#r%xZc-GVIb6)1>iq6vjvC9wB_1xvCfEY&Xq)Hu~x#=S=HP_Ay5 z56Pixr2h@wY@|2ElJF$XJ$I0nNpRCiP!|*q2j2rFtY!&nu3;WO!R-P`iF^jjXP!W# z7z!=ns7F<3koRYWKua&@>4n!)7iQ7TAu9rOvwOMOs$Bng6n#XMnhH0A(r7LelMlM9 zQOY>eRZC~OqAQsvLR~g6sr6FS?mFn5C&&{t%W5Ol9d2ra@Y|TuQOASPO3WHB8-?zF z{PgzR|1?7Ge)Pt9%AvfZZ2FlXuTK z#7zn;7q+`@C%uuVU(zl|?jQqZ`(~HF?@~cQKIUa~azUrd}Af?F*!;^V?IN4sC;0uxIn(DWmfP8I& z=lX?vRmsptg_7*ObH5x3@7E4|_$=03lY-3---$C8S`%}(onjygwwa23pgJhk0rK($ zI&kVAdgwsxZ+}n$*r=Q0bFwBynjEi~jLSXxy|(rzE4yUvpmw-UTg3M=FXeOENTP+} zJj_oa-yG73sWDM?*Q+n0Rha5GW?ituqq1rUYv5+3oh<7+0%P z-+Sao+MRf{T4AzU^-;`YiaerXap5QoN152H#o2U2NP?SwmBA8sExpB~4%`(rs$msK z8s8e2*{;CQZa4N8sN{(fws5na#-6cLlV57dlrTpdRY=Y50Dbh zbzw3u`OS^6+ryP^g`bQyprYzvwwf$yt3<9Td*Pa&;R7Z4Y|i<1QEEi4r_<1;gL>m@1L!GGz&{7Iy5M z7hA7>fm>Fvd+pazAvY(pP&x4y(85CXL|W*nRVIbp4Lyx5m*PYwPq15MXsPQZYyB(S zs%CsLr6tC!ko01r9scTaY{lHaGn%Be?;U?aF0k9PIWaaKnm9476mx?j*QnU-fpw7c z?-F7l;CMjQoCBg&fg}D{%ypfhbWWXYWdP!)DW=df*SpZO15}mP@P-zkP(_*?v?zGB z&Cihtf%?&60 zpPLiTSm5zL8=T86_utNJk2%XbO<$wC!godGgn>pU9KY*@dLKA#z1(tNl<+{(7agsG zz_681$hzNm{Httxy5o;!&!eo?0q6ay{`~qMkIyk0q4+=eB#?~hq{8IeTTC&~W0Xt9 z8stAf*TWy&0o*xj-O!~tq15;#hN&M2brOhYsETD>^k0^P-WfP6I#jgR(Y=vn!WPaB zxenydkS(`jwz@%~lb}yq5njQwH;8OQjL8?w@W!5Sfw5R2#(4vBb6S3KhB6{%$Ybat z8Fq2n8fefc8JpBdqL>X7S&KEqeGxrThotIsX}YvQvDVGr7_<&LRv^{7f$V%0E2K_K z&bs)izY!4c71wSgd!2Y|Rbv8(GKzr}v5<>3wyD~Xf2K1#Un2JJX$!YeAE|_ zE4Fuj%nHXU#C0?U-`JZ}Vcz2VRR0 z&okN<7m?}^$!51la^gVmM<$l&0L6edWHEYKTj#BS5WWtw0ESLcwU&Mong*t+S3VGW zClmuJ7!oXtyr8JY_#kApl-n4P>pR4aTZC62^CA{ut6PLE!arXTK>}(gaJ_2j{Y1TU zX5xH11`%z?&3N|6F5K{X@O&q6-s8DO-1PrUx|-}^ha2a0*h#esZb~Tz7~>15*fu^Y zAmwmZM^w1oSbR~1ugW6Z0WwK}O);ZKl0@wh9jbm#s$Z?RU4F^?rUZB2mqX%VWBiPh zFF3?q>xx%YxDB`)UOB)svJYrp7zum5dY*UN;|(Kdjs*Q_4apy`D#M8vp0g&&Mmk=7*HwORN`iuBJ&{S3ia^iVLlsb6~@g?2$}I`-2+@RnOi*N4D9El-5u z*?eW52vr|p--n!OCGvBsB!G|>oF)AMP{EO|5a6xqeIAucdqBoM7%?+4b^wZHC+GRT z-~Cx;L{G-7|9p?6O$V~NvD=SqirGVv3@Y}%?6dh<;k!kh(xREmA)1cX z6!cWvApNuq9g{X1x7S<34xqwtE+>_X zg1@K?Tt-)fl(?0+L0!T@Zxq4NsGu$a$d@w1E_t`{kwLl|20_Du|Jtvu*JebU*^kk; zb^x<}2v&%mc0~vct40I!-r4>$WaW6vlM{Qq7Lwf=0hrcA72#CBkCrTZU$0MrYgO8e zDI!UEF`w2W<;6Wy@yaq%GXAHp-CXA8OR|r)z!Kdq$I`s5Zat)0vK7cVJNTf~hGkL> zijP$4hvF__zBs`%TeU-w;CV@^rO(X-R_Uo=+p!)zqlN^3IvzgLR{kK%T)5GK;^86A zPC<6q4$(zA*|V6thG~-Di&-;0F1XYS%MO}ik|DxXD@$Y??SqVYDD0lV7Z((+gG zm2yWG+saCq;3X1$X&k>|1AtpoZ0DhlL!bJ%xB*u9v>7~qJI8{Tgx?GO*bd;Yt8{!00+`DOD9J?kMphxx{Gaf5@XXzTzl zyF&{He9#p6*vIIQ=oW|GB0Ih`q5C5y?n^$!fES)k#XkG$c5-6j8M>a12S$fAye<%? zuT$l5%fk#6X}$7WKzdW>Uc(!9!5+9>UUmFDF4njqi3l30ZJgcQ%h8PvtpA@RS{#BD z9Q4DurFYGX+`c5{#X_xKOW>x+){qv?26A24Lx1$mTjFZEB;bJ(8AI?~8)&2SgNRZQ zC{A}Tc_PE5?GQM~_Ftf0npE3B^3pJNKM?e_(D%#<&R*8ba^lFDMH;g#D4f?xPLYG& zpyi2NX1MEl?4~czYnE*Acs!+*O|0hM?z3!pK+7J>w3eTHei=zJF%FLzxgtNP9*7fQ z>(CT@|6iT~YY_4BU5-Cu84)8HAM7-jmUr6pj0Lgy2F^LiBYt>Seui|&dPqB`fV*|x zA?1FrrkLe|n_hL2E$+?or$F<#p4=3S9$7nm3J_uq9$wzZ{J+G*c)`5ydgN?+)mxv7 zWrg7&tQdv`+0Zb*&m&EK6gafOTwrrLh%G{qW3T*pM82TZ3rmi$e>hv!M%v|xQ9Bl? zJHhuf1maN<1=#JevAqpyC#!guyj!4O0}I%&XIbCSUM*R}%Tj6Sd_f0%TqQl^zefT@ zk#A&s56a<^1CS;>tygqxgXY#;4xYvu6&Uuia4#48(BLk9H?78zN6)^W@2~C%@8zNZ zT%HK4Hesl3l82$yfgNv*UZhbZaq!nq;(bY4pS4FL=vQ5YZSC~kKtXn6aW%csFNf@g z1o5FogYs_r07PUmIR%{Ck%^2U-+*28alv`w;}MVDduZ$us-%nP9KXv@OX)~+Wt*Kc zS}q5gE6eNFcy_9_@6Hu-_D!c<%PjC|8oExBLXh-3Gfbnvmgr-u6G}wFFb_%Za>8;2 znPJG#fpr6qLNyBHPr!49q}|a(ceVq?fhd8YSpns>bGh}G1jd-lAIfI@i6l625-86k zin5bpwquMiRwsic2=p5>=yb`TTzz21E%LyD(Y)ElGsbK)5&wBmfkvttG9 z#M8(g9utpeue0pn@$)2d-8>{`A(mAgh)k0M609>uj{bNCYTMi84W20?gAhEHAD*U* zX5LtgcKu=SjxcRd1rOGGtPolma5xy7#T`p^z^8#ijG8=MdD6YUCeD2X2ZS}64bJCUu*|@AgqVl{Yri!Q!UA0)=rUPlMQfkD2zr2 zALM7>@iuy_KfCl=BT08+j}_W5#*{7PQ4DaWWKppizwJ;4-Ns)bI58Vmic7E$E|cND z<}UEbH30Wpy!(BA+6;(-oQu*I;JK&Car*h&Pg zSMIBxDkAr>(`O|n#@;d|ui0UuY5~=5&dl_3PSA;~Pb}~+WdZ?Zm#`Wd)hp>LVUZA; zxfW*lHO1ulZqQRxUgLCyI-*+1@)TQ%7l+!5@qRnYc=_eB2Mev#Q!J5VH{3bZS~vBG z@+#dbTpOWLbkol`jx$3%+h%sxBK-c)ZaBOBKIR|J|$5_-M%^?^p)pJW`h_c1%RrBJ5!O19|30%s4r9s&=_TpgV z{S4`Y6@o5=}FU~VQ_ zi?wu5)U(ia@>0%C&Q0-==z1u0OcFJ^*7MsS>1#OB6jJ~s$48=bVVQk#e)PJ%LdgvosWf9!n;TvKPdw?{l7c`;-okemVv1Q5i^7AoLGJ4!_RX%!bF;zq0BLQ&y+o+K!ViRF-l zU+moJw4C#vb6)hk|MP6`^ZcI*iUB6B5-P4-Rw)8Y((u1rWhW#trTN9XBvDBeraN7i z9tdB!@B)nHg|~y6LGC1d7f`Atyz1aKV5uE&%>){za$y;F;ld6@o_yiL zLqWMT@M1?*fmz!y7rG&RnZ^1gYs4A|PZiCV9(>7fe(_~iLS@oF{ZaGeWp|b?$qU@~ zaBj@5kR6R4a>CeKQ53i>p{E7;E)5*~FL1U1Cvk-sE69$Joe>qXqNv5;Yo(Wkx57r> zb%^YwFAM8@wfr7gO~8;-GruMPYs#{8vnR)9xM~#mTsN~Dnl&bi?vPsWsw!l71(>r% zrQA~i=Oj4N3K51wyd{u&cM{0Iu7(YVu~BYz#MblRdoi4huCl@w#oXExZ)yuW>;|F* zySGt%Y)%c7cv;91WNPMDDx=oJnc6x&Qs$(n8jZxS}xd*TLg!Li)b8t zz`=*kOEw4Ii?Z8>qV;GoT4bA@utLkEhi#8Oh%qQI*ZW7)$-b|QdDCvB?wTkD*m@eM zxB=J0l6})wa`ODDV)g{~c%}JiBi1l_8t=2=i@gWswV``raM!42dLKj(&W2-R!lCKe zuKj?+G{1sr*SxXq(dWVkQJB4eS2IPZ-;y~gJd7(RbCO&h=zbmvN)}-RsFy~@Q@j(? zfvr=4*5mGZVKg#g4G8BototzBfF#*_>TRTS3OQ-C1a%ZsO_3@p?)@RB0cdcyK6Jlp zhwR!++|h0P=IZ%-z6G)2Y3bbE(Z@K=l5Fz9kP}|g$~H#hw$4tslwggboqt$@O*Teu zJi#Qm{S}(%ySfD6qKxEmVFSI$}609TKTsjrvlHV;r;T}R`h)yc`bH&^&``d z`zO0k-eKveTYvwn{}37wv|`KUeL+;ISU~i%^MfzB~R3*345528@a3=Z2y-(zZPx4 z$HshF7O7;1569g+A#FzZI6^T%e_Ib(FN)moPI*sQGAEl{4!`Th zrLWD|rYaGt`+@tmp01#8dZM(%!V-<5%WZ#%uFE5`P7XU|lXW~Kp{sxnL3tdKAnotN=k4IxLGza#@JF4|guCCc|Y9q!C3lmY=M!V|25O^VyXzFL-rexL7E7 z9bqwr1;+MN)yh7Ceb`$=?}pq3&g8wEGP_oV7`sYri*_udQh%W9t`MEwRjhM)%o5J zt8u}iu+PN_x;yHjp^))2ZAknAje~Z*Ui>dhfA#tNH0sh1*L`!@cMD#>{>G#EH~wkW z>*{q>+4tLlqa%|p2w1_pD$8=d4#c=egRX?dhn!TjaCW$rK$o)~*vMq#fsa~qHrN@9 zw0QoWQNN7Ep=IotiyrTtdH2H*1D>w`@YWtu;J|olGQv|8#Z*#2(HnOhn%kF0H25Hf zfY1~im|>^)7hwTp3h$^Q3pyh0_Ecj%VwvYD@dt~r_04&xxgV1KRd=jW>=&PRJ}+sK zmC@?Vz!s^#>2J3@3Bov!UrQ1UJN<3Z`l+2R=OvxOUYjHU+X*`?P-6a8R@j*w__Jvn z{>6YD%`L$b(mn;6@{UfG)KW|rMJ`csgRyB`?6Zrdpa!b^rSpd2OS+OBs$1E2P`sA6$3 zkT2sPkQ|B+z%oCE-iSETB25@qs4Nd#!k!myhN}xVtzGQ7Lhru+x3&u^Pj$Jz**MArYYpJmy9@f zwNXFNA&PlKkpU{MM;0%5KysBp&`u9UCi1WWH711YcZNVBrubY2IuL!zFtiFqs3Xz4 zMu9<{VW*W-Gj&AG-Eurfvpe_o9dCR#Z~NB=7p$3o?e)H|gO!DbbGCc8qEK1ur%|kP zdo0E}?>wmC8g{CMmT=fVX|+o?P*q(~HbQmo6{Y?UrBM`ksxg_UKk73{mT$MB9NL0q z0&@}Gw9Rwa3Db)(0}1|M(D)i!nGU-4LQ8i1uR4o^kKc6tOqxFve*v27_^(EB$@?Sa zOIXLfFLTkJcMJf0z4)*d8W27TA98v~20gWWsBhWBJfiUzbV+Kb?@-?Jt?}$os4pm&$K0IrV0s0=JgP#h zZc}D++vWG8kjh)VMm6Zx8GK@f`XG2=Hzf5?gsGKn71i?ff6OfB5?JTN886{_nlXeN zcNbjd8+^jU0KW6Do_m+1Ik1HWJ#a=O@~++W7L{& z{rdf215EbxmL-sUcCjbN{Q$^eBT((57@!&`qvB30m$_~ZZV6Sd{D0k|J&cr=Oeci(RfOCEAE=-^=0ZBJTc>eEPBL_UT`E4Lj`;@xp)ISl#)~uaMxX7Vc11C3@|Opj92g%Kf=xTbySdO+tA+ny z_Gf|jBB6^`k1WNxhS?UaMq24|=+kD|q_^GceL54n^?&-9-486wnLm7*baIYCrewTQ((naq{r1O}a*+NAB+$kg*l4Lms^u5D6a>TzGX2*2 zsgr_}ROPa^sE@q#_hs>io#LmebKP2`t#lUeuKb$ftoQC$#;@?$ABA?<=D7_ilGUQz?ft5>?rb=WDHm&`$qobuT8sVSUhh;+0_G5T;3i_bq1p3}K= zBlF*tXo~ks;}&A4?9w ze*DsaqJXNH+#p>^Hs(ca@vBzUM?8!^#IK%qmv=aL!n$Fwf3eoRu(3XESpmFMj?$6$ zG@aITq;+5^W1%B$-BgWdn_?j9y5~io65$ujKuBG1v5&e)a(C)o@XdQ6_PtjItb6Ar z9g0pN{8%Av^W&zp6&P;LO^d8tK8x@z?NipCvZHJv$&UNqnX?Gxtr$V+f5ZUM< zZ9`rWa0&Eg{Hf`%M%X0jlW7Bowd?pyCwhlt2>AZAmz=|Fgpg5a|@Njv{Np|Cy!X;J?F8r=cMDI!Dip zRVWyAy(Hc-ps&CF-3p({(SM2HRYL-ye}+Q?c5wnB=HWJM;$Y^BI% z>~vocaF!%`WrN+JrH@DB2H~B6ha{PEDY_^C0w%7>ynb00@<8X)JwPtKD%h^NVq|T` zEHy^0`pC2P{ApQgm~pcvy|U@QmYFtvWFtu&*ch?UyRsyDqvTYC{=aRi9$k;6>k;e7 z07+8amJd0VMBkR56Xs8(7Geh_dxFP!?o*Sz{`j&teZ20ENjQ(_ySzTq;8PE3IAkR=s8fkQIcnC6ydX5Fp4 z=2`~%WBYx}L|eJ5I3+@id=!Yaz)C!kU>Y-!Q29&*f(*aoo39=h|H)u*B&|q}A-yk| z!?MaK-{uL$3{vDF^PXg$3pQ^SxtnxFS-XdlJjRGr%{?;YE(f0Zt37zOK2M; zEa2owdxYiZ{qi*o?i4Kv$r5R4NRYGMiiJvXhe_z*2QEmNd%q61Kmw*$T6x=lxzG>V5 z#CZMxrW^pyETh1jD|Ekp+aKo|bXVJ(OLIw;Io;*J4$nCw2l^Pr9HGcZz+>aL$FuI0 zHEyV2@+JLT?4VE(fQ1TJmeQfh17Z?g=Op^Gz+@4 zm38@?9RF*odVUkVD>{K!PhXdILoR8u=yG^@)X1G3-)#GK((B382UgU-ptu-^lo{o8 zf7#*q1(Ica&G?b)!`BR0I{%#;d8C3}jNgGxCou+9GJ9CBWACvp`)ql@3-K5|P|_#JvRSB?3Yk4X7c9mgbQk*CEp@b*dKLt3PH z&gI-;r#+rp8Z%sQ+Jga)#T~faD7-cY6CJui^?f4j20&uzg>~xAjsP1!y5PxU8(%9n zrB(Whcv==nn67$OD=K_)l=$C;3kO^u%+`TJJ6)Yg>-a@4D6uABS5QgB9p{y^b_JUPy-VVB=6w|3e(=830E zY6sXno!gp&TmEnIry8Ws_5bMn8(A@hSm1%i!nK;{6>`k(zeAVNC&*zDIyw5T_VI#) z(G}tzJF+G||JP*oQlEd^ZX=e3y%}S|ajk^656-<|8t}CcDIJ8qj#v?=OCUQziX%=b zAB*orHqp9Xj!;qWyh^@O)(yLO)=6I$OU4Qu6Bznn1rEo}0Q%3j-#Pv@gW!1UM(Y-G zz=8LRI*bgF;}ipRyN$SC^azNfOGDOA?G>t9+>0dL5IVtjk&sp3v2soeqwatrwtm@} z;9;js`jWWCy`OuHdt)~49~O#k2Q@<>8HRe)8YNzCB%^N_b@>zbivfAzhl00nP6i*>1j6UZYia9`$eN@~Y=03E0EfAp9mr18aKLi$_%IMw7GI_VGQ&Tp!x!`cQ_QY-?N&NA=lzXL9_Yum=>L&1o4F<~+8Hq?_OU zra>potK#>NhLg_#`CZVfi&M*^2gw65ad!$D?S+kF)3X zsl}0J@EP-wWt@HIf%=cW223q|*LyX|{K|xZYmCrVN-@xgrhtn3WZt3#s)tVDb<02X zZHCY+G8(BjNp6KYNYJr5?I9dbwYl^iXh~Z?A7E4{pKYGkl`c<*|DgjXTs( zJ#9D|4_Llr^=yNsFZm_oxP0K|ZXbUGVjjFDT}HOELyY6vY{26xphlxw25#3KLaa8P{gU*lt&CD&ZGMdyX``1e^ zdnjgyTNyBiBe~kAbD`TUMr8VaHRTZ8)Fyy@8+h{YmX{dPxhwE^3?#TRYYlBuY0}#2xnvx<;10=j4A77Jf&_^@p;+zgu!qoM?XCbun zh*a}CT?#n4G~RiFoQlBfRszG)@>vfUOEX(H9xrKq$C|=M2hN+aP}r!iW84zkOG9ty zT;z19IjTD7*N(gD|VSJPL1sc4c792}2En2YpO!%mI`BsICq|;;1jQVsfNU=AuJ|0u7G;oq5QH0joGw2@ zZ{QYiHHwGcA47g7s+9C_Xa*L#jub`J$nHrl&B}spg1fGLz|4ASRenX>i^l=yn$h)aAl`%Cp|(a&^4maPTVVg9$BFe7m5ksf&-nmO_{DX1uUmsqp}W z<*yjm?Postn{T8+E&cY)oMKY;m9ckE8v&$&VrnUJkczwNc~O3fn<*aonJ!;PHZ%9Y zhQU$2@;VnZ=(?IWNNd4cIwjf1%@%F&?h)w~SY>p6P)f{cQSnT@dVBP>4csgSpDp0@ zLBu!PcSyFy59eIXYz?{qjEB{Mwjt{s4~9nRCG&^lLXGi5A2pxj%ranT{*{8&q|kv) z8?as0yDx0@Zwgdn#+;Temtnub zWs$jbO0-te4?C2oXR3pNHx;^T0B0el^<$I|yZM}i%91`BTcJK6nSojw`!XMh=(QoU zs?FhGysG0qV`a%q{_Xw!T+_rDht0XOkOZ^bYeT?VQH{$@S(*|u9hmdXE%A-nJ;HSP zS?`+AmS~NlD!NIuNskb69nIoBb+-^{NS#tZg~)mR3QKeL9{!fB_moH{{J2%WhS$RHOhtQQtA8+_~* za2xyTo3}pg#q_~ShD@rB$XRJkG{u3pjx5A*FOV+&N7J_ZRRF&bCa-j;l7S^;qv!z* z@wFM-r=N=WjDA3$XXC9wP4vNe+vy(FhNueJ?Vue03*RpJL5HfIRLBOzyXRh| zKNYpvYct5Oy5q))ot4q`AN?X*NPyn)VR%%#<_>O0OP`0Q+Sw&rx-;)%(SGqhT_Wt~IZ+ATJ)-E5N z$)v0#yW8;*kP@Rjg+G zScsL3o5*4htqdEhKlVCGr;4zRpB+IoYTd_yr%_Ma?{LG8O&ozioV;5;^)F=gS0;>K zXe4~LQp{$AGYFqu^Z&Xik3#X2C|&|(X9oj57dLP!{O|Bfx%;>k!ZV)5E+-*?p9kHK zCk7qkNItM1Bu3A?GRY8J%2FUUk4<3DLqoa&i5kgOZt>Ow`CxOXzPoiTKZVyP zTnE%C%fWMA9$qQi&UEPr#xOd6#JSg2G%BOt?bG9N$cAUC#$Jvv{OMTfno0khxcq&E zL6WFW{HB^@JFsC=XCxcSCV zbZw7G20r#&8ux3E`>9-bPOLsg$~^G|GbGa?P74H?X`-r_9OVvfx_mM3lAo;uDb`_T z6nt!qA9k3TvhrTpOao@#|Dd~?>@-)1=)jRypq(2Pf7?$ndnvL9i8-{iHn5hz2Ktiq zM`8DL=u3vl1!bO4|AkDYhF)d3)tAj(5sieTS3U1VLV&hK3hkeTSAl<_7BUjlb-}v@ zrwG<~sJD8onboGy=d-B0X#KRdQBbyn(^F_Y6>ruYSccP&HLF@pP21<|;vG0i(t-ht zy`)NIr(P}N904!Kz|Vp~`&%K+vZAO#ZWdI3?3ouI3)zsm|8?=#eYPHlq* zb1P;gLd;ERlQGK(@%rH}r<-=SbJz`H3vKsc(fL>8xqf@*WQwxFbs|JZ5_eEp9C3sU z$)Iuh0Lf7{aO#!G-u}-96XmBTsW-?O2R3W(8kr+kDdsXoI;l7f>6X>`4mlNz zTlhyvUGN2;8$MU$1Ekvhi!iNZvww%;q+-aa%(IKXQnkUml`fjr;8zYw*>%AiMUC7J z71FI|F()Kl^aGGNUHl$?hw8GpN7>KqrRyb^#J$p1=*=20zzXJTs^wGaf|CR#!Zy`) zPpwxu-LASoYC|*WI?oPOYE-TunbW4)sFTl|>df6w^cniu00_A^k2$ zyPV9q8rDQZxI)c(Rpzb>r7VR1Qadw~@=CQlc>_phdPtJ{sf z^fZb|rARUrhn0Epf~6suF6&4i?{dT=sK&~WoF%vAHK7yr(x1MnvA+GM4@}SjtGDmC zA#j)P{pe!2K{LtTQ*R@s4y>6@8sW8$VxXw7ii*2E>&&cGGX}X$kmlMiOZ3v8zne;3 zoVR@b@~`*IyEyN^YFsjDvsbF9ir@6i`>!e>!^0r|9?<9Axy}OFCq^^cv=~unyH` zD)XDk^SkAE7mk+$>!tNy!dRirQoC6F0kDJi((it4Xtn`C@7AaqNYPg&_u2$I2+>2@pt(Vmbk*UANrCydev(9&on?|vWrv>`Od~O1{B+ln7;|+vq6!-ia zI0gRa18RjwqibAhLT`ADoOP&{@{(g8L1Omi;AWs%tQ9U;5?L!O3|N3OY80ih7p7T{ zGaVpo4>qfDrklX&NcyLyR2yt|7acgRXMrQNoUVZ#0|?S#qJ9@&H4 zJZ_ItJ;?2q|J_FGK896d35_w#Xq#K!i`l--)VF2>4ae2G#DXVY_lry7kH8-)lOI=Z z3BM9rE8Iojn3XsMkCy>&)2W~&_g=b~bQ8pi{tjr+2aP1`j~13F8Btocfx-)B`SY#b z!=?hn!k3rzN{J9!(}~n~`B+N>TQMQm*s$oXrn*1{B|a}%}yIRuAU_;jk>@-rkF1%@&Gd!uy!b2ekiEYD@m{i9NI#ki=No7 z6%bO&t*1|hZw=3O1^i$aTT&syg0j zuRP~EUpuBb7{2-O3ZiEs!#+or_1pP>RpfgZFmnFYbMKNg2aX2sF~UO*#bi@NL&fR! zO%IJ>7DPl9T}is!)=JgLf6(rN{9!$`Ix8Gg9gthQZYEkn`E(ztRwmh3GqH8}Mq9h~ zfx^hFmLXv->7m7kH(;8jX(1VMLgwe+#CrTL{#&5BwdZ3l;B9NE*Om&a}n3CjdpNKPj% z{~h=5@5KG#&%gQIg8vpTqnITWiMPF#JOZ6D{tTC|2r?x|eW_CIT8zZaM-&76>@`5$ zq9_5TjS7FTX>yedq}^h>tGEgo=#$i1Iw5&qoF1h z{d~``HM|@SymerKm!pAhn{!dD|F52kd6oAjuj@=Vw3SJop!&3D%RXilJ$qzAj$(t} z%QDqMY=Sioj0+3;>Drtfy#FbhSKyxtPV5!=a^6vJznf@md4*0dbTIPl3@wLT&JEs; z+#XR@NUEq(QWUi-8dMMe3dskjy^Cz{-cAPP^|Ur3g||0)w`YdyfGax8n>g8!;fL)5 zJ5-mTZsoBUE5w_?_TJzPztmbmHD(k&r+vO;z)GwwTJfVt9D|$m$72tUk|hr8CV^Kp z%57dxF)0*TL&fDN(Nt0Io>L<16Q@J0u2@GPheamVE*tB&JJxk(J+YM!u+lk`C;$1) ze=)U9z9O>4f+e$Pi(>WcE#Sv>fgHKO*$1)!GpBH@L$!&-3$_OBbi*s!*^9k#v?K!S zmoy5^{Qr44uf*yHmxmk%Rwak^+6eH3jFKK} zDP}cAR#0)}!UO?!kWsfOx)m5d**#;zr3S9TSk~CO)f(T^tR%;zfb#NMQ$(@Qg4s4C?Hj*3L zDJGL5TdBCFxl1_N$~3=HZsLrk0mvIy!=&&s0`Jn8t_v~M4GzKV`(G>PCHL@&9lr0-PrW!zU{XaVYMpm%fP;_9i0Te%@)Xrv# zfgsO%Dh`XD4TdX*_ICvpL7^)!8^5Zqo0H5d7vjmFtCdkPRv$1+)UcY{nwvv^x8i4} zyL}EDD6_D$*W!-F?;BtHMA(Zvwov(VXx{yKY8=?+p5>AiUgmj93=vU{0#o?XP|(LYu(2WEL7Q_ls` z4Q_0={PT#%z&ZLPa)W4}nfz3;edvTw?mMq%gHX-0N z0n?bUYVyoPoU+5}k7}3xff#I%Jx*CC$#Qd!t^<22xkh%$Mv6(JNGcUq6R|P|x;}Nf z?DbI}an?$*oLeH)7l^()3%V(pq62nkvL3bEV?xVQ9Wm-bJ0Ds`%eNwa|FeIr>aRqv zyJB={?utJbHqq;Tv7LSr&?CdfgAb!u#2^`V6J0$YiLjRt9Npl(*S(3}{Nr(VBV@6* z?C>*jK-y3hwCQD6DTjid`Y^Z^rw?%auS$7_PRl~Vj9WWd`^_sqcQvS#s&9R_m8_aV z3XG)m7K+(K0ixq@e>+{Sl`ZD9hIR^X$kaV_>Z|*GP6rQy9hx8*bOkMsB*0f?#SogjxrH-a*N)~i`|mlrRnj;VCPD!@E$h6LMj zw>a+xN{sTaUlNZzQQ>ya6`!($gkp{Tr;29uf$cwoqF+ym|Ho=%ij1M1q(yots6fyI ze6)R1EDtb1%wkC#1* z4AXg#%cO%rW(6>(d^R&RJmeU*!?J(Lf{hUsUowmBFSA0$~H?Uy)o83z>*l zJ+-oxL`&Zn>k9XD)Hr%Z+BM8XqiBp6vj48X2@L=Je^&OH(h@vhJ>$T(sRgQn&jMRG zeRKgQ3zCx;6oume<^?oE%}hRhQt^O5KkWw2c`$1->2gnXo3tDJevRS+rvWH` zs>HYCxDC}s*G<)UTAx2+6SzDB&a)#cF=H7n<^m}n`JDa9Kb9)$3A-S?0Gp9({l4%@ z<>pW8Run=z(-gsgWMcLoM=$SKF*N#_i8%h=nIAnd%`CEzz|+cE(BfMuSv@sfep|lr zYjbUCujJLw++Xm%xaR#cQtR2zT=EvLN55!32xnrq~M&nA9Jg9 z&E+~&{qi+|2_&CeLU-}g!!ZQcD^HC^Rd^|&Tk$bBe&)T|=xYy!eF>X`Yg|?WJ^m-2 znbVOjVS_8n1s~%w*Z>z}{7vj`vn(}NZ(f&W8n|`XxLlEuXwgv27K&`5;vRkNj>{3B zFJza*m*kbQ8lbOf=GXW(aN51HMT4##DlECmiztZ5=GOS`5O+^htuK2puv^se-fbBS zW^7z)_3WY7-!*8etcOig$YPx;#>&L>9pKSYWZ}mt&tGh~knv^!|`^ zK_jV*IVE1@rXIPo9-h}I?$QZl&^11`O|d8NV<5vym*1oNqt?gtLxG=xd#hhwz-gjU zN8yYOoYn5hoV&rtJc?#~;=bNPD>*MY<-LJ7>{J`Hok`dAz1tI*0#@L1=S?1aWp}V! z-X<<)%Pbe%k`FuW5@4Pt(8}p<$_vj6*WH9fVPLDvmr5O$e9JN@k0Cf?_` zMxdol_rxim&K>dxkOeK9a!s{hhzyVvZn4V{S?%0Jul3p}DuHe>RiZq4Rg`_ym>J2! zR+yQY!Y)IPZ=U^|HAzqoEVwLCU}I7x3LoG{?}6?MxTA#RrFnFvR|=<@PV>9#s!o}i zLY6|FOg_C!w>h~qyo?Kuy1mJ}#$Wbf7i>J6w00Bsq+}%4cUSlalETj2a^UnzXsJ3X zry+}CprAB^ic1!0CEAF@8S6u<=yKt(6KbCfj+Q(kIpWV=N%a85ag}}uT#?C~w3yn^ z4zHzB>!W(lUby8Wwt__GCRv`&W>^02fo(0OeGDzM3qX1?@?y}kylNTa@9l`)9%_KiiuAtSq<9KBVYCu86mx(g`>41qdZXk71B%^e z6(@^sp?A)Kv@&6*a074$rE{Bu9zbSEo9g%vFT+~2t4`2Ra;4fyx5hx)+MG63YfQ79RXj(Alu<;oSm27>|1A%C&9)!r-fa^NHTx{Hq9<7$H@E?SJE!bhXP=MXH2P9o? z=~HbhG|VGpT(~tql}$%k;davXz^MOt*|YmKibMYSGb;i%Gg>}kAelGubbPuc91A3D zy4GWyj1?p%S0uRa4l!t)>p#4;hZIa9O-350ief4$Am5BjoRTp0gs^vZmk_%r)zclS zWNv@hPCDNwTU5X;wJ?dV|@K%xsBBlIj**8OM0c%lvT%JwT%S^ zop`}vXkySxKbD?Q)(CZut{=drU7q?9?J1bbIVx909#jkk9|Id-_!M z39=z-HwTec%TJ%0MA-#n)}hO$ZkiFkj`aEco3{Zk+PPtu$QE{Zao{j5G{zkjrrkv` z(B3_liW_uW1DPwp-K3@KVKWlt%O*}$&>i_wz{(wh0%(U@{!DehJef1-*6e&ozBmR# zw+J#VQ2HR8S*ut01ju55sx z5yfWu(<&&59>o6DG+D}Fja>`n0m&ldD9;t_a9hJ>{0Gd zH}fznduH;_=lrtK6ftZhj^jFh(5Pj!ek!FH$h9hfUQxo^@=cNzvsOemLe9Yo$Wp%N zif8qZgm>2aw!DePvyT#A34LkC?)u7SUcf@CSvA;Y593CWchBSK6pO?%@5 zG#DF0V)Ec8fBcJSD~*>*uUKf9Q68mhx`NqC+kDY6E)?DLek{fe*H7o>(;8(m)WIUJ zMi&HFGHFCsx*QRPk=@nqc2+av!wy2pzWA|>okzPmJWclpEf587lH3X_k3i*(WLsS{ z9^3hSntv-}dCTUXwC||pw`@j+JFZE$zoVsvZ%{7t&u#pOtYjBtb6_V1VrrvQQ3l0q zq(~YB`ypYbLzPDENe=m64SV#pE2N6*CxbNP?uFO#>m-97A92so&EXS>_OW36nF_Km zu1C-AdcCb`t^pivZ!XOxRSs-%F63&s`7Z4$eOIPIkJAB|Q-xK9+uReURLB~=Iz#gVmrZGA zjw-3gOQJ` zo{7P^bekfDhvfN%f`!P^k>z>MtChZOJDz)U3o+J}HzzI3&1b7IR?=hgZ}%Nmm?l1c zMWly?G)at}UUU9L@j$+d?iCI>#S0#ijjD~Hovv`s&AI{(4qnyLedM8C`n#K5XSQH6 zJBb|zkN@YwYEza{3-q4*B#qNs!wVvGg_~MBgL8UbeC)@H43FiJDZGdCutHNkK&oOO zH7GX6zdR~kejS{kBIuEmEEk_CZjKqh8n$oS z+$a2W%y??6&RBiAj=Qg#6TAM??-;Dx$^HNDHaYCT*6mdz8>5|KS}AfGGfeVl=F?f> zxq!-R+)Ls$4D=Qy*CP^xugDL%4M;SKOnRGVc|?QH1y0Rezw5U8D=916u= zej4QfspXy^hk{z9a2d0mG-Rz`nqQ|2#N#@k`7!>A43ASJ7ZhB!ZzoibVgI>pp2ZQJ zE~T;*p1O?#Q^D^Xbgd2TrT5HR%RQhhj;M)H=gI5Xkwe)O@K2F%>0s>N06W0{h5pVY z(>`+!YtvZh9QTOeo^Ub;sVQOeSl{xv&Ua^oMv={35xp%)OKYK`Z369BcGwvE+FE-j zD{M>-srvM{T!YMU3Uo>!iR{8Hjw?;c4kHPbPBH5!vX+W_KwtONa2v@--1|}8M4c=e zoR!91>xW&7^69*Z-V_=;imb&2yXOl(oLOqx*X1i>-&*MSQsh(Q1HG@kbRlNwLx|Lc z$R>D=+Z(7+-1Kgk2wt;Mi!^p**g1SZ>d%CS_h&unF%8f1-JBxc2OC{ZDlSLog2yue zHO)EnV{spu5Gt=RV#LJfus>2*VT8IqFaMmUL2k6YUAd5Kc3`=&&q!_*P)r^L#%A0u z!9``a9FvRgMb(B5IU#XLndf#%b>tvyzBJ30K+CVqoRr`l+`Rz{`glpu9K9CeXlK1w z5m};F1Y4(S>4y;SSkV0Q3*TS;gUWgRKfW=q_uF-Eu6(`Z?V9i0 z`ex^^3x0d<8+U)u_|wnlSI%3YPS>@s)Tc<}=AJhCn7>AWTYk3P(XkRYPXT0W!-W-Y zCbh?XFvoPm&tdhDg)KbHN7LCRto4+SA8u%mi-x*T2DiABbkabfSgGZTcd z!-dc9esI_{X!=s6Fj@$tw(y%}m&JEN;=Q*(6NY_K%xpcclPRcvu+{BEcmH95mkC9T zWxTZi-#5DF8!V8>MQ6H54ZEPJ<7VHH^G5c{Ns2j60X;_?lFDl78kbBuGeo1P4y^P2 zED$5^`E-H+^IX;4oI}AGoB@|w*atoz`dD1-Q4Upqne-z_tVVMfYPsl|P<39oHewC4 z)~^wYx_TqjD}abCgSYY8o14xu9?XaXs_ybX zyPMKiIIQJrfsjJ4S6ZY~`zj*?~OXvrNo9s+!PqdeOU6tkWpDO4P~vY5(>_K2>n4pLtAf}Y6| zZT24yvy&ZaacRd27LW-R=di~G%V7EUpOoaxGyvoM54x+#&M5@y<3_#f{S*W2hCQzDwU>5_yG!i{cxzHxYFvoR-FBiPU>y3ZY$5 zKAlTLswn2igQoz`sY_$^4=1@8%KniBB!8y98ATaqXD}esv@H0 zH!3^5Pm(R(klP+~gcNY{AUkzgWT~tsqC^kaah<`5QW#g`=rHlZp>~|J&x`5ZdaL}cgq{1k%g~K z)ehZxyZCrlhgS_m&eu;Zqw&cO-II8f2Q1k+&MOnkbHn*5oy!W15wH@a<{!@)&#D(Mv-3aa7V zbpck`;)qkq$KsxdZ9%J`X;W6XuJmlPD%r!PXnMBl*5m5gT(&&=U2Au`HEHt>9A2@I zHs2!6QDT!wV0Hvn;6b+(UL(Y8GwC8p5fpnZ56>fubpNvx5T08+*2s8n2Kx-Wd+^#F zml1VdpmmeH#^n=|N$=($-CicG)v@=e4@Y(J@dV@h z6YD3jf`)xiV0PA5o@AadO`3nHBCi$_>Ca0#g=u~Z)44UUZ=+{}M!$KYL$2`7$GGxl zSI3@%md%heg;lFf<7sS64hN<|vJibcO84;3ar)@W=xkB3xJ)!i55|^9Y5B$OsT0>n zc4V=~2pF5%q$9ukML%c08{gq&&?0Z0p43h<%*E3j*oxS1q$>(32GZ1WpdTodYbOYF z{i&f-G(>Hx6%{@S0^HMu6Af4rEf9{c^uh*QwcIvEF7V641${aAAPng%z8mJkpq(s@ zu|mSWF`VbYVOfUM>`LH88o=_~GjobbnFGhBg9>$n9CIDq~hA31ye6%F+K^qAC?ua5AJ`#tPS(_3mtGeghsE^nNNATLSQD=91K+1m)PwSj(xG`~I*NX@UTKrOV+pK+crpQU?E=AdkZGX=J&?U*7|X{ZDI;lQ}(f z4^WNbWV;00nPgEGl&bgAosdyitFBronxXlG2m5U?-Ol^Jmhes5|4 zJ8W#ug0-9G*C8tq>k=J1gds(9J! z%VL^p?7*Sm zhv)pyh+4XOnvi}6Hf>TTkB*NlnaspbCQqn1+|fyJUhGj%uOvvW)+sEbi=q;t(Pc9q zo1$D2Z+s0YK`(tDx>$27eA-n*PW!kOvZAP?L07_VklR6tQ}DS2k`1kw?~oQnh4`xU zEGY`FVepP^?pZGQeE#>IJ%5zG5_Z=WukID$^CQ0-M@l?uq0MHS>e08K-9K_2|19_X zM?W$0r`!A;{oNg^_}Gu;{ZIY8cGY!hiN|7RjY4m;E%vykdiIyp{{{XZI~0zf(f&VK zp^$o|(d*93o`H+Z_p96rMF*8g6p4X-Ox5p#(LzCsbmi1cFnm&YdzFu%?jv2^5Y#W< zNq;nLyI(789cdIxV?P%o{agoA8IAwgOWRqEmoJe%=KmjbF2*9gQ-z*$7QmgXcN9GhVnDKgS9@RO|PaDgJB#&%ZVOd>*;`l7)WO8mYm>lbIok zJfh+<0yF6)f$Ia8hp%wY71y|A1SY_y!IHqLnC0OeUQ4BS=q0hO!X4anc?JjDD-?># z=r;FM(Mor0SUK$UDbo+`SB=Z9$laWR`8PfLWkuYLGt>pjJnkW0zDsQ=PJ*q(`nhWu zG`~>Bb%E2QubxNhg|&WITsG`f%FW^|3EUr2C`yge6KbmQH}0a#=$+7wOiO>n)k>~< zek{b)Pu=z?&V|hnFN^c(;UK(G3%>!Kv~+ovbFJG7_XqON0`>6OGZmNXwv;AJ8_Y%hUXFftA|sy;iJ|2*%&AZO>Rq1jo(Ldh6`*qu~aa z$lgT6dtF82j%Jv&KjoD>vV7z(3uW9 zwZ~M-^63MC2?F4fx);^KX^+u;0Vr+?MfvoN*$^Z`KKA3X3s3;N!>x>S)e~og4+aTR z%Q)v?7$@E(J{SZ>m`}I-y7Y)M@OoRTA+7lkp3{n>2dw-q-quEXnO0a^C{6C-uO#cp zPPaVgqtV&QZg^8UbOOl%mnb#5LsiHvCiz^*Z~}v+LzTcg0THx3x>1!#{blHf?eiD^ zyz|$~DQp7RN5_NTqfuPuFAcz{`sKJ8g$Yy_MF(D8>Yl}9(rw}`e#PP>)qrb{>@cTZ z*1#EYbs!U%e~GW<=Z-gh%5DqsbWz5)BCI5m<8C@i*Y&1bQ^exH5~<8cBJH4_ZsaHV~)y=V1 zuDlp*R)GD*6I*%p-bwxyYf$32_$L-naul`>2Dy91ee@uG&#NB1vJKvQ-9g!ip70=| zBN%i&!_>O<(c5214!sjn3+&sQUja!dT070Y2*HIMCrYfuz?VP1yy{+pX`5XO{b_eh zyBWDJW&o(lu@&4v*p)Esy@NVGD;!;s6iAiJRX&(q6?1axCT}E(D;E|~4}b<8IRqCA zP~Ug^54vk5=Oum7SR8Gi*FMfy{ZzNW!I!_caooX5`#A1aBfoCX2~)S}E0WA(!TUP| z5hBbA#Wn!P+z-l1=sc*M%n$5__Eag(&C+__Mo|@wqZmcJF71|Y_Prt4J?9bks(ri9 zAPC2c_Hh81)gx!?XB;yvmUGx$Pz!}|dO`=x``h7`Pw$hcYh18KR--uQcEs~uWc&mZ zz_3_qR!e2=&Flo*kN=NbZjiyBo|*Y!D#@Ed8jZ}pN{WGIP`jzP%XD94o9Y6%GVQ8r z{#AL0s(@4GiAe=Z1Vc`Tg6=?Hxgn<^`DXukUQyIz&W@-dr(C*$Y32_(-S%ESt)656 zMN1}ptksd_@Bx?d9&l>7X`<`m6T&?1QCsUD7V9%w*_b)Dc14VzXSF_#YxaFT!MEJi z0Gz6CeYcgYa^MYz0wY9kp_ok+zyqqXv6*hae4FPNUX2iqO`0DcTNSgEv(_(99546+ zEUE{PT)dam^X+3B83RLq#uB#Xrm)KxJJ?=2$bI{_26QBDOYkL^%AjKTZ@6h_UtWSY~% z#t?Q~6Q5Y6ho);3HNI-xs6yV5O34m4HEtXw@)Ds-TQ6kx-QnBIvol^9-5Fch zn3=*ZLx&bDi+tJ5--AIKCA#l5b03AN)8!os*w?BG7!DegFLQ-jN^G_|aaxL{r(>5z zH9zSKB*gQDPm_ONa;xrT&tu=>*9Y0TL$U-UtMJ+hnxZE31~gkujYg5MB=UyWwvgRH z9je2@n}hFp-SavSoi=A#WEXUI#OtvC4K2b;g0=?LxIB~%M4cAt({C_*R~@lRzASQg z5VjjY_?3woc*X8RzUxSlP3mUYiir8+yQH4omZSq~ zo{L6SP&38A9_4XJ)Zw+br|`5SO^G{)75<=S1o1%DTT6fDrB02?rLm=Gfuai-d=I>e zbSgBjJAXW6m z?>->$>>%a9hPB2BQYjR(h9WDexNiSk|E-#iDcU0%&#p@`!;8JRX9*3%?QD??ARgZ@V)+y+ zJ_sInK*868tT4B1L!HYVi%3sajSW0ek)b|b5DEi5*%{tpW8}ro1Ph1 z;eTfW!@N2ij0Hw|m1tksn9hEa-`uJ&C==C*-&B)qcIKx81EbCe7-bYwLXllm+;J!+ z`0%Q{X5M|-T6r(TbfKRf)FM>RJWF;2><>ArI7+v=#|t!&keL#5#OERqpgnR;jm9fG z0_y2x5o}=bn;~ryThQx+y>{w#Gj`03GNzt8G+lA)&%gEi?5^c+;Pyv#1BU=Oyq@XudWn{gJ4VL@?J?&7L>oDa1zr69utC?O zUK_l%{3pBwXYg@nK6YL0lEV8~xR+q1LG`rNQ*XVpUDBb#MD;HI6R#Vy@%ejEE#CD| zP|z*Mz@LpYooxU#8d27R2WRN~R^(gO2ptFZc`XQ?BhLC1?3HuUAq=-U_>ibyc86ax z9XEV#es_0JQFsdP=68o9v$+V4LJ{h&G{0VYi@eW2gWoT&j>G`o-l>;J8>dl>agvNv2__D$%0pn5t~dxI;2yU2h?KlfhLfJc|`FlPz1d`cGG z4z!v(WBPz+Bzpph!db4jZ5DG3v#_#nsfWu?dYWeXo7n?!SfAcf$29C#gS)ozkuCgk zpM0=6mWBXTCYE_%CN^&EqF%u)(5&!I;hJ!*Y-GM3A#@HWLPG2#`T&cx^2b`T$*^hJH9DJk3Wm#lR=xMJs?g05nnIDrC zj_J>MM{i(>@*Hm`t-roelm~Rn@q%VXHHs_pyPVI3 z8Mby*Y_}GsZ_~PS%WGkdP{?6FeY?xI5b!b=Qxh?}h zpoYx~>|u7i06b$y>@Q7|R-V^T4(v`_NKNUTwE+I#p*k(92~7~} z1yZyQ)e8A-d4;UeE1A>a)1f*+Cv*C_)sYhho3R5%uU+hb4YN~P2F$~yr7JuQ9(CK> zl?%z{mn?@4SRF=@_ZCnLu=(zw;xI2M2P6iTaQDm8(}fF9JW|7vPaCWvTX(}V`a^0)l9hoD;raT zKOq|&c>AZq2p4%21HH$ysJJE?xvM^tYM|Rq0=dRrEKt`+mvN9UVU->}W6YdPsKY z2ZhkjcR6pHXNJe6&=hivKrZ7g_p{#RP$B)lpm_V8KmBK%m}Z;^MVYaYCif@(eolx1 zDA#{@YY!=y0!^4lf2UOxQ%R9>DsI7(Sz5_C_#g-A1scUevNvG)wD^$Qv(C)g@3SSm zW|hZBwN1PK$c!1wAT!xJ^~F0J}vjI4uHYpr5KuHS%1$ zs);r{15emu>8$#0jXvU0CpP$WEq}LWmdJ6n{nW%SekL19hcqP@paQW4TCeuUBELzCv`?HKj=pIFBu8h`c{Ij4^>iljA%$XOq0FZ-Xm0;} z)p6s*23UB}6s&OKxM^d5dqeqK(_92L`-%=6$+eKVfN@ylT-7L0TR{Guhd%6`Jo!M>Jx97{c(M`Yhu_xKph zkSENE6-48}z7B9Wjq-s@C}tN0^*nJKMVQMLA5tB8FRC`Q!Kc%uO_fP!hOBnU;FQKD zdhHh@4;>QEm&>xb=mlR2z96g*+z0fBhuv%GesW)y?g}J{aPI}7*4FXhk!4{oAx0U_ zy2Fu~Z2Oayp+p6Eo?7>^r@mpZ1Ys~71kFjYY!9jIY~eGpU6?UJg6*U>xStgyCjZIx z%zSIw?m4cLMl7`5YlJH1Ci)ccU!LP+ifVED!_fC+q|1S|4X}TY z(l&h*^EpLsQ*rf@#lTFLLJCEPg0P_?l+-MZZGfoSVJLt{j%M}c=uF@QDio#g@an=+ z86H5D;SR1=c2}MuNfSZMa2-hu9uBC%MmSi12qStv=Wh8{=SN`+Z}DFy3r_Q=EtM`j2c2X!ew)2?Mli}SWL!n92NT}nB?vMha-=I|T_2hc z@qvqTk_0s_gOKL35Vy9miEelYBWeeJsf{yOY*FY2D_ zpzGhz>%l6p30r8E<;)XP-=TvLE3;dj3(_{n?*5uMbTQ zej#kx$`tb^&-g}PB$QYIjT-u1;di+Ppg9FPC6GjOd(#fAhG5Hal%<5#Z4pPfjocS5?2BIRIIoQvE90DhypLJN%0NN-46y+h zvJd}1d*1@rRDR{}6>ms>81f>RWI#nC7{oDxp&~YnGo5zZZ99EzJKO%d+ih27JGR?) zx|`1SF&#wki3ln>g7Q!v^6&xih4S`+k5Pd^MG+quK!@T35E)eXpOb`kw7GB$iTho$=IPLyCST*RG+{CQfUiT-o4!Ksl;omatGOaE#=tM{j z6op`nsD=;p2$;ukhl6Rjkg}U6Xr(uYr1%bSDcf1Z*epQ?;KKN7J05Jgk{xD}*L*u| znpvuRC-2|iC)?Q77CCW1@rZ>C%BK`SRhLCYbj>YN;N}_$vvnKQ(8-aM(y5|8*_kO< zggR`Ul%hnwU)^V(U30NFK)-yCaFu6QNR%K=q$~C_$g}B9Qov;eXWbMe0kbo%jB9Kc zaM^+ume;iZOU)vo87`}nZ~g~a=fp|~;sJvUk?oWsi6UF52+Z+E02qBGbsW?}*ZqK9 zC5eMlrcT+(y*Q;>f<0ie>1ydRA4jb3i;sO2EB@jn4j5r4PAZgl{yfu+kM5sISCAc( zNV&x_6jF-46v=s(`Z1TKRDFJGIv4w= zVpA#*_NI#CGE%3^qPu|p)F8VqH~_sMupq;}Yc`6tVG{OsT_IL$pF7$ue)q9YicxhcL)4peTc;=%LMyWnI4(I!1Q6FJXQhy zvQ!@?1{pvv;XV2ggZx8mI%B-vzNjCqSv&;v?TY#d6MlPpW2v-2Jk6a+zcAb!A)*wt|jvB0Ha8CFrQxo5K!R zdr+Jf9t_@+Jk_3Hi0xp)8zwK|4hPw7%LuT#agX!H+(L@o0LG_0b^<=hlQ?|PUAJvy&g!aXCwq2w{N8J0lgBB-+AqiO$Q29OUy=Vcv)Bw zx%JBIf%`3N?}wD)9!1(rTcAy%qJ@b-$CBgMHGj>V6lER9AU{nKp`~V?O4kh37;T<< zf^3qyJ5p#o6@-2|+&Uq77+UEV3n*J);9%|S=Be8~_l7lqjh+UHy6b|_Wpl(oQ4Vx1 zP~?DTR=EMyBvQVoYc}~}&*Dgd9!ir^!U_X*YqaUptLRfE=T3^UL%hkiiAISRCBT#k zoIFK|djViz#%6oim0h=$^H`QBN7DyCBvR1fW%?ki{Wy!S&JcEseTN1MJFlhR^7@V$ zbBm94{)C)y;*Inj3lv?Y6qhMdM@8uA67>^Orv;W=-X+Nfk}&0>HbvP!w-u;JTeNq$ zm5_nZ=kgG0u=`vdc$K)d&+T(5=3iEs-uDRcJ+wkMzE;at0UvIkOE>9~;P^`K1di#| zbbh6GC&BmS^dEDmHgO*HN4)x(UzY@5S9(Jh4V(oxjq&bnuBO)x>81sb#O-r!7T4nq zm%)%2?wFOLqdac}AOFt*aP2-Zm3;on2((oeX*j)<;xR?KfG}wGexE#5I@!R}<*D)| zeJ*&2P9D9<&!;DPu$++Uzf+FF3Da*@1VPZ@wpWeRR9>Jb&9DK(&^h{;?|@sOIe49- zgA{9YO?>P>P!8RPE1TQ(hN4jzFAGm&z0#jZU*(cByv~FAIY|Rntaa4 zL9-;A+n_E4$+=|9tSBCeuS~kpZBH;tE->!&A&K(Gq+R%!tOzptJh16*gQT9LGX!VS zXQp8H3>?Z8HOT8Zki>)cVacY`rsEVpo%A5g3_ic*YxBrqcHnd3$mS&rIe3avoS;Z0 z710B!HFdx{QQ&nG`j=v6aj!UvJf!nUfvjga`9OSqw=j3cGqP@iY(SYJZZv>%zFbT(Z5JyPq$O7ejPIGQNM4T@Z&B4Xcc3}5$) zgWzb zcdMpn-d$O}wq?On@m9^_@apg^?p{&7_7>?TeX{$)Y*i$c7g#qf#yckvM&-Tk=wQO} zBjI_0*>pF#=MPC9O8lgHEzWm->QnJHRTaIAGR^crXnMaVJf2r+oZSXMITAL^meo?g%=+sr%YUBDO8GE{xqt{?!HFbI!COlcHJp%kaYX6n*~%6O?VAy6u(t4Al zim0sCXcSPen!zz|nV~RL8qc2p?wVjNOPtny%^){0sE%pIsWiB1QblFLNI@J&AA&<= zpfgK_{Vu2^6Xm7CFT8aXoN^ib*yIaYJEiI>*yh9mAzq0a>PRbpj{O#B6kxG3O6>Q$ z44#PRv*P*P>F!fx`6QBVA_TbL@myy*L4xlHL zqK6`nsE8UV@?AmTR8|cQF5);kq{1@U(LDtCjlh2*>Ge3UC=P^6zsoIoEI9^UN#Qz- zv17}NRt-v<6eXSjUA9r(3AMp{c~=)@LVZ@fVhEno*zQ)!IY(~5ZRk$wbNSq>Q&#F*gSkiCBz~hhMWH+I*~G1uw$sme z&v^OaE!-Wn-h8>KfFng|%u8CQSUIi3_^ENCn1BlQr5> zA@tjlR}m!G!30!@m>7F_K~iCH0P?J5(Yj0OX34#I%@WKx>oZp8V0GY5ZqdSimny)| zL-#$jZl~D5(PjHIdGrdd2~z!2{U4B4S<%7{u|ZxaG5}!#dn?*@Icufc*u7;fm|4l$ ziI*gER=(^MY)HL^RWE(AgwR^L2i$?H0`_T^2PTWkK~kOcS?$>q5(6SB(ftT8Iwpvs z0}qhw*-10?p>gUpnVu@D<6!H?yK^zbiOCmb3oxyU{kM+QUpVMumfzMvcRKcv)gm~rlP0OlkYb(d z?z9^#hV1oSkO+uH*~s>W$^?|nZ8|yZ0qnH(up!T;d)nH zTQ>cPe+PL0?w?h_Dq17mp|~xJ_gogF>!mXmRcLm37Z^uE&O44*Hfed`DS)`59(#%rLtzzta-l$8Rbuv}Ou*S|!FnVS0DYW5~ux z4osWBa#A|fZX^eq0@~m@=9-R|+@#BS$$?8cfEss62BdAkalICsH+9gh!YyNuhYf=< z765*(um+BXnRl$GSOJ(iKs81Mm^a~t?FCUJ`z4D=L2kvM>a9bR0=TUUsE8#0htLNO zi^n4c7p18_XCVS!ugz9P&B7X>>wd?%d1Afl6NsB8`Tu!&V3Pkw+XJurr3L0e$6jps zkslHt&_V0z4ql&2iQ4pY1L;JXy)x85LW6xVa{{dlQU*2o#O zx{f@Y^U+UuADAUe+T*H8y&C2 zy?sVIxYN49aatwX34IAVNQ?H6yo6V$J@>WZx1u91YWrMPOSjIf(?0q}=R1u*%~^~i zabXD}yr$w?8Sh4a?`$}ZMN6YRaO}xH6n!)QhZnWG==OQN5*%6aFPMjQ(`$Jces-#I z9~Tp9N+HV_0!;9!@LPD9j_S4L5T+`n@wz);fOdcr7_Yb!e034D)_~EH>8E7X?iJvi zIj?ytZldo5ul1O~=vPX&@lJYf{|3;4Lf&T+e^pS97Xn%@jD|#ay|BQ}#3gn_j$K8-D}tQg zUvxi428<$?v~y(TD-##Xv@n;pP>M|yNf_w-fyESw+!7_KuYQ+<-gu0g<5;)E$xHFS z@zQCm{mgbn@i3FU|6%Npea*H{!|ZzRP1z-^V(;m!va_aN^y%K*QtmL;K!+! z?c?TfR!_>NZ_AqbD|i<)m?;RggwyT^ubEUum(5rQwsjs|@7)1F*~i`F+vv7_;p)i` z=mIdi&XN<5_MXMPORA;M{FVo9flz6-U#;ieIcq)6YwERExs@|_(f6g;W%U53fR33J z?FaQ0@;3QRuWD(_f_iNS%#-C?>shNh?RN#(-R}q9aff|^I1RXPt1LcozqplOOXIGi z(e1#?!!|ka!fg&8U{*v! z?ENR4{nh^Osm+*Jo!GID>>kRtabnp3SzLp>$0d{k;^~L5@I5jR2?mXUbIf5kmWzUL zxDVm;HqZFijKOov?0e=B$N3Z*VpjM(^VV&fk_*u_plwg}Zw70pijz$@2PaM*__s9G z-=M_*%K|W&5QI;Uq%$NTq?G=2`u_Q_#Tp|HZNtI-odQ?~#+cMj9 zH!~K7k}rXk1zc;&FQ>%^1Bus8s^HrxzSWX)x<``inDEHK=>sSGbQ+NHDF^r~qRjESP2< z+c+!0Q2u^yD(gB2HfF5zx)GSlH831y?9Opryg8(DW}~nK+J`hqA_Zs0oBfG}={hz@ zu)h1Pv>&=yyI>f+8XG4!y6v8#*K~%Y1)Wu#2)VlGzOXxF{{rlqV(j`9$HCK0bfVlK zHxHMpb|^5V=pc~W!&R6Qj$y&&vmedXbB~pJT~Og2KriCr)Ztb@tEIiBT`gEzE;?JiSd?ASm+XPKwE_r6r_oN64$nY(N_RAf7$T_{~wcV?Kez!1VrXKh$iXgRO z>Cz5qlH~4NTff#P)9vw1lIl=7Cd%)=h3w1+pkW^9%S8EJ?x#=*oDDkH)W-~xd^;gk z7pWNyHv-;_1etYSzk?e&kxB2w*5c<&tD$8V>Zqj zZDHU=S?73=1B?y^qg-1Sw=o-1f=S=kTc;{G?GBkC`Cy;$h_IG!)#$L*6n@XUrk|VA zZ{%hJd4Rp!$5*kiLII0a9P6dbC>X+bGI8P$%QsFoE1)~SYW#?-c4B`$(8vt(!)~P% zfZ>f)L=%4tXRYtw18S6U0gc==aU#z^@H5Y97CbI&{3r2(uA27;GSP78{OQKS`tUV{M$kvy^0fe<|V2Eg*&=rc$ zd>-M z{@)@OUYSi$yM@NOK`DSC;tIB_L7G1d@@?{kqT9S6qzuV57WFR^9EB=LU0TS-&@!)f z`b0oJ-6F5#VRa+Kjv^=P>VfsK!>bIaX459$l2>{l^+=+;#~l+0%l!@V^JM?bUeDW} zdm%LtA*S0MT;a1?o-AKCB}LQB&!lz82hcsY2(qSHW`lbYu8tJ63z2JLT@b<;iv2v3 zB&okY(_Men2;QnB@Gr z&`VSIy@s3VQuQ?w>sAW8rB$$xzU`GK(51WIS2js9Ae~JwgrfJAZqd?xZU*IP&0g0y z@T?WjuU8g$)d34FG7PPvkjbh+@|3?e2%4{^D4%flK-n9Rk1HF&E7$!xCd%}JM~e}=Mva=I*FYk)okFUMcAdHU9X zeCVE%O=I~k5(o?yW}`vTAVAsoa1cCMS<@4Lw<|E%dQ1MIH_wS<&^5WOdmVq&CD37eDB@!)@IF%^23QEM-0Xb#)!~&6YEyO!>FPz zN0UW&KE$}>1L05#XvfxgW-YeG&+ckxWZ$;fEglPE1s^U%wpu3 z$FuW5SwOZ~d^6)H#afE2q9RU&Bn!?dd*$mCptlRW8r>ZIg&X?Mw&fWQeWznbnV<0x z-gf8x&PksBu+W+y=oRtcF*N1FW*1F#k*0!|#xrpCa??1_Stk&>k5R4sLOZwo9T|eQ zz}u)ijF4+Xf21W8!^WKN5(;; z3|d2*kwMwbA3NyNnGcV8|M-_bHmi^yoTI)@K6_=HCl4(&;cZHBlOl~&ghBRDSRjg> ztoM0-RHtnVOb%+%cKzeUH}Py|~2fZbaVVL@IW1^-ntJ4U2Dx zbGhAQ1@D>P2%u$wvgZb_U5=O$hG%#y|L`8)jIuW_Z#hQRIB|G-mj!N;D8&|vY@#Bb zxaayklja2MqSL_l(M`^ARtDAZ3uG6_L$Ip)Tuyr)l_3Sy2un2lgJO4<-Ey;=dn{M8 zvZkmf!mJ-y$EVpSC+9WC2+)%a+9qdF3b10*sE7lkQ{L-g>h!eIGoJSZ8odF*(3~9- zNOr4Jra;RaBRe>bm`hbLH3nP!-4W}d*vzrc6?U+)5+7s1%?S6#Z=?@fbNsv_Kw+RN zsuk{p9ER-K&@crLp$`~v4Pi_uiEoMw=?+PbPlviQNPEzOmqoWT2)k0E);U1-`yj&ebn z#>%O%5HWg6v5z9Vk+1{1?56r>&<*P7sof#H?wfo&z$j_vsKnwnc3i1 zOKC)A7tPS)o)FI0b%HlwEd z2fk-XloMMmX%;xxKq=N!B$kReFEwPh^R;oAN1Dd& z=UG`R6WTo%9^simF?;LxPmvW)42ld3P;90Y8z~Y`MPQM;iDTQy>J7;RsuG^=ysB1( zbX56NJ~giBun`zW{b2XiA7(;Uz?k@(T;=&iW?+Agp;Ki`xk~ycRDCX=cwZ3hhuVM*Bx%YWaRprtJyA;38%Qct3EU0CE7Dxk z#Lsv@+QV!1>XPi{?iY^%t89?>f_XCHz%bAn;nd_mLq3*mvVyXKU~`>f z=~A?ATfvAuF2?iFjte43xuQ(qUdW267zgV!lqW$~_%a*xf=vVw2`dn4_eH7+qkcFSNE%8q;bU(U>G z8JstO@+kD`AHvP5<~t{Tw2u_OGG@zF3nla^r8rHIlT^e`?$faLskh`kkiGVJe#d;I z8q|kucHb>-aZi-rq;~^r*xvAV*gD^vzu%`$`*+nHvgL=d|8nK+daYsp0iQc0W9CL` z)p;$Y64snz40Psg1RbgxM?CB4@ znmnpidE)Eh(=d8H)SztDteBtdc?RlAj>@v9SNNO^-Xg-e9?7!k6(mv+(u~1q zeQ`nT35FMEAN>d`5lPi73CgkVV)J6clM`oaGjz5|2FKe;pDV()*H*Yg;0Rh;W64pD zU#skbclPY3?%g2|dFYgjnN>M+rw2007jmuxn_T70E5fH^+1=Q*aHGbKO|y^wB4+Gd zZ`habX%;a*tk`~;Y;3xV$hqt}$c%^~>|Uy>%RP9C86^#`U;cn>a$=Me zTA(C@Qb4=*6e{925SH!;?+|KUTutIg{R}%H!;iH1T`g9|{dJY;qJE z*5O_&Nq2ANp9NFI(Lfm`Ozauf84_LWVDhggHPfs!<``1V`oDV})}vlq%H8f2r!M3i zbgQBtPQD^sOLlqGIjU{ukbfc$+Whz#XK>63c8}j_Tj~`BB-aF)#fR)qy6t4|By!H; zTUtgbfCRD#`D3Af`*V-_URi|%n8sxf=-Xah^LKJ}d#BX!u_^*vLSXSril$VZD6dq- z@cQRrbN5tH8I-5%lI6|NpzI06`Sk?DvY4k-~&hfPl_`K2oK#?)(11eAnU z%Z|u00Wfi#a*aWLcMe($n01y-V^jy_l^s=reZez!&4C!?TLr)3}vXefa%Wug~x;9IyWbKe<7sr8GkhB>Wpxxe%>FtZ+I2ArRJEC|1 zHABC$#rV*(*nc$S_l+z3&61^R@rELj@sdS{DlNp|eo6tmq&-wbg8XoxF^lOI^pwt~ z<9S7({BFy2-E(tA+4N1Iy498Qt3ZFY(}m&71UgJ?s#G1(JoVfi-0#xDT~0drYZf2@ z0tO$AU5_(0)y86@ZLUw|t)lGduVxF5;WelIK*B5l6@M?=W^L-SktxnAEeV6gUL)P^ zxnh3Hg6oqX$xg{Zy9!qYoeeU`n}Nr_o{XCuF)yyoh_N3SY?^8Mk57HybwHearTlLs z`jv?Q?643tnP@kWKs8#$ZS z(vBB>#Q$i)%N|iSweL+)#(2kg>k{4e%AiOcvf+3Srs+=l#PL3#T4GO|c*KWqvjjk8 zG2(3YyMS3|4B=Uuc)>NO&)qC9ir)TKEm_M>UN~{uN{)rRNTw8@P_SkZ)w1ROkTET7 z^1$u)0oB#ny4BK4YQ3;C&lE|GP_ft!XcumTHB8 z;sR+5uuWJ7bO_GSKy^3}uNb$bhA%p+)wZ8t7;z?clEeMZ z${$*Hcwu9KI;OaQ0B2^Q+*Oo9{ow8PV zN!=_d^Dahu%QHM(uSco64)WykgANDwNSZw2cy-E8h4u#?D|j*QWJZJw+Y8=)bnq9} zJ~TFfapEltgDddTRHX8H>fH^s8c$~@aW+rwaqr~b^)3(S7nVt&&QJ#7LUAdyC(IAg z%i`s|@)EadiQcD)-zqGpaRiD|>604MxwB-wZ>mr6bnIu}$}gT?=9Nlc=3EwL+tci^ z6L`$9Vha)G>mU7s_gm}s&94Y}7`jk5LT=H=x8-{vT!e#VRA>2kcHAs0JEiT9$`!t0Qr)Oeza z#+`Da92%4>jn#+op38!I>5N4c8XfYFE+;#{Er$%uqb*XkuRIwNJ);5$D_dv6(S14J zeA^7A$6x=(FG!^mLkXA>1|`DPQi=-{`IL%C@=t}n9gt#RTwW82m$71KziXnr&2y~> zK;RllggR1F1hf%yRIu6QNmuN1S-@Gd0DQE$q6e(vwj57LY%^uoj`Wi9F zqewP4Ns4xoCD zG#2n;2daAqa%)`7pvwQ^{lAdNm(24IG;xD~7EdYGQDhAjVG=&k(i9Kf4j`RG$A0d( zGW`#%m`yMlxOg02nPIZ!%E!yc;*D|QNFjqa1{tkQ(!t0y2s%#M#Lu9w1vJx-#C6Je zAkt2y+g?MRXU~Aou&ZqoZk`K>ar5rw{qE~L>k?)*Hn0=#xETtbjg8%Lk9}w%G7Xjr zD>zy7Iey%0j<92g#e}U+o8c#Mz}cgIsh?Pv1id1lV5n@ux(`jk-9;hVSes)_#!su;Bwnc?q+jF+?b6!YA$eZPp#@ z*hms5Ha-}7*=>0H^Pg`eCF(SOzMy66>=FbtZj1_;giU##o>dcLviQKK) z6Ov<^>kF1~ZpzOROg-N$I_jqn-#oQ5#Ey6x19%3DA@i|qz_8`2!j<>FZ|y~K+N=cz zzX)~?OO)3Na|NBUUQck%B9~-4-NKCw%N3m^=`#%Aw2$NI$8CQD51#FM+p<>nxQ3My zny^**>(K9+t(aB6D}F@IvCDXJVl%hR!scnD6hNAAnTkNQvrN$8gV^H z>o|$>wH_xlk%9+q#_}YtLAiP%Ho8kuMo!MAjZ7`ja}MJM`#2B4^$(1dN+j*co{cV< zO}^MA9NP*Zg-?biSAbVENY;A9LPw%rk9J6Pz|0z)8`WD!CwuU~$)U^gjnVk;R!LxVQl z6_f&k(MPF>BKI^v!RD#0qJ1iZ{O}vgNxOf`0$|LL|Fx9vnja}hQLZA_Ma!wHiwfki zJUp(`Mt*rW<`|hJRJKo7!d{GPCUCCvVv1s6I z?mwUXR^8&OaHp+HKg>!v1{K46ppBVM!8Qk7UR6jIadH@pA84 zEU#?{vbbej-u(2JDS6&zO#JM^go`AJ9VVPO!|t#JCUPkSWcy}N5lF*SugwZ{H!EKNE$#do{sZRq-3=h-pyfO!9$E=h^o`8lz?0^`mo*88X| z&Gn8W%ZF?O3zI40INXH)-rvPKxBEqg9w)YT8S=dc@>h&a3_H2|eY$3%8LBH4#`78^ zz3w__#;n{M*y}zn$_CeuEdm_3=9V)vE1av>cCN4{=XYBBgMn^82{_*_@hdnvoK%rs zb&A}Rq|G=X*&)e-I)HlZp76ZTICU?#TKZQ#(7E08(yKno3pE||hCL?Wmg98KJ-D#T zyOQjpyMZgDgSU#4CP>$$LeH4{!6h?pd)3*zSFwEVL%(B-5+;dZ93T24Mm5Tek|Baf zRQgx{@FQ!-ozo(RA#J0a$n%>Wc*F`I6P8}>ES_aHa9u>2LnPCQeJjT- zOx6RG0*Jx(QV|!wZsZJIMqyrm6V!gR^5aCSs>(T%G>jUqs z`cCFszs~;oil4>QB1JY(5yfxb5I6DjCGEm2-x75PuR)DL z1MEyyD2KWL1&)u8UA5|4_b*8|vu7>}>wZ!dn_lpbl zA-Dzla9NmcSs45rKelx?OO5hkq$RS~c~&xGVouA=Z;z#uofA7#7&_S@J7qj?yO&Wv zRna+$81HtdK+M#%%M(IZPlqZq@9vOPQN0$Ace~xvq$qFjQ)gzVvlpf)vB#8S1qxeO z4N^t+LeP-&$ZMyiT2rn#t<}mvqLQh3LZ6Z?oAOY6N18nwDdtj?Yh(t_mRYA%j$PFu z9;vMhJ;bdJ9bzRvsBbL@_fQTv=+FMVn546_Ih;6Kzs$nsD4-M&0?MNzu;oXTAaW`c zCnD(^77ZTsHpn*z9e@HbY*AQAPRXE%3RL@q=FPX10>?*OS-1( zeRL&mh5qs8GxqoK+lsJ3B4p@8JLB!mo?H8^b} z+IS~DTeKa*&F;^5eJ&ls+d=qR>Rk|uouR6~e0}I$?nw7MuHfk5VCVI>nvPlEM1A=F zdh1A!)9M_C*o}pwz79C@F?%Feu)$x~2|bug)kigzs{HWJL@CN1&K?L5>(Y!VP^rM$ zk_PRh$|k3HfuuPzWdqRpILNpiJuto$7gkGZpSR4w7{Vtz;qwFhf3hw#eMQo48A?zQ zM_4%63DwN7F~ayy;^c9+V_>aq0+4+$!D^QGPcQk;mtB5fvHBDGmfU10l)bjjqr?qa z4ps$3@qls1Am7Km#=*WEC2sfTHLCkvc6t;*qV+-V3*P&Jkh{B;Zlc!$i?`!|Vbk(| zDJot#V6~)dE@me`{y^TTwq|;AT4lq)-E=0j9X2?6;pg)TRk{qI)IJdsCn}R9`E|{X&=Tf8Y7nbl+{aa;43yqhS3HJz1o|S4>-A%7$0)5CvNN+4vSGzg_He&>g z2|;X-VB@u{05kDHQ0pOL-ue`}q@5!xo!CvEX|b8vLP1jwl0ZcucX1kTXHX^Y3&~C4 zCaEqHDg5aR-f^O?>HDGL$B}Ylh*e>Oj3I7!*bqBp=uiIUuytbNiyIs#-uf~mJ4Ojm z3xPR$m0LN}sC+6(c|ZZ@q$_M|Z;|aYw$Iq$50-7Y#K37(?YUQ#j;^88(&gMD4d09&D^tC*B=1h?;$nUGkUP^5aIPuOeEn+U{8`EAZMOsrN47 z-V`Ihmm_oyMoON>w``)u>2(eD)z#2i^4T zY7idKf)u;X#7tT-DNbEE6KiY?KF|EB`Rxv%7qwiVQ7^pZGiSfQZzKX-ViZ)@w?khg>gzPXO`o)~pGt8Tq z=ntmGlGIlwep6zx<=IOqAeSwNirD3m%f(JDUxY`7r9kzb?uxL0lPfA!ZxQX#)P!R{ z7F~~f&P+_G&J%p9Oq!DHxn@DLq@7;6^zW*>bC5P2G53WmSJcE$gN{E5A)Rvjn=Urt z>jlN~!obLYVR^-O{p$=6_XfUAg77)+$4l7NViR3%Gy}dYgDGz<2h*723T)Qdt}) zEeLIgdPUAS+OI5SfyMec@(n9&OpKnv^&X3l#ED(63_g-px>jiFz0pIEFgBH51+UqV zB~S*fcXd6$pbi{NvNdFYQMR2n0`ImL-lZp)wS=inv9xf_8cs$z^-p6Xh5*J|XIu3B2STGzN{4#I1kc zHDS7Yr`$v>x;#}_ZV zuOx;S!`my1=jjILAdPcC&-CzZTjfzL)R=B!r**zl*1bh$79pwAe)Bt$$WDYf@qR{c zAwn`K1!PdCVmF0{&`rVE;1;usmZICefYYGJo?Xh#qSq;qq&8QSD*z2-@QU(JlxNcW z#SM}R&|CpiUma}p+J>shZdm5IJG9fQ;@lUKBDr}MpoLE6I5J;wLj>1+N>G*WFz0eb5jvjCFMfUvd zUdX1Dp}|73Nesp@_j+NeUL4mNgL#aoa-Fiu^7{mQt52>UU8mklsbgPLa<(S zvtD8G5u>`4odo&l_~k3s+!Smih!Z=!k~FUpb-T2CZle5=tW%x=ExnNJycXn# z4od_dcs-Csa|&qP4$lPMBiU}d(%~>J>1g1@IGG(!YE)U!;$c8?Gk%G;@kYkYVmtFY3qR|0l{p*e&eBBxrPV1~^Sm;lL*CD)n&#P5? zUUl88cMcw{_r30QQ*zI1{DL>;Cu-~We_#Q|>y?!{p(;uvmj2Fca{RvLm$S*=o!I19 zGjT9W@KZ|h1x324hzIm0?&t2S<+>Uv7Vm+7T?&Uybc5t@NSzwr=2Q3&Sf8P|0_*iI z1FKrKtW;|(V0G{^YNKa{ zxS8mp1V@6~Y11qx0!-f6L?{Hh#<}-e!W2l0yy+DgmIWqHKKJy3dhdrJ@Oxp?g>qh} z*y#4fd0V&_f!Hg1c2>|K^Dnw)aw^@gIw-~M{E5ZbyE00kC%QgQgRl->M_)ofv+*EiB$tO0kV1Td4>j6xTq*mP}1GJx~I6 z(i3WXFgdA4nm(h6e@9xa#twqHZv8G-J+8TpPYvLx5Hb!@iWNd8{L2OM1M8wur`1~w zWuSHcu!fY*%%%$dp)Ne`jhL_1FW&a{xa%uiSz9164hVSRBUnLVVt&bozm%GV#+iTo z_I0wsiG{{K3!$-tQb5}5HY%c<+zAX-J zzvsWg{fe-LuA=XTB?R__REfp`HnxD|Mb+`bn%aGU6-cQ6=3VdcH-n^V@rELj;lv=R zv;fI|N>M{!oA#Wl_`poh)WK(}>JDzlMFSeNI69?-_W{KnaQ-&|8K|y2w4Q@+ZVI>Bx#DW7=XDI~`_@1C5O!*k5@+as_ zGp#8c%>Cn}*Bx_tAjQW^?^Q)#3D#lh48}?Z2D0fojuA-MraVJ)+iUZT>oao&I()%B z1#SbN-U{2&zzc2}R(oFXhNk9w!yb?vKeRIkrUD6ZBv= zl_@}YfZX?b=(tibGtx7kyJZ#(AED0x5ym)}HYU`u#W-86j`sSW#r|~Ncg^59{1d;Q zk+V*0vfQ;;?gmOxPmzmMM2Y(FjKeA_6TzV9`PTOEkNnm#K$V+&X8>?ZK}kl&cI+p~zOmTe_P?nP9Lrc1JQcB~+aKFK`}tp=;5RkZ21LGJSp4Um6}o!<_6LyPXa zkb9f`x~+?D;~Lm>nN0w)X(sz#)c(QM+^GY$&7yfXNb;nis#1y5`df-Dgis!(040=3 zMXY!0f@G8vqG!Bkyd?K!uAhOLIi=d=fkzJm&+)gZGQPWmzAkQ+wfom=?+52_vZvh- zHiSPD7lt1aU4^y@8z&b)l<_ce99nF`#8AS)G-h8Q44REus9aur`Hc0}=oJwl3_Cnb z$v0+IHqnW5kv#POc6fu3G9mOEpRW^`#m5Kw$^S!QUm0UI*FvytqZC^)EFRG)Y=)@P z1sd|TBg0l`F|TP4`CO;ZL}vZb1icrFRUFXjIXrK}?q_g$uWsK3PI2jDV$_4_xVES5l*Z5VLA}erC+jI(<>{LOwr3B0O}emv2-z%`@c^>h!{E z#%A`ZB3vd3BK*v>d&z};4weNVJ2YK z_Cpwg&btHnLsQ|uTGxFr;N)`?$<<#T$(s0|a*Ks`=2vh&_si$ThNki2)IIJ;UE_Ie zfvwP3Eibf3Snu{Jl=9u;WYHxviiM@BHHxu+*@jsTn*wytzh4$d!%aKpbF6UUyn;Jx zuJ_%slo&eiUk2)DaFF8562mmXHSnSINc#OSf#sRFnV&@$(T7Ba`Tc&-DGuJH!n-QE z-!GpW0{;t+JcN;czb2ry#XF#uM6?6$Z*tsbXe`nZzXM~2H^T&_@BQOD|6sOYCU(B} zcXDD9xoz>BFH(wHid>)~4lX<|Ms5kz;mHEsp5X2fY%qXs{&-F`jU^H~OnGnN?skJP z9bTR*=yyTG5$VR!{cj2!Bzt0}i0+H8ujlHPxgvXGze|Ru*P|5{Dp`PBpoVDr45+#{$)_rJXj2U*#-=E(hsH`=f!b)B)W5x&ypu3ll~C9b71M> z07d#ZZI}>OED`=96dg3;_U^rgUz#HK(F%*~%^pheh$0=(PF-F( zvwLpG%m&FyK^x?7g8XZyF>FxnT1)qkds394iE?axi*(02OsLa`;NGX5&J|?C{}36) zo%nj+4liA7SoGADL0LqH*JCsl^Cl7qRLy5Yw}wko_q}E$P{tBRl&ctyiVi&Jh6pT~ zuER(xwjfFlD+f@6i0YJYea&_SIl6EE zs8@b;A7}Gjtkf#C@h8EvzF`(oC;$E0-^p<&4pTN+=%YGHQA3gQR75@J;=%@rL4HuO z(?f^$(L;Lu!kfUT0RkrB^};|*9zQf~4F`hLXi-(sXvFOcJq?*tF#-L49nx$NQT;AO z;x>7?qz`xo8zi5oN>#1&&Ud4})fk=u9POE!QhLL?XWnWI*PSK>Ubp0jR1kCoW33eC zHj?Gu_A=RLTbDcvZ~j_iIQ~+(0*sHdvDH}XvayK^n?VJ?7pv$%^*-;k>1J4I)~zlh z+npFzM=fBrn^J(oIGc)i>V2G>7nm4w6Ts7DHbmhXn{&T*?YB z)X`hLf!3u1tN`b{VNHfY1MK`zBgL&BDzQwjr(rSPVAv$ev0h|gpxc#cGjCOd zpOR(FJg4l1mNKPuitko$T?b$zSA^MzNDi`O$D1n_9}K_Aj~AqJmgu56jc#WZtHBgX zl;0wod^?7m3-lo-9HLd;C(9s^);L=)%hKS-bcV1~Q-cG|9_}FwX=)+hXGE1QD zZ+|k^y5@k5v~l8n8AF8vZrPv7@6ftu-Z9?!!8;@=%4(w57@LX|cva3se#G^@101UT zE?dHS#kKTm>8L?r`#OylKDK|t=&lSY@Fv!N9R`QUQ4NY{$PbPT%oQK>u2UMbBbQA< z(VZ5Qp=tBPn#6Gt-=@En-7_8C^BK)u%wNfuDnfPby^hCzX6yEMdylUtJ13C}3nNlb zDd3y9n~JEGKIE!~0hxE9BTWx2*oY74w1Y!p^VR!oW^ z8^|h7=gj<&X7ZG?Go)VIL*mqjmHomtiGh@amTL}5F7Va{e8ROqoYCJ)^*LHxqxpj&G2R5gS5>xlm?dF6DFIoPkTa(XjlCft(Wo%0x)8#`KL zj{zmauJTwF zQ>5yX?c_Ggj9vyLLf#$RLv$GMLn0Slv+S&*z~8P9G10V{yz5&}q`m&k_(aIiJ=TV+_KTcMzr?*JsnE+DKcu($Ci!R3Q6!aqOzO3#+8#M-Z{oCii<(aWHd+Ya4krnK)<;0%83=3>+rW6|~5>G|!Q#D9h78o1S=LcO} z*d3xn51+Xb5FLHv;rM+nC0GID`E3qAO4e^YYV87cTKky6se2#xPuNUix8&F~Y(BG- zTQcKBKsL}l?4h%KBLx?2W_7X8<7Tne_W3{?XJJ-7Lxjz#AOGr_V2)X2ym2chmgKS% z8BVN@fSPTP3-}17I7E^CR0L-4$8jEb>BYI+C@ROVXd&*h^Hh4U`ja?&-nt?F=(hYK z*hRhE7|*lh?%a0z6s;Hk`6Bd{{tN#28~(@-$DcLuv&5|metw+e*Q)KFuNNESe{N92 zl@00=x4$*0|C-~+MzajP#w?&_sXK;$H}r?G7_DFatG*wZ!E-iT{XJ6a#EzT?7I0{y z6gMbx4V!5hl&$lyI;qJc-@9<(Kt)o&OQL5My>aUP`F$?wUQnL$l;1qPPI*PrDeM!L zxK;S<3I2k233{t>Dg(}FQZ!%8TNipkR8HRrZE@G(4-+HwdfzX+`&>%-*}x{gE+{!j zho98|oSn`56p!)%Or-k4+W=K8rbdpK=-CN8=Gn84i`OY`sE_k{Bs->?lB+O!-#$;T zNtQQDu8U7h9Z|J82sXom%6N8o46%|>6Z<}HIcc3gM?1h@)F~I&TE7VXG|MST=toz%)RGW5&F7_v6 zE3_vAO+V|jrNYOgMSBr=MEc!MNOJw~T$LybqHMVFv4`rhZJk(S#5?TqRryFqQuFj4mxAsB3Yw>CBT?$TI*@Y{C@Mr4}W z*kB!`6i}(XkBT@G+DvyhX;y*jF>_b|(7C*as! zi#0D;^zYUoZZ>5|PVC!d2#A~d#FhfV&22fP^kPycaQPN^HM&*v@6N4#GncdrTjZ5I ztk{m{T^_G4vZDdWFnF`uDJtaOT4tCfOY{d*V@c{OBUylBagZl(FQov)sou+28aS{D>KjZ9xRXtqvW&`L91}ecg#U$N{!$)NnlBIxB?4d{=&w;QkJ0!=rknE%erm{r&d09&)-ugo^+oP`J~qZGvyDKa@&vVEF7)^l@b)ob-W<@9pSwy7t{Lr56!buXpk zIG97Xms<{b45jJ@bzxw<>sCz^$$$`1E$puAwNE&C0`%PVxunrMW~BN<5tgw@wgIR* zp{|Y7F4TvZGU|YU5$3m#{u<2}VLU9H<8!x6v@X4OT9-cqC&1N3Jt5b{m@)8~XFII( zW3tL^6@_%cNCeOf{KN)%A?G>=E!P6hNmpGaT`sAB{-NhQ98SeL>UDRNK|c9Q}pHeN1RNRA3hQA&}cR74a1*tC`f$UdDAx>xj64Dldsl%OEA zi61H0rm6{F4|G#Kq*t0R!Io;NbW+$>RhD=?WYS*HR7=-zo{Fp8DLMeJE@(5GxkghH!Kp*_k4M<~ZrJE(U>_2m(EyJp)nHKVd6HPLo*o5fstm+M2p_zUZchex7+hDj*aIrmWjJAWIM;* zn87Q`AKFZUvP*$hD<$fFms=qCYNR{p9U$m>J>q$~ZVq~rKf?JjGh)5GU4ESWbs6xm zn;H@VJFKKvUgp&+N51x@pDL54V2xcV7jM)7^5JZ*XpOc~w6sQA98jlhr%k_nit>nN zcW|lNX}Nuw5&7*Lx8F`SqwmhI8b2bdUo!M%TS&F7lmfK$Mk)eX%O1(v=jqD%4eF=R zB)gmFa{X3&w#teiyRU_N(Ww0#vd3bOAS|U7!z}h6X2yo+65TTEmXS_t12Hs#Y>+ex z+dXu~sL@gQl46`_t%p6d56hLXZTVSFYKP02mw$+z_r#P>-+VUHEH1i#CS5^xOd{nL z-@rmj0TdAVR79bu*gwfXO>h~uCtZ?*p_hUWh`U2?h#x??^wmZ6+I+frzCm8)b2E#%)Ne%}{hhkhe1=*SC< zYTo~;6PB#s6WVs~U(Yeili!{0K1G&0@dhg0LeOob6!8>UM@1~zBa8;p&KB*GG)_LJ zBNQ1K(2d(o$~g`YR#q~?@~zF1%`jM|7A1fsGA-@yb;na?pFA zAC>tNGd0`2@N_d&u|OmftF?`v*Wp#>6$k2f1O(d*xRJCc>+{27mKkup?-vGJhqqZd zeVo?o4U{y46CS<{d^giwLae?<1SHCrO@l_pJ){VFO|)<$!}KAK{hRoybTc$Zxdx6I zJv2r{_fMxdi{6PTSGl4VZYO`uf+o6D+AKK&JuGxheAB-7mOPt|lCJeI?dbKv6+U*; zfcD4L@OZKHZ1guLzVnV*H2uEy_bud8CstUWTj1dur2rC$OH@QBH^%#*Y=;8z>T=5n9cx{AC z@&~T?s_XEo^KJk*Z1PPD0)(Wycgjto20SPD8JR%nmP*Ge_Nf3pHBvC8La-da5CvbG z_ve5~bGG^3uKQ;4-i*9l6A^^!!~gXWSe|1E?g0W&ENKNFYpcW zXrRq|Ob+-h*)=ik{VnfjzyIk^m(*(;Nww>e8YD{4$Q%Ba(KmV&oo*uyzZ5vJnwmNA*OKEX{-6k zL3$z1x0j2*O*9YPzUc4;b`aUBfQ&_Y;gaMxTfRF?Hy&G%W2-P3#dVC}VkLSg&y08f z$(pFqX?N)i)QiS-0}m3~ytTre+-!(f)hl$X1ZN320)59>uGo$X#wB#f30An6P`qe& znROAZ)8c}mU>3`i4DuFkQW*Aptm548)t!-D2s=>6CqdXdJ7a6>(?L4r^uY>LdSO_8-8 zd5w|%^B6Z*@R;a@U@F&0m(RW;tl(e>5scv|0ro*{rTBwj1_v z@Ee$>gY&kEGxj(ez}qNaZ3H7T&L(%v{x?@MrmDVD{x=dm6sDZl<4ty0U}_VkNT5g@ z#>8{{s$DO5-}dU7f6u>$gI;BD6bd(qvc`+b^ZDA?3XA7=jL#7>EIxk!oBwGI3#U~@ z3=n9hF|z@__h-p&=q0??!`v($+rW=|A&RXKa71uxI506oV2Dlg#Lts{m_VDQMfVSU z&yuL2q=gfg96$@7K^kZSrC3jqSSkY3VG`wgRNGWlKpd3DOB8_YkC}BwVyoaE?h&&) zXCqM9O=^d$nSt`oC6_Sk#1@8JkwoYh-5h)^u|X2 zPXkkw(ErH&s-jcAi~iGxf5Sa*x4*GdBj!EzyI^(7Uf9_fRo)Z7nRn#iV zadLzefM2vS9R_4p0G^QZ{l~S|i3_iYH=7Ho2q%nV?l{Yx^jYFURV^Mevx+6Ph1nuEqX{&pGd zr*Fb(1|n9J{QP|@6}FHH8EfjZqIFa1RaX@?Q?QZ!gj!3dDl!x&zJ5uP9dIt>eDG0v zofw!S;dsrw>*Er2uxTisesXr-!)+*>w(R4=8@s>vveBZvwe45$(i^`t&5nCb>`w{7 z6cKa*5qshFE3c(~ecjui|8U2G+ixe#A9Ua3of3L-e&=gieyTJw4r=N|%dM>%7Z4 zHbw0&MVlwa(yx2(6_w8!j%-wIkgn!G@)=+vsgkY-U!{#Q9M7XMBV8msrqn^T z)N;@JihL5Ux$S+3)bcj?9tqzB6d^jdT#rhB`(wu@TRGyXJo#yNTiL7pQ@xMT$o%BQ zgi~~u6T2ne^|D@XD??Ee zde8k45H2rfE=UrhwMv{+YeAFiIutJy^R9$JJDdH--8f)l9e%ry`^fLb$wH`J$@=!u?8A`_`-t z?F_>@vcEkcm!Tk{D5BT*G<`BOo?qj(K{^oTFdydWQM4W$ zXy2Z`-cftfe!T4!b8tM(Rp-Q}fdvACyyZYtkfAv2+8>@VS#v)qd74&$27zZzbNYk!f&}~YH|29>qB*w<3x!HEbx5UEu3|o! zIp9?kxZ88P=b!=!%%3A-R&_SVdG5$VH7l))sSf z!Liq|B-wBV9vhMuiwnsHs$=R7dY5bY+>*JrDLtD%HhT|?y~c)pSYD_^^Uf)}jW2q) zkdB{2R?IF1MUfQxdPEP22w%0eZxhk#E0wkzO+iL|#S*?KTrk4nc1uV&g)xsBPp>Wa6Y+ z0V;DZ<1d|r{dny$MZ#iGHaslc6`091f*N=lzsUQnVky6iI%)rE5KDidZO`4(o$R}h z{kgwTzHh>71>xp>V_c>oCr$#ebeUm9?uI8c0b$&~OHtzoI!DwUzNT7oLs2*#xI7dW zMCp(_Jm6m*g6EEpe`5p8?TcrYyTzUGDC83 zcUO4U%ZmfMpHFYV@(^v`Z8a=x{|cL~oZR5^4fB{BOPFh`N8OYZP+M*hpp-YP8Y2e% zk%3~+T^Ek*vnfFEj{iz=y<&ULDznI zs}olt95C^S_7F@ZL6;M;-LTk2>z@GxDcw?CIJSL345b^Ko+TkVF_16;hb_{!$3ilRt@ggeMZr@2nmLj|S|GR~Nz#5UC;btSZs7KGcFrnG=AZkT@B z8#Ui!;VW#~#R=0B`ahU8O=^UW=#J=jblMBH=mDvbQAI5}f&qD}EFxA1JYU8B7rjeo zub1|JrN*sv_D$~^&^6nqIxfjlSBrs?8`{XuimJt(f(D-|UiIwF<6{tQfsIW#;CK-` zY<}Guwa~m4!D-t|S&AfDWu2rU;+mmArZ=o@(`3aQ)oWZ8l6oH89FQRcKXPtq$(hZt^EAbPNMwC1$5n@$Jm=pw2rP`~5DdHn)3LYgK}u{q$n zViwSYAuU)*_#==Q&>(k8Kg4%!opEi!jJbJv=4Xyr89ALDGZvO)6SjvRoKF}{&0~+p z$LS##r!5|aHTr0?l0q;m2|9_0Ef#`A9(Dm8aP5}vkRN}=)*&5vgtq;(0uZ*{c#KzZ z1H_86f1YhFN8_}Hg+*ZIh9_34CGxYpE`{cS%Z<$i1>reRMy06`ou{zO4GY{1lq$%w zoe^?Pp(meg15&46QHsywuzrQ+vNvkc)c8FP18xkEU+Rz@qfwP6hs5X|q;i1cAIK4| zk`FvE3^#VvS%IKuS`3zhW5f?fp4^G@W&h#oWi%-rublY1g|{bDlL;w!q_A_p{eAPVkuc z&r`nBVqPTawDS;4*<*UZZXaY(E*9@{1;G}t({x4DZ9y-&CNxWGyW?r=cvy`KTfe}@ z>p0>PPR~k?2!ZRYXUk+{Tcg!+*n8D0IRkJ7tP%YEa}I47>_s6PtLH%d}K9xSey`x zyKO|w=*u{vXX0$->wW>oNvCA^N4seqw{8z7PBeo;+NhYzPJ-D^(8WmORT5Rge=xgh zMmN};dg&#}(wKp$A^FGNOZf@1-68iCRUw(dS;7TZ6nCh5WD|cWc*uCDMOG|68&gRy z<>Ov$KwIdZkVDD^&|@TA5^WDejUGkVwhzF}Zfb{*KmGjGUrjPXCZ_uj8ajRw4f*0x z&w4e%q!9EQZ3#%nz{7gp-(}syeBNE`f1KOf>a>vvkMHF){bRK; z-+Jv5D1l#9?3;H_-X3$>GoRd~stsM`(;gEiEK;9P7lahk-SWENYVt9^0A$5MZjQG> z`s8T0yqGTY-VC~oj{_a{K3Io(i!+gar~UD6xa04W?@!CxJX=bl<+m!F%69xklxWC1#vRd?s-bzK+y(d7`yyShz-1&2sAm4 zjjoMli33vB{ii)M+Oi77A31h*Sg11>|rIxX>!Ed)G`Ua4qF$34o#; z$xyKU$FX=AJqTQ}G5T6ZZga!NI_di@zcqTPuWfJkr9ZV)b>hU+OOlDW_aMPMAn3a& z8vpRsyV8o7HYM^g)RMZFN=26?%fi7QRck6GrBH%|akt&fjYv&>WS7#A4n!tWt$H=x zBUnDUjQZsJpg33Kway`401>$9(0W%=&^vEX@gAaR^Ow*Kk7Yhx=PhASl z4gR{Q^tdWd_487eo#IHEwuE6samCVT!}ANaUySw?{yQIh zGpsP$C_~LibwoayGhXBZTrB`P0*-w1NAAtZ0yyt7^KY;H*T4T`g>B0eHB5^5a`-_g zwQLc@3HQxD;9kYcqHdE}L3hXms*J30%L>ZVZ)ufVcqLIMd{Urs>nG>g%ftYEOrCU!h z=>)x+h-P3<~ z(L2-)L#ok*Ul1jS!!R!vVjI`63;s8*rUD+b)#J=QBn~;k$y!W!@bkzu=BexL5}0V;*(h;Lt1< z4!Yz22PMfU#M~TQ0!bqSuLkzQX_c#^iURQ?Kq9DUf}59Cj~kw+{w^()dY26bkQ_BHs#r>+_5tXGo4fbL>B##^IFOQw2B}uLE^q_aZ~oe9d`|qmFnBfFv>rksV2_g7&U;7 z+~hsn?J^dUZ0B9}vyo(Xll@UgX7`eZlnqjN|9PF{h&(*T#88`iu`nA90$2r43Q6K)UcTnJ3}bwXED>w%dc6goa=%wxK;Y= ztC)4JtzgLa0r6Ff>qgfWpengG>+{I;08PA?E<6wFKAvY`?e>Sx}FllS1f z!SugHd~KtP(XA=@{M`@eIBsr@^ZsSf-Z47Zl14De1iivgyr-#gYX!bG9p6yAXOEtq zXM)AzL938$@k+<;6eMl>H*X^(+J3QnA)Uz$5>8w@TV(=?B7y0C=N_PM9w)oW9|r6ew2_;S37a^^^D2rA0`-JL->%0MY{GU^4JD4 z$$WJSiJI5LE*~rVf-2;l5Srhnst@Xqf<#|>04{=R$v*Ghf~$}r9Cm5-tQGW%Fmsnf zLXWEcny4dF!87^Pr=OS1pu18m>=a-+TN_;-0zHMevN3$f03sf=L$*(;V?+l#ZbpvS zXmB{qkrN{R>D&h|dOx9#$qvtR2dy!n!$9`PF6x3?Tu8Uy2n7s4MK7FXIN2(2wKqO*EREgk>SkK(6Yv_2V?~VT?-nq!P?3 zf?iI<;*jeb2#%{xs!HW2J#ByYt$xauPP7IQmacR3K2C_3n6at1TVh0n{BK`vqf=k7 zf)?oD8HJJ@f&l`AOelbWT1o6jzfF&@NukuzPf(rjy-f`a?sHoLo#SAj8Add9%xfo4Bdh9Hj{`eKTaC(LU1HGNd zeE@euqJu&ABsY5()!BrZFTgW;!@fXb$Bg^C-CaRO%p81eO*NgzZ5iamD`U_u8f9;G z5lk6DmlCmSLz^W7VOPQur)36fP?|wsjJq0y3XwNs*rkl>U^>aG-aB|YDlvFCu#W2W z=mlmUly+GX(g*HLcM!6OX{uCx(9fF2&zxSwYk?#|x4wqWzNX23fU<9a7W2lALyI6DZxsWCv+L#tuJ%5M%t~(MmrG##vQ$VvWCHj(J zz#RmbNu=1l9(h_?pe~o=9k!A?c=zcBWeY5#w8~3@Zo$~FVn=L@wmi5!ChNm)Y%EYV z5lL5nKH0o!h;@dua04ecKshG3`=YP*uBcJ9O}Y)zK$=x6$uU*;b`l;_VUM&>u4#^+6V=zU74L+xRZ}^P+KwCM^V|(pw7;#RP zVxsiaJ8zo@@t@X=a$P!gA+$Pu3oKMfrlT~x2=y#FWuM}eVkr>I{ z`lG7n7`Shd%b^#No=D=f(JOg|<5(T6L6f=PVd3+b(PMstz>LU=Dx>QjKq zUBOcwwT9VEFck#oJ;W{vTrXWg8~$yJLBX;L(E+juWRkDbI*&twA(syCN)nVF6xZpc zvJ>J?>LeW7;BzjdRe4;ZlMF~U_>|ACr;+~cv~Qo{w)ZlAi#!FY?YqenKE3wl$F~y{ z#_E`wBjFjL6bp_f@lW#~e3vu==l(kZN9o0rXr0OUtRa|Gf?h?$UiBWS$fym42A?)+ z?c@t?7{ai3z~QMs>8TyoKKrLGbnp&#qi*5ZJMW~v`^ZJ^0=obt3z5?>5t37H1Ag>GcW=Z@#5T{QfDnNrh z2@4u)$-?P%Q!a!;39XJ_Pu7Z{6JCkRsCa0XKd{8LQLCeBAyrdX!wn*d zz{!8a3p36)6i6sH=z(>LER{w|es;3cY z-*4D-hF;4pitEJo34*qxJfJNE15@rMBDRn`0sc)Bz1a0?L>Vw`YzfEGmTR+GL^)H} z2d9dchWGOlC*5~FNhJlR^RLSu%-*M3Iw_AVpjig}$zvY#mW}us^Y=dQ2D6D@|H!;0 zo~6tD{YBVim7(Y%8&x^tT3Axlic$mgh06VkHhz^9ndvd)4R!F;mhesia^{p!xP0vf z;tQQyd^l=|-X<@=rR~aN@DhHos0*s;TV$6cDO8)RF&HZ6rB#Y!$`nwFi-#Ow3I(@E zX^oSjEj%o0#kciyX>F)ppyRsi{>)5B>0snttHe_>6sw~il9v?wR4JjgqEie~Gu@4X zV63uVen@@=&V@?*R@n$8|1!5OMGDnGA(vCW`XSjIyj+&XKg?eYW%Peplp=lsyUZ;( zO3pi()W|<9G#8NJvQ&3s6U8DolP_7vtKnDs7Yd(jwa2u@U^B_Jn7zNKdS~emPQ?tl z9G8>_RR$W)DHOKJ9w-Vyqz{((WKrN1`U;H!VyzPI(;3|k`G=2rC;d93(*tsp9rj(Q z*=V;cPIu3G3;o}9<{VN^+q711GGCPw%npJsA!2pH<1-3p41!795#6cGqV6aR z$)&RB9nombP@{33Z$3Fd_mPW%nXnd^39m<bJG=*G;rp*(`5hRI$;HB6*&w>ton-p_LHB;aidn^OhMrmc zx6?CyKJLS0M4a2gg?5Zc*atXYm=o(zVZ~XTfHNU|OX}-mX@PX&iHfBK5;c2ogtQ3K z=pOk)8VmI@gAFnxZJzx+Q0@hqAr-Z4b{hZWoEE`Kx|xqx z^vO=nIVEe4!R1Ad40nAPG0;3(3G@d(BtQUmhI-kJAj_DwxZgUj~(v9~iM zzrg0RH=h4{5zCLvc+0q8d+o#if2NN(aS-~t$&_}EV4%djjflkp87#QcVYllQ#bz~9 znJp&t>Ub4VyL~?7Y1J3p?uF^4XtJo?K8-4^x=8Xk4A%@93Oo;M$^Wt~@X=SmQ7;yw zh86Bf433NbkcXGTjSac6Bm(QmS9vZ#*?|T7gO`T4$QBd?RL=nh?-c4Xxc|5q(G*T^ zQ?`&v#N)s#p_<2mpUi}Rx}^(og6&uIDR#N01V7W>{e9qK1k*E6!_9*IDa`ww2}W?K zw^aR^e(b{O2F@F8rL#?-m`pG$2znV*2-0_>G^hQy2{CBdML%tSF@wh(qJ$FKPNElax3Nx z+a5CORaT$^4FK!)jyf>qqjwg*=#)}fUdQRYxee5(VdW}dV~El5PfM%akp!enJP5I; zt2XN)+V<3~9P4o(Z%*c)c>D@6J=!>b1TI;*jo$O6i6);mu@{F3rkS7{iP+`fiSC$p zRq?+l2e5l~u6S`kQgAg<9N6n~02l*>OPH3^IRTSPo9#ko1P->mF-oZ zl&%*y@gQ|h*93zUwm&$>ygjb<>J&;RyeF$Bda1NIU2xwPsPoxOcMJ07Zui^=Uzp5b zPu>|SPf`<%-!K-_!~u9lZ-c`R+ujBJRlhp?MvFP$(9`KrC!SzfIEXU%n_{#|jLcmS zbqYRJ@0MwLjEijGRQ1CoRt~w7Di&h50c)aZ3=h`grQA_q@iunf$!>nWa$>_J^Td_Y zhW1!eQHHDq4ta`rYdH2gF86()(rkmc%(igk>4C*fk@^8$Au_Cr^jf{2daaIH1vEor zdFibfC+>(Cu?tq*d<>_sBf>%~Z^E_jer=0+jkwdIgr(}co5Xg=8aGT}Rp{Fy{~j!M zIAm5~Vhm$z>!qA9G4cP^cE4_3jpekMU?~$DaKAsZ>g5&GkUU?kNe?)nzAkC?L7{L5 zTLlXvWECj>4yzPypeRb%Gs&D;%xOWv!WEVe#OFu`cAQR++U!>}_dW%IG^6&K=7_wO z%$%+Z&*$~1$1loZB{)W=j&WPdastPMfB&zR)4s-KP50ceOLX>^W?55hvaH!gFo3Om z41)lRA#;eV4b>{o%{{Iz;4O`8i>aS?ogVpgD&~NCpQ?(=pivdNNS)=C#_!{oL}``j z>I3fU+}mbVN?Id&gB)j&xMJ%IkQp{Cm@@n9y1cP!jX8zC)0z?%ZGofoNxwLu2GfUy z;%Jdml6eH1YQMmBJ*nPP{MhBrv)- z9xI$t;$Ukq0y|D9AtwFnZ!f;+I##{ZnpbnhhJV{)?!8g;y}R?U;Uq(mOFxoa@aYZ8 z6&I@RO4mfSe=X~q!@n$fqaJw0Fk6SShog~VBTgLd$a2HUv7&FkCNp{_N51{$MS6`B zXM=Z|_!^rD1{S;dc8x^T^)$12Zqpi6*~C&(1{D?0V+-wX5A*+Sf1 z9@OJwpmYB}ko@P*WB>e%-~RqTV1 zuM<}>v553y)#L}5ziFf!_!m^WLr~y7aZ zw(tgHoNL@4-akjLcjCFG(qyhFBp4X`Tp~6*pif`J-zYdweeQEFqK_<_S4!rn9?e@4 z5(k;JLTGE%w;)ju=!Ebk6!5vIKlUDyub#awdT3fdFL#P#vA|*M&6cx{4V`c^Aun~V zt~8IZeo4wkSmLo5|44o9Botwz9>Za}Pl6enE`?*!V)2!?4nh_$b?DAx$Nn5>UNyo6 z6i!@Q#v)sWKI#^ab8|H*;$`2^D(}gUOSVY5s4n2xj}sn}H1o$Ty=xd)>yY7KXT$== z(*HK^$Uhb`u%Z7pT@+X>%utkh@A7YP%?1K66q3Pi)p6Ievo5hW{=k?JA#Cq6*SoAW znb~&|46uV16R{U1p9xUp;y#(mLw8>j7l~Al`mWwqPdw5gy(POK(zJReGxd`rB% z8UGueNL8D!&iB6IQ$43pXs8$ZJNOdT8HTZ-&ct#6yFy}wYkWGqx0M_=s= z)^w6sK-4NobL|j;5i6%IQjo6M?W0ximE)1kHyg*F>t*bC$?)1Y|D}4{ zQAqBb$4cOwMZtE1$Xqe63o6eXYF#+PV$?*(b*Q{ zK1BM<=V_JOBZ?%2W)&DB`YHVSx_Ho1W7sdI1q%X^gba)UtiQb4^BD{XB+?m5kyQ~FcN|k^Y*-P+2n>m)IM)f z)4bn}$)Y(BLt~26CH(V+LGxVB)2@sYBZVculj_qGd#HqsIdH2-cO zV%NzY&`1rX7fG%rmjnSn6tLJ(Rt#tK8ceIKbKBANIr`0 z^*sWv1*+rfTW)x`Rl2ZKQG8JHDcx>MjO+mcu3X-XA3puL22NZjxi)= z{S$w3rj?j@a=|!lal*vJHGBS-e9`0k6{5w$91xK+{FkAqq;~UzVo z5Be_*`wK~}+7N23n`PL;!5DBKM=POX(P(i0HtVn81eFPI|7P34vBdYCc)`OG-|tnN zm3Bx^kopzz>Zlfoy`PTA48AFA;hh!XQW z#eL*Q%BA7ME-03X$kZx7zTLP zUw`98kG2&_@IOslboHF=UKQ#?^7~O+$)oi42;3?pHw0-6XE@ZHm^@*Qqiu4nBR_Dn zEi2q6KQh-~=2HFW#L+eub>^-99K_Rx?+t9uHJ;#FN2bI1)(4d%yXM&kA4k@_NkhrbK4}lYXtJG58ZGuTKv+ zB(D|q%0Bikn%WVFN-)W@v!hW8b@!aqfCQ?>E!(#SA}qD4ET)9Fewt$uVAI%-0|GeN z65^Wf;7Tji7*={TtC0)(hUZET419yTh+|a6OPzj$U*m=qlI0;qlEuQk5v^!L6$vw99v2h{l8 z3~Hs0_$3Kj{P69+6i}tsj1WzCNS7uQ0R;W(NP^_S9r|L7g z$>Ne9T%4!>&CWv@Hj$smFBSqZ5|+nplk8K0)I`MISF|8mr>TIx(*ILqaICHJy?3<4 z%NW1uc;(C==yYyb0VhsPmzzX;@(2bvLbXI}J-Gtp?rwV*^J?6XX1d7-)Ij`)T(W`| z(|Kf#cnhyZfX54|`=ABgK(&Y-NV^4(`S{tf<_t#KR*U9XkW82`wBy%V=Dsi&f5(Z} zQY?9DRFSC-UFDZTm=cmZdhj-jGV(E@x{<`#KsrbIeO#zk)ND> z(Ycl$QrXcpZiPWzP|emU$RWEUV2O-f1~^5=$9&|cxf!Na;-tA~(W3Dga*o4or$z7c z!NSS8nfS)NBW31lKTex8V$ts@jnXlF-t|Bu-7>4ruLsz2Fk_@S>9Jz^Pyhr~KvE75 z^pZu?0Np_8TnC^bti>Hvtcn8hK#g0aTIY%lDdL~ruGE-Rv1Wqq=}LEy}jDS#P$Q!}i_KxaYN=%-bF2C9jUHcK|6 z%2v2odEMih!AA2m(bF}QPV9lO5QSpesDT_1X$E;WgRqVX7*#O*@91%U`mtL%@=ssr z;63d2)RfK3@U5rb7YgxAbC;T9)pBlDSRE%B|kS;r|1YRqADOba8cYUj(fR+ECCY0 z>;QxM&*4{Z20_e6)S$}cbqiXdW;1unkUTp8wRqO~4!Ea8UlQyOiXUUN&q|D0o}r$- zcT6bUE}nAS#fYSm&)@xkj&tGy2dxP<(g-G*pjQyFyXPF26bnm;5&}JxJaq%rOCudV0a*hRORv0xQ8IxOOo+48 zWVM^oa=foO{YN^R!^7h~%d7H9LiIoZ% zFLpQ{MvW^cXpFkf(VLv0F=6ihWm8`@LgOc&oXn>8Ix#fbO`vgvU=9*=6I9KC#PjwD z)+mViW8 z*Z7@|$QcU@DJL*lGC0rKXDjHqfoVf=}$$-9C?Py4663jDmQo)x*ewPsS8}|rnJXgl(hNwhYh-~339BA zqI}s1nzy$7>Ro!H6B`&HR~ePfDIpk8;V2+t zF=UTPzamNp8Jk3(a~QzC7nY$Y@;1I^u?;ZNH)?tW{R{ za)L~A85Xkr3Y_kDP36kLY`*jLuCF#M7gQHFcxtN*Wv?s;$<+wscv zGk<2zc>;pY4}Z%l!%F`4yt5Dd_C6cMorFC_(EQtVTGCjVq+8egj}4!kSv z2pw_(F=cn1usCq%OT~e0F?YZ9(N`aSt^BpR@1KgPal5Lj7wCXoER(Kw*O4d0Z81sT zTU7kxy4TwO>D2sFF?RNO3B32Lv&n<5A{d}SDI;RLNo>DqQ=Sx+5ZEA4IQ^oe9J~|Uvt;^N z5oT=~s76%^HNso2RhAIw#b6r%dNv)>+pZYQ!2R1)xEv`VT0OfUjMgJhbG7#x#LB@K z{ob){+jVf*2gAbq=)8+m>2ETY1R0TY@U=D7bl#U{25K~U0d^4#=yI15u}8m#iS0Cc z0qVb;n_C~;9)nJQKW~3zC0#E_ry8Wk=rsNn#Q-eHvFiU0x!5BWGT`;WxxB|Ron(*X zPDrij&}?jVT0aeQ;SeW!LxO=*VV71Jdc=lLD`si( z=tfnm>?obDZU?^J?GX!`gG)%1)hzXjryhC2j(qj|gS;ZiV_5z-sthYnSP&*q7u@<4 z`y*RrjjF#6@-#@K8Yk=(VUIO7L^VvU0U3d1o>=m^c5<%>0tmIC%RE=a3`n&~!(E;e zakd-2*laHQ{%0cc?fGZF=WoQDJ7U~1ih7rO`LQ~c!u2W_L#SH zl2+Nk2dEI|Lb~J^6>X4FUF}gDpbcn|Hc$&r&@Hk$k2)Y|{z!CPmL$-^=wUU*-RUc) zYn4s>E~>&m-=mVZ;2+A<wfpaw_fy!Al7qtNYh?|*Vw?_eVD%(Lcb7aY7Wj& z=-lGNuT$;HGT^5xqEJxgI;17A-9VE?4M^*z)VQSv)J=g~gg%nsVTYi_<`%S3hTxfR zzO4s2!DZst7io@|t2VQUBOjUFOJY-P78IbMT1Z7yp1N2#;NBYAA;t7yjyPY^Dns5t z$2@RL4}_!HwsZ@J?Bg~Fw6g}mt`*#5erB9~y1x6?TXY9E@5YIj-w@>;n;>g`qNHr=Ufhs_2E@q$MHv^FEOCE~N*&Du9l8 zyIY?Ex00E-kOB8@;3LATaMr^s+hQJ4+l2Sz{h&7h>}ej^9z*LW+K>ppqh_n~{~yd}h;d4_KrlIM#rxb1YabVVVrNRdzCdF(GC z$CqZ5!LrLH>kGGH>lxZSTmE!|iW7UxEE#mnSmuZ|6+oZ}KiXh&!*0o%h)$*Es<$?v zRD_`j{Fok)5RQjh#n8gRMlMo2c$58gfvYgkjzy3ci z4}6SM)=y4MI7MeU@q%fO$s|-vFwkV5ht=d*Y}G3pRA}n?h0~D=QG>lHSS4YQf4Glg zM@7`Zd4uk$JfOXVB8leUE1??noHG;~rN<>b5eHm*p?<*^=z$k3-!znN!os-(oD4%B%688a)LPGvX{eyWM8&X6 z0a-l`EAR@ejKg)cG8NBrbqqvnKW+GBk%BC%{Z> zKDzU(FM4sqfrt*KK6qL9Znx}cEW)cN2c${F=>Hrs{^rm9$!8A8cyiG=Y;i)ygvfh7 z?JxQ)Q%J7!Xn{thBm&74HoL|PH5lyvzJ|pC5k~3M)k9U7&CeY<2$g~jUEe73JXrJ8e^SQJ(bZuxtpjLWg zu4Ayp2}5i%j|B{jbr(B^Ui)?35C5^$i(2GYLR;hzHNpDsnusHOJtbmoXbYwWXsuL( zn^=``+O<|zsApUrk2%UF=!_K?R$M=QL1U2AXi!9VM8BicCPA)$^uj?$Fq;TE3lq88 z=<*PK&lh@!og@l>Ul75){Fg^(C~8RzPGwGg0EPm|Zn`L_z}U{~jkqpp_0fDX zql>DgfJJu)7iAWko{p6x@$^h?)3;fF5vP%G-V(W6@BN#`93n4NeuiZt%mEsw3aUa> z5p@-G8?b#t0IszTEmt)F71>OpW!roB|E>HV8FQ%;z*5Y;CUR$TFZaO?|T9-LlyVEX&Sm@W+ zPQKvQA?*g9F^rEO2s0G>)G7QFm=+M}Fv-juU)4@dptT4u$L!<=+F$+ZK4Tted!c+? z7BTP&QM*znHi&n>3lb-;C;425AFxd6kiBmbP6aRSDK|DMvgaKcFPZ%fsm z>BlbIZs4?@6`gIev`!|N6$HJEh{bhuHSi^tOe^%ywKqr4bH?V$;pd{~$*JSI_0HYa z|M#K`$74}|K5z!qrq~7u@nE?wYVGWCH4lbSd+sogqaQIEN8H}=kCr?Le%)v!64!qB zM|7PNN7OEuy!Iyu<|BeWOvIuf?sW=1h?`**^a>zTU0C??a_Ev+O|_Ce>bp`+C5?+0 zt$LfV+bgaQ{*boR#^#!J)>KAjlqZ1XTTD1c3!MQ)F>as z=FQP~dHyyxM4deGitLvoh}yJx3Vmr3{n%vSZV}991pNsSi=;}0{+%8e4u@D|rCZ75 z6XF|^8n+@?8XOgVDEttLk5@ z$esutUxP|5oq~Glko-8bnD>I4ISA^RiJp(cP#&?yElxP-en@^sHR!&IKS&PAd!ZgN zN309Sdus~94P0*-id#Onzy;aq1~JjPV4XNe*&)KNjrWflmq*0^1# zZie9>!BVF*egpe#M3@gXr-60e#L<7J9{rAa#`#MUxMj&t_k6R0tO-T#GJSnTw?3H^ z7d_rA@OnI13vP4pIr{U5YR`2cMmu!etmh!mQm%0p060NZH|#?($8l(lGhwxJq*)IK=jX6jbZix=^X*l`JRSxhH# zYfN=sTM3lAqmhYzqtKRML(rz#1>u}MnX!(U^(Y$a zH*vfErrXNDPB-G>{SUhK(_6XW!g>4J=mRFW*h4Ut1YJ(VmW%fXEt^!}de^-n;Dk@o zq#QBM$-& ziW5VJg#oRKR}$3#^$EA4(gO4&N#P!G{9D4yCLQ%xqh;=PVY%R>@)r!z%_|z z3U3O34!07=$N8+&xf+jW-OHZG*u9s3t^Kc*u>^UY*i&H%@?teav9Qw(+6%{*48~+Y zpN|xi)5q;Y=i1MfKfj35i6;}5BFbaRyODk5dg>nV+3u1*=C{Y};N1y1CWsRzgd7A# z%|1aQMEF*^^+7}O9bP|7J%bLf2i6bj^MNojDLnZ(9%x`U8Q=fWTd$c5^>XnRcRu7L$WFehRpJ`D!uybeTE3n)vR0wxc{g#u@u4lcx6JENoHmKfQhSof?*XF7 zYEYs>v7iOuQW&aUGk{a6L|P)is?UN)YkgH`h#vLYstx$V7PTs=FVObFcU2bn|o z2B6&s`T2xEtYAnHpM@}OrvSh8q^L!P-`qfL3CHu=l=q@K=@UTH-xAdno)EZP(H@iO zS{aDH0q(_phHGxh3aR@u*Hf23QD)?C*hX!JP8b$WL{0!TGgA!_>wlKTY*(F$CRxUimgI%j11}ZQ9B&;(}3i< zS$YeKT>56W0V%^JL9?_Kh?ubuH&dO)PY7(Kip3pJgp(ei1M-#4%vCw6f!jMkYd<7d z8c*BzaWYjC{*<-m7hXo(bi8ur4|KXE@*O7*zyfqe6(!^m3^>hNB39@9$OoeJtDu?V zlbK~?U2vRGb2_wJYUmtH5w{9^XK0m6`DNrj)pk&1xCymz=jJ{FGjKZMnBB2rHkxCA zj@g~;`;i?wr?OS=z364R#lnyL022Fy4E;W((Cc0ZT|NopO@0ewW+edZJGxffjU51% z!YwP-{YY&dkNc8@TUerZsJMxJVr7aW>g4cbMi<^5Q|5gftVVYDadoHQviBhG(O2$$ zt(L41cGrF81H&ac*tFOUB71@V7M!=nY>QKNw2iMol|g2?nGx zHxsePRBxUhpxA{L*)r-><^G9cl^HKx0(M~#cbGgudjf_7G>x^yzI5z6X~vV z$Q@pV(}97^p*VQ*!Stf0zV6xC_twKXP2;R~5lgU_i|VqV9n$i1}yFk~uSTC)Mm9lu{V&Nul?AO5m90Lt57NnqeZ7h?H6R&kxCLJ9=9f+$lrw)hJ_-O{{bRfyON@ho22rUfKG}86K z=>a!94J6k`>Fp87r=o9oP7=lo&x*2w?m*-7IbNEBWHk^GV+58%K!rE``eWY4pBnv> zpYGfE4f-56|HO%*GGJngJ|&nQg6<+>4W=kV(J5$!Hs=0l1KCTPGK~-HqG|jAx_VBl z3>(Wz2xR=iL_vqNZ+1%Pa#<5-0o94NsX$bINyvKYJe?9+4W_D_)E`nqrx||0Dux}r zdeEeYd^UmYY)DgxG*9UGw8}sx%m4JNrJ^LaMT-`p%6}^R8uwcP< zFYG?cQ1s9>qIK>yZaSbmS@6Yi$4nDrexDsiQ_5+gaNcBp!S*Sh<{V#6>+`TEa@+yS zu}!*7dM)PKthM59MTIE-r6SP!+T^`aiJy`bMUK&GxA5lacX2TbZ1-@)er}kJe52)u zp2ivL`wg4U&}+HPSk4Q2kRu;$3%3xAj-WRYu}LpwhaaG~OIGo-!!@Zsr{*jTFB8`@ zEx~QFGI6Q81r?j$`S0Q%=YLJ}V0IrpC5}MSlY9Qp zoB#XPH;gFx-Kng1=ntJZ8Fa;D+&?Cm7J@!T#Ksd>A!}DpCIX{z6K^XN$F&H~`(>*3 z3wq?$bDDT}BkxBIf$ZRQ?{1$8a^(0O_dD*#l((Q;x;-YHK0_ZCR;h*qI|YWba>#Ub zPt0Pt^D@XX9-@0dpZOMfBW#;wpQ@0&LYIWsxQ$$yr`{*;=Y2H$AzA0$2d%4T;M^X1 z3;#w4cTm}m6I0YY=hG7aSh>Z`_V{Z4{bd&;s!Be8_X9ePo9%I8Z&_=C!8C$NCg>GJ z?1-ZYi*(fmQL!*Rz)?%@=M0!N0D8_X9CCQs(K?;P2nqS$zS>5oTFM?e@uUN#Q==U4 z9D>0YG{@aYh8k3m*6ezc$fm zND`KXG}FUh8X7PV8X09`QVC`iK`$p_^@6z2qkM%f2;btdSXd-kXzvOYtH!`8L|Anv z`<|5k$FRfSHX`DKqP?Hchn(1ly=F4T?F4g*pidI9SHE*{UgG>erOn^T14ayJCQl7W zpq9F?@o0}3L#%;q3Etmm-ONazze#j;Wq^Bjg^eeTt}_u#ZG3p>3fDNet~7?odP(0E^TvNsl0( zET6S}vQ}xh{TAO?DLO>x&$6U~g0`h5TF0YrYRC9TuZ2bkFzRrqA`SeBOu!qyD>AXus-dq_YG)_lv%- z3H2)Dg<>mkYdV9B17`6nIN^3e*{46eE|HEeFvN}+x4W|z5k_=8_IP|87IrkmB}Yx^DFm~Upp%GL3^pOt8N7_! zgqMLxar~a^7k>zBSn-TUpFVEe?9bT3y3lz;jW55n{jcWDT`WCRCDJCJda_=Cg~j{l zoP+FW8;}f_L=}(=j(X$-{RK5~;-iy;F`h6uqn<2-;^YlfF0TN{@eJ2qmfTU?gFdPP zrU%NH)2XT%#R271K}1nR#pD4yn|mv8W9s~|f3hp4IN`>5(;LNWCcXQjlO14*w1Gq` zMUVnaO1+|$9?47(XsVCDsk0KUc7>ufH{%4?i4{N1%nKS>tGu?Rn$C0LrEH_gytj*B zpiHEch)v{YNe>J2)L1CLUc6ad4K=KKCY?^GK-8*x^_7cP6)-JgX?1!4$oP1)f|0|+ zC@)zuFGoBi*DMw``t_2iDp?1@KIKdWneYZgN9_zPD`CTK+nyY;1`Sr9!wEJMe)zk; ztc*3n=0WKH$mvcehE3eW(e6?o!Q3L~&xlya;DV|lhU}0Wzkw=5358=JAoc}>OxJiz z=zhV9S*Y-|&`=oLLmranO--OKAdz2=7&H*qQ}M!jaw%O$AER&Z^Fu-Zr!1&Kl+MeN z>OWv1>p;{s_vO)B-rgQjNT!JMy)z~+9Qj|rBAMwFBvBO*iq$HIxp`58f{PF`L|Z*pSL?K81e+X)8ZaD|Ws0uF=) zxBZs}YL%6eO7TH5k1SBP#oXYp1oY|nCE(uIlV`kZ+y>|~-lt;r3;Jkepe>?+Fpz&U zvITnIcd9N+s-bu(MO+!^plNUHc(K~*Ba2Is@urauvP?TrAr zxEeofAU@!#UwieCE9h21Jy`(d*u}zbkn(`Z1&@M(sg-q0_rjbs_K_R=AX$8aISq)V zlQ|8N^RBb179M%1!g&}kU`666yMgmrU>*sjP2kso^>BoEP5 z;#yS}G#42541`satC)UXE^ns_Q`K9^l>vQ%;lLr63%nM|3Gr}Xme;k&u>;x?n4avj z8SYPR;gGGaU9vZxPj>CR$t;?M>>7zyfXJLQ5Kz|5f}WlX#oegSpufE?c#mYGSMm&W zTgE3hY#2F?5OBzFTe>N`#oH(%zwx5S{F5f>s62H6xfNnZ`RWS)Zoy-I6|X@3$;>0( zx4j+x+M7Rbt1)1Hp@ToOW57amk(jT#`b~32V-_l4%&TdkNwezZ6;yga7f8@zT5Y@^ zJ3c+D9=(Gfv!&Z}di2g~h|^Cmdt{CYr_B$tfE76lwtB#N+Ivu;x$Acf(2Aw4e+RT$ z1=exS3+i7YPQ_)D-t(AZPAtjk2F|O9TWI2hW)ciweJv5YBt$2^PSu9$+~UJAz$K)&-l_&eGLy~bV_2e>#Z^_mdy=`S(T6sv3g8ZDx zRQ(-ZN$X}1hd(z9U%+Vwao#1Xcx&4Au`F4gcWz``vKq>bx)eo}4q~25!}}F_ky~u3 zC>CnaUSKV&L4Z|%q*aCtOKU@$B^^K_;NZmzC(B_LjIn@HyX_Jrx$S=07}|Ji+ppfG zH(IJVabh&=HJLj~2nI?}3W(U19?h^uZC54)9#?A$$TWJl&oy2x=ui&Hi^cmOjg6E! zBO*@vqUU;k;dFh;L@GaXdZ`FmxlniVf~1}2C~S-c6pox^*lpnTF0Vhn=yCUINz?Qe z`EtcpGTr?EjawOt1m9)hyZr0P6_azI-*?j#Oq<{lWD>sZdqQ@U{$wU7bj!LFNLzx( zQDzd)aMa{EW^CBaM3~-)9UGPww_V>o*KO{YIPF4)rIP<3i4JEWd2iXdakNi?7 zM-95^qc;0w4I<1g8JCZo5HVq0sN`?vC5}#u2$pijzS(Keq!s_IV(~4}O2t0UT)!j& zQ?2peO6OgGpy+Kq-BG$c{#)Y*6l>5i%C8s~WZ;)@L&pNISFfA%n6MClw0hPGdPO*e ze*iTerSe`cplSzU@nZ2+c`vyT3BrrTNd8^NKjDL?X)t_!Oj!@rkVO*54h+Y^7&R`A z4MT7n45tkkygNDOwf`_WtrH*rcq-k^Eg$E^?(7ni0+o9N(?`&^ke@AE*iCjR_6uNH z(2|A4Lx)LiyP&bIz2ETlS0=T6BX zAA;-L?NK*bQ)H~VTjqKEmF|dS22^9j_}XcffdT9VBHIpWmUKY65R*dqZTJi96z&u@ zs_I0gq8`%lmuOZ}aUu7}Ccc69($F39g_}Kw*Jt#rWXElTx%|UV=Uy}?;NlX3b7I`F z&~+JxN8cEV`~;YMfW>rNG*Vy~rKUQh2V5O%O-IjF9HBA#QipEGfAHO5zR@I2&-~sY zda0##wDZP`XkZu|)kv_8VAc|J8WD^2Zn$jLg@Y}CWRvE!bO3^|P!@B9zY5w~zj$XI z`gFgzIhW!8;-YcbV4qmOws_NG^A*sS#AdLpml{>wii1!#3nQ6h1mWRrYd)X%q3!$`$Grx7|k~uqx(*{mh_(po@CBnu0 ziWy6UXQ)21Rj}O?tC=^dj?G^}_XgboUgPD_4L&GDxr!;3KlaJDLv5@z17-t4thw!& z&*6j+VsXqnO7jW|ve#%~gsUqlI=%?K9-1`bFPPVB^p%H7b5CR%)O6&~G0pl{r(o7x=K9`~6c3YhQHrzcZ?HOgXP7V(sMf^p1$-leegf zr>&iQ!OpFoY){Z+zjecGa>=-SV7DF?{8VDo^!amNbT1@JeM8y>!oj(cw$L^7I`B)& z1YN7L+IC9R&_KxqzwhU~`?$LLPwS)bR-+?EiQm2$YaF)Xvo*mqk(DJQT^pngI9?RQ4kX3x((Wb_Ru4yr6S z$@@Jd7(m~>C%Q}+DtR1sFH9d*z2}ArJN(!mg=K0LQ5lL4d077lUd?qGki7zv0k(Y) zdRz|ID!UZNU5g~cE?dI2RHdp-xd5-!DshrqFhH+}$&RjoQtAvvnc>X3s1nAwLD#TjbC>3fO~Sd zrd}uXhgd5Q({~ygJdWRFE@0f|JTt_hb3A&nyy{H(Pn(Lk>>=e|^yCf+k3j)%Z0rz8q zwumdBlu=Fg2vDyyp6ch_jj9c8=9h!I#tr@f&rZQQUQ%$kV2~V)u86ucqk}oauau;S zD`w}--N9QPjU+>zf=bDnS8LpJ#k=%Ju6nQH56SJ~E<3`0Su#hQ_i?gc#1CHbOnlK< zdOr1S0HcHx5pn|RrQ?`%bvLkF*k`1<41tv{6qj4t%F{5e0-Qv$Y;)RO693FdtuMh0y6&-t2hM^G9eT+LbHA<3hUq zG$~YS0E$GSW_AvFPpzpHHBefQ`rwc(_B+z)zfZM7q%7mc+_Yf6C`IM~}Tf zb2A+4&zJt%oT-|N;c#LS3l?7MeDzk~)M*h6$qT8obPjojig!CiR*HsTO1k4-H>Fp0 z!UtDF_o1N0;UF3re6C;`xz5o)Q`Hl`Z{BkJbb`Q%E7(}tkg;G52$tlxfxSzE0b8s< zSm`m&i2)ovD@WeZ1O3$f+097w{}$DoSE@K|xQ0cm4Oh$=iVE`B3~fN7Pp^9>ac&$* zhDSYYYleT+>EpJ^KKvGKf4$xwyt4{-UkcN0=|H6QwFLc-bMHP|F)9ObDr((G4;gW!}|uv`XY#ycW}? z*fgbAgaz{0Cx{~ChZU89L-L-G;mBOxvPp&H*6@L-q)BZtr((3qG_0tZi8mQgU zeqO0p-_$lwthQh2)}>fBX?5UE?+Q^uc+w=KX6^|olefs0O=^`jP*3iWz^|HtcgXTO zZf)|xO5{J~yZ(Rnz6GwS?AqHSo{($|c@an=phyHkbPx<9;GomTy}fOxw~u?<&b__u zH&fGlJAF)FslA=f6cK!bpnw7z0s;|4P!JI1rGxkuP*fD3Kt#u3M!^Sz3g22uR1%5i zgoH0nzxu15*WNH^|IgZMuf5j)Z~F!7z8yx$r_LR``5p81jfFI{dhY?K>1?JyxFT$e zx%urD2wa^3edz{n3xBsHmUqK1F)|U@uMdkZ$j;8W#!eL!FU+O0XN-eQvTZ1=d!1u> zirG*&tr1dA|F=cv;W8%c8z;_W$T2Zl8z@;aMG`6HHDN!L1l7@-Sk*xo8A)Yh9z&Zv z*5L$v&%@11a6jMB!B@WX!!vc}0z6LZ!dj5j!EKbw!VRIYfEkcp;`d8d6IDCw+>C4N z6t8iQf>}AxFoG zYf$3gFMrhTEu-Y&im!FN>V~38dRVl$lY2>6>vlz$5?Sa8`W!+{2o!pH-XC|)0&5_k z_cBKvD92s241rPByVSey|Kd**j7uBE)~dgd$IJwXoj72YW#UyNQnJ+)SwShW&QqTo zi7eLtZ`dVa@L!G|$06iJJwL2Zd2BAZ!^F#U;wZWW<)D-QeMJa~_re{X9kMOcy5|iC zY!bDIVMQexl2inAv$lyokq(AbQ=5aXO<%^t`|UA0(F4*);}dGRhS$qEAZ+1c?A`QP zPy@4jeaFjjW97>BMIf-@-jqv0zcnu%cUnKxLNRzL`x3vEUnN3@d}JKRSG3S;fEXIe zJUh9^--IbAqTgYBNTa=TMuRoAK&)I;{tYD=zSVwPT<*2*oM0?Z@_EVQBoEoMw=0Go>`0$Jvipj zGIIX=)aAdL8<6Kw;>7NT1*5XUN2@52obp5g9!y9*sdyw@|EyT$fuO8VqCFS(+z=; z37sbGkEo8yBYKG=$rjw=z(x|DX334GS{id<4HRv|F8Ln$03AoId6^;O znbjC^yf3+pr$Ko-=2qe7 z>C@wPP<{E2-=1f*AMc)?zk}?WM4C-L)BTjJjw1UgWy#C|_nkp$?3S>m@ODX3XsUmr z>^SS%+ykO@(mX+``!>4I`yxn*KK5C`ONCvryT7=w5G2e(<9(CZ{d1S{S|oM+UjAmU zc;9|@Cl^~CRj3tJ>^R=8$c;05r74jsc=cqPM-lsP*S`5+Q2~3En_8aWV~5=tt6u(y zFV(uku@4fXkJovfz(}f6Co*D)e~15DlFY1*)QPQ1iOHv+p=4lYGO>$-g|BGPHy?^<32Zm9<4g6gr85V4g}TFHQaSlM@>%8kZ3c~pLOAo zqrC4W^qHrQFqwFqI2vyu1Ey81rB4b!X6rzoQ3dL6bfpiC`v&;kB*@%X8qC58|&XjMK_d7;w+wu-Aj3o;en;8Es@&i#6U|&<6B)jDU zoLJwnIl4>%^AtH&99Rd9(Vwymm{C?d6Ge^XkIXd=m`prQY<(=KAC$6}u?~=$se|st z5o-SyP>sC;boV>lQe=l&wQd?&oJa0-EXS&u+Rw)O36s(|t&z(i_3WxJj&q1E)nC=J z?!VUUxP8O1f@^eWjQ#355BFj7a0(N+II*8%A@~CPC30LO7x4;(s@QoCBvfhM zbCNh96@)>SG${sYS|Q+XAVKLPHNbv?4h?JpU|l-;z7BJ5$BGt5W%stJ|8z!RoS|~K z>DNfoBvNcL5vVB{Y_)Epl-q-fMfjRCWbm4c1Sr3>d>&Rqp7+Y9FGnW3p7kn+4YL~g zZa*EZ3v}44H2Xwsc(Z0bIOdd*k$2vYE{YpC(|q=E+MPcO^GPaOOKYHbzh7E24;e0M z+}dS;V8aC@MnWQrG$|riPpnidI!2bqfME7TmO(}5tnbb#DpaArHD%>Y2Lam#96#r# zI0DJcaO~3f2ko0id-8~#^Al3XERf>F&dNm-vv`t{9izx$2&725qz$5_ytbIF9uYcAq4vI(BbgxfDBo-pgpF{dk9+APT$VUmDPI`0ZRd% zEoIeCtB%Z7fXfr0R$!}IH|*^cof6p|vyH~%As4)!>6NRtcPVm*QsU;ILEWQO zj6-%dR3|d>kz@QVU$MnI2h%vN7*!Lxd0E+QeHXG=F|jVXI}$?1Mr`2dui$2Gt}}T0V-00h>?SF zW*w$X<{;yVZCJ6k)wG)4H)a^8touI^FDJRod?4pt43GvBAE=CyK^a>yrK}+R(m{_n z*QzLkjA^25m!y0$vZkx$Uq}u`pZ05IgJ+E*W4&S&BE+J@<9scRv^vS8W?nXZTcYZa z;%TM`Tik<|$QovM2-yxAQy^Ij25X;Nl(Bd2bR%fufA71NWIM5+Q)>c)Qc4CPgF;Gq zG_rc(o{&$2%7fJMrIg_)j?+C4c_&x$$_3T*ioj1{@1R?HH{xP&mRAY~%Ry`1VnZ4^ znWD9kxuWiQ%jfTtFLlLM_>A3$Uq2jnz<{;G%IG;plM(f+AFD&(HzMYDt?$kvpMGh~ z%xV)q`Vl4jf+F`RnpJf^POYj}nm4J4dvvvdCSL_kl%HEgx{J zvj!rDTn0p2rVY8Qnw8))yH7~PLq)T~+ScAc;d+tkl zMS}gTL4Gq?Ni}c}(YK}R{I^U~-Im^!Y!^DMGyQr-?=AC0ET;_&SxBt{F_)?FVn}Y# zP6g)dK@X&Jbd1S)u^q6=AicN?humdkkS5MO-~9FujRxuaC#ZiSr}w_$0s49y7?U5cbiPQFnbx6xyJ2dOaxy^D*dt+s$y47rvh0(Kb4T z*F@HG2EnlH65NyQ_sOF>Lht%K_Q?c#=ZmcC^RJN%#ov~ERKOl&x5u28)p>Ms`-Kg3 zl^{X?$U6Toe7fkZ9$Fe@p|9!53|m7_geSjY0q|B~($pui^V#T-eogeJ!Lby|JMkLK zLZSRxb_tL&>s1V|@vlqQdpr_WMx7KUPjOHrdi3{hxN0pbUet#p?#;idi#A_#Sy-$f zms%Y?5OLjoRrH;I&LE{;8v=Xh_Dl2W>Zs%LOs^sV@=Jq9P)C=-&Q^7lAw9d&=LqQ& zLKLnpwA5=WRV&9^$ma(#3x+3FQWicXoBxBYecLbn@NE7KMsLVyUU1&!><8cZS~z8# zupWCnK1qg{RmVE-e;L_qGGQfAvNaTVmr8VPH5ltnlOnTRs=vzKkR2T`W>>AjwONyK zInAl}+cw7ti0WB?SwXfl^I)795JyY^v5%7Np-2U##Mm$i~|78EMk} zevf@{$6*WW4uoPVNCOv=B(h=aA>FqsdN<@u-10d_5`3=uoF&_Z`-1w!nZPckTJB#n zPdhbDx)w5XP+Ymp_aaNxAxfUPdS)8zDcA>W)}ZI57SZ$pyPzv>@tp6P=TbQBipWBe z$CcpI(>|VN_&ZIC$NPfRiv64`!8rRJgwkoX{KA4k@JQ}(`pCE?e=;9(yNu{dJ7F{t zQl+BPul?(jh)vJ$u9M@F$R{S>XzP~cU!c<^NRn6`gZxi$&~5S;akYcht0=_G&Xtm(^N@6SLXN ziPztiCSS=mN|sNNTuOQ6+eq$EEb8T+kzJqPNIs7^L9TEfc=kwgMLXD=U`E=?Zs9i2 zb^P)Oji4KdlXDfz$Yo)Vq&HA4zsO3M)o*LQ#0%gw;?`IYKcXAE9~RF2r#TTusQjR6 zyM~lAgNpNh1fJq@E+ozLOHtN0AGqeS>1&E2L2qF8Wbaasbyq^WD=d0y)p@SZ%_*+kOG-gh;$xrc7Ev6T0r&9f!{2(Gs*KvFKM@|v74s3OcKIgBICWUSeOA$ck4W3jedOVU?D&+b9!lzq~+n}G!$pmj0C20C1RIwq2 ztV6H@vyV6Y+$fG6EzfmnmkVFvy9an1ZBE-emG6;^Nu!iIII(2`0nQNuABB_*I3c!D zN+ka{?RnB2DRR5K@6IbHs!Xr51S7jzx{uUPSAFtLypSec=6+n>N8|I!v!Jp8D*X{f zsnc;)Zm-4BXhStdM2yy@U9MU7yZ&S0A4~q(h?SksapZ z)%KW&U%3L?1__gne!Jjbuod=}>%YbG?u;`rnW7`SLPg&^Aay3OQyb_O$!@selMD;izor8Irm|xlCiFZa%MsUl@GDyLNhxqLHH=PfcNFuwgp*wm<(%_eDp{ z6F@CwPIic{&(Ee$LzX{KnL?~&&;zwRAak-Rs1F1#d&RJoROyo_(}|i0s8P0xG5uRh z@7HrMFt)|a{*#!!-jUXr+0SA6wB@gI6i?B4&Gh=lXkCoagy_PiW|DoA2!!=Ue0#?z z8R)woq?8|YN%t^+XtQgPzs;Dl9Ze&VW%-yXik>{r2u;pww!fwM zpUMANyOfnagYL}`hrzJ*T3K)jPcw3EP; zj5IG*lx#Oec2UYcaSlO`4>x;{ipoMzJq+U)cinTyBk>L|lsAUTB&g{v;^iyyy|;So zVs&}9%ZqrZH5SJ^5LCniao&k3HTWl5^MIYx?od{nK>8`w;s!{ZEz}2ZlR4!eBp|SIAQ_R|o0}%>RT<@Z<;Tuopw&Lu-3<_jWlUPaN&&jup}R(R@^%$nZ$=-u}XS$>hwYOvcL{ek0jZ( zckZAlja@6pDwTq;H0D89#;@DPacBHX8pg_PC}Bs-rLguXftbS$41nmPX4)v^+Ho zmiis*bw+5JQ2MS%>08DX(wiUmzfF#gwteKpbI%PE-}eF~J4cZ=%nFN*K1No@sM_SZ z@I+YE-F8LN8Uuh-Xoa!^dtr&!Oq-#nO`gorfr-jjv_p;p5Rr&95gQ`)E$YNY0IFOaXegdeEa(a&ivT@1lR*CQp>1>Nw`e54#{L<9K`L;4#jD zTDr*<6X$oaR=*?{>ZkY*? z@+cY9n5Yp*1MbyPXo}F(WCvmzu)aK7N23!0)JIhcJ?T-YS1Y%kl@Y8mL@5eS?{dj4PJaZl zysY)aBLjVmTC{X3erourqmu@VZ_C!f80*V^S$jr<;k=ffUiZJpU-g{vY)GU_=HOIN zB*>=E^Y^o~e);rq{*ooAgK^7q&^;-%)~zZ`&$6hccltTde;ENAn~*!Ay`v5pA$Q_` zsV)vmjkC?hl;HCueKc3bc|(o_0?{Kjz6vPWR*GmSrJjTni?2U%KNA(pYh(3GZix4W zsVYdK3^=Fdc)62~<*6{GDqD~*gdI}MYk1_2nVQ2cw>%xHt78**p1B~7!p;QC;K>UY z&mvED1za>I$O>k83r-9nNcJ89p^cO*l_DvWa-ZTr5VA$9fmZ<%re^H-Qx(FFg+bmv zO}gqQciwWOoATTkGr`PrjU9L)a^mVA1R0SrEBo&$q`-+C)<%=DuB2q;6xnI0he0Q` zLzFNxK146@iF;ZqoME54pCWPGEpJ$ z^xnCaMwS@E0=GkEU&vS^8fyU{&tJA9WZ?H=6wuHA*Li>WiqYC6Z20bvNWBwVn@c9v zrkRo*r^pdXS+($r@UNIVljYw>>m>U*`{huM7t$Z0;Wcn?%+m#?Mh!{YWAf>hq=tWy zXlEg-=jF)lk*g(TK*xGqzAU7M-$3v3&K2F5cL?fDkhWFpHb5XJmwREs7GTKQF~7pE zh20kZnRqaIi+3M=@3kSx?#R>L`E)rDy+)x zJj$M2VT?9Lh@y&qnze)TWd6$d-eI!*OXJz(ncy#-l5LNZFoBUR{oC4R+PZ&xU2Fu#(QkeGGFi_I7|t6O0GDT^ z%bQEdvMJznP-3Y)X5eh`?tyAGuq(jj3#Gw2`Xspsks(zTJ2tx34VU73>AMkE!w-s7 z*T^j>Vk@>gH;MV6F#*XF!=eAzydyiG_eac4a5sXa`s?33MbiYk4up<1Dqui!M$`Mh=h zZT6P0Wegk>+hTgc7#%NYko+V0lvkav^%_|XW&k-_T78ksqdjIjuaK2J<71yg^kG;T zw}=OUwzpc?IBUx^X2=TjPeHLj002{A>`@g;t)`nblQIMy&swD6FaTr zVL@@YL$s8$7id?qcpE}7*$IVj_9>EB1MV1K!)>1Xfr&Hq6k_>ws^p44E@8%D%o#0j z$8pS=#oZ>72(NzgjY6k|h=onT4$%cRma}1Di3*eD$|Ll2-mT*D2%xitSbq*IXkk za{TksNy(SS34x&gi2dU_O16(8dnjeR`%P|h@F0kO<|(S8*K*eR4!iVmcfgWp*rhr6 z3+ZY9dhfK!8vUPPmorf%;n>39i)dV1#g#rtqJUe7J+N>dc0tL38!Rm1*0`cJR;xHo zs&U0+1-tq$*23-SRQ1vW`+)H6x3_+0rV&6c0!0a<= z&@J$pe~`sXls%q{Y0!O(V|m$hx!;z^g0TJwE!{PD7?!;8VR5`cchJT%EQ(uP zyXAQbgH|yrny7MRT@r+F5~TeXkP<#>{Ivks?Fwo80ouN@#LNKw_~%oj#}e3g-Y){Q zltxZiPXqht@9Ygv2u7o+MM1xV!B{b3P`}n+z)ApIMcElJPb|(T|1!qXD1A8MQuE(# z@0J;7Blm~*FOZE+?B!ONjJKAO!M<<~rNnxR({ILeaMy1qyU+a^sS3!QvelzBqzK~n zT2|hSJ3xPWRai_Xh2ouL&Y*i?z#aGc>6zRevP@aNy=OwRfnq%N&pZ*cp6q+d$c|7Y z-%E@zN6Pc0U`}i;ETKowaf)oF=sjWDKvQ@-EUvM9Bgto*hpJsMG;ga1x=u(aT_sXI zWR>vobkH40;2*iChgCR|dSeXe0UX&EF(Z4kjGn-^ug1g+{3{p!#x2!z|OT`8t)B23~2HhJ}0l7^5L?hG%T#mdsyOpm6wxEZvm5Uk`mwX3> z*|7JsEM%2b%W44Dpf1S?IKM6JAxn9O<(1&ZpP1IHSmk+N+DTuIJVmeI9c2A~XNB|Y zcivV0`e(mf{6`@zb0HMn?Lv5a|CfB!{2SdDOb)fft0xRm^JTvhY(b-n$Rga6^(_ zNAV}aKr_I~GneMsk)xn8;ZF-)I-a-iIcR5HM4uj zLD5!nh(0mx3t+RZjw-aLlK2?k`LnmLKeFEtSat_SNtt)vkCEE8;K=Jn-=*UlHw(ya zX5n4u{VyY@Ogx=NO17UOb(FHq`!>Ckb&Ra@$GmF|7?2OUq-8+1h`}0kg0rFz(H+uL ze2hk348|1inyH`7ZE-yWr2c)dutLqo`>$m{WqgvLHN45M!EX;pA*~X3dbG!&8`a6Z z4|(DrbxChSr$9NyuMpPE$__{iYmn?8>*QwNuVLiJF}5$`;pF~5%Z)IKIrg&}lHmIc2P3encfabw4z5-&hrc;+q=E91fAS+`kW{}ATbR6le*AM)(QEoDLUVD-(?{^ z;%y$7>VF7ga&7a^vyMtjqgO>83{LjUVW;~xK;^6k zVADXtg;KVLg`3)#@Zi`m-I&lb>bTqIqGj|H|Dib%*bmc`^84(9X!q z0S)wks2jGnR)O%|MOIZ*t|A8n$J=7sScsVtuO@OT^wtz)=5BJ`GhHojpfd$^+%29+ z?b|0z3rmw?4SsCMib>VNHU3S!m6HagJzTXsN7$~|4h2PRuosyLJRU9Fe(4qNevU?n z@_fnEpX(NWIz%#8^;esZV=c-n%lLFDpD5R93W;Iou>*0<05$$8x5m0_3Ml6(wz@vO!3 z6UDQ$UEZI#FL#IwtREMfkzt%^UNE?0x@8#|qs=^ttG;^PTv_CKi{ixdiv|6VVgA38~(%^go&3ACETg@6|H%ErWErznuM@9xfYEfss&@k=OQuL7BszfXVZU@! z&@S&@s5q<#mauI4A&pw<-;!|?aiJguv zlW{&n$zb27nNkjsZGkz0ZK5Ia>?loob^3J>WKDN}=4hY}cx@=U8d?6wSk)dHVL^Cd zcun9^MSDzj)b4=u3zx)EC=Fh~?uUo8>=yA#a)qnorb!R7cYzzLgFVr7-wG&!e*=Bq+ET`I#0`GA%{&pa|oiQNq1~zb$aZf=0z{;U*t#L>#Y~zl?K+ zW4K<%s&tE!-IpGs@h;wLk2w&eVO@7GVdv4yND`}Uely>JANAPAVt7XK{%n zC1PgV}LiWB=R7S=f1gG$2n0a@HY z?1n8J$8Hdg@@5T}tQmKUjU9Nyvc;?VKS959d16Gq_~i#Ab`tR1kDPImDOn;#R#Qsc z>d?PrL-_lVd&114B0GOLR=opj-m(MpoGt#2mf!LykqGAvSAOrk&xa=)A#vk>E`3N= zjmDhqyuUvI2WiB@CWDeeRon(jX^5>NjY^xmlN(1+*iJ1+#SPSCL@AFQ;dZZqv{%my#)$-Xx&1GymVXjvGQJv+w`lRYWi z7uf=-2OC*~A?fZ7qygfyO{9-g9D(TvXXg|Xs3?JU>Fy3OnvB5o1XO1K+80KK-u~#K zi!mqZaoRFf3$!~Rtm~1?d8S>L%t84e9sSru`)}LoHoJdvJflyRjZ^~z^jfi*t`V;WY4ts3om%D#Ie^ozhN-|$WNSfh#j$VRnip&~S3f&-;45Zi z$elNCpq+lI(44K-X?+k2{Id<*b95#*%PWPG$t|0IKjIo&BY=WZ-)8ZdsB<)4>Et4@ z{uN;X-4NLmvz&jCwb?7)w>@Ud^j|=CDH_9Nc6Lg~ zgy?73h9mpJn-QX^g5&J1=80++^wp7lat362gSn7-ZM9!SI=)IE#iD$5|oYW985W3Y0#`dgr%kqz0T+}bzVed{(j>d z=CL%Vop>z7&31ukJJjIRx^;3<%tuGx6Ji#01;G_?HYolWcyt_LuPi{!3eb+2K`eCQ zfXmE4+xOvN@$Zar)|IQ0qDU{Z{W~XK&a9a@G6MUUk_}Si0j11jCCbpn2Ax91c3zox zzob$!`vo(!jsUKM52s9NC^o0oa9Z1Bfh+?{#FmkZq*n~? ztaq7_gnL%r&~lvf_H*7G>O9E<~g^^oBv zOI4_laPT_l0D&0L=ll*)nwwVCDh3k%ZaH#0?F4Q(i1E@TUfAPt`B{+W%@_6&=&_T# z&i@Oa2D+0jkJuLk+WL9#>5kEuxuT_lo0k`Q({CqkOwhmB&9>hfkJqA5eA|BG%P%YmuIf+&4hUbl} zSf10G8bH1rCT9G`!nH69tu-A++;DZQYLedW&2{EJIr8mprKQ1W$FS(_Y!7i1gmgjMk&g* z(T=3SvHUiydh?c>JM0d#H}AA5aR>8`rkKk;FbOR*X?pU^S-+%lszcY{2d`V1)g}BaU-zHc;_tJ+Vdfl6WHf}vx zHx>Wg;Cu0nyFe|MrzoHPna{Ww3*$&LvKbCQhovt6;~UZmJfqF{efi|SkhP=P3@4sE zicDfI~FOjcHlGivN#ajT=5jjH^{p8S(sh9WMItoA12p+s&0xxJF9x*u_ti7P+s7 zz$PkNR*(j$=4#-Ud9Q!-ig0sKv!X>@!(B%1(3|DCuql8~>eVeF_<2Lz4Zg^;eo&3O zkcCgw1+A9Z8822r!OYzp{nWl7v1~O)$%LcAkN>ULTw;oe6XL{fs|7(S9RLNTs6fXM zOE}_qpw7_7dP1;cXOO&XXsrj)%X+f!{WIHtdd%E6b=pY11#h!3Kog8H&|3cEfD}P| zNUmZHYiQ74PS22y)6 IM$E$IC>fIyD>8vEqPu3uR8z9PN)X{pi8=x zx+l3JTmf5pCw(8f1GPRd?%no2OlvqhgIdH|*nf_dc7&ElPI>lAD*TrE?TOA7l!l-6 zN+DX&G14~owl_vX9Az^Y!DY0m#lAM-X>ZKnGQVWY6u!|{U6pn9ze(bkX484QiLbha zl4Vk4Bc;6US~OFiXKmO}ZeNs35eO#8e%uL8^igU!=gg7lkEEt7OSiziviKRW};a728b944xNS`g6ziM)M zP-ReiOo`W0*FI7MGP&`rOAblJ9PRsMjE~VealkcZ_{iv4|A$u{AL{6`kXA8n*n*t0 zjth&DOMD2Y*Re4bAl1J?G$1-5ACxAAc6pW(R4V=d_T!H~`-AeGU;j+`mv?^iyI(7f z@;#$0awkrTNm@P4=z=H`SJjg(Um6z#^l?Xo%XUyQ2)t@3WlLx>=MsOvpDK1O!DaVK1W-4Ew7;OYez_K;l$sQ+8bUkY_?1w4YL5duplnwN1(k)y) zd-2nQwwOBp19Cm4>z_KmS@!1Auc*Je?BC1ZIkvF!r>cstA{Rv?XTM*8|DKS?!FaSP zGG2OTPC*cKQ52x>>5#y{H(VHi0zQ~b)a2JK#j`GxM;`;CS8$jef;Olf=y4d?8%6-Fy?+X}uH++qx@!vYC8xO^r z)G-@Jbd1dy0b=5w_CMY?-^+7aKv>wFI}miA+@?|ALJK*|YM@oacqvw=j?;>F(PQ1lyfwl1}tr3<+#|P|uckNd%e{C~$ap6{W_WV9F5K$eq zNmjma$%lok0(wJW@7#XrLCL9U1MaP%l~E^!;Kv{-N@;X0CtJ`9jMT#}Mvn zN8ry|fQ>%OJiTgn%w_+k?qhRH^5xD7!0gpF`6)i+uh!2u*5d#iFh`(s97)LyBgp_3 zb|3l|b@+bpM{;vmylNB7GFa*+hFs*9`)Q$+I6kD9+UoJpwaZ&eUsy2i@eykW`bC4{ zh5CEog{B5K{X(WUGY@@%;BcaU_0raNZtn0Z282;BDvMOe3oj>?gM^L(|uE+`X3 z!4C%Caqojh_g}-X3$}>opuXug`V?*G2(M{CvY)LwX&d+Nci03)*AFKt{$wmNZ01q820?AMf2GO$NGDCHMEjhtFeA@J*`vb3|- za#H;P7or;=j8HBpj>r@0$XTyO*gL)%lqtULd&Q@Z?iUtEw23zgu$(vz08kF)!1~rv z{p?QpC-Npz;g;sM(*faIJK^AkV`|GyX7=dQOaJrm*GAObj9NF7++=2toOo5V!lXX^ z3rcpMBK?$d4X=dX$~ptv$>|;~;M2nH@_`wb_*cWVpqsv*)h+$R{Y+G=>#Ar7)pkoY zVL(d(73_6%uXs?JCPi8d)j58TL=(0OG)nVnP1r@EopqlKNeUHzt)nrDlTJNe_;BH^ zDY+mchplSej5lh(d-LsQmsLC27g;C?gqLu8J5AalsuDw8`--q$`pJ(}#q#Q>Ug%U_Z-1Z-pFGEV{tZpj@8dz=idiHE_e!h#g9 z2jmc4y|69jz@jU{zjjGWKxE-e43kO8Dnnr>i)HtnW@Z-avx}}}<~3m!%B&2c1eu~E zJmlGf^!7CAhRK)QK=@IRFH{-x-z{+Yb)Z{^iz+(g}=Uv_W_q083=49$$l2D)p>h)T; z^8|(1w?(B--|Sx=gETp~6`?ARxaC>IgS-gdZPzq?`V=KDO6dAp=^nH*b>L-_a z*L%l9em6eq_>s3|*uicw97N`iGqP_J|M&9`z09fKn3xggHIT#tQC#uDF6lw=XMqiQ z+MxKLB$-o6@9;bq)GevTWj!RxI37DAN0kANMmBT!@z(7@VdmUhWhQ=09wh_iP&K7o zJ}=8ZlMDJeqWykAt<)jXFJdnORaw7eXOPDA2B`@s7s1!sMAZHo*L{#ok2h}f8-)XY zj<(Va-mi&U;plIb*>~!+L`#2Occ;U{=-$42i7X~5P8{VcHNkHVC4(g0EJ}$X+73~n zaM`T$kdJX!tk_J@TTw_f%6rId~Ov>H|) zo$*i0LT>)>iZCUz(6fw%DgRj3u4d~3VcU>(YwqDljo`3E%~rFIj*A=0eAFH778wA^ z$b?X-(tCqbjPpytZ)z8lP0aR;oY;g^n#@MqC>gA3aw#RoL5hGPR+SW56TCc3Cn{wh znAR`p^48ahp`e&5n>IKz-O^0pMA#QpyFe{ZVhy{b3aX>>!L3OSen1@O!q^0?Xa4t7 zvt;js|6TRJznE);N{)YCIw@fW6(?Tr9yftX9VLSr#66S}gTh+3y3kJU_4(>=-3rNd zYm2E_wB)1SKr8^b&3_~;Ux-&@se9zZwwPGTa7!)62yzcs2Na_1F>5_nOv;|o>%MZ* zX0L~&2d=G{6wB+EUYxyhQd`Ub)Do!WXSf5RwwRsYErYYM18a-`dIlxaqcP(t{2R0P zN{s;8m=b)Rq@H~ z?uG>qs(ZiJDo&Huad*0XCf3K?ud}ZRHBhTq$3G)$<-+19x<3LL>&M4mwQ0=9Jfbh} zfRWoWF>=zTm4QZ6a_G(V)g*5clyQyBL)uNrc2T5+QYLUpAyBy^T%Q4L$cbK(uh79X zuA>X(d5SdYHBtzLqN;vKrB}Hp1u6FTloZ|(Q#g%@K?l%+{;yY>Kvcfz0aS!!Uit`FWL1W}G;4yt6tM<=@;2rBj~ z_CrMlbdPcEJnVwtK1E*+_{>LD6Pze(2(2eYyg~Q75qBZjk2h7Q(vt=WpPlS-{$Ys< z6*>mpHG+6?>)7QN?EED>{VJ>kyv4`lEAzkoh7m$Pxp^v!>~-RRT)PP(C>`ACA#?KA!OqAk_Ih&Dr`GL;cxCXa$Q}vW(h_zpJe4F! zl&uR@%QJ!-*-d_T6!?=06gOX68gZ9xv*mOYJf7gh;`qPh1|vgBwY6XQ_$?!t-u$@# zZF1C!!F0m}OcyBGIf}GV$~HMJi~EFWVU4hBg_79jbDLX_piawGW zRTrAWUKN?|(>+h^U&0@ldu=YpZ?3x|B0EGm?t5q0Y< zIGi$?jwb$X`~Q6VJtHQ{{Z2|q&zHtxB$yNsJfLKEDRPHWqLw$ZXJd6yf#NKvu_9^5 zao`n$1>-ex%R?Ya|p z9rf~vZs{KPRDtS-xR6yBn(0#^L!KkNln?cMSNy9KP}Y6hbvtL>?5#3Yktgb)6$y%X zPkyg7Db`e#1tvqfS4u=-Kw031(0-}PY0C!+(tX}G--dA7h>wMhg--68MSV~p@(5Cq zwGdFaD{A5HAl(ERr@H21)#Y7LBm0W<0NKw`D^MilsPyJ*?J=lohpVV^dS{SB=Oycm zl(|!4ICrG2Vq~PK4=C-Rx%QL=Ij1g3m*iB?sh|R=!71W3gR*3$PeIrz@BxsN(xLBq zl=o%z>mQ}9BQG&}{m$zYvj5X{^>>Yv*UI;6?vWGBCNC$B?c6ePe=bq73lup=DPwu% z{M~;2P-&*mtJygdizJ6#w)1jj`xHqmRS)DuWbr^>88~=B6crXmSXtR7#{g$tD9&jr zocHh{9W1WO`8a2sqJbdua^xNNYosM~*ahc4nBVv)qnfX%qlbe|N5a&z)>Fg6pMmqu zCyGjO+GOD1TupRzg#h`{I-%1RSnm~t7Xl?X)vi0gqzpz_x z9_cc-J5D=ASulvWWCpAltEEtrUSCXwaXegJk26^nlaYSu@YwMIW(M)?UtA0}cU_ny zVL9=(xCKWiBWNY>l-q?jp8RW|%KO#B|9JG)#-Am=dHQ?(3opN!{neFk?t16pBHf~2 zory`4;`2oT__u)Q}gIu){VJd0*%MQPK=#m};3 z$jwH_A*al2$;FnYRbvU+Ik7FV5VFI%%@%GI)K$l`ws|zU9-;?+w2qY#ypG)-vo9#Y zx5fSXd{8w~EQQ>f-;N8CY%w1WfzETCwtIZ>25mGS_rLFdjKn#y`Ouh{kM)#n9Yxks z%1WO>58QTaq1Sj62~aszWvD#7nCc*U6_r6 zT%gm~ERT&oMGp{MBHRdXlivUdx+K9-4rFC#gEWkeP9~qhkI9*C_;K~~d*Pa@9Rh19 zzO~^)l9b3=w}hFs9A<%AD@X77>D`tq3W1ngKj*dDJqbBcp$D=LbfV2z-7sfwVA-xj3=E@|g*K zu2Ql~6uCet&-#8U{e*6h%%kIo783c|Vp9LL4Sbk0G07afUc-%pN_)fg%lhlb_!|U0 z{KrsKze$uIh=~uoB%4G-P(-^?V7Lb?6r6M3yCf9>i#75L**WjU?J?bePwel2uwJxV zg68#O`cvsiQWp*k6WklXI`gD$1sUdKN5^xj0+xjgkaONh?uUKhF~}Ym9$WlU?Ai=I zTkPqGHZKF@;WrmVn5$Hjn2e2vl5L?#CZ+6>c9K}?2I-zx9hs|Wi>aNGFr|P#6H`G3 z=E8o8Yxle^etXPz;9kEEBlCC2Z}B=B`O+8jvR}}O`!>B>EBYPlBbGnxQ6z_)_gjtn zsu~Wp?E9OLqoCbpgS(|5`yz-Occ2FW-IVWtL2WK7OervXg15y zBrwcC^5Z;3RHV^tY}86Lq|%8aGC<)wVnx(M$v&n?1Ep;KIx?pnri(?NF6aX$v~Iu0 zKm=qsPGYIq=ficP2E{Q*U$`n9WG6x4O8Hl=2%E_9H?Ihb1cv9*q$}L8U8P%AGNfo{>TSg7P7h*{< zSQqn083Bn3&}>PWZv;}d^ruCng4sgKdAH@sX%pBTq-0R1Q;(b1HSSt4G=<(Ov^^`r z*1FvcYh@X>BKz5iGF4w>59Ifqi7E+iCWXTLES(7bu``eyRq4}9x6mu8^?}6{Ze!;w zibdP$p?NCI15T3`!anmo$!@<`-sYgcUV%af*!k|4+z{^#(@ZX=ig;(D&Ot7gCTt9K zI}SzN82AAycyImZkBvA|-{L(cXPnp|^_w6=N69)Va)nYJg~izcvW-rdl<;P@0JwN( zCv$GFmQOlF4@PhCu9@1;R{OUA!(u%=S2a6@)6Gg^l?C3Je3ErHA{FRaV|iM?>+?H3 zia7@b36lmySz*g3VXJ0wIqP$B7ZS=bV>F+4Frp(=J!f0Mx!LV8RnS)>f$eZuyc>E{ z%M&IcKHKH{jqMN9Nv!R`r|3prk>XQf@pP2+!rut#2n;x>mTB6gjc=Icx&gy_Bqq zBD*Q&QZG$dZ73*bfHlZxw{R~))l#Bs4!e)Dk~Ji$4Xsr)DH`ZY(Vs>4MIH)-@&AkNKRr+ODb!4M3 z3mAp9f!(tD`J2Pypq_F(te{N@dC}0?=^i75O!%qh-HuniT5gAMhvzQAGF}Gx{V`U# z&t1`D&L5s#OOu|a)7gb=a6V+$!7VulyhmwaE!^7B^`tnWMY2mU%-6tbEzAE&W6ZFu z30w+NDo1#mM18dO%}e0Q0R42-?ttZAS@QSG!X8NhyFUULCwwbBFD!T@?EXeCuuNP6 zcKRgNS@Jj4%I`vnCLG-dI+Ls3Po9TG+L47xENj4N6S!VvfvmY_1ppa=i;CAsNT_k{ zT9wwflk9Nfx$Bt8+*M1-_EMyZQl>;`>2sWJ$q`;35HNlML);!y$SP%H#9J-LT7)5& z7RepXkV|FAeW+B(2;LKYF!<)|3J|4z(iWd;pa(=rp;#5%!flWg1wh0*s4^&ybBHdX z3;eN%PZaT1VZ(}H8e_KJC}6SXhg$IrGyLrPmxgz|jreJMr}90L;l%g>S7(H)UP#Fb zD6*AO?s0FIa z0^-fhKA1&3?D8bk$Sje=ijjEs$XIR7v!3jG#0V=B{;T=%f6g&trFzz1R*>zZC097_ zh8sCzf}eeqY!5{$DCO0+aPlerX3jToUA)Cp;|9_R-^OdTZdvbM5vEaV-#q@qTDR2y zeDt-=i%>}67`gvidRRq};n@am8EjwoOV9Zyg;oT><~4S8&VTWc%VI-QIBpX!HZ+H0 zYi6gHjq_L$HH!TiL;U91Z)+J*YZmV5Fz3E;+Cl^f_KmPKPf2-h&^dHIJ9(xWO7k(( zrY>}cSGfWOl*&O~M}=$xSa+$VwSGms7D*jUUlpW*yV4DcC3le<;vq>*a2w0wPxBamrM4pYW9gZwRWR&gu3DWj+Y;#ibFG-z(1#b0{uU-*$ za;sR~l1JPMVUD1Db_QgHlzr>i!ZR`Tqy%;}TKIQn_t7o<46?)X3b#496c+rw`~vnd z$FIt41(h9e^TICfzdyO}Rj*MnFamUPabuy!D2fg208pm@MNVpYF=f!nPW8tUY{$r8 zkM@pj81m7&vHLYfc4Xpz=^p=(B&tkNmj28tw8O4R++*Ufc^ z9{c3`qn^6jKLa>wFxay%DAnH{eu(v>$^b^zcWQ?x{~kImcd8K>Z*BkiFUe*nUID>^ zd4yk5Ov#`=rQk_CRz%<}>*cz4?shzdLZLeZS_)5wl$Y20~QN2^VXovBBt_wzXWI{#xhu+hSa8V?# zswZ2V7%spEJOVB|C>ik6YboW*;A5Q zQWzEGdbr34a*mZ0oF;ANuZq;#b4$QvY{rZiliuui!U!*v?0bLqeATO;aB*W~GSkv) z#c9$(I&Z;##W>ep>qpQ^P}py@t?Y>rC?>xDwI9U&V^K=yv$JRH4ElZcjKech=Wm_y z=X~}Jaj)+m4grVIdLaARIwN~Vj$(LT>ip&Nz{=1S0ogP1**k;U#Mq()q@%I}9Wp5y zhu2D2jXp0~2@zW!`*w0g^-SYz<04R0k$fkfZ6Gmk#B5VR$zXw7igagKa%~{7yByH! zqfg#k?YZ{fjdZMe!0bBcjzk`+tgv>XrH_JIn2y^c8Fu+J$WXD3nV{g$AWa^UL)8Rt zi3CL+Jgwt*(RTG@$ApCE@L|iCkKTrnftesl4tqy##7)expVg3@FU@MF-UMa4C>fO1 zY^Ri|{#X%rYu--Q>M8NQNC;d&KVb|x$Qqt^Qt=3)Dy?D^-^mA3vhxc+UzjJ<5!FG- zscCtT{Ha3Gb<{wQi2g>>Y-+ zFdVj^z(QAcy=>Xe{Bi#Mbn~onCXrYtuDZ35W8UT6F3*a_8Am^7HMn06K+AAkp6RvL zEt!LH|Bu=Dx2l4)lLq&X+4?m$rpTZn0j^-I)y0~~1Wu;cmC*ZIftjYYv!H&{`zOXT>M8NEps816VE>u;;=|6j>VQkE;(d9 zryP=9knOk0RVOM3HS=+rZk~;bwZre2VB53b!~Dqd#PiYTJ(1=$ZA>N}C$<+mP3%Q3 zB?HRn%|K2Mr5{;P6L8(Vgx$>3y6tBTNo#?9H5M{-JLzUci@2B0kgW)8WZ$MyPVf$t z=^t`^_;QMP^Ow&2+~P9HuR_-*TjowH6?|s*jY$pz>fEvz-kV+{f1A4ncN|_Pd!ZpNT)$+%4@6FYO zA*VcY+u>FZDg~g_?IkXuj@jRix8N$vc#oYU;OsbEd!AZ@eb$DiT2KQS6ihmrB~j7$i%{1p5A zSWJi$uX-$)5Ty0!mg;$$wKRygd1i`k0aF?B@99Ic$g6E2bap(3jDyP*N{lgKV}nDX zY>kyfKnBwD4pFx>Uaa;|WearDLj+HEuzSOoWYd|VBfOruD2j)wdAa_o0fH5LMM5~l za*AFW!>=6VrG9 zvH!EJrVQ z&r=|L#|BfIuw)&!ZR&bVjZS?-k@%{oDm4J*pjv)jS{zmvlqsuq+a${sHIq9KM$1!R z;SL(9_LwxcQb{k}=ve~Fp9AEoa6gc?)J(m@X%XWyr}#&qk2~}Qhoul$hkzH@vN4Ub z<;IMx8?{q=;X6LYmGMuSC!8Uf%;NG+yv~6_pAqZmB1)#ENFJpu7M%;yC^mZ~&R$93 zpPsp`Ae?+Q)R3~UY}R>x>^u!a!4m?Dp0 z1`4Z{A*pgz^v&0{)4lvbwi>uwQ46aodQgf>=uU2BWD~Dj(&X1f7qB1lzu=ySYST{o z3(+Dg~0Q&R--AzyF@r9Ck{A$6gfT~B%9Pd*ZG9!Q735G6z#u3`EVR@=AG8dfo@ zNKhBrz&;+SrTfI0qN>nZV3a8Yo|%)1tKP_cwQSL#JJL7cZ}Iy~uUy40?~}sQey!|k zdY3EE^+RWWa9UBvx8Ay%Yo)2`3exzePrt*zD>}zX3hnZ42~MB>Ncsu?WZ*egd(7p? zt;BHcAwdG}s(_`e^<xAZNVGg%i8wY~+ zFX*Kmk?UkWfD8~a!eE&9V*hhS2$|6Ranr6>JzEYlY17?z3ok6F5}=rsmhJ-Gz#L%0 zR{OVwYFux?Hg{QM3vgOu)kc@31vJ%wWJh$}t(V>{yzPBMIN*g(?g=?8SIbK!ZFZYq zOh?x?#EbxsgHpB`A%C> z&)8CZn z(;4c3;k}V`OOsgpSv^pe_TGCA@QlZf7d>k14w~8DZ(^B;V_&IyLkmGyL-NWduj_Mr zLHcE#zn%sxOR5vakwJF5{1P9w8PcShum(v=M7ykKZW8MVStTBa`NlRN*nQle>(H_{ zwS0o#|Gw(09|IgZM4M-*aDM5ebEcgXo}y!Yv!trs^e5hxJ|DCD*_kAM3efk@j1Q<{ zpO>j}Wx8pD9#v5(kmJ-Uhtfm&phvS0hS~bKI~=34XB#@qXP4&&cgz7~jFHHtRsYzrmhmxr&l0_-wM6rG-D}7ZjA6@t@6qSA;Oi>gC6ezAk z#CEM$ne_j%_bzZzooD{IM?AyKi^JR)oH+pm86bls;o=Zc1Ey)y%XYiF?f$pj?Iles zZDa54Z`N+org%pq7Zp%J29TQ|ih_Ws!%Y#Ds9-=*yn&2h3<-iFL6raVFoQB9qjO+} zj_x*}hI7uG^Fq%1o#%bt=lMS0r!lOGZ1brCre+oYymzIfV@llH2_b6IczN~!SJwWn zINX2{UGVH=a>$P3u@{VHy`EyA0Z}a#iD{dqSmP1H=p@G=Cq3-Y&ToQ_;5WF-1(@=y z`C8n!_P=@ZyI=mGGy)c!Z9dRTCNN*LP8!d{Ea0txlwj;toI6*`>ww00sDq(y3tA50 z-JIYL`{XFiqfs0UuJqR^ZUr>RO2~e%G*No6dV_lfjSsJOzasuZdRaK^P|Z8yJ|VYd zLijK>IL|tNHXNR~gO$aZ?kQXUowp5W`dxGKcgYbuMpK&+nm(bJ28x`bB5%_}o(;0i zeo+BB&r)}t1m?L!)T6$`fG0lWi8&@v9s|7Hb}FE*yRmVOrUX@^sCxoC<}a{ARbu7VSX1Yi&F0&VH&bTZ z=Id^_mC{pluyF|fb5eGRlSZcn=+D~eQg>AQS>;>c+vij5WwRZ&@gzv2%6omxfBfvl&w^5Kl2N6j@J2wn~yE1Cqp<>dug3js=wdA2znZ zS`%RrB&_wzSd0_xQ8fJBCORub7`$7|8YMw47YF%V@rKt)39g^@hs->=JC!sFhK zdh5S4s9O;OL2zBZ0X9NubgbxLMB(`%RY}Gxj>kw^bMinLXTp^gr5vocMM%%wBVNKYd-kP0|lc=s52y zZi(`%;)5Xt0&BczbN``_u1{np9zATs19mfEw@IVO=%_DDC2`Ft&-N3f*7x$##9RG) zWKjV}-K)r7k)Pfsj}lacty|C^dTQ~fZrE6G#msJDwfkOYt@AyPhYlxI=Sa2tYWM!o zp4knshppxzdpB03Rcv*S65NKE`x)*Xu9eX=5!{UbMYb-8UJ%^;`;w1aeGF?;|Bs|= z$PPPRo2rb~rXq@g;7b8YnRdvr+ugxeZ_GtC1$8D@CtHePy`*4>>*~afUi-s7anm|A zNcBg9+*saLPPr2FD=!NZyqBUDWw&rxoV)0XSCwoPvrBwPsaJNlHYvt>Xtpk^ffbsk zzj^fjZ%x;u7a>QWm%7!5hEbhR$A|WIP+f=h2uQ+2NL2yY?$gCWBvj^t@w2kI8{bGx z1UBQ3WZkRgVYBM{ys?u& z{GmfTiJHCFWurXJFUcbr{G>tWhQMC>n&?xXB#%y88w^W<@}!SsTez?S3KjhQPWsuZ zxAt2H2Qms{VGj2aB%qy`utyzxv1GnSUW0*oD{Z7&6d>YD{=-JH>3nKQ^-4E0z;2K;^+a-wMUe z*V_60f@uFc?pmi}dJ`F>uZS8&8^bj4N!^s>4@aila^UFgjE!*e#IOl&f2p;QW{cJ2 zU63p3rY{Fp`uFmpVU5?ka#yTwTX;*XuJkV!A_E!|E@~|B)_a95z?!cBIy82ktWwpw z+A`S9_8ps8^e5hpwJw*jVzyIc8x@IC zAn1pnv~8dKu>YFiJ~`?=;Z7I&rgq4Wic#whg-hFM{2>4T0)X%R`JF#SzWbYZBLDL4 zZ~yR{NWG>$C0G~aFx{}$tXdmTONJd@5bTpwBkaUc%m#|AqaqO$nAubBxn{k$AAFMP@?%D0G)=auCU$DXH5@{p1Tt<33wMx=N|Cm{}WRh*B z)AW>CXZIZ!?NThA47JUU7Z@|twjoy(tF54QARnm4IbeC3vGuf1Ws3l|IwvrozT}jh}Qi?AKKyX3CGVRrg1@ z)Tv>+!Jr_2x|MFFHJn4r-9!U&7E^6kSj@jw=GD?c%v+S-{%6vC(^{`5kzvQ1R_L@n z#tBWKn5`7dLL|y*<`>1Ll1a@0U_{a1Hju!Ovj@ zSv`Yf8h!B+DP|Kz;;6_z@m2;}CP^*{x4gUpSb!LN%5cWmQ;JhGHC3;8^aWV_eH(pX zihi1Z)z4MHblc`H#WM}K==#;=zmauzOceqp^D!&MHj05T&1R&r#*(g(;*e^uY*hn2 z=$z_y%`Fv#vViR~dzJLG@~m=??5yil0@q^DPN`4IQpkGxOPMj9&wF6(A&9PFgSbw* z8!|lAoB?vmWh#;I$QoG!iAOh2;Q>24w)vZXndW3L7w@aj{h371AUQ_g%B>UwYP_4M zNbJ6&=Uy${p~6zR8{B64n%i?U}eHPVjbA!fvm8~6`=hkZ*xR6?WJM9z_6-%SB|oF4gBNrMy`wa8B^OGvM982T<0 z)6tyb&@yha-)$Q0QIGtL3h!Pvw)jHa-rIBe z4TClL(ap0dL$)-${PQbF5Mtpkp}H0)rcYdJYhjaU^;JrG|2K)9f~H!=TdCxqwbRCao`@{ zr~*MPmu}*ey6C1c3v2P6(b1J75K9E?oh==dx8Tz}?qPR2#<{kTC|(lV=lw zl=bVau?O3%S%J9_1YQ z)Y#zm;-VgA*a4;YbLMo<$rY^eE_JWuopG;GUJ_m!kt}a??NF%Gz*SBY4M_4urS9Ff z<)2L)D7MY;W&c3^%jaDV2NxD>mtGu-%~jNpwWTZPW~nZFtIyKuyrz+M+Mi2N!2(nS zP`9E}rBD9daLxc2VEs3)A)8!=#FgX<=*?mq2OAiU$zb5|s$zX}VScR{&r4}ot!QBk z&W!W@{)=SG3lpI)F>-TqDF%qp8B`>Cv#C&U8vU)S;vS;o@ANH{?Q}c%DmEsprz?Rz zpU*uM(#JhUYGuW;rr`T;^pI$h&pov`{#)vOi}#Dud~ATG$2gb_t{#tK{i8K8yLXso z0PU7=WG1KA&ae8#RcMoxtwPQSN)CfKiQtwTbHvy9S&K46O(0kDsjAv5nKMZ5CTB#k zBsTDtM%VAYc+@@eJqR(Y!Jr!xAS$fDp{6ul?|4g$>89zeX%~Fgq3}v zx)Q&*VVdBvThy3I=9ECw(j}KbF(6@11qZjvaW6e|Sm(FF4=Joe(t76vZ;hgT{ypw3 zNtw*1wqeY-nl(JeJZsBCcJPQ_QTxm1y-DR>)o0veUR@+_(NR@;z#q5#+uex&Uh|bd z9V5G4waytL;}5DA{|Q^J9|}wIxZ-<2(G<8jFxJ1H6uTWzq|7RU1i<0h*KJ@-*ysMn zn2^VZ)&~wdxC~s|?mpk(t)$6+l21xr7;ohhBX6aaVrnQ-Nkt|GcfY0$Oi&#oy}SgK zHt-X-oph97z(2|Pmbi{i3q|#W1XVGuzdr0SdbOB72JyR1WG9`%Br=ch-b4~qw?R%4 zxhNV1-VyCsN3V8|7Tkh$Q5%S#*U@p_yI#HGdu!6{1lC6CG1Q)ZG)W--=#!0Njl6g>LA#c?A?I`{vSR z3mSr_#y?<-5csxP9}eS;_<|rLbM4|Yve9*D_HTYq64|Xo_KST|Y_tw#Q_K#Eq*0Oj zOmsXi4Y;%rlhG*pK-oTDlncZKb@i-5*Z7xCy!4T~zDp)*RNyTSWgEk)Iat?qW4>)+ zKLNOpt}o+&)OwfxY3HIC)5J5Iz^xr;7tQ3)+k}mQ+4Sc@*{ZX`OfY%9q!QZrG;y+l zgJ0uQPvRh)XIm_o%pMzPm^^CJ)7izp-_AFfkj3Y>93ku3nGidU>F+c$A;}c8g(3-5 zWVWCVWQe=z^x!NYQtxxx=Y%pQS|{utUmTjHN)OgaG^$ElCwa!sDjRT2(F!^8=l%Jj z_g73MYo4UV+A##o$k-fG?&E?wXO=1#HZj$h2u|SQN&(7mQ`{cVvck@hMPe*qrr_AT zizNT_UuOjyOor@#)TyM!m3;mI*uo7A*_F+5(qm$7BcC!*4Xb zHJM$E9s4O}b}_M{Ub>wAM5Iw5A2^0d58menW#uD`lzOr|q#YKlkERk!EPw=Sj@fy) z`UfAcuKlAa;p9n2)sDj*S|c~5kYb>&FAs&}amhn1@jkhZ-tU+wPkar^%Av8+C1_ln z;C~et%BuSm zXx_k_YIAA$mBERf{@@2bL}$lN?8@n5OLy*4OfN;gpdwL-b%o%bOugT$PrgD>>EEQ# zg~Tv=flh4c*C+Ky-7v>BbQ<)CGcA zEdU9X_%IZNMbzrUaj12VqbAC;y^H4S=xC>GRi1CF`PHL_G?6a&=BeSk#>sNr0G zF4d@FmDS`v3eY;1$!Jwq}~aLi@GARlbLmF$>GMpSTT>0!ZGk32&CAD9L%$XZx2jKc~t_Pa;9NeQk_Gk7kl%(B0b z2M*BIbF5<;Low?qvX+X}{%p;g*S@vp%?od@dGlg~txUwTrs^Ep} zYq0j(Tmv57|DfYA*=ff+sAEPzJwP$Q$}gfK2kERu)lj!lL9g~7_SGng=C7NHayT9G zPr?er)W?*u5NJkaKYcgjo8Q6PupaTDgZ@!G&!x9J-VZ`Gl|(uASph3}iQuI~`98PH zPF6K^&&2eo4YTN{^#INa)6;5lAM7)Y@IF`J8Z%Mf0?Ao=%{<)3s81P0Rq>vP>+-cgIyj)% z?t0O`2C9FqiBr6z{p&f$yvnINk`nQzH#%fZ;i&7SQRuHX@RQ~=hijd-D-ZaWQ~AE- zE_q%zoR@qaarhfcPAiKz)?Rm6;OXhs$I2j^HSV|lv-?fec7e!js;%S#I=aNiuJQ7mhe zB8y>{0F`Cqd8_@iY3!PvL$}KF1$quV}Ar5nKQ6E>!JHi}s z9|V!OcK#_vKPeQ&h;jll0?N3#b8AC;ouRurPsM3s33D!iXh#X@lP5ES zF1XW+CY|AB0V}+VxT(r*%xP6SeTnSm<A{V)hXK{$xwAn#86oef=@G5>+0WEJ@VX!OG|vh9xOn_eV8ZJ_AVaW519aDJGU8A5oFTLd=<^ ziPZX>Ux!@lx%cIN#JADHl0NIcR!c@>_3G_6%<{DiM`eHk6J6iARYdaa*t-PRwMW=3oYJSP61X4Z_WV`kb&Dc`eO?U0!H2(&#FB4P>VqgxQcz2EnFm zWrbMlbQy9*ccDNYonYv~vVOk)(WU1?Ffg-p=|?WLsv4!1_x^H0z9^mt{b`p@-r7vq zo;+z`#$K@mw@g+D`dCpj+t~-X;_Y9w13i^LFZ`yvz#L;jjHR_k3b!v zE8ym9E#d2=M`6aJUs^6`3D-$*_VE}xk%cAswuGz zo%M--$OAb@CLsTAe3;5CqRz!G4TCV=n|s zrg0s6l#`JGv*X>585yv?(DDTu#Rd5up`N4AA@2@RkE#V0hjh}A+#CR^-YRL4@;Yc# z9uSv^V?n91gR^fc`pHcOR^tQJI#)&S1^>iUbj6IeNt2=hlpD(zWP{jk1>H}_$${mH ztg>N;LeCE1614*%sK#YLf>$QLyj%9)$NIfmHteMDVrA6r_uE!```_Px-qq6|$XB=^ zqfIBUZnUdXJ zkW7#nYFsX}YGDVh|2OI%=g`ZcRvEWA9dy(DROQXr_RHeQ`H2KeCpxqC3Z=5@S;xE# zZt4T3ejU-+v6~9@hGX2+JruKxB0EvYpn@O6Jd`0@8hPA(+>ANhz~!%esh`f_4svfm zJLOG*SG^0UR>@9IET`A)Bk2e!JHw?z0R^j{OH18zR2f_(YoAngstRn=TZA11#%^^V zA5Rp1PkWA)HJKKg@yc&ZyA8e|*2GL7!qqdfMLN2iO7+bRSu>+|-i`3JGm<B-7VV4B7-?$+GJ2S@cq*6>W zMYd3pW!(KjOyEUP%T9k&uUiWOUNgbn)&NVhQ>I;XKd8p%vM^DOWkXZtY_q_}s9~`o zDjpC0&&fmHv;iOe-}SC1E9`iwfdI&u8Q(-PaTM7ABsY07zyH-VQ32J&DO{-Kl}S#5 z6qJn?jK{BTyd%q|dAuiWdyUm{GJW0rH#V6%r*<1wHREn>BsI!3QKhg6zU1n{S&jTw z$?BK4`GXl)AwX_G2d9oMlI`)fRmI>q^Jy74M!)`3S6<2v91Dtnw#Kv+^#$=Y%oM3E z7o>^Tg;o0X%g+f9fO=!5ODnLIK6|4jyhDC?ULAd7ehtLJR`ZT}?xeFNc--aK>(&yU zM)%7u243WLdsTb2gdbDp(mnIq0zdLUY@k|wB4ZE&vU zWz!g29SG`mJ}&DIos!VrM1U~)X@2$y<_*p7!+!ri&pUwx#S9vyC}N<-V)j}NrfnOg z8Ym{hC`Z039@=H9yIe2$*OMine(v17VA$c9qfYXfxDC?3iE^7_VI)MEQaO zYNJbqccOf)Qyq<2Kbrz(yeYL5MB_bc%WLm0|5?^_11wgp4X7o<4&$N6ZWVzbuRC_~ z<0xhWMb=S~dqQ;Jw?c{GXqClOZC6bu%aVCFd1OjX%^MFohaW!~^0OVsJ%opqt+PNDj=z=3rqzwEsHBk(8U(N=9hYVK zoKg(RtDvqmJ2=C6kWPWBt_Eq3BdR;K^HV&uA*1*7g&y=nNCJVWiL=w?5(lLVkf486 zen<%laS+o_pjN-7J{PEi{E*i3Q&p*yEov=++vwUf$)BuXdMUc~wuOS#m^|;Ki`+m0 zVAjP&sbM+6xpTEVb)El!q)BlWTCT+es~eiR=Z7So9iFg(?EjQNn=bU*!WcE+UZ+w-z{ zhSg~9wtqWL*4XiC1YHBi?0=IeW-~?Nu`!Q!PAC6!zeWJdCQ=Haw5D*p(j1QA*TRKX zo%W}{eEigMU_5&)Yu>>Q7}>FJ{flYM1)EjIj$^fEDm1pccJp!sRdin-6c$Zfdtc!UB;lm8Y` zA@>e4Xu{Guntq8S%mpmIx*P|!3E$kQ}W~5RwoO?mbJkJ-` zl;e;zMDv1jP62U<=mtl1ydc}<|7hRSrd?b2jZJO_cctmw(xqhE z3{qy~735J2C`o5Rj}sN9SdhSj>JXK{B~)V`u0nX7#{a|VB~lUC#mV5V3+o=aq<=!+ z>7YZ!>v-4cF6uvOAa=lq9WbLkA=q$Rw3!G7Nl)a7q4%U*m%+$qc9_JQJQl1GW_#YIEn zzp4*>z1y!PeCOh;;#kf(@;7zD*YFfW3MWs|@nL=l}K3Phz?L~9+)5Fr( zAWEJ#uU7l)or?F@x*GuUqchW*NisWt*s%+Hzz87u6q8GlEGn{wRCBt8XUHwyG1sWY z6?6l?Tc~x?@zH_VFRKWA=#a_P&RMy*Oi}`c<@&P*>5chooO8H0$;Cy7{nGqmIX9eb zxKK?1EMvSCt6}4bfxh?K&XAW4U~v$s%1HKj99laLkQ_0BLFiN%ktHr5ZS@G?9?Q!@z;m^*U3MVZw$Rp_rRXMLAKdXy_<9pV6y;i0$gMFJE5@E zN9~K@kTl;$>7dKCV7x0&yh4x*^6l2p#?OOakG62$?L`~smetGMw8{Rzo^E>c6ia0p-PL~4^^690k z{!Jh5*(zy}wuGm;T^0ZBy8QjSvVCqlB*{R#c;qoz2vnn`%9bai=s)`T>A>e*$qm{H zG3cPK^J_YTfa8Nl&E00&aoNlLA|iq{l>gb)pE{B)!(|vUgbbg ztbZ}Re*Qg=j(PXyed4Bd+9*wOZglLJ*ZJ5#uqGJB8H+V;#~H&GmssJJ`kA2Q+ut$x z$p`+^_eZ4uISaglVaJ2K14S5@0Aujr)6iC=vdowF0=%;kLrBNztJw z=C%XpBT;^Ou|{EuNo9?=aV#Y3OBO)XJmM7dWDB2nSGZ4hE$nDOooMuDQ+S2&9=Q?z z>sw!4`OTuYT7Mk#m0vdlv3mt67M=>o_ss#wV3Bm4s3p9PPW4F>4LV;H$9b=RrBA-& zrD`urX4-<6ZgPtiuwgQybxzG=W1L71<(mI#sf)oH{p>PXLN>lIyY@mOJC#8(8VXu6 zMryyd_3ioy{lBreKF_z>XD8cb#jVPd(8(WvDJ$*8wv# z(GxEC9Lz`Bdkyk z3T$If%+Hv`xG+D2O+LB6{bqkPo`{@YU#p=;)}0_x#%VX~1bw zCvvrlKG_-feQqF<@7g&boHf&ELE# z-pIQEmD#IjteMdgeuG;OTFq;8jS^_Q8zixT7le&;9ld0~?5KOYu$EKowWK8+pRNep z;Qpm#^}-x(vLst|pT@g2igO{gyv;Kwi8j_C8U@oce?AEeHdk1Ii2A`_=l*!U0Yqu? zpX8GgcJ7HC=d?dDf?O@dK=Z>&DsnZiP!`WquLXun8l3|=3GORj=@lmiV==%8S*pqX zs8j09d}x4}FG%nPC56@g?F!ubs4)cCFHC?>4$^A8>4pbxcB(n|L;AQe%&-G)jyBKi z2HswS>aXN~V!G)zvya6W zgReuQXqDvB=REelq)ra6XRyaxaVXA0vZR2k)+-d}iw^q13{)t0ForA8@(XUxoc-xM z!*mU~?ha9Jav2hzQdGfsmGeg!2k1r&xuVE1>~?8wpc-3jB+B&?80vn|1iR{8rb6MC zk=`39$$T`c4 z(c^{r>^F){pjK)OPjf59BvND(DkLK^bkG$jlUu$3IN_s>#ndP^)+$Nx9*{iK9$9m$ zEQE`}Mp^SWKYc9oa5;IQ{Zfbl8bZg2OtP1qQ)9>bX{bXP1DYy|0aet4SaEmVt!rLq z&_HO~!i$Txsdg-0ww#LNZ1a!fj-<4qFXlE?*}^1`3{|Ws$s--s#7-brbV;F#R^Az+ z0XAOL;@^S36unp35`ND0EO^w&qUfdfLP^sK>bm^HyRu}CPFBqut|Bl2-}MAJC8LmY!9ruZasFIq;E7z6c#<{6M4F z9Wv}-=y*8!IX8X%R>FYku#-MFZ(zm=^iGfd`S<7f7>rK;kECnJj`7wbJ2ovMr{P?1=8wieh89rJdD#Rhi!-I3&R?}gRyo5IV!+ZECL)z-J0!%xaCaSCO{ptexO zFLK+txGChYut8b}4YV%O$7p=Aoj>sQc}X+qt+oO1(&(#JupLiKuyug+#ITdTU>-s!}zC8%RkyR{LlayOdiQ^+{O;j8MKHHZX@^XA%OR$5{=ROBr(1IpQPC(}^9j#tD z7v%!6C(da0(>jPWUfU*lHiG8fKZKPpGvMyxzptK0?y$q19h$eM(55yq?5) z<6RGU$)IF_qhV9>c0N8)CBX=1t-pSJ{ii6TrSGX-Pqr>dX3nTu!mo%=i#r1nfuYzT z*96y-Hj+EH5&W@-4)@^S2*+y9=C1lh76Tz@mMNGE|3P(h?R_1ZDvs;^f`(&!xF@<9XXus92-p zw}fAFtDy1Dc#`VYM)3Z_{C>E%OtICL^}Umu{z)%?W(P38E{@}1QIa=Wd%tb4SvTiZ ze?d+>XAV}Uk%M)CV$M2p5yd3=;8*9NW-ToK=Nzanbj_Y0G_(f<2H z+Q32jlK&>Cs=~X0ivgPU)x2o`bSB;ZiU{oxrr3}LUja+%6;Y$8Sc#1D8}t9Z>_e0l zc<9h0{F`21;3{Yhqwv7{%RYT`R6_uEezWg+;MMYoWk(_&Uq&u0N)FupzD3^u0Qa?VTd5`^N_k^sIf(PHB(5sBE5PagT z-ZA&u+#B3x8hw^eyxYCuj@3N$klXpSb8~1&z>Ij+>S~ZCkLSSy$TPf7-*dtL=R+HX z^qJTlY^TfF5dkSo5mnBgaFl$4xMN|0CWgzQnSb|s-kBYZAT6(htU79ajrQ%T0?lcl zigw76XL?z`@MAZ(hRbR zhur{LC3)hJ*6>k+Zs-Sd2JF1z<{x1C7-=X8$iix^{U;a=cu zo7+!HpfhM1{XmQwS?4@H1HA$KBo2xz<3z{9zLf0}FS!7%+fb^yPqtr(6`w~*iSLq< znS(R?#lZO~ktGJ>Kv(7W$P)dtODu9($HC0_mslbE>Fdb&zhHt_%_IEF{qgSqyV?)m zmG#Lt%q;f$2oz0v=}#e6^5J#)kQgsr54s>k4GZgn;u&+s1~bnX%BmmNl>D^zd2cU^ z;g%DQ-ST8pCs?D%7w9HMv6}EJwrrXwJp6JCeL+{@XUu8K4@Q^Ap6`JrP-=4V_7}`Le_DC0~bAme*?a()L3vmBzXi+Q!mbt}d!nk?ngX^<m88OX4q9g`$4rI=!h?1ktXH=Va<7B+uO_S^84mWUyKmD{CPnq2$1H^P^kAVd5l z?g38>@F&m0f9v?Mfn~CzT)cjZx84=x96S$r7Atd|Vgq~T?RQJ&A0vbE942qkQB@ze zI84X4uKzJPC>TTJne)ry239tPYW(n2u+#vYlmGPXD`b-$181)haCT4(=-q6mA|Wy{ zdzDoEc@UOUV!Qgwka;PQ^}yP*n%AU2!iCmTy-s=+I9sPZh8^%9J%z%CmO?*c69oy2 z`JT|NtUxk-_>;fwGZh?vL8_w6=%zpL?UXfn>i^voUj40RsJ4vdWQu!90}v;gJdb$l z_d9iTKb<~jwWt?ouqvG!>UWdGg?2z=37ASi_ zO_Y{!RP(&b%XZsI$BQ<&V}TSBWU#8rR?E-?f&ADvnJ7}9@f<54Q}O>_P|@>FO*jK- z+xM@_R{*u8CA^00q^l(xdAVNkg3^H1GnyTH9I?$woHrKK)sUu;&y-!Hj^6y0E~4Hj zIv{=!l&Y+wH&SVSsY+X6g~0^scmBb%Xq2 zqqN57vJkUMm}l(cw!lLg)g53nTwK(=0G^mL=rZiE!Ks_hb1UGngzBWN)gvfA>wp~= zx8J!}_`Fy5>TCFb`b1-X3bi2ixuwx3h#Dy6s!!{OjtOgo;)OxybRdpy3{!(BA3U5o2U+YI zMUTfB>C&Z>aoTJP3wBT#dp9e)G%YmdU)xOg*-v(Zvt#RHX7Al7jiNpQ!qW!s9>{WH zj|`{|jbRS^r$I$;1sA1>a0jl=R;9Y3>@0RCEf!l}aeC~`7>*|joNLv5QY0vns zR%lqf)}>tgC5f@)V9YL~trkgi&>cf7Fdiv@mtMfVVs1(v}vyf44HOKf3A#(nNI4hlC%-dB=T3( zN6Enua5SoU6?8#F92Z5*J87LvUnaLrc}b`PRH-kE2WKLu0!h~SPSI4SQwl1h9eq=s zFT#}!$_mJ;V5r%wpqm%eGb0E0IW0u3Z&ukMub_8LmTflyP(8~tF?>F8K;e_-LACFH zeD0YB^6Zx7Z6-8VE7<`xx!a(X&fGZH3mnW1#d4~>pfbR@Pkzix>pARjiGUP@PaS84 zV9;5O`T;0y6GL)>haIlF?Gh|mC*SXOPRPdjAxWuP$sysbWA2MWP~$}^^JESpS#9rJF@ zIJn2fI(Z!#*+gf<9x2%`i%xZG4}uC^NH3+)A1kWDK)e_?;i=Hht2p%FNIR9x|GXAq zbC6jEsVA2lTR~}b++SY(%l9TDv}ng5F(b5ytZGbKXVEMDU?=@rKAfq$Le5DBNTamW zy+jE+XY@-p@DjW)`475u2MtQjxZ9MeV3>>vVq|nGZT9o$|D2L-%AYh-_M8Z@)BUi% zR|FPC3_8bg)$Q~O!9n=`MLX#!&#@v5KwzC}qEoI!3j_iX9>O35a!2-YS{ZZ!dc*QK z$+J$A5yjjZWjUR^FwW{-VJG;uZT=NyXXNcRb5tX0?Uft$$6Ge~Um!_#yg%M=v_HVDhY*?3}REwpt{JZI|*=8x$S{wIady5G_4Op4gTZ7zz zry{Xm0E!0f4gpbT@I1?4b7I4ZH$QcjK*Idfw)=z?B&IF-k8AP%21pcj?^{7~XOLQ> zH(y3Eu;TBdBB4{b|B!p(f-9n!VEkIfxu*o5c7;5tqMR;Pe(HS1tq0^I@;D{RL2ez5 z*P6mJNv5cV-@KqXJXQpxv+#DlPBwOo#NY?KLGCVbE{)Qd)}2R7CdkYe*7UMfXN%PY z*{=%9-;T{`ecqYbaqe5)4=J;`SmtJ?i@@fp$s)$O&=O|MIw9i~y7 z;qLcEc@5~?Cdg9Z^=^8}712Y7D*7zQAUt$Pa^LEMKR3Qw6L3Y8=C{>-KypR&e~2c$ z-@N;?$ltv4-jY9yX~uz2)bHj9`<{2;6o=mpi54`-s=ZpEj&ReP`sRy&GxS@O6ayJ* zxnqx81*gZO8TwH`pHjE|z}v3jm59!hYOnL8OjXJ2p>I3)LIxp0US4W|K4g<9 zPjYFX2VJc1TU*T(jfaX!T%$Mt=})GyeKSGv%wUv0XmN+q91O%RS;0Ky#d$BeCN5pj zC@tVD_u2+3fEzESoO zsb4{G)o5Y5>BPylzOisZCV&tNN1ohmtPnEoLeJ22iNR$N-WL9j#M^OPZjTW{(kUjD zBFUfx&Q0;&=HI~IFB<@I^`LVFo#)f)ds(PmbX}M%sRvt9BFp7u`JkX+4!1)dyD-_= z`W9oHS(^YvtZy;K`@7iCOQ`o!{(Qx>-+Qb@~9RH9F+G+|zvGW@{98LmR!a1?dWKxo&WgPnj>+28Kpm zDcmE(E;`821~S4SsQGB8LG?KYeo2-}|7$jq$$4HAfsoM(^t6+4pke7HR#s;EkNv)} z-n5$u8w}bpd&Nwj6Kt=wN~&>DV}=NXzCH8g!w#q<2OXTSD|;GH_$P-9OJ~Y(A3Gqi z9DEYDUs)NeGj&neK*N4rc4Ee%K~eY{+zR?Su=#KEb3xL**A1*miBpMF6}Q)I16SQE z#SRu4@0j3rAX>%($7U*)D|QGm*p*4lyrqwR&~t+6!BX{%@XicI%5I^$*7t1}`?e{++`l{j%NACgR?w}1C|245Rfm~e zus0}_j1L?g>7k>)H)^5v*Q5!VmGQh5OYMb)_eobeTrlM`*{xe^hP%`yt%D+;Q;HsG zL1+aq5;n^3%I=2Z@l@Cj;}>n=JR3hW1!t@*3H54luI_DvCHY-*@^{IR@mOwl?2on? z*^5sorhy`-kU@cTd+hjv{Q5lC0ZBCn%a4)at98;WEDFe_5371fwjjW60-a0c%*2`a1Zv2X&0Y^<;0_4K%vP0yEEF_agHbYri)8~6td$>D zYCVxJvqEsvUjxn$YGRkVW6xRaNXs6n79wpt8^Y!qYiycF+HZz0mYPZ)+pVoJqs4q1 z+RK81;HsJ1eEP{1@k3cB{rSvN*$!sQ%MDIg=FWa1S-;^i53m5in5S%iXdVRP@%GxM}D>Sy$DMt<)yGLV(;VFiil-+pJli>aXTlT-scUV%V% zbWEyv3&kW*WFr-+Lt&;75qF5SX`MiY6vAYBirnQp1g*QEnh1%s4`t|JB+73EG|19} z*I}i44pfy6JM_u!24Ozu+^iC@8kqwpNUwO<0TVsgFHGaT%^f1NKyBjV-FUf;taU}ee(1Xf5HMa(mPJ`Dkuer6si<1IBy-rRVRGN9*y%Y)P8p#ung zj{T~}Qp`sbiKZe;WF4Hsh5cbMvvl;Dz<-P}hnMi|_ipK%vhIEC-n-pae$^F9ws;r- zp<5Jmm87r>n%MCYRBnXpT@AoNk zH(39Yb`iFWR--R-*-3rIe^8EoZaaU&!fyFlx{Ry7jms1Cv*)*Tz zDGALQ&7!5SFnVYTPX2cD7nJD+<^@ULn5hCvu-jmVneV>72)<3e`sI@nY;M{)Qt*YXmj40XkVO-1BeVZB+_3Y5 z7JTii`(_vxkZ|1})nvIH?_NLzf6N|sBgMp0<)vXs+^esg^Jr1TiJ-DivRtswEm6MPsh>VhFAv9B927>1CVS{KWejuH zw?mfB=m`S3bZlTOw=W3qZRd|2XM=CL6&in8gra9}K}_$MG{RXK7OK0ial5Gx!p1GN zZg1k7OlxFt3Mr<5B6(CKHUTV%xGw)|7pSUr)7h#N=n|*KrU|H`1>Dpo zMLh4o;;4oBqC25?oS~0Gg#KJz!96F*0hUM(y+@cY!o08c&8UUybcl}}Uu^YVjpev7 zwI9ak*1*EpG0Ynm>9yuS3!|nO6w)Y){Nf3`Zk?n|YW3GQ_D?o@Y{5G=`>ZWLF#nFn ziB{XM{e9!jy>FWe-axOlF}sCqiUAv)Mn$gX<+~zQ|c)={1s z6CZt{?+YMszhn-H!x@lV?$O4=gf9b}a*$^nKaALovx_7?vx@iw| zHXexm{+GdmGZyYC6tk5AFoD-w2c=n6vIc1i*ukD(u9vARlsgy{4n`Sa6dT;&4)D$w zOkE6xHOOrU?6E_?UbcAMPYjMo$G`sbKa*BF-sOC0G~c%_A*68fdvgV$4_6I8fOQS>tD{k(~VTSjZS1voey@k8A$)2ay478T_O+63Z@kW52600snW*(yO7E6pC!6 zBC*Ato@?JjK&9YFN3fJ&uNbgu6mbpPsj*m32BLYyixenJO+$j{Nzq10;kFS zM$4Z5mzuS$zc&En4~;*YOFnyH0J47iSbz*s3|O;%Dsn5eEQPu)zRyKzzYY%eZ7P&y za7+1Bv^t%mQFMX&t&Yaao%D`y$n&(rCMk_h3eE|J6I2q-R)I=|^Hnhpf?SkR_cKI2 z0Q~v^Cyh%Z4ADS(XxM*m_3H(}VvQm-EGPJ~(_Sa_?Xb1b%Q61{HQpV=;q@jV{o+&;x0voVz_qTQ2y1_OrN`oRw+!Z_ zc-GK6q|T0a$~q(1oTr#3iZoJ@sG!s#M@>+TLhG3myxC8Uf!Qlz4`r>uxQ`0R2Nkb+ z(hO9~TfFSRD*AE9IsNU?l%Ae>FeG1eFa&j$HlUzE8zdUKI0c+6vca`YekZgdaCH1n zRR?Edp-^V<23P$MSD?T004ZN^TV7AFrLvCJfi{#bq@HM0F$}x>gw>GsblmAbWr03c z$ePw&@XZiY35_Sq4()jRV@4=r@TaRH+wj;bUN3zf7#1r)DETNDyEvz(r3_fiknye7!IANp z);KedBcArv+L;Dx)b*>&e85Ls$a!2D2CU04v>6NA9r0Ccrg|H7^XOMNjJT86YsLPg{wZKygedX znc#hd+o8Doi|PoyJe(S%IO)NkdZf(4k)bDZqx29Y+V48ub*dyej>8Ts1bbgbx94L; z@~l+QW5N1pJbx^M**VfT9+FfwO2{Q)=;*X+NZcy<$Yt2!TnLIr4Lj)7EMpkZadBwk zB+6lJ1cG4&=t|w<4a(IObSwwcVF&$i=q#1pCYg=|f9<}h%DUa|xXj4EpHLM#VwArk zFjaKNGc%}9KH~M_!I!dPP~q6-)2Z4Nf*{ZYR+4T=0!BNr^Jd3b9X5yT7N}8Am(6dO z*uCs5=bqCJq8}OzlwU?0$dKFd_Nw2=dFi4U*l)F?^D^j?9;{IS6^l-H)b~`-C~ADL zZYfU;$?XEjP3!13d6WP@yX{y_>L$BLMQpi$b z&hipLaIS}-Ks-<~+UeR*h%-Aj$ZkV#SyXwCri%PNhQB&%PO|$ppWz_0ob`!8Z4`1= z_-O`-?4Ty6Ki0V$p!TCP)0#;#yS31ccVeLQF(xpWPch(TXHk)Da_n%G60T8PlOLWJ z8+huCSVW?%U%qw*v>Bdr^0jVREge0hSGwO3?~4`TeHul|tc|>toCm&{BsTDJ;K?u> zy4{n5mKj`F;d0a5wl62+IN5P7+l=FMK{5y`|08Y*dQe3XX`+X+JYf8_G8%7vFPQ}I zv%+DAld72DPo0|=j5PYK^)CQTx+ox+c1s84RlMA}mqz*m_DWGqSL;+yw)>&^HE%CRy*(J! z>?#5aWm*UvWz4x2JnT>`yu`^;VH69bA);W{s9qV64u9~_>nZYHAN^mHH2+g)EUVCZ z4m&Jgs0)E#*-6(ZEB*HhM#ANG>p1;*%RA~K!Vm{gQfjQ?Kn-nPJ(?)XVV(nU`|5m1SRv#Dk)i_%^kUz#1VO21`|Kuqq<3eo3bhep^@=_GmvSo+Kkw-tWqrM2@nhW3SKy6odzhXwULFj^JF#fTyHdLe7#M>jpnw$i0v@aRSOq-4zwdKfFQ%z30jUAcM-#nz;2c<9?1^!ecT^L20i$g&M+AEj7 z9oh{H^r0vJg{y$TOvEAAVGo0c6u+srpo!lFy1Nkcy=)Dc>QZ>rFPpOT-39-~pn3+7A zU1o=hq`c!fZU#Hi^ls@=vTX(_Gx|E_Q4I8w%A_KT+yH9(JfsMsoj zs$$HqJW}N94o)|HIZz)_ zos#sW@w8Yv&&E&L{*2XrI9djzo=uBXlV~b+QxyEGxoD+4UG*+~v z{%7Ui?2L$Cu#Y+Vllu{CzuG6;t4t2BpR%4LmLlfSifDVVjQj@kh-s^Qef1K|`OXq9a9H@I=E@%0F3wmQ%Q z*gU?*Bc^Tm;r1W=?~1YAe(9beW1d)F_uMQJcdbJPC5k>uF1K{{$=kV+{AO6>L`_of{bf)nMN7ccuD4x}X@Nzf>* z2ObtCM9)Mldw=BO3O}9aSuadJw{TLG8b#&&?ZN6y($B%W4?^dG_47e)0{MKobh2bX z0(PidegW1xR1e67v09j=@W09L@N8sNCJh;OR>l3lBsnh`faCjH?Gm!}Ib)%h8i8pi z#Xtmi2Nii)*v^L>;Ye&=od8m@m2h_1ld_(tK_m6T9a4Mmz_#r z*~2A3nV?){t^a9{B;FM6v)BS>(ulBd z0L{E#2AQUL*yI51H>3pJZ^i~*9yMj#Ngr|F!rbwBz3-)R)N&N)8nl;F0&IPYp7b*)#KME$8-pX`P> zCVZV#jnwR%;9QW2t|U`{0gI;miJh(asj}85<~@t?w!4o1nYd13@HT|Eg})>5GYE8V z9}@sirx?h|CR369uj)LpFZ)QO0xN8wJtpiA&XavylzfA5LD+zEoI8@fnw8DU4tX0y zNT*;TEY7|eO@_5z_fy7UH52x0eQS6AIK(tYYqvJUOgy$#f~!P3KOPGBb3h*vNYh^k zwNCrJc1VgtYm~8^JkB;K^PPHgE0lLMFC5+3uKZ{hIWU8q zGcq8@DW;YpHB@93=~QihtBW*+H^|P2*34@TPn74-XGFJ|l>zJgiXjNqsai7+<#4ru zy}W40wSHR9^PFDfY4acOia`@BL6zfu+Hn)Gw{fiY#a(30ymMHv;a%i*pAh_Xh4xzY@_B4akls`EKAHBl;V)PWVuUYIxo; zk}0m?m56Kyfk{k18oOss<3w*@XYNjY@3Z-)?u!{O0c(KrREb%S)Vz|JX)w(EO`lISJ+mvhpXaCc%Ca_zg?HNMsCaj`RKgi)4!(XLL%8R*_ta z$)ZRG75Sy?syLa`Lm!iU>e(P0^x8?|9vOFJc|NVa(7KI=&p|y1&o=v>6XHEB;n&2V zx6SR~w76dmw4Dt>)Ukx{SVzgqpfw(j6@)3z_*Jr34M-`U`Q4Fj? zB~&CzwyzK%QJ{)<2eNihBGk@5ISa$=i8Iw{;9+A7K6maQw*mI{s2$%Y?}JtZ0<5wa zbUCd&tHjIa`0M9qx>YaA39j*v2_7Np>w9$daq~r`?y*?p1wtFvLDeMCGZsl>kBwn4 z!82CqnRemaduvQN_|Fy5$qX+a_tBu(kW23pbkY|f9=lg*ooQz|E0&;y<&ajn-8@Qu zm6rKiCxbnCUw!V+B-)Pc39!D$gao%z43t4^f*n2Gz|ZHd;%Kl@<-v z^_HF>pw6VUiWZ=uwMx=N(4lMQ(tS?*oC+w-!lHmyXcz!3BWpQ_RHz2k4yF5`Yk<8d ztUx&#cTU+LYZPI|KgS(*(iiI`t5!VkdtvNC(k(@Wu3N6FUtSfUQ4Bj^)(5Y3$hDq5 z^M)N(yYFD`aS;l`4m&~LCx!t>1$&PibJUl;ub*!%k7K+!ndJ13H>6dru`(>v|MeTL zUpH+zY`1|EXizz3%}Jn`jTDKcA~8VH5{@NC$LUUfQ%Db8&1vTsxz*D*Uz^gk2ItRG zSXgDQSq#Do7L>>0^o`HE_62s}tCBpF^*E&{_K7ZWw!92vv^jT!vgwBomxZf5vVq`? zS{1F525A+qB^>F`L!ys-cM~1`MJTG=cPTTEuyW8Dv zTj{p8+wN>Poo@SUi;8zaRPcftK!FG%ir@tm1TWxaRB%wEc;gb$ac~q6kx}?RPZCBF zIhq3ro3a0ykK|n6z<$P||(>QD95r^XwYQg}wY7j6=A5 z-X<%DR7susxcU@ZuiVMpFcF^|5%dF{Ji_Ed0Mji+i|DwzMqUHm2?o!k2wUFXwMH$4rf`jk)$*3N;cV!lkZSso;`X$D<@rR%=s|mxo3)wq z+&vyDjQdh<nycbS|Fg+OFLuniZgvhAVFq%cHp%zXJrXU@ zr}|tS8;ad$-*tQB{_I`OKfVp*K8}5`FLMk*g)Z+(l{2jR(g&u2h=PziBZzV?X!^+PAo{!P#VGaGs{vlN70= zl3IOEtM9p2L&q!N`NuNh22q*677S0x6W{zmAlV0&M=TW1R24GyqF&FtAq@A9bTod4Lt&?M?wr^FP@&eJbB?s{;Gr*p2pAl<5wTRrMa z6bl*qA5lq|>zET=a|{cfuxkIDMokPo6Pu1ncW%E%t*dx&f#K!Sv3 zBsBeDnj#;|Dm@nXUXzu@70UL?)0h;+fZICxohalx(W%qGg}ERspDutB-vYWwS|kR$ z0L#ZQ8woFXB(I7Ds(kQ%wVB9Du+K9_b=$l*&Va}+)*j(#=1v~=ss~=L_6ui!J1(&E zHRl%Ba8{wsBJGt7dJ6amj7s(-D21l~a~ZKSZ|C70}8|i&>OMfgwX5o)M^B z=~oGj9nFiVTMYK8(1$lbr`Q_*^S*%`XGb-IiMT?lI+lm_Flq`K5m8i=Kw+WyB3UvEqL_@r1lGju{JA=U7*;HC{j-) z?Weor4yw!jUG?)hAR11LV+X9^zMcCRH?U+k5*a@Wu#7XKu9FQeycVyp0?{^#1w#FN zNYE73(bZ6#-LBj$KjgCx8`os2kdO|!w6z(UDoKWBQ@|WdhA=B6ya6ec&&?& z0jwi1s_IBF8hVd`0{> zEy`rykAit+%aIPZPx&C~}@k>QY??>t{DPA#PVTM`o%Xg=~o4&gyFo{UGbgXYQc^Nj2w-8kKl*eP7&S1#K=eU0A}p;6QniRUiQ>^5DqCb(YFIK9$C z@7E3Af)(LUcyw_d*-K)t_4CYicE3C&uT<`xY(dw1+kXBlvXLLUT-YnL-wIvD6brSR z1=#A!?4ihzR?Ij>7A8Uyq!pgkAzd=<6|k6fU{Ke~Z+d{)T_iqwRE}1kNstAI!>1@a*zv^q8a^LeDea_S+ah2C&*??O%{h0t$tCAy(Q$8^L z=XeM>LjIr>_zBmJt9i|-<1>CblMS{gm*4F;x`J%sr(9e(AEL%exs*}tPKs>D?iK^V zX5x_s?^YiqBPovS@YVYi2o}axdR!B4p0Xq%S=% zl_E>2By@xn&*~H4Clb2-Uy-=k<=nU9^>X+n{N7iWwSf(rZ>88Sk@$FvUD!(DSm~j) zp{H9^Sj@C{#tF?1_Z5NVu%2G*v6wnRO2G=jv7m_#Bd+wDtlY;0iPEFn7ixsOSjMI=ZZd~ookh~Df>{9pEstS_l!oIn~ zR?}WN#g^+V+xtIuVbgITK|iHhiWKq zGhc~47|zJjBC#eYQ&koC5XjFO6`3mY*&-lOUOF+Gu5&yo(5Qeg1Qm|P1h1))(trEg z;kIV=mkXU;_vyaoti98i8uzejxD05_<0k1Xgz}g%6?^s%JFNE=ZeC9 zQ9pMWho9u9Ies7cFB@%na$Pq4n}fg?!@=3oeA!hA#FH_7CntIdr1>Dd{%17kgT0Cc zVaK4?M;gR}=7~R*EK}vv9f~nf2lhi`$V}n{zlbfyctK=*rdGIjlEo%zdhO!-B!i!y z#D$kPyR0mjJc@+^106bHjbMe~a$Va@7Do3H@WPfVk3z<`*{Q0hYu&M-1MXom19Ca^ zlG9#Jbh$ni;Rr$H41e3xUkbL~|J|}5**Zq}*dQ)kipJsJ>3MCB@9t3W#WymT1cR&{ zW1HYR`26jH#Nfqa_}MtI_TOz+zAihtaI9rpL|ehvd0kdb8xmVY*WwSxHU#CEgR%_s3we=`n9-+9q~M-)e#xvg1TF z@S0_)BO1?ETfQGY>cfS-Bpgh+O8{4f#6odj!X}?nWVdV^-ESfZucl`=`ZoF&01wRU z#>A^XdN8ZzYYV?t_Jh_p@!$DG=rygXSJcskkg8d)ct|ce&+;7=w46BOG}lir4;*yD z5BzWs$#G#fNtKn6v6EuAQ=|xz(*2gif2i(=FD6g742c)~K2fdny&Tb?HlM?`g@(jB z={nz2lHIa$We;l(WEutQ}Pw@Ely@C|Kp12-J7=fzk%QM?l1_tPE z3uSqlJKo9SF$~)9)h?p}2d{hoJnLk>zXcml*i%c0Zn*So7q&VMS|Ol>VxjV5E0uK3 zZvk5uk`v`N{!9&8CSFg@fUb4f)F`si$_h!N*ku%y@g=Q}Y>C36 zK`a7Gjo*+k7V;cM6fYL>FgK0CA$~f;|6b(|+u$xAo#Da(8V)YWH0f2z{`f7jHs+?_ zJdKt1*y`gjsSZm~G&rn7huwmHF^mSSQ@e+B$c4gh&AU3)qAA>@n!P07h23OFtTe@L ziiP0Z4rFoajx7}|^wg<8V@f1#sy*?S*|lJH6i-c)t`A!QYmzHrYp169W(2GhR(fo7 zuMsX6_R;f%Tjgoe%E)ec-pn8#nwM-qIbP>OpMtZ@erf$Iq^k*<|Gen8k+@ zVC7OQyl@++q?_S+8dG~444Hr8IZu3e3Q#9yXgWfRq{*>+eD8Z63eag9pcW3h7&OVD z#6RMuYAk%y_PlVz%-hd@G7dP)OJlpP!j1T5=_k`H7y&A=Hrv zDHdF(dj_j|F;%Pr>cB9K4B0~WhIZ2J^x@Fth$0ytL-vXe`fB9T>5pGWO_Rnz`dJ~9 zA!w(~WxUmrjPNZO=EeB@ zkOSUXdt0WR@78R(NLIVBHv~$lhD;<|C^m);dTKj|ANK+&siboD8&L9KoymQG>)0k z=Yc&Br-+AScQ8;f<5|!4k~AF!cTNmc}=0slxgjboi>ACjJFok4oKHL3^d_SQ4TDajGLgJ)ZfI4bOo0$QKj z9$XtEK-A~Aa6>EB_u>}Ybr7Gmu?rh19Lu@+;ib|xAcHs|u6G{**J}lJnXZKiYa4wz zEID>x073}aMDsOe%+l$4I$2yK{wQWtu=rdK&)*mkdY(U|L(cyGXOF%wv#1*7-@epH zR`6Q~x$X)N$VP@vU)dB3alrLd60USI0*x`LEUsDZxhMOrd}Y8bBXE_lM6ya)LFZvY z8ur00lBQ1T7(2rBks!n&5O~2L#c%IfWoya4BwB={3+im6@}}ghZ?g&)$4kC(T~-2l zX~mG2wlrb4bJCliBLcf&@tiT7acaN+H;>!WQ@d>P5C`FO7IWmS>$3UK3G_g4p6Zxj z7qdRB9=e5A%k>PlUoQ-53SKzjiE5tLar%?cv6v=MQg%yx(3)_Dc!wF2+t3&$iHLfX zOa13-mWibz{Ev&sHhvR}3mYFNtR|KN6br>QyQ!p%z&7THZ@=5t@VVjrZVl{K_*pa! zJjavQLr>h@tU+1n@i@VtJooBpf=7-CO6kX{`>(Bsl*A`fQs5IRaUjpNlhiBzr2W$$ z?$h&puhAb(sq`?9HSkJacs8(^R6{@9D~w~%$Qx^enB)E&208o>O^O`&-76OK1pjB{ zY_jg9LC+p5^lYWre2Q$Ql4^uGY$a>{*_ilWhQ#xsw+MV}A&VwtOHaJ9cmf=PuE&|s z-Q^+KE~||EG^R2jS94UT_b?>hCYQr1!}dEfv^+Yf@P~lCcQk%o_zl~fXP2FbII@_t zpq2GQ(x@(k1QDQ~3GIY>tYX1g-!$J(LXi$1ddjh%L0{tni23yibgXuV;3~LNjL>fm zj&G&U`d*V~st$(l142f}D28XB{^;Zn`r_Fk9dd-%v@|YeGx=xMqF6jX>$yzUzBGzu zx0PbiQ!FH|<67e%RZk@m= zS}9G5=%CLlKBXHK#e%g{4}0BrVrD!7U=+A=2!6N*iIUIFw9Ss^NJ7V?9-aDxaGn>i zHFPSngE~eTe8Wehe;Cw!<&T)rAy&U&8 z)qPbe3v|DM$FV|~q((Q%p&KPMr*tUd?Z-2>Q95|Wk-sT&$T_W9@GWSGDwHR2tQr1 z>EGV`ZKMSmOENomlI<^zk$TDsN0k%{JqY$5cAq7Xby$fE)BCn9*XY!WjrnPS`&= zIl)mWXdDlJG_vOy(jkZWX`26N|6z!2{{fdp3P(qQTKcm$T11P)+hTel@_%2Y58vpn zBiTV0g_k4B)lHxTGBgE{J&+bzs_B~4B}fKL zR))((ifyDw1C?}{ojYm4gjI1@!j^+4b|EoZ3catGtVpO8e+`WgD9!ECDrz_gB!zWl@@$&F=AzqO!&4J9j<-&5F{$L(B9IB8; zvds&-0`bWYY*o%nvV7)HP>W@YWqW3vks%?0_J$8I&xe%7HLAe^+Mz60@0pS8=MWp# zUMcVahP`7t;<{f{A2Hh6x*R6uInP{amnxmTD!Dyv(S-J>t3iFLl|l@x-4WJ$SI^uB zZ%lVw54=%%Gj|G7gs0V=vf}8I%5)*n9ZK+wbDub$M{M)LIpRrjj=^gVaNW55z-yoG z`!~zv)b{Q7-Xl#e>~DEwHQn5y*jp4aQc3INccM;AN(cW61{g3a_HbAY)2zx6U|J)t zn2-(Dkcg!@xYZ(p{wAiYDyJ0C_t-XetJlF0oR)Buc4e2SUx`a+^L7DDKQ+vv*e*!t zhWjf)mrrB=~|XCOA@&)O^4BI|uJ8 zhVYU;+}s!!N5ITq?pJ9PuYt{5F```uy#{NpmaQKRub&J1dpUUhT76E(c56DxB4M56 zaM*y`0Zmu*8O>u^3xu?>a%8D^UVxJq2p5H8rT}zP+DrwL(fglo;6GGe*%LCi#RQ7--4a#M;c>WA22iFYuo9^Ie%uC#$ z8yx)goWt>&KgTP6{HQ&ewcK=eS>s?I##%ll%Vs5f%lXqN~FeK^t`Alw)CKaW`K*{>s<{pY6fOGghhd zAu{gYeM45z7K@|jJAtRjJQub&pv-iL#j%QFS5Ra*m2^^h#nj^jb-;idB$`C8B!Gai zvoD^?$v!bZy?(5=_)P0~xo|ltdTG41$E_f=pJJi2N;#FZMNmz)dR5FQ z5AOKZm08zh?aKM2BgBxn)eC*5$zR$3?$udoOk?6rU2ONgP zRo`luWjpJ*teMJDF1kywIILt+i|D*>q2LTj7xo9hTtd!-o(p{faUtvkyCHBXDT6jB z^`cXdC~H)1k0~3I3a}AL-C?rj9*CEMq3)!Wo`|t17>WC=Eo7$)&pKzUR7N$$R#Id? znx2?iu*Rz%5+N5y>(nUJF3Cu^yqmK zMg$=)wK3#4H#qD&!wW*=vZUw#YFmBkvTBB-p0wAqVg}0mj&OD4!pIT{)*S<#n2A?jh%}%T7W;4YC z-Qz|osa~`{epAH0fa1750T`A*wY4eYT5t>W-D+lmTB|0YGT<-U*?j?tV~;iWmK!ttkmY1>nWPa=T}6JBt1+d=>Q?P<_C z7TmDM!nVn@M)|S0PO^6DJO~%=5adAZ;ax?W3i=l7!3;SR(C1a@f!CjdI%6I5V;YFr zGsEoT*N1PHKY)s6r%YNShu2fUIRj|F-26|k&ar5fGXFD5((b}qWx@ENu9Jrp+f9-C zR8oTit<75hQs$BGQ6U<%7%tYS^EDf=!SD^AF4bRG2(*9eg0S)iv9>QELx4s~rV1Gy z=ZWVE_lH2<$w`e)(<0g#x+uB;Dt!(@s4rW3NKzs}LdbOCc1XX(Q4Q)$RgHgp%q>N& zyEa*zCGS%8yX6v~R{_#mU=G8k#aJLv>I42<%bG8zKpev8}gahT-aFVh%%xEs#jcxjL1f{PL1jSi7DHo8WhFM64eH^-XDM; zvx^yU>xcg_kIpF**E3ig0nncr!x;y@5$*ZoFM2C%Bb}VE*+j*?ck)t3TdypWv;gPi z=VfOLgvkEZsK&lvZK}^`qpUjEWY+*hG`$25n;E~I7DXI^^lc{?X9#~BFEuy*SO0o| zvZZ%=c7g4}#xMuv(>B?G(33%6%%rm=l00Y@2WxG#MhXHt0_KVLzXoRx2H%sP56Y$s zpltPI(3q%#VY=D;r`p?_;3`ZE&QVDR zG}~h^Y0cDKp;}Z&BV8amWj~{LlWLJRQ`MlrLx;#RuVb2(ab0nk772L?-ErETQxEwd zhdM5!G6FkjG@&i+1<;Oe^;yi6f#sT~K{8wu>ayeqVCKRiw?ko>s!ET&f)(KpL-LV8 zw~(11(Me;OK;@LqcqEm8W)Fq3$|+5gZY%IQJ6%E}ftIa4;=6ByqG5Nrp|v~UFaII7 zKIjLb`TU=z-A6nPVe%zA^{J-~;^yS@nR@HfM1sf8PAa0}ID#<~0bMX8(wWV)% z_+gK_EwTf_`@l<3Mn4g*j9el->w7$Yv3F}!Ha$N9FHH@5=)^S45kPC`wDT+`9CYM- z<-{)AJceftT^BZFIC2||uqpy?8^)2~16N6yHPxs_;^QVcb}`bn(?)r%ySZ=j)1OY) zG$TcVg9Oew47ZZ0ew=>Zmdo^6jJWV(h=Z}TLR})sRH1E&y{&eJH^~oA(WWq0J#wO# zKuSNxthEh_BZ4+{RpcqMBp3-64u@*Zc?;M{C|^@gW71_AT^R}HX{IVGBpc|nuW9Nd z^EHLREnW?RIZblRYuFEV^&Ce)YozEIs!X2Q=EQ{kC%Dm5e`eD<+lZ6P8Y~Qcha50ZuuI)^;g@JIkIGN6{PmBS1(EE6PyR{a?Y1i-|YSF(x^mfjhm`b;eE~V?}LjRkd1Fts(I06a7$Y%857-moh9O4GYaQ2I$uA<_L?x(v# za0Q$VC5N!kUGBYn+6BLUH@*0XZ>OS249i<=Al9ck7z>e}DW;|H$DZeBpu;X%tljE$ zj=tlKq*BQ2;v}f>Uf(!FX$%J~URW9b?auJ--&$X3RcI`y&Hix8FMG@`P7THK=S4mufZ$Z*sPGhwy{9d z5VAYiJVd%TBwA>e1owsJl5?hRGI%aUSVs?j4TKp)=Ye^-(HoDnh#*Q%3N`0K55m_X z-%KI1Xu{yvEuPS*STq5zTRh<*X`t&B)xvs39epJXN2y`-&@&CVzr&y|7w+Ebbv!g# zT&?LL50pJnL5rjgI`!Z$`#JPSuh*VI@2Cz8;?{lpPtt7-BbS}pIEWbTY%WD~R8p_!LQnM754f$JdOWmERXOFJ zd%62MNDqME4kqUoi8BHR-2QZ&u8Pz_0z$TCjbKLt8n8I8IPSsQ{=ugGkxOIP;T(35 zmo!{=_42RxzaM5x{o=AphJ(xng9TXrs@?994IzL>%6tee?Eq#GBl`N0)#Xo42B{UQ zcp=v6PUr6MLhN|O#ozbY0)tP+gX=;be0oFG;2MfuO_6jesnP@6(&z%~QwZ>Y(bJp8V9V{ECjMm%UL&i^`APTO7mKrL^x7b;@`n&W zyy;Oz>*)q}ZML*Sk*CRr(DiD+qeAW2s1zP37&JP^00R)f;?xCGe)?_0k1dAB`ByKG zBbQt_rSqYcD!5It5V>x}l+LxxT=4=Sy5TSZPFFdzJU1#~(wfByxtXSz0-nkzO}H=QCnO)C2VdW_N{ z8x*I>20!?wQjby(EixG47p6%E$1fHbd@xcv7evPI9!NN$vOr|* z%CJjh4L=v33!Ay+R)8v?SZJ`51Kd)+kd@aP)f}~8YQ9X1xl^W$DP;a`0G}{=miPGH z6m0S-lHL=+#DYaE?aGQ7Re~Zh#;7s^_ryCS*fcx>Mh2_T=cM5Wb->XX&8_Hl>eUlX%s`mn?L?_RM?m%f;yJfF?Nn>d#WPVB zK;#DItsx~YJh^;mCE_ZIJwTCtRMKveA!t)Rdi6xaHR(qA>`Yak;BH8ZsLlULuulE! z&D8R@*1ePZvtL((7Z#UCb&$(jM`wqodNu#k8tMb{`8~6cBsqsEi@~|5&$Eg?#dZPh z^06?qP!IYSGij0LVKarZ>!j5V=u=}4BtF*&_HT?DD!is1ifVooJHY~=MC1EfGS`Ix z1O>E1bi_)EO{2&%D(O&IxxZ-+*&hShFvaphnRaz#OH|>Q@OW6s%U(=a-82R}yf88D z*WSN>IK_eqP3n?EWRnYH0(8g_Ol+rESmEfYq+?+%q8yLTpxypky;e?oNUG(iKh_7F zr2FU!>Jx$f)k5aLtiFU=!BIgvxukkX<_WVx?n7RkAu)s5!yg~Zj|Hk^+y8Su)<)=wKbc$-0v?>;bRC;7h+MzrQbr##i z*O+5*X<<$(Was{a?0=%%V>sgszZcqNo764M$O+b1G|kDRzpf^`_-Ps!UMSaFX_~_n zdx#kS5 z1yDa=suD00fnf3tB-=R4)a8=qi!o$T2wVd=>G+Y!U$3%V7;~&Yi=@c)hH3HUKpLik z6p6?BLE>|;>tQ8Z%^q$fVd+6j?zft@v8cn=7bm|F+~S`R|?m#}8I~EmKuR--*a@hdOx( z0XiQFR&ui+7go19eTNq=s2y+b?6T!K<;-6hCeOfi)gs|wMKwo_yHr@Fs8d6%7<&ZW zVh)iS?>5y4+!dn|sH=jg@6(N2Z0!t}%^u+}fQ*ovhAb=MA}AGWHJO&$>M)&VjP(TI z!%iuW3P~fmnU|^W$ zU|@9j?7rjQJKkbeq%4X!Mh0BiYXMn9L(Ga56uX=vOMt`xeAjz?jq;pmZHedZ-~*wh zg8u^v+ovT)C#{j)%}-0ze%^GN`0#71>&^GJwYAxHt$C7WUwFZ`U0^ihFbD8dwh|Gv`AVSl1{+0 zi6MS0O~-?#yrw}->hi!vxZy%M%|Dz3@(9$NDUvVtKmhPEnnGZZ{$stRMa zGUiiQ#lByY?*m@bJWY{QPoGxjfYl4b_R)DWYc%F)O^yE(-&$x1qW378S`sr)oZ@*K zvXq7+oU?O6dUjRg;3Kwx9N&P0X$cn>zU(T<#i2; zn72YPPlsd~DUPnDd*x=n?M~9>U}wr`P-3HJ90`)=)MpH{zNHS&FnsTe?l?g*9%Q;U zG4mW&-3Z+kp+!bgJ$=u;)Ti8kY|R}WxNxZ4Uo;rE_80~i0?7F|$gr##!V$!^(jdHijL+IR|-R zgSw{i4z*<}cu8C?9E1zWe*0h3sgd0RYkRs>=cimx=#{s4LAHHJN0b(s6&lqA^dTU> z&GosXG94(VKlUkQYTcJ8Y)knZEJI$Pveq#U{!Ofh`_0b-OKc;69DRNsDEB3l`7RFI z?$M#DduMR#=Jdz1I{E;l>Fk&_S8Pb^QsMbKkbtx7?M*&-&ls2pBYV2mPr?~{{9aj? zP1^aJe^8EXIhV`MGaLn4we)V;$Iz?LNMi$+Ow}W$F{TZ81F=F28P<`u`Vv?g)xt_4 z9&09dyijGV*XYz|{2n_yBWYv^Mn&Na+&&$X7lOyve)QxQAr{NF?yXf7B#&R_kqi6z z4_n#6j2meGXy^*|J$vt7k+PZdoV2irk%?EM6|`qmTIki;CM~V(?f1Aszgg<%wo_`Ww}v zk60u0t?MDj!rFj>sRvd zJUf^ALor#HeorC0xr3t49KSl6-k?}zTw+67BdHz@NI?KykqJDUEt=^X6i;q%q z;oMygLa@{7YD>sU7r5S6yTz*%OzkH5<(b-6pKbI?@<7%A&f&UQeV$#SQPv6CFE{>R zvVZK+T$%ZPNXy?>JEFPbQlBbtVmBrpefI`@qG?jDyddeMS9|9}AH8$*dBvLW<-%T| zQ@A06g&kQdJr?nh-1VvSxKEeI=Xu`s+2EHw#gVGUMuPmoGF1QA4CcVy8MWU@p)K=J z&vyb(k$L>=UKe)ySi9UR}H}aR3#hs355iNxEO*=F!>Vn2b6|y@Jz3iZ|biNH1On6QQ zc5;hXk(5tqrhRxBsyIfqZ8yMpnJ?p~KVE;<+k&p|)@-^+R`XLiF1*x+u)>f`^DPt$ zI%pG>RPclAvIpSDsS>orJWwJNSgs&hy#AG=f_wDcs8;%x!jM=P)D-+6uGjOtqDNIr zjLOaOgH!g50lVo4F=aP4MtWglJ_tR z&mP;^i3cb~D&G;^xAJ%2nH)E$I`(eMA!RN+F@c+Hh%z`rvDFl*#N<1eD*R17sIa;G zJaHLT4KyeUnL4_Rl!03}8xCXAW}jzr!ov_GMgb;-JWZ!!e@Kyck%y^Hz+7^G{qSxI zk`o5pvY5(O3g{a#>B9STD)(ns#ZvNKMu8nx$g73^TB0LP<#BT!fiw(Lk z*@B|?w*CB9WFx-`$c1g*{Z^POrr1J?6i`WshP$5qZhK;SNs$yYBdP>L)2RS}TfLBY z1C8Pq(VbVy869a;9twcw!>@c_v zHxNGv|MKx)kH=b&^0SYoZzrWM%`|k@3Rg80TSburW}?dNFT;;6 z)MiM#WGzs)k3FM}G;&LtdUGbAO`JSG&mZ=VD3T4?I zM>SXd@R&}0JEk7a)Qk4V?@*Q~4}0B*y>#zdK*k!sUe9^%<&wv;{ZMlP%-e!e`V=_^ z6|}dY`{qf&$H(Ez!m*d8QJyZm8d*Q8(r@RX1HfZH7@oP3pFVo{`VwD}1xs&UTziBp zbzv-ly)|TUl|`{@DUty*S7eE#B&H|MOns)k9^9sGf!@9*VmVCM#DuIm-`puspEOo0 zYTn>@MvTrmzzZDXYrp^K4`1|hNV7B6R1Dc9zbb+3(;lWvbuerpxO0odMhCF)^N|z} zT<{ro)YtOEh2mdx>uo(U9FC9cU%oE;Q>&^PSVAGCO|Sx#ezF+dp@(JlqID!Kp#Us` zPIjKSm~^QwD)n(i;+4z-b`^O@il8=S3&93t*I#Y%$YFD2m65ARaol>p`Wd;>Hsu;Y zMf_%sAu$K+fn2DCPZw;HpAsJtV;0k)uuj#e-*q0sdck~eeR{Og;w3ECT{Lz6_9y@4 zVVQdVq`mNavY4MK?7|rh1y&Q)I*QGt$SPc7ArqOo3>U0cB&vcgXQ5iuPQc%(PA4hR zW2%w(A$+vBwg3JNrwV4<_SW2H^-MQWY!*e<;_?FfHJJ+xFzjb)>(op#q!lwvi;R2? za@=>un}d7Ka%0*4Y5X3nNj=yO;go$|s>5|%o|SnY-1fAnj*72*?HpM)92t!3{{93! zv_mY}4HUbM0>~sSeI>>3p8QViiRoF1DUhCEP@e$ipLU1~ZW9+^ya0{a&!Xl+@Apy9 zZua7fFO=;aceK}WL;1kLOAH{+*Zx@whE|AQoH>&$Q z!9RjMs<5G*IR;egj=m7&mMRe3hUN7tugBna!5SBxdcf^s%vFyTpy4)`!(q~b7O5XF zuOVMk99JN_{VI|wHo~P2gdAT`SsPHcd!vQY)B6>3X7~MlF{Cyqu)Q-{AB*44rx%22 zmy$X#pmB4(L(_1dK4k7KyJ5jSgj3G^Y155w+BQ&?5#eus`m?0p{Nz`&e=nn1H$u6t zsSS0nL-oezO;xLv_l3vgn;|&eC#a(rO<0n!ctXG1q6t_1%%8=@l%=C*qeXN?Q1@DW9RB;@gAeA4w|Hp5k=o}~71^j>6TDmq2ggY2c8Yq41w2$B z&fDf?5|2wdzGs|mC6CJjgQGTQ(S$O|u8E5$WU9KS+=^L6Hb62!7Ml{ClJKc4l|3Rb zB(_I%PpPKsm3SmYh`y0s6U)E@;(Vy^M~MkA&Vz?#WL_wtPB*0LY%^l{7#l7;sesFU zh!6hr{MiGLC4W;eSA18IE*x-cP-HQA)Rix-jO&UsH|N%=k--N_P$9$CQttzGf|;sX z_xtob(j{6t-MQErA+8WyPQ}(g^se}U?L_mEAcbRE+N0de>gjFJxc!zQV}{nioS&h+ zJQIE(*BJI4;!_?vdfD&-#^}G!?h(G>n;B%QLO9;NYSAtqL*j=(23iMQ?K@>X({cmq z=(WB&b$4ucTn|}0wZ`9&*bKc8dgHP~_Q&+b)k%-Z>lOFu6@vT1PEzA!R(xKMc*y+k z?01i!PIp=E!LQsbZ&X`+ABm@aa**T>mk!{%&OCC+YDy`kScunbqmm5X9iR)&(svd0 zf=d!CEU$vloc8loZ0n?aAji&BrM?13@A&GVD^cnsX!p>iN@L6!G1^061p*BABLAjN zJqM$}P4dH`mBQgV1=;;-K6BwH0B#L;)7~?~kN5yLz&m>Cp14^sQ~c;xe z!bdvArcz`nm4q>SEe7&2e5<|ScgMF~ng0KGtb2+UZt)N2?C*JZ+9cZv$7NSG95dkx z;27wZFPN}cRWHhsHHhYUEu64BxRo*bU6a<)?SAY042kpP$^Mld_k8L_3npBWtnsTR zgID#2ot6H5TA{p&iR2wd3@(I^$z)HxCzIFUF!F(6iO?_-k``;Dk(<>$&?a3 z)2za-Q$Vdw@0FG-pFoAp4#5TQR>cL<=b0nLT!9Di{cd+uz(6{+-z}f+cdIu2D}bK` zbdeyJ&JHmo9wisNjiQUp<(UR%F07%;;yzX9$ok#(MLd?_XG7$G1V4Ai-;v|zVbQKo zNb>f`6tM@OGefP=u9N9zC>v;*dRgMa>?Pyux*ePn5S!?)1Pn8kJ z7B}+r(H#?Q%^D6%ArqMPlBrfz*{l@00&?y9-B$QrX4+y0g8xv%eCpdCv6jW1mGO5? zl=1Ax2J*JM9JbMGmwe1R9Oc_$=7A`V3=bt1fF+-E^v$(IG*Jw4s(lPn5n{b z-|VuMmQ%KwN_;>*RN$%(;qjZ>N}FFpA?(^HONzJ>9=@-*&EXSx|6Hg2?WcF}Wn>j!$5hUxDpoVY}jS53jBcGQ{mwzrXmM4}bOF{caaw?=t(o;mfCR zZ}nunDm%oH$g&eN!&?-#M;!tNxrSrGoiRO+Zi+IZ7A%STysG8K zDgAEA5p7I!*b>!QMNaf~(l$o7@kco~>TTL-S_diMRopLr>SzND_NI{&@V&>HThPOowDWSszeA zeME2fh7P>{)b@)l)bh8^C-%E_CET2z9g@a;B7#0FvMl*U**0dgQ^;iPMc9xT>DkR8 z$Z*sq8}}Fgw=8<(u^{)yr0S)~fxcp;bm5J+uBhQ`P9*AOdo&1oVX5mt`>e@H9bJl=+z6<1{H#b|sv74at|FZB+E(Ff0#A zj<`LoK+`3_I}C~Q89l8JHy^t_4e!9Kb?T$SHmDiFmR5M(!H`1%Sg1TAbJ!t(dKO|< zM@9m&uSot~X`AcDM+v#G!NQRlhXifgLa`raq0GeCX6DbqTCZ%WHaWJ>>jvb?>7kRE zfz@ITEodnBU8Wu!&Q4&mH&xb3>gJu1$lm^YB##wEw!Js@S-r1C6bsGN^QffxW;&et z5!>FqF8f<6wrUgSN?Yjp;tFcJEX$*eF7)2TWQFw5M+N1|qrSy)$LU>6xnM7I(*LO7 ziSX&M5xr;j{7s{M7w!3ukzB*ed~n^_Um|$?@3w_oF1td8LYpC}j%gITj3SFIc?ZYI zdZNY7JT0_S0!1&zh&@Ps@hFbJf)`T9E1HzsY_$iUImUINPB@ex_P1OSxXNqo)Z?Ko zbQ(~2cG79g-H=bk7ojC#nzWNX?z_g%D0(8S^gy33Zt2vYDDU}|LJYZ0)g4zAnWAWL zru)SnXbm&ZSPnTJQiCJ@UUWA+(PH5mqbFyP11`K$zGS66PEqV}iX5Yox>Uu&Oz7QS z9QPT~1vW+&3l~f{BrT`&#Wy_{PDqKU1KaY@#A4ybpfX6>+z{9r)1^8q%?dg2=0#?k zyhM^LZhqsDVx<(@I<<(({Hy79^q7*5! z4pJ=8vG2jo3dz7@vq%iQHJ^m8oTR58d+w)CMr*qSxdGTs!6^DzROw+9WlSG{K7{7P zdn}hnMKbWVan?rap|S(iik@D_RD@>J1yJ^m_u(F_nMCc%vba6Iwf?&z2HYGLH+DRH z1{K8UK6$}sT<4M<|7yG3c3D$~V*!nn2j*1S0XHO;#uQsz17VKsRY{8|6BamYybSDs zTMd&sDLG-zoF+N43+vRW@f)B8$FZ;{UgqmnYjXQkD2mE^cON7$@7=)O{_QzOYa65`kW zbZhj1;97E#ERL=RE4o^_Alw<6V;JGu51PS=Wtc&naEKc;-!*<`*;ETa+@zYlB%fcf zz=eIgN35W;n__oSWCxX0O)n%nl<7hv4b3;OH6^4;cB%Tja-u(!WU9JUH8Yn^$1s73 zDJmPzVnN-KiKXEG-yZ|L%k=chI0LXjeJFwZuR!ZdoVT)}NE&E)rsL=K7zd(_KcWVa zHQcA;p!M1BzWZ0(detu!7a&JfDrWo_#1zJCmp_=6FX#mI0qL?al^$m$I#MsXE^?q~ z?Gc9k;i@kTfvZchQ1bG8QCd33N|lt*P+ucjj8c(iNQP4n0r1&f=Z8iQC~t? z+(&ewV7u&rqC$F!>7cRidp7-%54P2HEa>dV)Nm@q>X_pI1tSTHRf}z#&AF@%%h3hz zPyKGNdZxi|3&P9ugnA(1No7fRsn4mIWzhDaD*u%qR;TXs14h%=|DnmYYl6!H zgQJ~8g8~>+pnE#{&@xr!z<#eqUrCp8L$ry(tqZ!uVTMAn+}!QSu@eNBFnQo(aF*a2 z%Av=3;bWYAv-GTC6!^onS&v^ zJHpkGMbcIu3=bBGF+My8Es+~OW&XhH=KmxZFUW^TV4JE;G6%0pXQ2KVBFAkiM^pgV zjq5@4!>-TKqn#J7$0`5w(MRJg9+H$r5y!}Y8`)&#T3A7`%PF#iO2QmeQ<1MZID6k? z>~)Fh>Dcq&NBO;^7j_Rl5Q8yKv^|@Xf|ZZcy{i1(vKbJ zx+OY@^JZxB>GGgH?+=;gDD;abi}jvm;6}wx4l9J|5b_@+?lLC`=qBIyKqZ1H2+6VN zHOm76FQA8Hc84~qb0+SUADx0}18L+|Ot%F3+)ZqOOn#lZ*Rx{A25}XgEk&iX5Srf> ziIFpKF?Ej4fy^o-n7pMZWQ@RfurpjwulB|ygkF-$n)OkKi4g%ux+Bbw%+o+~Cpcd% zZ%`QNMuB`etTe%Pgp9DEVK_v7l{ z-kUql0z=J~y?-Q6+=lbKxGVu4f!N@XMBP-1T}qKf2u)-=H}}2W{*EIiBL14n9*XOc zj5_*zxH0nUxBfgM#Da&qw^mhy;|y0C0kmq?3>30uRZzXJVVAmoQ3=Q?_)Pnr}}5~f+UD=TK) zR2>#VwiD3lw|F>o@#IL=GJ3E)9f2D*ul?`4YTHERs%w7am4v%t#6c}TI{ z6uD0&WrbwZn|xYCeRST;8dx5Inm8QVpn&uM=ulJOl(j(c0Eo*Jb9bcd_L%)0M?=2*XO2 z-P5pJ<>YU3c6lx^~#Jci_28u zZvw3jJ+y7A!r*+mT7HJq%ryTQ+QYLx_zRjndiodkV#&c)%L|9&);+`6E8|Xo@P{8y zvLLVNwTthQ3>QWoe2rw9?7_=SBe&=ja;CsQh5^b6Lbj+tyJWJ(uPSGu*GM zI_XDWZHAXDYtju>->V&7yO^b92~$nh2UWvMe>n`RHr51Whh|OM#f;4~Fakk1RRDv7 z^Ad#XCV;TA-|h*tD3aoVAMPPJF04qZtQ5&kiUlkeQAzVAexh0gOs3tQX<-$9*|KT^ z45hxe6xcE>Z)R~U@ZwG^_vx1BnW%qie}L3Ky-`nsEBtbx2!5^x@9B4|r5E`2yQL(Q z#^=%1&LMCa0hXVcHwN+N$WMM)p8p5Qd$uftFNrP8!AaQfZKPX79SUgGc|H<)$A~cQ zs67x`Di|9bI7HU=LTJb>W3oB(e_EBco!P%oxsx3I*Nen0qC#7#JP~@Al-m#Lx2vo8ibaKLRB^LL! zsXl}4Ow~YWpSN?1aqN7^!@Pl}Lyml|R%3C$lf$h!UDikb@1-XSCR(s^>z7ymLY55Y z({*7N95BcXSs<^cSjg>LO(iAAX4B^+TLfLoVu6XVr-k0*3!W7*te6|s7_7mlbRMmB z=mcVg5StyoK(@KvX}jD&7JToMfglSudVZ`}NH)JTsr%JdsM$raJ1A02C1E(QT2x1` zmfj}y(A>R9Y)I^+>)*aEGse^_>Yz_nsm}ulUT&l7qz2`ofHcOCcwM$bnJY+tQ(OAx zorp`*4T*UgqdZ;ML08ZZWA96DdglAK#X25keBs2QARP`X-xDA9*;bzNS>d>_MZ!_1 ziv3i};#x!*f%%$zpF6%4GxTC@x_grZTyq(LW1XSF0~ZdL{qqK_y|DgmJB<=kFX{uD z?J;LGk7Z5rWnSR2#sGdcT}9vWxflqIY9S~mS}xcLwg=jShl4UzeS!*btRH}-ffl{* zP4c;kA3Dv?$>Sjp|K2_CNsZzhw=N&XC~@64&Fg;$ejR0*cya^t){qhxcC3A9HI-CR z>;VcWo|4QZrRE$F?F|_QTngwVGfY*aMmirX=q5Q5W-Rb+@-rDIn3eEQiT|{~bpVYp z*97Ot7REkQ8tF@uy5LHqQoBvOnZ?D=evRoJxFpd-dEy*%n9-7YG79wIt)oWv@OU3} zw^Uxp9yf8p!s@SB0QB&kul|%A=4ZLMu)TZ33U22p_AEs{q>{3wrYwa9MHP^SwMGp@ zb;>fM>)i_(Ojpvy2O@w*Fpb-QY9n8RnG3bT zO6b(vLZi-w;Nw;rbDwU|3q7HGB5S^M9;8L+)K7$IaA9h6PBccGw53d~yJ^-#VwYMl zp4!1Z(*;>qd}lzXLGW2<@_Jb&PoMK^r3F;klYjF&lIg-)smux{`4qdEBH73?b}%+u zniD-g{1Ozmo9TE=Dhd54<^CuxOu5q~NKjjnp#=E7q^tditaK=G+i^K<5CZxTt!t}}qGRRlZ;OneKmasmr1DZ6(w2QVD zzVI!g1!4JgJB=E^sTdi}6y{5Olw8X)c?N;J>c88C; zdG>DS9oh8m&KOk+-hP=0YDuV~imY37xAU(W#rp9H_IPYC58vtT${0 zX21At?{w}yw;uT~L#a?@(HU!3hFv0ShI7`s?z)ZuLB@AusY4C zOo0SW69n;!G9YNl2!s-kzybmp-#E%u$^Q7I$~MLrlghUGbjNnaYl{UZre~_sr1|u| zfP2bRV1vRFC&Q4eMSGNB6&>V2NV-gwZ};eWcWu3Uy?Z(2?Lkf`1~YHSis`3+0;XhE zMIMXmo_6z@_0y=pI1D~Vb!Z5Q;!yu`F(N?NIjK~OV`@z+J zp8vJY)MvAF-*_-AQ65*EIC`b6+0C<_|1nT;Byn2uB^Smx| zJ|y^J!J-MAiUS;+d)oRB_xJPNEz?%r>(zfHDZ|+}F03iQ$uh*g$)MPk6iK6!cKM*+ zEK^XaUQBhyw`-~*+oKv3`ys-ym~uW2Yy@gAENpIb`YtytbTu~$e`A5g?HMbklG`th zHM7X7u;dZN_E4mYO2SHea~q+IKt1wOZU`#%$r4pajZ?4`Bh#Z&XsM!~lN`HxqET5d z$??*unttqV*Ae#&dXw+I*}i@b;4 z3gw^|~rwbo=Xti6t&e45dm&jvpjK*Od9LCw^;3Zfv z*PzhTHB3ez8a64+RSyG{x9Dl~JTImyWOpEl*XMQC7w`||FG%N$OjkL=Jl8EP`CZ_h zy4Nh&s|f$&BC?Gi_FULabHYkl9iUifOtPCwT0jlFy@;w7?ec*fOkqY~uV*Uz@bxrl zGE)G-)!m`%fQEa6>h`NTHFo?pBsMZ>(o&x@GCY$P4S7!q*|efRVuETWQ>c57awxwHkSywOzwQ@FO{ueSi zLM(4Hhch!(?X-S!pJ$E~|1F(fB*kz%vgBTjxjI&PnEXZVFsVLwAns{wm=OFp<|jo{ zS&Mq{{H*6PSv#D?Tt}|}&3#6e9I?;quA;)*khnVJL|6~q36((i zRY#STfmI4aB6Ra$jPeB&QdAq6g%hx8ks)!xglnn=;cEguQ7xRXnBbA3;5OODpt_Lu zh^r%8?uwNG zd&l>x%rN%VMfRr7j0A9_x`!89$0fbjk!8z9<+7mQpvf{(WZiQwn%X44JX6~wzu>p? zmH!Kx7yIvI2g{>EkZbjx~2J8iasV`VFa94rmZzHT` zAp~HYlBqg0QD>sP+pW9@R^FHshe7{+As-kxmaFQu*2E}>Citj;s zvK>fvagus+q*o9-N&WxhsFfK3mYsj@Q(Q@Ue%p})TMi3tN4G*s)ID@AkVau0rcQ<0 zo7zVAE-x?}9ReuY9Z}V@y(zFTVq7hRg~!=rLx+^lL64Uje|P8(Vblp076n~ao# zc18vaZ+;wDFc~v@?O4Xb7;ElV)<3BGmx+-2JFx!&#$=Rr ziKo~YimXLOD1Ci@4qdM9CCIZ8%QUMd2QsEZZWVw`Z<&<6|64!nJ%8i05}5Y)1d=`0 zwBoqY8gj&B;P+81)c6-tQT^nSB$K`ZOd93k9fF%d4FVv;6=jHz$y@lj-W~KY?=Hde zp!>XPpx!+o!RTK9oMd&Zc$IUxph%ucuXYC2T@nkz;g5W~1v_cLjyzgeDZ*>h=;c8+ z#*kI;crw8}W6QRvv20&Fs@|br^uCW`WrKVtJtQ|2-B!~pgeCs_{PQF7Bd*RZh)9}w zPuv(>;dN4#38YTVOoM!xUoz9>1+**t6QqBRjUn*Vz>OOdPd&-D2P|V^j5&w8wDH_e z!i)w+JVNctuga@V>m7#{jxG+VC)l$aU}7k?t&JJ|G;SykTJAN&oS>3RjLm_gE5J=S3R-Ct zyOkozR8*UE|BSVa_J&)_ybe$mIp~)r*2|wI`oyrCTvrQE5Ue{o9o!||GrxkL27E>K zraCo@-nj5Idf@G}9m;h7&YbD?B|(S*jUms zZ;RjQ=_Lw10J?g^r3)545 z@SfOdCDSg5hG1R(UeN?Bj!l5`1VUz8W@zBBYZ*9Wt(xZT zc`ML-jPmAO!dDyw+%QQhMmrnB*g~)YEZ}8CB#dR zrEFaQ)@BZp_E*#BY+8$lp{jutfCvQ&)K>-UbH$=e{FAOMo!uX)QziP`n7+buRcH^~ zkQ10g^fV$hj1IW(vT6K}pfJH&O^52VK#O(Xi9QBuv^?<(s3Y-v$&deHPKD^O`6dhG zhM+}{!ZhXKIu*LfLr$nBSQ3F!X8@V?^9P}Wq%s(ULc-Pncm~~oVHbIU`gj>LkpgH% zKrO-bJi6C0k*V{Co&G%Jg!M7VFf@T_Y%Gg6ZewaTiA|Vx%IufpWK5^9KU4Xe^Llex zemQV1!UE%Wi6-AaZC0ANo!kju7L?0u;6Y<}U~W?7G{bN(gYBM^o@B020I%bX!aymJe(#2@)0Sz52MAk3yv zvmMZE3$>3|9ZnJ$r@J@KZ2pj}9c#LCV1Vp2al2C}b~8m1si><^nhvzQ(2xJ;Yoy}U zX3Ylpw?%{FUqEO3WYH;pcj9!(4`0vb_03Q9Y4gA4zj5k^uisK`k@nI9Z(o9MK32w@lbmPz zBkqMG(cHA}yBxE*@EAJ?~1OsmCq}w6H2krNoJ@fKrWzWh28&d*Z2pPQ7WKHBc z>2gryL(5S|H`0ddkUcb3oa3D?ED~?x)v(*86GIqB)I2pWJ>$THFl8AvV=cZUrGLL? zPPF-w%up6+Ig?45TZZ_oCUZdyTM;}6`LxZijdiS9)2(R%BsO~(&Q`dN+|fiQfcINR z64iKbtELQUEtW8+r8*TVo3txFmA8{Jb;68gL3hB|O$=p^pyG;DH}6~T)$?Vu?~FK_xt38Gr4XuF|<)EMCw|o zDCFy@gr#k@d^xaeCy8v(;5FsHobUC*CDFnjUYpc}q)N%HqAiW+_f)cZ9E@_5S`-T#ADNGlspY(L`9ePU6) z|NNgBXQE^8-tr-xFAbl^Qj-YtJ&L_cky}*M<;W$VVp0a}C`ZL9(jK}+c18|Ge$b3C zNaxJX1s;#p!u*JsY1woPds&j^T@Fo4nTj<)$%>RNJ+$Gfy^z~VQsBLViWI0&1D$gy zIBwM#&b9%2cpd4H)iZrOozL;P#qxVRtnk6-#Jz5~<{q!q1-OI-L-6T_Bdc7NPe-AY za|@chR?YZ`^upTjGTk#rr+Og2uTBu>G7X~Z!dj>of+`~FLF8cpe*VX*6`&iZQ+*O% z{^_Ix;^l6JT<~S7l#l@Sm<5{4ULpSCkM98$=A+bz0gx7ZoO9Xjyv`GGw3oP9| zq6W}x-sgGPKf^vso94ee>~nstV?4tihh7o~hBve=S!y&&qPwDBlXz}M$$=f10uwJk zonoP7JB5lm$Ln=N4sqyx8E`|_K8@ZlKKT08z?=L|)iLi_ex5L&!sc0Qf6>?P*u6T( zfKSdyM{8jZXq!DV)aBMHp%Eg#E1CL7vYs169N3kCVx3XdJ6kC>8GqTR^$}^}i;{b8 zeLTz+#6pt)qU0k|$NR{oYHlansersAFw&POv(-n%ljiSog@@TF+HE)gjp{#4M0Lw? z7k8jiIT{{G6uW^Uaa2?ny;O)5o?1*lrqO9ZAV9KqZl=0T+Doct+M7Sfd>}dC!u(dd z?mq|#e)oGuv*DM~-cMRzusl+~$>?9B*e;56P*J6lQh5z?GX&_9**XF44M>hhoQ)`z z@ALy^K|wb-nWd7eb2|k0fWWzdx6`j#_8Bo;do!elDVVl|Sw9~HWJ$eeC!5ayM3Bf! z^%;38wpxHIs@&~*g1$Z{4Rpm?HF);w+##nue8bOt-!9s)KeDw~jQe-p@LxmAMjc(u z;1(8>;a$xhXGKlmi08PL_&3O&X|y;1$`<~r$upjXNXs7AuiCzU%DjryLSbf>8pVo6 zn837~LihS>^@QE)6cwztM>H(?RNCtX-SQ|qtJfY;X^$-I6KLVa2vitBX6~e^+E3-U zMClNKa1p0?+_V0JPtMtF2%bFBCR>(=V2l)x`#%E? zw$W2oImMPyWFHlUD?9v)+%c#jIuxEyEw1r?5&2Y`v>s}2x@o)y zrZ|)&jpyG5GWsM%4qYN2fMgT0=jodtK#qsbTu`I`3S8xFQ)){h4CU_mHsKbw*|%V{ ziLt@=fZzU|ZRQmc4x4$hPz`a~<0w<-f0cKUsR#PK@)-yHbm~LCmBCk(ON4Q}e#Jev zM~XUmk;^k@3LCu3QIFs@+@o%^^Cl;wNBuPTPjk$}EnG4J4!l~j5H)JiKm~*-U7QPR zrFwBu4C>v%^E|8OKRv0i}@d9RQj6pmvI4#1IJq|Fq>gN9O@ER%{6+ddgpVOe?D+I z=SsBE;pM7Q?^YtNmh_H7VBk; zlug0N({eNb6?^UX*15js$KOE1hTXo0oTePdb?74fKI(TiqD9m7J1tq_3_`u5(UL&1 zu@qTHMXhr`s=ny~-5PbE`kkae{xD=xw@X**vxm(}Og#G`b~wuk6OQYKq`xcshPfN^ zLaAk0a5h$`monw^_PJt0rzAk#s83}Q{8wAMeH%?WnjYu`epTMS4i+gA`x@fBviSL6A`u<_GdiB zG9tzZ0l)H}vA%yYCscJzv-vd(WjR9V-0T4#)p;wb7LM*gI=|hl`PEu63 z7Ac@Kz#t)QNV;LaNrzyEvSvZ*^op><0ZB%Z1bf&$t|o2_Kh-XtDkl^7%BlI4^UYH; z4(p6qNVZ(%wLp4y>-1w}jp7qUgM5twsW$b?@>?ONr+=c*o)NXO$f7!Vc7iiH>}(iD z9?uOO_x|`mAgkC&)*%U?UkiSn^kBx`VTm~gpv5KWHqMOFNeh+QOm=0&4nA(w%akcqRZa~Re*Y>8lzv{3+vEWD> znj3+KupmqyS2t*bXtCQcQCv0;6O4s{+P;Xgc?OvPdrT@%4dZy=@zg`?aDWp$rf3iR z$Dp~tYQfRF@CI@Y7>Kd^NIF|8sTc0ychNl|Lrz$@4+Zu|kq{tyE%1~^%hj&4Qb5X3glk|k>ROJCH zL$$b~K|OH;3G+%I;xSO9Hpn;n8OWb-y_4jcBdy?Tv5guFYd_?5NtX%-y*k{XG7Wo_ zy24I;K38?vX9>`eZIRv#$${(y>UvN3i(~y)!E}mx-rhvNF7@;p+c z*c@NueZkexNUz2EMeu4EY)`{J`paR+B!_<{JpN;jlMO&I=5r@{(6W8pyhvSR&K<|4 zR={!dRKx<)Toh_lfg4!q+0j;T?U0sh%XTENr_E+M;}EJURF}RKoge?@YTR zK$Yl0w|r<(9d^Qo_ME`#;H!bixnQ^-udq3)Yyc;txtS<{Oz+?BxwG4RA;bkV4s3la zESpmN(!Lx$A7+#auSLZ{$E6n~y>k|w_EmuQI7`!y=>XemFX#NjGIIu=7b;h5ft9CPutRdlH(UKsvX@i~ zjt1OS?4@KSe}H=52bzEJWgvN99k_XhJGicM^E)h6Yq-L6`A9DP1ZOKNg0nv_Rzc8fskH0 zlinA&o6VvjdI8e%nQFX#JwMT>On>}iRm!vzWRo~Syl;X;YZhPBxX0k}_tWB|Cv?K{ z7>tpdb=-y0dmXnIxf-3^KWR_@o~(A@z<#!gL%W4ylPI!*iYj8-fE@W?m^RU;Q`G@f zPEfpv{seZ79S_j?Y;td)TUmP>i&1FehS^a!+If=`W~WrX^X?DK`Tw7uavazLv7iai zMRzWYg;1K=#2Kio*tS!YdEA&Uc_kI z2`;s2g|JJbtRtj{HUwX^*hYd%BzNcm zl1Y%ct5H_=_W5rf_};#SF>l0u_3oQn-#GCt!?HO^u~#o7V zV}tuir6@hLQ5G9qv!I(s>_E8EW~**HVTe0u%}BiI*xaCS{B`H6NK3bU1-6~@WO5Po4@}xYaZ0& zf+5HCLDd<9!Pd->hsMpd2X7S9yv3DPA%U%$qq69g)kksd$1 zVx?vvqFY?(->T{5LzS*FEuum6K-#EUBD^}c)P9w@6NL*m5Lkb;_t4K9LXGaR{QtDu z$Q}oFk58Hapo(H)65@&bD#2CyI$NydCQY{ux(Vo}k8B`4yd!E;V0ZJ?;27 zaPZf{q6Wn`-0NM=Us>0v(%@7>Ul@#@z6wBHr)f-kK?jE zT(L8TX?gCU@c_&+cE&I@ul(V`hJk-BAsXZWZ`EX}|By??e(etzf4FbqGU`6N+_2LH zJG)+8PVJ)pa39cywEbE1Kt$U7cEz~`I@KO|NmwV_0oA_`CD>^BbKE0?Z)>29Ja?i$f6Lp@@XrOEeea!E*dRC<31z*!ikY!Y1^-yssu`=^%H~}s z3Dd66N%7nHg)3l;;*`Z^_Eu(CU1O1F^ZaBm!0A#%a07^BWuPvHpFI|1-vOCnuZlxPIf$ z{>UVm`@LKi+7297w%`$OW|j!CC|IXLR7ZzzW-dwk=_`s7rUNRma$usF)Fg!CvdHk_ z!+GY7id>-KxS%I{OblNJ#R7qUDiw7N!ot@;Z}D_+JbzKFI3{$_a#qF|8%x-R(Yd6cRk5-;I&bm z$*3Qs*!>hKrJ}I92?bhn*-hRc6)oQx+NapTD}o>{5~y_11`!YK4$`SGa0T9*eO~n8 zjO{*0;K=c58-U^zZ;K}-q6*j4(0WhEPMuVxyJXQjLovTu!&LHmWv8GX(`GOkFC0w4 z@np|dkl;2c{^GS``za%M9=Sg{MTVTXJ-}fbG|5&IsKrw3I*P1Dec?RtP$5ZhUUXiB zV4A$R%jhAt20o*&pRDaY%l;#3iV+aYR|OpPK<)#8K>eimaic^vo;R`>90* zT<}dNiILv?Rao;XCzc8^|3vHdUy?6Ue#GDZ??o5aJOwhEMau2c2H7KD-CGAF)e8!E zhXt*g4IcO9EwbD6kh}+?BS(36l*{-Diuid*pioPa1M8$~6m_b8Rgb7n5F30}IpowJ z0G&0_Ffg(-@CKx{ASu)%>f=>F&_A8YXHaGc@{qih;YWCB^j-QgJ@Q!-n|G{&_>;MW z%a(Omwr~|Eh*PJ6MDJ>hwxug{MhZFPOwO9vmJ<{U9>P&7s`g7CY@ybC!T1C9Pj^M$ znz1HQr#b^$NIqUix6HdPzoA&8Xx1Eh^W4I?H(E49PL=-YOpomDoF&3C_h!xd?_T@T zb}Eb7^cJ3LWm@JL&Zm%T9y{hYYyR;@*_Zb%y!Io!x5719(<3_)+9N9UYSwI_^1pHD zOD&q`(vn+vF?iUHgJu}uXMFavPH`JFhn?Wxj`^?Ff3EyAvhXi+-6rT{5`C64aXvZ~ z?wx09n4R=Z{yNt|w_M-U>38|3K^Q!7#_5Rj%#k_0q(hKKcZGp<1M7#^8_rD39z6qI z6GiOu08GwwofIR8%Zm!LOFDpN2$Z`Mh4bgcs+tPI1Uz zr$QeedG?`L7{y-=C44$nAFtH=6aS%bWbluFZS8B2!d@y~D!$6Aq)SBF!a139D!f*B zLQiEebCJ0qDhS&FY|#du{$Z!B(_&^Ltywp{Qgc|cpLfXrF1?hL3l0SJD)E!+p~wP% zekx7m<%jmug@Hp(_#0jb&JM-jsu~Ian_QMk(}nrID0CyO@lFoV8p7uI3!PgqMn>HR z(B$G;S-#c`qj$dAc*MLp+hJ{!g|2Y~$x@+C4GQzY9~@uz{mYtMk{n;Ver>UlLN>0?wHt>4vM`*k@Hm4G5HOzoy>Xf zTgoR}t(qfbuWzx}$QhL5jPqO;)E2pPDz3$Js;h2yZ}3-K8s`ij>dZ{o))&mR?i?OX`m++!wQ-n5Em8wbr zwlh3m5jN~(IJF+sF>wEZ$6D96)evGs{D?R5iI$`-`Pjfoe9uzjEW~0 zQtWPuzOoyS(lRziTl0kV7z-Oo~^zW6Rlm9h$`y_=cn@Dv6E#ApJNfw%Xp6!$6xF>I@*mfkn- zp63xs42$%d%Sf{CC3|PH$B*D=0&v17xgmHF)u;Hf(Y8$K`{^g-r~`Y)S4@mv3&l23 zq#lf3M3Hhi5(&}BZZ@R2D11u(3^v0c|dx0x}s zLx4@Y`kVUcYa|~Oja8fsbeihz=c3IJL*jZ8*E@&hiw z<%tBJmHyj7&%PFV6n2LVISn}(whSf6R$I~W<7ymUFb6Bj30G6pyesca3 za)g_MDahY1RsDYlU!r;rj0Q!F5~2yve>w@!J6+cnu)Q6pffG;+UKr%DZN1v8cu zia$FQSdo)Y*V3)9X2QB+!|`U1t<&*pjLckCR)T!_Ht3DoB`j8cEc#TsVtx|v!j?vE z(?@Pb&NtD>m4N&ThG*o^?Ib4S+9?LA^Jls~jej8xIs?Xtik!8d5OWwDqnbX+3 zB&%`@1UEa~%jT8(7XHgcNw++k*B1eC8uwV{G|Bnyg@v(9vj%x7G0=m@`b7^VN8#un zM=rJxY@Q6~xB>L!5q8=7QR~_`iE*xy|NTqNB+h|-AXw*(TKJ_=EYxTvQ&D~3cC2## zM6%lXln@L>*mjLhb%M$B{zS3^W{?iSGG3)9+ouz>ov^v!hTF|ak+B$80^crUZ2nNo zMq`X#C#wC*jWdi9A;0;kzm6<(CK)CJv5{iqDH20P<@j~7(cu?@bwCoc1*p20f{@*D zz8y&B%>PGY>|6PT8{V=ERT6Nd`9`*MTd`t4zRMM6m@F z*+oTRks=m<4m-62pL8Qsn0HBI#pemufL2dU4%jX|;<|0BPKE2RUQ!OsuUYCrMT!8T zd4UZsvCyng&CdzMdvoZe!aN~1;@GgBcob@_g#NRaOrxG>tIgu_A5p(FV&>0rt6wMA zxM9YDUE7sYM$b$G6nmc{y@rrvt0tM+^2I-#d*klHI(h1}9g-p%JKsQmNz^T05isoA zA=(T6@0lNd@aG%K9zUo?TX-|*K55Zpv*l2io=onFve_GR;(S)R<E) zIs|wOvhbj)UM;Pr`*;~M>&cNPVwpC9-m`(p>Ffw}KinE}$jJ!3{7--JXA^8YpJL%NQRe(qY2V>(DTh&$0C-oTk<(@20l>=nFa;uOW6phz7Rbu{26?~n_`JLq^t)4Y8C3g=V)t33_}G{DoFWe5gV*3lHNE}S44@l1sLzellOw*i54D;ER}Ra`+rSje?+Nw7RXSS z(YrxYa+~^;x;(IWZmXsrD!)*TtaEy|`jKY_)ME@PR=SO(H8a&GA#1*SdeL~ZQ&s_r z)A!!|;_*R^6QHQCMSkPgFS?DnCOa4!EzFM?9Qc;%_|T ztmEY8v=ofA!zNP64wG3YiDEZUB#w%z3cxr@yJV9uM*Z_6s-cSoXAzxGb8!9>UvSj> zu;->VmO?)r*z=bdQNK2yr z+a7U@Sl=y&Vrh)ww%4J=wZb2L*Jv;{{q?Ock<$)rFz%TcjH?t2S?kMG z)CxYbHFN>pWuGF3?G8(k4oG#dc-aM9ziBk;x(z#Z(cNLSN<5k;yyFrZTpqA3aywlQ zoKts!ZK_Mp=>R3!iRvA^d}tPpnN}QxfyK^Q@%&R{5Y#1sE;yGtJnO6|HaM0k574O> z zpjSE|?QlwKov_ zfXa%IVmuVEMFn-NT#ptm57M@(dLwh0UfCcIV}F(OA{6>Tye&#n-2U9iLpHcSluS78 zXFte|1tjZ_<-Y$$pt%_7OEQgFkS@(sU!S8Fvo$nJea1_Kwke#vv}xjN$}>JtJ#(JH z*zqK`w{u5_^&>oyE5G|7;>+d)R$S2GzyV+jJX{C74|rFKQh`1?HMBkW3f%?s9)@X4 zBmXh^D<*4UF_9r2&%Ie;(TmJs;jjgp76$rE*ag=Py3r$sO#qe<5zwb>oIwaSAe&MfV4L66xfvp9!1&oUDCsQoIeIpfxoghaB`9RA894$95E%R(w~jMUvi&wc>fUx+M0yVgF{XR{V5e z&4CeMLBV+VolBB|kQ#oYN3`&?=#eCcsrG&ZXjGO6j;I&MriRm~^vHzf8eG}CHwU1cs z+9+!W*@{$gE#!gqIyj;#5WeXUq%#@36ASiv4!YIT*MR@Qz>L-)hfZtUTo!s(c@0XY zbSnKJ1N&JHjRoC3^gqJZ&ebMFfN-tivTq;HAZ}Yp?|{1T)l@nZ@Lq$IaGv)%CTZ%h zQ`6@=6(5mLU7P6TGVSJoPQ{Aw>vR)c=AJZ_!#v0tauZrcKL=-{PXFiMWX0y?axawK zu~0&HB=m?s{%zJ^g=C4SOmb>^t7bq{6p78Uwc>Ttw|KTdxThVGc&L>GuS4Cbi! z$8~cYIN8bv~HB$gmYA3PIB;VvatE1Rdifo~x?nmx$&x*_sI33z8 z#IQo_{4U7CwrVhB(CCuq(;=wwOjPSs+h48mZUkk($$91D;2j#*E988W<##?-K;2=p zDl%uML+0wm7G!@DefE;05r+T*I-g|EGLVQ|BFqyf`s850&O?ZKcFCc^2Tp+=UyuH* z_6S+;fk%!9phOYa_9R7$<^+pm3Ob*vfX>&Grmm*&kBz}JjxVYOlg$|y&%A*XXdHLi z|F6Z5n$7*)mn3px!4vKf0Lf5jwe-HcRkK=?CQk9Gr5}Y{7GW7$zO+HGdG>OChv0_1 zF}T@o9rRBjiz~{XdVrk{Bxj$wXHHJ_lEA|Pl~uv}mn27+BVL1a1s~YUf*J&M%rSWl zQ@8-L(Q8RH{fK{*sg_@Li)Gu`ee=@jeo`GUEP61#mp|lGE=U9F=fjZ|!6?FHV^jEK zjJd+a+#)`Cl5MyC@BIh0=3XopxHxdK*@AzYI=@a;5|IQgAV6|AtCmCyEBHtu1>LbH znTwI2Eh(rD$epz^Jd19SU7edx<=c9K9D4-EgOIVGWZR?lZw>t3yd2@_b&dn$z(Uah z3dv@NV*9do?c4_WN=N{sU;r?ablFKCb~6TJ4G@>%AM1X59kc8;#cU6utn3Ct znKGnOJ4S}&S`=KxESRBkdF(v8z5X}wT&rKS9e;~*m$5J{|B(Enu z^vKn%8jOhVo>t>uEggao7v2XHU*bI=jy5FE5@!i_xVLKd&+d};(Akn`VS}vNMqyH` z5oA7zTizcF5S)*3LXhJI#DDir*VkY4GN%r~Hm^?bAWjCQ(YcB&uy{lAgve%9lcGTH{0PAI6uoja&yoTR`~ zO)N>l*0XN+CK|tgu}rKu7D#}#gTnJdEna~i+TdxYBroZq1?(ueo z75FEUyw^tGi75fY-6?)|p}Ya_pQs~cH@I1Vk+p|s1~aba@w=B~@nj(k*%Q(wZImSf zU11WivEEVYNSse8SQjmpf1hEZfoI6DmKt_K;U)C>E{GC+v`EOdi(V7iO#?BT`mldI zIRiN=o$8Y0PIy1L1J#=o#mG}*W-HJ<_23B}nAR2eudf=dj_)BTfg$Lp2uw% z95#RC{?+PMnGqky|Mi`VWP<~nl0p;UZl~BZ3WT+y&aq3TVe9)kbu7PC^%3~A6_Nsx zb``%wxrToh$l{BX7hJV@LX@IT^y!l3G6O(HhqQGc*v)zvAu}HI>YdfGAZNp4+|c{5 z`>my4GJ<6HEaon$c3_ZPGyzEi#X{)dI2F|xwpMx-$OndIWzoHK$2&FPh%TY7DE9l7 ziEqug=ZTU@I#nH6$CN~L2=GWB5dODnuoC5>a4j&rHijMZZq|JEZaUTQ)fP>X;_BS0 zfYTyFTA)>PQMr!U&m5<7XO~gQGmq0{RAyMGyjip2>(>{4_~q^2?06gRNuwW-v|0DT zmSzdpkB<*_+>jb;5oF=mc#y^kQq=A8j=ndIkb3ae+wYOXFU`W~vIzj2D7K!0#Z*+X z`tmfLDw96S?2;CUy4C3}7pi=qDbfb_ahSl#fZ%xQ{_ZmW{*{Sjq8xZNZXpx3 z6f9UiEW%qgOW9MR1VtwOP?D_5qIZLJdnhUS_M?TmcTguCqnSgB3Vz+I;~4I7Sc7pO zWE`Ku{g(3!JyM*FW=GTYJ1tq_z%v2vh6DT&8BI`h&*1sUEO`0o#VB8&-a#g9f zp`CGZ2G-=2TQ|xkw@k_or>W?bd0$`r3v+JE7b-hxfx)svU|?nLrm=kxYqm<{&>4N$ zzY_f5M0SnEIYVkyo9>mwYbW`kAe_~yILg%d;}t#fYvHFo zR?W>)=lbf?50`?kOv{Jdw-zG@_CU_~;W7$Bc0i70xUAjM8}*{c=x%y8i3UhjV4k=t zperm_rNw|OHq_tr%%LwxT6vAIWYXUBcsL(^9G-PSbVb=6Rw{{iX%l7ft|`~f$E#$bFe3EYB!fc?}IEt*NqOia? zOWi=%!?d&y7TOszb$ZE^$q6Z%jG~ngGP!3mcK-UC8Api`6va;IC&?-Y_EaIHKgv_x zOtFa+Nial|I~9pOX+Xb!HSi`sD_Bp)q9+|Z<{vA@xSTy^3rpi;F_xo`T5KC_?*Cwdg?U3KbA`kFF5w2Ri})3M@@MHwCjZtWgLJ;XAyLy><_yCbwv zS)%*^o!Dkot`IdGyXe)f_KV+gg32f#V`S3~DA7NX6I5Oq+;BO`yzr7scG+=T7(fVZ zw8_~-u^TB8Pemp1Dj@uUB{56*mwdG6*ggeN%K^14D95thG*+51pznqg3MUY z{a*Rm`{sq0T!7)gE{6p+`7T;tn2U4Iuu~0#B?gE6_wdpg`>EiK_2La*;v|ZYU;O&F z+W-C0uipFpFC{A}b{R#YM{LLApB5irrxUlg>*oA(X({%~I#tC?;*7)@I#sXy8mUus)}n5C)!bM@=b1d<;w6bG8u1rTwggm+2qlW<>dT1ZZhjsKa)CQ8n{IaMfIBx`y>logBBc{jYa=EOr33u51Up-Kr{Or1K#ub-}f5Qi4UR59j$MLBY))Vohn z8C>pr2h_(ZyppDV@WBWBJgdkBsFl4h!F#nuu0^i;++GIRz|T}4^ur|&LJM0`^Thk- zVKxv9OowZSZP?_jMbDV_z~aG)!-*4mrudLWUwzRl1D4BvMFf zU33xM?1Ae;Y!^UMZYHT&SRwa z95_@4QgoxzHt7@#0v#z-)DeD-cN+b%zCY&{c&vHCE=4U_jLg(+(hqqTMB1{jUXmuR z6vaUoPcy28(+}JyA$P|W&qiQdfyjwnVj0g@Iu>P{*A{V!vpVh~j##KVsutAJmj#DF zSR0jpN`OQE0tCn|3$}YD@Nyu^X6UgliNNd+@a4TR$1sD-aCm!Y=s z2)i7%Y%9i?J6`!=_nOZuj8n|!gn+YTlQStbvD(=b3&J|-R8&6@W3*~gzwm$@Ce2|D zvK-e;^=2`W7i3U37p{6E;mx+U^s5Ll@V%-}&?(Yx!CBC9>z`9T<5*Z5Yv(T?m+^S) zkmxt;G7OI&%4s|t*Bpyp{mJ$hT?wk)E5kj_@2oT%%twBNwoGzUk)-G)RnWjt>AzQO zXMjFNgmJ@qz?$D_-+fMiox<1s+<&^!fOIVgxkOUO@=+Z)qFQER?RHV@PKspdON&51 zdC(2{*>tKR-d_0w(jHtAfr)+yZZwH@@b0)Y2=?>R#DZQ=_Lwn z7f@ukNqZwRcs=%H`Lf;$Jx0x_4Lbn{9$RrRreA49O7naB7L#NL4z}($8T(v{&89$& zFsehalg6gcBfV&iq?B{hVP18 z8D8jLFl~to)Q-5Px|cBarLLaP=-Y4bU$JaK#&BY(|K7W#DcFdWHA%gNWVZuj1*DKi z#Y8G77TPBEQ&Gq|f>q0Z-lohGqQf6}f9UzKyhBhGm__e{Y;`O@Y3gx#zO*VZ zjjr%20hy)`)xB^B2JunSX}|bGPb@l0n~FyatKc+xr4I_Nm%wupK#Xpm=O_vEi4Ed} z@H5JeJlUMs?=x~YC;YtPwZHx2SB&^k@7VuG^2o_yV}elFFd9EG6uXuptEi|GZcE*F z&bbf0n_sK*Z_zx6JS-Ub z9JpANFm0=9lPp=?>RG{SkUjFn-^1`bWI)$>5q+3nOCJmSBm{rgA*XISfeeK7x}^w~ zvuS=ORrx%;!V;9NLEDq5$qriY<(ZXw-&|yy4n9wJ2yZIMi(w;N#&Ur~DCksq&C3HpbF3BN( zAS{tVTxF>c(>&TP%k(MXQqsAU%$6Z+H{@N1y zqGt>%>GRA^y2O8=0CnuF$K(Q-h$pOMS?<8iXSnWjXW}xz4>h00bcdyF(g-E15O2 zcz&m%C+rC1j86GC`5W#);3X+yd>ZIHJCYf2z}6@W_;}N9INRL!-o}X(xH+&RVxhn- zp1&LD*Veh~sVs76EO2UNhn-gN@dPe+3~L-vR-9KEob6IIbJz)`YL*IfCH8MVA2&?w z1`xx8-(5H5uU|ASbyjTr`gcgxOS9CuV6xPyr&#C#`-qCdcu0pJF8Ce~MKy9giXqY1 z7F-;ZCBCcZlE(9|%s4$A6cmc|VYbV@MdD6*0~9UfM_iqIBy>MhM(y@EHG31UL$p`Y zs@d&vMcF5*a7`u`!*cxhGsVhukSkozzoHzZQ>L9FX`qsDS8-PKfZ$xHQwiFnIF}MO{QaLm-S!dE40}Di1>TSv+q%!zJ-k@8$xKE)? zqjw4K$xlmls_Szap5qw9Omad#rCH;dAqMDc)wXX5&abh z_I54kuju9%2i@c!^SG)x`%Z0C>%vphdc&cVA*e^BQw@JH@!cg~$b1uXj489vNY;4t zD_UX6-=?@Dx(4m4#X%`1PV`&zMO5vG2@fKilIse9JTt5$L z&oMc+Jv|Ds+Q0vXG4DuyNrE#L8f&qwr4$w~b&}>8xj_lMUS+E$fu{py4lKj`&~t_7 z{W*vI)BLLWIf0``aMVk){U%p<*#0*d`-y#LWS3s_nC^gUr3ihI4pkCQyWMAGvr}=A z$pP8*NgHFs8@7I=#~u-T9p&V2OnJ1T=G*3}A%|W1Sjbr+ArA7dJ%So-Z8MY1oDS}i zYS9!7$@ffKE!3%-ytFldks9F3!BEPj$R_ussjI0Hx`c)%<**nwQ+>xJNwJq~gq~CT z;K+KoSezp4f*5XaVQQmSnb)vbDE_$gFP9)bRZXuYca#l+Qu%?y*yY^UNg8cEuQo$`_3Ei?(puQ5fWO_)~d zy-RqJ0c9UUaCZ>e6%!(^@@oCk#7Oj394Vhz7dwHEpx16SVIP zS3ddm>Qw3cJaMxso}Z=G;%{$IAFqTS%bz~~Yh-bTv+Id<7uT%%=gJG9pzFg zl%XJUCjEd;hyXr7vMi{UKIC5nAnG8?Ll;%ci~WJbcF?_7*5p?)bMYeNwjH?>?--oF zC$PjTmN{*g_5vFr#^yFSQ~KVHiL9v|IFn>yO?}fN4~i|j=zIz};SzoFg!vS9+m}O> zuR35=XtXfff1P`U7dBH*+APf#9yZ5;g$J$@eDCjKpZ2 zcC;FRzU2+ew$Z@f^83_Q5A*W=s*SV$!)kaZNBoiQ)CEixPC8M~HaKY7nwha4x4 zCwJR#i<>LMWxXRngk67O7nyp{PHPmm2Z?4 zf(LO`5~C@4ZA;*4*L-Ty>;nupC%o7X9#1{lGG4ZXzVlae{vIw1KLg zzn4Q%oN^D;zo1%tnYhar4Qa>Pe^fs zkh!D!)WF;BK&32FW^P2d0L6h*iWbaEo_DM2tSFDSY34G1kz&Xxo`?Im7t0Pgr7-cl zW&FMV_QZnu=yAe^`OS9Rult?!(>X>DV76x+$}$-hSW7gB%3wYWCueR;f?+3EAmdJMYgC#VFq{zgI?4Vev zp4&!6fp9{2H-tC_LSh55pnbVHxX`~!*%pjl?pdCl;iWE};kQNWpfCZo7zRkUyk6ZP zTP_*w+ zx7(CybRMr8=<>@2`oJ#M;5N$I6*)7O%TTSRPl18iNwPKwA+ESFgPq-Wqr06uU-Zg@ zE_#m~n$cY`CS!9c%aBR9d?A+Ua=t$m_xuCr_*#+X~Cw6?j4%(;3T zw(`IN`wj|}Vz{&BUr$w^s9yhNFw#&USfWvZgZF zp!d{GCr!08TE+nrqeZbJM&QxhVDj_A?CT#G&B(cLt_UahxUF3rce{N>4m=;}{8mj>K#FD$h_@sG4QZ}#ci4W#R#}`+jd!c&V^Kb{MOv@! z(eti%hi&uM`4k79nyzyt=%f5zdbw|5Ia$PmBA z2T2rcZ^>+g1DjhiPR5IROZkQO%}Xq~q=pi)SwdKuB9bFmRA;7&)ZAXp$jQ4EKxQ{U>JmZPBImOLvxa~@L%N!*yiP^A#DMRrs zMr(6{m=s$M4aEr1D4(P_<{itw{kd2%3Wl5)ofi!|ZJib~J6hNy!YHoo=N_{mA19{l zUhcH`?RAUH7f@VK;=rD

WBzg?2#0|z^}B}}#8Fxd%si4_-41**G{VbNZ(PIcKg zF1V7nO&J?J;L)ndq<6c5rwX2FiTs}yCPnXVt}t&Pdr4qoq1~*(V! zQu`ohyjEq@>}6)BY_of5`HFD!1~M*iabTBqpUJXlC&gw_WIGk5*ISMjZuLh9^vGhT zp$?+S(~&=4ox&!bQMWb6e&x+!RM%& z3tUT>F&cKl@NWlPgek=(^ARGGQaC%lF@JiDY<+~2BRl2izpcCOYDCDNw5NYhR&xv9 zIj}JS9jZ~~`dcVAi6R?N?&DK=o_M`0w0x-zPHm@RP*KB_dt`^UNmHZ)(z|w~njB-a zxd3U5+wFFUZb|;SixC>NpR4{GS^m;MW4j475-B!;BC%9dWpK0V8iVT^m_8DH;`w>P zyKZ@6Oe6KslN~0p9u{^(z?cs{^tHo#C*poMaMis9_v04y03IRH!XpH<3Ulb3z|G>A z$kh3J{qOQ&S^{<((j5}Z47&HIP?lk*dlNLCB5fcF9}c>I$itw$U8|U^hsr45^zr7H zqimCd0WDcu)J6+(JnC;7NwEXF9}Olp;V{KkQKX!TiWT=lWptsiP>8vmJ_Qydp!)$L z34!Z*(2k5$O^~7iP0Kdv1<}DsK*$}JJ-m;BW#v5SR8-BpA2RIJsH!5!m=r6<@JJ5* zsk9PkGIBizq}mP|PxgQVGw4rD4D)LS>R~| z(f8|LX(1^N459-jAj+dy=+DTcqJ|>Z`r_Ya&5#o=Y_ae@ONb$Oo$5f)eUKJELiYN` zGA%9#fVd-z&X*hrYSy&C)mg&zJUrIUo(wt=WVZ#(M3*oVM8q?0=L8n&&FN(~0*qiO z?kQSIavT^eM@_)8pJJiwtB8t1Q?k#qOPU@E)sh%!ZS}kx0jr%AuNC_N=>YVU?(#1P z8>w&6o>ZZU+QVBjwTP*3#hZ@?v;6)4x9I}klMou2s~H4z;c^(2d`|N%?6pvo3{oKT%KpEYq05z017@pokXhe74GReecNI zC8eUnpv-~itLb*d8B!IXQ(fNS0xy}Mmi`pxsinXU_u-7IAYNDt*`8k6 z2_ThBh&&pQ?pDIXzQR572iQe8HBX3HA;TWZ+8vMVp$(s*Q7lV&>SoaW$mnp~iV0})UM_f!xOw`w>NsQBl!v43ItZE=hJmN!hN*YI%-p zLS)jc230-pvM)-(VCmbZp55~8(jliS-L;{;s=wj_=27ObCN zt!@s>We(3eD@yUZOP^sfc-883VTQUS@-{5b(tN7bJ+g6w#;O5?z%=>AuYark-yi+z zz2E;*vXWw#Q6$i|SP^5p(6Fs-%y@_1{60=LW=g`(!@bGKD%>gkBw009 z&$9!kjv(zkDs{A(ViPHnfT^Qo@Pr!VCC19h%gQCf>`=%ohqkIffCyIDcC{1cgTf{4 z{M5bnJYX3uUp)BlugQ(bNSpEBzadEuY*N62j6y*c#X{OBjfyH&R!bjx9`j!^e?PC9 zKOn95Y}ItDdqmazde0_$#cOMkj~9yKwtAdr3M0$Iv7R2J%Di+cyu#iH8Mg(oCm`V2 zmN79#XDd}SE3m=5v-N4(F9)_D7P?&zsk;=lu-d}(vmuLM_a>vU$JH8`jD4qlk6MPw zFUf%7MekNz7IX|Ky?028=$-D%VSSUVuKdb{h37P97jCDqzx9eT`t3`S2Dhto|9(#M z_o7J>V73MeX3nU6L128kWh{(kGIACioo-&q;;;c+3$-aFVc@IkwdeX2I-dd&q!-u7 zP6wlB0cpi6K*RB+MTzo#W z@2q{prJP2@af1kp{eJSvn??uq!CP;?M-Dra%O-wc6UEk3D#J_4(jqU z=&Gy^IKdqAE~8pJ8|nIhK3=JJ552+}nHP1cyL?n$S?QcccR>ed7w9*nGdJA!`qp?C zge56beNcWyr&@GJde3iBvurOZ5Jd;2`jq(hk`%T9=x|y!M*}WHvyd$IE*SujXtZxP8r^!j;trxLSpm1XS;T?MpMK7(umsRnSo~ z>@>y<=D^Cd=?` z-&F83^V&*>byh4?OYW85fubewOd18-eX5w`fKffAlMV!HAYd?46C6{Hw_Aq5SQa4R z$3OUpF&c{BeAHh@mN{^EB*VlgY^2zDio{^OzD{*`7A|fVBRPw{1H<=@G)4BA4dbL% zj)K*wu=07fx{Mnt-gEt*VDtVkhcy-!I=(hd|3timdE}YlHV{$6?BVy!zC&M~Tgp2@ z+Jj@6%mros%YedSd63QruPmdx#reF=udO3BO#hr@b-FAja>yy4Hy~}ImveV-tQ|J1 z;bOtzvG#MgUdIV8RCK{g+TCbG-oHQ=k%X6~Zhw!7$;qHt2(zUcD&~s);z2>D2fDnp z(cvg#2j#WkE)Tj5I}OsA3u@HZK)6l0RJIPBmXfdu=J zW8`qd%rE|}$dzxzOkncA*O8SDjG0Um%xt3AjTDKeqWb3`og^-ClBR0Ws`MMIN){-v zA|$%(&t)-s47mVf^nLc;<^+sae)HN7517v|FL%8)%1~}ocGCyuH(&@2do&PJMKdOS zIllo%}$WrVXG%q)N9lWeZG|#b@3!_ard3&&_a%hbO_4myKddG0?{3iS?{6y zf4lpq2@#dhwE8x!tm&=Bob_ut$7}`Gw92-~} zc@TI^$K}~PAAMtaDqN3ohwWdR{66^Sk_J2S#Y>Vy{tW^sSd-TaS9qL*x|24=15yhz z^CfiE%pFi^Un!c*DTos`xDWBT@0k6k_r9hv+M2G=87btDGdXMWB|Sm0brd-YJvDmn z-!|#F1;D}!WYXGne%U;n(sKfvRn1Tw3Pb^@f>SmxKLT~9hvWxb@<6M)!s{O07qM|( zUj+1eA_Fxp9kS>ZBp%Ym{fgN6DQ=KrgEC`G{5Hzc{JMFuOtY#>x*t-)a7&}CgFfgN zEv$Mqc76?m(XTpsK>oo8ABScLv#Ft(Zi&1&pIGs^pb~meaB4b!7i#4_ z%uc#$YL2%S71VbL(`Tg5*y6duGgsZ}pFT_IGO+0AR7q18Cqy7((4;|%a2$5(xo_F>1srf;`5|^V!wnZJ>#i*|FQT_ltB%Zf zf8M6d6UR)$cBe`@fv5A?;jKMBO{ePPav7Dm}8pWPVDb*bNkkqoR@n zug*;X<)HPv0#S0{64(Bj&{-Os>s=mMLT{RBUoby`&&Zb5@E*Bgl75P2zxrKsT4IOw zL@W>ur-7?mAi}jV@+KMUq_fnUUt0>o2&g7zh^1)_MM_DoInr20`=Ri|NlExvJ@DpM z&AuhN+1lquNe=S>lv0va*TOcRTvFuuWYeodYyC36I{5#w_a$&m zWm)<@;f3U4$VMP}1u77rf@mQaTGWoU^uA7a&rJ1FJ-tcyv^~?^H8bg|?wYPf6n7C6 zR6qkLi|n%6D7y%#rGf<-1vdn-N~uy%q)_314hc#kvGO2cqW%BXui!2B-C*wb&OP_s z?|f&ePTe9LcG$`tcGv(_w9n4Q_IrJLg`Gew=#uBT*DCeTfg0G$2?2Y3GKh7I%oGsx zG^X@dtYTbP&5YE6?;n$f7!V|MTE34Qc+T=bPZ^=AhGJk{bchPi;-v68`PBivkRPs{ ze@~1F#u^lcL#6m`g_BYNi;lVaKWPot%46h`mM76g;rtw6xSOM|^?X z=dvQ-J%x8jRl!+-MLTXoiSBerAV?!;+|DR>39f|n(QS~OGVG8LFaVUxW?2;@(F$%gP{A{;E8M zcR;CA-;rOIw>V{bOagee1KW5L$KI^fdY%%Uw-Q;Z#(r2J$(o?oR&l;+!f#q2} zPwk%9#}4s~!Tz{B7`Q-oI+9YO7dMk)fF_Vig~!Xmf-nCN62x-EL+)Mjb&d%ERl*D|lA@(!GW29h3b(ns19DUE}}pQ}~LT5VjqV z{I8e7jd|&vcQ8AoRRNc=^rrqq1GPKkJwT#FT`#zE6(s$$-4Z1?dFerh=i@}>OWGwl z3Qdv%HCyl$H%bo7gZmgx%og2tx-aX-k9+3~4W1a?CD+1^=XIzm1DCMOI4%M8&uEh5 z_|f-aPoza!9EyKPG`SH{%juFA$}&{Y9%am0qYK2bS}>prM@tC&fafT4`bkB5QEr z)rA;HF7v$TjQh|C4Rn<|Y1og{lP|-K&0^ zoi7S&C7RWNDe%^JxWnN^09+avrA|Jg21Dj|WnJ=XOdgW~!<#9AC0V)(i>N+WxIvCh zT7oL)YK{h=E<9ekE2-yRcgg4C87j^UGK1fL{|*+s!9gcow!868JcsrVo;&Zp~NFaCJ-H}+BwmTSLtZ}~n|hEpY^ z4P?^C{LVuGyh$vW*pC1{!4@5p`xN$1u|vn^{MRD<4d`fmZCg1>b0l>}gHS>-P?|8G z3a^o@$xXz-Ozf!DcI*(A$@vdLHtH+pF-6q_4yDDxb0)e|p9(>{ROr1fbH z%H?U&xw}b;tlsk>1Yx?EqX9YM5r;u?($&i2#&&ef+jI6K@T@1;jSjKb?SU{#eO6+C zjqr}}_ax?pak?QrVcZOsLNSnppGbu#QM#|)Sbjy>53;E5SAokDkV_A`+=4PS@$y>V zN}w7IhBk_BL3Fx%(PbYElKzmiXJuGV0If+PeS)5>d#PZ?m#SwQAkw{T;boG<4kGru z;)D&3@es+OnEezw@HQZy%?e z%=@7ME!vxcN2JY;y_9~V@$aUXPKsQn!n@=T=yigdat*qvDWFd|#ND_Ue**j5S%W@I zo$p0qj8nT%GbB&spHe*(CrUNfJX%2KG=&!jVl`)8U;ks>@-!7(*7_uPws6{DIX$YB ziaW086kn6)0E;vtu%4I89g_Fb<&en$ISgHLV4L*I)-NsdOmVtFG(FBMF^jU3+eh?l zNbFs6jlS=iEgIFd(yDfHRtchmAkWhMHXoNqWB>Kbl7d&Re3~FMW#l};(3FlIL&k8x zZbm(`fALpL)yiIwAha26GYmU72Eg*E%rnmCI(L;oGw@2iI9qg0hNIptLBvj@+sxK+ zHXUx;oqs=i$A9q+Cg|03JCBf!c5H%R#u?|OBvQ;yio{Xjt7h+I8tF>!qY|AuL0J^2 zQ#VUjc@2`&WTRTAKFHnVJmiw)mp`2}b9Q#acEfAhGb_76{bKvw-KOApLF|SZD19PD z5@R{NRo*Mt;M9}sdo!@YKa<`FeD~30l}lb_B8(E+%2zQRF}q!I|3K8d;3o#l;h%bS zfaqQrS9ie3&fK7wZi;kL;hp?!uT7k4l24zZ2SIoFl1IL9uhSJV?p;>#V|WcT)FqXBB_?zc5$ zl^st~$wuIbrI=`nY@)&uBtx!Of@>hK3(hE#c5J-M{p9;IKC*0-jh~s0GgfocjDP=` z_mzLF3~tQHx*?5Qg2g;Fed73FeSPw2bz==Uuna=eUok5vR@i2$5%PjKEM^o%t}{16 zKw=6K`7-HxS*7barKU`_m%)w7& z23@cz#;^-H=QyXFE-ZW2JTWP2U?FCno)}E7OzOj*@zC4lTi-WT|7U|4`%MBQX6gdq z?)I3b2qbkPJpM0?h=q&dXAGUKZvA|g!*4wd{zmIB%2tvDJN7pY8u=Sp6ayiVeN^~1 zvSW6+sC#a75HK5nHFHCpB1V!x75#{($x@+$^C|Jj!s1YLOn^i%~{fB`&6D5NtL0>=GAdytVyEQaEkgD;nvLtuTGdLd z$T>rMixa15~UNx(M`V{tDUnMesCnL_PL+ek@>|;$&!!Xno3a3jtz#Hpc=B< zvIX7r%ASSMoQ&X20c-qj&o1D`hUHR8!B>MiA*-%nX>W)v=Ni-DvRByrW_#spQ`xDp zJ7Ph8hlZI&9r0(>ewU>k6P~F27J-4+=F` z{JQ6E4Z6!c8qmo{LIKoZmf^N9eeR;0EbbntK3tVJp3jD*=@ z14v|ld!v*rkLFnNq_>0vO)1VVWx8iy`O`*Yds-aRmQ7TbivZ&@`%vB zup3gH`ii*D7t$J?qC8`{pH2su02Y(LlkLF-pKx!+y8~|<5R!7g(UE+@4k32zerz@( z`wvsh1Bwjl6KFAKq%5p5sCfQrpvof=C7QF_x6xlyBe@K;0Da0xk^gSrE;%|P7npu| zCY|8h^Z21O)e){%b$u~Bf+RRRi1hECi)8z#myb&{$n%efP}*72=aCV71~|Db(r8h( zUza#ploAYFZm)V$#oxlfE+XB7y;scAGctX<6yTrCA^vcL7SQ= zGw=2l$tmTEKKktJk7-?4=ID3PsxJHXMW}?;rtSyzr%wK9QDx8o-AQUdPVNl1b@8;g z3X^!Zr)A~AZmzIfo~`4t4^F&@6|4?&ON0qTU&ToJf20MEbBrbMotC-C_E1bBMRro*$9R#hne=TZ{H;04>0O8?_*>H~tpT}4Jk=j* z(jZoQ+i6;&Zk8bDiF0f?G8tuCoW}j)($7tUux6r{M?4cj5n~Ia;f=~-3L*ou;eU-z zam=YNY|}I85jObCE&oDo@Zrzs()_D9QMQ%c&Ln>#zO`R4IJ@m%`zV8yIFd6)FI+vv z)Ka7hL{%Y^SD$07>60c(F!xupN&pIinCpPKm34EwfjXMVDOw(}G=|p#KjY;X1l=Cg zxiCXj4fQlNr%AJPlc1B21vj;d0&Wd`og1-q)M1QM^(wWB*q{g8EY($coGKj}B-09ye4qOp^2^viiZS^Plc#0gpwT^)Z@m2!`Bw-6cK?{!hVth za(7#TM;AL?^3NnyLA7qkA4n41`PHXiNqJSPz@35p!H4}cz4G%C{l#m`yJaO_{ql#e z7Ku=1D37s1px0J_|2#l{-hK0ckNnAdwLdjr=%a-$9e}vgBvCaWe84@a$mNbW=#BIZ zX})_uL=LOH?=39lw2aCs03U^aDB~FdNt>)h{);=^1F4{)`sj8lAKh#y1}Fy)ylimR367Cu@Q)2Hgow z_Sfv==7!=2>65Q?3XBKCmB@EuA%$N6B9 zM$PfCvL*NjR0UP>hg_3E%yYy6gLv0G@VPWqF@ccVXA--On6VpWyLHAj(p9ylL_0GK zwM-hxiP-{3;K~V(^T`RuP^?+|Y{M7Lq8HzWpKq?~7X%V{L8bdmN!(A9WbD|nHdFku zTzpd=EozZ=$(zJSd80Jkc8Qi#YLof{<}jkaZV`f~kr*@f*P2IgWZ=s`4lqnpMZJZQ zB$M3^l^qBBjvD#U2Pvk6B861Ao_%fL)wP1(i%Mf!#jYyEK_2M zyh5m}2*E*lCGVi$Doze(ujINKHI{Ng&$3t;>z*hX@$2NrksaJ2m+hWfMUU{DM+#4; zPNwfK>R*I~I*VmpiW(?zblanuA0gQ9+60lGfq>giI`sfu6xb)LbVeF(J@*`)6PEY# zrLY3-AerRgSa~ES7%_{#e4~cu-|h(+O1ANk z@SU!@O-Bll+_}}K&jXmi)3iLd6dvQOniU5B^SfB^ z^M>e4e)U{L!|K4D;9CvJZ$Qb%2&#MepDwtnEB1xYWvNmozwaOL`#SWvg%4{HtK$cW=G}8{J{` z-P^A%cDc8p$u!2#W|pyI-!#?8;%ujw7>aD6!mo(0E?mVsFYKc``MJwuc*Vd2Jjey+ zrTG7Yd+c4g`%mqgzHFFT?tkU0zaVuljB9$u2o+~226nYSpu)%48W;h|@~e?$gGjc% z+5|>IZYg8kZ_uf9IydHJOov4+_GnH@uznZuxGKSE6XR1ST!C znB%6W$sCSB#q!_(;+*(nVgOZdoDWsSkGVFK~YB55_ z5sImyNF`(w2rx*99EYrc8y=P3jf+n(ts&`xCQ>1+=MHhBd9CV>fB|w{T`2l&{#8La zwS(Iqn(UFFDyPzxT?*-vMf&uIwyAs681~x_E_0#0KXiyL6qQqXZk^6i^FW_N*+(uc zD;MFX#Jb-niT;_~bDT*S6q6BVfg(1B#pF}ipT70K{y#M3SG^#vj2Y(Dd6G;c2S%q( zSKXl>iBE*}@j*=4Wxdl79|FjsIPug39aeP7%fk9YyJV>@siG7RI6N&nDruGu%i`UT z{Bm;E2~z6?VeH@yPqou3uw|CK54IAq+`I5tl2CUD&tix_XEA6=8yJy!?RQE%|-2Q?$doEC4n= z=@j`6Z>(N12*qY~>PEW9yUu%u_a`3LU2fYl?|y2-Z=#`pt-RJjc9HmCG&TuLiU^$+Yq%IJM%Qu}vFOwaO@Gj|Qulj>P?f)siDu#;WmlgYlSjHp_2S*Zkaq>VQF)|b$OUGcL{6!& zb6#{%rf=E27@lUEb9YFzsBc-bY_n4eblES5%_HDd^(`y$z$++Mi2t2bv;a#SyE`cA zhiYpbexPTTIS=ThkBxzdi4Ixq5jqs8kIQ%i;`@T22)gv-YOhz~Gb+ zPZ9Fl-~UnbKR^4=U;Ov)#A_*LHANy!T@ou;4R&3ynyO~b{Og?W|6}>U58d)YdTGQC zRs*S1wn*;^d%@Rjm+S-_X9hm>>R_}$B_j@_B~5e+{AD}D zS%@rCq>Y`kgA0q+w=zzB&otiqT+wpPM0HEa7A78qG;}!r`Xq&TIhOqdGXr&F!17m5 zkAQ?>ELe>{i}9OkQ}dR^D>&lM|6%ZJK3cN%W%7|76BpJQ`7obR4CwmzQ{geZJm4~I za@AlN-c^DuNL49VUO6{|jv;XA4qU?16dtx4alox1RBFNcz}GzPF2bVanoLmvsJ&o8 zR=k|y-Rlh%jbcOb^zSY@8I}RQuLKVtLGId`a{;2nND5gF)XWCbjPJEsF}aRPMqvJ;+y2tM<3>H3)3pDFYc9}M~YG6V4T>VqY^swv<9UGRW z>I=)_LuU^my0&mL;0#;z%T9ah z)E!V_I3nC_j4 zj)&o2I`$X57%-&X*>EB6c`q%S%I%dFxTkUt`qlFi$f>U<5KYIA;75|Y4~D!<28WsR zE(SHjVk(!KCORjp$mShvrXu=l$AB>-`gx%BKe;ECGTzF~yio&X4lG_MiV|07%zAef!&_ogF~zIL-`m z1LKrwZc@w*igZ)qE3Plrky_szLEK`^7T42$RYCpS?pIcfI3O7dRX#LpgX;n?%6pnF z7d7(-T{JaxGk;5P3M3b-S(*-8o3Psn+5}jtJqq;F;4aiq8+2*rV;wckCkwX(?{@{| zO=XobPn-wk*juEaX2YqRdo!?kbeRSVF#Igk4X6lMi|a8x6YK%O8W~CMtAIiXOE7MU z*Ii$@xk_-)d1VzhBRGSuCYlZMRe~02CrIj`Boux`%B-ihQYV2ngJrVzFJsX9{Ymyc z(;C)xy8~*bDs=%DrHW7_HBEJs*WecAxLeUjJ`yCT);hLAQPBIcwZT!2Hr>p*2tGZ2XJIgc0@i!hcE1sG6VkU5Ssd+yOb@wsYrreP? z%SZ}4+hoVQ8xYAFXPXXC3}ioLQ{hp(Q85vaLg}T~6O<{}V7cMy#aIjF9i#)dOLhd8 zDj^P2xlD7{1vO4k!X}IJ&Zzb$iUULldO`TD0X2_~tF)^0*UMf*a_Ku(McWhsW@Kfo zh_4w=znYb~+yI$xAN`jDq=Fq}?07o6U<8>DDCQVN8nAL`G`SHH6_8G&&yxk26+@@2@A5r%6R!lo)?`F{Ej>*8l%roX^vxooj?oUl?t(YlNQtOovP{`Q`G|Oy3hcGQfgDE1#Aw9GP z>*(n>j;#B2n)%m^U1s-fk3D7k-{!yV3490peII@X?CgQtq$y3{;@}||1^)bg+mPvf)&Tng05JZG3pB+v{jop1-4_dO%~MotoAcN z*e9e=pmwO1QyoKz>?~>0HYx;5E@jR!EgJ9mOf3<_U~UyPfj+*-PJ( zzhXem+m-5Cl55ALnBzt`sGt~#-j-3}DS|w4+A>g)LDGw4?V>z#*zcWEvdOcQ9B_u% z+7QrTJ{4*eYorY>n$ILrPA;pc?b*Re_0}q^5@oXzAf7-2kERRfU(oP190KY!*89yM!ZqjKH>) zVm4D`17%<8VTS}|J@=t!vVVbVk}Jx7=Dj>1?1Ozhow`o>30=T# z2pJ3=cECwx*ddye>749;o9HVRM+0`u z{s<&p@A~Hp_q%SG{ZM?(qg1(T-k?j->)EsR~L4mhO;?PCfSQZl`RYE8=s?Gj14$&+<$0Zg8<0imV171FEcrB6je( z*}v#b(}j|m6~+moRbeJjHyyF4MS5PC0kMoY5hQj2#ahR#L}mXjd5Q0k+la#<$b4>* z){||*Rq7AsLb6#r$X4FqH-S_3Ls1)aITLAKsqfbfm{;b;_sMU z>ULq-LvbU0@$0S2d;MG0`^f&EUs--Lus~!>UcwY1Y>YW8pRs~4)$+>A`%Uwh*m$>g z9DOm9;grwY;CB{Og*pO`DoetmUWyGHT-YVo`mAw_dTBopE~8$u+Pq`Y$g(hn7WHSX zMuM9i+e`~avO$I&uY1hYy+qa;l#Z5d_s9BfApfCH(r%Y@haBhGbUK`9%q+*nL{Ho9 z+Pi=K>n^EbZaMzVZ(SnWUYH;p#N5UO;r3EYGDUV%;Yp!`l54^GpE`B3@V=x)af4eJ zHYn~Brii0qO>}nlEjboVi{{jN4LfA0uDkSeD;71IJj;Ljym2+d^M; zMHeacTai95*0o=@!(~vcQ=gYaa|T}N56uA`xfMF~BjsgrF?XXRZ*knK<=_Cs1|3wC z$Ub>xL)iA&l_BNJ_tIB67W6@#nlhn)<3{ zT*6g?W?3icUIF72INu(X+%p@JGD9#^@=l400giE*ms zoFqxJZ$O=)6sH@Lts!FJXiWhr7JkCyn!O-M`S$sBom0`TvtwV?jDB4TCr#DKFaM5y zzpxIx))S;Vq*?mNuaDOQ+ID5mn?Utr8s?eDxMKzJ$4}Y*ppBXTCDW{XHr}fJuH}iD z9DOYRmlB-qyfRwUD;&>i5ulTSbs|mWT)bJMI_fqo>!fQr{oGXV&CG{h+NDK-r%94RQ^36_#4HXZ#%9vZ zlIzZ+1uRe{Ptz?!sxE#w>P29#2-ecV?WAbluoP-mCc5fnX|S%WR?#j$MBfe70Iipg zb-Xg?ZG36XY|RN0!|UU%=Ts9sg})d1H%RKOQyx+QTeqI{0Ev7w!HE_29?!qFKe!#) z|E{TBvfKRtGnNKPZKKqC9kW-0n?9OKXCy)=K<%{aOun#_zUg+#eY&#D&|a_{ZQGj| z!zZwsOlSVf58}3%?jAm!>}q}dN~d!gdOX^epyYAm zX*@8GlDCWgsYz&9KW|97@*A>+-TK*%<0?5u>*qZblSq-Bpu+2M(OIJ}lpOPtCWBr} zs!1<>E99Z5S+)W6I=W?}D&C>}p|K?I`?WT!bYndHmLSJu?lA7aMrR7aPvWPNEoR4F zu96%fHd5{?RtQVBUOBkfdS1 zZ`^|Vt3NeWG_+gK)QnDddT7V|5kLLUc=5f~xc2U=&5kH-JM`(%cmNLz^Equ-E9X6d)EC=yZ*1$IQ!0>9Y1M$~Shc7X$J#V*I z(HeOgF%+|fA{$Y+u2O_E8L^aqXeP?|kX)I(`S|zwUeq`P1?8gRf8KHs4pK}BMGC3#6sBBU5O9iK6PV)E##~>#%{ktCHSpPnWksAC`a)nM z=!@cBPc7&AVoU?l-*4ug7w#0r2RF<5q^V?xdxF#{rO`w&W2Kvg!_TC0KVc$FmSbxV1i2!ltl3 zas`+<37qb^K-L+50|+{8%(g(dx{Wl2!5OMAH_~lPZs-kZC!{I@$KMYh*e5QeBLr2# zJnovMZOo&oM4GIJ@NqNM<3l6RvN~ah@WubLa;7Qu)@~QvX2`TLuU{5ta#ja^3f1j1 zIiGpn3+V$2ZEH|2uYqWNnk6+hG?~9*!cJj|k2FX!Cj8;)49DtBsG?ZEUK(RXT?CeZVw%-XlmiJIx zrQ8Sx=>R0D=>z*KcFaBqWt5u2uq45XN&36y#+LGU&H;zrRjOMWj5p?ibe7ABdQHRfq~pJ;Wg0$PHmot8g3 zLKcm&4ad@4*R1Tt%x@q1>rPV@+vkd#Vn*2(^G`7hSLUhnTq9lYjAS(=v}K8-dB~be zS2cqt8b7PjIU%4!SUNv8v!V5>+$#B5BE%%lrFtUS$(8#ap&H1QOJSI6~a z-9z&T`p(tpyZ>WY_x!5ypO=$6?0747JiBZ)szkGDCNoTt2UK``@ILWjze0}&CtMNr z%41=zR0=hQ9*qj9WN{iubpW>bySO;ClfNR-{S@4c=9Gl(QC4wlWKr{0U_N3UspXyE z)$=fF^O0;Hh(zUtU0!oeq-h zdmU<^r>h`Wn?SDHxHb@~fP4F0n`A?fv+{likSY>6c3Xz#-`sPqn(k3PZL#c_A!TOY z5|>P)WgPdp7ILn6+=Qgo>7-gdF(DS2r6<^mP{<6+pRKgrEQRB15hNy_scZ@l^N^T;qBBEUly*jJfQjsHZ?uxF6Dc6u zF%K7n)yF|6nPPTRB!LP?_C0tk0!_2iK5;Ug90>bLDWTheRf&B4%z1}Y58X5bn>XSo_O#!3-yUZ?#Z|jy;m`YTdZ9Eag>^5Re%m`X`)8$KU^0o(I zvdAdqzmlC zl@^LQMUfL!II?2uxjnK2+z!qVSIhLm)<84AP1*y(8{1r)c{=sAg^l!z)1u0t76tAm zR0fqV>5^x0c1W6qjr0X(l{;#5;6vHmH5@G$^`v0e5{iqjRo8l!(>rGIQ9U;&>@&}L_>g+fW+B#EOn^{eo$Haw>>04c)Wj3n=U!zGzmb4n zU73oOr5#UIX0$A^Y*jZTjHQxoJiX_lsSb#j-wnO%lE@i+En;a5uW|X1OT^OAda%u} z;sv~{RiIVEnQO!W7v?Qetn{!BJSEgF(kgmDwb0r#n$=Ll9vIJjBIx0NA6sCsMCJ3| zT}N_W7)u0YK*l*ZhbiXo6e*>`?@H>qeY}46GxQ*?P5lSxdTt%RkIn@F-S_Uvc9H0% zv7wptIbIj!4#b9*l6W~*kLlz`LPkNw@_SIJDn7Uf+G!O{P;9D6ae~pQ3%Hq_LQW&S zSCzk9Q}}8$6hazuTPLsZ9Z3UGgFE zRy95gyIQhGe)^VY=q*;P@M&fASPqZRX<2;o+kgN3aIqbS`OJii&(6-Lhke&9t@m8V zyA0!65(*_bUAiE7ppnkxR?(kIFVEk>?E?gSOzXn70%50Bo%Wqe%ds$Zv&$`I^5YiP zvL87%xCHF6w0AIyzRiN{uSJWMHiCHpWV0^Ynlj&80Y!N=%OJm}GY-qo{ z9hjl$^0doiA^W~nfko<3EwebZF(8@~BRb%0`Z1qKKLld5J}FLG#YWOd4|uhhS+Iz?4bne$4NTF1BYp}c4w>fE9pE8F+b z;(zdH!>No0yH8@*n%!vFZ6NLU85yE-EW+!Krpq|u( z0iskb%lr8o7i$Wg3!SZCQ&|{F3w+zLCR4Yce-_oh#Wz@(SI_M{LN?m5-@D()W+YO~ zPKv});ras5hQhwhvf|O5rVVm16R0lLCEx0Ng6wnEDyDIAWINnbA4to3U^g$`Tlj^@ z=RGB5Y`b~|R_S_9sfi7J0NM`8qIzB}RPW8Cb&89yUe@&S5}1c` zVR21YliN-mP* znSE$>vZxyBCRM+*1&q)X><3#QjadfEep?2biCVD&(u{$1D(5){EA#D~U;UQsX6MJ) zaq#bukxj~_7@(AAVA*k0(QK2Rlhr`xQ@Us`eNb{(2C(m{b%4I_?$0b{RKJ&D8mfGik<(3Zf z=l5jY7OZRq6)14}&GSC3L^fC0O7KPD7Wf|*%oWZ$2Cwc>=L_%7)iFmrweTNa0Rp(o4UYptJkF%E z1-7PvPXQ^Dhk*r8vO>xXPQi(PF{Mn|EmF)-=+Zzvw}0UwS+A>3z10Vg<48T^n{5_W zhMg9k7M-Ef1(jjV!jI-aejD7~%H0I|nWe(b!fnpQbcPD=Z*xxeSj`!R_LLAx$nb8&$}nHqLZJlTF2Yt zffd6sND|3Z2-8%U?Wk307i{IGxInq!c@?S#lI_#RfLs+MA?8Z<_za4Z{c%0rBtB}> z_0ALwjSUlNS&EoZtFP?XX1Wr3LH38ttee^;n2Q9wlj_AN4qqk2ogL&k*?5O=f*0LV z2u<*)O)q_I)v>dtTJUx|`Y#sqFRNqtq%AnbRsL-339l zWO_3*zFiYJ{z|Xx1BFh~NWZuB>pC@_VxqwH#aQh1f8Tk3>({NmIqa4=K`2gjTOZIZxg`2skNS5t->?00*>e5WLectwTF>8EP-fU#z%oLa1t*^eM0UW6x#;Th zy!$d~DyZrjEyD6rTE$_=9?Ey`^G%c-gDu5)dAnql*Kt=3s)bYqr3B{-GwGyYR2Pa3 zjR0xjczI9AS$HgEmenoKlYyEVRG#!>?ITuDnfcJ`zke2Lut7gKn3F<^pR;6ZAasu- zJsqK#8j4gRfgw_WJ3q+5Jwc*!R zR5ST&TXd_DH2yfG+7$WIpp7K zJ85mUgJL>q8-0^jJYhE|cAG}#zGd-lM}zaB?tWWCR@rfgF4@S@h^3flifp37_0FZH zS(Zuf@=sc-=>e4De)Kd>ECyg%2v#e?(6X-Bf%Q>p#ePo%Cf+(dqm3lmF(yEnWt{Vw zO)*eZJdFw;;+Bg0=yRlx_lSq;sHpxFAvg>f=$m|Q1q?e>2Ax_k?9eSc&?`l` zI^@aq(aFBo!eYZ(rOCei{0!UKwW)3^TQ=~MX~haG)Wr>N{Lysv@`CIYnpxK%nKsTR zewL<5k-{tUJV8>uF|PyVFQ?t&SZHu;#k)2OuV1%Ep82D}gsk5Xy@d4Iu?g8^B)R;E zVumPkp9+r_fdY?DldMX3^QE;e)nxP1HM8%?w}8`A$Z2Pw2wF1zfda*PP~8%BzQJuu zS7rI>PcQ}9{ZgEFI=zKS2}U8YEKZTfF3BUmjj}EdmT1x{ z)<|1u&2{b`4{WSe;OCwZ?&D&x1U=$rX(OGg>hpl@0}R`C@(Vq53N2(IwMju!0cx) z)5Y48bK5|M6=0}$BZhXHDzVsYoW+bp49*Pgl2I|{+fMQFHNnxI1Fu{Pi*ZV(i_CO0umHP1&Tv@Y`%n=0!(a!4@0l4U^1gQRQ*@eLI8FP6EmX!hn0m~ z4Qg{~dgbROiT*jSXORDk1@sFt)qLa{zuk0E^jy`3FthZ*eH@J3=7$V`&S~{xtzt|T z!$uRyGhTBGjmI;3u+1&=#^cW&oa!Kh1F|8$?*PfOV+Z7fkpogmF_5%=kP62+;v6SI zv1)0t_Zh#B5zYDVdndnl?X5FEj{E5vs4E)-t07!)#C~b-kCNWdDq`l{ z zm0)EdR-U*YpL=8-?3l;O1p7YJ+5h>k@3|T9^Ml$w=g9UKhGbA^gtmPY13c|L;O=@y zx@JPvyl$w8l{~9kwnnNc0#L0Fh~y>H4z^n=&8L$g z7>C*Fx)qqWnta(AH!K7@=w2*q_Abk{40}Yo2z=COwljUDa*Nc)ra}dbp1XRwQ+8aNq0kQ}a(;l{GnJ6~XckSh~rkHrH);VT& zlu=1HTX5T{%rnNRT?hg^K)44c)BnfGW2~FE9r8tQv@bIFAHV!?aULnOWB=o{k^fOk zF_1WRm*fUju1?<5~M8I<(H(ixd=nsnDp8Y{i7=M>Y3v+LYZ z^~foq%rd+3*ri}~@ObbiabO-V6RdP*$l~3vJ@56QkIcCa@sVWz^AgQLCD01|w=h#0 z*-3oqqmNjEh{^UnmFHL?V&=*^GZ)oH3e~>M5p{B2|!c!A})!=9Pu* z;J)4K+2Hl3T3)fbMe$!xF1JFKQM~-LXET4?eK10h!CE42U93|#iu!20RqL7M(=FTK z@`%@}PW*1sxA(rb_0`7L{}lh@yUYJ{<7eH=+tlZjCG%sJl!et0Y!Uav^WW)%t5blV zWiVnjA&qPN1Qu+M_<1hZp7(9xbSS-nwKj@8M)j(-J_(-Sf`BNGrdYW>C|eNA)%qmz zi)HCF-qIwxA6IF8*3D^k>+?;Sn@MlvX!@aW9kxPw>#Cp_w^hMY7;1AvU>yjU>wWUq zSxrSV>Po&kV!F9wX1_$w5yJ1b+WTTqBYg>CRaJBxiTyudaaj1}PySn+Tl?;B2PQf{|;Tm zsSe2C9(r@E?K!~#LC2dTZ0ZUx_X+iTN1trrw9(Nb{U>%B?v;zq!)pCD{jo>C zER|dd1|`%tJNeZfDZDtJ?f$(BGeAFC{{*^52ROGuNGd)!N4c2+2}5-TDOfs4;<$q( zI!ON+mClFWM3%2k4Kzw&6MaB=l3%0L|8>2>5iv>}W`OKF%$V%pInIBv>EVnS-~5*| zSDyEh4G96w{EmQHMRf319~44I>T8?lb(alGk?qw zuba0<6GU66x09=HOl4o$ju%p9_LU<7>wtC89fI$DC&}}R^XViJbFlXJ4e_A67S6Pa z-M(#}gA4myqD5o5BY4mD;9`fV$9BW|ATsH3M^DvUrC}bMm;AR_lJUaKWA#R^PZ`A& zQ{(^@euVx+oaNi@hed)~gVw5bp*TaGQ||WBg`OfEoL1`goO)gY)P}!Gj;OS%EpD;A z80yhi8dN$7YA54dmf|!p%IXIV%di6;yCvs1Czjm-QvXp&^AtClO@_ysbI(&S=08( z-k`9Tzt*vdKI0b6IRmqh?M;MXIM`Z>DUBkFhuO_Ox!?QOL;qOy>bjx!Db}TfGC*{0 z=nxmxP~zn%DF`H$h^1-1?XdBLr;v}MuNl0`t^=x6HgCcp2bfE}JU6K!Z z*IjT1YE$2%PbyaVe@a7Nw$MAC96%oXs)mC}T7zdiW*#6Dt$QMKQpL~vq8QclHPznt z7Dfu7?lz{Mc7rfrq5P@}*auys1?R33DCth$ahZk*WOQg)gX0+kYRy}$z%k=T3x3rU zV6Z4fy@in^(~%rCdT|d@ObJB_sqnk65!;%z?6=BQ%RNuTDQ z2Qf6*jfoYl2~72dC=;+)V?+1&CesmucB$r=-+5stok?SB9ivH7tO5~GB)$TbL|>=* zkm{kErpFm7)>ZZGioR?>&i$`^^%tbhj_0B)Mvy*3 zF~ErW0JYoV<)Z`yOa<(c=W%;MZMIw9Nb1O*un`B;rA6UdtbbWddQ?!4Zsp2!5OX@g z6vC}l)I0yXm3xBJDO;qF`T=i6I-M;#O`4^V)H|!Ftz1-R8CcLgw*w}*WV&)L)XW|0 zvS(q76q4foPLbMW=mVV%(&)>T*K#`f_?bQQX#G2U6)bueY=riS-+{5K7Q74W(Ehir zU;c)v`Z^o_iyg=4fO0p^9XdcU;4NoU;aBCK0arJg^NIgwp7p#u(nXN#U#^Os6T>Sd z#{_qMYaz;iUzY5f$j79iR`qI57w~(>&OvHhaoV#@ty7PL?BwJw-%DS%f^_W!18WU{ zP0;O>FAX$Qe93f2k_}|+SaH(KUL^)wD9&2QmXPR9~@ z{HR$DjPXy|{`ie|uYLWih6$?SR~@%V{RR)z2N~HHGbVyAFZ9Bd=W+N0A~8#q&V<=OA|w z^3G!YqJvJ&J~bPvEG~(V?OM|2j{6I{mOy^BEZMKb_qeYHEm|`u5JiASd!zqq|7;P; zE=C8%yOeo$^7~~)kbH})ze?wPc|LTM6s!ZOGflyg)1rNTQJx!?4$D5I@6rRpQn98e ztWMrB=R+vxmk@9qwkOsHuL(XHS|sm;QquaKk`y`0jp}0I4e1{Dcw9~1bB>qSDXWBA zeOf_BVn0+M(Acd&Q0*szBU6#!wqp;;j0ATkeHxexm%;|g(SX%~T16!JjQioM@4TNL ze2~-NmFJe~U9sqH=vv;*`IWo|!3Da~Iho$=rc;+W@8?u3YNQi6jiPg)y?s^MB|Ar7 zc01!%ro14~sSi69c_g{V@{ptP(JS><3?NzAbg?v0=9}MobD3hk!F0WPHET1;o+!X< z#|UXMn)eP-%t4BjP~lme6P#_NRD9dHMbQYm4uc@{)k1F2bpgf7W@#_I&apkDN3|`q zBVfHFa)@A7)& z$94)cKN41E&^+UVSl%+fT>I!RwX3I6EzOSSJ;=fzw=~;EFIze6O6vSZZjGeS)~#cZRM6 zaVnjQ0-J4-Rq)Iqvlti8?8W*w+2P_q%ZVkX<%jH6V$4ikq#}$=UL%n=C8KLe_|Ty zvs-+a3ErVli(Xn86~=BUW8JHiiH!d0+1bT%t4|rv9tF#R@$7!AeU}w5W)#K-3n;^c z^2qhkDKhLpfO|4-wvVQmO%z$LC!>@tIVHX*ghIHu?K3U!-&0?YCExo~4^6`bcHg_* zZcbixZ74G>D`cj6&!{Qxq08weqN#MQ4j2$5)6tN?2B(+(x)e9vP{cs1n7;SY5@?Jw z)6)P7{YS8ZhW+lU92h=(%{0Z{Om2FGs*fa#a6cF%`~0^AciUcWZQ{|i1QHWJY1^wW z2qdCh8(uc0Zku69w{m*tN0WLoNKU#wlvjK7(s#gGfDok9-Oz4X5i9}Lsz2}^c8F$9 z*h+Q!=^?cQ9^)*=wCOFef`|Qv9bf)kN=lf)yV#j2OCe>Bq}9l~XrP!Q6se)YQ-UuF zw*>b92@3g9m2(j$BM!J%f^d$(8-%r{v9l|e|9)+&%C+a>P_yeJ;! zM5l;wmw%0P!~x4N6o;M+?Bw6%Z4XL^zDFF8UDW~Q77nZGlxUw&>bfGJB4CQVS++nL z530{Qej2uxUtt9#`wcYy@NjE@X-fSILRB!M3{bi3!NNgcKNWghbk@ZAbSXONn{Ghm zjG*3G|N8W~RHI0<9Nd-z(;6-P-uw1?#$aAt9t>O{JMB0cR%>GARa@xv5n6CGjo( z6_1FeUGn$txKuCR9X#T2OIa6?B=4S!#_!rfo#MEx+FOH6x{Q#lkRr~ALpPlfj3qK} zJ5?yKpspg8hE^p(Y>oI;pNo9GJv z1obgqrt>~=oX-aSi1V;RY~XsqjzH9w()X9e+3j@3ZM*+Ipca%W*91Pa9&2uDV0#*X z6GMmn)zp@^mjAl}J~JQv(3f=Ev7uRQ#6`JBF?|&IlnNgN-}?l4@Wm*{Xy%5zlMhJ% zVV6W}9h2#1nVuxj$&Yd@ru%rvC&9GVBKi?`t>dsmlw+rJswj6(v#g6Wgy>}YYuM`E z!XjFKmg89%W_-V5Ntv^8`#ZcYqLV%8_0s%$ZnAQV>-{i&A0N-|<>dIU4Q`f=UD+z% zGgp7FU%qwR=R-+8_}fm9fmf>dLy~6MYJbxi(b`CQ8e0}#F##JZvqil(=e>_igE4j+ zwlxzaI;_eI0e$7iQGImE`$&aql(n6vs^x59dcg9S4xq{RX2nyM(6HhsKf5bd(3n}E zK9uijFiF3>NLG+oJ2pv>zcwzSoJuheHcFzxqd;#Zoz@_Sza8jqST1B#N;fvNiay}n zqly&7^Y*AAEk>*Q&|_E@8xL&6Wa$YxsW7#$Sl z_~>gHudQ{&LLxP)MqvIw6knC!;dIm8;4xhhM+Y_Y4~1O`j(+Kc+g+GwHpuG&4sx;- z<!5!``_Yq+6JEbEay;CX2eq@H2NXMc{q)eB>w3JL}-G zJYT4jUstbljFTQzma8&Vz+3USE2)+glU=@39XYmNu*DL6>3oah0H>8HqTeRZQf9-Vw3Y=>f=nI>L-cr=W-i&L(I+3PAq zY-%Je(n9|X8q?Y|6#y4YH=QzTWu@~OI)#UU$|iA(6rVW{<}3*cu_e>jJTzJ25y-KNfJqU`&1TY> z^J@5wz$Q-~eLP#V&iTA>h)!Eps~Sz8gErYB{3DQW-c9Q>>0(2h=@l4pt(=R0*@~4d z(sJ=lInv1Omdfq;J=fR;2H=&?dv_hlaU{o#W`x5O1BJg!sqo669vP_J3UQ@-MT{Yp zWIEj=2MYIU6~{G1}doa%dyDTan&lXP4a_~sf|-oE|p7J z&(v8mS}c%ZNq1Ij!}HnTYbp=&f^15G*%)U>5-DaUMdCo~*Q-#z)#v>8YrodIe51N3 z@Qll-=x-m;`;mdC!Pw44VJ~dlP8&j4fMax4wA$!@PDd-+e({;7!EC&BdPW;bWM?+) zc#;9fY8*JSDJGL5X;gTZZ1<8px4m=&{p6@so&09+FYkp%{04ofczG&$D8A^~>z^vh zRNj?r3*GU}x;Hn!_R;bh;0d8B$_|&-#WsWvTd>&3k9rC^Nei|#{9|PZWztCI=npMl zlBVjCW591cb(!C`#Fkr`PkrtB3l@UH%+t2Jd~@aC8B@C53o^Z!VGSk)Uk$o0-l8s8 z+8Y8gNtJWgNVTpuGzVtK<*decE<6N=_e{Wx8^Cx`u!qPaS5~CvBdoq9K>d5KGDA zJ`|r1>6T!Ln%zFd%8r26kOpo&SEs(oYj9f)@(jZcr{BPzWurgyM}KYud4>k^F=!~* z4lNd8!g#_Me`YgKm|v9t^nX1{ObhM3AO-|PsK&V=sGp2{2GjUvX&IcU8~kU^_gD$JjY;KX!7-@7T_) zj31&&fMERNHaj+NVzPG@{d}HbmGhVPx+=*2=S&x&!3Z9Irx>8=6;a^{0e3yOc{kBj zL0f%310faUNY&7lLAjyB4z;`%w+KNE>?uNU-la0=wsRjZH&my_l*AaPKKi`d1rTDr zDBj_9YW6|VWqudv2R-2A2ofdfVOm9mVB`E!f_czZ#ndFAGtN#JA6gSf=AkqD(%~O1 zGeD>H%{A$y+>x9!8jWKV1H*HK3NKiGO^(cPl+)2~UBJFeqW}Mvc%_nUu>U&ZkPj`^ z@jwidGeB-b@JioywAEkrtwtQmVM@>{a+JA}EO9iiT@oP}aY*!s+xm+eYSJN)3@Q;H6cat3BNsd4~MaXY||3}UL z{Omt}@!!7_ucesP6p0ucR-@mO`-I6o#R?hr8{9BmSYfzRuywf#3R34TzX!a| zj(}`#tGZO&rpDqmQEnHw`@~1N3CenIiLBo9v>?-YGxGqlu&Q9)RpNTrW$bFKb*Cp| z3{K4&8uw+*v)>ka9db8JN$m^Hy-nimm{V6|G*_lk3?!0h!E+&9vWuRZd6`OFQX8M? z6`bN>)e<4t%)28#`2W~@6S$_%bbs6>D$6|rcynLC}%?d<#f z=QnrCEWP*6OlN*+XWCg1!5u^e7tnyR1P~WMw1|YowOVBn6~&ERYq1K7hywrTNuni@ zXiiAD(ci6~%E`$&Z}7a|Jn!>9%lFI65Y_>28*Y|R{?+b=&)SK%Rr3T$y!OHksx5(bM@m7R+;vzJ-Kam++`CwlZJ1b*HLT|MG~lJ zDANUpSXa1&MM6cMH=z>0-!+|nFm|d&k7mkJWAfb3tiC8kHF)A_g-AbSfDFKym{Fxu2|5;3@@bL9om)mS5>7^^l7}x>pEWubH!aN#A z*kMtBxcX*(WeX}ZzMv4&z3DY6nYf1!{J+Ls?yh^VT3R!9n&O@U($2}=n?hr?)0 zjyYM1@p0GZ>V?-e@nYY@v86tfTNWRZX zX5i-~zq+>IgJde@?YM;l3s3%}`t>slwW3z~;mf_V*YIng#2=ep4?$7(11Mk8U-jTy zUi`&*O8(Z#^pqX`}B-Uq&Yj(f<=48lVD+_qHylbTK6BfIC61hm&F4+EBQ&=Nt zy`GUS0pW%$P>G3k{}_`%jZxd=%RxT6?AQN#w`1-NLB9Ymf;w#wg?(48DB$&o2VQNA z%BAnRu5w<(V&_l~Z!5?g?c!Y)mh)o?$WcS;9IuFSDvyVe&}pg-UghzbXpu7#Reh-YZ?qQ0k}tkhNU~($QC?I2myPYF z1l(NKJ_pVjnW5``@Y)qvk}`CJQ&^63+H;4y#^Hef8I7Y9dx#=cRP<+Wo?ZA6sJyBbDZH}+NF%x&rSaT( zQS;P9=9csNX?PxJJp;<8);ecGsdugPmh+K-MBp3iku{Sq=^G)KvmWrmMF)REaTeT> z$xD?qPu<6h1;MrQxvTg+GH8%*p4#WC5gznP2cePAU%Jf)Qc%xp()}FPrYGjpawLts z`aJ#Y6CY>s3C(hs#irE@W{L!G4}z+f+4Mk2A&8_R2LJ?TKVzChu;DWup7cf0MNtE6 zUG_;aIMED@sh^8d-IqZjoW0Gs)z8=BCuQA*oSXs2t(EIu+O^_&mkL2jU@W)IRu)LM z(}QzRd~i_qi8G!@sl*sKb70DT6m%{S+}I0KjSCs=m_C*Y2kM!u~zO*V1y zR2R(BZeJ`S=2(k_b*5BoRuHET%gOjpjI&+&; za&uA~*n7Qf`4kJRiiUzWI4go66U3EX@CL(c6c={^dJ2 z7JShE&cK`9?{qJK*0i9GsMBv%y_NC$!#B@FHAqYR9^GC|ZiK)9(roZqjzwTFX27qT z9{Tr7(5luLRp+s1`ex>g^h;VBu!8L);4aT<^(=)=bvL;R@3*{N1=qEDuHp~MtW{C8 z6jD$3N}lnq70BAK?@r4M1EieXl>12b3j--2v>&!vE}>ZP!gf*7N1%DJh^dry(<{9< zdTk2qm1{-EB2d70o9~7hjZs(RMa;1Xt>_$sWx0b+m9lDn$J|}~+pZmRkAP^e{<>ZK zV{{P%@51ZHW5`xbvh^ZoGu&8#ji=r6U;kB=8fd^xQCDFs(M%wRjUMHltkFKJq6N7SX)ZogXpgU%?8&sMiQ9-jH^{~UQWjoSImidWAp)RPqG-_zfx z3D=j$A~8W@6oy?>LN(#V(EPLByVK3K58Y;uoU*M>1tWhyOlCH2 zX^Q^H)3pBYhA6G*9_av8-pdT0MWE*SVHkgO`f%601MwYLQVw)*)8F+eTK6KqKEb?5+ zoFdr3s8%F0vBC@}gBWs3WD1x#_hZx5ipB0-A^Lu|JK@@3wc-}t0?o|1P}!X-NvGFD z9CFz*@h)^tV}$oJ*kLW2-0!Z!qP{jD0jdYiRQ#ArqOH@0oSG#1?HJMus*u$~ttbm9 zpWh^@qWeTcPV1*_nvA*8bl*hgj-bg?Lzj4fFw~4VzvA!|d9AdPe_zrKE!V5q-5_u9 z44_&IvuF4)7QFp*RC0pZc)#gBt)}Z3mu;5g1{lmO#|W}7LbUFQZW_8hOob}HEdtQI znAYH~R%C{*@*0$#0JQ2gFfv1nWqlGzBNQ-IOcAhRUJ9-SqB^Xg8FFfH$0WO!w=@*1 za8wP_Qvvk>#bH%G&&1BtXVCav+v3c4?9sf+3yH5cZOP|?90%T@n`ze9lY^^T1Q&SS zu2@m5kM7HRdM)Ty2?&S;*qr2_*Vt2%(yUvF7>*&js z|0HqTHdYR7Shg9hcF7dGh9Zenbh2CNyjub&@mud+OW*Uyw(aY)RkxgP(_PT2GFEsn z%&6tCupb0Q8u^DSV{V&G>-a*=Qw3Y?o#bZ+KU=<$BO`;9Y zo5#Zl%}&(oZycLu8m4(c7AiAAom_XU&s_3a(z)R4%^i6+ILm`o@PJB{F}-YEHb+t{#q6pc;dvYx?*9`GZ#s>($S%*Qb&F4jiUw zG+LXEQ|wWS9KyA!olbPS5d@KarG{xENVI;Abckxa)b9Q6gH9FGGTo2Sbu={B^2%NC zC<{vM2D~16wJ{jD$rq{>C_GW_f@K^kwfmAeIvQr=w;S-v_EqJCCa}3oBD0lT4L#+( z9;#n7bT5gMmQT*63wYi1yN6{^^NBZGyK^25JS<#mtg=L$Y|i*#p64h3Z2;4&I}Ur3 z&!02rXPHrJR6oVuqsSd98tZ(@T~IY7oz}RbW=MNjvhVhQQUwYk-w`FSswC3plkAHn zoS(RYv|sQIes-{GN$^flV(=n9*6nr7J&Qsn*QCuL8;}Bs?UCzsH1zF)4})D)Nh$KzL@DRiDlblM|{>pT{ntaT@fE559K*Z z6oEu+b&Ft#IDNiLhRxE#2IbaYWaY;9pvGWbT!|_{-7QXG!6hC zew*EoKlt7`AJe*WhfQjlsT9YxsK{R%p=zhIl>7bbMNo{@z+i72a)95{+jYA?$;ThS zVXAGLXMg+~jC{b!mW_YqTesp&vjh$c3^Q?mJ$uY+ zv7>u*l-&K1$vj~8UhJ<{oc8RPb$vFvkjV0?$`{^(sLV$n{;z2DPKc(LlKBdUBokrgm+YgXH9LZ&G)4J2BYoMVYx?UEqly6fe!e*=0vs0E4 zx?J8OIxAB30H@j|a!;N(eXA=(7QyX>SRsV}pb|eusAX;g!taq{NTa_X{r_*sLn?g? z*c)I^EGKFQ#@+#=t@}=jg=|AE6}@|AtrP*Iqe9D4)_Jr`dU##Z6J$r&GOAHg=m|s; zKCruIkPG6W

R_bkUDQA6q>!HPs~7C#aTD_~$+N&-N}2S7SAN&ob=T=5~vdu^K1+ z#q^7&P7W9M*Kr+t&`la%E4rRy*HI*iif;cJ)(2MxZsoPhs`&kaTET!H*aFd==~awY zv|)zrk@D0jwj4%J9meKw4qbYA&~#t)B+a4&PrsSnPa7F>ThBWqJtx~9(Lz@-n5aC@ zuJ*mn-yN~puX(rvB4Z*YKkVX79 z=RKl>Qw~nSP#Gu^x6?_^O(GQA8gjyoR+A)tMxVsmaPLUqcdCwjy6`93Z+V8h>fpi$`X`(b@kXCnMX;%Z|_Ka4TvZh`q5sJ;lRd1 zYc%`2DHd42^Qq{ElIp;CHZiy-q~G=hC<2okTwpp0f*lEP;o?uQdC^9-uJ~j}fc`_4;nN z0n=Fh3liTj6WQ;P4|t`z?&Dn(mxxzHT@kbcw5|wthN0rxQMy{bW^$YBpwo(|doN!R z*abKvO*&UFjWmWm&fR8b%pFZdzXPv6W-9u#l-l5u04VNglRuy_rjIg=>2yx0mX|Xc z>K?0R$6{DmHM;dT-QD(bh`_MWEIha2AX&k!(ZhjZk!!RprBdt$3i!Kq^>V2AcV`Y{ zi0cHkGF`PBCS3+X?uNwANaCxoF|18Kc8;vY;CQtBSU=yK?))^VW(#WohwFph3uMCt z0@9YlTs;lN!me;D72Uw3$Qu>Mpui%{?H)my<@2H=ytvR7!9HGMa3ZfKB$rMiaiRI& zsQ!^Qpc_~!u$2?f#7U3R@8$?o7`saSaDK#(9yR{t^w79n%x#S)jv8GWlKwbkwgD;c zmq%5R0&Zr+furlkjR1duVnKdR57*E z0cgOFWlt);05T7diVm@v^D?S$!SHW{;g}oN&0oFzjh|=A4Dd;x@?U=>Dcs=Wz^hQP z5qz>K7QB^oDjK8cx*DasvMzZNe;-6)QG&Kdw%bLuGpt>Jd3#-f>s{AH!en1V(B(6r z{BM(=m#mmvKHp}b7-3pQg_9A+v(>5n-+XhN+<=n}YXZ-c^$xr(2L6~~D`gJFW>O@B zimsL2=En#R%-ieIC0_$Z1LJ1xf=-!QQN?sBmq*<9I>gWLSO(%eCtv&P-&KnjKK#~~3kIF=^q{ny ze|h1J1*KFP_1BM8g}S$FYCNom%#%yX$dhl~S#FS-bi^ax6f!SJ{KQOvuH5yGWEHy7C*ys^v7L?R+CI_Zi@qZy;Vkw zS24xzrpOK|I>Yl0)G?MW&;oH>m#ExjQ(IG|RV6P6d&7t$Ou=!A+{gHHP*ug@-b z*(JLdk-)|TZ6f8tg>KhK$DB@Os-WDZ%e_h7Cc8#9PKxtR1(N0?VU04|w7F$c!RUL$PeQY;i8Z9~!MJl>UAt9>C49F*q!(5qc=jP4Q* z!7d4%ophfQq>mIa7~?<08{l6jxj;a8kRFunk@bnoA`bBi!aMA%dNdk9Jk<`2@=f#W zluLH6#?=6o(l_6#Co8!b7zb|I$uj~`8pUp;V1A;{2DeMvbTD5+Dd_sbCXi8n@^^{yn?v3}P33;7<|A+JQH%4g82m*glrTzhy2=GcD8 zFnRP`C;pK;Ych<@Uz#@73vKY(o^i}27uAZWBG7n=nR91qu@4eypz;F{2TqBTs(`dv zeu&`p*-E^AQvhBHJ*Kh3&X7cAsqlszn3$H!)8xo-@Tv2Tu$Vdc2{=gk)D~bUz`<;h zRJ0gxkJcm0evTTB?brV8;WPt+;@ za|D*kEkUO}>zsFcZdG2Kg*^Q2vhv_O-j<-te#aCU{A@QQ7{&L>0_vQ*ycV$;)06o6 zJJ*s}=*`;_wA)TC88$%7sL?Uf08_l~{>fDA&WuW2Fls8Q-tL7Q3iIcxmEQxnuCk0YF!&aCS zid{>Q)l_u3ON(N6K=Vwb%t7*r0!jm9ZcUO);zr)sk<31VnZZd`kdkbM$PsX9f9gdCcEhJwI-UDqQeEH6#jSsShd_oQXoVr_u%2NS@zL z(V)|fpe8mCnml%R^pM5EQYyoC2B=1>}#|TY|rQ2oxO$!>}rBoqNS^$O1j@;DbT5*05jT~ipjOXW8pVhHS43C zQA=*;WM`-aA8lXv|5lxS29pum74HhT;Psui1^cZPMzt0IcS zu{y0|?vPWTYhCb0|8A1*+u@q*d&y;2WD~0@^6&JD5q8Q{OMDN23T(Dmb&@3ecFK~R zhn)8D+T`aZ6;D%b^e&VYd842%R=cToyL2kgdLQ;y?SCm_)+OF<5aDP5hIQbGoOjE) zMRZsvFWbm!=zQNaMN!Zd-+fBe;lLF6c405c^gxEmeZb$}?Oz8Ms%q#Oh_+~E=F)?K zML{POhh`5sT_kPtG=&DVf3JD&M&@?6bl=s!H>H}H$-X*0TC0qc3C8U4l9O>7cXP?i z-0vIE_v^yn^pm;?90ujNBP^#94)6$?-Ev0#qo2 zik+UosR{36R$!Mb38Gm@q^F7x9P%qK0T;P56tLR2a3B}*5MUY6Y0oS=& zfswiT;3gJPm*IB%nd-s+gH!ob3&LqC9XEUM{+=Jlmw@%G+vjd~j2r#h*v?B~-^HZEaJ1$D|?f0h} z>`(8{rAx>*2VVUujSOQE#qOd=0TtZ>bt?m2WdZTtTG3YC@`&8f9(g?|`GR(tb9F%S zyj5QPJT+Sng$Y2eB&qQN$$5DrbM&PnVHf7Bm0t+DXO~ZXbY@Ui#jJ6Q3tB2Hv`6c+ zHK2KNk+L~v#&H4-RTH=H{PS*5nj{C5-DD|WE2;)h3H0l*p)Ace#(QzlDeqxx>6rP) z<2Yi&_8>4bA}4H&cdh=-MAK-WnPBT@qOGpo^pdY5QQ}o`0o}_a`yTwkE^;qmzkiL_ z*gbgtzuV%+Z{{)C>oljw@3?;A4ww{z6x27A+;#e!bWFPU=z@V#suYy z_wo9m{keRq{gPxBWB$oCNPosoppU!tjr*p(?+)v+n(1a=>a~MZi_4WmbZbC{Z(~%P zM@~SdQY$D9|9s43#~x3Qr4#-5hA}$eX55+{!{Q?UB(gnVXIM$Vg_&z6?eXc9ZId)f zz7(yLW5Eg<2|QaJxR|Mv-k)>ITjNs;`eyZXrhAKfC9uw7OKOTdgWm@^%ThWfvePRK z;wKlTsTG&~(iC+}k1WAA+4r_<1C#HDqb0Bzp%w0R%mYyz(ApOJAR`Ct-uS$r@fiFU{>vco*<6R0r;pubT+{Z@fHa{j^%axqxKeYG1XYgzWKIEycd@3m{;o z|4qT_=>i5AuTJJIougJ}@Lvoj%{VeUKw}x>EE;fDP$;M;ynsRpW-P5+ zMCru8F+5{@iE5qKdzqn`;x-VX(zxWn<&9o(QhSAFEM4Es`7w6r8q#6H{Q7nwaN+-b z-ZPhI7Bcu9yqL(uU=)4YMZnW74&O5!qRVI=usXek-aiFif^;U5qtlplvY$Wu*lV#C z503}1(}8(BjL--$?iYvOs{Dq*-wBML^+R%m+n&jRO--B8p6M*bo~B4G75&kmPA{y_ zRZm)cjsAGhsVBU2UO{+2uN^#^EM;QwI!GV=ebA{)xg$)U^0>eol&La8D56A949%OVXS`z}Y~||Kb*9v&&(%t0 zhVHb77w?5T=3S(mUmT7)c&gM%XUYC4@m@HQ;M=-hA3J6_m<&z;uyFEhahZ8L^o`jc zXv7AaBm39u4P^BTV{?Gza#+o0I>m0L$R;WpnbM(yO1L9TwaO*O|8v(f^NPc3$mg!G zY^`$9&}Y4OPu8i)93p$Y+9l@#a-7Gc%Zd|r9@_<5pyTOdtT~g{tuP=Y>clS(kgX29 z3$HMm^FoRRjm7O$bO-N>ID_9VKT4_t>R^e$9ub`*TOH6v*GY?K#el^>CO!<>sl6|? zPlcigAqMO^=EevC6%9(5eTI)0D9I2CbNt2bQ#NDp;fQ5A^cnpU}2Tz|;E8aD4 zK}PUo#{acyR_ycMz?QAd^MgDQmi<$`6>a z28A-k9eSs%fGnNx)qP8%`oafgYa+HsESrEE@W!aC;!@=y%q0#;10Nih0e*I6TJ=9!cbPl|AGSdpaGpsi} zR+uM-^{IMl5?>p<*Tr7ZmZ#2xMTmIn2=;mNY~?AJHl&>$^aed<&miQvB51*80S&w*%+0_%MM-7Ym=3`?2S0a zoM7^NRx%r0^1%BkowqhJkLjMi#C;_c6JL(#BfEGRU?DEd-x5$M>u@c1*&lXG7C#3` zyz`hY>6Kvoj22OH$X?l!(Z_@>f{_*GWVGO_NIQMyTLu`tda&mw?EQ{-9Hp*IEK+yf=Cc zI&Jj25H#p?K#(rRW4!X381K60y_Tf=BM4UecFgUj*L#J}B?fPH&WEyyEOHkjwwu@(VY{Fz zVhz}d0y;4`pBKj*qwDDEz$@Mdm3Bt~hlw{b2sWBuv%_$+9!W0l(@Z;AxL6Md&P$u= zdr4)mY$Az|Wf58)urfoA8sfeX>IJEQfA;fIt0XmicR%W7Y)EDtq(*Yds>+?S4VcD71cVoc^1+=(^b91VQVOM zy}2N9G}@aEyqhR>cwbhFw^?E^QUiQCsSLHj1&b&x-74=RNpAaQ z*~K}w3`R!2@DnfW0V(q^%2<5h?-D~?Ms!>B2a?1syyw8qFQhDo<*hR)79zl@RP<5L zbKz=5qXM`TbD_xYi@^4&?O{iO)~ucW5J+nLH9nBjCT%2_Ug}aDj>Rt(bJfZZ?SdDh zxwaBKMq}9HMWeRvnrRa91=$~&;oMAgYnQZ#p|^s)-+FcDetGJcmcSTh(lSukW7?lS zrkC1p%8v|oJwFF@+J>B( z3A9TC1Hg_jD38PbqjY*1oX%3-m&8YErqsa=H>EfD>%)hf@aB(}0B3-P#uKWou08T> z*dBi>iwV6x`zW|F7XvrU!1jKqv#XOfC_qmdOM11zJNb2jAzdwA2LBMz(0H8+6O8y` z)AUXd7stv_)v_5Iy?4og8XODz4w34qP$#OtV5!IU(EZRCW6oW&3^-4$KDInDB7)2> zMI#hIjH_(#_^l}kjTx2HU6bkrO_B^Zh#aKIS45>jBZVp%$TFHF`(IMsgoocncYv%^ zzFXULthBYiJ;#5(6|*qnQ?vC+PP5>+YkcIk3FAkTu6Epi8F@;&TBl2oF}oWcSYbOh zNH#&BCBQbhWlUZc%R*R3RDGUs1?hjL^pv!6-=MSEBRnyjLK(WWC9d`2#ihQ>w zNwvGfhV9KGZ3H)FeDKW0BEG>}p1J8e$H)=~_Lf2Dc-X#qJ;ko0ND>vD1DMO9D~JmF z13puvPHIpz1h0>5lc%_7B;`=?JGMs1ViRPq9<{Hmj?ob(V2nHRNB3G&N|+~?D+i9) zncKzMc4rz=^2|dhBM*1NK+*^Xr(!N7pOKueO|E820>-X-oD*)) z=P|Md_IokoS9Od2Y*WNI@;Ou2OE4mO7^2vR6uD1D5Au&i)CVu(56U*UE%mAo z$e6a-xngpHJJtbglw|v(`eI`gIyxuFVYhv9q`z!r&P~Vr8l$#DfgWm1$AUWfzFF;( zr4vxO@oE6xxsp^d%O-#YicIEdW-gtQOJ__w;)>(Fzu{{~etP;Bn-*sNqH#}jxF&58 zJco|X%(bul%*lYjg8RS!n8a{P9XN2UxZ21CN}|{VimcFUF(pK#MXHe3KHn|bR|B5* z*t$J>gJO|UvY+iSJ2Ly{56%TV@^jPwwT9%pFpxN81c|*A3xfB%spxo6R)VzS!fOjo zDnPlMYM0U&GIzQlnIr z$ZpXXBW|HF1k=cD0#Z(&DfIU^7t_%=aYw zwgExhnh5>(y56IP&UZ_ZV|uvAUmMW{EFNPCT?;{L+wZpp$mRsv@#opIfu{XKPe$P! zIG$prOQ@4>r^}Qm$A*cce79OUPO5hAqVrvB?pQv1PArAWvq!S+73QJxkFAZV&pX%H zeHq5k@K{MtJw_HgaAs$#k>k0JVv{J6Kt<~d+#p-nBuSdA zR_M2(@ZwW`?C2YkM|i?dJ7qB{o;HXb-s6Od@rujK7Mq5~o~s4LOaRQ_B*#Sp`F4}! zviMWKPkmO6JuS1P*&I9I;2A@jhsoWH{@*;C<Sz&O_@dVg}0m^5M)9=?AAT~(WO{}tyUbDA@}iWU;9NrEVd~oOU1H2ft6OGqoRFO0FGu8%E4n7D($U>VnImH4JKGZv@rK+6A5VZ(tr_>MR1w z2>03h)GIf?t)UE##(?|4DPWiw$sci8nQH>9fy2s55-D~iMV3*~=##E@>6@rltO>6L z*VF!v=ZG`)>}enIHoKoRZzJCN_xHCw?=?j!voYRQ<$>S+I2KeFrR6-;op5ci8VCaK zdnNmhEgGH;f=5fjSY`S3_%aWI5lq1N1bX)srNL4hkN)R6Qp7EC?!eVXwMI6inqq;c z@&FZ$l8lLN-B6-f3^I=e{1*4ds9MPYh(Bx(E(+7t8b%?5TfE0W2);WMaF%X&ub9{% zO_AsO=0L<9tq`)g0n0JucN?P;-HLrs3GEQS4U%}f+|%4^q*>Ej=s0jxx3{y=Nbc-YmV8x<+Xl`*?B(jk{sX%~py9$0Lo3?xsI`^@_MLs!hI8 zawBAeTaNe>pR`x{c{lyL>GL`r>2_!ZJnB{nqIV^JN9l7U?UkdW)&aL#lqh&$$VDi<-wm0D1Q-yJagOf1<-9pC=LrFGn^)0RW9!L*KJ{%$qjd` z(XW#NvjN~^O?V@#%gM$HGu*}^yoz5-EATN&V`16tobWMD9P)kFK!cen>MD#Snvp6U z9XN7w*k~!*N3nY-QbN;WLZE-EVlrG7gGL8j5HmG^((f{=(5288 z9?Ih>wE|7%ZO!9RM}J@*P2DFJT#Pgz$8YhcR-c{I)-$kpA9$!m9*#yKOos~&0gsKXpCCrk;v@u$c6@y zQ7vmW1(^x_Mgz59Bp&HAXhM;FN8?=`=Vlb%I*iRq?GL zRVEK(v(W8VA=(MDc1Y?4S!K*AXDdr#|BYM;H-gea@h1@F%2b&FNE?_FTIHi^P_!wE z!3Z9mV|&f83L zPXlu{?5ta-tT`A3s~e-rT@DD4Nx9$k7zFzV_;?nj&VZ|NG!L88y&Mt4Y3$KlY~C9B zwmkfiDIo@z%#Z`S$!2&x%9N-#119Iu9t+DApxYRWTU$Nb-Fjcy+2Qa+w>EkC)J0LBfh0J}-)UUdx<5p!TfN9``FxPD((<-3*`CmHc+0tneE{;p z@z5oPr?foX`$*e^ZT1N)5Y2{n+JL2yJA_7Bmqs{e8B)w%RTL{#NDVH`@o&C;nXGkS z7iPDS3$u-4(jqQl1Q%%~n><5shLsc?H3s#NM^ z%rL%kjp+=jleV~DW?+jpR_lpw201~)7ISXl7@VLnF7Vuilh5{F9k@cpjQ@(wGkQuu z9m7~B@QPx;4us7&D5!RD&$KLjrn6gu?&C^|l-?)L_S1^(yc$)v-95f*2}m(okI zJrub`MI*Z`__3l6&}@PEhlMG3UMC3isx~l-gu5VAn?^fh zjd(5ZiK{ap35i$0$73UrX68c>o!9ZlwLpm#dHhnH6hWm0YW5_47O2I*uaFUn z4bIn~MoX>81nteG!p)&@x`oCqI1Hf|n$J}}hcqc80B_u%;x0}x?ND}DFUU;mv^ugV z=#=x4DBw(2c9S~kUN4O3k9EnW&MH`DjVv-CHob_GH5%6!^!{nn6r;m}!c6WngURO= z`X>c;(OGmsxZ1Z?(x}sCxGF}9?3FI@IzFcD5gl{?-Vt>p=x)S5-WlltawgK)RsjZs{B#Oxgd^qnHgW4iNsG32pIAr7VtJMqx2|juvr8XD=s=(xHs9fw2n(Uz=5MQW(x6+0|mn7Nr(8E?w5j#W%YEQE4UHJ zF`LNmk?m33b{%rcbT4FDJuy^sVgBe6&KA#z75K1t?9p8KoxDF@G*y3b*d@n|o(t}+ z%Uv*XixP$FAh-_<#%g6ZeOY|iTdmkJ=^*T}@|aZrENC;tlI~-4Q?T7E!LZq|V$#PX zK*wph8K?*@AINh15R<2?XOR}b`>P4A6+(Q*Hml@Qy zEO)_8xK8&2bBWc{TVa3GCB?iFb})VF+&)TDFRL&#f(J9wsbNs%e8>$$`QzC|(+n7j zd*6E%Nq=En7ATn-7Wdmlu>};#Lur|cxn1;4X(jB$LGw`A6|%<%TcPzwZSq5}9uB0Kini)vpFf)m#>|(NPfiKSBvEYW%Zw)p* z?`%zaJzeD9Zw_v8PlqsysxLeS1P@n9Q{)Hclm%>7oROwXw9CAX{D`^D_Q+${<18m@ zGX5t455DzvgEcww>x*BIL)@&113R@XM%JW(V(Tb!ii*yq&xzs#b%J6a>}QM% zQdRM@X;rE_M(w(!duPT9?}V#R{uu2=62CKSj~LAd${(umMGsh+m^nC3I^8-+SJjE) zX04tWK-?eJ6VeZ^vD&vevTy3*ndhOh3Xk{BJvp^GGT%=t(iPXJPEPINB>?|mbEN8& z^Nl%!P9R8yDKNFta{7O6ARKAo;&{$N#Jytdt4ME65fHc67>&fU<*Pp?FS1 zMdPLhD{6>V(33!$hB+k5eow*;URPH{M59u!MJ zDIWifxaisdmmE#glb4T#*yA^^7n(Noo9TrA+&kC3Vp@v4T?W?EKW?e64_><3tChiIOJ5# z2hG$^{a~Gm5k3^_#D?&pf9?(54Q!XpJk^o@MI)WF=hGkQ^DI5b$tpPRm-^xCl|M7( zpW+e$*=Js7anyH_GK#r zF9bG1DyxRhQsyxanVTV6QHLv(i_)Fa5p9K~=z02eqr?2u#~#hQ?M^SBHO(I4n(2MUy__r_rCgD7*|c8#1u-mUs>?rn^@IJdHiR7z6}qkG zwSlty34*LJxP>Z|)ex+xNs(`2+NZV%iXjWS*}2|hk^h%;xn04epHnL#^$b9a{%xPY zf7FjA0_ec2kC_PI;lK-%nIQfxOxzNDhDD!YOySLhWHR4A`f3ek7mMf8f2-JtU$n~o1GXE7QA zQrf7}T^ybwN9PurSJCbdIqjO1O=nW&3f1dhNz6qhK^ojbtkM7Lwh))?vIrPYGyOy(SLqYIC+9Ruq zctI+l9JYh|XD$EfWU8y?u&$384K-A?z!vK+QO8{Uru&xj2^PDu8(I8TAj8NdQL?W- z-z>#lVTNll55fC+NKt?a zJFM_TzOa?P=brA5Yae>lJ16zlln$n!k>!tIg@_e$VGsg*7EZ-z=nl(F3t z5kI+7rWMtAy&zwWANx&s_jxZ4Y?74vX=qfZ(9jQEw|i@#)nCi|pO96tutDO<1<2lK zICrAmIAr?ORb^#~PkFSH z79L~~bPAuC-)Qu@4T<@aijDrGj3`++gI0M!7LGld%eVpLO4cjy{K5c`UzBbB8oBVo zgrob6;Pg4gc2cB`icV#=DK{&wfR~o&R>Rxs+84fqv=LRM@+3jg&uvN#(=5L}>mG0i zJi1^Bf4~pCX=Rz;pwkLj8@Oc0Xl?KjpYo|kU~K}9k}PGL93$B{Kv_V%_bGaD&`vtX z7muO52sl57oU%PG`(Y%x8-)Due^ld10SccAaMV@(?pje29~W30f4N6C{CWw{g1pxY zmC@k>Tp!i$OXj4=F(O^&yy0av#H7u*_EtgD$Tc7Bi;)5Ry)Rzxlo`O5KIOmuNK#%H zuz{@OFx!<)vCssVPDNuK>5vmL3}A0L<~;Ym1bl4ZG~ugt9#_EOZ(!=A1-wt^rpOx< z`Yptlq9G?+)3A*+BUa;Nq`_=>&V8fPHDbU?kyFM=vUCE;HhRDtC^nfQYpCd>G=6FB zh^hnXAl;K9*Fv{bsbZ%HALL#yEXy+U9kJoBJA5b(k97ELc0cB}Dmd&m>*ssd{?Jt6 zz+typW;6>BnvlYT$V?+_SM=H_V^mOV3^R+cYJAI>yyAq_@!#yyH2l$EEdEsYA2Z1p z+!hrFuIpK06u5dwvG*y`OGS6m`tCinZcUOJDA!S?GMI{8AHGL6DAO=W{F9=s?qy0m z28x;_{?=(N0=42kn;|}~Tmn3g8fM7pT37;mEi4ZljVhl3ufySG$_{xo|GoE6nWav; zU0fEB=36K~u51>yN^8hbI__0G(m5jd@s-NM9@~XQo{t7tBP>PUBsl^;kHM8m;_nEhER%1NWb!aik0oM{gih6pF<4hEGEh&rs`j^0+ziwocny=M7!dNU?Z5s#+5DUl z5$`ubO##IM)kzK&y~?9Ee5a_KuNB=4Q7g(__INIK#&T5*7arm#GPusE6}u+Y2?m`& zv^wM#J?I2=f+Q&-A+#@~Iiy_xlx$EVxZ2l_#0J(u=qpUw3ryxA6jECGpQZ}dW~84% zVuF{!!~|UpE1rH*(dyY4rLVl{mt!Z>QF`~pT)(_ohr|giUZ+-6Ftz-SD9ubHvPuF% zfEZzK_?D?i!+TZSBDf1xQ=gI|S@HBOQ#1Hml&6&GeyO&Tea0s_+S&MOUToX)VtzJ8 zXv22gq+)LK#e1gFJ~Kfzy$69SKGK%~Nn(Sd9UKDOq27~0sh4JEF|5@Xqd*<-4(Li4 z1AA{BK86fH-b7SK2E(zsEByMNa-su_WGsos~?ZWhaGV*HUCP6`dh$lbv8n z!Mr^>s)SG-a-mexeYQ1M!x3QS*kX%j=6-vf<^))3+pkZRziPn6L*CY(6D_wrl;fI2 z0)#rl0&}M*_5?+aq076CJs`+=<=w{zjZsJGfjK#^B)yUokp9YH$uY%cF}$A-k>6cB zRBg!v8l9tb{nSIyAR6zL^h#q?KhW}Y(|hOL@v4Ns^`tzf=})F@b*=Im@Je-sB0;8O zZZp}W{QHNoR1UA7*BEu0)w;I`dc^zO;&oT7_Py`bN@s%V$8tG$6KgU2hOeC5k2#|> z(s%AUxDoccO%kN;>h@eIeGmWRKEcm;^0u1) zXM7Cjw>g=gaRYl74GIisS$J;4L9&9I`ElF@mjIN8nV(dO-9V9K4B==632eSw*8FTb zF}Rm#MP<%c!@7Xk2tzek&Xc9YkhaYMV}2&NLCO3$wti>Z#+JI#6pA`L81_b=sp^n~J%&?N2+)4J2PJ zG`0qX>3C;Ldr!S%upocx{8I<{%z@XN&y8%tMT$L7k+Y9%!c|W#y*{#2Sq#Mtx)mm# zAJ1Rq^117oc^xjWqFnO68kiT9Ab{%jbQtiJc&4gNaFE#wPr@3K5G6^~DF!^-2^?0#S#D?d0L)B2LZ z?3B*@>r%4afz8gRMrNmsVxa-51nOa;PJjZ{jRjfL%U#sEpD}ZA_Oqb|PuK5V?y}iA z$2pr$R;m@)--{>T!>oIrXN!9s@4Zi*p-~c2Q7!Zuk}?ro-V}L{EG<-3s!R;t54s7D z9ONz6l*K`i*#XI^3+DI>%Y$DRLnt=3d5*I|>{W`uF$-nj%t?bs~U zCTk}9c%`npm0Ni^p%(+Y!KpnK@&2M;)_nWVx9Kb5Edh;DaqnD&tJ^#a#~_+$B}Se( z$D_S~6Go_4J>U4&vuW`icmrgH79YEY`(2BnjRl&n^-^FTBSzE;ObE4LI9k#q0YMfh z+qQ|xuwdFaBgTR;ZE=Zt#AMI@Z`QQM`vqyDt2J7!9&5H=65o-O`4qbKioZHK$8^(J z{;-cXFzXz__RInP4qmOOO7W%i-6KBTvK``0%B^mjW?l%oB`ekIG#>u9NB+rjVxyNkh^n4$3}p?&1A?88ud4RQ-eI@I0R}Gq!!( z{MU5^C%|zn_vY{Y_s-|tJGm3y8-74pNvmp|GsT$5EC~QPC+z4~#p{# zO`Zii-{tZZup4d;DE8@yDtYtvf=l0TSTN{x>HAgiclEc%%)oJl#}o5trxTpuLCrau z9%@>Tz{NdrT;oX0lvQZxe)q)SL}s;nojZ`3D>Fh5!5`?^?RQ_}T^@$@6!~t+yat87 zzjB$~g`4#_;0h-&Id%e&@vt6Oe#-b6;wOLHb!auooIt9K9KT|Ug(&$BDtd8HHob)G zT7bM8pL-`p?TTzs?DJb3)G_ytS82pHB@o8LnX&*iRJ|627c=dGPTA(^rSe^owfxKC zqIsQB!0JqHk#@ngIqO~4lEp#A^jcCFxhvB4Uur8Jx3Q7ezcPMSHh9JD+v~7R09^-9 zPBm>Ka@bYJOk)sI=;>>)Fm$FbgvL&=0t$uT@ExRaP%E%5yC8gdM658*^}3*d>Tti| zUSM}e+iC#Y0HjYFf)jwpz4WW!d}c~h{#-@v%}|%O)9wCh#fAB&ymRTnz!+~RLO8BG zCGH35S?nQ8b^{LT3+$pvWdY^$bu`u2!mu&l_IA#6s;q>K=`c2bZ5}q$ioZMbyt8+< z(}pb#cA*}6HHBBMC<9{EPDs_EI%$f$QZgt@rw3&_c49> zk&u4O0lNG$$V}Q}b+ZY&IH1PTOW$7}b7M4Ne+R~knXo_F7`5WA zbeRj#0be4ReQu|hQIA-Ouq#dNUKt2+J^c@Dz&5C^LutSkI%WRh8d58b zpAqMG&R2C`h9xe@=!IQws62?&=c<^A;83-IHPT-;T4>n>L?as}OYt(|b+QfbCeAP{ zWlnC&eI%P((WL`}=%CTUS3A z?UH7)#a*>7SOYldcg=9ay&2*GYn|TvT+}Y9ldke8A?t!K%)cXP7t{%s@liSIrW94Y z)9Lc5Hnc*v#m$HiucfAj73=)FLVDyXG(Cl$STF~Kjgm}Y8@C}So}97Kz~sr% zYX!Tzy^mQo;RM?wZ;VQIy(1ZzwLG#o3VLTJp7vQTXbRpLRuJCL zTO5SM>x*>#A{}-Sh93=P9HWEM3EAeryQ@kg6ckY`R7e$2 z(b|AicC+^pk_4=uHOe@)5!xt`D`C(nC+H$MF1yH-yOg*Ugr6eiz-We7_w&^3pc5Vy zgztvr(>0Gn{Bjq(8tD@7e1}W7*bYX-CNQu=qJY!piSj-#5}Eb~IBYZ+av8%`m~9lB zPLa)2^zNS(e!cyT_=T;WL$59S`pzHUTL?1OZrS1_XHdtObV#gY-9Tse(c7tjolD|I zyKCQC*O3c0;_-LdUI&~&GXAAEqlQe28D0>l#Y|x)61g@>us}z3#-)J4D({@ooX{4! zNpgzp0q!{5GXZ@fT&7lZiH5|xWT$|d%noM86am(m9n(>4|2k}T+uz?atUkp)r(~ql zfmfe+BWj>~6nlpvx2folKHM(c#$IDIGaD59q(z=kT$vHN9Jaw~MOtXBEZ*y~xJ`au zvUAdf`3=(Qz)PYQL9Fn?{2?clKTQOV!HQ{@bi1GR(<)`FCgEMp!Nmc4p)PaC>ElO( zPx&$o12{m&sb)$7w4zqfPW}vFgxhG$vV!WDK6aSFAe9X6){Db$hQQvsfw?5wACavi zZEKSiVQkAZ1NSBtKmynBE_zx$TBLr1E`$Uiuc~@fL`f1X#w?~ z{j-!KnE;E7%Sboc`5Y%ej9;=twZoLlnagtJz_FPOBlxVP*wqwSg<894JdoQ|=j&kF zO5Z{~1`CqO)LR6hG4OCqeywpcoD`Ey11}Ei=$Z+ovgs#xJYnf)tvjpvjD zR*Y}7^DFaM+4kvQE|?M(a#>RxI3Q<+)^O3+u84Epj|D99uY{t&#;6m_R<~SwwO6lb ztK0HOeF`#%mny0EIH~xQ#7S$%h%1-_b{?4+JEKNF{4ZwyZC{|l4K3;_j3pWe-T)mo z@+0<9>>i2~qNq={XP2~Jj;hDVVwNH=3n-eqZsLK!b_k}auyeiKB}TYzVkZ!y>U&v1 zq0+6Nw^ErrZP2M9_-ruh*zWYr6Aw9^Cw;DmAtr}iuPO2pk`Hb4X|^rdwE_!|F!97s zTY|*A9T}l4X58Xeew%Juz~Hd1iJ6jwJCYMYErNaEp8T^_eo(2I7e66ycB}jg+4b2= zCTsw5|5(s!IzdXH+4U;0J}rV*`$4&>dU|UJo<)VH^CJCKWpLh3zRKwFVoQ{KH5Mnl zP)q%;{@%l2X}(vr7NCrZNOW5sf!gZGtO%LHGN@l!=?+93 zf_55h51eZCtn$7nTEyS)-|q?$kpd`t+3frv#O|Rt>yh)-EZgJAJaWFXbj?R|4Jb*M z|9|b8AFbx86IOFD;|+&-qtT)x3%A45^a>PZN|;YUw@G zw@izZ>fgsSFow&prwo5WwFp!eOo;=EFDt+F)=$k7Yk-*()=ixGsTpYzaM+O5^1#v+ z|8E8V@5{=kcG6g;(8j2cQS1V*+ZDbR=<;luo~2A++kuT;12t4wvbom_>pYPk$V&Wj zK$n$J=kN*ET{`vb{dXoBY|XXbUHS)E?!dmU#>fb6qFC^?)=|;NWj*qXax8m8d7|`? z+r%6-Lvjh|nyPb2#s z7#od7*f>tHM=5fMiq56CPP-)P3tus})_D>0&_fNaa`E05nGG{KL$*xo66G;X@}(16 zLmCy!CV4?B&G#N)dFSXZQ75V5-lTLldCCVyigWW3~6zc?SfEOmTS6qzi$KQHz)~-!JALW3Gz62(j%9 zSvMIY8h0+w`kB|7pNtW9M2`FQOb@=~FEV&E%hRs@mLza1&UN7I(RL&MXA8xqQe*=a zo$t0LyjI7ms}qqu5SYML23L@Ml7diGGsy`>0(J<_fEpE)MHPpQk>s#7Fn=aSZE!*2 z>eDZ=27I_a=)FKTIIzcCVl?9#ip`?PRw^3x$(tk(1uH{!6$P!HC=l5tzaiHKH$o@S z1A54*Ssnwv$O@($1kcU}w*zryJE)B>1F_Z?mlpWjwl>XVUaXn>M+VMhD4V|IWa7qW zy7k6B??nl_cy~jt&sxmX26xjt12zW?%C>kkMpXw@@o%}@4Y@18S1X8Cu{?5p_!f_J ze^mLsMPh_oJO=DvfNL#GY>J5y2IB-1s_S=m`sNr6%72`myMq*RTW}nADJ3;V=BJ8c zwG=6%qIY=g5wG_yowvvDC|$O2!)rOb*vQ7HdvIpGcZtXS?>8^#@^6gFrF%(>K!5F` zDeZ#I&YL|3VDnrnTko9?a<(@ox5{eVc7OtOzv~vyRJKN0Lk{r|OKQQEU6@}&YTVXD z6e-#vmwAfU_tHIYTUab>5%Ivk4(ZkEmUvI#w}e)ihM?`-;J>f4>KWx)T3|1IY2jc@LR zqrTUge{kS!&37++FGg4etvXdqr{DxxKG7DBb%cNJ=%4fmgWBK}x9_&Y#{D&4+;sbc z0WhEZ(p^e!IItJ9+^C@Y0mXuR{avhQKn{;eTCHefYGwKk`Qq>t=!Q>V6PYgpdUy#? z5Dj!MJs>5YBF9qe3-c=j|JKjL>)*>(mQOuKe-VIWLdD_BJX)a$d&mi|xGKi$GeR3Y z^1^y}ccxcCi8!VJuylJl|0r!3L4~C;*p^?xWcuNd_~9*rFJCU4J}AqOK<=x6sR;Z; zd|8a~!A()0^6^ZuELC!lr^2)Xj*P#)PJRI6geV}8=~*>DMyO>}T`pSYB)Oua0j`tw z&hDPRPx68kbx&N|{XJ73noGTv1NX_A@wj{BiO~Lt?Et{InXXGeAzK4_);B!96s?rU z%xMFr3E3oj!PL@_@6h~!lZD3Z-4N>*8nmW_+|0jce z4!q|!(-KfIH<4c~iUsNh)hS3ej&+IVqsMK*v_HCZOn;hRW&b)=+BD0sa4l@kTSfA@ zdCLwAm19N@+kT4ON0B{P{=Yq7XIQxlO06LkP{mwTgF>yyQesEIkW&M6CbdDsef*4d z!63+ny+-ST|9|$r1g^>STHjCnLh@qBMlksXR3w5djx2_X+Te6L)8*g&-rnBnHurXh z_Kvrmb~;UMJGU(^C@vrjREbO`)@k*n~GG=WFx9lYX;9Py@Swp7R#iG8 zI^+ZhA-e2_(WkS#_F0x6S*hEp;-yYoBFGe}H-!`m(F^^=TT72kWwPZz)5dSp@-e5s zw{GW0$Qih<>0b2L>R*pH*_CB0LTf-T&}Daiz{MP1q_~!1)=*?66^nE|kdQ(?ccbL8 zFxB4?zv58<$QG!MVpzK$VFlIks{VCTEICtbb~vt!bz+5Kg-%KuS2G2=z&3G=UJ7_7 z8Io>!rsxXb} z#iUXs1w$NLeeTj&Wswxt5O#vzMzZH8iEfKD3M|pt7yXHkGdG1#eiJOd(kCBw;u%&y zPS*{X-=5e1^9d$osPylt$zpc?sS8^T5OW!(?vg?=NfcQPnMe>&z3aYT+0KEz%TBL) z8eGT~g4?5L_iZn_Y=pMGAH(|j|LTc#x9lVGf~+J~x{g4iS@s21q?JS+3CRw@jRyXg z360OvM>dFFkGhrh*)M!o;)k$-dsCQCCHv^erYjkdeJ~ENPxP7B!*}NS81`fto zlSD{$DEBa?3O0id<$Z{?U5571{a(B1BG?*tiPIt)<+y9bJ+E>9isjTA1>Rc|RmaD; zOAGWXL7&F%)pj&6GQ*UWH#Q*UskbyWDUx3@!6@K&<2oYZNI_6*OMNl!#OJ7WRF|PyL$g zy2#2&aoykL?blwLG|L2)_sUh(q~L|IIH$}Ie}rPnDEJk`RtI&>MUtslG%FQhss5iqts$tO8#a~3)L5$=WxAYdMV|brv-w7! zeA;A6st_7|vrd&E` z2&#QF=Wu{Vxn+6*T`%0h?UrGgthz(m;9Wiy4aZiWd=Yk#Xq4=o0cOLfzHxqfG7tFx zDCDokrr}u=)f*+Z!{Y={V-&I?BGq4`C>Hfb;ZYx{;MQ_K6{g8ocpyP{^!j3KFmBaG zJnh0Le+H~DK2AIPrJvYS)#}3gXe*MIsOwP~oF<13zq|}d0k>a{*_n7=8@RZ}1T$`< z(0C$#l6X0Em1Jy^&Z9!qC{e=Zv8L7k`ZnKWZWdhJ@F7|C!niRz&5Tn9#cZHRDkhS# zz)YPRf!dd8a@+&yBddUOG9c4}fBq3~iPvdnbA&qA={?jQ%@bR2JbKW{2XiIAxoRm0 z1%<>O@5lb;cd_sO>CJcEeqX$tVwO@Q&i<65o&Of-06E-y%A^?JoNO{`u2%(LB*-+= zsj6hzp8LryehWV#ID7t4aV>u%Nt!nR$}DN}BKaxhU7s?M{Z6MX{}i@!4y%bR!`gih zt0jgC`RAQD|u!J*-SATDVT-W6A%l*?sxjgS|~?O5?%Aj3+Rxw z&t4^KFS)#-4S+{incIasOxKA8vLPz`eZWoh07e)tE!4KPg zl~D`?eoCm=hclNd-ptV$$KPqia!pW!J5=G`8nTO{?H|imVpwCxM7Y?bpnX2 zdxhx%x8x6b4XVp;r2Sy+8>KPlRYOl#kp@-y+cjTX`u(*F9>RIYVT&bpp5~_FHCv(a z%=_43=iZ-ANWx9n+2)^}P72vMCa#;_A|IO}t%71eZ|orK^XXcCySQ&sF83xGe%zqC z9lAMUdBhjw1|UwSDh}KqP(`;xUGAwLcgE-dgH`k@a@WWB^j$+dHfo3F^!tx`Pm-lBYHfxj^xEV_FWKDEH_|-V+HbUoFE79ZY4#E;U z=u9}3bIM}~I_eAWlLQxrPOcerHd0I)Mb3mX^P5Nr!dWjj|P}QgK6$;1BGKI zn$0NeKoK9(FSoQ8R(_U7$s=T1B^y$LKPBnxHa9LDnFo>WVV-O*#ei6U78P6Poy_Zy4!Rlp z2_d%)Bk?)Pe(p*Ly&VZg&je*Pa5>5o>m{YWi4j>5NO-~q&rXlgv%~uk?(x+=JU{E{ zDVfcSEhDhcRpFnNsH~RWC0H{LJ0F7QnGbbWTKb5xi>%;P2#-RKW^`(yrVIE4NW@<$ z-wa)!#>D!K1_#fYLq}Y?$AgK3je@yJ@oRJ z0K{xEZA6RNGh^{)?#dseS+XIYv>Glv^Hw;IM*br5jj-2SNt9vN(g-~Ym-CiM4ej2H z<)AsrT6(W{&s5kRW1sd4knw7RfL9BN_sykyrk+vs%M+Qc)3#1iXDeIfO9PX?3isi# z+K}tv-CR7;By19@@u0x3z|U}?6=h5NUjxiXWNk~!tK zzb9!fyx}b|^E+}VW;+E21Qd^(I5~78lNF(*ulwxt$zrmCm%7(*@&#L`EnK*;1y1(N zE{`0Lr9^{}MbGS-;AWu3o`ybSAg418*=(bq@n=toU6^?G@T0!NIwnRcNSqcG*!#RU zWk!-+l4yrlqoLU{f?}m~H@73wz~Q~^oi9ihr1}?!o{}UQZonI{6(v#``$&=q$&MNY zo^1ukwo%>>jViB)4?3l1gI%C`Y*I!I4r>px0?qhL?*l&Xn~aM{^0R1i*M*Hsl9}{Q z+<4|8Mf$1O21POtiA0UCL9d*nQ9KO47oNvCPP)VUJu5)Bs{)F_%ajA&>TY?f!YFDI zKVw-y8zh{KOWB|shTyQ_DYo(6&SZlK#$ESj`3hcNc)u5t#cM%x5;;tprnf!jIAog!bPr8B(rlEh#*jkV|%;4^1*2J!OEngHj``|aWEWq@(j=^SF7sWcz-YA`WRt#zycbTm|+Zp{-W3Mdr zU3ZOtambfIsR3r@wHgU9!6JKyrh;^X?zie|qzG|M~uZKjNf4UhLq=+QUvf&u&F*obsbDEQ^m^Hi-b_ zm|dSlQ|@{#jzABcyxo6St|B zh93DLK7e;DOrCjTcY8Dk;LhMB7##AjiJYD3PIoGP^5FMj~(D~ z9&%u0>)`46x0^RDW!zoXrdSbXM|MN0FSe?xbvFE#TXf^PqrCqiB>I zdO2;PJuw;W zSo6Fw(?ek7+z}AZxCvH$& zBv`Uq2V0~Px_pcpIh#$7ZPRS_$e6rgWirN%&ky~J#I$P`b_(Am$u7KWE;MsfvnU2+ zzcQ%UrYTwCB+zBjgLL{TlEg6*Wbewdpk>WQ0=x_U@&$ixjd>SZKfqP=PN7S#+@e_E)0=_W)R7z7*O`gred`;mYG7ubs4Y^ zusIUuq+1lXKt&ScAE-T=FW5T+gnB@2BlLz`y;+emxg75OFs7TYPWA7g%jm^|e9_%W zP9raj4lKjLGaP!3J7xuzasR&LpCy(Rh)?ciT-edDQfJspu8QNJ1$Z2FPo`68{jz4R zBJHHeXS?J+$>mmXG9)Fkqoi&u7b13maU39?HIj9}Z2Z-7o#kS~#y@r8EwGiP3AK9+ zaWym{rJ@JH)tnQ)IrKi!pxfdpi>F}HY0wRCs0Nvy%i_Hd)E%204?FOftjx1!-T~KN zJ{^43vT=;dZgH)&f@$H`@;ijPqIIeRq1S!->AlJ-{^6jr;-s+R$P=*kba|Ff(|YJs z82D(H9wj@0D2l~Bn5Wcnwgx3ZLzwbu4Utgb$}jfmmYwtMowaQmB&Q%PxhE9cX|D9{ zmFrZSm@LsYZk5vbX^z<7j0zRQm!Of39e@h!CS`)V03tlt+{k*DVX6@Vykyz@Q{{Pl4IPtZ(@85Q4sq4^$j|xps zTArcd9QJy95u=%E!i+L;WhL3_!W$u|LLHXU*-bGZOsS<}<9W#4#@&xm{LfHg)+#UY zE%DVTZh4-cYvc{2$I1#C_cEx;~elEQ11)Zp#&$XYB&G^4x9Di z16j8xBdkN?zxdVN9?PcAY}{QJrgU0q_uMLf5WLr;O{p%Cp$61qLAJ7mKj`-GdmFwr z@Otlp-j}l?N@TgpN&<>d4ID%vl6|jxkAcrST1X9_erFs@l3)MI43klD6Dbdn92Yh! zADS7J{S;G7kv&vwvCn{PP1J$Vz4OnCF4CP7b*i`b&^hk4VR*8TTO!e^y1>7E;9DzO zI;U9H!_}!8ptac2Imx^}l5G%kxd|fR2gRw=lIFz;j&YNDiHz}??n#Fn&Dk6+5}dul zure;xZzdO}|JHAU~%T|^}Uxb)2N}DD`d=WhsJ1fg2$0p-t`Op}=Uw3ozL`wo38(+tT4V9JN zGVSy>I)|G)L%kzt$&941>akXC9hy{I5HWP`SiYY7&gcJWnQmr-2p2XIR%-6i_BTon z2NiLF&)o(E2wGYni9x%uHx#yj0{oe=#Dn3pj-9w&Y$@2n1{5xACalQ2AY&V8+DW86 zvLv$FuUU>Q9yE$&oXZ}_pk=v%V@H6R; z6)pV6up2XzX7+>hZW+A=nuB-qSIxo716Ygr%RF!L;SmR+%tq)u=?z;w;~12ED}LJz z%QQ3_=(upi&Pvj{FT7P=5%W;oI(slcr`p4{Tc@MlV(}My3cnA!qer^(P zC#~nF&iUJb+XwArZ%}67Kh`jKVsFs$*>`4TOv1Yz1P;rov>z;%BRb~&;%`=S7MMKL z*^54DA>}WOhuUc7p`M`_5cofVr6Xzb_437nPoYX_ccgj)2k8oXWT4YP8%X_Ff&s+- z8}g&PbzTpk`T||rK{wRT)yrVJbkI{LJM{96klVcVVd_#}W8qLq*e+m3sLMTyeCk1$ z>K;_ZVav^a&p5$DUJ}nxu8sSgWcNm)9-4-$7xSQU35rUvTI#TSos=!??8lQ;D`b9w zCCs+oj96mH=6^wqj}`9!nKy2I{br2z&E-F>=56El{cFp&4}Djo7~oa=XNpQ7@pMg$ z9A1!g;NS2p@xpMLk@6qD)s z?yk4qAzRp)9v60T4w;#r0*V1vZZ0U;MPLX88(8&BZP*zcJHzC1C>A;jWmlOMEy8V`y9E_C-CMoemu%d!q`42@_i7 z?b5EGiXg`$#{sB8@Yn_g7I*!6&8lC-3>lf4AMYhcUKrF|GDFSB6!Q^9j)U@hXj5>r z{L=iBL8an?s1|-&M6s-vS4FQT1rdL*pxb0+bB{rzp_{Xg@r=ImX2p8&j!H*av(i{XQb63Xf9SaN%yY)N!J$mg~vs~B+Sy{6H572b+gX{8= z7$kI7Ngn#Hl{bN9GJ-~WPMJ&vaXC|yKv6(sF^+4M=?bdDFp3QK2Ypb z)qYK!znxcAbIy`v-gWuVsSnPpGzzVxoL4Kk0YtqN(X!C=$??FD=!w$H`nmgI-`&DL zFn7DM+(WC(48R9L8v<-djigrE$<;f$a?t@mFj)!5fdEHhm$E$9zhj$aMA%9wH8=^KNOv`$&%gGHBozk1ipi-ZkC?}l`{<_vpb}@4Dhr~mNC9etKzY9>)*3@V9P@G| zQ4k3CI_C@VbtJ}Cl3jEe19!=hbaWqRLQeUij%HfMVwzY@i|cwL`h#;fh)g@1ZT$3R zk~D!7m~9|56tkHk8;#8uP-P$UD5!f?P7 zPmbY?GpqpOx=!7_MFW*w6F}So-IkDe7vB1USo^R5Y#PO^qsUq+78Q9Ar%MD`BIGj; zx*Z9=!Ht`jFDeCVHbyvogfW`|`{;?WIba9ajO=fHBie+E4O&?iDV;zX%)b8P6a)0> zD(t76X^0_q$aau>N=(px1e&8aR9DCs(OAlNLs$rb%u%5iyfx+%Q649k zu9aRDXZr8(!s5UaAUvr9?kV2eM6@9-{PWV!=?-~HbguH|ES(DJs~-z@kTiMe+;hIi zcv_ER++Ft47Hys{%Vp=8@9mpcR`cb$0n6OhgDJl^0qGBQzX~CDT-d;@GUISRpqMWy z(o4ne5^QJMIZ2{U`n*)VK5UEsd15+6t)e1LC9tQMCYWX+}{w@v&?I!AvNHPfX4xQR4=`#qm z4n2ZNXpQ1DH2YkHG;NIFr^z!o=Y8{_u%KjSt1LgjG{Vpm4c@KZ#=ldeKz@0L^mwFB zR_^h!B4g6wpz`@JPPS>BJ={K$=UvP(e)vwWg9^+|?euyMY1|{!3ygnj_WN^*34XY#_EUVx+^!yF*C;-C?=I6DOBvyf4LX4lRo&3dog)pEimf(J+q)XKh<9!nTWBM zDa%-sSnRZ8w#Z)^ZK-`9#PaF9wQF-NRSI3UGT%zynj3SvXf5}SycSYEwZbmCD6&0N zr@A?_RUSX#Uicy5<*?-w2HfH&BrpjPc&t-l2b28)Z(pR_W~Xx3`oHZ|@5r8|mE~-* zx}B8;p#Cf6U%&Ui%Y+yMGZ+K7DlUk*4pPa;?uEbrlIc*FN;8J?b1<`PaKs7-cHTT* zczC)c*TrQ`h!ytAWiTFXv!PI4loj!)YY8&{^>m#Sx!NUB{azq6m80D1Qx&`-;_8Ba zt{PRnwIOS~KZY)!72x9It_zu$@iR6B8;DUubWF#Z7dAL}hIek~%$q;X%ASl*`A&6| zJ;ZMpc1f%dzi)dRI}l)@*>2KGpq{Zop_d$xs;>kjM-Rx>N|#9+K)S3^vf3+O)FnL) ztf>^JYgD76OsfB~h}+%+vc~z&)Dp5ATIPsw*t_|Hu3eV;V=ilAtZ0p4Y$jXTDQ{3f zkvythV!+1G9jV_dky#q@sW1&nMzBSNW6i}=CV}mAKV>L~USelSqOw0*VA({R4U5}_ zEs2$u*G@z(o{ z{SVE?4Zv7^fx|wBow;_|hQE>1b>}Tx={+eh=fbOnmG*k*c%TU3h8ZV=KqS(a4-ym4nMRG z>eG=mc11#5H&1RYn2M54cioNj3E)X!%}=8Qx=LO5N#K;w5MWh~ZGw>%W*m+T>jS

m|{T9vVe-c03BT8C+KBcg06uW>hcL4;a#(#%0i--q{*v! zYv8~1$*D>dAiT)AGI^;$r|On}LCV8cMx#ifPPN%TnY&9;E9{YU1s@Y9&BK0-wZSz3 zPFP-9fo8asF*~+C`J=t^`#Y<=D3hc5(Cgt@GT=tGn0?zx6tkKl2~_NAuUu%fo;FdP z8Hi4Q2kj_z0Cs=7E#D-&5gl{4b(1|(PRwQfkgtEz)%d&zJFwIo1&=|@LYNhyQJ}k_ zg*KVkw-Qzm5afppTmdzf%zs-@JM5?K7)aVM3Om%!%bI@4o^7oQJ0DiIwS_V51mzDP zTv`H$pGlK>3vo+^j_P$EWOV8QSeWCz=TS+<$Re@?6XMDSDA<3?s-J!9<*%6*llb*N z_z9^TNk7nq{fx_IE8jVa`IsUfQL+1GZif{9QbDV1n zC?4iLj`$c90P5*#=rok0tbpFRPsa|q;iXEyAy<|;y^?6y#avw&MMum~R7f$<2yrJBYgGLzoufug9_&Db z<>jSw(ET~Az`Wi1iI+WEpx5_tza(B(#7epGN}UvI4JyLg3?MwLNSmlrELMFyHErUk ztjaTI&0Z)uYx#<~nR)1W=UroXt(xu=s!@x{kk7ls?efg`Z&hSSGa#YsoEfz?%W_EBNQ;gOyD=Bd%*96(13H2qQn8nXg^@Y*W#Rqs zbG{YAo%A_hJU&U+(VIBf|Kk&%Jn7>IR2ssey_u3)lb!0V$K9daZ<% z5X?svNT5IsTPo|ALy_twumL*m4gnz}X9-uHOuBq(S2SwE;hF=Q)d_-R-bGHn=x9(o z=YXV>*W{S5et>3Fb zQS1X7v3v9x#bNiu?)jo-lE(px$|a^7$P35a^^#1;FWRyRb1IOAgUamq*$A@^q!A3& zxKG^UKCqNnaam%D715NT&9ouESt%?O-uA|23|w4vflo<_L$~{v%~i(R&+{qL{8 zN-j(wJ!b2`4T`x=k!w_JRzyWuj(3LOu(EslWe<&_4oYgu=6(vyQ=}%>3wJJ{3 zy#~sCa0m7@Kq^R9#9+|w$UWo|gQT+^KGpL#Lm$+m!Gmu7ujxRM0#}ID5n1jUMK1Ry zL6dK67=$6*gn7+EY#Y=MOTkXBdLSGjPpz6$!EK}44ZRgiv}n*nhSO`ju@desI5{<( zP0FJ|bXw;1ykR2nfoGso(J}09G+_reH^*6rm@WR(#`mFe*z4fv-&BmzgnN1_xcM~gn zMzyB?_Rlj->(hefyfq}>g*Q+q%@)o>6a(Skeb^~n13k;SzjYNf-df2fWfpS{$^!en zD#FS=J_^m_c0rLx8-EA4Y%bPXV-wdj`DcN394PRh&(T@DjU;IvW*thPUNmXmRlZS@ z@`h@w@~pBZ_!CJtM1<2r){Is`a-``T6?m*Jha-K$=+3bNkJN2e`70*yJm7BoIjL}A z@HCl$=PbpXrpQSuHk-Z4zuDau#Q~NEH$yg4J!as{y{2ygs8<2I#Pcuurc6dP#kvR~}gQYBe@>oKN z4U)zZO4)zkcE+-muazFSb;3LH&Hh~w*0?$Aj3P;Nh;GD)e87pQog5VHDfVfBhLqXx z{?n*)5bn4%zej{${1|-AtiIhx%eK{@m%}cw`klJ&&ul5vfB1P{S?Z*1z?cJ(YcK>g zQAX-Zmo!6CC)FrYMQfNbKNy+~ zKe<9!C0jFT6SZgywdnM(l3wYe^{-zQ_k}-n27BMiisMhOnDrlfpLe4X zC#VzsIgh$6-bMX+u`l5yj zD~-|L+3^$8Hs#V!OYV~fcH0ye&J1^%;i8pdnkjMx8b&GVg?jLrTlgPy7Yo#AQ?&55 zQF3c?HE>UI>E)0wY=B0Ey)WyhrpYz_g+ideWD}HS?v`Em(eqbIk3oagon9Bh^pdsy zr5@c;Ae#aV1!Pj)r%OGmXq_)s4xqfVdI#y{)JgA?Z1J^7bvx%i(GxhRlNNvgH`J;D z{pyV3ilI_%F*H!j7u|KQ_N$y$A?%LQD6WWa&e|KKPJ?kjR-92R=Cs0}_0&j>%`ki9 zk2~_fIzmIf?1{A0%6PKI*@caa6-A6Ik-NRqxfgk@vM#{Tfr)*bLSZ6vQoPJ-tPPCK zCdT?CIq8b^h3fTjKlQHNy{mnV&d3$p@B%~qbB6a!+R+o@QjwP_`({z*KHxPxvj(j{AA<>?0> z_j3?=M}o_j5^UCTIjl2Eod&9rov;j`ekr6Bm_2ZH^%I2E9Ue9|c)5rmKUat!~D(B5?&k2e-Mv-bN_OWy|C`QUUJ&iI5 z7lCMSY5s1eO}Tt_E`30xQ{^c$CiTxNiMllZtn#1O|9!x131xh(3VbN)=5FI&0HuI} zDck&qjxS8^@$F=~4d?NmBb*Aa!GN7~gQ|p1;4GiLH)tokKd-XOa5)Nhj}iFz$Q()R z-pMcBiLo?5R(b?3en~sGN-`kJCK;2?k`8(aQ%1GQj{EKQZk3&ew&xA1me+58ea$PS zF$cbf9Rrd5u2YqG9}j}|Fn(3^p!CR;#hgRYXP^ipk8?3-i+5aTKJY(kB-bMB`ff-1 z7dz_v9%*cQ&awNQyR6p2X7RQv%Zf#p<-b{}K}5eLnJkOwiR$rGBa?RxQ#maqs+KN+ zMGF;dwREz37hN~$n7G{INN|G!8N#h{!|m8h0&6$0Nl2@_`s)=mj^uC@Bhnm!WP1>6 z?O9ex9_QWOGWfi=Cc`ib)}sO|7HwOVvWl+d>*;%9W2bhYj>`svprJcZnD+UNnH(v5 zd{~=N$DU`0kD9*UoUxQicUgmCrFX(T&o%A`p*pdb^tj&?o>45HeN8mrmc;ClWx~1M z+52QpF%VX#jRi8Sj^T_8tdK!T{f{avX|yj$JQOye<67%u0e{ zVSNM|Dss|561!1?lDBQ)SsVw8-m-w&W53nvk>f!3(>Fg)uq+(2QgMY673wc#*r-km zwU|(x*(xuGCg*3yJ``%}x9Pa=$+nSL{cv43zV*{rPcDDno!3r!tIu}TRv!@enYSH8 z1M&TGDCcP5ANRWm`lFeFU34b|Rx)6xyi0Idcv99N?eRP*%kVs}%Av9P7~*)M{Q()% zOQL&cbxn8D%492&t(pQS(b`xcnR<28w^u*!%5rhDv8XFPWDWPTKs`|KuluO6+n7e7 z@mfCnvap4(rLkI3H-?-hmf*0>MmgXvR^S-7^TdNUETjEa0_f!)@e|5s9iFw+{e~(- za$Y(hOXn@0kO^Np>U);>^+f4qec{dF9duhHhS+b&*M~uCCeSD7iHZaEP?7AMa|?i3 z&WPEs)&XC<)gf5Ti0cM7SDw?mO)`aT8(+To9!X^vwsm2$La|xcHk)D~;jE!zQ^^g_ zlYry|=mnML4P=Fn6)k-1+pb2@1Jur5*iUEAuLc6a4f#=C3n&r}y5+)JmaR;5nkK>! z%&=5{}Ju)-iTVjHVuv7RhNp|6yj6$=C&!U*E6v=?vFJ-B31E}oO%G4#o1IpXp z7bcegJ4KB=lQeme&py9FH!WQeRtfEV?ul1;gS)L$5u9_#%EXQT z{=$c{Z-tg#x?$Uf6IYke+^^q-#_{sQ+Bk5FEqVd`( z8ppmP-|*E0as?7aIlx}pz_}-@H=Nri>l0rON7B>Vc*|WGHk!(zk$+cVco091ZM(x! zLd|Sdwe1Z%NLq*5#pdL^=e?}`LrC$RRX%{2eQrdz;RqWN?h>qft>LxOn5!|yc7#B3 z!JU4Z9MvrMlPbRREzc+;`RwL)G~ zg<4S8^Z>|sl}}wA4C5{ltPtFWw#*Cjl-M3*Z%|2;Hl#t3%rgj}mPF+!Yazk3i!QQB z_3Mdg9TVH~#HT*{1UoFflyN1)GW+bZShA89$M!HQrgYJZO)*()+VCK_gx&+f2N;jJ z&PfRFjsocf@pfhUWE4KA^*kWasWw5g`#tl{DzEuuLdDzF1-V?L%;Ji3-uHmyffy^{ z9Bz~x1)FCBf!!UfU_-f=-AcEFjg{F+52{4q5x`M#%=oaLf7WnRG4<5;(xQ=b*ZE`P2sK(cmwFa+Q^kBQBr% zR4y~SZvgLh)a;SLu!sD*zq=BoJb z*k!vQq8zxCI#r>p-aB>L33~b07A>aUb@&qUv>AH#Oh4YwJZ&6DUt{$%r@r-h@gB?a z_UDSfZKW0;o5_@Wpzqr=8@iC;0lK(%-Hpva)P1CEZmax09XHR|cm|Jezmzb;!K>Uz zlgkD;k0IpPRc^$;%PQ00>!(UeW|qWBOg%bfqX7*$U#Q@J{D;1js*`LEf z<@3|zOQ6cIDfkT0OR{{Lf{#ZwGUXm!(rw%5a4U^VpKAY=P8L!aIT4%9+N?)b^2uc>tNmK0 zKICc?sCtY}N}~k9=6oYGd=^<4q~VV_{aAVCM6QR)#QaTt;eC?u!Vnjrs~PW|cmJUN z)0^+S{l0i3#lTN_9Tj^t2)i|)`V-Kk48{dzCmSXEgVbncR`4o>C<8xs@L&NAv@1@B z#Lx%)g=v>rGHlrx4Hw=aSs@Ca0ojaJd5UVUuW5^;g&=Vy6vT}k7>2%l8$mI2`&bixM+?kwY=tDUwOWVr(UT#-?C(lBjZ`)~|3z&U9n%1~n25 zu%1jyH&7qbDyG*7dP3tro={^jO>s&5jBBd4E)-XiOt4% zxbXhiO3~qUh}P;OkBGO4HceZ@s|6{zYR+c=a?fp|GI5rmjLM$h9n}w|#{=K&iCOkm z)i2Mh^1M+8Dw{6mRLL#?u9`eY#70IH;Semtlb^uuP26+U_U}+@p9`<(w@M>K4B7 zI-bV+i)0vf#XhMm{1o6@f;1QzkYz<=g5QGlr+t3iNX_i@*hD$fg>h9Vu-X)lyZzsL z7e<&Yj_j}Mtz?f2M_106*^o+#fmNuCimeGw5ya2K7z&V6L|GB7(DW)>*(%>7tCn=p zeZYI{md8)nGq+p59CTN_)xyYD}KWI z$u~TEB+Dn{(0R(P*$tdddA+a#0xZS9*)U^?V;Xum6hS7t;s~3<3PIE_#|89R7Kpqc zPEM&A-W~}>VZo_RRnJ?(yC>T;d8we(V<+druROyQvJW-G z{F+f5j*J)MSytv~ocg7~GYS*NRHxrOLbkau#wyJ)RzxvSZnBGtJr%Jx`YfrQKlG}` zcJX-zPgzTE3{>Cdtq(IezgqgPd#>^V!Rv;z+0f4F(3B5-)dwWUrdEWl@zv6MLf20O z#NaKt^loO6BYmJ3gwj{xG2v6oY_-dJ&sH+iu=AlSVhZWZpf2bymB!n_JEpumIVr4B zes{J`)hOQr0h>(G4qk(*;TNSbC%=DQ)h<0lv}2)xIXc7gG?U<%Ype_lm38)K9pg=g zW!Z|*8jv9#AxY%IJ8-C68@2;qOEGIGvXY9$ssOMajgrg44lj7Yt@Efeg?3DgZ78wh z14j9%b(9Qz@Y3SxCPei9T)LcWf6h2kRc44NrWk0YSwO|M@Dn&Cpb-L_W~^n!;#Tzq z?~mqSCE7{S!yR-hiEN_>WGT^Yks77pGClyZBH<6i??HL@pc{&38!KP)1@~z+3NqlO zXLum&XUI0t*y50||LmtdVI4F=KhEct?LnSoU%GIv*-Cq?ynwM=7^$TncwP}}ri>*}jFW34TpWo6o5q@dK(KeAX_YDQJ1Zk?FHBA#e*AEc zOtL$|nZ@y-q@@$2R$ zGJQnd;N3MnYpPNAvVyycqfx5sq0w&>Nt5Gt>Vfy}NOc1Tw%goQFD;v#%)_8{NfiFY zn0%-R+RmtVL?lLI+<8ED&|MpXt?r9_&VmpuDnR6WBv5ry9rTIlmTitO<m&%@9ui*9kPWT@?1E5dC1K3DWI4@K+!}f%U@_;E5}$XP1+y0}Hd&of-1NtH7G3r= z!R6=Y#+@e_E*yC|Xa<0MiUGGKn~Lq9cVJ!@o&Jrh;yp1&4Z>XI%~?9t>91cE|MR-M z7(@w?;d?~6HS~J;eb9f#3n*9kk9)Gkob8;WK~2G31Yz*MZu+geH6v`3sjtr8`run8 zums1?{5Nvkg~8Hl29^sH^9e<2sn|*&I}`#;3IG-QNGOK(*d{8N@~J9)@&%FtB>R-xU*ToFD;hS6drXgmlPdwQgbb{Uqr)p#Vvi0==H$1!=^Ox4ww)%htf`ARC zFdO9$ys^xnmY+Ij`Rrpc`k3XO%|g7|F3uAF`BG$-I3@b#tk&5m;=YtqBs&3u?(Jf{ zx!mJY@nOfpk61CS`OJA6**7HA|v>TF2wwsO7 zO$-~1m>HJQeVolt{W0scqv57?X`6p`Iw>3}rsu+*?Z;+FsGyj`6gdb4PR}m-xO)b4 z|896w_roNr41|I{B1Jy;x%apyWsoq4z}HWObs!~_OW#msMXZ!BrEbYk`UM$Uo8?{8 zu?td~{04}Aplnn!xyU)sKfpDP_cwgJg@2cnPt_>SkmK%-0nn&nVul&BH#S4zX*XJ5 zlfoWs_^W6LK>YM(lEiM)J6}9b*J+`s&$Dw@pJj&n>-oCe2;OU5M`GN$hBCOW^p0own;d;qkcNDKmFoar-(%;8q{a)8V zw&9*f1)b@i#Hoh1Vr^8kNSRcVTSF8hIgym>_ z-|CyBw^ z*~*TIA9?_D3UiJtBGlPRl$FXCKyQIMX)3qY9WBqDd6&iNP172gQ}l7aOwpZrpjeg6 z>!g>Ee(r;>z$$opYQ6yKsER`$%-icx>I>R55o+wbRYfl+r>E*Ar@vYdT<+0Jp9$>@ z22we8xPJWnqsKcy%&H{n0+2I~2VEer(?_tEd-O%E^=$>&fV^3&eNRvA3_hb+4U$KO zw%N#H$Dy+6EsC|=U$e!jZAQudyT0Au{-(fWm)$E?4cM4;pI`WwXzanC1+D`4d|0ypl^s1 zRqD0=$9TrKXD2@pt6RJdh89Tj!DWFr@iR_I@LH*Nk*^vyPk!mV?f(1dee^c&7gF^< z`ngq-db(S-M(SwV@aRx9l38#hhMu;IDy+}_aj*$7#T`XUh}MM-N{tyh4p7WKiWFhe z$UtY3;+9zhO-q0i{)J3k71%A_9hoo6QRYERa^kPgDlPp0SdF+4zK2%09KMt)xhcbRxM__pkw6^r2<_LkMGxNe^BFH7WkUo{z;pWm$C zOb)rQq0yNcno|^Wf+EML*g^5hXs9XRULYDp4&5f(FyrRTOnMEII%n@3oeEvZOQFZ1 zp(}cuRr7Do%$}bMxW%V*D!ihX6ialR1i|C8Im&Ah7bV+4gytBpm2-6Z7U(T^GU%A^ z(cl(-j&fPRHU5RkI>lO1+1xB%@2qXpy16GK)(|^>1NNePq_yhtxzU539m-SJ3I1sr z`(s0laA6O~N~91}Be+$vWsDkW#rMMR@pI@clM-i+Sy%I6GixJ+hCMMhhlhKvUa}-o zxokk#3N`9zP%o)4%G_mfvLcK+nnj#OCfU7F2o8#|_RWspKijFWak5P&fbHm`zn_)8 z82`hazx=i(k;-MEVTBrny&69!-SVsaa*vM5$emLsG8MqJTn0Jp6%kM!fV{dqv6k-l zR3lRcm<`9Ce`Nd;-F&4H5{!GReS?0F^|n7Q#GJ4vjnL65*DJ6j7bSgD6&} zlLbD8OpKM09DDjlf13J+*n|(+KVNMiYuVw$b+>E;=rO|*OPLh2g(91%*n|niK08P! zzhO=neOrmLg}USQKin4eRqvRieO;y4hpls*qYxy3B!U)vZ1){3=P61T#JT zo*i^yOgcAhqB>vDEKUz7pbDsa9-F3rF|~vKAI$tgA!{}0!4yR3WY7ULSENppe*_7` zg$oy!dn}P06gRqKw_uZVdx`0WCO0|o5uZaKX0Jz^QhkA}lx8Zl@VPo@Q;z?w5Zst) zJhIJ=u?3fltr zdEW6~9cLd8gu%k0BB?Z*Nh3JwL_^yuzx&UhIVh0i9a*<8C3VGl=_OP-M7WS?(L$hu3D=W(~ZuTw7e@r!jNSU~@l5BNh5J6ShFc9sg7+|Mp zsn}l7idq8Y@ES#~dy(v1fT3aMq;75p{dPaMM)pPYQT}$x+xx})MLN|raRGfmQUEF# zEqtwVd(=Z-x@W_j)1*eULtF%IP6pHOB#~4H;KgPPu(^>HUZ|3XfuCFA#mcn&ZBeEu zN36aQePs4AahojJ9e83IWvk+{aDA8-%7gnovqbT}A1peoC=?#_)X55>`neeX#GX|7 zf}LJX@Jq)moO&m{W?qNfXbIT=PM`5zIAB@~Kk;XL4C~9Rrp0x0VOb4*sn2`V{;p3g zzXb>mI#rP@BQPT{hkJ9@KH2ArLE%H*2@v?nrXRo8srE!wk&fsSj_)%+0)$2d-Vw%e z^eI;0rRKhIhWicE8g%@vraR=A3$H;fW@}Ib#ne&cEET(s+=2CHyGjd$m|ebGf-(b> zIEyD;=Nt}7cWo>wIIK-i^L zmc+@Bd?Y~9{$RMCk98z?Es^<4V_=ixkuLuQzl&zujkzI)TuH!HNiEKS}AO= z11`GTZRV(8H3~Mfrp#g9y|lCKW@HaYjg1a&*i zKzNm6nke!qw5AtVk<96z1po<}Bm;;0_DlN!jy1t)@;uI$vJUxmXm^9xPte;u4^2mr zp$>XE5B65749Ri`-BdxPbT#u?03-@zI_4$-4&9|%$qf&P{rKkroVs}1RX04cJDwUij>7sgJt4|-%sj4~nZ0joz zU{%%5?dBff4!_ejSlBd2j>V2mA8FG~o$KZ(15DWY{hniM$&MEWJ5WkH3_B$h1M1uRl075gYy&vMc3y*?vA@(++|Q0ixV7<)+mlfmxm3w9pFBkd65jbZS=hqI^ecv zZu;Z_w{y_=e!%SrKT|y5c3ODPbHHu2|DE8>Ks|IjS?hlTQkUtI_j~_MeIlkKaB1lO zsB2#72!yqZ|B#0&5t*TD{k!NaCYK&?%VIi$aJ=5H4!BkO?w(Takv_GU)~QY^vUoY7 zPWfi2&)y-{O7a4lm@H2==Ep%u8MgF1<&Xtb*!fC(LlgIZ*MzHAKJ0#joN{4YwVUDU z62)AkNCPOiMyHVkX18}4)h#chjPEY;{#+WpT~*A*6Bl_UzQ_HZaU0&!pwfZz=IzkU z5lgB3Zyo(f=eJiZDE%28C4MdL`{fIYzuEcPm)}8cPMvD!f~(@Cp$EXG6?2Tk@Af_o z%>}jGVoohTecC>5|GYDb0k<;x8eh*psMzAoBG@>l-~!y)iZ6EHda39QFUw9BFG%*n zN?#0Pc}kKfF``j&YSOwuHOfuz_E;<5IsZ#pD)j29gT`c&S@!0#1Pd!AZkaT;F31<9L|>N`PU@Qe_St`5>v4>i!N;EB02Xr-=E6h`(&X=$jMLp^XkjV(;k1T6_nlMl=Qa;Ok)PLp@bQrO3WtWFkQ za1@YOEIm*CdxkDr2h#oDZuwcH2{c0Ym>p#ANY%A2oG(9X=KNGq4A`PWRP1rTWztmt zO!^Zal%)Smn#f>Hzgou>f_h@Uph2O={W@01qkzP9x%#~OHc{WyDq1tMCU}?Fpf{iD zua_W4Y|t%>w-{>vJD}eiZp~7aeN$7Fho;2)s*$!+>U+)?FAln)N|L?hw+@7mL#>l_ z2%UN3v#AqI5K`&iQ zQg+cC*G=G&ZD!D3Pcg|9Nu*-2l%?MbnSq%!G_l2ol-MZ)mn3^l0vr4o?zP#5upZgI z2e!ZO#<%R=#%AN=!l)w*NmbGio zRgX3+6>56eu^99`4oq3aQJ0Bt%a`(!ys$%CH&+i%eKA(kWMbyeJhrwpFIIztZwSUx0epnsT z?^!{^xs3!bZxC%)X3`1J<}?=+S1SzAIUpT$Tl<5Qs0w9w*3uhUuZbYeXV)W*DqXZlqt1)Cf*uIViaIF02 z@w19viGJQ?z|q~z;3P7r7nJL+0l2h~QxXlxWNQ2ig}dkuA+mh+PDlz`oga4KusV#x zFR%i~xL@?neu_lk7Fd6W%OK z+}Lva&w}e;`??7&F=hNVQtiTfovUW^Uq>;JnmG;gzX-LS4h3Eoc6dQ%Zb8zI)qA*E zocQ1z$nsvA-$moLrHHd?j=J~dtcX15JA!IIKxqR?Kw9Vx`3H;6Dj$eTeKiV@P0jyuj4nfHiuzA`ETQKr8CDaje2IGHwu4kWRy2fQ! zC@XoSd?<+0Df$fAxmNjNssu>Y36AFl*zon+gb5qQar7;$R;BT|zxqX!#AIiLox*oX z@(Z&n6`EP4EQ*0heFiqRF(hfwT;xLWW4_=9w^^yyhTNTUnqU(MjUqpwLb!2K!HhN; z3IpW}k_CMd8x4g$lW7y6p0FAAe4=g7=DtyI>sKa-#BW_Pja+315f_fsKQx2L=M-~? zA~%7T3)#Hwbh(E{Q9m`_H-itPwfi)>!u?)nJ=R5JV)ObgP>$*X5$4Kz^8o`9gkP7L93{hQN^ zEqB>2yHHu#H|GG6u}&HvToY6hS?!kxVY+4@HfpCp4;si5Z}q{mV_;N?s8&|(*A`UHbke(`%b5e9*L_5hLhx)WGa%a|dl=L%)jQEyv3>qW zTojy!0#-(Z^5|KrRUeDGVCDIOq#>~Z*>8!`pl<*J!Wa)T3$vkjjjY})e2!>rWiZ4is5W9l( zMKuXe%^Gm)2;I)$Yn))Y|EXF32bZV9rE7agmZiPJHGKRro}|$`@3RAWoVYEHf=LGXz`Y)HBL{CcKq=}ptyl-N6XjduYqTG&}{ zm8t_I7Y5xvC4HXI0SQLm8=}@ufr2p6RlYiIUTH`!xV!2*(ztnO!q$6Vh4G35KUSc; za|?fWa}_Y67`pHaE2tbA}?Pkc^CV3aA%{)FQ~(q40GzXOFBJLZ;9!)wg7_ zM$r-39M;0ex)|Up_j@Jt(&T8fZoQnrWYS3@3}?V5)jvm|m!BYpLahdc@yUasC!t}u zMo|~m2O6b2y(&P*G#9>LoWgFIF1ViF#0=v>x5|q-S2#^GOSn({-bebUkJ(&|^x;oA z_kvih_}0D4EFtxR&~L5$GLYi3N0toYF3|E518HEfl2I5j&u1BQkwv)cEMy=K~5MewFy&E-(Bn~zWg*KhD4Akyy<#E3xUg4x`Q5|%h6xwpuO)r5y7Kst< zoK)@_4~P-&5}?AoHl)bstauGX6B8p0eDE$Qh6pRd*7!c~-W|E<(dXId7L3_1`xipA?q)8bRWa|DOM)ObvGyfc(vftMAItP{$H2>PFB7!j*`}FG2TQm z;H<6(6)JRqT z8X;f`j5#&Sj4(38rj%LDpA|M41+-Os$+V=+*?Oqf^lxbSgBvVHLuo z!8_?(Wr+_y+Yr_Y3>J{kc(v&JH)DFHX9a5%`+XY1FhblfXcAlz56HJEkHJ#<5oz<* zsjfutAX<zQ87iT8dO{^UfC>2z72Qb1zG5(2V9A;tgL5a%v?BO4#ciuRV|Nd z?)8Q?#JdzFAlOhyZV6#a-|vMD&j;P|y!#}0v{?1=RCNRIfFzZ>N7f292-_5=kPaaV zI}FIGl=*^8x_l~#UluzN&HcXM$=-y_6A@&e!!K8frnAoYwB zPWmN0<5Qn~`HxHEwLdl)uG2ALKOprljIRXdb{K?OCk$qofy}yLwYXvpZf`dK<}=GueOHaLS}VTKQQoI7 z`ZxPvck`>kUqrX?_XlkZDU9s4mD*|T%NfOa8|QfxBir`~yJfZb7psEp>9FF$uDF#B zD@TJWgL?(T1RQRAx5~eOmB#+{h3&+-Dujg1ljve~#at{asgR*JP7z{jr_)=JmR z%MaKYaX28^9hI7oOeqb$OFKhn_322l5f@`v?qvLXZ%SIyd`*7J&(DoJPcqo~DK3nQ zgJyn8KE*(gC;R`i_bqTupXdJm#9v5W47m|ZQb3Ui2C?O0n5d0*y{z5V-MgKg|NduN zC)+vhY-in0=KkHfQI)DyxhQx+4FQ1&B7&$CHCz^bXakl*bM=Kbb*pZB?Zzf^3Ecf4$i=UL?`^Y409%a1=@wBz<@Xiu{JjfSN|tS;5j5D#(_gUf~{L5BuOO7uptw!D8hnm=1>r;%Kh6 zo)6snZ)UKxeednJNfR@$xbeQ|3k$GxP%>a(f)JCFHTkHRk>LkgV0lA4JAOV2?Sjo{yEd~iUZgJop@A9cDeeZ)>AjWRu z{jUhsCVI^j>gsoHzTOhCKLj;xmxN$M>6AQeQj@rrtNvJpM8NG5tP-m8R-X{<27+=t zk>TAVt%r=q`2cmc67zq0ug?_yv+&Qs1np|y!S_TLIL>PYwSuw*+R>rSCIp$Ev;oV~ z;Oq`-eIB>aQ7T2O2;A-jeMoJiqq$#_ku7U zdpS$JWzw3NhZLBSf;LUe$wLlT?}feG&e$;9_@EgEm2FP0tzk!fj*;Ixeoo4pIo3cP*o)?jg4<6;LG`(i#+>AhBSdf3Pg z-_5CyOpbZTE)(nuiza43zHH^Cl#nC1!9D%;1-S4N>h0&qKbxaZV z&4?|uynFt?;(2B-`Mu(ae<7=F@$$NHlm(BWVB0rj| z?(nbU_mc7%O(M-?i1#-O*9WRQr58o}B$={&o4%&igpdAkB#;^XyFPma`;hsr?$uvf z*CD&@%44TM8e5u7CFxkbi3IPdQf%th(bb_TL8%jMZ8M(znT_}k$3ZaSw;0ib_Ca97 zZjAT*eQuMrSUQssog442?MSO*MW2yTv4p(`w38lrr$9~c7-Ppcu?_)fL=YI=L$9?= z`VXmjDcO)5dXcQ3K=xQz=Nw7~@lP!k+XiLAsF`>-x(t%7cQ|GI3OYIFGKl8gp|=FJ zM7MHpc;4Y`;=l*Z&7rr`x57@6#Gv{q8Qym|DAwTOkKLh9cc)>9uRMSdN=_e0i9$2zKAK#^rGz{1qAE3D7Rk#F&J z_!6&nI0-E6M57`}cL-MQJ@Q6o zt9H7MpC{-gr+o%Ika9Iux-%+OTIAR4l_y9LV93{H)^fCHaTYqP$2t8{(@|^cAtq>X zU(;z%@FMOTSA`ku3S|Xd!$m@wV=>2KZakshu>OEtchW^zzCs&U%FFML? zC4V80Js2&gZtSpbwpeu5P_mCHvK$D})W}S>oxTf9WLW#@GBaTBBR8IR01v}H@2ZQN z7yIt;Hshi6^>5aZRki{jZfps1EoOZ)CEG~BXC1qQx-R_yBWLOK1Cl|~LTdO8;v3#s zf#u}#%yx03s#vmh%IPT={Z8@63=j@rz_sp!9T3?!06(s5YyF=korvU#I=Yc%^e>k1 z(!q03cX_5tuf!ajl@nPRXmTzvZ=41C=##s4XW+*7SR>IB=!zb9Pczs zD=*;JtIh`Ky*}h=yKDvsdhobgVzsY~CdH!lIV?{j9yIIKU(!7Q_=K~Ha)fM#mxk&!4Ze%sg zDF|uD3*V@Pq-;HzPB)@hY^oBvD z1D6;l9ik0Xj$>SNGFqycUUE$0UNsxU4r zPY6Af$}UmIEI1o>P`UZ#*w=So1BXC-n%RS%yKO&bkc1M<v(b7Q+> zrwIUEtR_(<-OF!-lJ1MZ4Si|eBVmVcGap;A$H3#Zm{W(bV)2yg5A0**JDUup)OmKuunzzmJM;0&UpW!`w%M_L<#5kW$%(IwV+;KtLyF=qQ8LhntfykJWf}W6fMfpR zg5yvjd1xARF4N~?GNxXO1ok#4ri|`mpZ7gJyG^Q}vOxw)k(g0Le!Uve4KLJd8t~Ze zQ%8f-neDfU=)<*srtx$$n#0>A2ZH-O@Mp22A5xL&A(^6nkJbLy{rWv>Jg>y`NsmdI z`TBYN9>*jfPCBM&iAWDoS8#MQaNmdfC*+(~m+=#Wz7TbTPLD%2%~2>iuq|@<7DrHg zhg}~&ni>Z;UZw2RIHX4I5f;WY5gIq8e@|5Xi*Oz%#g@^$NGstzKPWh05~*jq*f+LlAIeI)>ik49E-}&snbx z|*uN2W zzfB^n3R9!>A3DKh;R%8P4`fNerX&o z8=L=phd*kYPrL4<{g2v)N8!FD0`KklAR^4{>fHQJ$8M7A#^K{?3+JVjlI^2NF>=jn zLqGGZ5kgiHd3p}B>;2<+*)tksr?JD7Hpy0?mY+txf@+gq35FuO$a0c2BWXq!2wfjg zsF#Lm1vQ=ttSlvVobCzMPq`~n<4ctJItQ|UoqVKbuurYR5N8g9+J=#^d5Yot+q<0^ zsDC)g)5ki?$)uRdeZLjNPEHiFS1YBPLGt`#UkvdhyrBP8Kjl0SPGRKLNZMnn--Ii; zF5QgM3T{T_IS)H_=5FMWvNNEQPG8&pf7e>GvAXRB+YU`^+gn%W>%H3EYMGC`-X-it zm6ml?ng#4l3E_|Y%gI^)8g8Qssbh8Y%CSg*8|Q}cj5}#YpEV34BSO_3f9G!t%nQ)0 z#phayp4nd8jg3f?#d39;lAWMPH5D7rsr60bfO&|U1uExLn;|ij3wiE7@6?Ivd{Lbz zCVN_eW3H1%5_;5D*&34*X}YmZn&G{Fnnqq?%u1@0qf6m3`o@raHzo=&pAIb)XsBB0 zZeeO9QjfvrxdnJUm+;ynYQ-kH2V?4}6tbcTtRCRLXYLX+MxV`f>zOB?1-7TIbH4pY z-DoO&-FO!TbvZ*~H_4PNfg-D@SZqD2m#eXJAa3q4i5BX!kipPt&dO6CvE366V#4n8 zuDTFGC;rx&)%WTBvm09yJKVjU;y!Ri?$Di(@ikJg02K$h!Co!fGEM7=H8UvlXQB|) z$dNN1djb$0tzZo-9{a%bjQSCXJnoEABRf6lKN6&MK^m8#;~Hf$(t~S2ThcVN`X&oK zo*K`@pk-q4hQ1J$g(FQ%xgQE?IBZ!Q1kW=qhNCw)3O4)88U5(bT`!rJuJ`n+N|Nu! zdnZ_Xh8T~7lnlB{_EE7FqUD@!dg*Idg$1+EDv>fcC1^9KZft!s^ZVKbE#GNWApvpx z^h47O+(U{U_C!=vDhru@MuX(hsfJ`inu&=R?Cx((K0Yd1wKj zPD*x*BJH^KJ|!=osPWt3pT??@Xyi4LV!k?6d_&p|I+33UGn6SocnS(le6xWiO7bx| z0mXj+j2Gb`etI@&ttSXFSn4H$C4ys;&5`OPPu+|(*;!>exF$D*n}A~asCN$J7W#bp ze6Ze1BQN7O$jU=grR@=z5Xgz_BDukB@a7dt*l%?L|MjJLhuLi`BQ4b~b~C@vd+^Sj z$ZX~95RH6$Bx?H@%emT>*ZoWnmxW)Nw^G{YtrZk`TEo=&cyrW9S>0y^L`HZUf3#uo zP3taPCfjv4PCVFYCf3Iso~%=ve%1pYUZZL+c|i0rne=iUtCP%zsp-WF?~xQYHcY!M z3{w^*%bQy%1_Axyw8}t8KLYRE4R$Ad|Fx{*Xf#^=9@uS;_w{?!ilJ{Z zuuO22oO6AMd!!RT{Ire!Dl>cX1CRRKGt4k~|F3Na$@Z_z&i$wbs6L@&dnvM;iX9{X z@1%DqssydZLeMOwj^4q;MX4t`gXE9`SnTe4XMiFGCdl^^B>Tof%p%q)R!anCQnHl= zQTIWssf>-Sy1E%yHCjVs^&cprITw5*1)9~eV>}!^yq=7}pLl$$b<3sO1_kXjQJVPH zQMe9#hT^ennS1gn;;U!Ru;u+@b;QV&j3fTP`1$i5B-XN8xsT>{2HfFR`nE@Dl&co> z%{>aPM6TaO;5qE3FVOe6SSzFFtO<&9eL=5zqMwZqtMB?40fZ4gsNG#@9nU+>VH~eI zq#pJ=xF0HEt5F)_20LzUsMwp;_KRov<>++Yc$lljZ70*iPXVN{AD*pslWniP~0)J`PTC#$* zO_(2@EigFO%XDC(9YPUW69>j}<}JGyMSc-z-oreWom@#YUzt73L5n?10VRXv_I6-_ z00AQq2}%j7gX+3of@{nQUX=i9e6^9O(sEKIfLubY7%T5uAr`zopk(eI2-&sL`~2}f zI9)P7r`BA+Yfs3dN>cH5Q1Yxo@3ya1$IU(#!*cjO&wvIBTcDoP&f8AzDKswKp&dc z3@uOkDNA@7WjCEf@*_Lx#XwmDnr+2V?eth!hzatxbS(VtbTg(r1j-VU z%?wj+9GE?9fvLTe48nE0sMunhOcyINk#|QwQ0}M%5X(zTW!+ z-@wloRZDU_H;TI9&UR%gJBwaIkdflTyU0j^`K9ZEweXx4iX4(dOzkctJD)S6qLUWr&{MMg6e)pb6tFK?4)Lni1U z6kM(3R?w>_w#lz?bNw1sxN^l$KNF%CIhdU-c6#IKP-NmX*pDLKj>8E8Yu zSe8>c3M7-Ia6Nzq-{IG{BO0wUK5iRXv6I6|jl4Upm5UPO*>sv{g>w0XRuVUPO=OK^ z1*mS<$Qo5pF2L&ZZew-wZ$#XS7`&>HuZh__w^3CiuBPw7zPex1DaC7T^i^(+q>O(r zq5-O~;&?sM%QKy7d6us)?ktO?pZLnGfNJP`be@R$&ku0(jI9%RzBtdk^TY7(Ra zvtjqF_rm8s^1cZ{-&|ogJhvutO^m60PVcpHBJRUmxu+zzUOEz;5Q=F*6dkSw@1%_G z=B7>@6lueT0v+ATjStPH+iB#T@1#d0Q{YTc*@haPWezh&pc>a*|6ZK6)PURCBRk>& zrgjn(g;p0!TIpg5%8NVIfH*h{4ne}fv7C7uGe~?eFn)(MNf47juN#+t+o2S~N|sCW zxZ8YSC6krb@OOb@H3yZn-4c1a$@Y^Sjxw^4;Z zZwj@-HIZk08|Ji2+Qu@A#0&!VCd~Qg?c0?9_DOsEyeAK_q7B>XQAym`otQ4brk^qS z(p%1zWAktM%vc=RpMM)W;&DwUe!bC}J<4s}3_DCwm!ab$+b>gu0eBRSNEhg*XynMT zP&9cHIiW0?Jht#~3>n7NVJrag<$Ewf#(3@ezVBM6mE9H@b`r%&Gdeg}kO(vxsnS|; zpARxsX8Re52hq0}%fN>t^=wZ(hA$^=3|=t9#V-%POpm5B-isY_IZZD|YGcgnVcZOhBh%i!Hy?YMHwZG7$A69*XId8SZBb;xVfoc_^@7c4Hr&?? z|6xwiVr#GCX)Xje4ujeu!9^Af5cS5L3>uM-S&xLRG&&|OhN}%z;2?r+Jn7m~_7Qxc zYX4_)^OB-E`SXJ$!;P1ea*HLUh?0S0v4e`;<9A4yE4?bL^SJC7azCGj^m@&%JR~h_ZM+LhBar z!k8<;RqQlyQ>xgl-1Wg)VX!r|p_KuihO9(PkTJ4*lZ(H9-#YtZ zC&{sIPXF{}KK+v97!+V?RP{&;6`P>c_dMMd(86z~&rJH?uP9Ceqpzs@e!~*dOI{V2 zZAS)oeG5sjmB@GBtr*F-un-zb22=wZO`?jpk7*L+Lgo6p@YV=CIxVjdAyZ^NW%_T? zqHU2UB2&cJ2QwzH9*Kb%*;k&~Q})sPyG2uLC^JAF`#e4aJTJs;({r#o3|TG`DA_8C zte|3ZSvk;WmrkSp#u`~%*qBW9(-ZLQ$$$Epu{dLY@@?!{#&s|J)=STO=(I<2A!v`+ zGB0&x&eTiFEY>NYiB zA`vhWe}WVbCiL+T^Y|M15lM|?P^?=c$2A2@#?*Ib?SJv^tdCTt6br`DTxDY%M#~Tf z)8eIzpu-3)<6r*$3z@$&uPU$XJQ6@|O&}{Qgl+pM*#nB)rDBn~Vb|Ow&)P|uB8(N3 z%-tulb*^63uBsw?gKwxhL)s)q19$pn2pXWbF`dr#+Z?3zyX(JnRtc(D z7+NnkM6Hnlb7(dlA9|6Gy%{}HHOlT{A^^07)<7pkEA)G2`vO+w^>Xb*S2PEX+PIl4 zozqQR{*lmA*869-b%O0|>`^`$=#o@mnHp3JQ>R{o#qL?g(I&-4z5$OT(fRYo5Hdq& z{MqmsdiU79j+njhAJ6;t2IeLJE#e0}6zg?e|$O6e+TiL(gE9i5VA zPxPRB>*$EixsQ>Z7=PD!IHS%St_e#2Gyol?Y>QmHLs1ZgtW}_w(_A}zpAV*xh zEiwV5woLu8$n0@Qu^GZ6he8jAA^|Uu206>rJ8y!_6F2w?kMcL833A_+`-LMqu66$v z6Q9J5_cwN^9*+reOJgj+`*4!hTutqY4azeC<`}e}IgHb9=m_)sy06)q{IGKC1+wld zv*#_cFpAqK8C17zrD6|*Bp0q$Tc?61*k0K07W?1#OqHSr>$;dj^EC4GpafRxL{zj~ z8)6Xkg`NviW5-`Qo#XOckYgyZu_`$o7nTF6Uz;L{9z;pW4lOC3Kx!>M-ouovf+Bh< z_Wp|oqhceohu4i43sHWRkheEOX6$1@TqDt272=oLXe#U|udO=#LM|zlT z5bY2iP~40<%)S}5&wrU%<71>Ax=i|fP%;#sEmQ0jsN14;PDj>OsBd2GwT#m!-v?xz z^`oVDAMwX;V@9k`S$?3*j=m!6&wp&tmXC)0$c@(@d+bM{xzkeFW&8^G{e`I88wIs> zGZF-NtdX}#j|q>su)56#$e1A9#^WwK^;V|zZ*Q9o%qxd`eo9U-TZY^>s7!8J*vU(j z>;gsVfh$&aMs!nX`k6v*c-O0{UtRq8%`g7--uyprqSD{Fz5pp0kFkzQkZJgUB0ivu ze=G)Hm+|rbMpeet-K-Mn0Y&b#rr-*?SA@T}`+On?1y#;&7E(9Hfp$g?JCA)Zq{<(! z=0_i+ce9QX9OWpvJFSvc#@{e~7pFq9dJ-f*f^N;fDt!OC)DVS6xYo1!#k`Y0JdjqOgi1-RNMSqnw3P_eDtWj@Ei!mRh+ zBdnm+-7yAHmaZ5@P9Oz(I7P9a#OW75W9^6fr zu@YF&GX@;9mLXi<}>lXXL?^(%@|CyHyu_q;yj0WZ@Glxis_ zr-B*-(buxO`yX?dS!2@`r>*6T7YD0GX^LZX9U=5uPDV^;SxjMv)Cv?DCNFBq1CPOcs4RDd{P8tuJfrXYxC%vc;IBa?pcd)=C?m;2 z7HI)0YjuJE^X{Oj?75p2$Lkia^L-#*1pl*P{+Of53S3S(d>WzlZBzPBZ}wVyw9nQ4 z$By4xBkp70ouw5(|MCvr8lPlw92B?Y@%|Y}X9q#?^sKq~jD1j~dcW|7wHE0o7T@PQ zN~WVo78QF-(k20Un(t=5QXjs4?irHA>7;jspT7-0SmJZ>C<#PCD7 z?0!m zlsjRkxvqkeLvXRX6uQcOFy(8w^pop}*0~(F4Y}D#GF^#D@xv>#hV$n`}my+F~$Zgmkvun6{ zL1p}O`l_%GL}JxDy-stJ*;gcO@-C7aoX+cGHGB1Yv_`hg>i0M;zXzLRjpVxDfCuLM zbA|dTdzH(1YWQDOWL4xtZkLxy+g6SHUyZz1gem<3{`mmI+>G)z@H~cT-fp% z$3f_6C9tDoJM~I&75MA5Vm0#7;@DTA#O0m9v!k#Z&K^1UnhDm-VE2nj z(F)aZ15|9S7?m4$@RB&Spt#pFw6}A~ z1ZZN$4?9oL<(UxPBRM8XU>z|QQ)=Xkuq@UT{8)r7??}hh>{Sr8#nTY8g2G@%pU>WT z39Ln^Cz>3M1Lg^Ibe~VXyeGPJ4v1loc0F;~RE^S*Fq$g&P2+a_%Y=nNW(?&I{A3@| zP9Ol8As8y8WWctS2PBYhUIqRv)aApvvht}|k9G%St2az<^DCm0IGaLixYxK>{P1eO zM;kdxuMal8t_fZHGQPsv7JLF*=PLv8)o9USa}u2a8Jh!-@+2e3xbMDO^XtYxTbJ?L zseeU*B-9{>LMJIiRnf{A_b-POwPGE#C7BG5OF&?Iu9!i?_86`>%Lp3oTk7=3s_(9Q z-kV&PfuCapDRwhv(<@`35(PP4O#aNM*E445VSra=1CR1_x=0)ZQr>Qr{ zIXCu?!5tp5`dz1F*C^7ACAG^ZT#sl0#pZNSvdyL|L3CyLgy!)pCfp>As%<1Sa_QXq zs4j3!OJ+Wr+ZnK4P#&_~rxIFUTENWVl|~gxFMy7=>N>yIyFrCl@O%Zh(pWWBtVrOr zMD(%iqc$l|%5IBwGah-zldX!SVVCBm(_0l9Io{EzIuhE-MTdCk^LwVHD=I~O>|>JR z=t@>X_!3SA8~+MC`)u>d0A<@KpZ9n$MquK2%fgV5u$$h%(ubdU zF)pmi|1oEUv^d(peaz|i*fnQ;K&?D=YOUSlvo} zT-j;sN|~oK)b8u219AUQ#C%x~b5-~mca1FG|9*5eefrlI7G%pw$HG4eKcZwyDH1nEC0+JskJynr-o-uz{aj}jsPhES8PC=P^-J2MyXR4u37U7^qIDQFhaf{&%B0IP1cV4S%>thUz$o0=GgZQ#ZXkB#D@-=`hV? zq(ptlJ`XJ(_e0i92QZzGWU63)-5sNYLb(K1GRS3Nm>Em-ke|ljq*gel;BjD=;}jGz z0?)X`sS8uBm4u#lNZi*l5x3N>Dm5@#VKHcbq&`Yu&>2|uAzdr_J zGCJU%To#6fna4@Dut}6A#{Wa1B{00?vzC%f+T;e;ihIdvAJ@oaIt52gL5}4h_HlIi z?mP5rW`i@ol-o`!ZI$!7Z$uGFcZWK$wUq1(MNU$&T2?d37bVLEJhZHgsjZOrTL*eV z>p~3N!|WU=tHsP=6;wok4)?^3K9{{N(CXE`T4ScGYqH);7kC(Gtnhdhv^lHs1>#8W zg|8B&iPC)5fX5DArtBv4qTsQ5TjWjFC3d%W9&d-DNT}9(-GngIbg7uIve&eMdls6!7*cM#-{HdBOkf(B9>)=pY@b%9YvC`)3$sz zkjJ#shZO08a=u3X)Cs2ZX!(s~c*Fz*42K@o?ZbfJ>6K^vAO2R|BsBwLLvrXvvfhn> z0Tg*d;wd?l40`poP)xs?! zQ2aBx0gUtaZ1@a6&91syTKum+v=-QP+fB0_<=+a?UcuhrR_=Ab`Y7Zk>jguRHLsEt zKX;2~UQkMqfg49H4a*0oq>XhqdZTDBDfUkb%9^*Zj!p@3rOm)ed-lXXWi3o6Ts7n7 zahchT9s5sjw=Upj(!A)#ey+x1IZC2rYbf$D75jw25)0uRjnoN`A1Wb;86 zpG+(JyXJY%`ZaKOPQ{|e>#Pb%h2*S1awxWtKK5qmeSV{=o`nUNczzeV8Nza@Qarzh zUC3@%ZI)Kk?UE$VerW}?&fn)((~a_Geucy}5OD+pcFf4&X_ugZnGx|V_}XG0vzPjA zGg(NI-PlXrWnmUFC>hv^%~WhW2Rlm5)Li@gp*l85eU@Fv*8~A&7Xca7jH#=5n`RX7 zjh(ETAY|oE3993}-gt}{BQ8P4lR+6_WZYXk{lRcEMpmYF7m}UKDC68X2zSx~QF=-S zT_q({Yzt6stcD<4jZtKt5Ve);*+QGXN)+EbiS4b*ZTF(O3`G5+^F}fD&l`_0P zm$!kQo^IRb6b_`2T%!@iq?mIYTRyZmm7ixo9aP?B&BE|`YUo<e~6??&j{~Fzt{ZJBkz=d{f~{R zWt<&CWbOD&ku?o#T9$CC!6@B->4qvBepWzzT}sez6RF>-f189qlBwU`f2n&~n{;Ey zO>zRd8nANiw>fx8hd;KPuJ=d=uP@G4J_JsO!NYp-lS@zD6x84%Yr>B~>pA`z!&@*H zfzYPOvO4)R@`jwrdA+#pEiq2LaF4BtB63v_Mx;q1qVZL z=GF&Pu@A&FhVX zdbZ7K=cb8PD68pH@_5oGt%$iX^Dz|A*0Z!AUx!@W72IlCH?0;N=_*beOD8f!^?K(7 zH3IQ`qbf&fEVPYUOU^25#Jhaf!MU>V)&8cx1>ZqK3fPxtUY>blc7}JUk5yN`7H_2L80~z zx(L3NIxkd&7~YZ(`M8|<(h*2F^ZgthixDK;cb(XmtbAiMSq3+b7TVF^OBZAd)g_ZZ zW~pl@RY8k?lIP;ZPE%)c6g>_Hn;jp}Nq2_+gcodG(d)LmGCS3~rBRtd=x8*FfFZ0n zy4DxA?8ZyVS?bogPIS+ehBteG*D8S!2LQZcAHNT>$=|(o{rArL7`rr0Q7#kGfTrxz`@YCz~3`!CgS zZ$%s;m7Fa>#fr`=6CpF5Oz#v-^K(u00`m0YU|v>&f2UtaHa=&| z%07!_WjiIyrbs3g+c!6bT?y<4hq%R2U-(ynM|w?|COXcm3;&Q)sK};^qbdXALkEJ} zRoTjZNvCuv3$Jbm601)BURDv^P1=|S;+ajB1AwsMlbTL3dnaah+h1QikLCT%x(%Mm z`r^jhUOP?mND93(DocriYsMbOVt%^dAq@>_8fBiKh}9%J650~pN+0zcs{jh)jqg}2 zDn@`M%W~;yWAqY61mKn_X}Cw2sBh@%_t3+pTxIOiYR!hu(*2C^CvSpiLR@sAsKR zfaO#@Y@|%wC&?VUPH7{|7~tCxSUT;VZ)#`1DluD=^vVDEUnJF*TjItmO0k9M%cf*N z`H&86Lp1VT0!0&l2`^m$JC%dJ_r$Ax4RnoXjkt7v8dN)}i}`Uptc?R^mPT2<+z_<} zS)OQp)CrD@y;lZkcw*elubejZHU_e9?bub{t*|Cqw!>$fAgC4Na;?TX*hl`^^b%ei zFD0m*>*5d9#=pnm=(Vk9U3t#_2W?|nao=y>?-PD|a*Ek{DA%kkCtII0>v6!sdhDcR zc@(5(K%B@xm#{wv+o4fj39j`Ap`pZ}Zoekc-Dow|sjO%BM7K%qiI?&g_eP^=5tb}8 zNkKBRR8Yg$_+*mpJ`k8yJ`6KV(Sb66Ymh%0EApjTaRJKhgIhIWm*L;ccxn6I+i#O5 zH!k)6!U7u|lnj*P4ODEm_|9ufc-`Ecft_@g@c8Uv$x2_{4E>Y@){>B733?cZxDDcN zF0+D{}p1zBH0Y(y_?;?h2{ zY?;mU|NHl!j3#~T#xZ_7>Elu$i|-{BbcbkV%!fR!JYLu#YE<2s*~(qPDvs93uS;*u zs*Q2M_RZ#O;YmhFvO9|NFZ_U=5^241xo!H`&MLLl535|;WDi-UMBpM+ct(2Xi~J1l z{ZN^mU{tSHe>^cA?i)~mr<=;Iopu_8CDh1D2wpKRX1IPKRf~)0qZlE;uBmW70vOqn z@$=qTIqi9;%*`N=gqcbm@T63Me&LG91Xi>E32}yCMdYrzE2XP^ZcaTx_j_!PYzgm` z=1t!n((bRmDXjI^$Ul!x3Az?}L4>^~=^>p^Z?MYi;}`cTF3lM5$cZ!!qLHtNRG$yr z!E5D$s03sflco10ACX>?4mF)^fpGgRFZV4Xo0t_mx$%kxTxCN-l)01)DC)OSvFn4624?%W&FZEX(?#Yvh|JJqpcXyHJFBw8$D39;3LJ zSheJ;wUo3C4jAILtC}5oY4l2=F%c=>A$&vY!&4`YrNi>&j5spyU*0^1FYH?s!Rx!~ ztxN9RcJ}S`0(|1#2x~{G^oh^%EX(hrkN)|aej8h$@hKUsfBB(#K@s@>ZXBs$wxGCg z=m|8khAt@AC|NT_E>W>{v=(~tJ3-7Oo!&5g2?r!iVm^-sB~N)NkXWkYW+5jg_#}es ze!0L}s8brHo>K%p!eXebO%SM41m{UBeTd`=vGYv5Ly<4Ss@qm*T!Z)y9#)Y!UKL3f zsO#lPoNT25_Va6G-B9tc&%Y*Iqr{TkT%+Q&dX>)+CCXN0DOdaMVryC8uZz04*pm(K zOzS2ugYGwEt1FH&z`5bgoFjl7&P^G`wPC+v^6KS*16*m*5%-Xi z^-|;>#sCw5Yp{<7${?ji{*g~Jf9KEC*r0}C?fB50vu}!XS?y7}z?~4h$z)~2{MQe8 z09y>xfC8}$u<$qnTjdTZ)L1M5vBV}3fCt4VI>iZ-AO|5(*ZL*_Qu`&Jcw+bUXP&K~ z2dLh~-s+b@)^d!k1qeJH{pm~U7HKa@p0p53{Ygohi22aBOXKCxHzgEB0OqLS!5}}Mqp;sK<|HXjH8|C6fObxb2cuX`Qz|yvu97 z_wcLv$o@fgpJFltb_a361x7RAzR?T8iGSx?M~k1{YrF5?50axpt;v^pPVDJN4osAa zOcR|XM?!0$s~%mKld?W^>?zE=h z&C>2^jjA>&N&{aAN?_s4Z)71GL%kS_y!$zcL8s(tlS(FPS@{rGen5(WAn!;hcJmLO zJHrE!Qm$5T405CROhUNnsl8si*bhL8tJs!9a8tji0|!Ov`1(aJ*; zSfzr#k=xQyL&qkbo}3&hvg||0#>zzP$-X3xGQ)}QIe!}|oB-rJL%-1#luS>N{Zwoo z#NV+$V;ftI*`RdLfh~mqJUD@ZPX2&UT?47)bQ(_;!^uq9Dt~o(Sf;EoTF0xSZ-;;r z(J4;itd^_+CU@MX)l?0-?I@_)ucBf3Q^po9-=IuRx-=jkAk~qh}xhUb}QZCTjW<44iG=; zX%Ov_oOu!U;8+ZV6dwEvLk!lGu$Sigv`8iuu?5n%})Qzx2(H`RRhiEkrltx?l19 zL@4?C^woXy)8DE7Nh_q!(*>^bf1aK>hERHX02iHEK2P?a#B61DduUIS% zwufb8HmEJ3SB^|DBSU3)Pfb2_V`QXTAS0QQB~WA)6$@coztd4!(H8^Sp(vt4QowJQ zWI7<{Yy0yX%~^*qF`73R;j@e|F}~XKME4vsCcN(*OC=?4jEM#dOjJ{{DvDGBCd5m_ zE&w%0x3p2!{_{Xa8i!H=5@sK=}honcPsN2W|8D;v3p;)fmrZ^S1f4y_NiWBOJA>_RTrM#;0zm z8vd>0RmL=imQ(8xlD>2>n7y!kCuT>CkVO5UxTMayWAeF*&)DgUy#B3vz}DZdON-~< z13JaKqRx<`fj!ab!AZP>vjzZJ<$lJD)Wm96^yJx|F{X&IJ&NGRp+}%Gdxy{zzdkL`2TvBG zJ+0gvNFKP{yI~5TL8Ie*$XJfp@b3*7vccu(;r&Zw1G8EQH-^bxi^WSv$)IU6gNiMG z=c=$?wNhH;e~`O`+=1fp237w2>#&q{lWW2|^b(Q}YZG2U&8-cP*m%s@#OjQx7If2V zNs@0qaV=OziWESM<((r^8keTl1i5zNnQVFn&4c;3lc(Bh zIQxr?t+?%~ZxqtsGh37u|8?*lIqk*{Zij`9Xr^S47r20pv`r$E0$$GRkg6LY3T)(n z_Nk$DGcNLrShp7>vW?7yI*2qE&F|sn3EJV6F%E~}G}KwcTBoC)SZ>niy;ssjZ>J9^ zieSyz6WAlgqO1Evy_{E0uxv#mUnZ`F$`q899lU)B?~+e73xdw6QfyN&2}Bm~F%Paj zG_9Hc$Qy~l%lQTTbm-E9Mu%qFQ2>mZ8N?;omKfc zIqt^zYPZ1GMM`#_BIl^s8#8;s;9!}PdOKt#y8LP+Wqjm=zdH?oYUHh6t2qt+M+32# z4Y-4hS30NGf3w_Jtd~UgzX&ybrE`|a zpo|Yxt`UbuFxRuNgsMV$#J7cI;2P_(F!PD19Pl_1vu{qld@tD(ST8;l+DhUk?+J7S zSpOt!J&mNN&oIN*Cz~d7tmC}T)t1UmG#3>uNt(CJ}jjD|kx+Sm30^YYzkN38eTUy&V^ot^lSi4AkGvF0EG{D(DL&(SLKKMxGzt$37Z(VGQGe%y8pcmv|@=#g|hy_ z+Mqb&lsq9EU+L)Uk`lj*(j%c~$q~_j2R@Lc+zaV!tTVhj%OoQi$14oRyVdBAr?Dz% z>_6zF9|te-spXc4`aRb1PLhWt8=5#bO*+#5u~-ifCrMu!dsJXy zkF=Bw`YqF_Sd@^=gW~YT56LNzmehOIdA5RK&?zrKx5&MiWntBlMZHkc+$~Iv#I3%` zREri7JIjQ>lJ=Vo{ZX4{gQx$mOm1S@5eY0l%;O|CN6?s4!Uiov9P&l?+ zi~>6LEsCu;;1|cvyeTkSnGEjw7LqUl=sAaak{U|3nIaph*h30~w3-;ZIq-!R908NM z3{VlKItPARxY^B{7^2AR09hU7b!mI%{vMs6=w{p(}+~8j3*YMTYN!OsD*z+dP8~$+Nbk)P67-=FgW6dI0Y#74TueQWL!_w0{!#O6VeX! ziC1EhJnQB4knlrWF-D${^?W%33F~LZ;+PR6#s`1AxbS&bV$6krRdF=7Gos)$I?Bjw zR|aj2I=Yh82|vriPs_EQc=dl}1Q_9C+?hW-IP|=$?U{%YP-1+wFNRf@MII-|$zAWe z-UhBV5_*I>=~Gk6{nRT$TPLcUNv#-8UMBVOV)C(SFDajzCwSxy%L;bG7Dt~Uccbrm zBixpGf69f$D_dDZUsM(;bE7!J)}r4G7BN&3zVNgWF%3lP6g@4PP#p!N!sU~z`{^`tr)}m z*-E2u$>(275cEjbO>FiW@Bs2vus-b(C&^wRXqsSH&BQD7=}SoIGT>1we*99e2+26( zc%6J?bGXk+3DN}BO^NqUW)wnHvKTBn*aR8B-`(GDQ-o z*wwNc2~vDutpBj`oalT&u@rdn{yAz}4%?+sTN-8qFtSVIS2dja=`1rW-aR*aCn=gh z&RTr>m6WWE0{(GsV_GV0i32JTrXbXWY&R&s68PUfyIK^Zkm1 z_UuH1aqYT0tB!v{qOOHjlnn14X*QG*=PLI`<-m4+)OPL}(_?*^0gS94b?Nb{h3|w| z6DqoG!qEG&mW3RjDAQu3u-CG1t6#~&x5Ic4 zc(qf?!rOrxFE@BEsgYC(8~q#RG>6{-!U%A{`+VBzINoNc`h-IaChjE$I)3^I5t1U} zpdFlb;qlWqg{rYG`Fud(3&s-n9g2deCJ~eq*R%S3`h2pLC`_pLLRhO?>92;PJbv<( z4{rLKSF@j;9oIn8+;}zn#9}qeqhwGjlLaJsk=V|5Q`{>7YAe-o4iH;`CRIKwL2!oO z2*i1trVT{sy~?LHhhx^dMYIjZ!H+pDuZ4c>%|ZR* zi{Ovz*bm@h`s{ku?l&L9)gr%acIlkUtYSrt*o8P@R?2A&73X5Zax`X0nK7Fjv(7nu zMZ6t5Oy)-Zq1_=XrLf?nN(;jZ0=i>LTniP;2^u_d%+w6tJ$Aoekf;43z`W8J=0#j3 znuyy0n{92VpzE9FETqBDM(?ZJ`9SKJQBnrV0!o_|^lB zjJg@?fQqbRRw2-p)r!l)SNrdW@&+}=f0vR>W3sB92HLS5k?r)3$ZWa-%EIwK*8zBR zON}erOBg|T+=Iep$E|5)m@GB!8)+g97UuCw5*iZ|Uis_Sqk11PIU! zmiwxEB=;f?vyscrsAGWT6xe%qMdau@qA|SR$oiXD! zCkwW422nXN?uT&agwWNKVF_poaxRUwBfL9S|O`p5^@q}E8~3Q9`y-C-l1LS`Q^^9g4*{z&g;WN5~#{wk@q z?sK!#Aci%(Mu*Up_bU%=;a>*{1~pbGU>i^?7u(0i=+nnfhuI8_^z&r?9J3j4TW{sk zxc@VkGW)iVeIB17{T^=nw~{Rut4abTTSbu-&?zFg641lWn7Se~pUS7s&va4^9aBs= zfK;YKI^}+5NZs(l_CNTT0r74#SxAzZH5$2b?e{JV_+(Hr*br|9;*cr5Bqiv&ILWh; zwM&x3`IPiQ7)29=t@hW1E#yi}zvLqS0?Ty%CKUQ!U?qf(g$Z;3QBQjkPQlFZgT(Hq zZ%j60zufE6T$pURP4 zWn@yu)h^z>|3_xH{Ho~JeWdm|6W;pV0zxg6>mgk!$TE=SjGEoTS z@ro};r>hK5gYv_y4 z1|aIEG(&+Fj%ThRM+dl8jLbqQLFd@<{vE*UngDmhwq8y zq=O}<9sVbYZVqOR@V{qb)fu?6f}ul)&qAYp-2LIFFMMv@x#qS3emgC7TF+%1=of&E zF}&!Jl!XJyyYJm8P?!^G5LGDK#RdEXPM=RJ_oh%QMDnU<&f~K_Plv9L01OU|JKCG< zgTYo^UBl-SzZgvwgBu51>{Kxvi7qq}M;(bipj2l}ZROUAckr&TZi{s@@~59BW!$lW z&)_sOfYIP1WAghu`tP$w)4Arxz_8P~mM7@)Yz+O7hYpWg3#A1*dIwzuYXijKBsaw+ zGaXnVoH;cNkm1VU?Nk4x*D29^)+WQwcdb24H{(pedJd-QuL(1hDM8CPEu=<#KoK9X zNtwV}{J-Bd{CF9eZ1-&mosx5^YBa23Zfr8_u!`xS5$UGi$FspViG&)br0Lscb<(Z! zThK&MW>js0`lg#K)E4O-GhQ(!X2a-xKJ^86&56JN?mshx=IxH;AFtMvL^s|Y@3h$3 zq*F48v2CJav!H~n+^EQEEM2P*74ix}YOEI)x81DRS@AH_W=a+S;m1R$Fw)UigitWMF1$5DM}IDEc4bkT<%f?) zOn~w9)2=(i2pHowNhZmsnoY;D_X1XvbT>8~pyM~B`DYg;gKcdt6^jYoB+n*j4$D9e zxxHkK3_8T1^`tlYhWE<3aLOpnrv+&<4YY6-vM)oF1lwP=kx)|$uW-VS*PSt3PNQ?Xsr zwPcUJKDr{8k z294h?zde#(VF`PWBtP1~-6Q#wd`h4N)ajq(umd_a{b)Oe9meSw?OTxd_Wt+Jrwv+* zpNT&rnQpwkR9JlfcT+OR-{n)W*iTVPe-xy~riSZ&>n65yak)`n^ZStZg(O#sOGqgY zkg2OfQ-W><=)Kf1N3Gme1Z2 z38AnxkqIol7m7olm{c3EKLFB!rp!w6q&cVvQWP5SacfpAFyOhB*aM`}~twHFT+RwZB?72sQlOQ`^U&qH+Et&EHJ!|k|j}O4Her< z_DS>{eR#U?ldwm^`>YkRME0h+7lgh4gl&w2aCH=APP>m0T&W`UpY{b0qGDH7BH8A~ zsHn6+MKL7<7iA|EyE&+2?rKhAK({0Z>)4I2S5GVs)O*GICra^EddQYZRh&lECXy+t z6gAAL5?+bvCCD^cBSw~=KK4nndTymCnbR=m0*vw4|I?Q~0=XgQEFMFr)xt8RFu-uI-eO`7A`8qugSaJMQRb3WuGkqa!tE7$z4 z#oUKJc8whKmPJy%`99~P$Hr8T1W-1Har&uMg}^;&Qr2; z6se(N^}y7#CT8^AuYdPfXO-~pL{jxG^jn6z#3fCo0}q)L(405?_|IaX1^ zCVS2qAr6=()A%4IE>_1K1AfE5FaG;F-zL!^g`P9u(H@1Wdyb=QBv5&lAsXho!;oTR zd&b{+teRoHueP&QO$&jE@rw`5rm+u0ZDJk9WL!SAo_#~Q*5{GX7%_j6lqsL>V5y(O z|2R80j&v^U4Cl1djON096Y{UUG;!6-X6y5_PwO_5{cdc18ZE5P2})K?kt!X$v{zgK49?{57(E zNlH+1H1Ii1`waHZ*pI44DrOvTT{t$YV%=+_yw68RpD;5PBI~l=D_hwJc52dS|2!*0fJw+9N?f=Z>_D zeU_jHeat^Gd}-JQ7kI|QO&$Xr4L6L_uMPZ*->y)znc3Z0w1nuMvl#CY3)53V$@WmB zh>FD)_B;XlE%DRWg&S+{4~8NYdmgk5e>`&saHKZM7A@gz^UH}!0P5vMS;}Nmx}b-< zsE%L9-^ShsDeQxw|GOv~O7CmMEk^$a93r|V6rNh;xmH;`vnLv!bKH3ucJ3J9!_t1( zJsS4y9KLAZ&ip1*c{I?BoX4`0D~ZMya@^Q*9JD}A0VUf(k?mCMBjI`e`k-yFFKSfn zrBBN%qiR|0$|TRM;JC2!qEf|9b}_p=v{8lktzyN6b&3si8Ar=17kvigwOjofRi-D} zB~Wl3((S6a&gcQ-Fa{XiXdf{8==XD;cNWTP6MIRc>U=;IyUw>x0sP7Ak7sTP=%NpW z9ucGg-K}0$r|k7@RCREU`ktJ3o}U+308L~i>_$~@U{CZp`Q2&ts%HQDFd+ku<1M{x2KOeW-cmdfingLuW zwTY@Ei}Rz4S(_spfo}VRGEcBS1Tr95$^(is!foJAW1F-y0Ol-ASu9&lfR%krb$!rL zU`@;O6-iLq;fpJL<*IOLSn2{yMW(QO3D(W*kGUr?a4Q4Lr~YLLuPElzmlCD>V=jko zn0}V0^QvUwPrX+K&|O|34`LEr>l}{;Lr#E--68%cIeh&$W{>9R+qZ*An;UPrmsm&u z+@)l9C~}*Mtrh1)EYHS!N0k=(=)JL$ic(TSieo~7L9UoWrYt8q9Wkn_4v zin)0&E!qG*sFy(@@PML--3INkx5aJJEuIe})J9(M6hY~nY`Ronn4$qT(=r>uXevF|_6R2=Ukup@Bxe8@A_(HhVi*w{ z#!hedy}fP!{%z;BmG;)Q(>v|W*v|B}h^V-*DYzRzHbF!IMNvUuL{LWs2NgwJkj-&$ z6ciB^|Mw+PNrodiknjh4Z|{uBS-yJ@ z0~Wmd%33ADEm^hjmSgSLvLd4CQ;xlh4*#O>D;?j<|MEg>tWZO~-lH)&xKs@}s1**f zFIz*oLQVg(V%m7uKzbIb9tc|}s|wurPSxFj%)xyb7V|aA`0D7m1Q>1__hd z8v#B{?zhz!PWYG@mhW}K93P)|mF;*-4mdO$v_ae9UPD*Qj*x3Eu`auWJ6w)TvxcIPWwXf{GnNfI>Z{l>(|IZ2AERks zwBN=|=sy{ZoGlcyk%IXdELMz+94LX+Hn(4y7>Xta_e5A7hOW<885)>fuDBUBAS@i< z*M3p=MMn4c%?leoB&+P${RP(DL6%}O#cZHR64r>9iDIF+dth~Gl61N}btqBwD>F2a z!WLCG|Np~mbDeGrCu{VVv73n@T#t#F((<>=tE|S=(!x@C)eM#3DfA<{Q_?o~Mqo8i z@We?|=yK&UVAa{dE0e@~B@1^euR~;4uZHCYvcWPZK@0q^YYL_Yvb}eF2-azPXw_Ba4QI0AdW3;-A1R!Q}X_K$)0D150rLX#!>mo}OPVNBMz8nm z1(nYyp!1kDhpHmY(oXMHp1UBbPyFZ;chL&bf2DRq1DJ&$Pk z^xl#31KgxgEerzQ8dtsMG`ROIdrE)*L=V(Nkw)I^d zS;7rvc5I-aqiGPz5-BE*BCDy0Tdujl+JOCSgZJAno4c1hzEPdB2ot0Fg^}Gn?RQ^E z4>4lm2m5!XlA>{zzOFNw`41@u+9pd4LKOJsb_v>*7bLMkks#x^&%1wm?$j&p)%4D( zJE!(LWQ()K5S#)NRY3nSM2k~bJg@#$yQJyLP3i&(jpo&OK6+0OKG2X zuk*9f&WUlfbK-O2zn&Afi$@JFi?A{RGchuRyts>f{aVnt?1KeH6g}pr{fv}3lFKF& zbDClx0a!yt)fI1aciMrWmk zM0qXq%;6jTmsap!UK|3~B*w8oSxpbzgU2gn1=4||?OHT#k>W$4(Si%UXI=M^2gGb- zi8-FJUXj@U7znoJ;dhGIUV z$Z;y-Ge~JSXb<`Bl9dMRaH;Vv4NjA;C4J5%RIYcYN1C*Rx+j0gbb1^k=)3Njf7l}~ z><-->hV~Q{(lWun-YeTHYn%H(b!l#z^J-EBrJy_JE8J^z94B=#<<=H)T^&}cPiR7p9B{A9vE3Q#aJN9avZ(5k{hDF7@5$o zG9qe2qTfZb{+02kp?Y+XNz0;`bc&=>5lQ@H)k*P5F&5`spzA#bIx#vGD;1cejuHcx z1d4(}Qw^xsCwUDtDVK|?>HU*e@sPV2Z{uKhUd)SypGE~M3x8uImqq_$x_&eX2m4*8 z*%A&pVsOBq@+OrE$Qah2jllX`kqIc=BS{8mnlwkyWw6RY)d?t&J{;=nBHu| zZ@>P()%VSNFh3t1wByP?mQKr0oX^cWr7j1VfE20j`G*uK(kO0*j6}Li0q=%;HpsX_ z2OOQ~i5#0_5+URW1kQfRNU_4c`jmeB&Np0*E6ooowp}3WUYV5!WbFs7H0cxrv}fC> zh;@^igag^D64@Hv+KxTk?I!b{L@|jJiK8M8g|>nlTCYxk2=IMnI{!e}r=n)j z-*2d*sQb#XTPq9?+~SG0Y<{eE6}zDrYLmZn=aVz$>SyC>ePNMmyE3%`a+KN5P~Hs$ z?&31ZIcRm8xrT;GB!w6o{BS~!!5=Z1VOxFqX>UcM zuW?10nexv>l4Zv$%2AUQrI=!hD6$9ktU$h-p+T|$-L`34-dyH)$MZ6Qa&Pa$-Yu%q zDJRLPDT`K30z*Q71_k5ED!tbxkV`F{l1(FlBB;^mHQ2ztTN*2^lwOble^!!Lg5dqd z|5zdh{s%)hiXN)Lt<(ZJ0w@>&!TxWnpo86r46)`+oc^bKubC5VaxpA+yxU=+<=n_C zfhw3A?)TdmD$CfH08LCyl7pd{G>}4L zZOuO41OAy%Pl29nQ3#g*A}h3&jga*qW));TAEc=yv2=#f+Birx`$?uF`OxI0-bXQo z6xmHhZ02q9s}(KL)~Sz7OP7^NYC?0p$|Pl?`%2sy`R^bb8^!eXI7wh3Yn z_1Y`4T16De0EVA)TI7quAZe}guAswp7r%nH*{6rj=IwAv46yR^=34t^<2U*F^X3>p zw_8HI_&(n4KNxXyb6)%ma&xF0wjG<46(&tg4=JXHBHdKPWp6CJ>zUm(9SiCVCIOTe zT@3yEkAjbo&e=&CsP&}t!n;MQ=B7aYx(9q<16xj#7v6=!MtU`lf$jWBUfG~tkvwV9 zBJ6^?{c9D zFVCi%WDWe}Nd=Rzy=l1@H2l)Zvs0ufkCIJe0s>dkB(KY$H5$v@@;EXRe-1f_F|Ug) zo$%IK&oZ&UL)_22EYo!2`^}~1?f+aBIy(;SvJeBL3b4Q6kQ%d1&B8S4J$baCS%k+5 zdHeaL!a`oExJB8>bcY-hZ(w?zvCtgb?0QJMsEcILJ+s%k94B$o3PBn_)xAP(ZDVUR zZ=;9O3p;7ePl>lKT{drX`}}U)j$y&l`qnHeopxAK=DUhPYJiUgZO%JGH%RYz-s5*F zj{k7W!p5&=|7g`0^WFBjF7=NTU-qkjvYJ%hQeIp@>9n1pOL(hQHUh%v@nFsA|J+#Y ze)Y9I4+Qg#%Ub=9qccf~9dEuanD}YODdrdjz8kbGrTFLY?gH`rUGTQ9idsd{-f=US1LNA|F>y#jfA~*)n@&JWIw7_s( zk<@TqGo34i{oLd;q?+EqTLL9qostg4lBp-4-~XztNP25jN<5>tBQIHSK8FU&53u8A zi2WF)-&64)=3IMr+g!^+0(kYU_1`WI-zhuf+u%{kH}Jh~qbh!=S6}{l>93l;R~)`- zLC=>@GMAuh8;L|xpyI1ZVjaGRVdiCIAQ@&*E1c#Ak{{OmqT)+NgA-mNY9|$TY;dlc z7@RtaIZXk@NJOHbXlfj96FD}uSybqb)#Y^oY0d|yXVOPWqM!vRu1cVe^OM=}UU`BC zq01c4cr>dM#4XHnM<5w?ER|$2OP!X3Daz1bLrmVl7QENFS%e%Bdi65LM?WopYsH&6 z-}x*YuQCLOGT^#e`Ge4+kZcY7k3L40Ic8CJ7#vlvu2tkpN3yhy3{x*LHbb2n0mist zs_^#@i_G`*c58T8f(@x4yHl^d<6bLT4q*hW{!Jn^^6be)A*I46vo4T3%1nB{G>%sR znA^p#l~ky2&AP<&K(OHn?Aq{(INlxiNdKcADBAPl?m`}R?T!Q+FPk>2;pSyOGJBI3OsdJKfty66PgV<2^qTGQz!U{Z#Kt5=tVI9cICo1n|el*^@whvFu11`K0`Z8$*og9qfT?SFN zzb*d9d+#s)d#4W`x;t>_132^`ypVMI*m~Z%PWSVfxAOFw568)-QEnH!|7*M##@v~6 zXU+y@1CtfBU$K;2m8C*~OgZeS^Y|NP*ct`qPv7D=?dNW^@c~Y-nCSk~#NU~R;<&_a z?RTvtEI~R03tEDpMx83SOp1YctOT+L7ghNvA87bc6w;zPGWXm(-8rUoKB^aF(_N&I z(Jdw0Aw^ym&~D2L?h%`t0lQ;6a49=y9y71MWGP&%ge^i|_zhKa_F{f1CS34YKL3p9b6sK{0;ElqD2iQ8cSSdfIm^N=-)icy0GwjU#hU%8!kS+bp%>8{I%lkeYG z7QwkXIJcR1$v;(wh4ogSs1eS~^Bu&q57u6NI4w5zOL_6z1)sT?FE*dgBipfy%F+N} za8YsLxj;5DI{YxeybBc~AgyI}l)fBYmSgzk<5>GFJBGjghqAo_W88G+rtj5|pcgYRAMvgSB7^h$V}}mQ zfnnpFi5arkfwAJhf2;~Q1bre% zc9=c0&$&X-;1R8`qL<6jt~~RRSpPIL4E_r$r%lV|6-Za5KAK!eZKuua}or*1q}<~L5yR@{I&eD`CT+h z&y@N00hioSV3ErY+2jW*5F4pI-|r5`@{wgTvA8H-QbK2ro-NLTnHv#AEEspRmvDo~ zpMR3=6KZr)He@SPN%1Q~f>3Ybz8s?%s6;BKB3fx<<6@m6Syk_fK6$Ay(G8>`(2YqA zY64+fRj|cJottt#piVgjQ=ruzPNiEcq}p06xt3Q;nvQ z#D4vFHaeaj`VMo3oYolh$aBSuuwuHxO|MxqdyBf4fbz(_Ine(qIq!MPV}z50yn5wXPV2mHE3QL9->8ws8A6uW6u4$& zz6B|wKCn_%IlV!P zJaEmTBuy5O5Mq(jW51TEPrUW&QbA8hn-=fi9abvDKeOeff>A@`#bgZrGlakrG_R6=MF7K6Q=Q~hqbvhL$wS`T<=CCtpuH%mqg!}MUOgez13Iw4!~Ax`_u{D9 zHRy?(>NAtI>N>>$%WfkTfy!uD{00JydJVd@FyVm@LLoEV z=W>#yO-to9i$3Jv@$3xL0W+8j61t$GY$tfPc(4Kp7ci*ZfRSvWu0U8I)MY};Ml$Tj zs(sRZqoB!S;O0iv9Va|kMjw^642aU=Pc^HP`PU@%0e48N=nUBmk-+~hwt zYv{Uwx`0?0bf?m#rE}rG3fILgo_FEOZJ}8~FT0Kr_zU7P>}Q>9<>Xij)Q?c_IJDYM`mps}K2Jo|Oi0 z1sZ4(iqLk*dfku~I$F@8ie)1GYeD81ur#|$&>QkHjJ#x;o}aQAj-LO*$Uc1G^`^f# z7;%*M@OK}O$Z?iq(3_wofns7QvI;lpt@7(6D=6KiQNCZUYm~24*jPB2<@sAQ@htb+ z@-Tapr}(G8OIvJ&#HAlDn?vrsGNxs%38hE$L}tlE(oaPo4^4wMig^Icbyrn~d{NTz zG${@3n^O>aD)?z&77b@L9di-{kHb1WPm(GrNE%GvIz0v|zH_Cw_{$tCr(6J|v_{-2 zU*?FTJ(fk0Nbw0lj(cy&Mb)aX&Ef)n_w?nCD;%$@?()i=ql01`H;db~STb}~wSU%< zkn|wr$GPR7s%}-Ld*`^XnG`9m4}KE35~#p$z*W&8&v1M4%Fsxl)LTB#GFceDO9+i` z%%`#%dP~?v)iTGsf|c_URlCEm399wj6BotsWw>!id!F4`{dMQRf7x8b&2BRkEP`!Fnw2)K zSybTO13Rtjs#c|5eUgcCY0wUYdO_Q!S#%dXl4cs~Hy^u>))%pGHn@Yq!huJ589Nw; zN?|RTF}Y)waW#AMdd?b>%WXBY<66BM6W8Y;#q6iZ-T~LgXXA6;yB;4yNd@kHT2#lT zmQKl)#!1oQB&)D?uTzmO!}4<6DIMXJLom850O_nc=_ZI4w)$Xw1}c-n&P-bD5h-p_ zJ(Oc+!m649+@Usv5qf@Tl=vESyYI&8<1!<7l>dCIp2Twlj~(YEcA0=Dg<`f+WD^yE zRc3mx!kJYeaekkQx@POpBsJ5Ryh0R2iVb?i10>+jv?a&F9*qbggJ#lJAM+=LA9pb# z0e58pXg?YYP=|0i03b<_&5R?wgv+Upc9p53H;4yrQWk^V()M zXivcDR6z!PSn}*-B=zj9je#;Ej0{)}TcN`+?%y8pSDRP=+O3bmQt68Zz@Q&<*I8Gt zjAm~8;Oe03R$>ILfOpKv#*=G0MO>!Z^iJD;z-<7# zCI2L_nhZ*6#!<{_io{S6`N2IQjq)7eYliJkqx_iD{})aWi#nxk3zVTp_g;8sfq9W0 zmyDAgJF6@ueYXT#14@A3d(#_*{vcT0GrMHoKEYM56uMo5|G*&#-Kn0;TWsVmRh`1i zRw_;fAcC4jPlE0W4v0a91Vg>eqG;!vqHO2&0ZWu?nbH93f{TmIdG3p{d>DHAM|LzP zeB$pX8Lddie>VLOS!u@vVc8~jVH3r`Lbwjtf`sLQA_$AL%Au2?o(JT!s-?3!ox4LG zOt}@7FS+fr9B8=5AOP}W2Dr`ki~DT-iQ5A4gP$jag&48%Oi1_@F(`E=NO~?YoW#$j zQ-d(b2CW6nflZQT(KdJ8W7kfos)G+$Ooyd(WdWPzYeAN%fdQ5Uya*i}OjYBRPaW|7 zW*Ya*w_b~x@zkM!*T`IiWKgE2fjJmFFbeLJy7_yj-kXlMXVa~scGx_R$_wI#ITq8x za(Fbj{PVB=`-ZusqTM>MEFy_WuCdN3#;-+rDD)`51~@Z5^{obq?lSs8Xr<^VBny!t z8wm#6{hMSTsW-cA@jA+XBB*z5an-9owR-V_*?f$A>dgkS@>fpghWfuzpS|-(qq$kI ztICsn^2+RnV@y&{sZ%e_-s z6mlKJB`SbYISRP7+aT9`d-6RGR4k9?#e)h8Ubj2!gldCyix;M6)(HAQ+_h8E<$f(_ zpEN5tC%9Pn0LYsWo;{%zf=!xdBi9F}3d`wUc?DEgtR-o_e|s-4{IRT5SVSN4jePUH zJzsh(dvSZwtTbsM|Au>~r}^#Cp!@i2fUUkShS=(R_UrF$^*!;$7sCu#w~<4Y#WY#r zFrxm=uYYsOzZP*2?xwo=oixVebXh?k^1J!iLASy+o30J0m|hAV6AGJP&H%%`X_}{V z&S+{Cxc)`+nuOh|!*dobvaXYCVtz)66qB*6`DOd@yXx6SGxct%wu0onGG^+8iC1)( zVoE4dOhudqvebL@T2iMjS1y~G2j&F7VJ;b%yHzWpq^eaO?RmhdLs8^g?0e5S+Ot7h zqg>*atU?(sRGU8Lyq|v}v`Y}fs|&am)C|T5<(c-8XwO5wX}*u4?6`9JBYx==D>5dH z5Cs;Wt1 zNO#D979>b;4fOEhsW?F#5A{SUed`o@HC_g_fNCfRjAf2FwaQaKJ@Q}B(oFrysuQ#k zDm~x#I!U4&L9u&qXlyOT#8YGq6_E;ZE1B5WU+jSywh95fb;?gvSwV|OwWk;f=q-Km zMl$wj?&0K&+HWk(>zDHc`#^!b9ZGjE6YRPzol-glw8wp`rC8}SW(JQV21eb2WGQsm4V^|U zC-h(nOdb{KOePW@s=hd&aL%)g1y#4Y0{L zMRJ^?*L&bjPC0$&EGFg z6kL?-RqpW7K~!G(P;qm1oOA~i!ewf3tAZ8*aOO9J+ zZa9H!qQ}QGsdbVC>Oc!f7h(cf0egpD+8)kBkQB`=_WclaIN%EcW|1BX>>A z)eVZdMv*2e0-0M+YaYvtJRy+JR&18v608eY5rk>TEE;whCAu{JS;<{lFS)JAo_SCM6th?^QK!h`U6nbImPxuy(4OW!`u?tZH|(~L}t^QrnhSh^Mb1OsKZ#V zMs^r9iF#GDTG!`X56pR4Ag-7zPN5&sV__fRk&kTF+2+c6$2pnEi2<(lht1h}?G_^} zOuXx-fzj~LtB*NnL6uaBe;gQu8}1l7)Z0q5mwmb{g$4TvwmHKM7G4=WS?1}WaTRf4 zN%{aKx!!9NL|C$CCIMEKh(G9e!2MKJ2=1u<-tTa3UQfvJx!6FBRsR124Ce4h3cX9% zOY~l`K}Y<$mHU*N-PXYOZY3^F^@b+C-YAY8wSq?G_926@;jt+}anv>hHu)^K;=nlC9waj(A z4q3Yh$Zvq}t&Z0=ceT7hduhI2-2yr_sB3vbSw`RXxFRc=mq~BYl!-2?*10xlmpXr> zenc|q>!c6L%`!k@*+ve_=%2I|AMc={IK9sH>zefV=eau??TohdT^(7%&Cb|y5mSnZ zok^saIEt)Bk}?z+D58t#JRyi@G>a1XdarU@3MURijmbyLplx!?n0#5$_;34uZ^Xn0 z@iDKH4m);Nmzxx1_fpJ#igbY_ji4POAu-?6t2-5aDo{niSZgmiq&_^eZI&)$ekNV% zn?~*WMVA7%xNAJB1iCWc?F=4D6)p*?Qy}qNHE^k7aqU21E#`Z&>Fk-C<$bDpMWr7` z=I)3(>D!8fKF=>|53Ex}`sd2x-B!&lC(g3Xqx=RwRSG zg#mbz zRkF%%`Fl9o$qDkyC%i{fyluzkh(-0)fPq60>u;1-`j!fGS@d#}MJEtljtc4krMs00 z0zg8aYO^R=mCi>}7+rxobUI==WwQuMFGGrgZK(TdnUP}ma%^R>{`!XGb@P%+yWKXi z)OsTA?lt#q?)7SvsmP{#2++DK_s_DCEA_Ieuxy%NHiB)=aWWqh(_M~5zhhjv4!rH6 zAh#UJN|OcR5ykXU!C+Ec)m%{wu6NQjq+5f?nG!O-40GgDRdK}0?|RAgFbn|XX##NB;En* zk0sJ)C1)k+{B(YS@(i$jVTJT^#afkaABkmj_%o2ygC(F;h#5Kv1H$i=xs2&wK89O^3|5D**qMemZ2PTM0SmF*izpe4Y+)R?(b&|w9t7Nz9 z6iWsvQ?t$e4!sA$Df{I|r&OxCreD-z$78c7N8GMR4BgMytM#)grY5T{_?~sm5yyKK z3-#)U@}zl9LIb&cxx8Q57<@-{MV>FYB3?5y2Ph3&V>|vjyX;A zW}gf?$9oCygeG0K2RgGWTw7H6be30N*m78wJCz-CHmT1@TYOvQG-z}A*GY_P*PO@U z7kEwdby)lGyy4t6_e0)AAr(A4a^=m(GSJn5cvmto0dMjvR253o_|57vYskb|8*&B! zjuvneCQ$ixd(=h{X-~d=n55b9h8|EbC_G$1F(Bx*lZvSFi{qUKS7wXXhi@*7vFyKovWgov+(f2@OS#?dme^sc89-XXv@ALz2n~Ra5NkT zKy#DheMea;{QKfF;qTOWw$9$n)G9{5%-IG7qt9M(Gd$JF>wYCMVrJ!*YriG2uZ-c@ zX@anA6tkHk8>oonpuMvCxjS4+=I!-97{0i4!4;4)&f<56G-y8oHoJb5B_$ z!>zbbz5M;=j;5E7XRWh~Dy!y?Mq1Q<-FU(xEeb@bl6*;@OBw9&AfSlJ&Cwf+5ln|A zh#0}JqrT<0ZCyY8*XoBZ6=Fa(2V|6x6H!+~&XcEey7{{F1(5%qnaIzc+$dk|qSJ>J z`qwFtWmm6$9M!N(zr zEVE;WB;91WNT!&z6p5!IwmL1DT&&#A3~=D4&?l5z#CmmsaI0&hyiVOKx7nNy{YWjJ z>7nnp)lp7P$He=;{^Rf8GU8&_blzQ3ZpXL)&7(n%YAwZ_qR0s<0vlX66fxS@U-#eu!Bn4!S79)O2h5cXBtM`$00a%Wi0coe5;>DNdpPQE# z(#PK%cF$*};$DcZfoC{?@k2ukE#A8r*?^G4rPsMGAd@y|jUIQ>ZFW0P?mBn4@A2(c z-t`?&kX#|YrQ${*U^%2rjmAj6ae*HvNKN>kul?+V`39P0=lP6wX)TE&MUr?G@}lJO zTU3WdW%N1biJ(o}qD1~A{9Nw0RNSKM^u*80wY?$hJ=+J4w`uV%i0ckIc7fML=S^Pb zkug0+-02xPxk1|r+8`g?7)a;D3zNJ?Z!2sxS6rrL$ZI*xmHoyl`htI^HLvZp+aMlG z(Q)6@O4$m*71>=-gsInF`Yve309N--U#=~Xl?O+Y?Q|J$^gtXE1w$X?g4v-*u+33! zm@PXUS!d4EV7Jg<;bgcb%9xLQbLGkm=Tc#T^a9!Cil38J1xzY)U3GyR0ew9@GIjvr zf(}@5Yy*WMM&*W%ul9(A?nZ~E{+Gpz$fj2&iUKMlgWR7SiUGTmPDO0->Z0$+5Bt=+ zJ|g>kH_q&JJ_W&6Y(A-VN+6F&n&v13Z*%4!6`;KJ2I(qop?~fSFk3!{H`d}!W2Qt*nOs3B+ z=RKkIT0DQ2KBqXRwGC`WWiAZH!&(-?WE@U_q5in@XMW~@`FwMu{d(YR+UBWHC-Z=` z@Rm6qksraz7twOZeurg_hea3YXP?)(K4emOnc`>P<0qeyp7*@(jNemyib84?U34bs z4D>5IRkp#<2pgvdFFhu6Z?!M%Ao>m8x7M6Xh=sKun?VglPltR@%T{SCrMmXOI37wl zf9x9Th6T3ClWZ8}{QT44_RfC(ep?-3f80a#GukiV#q2Xnwwn{dk1J!sLJps)Nq5N) zhBTk_9O#Q@&}*2F)O&oFK==l!>FTxZ{@u#U-r1VcDGl0m0zBWd&uPo$d2|x5T>ksD z)N<-?zqgV77yu+UR1Y>D@Yz<7hH=?Z?b67b^<(4q=h9D$9+7%Gj^}onY-n02<~l{L zP!YAFPh|)EQw7(&@A0wH8d87Bs%l97WzTFD9q_-e><+Y?4Qh7mG(%xazPB7?O&>@e)Zb5x31f?u3LbmuD5g z<(CN%eE47q_!6pv7;Z(5#0E$)pgv<+z$tYfG{WgR6gj+WGW4SJ-Mab69fM5@k6m}Y zltyt>kTOFU=dUk;Bs-*rGL-k;KASh*3oma0TlVOzmB5x3 zW#}2O5EMh-Y{R3RKrwN4!cF(_ZhH(><8`8P8*E7j?g7k{fhCQ~>X!3+P+^xL=3Pe`uslR_QYNHwJo=Ftv5V9ZS*=2H&Nl zNS|{&U$+bR_?qb`@oJEcjR6U3cKBKdGtOqk!W&qD$_-}HuQgpUSDqeMz7UHVbvAv5 zhVBKjlcY(vc{MUic{}J1MU3Cd&mnny6UxWjHac{ezyUHrKK%HgH|P13{KIWQ2k_ZPHwm%dznG(v10mw2Xzp*?a+}q z;ujNkUD8RTxG+j{bt%??(9+B7#;9kJJB9|Elu`dY>j2%WW4~n11j0hNkxe(r`W+fV z`W^22 z>12-`@1ahc?4imjrj#NF4RYd_{kq7_*_ix6)wLQ>1!)#_(T`yHILhB2c5`;Nuw%{| zH@xDIZ|VF7Z6Ci%37aeMhBx`uigxoqlAu`DLk}FG!6RBxDnJ6@F2y}QUezor0wtS1 zkl%X(=Z^|Xt>ZUaitWJ$Vz|%NfIKH`Qvq|lR*puG&5oBK7Cp9}*ITDx>wK$clIDyj z5;o_B+zV-%-U78!x@NlEul@A|0pjM4`yFMoDBT66tLp;Vl_m4A4Yvz=GxK=R2Z|b5 zHv+3^%vI^tDPEc48aj@b#RsKYW`$3re}*@7$YC%5oBs9cbeA;uv-Aaeg=;te4$1Lu zQJr?_qEE|?Ow+56Oxxn6!{(N3`j*QHjb5{`%Dveo(*Ln);TqQt|E|znFYN1&*IRg%esR(lBUJj?ZF8shk$JCs812{3 z&P+H*HgijX+i{`z0TXMNOEC~2&!8f(-Zx6z=UlEi<=?DMo&?RmWiXx1qVrR41eF0< zPA@@92VJzH7f8pk;tw|d7`cfO7x|xn9$wEX*hCfjpOD+k4H~^&vxXJ>{1|KR%@n#C z!SaKOZ5PNo#~~JLJ5Eypm(`%Eh;)j9*8go(#723h4^k+BaGKK&x@-0s>GC-kKW@;j zB{5FrV4WV3UV@}bXL#Mv>Xsr*o3?N}wXjH1W%Ekn5o6)=(`VHq>}y>3$NzO{x)CKk zKT|9x+wIsTDmOtyA;kdea~>6en<0!2@8<6pw~AJYBgIPvdHg)NE?2zhoD((+)T{I4 z`@^6-nCz2XPy$uHs0V^9T|}3~obhP#Z1Sv9c7Pe#<@O;TEX-|1gSttAcN_knJ?_GJ zl12=WAw40^KhJstJ3xl8855n~*dI3DgTze%dMKr7l%u}AZne`|&3VuBuvA0`J(oc- ze6Zr=vpxNhfr4!`>mBC=iV4&2#-)w-MwCJb!CnixtEvs?q8oUd1U(^ob!13^L_aIZ z>zGX|>4us$i*Yg3%{Dy72^ZA6oge-D>&6wP$nEjlq?()4YR5aKR+IJU0>#u*q>hT{ zhGb%;Z!eU|v_o0|+fi=$*D6~29HAbR#@qlLXCQyAQ6~z@~B}goW(dC za*-PO!Qmjha8G!Yd0(~NR_?KMN@E7RjBe)FAQ@S)M`B1DAK8sS!UutpA}y12$&!OA zfPx75lVXI!PS$V_&7$)om~o579qmP&EYHLk`qpWg(efz&`BpuN=Vp2Ac)PsI#PXz2 z%vOqQ0^(mNmT2Q&m4OApUP{Osq|hHha4^a1BZ;n7v@3Yo3<%mJI5kR|B%LlvUM=b( zo5QKUw?>@(ToG3sG&Sz zyA5>{mTU(mHwM~l(Yzeew9=@m)J$V@_1Kt}L33g$DhAy$CLdh{GamhG@i7>RBL;9$ zr)ZSpmU+z4fKzJ;BE|#)WEdIEel#a7lTBFH&af)Jqo~)XRSnu7U zJT|qP9}k^It@8E!tFu~^sj_RLT1gRIAWaoF47A|pO0$9npMT`!@IE!cnhO~f=eBP1GM zzwj#)e9oe6120~6ibmilIvbe%6r?q`N(efye!>n25$prJz3SQ z$`!X7L^u!oEr;HhD4@Kom>wmr^X!0Junx5uOQ+;Yffoq0YqRJh5VFt}g&1@ln?*gd zH))Uw-WJ}bsqb9Ey^4|X74Ng`W@;# zu6ge#&%WjIqj`qUMIp^l-7;nZ6x;za%*h7`|( z!tPm%`kd1PZi-Q@h+T2lO3!(SVJw5h5I5TR5GP1X{8q<5v&MUV4><+@UPj+QiDx8p z!ivt5U+D6)Sse)b3|aq)#f0BvghkIVI;aj(&3=+;$CUyfnz)?%C;9NrUx7H)j_>N~B5hh^H%_cUl(jbyV=wAy<0PCs~D_=w`R8 z&{6i{!Z=<%q-dH!ln-?b&~c3yo+RlmR!5elxOwhuuzoO`_)&i4gqw+On)!bk@9Rq{ z@9>;U^WR^*=8Gufb}vN#hj&Pe9yVYh3QZOq*RDKfGb^lRB^m) zp+SnXUd@e5lI0-}f0!FQ24FJ$pjNno6FesJdf)6NMtBrDq@E%x97(3hENq~dB#I4#jh*U2rHJsx}CK<=z3K^gol;^aw#>by2DXbm3&V?KfjfLXA@6CP#=ZWCa) zDbJ*X2TBA+WX#<3y&AH7s3eUYmlI`}z<52ytfNQ*mbW&M^$T=Oq$s#z>biiIu%`}~ zyuBvtR<>rv!nQE zMQFuLBQW0opye>x!7a3I$B;Q@0+|C81AU~0Ald1c35_#YF|=wDcINGwCZDZ_taP={0dr4vhsiVFOfefMl0-$c%I>~?$0`1`e8`Lz z&up8ElxxdC+;0186;taKb<8Cu-m7iy2Uo|itCZtYwc1k~ZXotob%>Om{dTB)_~wPk zuNdvhv0q>QlpN(|SM2w1Mw(49UQaP~6gf>roKV!xOOx($eRr2@iEOF(iQv!AjwGu- zrq}axK)q}k(>AL?i$vYYLA9dNfV_}i=RK3)E?Hh^s;oh~#p~J4NIrPjXNjgkd)K*( zel}8tM~+v6_PBVX9HKK{y$W}rip<{@FQHzD$gKC==Xz=WW7&b9<9@seD8NxprtE8+ z3=+q;!^FUF{W9DeG5EaVf};PMMp)gPQ}G!&&J9*}yj-=L!0IB!oTtdg13k^6+krcR zmU&#_Z&9@Zt^ZEoZaUzS0^|2DMH%0vptq_+8GrI9TD1_#KcHh1?E& z>|7V#$AA2W?#V1Xlq zCqA44VQi!2`Bf8HND{eO9y?yppi^*AGie&dKqheu6ihp(NpsYXJ$AX43Jbs5{PpIa zwY;@+K{D0(y$5eaeYZHg1mbzvaM~GIA@8IULq2xjMOw&M;EE%n3=5LAVDpl3{;fvK zpYseOAJ5%I3hX#McgAGaD=4OnB8RAm+X}QH*jfYpIcWQVfsjs>0$Zh^ND@OCic1X4 z0Ko4Rx+M&`3%dDxr{0@x_zpEls4tB+C|Pw_y%+j;pfKq;DRA$VcZXoQu{{_$z7}Dx z4epdOG&g6zoWkoRk9^Rn5uw5oXJycDjN~IHJ2T;v&Zs_fJ7c#c2P~H3DHOjpXj_$i zs&>(4&3^Bt{(EKHL#o12)%t?iM_OrG2Yg>@0Bg zAh@M6=+YH2T-TY5|EB%sHr~EtV4*zn&Qpi|ZVRks!drLe-<29B&- zy0P%){1aN^MyNGt@@8^qoTVEtnrx0vQcMj+s;CGooHT50P-O_qsasT;uvIDwNmgC) z-U1sPgZ@yTOS|zpaCO14DGNxE_QEC^&n5_V1@|k<0y+a*L9-|_4FA!q4}}&X1EcL;AkR8Dgal>E1>qq_$*arH~-;$T1`=#gDp&wo^{SR{;2o`CD8}4Y^ zyF!6XVwE^gu2+9NGfCR0N|><2LrLR8Jhm`dRR>h)S*L!p zXFSSh|I+6-nv?9$eu$)#{kHxe$SJ#LUjJgZduNuy9Be|566eU0q2+52|Bk9;-fnP9 zsy#}lEc3{f-Vu$p8axNcjs}68@8ty93GULAUu;j%NbRG92&iYR6eMRq|Qu;{#`T#2P?4cbRv=ndIBD~0Zq;C`m#_0>>V za9iH4IWI|dYL)|AtkiHtvnYyW^FNkumLCTej9lqn?@df4*+G}fTSeu{3gA{l`gH4K zW<&^ihLjh7S`!=tW3eM-2=AHt_lWwa$wt(CYv-?jN49cHB-^pYIcS2hJc!hGO+Kf8oLu zIvQTv0{%W_H$T$9oURoubBq$?OG>CM-~%pq+(9pMJmtBTh2Y1d8K;&GJ3hPG)JZl@~-1fWu+k=Eec zv&NHu@peD^H2S&V_FCI*Ebecz6aVgH1W@%G<^Lejj%2$DfRZRCks@(a1nw7*08dxX zGrXkO!U;9f{$s4GYOt!7*3yrDra6tk2^RJnE7|Y1_>6fUJC_}?{r=4eaN!TCXIe`! z@f2ADMd+YBcL)S4GojpHhdCdEe3S(wubl8<$;B3$JWGbM>NQ`Q@+4nqv=#5`&i{qP z*s({OWirhhDFzJ2S}Gzr=&5(Apk4Xt?9FZ!9t+Qtr{1`xwJQ?@y-fV%Nbz0fskil= zQ}%hVaIy^x7yB?)Kf}q8Q9n&i?h7$O;|Ke9rjjB@;By}QqJ2m)AVN?|MeJ4f0{305 zat{w#?NBFn)oX2mYXmzz5Bnqvt_LoG%H$gLNe0E2;sn)n<@BviNnUy{6ng9;N2c{V zl!>}zcVNomq^5T*bA1xnphZEZn6P66IGw#tsq>(&K6TnwCk$|nep)$A)-zvY_&-NI zdE6$;Zk>@8Rg<=xS1ygKph~(4K0T0kwiXj;Cc zMfHfT9%IH(4j>*iFu4rC4&wONWPdWZE_R!_VKICAdEF}5WvS~#O&%vhQ>J$T^?Y5x zCv(=wQ-lTX=R_Y*ZP2#y&(U{Pr%8gK%k3UN)~WaP2JOkvF42B|j8lX2adK};(zKPT zEvifNyXalQBu%TR3IZ?nye?6z~yh8uHpbQAL>=K?5-B7LfJOxS1X#f8;dj=Nj zS-;*C`y=BT_5D-Sm&wO=?7rMJu_QMr<{Cwspv2(YSUX$qu|{x`^vu2=SgR~>-!8{R z6p!tLn!a;jr;akZD5u+?z3p~PoF*-yF3oRI_DGIRt=D#Z z`_j+$EqM6N=(sUY?5Hau+Tq824a^}Ta=-% zt&qcIz!~Gchdj6tjjRDylLk z?z@6v$6l=LR6(vQTz5X$n$=MjKy2s*Wyqt}16xN`=Sg!FT)QUH=l7ZvIo* zX1CMinq;L73R_+}W0s@nrNdbLe%^S)k2v%XPI2R!Ov|P+}v=D_59J@0Da}j*#^(wRBcc%<$5o zBb_D7urj3YFmQ0BzdPK0oUk(Cx62!QM^lk+$MH3mihS&Cgkra{07Olrs?Ryc`!*E3 zK^r7=DCLU(x3R+jFwrmHoMv9|_KNr$EOmQJ+_t&41JhfhVyXX1;A@Md(5KDu*!0>? z*$ItaQ%Md_Ugp>=Tq7u<&UqZ0y4-QYj3|NLt5&{$R;wH(kD*TxM2m0vUz}STauSF& za>OyNHv%L5b7f_G#J{;`WRQC>IG{{ zNQND+LX{?~&|ZpxeNH|VQMw?Wztj7%swg;(ufw7x6d~2sI;G1{h!TrnR|KJ;Ep|KV z%7a(&@E?e#VcqN={$*Z`a*5l3U>KsKAbjhasZ$OQNH$=VgtcpEtqKIo=GdaEI4wvM ze;1zlpXU7@cDoy9>Fwx{r3&>xxzw$~nES!d^&$5phVR(`l?Dbw_u2I7pj4+?d7H3T z-aJM$Dx5sv{I$F94AZU5k3 zJONHq@ZvU`f8sO+_S+#7_x91BnY$*hh;Pf{$eb1LeC@Qq;otSzbJ_;&I-dt5m8lEu zcSzvfqx&87uiv9nfT?(+$0h&MvQ%IzZqQ)-8j-cGL+HfR$B7a&xtSO0x+Bo)tFX^nQIVJ2^c7#U{N(Vvd=`%UY3&k5J4 zE997>Yf}dCK#-2n+9!fI5HXA+9l^`| zFbNC=2%_6SmyP^P$OM!Qq8>{GSG^W3s#nJ{Kv)X=P0b?1_2rr>*CM(!03#|Vf|3-A z2L5kUD76G_&zBaCA4Y?~zkl`edGqbgE8?uMB$G2W?V>czQNc&*1VOp{s%!}a;tC)K zc**~=-!Za^Z1t^GWD9FWd)?Osq(SxeGPm{o+lo}-e&106z@kYSAxbP7do&j<{^a-n z$6S)`6+sD$fZck}O!}~2YjC7E(X&CjgI6obcUu|OLDKlkJg`u*7Zd$yL&^Kj!>`#d+-B<8Q$lTsrp(nRa8lQ>7hqHn)lb5#}# zt%ALZ5G+h9Bk7iG69mZ<~oj6RnxMGCz+3`4CzQ?|55_ymK^j>rz~VK`2IoUJo?fw8C+dU#+NHbrt&CQ>0gAudV&kgCAT8fA9TTQO5Li>N7Y#63b)>oXpk@?q5fMHBNA% z*8F@Ov z&rw3lvL6}2RfoJEe`LMa@(G^>+woNtKkSEN^l|*mHA+^IMRw18ac2Z;q6>&4{imXi1F4+S! zU90SYhAk9D#x#G2tcb2tU!I&Q+bhGX%E4UR46BFi?_E*btSHW59=CT+9tV@k;P{3```9o*G?3qTuy#GPVVY1_uS#*w>!0Q0T?4w8_6>)Y-i?Ye% z-J)4((i`qeRdwn=8;)QFO0w#VN3H0^DQrM4qxZ@XV(1ffdUS`>N}dRSH0AB$9~}X@ zl%&^Nl)2t%(sEL-{p^cJzWCmw|Fd_&#V?P{j~N;eqX(J^0*C&Q5ggrElC{%(39?)F zmt|2gR1`p;uR*)Cn#LS(6RA^lDf(<>I9MOG#S_gsf^AN4vM3X({s}(3(V}S5_y3hV zaTtoHz;0>iK;C$;vl>e=t0=O9ioj*1+Gme{hyQXxf#r(kR-Te!F=FN4vvfFij8y%q z>|Kcw4=cA^`z?v(w!+x4rwabcpeXD%irGw&4OB#-JCatE(;ImuzMzfFtC*_8{;n2P z8J+F|)C2nA{o7rE%A?&aHzE$fQem&dVUf8W$!D0b!2G#LUBW2f&PP%l8 z?z-X(^bDWzd#Gymf-IhNH$M)X$?J+{^%dZ_8M7Gd@W3@U0VBpDryIEecfzmICF30* zn3};$h1Gpmf_l>JK)@)#I$=hjjYizLK}AK5A}v{nSi10x79ID6JWm` zHVYo>w)q)7(ZbFG5Gu0c;8>N3Z@HghppvA3iZF=fd1M=e1aueupE|?g#3&1YHlfB5RTWn^kAr_b&_l05 zy;CcLo(rVSSD(*+w;sSbL3(0D?R~L1Aq5MQL(|u;L)Xbh`4dnUDELM_eC`e03o*>N z?_8C|1$4WOjVXWrpe%p>&u;g(SObPA-2FG}u^?2vwk7D(Ie^X>;WtRfIe)~mZh&v`_OvqRzniiHnI zHC;r!B5@*$WtE)95gb&51Qq*hT4trK`l4ZxM)ln0$6jMwQMfMnCX4~AMU=irUqj^n29XeO&{OWef4*B-* zpmG{ltF5AQ^B^vgJvmyTyYY={;q7!9NNwkmPD!QjQK;|*@ojnkFKQ+00_vuh3h`UL z>skVLlmz$6tr$PCHWau__sg#3WGp8>T+sVxjS)B}BL1x1CXM`ea`rgUj0OMZXWEUl7rR& z^Tirq;Ytudpx^Kn&!y0j;x(>|GBt_(-Ta=A%Orc|L103GVt&E4X?ysK_REh>DGONR zkwx!fM#cPQZD5($otGhJm~-sFs=oa5s_~vkZWiq@GA1>u9*}jD3g}&PZ_Zxgv(mL) zWau}q=QWFx;hKCp%PWEDR3(OPknZ!YrWc;^$lv*HJKa10Soq>_G^`yKN>bh{1P?KZJ6X%wTUAng;8Bg6kK>Z3E}S7`KV)PXAt z04_EN-Q&L(wbJ$K7WEyc!$6~>S1&>t^<>o@V5J|Gaj;~{MvM_l1|P*WoG>zB#j3pf z=JY48NYN4ti^~o7iv)vw(LvXJqk{_g2VWoHoNfks>H#{@cDhx5%*l!_%&>FG5lO?2 zXRQnV`ue{eH*foTMXyjly zB)GtR@6@q@2}gh!ToG%b;3ea{Ra>`MZgguUcO~zT1a4bx`;AKiso&r{aw^4uF2ZIM zCDp63kpoqB!EZ(O*`i>mRVjjY#zeQhQ&AbEQI4vJ0G}2$s>ziGpBMuU>JdR?49K`< z+}zOiRUjjr+=W7-9`r`;&B0E) zR%y={y7{=u9|m%v2ju1~o$dnZg)Y6`!0Um!LZ1X;n1iPdMUuZ%QQ=NWP3e7`1{>>p=4 zY6#s8DpaYa7>L(YP!VxV6KsrI=>|weHp(lf=psG00K^B91-Kvq_ncF)M?G|=o^z{} zvj@K#ab;i_1jBupxv+ZI3a(J_yWrzKSuY7gNe;AQ8XG!ZU(!^~Mv)^}wJTk#p zC&he1k#;I#ooa~~$wZ^XA5GaiN3ZVm-KUHZ4nU<|{h6#*sXHHlQu6U$UG7gE5(pl= z4$@(abJ1ok`*&K=IN6l_<<&wbO=-7WPSR zn%=I7lSX-^NjIq|0+KiE#GT}_?F7> zBS^)K5MLey*JjhtO!5%s*!D=Y%siVT@_ORhKy!f(yWN$uNNzmqW!0^AsnVb>d!=uw zK!>Y#lcdw-sl(#Mi)|8@F@KhZ@7!j3=rL?@l-=eG5v`dRAF1AGu2cDn?6y)(?9f_@ ziKoaKDx%l9XEwCG0(Bg2yzzC<(BRrOhR}I7?Uq64+0|qBXQ<-h*6}VZh;_Gm^`_}h z193al=M1$D;C0jJq?835a@z35ZvIEGeQKp!ea39&Et~Hb27at&oSVrwk#i{XHKXg| z{m0UUWTPGDBa2Opzz&Lm3b5@|L=V#o{Iczu8joX|-C>2QO-wa?gtrNrP?J@g)Mun; zoo>!XpCebie9j(UP+b7Yw@Ck+vs>m=lZWyn`GFk` zL-?uGUk{Px|5~Kz#i1)bAE;s{$9vTe(4ph^tKYmhK#$%(z10V)(3?DV@f)w9d0UA7$O_qr^*s(g`mzgQrw_Ds5#6lR~~>fkHb1WLEOjsV(lXU+29rH`Dbeb zPW#Cl(|%%Hfb_S-Pslkt_Eow~7MoUzX{N|kDx#cjmgI$OkYdEF$hX+HRkT1fW>X9lwWm=L*oCxGaUR$w(Y=ik=1P-Qm?7(PxfWDTCx+BQ-Q1qgbVwEK z5w>ekI4W7NO?!iW;8N+^L&vMECL_l(x0WD<#U8MXXr05{kn*D)+8vHYq-a~;)sZEy zOfDeB1Xqa^6GxHN&=?JAQ;gZJR_#|{x^#u0Oj7AL+)`nkxiKBZA_SS9M`H+itsUrTy1!celHnoo=TGI(XnnL*$l)<~1Vlu|-{(W3Lt>(NLPDeczveaLbH{nU@8|RVd_Twg zt;GWTTrlb`k5LT0N{5EQ<%96XR4Y94W`6Rz^YWm#p%2>V58?YqbSsl7>Gv&>>y=l; z$hYr$Ltzd!yYXs%o6GmunT?4*{dFqe3i!ce7uK1}Z~lG)h|~cPAd? zX^jUPr6n3%+I*tc@p`yOke%n(4U-}E4?Xor7eHKtcfh}Z!bcjnu%5}d7V|`vZ~*0YfNm^>0=c(R&0WVBuq0yO^E}tyL@e`SE^TL zkX{MWw~=QewTC>NIQlOi8`ju(`Qw~+mK`=SzWb#~0VccBF?YsglKRH1aVyNM&^C&J zIV+orNmk@>ZqF!_V@UQ4S?`5~lG^LiJgDr&I_^GlI|R!QRs|izcqT9WdUf49t2nD5cas7X)@k%HdgaW@nKz~vfvQW^C{PMIIPPX+ z=8f`iU!MZ*w7$xy^C24-__U zci9Pcpoo$EWP+b*qG|ql`9iYMjUCusW>ZHV#pF-<8Z9H2l-_JWK$n{zpJuq|oWSyuW;(RD3 zkPUPnxh#$a?bAQrm#+#cmpk1<84<*WA%?@kv(RG#lG8wRMX_qv z>rOpa6?L5S^6yHs>3t+As9U;Jc8-I`8l&{e(y4_osw)NiNIIumiY3?&z{~BJ)K6B* z(mCazy`4s%XKI;3;Z3kWNkQpQD$+YWK773>opV>##>X?+PWFJEhy|zN03XE8vH4`h zwT!Qs>`v&CX+I!`*x4O7uCi@6vpZ)f<`hL5sF=Edx&XbhA~095(`zNSfWi#b`KdXO zx~k?rmQ~Yd2vi-=1r*fb@_T@M1G9(uf&uQTpg7KbdA9;vgfi6!3HB;t!mED5Y7X9m zS~12ea@FWqp-kdtP&+L5KRTs;LZL8LkTM7V)hkhl6T8yN{mY{=L@P;?3E)Qzznv9WY*T_50eXzp_j}vuPc3W8B!tq92XwrjtY! zv#Y!?|AgkHNqAJ=@2$n2eki;I#%An4Mfy6#4omy+iq~!PG<$s5e3Vm9-0msdZK>-1 zhTy}d2K+vtafkkQohZk%a%Ks4i%+}ay7aKPZWOpUEXK}WR9M{R^4+Z=vO3Eo5gSyv zu^VF}rF7Rborl7^I!-#s)M2p`HZm_300k7<4P$NPs96gfV0*D)^etQtiRVAC;mhBB z+p-M^^sI+frlwKMCW>sNVm=dJnVhWXQYUjS(Y2b(V!g6VmK)g;eN0*?#JyvXO=*d~ zMBh{3UXHrcdvvTZyH`8(f}j8IyK;lwTElHMnPx2O|Gs7I${S)mY*f0e4X+a;Eyj|! zHijBwy+9R5pN-rLyoB0YvO~aiY&<+V_(qJf#JDSAZT9Sc9lLSNd!~iVzn@O|4mse) zq5rF9cI_0!0NL%Qz>+flq8ISMN&AUGnj)zUeBytLmmLla$9w2Lp*Bg>O@FjY^wa|{ zLki^zN0vSu3SX;^3eas99uh?N_3&&pnwr`NegsSiE+LlgA*e#FUI7!fGhQR|A16hMA zDkl9am&J=9vfmQj&Z(!bD$dW%iOh@C!3yU*)1yd&$S<(1Ps|T$<6|+DUU^^M1I|q~ zzk%NpeTW!%h4R&!E_uc5gHZ9XbbcniUexK`>z(UWty%$X_bt&Ud@sUN`?0z~(!kFN zyFjY>i9CD?D|lHQWfx)2>Tj_*IC=2@9$31)ZkwyN;m9JlAl6&nlI@zff-}TlS_>NS zmFhZ83O`>241`fg5jOHcwEM-f8*RHEJK_sjS(q{1|NPZoziYBEOVhe>a%b1&46y<1z|C#+^e2W3CW_1 zIhe=noOKqm62+WuAn~{fN-Tx*t|^M(_d6o%>25_Iec8dD6L$Esdomk-p+i{IeCv3U zW$49it&NRfN{2WRm^aJ)4Rmb468|TVbFLTN3*9|ui_a(;7}MLgbDSL-tC8Hq%Fa*| zfBeW*#r zCy@>tOi1Q5D-(G}gbm8`MGf3@!s_rL{gGy%@hz8O)zxw+ zg?Z|+OJ$(j_%{P;A#s#GDH$d!ZFg`RUl(2|>;dZL*g5zw@Cb)CE468&G?8ApB6<_I zP^gU+!~(Zus&Z(jfNP+Wr!*)}a<2qyai>F3Pj<<59Bm3`D7g@r$U|cA+(2!utX(na zu?%wUgC3Y+@8KRJ73xHhHa8G+TiBl6>;0>dfo~WNoxEuqd>MWEFHJImPn{6Ik8FNp z{Gq*O7Hm7k#28F=dgy3__< z18}0=oKYz37nhOynk(Y{WGijpC3AbH)N#A`8zlWRTs1$A7)eewUY|cKYUPDXmX#4N z&TDSG4!2SNP%S(hh011W!6lj^CY>o2*U+jVz309d%V-kVSbOM zE82KrvLZJyLzD;0?}0fbAfxtdzd?QvI9p2P2O*|YBVRIg&||M^36xmiV_Txz_+aKV z*E!e^rJGY2l*n7n&7D{t)~7J|Rr~0bRpIG=>%2=LoZcRIDDckQDwavD5yQuR1N1pI z?AWqxh!$vnId7WD5Y4}ux02*rn+4q%SclEb%^r$@g5y#urkQhqPVsG@`4ni$EWbfY^z~t6HLC!+I1?gMse{Xq*@G z+GpY%h#GEq$35@<89HE2{PT6PQ{_BMdw55d8m@OQ*~o{inM*^poj5TN(f32bg45Y-Xvj%?b%)BYB!CPQbDFLm| zdjD)X3Da2_-aU$JxNERO?x%4XOfOl2o7V*b| zp@zCNy3iN-oV8nhk`-GyI?k$~ozNG7Z0;y5pjYly6>-}@t{`{n25_je)oJv-pwo`m zX}KQc!=Y>cHY<>i`R4a!^_HwdZW|-A!6Rf`m+6(6K^3#hXP*$~3r0Br{d{oj!o>5d zM{k1_CdPiyEX%Z133$Ki?vz%SXx%C5#~qBk*e7%$5f4|;5)i)fsSu%>sh05olK>ARjt?u`7+P_adW z=u@dD2t0ZwM(tG$$}dI5Nh?Du!O4B1SPZ%o&hv6#7;Y9&vEnPe@HnR(unm=S8PcZL zowzJblpZl8PJJjaPCDSTLw@3IRNz9T7>q7I^{Db{f==x=eg}PDjLn#RWc^!|| zZ#ky~sH_IOLHiLayRZpCN3KVr_P|q*MiAu6ARQ2?cA=4BThL{0X;|Te-NpCr{O)I# zq~EWVJF?$w2~3@&I9ZWME%k=7(W z#XhvyxY>n=1!D~Y%NiTEUG3VanpwhI!b5&4sNEc*MdXyJHb><|JoShb?3+*-x;aW~ zRI|bWK2&GnHNY7$yBZQkt2MVnbevqTld5ugP3Sr15UGE=L4y*ePd#cR9h0;umvmja zn3oGeCb?6~`G;iJrR@ssHt}B7V{cp|!SsaHfI|#>I7gv>i~?F{s93I_?8Ate&2V&w z6{2ig6e~{5*z~(!o%p&-3+xRnRW?B^xq-hYYI|gf7;7U?V&MGT%_>X{RSUbJZmN*Z z3EDpE#`LW|1sXiluQ@pV^Rvl{XXg#{3ZR#0-jE9;H`YdM!!G4W4aFd=rEd}m_NTFbFRE@e!rdOsw?eba9 z=XbxFEnpOJW{%HSppF&Js84_V<%^btHa2);S|v5~6L1}F&nQvfg^F+-P#woZ7 zNxPxeRP+!Ad_E-&5D3i~lGPmWIq9!OJ%}ubtTxM5Lv^|9=8X{!$tzy0I&RO5q@Zik^`fgY_Jj;T2<0SyjZf{ADrJqN!4v<)e(F+T6TKRS=KS0f z>Km$3As%`4D}Ckv+siI+d^}$E)RCWKH6HF;eb7&G4VFd7Hfnu`*c6nQ!7;62N}xFe zJxo{5MK%V?UR2n)-DQVap~8KqsDI8bTN7+DAOErA;3~3Z94OEaHyG*^J8tZv> z&ABn175j5bdHUq2QCBM_r#-p1&n`HY#TcWpcACP*wGkW9H zBBp{W69ZAcPX(3d|FNK5-lHiE?(|NFxX}`I7Ib9|cz-N_s`vLUi$4x4|Kepa)~Mxi z@cBAe{bS8K3M&IU3P%@in3=(C@~q-@dO!P{u=+9(t)umIu=z6Z+C?`vedBc(U%{m3 zfTB={aXu{;dG1vf3hxL%Q6%_|zGG(hK(igG5nj*A437QR3BT;K>}_z{IJAv!g>8a9 zk}_c{uvJ6ls9=)>|4b7t7nHz!XFPs8q*Xau58d-Nu(ZtGz_9+}?Hn(=@nUOYwDWh^Ud9S9R1=jt_SGp+_jTNejTGn(Zhl}r zy)Ix4**)h8{ok$=oj9jb2$qZ zY!eR%7A!~$hX1C8Bt)dnF?gRKC~;fm)fisQDd4vFI}ukShLRb?*aF7(^|qn(n}3;^ zs5Ch>(Z_yPNix_uHEvwux6jO}DWaG{ifp4|jHxl;Uc=-Vbl5d>`n{nj6asY^r0amH zVh|Ej?)GkxBX5d!o2oDpNLAj}c1z=>s3iu8)4QIP%F;l1M1!Q$s~PfIdS$<)&FKYR zPDGP!XgF}@XJ7gb^Gg$EwtLNdi(Gllx?Kj%ka3q{x+!v#ia8op?AJv1dqV>`)X{uO zIwVJAd*y4su$oyO`Vh*>uab1=#q61T(!WXCENhP5$!&*j7TjMZD3@1x^>OMnSEc(y zFGtl$D(Jl68eZ;HY|$$dZ!aNT_GsH_Qc{4#U@8 zlqLhjKYhv-DPfzX`C)ibS<1L{)}?P|>HD>g^wFgZ<%JeNurq56wy&JTd0E&yG#qyt_w#l$w&0nxh*{H1+wq0k5-L(l$-@D z92{WBZp)id^Vww63>NqMfR!Zejq!S_%wVyDVhSjdN5ymi?{v3xP>yvvt&#_FsM2YK zcFp6;bkAaF?8HLSZ1rY8%v+-NAxdI)L;dpS|75B21z0tNHA61s5oCcFo3o8$n0WbJ zOKN2Q^g|P3&i`}CBhu`~33G!P2HGj+Dn%}V+@;?>zeZ{L_=SzqHmC!{piS@e3MN@` z5i}SIIDfrP_NyADd(|jx2PNku(}v{jl>GD!nG0gB1N8s4J=z)JV3{%RmPPnF3GxAC4-NH_5u` zZs9VhyufVRVJO#ViH_%;6PC*R#ibz6n?ErrXmxm>W)PNSH>TG+zqs95u(37#jvi+P z8>)2K_s+cT*2!3&4h$4fd}DM}H$~KPpsP0zs!5GyqhrndgW*G8`GO*Ylt?BAoq>&wrc)>q*kKEHogOZ$U?|90HJmbeP4D^+&

  • i?hK^>ATkO$c} z%j#Oo2moin%nn0j)7vpZGxXA;kFH?>ciqbQZ2-GIBl~M zKRqD8La*SK4C}|?CyPt1{#cHo{j1RR$(ByWp0dCwgY6I00EH)g$c z&dk?7MlnF8Q%A)Z1H^6gL3yLBhnv8xCg3ACf~r{u$ZnpJ^oVt$0G@IACJLz7iDT`?Q=IycN(8-k+HCBWQ`m*KI}DFfgQr_sB-m%p_k;Hk$3$*mc& zf^Hy4#Ya?!S9807zXJOK9XGP9KME^T&r1(E@Il}$Z=YGj2{)lBKH0F7YJ;>b_RV+71Yvk!l zB+PE+wDC6(-JAwtesHb|>9Ie*r$~O1JPGcCt5SSV5WM3|V85o4UkEmKfSV+8mKk+A zA{<13g0or(p%ePvYM7 zOJLf_pge`ME_j2aH=x%)U7Y}Q+j*Q#o~@kCVS9rwi5s9XteLZui~MmLBz@vR`Rb@R zXdhe`;au=I3pvhq?*3&%+k4)o*{b5p?>CW8tYu5x*RhA_@$dvyD#dJ|;9Z;Y=-VTQ z{F-Wh0t2h#@IRgs4R}NGw5Z>EP0(H61YdOTW9K+`q-}rFtTABwQBFN$dw%-n%x@*B zOz4=J_Se-U=Z*1<51QeooMM3TVkZ^TC{0#8{zcb35cLSXD!c~8_}Xj%auR$R)=gKw z(=xAzzhYJozl+8x3j)TV(h5Sb@NHBYPgXR` za_OYm4c@i99dnLHA&H-H?@(lqVkNHw;>!7=+^MG`KP4-9+kndU*^`_ImFJ`66wEw- zn)5bp1-~V;jM1=hV%&I=vJq-3_CunaA)bQ^vRW;H~m z(N&XNfdi`}#}*t`AL4?8|N7zW(6^uYDO3NjjBK;!r?|1%IAS*6?4_996e*)((n1n> zsfs<3ABSxO&$KO~U)UqR26bM8KxdE!Jk*$Q5pA+=QVevjsedn0>yN^B}|)BQKqE zoEWUM?Z-6b|Lm-tsew4}8~ixVMK74byVN_mrq2^oK1o6OqAvO{$q50fX)l&!AtU%C z*uL%&{1mHOZNF}7xns8_Bj$D-4!87i-8QLh!)rBV5{1R8Qmz43I3>a=v33i$KvSYg zR&4jJh}0h7mOzrbOMcSdjAW@&Z6L<(tO zh<>>NVFlWkUd#&7V=qlC)_mDCB|UiWD?cau-8j&D#ccX%qL`BuIZnki0b5cYCtKY> z=P8!RApDlBz_OP{po?r)V(mEUGhlQCizKwws!Rs!uUaLT7tjvSS#xXETD%*TUzUVs z1J~O+@5LbkH;iKBXCPTzaF2~uJ{!A65@fD0#;bFHbrQAG)a&!XyDg`wkT8Q zKt)a}Xgy?*HN1AkCRQ;R=Ys0xkYkfzV+E=)4^PvHGfbe$2+CSV3f&k~pn^B7F|(Rt zpwntM6>|)tD%JEO@KN&AyC5^vENhpY71nafg6rvji~aVK_u78d^ZvtcJel`z7<=mq zOCRF+LS-wx@*uY~s#&^MRW%pRNh^S+c7R)_sRu?MdJ+4c0ZD3eiHpDRx%d&Mke_4G*4*!k%nt1PnxZtE-C$PBDiED|)zFeR{$XtBQ6NJCjO-c;Cw;TS9_ zE}c?8vO#3=qFIkOi)WAIQrjlQ+7eGRy&iQs>q}u zy)q6iH?||RDh&MX&@7d#sHfA$H%hbB_#1`3hAF3KqtX<5#v5iI@$QN|e zSp=szJh+_7h9AKjQ2zmDzYVHtRdqO4(rgae60t|UJ+j8s4uQ>9`J#GgH^xN8K~gR4 zknB+xP%oGx){wJmOh$gt5}xcJ_sf*uM+i+bTYT!3Uy}qkc9*xA%~hKzCWRvFshAD^ z=ffJ6YsgV`4P7lfs@B7wX(83Z)68Dks-VNFG`c2qSp>3&Rtr}Jy)clLFVIQ*tnuNb z-+k3%SmA?W-g7_N z?823$Eg81&DwzFivp&nigp`7Zzxjy7zA;GAn;|8UViG8_f{MZBnJa#m#U+|9x;C&~ zab0@IbF};uvr)7M5@t7y(vJS>ghR9m5`EtZIz|?|v3r_sHppuzW)($NVsExyiNQDQ ze8bj#y)q9};vKKsAHl1)XPBLsPqzDS%4$X#^Uhc%PGbVacXw?|CnfCcgd4Ly05RjR zgunrcsi8;}6*Dj=i8~}2wOEiM)#lKNq@VnCk>IuzWuzAF;BNK<{-OcJP4CmQTP0~C ztwEY1c_O{(ox;B@ZRYKgtn})ToC#QXRnbpONhi!TA&CgCkgJ$P-Ys4>BoER(t9^ht zsY9}jj*m*@VaxYf;gZlQrdjE5y52s_4VzxrZrY79^9yfY`9+172}OU@p80RG+>K33 zju|dCP)ssK)?l&aaG4)%MP#P>qHs?XhFG+X{F1;nAT`Pj9K{)h-C&y!h%5J48JMxM zkmij4G(lr|ylyt>wzlYXV>9x}*x_d63B?RhTAT68&ye0mZU~!BmdlXP^Ikkris&yAY6DU1mK(Tbf7I6`chkB>9 z@ngfvr|gxL1w#886hE}_@uqCL!4s64dFh@T)qPRgkHgYMZ4hninvoiDHM~JyBHRcZ z5vcQgR2>HaxgjEnHoj4=sz;+mNs^;c-SE3eOc5*S!z_AOVSpiO7jR763d;nA6R7h` zga#HoxfCwP3ZvR7m7&;*qeH_2uHYWe{$>c4TUeAjuNH<;> z+3tlT&QeaLL|cT)_6Rf?&Yev-iC3EC9{ids(P#2#*%s82rVkwte) z@y~lhe?t2!oJGB`5)Jk{cF*X5BA5oxMBcJ-2gN6)Idm6)`8XtEjN`RLAL3QxTDgGz*2WK=}o7rD%uf#v=#ISVXQsE<07j7(PGsb*ylH(7y2c1G5! zA-*aU=SA+U9tBcXVCxYw#3n1YLKPlrE`kokPzV3QMt(cE8CYkhpHm!$4Jp{zzYseT zFy9HKf05WThfDrm?+mDgD3xW99O!z|EA3hOI}=Q=hTHQiUIC?-X~SEI{f_z4TWx=e zFab>df7;DthZ_UzQ!{|=qnJvHR8TQTUeHKjI?`4fOfA}_z=2dV`nCYU4Xkko1k5yxdrTLS=?iy z3yM4?0IzydrfFxfpma)Nm^~1A#aOutL9h7PSDvrqJOwH z#e|5ypGcRI%yCe#Hr# zAi}%|p|88NHbqhje09y?_cb_cpg*xG;1Fq$SI*Yv9W$sl}RbDq!pYob1lVF-;l+V4Gq}!oJ9ho!-6dYMn<6_+98DTwHftvJ^es#BoIRW zXrSv-w@L=Qksu$7_jMGQ8C z6c3AaJg8cM%oEzc0!lY$wFu)ZR_fR+|M#7~Z|68W{dZsWKz63BYU{>zmW5DmiwYY> zPe|C>qsf3p$&$C?g3ku&z$<;?e>J>~|B!h&O$VYWC;iul*YX-=u>u_jwGv=nl&6OR zuaJKo=i{)Vc__kYP;@JtJo^nBDm&5gf;})D81^LRZn8oPrEcxif6e5igf5x(19HfX z7e4J~-qRV1IYp5MD(0jh*Q*u;+xmfPwQqVS$n%_hCw7irxmlJDB)_0R0piSYVb!Yp zGg_i?Lg}8iL9zmb5^qc|m4kqy5GCRd1?qTpynU0G&RiahOhNmnTvjZLXyofecm3j| zc~e@0ug+M+yCJUy^65@^>Roy#{NgUvYEfKRcR;M*G}AdNOH&Qgut=|I4j550goN9zB_T|k-x7R8RRiv|JGt&qqIbWd=7t;bO8>jJpf`i z^v=1OUp>FaGBIEyqW-yCphZfaR!Q!}IO$N~nI2kIoxZ4f#s8l@FVxj$-P&J~)fef$ zYaZ40CkrgaQQX#J$uygTp39wJHA}V_G`X;jL7N^C4=edr35tVbK!xQDU~R0~9J3#< zY@RyuYuVv-Rq28sTe1YXEi7#C{(x%=nie{!nk@mEmufz0QKUn}ra_uZp9B}Ad-`$# z7Qu`nK*7^HjO%Ny)$D4ScBOTeN=ryjMBBo$f}@1-*nzXekah?V_O)gtVxi?5_R1$ zakE~uT~@_ABk3pWBqup_p}BM|=c0HKkUZnloe3>=*st~fnP3myt65kfzNJ9vrM$>uU@u(gyy(AEf#XE8#0ofLuKoAk*{^%o z&@j;hqg)@N!36}$ul{Q7!^Eqf)K^QN6>ap`5oj4KM2c6-Q&C9l(2OxIPS@eyK9V2NtbK_De z8x3bQpfI};h))L~y1QP~AaCSvmt73+j{;^xRWWCOC`Qx{i|csLuE9dvYPvbHQn-X< zaO0$#T-@F`0$e!{L{^@(*DKY-aw%kE#epxZMQKc@K)@skruZR7ueIYZdSL+)I+s|V zxNny8W_K%p_=x->%^(#CHP{E+fiCBv%>9| zznr1N>|{i}s_1}j;58xrf%oN5p-Ojz43)sA(PxOkFJF`s(Igy+j?YV8wXIX^hQQWS zoPV6v5V&vTwfMnjPb}rkfmUl+jr(&QaASZI(y?J$V;yO$q->J*AVKNEeoZ|H(icX? zOCbPyR`X#S9~Gst>8iQiup%w;yXf^HN~v{7S|xf;nM%8bs#5C0H+gQ7;h}uNhp5ko z5&jRI^W(W=ncWmhR-f1TUMUD*!SbcjT|-8M6B>3GexyDjbRQqWk`z z5gXErtC(~#h*Ql<1H#FjD!np?n-y#z7p8qY=dg0uv<@Kb?-m{-b)qV-BB1)sT!=?=i>wkv9*a=14ji*zWkXU1OD zg=vF8IhP_iB3KrYp}sKfzPtlM#0TM7j(Q8`!`f^jOO4fTqP;ygR}{ zj}p==S>U+scwsEao^R!m>(w!WhOj)w|4VcvX&ifj~lxx zMP{(xLNP#qu$hXatj|nBexJVX|)$CB>zEL3pO=-hI zKp7MRtZ=DROiJ`IW|cRZKv*BQN=l$C^VUq*xI0tZBQDTjB6(+&exlaMAqE_Q#z4`o zi7Q;jJZc(FhcV*f`liI+E^M7(!btG%suqw9ZXBA2qNHJ?zLjFKDUu0&>t36Jnl!rb zjiC+v-E%H_ZKd_oE_&@&-ceSD?3-LBYL0IF=ILl$xFiJ`Xrzg3@sE>M@t&$qMGbf# zKoIadSF2lTo0{bv4Xu($bImDlnAwW zppC>%)GL=pJRnc}R|o75ZKCu1oV++13C1j^C+AH;W5)dXoe!JeL;|J~A^bCW{p95F z3)g&MGY}iBl-7gv)b0_4G8}#Fb`HIxk9YV!+vdev)7@^#mSTgHWI5-Ke>wLP(g&?m zbt3b<7ErS-3D~PNt{w+H776b7mx`;rYW)qoMyN2Tpi9MRG;TfNrFi50r=^8Kco(iY z7n5vmtRRn*z+4n&Og27wAE$wLpT0kR&|{5fjyipESHQY}y{aVvPk4CupvOTm$Rf}U zd;_mb-ZHg=k4N>&UQMh3V`cbsJ0}u5$ISA}9KpVQ!)SZVUaMH{2EoF2-ZRZr|9(2< zJLG^HyERwMW}j0O(?F3=shDgvYJZjjpCDGg8i{YQxO$;Mng&`R`k;h}d{J!JL0$oO zK(P?7LAQ433RK}(*eFGAuL6+Yz@<}mM27bOuz;aL&}P3&Kp6-ZtoEr6N>&{4gsY7L zdU|F2_u?n-xR2>xd1sH* zgsP(-eEkwxGY)jxhx??N6a!rNn}CiwRL8j*-YQ8KX+Kl#4^4D2~ZXT2Ibr^ylF%ORkT=ojs~dAjTynImCPapUNn4F(lmICiz`M8zNm6)*3`~ z(P60kT@`d+jzlQEn(FYD=yt^;Sd<*2Z%jWFcn;cowsA5gEz#QqX+c|j?$L|MDG{9E z_P8*(YcE=!%@h`Yju;iT9>5AM)Z}shb$q&MlKFjAbS)`(VjEIjTJ% zBQ_|L0SCWcSrFU@=y^Q9%Bzfk@W(sl{sTUVye%e5ThMoI1(w@2Ay~Jtn%Btn;rrDZXiZ`Yp@%?win}HxA>zZ>nW*w`%?~MvlNE`ahN1KMUdyRQ zX}LdU28L81#%s&{_wi2pchgQQ5Vw5&BOZOr2Rdn=m8lui_`4>xWv_?ZVkOticxTK25GwXDQ#)w{*jfgs{p>3r#0nc@|GfFo zn13)$L#^7p*}(;AU2s%- z=WqBb?q>_O_|(qd^<`8@{l5i&Jte}6w|-CylBv%wz61)Xcwz5+qmb&;870D`$)COR z;k9@E{n<5JeCp{n?|yh-J{-YE*3*siKU_Ee-yIFOjM(g0opy#Dd?kpiUtk+rwAW4E z>#jzHN#hl>4@?9Fv3AAkfCG>{E|UYVa+)Z9cF{x>zU!B?`8YbJ*5Bw$kFtZoOCROb zleS@CC0{jm_uBK{oNKZ$%|DFGCKc=)T{o_XI%{Toj!+CRDI7pehbCdZK(E{j2_FDc zQqXSiBoR(q=n=kseV% zwMq^^kGpnbXubfSkqb=6_-D2n1?S8Cn}pRW5TL?vO~bpF#T}AZfq|bq-Z)D-3RZT= z8)jJS-EeN5mIW_t_|e_i#@NszD;9m?)2zhKYm{u!o*8PR#TLcUiOdck zReK>c{GQR+e*e9~MU-jcdE)cr1R3 zy*dB-Jni0jqii#bgJ;jMkH#=t{)TPKarmF=$?HzkwO@tIx0sH}6QoBpE8CS=dHuHl zfIZ{y;zDZ6p�}J^t3d{;z!XoA3U8g<*RX2~4Y`TX9Zelq-b@N*dh^nqk#+lBi;~ zHh#{a99@!bx+D^Dmf?46h~>)|iQDh9(NnL)3zNrd?|N=zy6H{SWpO%hwdN6xVq`6x zjh<;DjPT+`&Z?`i3vBi}pT_H~f`{$FwVHv2ulu?;B_x-7fvg-~61ZgC2I}FrK7O}+ z-iIH4xOAL>PKF4|U#=VYzBebXyu=L>dDMXKoG|FkR8L}?;plB0mS zrNG7tEKYRRw`TS2upSe$kJ0KQopO-Xn7FT-e{asjos&#pQ76RjBb%*-Vcob)ZLb-! zw^IyIdgW3vZ6MxsUyLc6+tLIECLB!{t;OKS+VDDYHq;&Efzl5IFM3S1wt3=ix@wX( z8DvrGahQ{Zg=5ZmDXr`3ik%k-a(EZdfeIU#q1 zeVQxcT`^yK^vcBqCvu$g6TWzP-iMois#l)& zzd7@kfA91GZ~f=Z++sB}Rq zlOZ~*YK~s}jYR6h&D6SYK7Oz1ozu|`(kHc4vqhM zP^1Y~LeF_y$j&z=u7AP|gf$dXMUg$mz)Pj_m^aG#EEcSlwNKFjM>fQ9v8AJ$o2_nP z4sdhmeFSW3vLcH1&7IvdW57F~-tPmWf%RZIQMURKUFh2piAyDy zxiQ1unw=wIZo7xw@2X99BjL7@@p;<1MV1hGk!aF=S9`Ybd{NOL@0w{ua~iniCsb#^ zEsyu9=fq2O9BqTVL>TY0G4zRlcF9 z3H;kL^pJEx5jSn6+CY}^s)gXuVIMFy?CzW8lHs>GBe1}Q&1bmc1S?#)uc?l_Ie(q6 z$s1{X`|R&Y(i^kNDK>MevM2^Rh4i4C%WWSbQNcJ+KS^L9csC^7vV@l}zyi2jI$zK# z*+CLy4f2me+8~BkH7Q?&!MRcB9AbfXGu!71gb#n_zJ(VDEj<#qUCG$+rrY?ddFO;j zNH2G5NMX<*C=+f9IySdKp2xW^-4n5F+$TO~X=J{xpR;^iJExw;D@VoFIWp+O2gg<0 ztV|sBu+ryymJnf6qvXDGr!9B(kV+?NS3FUio(%z!#ezeg9TH3_YfCh_bUaAX)zet^ z)~?WT(j$(l6C%z8l(;g?{5cF90hDW5!Hj>mfAsBZvXmfvLm+a{Y_ch*m=X$TJYtpx z?NwqLXlK+B&M~GaM0-$Xkm3$9tvsEX!Nu*YDUW7c6dJdZ6)nn(&^Y$g17!-ec>gx> zUX?Cf$2lNNkwDwmF3G`3g^|@k*bkQO>D+=i$ADut6Ajz>9FQa3+E@4ZIM3vzOx^hH z!({1emOjifgU33GSxb>b>`T|Han*~&ZYWQc6of5=(D#)#zC#`d`f;N&|2!WZ7Q_Gi zp3&H_JrQ*-e$@23cdF+Di_g;VLScz!&qRZtHVqQQID;VH=;#hV45hsY7>!XM{u{Ot zU@cWS?mNj_ErTO&>#f>|+6;J~<*W+IQe(TsqZx4#ZS-NfKvE;gQyBbi1{|JS1M-1a z2~Hdp^yAPQ{!jedKpkPDx=3^#v=Q#oW%3Sw)uf83O!Xp;&Sgs5_TpvOtYE%5e2|^E z%?dBn!<$PdSgLfqR`FIFS{#?6x}~^mMU%4Y8>ml;1t;|(n`K>UZF+<@kCV=WW(`ng zKdyW%s+8RdyC0>+EL|==;N1`om6Pi^iv)P|s=5ra%iYt{M1{a5hW{*@vqjWB1JT(a zUmKDI?K!B>5HBrbs-V_OdjX0r(?q90Kdo8WAsO)L^Kx)98!^Jn@n&_;5#ZTI*pzP< zyk|+I?zYy)1~ob+XE9}qJ%ccsTYaj-tHU4BJu_~q%cZxa`GPe15nVu`#!KP8WyNg${6= zh6oQ(E27E^9T=>-?tzZLY8v{Ep@XMhWT5-W{!py`#CZhvtV>u8f%{VBC8pILwG6@82!m^ z(TVQMp=-jkFsO-+4bzj;(mSA7-A>AamyJuK@6p-ZyYxx2e4MjCNUNjA3L>UY)9Qnq zeDDo{$hP}m%Ly_?QVO2_u$-ii13HP}eRw-51~@tMsTgCGbEQbkV>@{cQBb;Mr!yy@{|5J2R#Mtt+nA<>NY;CIdUQpS~>I{^<&xg$Wr;% z5N#vpG&l%JLB>N*Jq%G@>I(p@*syCeNVz5reNSKb*uUJg8@S1NnRKsF|*brg>Dz}l5Qm~(}SfZBw z4>))SJJcp5{5C4WWMAa}r`=3;j039r;ZxE+im9YX1?29bUvCjegn~R5v&eVJxa;D6 z-$D6i+0ton7HwLA64Odv{2#s$Gs^yU4Lu0~vqam-Ru-;RU=U z<5ooPSFGaZzm-K7XwE5?j;o>D=TuIt7jKsx3A-Vvpen^z$#$7@rYn{MWzCLQ9&>PL z%x_~gQ10vZ=8L{}-Le(&#rU23nnzoGldY0OK{l{$VH)|0Un(fv=0I$BrB@XXDV%ac z+L_pyc&tRDgXB>*t%b(246<)lHxNiBE7Ha%k6%u?*d~n>+2(k3-kxn_i#BPBJx%7O z{vT@oLgK9D65ZDw19h_DwlRrfR#QZW!ctg(kOPhB=yn5VN*aBJ82r#dcZt{i(y_7| z7B78@^H17_#V=KX<$)$xJYkN-6Fs{^IQO-V1m@0R5_g3ZvyCEKshF(Dv%FefhNxHa zd;FtO`Cp$OK~ddg#Yz7LzVQ@-sv4A!nrDGdH}st)M>Hx+=ryE3(-MtQy&};e!IQtQ z!r=E#OiEFk0CP0vFKbEhI07O&!`}3MirGiO+mES{V6_=`ZG1NE)U@4!I*xJa1Ld{i zrb!23WxOirp?_Uy8y_8(0`Idy`}|s>n>i^V>4LN1_w-82qs~N@iTk4Bcy0V*zw@HA zLphr*;&x6VZ|et_LHPK#G?x?ulOz6CI%k_W>zz{~Cpa&5f5?{p$xHV~{$bnxkgWvU z-S^up{JF>1!b~8#{>@v(BySw4Hv>@x#gtQ|go^0~-xoT3qO<9B0Vjl;!`8obShYpG zH#l!fYw&d`774ZS@gwc^Ud_9vsG5xQ3P4{@$46ZhR?WR8tbv#woagpxdcEs3y3i^w z6dKFrR?|0U7DC0*fr)nm+GjceGS3Issn~h`)RAwzA=vr(>8t-U)r6fte$-w`w!UVL z46sQKvzEIlri>!RR7|-%lj-3m0>R`^ze*Z?PFNj&MwT8CufSN5)~Ggyn}hPS;10=R zL8cmtrk0Z(qAYO-$n$>2zsb|ahILHZG3O$X0we_`$?IkPqgEWn+ZoA^ItD_nuhoK&3fUDyb`Y!U^ z`N1KQI=_|41fieh%r(C)++xK|?|ol3%=_$%h3~9;chKY5cXaddc4wM9%Rw6vEG!@9 z#C_Yaur_rDek)lQY_c+c9}2y&bi&y*kq1VlnbRmmW|nqHbPQFsBr9SCn^cvdo1=_D zYRGnU&`5czS1#j~_#yR+u@(jia<#Z3L-H}>BHRT_@eyIn!bNi;#Kxq)KRaGvLe1|> z#{Yq=d}CHqK>Rao*4jid8_~T4LUvpoZS|>`T^hV}{)cPdy(I3E=Wu&qfpinx^Fa`r zLGtx_y5)VHtA*j8k0fh&JipVm8|Czm>MUb0Y#IUF*P*u+o!JE#_y_~n#heD=14*N@ zRNgPvb4yeW(p)cB&O+7((r#E-yV-@u)@=CIkH2cN8e@BY^%<#iVDh|~P+FzXqQJ<`Kz6Y}yOClsmpHV`0$Hql zn6aYRg&Y6AVtdQSZH4#`Y|`!Dt}K>R6}Zj|LigB$L-$U==AUMw)f;?k!dZPHox8@X-1+D30t z2WaA!K)MLC0YKW*A=mbMSIox5(QfY+dFNDihJeM(vwkU?p|<`6*BoVMNm^!h9Qmzj zhPp9()m!8SJHOVA(@M+64)<#xQcNF3dZ?HtI-Q680nl9%v|YC53+-}kqb!lPJTxmZ zL(~N2yvV1FGtH;UhVb1!d5Zg9Aa19FoWL^PX3$yOtUBqR8j%~AF95}I2AU3xcVh2s zvf`=&mq|71I+^wnrv*Yc1KxmCK6EF1syqVOclBjO^8z(H@A zecP|ftA&D)Z-po5nq?iPXte|2( z@u^eiaBq-K#eUuzpT`iL``eJa+tuji8iVulr?N6fWB&DV7&Q`IcsE{0+0cb=66)tv zMOArai?)Iak$%p3`N2u`yvK^&$^-)SKZ2~ei^)~aV_r-28o?#MHKa=IBG<)Y7#%Pg z7Ef@^?zq`M|KI;G8IAvJ{KZuA*&8!qtuULg9#hOiiu9ojYf{j0VFH7URnXy9>{sY} zelG6d6oEB*TYWBYhGrr|6dEAoN-SB?uJ{NXCG?h(6`RA>1=NR_cGgWU6xN44_0Wlm z{cb~*-vRDPn0povwLNLE78*O;u(_^Pg6vpNJvK;g&B)-!gG;UBAZcQ=vRMg{#AF3( z5Zn$~>Yc1Wo{^rp?P4f*&7ImHy)4$jRk?Hz7e$g1n5P~fO%ASk?UYqPtAZ-$EfU}= z5lcC;)oFAp=$LO-pHTNm)955feSjzD0Q?-Bh1lJebdEp%Jrn%?{dCHA$N_fnb7KQ_ z)og}4MKO^4|CEYZ7SYIW<8M=C)0GqVg!FK)D*8#aDwDy1hLxA(7<5OrM7QzFNQ1lv zz}G9uRc-Q(3&3iwWX1oceHE8wMSlCI=#_XI94dOJCAuQ2R??#ZW_-FnBsHiC7<=30 z>!@_5Ir56|G$@1KpV0sn1}Oh{i}#5arhyNNp+b!A1NyEbz)S+IKGRznwx)2@P{q>r zozkC)v?Q%^+vQoY8BU(lP^D|uMuEhLF`2m}v?Q{FUL4XfsSB8Kk!bgM6&D71S`~Zf zUQIW>&Zhy?rSUY|k!1OT9Upj>Kg9BKcHWV4+J*I&g~czHg}Skcu~7w`t6Kb4ucU#G zjLQ$DH>MlrZk(LYt@3K%V^A4MuJp>@DYcxv@;V;2PqaiI`j3fMR zA)Y@>H$r3PUB7j~D|v^-bwd(2MUhAN&71{!!FT0ogciV);Fp|}v2Yw!FPU^$BFNh1 zwo$dI?+aO$-BN7w6mDE{Y@_K5*Y@4gK{+Ne4+ZLYxIh~cw7Bb;JQO&BxHh)$?iZaE z=L>F4SpJswp?u?%E_D}O@K!F$KH}N!vx=cD7@H%IG7=ksoo=w41tx2N7|F5eSt0Md z?sL;MNLDR|HjB1-_ra^!sqXYH60MulrEU|Ji7V8#&^(bEc96GY_G)OGNC~+|R|+3Q z9hQF_c5dDSmsV5^jf%!PrivtG)LGFisO{aJK z99-O7^eO`ru!~zN)FuUC3JPl>Fx{0qF%BvuMk(a%2tw=`;gMd$3L#_jHXWzrrlnKb z#NYl0NoKcna^p}laN`Xtp~$8fXlY18{fQpM3ew0&Hndg=at|K`LzKha58g{n=rZVR z(DrDssI^6zI;T~FT*O)|>U&1K=&X7^>*MSQs-F2R)}QE{qep+E|MFxLPU3zau#%*W zBcSIo>;>zT(3l1-z{uZZdN@z>-1@T=$(w@RhwS9bKb4RQ&@e#?z^t|^S}Lij-@83+s-{l z&7fRKF%=Xkr(({?j?6+I6%yECoeHSJGsahP>1*@@(8a*!@~0kKCf=NhT41X+4ImAc zFY1saE7p#`=!Na)2~nwYFm$$BQ%yg+4oUF)CC4YlNw3o&Y@e*y1JO0Cv~Uv4oq+3S z7R9MBXNBuA#pAU)ZU zKvw;O(TLx3DDEc3Gms+9O<*@h}VZr zM{6l2ks=AuQ4hT-Ngx2#$z3Ec`HNViv~QN)`OC#uy%~F7{;Qtj?DMQ%zWcg-h6O*k zZn-LE6G(C2$rL(PhJ*Wge78L6>f{P)31__|e$GBpNMor)OZ2*c!<-YU<+5)2raBJn zPM!c0M6qG1LEv3+?$WiYO!y#NZgcEKyTxqd0*)N>@8uT>Sr=dTx~qGtovM4@`z9ml zW=;e~zsmiWlWO{!B44y0>N8zFxaP0a{-HL%&((XZ4BMC+F`v%17oI&%6tT`J(MZ+t8cRF{z)lD$e^9V53T_qDQfsn+vR5dSFv> zF~;o2)aV)z&EC#2mMF3DdaEF}8aH8i=(9hYqpO7vNGG(tLqj{gc*1oiKN!z9GWDVy z;ElnQYUYI6z`S4=zgzh39DKd5{$gL!PiQT|>9*>6^~=7n_DiPq%e)GH7pb+@H|xe8 z$7M5*qmg2OjqMl}vz;3&I1IIudgZ;aPx+}qC3LR{)qvA^ABQ#Yp9I&?cty1^4N6K3 z!Zl$xp(SalPrPCcSvvC!_$FJpr~HuH0k2v*vo5ql@&Jxr5vK+jk8Gu{sM4Tkr2E zob~H7^Fv;DrB|dhgQzM*>a}~-pjpFdm0)z=<+F*EVXzYbR_=7|A=?13Vgbh%)c+;G z$7E=Jagi(_tJy{H-Pi=}Ff&0J6a)3XsZ`9TVFj8L{#juI_na`pyM|xGuM5Qm4zjf! z7Z!lg!2Tsz>cBNf7~LSvf`T~q27>)4+leT1lfnLHSl=-a{F56aSs1&qV{2n!d;x-P znIK1+5OzWF07yYMgbZbME)ZZW15)`)ppq|>-=(!Xq3K!|`8Zr#9)$ZBpaa{9p;#LO zTYN_2%Nu$DY~QPM-rB35!pfk#uRTuAd1sQ;G}9dY;Om#j8fz}C8{3MV zX7foV#iUVW6BT2uj%?!(&{=b9p~R?!t_mM2Fv8;KXD640X44?%8B{!Vs8Fd@f;z@7 zM>bnTKbu)zjmeQ+VH?p?Tk7Jcm_X9^6X{Zt$<7^dWBAmV!KavFz&|N~6;b3~D0fVN zVp=TE*M%FTsGtX#y`-S5sRzXskyxTK#PX4>KxmcwLz`K;sGG*>`&_yfbQ_U+6Y367 z6lbx(5OoKn6kI^#;qaRO8LEc;IWqWI-u=}*v%e`cfhIon%CAYn8v`0(K^&Hq+DtJi z6j@Kj+T( zNJ*=cyBQ>{DyL%7vOwbUNoLr9u>!!b`&_+gd%pRR*YcZ|r0;BE_U@a_1oqV7rsFw9 zyix54b+ZaM(AF%^4+bu&Honn8`BbSBRfi{fwhF5zjf%+_8hKW*fmy`WaC!Fp-(SDk zWmzWhhJeFHwF45vq(?j}aCqu*C=hxSW$9iyqn#0iA+Qq_&jxGM{+#>4ts9mm!)+_m zY#4zy{t@2`f!VMm#m zu9}3^jU^i7Cmbzbge8_nZAQ6y_Jf~VX7$|GblAw?J)YAY5Esy=*)6UX)^dJ-dtS}E zpS<&5$qGD|9VaFC+ym(<>Y$l zIEUpM)oJf0dMaS*N3d^w`>E*VuLjUD_*z2C{?ctt= zFc(-9X^VUz5*B9uwD!w)=KT#yZ{5}#wG99(?#US8;D>>hB1F5*zQ25m0j1U~RNm;( zWPm8u#?S_S32opNXx0UEi4ObkRK_p=%k_qEI$&i_uJFwHFMSBIV>>)g19qSOqcUm9?sUZIP_3||=Z=$v(o z91ytDX&n)QZ3A$`XILS4%*=l{@*T^}-W#&$v61sD|1KnGa_Mb?p?5y^{0CvF^1 zwNWbaEOd&b#?`PEh9EprBbSi`rc~U;&!(^WVQdv+sz@CsJ;pvm}sYV@uo^KsM+mv7`~W*Q&hGK%o3*8troKB&*-5U0|_#r)v*- zNx%2gS6dGE^`mTcUm0~6(uRpJZ1q9;g{=2$A)3(~eG$a8u+$%q^$8z~55UnsJX36N zrB{N(X%`r*`PHt)!b|G=zwoCY{dT;`UA_MQUHmhNcVl-I)>y;x`Wq<*!hmb37;OAT zmGM?dp7(KO<_JX&Q8C4S9rOlHg$lWA?g$@ygD=s{DbXZ@`a&YFo-X!V zh+J673cMy&VVIJ_H(seNp-;%F!V7!{J@CfQrl+kBo9C}DyU0S9;P=igC4b0 zs+5Bs#+wH{4AKlTpcwQxEYF&f7IK$#$&uHjgYF2)7d)aJOAoVylQsMD>SK;Tj~$$v zezYv{b&st*R^UI$isbPI>7I~Y@0^GR;fe|IuII*jad_+l#f$g3`UopfjMd!>Q@%P7 z@5cVAjX=DS#Asm`UCmEml7jAe-;xb@n{3T}??nQ~41uF#XCGFKM#9nGaO6JQSn2-e zyPDsa=AO`hn=pZNzA@E_ac1n+eH7C}k-NZ|AxNF0J?i_=A1FO17S6AMZoj7zT5?e2ico22ew?h&;vzks5VScZXpBdBx zMQ$xXIgKni1KxV&rQlTMv+Yc1CeBjl3$Q!mELPLPujSGmldu93f3F94b?`TGB9=!S zgl3Lp=+eM@wh4f`IHCiJXxsS4zb*D8?&50kLG|=S|D)0jkf&K5F+{WTSO(hCfjtmI z%omt@U~C}y>e*|AR}91M4ef;Sue(EhRDOYjkv3paYv!a+r~sX)O;aGqf*FU4^gVEK zlEW&2c6{`V5ZlonHXttDv~7U?AA4^C*Ho6}kNdTDkO&1@C5xd&>{wmh z)9ZBi%=9w7RL}HGOZRWA->+n)wEpF%OuZS5r1 zi5+4fh8wY1(o#$^MRrh;Nvc%W4%nO8LXIo8%}fC{am+YC$;*6M9d{pSt)3uTJUatY zfw$y%K%K(TOJ>vXvIU5vV{OBzV+9au^^KC7mMkZ)NF2_J7F8~83}GQ?r7(_%q1vQ5 zPlBPWVwkzn+~^O#Rfqj#10#q15~Fnd^B?t2mOw6HK=9yBs9fuoz$dds8hBHC&zHG8E})6Fv{@Q{}L;^G3Cqq7uTvx6U)iS z54Vz{X{6E2e%De=6-COiaPBU>+pCeE%Pr-ellI8t!D8ol7I6-7i|IPn1*Stj5ZDBj$hq#(PU9l11ypi)<@dwC5J(hic9(Dw+r8 zXceao=m(Ihx{%YtxwoK!J^%vp7Zw|ZStoBAVh!?%kO$L2gyc=S=8@c&Hq=(I>OQo>X!Bdo3x(T|cnsT7I$pICpLH9`*eAc5rNple^ z(>68tmsh{_vKMJw3LhW^GX`ehisn89%i@+@w9JBhJ{84zr$=HXWiy&73pfU)7XxH~DOTT;ToxmX z22yr*hJN%$rfNvm0iiBqs~%|69yOh^LrXifo`FF_Qz+H~QqTl0;f6y`)t#2`2mb zf0S*n`T6G@eZ~4~9xX`awCsHI=KtlbFIHh_A*YqW6t-oVs!T5={4_eFt&$6iColAB zLZ@xu2HW9iG+Xq4#wRr2-()EozOto^WV|vPJ+)>AuY_WNu{58G#5f9;WgO*d4ZATA zO-Xa_12R~2ZF5D36v#}{PO?JMJv7;(0xsq}cJR-7=Xo?sA@{(eGGGu4CKes<=V-F& z${>RtYBbaqRD=%6;_bT^Z8K~vXPr^bEk=Gqr=Jq&I@@&-$jO(AFfLX>yQ;&!X-?ToWU$S70E@Flx~ z#X3Sp^RwUmyl2SLq&RIr&5HRLdh;esc*ehVFd=5qD+?%wainGP zjK`}IEGhU+zu%WkCZ+Y8*Rx2;OIBZg*38V*Q4H9fN=T;_mHRJ^ci%JvQSwN!)abY( znHFemfp{@gwaO}mj}?y;Sm%n8_IMmCIw`9J7QSv-qwFrIq@(r8fOMf!NtB>I@bt`V zU=}an;?)X}rA^=@a4vF?4dW^ofC%g=&C)K2?q^PvSH+HYXW0Ijeoo8`S+_gI@7=Gp zTrxYYjj_Uxgj#OMwWcW%cDY{iix0<%C`F_>?Rq$L(wq|00NIRyi47?0r>!Gkw3%n$ zPp(X{+~{*!R9M-%t5(%}q9En!kVe@d1th8g_g>)YkWEZ#NTPg5ehF5%ecTeCdyuv^ zX&|u~9d`NF+YSk}h2pH_)6HWY|70mzCa+}GntAo)iHp-li^*;?4`egNY^2CKDzZs>0JJP{pTM9H zoGVD2K8Yjy>`QBlv3S;c@{ZUUi|;P2`PCOIVG)_vu;yHYJg`{I6$|3A`q1ITk<>{> zlPxfs@L!r(K#@foj$4`@wy(Gm49W+h1=J9uBv6 z_|sNM*#K2ryRvILR#2U?B5+-ajhP@Azs?mHcs+1(*1?_}nhi%!5~I8v8(uxk7o$Ar z;7hDtz4KZLzRN%MxSG}>4PUDMkgOUlI_kW?KiOlpHc6nEc#6bA2`6l`XqCeP1Bl(k z=}%mu1&809#3;rv1Z~6c1sz@oh3U78*n)BY?W$ee6~=`W{PYCu3AI7KvoD={P&6z zLuvpPn2P)V`&&;T_V%Z=PgO=vGHS#Zs zmu5iVt6111%@x&fI$&x@q1SVkHA#^(EmyQtazc<5lINDd-3+zgkc@p=d|HgVckw{g z*(fU(Cdx5(qvzk4StB2i?C}OszB=9>@1DibMY|kBvRWlZ)KZ~DbiyEFXD4i*Jf4p2 zI1jA5FQaYMCI8X2{bk>(Lv^%X`X*$W>C~y5I&za6Gi~*>X7%5{f=qk$fgM2Cna2GD zBrGO8=r)fDOLW-W??fB2LI<_~wePk`Oy)&+SNI-@b7J$7Z)RRnDF(tPJE=%C3_4}4 zH;@2e#xm%s<%<3;KFZa@T!M@;<^Gt?d<_i9O8?@p8eSGStdrt!u!M-of=mN41DR!KsK{!1J-HBANFyYU$$R1FYLD%rI_?8mnP{wg zn3{ht?3TDh+&QzC_n247hl1bv-KvYiD?nbnq*fEs)ED+kqI#)$H_0l%H_bDjPUGK=cmUtN5%|1x-y25_~aeu|hED4%g7{ zG!R%N`(g#mO-0$FY=P#c;;b--GeGcetE5e=V9DY+%Jyo`Z~EVjSQFp?gqtlQ>2TEY!i0ypGr<(OsY3L$y2MW z465?dZ1ilBHhb>!P4euN_0T9GkmQ*Kk+ml-1t99G!K9#4Ziz5j5G{bbLu1Nn0T)*( zHy0e>gC1FQ?*gQ|YnOKiH%WJUN5MiEBXii|0@EjN6mA7P8e0qy&3e*6Ya`eE z12$RZew~jD4cot_FJ{ooEa|WT-6)2MehxV*cEdbe7F$2WN!+&cJp-WN6`8x^-Bpu+mlUYKY7!7q4V zxzCXwS_Xbzs(LXi!Jta|B208KGb@Fi!3|LD*bQtLdp#}!Jd%AYIfL$aU#kSeC9H6Z zCq2Sz(&B(dDJoP|^NYNY=Pg!*bn99r(ym~AbF+FSC!MSS09CjpEZV3zr*3rZ<8BK% z7?3lk)n23s_JYZ?x%36&Z{uz2rct*2(AU0LiX1?e9(G)|0xCn05M6Ufp_inFOnR9+ z44iFXv2ld$*z7x@$w`(=c{US}^G>bSs6Fn=vqDOx{hS2f)R1-P8J!g#QY85PFMz@} zOiX@8$Cby@jF&xM3gdgoJJ2Q13)M79(>xLu+1*dD+ayoOtRpo#v`^1s-mfNTr2 zp7|E`I(jh>SwV~1`pql9__7IF?_58dOb)ZluW{lPkj@NcsqduxqIpTS zWP_}Op9q_rx1}BY`?5ZHpZqc(o!=|+Y|cTF01!O&othexV-#)I?MND zcrPz0u)=NlsR5wTzb5UHWBdSn$WmROU$k)DbgeS$CqrL9@U6xrxA=GE*>m-hI39kk zR*7jg6R?Mlh2D!HX065B$NtgJGKbG;Hzus)>-B?$xGe4i`w|fNE{tIL~I~(J~PVPZ78HViQjh0kk&&TSV_xHEPTV@0|sBkE2B$*JrE%`=N1RxkoQZng_ zkkp&}PjCSfauMY@;!G8){6s~prAGTRe)x`wuZd%O%k_!+amP8HQAltHT~yY za*N#z<-FTzWWAZP)FX=doFe^H9>?vbX8wMwkuf8v7D^Of{TV1v+P(Q(9>-(Cs7NweaUc`M&&m6Xm}&w1j4 zy`2@NbF=ARPaK#pFuUWt{pYX!{+*d7_~{o1Una@ZNSWERw~t~VojVgay(Bvp?c*5# zMv9#rzg08)!%8J}-Y62Lg-vxmA6w|uMgDQ5KCnFqXTk*EUK-osJ&48fPOHuaIq1s| z-_8lX=5c?&U-Y2$@Bb)47w28ro=LeB{j%%EV4mm-s#D%UqGsv&pin|mdDVd}1lCs&>zwrBi3Ba{TbhLB9srYe?r7`Ck#+8B%P4V||*iFNW9w=im$b&9oBoY z?;UpN;BNY4V6Mrmy#GO0IoZq3MRww~)-f|9^by5CRHulFYzT{a{eauOFwJ$kXl4f= z^Aj_{=QX;@PkB>OxHrLfNCv@z1e23N6FTYg`X(*LrnRA z&ro2D0{f^8LhLh86vWs;$9Y{bQf~&G zQi>^|NFf!O=UG0tC18;LTv`Ir{(9~muZw{-%Gii_Q7k8kq=`22x@2Y4r*xVqV{xIV z3+z(|e~09;-zCA2yhJu6$3)d$X#qK-s0iwk9jBWBcbk~Dphj+qtWL29#N_R3z3hmb zaie6k2Ro>le|Xthm^=9AL!x{$>1JgduS*`wL7M+uQAcnOlu&JAkdFHbC{o>V?GV<3 z&l4Y}V{!#hDgz@P%fWat*4LFvlfPZ?^)84z4!SqZyaxC5{BHYa-c5AnHf&Hv8_Ee? zV>MMx8CWyeF_D^XC$=|MYP#=EFI;pNHomjDSkRE9%u-!-M=j1>z8ECgrObePSLg@C zzE)$*A!TP$#_YRtjqaF{`(nA0=p%t&_KqMm|>Cp|jD!uTi9kJBXfN z7QP4AwlS)jDDRS=;eQ4$Y>Z;ftbR^Cw~(%I+d93FzPI4-!VU6Mq+a9@BQhZ@I}p;W zuuMg3e>F6bbWSIpd92iN?xox0+OP&F{c3<3f+5*Kzol*RmAs`60g1Vpu{g64K^9Lr z{&ePsOZ|9H7H339_6r>7k$D+NNwtH$Q=Uid55&1NyitsBLrq zvM;p4?Sti)-#QR}Q<5+n`2VG6AY<=}xPyP|t%D-1GS0JN_Ce8cSv80&X3*C>Tfh2% z^h)n>YXj{IEgNBDF?)`BYF8{BznHtg+cf*AHm$27yI+|_GjNdb{_DNpY5wr<|F!gg z#Ca40;_}&4WXw0vc}<3`swhD(6z+Db&M{l&VGd_D{jsc9oB$b~8%RHZs1geQ&N0XN zXGDX*Gtub2LsCal=oqLwxv=;#@K4}#_C?9Fp`9>Zo^@r{qn!Dl-1afy7FT}=B3Fn}VrU?VaL3Dg~){CPi;;T4u&Vvcp>GPSmF|>=w8T4an%s@Dxqn(i_BnBIEV!6p!>x?_HjsUN6)BBl+i9) zMFg80tk5&HF0pm?bQ5~6{hzkKl66juo@_JpBvA|)zpYf{F;~3={M zhmdi%SqVbDShs>{*>}9|K>k3nq+$|E&sgE%**wB(Y{%2sIvz&J+MODD{=K`Fg&$7q ztXdJL$Lyfkpqjcy>+JFbSBr8KgIXUxLS5oaT1+NYk|1J?@T z$t9n=fk;n}>f zA;kb$>prN%Czk|8-2ON9-a9!dqEy#>&I-YCx*AfDu~0a|p(!Gk!f237k1`ki<&hEa{sV znZF>sB(5eacvX^K-dZShcf^>RBjEWnn=tYGV-CNv@imW_@0sk#M?dv@hn#w4?8tR9 zJJL!qz^ibMiahIGCyo_8R+Ne|>HT5tuUVcqt99xJ0Y|x4)qAKL-_S3?RQ(isBZFkE z*!IHWOMcDjT>+mj*`+KOmn}M{2L7rqco*i7SE%BHCZO=GiRX zW>5w{!>Jd=K>=EwxIPeB$X0UNfp_u3;!@EG((BbD+d|g5o`v&Ee#LYFkl3}qW;;%8 zLf6O{Y+}bYqq0zb;o-~f_-+mA2sUZXRMYX{_rmVEr}D}c6)%B$)f+;+q)8edh9{Ud z-NBE6RO@DBh>e=QeE2`x$I<2&YiKy(QD^sYHn(Qy$!xqvfBBCl3-m9YX9($IXMvn} zK^<>a!n$%Q6E&4QrXu^~n;6Jpnt4%}Nq0lRe+>sEL@`CS8Mq5Eez(&D>-@`=B|eC% zO^lJ~w}-9_*ux+xF(%m#$v$$;72Wj2aIjXHASzoFwLnkzg_SKr-7fT@Q|JOiA6li3 z$rqyAjm!8}NsghMB;KRaH&KqCfuD-;KRu*Blr8F#9|%8TNa(Gk4~wsn0m+~TuJ5vG zB;LgKS|vWQ9sIA&yq%Kxuu8g#Ukd!hc!WB=(IR{7O-I}u_Sr}=OL7SZ1AX8 zWxKt<{KG76o9rIcs$oRAPma*+m2UMMf;>{KGFq@xG9ZZ-U}wlIY+s`{VN>^P!n4_o z+0sy5TlY=N362d7q7%E+R^mXo<%kS0^^y*{gTK`k@`vd@ZrW^}`XOhZ;CIdH-+kh_ z333!Ve6P*>-8prI8(u0~)awr?_Qizd_)GvJmR;HP{Kw5)&(};7oZtVeSW0%Vo8X+- zv>Y;XvG-C8gpl`8kpu2cLC9f%%nF~&az#jkUYT7TvT}iP%#)1yW(R+XGIr|Wq-p2m1DoUc#G6IaiF-vZ19^Y zrimh_sYvvK`(()>`^RR?yznMGFu>3n5SH|!oDAU+nJVhTC=Zzamx z#25}lPK@|Bi{`-&6w*ON6wO0c<{D0}pu)RJx=(y0pwTr~aEIS3J?!5@qX&W$&>_X3 zM;xyPpw>ul^+@Hk%k}W7M*j!>k%WaAbSrFso?@bbREf0#nkZ17O*>;Bjlh6ShwNPJ zT;@MAe{O=6e_A`3a3_XUzZtBqQH-7<9aQAKuU`fNl~ZJ$JWG{A7lS%mofth0O!3z$ z)43J1@A7mY;METal}O;KQ}3R4$*+*^pbNb^_-P)eRX65tn5$FwbM%sXio@ZzWJmly zQN{vgS$z0|fL(Jy>4SHbYa}t#sf$4%ESrN>*IM7Ax%cUdfiZ#j>GdLG55~^}Yenl} zpW~Q&7NmH_a&G%WEk6(h8|TP8W+#)tiw}zrujkaOj)mR|9ut0(1+~$DI2PtuL5)iK zcJezfJI6;cT@i$~#UR7H*>#hjz2KDPz5C>-X5h!+Kx# zyu%6}Q%mx1_js8sQS*NsT1JxC`9eAZbj~ zF|qMvkF1dQMIYU4M`O=D|m%G44 z9=or!)0RP&2#$XEt_eRMY+m~&x#q-P&uX(!;Q+Id?D`U3wneZs-9rI zWNC%lC2{$zGh}I$+ghqoc9%EglIyq5bI1jER8#y@{I$x}Z3ny1w8n7}p)a!~qcQ?~K_rB)*|B$s#JkNnsJfehh2gM{(WD6BJ zOb?9PSeVdWt->_ZOx34e*WY*mWUF}1fu`a4L16=p=bv`m6;{xglI8K|*{_?R;UZF% zlFV0TxdmkEBY<>}Vu~nIKt)~#n))3;KBZk)NIwX#<#)+z!gE54I9c55tR~M%EU$Aop z_k;;&pSk^Z1{}TW(n^jAH@`j6f1IowEtTDgHzv}|(6ND1@cRm^tx7||BDJ8s)%-;y`-x6N)=7XXpx z(uO5S@7cj`;n;mY)J8--vl>>9>TmUa!p%zu6a1 zrqt+^ecW^ODydH9ut{j*cHu<>Kfe3y;PUb5=AXWbJ_%McG%GX0KAYLC&?>t)ZEl9G zB@B@`E&;;IfZ2=`E6+RXvUQ{||28PLEV^WqfaS#5qgKi>n}9anxLnpinKvii<5S^k zsO8mM1S3@9(+9hj{bVwi$~I#Ly$weJWR!m2+y9|>iplFxrI-GRJa(}TqfzE3C*EpI zHuF(7QOtUZtO4F{;0QvE&;~l{t>qu=nXH{@i+67W9xR@ioYT@TmH2(Jww`w=wvZ;} zX>qD}K-?#*U06+;)!R7@q8*A0^O}@+r35&~VnK4}{LEWG-(9DO<<$5Lxf}_o6!r** zTpEEGw#T8qtqoYQx<+xNE!ke!79=&4ibN7eU%kmi5{=@XVk~#P0mob{dst(BC%bp9HioIS{UQN=}>5iUk#tjk&+ZEmN zpR5;|K$6Da)=pyCfy9Z!XZy`SqNNy^adtp13lxW;aE4 zQjsWUhbc8H1i2!tFBy~@(gzB?+T0rX*XL`n{u+e?ioy#)IToZLfa%a-%ifV=JPwFV zinr<}Jo^K=2_oC$13x3%*g?dJCmIkG83B7DA5ey!l#!6L?{V1j zLv@zKQ?JN6*9uKls{|_>!DI}2T!Hm1u$`rd^pbY6+XD^;+_(6mBT_Y8>oqx(3=Ah* zY}gk8)_Xf7$Ov9+G9u0Yx+;^DIdOcw#mp?6pcoLbItmE@^tPF+IQzkOs0l|Nr_eEQ|@P2d2%xt|TAsQVu`S3*_(iLi_k(ZN<<` z9MOgz*unZs|2FU^ODdw5O6DG-CL_Gtc#4Uo$ObC1SqUyHT}3olE)*>|wCFhf7~ENV z4L*ypvJEFJIp1zJDI2FuME%T(<0)3u&(MQ0aJ#2Ll}W3lEObC(SR!jB`ykbz z50aFSv;$jA8h~tr%`;PCCse$!n_vH*#l4n93r?FPZG}1^N^p^rNtXxqKx{0F)*y)s znh65}Vj;b`D4>%*Pm)8hq~o}_o~J4FLb8={d!#RXsZSfQ9rfYyf7*eA4y+rCQ8w(R z=KL&o$5higXw{l|_2h|*(-tQY2$+qSKsHm%MvAP13`L`8UxM!@CWZdg3#*eS?Ylp6 z5N$!k$o7+U%nlJhUiHVNFS`<+-n$-ZW;*y;u7$E5x?R2o%G9wG{;Ir|*DdQ}u87YH zi>LR{DARn0UKxSH@hSB7*Q0?&;S{NZ7;M_?4nFSaTv+@_+$de);~;;;7RU@64r@Oh z0V7tBp&qaNdC4zLMr6SX{TrmsiGwK*&5TG7#oVAsH&AedkzEM#m+TekoZTN^2 zg3^Evi0|Tu;3cl{b5cVPowWfybQM%RV;DvgO$NEKZ$wXD7AwM!#ZSO!pjkqcJ@94? zhpexU04^)MO_6=;`!g-Knw{3Mv9fs_&1A}dUmtdEQMTKE<1yBJ>C}B(00oVKJ)Iis zJ29wdY;=s^&v;S!EQil+;|0lx@pQ-`E09o!f0q=)Ga041N&kGDtRAf%!+C>6B*V;5 zZKIg26p6#KuN3+YeNa}Y>gQY%cS=tzSxFsTlCG-cU*P1^yIdU(5A(6L9W3V0Ir_{x zSdQ1M+~aNnM%h1n`z+bOZvW7U*UUL)kV&SP9TbdDq;Y|?E#zQ;LBFOlC{aG>Ug4b# zMe^OUvPH=JbXJ%IYph9FHVu!v?YJ?{E5pw@`ltBuU*4Ta0)i786e|e`CskTiXYiBY zb`meT&pi&@Oz|Mas#Bv#+O8R$vOX^OGln~T^6~}18|9H7Iw?kmEkqC~j)n*J|MxA6 zn=REho!0%ZqNkY&A^QDc`@%c;rQCB;l*oJ{EuYl`3YwKctwBZebZU&IBn8xQ?{N=< z;6$>rj=Nn18b`unx{NC2?4dJ5S_0w(Wz@#7LeW5&g8(re2$=h!jsb-A*=3X<3swL4 zhljpcS#M3!4*pVHg*+g!oX>*~ge=tqRr`U3J7*yM*9Oj3|3d!<`w9@4+!=aTkv?N0?j^zm*&Ik&mX4^Ja7Taf+#@NG%na3fwR#=&?pn1 zKQ?*j)O!Q^IZO4Sr4dW9rz@dp+&t1{=uI!H;SKUv^SU^#YQ3yJ#R9GgN-q8`(In=T@opg*ZbbMukz%$`WHS|M z6mrynZaffN8kHS&$~K{<*2^8klLbiAfDX-!>}*wv<8)0l2E!_k){) zD!jF-Ub;otEz@*)MFVRk9`-@@0+cZC^|)ww5ceAFze-{Onz4+Z-QC~CzjFLG*ljk5NjMj7~G5Yg}Sy0;)# zP%1TMB%q9lA^X;W6zwKNj2lx9MbwljfB5!?RUsyn$o{O^O$ykd#EBF4KQTi|6~(}^ zyNrswLs~eRSY8L;NNCUvi{>=7X3Us@!>G4mQ_@>&Fs~mp_q>;a)OFnqiCTs=*!|}b;8?iz#fCUE9>Nq(xEwz z6a!4R3XXwbMyu>6UGzuFC&HVOE^<0#+uYNn!fhq#l*Tftu65#anVwGNotGT(Y!r5@ zpFLOM_GoUtFeB^~2$H2h@pu~dmOMMOTUO(vQ==qI0cbK@mu0DP2neS|RC0ilc$_)+ z4jYPRLp83iJ*x@Ad4r_VVm@~uQrp-PVC|2nL#~;Vs=p^3Fo52vmjqT-J_7x z>2+c86PGq&g6|U-+?*~ByXlGb9xe0(DgF=o!oaU$e+PC{2#p82qjqHHi>#nKHSFW~ zOv?-@HdRDUyj^4^p9)jxn)tBjF|Dk3&DivYW4++f2iPY3jQ*%YufF{=waSv4%L)Zn zZ9rAvK9H8%DXA3p&8hLxpxm=j8P1VfJ){5gZF--EX<_u|9d(u6_t9x1Z|`rt{XNPw zl|1%%e3}fgBTsQ+Ljm*5h{#(k#cZI+S}IZxRFF_(h<|~O{r{Gku&Kt-QP>@;-@fdB zTlcb)4Br6PIe$r~eC50rI4a__L7iY7cdy3)X$Ruz0{H*x!bzPQY&T*PUTusuw$I&T zWyGkSzV+Xl2(gY6+X^e1F0l4h^~q~Vb)X(R+%*CNpOYt zMo%o`s0O{WC&6j%eZYs03I8Ypa~5P3mYJn9J~s*zxxfZl&lXS5jwbTiV?FEfJzd^J zYJ8kHY+|Lxr!RQW-KZUvJ$EMu<)PPjLO!k`2^I)bupkzK^89Yu?Rsx% z`^tZ?Ooe?#Rwq^xXItfWTodJ6r`Pfle7E`=wr|h@0iB{hFaHoCX9>PtO7yZ@C40O- zS77}d`*>Qc>ejyQ74{LtOVh4_K2GInXrkz zL(Y4iSZDyl=aQI+osvQMX16-t$3QQ8bHM@56PFCJo}5_NBV8wTI30@d-H#Uw<|AkS zGwXgz>GGo?mi5I>o7QHfmN-|ShYA66ott&;OluCq&ju!ynNqdJ~w17k_3`1(W!UMJO(sQpxP+t4_oE8DWsm0 z8dBkQC8R~NN_r&(wfxcyQeA`YIesm`UC;g&W#n}bIVfjBEH4`bTmrv(h{a>~k_ zQ18C`#%;?9#)=)U3hWnj1670ucRO1pC`96r7yF{OV9SetQ6G-D&+5fH@47Jk$A^V) z+m;T3>j2ycPju7IzquGnvLPY#iHp{^QG#h&$)bB<2g&s}G#wrv%`24E0e?)Y=!r`+ zClNUGs_9tKfTU{nkjpB+Ex^^5H{&SxV8DgBt&%}^O{+ZS^^=k(E|`*^D8D(cQC8w( zSca|&iB;%Gz7V4it&-kwO;!lXEMsC=t7KKgMu=dICm3N1qAb3-<2hvwx>p2I179CJ zXvt(?g)ksmfYtR7yv`M+(5ro0LoegeR5$NDINe3j= zVAbj+S3;I`@Ugt68^nQ))D}PhaK&r47itn`&srtj1?#C@%4O@Pcll(hc6({?cde#d zq3$zLzCpUfV_B)B&U=TX(Uh6CT~y>1PkMxj3$BEkX>?B#-fYF_Q(UsboAXA%Z%coj zG!ZEl=XLOi6)Bb)UYlZVV4-M`P7XnNZLRY9d~l*jg?9(t$4#5W3RT3r8DzVViIM=ob zHvw;oVV@^jFzEjG#VFA3NXL!E;6HDn960=}tlLzXX4%}YnB1U8+_ZN{CA++8=XLPN zMYCArX^H_Y?ccB@&=f+s)YFv=g0J_Coa{hde2?TTt7XnmsG2A0bYIF z5}$i>G(+-Bf^?4y;QDD|MV($fbS@xFQyS1qSF1jQjpa;LG8FaPVTaYZ^KK_u z5*jZSbg9#d5OUq6Pi_TfPid;+iyvgIO(F$pbJNSd#b! z1Xj-V?*}e2O;Yank0+8+Cr-3KXEyCLP)r>~YN*Kfmt#(S2On|`wsTep3?y-@eT}U8 zxWL;w{S=cRIuF%3A1pU6{L%>SL>}cKn>*$p{AKy0hy!k&UI)o0SO~70`NZX~%Xdg_ z24h&DinPj)x#}gmTptDl0e|3q@miQt_R(cb2Yr416PIo{GyX{p+d3|W%ZVACF{YQ1 zx7b)4Q(IO>XNp_h$6CugTc@3!tYpq2N3p?~$P{ekPO2b>+3?r^2($K+b;1e|Q{&%j z6D~KImJ2_OSwQZw^M9Rq5xr&V$i(YaQ<)W0$q*G8Gi?ir3ArEE1*CREE-};Y(OXEg zN4~G|EO}P%f}UXGaURL$qy%i{-U?VfEyZ6?_XKyz*G${U%M4vR?Xvh9P<_NqGxl?j zUNfy!p2yXxK><-%COj|fc8^!xpu1JZjk$_%pa}apbvo2EVksSN)>V8%XSk+Q7s@ts zjlH5WVm4>n>{Vjp@73$xB8m1`4Zq`>Y1kt^W$bDA8J+q7z1tVBUo!ND*H_NFCs{j9 zr(Q85W?G|c(C4J8I&k&0EK>IKU)uoMk@jfPjz7%mV18nPq4uWWF**0jz_8D3KGai8 z2SqM}E)@*dBk`k%o|)I@m&mRzY!P0Um4FgNs$0x7>-SbqOO#_lei3KnlLO2(&ztj3 zk^cFR3KCJw8SY7^?xnFW@Z!ss+?RFm@uA0lXB0IYYz$}<4#*k+TaB)5!aJ}Vw?{o7 zX`rzc;^#P>9D#RF_*K&z6=wkNwY)ZA1AWJ9?X(woVgGwL8{5oI`^kULv?L39MV#*Q zW;lDwc@)Eg?Nq6ucIry_xkWmq*0?0E6KB$A0v~>Pz&*=Xs~pxsYZI^V%!aaR&ACPS zp&fLF%CLNImm{^A7AmYUDwW`S2G)F7k6**V-I>!w)9JNOut9Jp@NuQ zk+@Z|!NVXge$elH==jn>CId=igJ$%*3ts!_R)NW6{l0kmpU6fhcFJ?jz_E*Bl5paV zEE1LbHwle&SiQjR4HPdCm-$!3`{}##?78>2`$COZcc^a3U+|3|{Uwd(D%$}c&u5J1 z8CH%J^~2YD+uk+7-z3-A&1lZM??P6aIaUJ{bDtu;RAde(6(||B%4c-&XA~P5 z`xpPJbA2iw)9p&~wr~oqy(2Yz0yiL&T~bGfp8oj!n@0s;ie#R*c(k3{>osynmNchehrx8I_SzE zonjRvL1)ohr?_`1f)SCw|`luUd1~Sa9VtdZ1zQ# z<~tRvub;(zLLpHr+ zN)r3bV6dBFc2Z>UhYbhz!jE-?j8<$JRX6=tFFO}qnR~tDx}xQ^;h)XwO5p=? z6_RNh=<^}Deu=!6*RV!Zt3(;-GvLrpj!78@LhJ>CG5cYK5Ng@1oxImgcEv@cDkYht zc_L06b~|cjYYtLO5k(4+!abYDto3Ec`hf)!xFrrD_=_Y_em+PqG3M}L_Iy4_zh=|L zq+7LYoqU}f82}Q1!Z`(UaP|eH5lr~g#Lp=4$sm~!;>z`fDy#*CUX8N%?Fh4RU}+4S zGbTJ(-L#G+A!pvDZ<=s3weQ!rNWBx!N1bLEIY%*PDRPF2?1#7rXnK;USz2WW>E#W% zbdYL#$mPP~`{5Z7SKa1bB2E)m^6N!?5IAiM+BUOT-6*LSf&VJXG}xPm^lITn&SP#L zHx=S5_q>aNto4c*c)($YaOo-1<%v?EQ35QvsDSdEYDqR|+&A*i&1jVL2$wd%PGXko z`urT#*?`@i-73@<#oy-Ef;wIbJ;aiW#=aOD170kNMXY0Jlo^XUc~^ zM^q@k8ZF2X=Xq*U{67O)M3n074%Mnq7W0whG`TzP6F%I`3(Zg+Qos)00QtlN&Nanh z*$}uM^gy2k&h7YJ9`OaU!;p;V&3=chhQxVo|2KbhdcCFWp3}x~tqAF*Ovf$bREWhw zX84)FL3xfUn)LD4FGPSIb2aQZ=YaH1WzZov$AI*CSh0rf^IbUP3Om^D`L@f?U-lVl zpCCB`{n|b#8r5LDwoi6kmLkGZQ_v5o3`p_K6nvt*>wXq;s-Cz+2~d&3C}!U!hm=yR z|1&Z-H$kSTUQ)nq6AnM31)=H_3v)%FDga7-alA6RHV~5_`iZ87S1LKEIH_8}d8~ky zn%-Me8PvgttOuViWvmGI!&7(a?m`aMh~REB#WYdmG|1lYqIpH!a^XtB9`(rMMER%mHf|13BW-hU zUZ_*IaC+&)kVbw*(EhNEyd1?!K`QuH$-X`Qxqf+`rJR+5$Goi2oX~Rr72xZf5bTij z(r3S21F2X#b=jg4zxS6Xy?sgCEz9Eeh3V97Z(SD0a?X?YH5cEyxJ0MM<9^{I@fEM( zHkaROT(bQ3Td}NmA|@4tHbZdAlnZly?rF)W_KM_SS>akMaDUE0*Vu$qyp26aOJ$u$^Nj~fIGmTVY}Ng!Pm}az|kj} z;K$kgI4iG($_T#p|18%^ugLUbWo3n&zP+SLx=IR+tzGh+l2c@zA50#FUT1?eIh+Af z!x?l3@(S)+aIwu8Dr_DC)(zQW$>8I%{uK5P#NYVp?w6fc{*dAT_aX;p1Wjs)ku?Q* zo{PhPCMA=umS_&EHqDsK$%5rzvmdo_KEcXxOi@by0if_Jfh5wBO&7qClW|(!Z zh(95E$xi=l?gOZ{Y?Ph}aO`Zu5-B5BzT?DzZI7`+3iTg9Y_GD+dUD!uoRvH#6r(Q` zRjAHO65Nl0&#{-T6h#SI0xCh$9*Jk`KoNT%Afz&24-+--(PX9nvj>Q613b}I>;Q89 zozHH+?50EqrfE_%ff(>5zXKr}6#q^4sBmi&9u8kG>*Az) zU^KT>qN(-P!y2|v?zqJ(d!&rCCXVx7Ss`Ug*!*A?4|n14~b!UWGL4%YbR}A z#{w=dMykaFlXS`osZ({oKeN-vuTX z5JG~)mr|*NeA^ep%Y5c}{>(9W)BVn+%QCmtY28#Sd9>(5Ah8P+-N8~9B`O@}QH(VT zwv6#u&pYZKI}}4-1JsA$fb)t>o<^EN(AMw8s_kU3pJ%?}9ZPOet&Tw3DPc)si7uj^MKR4!uu&VX<9N#Tq&mp3BHndt!(k zbaFbsKKQcNi0y@3!M}CN_XC|95d6KipVtjV8#_Tb>6UnWHD7Os6DInXwl3u+a*|$mBL7(^O1jo5ma&;k=uV~N&HY-u) zz}NxqWf;=mFr_8YKqi;%(}o<@!z&rHVQe}2>V)XN^it#H%R z_r-mn9*{!cd@Tjy_;##bXE6vijWLVYXdHoD%I zLIa%7p15SVXSg@YYJ7~9_kDmVRIY=%wgKKc!8NJYW6ao}qhMnjx~zkXm6@8d;=pHm zo(W`glm7WQkgYpyrUilA$r06H+bCu$MdGMPJ$=Wu7*t1EC1kg6foy=9|E~TF1M7kjhi5^QKpN}Ky}xF#d`j= z1*LvE^`QH1uOk6k)qZXg_YU=Bm2{-9}+AujOD~KDbRArC7Njo&_f0u2wp07d$OCn*dWU`Gh($BD}b>=7WG=f zP~sO$%HqJ1ba9I;M|B{ee}3;AB-JSAoN$fjJrWPOTow<6rMdUXce(1rhg^DuwY*Zl zbnz)(9dAfpqBLF|;(a9EODE4f5T5OZcQ!HyI4S-&gOek4>N~Jaq2*#dNhQBmd_|l; zqmPS4*KI+K+-`_C8t-R{?#i(kVXbFo$dJnbIScv2Idg8pbc$_tYGg^=6>uqVAbj|f zhg=fnYurja?zu-%D+G=F&x9vj@5_c<(%cWv!QXj@WG7?o_r1FCo_BT75oN;c{25NW zl27HQ^0*6j$4kpaUEa%IyDUx>==mFYT|STf2FRuGbb?_? zc1DSnrLb(w&Z1d9?YQ%dug3NN+%)_7r?vMJ9lO<=6N97QY_7aUF)(>{04x0Wnhj;| zb)*YwoxAyQB$m@0iee+1pswH)vua+ZqH94G_d296SGcWSutDA>E2FN=Ze?0{$9$UA z$9|Mb?f&VC?>7I-Id#$E{foPl_{@WEtak5OP)llHUd$7(Rez>FD2`jyB|GM0taI7w znXYOMZB{q`VCj>$^Z$8i^AFCcb?OxVI3C8x<9JD`;r2!vd7b(gw~m|s&llcG|IgI7 z^h>Y|J6%=D2dX~keN0_lc2KiAF^=CHmJLWAg0lU@?UaKsVSTV|yu-?&ni4$g58?mUWUQ_) z+Wb1X?!>P88ngY4&nX7%Q6DIrc;OoVBA};Otp}#Feh5Q87K%ry5$2?g0rtAZihUeYBITkaUj*`p-{XayUEvo0dcY@!3O;ro{EQ z3bKgxQg{x62UurQEnCHC3b?0dJ`!iskK8pI zJwFy=yLF!Z^4tFV=v=?!;pZ5q^^4eHuf1o*x8E{hZ^80UJ4nSVGf}phO_ZNd%t?wg zP?49!r^Q`z+yKdPJt<3?8w17Ko%90_of;F$y5#9RyQkJJcHMATIhIrjH1GA>&$ij(WIQ89(G)u%2^sWSEokWkqWmxl8%UvfU&$;(H+|8 zePiY+#W9i&WnP8hm_=}C(P{D3-~*&0s3BsUM$kc)Z^Z0qqp7oKR(d;Yx8?qt)6Rxg zcFYzRk|eaNpc{c8ST% ze7F8b2S~XSPjMH_49CY5bAlqrFidHrhcRuP$EczVWJ1MZ8{J|#dJp*GgTylKgJ44x zU+$8B5GB~iv`S9T)GD`nUiOVP@OCBzHU*(j%1(dW^zY!K98x!E!95LMqj|tu&?hbs zRuAlc6F1p2Kz${f-ozLi>&NXq*$k%gO&}%a% z*Y{CGr&pu!y5Iuv2V*wJ)rFc~uLEwa@-BI%AeM8?r)b`=S(=2=WeEl&c@4YKb=nP# zAOE~M$FhclO~}cK_k^ugeE3EC<5sN}(2Zsdx7un0ljdZDtX#JW&+f~5EqaBNG1e%n?7NDjXc(t#YjvhN_bcJ9*J|}886>##!ICb zs1n>sMdD&V!B-y&+0|~CSyHwLjUjHn!qOkCV6I?CKxF{xlP1bfK-$<^x3!cb$NK3= zjAziFKJ35)R(5X6)bIL-Rxf^+Jt>QZiUcr~qu?SXi1BfpFEMVw<|MJ>W1mlD%?fh=}L zi_r0CVTqDq|I@)3nKFeHN}PAUXp8r|HPcLfh+6-94O!vDen^Vhv=L7+u@u=rMZ!YO ztHcLY05!Px0NNU`_fWOa@z{J0WVV6zxyKxSffZP%W>)OIM4K?t|1Z9$Nc1$4YWC{4 zP|RkEY@{L|MZiiEB&D~9C-`>3drzF7IPa=_@)O754Za%h-f0)Nq1BRhwFwmZkXbv) z5hpejpP8Zc6vZ5;NIh^?aI5KN9wdl)ZFF7j*Fdl2ZS}hb84+4#i*PlhgdsybDi>o( zXt#O`DGWEZO!mzPx%tL=&OzCB(Q5UOe23(*p>;3ayEs)2sbb;X>P9J$f*JZ9W}0u5 zZdTwuC=hddG`E+mhiBU54X`wRK*n#IY=immG|2krxJ?l&6GN3$oL|5(q3O3L`j3;9 z?DmzMI8>5mhP|y66GxFvROAWLE7mDfbFwl)6Vnd#Hwzte>1!@w1_|9YfEa znT(^a-oFy%Z$d@GSGJUqj8_H~wPvU&p%`$*^Qp)dQV?-;)*fF7o_sA@FzAkf|0gba z3-g7KJP-ThaR&X-ydG)aTvS@uDv@9Xx09avAcu=1%JEPEK{l|H7Zc(cwJ)qD_gSzTQia#C6 zcl@$!yigeT7~A-u^;xL)(Uw0gHGyPFRYVABXJ=cScL_(Hn8D{R#oVUIO)9dQUQh0L zH9`q`I;`2x`L_6Oj5rXUJR_c0O%F&KASuRlpKSKs=C_fzKMdQH`^WR*0(ENK0mIbx zlfE4uA1PPO9CE><{UH^Q7K}$7!g@}H8y+19uT!M>BL@o-S?{MYvHtI0iWOA`doe4aNWbq} zhFzY=gcG|lR$@ejki^m^T)P00FuUZ3WP2FR>438V_xNW$Cu8r%Z18LX1=@VGryPG` z9Ta~UlK<$7wF%hhnl8>z@M#q$-0lTv<8E7xEK%56# zW1F3l0m*)c1zsiw7RMPNecQg@f3c`TjYQ>I<-Pz-)Pg1{a?$SOR7U%dH1Cu!MN|olfDKUkR4S@?ty_9k zokCZ<)<6&6tCn>NyVZ9M*N)OdvT8a;fK}*7JzC4hnsa+&#W=)vK#ZNSV(f1FzHk1g z(ljTjPyKs2Nps@OqAD{C6jBVNyY0hdq<)C0-S*ciPtUyMSK)SbLAR=iQ_Jh0zYbUj z)`T?ikq-bZ!qPo-_FSMH*%45+5ZSY~gT{3VlpxV)}ZxY)og`){`V`P8gfXcK^?CY zy4dHEU$Rnj&$~;B(t%rpZhoa;#sOgW%cif-NBW7~UQk0xSAGp#oD$bWU|h=KV7Unt zQyx;Jd2}h8m4=?Sb8_T^9?f`qilogxq|hlUxEXVxwhd>iUVffWBkRStozOHq7g_(w z5^t=~H057^5%96M$pWc1t*auto!9~$F|$B<6q8GlY$`Guwxgp-3UHA$s~g|C0`=Qi zJ)O$Q@hsw;5@xHmi|T;ep@?%KFjJ(bo0K|rHhqS!;nmBk7oQeCf&ffiSbtazFE;Q3 zr)vUR&SMYMgz++V_v60H3NO^uzlVjq>_kcZ#6U{>7v!YX%W8Oo{0`wkahve^{8Hcy zLBebM&pl%d4;x^N(T&}2u>;0H`d5{HCPVVug1XHleYDL*CyvzCm>HL1ih=C4LvOeETrh{|wqwl|I8Ovp} zWOCyC7AsNSeF6EQ2F_sQ|1%N@8JxOq4 zpQOmlY-CUj?y0GyXsJz6Gwy^h)1P{6g|#$c+u7}IFWc>G zx1CD6b+?z6(oW>IC2pgRQR8hL?w}6 zK1kT;{(tlP4PU;?o0#{>dCz;!bDo3S<+$L!khp+a>B@-MIo-VaDBJ)w3Ddk+`L2&R z$~mNhQp0&?J*&K$TrPlsQ9SRaM}3%%ubqBJmH^k{WsI& zyJDvqZOeCd{rp#Cn*-aHgC;P@qvSah$--LfM^K}5mD9?sR5bY9=T~@C@yC0E9-e4-|$`Wm9*H@mfFkwEaBYu>!#MQ*RiNm5rf< zZ~tt+s}U}x@4kPItbT2hMu4RuWV&QZzJ-E0i`M%l-68$1=%$v)?$breT4@|-9Z)Z^ zN9k&2KlBU(!77&uZ#$Tc;b)yKmWCh2_GjJBL`?au5gos)`zVln?!d-mm5H#!V@m!d zMS7uSoW8l>+4+kEgFa=muX~^4WT+aUIb6jYh>mqLsK9_Hv051ws#8N_f_8~hLv<3& z1wmH1PKqV$g-X1sP?C*wQyPQU~0;H?d_d;;2 zdm)Cg^jXRR(4|7hI98A!Qz1KO$ZknJRCHiV4z9Ukr_>klDL|Yz^oEQL!rCCGd?m zP1W(A@Q{~V40(mK+#68w)IYaURqT;ur^F;q@)*%#2b>s&q^|vRx_S3yHnAB8HZK-> zHKXr?Yt|BJ9VjSbDyiNcd!v`nm^F}i`6#x(;~z7A^2b?51M;yps*2>jHU~`RrA-@GmIa$?{3}=BNDdz}G51Tro+bc;LE*ch<92h6bXXt`92{B3Y_d zS_xs`E@_tP^89i-OQrQjGqcnA5}g;`%ye=#i8`d9CqWvSxZq0FF*-36pTsL#?`#*F zEk)MA=jk$I&0kvqsc1{_!fzSj^Vgh%x5;sK@NwX1?=2JfG*I#j6gfvlpXIhHN=UKS zJ-wH$Ht0R!i9|h5swf=Q>eNW(yzAcUzjg+Autmb6@HBD-8hHRY@4j;_e?ND9c#dD0 zqSy;1q0o<3qvM;QI_Qr^O5F^NTQgLxpz+f08Urevm@)?zSJBLUk-0wA%B;}pfWzFZ z&_1#fc5W-Eu820!a;xR{xn`*@k`l5p7>Bl2amd!tvI2fqq3P1shL(whHqq1piBz25NUl;U$fTxH0(wJa4{eN9lYQwTQ=K_qzr?8^{ zv|QL39|uOrQ4`~{pOP0-qySQvQI)baQ$VMl*UC}XIO`-O$`WOhOID~liPOknU#a@w z#^5;7eMzxjKTv|V`VNMmy!1&~7Xdb~;k4G}xU5_Vm*v><0{TJ$F|>*kBWPvvW&3B^ zG*@9;+hUd3u}zteG?5>m2?*; z3bNjJ!^9xBY<59dd(ctteUA>|AoM&% zueKp7iIXDIy7bOEEHFG-@7@r2jP3{OZa1uE5cQR|)49t1(0}IQ^pn%e_&P{1qwHOG zctg}xvc|Jq*)LcJite>S-Si@EMSwD1l?9Z|ZitEt-sXHr70;_vYyv(; zjG$bY6^dfBC?9h(WH6)y)St_RneZtbHl@A!VDG-aGg__>Z=Xvh2OXFr(O_bWPEzt? z6gfgg7tt$#clSWn4T_zcrXGs8=C_D*Qnk}9L%v7Z5YlWt~u zT%SOb+h(R$_%I?5R9yyVKk@8^6aWTYPO6%D33Has#uB)OC>&eIPnffWm&h2NHr$D_ z-kV&D#aK>Z|@;*)>&B|3J?paUlm z7Msv>9#HZwihO}wzf5QF5kq3h-;R(?7}Mo|pq|1))h(CP?Mw zif(|{o&vf7pRRUWHKzrnf-g@&u7r+W*eL0WKp%gnTZMNO2V09{?gH;Mq$aXLkIipZ zEj&E`lTTMr_}{14zYrwTp#$PVWISWXdTar8GBiP2sXFSu5S#G8zbR1N4&7mL1-G~k zYs3aR3Vvs<9brLqA&a|?TQq+UodVVIrTV|iAteY+vw^d8T(%7skUrO1{#xffK7G(a z^h!{pJcrxDzfam^rIAJ9txOiwn4BUosR!lKPv zmay5zIk1tk(BNdhoTW2xF32) z=0Qp-HPns(e+I!(A^q9Y_B;4?RPZlf^$HhcgT_J{sGLrBOP!nw4pN1R0h($YSFTRgF+Vj$Fl@I9`!Zohewu?F7w`?H~&d znJUl>U*}l_9Yn8-bE4F3fWTw{h=oWo`Q1lOLzThk0P&Lfu?1dU@)o1I%nmQ*zvDeH zZ&%I6QE}k(goWnTah&@eu%8jNGHIZ71`%0kY`b4{28y^l?5`?sKBLxx$ox5bzn=H* z+x_O0JBM{sERg6>IWwC^_MDz-*(~k`dvYlZORHfS$_~Vq=7wstO4-stJ&6O$%nV|i zxXv^Qu}rA|%C>>gi!B9Fa1+x2xSk!1XmP9K+S0aIiv=r}9GgLeok1zfek+=5^ktm< zonpyS2li#sO?;U|O1^<2>wrY}Pbh!cMp|ZS6zjvg87P%Z@xMs+x}0(DqTB4IGi71M ztjC3g0qk^|9WEx%IrHg6BQBOK4>}6bB!`Uy66pRrgj~6flCPo2N-BCWClOM8=v$<7 z``s{eWEaokrSoM4B3?R%J#Juyh>6?Iep6?@6=vglI4)VtQZK;_)m4IP4oIs`ujkwj zt>Yl~zteW*L)K5K6%ZPSpxTR{cU8IDgt1gcJFq`$p)z_U^Ep#1?c^C6>BR`vOH19> zi7s=_xgVo@zz5k4+KBPg(zn*U{rJcG=I6hUDj|u?O6HdEqHsXg%$y+=oNl1^pLz4X ztk*ftUOYYP05g(r^5SIM;>;V7hdwZG0_?D!s)g>v9T4egWiD~BLIc!NRgflb2h)%p zSmE6i9LI~}v`cHDyZn8+#SH{jb3uQqijyqr318uoM~yPtVf9x(`gyndEk<=CtF>mr zPv-cD%)8MzEJ`eNo?8ZIF{E*e-3Ks29?NMP&vLEf^qE1fE-w)fD7 zo|_8-J9$*8_%4IvWYAT5ClQy33Q)hr)nMC{wXiYG1r7ELxYKZNHr!XvT`ssgrY*_n zCt;X*F+XMUwX%D`?uYHL`^sPcMIHT)5lc%q{=*MR#cM-?YclysoTcQSQRD;_UCRf; ztiFj6!WFgO)k%^(?m>Tu+bfgg%?v7&88n_?HPeA81Ij&>igeTQHc*QX?B(2tc$*pt zaVQ?28VdI%_~6w%ZZB6Sn=Q5mC1z*UccehtqC$;FW5*33>8MO5RJ6 zZshHran}a7z+7I4yhN*G1=z!8;led8t$LagvK<@cjq;u1wQ_(!d)OM6_OQ>UFRXV1 zo)Y*+yTjD`Jdr?tlgWaqUn^g@uudA!OY?5w;~JPRi{~9wWqRbgENm9uVwQVmt5RJS zg|-6+Gmmo+SlivcO)gu3_GjpVQ|A6as=hTxJH0S)4@iy2Ks`Q;27tt_OX#nc2Om^D zeFXWF?VJ{Vzw6ovRQ*h$8wj-B$>U_q%LE&p8rtcyOMFPV-6uU1-)Z-p+q3YU9F{<7 z+4LqlF^yj_Vz2R=x3fsG14nw#nb@ZzlpI9E%c^c+{w1mVH^;3YfGVwgM z7NCV{WgQ_^lAUgeOls)WfXy?q=}cPP&F!EKiEV=vS$ZhqXnlBs>byiNSA$fi9LZC7 zzs6^6s5(7#je8wvcyvhnnfBSKE~kZQZkUG{+iu;iMv0vrGP&C_WQJjePq;KO>(AyH zNQd2;S;#43=f2f$cX=Cwa|O5vp%7v#0|o5ow zQVSo6^qF)~7-lFpPHu9+asgP-3@1Kiy_<_~Xy9h)WXpp02s7z4XWjH%$aN!0ztgqY_agL8Po2Eba52}#aH7~3wE!NtD!JUi%IQJQNvgDo}+J@E5y6zvDf#S69YuMYy1?Mq0?L$)Tzzfb*lLhAtB|k-x<5YAS6c!pbMIFptrp7r7 z`u-VU1AYZCIRla;sGy1bDpWyil%QK=ltsisff&It|9aSo?VeJks+0^sET~v^#4|py zkWt?d<94cd)>c(F7xhW-2yUE_CrRYppQdMXX?$uWjo{{B7i8`96)pznXcW%(=m0g; zx>$3}9&)SzH6d6MG2vBDcG3JXL&ZA7^+HK0D~0x%|yktcpNVV{RA<5fToXX0y20Do9PnErOUUZD!@zy7? z6r8J^kAE>fhvE+LaBXo=5B$F&3WsaqUxniw4wlz1;k7cKe)8#WMvEn@F=5T3GTc2w z=9$eQE6YOtrqwmZTpf*#WpQ9Cp9R6iB*_Uuh6)Rw5}8(r{O0&2h_IeX-?SL(Rqn&- z{3&((aa^z1Vu=kN#&U1k_cqM>xzWDtc9}hc+;Cv~GGJm~?o;v(irfLcN9avl%fIiT z;bzRc8+3y%mi38`)9Xc$I3O|8uFR{XcZMGF%TksH-p8-b8ixXiri$+nPx&z8ZIh-1D(&N5|j*s`mi44bv zJ4!t2gI7$#4SU5KHSRkk2V_OVi)yh2(Qr0rtmlRU8#{yajk<@A%(*=dyXCgP);X<6 zbv~%-XPTLQSCnE*WN`ZSee)Q79qJPfy?cy4^WLD-4Kb2THo52|DByUMv&3&aEY2uV zGTg-4^c*XcOjx&W^>@ut!e-fW;Jms8Okp1a+)QPFSh#*yuA_4lSfSs_6#MqO1MiF9 zCR6X{#sH6%S%3QUuqqSeR1k9{WYQW16sL@lcW2Az0;G%jB?(MX2q(v7}ZUz49UcAe&tsM=5>=63J&X|pdp@r9B0L( zyKgket0BE|AqokqnrV+nsdy82>l<2^$DW%YKyfr6pIZrm@^bnV>E>?ZUx;jox=i|9 zOX18C9;W2(xL$~~lk+mRd9mNFfYrP>ZpviupO4HjXO26pS+T$pH_+)4e1IPA-X~Fa zL))|#;Q3>Lb+SN>jO|bZfK*^D8Z4j1gDCFYZl58XLmY8 zdJ(Wl@$Gc9&L2~r_*26-eFa(K!#VEX`*H0db9EUuUXKIswJm7Tpf`=Y1f6V&6wB^f z88yB{woH>S!DrA3ogXZ?yXf9C{qD^5Z#KAQa$^K%J(utfLlMNFQ={aXcrldWp)L)q zr8oi#Um2q_Rr3(JHZ;UzH@-vN?oq9O!#<>5^=&D~xU>50iQeO6F}n&X2lkiKO!i?L zDS13a)>6^e{R|~9jgrKmTZ|g32s2bwAiHhf?bWc~zO|?r_F+4oT>a&u!>_uwOd8q1 z)3>~@m8F7f^@X@O*oMDn>%V+!=GgktYlLeuJ7QuD z_EGX8itMGL>!CGCjaMAc(3Kctw9V3ep0~J*1b1W!s$9|9DF>zjv z)jO*+GC|d$Ok}o9Jupq9=#o}=ptwJpkU^(d0cxZ@Ep@Xo^5t-$HuTRo$RrI$aEyzzEbstRhUyO z*?6%I?2A}XAZde2+IVP>R1g2bPG$UJb1zJsb(nbJF}t7myTAS6o939X@MYHVcY^O< z<&{lu@rMW(Hfb?L#_alI8txmg?#r}umUWm9HiDIYmGZA=PJDlsd4u~GKfePn92VNx z7tO2`Yz6M+1=3)e-ii#+Lj(pR0U z%vIi-a&HQ#>X973Q}X1wTa*tYTKFlnE-Y763n@rLq;28CT+w~!6TDcrd;T~$$a%;i zbT7VD+-NKL;rw@1 zxmt5wXA7N#+QMp_6JdeR1Y%q_gZ(e@2dVr2zrg2Tzh$p+xXhZvQsTbV%xhKW2Y6p#k`+nC3Xa)hb$fT91d*9SeFOh*vFNDyt1LK0LcxXQxGnE&XW{ve-o8idX2D@i%2)Gls()+rO@3GR zQ6Tx;fmybzOcbIXQ}Qn<(o03hk}`ff$$~t7Jc%V~bUQ)X2gnz$b~_46yeUD@_Zb!E zR&a{IP#9W1=g++(Mbcy|qwlDOnqzqCB*gDEKBq{U_Z?{)$)?+9AK|I5l2)i6!VYX1 z^nE#AxF^+LIHZJLpPUlqsyDQ9eC9NK2)wNE>4#lb6{meR>~cD3e5xoMU&3HL;X2ty z(5}j#+eT90!V&$u0FASm)CS{6z=t)8YG1=AgWMGS-i{C)9noA`)izKT+Z!z{HE z6stb|U@SGS4!kM1Q1gl^##rr|=6cWtVzDh3Z`RluX)pDB>c@N(8Fq5nCr0dC`QFmmq{o35whbl=+mea$MH9&Y z6^*pVp78SEI8K#U4e1NHEh|zEcdkC_^O@&o9`LG_-HJ-2a=u&o-mbR~ynXkrhF`SJ zUq^lUt(dR2y_f&>6Yre*;YKR`r`qpyePi+8@0;KG-8Jvs`S#kk8=@X64|8utb^hDV z_ioPzqy*tOC11m5Dt=h`mf^ldKVAM|>ql$#<8J##yC>o5*AhKz&wk0mOamszZoftKG<6;xT zHD)vQ;s75V3Tqw6kY z%m+PDe+V?%k-z-4wT$c>j*I8Okz(lXF+^bX040Z#$wJVmfB!nP=1O3+E_omhav`dU zmj<%n3$AeXz?!gW>K<-2e?ddkXHZzs{od0%59;r1g=J&Y)FVLbUa(5GQ(35l4#qwU zTKEh4!k6+Ew39UPZO_!8OnRGd2Xl)#C$W*!V>LiV@^ehC*#Ol86w4fV z39_JA))d?c_G43Mp-7FefEF_Tt$HnaykL(|;fPacJz_>2#^x9Iet4?K-)K^fy}hB7 zWH_*gTy0`LiYYm0bL^p__d=f~RQOH~Z4S=$JFfU5Ji#ZO*C~$y+Cwe>U_@oWlaM?p zrM@z6yYj)T_1r^#kHyPFljk->E#@TqB!=RB$iBoomxtzsXMv>0laTF7tZK56!FWDz z)&s%BKD_Ya`G~eTvTQnr3E5DQT|3U3XQp0cVmYwUu#iT}rXN6H1}kY%VL}HLvqj-; zVcV7QZ`ikVc9=P20hD2$w%;ksKuP~tSn#S#?`r2k6n24D{*WxduIE*pvt+9hsu~S_ z1c1S<*sCm@s&(m6w0krM76CO+jauG@)0y-}|Ihh%c**``@nFyixfJ8RC4xz} z5<)hGhc$z3Rea@l=50wVboz)9+!YP_ef$XOxc?@`Og3+>`0*J5#z zXBLgg84Qx+*`n}bal)K7)nD4lULbJm_z80c{cQUaefn>;@_RP>RhqnfI+RWSviqqz zEPvv+L4W;|IlbmJu^|>%Lf9Qb-RF8#Utpk)2|B{dm+e(HN_vPPlD}6OD=1NdE1l#P z<9<_SA5@z|aTF*wAHpuL>pt@R)4v*HEdP3F+t$Pf841*&Y;&(cm@CBwG2Am_&Q$yx{Qxk%lP=oM!@YTz-D4k$M(W9 zPDLw&rkV$OEksBQJc_+Q{{;K5>iJKJT&+So?@8EL%LF`ZjkSQ23v`<3>vcI26Pq}*V~8D;B_?5JgV|tBQt^{W!iuKmLv|B zSae_rfTqWg#9}5T-$9WSD*B{qtMh7Rzf_$axR-ksgq=F*rr=DVgO)@8^a2R8-6w6b z(n#D9spCXTc}Y{Q2ZQ9Qv{6!~LgKT1fx#9HLzaxuOtRI_Shjd?{U?_<-G~#_+LaY# zy8~y74w+zLHzfyOHk*o0=a$V*cUvhf0S_X7?q?B|{5`_EywxsQDvhd&x6S>MPn&eW zsYBWep%;x}efW8gVz(aG4rai~&=+D0_oUC}8ON0koGiLJz+MO$12zmWv3j=I-Dnvb z!v$Bq;`*_|Tu$mW*%)m%LDxD;zJ?+zsc0SO&{fQd3Ca~O^}Hr-WvU|Vg^-cw%?kXC zJb<09{vz}K8S`-4iy?gnUTG`@)v^N{-^u}t&;~Jk^fJ=j3OF~RpUnTl>{~JQU?S`^ zzatE5*>t#k{2v461pqIC!hyHJ7HIxAJywx5ytATC{;io^bgrOSe41Y@*1CMbukkv? ziH|IX9-WIgaiDZMelW2P9nV~i5$4mjmn@^>or>k}nJdt-SwtK-KWIUl?)I$x+&IoH zAkJbrZ@MbWd3V^L)5^(x1p7fl?N7)`CXqP-XLCjP`k)HRld=Q9Aa`7Li@QO!chCu| zD-D+*%7X2Mxu}quN%zwk^QwmLS2gPSWC^Kf*c#?w)SqFv2S58_$xJt+xB5{NSwPma z+wwZF!O1uAKGG<;h9b#S^y#2Zc`N<#x~W?{E6INuUO^8=x|Y8hPSq*;!dH=EvEgzS zy$D(o;Z~@19IC`eiIL%^)u!hzFYbvo50||r#>GO^596sv19pkqT-v}>?Vzy>bQ|5E zNS?cNM$@<}rCN;(o6h>n2fO&YpNgj#ZAasqmp6!qMXG_eL5lpHwr8Y;R+vO`i2 za_F~b?~v@4x4>?;FQhrdP{JJ}Smvrx91ZUYPvYag_L{hB7POvGU4xcN$Uw2XLt^#B zo2?QrpSrQ#Vc8EEZi8Di=`SaxMwp!Z*0-;b4Gs*Gy(Tc(LCI4nvW<$an|dL#*vr7V zkDq!|cpxZIpmj-^+Xnzi5w#NxhHQ<}x-7UY-z`oFS^y&IU#au%kX^GCo8Iz-y&fP5 zxfA!DSDj<6QQUGZ@#>l0MSm%~E>8A&Jhx*i_D20oku+uHD{M`mYR z<|ej&{;G4@Y9&WSF+ojlT^FMYVMA0R9Q8|M1htX_L8v+h%r)049^O?1Jhg_XH1EYh z`=PRDm-rN2!7G}tg+}4K#K>;DGVeh|u{ecZJQs^$fvzv>YNO;wq|nIE%DvWGh%sI zO)e_bTdi%y}ML7lXWQ|^Yn8ndC}nMSd&T)9JD1fo@r^J&a`GGmIq^zQXFSBCnl)CKVQaj{bCKZ=b&SIWEoe(Tv@d9rw=9@otgH3Z~7my z(t(|sY!irVrR1<&Zlt2G%&UaDbZqu;5W3K%a}$|serrRq>%%TVyT@_*m^eAK%KqEw zA>S^m*&gzoy$>xLmSMs#6TbS^ADh2w#Kvx4PA92!U~F77!A315KTVO7R5Y-TU~7~E zLIxT175O5;US+CtJE`LAmo!THUDX}TU8YfTWu6)tu2>y-j()0}sm|v%30t7|YpwtV zG_ehyb~?DiS2>wSoH{Izl{L4}UvZnZSh*`RACW&jsrk8xZ=+F|`!8ZhXW z9hl`@AWNY^bgnI=PMXJAGxq{0ie&-c`U@eX9bu3b{?dR^$YqBJ*K$Z$G3M)MNAQ~JiU(iLbU^*pj!u6bD*(Hy3 za>@3R8zWDp^(Y#799v&~^G8DA2gZfyz>oYsB&XQ17aW*FciUvyYNX^16uCe}H%c(F zhk-w6A%VUPtSHlBf?7G%0fq(W6r?b<;ypq&>f7rBg1YH_1g{? z*K#j|xLh}rFFW(5dMAyGmqvjs@6(E#9_kji13`^Uj;bUK`ch;38C#S*qV+8B_=vS~ zT+9rub2JKkrxPTeS?BYmr@B#c7(SI6H+Kew{g-=4)}zbJ!x`(#m%#DglP`*A7}4b< zQtc<1?9k=Fq^iRv=qjP)kRH#cq8DV)-LeHsI881fBcDq;gac09GL6sLQ2cu}VDk)2 zR1G@ii!yxAIH!mPoeWf|^59i(+zL@65i7|*1N>v`Z{8QahPPIpqo|J5D3FFV=(Jk8 z#m$xyF}B9dOW| zRH!Fu1vue8s$7T-ZywQok_~RKy?R6+QflYfJxDlOkXSw2qr7oiprCUV zAs!kOYY(b+hjoktopK1utcR?PbJ9aT!*GY}Ap7}Gm#nnXWLza^j4{@AAZQy1XY@&Q z5-4MCh3&~6a3R`pLSA1)-g}S?a=Mx2=8h@7b%qWaND8` z*AlpoJ-fN3^aaoT(1QoB9Gg?X)B|a=W9miNP~(2+vFBw-4?Vx5%5y)r&-FA&FC5|C zrytC!R%)l$kc2tQW}}#VTa@A5uKBGC?x2$u8S{$8{Y)FbE$V2%&`-C7hj-vu8pU?V zq{Cj!Czp)$OJnoH80&85h3ALVf`#9#cI5o_PR?%YkTI`Hz;Z-T4ODLL$_ zZ&A^kIg12KL5)S9hP>-u;aLX06l`BIJ#g#O&6Lev=2HeU-yDLqBcK9>mNP+gNTJ5` zV=8z(op+~sWyE>)ly1m2)N8jwe0YtUsrtjfroM5wxFc84{&pwoxt2YFm*LTY${bt?Cy*?~Gq zJTG~wHmo{8qqrH;43aYDF#BJH)=;cj4Xxp_SMOb(TK%eX%s&(4L0wU5s6Ke21E}Ft z#tL%8z_f~){ZOWZ&Hi5j*laPMyU)M2U-5!LffiVFIDx^ugXj4W#!DB+% zP-9{|9`y$?d+=ZX;8|gwSa8_9lZ8}4A`=V6)W={_cS-gL8=_80Ka*y27rE>ao&&>T zzATu10XBbW^-SCRokt#!l|7-}j{oCVUv*q8^1A`a?3GM11JU@q?%SsAnhIXd)&THd zfmBm5XMB-iH7>??2R?xnE+%eBS+Hv?p^ZcyFr#=|Zjsv<_Kx)VmGP0Sl#nowNeid#m$?dRvq)GnBdRoub_^6TUlEiz+z3I_2J6F)lQ71u5qQqHE;s)6Cs!>I(g~n^> zMA#7nqnrDuFtEie4?hMwb5dFTyJYj~dkbaji41DuR!RndC41y+pMN{+qs!mE_I}lT zt;^6?2i`tMuaAJ<@N@wO_p)}JfoK2zQBB3Or|oyZYAPJpFc%jt|JZ!Vaae0&VG+>= zUs9%tI+^T14XNUS+5-O=ofx_+VwV^hQG4evC2;9(rh!&spiPzw%L0~#CQcr=mcP&JleIv3_P{tD zSO&`XPVQ`vT*E__ zTq@L4e|nQy9Fz+Oizqnaof4+dg`7$Vja1T01y=&n#c8S|yt9IAmnx2q?g%M{^R@h@ z;38-Tj8`#_P{rBn-{oq!=Qupw5S0~{6Si&25gy)~%U#0JaO38#7ikpf;wsKqnqOH9 zr>8Cp>%XIfpJh1hzWT@8=C#(Z39ZFKw~-oU9Mlbf^}+U5h5qqCt*cV*@PQNiB6}f! zloF&-EaRP3X?dxxcS3a7p;d8C_?b}M&dC*NU9el)qPg3ZB^;Dd&2~wL(rk$HF6W#f z9W-9hDDXMkt#LNS)F=`3;$uB;Rur1AP7dqBSP(Ca6B$UC8b!IXM3h2*2K2`*P%mw} z*?aMOvv9_`X}0jN9Zy>}9m9p}Kk=P^oiVOXo7V?iCYu~Mkyvc95@k_xh{dK-(d`g% zZx6$=`W6y14OQfdAkza0!xYhg?0|D3R3$wYmAM-(Le=rM#0k5Y+k_u{v*6>g?KsG`*YS(Iv@UI7Z4iA~=UF7gZV_kPYoR(Vg>F+71`Rqro;&EY*JY=3 zCr~kUG8~v?h0)89Fcy@VVee&c#p-(KTa))$NtcQP0T%3US{bm;2WtYmW#zzT-3r}N zkBg5&f4Ols^45Q&tiF&j-^|LuJMPc!8z)cwNo8DlPDX#akrX=c%2R8y@>Emu3W}6r z9Y`xkzG$c8?>re@)Ghv9B*mvfi0jSWpa&5*#7HqgRUmu?8oF=vyA3*($nMjqUDhbs zJ)?!Ien^Uhx-cC!I%|PYFOGqC7}CRoPRF5AJ_`!5%V{0I4TUg8ZDDa^u066I-DcL| zWiQ8wZ+7U;USIC*ZUoUkRc*gSHcleYYIz77Hl31#FS(tHPGkx^fbvr$#JKpysYMKY z*?0(5p@y(Vg>a9sUNPu|33;3cm<|-yfchG*dOIgl#xNOU{JKYeoZVzNEHiWW`lz5+ zy{`3$xa%!-`rDaMA_jhjEZ$j9hr^DWNOFtgqynM%6gf8?TTzYMQpp~-Bje3hSJ~mV zPWj`HUv=G%1|XMr@pnTV3Njz{9x2crmj)bFUW1k zjS2DRzw$LF;{ucS<*)uqV%Sw5ImLjXDXe6bmvCV2dI5k=HLM$4q zSJ-&BV#Pc;Y|GURv#!5ut~K{!4+zICZ6X%*=di#{2U0HBxO?HkIN)6C!jKSMz*#kC z;W*Df3*%u8lx%j7W#e&j_R0G%M}^}ixGkaL3}kXNibt~!LRm3N@Ne@iWi*N`I?r=F z7Zbw-T8Rpq&9=1zSw_Vlzqad*S3SI%5qfv#Ug1jU^?p{ANney?(e=PKTPeLm>O;H5 zm${t~P%UI`MJ-5j>3y@6spNIhJ-#RX8lt|G^t*2qZSl1)P^<*SaNAaE=XX^1Sq8;0 zad>J?vCmm^s_To3iUXSu3$*C7Q)9e3LB+N{O8bo_pqE^Tx<-!Cd44)NRdm44yXqIG z&1$T?_^4gaSjNf+3pc%G-VD`YlXe!`n_g!Id~?0N6eoJOMH!BYp_nh%Yc=Of@#DaT zsOwB2r-0jPPokpv>{$&2z1c8-+J3Js17SF$acspee&k`?oYenp-$Jt0fkUYKO|~03 zlst7W2d*811c6{KEqmpL~rk=rX9fNIlnUMi=Df6l!j%5d(|lof&-;?z)3i=!8$ zQ43(TN&J^>?>4`)=B@gu3XcwY#17bFn=f`)d2yW?+q+wSnjKbZq!GUyi`uONFFh92 zZcoa}>0XxO|He_{JFP7N972O6@D)Q^2IZjL1y{N z#=g_8ue$uCjyXmSsXBlIkVR{`>UL=kP=-*GvYW}o6DZh~O($^TM1|nHtd$!$M5oA6 zkw$gMZ>RDRaFMc9w`Xb_2wbj>Vi z;GQiw?1{>v`e0fuh-H_9GSdUunpvx6K(?t?S~-sCY}Vkg{F}Gw6?WiAqDpU0Gx|E; z+4b{Zk!_R6K@*oEkCH=o(<~H~#eQMQKw0h!>2g)a`s=1+gGrFU!NLvfr_rXus3tU{ z)N2y^WI>q?wxUF)A!-QhpmKkqQIvS?fo=>5KBKbQFa4%i&x&oV_RBvP8#1Zm208bp zA6O(d!bSGCuhx@w4(uB5Ho?S!YNgDho^>?MmETxaB=>)&aQE4teRxIOykIdva%HuYW|$_V{@93!$QIW z{|Q%b~eD{#)lLpegpt0+=VMWZuIH1&q}fYU|k8D%0j z#b1rf5)J|zX&45XOh-jKfz^};bmdbQp zu|InA+PrM@^@ojP^6F7e4yme@7>ZHn2vS3lc!@O3|G!#2ht3ZArIdCZ2ka@;E2Y z^@eaIubDCQTU{cpgVrxybhYmjzYhAW=QRl|dHhQpeKc>W`x7uqm6B}E)__`moap)E z_?Sjf?0cWDz8P{1T3>0sPmp}KV%bjr7JifPwk(s553GZl!RL=?6ia6;@f&o)^>gWr z6)JVMB#v`tsskhc-r88x<2Xxw5VSZaEepo zRZ4Y=Q{{N(B*>Xo1SCZ*57o|Vfa|-(n|wMv7efs5nirnShU7s*lwpKNAuAcYw^)ct zs2slp=)%zL| zJF275L8~tWirJWc<|<@4E=y9qYg};@^x}4mD0%t6g3PDZ?FBqOKOsIPKBP7pu~QM= zJksUBMl62fP!4YFMEM{^9#hfX+%7uRxsYQ}d};$-+FJS`?A(saGJH``=+0a1^Y3_M z(-*yN0`0$2b=18BmOMin5bS=|MlYG$Mq?ri*UZjuYz1Z7N`5&w)Fq^nkJtN1uBe-v zBW{6Y)Sy#2zeCtZ^xgdcuWEeTMS`sa)On|5LT+B8z&hGoQ8Lrzn-EzZ+$i56>Evm^ zQGcn6KK%u0SY_ZR=GuzCuwVg9l+c#}Tf@rSVl7aiE(S!j=u7^>sPQ0GNC z(9$uEbCH~*uZchBV`w_T`xu@1)+hh{n-P7Y=RaVZ9I=9tP2SPELGJygWW)ELTBrHJ z|3J=7BKJ)^nHEZZog!DMXdPss%VuLW+CERcD0QRcVwf7Y+=diZB7-W`XH~V(c{h>i zX4cAg2?{+xu2`c;5S2y_IDIaSnLFT=AnJxDhpi0Ox?#u{tiA7bAFT`aerfi-;Hz&@ zRV%sWf*m+lyB$?!yWnpYbC*y7m&ECk;?MY`YO8O)Y&Y5?F15 z{$ffF@yk7k{zOKv`qJdFSP~ad?lb6wg57EaT#i`1OPL3y3J2eWdd{mHJeML;BkB{G ze(qJy4he`c!}=szI|bEXpa4o};C1Mwf2c7W05ivRvPRy^_8qenAkM=byJWu4L z&OKvZ2WO#8bPBYCHn4u~bql*G za9ERaeReZH$Y@g9Ke$^+a$YeNvtuS^rj(NJqeu}I9V5_?QYwKerEbf6Bn9$DP@6D6 zQyEECdUf~xD3 z9Jr{Lu~T}TJGM2!4O6u!SJcYr9XP|wBuPJHJ5pTIrY9>3gVdEE6`P@I1#hZNb?=Qe zatK<*^XlC%3oj|NoCp1o8LdtYO#}gP9LqS8eiz0(b{}+V07on}d}D-qsXjJ^ok>t7 zxF{TBRtY}2BAukpy@F&pL;MSgPlouGzQYPP7n&3x-wF3)tKQJgd+Ca3Q|!GAnb_jFel*tYgge6e&^%RMtqEqy`Rcc&wm(JK40D;KTh$G@s zI!}TvDQszPM*imYSFMMI6@%IE7V=KrrxGJ9gq_0Qka!37gZG$BbSfp^PLU)kx>(%C z&!n44ZE!q<6SFzJv(nx0*W1b@a?6E9!Y%%TA#uC`r{vHJk$ALu#_{n{!?S+UpZ}0; zIJXvuvb+urS3>iZRh@pWMu?QY`~Eqynw{lv;I#?JutRikk}3HX3g#^OtZJF7UfgCe z*+Iua-9xJL0_;|kKNpvRJFW|0i?twsE*xa27JTLde{I<4VDb38h$)mojvCE z=W=sK`ioVW4vYZ{6?xb`z|gFGzBk@vyz*gTg zjKdNFHhkxihq7!=TIHSz-!__#+q0{_ASWDnb!js(2A3)MMT(rKq6~o6ad^EGohA&Ow4%q(*g6G(u<4HtCu-CG~bXl(kO0t9GhQGXA08Y z;zV8Iwzog|OF5n5lkUdm_dn*jvKlO4Fvk35EC#R)7U4V88S`G{Y}geJy!*A#*}PFw z5YQxCSjB1KpMV^4H&7dDRM2BUkSoaX%?eGS_xf1Hu3_=@k9aE^%#0^!u3d~S_?mGU znP19pBUO_KBo~Kl3+gEO8H${uqW5!)g^A3ne^IB0B1s85;to1hi!O;V?zB&~L!LCZ z$A6_%-^i}jBi6s4yIPgt-7~v^qeecIMuD4B$SD%&RV7*~f@XJ#Os#Xe7;`hZ0!%il zxA|_7CxSjqc3?K0#>Kld3aDB_5=%LKmcN(N4aDK4WV=!?ZMbL3BIiDGPKgXERtA98 zWQ=CjAM(tIs4%-|`3F3V2jIASichciKKjq6Zg1qe8RWJDyC%y`_%&Zr@?MH`Q_)xC zi>54{bZY8aP7MKORYa9nlS`TZdfpoOBhIo(4NpRVSW`B!*%x;7Ie zxYTliWyLIai4!fIR0%ax<-&D7+ufH<>gE=E85)+fa=N*?utlK_QHwdvWDzIX2j2(Z z;f8R1aHlKQi$4k3u563K|D!m;@<~0ieUW(AV$Qxu{5&fr4LB{IWH_oJ%k`gh#iR`Q znQT|aLAJp#61KiT6PPq*|IB#zMZg5=1uD*d?m1zt6vrO=s~OK9#gN~}@P|Ww7_5F2 z)Q!QrX|s&j`&b)QMe-(*lP318jFOj9WFHls2NbFYvTo2S%b&Z|8B3wC-BqrjU*88D z`^dy`4uf1f*8CTRAM`I^nt_v~H(Q1l6NkAdhPNrSM5)#{%3_LyO~E)KK9Qj>AWi_8 z@^WE3Su#3hhaq2s@zD0O^x}Jr<_xP@a$I*pRGs!CKO=m8n}1{-Nq1n=1r;_!*6AWj z4qe@LQ_-zusWL2uj3Co?R2}!K?P7G){I|g;!z=FIH=k996N)T^KjY-g85aNs@rArmXKo08{JB%6vxa(k<+)b*xEJW#H>qzC6{$fYS4r^G?M-X%_4 z07m#AW$n{Nm(9L1FP~_6M*~{;=q_b|7~pp0fUH2)B~6~9^#+!(-Gtvpg|4A4i;YqB zbhx+6-Gvj4cv-SM=qRwW9hUe5trLc>L+dE{8j7q$eqPMnRpcbC*2#_l>!nfB?>f}q zeELtd`(JD3*zO-}vws z8$46NyH9tHH<+w}k-;(yGuL)JmGa&AL7~x9tW3V~OS1O0Sz30P7?$mnJc%NkspwAn zCesk5dGoG2&>n(8=usPdpS~Q8pvbpPRwCNzR?XWPdTWl><@PKn-i4zS|0@A3 zV~9h(;o%P1qv0tkhWR_(8;tTCt3`+Ux%k6xer&WYsSl4$BKI74JG{muk~cufA5o-_ zimq2+DVZT+1Nw_gq?vSS5P+-bM!`kHI~6)v&c2> zAWRHu6dDEUN}wW&P7*6v!z)&Fg7itOyoJ9$yqv~jO4Ovl!pKTck*jsavshk=( zf<|Td2q=!RM|jUO1vtY;IUOM>qSR1zmpCyLAJFgCQb7s>gH~0XMy87QnpDa4Gk#~j zX9VES3x3g0>Kqt=UzjXc&6NBaMH;E-m2X`a$3my;hN#DLJErDw|CmE{(YvTWV(_>Y z`bFmYt#of?sv@$uw@H1}-Jn`oI`<)2Q1wpJ*HeC+{>xQA`Xv5apDbw5Un%r2@mtXN z{+0O)E<_bUL%5}rVg;ML%lJ$DIzpCB>Zf}=pYZCVHjs0XJN2>6lM#Kc7ov*22H@Q> zGFc=$q*@K#v+JWSe`~lA@E#ri2``JABLlY5l@X^0NEM2pM7-Z z`m5gcc%NszqBgi)dJv0!15$Zsxrw}9VV2$Y%_DvJi4-Bf`pxgufB3gw|LphwAzntw z7gHq0Y%WI_z)qJ~fntLEE%jgJM#JO!sP_ul%+9NJ-0U1FF)>rwlpOY7X;d__{jgD7 ztl%?FbI5w4kwCJe#qB^)o2)bvyB%UONnNi%CBSCYE$#&x*`TqIE@^SARbKEsVrRkH zF&P^>VG4(x9bfa2ixDY*QD6K$S>eD-6e#KsNf~URr28gLR#z+QAiJ5$YvH$rLFEBl$PxFuV!mkH;K2?Y*j9*jv{sB2drA?3kW{ZiPV!FYW(UYM?%(O@^vKC4LvDR(R_oPWV=NfqpTq^*O0p z#5t+5tCe^Wd)DLo#b@n#=5gdvFQ>>MD*7_P9Pz%0 zbWWDwC?{T^^tsv$=?%Y;WriC-q%(*|s><&%bv1??cn>7e8M6_acMAnQ>vbRBug zu?^tD%IZ;38}Ft>7%{XtTb4@pO(OLsQ&mIB4^yO)imsF$^2-bF=hpFyyi$OIHK8(I}$LktvKjF}xnee!3GBF(iMGtsdBeR>^J*orrk>_AYSb9?%1E2EwMgcrwC*Sc1LEKi4U(CO&=4fBh=x|us3 z-LewW3Qo-}@mlDjt#_Z+$Sb_F>60pTS->(-P<`T=>7$MnWYa~Y4YJ#y&kFl{6tEi^ z1Y?*)^GRj(_HHnVv;{4vZPc>ky-fQ|`();9*1G6?EUNj_tw3?EypvY1xdbOM3 zzr@M*7xl1&iI>CF(v$W*&kCj!<&gp3HRtfYSdHz#PKgCxFWNdiqopF?fEYy_HDlRz zjef>nj*{^U{SWmF}5h2n1vYK6c$R<;pzdTNxh64#cKXBQIh@ziT$f*5KtO zs~VfPv-)1kc)796zwO6HZ^b{Y8T9=fIDXe_vYxb4aveolfPTl@FG12k2Hnp+2h;&| z6Z9mgS7g&!AR7;XoOEs?GpP7n*-oEDhPS9hcAo}mUhVV-)xd1r?F>4#(AjjN3nB?x zJ9S0WEAY{UbITI=&y&1`2DQV~P;4foQ9KEG5R%Bi8!-*=3WtFhQzElsu1*5mbEtH;w7s?gtX_gT zD=rTMHp{RYu9&>`kN^Im%m}rVY5)0KlIXxtD>8vvCM5?6#1!z@RX|{i=Cjua&^GoZ#^jWv&TzI%iyUZK~r$H?_g`i|>`$aBJUZZ?>!B|+mP z6EX^T*a`t7j>8Tfo9D!Qbu7gwj=Q%b7K%~Mefui8%UkWzK`*G2Y7~VL1reQemd^sM z%Yw4VeJ%sC4iFf~_G)ok0Q>bWkQ5k~ock!y1gp%5K(NWh{|Zd}+}F6w^nNH^Ms_&x zGE-@?%oI}ce2U}&#R59#E<~9KtR^W6#~qMf6v7Y#sssU!dTDjQm3iBtu~IK5!KXR^ z+vX&IgQC`j6?t4F%R-AI&U;`6E0Jl`%iyaq9MlzIbI*5lFc{9?u*vK{8-$fnndntq z_bsjwG*13bv1BQ`%#i~#rP56xmPpArP-GnyUFb0&+Zu3$Ti~%~ZkOv;#i599k|x4_ zV5tGuME2PpJ)4{t4y;De3s2hjM9|^$B6DBl#qG2M&%Oo!Bwba@?<3bD(n;~vVYnYUtjdmTh$`^krVsy*8y2hvfzf){cj}*yZrPN~Ui2Hy4~%BxA5T+X zC+8h_t?4u|1UD)94T>}YrDEPK*L|Kk$^OZuOr3K>)Cs>6ej3FZcdhqPPPKAvNR8l@ zYeSUbZ>v{KP#W-BG^!P(jekS@h?@nq(=GgS{u%xmVr_sycL8gt+h>=}Ziu?*wM)?I zb$ixg=Uow*p_i1cvNTs@!0h2H2iP@408i=)Ep&&`Pk{w@IjV ziQ~m_utE}(T#2)C1)C!@3ao9!fDCquNMzzUwfZNJsgoVp7E&kOGi8aNA=%X@sqn`0 zD_k<6Tx5{|JKMoH39$e#v9}oJ3rWgH`hx{YPWkSc{c@35201_r#G2{^V>mlJkdLcd)RXYDS$Rg zPurI6jYwy1)B1KeyTkg3x|?ZL#EDSBOana&3OVXJDJn8{OV>{6pfT!G8QY&4#(}lZ6>oxl`f|Zt=qIjT=CenrFMSb~igeSVnd;QNZ@^ZqoGyb- zsYSvCTLY36SCtMG46b`JD;wQZ8xSdTquAE+PUaIm+IokV7*1cy1DMe8a-)vDVa|L}WI{uyO z;%rVAvu8>-@NXKTYP_`5A-xdX2AWh2QO(n@iLrIyWu_FI@1EJI&@?HY*CkHyVY@o9 zG7i>RqT!t&R&T`eX3I<2Cg*pznrpwZktuRu8)ZQ$_OCZV=}g~C9*(-jxgxBbRA+g2 zc(05|lH_ssa$A|0prSd@u8V`p#`?=Ss-!6=XBE>oWk}vZO~!qmE9Ro{$a1dmR;Tj* zAA4^C*HoINjn@@#NZuH-5lpV20uc(tlEu&>cC`E7+w}DG^fEm?OMhL{-O@8-_3~Bs zq^hTSL)=i2MZpC$fC>mAh@iN!St9O&qN1qC64A1-6ci~`{GUUDl9;Qxknmk==2t(< zo4dSV-shh8yyraUIp{E*o|r5e>JVQT2`#>G`&S#eOJF!TwK}jso+>O=zvv-f7?~M} zJy@T)trskkntQUY311zBO~P_PWZ#e&irtk9NSCZ_u-J2OBy`zQg{LQ0O2BSq2BHe2 z`Db*|m4l~OGx!@lbIZ{k2sHb9yp@)y{PDjkY+E9*bRKAvK(Z*ZlQz+q37vykZW$;w;a)@-+J^Zf6(7-?uhoqA>ld7N2S9+VB0CRPlIwcFk&;I!iI9C~};N zsZv&py6KI<^^$dyj|eZQE&&_3%sX9DDZ2pf1d3h3cAP${yDq;4jo4GE5B}UHIXLg3 zI#u{UcymSj4*3{a(wZoZl9Qpl*LL<8+qmYrzTX)U5xLR%6>sfwvh@*6s)e zQZttoT6xt=%E9AF5H1bO4X^PnkfKnd(=!pTo)a#}c=brmy~+t0qR0(50zIb{8g3i4CK`!6oL_$@OpQZ0al+5xIs`jT7;*Nbw=L_Bn_O2; za=tKob}0HAMj+lpF}o>Jj4Yf=8cAHGGaiv*VY#$GUKzBK0hbtKhPPA?z~9{}ET5Ae zm`Oj1O7Ln>C3!Y@UWT;zGq<|n4((RYB3YyMoXlMPhYStFl7K7Z3ao4(bB1?06?hII z%E8r#D{5YK|Jj4n&%Nm-EotGMP?dRHS{{UZpbp=ozDDh}(0XB~x&s=#-B1+LZIGt8 zJ+o7?Mb`k@bxrzHvhDPMDt%gIXo61}R9ZN7s$crLM=|#=jpfV}?AFAN>rK&Z(qfgx z!kqZq?_4KqxLFt;uSd|&V3>v3LNTyFZ2}#U>9@RFr{D3ywa9>WWV3gw{~+COFnqk% z2eT4+baBXW(yai>=qHma#0FHa$IJgSA$2c(kB@Bxk0mIEB8A!H@t2I+0@A6vB9Q|+6tC@|A$R7hI(#f{uPtxYME2#5JO4kc%CGMW9<33i ztI}0h1vLVfL3ucQo*#3-_w#4Q;$YsBUrx8>cJg-P%fi;|^T^kxX=KboE?pfhxYPgnN0ojSZ*|H0lM_fLkK?fwR{m!(#S~H`A9uKq=@v-`kfpFL zJWW%ptXFl@>zP_Ls5J=dL5pg@?a}0V)pf-l(x_^a)c8F0sixDJdKEOi0cvn@biJz3 zrwsJ&+P!y3m+4z1u0g|yG4R~TxP}qV$*fSVEAH3XE;}sN{ejXPB401ZDas1DN*Y7! zU>|@Rg*@GDpWKjL!X|yT?$#{a61e6|f7!lro`#o)aLE~N)4*GQ_3&o_`EPpjHPr6M z-WFKVc!qo;I}GiykQsvK+ay)qS)%)zyy@3M)0MYW9hyzyW95=^hVEF@kig5p4Z1$l z%G2KTj17h^kcgQfgPM`!zIRxZBS0g7`JEHU)~F{ zcq7v@|C&N_xy8q5fcR zRJJ^Uz(LA8yF=DZ?v%$$pNjH*)(Hp5ZearyMI&{*Owl;INna<2d^Gu7anyIG*Hcmb zob}93ulV3hU0zVNb6XB@fY}SDh2@A1G9x)Zr)Np6tsBdQ=EUQmD~r?HtVE6sq>EAZ z8;RSv!7-d9J0nok&JYg`+!sl6=w&3A24-b?;C;)-0y(nF;nww0T~PCn^zy;JHEEg> z?`vY`n=PjyW+bq9#WP!%{&>K)h>f?=6PB{H3h_2NNwbHn4K>#U+!q&83s1njuAOia z_C{k=DLwQhbC~I&N5bTKa&=m#1g``k2LAKy}KZ)Rqs)*JOdo!R^e^n8pc91 zvybO34k-Z|A=JOSp95h^NWun{L>EVwE1xKw@V&EtL!+4y_EATFh0~1ici&KW=vTyc zugt}<<#7<7g$s$oFOcyw*DRvCCd3Nilp!O$H5ApGmqh3IyA;tj*y99nn`d45@eeb< zc35E9EC2R*?=iBN$C3MNt2JmX#UxW?H5G$fjsdrN_pOShZ|?~ja4YiI6R6BqZU){3oWQ9^*y5|ojgXxQmZP%M-=mbBKL=M!_qV-btd&n zVxk`^2GosOEc`Q6D^aJtSy?r~=xso!^^^>|vmVlQlLnP#9~&D|32enox(CRyCfZQq zT_8mtq)pNS@3_5e!DRg~2v_KWxIGH-cy}urLQV;oEeFgLqrN!wq?J-}l{Q8-n4jPs|cjsZ*Q) zk)RI2HF1abnyguwC~r_*B)h~0l(NIwG$Riw*U#Kq$sJFuTFM`!V8=uTw|>_?evz$j^D;l5zrGn^@nCk%YfBX3%xz%1IT>S!tfQMo=s=(OqF6Ax;{O5lk!_ z!?`SNl}?9zJTFOzWvj?}>(H!=tcZX@dZTVNgUe1oy=`s{6ntPk{C;h-GFEDM3Tmwl zs!V$MEY$Kf>I%G|MKr;#TU|)Dg+OqPZ*4F>0P7}xXjT?T%~4(hu5sIRHG=`W< zIiTyg>2c^_lo%_&xu|l2#S^;qAJ_g&R&dKD@fbsSR^IeRih*@@?U3Ju(`+zL``}@k zutIBIm#|n6dzFk`T$n31MmgmUKkg?T?@w8<@znF_88X1*#sHhFW;>B$R#7CLiZN%P z(G@WORT3TI(Z0zJe|^JmbYzN$-#=Ez?2C(?qK$C_o2!!F^v=LJx-#D;?O{)lQr_#6 ziBw04p?G401nWn)_*V!^BRe#Pixd0CuuYL8;ILW?#t0pp!130K2@~G^_RyYs;*)nt z-FSN#;w>YKfb#pWx}8fD0}TTks2J>|gMtg_Z(xbvV9Ab&?xYjtnES!YX_`;f_a=9R z8MTi{scev=jU9KOdQ4kG_rW$h+3RRTmEZP=hR}+s6=EddK67iTLb+gk za0=6*{v4z+P^%18T?*vb78j&6r*+ZQbiKr=U92!gfZRo9AmkHv%h!dM>dd4X^nDyQ zYn;($1wHN`90e;o+5|ZlZyPMgIvM3BBHi446dv!5lC5GpapRb06nR3$?2_fkcS9_u zP}HNTrpqVh&FP}^q1&`k_!)iPw?PQ{Ok@pNp;_nmsb)Y^A-c@;DS_8J{V@n0w9z|4 zpQ@I5E)p9(`#qcVy)&Pxy66tlO3wOZ2CmS2s#)gwWO|eSQCJ!f9Fo1Nqb^SV6foF8 z)`Mbap1c^O7FsoZ1VTeX*sSe}Y7X5Uf+CHJ$A1=man6$Q zS>ZjxoB9Rmz+_DlUJw=dU-vDA!Jc@1s!0-71wRTZ*T+j27`02noAie@kHYY`$j}^! zz@MEiI!OxrSL-Uo^@<`HZ^Jx)`d)k47qLwK!L%AdoOJEvvRQ^IS&{74VENE$#$j{f z=1&_|5A^PuOP5>+6t5geG(?G4dAD3SLqg5_+mV6|743L)kK>FD>y zubME&K*6#qy=vf5USkJ^J*P+RNBTxnoyg<-Bufv_F4Zn*PLURJQvXSZYUdzV+sX|yQ zC>ndU6-Vcp(rtsgKbO&}|LQYO6uVIWe?Kc{i|*UeHE5d>0g{&v_jVcBwgW6jt`uAt#`R zb%@@NmyWm9>3O4aLo8}M)l9v*Pu!yE_t@;c)az*&!T^C_ghII9ncdU2)Azh{bwI8#?A#*k^eF6Nzi9O>7C`{+eqw`L_uX}y*nP|=t_NCBqrEE9|I7@k9RgyMuQG3 z#IVlu$S$xOorTH=PsA3-^&`W@-;?++%#J$O%FW$CF=-T8gFEWZko{h}V26W}@o^-Z z&J*mBHR+T5SCbY+JG9VS7w&9aymHpKfaaA0Ir(DnkG@msX@SPiuaN~L#aSgNcA*VKrSiuiqm!8vCVTv&BJ)Y;3icH+{E!m1oh!Mea|$ zc1GOP92PBfc^-%h5?pC6Y=OjXKHmLl#ADmy=9iZp9%F>168naCET|dppK}$|gfr=Q zSXk-;8x_yo8dRO2g%D3Nv^#>XDPqNG6QCPsn#SeGFwe*c@L@ZZF8R!kk9_~xcWo&v zyd^oY&{Htn-=(>%tOlC_sbXw(mMKStec%!3+P(AWoUz^Aa6y4frkiC@b`;nNguGSk z7kA6HKL~G;0Sm{ll_Z&BR#Rjpv^fhrg=#!=F&J(*LbAg5M_u^JnEDQusc;AqEYs%7 z)7&7D)VlKhpIRJ_j_>{5-;ri+z9^5Ena`~NbcbSYQIKVfL2_)X*xXPErJL&A(Zpc) z=(GtrIujV2qKMu|EG3JRVvf+<*JDRep%Rmg*njb;Z=xte+bD>gl@hrp2-;7>AEUMu zdQ@jZ^C5h)S(zqk)3!pXVn^XPQt#!2SuD4&l4K$W4X;0-+u zD-oO|*?tvMQ$Xr}+uU+pB`~h=sn6BeKBi(WM%EMxh_J94`=g2C18zBDv z{-E_VeA&vNm<<#_CZ<@JB^cyvL!j2^zk7O{XQE)yq-Em=+?I~-nS0i!DtMn_*?43Z zj17ztU>u>$9g`z`f*s073xD~yJ{ENRFp8{b6%++1M~DC4e}-=atqTqKX?8P@yGcxI=UU`DvnIL6Om+E7uJW#F+xTWYE z?5TnGrOsY9t3y*0l`FhW8ueMCL$lH}c-|z*rSTe`Z_=*>@N5g{(A*dGkrMSIQb(I4 zsk4WkyP)6hSs1xBJP8B@^96S&WkxR_k0Y;~ami=f-1zZpyzYdhh2uzt;V0DV-JDAyE&jwbuAHN_l^YD_Nf77-6v2w8w1iq{3W5Dz+^L*fe zF90SJoAiAk3Mq+d@!W zo^nJq$nveDv5FBxgof}uDpNx zqx_o|&>a26)lbP0d#eq9-FMPv1&v0EX`sj%D&{lQQE9TYBD5gj#Zj8(JiShs4-WfM z=H`syd%kW_VPt&pVrfozmZC|Y;UCZJQRIX_73D_cM;r=Vq-)Y2mu^sY(Zi^;7{L_CbEwGZ^m;Ht$e_?DPvYE$p|ASVO zzLR1KD3V9T4A#9*@6ha3n34hQLRZ#Y<|4>sBKe_`#^V&x5mCCN z3A(K2G7nYd!o`C9QShwtO5p9Fv%(|~=dL&}=4vDe7;X?;(*fXB=EGhDspiwN7TZ`c zOOW-_-?u2@NW13HoW0^MO-fXgJ_{PXl+Ne?70D|hmx5BEZ9}~(Pu!$W^-GLyjxdrP zQHk<3UK_}wiR&PccT}3ibV6l@t0n}dW^{A%Vs4#%#BS2<%?WSIPd)!}+XuE(-Y*E} zn}y`N^SkvBom2MGTkJF6~eeenI@?&t=vF3;R{d-r)f z6Sq#^rgMEa&EX*6h^66=y7V}>F|}ic#g(1knx8-l#*<@K9^(OuDW^yo6@zSnB3Tig z>;D!F0ly}mE z8Pc_(*fY1AZUed=3(&0hH1|1Y{c%FNc2Bt`Z~;O;D?DGiAk{z57kfzGP~4N1 zc%p;z7-BlNLb{_CIM-u$BJ!PvdK_p*dFd0|0_Zm?_|H=2oGviY8PUn2W7-;}A!lmN z)XLyCt)WcNE=UpPO)1xGQaN{iUo$COrvEiVIrTET;S*&>e2uo)lHMQqpCobIYzdFu z6i6ZtqmQqlm?VlMP%%(U5G&mW_5#Uxh6LdNY+W!E=Q4H_`%F3t7xoccbLHQUMzp$H zKvDOl>c5aBJPx#Nv6|*oib^3vz337Cz$EXqB zbJVcg1v{~uaZLX=&X1<)43GT|mR7ynJu!20VOZW~JFs>ylIrg!3K|EqHW&0SGB$rr z6#QW*!{hR0$8ktm`_#|p-bDwhV+(tw7TfWD>4ZO!1a1ook7+6p!5_AuY@(Qri0BwJ zE@*AiG$_YYNMVa`6Oe9i3S*j87XVA2dk%JmR6s=*ryepeKr_qI8cA&4XD7=m(KTZFA@o#aU9T#Gq=L zrcHwSx5ot+BsEO3SG}?{vJ9$<5MXVRXF=zqTxC~AfR{0`EA6+o?6yq{^R{4-CDGfZ zTA_L{t7aO|@%ue)&FT@NT6V=$L#1$wg=n&0ZkpPz28pIvB=6_ZdD9Q*c1E~ruAje= zBc1f;1GwmblUwrE^gY*(jAq5*@fL<<#WAllW;tc^|Ic}3gq)w2?~MY>mBJo~lQcm( z8Fx4?2hk7)xPocuIamLk@b#b6+6qVV)`DbOxfm%FlSq+OR7@s)O4}w$^#|QqXm@u! z;IhoH%JU4U&yLyb4$YFIvpjVFSpCh(WQGrTu>eT zG~jf|S@}93J(SYxJeJUp=!!`u_(srtc?yU&L`!cjccjZH^ z<34R;s1pF{DCDgDk=K11Juo#NNf!OW0-~MbS#OgYJdTnKSi#~R#dJ~R6QI~b)MfB4M+r0kvU; z=J2%2&;%doGgTXTP=}AgV>_fE2;{hr*%$2C*ZRVm zkNyjDYy?g(CV$_n>@yE{+e(FTu^2oK$FYcr#Y%I7I_cF6Dp4ZawqBN{GAD>X{P4qL zuv!c*BwD=fTBwn!qT#gb(asg-7Qj1()luB__h~c!bu{@u9^->0|A%CI*h!`Oqi0d- zYd~ENqqYZ%23?N=n~8E16gH2!{NttXE}?xbp6yfSF%}rH;bl#iID~kt;C5z9ha^(Dx17Rfe}Q7M6A6I4SY& ziQGS@LxX}Ug;au9^>6RaChBJ9)P zQybg}FiDBb4=`XkdV5$Ce1xvt$JCV&cmo-9_}g$a2yc&n+o(h#OIL<+9f!)x>j<22 z2~2QMPEZ*a{H0U$I15ykEDxzA&)h&HVtBAKiDD8cvf_ngmQmMD?+&~VZMMe1!?*J_ zIr7EZdCcXfIDPT_b@RTz=bK-(3urFaL4PYyGdEBKJ$(%4Tc$jj-YpJ4Tk0S;u60Y_vr@a zWT3@t)O}^mezMKpUWUgBox@f(u8d+pEvJZzY57iz;?M1x!a-@VEap?qf;LGTG$o9a zURNCTs8-hqw$p3G8)X}1nevm;%AkH}lfFdQGcCh^rvfTgr&aszQyinud4XV^6zadA zF~?FUOWsBo$qq?!q)v8bMhvBw@a0_F$ZiPt{=Q=7o8A}-NhYcO=C0bQ{umF-*X8S= zP5T5VPp<=$;vfPX7|FS}up_|UKIY>~M`s3GjLGhB#} z<^Jee<+-RDrg!EF4UUp1*rDjqENGBio&H3A$P4<5t_*lM?Q@?aB4^^uLFBw-DeRtb zB6v6j^Y?5m|skX zV~}`p|5zPyg2Xtf?8hFq40_%|f`t>mb;>8RP_=VURD~#yj+4T67ur1Ge|gXfxs~J# zdIjds5L;!E-J50RPxnZtdjFi#$U9z0gv6*B{^u&A;uv1+`WfbvE&>oYylgJ7l0^@N zSQoF_LH3O&7p+{wV-y3v;$bSLRCwLjkUGB-HiHH74o&L({D7O$rNU127UhR|^u}pv z8Wf`kF+86uLZF1DX*vXlgL}a6r)e$;GUZ1jJ{ESVpieb&kL%T?GHmaL)4UmC7`b#6 z^z%9*st|pMxz-OChji1cqw47EpvGv9o*GU;2@R4`4(g2Bkp=8VJ!7w!%%MvonN&_Q z#@~Qj{Pk}SNGwpAe`)>4WTm}W3V+EkkOLkbFHWbJ^%O~khCuR4S%GwW_(9N8Kc=V^ z8LFU9WiB+FYf;=)rTQm}45`q9r^BT|dK*}AMY7G~F8j!i7hdzxp7xq(Sznspx%56s<#Bjzx7Av* zjbgGXVx(fQ*bg+YLZDQ~s7sWal%@#uVdKSy#UWYpcJE`PJbH~U-eJ@pc<1Kya%IKT zc(K`$b-AG!F-%;H6HYK0*Bg~8oMVBB$HQZ3q@2evxnKp8I*K_$ks64?S_lKZq#J06 z2!`zuheg%uC5)k+*~y^SoDx}2Cqr+ReO@aWLvFZn@Cn=-J(yJ>HA$-cu7&Oqlt5&* z-s9$M1J+wq26dC-kOHYudtTNMnmm2C#(=a0eCC+qmY^9HD5DmiyQgdj&GtAjZRG3H z$dUBi^t|f8o{e+p95*EWzWXPBw&4|)C|oviro&_B>I zQ}5VNPQ8oUH_uxg*d?dW@BOvKxO`=2?PT%^kBv+GxZ(Smeu{ZSkq1=FPSQc{2AXJ7 zAgFpmr?@srHxz8eN_!(aNE65%RR=!~!`DBDHslLh!BiNvYdoJQ*30ohl(NVaXZsxv z%7sRAA1>3U2+qh-g-KALm7^<+ygj2sP~x4YsdvY)f4(4x#-9MGA3?>`4(c;-k8(a0B%H34J)d9ex4xFNj06dq0?Eihv5UR&pRb4TP@? zBMYQhw+tbI%FvI6c&8zgzAUQ@&5%Ec#E(Yp9&$U}uw1+Ztmtei&5V-&O7sP zz%MP}`%Uw9AJT4bgU#ce(jqHB_E5|{ige*>x7oW*VlF$vUOre?S`iu-XquX)IqtqQ z0@}HQ5XPWi{mc#9++qx=K~*5H5#1EVkt7hmK$^a}5fI2Etewy$>r*yqSLzMNK~@6? zYok+$*%4nk8yW$jUk*1mE4#f{sTM9oYIhD4gW#{wCTURh4fYk^7+fH&Ro)0+xKvcE zP8RKq7*LzN`n^8=p2?yb5OTrc4YySJf&{NV50sF?8pR&f>WMITzbC3Rw9yAZ)D-F( zEsLp(SH@SZnk?7MW}e+X-xXZh=wMIYYGt-Zwme%`;{&lh1$N`D^)dM*hNeus21?Sk zLUc#4S8p4gtGhijUAcOUD)mP2r8{-;V;hY9OSiNC;V%j4zz=JCZJBmlVmmwz39|6= zhDs&DPmk=R(ZCowBQ^)+O<4lAppO*D>xGvgyz}BR*n>D}@}x1SA{qhK2Mq%oMxAy3 z8vtb40`8y(W@XX^!jt6o%r5#bMe1a7Su=K%{$f@f zn*JBJkJ%CX^xIqQqKm#W!B%DT3$m!NXp-V)=U`9?bg0B$PsoEXU~I5e(*)V>u?CVu zAi{-X%;01GLFxMcq--@VSi~VPTsFuq!F(My?ZXda|4qX0?bw$&#@{6VUSEtC>bG^< z3#M51#ecWs7yn5%jVA}J0;`1-1LU+^XqO;D&RMGewm}-nV3*izzc$G!5Z_FnSgigu z>=vmOjvA@V~Q#E20dzeW{03b%7N73YJ0VrfwBhwu6>j@x(UC+Q6=`9u3@# zugFD`L!Z$-QN+$F7M_yfaHzMF?lc{JCqUo&y>QZt@YhN2mdY)d$(F2ZC5hZH!{5z2 z^t~O986(AHP-FuY(=Cb@#A~iA8pArFC(K{kHF4BKb@jXfb%nSXs){}uQ1=THMJLG~ z#eRCT_X*u*FV`4Dn+Y5VKx~F^$;W^2UwTY$0f;GlN;)~n4In&@fn2r%$VrNURz$T_ z3_7S7aLEyeC zcmiUl30`IDBR(182h(Z<8K5V3Vxmcj-Y?Q-R%l@P&C%7deYoitJl zD?>Z!9t~1FO(0#kKWc=G;gO)p+P=L0(@5Y}eRpTJE$yF+uf}6@!%`&B^0gMlBC1`} zq_38=fX=nir%c_a*e$pcTpPU4s}BkhQIq9g%*^1VaYV;RCVE8A{OG?I-(W1Z z#^Z7C6|$bkNyt4`)+3K%KtLlKq$LF=$#G$>|7!PaAVcMN*7!E**UMw2{T|2E*?xI~ zB4N6uN4*y+(XfNUT}V~#6=&;qNP9_fbTwV)eSE5Galmozhns;B&8edgaq_{bx#zyK z^UD^b{N(nz403?SNNKV{$_a|8qsS2|CRM(FPA-itC^t=9J)=jHFQ|jH23acwUa*2- zKLv9_4rS8QG_kYrflS&2DoloII?KIBSmyg+8sreuG@W6If__L14wBFh2bt(tX}4;v z*CSQhgoS(5_h?LRw|VXh!AQ*Rz+__B;=g~+3Kc9kGs~6eM2|3fFbXgY2i7Z(9R<+v z`#6E=t^fGw2bq2rFfDn1astV;mwx8)(o5skZeUiEzMI|&>Q8@c*OY=YoZ(;g!>99;zP|Y1_Rg#N z=9#a5c#^DeeigN|P&6C?&NyaA(W3Pi-nSM0;^M~e*ePa_2}8jSBzF()p6lo~VKT{^ z(keV0)JyKsN1&~kAyJN2CXcRne;W2M%(aWy(Cm%?IYW=S^zVcD=^?Mr3Gx^PEIC1B zf-DKxHW#=jW+bfgtkt1<%PLR9GKC2!PS0i;pn}SOgzL!&CWSKuM({I6@(8=}IMP#} zqP8qInNxoCTaq@Olv-Kv9EyQdN+uPvPw+t5CW(XXLm!DJb##T_j;Oqde&3JQO>m^r%tLdkrqrxM?tFosG7r%s@zxwCv{7Ak~$Dd*U>h0y@ zcuUFL(AakQR~DSyo|E)8xy@rAcexbinQ=`o=psj2Hp}B zh~ueS;N5PT(GR?-Bhn(7VH?4yONVxs>}kMa$YoV4O;Zhc4@xd%iSm4NXq2l3x}^D3 zAGsxnb2m**iELtE{6sm73Z%%qDG*tymekQWbQ^fnnRF$De>43PA=8*B@1}1F`aKT^ z;nlpLT406ud!F@y=iBHW^;sXJhz}CVZwZ>TeabxFTks>d(u<|=o8ag*f(}iNEuDww1!=;^6SuS!NN*MN-p%I!|euyhXW9 zmm|Lq&|Aqg>DweXWOu^OO7p}uf)%36vToAMoTIZ~_r835mivmyw@Cc>(n#|YCEi<< z*+Ns`y2(d`F0#ENx4A>c*Ks?_<<72_#;g7oWA#6yId?g1lRj$Klng47RntqQWneIb&?MK?G`+V?wY(+aaLL(zB^?_6do<;k49!NK81H1wVS;!mMUMf_J@qHG?ON+5z=G_lsT)lESx-&t2}d+{;8) z(?y_V1-$^Gi)5dX&q(8|yP!CX|DgrPiGRQ9rQwGE9~Crw-?rxVW!5f_-I(20(0g8o zyFqtSwj=5qeIM$cb}Bk%ZSmTutBZUHRSMaDosyQ|M^Rl69PWjFLWd-oiVD9yx(YEq zgGVm6q$7pStESx<*Vv&GaWeM1Z@PY6v+~*?OAw8M33@(Yw&9Af!1ua03e=2Uq{cE+ zj)TA%Q|tH`>>#icH5lh{HTr9|&dba8gTKoxq$=%tS}Rl+Lw5ufMZ`%nw1$+4Wxnf~ zTD2ilj@l0{w+}Xx4ltq`z-Vj$?715H=u z72-x5!=x@#H|fzksfp^Qv6{9?-=SHest?6`Qs-2Lrm7x9BK762 z3KN*QP`T>L9Du{fuJN=o9J$X8HJX!A`F@|D9gDg5t*)3$K9)TBBEn|Ipb*f<&MD(1>+o(C&8{ z^@qN7=|^|wUC=)ZSRAq@@{m`8s8L_~kBwi6{l=oN;#n71rcpx70zBs3p1azu8BaFH~8y3&sLMI8a&u+20 zd`7+UDNq`qs8-XeE08`7Ytr9RmCv{skrLML(dhM@iFEc~~x*>W>^3l0{x zB~tzeDGJa40Q3%w+KWGW^tBv%sR#?Ciq$A}(4ncAS~a22<=SM==Y}gfoxv^Blu!r<$s;1{aM)QXM4gb=}Ri1b^^peV1E#Z)at@G?XeH6NopfeVjL4diE>T0)dXb4kKS^M*5;`E&Ano9}A9ta`vqYB;6hZoWJh@ z*=L280*c9_$Tljbjn46G7;FQ-*QY>=iwZgwhK6}Pl9jXXst&w!;2ooO5ml(%5o9yGfwFzg@S8`!5T0HIb_8P8EeVx)8an&;Rpk=vf-&4L4NeRPSiEK;oqL%b7Ri_X{hug{ z4SDMM^b8qrvlH~?u?^W|WkV7vW)(%^sTj-#6-e_HX%k%AcH}kl^m;(NW&jsmW(P!h za`O&@1rR5r{6wUin{&eB@P4ut+c$0;^Nb=-&^c*h+9a9uWBN1M6Xg@dgIQOgRA9HV zQ&$d=-G}NWf{WkyAdA8b(RM+WsE>Ryt6#WP(F_&Qho^y#VsIwit3D*Dq&sGPGRvs# zm~}?C0=UEJl6=8YQY`G4b!b*O&}#O>DwRf6h|Jea&^k8D{T8{QXbn3MXj1kFF&}kE z^2x0Hxkp3;kf}PM&XAVQKlL{!=dGnaIQ-4^ZyyT+iA})<0dO^^_K_1hqpns^t2-iE z^^bz`tg4g(zQfMwc#W()S8XO8iG8hVqv7 z+w<#7-+I%7dV77K>@2`gui71gVjfT#*)QykL{9l<5wKv@h_*ug5xTMGWE+EB<)_$~ zkJpz*h340Y_9WPU!F&f(0F+O7oFh*jd5P;Z~y!~3pD9-S&kg~F0}QL7Ml&PMI(sElYZBTd>8Z%8IV{5DZA z)a0Z_Pz7{I!{MMzIclvn4wk)jlS8_}dJnu8che23tE5ANt5T_KL615ke8EB-d9cuW z4?%)uO5|gDW3W-XB)o}WZ!q|AP&)z?o7K=(?7E_O_9<;W3VcKM@+ng_$?#Cz10@A% znvFpflNL=fR4egHji6E6s5J|GEyOMf{T_#dFgc39u{oFsK|>v=gwGg^Q=3omc4cCR zW5zwfYTK?kUy$MxmL5CXJ@?Nk3n-f3IPb#~U%5W-x}sB^E9{AE(r@=n6kH_bU)}Z1 zL;>nWTqM1+CyK&(AO1B_@Q0z>T)X6PG-tz?Jr@}=^RU18X8K=#YcWv&x8dj0$)`MC zyH;8mqbC&eIYoM@m`3{kgexIe=AO~z3YX2Y{$!He_gFP~gHMC@{)C;vGU#rPPP8c^ zHYAD3Qp8T_7In&v+D1BQa@XW0{bSNCY6-X%k;L4e@M!XN#U9Wo&+xAZT|*X!v zH6-4B@vK8WNzBksYtmN=4+i1b6`}F&+v$DZte14^PHPtd(8`07MH_ssgfwWk)8^k{ zks#Oqz6Su39zI5!1fM?*`e{x2t-_nL56=H^%~$K^nX$V#`j~dRXR_DPh|BYm1WC^; zqDBFz;ox%Ku?42=2FZ@=ME$HJ>ujh6ek;_GQ_9I!p0T*DXr&8F1U6`&5%HToi%;EjqyC_+pL~ver{fDn(ML7<8A=FKSk<3*Qq2embM1tVPg zM>R$4mfW6qz}qOw!+#NXNG;3G`=t~9KoZ820xNsHiDEXQ+Y*D(G((=eSO_{~vkz*U zHK3nMs{9rf`WE``lI{XR1u`#7BReB70P||r(Wt)q;SY}h9>X7Z*)b=0@OSNg=P$~o zVHS8~`)^xEig>)KI%S31N{Ts1K_WaRBeF}d#jjpD;MN#A;Fj+_;C75Y=XFSOeo)N; zx#!zyqkm&)p{U;@9=PH;(qnXUXp_G5t1S>O%oi*j{~#o9#_5n{;}fJ1-;dljH(lDQ zxi83()<)ueNt&ZNY}=Memn({+mqd@w-aSX;2$206JNSI)Q5|qXHkCIpL*r=y(9f@t z1tf)=Tg&4O6O^zFD`v>181QB?sF({rPeu5rQGeDaBmAcRg1$+grokn)-8(_F-?u~k zL9DbQ^fqkdZ>bgm0S7f{o{k0mcw~qg6<8d4xL~p$(O59DJ|*}vS;u1|Q)V^mxfHXN zB3V!+2T9IEK~{JUL}^iQ2tOxFb3BdOr#`Kr3l+);xs}b}HF3h<^cgZTm z3F1mYY2-=@}*wR<;$ifTU% z9sLBE{>MmP^4KU{*H?cv{W`h{Y_N>%V8 zDX{ww1g2?*n(>(%h<1YN$6(V?Obq1%eWF1%*ikgn+)EU>dYJY(NwWRm`OV|5L9 zCR?RhrAvgKwNMzaLX)9skd*qOn5R+un3y0ae8%0dO_8eGrt1K)v;^jQK!z6XT!>2a zyUZx-L1M4@c*Wope&uA`ldL7<+a_+dqRws?OsAEw+GXbz5 zA=#|FGIysCl=u)q$OY~IT22ON)U5St@i(-B0%ay08;#9tJ2V?+uMPYdLI&Hwl}Yj` zg_Z4?x>U9h`_CdR<3(54SDQpmSDgeHPCwI;o4EnzPtm`+`Ja|0E%?1jlgLeO{xN^I zj%0~dj;ohqdMI*lh_RlgX;582!N#aIiHUxwK#+l+V;@<~;32A*qAHRl-hfh#=E!ir zDn+;?Jke*6l!~eJ8{UfpOTp%ymQ)9!C%rm6BN|y;weobRDK^|t6b|0fN4DwqM|Eg= zNvZde$;;uUWcO}5PjOaKH1Wz@kbK19(P1~I@X?AHdf&l;Rx@Zu6THk!`$kYg1j5ld zuX90I=XKtDK%EkWA{qEg^+I#yr3Cey`aOo?A5MblXiS*Rvt|&l`1s2?Z+hu3#)mU$ zAY|Y{vHHwA26X$dAo#vT+vYMDsU8SaxzxEHZWXvr_wZ+3dCbWwjT8P;$?{1Ss}%6- ziUni?k9TByt#$z0DF*UQTd0_o^IH^og5{zvaE_{@n)K)9)JwW%Vw;UC!S%8ex&gOX z=~;S{`w87KpIt#6njVl6zgoI@0dUa)rw9*VyiWPCCg6D%9y; zzo)j{lX1yT@pu7a*;+QJKyuI5yc`V517q{jN?|Tt9o6qqF%_CPP0OS|@yHMtQs_lv zFx<@BGHOklI;f$lfV!L4i0X}QDo2AXmo_^6vwD{;M}oKBF$*6eL-VQPKHZ{d)L)(7 z3Lymy?;RnxW|_!7?TV=N@*_aX!46Jqq=(drqLq^_FjuD^pL>8b>bH_R%58J^Y4M3N z;l^o)qD*9)B90uA?9*ljR{L(7dyj6^Cr&&RICy1lnfij>S?kv55HZ~HSdK>r(C1Hc zvo#Uh%d%`m)_98_7C|-ajZh%JCxIPSth8!UCcRT^Xr9MN$=^-;eKBSpL2rD)hmTl8}M%#QX&{gKgcx~z1E zW)1a7{Rnz~VK~%)DWCOk_XxKs3Z!+xV@~FMivtL_&f}==X9s~D-r!p)oBpZXmKFZ8 zW#KWymxcR{tpkTt%&>y@c_Ett!?*?V4Z^WgxWeh=fEHJJr!OIk9WC!^zizT+h4XeV z%);}|(dBrb*RBmT%W)Y{y=W{8b6Y?2aKZ>BzNiyvEtwtt$#)NtYKuJRjc%%Qs z4iP)c5mo*B6_vKTV%{!CEIZ(Jew{L?UJbktMxlt#^^cWaP*u~3q32{C>TEZ@uIguu}S<=z2Xg;X0+g@6iq7iW`(zd6f{| z8n#t&Fkn59TTVdq2*z|)!r6H?W(ZxQ2M>|7zx%qJM(Zt&9Dsf%&^w=_21)>3JZ`qL=b4 z;CKKkKa_Y^>n;frgPZitnr<4i-aQKJUY;$i)?s7w&mt=PFxR~!$`!B}2@0N@iJ=3> zCYIA(oE9N!f8?L$*!DKzEoNA{m86B6VBd0JO5C8DCU&nr5b;ofO#*UtP?w?0q;qM* zCill-nCcl52L&gLJ3cr~^N4TegparW^}l-mYTF2sxAMo++=Cb$FLx0ZzZ?+R#eqit{Y2R8R!n zkn-r;GYyZ0U182OE!;sfWJzB7d&B^-_aHaW9EhKI_#MmUXaDp+EGIj-Rg&?To^sr3 zt8|cJfET%kidp#hpR@nZGU~m%3pUMvuYl^EXLI7_yy7`o@)pI403c;a(=-|W=FiDq zD*{$CE29qwwJ4q_uok~be^ylp3*CS^M-H9Qp+dYujFKQmZ7X;@^{N8--jGD6NKiQu zYcIfO4?G;tt*Cm|6EU=PcAvME>>5wbSbgo)6jMQw1E`pbYUKlcmI(H{Z985_$dlO^H3GYz%Dg*0whcMGly66JW(sJ#|yWWWaG z(L1D2&J3rz!upkW17RTAfE1%rS&7iFcg`6_9_@tWuuHys%PesGx~*>K_MY=L#?_j; z_K&t(Yu>I}EM(an9rEz8`qM<0Ms{fWz3xd2>!MTr;-pvRf)nh9f`u4(fuh4k&|k*v zV!05-o@1kXAceF!DmT19-mJ`uTpKl10cthZPWUi-7(3yYurmZV7+2nI?;US3IC|6j z2C|69vARqvOPE42i4<8y#Vi$dLXQFH-*{r;=RQY7hNl75Q62MMDBIfkl3aAkK}@*l zr?U^s2@~TQ`$I%33notd?RTz|H9W?|E-OrIp%?(kCM<7HiE0-l$zx$(lm z$)JxrD4^UUJOr&wvfSH18@G+#6HrHk6f@lKT$k4Fj5r9A7k!3dXf`JsD$3$ z_iVF*yj^BkGJqK%g*_PH7o%{?V7pjkLzbwIh%8NIu}fu%`p}#{2=9Y{${3(y1bb(| zhf{%Ie=tt)c&jFJ*4J$t+`ODL<}tyErIYm@5TZ!a^k_2WnV_V-5xNwCkKCZb7O{Bx zsV2_TteHG!u@>&p7W?z-{Il$cQKu?I|7BTWeo^-i^T>UB^fw-Fg_EpARThn71}O5F zia~CQ5hzKeL3@K5^<5x+a6x}r)<7>G|GCe@h^pZGAV+xor*ZFYo_|3fKfVsm_X^Ir z$AKp1VX~bbfd10;z-YnFR;`*ty8EJjO^v_^%|i|bVAWu<`x^IL-97J0K~->zVgM-q zN&cVvToV^X#E%EU6aX+EdU{m_CkoaDB#5#-`XXxt-Si6hElv_s#Pab?UKdn3!Z_%F z*BNmuVlnB7ETn+-5L`X?rU$ALX9_cQS$=Kgp}H$9etf681L{UF8J;)2XyQ^*FG++A zBmNtV(2`~UoQBYfsoNnEbYxI7ab$~{%^K(Ij{xacVDRbRt89xsxFnT$OnBL5pB8QZHMj+6SHAbLoE1%FrzFNisloK?P!M@OpWUZkOO5 z-R7AkUIf}(D-?0cG5Ufze6yBUZVbLw_z~HJeAkW?)? z16y?kAWAhx5U`$5M*v~{#Mpe|1PJ~*i}70){lyj#FFU(DUNBhD+pH{weGfRem^?&& z{T-5P2urOd^*;IHi?EW*ns^;PI*0ph*Um2_D^zwp?k{%GVU^a~S3 z8UE`6vgHku+fX*t6x9eK2q4X^uMjut*AP6~Ko(7OF|ZslfcCx5Ve{>R!@qv(&M$49 zP~JvSSlr4M&A_w*-=45~Rj1y3l*M%F3w&1zdct;VQlB>%bHUeVZ%D_d+TPQyJI8J< z>^VQnmsCD}(}A%=x`n~kmmPw`!A5N=X_W!L8XA52R|-)a@_+XmX0UMr#yIH&?;lOE z>`wk}$1nbqY~pcn9(vaeBUcwv3`Es(shER7+d#U_sNL+pHGIG=T?v~ew{xLepjan+ z@I02|IEmm+zrI-|oBvLU2Ug=6?v|)gL>Dx;}413jEm<|X` zRm^=HHYPFoVXK1Uuo(8x7#wnf#W>aKfgQmXSnTd9i6eRANv+klUrsUL>XlG2u_5V# z8qq4xF1klp=6go>MA1hK80TK`}r=2-l5HMrjsX;5c~SO-CVl#(^; z&W?rg85y3P2waw-*zsKWk^FDIW-%`F_DecR%@>9O*lwb1*Kanh0~Yy_XC*{?V<9nW^j_C}>dt`oM*xF_D}h5C`3Cbp~b zzBHiaR?@kOSkUi6m9EWGvDag}cbnv-AU=45v_yV)BJ`z?ND!@c?@(V*b<9HEZmhIP zk9zCvnhHPYIYQRS%|A13L>sS>p>4Rybl{i;QLj49{H3?Zc0a@=qQYN;N?5klM*odI znRI4IsX8%usKY@Y!A2i@eT<-+;o^vi?PNF(4BIDN`}tA9lj~jeAeSb}@zm>C-UhdRlfS)^U{Bnh86DIq_6!?azK}OKJOpm=6{L92Nvs%}x}R%Aihp zL9BEuy;!>7V`0_odKG-CSHq{wz|25^1r$1i3!bJyuG1lHnQz*Jc*>>5W#k}nEzxP! z^=JRAu$U6{U;n>GlElrF@HlF})5?@&QVe8#Hlke9E?`ZU%4(I}*e4p zkF5er1sBFSAh0_4*Iu7wK|OucHxYQoW52LCIP;^NNA8cauPZwyf0S(-0O2i#ge9ag z7{tIh`!!in#8yv?T|9Hk3+!N^`>pWsv}bNuwQN?Fy{S12RpHoDr&-xes{C9I5eE@5 z%)vAta-U;vPp~7xj`KP0-m!#FzHC`kescR<2037_;)2IZN0Zgkae`v%C~^dMGd&s; z>{%W}W<%K@N)AJ)%S0DVuMnrodlV>Ek*2|-o#mo3Uo4quVu~;#k_}W`Y&D5b-%^#& zFhO^wbm`(~19rs5`z`$wK!k~2F1-VS#5;ltl@(&tFy0Qko^t4$kp~@Q3|W4g+;;?l z@wzuFXJopeBL&z?h2tn<1 zMl@!wF-eUT5zWri&05byX1hC=_5& zWUJpLKadlUr|UYQV7O6`?bin7P=x|igWNc+Q`e-&J1~KlA8?OZp(D9xL7BYuN62Re(DiR969gVeDmoG1R0)|E1jEy zlg*(uBshQKgrRY*Cw6>iH0>vO?E0{@pKMUYhO~#FVsRyX$$t;IDuALsz}TT#sX7R+ zK#$`#5KqaYfezq(M)Az8OM5XiOVkMcI%)*?Nu?`ldcsdXY3Kj|fx!TYMh~OTBX$=e5XYN8Sj*&OV@%u53L2 zUmYJ$G?&wmAMojVnr`V7)xeFPjvr zO%d2qGE0y?(WtA{<yqkkB3-I3Rijrdwb{G%9Xy9edzECn$Z{Xktue?mW&Bnhv|J_UNqBJnZES62g195a^pyzjwQf8g!)GX~t^HI>={ zw?0|9vW{-mT$y`CxCDmI7Yu*;=IkRrXX$#--Cgp|l6M}ee>p;|**HMX|Hy7w>A3k{ zvSqmNc2Ca2S;^5orm^fM+>{Os5N_g66rnsFl(WoKJXExWY6;7xB$mkqksiGv&O zJ-X#m2<7dNUKcF)s+7cr6etYo%HvY(!#<#HCHaEf@Ju>4d^Zr{T+V!h#bks=!%52! z`!MWA!wyEq1uXmOuLCU0*>8)FB$2J-NsZN>x0GVw1>Q-;AVoEgJ{(jbGMv`-L+~Rr zYv!g{^|pL4x8lx2E|$@UnKP`ZXXl+8UBDnB;lphH0Lx@D&%n}vnF4ayMX z#e>Odl*ipEHhR{pfEE>;KMjR)P4pP4!dalj4aFnD7M5FiD&=n z2d_Aci`R_f?91G6(SCW_$~V1d2bO2#(uKf&OAKxFF~Zh5-3OSIbuaD$;cO(>ACDV8PDgxer7c5>OURDL9tsP03boGDG>fB~l@B5- zp;ZEsXa1k?DIMGY*q>yVo>jE|UT4d)OiIDDniKtGgiD+9uAHKTx)4 zHYo3e-G}bkedGZ|^o`m|VVU>Z&@qFRV_yVAY?x=m%gjoz2LC?lA=Jq zCNz`2J85BqSF&hbK$c>~gg%lL{=^F}R)${tdcViQ)#9B{7`U(j0yssoyQ)t52E9oO z=esm1QB~gkz))ycFT{LwnS1)2QOY84#l#DAy!Pm&dFFD0lYx2b%g_G(DzPAFw_Da} zvYf{V%CSPwdWuP51Gw zr$zwneLSF$P`yiFI7xO!bZ9n(gLIIpOxC9?48J3C`38s!sH_dlXkqi}=hsaN{+(?I z6_k`^3UxW?Cv$7-sk@LH?Z~?OTd=-UnCux)lnJI1@Z$r$m}5t8=1BnwKd90VWYMe zB;eLgxCYf9D`)SQo2H&6CS{JUp3V%Y4>#bANUL-m{bCJz4sb$1jt=TP8i-Li;I>Iwt!`7cNgyLENesR3QRUsFFBVqQ zw`V*ehlB5kj(S}QX=d_;2O{pNuS-`d_fsoL*^HY&If(N=>bw8{XYWhkn#!{Dec}tr z#gGj!c>=OTkj0WMjEEJhda3EDo}RAhnOZboZf0?(%sWN-HnteDk3VlpaxJD zK?D_qLS?f+EVX12G>WJo5tqVZaiLJ)e+~&sBGJ5%Fwy>Z{o1_cz8ih_%em*Cd(QV! z?H(6-b=+QeZFm!)Hz|6PD2emzj&+b{HM~7TpjD9f`I)3<;{=oErz1PYiHW3><|Wzm zbtZn6Rs!j=!UjliYY$1_=W;tj`sg^4&AqUsfBsgF{)pAHAQdUzhOkCZVLj>}C+LrYUQlxRBHTs6Wr+WWoe0Ft{yTHhB-k@1ue8W-xZjV! z>gP*DwHg`}aGE%+kWQTG+)rmMjpLo7n;_RNNrVY4$d8ME2Q%q2ErrrEBVx(X%h6^U zN|ViEzQ?{1Bh|wzecmsAO;TT*()eNnFlJK>km06NF|EsPk!pH1BwZ%+)`X=1?;B(^ z(W+}v=ib;OdHm*9Zc#K;%3^5=`aZ4y_`4hKMcfQdeBh`b>iQg7W>X_I6;!<(8OF>Bl{YQzdG)DM#)_8Ajqzb5M^6TII&g7fnG9=bp4B*da?=SQ88-_sXlH?6Gp{a#64Q zNrKYzgQ2&j)}5NBhBE{2#v(kKdgxp;-Sny#%MXz|o_FV#aL`Ll7h%*!E#K?X21Otr zIyWc=xH~;g3d*U11w*hZ#_Job9)+3w?OHJAlXqn0tL$t`Z~vUeZ|l9<%l}&Zn6$9- zYVEgxj`SONwOtg`L6NIe%)x-0i~sPuKlk`UYHbT}W7mSvM=hzD4{W5pj;S(CA-@Z? z&}E{O#U%i#uJF>JQ3tG@9uFEOiCUKwhUe0mZkRJQ7T9;w6 zA{9fFket;aLsF4+I@`S=a7YHbsJZm6IU5#s(2JaWA%{drQ5Vj^vbxNMW!-9KVCI0{4Xc))H&(9r+^VVCRVdl(QqPF9Z z4J0;9AdoAhnEe#lN5y0<&7hmYGera3jlP}yBjS>UYopN}u9hawsc^!(Ey{C(E^>M4 zX}VR3E!FZwZXvkG<=&rp#0n~f$jpze!CMg&3a#szjK68~_nJ(afTvwq^?==U_0=u& z-=3)l%&IlPb!61RZp%`DFMT3lHd0I?Mb=U=TSDr@m?(+mNy&4&!%sTeXvkkOPG-y% zFX_QHw^<=#=AwUz{MKweB5wVx{ZC{qI|s&&XO>(8{Oq6@h!AX{Vzl5lR4pELz!=-8 zLk6!SBsO}CFDg}r&P64^m(HKQ)NRCo8AzI7RcwQr$$DMc`nO&RJvL6q{CNvG%nloN z?A$gQVB-YER8izO71QLV204ZeOrI=Gk|IL)7K=Xy36iM~$#zB$ zZb1MCS5+F|5kVrUar7Vw8UB~8_&lNzgdB2#v;`=q$gS{hdQ87PU(oJ!#CtEdI}{Uw zlYr5sbUyf|Q!7Vf2cijhGV5?Mc$jzXi(e}K!>gXbjhi(C(gu(2&?HXoyj8BMIDuMm z*gu(=SVoxlW-MxD?l4=_`4k>vHS5d>Nd1>l;|%WSH>B;@ewfJ9K0*c{V;8vOL}`)% znJO0+q5A_0!*QXjmKR6HMjG+Z8X!ClH}@>|S>%riqU1st$=)HGyvc$9s?Sg6i! z!}}&g>5veknxDVmAxA4I;P%sd=!*gIf!(26euYo%qIw!Xyam1L0!zA0uRQMn7$c+^lUXzH3+Q2tSrkD*BSx3ddP7v-DCy6SBH(*<+ zOLj_JH%*!@<3ptTGz&n6k4($i-|s#2nKAjk3D){t`naN>+?MVYuXTGUgGt@#At%N6 z^qeZ62cdcM+x(D#WLkcV<1fancd9qgQ@~qv0^&burCM5v&F}= zBiloT9YFSfz0cV=R%#;1h$B!%jwwzA?3e79s4_SiVHmTH6U2J%lmXW-t_V|nr_&s> z90J?Fd;K@C8Up+EFcz%w`>~JSc&vKgGm)gvA{7R|(ISe0A8j6pz;cg>`{~jp_dViy zdzc!JYTy$Y_DFNqD9DZU_s;ke4P$GgT`qrAeo zxwwPuP#j(MyOmT9cgQvZODqxd>~}li$QCeSg&6xSYH(bd-TB$WD3HqIJOg+b_HfjYb?kARoW^N$`jRUdPW=AKm9pc?a`J@PGUU^8}8NRh)hr zRZ&+*te1638-<%9GsK%oLF7rG#n{Tdwy;Z@A;w%H{Nj4ip1D&J^{|FfoM0b~4krPE znU_o(#k9p7uh%iw-ULh*C}cYzi}TldT_Gm|(&+*T%KgC3T($yk3deWkfc;|Zf3{y9 zSqKj^2ifvwrs1*i?scy5Ca>Mb+D&ZRrqd6k8s!mgAFq&m$i0(at{8N?F22m|cdq0Y zfi|-4PP?#T=@rfe#{sAWTQjRnavI2TcKMc5$z&tj3MpIQ!-@h43KX0m8}{%^lqZf58e7ocsDA?M-H z2cd&RRjsIV>!A<3XS*Y@Vj6$c0R;u8V@>+>UkXB|1nk!tyjvl8$5CeLRICYTkm_lm22clyI4?(~VW3;f!8@qBv(NFIK_8u$(8gktZ zbsNYa*DTzNt3uFfuO*e-bQ%+zPb<_4Y|#Gz zi%y%UNeIY5xuPavnOpCKW06)$&!&xA`lQKng5H zRK5hc_EkmPc44VUith@fWIF7BR*nfM*vwYZ#VL&c_@ruj5T|g5)r2{t%IBSEU;P5R zVD#IiB*TuQ{h(7gVf*JG#T=kWJ{5B!Aemg`U}234-KxFZ_&`uS%~oJ)tLhkJGIr9R z$g4rga=pu_1IlY;De~L~C57P+Bax~{1)_yO9;s4ujs)QI045t%SXcovHu1k$L1yN6 z7d5rc)kDVbmley&b~}blsR3m6Q4H_|?xAAFtFUmLJR#0()~vrN<9*tDtMcKp*UNirQ5|rvu(-iHbtmR!G65>QC3iyF|Ta@E#s{s zyIobA*u%=DF@6FwkQRuGi-kkZswRGhNR=+C@PesHgesM}v<>85OpgOwY?$uJx_8)N zqf}C(HBLq`K@i?Sa)YV@OT4u->=|~+@OQY1(MYFrd8~g^X1;pszE?APWB2OqHo>Fe zdo%dP8a#GuFHGp|^z*iQb}Oo-@j+OX*U7&xYZaG>JL$pDE?FxxY?T$N8HgBfVkTn2 zuNZ$~dOlepma6>D&)zaq(hCy|+9X@Ujs&PWA;gAcRjTcr@&H}DEt9)8YIi^qNQCYc zALORa*5#x_%Px?-9dcglmMz}v0!iw0MTiPRaJrrfD6E$*hRPM|+8px-XMD+E`qP@j z#%gfvH|Q4m+wa|a)pS~%a6I|a;nKb&WHoHZBbqn zRx55QSHG3ZS^ZWLr%jo|#q-K#xtv%kiDMJBTkD|NnmE?=#R|>Tzt8+>iLqb8CQ@s^ zKL=oAo49`7L@`MeSwEgFH)*xbOrEX6CNCUb^APmHE^Ko1&_Cvd8mA%HtzTjy z{@Xf7~kGPme^z&$=Eyz+LN-5xII+i?UTY0{e*9WXA%k zC6%1x^AcyP6(7tx=h^AfDZegX1v`Y--7EQ5NM%T>Z1iuMd@^{cf;-OX^i3yixJFqV zS;NVdXu~soy7_tTM}U6htgpsBKNa%X*l2W`w2$2K~eAOwaAQKQ>0` zP0zMmX$C2^U!MZdyo#f?cBB z?j5qR7Fzl;qboyxB!%#9oFJDjThbT?jZ#E;a^1QQ#hV{VRTu@vB7sCF9fbjsEa`)O zNPf@`tN2t|oB-?I*9h(cf!6QF{>R?(*aAq;AocW!HMUp*X~ysWGN^dfD@3pi48n;m zU}5$!z&HxvsTAG~24ta}ST3MyEZ!R@7;?Ty4smyebxez*m_-0#J>2XcUH^4UD&geKnhG=ZtE>%`8z6)B{YS^x8 zrUyZoWe652xNCEWyV`kY^akc|z~KPvt#i$P=}!&yiw~?cAgq47_M0R0`wwf!h{9VaN#=;Qd4zCCvA>S!Kx7L2M<|8g;`C#0UTE_1jmS z;7=zorowzDOlYfd>ylY_^TFs1n>sZM5vCrt;dNGsnCY?2Wu>w8VB?b5aa7lYttl3w zUZLu^U#|rJf6fu-zI4&31B$ob8Yt&(^gff%cqaj@?&Cn41uzOQ#rMRM zB+I)v+VG+1onabfqT5+vYy@OuLXnY{&Qqk#(E>T3&0sNMSS>=xgy(H_m=zY+`Zs!C zr0pxWb4bZ#Wr_CdOOa*+9Mn=w4Mi%cm=ta8204Hv0uamP5((M74tC|G`*b`FZ|7 zJDz$>h%TQXr=)|SNrB0{4q2?=pj+#l5vi&~*yVWNtJoWP>~~CYFNoz}ut5gPuBX)8 zG&1IV)9;U?`iz`!c6?(7dE#KY@g${ac5Hgm3`|cV#jK^sYA`)Od)_PADEr_aM!quz zBQT}$cxf5@)Oxr4>FC-2ss8KANkG2Pv5bVAFq}q^=nrR({{v*0XFMz)TT7oT&D(rC zQtyNOQ`!CuQf$ZZkS2p4{YMm2O#v@VOtxZJyw$lxe2T6NQOj{#aG&Ev;DJe%Wi4$9 z&k$rVPk39o&#t#96W_V}!|Jcbzq|i!_1|{9o%prO%aApppT74!s1$S^h8^TSZ^@z+AG8idnuqj}u3(iANkz@AjH_gKvv#8w7jR^7En_ zGL5oX*yxJ+2FI6ZrJ0J_Y0AL003T1s(3S`6@UiZ@A>Z}V<3q7wZ57#V$M`sAfR6(d z1N_vvR18)kXZm!>)bi`%D}ha({iK#x!P~arR6s0j@12qsN9NMGLCK_SiQ2bKYnfNIChnd%mP=7KxiV(ZJuQ zm_CZ!qheZI2SPFQgy|glkjIhCd*D<>RG^YZUj?G9O_3m==6uWbDAWoszbNeHr;9Se z9+4j5@(oNqUBty)#{^0v-{evWZ6DLi$L^)d);Qtjpb8H(a%_l6{%S=6^)Wf*u4)2z zQw#LIa4$KVww(B+k&ZJDwicskNnJYYMz z_PhSPvo?GNR}VV}e}^~{Z^s5|uYrL|rI^hW*+|8pquWof`ua7X!O$po`fl{ib*gqu z;-omP<_-JVYJxIlILrpDsSa#(os~JF4iAuzjQPP{lc2u|m5zr5)4Vf%n&a( zpp2XIKzdGG$XmDYBst4HE<7Ez4g$e@#a+|K0&g~$^v2Cb*mG9}eY_g)Jiiv@mVdAK%Kfhnyj%CBR^=+*HqqGCka{SB>++5b#&!ot7pZXSlVJqE zSDF!sX6CT_0QU@8#mn>S;~`^Z3p|1As!ZoCWb_*jD(S7Zh5cb%vPZ>8h8h3=YMyTbbL74STgm#XDj`nott zgcbalueZ)^yICvA|J3T8nWXZ#7qMZcMHWNQ)B}bGZy;C z#a6y(9cR=56FGB(>H|}J6P%_!3Ce5?O~qJP@N9kadl$c|N5voWj@~Ax?06+}(_ox6 z6a!U?%~VVtCxchbucwnl>nNzHNThHbd@CZEhh_wQkFkml%y`tu$I=Qwsuwbtpr9k2 z&UdQ}*u&(Dx@6_xgtP~ZWl&}*wz!=q`A)@=c#1_Fc&;jnP6yVo^)5QO2S`;I%WWC} z_omciEP#{Zn-2PviCpYXceC3GzY4$fBgM+-0?ItxnGF_m{S!0VnS&jclhJ$MN;95} z*er(Z*x;C8`N1?epp^pwnH2-l2cbY62ZsZ);}KbkRM}1030{R$jrXvJc_1)nKwjF0 zC6Zp+pS2o%X+`aTaR7x4HtaVv1R2^B!;eqHE%~B9j5z#pC^}z|=mZw2!U?UGTE3rt z6jkcc;BjMmt2@}>!dhc)PCxTD|M`M(j)L7TZA>sB=y?BhnQJ&zWGSXy0#E;{%^GEM zhQ-QlSj{6d_kSTi%9yr?jjgcbOj<~ooUj*NOfd&3a)63KPK6XGR{Ds0B4DM{P3I^4 z47$L%+IuK^h(16XM6oc-Xq0LGeNS86f=#1SbTzm%cs*Fqe`O^dFNrfV3ptH zK($Sl)MjDq<^G425jJB+o6iOxzTUBY>*BVL$T~ZAZ1)-9AdO^#Pz_Ov4i{05SPO~beq)DWXDm5WL~CECoH=Ty6d#@2Dn%Zf@HX(4#xne&Cca= z+yw`TZruhY5~J!8zfrINNtriAYM2bDCc*26q?iHjieI7yd0u#?Ts9)jrBjwvhE$P? za3D>kPxDOeiL=Ay%k7Z`EhfdL&kv7$^r{z;A&;HTsZn7H!~c~xW7ekjq=R7Ni>z$V zOgEM0`*OYQQJ(qXagzCpm3>qhSeQc;1FG%&shB3$9@!-c5+pXzshSLHPMEcdcRjo}Vm;dk*&|*b8?9ueQY9+h9 znxRZH*E`;^!l@Dxqt)^P?zROQWnE-H$p$Y!&%N1eXK<{bPP8+4t($cR#R`Kl{^Nb| z+huvz^k7UiW(1%vXj4(dLTt1nNEH`uQcj_+ZoPQ&(B>u zY*bld1=zC$w0X_@g8yr;@x)}e5q}et5PBT(3~E`X)5YF*6lD zO_gcg(*=!znaSc~_CavgyX}yKtCkN%7e`vx*i9INsRCd^ zS5`e_XZPIZI=$t))lZ84#MiAx@=W5zlnSoEMU1ulD1L1|(y7fMMg=}%=q z#~(3%q?WJbT$hbEK?Mu79F;3nSi06GS?`pvU@Rd<19>qU!rCOC3D?VZg>4e8;Z+K& z=|(>0?5gEe!b0IxEzuP5GVWwO|2ZiN%)e+FFOxW0Gu|Jb-}C0U!MWYMlN`08Q*#bDxOs;rVu;vkC7IpzzB=?9>cx@%6AkLs3d9tX*3bLlui)oitUwj$53 zTe?T0%5)x(rOJ{-8^Sg)Iluxn*19ebvDfMHVK6MA<{b@q+J$a?h1E{Thg%U|!Bcg} z1_`D@Rd`*X4?4lZ?2ywTCksjEYzC$Y@G>7l&swqr(@(?K?l5+a?baZfaCg&1w-;@c zUH;1TWzUXUlz1NhUg4LplpcxM)1s>b;sl4HG8MNVty|YpqpXLOqVBj>toxj*k$E-} zhDTGzi(z+GJ7a|xs?2Zo`wR5O$3dhhBiVLre1KegLNdb+!Al=p=ipunP5-g@c#xkGkqKXZlCdff!*4f7i4eWb;E>8 z9f@VWARnaAZqO~$pS*3>&La-I)rQ;5j#ycfnGwE1%2{ts%D?<>Gg&{`+R1)HO9Vo` z6D-V5irG%V?=|LvXDxS#TgW{uEv6epgUpCSX;dKx5}`QF&hgGiJaQC!+;%9=)1UCI zXD41ZKrHRZ%eu4Lebb1MG<+E5s7FWr-&FsZtg_>|W{&|9QYdB+U_m2V&f7 ze&ReP-iKUqsQ>=N*_JEg$w$X*2u%LC&90h;z$ELRnFj_||LiAvSp2GOp9i@yi>x$Y zSQwy~UW)WkF{!euz&(;WZage4kMqk!6?3nPlY^uf^;XxYivwK(LSLw7@RW4G(3s4iI+-3H%) z?Jf~q_Tp9DbI17=bFq#CuV9YTvzF^-r^7ApC&&}` zz?%A`BL?ygaWd&_ZX6ll=Zo^hn26Lf$(FO(gKOvySo`cbkgytD`wf%%uH3OcNRPTd z{ITme*=NVb=%j%$`jBFP66P=!Qz>i(m~NL0hIUII6I>MCa!nUChuobz;&6FsyvJ4G znnKB;KJbI{=C}E+c1EpNjdF-P82S*}Rtisn@Iw|H;ayy&pB#hxX~k$LD#KROgA%Jw zXx7B^bHJR!1y-1*!rqzrKgQM3Y&M|mcn{n}E%g=Q6(L$wjNV|r6s9@oG>GaUT7w0T zR)xt+N0J4OUfP#68+_+o)xW&ztDVm6x0^P$YZiFr*-qp9cN{>{eg>DOeYF|pE4 z_wRuuDZ^|lU5o~jt`550_gd@rP?iPTm-Rkf^Q{Ms7Y-DA+`OlvGmT|VHVhslrI$#SykwJAj^G5}R3#i%K= zQ>PSOEnUSRbZY29$kZ)gqB^eVCzZmDzWc&J3j!1rcg!CJDUrfCmzSz^beYo;zp)Te zlYzjh*Ajm7&u9Kgj}ei_FK3X($pW`_yl}Z^fQ0K5(@v4gx;oe0?h{p5vD8SV6JmzN#V@0uo2q19@W9HrZZ5FUndv%{RL9IRCX_XcGB*0?2cwuJQ1>%NE{ zwF+zG^c7FK&4(2$R@89L1vCc~a5Wyfzn@wT7fo0=3xV|1K27~#a{QS#P4u@$`i(^u z+4wSc>>oo8#Dux+sWv1^g%5^y$qp)+>U;{hMtJHSP< zl!U-DoRi+=)MdKXqea>AcX?lHRjTE;7j|)~C&@9e!N1DfcRRsgTir-Z_Ro4v3`f1- zb4sd*Ncy~A{FNz^RkoRwEVwvX)Cl9(XJ94~(Dp>J}rJQl$ z@uE1xz|G!FF&in8NX0zOciJ!5!{i1Xh4{`T;Zen8l_A!A(UbLMCMYI*)TY;1fnvrT z&udFx^?kmbVI3eie|mngx2jN@?pHGZ^!!nWJjIdly)JoTo$8Bq-N~1XpP9(e%^ol5 z%bK^@A;ast$98;0Z!TsI{Nz(oXUFE^hJm?grI==lG*B^X+;lv+_gvSwVVS!Yl!B15 zCmXT|)bbJOhmJ_`KO)@}l`lf>%29_wvVo~3c#Mma13u^87art^T^d$-#l>koo2I zK(abqOm}$ay`otGdT0ct|78=&VFwNSb-00za$>?%DaAk*%0W;%lj1TNNpO)T1DPsO z9BM$Oy6aXpZylU~+CyQup2w2A(RY8q7|VaL@X8o>C5Gp2xT$cRi_p3od}*;-eug>3 zQRNE~oK%>GpX}H+Wk%DdBgP6iX3SMjJD%c!e_&LsgB~~qpa1-iB=)rd4%7}#sNzYc zm<<$JhpI!JbPH%ez}`+3jJb{lc{**#_=uSWiSfqM`04j|EC1Qp4S7v+S@#=2VmHO4 zQDhqyEf2_0T&`Mr%&!=vV#bKT+9d756kp(#ydPN^a>27$wq91|3Zj_GzET&7F!8vmMVhCX|?aTpRovB_L(kCh3x`4_oC50%fRsm?|3p}u?|^|N&{9k%MXmuMUD!SUI}xj$TU{?o zyCNP%=*#iKAx`FA$lVgM%eN~6d9@Bk?Q*N&0VN&C)uhwu!6}@SC41a9yA>|`T`YCs zt!sX}d`l(W%2Ral{D*H0NE!NySU1F&4-%vMWFyQKSoe-u|Y*)+=7Rs}VJ@mj z2T~?Cd*yn^2IqrzNTamRqSf+ndooB697|C4IG`)9X1 zJtnp#Q8MT@N%2K&RnM+<+C3Zn*O$x^lfJvDUNj36Q~eSPo@a-N^nka&;HkGX+NGh_ z$WA-9H01`CWXW0s;#vQl~0OQn&g%KG^S zydmoqQ($$;(;cL23B*rutFW6+=IwTGcY=oX94qsQjt`(&XnBbR8Sl%QpRCX_<9lV7 z-w)HHMfyLg-K5Bl(ejA_TB<1KI7P~-n1@8eXq0=Yn_oZVpB&gjwuTSAUnfvyw@=95B!s_kYZ=AyGm;7HC!{RlG z(3wDYg;Qem5D3KT=+z-?e|wQmiWmR1j>Rw^M&_cMwS#K@Ut!XgH_!ZMHBUd^1nv08 zQ)Kll5DJ`VEw)n37K$WOF^@p&dc6z8voLr8uQQk4FQ{KwBJSp+M&KjvdY5T^sb;@X z3%veGUzM2`S=kHwjX2D$)buaY1LJFT-#kc;+p!~Z*#NO;DduB}oTOs9LG*bY(b7kN zMi}YJ8@OF^jYnT3#_IP2`D3>N3i4nlu)+(k>Nbool16E!^U?XJ=79w}y3)PM5KNkF za&3XVp9F&GvlE}mQdEZE>UY2!#DkzHs8QMx(kKN=K`%^g$QMjW^P3IMjl2)DPR)e% zbIsV{{21mC$rrWB@&%X*H+`^Shnneg zQ?~!vZILnkxe4CzO_6zigOXZE@mLFi%8als*^u*&z%l`9CqodVTLJpIxM**a+;<+9jXIo`B4fc>PDd!F z=%@Im)3MR}13nAS<5bN(>9!(+SH9#XNKeH^my5JsSe@R<&j>?LNDChQRSy0ilrJ%+ zSG@2MDdLt1R^Yxtzx#Suwv}cq;cu41GI=9*)0S!Y`z`zZx2CdAwqqB@#5#GSN58B9 zcmzB7hutykRw5mdZ4>QL9A0*9+G>!Wj-I)IdHTRK9jGGx62y8#BmK)4Tgb-A42>NR z{s99+lTI`D zIhF`>=nc&DxFqJ{V1hB279^N996wqS9%t;tn(+QQ`RRW5fdllsdkrLXq|=Y-n~P5b z?x)vAwbOn40olQTUe6jUir&2Z&9HjiFYnH3Elj`eN!%3s{q@bb_ougwtp}UntNs4V zhzYxL6DE&MvP?yfY*Um*c}twK_)*jeK1dO|7LO6w9}%qyyS-?)S69SZzjomxAXNJ6 zwkvNgxN~PJIn{RT;+n{*E)0JdnM>bZ^qFwI40%g?`Q0zeMYEb&G3dw~+h_2=G@S(9 z51HkwcR311zgyZs^3niKUQUu&Ad)JPO#g(hlVb-dmA**0mMS|g zFM*^4tD6>cATz-;F?nchBzO!1L*_T#zqv`y+HuhSjsXa+P|PKYT%ck)`PCAn zNAD!D^LEM4362HCd!F}h;vSN9z?M_KK%+bvJ{Va;7sJvQiImbH2VgMrL_pD!Yhu)P zDx9+t*tXK?L}q~7$j|q!gsg%gSq*rxb%6ztnu5Kn;Z8bz1uA|UxiL`Tt7; zb)0JHP3JYRY~L{FpT-$GP%!fAgAC@Rf;e8tmK+FOROcSVJH=-YZB+L z^*Sp*7?4fJ6IDj!mZ&OlZbu!^g*g~dOLomiK6A`qKqvWPK#n7%-d2X(7bf%Yq{1uK zNrf@~0lqGue`{n#cyZ)*l0^@3F<}z>8jx*Zbh?8Hfot91FIws`iHrpn8ycFWlgSeg5;oBrv&PmOKlQItUUePaWTsG9 zuxzE^o-2w~BC?M2R|=AuWAY1bYZm8=digz2dbokXRG3N_V;tKgkRdz}-e~*6p(()R z>5$sufE`RuI^Ior)twtGWJi+Wdt+uw1?-ZFB5u1-b;3(at8l{+Q=Qmu^i5%CSR!Kz z-^A`wcF6cqz@}g`b$c2_E2B{pRLd`s?sG{FT;*94&PV2BNO1tpIHah7BAL3o-h# z$5>%_#&*|tzVWIrl~7W$4|q*SWLE{K#I8E*UQeT>c8Ry@hC8~MCEktBH9VBq*d>pR z?sFP-(707CwiPm-jgW;HdDelAj#y!2ru>5sTK`_ZRJt8e^BMWrj+aW^2Ik}v#ay7s zd5HLX)dgylJAM0UOxfEn*)LhoISVoUy)JD`4O9~L2z#Zq;tk9ma!069_CrL!lYbNh zS++VKmL801b!~O6a?b$5gKHo=baU|qA=D;EZse}#WQ27qZYtNz!X8gaGNA+w`)yUG z$}~ailj9I=T)byZE{U;yAg_X7o{Ez9+})#erkD(Jcq6#D)=9pF=g?Z zO*CbJ3KKop<}xc(%qa8vZkutj5SugxJKm8rQS;PIpB1LchFm_LyN6lt0WlJkfNFza zaJ~R|o??MfVOnv-0!Gu9p$*7vnXzz_vAK9nfMJ3U8%Z*-4o(djEt{jTIpDe`WR^zk;jUE<*0aWAQ)?{iD$Ym~j7I|Q{L^nIUu zRH0FB^;zD@U&(28t){E|k4KDJsd)X_YoF|hjp2CeeN5Jom9KH$`|tJMdLMR}IkT3i z+2y#{Z@o1EE`$l0{e=_*M1uP;y`dY#1T%QZAetDhfyMN;1*ZZEAPfKgTNjan^a>}_ zCl~lKQh+SGRJw;5bwKO#5c2rdieo_1Q|f`)QYbKX%l)FSttKFg0P_?%V+U4-imto# z%srHG9U&W=V#k0nQDKN1&#<|Z1G_fG?w7g!&XwG5sQx%b0}0D5P5}q$YB8LfI(yn; zwyeQn6DU02X=e9OtZ@pq-3Hi9s%Kpk)cbI&zjaE=Vk zE}*7~O;q=W*^Tz|qhhNo?A9*-+2|?ywSK1gP1BEq$ftH3Em>!ZE%eGaPSl z(rFC?wkyLCD>GHOG#=o5IO=eBZo&f007b4k-G4t}2S39p$d+g9;J2#FDf(46O`pg%dQ#c z9&iJB!20JjH(?_b=t(EK;CRYt$(nvAD zEO(lUIVNukf6Q$V9SC}Mlq&1tedwJCMgDkQqrAw0U4gJhenm(>uOJfl67v1BmL>$A z6#O>Dxl7vaktsUudQy-f&I!*8KkmO0R^a7JO8nZtepTEp&EXEfBOS8KODnv#dLD7R zy!45<;_Yk8di*uY2GJGHPT!sKB2+OH9MI@$3H&UE)Ht-v`LqCU|D=un z$yAo#b_^+#%kM$b9DsTMo1$>+&kAceoqQz0)6v&z`4wKso}K5H0UJ|wf!T^mIwSB3 zr-qYD-;kbGs1>onTCZwFom*1K5J+HJf(A1NYU+S7<4Y`P68yf`PmGt}&+};8F)U0h zvXP%XiL+O%mR=W~bME9<3R{_zB+D-e2L6^S6r(^Y!#Qzdh>eDOvXNj1fysKX%{9}$ zjZ6G@Zy9S%+3oV%gnHDc3(qh7(EA)LhfdMeQsg9s>eaV@g^ZXPj++I#%hQMgR;mBf zx#wdV29s50zp?Y(0%NTyyM=}c^{5-}iM$bMEmtjH6I~`jJ7JVqHK?)*CyJIOT8K%IH@{&W|5nCfg<> zXR+hZ{}F@rPcFqkIzlEDlO*b=vlZFyolpp}#lL`RavcBxVB}(o736xW@OGjl6c5fNT;*iGiSPGP0eKEZk=nU(epPRGqR1taIZq1yWCM+;jHm!pc{pK3zI!h@iod3Xx$67*U%0UR!7>cMJ&Eq|yej!ntWlnlXYlIj#K^5a z2b?y*!Z?wK*KaQZr5D#+C)-X@tdKHpR&<}{4!T9{ez`bt)R>vW1f@k`;5B%O1(IHO zh%)eF1@}Gb7oTBT!nKlQ-Vj~O9dRh6b9~3IrTG^vyU5Wwm9c^}K}+}$chs*X9OQOf z>*?ddTG*F4Wt)VNDZJ~aZ{DJp`I3iB+w#fQ8FoulIrOi8=FZc*C|}$E?>{HI*adIx z*m*f>;Jg%24A8UYpbqnW|NH*wqNKTnLfr6E%gdL18m-&Z>vc>A`m+xmKLZYrc5!}C z`4YU2B30)cHBbhSvZONP!J>GNr<;H;0g1)8r;2)uzxX8&-u^u6Z;e@RpO3TIF(^z> z^n!R0XFq)iR?2N5PsHsZeUV29u$P|md?K!bKz<{?jC;#9#TUy(Y&*3~c>ioMFyT>~ z9-6j%Un=Y8cRz9z=DNHwNb{BaLGh){wO&PABxu3bHe< z{`h_yt)Fxrego*w;@C|(cI>QV7}&zi6tj^ciBt@R(sZ;>JHr}0pt`wCQtO2YP&S%A zCJc@7p_vdee$3|QBXq?Njq8-3_fzfI-!MV*blh){TorUl<0SElTjT`UNAC!1gys8A zc}DR5s1{}NH!A-=h06Gm7NR9;d4^Ag&+g#ci>3n?W+LIKM=^F_a~uFgnaNPu%5QJ| zp>edtZukC7gmJKJeYZP?YGMVgt~yfV8eWrItN^r?qY5HB`JXcU>WE(nmY*v8EbVBjZ8HzbYkvd=ob7_F3ZxxUt93Wi;^Nf1vt=Lz$epYDH3n#b8gcZ90tn;_98 z&gG`FrN+t{2r~kl;i-kj$~3~}_+NQ!y!xED6oVLD7rkAwiFaChg;VDDL2##Vt8IBM<=j@BzTsFUBwKNT>{%K;I)&pm2^nB}^-lYc+*Q}G#cdr|x>t#loE zAl>b@dX`3c*nLD=#%*=o!ReFbaj+kJs*$@nq+7a2lJ0$mSUN+Q_@!G9W+uMC!pj-u z^}T;xic>d090VpGzw2`n0@EZP3!ZrNDGn}OM`8sJZ5uqCY=q2(#bnRg^4eF_MeTC^ zRHQuf!{a2=j;Eq3gQ@5c#S~IxKPE6?`oMM3HU3`lF~1zB5Ja&fWVs!2sG7UMB?mOQ zd+5E~@+E5dK9|aX3MW*kQ$b}suOHNQ3&!?}cFZ3Qt#HZ>s(@6v!a0|h+N9*&Tof43 zH-_id#{;W5Xr_PbtlTI)X13)@Ge{}B7@Zxjs9FpVc9LQ~qDVCrGbHPZxEHZU^2d&M zTa?A$`fc&IFntBd{Hu8>K1E;rfW-M~V9^a)h*0HoPVq#1--jpR#M=MUEuCgHeKhPD>>=C-X)IL6=C^AWgXZV zf^fo_#d!R?Ytmo*tsY1x|NY9Rr1rG|Ql|lsS}3N8BIl@>tKtW)_uOMe`-0NmoOq1I zh{F*%ZzjK)6M*+l$??x2TKAa`oMJ`XYN(&Cr1N0Ev7T=BsukCH#(N~csZkDbk3j5a zGuaV%-{XMzfZL|X8joR*G-r+S3g_nH4zfeBlKS1Ba=1fWjq*HG>C(v`5LeS>qWG~k z8+||bxIyoA$)(q_rH{8UtSm;@PfMi>ee|$e^}c5!Nw?!Ut-=5(MHEv&kvu9UPf#!I z;bic7-A~auieh&)SL;9Q*Wl40$aP91IgaTN;VF^UdK5*US&D^27s5XY?{K~4RvB2p z-Q-fo?R5v*K}jR;1RwQhHoH8r?qcfcPu>a`c-ogW4_M7;)Y0O;yuU6EDJ-oo;#P5X z0c!*N_@Tu0?2ppE&?OrP9dgd$-iClOmZi10o(R0l8s!3%%rj`JT^_N)i zz|T3f!!PwlB!z(khfnz%PYg>3%0bieCS|6=Y& zU)6;Wtpv+hk4IbqR**3Q5XgE_U5)CJpBLQ#tPFds5kT3%dI$llsQZ zG5fEFZVv&Ti6+-J$&UG-h1)F)$d`Zqqep*TyW$t=zS&4Vlcl)ph6F`GJA<3#Hd&Ru z#2T1skzUe^^>3$~`AgGOs!Z&dC)GriNsGKo-s^Zc;09g6#iZSVNLp6|QPx%$RGUySj3-*F0Vzr{z9jUjg=?>X_ z=V6#J*33#5p%P=ltVaQj63Z64tcGI3+JeQQ$=Cg&8|*w$yWIr*r`ViD#`zU&HXiKQ z|1pt`k*yezA^Tarpbxkj`e}_~Bx0{inM8HbWrNFLXcG7xO&l~isw{GbByrApHUUo= z{gCw2d7KN9BhoI;MUaSvz3|A#Aca%qV=GBNDkCc*qPyOlUoQhll^Z>*iFb z-R&3=CaB$?N!a8IRLHW4951+L0)KYiD%a`tR#-`VP9G}Zn`X!QoOjooik2Uht{^+? zIAm95@Z-;;7+`tXOU1-_q%Fx`Fyc_OME7}|pW%b&v4RYrGH!F^J%9DW8t>xB+#p1N zmj6VS?sY)iCA}}X=YC>gJ*YM!A%iUqh8bqTv^Qi!Jg}M-s73PweiEq1%Hf_vaU}N@ zOWUh6z)u;)lu+c5E~DmuZ^+g_K0`k*fdhVc$Vwi@duupJq6|T%qM9C*4D;|RR2?v9 zfX<}{G9!*gt@LjR2bvd1#wq3;3)l~GGQ*+Qfl+8Uv|2I1EnHSfXL3K`<1;}!E>15Bk-%w~#gq+(E>6L*bn$aKP-S&CFy zhWl-D2Nqq^J@>GfeXI!wGx~D+!<1ZVNmVl>jMP2hNmG)a{JJM|I7|k~?IJF9h-u(Dp<*Q!cs9T{x z;C)Pm3daxWOm8mUHNOkU-Yb?q4xQc^gwINuF3J|GszZ(e^+t(btUy&wpM$(8 z-7*LC%Q{FNXEQI8n@zR;FM-)py5WY*c&qXb-7554;tOhYO zl$_TI0iaqy4QDVkfx7O#+kG(flJEpMB~=wiR))j|*YfU1YNb~K;{#RqBR4u>-qm^+ z5Kn9N>XLyty#RX34SM3XQgAR})M1~?6%KymiTsz zB0#ayTSu%vCc=R7m{|%OFc8))kaWWK+w{~MKK)m-dH+uznuY^*@852t_;3E_>a)>$ z>*M#q`u*g?$=1$x9A>*Hbus}b;&Rf7}zgm zY5Y-#Yau)5Xh8T?Kc_uAT+sp*OY3H*)1v_uUZ^w;44L!5ym%FRBa8Qkj=A&`XXNA_ z_Q(b`th1bs(7Ly8Ez1pR4{7A$_u;opWjbVodkbuf#lH~KxL>*R=zL?&)Pn}2vxj2R zDY8pn;ttWyF;kDbAxWI!&<9W@c!q3=Qp-DCnz$+?CT((6%QLxaqd;>CQ!>y}TQy=a z$y>1ay-(7YH7}Wd?~^$dc58127bea(t~#;Xs_Z=m=uM%RO%zF@V(xosn7!OoS&Lf% zxDn&*9_Ry2J6CWj~+9X5HRh+$$F0HEN z)p$e1-NqYm<_^asCfnjyZO()2hQn^n$CKS**WS_N5`@RV17)6Pnv%g4e@+z16~XUWqO3@ldp6< z5r9JD+x`2=>F@U~`|ZQ;)GX_YNbzkB=?jLLbF90b&>;z?eCLxbvvs_x*d>;klzw&9 zTYt#*=!(GL%}U39@s`>9#m5358xp<=8-TK!kbRc#6n+Ae{3>2>$sF8IXk+>(aQ4GD z{=;~)@p(EKJ9dIiY(Q%10one5!thM8nO7XyCP|*F+VsZZ;1p5$63m=S4ZbM@6*Uli z1phW$fn1sK3$}!4=}uZJIW4WEKXgtJB>*p`?R>5&Vr9bUTKCX2R(|95i>m^?1$w`D z_U}pJYZC+p&cg}pA-gDM2O3fc0w;5>EbOAkKQ+o!Sq{BYp%X{zqFWa44!FIjM4C)4 zPG1=3Kr@r+f5UbjFZ`84GzR4kK62$>6v4a zb^44itvj&vuWL&&)3=NZUMX!#36u^Eh%4k>5xLIY{C)IQziNLx(kS=QM>vP*-JCvgjrXtzX2n<&ifoYf z0&i7!%Zj)wN{$bjiYH>n$txy2k^0DZ&@e*2_Ba8EK+AgQ6?K6HkUyEq>z%iv2^hS4 z=s16^ms)XD3iqrD0@LGW0X$51X2mCjZ5-vdKY#ZpuX;s1N>uf@4#~2_+4RSY zbLl-^uu0HO_t3`_Re?zyRI3N4B#GP#KN0Z6O_%nDB6un!jlu=5Zj(mUAu9{m?YlSZ zsNz7lE!Mhhp=@Ffc_AzrA8$2FDGWLhTse)~9u9ZQ3Ky(|^<*^nd z!n@w+dSjz6(yw;O*38ObK66c{6PbfRqq%WzKODyjioFwN0rN8ugIftwQw7lT*2)}s zyu8gD1xe>b;{f7s?I1~$Mf&YHCJVGJ6JVmIn4J{a4m>Wv;exebDZYb|H=Wl;XDfEQ zZ+1MqxEsKeMQ@Z9hChr%UrTp4UtpzSe%T9W4}+KWXr23}C#l6F4Qq|Hie8iYQxhsi z$V?Z{J3(rAd&OO_qm#rrF~42dFHQ{`az@IFY(=*e>)*Auk!qWCKw08u(mrf)@qe!k z%r+JUe|}5Hj+ekDgu?6TWKJV}hMWk%+!!q^+a5_1s3L9#1m^G#rX5X_$f%NySp0_#xEI?*kiU^ke7`+gQt>n2(6(mmrf~c@g|r+%r9~OtQg4X#@-3ans}H zpK5krBwJ=dY4609!Cs01A*kI{%whLYzka&(i`T@3%WlYumOX$%hKD4Dw1ylENT;>o zX_AXTUtKL3ksk4@r>}70g9f5Or5D6qto6!GzXb~i!*qvMy2EZT>^2-5H`@Jg#wfAd ztXLE9z3yJ*ctcjdEXDV8&T!Zj*{H)~$2KNDGFA3C41=$a=}`wvs(%8T4VO50J%{cA zAsSnRT#eE7TEKVap=y6m<1_*`E{YwmcCrmj;x>v&p~xmGra1CAaP~t5N0y?=tq@AZ zKxGKnv9sy&B^TzXH1c-g6N?3fj4@))IV&8Q)1j5xkmd9$<0>AzwI3G@#`YN zPR;#+F^S~!ffGAk1Dl|V?4~y`DZZWjkH6FUlWbab@U7p)QX9D#b!~^t>o`Go_({j< znT*DVk|kcoAG7(VX}lbYlQd3+zs-&(AQSv;$K-{=4Ra0$tF-(}a}I>72EbvNijA;z z@XW%BrJYaVx!ij3L9u>$BmK)4TgXN`ULGGXn0V4DW*0?vP%&+i69FjlvSoI=6YS$c zYCxkSsz{-2_*@bnjdxRL>ncPqgd=&*^dP{jF*lqtrZAG`KWf@OOtJu;`Nt>gwj0|F zyLC-WS%H8<+-ACq1CD4eq%96PXM{};1dIm797GsBGA*a<5YhA5Pkv_1MZjhmW5>RS z2}%WAwTwktvCyU=a7Z=?QFvse*fk$ntEXcHJP`-xz({XbCO&5K<1;6J@%^dn=-RP+ zVPZ#j8)v1UNSe&kLe`q9-!%!5cv3O@h&A4x2`j3PTW8)^f`B$;A1P-GnygXoCoAwE<(9u~AA(*=Wx zqiPl;CO$At$7!<;a*a3nUz3@{#P&ZHzo*krAouWaP`!JO9bn-X8IUow62#=ge1#*UPuV@#wst{QrGt+$ z$p=kamwH-L1pS%U!VU)?;BGj-t( z0F-Tj8M)4k_S*jFxkxSi$kK{Don3KaSXi+YNDhfTgiDt$t@K59k$k$20OcMqjKCoW zOtAH-VKepCN3s7I>&wkgVjln4viieHk;RZ|Q{T( zOrP2@=f9Sk)|v|6lQPoHZmn_R_(!~%-snS$xlfUMRLowoF|6GClAD2XddMX{I3Fy` zp!<5#^hG?CO64rc`=Zuw3luIw=O^f>0AqcV%%Dy>=+RAIa!U`Xgvw{U0yu}$5`^6>lCRy*qd!*fFeoGF; zKyWRSis^r;YxY&|Ec$(K^j<0=koY-UWaw`%P-0u2R)H}#LzI5I>aeT@8inzI?`oERfOLOWu^vz}s-D6$qhI7Nw~SOFG+441}F zes0D%kv3t%0dqdqknAw=v#eRVU=t>Wb(yS49p!Vc%kR-~q=x?hbU~N%E^>0A={1f? zWa2qlfjM+Z#3edCq*}E+bcgtqDBZmP+@EdU4`jzl6K{{~J{PZT;olBj!8<^gl8+Rr zb4nr#=ySqSZbA6D$b_g3p7#AB+cZJ88RoH`Iji%mCdhexos!b%vhSLZBl7v}6w>0v z*5rWeNg%e)}zCvN3q ze0H7ZBM8nNc%=_IGz!Ek-A{V9@=to=@}yUv7Pd$S2t@fk@hC+4_Po=-Ek-`vQs^B-%UHr#9 z)TKQf+$qQ3_g}u{P?^EeX3u66jdrsGkKOF02$Sq5S$5|A(OM6&srAcYDD_ zP6jA_C2@AprCv3zjlnjp7519|X9dQ$rp)<7Vscf4_k_PADNgLFl$xy!*%Si>j~P_V zIZ`cd3h9q5=Wh4gFFpj|P3C6v?tc4E!Bb6WVTrnRZ@ z_Zc-%Sc&q1IW5GpX*716N9PA3`|uN1Eo%7R80Du8}8Rd#-G^ku2-KDRnUj%9X?60 z-%a+p3n8O=dJXs{cfE_bI=2>nrK}0gHTXVp$q0-i_h_9cm)pX}^M#!Lm$5StM=)%I z;Qa(ZGTJoS4>+TZKK^6ufb-4At&J0DHg@84la*#;ls(LlybEa^O||$oWQ$Py7!pg@ z$Tkq8JWDc_C6t|^^32I(g`;PVWZ#nokNDrR+#x%yJ7b0WCQoI^^z=bS5T`vs3AM8- zLzVUaLXEZU6K&s#9BVXI24iYSPu`!OccvVywI3iA!iyXOA5z_PtvBi{bk(-{g%5NIwE%m81B+!}<`jp*%W(V9LG%hoa5B8ObAk=q^@A?8*uIqNkrAJa-$ zG&SgPK!t3?LoUnI@6EQ0J+U1>BR!C3za9%4wwz~$pDFJ*`oAYMc`BN$5oto9MD*#SJi_g~h6dbVCxNS{y~j(}LKR)ylr zK+@JK#XWejXQJ1YfIe;=fzC0E?C$`v)GW`~kmIl?D&fKnD0kZeRbeQTFy!*thfwCB+JAg|)(6ko`LXSG{vxK1?I4 zS0m#X9B2SVcI+4^=C0sqA)$Un4h2heR$x7^n%-tN0&}}#hb>5~T+a$S)Y@xbre>II zQNV9?Z<5XI)+#5?1^}=|AZ-W5z;0SN3!&{y5!ZMwZMoCJQJ1yB`#QIo9mIS1jTpQ z8Kr?*^J;mczveh|fm<&-7@)bOY~>Wvg&`;fj|USOESv>*r-*Y*rB&_m?t*stHbsv6 z0e@6>#yo>&H}^fDBW(e1$k;QTCQtA*9y4CN%4S)@aUyGUU&f}Rw&9KyvZjP*pL%_Y z30bP#eSap8UDyraw4|9();n^kN~D-I6iJ|Ba+TMFH{_SX?tw&j0sZ#dZ#&)yjWThz zpl6iZ9DZoem;URQFQ(Qp7VTo#CQXi8q&9_?vp49v_n^mBurIY?2VOEZC4I)EfaU^& ze~C<&$_NKi*J|go3)0cIUCj`xg>1x zg>?wp6SYx$LQDNl&O}yPO)Z_o!PD)Korc&y3S#%VCoj~w<%PuijAMJuW~wHF2{xa@ z_FC(h7^P=w>XK`Jz5S|bjfvm*?H`i57iK#IwY(#?L#HSP)Supk;$99m0o|p`I7OTR z*i`>!_gAZ;SH4pH<*HX|U;RxMy+YaMnd-MnS>)X&H8zvP5-g=UC9UM-F@tig;?~Pe zy!E0Lp*=t-c8L3Mej0C;vQ3S|4CZ5QYp#g+X+Hxc| zc7l41yw7eKa^eiqJ+o!#I>lV2$R#SKRhr|DDI?@!cs&2Xd_$f9SEfu*u0(-TBpZwq z>;Q_W1i@`c1Zd7IDh=-r-xdTp0KgfhB>UzY9p+TOhaRa5^XT@l9@s$RjjP-=1LTAU zrVk3~oq}$9U(`BY^2{FL-!-kk#crf79CE>djAzzI6b0Z100E7>nUuiZy*8{slmPVq z#%q^{oEPFg|0bsdF^qW2Z zn-l{<*A6uJ*hIL^yBn;#9(vZVEi4scXch9pAmTo2g%mp42M97(SHs?<0%pPBVB+@~ zFYT0LK3J<5bie0b;`30pA*`9B=_gNQ*!tW7^a)ua&3S@XV>vD!G~9koT1XeD>d25w ztMp*dNilx|c#7-pig_t`v7=+%gMGrqJwd!u+F&g)2Q-Y)lj-fQ=aOV#N`gBa(RRS1D#Xlb0_ zhB#5(LzhsKTcgRMU(L2A)~A3s&I9c5y5$Ga2Q?-ObUbnnkMvF>DQ2WBaZ{NgihP1( zgOBJBRM)}=xrb%FbPYcpJU+YO4rNYy7q-VudMt^&G!?R9R9+sXdPyjmqp|%^=fRRK1#O)9}WO556Cs| zVjdEZ;hu()xd&xOX*{}%$)~RY^|n@F{M%Me14-_{y#evl@LUdkEo|kqjKCVcp0tGZ z0*gzW0RL?e23&lN;47!)XJIwh&g*mZyyW$Dg$ab}<3Ft-IqbG?&igMTb!Mx1ImJLD z-cBm!hPa436kz-=}nQU>TxcBq%qI$AY-N5an@!EJ; z_i>3%UAS^V=YpXC5O0jAlzkYK=YM?3s;`{=f+IXDyDZD&G-c1aP32ki8@>rMOU`XN zMAopwjPtJ4K(;tC@|;02n<$bBk!SUB25a|F>>0|ARhSGn&=|E#kcJAOp_`351CD^f z=u^QG9HS57fTN|K|M2^kL1Z=@&`t~vE8*oSVF;#`e95`+rSdi6@bl&F%3;|zpr<|AuRz2ugxfw!CJcu!<_u7R`B^Ez~1 z7I9KsOTA7I?YuoR9b#H+0sdHmn^gL|_AJz^uZnZHv4U1Xy#u=fo|>M|4%h;q z&#s=lR|ovx`M?qqR-R?MBwe{jq-l9#l03p-@#nxofBf=oP0aja~SY*4a9Y3ixdkU~$D5anEGD zWCz%_M|WmD?;dD^&%Nmhyer}o=6q7=c zL@K7kV}EFzpvtq{2WtJXux%$jAUVX{9To30>AsV#NSWx&v%TLsQVRaQc`(9+itp~* znN2E2LxmGpX1AK5;t<8uQ$$C_U^8KwJcDx=>aefPPvR81eJsHZZ@p{<1EytHcs{*4 zQgcJz$V=h%2(cnktJ*2p&UDPL2-mq8RX>oFyB9LtTE%;c46oy47gR!SoQ~y`yTXBt zOs2sb3PV0|-vtHz+d28(4IFGbnXu1pD~g_NT}C_07B}YCTSw6^9%^DOUO?q#M|x_BMhE`v6d|rf%WHR~z$ZKK46X(K7Z(_PwICxH9@h)J8LUFT)CHn(cno zi&CJ*Zo=RCaXft+r+yqGPwW=!hQcp!QNAT8oObJLMI7%WNr56J46&z*suxv6B!%7e zy33#J9uLC|*#>}#P3Cy+whn;N$}5k3@9Hm9CLiQj%s)1gJughIuEoqdX`~n^9jc*X zu8MoUdT2|jxGd6V12t>MQ5S8guB&Cw>#1KFm?hopsn02p}N zxXYr9_uIiCwheNlIUmnDHXe|f-*fQKXEu15K-B)yxwl9vI}kas3$n)yM0pegJ0UF< zg8{f=-^<>Rw#icgRd@f()r->P)!?LN(fdSsAv=SNMMcmR#O)<;)dveog*vx<&$SGC zDt5qv`3#u7n*dDA#jiTa5T)QN@WhY6_LR|P*S zbJ`S*Ve2E})PwFxoUL9>FKGrOgC38ScmkLd4$R%ou^RF4{BBj<9K((8f_QC# zUp~Ekej(jqI*~Sg6}4(nw<*rqtasUAQJqewl-Ir`T;3g|W9yqBRREA_lqVy;67>`bT1C4br5xN}C8WH#hp9 z7q&`Mczxnt@pjIjdp6uKRRqLzg*=3_|pt*tDt0$q-U{i zn%lgMVc9%fNee^Tz`5#B#KIr&G-B<7x6drkcwaq*NLF~8@}H0Y@rJ(%Z~bpbSCZ|c zMTDKWQUwCQBRF36Pz-1Ymw?v<)T+xvPX#5<0%}!woyU|08Q(vkkEzb6--jZ}CbBo+ zxbRfag~caT1>BVj^y&{_hqgQVy8L|f16hV7SFBgp!I4rpqJtyHRGkZ+#=%n)!1jQB zfW-Dw$G+Y=kg}6p7F!lzzaRl%D|O`P1!4TB6KD{^$~hzkb5T(KNNlI!jL906nKZ1x#fC{zDA8b(e^Pu zV=P`{{~1qRk@n;2?B~5YCMm38u~CQ~yq_wcOJQ0DWGB+(dxEahKpchnB&1Ho9O9(q z6WGu9Q&Z!R1J)OtQM}|S!LMI^&5}mS3R}~jz!ZWkDVlOlJJU<*<}7Ut$`V}^UK9e! zBrplCmap}g%rqXm32?}>9(7Pw6X3kz)9k;!vgtRbh3K8+RWFg7)5uD*X7~Y$c|eg~ zDh8Qxpd3J29=%L(m!IgRxfnbM(UD7TrQyY*oq~RXwMf{E2h0~49xbkgbP7*XEXwd0 z^uU)xE`uI$J8i5U&?@?5>zFdm9vBQYVqnK(WY@SDylE~{+G?)K?tZRan!0e6XTDOS zi_Q`~ku?St2Htnqq%XlYTHxq_eF^GA)bh_PGLkN#4*53kPsga`K2X(5RtZvO)K3R8 z=0eyxY5L?VK!yaW9|H5}ZrOU$gMTlogk!nGz0ebU<4A)ZFGdTz^Lh#|edoJBu~h40 z;}tvazYIhnM{cq{)%Hu1Ls?%OkV_2$&zQ#eMR?b%WUk(QyI-*=UzJ7ofR%0K7}}g4 zxNY(8R&~>fj3$*^NM|e2JJJ-Xw0?<7Tpp7+8D5%WUU+oKrAnl^!1O_SqYyaA?d58- zGreQ@GTx+tcDB3Gyw;2e?HEC4+% zcLVWVk|;&c39f7weO}l@gP)|Rh(NGt`d=!f4*+M+nqS!Dz7_CpJT?%og%Nig`$p`&3MyJdsHPbAxim?UEk(c?qnd zxHu2%9;bV>&_!a+&Y(`9&#H*1gziWKQ*;jLg6S+4ARQtGExT1l8k00$4uC~E5RSKHE5Vts4ZF!fKfY7G)%DT*4$x3en?B+m3@u)cOmW4+XASFNl5J;@#RF$rlc_jPDkA!!Q29K^UaOduFZlQoVP!x&9kz~$(KdG8i{ALEWofC? zrp&Aqi&iZ<39Z*lcSrSvBd-|<0Csq+hCq&k0WivZ*@%fz?sV)SR+yN&>#Kj9JIl0e zePib@e@(VJaZ2ofnZr{;G0?kTK*gLV1MYftbpTLoa*OA7@VodQy4S;cFv#62KB1_W zAR(Jx-7d)%b;^gYdKg&`+su!Iee*X?FQ+pkA4Pyfy8j)q!*FCHTr92rr|%sF8dh+j z^1uAs2%ZTpAzQw4l&pMVyy-kMaHUhsMvA0BD4LVRITd^q!p~hYt)gxYP)XO*Pk8u3 ztGF6a=|7qAEf(#W031g_X0&lz&;C=THUXncIV*!4oJP)@J<;P7bCe>7shA{D3x6NC zEcE2eQhDi2V;>3bCvab(xhXB;)T<7AqJnS>f507^Oom*H?%E-)R$;XNnoQ-ju*>4L z^6e_q!>2Q;LRMC=+!61ecYTm9eixsys;2#C)x~L<84yh zK@VT`h_e&4OOyRBaC_bBgA;)4vrU>N&lO{v)hTk%ZW4ob2m6>Xw(SDzVE@9OY=Bk^YY+lLf$pCb3Dn2qvfP`!zZZl%XL$c0uZ1^ zZ}Y~YPzV8ERpM?Js?1?BFfLLA+9iYT8l77#1niE&LeWKc1>y-q))H%7Qg|sL(Bc45 z?;HXlkP-0^ybJ@=J@A=0yjGE^fWPIlw>1F=Ueas~>zLmu*UjDp=CF-}*I}jV6PI?$ z*_mZCAatH4M|Jj^03fO^;$-uT8>vh|A>AbI2g&44(lZ~@pD#w7+IcP4RqpnC9wy}d z;u3k2taoC68mf6lSg9O}0XCIPDyG4AOK>@7jRYu+Q2Idgv8-4$=+RB5DKQxW;KgiT zMMRw7KsbbB^1aKwYoHLdRr>6}oo(OgXJ0jLOZ4nvCw$hrB^qTPKJ{|(oUc9a;&E34 zPVk?|>iNy|o`8?Utut@JT85%s7lBMOi_Q|2KrLa5aId&*Zr${%$RtjSw4QfBRm;y7 zEr;XQGsSrqmMGaf0VtH-zD}t(mI_O7aH}^Q{tC?t7q@LuE*<=io%5^*CR{*pT%S^yV#5q=ce8_b5ssfOd|!%@iB3MU9%D2 z$C=e~V2KAu9U?TrN5b_o5WB_*;A-A=*;X&?u37I@>AT)btGEgB3;EoXxtb1seDF3u zW0`uswWy6YS$QKr>Gg2dhF7nG>XpV@Nz?(eDA*=kSRkmo3b0?h5x?{E37M zZT&h}4-q`$%mltL@6^B5=xf5!W9Im3q8%-h>BO7w8Z)$%Qp`>YIOAh_d{Q{ok=s@I zqLR5IU$?2xsMF-9d|LP|!uygO(VsF%oL`QpRe?39wREMp4H|KpAzh6(J?6EkZ~ip* z>s27kF_Fo)K0kJVh4sNFaK^foOscB;{qye6HUO>72LWeEsn;^s*05SmulNcNjf2kZ zo_pP#-oQ#a+qI0NSED2p3X%?ZRmiGBv4FTia(5i^vg{^e+?ZhZF|4n)j)_tBJ(SG< zlRpKU49lLL@;Fk+Zh>-QlXKY2x$>FD8BS}siRRIq95m6N_#bi=sGZlkHCtb8&fx@Ah z4}%QJ3|LTFAg{CnLMP~5;Jf&_1H5m(0tQ;ZLM98>qtLtfxCFj>?`mO}~xs_byz6F+w&WS%swo>;Ko7om9!rSe&H!UD{@ zqV#1iFfy%z@;u}JcZrj{uw<`TfC8rL2^MjdLz$>9`t3L>nLFgNiL*>_F#7FS!<+iZ z9{Qhezk|p7WSPpPONU%4BJw?ALyJKx^E92!yDRU6phOBUR)B1PSh2cvDVD0@e<+Gs z%SFQpZaCEE_0UU~_VSk&h~dX)iQ92v^XVnaPDh3Tu6xUOsb%-C)e64R{3WMl-6mH0 zg0tvHP)qrq@J`@Lcj&BpxlMhWcf_+#mi}_l%UfSQqejO$*>43$=Y|{xTj3}&8mGxI zXpC1hi$J$hTIhC1s#lbICxz_|Xr}SF1MHi9Kr=km_WRxO_l<`wYxJy?!VMST-d_Oa zw}muT)ndBF0S{}ef1Yg*bF5M9f1cIq;Jn6t#|J+9p7%V!0~wg_msbh0MHr;pD?_S# zEcX+jr+5S8NEMW6YZ^GWtVFerNt^4jb zE&Di~HnC`>!4!-6?m@H1Cz5^zY^R?r*aB31*XQe@h2xI2CKv>-qi=9yqo6S6ILVoV z)j4$wR|41Lfmth+XnH#3n4*5)`^G#dCe)KEpVLCPwl-)D1C>Ci1)nBgsca0^DvHAQ za0Vm>1^2_jSZj!%jBkg_661sO4Ri%szfvL83*l*$AH@_K^p8Rx7L-mv(L^7&mrfEr zB)8^4l45$HFEq)d$?;sD{F1Q2w~N+b#TWjutS!FxYO8*p580>scA3sfA6O z@+QJK>b$-ivBEgoD($9g`6w_5U;H$AEge4t3O|`r*aJhRK2?Q`Ufb20zKH5Y##+B3 z@sLa1!X#dWa1dH59qZ)V4yvaXxrw1>jJtm@<9FLAle_cS&=bEa918%7MJAwNqla#Gmgs%oR&MuvWGmk< zZg)+DD65fv4;VZ!3$$9=1YucR#gO9F5IfCOp7OZ^10pX~vR@NVb4;aGJ>uwz&aIVS z;j2~D2KACe2IFs<*w9*Gy{uM<3tAoUlK%^@LR0q~SgW(Fg?d5i4j^MR!Y<`f3^cuL zqhf|c8KJs4#-DoiO>qU?s<_{|J8{sGEN*EyevSUTw_D1ADq@S*aWcPE{ zN_N7`%+yg#4MnP{n0m=={#gQC=l&1r!=hfGyu^x?$MQIjQrxhnNQ{0{Pu@5o?Y2v8s5Q9N6-FFx26eyz$CEayo z_4EO#(N0wrxYo_l;2VTyt8`GF!T(s+!7qib*%Hv;IS3j&z|1Z_7=!~NUn<7=nml`@ z#=jNs)O1NuuNRS|) zS$~zGsR?E5#=8w4@1E&dps+Y+juBhmvBT1TeS3!JUyCa-*Q1JzRxn|$xG>X|k zkz`;v5EM-BnBGHU!E{+r$=p^TC&*Lv$so%0zfkIL^8+2>gs14shb6v=z>= z?)j}!o!bs&pBytp7^~?DA9BH;{iMRXKvglxJ#s8?WIoXjei!%^m2i>Cu(&@snnw0L zXVIB6X2zzOVn9l%0r%U>qcA_+|MF^iM^tk}f5Z;@{g4yYE^^KxN5wofzv%d%P-esmFG#)E+2Fs2tE~ZI{Y4VgzS!bFB0R( zd%SbzTvZlAQ>&4!H(!wJVH{~<7KjC2teP4tCnkg_>vktQ`uiH82}!Fnul|Z8vh!S= znArumqei40w^2+6MK&2V2Xx;$|7zlrqp#lodfsbqUH!(T=r`}Y@rM<>GK-u#o^`;ih8P=xn{zauFx3{XnVfRarykkrngVqyht;&Rz3 zNIIgx5fF@VI!G@s-F*j*qQ*~L4s(uzvQ`g`LME|59bXfY72sfS7-wlb5E*lR?QlVJ zSUUNE2_pY?f_ja-&kiC^yi2}k29fI&bCn{OsF+siMN$|7WG$`IcFB5=T0foJ?ZxSC z*|UvxxtN&5*skWXctCPE7z9d;XR~>KF989xPc<_JNcAE(QY^|_oVU31rOuZiPr1z- z_1y|*4H%T&u!0jnm>{?YxXtx#*0jWI6P(u8KmuFuMozoZiX6q zSQ~@h!PB6(h6z%m!WwYYLlm%8S`o3{J1b)GHtbDMqfjQ73A&hcxbEUnQd-6|=g+o8I;fs?_>_c`QKPalh{gGjwzeM0dr>Ei3v?+H_9T?&i$ z0bO6idB-m!IFU9X#N0Ko!~Bmu1S^D4@(Ts=-!M6~uN><84|0s1Q|rV56d)oTu|d8- zF;KS91{r}lm&FyHM?foDug2nnYOf~l$^}PMy{bZ$UJdT23dOImum#Gc=ziHgSra$k z?-=Ah_eNqPKzZb4aS2zi&U`u9XOpObQ@!XAeLn)^mpl^$di7q}QMyDD=bFmR=JhU4 zmu&aEE$Lm{1kP~6wEGdOBUep3O_%yX$^iOflL5<*Mdw9b{98cW(#2oLtK+or_C}U@ z<7cu!@i?Gp4op_>#)jbV6`IQJk(NY+PV1*wp%z3M9Q0C73y*=iRy(s@0)(on7UeMkP zoBd!gJc0K4&B`K8nf5#Gzd!G5R;s9aQJVbrf)6364^6D=BZ>mrC0Vf3*|;P_5*q~v zv7M*`*qgFN8sufRx=ppu3=EGlK7+>z^o1E6HSW$OH`7*6vq7wPQ z&Um#*>Uo7Bdi7n`;Le*q$quS1bC7jSF)Y70^34mm@d2HgPjO)4j0$u!(!0e+oKV`@m^U zm{u|Gw#N>r1=p*KIIa8vvSU(8BgciCi9y1qyRGBqFaHyI$&xS73N0V{flVaEwOy#e zGWKpmgE17)JO1nV^uN*82mSP&jy+=iL5~vPbl%FJ!0#tjLWUs^aAc|wb`Sj?l;5y}p%6~NZnSv-KLqDt{*-l(2S!)&+ zDWe#mhbu;f@O$!o^K!UH-0#wjynF8Po)@?ooV!5!c9~NSM?T~?i*~3^LD=m=aIUyP zv72|4ZWfnFj?uY`u-`RbwS%sg>=k`1J>^qK>v%wO%O-^J^b}Y-VylI-bU;SFw>I`f zjnYSvY|Cy{~7k6`B6A54%8F$-CF;S#Z>F>S zb7o%^4RSAq9S&}hLbxitIPA)Vv?6W)lZUTl@kobnnUvq`zMoFpt1&;hV}+#xuhV)U zz>+c|1iyh|k}0x|ioq@Gu*Q37c)x)<2f}3K-jxm`Kd=Tdo6%}*00$jsh1RJJ)4d-( z?;Kxe7L|tMLV!i|v4RA_P3dj9rckt-tLL0qgp~Q0fc+p&P&5fuBExfPGdPUC#qezo z|IG>ZTj4KA(CYzd~s0vk+4gW`_1GfI>lkm&;pcoLB;wEsip)-5o+lcSsAB0T&qg> zqLKd=WfD<54pVzcyjpH6A45PD&~uD(afbidZ$CYL)QkYoXD{srrFp+xvFNwM^#i*O z&mKz3Y2tU%s2Qeb48kJ?${spauoY@OR}l@K&E`Q$o!ccb zwsYSDc2^!&klv9N3y|!@Q0&<*$2R2E@(&@m_K3bquj3kS#S&4iLhErZs9UyOq6@pn zgWPNf-8UadWcj+VZkg7t2gq@N!uTY?w=I$zZpl$6XAZd_1z=x$Y^#l0<-AYD5FqQ4_yuA>_(vzO)j@jwu{>nUZ6xe5(87{ zMjv~ZYmE9l#(w6X8SzhFjXIKj4_bfbqr~r>_Zt_{wSR?W=Fe$&&sMU0ST&s?c^AA| zjG>{gi^>bIE+L$^c@Gmlb3_uKazPX;%hl1Rj`eUUoKXsp0HH8kPYu~$tl zP5j1h|B%$NbA+9^B=VA(BYcWt-lNF7&pyzr{HfV)@R43;&A zj`z&vw(##q=w^?+SFdiS?}k1kmEt4r`=DxWFshaw?Yi--dE+vkpIT8HUyOpUu@3z{ z_777m>mZ-@Q=Aw#R%#=cfANYqn{!&&BJH8$r_~21C|6F?Dpmq(piy2K>n5>D3{?`7 zr7NVU>hPY>SazGuDdAw=Xg7C#o-#@;GZaIqm-&=p^qc}6pn`z7LmwoiS7b6@1UVhl4YckZwnLHH=x@eF@(x@+_ znFNfEKo<+Rj=0P5FImBL%D!?kd#=fSaS^HZk^C3NeK}<2zU-wK(74=9#k9z)_YZ@q!1kZJt`iM-d;-uc!N8?)6WYhG&oX9`QaSy+w9#N+OVr zA3KKE!rfX$66cOIO|DmzdqatgVLW?dkOj(~?UC3Scr*U;E4Q~zxViAX_yy$t3xk^t zW&z0+Q<-H`$rBW>LY9;?d8S`4kdUIjR4JVmq*V;KxA2d7-}T;1lA=zjFpsWRAsC6-0Dl`{(TUP2xOpLC!GzjWYB2TgOz)TNRA2PR_e5xhKR|*z%0O@63u7tm6$r zNU?}>gzV;K15b+~D5=Qyz7P5kdUXRYpS~EpA)-H#VY$%SD2Y-LI_=dVaGI?e;lna;5NCzSiom+80rz~Yr zwpg$3ly`zz?FZVsqJX{&#BqM5L^o*o(FzWpxOe$}n22RkX+|z&lQHEez=g zW~c_=!;-7Yvop(PWXcN{+1m^~%X+XfQO_F4&gXvjH=o|;-Qn35SpdvkC{G3z59df1 z@F8yu>zLmyJLNOU?vCx%myVCnm;;5K!xQ4mE3>5TaoU)k6*8VedW%2mPvvqGfE1gP zEX7{IyRH=g6r&X4`UpMqQTRvUrG9m!JbF7_A+*A4V*Vs2{En3=81`2t{HbFPzzQN$ z=L)@kXStVlT8LQLdwT6~1uKM59{=&qT1%ohr$vYr>Nn_1o>AqK7?~M`N{M;!F5C4c>6HJi zovDZ}q4q?71ckR))^Lra$>W1#qp-KR7feltq{HuKsG~Kk#sQQ`F(|BnGR6N#!e9Ms z=?`HM%m=C}DO6|I(v`k_@=7SgL&k%lp!Xv+*W|`c)PY$m4H?yYbV)=;U~TjhmxIyE z9o}0uZcN$1z&vY+U%skrnr7NetGnLPkY%GW+&XWblw_HOTGmrc5=GWhF;Gbdxt?pP z0g?-bB{r&hQBBB=d6Toym_KiuX5RdkNqJ}8nuPoy{m&sLO#Jno&Kk1AiMxbA(Pc#C z_I`@lOOZWP%q}rfsO#L*xgW^y18sg6ziv*gum=pnp=b=AJdt6aG6qcQB~N7U2}>YD zeafeWuT>y-x^8xx;rNw#2RL!AY4R)cvgvY=BubxyS7d?Z804N&+Z9bV!DM)u8RNG- z@nan(qog3HzceVmU|HY&f>5bjQG5oeMAd0%rR|~b(JTElw|(kD+9j97daj|(HVcjw zhNSvs(O1N8+Z8X*m}AyZdd67xJQexxKTdqzWLJ*-^5Tc2={Z|#I?QZM8^yFz?R9I6St;*7w*KY?rj)Qld6N>0q8 zYTlQ;FX>udm7m~pg<6n zr-}s*jmb<5LSsW=yjT3ThrT+IR#qpD*js639TFvT*Kp#erO(WV0Nu)IO+N33RD~4M zmw__wDloKkzO*gm_~Pb>ZrPh_rv0lgPq)6nKl&xp^7Qbv*MCkL+2s|TxBh|vOZEsS z`836xqR4wxOigg70)-84aATu1=nwZrR4>vhFuton#=(A4E36CNFGjT~P^-hn=u*E< z`KjPdk(%`pPh>mANl^_P6h?&R2Rwi^o{j#Ogk2sVkSw~)&(Jr4Vu_kg`6Idr3Fz>Y zRs}E0-OAmDTw0}9#P5Ts=7q&2l=1&<92}ujj{1nUELcWsFfp5B+afjcDqC$(YuA@< zKJNkDR*1Yc$*#+;g+XK`P2MUkP-;~WPf3|EOy>ldfL6#pH~J5H#0oxe@8V}WqBPF( znU8PIEl08cd+V;sC|jnf?O(VZW66>Ig19eM7`4%pO_QfnA4T^;U2SXhF1IVh(Zy(%Nzv>FLrqjO2w3$q%XG+T}8DMm+;15`{Gz+$~u zrLWE{hiu?sgIe7}R7ptT-Qb?%R6`X@MMMYRAg|j2JFNrZnqraB|7}wYx!|IO3kMb` zB>Oe-G)1abX}+pz(ZQhjnMQWg4jOylj<|M7vgiWkkc;sHCJs28AvAIU8T<8wf%c1n zfQOcr<>_i}Cw5z`*bh_ri>#4~-7Q#HW+pbGLc&*$q44tF13W z`;Vw1En(rb8AK}(z%qnx6$Ep+7NvHhe%uk?rN4Z*|ME<{Y2%hb+2140y!3C-@nG2bbtX0}-EmE<=zo?Ub~Hw?Fi#B42P6 zuE-Fqa=;GM=5Ob-&GCf4jg|wjni}VgI~*347tAnWCg8WaH_7JFBD&7|_k-~6$c^z1 ziUIDb?Nm$?i0MKptty4HUwkk0!s4vJopfv|vc<82T49?SueEb_=qdc9QH1>ZcfZ&C&rg5*^FRJZypm#;QzZ7;C^nt79Lt{X zt>Zf5!B1auGePCAnzMf-tEQ0xv!}nAV&Gr3fr`02uNumNj?M1ipX0<*SHuHEug+6d z%6jQiVN393$!bB|nUW zG+;eh-1H|dt->d=9y#FeKtV(I@TR_@w;V{Jq)3Hjg%uITqQW7UL60nY z&?8@^XRyBp1u&440IL#FLGl`iOCopukV`S?5pMOu>$D1NW<^y5+<~l+YCS+eJ)lok z7PN9freCa}ZU$y>|JDO#^F!fT0Yff^J&OC)$gF_4z^>WDe?T9U)AvdI+bvR92+p=$`B$!av)^n0pGIr1%+?*XAYBEwczj=EPDPkAT zbzb+4G@H5c)f5A}stPKm>eW@RpZ=QhXPbIY;NF1RA3plQ8IV@JGi!AurjBaq2hc?r zPc?pR!_O0_cgi77hhPC1Z-xu=1g`ec6pKb;6$Ao1zNllX!aB&Klx zn7G|?PyKXm(21Q6D*=G?Ir}B?)7HvYPODY)(wlwqAdb6(KESyw|45h>xPf%bVg<{& z?+YJG9j5y+%ADJRna`{Tyc%T;M;x+_84vd_{`*8omYo8#{w)+mAMW`FYnDLN6nOGf3wx-V1I9hJlKqe`wGnn}Z4v5RAD- zZhPdBgQ~`$Rw?db@NlObyIEm(Y}hOd`uOc(T184o8utK5E++_jU$!>_VB+ip0y`rB z|1LJwYg2yz=9Z-SCf_8-FE5>xIvsXg#y9{`Q8Civ10 zcwz-0O7!j@_E;tsoi=-9CF5z(nK@2&g|GF{H27`_F6XR)h_In;XVwa-rb*cZEz8>` zv9^gNH0&_xwhh4w8dJ`kx)oq4LBz&5IdN&66~Q80t_&2*(1QqKpF=LjO3BICDqBvV zjS#ZD&EfA0JL;zXmubPdwJ3Qmx#h$QPJ&q)>tl-Pr$`?a(@$1_I1F|nchSuxNwmis zi)SCtFNY9{q)5CpPLM(ly!5H01om?1+&W0?ETrzb$14(Vm1salLatS0DzAlA1XTqM zMx}Dm0ecT{agcV_8s?l$i zAn5@8iKT_J3gMG_`Rxtb<%tF(VfmMBw^B)T?xo9d)WQ}7*qbgMK^g>EVy9CE>*I!F&*_r?Ytt)k#N zEz#A!LoRrQBft1Ie!(sLgHPW&?2FE7jc{HQe_vHyHIcSf=S`LpD{ZakqZ1+VZt#WB z6df|WP4{S_H^Q2hkNvD@XVT>Tk(rVC+(ICw)D+UI<(Uc%E_=qRaC`X0R**R{yjvO6 z4u{VMcB$Vm8I?2hE=(ckop?b5g{2WYs9O|sgCf_d7;Im@!tdg*b&L1ukLZfhx#`ur zqmIds`W#U;krkX|KfUknD7_jhcUqad@}#gL)dgm+CrSVey02q~7HIv-f;xlh>FYAR zdU&%I5G+wqfR&{MB|JfoiI+}1PKISs0MsMJ8gtWw~s zOZB@hO%T)(EUeJ0TZBh^`v}TU1Cx*>oqJ_orz{&7Pw}dH8gt{j;hx=JxGude)A~IN zd=o>@r&e=(FP?!iR@RcL(toq(dCxYjRAxiH`bLHHIuHOjCj_lX!~0BBs_rE<3pCoO zJ)s@2k2lmy9PwQttp~mZ%v6*uItF0^JPnt2${&O`@m5QBc$}s+AdAF-pyNWQg5~ZC z2f}Amt3(E?UE#RbzwWIAaRS?{-vn?tDmZO8XcK;gKbd~YQiJgY2?knGXFNGGhZ{Sy zkE_Yyu8#az((7*91lsg8SzKfTAXtoMze|7k{;eF#sye4#ldMqk8=1Z87Jf>^(Jn!q zXjJ<}dZ@FWSG_I-oU><*`<-XE3`w}(d1y(VTW)5*b14R5@Y|>u zbQI&~Le9p!()TP`8P)|M%`8#7#K?}4DS$1a?Bq-3oFlF+u6_Ouz6VuF9JJ2G0>^(H z*-gf$^J2fJZe=8Pli{?DCl8N5s$opdiuEB zuglkPp2)5Sbj?P>D14U+k^D|b93;z=7<;nQA%9Cfwy8jfzZ=7Pr!S2R&Z zIzk8Hew=_eA>RsG?Qpp2wGA9gq&V$bV`c7lDs^tH(rnQ4>6E9>+~%3-rs?JH@`YbJ zJT7t)ne6#_D%|hD0Sw=E%5}33yaaD_Zm9QK5^(}5Br+XXvpKQ3w@2Z^=s;m*?4~4N zR+n>4HpL~-C62^DXD#))W|kw3Vm44DnTk2;qYDGoGuKpAPf&H_zD2krc`VKn^?Kds z_VOKQ5=NaAo3S$L-HtnG9V<1*z7rc}!o}*e-ZHYwiKG0-%`l*&m;)5#Ib+heEkOyq z9Bw`E3Ig#qU7)(*K0pS!$H`5xE-V*{q_+6iL$m)bvP1bX>5Do-uJE<w$S_3*l$*b7rOW93u-X^afbi*3Njg$kgR{MCk0MyRGQ3;N)^RaP^1j@U!g7hKUGB@mVG?$jPxT&-I%>? zQ{Q|o=|8%o|CIb%>DP9@mi4`==rd{%efGrbbZ*A`I`~(=aRr3Jmxtc^!4>godVM68 zZbFWT-_6G{?Fla%fis2)vA@I03OJO@U#_ND68|`@HK{TK&Zi_l8InGEmIyR5XKQu_ zl}Jpia)nTLbi_A_X9%7)Lrk`VdsK$)WLd8xd z<*8v^qUs6a{ z$hJu|)#BUoPN?ne<6;!PTvoSm09Z&6L-w>R7V|VQ^jHjL$G6teGnM@A-1A<^T~7~4 zpfwdJ6S6|n<1CLklyw-4=DkL2KKmm} zevPM%h!Z!RS&TJM`*P{G*)OnFDS&78Y=i0z*w4DdfHn#c&%%Rc>{p6hPFh zyh#!Q2PX+UY*1r=U&K0itfMA%+b0@J<6))SVy9p`(<#SvezJ6hG+7uY*sp-n!eSA= z?WVDq+JUS3DE~NCPkWU69d*e1(;nqJ>bxuXfsa(nUPRp?>``n`VXU=baj6iv`AZ^t zq=RxS)ZguuB?96nlzPV^w^Sfl*aB2mu>v@_+p7>FFHlX37pMA})Pn{+)_LNA1FA$X z&1WMQ(v3j}RGJMQpNJbh4)K;23s$*-)I+f-!Sji1edK-qX(&XfhO*og&t*zD^&Zf5 z8}Dum0-<;1DIXg(99Vq^Y}*};b(KwD8|&YJO!bVg;K-P2BW3NVnM{t>^VhF-F_XPY=oA=+#eTM_l71vA(f_ zjUv4oW#12iz+ydTnP9+uUsRTVoBDWUmj4-bV^Fpt->y4m(=^!!AU4f8tBb4yWE3Mw zRs1;j$#+el5&8Uf3Tb)H!uxm3fOM5&E>YwH6?1`E=5;xso?gZ|rm6{C&uddxEV&1v z7AS_}Uk*5?N>pAFwsIRJXBO$z39qz8ZwuWI>WllLZoORS*1^vg-IH&YJPv468;=yy z`xYHlw1piGhH8IKs^3md|I77}H|A}cdo-xcv)BJoc()2Y?XsXV0L(`OKWm$3wp;3= z+ORy|c#^`yaV`h!6MYi>Rw~sSec{zVyqQaiI4Q13p{G^ULTA%`!0p)kFZ;l5EK~H< z`7t4A9d>JKzV;i-NH3d0C+A(riIuo8wttu+Iat6pT*fxJPMmSZY=q00(`)Ywntz>< zY{|^P1{6-bwW>9ikc{!@nVYg>o^}ZtBuDVCZkz`JntbABcS}ux_|V@fV!8oO$oPFZ|$n=M=-ZY^yW@ z+LN(Z5_+S*^!A$XIOID%26Sw~>KJ1<;xs#~9@y}YMJ zq{wC}1{(}Z=~U7HRrZHPm)uYtd<&;*b|ol>L)wtDJaiLhr}$vdP3bOPpX>&AlF(uk zKt?VIjw@#}!x<}pP)Ro~U;VrZAaC3{ok9moL(dF{ zw!ATel&x>xta|lJmW6U`JX$B-=2$7AJE{6a+%5r%pi9CPGn;?)_FoF<>@T;-$~cF( z(C8sZSu#n+$T&l6aAE75+T%LExgUJl(lKF!3+HvYh!tv717m_w)({G>H%o5M$HIh3 z1p^yM*kh8}jj?O#U*6!D)}8-;xBnEU|%8vyNK|9i4_M zkXC++?5g-~=wv@po4;Y~p2mc)X0_-$@7IYl_)(~3#?WbNn5?8S@A*`ECv&feA#nN{ zW|FbCtrBASSPh#v|1Nzr@~$Id@zx<^GiJdrus(|Y&#=NQ6}d8KD=`_2JucZNNCG=Y z)rq5JAaFOLv1b#-q*7!(6?5oyEJG+36(~DMFRxNuPoEM!=G|D>B5dK`_0AI2PAb_x zD{@o&gXI|`w#-}mp=HvKjd$V1$gq+kd^r28a;?u38Dwevs}_R3sw{_li5u^uSEu^L z%D2rvF}u-cfK*TH(JcH6M)r%?vwgxrRA4jir zL~F9M!hqcx^sDdvsmwB(_FR>`S&6WHEW0A^fsVIGYk;zGMe{eL7>~rF9>lrh*qnTo$ zo1=k>>EM@(u901yow91`Lv9~AO4rQNt5X)O<5g3gq(1nF@5j>hPzM=5ZQJZaj8R=$ zukP_l=Ux`%!10<{t36Lbx<8SL4gE-Pj^5?=7+81ChWE|atCPa^1~k)eB0dXWX^bFLDVna4IeD&LNOh{Vs z*35Mz%ZZVsGegpDiUB)Qgfu!Sya9sU8~wsm5t1@!4hNS&Cpr>G7SeHoW3p}^XdAn| zpp3I(W`Yt*;#AGiS)wO0O)U_8rO6Goavq&P?N=;1pxPDQ-Tp@+$k3#$Q~H!1*>9wRqM$rQ7WBCA2oZqZ4xAsU2o4;Ur-->5Euxw8d7 zR;JjYhYP+PrnB5KJFSmvWfQ!U&WyS<|1o5O+rrYLkoXGXsrB3hL7Kb<+I70b z7ObDYEpt7*YOo%B0ynVws!#dk!(R$QOt#{$?{wCX9qfWBPHZcVnAwW`6a&;5doYM> zuoS4no3APg&@-h%Y+YWzB$->p*~eX4&r6+?CCY-%!z!N=?xwk#9w9WtyB4@Uc=^H0 zXIyhcSo5gas%Q~bi!i~ro-BPJdmNxiB{zV>wv^5a+6(;Ns0v|YarIbp%?4=0fE(lY zSRZNA`JlS{PoDP>ZU0MMvv&rS0*_vr{2EBtL2VRjN+V~N1_>mAitk`h`D7Imz+-1Q z*)~m-JuYHpRi*~+$o{U>WL1uR?dzAx26k4(i34$EW>#f8#Q;g|Rw^bNA}==}#y?1Z z6n-&S?|DX=!X%0MNDln>|JeHyxTfxGf3J8$@{1uGfg}YLiC_RLi(y0@)Xucic{9`5 zmpAXdd9##Re6w}t-RaD97I5D{P|yMzf&!6M7DZ9nT);{du&7i-1wpiyRzZ^xA`MJ1wYdN6XQR~B6w zF$qf+OGueaha*7qY|AO6%uAY5)T9;H5KWG@$0eo!2N z(LsYnoPp82W=@3)QX5b#f`TpjnVXk2E_w4Wxnf*%FsMZ-GJII8O7%QXPI-1oaflu| zPtpmU>&Cmasv-T*G1dxLVD1=&8U; z6l{rSK=m#(SDY)vhdLG5bdjvsye!pooiNq&3JkF~s6o>RZDYeO*SWDm9q-uEwk4an zD5I7@nC;M2lq+r^aKqB8ON>K29||y2zrEnpa?3D@&Hj-STO=z%Rh6hvRKbsdZrjpu zC$=_G4ksDci_R_XZ} zrHV33B|hOJ7-%%^&4 zeM_K|8)z?fdgTc>0|X$_fL*&v`V;TtJP@74V4|3Ffm7y<6c|~WM*8C7VV6sw^@S{4 z`JwHzdgN;tuA2ZxViqWPQWllf4BK)uJ6l3Y%7s%*n9=M$@MrSS#c7$?33P9bDr=0R zlp83r)>zKH94!LmU5mZi`2Pp4C`^nkc(7&u9DNTvJjBfD{GuhVfR%=PJzW`GDe8wb zcV%#Oz)p3pIG%@%Pm{i?+xFd0vN1MHBdhP;d3%-aHEQFY_0qUd}uSA@6XI2Q~=beV^*Ye$=wPZ+sZ$7Rl z_(+`6KqRK8C#ZtN7EAJ^CFRk3jWFfP;1E^`_Y%b)0m1``i_v+guvnO~*l_XQx-?j6)#m)kQbS(C5Lu}l6H*@d=d~oyi5=H>%qIUZr94QH{Z#a&fL_ulI1WbSLnwqU;3sL2 z>&QT^rO8~ZU2sBL6j7vZi9$X1{P4VpRlF_2GQN&?Ub@Y#d0`Hx6dc)f`i|Sh#a)6d zdWWl)#023;r!?oKn`f+y)T%BnKBLLzu36x?^>sXWcw}TgJ2Ictah$O-G*o8#ly@w{ zp6A5SSP7D_+5l7m!`k?Ugd5fxn5t2bC$Ni707Ek*Kj?}p9TrM5N#gQFDAiIF2>Q21 z9_(ZTdo7Up>FHddp1v*6suII<{2T`^!~8oAVdH?)7>~ZS=B;m9N*q5Ym;uheQ9)7$ zrGzwo8U{(maum$oLv>20qCDcT2Bpe|U4~tZQa;F1fGF8FhruPwxQsa%5M#`RofsKo zB*$KQJ?vNCw+w}xHu-8LMr&A#9e?q`)tWlc6>Sv60|BHSc=jfKRv()S+ot)k$0qy& zyT#^Q$Qzf1CRgQ+(&>LDo1EBH$u%wDC%r z90s2j4BzsZWr`S)bCr`7R@{0C`?xU0&gg_GP;t^IuM~ZtS;hH8c5{9_XMMz7X@+>a zvRb~$vqMkpgjv<|EiNx=-r-OT*)X*YV;! z!IP=rSJRCGEUxVreF`l>9U$Z~B!7}^V~es!cyqwf6HFN^luhN{y7?d9HKDA2N$@ww z+2^J#^0pZ!J1FH9id@1LvJKP*YVV?y7plo7Y7UG1b^+fACsKBdQ)!jMu z1j>|De~*d8w{DODH+-Wd3eUPl+uV$gdBg@VV5*jNf1^I;1Jn#`68rVOy_6J?&GeBDha{@R+H*+ zD5S}de61Vm4;*11Cgh8^=bN^8hTU)8X)6UdKTcXSp1vU`Hbz$ZhWf%W!>gxt;&ky| z&t1L@=7_uLE>NgTR-nQmTx#Y3EdmFv)A`^oX{)r$y=rE z3DY9+eMENXgwHq5ls$9q_!)kDL{|1~>Q|TV`@nz1)m^rwoMf_#vz@o1gB&$;jQ3H> zVu}<}(W%QaU%B?mhGktpdK3OB@qbsfLYO1zq2Z)#Q6g#MH+jS*`9>cseHGQ-C)N(t0)6$VbHCCE|m0D6`6pd5x7HJw*3_Ywy)YS^co%f%!eUatt& zd5v7vs@~eniJ?}KT{O5MAVRj6TdFQoYgH!=u@QPI{hUUbp5F)?(udWnXS6D6ctdly z&#U0KDu4-Ho-}*hSYwYO+pNa8AOsJyLs1Uz+dr9RLQz!L8w`ncViZAh`>0LnL`oS? zk&RR|hE(X2VCjpIBOmLa#y!DyOoUCCuw&fuKEetUQ@<8??~J8O%ciWsiTM(&IKb)R zoAWm-^8c$DSiVM%nxoDE0bDwr8hKh!tzIE4j>zWTSO6F^+%Y-mvI$S-W@WrkvFZ6A zlx4XtHd&44@2|=t$!yt_%-vhLk@2YTjG#e=Cpia00J_0Wl%FlY>t#&jg*nuT;WZ8CR z6MjbD#0<>e#WP_kFy)&k$r^T8a^mnP(+o@7DCJg)BtUnPI7`A{vDE!p?U7h2)j$uq z49rUN?x0(x4Gzi;j{b_De9QtEPaep@?`nDy!sI4kq|f~I?@019rhi&$HvL(Y684eO zspvcOL$5o$72dEZ%q^L*#eGPQ6d9-()g4s9Pp3bnYkl4q)%X;KA^$~7R0ovPrTZM2 zrB&gE(?l;kEYM(X5F9-FmcvR*(C#4nclc0_h8Hw_Ethu1uXde4ch@ zi3i)xt)pfP1;r`DtA4%1vMkR^^;?PZQ~D$SHEvo}8qm~~ihvYWo}+VVs6Z9=e(CgZXsRr-M+s-NdBECNc$3!L}> zAk$TUwjqk{Fkj_(2MsQBBD{x`;w z`0sI4II%Zjh2Ae&fjm1+0*pmytp-C6%zo&^kONDej}8Vj2$=qv43yZ6j7Q7gq!v@+X0vh^~9*!6DrV$L-$ z+-Jo0!%BfHuhgh|<>XB_yV;(Uz{CJ0>)yr~DblGQE!F(pvTx^c-_(f%S}VOg zT~OeR!aZ7*o?b`AL}tz@4ewN7SI_@Jy31CmOuQ*D{{-t$3I4;^UbAGRX0x0)F<`9l zSEC0}#5Z(?U~xRu!OQ;-kOK2#6IjgWaZ=8TnRhjoQLNLhB32^S1Kf{AKzIU?>9EUX zzmL4n(=g>H>89Y4pi|Nn!UNJto^qS!+7w<)$|RdU#JXV^V-1-SU+cZ*S=W$U8(Oir zoZjixt{5P=d3%~{3tWyGO;~}7+*{jbT$RRq>v-vOtT5Y<9EG$;s{b80&3cp`#Uaw6 zn=emT0uY-3XM_!6M#ZHThmOY*E5uAKy!wUTTJF3(Q(lLa{k5~Sj@JSCuzQkbRRS+} z?uIDbo}U01JaqzX!pT!dar_O|adJg2>bIm5W#e)LVpnXw3ghgtvpN-HyulJRV~j8{mt?mk|2v-b#)(5*E9*`E z++y*5s6xaph%7H$TY!HHbEQaj&<7O6$PjvTPIpL?Y_0H?f~gkN8Mb+FsLX}@1akER-mz7 zvP;k{FM`&+T6K^5fH+s23O&8Zqi(!Ys~QNe6DNvMW3V(FDqXXeA_HZ+=(+-HUU5q& zT|6W&^X`?M(lDPa8FcFtY*^S0P{Th2KRZu&ox6f)RZv*Fe)<5xD%%b^M{7`V9{2frsEll5Curo{f^2gT#r26SThS=k1}SvNMD8)&zT9&kE+Bp_AF3`Qn# z2i?(V!48l|X5eb?b#t>PC)%Y71qHL4A%?@ou%mO@M5x}2B$yT{BzsWD+RE7Yz-avaB* zZ?Z*iR7BN~{AuKrStjHNr7WjN85Mm!>h97TL?<}}J-KTZ++11_+!fX@x){_Ku9I99 ztXZ&8fdaA*c`Z?g)dyhTKPyxx8FI-Wss4>3y&8{K29~>JxZzPn@C9`rDHadXI?2kQ zN5`LXugtsVwT2vNy#CqJIk8=m4Em6BeH-(->rtRsC^7EoIGg9Ul?<|RE_iBkc zPELmnhu{+*$u#5|MLS=yen ziR3yl0mex)b9jhS9-yFbG8!q7TBQjuEJx=CRSU~$6i3a`^e!-DbI?UVE$vomj1Vp) z0fSq-xLUqC5V<0OO&jAH90N-piaE8e46;skL%m#|0#iWDHupx@;vQBPgsr32Ifm-Z zQQ#SEEFA$LR^XYk@3qZqpY=WNW>pW}0n@7oih)K~1Eql>kDM#6mR5?cLHYhf=w4c% zD^|#`K8(ZPVTFvTVf>r(Eem3-RFWXkJ*uiO$j6!v%xBq_e!Ro~TgD#Hrf+`iMJj1fM%hEZ0Z7TF`s=x< zER^@vHbBYWt||U`?}@q_4*34B|vl3uS!18JA79#?#!9SnJsj7d3Tp=GYn;h~Jh#MliIi*X!%$_gY?e=?lJ zk1*Mi?RvS6l(Eb0I&qe*+063PQp#f#siva4=xX{H7nn^_0vZJeB5%&u^N%XB=|+g1 znRNPnQDtz6c%yr~XQ?N?x!$8+G3-(-PMOVQQ(A6fc$#qEoSm~*$>MnlygP15zQDN; z#aZ(05tjl6JR!9dstfIc%_$^)Px4LR-Sm#<99fVB<>I%zO9ES^jk0QDPBSJPP{0cFoVVcN(qgH#Z+{$qJ;Yf5Vi42qncEI zz%wJiUxRle6l)qUV^>Hey-(7-uqCQa+)FnJQb`H-A+I?qbLsWhE`Rmnl5RPhxwbrl4e)(LKBh8XSu`qfv$9c@k1W`@vR;n!fRh@R7veJQy z;$V&PO9AM@psG_Z-3W4Kf4ye#X?p03L6-tLIX$8kJe^Rl&JeGoE-LRzb)XV{ePNw{ zx>z^CRO#aY$~d9$<=kG|{G1kP`4U@43bKo^rFrj{5_{AU0 zHLXN0V$A`P#m=R1;t(18px;UKd2yL_;p<1H}mYIen28d=%ov^Fc+bvd}Zu z701Lv7)PVdZj@O%EDl5%4mEsp>&)@E!cL4EE3R-q)LUUaUIF(GuUlTeXb-)TqjfKk z>=Yj7)|1V?4cx1KEm4oIX;ozr`=O{$N7s`=ZnmaGj!(8kWzzlfEN&`io1adyg=9k| z7hbRAwFQsdkNZpq)Meg!kWab}`wHFiLz=?z?wUFbNS44FXBhvb-Yr9cCIj@>w>poI zJx=Tj19R~xlX8$!?x#oz6}_DkD_qS(Q9=gQzq%B4M61zr)1mUjsN{`BZ|DhU`zQG_ zCrO726(~{47qvJ_!<(T+HUT1?bTMup?VP=H_KAQz34_(c#T>j1MXtsb2wD949SjDG z(d|EFwRTOZU;Kxn@uYv9I2^Z<{=LQ98f?t%7}$>WbX@=d3c-_37w~h%w*^pYS;_4N zs2zyh#M?YWt6CenHndp$iZjw_BUUC)@v9FVCB)O(A&u7rb?Tk z?oI349NZ>q0`~j?&(iQT-<}^;@E?5X){<<^&G{`+C0+w=c$TR-rO|30J$FIWEngil z?1y))4mhEHS2f`2a8!%~2xcpbgK@x8#N*c2)y@2oRf#$ON3za|SEX#TRq1U?nM{!_ zRP+UMD<~~=51lbv~l|qJzpi{d$AABDhKQIu^>s2j_;6=DZ1tD!46+w#+bcl z4c0*7^n=e-c7 zo#{4@1+({B+)=AP!DL||Ip!+^k)8v}#+abW-w1cjV^7j>+VMF^`$Z!aA zRD3}Kb-uwz#fSWp6}x5~k~1a#3B;hiiXHwBJbHi>q#$g-?L#*d3+hy0i`&K7jlNK+ z*-Iag7He93uJKlZBAG+u0tsa*GH@ctDKzwy7q6dUFafpgp z18$fIEDi4t)6ZNbV^DXqgYG4V{23!lT6Y+5q#0GgV}w1TJNywi+Y7qS7C46=7_s~6 z=YF}-vXR7T_m`}+hG4|A&j*=t_j)!7&d(@SVCN8a+b$pZ`W^a1ebQhy&-0T8nSl4K zJ1=9jC`|oz#EomeH<_Y!>*E)Zo@dPG-DqZT9#YCdirfbVOJIvZ5wH%yAf36mM$=7~ z!vCm846(Q651GkOt22!#Nw>~JkGj_C=hJ)S<t#%?KeVpkj5uSTVm z^C;yWieypIyt<_AN13pe1O@LK14#}UbyYj4WDIm%Jb#5hn zO#vRqC4z1L*STO{ssj3jy3iHER2igw(MUo5H%DkZxVU&PDO-3SWbfW=lI?R%HW2bw zPtVfi_@&BHWhJxvxN%TLk>+hm9uK-7=k>@OgpYB8VGPUQ7<^baDq}5l2VW^peA8rT zE`DeA0&@4cF*IAumcUh0mC(*TL`AQrhF@AsZR75x2kCU*cuu>bh+FGZ5(HEuUbmr` z{JNquGFfp!-t50iI^dQjge%;0^i{7)`q4ABB;B`4guPDbbP+Ucw*bnF_m+6&_~Fqh z%^7gfQ`9)_RY|Af=DfRZ4;8ED?uVlFCIM{0L&?}W82dy3j@%`9$V*qDOtqfwhpJ}0 z1B*TK{kD7_OUgbq-0>*~h$y1;|GnvJAAEVpW$(XLEo=Gd!yi`gkG=HXS1$gSR<&Ar zeO_O9OVp#kJwCWv+R3@<6$`X0cjw=n=d>%~FXI;Oc-GU;m!SV1N!6Id@YF~>9p`)2 zb%hjgA!YP|cLirX2HhH<(9=QI#}-?`&O+HTqGRr1XQ5)tZoKlWZ)o)?;yD|>l;;Qg z<*Cv^MXtD>WX#4W{?N>1E>j)Q{bB;|kmi9$L0D|myR#T1he`6qUCwgY*u18|JF_(z zvzQ8go9Ghf9&CeNZ7$aTKz6AzO4WeXA$mFv`RY_t7KaDNwg`CCe#*~Ji2@Gj;zCEC~>i1naS6Az|Hti zQpwKOabmDsG7GBCP|A9WoTQ>#l)I!TRe|oZR)smj7-2WW%y>PYGIWe$DG!RR?Ug6+ zhUB^8Vhz;ZkoTp%GGx)pr!e`q!!_GAg&-$ewHr`gxiOdn1h0g9W+~G@zd~h9V^%{r zUm2V(PNzX11zT~LPEH-2CG4M%S;by*1cYPrDSNiiY=PP+qhs$OYm9#3yFb{kHUTOy z{hx^>n;oE>I7&Zm2Bm=25E_a{TX74URgYYs*1wTj2)${e7 z{`oP&8cs`8jz3)J)>%L8y&9pgey211!!Lt1b0RwtHr;z={)(D$q_Nn?cMQ@)lwJDf6s z?glC4eTv+rq6gft8n{os zBb33F=Ma52eBYcwkLAW?=$3aqS?ygKehA18)9G@EkV?ao6-=IRmlD~hv!E3dH2qcv zzAwVVVV6GON^Do?TuV5{oiJc`g`M*npNHO1`?b>x4{#d^cgP!PaCo8m6qpi$hS(i9 zztU;+j}3nh{3#!JYe9SAj+@~V&wK2J3?BMC$ca=Y#T+!GwA;y``-y;*Ijv9z(y179 zyBofazkI+AS5RY-b9q<~U9T+mKfvu&l+3uqS>7ab+H~DFD(Ai&Y6749FU(G|mmT<= zIFdSTW|68WP*!ha(yw zY%^bDySz`hI&N#Wv3{YeWx|&Yf2R|NPFAXu?@TXVlx5&5Nl|5MuDhlBXVN=-k=ta4 z>JGhvT&3@m$#a!q0@*-g-1E(1fZxn>{fp&pCL3rt@yceUoywSd!DXOag#{O=29-xA z`Qnu%@p-aInI}1{2FAVM+nx|=Yj#r=DqQt4HJyqh0_gsg7jbmqTO$xmJH0vt_LtGR|bhEst7PtqT0kt%_J-Ce()YIAUXoB~+efE==!_A28OTGDcEn zD!KK!A6Qxzr-hFdE7ELWsK4cr%-4e#YXC$Wv__SA54yz&@nXgJ$cI?Z2`gx@9^L*2 zSV3dTj0Nxi+{+XUHUFe+IZ2sDfK+!>&NPQo!j$aBoN12VNB(P&xm|T+QMF(%RG!qr z&J5Ox#0TGiY)3pNWe!$;U=>PyWU5N82JY_W{&Kn%0cz^ZcR1;QKtqnQ>K_RREGA9n4p{Z^Z;_1g zF8O_VgLl06?yQN#uT{+10&J@fPv9G0K2_%LV?xA3W&L`hWtU2EVrLdIC!-KiNGbCu zvWJR}^=y*ekfqFl?ct*Vshm5CezHPY6%eyvMJSR*Ye9H0R>O^pS*-~nR9b~9o;{_cg`jLRmjLq0HkaIjHLm5{M_-#nUgs0mDfywIr#5> z|4Hhdc<=D083bA>WeY{lQPCsq2N>3(n~5RKY1JT>{-S4!c~q?EU*mQIQmM>I36iEE zH4%0UFxbP;Pj1t?P`n2p)2fWkN6>ndE5`8{)HB5zD@^@X@Ds>8WTPl00N%)=jbSs^ zRN`f=s+y+{WUhJT36r_!1y!6R?vTqy_w^BYbXz(cmhPYKZxg;8gO|q<^yDi~#}~Wx z<$wLRWIXGBZk=SC>|>F!8XU8=M+44+5P_kk8)F_Ud$DOopZ?w2cn2$J*f^#mZes_H zfBt&^?=00Tt;jyU%_&#ce{~PkGRJdPfJ{{$T}(eztZ-jFZPPc_@M2%U>$PDYhIF{^ z^jP7(W?BV*^|YN|&wVL&*_vsC+;|YQ-m|oVpS<*nBvty3utU&6VrJHpoAdvL%si{P zabAb_=+a{1kxCiX(LO5kQZ%o zZ3(Nqy(5f-&10}a>=chbhQDdaD#9kq@4S#FR`^Dcn`ncR`bM{)d>InvSf zB(-$9f7=Xbu$0C6v_#=(tAIVNOVH?wC3{`+s(@I|RfWCavjqINTcyhkY;DRP>MULnLn>K+Jts-?Y<-tU#6hA{*{%l`hp@{U{8 z>{LOkbffz&-@-5~BdA)~D#dby)1nOb!GJ3?zXZNZHoYrN4Ng@xA3j)xj+WIAX(?&YlmNNw^QQf!?bCx{cEI;mZn zD-(_rJ8c!Kg=}i$%MJgq^l;cLWKNvYv*HgSC=Q?wT&uKA)C1XG0Bj%$LLi@(bXvYH*loY?m4HiK9)rQAZ1%~W(L^gklO4TJmO z9rQJ?y9-K5n~1?unY#-nUf0KbMp)p&{J{x6%etPvQ)$A5+o$~($#y69bbwNN)HZ>h zQo^#6K}BDJg4sdR{z5v486Jkv37H(aKxFd1yo+AGeEDsNl2WCafffAH;5^mxLz;rH zOig7l*4*9|RSK$LXAS@N)StH9Gr?l+$Fc!-f9}&BWaYg~`DDvacTY1RC93NUhQv-I z>1Gp>NGanfvXP2Dw=9`k0Zl&1ifZ{`Kc;4JJh*`q`3+nBFSYHfxBB>my}|CQciN_& z@IUN$5E@)nmMGZEY(3i#V zY2xh1BNW;OXA2|p=Zj$%IStiVA;?`G@olcCI z!)BN%pp>~3(W8`3wX~YP0cjG*xJWuV7XvN;7v_O^UD6%E)oGN^*+q9rx4CT>=WzD8 z=W-L4CPu7}xCim_9oW^<%|n{JeUffLm*B#LC3nY-mkF{c?C_Gpk<`3s!b^GJpVyMS zY2+QVX*ojE%~b$?(nGMhczS+PrH9>q<%q=I0<4o$7gW z-s`KqQZ;wzm1NNEnygaLC}@cqbkE@&5`8+Sj?R?aB$-&WN?|petx!KatKmv806ZQUd-|~!Q($0 zrko?G&ka0aO&tY~TuP~@NG28C3=D@&CNZWX0U*l*4b#|_)I*~H_ZiotiW{;d-zGsR zgip6!bLS$L{jC?)%^g-nn!SZ=ATSp z`PQ`tLi*T&#fdkv6U}<{S58#|)BQs#I!#qJy;1(o=SN{clh+MysYjt95v>PKrrNT4%%4OMcoEL<7!d-_%HM{~3 zJUYPnKr!Tke+RG;dDr@sMc}=opDvK-l%vmcI3+^-njv|mpO!dj)ZP;uKu8@G`sBCdVZ63*)hT}a`J^FP;dSb~z!Pu6=O0?DVg2A`0uo|_WD)Vg*__sY+_`2RO6MuCJK zCRTl|BGt`gD#~B}x3gs9bK}C~m>HHdO8GVgzvJkOv|gR^LOIpJ-v<4cccqPj3+hgJ z=3?|q?vrj=mtYm=usU5GKVf;iQIlmlwAsE)vX!azEEC8~=s5M2uU;lwoOmHAGMn#R zloCwH4y3HZJ?32TY996+T<69Ln^irq_sk5C@9vYkb=H6Za)P_LE}wtv>engk50ub9LUFgM8JIX+b7Ca^tmIB_wAoZ7%2i`=QW( zbb_s%QiAy?h0<=<60e;#~E^o5f(%Yx!flgNToO(cnf2MrQtQ4hrFcO zN!$U?KJH2R1&-FYIT$;VP`IsFQA<_^?dCQLx`EyO+L9}h7%2X(fZ}iC*u`GEq)o0x zULW~i^TOwFjFza~-10?fz9VCuhUL6Ds?oKM{uCq!TB6R7QX2oR7~$KC^s^eJ*ZIf* z3+hVZ9@&sfH*9YXxoq>hEY75>$qvt4&B~y{fGZOG6dc`>S$+b*IHpyz;i8%0vku0q zKk?;T5}2{+GjZYt(F)yJ2fs038#h;+4GjesP*_qOBdio%mE_YMg5uW_cn?4|;4GaO zfgR%q=52^<@V+dq;O~7r9{@K`{w1F#m||%wt}bjCPOHs5>8h^SbeNQ!3QRfjDV=3lzu%tl0v z?Ny91o}(^VcS|xF6640x{_4aA#7g_?n$V*GdOFtUl=}s6Q5$6ME0ErX!FtoX3kKaL z+mNt;jHA$C(>TYpA1b0vfztipzsO046YmK?)yk-vr(R0=F-1Pa1a2jL%e#D$UYM*X zTa+Z;6tN4LQ=2?6#>S}oplcGhLRl2?4#!X{j6~Z82D47Z!;pI+x=?&4PZA$oq10+X znH&X;Z^&{usB(UE&ikTamul%+9wvoxs59hU(bnm0`~uHXNd91i1A7l~O4JWP#@`LX zd{AT?cEQ{6JehYJv@TFIo!L6Q8`Ld245AdX%x;gtiLFtuS=O+V zQUXK7Wh%M}itge$(Aey;!<9+p#3`{DB3_IXiFbH~!ZyBEb4U(|Xyzmx!phpvLID%1mO6YU0a0MtIkr;}&raF2Pn@%Q^M zIUF4i|4jVTGG<>5ebeyI80)HYfbtO-M~;3jOaC`N6U=_M_xNVA`?-NxjTy{JDJ5)| z7EsYBm(wckQ8cKG!d`$UEoZ-aXGC`BS>WhIuEjN>jq(-3gB};Wd+Ab9`65&%Wv)u@ zC{i?iAXCb;iS%?kISq}uE#7y+Gc}ouGZ%Nh*y;GH49B5nWWF8q+C8hV)3RpRO0n%m z_Y;~kngUlXo4T2*3rBb_4|cEMkHF{jqm z5xpv7{^yVTTOM=N@n2-MU^wrda)|bo{6CmXj3P@RBp*Am`xIy9J`GdK2Nd~)iblcu zmZ-I_++5Newe!{Tm;SW*<$Ft7zNvpTW$DUSs=mB$SqjzhYWuQtQMYHF5AP7PaJuHF zsB`BIxeU_#=d^?$06wdBuV&fVSw}eAxSNzkoNW5`%ul=_7yRQhKWmOE`eDvD8h+5b zq~pK-bmB{GKV0+uhd=()TO)%Fd1bh@gb%CF)6KHNxdU!R>OqgX&jv_*=`Q?D2bW#? ziSbWZ_aDaR^nvlg7kt~jPLiU9>;A*|VfViD;2X_R_{5(wUa9|D!Ixsce)B###u2A7f;rLZuvC=z7Hp;f~NmaoUWEu%p26YwYjC zE?5juK3!laF=8|5N~IeVf(&|u@Ei&S)sl?wW&?D`gDi(4X*8X3(I17Hv4AI>tR~s`AMhWRh5P#IFLCq*}oAsQ@bX-#&xXevk5_CJvhPFl$Z8o z5R-K&anYS7Yo`&Y_#Bl%+)gQzD3VA;FS}S!g}IYt|t-qy{lnz2SulD2GqSs(H*cVyi?p zy@oKDCB)Pr);z=W=$#wew9r4tWB7mIpgW3lpsTMB#HVAaO8T08B*ZLGZn?17d~S|k z+s@db{KM#wKf0K}l>f=E-XbyVvW(9AGa*_t=p|6fIErkbqFX?@uvH3r+r^O4X4?3L zVGK5%kH-tKZDK6%vJpU*k4^abTiSEeXPO}KwY;DGitLz14w-$I`IHhGnzE_rQ&4sH z_iKu`u;%VpgF7>n8Ejp+DWf0gWE%6nW$EU zLb8=K-f_*VIV@XKF?)ihz{cEO2*d}pF8(zvl=7ZRTS z5G$-u556?;d9}&x1g8Hpkz})TW}SDDg_vn{=wC)DODR%BMZZtdY1D^mi7I~WilicV zpV##T-I@+Tm%LZBGHBPlK_H;ryXX^T&-`vp7QJ`jK**W}<%_hwoeP?y21seR);Csp zRZ^rz%A5|^8Yzpo$Y~B82#F79a_w9&?B`&-jT>g4nWU#fhs_;te_nUOvTfLDjZB#t z6h3PN?&p-dWr$b!?o>C*>gf9s@#4#V(5NTY1>wKD1iiAm+;smzw-U}#UN6MYM;3s8 z?uN>gF=YFK3`tRhF37>tZ9WC#nSAq+9DByj#)SNQ#ScX$8?!#`>Muy#G+_E2{o~w8 zDN`x39TiBkG$WEM$t&P{U#(fKNenOaOaV>@XeX9r00-4o&w=prc@z5GkNU<~ zPs6BZC-RWpG&n7>d0|QH5lc?T$LrdiIBjBuxe+6xMiB^vsj!E$msHJWxEk&Ckf1~u{@qV9TwoXj^)zz8~q@a6d74&lnc_eIro6|_l~iwE5p+-Oen#R9T) zF$yo2ik2S{6!0@OxxQ(Jx-z(Dg-}l?a+;uz4r(?NsIn#54)A-~L*+AH<4FfBKr!~M z)}b<1?eV{_-0W&eif@H`{;(SBXHg{=i@CAp1?a2e#l3V2DfJnIQcGRvNU|JfOs~1T zaGen2P29#s4(m%KS+P&lAwVi}dZkbG+W6lu#jajEdgN)p7LnpqmbwdYRSUmnEHwng~68T&@e< z&23g4kTEgBgOGym0`M)>x2f^=V)%(%{(g{Wk7s3ku7WW4d22((L<@;0vyqFpIv$FH&;+UU=;|*8B z*~42bNpsj#DmZT;7_q|rlPfkB33fui!pfi2D~tUN=n8BGf+eV6fNURQ=t!78VMWl7 znJ~8Ccx-?Px{%WIL!D*G!sD*56GOmCeMP*uNrn@RlqQ|9Y0RK0h!;b_Qy3PMb^%qN zoj-H?CW6iU+xH3fJ!O5p{PjOyd&6=|k&W4K;>4VlozV`#L*A{B8jdkJ*E4^&WYF!3 zBu;sft_;4eILcc$FJ@*f>7TcE5t0U=#&xZal__gQ)MGS>^RRlTZ8mrk@6q51$*Kc7s{_ z!vjkB2}Sy$$WM()+w-K?Yr9y-d(R8mxbng(A>Mv{(5=ILH?LjPC&$;z{v+6ADE$Vom0bRDqXA#WqP47>b%dj&{CgmJVWng zhhTL8YUE~tCQO>JNp>xy4%}Zor+e+e#@DC=v(qmf=w0Zuj^+)3T zm<7m}pAJq3Qb#n&2BcfuTcx?ct%Za%n4QZN; zRkAz$CRwJYOPVT6mF3K7wZG=YMmUWXN!xU76hc@X{!FvyOO`7So9v4dZva?XldezC zr`kkGzWCZIMb6h&DehEHzKOBXd0{o54!Qa4hf81On{3Qqne%@n>)6>CCk92fnT>gy zQYKSm3#>fyQZHs*2Xihi2-1x9;*o$sIcAiG!St@fCI7^K*AVQ8VFgL03tRk31mL71^| zF`nEA%Y}`gqv}JzTsG`7^1phr-ftIhjqjDmdj9R2;^>ka1dA|`mhVJ(5eMJ8#jEu= z42_PrcoqEAsCqloCal3@lcQmF(+D0|y;=v)7}?yZ%hqYbpe-v>K;*-CG01=-3#1@T+*SV#dvhm^ty5Z-ro|{EW<1D zoO4mwdw^rrf~t3y6zR{=1J7{lQ{Z1%N1s#yb|(RbW|*|~gcS^_@cFaqUoydP_115E zpH#7PiJUkXZ!@!k4U`h(;NPL5Utc}#s@FaFz3~0qHPb#8UZ6iz4=u=C+~ZwNb;@gf z%H2`z{d&~vccqESV(|rDr~DJmN78mhAGah(t19O9&ACbPIM=*3@ye-fz*f{LzXv)A zpUk@{ET_)QtKc69>hUk9&Us(pL9v9_VQ6+)71=9k3a*5sJ)rRLZeXXPQ>E1mx`89+ zb%DM|ALrS#?oJB+SQ$X$%}0LrH^#$+sBhQoyg;_HLzEK-^2KI|+D$1T6x&Hf8!M$^ zB0Ck=xiOJ3ktjV@F{>z$LFa1lWh{xuEpe?1H^B?T`or}Afg7@7dA4T43ux>{i5)Y~ z=8=6%i#nh67PrdaWX0;aTR{vcSDY(EU#Rh=E0TRa3<|T9F4^gY0`81<;Xa={^(kov zG3+R0(QPz?g;Xouv@S-fo06qWz3bO z%-Us4;Zmmc3yl5?O#d?G;1`(l%b3oWnO!e4?|re3zU-H#iiup_=$a?I%t3+Eb<_la z$TqVvvT#^?{cLU>cEAoKXS&}CczT}CiK7}TdA{r1a}ZnPhnK<@PpYy((-qe2y$j0y z6Fy%WzS;K^?+(G)S;r-fqLM|cy;I#L03Kt`Cu_`%If~;iTF1KQM_yR@oWb-egE_cm=5OT(u@tdI0__HqXZ-VqaY-1{M~QTEWUB?`}hy%g+2Bh~Y+Mt`dUmY+sK>ZIo;kBVJ7Xl$ocE{t=|8@5_gQya+xVwd!@%&(y(l#)n6^wAE$L&t*kXDlskrqLO~dGo~?!U zvvN9LdRkTLgCYy(XO!AUHOP9_Sm0yCUa|fRdmdngkEz#x7db6_BztxsPe)3ho9tPW z*)n>JQdU!>f{F(94cP!0nAJnS1wGL~9nYZX$weq{!To8Bl&c~6j@{F_!sXjv-AwIR z`stDZ&r*4+00n44cUitm*&TK_yoGaa(Y`t77A;3YZ`kRAS_rKwSyAGZ<7d3x$dp+b zT&+3bfxW>E0fUP188JIBH-f#VzS@Jek!Su{GV6q8%e2#487sZq4IrHS*K3M%265xf z!FRbw13CoNzHjb57WgLJ9)7iCcyulpm&Y{n{hGfz1_kD0I^v9#<)LoP-u1ollw~?`-rh=Crh%g! zSR3dnDEO`eg47t{W`6ISR%yC;0#Neg8MPHdPaehb?{3RF|Ff@{)}yy_4&5XtpPTjQ zy4iZvLMcHn;4BKi^+9WU)$F$5eCd8DR(RL9#RsSgxdYM@0rx_(fNeP)l%j^@m;I_1 z-k^tEI@}Ki70o>xHsqppzn}&+fEl1MC@&T1-S+u)g<$}QN+{4Ml*4HwC{R%vj>m?U zdL*OJ2ey&du`B8IVHd2{8Fa_K68vne+sOy2=&IRizHoc1^aD+qjM<^t&cT(@=4u^V zq4m@i$>tkugCTuVQ2MOr>#{VclylxNqZsiQ%Nt1leBA3RQ|@+;Q|5;r4C<5A`0SxS zfvSZ3@C%?ff%OiPU|;N;cH5?IJa@5Mm0FAPBO^@qXS-gmBV|rpCev(YQEDmWF^W_p zCt9qqct#f8tV-Zv^L&M*P{ef20xm*m8A07htOPeS*tfvGAQI6qCH@KIvSgzoS9kzK zMsk$fd>AC@t>LtXbjf>U3}%i-nofpY@`SsTs9^%#wO`B+ALJh6c1bI_Sp54bohxpI zQF_LaDY6ep4mK8JAdVG~rf!O8zv5;BQu)jOc9v{(Vjuz2_$bkfG)nn41+x%cPRDZw zNVX<}yC!ObqEYZbf=%>z+%38Q1mtT$KevZIA6Dx#3HHS_BP`J|#yn2UC06L5WZ!GK zGQ)(9uEil&NZNDb=9HUZDUVVD=}Q(BePh8HaAj)dRDj9J=f1P>C^;!-jIv1RU96*) zf8^f;s})Kw8+GA270BFdx<^m9c;5}j5|&*4glsnBB`te%#+xK-IHFQ+>Ad!z*(8R9k5kp3w? z_(;H-&^kJkUO|osa{SUl+b2g|!XAiB3@FS;6~Dw|es6-vy7lpkNRJaQHyh2$dmmED zL5kd`qH#)5t_pS$=Wdt_lnSmXn*IpptY|kEO#~dKXgX+PEb=}Xa07-h^giM_rQyh< z2OMGCmBJKPaCWpno>1$S?TdmmD})ypgINh>FiociY8do_F!V{M+cl-(z+vW&3=WJg zv`rKj+9XR>G;ljrAI#X}UNu`6s^ef2oSwb}B+xrl?HUm9BXOZvXl_uj?4mK9h!pfq zGE5_)h8qTKmGsEp0<3=qy@MuMv&!f>#Rz*u#=jUp==rGYowD|)UB7zWguf+Kk)fpB ziSajVW_0dQ%8w{=i;C`lv1?Yc;zC#(pV0?mpsc4ic_0m^Q7);At|v#;O)@4)oTJK9 z6-WvIsmNT%6b0_#?%@KlM<=+x*SVE+v&RnK7-*5K<6QAwj{1K$fXWlY^3AeUvTl0c zoIww*Dh3v^yYjAC_<_Iz3D>HnYk9e1lqba~FU`R6iSHgKdPqhe(me1e2-5$@IB_L*G{ieP~tT%q}@QDWuAAo$BK=y+BxLXXgkKr ziJJ1`pGCj@1rz29=5X$i>gQ%zyljThMoM{_BB!Wm=yUlS>HzKK9+}z4Z4TZR*~i5; z+B_9@T> zDyPaK;-u+bH-+)RX`W|&Iu)@VH|T6`F1N(*W66+9HTSH~c~P}o=h`8y;J^OR`;OPm z`E~T`>Hcv6_X*^upmE`>X265B`#)I#91Sj8j=!D!s}0X1voxj=n`G?@O!1+bTYh+s zMmKwxFN5t;CE_Fg=v1MoN(Rsov7Sv^zS%MEVGV*Y#+Zmp)~(RMtM3;~H(8t;zij(E zS^wNv9KD&vNuiXmf43FZM7nk6S$ehihk`g|Owfno=Fr&4{P2s58^nFw%aZ$$dyW;R zcuy8$aJE^YW+LEWWy}S6ocG7mhvLM&zE=8Bnq;UlmmPXmGyo3DHTe$DxOqp_8#!Iz z1=SI7hcMfCo!h54tU0RgCpl4CRgthrm_HefzpPL)%8#)!xemMhwqnOGW}7U?syAkA zBI!=-Cs&x+guRq9pCUO_^hiBUt8}9~gC_?it**vyqYm&>8SGEOwxd=>GSpObNx?%d z6y;Gx6jeq&qDRhEvs7 z=l*7?2{8-ae6NF4uv>GSc%Qu03}2@yC5R!^QPEqcxA9vQ4?G44Q_0=wg)Z)qz|27A zntV$zb0Q$0ZXhS5Osmwu39VSmucI$6-p4WAYXqJ#i^v!i(>nyGRS+0?;i$|3>G4^W z!JE7f0cx2h85Fk8s0iEP+7*#bSIja}P2(Ap2e~DD8y`X_cS9RHOMp5EeHLIG>vRNH zw!M1i|NOUN6P_{z+uBJyJ3Kk@ItAuw)Ha5eQl?SlZ7Lc$23nj}!Kv`KQ5%pms!fn0Nv?^#i_SxvcQe zV3P0z=cGDmVcEjdv+CyM3O^lI4v}+Y%SD_I_bDj{CBs!J-Ed{9Z^RcueL$APF{MXHvO-pQe6P?&xSn141Qtxy+ zJqTSCEYB?BUgDrzIw@<6IozBT+wxy-pBZJc6J4P*Q^{c`PLjND7If58%99j1K}DZi zv@H^OQkh#po4tozHhbR+f~{)a2a4=aXgE2y2pg5^l^3~8V#L*mD)(wB10E}{CuC3x zv>s5-tJU<i|Uok9XQCJtB3bd>jK0~{1bV(GaEFWVcD7ol64fuwz$V# z$=&BOB)>O*R}?;0!8bl*54>!{$5I2)>p=r!l^6osO@i z^cJ$#9~jl>8n_#e?ZMI$Cf0EhflY$MF zxio_(;#d?|8;u=0LeeF#c$%Et))z0lL6V%<$t^K+ax*C!%;h#>DMYk)C!VO1BbpRAER`RfDHXit-C&b3s zixjgc$5HSYH4ZDJO!?F9Pe12pLdx&<9^XuMJ26sf%#cz_DT^pl0IJfFtG)At+c}k@ zt+Oxt!47SwPif@lMO*yyyz>HvUG6AS<`lYZ3IYx?PK^&!Agt1qL=;D4MDBCFIe*yY z1Nu{lahhn{>PrxV6KXVZ>#N94cH6;D zoPaxQhOq)l2~m%picaF{>4zc39E@;)K_-Eua?tZ_R<)=Gq(@Zyp!@pv0%)5JuJS@| z-Av7KNyV&;StUUX_B~c|I~7$6<9Vf`T=Ay)@A*#DRnk%UgIRyac9v z^CVfr4lhm|6vJXQs(fG@rQAx91f*LwF9(p6Q)itj{_uEiH2laGF z*jp9ww$f0v0)>$N7BSv#i_OjVbAS%m%W5DOTYxjD9;*bA<6 zg*p$e2}UmdE-T3D|Hs~!z(tXs>05)WD~>-JP;IAP#aJVK|DBwK>>jg@fZ~tR1{G;T090v!Gl4C|NC|~ zO1HFjK{s1t{yV=7)n8S8-PZTi`@QG$zA!M_ZU&=did{#MwNz9WB%?sTK6=;G0zo&u zaJyT(p~0HgSPBOuR1UY4hOZSDTnxX@(Z`Ec%?O-m!bODsJuO*0k))YD&_s%jr^qTQ z>Y#r-ke6Vo12`|SQ{5KlM(6(@JRY_}+*CxuSFh%`-hQjuGUcAHzm-@?qA!re0Obx! z`!XPpGqya^A|ud)pg72%#+2#!KrBGSLQ|>bqS8z#ZB|8=sv|_LA+;k#DORrJK(1S=IBykMPPoY!_uU0 zQbbnVv7~U}?dD^JlqFeMp)LhV(?Y)vdgH7kz$TY7d!?#VFz9!N#u=u8VOq375Bl zfR7~dSh(yl6D}DPtE0$f+=SNqwkW`gMLvS^*=atVq*ba#KEe)m7uy#U@2^)t9uvAn zx`N^b86KEDym4bJ;4Ep(7tn#9eynO; z{beuO>Y^)wWdIDDOihVrlaXKv6O^!z8Is}{J@AgxG@e=>Zi3*c+Z{H*ZCjbR>F4E? zN%eT-_UI%Ta3WjGcKCRTT}6>tDyk4NIPS~?$F#|PHlACy~7Klm|AHDfq+V$1i`p0jh?;=Iz}Qvy0!fwm{bBZCy<$ftm%3hopM-3o-Q-+ z0V1!zp6++Wr~^>2Co@P7c+)@)sEv`JGTTi}qJOKETwA@)c{wwMrZR`8GdL`#me0X5#e9@|W@T=e-~O zRBYOOUOl_nCQo*p zG)^}#mb55NOxJF3s|OAD*0@!%+Kl9(*WV@_j{PF*NPhp{syELxf#cqr%H?Dmk2Mo$ zCp8QRg%k^JvwSM5m%&XSdS0p^U6e*QvxfAI&5}+gO_oBm2mDUDtPAc7z2X`do-Mpa zE|5+pS^_6ZJPqe{ii7hZ3E%?BlN7n{QNeD3%zSr3&q1mhUSQ9y!*)?O1dwO_9Xl=_ z&tM%OBZ&(6A3t@s?99wXJK?cg#7ZCMF3rIihT6~v(xu9iQ#Xhlqr5&|4{Hau2pDVk zI&%Dt*6^d2t)+Mynz7O>YQHpH-AD8xD@0Y|I(mnE)!a_D);~V5Uv5X`(Z)rvWt~AZ z*f^X0F5@I8sCHIbV~L6vL{3=25T-R}Y7E__5U2N1Ro{1Ve zpX4SmE*|+||9q1+n#TM*kL;UBJ~dPNYALpcA_u9crGi$Mdgtx(7R9nzKr`2%H1-1R zXBLN-sViI(UrrW2d{v9d!bo}9qUcg3IBU0gmn~?4j8A18eI(7RIih^t z1wr}L1QCAfBRchpi7O-G7Q{|$j>z@NroT{q0wKV&-c@3IiZr{v0;5O#Q+Fo6{oyWW z6L#vpQuSA|l*bk^4E0Sc?H13=`jcb;autuf*0Z{lZn6q#3cK zRh1)lBbKZg1dvDN`Q+0 zda0sRSS&612D)Wd{;PH+se`Ce;p{q(zM!@sZJGQAlQVm1_-c_({ZOt07KBT3(=~CE z`bgsBongoxYnxoM6^J8ms-qgU3BuFYZ*8}vHQ{0=@z`~1g#@L~wQNC?x>Id3m6`-t za0P92r7^zx=?Ites!nwVY(!WfnnPckYuf_b^t7LG(I^1&jM?mW>VAh}ttCLLJlSOv zGdylES4sEW&Uwoxwu@7JD5@)W3h4{I<(w+XqD$xK73Jb3;n9-R@yg@j{$%a< z^qw^nr#*nbT46~+ZJ;HyNuil)@R&g4pmewN=vT39;pkTfouVgc+v(mQ>?#3WsNqzz zC+Z4V`H01E`2QE>U68Nb1x?=^uRCyuiAgFyYmM!2juTAyD~{)<{mAXh-jDlZ0HeXb zhP7Fp(1K7e?ow5_#0$2&!3LtJnBK`epc59{gH8s4{jUA4o2T9u-Bco1g~Rud(W2K5 z0Da~ZoY4Eye>MGMiQHuWsQ&h~CbE{t_RlUeA(uw6kY<-cMPUcYOuAp#2)2(lL3G=- zC^VUA4*hJ-r*p95ZLD~mZ-vYKpk$G$)#Xy9&TTxX^+t)1XRfY2PI5xXgg>wPtzaw! zCj8w;i4_8qTK_%5D&P(7WX``*uejn;A$l)9Fh`N(z18E0Dsgg*s9n5`YX8=wH_pC( z@6~IO>7oz*`}aC^y=sZ?5!DejoYdR{!=+SM?^+G*yJJX!vm=DjDA8qZ*w_P37ISQf zzc#~?gq(|g!(Tt6H7<3mgx5nsXwK|H*($eQ;hJ%)q_BAOwm`z-mT~#`o7aMJEMq_~ z2qdiZN4eo{$WW|sxgfm&Y~;P-)xxbaw7C*vf5;qKujqyz$&L^URu+D9>da z*W9#RHZHvb7WdgxT318##|pPuI^|ywd11wck7Zbx>WnG(O@d@*@iZhoN|Ry#_3=Rdl_0!0AQ+{_h<{y*SNR zZ`X*MEz4o^O4~A-9j~!28b(M^q^cLc;`I}g4f}%=)Enf}FImd-Z8N*JgJLgJq?L*S zR%o%FzAMKzuf4!hljF5~o=#(`vgrZhns#wKxf+rM0Y8)$y-dH`0ZqI>nW|N<*0L&5ECr4(es{srB0nI&jI@B zL8qO;HGU1s*1+2^T{e8qgF1`PrK^}U-vo%$=xHcOPY`8kFd1sl3I97pGTaW49*<`! zI5s2iiBe+EBdb`oj=b4_`{|%1AH~y}3V+>!#0p0RwBCeeGJ=PP}(NeI?vwN0F zCOOG8r6h~TezlLyk3pCUG!KbSPG7kVmVUaq|u3Cb)k15>TtvJ&fG5L zc6q13-e?&kMxIEQv4XeJuZNW-^UxLay$6#_CC>$e^K7Fztno_w_J z(LcFCb&^^fPi?7d=BC!bRw`%HyqkgoA>jyOyPRC*}KSV^s<;srN-S5vF0^THli zj8I31_Jmzk#*U8z#a1{x*)_%o8rI=3(q_^Z{F{INu}KLz9vS#O(#T^ah}UyEi@i}mK*e0&u|RqY?1+>K8)cj zMs+JESiE%kY|S=HK1SZc!V1G;*@9x9jX%=Xzj`qeQfpP&!rsv4h#jw9f|h+}$Oqa3 zueL>MTO#WtB4qk8{EX)!`L%+me)mx7e^kGlAS!RbbX^mTV5c z6M6^u)W)gfuf@vQh6szB#^EC;L`+D@Sd?t(C3`{akZ1VMFbea9{1Jov16ci+%%CDV z{(7fv7#jBtXF`DY^;wq9DW3M7@;FD;O2djzAm;k{YnS9Xg64>HL5g4p-O3z$Id%Sk zQw$r+G{fH|qFCl)$YuFHpE|m6LB05dIDSewv~OH2*)YS=_^=TTkGGh|2M(4h)+NfR zNh?0rm|SJYqW-#`6uvMur42bqeC#?=d^*J>e~Etoz!)^^-nh?0~&d!|O55MmiWf_Fnn^FK1a2zrG+_ zkQJKPi}DSjLyj=;%H9xV8LEbRl&Ok58Rms_GDUPn;3H3Tk7{ukGf~p$<=$J|kCVJv zNC4W*KJskDPN9drdOTJs9np@>K7biJKrCnY_}_lEakfe7eDGmsIoa{TNS(uGQs*Oz zh2GVLR8%sUKKYRy4DMR-frWw+mpb~CXT2C%EUHw?X78Z$d~#>zhvtV~ncqv!ipr=v zidg^KQv7^Slr8L#?hzfK3t`vm3i?yGZ;nTXtc=fnXlm)y>jP`$UntH=F@r5Z zbl0Uf6sfwdDsu(-v$nhC(3og9=!9YEUR9+xl4$7EFaacLV##dI>`n%Gi*Yp1r&FFL zOAn5vT9|%UOrJ`bogSPe?1DCOwxiFk*cxd|Gas`VbiewhXtRe2TTO43EF>vBjvJPm zp(U4M!FRiziW(5#@vNfP3O6aPh3eFILSrSTl>NeZQ4gI#cQ7%eTzpq`n8c8K^j(#9 zXegJ-0(<;76>jO8%i_XUezQ}sO=0_T+WdTt7mjuHIewkZms?koM~L20KmF#7A6XWo z^VSh#rAi(nxW;aNN#3{NZARyQF3`)O5NY;Ys!R}J*)ck6ib6Y?n?YScw?(@pEzGi6 zO9LDRkD>J%Elh^)chp}_FnMXlHdbVM zsxKx2VJ=8PB7i<*Ts@U1jjJZ1N)-kIgH3Ax>Z8uZlTBeNv$qu zt2pwF<$x9Qy%|4^xXSh!3~YNs+fdoxh6|@K`Z9WhUOLDYlY=iu0&5 zfqEc0+9|;{7XZ#sR4Ac11C`ih=Ef@_W)q_-u1&X+2_WQ9{_H~8rOc2|{6FDK0r zMRY&V>!b(l3`_G#RphES2;;&pf`dO*aYnJhXM-?TnC^d3W_KYm(rVa>gJGMCBlW7s zKhe72IM@CCw=Cn*Rsy;=+_wwM766qM)?J+l>ojn*rP2D3`v3pJ4Lxb=4|)Xe1b=IK ze)&c3u{TXx$mu!fCy;YI)@1LSX&l!m_6kKhsHl6SnKhCY+ylb3c2G5sfI|&v3jncQ z;BJMow9b%Dbu1|WGQm5lWabOcLn3V*eNlkD-avy5$qgwK#8Srp3OtIv@KR_?TS07(p)6joR2fV4 zhbIXWLwlKQxTPgSdF6(WP)@{uaN#6q0UV?JemRzYi7}j&Fuo6KMfGMxg0oxnRsog;m&#}v`w=u~b|oSX^~2=`0g z+AEXvP^^!HA4#$bmp<1dSr&8->11{Y;sq^=WT?7M4!FYniZ9@^1Ti=8{YTBTcfah3 zdTHKzrjDdTfvt&)(vh|9L+|(qe{Dg;2)8+Wh!Y}S(yp%irKKzQX_cA3Temg8bsqG+ zZBkc33^CU~3EG)LoDy1YYIWZCAZW0~`9aVK8;m2D#ul^;z1xob9hP%jM@ymXrT>gH zZB*Z_{qbH>&Ml_RV;lRtnF{p@#U7=|VJd1LS;J;x8`}lfU85y&3o53kfn9Yhtef6E zxnc%1=O7hfr=*R}cjoOC776Z!_5roR9$}k-cp=t*X+RqNfW8fmoJ@MJU&V|ZNXj}S zY>sGSO8wf!m};@bJB=>#yWrmxQl~l**2C1vn#)phrSP?_wL2vnflSvpUpoXsjjfo#^^EZrD4U&)L^7yi-Sb^D>&bE##RP)(HL<}5_Z$Af%rW+kX~5f(x61jqT}i#>MF?C7<6h-UJYqdUxl9) zFFPjEVgd*TXVL0FlHUd;+8Mmnl)m`Ki?b|AV61ShoM#tLUN*6tc_hqHJP6wAp6;@I z;t|#2$(xy;navT)CRQnO1b1du3o50{CtiuXBp*-#<1BE~R*825gZul7GGDtC*&G37 z&uuZ{ZU9`R+f;~PKX{G%PTzO{_t9N_P6@qW&qyk0j6u4U`L z_fv9^#|B%g+0J%~Vn3nCQ6%MV5Epr2ZS=;k76oWaJp0`eMA}^;38F^2k6=U}>+X}8 zOX9)+Fy*jf7Ed%vw+k`t0&iS+BA^@cz3SHZI+Ie4&XI@XoX?h3KaT)C`br_3}f62wD;zgEkMdutiB#cHJd9qxGeBk zuH^)dm;Sr)Up9TsByM)i6x=3NBSl1ctPNc>6G8P9dx9dzsHmcl3?Z6fCxunM-O}FB zJQ-r9URt<78v~t(_K^F2{R|G33o4tB6dK4X8G|?37#-G^lmn+GcgXTM0OiK9LSOwaI(wi$3|mIJ?mu+VM@y zCEZ51jiSFWSC~Cl(2RTmXT({D*_7PQFw39^muM-Et$!=A8!WqPQr{4*4{Qe&6*B`H zlt!vf`>I~HERv0Qv1KZ|UXz+K?Yd>E*9(FQD;aB*(uNs2b-bV`G-YO&#+aUt*~pcF z+ugMBvL1SibEA8=vdMkh^e;RiOC8hK+sOW4>^{GWxi%NtG>J}mcGHf-kd0UwSz2iZ}S+=f8WIsPMV= z7atQmez)h)TC#m2sWyAAMHIW2BD<)l^x$Nszyqk#k@(L*s(!$)aQ^y9w}AQSQ?yFg zdM_2MpOnmeOxH`{<7E%NW#YRAPG z>)UAlc8>im9Gc+;*@FI@_x&h86KZCs{VkDXzc51OkQw+(C>BU4_EJ%Il^cW=fd#V5 z;v}fdtyezs>7h?SkO{Vw4{%f3>^*cvQ^e_r3K!#bcw7c4=*9A;`90*ayw<;7i9xI? zdg$spx>(gG-}{Q4`(M}-|3lcd$KS#Q(O(k{0Vn1AN62z+z~QlNoM{G}4HUbcA_-I! z%DCImj2n@x60Fed%~MBvgv%vbDY=T=H(xkbn++gv92xs>jz)U|FRRn=X@{k z->Sc-Q!n#hKCfN89c^0oPNp{q>aR6>{SrjW{P&09xy6Eq?nt_gqvHY$+u`s;O@6Xy zjytjrhY|cYl)UAQk1P`qUJy5ll{|{7e`*FNM{~qoCB%E&SBv_B+Jd_2Yjf@soqA}j zRa)^1*1UGuA1dqZzD zD0_mpVMlfb+}wzclnEvkRkOYH&*YI4C*bi|Mcr(sqOPIXl@wV)MPc|xdwKyTK#gkz z*?i%ij|cPX$Lmwf`OR0p>`c7-NS&-_?kQOn6tpe-`XzbK+#t(d$S)P(K6C=mOaa&ra1N zkpr>D30lKYW;}crIL3DFTg5HMC=*y7xji~b240wrWs4bD;wg3&MPjL_zbuNT?nY`) zY97js~4zeq@v;jkF%S@ z_i4J^pujX&PzMXG^KJ_41uD8^<{pG;H_nO?X)FDg_p+Im%yG=<$G%w;!taZ6x*4mNB;!Babgm4WAc7njqV?2QTOe+%C%Ghrq} zyrGT6b5lZiY_Q~;AxuZHV8Lv{QV+~+X;5NwUZ|Ro^|`L|&XO21V{@SZ78}z*!)Kiu z8LyBS{@{GO!p7QCjtWlJCbQ%Cd;eSYC(G_bT-011?=x0<6(KxMzlJ5qag!2(=Cr~^ zuY!o1PnG|Q;4@@%@L{j(K@QYFpI$pI(0KZ8$Bq1U;-mkXZjv3EHE{>Y)``H=HC(Lk zqFBhY$f2TgrX{+i(f8P8o;r1{{~C7CuUQ4w%nkK9eCK?QM@>WuWZXI5~YcP70XbcrVSo|)N@ zV!KJv9C2Bms>u*u7VmYbcg=+yr|X&vK+0aAx)01P8wLA(b~+86yDaXfbNr$uTSeK< zU790yWoNg|t{qv(JvC|#2BR701RYA^d-^S@Nuo+;#smFR}jakVZapZez=Mu}@AuH)P=O8ODT9O~yFHB~tR7+W46kg?)C)=Y+5J6crwlW_V z|H05xvHgj{Ha>LIIQ-=#xB2^4{pzz#-Ii_fc&n>gX@Qp~_{9s-=v<$?X^Vjs2dM{4r5!nRqjQWxi!N6v@lbaPYW$5H zgq49GLp3@A5j&!3k?10ssTI@$HDRvL!TDQ)u}*ZYcLT8C6bKGGHv|qkrHk4EHwnu8 z(u1+@&^z1x8tFr1yC3{4_<`?e8(VX$P)>^9j+BJ> zFK8FG@tW2?B&2$-2BCQ{K>{Q|MB(Xb-?Y-=dJtgXT{$hu%t4v z!Wfbx>G3_F*%gAjL5Ihkd8^#CNFS0$f1(EG#w2Z~=7u}^zjW$x(`{@7$CC|VoY27u z9Q@U5M8EnZccMvRMCjkslEvKiEB^k=NSc}8OQhI%imU<>=9jTIQi&%PBtjbz+=*_u z|6g=~w$T+gPIw;nH9u&*zy8baTtiYnJ&ggTEZ{SMs!`zsX!%^|*FkFw1TAa_)X?0O z?>CS-R!ko^^Ov(Q8z)PRn=o6Q()@cjlgxO#mApd|dF*`MVFAeHd9fT=`t#f z&Jx}N%JV)*aNX*DpVW|QFx^ip_2L7nE$%zRHc8q;py*O{YuafbhD;ZYPbCF+jEs*# z&j}+`t?!Z_TVjMu`T&oUT&*A~HX@F>O4`L`^kNAT^R)&Z3#$j4<;Ik45dF!absD6v z0m44NwxGp^PKiiIIUd?JUlCOx$R?KcSda}iEIllpVz*Kxm5N$EZxvG!ST9WsO_!(3 zdsLR6jpV(5?O-5xR8`=yP4RhXlK0>9v*q@-^44}auzATZEK^Nh5HVsUW9HP?F3HnD zRcT?sHs@xAQ z5u+2YD`dB#d_jrB#LK8e$`c#SVIF9|AAa|JOI!Ep+Fu?UEmjQk0~)0Agnk?Fzetgo z9*kk)9(i&wrYmByVysI${5KvV&txfZgTOHJ!oi2uCEtmd_cARDZ+Pp&w^B+2Y>7d* z#JQ;~hgOzlbSAj%!IN1&J31UfPeu~AJXx)37mMxLyMp2c*wWR|(;eBx4UL1K ziI9W9!VMV1g~HIWV-R8;7$Z~yPpJJ#_Npl+Bgf~r74MKu+#Dl3R$xoabgUf|3w@=x zQBjK}r({cB&KBMXcyiPfQ7OGot{CdIs^{eS7|++qbn0zWs=c5(3%vaGVx1aD)fgOc z5TuNN7z?l*11aOIzq|SKfU(qQ@i^vVrA7+_GPnUHd23@Nt@3W#k%2+$OJ@rptl#ap zGj;tB|J{<8hPNAy6?)#};A*d2ADtS_*QElzxLv$Sb6k2$nm#UOmHFD(vd-h<1hM#C zoJ^7K`)@3^tf;V3ihxNZyFwoQDutxWy6KbOES7dUYj;5NzE+oBrhHxweL}rWFrM$6 zjVoZj5Xd`8sLz*?tyS;1tmjvYaB`jDXsrKw6$H`h?MNFf+#QM1> zX6w|am1k8Bg@f&jKV-i?`Rhm!SQi%~*t1k$j&JH~CVTdwFyl>9$>X5i1v3lvB*h-5 z$Pr`czBjaNK_~No-sF1~{9%{|W@uY`Re&9ikc0}mB*C`Q3SGWh6q^}6kf0f^x~j|- zV9zvciG~xT$#A&VAEXRql=U)LU#KkwAM2LMYv$DWC5Aqr``s!7F(`OMRViKU+Z2Kc zXOKjONnq`Cfk(SI$$K=R0QFMbUlJc$b|vMl>;VtQu&~lbicO+O zB2@KBHVcgveK)?5rKz6U8H zFJE?K=nV`2T9h{B{>ve*lBwD6YiMS?XL5rWbL0A5#}gYIkm0c4%18*~hK%ct|8dOH zF~UWU<*}#CiW?;cYS_~Pv^n%n$$_vtiX(;$VjW~Tw~lA5#<+5}VrZmwcI4QfYNM}w z+2b-Tic@rnC+Z%ZOswRPOxy341MOXtyt6b}9yg#NQHdvVoULWLnLJr16E8>)ZcxTb zvJBl>@rHaVpUP(b9vc^N5+M`f{yCnu^rCSQAv`vPtvCZQAzPbg^G}7}FdzKjPT@Q9p^Ex|tbt+N8Jb_CUd)Pnwxu2<&kfnz*n=3tPap)?NShl=S77urwTzHr%*c&yY~NkYLS%WOJ@cE)zDxKs7J_Pc6hM8`pNWjl(|$;R4> zXUGzJ0DOcwG|S&iwd~T!1uQ&HE3wjSvi$W+^8CmSra7WsTq8OX_8@4z=#(VKrCLzo zg2^Kff<97L3$QuT_^B{lBqCsBgmrYxP)7W=fw?9P=DmuDYLd?_>4?Y5?J+Zbr<`KT zC{jX2ReL@1Y*EzsRfx2hIG-oOz&B>CEeYtM!6DmDXKH|Gen~*OtOHb&ez(iPnE#pw zMaPiphVGdMg5A;<#daZF+M;Ncf?`}CSUDe+r5$KJ~k zKF^RcW~auCsZFv&?=dF$Y)te&M>cSS508aQu^D`_DHh^}=~NVQZG8N4xAcbmrc$p+ znb{n%kE~%=OiX+^)jN4ct2A~ZbQXm3wSsPGBYi>I9C6qW=>!jGYW(_K;{^^DIim;4 zn6dE83G|bG@$MTYp#1mg)E|(KUl^dYnE~Zfimj)}2`G1`R=@hE(pQn>;x@gD-A9(Y z_QRI6PrgsKP17LAHl8-t>GU%_^ZJ5%TrpjtKy?bDkrzm!=#&Pdj=R|Ah#orS8^%3t zMQ~|2w(;wsO%v$UIMgf5n9=3F-n}`ZkMs)jAWmia`VNnDQH_6$Z>FX)aQEbNQMTsV zT)crB3`dVDGjU_fhi9V5^5o>?K&P*mP}TB{YuTiDB55>Rjzbh%O_53}s#5?o&EzWh zHP9<|&_6c3O{#_XFAPDwW(cI7$#dv_Sb;S9k~|h3;D9PI3}Due8DmqinoD9D~V6LTuz8P0cH0>D+2Gj)~k-stn=sAZz*Ts4!hTphn#>*5jXAMW}3vzy*HK1$+n54 z%1q1@QY;i{=2KBCC*1~R5u2&r2~84p&a0T7%(RPpg;&G6G&*&g@T8_tkt;Y%E<+=d zT+hQ^6)xSryWBfGKBWs4jdWsIa|C+SP-b^{^pP6>13nMsEzBB6q`-1sn-Rsg{a1{3-TcmH#%UmcQS)nmUluc19j@=h@Fyc zY-M16a0OeeI^@|cZBp(LrVFx#=l$aaIY1?uL!a`zHn)y8&VcAa&aQgr<_N?TMyD`q zVqe%^5Bvt>IWEf9)9d9pY>ds(!%b5DzP9*WhzUtj=g943FE38b_BdZcVrqS5G)X?#z3M?;- zquhKIZvq>jj;v~c+afC#2 z9`8Hg=^K{hTSKwXYpk4#+6Vdv>au$IYT*%anhZG!MBb&6Ql4A1!%6 zR|;vwzJ+4c3EU%L^R_*PTIfGSH?K%b5na4Md2FRUrzwrIDc)un(wOj7(m zmm1Hk^*sMdQ$(xFKCjJFD?(abDqOb93+Ooadko&Uo?3M7)d8n3=A=y7Ak@EnT+kHJ z7g}eU^ubrL4Qf-w+68fwmbo7%T_MGCyy?^Lul-t6#1`jI!i--%e8ZHm|$aM8#3 zS?9KFa{7#l8BGy8-10;Fg4)HgY=vuvtlxEsq%3Tk2It80IWDz!zB!^vuvBm_^axZ% z+z+dQ`kP-l6mU;N(B={AfEz(XQJ2#WXyhgks{Z!1CbE{tAlzjpwbCdSYVK30s5mAL z^1tq?dYQG}OTjW*3KlmcTC4#RJ&Q)B{MfXVr*uWjpAT52(fYM zF&xHIoEW6AmAu@+?|=L?7py^7_S zaG_H-23}O`aH|AfbxbWyo{L!*nB7$y+Cy&4+bg&yNb*Kr-FVqfI>{SHz}9ze2q{&J z*)jO6#Ty$~o;CB>&UyT}Zp!N>VY6)g_kKzazA(b3)lAr&qS#L;a+Hcf>%2s@jcpJw zTC@m1^vqnUG<@h$rYdmfI?whp*Z`|bgLM04A#mNI3j`?5;5^20<7J)dW&T;$HF6+$z+m1`{-F}Z+($C|Dc zuFMio++d)y85&M^Rox6wBc{>Iy`e#ufd(-dHXbBF#K`A3i5eThF>FJyf-*Y~H_w`N zYJy2K(QGgMGkN61+cFDaZ6CgmtfAPI6j?zoucJ(%kEWA^HA|vBv|QD)dhMa+O^m@YX56GHL9eELY;S&C}%d%qUQP>lA|>X z+#cvE#sN%$Fb?Jm&{_P6I(0IdHHLuXT^CebwUy#EK+)W!HtMq1=4#uReBn_bElLJ{ z-Z)5!h@RKyin6Gk#!HMLD-2{l{^~T}9_Yw+B4GbFFt}Ao{_6QhNu^mkgGG?dWCU@kmr64RG8y$ss@UaYQZgM9`~rSj>xLJZ*Bbh zV!5jFuXyDPMUN{GsD?DFw7E00gPYZx!*42+nGG|##>b)eM7Y=>!%REWlY7SF=x^`m zw0z67OBGFi_zP0UV|T9JY?nGqu}u_dprURnF|^$~v)o-e#pKcVATyCBj9v95inc=XEVoNAe zL`7wWq_b7zy?fFMaKWU@e{cNU6!G~#Y=8Uyzo-97Tl#~oKN$GI>4;<|#lK!WbX_(b zOD@VP0@sli$O=O!S3wxQBJg9{c*T9_M6iyeg=P@FxL%s$moB#vN6%kXdn4`n-)P$v zbC&H3wCtS5TOSOhPz(zSJnn^dfm8_Wdt=+GaAd)#KXSL!w*(s-b&SUE=15%abq7xdW-Z zO@fEa!TIq57?I}#)n<kBDX2^Jk$a!JV*zB7>=;=cwHGKxoQD%q zsJ8as2UwCtJnfF*u@Y*965?YvC$uSOP`XccoZOpx-8GKc=hLamRo_tnSF6V$eOZpa zl1lH2>5fIo@Fn7inBliN^w>ILjw{~(X|QQ$QN6F-O7`$r**IagvmB(@a*C8uQMLZq z6!ESmLG+NhF4d{`2=3Eorh*F$>QS%)DIF?RZUlTPs`WVU-yCs|45)Avv!aSseR4yx z2{_JL1BdRz>yRC5FG!zSe?(D@-*wF`RSzkLg1>W$i+1vzVkE7q$l25yjIef3Oxo?AV{R(1)&d-D=zwwvn0#yfXjWC>#ofa(UKxBJQp9h zhg_65M-+MPnua+jYrQiyHU2qty;LvO3vV#j<~VR;8XH(fSYJLn6gqy+CW2LDq>XTrgx(WN2!Lfm&qJ)2Yvp<$-!| zRUT7a5L5eLQg3wU&6SPnZq_>rgKv>g0R?FfFf+V#uc!UE&YIyDlr zYLAm$AuWnkZcra`*h>#It=%+c+}>2KVIKt*KpvzPLwSbIX_o>!CfCT2waV}j6A~PS z#T#7C zkSYTk#Z~u1q5;(bKb;20It5u!#Q&N2x(Lg*w<++fPJNb6g$@@ua^4q@n75W%PV*fM!SOnJuX~Tu z_+7}0lry_sjPnnuE(IPE;Vi}(pO54*W71rGw)N9JYY$p_O#K>Ux7-jP^B67 zN^MbSGJ`JO#L!%yJ>e}-P6Fvs$;>fDba;i!%81SC=IEi`OnB~16 zSos`jkB<=!&t2BhUSM5BW&bkS-_kqtf@IcN@!OONKbIQX8I)_1fb$GnxMjHRpNR>V zIW#g&FEgZHI=Uv?Y$>_o#_VPXe)#>T+a;D%p1d`dtdMiw_AK^@A?4yr@-M#L6nO~@ zqB!P|aHH%Fy^iGgWeAJC8qn`62ReGqVb?4gQw(y1m2|r_U3m@S!~Mc|*f2Wh^@a{O z9rSN?A8^VNZc#@|(!B?qZZYd%bicGnIN;RbXUDFhTmkXCjl*6Tu`WnPa?LDEnAPQM z(sk;-QuSA|^o7xNz~~$1$VsBuM2f_Nt`o2ha^5z%HAg)3JS5y1maS>@s)sa%a&goA z=W7%8S|!swwt~pcU;EtYZZ6x_;5!71*VdJhOm48?vD#8&28&{fEuhG5bPk|n;0WCk zn9Qu1w%FyAY`{S2JK%&Ij%hMX>%{hit^X`@oIH6y-5u%xPhn20c884YLgeTY7ML)998B_AZct|hprdJ7&yD4dY zCOr8JNHc9vZgGAPbk)ETRVcV9Ku_vEvOe&Xe?LdA$AKasmKvsCY@pcn6iJ|>y2wT5p#Mj5Y~s1ryOi0YY*v>t zHU8Po*XG6x+U>s?Jp0k|b*VW?vX{QUN&3?$6EN-vzON$fFN{7CZAK2$L$RMz|?#qn3$b2yWTlL zv{O>((JJi*7x8wt0zn5Hw~LWUXyFcekMKl5MM$~$9=!zU&MN8dfNds)YoNP* z*RY5E5Bo#K3Q`eaIf_ox2jtlJMGO1oYl4w8J3UythD{F!c2BqgeveL#j%Q5s{$SDL zp&Hp)`Ts6@|3<(Xnf3@d$8MZ;#19TmPrl_GE76t; z+F}ekwTtztl|)b91!;$O?qaj()6PbgbZrqSBDmpcw?n~fy{bs6O;x5s)30nAh(6H8 z)1olI0r;z@W(ZGC9fY1f@v>toVCsXZ0JXxh1vi7>d+{J%mh6(FI0yLBHo>n!tuUKj z4oQw39{P|*piahu1>8%^{gE^ayDH->X|jE)TeQ{){bc4I{Gd{BN*b%32A!JJyanGM z<0d4G#RB7Tww4tOtX-V)L-hS%PDQoX!5Nj(Zs`{1SpPEWI#^NL=pFQW=X$ZuyT@a@ z{D99#@*W6G94B$m%MyiPJ~W=|p;tIBo!T@l7XrCQ=tH7yitDc1U9iQ1-R%Y2x7BCc zV9(yXxt|UGdhaqz21YK&w1~l$zVRub+WDiv1!i^dv+}EAc9R)ohQ+{pQ*!+JpWeU0S!fb3y%{I1s z6x&OYTU1mXcp46~yQS!H*zUI(b_lF@gLaAS;%;ScXwA%4*!9i?Li`cCC|9_|d&dLQ z5MG02vbSBZYi>S;+DunSG;)zQsc|};x+!o&cr#4bqUdBYezx^h?H(0&QAKZ&c9mPL zCdcIhsqsryRQax)Qo%y?EPm)`OdoWb24x>K5^5K3m!~`Dhc5FU^vR+tSe-_j%v_rb zWW(4OpdBJ052q}TKo57H>DT%C@es zvGfDtuSbKlnOTF6`vLVbYgnxHu8ZsyuY{6qbYmfTc?SeMt_W|_F`!mg29_#-S)|f$ z8`~(|4rFP`43dX4+Rhr}UNxw$F;~*!RBdRpFJdAlZIesy-biy2< zmiH+B|fJ9jM16 zUSLDJ?q@vOr%%}r1W(UBmJ#d1WrSe+gtAj7-?gLx;BBskmBM};ya&@Hn)sGVXjQRD&@ zg_-OpH4o)C+(CXcK^-0vQ5hY!NHEjP8u?-dosgOQI9uS0wB(psQT4OTnP4I56%9)4 zjgSp4>dcu)0(?{17PQ|NDOrbZ&DCVl%luLE!M!dXfVIrO)~~?hq7 zpSp0ifsGmrum&KW-s^`MtLXwff}{_~RxR-~BrA9u=Y6h#vKclb%b8J+J`Zo)B#+0^ z*w4Oh0@>kTUige0;xWkD%|O;fu?-YCNkx6INKfA<6)qTd$e|yKYC{K|@Yq-Zd?&PG z`puvs%}&WhL61DiyZ#@*?*b`4;?}_1p~u*Ghxr)qDg~J#{rSp zTl~3bdc3OR46DcA`@92gu=>39)fmg>308U(ps|t;lrb@GJ#@XYj>haNY$}2Aq8?zl zTH%86&UV`PdZs(~DQth3&-+I`>-;Dd;#t$#?Hs3t;I9=Mlkv4aOG;TQT&xYsc5%8# z76d30M3sSiUj>Fm3|fOVVyb#-Q+BG`>EwVb%=kcK6l$m4mjBrkaRLW_!$I=H|Ms$n zX;*4`t7IWb;qk6iYPKuoQfxLwwo_3p3Jg(ck&Uxd0gVJvjad(_-a#jf5p;OmnU|{6 z;$b3m%}ETy@B?-#kCya`kCV-6RF5jfw}5@mVVE9;gfr7xo(;v(pXCJN312JkwKvV|)(;B~n{`p?menQ(V?S(vGTsen5cJl|reG$y>cho0N zi@;w=vT^msSN_8!N#-rq&mpZmmLw0%BuO{LUZ+SGI(8v{Jrvz}|C=9b2yO0xeuXG3 z@~-F zTiw%v_Or`n;lhPk8Z@e+CHKk3h>a0Cbw8c?O0@>sONj@aFm`o|oE79iv-CQ;Q5lOj z413AAzX)c+$j%?0yjGlEgbB^lJWp9xY;j4^;IG0)tkic^iW5Y;{kHgF;Gs#OSEMSA zN{W3_*aX=M_XhD{NheTclrF$5uIPE#y?L=<71Uz1DbMq`-!RfW75~-UY?*cZTgYxS^%@M`IEA#cL`_K@xIbsLBQ-RmqoRX%#GQWq`(qTQzpurP^D&ZsNNt5^)`r;`>~UJ#$vmKO z21pr|Ui3hE{xJ~Ojt*S5?^~k;csV}1kt@WmCMPt7beG|%1lHRQtTdzhF|9R=8|@^XRWS-@YA*<>@tk-SoPj{BN<>y4M>bDqL2-{^z>L=7_s*biY>Y zRY#wD4F}J?W*Dsdhn=q?O;sJ;Vi>-P!=NY&kJx+vvma=lEVw<6mhRJj_sh;@oT3Vlx}))`iUTI!(3W^a9xAv6l_@ZQLo@2H_eE z1LlJjUE_xatTut&3z>4z3gdyWC^S=}gHR(>u56ugOOft8pgK+tXf_9*UVt6jI~jw1 zAvy)5ppO`sImsInJ@@);Q{Wf1(<`w524-_&VyE^H0lqHCEjG0NQR5Z3Wf-oBmYiO2 zFI1Z*!#_o#D}pf<80N)U4G9&f;v@~&%WZ*q+Yao}(gn9=Cgf#^H?)y> z9(#xK&9)>R#X>K$O~`Gar^{WM)TqorE!0v`C$l9u9%7C=6^-;dDTFRpd!F~N^=}K* zBD)S|dX0yyvPCpKrTaQwGELxy+~B~VrIvBb7sQ=oB^-Mr0L!rI75&1SzNaKPF4cmj zh{~_s{r2gIjk2Nfb}?R+ElgF^%6f(4fu>PJ+$d@s&@^E}+Pl&%mJUDOdTgw?@Q@xw zPoICKTZt)qhal{ZWtOmmVhKyUEH-#46a|+mwO66DU6UI7Be&Do&u$@hszZ-&fneqQ zOwE4ZY{>o7#*zX-F0cx&lT?b(;j5?X$Q^OL5bdPOKs1VmyM0H!NTU)kPe~k$fl>dO zCpU5vFdv40USqifaS<^5jl6)lG~CPcc)!B9MaA^OfLys2gP^rjE2bxg>cR^H2Axnb zGeqG~nM$6EQ8)}njEjI74V<1Vy5o+V$;=PkOe)ITt>hh&$YT`+jO1aCy$p(l_CTAd zsLkO=)G?BS{-7pT`D5b7=l^g!5{p`lf~Umu3k3=W2qMwgxiC+bC$W>~C2}TDpfqk-D+7yxNpUxt~+kLq1 zbVL=si{uFI%-V$~*JUzoJu@(Q#pVakk-iWV)xxme0d$yIj{oR^;i3H1Aw51Ugp? zT4*_r`G~*Es_Jr?NlX5+7gR+{)_d-kH~D@pISc8+eUQ{;{ux40E|=zP5O%5_iu=Qh zLQBFrRb{R@z_5}=Kav}Hl0VkOF*wJkUYitIzSV}+w3h^UJdZu#@pv*j9$1IR2yQ{D zGcis0WheDN286fC!P&xVBwhx>LexFy7JX&@!>{b0zK-mktPfcnd=H}I#?M=(t^;9k z)On+1i(i*&n_}(ci#~X#LwB~ZK*f&8cxozcQ2A`;o2M+B%2*XeBVmoV9!g@R1x>o3 zMe&6yPgbhPht!GNqQfD`aC%LeDAJ41Lf@(b&YQg}!iqF2CJvoF&h~&3RUEd}uMyHs zfowv)MG-sEf#4bov>d5p*JVj^Edx+oz`|pb+e+(Aq|Yh}HKZdOGv>$1rnY>mZ9#

    *x!vH%Fgu!{r)q(I9qUSg7h zwmZh}aMjA8k7_zdC0!%#^FUdu2T=xfacmlMW7>m2SlkMkGr;}PsRF_VuVfK)#@0y@ zI7NaR0<05ml-`5lO(^0k2;x&Ui*5w z|0RNsqfutAA2S4R4$PCZdmRWwd;otn1mB@6eskcqDeV(4C~q@8KENMHFDK9}k^x_= zOwC~^t45+8h||Zp2T9Jt2 zE}6EA&i`)xSN{8>*UqWChuItT6kyT|U&8Q(nqr*BzZmty~?KO@lN#@kI% ztBo^qnPTfHQcLM@qjODhX=1C&ylcK8ZGsKp#!_4GKTLF@qDIvnaU9C29?JDC>PBf>RIL)IPn9P`(7S|%40yPU zen>i1`9l3Jx<+L_xDLATs|8KIy&gG1O};Hmz7PpPdVG@vSpT~=7$n5=WTo^zXnAV# zbq;vWVdVuzkozU40LRS?w*2B-d0|#$6`2Z`kQ^Q(>!b~`4pJ=8I2TenU@R0=LhXwY z=C3U(Lx5Fsg{n`;Q@`_|Th~4HwAg|ugQf!1i0TDJ>|%+LzAVU)Wl2_ec4!hjGf9Rh zkuBDi%Wlb!`aMzJW!DBjWfJ|{q@7Gj!1}Oos%Q?-cg<{_{DCx$DUVF{x+eV$_9?lVbl@n$`h7!ieI)P{C6L9E;@A~_7!9sb z?Gxzlcy|DmkU^8krcKZz5s|)0*g;yI7ZY8fq#?uu$*lOj8#dVEsJZ7Bwa^Mwsi zu#4rrxsq~m4n+QUM|a4pX}q2Vh+C?xiL`9i(fN0785(Xr99$m_d>jnvlGBdilRj;J zn*HXQw?2@AU4MtXBJ7a7JaSE-IeUPDyg(&=->*?R_R2a=Bfur|>ii&%5io>h7~eiw zy6xYs2ISOt9t4vX9J})@JveEV0I2O+puWx+(dU6pM@`~C zo0QNO=2bNnqC^ETE0y|GjVW}o$b_pEGCSnL@&eTU?d-OI9Qt8QpT|2LN<(mBbgyc$ z=t9`1jQ$#HP~*Xe!4L?<&Ns|V{XeLa5P*IFy@rKo9SE*r7dBqdCyVf>W_uN?o<%*0 zs;2h^fCR=BP`1$TmSdJPo30@G1L zW;;)LVbpomoIa0w!Lg}d48n(Tx0V#2wC~O4Qia3cXceqp15*&G6q`bkBn!tVh^%44 zvnT{ON02ISiR3V5H*t>Lx>yY|9!}e?F{TIKzi`8Q_M(jr)DFAZQ|k@} zX&cnxV_^O|bb(@fX=7wX7}BR0V4JGXg^f*oM4lXrm<|fNp$hk`v}hiDrSse;WX0hn z+q#d|PlIO7MJKo|5WJ0({7@J4j>x)lg>U};X|kNhD_5qCQP@DS>nV~<>CR{pqc2S? zW{@{VQy03%`{KOI^k0lJD4bcUY*1A?lgx1uDsUK1h;`x59d0OYh}HkwQ`O6^$d~l> zoAUhVWKpZ?aBz+4de}10QlHM)c4oO}z4mfwy7zWjjq0NI8MDMU*RM$MacJ-Cu~|Sk zA;e5u9n02#?;lCUOSWrnv@s_aC>A=}-luepbdAWOy|>JB zvE+er7ciJT2wv{l>#@u;fi$YJga>KMrxbE@_C|^2_?G`>^;WOuImcv|KCj|AbiD;P1>i^aySSJOoSX9C&-TE##B;dejk z-mDX%%`-PB(iMrUlPGg;&Yii-S-d(kmMib!hS#m(_s#}dot&Ru7N76g$Bhtj7A$ra%KFij@UR_=8(6;g({}>YR5N@>fS4qzD2J`;U^FHsh{Qu>$4q&~!9C<&cRCEAI zGJ}eRRe=wqP7));eXyXn9y&*`ZUI-yR=$ot4PE)IbcW<3!7;fm4#TZ@y+*NHRO1vv zy^&$YdC*~p8SYosW{kIBCTU4TE$MS}UH>-A3NtGyW(7r-QVFYolFlH<0^3Xvhz9m} zwK==X#QG`if9D7y`mY_!pR5oup84rtGi_C1U#MM>9m(1j2__i(+^!QHG@z`BG|U`f z@xgGX@Wq2k#P>fTFFtVx?0GKxR}!QC&MKrt-Th${m%lSbfbcA~HG z3^tAc!Pet=cDM6hv9lK~`+k{j3l26}K-b-q5jz|(mzgxqJ)!yU=J?a1GT3bcVFzS= zKO^V-K9=v29rjJ-)C7zPh|Dlx2PNl0$981D@qfN#TlVj=+gdx!cwOYG_@6iAiPS?G z^v2U&6{0#rZCR1qS9j}5a~9> zU;Or^*;mEbc6izMg7BcY%BLGjZ+CDD1J(tu39j*e61);fT$5%V6qm=IQgxC%ZgYGw zcUyF7S$ENL~KZ6c|zSol&@NmM*Bx*cV?%OWQkf=Z7 zo+Dd`VM@v*$O`1kt3cys0VxxjZ)$~tyaj?n0mzV=8)wsE@_bt1P_{^oeUOiNwcsW7 zcw%Sdz1NF@sDbT5HzEKIVY}=eVgI{1fJVsq_w{3iIh+Svq&m%o8fLS$(NV(SMOy zrYK#0aaK01zBRoS+G3Fo)1mHEmft*HJ}Rgf#^*lo7W-!Dhi|`n`CSWMK3JLj8oA}d z0k_3gb(h@~^N=DPR6@OQP2^s2i#z@_X_sSVwkBxFLAI5 zAAns4=3mt;`wg|Dtwig;MA=T?Zb0 z?Hi(@Qk~IWpVZB^7^m;o{%9vT;KIh~qSeB6f?__R$PsLpEKuxn2d#4NbPt^*m1C;B zMY%m7ZSZJj`rK}E&qA%J!9v!60e^6PWJd6FuoG?tgIawW#m>;8SNq%^aUMmr(pr&G za-C#(^|_&ONSD`v0jim`E1)vEHVj)TQQ5W`#z$3^IvQg_TU7eU17VAm!v)_k;AGdK z8S=z1Va5thdCnBmU% zFsB1cvJ|YvHni96Oc)*9Y@D z01dI?o{3OCw^(TwJ0Dv}z=S!9is8-e@b6-UigAfQnJ2bY0d-ka*wOWD<=6A~#V%De z2Ng_d3GR_+FsY)Xh^2~}fMVd6>fwG8QVk^uh2lNjGcoz$sgYHZIyxidpPz)7FPf7o zO66SQAn{bE3=bJNYlG(CZ5qY0iP1~FrV_e6YbB_xoG0#eGj$Vg_TMLaEXJS~beSG;$Bw-*tq))~1B1ROryXE5 z1FjqMI=gUwsI6{*%i0t>N&uLUt`}AVU6p!)yU{;gjt8Kl5PxVCc*y)(J(`x`3xUNB zv}5OrumbJ4fBkl>*0!jYO^U#EQE#t7)m+@r_Ey_$^&;P6P${=LsMfF657(A4Dry)2 zD~CZe;GVH~`RGrDe|gc1YH<=tjH)%*W(BBAURJ!mRK3IdhC$3&u>EGvuplhX}_AWeK-XKkd zU`rNfk>LJB&7_kc*p6X0V+blMEf8Fsl?D#AMu9wLH>YBbGf{v_ZcPg05wG-LHe+?j z>X1T_i9Wi-=RV{nvSRaw%*L( zD4PqfOLjb*LID^jVDVTR2eDLDhxnrz_2LDBE0Dh_;;w&ni>loBuCGRAZ?^1wStB~k z3IQYfBz8C3w@&jiwC8NS8JG2R?X(EZkE{qS<&=BW`0nARaz2x<^abgWnDfF6;2a#w z**DFJqYz*>fg``#J_Lpewe9B=>(H7Lv* zpwFNT1xj4asPmNtutP${&U@|1N zCoARcay90NFOkz>J@P_Px$hm%vvF9YV*Zmj5g9HwGe-Y%-$KFmQ4Put{N~xuzl~uT zIM^t@W@S#sEzr6B=tWmGG9lz1Sva$rrwhl;4gdzXFeMxikjV8a;oR^+jw;w=*M=3* zm%@6yfG=)hHz^dPPn_$lO|*GNM*@;j#KfSXSb=0bHD{*Gmh9;z3CG!?hC&_SR^%pJ zKzcko`6wFPBIzLeLM(}hCa3iS44NuDzy?M`4&bDN*8Y<04sj#9Q`W|J~?($d%!u|?b-&t7O2*)7WqDw}$js}Di4j!teJh$-HcbWF<> z9iR7YJC~@CpvZ`D+5AZ`;*}(j0Gp2kGd|B^)8H) z-Bvi!Q4H9K9N=jOEg%gro%VRv`J9i@C~^WX3(*fj=M+O@&5CsS{?P421HrXEHv>ep zpaj-e2mVV9XE!NUvahyxY$p1z8a7g#oCIQpl<|L1NiN$qF|%1}T(~*JPJ6QsgmKIb z&Wn7nh|To5W3+|ohmQHDBLK3#V{G2--MQ{Jf3X;kw{M&#r1PcO@updo_AeaIB#tLf zsDvU|YOoOYMpPdNk9F`iG8bn(7S{l;l#yTQUmJFs+y@S-3&Kur|G9MehBqDqdzF^c zYAT_{JI{y;;7B$91b1OXrGH0k0jD7{$*(#763O6HMy~YjkY5$wk|sfp@iW;`*%9~0 z-upt#qnht1;4F@B19^e2m^wO#dtTNdESZg;@nAa2^j#M2i*5C4;GS|XquZi2s((Bd zx539>6JKQJPcnZH{(?9Z4s%_uQ}xQK$VTSO*Kv@Cp626tq~uLJQq4a_wsBmx%D-pU z>l?i-DEzmYtn*|II~2Nb=pR%X2Q(k$QjC@&SyaL{@&3>rD5N{*w@sNY@8zuw!k_1Z zQlXJZeHd!DvBny+0EM6iq2&}ofTmu0bB43JjI4}}Beu%Q9kRT!8pH{2e$=S^o^5o_ zWrHtv;%z2gg=4CX?&?(}N!bT-9t0|-0Bd-o=k?ea-y+vh^uZcWLZvejj zPVok4X}u!;C>X+YMt+;D-1jv9g75+dTEj9JEa$sDO`Ry%z+l^2BXB>du@B_B92ouM zFu<42EtqwlbDcmJYd#!;#Bk(BvWka;tJg&8pzEX04M*yZx*nyYJ4jXFFu*rFFb!o< zzUY_@QXF{83QXhvR-CA}<$$t7wS?{jw&GOs)H+%hfP$yP_}U!!7d+<@N51&y4B?d1 z?7n!HW!l<#_>FgL&CB321Q*T$*)ch{zgG{84CvotV&EXD;T@Ifg}rVy0abx|VWr1> zK|Q?JC_agLV!Eo8zYZ9?HHt)mURX^do2wQCq|vmp(qzz*wWm#+V!+h449s?R0480LjYL3 zYif=N*|!%2<%l#2RCHYzvlG~&GB_2nJ2@@V27VrGlx&=W=da2)g{@X;R2Qdi_pS}Y zD!2kcV#H&48*s*GIbFaqwI?oP>Ur_zpo+L2?h+2(hffVj=XN9@m6tBZ{3Xykbjp`-PQg{H zc%$ymjUH$m#)Cr_+6Rr#_U7Nx{h)a}=r~#6!dV80u?z^htfrWi6j?zf-0VtmcAJg<{EsY`UIbK|kRoPuR%RL8xU9=)&meCeBvXs_>)q z6JJ!8XpV1@ubYm#BCRBG@`EWgu+P~Tx---{*l`FQU+hyj1u^!~F@zPNZvA5Ur9W5@ z^T&o?hm()lMgLvcQ(0z}gL+Ic&|T6+B`lLw!Jb#0N^V8prk`@Ia!F2g$VT$syP^+vB+a^}ug3r^a%%7%FsZRox6<&^o#+Yy_ zB<-;KFX1ekQ5$x_^a?7#4#P^CZpu>VXnkZ=cs)c9jX*{NdIq%GOc8^1$694Fm5fgV z?Xt2hexCAXkS|xTR0WSD`ijPBY(WjxP z6QaZh#kT3?;n0Ndbxm0bo~(g+#KA&~Y`RHNC_+LXoJdnG`91e50mf+>MPpnM49o7Hix;7XAKTqb-mTy3fufJ6#wur>r1TO);?O?V}P71a}i{#I{(d z{^mXQJuB5l93eU4+rHX}z5X9hKkKUtZHG$mt5YjHH`Cg{J5$=?7w}ffm(IKe(Y=h2 zEp!ieTdbCMPO?3)EU+Ci->rPm1ODejS*~ic_m)|w>Ey8H_-o=;K7LLIy;-%+?@m#%In%_GTcFCA)aTY9%?io|I?8>j?a?M@W_xr^0M2H& zX-Sz21FOLbScfPEVt$oW!kWl@_imEIg@pJ1P)(559TAY^w?wJlO;VUrAEZ@Hm*0b+ zuNr$=N@vd(>{C7V-X0BoEoNNY@z^f@lq`-fP^C{SqBcnGMwLakfkfMO;3>kA3w5ce zD<+lK%Eymtm$%XCURhdf0rcfUtJH{C&zy=P2hd{g5_TxssXw17>34nZ3*RHFUYgAs zXl)IM8fqy9()}4`CF@?Kl?J#wJF zI}lbVLJE;iY5Q1rYCAsrBbdrz+-l#Pw9fo>o^1-@B`H6*lW;KAZ|&iJDl83IMdn(+ z8v_F|aQ0jkE?%;o24Y!HO5L<4$&wc=8JTCbYOJT2wG>IC5{&dAt`Ue^4*FwF-L}vo zpq;{X0g0Q<vH*d3<1(yIVSkf-5Dbwo7+29L zsy2)dF*XVu077oJug0!Dtk?w4ix`V1J3vN2CJEy31l7(kB%;5U527B&c>+ z%UMMdXQ~SYH^Ck?(rK}{dof>wSM<4I@4k_KAWVT8^Iq9r_p+E>Q((Xapf%DUH3w2g zWw~+)Y%F6YBSsnwFjc?xpe?V*OX8Z^VF5A22NKm%2+vD1fd98iF%R2t4RSL~4Dvdk zzR-CM{IY-+Nmoe9tMklCNUi)uVR^u_SVmVy7K+SKv{Avv;CUQ5S_Y3f8b^M#^G9g` z7Gv{-Ilh!=CXfSGPiZm5Y@KYmh)LAs|2p6Zpms8=y1w+J@Rkb8A&xD;S>!oGT#D6sQ zVpy)&+%aJ93@}g)x+|Zqg9_nn`j)g_n#uvqwGh4VEYlo+p3^}OzWQkF9#D%ss?hSb zi5uK^`t0{!88rOoWi%d&FIdZlZHY+8gxxPXF-W_#M{c2?YLOtPlsYG{mA^I!`AzX0 z5y2;2?tq z6f}ypujJ7vkGqt2LuPf^HF1%NOrw-@g#!yYWYx6FrE;~IVFM2)ag2OCgvA0_i7adj97y@H1D=h70o|GoN_)F990W(ukyQ&<+$%EvO_Hmvt8 z@wx1Q;;=oQS`kEpTO>wVpIaM^hf^55${^q5Zj4ora?9lCBHA84*&ktg;M;_Z08~eIot)y!rNx6#FmRp*tK#ibJD~-bMD2 z^0RGU&6#8|CoQj?f1j+KKuWEi+AS0V-@FDD91et}%Q0Nq>y|@S^H7`tb2DnJKs*R) z3_wB|ds1=FJr7P~aLi9F21HT()Y(3!!{5B^@-li~@{l9h|K^9v%((2TbN4TQze8DI z^2F=ODbnl4>R;x%p(V1(3hk*Bvz#KyR6;g=LV<1hm610@%i`1-9A_hVNccEHt~&IM zPCokQUnpM}SOD>U*~Gt+6)#OhuFwh)ngaUd@lXTQ?v5E&*qS{jqvGlC|4e>&>w6Y>?0v^eO760QhYJ_CEwyS% zdqgqa6nRJ`U^?eG)X=IkgF5I|em=c4PK~=Lq$4a8RnFQ0uBx7&&%F(T0GKaQ-{q}| z#5!*%=~om(SJ-FVbEHCK^hf#ddy_Jt+y!@ThUo3DHbMM0pEf|*i$RWe;I6KbMr}|0 zP++dw%&3o(oIo6>n!XZzC3u^ll{PU~b$eDQ@PqQ;cC4W??GZ6|k}khMx;@u|sL^Hd zmhg(`Jh}?foqcY%qy-%HRxf?1sZFjys-q9fsweijWq75_x4=#mKOZ-R%b~x`WjoRy z{M1J`6655;=(A&-76>$|MX#2IwA0BR&@c#pw4!#pGpfgPw+#9Tu}r&-Zh_1s&hc1T zC$`3EtO}<-{q(<5Y zinXv+f^GDEXc{SxJ0)K3H6}=44Fhn|pDrX$d|}^#eeEmazu4+dut}u4aJmtckp|T0 zZ>AWiui8W4Idqb!@&K?Jfwe4LZQWGb)tWGcHrNBUq`3Zqpv$LFaY(HlZ;k<-K| zF)c;XhOwyG9u9oQ0zk*s{BD(PpP8L@t{sy$2en9$u5AlBOqP*j@Dvul+7)ZogsGR6 z39pIq8%34Pj@Ug?^&*~GWCocKp{fBCIB#;1%3W=0GcNbV3k}es)#C~52w_X5eQwy3 zzS8>w2kIb5vNB0|X!1_wQt2t8PEt0%b`5BKsyG|NmU}FFtik9?IGvgMp%*Re=wq z44h7G;jAk%XD&}80K+QwR zWL*e0m-V@&$F7@!ukLy$g3k*#WO%hJG^%6)B*E@_qaUnBNjBVA^|>vZm`!hlJ~SQu zDReO+xkpU^+T#cQWk5!2j80*&z&M?|jML{fr_?78c25n`9kT7w)sp!z64oL2xvdRq zr?=1t`11wbq(O@Nu@gWspA%>n05cV=rOO+=8$)y9U*WP_Csyzqw=nWwm)llIuyK@J z*vGb0Pl3{1W$;H`D>)OO#)gyzsLujf?8!;ec}Ut><86>TS+{GTO<`-s2Hxw;*YE#; zv(2`~YVaP*g&k!(f=s2N10kD2_l8w^^hJC=HCJrlG3!ubjRGgn{J`PuqhJXCeEcSIufeu` z%`S|F7{G;Xi=Ae-6Ttg?an^@_DW3fhSbi(RYDHzU|59OkQzqIjZjo%7;%ozAHC^cx zA-{X?kLo|a^ZQ@@>0R++idjgJ#HTjH>czkto&0*mAOD$ZD=gr$Cc}>6!P&UtXcP1I zGX>DGoZ6=|O8)K1aMZDj)oIqQ%ju$Y%JNuy}; z-V}6;E}TC0FmeP#U@OKDJ?4+{Shp?Q-4aKsd%gNYlJwH#AVC~-KooZ^#iUUr6>^Zm zJt~yMf}-AZIW|U{cD|b?rA%#O7vUR?xpEK|V~&h{2Km}AmtV1fEol^Bkky4>orW@_zAI!n7W$bYjt+8tF`RbSvigq2WghbaB}r z=c)6Ag1vO0etSSKg#XP&1buFp0Xh(T57G*c=u_gpD2<|Bcyoq2U#X+>m1QxVvRl%9 zAQO8X@(f1$p%ilu5K22X2{~(<6(6~G#4cBjZ0&Iz7gnC9BX34z&sg~)<0@K?Ca5f{ z&as3K*O1beX4R>;T6GRmOce!fy9u500!|(12;cB|FdemG)eYXOc-^4AhkfrBmJU6A zORV*%WIM*YZZTw%B`|;rNJW{|m+*$tnAm zW{CLexbvqxe`OnWaoISWojA&IVPEKFpTf!NCdF~)9H@=!=xqYfBvFivs&-hD;3z7F zHTVc_Vuy-XH@vs`*Zpf!)`CECmED@;!nLIS>YV5Ub zX4Z?aIt;H;-{;jslW^55AWFj99Avqw**8fFmu?Q4i=!Um&c*s|eIy=kqx%`IQj~B> z|0iw}T!GdF5FpWmJjwx82T2nh0QgG}hHZngQ5+h7FY~`573v+H`PYu()3N1g#8*8< z^tsueOq^{&-1lpLw38fg;d-fyRv0=#F&|On2$g_^{{<>6fB|MP`XEqqZ}jaZ`xRSM zYlDtW+eIJpHj~_Ke-(vgI%3;Y1*(7Sjrb76Buw|LjzIO8WC{xsOk9Z#(iPldVOC%} z{XsrV^gVt%y+Cj*OcSJr+wh*OKvb*#06V|6pk$MxY*wU#t8I)%WS~~fvx$HDO+V+w z7y;xtzs2Cw>^JX)!{C?tl6mqKfqrXSkFiO^5%HlaU*yCPOp{n^{*Dw`9T-C zol(dwR4vO_<#1~`)lMz68!#X20W#obC%$9_5bC$@{XK4+#onlL_xzPSadX+eO_FJ4 zZ&D~`8AXcsy6E z&{Iz_Cn<7_O6c~)o-&Lz1A@$KCT0>ujPGEzI+j(c*K_6zdc3OOfDt6uK9OSAn-QF# zy`F1DrlvKlk;DTvQ$F_W_AF53(Jhg=+|Y7L+keC&4 z7V)6ij!usN29x4R*Pz}8wADDfLn~)(j~?i(I5y|c2jGa^6!QR%{W?#tI#lw2*lw*} z?u0gK*IJXbsE3cD1P|7>z#>r7sJe{sFda+yagWgDsp;?;Vv6F0n!KmJk zTbYweirGhzJrLAWro<*rFYqXsivMpBmxdcjY4`%=4Vl3tkobg`%2VV1N=J7QED&GI zSqr6>y8=4pc%(_OR#XuO`qPRAzIy0!aBrnipusw+Xc%V5&UTU=MF|TFV*3%pdT)R6 zT5-#BkCz2MzrIZ7khGU(p#lN(0R<5`6a%DZnc(KcXqD>%3xl%(90-^Fp(R|P7b@Vi z!QLlNwJQK!`DrnoG3CDbPA}7ic^w{lh6M;#H*Kk2D)X>l%C zT?D8P3rYvd3*4HRy^g9GYSgh`DnB@7%-m7O@G;Q27|v|7zlRk*#{c8?+P`^Q@bPao zS?9?b7sf}46+Ut)22`oCsDvEPP80RYgP3N{*|>5Uu&rQE-wF?G48+aoa#EyJUxz&6 zX^6DeaIkE_yhC+9SlFY)f*AG_&*~0VC>i(r3#Zz?TzoBQ+;+Nx&hpVf)2RBIPnHj+ zj__bMT>vYRIwRyhJzszUl24?GZ;a9M#2!qnFfr_lKb?K(ueOEUE*p`zQ<~i(na?vy zQh<}Ch+Y>{t7--I(&b(a(sX&{*s+E6&w!&7-GBX9{bXe|#&=b_EwkOK3@(y;@iru| z5jo5~$ls{gAKOh(UA}{tM>oe`m+uK}<(JXPoEpwv@feu|)66>x7t_)*R#<`0u@9Hm z@~?kjyH$Bf;-_}>(6Dj${mUiG%CokPU~q5b0Isg zx`flhpuSWU0T&~2;#|}R(I~3I_KMRdI?Z}GHu(d0E!MA{1q=J8Vu)Ss_#;0&H!r+D z2=;;D0NMJ|1i_A20c$VC?52S4IRR-6c7N@vxIL;yf@D(*NS2S0zag$U9&HISf0e}` z#ne3s9>tP^)2e!Dmj7wd$%)6I$N(vyF9!d#~vMm}Z z&<4V8G(mUf&W`uOEYK-!FIhlz6F|;sVBBF3#q6R;2?`KmiC7!GgL^`Vv4-=}mpz~g zY%Rnhp(#g*(tp@|EU>aa&qrg|-{(%j^C449LWJ%F^ck?sMd zv`+387-1Es9!!HSd=#Jb(frDXS)M%POtE^lUAMgO^HX6DzG6YnFK#zxl6@|`J~dlG z`WVI3QRFZ}T2HTkbpa1y-A?BMXWccQY$%T`i`mLa5v#ynLsb1fQ=E}mWy2i0l; z_&Q&CFf5g)=hnNo)2)(vPyCb-&~1z^$_JVm)9WD-?8rfOC||Io=76UKOcmewP9s_V z(yU+wR?y3&m<<%nS;8W!@2g9w0?v{NXWf%0Bu4BGZjRT}C>`0r+2wN+hvp_Z?23y&9!Mbvd^uPiP0G(6+W zf&c7u*yr*yL8G|RJm(+js$VOsN7Qwj_jF&KD;`OMyzBnUh#lgcH4{-+2)0znr;ghy zbFBy}ikq3;#tz#EeHQqB&XY2_r@wJT@L_~yQMvUmcS}jZOS7ne`!XOkSV1wc0WYHx z)UGQVujEyF)PU_RLu;HfM7Z*>ciUsC`(VY6{ z%@!+YJa4d3JfE65|FsvthMt3FnCW zSj{@C8ONu86le8|hGOR^ef2j#GmfMf+I6@5#7nlO~v4p@QY^!X?%QNga4K4HM>TcTd zbG;LcpdO) zW@=@P{%z5PfS8p*jlzv#ozAYWA3jW+h7i-hDc^i++#fAQC3$IT9BF^S8upf3X!CJMkO>zvB?+cMWC%NQGmi!8ij#*#8F=sw(`^En7Lj6EIO!(fzncV3c7shZsB8b zy8JQnM&2U@!JmXI_AHAr$kn((HeYkf@3GiChPix6-5@=|xgM1<=@ikZc0?2kQfH=# zDtTJZJ5k73(k-b9)C=*QMgd!-om_()i?ub110hv``=KWom@aj|OqeP%HvpP{SNtf0 z{A%Q3UY}e4hoNHyzZNnR;-EUZ(R~4M4(e$GgW6iU@QqVmK*^TzSKnGP&4R!8E8}ZO z5xYgrg&pN%R*TvJih)(GoJwfrw?*#~x6-=sCBOpKMVd|WDLMW?iV|kVV%v6Q&-(Fl3w+k6g*K0=qbou8pV>IeHqlLl_b%Q*cz5wC9YF;G|k6kBsy>Q>YQKeC!?Q)(cWSY$3 z17*Zm_`(6!rTF^|4uu^wfU9=}8CTsS~yw6X{1DW-`c4Oq|A32fIHoNmvqY1zOESSZ5S%TsQB zWHiCUKGmC|%wmPFJEq(F0jZ$Mh{z zu*}K4u;6=5P_YmR{99=>Flx(i6V>JsTYU6d~% zJ(Bk=3$l*<=F-RH@C4FoHGxeO(?F3^R04j->GCWe6fN&_%cGG7;xlQn5PS5Nd)<(M z&xi6%c%ncx2gS`hNDK2QYEBst0T|>J(dtxk3rfzwnGMJCGZRm!E_{V|9(3?Dc!Mw? zDfCd|jH+=;mwVN!bRH)Z8IyLv&_%OXa@C!{G>fMeQ1ku=O7qQEY*8JXgi5V_Zsoqq zW-Re`Y~2{^*@H=UXad>Ip38D&`T6&1zg*0f*nWmCZ$D9Fe`5OqK6IvX8g!9x%(O|n z0+5vuBt3oa$5eXkCiSugzgB)W&<(1uiHqnm;Vse%y^qaIqCkC->6Aa=^*}L2CpU#r zvN^^Z=uL2+5AjadU4*tiGp;-I;$3qw}z#?U53KD4cB1h=D~kEyfVG;`a|4$ z=}q?wrh8HQ1S{y(dnAYbQ4yzLRwt~UYPI_AtaTn*e|WJm^o);LkvltNfI)U0@F1q1 z0f?Q|owX?Np{-i^3zZqOBdu>YSwy5^2w$MeAl912H1OK^m=WbJ01TP88^?;Yn(1K?QZ?)0pk4EuF^Cu>Q12O&7l&I38b5x z^fEI=;lg$lR!Soc(+vM%5U4<`h6cQ-w&(&e961n{II~yQBKe5l5esh{fMlwhn8S~H zZrkYRNJa}OL(TR`j@d`by>h=>N()kCe^Y0X9WK1u0nz0ER8&#S0g6;m2_>+T0M67x z0c`qi_|*7a60VuJ&J&}9nkY-xx)9c_jdCZ12%rboAT9Ok0E%OF2_&MC=P~Rj;sexh*998A1$=lx-}y~ z3f^lUEkpHX{`jjue)r22&o-}Jc{J`q+dy`_GE#j?^%$gJmwTDkIy?gMNnni*PPr`9 z(T8P~UgtyAXW#6aRuN|a9GwD9gVm8&gVd|SE8<#d*ry@`#o0GC3X~v4r>k9l7b2c0 zaDQ=@Lu=G?=iDh^d+w(`|AM#XDj(VMh}fa)Kz9Ij)UqSYnIm)s7KP#VOf{rkH2$(w8*y7htlR98PEKoS$w6a2(qhPjOX$HM#^{LOlm=&0*LcQjktp@c=lJT&k zSZ$6Gt&P)1?hf4@s!?1X!#>J^v*!SQ23|g9uh(3YhDj_|M0j8L9$EE*?SsIO46rO( ziUA;GPzh_f>GG6Gi8CKVup9S3havK^YYK?a5ab)do%I&g?P+b$EWQ+^Xdyg@88LGs|60&-MB+qs zc1TrNK3C6AVG5?S1grJj3e}USfjiHu4W~F(Z1E*2BK$9=&r+JLFuq^|tYc zGyZJbBR4pF>%s*$cDQ$O=V~Ui@>GoR-roJG)M7b~efN7;$m*BIa_qFS9GfWy81Obx3CPlH z=4Wn^Xho*V&1Oj77e}KEL|K5kTWF9sd8dhzqz$mo!To%TJY8N9tQC#+j%LuL4QqJ; zAPpL5ln>0?F8JU@-{|g!gmk@U4KH!#7H~q3Dl|}&lP=%xSHe*nfl{@MKv>K)*^+&) z^|{qS((z36vKgid4GeZ&3@f1*y_ytkuEN11ZK?;7PVvsrQSYl9E()Gnl>Qf^i~{>c zWGIh9HGMYuav2t^Waxpk&%w~Mg0qmWh8hM0l#YJvndp}(UoPJxg#au!-4T+_jyoQEWSBCJH)Cc!GSp&E>`-j4vU`- z{5H=nVP$@(-}YvF5Nz>l{;=clN|HN))L1zLWfTJ*!ge44dv#MJ7ED6HP@A~ibJy!n zV7}Cgz`d+^eO{;h60H76MYlNimaN!)k>LJB%_NQDw)>q(c*M5}Hp4TokGvH1>{Jpo zmB7w*)1tmSWS-W?lz?F%~if>LCs@s&M%1Gg(Lcxtl)f# zVn8MBD3wqRZJ0*dBztxwG!+bxa1kwQ)*cl)5eI7PLL>^zP*9!k0=9^HQOzs!;I2zYLz4#0afBl*g*fpLBLo zy#ysu@I3wsmWruiEpVfXbh|c*J@DBfU8nrr3|l_&Y^!-oqnK2RET6S6}#DXw$!pIO8cO z^;!5^+bEvP3Ypl6*y(62elo}_{h#n*f-;wuiN%HKl%bDANjs%o|JAm60Y%fZmm`pWDfiYZ) z^*m1Ry0`wm*vx?Soz8v7ZU$U-wfUM)N``Hjt;?DYJ0-Q*qLUGuxMiZn@rS~i;`PGM z$j$h(@pbe$Wnu8*__~NXQU6)s6b6ETW;j<(*Mn-skpP2y^D8w0&NU}&AZ)XCI2FUJ zAWU_Y?F_L+Y> z8xOh-ppgl5&)i)Bmp!Usk-s2ppgU#7vf7wSoHSlj{H<@^d27X+`@V4&;vh>Vl*btY z(juGVn>d)O0gVVBtdz!B2u3^}!34GokAx-ywZc(ZSxku!Qn%lp3i&;i(Qg&6URp=v z{YOI&`!-4O-d!@JDcVeW$rGC5i+_~%^OfJu{pR7fcm6EtjjEr1^xLC9KJ&(d z*(ZL8uhjKFO?e}aTKLV$a?z=0#_sMK6wfxOH|GJrK|G}|eUwre^uXnvu@*}*h z<6E7xtKY!eUO+a7A;Pn+>qD+$x0xLfGe4>SKgwX^BUjr~|6@Fopvi$6_5+ z@2au?q)DNp_einHfK$t>5fms(*Co?Ef(EOD-yXkAfkE5F5cB5}<< zdkFn>_hZUrjst82U~$BX*k0o>uGqO(zYggBFI$RHmvzML@C#q(tYmaC=D&d^5J@uY&!h)KIcLQQSqEnM*lQBR1j&~*WR}n9PMYobn=M{8=MtZyMbPcc|?(J z;L6}7LerUEcrdI%dR>m?SPPhyag`p+W-RbZ4Cr%fklu)@L)JG)eQOk0w25tMKuFRw zxf{6BZcjCE)EckF(_174Ig%Kwkr+-N`Y;N$ZVhs+@}|5JDAzDrfm@3QP`~kUXzz6N zWQw@zv`IOnYQ`odQm`SX2{!uc;WBd%`HiS%MZSA0)Q9%Dou|9S$;viaMI0~?XnB`N zlR~4q&N;%p3$>>8vV|e)R1sbSoOF=!tMsq*N1v+C?Yexksz7A`!Kilnl(1fQOIi<9 zPA|tsVEq3q6qdi}p+y9iDPMvr2^pLvysXewylOftG+$Nbvn%!taH6MjkbblaJZ4t@ z>`2Umg9d8A@-vc8bNpUb25Q^_>AP9BrNu70Y}qLmKFu{KjIykVHeg%(%t9&AUx}kp zHPYSA$tKvH5*Em?JBBk(utUcAq}j0}=}K_nJ(ZoVgxWA%3UJ?)<&!AL=hkrU!os%u zTNw1h2DvPsJc!(5d{RYfZ3t77#rl}5Fkvz;6a!-kV;$vq$`>!;YmjVNBE#V{i3p z;r#d${+IwTv+9|zb`EYVQ$8~IH~}~Pzn9$x>aBFxVOzJ=Wp})GTvTjOIx4;-JfN!d zC={%M@LZu_ZP45@-~#HE=_8SP6KL$ma4}&A1LsYS9Wr->}F@kXR$}3yT>9Q zzz$OVOOh=BM*sbff7MyZwJbGmA;9sz=mlG9K+<3Up+`E!tft6HDgnzi^HiUN7=g^H zMRHQnfU+%XgL1u~Zb0JHHDstca(woOy4!ij>>HHVzw_I#*zRWRYyuv_O|1JG>+YtNUIGOAI z{vZlB&|o~}=tvZ_sj33E0i`!|zGCwcG7D-Hk2u99!a+ks*VxA09XA=ysp7YT;xFK` zZ2U!Om2DTJUB7h*%)79^VW%0h)aT@sb7Y-l5BIas=J?a9#h!b7E`kK*#aXQ+Yf850 zo)C&l;tldg+;dR8vO4?{r;I#^0VX>Bf$6~PFk|_Qt3i8ISW{XJ@in9o+AY)3sMR$3 z&Mq4~^}8QK{Imy)VITMY%hzVkus9&u@?RE`-Ry#^E{vKpR+grgVh&QIib~LOu)?4@ z{t_{OZpsark$;@jiYh&H;?w+GaV_^!SdF4N{u3#5pYXb3s>ngUJ~W$t#Ko;KzSbzp z<50FTm2=O>e62pTUV2O1;0s*Kn8zEB`JaK(Dsg9)KmkRDVgPTx(oDm$SFmnCRrfu=)6g!J7?BzDSbT)+;BNN++;_}?;ut)hS%gu?ft>+$x`ki0b!FL@ z3>RK{>^Qkavv0^hcnBVfmea~l6=?;{P*~j*eGfWNdS+(%0P}O~n4tnEg$*+7Orlc` z7w(+1(>8QwC+4(3urC@CrC5T8ItZ7A?OuIu`bcaMTH)Qmhn>VsBox(1_CnsYLHY@F zH`H+2=)y7Jcw_VJK70C{zbAf4cEWyv?%uX~kS-f~x1;d5TUH8{?#(f|UZ1~JNAKWX zmmg5AV8)t{()0{FhE}IM?q4`4E3{H8zn#8ZY;iec|MRsbvXY(4;lhCvphp-GY0ajX zO%&M(wIZ;108zH^5eEew7lri&qqJEE*trae8auR5zv?JRt+mqTgJTqZbqr4f>>j-Q zLLiLN$=PWY-~DoJRId5dgW%F$hxYi*+SlfuF?*M!jtI(H=WKW!6sY&I<}3?e(JJOl#v;9xgI z{~v?AM_$8misaaH=F(xrJZGHIK3(+g@o1IB@i><7;Tlrv!c}I_OFqDrI7l&76gfa8 z+~uu_#F;Mf*&cn)`++}d5pCd9K*Y06cG}mh5>@YB5qFnYD5{rMLxF0&d-u#_L0fcI zXrbU!qURzv1N52mx;@vb+URAkVpkLr@E-|Cl6qX`~ zVe!;_IPzlTAhB;qhRDE-k6=oE`mO~;AFND%jof1Ay}Gb*S!~q~*iA7HDbhhD;7SCc z5D*^Aqs`J|h0vOJRSaADYd%>%=(d=mtZOD-_P`FmP43MKbqS}1X$9{78$LU@S`N0W zqc$GI=xGpm&?>b`)CkrDr6`j<)Tk+psDv8nkaRg7RO5?lU%MjSz`%J_%WI=iAOc%^ za6C+#m&GhmVzrzalLW_9DM~z8?z=WfeFD1q`o6Yf(m8@EH9IC{i#8JAOwcGcf@BG{ zHmcz_lNu#epj-lHZh=B$Hhn^YHCAhbP6%DL5%8N^jT+nDQadf2`?0GHmv_npmUeDn7UD+gHxIZ?vT0{Qh~iP z)CJC_TY>-EWoM3^iE&x(smTf|`$g|9%?#?G@AI;u=XjI*F5#Jg25yU_K-?UURv=NZ zS6mjK8i~orbh$prNvwoJELp7A#v#UV#vxX~rM}_8xcgbwM&qogt0dEf9jpo~2WuO} zY^9*^C1JlRPhgbS5p{vUQn7_Dmr3U}U5=`+eQrBF+EnOqn^ayKy-(AR#V8S}zD{)E zQ0Nb<)kDwuu*~cx)%;QLn+DG-YqSjENDUr!D3_`g_c(g`q3Q(TpA}g3ylC( zBZtlxj1ZR|S7f`LciAL@oejFtzc%bxcqOkp>NI$$xju^w#s`zaD4#-(I z$LDfeNsqXi6p<2Hchr42TLdL_>B_s}_E;p3J1^65i=11b=`gUyXhG@+=|ew{leR5D zY@p)8DJMG(Lzq#z?6K2BM|Tm}trdYJC+_u1LOvt6l?@L7N7MAb9^b$u#4xoFf0jG}EadW^zm>ZH&Jidl= z!{yJ-Prq5NvbD2+0#(Di1qH76^cpQgB!9wAG6hV1}?K z&Pr9z{L$*ewp0f$>yFqVFW5EpNWev2l3zPr9{WJDYw82Zc8`<5U2v6euoj&GFx+8N_4L z<(Y~UC1xv4OAo#PPC+0+aP}6)YTnra$?8_8zWG|KD9h8b;#7R&JB?)dOS9q>Sgkmj z6tjVXSx9K}_*{m<&pHU>x1(AfFk3&va*e*ut%F9M zHKGIj&GZ>judI_>3~tFOL9wb0_P3qfl@QVI5+sK$jcph9K_tI|KBGvBHOTMy+=9Z* zHs4G|I=7j7N?ayO=T=2D$7j+=qd2%o;Ce`@%N1xXkGz3dTb8mgDwg(IsSR(LT`43O|&p%NOTSU-jP zm_4B#G%DX9b?kni+ydE5)njpsWVzSe25E;hLxSOa4QF@gvR5h@zG>g|DLfsiBh%~RFW%@B?KVBXm`%{{OnT`#Ge^(YFG z$!2<%UBNxv$KDwN%ohA(UjIcam}EhTSdAIkL%?9ERbsmu9;yWAyF{GAtPL6xbxs7+ z^T77>{7|T~Z^Xu(+##}Ujd$5a%udh#X>OV9!z`a}vPG3H&r>xs>Yg`yJS+XxpM@^+ z&88Ct$z<-_v1{9B3xc60%o)e*gJ6j2nQ?Dl`dzqfET2sv&V?m?{j6g#gX}b?T>bn11?-@aM8+XIYBWW zQRE1f(99GIvqgE}oIvG3fEvlnD?JSEr(P?Ix$BJubC~l*z8E}XroX@Ac@{V~D*Y1$ z7}M{OXYldeank4yJfw~O=8MdA2zx!7Wl2miI6zp30A+M|1~ot-upq;rFBLYr#W6f{SqgM2a+#STlq+e+$LP-SxuUl z-Jn3z#O#<=>zl?mbUE_5=w9=zNS_GSGeBCD7ENK+x0&6yM zbD%tnJ`&Z;aMz_b^*n)q3*@fVJIF<51s4}Yyd1k9FyD~EnDYy2j5cDAImj5vo{bFB zFbg9(An4o()#qn?Qj}Ojj;&J|kj7zqytGPG-RpA$aW%NL8avgqp-g$lq#SOh5-+Xc zl*wB8&7@dZO{c{|{WzvZTlse*GLn6+E@}Yurr&^F_A-!%Fb*~ z{LRVsFINg`nYUKC)TCF1K|-(<_{ihV90jV1sDXdRET+Q_OW+s}HStb3)-8$uxsSzk zH2tc4F4^e9relwl=_sHWD2U3X63SzH0}Zmj>{MOjZx_QK??DHe;y<2xX6oL6OeKm> zr^~mCD{aOY;7i3#iqm15LLExNoo)T;xlNHL9Gi5mM3ia2_HkHi?g zm0lf~O&ekRj>2^%lS?M6agf!48#$l*w(`xFc1xsBq>1Qq8C1PcVm{Xd?0S)3h`9b(c3*3&(n6>;s-KDTQ=5XzouURZHVjiS;2 zf^V7PiUJe>;TP$1EB8GY>9StV!Ed)DN-Xdb-WR?{Rqd@(0tOwDHPmuPxgKggJb}}pa~t6pGKoDVxL=yPb(jTMu-ml<2)<&%nS|7o;|zW zX|Fy_)cn|%;m&14%XT>H%;I+G^7GM`J<5H1yjJm0Sf&-aDw~-?!MW)4i71}AMOE(G zG#Z`Sj?D$z33qJNkzd7X0UICp_RT;yiv=nA{C6La#Fxf`Xsj&ADvC*=$TBLyB-Uy& zA=uEiD*!~P#<()iGRF=AXyk}sIYQdQZ+&KKF$VK@yYSLs$BLk&Au7vYNjG#)Z&4ZK zSAsiy%K5m+L#G)uYU}AT;hb~PbLO-~!{1-#Wy8J|@jv~WM0bnDsC&KoLz3je+oR1^ zHefBqq){Z5NR}W_gpo<4JxWKP^UH&p8Z%jq=@iO$V`-H^j+`}ls%$!A=1O4xPM04H zO@bX2gk13@4w%e47K`@`H(%L5yACcs)|+le-}QgUFn-SD_|1>Cz-YZrrX}SsScP7b z6T_#Sq%aNAsu^IP4g}wW znyX6x9M65g{(-@cvY6}gLV*tIba1FXw;az-c`g_AkTAWY^_Z``#knrWqre?x8V++o zXyCJr{o zqcy3Zi(=ZPJ#zIfcXJ!9Mse4BxtDs2WeWxd>7o}X5zvLA49Q36Kdq6^7hu0_9=#|m z8xq#)x^G^aeL|5jsZdlMyNcKL=C#>4Xpg*Dh>M=_8=ZV?p+f4dvY3*|#%XB6a8n26 z=rJ0_GfnE0N#H-36-3pk9Q=LmdS4fI_^8{WmU!nWb3A+Kr4zeJhGf|*Nq$-0T8~ep zx2L&m$k#q1CD%poi!A8N;jg<+QrUUnF1(`$iQxeWat*~~Qe*>_pyeeKKm@nP>zWdY zWo`u?qt)HSd~grEz1@KT(l`Df5O?dWt@q{K*Vl$h*$F%Yrt&~<=ffFiA&N?1SV zvT$E4vTW{~J#Wd^v1#|DLhGJBu~1MUzAMKSNE1{VZjh(MX3k8XIQKgAPPOuprV|h6 z(bzu&cYX*JVaj!0^>h^vE`seKkV)4&bHishNnu(blG9F)S?!)>w$dp=e)rxV)qj5H z_rLnnyW+(ZvydW*gWZcUtKG98^`AShgOJ_l$=k(nHx5$<$H{<`d@{Qsj|NZgIcE5aXh9qB=@2(dfjm0bbk0lE9LVS-u zp{!EaB`=Cu8k-rEPcPu%wPheqd^3LF)!k7~g7NBtDJ{X(V7uPWr8@uhznkNCaIvd6 zS8M$MBjdjRu%PZg_HX(ox#+^{7=(%k#OUr&OdCZERKj7^WxwY5 zR(@sVKG{l9z3~3jCTVZ*li=*o6tCk`P^AS~ND6}2Pfd$zhW5(&0`ry7iv-siOzYhd z?{@lbL_IL{?TfxA>vc<3rgD1S){729k?$^`gKZb~x)lcNh3!Ht@okRJQz3VsL4Gq{ zD_FtXDyg8>agTDY2y#T&@Of398i`}~|4yi`o-XeOCkl#%DgCZ@3 zywhM`a$HNm25s~!^SrDq;kP~x9^;3E^|sxniS;+ycERG;*H>E7fHJUcR!a0f7!a!^ z6WyRhhAh|Y^w}% z(!|eV_3CmraZ|p?<73N(*q6knFf7|LCC7aeg36^CL2)6U0?4r-_4pSY2am0WsZ z&8!pj5|8e%J7I6NYwv$VU7|8YmD{0!npqqD24~mt>V@Zcc&}P&IC8uHhTvrSGsTS= ztEoC(eng=oz4(c~0oJFii6h?f%4F5jKUy-#a&{rcz(O7`$mC&}UvVVnxFMyD^*1%* z5==fHlkSl1kRff`V>}DXhhpz%c7tv2x7mI@J9F1`cUPQcnFVK|!BA3x@f(Z)pZ4tz z-wwR6+2lsidFi@P%ofGZfdG%OU?B-ylU_v^sE%RT=9vebB>2n2yf+;v0G>V5F=y3hj=+2kHq!|#D%Dm{6` zRWHM^lWxQVe?WamKz%g+YP}?W`BPX@VJ+w!rp+9*)ZO4L`YOy8W!nilkD#_-aPcIx zT;_mCW{+}+supkZN}6jdB}Bdf#H4ah1$EjwEouM+H zi)=SL?9XIpsA5VEQ~zAeRc&f0;DTzxY>5teNB=>|Oc+<20K|U$?Y_d}tLa~xZVKsK z`DN$dsFmyjiQM(|sRA=wvw@(&Dy}19YZcJ>AUe;(3{{qT1+R%(5D)3^(z zGbdU`JfZQh2^$Y=%7ngRLKiK)+q~2g8=Re47)IM9hKp62M!i5?o$ozW7YU`GrJ$A@ zA5;b&2I}QHb@~|p_}Pv2-eB0qGU+{6j!#$apAG-rAy*&$Ra=+12)67cw!$OR*)A-Hq73n z9^r2SBpp)kg_V(Eg|kFdDAf%E0f=7siB}%9&=mR`j}|Xn1N3UfRJ*Z#uM7HJR?LW3 zb$g|}pj!g1G9XO}9e;MrR<=Ru!Q{hwW1`Tz_mk(RS?X@EiPUk}FJVyW*b!I^wO8qe zx~%O`WYIwvknc|#|K#{_vI!I8FPNn5Ym2@W|If92L2@($H3U{-BFD%-h8cpp%NP0o zy!$%OG~@jKgP~*8(rHw#nWZ}|DND~jC<3SACv3(h0sP4PhXzfu06x+C)2~5g=OSg=t;URKcz|t zYW2y11oK8e^BZ6S=TN|6O}qA{dp}t`r%zQ7Sf|ocy^?nASOf?n2dtmbr}|SSsIaUJ zX^ws`pFa=@vOdv!{r|G)^!Lw8dlk5c-8*FK3&7gnc=Q5R?+)_|@KfQRTrF8>xQc&@ zfpNy5wWIEkq27*8Z_wp2QqfNC(3x%po45x0Mcym6hCvID0%xqZM-G8EIGef~z z6u#x(4*$|J3eU!0OG*)E0zI|;dT<6 zjp_oofKUQ9_>>xov!Q`ek#rnuMe_g;qv5Rwa>;p`Ti3dLyl6yEc<|ED-wl6bV6%&<6;`Xi$7MD(Hpc zv#hywkONt%+^B4JA907;AnzTbv(L9v$6;jRr_}`Q2Lru)T5@Rr{9Jzgv|;xys@T6e zbosP)ZC_abe7xh3Ych}%o)>NL*~)8BEtX%8xDLY>Kh3zOk%#9d(W|KY=Qe;g79PjX zCsoG#XWPC~v2R?=EP+iQY#k?7FFbdhhkfg=y#BNoDircAD9go51#6UheeUqiL>!6O ztB4aV6}4-(N{b`6N$>KuvUE6X1p&i(Y%t4k3#aU6huNQgn3H_Zw5WKi>eRQW5f^A% zA73WBilA2z)G{JAn=Ekuz;$N?G-r(ekDZri-#?knGoQ?r602v<-Fe+T_bY>P(=>Ga zTd$p`)_!S%wY$wQy@jA542o zMl@vO%u?IIQe`{-Sz-ObtY`cECwk`c57zwEvaI$|?oIAah7>~$@cF17^^KY7{FR!-DfI zAhX?*{`jk>Ed`M|yEZ;$W*iJJ zvuX0d$l9RZXS;;Pt9nWNrY!Y@x@IScng_PTrVXojef&;l)L2PDV6r^ zSwa)yi}giY@d88WMy^byv;rh4f&aH&$*tuDROr z$ay>>Y(2^jIX8c_VD3LxLBfM%22!D6LZgek=7}{;6S5z+j*8XuY=w~3i=DW$sc1t2Yr6-~BhoqH2K@;rUvg?B&0>UX9c|N;VsW+eKC&_x zQ~CX6Z&IpG;)m)tx`Nw^yWbbjIUMuM>-H*Ba#=T`x$kZ>%4 zz+=^+#!pBfR>$jsB~}Af5IpQvrbuwRPvnZ+($#)RQ2?{E&+p{l-)$$*CXAzph+^2X z4`>@_{v_a^OXAO1{f2(YdH>H1;%?wPn2d;t<#V=im@O}J`e$|y zO7(Z%oaS$udVasFVGXsF-PFV3s2m7Zj9WRE5i~?>cM`GZcvV1}vM#jWy~-`mvliH+ zjwvotg^--w!mElXGc1-1cxBRC^HPBX0jH=tfuQ7MHVA4Fd&kSwxm* zhbnvanwiMiQRka2L9wiOPv~dQQXkad58Q2t{!UW>N!PyXoL;1cG*M8E_GYk$ZndvbYIq2>n_hbTqcwi-%Hf$aA$DPM+ zQhILs>SrulJU9!V(`M7mgQku90orb*?qk`26iIY8xqdoBmnn-^t@p{-)KJG2)#P%& zL-WfbPf*Q1x>lc|$gTY1$VSBpD#=UNDJ=@Fm9_fBtCoi%%jK9$V^A@KeD4N*9?$}1 zmp6S5t9FA((th`C!MojhH3tJv!aAr?VJLn|o7oxKC5%_07U)FRI?jNS8N?okm&Y7q z1*Iv+PYG5(Yw~X@ga5pO+QH#i{|9DJI!Ms_396ikjr(END?=~7{r3HzZQ9o8jp0j) zw{oTy{VL^GTYmNS+bCwe^*h}!j=FTdRR6D^zB20a`HvTU{^LKbCqDSbHsEkt4$SVT zQ;{)eiQw8x8>r)&2Ei7;cI^gr|NMSn=WQ0Eti_doJ@<0_i-*6NM@|$p55~<|=y~u3 z6Wm~h9%6dly|b2FyN_n3IUJ7W8u9%z+W~@v<{F|z8nXJltg$J%u$2e?NE}LPXouDB(-HPIDUBb&i)>N{v zTUifem-%E0h_E3^x2_v@zuT?+zgvpW<)}`h+X>~NTQmt&H3`!hMpG^?fXjh{_pS_* zk9Z1Rhn#2IG&vH?Ob#eP`>#Z?1*}fG@^9Fp2qnA=OsA$Rx4{XE@IRQPg7OT+8(0`~ zheU=nOi&yBq&chRKn*CgPv=OQL!YElX)}z>W+8`scq|^7VPq{i{lnq+^DO(%IP3c` zs10LnlOam8(`!TEsG>u5dPbSj?kVN5583)bn;>ERQm5`^1&OKBdaaKoB-pIBIXpix z0H_@pP!T|BemWq4@+}|_tIh`he*itiJOm zyms-fWe(`6vKtI}p=?PWbRea;e(HWII!>fJr0J(>$(o=G(#eE_ae!900c6P>SSe+`9Lj+H3yRb4;%3?ZVJIo2MzjTZo`ZRYs(TzT;UAMhA%cgD2JyT z1|Hiy%{9Rh&l2+H^epvfZyoq{n|8&IPid>%R{nJ4+r10>#A$weeJ24v4@Zy&f&$=!om5c3@Y;+E=R?8l z8@B0$pIBD4aW;~{P^T6@?Y3e#@}wj?@CLcrbNRF^b(>5N(i%OAO851A)LKPq;ABab zUo+(W?uG3EwS=3{*3-s29dQrnxe_CL6i0kl`}LDuLd^UjE%s2rRyt193wo<~6{Kk7 zhh}0a|8n_-d01kQr5^D;G#}V!{q04P&B(Bre4hBxY@FGDU^Vp+JA%*crA_XM`<V-@#$f_vz z`o_&ScDN($(H_X1;TcZu!?R~V*gt05uV6JG+_l%g`rW?$Pdj}|%U3T*|K6>PBW?j9 z!%ok8bFs##p4tx7)P=nB{tb|!DfBKD9E-X-uR&n1fv^=JtZ{4WjqN;-6>f>6uik#w z+hj=Eepc}|wZR%=6L;ZF?K87_g#-<*>sAnJ2K^`l(GICY9!RXD*I~1_W&ykS#WqF{SioU2`nCYH7>tcAj zwg9Ty+Cq2nkCItXvs4P1;b!51C_M9W*sZWE_3glB*=3NrgA}iDET~YGrM@8Saj%7Y z<9S=5gRIcMM7+lD@`85lJ>_Daol};L(0IgtJb8zG*kQILR-!3H;IC_cD=+~mc*A#( zQOh|DByg?A)lsh}=ye2@Lc|`|G)b2V*3C)|HFgzOg|1Ou0!6`W@(eFWT|e28SKEPM z-$1jynjJ7!#rnN%8B5{pnwcRwg5BgebLh5u?BR8LpM74}?Oi-)k0eXI543dp?5{0j zH3YWfViIDm<9>5ixS0CN+9}8FVXEe^Q_H|qovA%RBi|gdV;HC}pn`TX6%-y0m2C)m zc>ZK<6^bq#wd^hCtgT?^7QX0xRGQ2?rZ`F#FI*Z@vM`y~2Gl3WFxbYsBfB1!LLGx; za3Sw~ew~|Mk<7agni-uSN}&$Tx6_ZXVUTPa_eUHH8}4l1DXc69p)UyAY#D`RvykC% z;)Wq+n*+RlrMwM6IrN>db%GIciRhl->b$**?H;M2?b_T~rMzv=56rI9^eF41Fhh8Z zJSrV^xfTN4lO$?*_sy;2;l3jjWDUcz{5}gqPxk)VM=KF5m0}cIO)~SFb=kjv7G}pfy-H0mZ^$+-~8B;7N3n};c$qAf!%-9 zf3GhvSNABf;YPP=L92fOnJ7 zv(Z0I|wY8I|>yEiu%n)PbQOUm8+aU&0}%kmiUxU8%f;9{x=_#}(NQ8xJt z%}MbYY7hSy*+RC4F0(g>=sbk8NA>tCSRtH9B{nErP1fn?bG83REt>|C!Q^_LelE&N1V@$#b9g{8c8(|3`z)G;cL-?pG_!Q}hqS@W>Y6P_sd+%Mw~Cn0-2#j z=f3>iN9tNAi>+CpOZD0dx-mv)T8Ex`ec%vq7Ec~|EL+q#87vQOKfU+#u79rj5|D9% zRn;YNOQ>!zQm^O^8w=Y>^-zXHGFJ-e0*OZPTVHtK1G7mNCUZFHj5g0??^9_lR%YOrTWb9VIgZZjv`<365|4{sIUzfuPHD5${feBB?PL<>P$U#0Y^2z{&C0v2w@8^iN$Z+xytKAso&C zGBg8j4!Jrn$FJwvib*-k+vy`_?o=skzw+a}D+qL+2v%FR? zOtouw@%2=KYp-TEznR8*C)~=A+@ZEf_xU@yZO|SN*@=_KUVivDVSlyk{^M-9EJLSX zk9P;%?VTLj7B(iX3E8b_kY0?qC{BuM*RJR9^xVfE@yO)0h4n(PB|%ggj7>OOJ(Ij{ zkh>vv-xoF_t8%*&w%E&FGx4~`Y|rRBGe^m6LafD9=5=1bD>b<{iod_yMy+AD9OAH9 z*=c50vI%+$#vN>Kv~+u-{-(Zc|4|<9WE( znJ9XQunFMl;l)nen)}A)6r~9v8L2^^Q0qDD0E1-YxDKv-f`-nW93s|O?KI+E7KuWU zx7-FIyZB{dUHAGym9)7t3nd!61OPU7g>bj)S>9%RG9S}Rx+tkC6@q&YaPDH^3 zhW@_40Y9BL_N+|Flrlwt*V8VAfFi{yp4+_Ye3L`V{aeYS)BI_lOl}6=S1bZ82);2tgHH792rY@Iq4FUs1hKkwc`%s^hVkjmVICh?N;%sp~xHwNC9U{MioVpG(QY@j=5~p zREK^*4Z7C};#HuOheV2t3dl=9#f?nei$z6}8$ze&X(#r0@#TE%)3N&?tM`?=uiMR+ zR}Q~oG9*FqbN`Jx!p@LzIKtm!W=KvGGz4dwiP$C4dLJ|sJ@g%(ZnvsOxk9j9vQn^3 zbxt|vQb2ZTc1TamLDq}P=kO)?0IHwPL1k05&j>$@#*4DlhpF12bIN$|kBsMKsmEN< z0->~a&8*_Ej5+mlkm5j>8F4SXKBzG4z?^;xQt~(UqzDSg=%V*#?|x8U&x0curFWGe-ctxA4)uZAk{WTMTODtWM~PparZ{qhZv?)P9f)<~gWUO*y;ZH>4D?x0&5k|Qzhs1_G$b3kkX57>a7 z$;He=cs&U{>}+aj!^ap-O&no^LisYiL49~YAxdczCK8)CN2P9X3P65Adn&+`(9O6fIaA#a%4D(&H68%&q* z1hq5bnqXaMj$qX@RlqHnPp+HZ>4uW+HelmPV`Mn)1OVCey-zxc9e~1ru=Zw%2|!o= z^;#)a$YB5-H3Lv3K|>(8jEF@&vDF^P&C{)fB=y7oq%A&qFx_NJ`oy)oF_)F{erbnn zi_ZZC)WZ~$^|LzxcIeOo?lcnAf;h^Q0 z%~f^}t;Q3!o`eZokF&zg)c?Ge_?~5ff~ppIS!BQbKtQkEv?*-H%vNj|>iqa(rrXEkUm#s6-;R%B`GQ7g`9i$s@qHfoo-uUWba8Nrn#&w`~w| z6fkVR&s!%JEVNwIvH=B$YZMrkz}PsHEa?~h`>N8v=AEC=8agYs&(gL6D(gVKrGV+yNCc7rl*J=cS(sdz3v3 zAD-(tklBVEOFx$Jv+dw`!kNJ@ZHl(Xpm5kXWiTih6f;PYB8X}g%V;cRzzCe<2Kx!m z{J~)xLGj@H$=dj=d89*Vijuu}CgU)*>M3h$-)=USWD;}+L8TF~J<5B^Q_))`RSQ0K zuW~EoApm$&MDOyJNYt*oIob2x+w zA@_0hPCE&@h@c9L^-gWtMqa7}hy9dxng7ASHQ=efAAF9tOmLZh3iL(0gjqz!SF_Z3 zY<~z=P-4FCLvb;$EV4&=(O)0Y$jjxozF4u)S#V*84J&hnliDPLMmRfOivew|9l>q3-RXvVd0;gaP@K=)CZ>pZ=+OAs%oZs%!_4 z6`SJ3A{%A^g$})Usxld3>N$g}NlxouG1OX($ zyqziwtO$hFdf;iG2*zHyR&tm^{>(8Kz49ph!hH}9(v#wIw7EgHphDHA2EaB2wg!THRRO5Y=s}rEFWaSPdY0`f zZ4v>+YF6$1=$yMHT_uA|&n^XWnH~ztlo{$p9xPw4M_iA<7iTFo;vN^H!v~3M7X`j; zYCWk#iOFpc{TNVUi&$@1i@Z&pK$%Jg_V7CBE}&H%oL|H1^MFyv(^NunC$PP#frdXI z*L^^Vod5>|_waT2wicE*VgMuC(J|K8j))TqlzNmb~mi(YZSZ`5VA z-)BMD0eWG(7Uib;$@MeJJuwQ{7qB$sZ-$>(^l%T_a!1^aXLL&MDOSxbr>;LY5RwyA z3e@mNsN?c>_nk9Q91EwvJjqSvM!#gV>EZKE3h zPMnTEYeD}dnRmmxh+pN_EWErRRf#&@se&@!BI?H1FGvUYoo-(Q6bo=~{IoJvpEzm8 za_0fa#suZ01L!fqN@BrY87wg1B+PAf=%JNE)v^44)a$k2Y@bU-E@evT%C7JF%+2{ zkR70N+N?7E8Bb(}8GpUAwpqb%cZ6Kim0_O zG%HR}>m}{lW%Au_SLfydoqL_2Otsg0UFZnEL_89+*?WkW2JJ|N{(HSMg$@PaM1aOL zEjbc&>A(EXBzxFAI2=c2VDrGhT$LM0I|H-f9a0ZA2n%uV2_TRMF;_ZEy-~D$z78!@ z5&xqXjJ)m+g#yFyO$Z?j_j2U!AN@A-z2{8ErQ@4d^QpZYUhTA+nTrO3t|O>wA{HCL zu6Y(gW$l>Dj=++L7Wtu|J|G;+Cv^x1qrCgu7?Dj^o{#DS;;2Iq``_i&BJ7aG&si2S z>QXlMQs`=4T*wwb;17bL?RY_oU}prxVmHk0)nEe~#(dl9F&C7PDIn{3+o2C_1<280 zpR0ZFnGi6TY=`-NyKus4j+)w#C!A)P-DF4@_Ch09i@eyk7>F;1sU$jwmlK}D>k?vv z*KLKK{KN~3RIXLt`75j$nXM*A#8t7^Q$Ix2Q@W8#h{=E zn`KaYhXQ(#hRhDY9#)i?*`PWabds!sWKyxyL^$okFc=cMkji{9D-2If=6~V(J=63g zQT=ub)y(1P>2ouvTqNjDf;vmY*7@F3b;wqG7*czfCqy+r_yEa7u?~wRiljAEi|3dN z&SRjk8TEd2zjuXDH$rZEzD_gZp_g6r-07AANkv_jUx|1bbdR7`QFUmxWLTD~!JeZr zmz3x|bli*aFUQUA6rbbWlA^N%xy2PQTMh6H?hk(RnwPRAySz|eu$9DSdc11!*HN0j z%15`$3%+T(cwW}NuX+oTNx~Lk$_t||$Ef(26VIoH z;P!HDf+)c?C1#&?)Foefe=nq>)cMlw!I^MILEz`!{lQf}#!sUzH_5?>7P?wK>T*i~ zH$B4N5Lh&Cr4Tl=?dBDC8sr{_CmW9Cjd+(>#`QQmr7;wWS41~K*9lH*=zKv=9aG`p zMTbwGX3_m8ZR;@KKYy}*vYWj6L!w@>r{j^k9vsvZj}P2zAZP&lIwE!lls{p?#u8DX zRJTNQ;91?e(5^7lQmYQ^6zeYYkx}bV@a4IWomU(=nzkb8{^aAx?|$69?v0I|_PW3B ze)h~eRN9wj)+sf^#x{b6G<6opBT*NioUQ~!k#+%G@{%{G3s!9Sk1;#;!ppI9_ zLs8ZtvR1RzAIkX@TV(ZtwVHg*j;K7RH_5Vv$$hvu%S*lwox^{<_@*UwGXq0%o~Daz zmLDarg)|FuXF)Z;`qLSObJx%MOxaJK`St}Vwtr-)Z&T&|7v`?#Z+PZUMgGO^f4Tv+ z0(cG{Sr(o|ubAn)S98J-+4!{YKNE-lWM$R4TR?PqQTK{KlSv8A{(CBw&*8OFgPD=2 zAZSqL+zo{bxQ@oY6RhCF;2o}7aCr@CpXAEbs4UZH9Z-{|)Cga0OucsmjO zz=|+zOc?2`5IrUKlCCY<1RVEU$I_{N)=C(-TN6l~G6U05f(CTe5wXV==XhH^_VBvB z(XG{uC`-iY%0r-2@waZUcd(dl3ms9W1Xo4_D1?3D{UHUxx8_wwkZ)T~>d{@jI_9{~#ac`OUj1)QTP%ddOvRP(W*ym}gNvtz%7A}Lbt1hxgf+eYT zHU2vpT@#+Wn9(JQ%3N>RNXgKf*bv?jj&EdT#IP(&jbiOOqxeuY`RR<&m@M_37qWqe z`-an}{)aCa%jf;@LMLxAKW}UC49-f5|3LS?+Ojc@vFq}w`j=4*ZFG8aAks*GcwM|2 zXj3~w!Bk-D^%FyDPC%5+=yt+)Kkkb8$J=|x*ve3Tr#j3PTVJX+_ScR%w}Yun_~SIPT6}lrsq_+$uCj-MYoyS^`%)O ze`Myd)DZL`f~v$WxKa?uDx7=Sz1w?)#B!BBvin;X!DU%0Sgq_)R1!-C>-l-%?__-a z_V_oe??9^R~|CF=w|ibK+bph{x5xDwcW_ekmkF?$+6?YwA@q(@;a z+pKcis~DCppN4c0*xKd@q_IK}gR}d{ZU-Xf z&8vG~I{tdQc8F@H<3y>ThICO1q|8yn&?#B2>5&%tUUol2R)h>G`@|()i(UK3I&xT? zrEaGOlsiNzFSpyz7GX2!9 z5wTrV+)V6@ZP(sX3@Aqx1ppzOOxm^OR2m;!xbY0Av4hs=QJ2%=TE(bKv*M=UdgMpc zb@5$T=dJa)3uj~d$x@&gxurNIzYgsws27P>o{Bz04$iI@=1B^IPf(3e%b5aAvpkXDHO~hsfWRsh{lb~P@m-xL8M~n>{cyi{pPbA7AvRQh9D~!I(ZlM>yo4PFe7p} z#>>Erh`o)-*h%$J)!@1&dUb~#1uldx;prKlDUSsp5P=oLce( z)#GVA+5<5n{2kN7AO23Q(D~4=-5q`_40^d|-hjx^u&gCCVH$S5ER(muG5k$()SRUs zVs?b)cf5VhQWTQ20UZXJM?;P~9}07klgiMc0$ItR<>6WCq^ONC$oY(2C#Y3|ZA^Q8 z2bAg46NpJD!eeQB?oUaNHyjQ)$_xi9S)wVgW`AY6C3h8PA;QQ;bxV3EC`EutWdr~a zjunZjxV3c*^bZbp8ju9cMrREL@xa4|Hj%MaH_ zB+;b-y{g^5$Dp*kR;KsS?GSCDQL7JKCcLo94bKIl7~m>{)WrL;Yhg${o~KFEBti%8 z0Fa=uoChZesQbWT@sl0OSOIG4?@vc$Ex5lRSN^1kDre^yaoDK_xyW&jQ6oXu6I2Zm ziwUU#{!ag1sNwDxuMNo#I82Tzpo70G5b8O}Yo6G}4;2(xgM)Niu(*dZ9fW0LBozZ4SL;YGk2*F3|{+sNu{ud3{it-{8ER^v9lq^DZktmm4L! zSR!rw_&XPMLz+cBtxYj>GpXHx`ZCvda`Zvn9$lA4fpT?jbCG?LQt{;EJ zcJluDZ2%mYN1@Jkt2CK+|Ndo>=$G4Jy$m+RX(o)F#d_rI>ehD(_$CwP65z6ginq23 zM&r%7QL2$T5{vI|LWs~}dd@$}^VJVa@C%g|>a{I`Zg+-wDyGzj+c~Dao z0S-oW;1=3Qmy6e+=qvOzVyO&%tbqJaX*4i@;)iXTT?)8?JWFl7coCMv7fB8TfH60o z0Scdi#WL(nfnC{P`O#-Nq}~f{=y20iDo@i)9*EHO3wQcLad1B5FWP}OxIvLb!}XDv z!+5L^yv>8g!mG9YROS(99yDuy5eo2BU^k^f+rs+mHGAmn>Nn#Q|HE};$yKX#sg2TQm-GQTkoxP3=h z!G-u;{Gkn&&D@Xj0(02QWoYn55(;EjH^glWEip)mjx`tdu2)!(j;%mpePcT>Uw+T7%mYJ!|Da}PM3d$&sL=!I2Zq9yi29?}L79uL+;f)~&gz@Jb=abr8d**5cW;L? zy1{vM{vg=;)GgjiJD=&3FK2%0R(w?MzIXl}uipKk$wxi=?`3zWHfyV64u|pj%{NR^-M0+P&b=&UfmDPWb#m)&m zc8hyB<0JE9w_=L8yQq9!|0kWLp!QSc!fZ1$Me7NA9YLiS19iINQGH}%#)NPDFKZXkj|e)$H9&bipO2%^i5_I|KaS0)s{q!PZcm2SS7oJ*ldZR!D@0{a0zrE zoSU7tKu_)o&<#DS_vumIB+mqHn~l^0NS@Xq!)9Yc&b&;ue#W~<5{rb@2Q?$porNGa zV6i%%I0-h)U>WU?$8KplVyOyu#q4NlE{#pMsywjc+y)b@*AvFPTI$8VQN)l8$9n@n7YkpsC8c&3$w?}!fYjIh{|jxVlj@_BFAFm0VN91 z4ZAlhvL!j;x5}ga7$FT*|?t~Xz0m2NyMHI_iC1kx`as2)dMTiTk>4VaAb<8M~b5l zVoT%&gTUKBWV_aQZkx9D^&LOYc5o#BD1WXMWo z+1y-7oA%-_#=gJ#Kk~nNN?XWV3*vCi;dwqE*5bPpM~WG2p8SC&u$V24wdLE&g|`XIq!F7&U=D|J7*b4_ z0`HLBaK*fLubb{u@RM8Qg&LI5%u?^<_eGRNVj16L1S9W{EK4Zezs%`BL$#ktd%UlJs`Wj=2*1+*f^ZGLXbyJuA0&9$Jn3Hakwq8np~>TLAccV@Z{>>@r14ZP|Dcy@=Gq)^%+T+$4h1URjL^I>YmXpz@Ydc`VAPu`NQ@XXQ}sw>)fxQt2W<+yE_ z-6F>xrNo#|#HNHR(vS=wRom@t=+HA3DHjA|=(yc;fPcmNj9(H6(5;j= zgd4v|x1HA_-yhNfVP?*Dj#OV5Ykk_anl8MCg>h-hDz_yBmd0V+y^mZm4fn)}@?C3Y zB}_|(V9Yl6QI{cdFtXA;J78P*HuolRz6MuCc!pE$kgYc75&s2tSeP$|eq{DutVL5s zw&?FZ?PPnsnsj~v)KLyW_2de{66%&BRWqt+eD*_)QvqOg)NDkB)oY!(lN~DNhy+oV zLKly&hB)lamYRiD9>`g=c%}&SB+eCG&`Dy*Ca()^l*Z2)iqZ{-T?#ux4MW2;_EoPB zFO}#TBU^+Ysr9nG>XWiDm(vk%JGGzRM(9|xCr-5}%+M*k+OTS>$-P;&BBYTTbAfEh z_#EvTf=(o;l|<~u@Fq^Y{hDG_r<&W z>d(}3O^A5$Qeh%hWUZQ#yT3nm%nZH<2pWW2_7bs`-#GXE{r{0MX9;+oTlu%6%Ajfm zWFpg*n`R#k(4D51hw8~P1@bj4pPxsjMCoM%z=Dr1?4h8|QP3HJmD{M#hy82F3j2{7 z$*qtpWQfWR(34mebs&Pd;*kM1kJ)V#(ybd4>w6#bU2G09BYpM^exzma{!0>FVF>kO zl~cMBIWbWX5>)vhZEVB}IJI!BLe&O2zw2V-kz?e>m}JRm$P`Y3Uo$cIumQrv&i$ZD z_nIa1JOgEWo~AhR@XRfKUBXWxd2&%$syQxi79NOd5FDP_DCrWHODEwuj{8@#c%RC6>1dVs9v~^T5hoN!Y3>t?B zx{{zOh}igP$mhL$S|RVE_XX)ap#Hiby(`@nGe{=$>VP8wJ&VSmt8>f!+qI2B@dnyN zy&@xOJ6S^J`)&5@)hwTONqKedZJ;y+Mjgpk{)Ql=Q^XKkHS{lE07>e*pwyP?$OaQu z@UiZSn(cKInpnYyxJ3Ei`cD&l{`26buTy8)h4wjIhBRyjwoeJVm!Nux*cM@#FH$=- z@{R$~Q=xx{EJtEk1Le*_PI=s=YYZv|dOIwQEA%hp_bN`R&Vr@@5`-Wc@z`TIdxRL{Q#E`N{j(841oZv;64S|CVdxDr(rhCA0SZ-&{<*VL&q>i7m4yV3~$qu zpXNl)V26x}e8a@9VTBCtM!A^?g_mca z-x%dc$g@NA6Q_}h!Pm-FtPnlr&yH`f*@Z>xo zx$9AG3)ZCpSx5r4SoFC(O_Ai~cu?5{nJ3x~$5$q>g3Ofk52AMvCdl0LxYq=+ZYYT# zpHWOE=+y+3K*XN+t$=04agAe}yy1X*_b zaCTN<=-*jO-BoQPKctc+`Qi0JAR!P@7*Wd8#mz_chs5dIqr1Her3Jto=yoCOC1W8`oc=mNNIk##G(s3TOj_j!qN`v&!HHFNnKLNdXBG~xhl9b&4O!uXaFu*=#&VU*S$r@gF){tSL=-mzgJSrn zLe(_mu2hGd2E)`x9)08_QysHGhLr(y+Adbem_qiQ{CJkh&sg@(j6^D%!>(zSnFF+o zpoS4N3`Gy+VTEO1xgae7IoD;pp~%7cXPYu|b+<%M&ae$x28!&INLOPJ=lm3MnaHNSuybWZ^Z*;_LY zli5B;0$EOW4uKjY2FD+86x4q9?|=M}WzEy0u;OsCh@rA+M_@@rF11!*h!nT+w)j;> zU*KOM;UiEu2SLN)IS=~BWNChVpf}Y_j=1mV zy)XM12=kMn=(Z+kKhzgdO9{*>%^_~tiq@F(Vn|5O!G_MAVMWKSbAho>5bM&?TCHItHOyz0EPfUE`< zueF+T>gwEoz>H%nx*ZJOC*Ce%O5qolEZJ;I^Eo^bG4N-X27Cx5aZ5yN{QA9tDf_&1 zRAFoYt`A%f4Ddz`B0cbrn)He;DfJQCW~0lSJv~cZp+dTCV=5UrPR~m}6K0cfF$KZ7 zRLz1z!XfgeyAC%WT;Jr{>}2^jZjk>hVu9Xg=c2%<}4E{Uo{ z)fHYb5A}|+WYr}0GZ%@fe6rN|Ye996TLSS|D{L#wm?kcJ_CF45A>VuLPYTmKl|Az} zzo*hU9F{3Fn?Umj8VpP}5xYbd=X*%Fgt`gUlZS-eN?a!AkyU^zSY?^U|`aGb8m*Tk?3M|v8Q=vJts%D+{&gWmHc6$IvOXME%7l za|a_g``yx@1Gf((6>rSABpz`;EjC`8qwW^peQs#pcJlspSddxF&zR#t{7o2YR;|lK zf%WKh|MkP(H{LNBnVj24r%|7C*vKTB(QJ$o^c{j4CSp<9u9eIYw5d}BsCk*p>jxFR zbmb7Ni`GCb!o9FPJYY0N!B$Kllm@)dgR(uO&|WD=#z3r%MH%WM2?m9Y!VSG-wXgf!PPY zYRsiWW{f36Ev`o*C{d%*Uz@sBth+=R?}M5V6)L>JB`B>jc13QM;SwValAvQQUF1L* ziriqya4lIaUgOuM#$fg@CLFr4iYUkJoWja_P5F<8h9Au_0nkOF*+=DZ7=VY(X1Q{L zhSvIBL~O~+7o_dlV}US2)INSW2()IYOQR3)v*s50WADJI3x3KIm59whZJOPu#68=f zOuUx5XE?WA+bqB5kLTat<3KB9c>f({iXOElcI{z>8{(Tszi0_G;U+ly@2ONihjG(j zhMNk426t(Nr`z>HmkGvSvH7PU2>ktrSNz|izv2b?-^1HG20{)4&O>8k7hvvR%nmq3 z9WQJRFahW2OKU5sZPTcFv+>zW&_EzqLc~@J;zWI7oGNd^SX~_PNg5Z7p zz2ZbV}r0ovN`eyq=-R1zm7sxtl&pvzRAJ)|{#+ce(sM^tn|9~-Hov(1jk=fPUp?{NQ`#;vL$Zno z3@Zw!v&Pv26~}i#Ug=Wk-7HH6% zO7v`zuk$} zvA=@d8_e0hk_7*o-?wx{IP0)7I2U)t0a@!BEWxL-ibTVGKuJc3%n17dm6Ajw(|yq3T%HRfMT?UU4! z#gSR+N~lX%<{AGC%6=DWGQCm+`Q$#yn12_X(XKU~Sw@b!RE1WO39fi1o|QN5oJ!Mn zE14SAg|wE>;5o+~==-uo)HrZi+q4lcPd<$qQ7?FZXVlUi`I7jt3|`d=&zhioevt&L zSyn&}5bM2@qBh2)EAuob=(GG5`G>qqz_GtZTtE)+YpBWUeVhPxo;>glIHT|HFMiXW zl@51<$P~j$CmEWz^%3ZOV+u$&L=Fg7Er@f&jP?DRs72)z$ZdLF1I%C!8CO7feo}a` zU}e;>`x;epOulAi6o_q(c;spF$&Y0N(nQadQT2hvP;t{oIH6mY)tzcfEd#) z+I{HhmaPGt^XRRxKzX-w)`gCe1+^Hx@Px^0@OSYm<3qu~J(+ zWwHOe-?4#LXY^xClI%!0|@5Pn37ez9F>*K=ysZ#I9cRN!xVG>4mda zf?>jFkzar~R+msu){=RiAA!qoNEjER8vtQ}TU971#a7@O{H`aQiQPCkOJ7VZe zM5mh$W3=~+-5duTD`RHZH9o_1}ndq25X(V(iCwb2inq^P)=+rd^f8x+F36-NU5$i*OHc3PI?KOYGq@iZS6l6IhRpW+)E$ao;7@{^^X-9^Cv3~WTe-~X zyO`&km)<$F=-*BAi%7KYJJcBt=YMaR%^1A|-9u29h}a?#0zlU#%;zuW=13CvGN|vxNButLe(eFan-fR z3&;(#do_i?%7Cq~WinhM;pJm4@q!}J{p#;7@*}T4vQMnrrD*oV-dNBOQS5icyWz=r zxv6KifIJ;q$6FhKzm!hzGfd%aKlZ!kVLS)VW<;zlsIqid9Qm0Ep3JDSvMQSTz3=|& zd%h;5-J?&eq_Ws~G8~?!4w>PygrI?HYCC2&?~pkm=cu#vVlNaoyF|5kHAq+a-t>9v zty0OMfDIrYkSIv=%Z?cavvuA-HL8cV%C}vyH)1bT@n!NFgEAy|1qz3kc%6NIX$T`k zn1;xf{f>Z_E#JfToUeTT=IE0(@Hh+$DAXU9@P43ymrLCR3d`P*PHD5SPu!#IQD0ur zr|hSCpbu?`8dW5RwyAr-v&CY9ZW5kItWnj2zk*FG2W73rn z2Uh4wC_cZW>g;uBDDcQU^HJ|7Wc=nMP_BV>}Ws~I*2I8PhUO6c2N4u{7t0orN7rT?F2H5zXF0CG)KQUSr!?6s zPLwUV-;$5~?i<5Ts*qP^lULGQuwq78gE5yoymlz70WpypSwUbIdCe0`>WnRtc~BmP zb$NGq8=;`Wk>Nta@H!4ktkLq=3z(slUh{({OZn)h3O2cBhG`u^rw~*UMga{oRfp^d zNY@+Ys68+tR*SnKpyPBW$YLyPMuo+7&i=@Zim&@T(`U)M`XvboGO)Dn@ZI4XH$P95 z&g)T@g`C&4Yx6X{VH<*vlf6`>`$72;kI6-i&6sg=$j=gk2lnymBTP<<*mdDnYB#%8 zIEN>oCNoP?OVB{udH^j6ii@Oy2tbw36>r`B)WU;|RI1lr-!`>A;+kiA$Yp4vhT^^J z;&kO!{wnQ8h$QJSL{t!*=68%L4Y?;6;b$YqtE59&Bm!0%oQ=8{4o2ouu-OAkw!1+o zqCD`^cf0j!Cg97AA1)RUV(H4*7iFH_rpVLkrijZkVAsPn-FlB^m>F?R_jVpE*EV~XMPPH&djDL>MXw&k`(8aX%46$* zkywd~H-mIW8!Q`_@jBhQ;8J7!vmaU+QYqb2xDxBr{PvfY4i9HD=?unzrBPkJC-& zG=%eA^ zMmxPy7SLe}S{Apq>r_^touWA6`L@!8jf~WwPpI|Q++7Y&AbZX5l26buqvQ~=jX~Qz ziX+eP`VA#ASEOZRi~5Q*N4i(Ken#1>dxDQ=4XWZqy_zx-T$D|&iGpU~fv77WFR&IC z)b$JZN&dgZi~aM@e*1^t#7arpg(8So7RP>WS*pib&y=C~E_YU!uw4sGLB1QF$$^%U z&tCl94(Y+bjWbpN>vS7$t8@=}J@O!g`p=6JgEM(Il`XEjy`2jL`tiMlI#=Jj1YfF{Xu#OXLojQAKuO-VOo&K8II2w<*sE&R1kKT>dXmEET|L;xzy3f){WfRoraOjr7 zrR|XI^6q&C+erqL=ajnhP;k%&VY6Xbmq!=brHq?@b#7-^k;sv4W{;xBY!6F_n_cV7 z5APiMTZG0mu^f;6zjah8yZ8r(vC(YizSa|T4M80uVuvG_LQ!T;INCt0_`IY{@yvx< zslDDyU2*@I%K@^3UM6oL^T^M`fGhDpK)Q0>tRCsZ6D}!N2)cyvLB^6%9YU*`%#^EoAe2&~S#6 zC&`Hqy`5wGgv;4Ykxg$Z=UHl|aTX{H`e=##5^=Jm)Ef#7d<)2H9#FSILoa!!lVy#}pAM@bFu&BPpV$onXXCBkm{RuM z*G%rp?XQ31XH-3h*TEOf9FSImK1om?LS2DRV^FF0sN(Zymd?-d9)Tq6y}4(odMN6~ zh)EWGlw8CAOj+dyO|qhOs%GJ3EgoOGU`&jkMqS3l{Ztoa-0t(wi@LwPkC({Dy`bnN zEcMRgXL;!x;Rx-Fu3nNb7<70?jmCKp8oR1^?Y z#LFm`K}AtfB%&fS3Jfx+@I6lwMiM!i0|{To-JM_MT;B60oA-a7_qqO`=a$6gg?{cy z$6jC;4r3&eOd-SHdSYE7{fao*Sc<266?+4d_&UE<*)?&ppSDwx`2WR~Z)00%$6{s7 z*<<4J-}W`@6|3H#v7Tg%rB^tw-?r9`Off|i$){q#mi6Bz>!IWS3OfmP0QGL? z{N#nJ7l87rP6-?K2R;Q~NWvE!^}Q+H9iFM)M;{ZwICu&Z)`0sIvR=?IQ0jo$m&ziAA?<;Y{ zK1G>-E?+OjQA?MaMpwF>B8?1QY0+F0-&N%DPYII*Pox8+L0(7Wb-ZFvY;iyiySN#- zV@C|BhvSxeo00Rg`Y9JkniG3WfZBG{hEqT>;P=Z$J_IZU)MAcF6+br&JBr;@54)_B zUkvTw9Seeddq}j81zWdL(Brj%IX@$lpFg|RA8ElZdLD<^PCMK$uzwe;QOh*JVb zzRbjufP4rhWy7A^!AsWBMr2I7cY_l$Ufa8)PVHe97r(kjmXahV78k`9;v$n`z^qNj zy(NP#nf0mH4e3KDn_4Tl2{GT=pcG}3_>7`hQ6ozfbVI*h?5ult9WhW4=S86LMa~p^RKOQ)Sj<1cVhjNrRw7neN1>fN3l3KQc~j%Tds47Bnt;IHE8K!_n%Topgv&-8pRWF!pB3q%RP6?Nb9h{E)w zPIb>cX+9>oAiY2I7jZoZtUPVf2MZE-SiXu>cBbbsjCh(PNREq+iz?lab*^2Q0W|O} z8f-?sGWaqO1l91;rk^LK@jBHh>DRDG^w!XOB5MWsd1>Xu!5tQ*7zvw<-N`b^CIOcC z4!f90Qyeaeo|73i%BJU@bMysv+2p_Vo8i{=8C+cZPF&H>Qq2)BNRw3u6i9j%_1xN+!|L7PlYk`~!877VnKY&7cfLLz z)`@M;%MB+60!y+R(i%LFbr}m;H$YjBk<#rf$@E1BZGi~FK3OyGEohbXL?-(!O9mtR z!`*ILq-sM_IL8(Yjt>eZL^x8Gu*;FEqyOA*cY%(pQ{D2yPue`KQ`j4U z944jo!8f`g>+>wR9h@)C1A4%G-T*;;w2I#e41_seoSlC*1Iucs;eN^ujGTZqrSM!^ zU8osY>r#65lHE=Wtg{xts-&316gf!6G|D@{yZUj`*o%(!{7O{Br67qmRcJh?2gZz3uMb9_QtWmN6TdxHhV|#2zzt3-=4pn^k7{;YBAcf| zG9IgP&k&9^R2Ks8klUK2CFE#uzABAu;V*60v}ktG z=s?5QzYI`yPSsB?XzcZ(vaHznQNl8zoi4v3D7js-YT88dT%6dDV##ya8rh+G61~^0 zRkm)1wt5<7Tdtgokx6VSft-I12DBCaIN;mrNynaQ3PS$y-z(2$kY6S>F{#9gjzsEI zee@=h5nzBSj*Wh|K|F4mQRewa?Y<@50V_lPOI!O^cL2l|gipaud|rwp+J*I!L18}c zB&nD5zuBTWrP2#>BoCG8ehI+3ktfUdzsl&{?T0Q7$gxEkwi}%EKY3U5zIo%(KdPHb zKATF`TlnOMDCRLm2B;XcEvo1Y3093kZV1rJ8>ttOesM*ZKCHvouSKV7kl$W-Bycww zb}8mx6Jx3HpnEqM6&FJdz+Ukf3>R< zJn49|U6>lwAk<>DYxRQis8ZJlt|>~KvdpI{C?iCts+IK2J{Fi8 zb}R_LV~29ky@AI6=~Tvt(5T9zckvQwl=nBDB3C5xEy{?rt$d_i& zC!|Y5c%TQWCrH`Ln^h$_Nj|k}e@>PyW4s#^%XD_T%t>W` zZSnu!I8|ZZ#xiF9+j}I1TcWS?%12USVFu<<3=}72po1qxdE7h8BVRQX1bKTH2>%qS zW0N9{oAP=-)+d!jAmQ<_3vv--@oH^Er=n3l=usfq>1sdKW~?P2FRDQI8EZPbUEB7< zb$@AHB>%MvII&dpH-i7yAOqTUzlWefB??xC=p_k~9=cJE4U8~Lp#{QFj$L!hMdE}2 z%SRo0#x7Up&e1)wcH24aF2mwaJQkQE!W1g>?`w}ha2YcgN+P;EjhURLd-|{j*DNVg z@NZzSP&Hp_RK9_z+_hC~YPzK==G!(9v~%sopz4r8&#&3s|R2hfL7~-h=5nu=m>h4yv^+ybDo>AomJy z$3}N3aR_j~JR6XvIm@(#-kQD*gjcro^ui`_7XKn(`Gi1=qz@0F|L{#^F1;oIrs|=S zsPd?7jP`C=BQKR)=3%=Ca(ZEorIIyAaLl&jK2uw?fkN zyYo)Ho=9>gxmwJ|(o=lv?frVcWIk?hzthsaWJFc}!rd&E%D?%YCh{>ivE;;wdifUG za5}|oqhKqE=_e1Muc9gCyg+h=-{XEPFkjSdbQ*PnYMd^DejY`TiVK;1dZur_)mDCtxvmYtar=$0? zW=4khm5{4JXVmxS>&Qms2Ief;BDu^QkYV`)Ucq+N38C0?R_CX8E#vKjcGYE&$DG8J zQ+a_WXLUjy_c--%&VYJ>-Zb{D@xd-(sB53r45+h_z5>$24Ny*Od?#9HL2<7!!xVXf z)i2%hL8adH4y4PR_P;FF)z7D(e$aQ`jz*ZX>i~Sa{b1Z($kfJ~|qx_`iKY@b0PPOGdeP>ZJI)A{>&SO|P&(Xb2tH%{U8Otd&k zj6M5=@f}sW{pQivKJ5#tV=y!DqNhG=9Tb1-RL!0}vp-UF@U%M_gHN+}H;vW38ImEz zA=f6Kc<64|0-PT;tz#b$r2-gcBAI)Q3v|iew$XA`5)gbPZofT9=brxtjex<+J!xInPeFfR44Enc^4ocQ6R#M zkBbEKk&1u^z?Rk^htLN6t2M}y`DwhH(Pa{yYEO87#0I7rn#sxxJ~?XB#n0xq`h)31 zC-aiQ*?T>7Z@>!P(S_P}A*eO{kHAY^7m>(di`N}qJ=6R614P@Z#2R6|0EU+GqlGJY zA1k#ZmlAjz{np9vk!3i0{~H++ZIyUKWD^}LV)ZcD{;sg7G!wmS^^T03%6@wNUsYz1 zY0mzlf@E?7nG+lEM=e0Mk7A&7vyh5e5uGULl&^@!x|syQkX$>gY^M{+;WxAy{(0(V zRW+RySrV}ys=BqGdaho0$h}1YhxNkC;&$P%GKa>_zNh`Q?ff2+?01vo^8k_r`lAs4-AT%;N8r_<=1Hj89p}2+Wo=(H38#IPx_fxm`mcKf zw*3K#1h~N6F*C~$FznJ21jHpC?H+b;g}ms89`l}!s2KBEM_%NFiYaHRc^`(G_ZzY6 z;w-Y)i3{w`TkIg!6a)0Ghp8A$tVmJz$kSvP&er*@nYlL*i|5i+`ymrKMcEy&T7yTi z3y|!08-qo3=ck*Fb07)212l_DpVgYx8i+vzW7-9l-=T}Oimr{=%X=1(#y0e(R`<)7 z+oAQAjS%9hYD^ei+#qyIUVGHqh~`2r;l%D<7L&U`6f1%Rbzopa+Sj9A?QTU>5oKRx z!?s1V2|>0^W!F)W|8TWWY!)l)(r^4L*)SD!z0pN_lp^+Ni-Z}X8naJH+WqW5$){kt{4?FtBwCm5uJ{hN1(|O;l_vz#H{$_l< zSA1e4$@-d+Wz|@KxrAbXeq=WUf`CB09L&_zpi1|Bl0o`{{Bs(J!xg)wsp_qgLHAQ& zyNujR;MJ2lI#<*ki3Os&Js-H?6Hp^T$|VoHkURhhvg+uMG-n03?O6Xb0xr%F9}V|s z$0O6(?JqmTOYhvNnqx+di$r~p*LOl% zLse+!{2pnt-x2?vvVLB&-=KRH-93L$c9jP`%e|U482T<1pCh-t_VY66z9sQX@Orko zDild4Od~Oba86Y%>vMmqw6`#|6Qbc6x29e9($UnN|#$qI| zXpKA-m_b+5Wx(5+t3JWMCD}wK7qC-oN35M8`I6~4A(mPmToh_8FkTUph9xQ4Sd=i* z{8NjWDp=NE%m?;QW3JTX!pL?MO>QV_$2S({S}PzIEl62~0l!jK_ei z&u48kY&^<6(|h?WDqyu%40)b+vsZ;8yN&HE+O z{y^3{u^+a;!dTf(F{x--nL8m>(*`g|O8u)sAr6Z4GThX>nC_Px#t9A1IMn&I(o;XnVHY@Z6A@6rCWB8q{6j9e-v_4O6dFsM^K z^0(QTHTERBdqxJmp6OJtUeKbs1SMKG=n@V7?+7#)sth%RUt#)Wo$3}%3B59;+rLMC zG{}LJKHOKuQl{H?EuPKBX;qxposj=8Y2UTpZ=SEMc3$@&xY9;rJA-1jQzX^oD<1S% z@1|$c!?N8mfxcqq;lL{)CG?ZnelT{gTQoOb|UOV7;6&9vvr@~ z(RkrJ!H<3E$>f~ z=vYrX?i%|BK8B;+d8O@d=I6cnw`Q54>k>UB7o6C-?X{2s28wB?NE;QiE~;;Ky`(MZ z_S|Pj&6@NGtjkGzt9Nnrzn)#3L!Di`lREPk+n2OxR`SaHV@21c?+*$8ub#o@hFtI* zbk~RniVF~|Lo2ME#LnDA5w*_Qt^{%OU+4nWk`tCMZjNv4pby1+X;4>~QwN zw};aUK)vp{t?dmn&dP)Tu!aV+g@GP$FOTYxri67Ui)4r2 zxE*{Gaw|U(XA{)6Gk`4w-Gq-|+>2!t%i=l;Xro0N%SXTPa;wkx|5j**(~?VDkCTs_ zc+=Zu0gyC`*-DXQDrSxAw&_SbwtISs8l8hnkw*b}6R@uy9=8jj3m`-a$F*`V{Q1~! za$Ihe znup?>;+rImcYCfunCsi^+ZcK-{FZpYJ%ir<`o_20q>W5I;XPQMU!+ijoMd8xK|dVSEK@}B#JsLt>V$*>Ey*}CVR3T3I-&;{=y z_f7+_DPoC;wo8x`in$T^0-Tvw>Oey134@+1nr$4f==`ql>t?k0|9j<9vTZEu%84Tb z2P`nLi(+ypvXhF*^erIwglLT$6g|-CIGesN!S)oHzU{(&iaIEH-WypIQBBuN3jC|Y z`LkR7PemS#J|NrSb2K1HP_M402V@S_eXv`%wN=ufC=n02Gzv>*4Y~BsYh^M5(6Gu?-8iS2Ke$U;dKqD zz>lc*h)W`lfafmp%m78cpknl41?oh|SEv@&(WsAnh8{Iw zi^m?dT6D_A3o__le(9_N31UTCWaQo*b}_X&Z&qP^p>&q^hF6=kB%+4b41tB4YHgwb z=Soqg1!O{;p-s9LoWsK|1)@CqbMeQ@dR3yJQC`G@7+plWadMb=&;zkG=%K}ri6gyA zWcEFxPT*;8K?hRP$LLhZTaP`ahF#Fbhy9(_h`PXDOi^kv#%GYliShx2T4WDjBSMN! z1Bf}aD6xZFrCVNbe}vQWTC%Gxq`xMiM9&D5RLL2_5JYs zio5d)Byj+~CUGioU)o(hJ(DXFlT^>uUq}09E>787>flj==v1F-cNd8<#Hb9^= zDWZx_C09I(1RxfWHymA1RkXH&e_4#8CP!Tb(gZ0RfFKIVqzXjZ+^}=xy!wf>oyMja zc{5i5YpNqqGcmZBcQ<=u1?C&5^HKiFEL!4s#QBo8u?%r179dY81jy$UbDtu2shAXg zr3zgvI@Ow)*|YD+^JM6O*+zZzc9+1^KuMdU#;ngwFphA@{lMe=1KyQBO`#7%w0Gpt zMVpsByCkArh~e`tK`!hrIrL}quZ2R7I@obilv^c55rcdj)#sihC{g69j{<2X)H_xB zeC*aA0r~2955EKMVz>Es=e2l0jA)ei&+D0OLZbns6wZ`1qcZ%Xh1xjLD)+lWfG#Gk z;GqEtse&6ld-)rfB+$gh;_0;uuB?=&ah+OWlYN7GA+Kz$IPdWaW9Rp)1Ij{ZbKbE^#D>+V! z3l_B{kN#plywBjjF}4K6#-_&2rd?PTcqpV(Q67~iDdukp0}}4MV8|2ZwaQRS22&jv zhRj7Rnf!H;$$pN4knLDukpd2a4L7X(^LyH0>xOTyh?2w73U1INSDnefI(PppgQ;Wn_Rwngm-g!72cy#BoWzjgqi zwneR(d49$dY31UTq&vVN5aNuNF_afaK<4j%^TXz!oAJ`|{deCbtxoI~hE|V)U;pN} z+W+{)ub2K#+C?#UC}N;u(7x?p_K2%OkMncX4Ri@Xg#wZ@n^he~c83nt-Z|K_2t6tz zEr|v_@C@^MvY~s(2%*Ck|4hk^=^ct~(>oX~#t=+8i7^uha~+UUxikC@R6>D`+p8>) zv?@V82JS?An9uiBZ_x zN-@b4NupxVcv^<(GddM^zDBIU!q5CXJcG^%Nse0YJ-OYFIZzT!+LmQR)wOTmNt_5W zoY=p|f{Zhf-H;022pNe+o^}vf3dkyVUzQtooR9w8R#}>SneqP|vLQY9E{{XPJU53U zF1Q|#{A=s-eJ&cW6Uz*i!v1(c+Vr;2ZL*wzD%eN*NtL*W_oWQFM(lIj=Vo-f#6kG}_2SO%Ah+IfWPqGdeIjz0S5(3)qH2_vlar7DJ_t4KX$WW^- zp*N9J-dkoEgii(cKt1hH-jV0h)xeNi8&oYDQa}?CAtY~-L;)72fLGrqX~xFTKEA`` z#kc@4O2u%5c)&S1DIru{^S!6myk%SjiW6@&EG%cAi~A$uL>bU9Pn+r93vEKm>1^m# z1M!{U>S;~l3_6jC_l8a(1)_^Hy1ca*@i`jQp~Nf%P;~PkgtIv`PSgZ-OfW1>7VrIu zDs4KZV;zNTg~#;anepCsOv{=m%0>YW5A6oJTe70>GiK%0{c-;lvXz_i>cpmXnT7J2 zM=?;So=L^%d2LF_Q&`kCH#@A-T`$cNJY-6xNZQ;WEE1ej6$y?Bb*l7`c2Xor;H7D> z0=Iu&#mudLsrydTTN!j~sKZ2&aaAXP6uY5ftcha2D6Xb_C^ zC$j350mx1ZfFx5)5=9cIm{no>_}H4D1G-aT$hk^(YQ%oX7Hc&ApCMqns>hi3i2f+OV{nWcW3!r1h|NNZh#o{Kkownb@ zrHIPzyHW1lk%67 z+;CnedNW1^vo(o;lK5-umgs0`*YUoK4VUpC!oQC-cq(4JnXp zEYY<@u|sw~uqiOt_3k{KYJ|OB$9~A4-)B z`Z1lrO9;*9?eK{eJaj+m)uKt@J@9I$6Cl67N~{mNA{O<4;&Dx;do_(?> z-Z!MUttNzG-OUx|Or(8f$|%rHRJ@Fu%C3i(3cmDLo421cF@M@j_B(OHT7$(_bBtn+ zQlx^4N#osv7G*0#F2TQlx~1$^CP7RWn^I)*Pr5&#j{zQv6^VjBYxgX04LvGIRP!=!N9*34_ z7sVq}Ob9X)g~iw#km>3~aKcF#QtyU%EqUJzBK@PfspK;!2GM$pV)!A7c}$T3D&|W? zfdq*M(PY0V_}HV~J3S1N%dj#O>*G_Dy~;dUuDV*WiqRt9fvIM2P+7xk2u&2|Aql@l zRUUN;sLinaeZW0`cHiv0;3w`#@sq$C^hkt0HNby{bC|pWEl{{_s)+}lOw%Q#)J;*| zRMtV}W0iQLAASd&3P19emmyp)Si`~rEXuU$=b_ke*ri>F zL#h>3Vk|B)eT|(Ae%6ifcA;JX{kKMjtYSJ9uSPxmYf90*j9@bWKlretg6wi)0G_mv zVuvUO!o~Zk82#LQz8<*qc7$B_-w{&j)2oaV^v^3+)X2K&(f4#Jy!(hGLL{uSSFw!5;0EvpU!0-s0(zkHuHYp`XF@lf z9=&REJCwkR+l(Q*p~JR`wybU&keT%m@k8-%NTL(#A$u&=Sx+%LD3V6SG|G+ts}E~g$jw~CS_x>dC^B1MYb*5#71!w9N*b+ zZiO}5#VaB>Shz8e*(*_y8nT*5Vu0H)U z>-HkA2rgI}kC<#;eb`aoERV0a!7G5It0G8SE;0BV@NNk0FgB~%%fAZt^<&;1F@(>1 z=Xdaq1?f~reZQEG-JDuw;MZ=Hrzo3&t1y=~a_+!{o_Cm);)BJk}iC zwx8tOnD@VD2FhbYOq4~rI^5Poh z?Q|9N?X2QI4u>Q;q`vHxw+p+)y%EDM`mk%_4Pn4qsq;$%{z3HTRfX;iZxkX`r8W!V zn1{o`PX{xE9wisa+1Ul6O04<_+C0ss$lF9A_cF=Dc7H6GjlRYXxwRe2dA~KoYwfy( zDAGL^yqq}3^s$9^_$kEmzCe&;l+X14bqDd;yrNV6KA%eX z?(eVv{J`RsRC4GlJjrtesj&{k}z0#ot92M%IyRs3F-tyOuXV zT9sQQ>mqa2mA;)YJ<FEFLyb0M) z4vc<_Cw|N>2fkJz^K)B|&$PDMUlG-t#m0}HvpE#)b_BMGHF?<$X%U?S34eFI?s#pO zvtbUvK3n~{yc>(=Z+W4apRIPhr*Qm8pE#OdG=oQ;=C{^Gt}NA|@nM(3`*}^CheC#2 za_AbSL5RIM>-cxYEt(tBAs4-J$mI_FZxgg+xDDsUQ9~{}HPCjEKjd;Ne8}a9_mIm5 zpCOmM9z!k{W)8V59dgMIk0brGUYa84Apav;q(ty100~hVf}T!`J-Ij$dqir2@a` z?H~6&ZX4vaw)X$~b&ob-5A!tHX%RYObZStTARgU6uRO|>#k@z+WUx)) zkn5a~@!F@e=KU3I79;&X^gB!9UYSZHy@k-(L@^sFvYv`*l_3q7HU(O^LgEc%;Xo)t z*eI-Y92*DgGTR6t*5@5{Qh{lL9HW*A3blG!@}J*BIT<+=eYw1K_&ct zS?52S#m8>91#`#^C)PrSECj*>is`1vJu1dPH+q774i0@ROKJrbzRV7k$)LV`kc9Q2S1^{@+KcfZ$pc(_~KMBfY@jmQau=s8qP z zBstjZ7NyO9Z`v?U_;@Y*pq#PJpK)4zuw=xT=txljn1fR&WXA5w@$5Bn(2=cD!$BX< zIxQzuO!>U^qjM_rMx{CXiwctI#2Xb9AB-}W_fZTu+X|@|UG)Bt*aaBO+#ixJMNWlo zI%S%6f50ayOqkg-CpB14Q`l;IPsH3Z<{d_zxn$=B}bhYGuJFIQ%^CUP~444*NPk@pwrV%9XD9HmnObJw{JRJA@9ap%|n#XuC|C z7&a;ao!7Ee z(C-W=%+Si1`ET!$l&Pe|0$Mo~1LQ0jRLmZrAjK3ZQ^H(DaE5=p_Z7a02LKC@bdVag z7Mg$^@P;izHIjW`Vqv%cUSYjVU#IPw4Y%B5u6E93Ax zU;+1C6a$QjJE@pc(xXBBBBRceFVcr4`*oKC4oOv${TvGWaUyK1xdr|!MIDMv_sXEvuqzIb?!bpoXqiE0N$gzXHtKg~JG2=WY!f)>`lTa57oyA} z#Qlp?DdeCN>&2i=jk2ZcDCPu3YN(j4yphl&#+5ph8yJJ^BuVxI9SFUAI@KZf7R746 zJz`8^ID^?cwLwKjjVW1jkwD@dJ~ZHtW1Hp2f)2SSOEA$7^f6S>9=N6`A1QW+54-e} zAq9$_%h2kiiY}Xl4M(t*BhE4rp<**OEOjDV&&G`$ZoA}veD5#*eX1Er8pHcqvVz-Y zFt@aI+0mwgn9v~D;0tgD__ii>u6lt-2NA6~M2|6<@h zR^;+~BU&`OJQ8>p$+>@MUW{x}7fHYPi8TJL<$p@x{b6KKv*s2^{SR&_Kfq$WoHz#I{he#muM=pmD$yf~xGtw&lTXntcC-8icRs+Su0m&HBuQrC+!HqJQ6Pj?U>-nans zl}=(jus2qW^Kk=A;hYazmYIc0$*l7V(#g#%cHSrjiMObNdqgpP6nRL+wD5|;AlK!Z z7(6hkL5-o9`vS!wgB~}+s=<4bAu#~O*|LG@C5jVoRk{_u`z#=3jA>nn+GFg~34WP8 z8gudXe*4PejDTLnpl2(yQ>N|mL`D&eUgb*Ggmu$9sB7;+9Rv&-a!IB(y!~+W;QPW} zMZGtE*09SRc|K2j6*8=n{WdTo9}do-^=|9kF1&?dH;l1?*9X(X%E7~vAt{^IM((*^ zfVyFzS;k7^D~n95%f@t$5=Fc4fzoN~-(uG<&b97s!o~P@V&t)OLNRd#G|J6s!6T*N zn2~pf)*@?Obx{6n-~|qYlh}{~=`N3|K!<5-&xFtTB-yj)9d%UfnHp{FvU|R6=IiH} z!>pKYkRFKwbkyF0j?oW<{~uaYz_Qu2@{U^G7tG7Phy6%*)z>=Zj*HZAVwW9Df*!Wp zi=TQc{4=jg@p}I1snzsN@i8cY-XAgvqQPp7Y{H8j+bddqJnQMS*y-P%NELw-qk^T1 zAdjvGGwHr^t1Op}54*?Q1D@df%KinVpeNq|ZCEcTipUS@oSUlI0@eDZUWdZ-_;Cc& zuW;-db*F=Q6m3L`r9NyrV_ju?-1y!HJKwV|LSm`jiG#*vn2AzE;eV|%Y;K21@st1B z*(Rs$*VWYKADUfGkKtKyUWG0DH_LWSGaEs7{`a;2BI~9SC>tFW%uS^jFrzn9F-D&D zfNuIWe}zfq#^5aSCurACs}|&{OT3__=jg(T#9M9t^-Tl0zG)n%ANTNA1RLL-UDW1o z#zy%!ztcoM9xDdqyx1oB7C=s?m~9k5CMLlz!B3~!$vlZR7DS($);I4q6h?ok$e~Lj zvei$bpG50aMHH5VOq#&^`awCO^BbPlM(pz@iW93BET(FPy zrL1Yu=Y-XPaiV>Zx%@lO{Khn5azMfrA)rx=TYzzSkX_hq{4{-?HA0-$Bg=vYlfwe5 z;4ukFn?rAzu{x|l(aSHL)d&8H9CZf0bxxeX0FuL$D?T|zWHzcrmO3VvRUPkGb~KEo z7JaM!-(R=hVO|lFmSykJ%`G61g8&Q*(8gr0PIa9AOnOaRxd@3jppgkLSABQh9ciQY zBysW@mzkqv880+AKX)NvS)$ggf1Hi<6_6fo4tgivdlD@i^l`5-!xVW!#UKjer|$6C zsBHJx;aMhGJr%Q6D#W?;Ze}Ggu@pyN@NUs`DC2|bh&~ManjC+li9QupKyQtFK<@<; zw?LAw*7NdsHz7!LN2+(LbgPjiL9M}|kW-KmnGo7LJr|7UX5Iz;xg?oSMqiM7DTdkpFS@)FGGnsRI!km0A9)r)C7;6d90I*P~; ztLLwL^e=&C5C&)bIf>-HG9ZK;-BEVl0g8c2mc3LACedQ^85B}=^Z?Y>>r}OT6kK~f zFHCP2*YX-1DYkTAXnfP%nI7>Bb+0ST}NeUO#qX7d}h7wlvLIrkXyzFi!NP zH^k^N=t`e%`8s)7WTm;MlD3MEg|avn<(8!sZUYqRGUZk0Y?xam|$5iQJU z>Vy~mo}6yl0~DM-(8j^+>QQI;&w6~#a%R>0GuD%gS4PfMTF9AVih=g_`BcmUuLfZ@ z9S70|6CT!n2NNC=g8Jx9q*0X?kjK9}uT=55YFPXf6r(~qF=+3itHJgBYG|9J|Eo6X zK1G>-T0k6mNIx5AXUSq-Q#)YEMs#o+X1JJ_Z=F5&iYU)4Nq|>DY58NfxMM+KWV5E; zd$;F$0Ui&zT%k|WrlanFyueJ!UNxRw(CAcuezc%Dykh}&p1tMO9PSu|*a0ZqmY&<_ z^yO<@tV2XjYlpK0eNw_UD!0zLsYJE~%mCTzu~AtoYmh$`q`5b1%D(w<@#&>)i+{hB z+C_hLQuy`-4d&V5lb4wRnIji}LPCPua&j8YzTu2vk5iYX{p3w+k>a#SVG$C@8;`Zs zy-;zg?cmkX{UjfDDJ~@*H;Qo_`rSOd$6Hm^WlY zm?LpAF2~J?mda`6IS%Fol-FXvUvV(GDSC1%zBEfL((`C8Ju3C)gDpr4o9MQhrXRG-`l32 zlI2-PUgq}FI&FHtYGdr`cg=`+>v-?qkuy`t9ScNUp_oe)X$GbpBat!Ko!F%pSOaFR zk}pdMs}CAfBEkQ$+1JE*U=Wy-Nzt+Z1GP;$Ku!iGK&siWOCvO{!W6VW7Gt&X4XvPqc@Ze)Zt`0cwvj&B1g??y(0+ekipy67Tz~di@%KJvaH%(}J4$v; zB}Xj4wVPrJD9Cz)gUVACh+gHEUyZG@>cKoX7uR4G6 zf3Q0C1bIYm1rEBe4a#~m@y#I@(^IuG;=QxpY%!)*|g3A`kaM~Y;>lxcQb zJllU%+m_k(`Rp~4lc<|=>G2Q$w{-;w3){)xua!q~Eiku@Vp1rwiHh0jb6D|2+T{t2%$rpms&@L|%#V1(E-1FJ z%oTw$iB&PM(42rY0bp=_)aS(Ch4`7FQ8fJ114Qq{0;bAhvG-C8q=*+#F`4uu_d_9i z!6Ww;%@*H{{O;iEBs*-2Z>;D^aF@6p@*KDLZj+S&hx?<*ExucPFNzNZ4Y@SU?2!z) z90CrQD)DXiwVnftJ6;WnV+)}xx}Z}y5uw4l2An|fyu2`8VV!&eH#kWX>YEhB|GLw< z^v+Kv34GJPSW42O+343pf1&7Ame6Z~B&rz9Ka-nf%$dJ_=Q`Qs#4={Dg^bxr zF<|0vr(!O6uTvrg;Zx~ZrVawgwLvLi21vubF1;_@GNTMOnkUjaI`=2t0nM5$QvA!B z#a$uA5(n~|JzJgeNtI{MJL-rVNdEZF+}~R3C@h=%(p%x53bsm$p#}SDP1TIkAz+tE zHiRV!I(QB8_^I`D2Gqnl=#3on$KtdCW6t21i`-VgX@k*w>R+2GFw2kN)E}H8tDRVW zWLwCOEflkvB8iZ!LbuA&JhuDYcmFH^xpuM%>Y$&KR-dyzkKRb)O(t9cY_=oTG00?_ zl^bH;KbUy<=VoAZeE;2dNh>$I#)(t?zqEi+7sWvKg@KB}wwzeL(<-|ozau{!mMB^m zc`~pHY-vcXJImzr9*}(hN92@BTg<;E#waM7Jf+g|s2uR^mC%);m~esV7MPBQPY=lI zy^H)41GSxy0)-J!40qNnO5oiSLyBG=v{i&t#jbFjs$2f4S8Y(C)MS5Stw|LhdB-~G z9Cf}wejA+%ze@)FS%3~G#|AwHJs`(#pO@)!hw?Tb8~Nk~ALq9V4Rj8@mKqcFwqtWl zjmIt*$HHfde{!72jJEenru~7eAFKG@d41JjDveI*+DL@0iBHO5#{GXn9=g#7sw}!rz@fV*hzVqEHi$8z6^sNl)^1meh zZT6D5x7SizsT?Z%A66~N|LI!l$oIw+>?1#j{r+;55|c#0Foj)8jFIC%CF@voO^_KN z`@8qWkvt~`NUa4x4pIyx{qHlyhEkL*WVQQRYQ5(P!K$$Bv(v*c*&nZ^C>v)Sp3zMo zeSL!m^yNgq{6o5uU+GrN|3rL@R0koQ!&5{43tj*i*|+HbEq6Y7MK2g zV4C*`nChooAZcUiw@z#`9kPH-0mT4qS~eBa<=M@<0=sK8^ZKqMHF_|tERg`5+r>-K)*1(be7SIYE{+LGM! zY|-qYbNv!R*GygIzjo?Mw;jUcUi&4rK?!qfAoQ`Dx$a-4crX!-#MUJmA4qJSef;A- z{nuZ`&M)3E>N zRE6rHrLXa%Kyo-BMcGMX^Z^s1v?ywzpO#9Cz8mn?6i?sjw>L_A($xXtWqfFPZgUzx zELfiWY4~j~Yl=IkP3~tQtwVbmwU%rkr!yvFCQHhrQ0>U{$(CpzN=wD9vU>3HuXUR| zFC6FfW?3J5U*@EkycS$4`$reEEGhc(*B_GDS4NgV*Q-&!y+n%HK#`BA7^D?1kd!F$ z=v*X^`b34=u7eDn7p|>M0D9pJj=9<1eEdJH%j2Kt4Rhj6gr$}ry9XJy%`A_7yjxyZ z7EPY4uvjQt*4^TPNjhN{9PfQEDk#`2BtH1Cqk`<>CM2BLiE`3HNF1UVp#Rwq{@vg% zL3~gSeU@nfCe&id21)FKYvRL^c$7zP3(#gry6IsT+>?y`p2`9bg*3S8RC)9h_X5$d z%Qjip{E7uv7G-@apNFC7>)%7&1QZyYI8Kxy(WWaiBt2fs?4^j<0;=b9jTc_959+?z z7Cq+;Gjz&>|FDJ>j^$-@VqtRH0w6~y2Hb9C$i|XD?m$C|Zn~V_;f0w7#u|{4hyzjQ z!M|$svwZAPtIl)V_iA;e~j2CvdAuYhF@=+@+&iX;&;UPk~SxHyFIl)%I6ewpCWgu zm@>#5>7+CH^+7vjn4Pjs)Bz>=RdljsePou$Q80QT7i9y3xhUD{Ji1WSDBt#uP8Az+ zLW-;+V6}7!PI}_zg?;FTU9NZ(36R$Vo6l``ZBb=;2yA-8201f@)0F6*beexSwavG0b3*5m+my-+^A=;mnI#g4AA9Cc}jb`J^7 zh+A@L>v8fCx2?{J9WY=@kJ5+JC}t}~lBt*m;ZtZw9}kOOB?jVT*^uH9DHdQp6;!}O z(CH?0Hmnj4NG%q7^bc(RV{TBg|L3#KxaHoe>aTkgpk3G%(#qsRtJW?2J%J}@ofK3G z_DsLBXoct$T@9|1B7Tea6NotGsQY;anVy#rs`u&x9VpZNb5*tCRNx8kV?k-t^8#Bm z?ZOVl9(UO50xty)(hb6HEOCPvWi=$oVGs!ut?UIFZisl1XfWSnTVjsaISChP#oBKy zdlea(TcK;KDY}HKewJuab^cOO649Yt-Y7Rk7;h>E$Z;sC#vSG(P2cPqU^{G{d6=hD zf8o{}v6gxqzIs7BeK>$6OOoTdj4^!T5E^4rJ6vLi(9BQci>=cgo;Spu*tlXzYUrfX z0v^)oqWxY5&@qsze|6ZgpgMZD8>B2?Cd{Bmq9B0>^vsGXv8mAs7ULUeG`E`Qd}-(W z1Rgf;#UkusmnWWPB|HN4oi(7(WUH}8w!kkJEH4AS4FAt^x|C^asgeKiE@v?{aWL5+ zqhJVEe6fS6pt9txpP3b#3v;hbA(y9;M;02vU5dF)ky}(uA-#|8;B`Z0b4ExWeH@5Y z^8M4JS25aZ;eIb|CqG4*qy8)auV4crJ#?cf5tqP^52|WUB3`C0={@ zX#pP|h-~uDlpK zqMyU{c}{|j%Kc&GkH79AOXPF~!Wi{w*GHt1S>w691KfHWy-UQa!nEjKkA%N_l@$wmq$y$OxWlT-BkB#H zf(z>PCpM{S zEHK#C}7IHB31k zIpEX<8tDv(Uvz?gouh66Fc~1Kg;ffeGF%`U1UhNEM@j964XeeS;3>Nl`oiOHhIu2X zIo*GXtQ-q9PApe4El{(WViGB`fr@FBt&=D4mZ1fP%u&mZ1s)ae5hG`?-533(YhyDi zUOJO~2e6~!J8S2;TNk=sXSP~0B?$Irgn9AF%+48+>P`HGQ zL8+?ibOBx9pR2wNgjJWsD?@6$*UTJ{rp}(Um=bq@*k%1+I?1d*J@`ap-d7A^Gtna*<>sJS}Y$yQVmR* zxf7lnb|<_;xfgm~Xt7nS$r9>NJ_hS%HIQ={gq=X}Q0cSLZ_fpdF|R_pGYW-E*{ z{EFaB!98@Q?=`X0_8w+ee#Y`gzrD_L24x1|Q;(XFr(zWmUd1^W2_rW30e)(H4HU z_vVFJvzO{r{p1p_U=H@%LBii28ulaKV!ob-;zy@ix{+Tn=QcF%(Wz?aC!iQzSp56_ zi+^)Llc&y8cZT=!SBMNgOOXlKbbHU@<*^jp9{29_90F~^xDDHqdB}6~kM{yQ#`i7H z5B%@d%cR`nI+R~ywv|qGM!7DknjVlf0cTZhm-4+X8VSkQj{>y~n5NE~$wYWLSgRft@F%kReG5>q2^y6SVmI0>D10EO`(Q+^j?}zF+doZMaA6kO7=r1UZdO~J4(8| zyXAOh)RnFCs=fr6^GYZh@qVN0HilVuP$v zo~<^>4nvbmhv8t1h2sQ`F`jqSrQe-7`GvL1*J-OJS=_q@x>SnpI`GYzSRH^I-T)mW z(Q`MMYm94T^Ex|fL0RTy*Fwg!aP+_3^u{mDo69?GuD#^ySb+d1-a-Z~wvhW2bC)8w zsTd@#!ijcP|)z3@zGgXJPd7W*BpzX>q*O6m_kR5`*@cCttwGetm_9qsRg?^I4z&+s1%a>}g z1#XHGT?Ht`xF>0j`D@#L(Pq5ZKBZko{;|&xXq{c>v|AHPzMO$Rqqqu{kzJk`C3?(D zkKPitANo^tLRK$sPeU%*>cas>4hQU*vx{z&;|pG`SOvRRZA7}rvHO?JOaHt8u*(v& zO~NiJ#@ek?kK8J;uI+!`{o=%7YnBRrlQM(p4I9EnDsUZ(2bT4)1tu(yI`lF-Olm($ ztg_xwo`-}JD=;kkh5^bc^Hrs@2R*gr(k@`XyXI!f24_Hh)m2d3x% literal 0 HcmV?d00001 diff --git a/tmp/epout/eval/events.out.tfevents.1586543049.iZ8vbescrakld4m4drzcktZ b/tmp/epout/eval/events.out.tfevents.1586543049.iZ8vbescrakld4m4drzcktZ new file mode 100644 index 0000000000000000000000000000000000000000..aed70db3f2a554aaf335abac9d105576f46f2dd8 GIT binary patch literal 2277221 zcmd44ZH#3{l^~e+y>wPNTve2>4P;-+*yC=n=Ye@uFWnfYU8bD)vvJuqRb{u--OSUg zdaqoM%6ys5_sXt9tcK6&mS~WM-R)%si(2wBA6hYDmr(=LEw*=|VRvD7R@-72rh#qf zZZ!*9X%Rgw@dL3Z;@-ITM#PD@C*npXSTe3k@!)g)^loA| zj2G8;CTZvRjqKwIyQ%hnrZ+t|+I?a)8*fgY+2Ow_c1!tI_0O#5ImT{+FLo#6;cT=u zNe`#>#T!rd3U=eg?b!mbH`z@)=@NUMy|H$<{*&_xLE}E0RQbU(J85eEEK^Qk!6V#5 zTHom5`F#5PEIpe47Z#n4E^v$}i)HSsYlOCrQ1k-e$E*xKn& zc7TBU6Cl?&EiMgrCl|MO$LvLR>!r=nVt8q~IpGvD@8<-UAH%VKx0e|#ageR7FQ=V@ zN$+-fhF5Leda#>{AFi{X%YPDfjU7?kK`YejwKw4zT1ru-Jj#A+~b9 zV58$*_K}38=O?rI_U^e&!0aQV#b`cRq{z=Q?TGs3uVjZYk?<;Z_~x*LEhmQ+`3CO`w5?XX6cyHAxE%|9_^_N)6{3f}%`=NXL~$ zKm)hmy0bf(Pj+9N0Iqk((-}OIi|c9U?xeS+Gs5q%-&#-8n+{5T|9O&j556Zse+Exf z?f2HTzap#j9^<6?@Y%PrlXX%BnKlMq9!_fL9CDI{$8?y};e(r_tqbGPL!55!e{=6w zn94FYvIBz~y8NG>!8Yu_N&v{6#dK@>uF+x&I)T&-YOlxvhfM7Do|7{1^v2skE_0dK z-MCdf;iao{SbO8>Y#XRLY zrvD&oq9PSGD-A!!ZmlC>2-?D>>GSjSP3(R9H>;{7sDt|ryiFDvYU;CahodVK31h1KEmLo_4*hMV zrhW^%IemrD)a4@24cK}1D*3nb+vgU$)7kU6YMxqtp4At(oSywl;8(fqiwz#$Ia{&+WW#6cx1ld3nXyeH=N_PU7G(V(=UbZ06NyNu-#l5aHm&jpI!D2 zcFP<{XACCUoGYCD`C@eO3i5im)=Plx3q$kp#!aP04I0Fdy%Je|ctUP~f0xviEa%3y z4F=IR_$j=s3JB>T&TD`$p`d5QZD3f;aMCP7@PQFp9P8;u+M0CqaX}CP=j;ju+T# zFHCm9D%-j+8G}ys{Jan9AZRR)OtD+nn_s)VhsJpb(8E@bkqOTo?A^vv0oq;NBho zUv=11t&T5WlJV1ms?t|zKgc2wX`Pgk-ajyI;uY-h`c3Kb!QcCoFrQZVV>O>l#zGC0 z3=@h&zR-ZL#QtrEooa=vt*e`9_w?1xEGo*g81zTMYJtfl1e08ud$0}i((c3Kan>6D zS$j#d=cxMM!Nb~ss+37@>vIEd?m+C75c#ef7hLnoy@x6{vX6Dx>l@(d7k=!O{J&Wf z{8=kn>H!A0#SPEc6;LIFYo-lx_iklJz*{^8x5^}sdWa>P1r_`66ZUIOsNtT(^x%_Y z@OyzDsS!0&LkMD10&TM$cAI|W!XvG!6kghvLqSN`f1R*bTehsc;SOCbb01jsd(+Gz z`Ko}{Z!G@!XmNS6``n<>jIaBPG23Ql!!%5d8g9K>MW!QbIG4zmO2 zxoevL-(Xp8vsdb{Dwd|>k|o^>`<7^ZcG(vDReeLn#8Qn=mvMpHkLo1@e6J&H7ttss zZde+r3Q78do3Q`2V={z-QO#(@?>nR1N=Cya?bpOJY8swaQy8^+P2;$YDa^*~AL!Il zNZR3vT-NqW_r)#J?qM4T!GK#@@4QFzUDzxb!Y(CwAiD^bmtlIl82bvM)M&t@%V)sh zV~Qe$e6nVgH2=z=5*A=8eE3 z8gVEB6(^nHExT2E%M@m+*Q%tC>!e3 zrAy8GfFfDD3cCh{I40#jjUH>;*KC_L#zY_x+!H5${;Z~gU{gC-9BJV_-nvfoSCEjsm&#^YQUw0o26QedrIyXSW%cFFYHkLHkcy%=sz=JVm? zg{zR^a!txOgnTpyj9pqAo@jHQ>yCYv>(bgPU>Oj=8w}JM06hh;J<35#Z!G_4y2{N1 zUZecdb$J=~Q@+!5xS5-YIxvZmo8?P+aAO7B(mXd-Vfa-;@~*JubzV{RIYiAiH+y&! zQYzqoG9T=jek*ffcQqSm^H(O5onao=0#VCTw5DT`v<80AH#J}7!{T!Lc4YB_Bn4Ua z!Wa;Q$NhG!dqj2fKCjoiEH7@j@eWZO^_y_iF%xMtX591AbUsy&HASBL}Fvtj)){x}FB`snp4st8YqXV~_~f{2 z#*<*F(1=&|IV6?X_qe&)73E5@II?L!Fb8p4!y-d4r}*9xH+d*C8E$-V1?EYwD&r_} zt(hMFMaOg{BT(sSvJ_01In&s}u1XQ7*F6-Gc*<-url8)!uq`lvG?zOCQ~N=_pG??m zbRjCLw?NB;8mdCI&5r+L{g)8kF}EOdWz$l?!&49VKXtk9R#dv+MPB~tM23)@%o^Wk-=iOVmBhB1RaRY|8RXS4( z7&UL?uNhiM%8srd;gxS(2`Nf?WPU#T073xZganU|z%DDxH)ca)-Fy z=wd$h-yua-IwCZ7X+lKBFCj2EPy4=vy+v34s?%<5iw0G~71RgGc=aOp$U`Hlyribc zPr!4IJ9P$dDPalEjx|d5@I$1_r=j{fGgJZbBxk{paGzj3!jjY@EEV>>>gaiJ!HCS5 zP@P+B?>xRc8c!jd!4s>izhy+6c$W8<1}`Iph+|id6eHCcJw>?lble?bIad z@^~kRK5%Nkj9{CYO<(_$-+x10a9G{6ycL2`IZz=%*)Z4DJ#HqWgHlqV9~-idCNzAV zoxiZVNN*7TF(b(qfI|-jQE(DYOTQ`!Px;x6lF{ny`D4x>#b6NtIx6WbaB^k zXi*VR;n>rb1D^uehwT$}gS9ZT#lS zq^HD_P`I=pA-!hXlYU>&6AkZx6VFqZ*ThCF&4ugiA~eO@?TeB+Sy>mSA=W!(LTa&v zlC-9l(%GFqTwJ3KvFb9t%Qo4a3OLAK^_&rGMJ=PsaU{}iWm-qwGvz7=v;dm$zAc3} z4@0C@JBQq=%tgh%&aWXOAdm1nzt}oC%&Xajtms=6!Vq-MNkRO4ZG-7qDXd~ zaHEm*TjfL2<$|eluiIMD?MMCnKi05V__6goX|uJ5c2&0I(xG$xk{rUbSN}~J%qeb2 zfuKl^cF4v|r6T2PNQrj+zv4a{tLZCYQLDpe|9PSbonqrE@?g}4oR9$<4A5#9rNcSR zK9FGu_Q8bRtT0ur5( zZ262OjX^i}qW44nYo4xMfn#3K4-P2$m~9ufE9+16#nWZJg{(fQ|G&9HY&}zXaOf8=)@%a<}RSjPVAvVBq=C;90K0#GG+1#(T|)c?QQV^))E9Mg>w zk^K%vu}Ye_>7#)-F0uOTztn{oKX$tO*c1{`5faJqgi%$VA1_^(ywel-peNU=p1&jk zC*r#x<=9nrRFSPfv^j&SBA1^wQ-8#zu%uy6BQ#ZqRxMs*Q(hh2+$Zq(SQ~cB`~rhXH3ZwG-k))M)}K!{FR9)j3wXLTaAB!D zID2ox?o%YC%-Fo2-IxA)Eo0M7NA=SvinD`T zI_%qt>?THE-3=?Dsw_5$_8bD3z;r(HRVwopY>p(;rWPe;(E(?F(P8h13t(d<+FtO6 zR7ktMSSa$ZQCl%<)PB}s|0phYOGm`s-deUNR4;ESiE~Si3WxkI*pE8wbX5H3i9+%3 zG%Gp*#uTFaj_v=q%&n$_p`Wo$A>t@wLxJdAB`?Y$`q;3 zR0PSgS?cebRMLxpL!6gg^#d;X!u~^FIlCv>?05?F0JPhsG4itZo)jO2CFPa$^nedb^1kZ_#;W8-r zN_LZIesqbChFhd7>GIQ0{N~%}!(`asBs6vY$|Sy7A4;&*S?U7ZRc_Pae0hfQ^j2#t z;>K`a5KQ{G;MooO|48&#P(f4M$5_nv{Q~YA9roKX@2%04CmBLyREqvGkf6X#Z!w)o zddY*Iu4-#e{Gld2`%fMAOia9%M*Lph7R{urFyv<8oX%Y>dI}q=l(y}Op8Zvay*(n{ zbK^bsO=i(W+|?vCsv;FdTW_@FxH18u`g+}b42CxJd4f^Thu4sls0voyVE05- z7L$MpgaNp=j*@K8k$A5G_8p=gU&CsM9$$T4sk9MbY9wF=%`P9J$|0e5*4q~;Cc5Hb zav2a1bSa8jkft1OEsgnKXKU#RvN6vQ$j`_AB9XHd+^V$vKw)?~3<%@5Jvy&Uy|_JF zjHa{6uIS%unDOk7JG98DgidGiV(bJBHCao_z*)-K#KL7$_u3+ufk%pD)RoxNSlJ7E zt&i(QlvG9-#PP_2n`n%aY*CJ+BETzb(a08fDF@l#CbBVXGuamQcV&osq9*gF59pcM z&i1Y#rAF7qgba!Nx*#v@G;~9p8eOli?$+Z8rg*5=zkY~1h<0O~8eOli z>;{iyj_)6U)aW`Lz($jt6kVt9CwN0LUwCxAzF`|voL3%6_W+|t*Xx_R$)_~TNzrxs zy4tPLqU-eK{f+Y&EI^K~lL7IFu9Mf~rD!x)h{ux9Vo35tcrHVIzh=vhQKReiW$k)< zqtSy|?J=mvRN!&Doo6PIqU&{dbxocW|E(`Ti>}wda0(QsrXB^Gcto!8b{RFgu7Ij( z`rg>xcb=g}*K;V{rhBlp9_pf~4y9;Qj2d09zPF13Cs2s_9sw4D0Z)ytR}kuqPqJvL z&uP*1>h~3<7Y-gLr$@jV&gZV{BS+V3U}+5QiA9<%9_;(A&Pk50=bsP_dXi%}UvXbd zdTMk%Kaye8dq`)W=z4Vv&^8YGaY+|TYYU&fNb}lhT_0o<7PE%^JzCL+CD$5ieMzny zb>Fhi!>j~9R2SFw;blSFmcxc>PFU*}NIQa4>@yvjdrtwks7zN2G1bD8K^B0;6Coc3 z?IcF~@oM*FM>)f?q~g>>>&ixQHH9sX`PS}SMHlTz#%FLkt&SZB>GT`fEz~b0ORy1{AzVj&W=gy}Z z@#gNlLPg~bd%4;&*HG{AI!$`xc{y-Yj^A_pLJKrz0YLXiXzm<5l(&rMXa9PlDf&!_E08;0T;G9lU4UH6$=K(K zaUl)aB;iL=9<+s_Wl_*)gRsxa>86Tw~+V1d0Z>4LFt^Z#-^9 z_GxWg8TwWQAW@nc?AB7C! zB>kgWfdfNDUp2BO?^w_Apy(e&@`Kj2)%a?R{HP6|{)|X|Wq)82cZlOjHT}uxXe~=xZ-J%h|Zgo0OYXX`>@to^lfHfg^ zF0BbusaLHAt5WZ2O~{S+-n8=gr#&&CkyvR)6fPvxidZT!7GS=v?J2qaz}tLXtCJfknN(h%FSX^6m$;){ zh{E%ApB`9pBzoa$)Syxc!{Zj8@K6YgF(GaTzR&R`*6iLkPhDh9*F`K|ydHwE6-E59 zZxYK9Mbh)(ECH-gR|_)mA96Qb#}S_B@X){4!?{*HoSx!o&F|SPfTP#a zN+f8lFL{>Jvxy`_uacEZ!g;;9+3D?V&T6zckn08<kQN_O2TohnHsfz<|G-VMe#>xLtu!nQMlV+g)t|?PC- ztc8(O>-^!8uGnXq9b*O5Wtg!kcNN1@W@*z_9)#`<;q~c{(Viu3JZ`N(kun$xNZ9IGEHPZUy+66=6wS(@?qcs#~XyVleYAB9$bs+!WV&_%LH zl;nJ}qTxN@+4^DgycZNPqF^CC_#}7^7t_V{#&7|;NX6aJP~d!{!R70HpZU1-x@Zt% z)o#cqCD9c(c z)mq4gVa;DW;N%pK0wTA-;RLlFhG0#%4MT7xA>t9BQ$&0$VNd(QyS2RqQV<5Uf-?)` z;`r)CJ}qnrLldVLVFIHmDqL9w-1m=;U}ql*p!NFE(&f`|z{cdRLQ;b|mqo@zb(wb= zS--Djx2)f+XZNV1T}EBG(oT{pJkXgrdMv0fC`tnT## z345#WyJ9bQ;LoBBR7C z`J&htT3x}}`3t*?^ak;ttQ`)?M=pV6f7j~XAU!Bg-fn^Q9Pk!8PzelOEb@BI7DpDz zb@uA>1Z+8 z=QONCP(e=areZy3{r&CQryW?oZ=HS>7X-TsVV!mm2pXxq;(h^VopulioR(c>rd$0Z zRFewaO>3+e2WfcXRAZQZwVu#rpLP%kDafz=kIG9qhNJm#0ZIDXyV9L(&%IB2ua+LR`)s*>dD}y~vcqi`kK##52nY zsXS=gbh%ZwYbhl}wm-A|k}IUKv~)`8z-ISHK=&8C$Q0i+?ZVEMMoq{U#eG=PHjxhm zTx8eSYKeRw+=2D$>YQ`SUH6@=YbbWyLE|Pe?n;vzuNFm6A&zP31B3iJV6Mal9f9mi z9x#A($W2(hauqq0yni1wg+Q^ebYw%01Zai+KiB-V#ooqKZ&m1m0q^+SCU-oahQ2st z>=j3)(*Rsh2=@NK)|3H3-}%w(a=Og_m(J7zj*x?ST&sAws%yoReQHCjuXtRm*w&LN zilb&^y@O+WkI~1kXy8WQl1ly7_CH_sK%dw)fGgmrbn+AU4#A%`DbpwtOlMIHb@}tUIy;1fwIKjRTMtU@*f$ zZ;!K29teV@Z2#cRYF@uvn~5a_Pq!62uvFUfxGo*7ZAh7y zF+LiFdU9m|PtOavt1O~$leCz$qQ&Mr;p2zzreafZ)hU$_4#$6S_lRe&4HcMG1$uBP z-DAK|1|)t+0jr>Kbp2ea)sJV%XE1b!6=AF^QL~#Gni8TbgNI z_>1bXHKjDrTGj(9$Ni-i)o1slZy;(>e80L5z~ok;tQFNb)G=Ku5ybFk4*6uK-X4a+{{Ij-+TPS3D zwABMCmMSoO8-p60np;bJ`oAOG(c)BkRq_!T;aZ?v_bf~Er#+#uy&X`vrd#eN$*^^CTZqa)mPvnW@pfyU?5Uhp3@gD-= zC~ArWo;OSSrQ=!wZy}(8Gszo^W!12+CFsxkA~z)D6$HuVO?txPlgl>mZw1-<-x{w0 zmhy@YCi^7r0zAGbD%sy&Hqv45 z#d{YU2ULcY_veez#VbD17ztHt(Neev=R1>qqaIl zIfkwEDL&mr$IH8M!L{_=>3A|ew+Y>+(B41B0y(K+eXEUMMeaVs`Vde)9G(dObO7XZ zbbfhvI=eFD!{vutqxlt>`I5j)o5PdM&FRj(`r#7Pc(4nT=chCIYfJpsW|NncZ;cbOr3fl847tKXnCB1^*+9n&S0}sI{Rb>r(ljAOQ>1AUd!0K^Q}H0UjxhbN>K@M#}GqORJ4(oqNF>t|Yb}MT@^)zWgK46{XCiULlTDB)tUl~YI@a{-W%kP~1sKZW2rS9COjDM$Sca=6|qV)aP zCu~DoV9Xii1Vf2jHeM)}cNPUWBW$T>iWWkord=isWGBfyQr>wjJIxzmSzLl92u_F+ z+Y-*G6GogQfOn_n?G7>DjdLkIOTinv`!1oN+8yg3YJpNDYZ-|p(!*g*U4XPxVmlYd zzIC0jA^j9u>SLAKj>Q!*fTSfvIKIK+ZBmY)3dVyt#kFaS?}?zSHI+EC86w~}AfW-g z@tDinK_@zcdt$L-?+5@pGaKI9@p%iibXJVCj`tlQ1L5CA-NF`J?n`XYZCTUp}P84OnZZS9CDQ4hDfYQ+Yar55R&(lRwA!>bvDzu-cYrDGf zx7dSzo{qh7JH7c%dHFC&Gpo=Nq~=$jH=jv921$g+fI*K3tV<+{%5V#w*r3k3tNL_0 zjTcC{!MfR4msS`avEExB-QeFgwmVdNlgSaY+`UkPwr8{=QhA?5NwKSsrYJMCy5?FV z9X05gr9Gn^B&`a;3wv(p%Y3G`pz<-s-83F8M#D=mGHSF)SJLIDpZLwUkz?~OjyA7OeraO!_0!blc%0aL zeRV|w=?Vs=rBY}(rhbSzh<00@8k?`L>;{iyt?wUz)Yv>7K+Og;TP~fW7%TMsN}Ihj zmes3mv_5bHKE@T0WAk)qOS_#Eo2Rd<-5M=6PhVcQ-D6<4Spz4>=E;C~#OBFsiisR& z<1&Sq$+163E}ug_vX|sfuVY40>)YG$Q9w2<0gEM1jm_8L)hms?5k`y6D_aKJ^HH#g zs`v!nrN-tJP&Grq8@nYj2WK7Bk3)cx8k^6T4Y%w=9bDDE>Z|Is?WwW(YH_<5aBURR zlUJRE&^UxqWAoLjI^&Zpn(A{}Y`*$^MG4KVd{>T51oFW7+;x3?XdlT&t%0R6xF;4X z_KpCs6EtdUJ_jw1y^n?T4f8tRa9=EbYHU6~l40|E2ydU*yhcnBWAiW4sQK(fm@N6c zoqkz6c8jc1V;cvdGx8G5P0|X-EJZ==YRJB25lJfdy>a7!*f%*;7xi6h*Y2Qg*JL7D zP1mz-(KF9>ihTy}#AFbEi*_OA=T&CUeQBR4chfu0jtuj}qoLr8dHPy* zY&}bFI{0(>Pr~!-leBa25d3>eq9W-XVk_sxH@qN@eF*QYV%%ZbHm|<}GjJtEzD_d2 zv5l-~O}>2RQQj7<1OFy@=?l5oY?&vQVm$-xr%7)-zdM@Ecedvfw5y)+rMxDrWZd{= z6=)jLIlc?Ti{UWm1-<~)c%!gy;i_8A}K&A&B)YFu04y5~K(u>9plWIYZF+ks<^? zoUnV9XKr(JeX}GG;V>=Maj=?-tJ(|I2mwOH#pq!gz{r zXSc1tMtzD)_&GVU8N)iP8V5NIBebr|rxqM)LJdz3Z7^s-x+J=fK2#Ti_K*yW(7KC4 z->_j`#fG(!G>fF0^#omtNaRX^g6Fm`Ew)D2(gVdGG7=yF zcB|8QS`&j`m^C4H-gj$4ZoL22M6q*xYeItLX-&w_Jgo`&*_yW>MN)Q`XZU#OrGv_v zAY~>g9#JnB6O7~&1~yHT-5Yyi@JnV-41(>6fv-K0gDwdrq1{5MPjlB6Xm_QTg+Z%e zQZ8d|LO_Ee&7vTA{iJ~Rw9*DvdqQ|6&x3Xr06+6LU)T1O+6u8Syknk9Av zouvvJt%~}r@G%-*KBQvn;$k%~*%vNRy)&0mYsf*tjE>9vBH)YKrh*by!oAVgzPAS1BK=!wSWxjJ=k8Eh@1?e(P8Gpv| zig0LAuZu~FDB$aIf)5S%B2{sHTK&H_nz9HKQ;~WU#Ne#%N~uKz5QFl(rUVp&X7sv$ zZ)`@!gbQX~H0&joio}AI{J*q}0-1@jmU&TTJ%yNFI}1tZWd_4=Br*|{A6*bXDx;J? zY4(ID7hI&^R1LlYu-BNJr0ggdR1^^7)xdPuK}lD9w04sdr#M3yW^D2oDQ7OQgaX>! z1Fk$L9W>6UkI`POYT(#~$zr7V3CQ!$A1{3%C)oq({T@CLb$5@~>qpI5SH%^mTkPcv z77Y!?D+ZNsP0jdGXm!Q5DO(F&B#T5zmR>V1d5pY46*t5}H~V4pC>j(o`pXuC-Cis( zXni%fN7(y5@lmSMnPzQK3n@NNl#<~g_F%dmBpH0??1D#_sH9Y)!JN2@w7iK2jLzUN zw&sRBt^UAzF(14Z&1g{MzuQ5lBqJVY4PIS3gkjBJJmBQyre9YvUA%iiwPaprjZFz9 zAXu|_!w~F7j8xs?u^RhW!k+e39^|yrJ)Y(JkpNn+AFY^IO4s(X$e5rHpadCV9ZOn1 zg6zxqv}Gr4N`yFO{$gX-No$Hm2qjHYr-HKgb=Z@kBzKN72iW7r9|=ClVteQD-O+do zL;lJe@=)3#K7TfJ?|qmO3^U%F8kQSsk|Z^Gyn_GY-F0Kj)5ar#7B{|nkxykDPG@5n zr8nmV>CLQ1=4AE(e6+QbpDg2qZD|c=g8$e5sjh=sXIdpx~!Z*-&yP0ir3Ydokf5Dc^y5y5wi7Yb+8j=9i?FX-VGY5IU~7sjP(!- z5Vk~AW*wzqUDpjVWNZcqVj3goZyoKc==B|BU29CD9Q3Rb-EW0VP8_E8Gf@7;I> z`ISHH_$YNbNfS?9o5i|imRUD!ahpD3ajp7#w%{EpIb&LSbyIsIGn+x2uMer z$IaAeAEjUiQ}Zi(Vl^iH6W{}-0e7Mz@QLVty?>^NLFbho!_j=Wn9SzeyH(^#Z)9|k zt*kFY%c7+BYFX%UWaCiLT_}MTU#mTGV|mN*>P*w+wf(ri(DJoHp*e7G+n&uA5+u`+ zBNVxqhqfo)g9d~Ty&waVE2CsFvS<0&GJC`F+a|a9Cm{$d zwx7shIiY>_hUEi2>A(wn?S-F{+@$u(ez#-kc!B_>rO>r-YzX&=r3?+(I3%G?5Q9%b zoplNaav&p_j{L3{>4$i&JF-)iL?^GTZClP&wri<0M7BS({ZduWioJNJL|yi{kihl_ z3!1~=HK_QeX%}|3^l?JIkZF5(1ts!Ho zj*x?ST&sAwT-Um^MD3UL6=zgoyBPw%m(t|F(XOJH}f=<5mljX1FJ%q#o5AiM~%K(aq7FgEk4IN*>%LwgSp44!y4 z4nSIf!7R>?N#k__bk5XXO`%bZWu|sp>{cP5@7U7&5Z+O2b~6|%gFkz7D_K&`StNC$48@3X|4?5>3Jb{l|>Y8`-V-JZG1Nsn~JMWF~=Hj4fD>mq7$R4K#%$q z)B+AWT;0QNSgVC4BPNSOkj|h@D{Og(Yb3APG(-%0TA+rEy?vu~XcGIDW?C2iA|os# zwGC?l5NJ{Nms(Vx-IKn7s73Mp)F;g0cUus(S=Q-Eiy7-^{sm=XkVo75}B9>esqm_zB!*8Ca4-kZc<|Fqq zxz;SUrh=moU$_`m?&AyphONw6IZ9zyRwd4Q|=#s^@Hp zLiCYg;An9wy(*#UQC1JEwWR7@oL*f-3|QB50$)nFB_!CJ$Ti(WE(tu zc#y>)X|qr}vKH4CXq|x>l+EEyQE=FFe)Yi%m0@v`aH(3YX#r(Qr-DzGf;dgZ z^Ff^RAeE-jk~Y{!Y{_Bjx@|IGc1HOV*YrtQGdD#eUB6P4*dE zF~9X(jxy;0Np$1=;RS#8P3z^))EItcMX`@2>0OpKD+X*z5pk9HLl4uM zwak!FiFe>~Z=RcBfftIl?>flrPL^g;3nXl&>PcxCkUG7fo8lJ zN)v>cw@OfvJy|Aki{ATqB2Od-tx>XuV2vC(uZdJBMGS(Ys3{J3-Yn^tj%x+HgH_FlZTv2j3UmU(}^fNAG$okL6jl^LyWo#Xv@KRU`8AWQEC zQBUHS8N#zS&_R`oR>azzqMXu3*~xWO_USIevYOqvNDrM`>`up%@wrXt=YkFyH5SN8 z4eQ8i{3>$y8PW%E>es{%XOowdZX_5VSfZf40LGK} z&#-Q+#y*y-YxD6hmgOav^M|73eICL*hnUC)I$pRs*}dK`B8~Onk&)yz&tL!iA-hdJ zR<~}XMImSB&VxKfQ>1AUyQ4c#qda_$=cvQJjkjoXuHKN)R!iGd0#7!%i`C0OqD2fm z@V3$CBVQ$D?oy1i*YD7tp9+nX-w_wU#~vJ>&TUGpPfZKtfW?V+S821PGGQqO zEfcT}ZGkaoloJdk&!CxAk?tT-fYXDNQ2kl-pFXtCgp|ldMg2iyAUjFsk@C)K*=gPg z%ixX^2&u313SFh9qS)zfl?%E)ip)(L<(X% z7stNptLjS1;4SsBN^QsDiWoqcJi_q}7H^R!!JrBTb1XU_)YBL|5tOy25+|oez%S0{ zPC5Z^Jm#`?VNuIz@Wf)pyb%C)g7)5y&s(Uavtp#x6juYk8TOe4m-Y14QYj;cgE%QQ z=ehDR0VG#t{yrLMt{%d>X!j68OchDAPJ8_uyq~9W1y?5Vt^QDYuFg`I?=E+Ihgtf0 z2E$n|)bup+4x+do6`7f{i{E&DT37vy$*qwN&) z9&G(Q9avVB`MSkSIC(_fzq%x&MVrOExApUMz#631S0_XJdAhc%8-IG3U{5ss>a3Pd zagEey>*whLGepg=owQVG_eBmtj|Z$vBnn@+1yAf#Xx+80-!%+Sv$SWlF0C-U%Pis) z3h%g5AJiHyPs85Up3w@O&f-|3AM3|xdq(STD;kjS?KpY5HR=Lvc5qmcj5^MZg4i`y z_QD>q452L^_>^OkCxHVM$5tnqs+B|&j?L;>k8P=2nlhqvYS`(Xn9KahCm_?)(!P-# zn}2b%d3E9+o3Cj*P8+(nM?w7hvB|Od`bUZc(iMz;-BOPncCQ~oeX1-j0dj1V{_OBm<1|T&yUtie`pk@P_Etg2KdHR02&EBXX;$t^rAsWvMAjjtE(3W;PDK<}E zSGzS@Y@WV+tk`^g$7V~99GkB%>!fav*gSbnF_Gge#e(-(GD>W|z8?YG)YyD|Sw*d{ zy=>0@X+=&AqUcBAXG7V?1NG$Cd>vk0(I*8U>I=|f^Yt&B1nt!QqhJ%){~P5sIX16= zsu=>_*savxD6ls}NsZ0tP=;Ifp-zfwUsa!QS>WW@e6_eD23%`n)_nxn43>8X?#Z$F z3PPRnNfu4@IW0C{{l3EVxs@LQO*k(%-Ir~S0}t|%Y}6W98iRXck!b6;#}uQ+=5x^E z*!!f|e7<2lHXqs0FGVOhHlH8Ku=zcNw@++dBc}4W7HgjF7ho^asQK(fm@N6coqkz6 zR#Seuw(u%c6ct5eY~vv0AbgOXVyUchB5O z8B5V+pTRpZ8Ga-M&FS(PC>G3)DO$E#Dl$OpG2P4dPlS9Fw9}d-)GoJVdF7cLAo(=> zVcAhYvCN0Sl4^7cl%)rXKhWH0N@@-56Xk9sU`g)~TRAUyLzK52XZoPA*Ro^lS$fmK zpUZy|o-v!i=&?ia?KOZ8_(~KX7iox`2_8%XM8EI2`d>lepv;&LOKb}9K!{w1v{P>h=akU z>>4deu{bKP<>GZlmDsb%0Zh`>RnRVbDQxFb@R^e4kg|bI2#0Q5iav8=Rm530{(Q$v z(o1BMgda(HFm92S52$}O2>Yy!2!!YZi5bhIRzeW#TVQY(52F+sjMtMO@2KL(wnFg3 z3AJ}y+VFV zVfbUb|D}QZLf=2UF%$XVxQ)i|99~a5cPACdCyZzI zc6QtPYt(19l>ht@kPAT-`rS)x#;^{n#z9WQ2(9b#sYT;D57niwy_yh4Xx&AjZ`d%; zaKqY2n#UV0-~(FM0s%@J1hRLd9pH>ZJ$h|tG#fuM-J18u(^5~dchVp>nShqFB)hrV zK4sOPv!e?3w*YG-{i9le1LKXpLe&%W2MX|cju5VY5Xlc(r&P0wF!H00dipaW`IT*N zNZcWgmwDbwkI1@p)XA(RwnH-s;CkVGC>ghAELu5~C%y?IJ3X|)4kqbRZ+Yp|v9B=b zdQZZI*fM8b5~Bm|XzFqSOB>Q-UnE|C*?yV?!fC4L5g<_Sy-?W&%ar&2aiT4FZu`<= zYjiC=Q2Zey0RmvRI-RFAG0?4vMytW9TYFj)a_4=wCgjHZZ%q_C$G0XVNW|(QsiMoz zJgo`&SsQBtbXqUHbWm9nq|7A6BkE^Q6yQ}tZvX9x!7rITF$lIN2EO)04tlSh8ByLH zgI2+$T*lmlfCfdHMM3iVNdfO^r47QMgU z?n?^sPnta;%Ka26IMp260PHm;Cn-A$2IVhFMHa1_e8CXDCMQmDhBD09;y?-L)TDwE4HnSOHf#w#HGvcQ^RsYO_HQ0k5}+tyt{5}dD?g+(Bj5dFY-E# z;dC~JQF?QK`gKp^{s2DO+DTKEal&S*uQ4{*nsNElgq`#yFXb3{%%wFAWZ1N5A7uF7 z5Box^D^SJ`x5Ht;+5fK9y+Qg-p4;65>jM6)g$`6fP0sM({^=G+7Kz(A;2w{Ci@Svl zL!kWo)=(CiTCHB}tR>*&TP;z_#H$z^ty-dRCWxltLh0L4M^A`(ECb#jXfEa?y^lsXf`MyuT#E{n30+Xhw2B}&8gI-#O2G<(5`v0wM#Gtmrc$leAF1Y4s#T=%n|+jm6;|EF zj91)RKevohu>QXDA$2;KBisY-S#GAF@7QAQb}%)+vL{w!N}0wqwN(a61MWmc;1kjP z`aBE~gU%~GhNJlqI%>?fcV*;BsrayKfJyJwGDLM`<51CRD1la9t37g;dCOTD)b>9@ zm0%X&!~}2&lIh^#4Ut-(Q)PZjaSQ_u2p@W3fnH&^JX>J-vFYZf#c57nMZa&0OP0P~ zK4jey{iwaXM;QY@;N;qHaeZeZ8e_F466@>@3k1pwp=ZBog4_ud62xo^j4^k|K6}Fg zo&F}^C3}{SEweW)zikqze-eUt32iMkTQepf7K@iC+d$G}zZdpeAJ-Kqt$2w2ZpYH` z1OZA*p=;sT5bhC685**2J)usPlabvSf-1`1K*NzJss(ndesqEVh*_s_AO|Ya#HwUE z^1EK7ANmtg+a{c~JKwcb8X|z6*?y_2XT@H;Q%Zw2yFUV&!{9Zj_@-$WcDD3!LcWk` zdw2yU@_~S>(ypjp(N*SwJFtFTjdgCh>%Nn;#~%7Rk#Sd=++2a;%<*Kd~6N%J* zSzqzERU78x6-8IS3S@twhiD4I4Yg|1inMa&li$4 zPe`P6^b#V25J9EVbp0MAD z16zHLSKzxKTn%ktV=y-JsW{+}LPKrNTE7SD z2ZQrm74*9#O*i>9~df=sCE=3wQUM3QFY4_gf@geT8!?k$qhP7H)GGekg z1nCUgwBjo7a7})j>}xg+5yPJ5tRrJ@-)J40#J;7O)-OP5xgwXzrGfXCno^(LlfHqd zDe?W(C(PmIUSUwb`Fa`<3cI#Ao6b``-AlhhKR&|n?B$NdI9G^FT#Lq_L_}o4x9)NQ zuQ=JY1-?;cWq)dO|Cm_P7?Fr2H>}cu(+dZTrwg61RlGP%|)wx)ukR=#j?>y_i% z>I?sdt;||EN?})4UT`vgEUqSHb-^;67%BJeG}T5f>-al4y-Ez4S_23qrwiK3Tp8YwSeYZ+s8bJ?TP~_jJSu5um^* z?}DPQxgH)wa<>Jr`C(z7wGHG#A=9I+9!SYrQG@d-Xh{a>bKn%#+|lAxdR0QxOS?U= zR`OF@QgBfo*FuT6gIC)&@+$b7A1ZnEUN%UNd+@4n$u@ZS@F0sp(q^G{D{gHIB7t3@ zYz}XVg2Sfss}EkN42zS5lVRH;Du81_tH&EuSrDhGcs_`$c(yH^PNw~cEm_QfNY+j6 zaXB^gJ15d+{Xj(#XOqrj$=dRSHBOy$dBN;6wqkzkxg2HE0mv&)LI#ZewXJB#eq>5X zbmRTu1%LKU7u!3J?~cY(n0iJI_|??Uk0$I1ml|EY(2^KsRk2yZE1$R8Dnx z44uOpH9K9e1Z`psq0aN;ad4j{{SS=y`^uqj2vRpL+WZmjVa zBzb`sJTu6s#5-`gH_y$mzzap&cMbaP9jnIo37z3TW$R$$*h{HV4#7bf9aFFm__7syow#2U*4V0t_=A& z`{CATegzUflHM)s@MLpyx-+kSxC8+$c46}TbS8goiT~Pc@{;oH66c@EY+U{l`$OE5 zW1vEyn^S8m7B3?(8XLO?<%F>ppBo*O$v>E&Ejz90L9ZD(f>J7t*@~~h+1ELj`s6(pw{`2 zHecqR*Z>tYxcSd4{XAV@YN^X4HNSR_($5%y)S7L0=<$GciA3RZTktNc8bY{5My5Vi z4)h~p6GQ8fly{XU6Jp(LtY0e(uVgVg0YXGWYd8$SU>N?k_Ka5O^a2Dg)S&Gd?OPeX z9k;m&7;M{!(RvmOGh5m-T9?r+b}#HzPhD%ma8V+ka!eL6acp&xsai=ifyD|N^h7hD z5JDlUr70tNsfL~IiMfpA6X2i=St&|W+g7BOm?OvLUmR^-o&3_o=If`ai)y!>PL0ji zS9dBH{d#O(k@Nc$sIQ-t9GkCyc^Cl#TJ1I6#Ust~*Qbr1W`B<^}`i{+}2st)i zU)D+89b3Tx{=djW(UudJL+-Rs3e! zB**3zP&J$08@o{mqZznLSER<~b13)7vJdrOREJU?Nk)y$SKljQ0O8jNCp3hMVYMqj zIE`Z7p^_8{tsvAHpJdTgpVMOV)$c2>n8@G=_T*NJT6o!I_PXQ?AKFKL|@j@=?%uWaKWq}(sTs3fg$%whnut0DWAMI@=*_r{F_V&CLYT@dn0 zg(JhflyoQ#K2MifXMHJb+cjA#XKt&EpRjNH8N3sd;YU)?oGzb%V!`Z~5_nTfMFwa+ zrhD1`iI9(ic3P8!+T|7|uO*XxgP(>!EIW#?mU+?$sYa(jS$d%O1I>+=r=AwKyNucgC*K6rtj%yq1gC8O0ELFNN*gUhoBO z++Os3zT+k7C9+AvkEA>pw@Awe)IS@9eYWN;`6QWk`A8@dGnPlKgdoVnYk*0LbbE9AG6 zitNXD|4Re+g}#4yW3;$NnQ z&v3)qNSenRE#L!M*8%}b&mUDKn%TS24sgby9=*0RnvEZsZq57SX{o2QZtnmsXGwN* zwSCH}KW9f3>~8_qNcu;$0tbeQzCy#}l*Xn1d5#A~{~(edv`%r)Dx!{hf-XfQ@+;fk zkhnt}FY~;W5FAR6Kii?11aQ6ZK9r0b8jwA8>?;hq-ji@4w#-?VifNjm=n){4a)b&=p~MSu{c)l#cy9aBVrz6QJy851BLMZ2?M)i zsn$*ve|uu^OJ+|Dg6)ZcuRW229?iJ|t%6CpjJXK`4T?02g5>p+0^ZX~8(8g$rHc5x z?J4F@Zfgp;)ie}W=Gtg48ieZJI|L(cEPbj;CcuyF3>?$a4vB^*>L zVR+o)6CMg-(Hzt5z$cMASE1newy=hz-L9`LvPRHFlmSKd5C@FW@bV!QTNf`gS;_W% zI7Tmv#Rp zsLHlokCw9~9$@#U_t|Dk)cp|1S?g|^F8j;o?p4SdN_v8xlcb~oqpqT*UVKpO9;fdo z?78SyN@>xxpFn=E!-lbe)RIcgRN%eg6fU25`mPL+=7Uw^Au}{Jn0>I@*tE}|SEQSB zuWv8gBOdU`mW#yo>?h5|)V}wSCk8jAw)?IsQiGdPvHchIOb(92IXpPh~KHk)%7=c;O6%0UI_Dp19Ug$+m%%)aF5GEEtO#`21A zK+=YzfM7_AHi^S9G~A0+#r0|R|K4cIB2bJgkvAl3-nh$v@_AQDLV=V zja)zsL-?AUIK>&tFk_RyNI7$XB^1!+9&i=?lQYLTqdrD^wW@(*7bc65;wK=fUaucDXI&Lnpl+eOU^U;h`?17I+M)s7{NC}(x29(ND73m_+mx+^ zE|Nu}B&Ry!eD0#58r6Z>51U8Ppor05wjk{Gq5G*1bp z@RuW~oG_U2-qf(%P?IF7$>SCL7w@hcTb?!^3ADKJ)r*Vi_G~zvjbW7DJT{Mh03U7b zq$$fdVKX|*Ck5AKq&=+q(}bP$B`@U|dCa9X4P@A~Xdh%$yhWBTUubm&%Glv{I1D)Z z-?h3oNWaN*yIWvgz@N3yfdbMzxPQ9EkwxNm4!Fl--{Njz!|HD7GXL*eLs?{MwR*9$ zmVlGtEm6e4)7WCuvf6_f?W-&w;pO6bc^fe#a@tB*=+wlCM_O+KjL@M}}z#;SdkLKqPFyMw2> zgsmvDt)h#{#^jDw$t6hyF0qeNu;N4sK_PDD2*hFdm(hf)RYm?aM3^G)8N9KefEY0I{i(+OZF@u zTV`)qe%mBaS1@hW`9ictI{H5FVX=6LvJE6n_IqKklKicqagxOEb}SuF5TLXax)zQN z;U2M+p&=XB6Y4+>D;xa`FNlUcvU3O7yCcY_T4lsn;eD|1LOSF?MVeTZOhD-#CIVelGMeABcGJ6rlVAz#R}J-mVv z`9Q!`p_J+s^}-!kzplnQx7>B#$-4Gpf9#>J6B&1<$xTLuIHsjd430SuK9j7;zT^P| zNQc~n_2ox86aaJx6brhDfE}Pc&^5cZ*xPvOt+e|G-toCj?sz^8ZF9)jD~?LX9$Zie z_Wr=uvjHKq^P}12beaFJB~=^wM!*rm)E?I=UM|&)BkLU; z+adSrfj)jk12_7Xh+Oei5A=y`1GoZ?N+&;o?-26yg=EbW67SbhpeuwDzLl`Iqc(Pt z+YuNtvP6O)3d9=^_Z}!yS5L5hdXGJI6*)y+0=xVN9rpEv{YD(v>T|p-UW8a6*_8?y zoB326aMaL{36KZ@?L9y+c;eYO0BHdRvp7E{jn@g#Ia7Nzg_blc^>~exHQm-6P)3i~ zcWmi>2=6F1yBQ3X!Joakl`Jatw!@&qb6nh7sRu)q_E;)-k$mE`@!~ zrXgb3)0}l=>`5V<>@m^vCb4g6rgc1WA+1n_{2YG473a9?`%8VR&+bXzK-9PR$r#jc zzMckz!mb8;>0WT9eVBiRU<-&|?pTa-g~-IUXbeh3L>7GOE)V*O(}Nz44_pecKef4k zOe|@PNW@abhAqzS;s~BTKoAO=kKD^|vu3e16&%%l;Z(P!xNObe&F{bZk?Kfi->{Wg zD@Q5p%E}8)#*f9-q^vGjh7&_|wJkm0E8GixyG#6l37_j#cqX!Pz5@C-`JI6tgB2dj z#4Q;O)4VMN#KHSmbRC-lH(@VQOu;xCZG}&5Whijq4+rg`C z8+jG{%@38ldM_KK$31w}w`3bUe0VI%AZfEuo34E#eBBnrX2#}ab9hq}95$U_eegnM zSezuP7qKl?qOdO4>hWGd7Q|^Po)6-PJlhg4>?5{hF#{r5H|Z{?u93dW%go)iXl*rv ze&$&P>z3HL)-z7T+;Bs%C zn_+<$ini}s>$`V^#`mSlN9BC>DO(2{$6iW}atLk$&I!FSy?25ra2{yJo1rv8n0c!N z6hAbHrox&vnuFya^u~t7A?dV~K+H0vJ!?Kf}7Q z8v9tTuFc1TUnFzOXS~b9Fs0n#A-{3qSPvcYik^KEUw9W*J=6qW7gj3ksjFs{x^d$30dFQq4G;f4uaS0kvBqJ~?;KQ0u8si>$ zyF<)(OAtGt~%M>SLAK zj>Q!*V6qyuFOg#hQ!cbfZe>sf<3XI_+BC-ZL{Qe6N}QRf5y-nThtlc z6N?plM*!H#0`KkkyoFjiD@IyPaW(LpVV_xWSx;{*!*O)?HE?K^)GGQqAbGA_AHW(g zSdLUAdRA$!9>Tk5_Ygu%6-l&Cd;J@{pQi!ul}UirUsG3{7P+gRC$HiRCH(5pwtk+$ za1shN+n-u}zHHuFOFz$G&`8$)egS9c=NSxKqoq%>lPbb;TS-EP$LliVATZS8xApT3 z#%pQgWBojXSJiHdW^s})M@$p1WNCyOaa6SBo)%j_&rl5Iw*4~{e*qOOaFf}Ys0%97 zMD_*C;LRp!Srtz9N|~>AFqQczM!v0|r;DOO)cOh#w4bMIySnkGhY9vX!>`U7>MCHC zex5EcwbW%#0^Dag1U(+GE|DmF;TAlxL7jD1RRMAuFOYJBb+fT9tuTCZ1pxgdoTA~~ zhk-Q1*=P%zPp0I~pVq_L+A~@|&!2llBrtBUg7uf$S)%q=p}Y-O&QTkHSBax%w;5>00&*j%E+au z)Y$xsqs^<6Uz*r_{WNvy>GTAUWApXZ-8NX=Qm;8AtnqVXQbnr(e*L6;q%$c>S^vll z9?4n{4Q|wrQNKeg@_2lS6y>e2>;_P?0nL_6q}V)tKZng83yr~6J2f^>hqkoa_x9Fk zv3dHjbE_DF9JlfEJstf8iu(r|uu+K5+e?;CUm)<`qyiL%b5|s*X#J%~zl3j8C#?s?TY$ z`Rex-rq8W>*OfWNKm^)Bg#bA=Ujs{Ha8E2&>>U9n7d{y^HlKqQ$KEH!=JO5fvH8d$ z`4Y$E*nEB@!{+x8-afH;jhM>gf*pC&c(fP|FKzE`jTY%jy8QGLzxg)0LxjCZ%QQJV z5hhDMZ>L|@j@=^LMX-&7Fjr{_W-4igW43_u)sTJ5B9c_@d*j9dv2SvyF6wz@(~)6b zN;(t=pQp>Lv%bWFwq28H_vW_BmG3)DRDBjRAhkGW4f2^ zp9uLVXs0zEmgc2pP*t|XiS}vu!?I&`{IaC%E`y{IQjJc5vh+al2bvo#PdzQzm-dNr zH@!n_<-DLQQBil?Ah*}DW9wOZ)4`w1e-fTuo1~qChv45+63&Eu2=A<7++mql)Zc*_ zxRN4YCmG?`Mpg_qTfXxs_a$}U-vs{UKR|kYr;y#jbB!Q=3tx`(DPwiWqK9~`5f=X1)O5nXi192QF$#FuQQ4v_Ff9xxxL^E z+_=5y`+Ubs(o1BMgda(HFm92S52$}O2>WbJ(rMRQ7V_u=i5bhIRzeW#TVQY(k0nSE zNQ%3Z8|4fUPY8ZEVfQM}+~(-|WOrzsxReJ#CUr6B5NW*t5y>kRNf@l1r!ENnZY>J} zz2`0VaPi50jQ77Za9`;Ahc`xxi$-eu(VEML_`2g_0Hi%) z0VA~TqR=;Nm{+l3Z6wX(jTZ0$t!se*r40hvyU`AC#-Sd)wlkWIADM2=`{QY;r?hVG z04--pc5}6T%BnwSM-}XE0oF+RN3{Y6hD-4C94JNoAkt!?|8ZKRk{`5AanCBEj(Ykt zBKehVZ%EuBj+c4f3YB$h=qjm}?a)jDxL$Z4T#S3_*jH$b3tjI?xDZ?BtV?2az#UCp zE?_Bf5P(1SMdJ0B?Wai~oTiE%0fILGg}N^I<>N$K@Z9#L#n$LrdZ742Mgj!DZgo0O zYhv&VvnJ%u`)*ChjrZT0D0Yr-O-PW4)kRWsmY;cA6Y{e*)&%IZUV7=EvL;BGNs33* z%f$pE`GkR8abZjE3iG!o28ulqU^BS9S^@UNAlRN5_}UXW=oRwY85X=%Fe#TYHzA-w zk!Delyna%^ds=COmQZq2L|72goZxoJQUJNFDdbkOAzYbj%S2n2{lMFNU8|Fu*Y){Q zTMoJBJrth1+;v*_0)i?{q8F}%3@Vk-5*~myXGQyKTRek(!b2e}nq#^h_&&$Cg*7BL zyS|>yHG(dp%r>%zIADy1mk+7fx_Ft%O19_2Sprz12y?@A9KyT9ATz-0;asa8PEYZ) z=J#wCz)8>&%^ydOk@BGDvx#rfC+Ias{dx*8>MC05#RtXiar%D3o{N5^qL|U{4`bha`N4e!@_QXN zj16RCD)8QLN}WI{ZFj&G#WQ?mVD-H4NyQ_uGJBYPu-n+Q&!1PMn{%&kFWVy?s6=ig zlsBgq{-nd6iA?KxVsKMxd&uoJregaq%9Dfl3)$uM^5965qZk22$qcM1_Y}rPD*GaV zZKNpTNyfZz`hRaU zWf3UGJ*^8QZOl%6#zo_O0eE>{Q{IZPl-}OU8=Do=$Djx&UKtb(dx@nYv0x?tFD;`$ z`qS4kFUqW^5G$p?>Hu1L{w#lC%It=r==jBc((DOQF1Sd+sTzC*V6QPbN!d{_XygK7 z7{b@{~Dh8dguMar2AETMol_kgSDpY9a_MtzL-YE=WrE=(39#ZN$PMEQ8>=Qzn8 zNbmRXbEvy}yk0+Q&blhDK;1%l!6H9c(a@m3ti1BAsTn^Ct*+QMWow~}WRWPzsnJHm z9j?c^Wq?aI@el(xC6s_*&EgG1a74toBtt2P*vAs~w6F3YrE}=!*Hs(6cw#gDL)cG>-D1*^GfO3UKSY>C6?n1DrH~Br!6~aQzE23cl^p^vNMFlYN!*0EPtM5Jsm5f>Js(=e6+QbpDg2q z6^Fbn6T}SR^4Y@drwKdhOJ2$`@|a6&8pyC|(LTta-;ZQM_Jvkgpo|@Ehr@ug|6Qwl zgY2O^x4Q+_1^ihH9f&Rj%lDC&bGA6LNZif=_YnOp?iMyI&rS!I{e5dFi%hLnFLt*7 zfP<&8#inJo2Qk`LSw6zc#r5(wVo0RqgnMfk zYyoBAlks%4nDhg)v>XEz5BW?;u%4yvn{*|*cI=}RtZOS?S7&xs;QJ4C^d3_ptmkCb zQ3}>i+@O(~Gm=}!SPxOZJA2xhb(Dg2T{m!z*;x1}g}~kNDg=c-ni6wXxw>j*V#UFC zm4)@6Y}r47kb?ZmA9j3{Lg1>}ZCUJ!z{QJpj8Yw?cf_*p*O;ZTY9FOwUDQd>`MZOs zxrD7K8Zn+hO_I;BE|frvuhky8>%8T7b!DQ^aU|)sIA4o_z`jav&p_j{L3{>4*M=REfQo-8fg-uBFlt z0rbrFOI1B9_TrsVdQ;jTENBjc*P!B?rd`AA_je($ha#_ZZaywF)eLkaLk3^?Mogol&PC|VJmVddH+6Wsu+qa zN!Y^Dunjp9pcUE!U6b0PMh$k;26N~Zz1pb#4P z16$7q1byd6v&-o+|6fb0Hu8;tV~?*@yj<0_Vj_{+FY7BF*DAKn{W|1pEsmOz^$w2h zJw_kDqJbNIODgr-x)HFidZ15i8^9HCR66+ye24IbFC=T8kVxt1B}4`xkbEm)Z%4g^ zMQ%sjkP(E7c;n&T17+&!3D!^Vv8S#gr>IL>cd+Q|3HyyWun~&*V}WE>$Y5;dQ*po{ zg@*PXpf7bS?In;~SFBCl``Fh44Cb7!m^6+@cG{w=jOo-dF^g zo6D(Z3Jb{l|>Y8ad6dqC)^=_Hx-+T zs}Aw5wYk>jiP!}ZQG-YYH~w+QVsKzLtkuGj5tGFsNM}f<6>`yF41qenX44Qc>}k$A zGWPb3)}cx4TbgP80+g03uN%2E@cvSZ>a%;&HxRWbzMuNWxGEPI)Nj6?286<{3fs6g zz})670BiK$gAx&u1>d@pBU@kv=SSDIus^lAe@rZCj7Y>%#fB{a zbAkZr0|cR<`N+LYt~HCTso<#Y3#U4+R#bntC?HpVq*kW0Z`jJLm7^4PWmQs6#*f9- zq^vGjh7&_|wJkm0E8LTIyG#6l37_j#cqX!Pz5@C-$*H-=z&oO`!eiNZ7IASxEs53$ z(@U?UwICF%>XYSru*OcL{l@n|-P7YufS!){AOaK^P|2(J zvO#*>gI9e^w!y=P2U(2EF$=ZZo-!tHathS)D;Si`;Z0F+*mQpN!3&jPaguPcPitE6 z=Vq?ezbMa~^5ZFg-$zldXvTkyZ%c-H?IgvK&2P%p^Zr(lt^chB)&PGJcLDx-f3$$o4&E}TKc8$~>Yw97i_;s#fBccJdBBdagFt^( zKEJ(eq{H5ew>CBosLV3&&lfQ5T!#CXI}l9_~C5wlJf0R4|r%Y8&_Z26th3XJvjy{1iCr3?mvqMu@|51 zU1xF&@7A?fC-bupCTL4f15paG$x}pR^LFPb86C%mhq(D%7k$s0@S(9f=JYX^C}=N$ z@g)8;tQ)JbkLBvxd_3SR8_b=1A}{!Wz;lR+Y@p+XtCQX9{UXv>4;~q*a`XK4&mS^J zXAhpHN$d&kJPnm=JM7zdi#F%#O5%h8Ydvk#B2t6C+W zodo&-)_}osq$1I?N^|uP-bK5I5Mlyb@}}`67dvDwPPgg{vINW!lF{cEthINo4E>!ukaM5

    -xaf_Yi?pA*&kYH2+?*%MW0KhIF=*b@!E0@c#!t&tjS{XAV@ zYN^YfSdB^l%`Xmv9uHWTNEAM|1@AJ_i5)MkyDD#%tMA>4tv#c4X*a z7p*5TIo)t3*xEB%*VI`YYxHCN7;VpJ{cS~07QP)PPq!wC0Gr*?p3x4H#>!sUs~EX8 zs4hz6Q;x|ZCXTI6GF2>}cAPeJZ;yib^<$G`^YxDu3FM5;TP*d^Fhczh>QiM4T~lN8 z^_AV=k*xLo1CSb_DF9JlfEJstf8iwPP=niPGsL(wa%^4!RWk&A_C~9mzhceu<4|Psg`>G0))6}8H=BvdOF`&HVu8lD3J_4YIxG?Q|YHYrOP-lFS zMN@rFi_KTRuP}YHl^+3X0d_t$HeUlvV{lI_R_q-CU>9XFYHU6SEsniUip}R6{!+!} z^CKBHzlZSliOp-oRFOn$yZ=R6rpeifFj?|>JN>eD>=s#=#x@Q@VcHVRP0}hkY)0zU zkbTP{l2q<{Um`op4J_-?V2o=Gq+X7PuREp4Bm;!@FOW`PM6O>v0!#g z(X!Q2kpWtd>0Y*fBIKi>oz^6wc6pm&USkI3WlQ{rPs1OUd-kHgWu7!bs?jM>mL4en zKy#zzsiy_|(mqk{rgz*Ru-CF<>sfl!!Jo^25*|BC($2v{@b4*!illdlt(?z;Y@Op> z_948pigAZ!5>bB#X5dPSbYQUeI>{u*HnO5>botJsydb0l|0eJ+l*KG5e8Ux3tY@J8 zH0h1!cSp1N&h~tQcGWY!l-Gonj2pkK0!>3Y$9I8v!JX#?z5vzb4X|spB*o&Wyq1gC z8O0ELFNN*gUhoBOT#7!6KsJSlvdqs7I``CpzJpe7tKDB6E=b`%ZX-}Vo5n6Xq=o>c7Gu*H?lIHP73;2N6wLpMUA|L>*?A>Sw zIO9-{UfUVXprP2-yg!~2^b}in3)PO`EoVt~^Z!rX`^RW{m4~78eKVe2ACH~&=MY9O z3sgwN?ojgWe&b-g_S(*18ymc4<6X=@H~sqA^DQ2Bc6OPW#k*-L6@nYJAyiGJ$Q3Eg zy(r3Gkx){J(4gE@5Odqm=Ju*liV_G=axa2gRS}d3sZs88&ikJCea|`1bIx;qd}D5; z#j~4v-}gK}&-44~K0gxO&(#S7`&R&KEc>H-g9L_)wZ`yM%m{YxMTrN){$R2nj85_I zD&nqs`U58WwPSBs(xHx5MN%c21$)R8*LW@gkQeR;pWr@n@)e%o!qj^fF4QM$(Parb z;Erc5H?X9zo=#ke-(!-$9Y4(i;T%`>C=jCQ-Wc&xh{PwDzTkzcmkzJYZj6pKKU6G0 z0BkOLdxvTPiuEQYwl_hU*1_P4h&1Z;myg_=&>wESiICSo#&2T1iRR1Yy$KBxv%5%J z&ib)PZ$dwo6Ho!0Hp(uYblwE3GRcUDMunJQC0|_-p*5BnBH>xOnf}D&hs>XtB>NMS zSbw4hormSt(v}e|Ftm~23XBS7)iTyD1Pmy$JPKCSPb>IHFHH@?(*A%7L|z2TY>)trOjubT^Hw{JzeubX{#iy~hvx7CpIFy5DGg4_1^CgG$r3Dd9KUyU#bo8g%L06vS| zr3nSc*ER>8Xo!q#ZrZj8VsjW>=Jc0gjYd=qsrhtsGqa8C$cwWDutOb^i5PliAl)Tq)klpZGGnX}SgKDojBCGuOv3AkK1LFSC zebl0feoz8AxcFLldKL1RQJ#=e0$k?wJjeAL1MzHek{TnrdNd^iU&uf*@h6{paH9++ z#Mg&nS{g{JsMIY5UW}w@Azm(o8_6YYO|Qu^!J^3iPFV5Ma&Ty0ysYRj=c1`sEl38! zMRJ;+gF^Ha-yMqg6z281F!(~@x2n?M&~U(bmNANh%jl0V6$KfzH zNu_XKp&`Wx%yz9{0@|w548D*;;WB8K4OgKdX&9ERTEYY@hq=DE?t=9X zTQOQDy<|9{0@}c`{IV&U-ik>mmkCw;Ij=XA10vQ%K(K`xMtV1m{m0JSBFke8uWsOP zdxoXqQLL)SXN~oWMpIRQV+i^8k+gJ=fJ1&cIyY3e;yh%J_lm}5C(=dO%ZEYPG#sUt zn!hjXnNwSRCRF~Uy4W_0>wa2 zv^wM=@!;r8k`7Vllo4U3kxjD*WmY-I9z1k}Y^`?U07Gc=dp@ zQ#lSOY&uJZ!nTAm5Nvt8X$bB#L^%fZs)&~d;^|o9L2oNv@GjqvC-8dzXwAM-zO`2s z#squsq|_1Qg5Q=U#24^w%U<5p2=VOv#n!H~-l>Ke-t81Vzee13=1#ltxmVQuk)e1p zmF1kJzt|q&2#PuGRCiUA#uq7CFyoUysJ7$t*HGRBB{>srhIPx@jB(UPP zuRkxhGER5)wqcdtK`UKlgjIYLUv2G`rvA~z*sw+hr@W@65q?B8Y0UU&+wH;3_H~|*igHQ2JVydMdHhwaf}5xIaC0*u1PnLwJ`W9g)|`Qdj~IbNqeJb zl|t$ZWu+A#MJ2eKT%|yQK&PO>Cr6^v1(?doK%sSC>w;XRK*DN=nDLHV@^NC70{Q!1 z#o^6A6~H2%c?!~$UZp^TX~mTzvD&JL0{kDAHfc2AK~w}jmECuiA=aH9)7ioFaK3kN zbzf(mUaO}B-LbXSObHsm7_apW_GEeD2z_hFSM)!Sz5RH+ka)Afp(?>XRvW-INa&=7 zrJ?|l74HHpAbjZs8<^3`<2$>%M9`f57MtEx5fpc9e?kZd_04Zg4{u(ZtMLZe4C6_t ztkb(U2#gpvgkAl%-e*p4_1ijDwvY{}eOb`EHwftLI{|OmBOWGpZxFxjN?X1p;cDJT zf8WI;m&M{O$`nYJ?vKKruU4|N9J6sGeql&VJVAi6O6W2I8!9|Pw4os%&MK%=venK8 znvw-|BK(P@YPdj9H*T|X;!<2|Apw6`1eOkY?q z90u<}#XmY)hA$IyoRBAU*&g0Oi9Dd-V#m5Wfs4T%c)z~EI(OZD-&ym5pkg^N*NKh0 z*5ziSLLHMB6N6)}V?}QCd|iA#0*1D96K`xq4yEZIgQgVf9})xGkRt(FVLZ?c+FIP$ zdg@^w{XuqOs0qUkme0Z195(ivqw;wG2$aGDze3GyK+tz_w)etlMgA{Qs*OCO;8-wO z)y?H2>(UZ;Ue+rSWL2N7<2r1zHb<@4c8A1v3SA?mV&X^N5>qO^7=b>?8BO+$jXw3v z_y(W-6uwi~FBX!`Ncq;5WQgS8;f)8{V%5`P zJbEi#tcsklubr30mj>cpXtD@-((qO&eseHuI@6;IL9diw6i6PrSDbKvsgm zZO+Go_0fFLCW1Qi*%O0YuN&&n{am@&ZF~5|SE%y7Xy`zH)_Q6&RH=$`Z>Q|CP#Gkj zfG)?!ig0Od4B+K?DYU8zDz_uU=2r&d9xgVGP@T~U5u_Rxoy+1AJ2K~N#IA*12JCQq z1*5Rl3(H1KRfZvbv9r8u7?0xxsV4rAvJf%sS;0Cs_VPdr0?~?+fFZGO5A*tS&{}R- zAMDn^`&&zCEZ#bL3)51P-|O2+-Z1y7fd0)N=K(>Bt0#cFib-k)V2+&HA^uhIsUhK< z8%(CiVhJb{5mhqPeKBN(sp|*xDbV@y&ffct$>kdvC_{x+{5}=_aj|3xA`?rkoaCI{ z$60ce216-WG4d#vYm@Lb4IK4j;oR_>B6c_g-r`XFG1X>m9HsFqyOwe`euz+$@pdIL zk{Q@8^=b9*(sw?bGM9~45jQ8aifElQy^czfZJyvcpKb4hO>ufO4&DdmHzV!@nCXZwB0zyL z?uE=VhSfVC4CLk!zfCRVLLsxWtr1A6vbe!(xn-gly$-w=lMJ+Yjb4+{?AGoGtaa)l z6G-p90ahrJI(WCOmQ~+JV6x~fYBKid&aEFgJ{UE_cUDv|`k_{k?-i5LsR%(<2QoxJ zK!guZvN$ZwYShj>TS*BP&^Ct;MZsaS<3cIXqf&q)yz+x z%hM*EfIL1t+JY0`q?(54BY*Jv4BA>B-MMsWzCYi4evY+72kD15{Lvrnw?Ffph!)IW z9*8G=W^{K$Yi874#cm6)zTSRJ-(a_I-PTr!4$S>bo&nke@$yi-E$vW^l;HJDG!kvf zRmE_^n<&8MPw}(Fm=y;$qlvgH{2>Zckh#lW`rOCfVj2v8O3f+rf4p|X5($cq@A}EH z|BKz=`vD)bzfMiC@qEkZQO2i}d7cKJk$t@OvJ@~IGx|$y0tX+4;yGj8g&b`X@Eza| zF!1Q3k0&$xO^r-Yv!-CJJslq)*@~D1M^n|1aJ>=bmmabL-cmq=U{W*|+pghQPcUBe zB{xj?lxi^`YsI5UrXoJIZi{|av28rn5_^C@iH88+Hl7_qYlj@dF+P~@UK(GJON&Rh zsQ<(x-->|UA&vw8b;bUk!VmFbe6+E1Kx3Er_}~z>o%?MLwa!zgs-+ zx6gNXcdi|D|F{H0m&J4Q_wVfKUt5yDwl{y#cy?(6WNW^+-TspJH9V5zphBTra%+}| zdnCElsPc$?@fS&qEq8kt74nw`=txfoQ3kW=Qxp{QymOQj^v^7#_jw2(mZ)PvALkVf z?{WcU zM=7j#WuAXM6xd9Y4E6!fX^Ku) z$zc=D+>c%%M+th!vA|d`$_s{3XMj%Gu$G$yiH=kVl({hMB&Zxh&vK89d)}%}i$+*g zmZ5zK$1-Tclb}H|ivNnoeZ1UcS1CkemvMFI^h^=#$ZmIRekcK@D%R>-itMse<~SG6 zcg-u-o7RPq`dFv7=jDbPFsjCVAHqbv0HC&@yZe>|birVOMGu5_8ABw3s%$RlF5dzu z!i73SE2|JzM&q%NwGWF{O+zFWx$}KIC-4?kwKx_Z9>&~-RyixiTA{oe#LckB2uS+n zt*t`ir>J4-R=Cbp+5^}C2I5FgnQMgdZpJ;NFjGr1qL`JRvJihU;92{Z=jFZrRGIG1 z(%0{PpP!iL>Cm-%cecl07ChSDC+2x3gU0gqj|wdo<1;B%>0f4OP-#ghhWA7HYE~;FWrM@ z&M73nYTj@@SB)CaNWR(-_$==|8%Rf2xUHnn?agu&BtI!~St5=N`XNz+j%Or)+qg|s zY=@9(ANjl?y<7xS;~8m?EK!caUMI-8=k~hoa?3GWahLJc*&KKzo;op<5hbeQr$=J0 zV#Nen;gjr>?V7iI#LmsDHm@I=+ zeE-dU@L1k@YO1V%3?6{o+gvSf#a<3zgPS0hD8vzM$yqrrLf*Q{DPaC z=R+Hpo9DmRq8cwZ&wqW*^{t9b&TI9*$grBZx%vKUz1$st{=p!z5GQ);JObc1QFP|`R*%r@l7NaJMk{SV_p&9=H_eA z%J}=N+54R9v6+GNX-2nR+@I|oT)TQON2lsJPiit@E#uBF zn?Oquzo?)eo?7lGz{0sicZ2sw#0_4NVslj4%O(4aI<*&6Z>0WtG40FwiZAiQt?CP@ zI1=LDVP=&qWV4hX%XqL}4C*iGu}Ruvn?5C{c^FrZ0@z>zMQOVBEi zm3L`BR13rgOY-{KP6KSIP4P zl0ffJ$)ZAE{~3Bz{3SmB(!qUW@87;PJAD3y`Y+E;(2=puqb-`yr9kwrPfu^0b}rO6 zir1fcoq3_Qz)4<{(9O0#FomA>d(_P*|8f)lGH6jI_5?Oz)j7yHSfO=OU$sWGp0Whd z#>|HVtkAlf!{6}1x{D1PBk3M*b_fsfrWPnrS_CAZRs0+}0nRzpV>hnN_F$mcm4os2 zPHU&o{r57D6&^q#N6xbR=I-+&(fwSVFtC3Gu*R}Kx;IE*qS4owdWtK-0KO;@0;UU> z{a|!Tx2q_Ia_%sT{o1iNEa_0kt0JkAgF|tR=Mn&U;eKcZ7m>+FPQKFj7wKFo zOuc8}LVdCpT`s11=5hlIS77@br&B57_n72|X*hMRhy}tquIf=BMAN-7*QM-;PcVJK z3s)~4UYXq(9czB5Sb)fpW*S#qJoir*=}k;tHE%+H`N+Kq{o$kcCYmpo_a-z*k=}%U zEYh3MkL3hZz^09|ODCN-!KzF$BBD_trfCD-tLaZne#rcZNwPmNiS;LH(0RyjEo}&0 z8E!o$H~Z`sMZQ>Wt08aMg@h%sk^{xYo5h>9UKpKfkuwvioqakcEh__}F3!#F z;*3;Jn{n4>1Drmo6;z98(1nKN_Zkke8-8o%vLR7M@;(yx&`s#3=zT4<_xneq$h>ElyI^%&6mGu3IhsErwN6`xSc)D0(#XMGWz>*Nj)Vfm2q;Enpi|bzDANs>i6aGmg?Lf6kB5d7 zBQV>wf(dA=N;5c7-xm}%Z-H?k&Vq1EXh<4{Wvf^?0n4v~?t=9XTQRyTy<|9{`rE*= z{IYp%dMhTOTqacU=e*uf4n@xEXqdu}|wid7Z)tmfcoG*tyS#w*7~P>(kXczFie z^FjqE4$s&n#L?L7guV#aUI3wKI7%%wg$1kie|Z%Jx)NoxisHp4{{&}M;h z%8SK!!;uh!z)cBGx5qXCd%@x)ZKK%LJMt!8*gJUBX&q#2Ms zyATm(8Y!J>a5wH|>4tKUF_gnshv=4MB;strYioxzti`JboSg`V*UMeL7Q13JkbPCy zmQV(QEsr-1!QL@9o%$#T@5=-6bgc29x0Nn+gB%iVA z_Y3&8WiM}Pgm`xTVr$n~?^J^j?sf{FUnA~1bEjSS+$(DS$WT0)%5oR%ZL-)N;0TI8 zntYJV-N2Bae>dRx5vN zEu8TdB^o&t2jue!-C~JerLePi2yNl>?VZ`-e4JRNMK@^;H#`+wYY?Dhk0v3{*$+)F z%PA}8(R-GB-iW%U3}`G10$rI!u2M(@jWwL{sDL9^DI@|Htj5AuDI|VPwfRd%Oe~>V z$M+}ZBcEQSkO(R0uli!gS1BaE)uI;RS0ujN9>Q!OL9I-lzsK z7FU&F%(Pch0o6~hQXp|+q@d7(s`dVb60Uw+$I?>;Sj>(+)2kFnSnUuq-f>GlPOMTO zf8TqN7N9l}qYf2iDFRX380b|BG?-RgITEW!EFoYt;6YRbK9${fM(Nd^9@E*u^l-j+ zaCKj2o_aYNU(x(v?iovPV2j?A{=L+m$vX z|KHLnU&^BE5!UJ58x%;E?vKJAk6*L#Abw#;Oguq=vP$SO0vjqkLbRbFAMO{_aU(cN zg4&Tl)0ry0^Hk%>Iaqij9daP2n2vp~H|2+TuRHRk@Wf`JD|X{O zV$7hV;}I|%2Jb<|KRQ~5FB5Z|kSBE69^OHTJfPs})CrrZVsHoEZ#P)S#ob+Z-*=XG z*~45XHtt%Nn~e%}Okzw7j=2=PeLez)wsbRA9{ErJ(Coly&pbHfs|ff4j0d`*t;L=J zObQ7M+--_8*S5s*!Dwk?Wn=59hvi_U7;3_>gXMEDHiwP9=BRw^0fAC@;8&=b4G5cE zobA0ZT9N-tlxic-C^%9M7C}~Zb9H2ODYmp^&BY?Kyst!%ReiRO>#)h%9JONG9TM9M zMxP|nq(dL!-XMYiuGPbawiI8CK%e?-f>gj!`Ru3gouXA@A=!+CB>Hw#=t_a)%LDOl zG@;Veb|j{Zh_n!-gqD`U<$j>8PCYHgqqox4smLk%Jmrf_48@lQ;$3B6W9;!N1rmAJ z2w0o>R2gtssiDOK^!Xt%UIL-I;cYr;JS)NAHs_b*@je531IX(mm`}5c=YDdr+h(G~ zSE%y7Xe@#fsrA%*%dnY6m8vNBcGq4ODm~%lntTGf93Lx!kJ=c(%kxrbRTWfjc^Wy0 z#FkhxAb$@Rn?|S(0s)YQg?hs2-s}oQ7{S)gBD@p&1H-VM*S~f~NWpES*Kh-ZfXaX! zZth_ewt8XNh^fjjq&HrdgH_qqi`vDEU3}3iCv@R|WKM{x}Z^LtGv4D%*R$kq*1~)R1t_4JK1$ zu>_Qfh$@-tUWrTxnMkdA>Hhmv_{YVPC5TKcbz+zSm=^>XrNK}NR*XC`>PZ$H!u^1Wg*Iu#+vNXs@v`0!km!_tKtwR52e3$}o^IeaJz z4x25nF?it$EY1>Mfz3r!g2aMVkIxDs5NDZqF^KDAHWyBxJc*A}#bORbx^Hs9?bI;u zoXMO012s*YA)AM&+KPmA14Jy`2hDD)6Q8D<`N?y6+N2YZ$3AJxlj56H(-3`RPD^x< zet5$l{n38=bAtK#%LDO*&y4PFXw8hetJrPf)z=$mKa*#Y_CUNm6mLsAR3jyLJrj*Y zn{rh#oU{zZXF;FriJv9LtT?b4O~k&!AAT?MpHg$m{2#B~utb8QUMYnx5i>>y0SC^pF+smI4|ClcKTMb`8gRg7Km+xnYt57ihL< zlBtMKt=poXRcsqiHQ58K^%Y;j#3%6(;CiCc zc`xPeEAWz)QrL#rCGpY#9qH*H%3wBqih^REcaB!jGx0GHpIJul^AJ8PQOANl&MO+; zOAtJb{|cMN>U@VFUCzT3Qn*o3MkPL3xU~o^%QQ@@cVq|j1;j;m(DA|R^ZlFSCezqY z9yz6Q2lee=Uu2%!+BexlmT)rXg?mrhPdP1%Kf_0~dEekK3uKgSmIi|^--NNz7Ymb; zvv=v}0E9uxKT{S!tANlg>s*9v1UtBiiq2AqZYw+c(lHQ!JrvkXlN{;L<(`YB8oKi* z#%h28&XA#umiOwLY1&lkyVmDrH zva1xLvCFte+8|E}?8t6+Y28`6lI<-B|8)^U%@&YGx zL_DLoRX`U^1am5~S%M#lpemb7xYK?-R7xmG_30NGjmJXPo)g#t9*M>7!YjZ~-pYm# z4`c2^tDF^Mtx#SK;%3-m1TK%HycO&LR;aSRADcqYuFN$;c{k%8QkaP!tM8bHD0}^j z0nggMJWufYiRMi3>Q!@QDnEB;>8o&G1aC;q^K@9Pdv~_SUlu&t-zVmICWFRm+B_=Y zhhmY>As5)}i}GPs<4n{}U)43GQm3@cEG75OZr5tgWVp2>KH3}oHXiOus& zeyc?-p2yi?%P+#Ve_w#586LtAZG%t6Q0?2lhZ8U0q6H{foiF-AxjDJ@K#l~<5Y1*; zSshLcW8N{l8J;ym{tEgC-NC7Oo<5FBG3&c8qVqg`pX)n+cACIvYnRZMsh&Kv53?+@ z4lPM$e)o6@V1w$X6+=1W0hU z8~x(Bz{qmx=dbP^&UW_Z`?3R_8qY{VXEz{VgMP?|(eaGrZyT4aitTu1IvY2*M0*`u z2w1V4Iw_5c*{v>wDD1T|#+|@OV%9h4aaF*UWA+_a##c{Kh!L(T2vWHJw8btplo7?I zbC9K;cR;8>i(74;b?WV^7Iq$dj}se?A-i^X1mwt|G&u1_fKm0 z@a)`t{|$x)vJH&!0(0~IL-fI*eRXbbzW-)Fcr0%{HSFI%1`j}PZodC!KY*4Gn3$XA zf3No03;Eev3oIvREBgm$=jQtl`LWyj?X289|GgH~c)5B0>l@#lZ}67EKQ}ke1|%Xk z&wi)c$ngxvgO7Nn;lmeECqde=bMt*L471*FO@iOUEcqgk+Sa_BVOdwlH4~tk$`5XC zz7MZ$=(7?K{TJ|Z^Zh5h3Z;mg5ji^+1<4H^+}ykYs^thoW48=C!hq}W8ap>%L+SVI zeg1lC7_a-TX-|0lH*RjedpWHJd>_IzKmkA<`FFw%?A&|@p)L320^}d~PuXpT@Y2C3o zRsa8k(TW(Jk)1(JFN@eZ4r`Q_V1?44uZx$(^_1_nO(Y}z{#&<>srL`ojMqc+1A4Hu z&u;2gIrp>1`U%-kg@5+b_#`GLj%1*Dv~nJr1*?;Wm2I|)OwfKz|7*vmQyxnC(x#_x za78omvYaU_`_E*~SpMVF@WYB+0?X>rX;3yg*8IQ=qZy?&yhjY$=)(^-*2L*iiEDhn$pe}-8ca*Nk;!@z>zMQifSyFs4GF!NQ0KI z1b=BD?lrEt-Pz6g{?s{fs}JA^i&TS9+rh$M?>v1)7)P}#33O@iv>L6WM6_lUe~Hh( zba3C;`?qh+4xfLa{>!rybY!gaXp1IvDG>eZ)6*NLoeTAi;`L`Y!>MycED+9dRgVH88i2-jSMu@+rZ0Hm>ZQXg zvm2vh%?}j|5INF}Da5@)wE)F>6O&iXo6uiAa&JO^_~^Zf=F8>12@MjnyGWbp`msoF zLO+({O@K`sWtUDmZz4V-8s|?m;I$%^9aM?;Cni5+{=_8NpP0n@6E)}^@}pe{!xa%Q zDwtKvSi2A~pvdwlSW!Q%;3K`X2`@m``V$bKBi#uVtGo!d;6TK)+vY;q?OW0A>t>(b zqR1D^Z8hYP+}4EgRK(QewmrT{IO$Bn^Z?$kMi_+6a7=#ypT+Ldgo5L1tkH>v$jIg< z!xmZXTgvp8VT}f>c30no&8M52nQdf8UYsp}9qN#5!_X6k8w#ilh`Km8yNfeYJ#EHa zn+B(pZ*F-&-WS*vKxMD=CUSkP-VLR!_b_r$J$*I4~Y9m_ffkf`aucg;NolH z=~c*MMtMR?32>Ry_r;NpIDKOvo-IyN&WNs~1oG=cF)a%Zd(jE}DAPf@DCWd{7GIL4u3#4#j&4^LicZxls76sx&w> z959|`WN~m1OQq=|lOv%(F#?KF8R%5z;gz|u3IJ&dEF=EnzYaxcNHGGlT`QP?wyHFP zdz%EyU@NtZSTw3gj#w7%ks;nB>HN3Ww}hK z;?H@#p&WXKaiFmok@q6NI_k+S9EPRgQLL)SXZ8O_qp2#uF<$vEg1v+SVrb6`)vY)@ zWB2_>W3vVVFO3Rq0hG_@T& zbBnC)G-8wf3sgYoTWAQxah55&{4#)IBg>fhztQx$2Ylpjyv_5RVCeWiSBuPTg*j+l2l)xob3cb&P@t_~h} zcLR$r;McD;gME)~ zbY{Wc$7e@pzR?c(`Dap*5Z)3@uNWdLm;2GAWx3-$G_c$;lQgU8<2CYEj@H4Er@!+Hv{UFQXsDS6h2|Y60tne=rbdW7*3%MxArX=79>^79E3( zc>0LXW|6>HJDi8ZgtLDyi*8VUQ`dH9V7*>SoM8h+^tJ1|L&m2wj;so|YrrEe`wZH` z2RqvGb^iCWp=>I(vTp1u6L2=XEs6w08Z$oHc6%_heVyl{qFmB0&rw6FrY(i3K?i3r zHCQK5bGrOfV5DTcMTtfZ#R2&(9r;L*UGYDS-ak8ghtL*2-`<%W&c}&WTAo!3Yp^7d zJ#U3Pi(w0SYwlpm%K4ITG6po31&^-GB3CJpXfp!JPT?7AIO99Gf%J84fl6!tu+&b1~9H`oj+NgI6`0k zyn$-j+y5w6$r8Y+4d5E2UWA-9!{mHacbvmgHq-(vAbjZsfhI${*VO_mkMHd65Lu%*q{n(E6R=Jm;(+rT)5DwB=4xC5dnE&LAbR%(@qlqd7#!gR0l5Wo zkM3j}oic`XdiMqaoqZ?ZEqlbn#O@8^w_OI!780UpHbfY5>@K-`g96FY{ZZKS`8CoKn~QD ziFL(v?0dZ_Kg4_8kuRy6NHS6cye*3Gayzk4bQ&U`KY#VbuAddV%pOs#%iE0xCNDw( z!(s3qRQ#i(W%x2N#|e3&p$Adbi9Dd->TVKlRCGNCci{c{2J75)_kCyk)(ee0nCryG zUF&k=&7=q_)G>)MG03k2=2GzX`3M+5JJc>L-q?y9O4C0EO({?u5(C?iBLP}rJkSko zExNL^vGvr$Jo4-CP}64U>rZv5;&=LZY=}loFYwK=S1Q5y)(6JIbbvAYG)kZ$Ho$q@EVz z(Oc<)ROA$W4eW~_48@lQ;$3L`0C}t7%{aLfMw+M;NVHJ_YcroJ0}d-Sw0M9%KP1LW zMAOGgFu2Y6c(6YDJ8IsuR=CR>xhRTH0j<~F4gD0*SE%y7Xy`!a)_Ur_W!TK3N>!A5 zH5CSVRPZDrYFrEGa(t`^K5AnCFV9P%RaH>A9T_$uAb$@Rn?|S(43B!8ezUi#kp#n> z3lbCS9>@}6Cyn#FL&D*}C~Wn@vJq32VMt%*Bu}cMQ#o4JC<_t8o)xTPW6zgXX}GG5 z$K)`FfFZGO5A(W;7ddGe?S=Xp{_bzBsIhqK=q*euN`9|zLj~I7pnvnnc|aKATDZk0 z15`&uKQ$zrbA!ngSu6o%BBDyB`bGwvxXgKC=c_*X7r#%1e_SkCg2=>DcjaUN?E_Di z4=|L16(f&wxi$%3)4)+b7S0V@D%ylY;4KctA5(4C#!(u-vWtSV@k4}~jJGS1QIYRf zf8z0MwxhZ#qtI{ksUL9RYr6`sL^jUXK%Z0JSw{?7@Yv*(xoo_OxH+L!L`&WWK8At-=oN6fmp0C}m&K+yJsJlu3-dt6g&|<3Bff|L1!j@=LMBws-SFGg zLM{|CJKGw8lq!oGe40C#(((6*47A{(YaDCVeG!4RT-dzQqorE>AzN1QZd+};8NP}4 z)1=@+^Y%ln;NB}Hqf-%rTeoaOgbz=$IK*)qwfPVr^%0o|50^)*%CbLcd*7|=V7V(g8Qq9q9cv@Zm zO{!^#{E*ZCjI~4u>4!J`(H}j$dhMP2v+bSv-XY(5?kMxi1M!5Ob+ww<2J7h~vP2T|GZN31uigjE^>U4ruH$A0HgTwsXJDq1Jio zidMf>{`2@eI?fqD1KmU{A_kmHt*wx#s4?!3+C4 zdzYtjHU0F;?BFu2r5`}+hInrN{+&JjFH0MDiQDJ9yF1qox?fZOxHo^%cy?(6WN5y( z-91TEC5m6eBRLK#6uKq19=(VsF}BFuOQkWMI7ui9_ z2d~ffZ;qQxV?TN1lrkRFw|{++c_w=-L(Ygf6Wn_m9 zhEZo=my0%{)$^IcqjppZ4em>0*mMc-oYbZqLeFxKjC2&%HV1d*|h+Y8_q7kwX33W&yI zA!`qx*a9Aj#i9UvLHqD9<}S3#SuxfM<<%f=hCN2$vR~fn4r%Dn#Lf^&?y^c>r9Hq3 zRo3@oQ|Q^1xkf1OX52#xGbJSEq0{)b?b+dMdgXPn*;lvdTTu8xaFp& z?j6dA64mk3BQaO8Vgjx3N%ra30QbE4KD@MME^sV7HBLA_H{VND_{zJ2E_-UiqJM05 zZodDJp@D1zW6T(-RN~~uZ3NX@cTDSY*{L$|4n`|tcxFj(bMt)g%-lTjU1ZLj_>)gP zn4Hq=AI|_>=9gGm-u@f>^s{`x#N0gpd$rFVv=IB-F|@2N7egrSj4{8E-Zr%XZas;BW z8|HF-ccj-B^$*O>&DRh4J$rIk090(I#{awT8g_$kru`~AH{ZS7Py>zx7oHiFj|;)g z&36#mg3t13y2p9B`R?}(p$~ic`6wYGcV3s6o15=}WeIpB7CYfB0I(M{Zf?E?t&G3V z%FWjwJQ9zeo13qX)dUP6;jgB=x@WNMQ@iqAKi1k zV$5z+X~1GL1L@O@ZoRla+dH^+^lERA$U&B+& zEzQIYUXfyRRN2cV`;1Dcy%*EI+$ugqr}6Y{;)h$+=YH5!aipUE9cEU^LN-hJv5W`n z#i0I@9-E{+wyC)eS93lEDOW{^si4(ODJdCADG@84euFr{C3*8X5T1Yf97@OHQQbU z@;1;ff)4n5mc*XGCagLKIR`7WZtANRLUPIsAFXW(m`A#_AxDuu)Yrar`6R5+x|_q_ z@WHx^4I3lr9&dIC5AdcIC{S_`D1Hu|0OuU)u^ZQBd)trhTsatT@3eMGcK11qoMrjV z-RDQ5`?)${VE+nWjb(pyZ;-%9L-nErs&7neFnO^t`y?-N*$+mi_;(e-K+T>fM6(}E zDPpl-JNAYp9qM>hBvlg7)pD)4#&ZdPyl_AG1ox4XukbJjQ}0>0P@k+tmy2nhx!l0g z;vfNk;`f;3hiN!HJcx85DbgcQIVgX`|G-FD0 z|8$Yw#KiU{dLm55OJD#?bNlonEh5sW*I#bE34T0dtuHRisUL2=2^PqR%OPUDi3a5I z-h{?wq&J}-i}WV+V>#Xg*tAi0>7?@}Sd~deL^RHyXu#`qL~c+e-k+El{zU1pO1wWY zN%kiuvHnC2x)kst{sfE)X4Nv*E(8oHvOEe_)K4q;NH49OiZh6Jlq7c!P4KnFWLhLq`ALz#J z5m%%5)d+*I8II`>;Kv-F3bJX%xf~nW+_Y^G1TjYYH)){AH5#niT^+1yKHc2RY$H4J z;%otIQG{h1hMr*1P*w&+U7VZU#TludHsh|%1~?6x9fo5UL(3wl_Zkke8-8o%vX3kx-x* z0mY~c?3tB`Q@YB)!tD)~?c<>##R$xHtzZJ$s`nID=B30mu^iY}rz{(;LPOFpEL*jN z30Q7bN-69Fm6%-ruoXkv^pfF(>Td(f^2>cYGO$k;%4I?of6nU-<=|H6r?mGt-*9C; zqa0Z-4Ub|~MLuh+S2UWc0vyw^dJ(?B)qt<07SRB*Jug&%;?UC`?-h;B!kF;EETCyP zN-Z^o1*`Rcc@+h16rqaZ%zg;;RHj4vtyE!;p1OZY-ToaPK|nlThFGS1v!2?aDf1CHMb-O&*Q^>N;<)h!%*Zhkm3 z;sj*|QNFVcken3{jP8#JkeEj^t~T+VaMe{q0`o1#4Q80PKD{7i#_C4fCede#Mn1m5_T7W0v{*Kh;TGL@l~q2 zxQuLrgA$)7TH$|4JUBX&B>d{@hV=Wp;UFT+G*UX%;BMRvOC|qEnEEJ&8S&L2y3rho zI9u@A+93^V@#+C*C$D-ahOLygIbbMkODF@umdBfh;7&u7!-7{uygU$3#~KfMTj_## z`F=cs*ZW6n_LcIjy{a%Kq<4FHKEF1FT1@B~g2G@~n7OlUY_}S)*J+9`;M3?x{=D4EINjOXhE;k8eEww#v-l{!+S)6l3s@)o zgMm03%U;GY>YPh94^-H;=onqs)y>@?hLHgONldV zpblzw~KCVU!}DMWlHSvBjj27-oAGXB)v+3e76yG^;T!0jeMoSql=cvRSJoqWesO6OVfcM zS1BX{7p%rAFAb7O6HBPpr6Vez)uSvfs#d{#B-5)D67dT9tG?LrRSJo3wWvk-6^Soz z#xb_T7W1(hEQwpES1F_cQQtdwnM>LmRkhW5)FnxMp)rQ>NU#jGFnAj&34)P=LWd=; z_b;1p;ng}8UoHT=9RjCUDUh(*A?8R_5vvsF&v{oV6hH}{T?$nZDFR(Tm`@@9ad~I& z{VMLKm4P5vDbQeAapg#?hIL10-sR&=fYE>lQ4#o5cE3+Vm4mK3J*KmR>EV3u;Of54 zJiS(jt?t-bYo-JZV2s!L279tRafH77`Ha1RfZl#QUP!#zF!gHS)CO=35@Fz#1c)~8 zzyiXTUa)}~tvtT7yGsPk*>ADwT@^vGFKf8oar_A(APzXcF+IF_ZLVyfT;$WcHwcUv zH-ugNW?c@ROb&W4#78G#KI9wFM$5NxIt)G3jQDX4=vSa>5Hav-Oej(x8; z<%f8$JMyLSMRe<8pXfA1K7an|i(Nk}cH=!_3}~Y-EEo=h_n_h*9WBF`i8)Tl6S`~< z@1R5;P;hlJ#*KpPKY6b%4It9P(8Jd;!J--O$z|pZ*(LPniRk1jtSdHDTDn@;Mlr!^U27 zR6h2AKq)-%E7Z&egv~C__FfpR$p0luwUK8O94QBjAgj8$IgYH~8$Q@SVbbv5;&=%6Ihb zs6bEQgf9=oyU{3NQ`=ECWd!LW)!6?)Te^B$j7M*!OIMLo^lw^U{9q`)G!XA916zNO zmr>Zxza&&ge-7MoZRS&Dz%f%pi2w!e zDxL_+!s1t`@;*d%44>UihU(Q*J5Rkuzi%la`XJ{!;QlODDq^T&2Pwae+vp~Dy76{nk_6kN}s~47yn5qmzdbaZKNLh#&_N-tX8+%8_>o6qt?O|S5@ggTJqrFhF0SF3j?r*KAv3Tp~Elev) zey?vs1zOpmfAhzAK+xjqiD#~2`PB@-967Z^{Hx+qL&7;Xm`sty5>O^0s${BrB{Bmm zxIFsEB7UC=|F~GP1d)lQ9QO5{ycKUbN`s*ktQdJz+-8&TH4Pl~W8pe6&34t4e1wDI z;*Y5|YvU-5U)i;kv++ZOnvAzAk&(>Mk^pYW+qP$iv+1R)`&VX%qt(&M(@(tb9en6z zB<+np^#d+^ZCBxy$j12^=yQrw6C(zzGM9~45jQ8aifEJWgR|{@uqjTD#=-l*{AR?R z05cu&MFc1?l8~S9KH)gTA@tp;N7-C+YRE)54Ey-ub7NZMaXJw%Qi&#@LZI`(r%-+;dJ?M zhFNjBATq=i+UD?~C^&4kyvE>#E3i0Acn!N;M4>|IRS^QLdVE$8fjG;=i$Po`v$=5k z7*KqiDi(7f(tVQ)Zl{KM=S<%0AE;^K4B0$H)m9{|9ck`%7AL(%+;!$odvVgh+hQ#~ zO*QkA=kl~kCm=5#9`Q}8X^1}Z2d~T3>5uMQx-{RP?>#>s-RX2mS3&yW4S)1U`|Z#1 zPK5a7fq24aMt3)~W=7pr?6&ag>y6Xc>=~dv5HAnK+tLozNC{rgL?h9rTvZGw>~2}F z0JZ!pewG-s;=pD!5&H^%zVx||y~Q-~iQJW}sKh65`9EH}VTlAq$9KJb?Em8UCLiru zX2h>k6Kp)+GJ2F#aMKc&ePB5oGx|%A1+Ej#_%M_ZgjKXjz<1|LJ9)%4RVvxCdf)H~RK&X(uq z@88+e|FQ&|`EH-@?(SSW=zdN8=w)VF_qk-c6wdH(3usA=*XzUjCu(Ntxc%<1Uf)AW5#-<3TSe}<1} z^S(V$&;~8*`p(AczM+PqG>CLR4UCP(wx48xSFVh+1Qz5V<)10btHH9)MaTwTqFDk% zQWW9~M^{zhsQBw4I^idSeZV`ZqXmnlTD1_aFC(~~inJQ-turIOGZg2Fa(C@hVsmQp zXo^l($$6zQ_oIj$1ko-x4`RV6FBo*4LGXK6`mzY5P7V{Uq51}1J0p}}*U}BTljR;6 z_q?(z5?D|T}E#*=~nF;82$L5C;P^x0B zzNJ`ZBL#D&iRVy#gK;27>SLYSp66*b;L9Il8lV6yj{LjjsL%xy!JLY0mf%MssLJLN z?#Mq6l@eyf`}B*9#$zFC&k1Y+kHlhY8I(YVk2A%GhcS1dRnCgBRw%CqaWm{O0+&Zp z-U{{rD^ywEk4>RxSLPa_yqj?kDa_QdM0VNhU*PjR9THrgm-qTp)n0d&z6$sI{2;dH zIaYP|?re|0EO@lPPt5a7Cn0Cc@=*at%=1hJF6imA{3I@yC<`3-^UEMG*~|YQ_%`2U zytY0*HqSHptroR-9w(bPiqB)L=X{z`^E^{ARQvYtDXQ6EK1gv@+1*(AI1E!vMzdK~ zR>vxaF^~Js{mUji&lBx8D*(K0=u`7NeH@iy)_15v=Xv@**LVKxG(nh{K5IPLB<`TW z{R3*Arw>f4bUBg%k61!r#sfAb5{1w2!TTCVnFH#`ubLv~bzY$42AgI>zSYE8{`&0EZ|bMqgX?Ovb9=jMA_$m>IIEGUWJKQ=oz-+w6EQ%}tq^bcWPRa9ja zH#gsZvmZP&H}4i7sOd>Z2Ou{$&j--*0o{>HtlT{Rd$rHr6AF&4kFMP)h4t3SXXobm z(6)X%D>u)7uSGRpZl3>odlyF$REP6>+}u1Hkciwo`<-ed$1|LeK*TGH+qd2C0k${1$YhvZ*`|#R^J}Uvye*rHy-+#h;A9-#6BFM+>|FG zhrRrKqT_Kn|F8jWZoUVWCE$@*?8LhOz&;6M=jLnB%J}=N+|8gK|=@r+l|{jVLLPI)NlOPiWemt}rh_L%9Nkx#=9E1tHe6#+#ftR9^P zWus%w54``f^c%HYIC=pE zy|_QyJGgfBV2)1JbDq>>!dk|iUp9emk+#9Yw?uV=b0wx172<}cmc$cleEV+eM(L< zG_D?H_JJjg?NK`=h|MiYs9A3y{t};m>EOPx_ix{t9X|g;{g-Da=*U>- z(H2eUQXu-*r>Eu3HQWBc^d^}2gZ9iDr=9!ujpFrZUT5C7E$os5^^2hK{hlSUC$I^t z&Oy$>3ay*^s&&&IaI-$tzgp??Nm!wEH;2FBgLM}hHb&At-s}(_;7u)1pyVJ>{2V#~ z&NVZcQX?FiOCO{KQT%6CnmA}L=Ad}{LJ7Aj0$GeGS)5x3@EZZ3RculEBHt+ZDP0q z`Y+PC=;aTh;ZI<<&4setx1!zG%|5$DkuR3pYRKbdS`w>`;I+&OrZkIRqzy9ZOhQ%| z4NGo_UyU#bo8g%L0DjEzHQ4mU9BGSeGHj7mo=E^}G+4E}`X+2X-Q3J;&Rt7{}oSWUn8L6H&jkzLr(@vNU_?#LeM%Tig-n_Y#$E|DMnznYXuX~R=uaNG7l5jTS{CA?`Scx zY`6*)Vq;jgY6%mt%s`4Z3`^-QSpTpULt1~yu&)p}q59jvvi!1n`+6%Tpj}9uJc?Bn`K={-_s ztCYZrfTrOnwbT?Atk(bKRTQ*QgerASh! z*~UQ5iU&sbN5nwPuSeWo-w9V;H6$=UvB(>&7#f^g3_9OJ!}w8XZNoNHt%bg%D#S_N z8?6}b5^l9`hgZ>{h_P?BB;#9vq!X(hRYC zb|E6nG*UX%;BMS4Y~7Yagrgk3Iz+c5BN1l{URyh)VJ%)g;Ox{{a^-LpuM7|+B~uL8 zmQV(QEsr-1!HYnQ*E4*1AfApj9`v@-1@H3xcml8YkJjufEku>SB}=fk*C2UffcuX{du{Sak{g&4XgAHO3Ua+@zvH|d0N0a z;U5gd*;w{6j#1}avU#AwwnfJvBSF7>Hj4zt+TlDLCY=3yS#*Qyp}Mv^1M3R@D#He9 z6+J(eA=^1Kj;so|YrrE^e+F&ggM8pxe2DL7L)lbnW!>1-(E|>V#*B})-5$(rU+4L# zD3`R$bJUP3$x9(O=->>dRx5vNEu8TdB^o&t2jsU5%twOkipPh?@9Z5yTljo?XLdLr zm#@-VgYqTz_!06f{n(^;45ZwEavr^B$#)x3*Ocn4vfz;=QG#5hkO&%UIO90PK;lInqb&#%U~3R>f6i30)t7Wpb4Qi4!9Qh1M~x$I5Ze&^7X$(hYw8 z>fYgOXK%hQ2f5L!6v%hjA!fYemVBI8r9l3^S8*Vr-BwsiFby7T4zd!xN`VH`iYrHA zHCRQ}DUkwrkwya^L`C3J*?nhVe%Vciy<)Xut8 zf%bIHT-Flg?hOJu`%b`H_K1gx-5bPjyV90T%=o#R3hUPo=2OUiU*6e!zcCI*E{nxm zlqrxb-5-TLrU+!S#4ikqi6;n9Rta53U_*sRh&D9j!~KG~ww8>WwoycFX9G=Vstwnj z>`8CoKn~;-)3NXMru;CTkvan?vxsh8>=T`a2%zV$zS#A%VmIC+#%xOZ!h+#2cn>Q6 z(a|z|nV92*JfX|>@D57k0R>lwQZ_He;0~g9-C!LTcX!==-`T$PLgNnRI|psEjLl(VuQ@882Y^5+Jn$>j%mxH~7iW7fj8^3T5~bS6GYXEBgGG>4-CRDh zI&-neDZ#HqkX3!Qj_a_=+8niF+Z__y3r3$L(WFD4(X|#cdu-`p@x=)Asm~@z1ss*n zehS|y_=|;PGZGT59ix=UBt<`dc_7}6CRCc*jEd!91U~rrBOY(T1 z0lkT!bvfp76;A}^Vz)~HeT6FTLuAMB+1+HQwnP_0m8vNBc4|8dm7WlCO+MM>_*fAx zt&IV^JTHY-RYB!;WZ3-5K-|N{rV*+$438kyuwYykpE!|1JGs{`m%_#r|~#@m(1NM`6rn;H3Dl_SdE=u~bHrWsgUiH~lo8L6lzAu@zDqyA~zK8$?#<&-J z0*e`KnK0SP-%0QqQ!EGKx2c6(C}eiFH3BJB7B_SPmWiJ-cLfAmyhg9dXm)FN1lEed z?PEg$Rw$D?c(-j(R-t+Op;lJ!6_e4a2w9D7*@g%oo@8-Y+HKU%J!Pc$d9@Cr9R#$^ z;X_ey*lc-?!3$Slah8yWr@WqQUWgZBY1QMif(XP}CSDBUI+@MoCa=(boGKP`Akux4 z3vQ=|dFM>t>>sFU;tbh5MAcR#tfiYVKTS3BljriZNhct$W3{`TMNqF1cb&P@UIaA| z-=vy`=p%pdx?I8h=+31}^ZohW^K+~vI!Hgf;g9}kzx_G7pk`J4@<2S{Go!m3S~H{W zDt23V_4Ss$yXobjcw5?`8Y#i+nP?>1l&gy2gx`7!SK{UQ{Mr;e8)cvR-}sqyR(dOPG4j`6{K_tN-++*&-kMg1op`&J0- z4sje9uq){I?N|5+o)B?QXlyed9~{EIb8U6d@0W!J&CbKyw{B~DRtDypYXK8OFyo7L z2q@IMMYWU=pR$(Rd(N77jCNBS4spaWerp-^QUcf*tz9_W-`SpTU)Y7AE)^ERaoJGCR1e7S-P+ZlCY&?p!yYLx*oVE~_?G<>sSp-1>}o`{->{+}eMfz8~r@;e=Vs z7myHv|Mf;Sg#&HBmecL=05F#}iUUd4Kz=^(=HaR znI=}w|9ptn6Z#>xbDmfqtN#t?!~y>!-W%AP3oWBf@r-C#ivRBcnvnGarn$4!}aRzUQlaUhxWj3C!Ul@1vnibYW{He3-Cnmtry5 zNqPyGz0mzUUdl#6XO+Q5$GTNmxWdwZ)EE18vD|ikW;tw&Wi(74;b?WV^7Iq$dj~(=A;)6ytj5dpg!n*DMS5NLjX)qF zng6>}%js7?&AP?9XS|k)v8=fVDCxc)Pw1l zTX#C+9vA;*DE^nCvL@Q_kW|8M;!w2FALv%>4tMfYA1jIr|0~Sg# zreDfEv>gd3f#$FW@&63bChcV0*HHj$kZ1BM)o9T2db7L1jt0{Z2DN1n(Y-*svNq4^UPd zFAjr#HfqH=8x`VjV(z^%UVV~&QG3>W zZ=Rj+|Ilpr`uv9{-`_t@-%auc4Fo&i-+#TKjgB!XboGe?0kR!i+%)0=KJ|C zwkXHU_w(Q1c=cD}!cu1sWrm+!B?_`{X& zug~N}2V#_Y6CdD3M=YMo4h*!MC0=Ak3ZI!PuclrAG;TZD?$QPSV(T~*{4c?hC#!#q zNEM3@C4IQ(T&WNB?^HUm^96j^F#5&1wGJDihy42N5YEUPv7n&I!KnBL=;RmoSdZPfHrskxna#f64V4FI>HJcx85DbgcQI;=ux-XPqzB<(RyBE{Fc~(YqY_%ddjV(R_LV zE{6sx*5%O8#kw5&IpuO(R2NJyydo7vFfb#HUY;eVU5-iTa`2)sMm#mr>1ZJAbY5;8 zCc){L{IEG4lXRzJ676)<(02r2g}z!_97136f^7pc;=A1H+}h6o2g`ChcvrvH_&le$ zb2=tz-WBRk*;|c)Kkkg5hSQ}-NVWSueGw4r4Yi3kVW7>SXxv%+Vc~7O&fsDLM8~7rsueHf-WP( z;S{VEB?^qP(N^siB8>PeM%#BRE$Zk{u)03>wW=(6F}hY?>w}F#cny=TYP&cw1#{K;(z<;iMaqQ;D5E#QvDp}=t>VQLOozF@ zxWLKH3h#)bufJT`S2YP{els%u&aEFgJ{a91GdimqE0Fg~8wO~QL7vOFyTxBHni!l{ zg!Dv8Exhv)cPxRwF4^1fD*lKus?a%ttr>!~n9duj5`0s<@4z`Gp=^ANDxlOoSmEM) z$R1P>kI_!L%W#<&#(y}Byhodwh1L4M%wh!{;@(8}Gza}#sZ=|x6dqhAzR!r87-V)T znB%m3gSkZ3>hO>~wd_AJf@uxlZPqdvVZtzlhABgsN9G!^?n3$h8?kvBe|8I=)u7(QcH;hZosBQ462&^fm9N3Pb>|BAe zC-1>Dl3MdQf<1gV%BSw7Wf+r43e}rEA#!#G%))O^V zEk%)txCBjsGnH3U6hwc5s(|6H47zy{|3IS>!!oukg%1jqIVab9KPBc4X$ck{?#Y-`TM zW5VhGDvx+j+*22p=U`r;pvtk0QdA?fo8}x^)r{A$$6N(D6oyZB6z0cuevl7nQx2AQ zjaRP%>=2;L`D)vDlHucLDJ4}rmqK(<$~oMv_G8p)Ip;M>NV4b-6i5+ndaK0F-XZj` z&$oAGhx2jyQm{41#PNrxkca8}?*1|`daDHaawA@wQUX>OK)U>k-YSs@oH>2c2;>q? zYRJ+IRUcLW2cE4GiQolGzMjDiwZzY@K>Etlf^wqC0*1mQxm^?o6lXRAaatf0#3 z+n(4ek@#ARVwB^N`1bUz666PiB^K+l)aI#@d{*B>L=g>n(14bM=D?pWc}TU5N6W&X zh%AK0gs#i=%0vS)i54T1g@mXJ=ry-E11;fNkCboERtXYXyGSDzThvww`h)$wgZAC( z4NjLpJRX$}^AV(eiR?5Ds1=ir#q3dw3YZc^^c4XSFVU*|O{TMh>EV3u;Of3kV7*pX zW!`Q`ZBqW}umXQ#@?sAy-R>($2 zeD&VCpqG(QKv|bT90v1wnkP--p9~|+C97I5Ck&#ZDO6`1g3|u%QJvCaM?sWSb=|`= zHEqb)S@N9xC(m9I;(!_`fPDY_)fc;dTkINq zMD>qvH!7LF#9+J)(ISg~b+inhre;zhk2FM~3Q3VC6lf8gq?s=Yf8Z(m#sl4T_kCx1 zMgR<{;^VLN&xy$q2sYF~sX;f$?*sO_o0%JbUzbUBJ{WDaDA9R~980qS|04i+3KLwX z`ty-P!6#tO(GA*~+}L`GKP4)PrZAsj`5d_Sd<-^+jbH^JQwj!OCMQiGD7-k^dttO9 z|2IOy6(}iO8be%lpLN9LM^2ZPtkGcN-;vGf2u9R*>u3<4xXocJzD*)gzGxImI!!th zCPt#cC3{%>`xq3e@21EH95#Xt71C2ABpR5_SYWiejMFKT6o!Bi{!igvY*WfmHva^< zCplJm2NcIZaN_?Fx zD2zrVsNh;>JzI#`EXq`6nOLK9kcm!D6uPFNY{PuK3`%MnfhY(}Ay`#X)A>?p4LfVc zD-Y-5&j;dfvoUIf`H;*|mL>}4gc`^V^&jRL2=)=CAVy}ZS%Ht6st{v(V|{rDj?Vow zQU}*KVG3awv;uv63?7{tfv~~-45n9Mrhc}L-MM(CwXaTzpBVieu6-pxXkKfc*!Ze| z7Sp?#U>K5{bGl$~-I}gjUkKI6lpA3<>I1Br+pBd>j(2GQ2X-E-uoeh zI_OwB2j{5ViqDflBSR_`rYxD{VyevXy*vz~KsXx03bhmJ2nh&CZ+4S%OASP2h13-< zGeYVx3WTTyHqNpIir}YUmi4jbhl-CQDnwhyQF`&gfDTm=XHq4ol~S4>8y!WW2nBI6PfUnW;( zpx_vH1!Cppo~`5+B0fXzFhgOp6S>6B3iR13k0-o_T`o@1E5dtQi9aHT>7vlv-RLox z>$FTRoIZgP|A8;Icx$i@Fp)!)AU6GQE15qjHb6&Sn=>vc>;J&fOxAXIX$z$n!(r z-mkf!ihoM>7CS)c2G0d|fYXBzy(engO@rnTRVn6u`QqC~e~TPLKpxD~*-gxEB3~s* z+FcmjMpyy;JVXHw`K1e-F77Z9iMB-gvxir&y>oxIy))lCWGBE@Ta<`@J`hiO4eibr z)MBFhpk3bSt~`5*>~XPrW(Br@%IcM7rr_;cx3$%)19J5fTEu!>)vs(i&;A8hU z$ZQ9GCorVD9wdm)ADd*9~W3p{@_9s)iyGKPv zBrhc>0?8E#^%Sd8)x{G|#ffTV7yqzg-~`$wwk`h}5kNdWo*hEZiJT@fKA7)b8efnL zp+~o<|0Gbrik;mdjsqWdJqGm7%c1xPG2PiYqOml7d~gVB<xj$ zhX93ii|TY?Sw5c4DucXZwEOKi2AKC4zqKsT>S{kWMr#)i_jk7E+ZT3WIt{+E)p)#O?Fl-JNR( z-9IjEK-0~0^Y`!U>0eutzqU7j(Rg-g1EglYx83~`Ww$jh^s=cWFL?C--z&! zgA9do$+UUG9wFoHt&_;$axHgX+;IjsMcT{Y|7d`ZF?Dcdu$(?dC-AghW4b|^2Pr4= zpIgQ+`4C1dp+^Hn&N~_oOc6j$g9;r(>wJlzU)IyK@S}3>s)GdxKlE*1#N25IAs@Ux z-@iF-l9&Br1*bgkptcC=o9*@5r~lEdQRCAO-*j9ShpGZm?r?PPar&Wdf1xOtZRij> z@9X`=i2Z_5tVIWsY{h}38-2b;tm4W5&@3QzK~ytHju6DJ4#nRs%&}J4qFYe95E1FK zt+%jr)l*nn4Dnro>4%pEI-DhghroHM3r!bHxeDEllyUH8Y)Kg`duj+966+j?`0qpU zbXg{^{mumt79SLofpZm@=i{o3dX=NL(eRZQ5F--`s`eKM|os~9dGXH{Xw_NGP4pbpKA`f8_ZaTYMx5e@WsdCD$;h{tg2HI)$! z2B80WfDg|l099pSeH)Tp@X8%c~J={bansA>~OR?T6y}3_q~Ikz3D)vCYU;8xI8Z( zAV_B2?l66Y?+*`96HJpK?FR>_psedrtMOKa0Urozf@v~vW?+C1o(dQg2k*ZO1v(Ei z8N6Unfagnfd}!Qac+{*vI8+NDf10TWR4D_khu;?nM&N;*VCwmvAJ#S;uuzII9aA=e z)UI|-75-dJ50)5~H8sK1^HwkJ>oCB1R&0Tb86YP6H+Ds7j6oM7sK)-H_Z8LJ%=;S)@KAMCq`w!$nch>0FP5_^!#D|r@MhnX}tpnJY7 zE`l*CIi3QYiHXuO-Pz|(93($$UUXhJm>NY%zT1#}TfA~Sa;#4x8xw*WMM=VEOFKHW zh-5VtkusoWt$qTvlNeVK; zca3!FzfH8Lp`>+O`dI8$ya0k!?@D*7I$50=%J%E!r1py}^Y25m-Rtuoo{WG0IDN(L zcj?&~|NiR@bz~bJ$~e`9dxZ0zoObA+q7P(&jDP>tegI9!-UTr)}X`|M3_Ob9-rZ#m@NGfTjlllA~YUhfOQP=R)!` z{@vUC8X+S!j|Bn>U_-Do{vC|A4&b>w-Sf zKrALZIW7US7dm#vzlN?*XaKDga2)aalRsP;|N2Z$cpyfbM`ipirs_B&H}x-GWJa)` znX7`wdK=Za?Pj}6*ZhmE<52Uz1RJQ#CVpswvZny7h~b$f@u8#-x6Ldg{oz};j;Rk1 z*7~}Cx$t*4brit;uCbSc{_ejfrb;>dZeV(}avoY3tCL0)$81=c;GLrW_l{2|J(c#Y zP3h{&7C1d9`DyrJ#nXJXq8WTye2d$$)}U>4toeZ%dNXQwn9mp#(nlR^tclf&bycuj znRUwH`!|S_XR6U{$NxtE-#}Ik4`3d~8vJ=gd~6`lCU1w5Z4uM>8f+ETocU%g5hu5* zru}B+O($f5$q@bw;E(LYF@=QgO_vq;*%Wz`SryqFR(6-kwx>=B5-JWFIGZ6SlyVd|5c2cY!Dr+;&$g(-mnYG`KRvy1+PP)lC|-Z&b>=PG0xI>t zyZ~q0ADBWz{5|S6lz+Jif9WPGi>vrVUFRg{VCmRRegC@26S!_4>fg6?W(TH+-^1c_ z_+s5%htA4$4>&u7CzxXv6f`aE6HqJu0n>na?8dd(-u9zAR}RM8JFNkOckEm41Z}aX zb+kMmyZin~bU#-o3=&)c#PRHs?iCUosnRb>u%s2(=5m>g+L(lrQQ2GwV_&+RP8b(j z>#s=H2&OIZxUe1EgN|KxxKJIiiX{z~0edKG(JelRk2!UY21JG%!zaU!oP32R!+*dX zqoclAOEDYWOnbTkCku7|mdgduwo&Jucu<^UD=jd`y`=U;NY#tT<39bTE; z7#(YVsCcjd=vn8BbvY)s%Mo-fNFk3k{Pm}g-sR9=eidAf=FTU(D%%3CXT|Sf;|3|N?&de5o@DtGHeuMlN0@E zXwKx74eJKrrcu<@fy`9{t(f0_E1NB`$fAuktH8_z7`G8UV zkWzoV06^(D)i(#?edWojx-IE2nyNuQS}=#=?-d8rs%CX3k1a@G7zJbK9@~N#Cu))Z z@p8C5Azoy3gmmE)tQI8-42rdCw-A-YUk%0i(kx#`hl1txDI^Zpg2%sARwjoFON(Qk zS_l)knT%S*g%u+gVlQVAuYhLv;86R)1{qdB^d>m_ z4Z)IKshomoZ6yyjN+BT(E^H!rN0cHj8iw5!eI?hw@Zkq$mt^QWZq=fik$v1752Q%C)$X~L*l{F znV5))`T2+&?Y}YVyBac>-!N`9XcvKs%>DWHbZ>TLJ{pdeHp&5G-%dI6K4e$kgJ~qS zZnt1v7>c#36hHYcn}WEJ zC1?tqsWu9tDPA4;^pu?MitbruAld5dOGI*~4+<%CnvlH1px zmn-3>JA2!(!s|e5jqK{pf0n}Y{j)VYUn2~`sx&f6?|6NMSYPdBQH>MN0_C$=Ou*iv z+A_F!X(-;6{7u$U@YN3e4Fv%~Oof5)`c45!c62x38khsTfv z59MmEmOJ_U3>Wp~`TW`x%DbVN5n2Xht@qA$ydNvQ z%-0*9b8MpyaCV4G_M7G$S{04gu*X~lITVIZrU3ebd_bFeu)J%$dKF-Y2xZP!+s2a& zA3sYesiLtIqJvV-;cmDbpw4yqr?$7VkAFIg?m&SQ@ae@8J9~%F#6I8NnH|o@iM3#K zQ`m6Bvz2lU+Tr=5Rmg+(Z{=k{FP0!*z^?%#u6&pj+H{()ZHeOph<~m??_qJ^e2XO# zfive%^1)LzVTEZo5cFb+MDT(&U&LaG#Lua@1f^ISFWL%L$gE47SU#(VeOy$Vr++}! z7bU{lS~|pHiNx1h6r&uE#J4vSC))&I2@VdGM6AnNo2PJxAPtQA9wMq}(q5@%P5yMr zL)MEWg2?GEvjVh`1DA<5R}w8oh}ziY7Ct+Y_$mN}Q>-PgmwK>Gf8FJ}XR!nctsS0@ z#TK<#g8pDX7i!m6`@+LT)F>XY4ieV0Sb_%Bib=;}wx}S4Zwa}hiio}yJ|$IF zyKXp{&JL!B^Sy(s`#Kc;T5S#8v9;DrLK?tevh@xAW&q+0hN_=@c+w0Y7Gb*oYBhTJG;A7$esNfo9I;;G1DR|awiwQV^mMyM(_K^_q*_Q;-Nk}|R?IR(6 z-IY7ihHphp{1b8?2?Cc@u9qE6p<3gRPxp)WxB<;| z$wuvHqv=eI8#4asIrUH47m$O!i1t3@L{8xz|GIC=9f`(#j}3KDYSImkJwD1>EDK1g^TBA#MWbai0uTVaJsuw-aXms| zz$akP(GA*~+}L{RVICWR7)8-k2CmP+@E<+~o5Mz65Xi(wT#IY*GC67jVYiF3y%$C+ z@_(sXapWNdO8Y_n8Ka1+vI9QiE-i6qfc?7|;;Qe~(I7r?o5NOon?#~~(I}L3n)p$; z#FUSJAA>^m-4ywN!$z>7LV5}gqk-9s1t$L1R5(ik28Q?(fzYOuAu<2d3)^{z1W?H+ zNJPou%o`83)wL6GYV;GH>RRNy;y(_=yYUDr^4o^nW5>lfd(X1Y%DZS__vpDo0GWBF z9=%9w5y`2kq*E5$;*}va7bBWRR;I$8Cs35tO^SIUnCl~$wzG;Rf%5oUo}kT1iLa9d zh0%xvbyDlBXA3c#MVYEB6L*R)ixqt#+M0&45A*RdD5-4(q98DZU{y&?cXSy2^MUx= zY>XOVJ|`k)=VThWGgMnxOp)ke7FWq&A7Kh&WVX7R__(PGF(%Py%R(HvUn~!UaE%l0 z6^218(8tGMrZNs=s10mE*x-H!)2lF(ljG42E7_C;X49G0Y&s==V)S>oW|RD&d98VZ zu&d%f4aD6{Fbv5p+;)@$tgS@EZwy(yrbuJSDi;@3I_toFi6(2`bh{||>%M{rgy;~L$5QO_%UTp*(e{FR|U?8iuG%>`mD!w$JnjIS~ zb}f$2%@1dW=W2Af`&b{6^-YQ7om)QwC6c$)Zc1Gz`A~a-+$$!dQ{fB5JdtsUh%b{n zGf;4hyTVq`PTtiRpy|!KDjBic;xhpipCOl+p|IJBTw-Yj`fNoMx1+!8@`U&9$V3BS zH2T<2{1KVm!vV7HkQmG{d(bnHApQeiZ1L9MplJs&>xWy({7JDnIvpdKv0dg693;|0 z4u#y_a>MQNF@oevTPm}G4665W2Od&lRa!N-W=pOU@B4p6$mbHN?p^dv;@iJErPpgBZUig{m-w2gwakxif-@$Pmu zQ@uvqb>>ccGu1%+7CDB1Jea4mn;76kzDklb;n2{}LloeUUt)D}hlxnECDNbG?AZS2 z1M#HS(C%zOEhf4T+T|U65g84<1rP04l|nf<7%Yt z`@Z+ryEPh(m>CVRh9L%fn%3V}QltL6f4ci0NkfbWu!9L?C0^fDqx|s!1tGlYs-uvwD?7dHY z`_v-sOjrmOUeQmPIK9-BpSWzFXt~&xzd3W@gt<#>TmIITdV>>YpIdzSc~^gtEc=1{ zXpP`7NaOk4JJn_Fa7Cv9)5xq2)jJ(g+^@jvFL#{jFU9i-hWP8>Ep?>c9Uo1vCAVZv zD*9`4tKGFbc5csGZdxV2-6@qT(kabMFJv|sn!*~!8FeE+#aeacmb^n#9H zLOLQdimvJ}_8Fg9w5Hgf7F~Qr=dRP2P7aP<-ralk;B@!e-s#J__qL_?4%PGX7Y~li zZ*I~bB)?;RczAIAw74w$<AjV?*RVK|IRYF2&y)`fpmEGk9uAb~RdKi>w(h2NaG*3)WnG!B>$*b|_e6hwLJ|5}92-jaog?RsANG zP2-!J?urf;a)nmF{A*+;3KJ`!7lO;K<7SB5C%^-!b=+Ni7poM#nykjCu*e!Idti|@ ze_zWYYki)LT4duZ6w*X0#tkXa*Mb=!^hdC>mp^}Mga?a|mxvMN;6E3nkd@Sv&1o(?2yl7<0}y(LUu4Zyi$#`PB#g7mvd=p6VYtZZz_9K`)(nyR#2Oaa)fg2PStBJ{WX<2#lF5$v zMK-=tfkk#DwDG(3EV2`BksY4PB0F&xS$Yw27gBZ+;`c~bqZNy6`9<9=79)$g87~JE zt)Kf$R=-FlI}t3h6LyhZiOeptMlF7kHCNTO$kv$m!bMgGhIKEpW{BJ;*09K~#;CB! z8Y$5tYyQ5LMRv?DvhkG)EV3)1yUusmOK!#HD>T}ZopOup=v)@rX`_qmYP4dJt-R>* zi@F&v2UO+pB8%))u*go?MRp}JyT}@~_(j%Sb-vpLW{1GpH-L+*4h-vFWX%w{Ppn~) zU5!y;ku_4HMb`X1TV&;11oq|ve(^Qo7uonq1s2(r(Cb}fXWSw?jxVw}@7u^1c%OL? zVHeCAU1V3I6^pFJuIDzA;9IFkCTqsa0k!T$b|zS4XY3-o5}93OjavL7Yp%kJEV=z! zFOR_ZrMo2o7g-$`*1gD@A#$H^d=TxdP%AOslU zT?w6=&iK0D|95Y(vjIxc9~ULI;1W@0ic;% z) z$OV1o#loh^zi4W$`a3Q4W-t}}nj2rYI&N~}n`X_6uAW`I=+4e((bdH8TKg0t8Ru%a zV$rRfaS8{_kyUxX*rMB)EV_Mu(OsE|Urlpc$}T(KMc2uy%@(eL3!)AvYhQHD7`bb# zW6@m=Qf1LKa`w=oYc8mg=i)`T$1l3^bqg)JE3r*9s7OQjHn6Xd_lA+pZSj_iLc9_? z46%&`%_X*1U8lM#p9tO*tMgaGHMi=T0dr`rebx0F#K?_>0l(_5q~@1Xqgo`@HJ4q) zCNa3`>WH%TRo9G>yT&?J-PIsfR$U_}Ty@O_RaRY{tb&;7o~z-CWw+uw--@mCX22XsWSIgH%~|jht}V zH5XKQg@%`1t9dMO;X=#qO70?^tU?S^XRF<+>&M53^Ao+SWxescr;FxzcaeLCX$-x0 z=S{1V05;#b_`pT}KX+dVr=OeZGwL1L#ta^{SnKe(t<}i2j+^^6J?>&m^Z)X3Kfj^A z1S;&6`5=xDRz({&StoAO_$7Vsgn(PR9}L#Nj=J~Co7|@F4Sn)z_IZfBIoqkz>9=3t z+yejY*DZ^eWWn1n_027Hk8M+Z4B34C*?Ifc)ejwP@d7*IV|nxmzpp!8jqSoMP~b;Y)z zT=vD`a;7)g%0(7080cULK5|<|cJ;$8^>CWF?nh!mS+eiHl##`@BUD=bg$=d*hAlA@ z&cU|9XL`I3jf9_oSss>G`ng$pK!3uetor#ETU=W(pAWc{tq|%jZ>+u=P+EF@M$qzm zF7;ToQ^}i(HKdf~;`aHr4K+$~^X1V?%TxJNY}=`@7_7}2Kh;wAB5OD&Y&*+P)XLOp zy`|nOQxDe-mx}j(AqM!=1Y&bSQtUwE7QQ>MgYR`gGX6{r2T& zG}O-y&R**PEptCCBWTQ32oQPS{Gf8?RjKKCnEf3Od6w*a2Rr_0?C)*4bk4;DfO3;Z&%wp=_QTKQ{ zc{zxuQ@xX~J2Bq68a~VNVDkJWW()bQojdi0;mwN$rn~%7)7$cJ^9I#!-(`oJ#WUUY z3h$O}s_$s2+bv-nHzYpM5*s)0s@EH@x^!hJzvJ7SWIxeTZA&uW?QE$#ygiHQi-}MB z6C)+zz`DuqE?aKTwQRYWm#sNjv(=3Vy$Ig7A0D5cZh5IIkw`Ccuny(>UEyasHWvz? zo$MW*UOzsa?}C)Im0Mi>Y4y5HwjfEhMeYO!lgncQ2S{JlSFQwj?)ZhXYkRNi{qE%t z9Y?NQVDZ4}Ppa2vntkl}Me}S`{OmnmY+j>|`YQFt?D5A(^WB$@&$f<_c0aRscw=r7 zS1KdaABu9|1#&U<)#^>|vtM|@eK@$0NaC_I)E-oQQnj6j@6WHFy=0_rn)id!+5BYh?D*vA`PsAcSI@2tI>h)L z+L#?+hr%81S)Ji_`@Mf?{@3bWRvVvuCOdpuI;&Kf>T|7?CVHIqrv67KSFFPbP$=}SAbY-Bz5dD_ z&I>aCvnDq!HsL>cu%Dgyt)5;<@a+7>_NM;FN&+oQxly#Ms?22B+3m(fY3@b|cFJ7c zmOb_3i?=>EFaEg_a3QKSAgg4d37UxiOs5f7Uz%Ti`TDW`0_pC-(ZSil-r<4%66DtP zllcn=udWn&=N()7^Uvt_(^sDS@~^nE-Q6)RFx#12zO{YRPVd33t1lfK&GmZz$}87) zHHLOS_CT+{`;iAGH8Nc@Ye1HGo=*815_X`m^EWHni zmgT^^d-~A#?4^EX@tk$qh-ihbcZnoh8D&uXCa7_@1x{;GZp zK~K(Z?}C8qFSXQrpoju^4Gn7U-}yj}npJZVW#-^3k}N3IH|h^8L-B+VqmkpTJvcRQ zsh`pB!jKF$$YN97kr}!pSGXF~{&yXLmr&EJpB3m#ivpaw4m3r#pX2TQ>b>oENAR|E z1iDaK2jKb_1aNvyKz`8Q5&&*$0JpXCKn}Q7zkM-lx~kG)T<7}5S|@jS26QX(aXhY7 z1>MhK(DfytGt&)0=NcP=E>zeD9rIrI=H)hS5afQoKwdvbo@5Ro`uZSr#f4+yS9hEv z`h%+5em{w5!qd38H@F}$)GwhC3RT%Bl9!0_z&<@r(eIjXvWxT2%3 zABg&!e`v`cNPuW290JibHw0p+un$DMUAqBr_=^P&2RRO9lL(PG07#5V#{qGg1Mxxi z!FHDfBI#=!6o(fCicAVRlt59Ig1&NmvR^j{b+^KT_`iM->MuGX5Dq0EG*b;h=o%V= zFjUwF;X8s|nb)g3uH4?f`SPFp-W69l2!}sgz;KwuFsLEC34{O~5_{>zE{G&a^+p1K zj%&(+Z^VPo*|h*WH+Ag5s2{~~a3lfWGbRo`zVm1v2dVFKZg1_K>GXDyloBT;%+K~} z$xl~Dqpu&4Oi?(JAkxe~M5Jqdh{#Z3ACYV^6=3n-6j&VPSWKEv2+&c1#;5bR_K13@ zJ+tE4(N(=mc(Qf4_uBkqH|{&N)R7!s$2?ve+YHrTw?}#FSc2CtGkASs=dm2GOXBlf zZu9zfaaPA{#{uiCh|G44`);Y&C}jXUmH^f`00gXSehAo5;c9^Wy#lb~9I$B<3K2W* zvWp0pxqc)^_@nCK_LU@rGh0O*WG5E{vOk0&JCQ*4DFfMucdq1+rM_cLER#!8q)+8Y z-g8}@*BnP{d4M|MpWAbtNYHBL9-`H?JVa}#a5c34MuFBzj@Ht134uCcbKfU(EY8%V zJt~1k%ZFS4zsfhg4yrKn4L*5`-H*lbZ4An_Ja}3p3Vkp14Hfp$%4VAZQ2%)W)L9PHQu7EQI%6UFSPs#V8nk;7h`R1W zgOk%NdkR*v?^Ea#B0kMl4dF$dNixqxPT_ zQY!{h4zLjt(1ieOPZCPSn=|Ri)`x(cETgXC0J}Pra{bX%zZZqh*b@QP%svLJYkLgX zP-!2qTucc`c)xC8wV(!xuv$zL#%d3|QDKGTHknf4T?os@Rl_9#4AaJ7s!w8pYr9Kl z0g@)|BxNbf(o%Is){s~Ogsxhoa*@@c62Zu7<{tyqwLS)FsI(8%Xk`|n^@|o-3p!D3 zQz@8MjMqLJTN8Gbg0aDZ5H$u1Tf;m7e%0738!jDo$uB_C^q`q_`jjkHx#${_aDd2F zLCZx~hjs)(Yi1vV*0ntbZK$*lS~hU`xw5jOYe72_Mb~0JF(WwnM%R$21B9*$R4%$Y)FcR0GyfQnzqWLA{AFQl2rR&_8jG%fgGq{A4h@jB#5S<*kDH%do9`d&oy~PK z+|lX0#@HG%d4RxGA039wQ}KtzeH~agW1GL zwTMgNtzR$*4LFWwaWAn5umClL72rjU#Z@l-bvZLY(sgVu5!12^zwG8x$lU=VSA{4S zRUNhxgs7Q)3{lti7^0!lK1BTtHyY6Hb#{Xb+7g`u!@0sp9e2fBLh$JTqOlZ$!f8%; zWC3e9E5N53i>QB{iKs4Z21qIf>|^FV&v<8;zvt zD)*ws5LJK&H3m&hd4M~>E>{Lf`nW;dWM`Bk?tU5KN|Js#Y=(p#AaB(`%3;%?C_!wR zsmIuK4Ue%ID(z#F^UpBX)x3xNjcUPADl?fFoIE$_(HxvZ)o*W?04GD`<&xh^Y~VD+ z6yQURq4Rqfbh_LaAn8K}o&BBd9G&mw&{=m88gg@hsNxt;BUo|3bl6A`m}b^7FkPEt zV1`Op1Ey{(&4F1kl7#b!G0BsWW;rIe)C2ANqnNZjr(POTKSD!l436K2;izR4(2yQ6 za2)RRb2u&!$@WRTt1Kjq%F9GbE$I0ZkowyaK0`k55RjhLAs`8*eLyl**u9|H`&SB7 zYK6>sC*tXZxa@BeJ2QF3*L^6*WlwFj9}wa)m2|XJo{N@!LpuQrdt>m_oopzV)iMg; zDQ-RO)z^B;K{aII029+8(cw_C)PE*+_C?gA+y^jJ-5J184VCtxS}v*8?QAA?YF!LU zsRq>OtYXCWSi}mOQlsP7T!i)w4Fz~rV;j1Eivg}ngaML@dr(;u*H?~e%Qh5bIpm7C zNVyYWc*~NqX0xShFyk2 z*mwCbKvFTufv~Yva!AWZ?sbe7_yWLI#75GhF3(cMuyw7EVH+y#!^Q4%(%Mn}XACc??{Z^+q+(h=X{S0PYe>-nA`4*1?mTgkRm4lumW@SL*Y+5qq0&A?6C$hC zrP}9tNz9~Tq>kBiSIAjnMg|K()DSFS4Lb$+RAUj91FXxB0g|QzwrCWT@6vKx#UV`x zi0rI}QxA@TA6I%55>ZR2d*?(iT(LQYeKEoBW2H9U$~NBddt0)Hm~wf$CZx12t6I z2Wm8+g!@civ})FZtrXj53Z@m~m8UEvY$XL_1H?v+!NS(iSb$$O23sz=x;?~9stpKL zC3bv?j;3;l6*93^3)>K6nb=~5lI zTpA#$xcRhK^?Nm3JPqkQKwN3UjlO@u#ZwW7Nn14rR@d$rtfA6ASj)syr%QEEaG0nr zRWO$rqXYJhjj&4I z560xEE9^uio|*tTE-44zE0 z+rTBh07>P|Zp7!?KDPM$<+H~p*S4N7^@UWf;tXjxK=P_s<+554_edK!hE;b8FjhmQ zeXN$NIRBtvGYksu5!vh(omPxlo`NK8c8dxC0LuUbX{UO!?m`Pj}Q zIVK0CFj-?!t7R5$|8nb$0L_XiYKJ0rQtko}vEBqg#1cwJ5z7>M%LZ2UhXri4LMGs* z%q#@%aJ!&Ad^ktx!|FZNOG>o|{c_<<&<&%$*>zW=TH&oh3Zn-S^*DTN$B&BlG ziQ&p0!Sr~J;ISIE`yvFpenexJvjB@xLsU6$jQ$XVQI{$MB>j+yu@CLc@))akmrl;M z+%pZG!XvkBo$)oK>;Re98DB*_rJ9#m|}5?^sYRcuHkw6@!&0 zD(&Q8ovLwrSPrb!LM@9~!&d=b4_ zU4t>WW+SVJuB1hc>_6!%?4kr?HdNZjEEiY8{iols;9Ag?WN z7PleM2gqC%x6DrQP{eK;n(4>5b&Zd48!GMNHa*4`>?T=^E#?UWc0}y*6gyDKF}4^Q zz?HKCi&VpV0lw52Qn?uGGH-ySOx9Q>tU9fT_G-XlHRJ#RsjFg@i?Jf&)6h5o#;R+6 zjMY$SAFI)z5+=oVj4g;yvKVX69fmB=fJ%NmsrH3?2@^$ZE-)ipXNrkX?WuHHJ}cugRs`07=WFw}jp6+#e2UKR{;j zx?gNTRqtRAj5tLTp|IaEFuj{1^9Vm3GH8F z5b5${fTWKappAA0InX{J09uXZ=O|?A0BNhnOJ-|#B%&v2QDc~N&5kh{D(z#Esku;3 z>#thCEa*w{`NYud7q46%&T-gR+wBLXI85aqt#l3di&%sjP73g(#106R9&;UXBbPV>B$cz4NZZyKUqb>9kXaBz z*>}C%wy}t{q)i(`scU?U(okt1r3qa(7Heq~VJ$JUioseBYl#^ez%?3kSbcqmG&{q#g%aL z*y?H>70e})WEfzkm{1JY8GsWjC6m2smbuBS`Mz~VS`Mk>i^5=HcAT_~0tQpsj#Jox zSO-?M#0_P^One=SXiT{SKpK;0euz~bTJgnff)l?G(%_gzR{`o3gbO z3;=*1SD8USi(JET0a3v5<8#W3$0BM|g&1pfB0Bme##k+{5Vz9N5|RNBWa8@>FbSy?ew>tY})i?PLoV!-lLr{sWH3=N>nS%F2WVZH#QHkJ)@G1leY z07=DOUm+OVor+vOJPuhvKp%K!*EtHSch>Lh@P4J?P!VpB`6#XXcn`E<^qu5MXU8;mg#C8 zyF?oxsk~V=|0>ZMh{amghVaToRT0BUiyFhLYj}*;P-!2pOibasTE_*$iR@|xlZjzE zqCwJSB@C0*0=LpeSRg5jPD5z{KHnHRnXcBcOR52q%DY+#Gu1~O*1}!@c#7yuTGSXk zUCU#5hD!VJ48kdg24!`%jteFe+0}|>6XSE7jznn}EE0-d8CZZC<_d88#sJE6wT@lx z43JdbteRV|W^%BW^8#Q~L|D?I#<1xc9b+?8+Q%jb#qceS)zvyK2uoyFD>0E6mONES z+SN)7NrR4E1Xx5GmJ0Cm#x{zWuGX>3lL3;-yIMV~)>(drOdTNYI?GQHJxLoihDq1# z7?YvWJ|>y%O+i=dxS%JIU9IGNVrcTjBx$27IVb>URDBkqhLZw3sj-;(W4P~p?6P8j zq|cbnxW{)M&HK)WMBjOTYwt`GC(aIzkLZJvoDEZ!#>afeMhkrb0;q_Wq(zN^)tvzh z)=+65tmOje?-fn0P?S(RlQ%enOqcx64Q>ZWfU-! z#O>&;iR&v%wbVl7ir@Rv@tFHg#ze$Y%AEkRl)Nc`$R(7nhFq&1UF%}r#Zc|&@o9z7 zoout4xWabyxBwvXa%&kEbSH+S0zz41Y4H!^psQsRV3&Iv`IHIB5AR&b1M;>kAcJa_ zBR*(>g|Ma3D3AKgX75BqQp)Ea!`8JvhHa>HHQ4@U5pcCG2DYVV4I{Q!?A1P*N7k8| zv_~Z(tK}Z$qbjjvuwXST6#&-8GUUI>fYs&607;)VfP8f4;T(|tGJveHn4N@_9Uyj9 zxH3GYiHN79MUCXg`U-oo#BdFj_TkD@bEuc~_btp8Jf(z*#i->!BKK4ds)fA(QC37$(xS#->slVe zHdNY&Etd+0`@+^cn@K@bqTboS*~C~K6cJYV4h9aAkoT62u7;-q+`ch@{vZ=wU8W3> z^hpz4uk1|o==xwJy5<0_GqQ$+9U$^LBdds|G%>S}6Jgi(7^0!lK18E|B*f`&S&_A1 zDV3a5jMO2(S6ot#0`Wv1jVEf<9!R3G=Z0)VHAEHQQ;ns=9AI6l43IP(utmqCyl7;$ ziYFms2Z&r1uw3OTqA6)nV=>jWJqB#3v=7*FG4k3kzM?Ss~bBv10Oqiexr5=GZyJ~3K( zLQ`sVEk*`dO^&XH-U2+SF_?1E)g{;fNt0gEWJ{sbcUTKhL#7W9x++k)=qjQ&P0ajb zpt{z_Kn<1lff@}cVclv+*Mi<8imvvwV!ZM!r_|_bj|~tTH3l194bcVoRb#N_c7t82 z4Un|NZZOfGR9oBXB;@-5fw{z80B4peAtE_xQDc>>Yk3UWP-!2sOk5$_R_)#3g5*T+ z27}qeNS$$U{a#|XHevS|2t`Af1L*XkkeZnESIcO>>jVB?BK;+-SL1H57Pp}nHcV8V zir7r~?8B()*&afaP}+xRkp9LZ%e~6l4b}=7ZUmhJ!%2lmo$_2JaSjYerr{XxlVt%L zVKOxqQU5xAeKU2rGyt&D47U52dCxQ68RqYKy5xHv@{>7St*S+B2&_!4Iu&u4w5XAI z>e(HGHB>qZRwlq+Qomg6Eoog0PD@M?#%Pa=rkB`g`bhp-W}tfQE&N)B^cotxEHn*k z1vvZ~*$&QI5ydoABEpguHHJ;s@EDt+($%o}&n;{g zTqUZR98D$$r*)c=unioIL8GlK_E~fqatiQyCmgL2s_J`~!0FOsfTRzZc-Y_B&g0>G zxp-)?Hyt&Z8j%f}Av*_%DkwDBb{{TiiYQ4_GwV3Rb#0D;87f^3n7?QNv!Emi=MiJl zA|uW60Jxxk%s?&)AVM(eX74=p|T(#iKi3elK&FiLpfA? zYODQ#5UQy(-lg&^JPqvx_)%l=l2ry}Z+ z7B!aCx-)>G8Y=BWl?^Ta){(NhRJAUq;H0`#qqB+;%M*}RX6Ea^GxQ2!T;LsRw znJ(3-ON0TEin~-<6W3UdhO`_YdDS4xWxXOUk~VM*x$YETtoo4O8c-a7jUgE)oH;-qWBH$L}A2ExE`}}$^ znlBz~pf!vX;7x1XLUP@X=~A7#oERXf7_8!URyaq?f`x@^NY4RcSBe?T}HB{P%D^tn}x>TnHCyDA(jm#HD?X)ZGM2(C9a2r>eY0F{3Yq%-Ew;V{U5y@Pa z>eOY&07yeV*oG7xAh2UvPI?;OY*{PfC270HVCz~Q!!}gf zhi$niYw?n%JTD2(7{=;MzeqQqa2u&XyN5QJ;IjcrSsfDrI{PEQOVTpd$x9;5UgIXQ z7FW3QL?h41I@q^1Ok|yjcuD!}!%bq(_7I|k(mq7vLB_9eZC;XA$ZS7Ryd-8)AyQ}C zT_N{~c{;#qT-j8>8g>d;_SYCh9miA*k#JToFGO#0 zK2tEQ7_a=fN(p;V!Pqn?qpaUcENl&p1^88Cu;rqwOQZpkrUy-{w@?Zr7PKLs2Z$^k z4S|-6t|AiC%*;Lpt!sM>+E8g9wCT~cATfzD;bJ~9TKO-)r3TPqWPnx91}sbsxdnJq zV@WXc0%zvZYk;Kb(bZ`wltPCEYRK~eLa(!K6_J~?ZDVP$YkdsVP-!2ie(V%P7(2Qa zJ&!jYU_cOLgW_ZGfcQerk!`V4|a`)NZgA z_5xI{B9fC9H3nJN@))wA(mrIFxWad-&I*zf)ulQp{cY9RNWL44yrDElm~DP#0c!Xx z!0j6YD3|`aj2a-RG$87?;kRUqT03^QAvyQ)_pGy!*aJjfXH*rjnY5@eh`P4N5Dk^~ zAsVY$37lG8spcW9pa=Xwh)bqD{WT3?*g(YnUv+ryLlpF)LO~m+H*r z(nuub!mGGTbrAN*rfRbyav?T*13D(!=niKoP8Rjn@7 zS;1kVx>UgwVT|&pDG3`P-DRQ_Zw5#z?ouTTTOTc0%X$HF zT@hhPiyFhGYj}*!P-!2VL3zqLtktDDE4WHjmnxb}49;oJ=j#f)RM8j-P%KFSo*8lq z@cG6vU8YM_%P4q1Bkoeoja~40oj1BH)qpwVSBrT8fGMIR<+F}AcwL)gV1`Qjz~pck z9&BcHsm=;Y64j+j%p=C+tR5&8$QJgx5<;r)ZmD-%eQAF66=^Xen!%UF!_a+>R>6_xBapC>Dlqg{8Rl- zeC|6^%Xi8s?&FfFJW;diueOwW;`H?3#iNC%=Hv#dTa}#D4r3xBa2041wyTnZ0^~>4 z=W%KXC?IY&meuOEH|#C-@{h!wt;1Mnt70)5GID^Cs}`R>&tcZ#Az{oKKVUHH&H{tk zP;no#<&s?8`&PiU;383NtI;WApxa}iE9_g13IJftWfwexBg~}65d7mDf?aY9z=61P zl{Iq75|xE;$khR&uNA@$LkUCJI06G;xeSK;3GiZJ zFKmO23(bSrFjatiHCDd<2nS-9ECWP!)*Ei)S4@>4Y+IDPdoRA6sNxn?Xy97Q-H^otRNTicQ}3aM)<3ktTkw~t zj@HQZGVmQrTUsMS0FcJj<{EI=tt!K00WP-2*R0Ndt@;5D#V(l!h{}aqF^$QdAEv35 z&6`6?50JPje%aF3Au?h7n)zq&>sp_|Z>YGB-*Tnzui4P$$xLwG7|@Q`Luc5;SlAQ` zrJ`}p$6G^a0nXG|y!{~#sV=7mh$^Njk@l-K)`r|3Aahl$va!}-G-0e74`8tB+MmH{ zsCX2s(I}*T#|A9VYGS6Afh_;II3cy48Jfmrlr?pUN3LPF0N1)0$aM)eKvXWqrlZ>G zO|%PWzPb&GK0xNG$YrCfLvO;!H6Fkq*R?-`+)!~Jx#g^YjP@E_#Vt|}t<`~dU z0IFCqnS@qKLu91o@oE?^z@Zw$E4v@%l5T*gN!v_ebE1t(mRZ5(#UcF%NL>}NY^-%i zPZ+Vr2N=YyS+Pf%pDK6e>Aq8})^mw~8Z=JQRmiPkD zrh3lVO%SoU4`9T4Gk_6GDDERR9&-ZN+VNH^WzK{V#anxFF>rhQr{z*Zust|{Zqyt+ zehqyF1cb)&Wo{3-=Ms2;s3rD^i592&qOF#;A?$L|wkM-cX<=6#0+7V!S)ajfsJM?^ zCe9FjcJ>~!*2ln>YKR@ehFHjW?D4cI@isAXrt;mgJW>ry2DpD?NaZr)p3Bn#qL#>v z3FFopXG6LWkXc$rqcdT+y0$0dNonIovFh5N!D^_ukJVUUApmRdA#0@!VCnQ2PAwB< z`NN#V=`kESF~;7@BiHa_fNM4Oo>n*D;Z@#c=m1g0J+DK$=hb;{oLtZLLXr@Wcb)ZY zPsWbY!bVZ++MYpesJM?>Cc+XOvbqZ|Cn#%u3}C4iSumpvSgjA!2^(U;7yy;5s^K0q zk5R*g0glvIbk!%qV2rwK93U$9o=4n4OUy(_Pi9~8^g>2Zg1`a*D>e<5QK7W3QINXE zXFwV%UJa!BknjSe1q+Jml||FWpp^eiov?ovjhld69;Sx!0$iyvn0_w@Q0Ho z=C&=k!!?D`kn;m%6`+s}<}5;GWT&2)d1jNzwL62*Q1NOA)yaK<(1PqlHO&&U$$*pp zBAu`WmKd^jTpD@{aG=I;$#u>4T!IY{Rope}nYGrkG-UVyX{#cWElXw8CT-Se>~t;9 zATw0lN2b5I%MpB5pFOp^W_ty-iRzjqCzS!{I6gf4Mn~9JONvE&^m$*-&vGvuvg2 z1A6hr$Re5+q$a9sHafQqbonEkgk7^y0ceo2>}4LohTH<2tFcJSbmX|XTDn;RL+Nedf=usaP5ghR!vLD=q^?G*$ks%thrxeSJ<3Gia$v9N14 zE;J8f!*Kz~Xbi+GYrm{FYNl_W@GVkyV}uum!D*=hQPkz!bFZ z3^4c&6|aV0yKA;r@SG^Z1Up>}bY}&vNyv_1N2+HbU{TwTaGlneFFT&8xUN|(q=4%r zMyYm-t1YQ!EdOzk6*vK!>^7|3;5H$c>MF`kpkkX+!2~kE<;hS9`&7Z;0J>3g@c1?47~o!OEMuM6rr1o` zrRM-qIsB%_VygjCj^20(WCmt1DJU0+MhvhsJM?_F4BZavfVY?FX&OC zcw5XI16!UZl^So0p#f+)EAV(Vlo{Yq7bD)fq#Yn?dQ5eC8*NmkXUHLY2uNKOv247R z(Wd&w2N=YF6dMA9x~Ahx9EgsAe(R^6#rpL@g6dem5;QBB?FwPF{HAYvCGo|qL#>h2}{+w zhaA#{fXr2~%Enn4PpWS`fQhrN{TZx=iu+iN1r`FZ&K`0>mZH;RIJFF9r!3`3oF2oW z6JspjHQP7*7~oosA(!i#)j|r`7UHhi(K@?kwXDK7KUGo7Mpzj;DxZA@wXW?M)P{=t zsAVE7L9np9X8Q#@it3sLGs=K<))h9(f=2@|vKpqo=NUv8QH^1g>zdU<3NR|}j3vxo zBL!ItD+DRGQ6>N|+eI*tQK9nb#~}5Lk3mW(UJayn*R0mZyyT&}X3>;lln%C~UzSDB zZWN}5@^apxlfsERe&pi`6UCT4b3>Ej0$&{s%Z_R3@3_Pi>S#nYtfc6W33L9X_5dcKv zx`z2Kk5fZt0XS{!^(xmjJ8v%p|BRNTib zTeA7@bDXZ(K|yMwx@Mzu%RtvU)JfPP8yyY6m{q?lf(^L^z@o82mFt=vxbzwzs<>;G zwR@cZQdw;-5Fr;8!GPO zmyJO_xRu>CJ1BTg6k&p$R|dMH1W>VWe}o(hR$d;phUWsD%3=PjF_UJyW(O|Q28b$d zhK1IuwcHI!KS1uPahEN3Wo##H;3#h0Ibd)bD(>T!DFG$lnjI8uC#q{UGQA9ZdA^gd zYc?`GSRtY&;ZbZDFTlkbi@hA3vmC&kjgQh1}@_U zh$?10k%q4|)`r9%Aam6mnT@qF#*;Q~6sxZN8LWni`&f+zmarkfW;_je#uGEO3}n%{ zP;prq*hxEa8IgJ9UKGX?i(Hp=15l8T+@im-rq{J0`3K1C6e8te%o8ePtfztT0OqBT zYkvm0q2fMr{Q{QXowB=l1_kShqA3OlC}x!bZ7tuvV}?$M!Yp2g7lr-A?MG=L1+QgcDAJa}3k53gH?gE0*}{=LYHFWtxMVg zzMmd%o!-WJ5F4_GfYenH%f?$7ZE9$IfQh&63^0fd755SAgR*4T?69CsiQ=t2xeRdm zui&M|O?z+v-Dp7P@oVTaz`Yuax7;4`&?WEyQA_M06J1mFeQQ=rE6kOpO9lFVCAWt> zl+mZOu+eDiTA#sgsJM?^Ce9E|c=jH$*2hF!s%v(PI-wcJ@+U%x_m`0~)%>m5p<&4Y z_iqfT+)k7hQc&5J$czc&*13nQB|uJIX=hKg4Msoga@ELc!f*DRV+2Bkdx zN%*ZzW<9q#GcrxNA1IZQUeeEvFEnRg+&fjLOJPTG%K;UAr>~ z4Hd73P`!zkcg+q9vJ=%cOUxz%&V+-VF0s^NVec&=B_BQwy#+XXW4PqHW`{1p28b%| zn)Pf{YgrmHe1No7k;!iV%BW4+tWjjTmS>O|D()k*%=Rzwt=VBgZKAqn$w_4Z${*t- zY@{Uz1)-Cv%2(QccT2tF>Pz#hFJC{_HSq4i(ZSil-r>P#=O??pp&_#XcWMl$T-U4? zQm~06?wXx6(>1H5z6jA&Mrz9afZ0lNXMw?NsJM?=t}F|gi1xQ;hXtvL>Y9ztEdyPi z+9YhGjS2u@%vpquw4;kcaN@dVwU7b?i@Ro7BiC56YO#e7UT4WVlChj}N5ByFrU65k zP~3-b5O!G)w!3DvKIY90)ioQRT#Vt-cFw==3R`63!zCb=NV$0s8;;9)hhdI2F%q$y z#Htmox zF`nR@GN2u>-9$tdn+HZ|yDgN8#yOAFKMIVek;}LNNEOE|F*S-b^fD;3ojh9JhOo-U zS{dU>3me6%Ykvl-q2fMP<3Y!-f|bp98u5%LX2uxEiWyJL<7w>0dDGm2z$4diUH}{# zi?JM6apbaYfT-z!Ecz>*7{(qi{LL+OKia$3@(Pi=*6LQqdeXv1k?Y!@L2jtHk6a&@ zY@BnxH9IO;PZUitKvglT3}{D1+!VI#FyjF9vWZ2Q9P@ZJ^cMgECw#5*JScWQ$|c|c zQIqzgf*!@15F0XqfYenH%f?z61!`n`fZ0!WXMjO$sCX2y(ctTJ&5jBlRBTf!m|O<9 zJQ*rs!zvgYpgFER_gZ&k$T7gZ)(BY#DeSjqwUC0BLh13=YI1beE2ie;Gv$ya1k5WP zE~SrWB%?>=9>Ad2wLgR2P;noN2uQuwcq^k#jf@X4h;?UxL2RhFkJxB{ z310%)-E9IjkosT0J>3g@c6wbG%9uv*(LA*B$U`gCOWF>>zW;f zEFvIrRqV3wZe{c-Eo`*9b*;}}H&oomE)!>nCOl^kxu8$cd&po;8OZXqDPh;_1hwk| zxN;U?kvhI8bSZ8pN()(UW=za#o$qV4yuzI*X;?>@F;~})Wjv|e18|)6?2oZZDDGo5 z7FYaEJxWG5|b6rrx&8H9$4`v~O%Nw~kH zll#1DR_kLBN_EW=v&n#SP&YV*U9*IgJT48r1vqtbB?kq_kE+k( z)R0+#^RF?pbqg!5Yj*6?X@IEWuGwibU9(#1hA_((r7}{J7B-4mcNQ4ThKl=`EmxG< zU9;nY)I@d7M(37+E`O+#uxmCd0Dv)k_HOm^g&XP4G z`vB3`nh?uaPTIy%2)on3KsZ#q8iehx*>OQ|qPk||lgnV3XE_NQY2y;q4&n&Qsj+mK zV@-@*mJL7xw_hV>R1h-&id}wRIHdaksn?2M8P7?ZH;P|(1{nN?iu?HG0!~=a+Fi5b zg6BjLCfK=SpgSp;#X@!jJ5mBfiCru{^ctQEaH_^o%XQ6;U8W5XRopcTZC>kJ*pT!C z`3me6+Ykda4q2fM% z%ay)1<7vz@p5UA^psj`R1P2BP<}wd9-cBNnr^b-VF`l%L0^WqU7l1Twov~KS+YnZ{ zSUZt1o^lVs+zHSA7^{ThK33zg#z(W$HLH~}Z+}4 ztZwaZ-)VhHfwq`c2DCkvbw$*jMKYa=%t9_Y4D)z3^cUbzja9bXe$+(7esYomu3Svo zoC=yC+5IRj@kNNWG76LyHi}qx1{lPKiu;I7h_!YX&!pf%#WtmaIb(oZ%eU`>sJX=E zDvP5(rWG{@k6%NM0r*{G8SBJ0wrh6c(sO{Q9AwjDvDE-+FAzL>LzWPbd9CqQMvv0Q zjiT4JKZD*-aUZ=E;%z~X5+%vS%rUSXRF9vo5zE%LGTM|DHi}qx1{lPKiu;I-2AD8ScH(V8n-axadvY1z z4guKW(uTXf#Tt60cb&Lcbg#1akS7Iwit3u3qUKcwvLopMityn8^Gr^W{>93P zVaWjZZw#sIPL#{j0iu@3ehH-3x`!Ong@DX!jk7YIRPF)H9)3f7{ zBn0HGidr_p%Ggobu+a$X+MYpesJM?>CiXC~m{A6-{0UIPu30cF04k?F zk5R*g0gludM!BxpiOa?TqDl+UuDHFHn1hU~%x^k{j36NJT0ttKLTSrJLFyWx0cogs zHIUj}vy*}aMRm=hnPN~n?F!pm(X$hv$h|k>VQMHZz?B+)dM_aigqfKcZx1W2ct1p)DkC`*$j%Hsy1V%Tmws z7@36PJ~EkFOwu*0l`<>|s%w^5SJ+oeibbytJWdUn1vr0WIOV!# zr!Jiah$`-yoi)=ntEFxTvs_U+m64jXu(4?B%>skjP;no#Tv-;rt+BgiwLS)}RM%{D zZW-wEhdK$nW}^boAm#SG`Qx3YhTH;R(HMfcu30UlAkvDvW?3WG*fp!gz6cPOv7B;8 zV8X6D4Ge@s#j8Qs?wXwz1ShI%Ha@uwhWo{q-vs0er#%*)g}lygE07L#U7w#Rx~Q z89Z_g=LNvQ2?J{kxNH~C)MecOQMpVx9o1HkqiqZRrBF!z0W#Z<;Q&6$%2-d@ zxKZS~_GgeAD()kf4PXIi?XKBr!Fr--iUBT)nPWgZVfS~5jI~InQ}J5Jm4a78e*q5F zSY^xZN4W$XAZpTnRM1!{k8)fW&opEL0jbv-Yh@HDEo>CA?hG)94Hfqhn-FX5E}m(@ zgNkiR1(VAFx0Y|;fi+{5#jT@jc528mz`fQOVjZOL@z$m108!Ips@3->4S;;S4Ov1! z=BnsrU+Bu{QMm^&Z-ZR>Gw2N!_t9G+-WK#IQM@f?m4R*TG^x}mTu962$FqwMo zC{et%Cl>>Ew$0|k2`_yANo}`83QbJ^vP|LQY)BUZGFQbaw}(8F@uakIqgZwA z&tNrF+{bDx&JuvNzcs6sGJvJiV>n|BWcyulJ_{cYKpRM%fZ16-@I7|V6d&Rm8L z5LMhYJ6mVhtd><6VZ|#A4R3b;cqU^~^bHBykZuonTOGAfi7HVRVL_zXxx z#jAnT?wXwyEGVjL7ELLGQvPds!j4)rE{L_@bPufU`FiKDn;hnM<$%qKZ3XJ+szHZPJo9giN+9l~J3tuu){XmS>O|D()kb zsnaA~v$KNQM0L%Qlga>;KgLPe6H5*XAR1MlkDrFj0-V1woc>2G^}eew&9A6U_?xWIYCb6_3_21j^vY;)2*<;|xAI$Utb!qqN z;oj+K`#$y7_FE2*uU>f^I0(&UNY7IZ1qC=rV>tT`4reY228jBA!P&ca-kIa<-q4J- z#=nqt1Ei~pG&cUp=teU$*Ualr*U}8qLdAWgF#!O5eC~g?;a1R%;3P7D<7q|r=YV^M zy086K0Jw_up|4hVJ#q5l$M;^{+CR89KRP`)K3X}4UUXJ5p}lk0<<0g@SJWSx|F!ID zWwQ$kMfIoDrRBr*URX9r=XHn2r>9#x2YaVa?C*Dip}R`m{K+|ey?^1W}u6~#0cBk#T)$6a^(Z2n%5m9$LeWkkPsjs`Sci3M3t7C+? z`vR zo$B%@Pi`Dt-8<6)zyBuL0QvzRIz3xFpr!s?OTB&X>eU-3dskoEIy*kwJKTDvi0fNA zwjZCGnR0O3YTwko`6m6p?mKb^db4_C`<~2!+(+8e*n-oj(~CRZ``YSLdnX5b&mYb| z)t3zOdUe~?6%m9_3rFp4gqJMOof~z@BvEL-;sV_jvN3%=h-&m!Hwe*4e??YaLI1 z4Om#+G^{cjxC%j`#MjE*8mG<_9mn zbauL>Ge=!c?B;)1tR7qDkITj2mgipDyFTw^i|S$p)hm3lR`De_ZW(Lnu=`M1Bdlbo#Rdt( zADr%nEzsizVY z)Kh%0zncMV>$z7B&aS@HzO?*zC*58ZZtdP1m^!`|$(C?_n+mE^leI8i(h?_4|7=P+tY^Zs~iGOMDba%Pp933Am z*Y?%gvk*XM@Vm=clND&4!$wA1Eqk4#CP&+iE)$#T*EUq67Ra5ubeOa9gig)wz1e*> zckd6Wo%Z8N?rjOkIQVaDsD_;&H+iMxpnjWwBDeWd>dE#KNj5i0`~uO8(SL43?XC@Z z%XK@qax=?Urm<2+q=^1oSU;`?=KckW4*uF?7l51bYG{` zmCK8F?2G?b-?FK`{v1t~2hGx{$tu~(_7pB=YGO+Wn)@=Y!I<9dPo3jTSY_vV zIz(sfF~&r+&2Am}XUR$a;SKd#bCZ00G(X)vJa~D&m2GY}8BX46cCp{E)7HD#3m4yb z=jE1uHMsbH``zCr(;2y;{>nKS>`A?IaMq;3lDGYTa8sS0gG;q<{_w%!Vdwn3`Fw`X z{2~@rizVhRr4Ekv=db3klGJx@s2j~qmwxlDX#iV$uO65e_KhTy$Xn6PxU+JlD~G&1HPT}al!%_y25_8y>ahG1WOw>Ek}L-hWc!C?opH#yS+bp<=9Vd zs?VH{t(=1F8(Zfigk6)Jqt8U#yG2%Q>f6q#7SMj~ZDv+uRn%6P(OgNZd@2@}FVA!oK=_Vmp0Y8Y%AomVbx~mQX=MpG*F$L>r8*{X7x=Cu2&z@q}TZ;kF$S0 zH$VHt^#wI_!GgUP{iSnGOM0jAtgQQRuyfT_@4s@lL$h5}l^*jp+fCWf+u3)V&d+v_ zbCcfE-rRfOdHwslvh7yQjnzA^`v+I&rh@vDa3c z>hqgxd^x#HmaOpH%v-Al!Z3W>YZe97eKk+Cj?`iMYm*W!O~}cX;}t(g#@tc=WJ9g* zjo8QMCoj(1w`Bk7tov2-Zm-ka;!sRoUVOu8R^PLsK2&kpaxAF|ylQCI`zBJ8P! zy_S0v*i3)9rAB2(w)d91Q?pUNwex`P%DZXjeJe`K+pq2&z5VF;?CsCb-=6d9-hOuc z_T?*w2kkxdThThG{@s?EB*~{2wOf{(o%b0D-?a0FOu{=?51h6&5M6_o+=%LZ67f{r+$7zJzGU0{Zi`e1z?qZ)Oa)V$_YPd5MArE1CMaz)Kr`*i|ZzSbmG=ViuK z%cGVHeSMGW)%U1==c83Jg;gY5+Dh-0)VIFW_oz{wD)#G>Yf`6NgRVXh;q=C}Yx?Mo z&*~E}?%nLR7k%${sN4Iu-FoY-n~VQbO1Ih6F-6Py&bh(0B-))ivj2LMKe8UcHRI22 zs5-%vo44XC{yOct*{%KUy?}ZBf7WVPeOmr%ZLh{@uh*{Uz;*r~Yc;Dr(|l#t^e*n2 z*MO#4qV?;JFRv0s!+Kk9?iL6AlwMx#Z02-%#K8-lCo5Il%SYC`yi#JKs$Tsk8|sNF zHg?{1)IM9Cn)h1@LknJU{CG=s8Pfz!23fB`XLX89XQz^1EceKgj^p`TQnBsVF|~*b zwjkY48?TN^``lyp>x-89>6Uu8#MmqPeTj+b$gW+!^I0)JTHS!aE;>z4eZA^nE!vwO zzj3%cX{jSp*0^g+HuzkeX(BO_PIVg7CFWIEf2^cu6;N4R6>_Eqmo>q^U` zi!et@=|hwX1)jTct-Y!L(P0F)&T3mupH9|ap3krEE_zC{BLMa3e&N}Mz)#gzsk?If zWzQ?#1y)w=)H0gZNGEe|>eDO> z`|0yK53b#~w)L8M7YNSav$ove`F(0`x!UT%?ORtrbczECGby$f=g4&kL7>)V71M`i z-kF;@duQ9e{Go%-JbV1?-jOxR@}$dbl)?j7!ombnW17w9Q|mk$-UmwplcrPt_$eceq!X}(5t%(dHe%KEm1qH<&eJNDvY&azwyf+H>?YZ08f>~STXCA-IQ7Rge#FuvxA z*oVksrI%u{9#uHY(j-eX=BLjtji3E# z4&(wleVS!nnX{(&Q_0K`LK%z49D=z#2Q-(YL#c-ek9M9A5{;xuN2v`Fh=GJ4g)8>O|8$?)Cs-(nf7WSZ75Nu7RAZXw4qPfUS@FD17vAiCS?3{uAo4_9LaOZzmq=f89`Dy%tn0*WTQ!>tBvK>mS0*0hh`AN=yC8wHS1j zE7&>CdInm~oFx1Cr8VnnGt<}UrNz90j6~_r1^fT%iIi$qrT48qV6KI z5S+64=})g^bOWIpPT2guO?7=OGBOr&4s-eXrbo9=aXA;~AkSxs9qOWfXG6VmPNvF( zU#XO0l^kVTV6~(ac_v#maj+#5aJl~erh4t1EN_*3=Qa7AwZ;pxRBv`0p49KtXXZ6C zsR5ZKd;i^8cW9{Vw9M}~2SeG^hqW6j`C!4nx~YyEcA*tT=X^Z3)A+DU=1IsGikv{5 zFDmLUY^bkmW{R$cr(&)f;mMBs?Ki8#Caq$fhS_+o*=ZW0o#q`b_xHsQ zX50atez|U>INe@JpZ?V+BA+yGBd1XMQ_YWWmrCd4B5B3a@9HDI&&^zUxs*;BsJC9C znhMHT$85A_rJSH(`Um<05M`dP-vsHBEPuy~-`X&H3i)}GR>^dJF3?ob)bz~$>rM3y z&CD!1t&%y%@*EcEPNR90ej%QA0%f&>FV&AVmLT|N@vY9T{9NmNold_)@aamCt>;bk ztsCmaW~9!~tXy{*6&4&`0|GFL`gQ#Q&PHruUgbAs`A$Pwk&lZQHD*)&cbjW^eYbq8 zW<}ZczvpQtnyT(u{zCd|)l2OcO23d6=P?WFi#A*K%~*yw@G~3gtJVS=r?YUKFJJZf z3e5n&sC=Wu1GQ+(e7u5P9qX-ef;vxC5YoVO`T9F@Buqj$U%#S7&ED z8e+Kgpvk21&B;i__&Ezvc*dmuNb93zoKW{rpz|dwuR)#4PsoAKM4=zihXtl;Y={aL zFF5&G%N0#R=lP-P65EWOit`OYzj~jo7Hv0=rCK=T-IOx0m$EE-<3SpqJB#|3mU=i< zVC(G2CrYtnY$O@KQvb__`m$6>^eY>DW@ITVQlfPpN!BhqE7EvJ_qjYPeMUXqeyVg< zN(qO2QvG^MmC8EohcAEj&WNl$$7??yQlFk!i^?X^&29ZL?w#kwl~XlJJfGXV|p(zM?g@2?xnr!^G<$dN9WOG4_NLeG3U}59a{bicJ%`i zf$i51mBpkE9rMmoM82&4%a(e3%)Be|ws0Cc@#`|}6OU8O4X;ksn`p`HGZmxC<pp%=GF6^ zpKutsLTC9KOM+Z&B)|IV)$^%S4LzstIJPbh=72h>KhsigTllLs#_|}+;Gi|))i!(_ zFp=JBzO;MwaPRcA{YG_9`|kYpf*ExIn`}8Szi^JJBTk@;J#evFT-C2$_70BbC+&@P z#vxlCl(uojA(TN~l>YMCJ&tLXD9(?na|G^0tQg|jI>RukO}PW>t}RbvG`7r8Xl3h; z+=cr^g#aPOcA+Ken|7gWRNIBJUpOvQ-5nMas2#{T2az$^QYOoVmZkwNl>ISxq2;B4 z6}zRpAl2cPf)_5d6vjLC(6~_cN0AFHuS72NadrC&X^YRun-wRdg0}1z9r*mWugM>` z_`qm8I}@>EkFaR?c<$utC)*p_n~S30lP5QhuIeLWvv3Q9T-@`a)3e1rE%k|fxpH#9 zyG`zQdFPToL3_3C?4P}MT|Zh6MVC(cGWF6j=4s64Y2R%bZRO*7ab5;G{kLVVm*)dW zQZDIJ-@eLF^yBih#LiI6+*;|VhLisc< zX11d~o!h8x>9Wz(DrL|H$nzKDom3@vlQ6XGE<<|p*dk@f4${?s$i`nLP+87UT)XvJ zfjHm+>QChU=_gU<`O9dZ9U7e_4aM{0PuIt4ucpjOWa_ZX{xU-=WU-SeGp6Mx9rHU& zj=1=H`6yIMWvrbKv&`9SnV_#3WU7xU_2S9#4SkaR{QBwYE2j4pN1o^7c3!{C{EJ`Q zo&7?Omj6nBl;ACMJ&n7kXY=b@L0B+$cUH^W+Bf|%eRBVo=6P3J-POJ`%e(SN*=y-- zNG0`cEp=OdixX3=oEhFuMNM8PHd3pyl@z*yUy7q& z1slnBd)rzv^XqNrKYI3G`u2Ye?Ee_r|1q-vV{HG&#Qu+|{U0++8hcDfw4U=9TbiCN zP0yC5XG_zwrRmwy^lWK*wlwxIecLvDTbjOenJrDGR+tT!HY5KM__LB#; zGy_|jfi2CzxyP1fU`sQwr5V`L3~Xrzwlwy`hPE_ATbiLQ&Cr%+=sdudW@t+@w51u^ z(hO~B?E6QyG$UJ@kuA;0mS$v2Gjg70OEa>i8QIc|Y-#LU#v!$6$i=ZW?6)cSc=h+sWZuPXVldQmBUwnWfNO7VM3 z1ihdXzqdrt%SrKjO9Z``#qTW<^wLrM-V#AC9L4V~5%jW=Tfw?qFBk_?EQMY_iJ;ep;`f#adSxhnZ;7DSh2r=2Se6LZ0-@K0{Bpe{6hB)c z=mnwpy(NNP4vODfBIw1S_`M~P~fOk^1SHG)w-}yVdK8ywm))bIYd}7j@p-wbU>&+%u>8 z;n~pEnC&-`zwX@kyzVU8nqPZEN1blHa(uGCTl}|OM8d7~Za=R>->=@=es@-f zMjotwrlrOyn&^oDy1d69^6dEK`O)rTch1&mnhX1&`e3_D3QONhr03^^soh z^7XoYe5(EI_{i33oNM(#)os6@)QY-Kq?7i#w7D7}o!;i_^s&9y<|oe{pIqB|etx#s z?d9scPrax8E>a!u7Lg`@H&v4%(j-7oA=Zm~`orWRaQOF#botd(T}DWk?d7Z-T|M5P zpKKj2ret?}YwzrAezfqnt<(A83tL}zV}A15*30vgqxoU?$$Xw>YSJEA^Hky?B5n2g z+G##GM%q3o*LD#;x=-Y~PSmJ9D5-1qi1lhZsif)4^L$oQHMJfg()6FE&gm3s`a#K@ zF2u~Hg;jCtJVd1JFSpbrEv|1Pbwxm3N?Hl1 z)+0o!>PE*zRbdX(kNEF0o!#1>ADzz4+)8BaK9#%rLu%HZNL-y5YjO5}qRKG6i8-?) zMmE0*K_`5O#A@m{lhuWp&rmkQMMjY2Lr>?^_lSC^Ju5MN-h)JH|8}z4FijejRePB{ zQ6Gp(dy+`;^u-J2RP^F?aCA05xi;TF(5wU(&EnXLA0blp|4E)>m{8HdmylDO3w6a2 z8jPa!w*nIvESOEbpL9THg;{0C59O%ssjc<{7`4>RqFMaElGTGL)O+LV$>(pDTVP8) z(7s=2fyMLHFD9!7^QJDRnzv9sl1n~Nz4lgAa_@H0?0q3s#XiiH?w6^!TwNc{bsVaG zd%J{=%lk!2CZ`O2m?cpuLyxUfC6Y>uS#~6q976gqL!v?m5F$gSFobZAkl3q&sxUhe zsamqklosxRJ&JKua`L40EHHGdn70xJ_{8@F5~~ zlWWyJ%$KNIm6#eY^%2&p@dt_2POeo4Fo&XQRaV+kwW{DrBE^#zuK~=f=*5dDtycAo zc=5uI5UH9ReFre7qN6V%XSwK$9w1UPxmF#(e2S`7dt1&5Q?6E}Zx*SST&oUX4n@_f zo`l)_0Umu8RR=I{qH5J#4sXR3)~e`sTNTJWNUl`}Fjt~#RYJy6wJN+{ zq-1i+Fo0PSl`{0%T2+`b@DCBGn;t@7hD3!BAZdn7VF=+Kv1V0ab|g}@s38%uq~&Q~ z0_)MX3DRrTA6u2qLHpQ38jewA94zFDMRa;-XoITTf^dJ<;yCy`HD zxDwA#uT@7dZ=!0|d}wdQ71pZgc8PUVa;-Xoxe`^Y5;Dr)i*|)CU(q1x`R@p3NmR-( zU~5&hu_8<*_=nqbogP9)Fhin32#_d4rZE5I9$~8lE1Tq0e+085k*ZcfnX0Pw2$8DE z(Q*VcA}U%2Q?ih=Oti!g5UH7*-j87ZBf_dvX4Z{W=i#;)Dqt1nKq3dW3e8jpwjN=t z1hJZ2tBznEMAfRc|2Z|+D!UcdsIZInG?dVusn?qV4EYoRvp6}imFvvsY=zVf+yJuLPRG=-!aUq=*5euzsl9BgLA7@ z$1ta&qc5Rqx#)`?AezzST6GNbDXLZ-RH;?zn?>p+*Q#TfLs7M=Ct)^!!uO}tts>Qu zIlyC>H&L}}E}ysJ3Tst#yGX_4T6GL_C8|~>WGwYw72YpWGCgI0SrU~p4B1*$m@@DW z5viLTLMAXnqCyCeG(#pz7KC{*_h{P$>8bt%W=A4btwJ+ZCF>EkN-!}@S&19@jVE!Y*s#8v;V%2$wNZoX-!W>BCz*eD|>cG|`M5-p&suP$8QMIb= ze@@L+u2zjbMx=7OGsD~{=FE=DWolKZaP|<9y2-Wb1m;Urtx8DhWlkvk2y4~&0Jb^O zYt;$Np{QDwm8w*&DtMBuAPhKw(l<*SK_%C!QzBSfku*Q!&P5mC`Hn3A%!Dt>@S&2+57{6~aU zr_2=As)4G`LqzH(V|4~|Adv%Gg=VS)TaOT_np~^SU>-!(s`jKfgh%E`_= zgSk=6nH`n0wK&0dazttvZ7_6jiHwl9Z`csatIoAQqEr)fvp2s9H6b&s%YYwJN$@q@q4s zbeeb$B}|s6VwDiGRIv*07isw~(zS$n5|uQJ>!?=whluq3m2`b!l0-!jAZnR%m3zdx zb%hy{NLQ=a6h_8CSL+cXUH@77yu!4I3Yny!WkM!?fJn~@c!gPz2(M1DDNOPMU7d%B z^sRtbmEeb{=0NY{Ur?!z!2qH0!qTAbReT+JGLj7aAS{tT0(m_Iu@XV{g9 zDm~o~nq@JP`mOXihM5ypwGvWzIh62CYWzVGLD!H8mRg=p~#h94o)H9ZRVV0J}EVM5SyQ5Zcyr0374do#?esLFM`j>?sOsI9m9 zhw1vkY>FyfJyGOqvG83i@qG19(shFw6ji(CV|(kauy#ea+w%m=^Yq%a2eT%sb|r)? zRlCCbMOvok4n3GJQMto}tzFSRk}yZ+A8y-0Rfr@oSE3>b5Gg~aFp_YOuyukSTtQcu zCy8{m3eQwstw)G-O^=%R7;h?S29vT7v`o~*4-n~@j@NQ0c?V)TWoIfborj3@P50q) zM|lH%E8(@d+O-ceBC2+^b#>~va8T$%$%s& zm6#kahe9z|SkuM_u;;%*?b?Ug6ji&jVwI|01y8ayTtgJbr+(8>m?+1})vgojRpjas zB3&y)VVGUfQJBy*Lo)Je;mcUk0>KiY(rZ_I&No%NPRQC-G8J9nD_C-b3biXf(3`4V zz0os63(HdKR*7LM)UNoOR<}K3IDm+tlwH_hT zH9czLW4)=U8B9vq+7&-Qq-O=ZmfC9y#B|C|VH(drM5J#8ABNeG$cL@MGu4N!M~HMy zfA2bgdregBYEO$(U6rd{V~-K(obJ!~kndRhadggB{ky)frg=Hyyvq-=6_pbPWZ>o0n zM9d}t`Kqv{r5 z_*`3iRUvnP`4W{o%+^`E@(*|PP0xk#Vct|E0U~AS6h;j05w=dygDdD-imem~Y89TT zx>}DA>6#uj@iE_2)FgE+6E*PzM0%#2JzLyi>)hawweb{=0 zNZ0h*6`%D@)vormI3-xQ+BNnVk6M4M5J$e?TSzTrfOG0 zR4<1T)~@jfiS$mdT}N;aimF{%aZ9~-6+B6#d4&}WW?1wJM$~gFM2lB2{0Nb*=}{OT z2Tn&}LeO$i7(GCwXL{|5&jF`u*IAX?mA+Y|UwZ9|5B#QTS5L%j28A^$b*o6Xbj~n7 z)0?VYa{;||S6I8E+eJF2*RJ?5Z>n}BgeZQ*-5nYLE-nXIk#40{e?wf;-#a|<)12SjtbkTjc>$McyTt6{q@ zpV|BgB26pc7G^{;+&V=p+Fwh?t@Cu-5*2WZ4*{2XvQ>j7d$RQmTPx_v|2o-|@j2jw zm?t|$-74+>EKjN=fvNY z7+z8(eAOs;l4!2KnC{_lS86Qu@IrI`HNi8v^FO8@X+KoP`TZemn^alLAafk7Wkij( zLbiM@!_Tm_f@@iNRGz?$3q)l?(vndbJwc@DKTh{)m~DN4_GxD+S3rMb&(I4W-v$VL1oy7tLY% z4pX_KzbPq~>#dL-fgd2!Gd-KYXMdB~M31d)rHKOn6p_a1)gC?$oQx|#)(mOVxWYYS z-P*z)JDIjtU5YB8WGh(DwymJjqbNQsoQ$HuyyTUsw6?`hur-4DLIvEy>_~=Nr}|9B zt@9L-#uac2GbEWOTXku&$6C)2X`5c#PGOcr*0#1Ere-`{X#pG=tinz!K0-X|)sEKL z>YU=$;ZsEOoc_LY3bQG)wk1qolwZ%!&-SD=+FtxgBF)olTYRuMS=+MmE^{qu4J&w* zNcZ%$44*U()-s}2TUA2UqcVQBZG{R^86PqZqB0>-hG=AlWw#Da*leKNn z)JFUBqtp3T$=a5_TclxnZHtc*Cu>_z#%va)wJmk8NW1j+nfSbLvbN17^wwaBlwD~7 z3T; zXQ$nFYn!9xh3-pM7FFTC)`IcaxSgO$SF6s`YY-7ggXGY&GkOS?&;b5diP6nW46^@?K?AL z7Js!Yc^h0ef9})j-I1>yoSyAneYt(}m+OCadXZ~CwRiH&=`%+MX9s(S2cMnqf8u&) z7`f${Q+?CH-t&iZ>%Pu7<*wac;(h(Z`v!^k4HNGhCEhnqyl;|t-!$>QS!a?q@7t+2 z_EK-`r`|Y7y>Xa&<0$pUaq5ke)ElR%H>P?{Kh<;ksh-nM^_+gH=k!xOr=RLM{Z!BC zr+Q95)pPo(o-;`GoI$GR3{pL3km@;uRL>ctdd?u#a|WrNGf4HEL8|8rQ$1&x>N&$y z&l#qA&M?(;hN+%2O!b^$s^<(-J!hEeIipn18Krv8DAjXDsh%@R^_)?v=ZsQ4XO!wW zqg2lsrFzad)pN$Fo-j8i>loa#B_RL>cwdd@i2bH=HjGfDNFNvh{eQaxvq z>N%5C&zYop&Lq`yCaIn?N%fpbs^?5oJ!hKgInz|nnWlQqG}Uvash%@U^_*#{=S)*Q zXPW9cvsBNSrFzaR)pKU4o-<4JoLQ>p%u+pPmg+gPRL_~sI>{g5O}?|2d}lxT&O!2> z!{j?h$#;&E@0=vxIZeJZ*@t@Dsblt%eW;i0L%n1l>LvS7FWHBB$v)Ie_Mu*~5A~9L zNN+4Bu3&m+IrYwDAJTivsbePlkltNR9W&X7^!{?{n8`k*cbF5$)LYD{cP9Ig-egW4 zGuemqHgoEj$v&hvnp4M2_94C1oH}N*59!V3#4+`5bLySRKBV`XQ^!p9A-&_AI%cvD z={@JvF_V2r?>Z-rskfa|?@aa~z44qnX0i|It>@G+lYK~UKBtbE>_d9{Id#lrAJQAp ziDT*==+rxteMs*?r;eHILwXlFbL>X^wsr1z#% z$4vGiy*r&arrw@Ty))T|^age6n8`k*x2RLcO!gtYNu4@ovJdHP>eMk)eMq;RB#xMgTAz|mfjm_!pcV zzdS$MJ$vo?+`jBx{$j2xW&lzdCrnIi^G=|MJj^WplAip5mB`9qNk!wmeJEcX}wJQaSfzR-(?NCq^lz zZ*kPY4s}8Td)IQ(=_w>S##bhzr-ZmZQ-wO1J3j-+5dJPhMNKrjy7u=n?W2=z2Cbsdm`oi`E{O2DDNk(#oj+C?fu}6Omxco zgX<&SPhOuro$~(R>ZtdFJF*cd@3*g)dOy4;8-?HXyOnJSd` z55(T@U702_%KL|6@At0Eo=ACrew}9$%KM3HvG7u=kHj zdOvx6rV8c#BeC~;S7uM7yg$FrGYRGW#I@M_$ECd=+>wb+d4F(y#QVwXv!_$uA6y;v zesD)N0_FYo^-}ML_hh3`-cQ`&O*Q5H>|NOV$0fa=ygpNf^8T^d`@Ji(CsN*@U+0;G z@_yo4?ERC{-Vg4`M5nwzxIW_jx?(n9X z@_zO%?ERCH-cMeisX}@GMC|?EmDv+1@6WIEOhS1-aV_@#X=(2VcVwbd-XB~a@qY69 z?CF&E2UkbEAKa0RKzYA?z0~{RJ=rLf_Y-$`Q%!k4dl&ZpX-V%Vug_GWynib8e(%cc ziIn%}*LfzPyq~xhd%ykNTG@U-xFZvt^8VoZi1(A%XHTcRKe#&T{osyl1j_sE>!scg z@5x4?yq~zkn`+AY*}Jg!&q{hfd3~k|<^A~=IHNnB-j&%CDeup(^Gx#pv-d7QmLAu6 z*xs4hdGx;jT>!(M1OYBZ61)U3dmr<-1AqVqAVH7>2w_)XB1*hGvZ{l0)s5+(#*!b+XgJQ$wgGdP3kba)^y`ZFDB>^WK- z-q|hLeqtSvYuxlEgV-9RgJK6Q@SR`1mP7%qlx;LsV`BKmz7(Wx!?9~>tg%BH^y@>W zG!R`B`RI~c9emP&c2MJ^ptG9$rGe<8=0K%E?@$7E3LP(DrA}%dR2qHntX~?8rfN)78uEt5!l2TCG)|U-tOveBRNCG#s#x#nm7;fwBFCyA zv-WRQId(@YES&y5`FU1=HtCi9EBT#P^i?bCy5vqj`<90A)oZ;fi=ab!-dlso&9%!i zfsM%jdLT>xI=yppc|KZ*=9U}VbC3Y1ku%?nt!%huvAC@oi|=fRr#;DVC2A(q@n~D;J%-ob2&sxy9bv!XhX;Gb zSPrm??`nvhDqH=^_;7D+Jlwmzs}eUy2Yax|lbHdrEZ*W9=ftI|)Im*apfL^ya6E$r zF*l$3h{GANa&tI5>hFziK@RdG$7DY8mC;@qB;wl|qElrVlo*k^tG?2&rM#df{5@6s z-`x<;RavakiIfpXE*|$_Bu;^2Vm9Pas-JWAIM4HcJ};iAv3iE`4?%4W)tl{JJPQEK zVi4;PFje!aWm+wYh3&aW>A?!e1$sD`XGo)o9T$Pxg+k!ltzFsy+NErpIpw3C_@=~> z+65Qi@}c+^kVoxpP36}R^|VX7zIJKX(Jtxk)lm#N0A?|WrCr`G6SL7I^_cjHMzsTF z_wZnMFkMp-o(Tv_@`aMhR-kEsr4PR5OAo%_hcAO|-EfN37aGF%`HGSclQN)Ql~JCp z`MPXD%xxcQMw#Qdvam2RXIvd+qt#jEV z^Ve{s)*k_Z+;3+@1Gm88A(B54LDBjRSVqK`!OAlpmzJKRJ9>5V_%Kocx=>Bd)tXIK(PG7P9{1=I1=2=J*t;;?lx$^L+4mcniB47gO7p!I_2Y#bE8`5 z>xID-mj5+)rXpVvaOj{_s)t!+8TOEoJdbb4KLndObAd#&lao{^v;f%4rJ{Lvy&tm<_I=Ns!(-59-(7}Z0x9g(& z&>zRX$MMZ}{yD!##QUSKQlIlP-?$qO>h(L02Fc4Wct1l(@d9Qw!42;ve{(D6>2dLB z)I@ngYXI;#`}uIr$d8nAiVrk5a=soF&FB%7uOqu?0go%do~IrOxnJbxOg+^Snys9n z4~X|g??o9Z-?H*}`Fr8Kke@xZB&i@w-RQC;$?T(rENDsky>M>G52Rv=ZF)+Jm5b4Q zF5j!3*o@Y-y*k)E+=Gu=?+@+{$9?x8EPj7GoL}-ot5}*KL8`p@Z$6uIyD8SAPDO6j z5cir_@?)(mUW*)Ul@O@>&{%#qaNUof7ioXkDU?dKQQiTG`STjPzn2@dq`SCG8q430OyQDS0bN zk76B!&5z`Oe?s)4O%C9Zxy1ilgK5&_@j<#oh}z?TBgh_v1?ex=V3@Rquq13`mkS#~ zCRsLL$hYt*@kG?Cu!Y7zEXMy;4aP~+3ClEeGM*SzAImxauy``s4&po+iN$&yHAdQ5 zxEez;daFn^+c+WT5&yKtK9fckei$P}C~0ON#?%0;UHz3BjFJ`>_6EVorcU?si5$Z% zu@*f>W4NTKDu6-!VhsjKvx-?Ltl7)iTN95(A7Zk{KCs0OJVOl^4&9(=2JO? z9np%`ybK!MSrGjv;Y_wk`w2^3Mwm?W7F zN!tkXWpxx{h7QwB^$nvlM=fFk}R{8yQtyIwn-Zb^JNiQ4ReU; z%lZdm?Q{)awnN%gm@n(3H?fc?tApv|GG@@1q|RyFA#E*OjUnDy>b@}DRP;$-6%VL} zFWVvQEc`S^m{o$vK8>jX^xY#>R}EjbL)uuFFPpzw)Spc@nZ7LEnZ;lYU$#TqRhTbp z(5Uv1>C0kWSOuN`Uh|USj-kaCfC@v znAL`0F!kgJ z`LYNBiKb=xtYm-AD%G)z*GbC>Ga~fQg#`Pu?m<`#*RYG%NvjDvPzYL`63RSKTtn-N zQ0MeqCoLuxvzjAHsiRP4wV_MQ);K-aNh=9EJ+0-J6jeDr=>c?xD9u(&W=WffC9@?r zlxg@|%$5dWF^f;1VZN-sF(r?znlI}chQ)LZUv`7E zmwb>eG40gL0Z_LG*B7GBz3c{QLt(xwvOR`1Md@dzkL#bH#CQ!~c7wF5FkjY5S7Kr5 zgi+eI$#iV})73eRH%MCxS7V5Kmr6BTjnPAJmaR^f)COs1;ioY|X|>as8i2)U4PSPH zw6QQ>HvhU@$(QBO$i2KpXQ-27yg}Mkm@jKgulA7XvtnI2D^tgp-5_l#%$Ln6WNKfg z`$`R~GhfG--5~8J%$G%|aYqfAJ}x8Muh(HKiik3&Qqol;U2uia1Fb7 zleC(!1BIYTNhtF`aSfp}MV-TSPR~u!Vq!6?Ih>))YD2J?t>MdVl2#ISdRp6(`?4H5 zx%aZE*L_=Y$4Z&izhA+EG+D4c!s{^d$ zc~$deeZ#Pru5k=)lJ=4h(j}&qUJih?JqU~88uzlBqz#4nvdH!r))a?c?xR}oY@&uQ zyGhzrm@n%jC9$xWZYpCW&KlNIW2CKxt1-kotLDq1A5l(+U@=?cG~OcZEc`S^n62W= zqO7U`Sd7;2Ww%Hh3-e|3Z$+1US!Rwg-kHT&jxTMvn0o zX-i?gY)+w}eVINh)t$v;jeFTG(tg5xS%k_;_p)Sv7MnGUh%M4`!i)(0Q{J{O%QPa~ zgRmH`;mdB3Ruguh5VW$DBsVwLF;AV-bBnZ?Sj=jUC>5t*ItLbi+R*xr*NFsgkya9R zdZH9nIX&qCbcV%Vz`FP_DH*k1!K(-4TY;V1dG`kzHE=Q zjWAzUM`6h$OHEcvZkf5qzTx%7s1u&)k@k`g(j}&q{02bU9z%{|Fkcqg z2E&?S-oN?>VliIBm+g^u73Rx2>8<3;GDc#tUPq0Qwid3&5cjT{FS|~EN|~2>Tf>*_ zk#-h-8Y9eBJB_IUSi4%om+g@@7Us*YXI~Bu@n!MOECy@%vOUtS!hBhSMzx1bUl!}i zVy=cS+aql$%$Ln6G_)_%m!-P1n1q_B+wN~iIBI8-1{CJZB4j42mP2Qgc}q)8BK}z& zK1tgNGb1*%?|BD%v+hASD<=$FuIP_) z!%)PcNEC@btb=6IM#6kror;oYmRi3|-_WH$yc*SeR>T5UF;N z>DpplS={}49qvf03iEAq5)B>9^lho`EH3N#wk^_dq&06||xCRSMidmo3LnWrSNL(Hux} zp97%_B4-hGd|SF+Ip*7L*70re&YUF@b$na8FgfPi8bqpHWcs#PR~C15d|SHqIOf~t zBpN!H>DyA>SzOlfZRw)pm~V@aS;@C0`?L70V@=RC$1!VSYZkt(dk_}Kb!=q12sw7H z5X34eWqNh4A?`i5r0vAwR&zi@xz&bXaa+f?rOT6Jr>M0tRef7}02ZgU+bz_$<$#Y{ zDGHsTmTyZJD#v_VV}7*>Oy3sk%2^sw$G4?xkz>AXPNJcMsn|Kw=cR_#xfgYO+b(&# zhxxV$IhA}{vcJv`X+zbqk?9)bm^IOxg>UN~gvD_k-)j-4w6v20bz&24ndQ|lC^ ztCX|2)f`i%Z)-844Z+&mI=(Gkt{gi>QBwIk%RFkyS!#I!)hd=-G?aW>rWfEIgtfP| zz%EZZ@W%fSeS3?q@|K?%NU8Xi1kioGIE9MGQ``f z=G*q(PT!U;YmT4F2)9b2*#|B)0Bc|C__lOCbIiBxp}sA%aunW~#bKR$Te@&L=Gz)X zs$FFIwpdpdcXfPQx;8oH+vX%1I+*FhQr%fx*70rWqU4xwi;!8#wyp#CADiTa_8M5mt|1(thIct9hqXNGf<8w6S&ch~KWWzogal7;Aqe zP3)a%K9fUWLv*7y9|GAlqL%!URuWHsOHOXEXqHCN_mPap)sbJim^ly6nvWWcTWt&$ zzyGoZp6Pn#9Uq=cj#z3Ae>4Z|lVUsCVgVK#hsAdtY}2L9z1S|%t7JI@+x93dp6lFZ z(pAnekFRtdr_>sT7BdN#F%paSe^UqNWb|r-iSvT@#Qb_Ohx*gvspyF+P&bESabHWT zkr`HmRzp0vRKhv5HuPkzUgFp4Ae*$nK8Pc`q+ncgT_0fZsCIHY=kStvJbIMNA=_4~b=$c~T3VP#2>KM) z$IRkNWPjE^*0~?5R>B;kGts>qLb)#p$XROH&U6iP)V^preP6D9;U0yvXtjJnx~w^R z$`I5liL2`iy2jSmp&q~F+ddw@nt#eYFfG2cF?7C2->8o7OBX#yPgRswwmxP1zT`CZ zsX0&@pB%YHNQ0&&)OI)e(U(Ybp3PG_tk-hSwURi*EbG} z?>fFOT?*Zc?GoQgmP4>@kHX@)j_i|kS1v|Sxe2i$MDw)`Wpc(3F8(go5{ z-`7cTHQ$#p6pQ;hT8*xlPG~j6lS{oMzE-2hVDVe$l%|WN6Q?vntCDQ4@7u!ew%O-= zE#H@}laBhnt;}yKl~?BMNR#Uzqf>}FzAs%89rb+;0@OZmeP66EXLV}%zCF^?LVe$y zMnfNSeP60Si_bc~Z?%%=aBf&WexiGsSs3&HEKYy2Mzn^mdA_QmHAjbs`*7`B62cGV z;hFb{hoT3I@C??N#ler&;ed=c-EVWCMC^k({SS!;qjwVY^R?m+)}fxXbxtRk>c24@ zUB5A%G~bg`{eZYHI{J-fj@*Moa*1ozww?zt7*vlsV5k#H%zYb(5KJG^TrylXeS>w0+Cjqt9U z;a#`FyN08w9ge1UIGWnwXljR}sU41{b~u{a;b>}yqp2N^rgk`*I^k&Qgrlhwj;2mH znmXZV>V%`I6ON`%IGQ@)XzGNcsT+=_ZaA8{;b`iHqp2H?rfxWzy5VT*hNGz)j;3xn zn%2Y7v>uM8^>8$;hofmd98K%tXj%_P(|S0X*2B@X9*(Asa5Qa%qiG`?O&j58+6YI} zMmU-_!qK!5j;4)pG;M^VX)_#6o8f5M3`f&uIGQ%Y(X<(krp<6PZHA+1GaOBu;b__l zN7Gg~nzq8xv=xq~t#CALg`;UJ98Fu{Xxa)#(^fc|df{m5g`=q#j;3BXntI`A>V>1J z7mlW0IGTFlXzKOi;1BLxx>TcYt47~;jlP{4eY-XKuGi?hQKRo>jlNqo`qsczs|K!G zHE`9cfvZ*xT(xT8s#OD5ts1y$)xcG&2CiB)a0LPKU|k0h^6#L2_wt${0uln2iXq4My)HE;#N^6+_U;0nU!;q%tO6$H$~=dFP&2$_e^TLV`R zG!LE^qUPa!Yv2mv=Hc_!z!gN!!{@DmD~O$k&szgm5Iql`7sBV^eQV$f0_frM*1#2n z(8K4gfh!22htFFBR}e-IpSK3CAdntBFGSMA`_{k}#L~m(t${0uriag416L4F51+RN zt{|cwJTHXQ!~5326$I78=dFP&2&;$BTLV`RSP!4K2Cg8q9zJgkTtRR>cwUIEhxe_4 zD~PX$&szgm5Md9Ww+5~t#vVRz4O~H#J$POSvxoPsfh!2KhtFFBR}g9spSK3CAlM#0 zZw*{QxIKK{a9ly^{l!!84CD3D?r{6LSH5QZQgcZ(u0(T}L{o@`!JW}0np-{ATo#R& zqxsdJZ8YC0PWRKK{K<{M(Qp#2h^6i0(fsnyHpI6z#D~0j%bz^FKH44ZzkE22MDh2S znU!yn>!jy9KEX)#f9~*LGL4R_Kg`j8;N#-BdXF4^@5gt(>*LEyQNroYiR=4^R|ot3 z$#i(M^H5TOU-rnfo`ef~m!|`5!nE!j{eP)B{;mzSee|q<5s5g7B+3e_W09WHUOQNcTBD^J5i9nZw=4+XA z`mOdJ^rP+7=$Q0ct-a_cdZ-yN^y}?QNKV(tM^&khngv5YGK;a*82X*PX!N6RmL2Fv z+bdQu^yhCkz34Yusu3~t=k1Io!O)MirP1GZ6&^x$C(%QZj!BEv-h+O$y&4_Enbq2h zexiq(0YksuzJz24`pG_O77YDJ4`Zt_^gDae=x<{$_8sU)+bdQu^tZE@%U<*wE!Bt^ z`tx?il3?gZ+S2InR7OA1Ly?Z5-)iqcKiXc6j-lUb?L|M)L(PDpUvJMvKiNmkf}tPj zVQe*qerGQl{hf;FN82k_F!XoW=r>xb5i#`V?TjVC(2umG(ci6%exipW9Yep>-h+O$ zy&4@uzt!4{exiq(0Yksuo{N66kD3KTKhne4Y7G6(UNri<7158jSFB*@@3PTvv{WNv z=+E03OM;;vX-lJjy)ycV9*T4f{Z@Mq`qB1kbPWAgYcKkV9%=>*{d#*Y`pG_O77YDJ z4`Zt_^gDae=wGjhezd(}1w;Qj8~sL0H6n)oyq&Qm82XX6H2OCxqo3%ZNXO7`wfCSO zZLdbh&~LT&qMzuYX28&|x96gt?4xGE(2w*mwi-jfvlos2jf&_;+bdQu^lz}yZ?se+ zV(8D?8B2npA8AXYf3q_Bi5`k{4E!aVg*D0CL8@mOEn^f{=A*BBpCXUwlwG4xxlz33-;s2MQy z>+QMdC;O;bF!UomjIGAd@9agR-|_uT^tKCauUNs*pMPS{i+-b}8WBT(-p*JO4E;#k zWFgRM&iZws4S%#GF4091&+DKG09u6Bnn89@GvZt@f>xgb9I{uu>;A z4~A#>`ov&59UjPw{>*|f_8cwJ?Ch3oKd}zTHEwzg5L<(EQ0$;3cJr&(k|>~+YK^99 zObp-H2RhVkIF>npHFijYetpQ42BM21A6;^*gHIaJ4r+W9bXIe}G!R|X9H=zt9ZJAX zq2nd2)Je^QN&_+vM;e^n9MXVZ>$;=??V#8}r6IqH^-F`%RE>#BL*CF>7*raN#>sM! zm3ViE3b#8(?bw}CbylrP?2JhSuqwlgsnb zLNvGB*q(z-HI1P8W^ARcEsMo%)mZ$6hIrPSwN|2MG98chhI{?VKIF}Q3#P$f;`_;}iGY4e?xU;`%LO@o@4gtlezLYZK%-b`3snlV;|XR<=yK_=9=zM2*#p zPF_xS`m)?;Mk(MDN)&42`p}AJ0f1Qy(l*m0?-yq$H^!rbn|+xns$^BAi=G#!hWq=Y zqlx~*vqD@QULPeQoi#fczG3t|EtZD|d&O8T41kEgc(6CT1No}+S>o!v_@0J%zS^e0 zGCtf}8xQwx@2YIe(ZL>U_k;i?4!jB?{mpZtU7bECM-5cQb>JM&WLazwm>IzfVs86b z^Nd)zIUFAK_eQrMiTIIYGNJg&XwTTJ|EM9>t1X5SD^jD?KMrcDw0`T}9thvt5Fe?w zV5QS4Bad9b+VGwgfz+XZ6IfRK&h!3X&x@yOt)Ou%gyFZALe6$Eo(&XcGmLdSq)!Rv zrL|M&Y3k50&xb}6Jx1V%{zGHcGWQn6!uDKrOoZwi3XN~SzG(;On+noWSDaJwzI^pf zMj*ArTt9-cNJU+J)1HOCY1h*?=_1ur6xm9*W;2YXZ%S$=S3X>Qb1YcjG@7)&`CE;u zht2Nc!R}zXrb0Lq5S!#PBmowqX{f=9ybN}C!*w|3R7OMG)f_`gR{@_kY!)B`fy zGmbLjFg zwG#v&oY4{7YKO0tSG0Ow9jfPo(Ng*s#(O{5uE{hG+G~VU`KWQJc~UH1NeGcHqx13L ze^ps>Q`V$$<_K6llmY+#oeOzLNPH>#ufov`UjW^#CJ8x(#XeMduvrzl`E<#Gn7iUy z{!9^x@|KAoZitG(!%yEBjv;uc{)j`@%OpL2CHn!((!5Wc-#!9eZcvaPHb#>uEu%V%!sDY10r zXfPhgFbk+S)((TIbR~J+4Qfz&UW4iGjWkpP@4X=icO{F0&0Sg%cZhe?k5;DsfC z$>{ol{`2YKbg-ZO{M=`UU}UN%?Aec^&)$MdmvZ;?XQi$eR6E#`(nR@B^|s;xaYm6j z*x&C@MXA>sl^e>(GZ1R(0Jn`>}^OM2r z!?kOp{o%pj))3PCj&5gf`pbUVQgm$R_^rde;eHu`l(l&vPL+m^?tf)}usghQxW6|X ze|j{%@%-@G;P(DB#wd_eq}B!mL^Vfp35Y_YreY|BnlF(iYO*p-0n>bAL#(QKTUlmI{bWmLwXe>;WvbkL-~qJzz-$dLNY|lP70k$&YC3baYnxr2K;7z{j(l zvBR1{OZ{TaP|BylSBT1y!s$U;rz?cz0_J&AM#2Y{9fGt?@;$hvH?Z`^e9%%B3xx9iRRZK(`me*d32~yE@1Sl91-`0>X zVA;92XhvgBL+LC)r9syu6{Nx0EfHsQ!;pSHX&}3}q(SeL!<*h^w7E+ftPVbDAUaf< z!zT^a9DZp)6}@YnDoR7HmpRgaG)|y|N<-dS0}v_=M#FT;G0VtVsfVLJnZ z>_wv=>83G)p`YxLvV@^u@8d^5+S-9ltGx&PMEfc;c+hXn;6*=OHwi;O+B&5lLqAgK zh(tfqmPUWxSOX1)exqGVDu#Y%r5+Of&R#V7k!~6z82ZT`DN7jo^*(;|qpcm-wAy>n zPqeQxg9rWA3|{opb(1jkqpef=G4vyqj!5()ZE5u9jWy6<=r`J>q+;lIR_Y4-!> z(w0Vl-dF<-hJK@6N-BnaXQdtz{mx!A`jKuLBN+P09w|#0`t?43^rNjE*tFVv&`-3l zGJ^;G)(l?s({+ zW2cm!tbkrp$JpJrvsc0yc7ud;D-=$Qol<0vMD@hjDW&%*ol-HTgVEOM=&bg|DHZ#8 zR@5ZztISZGQoI@TQ;Jw2APN_Uaq5!)*rD80n18JMUhncSzW2aO; zXacU3R1E$4wLTyH&R#V7k!~LJlRZ+FF!bwv{OCtpJFscB_n@C>Uu6al`mGtf=qD31 zQ1qj%Q~ELV=P3~k^doI)^yiH=&|v5{+NGpoTHIfIF%Hbvi_r`vUDc-2rtuPzuNM<} zXhgnVOyr><&C_M$bI0}z^_WHfY1gRV*N!vK{AXSaj}biSabNL0af(0G@#8>9#Cn>^>P7lS{jhX4r$0+YXCx}!DyJ0 zi&;i`IvFku*oorol`w$W$VfMh6U;9ld!#I3Dn{>976i=BM_Z?(v)XH7;N#go6Ie57 zLcp4#Bm?*gAsU|a3(&73rYnS5MtQoNApn@PAZ-))V17Z~SOX1a85!*y=+~3I_~>`` zqS23Z(-^_fPxf%2U+?2bKib-XeyhC){Y3jJGkDN%&EQ2pnt?~6A8nn|kD)(Lmt&wG zX-lI&Z>)g^L%-23B^A^Cda@TE{mx!A`jKuLBN+P09w|$hX4Cul(T}!vVAE>vK|j&H z$_yU#TQhjkk7nSJ=toS{Z_VIEKbnC@q91LY(vP7(PnTn$A8AXY zKX0so21CEmE+rL1zn<*HN58Wdjeew?#t4RfvPa4ihJL+|AN^=+2R5zt9`qCKtIXg* zzcqsw{b&XriGH+oNK_TDg7Aw^K>}|`jNIY`t!yb zXfX5}?NU-P^y|r9eDphe(db9IX^dd#CwrtUVd&TU_|cEHc3{(L??FG&zRC<9^jkA{ z(T`@}k?2QTr}Sg!&(q}?=ttVp=+7H#pux~@v`b0F(61+Z@zL+>MWY|-rZIw{pX`yc zgrQ&W<3~T*+JQ~0y$Ah7`zkYd&~MG)ML(K>N1`8Xozjn?KTnrqpdV>Vqd#w~fd)gr z(Jm#G$k&U>e7*Rg>-&dSvucU24hMT{@_$|%j7PGhAyi+~bNFK8OIc?1vtNSJjeFC( zM^GN^A2h_zHQGPh__yT0l}1DS#fEr5C7wZ<&FT1XU)FPkPx=fGcJD^ZFT$VG(ey5n zhIaC^TFRQaY()LJtS;?R^8;eBzq>z}Ori(HJEM2(AMS2HlZ;m`izoj3hWL>B#Y@BS zbtqZ)+>PPx%_As+IGwER=6~-WjfdApcVxlJR|dN`36fV6l5J*9^3$5ZOL}Sbtazcb zD7&opm#nBwZWP%+i>vUa7g;nZzE7NiLYc$qun*-fp*rkXe#oaG{(K`XikWSuwymX- zjmgR&1Xf=&V(-NlJ|W(zb`ok$LwU!qlpjP=Ov<`<2fMa(_EK7S2fJWHx)u6(2Rp>c z9*Jf~KgFx}DIMZF*d-y_Ivt(WzBt6aJS%FF_Elym4sqTL`XNrN5OHlutPs*Vr5`)Q z^9&LLNZ7R{(l&t)c5RtA)o%G%?`KQIZ2M)XL+8YfWVrZeuK30(MDfJ%d|Ni0~a zXJe=|IJ-Hd0qN(H2C|Dw8uU&KX+WF1q`~UolLn$gr8#`kV9nu|22|10(ttL1Nkgue zje`L_#F54hX~V3+B zfZ6$I>vVKhdrb^{JlkgiYX(gSSTmGl0AC?Q!;_YgBb{*vvyAd=LPG$svjb_Hzz6dS z^2QoyFw4kj=Rm)n3&%&lvloqiq?^VFhJLb#1O0j*Kl;(u4)j~?J?JOeSDC?serpCV z`q2bM68&iFlzt5Tc{U*f{YYCH{dr>zG#L7gb}6Zt?$>kS_~>``qS23Z(-^_fPxeSz z!Ze%S$B%xrwF8@0dk^}F_El!^px>Ioi+(ggkwib*I;9^&f1XXqKtIx!Mt|N|0}Y0L zqg_fWhJHO4j*ot4FB<(wH;oYt{bY}nB@F#~A3yrh)(&i1?LFux+Ey&;B{dqPa1N}%_8vS`=4Kx_~jdm%i82a^GI6nHFy=e3!-84op^pibOmN4|| zef;Q0TRX66wfCT(XkTRp5BjYcyy!<06iM`>tyB6j^yk@x4D=&yY4qofHPB$_H`=A7 zV(8a%;rQrx_M*{`bki8Y&`iU4E_!F zK=>Ui;&d4h(Mq1Cr{^8Ut`xr52P!jb@h29oZ)^YHqWFQu?YY&)SN{JiYw#5exG~WD zw>v%k?>5r9sLxI&qw5FJV)CcAlBuD}nsh1NHy4}d#EILJVIMNkZ;lSG%dF`vOTE!- zN`ccu09znT2I-HEAqr-!Kv3EG<;+GBPIyrO9;T$*V_lD!O{lVSgxbLM@{IS9LM$S2wC*&MYDFlo4jajq)v7Ge}izlP)Al4PB;-4A3 zZ|1yXX*me*g%@YBz3J3}PUJtAGuRWG(Ru|23wrQH7yruOcq``^3tf;LC;q6#^OlpX zkK}AUA$rkfWww$@amHLnz+xieziF`E%UQ=FFgoj{2#m#akDw@6w4ct`Xu5#nRJIR?{6mXW9>!oLl^CPWq-;k2;k|uerUyXEI$_Ni3#|7+7pE z?Y5dinJ#7wK#=&N$!}pr`IMJJOENpP%}$#t5hy1R0#ukq6}~FM+UUZT?o2J8bq;VgI&?@WLp=H8zn0Ycsvx zWb0UBOoAzg&R=!Oi>n?2Wz&0f9$YvQr!L*4}_W(wKKe`@g8E-WkT z7RTU^5YV>Urg|dhuqD=_$Cw<3T2rfuu?${eX(QW(HHA6E`l}sOY^-f#4>0A}2E#f2y zUl8%HO`Z!A3v-LfDJTK&HkrP!XJnjw@A2*oe+tvgYP-!Woy3ycX>azrb3d6T7&#xY z7!DI`(K55zZZiv=o7!Z*ndRIEu9@Ya311P*VQLp<7G_&o6BK@`y~UgOoZ(W=G&UVX zS`JgYFt>af%<@5Pj@>b?03~J3Ni3<~VKP1Ac42+7HoEXrZMJ*Gy;kTJ6~hx8x&mE9?SaZZ-QnC`4|yk>5a{~26&3qvcU_>*OJbT2hOn)Cgn*p9ZU z^PQ*=f^&yh`!BP^@0#ov<`z+&3Hi(>YHxMjQLtz7McEMDsLfiGl21-7hf~D=g~@GU za1qs|5N-*D-fcH`a-O!tM%1mw(>>f|$OS1yJiCR-#bevyCU@QUior`#Oc}D#kPA}g zxG=uFy3BG~ZJF){=`w3zg0p%OZEbd8jqx_S@MUer-C?ib%`T=Smc%L5)@B#>8E><7 zU)JWiZtyGQ*i%r^$->vgR^NriMyPMZ$HUe4W?_u-bu}?MPQA+(LS08oeHRuQ0d0gH zN_}rtP~YUR1fxEQ#<--H`Zp{ z9SsZK>|#n{VXJL+VWIIhTX#`yp53Wn!JDQ}N^tFklcl~l3X2W(ocR~c8*WRDuf7p4 zlfF!Loz<(B`rarkGy>Yl(dd?e#F@}21$ke#uy`foQTsN5-Jb*Do?OUGnMqzoOo^wu)!^qyy!d+aS z7a4a(9GafysVulF(vjnh;iLsfDy;+!BSc)4$mqVU6 z-S!qU>&-c5z%p3b*qepHg=J%t^Ai%j>9)9-ncALN3BDalgn0C}S(sed8Kslqc9=Ka z78mo_BIhF*c8a$7PIY$oY$7UmX<>B4ihe2L9< zmw}4~Xo>?WbVUks@D;I=bTK01{hNs#&AI4U4`?kL=fh}(!> zFK@c7GPcSl<|7uvp|V%4Z0xPV8pA=2%uT55E%)^!8@J?)1iwC&y=rA+Zxu!umW`c% zt;pjbZ_WH7l9`Q75NO?@Zxx0Y770Lx!42%KHnH>$sJTxoX7GTb(8|W%Dy%Fl8#^Dz zVN7q$Jhz_tv>72znAzA{g=K|hV;e(I0_wJ@nAzCqTm(KGqDj0{-YTpq%rn-%EyeSU zx7M3X@7Fsv!DOfqSoi5$g=K|##!mVO!F8KgOlR3Yalpz5(=*;G%qz^?K!|V&vli)M z+8xdb0vJS04$Cv%DoiZQGbTx*1Uw-2%!K; zTx4=IQqbK&Fy`&CXIf&veaHQxOxQEncUxRK9c3rk-td{2T;zlV%MO!kQ6}se+`4Tp zbaHZ|Lyaz{CBZei9Hhz=T!Zbx>cUJ-YplYPjqPo|_g9M!dT&5h#$qDhjbfHV-TbHG}Raj^^v=MG8_1*GUu%C%{ zjIM}FA4p%)V7oBPFsHe@#x%xT?hG&H;U9z3!WhHcW@Ku@@NT)SFs9#3 z%@@Keyi_)LEi5n0an9#<8QCqj&BgSbk*PvBv^?ilVQpcavoS3xs9Nqk)7?c!CRlRV z%%uNqd(N%G!oob~^|hh|4p&&+?-!YFuy<)Y}J!k*S zA^cmObE`19Fardk!YR~RDOIMm;hZ6WMeJ_b@|?BW_L%2P(nJZmr_4RmbM{P2EQUi1 zqyXVZ zEHc(+7oMu^MYh+uksJ#{#jpfxt_^icwG}4O%Gz+Mcsk;2!gw z*SVf^+kHdH^qh(5s&6pjIcufuG0z#9nlQYfti9`-)lAQsnlFS`%X8MM*<+q_KDW!r z?zqEUOwSpaDuhGJbJl9uW1h1yEhVT>zTUN!OlKOMjNsCthtqs)>&{uLT#tFq8`V8$ zC}QtWj_%tRrd#c$3SS@l&RMHxk9p3{c_;yPn_NuK**`PRqI2nGjjwVLOsjH_86XH1 zPN81UJs56sqjQD;77>ZF@0_*T_n7BQ(nJcn+wfv~&Yo$B#c+r`u{~$40zPJXp_7vvy;biiBG)Hz5QVRaedpXMVqz?|3pdtwraCkK zgf`RJw$`Db()OISqW3;sW;wAo&rtGyNi?oRbC((+y3CpZ-?tz**_+)ht|3{QUAVXi z=h=pOE z^9I**?yeJLIyS~_qBEUsfu-(l=L>g;qsts*{Vdg(NJ zJZDcOe5N%(IANFRcPTW_qEMlN(*#^m2?(uotr1wlk#_zGtyr_@uUDvCh7R zw74yXCAf8Mrc0aHp0if;-lxkfzZT)yZHEQuGHX&|MZEiuY<6Lhu{OK#Y!Sva%Z%w> zSi4?~Pw-fA*OARGEHu_;yFKUix)HHZIOt^IONFTKA|{5bZ^YNbJ?C||rN&m@#Dq9Q zuGl_u=d2aI$Dxf(PN{DwdGE4M*k~uGBslg-^s0U5tQEY+Jm*cW=e*%I$(U|1F&)9K zOSo;%d807KFwYs8nlQXK-mdyz6p6OytW~zhJm-8MmyzwMY0q?~k*Pv<)AF3PYWA4t zY)nfEYL_6!AxX3IoKYfe&snQrk9p2p)jelVEqkWt?46icTc_As+1grFd(3lo5RuLaFy>vP}p0lSEKGXCfCnQ*Qin#U8S*wAMnO^AR)J8+~`x%&CjwuO_ z{hMrMYiq^tS!@?BuI)@g?fVPjG+6ngy}@+YpWGN64Wqe78ZIN)8O`C|y<00)P-#0; zTIqYAF0-6igy)%DLlO)m_#zfHWV79=`4$z0XNxedT|_c{3CH*Zk1i2LL=4&N;wqB0 z*}990@El;N`HS>%a}|i?OF&|K&RP+CTzw;s9qu_p4Sbh6W2gEc+Of0d|Zzq zlauOugRgkVVF|8$Qj1+R=eJ?D=d2aL$2{jPuIId2qeKPMttM!}*U9#rH;Wh;<~bvy z6NdNZR`_|uypX1dA$Ya!oV9BAnCF}i?=rGIh3=W2Gcpyh7n#eq>ROF^%yTxT zr3BSe-ky1Pj7~;y=@4HkTU)Dak9p3$>Yg(cxp(PfrswRPm{?(1m_O1_S7`Z7$4oDsMvvzVrSe@^XPRE*gv4?4i>CZZzBU zatupwRd16m&si&u&tkjqNo~hsYvvI#rb}k6VnLZzTL+7w>{@x z5evgS=N{K{?#;YtE7PAQD6~E2UJ(PsJZEGa!tm~C31Li6m>QheS_i7JK(T%2tX0Fu zJm;Jomyr#H@GH_$fS&1ABXr;^VtdY74SdXVHby1|)qVTO^qkSja7J9Orc%%OC(R&% zR{8$^tHbeh?db4ue>h%qZr*0|gJQA2yFZvrqKCwT(L47KcekIxyAz~_?!Ne6Ee?vH z)v8Sbxj9?S59K_(M?4fgNbryhKycwxlbAr|Ko+m20hy`Fd~bSJUM^i)5kFupdXX}AA|(Vpo!gh(*L3gB_uztvqO)ob zwDE{#k1h6>#p3otG`Bo2{<=XbP!KI9yvL{0q2CyeuHTqWtTk?5az}mMlT-VExGy@d zQ=1M+5d4ru@F{U(dqol)%`d;XDE`SiMN;~7Je*7q$LeBlycaFJeE6xO_~d7`EsgKj z|D|o`Y5nK@R(z)X?@RGn;|KOnT2`C3^^=bEldkoXb?YY^)=xIApKMt_>6uJe1hyXOnCRH!SWI**COQ@q9gB&M#YD$q!kVgUG10Y{=vqv4Ehf75$QBb_ zi;1qqMAu@XYcXLBylyeEZZWZLF|lqjv2HQ3ZqH>gv2HQ3ZZWZLF<}j}VKK2`F|lDW zv0*W>VKK2`F|lFKZZWZ8F|lDWVRhQHnAo(K*tD3~w3yhmnAo(K*tD3~wCS;!*tD3i znrvB2Y*|cfSxjtMOl(<9Y*|cfSxjtMOl;W{T1@oJqoijs(X*K7SxodSCVCbVJ&TE+ z#YE3yqGvJD>&4|0ylotQ)+qs&joHLJ_2JYgn*r^P`Gc^+@dJ}`I3A1^ z)}SWka5flSOv>S8Fj|X;5Bs za|k#K%0s|WVEn+O91a4bg-JOa14avLT9a}(1B@;vi0(n*M{Ta!S1j>z1D}^&q059=V&+{-GbNwIPtdzNBzOUUjOP~ zx_hJlo9$1ha%#9E`6|`j=Lh;*(-|l*SuY=i}2tWdG{7jDNR{e|L<3 zcj4ck5v#8Z#*^XY(ZTio;YaQaZXN9p*Pc5(m`tPN>JPXPIj*i3qUD{%WpH_;W2M*v z70ZZ@#a}OCYH$dZ+|%LP+3i-VWc*h7Vu+FV2hiA&YPSD2nW##VOq4 z_t?Knz_Fb~Vzw{E@1>gbTbxF1PNNQ|Q8#bYjKs;wk%V=B3K6hLG$)QhwD6>giDm(} zcZzpxAD6+KM)t?C?PKX*TqARJ24qv zmwG||qxqN>dvcK;e{(}DT{(R5U<$=P7oHnTrg0b9A^S($3E4{2C)<~rLM&V!@5(}9 zE6vkl`O5hA!7gO1L$3NUSLnYXJ3c>|%8v8#w>HE|N(@MRI61A^kn`s>a_X#3+eyy- z)`l&Jx$R@6h0!VL)L58~wKt5}b?Qr7)U82DS0B})%5hy#M$8K$520(p^b;wFw(kNm zonxgT_WAe|LKM+(ZKzW@9xDv?QEjj`HdtE?My-8Lls--e8>hn?M`x~!jnh@*Wb-7!{S2!_GCB)bv)c1T^kMe`Uits!)Q55k2CRW3t}KH>9*MO zV2DkI)Bd&n!Bp8`ppy6chX=#{wbA}inqTk~hBCkI`$D7fUpLNN`RJ>|-NW%-I#&GE z;_U9>ts^i6C!@mybp)a3`BHRjr({blH3(Es%@@jq=u-@YNX#r(5-dvNjBxN{K-a>Ac?;m;WU{5<@5@tv|t=Yq6fw$F;w z?USd^H|7==kI%0xM@!-c^aa;rMI2YsuqvW=h;!rN-tFC~gxu(0Z?rp{TwJ&^+Pf12 zd=qPJRb0GJtlS(9kNSI~TOd4F;jTD_1O4KvP)(EhVmP_C;XblGH@_mjt05Mj9bbQG za0kR7nN3oAPOQKX!vl%%bej9JX)b0rijw$?4;OsgC(gIeHRk3Qj)Nm{;^e8*=c4oX zKR{1@@gboPl~cEpQ}+rGHvy5KMyJHe?jf)(`G@6L5ew4CYQsXFmjsCY(Pa9FBW zG%htRUJxsj+qZ7P^(u(dJY?6zbF1?p&f<<5adiD4Iw9uA!)w6on~jF#U0FtQhu|OS zJ!Jm-iz_*o7vXXfn(EK{vUP4;d_bH44b_K>$D81Ah`s4C&`Y8xq45X~dm4^eg1th0`Z6`8v8ZdNwA6%G37p<;<+?+~(mU9`O6(`f5 zsRMm=1@Li79>|CJ1nG`ke3v+z%T+pPj*G^{`wLi|5G(B^sVNuEEr~6${QT(kmk(ba z99X!_RPBYLuk!5xk;@T?#_^AQ?)G5c>UJD7G^o;tie8ZW&^}uz=#wW-Es5wq2GRJm z5Ffd71n1fFQs3$d6a*X;D5y^O-!f!1RU$VNexED8LM?y)oOm~koX3w|8XP@)u&2Z@ zHA6r}pGr(632#`^)%&6c#hJtD4X`30nsi$_n3J}+lK&m>Cz9}!G%{9KqWeW8y<704 zA+fuBd>^xpXJ2;A3O1K5e-cvZ+lCONlP?~`!67A;%@hY^&aK0!$fUwpSX znd;HROE3Kvbo9gGRQmhW*gK(=xcWz;#lhl}X4Y{0`Vfd4?~M*X@lJJh@maC((Gi@K z5u69G1sA~s-y7aZY%)!MkUQ{hu#JlzN{~y(5qc#2Tr(auL85jJ6!m>CPO|ix7Z1QN0Bn4I zxEGyC&hl^kZ-kCx;h zf4%{}Y)bTNZ4G-DVC7GMaDf?__)VGOJ+DlwDTw-ToSl!eVQ5<{d=k7QO2e8aaAG>H zoIVFuqj*}Z$i7Y$?HNFP^6>iQ!)w!9VBEs#x_@|mG*u?V86n^wljK)$oFz770#nFT zYcHHX-I$X)4+a>4R($yoqUB%^fF6WD?yd^)s5tr3;N~z}Xc&H}P{SqI$UXU-SWJv3 zxe^0;l)|4^I2V_| zrXSt9eal+?>{z_7IN*d>ZZ9U5n)tScSa@j&$EhZTF88Bkw-Ek^-%O+JnwQ1#mxkl( zLuo##15a)tm`^3m07lOVWpzk! z(`zU%N(tKfLY&c83ob6iza#E{Wq+_cya8|>jz2w`-gtg^Z2%TLOpyBh$&qRM59P!1 zBD7aFT9o`DHpKiZhi_yEk$;u~9hO7bIIkQ+rD$FO<@e>HXr2+LUxnykW>Y{;cF9wQ zyVx{FKzhob5lgQQleGb@m^xd4YjD8SPcJTBQGb^AP;&@kxP(gjXAmx2f&oog4K2|c7I@YDmx&YVMvha|)ah~i;BM*RlPAnzuChf&NWs-wTrGGO6C{lwx zE{?xC`5eSvU`J&LE0=xj-~fkeUo2i8ydFp$NM1GSUJ=JHtKv%9F$JbxX0R@U;i}vK zT`1lK>L&eht|N*C{+KueBaZimd*D(GRcILia20$j`FB`WP+YLA<2_JZ;7X~BLHW-K z+_OVN8Bk6dicdk9Kn9*>M%nDZ0}b%_c4#bJ$->a2 zWi<6h9het1r2YVwh+V$0=mPdRM&TVOX0}T92ep1X2 z5B9PRI15XqdQQUjKsi@(N8zRdZjjTkYr(d=!y#5DL`h{%R5W%C|rBqp2~|e zrPxbG%R{34OH>2YUi z2YWBe@a(?aaCLIYHALc=5Q+p3KNpO%&>DWwE`IVrDSEw~J97|f&K;#x<0X#S2{`T( zCtE(`(Iev$*1VB6d|$q>U~r4=`^ANJWZeOY=3?vdR`>Bs z?fCWNFOQ9PTaUHdk6j;JYj^i{d#$VM(O>?)9q}KE2eoH!YsD&!*FHZSA5OBn(Nf+Q zaSPhSUax7RLmp9@`nIyOq%K$+$K-{W{_rl+0YY6%Hs4i7#HmOh!b`>(9cXF(osxn` zrj_WNitFN0je)fijMflbAy_NDE2KDreTejllA9~vBSrVtdDo%DDQLZibPA+tzGm3x zhP}?>tQr-E>7_~vyNjpb;qB{Sh;2Xj%GW@u%7-!lUEFziS>E@@kFQN{9jy(IKpU?O z!LfZDvO&ixeL&wxG7~_&E-t8Ch0L0qtR+ZD?sm-i>4%n$=0h26_wQVQXN>aoWn)`p zUh|h4Iz97nFM7H>Zt|nmx@OM2_%9kps4lY%Ud!~}OFN$covdD+ z?_4P7-(R~e6I*u99vs3o=WrjQf=PIH=N*$9kcS1CVQ}vPdi!VyX=6!aT?&>Q3T8Z{+vuMc;@0ULc@g)O&b63Xt4LwSvG4N`<5Y_zRh z6A}atl)bh15Wi-|%}Y`A*j&iwmks4o9JL6(Ocp0ZPW(#gjxN&f>zbA=NQ~j)TV}Uw z_k7Z#e2X{nYmDo$(q&m(-cgw^<3DSiOt zHJc>6`89t~8tA}E0Kc~?^UDU}5aLa;qmoSjbH=J|>v{9@R*rb_aG=qAueF5d#2MrG zfbC90yvZN2xhYw?sG2Gd?7HpEIq~)4d=jPG>#NZC+GHxPS<>f9nh)YMTS2x{-|^JW ziQK#f|FI7RUXt*kK78>W3h!68#vR8?gPI_J5mkGFKZcXF#6naGh{ow|V5=D8yyHJ^ z!y4N0NGOuiJa5J}r7{Kes>0n#OjG^@nw0~c@`v!2B@dQ5bg(?mlQ6i%OO{9vH}&sU zHHwC}t`7Gg-{|^etw``%let(p7avsOAnQo21_IUZbR&t4D6?k-VM=Uqsv|f;T<+{Py5I+v-LA8jVO!pvkk&z#=4>p-A1vQ!+}NziD%$C95j;?(wuXkqya-`|jevnYi|MWEbnl77vWMC+h$$l8-jx+UpB|FI#y%CpsL@@1U%)X=vy_Z*%_;z5(-G`Xg1Qq0a0gF+4u z5K|V>aEGlHy9v%#Y~BctvwWexy&73|_nJzJ^VlV&kn4POY`Yv0ofp3Yk3dZ3vl*g1 zlRRd@D5jRtUZHFX#YFQF>zIZ6L?jrwC{JgVqb{GJE7w5P)lf<4sTYu^pUW1If|%u} z&s}PWLvf$U>t&Tpl=?P~O_mM@)7RAeid*8n=7c2|tL%j0E?mhZahbZ3zxOewo1s(6 zoN#td>d(yC0#RB~!)e9$JLc?4%v(!dv-UmkT1h_|b(zaqLm(_-PrS!OOR8!6twh(Z z<>g$?XTFNqSq5k9*!J9;M(l#{j5xDs@G$G=CM6b-*_A3#>5Nhph@VD`8iScg<<;7% zs05$bza;)9suKLmP&dRSt%a75E`-wMa#Oy)A>KjmN^u;KIy<%waAO&=KHYeaCd`&A=B_HYVOv@Sc2JL0k8eSkQT%*Ee9Q#T^zh~| zsT&a%MBB=PXzKb84WglQ5TUd=8E&c+N&_wX!VI|518U(+*>Nc=d`8Blpj!QlGhn0~ zT$?2;{1DrZ&WTrNz=}7}HZwSvBW!dD_BlsgyUwkH}R}=)?2cG|FB^@ap(e;f^?;z znRlO&Oo8AXw^J9Oe$hmE)ybz65^RaTIw#UEN)=%`l&>`-SZBFmXfOLB0Bz<(oeXvA zbzEfnRDF>k!hya`68}p>q+e(9EtnC)<1JV-e8+vH=ZhxBwr!cYBkgiHPXn;aYx3u} zivQH%LA47H%F1Hs(ymiVYsOblByT7lHjf%vkLH=Dx_K+frm%##!RJ`$BpXNb}l#Ab+adTvo4+Wt{PR>cQ*plbau~j$5ddwSBJaQO_;UWw2_9pJ<5WYfsu0U{*mO z`O^-urewtxbVWX)RCs#j0a*l;AB!z@z8XPiSFoq^mV}hDS#tl_8#*EZ2tEQCOG&3m zLrPXN!+*B1lqtcC+?Kyd{&~T9JzgZY(i>?K-{(+?okArN*rWAgN*D&Y!BIjYMnxF0V{I7&rK zVdhYz-zs%&<+BQ{<%;jbfaTS(tCz}8`=y`3<<3|{L~msG2K}9{S1+YL4V)+C93Q^( z@RLtHeD~qK(cZ%chtr2;g^Guvw2CZX_ShSv$#DD5!|7)pUGv`AzVyL|pL$rA1Nx26 z_f%&kKS`d-Sn2Mx6)TV4<aduD;+Z z6`QfhdCW0zVa1Oisl-*uP28@TY)L#D-IO^HjbN4fuoaPkI?&rfLOa2?ytlPHGGyiG z*t>TM8FPW~cwUZEKGO@_XFK!`>Ng3!m$~{&b9N{h;mVA>b5_F6nZ^cooumwbaK+_A z*Y1K<&mF@3^1X%Sb6^Y^A$QJ7y%q5N-iCOmsYF%vAh0r_z*Oob#Fsv6HbQL2#Dh${ z4E>mJh3Dret3_VT!+!!CFe}M2Ay6w$@Yd7!CeS9|y~~>DcFLeC!Io5Fi`nRiT^kWQ z7NNiioR39Xi$k#_FHyvQ>j+qOOSdfn3rt(M^YG{(%|Tk-xj%iWHGlDB1J+cot8J}9 z!ZjJr>lWcid?V;+WXhwF*Y;JG&X2^|?bGtXo7DU1E1vAg72k*e0D_PZ(NJMXEI>&I zAiKc<$R9!i6`vg39gbzeED+}Dpk&LJY9zIChvNg{ebIYShod|eJXB0~@Z8QBSQqsH zwMFDTn(K>0i{d96_A|=zM5S{xfKKZXi7W-C?wXg9Z@iV#LK{wUk%L(LnK?TuSmwYy zBdw>H<>s*!IV=C(-bF9X36t`Fc_ zFr_7i+p@ft80SSz!L}NU78LV-)=4#7?OS(7#!MwiK}^OF;%6y6;?E+_t_&VbS@H4g z5*B<|L5ZFCLFAE@F{SsBmC?SCmym6CaU%Y7Lu{GGUorcwL-qL( z6$%2PjSo*{LFUq}3vC$4+Cx}!@|^UttS5Uo5s|7%A6uq?I+EjHP68gr6ZbSlhLE&v zt(zH}t5Rj!Qk~kExh4fVDn0L)Px?Flt+z}1+Esq3{iw)(uyq7Q_N^D1_Da7Fp8*>{ zO=23z1U|XmPD4P=3@h(~CL{%O$R^k?Qt4@Ieq9%h1%Rswg*Xr&q|u~4I3z1UtbwGi zRQi9#S zNhTQ1r>RDryai{iX&zT)t&?KGell9no*d@VhAaqw%<|LZ_mHU^5&I}TvV)g zvJJ7Hoff0b+go@mjj+wj!#f> zDm=qK$f|B&Tbj)_L%j3m%MoAGeQSuHniKCT?OI8kfw6MoICO_yc+Hk5Ga=l|$ZC}z zBxU@8hx)daq7$DrRxwQqHuI$I40=Nq5}8*}F|fQ#znf zH^wgdTF&39l9Sn^d8re;Nzl3C5Z}=-luk-^ar0oSo$SH}W0YreXZH;`9p!rp@l6eJ zt|U~?K?UC^skm8|pp;zc)LkC$UQyYjP@N=88Ig(j@^|VR#ZdL68hPStvxogDt1@D` z23Aku#Fg|10X}#Uk%oS>f)Vf>D7C?ju@j@$g_QKoxzTTpkyQMd?5@e^3rZ$;OM^x1 zbHrkS4%A%^W9sa}$oR3Os6Yp?u;MdWy&k={%-k;6vDf~$mUJJR&cGgL?@!FZBlShk z?DdKM+r%JF>rv>xNNDmi^!0}MqyTpWwONOIv#3Avb&lb49ms+-f)epp^YTeL>k0J? z1z3iBiE~5a=)cnrGp%%pTxqWs6kmrSwvBs)^yYK{?zBvicV-l{QJg1lUTQa zy2QF=Ih@KtEhwl^hRW>+_1iP{@!VEuI$L4&mPF3s%@LOf4TwLcMafdhD;uHNe7T6@ zR=r_LIq2|Ny={io+jd&LDalTIHe;*BYE`u|3a@Zftx{Iv7H*YFn?H$hB%+U3|CKp9 zRz8WN%<`PPolBp!r7%gYCjespr5nB+;x=hUdm*Kzhi&{PmGJqcl$Mm(eVOvnzqB@( z!Y4dsdivUAxPPr24I$4X(WOrGYY{G?9>84@hfnpgremEj=dN*y-S z>S^H5K2aID85M;b*FLuE(J?zo{*n3Y_T7z)l4gZ*Y?m zF?j@M7Mme{cTT)C)Adghg=W8}aasv*mP+(3U+v$^bF2Aml+u=GvFH9bZYk1AXN%*p zOxoN|sg}Bv<^SjhubJYUru*|3ES@c(Gf$*3njoi8nZy&={s%dYXvpz`X@ zcsabV68}dd`-l$*0#`?aNy&ODz2hk%+LhGZd^R^BH^q9?sc1qVA0{Dt;sFU(Sh7&2n9HC;ndU z7SE+*Z;rPXjp@MG-!t*+bK(;-T=&FzpY=}0GgpyHb43{g5vLF!!>g9_{F8b5CYHAH>VkQriiDXo!~7^ToW11+G~f(WN=jw8IfW+ zb@}QUqFib93guLL8JvKJdpx${W;)|6UsNtqdcj-vMQ7}ji{ETyg<<`$P5tk}5wEy3 zadmv~o$LY;xVtXR71j;n>x2E1Q}r946{i0Up}7J6Bs)8+Bl1^0GcACvEj z&xC_I&N%mWozGs`9lu8m@*Q&S?K+=9uSQJhn8^0wXR;Z(7caQC>-^Qt z9dp+0I)7Ya*Lb(<{NWSSS22obu^H^!bv7dvZr9mk%BN_&P)LO#hl7LSoSMa2`u9!- z+gaC{IvYN_eI}$^w)<>cb?)uD&aY|7S6(pQ?RsfSmuV4vsjkmJ)F1P!VMi>0eY>s` zVpb)qyt;3>%y4hl6E5AZduTkf7<)#}JNI_In7E_{X26Ff@in}4tg0sWcD>-x6nvIP zQ*`Iut{3e~Hns}z1pC{;@u`9uTtTpJ*NcHn0k61MLmHD!1zfT4w0jzs`%=m1;)J;=%Xrka_cW9`+;e zle)V^Kb;SnHvia=J?%5RY5&?BFT7V;=JaKHx%@`?;YmAY1;k3juC+os1f^mzylXoye@#>sG@n~O z3DwP?5#Q1fZO%1Eo=miMG#uZ`svx~DE8y|2oj5J(zHsrZtmiJjKqQO1%f}^UkkZem z7o>&ySS4zrl`G@H!Q|+0GR!`|=Z=M@%NPeRT`W3lndJ(nORdE8idr}*OCav>?juuD zvxmd8PoTSIg?6z571NVu135Ux8|E2z<qbjz(4X2FOm}Z2|9Gt(vslfSDXzMz^5h!qeM297JBFMVf|<&UrP%3lt5a*);nQR@utF# zWOZ7-BNuh68qI(;`}cTu%17=T4G#96AKjX)?TvCfrS=2ypJm=97t=?g_acfo{o^ENzshx-gfr_>K)5eo&gW_N;<*H{y>LF{b>ongd*)p~%@bVsWG8Da)Rc+nvIHI0j%O~XywqB@pI~l36tPM! zgIFxQ1VY4Ln*fcI^DMrL=?h*yyf(cxxDzd=zM?;&O6_GX#Ic;_Fy3>&Bmi!mdkVMTs_edYbhxpSNwe*% zinfGv-iqsiQfaT{5xjioz>{O)MVe9V!X&h7@4ji&m?Tap@!b^+!t^sC?f`xj@~Jmy z*>Y5lY<5gr1X-BU#~SIIm0w^!N63zsbn5hEW*d?f56+xH>ARVg$mbjv5Oc+v!qtDJ zgR@f{oE6nmr`oZZGM7q#Eoo*vzvuI&vE zCigPG_apgJjVDAe+T=XdKq&KN`2Iis0#nGSrThacpILuyJ0mU-$JMyZp(9RkS^wJ1 zx3pA6>z*;m&-DGZInkdrlSbBPWL46-HQa@VVls&apQmCULSTFU;N&6@gmQNZPh zFHnrIQcC^v%Q7T0K}ZNbw^eT9M%p3P*yfMtt6 zHDhHyYazsQ{`@dPssPCsRlYiA*eVKo)MCqN7l?%GvL_z9sl=U(1kIdw*;+jA>A`u0_PD1A>YdT znS3r?R}Jd3YR8i=CGn-`W4R9buy``s4$>it7FvnC`LqoAGLMWxHKevwrbVY0}dyrdYIkv1?UF=)OOec8@6rO0?!aTF3pfDBJrIK z@p4suZynx(4}`&oS|OMPdU5aeuKZe9A3mBm+J!GLrmCWLuH?4|U|lCIT6hTv^d@(u z>K@Fwc*ZNy{D5_Jb)PuDeXe|U1>fpb8JVL!$PR{#Opy7T!{JdMGBhWp3}Nw84e@;F zUR|CZ9=$Lg?2TaOlGx0OKRqX23})kVw`GL!`O&p&!!f)I7)IwV zn?I_u<1<;$Vh#94J6yoO8oCIt55}Vb@J`E~c1*hp5y^YE^vlVoc21-ZuNvEphWOz* zZ)thUOv+%9k(bKf?DpDQio;8iy<$AkcfJXMQ~o69s6 ztJaBK2h@R#LnyJvYUj^6nAdrQr+0NY9Vo}WjvA8aYV&tF2ZjJ0Tm#f_6hKnx{Ig(| z))+;JMx7mp#Sh_~{My0jU<&WPhkK*JbQqe3MZf%>+c`^wF8<> zBd%O=Ro`L>0iyvAq9Xj^=~2m3j{acMpF)!U;W!H?YX*Zmq{S`}%>4En}%RrM7XyH}Aj#!oNPiW+qzr=xBeR2%57;G50&3 z$r8J6__Z^k_70pC0b*rGApYDO@p%k{ zDI-U&MqopQM~HW3$iy?sm{Yr&mjUV<8FS=!G!TNUY&rJ1o@UD7kJ*u#)G;f8RER^( zEV)%NCY}-^)4y`~hU-e>*s*6ywGqf*fPm^R_#;z%(r6xLCJHqnJ<|Ok(l(I}imzfx zxi^26)(*U1SLK}ZAJ}}VZ0S?iK~T7fy>_|FjlUK}mO_z9)PX^o!^(xM+uyFh189fD zg~ea4B1C2TUuButfziU^5L2Qe8$uE+E7bqFlUa+z-Xbx+=FTUdF5yDTPL^YNFVu>Ud%Ugc z3Pnol|Iggpg;;u&jiS9)cRG{GBwxN2&bE0S4G5aBB&^KpY-1))oZTjxXeJ4nNxqAN zELPGhrk$Riw!6p7h!;i;Uig^{FC?IdAYKXqFI4u22>L}h3J3N^4suk~qiiu(qFeAn zJWtj8)?05?Jx{$){j64DGC%fPRnK45^Z!(Ay+dL-gC1V;`GV$Y8SNV`adxLqmlD z;@9EB=!mDw0OSN1{ObD;mD#5b1T~?C)o~W)x^bK?-wr}9#Ep^cywQy!A58CYsVLOFC2P{#Oh5$>s#0X#h~g;rI;%gsE3Ik@VliFC-{#l@!Ostd&A z8lmumDtQ9xx#!^WRu{mi8Ef6_+jSyJhaIl3U=((01lWkF$}nVn(8gf;>fvGrFZZRL zy^mN)(^)7NKc+N94112Vj*Y!z<8|l)`|f^TSN@`I{f|`|cz>%bO~l=!cQTbFxh7r5 z<0>4X$_4R%55zloKv?1$T2lu7TxWU)oSvMzZgi^^*+6`KNEqi9lPR(|0?I^0l}z<< z1{$JW=If#0tN1?U{_(No2qF_pA%FS0ku$VK|X ze@fL+TSsZ^N;6HGjUU3*WUa15rl{w8U3${)2wgv5!q<8go{4OouYo=%zq9rjaGlSF zjb$tw&mwM4=p@mS@4?yjJ=hedN0Z+7z+Sh|n*dE4@ks+HFxI<}dH%?|K77ac&ye^Q zHE|1t%#F52Acd#`!?*FcA#`((Gf^D>jz~vKQ0cXqP-TdZPCN>(rK6pr>s18Sa$yT< zCE8q+S}0Ri@oHNgL3ps0)qBNsbWen=s$b^X;LumpJc@F0uFYid_ZTDzuAtSpUl z!S9!KSvH4vJ;7mfr40(45Je-2*j6oP9bEk+l^Gt04`y*e_VVDVzD=g>2-xx`okBnREj3;^7g$ zrHY2=BfIs_SV?p*{W9=mpk(Wz;)?_ENJxz~FSKSxT~+Lt@apG{sAcZ@vV)nU;?<#e zSK6T(DZ%rZ_>`-P;Up**hG1MV6Q3a3taz|3p_UmkD)AOu;mvC^ERmqZl*I6k-N|HZ z%zlIFU=!HOs!>kCO#%l%^ULU+AZT6(n(<~ReGpdBDuIEEe-(j6AALNPCsKpfd$Xos zEd>rJMNERD$!bV=-iY+e09gTV?ChG3vjUt+(O7J&hGQ+kc+pq6At`Xd%@$2Evm{~m(FX6q5y#rdq$|r|M^XD#wbq+i4BXEX+_zd2U zj&lac(ra2R(Ll=V4Yb3Hr4?nJq8-D=*(r#NZ?2-_<-@p;TKd7x_G0_OE_9zld;fR~ z)U1Z}t@eHtp?!t*A>e#CJDvWi8u|37EH&wl?UZwDZaaM8U}x{rOpb+_U7jCa8VxrF zFlg_V#qRFT{-OEDl`&MSJiGYF&Yu3U75QU(ix;hLSH>Vii@k00Lqz7O_!{oX@lc`A zEx9#I#3LlTy{*z1TMmXUD&#K@(3YMaqAX_9r!ZokEns=)C@1LOUPa&YK72T$jyZk2 zB^usK5Il|l3hTz|?L&|*=i@0U+^8J;;+`YiT7*{RL($$lGK6^nF_ArVeC+Dt;QFMA zGY`jSZ zcy=1+k^tnVb3!BK50~ZD-n7nT$c7Zaw!ubMZ3`x5u6v1@MxsklM*M9ku%0Fv>;u+G zZBj3nYSmJ>USoGq=AH^<7uJ*&Zw$rxqTF5El;vrPc2~(^6FEYu1Ph0z#M9)1nbpm1XGobkc_EkcUC0la`=@ zYAp_;#Mcu*#F1zWi z`u&y+Vp3J}rzA;z0BgWN9H~k4G-<98mv__dA%&TsH2mi6`O$p#JoJ{HAC1;WYmYzj zZ|~!~@LQe|zZ>wh{Y#4kt3Q=t&1NxneZU1hr22V!bTwOBw~^=F=#En++^ES9fH?!w z1HfwpC5V2W$)MTR{;>k*`2eVXp2@)VTKX(I$xtZFCCUd@#I!FUN?mt@Ao_VG|+644?z`8`D@P#dSW(z)Y*%MD^_SNX%6%pMV5i-IbDk4B)xOIy*M6CPKu==2Jc2-R+9m>D<_S zcXR6oaup0U%uwgk6#?})V)K0PY-@c<0D5*g*?Nr|o96@Q*nob^C01;nzh7;$7eQp# z;*PZ*u&EyG*gPNF)^2CT=K1S7RO7|w`ODW_A7fRPASgRF&jutSHqTyDP2_lrW5Gu( z8EL~8!8SKG-vxtK>w9gYlCk;jkvx+fpY5?O2gI{t^WDW=MV}Rb=q|vE&3C^DKDUVa zZ8V`0T#;(zvt#oXsE#2Jjoreq_wukbby!i?8p^O`A2L6Zy`^Sjw&5?hv3aw$bq(~} z*Y@Kagu9?u&m5N>n>U}hg3q#O%;&t=y!m-c=xZy#h;42JBsIce&M1R?EF0B);wIlj zVzK1u!^s=m*nACI8G9cK>3imNz2QGqY`#8{XY)r~-Z8OxhnWmZWZV58=W+8>3uUs@ z^L8d}?bz+QK1^(#fN4i7XNI2|bniJ?V&s<4t%>286>&9X-?oTkr1!mP>qfQj;N%jw z-PESNzpXWXf`;}ryc3hh=oXqsYfnP4V0FsUvh7llDO!&iUUqUi<)fsXHZ`NJ%9aAM z#th2KR^&fE4u4n+BvY)(q!Ct)PJ^=1jm;l;ZnQe}v=Lw4Cl+lqV1w~Vv3{{GLxBNK z<9j@If2%lkrW)OR;_v!@1DRYufEv=1@avrTJlsBQxQESyVpH+VlMuJMu-o1@BFE?H+(;@W#LZRh-pFLC4c>a!S)xYWnU z{V2XOjJoMHvRRiOOL_39$jb+GKbxj~mU9I{^noRe?NQAoi1jV-q)WyU^cBd`Y1@rz zfiNPNdkKDFAnvuEx!w8o#R2ewQ|;g02k?J&uR-4JW~&Hvx!{_NhOjW$IZyW@jH6nW z1Um1j>0W#I^(p>{_rLUT-{||dY|W3Jd!hczGgGu>toLb)CUhte{o}LKhQ&^MFY6uR zEoa_rzpO2|@~W5c&bL20gQEAlR8s}{m+SDCf!GsRhgI(&=V64_b^X+$A(;oeJVj5t zfDu}Ea`+oItgG0tHj@7F=11@WuWNwpohtG6#Ppxco|q=v6Vq6Gq6Xa{KiZ!n zMYr|K2xt||%4Mug2xw5`SQIR;pT6KDtu)mL3$+Q9SMnlgc7u4sO2o6<=2F@1_tEC- zW|!Tf$QR3PHRL=bcO9h`leNqWsx*sVqz*DQDk1io@*beCL@0z!bIh;~Bf8E6WW! zo2w8>K%HxQhKZL~dz<#f%Zhe$E}MGQl4M|92-Cf}cKQ_>h(8U*6NP!bcJ}n8wx>|H zHx)a6QJEa9ebL8;eiSeb$SG)B_>3h|ehl?YEtovhGIeQlanbzD#bSw7y^ zq!@wO%@s^QTUDCDgNee0P>g3N_7YvXWBL-YF)TaD!U*3`a@I|&^os-Z^a{-n?f9oP4jdX0XqmF6b(mROHE?Idi^^uqd;e(Y*ta6 z*-as~YK5OwIRdSIVf}xR9lyk%`aK~QftwT@)8H!sd%aNzYeyxZg$syb2w$HQr#V9# zX6*A9S!ZsrgaX>%1J3Y2f3^<@>ic=GmKiwq?BZx{`3bB>lpko_AZNt`qx&P=AodQS zH_0|6gM#`-f7Vq)0(*Pw1w)kAVqkjeYxtoGDg(0et*;qB3azWy_GN3Ki&TX;iKt8q zkvaGXmLiq-L;olm6fy40mV{j+-KF3Ovl4iOqv?r{Qq|5hzE1l{dxtzI-ZMIrbcgg# z=b0%;6+dl+o(z`848;c z%0RGV@unfz+z{me4PCn8)q!|CR(a6VN|!v#_mu=*?;fq0SIXD+s=}Bs#N-ajRGFJQ z+x7S>xDGBY7W*^Egv*`#UXrO8j*o&c?Er zb&NXZlFb7ZHZ3{^8GQO^r-=A&7747e!+AJtxq=OZ_;VKBAbY6J?asiug1^epfeh4q z7Y6vJGmfkhw`;&7uKo<#!iFhO{v{jACQ~cx#jcJYaPTx{Y}!_PFr$6O@=;zcsh8(m zLn!h8!6e?E_j8Y)Cw!E&v?5qM{=457(qZATBvzs$61vq|2)=!55Z_b|xTyHiOK1v~R zHVwrAFq$;ANyFPOK1=SqzW|EYJ(o7K*H*}m?Kd|j8dSt3yx9%zdxuvJbmWQMK%UrX{3hL5 z=+>>1O=qD2^a0!GKAjV$OZJG5iP;;(Py5m~&c+>n&4=h83SQE)Hz<&tyFUthK|+?3 z5AnGnUnz1@=rSA|$~{7qp&{5L6YAoU=9E&_b_tr<*?5AS!hsyfOQvJL>rMJ0Uh9tR zR6dDrLmUv5hRF6$UU|{fvtkd^Ijc7%-J1xS!{9Zj_@<*(*qP|#gnXgX_V5ZyHM`|9I=M_wwiuvJ*p1Xm+rA9@^%xvDX}x&jUap zcGTQ>O2n6_o(%~4F3$H}7_G_w5~4y<8!(uB7cuSbwoZCm?w z*ko;vTCw#GiS1*MifI^qD@>{QQ3U!VXEf0_Hu}^zlWTnTQ}|B7Uo0e>k@Br?$EZtW zngYqs281KC$?ZrC89mTuqn;L%(cN@5DzXg~EP_ChY)0I@fkXa-q4>!_d=T}3BAZ&) zOfXE81)Be;08kc4v{C_MGaoAhjvX4Z+@zF`q%hPnawG>&JY5E$-f$4^tdHh{IuO)k zzO0V3kdhQm+89CaoAquuN&JFJ?~8^G?$7$3S`JmJqTCyD{x(?B8A@4nl2wim7va)c z8Nk!?QfO5byxh#wk;T)2AWKaXIotRyE;cP!opnp~?t=w$vLN%kdT^6l74r>z;IPB> z6^z18Ei4-`RT+j1`OdOrhNwpBrJDFLr6FS2bDVW->>V4gLzCEd_w%~)7kObBZH3<7 zDpC`1_voEWMM|z&s2C9HH{Zbnf)>|+19u&h)C{0KIrT~d#n*>~ac(i0B8wxSOhi=4 zR1ZpI23Df|>g(eBl>5iWk|T&rEQS2#>qbV?p_b_bj7!1skw=+an}n@t;i&5iZ&ebA zJQ=@soKswhjQLnO#l=slGHdH7ja_L<%GvlKTus*MN@OH680uxup#o3Z9ii(7O!!)_ z!ZVSr^EJ@tB&Q~N4Ax~V8_yzcPUs}jlJCLU_C44Xr$>|C_rPAa(3=1~9q~Z~C@|K$ zka_<8>%Hq92nJWNmq&by8pwr0=0;m1kWyuFL-_6t<>F-GAALxqqa~>H+Dzz{c1K{% zTr8PzgftSVg)*swSKDgc%`#a+JB}Z0rTShm9o-Wl)%_*g;NioAEFMeyh1y7dMD9^b zO0a;kIlL(f4x1~lF?it$EY1=^#V!}6s89y42mx9>-YbYeoTK8!AkN5aE}S6|60cCn zVje`gZgR=x)X?vo$(!8+HAS3#IuDVx6$z^$O_8ch@fuakPo67KCY^x1czDEbsiGnJ z$efbsUi#q$fApr^^5@ow;3h$=i!TnuBOx`~ywI8%bycxn!mFRR|DqE-Z`r!JP4FBz zb6)9bY6J1=P`oScP>qz}`AjqtUCLF(a1zgX;noA6Alj^WuvtaKCh>;~zm^#? zD)A0n;mvC^ERmpS`>xZI@Q#z~@n{#{pgPzD_Ofb}Q*gsBNNxd=y)mPAY6CcUGn76E zt7w&gow{KhI&$b(^zme5zsZs5y;)PRmI4P1d{2U-$!bV=-iY+e09gTVEl2TTl+L7R zEVfm{v6f)G=u2*x^1*2_Aali|Nv1qL_1zZztzz4Fs>vE)_S0^l`zr1NoVvqOFUXetvVP($;-2gkL>K3bz$?x?(WY1VP|vkHQbZqp+ccsa%+}| z2PC<@bHJ2-C9CU?@Y~swJ!ki@&Ik==^@HuHhl^s<{kfC@KT}fgUoZ36ZCJd zqVIVhJ{(cUoIc(X4euogp2mNLbz}ARAxM|=@syNLOd4^aEqr6h5atEMME20}v8#)N z>yswZ*bN?esf>s9>t8=)w4@EWP4a_Bp;9>^jnhlz6hFsXv;{Zdr)@I(n+JD8T7mJ? zH=%9x#lobt<}NJ>Kz@%8Xr%n%vH;PgQXQ`Qyg0}W3LJW9++J+KGI$DIWdxc82 z1;(6FK`^X5124?z=4&DojzTHL&5t7i(GnPuP3T$Xk@e0y*=f-TtI9HTeAcHyD%m@v z2zH^}A?CXYa+8~-5RF}9B$a@npEqT-JJvsxfKnxEbu~q6A_cQ$OJMCnsYcLJAFI?3 zY;0Wv9TP4(8Hn4TD{vOS`;VWF3RN%>%&Exc2!14js%$O^W-}CEorHu2@Wx{<>i}(B z0guFDFu%gcem*wDOT(DA&`D>-SgS9t27WW_GXj^#lHLmX0LxWbzmH9#r%7{-xV)Qo z4=Kz9=HboT^P~Cf`6~yP=SQRU(c0sW{M-9deEGi{@U;C)iv+8`b18=WTEpZE(44Br z0U$tZpX6xOX2l+x23q86FMX=$nZ2^?eWIVICq-tvFlcts=2C#;+xkR5&t%|wEq#`q z#73p`ffcdy3y{IG^8YtJ%{LjZ5P_uoc_y#wP>W}AvWcU-TK)QN{u^sfBxl=xsNg+R z+xAZsT#@YOnM_36t}npI{Z>8E&ocl5t<3lCh830-fRgE0p!#_RV0E*TSqiD)VYHuT zXuGcQ=Y|PN3RgkDfGi(oX=Vm3Nov0NyeJ}ic~(G=2dqmZO2)7S-%;{AsTFyE<`1oGNMHF?DR;?RV<%?yv)^I zS-3Qn8=F_DV)3#0RIhP&yB_nqYXSv50qod(cXR6oaup2KPF`E;)>}oTR29L8pZ_ic z_;U6-D>lyu&$iZ=bauYI&W+9U0d#D@(Z4wVP+j0yKm7ev-b$BY2MiOjXl%ZFR(X*4 z*nD?|9_i3_J1aKNU)LqlQn=ncE=BmhUu)*Z=Gl0Ph|ROtR2KYGWrf1F;g1R1+}L~< zGnQIkFo@pW!wP+ZK_G^u9N3-Kh2*oa|n&vN-A1SzSsxQwKjBt%>28m9%X)wQ28fdycp5 zYj`In&yQrGd9?N<6bn|TEG^qE6`7*-nBiq7r&B&k+G!KsW|s~p#QK@F(eT8YX5v-Z zJU|wv>3Ncm!yncJB#p3YbQ+Y6ZfyR*bE8?QHM~zO+Klh<0DG%Ab*38KeB$r=e*@Xx zcmSZDgkR@0xP$RYv3{{GT!DE|;`4ZC6_18(kG#n~%)r$g`F1S`r?#r5ZN=K#Zj)}H zA^aM^FIk;pGiiV(Qh~*K2GXY)-E{F_zIV8P<#2&^)$_j8WWq|uy&txL<{@3;yFt8? z&Wj3Pa`cvNH6pI@6rtv*vX)EcSd_T-UQFA$z4$uLMe>px``6S-k+^aG)%P8I=$A)o z2idI4kEJ{~3&Zu7^w~7+vrWx)_%#>0AmvCXmN2#unoAJtTi~Hw#u6G0ChAGh0>P6< z&L#MTfwYZPA1d1)_g^c6xl;duQ(uZ#na3`<-ople{J$yZQD< zXHc$wmwNf+U#`Pn1})0Op1?Y+dIvcVBebsTrxuO2JlK6&>1h`*LhDWrf5V1#h8xyK z(m&q(2tMF-El{AOV6%NRKZ$mL^A7d!wf*_t_Cq_D4=39@t)7zI-T_+9vg~HF{YaQU zSGQT%zYJJo=^thV2@Dr~jpC=65p>&&5)YRC!DK&Zof4i^#2xhnU5Z%j*S5W3NryUK z6-kv89JVaD&oc>tyzoA>g4@xVND0M~qtBgsT|n1+7B1A5wdis&%~O{fSlX*A0e|9q zO!Bwwr&%DJ=UzPuglGU7C0+`V_%hQLym00Dqs#MaqZ^w)R4hOM?5xv8S`*WMm^Gny zZmkK#Q0TsAq=gY{P3Vn}-6Ekl$+B(xU|rE8<*%R>7=X#@d8{21SlV!Sed) z3qI0Ho2EgDniDJSqaotW*UhD}+wY^z*Uc`wMUgL-+iJ*L-0=uv$$XdXz>;I}i_}4; zMkP#-TYM!#A#9psh8_4Z$Jbb+ALdA1WRszbtn*X?7^A_8i)v?8vvu<_vz6?~hqDFH zpbm*P3!FyH2GU6?^;3bDBPrcWHaQu%5VnD`CLQ?Y)!wFk@v@@boXe(OwImrB7s7OJ zHn4uS2I5cs#niEMs1t+xQrlx!RjI*!so3$0LNA0C8OMfx864?z6eFNmnSn@xa3SvT zat_O+2t)aJUz1`4W;a(b0d3V2g)j5WMBzeY!c&F89n;q&4a2gNOPGLV22!-cy^x&A z@ef-uqz$JGhjJKG+9O(MmRqwkieRM_N3)bu3 zc^L(*7om#c%x(%XyvuD~!YqGiZHr;?r+!a}Mc^g{$29m#z+P{1lC`4}(82}8Fodtq ziPM~+4Kw!ni>xy@SV95q?*Z3RhwmPdb+qs2y;^4A*t3hHx#cIY8c}|rb%UH04~*`Q zaD&)8upg-KP4q^8)>T6Sdwc5z^OVqxmj)|*DTOmx`PSErABEOcZ2Pjc&_$|3oHQMU4BhC1KY{cQNz?UpVjxN7EA@rK*$59%*le2gQ3vXOeD)-s!m! zDj#@+nM#Ve8vKd7mWwciQB3g5p``zc62BaxQ<4#fv;(huJEUPP{(8XKDI}ufatKY( zdV!&^DWMDmI~H#mf`bpCxju@aNW3}_kH;zxdRpm{XZgO8!0X+kHS!4{$gX-S!>#j5ZGXRdMF-EWw{IPG}+O=djHg$Xxu#Tpx>ht zomsGV@o7lO9v~ zXA6tJ4#e45_Ogyq$6T^`pu(m_#~{PV7W;nrZWalwvBP;dOgQ_Wv*-rdLv?O<2G$k) zRfY~^pyuC(;GfPovP#^p0grg>GiVDN^8GU{KEz+Lp=>g>vR>>e6Yw_Z!Y40y8Z$O+ zt38;}K4bYPFPGHIbFLwkl9xhmxPvp88nhG0IbHp4prmB1qC_Ky;(+`V*vU}}J9|e^ z7QWctnIA1CiCJ3g{vzaNp$x5){5td@JxYPx+VZ*vv$G0-nUk4Cj#5Yj&2G*}=`%|K zS|(vxP9;H(Qb+`@HyaBdrI5IrD)ZOx$|`?i2)Ug^PMb<)VX8L@zagz$`K@e=A;?h* zsuoK<@4ao+7cTUx`79SE3hW0ZIezL`C3J(fuwFRSdeW^q9>LXGe>@!z%|m z^2BZ+PwX^)lWswD>(fnhq%uW0i2ouUJRrh^$s*3eCP#% zKkJ>Yvjx`Pzq7kbIL+Csxb!Amvi$YxA?w-fa+?s~B3xXX9bMmFsI~}!z~DwJX7ub0 z0s`xW@GfUJNx|aQ;E0h8?ET2aiXdli5YV}A0$#F5d`!&VAb#3s&|s7?7Zt>-5MmN~ z_67x#6H|=BUdxi>Y*2$heQwBCikuX>49A9Yj}T>O2sXP3b**Neky)pt%cSK!D|`6J zK3I4m9de*1P0S?IvETJ3{SdEpM|M&#QPCv24RJtJ8Y0_2dF4e@&x$=*=Tw3Ac5}_r zm158w2CqTIHyy3Q&O{$43A_If(ms^qD>6)k@6k=Ix2LfK=QKzF%!1Q?MMt6bso=2#9xAE?5hdB)So~xlK8X4ckhLspCKx8_fJQz5d0_c34*;|x$7ViO1{^yy zltlvntX}YkY`PAEC!Q_?kP~3=i}OEJW}i9`WN|>iN4GJ8{xs{|c#W3O)jS4)1GLd2 z@e3-wFP_V_Cb}G|R7JU0qjHdk%3kuxD#wS5a8Ioa;OTiOw5keTZlSPa!^W){pL1VG zhx}b!Y+A0mNTEZj>vPWS7tX@PS2ycFc}7S%udyT@cDTNRQP`=4Wh164!;nFLT^6?R zLI8%WQ5qtKJ;zzc#@?~r9MB~8-Tl0-{6$_^Mq8oxxBAgU+&y|HQ$Lbx7AgjW`ptLn zfUv|h%EZ&3reP{`b9 zYXnlNEN%$C6PZ{_cq_os5>$F^)^am!BCuwzi%cLxk|Tc1mQ}pkR_ksLwi0-+n2zp= zkie29+u-5DV^JPU`-R&55@W|hLTE))v2iAb1eDF;O;K>zTzQSb3s+!qmIw-LCKlo| zBi03I^?0u!0&$Lt7lSw>v$=4FxJ$f3C5w3w>AJ~qId$iu3VE=l3-|`r;vV{y-`wW8 z4F&@Goill}d!VL>vrp$CvbG{&HKgGd`vps^#cNbCKY6ZhS?}m%_iw}?9&)9`;^^N} zMMLC9Ui~vx65UHbyx@=C^ytd|2M*@jJBz&|Lp8RVExtGqkA&1{^FnK8)K$fP39o+M z=y{Q(P;_65Rq^UjyesWcjg;W|Of(YRV)m*SPD+n%GxWb#)W>_LY(e_=_NqEP}bw1j)_gZ{|>R=Pt%c@aM!Hp4F?nNtmV@B@;QQ$hz zj5kB+gRqKL2@G8Ps|YOm=;Nt8ks7q#n>7V%DR98R_ar!)tcC>LfJna#kQMNj0vb4z zqOsUk4aZu7@uHuY?659xwrG+mk57HKMSrW^?uaAcq!@Zcx9)Bfs{5(Qp6I9Q{^t0(i4c=oc^Ht&P0{TC>b2het5&TNCAS{Gh$o4coEzO0a-6ZP0ViScdU=4h^z;yAF`GU`K{3xeM>#?N_A2_G_u<14 zb;{j# zRK~;l^{*c?=5spaw$Bfqi%QpoR85tx`E$HQTX0Q&+9tEVd2lzR`kz1x8+t3ym zb4CThpz{oR_rqe7MIiM8sBYS5E$OleJLQ8$DQaiA* zbqy4N!xLe*M?_{iDpbKlFsCA$BlwXBs+28 zVo@9fMc}1j%veD?g6gUj=y(fVlZ@kjpceS8;w%TwZa1D>{jX|YVJKlS!Br|Pr$L2RGoXw_!L zut_YlK(v>>!0G+!M!vmI^z-zjhzpu+?Jos5zO7I6^GpV=*V1R%No-U~ADFF#^eT~f z8D{}9bR7g#KhI>mwl+T2&og;dhgv*~lT92oO*~YP*my_U5NSRUsD7TQ7^-diCyL%| z&>y7uRoOl`$%>qzYCpxKH=CtpnMEySzT>Utj6z2W2=Y0lzYw&s=>I2Etv9f`*~yWp zLi>59GBBcI7v+^yOCHLHS(;e~t4`{2Bv$*mx+Gp?j|Z$vBud7x1?@{>t)6lrN?tG}q$Q^u6HOUWKK1PMNX%97?2}-O6J|X-Hm|Bf z;$!ov9`o*YJ?3}U1PW!;3ej`hRePvx!PMtBkIcFpjvTreyRc5I#xZELr)V)OiU9jftS^Zey&t~XC@ z5z6$sX4tWLHXsqPdG?xWA_q2u(t8>smXxvi<3VyuRKbnScfqjK`g}XS2*{o#kQtlr z!s{yftN=uJ0bXpr`$h1%4eI_yu!-yc{j$rA&0C;4hCnoSW32iT*kQ-!Ybe8(eX0}7 ztX;^q=f>vE#@01}SoZ}$^Q1Z>Traq>c>|#<_$-UYe9nu_o1eFYUS)6uCA|QCaXAl~ z#N60?2P{XxBeB@NRmpB`bClqOc)^X$*XxzB_gS&|dc$sPerXmzH#T1%$+P(*F7KGw zyrG~<^e`-u8>AN>=eeMtS}2pHp0_h;YsWgu&nnd!7YY%Yo)Ti~1oV_xfw@ToaW!S% zwuoe;m%M4~Mz!SNWY^3g&yu$7rVhgU+gjr%=xtxaJ2836vW4c++LKT$Se>#=PrFoP ziq>O>mz|tW`6y|pP0gsQ($C;GFEk3|!5TZ`XowYO8AcwXD7EHksEkgkJ;rB@0q)*$81B z6j-chAbpzAO&1U5dx!g14i{)wJ?~3RCah%K`(Ybs9?~Vg8^kN=yr}R6sBS4BuJMu- zo1@BFE}3Itu06`dhuw4*?!}8~JLg|~i5ur%ecu^I-SisStjmw3Ja|;(emN2$QHJ2dPx4@Gw8H#E&n5ZW~(}*XJoJ;Tv197kQ%clkhNwfo;cc_Q2?a%kNAKJNmIN9E5^%T4cUeQVqZwtlj)c91E-E6iW3G?UbHVgZg z0c$M%!>k~IY59kX$3_EGTnP&BMTrop^D)^ETBn#*S5W3 zNryUK6-kvG9EyFONdV-9_aP*>=T5!O6I|$e&%%Y;vKCz~rg`de14|0WwUYZiCi&a; z(<~6qbFUr+LNwhQC0@#o_%hQLym00Dqs#MaqZ^w)R4hOMY%Y2`hiU#B}R^^tit*$8|TLrUn8EX>) z8WcGe1%7dX_MP=3sdpwA5w*llyE?DqR;^L4Y!Zc*fm<+d8~NN(#2 zx6N&Pe3fu&RKj!z-d7?N!lpT9*n!Vtcd0_b@hL?>V{9pZp78m!!1 z{Sr1?H!m|=$&P$DTL4=WVUdVoAQCYYkQos5aBiB1GxB=cjC(d4;52AXWPzU?ER&#~ zwhUz7{nqqlUG@akhubL=4~Y9m_fbH|#$Mv()!wFk@v@@boXe(OwImrZS8gHv_oo*A zX(*m3%8r*E#}hB-bd^Z6u4vh znxtV^c5(?5u*`UAwCA`G{^aGXA`+t;?YZES_Lz zWC)&}91+5pMisH;}3UPbM70x9Vf9m&ySOjiTa7=@*1nl)DCs{ix0hKR_p?r2C zF%04BbK*2-Xv2(s{vzwl4VF+q`+L9{K>4$MKv3V$d$r8Kv1b=YbIVU)HKP1L>)kjj z9vIyp;oY!zDD-;0(VunIkig#FdciF9HdHx*ye#8l22|K}<1h;`(lgxZ)4}qi9gXxG!50c8zowLqQN&@CZlK6Cb6jooSD>H^YPCJ)<*8 zH$%xeyWkOKDkShD~5TCXj9IfgwKsRO;USQ-ZT)4`#ggH7xhkB+Y92c#Zs(qjhh~)7~S27Po!% zIXRVawzIblqx25jn~=}QAz>At#YbBQWpoMSg#S7aXJgsRIz}CH$>xCyn-(2|jClHp z?`Dy}8atea!-TW{Ig4(PepBamXJB2yUuEb({CtiUKXO(`#*tOxb`5yM)t^CI*swi2 zA6)U5Y$%&dt*jTj$^@M4m}*fZJ&hTgw$&cYXrHlsl$T5DK*h%UfR8V0IRj{K@VoL!JSxzp_Ca2y&DHxoj9TyE)@hfZ%6jk)sq6f$Po2!bd42 z?xxE8d*kC9c;m(bn5VO_{x_BVlL#r?U-iR|k5Win)u9$)S0pa(wqq=S@4gI<4SJM9 z8W8ongQvNqt*9C?7QZSPGN9cK1zrYvlmdwpIt0~AQ!EZufR{WcGn!qsA;?h*B&@EB z8Lzk{KPN^hkUt-MZ3T18DKJ(M{p*t?kGCg~XN?hpGVk zL`?wKAW;flHmHn*d;uB|KJ{slbpM^|l0-n~z6p5A9`P|TdxQ9C zU)l;XL3AW>0|{};hO~HzG6j-z_eWtbNXT-i5uY0pvo{c+oD{kY$A)r`5M^k{#@&Rv zj66C;)wWB}>{->iM{Ag}4;EfXhaAXDrenYBP5NOxBN>`!Cedw(1ESIp0rbf$FPeH* z>|r{m>aulr9lBBsn#15VsQ9L%RoI#6+2C@Rom8n9X45;qgHIaLt>kcwHDDa=#wOx zhS9geMBk4h(5JRdkP0{|pZyfRQ}7oH$z~)Z(XXRIS4&?$NaAM$@gX#z(&Tm|hKwF) z^Hxub$>?r6Zxz{w3J5_kNOmpm-oPRM!BG5UAU;?Iwr&wW*?(*RSPRsF9Gm%A z8F1{-P~n5~gF-NP;^{H~IROU0IKL#1O9sS?;zRE7jS)1fS;s>}ITR4Tpwjyg(XwoI zKNzZmKg*#?Rg`-fxYmZ^I19f}<@j(BF0GXTJUuUkR#m~vEek~vq!8eczl)1a%T;I9 z3K;a6rxuotn5qmzA}%wdLS$hJci;6=P5hYB5HajI z&N?>s3c=V94K#^;cR#P=VHs_O-rp)x6LI(GolHeau34xU5b8JI!2`k)S493YK>?2*md(_m7VyM-Z7ToHMfn|}>jzBuTCc)0k*)JJ(B~wl#(50ZTSSQeYd{ndHYaqFXuUAKeh<#J z@4==xJ(~2s2ll#!-UR6Bhz}w_fwA6&%=4Fg2#|yy@hxf~7YdmhZH+)mh3Ayx+j!i7 zT-YtEjWbak|9nVCOHk>xnb0lmj=)+d#%hEu69ZBUWl{&Pw$-xwU@NQlis|T{2w5#z zvJD8h80c)=gNX}A2DoBLp1MDfLecqF7on-^L$qpm9UOL+D3mRQKBZ z?NE)B;Q3565?#tw#c&doi+C{;pCH<-c(7ST#G%BWpZv$azq1=Uqmx%4qZ04H72doy z!x9OKw(r`Ugm;9-_gaqGk}tkNb+8HSWz{IB;3nam&>J&)Cx`;qfo8lJN*{z(v`WAZ zGGC)?qTMB64Op3;0TQwYO z3C4?lo$SD-REq(bD;`ZUC!6`N;)2 zw0Lxb`Yj&$wFfLz65QU&w~7>bdWWj;APf@$Yrox@6$2<3zf#Aon+ zbeuE54RnL5CxLM*&E)!9wr*}qTLuwRD|1fKPHD?N#Wz>c^!P9?WHY0Ko$bZ;g&WW;C_?)R>qEf#aCSQV(3>-n-!5)h?C$RDADSOi|G2k!(fW2}3=*-}+csZPxkQZ!txARMQCSS5#)Ezh z_vCn}Q0SK2nkC|t7=sV_)>-}L%ns#MayU96*3YcLF#BPec*(iZMTPw30ou~jLzKmA z`V~dM$mHon$B%<`SiW7qRmTkioL@iGH5Rc!daLoe_Zb^GsgVp%%~L zWD`euwc4vcS-~{>h3XS)`;Y49nTny>wts@-FEGap^DI!Zda~$((ys5ftJ*M7KC4sx zJOfy^lzH5Cj+NtrCV`=yXA)5TJOi-0*~u)0)Eoir=Na0rYy7!kf)Y}PEB5^y=WF-v zWgyFkS(=$aOOl#zK0js&fgTT7mq?V1VGEww$BtgKHt5% zh>y*)t@PADMfY5Nb~-mU-(AzXfm{V+!fB~g1TuyV&yLOW!LzOPB>}>>*SWEIK7fu5 z=(k+Dz%f?%`_(pkn;2IFqg~V6)@k~+R49a!O`sqpaD^S4=R@1t?X1{5e_e-ayx2T{ z`BJg@?vDK$I6F4qT{d_-Ma1UWYpRJHPn8vlGO_tC2n=jhS#I@<8vCH7P z*6&q<*QBY#in`WNhAsP4N1R!^kZsS6&6|y_Yanb}hg&@qbmE!Q`A$4;Y~HNl3O>uC zF`x5d^XBI*q3^fy3$RXtozIQUcffKi{YWgfw@b1c(;OwphwRvVy6DL>cG}c;*Q#tOAS=(Lg(?5> zarnbpp!jM{CXKLabQ+Y6ZfyR*bEDO%r;YgXKCx({0UL}@iuH?i60Dr=bdSdXZxyG` zRCenT=$$-(0<)9w>zw#J-dV+?VcR2bvJW$GHAlW(GyJKoswv)Fd)sZY-S!ZE4eH7i zc17u$RA8~5f%IucH(flK?;Y-6Ib5J!^}H`NnXr;^?}u%m4boX?mKbhOt;F=ALL5*@ zT;nAvHb<4UT(UH<#I^Tg+Rp98*ReLq+XlM@MdHT!SKoJ-Q6&f2tjmw3JU9!(^_TS7 zH0`sTD-fa&EMaURG?yUOx4@Gw8B5StAS>?DZd40|k;mLi@CyTRul3CB&aW>HX5NY0 z`v7FJdktD*LDvDUjd}N_%h=y81{qKDdAZ~+)+>e$z;E_?F~yh)bXlFs$}79pJx&P zdEtEs3GTU5uk!>Ky56&Jp|-3=mnG4wto<7T;r%zimIw0^vOO>QNwY z0cccqDJBMhFEeey3s;^$x;(!&y0Q5~#R3Gt&N^MBH8K5%SrdBa)|&AB8MeEG^k+m` z6MEy;ny~aW!fAq-G3P|AHPP%`-kQ)LMOqX3vq)<~f3_LzN6|vGDs|c@y>x1<307v3 zT7o?>O|~bdvGzm_dNF$fS_QLm8EX>) z8WcGe1)G67=pt5|e0paX zqY>ppYPN1(X10%g6hg6eBRZxq=C3t4cHYGID$i zNxXCk^fgJtuq|iK&JbH8;4KhL{53a1 z3EmS>G#qs;HHii5_3yll0-cGnSw(SXH-*^B9DY{i7ql^5>;Idp|Br7?|Eb>-V$r8b z!7&ZK60p~soMi2&1XL9e;~l$F5{Y35U!N1FIYS#}?DH2{XKt{B0@~jL&agP2jv%P- z=e=5H;MlW^qq*fLuo_W*p!IH?6%UN=kMM5TI~01o-ssP|YDi#j-}izQ1GDHW235m- z&G=DhUB$L9TMJ#JD#S^lmt|ZG?+Ilwf9M}YgCfR#*^;nZ4)WluuAc|#5ss!OK1x+5 zmp#(n3=fL;jLsz83?=96f=8ICq?oJ0pSas``wqey2x26cLrMQNB7QkUrz9hOW(QvP zc1Xio{PlpdQ}B8yhtP!k0T>FK63RfZWAUaT*u4Mcun@W{#H$1Gc&zfErwmhUSG zyxu)pGq04d?Nx;_VTf5~((e!PY0E+0)CdX8{Kdwuv(~g5A#h<7pB{=wQ(4X%`isp0 zj-dEM$p`7q2G$QQ-yRVVpMNTK?_-z}3^U&QO6NT_Nwbhs9q9;%qE?S;wejF4;U#Vbh{xkP&}}itlETz#2Q8heJMZ=?g^sIg4(PepBam zXJB2yUuEb(hGKjd2Kc8lj;s>5YrrF}{tVi}hAB|~B^$~nQ!DGmu8tpY@HA#@+E#lo zqkYEmQC=>om*-qVDyJ=l+;9hHFtr-_Ti?PNt0>XPp*SEv1$J_j!p`0il!Y(0cjiZn zNn(~3yP*oXSt#-9FgvRNn7(!(OYZGJJE2(!a+E?MXm)c( zVpo3e81fV{yFlgmk4|oh6NAypc+Jby667caa@jE5db6?cQ3{E>)l~@T*-{&L7FRyQ z338M|BBXGC)ek#9N+EGohgyVPk+`_qjxisry^Yg0=urx3K-BLJp5~IaqH4tWBx;hR zE@+KNJQgfNE(~5qN`hd8pgjLB+*EZ-mhN1_4f}fT%edRXS~cclgC3qQ4#o5biYeP6@#uT zJ!bR6+0kO}@XCRXJh6M_6FZIHq}763w@x-Khz8IH%o4YdvNsS=+mHJTi7hRap$6`K zSN0B%ih<;7I(+pGG$4HF1sj;r+WU8QcL}FCdli@7RS^^$w$J7p7)1}h!o{`O(e?d> zYN-$isB8#w_6C6w>xDqVi`nP*UL7~Z&XRy&{Bwhzy+J_dz6p5A9`P|TdxQ9CU)qQp zJVd@=U^15!^z01^BqydAg}s6z5W<-F+>oynIVp4*jt%7=AB> zdANhVPHfz@CN~=u>X<~E7#wpgHuwk>KZt;#P2I!`TaiO)`p2M&fA|(&aYQt1LyiP! zh4w(#w6^HW&Uovw_wwiuvJ*p1Xm+rA9@^%xvDX}x&jUc97AJf(iZ4+;8xZteobSCb zT9f}JQnit96dWlBiy*6dxeQqoqr59CpL(x&J%X%i+uEK;a6rxuotn5qmz2AywN*uo0|dZ{LU zOlgQ1_8ey&8+-j+T1FZKG>Lt8Kd&o)kr$THR_Oh$A~g|rkKV~tq~w}~iUFa1^Bp`O zXmJhrGu7NQ5F|(xp6$&qgTo&ZNHT!-PloJGZdUd1)j7!Le~$N@U>orXChnYYoO0bPL1;ztQ-13 zEE~@vZcgYV(UR{$h~dX)+s9&4oE}Yj9}9a^_Jbjyrz1Xy00pMpyO0SV<8AmBHINI1 z%#F52Af?LUhL*2#Ngn?&AWxKZv;>u2o6p<~n+U8GgF7TfQVV5L2d}o(x|?OPgmxT1 z*b45wVmi7fLU4PFyQPN@53+bH?H6i$%99=b2*QL|V7Xwr0xY0x4sVKr!{*9s3|_bb zi?akGHHMdS(GoAj0BvMZ#)WkXuX~tjZLxQN{e^xdLU<3CK$*0Q@agG(;cyV^^iM`q0kv&o2%Z zd(SPflIULg;RS#6rrq*qcAZ~KU-89(cqF7on-^L$qpm9UOL+D3mOZ=a)uDJ-+MyaL z!Sk7DB)XKVis6LMdC<@IVt-llCx|vH9&A<-aVYUemMPn>HxbH}8I^bkuJGoy8J0*; zw0+mTlkkqF@%?~}fAI~fgH2#Bt428mH?1J@&k4OTqj!R!c^zoRo1yeUSVgM@>_BtJ z5(bMt`gkf&qz0|`W=+AGp%43Rm8FPDa5Oc=3C|mmeix!@9F2MgfnIA!Ehit+zIb7^MKe-@> z7LRUFzr`cJ_JD}_&Bbas`);X*S+fYu(Kzs)8 zN5?q>Wa-_Y>PaA>p{aPiW$Wg)v}NE-xTjH+cZzmOTlOiwxhk-%<}fZsCodcw>})T# zFYH1;7j(#|w?NHmSVvaxM-kdrSRVq;hqKe^pSHxXILB?qY&74nw{XiHBI zQ5Li5Qxp{QymOQj^lz`C?|C0S98t%dKHd@y?;{j#Brheh`XOWX^$u=FrE5Z}W?Z`F&+!&*!OeRT+Mr~;D5yE< zo6t7;VqsERbC-(H(g%>A&Iyf_KU@~T-n7nT$cF7C3FIu*gOKoC&mzR==6s634F%TI zBu6^5w_v$cLudYE9NX>pQ3Tb!WEO7>#rdM#UE7r9X^M7N$-uYWb@Z694Q+ujXH*aj zI?te&x-vWm52KZYRs>QvVo8i*5^#62%p>cace2x>5muFD=wOVq2sZIRXD@XHv^&In zH$iT4vlODS+d5{r(`JIl2_>yI6Z!tf0PP`5EUS{Wx|*Ujk%HOKCa^X;n<&@-UbfW7 zDzyU})7L;Skc`{_1(4v$zn4)Vpb92}IThI)!H+~xg>{(?Ak@;a6_>;}dQi!WRF@uZ ztDL|Yr8gdPSqDCG1w0aqM!XBKNr*GWOT(DA&`D>-SgS9t27WW_GXj^#lHLmX0LxWb zzmH9#r%7{-xV)Qo4=K!KSR%9G!@Bt0fT!(WS|nKgsSIm2D`xY9*gnb8s?FBcZRB}3 zx`mnr!1T>wez_*m&(o74E@*bqrj=VrIE<6;RdXbyu3I_;sD7Tw!1Y@CEIY}ZJIp1@ z2PWFBNRB8+VK3-rGBTWt2RY@KeZ`cq)!7`_wba|}4qysXR2tWm5vxe0nlhq9_3ZRW z%vCI(fP4#-@Tt<&7VvBvog151!x7?R^MN#*WRG(9C^0s?;Kt^=8(23`&{DVVdCHhg zjUaT7gqX9^pqI1PS+V)m7^IZ`7 z?H$35&3Biz)cO)MUy|95Bm%Hw^IdpdTb>nw=q|vE&3C^DzDGfOz6d>V{lE7nV8`Yy zP#r@c8oOohABN0uWAinXVaq<%{bJT;o9{uRi5;6a8(Y^vzkO||;)QHUL~fJaG3L1J z*u43~6?~ROV?O7_=FQJrLSI|?t$|a3PjI2`BZT8*V8`Y=U^xOFiA5vc1#l38mK~d~ zK`UeLvtsl0hR0&@b7S-Mkvy9};_{A(%{$CguqD;Uc|O0V7RqF)=j}||+OgYp?U>j) z0kvZ*FcQfr93u>1aW!S%wuoe;_q}QBMz!zYWcPYqCA;({ZM&({?mMPExy9H&+d1(X z-igWcBUxx3tvw0Fg4HQY%eG5Jrf5B8c-hJ6l#h~j+SGT~s*Ewq8Z#&_Tao|xIQ(HP zP<*u}lSWuIIt|K3H#UFZxzViD8r~-sZ8Ttm@kz0MQOSm?sJq82Cwr?nb*38KeB$r= ze*@W7c>o1wC*jvQ@p-(nibuoNuQ}O=8MvAw->wDW)K=9LZ?3)VHt7Z$!mk1Rl9emA zY|y^AHL1X2Jp<{}jBdJkFyA}ezjC-hyXtvgYBFIZ~7KZCD>9c9tXE|3OjD%tdV+*0V1hKvao^;7rLZiV%JqcPMj6CLEf?pVjd#z_~ zcYb|wF!N5_-Uo1mWnP024PjxhbDr)+7)P}#33P+}%**hPc>hZe_l>@P%hvqpxfklc zJTpaG#(JN&XhMeq(LX*rJwEL{sCS6BoO!eTptd)?yzb?-G~fQ{3`*SZQZJwU%XRq6 zphcP36Ih2;?;z)4gw}Qa)S~H=4|XrX0jYMv2(3Fg{0$q{8E#k`N&k5BBlv*VwLpQ= zA|L^+;*)3xIPXvoU)!JWK|`_2hm-A{R!_-p?*J`lS$4D8ek9DFtJ^H>Uk0qP^bfOw z1V;6XlR$wE5sk`5?UNkJWj|=0VrCU#?Dr{qCK1Vg|H)*(w(Si|I@Iy1NUCr-uY;}@ z)ncD#5&(JOeFzEexl^z61Q)vAvv8rdtVNe4=zu$(y4=9h;vfNk;(JW;x9z7{Ae`r3 zJqmn|#B>FQ>##$44 zNi! zXu#9w&UdQB+Y?jEo(Q-|#tAMLiTg-=s47SZ=E!AG7$X@!ly% zDbAmuO0)Py>L6325~jy3z7nAjHq9}^4*ZzoQwqN3>S)jArmKq(UOedk)-(w;fu8lQ z2E8iBXs~j3wX>?(x_Oz|N_OPK*#g+22#Z7v1JX2ov z2q&0d0Fr=zGZ4=dC#ltvdvqUut^_b$e5>;}?bQhoy1aPelR2 zkv>N;0*aLxh*a~HVTmPW4GSnA?`u+w!0hGtGsHwBD;*owKk!YRX{9LAJV8Vk*GYtS7VYv*6L{^i7Abc0MISs$-K zCXJO!Gmz;reHnk=>el73?6l#8YY+K;$0FA$)yKoaPK|n6b}aWSzOe5(;R44>+SbvLTj&`hMQ4 zWd@EtyEvL#egdlz7;dh`j^*fePMI;*I{StA+&j_SOrAsJF$yEcrUh z%D28|{3x`pV%wLkg)UMR;-nx!L(k5*7*WvYQ2e2P6b*_P_hn1Mu95CyP==m5=@E{m zCq7D5Czo{s;)5`aA8GH92gQ3vXOixa{yDqg5oRhW=4x2_#NBeByBCZ2Pc$77WTJ*`w0 z!o$6|ef2px=WMpKw++Lo4)s-$ol5yi0p+~1 zWiRU(b<8E32P$k@bPO`ggUG#T#CNkuV2vHl!(qbN|C~iP$R4V5yECw^;IA@tAOkhu zg#rHQj3cYW?Hcfit3QLbuwlxq*uP{$*<@;Ez1Y?90}h_Xj7{5W4`#H_SU$?jCH3;0 zYe=Q!rH~u$;0&f#BY*2#IAawh8aWgPP!P8vQRt$_%NL^57TJ^QMD~n~wg~7{6Nf7!% zQ2C^1W>;<4mwqqa1UX89gw=I1N1}=tr9f}joulVIq(KLB0i@%e<$kP^qZDW`9lvrU zR!iP8{n92tX~3PR2z)BK-zB1oLD!WYv-#odXt8&AtpxP1*s=7&iQ2p@XE24=MO{+-=j z!fDQ4#ie&u1jUB!v+;ME5D?i|T$>$T-(M&_DG*TEpl5Fo80mW6ma`{Vq zZpc@PoD{kY$A)r`5M^ixHcKbeZAkLA?Sf{{${s%S6b|G-UNRl~U2oD4@mhCeCzC{% zIefz~jW{4G4Uz4iyz-){XT{#4bE@gxc5}Ou&Hw?;VelGMeACe?>`e4=LcY*xdw2yU z@_~Yj?G{YK9az7i#yYp(dEZ%EttZN21@v`dje zo4Sb?wjzhp^p8PP3KT~~g9YSBfL3S^bWLlEzU+**9^*GwilHVnJ6JsrZFAVzYmUn2 z0U%Hc_P#{*Y(UU=alZG$Xiff?NYzHZQE;RPbp%<}%M~K)$_lq%*6UQoHv*d4w)X3= z$=V#XV(T3e+bML7K%XShG>pC#rd0eW0)1-R1gU_d^4U+}I|YBSkZeXmqHo8jOJte? z$v%c=N|uAp4JvbQ*L)9muhnkCg$(4hwD_yGT3scQWfRikaM&_m(GC9BABdle7Fcc zYGnXV&r6|IRq%2%Pe;~m5OB!XG!fvCzl)1a%TM#~`psV4%u84kpNO;pMCR1c_1eA%0Dw*m*iOfJLE{`F065prXKR%Wm zL1bddTs0Yh1wnxH0mh}^_{gJ7u1&($v~X0~7miM({wY;vZ5^etD@{o`8$X1r$y!~B zjARC)7M$(qS#NnmdZF(KT|Z#L*LoG6iEN#(fj%cWHPK_RE@Rnv7IAYzCyADP56-r? z{H8cPn)KfC_PU{)-ZC~r!K`)J(-9v;fC6K^3;g%YSFXmH!1q5a;#<@}E)+61+8TkB zDvKL}??fh+65a}Mv;>u2n+e_0?g*^4_d+f!1m6wNLYdUTt8KOJ_FyZk_loK0o(Ng( zEyRuxR`BrQK^Bjt{X*?r2>K zdj%1Qb5y(-#2J~*JaqU?dWA|B^B~f7lS?kAhJNQv-s~QzDdOzYd5Em7NLWi3W4=Ze z^ONTalu0Kb&#)T1c=5MX(GY!PPDykx{qTZ6dehAE=PwS#BOx`~ywI8%bycxn!mFP* zqL#UPqle-%1M%umyesWcjg;W|Of(W*%2ma1Qd*KeL9|)%V6%#dLy12>`Hz2pXO}c| z;?Ms}^(hN)UYlWw1V!6-{ntr&$I0~sf&T{8!6vYmRilh|CkuQHHe$V_k(Dtr8fxxOqkE(xQ((p3DYla%6gM))cIzzyYO*NpLhZ#R<riGKf98LH!nwe8&TJt2hDtXL9xF2}wio65iU_JD@eod~$dM)6T;> zhm|N1);T_d_oL&S0dAlhR6PmR#ijeaF7FiWl(uaB>zk_r%W4keVs!Gt(ZSC4V*A1_ zbe}?pjCu>ytcG=D^?nqgeTDTQ;Cwhco&M<*h{oAYPd)|i+N7ON=kGDj_dkKQ4@n2!xSiOA+(&cE=OEq-nPsXvEejjaeX9?_vvbxq)?yY!ZD9#t< z?%Jj-PgAtJN(R1F*U{s~HnauCoKZnAtUQAvT;YW;-L%n~H@bG;pr4hnKqlOsEc3{E z=bh}dXoOW|8QNF{MUan2OanQj4KW~Z3aEmK zU`|CgNAM#NRAqCCk@Nz{#N|Acpo3&a$i{$}%R0a(u7F2ku_z8g(2AFaF>j%h&Wf>C zUtA6RX4q#0F1zWiAa15f!yNA3l?X?+^|9*k(BfE*31GP@>-Vuq^fYO%5tnz{4%jxJ63$j?U)+Ci{);h`gsPv?~w;><1dq1Jd2Y}9Ocz&ul^D! z?ll_ZY}*fUta~O9wQc`I!4=7Vp6RT?>E~%GF$&N@db3$tmcg>6%y;53X=Zi-7?|xp z*QtJ29+=tFP{m~za%V#?U+EK}9I6Uy&luwhx< z`6fHnlo91q&rXlTToung1#NV0Z2sf(-K&e^i_LepD`uy2WAok3dn}VqOC?LOBjV@W zN#!1+b&o_0=wQd@yK9BPv#s@iB!JSWmqU(OQC>cPy%swwHqYO$w%Ji5JT=_Z-P|(6 zBV7aR*nD@xF#AB}aNUY&xm4rD=K04i6`Sww*dIq@$L71s2Js&en`f`7EcmB5E`f+8 zr464QMeOd!z&1BF-(A*H>ye;b5fSzj_9EeWd3Um6^IdpdMV}Rb=q|vE&3C^DGBiKFpi*gbZc;l}1`D8rV0s_WmZUC6fQ#^%k&)-@0^qcDTM z0NmaU5f0q5WAg?=SMXUDjrp7xn>Rmi3B6BT-eD%g5}8vO*2TwpnI=yyl*v-h+nKbrW4G(E zNMh>*j73_3kw}BCy|GFrkY7#Nw=E(W>3wh7x>4;rIJv}aH+898$MPj18~bNFCtkxl zF?r^_h33)PlTa*JowBrSyHsR~)?6wv_!yne*<#wwJ zPKfn0Yopu$0SGjKIWzFiB#sjaG+OSbm5+hn`#A^aND zRV(a@(lx2TVm$-t(~NGqcrf2P+`n?TK)dRBUurU8CF9-?+dz+5*p-NFFnms43Ktc= zhWD(tG!xf&T6%L-S<5Bsj7nU4FQ)C>UVM~4(Z6);3sEF)++KbD1`4@)@(y_&ith}g zZhDPu*5$`i9y}`Y@&Vn?rfHw$T!H_KDU9t=%_WHSE%2mEhEW;~ChAGhG-?GZ`x5-Z zK-_CRbG!5Fi-VbW;`Tm(BP{b8glGs0gPrqqFGAK8+ET4b0=?K}_(#0|rHA`Q-@j#R ze)QZ6^+Pg^vhLxJcYpPiN$_fvkq^S$kdb}k=Iws%@RCA+-?w47zx&1Ughk3?ANxvVM&KNUKL4|F;iN*sM9e4 zvClIJfV}WNgar58sn>ae3tjJ7xKLZxqRSF=z#UIrZeYn_J@Gvz`Jo$59V=piaGrbh zC=j@GZ*+GhGA-S|%(MkBTzUTJ^8DK9#^w(d3lQ1TjLDVg=_0L(scTKJyb747C041| zJGa(^uhiRC1L?hqv?lb%$8SwEJD0a6G)TPi z@`%RS6AgIglFN0f#M=|oe=>Vwnru%@W9^9=^kVh|vRt)n+dznN$yKOF&-F_c!zHWBeEsA`x+*U)*|)ryQk{a)>K$kFOF=jY^2U zrd@@Qc4vJhLLqFLV}>30EOwVF6dYgkD1E$0U1ZbMMG(YT1B&RKVT=YVcUQlJ&DPD! z%vQ1^AI=s)i8>i{7zTn4LxbN=xI#c?K-9yzX&%nV>uEFY*=&H*r*>lz?5>-#Ao*#_ zK=$2lO<&f;391jb!y_IL_mA$QhDUUr3CM=U5`cfz-@OXi-bzn!YmBU^s)fKsJx*3I zhxR!AW+0v^PExBy*V2F#kmK9EjAfdUiXFeGOb#xC$_U~K2#)kQ ziV;w(%)lafyI2ZgA#Av5gtC0RuSqcivzsfJfVQeMgEJ@%iDXVCUunt3vz2A}cwZtm zhGi$0FagWO;p@%27BK!{E9UMBrwk`#YFn6;s~1WdhIE;}j6ZK{>vHJ$exqlD)Kq?|Iuiw3UEy6gnu7<@j1n^Jg+Zr#oNbi?-h;B;J7h!Ba{#= zSu`AVEj5V+>-F!vjDpsSP(^X(v9$O4Q@S zn4A?4jP8$cnAm3ty>sG?{;aEp1oooV3+AeiFg@aiIv@7NGdIXYOv`^5|HzfXgsgn) zYsQa4>ngT=*;?o#RUu9azG$UTg=?(ACq(?Ae-sUh824pM!fx+P6?75yitk~P%;!YC6s|+$Kp*xu(wmhUSGyxu)p zGq04d?Nx;_QF=b<5AkWsLEh8|3C#S(#;&v0w1*V1r~iI>C>~8^xeM+z+0nmx|J0jk z+&u80-=hjQehAF`?~J0q<9SQ`Ig4(PepBamXJEY^N}Qns zrMNJ_Kb>)8mAG949`V>`&=xi{wB_IVf60cj$<)evv8&?;96XH~o3_;+%xIsne3X|< z>g74tkjiOGAvfH?8B7h@3FMrv{x?ukGFDNdkwbAnehTd5D21K9BPa`BZ12pE7L&v* zEp~qqaH@gH)3*dUN+A(6yE)@h z&dtxrB1b7C0@s_3g^yB5+)eqd?bNfr;PwWdxK)4*z@}}|7ClNK5mLCn>W3X4rI5I) zLoLFtNL;*W$JkF0t+4VN1n;5x-NDmb(pC(NQb=8pGfH8@JYM;%PLEO` zL9jwl)`i~BA4tc>d>H&Z3N-I)gC3r4tk3VU=h!!=%uM3 zM=8)?I)3FytQNRjdb87vkV&NhccLQjsp!7fFu$(!n9UDoM~l6~D+fCA#BQHY>@wyM>hF$sNDv0eNi9moOadB;SbbWuJdM1#wH;5(Z z7a|A(@;GEmNN@0RgPy%XK zDC`-{CmYltP@fwT9ZwLToD{kY$A)r`5M^k{#@&RvR`1Se&bc9Zt8EuFQ-d?Q)RF*xQ@ z@b-fU7~poOO<26J6*-iqe+-&Zpg1BLwjoCXv_gBJYg$|MWoNwg*n4^O2ib|CrgS`? zhqgIv>@`Q_^8gSi1$$qjdNv^FyExx_VYDXyOQdQe-zYdz4i-UH^>T&Cy0XIUm-TuC zS=F|YLeU(xV(T3e+vb_!M#eTo1o|Y2reXB0Fs0&05$IFfCP)PwmCt?(-zn@D3(00A zB>HxYxAwR!q*wS zPh22qqb_qm$9ZKT-3kHyf=cg;h7RuB`ks2a44b(`D9XJdFKL7L*6_?hwybh|xClOK zWdKjlOQBU&@NzRxM-Hy~X(Ao+cX6>vyXpdt|AzL6zb*?h&sPUGc3sYN@{EvjUhgX% zcDNRgQP`=4Wh164!;nFLT^6=*UtKTN#E&Tr5yPJ2tYc&E*mxb9#J;PE}(w%9Xuc`am_flCu(IY2gTQigmG>$nIelL zpiD$m$y5(YWClucc?{D1`;_~~$C4w6Oe`5O%m6G10<_X#TndhlJj&$SBy3F!M_pgI z@5MG~u>QZuA=ToiRGGDPl*XjzBuTCc)0 zk*)JJ(C6fL#(50ZTSUlMHl9V?oX|<4^}_VpD|xnk4>rZ=(WLi1u-7g0CO}U|d=LQ& zjP)*9IXNNel~*AsSWrMR+nZm8Snp_MoC$pY!y>*#4dg;0bEB;hNU5^8A^1*YVkzOR z07pwu>9v{AE$xoLS}RYvtPtWnsf9ABgIC*XS$(jT)qBNsbWen=_7-ACC~)B6!-Fgy zOZ$b|Wi_vHE<#)Z7Em^aH$}l=bLBM#FI<7eSt2N~xoAm{SkUV6UO@!n92GAHaYkly z;S37xD^#+W2a&FuTyi-z^gCzrX7@l%5oe#yLu745!Wt~CG<4%uYKqsWVt(>m-Ll@% z$?o5XK|JJ2iN(>srHY2gjlBA2tR%XZet5wjz3I`F{SO?>w|5qMM}{!mfZw|K;y^qS zQlrfat(j3*75gQ;`g!Z9a=|=Km=QKBZ?NE)B;Q35665V3< zsu)fTO}N4mls)kYqRom2n^i;{O8oiBfBgG9ZLNU_WC_Tq#5-_>H?PgGM1rF2yWTYk z?+A_W+1S>x0Pzi~gH2#Bt428mHwov2-k8xlwE-Nw8A>07RkTXLj-7WbVX)|7V%Mtu2KDwHB7!O_$dC+G%5`elHufVUJ9fio!@i*40#tR)yP`jQ(Gk{#CN z%@$2EKJ8t#jb2gq0`})J4Q+@P2fh zGr$dWLtX4Z11ZyKQDvw5^sZ)|qMg#l*(pGef%xVsnjRm)yzL9Dj4$;R^>ldzWT%oc-+b z{O}S?l^Mu_m1h?p+1bIqC)=i0Bz~%A#((=ofhiVO~H?WDgx5ySg~IK4~J2-Qbaz z%6M46{`Et~?CTxej!M^rRL!__&7b2f+Jc)~i4z{I(kG>{8t(ln`eL-u#SSeTU7 z+@&P}$WP~lM#>*93!sxg=%;lqLpHja+8~tCwjN0DEJ6&pZA_<+zYPV}(7s!OWlVu)R@4SD zO^}=1EQM(7wvHL@Jf7fj`X3@$?T+;iC6-mmT3t=iN;_q?YzeGgDAfpB>SLAKfsL(e zz>@b+{1{6&Du4u6(ETYG0;*slm{XC>5&TF5RoPr(Si1m<;=fVPSiCfhc?+F%R*bd!;%eYG!#*P*>87`8ZAntdbZdyrWwk%9TRnh2 zfHhzsj?^T2nl#sl%e!g!kitwY$;R1fufH4cwEatq1gpPuDTbWo!(yl2apJ(!_4z+j|C!09RtEH(kC_$ZG6NKvLnTny>wts@- zFW{mDC|NyObV1qAGgW^xr|vgQT2fN|JOfy^l=)sVmQ><95PC{H!?{59^9;c1W+$^0 zQgckS#yzxM*Z6Z?=Oq{eSL|6tq}%~iKhF@DPU>LSnE9-0iM`<)euyBMiM$##IZ&{f*-;Xag-`%dq{O(#RK~I3-#^$@5TQ`uaU`)EP zdCF2xb^dox0n}>}@=k7v69cFsUGrNjabxq{g~Q<4*7}kFDIc5X1L)X*iLrV9eznbR zX#~CM6h4Ea2r;?_zu?B^`Ovm@J1aKNU)P};FE-C#zKPAZZr~!Q?j9)hW5kluhCgm>b7S*eFf6scU{t-k2XoYhGk3n9ySTCW?pa(#pA~@UF2IY; zcfSa}M?u}ch?L^`e?NA)v3Uzr#}J6d?y<`ZH#T2G8Mf?GeRgK;Lbg3OHg7hzt^v=S z?y94MaBvj4LJFY8wfNbwc>|#<_$-UYe9nu_o1eFYzTe6(fFoSagC;RIHs1lu5%5SX z8u2awa1b_>uBQ}*o?k)W%NNBTTBZQZEOGdO7- zCvDqJopw(EgKTS!rC1fO;hmT~W!XaWXzfWT7OYNLaWcD9WQx{fhL@e3PWdQlr%io# ztxDrn){;RP*^2zf$KelafqK3*0ZAjQ8l47ZqZ^w)@Z4y1>S-gsyiY9JjPLRI`mN&B znQCNEzN4{M% z{Hd*~=|-{kw%epnXb8Us@Jm*%STYdAIw-JM&p`S#qnj=s%=ZrWuN*GWu6o{=noL;9 zxc9?0&^)9|d^dXG@Gu#R#L9B0qCtWg@ zpszqy+@;;976^S(=St&=Y4>gvUl@pct!Hj`etmH;^G@8}2f!xFyapi}!opzZJl%^h zj%rmB=mz<$r--~2#UJthmmcmLegBrN`O$MP)PH$qinfgPK5fy24h5ore0EyCT=VUZ z&aRJ!cMS~4r@hzr4)K;VZ?<3G7G%k4dLgKEziUP839Q4acaZZiLhHJIYSGZkgWac< z%v}^PLhDWrf5V1#h8xyK(m&q(2tMF-El{B3AW(b~?EvQ;>fvkq^S$kdb}k=Iws%@R zCA+-?w47zx&1U#WYV{ZeWR)+ztX#Tkyh_=Z`MWuZ?bO z{!p<10kFAT**R1TP^>jE{fAi-dgo)eCiKR~Z%s5im$xP~NX+UYtvTz@BCQGiS&l;m zblNDrbZV>#R%Vjr5sh*&!AQP3A%dmJb`6JC2py+Nygf1fC$lG}$@auF)}E+Am+ztP zD&g63t)>lrE&;8AS-Fh02>}g?9E*bG_0t!8q?Ia?v)krU z+3oky=Idsc-J-}B%WXB}MsVY0zRRd!$+7rF>L6325_XhnGDHEr&6-@hPH;nfB|;%= znq!6?_%X-VSR>CItBY(hbP+4Xgx(p(Xhivtnys6cnXP0;KAbIp26adzVi@p=4+Uff zL_M6F=HZOIo;Ks2%?3F8?MsLa$_6D*TL!Z4erx)&CQeX&xE&txfVh8jA2mFp>r6m4 zES3QLtN!j)$o5uxLNEombG?v0p>ZLcU?%^{iulbyJX4&cR*SBsonJsrF3&pu|EYT) zC`*s)F0kM0R%=um$;uOnn`b#D4z`qMOFjL%r+cQQk>s`|OR*#?p0RVz?n2X2KUp2k z^o*xl)=Yr2b`q0?1hc?7tl2;|5JK2p!kUD$A>ak}I5-Q0e+yYK5R;f>H=Drn$0j*m zl6b4W@2je>s_wmY>(6V=<{Zbe-S4Zqe}4B?-MW8%ekgWp6X_+DhN-~wK?O>)Bo-(JNO3X)rGai=`@7ky)tYxzVnUE7_^*Z{)}$Dq+0PZsVB5I1f|H5D zmDmTT5XLo2hGSw)(jYE-xr7;9HYVl7jF^z)A9hk~7o9Skk>TxdQhmKr(y&aIiDmo+ zOFPEFIVn=fnNT-?Dva>q4>a|XL&K9)wGp4!)+?D!;|dyMRlN$UV5rBW=9T}e*155~ zl|X0w_Fl={tjr0DW+97)lgv_=Sg_pwJ1?U^$J5uxNt)SDA+B62Y&a~w6pw^B0`5|9 zY=f^1?nRT6oD~~@HU-3`y}c~eatL26h|}%CJTu1PMb4f(E`fx`d%)#HMI=z4kb8X%cy&t;Y9sM<;$cK8glpjPKd9xNEFXJrI$qBqPGf z@`OjJ>g2Mg4EP@vkBqjnz+ZFDu8asX4=L6(gcEm*PPg?~5;r5_3qx>9G7ymV==H5b z9@o;%1FTM@d8o(G4ErDz3YrqiV6b=N&10~&5cRk~nut#g#B-^K2R*G+brX?xasSm9 z)ts~4gTs9}oa#uMisDqtPiE+Pe{0>mQvPXgT$vLpR(|PVza3g7!Tri?c5PQCS<337 zev&YEoTm6mc-k_lo0=j0DZk##Uuf()>znS70yY`nF%(bdsvK($aDv3&lf9AtY~b|v zM(Tb31G#I*rvw*cPtJIcH7pN~Nm|wP_B#C4lXbM^X>?1WwXz16tG^_w#dpD@tw}u1 zVVv++2I4}hdO6!@-&~4yU<^%*PGLs6e8f)_iNF~fE-x`dO-f!)Tu$2H0;O(7{=a0sRs!w`evNJW3&RwYCaD-Y!+Plx|g_lEM?lDxga%?a0i+TERytpv%#G zmG-N3?C>as%&&UVLhcorANSiaR)BFVOJfBerI1HN>v!NZm%J}FjToN?uOzu2Od6$t zE)pH3uwssS_$USR>#--O3d#{gjlo2&71!v;HR>N>6Vgyx{SY%;af|*O8Kr>!J!v?| z0CNSV!;$4USAQ0JbH~Z_;>w9!O;GO&?p7Td;21RsMh=meFB`KPyT>=L&6+b7@u0FI(AgUZ zM;s5~oQkPspftN;N&`>gHw#_CXKx_X`91+I*+YJe%-%r$Iz$`k1d6!@y8udP`0Ndg zB+vFIajzl-NTR-bh;%#wi1JeC3IZEVcmyd!0~W57P`9E~%5(|ZZN_98KC3=ha3LMA zp&*%#?bo~XL%7x*Sji^Q>HZ{0k4<)*XGL5`Dh+|L8-$9xkZSCH+2Zs^luDF4aMGmr7)FR_y% zP1)>V@mbk6ht0iit9}!6r zdV{Zi4ZjUn_>2^(l&mM>JJ~xnq~}ob`GI&J7$xj-J2Ib)9yWQaYhrcuAfC4hEQ383 z$%i7<-%|U8aJUDs>^B|?Z$*%V!hfoPwzxj3^5e{4n z_U&UTtUPx;FcB?X1^_R9aEM&a9VKhT;sKP&h;g>m$q0G@OF^{Ja{n7x_$TC&2Sg^9 ztRX2NjFP=dA7B^-FGijew^>K-nhuZ9ShyR`c0=8JP47v5M(@vHW!BDCx_hN9DQEKs z2{k!iZy=+T!HTxDj$B2d-$I!md4+G!Rd6PHfW$a{!8xx{=mjkIq>%WyV-Y`w4denT^IuyNm@+QX z2GUOyaw)@C$UqBu=yjFQFYQj?nl&tiL{gfkPN7Wcz}2?;?DnX^>V}w(&Lm*9X34gU z@WF3Ufl7yk+CUSBQljD@m}pNi0S9iZ>pxznV!4crNrs^2+N zH~Smvia29u9wcikkyf&_(#nlnsVP2<74xI}BFdyQn3u8TU%-ln@GXDj)je5+|DJ=F zUYcE>9lkgleN9j$J&wu`F8IS2?Uz5B=mh_I@J9yXDe4$)J+xLvT~!>G@M`zQL4iDQ z`V(0VjKrsgVl!`39ZKMQCY*`B7-m;#jN{PbtO_Z^PM zRT-6V2QGT?_8OKcP_TX1)+)W?>dlOb-ygy{*br+uHOe`>$=IjtjhVcXdBB02p~NU0 zCm#vePv*Ms3rjrfPsQeEj#rA;5F1d&jc0L$wTA zJ(^v9Y4xHST0FX~`L}fDdl9g^#VM(OHdi0__e1ed~>Bc21u;oCFFg zCuAVL6YfXH*@Lt|w>9-7tn)v4@+sMcnbL-{nt4O{kwpPzHPgIM&5W)e?9cWuUY6ab zz#*f-2U=BwImvgut^gn8LM4A6@&T0g!B z`tuk+Jg7rKpWqXn?qvW^^M4F=V-40pL|5|TTM_c|3G*UwMGn~UkymHeZ?1MPjs3@? zAXV725W^6#iipqD@xrghE(8``-oPO5OUpR4J_S*SFmZWc_h7?g(lxPt8f#3=!{BNwL64jH>5WC zSqjPAW!QHU< z-L-qjQKr&954}gSUmWmG`>)J0Zv9kbFISoEMpEJdSU=A|)$Hf_H~tEwl-&1aYpxG) zE(~1K+@j675XZmk%WhUshsbQXMQ{47dlH+KDh8$vIJ<`@L7ckSXZG-P>O--9p2X4- zJm@?AJf+3Eak7QeM76YZ2H9r~yI%Yxj`j1*<$HU#e}7H*%eUn5^Ho3SOir#~PvZA5 z8O>&$vaINw!#sRz7&xroDb1rCNI{==n?iOsZ(rB2znM%g_mCvmSL zsU%Ty$}!u-)VZs(uBz5gG;t62vR(70jx=Qi^)$GrCvt9_u09p-=-jvYkM3Q5b(a1% z-!pzxO4~5|!KJ^i$faFy-{$+@bS#j6z*w#EHs4-YEST=>~<^ZhT-+?BVzwiOfW zZJz&KPinlkdH(0SxA`jS9={*x=QH-(Je!b&w|VwAn~5B(h+x7eojd%=u$}!j-zUR4 z)&~dcxn1mQa6}D=XT8n$>Gco#tOtnx2Y7Gu{a=sj9QaS^oDQw=`CFMjN|d52W* zB9P2o?F{m>Rr=eBGnf4~-+q?9+4l~Hk=N$jxR%`o6$KF7w|Vr+k8t}o%=rPZN6RbmU^3Sw-ns@6O4Dt+q|btRwasCQhk({Y4U+t zb7g7g?W~&Gv3qsz2(fcY_KsMPBaz(35m@1ocr|BTQ$#Y->)x?*d$aD~bpJM`DwzCA z{UtHnUdUT^U6;y*%R1vH@MS*@cVY_SNDiAvOV7(CdjN<@_jX zrFE^Si>jr7swLC&CZCi4utZc9E(w`5!m82fST?%7`v+ba%}K4{{lt;YSg2a?|&hhe?)RGn-}nSciu|ioAS4|7X*@pOu(E`TD>DquHu8f>7Ur0A1=Cr8~hS zxdbgE1E^AKDE{C;Y&dT2^4`ta_1$2{#vaHm%+n0Y*ANy5d;94d!ZfLkvq-OI4F5gc z|1#ixciz8iXYcsMm)rmFp()rhHu$Mw6F3wI|MT57<+=9uzin3*s6WuSG4&sA%KtF1 z%C{Cc-?xZ2Z@qwAb|FV--Otf)xnLV9 zhuTPn+uJ*qKj3vOXh@l)v&yhp{2s6aT(GGpZd}_tlnuqM9Ifsjm~)Ebx9|f9+Ur5J zoMqjc?eZgG|9O0m!~H9gYpnB!{emLK@Pf<7?#!oDuIZ_z7$OfMT=kQ!Q|MVm+)+<| zlc|2D?F|bZ&GyELQpsVRIk#QoT?wRkrGKD+`}lpI;{jK8y=U>FS+a#L3+R$}yyJ35 zONr`*rJDI(` zZHmO(#MHk{IJOZ9T*#g(U6;;Yq$TxGuU9_t+k{^DU?NlrvG?LN_wr}sK|9qtoT=RSd-P?jbcAs5B_#QVfF z`#v#Ey-&2HSG!NhR>7=X#`X#!8x(mr3Rcul8~9YN=1~dWKo*R;W?Wy_9d*l4DKd8Y zOUdfC-dDFI@ug~8OS$14Pas`9uMM3_e`#WR@+i&H7j*`i+9M(4P5TEzG`1%GM8biv zYmP|=@Kd&rJ!@!gE~cJE*8Q^xz!;|@9A}IiqruAE)t<0hy7SEJLw2I$Gz3D9b4Y;5 zw1rz#nE^?TbKN@5MDw(s<~BRxJa5VFqph`Oppu|I=v~NlzeCfO{u;G$u5u#1k+qwIIGY5Fe^8Ql~}N z-1aa2$qzo#mnF(V@$*BmTboEPsWfcUH6Nt71jo}w!DQe{B9ZEA>VL&27Nbr3(y^k$ zob#66cup8Frn44`qXHLS8jAN<*7Y_qIF{O;Bi(2!cKSzEa&VMOwdEq@<47z}43Oev z1_H@{m2es)09HvoR^{WdCdB~Fey(5!+s3sOycRm)G{Lww5?RP9BoIr)hPdqI5@v9@ zHb&y86mtB-PKxcKQ-(A0w;fKZuT!bOt}7-pCS|%zEaNX&+A$7v%5G}=t(@w)V1Geg zu^Jklq^gbhy!!uSHjOK23|S;xMU7y~KJkf<+kfX}6qqQ&I7u`6Da879sdn|S_)e^ z50B0#ghAXDDtWJ8iDzAPB5;?Vrw6OXBL8r#D&Jzw_(5v@gKaEZ3;f8qk|s^6@h9DL zQQlPia(omG$Qa+VWpUS{yBd2aSEnMv$?}9psp{mir;KHIR6H`;&aw9+Lv&@P#2bB^d}vd-VF&A&+b6<^fixNj(--whvOFpedmY275Q&JOfp{+U@Svxa(zASv2k3k*kg6PkZCaoUrDeJIuGqzC74($6QJ9 zzcQO$+m+dvvL5Myf%r*y+A^t|njyr@Uuf()>zmDx0;7{$8pU@E#nZVe7YzM{<^U&1 z{5{zlDVz-~-{(J&yY>l836?Y7V-3rLW0F=iy}b^9^<*7wc^ci4JjLz5`l6c3xO;H8 zFGuMek%&tNu=p-`v^9w*WSnsLVj7GM#u6>WR|eujs(LxwXy06lbzlrli%wxih6sM5 zNCeK<;W8d(?ET9{a+BFZZEkk~*BkoB1!o{jH9v$Q(~hj|&Mm4xcYx8EGt7)Kq&gU;FIMr-^c@b9spwYEtrYlv>*00;O(7{y2w7$Cl-c7+OXw(tOw!!ujB_E5pOJ-*QphAOnvDgIQpjAbDf4%qe`AmQE9#)D zvZZuqi%ZQYxTj{a8ceHyvJ3P92UB5iBd*vf5P2Q+0Q}5n6-L)VZ z$Ua~zeLqU|{2r*~qMptCm+j^G}GS_X0s#gzK5jlt}V6@2yv zLY?mu;F3M$$H?prZn(-p}o zL}5416>%M@Gz18J{_1PCp4B-4V=W{gulD%kLNmHu3!SzHS5N{! zXy{@`fmaOkQ+OwDzqYXspI!IfzjdL1su>8`*NM$LbLD2UqS+?WCI;r5;Rc_9;wo zvTY8Vd)-#~Iv@pX?1*P~4s~`;>^LmVJDyCRQ~lz*PKT zD1Le%zPS!<{W)Gmq=im}|3DUduP2UHAlr3vrC4?#stcb$^ zIX+Q^AMIfPPS4Ac)wrT@v#w*lY*hjI2e{mHLUl3_AfHW~Svf^p4T5jdG&}DyogO$> z=&0JLfE|3?gCy+L!m=4NuA`7-ysn73aI7xQy}yH<5Fze)!8$hgPE6Njlh|`{U4KEA zojcymDGfYt4pOV)!O=UI2PyT)ZClS>wc`Mac^hL%+;4Te$hV&q9z zu65+D>F}shEWAMq8Vmm!tjyZkN_VfcCFN}XAfYDb>kVX-G89*h8{JS*=(kYjM_%FE za}}J4?Ch_jz9c!djTo>qaV;J4EgQ}v?sn)U(FV`-Ix0z;f(yo3tc$hLYSdY{i?SVt zkUbsYK?EQ%j$a7Ov3r{$$pemng%tbVLi`vukPD>De{D@*%D6}y$dg+kPvEl~B^VWG zArHN-p7|Lz30x!Xy#d>#i~~ZRLYdNmt8Fz_A2o2?5Yy3_1US|#*_IJL_>m=0>9A0{ z=oD8fv5N2;;mXUJBzTrQ}hj z$6#K35=#~fDAIM4b1tWr{mz-X+22rC#2GvDAX!_9wAL=h{4`d~kM4^olg?nC(S`}T zIQkc`q9J_CA9+=Uq26=w(o3`Jv%?o>P)T%DesIAbzGz|j^B)JAJ_0U=wbyaa# z!mHg|al?U64aH{OraF|s`Aj$yeaVfh@g!aINbHGkN7}3ev^hn@Ht~o5OdHfVtBgvx z0~ft`dkxDJDA>O1+$z1|>P)i$lo*BM zAzo=*%)-Jvx@t&gnA;J11ly zz7y_8$Jv9lK({sZB>j*Tm4Sj!tt{DvnbKB#{`!$cFg>2;#pv|Kh5wvT#1|M5l8pU3#&K^+SE z1fS@1F9Ud*|6`~dYp@O?x{@DHQE-D$1`(ePZrI$YC(Mh$6**wXM_!#>zq#7IH1;2l zf>gPq*88`PjB?}5KIr6DuvAW&#u=n?ihlvOXd@SJ5+?#$`)Qk1$SFzRlx?FgRTiZ) zcWF)p^3yqGBjw*wmq0IpFih*52R7h*%_`iLM+!4Qsegzhvq+o%0L$Vxh63toQi2Yh z?m1tnfir&!w(a-(s3I96HLc_SuMEYrRkgdlQ`VO$*j=T7-%Z!ilVP&CI;zQoP%w%l z!^txMas{f;5IjoyX``*Wq@r8ePS*9v@$+7GS~A1NbrqUQh3~i`rGn($lg@xR(s&WbTiO#zP@%ii=)OLn0TA zui~~hBDO-4_VHngFZ9w`A=Qe#tI4<-^fQEjetN4hXwuNzRMoVe4^W<~IuAe%7?3Rm zX|4&zyKDE5qf8akUVm}GJMF(R%eeLDBGfLHa<7-R{_frOjcT_EJ{*lsiRJC3(eM=7 z&ohv#U7CO6N5ZJtc(r>_J={LJclp)X==6>_H5e^SmVz=)f%NmtCe1C{tRTaX2>x9k z>F1eET=b^Tx+hr)(OS}|p*G-)H58)YX5WuD2&A89HeclnWLwe+_4CaBswXYpjgu{$ zCaR^SGpHe*`0ZZKwtXMxcF!cDz1zRP;uq0=p4mcd?)6d5)RcpYX0uLNmdnmz-pNLe z?;HpLRIr-C*&hx7VEsHPae4RoL{@?QJgMdSmp?yr&bYi%clio@2JZk^KTnFxoMXOK zKn5esp|_lBA$vSPT_QpH=o>t<-yZs9_7$DJyddov(PbUL0|+iAE5L8b-Cb6Dvn~os zpBiy_5QiH5psWGgGopX(oPe6x4tZ-Y<|~L-S~rmPj5tXil#{q;9a%A{RB%j9Ic6K~ zI(K!p2A;@G9cju43N^T=CvtAgiU}C-DfO{xP}r2QDKEarIpH-o#q_uN+<16@xoY=x z?%RBSamNDr2aMGkZ}a`tICpZiUmy2vzP}(%o_kw&POey7qKd%bOT8L$%zEYJ6X@Lm zGvDU<>$P|GO80DS1s1bkUiuqnzs>i5M5~Xvx6fB~A(V_z%=|7|k8t+&`)qkm$Xd{Wjm{jB~6Hj^OYWI`Iu%9dD(Pa5_ZQ;E1xF zthf0-z5Wv&>j9$w0p8nu{}-e}YlAnaV(^eQHBr?8ZXoQpd52W*B9P48%It^|&VHM3 zsibfAxnTzT?Mioh?%TXw*f9eDUMr9m7?+a!J@;+iQs@Jpb)&IA=e^C_zjr|2-tyBg z;s7iyK*~kJew*)+M+;z$mgM@!GkV!`o!&MDh16`6wd znDl2?*K&T8wbHtJ4pyaQdT!)%@*kFn9Ev5a*^5FZjj(ETI+l%Y@BV=oMr%_~JNWW` z;>bqtb1*qAmM=BfXiDqO1nU0v;=b+i=+0AL)&Dn8jg$v6VdS*@*W==Q;La+63{3{o z>NPn7S8L?kG!X9F8FzEZmfm`g%HJ5ue+}fnRCS7r4LpTqNgZIJo`LvoMt5AgzIS+Z z?ds7C?5Y?1Qi};H84v#0CE7Bb#b(9fj@6GV4qR#&CueYp8@wdNZmX)6OVt_GFnh1% ztz2&SG8Z?3LNyYK1x7=m zHG)vzf&g9WEkT<=*1L;Ys0D;3t=tIy!GYLt+}!28o3rb?!H$hRkeHmO8I-vlEDrYe z(>H`^QX6NHJ}I+-Mt0)g!~HJ<-goEyyLR@DUwpa!4

    En|b98a9DLf$%@yT~jt@ zZ~xnNW#Re*jT=+{;imi#14F=ME%3Q-5pUjplk0PhGf*Bu915tzYOs-Kwb=Y%LUsEH>izdxV^n&`2$|pf`*hLp8X!M16;7FCvIHZJCqH@ zt{koIADDAWarL=uIm@~?+vP{X{`2@Ahx=C~*I4Hd`vpY|SJS!^p3*dB^e-tssL`J~ zw#`*P**e9}DuUFHJL>6gGS$ztyG;atjnVDhKa5#~NU{wyGq#y}o0z_Cw+X$nxlJHvb|8e|hF3)5Z9*?>ZWH{4o15Fd zo}Qe1o9I@qf1A)GCEh0V&k}DF`e!A#33;YXI+srEZGx4VC@c-C zq~9l|x7dARnth*`rrsx7(yQGkWUF9SE@OLzkPV8w8wD%srwx4KEv;#UrLPbRR^PJj zQs!PF($#H!uIlzn$=BEQzPcreFIC%G$`e`Nh4Gq*sT;~WfAWKm zh@T&d-P%M(Q-SA$l)jN=bkI`fOm*MV5384&08=GyOUatfUg8sr(WZUrSkYn5c}s6R zCk$AVTZ!fIq`NN-#rrGkdYc#=EBnZiZZs7;{iCzJ!+7#^ZMmq&kyxM@ASIR=SV?Ju z>_F)-+Ej$o+d@5k=e6&W<>Rp?#Q@EIu3!e+#~rYaV5GQu4;tEJf)#VTDUmhl&S-7yZ%A&MjnNHESQ2ncpF z35-aPVl3p)@FZ1j#ODL1U2A{urNYcJ&JwE@XLLECD6>QELCb-~$%)Rg{L} zpeh^!iLyRU(##V{A^uW465`0GOTn>oY%{nQO-^!FYyeuhfLIRU ziv@AIJ(y?4SiH#DbH^o+(0C8HoB=Bms88~|mK`|u!t8j@i4!=DDBo+sAQ!~Lqw@)2 z5O)Q(gQD~Oef(mjWHE-jpCQ%qi8_J_?|6`yVko^V-Lj$6%kIBCp=13Czm~CEW@MXk2ho3_)(RJVrGZ9?#3l1{ zVv^`6g@eOmSr$IqKiE5-t!8FvvAd(7i%}U`r}=g0#3%)HX(#GRW@l9(vm`SM9i@;- zn%kTatE`v{(JBe6f4&_Ebd*9SanWonc$7lsYHbyQ^Cla6$X|gou`00J$-+WMDP&SA z?N{sA;ZX{iU-hJg+$%Ca?zdyCz~ZQI1v*M0kBHXqz-cadUnE8;arf5n&gxC)D{*)P@`IdcBsaDR4VR~PqEZ!{|coxOo@#PJZoLuB#( zqwkLn0tPg7={wG-+Z@tSr)yDzA=KdKKFTu29OC`hJb`}HpU zFkO+@ah%?f0Pfo>$GPrf;ggWT_ujvCLCC&NY~GnGH=7mBHjy?lFz0gk z_Spm)OzLJT9{E&AqB8&`j`=1+t{~e3-O#7SO!-fCo_S12?4(FjHal2+R<_MybFbSf zpLWSONIn#)uEqVZ85KVmik}{cZw8Yx zfJL3#4EapNLO>DykcnDYP-8QnsY8zYH1v89)RgJV!4n^>Ly-5t5YFj(U!8T*9)eby zT3h%bs7OJi#B3q{&tmC)=~}KIf94~VaaHY|6mgIUOm|W3KP_ z8do%K)^*I6ttue@0GFFisLrucaq$6>y&}qjsGc;^PECEQHbMsM;Nu=7VXqdJ&6sf= zg#)70*o@hMq7a1AVB=%ff*WqUweG5Hr4pOV)!O=UI z2PyT6UC1 zdrw8N+1JkoHulF&2;9>8Ggz6mvz6{%X-mr4{6RuZ&et2rC}ogiitpUtJKo!U>FV_> zd&i^Y(b98Iee1guNHwP~8A*E!Wq#xpzCBmLnaIxmI_gW3Q`?Av-W{JH^_C515qCTE zl4yfxdhHJ`7=N%X)<&yQf8c(%l24F59pOO)ATf?#aE@^?yCx4<&ZN^q;^U4&{1`Tn z3#80{ZB1axxJVnwlUpH|;IkVgDivrU552Au`la0oT$>cqMp0pHsXT=;r2|*nYOFqL zu(~0pqcaIuty!`yBYf~%RG`wVLhVwb;vjTRYox1JWpi*-6qq(2uPJ=t0v1+@Nh#aX zQAybdAy4&i=Ou)3-VrZ_amA$bK>JB7SuCJP*G{NalhmOuZIfq05K zMq3Z9l~Gp}hb6q)y%i(mQ$w+tx2X;#a6S{xL|<~_YCPd_Hm8W# zCjRjARet(2-}?^7-;LcqHKK6ZGvC{|3g>@8)7Y|MmdK! zW_k98Rd3AXmEFoQcKM~%i)w1|=(gtH(z)M;!0?VIWFR zM8X|9)67?oj^(^_ed@p$#*X4{#)#?|0v*0sn~+B6wx*;+&ac0%8Ehq|6y0g6TL#{c zdt$F;+vrVdE8ND9E`qm|2scKjFCJe%*q`lRynOKDYz#$k!Y{P?2EC1he+wu8*dv4W z%iXnt-%h2d?>Tz;`oZBVyJ}?p?v=fxSLAqw!9*6SyfFLl!J+M5o$MQl}IU1+8Q4#Nj8>xr*&n$v7wWA#IFc0OZhTrHE zFb98H24LVVTnUk1(XVes?91l7mw-!i$dHe`I=g;zwR?Q*zb6IRjYsWypml7_Tm5kI z9uwW(7UxErSah4-7?$fP^E_3qr)r9Ru`;zP?5DkW9U9<#P3KZAoygt0?LmFFVfWE! zNes6agd92b!O8@>*`4zuG3fDuk|=wq{?5uIM)N)A0ud^4s2iyQj$f)H0trl$^y%o9 z`1m5&+mJ?HB;i>{66P$q@qCouUV7{Bp2%El1uy>X5G=$mkU@Lr{6J;o$Sdi19v1a= z;5b~|OLH90MTm+R%37^MOGgfuSH!(^pX z1^ikX6|wj~2jGiDgjnx!G@ZjI51xNK5bsK`EKO!t{|xDjuh=ms!E<7Bw2^ueR3ERD znDp~-A_^uywK8rI$aiNTzq|$;0#gv=vWlOAU(F~QeH#FUXsO?i%XOu8W@`mzt8g$h zR)XS$pSzUGNsIIf1kkxL-wCI(VbpnPu*tFRCM-OLp8v2F`+2b@#8f#!7R$oG_SB__ z+637YW8QU&$kM5bDYLRI2Jh7P$}A&3kgd0NwaBHvf4QMjZ-zp4zxo!I7qe@VCzI8F zY@i0AC6R%D!3}+IiF4xu^C6EP75Jw^@n2MxT{I>z2T&+gg)il~v4RR>5nmaK=c}?- zUK-oHg1JhJc`r><_X-ZzE*KZ8_`il=k#^{B>{m&X)J6fSi*sDxZpN8YRTe@9!g;4J z%W3DV4_~tT&WAH^z2-f@ZZv2iD*g%g)NenMcj~vl2kW)76`7$Pl9y;DTLrraUqMs^ zo_AuioXNWd*xxgFVGhI<2p9hVPI851N#FG!H+TI_u`!zR@A``RiZ^{(M-El$;UqcO z?>%yn4H&^v)p8avZux(Gn4p!7$h=zBH12>kaX2#_m(|3ACgog#gLbhf3}!KeOk#s_ z>YKq^hbF0s?k1EV-%T;UEy2_cNP^Ky@+2BUcKQwgr`3x*GnMyh-2XBy!|2_9E)dZuUaZ8tO2GpDdq1DS z=7v9K$uZyi`5SMQUrTCGNmB8bLiT$t zzfXvB>aSz{gTwp00ttfs-ruL#zxDGTF#2C$zW4WkLR#i_zrUJ) z`%mZa$(t+odw)wP9RtV?N!X8{^n1Tu*|7*UC<*LA@MQ1C6sgM{kYmJu@3%DiHh_2K zvA<`&_uC&h_;2F@t5pNOiFljU&>>c4zW{=<^^ zC2qA^N8af6?jOdy$Xi?J(jl4o8;6s6|AWbCv3#kmhLL4pij*W??A*HE3&M z*u`&4>Uax{xgoYQy5rLIy~CqxSC3}k6oT-lT2@fOc=*>Y-8>E}=5}0efO@H6nG%pb zV~Cr~wB2sAs@tb7|c}EKG%hDL;{cf81 zyLGL6u+gK;rI7CzJXjhctw~IN_(*dS8X20pa+4_gazVMIglUj_=o>}ne1cchFHLyX zQ~Yc0>yfclvI@9o786AWHdu&LZw0{eLa2=tUxwTNg*)gj1b6N19l!W;`yW0u9rLI) ztO8regln0<-(6Eyc5naNc4cL|Y=+zYhnw<03@i)Q0!RE7@#gI}xsKSdK=}%B73@$N z?&MiH^6RGdzv1hKbh(fA{Qy2L?I+=8!{CZ-G9A=J5BT%FW4T9Wi-m@oNjkd>tHtjH zJHLe6dg8{ly~F+Y99%hC-9In~7`R+Np(_T`Nt9ZDB>CX8Lbp9mw)_C#iOUGCCZj5g4{$b3cMUq~$ztk(o^!2-P=%r8o%Ar?&9j+YR z(sQ_SXrfZD9Qt>uR}THVk}HQi^zu$}`pPl2R}Nkj#)+pUUOGAoEgnnbFd3JQsdMSz z-6i^7g{o9q%DHq*^DiCKwG|@nL z<>P6MI-5-GfiXYy_x&ly#jfoq{S1&q0Qy+dg+&*W!=mf{8O2dZEVB?`;Ib;sOnTGU z?bh!6HUAi%>P-!eM2OC`umpp9L-EB_Z@O;1X)+)aeQ8G@e>9E7pW}5^zSW1E+w|qV zo3rb?&fTVKJ)1cs_DNGq@`%_TonzIK04r8YK9*vQwj%=8cak5-^y69tgK;Y7T?mg; z$%gmeT3@Wk`}?zx>M$B+DCS3gn1uWbL-D(+lNn7zo)cjD=S)sfBG`qj7Wbu3FeT68 z_bwXk{|Ud}#5pVWJ=H zau$}`e`h8u=n(fhIHx&!zg0%na;0*Sv-nj;+{6)ckind0zS^q6^gWy`>(#JhQaL~9PZ2MT}Rq#;JMY7qHX)5JRcYHBg54E)iN*PY@Rf2xyL7NJH$HwH zXqN%Xl2uLJNx>bz$S5`0e>vJMjdF>0H}f%Cvt06-1|&stmlDX)d3>tG!Qrv2VV~_E z>>bZmGgHC%4N%b4s5G$C3r=7Uv}kNaL!ZB7S^XHR5d>)tufIsyj+EB61}W-^5f2lv+PY1L=r{ z{vMkuf&PH>tz>Yz2IJwVbXudq=y+ zv%{mS*L5hmcZ{IUYJ-HS1IV5xZ`3`P?%p}w^;{ar?q(}Owjw~HjL4@PMQtAKSDTUQ znkfEU0AvB6`*v`^Lt{#xABxA=))|)8Xz9s=%a^f`JF;njD~;i@LLb**GxZj;VzE&! zg?ogMFDJ7byT>=L%~XpBfe#}=m=bdt<(mu~hW~cz`3mCgbVcC9ND%sbCxH|aBEQCl zksyB_Y8b5w(nWinL2GIU=rD+3Bp6YiGf3kg%>dmtt=xKwKQP3GZ~=6A*?J{mkS3af zRmK5J_j`{zDSIz8ZHEj;oc}FtX32``>xESJ0Xquv_W16;%h^Nf@`1G)bMEFX7Ut`% zh$F0_0I>e^S6{RBwosJf@yFehD-t3piZi2!h>Tfa9qi>vzB(-|x(_f^G-D|K-axz$zM})H zI@ccZ&8dty$QHRoq5ek=9(}eFLGC+KZIXfp0%#oNb3?2xMzV~&mx|FFy#E3Vl_oFL zzA6zx%ataC+<)%T;9Q)Pu^N4S5t$vtAJk5;NxlqIJ3Gs$p}>p+wUl!Md4OUU-IIq`$| z%i@QJEM03v<6)J{i*df}NpO~M28Jdj8}V5*Xyl->;gp9-E~hNsN@ydgkgY&C7Qzd) zBjyMQ2&kLcI(98}7}XD{*NC4t=&?OIo4~wk1%l)PJA2uYM&ezvm-X%4KaBZ2!a`Ke z_l;7pgBZQ>r*NoU#j<}JX^rHKzddy&aG)qPHZ#Pqte(n|MCOi*wy2Y?Zl_d#@~Ckj z8)7;-6L%o)s%$4BK97#f0LgKD1(5$!y6?Cba07!B6UC3CGt5BR{D)j*W(D}$ahXok zZo^V22J!dN*FBJWzZyM-bJX)~&#;tI^mqBNg-e6=Tf}1qnP7s<`7qM9LEJWsk33d++tV{e#)zamM*u z{Lz7UnmlORbE1Zcu7eJBN4vaYsC?g0Y-R4MgG;ga#`V}@O&(VP$=Yr4yRm*TftJjn z{Q3)M?l!%8dt1wuDmI9cUUT*4gw1hn_QVgP?RbdAoU&>px1IR8yj-R$iMooVFJ+#C zkghJoF5ALsq)lS*yY}$YB-U}0m+bLf1pv}FsNyqSHcs5qX0a~E#LAO*9w}YoM2^qH zvOfbUSqmYVXLXP*F%^VL=cFlgrXN$R?`a|mFwl(jeW|RJzEdiQAevU~-(( zou-&(;I@g9m;JkUTafNH528n-Q(}30Ne)mSlx@Szw=3Spk1h%@yPD?4==8j(R@ z{fn1nUoE(wQSgOU;lTFZ!QTQ3Fb4Z5aDKVFR`A=Y6!kqvFJC`6d}Vj<#=+6w|!cIQM5fAiWpeXo8r-3;D)G`=@ZD@m)5cw7T`c?$LtlPc>U8e(veB{;H z^_#2R<7GctAxNP-YEJ~MXJb0p;e^}mo!l6fQ7W@ZRYs|h$2PDD9r;my!e|{DN)kt9 z`7}5lrhq^k59g64bDtUYFoP%KDMlmJI;NY{}2NJ~^)6ONLA5*|@S|lW(SVI0`?U zMn$X|JA#28Qm6a`h;$CyfL0yy0o*(nA+$*g>z^T=gjeeAmfg*WSQ*cn@XB-4 zhyk#=ocOq75rCv7Zf_l8iAezPAidRb=ww!m_^Dx%jy9t6Y(OfH#$bP4f@0?c-slPm z`WifjAy28MkimQ~+)WYYrAwr8vI-n1qKqjK-!)9~k6t1zXa>)$5^oMoU>FW4BMj{) z);j$t>h-EvR-+Sm5zf|Wc6qfn!8XO3cl|1|l*$NB*0SHZzjwU1`_k3xSN4ua%cG^| zp8D2z^Ixk}#slxz_{uCJK#&bqyISPd-#3O-bPw-f8YrAy+c8Z+C`9F$=oo(uLQ6se zSO?Q=;@rT%e8}U61+WgL+2loo0&@UGUsbaQR@ab3KRSSRFwJJI2?u~3OtU}hNsM_h zSFw)z&e-`oSxfZ z-*P@Y?8CSDM9YD8FeRw<6Qg;DRZ9w7mQO9%zasj+bv`!$yIZ7;BC1eocqQ+|X4mqs z{HdHZN_;YyQSB+fo|vF4>B~O1VfweLmX3 zrbXzl;Z~IBpF1wAz2KXGdO2^?@A80gL0eJcM0uC{G!9x!motF6;j7`%e0x*rTAyVg zEk~T&eOk9kuv!*tKnaRAxYDO`Z_EoIDEH9V&L&URz@hj$&wlZ5GSbpt{Ig@u{q?F{ z>De#-{nZ_F=7}tGu+~7TW@(FY=_VXfzf3;`7`148hu8O%Y{`>`- zq?p7*dgbT;{$|xq%=170#lQcF2uRXi{P`dDB*%R5=YQY7=GRgG__cxk;?E`}<;9=< z(Pmr;FNZ4On{{6N`wKEcE9@8l{&$^YKk&X+M_q#3K%GlJ`^CRcv40WZykNrLK#ozwG!G9YNfFVg;ZAm*(K`i_i{~lo;)>AoYkzRoS zIwZz^@o%ZC6dC|C1%iusyJY{x-&~Mu2M5?M{_UQE@IZ<+PkQn9l*(#GafskGW(4~K zv*udUPNP~i)ol0bYBsTRN>;Ni$gxmfDI0Qa79Y)8*c7vj^ul-S+}3caHS~y-PN_MU*)Z}uJv_M zU2|0}oSK%Y{>$g&KP(Z|j!WW~xD{(1+eWu{|G*5rIk`K`-#8M|v<)Vw#qy=LAVv=)+dFreB{|2h1^gtGMo|gZ5Tzv09fJNScBuxjv)oXI9u+~gw`OVVg zZ5lfF?Tot$)TOuHqsF8S<-Z1P{mg>OXa##B;@cVBaq0Tr;nB6LM>Du(VDM8ds~oy5 z<5{d$T<+N2AofzjJ2jf@CNnFt+pMZCQ)N#zjNpyTrOVA=?#jjxB6fWeGmzpl!!*mU z`OPwgydy;*NoMk)|GR1C@75`kr}Y{(mTH0)50+-L)+B;`76MSJ*NE?nZFe&ff8hFSq~ULsPg%Ztz>fDtLGj{`b3UleNIhzD2xw`%SKwHKt$L zeM*Hr>qvBb%A!?||us?=u@h7>r#*0^9I=-@ZV|07>4`UuJ zl613J#zvE`98>?wfk0R`lIgk&VeuG|dhpjvpZt|WuWYUyv8x3d^GUvPbW6|S%Atu` zOT0Sh-=$tT^zTYyF!Iowc5a>8D+ezD%892YUOGAo(*rIR!A|5xWL!F?x7wv+nt$n- zCSN*Q>Mc79FCDT&GB1C!y=TZyOWw7E* zTc4}e$>hP*tL}PV>(XSHtaL5$JVy7!tV((&_Gfff9=3T}qs}H%dti+EThu3gf68&O z>j6qX10)fkDaRiyQp2L_{uu?}&2s-JjW7MrxOEje#?`Ie`D@;-q>*Hz7dA9n_5m*% zLn6F46kkmBrt8+5CId3jmv;1N>ihs6zaOkjxcZQDn+5`RT}622INB%8K#fPl_UIfs zP@{k416Hh-eJsTqP0j_@cd{SI{NvvH4jM?B)Zz*l%vun?Fc9BbU#!Rb`?HVgFxvbm z;AbKK!chFK>STJ^tYIGVoB*RtuF3!H+u!V}<4PhDzke~_kT4xGIzl>k36AFl1#5mQ zahj1K`)fn-d~KC)qeGW!f40d>o{z1qWUD2zl4V?JM66mW(vt4-zdgb8o!|r32jNU|hzX4qG z3zc&?ZIrwkgQBcZlL+3OWO3+x<(wRIhnwp2l`@upkSR1z!v3y#Y<#K(tzk9fM zWi}d)7P8u12HOdn90d8zY9xD_#x^4KEJHLGGuYa=hTf2Qwq%xa9Zdr*i0W<^&5Q~7 z;ah-oj?&C9Z=+QxSH;a~Hmb{6SZ@EFnXF(U4&yWx?dLC7PFprsR{Sa>ZsLf!OY^Zk z%W^mvkD)kgHejt>=qzU!MuR@xCd{j4v@qS-b;l`~)_8Y2eh)S)OIGo?fg>n+-T-q9R@Mwi90*!qeh`n1DyIZzwVu zmtiSjr6P~K6(s_mduiy%YACEF9bz(Ud6_%4a6VL(fR(~x|y#D2f=Y|W;D5<2m9^REomfP zna!^4%3M%c@YKsMW{)EkpF_r1lPaotLJUy8P^1J#pZJ7)Vko{jdrj6*;5Ru$Kd?!6n^k!q`@fu zYMB_E8P0`#%rKO{Rwg`|)Y4{_mvFwF@LX~>$_aPKdef3kH%a3y?J2IHM8a}OD`CD5 z_)0mUT|QWu8y}a_&L-J#Nku42RyBnu1$X=+qtqmgr0XEM4iDq-xjoI5FV7rzDezUU__ z=Ya9M#h^QqlSH~YlG8il)IdJ|Evfzh0v{}aeu^f}?LRr6o%8d)@WB$9ZdH1iZpoqY*B7e*rJr|343@~G)tq#Y!4jFD^(2N}JTkxTcbu%D z@KFg9gC+9FX#Eh9MU(f*Ca0^u-L$UiWH^lvmO$C!gs8z}sg+bj1L<8g$oGTZ#9#@O zT0cCE6a}Nd#|BHFKOo%?8Jw=cGB_$7M=U;A0w>gqNvCr5q@f~@5=itFAhF3A2wt_@ zf|K37qut}#;nCIW$)RX|yM1I&lQ-&f(%n0!yIx8I*#m7w^&&u`jL4@PMQtAKSDUfw znyB)uV5M=u(*5K; z9cNZrVmcS?rcEYUaev$lEme)Zkn%oYM?vBq-`#iVj!0oXu$E2O;|BazjEIquq~Zwc zC;+Vg{MFZ}hmnPa$DJ0OXmOnCZ-dmx!mExJ<+Uu;p|wd#g=$A z!C=TGWTT@S`ZVdB8jwRD5Coj;JoA`>QIbqm;QCqF{D;rMZnFs-lw!)^;HS`56M(v1 z+Bm}5T0-%ep+vu+MPx%(M-?^Fwy&=4pim;zRRaMt;TP#%Q{-cm%HmMr*m67b= zv)q5F5RMeW;SMd6F5|RbmFC%pJ8z(N!JhlXY0qo)M6Qz#M^%Yg#(3 z+k-;gCgS?728(|=5Z}b+s1xQ3;%xp?QHK%dU?1HO6ca4dnZG{7E+vqdy_#JnUX1f)Pd#Te z!=2H3!V(69@pB1^VUMBbOx_yUfF@Ivi~IYOcx>)5r_VN~sq zIymfG_5wk2ft|hVNF(ts+06R(?jOc{9$_IW=lez}*deNAs|)aiFBOFQZKO4lH~#k2 zmB4|b+|tYt!?JoRM-rJkF4_{u7iPzM&aKF)Zgej8gH^8rlT`)Z{lu&@+ji- z=*$d|9LHCXn+6<7DIJN_mj>&j-D3ufXT|zxEd`7ef_BG&sq>+m#2PksVa1>oSeii;rNw-M4}Y?~4AT4vVV-vaOk$pO{l(N!)=vB( zG(#;QN*8$272w75d+<(()97&s%M>F&4wQ`=3wnPreClh(z1v^o51$&4^`qiP z(KZC=#)!&pq=6IqDJ4>?aa@X2Jr9zAgRaD?;*JXuDNDqc?N@Ab6D;icTl~?1c$z$D z+oPd|iLQeVbw|6rVyJxIP;6!Hs)I|h_{R0v!msj`#VVJg{T0OT#`?tsTH=&FQ&~p* z^6M|4P1xww+uK^MRN=;CYpe8{t2Z;S|HEiI9%3=4tQu*(M*Lh}F8NBL4Yaz7r7vZk zgOIK+#4g*y>GY$56$0BpsQ^Iw2Gz@XmyHv*v{|gn2yqB`^3EfrD}&DQd06%*AzjJ5 z96`w-R5~Y3p)-9ll$)+z_^zEhO&!c2VRk2PJu`%^c)gz-3Tm3p`snn<^~VD zeOgte^yiYQ5j^)Cy?p)P@D=B)16jr)UYLFO;Lv_oE;`FMyDgLEQDDMPvU-_L*`&^7{Z#)}hWaE|!F|ysJ=WAro zQbu+fi5^HJi=%&t8sLbLJ$s@?wp;DQjI3VI01D?Ic2ajCEh0wtEW^m2C5>#iWzu|V zvS=f#|LR0fve3CTJl>*4R=N*$=UX)=S;T}wM$eNJc39R+7@Yv@cRn<+ zR{!rNlHeRtl{JVSZ}F4t27Qu^TPnoJcAK8Bk=>+>?E33sWH(RL$abrpn32`%89>e3 z$Zj%>>?UbsyDgJOR+B{=S^Zb9?@6|QfYm2~MbyYj$1rartCui3F$W{ttx;fP^)4|Z ztN-`87}>M5k&Rm_#K?A=cCY=pCRZTgf-TC(ZoDo=cI!lqY`2;HW&p<2W7w=2*9gAZESuW6`} zm5#wRvVaA2B_i#(|3a`NhT99#C5%qYAtu|AQD9^ZBaoP^{@>?fWVdJ|8@E)5k?l5} zdv4hx0*Gw@OmDM2+J5T>Uuo`sN(e^ zM)o|z$et&SY`107$ZE1^Bdh=F|4-op=@{m1Wc3n8CjdKM8k6nTC@`{mm)J>G|L=1# zvgc?c8@E)5k?l6E{e0m`_74oi*E-1&7nSPluAQN*>{i^$;^u~U`MM&$(bT zxs~b=P|Ms#*D$J)Z*Twl!1KU9AuNL%yJEew_krk`f{b5gXVocXCEoFKCbjQ~BMcDtsnt|l#Fb@iXUzDHfh z(Ca7zZgr)rn7h^0YxGnDd~O2`ZNEeDB4KsAMM|u$-Y05x^&j;7s_cfKY611=&ZsF? z6d*_4skOS9h6^*h-M+0|bE>}*pV9)mUE)XSY5;pyMDOHoKa%h}qSD_WGJ#ZE2l6 z3Eb>TUom&HtJg59k#fG}n%!=Z60@uKiJD#g2lFwzz2dPeb7$>(Z<>weh=0syN;?a^Ijk0j%tNUDpX<@vUlJh50{6x`e%36dx7e&;(}iRjsXK znx4f8zQIAaMb5fLui?+BPw9X0>HUX8@mo=Xy=y+=$Je_{8_u$EoTu?W!R*~akPmyC zy0K^F!`?P|XOI~+7AxCd=hm0YLh5!rgn1+ApF+iV4#d5Ep2{|4v)Q-LMz?qWFvdA9 z2Vk8(YX5w9|KQ5e=r*x9mQ5|pLi}z$KcSnvb#j@*pqY1=Hk#ic-ZFZ#ZKDY+iOrRf zGiAcLGJk#`w)-69@4Y`pCTV+&=`d*8++P@oNAkAo4v8t;Fn=b@A02k1f&oD}&5qMY zz_iWslE2#=5BlR)w8}3|lDHy)cr=^sPn!!$&p;2T879jOn#G?Sc2fhAcQ1Pcwewus z$-v3K_s6#sX_MyMAdPdxn23%)-!l|vv(&t8z0`K)Pv(u}?YO`g%#9oW%RsDP+;BQ< z`;Bqcki$mZN2j+%qj{|}Nz+!Shi?i}#D5yN2Q0_NdXpDWLc7QHRH2eHP*SKsdTz&c z`TO)6BD8g{mo7^E+;r_G*&V6He?1Uyg17Itct@zTYMnp?+4xzi3IHN+<_C0U?yj1ShV|!oNaf0%hz*VJ zFqDfQ8o2kO);jL^V>?i)#*_%mGRUlIen(5$*aQ$_Z)~h{)o_Tgo8aFmBv;nD&Qh<< z4nF+y@ezGmY~5VfuuuLw3)&pQiWOzcnzn zS#nqSZwF$UvNiXHfQ7WOHXYm-#FyorF=cb!xiuUJnXWDH9e_a}e)4}yaHqdZ+_`;+ z?1#449{eTQU24iMORdduzl2%i3^kac6*9XWz;wsh%#KZy!+zi zy`!TMYAD!z5VCdj-7b;N=_L9g;P}GWUvO;sE&9ygS3KviCFID5uXT6oPsH#D8K+1u zMKh37HHTN;oDA`M#JwI(-r?G#DQTOr4GCF^@#b#%=9GsXp13_0GyVXzDE^y)xU;tzM-5}^2m15|txs)Z$EJOsp36KSeB&Z; zZxsLYK#Y1j^VP1u>oLtw@?sYw5AvkMGZ}9lxx1 zjeTpRP1zkgx61?2!uG&19*aT-wpxL8{1`32|8Vxy)#LHkJbgU7zIS}}`g>=`&&_Te zZ?BJm^~W$X#|R&+I-C;2OHVhgqVMgJOzvFys{X$LUk!hkIMr~)oYEX>#=a`q4Y2ib zCdSPQ7hk=C4_tYZ=JROho`%mt)ACg}{aGswx_mHND%liuQD$>OqZW2#T4w zqxtHya->iyIBks#FUo>@%{1`@#4#~7hkz{RZbw?Jvcl#KG?f_ASXet zUb{Yf>EK33&|B|X-Jg9_zE5wz`}q%UpPB4v6*#}Mb?Wxf!p_=5t1rHMa5y`f?Y{Qf zm0gLU-De(Lo9;gK;MV53-FL}mB#4%v%{mS z*LM&1uFT|XV#9uan>{hyP8=NulXXqcWaj}%Pt#F)r)x5P<%QY){=wmgk5=t;A~icx zx~2P(tu@hs-96b!zIoIy-kxJ6_4XEooErFb@D4_I$npqj&t@N*j%fjpIqV&C3H1d) z#m**g5X;TMu6YZd4hG_H55%$;u8hcFjL>gkv1ZP{_lviUzR?DcXP8`(I|50HJn;DU zgCCHu5U_A|*JyBINUjY!@h1l2ZRi$j!fQjr>m57aWZ>0RAwgnjMj48a%K^*i_LvZZ zk>lDP{9HT`KO^J2*dF>Ii-~&7WAvC$a2eG8f%L#F)MVCA2RdD%fJ$VHvclqRYVh`e zc>CzLNAT9p9EGn(6#+NBMF1yl0;UJkRR-YBX&kNYJZNy#)!SRCUAuZD-MBpKTNB4u zcmi}?@^Rd+m4fb{;h>vxK&Qoxg3e_&3c8SBA9TJ8qjE8p+K}weTI5X)@;EX?(KjVy z6PpH*?-Y~K10WzFje$W)ygzsJ5q@u-hM{v>o1+sH@(~(?;W`I~+Zu*zJF5nUT}KNG z!{e*3%nqBmRJDQFrK8Rth;rs1X7blLK-2<9f#|Xu1!73B4@5j*I&5vp8Hg zIOGWtMdCV%#B&CT4~ch+qgfD8+@PUY`>2WD?}#Jl1_z3~ljUnyukY6e zp{!O=iU0Elp`7T5fpCKZLM>_(gf62|5QYT%ApC|PJo5%|*Y=&GrBnaa4|RFP0Sx~S z3&RZq!{BP*2@nO~1_?mU+0>~M~xqZ=5wA| z-8+`YTXiS}Iwi0VR zi-zopoo{~xhwK&yvIMdwPiQzlva@aA49&O(>}8uU*<^t9kxZer>(?AdE460^k}0%q zanPz|kD}G(Jc`zk;4-xS2aDD%gVtnup+MatUF*9IcF&8g(b)v-_Ee>=TRlf(w6B-g z{jtVFq>wq>m|H!^!KkxQO8MWb0sH9ABL=Y3CZj#-e_JWE8R8U5yN17UlroS=`cxuF zdya!rEqxTFF7Hv4h6MX4C1RFs{a5%BDNBg-^T;^^)eK3YU_D2I^$7#khsF8PRx(&y zH9_vKQFc&A;qx?RPt33W_c&(Hb1+N6^j#XJ=XN#?Oy3!S>Cx=+ORKNUZmuTz)UILd z{LxChW1i=rRm&YktIKs1ts%iaTB%D~2-IJ;Ks|4Onj|X}qUTA7K4u_#R;-WKI1qJJ zB*T~fse$-y*kkmWTLhT0!2Rv`9vN?(Y;XE*4b$zNa|Wglabend((f3E3BKJEitv6x zmSmxu=%X_#{V|jG;{>98fJXgCh|>Q!5O?TogEI!EiLx5wJR*_$E`!u9adxzxgw#Su zN&z;K1axAalmhG;$CaJ{U}fKs>#b)M6kxlil=Jtd@?8{F#x)jTwe)eoy1d5$8xrjU zHqE2{yB=0;Y7h&nR!lfn*XDv%yE1a>*BdClM946M9a5M>9oq2-j)rrBD@~&?33$oMvQ*QH&8D2-ppbq6-!z3JJ z%cY>DysMLTgo0K}9|x_=dmOYO(LQJ?&k~-5``*>29kINtl_!o?Isqy7By3kka7^)l zgsIL=K{V;aU`lycH$4W~DHTu3_q;9R0nbZ%*Dz5B*>ovTDevl}CZRyp^2dSdavuk3 zNVE^sxU&foBKqFdW+$<{YhQFYUN?+4%XYK{K`t`TmNR1qnZS$PFO*-=)?d@9e>@N z8Dysk$6tVZfalCU{)V|b$d(x}BsW2MRVP~sg{YQ34pEo)I7CCDeTc?<3xLypG}xvs z!DnDps5nwLNied{!05^_9I48m=9PJc1gy?kL3HZG0Q=_%kLsq)AUl0bzshXxY?xP> z3HU10HaPLBs-28S5+~GMHB8+>wp$8T%AY#PN+?*h_+Bs{W`3@d56O(=^Y=#Lt$bRREO(#VO#ikZLj!l>GI5tC~eQf%dPwKq6EM(q; zc~P|)N-4s`!Aa*PJ!+7)A*Q1<2}sLL{Uy<<6H^c!ICNH5*HWCUZiGa-*+*0@CyhC!fK{9R;Mvbrg^Q(LNxFE9VMQ?EQaRR7!$$zZ14NQCv>X zkeAJe4KCNj>gYixE_0KPTH;A~>a-J%VV|SxRF!PlS$66sz92iL;-KuZ)U~9tRLXbC zO`i}{i3{hHm3kz%+O~GXq3TKihiXW)57ji^Da+Zk@07G?ltKllQOV(mT_X?HB%GP$ zsnN~BPL~9(PD4T9aAKFxA0vS4Cc+>)Wgm0d#i}H`n>z=I+%PQ%**Sce<$HPxxvX5I zsa66UxvmIsD-#Oa#VPn7bktVE@$ zmOl@?qL_G$~OfHutMLAG29TI$BiN@AL7>Eobvd5?oOB-#fpd5QB~uj5^95);e2 zT6yAVrIVR*U8P+a!79ZA%Db-LDoIU*ca?i&BsJxGSEtIb22kaOnHdw@)eu)J2=BVi zN^UapM}g{b9|bBvv=7uW?XJM9i?TZz~>pGp}lpCmRNp3X~YVtUbLpCJZ zhpg{DshjIcKN>7)(P)LLRM%D!Y2rv-o3Mkx$hQ$B626lxekB2_^H~tJKe1aYar|9( zGis2XCLDhOwpWY>%k7>JL@BSz%4U*VjX~7qJr2>3Xdj|6M?zPruG^F*_za8+6-O#P zw2gfRMps5~9AyLvSe?m&z~ICHOH`?@ySX&TPT5teYo+sQh)OuQRVG-Sqe}Y4m5Kzf zvvQc^R%2jw`Hq7%B-#gSia%Wzy((2ni}t5bl`0}j9HZ;@ZIxMHiY$SEk#MTZi7%_` zI%@?{_!EOBQKc&P$hf0Y>dmfT1>_J`uu{QaxmPAO=Xus;B`g`yuFAbwFuM1nD?+a?0zCPNGV6-A#`{cFHb3HC#uM zd&&DBoIZJj0tH(XyI!b!VFc$~eZ0RYMx^l7-kR(rIfONTz z12QDq2PARRVpOTF+XN)2N);3tj>}C1m+?JPJy)33t>9isc?&|d`t1l!?NTr--n1+Mwe4fsem3t%^9K)(B0vxL$(LPqwPPM;j zvl-TH?h#n*78NUw*)4OjoN&w3qN^chfFwV(03>#GCJLfgCx+d>gxAlkySXsPPT!~V z0C#pCHw_>*fi^2^B-_JlXZFcAOwB>IK2Pwn@{uHmW8igVfP*(Ax(vJ@8=8{0b(@Nm zDOVi5bVAb8rZ2=};=<@*tSorev4 z*O>6FeP<2xa}fT{6S=IUB%=U$Fv(>(s+Gr=YqlOdJj?lwEI_o6SmMHnjZ*!pg{>q=zt7}I76tAGouKrjLE$4}dvq=xg`Hx-t~SaN z62Cf61@Q+b=5PN;0>7*rC2yxoqv5H+&Vw=h)-PGz2=jFi-pxhXbAQx-N|rGZgd@k+a)6IBpB z@-%;8XYHZY7hgU&oE^<}UwiG!?(D|#?lTXrO?RJqaBK72?z`ka%81VO$=Hn<0iS|(%btYkL&9C*KYV(v*rHX@w1=)u*dEAilJ1&YJ9m$Q_OO}kE=3o&^DjwCKQkA`!Etso*!8*cs#qV>*B z35-qhLQ_FSB8k;7lLy&#DOM@}%1U37%#C5yWj~J9kZ2#P8UFP@^#E&gn4)2)KwLOv zH-KC}dqU2MAq9&gIF7P|M6OP6K{TtjFU2}Gkv`TWLnCOG- zxfHjQk7Z>yZD{f1=VF)nIBr9teca~zSexC%^RZTFIAF;Fzjl-C!j?ZD<9+vFfrP$7)El zkJYGC3Gc9dA8Qkzcs{l-R~)i*^QCMb+g}{UWppmwGKpNB1%qhTi6NJY54(9dyq%Kq zVW6m~S`q6;m<@#Nn1M{vbWk3al>#NV8ao-gtjAFs678dw@UTE}Tt7Z+Q=pLeFd|MI ztXt%Pm>)7m7Dj+dae&0APIf``=)^Ee#fRN=8)T;`XI+4|HRHo!+7GhlQj}8nS5~gm zh892WTV3YkC=H4BQ5y5D095<&VVm%T9EMS;;$S@|gQm;7FJ9g|IvT}d!06`@NXE$x zvlGo(o5bv`lIKK}%Wb%cHi-W7OXAMVv6NNr^@Q>zD6h)OaFSb%c~zJ3I9@}deY_Gr z1sJ&68|t`WGn~M3IYgK^OzECW+)_GZSp>cm^%OeK-YQv5MCTT{$Ntkif=+(t7J#$$ zQitVcUw?R>Wu-G2*`w~%<2(vafao%K_WF2BT6888*trFhCW_CqbP5x_l$UG^iCqxAykIBCn=r1Rg#o`m_VeP zCzFB5M>QhP?yMU`eiH+cz}1w*WSFUg>~@aV&$F!bBqMm-U%G6^F&Pr=W0JU!VKuG4 z(gUVVPvXlAhbEnv^oT+0RGb+-#6@dvN>Y6G>RFwWf@qTSBr(rK2coJ~?vZh*b#_WQ zP=5JqzwdcxJ>BwG60Kp54#MM7v=W!kv#i7_$sjv@pZ1{9s#cQ66|bgY(hjoq zd4iXfsU$fZJN~*dz`+|5T?XD>RqI)sqy$#A#^s8mm(En;R<*`=P#eBFRRz(k6N4|2 z|8Uk#mO*w(d0T#N1RwxgZs_C+lWsRm+d+0cPvo+4l_Yax$aUF|BR3?v47t6k*0VNO z2~2$;#fF2nAX|x>7eVd}1~VD|)#)mTKAjj|iK^DKZo&+*Q+`z|%0l4kN~RqU?O)`c znV}}i&M?qSWh|X#Wh}|9#t`cY07q;{w2xTgTFa0V z7;;_KTqdCXk7M?}$1 zGKE<;c?Q|(oG{xfTUx>F7t8BhD7_AP4le1s@AhMbBPC6D>NLgboNpnliVZYu_rI86c!-%@kky(iTkVEG4uXfidDkDZnDyti~^uo_1KSM6(HKjYKDL9 zRoqB|w116;odUU{klj2p$%!5X3n0LZlbo3uys2|s5HD~moQvfi8OSX-`}T?rtKc@w z_CdH?id(|RZn9FGjQDZfy3EIM8xrl~mhvv)18vX8N?J5zc|O((6bI}Y8BEIdv36+$ z$`lt!AA75$JW)Q@&AmaizQD&i1&CF!8fF0@yDr5lpwM{Hj(nUC7NBcCwhV@fs%kAp0%FE9F&L8BUU|F|X<}9>;4)w2xQ9tI}1i zn>ND|=*}Gdry#0YH{CoLWT*VfRgdE}ub*M2 z4zgPqYR)NwGz?T|mzAC*nHs~S%XS=-A<;f2iC8A1s&&(*CxKP1pgeJC(uqmjs#b7O z1h`T1$wRHqNkKH}#C+zD;&tbnZdMGk)2DT1+~Yfsn%eRWptgLvx_2zYPR9pV59ioX zNpAWUINY-GlH^unV09&cgEb`D2P=8-^8-oZ*KJYjrp-&rk}3{aIyvdRrV#ytczARk zD@2cLRdVc?>@2!+#cjQB3Q%A90zy=$r5H50k%r(MSdQ%Bz|R+f@c0yN6& zi2y||Ky(>$f72pY(xQ*PP&s;BtSEH1=&no5a`gBH5YJ25GRo<0-6}ase*t&8a*qt` zKB1lNBRkv1>7L;^-O4vsxnU-3OHao{uy>1QwyN$CwFSbT!6P#^Nn<_h6ouz`DJLh<9b-l%>39!0(GRRKftHJu{&Lak_(-g2) zyv%NeDLcrvO9M>AWp;~|rzE!;IUmcv=o?EMt|8GrT#4HpR?GSuJrfZ*M}nB)c~}nXAKi}kS&)UgelL;%2L|W(#H?NF7I)O zhD7@ijr*1${PZ7tp4Db4B?=Ws>P9qCiXKMwF5`-~#M#k0#})eu2fASrusTr%(WzSk zux_dhveSIP_R0>qhf^x~cPq@;LAG29SnB4=N>h?sjrmlU_c&lfqJ6-o`P6^c!>UbF zVqw+F6UXXV%EL`^k!V-W@VO*Rb;1gwNsivm@d_)b5R;oSgY1+FvSx*!!V<+kGf1F@ zxjPw~W`LOLzqG|lSlZI^$33gdeH^GE(LPW!JnL6_Xth~Ou{cv8IvlU`z_^S!Q?NFI z*eEkd*y>ysM6YfM@9JjLAUo}2ipn*Ze;lYT_i>q6?x|juy}JP)tRG-BcT7rwP$upn@qVO(oqL(a>cp%=bYy zTnbt0=E_QPl3R`4TwTuNkPV6UAxroiplr1t4Yo;6a5NZ^CXUqe_T`3|%0FF27J&gr zog1L)RSKZzB1uhi#H9PyN<@R@9{X>7h`uMBY5{Inj0Vf?p3s90;Z@JEvYCwZQLpOp z9)&1Cv=7mk69sVEiv~-AbV?I^21bR7BJ~`d%fyZ_p(`WsjWU7+?5&c?^yl#Bn{#e1 z4FcH5^z+T;&W3runSjqX;D8&)e&qU7xz!T_E8$Piv2vK?RwMq@<2w%4kZ2#Qgg*re zR)5LHmLx42rwOveF-m7IT`=!4>tb!RioeIe8U((@Q0|pX5I(1~RuF|hu{$bJOfL7x z@TKfx@@bgQUoS0K?)4VHCM#jdh#q&OF5_`*hD4WP^Ot+rw7E)9F*zno9Gp|c!;#h_ z%qOx3W@a%two}q^>f{szohNn`{WFB)bkk#yoj$A`=XB?cahz|b9A__8(;<_1>7n9A z=fdn9WUF($gq~xiB+0GDoTkfj9GD@|Wx)Ko9x!c65>t*iCh2)`=Z(W$6%UTy5p|e7 z&8h4xbs`F)IVa{U|11H=|Bt{yazDUlLev0YIdC5p>V#brAs zxssIZ*ox&yiexL2qSD)zEGr3f`~Lm^|9zZu?)lICyLVj`B_A~8t3bddrH^pI3Yu&eL@nM z7Ishdmn~M71xV~A!+F_nKUr|Eqjs9_V{%Vgy}MG7@V2D4%*##?9BquA#4XjaTlj)V z%D$!A+sG|dE%aRw)UryCw5T!P>8=1y)lg}ls$|IMe>t^psgBFiBj}cDbX9T2Cj1N> z^9pNJI5@s0iJO<9AavLmafw^1W490nk(3*$w%ONNNpqXLm2|RUSq>t3s*|OTideNs z+Q2d9x=VmFH&ohZu1`+-{N(?&{A*c;1l>}NuNF>Sx)zCfOEoS4Oi$VX-`7xYJ$!X` z_}a}geQj3?eBT_w4Je=+b1#3`y6pJO>dQW+DHp5S7ZZYLc{d zV_w#^KTg+BX`ik{D$BT~IxcIHpj)bu^}<;@MtByNmdH!F7hWS$ljO~7QxI%zjJ!l$ z(%7wzK_q3@CE3<0?j+S#m!yTg3({IvU6K|xMqAhNIBi3vecD#@vQ}Nvn667gR}<&z zgp5l8?^k+cW*rkM6ue7N1f4|ck~T(AqAp3x$OwhmbxEN0fP-LZse=}mNm1%Vo8fh- zyEGG4T~e|7Xdvv_9;GOtv`^8vuktfoWz{8Vh4N?^s!Jjk6=mvVw`3l(2#J^m=ANjU zm?Ugoor3r&C&*l5RiUB?%c@JF&@U&mb??wN1r-WU!ZIDixC~N~y9ROKG-1^yX;C9S zRkyG=OPsKw(mr8*I+A|PzDJ{Vsl?TC&2X;L^+{6tmNO&trQ!;brguw?5^=;daZ6EQ5WI-fVm7taEhnsHM4Id+w#6KO9=+q{eg!POcas&k-1dRMoQ z29Z?mL9Kg-TpE$yHLT}BL|&(NWfd_^^3}(~Vb}IJX+x!b(#pMSS;WMOgv<5B*-95P zNxf@1bKTxGFK$7Qw6Uy}@~&=q4I-)BgE|)mYawb_=Yt5HN>s|bvWlA~`TFBTb*+yR zHB{OsYNB^7i~zcV#)OKmukQW0{6 zV`JcY)AHDu^sQmN4+JM_{)@iWTC69%n zTdF5te;}wHIEKyEa=)T~UcwAGD~< zPdBM#rA~rbHB80I<79R1j*~T1+9zwZT-CazIw@YUk3l45-%>5CUUT8E z7V|D3CaWY#iy9-QYjd2Kq0&AvNnhf7RAt>#os=a>&@EM99dS<5zW~R)r3%a_psM=* zhI$)Pk>A;k)FExmTk79aeJ3F#Ba&s`MfJ=IB$i~p_s1LRJ%pqi;-vl|;1(`L^Xk{k z=8LoQ*-PD4eC=CY%eSg1?(>q!=djDFztT|Z>G}Nl`ssm2YH>sa-Btw`wMCeSEC^&? zWd$ciC>m9t=4oDlf_S5i`AvQ84M&Sx1A~alzOdS_@4_m{*{~u9kvui|{Ar4_P8|}? z*&GK1XWdmGI2$VNbJlm4)S>4krSHC#bS-O%*r`y$DVzg=ps>3P1cgJz(@n5G%>=w%)qOvO>fEAErvLt`Q zvK>Tf=SvuMb`29Jw4V}RhVy8;A%L&bgml3vaC&+2z<7hD!BK_w96x)Ri- zYnGT54&>wlM$@d#%T^FjZH%@5CB<5|at0BVeMJ>DaINWXSjB_L4b##2Ktv|pWV-9r zF5$e*Q9$t4T?2x*q2fMoiF^;c(E4SYyk-3obfY!0yaatyOPH9y)*2BIrZ~;wypjch z*v44=LlleMVj4tLc6E|%Md77Z((i^PJ&44q{3ZRaQ^r^!1T+I8d7TX$i{GPc$)0`-!O$(nd`|``Acb z)5d%($y~R<24R7mxz^=GE|nDNWy3-rMCP@6S*N@SXKs!Fg1N5!3Fd~1`^>H8Wq->i zZCT<3l_CbrDOZ&stySkFWoS7M!cbD>@R*wBYhLk!0LsDd8o%W#@;u5dxR{Tzq15l&v@NTjVuw<} z;u6Filb>@1hf=}h2%Dqkp!u5@#&F=ZPX3bq)~(G!M5Vwl_r=y#Mw@E1zYS}I5SiEN zZ=K30oV_^$2===6C)gV*?z5M4J^^tZe_K{ZvHWehas+KB+&cZhotMKS#3VHh-E3LC zNM*L2NV(L;;$zC+Y9SfrQgVOmTt>86NuJOe!*Sh1B9s2Mlb6pR4#7e5I{htmhOC8T z@VCMlGH@xY+QWz>cf%?oj&Z4Q0d~vex7!Y{e5zP~g1fHu3GRl9``jgb5OB}VK10^} z)vJVuw6@Uq<^D&@z?=sZFn z3*KcXW8}O(2BE{oVq|>*4@Zt$p@WF(gaeevV&9bQ!#8Ege6|x72_f>rj+1{TMt!C2 z@T#M6ygRZ!4fS!_uOmiEbmV&+9!Y~I>wnwPvFP}&$xe}JN?TXchn%DygJ*tYH><4#!f zgUGs8hVqJ?w6IZzx^^cR8Y-T~Q2o)WWN2CJ1YMd1R+AtnUFO8RH498xJ1_I{76d>W z<0W-#w&NDqAfmEw&3b07H7yM*d=P0wvs$S<%7t#tf{RKJbQqtW^smf&wV+rqiRmymFJ?jDv@xDiw`MzTISnEz`_}AW zGq+~7)EhcVB9l^HQIi%n%2{_62+oFz`<$iHvhZt-eQUN;7BxY)W}|CMPr7e0VjslHscx5~Ft5s~00Qyl#$5)2 z!lB}66t-{8cFF=L=+MK@>H352{$e@?S}Ij7X)4#BQaSu;a1ro zqOz+d0F%|8(1ztch}7(JY?$cumseh$Qzyp&!C!X;2>ynOr}5XmHQOocoS=dUa$O1P z(hov0YbeOc>vp}}yQS8Nx;3kXWVl}H(+&GZEoyuq6(+yuXqh(@I#cOxmsjmnpaAEs zw+1+G0mXgZ#=gP=A9d_ovs$S<-Gy$=MwSq#Z+DlBo|$)IBhw=Uj+%t@z1>K~)5d%+ zRq>>SWU!d+#n`1$_JTOqYia(5g+GYjh23dA-($PHil<`z3I4j)C-@sG?(>(7ME$cY zWmi0DeRAqT6;J3=5u~mCp(tjN6e=st)V$sW0n^5qN>w~{Srt!grm{<;K$B*Um*#3% z_=A`?m8+z$eG=r+s&S=PBWnKFd%ncRy znOn`v+P8SLKDn2L$`k`Km8(jSwpSh~F>lQh^B|-pRSu81XujsTBH|3+lvJNVC zC>5+3LEL^xT)_!cFcU)Exb9W?n-|9*@Y)!E$w=8P&p||$`&;V*BzF^_*&9{~Au^}3 zm-M&1@~Cc(0K(t8_9xgID(#{%Lx|L=j3xaouWYKDV}M|+y8;AbL&bf@#(hpe zTgTs)Wm7DFYcDQAT>4EYsSCCzN7x)S2OTfx;qwBVuY@Q^QCXicJG#QDKav4>vRy@SBQ)O;{xAg#jjInquveQYk7521p)}Jy7niy z8Y=E{HRf0%w`RL#u@oF0qf1LrcFesjJAiJ{uaE;@XOF8JhQ>vXP5aoqJ_ez~#(XSw zYqslF=pdr9FT)Pt%dq_Kqtv?;TGkD*mh`Z^>L@L2^fbz~J;B;gai6t>e*p?C>|3+l zvN{U7HH)k$LDmW7-J#!pE$Gx{Ua3W7f?Y4k(caxs5k=ja)j~3ADcHAW0qfORf~^M z@Zo8*`(c-3f+QD(Z9Czu&3?lY6f(>T91tCh+fD0FKUTvURf-8S>yD>yJh(YT7zRJ4~Dv)pN( zQnzNckc=>wI__oPs2x_iQA-~CYN*@rs%X91!uYy9xwnL&bg0R!d6lTeDi9 zoUYKV+35Na)TJMFV&0mK3V=|YvIy;I^Wv5}&0y-*Y|kyPK}2QWnk9`~XUZBD`yiro zKk~NWPdwzUyB@E~N!vK;cHLzlC>$!DMq&Hbtkx%|Fm!7+zPJR3`(!Ffg|plrsf~*V zGn^)IUgLtmYh&RuRW;Fbt85Ta*;NyOA!|=)!*U-)>b3HhSLdY78|AON0tA0U#nbp} z-s$~}ZH%?lt=XPiX@iK$z8Q2*MQ({sJPGDpkjhZ4p%pM>|3+Fvf2r{H5*x8g1+Q)QLA6U zC9+gqw0`N}@MJolH=k6K=IG??aPPy{X6K8Y+3Dfg(d>NZWO{pceh?8*!s5K*1%cSc zSWHzs_1u~pL{xTFlWj%kGAy4t)c05|?cI?2@+zLRuu=ZH)+hKIDjwx8nfckbW_xtS z6S}4ZX-9IO%PfaNrIIl9--e2(o?CH)FqK{L1lq4wUmF(wATp=M$fU33RXl0qM!D+R zpWte!xX;z7YYlJm*tceTbj1^~v;<}8U%zAC5g?|9dC9mh{QV7u{*F>FuX#bBwJ{${ z)hcQs8OO|W=9c#>Yx1&T@eg9&)F_!eVCGdl6$l^>m|go5%ncRynHxmOq?fgC@$|~7 zC#Xy@U{1NJ1Zfj0pQMZ~<$@Vao=5c~)lX~u5gvydLtG{OEU!+gKmZXLyY?rz8Y=E{HRfj{w`Tigu@oF0qf1Lr zmi}QpcC3p|E%dSUefNG|AA`_gV?LI;HLHbWPC6RPFmO~LtVQQ3=I`eW2g@9i(9k(ve*f_H4CgJLCz7(5jkaO0A_lvmoM%^GEiL{h+YqnpO zHbJ*$!9^tqiv8GKTuQ2>n0XTy6qe@c-BQj(-J0#YUluh%w`QYjOHh~ou{-9i*{A>l2Tn40 z5Gioln5U&~&1xYThe_;Pv!s!0Oj)(q%z${E6dv%ZoC+Y|=Sto(;1mWFPouDXYgX%% zf4PBf&BhlO=kS28a$^2)Y+QocNgSzi+8Bwcs)>PHWrH{~cGU!6<+>AEE%DvJUtXP) z7B=RBy%iw%8!Dd0U;EbVpsaI(3MR;PC8*meK6|#A)e+=O0*V3^8?@`q>s$~jY>c(k zt=R#q)`>dMVc)1l*8@m}$@)Ys^M-g!rMm-OwUZV$%3F602;PQ@`@9V*CWy=uxHYSl z%AGHCYc{fk1byjxC+3~l$n*%O;}-aS;zrLPuXwrBET$@+w2+JdnqBc^+j$MQX0^0T z{yOLL)FuQ6*yL+7175{bvHnEd>{_4TZ>YG>-)gC^eQQ?hlT#O}ctY2dAT9k+6tfx% zofx4jWdYjX{@YOTG;k|!kWGbxI=dPQG<;=W8y5Z`GN*Et^tHT-CvDuQuXXKDa5Yri z=W5ikglARut=R!x@kA^iLD^a=o`|Ut(nd`|GdEK4w6RE;s#P4gRW}F=*27%uj-yR8 zI!X?Ue-N2r=jl{7k;mNB0rP-Y^`wm(Wv**~g1MpMK68U8nRGh#Egr2;?q#7e#eg~G z$`PccE1#HUip10iX-TEYm^d$gK>)SJN(u*+5@3B)LMpCZi>0 z#>J_OC4DWg1WF4VWvsgb1Y<+Rql}Gv9Q)Snpsa(69ZCg@OAvR6Fe@q?aF&#Gz7+^f z``f%Y1|h}W;cwmY97I$u@${kLJ?3^)b8mlq<#O|4bL&cDpS6W&m#UVRLuh{+Rtcrb z4AY#>YI)^RgB$@wR_ofIU~j0n&tA&Ygh8_7Z_Dziii<_imj3-aslP38g6s;E6==TZ zB{K-1?hb$J7VRLSQt`3e7d!VJ?Ja?3Y*;;nNDULwiA6}SPx@P4+0-D%0Kr&y1qjB5 ziu;TWBIKw`Q;xqa%cfZV)?Pq@xb&M)g6~4cZ`zY1)QxLT#>;v63<9r>@s~V9b_;kA zQ7fDw1GlE?yEQuqtB4SZQ@KlKx4iNxEo?MacCAlvH&oo`F5zc@OL)#1a#=nFpCKb_ zN>Fx!Fp3nCDt2TaJ0r4^o_2V*luJ=3QCdjG*B;vqQqHMLWDJ-zJ5VP5td{pK$ZLnZ zI;jEyxS#dxk8>4J+~;b{u>gd%&yclJ^l=orVmM`obd?l4JVvKRc7&)YXy)eiF^C)3 zSWZjbnjNxgq}aD+N9(*bt7T=fmWt^a-j~d0hrH@2Eo{ugdbTH68!GOzmT+E?TeDi9 zoUqWXS!6{CvUbR$B<8JIWLku%l=d`7^IDiY%~9&s?9i>oK}2QWngvW-V+pbrmPt|u z5veGFCMmBJN(&n$scU?Kq@m(zB(-nN4$CSi=+-Q@qy$Up@+aoSSZvw$F0ulnAn|7ahc?$yejfJ?>t=XYlV1tOtPEOg^YrQ4mq@`hn4Z3*fQVE)Koq+njOjmk_jI4^ENXt6N{Q@3V^Zg~wND*M(fX|Xy}*09(I z5k1xIQeSR{yecPc<0ysQWgsXVD(+J_`07e=TG_W|wLZDWg>KEp7nk7hko%Ld@mXr# z4QLYQH7*FeHulw(x-~m=t85Ta*_9E%3e=v^hUGqp)N4I*;?+56^G5mWt^mQ`P;sBX z6kS6XY~Pw4mUT{0!34Rk1a;|3CuYF}Ik|w*G;8xZ7X(xrbG_89*`Zr$gNRD`FZQij z)c93z@@Sbi6grdXF0a~23mfIFy9NYrL&bgG66x;h7hQeNtnA;K9hTKj(5>0X5)$;K z>z$alW+T%h1df`7W^rEef>2{)ET$@+hHlLbA}YJ$$+okpD|>Brd#ApNCoS!U_)Esk zyox6+Y?QyQ^$Grliu?Ta*-56C%C2}C(iKnWni8a?ABtj@FQF47bfqE>lBuIe#nZ;3 zWUAsx3(2T>Vwd=X#s&N-`9YMr$)n|Ea+Q4$4hPeU^0gyg#Z!R*yyD5TKh9M^ai6PE z#}Wq0cEyucDo=Ew3jxH^;*=fjmS2FFc=H)L4UpE#LQ|D15w*UtbRdgN|+);E`DfhKuB@iMt>_Yic zkiM2z0;SCxWvsgb1Y<+Rea6OpjvqNYw|GWn9aQX4Dp*{CxLxXV?ke97ePQXM1;g7FSm{Ik{JY08{;eKZ{4CD zL{x%ZZMiRY?mgOUrWqSn4+*4Zu$j(nd1X@t1_;KwD?l(dRNQB5s=qDErda;gUOc2HRqF;n!pCMyls{T8)G*j~` z83g_t^R#4S?AGZZqE?8E0pr$sh8&g)Au^|OmCS2-byA}o0fe7*?N4wuRNUuk%yR*R zbVktO0MwgcGvcqUrOa7R1n;k%-!ohByW^P^|gFtIz%%yJ4j@$|zL{#>z+3`AW z&1zYh9@Yt8r~n_PH+dm%#H)_d!bUx;YkPvVq2fMk2@e~&HLLZ>&yS&7v&f1PWF3{K zNz5y<$Q&@|$fULEwVzR53xfcEV;rS!&1xYTUufBwXJ4-cmc`U6(mR)+G5owTq~H|pA*U}&g#8bft(U)-7< zmBmibtyy3-335)_%*(RCumbJ@UbFDwBrnI20;i4flDajkg=Dx-_GMVlQ7&|A7F<-Epku43iFpAQoB`o#T-WgM znC59-%z`+jjq#MaH9KaNGqG>YPBwFER!hAh&QkI2m{-)Kg^hVrZxsm6hKl=~t(KJ9 zw`R3IIbES!v(dFBs7seNF|WWz1wfchS%hYAUfgo08BE=p9lPZJ81()dF!qL!P`)ApSMH`$hkE;E~}lOTeFenCFtANHF1{?4o{}@dGkp% zX^u|L4);EMZFaucnVlY<9nH>nPNuhK=LeA~5CX^54l`foQB_{?f^XqraIw(S9nz?z+3j(c;9W#^6b*pX= zQJrvrqJ*5{r{~$7_M#Gq37N3mavwYkz{dq2fMseO}T9751&!aar{Q zl_>@cC>M($E&U>tlzipX0_RKf^=_$tBG04T0vrVUqVuTm<|3E&G-Ja`AVlg^#*)64 zR{}N8F+hCAbyt93Y^b=;Sb?v#fBSA+)yo3RFX6Ol_J&nLh|H<%CH*b0JW3lk>Tg~96YLEY z_t{IPdA>7Nj=wF-qgej7TvdX$lY%DP3je^bm-8U(B~=cIOY(Jcx0Fj!{#FagD3{_= zjems}S(+>fCK;d2X<=jD*0Vmr-B59#yM*fjF5%f{$XcJAw$QEFN%dQ^6S{1QeXNXu zspea=le|g>Vd`CR5~YP?cv{M5Rfw7a!vp8bw4V*jg%I;P8zH3tOlsQC^6I1t1Q34K zwLihtP;sBDF{c9%);>d?l*LkTc#JL|;brL`#$z8xp;HU}@8h(O&Ff>| z%!{E-l5U8Z)aTX-ue3=E8)c?zd4idt;yyF| z!(HkfrR*B5`W2g=WoZ+1YZhEof}rFxPWAoPEGQOCVmi#ti&+rxZ;Yq^VMD$1@YUJj zYd6nyGJbG;db~KEo*aKsmsxivZaEDi>U-2Z=H$HjmHThKa=Cf2xpk$n|ClM;zIk?b zqN}!R_m^Sy4k8Q|GN$)P|`kzp5Zoa9k>kk1^`LOm9&9l4&1p&y$c=qQgp1DOZh^Xv?&-dkoe8>J* z7KD5}BqZnaXIOQENSDeq(*N~sYVT`Rfr3C;b7 zu533i?5Qv3|JSg0-)-yxVD*jaQZ~5$+GHka-FI>}pYQA+Pv=h`9kqg~m)Y~#t;Nk- zi-X1Oo3rND?sog_>VbpR-+qLTc?N9h1TL-7nc9++4;iA#nMRq$%cCC^ziW3`SkGi&f;t_ zJ=uA#^y@p?x1Y`Pm3mu&lsBqJn~xZzxP9H~g+}v2`{D)tfA%sq&}G!&#RKkeP4&|B z{CIlpWcHHzh@cnde0FqjI=!K_)q^txdQe?GJUd-Xk56akrjOnl&H6UC_Ly@0?S^{P zm22ne`TY3$Y4LH>6T20)D0zIt&vSOmmg!p`u*`4{uqRyIt&XVSpR+G4=^=7xHzfb+HUv+2>{a+CbR?D+bti}{Ye5vp&2 z9u)s!xqIy7|1o3!JukdEy*X={O|`_j-b(fUUpfThd$IZ^^#Q2}=hM^I@(s0_yS}UY zmyPRdFg@NMYg8A&Bwq9EjH0$*M+J$xw?81k)eS59Wu{lj-@+>-iE` z^> z7mgQ)uQo4b|7^+amC!ZEwrW2fM0l6f#k~uA7u4TssE;M8Iyb5wEG(~@I%Hch`yq`3j3;G-BPo}g@1KAKgf2R)3eiTZ(r>_OG$K|aFF>L zlc2RuHZt34+iRU}a<+qFn?T)|`n4^!Q418{uI%QlIH6S+ywUAH2jq79`GNt@s{Q7N zMGUY+SHHETHVlHoyZDyLB2UMi=n+-Kxn{j~{YJHOx(G?O@71mI{z5$)>xNF_$jYnkvvZ4Oy+EzJ? z)iw=pWX08gwxwQQo7&4r<@hv1r_M7??dufBMue{RG z-~E>V&;IteO%%{iG0%_Imk+DCTg#YfIhQ+;Zv9+`0^4XZrea{=t4D!DRStg!v8YyDzGno3>8# zie6N0rI)&`I}*RP?gV{XcLszdb999Id~@sZtu{@VExWu%{pBt7#m&K^Oe+pLn@-1A z?trS2Qa`$_K7VIyVoRI2rR()7{bbL6Q>$28H z`vbwzR$IMm?_sAncbRs2^}6ODXG@QI|9rMMI4f*^Pjh?vzH92g+h~Y!dh3RB0NlLz zg1$P_`j)!@gu2EAITJ6J#w(9ge`ZU)wmEm7*K6IOl*T*7K#ZmS?=5w_F`MKyWSd-t z)&|2XpD$5}s-M_iqX>ss!p0;FN(Z$*-iG1E2{wH)#ngj z^M7tjt?!H2r)KBZXU%)eKU(Yl)O^rsZEkZ|?%P?;9W8rN>Ob95&n9kLjwRQ;V6Ep8 zh0UpKmsZB@rM{=3232Ra&z8{*Y@8pU@9e)%U&6bv|ISsZ%Uchpr*AzyTfFt!?5#xs z-dh)EZ_U0Yz281tzpVa)h8l_F)0^5o*2T?)werpVHOw zLOHs*9EgK$zcHWsf&Kf;)TV%aZ`U=%pAZ|O7Zczuxd4yuKVk&9f3*~u>aHAL63fxg z&yW0bidy!nRxP9C;W1S*_1`qA@2Ki zsRkUBHpk@tk%oFE1#e-6Qe}gCVUTf;7(WVy)Sqss52pzJ{Mn~(HZLy!xAl?APT0ms zjyd(yTk81~iS$RQ#rfIE!O`sIOn*E(J-l5PjB^oNPbX=*G*+nP?e(@dxGrGJ4e&S^k8?jW7o6xue9q4OHkE+wno+S0H!Y0V<8UD@BL(r z8s?!Yvr9+oV(?$0y&}`vPr^lZM&^|n5Elj-xn>RZUvb5@KgZM}uJe@P>bW-JLgS2^$n_pYth{P)UA{3rKJ{2*3TS}teE-V#z;nt1jw*H?M zA-MHcTef`IX@GtSuz0Fa2k)=VW;YLX@#(Uj>O!Ht`orp>!hUAF@>?KBd`f@)I@kln z*jg(+vQe(ocL|MPdTiUai^G@M2%^w`)KJZ0R_zw_Gky9A$mZ4JkCp+=bBETy8zBqZ z&~KEPzNoChFZ*3k`o{;}5DKP^0*YpyKx>tzTB73qa(mU53tx0Amz5TseqQafmi_d5 zoyRwB-PpOEe+z`RJNJqgzTM3%sI3NWUS9p7RX$K?q}y4(BiAYyvJj+P9R&CK(aaKc z=G$!Z%CpCxfBx+G>8UkK_R^)9r9|Lr7Ianfol^ZFYbjTqblq$%R{HwT{2C*Loj?e+ z%D=UwiJR)WHP@}1`e$Q~q5sRoq_sL-<$Rm`zV(4~)LK`26ySdbs(Z z`TSD$OL_zHJN0wAbE% z7cAJetQhOMdeD;Bd*{)THzpCn+XAhs086$L1W#P%dl9O6w!fupRRvhgP#XkGCOQt5 z%p@5bi`Qo}fmLRI6nv(=!R`t&YCz*F95)%T6~rC0P1oBW0re$j&+f zOkd@~s#qMMeozYyDH?*rdwIf4QOly|oy06)V^=&Uz-wR*n<> z{NkE*wJX!tiPCaik%&a;p9}sYYgyGWmf9qYF?ID*UyHS*ukDTD>a z{FgV?4{ocQYmt$#kn@O3UJ1(k|O0kXK^7-a%uME2O_@9QVN@0huYbV*(^P)oj4>;%xIf-2U@ zk6=wI7tfs3PwF3ls4~9(5~Opp;u|mi)rL4zDBhE_Ql>ixfi)*h6|Oeby7ua~x79ao zW@WM0Dv5V2@4yD#+GrfLp64mc6dZXKQ~mJ90tEjpzSY~6UuxW0(CKdoez_8I;#?S0 z{i!W=eKS%QZ&t25j0zJDw_*u9j#j^+e}Ho%w#Yx_Z_4tm4MjyhFAA8UCjTF{*Yx== z`>JMn+4Xc}L7TP#B3)lJa zRll#W8RF-JP|H6=9;8(IV&;bv?CMx=jSJNJNP>_#P3_O$4AcC_4b`bKt9??t?uVT8 z8-u;dvgwU#_2y(fZ!LXMB18VD67O9}bWi_gkpYUgCTn@H=$j<^=T9$Tnz`Sr&D)Q8 z!MOkCrfB~BWh7vpq6(4u#-#p{)(=(zp}s?by)T)(2epz%D7Q_#vg)tv#{#7q=YaV6tY**|gu6JBseMjpm z3T(X{`E;e~7@J8suGIgur9Li|M1Qiu-;AteMM$*PQ)2DRTahmmQR#E)W6h6Lj!Ge! zumvr)0reXVRVnJQXUYETy%Cwb#%q5cQoo*9i^_SRp&tsBB60=mSK9T&D*sDOn&sh3 z9hzaAH}(Xi{lvv(^-pTjC=W>L(5RNM>L1mlSRU!rp_ud4j0(!IZAI0kRq^^`4b%#? zsu}|R=p}gwnC+3ptB^$gQ(eN4KGuKZAD8+`(iYGm^kX1|_c z26d^`tw)|wUGgwrbNunGWM`YLYw?k_QlJKfyRFpNyPX=UK~izf2lZ=sb&;enXTmq@ zWINum&zgU?Hba&#HOgzHlkHf1_}_3%?T^%E#`0xG;Z4_M8*bosYngTZ#FqM`)Z2ww ztUmHAUt+AZJ=&o`o{O|TBu;0t;gYvUNGdGkEN%K7Tk0vPO`Q||T34o3<7q{q>+%n= zZyx%L>8lZQ1O@6ZnAaAIH#=G%6eGYnPih1o(ObdKZC+8A%#Ud{)DLk4wm&~4iwV8t z!+m4%zip_u#>~4a-exXCE8fm%pQAwhuHsxh21{E3x%6dD$a_h_EeXw96KV^+F=Hp{bYDedSo? z*DULIHMS{{F`d=*m4U1m6N@EjIa6YyIc3NAONsjD;vO&>=zTnVB{4Hp_CjLz=j?UF z;E6En)j{+{F7mNN^=BIDAy*J>&Z~t~t2Hz4MYS?hEMG&snr(i|mk_bDE0L^b9`(Y? zHmgVO%D4Tnxu8wTdl^wb8>0`6p|$eU_N9Zvlj(fkd{jNse7N{lkoQE)W1p?! zVrbeJIOT&m&(vw?>#E@M+0ntNu54*;H81G1>xeM^vvW%^OldLuVML{WdF>I$G#QJF zr|KMmTM;X^YkQ9_s=K<8y~dc^(!4)vMeB~-QP#w#vmlhQUIHQ0vn~jk!5j#s!vT6H zlR%*n!G=&~8U#Y-KNb+mng&)3AIU9Ww4_$a3|9?*7KAckyjKr%2xUX(8*E$%Y{&nU zx_8C2<j|3$bhX_Mzf~s<-$g#AC0jR-SWZ^D-Rj80gCt~hE7iKbdb69a@aCtFc0bK|4{NS ziy6wcJGW5{awe)jQsC1MqVn-GZ*LxrHbF!=erZVrUrIl#y;?FWkg3BiM_Ghc%3=po z`J5I^TKV5ua>R`@EoOm2Difh;35sca`|Gd@F-&1O^$De}pP${*uhh?O&R4%;`bhcY zc`LT>l3C_=j!y-`GFk?<6UL?aD|+)Pr12k`F6C@d2Fq9JYMT}}HDXgo z>VZ;L&x4k-a>YR~r~aUNuz>1L#`aY)s&kdXthw{*E^cKV(UmEy?Hfv|TmGcsCSOD+ zF+*dTeXY8PTCJsCF0^jl`*KM(49JC<)syPVbNbhN7sm_x_j}D$MI&q!^?Q$~i^r$O zI;b$eOniAIh0fqJarCF)T(X1Cu9htS>s{wxI`)6*+W*qC|D|vL%fSAZq5Ur-`(MWP zzf3G?>^U9LI?jKwrRmtxbZlulwlp1EnvN|^$CjpJOJh&dwQbY2rRh4&Y-zf-G+kSo zt}RX1mZob<)3v3sNAB6u^lWK*wlqDbk1b8lmZoP*)3c@N+0yiEY3#xJwlsZPn!YVf z-H&3W8T#p*)C7O(9qPU^?#%zTbZ_aYFu7@DVD*n1YADe}F9f2?SNT&e&R(0H9xQI(oY@+U3t``{-rH=0 z!ooeRvp&=o>L;Xnbuqo(?(22)Y_4&3c4})iEVO#BYB%2nY6bOU>7-F7TdN+X(_4I< zJ~_QTJAeM{{Kn3;*<#x66zaTFeO2@ApgLX`mL@+Y)ufMU5-_L~>-tpx2suXOHitC+ zShkMpSERZOFkN=DRXII8JDQ#EoGh2*V0UM_SS-`&v(ugV?Btc5&)u4x-`;s`c78fL zX+Kk}(?pG$18bcs3}S1AnM>*SwPhF_V%ol+ukF%(w4W|?9jQUHS5a3p#CkOyC29KF ztXNg4rq&QFP5-%cO~;s~@8zuNQcSZPA1YRKpnA;?dqs-@SXzpCI>EGL@zk6Z<~-#H zDkif%llrBG8p-_nE~YMvsF@H+M74(C*lT&wQK%{!!*nC~&BECoeZD!LsoK#qFcQ*AqVztpQX;4+|EO>%S=w%~uESRu)p)nP_c^#iFX6H9%N5{HC z!DTbs_u@maRQ(s?HAVv}*!co-R&&1C04z22Zv_fBEHs*W7YIOShgn6(90}V`7Od^4 zo#y*+)&euA|4pnO8bW<_Ts^aXnA@2Y`s}FpHQ&S3#|&xPOZ^{W)zG-9O{wPYfxLr{ zrwUzrs?*$w>gvv+eqO3#7mbzP!&5QaT|Zce-&ft{ZUynP{`PvIK2#hsbkQgY3K=?e z=fT{Jpksai436sua!qr~2_i$+9{w+yUZ zWSZD+%S>VMfo)LB!}~58|FF1fXOO&|mze_udfPK402qydSOB&XO=h@<=)tDDJXh_a zaS)WN+HwN%3O6T>4P%2A+u;(xjK)SWFgq%n#S!2%9E7E=I9KhW@e-7)0!!mYHQY2b zJ`hW7ajx1!V<;$BC8e#Dt1?DnDJ~8)dT3k)Z(czDwaRbUn-@L=OI5M=?V&Lh?0o?_ z4U3b#FE+5Pz7prEJv5$za@Ed`v%{<*R|RK~=c+w4hJtccZ*sGK%$yYJYOfILYQ(u} z4~?6kT(ywT+i{s`DAt{&qBvLWp|KK_s{%6ebImq0p~WVVhYUS5N`gX$ZXLNQJqT+} zi5v-h^16Z4a4d?3~;i*waJ z8bd+3Dk*KHT$M2rOL6(;g~nCz<^|+g)pFJD4n71+RdKG`M`J43`vP(r0y0m8K}Got z7U!ycG@gQT)$ZDJRd5D*t~x+tC@5F;CO771=BiLv7C^Hv+Kpj@?(&)adCxhmG3 zrJ^`j9iXuil&by12js4LWUlht8yge2Fc`1@&s;xMo5qgL4;s$ zjBREDN6eF;DjFTJR4sEzu@=@49B9aM)d3n2LEbX3c9E+xy(K;XOHDaf(fEhORVU44 zhWCN0&LAvx#o^unje%GIwsKYM#9m5IYiQd9^6#nxG!BAtReMpK#BCMbGIQ0~Ft%#w zNrV{8Lo_yuf!R^n?8O2whl8-x73Zo$G+u&oRbXkn$cZB@_v34HHhB&35RIXrT$L20 zQm)DviDN;z_Z^~f6}))?30$>YwRgvI)gc;F!QK~8)vy@brP*E?8`xfH`FGVJ8c#vF zY7fj+ndQB3XIl+)KNi;j578J3%2hoH`4KTQp#{d5*8mUExCzQti>W0$E;FG;=8)F_ z57Af&%2k2sE9I(ae-=dLAp;sEK_Nq*%vG5o13d^!U2*(7LL(%|g&>j|W-{YkYKS$f zibh8)Rjou*s;V^vOI7j7?g)*DAa5BgN!45xAAqH%n5!c+{$X*oa+=9KR|@Q5&m<4` zMraJg0TRRKx8$cgz~H9in)mBrsxM`#QM<*KB#m2y?aNG!$Wfd(2^!J8M5XRX-G zk&}Cz+{Y7I`AP8@jj3So3#e)c$jnJW9rz3ue^(u&@f4J+_EU0IxHAU`^6#o+G=_q5 zRZl`=e(bD)nVv#J+A7O))iD}3LAh!nhqvQ0(@?CttqOY5ApWj8Mq?!?R|RCuc2|y& z+!|nP5_t{q7>$ykkYP|qu1XKWF_%349itHva?H2z_6)y^QfX=tFTGYCsvakw`@V;~lQtwd7_ zU~34Ls^VOAg2q8mu4*releksORb#`jR2GAIg2qNMFgq%ny_k7OiYm;Hbj7*q1dW%V zTosVii=42)keS2AC$Lvv{#|u~#!yhMN=j8JS7nT3D~JY<^34m4tKiKGNZ_h{R~_JK zDC-iBJXf8dF%|560ae)qY%>#D&>G8GbFJ5!gnaqUHzmcG9%*B1?vATNT5YRJh9fvF+Ztt%QKv2?YPO{uQd5G-APLB6hN zSOmGuU{R_jtoQ&dJrnqfMnNpTI>{z8$PaXN24U%&z*jUBVqw_IH>DW1hG6OXn{o`J z@eq`=+RNf3aMg0w*f1=e6Yz`%NijS-I-4C?fah=!mcIW=zQ$GdqnB z#M1kB#d>$pXbMVONpUNst&EXan#*^v4jN&>I~b76t>GK%5KnHE+zWTm=nD42fUbtg z%$yWd!VJJ#)1Q{38I7!<%ypQOxx$@U`u$V6erPlWrLLZcW&yC%G-gT*4QcDFeon3% z8bLw1YaxiY?=o{&th=oXdQ>LQT{~#h1m&)Pkl7xKEh6(`D>|;Ny^_Zc9W-8oVuw*3 zxhp-0trxng37!Ovl^{=oh-%o$jD@Ko)~zcVC$V(35>KhF)(|XRq01W5cl4 zHv!LRloaznj?QL<1$YhzVd*Q+UGdkwq1+W%9504q=C1L9SbEEI*De}OLAfg_Zl&Cn zF%nDjgdGfxuwX9?^4t}F=Nrmh zM=7~0+?l1HJa@$(^oDX*Peii-%p4W!%F<1qyW($lL%C}qpttWbb62c8OGkO`ia*v3 z<*tB`wdAhoxVH8QxhwuwHxxUJ>&RW{L2SL&kh|iKcSD{85!JAhna@!}tXo&~go&kV z#l(|)t0pi=IZwigFaE4IECZ{IV*W%cV+Kca`3{D^0uJwB zKt5M3cOBok+_jIMM8RGd(AAKTnY)6T@>wj;UGaCnq1<(xlDop4ZET`ZvOIUiANYoH zS5HK<0LdMkhp1bzZ2nxzw3jvw?E;FaahP8E9^4t}Fyc^100b#N|mic`Z?Qd&? zqBJ3PK)=mFvBRW}+?5`LrLWwR;E#Djo&*uqu#*`JQ$ws<*Gkn&fuLnhiQ$f$u2MtW zHb}@_@n^mvuL%moc6!|06`ds^cdho)cOZeCQF7B#dJud4Fq|ahuGOCUMnp~G>*jLT z0eaR1<*v4_P99e+ca06hT3@;U!5{sO`5#ATK6jZ3E^3zLObNN`0F9iW+!c_*i=mh~ zYJ4EpI?Hoc{26d4cO}KGC3j`aU|S?*2SXz)IN|{ExoWxVKEl}i+-hhA8T!wLYP(gaIvyO)o*tB zD_e}fv6f2UEgBK=coPKUn^4&`)?|6R%I1yxwk z$owA3n8CJ)`aLf&`RZz~S_eyvVW`3_k z`?Iu^SN2zX`Wuom+i#gCO!xpSJ>}5^{`NN*O>}C>ZRt^18q56({v0^yR}fhZ(^CBk zHMVVq1TTue5YE%KOgY(UYoL`ihOHIqTjaSd{#ZEZML|g`c~N`>mZk~3MWZ7gZ=I}^ zn*{{gI-{^OPT(yXA@Pu0F)x+Fy_`tNbK5Z*C4srEy)sb5gY8t@oHsVUz4i$wHuw|b zQB*rxle@D7?cwN-#`4@2e@z_BZGm+tGp=j1#guj9vCU0%qthkiw)n&1U~Wr_lEqrw zw3RWGy}GEMlJ8~sOXJ912IOk1Oo*Gqc2d9A%DpoF$T;Ga0c|UKWo!i2y2^7~{C#mS zx9#M)(b4R5KHI67+rqtB8p?B9{3&rTxAkN+tH{l5p}s8bJtcY|i=_^}Fgw2f z>SEr0TOsos)uYWvjLcpWmaPAYRMze;`p)FND?yCKFjl{%P~Z)!*?h=WAh5I&TeQ22 ze(ic-dU$y2e0q3$XK}Weo`4qhny@UIAS?Q&V=3$5*{#zBFIcXifD{ z^8sVD^1oUZyNdePpL$unBl2qp^TqV=wdTdo=>KhXB5hxqo^Bnj8w>kQS|i!I?@Bv% zq#e7`jy-9|zO>^&+HolDIFfc8OFK%D)0HBpD@9IMikz+#IbA7ox>DqHrO4?@k<*nT zrz=HHPl}wL6gfR9a(Yta^rXn?Ns-f&BBv)sPEU%Qo)kHKDRTN!P(P6gfjFa)wgm45i2!N|7^^B4;Q?&Pa-!krX*2DRM?qkUm%z?qK?4S=v*KA$_(iol}ef zzAT+nj3Ir(ESysxF-v=jF{BTfrE`igq>q`UbBZyf51OTOiZP^*nx%7!F{BThg>&lD zW@%3`hV*%}bWSmb^og@{PBDh`nX`0GF^2T1vv5v*>@4jm#*jXEmd+{0kUn~r&MC%_ zK75wWDaMdKewNND#*jXM7S5?pprt*<7}96Z(mBN#(x=eUImH;#=g`tQ#Te2j(ZV_P zQM9zD7(@CnS~{l~L;5&cI;R*z`aoJbrx-)}NLo6l7(@C{S~#aZm6rAtV@RJ%OXn10 zNS{ng=M-Z|pG`~W6k|xAP7CMM$J5fDVhrg6YU!L}4Cy0k>6~H==|gJioMH^=V`}M~ zQVi*9C&D@PRSr5Nh&N-@;km13y7JNnA|^_8&CA0N*4p83qz?(McO ztHz7XtzFet>f-eE<9T!I@eA!Ms_}et`|+P>wBMv2I9Od1n!h@|Ih!}Hs>^%#G`Fw( zL___lhI)6Tp81n!*N+dUC(oZPnyUPBi&c~Fi%QD!8=r_4`ag4aI$t#JG5_$#{?<>Z zd(6`Y&Hv})`~T+SS1va*9ry3OesXqgbs=H@iR^`=8`ow>N5`kv=R03GJ3l&@75~2f zRepoT*=w`YgT?KeGrQT_{bo1M=K4;-*{R*=Eq){S9%d*^`QObA#(n7Rck=nxedsN1 zq{s5|7q5N!K*~P!eswT@{(vS={_H)9>~b+rp7I5YedrefYBR$Lv5c=)* z4w6@JiAm-3F|$DE2WIh>8bUwW3r9ckHUzjKf$fbd5c-Sj-%<2?EzO7!`ipj+Ng(tC zZE^H>spri=^dmit=+H(Jw2z=4Y;Q(~a2B+Vq95sDW`NLdw|9_CpdamHW`WQT^zfD% zLOQbO`-H>nQq>9%cpz{dRjU`q4gS76|=74{xa<^pm}C^babcA8c<_fzUr-qu*<3 zMugB`wDU{?p&w|AqkmW#{YVcZI)wh9eFXhrdowzO{-AXf{YVcp1B8CNJs16GA2SPt zexQf9)DZf~UO4)P710m2H>yDBAF|Q!wKO9_=r7uNCV|iow8hass*HZ5hY=k@f6zXH zez3h69YTN5I*NXzhnWFFzulgTezcF71wud2!&_qa zP=F^=*2T;L{=R@M5=4TiU`{794}@p^GdR=5Vs^Ux(4X(2-kO7t!;{^z3;Jf(r>QDiEDcfsm#)R+<{Zf$m8V-G?h8h!Wusguj zb}-|EpbMJE*?@F0bAW8HJ2=2zK_@Dh)5**OvH_Tf!v?Y&!3OZNF2x40gHZ=$L-8p# z&IYfk853kf(awWvB=>77MWAh-h_3!58*nR6^;nknB zzt10_&3fhkmi;&D(O2tP*UNPI*|#@T?7h}&%SX`l^Sn2v^VfD>SQfA$^f+i(f&e1J)J1pK?gA$EL;V<)gfGWUBdm+wwj17*SznAY>i)7 zE^+j2VFw)stF2T(wbGY8eTc4KYx$UZX#VQ?@#$*^%YswGO;dIHh`N7va&mlgZvW*& zN?n^>Kh7weH#?nu!Rz~gx-vUGD#s$2nwPw8FA4PA>CxfBwT#Ip!lNN+PknUKYsD} z$Sg(sL3Q=D+3e=Q(eVvkkNmC+#fqu#Y^YupGZ-47dsq6VpR{@Dx8K*g?SFSeJyXS4 zjz`LcLyZS-{1qTwk9cd_`hR0vJ(*xVZ}XR+Heu_{SFcqE0O}Yda007)2ks`UpYcy{ z4UM+;AaU35co4*0QWy8Oniqt-3x>cuy}NV;yGw<*Esee_;qKPmB^M7{TX^%UV7;{6 zrJKIHbW?Vhm3dM&hTOuPItB@L7i=E^qscJnWb7l2HoA`#;ffnz{_RFJhjMs!dN^I| zm=m74CQ9}TC6%KUi-qA{>$feaNQs?+&|e+v7fJSa+s9@=2jV;**5dc_Uis+ zo!MU4^Rrsy=XBKXZmDh+=bVYsyrS=ZX`8;(ZKb&OYMs4htk&5L$>MJ~V)pkzK&3}^ zBwm1J1(0o}e_XLrEb}!pb8oPDdG&|r)jz#N^F>Rry1tGp*)O;6QM{9DUj7r zRaWDND<1HD_SMbkNT$tvS{7)^(u2rp5zJQeuTypu0*+dJe zjrwACb6f9j%K&G(IDB<=cB1cc>yxn;^TVmF-K*{U)#Vp&PS2;yqm|vu>hs98>GRqA z?AEzH(b_Q^GP0l#=Ns0Q{Y(1Jv`(VzQ|ZfijHEuL?p=9C3op>x+XHT+zw@Spt4rB< zmXjAj260fHJzBQ(IcY90|FDdsw%RNTTAo&&#@wp*G2ySzHJjnzwd^bFsKI`$p(4LQ zrzZq|yRP{t_}8u2<@k2D_&L9KsjqIn!~C3|f5+XsP;W1O1J&gxqd!B4>7v7GL>t~s z{`R=g>HX@x%{HhL_yzz^vwtMl4E>Q(&f^2^VWHQ1RlE5vP_LWeqXj&zH1;Czk)Z2^ z{+wy#wS;!3Q0SfNtD0{I6>`6_@^twJa$V4$Jq?&t3TD3OvY6!ev3xJiq+gY5gZ@A& zRM=*hv;?&n?PrR$ny688V6D~Z;n|V?QR|cG?b-POJqSzRUzY2O{?IB^CWwf-{PK_O zYVD^BD~!~j*{f&;Gqi03m7pv7W34P*%NiK5PSy~%PMGZ&KOc|oFzC;cLq!+Cy5_}f z?GF{JH&DH1hrN2m04zNd_=^67Ig77$-I18D&LAv(|F39Lg05?47jyM25NEO)R!NCz z4Z%|Ow-S^^;~J<4*haBayp@$lwLel|`AIctjyPC`hG8lFYY9rDAx{j_j3^M|I2t-Zm4>9^SwQs% z3e7*LKG56~X`YS5QaxoGL*p!X8v`nOSFC1lWB3p(W&h8Fbw-0M*ck(o8aA_?F*X28 z(Vt0B6pgT;I|yDi^=dtzEHK$8dfnM3Tsaa^>)tVf~9PN_w1rk668HWNvn8Id;pfBDV#-PA{J+zG?aPv6e#Nq!csVevuIqzg0Yo| zN?~jb!BRFMm+hjl5tPf?3j+l@%v?4$3`^-0NTcym4APF$hQ9(xhl8*bPRM1uXbc7A zvcU3q;Se*IjSs|9JRz6up>Y+I%aZc)BcwJn;f+tH61K4(8f(GZ7?7P+%VohUeN{4` z1aI6!<1E-41IkwO#@GNXKqus~Jv7FGa@pe6qJ3?$&CF%t&MXBJa@iglS3$X~C!<+I zj)dHV7Z^V!m+hgk6qL&rlY6@_Gnd7>bL^uMa@iglKS8-HAZ4Xo7VXbcGa-oRp^+05 zMA#1ui*s3e5SGFz4h4;xAcum;Y8c9li>V>jt}Gfcv6QtEQLGYAY6NR&+XyPbd-l;N z3G$x7vMeRy@4U?8VssXjz}Y?;6R|k!q@m2fKhVP l${OIS(}*hk|c7L2V#R0>yX z2$r%5xojVeji6lCUKl4MTZOF5TsAfgOX(Cyqw!J<(vH%GzXC{ygRm4%$YuLz38N55!VDA(!o=aTS!ylJaH|7Bfv{jKoqsA@t~@u@=0I0ohr#To(MHh7IMi zeKgL3y)mGyVKdv&Vgs;NH6fSnqcIkg%ND;!A<1Ro&MXBJa@heIS3$X~CnG<+8=(-tNm;f;KbZMdnD!Wd~^d1m&`T0F`oCv_Hp6DRJ=tjhvt$!v47s zaV|>_!csWFp$yQd334citcIb?xR@GZ?aHDN6H8eu;gl+C4Z%`2A(tJXQ4-`mgJmgl zS*G{I2Vg0h5cm(!n25z$CkUW!55QMyVli>k~Hzfx+9hiD81<+6aRUNptbW#a>}RyrY< z9inj+l*^LxR?1}=Be7IZ*v5uvtOajlKz3Fwmj!=zZ9}>25RJ27Zwx5QCL;G%5NMC( zEGi+F9ilN7l*<;sF6ZR3*p~w4_gZ*R8_p^rmmQ*U6_m?*vY0hwrm0X@TM6`-JfX&T zh{jS-E?Y?G?Y_)373Q| zO1Ug!B$nza+ZY;a!P^*+-c`$G!5>lHP%b-0<1E-41IijUV`(1q#2GY!8Q5O!lw5X< z##m4;Tl`kElgl!5R=BgR2%7FCeMN`aNHa-wb@q}D< zg2q))E=$T=DVJr8#8N$F8$)9)cpC%Ky9$XYe;Y&RNxAVeLE|jg8v{ZZHZyb90rZ>c zW&qZzCVZEjpfMJd%MS8i4wmGyaA%f+3AyY9jjNzs)|1gJ0rR^o)Rm=N!gtvT8cRXB zY$2hy`!aJ`tUF7|UudX3`nMy9w=>ZI3QA@HF)JmrXn&TTzn-Ee8aqK@#L)UZ?=YKX zh7t52ERFTAg2-1E4V)mKf(UC!%FOVnA=a-g8a1)BwUSY(w$>0VZGSOkZP8E(a-qS( z6bUWUh2jISG)>|y8X2*8>m;30-a3P@G*03z8XmEbY^AhV1}{apHMDI7_1h^(Mq?x> zr?r9TWM&C zy)vM!VKjzqwztLxw%1$zYzn5)XbZ}0hbfsY+}YLyMXvruLvi%nK_e_Ewe^IolG;LD zS=#;c6z$Nc3d(H@5xqT_BPRF$EHcOP_deunh(=OSZVO0IDYr%YbIhbtqKOU~LqXBR zsE*v09)zWFl3zjNC&;fL!dA&`sUg;{EgCzqw6zjbX5=4eYYoBDHpPqL&yho36cp90 zEYn-!1F$qr;cc}S$pb;1w3L|z(1WlvPT_5}C&>|ElX#2duxJdClG}FCC<)4KZEc+d zty*py8-}&UNvK96rOXHN>wu?qnP;Lt>kQY}mbJ+MmEX`AL+b$Ym zLAfm{aHZUqF%nDnq`eG{x?ryiBxNgRv%NAt1WVf_uZ%`quvZ2|t>%@n0a%)*$#Woeg^+v0DJL%D4sqPGV#zuRKnSz4y# zw)mstP;Lu|St++g`?K^+i6-zj$DwFqTt{w855m$o#joIxkVAe25w=Q6y-*$q^}xMo26qTZt*Qx>AH&L))fI`fl4p&!nK-*0#EnpjFFlW5d|mVL(YjHU3z6%o{mc z8xAvbTU2K=2utIX+!lYe9LjA0VZFGDncKz(Vria|+v3lcL%A&}aHZUqF%nDnl)Vgp z%^cp#fMl&&Zac<5$+lVArg-H(dSV58Wk6dt8MzNdfd*O5lJwmcf4>~cZO19OE!>&4 zmMOU{{!lrT+j>IsSCN?j1LLcd+!lX}9LjBr*}Xj&i!YctFE*^bwn^V@@khv^+!he0 zQf`a(xAj?1G=auYP&6^ABe$goVQHM=SMW#4A-{qMYgo#Rg{dLduWhC3r$AIIF_mg- z4Z+ej#f#$4l|x<>)Yhyl^SdoR087)9GpTAXmIs15xh*pbpa)@ToWxr+LSpgON}Ynrb27a>J3r{`lF2O=G(M{?9UPua=kw;C8aIb0XNP+qu0FbL3-zlhI-}K4I~Ald zM73qYqP_Nhq5fD6oBgWlFDFXV7p;Ke>1$=DVpXg$I5tzioU*=X)bu{q`YvTPtNY<% z;fAW;?D7|`7=fj05`WRNDIR~FBv;H+XB3vsDg4DBGw0#i%0zu4JVMW> zdNDjZ+1e@`{%C>f52(H7n1yO+9G2cG*v6kWk7C=A*Kk>a?Qj&9&M6rm{>nL&@j3o@ zC9M}|VLqAp!>)`OY@4XxPr*5QdNpL?ywp7lzn&|g{vq{L^T{eu_lL5z#5{eHm&RW} zM|L$JgIh72BWs1I1f3|!OOMbf4Dr%{7>3yx!uj4C8`-v_`neQjqY>DeI+~T_$k^U1%&@E8-+V9E49f|UzIKk#NDImc z^}Y%3$1EK`Sm@YS-R5pw$8xv2Hd{=SZpT-9!W_~w+k2VchwuS7mQpG41^$LP7+-YA z+?N|)(4(+)PKllIXU#z`gGkF~D>uL6*k}^!p!=zfRUF=C3mfOXnp1qEQnM&sN?l#P7mflI&Mx&`HwjI6APA$Q9I0{SWl-w786&=idfi-f&Anu3EPJASm?kTx1 z{y;jI`;rn@%Y7L`v9wRw)$kY7kzEbQ*w&HzcJNQaNx3inSUTdR0eua#`Cb|ufwi_N zxi9`sI+*)*^1R7;UYVQw!o68KrsTf(ljvaX>q%)=lAG2-eOdaYd<~tTkrtTy7BYJK zF-OieH}^#*Ny>ezJ$bJEMv&uI%YD%SIF?fXeZpxC{^t4P=CtPK+1ZKy?CX&5J;kZS z+td@yHNl!Kn~xZ&y(TPae?Ts6cNcwq{N9x)#%gaVM16y5HXpJ@#~?y!aXYzpY-h^-I(9=jP9y9xslkC&yo$9ew&{YZ&eH++24%o?bhdSv^~0 zS-*phxMx?~vnTG^7xx^9dk)1tN8+AianDI>B1OWkylY3^wJY!1lXvaQyAI@Chw`o? zdDpSLs~k;TIhwk1GdMj7m7}REM^jghrmh@KT{)V%ay0eiXzIz))RUvBCr49H zj;5X*O+7i9dU7=Nl%hA-Aqp2@P(?E`< zfgDW(IhqD?G!5iv8pzQ!kfUiJN7F!#rhyzyLphp;ax@L)Xd24VG?b%hC`Z#!j;5g; zO+z`FhH^BGzqG8q3i%mZNDbN7F=(rimO)6FHhDax_ikXqw2;G?AleB1h9ij;6_^CH}FN z-Q9$~orJ#KgucCmzWs#0gM_}rgubJMzTLlQ*lYlFIKrY_v^a;7VZvw9L8M%Dk1YGG;a{0UoxYFn3@_7?*rBBMm^Xj8= zdEW$F>BDmQya~9{$K~>Q6L6&u%;obY;7T8v%jZqNl|D2V&#O<(<$V)yrO(af^CsX* zpPb9*O~92tJD1O!fGd4^E}mB(pUe9u;7T8$%jZqNl|Dk3&zpcNeTXifHvw1r7+pSZ z0#`Me3Z(nss^c@uD@57))>>eF?3-vnIg^L6>W3AoZH?DBaNaHY@K zp`f`2Kf&{L1BKW|RGUub-Srk;-cm#8Cwt-OC$4FO=m*;y zRUq^imo%g3_gb0}A@mpRJd;4^2ioH3?@~_`fqN&?!-x*CC1@W(KiJ-k4&f|l9YsIV z!^{Ao-)`?9nLt0<$IJquAL!vNHH3b$7mofe{zw6cez3h!1wwx}zvCQ5zt_@?2%*1d z=a~dTKhPFOf3Gt7ksd~L2>n6(2>QYHW^@SsLF*{`ksf9S2>o_@F8a|vW)=wjKo4)J zA@q~IaP;>oq91H;RDsanW24_|X-0(5U$pZ~0-+yhi=)3^8U08PBRYispnU}WV0$w< zg#Ms)6#YmKGXsQvyFC~EXdg2Rgnpohx6}~&$zC}6`xVg-wl}Ik=oO)9Q}ie=m*;yRUq^a*y#6Knh_!N7wtTgK0v~N&>ysqpdV~+Mu*TJw2q=5>0xGo&~LZrq95&JW`WQT^zfD% zLOBR$Lv5c=)*T=b)T%q$T4fgavcL+B@a;phi{cAhxz2iqG}AoLeMwH-b4 z@miV@A@mpRJd;4^2ioH3w||6^hkm4o5gkH*&_05Gu)P@_LVwUYihiVrnE^t--JXkn zw2zquLO;;MTWSdXWG@{3#4mM%C!Sz?qY8xn;>Y-+==WNh5h3&!?L3n}=m*+nh5+Av zwm%n!@CP5yMYqrOCEjmst0xnz*SznAY^7gVE;QpQU3L-SY9k56AaSXPP}u9|Yuhm^WDyMCOp*Zj>R>i*fu$??s(-Aw=G>Ff($ z-v`u{+38U^7Qx)SR&{$xr{_+OX0PkQ)ot}X4fSlbMgPqC+0oAV?C92^si8bRJ<`iP zuLQ~pl>W9Y)vYd{evVp8*-C+Ps}{5PAXpU)Z(~3Ew=5)n@%YHBP5VK0^|jgT=E2eN z4P9CMt_$iM0 z1nExe5^!x!;ooelr;;q_xdj=1lQ!gh6TC|rP7kMx9dj@<*R;ughC~oC z`hgl1dK5vNed|mAVo;sQB+uIJwVC33GuA&WxS(9Q%>&W}ZG@-M_5U=L>s{ z3+kJm* zdBv#bYi8!UVDs|o4~+YMdb#F{mT0XJO)t0aQ-3A0B@8)!Df| zJT(7kN$ur}{>~t@eN;WN_i*#xEBQKpxv@Nn%-xVuKUSGCFX+o;hJTjQ1eOZo=2&J@+B)%}_O)H7JvK@Xc^AX0z7p|0d3S|NcDK!_U) z?9y_z$hU>d%Rek*tSuJt93QNIaqrTx{?)mIZu8>vXZlCy9(#FuczEl4dU$*1@a)#< zV(01k{P_B*{r8Kr#q=cq`-fki=^)cwOwa$S`SK0@(Z%h)`a7rV8%#UBB+f*|-`Z`} z8`XnG%<0L=!D9J?ck@hM^}tI0-s<1S>b~peXSej>;OyqS(CYHbnUby5A9}5avcDtxY_Byfg8<mNTmKFy4_%s8^`x&48hjlKPjX5)qZ_Z%!{H*OwGJN?6>Yol&& z*c;V28}9rlcS-GV`|#p1Y{=JCE7N2--0bM1ITFqA$gUVp=AQLp zh*|pIH$byBAAG@`zU7;@MM$}7TpGo9G??yx;uATTGzqnjU(8PDXXgh?Cgz6qhjJVO z^@a!jutxIp7OY1!m=AJ9ANGK%+Q_<`ls6opQb1s`tkA$z4?v}TXpdFV15`v{_c1jx zMR68X{0O#Qjm~PHJ-=Wyi0K?w?679AO21e$IQ2BJg|HeaSRSDD%0f^rU{NRK6?{O| zAwb)#-h=z}22|cy4C+*|KY&~NvNqaSQdpg(9IK|j*I z$_x?o2Qx&`kJm&&=m%R9=m#p843Hr718s5i7mY2@Ksi$OBb+?5ACv23n>AeF0}t&)=})Z6aecT?Kz6(tMPdqED}T&4}Ab{4vPl7 zIUF7!b^mR!fg&TX9f5xGr5+Uh zWG@{3KsQSf2>oae0{wQMIQqfX1p0&a5%eSNtIQBVe=tK7{rI~{5cI6qOk=U2>o8Wm8cN<$(MRi^pm}C^aI^2MIiK}Jyt3~=(qdC(GRvJuo<+EpdV>p zWrhg)gBhae$KOqY&=0m=$q%6)c?i4TOHL-AYsl{p3qMDEi4>IQoHZ zmLd@P(H<+6AoSaP;^+rk6W9#eN6?S7uQEdf{lN@T^yBX)LFfltujGf&54?1Qq916B zqrYfufd)dq*KQ>$gnsg+9u)m#FC6_qH%k!+{b-MsN)Y<(K5_JetqE)f?IY+%+EMOFcu&fTt)C3+(^=*3dPz6?t^;fp@ zLchPXz5hMidz4b4k07*w>t%E7$6GA!V0k&tRd zs_->Gd24$zV_nb*~{(@_^a)8_xpaYz3Y$h z7i+*awq->m`0=Lu?QCay`nCJLv3CqsOoYLZ1d@=50D?dW3i(n(NQ8nSLLwzWB!)=* z5t1lDiNGNuID#M(oDikz)_u3?*7=<}^**|L?P*K)y5Clv`kgwbPW?{ZbL-aAs_^Jw z!RNK=Xz%e$5a6TdI(lLq?PWDX&xJrY7e?x6?_qn`fKo?$kKd=V_ta5jg{=$eN$uU< zQYNqfQjnf8l}Ab39+ zLG*sM+yZ((Y+c|--d`3vBzQk)%X@#>*mD}>{Xx3|DvILrjf(XptMKY#p;)onJYANl z!mBgYWlR-bovAJZ&5LPE>)bS%gS0U!>wuXj5ZLo=fPn;>7xRQ-J1VTp zHv$;8%AoDmja>_{?x<|-IS{Nn3L3^%gC9xoW&l@1w3}4{8VvOFl!2>(?NX=(t?YNI zZV1rIuz74XkPgynU^*-`hO`>U81ia>fu3Ituz8^aTn%Mhb`H+32GBUN8p_t51Hsi0 zG%VmEX5<&4;l+SdOvJk;78tH7QrQW zKWNK)f7#e`8sz;!y8-hb;LJGd%^pe_KS=lct06I^nO@Z$9g|(UEoLFUlze7ct2>% zdw<#3a~kCRLAwGf@_xUBE8Y8}y?E~j-8@B*_p?0;m5}%QePr*4ts}lk+6&&#v|nTd z!TZSwqW8nPI@bGP>jFRW{;~)z!TUj5-uuhOp3@-j584$_k@x#0Tru`x#2;NUd5WOGP)v?|WTNn6|_m@R*3EmId^4?!I z_M8TJf6%UgioD-1;Y#=ZXfNLTK{rnkb+!tx{`igk$@Tedi60n`b~nxce0X%UXRb8Vw=epYdG+A?SK;kv|50Dv zxI4RZsIN!+w>9IXJnqs|X^_D9oc_LzD__LcjSo$a^Ue%0&m*bDUjPEEZ& z@9+)dqZ|6ly0y%$0-X{i*6RU#DJ?TT zru`x#2;NUd5WOE3q8JJby&twN@FVXpD@GE$AGGDYzijL|4f6h=T>%wEak-pv=?_H1 z0xmz)7OTzE1r~6zT?(~e0hiyYx&&YW7i?b0PwU_<0#R|g2c(VRErYZ%D(irmClJ{4 zg}^`p&5LkS z!fJrcW2=F5kX8fJVWBak)j-COR|5?6{Az&BW2>Qz%fZF~H*wH7vKq?Po&&+v5Hu{{ zB4*@Q^WnvS)=;z;&x}AfPZPw9*dB#SXj1(?)rEj~K5Sh`PipTi1C$F9`L_KcBY2A- zX@trez%v9QJj;wC<&4)5Gb*bIg9U*04$zioMrC8qX%I6C+C{wIuZ2tZ{%9}W`$0EP z5#;@BkBImCePr*4ts}lk+6&&#v|nTd!TZSwqW8lBMb`Ua>jFRW{<4}d!TUj5-uuhO zp3@-j584$_QSA3?;nKZ7+Kcyo(9Kf>c|Y5uPzgmgzmM$wuyw>YNqfQjnf8l}Ab39+ zLG*rDpvZbZY+c|--d|P|CU`$+%X@#>*mD}>{Xx3|D)N567B1cUqrG_V2i-hHkoU7a z3YC!e`+a2Zhpi*NN!knE&$M4;1i|~s2%`7H0!7yQVe0}v^8T`#Fv0slTi*N2#-7t4 z?+@A)P?7ihwQ%X)AMM3^Kj`Kug1n#YQK*Ex-|r)PKWrWGP10WQey05*BM9D4Mi9Lp z7AUgb4_g=bk@uI?gbCgc+Vb9CHujtbd4JHZfQr1|uZ2tZ{%9}W`$0EP5#;@Bk3uEn z{eBlMzJkhXsnP z_rul&e&qdSHDQAHgSNc)myJEALEay+IT zXS~LiGhR!+RB(X{=f3xxdORq_O1yzEc8}f0+4Y29I@d5(yhVNgMfGsiA+n2QnVw&F zSdW$P)g`A2s}{ekt7|uY{=9m4Xty)-?P^m){I}1mpE|#N@0Hrwf6wON%m?B=p#D{@ zxEJ+p)9K!ggY3Nhr+AZT5Xv@QrXJR8WN&w*%cJv!XlkQ3_^s?y=Qm!WF5RAvhq_+< z=H9`LynRvAUfXDtHSMoaugbnUuW7$NKALTgZ(Sen?&|i_%_{5AFAfxS*j9sKqtrzz zQ?6gG#^vqCyrR0c(JXOTbfG)sL_s;Y@j&5b8WsC-yGxUI7L zE_~{o`JcE&y5&T?L2YEOg%fc&neLfy-430V@MNg}6e!m!l}oDzC+fAL?$j^Azfrv* z+koI_lbhp%p?MEQgun4PnaYRMrR|H^y%+AyXHxxYpkKSxk2qsc#!fVAM>R|F&tL>> zl8q;cPLHUEwl8JpE_~)EYgVg22(;>yTA_^7FFb@9CsDPte86~jY2jzoo3ept;e(yY z?)YePe{^SjG!!x6$x(k8Xx=R~M|DEcQSul;RPW+}m!FWYEhps7>gjBIsR_xUs{a}2 z-Ya!SWjR21m&Fmadr{|pd#S@f^|HWlWM7K&tv8p7wpBCRT0qfd zlYNu?RyB1bEAhQhYRx#m8k9WC zw3rpfE$}v@bz~hTP_M*%T0l!^Rx?hsd~uabl^VMPR+n?xL<+`^$INMLwG*1vjMFSQ zHfQ9u(5%RajJ~1|k%Xz~Xcpp@1cFOCsfl~3{4pYxQ7mT2m}8NIsp;sJ6v0TAN;Qsi zt|h_|8IzfhB@FG@SCXM!o=Vk@v#`V@f;fR2kx@G41jNuS$HtNj?dPCG<5rxNEo*3h zOh#wf&|zlWa_kIc#_)}2Q!7Q4Zlq?$tV~9!h#?otj9ZS4VW(#ERH|+C$zi?&N3o^H zJ{A+6Es`0x90^02art(*kQujH8;RL3BN4-m9eYF{y{a%XZaLzGGGl-W6YN&}(PCm| z%ni51ERf8&(!pykkm=(&5gHahy)#}DhCvB=4qXHX= zxidtAXMtpgEyt!{cIZD{NlT4ey`>k_Mrs}?qGGiFqaagiIc^2-Sc#6qtlc=ndbT_c zd5d}`dlPpY62Bs8=>x9+73%LO7Q8D4=pS3r13c}^`@=iR+1gh9th>l;Ih^YM23ofr z(OTy6C$bT@w>I8Y0)1L-XZ^*2E_>i5C)(<_L-ifqVyj)QfyA;rt3Fz(S#&$H$2Mz1y*dK*f`fC0j-QI9va!gJjK;E&!(^)Cs260M1cS>-shQ)?9hX+y z4auk-HFTIccO3;2V;MdHsseCYE?%E$b zCf0O|P|5T5p5t2R)F3Jluijq#+$-@2KpvQJjaU>T)!02pw@`^201;Wwy*S@WEOCpD zxs1+XHFnPtE>w-pj?b*{UYv0yR#gk5GI~eCfPD^4PIr5ba-luShY?Shd+~Q06SpmP zJVxQDbBFIK_Z;CurG0R0&d962*O*w*8o`8Tgj8eq9NkJ%+GVbURQKYCV9ToHM#L(` znEZKCq}y|(ONwA5k);}UHWfQD?x&Tz;EIUIbv+#94;W5X}{@TqQW z#YS^)@&ZlYu`%q_KvyPY_v17yIeM`pE@@!;o^s!jFjS3QeyvDIko#-DRwA*un*m5p z0{V`4p_2e01{3Ulb0hHpfD_4>ZrN~0s95>|vv_gfo@71u!8y7ddIvHOl$ zp=#^^5w}3&XIP1~)^IFF|48g*-c#;7Hia@{|Jzb&nQ_0vU|ZJrHWD*q5h}@y`;J+m z%s48LS-A11i;0=BJn#}zOfuuXqgRM-01&ao+$(upD#45nu!Mquu95eY`;KCv%$SA5 zE%5$EV)iRYX4H)==Y%Tk@btDHr&&J4cxK#>KVwYHjM?#+$%w@dLbC>Ongz#Zj6B#% zCHoPnj83Z=I82xZj%FclN#MCWm81`YMvU^JGCQ~a6NzU84xTNNFby2tk|G$%V5!E< zWdawGi=m58m7+6Bjz9*EeI*&%C7o35_$k-2wsa#h6B1p`;Zh<&l0!Dg6_?DKX~=eL3_CR#n+w^|*C6d>lJj<$)t%C^IgKKa$^RI=_o;%TX(o8V8Vy6`GN2QR9bK ziMcTxjG2K*WOnt|fvS#Bq2$>A^ej$}soL=;h>5ALI50B>(M?y^TL;=ZdWF*CCw4=zf5);A>i{UcHBD7-rP7Y5;3YJRE^qy&a-g;2S|~+k!LbYd zRPXW<;F^NHK>ax5^5s>NWNPrWAA}{cG!~;N*bB6cb1o=}HMCdLm6$d}yVMmZ9zXe7*6rPpSgAMzCB|+zhL&V(Kdq%YH@xcl#K*?O)BO<`6_;As zBz!jhT+>5R&e<&7RH)9~GJ3=O+Bz{7>_#tGf+1x;W< z;b}6tWHp#I95F*lb9*z9j5p$EUx_I&GaObXzLNFvPoSwIV<>F~Lo>v?(TAM3^^~16>{SLWy%JK1Q+|UI~5T!(|{BBVKd@Y3AH;Yzt-10TOnhMh~~-;^@>n zz91YtHRj1`U`)rbQ0Cm(bXVY{F4h$;`ASTyMO2LTBy(;!wuLh1=s4U0jvss_X3p}+ zSjDLCn>UpM>pRMY2nawCUDShAz8Dp}gl4g@Xx>z=YVF-^k1}T#9JkQ*HTUI55ECCW z3u7=-upa8=-FT0vmk&+Iob^@rYyO60B7tKuQtH)=Q!glqHFW$`EHSHy49QHVXm^@S zH68UrjFVu9WzNl3>Vx%&SVpx7wT?e(CCOCNQ7|c%krb6CQ_XluSYm#~%twqx%LzL$ zwqs;T#&&5eP<(|CIxf2a93Jc?=8mr*ggb_oWNbgL1v#;Sn@*_U^t%nKKK3UFdP*m6%Km!!nvnyK;WVSFl&P!!ks@d>lgNtZ%|Ej|>ud z7Au|6Z^2bVsh4*HKB8U_4{PXEA4DdSvj`|WO*ENmIqH?9b}Vylt^M|iW#_@6q{)Ae)o5;A9f<^2-ziU`K$ifzGpIy)HlrD+&j3D-TT^F z>1L_#Hg~y)+>FW*Lxaq@?asrnYXfb$@ZFX#oj?!5-HaqVKMM-a3C*0_?ko&t&fV0^ zx$QlnOiYs**%*2BNe%EWnmM=K85qi(0UCyQx4Ws2jyaj@QL%Iv`pmiQm>0^NOI$IM z-Hso3B__vUs3o+gnRDA6bfL^S7?xeA@guIp%oz^GXc>vaT;}ZEV2?8A-r||FzRP~i z526wiaS^Mz9lM%Dc(>Z4%sDENS-`^m_KDA&<&l?|WSTj9H{Bxw0#L*jwZ8K{wi723 zkm#VxXic6ucO30PnKKK3Tj;_)_=%abFf5~aWXW3&POG)O8}SkK@^J{6v%U#G!h@pE z3{!a-iw4gJ&78d(@DcTb@>xT3r(&WNq6;IDkVT-7vJOX4bnE8Id|d@BoogbIB~k&H^Q)f$^*nmKnJ8AF*f zfW;8+Zgb68QxkCxBF&t=o9$8NTndVj>~5=!4XH0Y1G(#&Dw;Wax7nl2IY7cL)FpV< z`f9i#NLsD!-C&P0=l@|Q$PzPWVOXr3{Pc=u&fbmq zhW5-ML2@03S z0>#(V{lCk{+7wEi}(tKu`S2Q-soV`o%arg#_9nPHf9r%&u7k~Vk z9QPSC*!k(fr*{QDj$=SR8@{8jPpO@y%$wuK_=%<592iDdQmx%{gQs?nX`AtJ<6N|8QenE*V`vPSB3(a35sm2a@tKZXYXcvlsOL;&z$uo_py_`#H?Bz zn3ILPqhL)*U+|Y?hLRU9K z@~+00Vp&4w++VdYe~D;EVBvY8nX`8>ev+|WP6~`YNIfPY26bXF@Tfs!TL(3^cRhZR zvE!Muz9E0j8&_^he^6#dB5aN;8S*a4$Ke~~k~nkLcjU*^PL9ybfQ-tK9RpNrdspOR z*9L=g;d{_5AEKoDI4hq~PqHXFkf52fcR@bNoCm3y^I(Olu*A%n!GfoUK6Cc2#z&bm zkj@bALHzhDF*)YO!|FsM8hz&M-G+}c=W^^A$rdicPfUZsP|IleTVWCJz}4QP%sCjA zU8r8cS7HhcB{M2TmMEJ-tA9#Vb%kDEaeaI=+dQ01_Qyw?QR8{9H@;4tAMWgrrqk?k z^;q@_{ld)lTd^EQjYuFxEL8m>(ZC&AU7;hh#)&^Ce0{0blj`y8F-9#boKY-dpJG$| zKv6sXte))I-t5k>kzJVoe@bceb@atcr587@sE3F4!nWb``O)Ecnw?kYw(q@CQy)@~ zde_8%cyzQky1qYt-(hyyUL5a!sb9LWKe;~IA5LfE!@8|vpN~FlY+O?J@6PTVjE)_uF?%dpplajA0=lNwS%P!~h?8#%MwocT)pw4ff%kI5!pZb{qsz#8V zx4PFKE<}HPym#aI*_6z2b1Tjzo-EP6Ts@k-)JI#0WB@--06(NIZC^Biv->W5>b&}m zS15b+>Ct#Pn;eaeS&w$JbMKwJ@38*hM?90(zrz2^Gv~wp@54s@k?P-D^~=Ez^p^%{ z(K zr%9A(5+$0n7g3@~lxPwqnnZ~vQKCtdXc8sJ;4Pv=izv||O0=jPqC|@*(IQH;h!QQL zM2jdvhH4Wf+C+&qQKC(hXj92Vi8fK9O_XR8CE7#@BDh18=ny43M2QYjqC=GE(6NXT z9il{sDA6HG5HVe%M3*ShB}#OO5?!K1mnhMtqZ1{%M2Rj@f^_N;C3-}O9#Nu4l;{y9 zdPIpHQKCoj5G8s<3DTrbl;{&B`b3F7QKCdfG9B_N(=_|>JHuvHb1gUXl4^O33q+HYgB1Kvlaf(4u9x`KXk(% zdf^ZK@P|;Ba8$inRB`FOBKRRhu6K%{MTlJQ6G4j*x!xs$79n!IM^vp44x)F6stCP7 z1V4nx_4W|7AVNdrdT$82gvj;I5VQ!9hcflHQ0<|2h2Vz}x!x0k79n!IBLpo%kD8 zCTH`1Vbj0geYyIcm#QDEsb_}6y@Th*N8^K?@o@IsaCbBt4aYAWj*s?k>9Yf!#NQen z4o3&O!|S8j&hx{k_eZy`?~dL)Z1e{DFiWZPqoW(s;ntN0>JO?5d&POfmHX-sD>eUX z*a-e^27kALzuWrnKd!F4e{?h*e^B3&vOj+33!__y`{T{GPY$ND?Ed^8`f&dK{1HNS z;oA8NI_1soshllvXBpW&^>0>nPPjfQIacw_xp?NJMJplb6-D1VUh(a@cmZQ%?!R!K zdKH*_`cszC-|B_C)t{vQ&N|1oJ&D=gs(&@tWY|b))J$pAN@>(C8=2=x(?jFd!<*wf z)9hY#&&EsD`48o1qS;0DK=b~K=06*kl{&b6YiQeM_g=YYaE$uy@SKs zv*F(EG*kBf-WpABW{>Pmheu<*aL0DdWEb~G*T?&LyGzsY4YRDx9~)1RqoY@-uiU=h z92?c<|F~!Sp5p&*Wa_~J`b;0*S?sS@4_Z*gSz$~-{d;Qafe%fdJ(%gsKF_^29iV$U9G z^*>uv7Yi_Y#V-%1u`~o|LI)CKA@H3ggV~ z%V{Awty~V%z0|z!E7wMH+EqNUDZl!p$03^8OUqevp}EqeSK<& zWG|TyQ%3EIgoY=|6 z;lb$EIJ=M)+nM?==hR4T`CRNBnutxuv*C05qgl?ubWGnJP7cPy=l1r;hJNW+VL1Br z=x1uRzgBzXL+|>)cxQ67Tga+^qq@8^xpkE0(wX)l{IvrE@5YG!uN2k!gSr!H5?87FY6LfBdOLrDU@GSsFtS^Z@m zs`6J>ixFJgSAzN5eqQt7{jsJNmp>BxtN8+Wlp8{DF~)zISLNAA^|7nxF1-5kgI8av zE{^q9nm?P%9#Q&fQTgZg*45uQr!LVbJNrwu_2J6a=1$pXYAbm_wl)uFs4sk@de0Kh2p62rq}Db^>soTp zXU`ha#c(V)7O#ysT$`ctPky!QNt`oyUK&_1d?@nfekwjTFXCS4Qo((2!Sf|0;^ zU5Ow!{I{LP#w=-T)nKe>^0W}#1-H%Bk*=~PweOt?sg%XDO+JcXweB>LNd z^_VAUWb^#N--7h$-+FQy>w~W8aDVS+@yPGA2LGU`f!2fHJ_XJ7u(wWb*cbG0H&1Lh zJ=npq4W`cnruukwcr-e=>0JT*qMpj@$2O2l%8!q>7wNiA)Q*w17lY^n znDre9tm~k{X#UKiCe*pX?eTQe__*%=<|Vu54=a}GKR=Kw#(&GA^70=5~IE-iiyrx!mw$siRUZls8%^!L~{lF8) zxSP0>`4iZ&>c_rg+0*>reC)+>0&{i{-MUWDXWqOkh{fYyMyDlplYcb`Ji}mdL@2 zbas9aaVoiCaY|{P<7EG04GHdxfQDp$`rcE2E|;B(G0Vqtx@_?mYbO!jvJ(iQ&Cgy` zzjRUUE_zxN6zB0fLwPu$tB6iOpW{CsFqKDt<$+_|dhFxp{87X4opqFhG}PJi46n+R zcVzcGBm3priA`E|0xt076wlOWAV8q9QkyDmR|_icTHl2>NDl9=)dTm4e)6 z);fimBH6VeGqbifVeNGa>#S2)PYe6NDy~c8YVN=5S?z0ytfmxUrFT;n4(~yJFt4UR z^|O>0+y0XM`8AvitTD9d0Fvo9v{!MKY>jYqq|Lr>_0yjqoeX4dWDj2u& zUqy($4R`7gUIeqyRCi9s4z38sCHLQYN<(W8Z>W!*_~7>Sr=yb|S4X20pJ?yFmipvL zD_jnq+xdl47~WcXcJd=@TTf4BRLx7g7u*g`a!{@02^kf<*!J<0(UiW_cH#ri>zF=z z!WwJy>I+5nRXfXNoIJ54+`zoZrY|`>fuZz3fAUxc!Y{DxpVU}6SKK-F(e(6vQRg2?BFDBg=Z(a4wmjoJ#1o}J*1ko}xKe_zGFPcwsP ze&%uY>Bo<02Pn_ivSZ$1>c_nS=xUB%f92V^ioWRR>|Es}IPE4m(m6XIqQ15&(2kZ%TyPb+Ou;N?NDduD(3Y^bS-bXC+TTCX4-cA zdHX8)?n<2gpRAY@obCpceX(|`&o#2sOo_5zs2#WCA9p^KeW7-W&mOZ=JmUDr3m$O< zhchSccV4kk^G7bKAG)Z%Ws#$Vph}Go?qw>jOI?mRsvmDMZpveR|NJrS9A#8n|*Yo^wqS@xZ^<;_&iakd6!e)_U{`wA0WzJ|3rl$AZk z>^lGTVfF51)d^p{dUh>>@+ZV=5@lW4VtaG+AO2ecwQ_Ub^aZsMOOa2E?{1uu_hboD1G%G-LD?m zpXd+!cptv|@%ytMy}$A~y0Z`86(7sneW=Sly!59seWZkqr3aPe)(_Z@C; zy|&hT+;0Rxs@v+x>-yr1&GD`4v0LxYTNmc zSMaRo0|HY!AvZ?1ZkZA1rq+R%nmTAcACxn-mZA0h!J2AXMGtl+yZVEo`$o&**5+t7 z(;qz7v%NVT??1Qs(c9yrJDWGhM+al`0n5XA zzBiis$jlsd7V9qJvU8S-OMj{!%$UB8A`O@8F5=QM^?d2~xZjwfun}C$*4&{68dPhM zzH>;NcWHqOcvpU?59Z2*v?TBH16B z_Dp{{%TEi z)3kPhvIyk!nbWCx{Y}frkPvFY8sU%Y{BBX{VdU>>h)*viM8c>gYmDrw&`e>h-`3mR z8_mX{=asvGL5Eib+kFm;U9A?^U*KXJ5FD;mV^$p8n9d?5YtQelse*6>duPRi$)4FX zJAmd8hS$j*B`ijow%ey`>b0`SaAr){e=4n;gri|2zE0T-7Kl|NYp?Vj6B-&ZEY+?e z9vc<(o)}fo37xOB5|D|bm>U)L4f=|RYy@Q(S0j;eH4Q8jMxp{FY(^@-`WexIH#Lm~ z(dv(M7jaoDGEs>Lc9h9uf=w45nkHH$`IqZ1;W9Ect{=gi%QuA((+yKaq)QMX;e!KM{_p|2UX zD%q3>5V2Xls#y)f3lR5$ z9Yrj57ht$oATpd86aNu40o_C#Hu5vvs}#ttzm|y)hr*F>;q*9e9~B^Ci*kELbU0%iw~ureaanH9 za4%nQJf`=|&WE^{AVS1vcfSOiR`7*0XLdj-W)J})HoFTl+zTkq#{RBh)rfnELL`hv zvW{yE_lk-S_k*5EN)Y#wrAR1^=vee+$&YKd)aWB|W!eXT9`C+(MATpSlzKQ>cdx35u4jcIy?o|q8*I(bn z2gJRYt|A^A85r&bgobOaZ^9|!UR+-hQ8bRW61msFz(R2}DnPlJ}(`;J77y6oE zt&&ZN01=zztqk`9iu1AGWNnChi9$q-Hg{vVS5!RN4@}aFxR)$NL>bL|TZrQj4;&bH zcy$1lW4lYZj7;r)Ooj<@1)euBmPOo)?JQ!l7X35$nPZFLUJuUebhgZ9fXEl@IS2Df zu0Y3#(=)`duRvxvH)aPw>hjv(f4MaYY1_OlFJ=Pi$`)m5ooSe&&-X&2`EdYnV8d0*DwJVlc*wW1)XI~(a73Kn|g-@so*C5E+>ocp$jG1VaAjaWh$rj5C@mcMyjG zo@fB#bpjYv^|*}9;*08GHtGS%p$F#0!-$(fy-aU$zKlq{e=*#wBXP65pbl{}*i%GP z)Y38BERY$_PF_Zbcnjz!BeA;y!_7*G{94QlaS=CT`ih8(DqMz}0jXr&A)Z+^oYzsYrhrSu{#P$z;a3D8=@cQQ0WP?BGPM&#Zj@(=pL1 z$l&2wF#E3AS6C# zn5|@4B0)xHxh=!ZfbNm~z~p#{n~73Hlx>^oG2ARFpX?DGlO`i>CW{eq8riXU3^)w+ z#I6B|*9qHSL}rBY+^mOKBk@3EU&PJW-Xf}wwJXETdJ;G5n*9NBGuTr^W5nq>Vz^l# zGn}2S*%J^q1N~$qb~j+SSt*fUi(L~f5I1A`%7|>7V7M8OO4eQ11X#q)xXv=BXk_iD za5;E6cMiz}y5Lq~5 z`AcRfO0m5~RK?q6%nnZE`pnAbX88-MNq*)_5YZX23;3DiRuya}dL>!Y^_P*^D8_Iz zf8Hv*1?-51)go>t$Pf{BEw4?mYy~0lIm2uv%MuAPI?HVtZU%IZ><1>tL)=W1BBQjK z9>dL|^64I7(qzQVWHB<%Xs+Bx42XJS{#rg>C(CoQ;;nq<+zcWMJurVMA4ffa%k&oK z%ZSu*h2ds>iJRpwsv>R%dx~g^cT^c}7RV&CW0))AW}u&lqoig&&k2E@n?y{d2Qh+;vc zzl_Yhjv+ zuqOGLFF{0S#PZyXK}TlIM6V=kF2K!-H(42O<|C`{7P=$mui7JSCdd%yu6XM{!Lk*E zWcCcRl`Kmn$mlG$Ww;s8o!<}g*YXiJ6Q#%~ZKlU?v#5NsN8~T-BW@;(k#RjZEh>&YUr;;nr)>H*222j(yBaL>|;rJ4FQZRQ3(>dqKbfv z1&r*_tBf6CEQs`%kQJ2#$;ceH@^>n|dUM=7E+;AVN{6pT`AZxNM^QptYi zGb?Y#@`w6qe&$P%&>6Az+zjRbL~F;bd~7CqrCHPU7m?X0#&I(r8Hm;%y1XJ3ax+1O zgs`XyO|fhRVYo^SvlT5%B#7uNx8=AQ(A`DfU_a=XTpe;VQHqGtW_lbqi^?Z^geg*j z+)NfD;w+#2GwTF!7{WO6Dk^53u>B=uMJ*M_&DcBny%myIR54Kx;7Z{jMx>4_95*Yk zd=EZ%HmBL9T%ZoQ8SE*}k=+?MZWhQ4)@Ro+SIApHKM{%D4LELAN@Uk!*F+0+NY!Ba ziim8S;J6u(8m_ypIW~aYjO#37vL^;NRXwNDhm*;^{)%PbE9GnC`Sog#b1mm%u`fEC z?%Gvvs4G-N;RdZ_49D3p1B&IhafR*dwx-_9TA6XT=F)X2ViMO`qB z@0ic+KB$zspua+|0c>wweUCan9er$kG|k?cy;(i29yEV#zI}2q)&DkoB>(GKZPeM` z?2ew+%lQvSn%BlV=ElqNID*XGe7npq(m=$7={gTp+>-$#PlpX=mjXh=EumrdZ?sEseMLkz zNHh2IM)_SpX3xG$mD`r_HLzm`^LN*LSBUYo=+`NNxynnVQNe*IL|P9;r8JY>Ma73} z_4vQT!T1{Z_CPRycg=Sc7+;It#vRU;Neqy$i9#e_>qk!#qT)ke%j@OPNrEgzL>b*f z!A;A-{M|LTMl+RE(ThflxhnIZEpy_BoK2K6HzS*l8PS@1FqKqM`Eze`Ff}`e2UoHf z8K*rtWh!8p8!dI-Q-ZsjZo_1JE_#7l=yPYc#Wf7Aw!9qOzzVL#3g_@{y5)=Ux#-0( z;e4Ilg3~ac?c{w9T=x;q;oWq53FC9o7v9NynWwgBZxyA8d)AlO+x^Vm3SMIl=kRWN zE1U7T=!<-0zPRK;lp^|^d0@fwxu|@?=kg~FD1DH{$T)5O!1x?^Ju95UyXkE+#^<6h zD1`HM_BKYtyy?O3t)T2YoWr}RieCHNS(Rz@iAo+X=Yftp+c(#+TAw?s4%}*Tvs6?E zaz3}n1BbKfnW4E}F5+|YLx;Pl>f!4D_??=1>Fj8GH4 zqW|S>^<-XwFurwtyt})1aAUgp@yXHd(EY8CndXmG^~~2_Ei|8pj5_mWKx@5kuc_CB zf@YJOx_C_0=3$9KI9cV>-mZSrglQ+^muM+b0$kn9WOGwW4aKEOBc4<+Gp z-9=pX)ghhGpyD;6uvZN|cPMiZf$F;0zH>;NcWHro8%XsozmD0TD-+U^yvr9bhsJuB z4RWbgbal)=G868StmwLnxNP9lMaHN=)3-!XU=viy&24E4GSKA#f`mCd&Z3q~>X!j< zvBzjOGxc4QgX$YK#j2cRGevchK`Yvs{t_}vO&X5-<8P1Fe`zSXA-6R~v$D)`5kW!#?%Hn`yJ}Y-kVH>k2`gvxIahZ=b z!4Mi1AX%^Y1QfIS8PS0^<==opt3T3R#ARc9DlSKvyuG@qBM9892`<-N#1)N7P;fX~ zTvTE^i45@+aC!{^hz$xQb6#{Rrk< zzRDF=LUz9V8=i=C2_i&%ma8V%w1O}6HN#dVn-T#cHp^F;f;ONy8~e`+b@M%(T^)d? z5D(11cZXL8+g-A7@zFbmd!@W@mVfyUaWA&Bh{@*Q4EHL}7a8uAe=Q1eFW6DUVs`6oN&79RsF@p#YvDsaa;a)&-Huj&rxDw-DO?19QJTSkP8?TO~xmR&HHyh=E z;L!89Lw%@}=`4DkJ=JHpSNW0;!@crbZ4vi^9VINqRkjTG3Pc7o(>L)SaWBwK#9<>p z!@Wv@?E1TFUVLKWsQ)@pjCJ-X3JcX4Y?QFS;ACYyPfJ^K6UbEi+k2caQOlxZ04d_ zeLA1&UyeNa$V{|KvZL!R;<8bT;a36vKEqfM_Yy=%_==ASB-pfqFZ4CTTBVpl z1c=xyZ)La_P@Iqb<_rMU|BFIIj5c>;irS*$=d~vD2Q$b2A}6~#0LLL7nBTjNR|nf& znmr>^uUa_XWWuy%IDh0AD@UD3kSXY&xx()yBSj4`Z|8PeWQ3-SN@qT2_ zCUjK^x&n-mUHFs3#vM*X`pd{{lut$GD3sp09Ww6GBSGf+%gAhWVm_Y$C6gKB!V}wD zMrGqus-O8J%Ac`@iBo9HCdhmVB03{hNAy%{IwIv`+cJR*Z8Z!wKdkdBz(z1#y@aa1 ze90A}QtpG!*9?J0R7{W|&fS*funCr}AS6C#Xsl#eB0)xHnJn{~640IO1#Sk-Q5~XV zq7)gW&F8*Ut~y}X31B|d6Z4{AyiVBu;tUow!EDq6l0y&7OM($8YfGezUA;!6jw=i+ zE3by&SXo{$hgccxDbJDJ8JLg81Tx9&m^}dOic)NE8I_Gvsea}&tICWSHk;&Uz62SajasSJ#N4cDqF0hNU4I#wjbaQp z^X*aLEp$iBOMVeI6J*E;Ew4?mYz1NHbEbG#v@DSzqAO~G8Eyu2kL(8~$3xsqlp>?F znI6NO3~;0S-ewG5`20UMFmS(ay!EU)iV!B!?cDfA|$gJ*Kye%4YBkH!H6iR+B1)@TLd$CtEL%ZHe9kai$+AR(jLvdfhMNK1Bm05L@zA-MC`Cm1C3eLuyG{Uy zp`Mt3{1vYgw!b)o#iw7{s0Sny4>b10QIF{@qAEW7%5bytN?wke<)1!8=Vq{{h^F}H zA;ZlAnc?i@pF2d{4D^$c*xi8PW~D@aE#6I66*H05zr8og&GMJJP-Klt5N&A0^5+?h z?9i)v+<8W%zl_Yrd?rc}$&Oq3Oe6om4~kNe{xULq*2qLDC^?+5{3SCKrP$sgs^YUq zsea}&tICWSHk;&Uz624SdGB=rKXcrwg3U16Bx}0;SpM9Ml}>!l@npGCIp`8Eyv9MfL-e;~{P)N|908Opke{8I@0-X`aAU z>}=En9EN&g{#rg>Cv1Ol28*}y*{BC3haQ-}l#inx(_2Pm?aFYo^6Gkyo8>R6B5nqI zifD>=R2gm-$Rx94m@DFDpr44Or-kX5!Em!uBEJ^zrYrrKJ=4F^Kg!L_%ULk8MkRk!=xijD88f>Oic)NE z8I_Gvsea}&E1&=5FR-E=oFMZhi0F)1o|`eU$*h^^m1NBYxLNTgE5prvWEI{*cf|Zv zd&JEI8RFa(Z`~(Ywt|q%o?*6TRX2I=-W>h9PCVX$48s{qdWS$ z^Kl7tBXaNA7F7f~7NBp>bFvG!DzTZU-(3Kk=`SL)F+UlZBUhSgl*ml=R2nkZUqoil z8g-OHtD4T4N6t!I=A;q57s&S(QQ0V!>}Ni;@@GtkmwK9?`4U8QHfkkX(~&D5n~7d& z)^z>QFCmtc-O~uO=VplQIf>3agKPUu5kMW?EW!1a=BB7B zVD4T)Qp0s;KIjGyG}z7(rs5NETx9jHx{h(PyoLbAm8b*>T~S5A#R5il=vBs!Fcw7m zOUQ~!f@EZlTY0;AUOknD%=H(M#iJC_8E~__atcN%wzr7NMyX^!^O==5V_m~$)BMbr zAfYp2?YSAu0f^R)Tlv^b^h&d)>n|d+QH3<+UT6PjY#3c_%e z8fGh6mPioMS#Ha5GoU-WADn#xbN&fTKI#DuLl{S1Ma8TWw!dWOqLzx|X6&8(-U`Vp zs+gz;aHVh%BT~l|j+>R&*9Qk@bDC|+1?rHS!JgtA*`0yoW`WFLeRd6Vh1?AE6Oq{6 zK%bjIZdOWU*Wz9DDOGh-)pII+IGOD0Gq1o`%J;Je##tWcTFx!PUiV_UYgfIYu22z$ zE4ZPIFFc0Z$pcy2e^gzz!F>4f@!pN+XH&frAFckfX+J_*4}U%yk9N1Wnok9-pFPl@ zP#x{>eL_#t?CAEv&S*B?9Ue@MZjJVH-M&dZRCFm6%Dg}4gH?VjMta9Fw=v%h%d?f2OK{Pd?^^s!=ibus^Se=?mu`g>pe&S%smb*}v57v6mF58hdb{O!*@|MxGx@cT|= zPj%EI-ltm+N3-XzUA^!?cF(m}zHxJQ>u_^?sFmNmrH@%xWaHR$_JZVEnUe*4QB31M<<%$mnXMl~$v-(Mm!)5|!LbPMzIz zRvI)rGt8Woc7ZByc-4)KDTgE#l0fnYB;h8G9Kto%Trl8AL3rg-!Pi{JK+JW;CdP5a z4=QD{r!IX+n-)KQ2+14zx4W7UI~Bvy_-Ms zjr*#rPwh@NwtoJ>7iL@Y)#V>NvHFjPx8Ejy@UvI_61(z-k6tc*?v)?>?&o^!m)K7| zySaU8vN_(H&vr(SZf(prCYu{?o$Zd`*Y@7vvirBU_U6L@yX=L@=DFG6%C*a{xN>md zsosI%PqW4G>CMUB-r#yR7#!Z*KD~Ne8*F$LyKFkYurnLanY|j|NvgGueG)vNheDJ+U*qiCqOhygQqYw|?_( zv4K9(eS3Qw&u$G4*T2J{8+Q|pkn144G zlQf%G8~#1^)A}M@Zsv3NT|M@4eQZ#T>FDH(8}rl84HoO)tH|WB|9`mTDET5gxOxDz z?Q+&z2RmSF!hXV8#hzHjcs{t|(J2`1#{5F%{!qQd)uU=CFB8Z9Db8q9__z_DV8)gl z8g{sqZcMfDBg~9X^UFQzy}Xt3RU>;~sBQ;drv5iQ_L_uYl}hQ>UwPRU*PQEP z2F5lKu}gi@#6n`J6pr6-13!E^IzFnX4|}KpKN@?vGDfwZaqK_ru^SzUotW)DI~!ak ze)AgH_$cy!NmjP3fi*rp=r8pHYdFcS)t4}tPDd9)TPjHvPqCL-gANw%n@%eawUxVs zmImK`w8QF+zgg@nuO3ZjFKnEit=|9S)2p`)Un|3bpI!TzD=r@#So_JJAI;D1jAlEq z$B*VB0GXc|@6Jwd?@o0CyWDGFFR<&+Y)b<&NhrNJT$OYWEcKsC# zgM)+q6}?sVBl*9cKIg))a{TfYY;je5#Lly8%y|cwKQ$9mSM}5tx-NS`huGqJ6I|RD zoKW)qa`wI+8?4W^_O^FVZUSo`n9L`8v-tq|x5`?gJ-J7U!yDMuNX!KyxO%<$cOIWX zkjEC5dzZ7X_t-TLfecPecI0BLzgNiO?8V%`4k3dN4>Ptt*?sohS=izB27PwGx4PG{ z-kHJCr$+47@pxnF%xrhIb$T{#7Bm*JIy>0`-I`3t+gmfCFyob@Ud0--%T8~f+nRe+ z`6s0&4b~2e$kwApepd+l)YVVTc4m{g`pc}XPe{G@S9BdY;t3Rc_|pPYo@!~++5E>A~uCg z40q_p1Z;5N-|rbdoOvhs|N88wA3VP^+uh)vWwHs*>%Oh&8t0lc$H@O***c-|yh2!1 zSrpm4@(HA3*F%)NH`{$-2E5*#ZfrrOJ-;9n;JEFn!TqgX3A@^n!NL_6d-@;wwsk~o zUPrWl9^TL0<~>i{%~o}{37hZ~_LcEWbtkahE)2oc>q5ebwbw)Bz=>H{yWU*UnR9z! z7;A^Own1@bn=rwam|*QEqzUoW?+TYYu76n9Y}i@*>}5~XD^rGV|3eki3X`^BcM19Q z;vsgu&c-ohre`*u-5dNod+lJzIrGBo^n81_;qvOaJy4(LwzuK8GZU!V?9ErtOg8ss z0pYQK(PM{BY)@yK(j<((RW@(9uut!t8=pEieHJ2;{?Y-T`2{BJ&N23}9%GLRN4dTG z%<@3{b7>?9;8aO?9ekPYaF#cAVeAfy3DOF9oj$J!=nW(18L7rJ= z>+DAPxApCl^WBZDXX`TA!1M2wKDbBq*#FjJFO$PQJexm0J3k-T|6mN%ossuBv3Rnp z^Lr`S(+=JbMllfot**6bTU{Pr!>(B0J-2mwGKae4Ee8U=>d26>gAeS@dEKVR|9o=y zH2?Dgdkg#NQ?uRqXm!;fI>Y{x9;^Iw z4p=*X<;X9x*x}DQ61kSWZ1rlj`M@6p{p*}EpkHP%3+xL$wq~#DMacNZ8fV&y=`JLO zHRqw8yS;mMvT3Av!=I8{!}V)}Ca3Ga@```<6x6O44jdRRD+9TOy?XF-!ayREQw7jx z|EkCCiwtHI#~n(}&L&$@>^De44y~~(df+8)_LzZ}>Bg{k)7_y?WK);1&Pu zzwPnxZIOEqkvH*go6^=o4nx_$>a*jWcy;#N=3wE(xy`C5%c~OlM`2{cyj~XbKH6t@ zcVS*#X6wxVs=KGzcUFBBj)Nks8Zdb@^vrUyK_D?&j(@kKF9=hcBRsxoQ-W@@9rjc9lAEHf8Bzoazr$F}rR($p`;X!H&@*LQFb;%hRd~UY zV+%vGko=R*-U2E4>ScUaIl{&+7apF>pPNChG@6{>Q2Dq~TUi(!ScSuTZ3kQMC3o!K z_1Kb4w47mZh5S=RQlHAlK9@6cU_k`WMe9NZgr872h8>j0@K8&FnmPO(;?b+wWqVjzDa8 z5q^mv)nKb%`xe*-;rz^It(vygC@=e6_oTkA7cDiaXKB>WIouhr!>iY{(LU!8KMHun zUfz8B`PtT(pBUuD_e9?GJ&8bH^7tRCL_kzX&_n=TEBkzpJzRh)Lad*?M%^&|5&!?% zV~-Xf#Eknj`FLlBLz||4r*a#S9AwkWKG|na6(Ge5`Q03#m^yO_c`|`lIJR{)yGj&k z`?cdi^MHKbeXY+PD!@CH6YPT^s&aY?JGR|JX0cbsG{bzBex~o1G*D*}ECSqC1mp_Dg*X!dhAssoiy?XCy>8h2J(je7%nQR4oH%bO*nh0&t6NS%g!Tc+;D3Y z>00N>4=8-{Wd|KB>34c;NMo3lUC}R~woH=98 ziFR=d`NfetI28&N@%^umFSKi2f z*-eIw)m2ehRB<*+HRuhmV+TbVbk&kz|U>eRv*vyjISsB>rl?OV%l}dobSI3 zSoTV5*&deq&AVB0%M-JPK+E`kiCt%l6518(K*p=D&?sqtB-7B9b~&^KoMAs}k2TqY zZi)GLbGElPp1tK992lLK_IaJmL)h9Ckc%H!?O!w+M*Kw$(mbk56Pf3rVenp0^I_;1 zjuZ44(DooFZE&UfV?|*E<%cf#7AIx1z~0PmjxSQ&0g;@)R&xG|wMF23bFo6+^jPq8 z!@`fTe~E;Ku88m3gcAz*PnE*Gib{o-BM^Z;Kb!50>nam?7|7MuH>T&gCuG~~=i^+0 zx-mZ$HkhRatlrjFaBodc%{j`uX%8?+@NY|-5xTCMZTLAk+<5IxVz@WLlBRI6XM5|& z`6G89J96R3bYpsCYkPj=x$VvAk=f>qw@%&q;>O-=_56`r;`5$gz3m?zIcBtzd8uO` zN+(4&0uGN!Qf^;c7NorP0%d&{Qr7n(dtIF3Mz+mwqNVCK$!}MDx`XgD2AVJOW>%8+ z1u6e98o|ZLvhcQ$A#N6zdK6V3eIN^ zzCs!N0+i(8ev{}Hy!`To%NK^=P4~{7Jqs7tpb5FZw7~v#&n;C!I48VPUOCyHarr99 z2Y3mYx01Y0RghoHhJ#zs3No@d-K-bb@AlY(wjLoL1?xm3LMgexV!h>D_WMXhJedfy zjO=D-qC&i{Wc-Uhd$O}Eq9tUfo|n!ke)UyXqu^=P!9dZt13Ojmntid)9_Y+ARXr{) z0k`m(&T!G$6gm=Yn5^>2OvbW*iFD?qF1OCMiio5t8TDeMw3{h7 zzEwfTwRTM$99Vsmt2s1oG(-#K30BMg`@UQLM5wZ>#S~6dP74w?aH#Q;AXeb8w5y2a%D3vTvT)Xx%X!I6sS&L1#_feH&+{|7z_*pc^Fz3-N zx;G4Tel4=k_S`E7=3vHdsaa}xU|wm5y-y*o@Zw$uA-~>Nr?Slm%UoiNXiXxAhTv-7(PRJE4r|)gp4}Ee2%-Qq0)Kn% zG_TJC;6dHis!r6KlYy((u4+0?d#gS3#5r(;H?i4EYz*XlxGe-Mbt9X>v4cNA$h(tCAzI!WKEmt+0OO)9-za zxiNW8QusFR9F?}T8xz*yx6`N5{E}j>t3RgQ=H_d zHS%a0FUA~!I?P4*l~#<|c1y!@PH1zRZl4{W+uAt8{XRgS$OX7wpU+DpGGW4rYO%s5 zwAwa(-88(Wce`J528mAzil43t9$anw)oq8Ad=j!TiKi8QpcbMIz|ecy?0-O_5ILswaFSu>a)=vW2&S8N35ko^g@?ifQDK^Lf z9)`^tV2SF!%|rU%+!RxOV|tE=Oi23IN(fGPyu^YH)DaVDJFl~Sr)naJXyi@Xz1#JD zpy%JmHMkxp`jyk_9M+I{n1kj+;~R8#N?acXAe>(75wT6pybyc+vgEX-@@iWi)@_{z zz^+}u;1-9`##s_vk#Fik&P`-_I;3xnq0_kq2Jn^F*ITtYV|ys%Ts1i=&(h2<>NQHS z+qXADZ>l$5=uiCuVL3jjVF8Kxxi|b z;h0|p7a=U;VxWuQpGs^2eO{xFxQk!?&w3V9MQQ%vLmSYk*xVM;=N0PK(kpLp=szNs zCgt4YH}pBQ<9Ts6T-+Q9ah}VWLrMW)^C7oSW@l7(GAR-I{G0Cp)kS2>di0#VeyN6IwC_bMIG2rgySC z2EU*knKnEitkf_%_RXHWTuvV$Oe>*bvosn0?#*xa*m_}68c1NC*a(SLMT=)~!r349 z*_+A|7;W<058DW}TP+5fJWDaOO?dTGg#ED3-cXji%7yQ*)T({a`qrfu;d{Un*D`0{ z>$4L@`Cn(pSlK? zaNl`HvUWl1A%A#CVxToWHFSnmWfeNubx?$Dc+*bYcK@`;ej%B=yvDD=s3mjvlH1Jx z-ji^X6csqNG1)VsCvvmgeJb+1gZ;u_B{A~j^XhS5xW|qCpQ??3N|fOSY(ssVmo!76(EgIhg+>pdr}*)niBLu6Yfd$9({F zaC!3w*&0WYd4qA$bf8ea5Vpp(q$T?Xc9pnK%+LEK^TFca;Ny?H@%~UG(IoOfCn?l> zewLBO<8=_8ri)ZPS3%pg=uR?JpzJfS!zU^7Y{xd-i&L)9Q+OkIqo=#Tg#*LxJYZml zq|lm5nt2nGPLD|6yVIk~w8#EUpZ$}Pa*L+!<*h+l=|TZvbLQZQc*xe>@&NmOpFL5M zuPSNZPifIo$TU5d(&?3uPGI+G()B;>v)32od!4M8mSk1jRT8vQmEya*5MG`pg-xxh z7}2@m(4#Uf8GpA6ieD*hH0tZbIK3U(ZuC!NRaH`f_a$U+N%e&6QJJ!D5##oDDroUF z67xi^USdB4xH0!o{g8+^^_#%tl>A_9n;SgF5 zZ5Vvv?7K7Bc6oq(sxPkx(5SObyg?nMeOz_ajG#?-<_%tK3udZNfg&YyQa5ZV$IelD9YP8dH`tY^Z zOUCrdh%2xoHOUk8^qfpc!s$7AOwoc)D)A)@$*-uYl)H$(jB}fMdajdU#OX!WX$RN5 z6KP-0t6)jbugGOA`d*Ek82b4^xPoTO45_Ld&zBvdU{?{ znV+iXR^8fp&3Gc1kWo+1+rbStVASVA6O}B@wQF%pb58w$e0tuFY8yT&lW9Mvou0Qp zZy3EO=magjfO_C|9uglvwIk)E8d;WsCvveh?*aq|Nu!>g*Q5~>c+%;4Jz~?7Esl`` zzBVOuSIi@yp4UrplK2F#+@$e2S`s>wXvkIp5}xYwYcHe%0($@ERr3DLO>0$y%X09Q zhmXLi@J0A_2!0)hU%FQfW!eXMpL&CKAFi7;i)pVAkVvRI`5RGHq)!a0m2y}xZ#O&d zaoxbKTfMe@O0a*$A^6L21~>bbDPu5ybj#}ycR6F5do;&6;p>85{G|eTJ&0sb>N<*fu?*++5#aAd z_?52Ma6^H_YqgwKwNnd;RSw;m8m*+nepOv`537VyZFp2vS8gpH3c83PE*TJtVFm0@ zh!t3EJ-T|Q39Gt2$&MP*74wF9D*jz{Y14mMap4iaQla3)PR0 zkG90!;&{g!+4ZXf;T@IzawFyjJej_E4B(@;2sg(6asmF*Ytexsl%CrRp#y=A+@rfk zf5q^(6sk#d!unopgFO)WY4IeK1UHHNEGK11=bhZ<;pQmG+dg=H2k<5yfTx5<@Y0LC z%n8qii+E;W;H+5>3i)rg^%&p`lET$?`T=YITpc#-2~pbgv_H%Bd?DI4HEH-JK6l~6DT%?Mu;VvUl94poiUSBvoxay`hCChG>>Yhti zTq7GMc?hvw+8rA?5#PSzC{=HzU3rs(!mW^U$W>n;7G12#d99;HvmzCLt*;@)St(oM4&oBo%jvL7{Ld$MPp?DX$`~{^C;1aC*GaMl_%2P$Ow1uNp~(f= z<^3(lLVb`_0h0w;?ZtVLT#RF(w32K?d7BWK$VQ=JVN(mQPaD!pvdczE7SB{5K`|<= zdO#b_RMJbb%h@H_<R!NFMcMz%3x~{?k%CuFKC8?J5T*BmOdtY0aUFPzV zg;kZ7^pXfyMcRz2|B9hvrK|qbzU!7BE4qp$$u3pHYRWnL)hU%yoFMxRqkb1L#FBn% z#dM>dzUsqZ)ciqMqCEQS8;S07CGk8<6HuD-Z0Jon7&qK2H{quv*%rWV?*Q`l6mSq{ z#ar7SRQHYUVs{Sipxie?)i7!H|<5%3K z_%FxNBpoX4w)kW=RRuIgMn+lp(j1aJ-Xx?Th$CUp78xt~J5$?cpUlm|oCwj(zMr^f zHoba__3yMMJ(cQLxlT0Xzw)~0zZWl47=m^B=aZay!>v43YSjVnkdHHmy%_Delf)vA zb{THZa}bUojq{hi7@-|#HP0BrHyR&kl<7u1Z7g!6tBnNt(I+~26@alL2aK^D|0c1Kykab?F5K-zItmuz9@ zqt{3*nv39XV;~;nHx)UYQL)5(})E!rOEJ`=Z+sTPtGi-D6U*;vXuhPRybNaR; zg&Izv@b%D!(|rk(U9f zp`$706i4zodM40JqIIUkdWbA;>r!`mQ{qz4o&qy^QoceD>KhWYu9QT3;C(5H9?9%X ziIT}a8ox%+@KrVAQ*}!xzw+ROr)<5DCqG0pSL-$9EXc!$)e+<|cGuv}gd<3E_JraQ zJ%k$aE*|QVl1Yf*_cv=0NVQ9KxB`*kGyY@18tMan?(XNBy zvfCrdZQR4|9;{}$4Y4%CxOhVbad&RTDTXn81M=KZx%HOx7wp@8Mhwj?dhIfE9@o;{ z1FlX%_fU@Y8U6tjiWpDX47Pw-@)$fUM|{v!RSNro9(z1>w>i6g7T&v*ot$s)JiI%ZZh*uF zhxETy$&N&4E}b^^gC2V{d--oUHvtWMVB^f0*)HJa&IZ?>wErnZ!bSMl(jLzvJ{W(z z#ED%cpI?~lZcKn?zl~}2@KAfT*WlPoxVM8x3OX!RD+tB#3bakT8-G)% z`>W2>E!TJBI_fc-Uz|}JYFVgBz*o$MV$9}yiDS(I|}Zlh4R)QjVJ~uKn?w5KBCzFycbA|4c|1Rv zR{b}TtC!e90Nxibb|UoQmkjEw!{f=`cn-}H+q<&t`*K;@y=LuDgZkWC?Srb59|ZF( zLARAqu-9W>^3_W{BO4kjpyEg3v`dndS*YP%r=6+h4+P`*N@QXN2Orwl-1HQ;$)niv zuJBSSeZ2ZWQC3QZhZ%zdgxUG={KC#mpmTOsn7|3hlK(wL4+X{!;rijug62m6r?_rb*YbhzF$snKKL--^*iW1)kweW`QH|^)=UAr{ zN&esf&^;e3Bq{@k%5!@WX)zJ_RK*D8Zv{SGh9K)ez&*YxspHbT7CNhO`jO3&s1X}S zHs9)d+UU}?Tt<}pBbBPC_DU)Ax=A5WQLcH-0KDVDB3fE+`gl8Bj;yMJ&h3(M^IJW3 z3zeHjY0U_;qf+DfKAfvSi|BZ3au(`_P&e!3xST4SeQ?ug3l!M!_ym)%bu)*|n5v9I zrXR92?gyBAZf`b*ZTtC+t!JT_(kr7s990|-bujK(#Wgbb%B(dM9X7fEm$7klaD$pt zfS6=NAL67zo-ugu&Oj7(r;4IR?AF0+h*1=uPmiI(ut^Lq)xJ7zuzXO%Tm!YoMa(!m z(Bhb;T_9gH$YmeyQMXG`FbGAi2!b(=K zEf|rKC%I#plb3-y0;R?ASfpCvq(23ZPi~)mdTWE9e+;fR{_45xV-2`Zutt!X=ZI1@ zLg#$UA&j!V11KAhQnXeG;MZAW%LKBsbhAj~JexwbOH`sUl9tr-~_04@L^rh$?9^gC-J~(k+(47 zm&6)*TX&_hkbI7*vcd{GIv7RE3g(<<_!-R4;60dONEqh8MtK1#?{!>uVijQ+r0nB; zPoo$vCacVlz!cAAeS8{68-nGAj{M9hIly^cuy$(dgN?kCz_npV+Cs8#eK`~}iMEP^ zp~wb7&3yW0|&*yO|KP>i=$K0wy@%1M?%>r`W`H1$Bt@A zS@xR37plSHD$!Y_=~fx?5xh7Y3_mQTuuV3Fah*vInvn_`p`!_JV0TnO!d-;XZYYR0 z0FZUfxE_BjE(7fiu3(F+2RQ-^`)(vlZCR`@sKMTd#X6p5{X|-o%ZmdbfG6eD1?@qXXsWv=9aI zCY3IZNfS_V8t&7t^}|IeSEzDPjWyuy>VD?g603^wq?JV!G9+j*zIVGFnmkZ5ZUvd6 z$e1C#uT=1LAFyk9azg#gFUSfkHG@kxX)gUF02lC7$tor)GqtZu{#K}Gx|ECpZQO9+GW$tnk3Kq@%;DxOM_wH5 z%{I@B-g0iXd!hJ?v&5&ff~(Q@nQ$jBgj-LQrIknRGQoCtL2*U5G>QJsUO?)`-g^u( z&uAHfO0)Txt_jxS*}{DtzQqaW*wt?;E5CI~@#Fk1+@{;{yC1fpD{LHDWl?X_m0QZf zuoQPy^Dz5i9|ddbpg!Qi>LZc;m0GngS}$2mezY#gsL;k)znRLu*Jme+YInWwV30$> zrG{Ms$P@+2=`y4N;Zj7g1N&aXZ9rni0@5HEv9p6BOLwZAc8hRLK%x`jz7u;-CP=S+ zZ9kb|Rau1&6%orgM7*#W*Cu#?6pQb|+_f4WleruBDP?^a2QI)`ay(Ruem>{0NJdNE z69;HLBO>es<~6M&RI}yR_OZ+Nfx!(oV6@4FCc;h}3J~hJ3#VrP9ZkyXd;n8 z=rd^nh8jDL=v(*PDTwjc+w|O>(d^&i>1>EB%fJ)4*cxy5)rZfS|2}Civ^t$J)w*pI zC?klR7|qu`yxiB$XMiZhurxeNX&Rr~4(J4@-5_^hH4)EjB|D-$P^`|(&(F$d`MLVr zu0{9f;T#{#^hrt--HvT|A14JSWcX#t^XK<~szxebP)Rehd$)Wc?&)4C;yh5S+4>pH z>`8eMnUx%Tkt?Z6r*)0ZPV&1m-v`B-%~zGvV^FNwqgv9UWte2)h*0)@}#1;Z4y%@kJN?(4)PJ~hP&T8(Kp)}8j ztmz%vF!)Z{dS@aDxw;2g(@ULg;uyA*_hS@U)BDp#oLpz=p?k)ixrNu_g1X^B*7TEP znQ{{MsK}b$*@H>Baue+P!~zv}F=cXdQi|GoG|`W#)*FhAn!7ciHu1~JK0Hm7olag( z-s+mHoKyqgB)2)a?BdJ5uFR-s?=MU?&&_^p&ffLa=zcq#ms8K)_0bKFV+R!CeMnhN zc9YjmXYcwVbQ1l!IrZ#aA301ODLcja+zE-{vRV*N%z3U@{xvTPY(%CzGTuW-Svv>ON_OmHP6g(6d{nyF$y*VUop>Hk^vP%MI=yyapUg1Y z*}E}iFn?F!P7&7O!97Wo)*&CGp1m8ST7iEuca<_e+iczis%WOs4&ucJ|)>yb(pjv-$$+ft!9%PpD__ z8d;WsCvveh?*aq|Nu!><*QAxn@3A9$C#|k0)X(0HwB3CCA}i_V&x4Zk?s_Rsj-TMP zQ_kKkC2Cb6^WBU^_RB;%pPGprrG5)Zrnaw=FJ9fWRy8kP9em~CBaq)-gkKyA`#Aj4 zT~a6!Kgb{v$1cn|@^2Qy(PYG`Dw3$`YXXg^XrCBV4-c`$bupiKr0HhoO6v{my47pp zN~gM6aQz7tWQidr9782KG z*rRJT|4bdeq4#UqpLLVD+DJ&vSCS@MaR8jT)Au^$uV?#Y@4#PxrV13iUG4shX+A;fYkQk76n#J8_Ft~rjEvQY;gw!oq>JTvb&rjyOkkBle^Ks=KesJv=QK!{?+c-hb>Mk^<}+xUIr z`MuO44Q4u9;(V7rxqEsY@<`@DPm`C z)Pn4?S&;2mA3V~;)$&M!*9Vhoe9LV?mLwPBScs(sl$u9qxm1LK* zOR~$UCE1$vmgjk=4$wSAYSXT(uz(kD6=g}PB|VogdD`Bm>bQU^<}0@zNGCo+#a0Xz zqZRY@L@JyNL(c}u`?dU7(N!!-cBvXxQ%+-YbEZ`vAA-QzbFFs~LoDe`R!lcq>R~6q zwNG-NE3NxX6;PVqi{GG|hsjiXCDjY^zEuc<`=zA+F-xvvGTdYFiBd4`Gr0)J6uE z{{ej0ANJTcLkOHHrdV*NX< zNl&HvRjv~aZT>0O;uUAv_u^#=L$FQnYKvW$!L=xCpE0BGt6Y`TqVw;{zepBx;Z77n zHAoWO%xv_cl&S{3JC40-EyFzLD%FM3eH~@`P`=Y+zewEUl|mRVpa(lqV*T9~tLt8r z=EP>f+|UD|S!4GQvA3zSo(unFSDC{JcztdjgYC8Gu~i59P&$sF_b2!cgfI4^@WPRG z$~?cIW^^PRB+xp)s4B3(!5@3XDv_vG*C}>KSwSyK#}RaC$B90Vz7rFmSu}Q(=tKG4 z_zAG3amu0+J%m_twA0#}jMk0vc5-6Zs7%p*l&{jmOLO|RBZV4LncqWCO6(}lVy9t? zJ}30&@lG&n*g#(j_Hb)JuT8g`r7clwQ{ITbx|zs8)s;e3e&exoY8GOOS*lgcQBlU-YOYeN&h0a&s1U zhsBr?vKPgBj#HQ{PyAhO=^29adff&zUq7PU#y#xr!D^P<5KA-s2FGh6Zr-gp#W03% zFb<}IBqC>05nEvOe|`~xTyG2F5lY!#;S z7)%3REyp$XfgXE2b+?gEV(;_A@bAgcb$w~gquBH)uPV%m_^ZcY@!r?sm5uN#!Vl!i zDDZ%MvLn%%t5z2!lKr5^9?f39{X{wTsCloy!`t-G zjdQHiGLp)%e%V)DnL(Ee`&cK|8*kWm<6Xj<;hdzyT&k+ofy?y^_Z?!ImfEtq43gSGIbz9?4fJDjHpiWc)E*w z;o7vUgcp}H1E1Q#&cxe+wO8#hiZFpL66PlfQNb#%aGb+8p!>+i)*Qg+;BqirJ06w4 z>i{?L=x?cd*P_<~LRW+O{E5sbOVX*|alo#d7bnkqdxCl2-t)4}DEz{a+tz6i@^ri< z9-1GK(EZ8Squxd6vhx$?eS6P4ZkV{}ordz(fP#1wQvS9%!zf_|9DwNNUh==am&sQZ zso-z#WsYh|%aa#m4&LDT7trJ=yIX$?9p$et=!fe425%P6WD7{r^Y&iufc(C~rRB2s z?L99EMmgB2!=m8%OQr{lD*;kF?#>qGeS6PKt6d7lQR2Lxd*0sj{(R6OmrxA6@Jo`1 z1=!!(>9>4`-F`By`fnmvLqS{Zv(@4X@Qeoyc(D_q55I#@-z*+a_QrGQrP$t;C8?K7 zQ0Qyc4mF_Az19B3CHk9Wo+aq!77F%y>`T6Ssb^$ELj}P|JbSwxw3%4>Jap&)NC${KC#m;B|Io zzW=dC4+X{!VMJR&5)h}~b?MHv6z809iFqGu^iXH-2{;yl=VK2(nCGWm1x3#3ThVvs z7QQ`%gvMVq!5fx*Bw4;ciF*Y}4HESuA2=^UlywoMLqS87MLZx5$i(`i4uKt9w5Km| zy=k#LCIUg)_$Tuv2>-K<$U>YeDdhdUxxACEK&>KVKIgE+c0ITp2!T&-zZk^hJ#buL z>Q5Vw>))V$+lLK{PdZqDnLW5Q$QO+qJa?gx4+LE;E2B!3!ZXN?M(1TzVC(^&o?Qgm z;<}sfIK}|O7>7?o2vVsv4!MGV0WHYn34mjv=(DjkSvF>$N?5yUoy1{ZkhM#BlUEPy zNs;#40}+HQ1x5zo7bR~u3LLIIb}uc60~<+^9za7aoPa!>%(Z5zbeV%-Qn>a$AF>s~ zw)M%@bAyBYzjUJ&>rYuwf*8dP)QXXAsV-LgXA;CHrmgEUWHB~Nt;q2M=W>eeCa{f* zWVvOVtNj4cL)8AyCa_IRo1y)%RJyteKBuT#Dg`Tv6iBa^zV}PvNoYr6_yE z!+{n#oYh^b>+>iZ4Q};!eIl#SJ!>f6?=!fWfM2UaMm4ez=j=m2)=6;`1xXd)pO2w6 z;$(72M>(ZrHrNO7m5M&()xR6wMaXp=J9J60nz+TAKmiU_@i}*QDSMI749`>!i zr;RR|I`FZzf7&0ZR7JJ7b<>?#8OQ@_@o{XA2a9NFz3Joaa5=K73Oct-#!V=o-$Lc4 zQCdR^W}O5dJ(vGSYAc&)`% zUds@PTx!j*gJJ|#7XViXPQj{a)Cd5eUUGP(804p&+nu0^U9PI?gkC~O6~9Hp|ALp_3LIdVj) z8liJOhUfIEtnYNB8;=kb=K#)+(X&SMKGj--yLf4F%1lExr&Qg!q=~c}>&$al<-%hW|w$T&p!$06_@HXtch! zaD&3(h%V=-Su0K8_rmaJQmwV9vR7C!4P>BjXl#mAttcMhv0smctj*9G3J{KTYSF06 zt04BzOR?y)SuT&2o3l%*NNBb%(0vkg0uEeX5emBC1gLiKVO`M4>T(w;@xW=3w=m+D z#2R^9ccmWP+L&)lHaDQ8Fd~-~j;XT33OhO&Mav52oC%3IJQp(z3Bw%NC@*x#|9RDG z@3{%wogQyZ&O+DV;J~mW$#!{yeZ22!6vM@2l^GJ4QgzaXA)5N<7a_W_k)PKEYp13@ z*vLBxTx&gYw-pCD#h{o;lnxGtB1?dw8pqIEAB3Fi9%ab4v*p2U35JXy>cBy9aMNpr zcNH!?IiY^$7bJ?BdUOJxEjdFlb5BH6QUk7hg+?AMoSCqYnkhBAYU@JDbKB$fEa;a zDI~%1OfuJ?x>#3tQ!}`9ljhQo_qZ-9in*ssRxweTseM)Qw?aMBrDPOn34y`iL zpW>{5Lp*;r;Brbbf9!s#B!TY3Gd>7`f z)$o|iUAYFgV0yh0t_{RP@yHk6N&;{%Q&O}j7Kb; z4UuITcp?`|XnIf~RpjyaNqeEy>5QpXtd=%DistJcUh19C0Vn0{Zsy9=0(n@d9RN{^ zVQF}l(lkD|9nc9*yFu>AQKAAJ(H(Xs$Bh!=KVSOK;3hdKB5T=Vw;v^PwawZKXkV0e1}D83z$Yrk zaK`TK5rj~O2&3qo)!gAiS&I)@(>t_b@J5gmng(m{?M$TOE1dpgAF`&GI@`oCY$xx> zD6*#ar;SZEo~3u@7H;h!q4|(C{Uli#dJ^}7ZM&_NcjW{7KCwW>Jye<8JjWrn1P6Rl zN{a5o(?o^oe;*gQP6x7 z&fdvmj>u>42}2q=dl5T`%8q>Yu9IPe@%xoy7f~h$K%ac}uG4D=_N0RbeE{0oyZ%K` zq06uakLjZt6FLqgpS>HTT7iEuciV9H64?y(?7gNkKa$+Vx-&feRfH;g`>)feCz?gN5)LOpxe$g&JPk&CT)7a%xD z8ujeGCap|~o}kHy)0y%ZZ!Y>YmeH?!2PB)Z@ALMrqu?w?~{F}vaG#Rm~ic})0 zuL(4wqJ3ge-E#X7TU@UjEc>@D3~y%FjO(i@WATrBgB#d&tJe zn0d~ZQaacTAm@ioqPfnj3+g$CRXHofYR+Xqqvq^B&fZHrI%$>&uU*vgY_5jjEIa~i zy9mG16&tQUka(?@)2cRh+sI)aul`BJCRIq)L^a(lhci|CS9|PsO1V%;891&hH_mY2 zwV|_gPx2=?(1yb}Z9TerrwOadB0cCgbY;eEfqfUpmT^dI_T{VBCiBzJ)qix$gW=<2&brsEGN1HL)) zpMXb-E=YHYmR;}BHOOBv{Vg0<3!AXMM1mo+xIaCa!wM@5cgg)(PKsgxoty0KD9(p7 z{)6XtCR@`7HqP#irWo zN~Eo$9f1?W%c$#|C}Zac>tnCpIk-Mjl}*6+g_O;L1ogu9eVV9j>i#K7fyDAEQfkq% zkE(DwyLMbje7N*D#eWsoBCliO$JfZIB_y!itP+~{zg0|clTFKns$qR!E-NlhCp zyHbeC)Ralids4r2+%bwj(uAKE;dlOvy_48OBNt=BQjRpNWpu+dk~jZ`*>x2y=rXnn zswag$aty!3Q8y1Sg5b-f3{ws)B#s~FY4Y#Uw{`9q)d3#+vu-k18wsiTO44L24uCUv z`d)|p^|W-ecl2ouV!k!XJHExVq7%u$_XLScf5`q0$BZIsb$!^^Nm8)6bow47P4?+( z8(=Tc?!HLTi+HC{I7UjjwTWw3kR%TwmRpgkgmNOjeffgOmTfp*l2=)vbEO;@^cRRl z7fDewc?$|nD*jqu!|R8s@>1OLv6FW{VK1C`*{FjL4-V;nt1vt>?>MHD&CrhwLjw@c zBnCnZSqIfA0U?SN<7Gpm7_FShWU=2T-WQZwq`^#QOMG|@yFzvhi7EzpxjaUAU~j$- zxoVI9`Q+|t{%6Kpqm%rJ2<&UIqzW@l%LE*?+?(irJ-HyeyuSrms1GL9_?Fv(EJ-fL zu@Fu9Wl6T7yjA{7R+3%*F)PU~XP0D`Q%kZn>0L{*0Ng=p)2^$qfERBSWl5?fJ(n6JJv2P$T>B*Vxzf7NQ~{+q&xYPsmrdsx=;mtP zmH`uf+NeTrPA<%L;XUxld=3z^8{1nFi?i&z(Gqp;L4|037rS$ChY!&_vL3o5VIzzI zvo)}bbb1Nh@8!w#7pLHCQYgJbLScW~XHOK?^}0+JyYZAG-RLcZ^g(%F zd?Eb~Y>3un2KLwS?j?j2e^YFHY6O+45ZO+5EZ()x04d?l3{A5js$z%O2-~XC3La`x zFQrU)_iR+oyxJDS|3g@0-*wO6aw)>2Bq|rk{;uEw2Uaz3`pi&<**FYXXKm z_!{|Dhgdn=`N(h%h_#nkm~-XDMWHhTj$ljlm*Y^C4wZJ-d@`G=0vZ#v{~~sN7gCx{DPX%k#OFR64wj4 zA+o>0AA7_qk*HSJDRxI$K`%D!JJYDi^%4?QWdqdZr3Z4~Cw z6!z!wPB3d&y%WXiOTiv)CEIeeFW*{+=v|VYByNBF)y+f(s;(5O@*9txQ?n3L%&o+t z9)*35o(VLQXq_ps9wLj|y40QCl;~2uJt^>bF_pOu30hZ5x;RsH%vh%DQA#Q#;({kqB5s2>PyCE94bp-_=?4db}G*<&LY5B-IXOytL&V`vZ|0K1?x@4 zxS3IN`ima5q;KkyU2e|euCm!-F&2b0nd1~D%M*W>TY84jbpE$d;0DJd%5B`k?jEdW zxec*2!*6iBCgSGZic<_@_=a+ei`%!c{-EDO91;6=pAjQBi(b15oyWCw_kgQY&^?rs zMuvX?g(8MqHiNCgbRL7l3ZX*iVk}}G=&{FBcN_U6HgEI+zc@X2nm0y`H@2o5r)PW8 zRdl{h{GJS5*O%5jicOF5s=}P$XHW50;iGROGq@uBK(34e56CAw5}moxRja*@{h-Gl z&0fC!MET&_llDJFuhh43d{u;xinMr?&OR7_*wu+$C7XjUt;RxV{^w`HbvEF#Yz8min)(qz)9p+M1tqxqSU%2lO)3nsaiz?tPCIYak>IEl6-#*GrNdzGq*LoRIo#OFy7x%)oX*s?_ zx7319?O81pvtJDKr%D zjvFS;ME7@t6!wh|-(5f?G~&E(?`0Afz0>f9qO25e?-`ZnfJusu>m|JTqKq=B4iEp^ zdzpMykqZ9yUgoHlv^;r1=HNs;|E1QQb{7KgD1UuH4XgJXyjeVxEg(tH+k3eK^7{&L zwO5-~0WN#r-t&@Rl!Fru7U^8>PJ`RGjNVaL^1r?3rPVG4<0x_7&pmJNd4E2*+mzTy z@G5UyKWP4PysYQx_^6XZ!SqP{DG$;gEZ`l4ps6SYo4|7 z*N1WVCC>*gF@u8-ZES9OireH-Y(SF8Lp8^iXH-2{;yl=VK2(nCGWm1r44CcGd!Xdk6{5 z|5&4sB+K_Faj!Lj&ZGv3`jHQu7a_{Jh|;wz;sJ3$Ce|NyP!VBZD;M4C!pd4ukBI}TVVi1q_!13eyj_b!) z_-!9HEI#RA0cQ5#)*xRra`4=RLOu|5g`%lOWu$hc@CINgVbCS-X@sdG)}al-18Y5JAXNU}V6o zJ};v{w-Pp7d+c7ajigAA7m`mv9!}<3vsAjwK`<#?d!G;43Sry&Wb3)XLH=L5(Teq_ z7;S@OQm2Xkpk)UlAJ{RaI39dvOns13Xe@MMdlHY=j4Vzk;4(i?owT! zN7-m_tH0|LS%vOdL-~H6!OaBxS{*W~k$nV@brKW(Ql$1c_~&D2jX0Sc(os&ag?pnN zcxmwIG6Y!%0`9}icyT?Qaq{3sotWs_7(}I7;teZ{_$f{oR&fO09Oc3!K!KF{s&3qc}BY8$t?{Y>9y;C_eeofQoLgT zW`Os^8cSFM^Sp05ZGn`?eDRcyjM0Uua3xU7f>Y0{rjz+(e1^Bh%?FEvgO5M*#`{Cj zR!u%T8xdJ^g%AC9m5)vpa4%Z^p3MV~?e-yR?U5-*q7s$C=W1667ZQcjZa`3+l`b!6R zW}YKT)d-#QF+8VNWqk)oH$Ed#XCHu#oi(;hAUjJpi?qZZ=RfdaZv?fl&f+q;lSLL- zyj0m>g`x4FzE^QAs(@MHDlWD7{Sn{WmF|Q!4iU*x#qz4~Hl@4-1#`%E1Zu4kN`K{E zzWR+=4AoH3OxofL;X#ORsh-y~tsgh6!{Y(i7kzN8)?hl8jD;YCBSu-@n{aIja*hX3 z%89|fF#MTRYb~nm6;@0G87LeYn_^8d-d~S}tj*9G3J{KTYSF06i`O9Ki&EWifY?j1 z=(AZakCmIVOR7j{I+2R*15F992nAhm0#rNrurBCib-9a_c;K|iTNv?6VvW45yHeh= zLMZhMRuYb>O2P^|Iv7PO3Fd-~h>JfLGYkpC{MINhIPf^e$icJ_t0r`$$}V=WkM})5 zWw@BEGD8AWs!rOFQ5OhSD7vwcpVtLzr=~vG$U6yKYdvze6$kZ<7c+^{!NE{u3DAqW z1rgdk${cTJ%Y)kz%+UqS2~z_H#lcOl6^@IeW7|jRVqixuuut?oP|c3H){?U9HH9x! zgT+;%v%J#_fgywE#o=K1VJU@evMG!UYG-TET~z@(ns6)-M-?R8#Xh6md^eJ%wk*~c z)L?JKVjWMjej=@RyaxZI z^X;97cPG;gc%v>}d{tJ&-rr-7&>}Obba6}?txOwg+u;qxmx*^$;Tg=U)J>sYkNsLd zT$FN!Di_sQ1KzIgXPzywsu)k=ng&~D93l$J*EpP*L(@J$&3xU5`x>5{P(Skv5=Bj2 zmQOFX-&y3w*{uxwLq6cZ096Kr3=Rex_!@G+g!MtGIe7+E;7C@M8YCupi&u^$?sFJq zNQJAU`WHaJo339z9y0efv;%+wByrE6LkmS8pe zVV_~}G#%6jJXqZcyuVUIA3^P~Y$9|^yET;I|M&XrL{aUo_nq=GMS*fU@SB%Dq%QZ0 z@G=xI)^^09y|I8aNQQBCP{QVjFf5CP8;lc&oK0}wi9I(Hq}RT-pUkkTtU}W$Kesm< zLk0W!jjdl9-|w?0C|M7X$c2#Td@my$ytLGnj}i z=$%!M?|Q-H&xfq(9Xe+4otGZZz!Rv0i~nSG#4>dZ+sVrs6j{^z(?$eTXX&9tyfa_X zN?TYrpg=`R=D0Nq?pY2C;R4`f(4_g80W*& zMD^3j%gI|^la;5djaq2_EoJK2yXa&4u{nF!SEGBja9&P5d)G%dJdPbu7>}eRCgZth zH~Qjq(tG5ycYTB~d1iL*!b5gCW7n5q5aPSkOFnzoM-CGh&CpL$@>cqMb&eizQ~Pnk zt}l->8b9|ZpS|nDg}JNpbkf;7eOyaww6k~m@OG|V#8RU&XVkNIG9d|P@8mH@AX8dp zBwvu3gl#cb#54cPAY*=mXHs-t{kn z4kehs7hygg@5i}IJ$pAuwF3WS?zYii5!&NaQqSIND)Te>+_GIeuNhAS4<4y!@9p4* z8z`_=$KuOg2_~Ptw-nljPs(K4&uM4x?av!VAJ6IwupGDZpq^0A-Zio;15e~)Yu*J2 z4w6Pad#_0=liy=U_D)(|Pq?ocKK1OqUW$|BCphhtvv*60veSE|L+qD{WvQoTB1fs; zLXwBhew?s_!&k{WZa1w}%^kOcuRMGN4xBH-FAkG^9DeBzd4V^Sh%K&*iN$M8H#=x)H?ZqguLZO;(c002KOBNz$Jwvo7hN6l zRM}~?1254^X5$P69vQ(c&(Pxsy24p$YCC@uS&%xRY=uDHQmDymm3ZX9d6}f z*Z}+09=n}V5>ZkHj_;>({|OGX;V@2HkFMTn!m6@JZ&?{)zF)=R+B^mb;*i+v%U7>W z=BJ;l|MJW-it6cns#7j;n=DSYIMZ5JZLHE@2~&F0>j#<|cbP ziu2)&|KRzZ$=39Njk9~B>4w^3JonOdriCi$J@z}q^PqP60c-zU9X4DDFDa3>j&=l2 zOhom1^^PG_4=yaNbMj%wP`supn}F{NDH|1C@Y6(PQ}<6v3M7_Sky4A6eToTm9#h#u#6VhG_=r}r1;8ylPGM3(nebO z39uBBR*03W)kFwrVk>Sb94>ur1zAN8FcX#aXaLgzo6Y9ghPLhJnrPKF1xHH)d z*bB6FDXE?!-jPn$_3q#;iBI-HI7UjjwTWw3kR%TwmRpgkgmNOjefe4-lwNu8lDx_S zoh#+Qpua#Yx=4ze$y-opQt{XN8eTt4m6zg{kDVNO1bgAc%SIi1*aC~j@XWm9m`*lB zKh#A6x}gDxXA%RIS4{~BQLGp*8ydxEUv$dcq@91GEuUzTJW z%3I|(Dy`&7ZAM9U`NymzyPRE;T}~~@)}*&Q&nw9Sa0jVPyRO0lUc6P5C8?J5T*BmO zd!MS~0;-s=+poKjl;%7edKJU0quM|(Q}ebAnDEm^6?$`WVYWNg zLzM=GutBM z_Y%S%qbasX%Q}J}S8Cg^J8n5r!krmj)p|W6c8HCzt$HG3+Ta~`ytOrIY{B08G8WIP zZ87{mghlpU_Y5w_@8FlVGxmo)ww5B={BHS-=6{3Bd}ebi4mJzJRfkwP+Zo<)4n@{$ zTX5lbc-WWYP?ZjqcGr9|o2mjD6SU(Ztj8TpP-mbmGFI|;xoLIJCv%f{H@*Z<`2F}K zx776NE!Mx&n)FnvU*$T{kQ?-(({Y7r>nNv0_ntg`VBd?EDGb3jy{ql(-5Feq!uA<6 z3ct!#sX!0?i)0}e?nEI}gCx<-%tkLtscO)>_mz6cU!Ekdr=z8EcYPzUv`zb>OzUZ_FDAVssnu}9Y@gn6FlWAcCjCY z7mln`=J^FRqa)!)KT04GXey(BgFp6&RU%QXu2bxevVvZejw9&OjuU-83VCFSy;vX0 z@5WDn-KpzNltd3v_o7gDTqRpdH_F?|iCr_S(~sgYNaSG;r^SbTl^$N2)3+Td)R4;j z9(qz@M|rO7+Emj+GyC&+Czv&Cpf3e`xRq?n(Y}0Z9in$hMu+imlfSx|$UxPVLREg_ zv2$t`Vv0FVh>|J$96b|gCeb=mVm(9_w{@xOT_sMRO4R@9{VDwMVk&bR611+AbaAHY znAw>UC5=J1TC8Wj;j3!K&mkX8OX48pDY^w|^h30GQ+q#fHlMin>*ax#)}1la~8|0LYfq; zHx=W1M$PFjdeoA>sY`abIg7h$KNuEcL9oCE4Kc~`#NXwXp23T>3jNiOD0{z$-91>% zvUf3iLfIMmHHe#cD^4+o;TuZYFW3CN=O(+e>3C~$b~fk_4wT~z`*xoZBR7j)y9%Ah zwRHD@t5diL3mo*mlvGa1Q0Kt`z=+|N&0woAoyXuXV^j!Tj0@}oJ@$C&ZX=(>4nGX< z2cPi1fUO(;0((z}uIo!{9>u0dc~xOf#9uuI$3u7o8}6h$kSn9W1M+TSHOwXO3Z)|UkH@2o5r)PU| zJMMgA|0Qx^vvqg&@)|oUC*vrRWCq80j2<_&s8;h1+`x@N` z`Y*vlH_ow6%SbB6`ek2rWkzdDuW%qku4Y%QUdizqjsR{*kA18Y>y0<;yYVinC5k8p z=Oi8GQdO-ET&`cZ?-0|Zlp_APlZdo>QG!eF#$P3>IGOCPI#ah?-_h!{^H!IRU?{wF zm`ok<5qs!bFLR1lr+7Tw#l3KCT22u!_|y(|23111(X(*hwDY2mc9AeY)r&?YiYpv5 z;SK0MvavM>@Hw~~4A+iFS%=`9UCSO&gg1^0&IjSWsPhOBYxQ*vuK&7Md^uN8A zM?}5f;LYNBBZ@lCkbhOAz&>-*Idww#l`fnmvFR_I{(0~^^5&G~u z2=&e4@nmm2hhB>9U0K!8&W=MNS#S7qi4c9w+Mxy^y0_ZDxI}-k&$CwcC11VN^MMgE z2%!0qzyX#dc>+B0^J$)GO-uK6{XhhS!!LP8b%_}qd}w2H(^K3gk7CQa!b_?2@oK1L zStW*>8Vrh?v-9Km1%S)rker?Q8#o(X@;}z-p&+(HbY88qKiKa*SX)DoI`3nR9_s8p z0mnk{eC)vo^Zc}{puuK=&V0qUfFSqaMH9SX$w!jq`;)lWdj8I&28sGnLQrH~MCn=< z@qjoW+)R1YVadFMG7)!BAfE7EOY!1^APaG>q>%UX=JHOu0=0^81G{QEnNP-Nc(GwV zSR5RD{E;`_9|~JUllBtZ_260kk7_Tbha zUo>*?+=W6u5OlSy45Vb~JcBHEblx8oTzB&wL=0KLFvj825Q0={jYF>BUqA~oc>>^A z3co&;uy)ltiNn4iYnSpSuO8SFHw;A|n|mOFkfp%LfLX2N?FrD3R`lW8VsrK^|yo~ ztI*TqP$i+yZzkZ^>X1>5>?7w`rxZ#4fC|7rA1fp(3QX1h6ZCt152#wQe5?Zj7m})^ zj_V2-1fp8vJt~X%Sxr~&s*@x3t-hy?E}4tqV{8BPbQv|Xq*4{tUgg5NIg>yhP>YXa zdpuZ#AN8h>x5MSgsw(K*_L+%FSPTm2w@|qmE3I|1A*ZZnAG<;Nb=}rvOmN zh(5$ggRCZaFV;X5b*GA=M(ozXYlu-4pU>RF0_DL)e&_bpaf9W98s=)4O=tq_UQ7qV z9J98o<%{gYeTtF^u~-I5WJHxM_3$;1kTG^(8E%T7^ObX4<+Tiv$fec{JFGpVSquPI z2u{JOY2*F}sb}zvbj6cf8a&c#*Xmd+5O<_tD{(LbyeHOJ!Wx+8ebZ?Rq(tV6r*veD zE=+|hL1;s1s8S5TYuW5y*0wUX1aaBM|vB3UC^;AQVkHGLQivLPF@BH z36vIxYmsV&lODuB3R{6LN9lOVp&mhIo+C=t2%YmWJf~M>eWxSccxu-i8#`-knLu`y zZWd|0$65K8sDovH;KSYsYGIwlWpXErEU@?l$Lf8k)S?QQ6|Uk^i{Bsdyt!n(G0OVhglo%@bKJh+UKsvNs@m-SB>NYMWOccVlz8B@$XgikOJa?@t-Dfsi!HgVa7>jIR@l+OC|Xu9=S(Oqz;iLf zkTA@Fjq(Dd!Zw&j$7RP02MLsYyzglg!^LEk84{RMb<&1mN1maTgX|gxKd%ecPECEV zk#`cfru;!y1_a9wP|PGs2M0rmmH;6_yGPmT?QD5)TY{~oBI>|Fad6XXg>&g~bUJc@ zeWLGyYIe-EmXu|$DSV+CEUprrv)>=6KO4lq47~4#*%kE0Ti2H8iF&e-4>#l6W{dR z3HL9Rf>;xk7mnB9pLD*x^YHFux-r|Dr=%h8@3BW{k(pGwOMpr9YyEIh$`z_yRAUWz zySkrww#2GpJc%QPKsXwQrSU%N?a;IjP%~fm;l73^C)CgUf<#eMS4QC(M;GGkmiq_$ zkPkR8FoyvlgM+~azJ?qyVSP|)PM$#(IB@K_z}jk%nB* zw^G*swaoQ5kS`hAl;_q=K#ahz6q4Y0CYftcU979SsTo|lNptDPdt4V4#SA=EvWkhy zOzo?ZzZL44E+t-|m3}C`S@X;6Cy_n+_R(YxH*YyA+-PsMd1myMbF3Ic4K1fS^2PwE3y`U2Aum?{5Bn%s zQwQ|{4_0>q@2}LVebM^Xi56jXAcScip8Z~*ohYi^^}d5a4t0|$3Y61jNCU#<5^)=l zn6ZE~NQQBCP-LCXae>`F+_(Uv4>GDsNHZkI{l}i03DRp{+fQa#RaT*cH~EXm121ew z8P4E+QY^j;bJuEkOy=$-_fP;_fVJd!s1*HtdZcI^9U2-PIsgei0z$;_39OL8wSDaJ zec%(r4IoE|-5146p{KXE=97)B*{;C&clwXbAR@#-1{b6floRV%PGk`Jidul7Mi45r zBxvac_>0?lkc)uvh^4b3vMd8nmFX}ozGxy zJ^+q*vdguq{aPS;aEMY2OT)7irvo~{X*b9nIZD*30^aW&C*23d>d^W8tbCTAtH149 zbZ-~V@jXzi*>vg|zOiTdn&3QTfbKkC9w^pq(n>($ zNqJGLf#KOi&C4V-8Z_cJdb{fhUq+kNd!Sge`Koey42m^-R7+a443jJzQN?O~XaN>S zh3kW2&E-&?%RfP>4^Y_xlB}LAIv~&Eqns%z2Okt`fXhbkzF*lk6^AykPA3KML9qtL zu=96P0NSb#3dI_lPEYo1hg?$ZLp*$ss~*IE*3)17COIi0YuRGAA0=|NC?ldIGdStR z06tMMhBNli9bzK3pm$bt>kK9<@CXP7VB{UzFnAE<2J(mV4NH}arvc#t*yBv~1H68Bm|?o4WssP7XCRNOdV#|$yPfEIz?7`DS zh3VwwWT)I@<)j(_=dkdR;+C>HWY5mYsb}vmOg7KW($C(tbTybW5BKSiUSBq&p1tcI zg>x=FGxsFjxn~0UBFwGHll3#|*}Fb+m^@N;e(4CLp1sovv~s@8vv>M@b&lRx?Lu~6 zhCZQ@YMY7@@Ph8KKBJz!>ywAoN9F0Hvv>Nqmegox@ATpAXH$wOcvKdQdiG8xB;o9x zJf?U&h19^5FeJ~}FAbAbguq`xCe_V#GK?@@+u0}$G)UB6gk&f3$Lk5?vv-|dJFrh? z813xcm{Je$o%|gS$D^hdJjp8|1@hUuL8=w_Cv!KY&Q*j;>e+iuWqu}~TO4iYHTS__ zYC}DHZwEKr0Akk#mhV{gr8s+UDYOlrl*zQ8)6U-8pEsh2I;#(fENIRJn2&lHMm>Ai z$g&JPk&9}t%C6+LO9@^%WYn|wdb~3EJ$7X8q}BC=KNe^2^-`Q1Kf!6IoV~Y_6P4ap z6|(dEUnce#JT((JO8pj+?4W;@e8}piwW@i@>fkF69{~vLi|~s>VIPNIy3-9M`v-X< zj$N2_RiqM0eNCVd73~v)>fz1KJKZ<1>sGIYC%Q%Ps0V*I1iy~AVmZVX z*To9?vrg<+@Qbbvd8!X&v;!~EN@i<;Wsr+QTq+=l1AX|_gJ1lL13$9IKzRJC8C-i65z}SG-@t|fH<`Y9d|o_a)%@iG z{H3Rq(gXPB%zpwNDY_utC0cg9N7o>K!_RNwxLVkR^(B%7K~xy-5;ihi>Q#P^%uV)o z6z9Vk|H1P+ldb6k8)x@M(+#!7bcS;H9pZUVJNRDpZ84`l8O!O10Al%dQlnGBstA z^Pbc%9e0f4k2K-uMfjcnV(%pO(8$G@u#_VWYuR&rh`jO!^BPHAuk$Ke(B+g>SRR)5 zkz@EJj=Fhx5d>c*WtehcA#u~0k#}(ZnQDT|pa%&UUupmhuLM*o;RSD%peEU*b;lnxh;3avL1&-3nfkA(PSagvT zHIuiX(4^w8^)wwV?^fYElD*q)b$u9qxm1LK* zOR~$UCE1$vt|eJ$9wN1A*Hu`+i?@ohB-N6hOPD-u?^AVLKoz4Nq3}Eet;-%znwe4c zU)fg`qqb$Y{8-UdEJ=2$8dg&tWPfvdRi7aO``rUD_AX+GC7sEN=|(->)Q2IV=Z_W> zixYhYxb{izbES2kr0q!)P@3~>=v68@G@WOlC9LlM=ly?A8&&Ad$%WbOSPxYi7{bhV zqbuiN}+v`}q};r}5lvhTWQaQPp=cl}|Ht)*yGez$x^Yrnx|KC>Px z_QK7=05cdXXFJ3DokOh88;o^#HXUzG&dvk|T@hmA-SzCtai~g%O1o=5nN3vzjR`i# z7vUai+hLu7w#Zn?-{q#&J)g|YoGWF{DuQ$=x5?mse3Dygdi56T-)T*HD%G!YooHxp zNVyi@(*3=7nZjUc)4STf-krgxMlVXK zYS6pm*sInu%yTZi3k7Z4G%8d6M8Eqw+CCI)9ag`I_YC$sJ@$*lJzgn<@dA3V6D8K) zZLzxUMG0hAjfF#O-e${K`fJ1j>@T~@Ty>$uV0$fkY}J82l#U}ne}b=oqf*X`{V2R} zWSuh4FQ^$E3FrMNlEC;D!2Sk*>=COrqFP<2*d1jBy(k?=(EE;}R>96k5x6X|7wbd$ z-S`QxJ9XWOlIS7oUKHw%t7J>*MtM6qv1^8P`caIMfzM2FXU)D!4=>H>+l~}!NM(Kx zJt?uHJXdycC-mp>PB3fOKwk>>a4XrCqkZ|-I?UdsxY@z0n~4lmT`5%MHy%5uW+A4S zQ@rQr=$Sw>iPo7C>mjnZtxMhMM~O0D??cf$TlnL}ROU7$Xk97k;!M>svoj@1CWCIZ zSkHXJSJjN4Lq3|8#6ieYbPLkxhiLI;`|jYW*?i*Oua^y!Aw0#0)hXmLcGuv}gi}a! zb|an)=F<$)$=DUUMpWh$Mt!N+y5nUcI%&KtW7V=aB+Fe{;R7?~Z zHK)JmQA_%!F4^VgEbg{eU5o{xN-2LpBw3#LyWCOF0Lk$q$}`-cj!%5@zmW$K8ej6eZVhH&zbcd(u^CsapuJg4gpd z6aBK#4tZ4rBhp*#_1JqdbX{Ls^C&hw%Bu=+XIV^O%ml zfd*}ly+kf-w(ibeUSntFWE_2CBZFgceGQqQk~r_cy*E{v_u*}N=*BtLX&Fi7SikJ6 zuFO!JjS7f;tP|^vH|)FdE@91ZPSRm6Rn_Xi<@$yD4lzwjZQR?z3kV&Ss@01UHtlZw zO`-0uI#ah?-_h!{^H!IRU?{wFm`ok<5qs!bFC(f`Jf802Ubr?br-&DPY6s0Jvo&gpZQ z&L>OKNx0)ch8!i4=Wl!PzP;xi*C_nLk=xd35HfVriHGJ#Cb~cU%uU>FC`>f-zP*=8 zT=Y&ud28TG+O}3`S#kU#!V*7#aV{bhFdXK6doPo(DpJAU-pd@-l9neg$Q-=E^A7|y zq6>Vf-9ox6xBu%ZXI8@^mtdaqeKq$<7ri%UcDGi9FT>9)DR zzT~TydcHO^Q~=G7?%&?pn@f_g{vwE9OD@_U-a!O}!!MDE8613QV{_9}+$N7=%iB{* zrH?0A+^YRy#^3;9c78m+umhJm0QR{ve*?!!OU(ONqlW@xhX{>~oauKbC3=WxZcohn zSfhtJdr!cz5Ii4y@WDJk?OI!R_9Qm~EQO8}nfI|qA4!()PvTx{0{g4*Kk|X|B1Bmi zQM#5zJRlCp#QLKSfgnIMtqU2pP!SV>APsRr`%kCDKglL%-VTAoPaq3%uB4Fn^XBpn z-tmUa)LuoA3x(^bS4}7L$@mP<$mfH_!NJELdE@<|z_(@y+x6gbAk#j%{bCT0_rUSv z`mXBVpnltj4U11YSb&*5xHZTZjT}68p^y&*U9C+;l_-U0kmZif`=f&EZocCfjc0&i zjKilP1gX>-hg`wGfEHx(1i-PlYAKGgPbI8fwN9qm9ewq{9<{bO8Ej5>0WdOPRx5ea zD`CU6$L=-sfPsyqNRJniPrzLVGS`}=(zynLN#WZ2e8^Ua_N`C0o*NwG|D_wPSbxfb z62vG9^EqM!h-FKRpGgp-n6|FZkj2<6wIat4oXh*fHZGFoJhmOM*!I~3wuxyov>%pA zS2w}uxU*24WvLXbBvO#P%Lz8A^~FhHf7D~IM-M=nUW&{k9<7`mYBcd;Tk*GaIP}9SH|Q}A8*;ozT-6%v&JQ{|R1#F5SZ z6ZCt152#wQe5?Zj_hIIeI&MlU5J)Zw$Qb;rrYm>V$r1Zj-_u5y%tde>wSRiLjG9?e zsfucE>*PAIGAcff?eSm{e$<;j-VT=|tE!-LyJXz_R*&66<)%?ub1Gf-X-s2n+RhU^ zjcI-yIgRO}PrSf}hjo~Qt(!Sy##Ch#vXId6B3sn3kV+c^&~ZE;!nkJ@*T~!pMMaHP zsESJ`U4XmVIC!{0ji~eJ?^uhI23buAUAenc1xF)x>)Zef9{^&*2ix37*H zEFaV`SH!M5EY~sqE)N`kxKB|sAr{L(iHxYSrQTX(2WH{+7!opOif}U-uN1k;YZ)Su zOT!K%28ofz%K&hN;1sNyHtv6rdInFmRy?_-!6UtPE!P(EK`Gcu9Lxaki8Yq62IhI+ zblL(bk@?~&9T}qwQ{hU`s&8hcdXb%th^)Cnu0Psk?D+I*+*iNym1FQON|<}ibo+vj z^fvaopkq~}8X!R0(_EQ^Lu!41WlmlO2?>-Ihij2)g_9n{KcdS~YUzTx3kCv~2o@60 z%yUGk8liJOhUfIEtnVneMu3f-HMUG3J4-i<@g{8$m6sv$#y|WRV3H zpF)=Jp;C(~U{+hjbxS}_k=t9A`iSrCN_WBEMvbBi=i3{nn_!HAv_52E!FdyruF08IwUuI>|$T^!L?e0=~yz>%T_pI zl=Zy{*OnmX1p9`2VfZtt)>>5AE3B9XGEg`)HpSX#>lv8CpXD!jVob8g+T; z=sn$VfY?j1=(AZakCmIVOR7j{Iw8<~qHI+X;0La+2nAhm0#rNrurBCib-9Z~J8)WT z*AFd>_$9GM-qu~Itg^x}RaRJGM+c*5S;3ss45~tj8hE+(tYkie9xlQsk)AdsSf(!*g#hM^@2eqI->otpY!Bkv?|tr#6a^@d_5 zQ93vnO0)zBZ3y=$TfLnv4{l4a)l@_sI4BNodaZCsI67SyZkyb&gGXSW=zE}=9doTE zW!Y;AU#JF)t3=QyyRsBD%Akwj#o=K1VJU@evMG#fZMG|&L8FC^CL9aIQ3VNivCn8X z-;HFcEsOO9HP{=mSjW?>pGa#ErKXiz4D5z|)Q7R;T~7eTCYXlcOp8KIC|cMzeRsnB zOSR!?FvOauyl}h*|D^NnoriZP(~a5ITrECcOknTtu}5f;nN+$JilA+IpY?WV+6SnaulsOc!;=&0XMRDVsHwM)!ZVI8BpCLG ze87Q$ISdFH91J$_HRON^>w{8r@(ilLfn(1F)>ebWByaJpdPh@po{=XnODiWtJ!THK z;vP~_$i9}j{s!_TW1I5anhA&z_@x#cr;VQ}CUXs{i*jdvVz@8thkX>R zse}4}2dg`Q_g8AwzG!_JDFqyGBMW($#=h5QCyHwK|4-e!2U&Vtb%K3wwaRj;C8R5C zDsCvk3P%`ui}hT0wZ=lNSF0?^GFY;#Zi|1GslDndjS9D_+g&BO8$1y`fQcAvSO>Fz zun`dZhlv;YXEG!Sl4D1r@$^5=ApXZ!BnP1+r z_m7ggJL=2K^T?BV9{D@nxr0Ftjgu(~lrv^X0Qx<4ZUYiCb|GcS2yzEa%w7_IjLVUG zB?-=!yiTQMlN$FQYd6Ev8$a7GX4t%~LUWhSo>4$K*=Dwh#L)1jg8rn~{f_5uRPb2L zU3VF8LRnU=G;je{l0#5w>iPVeVo;GBXfzi~-Q;SQg;+ey3iY#n?D0MGNni#>Bn+?= z;&_5|sORG4gQMmC!QNHfzE;5lsRHH1DwYcwgmzJ57#dtcjh3+a0ycfofPwfH2aHF| zo#n`iHh3Wyr9gbl3V1f!o5$Z3orO`QGp1U(SlZ5M)L$nkS!v}nK$K!w8pKkP3g`l( z-6MC@C{ZnBhgb4JK(QutzGtt!mtVUl-CBGHhxhRbDAsB^TY327d7ua=)@ss1?>;z? z<>iVd1kivu0*bYoxU6R{@{61X24@r=^J?yrSDyfG*%TCOHD9eyk3q3kj~Yo!=3zQ2 zGr}?N)Eh8U;D6XN^-rfLB0?>PI$!>slK6m>Eg;GHGffB7e*D0IV_`Q1#maCwxV#58 zFYEubystYUCHHkW>~DgueGo%Gbyk_T|K>jw6e~k(?7xdHpl)R-6e~BK@9f1UvJ#4Q zeI=KDMmX^A3SbF9z_5(#L!3)lRw)c#vF@fBBnG$eQHW!G>?_t&gTG z@zn`rO_I8Ji(}YM!qY^BnfT@6qptVLMKJ(Pvayj`ZR|p;opXJA z5#<+6|H5yW+`GRT-!$`nxit6gj~-Z@_(HMpJ#)Fgs-G=aC!74m`J_+8xp#jAp1jaI zZ=8Ggmk|&`KIxO@-u;nz0wbTVGWRZ@&)lhpUAxvy2U;ZI#kqGmwboA;<=*AvMpBdI z-sQvhZBwilS)1ANXXK8iE^R2zz5Bzm+Fg))7msP0r-%xe3Wf}P_G99>esRqlKf1|*kyY3|*pH@?`55)J+UvfR7>1*_0)>?Q6nh(txD^~JgOfYj*XU(DU% z9GSG?OO^xT+lb5a-_A;DH$k!gt(6EhBesEotN~2``9q?{3t1 zWEeM&D@ed1Y z44&OHK1$aMX{exo!_d|0uI+i>)#~KyZ@(Md*cdv!w%%(RRJ(|EDPQlMP%iYaW zYW;%wRo&J}eGq@>nviETd2t0g(elg&{+^h^&jfySds#WNA!Uz&@X)K7z4`)x+8kcF ze7J{V<|SVa(!pK;CBGaJ?O4b$UGsbG<{O&vre(_o!!OP14+*VuzeMPE(TS;l8bY^n zJGoh*^g0A1;u5~wNeI-uTr_c zfCFtLjPusxt9O~O=2fISp4X}PS8!~Zgv9<_zIA(fbn(URFVAhEsGiBEmOE%a7W&tB zPH(<3v6Zh?-+XpvY^B|{dhcq3wcP!looln{8;$kSf4L5SnOFuX9Pllf|0!rvbbT*A#EEc>eqz%AUFL?xj)Zu4@`(I0udb@ zZUjwChFO#D#4xG{Z;dW!@^#O}A*?4=HUr6yV zL@R$vYF5n3Gs5k!RAz(NO#yHjqMP~o%P$-OZtLt-{U1=^#)R`4AV7X#*hRf8dnUEp zW7+wL%JQU~22l?5Qb}E=VczrpqhtE^sNJO-{f|E(?=t;>u!d$>ObAQ4@UWKA12ga^ z=mT@?D%yZ9gU^oPmn7;&@PY+j7I~P%z((Q)8Vy$ZzBgYAiY=wSQY%~Dz zEMlOwswEdf6f1^gL$8%iPGYjuZwlWF4qD{FOjk>sg_{;HU61nE;n4-~Mkl)8LXkZ& z7yIt8D-~v$mKiwe;RY_Z>%{@ttuP=PN%!XW!a7&XM7bal3l5Wb(CA)v2y5h zT3s>`oI{t_`a$6pYcM8fLNiK0p4|>y)sSai(FzxJm}*@6BJ+91nNQL2i*e%6x?}cJ#tCH zL70Q6DyG-3ekYZ#Kz_|@zbyq$DAX1Y+D zXYT~ibm{|Z**7uedu51Gty+5X73~3cqBW9GzTWh2rs|!Ib-fFdTS41RL^ELRY%)$_XMdm{O zDK`tB$0P&PyoyN&VM(>#B)pF=GE04}-g@_cS&^P~_2;!l)aeFetI1M%y*Hh|{=6>Z zQ(w)8DFT-EwX3dscLmq7uzkUd#;%&T@_CvEaT2fT)%(OkMk-NK2RW%Yp|XBT5DglL z?q?2aQHW4PaWv=ydF<6_8DXBwuRc%t+FPvHD;O95gjFE%~^(a{@sQwyn_J~CyQLXM=?2)p9T9lC^_|nc3eJ%UAFA{sR zI+S0}XMoeGyGoQo3(?i0NKafVTc}3KZkzXu9lL(mq#gyc`_jtl&&aQr;q9d*iEE~_u+E5V{+nYt9LajSSQXY2B#b(q>Et1bC$2VFM{8EC6gq{<&O zJC}MPp_sYwL4REyXjY$;djkC=US&$IhA85;FLl8hy}A>wl0`Q!mNK^|!K+Fs7iX!C zl{QL;YPH-;ik`2w7{7*m)Gf(_klS6K7oc{dpOD=*UEjgy_l}n7`}(1PlJ#ymh=#Y> z^yNe9f!X~9>C13w;l&=@$=nmWUR2>IL3|0^)Wa8-SNC>z4wf(N>FxuY%~o3&Mq=OJ z>vkv)REaYRJ6Ex6-bj;(U@)s#9JoU(QK>(hNJILeF2&*ID(S zRl=KFL~k60uH#y{dBD|)O-Hq3XnB}gE0plM6*D*rOxH1()plngnz2az_(VNjdfE&v zv2~#j*u~wKFY0ETJNpN_`xo~P2U9WnHSrHs=(@kOzA3i+mN#$A2`8^{O|apK@MEWF;?Q9^jpv-arj=tZ9%BVnD@E08c?CcFzhNA?en0DKM}2g9@D zMf zt0XN%_YXoNM0y;l&M51>N3K+nz4ypZc)4ZYG_>ChzLMt4nK$s~4`Ogzk}rB0_UQxN zdyo8{=c~P`;Jx=MM~$RK_yv`N_jvvdILp{u1LD77Pa$1V{+$i&ZFk<_#p0E00Z9_Q z_i6_WaQu&h-E{9gk_0*&oEI*VS!zIK5)6i8213tK@4XGW_Z~@W916x!;>gd5-h1Tl zvxb8W1UKL!9s$nEl6GI<+x zvqpjf`h?)7!CtP6rYkub;#VshifqU_vAwV3EiT$PyL-AZTp0N?WL0RfIDB(YK z&r+Zq0*9YK7Lq(k5ufM%;~l)>4VlS3MX>}|_I21$R|#AWWZGvgzmLV^J#p@wf8x>y z)UQz3u=u32HJF*etwFx%BUqA`6cmm*z69$8VII@Yd=aMTOrzaVR`W4?4{{pvI^g_a{m5Q zzf8cN)ghw>-Upl3ktaLMQ`A7(j8pbeLnq~_pZ#%y+B|z1R35pT| zDI^shTrZx-1X_pfwKiIfZQjM#X*L(Mq1%+K`?(T=$Tmk*AX`v-eh&11ee2EML7GLc@P>kQ#r20Zd(*No99 zK~l0NOaYhyUK1NEVH>vRqr1BfNJ-2WB6JjtZnlb60=p~%5-ICiudA0bBI}=!w`$t; zy&3NH)1CKwb+y7r`abm**05Sq0}!A_t+^^EZ-aycN{hp_3~Ggw&f*{OyTF&D?>!9~ zI|>0wx8|#(5w?U~2w=!roxD zu=C<7xl=_J?D!J0sIIsXRlqDhiYpC%{}k1`GM%u)ArkFWv%EQfOj#=-!5pGepwTKJ z^cQx%N~2@)yui=q?od4i{iF_G2)~3-OC`E#M!z2H4#EPc-=g4JqrrSEnW(cBju;hv zAHucO$T@qH((l6XHmT8CY}kue%m5jPzZ!>PE!n(3le@C^LwhJdIBsl2qc5+`+tkfN zXwH;GWLc?C%%KOB{`2{(7j-`>r9PUwefG=McjZ2JDHVxKCrZ(MpeW&W+@%XLK%If- zb-^cV%v&T`!fD-VkS<}wABpwyj^0YGIIQrH4J({cr)P_7SRq`UZWpu32_yXVAiRJC z4aO}-5I6{f^ju6s{oIrQD#OJTg&7K%G9RT4n` z`)Mo#I8t>mQz#uA3}pxbl7T(Ap*?JO;Js>V_U3{eSc<3viDLH9bHZtiqcfI7K}7Hu zrUX=TvRyk#MfF)E`lA z!s(?-5NopX!tomDNslgHdE3?H?*87vQ9dH#$osL0dQx^X8e5p95h*vX=7=d4Q!Fk$rPO+G45MAt|yQLmn`IgqwF9L7@OI565 zCNk66s^V{rdX`Hm2+(F@a{4j*dE|_K%VK#1FK;#S;^J`c(hH0CzPxw!`l3%l`vD<{ zI)~sT-81r=x(Aptx2SxN&!+g)@_$%9$&C5 zs_;EN0s*`FyW7fd2w?9%ZGd;uir-c24PRl|GF=1j5tQi;_IFL%XN(oTUWhFoR{tQ(pR%&Py)Dg?>p<5W+*$c*IiCb&|_0_3*s;PE&=MDxrG=e!GQxqs? z%#ejGXPa;vkeIOxDN9C>I|xcD^SXkh8d}hiCLH35;!amJ?myOUhNU-twqMM!d0T~M zHtzi@7@-29q|2L7qwgF2G812r5lIpMO&fDw4@`gdE$B znKwLQ+58GT+s7W?GlK_a0NK|v;v4y3`>JpN>Tw`V6=PWjv6JZE`+ws(R7M#g74XD@8#F(uUm_6?RX!b zfMQML>BbJMPZG@#c=&C}qy2jVinW?Fk#~Pw#1T-e)x>2z`}VuRSJG;~z74;fSO2xC zdZ)yz6z_+@r^ldJTS-giVM>?W`|}Mr%bYM#NPvQ3t>sYX%eR&!3dLHxM1xvw51i^7 zoRcb-W9YqFh+d0sRc_GW@}3m+;LdaD#%|~nxP?0>UytS6a&yk+}KF- zN;yxb%@5&-o8sL2E6YnS?|sAM-u>11cANLhrMY*1^wFMqZBD>ngij*vn@e-={>VIe zBJce82$bgD`G<=*AP_rAKdoJ|xw&wpv|T}((p?p-|QLY#Xq7*glnkB7;_cAMxf z$fCHpPeyQ!kFU+TMO0UIcIvD$*mfPj-PbQi?=M}Pd-p$zFZRU@ljYunDg6t6k^kOA zf$;l&e(y?i?*XaN#lM)l$G&Hzx%ZCBdQZMK6yxR%&py^)ljh#t;DH%HJn{z1PmIfm zpn;O+-W`SU#us@q?sHl0-Ti)Wi^zNR4fu=ib5>8Jxp$AOXoDAW(VkV+VQjY)Hcp_V zx%Y0oHvT=9vQOOVZbCoz&Y0p7$KLW*oO|z_|PNqF_c!Wx5T z_sk`x>xDGD+xG4-oQHjK^9@65th=`7eQT_fufP3nfRDZbKN>Up9Q^1vhuDMJME?C^ zG@6WA)m#)J>7EJnqLO_QQ0=&VO08e$?sU3^`aSBlo$hJMj`_#K*=y9DXYZIzZ#Jqo zfj^vrpL6Pi_(Ru(JR5CVT!Bus1G8}+0^3eZ;b#Ir`sIRu!m%<@kg~@>c<9y4UVQ;T zZ4R$oKHNhw^O7$I>0mE_l3xyq)?{2T)CW!KIaX^f19Y@z_c`?;Vd|t`B6Pdx$B z!Fhc*@b?D%4EiePHH2V9ytd16s*SxiN|>jke}iI+BBXYrj_%Q_{UZ}~uOuX5r3@Tj zrE-4(2iiy&=dH(A?=oS{t4L?s`NqU3zE*wn*_knlmN>m#HFe9~@7cL#fSUThT!+6*tWqi*@GY7DDQHr3 zeK408s@aK${VhvJOn(u_)xsw19+6ZaJXBX? zS2hqPn0#(`p1bQ)vfcShI3!!5<7Ucke`1hq$?pTpWcQk$mI6f^V#ey{g~GNdZK9Pw zB{eH%V>c)@}%i#*I>U?Xu~h7-9~>(5-S4oeBDFO8GA zJxD~wR|8G4;s7}FsPFd$AkE*8cYu}(6s%om{+nsVD3XEiHCFC#am*+o6S`#IC`rNL zGU|JfbcJz$MfUb(P)~E;X>=VEx!m5w9V|$Zg)qyVi=u?XMEdx0jHKX=al9l_Ss;}b zfVSQO*iasgS9T!OTaO!|?rc#FB8`J47aap$G#~EM6 ziey)+VIAcx`-5HWE0)zu^@GAI)?iG|gl3e0Ji8sZs+$&jmp?A*Fx9yBMdtI2GoPa4 z$rDhS@$Bif4BHHBBKK;qD-+OL_}xJi`qJ|H-qjuds?x&{w(T1|pw2yR;m#jW_s{O5 z;LbDakxLQ|!W=MLF}?n~biz+TVrzP9Z?-3YqM+5&bInD1w7a|a%HVIxiBzBa$b%~( z@{3co)0)T&1AHQdK)0Hm5=Ov0n^gU9y^U6>53FVJD9iWC5TjbP^yVwt1J2a8;WB5V z`xgm+lgAAxl|vUM%f0F-9o1CDuf)859k~8i`RXNvKS5C}Q>E8q6d+}@jnpbBofXvO zYovtxE39htwmp}`M%XrgXT`K^jXYl3Dv}8Kj<$_x^jpIm{vX1k_^o>dm-Du!^*&R7 zI8oarqUEdQxyod=U3SH*h(H8-S^gz+qG1Me;p}9Dz&JEnQnJfO6v!j5>ecV$p(+z9 zEonl~m1V-%D{=0g6-NAb2V=BC@yO*D?%g8F@Yky+|%_13%p%Zl`@ zt3R(bqK+fXaZpl5ufCcOQv`zTYgb+O?h3ADVf%s^ja@Zw~iSTrnCmx48J74PM2U4FEVuy!fC|8&>QLI&EZ6shtD&Ccu8 z#F~k`MNyxWdjh>Kd6g--8ls5XzSP;;l&m0vniObWEM;y_f>)JNF3wUNGnFZWq)!cA65^}&Q|SrGkb82 z<0*uC4dLPK#A)UTY=hk1Ay%Xv*?BXf{$#3zH@Ar1I0{|IwQ}=-s}l>k+>W7m(u+IL zbODg?x)n1xf_155u(J^DxWLl;;}i9C>1i{x#9nb@_#dj!b$@AnQ*8MyZ{C;_PMEJt zDjjcCx14>oev5oJlsiuJ4T5Yiui%OBW3@738pszL5}&zp+SK1q)Z43p`iX$TxOP?ydu$hwkoh_t<^;qHcJ*vwyI=e{t_{IF5;j`2GRP zRdP<$N5z3nXYS(V_4}-yjI%v9Dma$!K+bJJD{-B3d8AaCSJ0ats&O6~wTz;0EG_%_ zSZ3tqGWhiBW20Cf^spbtyMi@Co@B&aHmWs(%Uy^25qBD^`}vrJf({5Hmg>ce3J&dY z{LK?(=~sU-n!2s-jn=4>cRrHfVL#<5*MXN2li3uXu!hcA855n-_vta_MQhWR6`p`v zu&E=I7ovplqG#>V-O-CaJx0QXSv|7%-u}T6fX~6>V0d=CX#ds$&@SZomE>Ba_nfJ2 zk@*aIUXo)6(NV}t9b5!X?Cum1m%f4xNS5sDM=$F9+tu<13hYk{itr%Ct6jyHq^&4dGy-L2?n+o20uX5B# zT7+LvIe2OBJ#rfEXz5u$oVj%GJ#utbpS$x0FBY#G(Y#q`ZYm2{)L@tBigFzKjN}O& z4$ki@BqrRzJSM?l^ufXg-FuIuH4X(AvWn=vM^DGTO91AP?+Y4m5sv_8Wva;Ddo-C* z{I`&+$Gk(}(trm$5&G~q2;Ix#o#o-q5!6y#zB=57yhDi%J#Bu=0Izo2_9+V@IlZy5 zDQ-(ftX2Is6)z<|2u!R2nqR#2^1N4G4!{5)(5!`NaOEOGI#6JW4EY zB9tm0&#-v798ETFR|rVhiUnL8TITa1mcN8{2!NyEj$!ALx>|B1#2K^BrcNfDpt z{o|c-1rpF{S%SY>QCA6E4ursGF29e(<2`Zioa})56$%>`pLDhcGZVNq$QQjFh__J4 z2Zk=UZJ4WGr#mG)121=RjlLD!dCz?h$#@1B#yET$LXdT>amW??3n)PrPXHV%;ny!0 ztlgGQ;;=8s+6Q@yR}btd*f{ElK)XYh0wV)vH6!oM?RSJF=Tr^1f<00qJsn7X3f?-1 zxz;aLE^`n}iOher;IedKdGO-wr2emb)0*{{ET}+?CNR&5F?l&ait$$q#Av4N;xl3~ z_Dh|M?+=>GD`J}#$rjJH6GH9(L;>5(v=!P9OO>me;d2S!N+~!~NWozBii|`-a%)3< zexkk;4Q=;YRYdRi@JZ=%aZkrn!Q002ZbML%5SV>a^bbUAvRDreP2$E2WwF>(8@i5f z-OsNKwqsuV{?QYHW|kn`miw=;P&Z zHL{vFG;YVnO$eaBQOeCAw1#9;#)vfymc@CLUWv*!yAH77*E&qX(aRh$W9DrXaq24t6HsMW4&*$HhCS0I#UB9{g*8GOmg9h6**fvbWb{&g^k02~bk0wY@>LXK$ zk{Pi?8&t@Md9~Ep_b(&3f^ShyEc#_K9&1UJH`<6oE{#!NZspdhDF9p%7=xSS_gm2vDQeT$PizK|%th#o<~8 zwZchf@sIdj;LDLa|NBh{r&`c38BBR^HqAf$nyd}o4Z5x6!eoid?EZ2LM@f(rWyUZ z-W|k#cgOEi)Y%G0jEcSw;o1u1oWDTXjR5b$@HVN@T5Q;hSIhtzh`$<# zVzu1O?_%{cxhrcww1)zOBb`Px`ts5T+WD&&FF-MsO8Qde`3qk^c6g+}hG(L=+h@OA zoxSOEmr{|)bfOg92NJHY<1Ss00qP7quM0j|W8NZl38y^ItxiJ;BmPLNmv{75Dk-Uw zgcTmLVTCj5^lXt0D}-}~_ioNn)9qq5IbnnY2jPXptCvrU1sq0_90wiL&rOL!Fq6XOTw|$*MO5GlkN@!BDb$6>ew`+r4_P z+M2z&V6T=U>Oi8HJ@lM#QXHMeyLBwl$#(sPDFM}-Y}Zaw(S9xAi*&={DluB#<-3xz z5jr@`UOytGh_`GBBU@w=wcV$E%o5ax3rdiL5=IW0!vk_VkR=v*{b4ije3?#Ne5^PLP7AD z{XB9;-?vyE!OL5X3b#1iyY#~1y)W-wy}sy^(0)J&qRt_BiK%WJyiTG@cW6i}Y?8#T z#*f)>Z-S=^u3H$E2|IdFQ#e7z8xpjQ7dRTw-CJ3?l642j&; zN^QOpt>=trFPOoPYa$Hw)v0=_sdjhgPJ5Z6KsjU7WN+ye6SS}tdD&?sX6!=Bk`d$% z;zHkECjukMoWuREBlXBl7?1=f!FFQpW>|XTXZytro3~ZyCi{%&9Uu}Fg$bbhlVbNf zp1V=OV=;Hzup>|SVCFP%0alVjP-*J<{F|b8nI+Cs6mn6Rsu-_B>hFrq!l=?2Q>|PqZRa%VuM@mnY2`COlww#K z#8Q$9=mMkNBX`s&5qF>IYfT3uO%zWn0*W=E^F4bNUVgXVx(DS2|fw*%UuGBl^HiMtV!@9^zKvQjN{%`J~VLz6l*ncS^3uzD-!Qp%e>J|{Zj*cWM-MDce4&suNvTbx_>1sKq}^9( z?%f}mCr{*^A0L6z+`F8>$mdJZw$4kO6Zw1{>Q-;2>J#YxgjzSOs~su|V+B+9Uz~gQ zC+F42`st$FyL{Y8YO>tBeE7a?iaku4D0o&D;@rEKkb>O1cudngMO464Fl6Af9}~Bw zxp$w8;2PgBYwpjHNAv2pZs11bwM&|N_tz5p?xIA4KY%Rv?tj4=6xKU$qBr=XqC9t{ zx%Yt7$cHcHuDkpHfH22<7S?|f$My}kv;W@~oYp$xU1k!21(Z_?f_uu4xDap}VJIAUyPHX0N^gpf-nBE+6ip zn0d*UgLJSLK*=wMM6*oQ`~=)-%~;c^<$__87RBJ3sSgP&qWUF5w~J1mH+8+aaDLtm zJiGxv<%$g<7!j}Sa-3>ouZFD2}*rEujov5SRGH$a{wSQ!y?v;c@tdxP{zH;LX z2VT<(yvb|`^#vSgBVnAk9$&r7gf*`seXNSZvv~{V~$P^-IS-zZ7J;WFxbkn|u?`8H6m$lksTqBr-Q z!N^jIm0I@%=~%&aOyqKV6L+v6MHa#=cP@$&4io9)%Q2E55#ats5!7K&s9b3W2K{Ma z&?TDOfkM|UUFMt)lj>baFwb&V*KT7SPRKpygAYew(HNc?IZkL~$LNRnywM$riWDAh zzul#}^oaBB}_+D^O2M=btTH-9+v{(>&6bm~IrO`$9#PfAZg;}O$ zAsn@sxZ;59)@lQ?5Fadx@ol+)tVk|FT1ck+LrAu#+^Y2Pg=DvakgS)%X0tZJVB zcB?uhyHy&J?MQb#=W;qWP#LEmG;oKgOuLK10$zL+lohGgV7bhemz{k!jti(_C6wt1 zu8F{Er(_7{p0Hd}i*e%6cl=9t$~!aRn9?F71Jwtm)@QfI-3Kn zo}Ozi(xAVzw{OU00QkX8=2kE}wXJTR%uZ~cRG<6EgPUzQp}sg(JFSU~>Z#Kz_|@zb zyd7B^K56B5(X~Ks`8vlBtYz^i%lFFQO|4pb^A+s@XGj}KC@+VixW7Wv zqV5^FBsRjfd20pdZbUuTIb5qY{KV^Pb+j#q|A(+BE~BmBvQ_S5+fM2aCu+Myqxfq1 z-~eN*%(*Vx70(0$dFVB2n892)JNYOu4s9OJBy$KV;r$zNrdRmy)`}iWW)Ys~*cmJ0a=~-8QUTZ|P zZn%)h^Sb2DF^Rd3kx%B;SMy79Mr^y&bsAUNMUL^RY6AJZzUY&%yRa^u($Ibtof8 zkW`}Fm}cRHGwU4u{Dzt_l5m{~&uhS79@Sst%^tC6BdXP%i#<|SP>V8h1iu<3OTD_5 zg*=MHZVEewIuy)KMaF`vMNq$<&j6=WcRQ-kLYO6IE3Ko+Of?E+B(2W8U+mcRD$}$c zr_3m0;1ZOkE1rxK+HDtF&d%ioDU$T{m;N z(1=o7l_FLCpxL?93kk)XC&XT()hFeiKtG9BnUbp^ig@8meO&uR4Erk~i&8t(L2q@A+zr@oUIO-I6>Ad5La;8~udrzUlf7KEHRgOyAcJ2Ul(uUK?ulJ5s_?@gzLWqUgm+<|DjNc_U4-mT$8eqi_(GoZKQrNA+hDX-GfRr8w?X#oZNfg`;2N6c)=9@5?<{1~QHe z_u41zet%d!I6GUl-_7juovG_Jgon2irIPi!{U^k{#DB*Q0W^jbtRL5YukcJ#$O~j8+)YGM>&Cn8i#f{;As6yBMrS(m* z<+r?fV@|Ycf(>s{9;=lR(?Gu1koe4Xikj^>{r3~~_Uh%wEy}T`@&yuZ!pD-!`A_5< zyC!$nfzLyCces1(zI;(PyxrM9*xkRlcQ_o!qhIWQR2u7@sAUv|V`w_Nl<9JuFX2_F_n9D}BMsT_7a6jTsA=lyu3b;pV zQ@ZyaId%{og}l_kMc~Bl&cWrYFD)-g5dOsooa7RG^2mn;8p!Wi^eRA(%abNT_dhP; z$liOE#AV+!ysIfIP4B(l;I1`N_{#SS8~(crMfTpSM0eic#o~1%8YKT`T5bXzCE|{xjx+X1YOqV_-g_iZf^aaB?rESL zNoW{-{jd=l{k%qfmTI9*(i(??ag;dnbE5Yi`TOi?!`3A?U^*Th&U5vTsbn`zW)%M| z2z~e)gzjbW&hl{Q2x=)VUmZeH-!i;&ylwlG1(BTI*pRnSBbMNA z4DPZwQNK;aONo&K4K+aX3xfkZkc8E7!TnX-TaB33Cy2h@!3_w9Um`|LiJ6^zV*k=5 zBDgIcC6@Ob52eb->!6mcH+s8L5Zv6mwqub-NAs6Sd~7JPStCI~a6%-uJWlgPW}I6I z_I_56RDNKhA_2{;x4@iBo9M*McFpyPKEZsv1bk7@&~38tGhQY1yTzleKw+VO1Q z2v+|{f%76nMGsNNhD8Vv2V`QusRMILo2Mrm;lq`sdAP#7Wue8;Tz>$5rVprJp|D}`NoQ*?Gl5%!e9_B+cngJm zVCdp*K)vn^2Tuvlz{_1+qi+Rw-gDnWvO^s(jB)rhgdpo$`ut!Ryrvu4P!CMD0 z*ZQT(We$QV;o6T<$W{p3E-VjToSoGFm2XY?PN`gyxc;XTDi2}BnX)ClJmMT{_!{-vdl~Qn~ zkb=SL6&Z?d;RL+o)M2m3dR`<;--?rQs?+N=^!}JsxkukM7XN z%i(HdHE(F#MzN?s6bb_9ZZ*Nfx^`_ak32#w)KK~Atfi{9Sx8IuIuzJv7yN*~}ven`s`nr7i zNDy6pWGYcIBbI1`3K=o4mU;zR61$1(rJ!FXoqOZ`oZQEq7kf)o@3%_1Vd%pZu=33Md<{YU-to$oePbe$%etW*~QFbO$f!i^6lS zpKc}iNZ+U4!WvdfY5)S%s5MvR=fzcWr;04t@q@}v z@}`t*R2oqQ%yLoOF`N0PsNR+7gk9OcXs4Rx&G}=>S_$ZYMQE~R9GT9GRtcfMcAc+& zHg|{WDd;D4_(J$4gjy=mO*8s+y*q|umSUIsEeftR8qCL%i8@>17>&{QAzWL5oEaao zza$MbDtH%$w@HoGV#8j%Vg|@S{M9%VtK)ih@#<%CSJr-L4+RKEI*n-b<>jOIjKcv^ zAI;r9`{nApa-X}DibSRprRYA8aD5$j>4FSUXW)5V@W~qU7D<+HTDKacOBnG-V!gbh zw^CJMg@M6%>SFjFWU91JBJ0)#SzhwWay zS8dJST(DP55p^I@%pQ79I4O=!+W_8ZqT{Wueql;LH7DD(lT@@{OZXz)u((RJjt`Dh z9}URJUOytGh_`GBG1|>nGG3}`mdn0$(A&T`2 zX?3E>VZN|v8|tSij3shC42liwhTu%gqI_6F)gMuB!s(?-5NopX!tomDNslgHdE3?H z?*87vkzKqXhVWw(^`z`(HYi;llV+}N+vyg%W19OM&TdxgXHU+iH|vs@z6si`+qc-q zL6eDpmt0dHp7KGdRH(zE`rUw+tJBPjB{pxy6DNytJBDiOozS!bsF}}FPvQ8<1@SXl zkSJ>EeG7iZ#r1-X{gV`MATvh*A%la#GGBuaSg<|_H5bod0~|Q^9J96#B&PU?&s!L~ z0Tg)T#mn--30aF-!!5hIJR$1$D%ambzEm7j#H$%bjKHrFk{~@(%r%HEcG2BZ53YPm zYv~u%xUMU(RK+T0A~T(>D*o1}XStMu0Bt^02<_3%BWLutFP2B}@>Zk5Ee`iCy|8%i z%X?R^FZv|39}t45a|m8yj%DX{QpfdllL2ti)xE+7O6*F*qaUtY7?ue;!sv?L)@0C? zy@1q>b=w_5o(Qbv_y`2->hEr=t5I!fTJgJzy)hWkwPq?_UIE;9W%Qfw{KHgX@H8XT z2P{}u3A|dVp-oUnEIWp7X`CW`JEfl3yIZVjcawc=_OHxqq;CbvB)y*yIx<`hiq|a$*(Bg$zQws4)x; z?$}04teuwPbupv2$#!UrN6ejNCXVwIgs zJ~O76tZvrF2GxO9LzH4z8pKkP3g`l(-6MC@D3Mcz+R|tWiZ!A0J$n^ie#VopwRvtW zfgcYi0vS}GpjfNv)R)sCn1R(dLZG58d9;5|K(SVnCi3o&i#P&`wVJrBXD{-LoCXGG z6dv>Pm6R>9O`Y#ygNspI8NTRc*nj{9#maooQ)+#B42rdS)JR$~4^u20^G?0wwme?x zDJa%j4t2i#J056i!Pv!5EP1)o6dLkViU>w?hSOM)^mrKORD;sFhHeEOdH?4#vpP|Hg}MdUOR=Si3&6E z%f&}s@0E*U0Gh+bMw(a3R>IXjxit6w%JS07d;h=4z56TmZTgxnbrj{^{Sg9-6JIDK zO;T!e0{$X=GHBmintS(0=E)Oz=f_8&H1{qiF!K2-bMNx`OkGMSO(vSi`RJ{1*_2bjXS@!ST~@B z6vVmrfYj*XU(DTO-!szOdq-uxCtn+iar1^pEaxJMbMJ2OzzpztZ{9y^82WRoOG1C* z+`Ahi-uNO<#(gf!y}REJHht&S18b<>71g;eHz8q2SUr*E-rX!!K|1$2uRGTB8^2k^ zx%X~uZTx#IWuLg!-GqMbok^o}=K$^DOLOntQW77(z-X7`-kla0B@&Qk^#P9rs2>*A z7(BaYe3Y&i(vWnfliepb-!MFQyK8&iKX^O&`rGdYcz zu0DuAbWO;!1tE(o(1~_nHqKDs;fX2yOyEbimkk)oR3n_!jnFU<9(pyiS6={7o5L%Y z5BE^ayyVM4I@k-KAz|vIUm|q7=)`nW*PE-g zVnYZ<#A~}8r`p(Sql9@n`Zp-HC_-u{>gaat4m4J*@#;q=>Rw5{)=C*TzDni(0uHp1 zFwR?#uijb6Y5?XY#4#4%&}}{`H;H z+6*mszh~z@tX z-#pS0(_h4KwXg}hM$SCdk~H}!rL=OY>a@oQI>2fL5$zjV0R z-M2@KcC4x}eF26T|Xa zs%!?nk5jg&g$@Zw_i>@J+4rX+1v1NyE}s@8HH?! zY~gcveM&(KF%b*Ql->TsAlZ^70?TCgnx2*#ZvDJa*cPQtwDPBi>WOHzu6d00F~P^3|!AWzVE`dn`L2Q8`sD4VElB9i%SP zFz@*nmsj=%OM&0w?}nSZc8PBEKmJ0v4nQZ>4+v{$hQ)-il#51K%VSsJK$i<_!HLMi z>N;`^za&vNf)_0KvdF_61~w8OG&%xdF1iBQEL_kJ=S$;cZVwVs@zp?6tT+J9JnH*> zuO5>Sy83>+1GH42VC_2d-%Kk;kqmqvbYfCq<^C4Oj1n@TOZJVD6dW$2zV{_9)_(Ph z?Cr~-p60$Y7}?aigJ~5utc2^B$mRAX?qETREQDF^Toff7Cep{3(;sTf;CM--vOwiZ zJ22=^3xh5LMg8Os6j~JgjlPENho$n$a}lbauns3A8-4KM2rL@IGb6_djqDixkZ&{q z@hoDXwW=i-LKG{8WJ9kQPEK-aepC2faL^(TX1ZG9EZnqs>3WpM4v#K?H#$*J8eL>h z%*DPt>`H}Mrez@%0Gl; zd&)QKBC&SQRD@)=zQIDWTh$@it1EC6?i%Cx&EEa1gQL0OS%4VKGndD+=# zyBGcPK~MIELZ*S^Sno^j^0R6u3Mv!}NtF&@}i zUH|v)APRkHd42Edj(=6@VF=syjn=taErQ1_-1!6Q{@HyL+<9g_a!JBLm;+`jrdRMT zy*-(9HV0ZgJ=a{ML4WDg-&CLb$b&18B~9cPr)sA)kx@N$S_QwFozjqRE`H{jw%Yga z>H}+8Jj(LDGQ_A>Exq}Q_5f#UW2C1xyI%dxRK2sYu6JRw^NmL}Rq-n^ul7dJYO`c3 z)zx3+tCtY%4~k-iMd|+OCE2V?U3Ku)cW`E}B&U z#(@a*vVv`@pE~5i*~tijacEpIU-5n;SzlT+|2ui8%4C$dYQC6F^9CBj?&BuB!ci_b z4OLJUnG5-++$?+^lMEoNpKrkv{XV|PEcLZ|>)rolMS9lNpVt~u?g~d%$ZLgiag^tS z2$K40K1>nFrLSFe-McHemWAyLW;AxS(OdcViG|#_5=9UV8i?*^4r);bV4 z7p67!=VN7Vt59;Vy%9aJ>OdXJ$Pu7Efyky0Sj~3D@-~%>BPt zg`qA_{Wad~5sO5kTHU$WBV`4(C?iMkt5LGlt7};Vu1M_7>QH_?p8-y%?kZ6VEksv~ zB0X`fY$Ywp-!FFT`eBoLl)!LN+Y~p}>NE1|WjMX7NReKt!tbFbC6~%eWtTTXU(Q#8 zMZ+?6DOlrH@m|i>Zmb}scoLNVLpA#X<2C*__%KZ#eF zlB*$#xa~`wtxd@aBB)7$=EYLx_9S>!DdplU)v>TLB}G*h@pQJWwUhDC}ItvUwv-+HA(L z^0*Qg>r#IpD&fs7qBo90*Kw`fJmBiYZPA8$RBk(T z0g&*z6*D*rOxH1(JqVroXr^K6$0zFP($i*WiM`^+@IO?c>;BUErr7dZ-n=mv7suJ&$O_)1JXR|srh$C1A@P};UA0wK_4gC?_Uh%w zEy`zioR9x0`=s98^exJ3Iakj@s1;<+LMaa24QmV`==uHpRI1i0lMo~DHmi>GzGwe>u z8Wi=hQLGPo*pK5~cBz&qohKPFmyK$T;BwdDe#D)o#M(gzgb_>i;zb3A_Bj5QQ1=(3 zsoUz_XpK5~=f_7d1YSl=W>b8^8aiiXOms@$r^lEVtxa2YC+6j>D;!5CFGLC9MbFx! zyQ3FV0d=CX#ds$p6tm1JIS^9ZaV*B1zq8K$}jTRiU-~z zb5}ll*idBeJ#wlzK~%`qW7!yX8*21lJlKiQhrdDSUKZ~x4|k5Bmg4f&Atd!J!?T0i zwomoI&y%xL*goQJW5g2t{l{JQChE7TcquV*aI_hq`Gvs&9!SFKi{Sp+k|6qe2R9%b zeo0Ks?Bo;smo5>(ZSg3vyope%e7p{7+4?HCD+R&Ly=yy1*TM4Y0QqSCGP%(WMK)_B zC7;D#b*MJg+P2v;DZsro$fYkmd1;`IOAu|>8Pg! z-K>!!Da!p7ajzluoR+A6EVvX!4^hU3MF4(4fm7-#fy;qR`^@F{IXqsuV-I~B4A%Z` zZ>A^eS14>)eA3w(%uL|cAYb%yAl^bD9~in;#53@67uV=p!JYTqC&Z8i3}YNV4I#+7 z);Qz}{sojEizfh%IUAJSYwDK^)^1BDao87R?Ss6VULtZPY05pg0~K0uJucm%Nzt#!nGfzkgX85U05ExI6JBTE8nzc{Uw{WK#V3Z z&xuj~vg=n1#Av4N;xl3~_Dh|M?+=>GC2}lan-(`qq$6I~LD ztU~JVPZhjOz@OD2qXynblKT*(by|@`d;9?Y`Hn`S3Sg?_yVhiHnh~ly4Zgb#K~X{= zg`{e!W`;H8C|)S506n-Bb9km?d5I-J(mXeXOx^2V0%2? zL`%CvA1{Zik=4ARaT|1cGKyDundkue8>QR~LTkZAgYias*&~m2Mj5UoFc&9-x_Nbg z4Zqf55{_Qxh#50)qmXR1{U~nX;kK;Q6o8K7R~?LdQE*Mny<^jLc&m+rhX>S{IFFwT zGns}Vn!r7|yWifT7V1s2Zxi04^nCsuYQnxPQ1IsVTk{)M4;pM&WdDs=ZnFF0u=sib zMIV_;l+1`F+Mq&4%&VooIf6&{o?}A4>)ig=FO%_DORBulMig?1RXevw>p7wTa7ADg zqL?;re~@^F_DIhVzGc9ps@Sy=jv#goU?m8ePF<*$#Q82pMDq0Dg z)G9;P>*}S9$oePbt(tayZ-#sQbSuF}`abm**05Sq0}!A_t+^^EZ-aycN{hp_3~Ggw z&f*{OyTF&D?>!9~8+GGiZYl^Al5WdUBT8EcUGp)t)0BvKrmU5abikrgpwTKJ^cQx%3jEG8KB=G0-JyC4`bizW5Pk`v zmP&NfjDB73jzRY+P$BhO6kKaGn2#kBb+*D0qoVIaxV9QOXE(Zj7lyYzg&G+?sJz?k;rtyp!?+4Kj^(KN_ZW2>4FSUXW)5V@W~qU7O6`(6|5gh z81YA9y}YBhQdMDvhiq8kj5o%yQz#uA3?&-^gqzsIcCX&6 zwq|cG*sG<8I*=%44?QQG6h|kxL*=w9#@HH5Vt!+*Uzie5&B=D{Bo*z~623?`EUpq< z#BQ`Mlm`%XaG1S*L`o5F*%HRNThv&E;1p0nMH7w%lBmJ}cd=r$o3CWNRL5d}K|S_H z?yf@=>lf0R$31{OmPdyo7h>P&PTjWCtzgHzEj*mvs%|;^YF*}W_iTFegvF;(R9o+arWHWVe3rs}D*o1}XStMu0Bv6J;4%Ao)moH0Z0 zExpwAF>xD^n6V2fOGc182%?c$=_0S_H=9tJ#3kd3;$FMJc4F;jSbF1U`^5~Kw^e8& z(I8@e^a#K0?$-TDvHKm*-KgNPn7eIh;JmF~hkhR~xFV=D^?d$KQCyKM1Ttdg4Ubqh zR0~}Uu*dhz;DH%HDwG>6KiSn|DwW}ZRDp7070ZPTLc6Fj3=J-!MoakA8t|Ol&JCE4 zW-t_BJYw!FM^?1K3%TgFcLSQ(E%0|mXJJ(7jHy;GmbP;m_16hr`jyX&t?x@JK{@gT z&z83eAWAVT4Pq%t1$2SY?vXocl*ld@O6hLdT^=oWUeNx^(QJKo^64kP>#dya?(6vo zDAt6|_v}@8`ST?Fe6w0VdlDkg)6Gl-pjNltI(=Kernl)cl3TR>FvY?#dhTp!1Jz}Yq5iJ?r&AOWp_W6PFaOS_mJJ|anw!dF<}v}r%5XWj zyyL!d>Aq{|0yve{8QY7O4~~}m2YXj_mq0~8u`;yA{=4V`>Xv{)v2xS-&R%?wvNcBy zb|iXf=k7)QK82U$q8piGOI&_b$kk)sA&}|C06tMM@*X?qLTqvmjD>KPsIwDNA$juPiUUy!Q>0d-qr4o8~sTcYpN2;=~t)dU%Oj$i@hw)y^9Gc$i0ilG|f{)1xy7)>fCz^7Uo?oD>B^$Srj+- z$q26TU%%YDzof)P*ALx|l?htAtSf2m-5)&m-HRC}%e@CvR{QTw6bSbPvX(%cdk;vB zF8;;bEl~F++e^DP7#v1&8}6uFktbgpr*Vr4E*ly$UvyD--_vFw=QP5`ght0Z2g{eB z|Hn++ACO^iy8n|by#^=65~R$i@y0J?kSzD^7#iFnIJ2CYL1P#ku#6_Qc08FxtJBFA8ipEig({opN{g zogWs~7(BaYZVz2Aq~YE68-|WocWux6j#npNfBW44fqeshG!*tZ`0?#3=ZK%wMex{z z*~B9Ci_vH@W>s@hh@^Wa(2Gj;NkH}HJxM9SYt)@*?|@PQqgoUA!zuVVXSut1O08e$ zZe7sUNqrE1=$epcYeW`Tpc5_6Y~b&SDf~>}N55QS79Md(w;TfT3ylKsH!SGA7`a{ChNxwwscF~Ece;U$Su^|K_ z;Q;kKQqqua4N&{(m?s~?%DdnE~AD`nvLDwX>S!T_ze9$&r7gf*`s zonUpIYvq{aEN<-#NYc#>7^>R(^`>t(&1uv-ySiq9nSMxw{Ni> zEGI@pL)Rjh1|?zuSJ zAPIEAgo6gYk5je?MUaeMe_W_+_Wh|yfz0yeA_vW}&oV*ILv=-VWdmV?$+x?+ASrV0 zu1^KKR{atV$(BgSOxf*E43aJRePEgFUenW3!>yke3frQziB|p;>=cn!n3ZRQ+h3{7 z2Cj>(s8)jsC~q3~TBKgf%q7VnSHTMI)@`v8!kUx(s#X7=B5jZUiq_@MV#QISgzh zE&yQ`tGbrCP2j07jgz@ONJPa~15L5w066og@AtiTmBeYFz8~)ZEfpwOyUzSK(~40f z1K(>*@89B>Q9>qk$-YsNg2QFh_r9dX8n0fFy?q(f)7*E4>sXM3Bb19kepYZD6S>^p z#2qY1k%chJor|J`!$kV{a#s$5DgrNwR2HaQ3BjN-w@G@F{j@OXGEme{?m(eM!Qbd> z=zdr#uRIr_`U&fBLbA~ZACADHF+4MJoY2Tt=!bZJSsv()L`4j=R<-0pD2qx(qOGx( zzbVY$4^w$C)72990izS>X}Lte8yz61b-#rodtxs3-Cjvg=Dv?L$X_?A=!>}YZO`L z+wf(Qb%5d_QJHoZg$2C$C@3pZt-*4cEiXI!Y#bL*#i(%WiE`pAqW))As$$ZzY{!o? zzKRveu2jQ1%4JLrwl@gO<+XlLc*Pow$(hiMdKY6ZkW0Rdz8zu^kY!YWYhPqO&p7ib zI-WcMl^M^TUdQGzpOvi+(Di@s4x-SPme=>L?)W#h9)_@O-{=8#?r{ru{(!oFb{_?I zo>@;2BZ3D|F#@Z70UwiUSy#NSk( z`^bZhIHA5cRXeSTjOwY=D)`mx6g+U&PGf8!&EF-0N9^hYYgs(X^1U*|s8%h#`HJ>H zBLyxfT2+5DRqt%9>s^@aeB)6~Rs2fKt2^Kt?F;5)>|Z1QRla%&(f*()HhGSU7o`{O zOANoR4ObK(CEQ=36H)h!ToM~$+q|`cuLPg25tkv_mc#!;SQMAhR&cq%=rl5O>JKMs zyF{Z_R4rfeDk2bpURJP8^;3sjI6E03Fb>H3_a>`Mj80Ms9}vQpSF7L2LsceJ##Qsh zY??RF7#8EA37^YL^YzQrT*yD=X5sUgWB_+`HX)mPd!XOP7n!BLR&TxgzpO~ly8824 zBkDN9+yaWK=f9c{Qv`BJ>gOxCmWAyLW;Axyyp^wot+QNH?-L6dsYI#t9Y6ECzx$99 zM1uyR`DI9m*Fc>N|vIyj6&x13FuY zl6xo;v07S-lHdBEsn&H?e?C^`whATpbTFbPRvoBA899PqpFpM6Z%ngK95d@2{QQQR zF_Lgyk76@-h|DI(1ix zQfMK%S`_JtYh^2`MtQ&3vFnFT>QRUdi9}POLh=#+%$lgr$gh{-^sXXBdZh}#hnkdJ zDle6t+lvIv>dW~`uxMDOE(L4cD&EW4y8LJzVeL|W+dNGN6} zu{=eelzRgGBwl4ou7)V$wlDRh9woa}9ey9`Q2fdk-Mm=J+@1ulDy3YUr8-vHC>^TR zay9ckUu`jd4f&{Bk_RC#(JgSJpOD=*UEjgy_l}n7`}(1Pl7mcqo6S%@q#l^vUyz{; zXAf*OA9VqDGWW!;7gab#5MK%$EzWDl;>pd3&PSs(B+CO;;*7%1RVdz+9kbbC3ak#mPyPM6J!7A`@tJfeV#qz}aa%U|AT)v*Pmf>Ob;OuObWiYcB zm38R}1_N_caN=#8V$bzCbq54buZJJVT=cHAvD z8M**Sc-@K_90jK97|b4oD`1iO@rio6^t2gTV(UU5u#3AdU)0SwclHl<_b=`p4yI!C zYvLcO&~<-leN$}tEpOhK6Zv-Qu>A#XgC5R z%CV;M1rlz;M=o#N=SeK_Kap?jn%rFnJ`dg9SGdPMDh_Nqa~ChK-)HS)ob9nu!Lhu) z%Jw4UTplS^<`wj&hiaUMMlGW#981f7K9(8oOnY5Yrh2Qo*kE=kQ(VIkJCn1mJVE3(fCqZ-+ z@=^yEffE4IYPzup2>-+z2QdOCxrD&=Cjo@W-h1S@JZT}izbC`uw+-VtK=2PIk!DX`d9!WwF4i0Qye&tAjc)aD?Od`YR zV~GvA_Z~@W916x!;>gd5-h1Tlv!{(fF_^Ae4Ri&67@^4Cdo-C*{I`&+0!FmlA#iEH zgPjO{_#1@oW%17PaOVhWDK1|f!oJ@!fJol9eaeDJPH$|;yNwY`@XtN&Iyq6lO~p%z zkpm4iK=X^YUOqTH8b}K40a2ZURwTjQRosV;Sew29;qXhu2PrYLlTYklxJWKUCT9 zcff?Fj%?OQP#2#GI2Ho&F@X<8{C29KY&LK-UyGA~8+5ZqiloT)7je&RfzhNgdF&1K zj}$mBLR9n+Wo%f40C7Mj_M18|H!qWA+0c69pdywAK|1)S=gmJ2=cX=)z~Lv5g(Oc> z#OHbccn7a|LuPVMQHF)WS#rzn@@Tp9f(|wu&DLippMLVY-pXAOI)`pLDhcGZVNq$QQjFh__J42Zk=Ur?N^^!ZYxV7T1Pb zLCUi`@44@x&_)xN1&v{h!>1txS=SneT*1G95@hiNz_Aj3{c^$DZRup`zLT#W*puHF zmOgbPHvup*U{*8o-rRmiSaMFr=wOeONKXfnp8`Kz%(Z^0a;|}3EKl<6Q$I=}TOrza zVR`W4?41D_%yc1<4v@*JM{5#xEfi_8ydG` z<0b^q-zeo~5Lyd@ty!t*5Wr#7K6R9yBvH6H?b2%t1laKV2`1s_WsaCJ^EL{}#@9x1 z3y-a3rKUm0@yi;V{I}GLo|U2-K_(H=S>DtdUnV5 zH{Cp$-PD~oC*dM>V)K4`jasNT&Av@|jnaz=&Nl^bZof6ZVfA;xc8!D7>%pj8ka`{4 zyCbNHcdq)#RH9@?EYSuPGGbmWbv)~s)hVA+2?(XbW{U7FHz7V0K zV05!pv=TVWS{bQcS1)Bm);}Tln|1{^19?2cKqY-|hI{>VE5S$lKJ^yXuv$_B5THh_ zxhf}bgMtNyjo3C`bEk}(gZ6S2c$Iwo1UiBRy z-B6okuuclFv8%=&9LUkqts+g=I1eP`g{=O7!roxDu=C<7xl=_J?D!J0sIIsXRlqD4 z#Z5y%^5FMRQN1hE2|FAj(M~nX%N$eIN=Pt=s1#_lN(lXhov-|jPHr5Wf9hv*cc`9% zeo}`ogkM6ar4rpVqhAkp2Vnu!Z&7fq(O^E7Ow`#5M~sTT58>Ks33mxo78A6 zHtfYKW`GQYCsCt;d$HzCk=@?vXL488USm8IARIR~T8b|(AH8QB4v_k2?)KR)SKpQU z+@(|`GMxzfPUUq160Wb~E?tlT>I^)u3qDz6-Xe7gr-CyJN*M7+V!gbhw^C$Sfr}hG zWFrS>)altG8#xH)l>7O*UCbsYjBwx}ybxGrHb#O30!I=U2NX7wScm$#DFIZ5izx~- z6fk8zN*j0=95_b?7DloVjSuR49UD6h^&v*yDc~AUx8088h}6MMp>%LClpzF2?p46n zdD!mNd)3zL%>{e46j28f#q6Qygp=avk!5Ig|zB7NqM-MEINYvX$oVBTn~d{!@40j)0~QBF(>|rdJ|4BRf1TPl^2fJ zKu>yf`O4d_E_e6$4vtFPkRO|>!C;xM!3Qi@ zAB38VXRrYd9D9ygTL%(Te8f9?L>_tZvb=CY)?(IhtHf0PUWLNw_4mcB6~`3uYK9Rb z%vVAZq-Tn`2GPYXx?AeOm2YXoy%H4zOI565CNk66s^V{rdX`Hm2+-y?4uJ4rJp@m` zd9gf#r?(n4n96U*5ZVebHy3{fH36CqwiSsF;E`mN#`TFe^%Uhi7G#MksAp=#OR{l=7iAIyex%nr?pp=*zfaM^Ql8@YDL?iQnI8d#ir_zCPhd8h*@nGnqCucjU&0GuGQ>uOkmVe1 zV}tVa*-fg8U>+^7yr$I8ze5Ja%4DVj>XEnYo$%Y$TW0r`-U;=@{*92uxmZ_F`iVIr zl>GGQfkp7Q8=oygEJF)Y6T6Tm89$B3!#Ema8&Jxl`FLGnH+>7v;ai>!`U5;I*0+P@ zWVF{yIcaxbla0Yr_y3fgk5N5zvG;2{XSC3b;^#^Wq>20qwU+I;Qo6*5_Yfadx>O79 z;V!Vc5uEIn_@2FrcmUBBcJTSDx2{`^Z{Im2uc-o?KNi$J$nEU zz+OktBn5ka{7RF3n>gm--GFn=5Zj_YF)h$aBeJL(v5Y%t^*YVW;N&&2AJ(5A&u)Cb zb4*qL7O$9Nh>CpqG7h>uJ)T;YuK!SS?zQpatL!6pSt@zkY#FPcBVsiuxI_~VUT&m& z|9wk*hc^4qynJSCaZj_okK%8m^|j`LXkr8TP@d1v`0v&6c7Hj(>Ek_kaX#N4KCnJX z^@1+M9f}vQ2YC>LZcZfa#f$U#asop?J|)q(=*LK+IG-;kG3(2(WQ+J}TkXv^W4!a{ z#rb?WgRMs|%je4njwC0{=gY@G#(ciND0x2L|B5wKCHZ{upbHSJFpg3%rf$(<&#g8J z&K#7b`Fx*{;FjNJrO8bn_s~m?N18B{=ktAvV=rHpsPIP+=JWkeSjBF1yy92>tln^r zh!W@X17ah;pq#_pRk)7~x7}Xye11o$xxgN8KHp6o?i|+l5$E&W=z&E@_HtXVy*~2k z)~m$%eD_Uk17xm_`(BvOcOL}%zw`g`?aE1$+o?@BDbDA6ghhM3l#}k5tPT#l#jq78 zN}SK{25b`vzyNPj|GS&f&*#Hw=UQ=ueK$DcNX7}2IG^9GC5Z-<81bTfz7wB`5<7*! z{rx=k>Vwii*0X!2a=&ZEG`PTCo|kjQPj0?pK#{y_d){~EI{EtB?*}dv!wK<{v) zWfS300)IINiqFxZ;Qrg6*#!!nSI`VndVrC%g8-3T72 z;QYQDczgqX2ECQ@8}1k~e%tjp6)z4?&SjxhJ5@)zV|k^D!}E5mpo!bUnZ7E>M{!JgAfaCldcceGYx0L zCVtr8y>zVgmvPiAbjI#AONlh*7_tmHw*S&t%u)3((ASpf1dm_4vOL&*Z2zUh#qPd6 zbzVE1-nVYwV*6@NjGTrZUEDD#7TFsXtfmUoWT|3bPhmuyqgoPbI|$)1JO;zW?#n2)W|Fl?eLTi?O0gDviAz_ zb0AVU1h|cQ?ygVCnD%iT@hhdw%-L;KF_(o`o}QcFOMR{e5e+o8@Ogqft zrAzNGUS@y%m(tt&`O7aH0k-VyRsA1O9KKXw8pVL$t=?)Z0qPYLCoLU*w@2Lbai7x@ zLqj^t6jAJ24HKW2J&oqPDSac=H&gxNuX0WOptMY9SWO6@xoCt5J$e~!(4nD>{q4b- za%p*e@9GYx6ahju3llB#MF&0`sr#~;6eu|yBQ}-iKhqtwe@#I@dRckau%1Zn_GlS3 zibj$iH-HaNOjMMs4@}3=3YM=s58CO*II4m91(u~=$@X{E32e?C$77i~WuIXObQ90`!0n>?iKdcm`h*>bUfkNeo+(FO2vC+dHZ z2KNU-%+Gs5L)`fEI z=FyO)r7_*DZ?u^1R((u&t2n0HQP1XuY|JyxS~#}<-_nN&T^JX_)=^YfreVSM7oLbN zvb$A8B^ua0;YhB~v@c~mi8n&R)W?)?3krL1Set`RSr^p6wIV z{i}bqmVcsCzGa5M*2*O~U)5G{X1A5rGX}e_68{eZtbO>aZ0*;B z^M4~KlR{V&)Z@0&lccV9Wi|SM?|^~t2QfH=>bOETn;S>Qm_O6nwzCzutz?522IqG+ zgy^sa%2`fF6}&Z^{(5Yue>YKYEs}42&3?_hm_Q%()cy@KgTnWJ3W9aX_<@~hT&CZK z0kP3h|FZy(GodrK4B5OvGQzxt-c+>xVXa3K7Hj;*Y|xilyyeb!>$N+(@gCJKB;+m)i3vqQWB@2o2 ztqQUe-zh!#ty0JG!p^9B}U@Z$88?7XgyUoUr97or= zj2*(Sd>~}6eqcJ5Vw;3*sUKr$2Qk>QOYnHgy_%hFK1a&=M zED@=@DH|E0i9tk)&@j}@-FQU(kAjqPI(OH$r_@?>wJzdQ*xD~itM-2(_y78FlX{mj z&IQI!gQkfu3gP^&;zjJhh66y&OExQ*O3-aff^79?f}*o%T&C`2+`i5~4~SM`N~DHT zD1qhq{qk*K4X$__1l`~()y2szZ&rT2z**>Z%By_IRw_j-_l2M5xDMD4buC<_jBXbz zW^hl0SM^f5GvL!=T?;E;@?!Zd!GngLxAsjy6${YZbFT9tJnSmiH(!{jw+T;d*P-$G zy`yE25)?Otc+w_3A65^}&Xy!R!|4O-&HCQz{}PnS^x_KV2!c|<8Jz<!L7)pZXoJQ-wT)9y3X6D0{m+Tq?4)fcsk$6NsG<%@VNSNshtOU5M(i7du z)JzhXi4^FmB{})sUqJD{pNgHeh~n5#R5Gr@Qs7F(8iR(~gjxgEOerWBl-|Q)CP%iS zlF2RAk24b1D5?KyqTX42r5IX@>slfR>F>ULQTH<4*+1CbzqofeSdG!|p8r*i&iiZY zyNYKYZo2#5yfrg6bB*dZ!A!hoZ|}+uJgq)7uk`gh>-}ALzVE%WR=+w?k5|j-yw_xY zcFO+3gMLqX>eGn0GmprFH|Ql`oFTwbRkb91g(rq65OkOOFj zy8wUFJ=zak79vW7gcX1H;~)Aq^|urC@VIr9g>q>Hn2+X%+fzBqseW!8`=330@71H( zP3C{|k^IZ*iX}^@TgFUg17Tx0-O)Veu4C!R2cL9j@G*<^0!$?*`A8<05cpT)3Eb-5 zjg32wM`6V522SUYUVEAe@ts3y6-cW~6xLgszu^7~vu0|T@> zXBrF)sMqwN^+^&77oT;u22)eeJIE)!FsKJw$QOn*$Ih%S74Z!`;>9)kx^U+`_X$CB z0lk?(g+oZPemH>{gMR`A(DEsOYuWt3hUKbGa?8iI)tUry4YK?}?D9!~M|OF6O4Q2~ zmLIYh7#cvk8L@9}ze9-s2-~GdhB^@a6x80zIoB_?K=dG<63%_N&Buo?CcfzcW_QD&HUn>!&nYWAT$c5Q2b}qkLXjWUsXv%*?4K}&s>l+<5W)XMz6(V@4Z;I#jtEu#h6|_fTkxRjsaE2W zKyNjZ1x?bYrW9OUv5cb3LJG6hRMm%Vwl-d4Z4$>3+`^&iIA|_NNq`idU499*O^aDDQ5!{&7+5h`~Uo4w%+}JF6Yg>kz$VI z3LbkCiRb0Ry&X91-?M-4&rN;x5HHry>%doq(%>Gm99ZF9a9zjfSaa{wi-y_(&demU&rV;E>3r;p-99>^8py}UC zMY?806YW(gFXjY#zCiYD;CzhG@omdWAG+NU@tKs6XirKx#l8Y^+jRut3Sey|79t9b z6XuAB9-X1w(11~X2%L##XTAD$Bs{K>7>JtOU=7=_RUd_pYp?46AUc2guy>I%K(h6g z-R048=LIORUmnfYXD6S2^1I$DAkLl&2RXs@@7`NAI{UsH2mAR9`j0cxx9xNs-Z7

    A zmBBeKPgAm>Q(xC`No|UT0m^L>VjEnhB1|mrA*no*?TBwCq zRY%*gEn$hdTQ3%7Zs%*;?Z>+q9#O-n#0S{XBNrTUAZA%R9A2+Xj+Zn zJsS9w7A?5LsdUH(P-MK3+l`J*^CrG^HK5 zPb5e0A$RqTB_&676DAo!zd}1{bpz^dAMlN=6xp)Dc;_rUDq3N6-YPi_by-Uf_Z!RW zUD;B;)@}XCRPw}M-uj?#9I%I4UZ8F?L%OkiRtUYY_ZV`azk_?WxRG31G+1HbJO^8A zcVtH><7p=ApQh}gN9GVUiuO0)cuQn)wFB6_W-AKfou+K3!C2U{L8G!5?-`?MC$YYy zr!=JHg4HEo_Dxf^e_v>1!uF*BDW4sp3>>N>-je(E_X*z=W&6$qT4FVvjF0r`eis%ZqiNT zzoj${wUb(jMv(zKY{)YM=|O3tN_90Q|-)~8EsF8%uc_Heiq3 zcjfNVH1XW{`Py>C?DLh$`!r4JXkG|(D3A3jLE9a>C+`_IOcg|>132!Vm9;mfClU0y z_2Eh_vR>0=j4f@q8u#GJAQpH=?FFro`i*Ovg;Ppdz9du5?}g+|X?h%wCz-%n5qxFZGi zBJ?|9@*Do@kDEZlohgrN)y)aT5lgYr)x@GB|)HO*T`+!3vI;{yoX8!f6nDA)CH<+io;gOF`i+dWwNC((j2xz6o$iAmP>Tl z>I&>NL1ZT9LrG3d{i;mvVF`=x&5%Lp6oQPx0*=gUi^Bve(UB`u zdDC#U&cR<&=RHfu<>I^*x{SJvoKH}XA~m0&t|4#LX~UAz^9i!BQ{@v-^wCm3HEUCh z@}%bzIJbF_k`U`CGIlIIpHLHGlu}!!AkfRCg(u3kFv1gMxy0~9U4exsmT=5+fg5&*#q3+efUCz2IcQXvD9M4Z@c@qG@=+SO)Z1i*( zRYr-ng(HbxNp3$ipD9;B2t|D5JQur;@xn&p#jrMNLQTU%k-K+Z7Grnc;II9=_6cfV z+1yNKd0Rfah%m}#Sg>IWRp`v?M1#m=un~gf@Q3=eY?ZR@)E0g~IV0zrpa4Ah2(HV4 zv4PM7Nn>J^00zP5?3MYxN)2+BVJu5Y7XsfLMjEk@@!I4{pQ;of7Xtc~!>=mj$k9?B zaD?Z2^pJ!+o}?sTq74NMd&A0wS!eEinU$8A(P@(5M|v2P7H|kUj7A+X;1D0GN*4)Y zq_1#k0uDg~*ns5Bhp-n#Hs?~Jb#c;0!yXu|`^^hrnR84zM%*rT zni#js-*;6t{j7L(QdvrT{pffI2<5S~w#s8Mt62pg`W4k2B&F~B!UBN~SUJa1)6%=L z{vqBrB_uSPoQP4@xA})$y6oZ3_VE9a_by;|9p!;wpL^S~RFb^*1&qH#20TnMc$y@w z?$*m466BVQY_Mcot=Pr}cdm7>cKAMuFQ~2S zVJot4W0@4Lh9~5Bb+|q}Gwtmt|2w3{CO&ml^ic2^qlX`ki-zvnI@jycf93Inmz;G9 z0ePQ$O}7vm|CW;uqP|rRc7G zPJSR&ro^>r-|u7d4MsS(?|6Q;vBW_$8g5-7pXiXI(Po9l(3NuER&k?7lhP#oQeK zQinX^$euJ-ne3B*(-@pVoln=Lb==WqFVr>r81T*5Yh7=4Pp=`N;)6>DecJ|hN=1E^ zwsmb6xpw)Qaw&&3-2Nr^BT+b-Xw47Tz##7PtR2#VxJpvi!f;yVXT&M0^;C?x5;8}mv*W>wNFUmFHOdmmbJU@(u6cs}x@%K$KIoJ4N zQECWo*bOsUz)NC-6l_oM$t%~xmtF{qC{UP|ZOm*wm!3M*A(mdT1+qKeM)jh&!b+po z*a8XrJCoA!&<;xy=}=(Rg0<8HW}!?gmO+rEK1}O;laahLQNXKEkt#1wykCGVAgB-W zq_K9o@?)YK`jTIAe#s37sXtv%K0lixFjn|!)L@T`cRQ>Rv5;8u8TZSvSfe8hy=;4= zrZrV%kC$VyfHxd6K9c-HL!IE)Vnua$!DOO!tpYEs5x#gE6?ivqjEk1=x+r3R)YS!G+L zv2vQHR3R&y`jA}XHD>r41ukU$kwhVD9ZDIpf_v2I&`NdBX6MQd+BRpeG^s83WTC2n z%xG3x4OPX_2}9NTH>DDXg{ogj6xKv=ag?9qd zcEUUH`DEdp{G#plpVNMRK2a%Of+J&~_JlJ@pIm@!6i&5I4h{BN$1XmSC^N<8FoRxN)Yz0>=tviS+pCskrxY^LyXF)Vjl4~XYsZMS z1#%%8bA-&E9t|Gvb@n|qjAVzrwnH9FZ5CcMT3??p_l{K4{km0n%IU&~x-s(_=7@z6 zqPn);mM6+vadAqutx5vnHd?2z|#Nkm!Dtfptzx3{<7DLR?!8y&{IIrI3M%=V7ET>GrY5x+p_TvS{Fsg}(x>1rvC@Bk zomsdD?J--?LXMJ0k&U%e?rUQ|pOJ^-!zUnVq-J8txLpMg!>k5k(Inu6Y88`{Yc2z0X(q%w~XY+ zLYsH1)2$1~tCnJma*LoXjo%We51^+s$l)Wgt?EMF-q)#ka6_1Ri0a?9DXMgip}b4l zGRN|5cA1p3Y7D^bgqc!C6&>q^4aY4jGQu(*@ip1M|Z1CX3M3mf!N>-H2}EQE=;Y~;&K zNdYI$@6O1*@mjGl>~FR>nfBLF_v1L#6d;wUwAyqDR4UPX2c5Zr7+rgiJkYx@(de22 zBpX{Juj-P!Q}Rblqn!*5Xc>ftNwkQoV`c_%A?|Jur7g{|jJz%}7wXezsTu5itHz_A zsK}xBMBp}QL?^|Jrh;EN*2^b~pdqqtK~AkR*zaY3BTG@3l~nMYGP?>?0Bka8hXT&& zvZoLQzSd3ayK7Tr3CjLP=}K>$QWUJ3pz4!k+rxKTznRuoBQ`4OPF9~F3yg)b?eWm$ z-QAcajb;K=lVb@6z@-RPc}UEHBr~w(yq^q=^og;nEu|Uxwb>N2Jd&8T(Xi&ESkly` zFn@W{V2n+OC55q^4oezqGc#ZbewE3v?GY|974{EF8sk!uye7i72S*=6OWu*>+KYz2 zoCHfSswKZ()6*0~OY+l#ankZ_51VY>m|hi2Ft;F)G)#WAU^$ZdIc?(W>T~7OiIYRE z0}v5`$#*$eldReU%!iZO(?#c!Nv>6mK``?1sH^^Z1f$G&B^5(X!dBFgP2~fVTOQr1 zw$d$f9`NVoK^Ti{GsLETF%OL4j@$ za=qqeWdcLrcNHgY`@tI`i7O$KBaV)ZMRVY;Bl0T&8#lOEF;*AQ7@N}2@M0piW(oZg z(2x|OQ&%za<&2y}6GoN(%9%rtwI(xN5&`@huR91`^NEL~KvRF&Q@nkO|DqDG7uJtw zJBK_yZvj!OPsILMh@t?5qKd+wiTySQf|KF;*?F#CO9|AW`j(U$i!Ui2jx{`{n;BlB zCGW_{inqel(K5`S!8P33xo_bMgT84cuNDT)KKJ^YXJ%%u(0}&Oe~!?9YC{?1Nlbuc zg07;Q+X(>&BkYtr!C(=pDw+bz1x`4ANJ6Hju4up`qzF6 zFs8KeHo34h8Llmr{JV4W(r`SzXE@%tIPJ4_XT2M=U*-FKNF!`{Q@g}U<|3RmBr_QS%1nB zt-ULKx(lTIpYB^k7$^q^@_C%3EY97c$@-1t>C3%B)&)HnQJ7`u559Me?i#wKFca(# zm+2qGn!JoqbMK!C6sVY}Lk)d{KCU`V@hLUHT1ZDyt|}QybU391*%+l}*!}0*aOS>F zpu%kN={?w@Gi!dO`ln{mZ8m|tGb2Z%fuqc;Ry4W6J3>M92k&heVmjC$T$AZ|G#C!n zChK&YB-=4}<~#@0JO}9wTSnfC>3#UYCJFf*?zwMVdAeBjNB;2})qe^R8%!@>7#5}q zjsY?jZTNQ5_R<~ zCjU(~_cXG1xV}ERFj4<+^p~5%SJeM*T1j48NNf)18f5>DlZ&|=x;6~LldS-n&Ln(z zmb@gHNuUI*8-*0!E}_bCMkAKD`O8ilA6=x|0nDmxRAi&RYpNY@aM(LGQHBuJOG@Xm z@|2>^!`85l5-kY7Fh=i0Qzb#N6sKWEULvv{+gkyg#u<5{6Bjybb&^=>CkS{DL=p=b zYL~6ahdPnjj_fETNh{-Qpiz=&E5W;>;3rBy+99{cs9q&X)Y*wQx8j1|;v#bjS1Ht` z-ZbzgAZyPd<)Bs*C<|_)&mDylFHr6Cskd+_>N7Cjl^61tm_d>vB;Y4*ur`{&o61C& zqMs+h+^f>0z!6|hqUx*F%&AsPO;BH5MATQQC-tv+7rYbYR6aH6r8_XbAH_ql3waWF)XN@g_uUjzCJbt6P9w-7 zZFBCL6{!(}4@-Ib5RQ39K8)^QXqbma_u7Y;GxDEguTg+`Hdul^$ECc3LK?6t(g0dT$w4xy^%Qu2G?gk zf%ZR~0##gF>jc89vzJ~2e}$k_ z=Be~i)LE6GLaek^^vy9AGQ_sgMj>+aqJ}S3Eq!Ye?-M?Yw=;Vy*VA_9%rcSjT0IU2 zh7Z~JT?{uUbfwLW`3PkxHhFTBNV3YasFXDWolJcZiqKasV{jC^V^fK33^0^yl05!i z`YRbZ5~sO5Dv1G}QWB@L-Uteskl)Y9T{ujO7q$4@c@<4VlebCyf&iATp58s^Py1_U zx5gX&X>Uhw_T(M+ADy#2D7}NcB_l7uAt+v%sCmoOB>{TR3n?3QxMV2+hK#*i&k7an zpSqER=RIc0k9V*lAs{nGf>2=~iY<90@!|>rKhYt}p+d#|NVI{;k&0{Q5~9-T*OGWO z+KmVRwSX{!*m;zgN!wFvQuz~$P*DknKz6d?g~k7dMG=V#UKmwcP42!~tElzJ^~=}E zx#waI)Fw0J^&PCc72xqyLTOw@kvrP104bf_HD~1nF{2b4hebwCrAMx2KABmXFMk#{ zACkCX=he(YZc13x>mh<5mfH`-guLPQ!;F6WA!GgtnIJ+0qCS*g4l&(3U6`jg+b__| z?RWJr5635I&@nIX+cODVz}AttwANRfNF!q*5$b{nKImbUph9$74SK#09rV=Ul0i>+ z{{tW12pEjEFj06T4k;LbHAD5ky|%zbDr}JFiy1b&5$Dq%%A!SRe8!6vu9rs&YeP?) zzwGafIaTOPGdh417>YswGAI!rpg&+kH}Q9k64*@VXVi3F;OtK(f)CvUfU%*Q_z)#C zHguDJmn|%oh=+!5UqYvj0s{6!H8(Nm)0Ze6HdgDKQIJ221#!q5(ebH3lCp>)8N;c9 zPB}6*6#y8uL5)p;TA)>72;HTR47UY9+`orMqvBHqaj^m+|8}{uTd=uanFbpSKZFj% z4N+&~8Wz$K2o$&@@EREs(=OEYRaef6yHk$Is^2WbQ-}OA=B|F>qXisge5uX;m`1ByYoh8OE(fcsJj1pWewMr!om@4*7k^a8 zs<7swxXxLUqcZ$R3n-ufXzIsUaIh+Bgs$q3d+`~0MR@@+Lif|?EKw4*LP7jO<#hsc zfNcQ@IQUeiM%!3^zZvcFxTn)|MrwY2?iF_C?EPm#`~OH<*I=stAm^EpMHNZ(0!8u1 z9I5Ozpo-j1GOHe#T9W^=K0?C#|CL17l6yP-eftF+65ju>pwd`{Ebj>3aQzU!2<`u? zS2){5B$WhV($1=mxF5qZ+EP5IVpo3ddiEs~f}Lj*h62&pu-FJfk{r0VuA zO`f%Ju?oe0)wK-Q+C17|<-sZeRlg}1jx3Wj9O6gqL;Tr-MD^l#C5CSSn z$Z8i(ssR0y9r6H6V!%06Kj@6rR;c@xj7fPy!p05#(AReCX4a34w6I&&%nh2-8k?*$ za@KOVP)^n}OTh~*+e<`L$MP;u=GZz%);Yu1y5eC^0Ax?aARiC5rN*>rVw=zrLg42a zQ*czg2A#x`FsfpZpmk#Vbe$hJMGWp+hVuI?rA<<^C`NR!1t~i5A7K($6r|p?aQi(r z(CX`zJA~c}j$mRFy7YeeGaVA4&6Mm!EUH;rNvksT`}De zji49#L{tQ<2ud;BQDy-NDVo15JUmo@GkOymAi9l_R|v_LN88)o2L8*rTLn;ITxE2X zhMEO?V#G`Vvfuauc`z_qAN>Kbd)(UPu!RZy_$QucoS`v+}pd` z{U(}ILR8GyIqWU84x!6{BQ0{#3$7K9;35q!SfwM|k=&H9$xs!(=atQQmAyz@@rWrx z0hiIJv`{<}2Nq0G3L{=~07r|wUk5OXe5fw85lu)L6u1HeQ zmV7deMW&L#14{Hoo>P;@##60racG|a((TC-BwwYak;T6htSP7xn|#66n}&q8il{Z` z62xWqBO?-Rp~Hq)Fa-(ANHC#eZfJ^?fV?b6ejqAhE(E;v!o`orX)CC=OZ;e@C^7gA z8G@G@lY%G}*Zd#gjtt?%g^5-|xPKcJ!qs7uAzb)Px46(vz+-feErf33!06De{-xN; z;zy!F9|;TZ`$Z&LCRL6Jfdp_y9t$IFv-6JHT3iSu;7b_-iH}BxK=M1brlBhqBEJ+> zxRpT2*deumMU-y^ct!z>zN$rMG$Vos0bJVPL3}nkc#vNTyUs=5X0pEyU~u!i(t4>s z9`!fR3=Q)#aLEI=IPPl!{t^Ih^vWUc9`^V&&#skgi%nr!XYOH9U^-1s_3jBOFaakQ zx?j8V09-mGU+p+vD)QzAm5W~Yw!t93!!}tZ2cVju>uc8Vg_7nsD>qXM-4Lab=Mnd-hx5j!aP?zG&i0~?5|rZH<7GO7+og4S<+danbR%Z%(hY_ zIo5;W`n1on_OYo360|N9;I93eRY} zbM#rU5qoQJO+0~lH67=BW*txedEc*(qEC%>nX8z}`IV8M3uwKYCHC}aeb}32|4tR&aDt)h zv}ORgO0k&cG;N4ZM;F#y z0)-^Dsrttm`R>G|o}B5g_s3O4o1c6*^z-%32Y)opWXNWUZE<|t5p2llE~-8VHX=LPop2<&BP3IFWCQV-f`Npr!4y2b&sC z?en$VCAA2#agfx!D@sz5s`U$3PNPVEH@#jLLkVpP>egHu6a+%m3foj%aXfou>t@+HF^i>Qq;Vqp8WoD*D7LY$kLf!hTOd57MmXdjKeU;56@pATDi*&RxlKg2FJ^>~r zxQHoa@dKKTX%k7$f=foe8Zi+hF*eM^=zaB>R@qv^80eNI#kMvJ0~}~<&utgq$w#^= zJYsDICb-;dl$BJ5BfJ>W1go5TX_G^?>L($zjWM7P8|>>6T7N@NPV27AZ{DR%e7*GY zZv$>{n31jKT!M!{r9f0x9KU723|KjJWb;1K{pqtWEpmvwVK#hfL6WQ1WpB>4BnZsQ znU(|r3!7(208cr~k|4TdDk@puBBR+M#hL|+5O&Caj%YJVY0&0Z^l%`%`r*LWBSz3Q zT;<$K@Q8{0W=)&I2QcIZ1w*z+)GS_(U9$-8?n;DsQ(c);Nv72aya%68X;YsqL>62o zA4nqOT5T5{o5_LjH$A#! zqKW&ED4*U-?(UrkDxcPkoFo+wy}INBt|u!+wQMlnx;(_p>`=lQU8 zs{6Cf*OVRb$QoiTi@r(Q1G$6DE_dkcp1z8LJ;v?ntGSRcfh%ZYZ0td!QBj0u2+;{%*3?BvR_k8lO9SM-`a`xca4(g zVa#nK4_rnv^7lbjtxyRq4i2Rbn`H=)8ELBoI*kczw^66!9(pq6DL3h+=f9;i4L2u4 z{!Cg3o>Dg?o^H&XPbEtL73W&gRBq>pB98BtPCt=XrzoHh*(USur;=fcRH9D#Ee0h* zDN~HJq_XkNoDQH1`E2HR-b=v&=uJwDJTUa!s{OS?)i|SS(F1o*R#^EQcFC zK?(AjU_4DPb?6C!db6iH2c{B)t~<8Byk|t1DToROrFivLmaSN6$+z}2-V-5o&#h~& z6p-}_E@Ny7Wi|GPE4@hIcxun<8}u^AqL%dCOrE{Gx3_!eH~4?D#vi#*b(!+mUqyC7 zK?A?C@(@UnC1GdiWTdx|U^f^MuL9k~UcidOQ)E4HpP}S(4$uckDRdFfbCc_>~Fg2FJ)KOTy&sy`TSLc2e4xE7&% zIgK9op@L}6@%;d$P;(?pWK}#1dXrUC$Pc@_KXw2OIa3_f%FHQ=Bj(~A>doeh6a!Cx>%DoR__R&4ZMHSVxhu6XErQ3J{}|N>V@`*%qe+ zh9?SIDZ&%wbAjQB`h_AqG1m>CSdLvP3QY2(G6dBV7HB;oiT)rwDw`*b!?uV7qACh1 z*+MG{HnpIEm`(1b*0{ZIctxQCQ22b=g-t;I5V2{hiA(_T6@^y?*Q4bC1aB;a*At!y zR$K{1`hbMKzIqS)zGX+d!Lc#!6i^urJrs@vdL@Z|XgyPYL5fjW9f7tsTC-^Y9UAr zeOLezty1-z6{HRiQdm19=1i&pJoxbL%TZ%PLytrng9$$9L+~-UDxpWrX9|*qlnW`{ zEJm7!9nOHlQz?EZxYMU91(gdi`V~X3g6v5=g_}NHE$UiC-W;SP)I=K^7^x;1!X`cH zFv;)(oaCUBL)Vlq9_(D%LEl?qbK>eaIeYFsAm9*i7>#Xo9DXpQJil8tZ+l#@B9@p#Af92jo2-P^6bAZth!>PDoq=-J<5+g-S z;;SW>F{GlOi9(uUkBQZ3Z!-p35vxa2E+kqPoHiPE58iv#PS}&lF{V64+%7N;jN9ek zFUp3sDI*U$2|Mt*pth7a{rGrDCzQw1+CqlMWMJ##l^NTB6~mSamfp414|v-YP|(yV zfua68+>cz^?BUktaI&^OIzRNjLf9N>wa%8jF%@*R>((Yoxbi|ZECk&&dk;Q_M}l6} zrXJU=b_DI6j|TXCtGsQ^wbSqrl`{)r4}#prGAUdQPss7=aD8}Y+S^h7cSz} zQ1BSzG(Q{{4c)bMuGgji%Hs)-f%<*!HTgEFd*yJvRhf@TG2PIgtWD|ViY>k--|`~D z49bvrJRjl-sMC@(d+eNKKLRi?zXY2WW-^9@Egx9@m2KmwucZ>t*gRx0Pa(D#9)}HX z2p-23ip6J`s4RsPip3XOF<_*`&#Sbm`z3=}b2Jrx6p zMgFWqw8bQg)s~z9kuSJi){9TjdPAc2NL2J!aY|Z^Rk7sePIA0=yU&svgCVM8%D&F_ z3U=KJO5Puo&Qx`)s&^u9W7n@{>%5eB<8NEvM7|P~mg|>d#WiCczU_E`t>ca^d!epNC;{J$oeY1oyP1rHiVrRs^lcm1IcIsiXou%C z#xJ=aiNeuDYks);_qflqc1R22DouAiJhxGpyxqO-0gW89;0#BDL=DZAoNIisC~?7!+E_yi zcu8!Kg6#=DdF6We(hFe`tqRi;o7o^#`*Z23Lmgu26Zr09!y%ALL16?R4dr zq;RS+d&w_3zvPC4)SoUWpPx+;7%TiVYH-5P_%rU8W3fg@7<$?ENKI?1${sJrVgct4 zS=5b@Atm`kL!ESgRVPhG0$DnwByc?_izR{Joz&BFq80LT@=))kKCKY%Q&n8piRH6P zZp-kVM5R8aXiQ;!XVP#1`@0g>RV9=%T#qkFrYiLbIvUMYtEE11biz_!{hP6+zTZq# zeGpKc-~qZ;nAx8%W1arF{>Db1z6ab#)!ILssC+4K5@B_ca)bCRB~_lJ`YFC>lyJWS zr#0{!YCs5hpp>#x-_SEFD@l-FN?w1bM z?e0>hu_dKt7@U@c1u?4V3jq?tHtyS}nB}`*gouhIiZiJ>Ka2fd~3nr!BX%$o( z;e_y#_a(|73N_vfH;>VadiBbd8lB}|=f(}kQMN);VZ0bXDEp?6gM-qjs zbtq-X>bPTT*|VN1R27gJ%}T4GsyI4fs9OJ~>>h{iJTpXj_8HVv*OjRB5&(4c z${~+H0O^|3A2jdGJt|6y50jVm9*it00mC%R;ng-({+hGlwmw(?QSC*;8idgd0Puy+1MIX@1396 z8su#-{l5*~yrDzxPpvUkmqpr*YU|G4mEB^lHp{kAI8_^A9pL@`4tXdwHOBqTbHn+` zv_GD5is%+|=4w#`FzTc|N(_eU(>_dzt0or^C9cd8*L7O7Pbp-gcg-m#8hM)%*Nzcu z3*xPf7O)p;)ND>5!MEmar53$446% zH{>p-e*)(51Ksq&bVY~Wwbc)KySkjW$H=Q8zBZ|SPnSfL1%&PA+wl=n@^<&$D|CKXTPy{o0;bhff~~s&tN_yi3|L$MS7

    Dc*4o%UazxLdsmglu6bZnTaJsuZD zMJd$STWf#Yro|{mb+$*yFSO>Jjd3EKyRwW(`(C8}O4PC?o*)6{x{`2p8ofm+Ebe5s zr>@o903_$m!UlcRx;+IK3t=KI8~HL*QoxDxyEAfcyjE-s`nZq^W4(N$=tjQ9u`z9S z6>Nfq=;Rw&io&dYQ_-`$N_(r6|?H96LXM~PXGWCpgJ z_md7+pBTH^5}T1pAAfHa09BLhahzLx+%fXst)gE9zoYbB!a^j%pygnXv z)nAWblo_w2V#rC@iaL5Usw}fzZ%tfXt<80L%x{5ZMfzZ1QXXhCT}eZfVxB9hHEcW~ z=p5=IAM5HbTpg$uHzwQPBbObvIP1gJr49Z|TQd|T^oPhxd-q3{(CemCl8K$?_h8lz z@-g?*Oj=!ZFy6XA-&dM1rW(oyhnT~;L6+s|?e3FnhYsY6el>Oe)ua8%)Wg2;(GK$Q zjNBm)#hq2hfRAVxa9BWp7lQ)b9^`t>&B_FZzV9kd-1dVvL=sm*CPy5@cNzUkz{U+O zR*cmJG{&YhG`yIIEqRoLszi(%Mx+n1x!%@x3!2dNDB?(Sl^?(zbBl0EE>-y@#%Wt@1a)iQ;6+Oe6SD9MxRBpow$_Q&HzLoY?*F30>%qy7o{ zboX$+dag2#3R3?h72E;e&vsIHD!D-_o{?0QLiTz4DO2>0sJC{=icrA445VKOFSNH?4gCa7GKWDNi<

    93qQ^jK>$(c?4Huz!Q{f$WVqJvb#5`e3@eCBek?>$076kk z;m^c=n*+hgaQ*B&*RQ1n>QH^l=7o#5WDQT}+TW6WM@Ck>6{e1sVFvA6_&moqt>o4A zpxNhMe>1%~bcO!2hyHVf{-a*-u)TdbOP<68SSIKyy2;IVcp(R*!TG?o^s*+vtV6=j zBiF5!FAc9v*uc!B_YCs1MRs2l9EG zq%1O9&>wv78r|JPO?mZ)%k&Q!7m;l47EK*)Bu`)N73v^J#^}b(y?-W9pkks9HS`Vo zxau^;r_=yzAstCk62nQEsid3{q-6J>Z^N1UI)Mta#i#dbb0l#4ZgviK3(O^4ss17F z%*fGb;3)H|6;19_kI<9-ozdGe#B{JhxF*x_XfPbCP1fl)Nw#C|%y|x~c@EMWwv4Bk!Lj zcgDzHjlsP=t0liz8N2g%twRjU_)!iqEaUemXyCh#b$OXxc`QHBuJOG@Xm@|2=3 zs1edyBD5gscDm7jJca^N`w;RXqS2p1H@i$6WH_f_ zWI;Z}^r2ZsjmkuG9AMtZQ+eKmj)62v8M2yKfumHX6?6ZQD>KEoH>&5z5WR%(Jc0H< zoB~x`TMKgj8E^bNeWh15emc+HIdeUYY*4S_(YZDxRU1%gHZEE08#?cylFK%}Q%rX3 z04yQ-lZ<>9E-A0fQ|Y6qvnn&MSZS;1n`10wh;5_o73Ane4PUBS`qm`bCu~mF%#zOX z3|${rbuW7>*Ha;$Sth)3`HgxLk&WNQaDzfu+T56rP?lnoCpU?t_<0sbvZe*jYHu;o zS1x046uV zqG@RIHi=)L8>e><`qTc}*{$(Lf7;v8n>~5Q{YPz^Bc(|uZ^_6Da0rT5CTiX?bxDBU z^Fqo-9WGf4fFWbs-&vu8{o`M(`0)-_Bm`u}NDwLvM6o50Bp$a4VQ;+uM29Sg3KjPw z(FP_*Dz2SNO`kBM`UFlbOd0moXg4AN)B?f?V&_p}7Mj}%f)Oez!4Sw!R=lwI->@iB z7QqXnN_*dh`DSejv`4OAzOI-8Eh@Oh1nBEKSa&PHX&6humn-59cu=8qWAvYx~tzHkI{t1~NLIk2dlwS@p-8BazaR5RF;p!=Im_)D@bXAN(8DT0h3M3R9(eH5 zcRsilc98Ew2R(JTWY81d|G}+dW_WKV&tBf!+r3lf z$BjR7pVcK_%&_5&IG_Gd7A->KGhVE4y*yG_8+zLOWeztcIH5N(kTHe8P!s|}KR~~P z4c)}yMhWy!_xyr9f)CvUfU%*Q_z)#CHguDJx7Z#oOlZBzYWKv|U7o@de+ivB3JBN_ z)!f9KPhX;R*jTM^MnNH(@J4ieDv+csVo1hts-RPjj77sYkXk{p%cM_NDu1wd0K+A;G~A4hj`Bt%tz+>6i1 zE6NK9PX9DIOVm&hzfgIdz#L#(Kmv|6h_=$s+l*gPaH(#g*Cd3z8SV18r_*!BRBdJ2 z|3}ig22=G1InRtNsz{<2D2hMkNM)}9RpfTz4dw9uetTq5XfATvE5{F_fh+Q;r3Y-$nP<9=kXkU-s;+B~XsI&KgC}f1gp|vCj{EkqfL? z)&)v(E}(-*@WXp0_w)cz6d4n+%&!VtLegp#mTql8swqSs8XreiVH40*%8Es=!Gf!y zf=lSmJfsve^{_|}@@w4j)OEsdVwgeFw)7zVv6(pWvXg|WHkXbO8X<+`Ny9efgVi== z!%kWxF`4D_%-p%{U6DL%<6>jdkNm1@8LqW?w86@QRRYKh<{DQI8gWrA*blUhy|)<=WLYhCfMCjhdiVvvsq+frlifgVeATxGq~Gp69Ecnvy>rR8eh)>ddj*>HPQOa$B&CQ=n_`LYp)sWCIonwO0DU zdO`a++Q`^mvU_Cy} zK`Dki$}AuuMe~=1hldJqMsGp`L@`h0B$QR}*z#z5yW7BjId`i7DvYa)uF_DmU{8#g zDL``PCa70eIwvj3r#kw(YKN*yU?jS-1Vu-g6Z3?L3wwC1}Lp zvGG)ITO5)&xqWne@>NRj4aoP&%?Gis4CrS){Lx$j`#-t!h#Wnf{L36uDJ@pj27#YHe z3lpt`aQ`+cgsa0QL%8spZgHWTfXC<_TL|66fzhE`{Y$d)AJ+H#NL1(}Vc~thh(ybz z$}u630M5u`o2~M#t;K~v0=|?Xkoah12qeE#TsOl-KGLH7QdHqq0wH6E)B+Y!z7^mZ z1xzg}0audjr6f9|84)}P;L-*U;WP&lL~Nji(%Y-J;EGFXoU%OO5qlxIXQ}SZ7ZStp7tdl~*lz-Mi)#W4YtOnh^O( zX<*WVS1`J^V61Hq4J&Vl{JHD4t2gmqG+JMuFNW*66a%U=r|b0gN;i3Z7vFC(+iHe% zZCTl=HqsW9&|3bU0e5TemlD!2mcMzBy|Hb zGj+Ai5wt6wysDcjq^xl6U0eO&c-PBY8|o^SnnBwqf;P$fyMP9SS*^~pJZdnyRAp4h zMwilRei?88xDbBZ^biM_bV+F8At(zqGCKkAS{Vn~gB52$(C z3OgOo`veL}Y*Y1*GxFVuNj*8!U+<5rh&Dg@aOmgjoe%zKn#qvO6g%YjT_xC%(Op!h zLZJaPfX?=m_pRHZVk%e1sps@R!c`F(?W86ao%(0P_K)PQo9Ov6uA#Cz-MVnRDtITC zI59EqBp!=)QriwTSyf%>Pt6W0yLswwQWj%;=kKK#GTQAeZ*;W6iF{ielLra??uygV zqtrfM%Ux275E}32!MhDvhZYSwQ|E33cnUGHKZ5SW4!__0`&3Qj2u7F_Qdg7Cr$cCAf$wWbp%< zjcF4}@41+-Moa`rj14m}{(beCx$U9^Nr|m<7G}G!0?V@>>89|AwHcV;a<5TVQW=i$ zVn`FLa_*%~&RM#6>bwo1^*8k7wC=j}c^6Y7xlMd)mIyEkHG**)2*6`Ogt;Mkx(m^D8Z)q4qDkpo9E+#0a{EtDIX29x;*MtZ7sD z0EYaaV2HFpZIa(2#_XC!aCcWC#GC5MoJumSPT)OI^C>NIf_xx}jBB-BbZjDTC$q*y z{e8_mwZ?EEM^B}^J4qm-Or#_bB|eRkP-#3$-0-Xl`N1TWwq#Q%^c|W6RMsX?61eta z`Xqr>LQkIjh+#SX;c6vSwn!};@k;H9_=VBByMy5GV6wzv^o)-;$e!g)Sy4Ua3~?$0`3Q+B{3YlyWh`X+4| z#||>P+@Z63`YHn(9qjM|q2eBTGUX{Z>89tur8EsUCqw>BS_qy}H$1q`6@oaQ zN|pdB&b6ee+|Ch29N#USej@KZMFAC7olhmh6sbg=@>>i_gi@v$X-Q?{n>ihX`e!r8 z^Ij5lRa#5qwKD2Y=>CS|G&Jb#R{ju-L`6-hwIMIw(KTcC$C?``S>7u1{Z3O)c#Lj( z;ka?S`aT&~QKvTFRDY?hU%d8w1KBNJj9hz!{wgD{cRhsN07fST6>R>hof(2Q8f$Ov<8OX+05GBeTCq`EP6(hzB`2*MMu>0mAWQ4CS6ANFlD5#;+ zG)GZ#(L)-z;E?~6IbNIfz>)i|SS(F1o*R#^EC&Y9*97BfdZ|NC2l?9` zIjHj3UqyC7K?A?iKm-zGN!S@W8R=~#*bPR+TKk*W3s`Y@imWH@Gn8D;0r~(bmC*wb z^jH%h@@m(Er!FbbvY$9Au^cE+o>UtnsZt+}1!%-NoN=d8?s8A!?FkCQIRAJkzN-FU zyb108#Nk?m?&UOk+=mLHImh<{ltSH6XbGr_$IouEE65MKyFYdS4LMUB)ym8%i6e&8 z#7Q#}7LpXkyEOZ?;_7yXBuzMENMb4T(nzBoC|M$wz>q|NtL>0PIjS662@t5}H>6_@ zE4tFP;sM@6g@Rw_ay9BaRW`(7E8rM;Xf-!dAt6r{=CBo@QaEhobAb+9{X%h@fG2g0 z(j`QOpAT|k>Z!80mr7V%cZLi?rvPLW7Vl!8+hT_~G6ldij!gMrup?8yQSO{rL=(MOK8)76MCsAFv*k35L8cCp!I|# z`h!u4@QQ*|35IPvv#3(gVq!obrZC*77Eg(aLS02D#Y7i2PS}D2F@c&6xs>f{P zRl)UWIRL>M3*q&ICxR7MLXkcop|7vr!@h6X(Qa^Tj5`HXMu~>PkwC8`(GRU>$}iAR zWKo-EyO}(Dd2es`&TsJlWQ{*^i`8X)L03D^MlNbZ%0MJAF>I=-e<*U#)+;y`nMWM| z7I*L8h9_0~(&k1I%enH|282;IB@ux9-x`guH?IQ?Vj+WC2oggd=F_583Jl83RTI|E zh&hug01rOA`*PIS(9k21#z_7<-3g%&!N=fQ_`XsNVm?!lETmjW>1Hv~H0*Fvz^V&^ zlH`YiJAJBBs6j^zy~>ayMoW3XVO9Y**~puNl!TgSLjxmK+aRp1Qin-~AGP@kaIC{L zAs!ZR2sn&JJz>Bh4k}w00b*4D!qo{l1Ps&$1Xn(U{ScKv|5*wO!2D&mrbrt?#WO(Z z1)g-@pWI%r2q(CNbI!ED^U4d^xN?9`R^eb4{}CjmLj1UTCZ)A7RzO_C#`a<`qsR5S z^k2F65JEMM<{V%&#BeHZ5;Ld2Vo$fkND-6xYRP2`spw~-kfzvUREPF9W1tnWdNk!i zqIJP(qha?TgWm#5I?Ehm%2UMc0@J{_UH<)|Y*?Q@a;k>VBJ#SRwv;&i_;^Ssl*iKA znha@uo&hw%Yy(ydTPj$3*H%B^ZBsx&Q>TQ7lyd7#-r;`a(q<30Hiwh7_0joZ{R*LF zUc@osTAMo|Z%hT9^Sbqw7pf=25<-nz437l8s!ctvTkQziJ0A`3`&N0|n#+gbu_>o& zvlZF5u}lhA!xM76I$R%~nf7*+{~c0e6Q8;&dMJ2|(ZdhNMMHOOo$Gb!zw&s(3(M>V zpS;h#Cf_D?uN;oID)TWZrW^W`wJE(^vBlTqV~mAmwPA;NJRjl-sMC`8`Wm_v-L=oj z52VVJxHj$keQdtL2Y7FZ1dTjY zHf(r7SX7phET&B1&6mH;QiZ|uZ{kz0r4rBBJY+FXA+~PT3hD%21Q9%rD-?^*Fi}|w zDHMw@7;`YSKMm9ttaakR!3_bFs6NS_bc}=qmxl#Fj`A=-Bl2e*qAe!5wc3&sfKb8h zvR-_G)*BMFN1~#?ic@o|V*AaV+01W6pHq?1p&-`@*BjMt+L zsF`=7XlGCXM@RvAG7*1hN5;_2d?tVY@w$~gJG;HvUT0^v+@Y%nl%7r3l38Xx&m3wb zFzJ0AK1$~5=D4Nb+O4OWHaI)vOC9ovBYVQZX$;Pw&Zq0rI_~JQ7wVdQ4ESd3WcZui z(`!hm_~4R3-?o9B)<9YsBTFE$gZz^FktiHZwC0CvU=a6t)(&YwTqU*Yof0eU-tJ!a zfJTm4aE7Bnq6Y072#+zJ9ez^BJa<~oiwB(&v!C6+lk9hU?q7AkPKV>VA!SWYIp-)k z4rHFxI;kZIThps#ewoE}>b?c;m&00I+JnEu#!t|4h2>%SW8V{7Rt0@86-h#j>Y{H<2w@tyb2Ym^76#{1=s?D z`XEmlYo{weCZVS<`6cI<+;EWk(*@=8vnc{&g`Y+ZP8b@0#{F_E*60XBFWVleX-!qx z~^h91x9_qc+rzhfls)`Fc zv3%RfNq_R5M5R8aXiQ=2Gija`y^dw(ghlVQNX&-FJp`t&{E{y?g{TLC|tsC+4K5@B_c za)bCRB~_lJ`YFC>lyF;v)E-F-qLYs%jCKlZp#{`wNz20NWFfjVZ}l7S&KO$K?+dAY z0+k&eN=I~QPfZ>hPxTV2`wpY*Z<3U0Y)NSu2B&3VL9=veYyqrMA4wFl)}fRkYxok9RH3SX%xG3x4OPX_ z2}9NTH>nawRg=GvD6ErE@oH(3MAc*@(u8*cV53w#mBKp#X*=Pa_Q~C9WMK z))vTxXxOS3bA-&E9t|Gvb@pweb0e?qkOxzng%^$1*XJ+w*Ds3rs%E7CX`Itd$2KYq z`A|1zUc(%*FhW$<*4y$#d5heW!IplZefy$gn^XJZEO~iq={gyfUgbjV9nOW2A28*-P^RkeKe zJ79ouCLidg52hX55_{yB z>Zvk9EC0#)F&$r|Pr+YerT_jqvv3jGW45A&93_t;8{6#MBXB>Tk%!~Mn$^cH_G#hw z*y!xp;dt0?3*dtp`JQ$mmW}>Y0|~F~lFim&Hulq?5rCGNowv>!?h?YLSn2x_MGqma zc>;+0fWi;}4ZHg+LV<_*Ew#>17n_Vb^!1mD;I`r3`Th#Kr_JocNQ?nq~60ZCZ?CRA+mH{6cHq*%&9%xht;?{?_3^@|CD%OFTgW%ylK<>NI+bQdr!{ zY)@UQxdG>-HJ}LWQS0^;Tr7l%xNPLhOo^bgnDg$8+#9bI8^iu)i<4=89d$pBQ%wQV zF_ktDQ)v&92YUA`%g``PqD5RCGc$+_ad&Qe*a4UN zGxEB`T&PcOL5feJwRnzTa!=XBXqhyq{hruE&msj>uRf1`AzH%=)ER!vaLNiua>YVteqe**!?(-`$N_(r6|?H96LXM~PXGWCpgJ_cOesv^`9M zg)um`G}dNjK(KF{Vr`%CF!B#c z8sk!uye7i7N0YqF-8+(8d(rThlVAx(wdB`pdYWQrNq$-|ZkwbNwsXqNBE2e>U~WMo zX_)+K!Ez+^b9~}UOm%IK@>tkIKAku@)H(nW5tw|JgEh&jJ-~c8sXbi`UYGOnsH^^Z z1f$G&B^5(X!dBEVZno>KiL0x%xh~V=w2k>Ka8!Xl7?_j?+Duo{5T%&sN@@+uSuSVi zi^oM7`B+zb;p*^WwYV|a{vNsPu*F#)t}boxe{Lu$=MRyW_U@0YoYyU;qz+??--B7o zimZ>hpJvkPqJ#0)1^T|yd@L*K2N8CNT8tt2iyufMxrpwi$uEA(FTfGCATH zzRT!W0yb`Nv0|()pfNV3q2a|uY{{b}R3&2Eo+A1Xo9k_Dx1b4Kk0Oq=TloQ;a`C}A z#BdjVFFPJ@F?AIGDM~_o9n4RXJ?xIx_ywUa z^im}5a?I~E>Yt!bcMs>Q=PKi#2{h z*Hg)lg(wO@D5@y@nb>c0AUGMWpPlFWwUj^|s&CnBYVm}u;hkL56S6W#ct=K7ycMR7 zmSG0%Tln^%Z(7N#?Lo88z5Zspeg6vmXAk}72>nOB;9-0Fa+W-a39wAiRdkb^@9;tn zNQ3i%Yw2Z8fLVuxpQmkKdi@(pzH;>~OWV@7}2L;pEKXp-t*`zgSf(#G55!q#NC zwp8-(&dp21@${bIc;n)<&(@vwZqR;}@Ar{+<QF=9ppUCgQ+!GduolvhloN$fDC2)%6q=C~@)jR@|M@nYxvvwbFk5_j54QHq(qF0m zVIU8tmoE$piH9plv*eu_IT{TdWnT3Fvm3l46hwdU-j*SzgAKwpnT|(;;b3jDPPa+2 z9dl>Sb5PB5klwIm>{l=B2i=BYvAHPxkrw~dX4rhc$8~&XIPcRql zL_|RM{#kNojQrIY+}pES@_UuBJCD~o#ITGXb9(MtSDaAOa_8n?Sgx3> zl7}wMt*@UvIv#DFU!!Y@CTko0$@yOArWw{pBE=<(_8~x?x z@D+`}dmY(1+#K+~Ig?hB*A^0+gZE{Kj-!jY96D(bJS}=gF9%&yZp{NM2YIp;K!r(w zy-h}?q=K!n=~P9Ie0Y|;B$-K|1gsl{6y7eO%5g>`mbdxKP8%Ox+zwT@8X#eFR7=F( zv57K-s9sV!mzAd!bsn6Ct+${B;TOi}ooK2gNS5L>%*abl97JwotqhYVI&q=1Rws$2 zeu98U5~+TL+CLrqP$x3mksXC3X=R)ZG)j{C+=OoyXcwg)?U37JRId^x>g>duTXDf} zagjNNtAqr~mu>{r=Pq1httLsADsF{%yLT2 zE)=jBoff`35t{1$$$kcKa)I-%S%U;YcFL!n*?0D7$7wYQ)fnrM$fg`7pYJ zpjZ3f=57aggDhf{_LJ5YvZd z88s@Cn&SY2KURGsZ$igF8l?HVToW7d3pTYUx{(c%SfDyzt*!xxRQBk?>CDHyX`Cz}T=?wqv+K zp(|}}%tt6ovB{I0M3PmuKs$k4j>Ir<816XMaBlXLVnbhzb|V5nEg*~_b{-{W@kuTfm0$>DCo5i9{BKwkDU0BR zQKi-7?whqi&mOsc`8qlG%*wV_=_0T1VBM_%kEaq!<0^{W(RRfiWMsXfh+U9`=Lz)$ zgN?%?Bd5|MS2Lf?EX|ici<=Kg+_3X%W+68vEUjJ-`95@E03s0eq5N{ljkW5LkSc7U zi2ud2ROh5Dk^h?;-OdM{MKtvSccb^*{n+X6b_Q^SsP%2R} zV`DP`M4>`bEb7%$y9chW!f93gTh~ zLjLV?C$|)cx(+}^&>uoa+J>;QanB0r2m}h;5qQl6iE$R{`e@K%ynPmOw~7)3*-_OA zLUklDNxZHq3zDb=iN0+1%b3gfg^w0+l<}oD++&)Za_xbtf4H2Sg6{>)ay96Z_oB0> z2KeHS>f998Tom^;OLA0(A87#v6aY=b=A<~8vZxukssZlBXXG8^1;mWpPovX7Nz@7j z@e7sL3Csbu$0OiagJ>%_7shz^PQ-u?9bq)-{-_Kgd87Sv8SFFHjVJ z%#q4o1IikhI$OsP4gN}SLXvwYqrw~f6;v9lf*brz#SLxntK^cpRga>l)|{k5}OLmk;ivOT?&$%bs3Im{nO~W;grwCnD}6l%siqKlXngD{vnC*%6b6gpV)pH@t#?Q< zW%6$k59HUlXQ>Ol-^?$AL=_JUO^Wq2h8;wv-AsbW>~SHa!!;AfLpYL0l zH`F_8UhtFGcoUl;SKsZ76i7>puS+iaQcr?GsaNgosWlRB(ppkG$n0{b$2>~4zhURf z4yr)(SLE*vsi`gxl&)nd#8?uBT6CU5ED_2*1xWsl*gHl4b4I?MeNDAe7W9Rpd7>uu zGk*>wv)a8NKe_!n4su%WE?WsmboG4PLLQ@ZNGIimn6vR` z#I~N!df`q!fip%_9JvLp`ni%hcd<^`Nt$oY%U23mLuuUCqFE{io>F4#gxK{ zA$NhiUV)GlP#tB#AftGy!&E@u#x9coXGZoTOo#}6Wa$ipb=kMV=s*!(Q9B7C2Mj{} zhM_wvX={=$<|L%hl#U$(XhoU)c}+_R9EJ?C2z_8fYnEX!;;I&c(=(bl6{?6;tnWwj zgbuNnj=L6QKas5p-i(!b`A@ujJ5#^Y_u6i z&qE3(=uKzL-;SLdKj zO0X7HXhS%j$gv0`N|x%0qqUwaXvP*F1s^I$QbL44Yr}fd$tVQggD8GM%g-5x@;t#i zIXu_v(tm~Xl93hb&gc}G)+_S_*bsw+iZDYvRF8OMVva!k=n>$FSe&9>G380L-4n6f zJih+1OX!3!k~vMNq=&2}_i^UC@MHM|QKr)v{7 zpf~C5CA*h*iMue?fI#}k*Psonf|nK>L{S-QS`|PFEt>lm8cA*-{~eu2SDr;sw#L^= z7KC0-4Oe2yU!E`h%VzE)^EveqR$HY>{ta2o(3Qo{V@O87$8bl^TLhJ=MI@Eb6!Pyo`a`SxtG;6Z zYO-KLD|1>O9#3Z>-E;Sfr%sQP+j~ciPo28IufoT^&7L*w_|c4wy#-i5SxaprFr&NR z0=Oh$l3`hV5d~kaV8RTSGNDi`iU}buNF^5+$grix??-24f<)8u+rI=_Zu^fJ2^niu=DqmsSLt%%~)a(>k zha?BEG}DkMxKb#HNB$uwsV|^1#6h9tDWjlJr7GlE7Ll`tMBAuD{>ZQZ&2D1~fh;{% z@#AsYYVYk5Kk{3s9v=yILlxx(7rj!VfY3!;7Az>-M(AR?-EF5_9VQvN=yM?T$`&|5 zLhvDrfWernBnnx?AqDZVvQq!sQaQpx7CN_zbX4+uUROaT0wy)Xh64f!BcGg(-Y;GV z2nPgQvEhLDP(U~!|1PU5HEgC&6d=qGc?;UN%EP5^O(7u2La+eCC?M3sLIVa~*M*7c zTCk^ei+8R9w;<;#|8{jP7@~wZj3hst=`XjKQ0LXMOa1YvzjZ?J3S+G2fV)}1>g$`f~x7xj*LmnVSJ3cajh4={oJFCFroEG9FL{K-nA z`5?;cUiY@aAb$azLN@=VzE{)FLp8rs@tYEso0E3P2=|yW^1nt=Z)0oF+xepPt*!Bi ztwA2ghG-gE&}@~yZ2yb)ybDs_Wh@81$=14lS;$9zC(|Fq@yC!pNPMn%?Q=YZEy&E( zvmhz38w}T{eU2Sqs^)aNnYE5J`FGtIZ&dKEcg-or!d!dsBuEBg6+Iq|f?2Il9#*2R zx&e~>wqaSnKQUi4T3??pt~7CpyBd-O1&HLgJ^OHGf5?#Q&9Z;%52*Q*sH%(T_p_BTdnR;MHS_IYn_?T`5j!@paY ziI@O@MiMdSBkBaO)U_HVu#cC`Hki#_@>_=Yzx^fl1bysxgYJF5%*}#N0KHvaiTp>y zQ-o;Tde>G+=Bczme~Jwalw7K}-2r@;{O_Lae6!e}WwHBUbg7EskBu(HYhgqzT=CX+ zv947mC2x+BQLUuqm2b-t7*TkrK9o$xInkQ!0iRMAp-a8I@Gwi*>CyVIH_QGV-}*Rx#cSlV>nZG3bOV8?$~SH{Tx2zujGeUY%}TI9?UIlgNk{0l0o-EEY(i z!PKAfL6}5RrY~{{kXFhdfvV?_*hkZd2JKUo6Zpxfo401#{V3zvK&Bf;Qy6kB@=hUE ztg)PeA4PFO3Rb|lGWq2+dRMeVQw!RI7*BD(ChHw|5Q zd(e$b3x3^m0-94q0ZHk=2Nw3ab0ymRlMi&rtzi@vo%h4;u!o&$ z=(>B&Wtj=dWh6O8EIDCDJ{U}`dZ#^{sCiV^Dh>Y; zS-_E)d3MxW#SF+UNi0^iPP(bDc~!9b$4;mxsddm%1v?|ZV-y&dIya9myXKbemD=G> z?r3P0#p|Y^Zti6tm$${9GE(&!&)H%#QdO56HFhPL$~)xIRF$q&ZzG_%FaiCf_5pai8E_6szLFB?2 zDBD_t?QT(Tx}6;D-R50ys+$jqB$Hk?M%VQtJNegS4vq=gyiN7#B5dOMLCSg`z2*6> z{t(D&g&79c;J-ru!POl9*jjT`sDNtaBM6(LNWd*rSUdhR8M(!e;JoLiU9dl3OR`<$ z@@sr~iZ4aRk93-o_NGCO;3MZubB@fbi%cjoo@dF6kp|vajHHVy8d(rkDM)BzU z4)R&+zIFwWiZg1{2tkD~&GF|?flY~pH_SNO0U{K+eg_?Ne1nRmdKeyyQIjSFrP3T_5pV%BT99 z=XtoK*I7FCCJ59j`0s80Ksl@yoSBsZZ)`Sgo>e(~GVPx^KQEqU(UC142_A@>K@hV4 zo7MX@5A4m}Lyyh9}6ADt_3JtHt%66{C&b80I!a%OAuOn*9GXl3Ia@xBAflp=7~aT}Og-M8J~|xVXWEwjEj)(* zrj>l>uHG!On$Nxd2ToBd$NtkT|BEi+-m!LOy+4`sZXy4&_svB;r#Zj#3VNk%Z75Dv zV}V^Z$HUePSWI3Wc=n`SZ^wcb$V7SA>)bM9gNO;tvB~sw z!3FyB>i7)%c~=M-$d*5HtIDX?(>P5RT>Ql)3%!3;j)c zH)eyrM>yJm1G~n`d??nTGdq^|B%q^vM1ocT5>w|q?M(EQ)mJ+QK4jJ8;A!+crO-L3wO)Q5{ z5wBi4+`F>7sxV~BSv~S5pX`zYz5-w4O>fU?$*RJ?_={d)ePYDHc4OXyhjD*>ky7>g z$)n@Z=J_=iP^@kAC+CY;pZ1K3pj`guX=LwkeSLIcqW<6LFE@v;X#8DjU5A?k{x>t5 zGBGQ5CtIsfQ$PB0=8`WLl^}piksb~;WErwqbV;hEOjGAZd9D%>1z5X0Ojh#{;ur%T z88*S1*CGE`M&^ClQAIm)Iv!nhvIUDK7GUce+{kC-qZwI1;-!{wpVn${)N5AxToc#n*7@OPu?s~B^BOz0p_?1@C9j)VY)K!l-94p4 zQj-f?li^ywh{S5YdW8NeW7DJ-t<=$^JlawIHrH6RwK-gSbZhDpzSk+hd5~G|qjbtZ zyoLh9Dyt*Pu$mzs%1H0r`qm@;^|ej*q~~3WcWu#-dPn)+IacB3Rzy^Om>lfwcXV0n zic|?M6h#W$8PHGHfo24+z(z94dE(q{07Xd`>07`w`kF0aq$T=5uDG_6wu}Mv0iwp09*f z?)J^*Rt5F%B*%NVJE$-BLQ0acL8Zpm0Sw6>nh`r75WA!4TQ<*Z4Tj_Sb@q(!+P?Yz zbUNIejhJ0D$bo-6Ig#YS?%2$j5T-?fI0b+8R0_$VeMwAo%rX60AG2@X6~f| z)w{`w-b(^d14#qZ}09v{J_rxI0{jMboEw&EK*mC z#d~e(s^#NOME^7+&oDXYeF9UYvX;Prl#ka@#{o`^>C5CsjAcyXFUaLfbMu4Y=45E* zuF%*hiFTg@piRCGU@L=}liq4ug97BRz_>4T_oVjr?FU@ zoWB;_2;9#J++R)}>b*1wcdrv63cRXgrhp8i?n^-dL--?TFcLIaBQNhgl#B+3Ga)Mc zKY$#_n9JpvN1M~(cw;yi(aOJn>2f~*235>ce zqb)5#Xhd5z+hUBLg$(;72A6r?V@V9~@strkyP!N!>m?d~o##Z*Yc>$5* z3httlUCb$1-paVj9ybFpfI17n5y4v|1O)J*ZFas5wMf+0!GXQ!Lz(Os` zxXEszj1YiaMH?pE=K&;z(17|X<0GWLT2fKL)p&iioC2UsAX_YBB&5DN(olgdI^oq= z>^KR+7?NZeC;3Zq%OjZs>^swF!{tF3A|np67bPV~DndhNJV`@G~JQ0KR!i#$7JoY{<;) znNfgm_B{i_60m4XGUjqwGy?>nehdBvIBCo|5Vk;`yCmZ+d!D;6@3yAs^w#I207Akr zvbV2HR%=PdSoS=(0fd`_=sef=1OQbiid>R$l}l1Bd6!mo1fUQEp9_#+#Q9Kcwd5aL zSxz%pUMIc*F&7E~C~jVov6Ve;UPR=%PHfTfG~#rSw=%A>$IZZ5pw2q6Me`O30hH%1 z$(YLSOSSV_$aDSA0O$!>sQof-vRf!41Rz(@1_<|g07+p0A${F1<0GWL;xkYI41PF6 zBL;jrq_6vBjD*xzNA@YAqeQqSM>OdEZY1$O)kWEt-tETo%m$id(+}OdIAL z0Br)~x%*|jWzTaLJ@Q=OY(a9bLC{*C*W*+u&pjYxEPI~YIK$0Bbe`*b0zf}VQV+qZvBv5R1z(2OKfMc_~0kIYCBmi|NZayGmD|_6$gvfJ+pF!ud zm{TB3f&S0|8CTikW&j3IX8|~bu;F|4NC+TbdO*fhc3-NU*Fv7_e+EEL&_c<$$!?*H z5P)12GK@A=xN{(Ef%%IAGCo4;E1YRcNsZT6%PA11kiN{DyWSrzL$(47oqx0F&6Cnx!{oI2x#rhJ5ou8FRTTngL6p*;7XjJNE0?vh8IEA$xM*!B53gc;D! zJt$)=d!E}k+s#3Ap6h!8gdyO256Za8C8?HtOv`fx7Xu^}a}I=T`sz3d;w>b}GEVZB~E4$=xUkcb*URa}UYb$)4w``cyj&)LkdW z==>KA1JutwB;zT2o@+R<ZI(A6w!)nRVGP90hh=PKkDCu5@?3Npi}^T!pfCU+Z)IF%kDCEtK%Es<7tLGj zNdW3lo_kovRCZsg!2#yE00e^g`0Z5QXM+|>#!Yq$Wjx8vRkSa~odZx5n!h+K<0GWL z!kMPzUcBO3PJu85YL2(c7zwGbj_fm`zB*2VFb4A6TVjuMTj)GD{7i@nK%RT6jJsUE*^rsp^&Fk&9%w7iy;a6sE{kTs(yiYC77cR_#8V*8 zy;a6r_B{81N1iM67@g1hd>z6J=;z)lV=Q}~+c?|JL3G0FdjfUS)gTHW%s8VL}0FqcBB5EcB1(;2u&G0+0B#@ z7Lczlc<^fkQ0K!sE2AW&&f@b>A!GPkx*D?Z>9EeqPzkBCjwDo=sCeadoCHV_0fkw{ zOa8)KIRi^y7wuA2_PIX~U@63h{J)Uy${@;~@PhD!>n}PHM#Dg^bO6gNBP)BxYdEsy zj5qW|Km^FE0TL|8Xv?1RdIDsKyM&gbQ{M11Au>QddO=2DE+1{k(h}#q2cvUd$gC~M z=*wl+3}}KnE;t-u-!SJuT>S!I))r*sWzTvKwwm?&d>umV|AtVL(Uv{!Z5$5r5CA~< zjyZHX?0GWigBN6kjOH%(BnV~Dr^@Kc?o%~Dz`R!=i%x+3&w$W` ze5$fI-EO9gAb@-Yz>xMB?Q2ozgL$v_)^zPSIP;W9p=(u`Qvi~J7I{%dOGuq{WTQfi z<8{_?5`;1+%rau~7v{=ISn^(haFo!x&w=? zAsRyC@sfw=yw~$&Fz;3Nqq~IFlAdXKZ-DE?d>&w=pd%P%DT~rwQI-WBoBa)lvT!E> zxPvjYvMt@7_a5Gsyccr{gesKxD(lnj+y#ID^;U$GXzpT9f>4Hhs+y*fFbM%-lw9@hjms)N=Th;CGXXcvYY~t6t>8|wJHDzkhk%w z8kX(~;emawtX267bLAu~c`rbD0Z*JTJI@FEy~;v$d)}*ZxON=KV|3n&Kf*;-Fz;11 ztK0Kl157RFy`d)pv;gy7W#zg(@AU-85O)C}NufvQ!@;nCGgZ|AIy7|&Fl8O_pnFaiym(F`KXI>VBV{&RJZ574JaTQ zwvhLFo(%e6Wna2WSS{&UeBLYY@G|*MjzXEye*Y-Tg08{-#$-mglK|Yo7+Tq$ZjYgF z-Ilx;a|(ni#9d{bx;^g&fB^M2oV(bQAe5oJSJ|&__o*5nVBQOWAVUdsAK3eB&`c?d z)$L}=cofK2C#q-mC0ocll^TLTJ?W02l20X#hLPZKyV_ zEM#|?H3KZ+Iv!x|5a)w=ud<2Vp7-ABk@pJUj81@kz7DVy%zKp;>-N02aX83BbUWAg z!vJO;!YXS~E@8E#XIkDX_!uCRm~#NU!6-{vr0&X$EMOSyZ$J=*I{?5`a`=ydVSQzr z`km*-{lRFsIi1WuG8|9mhqd>+7WRcCzAqy?*3PW=CzIYOa!>E%`qr6cyBRaOezVs+Oi0ZyR1mSZ4fA>CCL zu48rA(U}$eO`yAuqadUqkyduDBSgA#9Wl+{I8k>3K>iZYVP*aL-TpdUI}fW~!bi?M zBJ}$Ka$oN>Qop@`ApZuq?5b>H|86wmUWqUudlLOe1^N$@m-QYDL*EA!@>+-p^5)?E zNr+$w#e}whuR!;E$iuytr9;FD z*zG9f+5%36rh6EH@8Fe{Dwj!mEQItFj5n0SZ5W-O8rL0|dMqU;; z3}#ViX-MRSI|@P?imjFH>yFrZ0g)HuLu$;|0UU*7fqAjAkln#y0357x36SRSoJWEH z`q;y=w&n1#+KFJwig&I4-*$2rGFZxbc89?-0u#<#ydgsYKtSD<(Gym8EeWeIRtc2Y zaty#`Sa)Rvh1Fe0vMLZxprekXAf!Q&mQfT!q$`iHy?L=J(oO)ty!fz;tQ>hUcpj+7 zg3tK;7k?PUX(%sNHn%(SVgoF=xC@{Eh8_tK0m_S&rS6Wr*b^dKEQZgE!w-cB0p-QY zu6L);He_+e97>3Pi1YS61VR)taLSr@r-3uT;MVmJ-;6m3LKw=6l@0HXym-N*Q4E|e z!%lo&>~SoZ7c0x#9eHu%6gMC7d9m*i01d&sSlP|)6j@8|rZtO0Bo*_008b$+VC1E& zX?I3m7Jv-4*acEh)`S9q{31}|QMR@3Ym|5{Y;CO%$Mcr(N|@TCqMCCnIn+B~s^+-A z22hE790Fa&Q2TA5gGvV+C?sX>xIpVW$#?W_b)p5GM1BlHLDruBg+yU}YciR)`?iAu zvTr4Gy;~ZP!5;unChs!?xVTU5?0jAWKnPj)b&q8M{66ySy>B%E4}F|`^jpXc>MHli zw14J&Z}uMg@4Tn}4|grzMah^yUH{$Mg7LSD#@{X(f4krK+XKem9yI>;kny*NjlaD$ ze})N*edb>;n18)!{`Hdi*Za-CK4AX!LG!NljOBha zmix_E?l)t(-;CvcGnV_!SnfAtx!;WC|3}@sKUscU_hEZyfyJ^3lG2rIt*`9JiENp5 z`Ix!ydlw)cOMnna0Hg?tlq^M~S?mxTVP_YZnZ3kHWs9b4mMae>b`q!jQ>x-}Quzn+ zsEQ<4UX{3L$+0ZaQuG!@Nfza@p46j?O85Qtr>&k{eMCw{(2{Ct`Cc|zV%Lf%nA-cdr{Q9|BPLf%nA-cdr{ zQ9|BPLf%nAUU?vj+FhQA;y+KwE6+sn-%rRZPet+HPsl6JMe*NH$SY4q(chOxqxjDg z^2)`%HvV|_Y?BU15*6=6Y|Oyz-nB|NVr#@}w01 z{e-;otQ7zKguL>!6#ad9T#ElZA+J0z#eY8`uRJove?K9wJT%3BKOwI?HpPEGA+J0* zMSowOoZ>%E$ScoI@!wC#D^E}H-%rRZ&rk8+Psl4zP|@F)N2vJE6Y|PKRQ&f7^2%dW z{Pz>`%7aw=_Y?BUqg4F&6Y|Q#RP^`dX)6BnguL=R761K&yz)d9|NVr#@=O)~{e-;o zR2BVwd8~^6JRz?JQ2`wb4efBLC|*AGwoZxN4Nf4G17s{DUibX16k4&FRE?eAQ>B)-dsuik(6&gohI z;mtoJaQ5&fuik&#jkjLC+<&OQbG3K9cWL-m@$jv)`>!ANclIt12f>h6#g*%u-xc34 z-T{H4m+u|-FU$WKgov5{qh@$J-vmfFe1~|;Ylnxg@83E)mcqRA(ohQav7=jWs(Fd; z5MKw!-h6U@WJLROMzlw7h^xK+rQt)DE?s&GO4du^hT=c;rIPr1|MH9XZV%eB4q_8X zebrh=e4F?NIC1dr{fAzDiXZC5?fK%-ougwJ$B*cL z8DLgNjlDlg4Q^+bvZ-`>=d=W^rUnw97T*xFsSW9e>vDx-!2}#F62J`nZK<5=ZZ=~4r8j1{cL|v`jbqX zo*kXtAEaDbhNJbU^AX`^5DSeP3%xgOp;$hWW}v67ffi!cpJWUa8)x`UfqmX4-g^Bl z@_@9{nCC}1<{44UGj`@V4*`B0G0vnh&O6h_dHMFyD~C<&eB|I3%`A^uvy8-U|2-MA z_#{SOdhOdDf7kv0Z}rARZ;w^aUg^A_fh5hKnTBL=Mt+=diWrrOHj{v?S6q#cUV-Gn_0`!ylTvhcchA{xhSTssoCAC zSaU!FO#C#*(o2e^SLlGw4Yn(__|JCY7{vw$*elws)75pfknX+}r)NM2YFQ#dU%Xeq z)X~Fbqeo}X3l{G0W9HVl4|BW5YYS16ci)Kd!D%5Tq{ZJA1BPNh+aE!+C~fi)ZS3K; zRUrNg%;FkHqA%sTiXmNOfZDsxQl6ayA;^E8Wpao>rA-bXs7{M^4f*~7W_1lv;eC$b z=IPC-+a}KLb@Lp%=AqjS@X6z$Iv6obvRQ{!<#+Qe%t zy_c`&N8{O^=_z>z?#Ng~U277TKglvK1mV)g1t{DFNhSNSq^G9=IK1n9%dXv`m)#nl zI64-$SH!O9tfTW)pM_p#3BavQq^OBiDgxRrUUGYN1T?>Q@9 zq-Yr0HYTENh@P1zZ)@!W%!xG%+}ppXnzUXzRX6P88g84XPqDZ|2T`d3<`?50n&EC(xDNqaW))JPiVVnzh?VB zb8s}U?UKvG(`8gpOg^EF&BxQ3DiGK`daMyP{2m>QSd;m#sEFr?(7M1j^)9}8b`zP z;4s2GeHKMJJ3q`G9u09DjJ!O!j1S6QzOtVmUa*`3%Y#oe%7*8`BZ}{gx6w{+yA>@;%mWO;dT8AB9r0Jz-wNrHMH zdVim2gf4e~k7Opr~19a{}Wy9|86OGN~PJ9tf7w!8< zv;$H=!EuQ(LEimc#sbCIli_wJjRmvQoWSzn6OGE@dGJEx_F_MksW3)`(kfbw;w1 zEPkrBZ^2r|m;-KQRvujXKE=uaaJxv+Ftlw#MB5NOGf&>u+6Cz{)+}&)v(UB)32mEn zL)$6}hvmU18g0Y#;A8r`jwqgPt<6Tu88g84=aM|Q3<^qkFNYT^l*M}l8sQpe!}H(? zZFk$R*%8j1Ma>8&VQdqCv0bciSRQ<;5jOlD9q76!4{qa|eb*$9Pl#?7pL{}s(x!mK zcA>IiC$*_Y(D0L5RMXuen|+*N%>j>V)=5pq0%0kiQtL&xD>s@4pOW}%fu7Sj=JVRO z(=#AMHY*PHOm5&gWPje5!FYJP)3nTDM4NCyLqg z!1r=iJ-Cbyis1pacG2QtdGM)5+3-AgMBknB;P%YU&H@+C2KND0J-Cb(iVw`uA#Gst z^WcRV5Lh03sxde`4^AAp6pdpih>PcgnueDLmoY@~27t>wC}N`bca6~H?(dQ0BRg-F z{DsVXO0u)Su1jw)45XVC6z` zhV2wQYFrn>fQM^;HYhdCqxo^7w@4X63V-T-j92SrTm{yr!5bDL1J%(QcU!5ue}B8}zYv->}@ z9$bbC#pnQ?yHMG%`}>^K&n;@{Zu@@0i6~<2)Euw~XZA&xu|P34z~wGdG^`$cu2DHW z4?gLb2S<1vjh6?P?oTl;K;pJBXKa0HbcW`^=Nf~<^WZJrr6Py6R})|@>97c2Xq*ktgD2+GZNFwmICBNK7?0a-g*|3w^g49CkDd^UK&8_!512X!}H)o-G`VFZ4Y24h>Pcgnub>oE@Oz| z4WL@|K@k(ngD*+_+@=?t2Pco8-T#?+a2YNHqb~tES8#uiWPJ;B{Mh|{N$Teo)pXIm zZy8VpC!&Z(Qggr}oS6rgu|P34ip#lLbbgBvr^Q$c^T~ml}h^^WZJ*rXq*7SF=;b#L=<1W#+-9w^K|D7-|y+k-Pz*~jG=8yBHD(ivh?I#kSk)% z0#^zv556R!ZObj9Z7j*ry0WE4+weU2?85Wlj2U42WJw-e1_dR&0n*0C)9XryM!3e= z@H}{8KHc_fc7!u$Q8U6x7~2x?*Dh8#^d7AdHoUJ|pzCwagOkT6L^mrBz9d0uOIFn= z`=(8n9__2PB=uE`j@2!)*~b~y9Pqei<-wOEq-+V0vWpZAn5levrvipf!&tV@#` zrvA}-@D-^K8yr#J3eLSAoHYkRl(X{SGB7A62Ryb;OL|tv1o^N$_)6nwcpf~d@l=e_ zR(E!)n>`Oa)>(bgWqeQ!KS76di0SraNAut-jk4i+@QA)U<82lzXJ>&6XM-Aqmj{>8 zLh%8Byge8~V)fuFjlto0aH8%@(KvR3xOhIeH?Zo#Weicgfr`tWpAJwS`-;@hZB``@ zj?|BCkTZ}&EV$!Fa;LFuGOsVX3>S*g=h=Pvd}PBiEh|z#x2UDN?fV5MqKGY1bHE~; znFp7#KruGJR=;-Pu) zmB!%kJa|jZsmP(7+Sw^%;^T>!YTW`Sy7kOj2{A4+XPg*^xEf+@0Iad`fSS_b=QdKCwdiz;;@bkM>;oKY z4rrLG@==bxA##S;yGx_Bio)T0IEm2~%qE?eMsHy`#0+E;yYGVOIjSDY9D76146%0} zH9If)@I9PFYD<{BTa0(#$GRlPnR_@y&k(CG_yyoh-J)xkJEHuT=pL>SG`v3>oLDw@ zi$;2#Wj#9w!e0K1yjX{@8e(;Tt99DAYDoBBqlVWI8=fLh8opb+v(v=vd7!Z&et~Cu zh`S-SU-65@>&oR~{vTncA87~<&yq)`))|bWRi&`A%Y{=_^}(F&o`%bf02S`MB*DG? zzek-xBXoG8oTyT=p|z%+nE+({tIDPHt~DaHeTxp(?F>+GJc?L8H3w`(yvDrO z%XlEgIf~jjZ{sIqSlSn#LCvi(J3M1P@0c;Sr+jv9nJ_XI*iGtO{1k5;r{Q!XfYV)& zTt9@GRwHzH(!8bMRQ#ZqINH9=&O#HXCjKSQx-?X71W>tc%^X{lu4I;JQa?AV!Okaj z8^Hz3G?GST<K@*f?oDwg#roF6mNC>)+Q zPt2@n?=FvX<}A>VWAbBUWRR>5N>16!SXJ@R`}LUAlP#FfIp@vE;}c?>dB3IsZR5TC z$TkdzY@mTGKYnl)4&B4YB<@?F>U2&T zy@=VxoShYC%pf0S%)Ge_36j0@_*s{{IYR^Z9zG^<-x9+W#W?N$-F*kdlj_VpTt)`T z>Y$C=wk5z9F*I*JCiQcJM%NdGbFV9B&4Cc>yu7wBUYw$^EF&uW<$LrgSpKGp~wo;N3|zZ9!uXN-&IgF6H-Z!Y7AQ>^BiLqq#FxvzL7+0EXB^uo6C?P*?n1ZStm|8yZt-GW$yK)HgQp7cRK?V z9FHQpNzDO^b8g;zOhfI)D|ApBeEhTxOZ#}095W` z#6sy!6H@;-tKQBh^@7zI$sn@QHHlA%E<>IF)PV{sEA7KiAZc^2Oq7RQ<* z6N{5kK{9zaERH1szI36O`$S`Lc;0+@;dyh$46seRC~qzUgJkh(eq_7O@Kfvz4P@e6 zqi}fMJTY9iy_*e6GiQMZ0*@afBZFjhznLO zIb#Nd_~zA{%a9^ySeBv9+S0O94$WyFx& zU_5-l#6F-;`yMS!K^ozaYS+mv*%Y?Zx^X4)HNcIJUI5shWFl+2+-)Ifb zn@=@Dhv&^(6G_F-mh$Gr(Xr^|=FO$YldMavNo6P1i8ABx>r~@$cwaa>ROgerJ(?9P z(@36?l|X0Kn@j&E8G31#%98cEns8{n`ILyoA+sqxix(uEShK)mo0m_MQ9&|!mpW85 zra~p-;-Q%POk;6)-h5RlZ;tHIO-9>=2^*7U%#m5ECIf?F@!1|f8P1qAwpF*3hhy$D zjl$u1^Tf=$?cMA+XU+l-1YVrW$RJrAAaEBe9-23wk@~p>>OSYZIeGkiFfi}eGAKya z1_U@Zr@p3i=>2*|>gN`nPEmZ*a%${T9czNjgPV*8lDz@@ZM*Zwmo9t{Cvo2bZ5HOu z&vg$^&yl%@%a9=18`Pk7KY^rYPguXuv-*t0eM^|UD8}vOtotU{s3kQEJP??BI1NRc z&G>n9rfyNG?5S6vdGi^m{~Mf`z9gM{-kdcDLM-s|<}x@)u@11cPHTHs%LED0y!lL{ zYj67sTs$9KQ@na}8Al|C0I1!A zA}D&VpOgB?t*YeB?Gdox<{L>l#`4D5o$?y@UQfgB=HS7G3zZJN*Uw3PIdGl4ryt$2Pc7B;KG8WG| zzs&2?E<=E1Ux3qHkX&fJ`CKD(c;39F;8cj9`!+juOdJ$*aK|rW)|*R@Cs`Muau*^W zo;RP9`odW~pVLXbV3|g8tnH*acbTTaadQC2ZFl~tJLCDZIT4FPv{`x#s&e`ei&BK$k-NQSt+`V;pvU~gB z{^5yy)V)&T{JX`K{a0=uoSybSDn8PG>Gs`MuH&Zy&dAs%-OeoYy8p@0AlV!saThG9 z_^(i_Yas5$BC8ke-^dx`$drvYwr0JvSOcz+MIw+7liRABE^1fi#N zT9ki}b%>9NkM=)O>JX_JGNF|@4N{x0hFhLD8rXJ$5$Jxt&^Q`*KToH4QG44Y9?&k{)sB!mmJcCWMe1$C=p9UHh%>7)32W3BB05aSK%l8kk z^?jj{HOBg$P4c2vXOL#PEtc77ieF;AoF^lMWO#tLZCL=qqV^9_+iU3U!}Pt~Q)+7W zMD0)8-*4Le{X4|R`yXTO@1$7@9+g4{3d#JF;W^A5UC`V<|Bq4YYi#b@+1Bq2)7d@$ zQ`XGiBfh);*UQXInGUKj@oT(=fQHyD0Alxm$oW5@{-D9TPgL=T5MrvZDtw=Hm+uwd z)Bo;z?h?*c{5sDkXpr7wHatfeP^WjR1bW|JXml@o-;X3RV`**F3rg<6#BsBGB}hf` z(A)(L*;@c)??UAi=zV{oalPz$FrxRO^MKt8N)AhmNnpXirBcYCAe{x!qE_c{MW_Hn zbn#hq-PefT)46ruJ5{k`+uHeKb`k%e_}2ahNQ-!OXe_*ScKxS$3px$OTL2XAq6GUp zsC6}d_bki0-6Hurtc70^FZ4giw{U8BEQYTz+YT&fnB4-<;4VlmbiH0`v@W||>ky(G zmeD=h#&OAd&6oskYu*#83mRCr1Yq5TN{1qKODa-_C_6mw(rjLWWU;1!Tb?XE8mY_R@}o+5b{pq>*qSi~Y~R+M-FdjYj1H2~0ZMn#0^-rG zF}e&cpHO*fBSoW~ALPu@Y4OIyH+*Vb9^X8T`-dw`!MUhU+xK-E;@gzKV zKc``A%TByGrkc@3O`v)9rAFBp>$|=?bS+u$wgP4+k&ioOo?S)=X^{uW+r^8B=Gm7T zgUjUEp)-1F-(uABbJgrwGV|;*P)OznxZF7ifk!@?XJ2WAE|X`E=(QeJ$Im|(P6yW$ zFV8N+hT;h;fZ9DEa-w@t{0J^|Qw4~(SfeSf9V zz3hEIlJ{fh;!5tp#BsC70hjh<^6F*WkW3Gdy$h8S&$F*Iu9rOzM)Y3$Jm5@Hkcl3} z8Y-k_$mG?_pdgtXV0tA-L57O4wd@+6XJ2VVFOz2__a+U++BnD32VaS1TW7n z znxBc($v!+omgVeS;kfr(a@0yJ`m-20DmxOkS?R3#|H)#zNna3q|+tiYP_d$uy1z8 zx_$TbbeDa5{U-7D{bsS-KmF9f>xZZPw~4o2e@lPo>gE3BtDga9g}F7}yT!2o&X)0# zql^EEXATlhCFQpZ(6De~!<^{i|3)pKk*B}bvOvmb(esIokn2{-d*Z$QcgK{R8eljG zOnFsYx&DwcrT8B?Mo`{GSi#1702tcV#Q$Yad~4GI<@n~|ty@QTUOnCY^xczN`-kqo z4{jYC?;r0zf9L4z=-~Fz_sfk%{@>lx-Hk`@ADr%=9o{*;d(waS{de!2o(&U z|MkQE&faB2YdgcgBp%+R;rd^Hqc7t@sdDE(xcYZLdg;=o%a<+7-9Z=VD*Q@&h<;fKWFCt=nZkT*S|D;NY1*;oT1u!;(y-}%TfXQ z4}IY1_U%D6yAC?4a`^8DI(&zC+x55hAG-Q?|8cLi41SFeM|W-=zPWEbsDm3w#>ThM zwAdFPKR7u$xOw~Vy0Nrw`6Pz#5N~c`Yj!RxakpeCM0tI9-4H!PoJJ1y?UpArTOQ8sFXpm3l6_tyj_O&(XE5C!)D{_ z@4S;Dl+yjZ)5Cok?61kiX!BQg^)5Gp_`iDMT|`}-93I>P>oZYTZoe#_CB^9E1$!M? z*%Ax(w$|Y_T+`64fWBbI#PSwwFN)4<^@{)0ju^gtgIjN9t*0;8>)oOSd)-B1#(_q>ARc*9LN$}! zH_eeUJUV{zlb?JNu9dxEV&M{Z#KW7f^)JhR49u4YoV|?@hjueFv!A-z7p|}W?DL<0 zpLj+51vUHW?&;ad(JgttI=y}L%Ax$PgI8WN+L5QD=2Gys8*kN@g5LGsC8!>3E9r?t z@r}tr_DA?Iy-LH3s}1u%?})F9h|_=cU)%if&f%xMzr8ekySRFI=a&1+L{uoFJ+Ceen;WOzye^Ah?lqw|-M!!M zf4jP&o*mwL`S8_a+4uA1d&h%e3|5W(Y=3VsqAK{Khs_~tG>Dmv5_^5|=*RBdJ3ehD zdb*KGD&k}ES|m-~{I}Vl#Rr@9K&NXr>w%2bW?_(WQ^vF3`ha-&;MOe}BA1`Ob!*V9 zGtEChLNmYgnKfh9dCmCrkN)73X2sJf%Z-Hc%fuD0*&J9}xFrTJzj06A{ANv2j*rm#>PTra&Wai+iR^OW!u~tQGriK2~)j9Lc0wcdX z&{MQWA;Ju$fSVOm(&Ow3I%`%?kVPh}pihd2UVh``tiSo+LGw-4M5Kn)Qa+J8UkJ)3 zRS^L-e`N8wnRu=dOR~U}Sk0u&_gT|=NxbFk`){T6l<{uM+ zn(dHkr#Y7A$%S)-!qchv!Z$xIUKI~-zNXLC`pbjA(mqkO@o{zkU`<9d0Y0Oge33XR9pYc_ioZ8ZYB$otOUQ7IXBTjkTP|> zvT)=#Wi&i!CRg#IdCBQ`Y_kmS6yMVSE0ipQ_(>;+O6G83)s6K>A=C+G)d55-g@})yB)z3IE3O)ywvuuv%9Yy-q~ary5zayu<4G<1F7`Ylh(0jV$vT)9SfTX z+`X_H?~eGVJ@HIO5j_Ys+)X#Rt(pQQ_aN~zJ7MY$Q&R|YYdE#C)dy8njAX@kvYD{1 z22jV$cXJ%Y16jSTHJ#koVnlLiU8f%cvtMsAX$_^K8Cv|=8OuR>8~a_OQet9$Kp!!xds8T-)3Y=M)ciWJlnt^ z64_TU$R^*g8t54?K4X+H!jzy=2{a{~X(ed>X`=-BZI&f<-oY;4SxbJ(;J*@sG2kdY zKgH3yQP&t#7acwQkXDy232xLSzipu1eL!N(YC+?t5{gwG(YV4${}YtTX1^m>PlcaK zT>=CZK(N^lX#8~H@YC?Pqo7fmOKJR6*63`+YNj^WAgFQ#M!dh^@Y8V%Kb1JM#Qydd zyueWy=Lk>}=nJ4iOg-s^Ts@l4>D9x0doZM3B{nc>&N5b6Lc;b+UrbY5;gmqpxg>9;3t=c2J z(DuI9@1DwvfxCC=W2~1H$f_{b_gUASh*5tx>bmJ^!1W79SS5YuM1PTJx&=e|ti|GVN-r z5Y@G?REFxRNt&05wcbaNjy9l%!>D!EdkOH^7Br0>4Sryta^>4Jm?(j5=%7@VmA=INw`oXVU-Giw`a`46iY+CYXy$)gZEy{!>QojtPO z9=wxsS>GV143CzW{YGNVQ?}UYL_?9F7fU`oo^7T+)*Vwe;Hmz6UXNDg;;4a7CC6C7 zr)0648IM7`lKhN653$CX#p1r1py~!YokV-05dJ&?-DRg7-fA6HRQxOyXKuxzpCdGI zy4uFxm%k5dAX`_9`DdxXgTwfwvyHuI2Og98^Ey zT}%8oJ0V3fDo-2YqG@4JYVU9f0KitcI^&oYrZn**5wtU|p%#)w(-7Kx%53 z;k3q8WDcd~&N)U&I&}QtsJTRqAxTGlJIuSMS0=ZXpv#@U{*6(-3>`HvGlzT`^=-!O z6?70sXD(?Cn`6GN+uP7-#+ptu2Kj&VBSw5c67lv|`kii5&27Zj9pyK--?98Q&2Vj; z0G;_Q3lT^@0UIKqD%t1!vv^*EmgDeznOyrLY#@tqLLTQ=5vrnHPLW@9D}Hs4>}Ij)FA;n zQuJj3yvc{^M-zv}8phn*eeG?Us;2$s<|_@QZoa`Sw6%b#GnKydQfDa-Z0-)Vw0Ed~ z55IXdSKaAwWGNPZP5jmVH%BQJ{u~oGButaxUJ5R_-hJIwjgoe@V{Oq%b3!$N5VvcAlnG!p6t*E~>Z?l28}) z`HO$lyT*DuVa~tXS-lQ|d zOLX5fciX9Vw`KUw<}$jhd$-lqEKRC=s_(XP45DJd-FDW#+sf!@X6%+kR6qJ|JArv! zd$*-tJ-@fP+sgZKw69Qm$8wk33CP!>)BG`TxgEKtH9N6TaxI+V`qU}Ee}l~bHfEsrFS@iMo;~CFa^xMOX3fE z;yW9#pp-PaCJ?i}7uCT`no zSaSxCH;_j*!}SFBIMLxovf(~& zNz8OX%;zqP=c|dR+0xr{mO*WExXmCWRM)fxLUGZLfc>fELa-e=jmcz(Rm)}@n5iju zfYeC8_o{KU^k24Juh`PJ@f*!BV1H*&FAl=Z2!J>>)f->mvsM>>)H7vpY&Esp&6lcs zSW^#KiEW&B*vukO!i8^U9>}z<%`7Ikx0%r|%9S@cPyC)J-^?N+cH7J%A8b2I>1G!D zbW;>W4}(ZUrF!wW7Ui3n4{&aEcZ%Sc@!bg?7<(A0xASHefT4w}>$-F^i}=<#ZDuL_ zH0#qHHnRX#EmgPOUMlWtXi1*-G>uSA0Ii6=76+iL_)M=sUnh`Pq&{wn)XGi^`H->t zGjQ^x65U7_DNP{VNPU}u$pWk=?~A4$1y(6dH0NjnI7j&}8PfvdQFv#hS1DQ3Lgvv# zF^_b;IoGspyf!b9t67po2W-%GR^ml~l@uzEj@;^+r3qvuO&}|O)c$kV3H(?nB~ zCIFkxYt+;%%^-$LECz8ah=N5WdC<1X4>d(6Dak`b^EpUTnl(VK8iQlR4?CvPyp87S zqNOqDu9a7p45?#D|C*phoI!C?63@TiRQ^m%5vTd~V8#Yi6!6PYjjmuCw=0+=$lc6> z^D$k)G|?$HSHUDnl3As?m3!S_-SR66>n_uQo<6OdQ+x22Ab#vx3+-`@&_Lqzp#=Ul z(&!$2=YA^z8|SqdqOVTDe*gt_bz6mJ7K5CxDjr!@A7~a1wq1mh7?g_?10mkqk&w4Wt z;L6H{9=~YYPf?CrK;AaUS*Eo?Eq_X^Z!M zvGJ&)MIyk9$~0WjdC>4|6Wu0&=oVh`0yqe}_$bQJ1QOlUw+BT^J>F!L&{fW^sumss zoHaV&6;!Fot_vHs;RJHlCW^C0r|f~d3T6zUNn2xBjW*GpN+#eLG4vw_6hawkbWlj% zU^j+@jI@bjq!k&mjR>KVxau@{qGy9aRHbWuP0b#6I z!0NP#ZvH-{Qojs?Z|cc1SBZRhcZ9wvK&sFIVJGO3>-h0r z-l4VuGk}m1D%7{Tg_!jx1-H8@+xaE&Wr$+9YP_PIRAU(9NqD&c?7w;^@i^TAHK!-8 zE8Z&-*FYC|^;<1*Q@{l(?(g=*Cn%o!+Tpf{f~DdIGlDP%&a9BlU|*~yw8yIm4{00z z#8yO1Qv|=dL%U)I8caS-xGfhM)HMSG{f_V{s#!^Vg?IrD4B(+;-nE{D$hL`r)g;O{ z>DJi(7m0XTAbAm97JP+v6%PI^^!wr)U-0e3E=;m#^1Va3YLiaUmgJA)<7viXa_rW|u+IGMx5W(t0eo?QB&sq0e)x@@;+Mu3OloUtDZ~tZk`%NJ!PJNpO z`?YnV@mn+5{Gr*3gHqc9ur>)tYmVO(GTWw`o?5TmK6-us=KcLwZ{NLnaC`sG-IL?p z*w5s%ag6E9^>l+n!Q$(we0NPjsZ-E5D{Ek@2$ZHqhjWcxAx&+nXlflIhNBoq@RPU| z8`EXkzyP~|%0l0ZsrvDo((Favt*xenaS&LaJ)LR?VNnGQ!EfO~ zyj)FfytPuLwjua=mD&r9^Qubitk*j|0fOJM4faYVTQhCW2c=gWs~`luD7;%Q-9;1& zP<*-}_)Rr2&VHK)1A}$4l-{RV#J22Q*7!H_SsmuJBcGXV{zzx*0KE^i3dmYboHanW zv^yCHZ)reJo)t>J0?=Yfm^ID9NNBhlzCpv6;lTo+#H9ea8>RGWE`A3J>1z1PF4H=0 z=B2Yil~n60aqP?Ns{V^)Hn=#&voO+ME$I>`U-d}jvs9cIa=(HOVg_Hp_%|Q037^Ns z_5PBa@KOE@2KVn+DE9xGJweKkwMOtj%HZZ#nYL3&E8+_~^lJ#vW*e3)(bu4k521P3 za1n4$aaFo@H%$Q5%UY(X885Yl(=)OWAS|xw4z(02^(%{k6i!=tS9a z-~&Ms#HNmELfW0o;Ao7A|ByoAau7g21x~vzSq>sT6pzM>UL@9q%RzwllxVinLWlRy zvuln80|2!x%>U9s9(j4$++rcMZK|kk%aS9=<_3eVF9txdasZAR{oWOzK8B(^~>tp7RR9yEsZP44_Q~3?KKe2Y((@Y zL(p47^kzaOf@y-DtX8g4nW@{yt4ipNL>q$MvH@;9$mK!d68ep9;?dF3^JdgJ{|vmVyXJ21<}9iW9I*-y*#M4s z(q7+YV$vV6uWtqWx%dqzo2#TO*gG{bPj)fH?%Q>aa;c(2%m;l@T`gsSs2Y5bVZD7( z6Tku*DddJn>x0|(4okcD0hkpfAPh5TEnBrABLeGF603Gs;S;MojmVs)b4#TY1NUZfA#rWs&DJree*ILeqM@v z$3yyFJ&TWtK5Hob&?68@|1V7)+%o8(B-lKN!BiXjwW){=g zhJdzzaJnyh9-rPlp>^*g&9ymOL@4C-VvEpB%!V|%dtXGj2X-@y0KU~XvjER6+{}pG z`y%R(Y-R@CDsE{K876VL%r4)Not50ZFEWjG&Zkt}*7K|CD60(%6{0*iI^t_uyv-mH zZbokL7Am-y**8N8Z2`ucg!44BZw3i%>e~$Lca>+zSo0g=)&$mQ&va+D87Qo!n-=iX zLR^!M0$bBUg4#?G)JWa?LesWuweu1=a-XbXWaGgAJ)ob09>z#d!;dAuZTNu z#E%GZc5?6V*3o)jzVi6UM-LBfKXYWjGF{3&VUZ!Fu!{XQL%Ue8+xN{(7Yp-kTI)J5 zq&C?xyV2%#Bk86aJ!=LrP*lT_o3R+4n&3CnnLbDGn}Bu%`ANx4)9JmN%2@OUet{PB z{6U%}Kd@wc*!Nk-uD=_pfghZ}K+iusWdK}V<{XN;`@w-qk3Rx-@Eg0@{=##M0ohY; zyJj_yDyrVywL#J*?VX&@mrn|PysUOkc#vEn5(UBo-$vI@asc+M2KHHXLF$6y;-qPSe>dt5& zp)IK0Xbi1>rgH@42B;9AIjnx3tLkSb4K2J)OGni`(n3LV{n*DGJobS))%(OF8!DdE z;X~?QG=~n{pwT}!2+*nyOHfj>HpqmoJTz0&5Re>%sG1J}^k>vYe`MFoW)R}Xo@pu= z=Ni}E02~f8q4FtT*Z78v6mkV4h!1tRY|xGQggEbCfWU_9{YENYbR%n61Lv2ORxi3? z5}*d^e6YElWKEmu%%VNjpc#wsasc3)TBp|{(x4|Uvrf|~x@ppHK#XO*Ls5YRF51>k&|2@_JA3`!*{)=hoL)F}P=cx&*}Ea0>OVnY%%-O=zStArR#SM7aX~-hZ2J9G zmCVC-#1CHZ;HU+sw}%GKV2-pbYPVzH`t`J`&6?Jk1F z6_N{UO6@^(IK8mTb{7-g+wQ`vd-ZM26E#gS?6TcO#O%7=MLyYfoAT{0_VJBbB1FTw zUYB3iOs>4scITIh6So)d_>*51w!JN`MVIX^z>AC5-F5kP7xA<3TH*3LRky7~;x%vA zPTO4oBA2VZvNjlzXHYeisz*nZKiku=8szHD6<4q0sUmMndt|WKGGN_x!Itl8ks#@A zu1I$bkQ1O&S>p(aZYW?(y)WVbMcp`V&gDvNK+f(IIt4`wYhB2pn=1}ov2``noK=dm z?rRQjmf(w~AhJxs7|etRN;ac|&M#_LGGw!janzGZ7IYqQ9NjnyEKi@$8xrE2MIO6h z;O7y9pW$mIS*5A4+qi9|N_G9Xt!huIqCjKNqXvVPeM&*8dVX$@hB$5k(NkTpJ_T5P z{)jApRIi>v|a!pJp_KFCVbNXm>U58Yvh*>utX5uGc}a z>>AIP3+H*sUViqKm;L8C9^ILkSirROq8d!VzrUf`DcM$&wRV6}JRr5^?1NisKpvtC zs64GrPyF>A9*PHKZbPS~y31B#r%W5+9|f@l?vXXk!sj`O-|LB7?M+h97Ft@WX}34m z)CC$-q*vv2vIT1myQf5RX5yA>J<}CDLd^IFC764Rp`+9y4Td?BTENn z-*AuK%DR%+p*t)c0rFI;>1`|>c%wrD&MW}&FH8*30VjEHQkUojB>t&yXDcETK2-7y z@d73B(d%5lPs;*HeCpdApo!YROp_R=uzqo&JKHS)T7|`rwIxc)qmvkoEg>aup(uHU zJ8UV9QQr9Yuo3_dNIC=6-i5fqrm9ub7#FbEd!dTGEuzHr2vhaB2`l}c$pXA*f=Xv; zjAIW+nh~jT211%wr7^0P8vukf))0~$gg7loW1O_nBMHoIz!OkqiKkO*p?e+$tN zSB6@+!x}6Ef2+6N={ALg&_$-#qXz2sD2>xuc)S9bS3Qvv$1BvQ)$`N2M`E>jFC5%I zJb6hy8FlmU?4S|xI6P=lE-^1zobMTNqyH2G=Zj8)4iPAr>oFSC6CKIJLA>1x^Q#Cj zF|4hEJdh&3vLkMEkiCtEbkSq>wMU-Mp)MXqAGqAqHbWFRQLGLkg)i!L?Y=%;G>@yw z@K9>IXUAox5G$!`Tn=8+ZJF^gliE_`R|@)i8d=#oE;9jk(!tr+U6z@M&sC0!lKmQ6 z%sMVJ0hZE5^A1m`?Y`JXYwzLR2`>tG5kIJL2uOHYD8ftU)!gZaBoCmigcuPYSAndR zg<`ERSR>R6u)$Seap*$#g<4Rrh?;e>Kv^erT*cTEvQE^uyV$c7Tao!?KDI(tzR_EE zPKJlA7*56p5m9faSZsxQTQ^<-OGFp4{;~NYF)A#!GV!n##v2hF;V`bVEDA-!`qQjgoVm_%}WA64#wyJvct@2E2UO z`uVfs{r#tl{2b1Lep66Ql!@63w|>B(Jm{sm9NWGW{ppT)eKz**om4rf z?0I2m1B9izsiF`I3TueZ?ue&5X}?8K>7;#hs_q0-TW?)goXCBc0`*oFg9E~JTycC9 zCghe4tBY215fU9&oB)UE;`8fHD^A4o3eH*Qxk#^+owg_&d}&#i=dy%67xirp9z(s$>6^EDI8$9yH#hF3=D94HJeOr9o{P08 zB)X`Ncd}?HA|qMvd_;yy8PJH#60luDB#&M{rP(e^$aYcRc8CmhV#Y6E>C{rQU6xc> z$09OQh{%v12GwkrS*3z0jmW5XGyx(rZ4enXh+7a*7y`0gW-UZUJ$|ZCB=SR<8j)GX z*)E`%5uK80N-hy;g~(hlJJmW!W=UN>V6Uesz1LHa$K71%G$xZFLj`1az?zW$uK}AU z#N++zf{+|0f3Eq6XRfaDn)u>Q0`|;F-lkiX=`YbAKRbBv-a>qz%$q)3 z?*TSg&3Zzv+VQhcR?-%+PsDUbuF|b4B3_rRD)PCuish71(s%PYL;Y#wd6tHpu5yd{&D2N0zE{N-A-0k3qQn+6Ql zYiFN zUE`^HYRxo6%zE1Mnt(mE=53rVp6REx()mBvPiw|nvhHbCqnwn&&`WgC z%zm}UnfG4UHc@$J;b12$C(fTi}e(jX7tARgYl5z>GA zjYqFu?myJux!SwFBLO&>-P5U`$Hn#jGM)N~I5(IH1oAM%p(p-TFYzV>K0peSH3EiP z{RZ{kf_Q1P;dge#vsJ99v4-uqM4}gSnfmQ^5^4KgOZP|$_nP?4o*-{Iz+XKi+`DXG zqp28f^+c-{KmiQVZsv4Nd|`(U%K};3*msG(26Y@eTm%7rUCUqCl-6rPJ`%nxiQE?4 zU7}eRLAB(^Jh*VDiQqyPw9|ON?>2Utn5YzFyN>ld@oMih5z)$bn#jl6E>W`6#6H^~ z+L&9rAPO_Hl?+q3)A&I41F+K&o)|9OX#(iFU_D%y>@*Rl!6>NL&S4u$AlijHO@LOH zXx0TurI0{%Z1A`hVE20NPX{(P)#DO%k!Z{Q3bK3Ew+F==x6=o}RE^ULzOR`lWM{Bye5Id|U7-!!LQHyva%}_bd6+P-Fm{AoTlMV{J8DGQ_E|yg z2p$xcr~)61rYc=$Ggj!+Cv+TI0x}!F*G$)|&3{OJ*=lj>#0)nXR*SDBpxS(q7!8)2 zJonIN+JnN^Ebt7f#S@Mad6HWdh(n;EXR5{3OXE+w#~kJwc+OOd10RRnmNs}mn6BSzizj>c7G>?&M!#JZk1O7aRI<>+`n3!z5Wt?k{MaW$N} zz!imYR-d$R)J#nJBQ_irN9WY7$Kj~11ioMwH7IHa8SCu634qlX)fK=Oh-$!7y!yM2 zG>PL9>PX^CsaJHp7J{!Gp4>USo$RpJzO)!XF~QZ6JgrNYC0qE0b=0Mp_orz`P0s}` zW!O!2NBr|%f^nPvcp+X|S*-L8_DiWEVvb1q_8Nb)m-C^v6C{Y9tWP$#R(m!HFQzqg zjighqZD_VaJ~|^?OrD(W9^QH7?ybX<-P@}3 zEI+8F{rrxYGo@|Aqg-tzs_W9~4MP*XkJiw391_ht&oA{vyiE)uT?-TCs?(mMsw}a2 z5&4@)1Z`-qALq7-cz9v$O(Z5L1q$u%v7T5pH<5@)Wt&LkLv1JEZX&TyHk(M+i(Sxc zW1CqQY$87Hxs^>Mf@EIQCK4c@%vEpA-9#e362JQ7_slJCK5N0>L;|>zt4@N{_Fz_F zwXzr?V6f6*jew{t-qe^6WU#IjgSBWgc8f!H9kGG{0&2*tpo5*7io1dY)RiKjGQiFS z!9YG$xhco1Dv_f$R`}*KL>j|(ApUryJPgDI!Rh)D}yIEk9dNg&7yJAE?mm5?qMkZR3s zeqcznUwh*nS9|@3Hvrr4p$*aWDeJ3qG3}3uz6yr*kuD@vb4C1oPY^;Jd^-o|b)S3y z=e603R~YjB*bW7C0`j#XJf3z2UHt>gQd~i=SKSpKMfSu`^+fztaSykH8f>9CJ}6iW zogZLf<%B@#QFO%Qwh=<}qqK??L5hKy(C9c-3G(YaOKPwLGoUrI0Q@k?5aHt_RGnL8q1F2V7Ls)P4;qird?T$aRljd3 zV$``Ae=vo-O$1GiKV(`Anzcb#lGxHiM`%=}D;5~3`&{c>m(A+x zT^&Z&0hj!>gNA0;!S764>J>>E%@Tc;l0;BP<`KZV@%!hF2PIWU8h-;qUvb>K7o8tim4r@RBYRye`}3?@Mu%qHe4=Apf65}o`7;l(~Br=9dv>KZh;5n z(eKWihG!Nt(BuLwh+-EoFaxx_<%+em#SC;&nT84$=8j0cYT9cspbCUklK>k?UW2VQ zHCI3@L6HU6)@+k3P`3oQfh@P+;|A2Lo>ABcO`!8}gOP3k)hJAT-Qfn*xf$F*I^d=b ztHBLc+9yVP#K5q)!OFu8s4gYRCmS@YZgJV%FD_eY+&~qV*^>?GAfUKxyw}DJbaC07 zuIc95K7y11%R?AbHGSOR268?d+yKq`w$?j<9gL_i-TXAXYP+6JLkv6chi&i!DPM53 z2Kd4I>_u%NM*UspMJ;}&MqO8Y5enw2=GZ%_!4Vdyd^OIaCGdhTrmLT0i)ru#ocB>f z%#st9kTsCc&|DI}r5V<9Q0^9T$<(02V`9}`Fnkq`%|I1oj~g@%;^GYs;K2$(c8@-M z3`MA^Kdv0$3Q*EEuu!0^K^5-_mjHLrA6C*?ls1rE>J|mT(IOTUXce87 z+k@+9rfdy}+!N1j4d4I_MQebMFjP@9g*nh^Z7Ev=0*rz0t?Pm{AmTlB;=E)N#8w}5 z4&bi=0nSjM)$?2f*7kxvs>DoqQzH!{-SkCYwN%#^zB=3e)zQ0$y);&)CyYFL!iYyt zkf(Us%}K5`>O!~B)f^4fFdCa{CE!TYlmO@nU7!YQPLG~2QuGAAISV-o^5(|J8pw*b zZc~|~F!DGGBgj!8&&@c6F5=KmVK@pSD#^fL4N{R0z5K?>S@Qtpf_kIXSi=GoZ_rV@ zWFwn}SG+OaYp@122r*(H)-Y;e4GUeju^1CSW;hC?C`VyTRh8~!GnNKZ9ECm3ArgWO zRBDCH47wtoCBU+aqjW*!#D?VXwa8ohv?{RN%C2)f*R=1+b8aGpR!QW6d-mt*& zlqzSvBEZ2D)Ky{d1=WayjB9L*2oF5~RjQ(e#8FKAVNZN#Bp^n+2WMx8cg~LP-q}4p zy#2}DH|`yt+~2L+%SQIG+4QpcpmpS@#1s9;Q;tlT6Fjuz?FeL5@Q3ROPH9W9B{9?O z2|jmOJYP*j-KJoB&N8TN4y-O|^ZcVrDn+`7SG1P+)A+T8;?N%cHJlsk+_yuBNzIhwreNMIexNI~4<{B^tBU-^@H81ydMJ&P&JN1HWonPHLY-T=?yVcz(0`SInCwO2)8@x4pYKP4% z0Cx*l*LCS;7V)j{-m#UfDslNBI9vH<7NEYR>K>kJc-gJffxwmMW}odD6xidNjTGPP z@zV2|!bj|s3HG$W9-y9uP$<1z?F}Q7M?D+ysAmvK${#YMJ8VE<-Gf8pe};cH%D<)P zt{u+Cl}KpL#vTi8q*!Q$Hg0p#@(OYdeCJ_y*i%sU{&=K; zW7gA~SK!l|<8d23)=zIP^3R_70d}lve8=Dwh;Y53cm4SRT{fq{md_E_)$w}xP2TRhX4#KpD)masg+ptXGJJw0&~wx`Ug-_SNvFb zrZMO{{uF1ub59C_XUgqrBn#y>U_7XntS4mgAvfZ6s%PbM^fMW3q*By9i1_MiB8b57 z{&fmVd5GWc3C2Rp+y>ii7)Fa1@6*;cRqHV+o)rqTKl=sI1u!yRAixU;KNDrzS zC9I>vKy%YHPNbToo32OFFEG)aL5*K{)r;eq@e3;N%rXkc9y+un&cXmgpn5SxN&@C! zvL05x%e8??z3RpBSOY_4)r*rhCZwwu7u@Q_v4+6?u18}kq4y1qc*rGCAs(@}&9Ji_ zAQg^5cXP~H#4kVqz;)AkFYl75fsRR)?HOPrPg`tcA!hvv6B|)?9r0x-pexZV+GP#C z0hC+oh?asfFyn>SpG&^Du5hzRTmx%>I^|m8ra%p@xWC&IpP;zvYlp=oLqLu~$r8g1 zfumG)!PHd;tQTtuP2przo4JqTBzWcGt2;D|Do{*_DQG^%wp?UT*9>;&j?f0{*&q*& zeuapj1qSdCG!Qog0z}R0Z!j-1rSOYHEG-zJ2ulmTLc1mhe-`?^^Q3xy6J-Sw@ublYnv z5z!6)EV5PA+gGex6oa9`e2cfLGFUele+wezw`_+5qHH=kw)+`F5gdDTy0N0ubvOv* zC9&#rBB6{pI>KuRf@6;mH&%o=2Eq$~q5J_tGrpEy)%aMx(AD7P0ISM>%fOZsl-@?i zdaZ>g9=}a}zRbdHus1Ifpa)7K3%g$(1nKgFS^c5>uK?anqMC*tI5GUMjZG)=>kjrU zau%DewH#7>o1a;+IP+46hXsdHbafy7{6*^mQm(MR6XH92spX~~Wp zC_58q89BN(UD9lU>I+5?=1c*Le9f zg5X*aho#`R8zuqA=Z*e5P)gTofP^Bu;dc*i9USj#(0sX9IvZ3;wQe290?I5qvzQ(P zMKWvTTrQCT`T?>VwdS4z)AM$gNb<8(BpEWtipYR6bEk=}kBjU5B{`X+910BXgUDFI zK>TJ;kaB3P5z;UwHd^fdS4-qP(MWt@hmNTMZMK2S5`7Ko_)r)tfX!D+lkwXoW>>`q zUzP+(3#%;AEG@-8tfU-;))kmC0umWO(#rv0N*-ba!8+ot3zq{QDxymi;qz4zBur-n z{NcjoAOMF7oOWHZ97KEw_x)gBt-e+ycGbe=Ai#r4H0#cWwLA%nNsF0GP*xlr29md! z=348q;?&m%#S1L%CFY*vL;>73FO(C>Xw`dUeQ^8UVgJz^LcWcBY50)$G{MB3yJz2-Sh{3R5E4#MazEJPTrISbE>oLwBhC1`V)N z_ff;8u0?DfH5^Yhuv7G?;ds`@bo8T!%Q;a+reWYj!`Cf(A5~R%HD)0VrZ9_Gm-&F+ zXRjuJ+&2M)EYw3}-*;QbxNzO(-px79YN!H#Dep;pwVR1af5g7pC7=ORxZi-nxgyPi zebpco(^?1V0&&92j{^#XC#tK`ED+TI6Sz5Roh}M4ThzXRn@#xv>l+xE69g-8rUbQb zU}TUmsk%$yN|ctUMrHuTOtU4ZeFGx{gd%cUEs;JLFeXTS18JXbI1}bH8lm^qv_&5F zqxKDq2l!!QmANfYd@T+?3B<(}L( zFeYsaLMv6byEoSqHy>x^n^^=8X|GQPW6O~K>PhTz;51pAFJ~4?3CQ<@DbtY zZu2&XtZT%VE6x`UF>s1@&7czlP2a$&>6)>j!qprZz^Sc{nH8z@r<AmcOM@@B>n2brw?X6aXM^j>ouD-27cg>5 zu*P*DzpRKVSjvnHhWCMKG@nH68yJk?-6HGFJU~AKk`G2!TWg52UnBsD{U}9*#ym*m_fJ_G5^zLG) z{xJe!Eyvh&c`Gu8Pa@oMXjilMn%pm2xq z+SpAK$W~sB;MDYon^M_P)}km!jt&}Gi+UWnsp80yiiO)2Z6ijgdjeviJl^xcBC77K z($Es?01rwQkQ!p))T^+bstRjMETqoO_=PF1Y#=A|63LX>_|Fmx=MWuQQM(svU}!bj zy&pXd&_hc#7Zj4yq+rsOu4B~1!g&J>sX^RIh&l*-iFnb%gjS>GwdEDSJ@8(IS$_`Sch4$=)12Ql!& zEbJS&aD|>lyR61HRvjMet#yP=Kp$S;zy-77(C_|&k5jP&YPvyvlLc=o#_{-h_q zE5m}ld++S^duO}14)2^EZi#64f~C#g5Kr}=pwMR1vw~;<&d&H^PkdWV;XQr^eOZ>T zG+!ESO$9yi!cPKP`2O#y)Ax4UH=rn zzayS1)<7F~E7#?qW?Y2JD=5XQ>tb?PXtzPOq!&$BWC=p2yi;v{qGIr|m9+ zrnK2*k@Gpa1La!f))am%x@>nb;l1rHgf6OYYeYg~m+dYhX4ma5^2xT_ly7&jk2kbL zd*a;gc1LJZD7tKSK3*K**;fsbE%6mQZFd2JT)ghC%eT9TpPk!wm%8d~ce@K<<#M%e zXEUQjo&ncpfYlc+)O3vbZ=3#wGmqA%K0kP({5;3y+gPyxJFg4cY{-f;kDNDCXyL5vaqZOS%dOl7pYzcR=!z0o<0H@jG|UfX zT?*x+(P3yqSDYC>+6FtGDL$Hop^F^GD{`O7hiUbP-6E40*3uOh9w5!)cN$c>Qn|aP zE3W)<_k~8KRk^!8(B$QVbj4*Gn$|V%D+;O1jH&p*8$#g>?EpfjEB+1L$8cskfY`4O zaullA1&zCa>RyH`O_g?pMcG7E&xEs>8N%XBryc=ekrjmBuloYX)Qy7gOEqC+V9|kp z3slj|&a#HcH|FnZ8q24=Q;(4rW%5aPiF-_}`inwLG>-U-KZeq}h6u%H90Tm*jT~#9 z)f4!A-fex^UB854*$v8%w-Tr+d-;i2LIBTmJi0T1v4FGbEpT805ZCYFIV;)1lC^k% zQ9O7xa^I{eP=MZcU^a?h-{E0M*s0&pqG+p{+HczZL&+shTmCTyw^Crk<(3j)v`CC7;mHi1%4Y(S?rw0NA3 zGa0($%&Q%rsoHUiGg0SeJj0ZXHcoEdRhg~OGq+Qvr8_P?bcy=s5^7QW#L?cK<2#0n z1^u1mZxxT|@AWS~ee2dh6CRfXH;kP-Dxr7*RKc#=u#!}h9tQo09ZDF=<6J}OEy^QR zU8fo|kw&8wd4{n9NRM++UW$70uR!>~_2J9acLsxNsALsx+1;u)gMs2O2x=+uZ=sN` zh_!IvH3-SF;t?c;fnc`7RPBpQuVxL@?b9D|^Iq_YOwwC(kJoDPUO2dac=FQSljGf+ zhi3=+O^@*AO1av+WU;+x#Et$_3~Vnt2?jDEfxpLS08n%!4;{(hEa(%9-GVA>Nql8T z+~^>C8zt$Y$LuSVJf9m@R#m}r8Llgi{iYZoM7m$p>)On-E}FL+aCMzwtFEGzcUWd3 zx7_nzW;iHHw`Haxl2SJ?!skb23)6=sdh_~W|B>qt_b*@l%$1JIOn|6#aQ1bVWhUZt zSd3Q=r&Njq4YF>_On|U-(fqcS86wg$2h5o8@6{4GQHnS0b2_4z^P72W9wwW^_J z%mKY6l!*_w%sqCC`Z@=0q1K1V+<0>@^omR8rq9$I)CSSC00=7d5|aRj5zew6LuIZQ zDjh7_rum>U0nkNOTJb>_stv3Mq3jJl=rT9#jSW&VSMRr2(1m(`xBlT3na(Gg*fXa- z(D8)Oz{X0WF52y=Ri+kp0cECPloh}(ex2#cz%HuL6mt%{IEWV=>>?l4edMEu2e+R& zS{t}U22vAW6XNGA?gGgwmH{)NLLi@Ra2NH=ufko3?19$t7Gf`et%B)A{A&nHxXw${ zlABF~zet%eBFc02vid&pSbtS~SuMC!_*@NFUE5)2J~sXWD3)~(iIq?pUv^h`+d+04 zG~)tWpgvMpgKP0`dg3LnL%(`(eB2F|`LOl$XT|&bPZxPQoCO{JP|cf(X%4r9z@a=~ zCb*N(C!(y}&PG>POB*;6f4U>y-^tuhpPnAQddIT%I;nC{+4HKzngdgYwapbbX;*Fw`B@KQ`wdrh^FI;<6|`2Uo|Mu3w@dEy5a;VO&6bEcUo~G zp7(&SIP0z}PJq;O(!RZE)Lwe1SPfvs&~Nksh9hi`@wrEfnJZdM_x0Ur3B8@zt9E5> z%-rL~%z4}x@}=DnEuzl8&-cKq%1C7^=(7Sj+hphN$1;+?=kF_tY%6|B!jH=59 zZ4sLc6pCS3mCxA;h{u=Sl~r|--eNX*c~NBni_&ku+Y_IN-Xo9hoE@GVAKp4TI6K^0 zAF)Se6D_jq=Vkc&?I7bt9E+xkw_{ zU1|omKZOlcs;M_yt$0m*r6=O87IT7mh+w@Lx3RmSOcFH&h)1uB4FN*^&ovl8y}Xza z*in3OCjp7(BySV7%Ji3Lxt$$6c;_LpjpoYOuZ5dD00FBRRLE63ejLh5+JgPc)Le1R z&TG`32TQlA2-wnY=44<$`5z`NmmTpxM@w#kPkLBP_lRK$x9YNo~ubjYzKC&c&iFva9`$LUpu<`UCI;NTkb4!t!pT}nXJU`rw@}=^qO)F`)tC2L*St$&ZUtcRC~t#^ z{s6zP(4Q0u(7rV)pDw&CseOaDZ1{Z($nPuk7O&`bVMu<9sg)%!ws^m0`boul*m1t$ zkneI05DT6gEmYYCKz`B0gDDJ~m!rzdo2SArp0zvJ=hABt>ZyFp{YF)3T=3(HCVyI6aO0 zXY7dLIU|bB9Byh+%@Fq2@?R8d&5D!@vBKF)^k4)4kF?~Mn_q>nv+L;;e}yBtpJ{k~ z9NxS^IV$-Tj^r?l@EUJ#bIa2GSlR(S~F69aB%5aN>wqQE`&}c28 zi%-GgH+tgRscQ&qEym_sp|MnY?UcKCZsK0ZycYR0VcJg^3-FEJA zalOAxb08z`4Q2xIKMa-ViGS5gyiS1+kb}>|^V0?y--38)MB;aL#IsebsPTyHxJ05W zbeZ}i07EU6BWY)Fa`?@jAn!WBUp=(kI|!k}M$m^^;s!u_o7HohE`EUC>VBfy3L_X=0*! z=(o38(2yn)&;xv_ZlhlzOTVt_Ju5hOb5bYApVhlESqN=4hhjOCS z;DFbhL^&n?(%^C)HF%+@!Nq&G+k7#DI#B?>H@tB0OVZ{0nh-gzzMCC{F*Q% z){$3|zA*3FE|y#^phR6dY;6td5p@@es9RvrHfD_))K#KSBznZIL?^4>WImSYgDm_? zblI^UIF9ZaAbRec#6O#**iC>9V9H->`nX6!ZQx?Z@f4 zz~ei1%iR(Gyq93?raxYYrB)UzjiY`kRiMlfN#AVaFAH-%6x?d#1ZmbMvp3orx<=Ay z*P7O4@(~=_-16jX_wdducW)h@?A}&AFZV~gX<J_~&pW zH<1Vu(Oy4HN-jYduD6Mhq0*8~BqnGzn6$ff*Mh%^L_{jvL?Rz*J3)fNnGBuSC(mIM zK_4A0-b8#%w7pIT6($eRuCgxJL;_Tlx$3RCn@Gf0DjZ2ur||45ifv$mzlj7;C|8{| zHxWvFSPTfTUqkSaj)Q)`#zB`J`*o?vs4XNa;E2wkGgRXPxAn7m7mxT?O+ z0YWDn5e$=e&4#G}6kf?S<71)# zzse706^D{D>AQhwb5qpS&9H}N_@EM-Fj{bjd|d!tgl5+2(g8=_l}gbQ;4EN3!;xC|kzy#D_3(@0g5M!iVaR;oozhG8G+3F^|g1o^y~H zq{vhZN*Op=qqDc%A*nJ|kZ2TW{)PG)hJgWAC_jphpBPihR&+oM>gzmHw$T$*8GuYC z>(%(c303JzLQYvOmu{W^`!!PvTN#-9suFaz0V6j) zg9u>v{2He9csyt=*Oj8c_mP1RSE~}+mEZFy3H$&dV7uv>ANUc!DL89HBDHSBtcwCa zKn(aAttda&|%2P?%p zV4w#eosG`kTo%V(LE6f+$68S@omLAdiGYqCSaW(Lf|Vi>u+6Cv1WXUs)&d|1vc86o zAW#)=6^7>11Ee5Xl4%hHlTUAM{ptl8iy%;MqSh(Aa2r~~s7;_&+ z9ak<#8~f#GGlL+MJ&ZbUBM4(tjy47~7gSN^sf`un5jb+f99>GC?;dLeLAroK5YR+$ zYu$q=0^kv#v;Jv#>2~d*h8T+A537-cdAO&7JN5hQg>526{axmTtzd^&CpU*CxGFpL zj%pBvFpv*BjAyO4 zJ|0xxKMy4T#(r&utChKn+D}fR8oMxfI)4 zEhw;rC|Uyow1Mxf>w+~P;yv461By1tqBS5u9165rXAMBRMLF`4!yGEH72Y(MgU41_ zDYioCl4=($d@HoKM?hKdBAA|xR>*`fl!cW?Sy(YC3)Fj|yAI~sM6*o8tJ_Be-02Rq1FSQDb^P$?sKJ9n*9P`*U#5~Ohly#MwS z_9izKg#=AlkL- z<~EXOU?4l-Y`Wll(7N(d;)(v_DOaY<2_h4CHyl|{{NZ|ndkqcfr<+Zy*DZ* zp06gNZg8+YXBpHs2UZz%Wz*KiE$QJ?o)xX4UjL1qxG4^uL&7RfllZ7vtMmpbf@_p4 zzPuxzD0Rc9Pfw3ty<=aCF4f|oUK|W-YCOJ}XH^=m=nktBf7FYl%!AmrwVA~P_cpWe{#AL?^2AKHdMn?|A|iI%%pxCbJ4@+i7W;Iz zdr!R}L%Wpd<-B||Lm=I3Y^x=&RmE9xC60i*((l99tFLs}%mSFVaCKdmZe|hR+Qvei zf_OH??I-T{4a!VS4;aK>=;%+N{bkO z0RVJX$L@*Gg^wPC&iwh0NR>4ZGI-)MJm&HH1C32dU-QK0!a&HBCq6+Q4E6_F1~{^L z;&Z%Wt~;uM&Ug=fVuIcu=m(nqK=ONfEwx`C#NHpoFJJscwCNKx(WQotiYIeadLlL& z^I9^Ge2({Yq7d{6nt&?9ykotov3B_#g|(My!id0fX$aWLw@01 zFFqEl&X=)9;t4;7Emw;A3=!Wr%?}T;_)Iad0G;L&=*ROJf(;tmI+1M;@% zf@NBZ_7u(=G)(wlK;*EV?&B~imX#LLqhb%*6NGG02g@|YMK;MvpLE zIFo|vMoBN?14o(GD7FWY!Hh;|qjf&da0DBzkJZ;*btBs`V0GhIy&>0Ff?AzuAKP1c zNRfQ>k7f%45=}sLBc0e1fJegc3@3hd<5Gh}Y8c|oAkT2s#)l@Rx>3H8M-$ZW4Am=u z%Hvx+Ln_<%4UK)srBGoXv8U3o!zZ8`5}ygYB7Olv0j}-Nd+A`lx>;z*18=+Yw1qqt zV%DE9ArCc70`S0Wqz>^8>u@EUMZ2v*Iwl#SE>-d)dL40{#WoMDR2H=nZe;y*0 z9LdD*_QWSB&idNnwy1}t77b3icq!S zE3^x9@ModlhYPl&_Z8n6gK-lZa@N*r)h1m9f)U&FLejn#e8ojkwV26=Zt+$X!9&_y zF@v{dSn{=`^&q$HY*jI#y{#$)9jfj^o``8mVd+*C5wFWu75QA-Mas6S*oU)Z!g_9& zUr9>0Dj(6UH{$@5jgtW3Sr2`-R#QPj>SFO$Rfg@h*H9w18~j;htE#8`c#ka^BP0)IaS8YC;8x)wuK(vnvu6!t%>v9{? z;G|OBxp`D@;@6$83?8iN&a;7FmrnsmgVVhL6gI`@6BgN0ddaa$Z;jA~9CXs)eN7r1 znT8zp>om9+>@OS2pvhUZ-|PueMy)kM8q&l@i=mmdL=LW)!E?-$k)*GKe3*>d2+ zMztoEk7)Wj7g>1S&}uD*h-3zvfc8!FK(jhiy801!nH z-sn(}ya8a7wxT)OpH~bxHeKXoD{k6%41h648EkX_$QsmRu&J+?7*wM__I`@V1fDdI zmq0%7Lsq?28Cp+G$9(jY24mH(A2^OGr#4R-z>6WO1JiPHVk#(`=o)Ldhuo>(kp845 zA5uk_=E}x>6meSlQABmUe_Fl%xv315m_t*5s)ivtJRto|M*+TL9z`TOgCYHmHKgs6 z{-)GNDDP_YLarT(^cQPQACUg+6$McK#-L~%brjhr+Fj^fW4d>9PPH1ez`OoGX|H%Q zG3k%kSG4O1}qP96AOum{XU?f1OZJ8o^^H8#y(+{r3 zTjNtudkjXTZ{ud9g7V0BXlL{-b zbC)6a7z{;^AS3Pd$=uD)ZNu_#HJl7szL~`Y_cpUac1ov4gc19Y8|Fs-HVoo!n^{a} zn{&AVPT4`?rJrA3`T_WHnWg@OHw}b`LIPiq++#2z zeWlHe*kdp@fgqviysF;OB2{gaU&q5;t;1#(K*5EZ*?K!V%4)-bsVJ|GUJ!2E&y77| z-I&~dF7&A5&4D4G_O~i@LZRe1Iznv7g6f3~0Q4zt90MU9aYQimF9~%uxLJb-nx1j; z4UMdABlRw)KLLy3=qDoDxF86L@+vZZ0LfjB_Qx9 z^)i!%j?4VTD;gimUU`k6qzs8@k{Wl#hp1A3>RT@&RE(7o2+K zCpwpZ8%wewV^sq!3(*0y&1aT811e>rwf`MgQ~6K#!HSB8Bu+m z)>+RBvkhS|D}z;`ba|};O;>h=L79}$6Upn@ea{dERnI~t3?K|nr<@Pz9)nZq>Es@R zfhM#FyzYju$KcpCpo_gI6oCG41{#0I>#|FG48nm*um0+e>wtLISfKE9W59TsRmnSG za6WUGdBvl90?zjZcMS&$1vkjNTsh-pr3XbFYjTgl#=RixEI98Q3}YMZ!-C=cs}+`w zQOy-d6wqJ{??xHW_2@H*kb8=&d4ov?MbhNuI)kWMwq z+uVZW|Ll*|*J+iBZM|wd$=(%F;~CD_1kmj0!UPbLdB}J?@o0AH>l`SF+QQAS%BEfs zq3V#<=#csys|A!fM+cCsNj>JA>;PA2QjH#Amb762>4x$S$F4J#XL+XUfGX?|0HI?> zhHj|phd~^uB5aEpQRikI!_;0j{$QRknM{eKJWDqm`$!R}vc!y{)`|usTI}98CjxqS z1>lkGF*wy#j+$;bHXxB4hG;!VH=G3cki=p2yosMPbi;{;#anbk05hTb(me(h21f2N z2-ZJ9IGljmUvx;=V^D(nhLcCClXOFC$YT<2)8KkvS~}^5DLlkGtiu(87VWkc>6leK zlE{{)3jo1PVf9B|IY7klE{Y6|bMHp1O(L&(^gsA&@74__7^swk#rZfw(S2KCM2 zwxY1RU|pMJVN3C6J+Ue_hHRB>& z_SIzNFFKd7C)n4|k>GZnVhULoDKCG*atlT#CMVIX^g3}ce6l;p_!FSu+?qb4w z+nu)((Qhf$_r$g6vfV|*?7H1WKG}Ag^6f76@r_xm%SXtHb=@x;eVw*DzbM^vvh2Xy z-tI!cco=8!ciQd(EV+2yU6*fn5kEV(?Jjk%$=biub{CKr%hm2;@w_9?LX#-hk4|9x z+oso{YD)%SHbwS(qWnC^XWumMu1rwMUVtwr*;9tXIPn;J>gycb8NgI|*R&KyuhM-o zH{dg?gPk=qAmGur-~gEx&deSIZ=x7@#b)kAVw5Z!J{CP85NZrs?GiJ9p5|>gbcHI^ zS{M?es#yn-qUdQBkY-KLdW=^DKasWF4Ll8AV#&LoB{5EXXqxpAgHO*YmBSkn#!Kwi z2RT+%?4q_=NR3P9z2DLpXPQXuXpFKYq@D^#<%Y(nh}lXUKx3S-D8=8eI|s;Yj)L+_ zbzy{Hjs07Y-sc+RjrseV#`P%#PK?<1tVKs36RZBB5FL%9|LR8U0Djjpq4>08L4N$@ z;F>3)1YMEf|4?{Iy)|fmvG5l0U&XbZpHg_L1}UJLN?Qwwem~Fg>CUvqf&wRRkpmNe z7=MPd^gBJlT0X!iX{a@qUuvVQHAN3F)_`;=etn0B^8uOL;A*Mvvi066(=yphH8mTD zRWVzO-|J3zU zEbv430|P)!Q8r%yG$jE!YgCWTr@mfl)KU_pta9Q5Ptdn7Qw^=RNRU*0pK9bpx&WmBG+YA#4RSgsGF^WK zfC!R_Ztl^!+RtEc4VcWrZ5!Nz3u-@uffCnq)ZX$UX^|4iP&Yyr?!5*lnE_UEy#q-K z62SzFC|!6m(@R|geeWLJIyl}p@Q4h44DYI^hE(bhQ-k<&{sm$!-wOx#4^LjYdvd&c z^YHATk?}aZzfvwZFIkN58F8ck6a(XnPJ(_15P#egkI`VD=tv$$vgOOku(FoKS9ZjW z4zjngk}i78zG2Dpxq<~Qj7e};s;zlRfg}DtYgmw489s0MbteHbRb(K^ zG-8>-D|+w|&nBREiE}~$bU|{Lg^{~dv@lyT7k-~vqq@_D>qfNz-Kcn80E(5^3w-Dj zMixm{7Dlo%U}P+kQ$k?7dh%kVD`8TYrw8)FeK#K; zsQh%XwL%{)SfO0#r;zgGWr@iPH>vmJ1(cu0@g*QHxcqeK$crgI{@TU-%8I&f9s*Aw1ZrE)Hw3ui33PWb8ibiO@Di=UK!;5p0s6<_x*|H zxVdYnm3`=nlgO4Gp`ny6jcu%vy9``$Fr!J3m*%FG*O8j_cG!v&QJNt!zZ`}59Bf63&54ete@kU@ts{x}<=1BMn!WR^xEGh}F`v_@wm z^|XeYS$pR-*2yf9PG*U8GTe9b0fNf1mQE+LxkC4I;<6_!y*m!v?i^S;X)tRC|$!SLu*Q%YBQ-8#5LR{MSu7U z1cTpVD*XnQ*8r;pJ%7`WZbJl=K2-*DLpjxu4cZeMy2hySc~hdP!PbZJsv*{UW<%d- z+Sq#0sqRN3_0?n^dH>PV6TbBO(8PQDgcU zvpDWEJ8RJaWES6S^psw;)5oErr9JE?b91ps7Rutt);_no9~i6nhXBk3D(5D3nxFbOF7z;>C{aN(^Q)sT%jV3rSbDtTlA<(tzE*M$PkFC$E+LatJzk ztw2R<*~dGpiqs3S;p`GWI0LG}`MUr+>+WieF|+4?%bKsP4(+y1I7rnaTM2==dkxDeKd>pSyl_ zvi~1%jy1shg4ajBsBTU+d3|Ie_P|?05O`Jn@hEd&1^tkivtfKLJWuYyMI$@a569~D zCZlM1#9=}r(iZ+b-gQ7sB3j<+MI)BbFBa@EapFt(e&Oj+ z5B0&Ak8XjjeLgPn*HN3mj&~NIjcqjy)som7N?&ijk<8p45|?f#ohD@}hkmyiAY@(^j0ks{glY1fCh@R-NJw^?)MJZxUbON(wIB<3 z8VtM9DTVqqvrzVy|3=5L#Jv{LHYMC?BD^hF4|kHCCJ{83u+!wgHZ`uoohBmM63<@h zPP6m1=_^ti9O;^~y}dMl(bz@v93%}kPd^gx-C=(v2h9UVpzn=QQhiCf;W>Mh)EQ>v zy%i-P91&y<#GKABcCr8Fv~T}QxMbE*rBHI&y2_E)D{ zUV&zgZ4RaxR9b5^wl!4O3v6{N;#)g z%1{m_9UievL|VUzButWRB8dm}gCJ9FkexcIM|PV?F^a>ai9gXb7!AQDf|=++mj49h zY_eGiHj#*mR<3%x+D#-8D-G7m+`+sYe^L4sHZZ{^5~0vqcZT`UAo?lkPS2!Wtw|*+ABnoLT_Yu`d3x>G_Hk(0F+jqrOPTAL=vER)Yf_xa2cu5H zyLRHtJ;O9l4`8%vWPO;R&JXKT8!v0#rK5HU@DkoFMx?ba+e>(LG(2@@$T(ByckFZ* z@|sxNPlQgwyS1Fu{EVOY<)KwT!mInFKKqrkyANKw?>Uic1hGCXKg$2AS#HkFC!e*b zWlkjRRZpyhckMdn_@^H|&&~`@4&z_3Y>hUeyDP?ortz~E?=E#P*%zpJo)hn`?RPLB z-qlfAoag?h)H@wWHeGEENUrmMdv%^9+s=azAZ!gSERk_3HvylZRW?@Q8ZF zq{J_k;}Xj3A~CL8rKyg9FZ)Pdsl{Z*jAZb$kQ5hgy!_LVVoY-IYy|rNub_V?rCk|r zW*_GzZesZDF^5P2`}%mV&O1j#;9IWATM$IxFb1MWn_uPd^eYhQSTw;COOCimCV{ z#a%7CMq;Y95mWI=3j6j6Cn+ow~ zeN@#@_UQ>ZltE0UQ>l7z!gjiM7TFnq&rE*CISVxgSewQGPE5vrHy;qL9@q|p!zDbe z(KRn7)1G|B@CaMLw+3k@ITqSS9VBUdq3qgL-gXzt>dZa-40FN*F`2%Kq{m0r?n2o$ zD8D5Qbz-u$rNCHB#=c$TzGVQKO&A97S!Tv#PJNC5@U)QE)bGL+IW$ZBN#0M|)E{8C zt}x(&Z>r5?RgBf-PxH^y>aRyC ztqY}0L)m`Y)BJH82lN?$`n_?cZjZ9FpY_*!>ZnEB#IYH^i#GCCJsnVgIif{U$;j+q z*jn$=u++7Q6mnscO#DQ~#~}hOtf`>_l8KMWK!1+KoQ=N@nGo++u{H6CXyQkNpgK%< z^%Fl4HwT~ilATgC@grhT@6omspGyvad;wu?e(t%KVtmtOP)NS8Hu42hX_{X|gFDC- z%%y@GQg={K4>m}5V4hw>55VD^e0!fPqJp%wYmc?zUKZfJK#B+=N|3EEbi=tpB7%(( z5mXPSB?y!s{8%eM5cD~_T6)3G_PoQu_w-moP)`sx?tA(^>))6gXgoov4hmF`w$Z)6 z>_fe~sJa(G5T=+QfJzl=P>klFBM6Pl(blLOZ8>!W!IYzM>aecCax{JAPoE&TaAwJ z(OofBldtiD65i_>*MD%bjMsg%zM!^~bw#wo-{zm|O&0@_r-vB_e*ggnbnc>7m{bzc z$l!9WqyAv5(sw;agaKOGM+^dI9W{v_dR2`SO@Kf&a&Zzts1HYKEtniVv4Oy=ej_<< zWFHl|vNa%)F<#mlP=gvoYry5Qh9+aYldS;}ZK#jdonQ?}L{GdswVvjZJyW&@M8rYh z)y8W8lR0b}sTJO{%wdDn3g+oY(j}GNBwTXhi$6AvB?}uQSujtpfetg~+k5DM>OMA3 zvakVN8^|@?=g>g15a-Z*=-|f_NftInvQR&sQtF@+se0&N^E1qO19c1Yh&s5s4mPv%)FG`~;OaacTI!(Lg`p0oHHS%v9k}8J{mCd~=b?29 zsh8*_L6$qr1#lqV!9(f6@hzYHE!1psTATctTrQXviZbdqR!<)2>V+nUGLfJGSsIEo z>Mix`VcRB+ker{Wm{9*RQood#67!S07Z+y_FYZ5jc=F`z!Pif|@$T8TapSLHrZ}|9sUpF0a_FZs(#>eH$>el4NTwHSA36cr2SAwZ0 z{&G9RISmaCWM;lQ&Tn2-Z!|N}HaIw3upG4>j&OR9+ZnVR3!)V2+P;mBx&1eFLgNH9 zB4Pc;1l{9d%#WzDKKrjHKDD=WM<11`AB@#4X&81&Ij@+Ldhw{;Jcw6QkR!MBW0HrJ zhgbdUD0w?7C)oPx9I3m*%xme8@;sAoW(gRQ+sud-YqweW%ud{^I9tT&n%1YgTc<=1JI~PyiIh%}6uO7FXWaV2uuMODD zFw#XyYjJLyWw$R7YiW0X*D^&UsAHYo%H^9`gn5Ol>rT3vC1UH6HnUvTQXKUGn_0wv zrS3k&X2zLk9)T&VBe98?!<|7Iy)zH?$4=*EUQ?5VI7Ae)XNU2u!f<9hQ1?1Mr z)wZ)5Yh8aJ?N~^UH-KUn2g(cAHsFd+q09CN+4To9`Y?cy=MEw3pYh{WHWu{YnNao12-t>ZGb!3nzJW#Wi}+v! zIQpxeS8mmMvJ~;k4BIbHUY+9x2p@_gg{K=w;x|Bi+wNGMWA8;P6Ye*p+8eI#Uv@>5 zLc<-H*DI$Gv?SiqPpOtEnMv+bjl?r?3|p>b`wUTE7y_r6hrIH2xP{rR|t(E)q=azUBbl0AjXV-|n# ziRS=-q3wL1o}p9_n}|wXXwNXRCmob|__&?%l$n0P5ib%6DB!-@UB)`lgMXB*;zfeM zPSo$2*zOajR-?*kiX>hnBG4oX;hk)eNW_y5(q@w)$CFCK`R$XN-SJ4-V+m*%1$&92iP^7K9Q*8@sRd_>6O6K??LqDWQGPgC7Inj_xb%wg$ryd+} zo$o^jG+K|*YJGL%)=Jg7%l5XW(Ymj05?U0a>uma9YaI>QAfMJxXF>Nrc| z>PDl*1dwRqAW`A6y=#84MAePUh!45y#tm1H>2!u$i^Y5Wk+==YzG2yip3jDTq^?S% z4xd13xaCSg{|sitq3zC3@nOEnEG+U^9>3zT##${W^Gd8?9+gR@sRb;~AOdJH`*TS7 zLkXv7xAmlBN&RHABT5c^P`<$ShRRNny&iC^F2!ESkxcz)q`t}t*0;|NB|VfL=%Y7+ zJcAY}m*bS&!^t+kOx;WNm z)n=sv(J%Jpg{*xodP>CyDe-O|{it}WO7M`rR?P9Xf=a&C8iJ79VYaH2(U-YZArsqg zRSENkY*mTJ^+QCqRiz%>i3y97{4zGPop`IlM7JTQ?a%uJr%mrY@c75j;4A4?C1bmh zS+cDv5l@%0RpoMvVib$FDjDF7+#F_6yT!X~w540~Q=qYn=@y^zzHfzt`hnmc!TsZT zOt9~l^Zh}fVjp+ldVWgEj4s$4s<<9hl{ME(Gx zw|I#8)He((1kltG`sxD0cx5URoY( zekLrR<=2Ck3a%n3?26B|IJY7swaG?JH$NpCa_~iiPdm||ex6-*{L^{37_6gG$)J%1 z8*c~6pl)DbU!%q~Rx&t0wO{^#WN=%xlI<)u*Izr>*u3WlXF$~i`z}Pwp}8E3ormAI z?0h9==iyP(bUZ$R0f4Sk^#IUAGXh!S?S!|R@@h`KRJ)`+?J{%d3r2@F1tOz8Mc-bE zN%E7frpm8#sWQ})WhH`?`sWM&OVNw!=48YEQe>hX{5VdU$S)(`A1T%swcalvq)GjN z8Qs3vrVq+#uK{oNA*RG%M{Qy#o+|)gwsNaL6sT3=8_6W;NdhIF4V!B# z`>AkNvgII=;5#^Z_~oDn?MU-v%RvzvwHcZOG@@rbsmKs62N5U|1no|;93)~Wy-$-? zU&0EqGyRMN%R$73B%VEP6>8Kp@oP@tC!|)6kA!q4(`C3&D>pwiS~-D7Ef4aU8D zg%rH;QIM7W%}=4=P5)CJ3gW7&oVv?Y7w)rO>qeK?b45!YV2}znJ^=Cq3KeYgQ=?!L z3}~PJkOK;?G|-3jdhk=Jl$@WE_0)XkV=qP2&qEB+lv8`rUv*fx3d$}lG`8se3m`S5 zFTLn*g-Fj96gZQA?;%g)OA$?FA%Hw92YK|*_(h7`50L0@-RC=7cPZj_K|aO*ma?M1 z`P3rqp6HJ|4If#0q1O&WFH)`PBhjBvQG(=8zlMJ`(g4#ZI^?#YG2Q2CPPG-^yr(l0|Vn1D6U7sezXOHlXjrQ$xg=`IFoB~@NqR^9E zC$YH1q4|sF00p_KaN9EN?ml?;>=V{wF!4?hThRImZjZslkHo5=6OMvHTDzNg188Pm zL&5DanD~L%#G1BDtOrIq3RaIn-lH2RDH@IzyDd|~I7G=Bz7totha;`{t;Z~LJyyr= zF_^Hv&&^l`<%#F;+A<~nFrEhzzzzso%OOl|pX|0wDQkN$P3mrjk9`%Gxv}X9 zGSXk49BnH|0o&PwpxdD~vy|azGecrye!rf5M}mgFQ1Z#3YNM%Wd9A9 zv+@q>kLu;v6Flgn%79l4XoZ2sjP%sEu8k9%Q^%&Ie#qKc} zdBT}c71FTB;5sy*WAzw>Z(#X5uv*$s5I!{nx{D4?Kx5QuZyg#72v0X2#?zJe?M^* z%pS#h*64@g1aXLNh}&Z@dW}91WeI1ar|?S5qCZRIkI`2xfI8a+QWA91A7i;a24f?8 z3{>V}pDr|vKyn6g8lx;|-p9f`D=IOl3#QtK` z=Sk$?`NW@;+0}~;r$}ZuH8Q&jlms#d^6kCY5ET(lZ7aF?l=~d-1yVW32S|QMk#cTo zlyj_N)BccJdPMocPXh(yhFW4&4~}Y`HY2OBBQ}(6PwC`_GgO3ao*tOdhTPnO4{Tt} zOhGrmi}nhW`RtVW94|NAKvJY1>wk)TkRi2q@7=pI(`DUFYD(>0#`<|kZn)_n(f((c zZwBu&-bQ>#JIcN{^KnjYII~#1CpYBEJ6(@KJ#W~>l3=m`!Ql+_`={oOf;|S;p(P!w zPBQ$oMILk!owWa8_?Zj?Aq|8*2G^kwRJ7ZA(g6mv)hRXyH7MiaQ_&UFV{lzr5*j?+ zT>)$XhnetEk_M!Qty~~FfBpMN{c=Ga`{><^ciz1?xp(&P$)V(jU-DAh+v?@XEsoSS z|5gxs(3*`8N9yNW8jn;ParOURkZ~|m^CHCGx~hJ*p_x{>amcVaYCkMx$`>}P_zuaY zqo)3Iq_*M#LA*@|S>hT-I0G2%mdxtZQw$zAH8>)7>YvBzWw8hP1WxYDqn7g`UgSZN z-PWIJ8dnDRACHQ6P3PCuet$x_+iT`xQ^%0xhbhM0$o?SfwSF_8^^5b8h$?3(+i#3)W%~?? z{mkQ!ZpqJcd-Y9mI1^@(l080mcKr@#NU=9FioFW%ysbFaI0@tI6ji#ZRR=pq zpl3jU7Y{R^K!;&Q3Ow_8`7m1yT`|l+#HdR^>sd6wOvPUfQw7M=_#b8>Q0+yGGbc~m zix_7{p5{f2vWP)N@H2grO-I{JdzPS!#?nx_9I zBlcYd)}dvB__Sjozuws510sOlEi;T9`!B~oGCs6a5F77T7qE`Dv{ZZ)PK<@p^!tsP zm>w)^6eu~-A_rao6He+QkNV+ARV^RjR}gA#`Sz7P56REOi7Z$B&A5*90h{}9RqDH9 zy?1VC6B0*ieB;Ol)V1Z9C7G-Gr%~~$>a3w=x1L;oLw{JixPS4qZo>y7CMm+qy%Qa2 z?qo4w=}3SC2Tc*&A%~ojMYbFA-GHS7W2=Vm6n#q3*J;^LK+oQ^chU158Gra0uyjPA zO1N~Ct7|jSn~1ISp@)w0=qPrO1!Xs2>4?~r)YG&h+;nfa!0bYvO=<@DNJ?i=UH*$S zgEONUd{KJ3!D$q=6a{lg@gE;Wxqgc?B>tNj@gE;W>5RjV9r^ZN>4pj;XRfi{Y|9G} z?}a(3=Hmk>H>h)@<}**14ysTDsT0t8@TA>OdEJjqa2r`)9L~IVvxooAPqU zoa#fvRg~^$+1y>oGM_SSbRtOY$!C7y8;>uXHV_o>QqD3y>LAmmu?W&uZa!;X>W=0e z;b)kZ0TJYC8u6u-GnA`{bt1^QMeA4ux#If!i6tjG1ULnt@m6XUKg&(#Go~y43POa? zuCG-48U45gCe!#(572?295nhN!MVVk20ba+gaTLL-s?HZs^RvUBO%lbqkN5R=9jug z{_Y7P*VKgV6g#oKYD~2vY!A3NaBo*fUZ$43n4j@KPQD_% z?P4{H$crmKopOW0lNYF32~rp?1o@^VFPh&N^1>7m^p3ZXzHkxQ-$5`D+A=L9nT;hd zw0Pr-K3V;|dSS8^C#w<)qtB6n!i0+30XwrNF$LNkWifzBwLR)&FpfgDhhp3TwmZR! zUKB3RJkd$+A4lrV+JJuR?)mvJEc0a_>94C-C$ESi9lk{cm-!F2vLGxW@TU-%8S+W^ zxVaPh947vTt^Q-IUL7>tuRM8j|E-6UYu$fxKO3c8dDJ-a$lWcTh#;kEJ`_osqrN>> zuMF~jPujx#5}WpZ`}pIXe+I5NiENoeXc~`j4Wb#i;vAIGGabuFe(JgYZqh|sCj%-k$j8`5 z=_@^KhSroi>qTE*yHpF{ngusvw7wt^41SBL^cz@SLtkzw=rEQ3H3_iMr^;Y%D5n~- zK|5ljFZ!UUMdQsoLuJTWV6JKk)`#+{A=Z0lQ&<`<(KVSPpZd{AeKna!-hX&;_W1nl z-u=55X9w#n_S}^#{NB%{H-hAbZnN^8oPMbE6R~oLb*U+Pc9(u1nwW2|^6^-GvC){? z)+&bzk<7R5k{fcv;CD2QtJM96_s-tCbMf?@v&s1MYQr!q6)oOSAB|MH{bI>5L_>D3Yq?b+xYPJt4ck%>;C1^!UJwqIw&4rlMNFvLLHCnTPGTpWfNY5NvN+b0@g z%ZZC_2_}VZUYfZHU~#!4+KupIo)`mtpU{7MV*F)WWxPQ8z6H|v$)W{S-Y%TTZ&7je z;tk);OFI_USWZ`!A=d%qwpWtQ%*{0W`OwMz<_;iNT}gTab@N(86xm)$s(;MKK-qRP z%g)V}qziBvPpqKjck_B`yLm0FkT5*`XkLj-OeR%lO4rHDT%BXQ9psC;L7X+R=6UWy z(uMtU2s(Mei*btcU+t_aQZK~$OF8n1B?Ro7@{oEH)psFy4XvHhH*;iv-{SSfG5hw7 z?4vU5yy(IB1Qv9M=KK@zawxnNon|~$UwSV#F*a}HR_YVn4fXwzV(-2nUJ)&iaCN@~5tC~2g;KPlkt5UxV?MeC zw)XkB#9v2k0=wwCcb}Zy(QDA#_aDAB8Gi<7V;lJ}y-zEBz4=Bmb9+c!;@L2RTA4^F zBW?(Hngl!QLr%BT9H~#=>GTD6_IuHj$#r$@=6=IDKJYGt!0}b#PLnbf?KE)?(Qp$% zCgj<-Y^O<>Cf{ii59^17WT#0zwn)9(%p9`TR=U$**iB8%!|fai%1!TZ!ks3<+k*9Q zC)sHdL6ctSf86`ePPo%VL|fun+Ias_snKK#l#`SO^J~*Jf7$D&7f2esFw$W0-pzA~ zH)RR*z410sUy{B=@?P`wnv!&0DftcMMXabKec^7rUGn+~-Uwa*LfzO4{62tgK$l3U zyEH;w!GKn1n{R-NkXO3WvR>P~s0|$QikA9H^re-@b`=Fnb3==-MCZPn3k!3pa`>dN zoZ3qCB~S5sLwRlU0`Y13B4teT_h@r&W0iAg)VNVG0K5PXLD8 zhBk4ZtNCGzT9-|><$y?!U-I~BspgYuHNG~S>%H<|1MmfI#$0w z$|!d84==G){_4BWd>flPQs2uJDA$-|Gu!lC!!-|zvf8vjy6TZ#MoWAhx-q$W{rziU za>aDDf!CG33`aM&e0*_o_V8C8-8*}H^1$@GyfZ(^8-u_pg6?H&5}M-%R>V#7^`Uf+ zo+i)TRK|~;Gu7{m)vD6kK02-SR;F4n@7{Q7X4XTNx3e_0f!ozzj8wV_4Pjl6m)5$I z`b6BHF)d82-$W85qQ8DPP?`}8x7$PzOQb%^Nj8y`p&U#)d{~fKc|=;ji6l&tZ6b*W z^@AWI;ml5*)FX>nw1uyW>x^I%!A$fZyFvnargsH+&>EH37r`bHQPIj(Z&$mCBx2~?mrbQ_0t0wt$INpCaCkna^>hK zPl{?Y4|gH&YKajkPgHClqFvT;QqTi}zmCM5)LqCs!#yCDj+2@m5FYN7=hatKP^GT= zu+&_AZC}3TncC#Nx^+pm%;8ITpI8a+ax=NA1LP2xL;riVY17KPnCN{CM8rRRslR zc1{sZGz?dLsT`M3W*1e0yG)Q6_CG0DNR^RSQl zos^JexS2(f<~8-(V-Ar5_Vw{zop+9g#z#Dbf#7kt$C4A6`c`>O{ozQZZ$d|40(!7V z^y)kqw$b6%QW04_nG-4WWlZKC+9FM=oy-`zAaYOr&LJjq0?zdHX!Vmh!YoyxxT~AY ziTF6^WX^4I)W}s&=7_A+d9dANCWV;Sybfsv)L8sPCkK#+p z{7@o&lzDpHP*#Y^u0b)GJ`mS~6HpCI&S&o|&>4UqG7(GhjWSJ_#CVdmw>9pTIlo?oFOt!+L1mKpEk2=oK_(Iv4CI)+2*ws0&93;^{#zHk0q9{f4B)fOj0ecbRy-N%HTAnNMGnmpf0Fl; zHuYbVPLB6rAQjvCOzT8FKlyA$rj@1wCY=g7e<(t(+BJQ~Pcx0-$74Y}5Ws6bFmI~OWK|5zZnEB#ETWk{@eIkFzNqtM2n=7!PmdA zwcev)*K1=rH5pA|VD$hdGCoHH___3du6(lj*eT7NuuTU{Sbt{gu@&;z79SQ*_x1wG&<9X?q^1!*hS9&5!-I(r6NExmBy7@IJsDi_&JFm&=Y1MCSJTN!Wg2HrZjL4_PD7Pl^^WR^LgDfv z;@kIei!)%RLJiv)3ht`ybYwwK>_eL(I#21U$=7&63Gelc>pxtULFEpw`_OwqZ71uB z(1X9tKara*2IQxaFyr74lmJ5`Gnx{HKNzd@UC$9=fR^?VgTPrwO`<2B8vrsi^0$C8 zd^l2T!Q|+P4Fq1Lmv?tiI&R|`C0YX#8KYYR+K`4O#_cA6AX$hliQqyuGjfuJHIgitr`JG-4R1Jo7o0h&`&c{4!kU{lc`r1QEco$6 zl7+RAEYy#ulsf3Im-W+)N zuUpt)>cIVbPaQyUN9-d3b-=|P8%G^v7kqB})M4)GI_6vzfTIrUv~Ho0w7LVDdT6PG zW*3G!oYovB0d`n(Ut51N3fXyRokHp*dP$Jw4s>^zTRxLpsM+MSHj@)D4U@|S(?U^3 z{l@CZ!<@djv&p$FLl?-s09zW0H0mw&>;X*gCT2+vQt|Qbiu#w4`lZB_Se)Fw(0#!# z?mv2X^5pEn*H6Cj?%Cs~CvTrUet7oa-Fns07llyNyoLuYv zGkMPCc=4#+JjiUEEH7uf?6h?EzmAf(qjKETSLaBBHl&FS4J%LQ0h?I@hWyw!GgRYq zsLd>8_~mY9$jn1*W(gCAZDxrF_Jc*bnWY}xb?=G$y0x2IgSi2l8AiJQ1@^93 zou=fQS%i6otLsj>nI&TDxMd8bzwQj$7?pf8i}lNpnZGfcrenS zt&IllMd|s>5R$Ol7p)778_3aOdc`Nht}W0S$briDR9%CxmnrU?hxlH#H*!9$=wkS+$P$HhS5Q~cEsQH_*pu0PNkDb?0U zsaExZK3-4-BKrUGJn>P8}T=BUkubANX2l`dpA82hQapCDlsrv`%x9~=#9mwCdPumLj6^7tY& z9h%LgXg@g?9eQ7%mc5pMsln&3dZM{i>&a3?G}EBLBf`zC23E)V`2`+ zmxmY`9*QW1hWqt|E~ji~cn;W?4cDYP)kr)O$FSv!sLv4dj??nbLl!@xzqkCuNL4Hr zz(j;c+r4=NS|h0s-#9a|G^BlecJI!^yXR+<@#Kp7^KlKM1NQdif-bb;#R}ggMXxQLb^y`>~!F%-E2sA;zc3?O`=6&CtD;E@svKu&@&_w>K|a~ z*oha3h%L!H9L|gFrV}}X98%hlx;(zZP-j8=rY#$!%QH_u8nE~VxA2Ss@;db4v#!k1 z_eI(udELgy>-d$8I&+|2+M5!oH`x(0KJ*X>GEd~1%$RcKgEan=EQ1|$22~! zgObHpJkVIHd=5-n$ zk9mn`V@t6(<8*psMoC%5l9~Ia3GTNzogh`i3D!s<(|%y@}3x zee7HIPY?}is3%~*RfFLHFewM)17If_+&DQ{H#++10RUIMppO<*ZNAyKXsJ@HI?J>d|$V`%c8RKy~NqspG+>?!5R`et|`z!C4;)FI@5YVo~!< zZSr1CwJC0DS~p-q9vXiw4SiC+p6rj;LZ6usyN zrK$4kT&fIoO7V%_(FjWY^9BDk=tXsNvSEJ>GVuz29G%I?kALd>BgOin*82s7G}~(U zj$1jQFGx}!jQM;Gc(V^NCH^{U6GQ3Mh#cW~dI9Fy$knE;$3`+qdXhkiXTw_2$QaU_ z!sQ?Vj4p3E=s`R3!=Y?Bz}N_Bg{WUk;iEQ#n1DyoA|+f7B2Xj<+MQ%MNW>5ss+TbF zGrOvAIf&Sh#IvoJ1Lk!GTcDMT1y6h=q%)a)hPy>tIrH=*@d6uMLQO`Gg0`gKjbB0P zdJArm;LSX}1_gmLG5Pkci`)Vg18!Xxx$T^uD~{U-FOUj0J^=CqiWF>HqhJ#ZsHH!Y zC;Vtb(Vx!4s;foUT-Q71vnA`P`OL>%X|P4*)aL1d`EcLOhlQ)4>}J=(;-2l285Umj zH^Za{z(s0MOjzx&G*~pg6melI6`$Q#hU%t?;b(jxkS4-Q5p_4mp6G9(DXM*M=Hr~` zZ)=fuPxJ@K80oJ^mR{(!!_bRVYx+p^=Tj6Q`P+h`aqw%GuqV;fu?Iu=Iv>b~v0Oa-dL3vQ7%&OE(_5Cf!CT!?&+!3C-g z-@2Z0+f8|FeMp5tCdF=iRvZQysol0l?Z)ac7!UFwxuDjy)KlZC;jrjfUn6?*o;#w& z(}a(Xtz6JtOB!y&-9J5`<9W>xJGtNz6^(CoEwlm2`pG^T%1>2ZE;vUJdC6&t20u{j ziR==TkE7kVPCMx3f=gUJKDQ*<{7l(2?*$|mTtqn8)sQczHP5tiL47AH6(xMRpzSee z?s)B@#OU<)Pv5DYFk@;M(^ zuE(JMP{r;s7zE@OIf+lO@#2UTa$dEaf2 zlz>pGmyCWMLu&Gi3%7k(^cyInq<$imKen?p;-ZmM2swk@J}k1gpF^^Qv(Zy{)zHzO zC17YYcYrdmoqLm@lm4#7?J*b|*-IN`9uC`YJ1QDEE_0FS!9UVO=^}xp zg{{UTfsE)$PsVFvc@i%Yk-Sh8!pu5UMm&iNETd;<&QVm<#JhV~q1ZhJqZbi%ONgG* zev#lw4Hu`RK94Uov<>O>^H8qLZb;Ynh~-J-c=}|+C;p_&j*pV8*l=+QWp?_X>QEB5 z1)NirEnyL%cBG>MAOoS?;Q-R&lvK{~0g@k5sGQqBomXtyAJRTOB8~aF@-6vvZBRWp zs&*P#h0SM1R&KaBg+U88|O&=Bf3J7})&O%E%R-I({X^T97ozjxW@XRI~^%c}} zZx#wcMdukOJ6zJ*%2vn29Qq(b5%yG^RhEPX=XX~Cd%&@Xk3gWjwUxa&yg}Og?<4ig z1$FGBcQ4*~_u@o%e1CFwAo<~!yqxy7dUPPOwp#o~9D( ze;%us#UAJrIJqy6TF#4jkq>$9aod4P)3|creLO1OHJx8q`x&>~?TnalC#<%PQ$x19 z1gGn3&fF=zAW&=>3<%#HZo5kvf7#m|GW9UqUBb+v+g;+3{jia5cd3VWVnsiJ4!hkk zbc=(wJ4}v|M54IdB{rU>%SnPir#DPBWV?%4vUuIy$+x>ioP`5b+vQBSwe%IO|8Z@1yMQK1tskE@|Ay^#xHyGkKmAV{v)z)P=l1FwzhD*Tbuf#R?D4s?l`t+& zp<++}Qyq5(_@-=vm!!c;ckM_ApMe;p{4mGir$B&*mP)>w%wTHehgsh$0Tg)V@$zBr zW9YJ3py#2k<*jGY%ZiETV`@su)A&Hzi5RC&p0(MCzrR_oPgmK$?%C%HYCBmA*=Gj#mnX5#h`pF`7Fs5V zPdgs+ubQ5t)AU56;f3S{zLQlt6l&0Tr)Wq~)S))M7i55BV z0vJJ(UpRg^QdP?b_!Wd&BRJYtYa}Uc%6Uiq&A5*90h{}9RqDH9y?5^A6B0*ieB=2I zsB6oJOXj8OpGL*2s3iDwt|I`i>Ym_x0nJz z!Se#C=Hp8s{h%T>-_)r2goE0rPoz0O1gVpPdho<`zn=@BzMPNhA{M2b2yzO`%NJ8q zUhY+$+;{V#;VMeIS*Ec`_g}aQIWK}-A=0!0*Ab9o;ULp$)>Z_$vXz@w1i7-1$^1-M z70EY)BFJ^0FRe6}+lwGq0HBbMb0SDxJGp1BJrN|=-%l(#(ILPo0FBo`AcLH4cxOxX zS1^wb?fOc!pV5z7U{bI(s(1<+{js1{Gts~7%2Hmm&!;Qdr5sXR<{S(-e)C4{8+rhvM%Q0 z2?nr8V>hPPOL~%N2_#bhn2fr<7Hx>qx^9`Fe)X7wEN#ulphXQ^ zW-w1_{h|j(y91Y*h^q_;_T7+WCJ}R&xXg$c%dll8;w(cvf2d`KS#UZ9S{Z(h8F3i( z>rMhRMPeEA_@luqdhijohHwcbcq~v zow@PFlb1qP>v*H`@qx-uXH!dFW?Z1)$;%p(7j7A>Cohf5Pgl77bnVECto(Enkr!8f zy5iDHM_zQjUy8f{fe<;bziG*f<~N4CFhvBt<1L^sGwy9Se+R(CJhWw6NHQBsV1UhE z^qJ}B)eDoYI5QPw3GwOTiaAGm3NsHCw*z)&Ph#fBZ4Z_b;*u~qtoX?GP>egkc28s$ zjcu2l)=1m`aird?4d}P-o}UlHGGF$=|GIj0@`@<<;agO28GT~`^IJz)Lf}s!Fr#tX zlmC17@!7pQ5AU9zO~#We$IVS^nMJAp7^_zY4fiWgp4@-y;pAHPpWM$zX;&UKj=VU~ z0i0<*E&>Gl?Xh}gkoSAimO+8+*T425OYY|bSDZw)Jebflj*T9o8MxwLMuSe%l%LIH zIPlsO%Uq5lNzTv}C!#b%VtzMh#YsfoWv)1qtzQjXaUxPP$omIdahR-T)<_%UT9Gfb zj2)(>MjSJIeGguurgy*;jTJIltPX9eRLFn?hkSdVjH5cOnN!HjxK~el;Lu1R6NVN^ zWM)PpGh}F`v_^jquctNKLGaE4I+>XFMS7TQBP^$rnIWCb4C!RJ@8$ypm1WJGPG-g> zQl8dqFsU1((t%;{KX`RfLNl#s) zbWO1ht+^>X?m}F{jTrsmFAxlVi>c`Q?>^bHyk;A3@IVwJ=rEQ3H37phmlXwbLpjxu z4cZeMy0WVAX<34Eu=Sz5YKZlo*{oY9M3R#m6?p1LBlXqf#Jd0R;_UJH*}eOBFV2Q_ z7CTa(zS9Nk9i2tIbM|C%U0u8R%w&9hd`z@qdG5*;epB4i+d`^yRedL?A1Xa2Rt~W) zHD%B4((gkv^6gbV9;+`l8dKX^Lv9$tc{gsU(rye&1&uen-oI#=`%lkqtng239`-?vw-2o9Isof}~DId$t?>;%Z zqsjK$_aDAB8GmMER_*k0sAy>q`^nsV2=5gAAl|AHu%+ML$zeZ$#K=Znm$h&0#9LL$ zXw(f1n3~;5%+1C7mZ_8HZma7f{Ym1PEBgPMZOc0)-KrAi4cV#^kL!nsY^zE=xY+%x z)sj>0{3YM2FiJkue$p<|j{AF{c&myqvS-5gO^z!UM*elgA^8O0?!&NSp(Rm6*B zZa#!}ihc-06-fHxwKS6MWpUX1PSQ6+lD?Ud^bI%&1n2Ngnm1DPxsG014g(;ud3uea zZ(D32b;BF!FC|^sUOHW!+|0@P<`tJr-ta>_Z=~(>L#!n;&>=?KCmP~DCoT^$=;oyh ziRxi-MgJh+KcMf6rwE@IyKY`{r0<&}eP7iJy8dr#RCzmhBENb0(nk<_zMIz;vEy}Z z$n8Pw8eB;lyNrL^gzM(Db_Jxr@Que8ch*}Akj>B3OfS&QYjYY=WOpU$8kCimLsB#I zeZQ?ac^MDCo7Yp@&1-IjgyHE&^Gal5GAX~dl>MX^j1Hr(ZCR`D`r4srt6r!8nJ^i-yH#z3@51r0T)wP>3rs+QD?aRyHFj;Z>VC za;N@wq<)dRhQM}C5q#>Wo{moBLHg+UCtRlf>D$j;zdA8=eROmD{*B`=czxuH>gHsV z*GDEoAG{@GgICobk23dF&=1LRTW4UrneB1I0Zhmu5B0;bdcDagS{`wjkjV6jmbst$ z9vjXuT9BpgKfHJL-kpo5?`VPQ>D5-TrM^E>>|F=MB%cW#8bf&uL_Yifq7&;wATt9o3`cD>U!&%qTf{Qe9P z!S_8T`G| zHm@63qOVaUx+bOnts~&35}gW;5b!|TyqWf?^x9RTuX&Z|HXBuxn!For^Ry~&0g~o9_@qJp;GaMu7uv*quI7g=Y6UjF%6c2Vx-bdXb9_iytLMx)Ti?PjA`H+!6uR*5&iYUeX+{a zue94l*r;?Y*hEr>B21e3AUAwikXiYUarK)>!X()yl6X)*2r|_M*{PFyWVeZU25^Wi zf|LQgiqIq0`62i4gs5PO=ZABpzkaL6ng6>K4ut9%gZl)MWh zTwNI9Y6S=dRH1Bw*Jh%g4-0qG>S9`P$O!%*m8g6q>c-CZf0spF)m2SoYxpoDi8!BW=Kl`Ze5Tq zbNCY8Csx9{a2<2}(~q8KXNGL?u00=l@vbc$2;yBGO!ft8oZ`g03;P`migy9hEsxkc z79{Jn#(?BH54cwc@VoqrA3)d|T3mjH>w8O)XYr8&s4`#h^lhq6CeK&Vw+ziJTc&Yi zz{sKPN9EbXBjgoh6VmVn>!N@UhM$NtfZ3$P!O+m~qtZ~%kJmqTLo?@ElM=sFj!P)B zi^RAt(?=N&LFliM7dX7^<9ekQlNmFv!3(NC9Vx~n2Y-QJAK!`u799NYF%;~pHm|AQ z9&?Bku&b!F_G(M=$kk{fh4sgHW?o3iY^@k&sz6l+H3FyHd(W~n?}GSnvd`s)`%YJc*Urm%n@d(3dLRBWKP7# zLFa2`lOmm3J((l2Qs+VPWx@e>aO8!gn^a@*fr`!wx@-}tu@**+C0GoDrR4PW2{0+9 z;*%72wd?|ksTM{|RY6if@gMp2PCH}@|Mt0~b)d_REBYFCe+l$a@rMIGKk-AUb6Wl2 zqfQ^SF#4#fq3qKW(iuQZrgNI=+UL!TH*-SVlbr$h%)~XyT%g7P3)2|Di^-}3!qo%Y zVRZq)ZQSU^WHUrcnBu5w#z%XV*QW7>vKw1@Yaz+ZLK5>c%n1+0Wb-~BF?GE5y_vz_ z#AHiLfqP;y?(HJ?Ed$Wl%rJn@GBX|^BU^HBH2PhbIfiD5Kgs(^oBGd>w>}LzQO{34 zTagB3l6z`y$D4XcvdsxBq{5*HxoX#R8GoS&dFZB@$>zye!}f*p<0|`HqOT|LP#Yv? zAaeY4C2IN>3Y+YN>mmkU0Q;H`+M8-KSrvme`P2OKwEF9jO6x)?(@?g5;N$&q6OX9h z8)xeFC_DREf4!%UTEtBpn;^k&;vT_*r~Yz8i=>i~*}t&0-lOD76n5AqZZU;{)dQHw z_&7wM#S7O^0m;P2WT3Pi_HXlOV`1fTBbxXTA*c@1UH!yQ#EoED9(=9Xi6(wT4C+1F zc5ag;Y5VM8pO-92`2wFDeAAMHC6X_gryr3@)BGYD+(FGlp9(I4?f~(5zJ%K#2Ic6% z66p@i(`)Df_$E%iy{8ALAZ_W|V=cK!r(-BfQbfR~2Yxt_h`>C(emJEBL4P}}CkWhZ z&pV9#o?cH7mhO9cH@nOYG@c-E-^~XFDo0!DDl7X?-Vs50f-uJf0rZz5gVPfPP>v>? zTBeuC5d_i2sGCJZ5L`Lh27Hkx4fO;8>Ii~3f$OhxOAz$18G=B52=1+W zkRov3`j24J9oj=3QxpOJtR)GHcux(v{qg60!Zuga$w_6xCfMOk$`)ussIsfxQ9V(h z%g|aJ+eB0W+Nt4{yTY;BPEQu-fox%?gghu+HTfDZDB-=Has3A;%Xr;K>kDc-Syx0W z{Ot<9p!?;JJUz@f_yY(q#Gm|*8<|m=OOX4D`h&4b-}M|3254y?F$kP>)FgVsd9smn zlez)Yh4DH&QXh`gS}-|!VgrF!={NO?gEr2RtpSOQ(X9d0eoktKn%ILmgz02bKLD6D zG4bea?Fu{_>l7%IbEG#P}3uK#-P(kgx;A~Lc$5IQ^yZjPp@p)co zBw6s|i6je4BUz{)PbqcKLf(4nz|9anb!e}66<55VKN29v&m^~+9se3@loK4 zImZRlLQzKj#_Gw#%12MDl&(eA(X%uZY1CWl*~2>F(8k7#&Q||2Qood#63dgj7Z+y_ zFYZ5jc=F`z!Pif|@$T8;zLEOZQSx?FPO$aWIZ}61rFvaZ9K?%CzL_OpNNzLBbmP&R z)xM=$yQn}EjcQ;%9dOCuuX=#YLmdH?s)y3Rl;ibTdoDR-BXoWogT>67FvUHnWKTO5OeA*~|hC zUNw@SeS2hi@Cs?r%;S$FpPQvc$mVB!aFmTzus_f$;>hlb&n>S>lg&YOl$^i&;Lc(Nu9!dv5dJaOG!at2 zYWoANtRyZx{U~++Af0^GFQOev-`Y~cAjQ*=-I`hBaMvejWj_>xK0#Zcuq_AHdzQ89 z=NQ&5^Mo^j>}=D^A~Bi{&1O=xpBxK|-T{|>)dR?_T2GcD0GZ+MMTf@6vnwXY4G=yQ zM+#3jp2TkxzZGJr-BB>71T6Xj8y<=%g@!vYuUF1?#^nfn(KyQaCb>^F63@gjY`Jpa zK0`n4r^YSy;^b!K)BrE4{$Zpl77Or~2#-cUvTdUg?k%q6fl+@xu3>b*-o9K==Cx!` z;qq(8AAI7t1^HvM@W?Em)K_T}QK<{<8AkS`gE9}dixuYAa_J(GfCAyG-KTbUAG~`u z`5(9c`}NV}Q{6=a&+IW@tFc9llWzSZUL*+YM1_!v?Vi4`*{CLXQ?yoz7l{Zoi9&cM zTO<~`XZ566ds;xlxTb>t4$sE73(GMw- z%&m-MPBf%aouNM9*MlRFwy~VHCL68SS2wP$RIO_!Wu6|Y8@cc10|Qkzt~B=8Pqj55 zJ*16ZXSl+oh^u$hYsFr#w*}RWF{w3$i%?gS7o)i9#x1UHTscToxNHwh9jG&0_xaG) zRX1({q$3^WbcSn-#e4mcxVp@5SoWcpLc=~%SEW&hParj11F}WF`JX`)5Zdni6d&fB z%)%lMTA=odPxaPnIhhMmJ>elfu#Ql|DcWrv>DbPjo#JvVgBrB>FcfM@&3Po)8!9_R z_Ikh}xOgwoQZ~G0GxzP(k4EaNoM3(XOe{Sts5PVqujq~7B1mgBzB5uk*UD(5)QDH- zL8%&PVt$?fc+8_!(c@#fU{xB23X!9>!=SHGSN9*@JA3cW#nX4rCganq;+Ycpru@~! zADY!iOqLe?2x4jI;|Pmiw#t>G_hM@^K4en0=ods(EqV$sVUGSRj{A6zXh7AX-_2~u z#j!T4HY*i~ez7kvWbJFwQ!3^{y9p@XsuDb;uN8B=?UJ@Cgxn6ZRi%u+%&iKU*nX=@ zm^WmrN<6L~BC@S2_28nIu+@o(yb#cSm6vW+nCLd-wEe{bt5fLaGw(Q=Rs>IMeZ zNdIq_eWaXZaBaW*LCGNKgiw4#wX@hdCSaLF)FhFZi!PFRGi94f|`5iA?b0IBBB3nyBxO6zhvx?-vl# zOyr5J3~8x6R3D7_d<}TB4>2YFI%*R`@vH;@vz5=&C1qq3=>{en$t3AX0wtd1$@q>N zIY+o0BocfFCl9|I^q?I%kh0|fW1~oL$6t<2eD^CY=I0=`72owo|b|+a55;2rH z9Z@5E2`k9X6paX$gNO}DJR8p1t>T9AX5c44E62}wBUGexCevlONGrECS~-D7Ef1nB z@cs&>kb*Zp3Uac)H4?n7jo__9_P61!@2^yy$NONzeRr$_!5ndFH2WFGbYPv$ubCm7$vs^6Y=c2LkzusJY;_ z$9ME>>T*;4K=LX6x0Dn8ncEsoy!S+Z+-dm8(hI$I7SA;%-o2bXuJ=1ZfwyEz`c*NZMl%2$o1u!=V}uCSIQ1mMQVKu`1|1{201N>={K zMDp9x=7^;+6c>uyV=!TTpPR7?$`jAwwPi~DVLT7YA7J989uT&cqmf+h*=?Cp*7jhU z)ZISv#r6A%mn%D--D5C*dV-Af*C$8Y3R1v!1B{>>xo-~G%us`CEMVL~D+|@V|Ci!ND882h&C{Otm?QJ;& z`y1YX%`Bq6!qt`8V~`bF4cBM!=TU0A9I%;1C|K%lMiBZNry7P{z=))-F5oFJ)vIWLI#*0Y^>@lbdCF`j%AbjN?^ByFHV0?7!GD6>D zaO1jb_^0za>!E38vaUK~_JYCrsg)~x!QfW2cKaeV&vSyojs0R61cUk~Y93^J44Q75 z>>h)WC!7gYa!`2vjc%3b#B{74gYXSH{;nI(QMgvxb`bthhFKS<32L>cd%HMGg{K=2 z#yXgDA=+^;9}#l!6k7oU&WJqEk*f~qr-fKaNJjD8+NYU+bCyM0*n z8wihr&kS_CvQ*aC&g_29sEf}(wZ{XKXK!|m+Su*GB76HllqH;vp290Ji~cMDL)&>R z2{Y+0#?cxbI4iq-SZrjEfyz7_cFlGkB4-e{$6(*y1j}-uladv-vZ@FFNE5k@NFx)` zBGK;6%k!MS0wz`@@gfm{CQ%48dkjX-PAn3{JQWvTM!%amP?aKzRY4d1n5bJq^pu9i zQ)c?LC?F+LpT|c^t|6VdZUN-V?1prGk651MQ3GX_H)c!9?D#0ji49Hm5{xlVuRuvU zbD&;2425kaH$O#1gd5jN4qh5SJurBIRL=1Mk{eQUUjXoeQO*sRD&B6=J1WQx^&r%P zBewG$Guo0>*nDQ>SnOLF>i-#DL_t~xdJO6)?eihsrPIDQ zGw7V$&~*DWgNNmY`YH?RW&Cd_+hfqo8+NfISpUqa(~=L}1&_A^!XAUG(2|Z-CmDX) zA`e=jCarf2KO@|qz35Q)?{us~be{2~<1p2N!BBex?_hf?OG5iE-~k7i@a@jNEwZ&M z>fcA|mka9HNAF&|^X|pTJ^fYvOe#MtC^o+2wXwI=%adCiZEXIn;E|u!YLS3+loJ&NO6zb z+8!n3X^DyY=dpTO?14Ukll$_h<-CYLSOgf>*v;v+zc~Kt<5BUh>HNCd&$#7o4|B0? z^pzbjWV=gnx|~1}OiOYHDyi*Tx@V_`Y~~9k zo~u{ibXSPz%QxQ3gE2mLb`nN&`38W8MzP12FxK6;>Lh7XRO!BT9qjb=9ObpnE%AXs z0|6diS{0FL6^0op@U}*QCmv=iVyrML^gPr^Pr#xir}{^D`MPMwcB(eM(tRLrl(bJq*s?c=bbhxhj${zDJqB8+&%nE z&8z@D5A|1p9;BYSn)qqe6~0cyxCNa;;wXv4idjYYui2i5rp@DinZBU5leLhU8UP#se`uK?KJ7Tj zKcxknEmJqe4UzpQ^!-w{+Sb2fR@e(}QJ=0wpI}3MAw~sJsq)+4Oz^13r%>4-p;FrdrTwVCKm#FjtmHk>3m9!0Ga@wPr1uyjOhO6uvuxQB9t%d?lK zq-Kzhq;v*lGHd-jz;57kQ@TxI zVWf6m#}xESga9F*V>#neH^w7?Tm+h^w745osOH;0T{@`!5?az6AcEBSUv>2?TOjD& zM%I_}nUfVkE=^su_FvbgQvVZVPSv5|DoX8uxX|D#9QoNJ%Jg74E&BlQge>jEo&1 zheS5>OIZJ`VxR(elJANa=o1zJb`<_ocf}&mO<|=<)f<*Um2PS~th}t4naG z=sKpdpEo_~drjS*yj+3$l5dFEg8H|SdVz<5k`F_Sq%FocZf$oz{ZSu{)$IYc_gTph zKNjz>tc$sLf&npFYrhI6nbw3b1%Sz@>ub@5cs}fLt^Gnxcr|R9N#we2nc>rh%Ug4H z1D6@hQ$EqlOe_%^xXeUcWk9g+hAcCQn7hPfMuJ_`uw^FVEJHkhsAXnHrYnKBkXji& zZ~1jA0a}_?LmnP7gJbw zs@>@_Yi`2my#N#|Xzj0)0I4vtP_nW-HIkJ9BP%5@`df58dEt(M_Z_L@5RezAB(5dDJ-a;<*OkOmlB1*AhDR zzCBj24DxXEP{*p^g+>aQFtkV_qf2py4Xu>cXz_eKt>MZG z9IcrGoebJv_@v7q9p!W~rtAew3#5~&4iH$DwVdkunLPs1WiL)JLe3ZEY0U!Bni;qG zm7_H?P>dz?k?AO=e%h7>&2TZ+!qOV^Gfc}Mty%VI&CJzd&A7rOr;{<7wFIr<>WW;W zbWO1httoZZi@v^gSslPNQ*OlQ4}T$x=eL+jzk%g7z-mE-ViD_8jaT}y8S$qm`lMSUlyA1eJstQ^9Zn(}wKeODFsLQr}yX!=NfJXT+9 zG^Vz-%ArCe^R2sNZf+RDY>Qi}wDWbTDD;NaV z6Pp?>T89#_rQhDkVLze?X3MP#A-}_HRVkxUH?X)TLenh?nV4-iCf=$N<_+1Z5|8VL zh-|A$J-FEYtJRb#_aBmPRTw27aH~p?;UDi-6=7u28k)N~qOgD`;^`B;Rqe!ERm6*B zZvJs>RR9w=lJvdrBz;pP>6;o!-++Ta*rkgHn}A1wqA%V*Z zeSoTzO>q60X114(<$q4rH(gZ(OYjHK_Qn3-7gUBJM%uoq(e{akxX+2p{sG;*bcI+w zEUp?1PmCMs`)oI_`Y`BG22ZeIF% zcGdAuKblt}6O&2REf`(7>CBe3hE86qt_hZYsha26PF_p%-OE#U_Y*Y_+F4bkelNq> zC4R65Xxi&}XE~g|3$b%(?UcTmBYQ5gczxCG7z9*x>>E9@uV>Cb0j`I_ThVF8Q}ua6 zYMK@w6^Ab-9l>wd=un6>OuR^7YL(sF!NyvJ84xe>+mY$X!EVJEsUf^;1trC&0%- z3L1=#e?om~f78k2)3=|yes%K9WPE*ebNv2|<1cuziZ+b-gQ72 zmS}macaKv(zE({^NnQY_K>*5 zvtb6cvY)h6w$mioQEsO>@CJh5Fu9cpcbb%`9Qxg6^pJUZ_AT3K5~j&_n#9BUAtBjm zQjhJZHx<@l5*=zA+6i|W47*!y2WBnp7jrZULM7X2BD^hF4|kHCCJ{809xrj$Z@vxX zPWrN)CL-Dr&o(;3lO^eZ={QO^Bc;Lo+H}odG^@C?oFQqjdHRuf?+)t*wYqp`V5t%&+Lu2`e&KX@dFBkxd9%OG&f4K2PBo%?QH{EjNoXBxJ4 z(~MNSf_d(>%`42OH{RW3e3{9$@1-|uT!}uzmFO#nrVEwmz%r<9UPn~gRie*$mE1Rc zCAwkeX7GBI=-`SEDu~r1%UtzrHq3QemA3#%Go?nkfBZ7=W2e~qE`BD;FmmN zTB`YET8)@89G5gs^Mk-xLwUVmfAxfwHnPnTEq5yd-_}rFFR;~v)s4XCF$4(sx78U% zeJ^+Wua(0Oy6~7uQ^&a*AhjUu%2lmjzU>2kqgIneKH&2&t$u%$QS9dbro>X?O?}Ty z8@w&zrP=TM-^&#!*O+88+w@(-H4loi+O$Br>XAhqR|{Vg-(ha-uQ62mG92C9^6|yV z*~4FXbnoo($ph2#^3L)kZwvybI#AHxcVdl@ScL+Zzx|q)_Qtu!*D$MVPemgWT}oM&^!G{U(wyNw$e39@GzlOtnFF z>ZBfd37ZHh0~T*0n29DZW{Y6xFaOb1zM{=au!%%ev~tzk)ovn*Sjn8yiz{G!R0p86 z!6*qfkqCv>y0hgb!kHhQ0g>ugJ_r4lBtzK`R*Ade z4J}np$!pLg;8{kvTG5*mkdv|rPRTn*Js)OTNY`Pns)zZpC6%arC9fab4f2OkqE?Qr z#i2!i_|aED!mA-qJriA1!a=*(dM5212gI3i+Q|o_PQq)T9|$D<^Z-VyM%IT3>ijU< zxG(|GujpgtCA@2dNavu&yZjKxN#~#kMC>xt-T^0InS+>e7xK<=4~Vtpq~>RQaFmBu zK?(1s4@>8+5yTucf?%0^Za(?U$(A{hR5uG2o>&R5nfv=F@lQW`o}C%)#k($$8IEz+ z@Xl@NKoIZhV6rbzevBUq<;1&l`yC8i!#gLh;qA}CcPvQOYmEWPbsli9UKPS00NA<> zEiOO9^}VIYbUVuzJXD*ilgaZ{P)&MHGMS>s$mTxtSDrk%|JK8!%Cm_>$n#T1K6_3In1CMa5xqJOQXht|`5o_guqzDTsv4|B3hgE{ z)!uXHi(-UDdZ&6aW9af@oy>{4?BSC+!YoyxxT~AYiTF6^WX^0-bcEDS=7_A+d9dAl zB^74ZJcQ603Zcg00~MVW=wv`=8Ay#~o_-`)49PL^J_<$vF;ym0%fJ)7w>wFh$ujn- zO=pr3Q&o@@o@VEomz2#>f$QA$Mbvo+wap~F0Q6C4Y2v%YKo;eP66vGnMjur*l$M_O zERV7VipjKqq#m4bo$i>)oa_w1XC|&u<{UK!Fi#K6g!^tzOlF?-%DsD*lFaozsa+g` z%3aczUQD*Zq=ftT9+H3yWrYES8fDIH<*mI?cH_XXX#7EqPvUBeJK(0}yc`-BJ zdtx$xr$~R_w+ujIGs6Hr%glIyjEw!pzYA03&@AyMc|U1W|1G_+P&x6|CsQZt`N?N1 zlBqNmFek8(^M@kjs$J7({PdkHLl2gjY@T!`wl9<)SJ~$i{bkyTVJu*pvE@}9f*?w;Q{j|CboF+j@XV7Pq2;05(fM-GsDVQ%CLq|!9Mhz56%E0|9Of$o5xDo2HC z{*Kf+(jCl=?x2DmaFY(7ETV$6xoeL#-_Yj(mz4*GLW&6Z>I*-dNJKC_6k=>s5P0i*=v;)hfOp+nR$1rqSj9=R8e zw_=cCe%_~Ib2XiuRHkBreAlFHfhdGZyXsxlQw6#dt;K0gL>7>pq3EvKPEQvEWoXyP zv2Wp;ozj()ukoT1UL%@E3{HD4|$^b3xV+Mhpj(Kw_-$dV zCP9D3{I{d7tu{Nk)n3};2R)fReuyJ0!VKAxhd1m^pJu5r5$|8m@(1%f?70u%z_iaF z7Ouo&0ZKf`wM0+Gjz2&N1N;f*dp&aeL64mgF`U+%Dj|R%J0}G_s``InnjZS7km8A6 zHslDz^!S#~+ZJjzIjzmx1ehT{3k2w4vO0$#mO6%AAw&HZ>llOvca$ol$q7ybA&{k^ zWTW0vMB-w?j~|{r zILXN!`jzZ1FUhns&~0w^t!{qChxyCu*5t)pm~-9T#=hZOyMgU7#mx8A z0h?LE#9^CR;(`5Ok#1(GM;Ckfv>c1_B$IDu81A-w>;DRySpnI&TDxaA-S zH!NyHRr1X&BEeF3GY&moqtVcvUcs~gg0#%yn5Y7aO5D`9M-GZ1MVfj1(TmdanL#9B zw=XXq9`pfmTD-$VBD5^RdhBdM;l8l6OK0qUa45AzOb zq*Du{jD%_nBUBTPa-SKOjXDS5Skn_LTB{q7;;;Cy^D|}HNV@iaY)~zcbj>^+&||VY zg-wRsy5KMUT%elBg%hwX$(|nlhrQ-G@)UFAid#STm9ia7ZglIXYh|;rRela8PaRA) z*M2&Ol-E@C3sBKEi-@wj_Y>0hV#t_C*<b_Rp2-^BXfXDSmD{E)j(|%BV_n@WybO^c!EkNyI+09QKb=MCv)LrHaX9M_VQ2uFs zhgD$Ip$AQh`;(*LAp9&oo&+L({HiCHTeY4nMdUI=>C2|-Y&V!1iYA5U9M9zE(ib8< z)$U|;u8`rOxKe1iNA%^3KxtiII}S-|sGCk?3JS(nRSZ zfdN#j<&C2eTm5MX>)?syn0nm}0*9}@y{F*|jj9-+chwC{T6LDxdIIw*>yh{)bh_bIa1qYm(iqLp!yV^?CJ z)sIH%tDI=B|pjs+7@at3u44pIYqIh|O#jZ&eBNhHO=d$Mr)* zwpFDb+zAX1xBrkgq4m`b(ya=!UB=eHSTkZ2+iav}qAv&u=LiZf ztK_iSX?q#kvQbsK6A~H~nwb-h49_hYHosG2B@z-Y`gnNhO4FAB$+PC#o)NFet{r?K zp;pY<=fT?t&C_|<7;H0p8DY@^??L5cgyz$o@i!pdy2XKgEB~*TS7}e_55MmyI^DisO4!5iWC(jP1WeK% z%nZE>SZF!?ysZYe%`8|_W__;CUQgKS%#(rb$mnTkVjwcw@pQdlUWQ5XGq2{$uXFh_ z)R<*ugD6=izvWQcp)dHaOE0RMlMVapk_kTWn(F%_#rn9``#S_T8@)?RDWmJl;Aeqq zb)x!U%qMNYM}3?r@z+tC7>ef#K$)$EAq_IANHy(EHj+uwBbXA;hV8YL{iLO`*avif^bFZ=lF0)Co^3(iu7~l=||!PmXz#xe|adh0vfnj*T{!L zP71g}8aVUxDk!vIzP^(J8s%Z&K`Ym4ZpCdi=RpS0wDCWwM2JZF5k;D|mC>{bMzqg? zsHu*GfV$R4fPiZClPgegjW#yE2Mr=JMW5X?K=sz<>45%JM}{r0w2xjn@WT~|1tH)9 zGoK|hL_O{SwDWekt8o4CIJqw3>%r%Ohh?Aj$Eb55y+cv~OZPFw$1Ax@qwq-Ki`54i?KSq;?+xtw`=6)Ogew`_yggQ42-m1``wJ{38Hl)puEIE(gtFTyp_ zzpah_jn#`V9%YM;i;nSTE7*%r7sLwqan*o0^1dd8Vf7Niwe88)Rg;^i1Bkq=*pYoN zLZjRb-x$_YOOn@BI=CPqTwAe1XXtJ68%D-yaDlZ~5QehKRKt!lNQP?*r zKga*17ZI-QH$)&JTvaV<`qNb>BGlg(418w^y&_*Y8#q1U9O_Z97vVJYsblpbgm2(T z{EFc~3S&#V6~cE(^Lo^hZyNdz2#=K>%hPrF?aqlgG+&&~*znMfKxnvU^qUy%fR8Ti z3>P1e*=@$U;DJiCtf-KZ3QA?6(a&RCO@5Eywi%0l1A)>3>Yh}h+0GFW9Obqdi|p;A z(TX(|1O~5GI{xt}p3-X+`JFAEZh)jVb2+=sSZri3fs}dJZ?EDq#^}XJuZkCm1l4lL z^D-6*jQeO4rHcdspr}RZkV^tbn#gWg+K-e7If+7;S*05P_ zZIpL4RjJ%QaZZ3}Di#48aQ)`i+-p$18-SKBmLl*`k{?r~rCS><9jn~9Kc->{qCVma z5K^_;Ud$Qhym4liqpcml^F{2eAPvip} zE?Uh)FTyyu>LrM#V9^AOV(nD;^L;-qNxO92P~wn@YHw8XU|dQMcY9+8l>^+(>_OXI zg4pFmi~mb~YJV7*{LkBy>sKfHW`_U!`W4r=uw{H8kVo#VgSNYr@t3{bAye}@{2|+2 z!px!DUE-1bu#s+Vjz-6i5I zvA-N<2h5}{PI#x;i2Ra#3=sv8x{J7MFhS1iAFq#6~=m#sv-5jPR8b`|q0@cF3x zP;K;H@V6n`iSUJuzh-+QZd^MN|Ma7h%qo?y)vl+mKjO+ttbw#q_i457QWIKVboSWq zlTg~YE^plar_2Ri9n{xfcMQ~*dc#>!C3OA)D8iqGK63gLD$A|$Cl2B1do3+t|NZH z=00YX`mR_Fo*UXYw&E!aG>k3%n4Bl%S3LDkqvAE!Swnl$N2#a%zA35I%)K0KwS4JF zAcf0aI(j(qKuZS(SZ&H`GeebtQv;SNC0{xsWF-vfa_w!-6N#n6(SwF_BS)sFP)aOe z1D1}6Qb|31h^2!mirfGlVXSfD!zrCaby+ad5t^qLOEEKH)GxTD#?==PZ%92LA4)k{ z|=ZZf~)f1P|s{c3=lY$MX7eP!RA`8+3# z+*;P&7e;b@e&HC39y`NOO4_}6+5l%ezmLLIvLl`2D{AOHWUd$(W9ldH}% z_sMcP36uO%F;o zeTF1Vu6}OXDvbhb&*obDK7H@q^T)5h@%ZBOi|3d3tP(Xnj&iO#uX`T&s=70NseuP( zJJCQ&{$-?|FsF&LD8~ErYJBQb+JBei08v9(%Uc84l{6>5fYWY7Pe58Vap6AE4{6#uN3_C)J(jn zn%cl+CLt^Xf_*z=naRXlmwo(TE7R7<(FQCt32_>wX}t-*fNKl6^1BPRsUsGY9UE7&YSWN7 zxZ^b&pkYbx$9((}h8BrfO!GsAR?J{@N|u1Za7V#0m<_4xXlF1Rr|Z~{jd>;}Fc@y< ze2Adr)XmHix()XckicLzh{5QxXjF*$Twaw@lLfx+mf4?q05 z^Lr0pecTe;T z2-TbBk}Hm7F_SaW*z}qabQUL{^mp3l)wAP`_)eQYWGH>23_KE6Xf9nhdN1UV#B*`Ck( zuzyXxGJaVU_OQREEHv};M_5B(0|?Qi*OQap9ZpgJQ}V*UPpKatsaFOK_sdV7JowV1 z@s0ie&;62=cja;C$ZK3&!A+_Apk*PB`s$H-d64%bxyztH4jXWH>#5m3bj8Wsn>k3h zvyCmVX5fm0X$^l>J^b3lkc_oLuyVt~d#^8RY$g ztvF1*%%+oG#@8<2Ai<1j{&@KMP7bJdVL(xxRWlo;n%Oj}W@|JLJ62Q#=Rz-wnZrE0l%O-|y5lyrN_@!&+5Md>fE(Q-rU|<0 zYCswMN>uqCIC2B*7L@$Wx1C2Z> zmiqQl0gl!zkHuRu{}rnISAU#az?(G+(bH}|Ec&foDg^%yvsLAc=2Q*qwPYX@`)^el^M-6yndagUk!@AEhO1S=`i((# zy%KL#7%DTj2Yh^~cFx>YBJp0g#Kfg-y(`!f6g?Vr_DUv3+cdfrs>kcPs4Oozsw^L5G79>N;c1#R*mx$ z2R&yxnI=5hJ!iVAKWB=KzIglL-Q^j0#RpxF`LbzK8ce;ruac$Pv&*S(z3_NNc{w!~ zDsO$=QT{PgQ^(Qx8_;)MOwRVB-j4rC_n7JO%zjG*6~ESJEe-vqQ%dR;x%`X=?FnZA z_*zhn9a97;@aIqgD*u*8@%J5QpQZ;Qpn`na%fVJ^F`kv>V8!1lAlU&FlZ=~|CN742mZl0kAz51ST zj=0LxEfhl3m)KkC>qmTu3#U$`pq2RRxKCiG%PBw)y?mYr4`dX5ccj*mnH#}ziD$zM z>SaG^scfgoP^8jMbKngGVPbMC6YexQQ#l;G&*&lZ$`o9-(_~DO?=+dl;*gN+G`VK0 z{JWR$$(bYDX)yL4u=<;&GiKYp?2dOS;ZBo)ZozuEmFzT`pt*vbMj(&ArDQuz!nP%z z?Y+~G8fbvDfm|3|LyC)Are6Opb6q1C4YD=TfpTH}FAeL&dv|aNIrTkvBqhJi^#Ye? zQ1Pq()rhjmR6hX2obuVcnPCW$^P~G8WMw(fj&}o)?g~fLWVIWpwjWU_-Cdp;>8@Zz z>w0Hk3e-DmfYhF+KH)mwIc);ob>UxDll*n_Nn+qhyAc)S`T^Z?9TbffO+bA z1Z-q5Mcuc;Bgz`!LQ=hM@X`)auRG_doA*2*q@G$xZGMM6`2lgA#e910^3>P7N0e`w zJcJ9gjv1Y@8NjGk;Dp=Gn~uI}6AgWxG+khT%o%>+AeX^!Kpq*Y$$h-#haGgCr5iMW zt}lAnv{1A0q!~69a35)%KHmk}8nW*NJFR1_`eC3xC$thxrR<^nYp6-EHRRq4Y;}lL z&-6aFN={Xzt=}oV9Bd`>gC0vhRv)>$Z|Ck#-S}jDQ{A||dqnc2pbgU!X7jq_$sHf# zYqcED%Q0T=CuB$0)gO%tD&BGfHL5Dy-FJw!gMI2drS#<%lk6Qg|4wmBL&^O%FPd&@ zvU<1e;7013-mRS$*OVU7MKB*^+KR$hEYGdAAo|cV42(?Eq<6Ee?2q4Mu zzOOLoM|IQ8+7R8F)gKd()}eWyQ$DQR1_lb-MOqNv5p+a-{M# zY6$Bha@y+7aCYkvc~$dO`zDfM67l-szFi%Bqd-)Vf~H4;O(bV12bc~YIAm6lCkQr? zj7hRhB-2nF1ceNQ;?&7Cc?FvYDPk0FBAAY%(?)epW)}7iD`cj^M1C6*Y$6FCZCv%X zt(!R$7dXW*g}`$X#!l9&dEklIv!HBUdd zNSit_+SJCuwP3Vrf**bbWxZN=CZMALVpV@P#1ar|rV7;rg%+PvakAct6Kn2)scs!0 z(xx7|$9&hYF2r)8YaQ68z3=Mrr-H0^hG8iv4#9HTo|ty35aDFK>!?D+%n?&mA)?+t z_?Yixo?uf~CStv;d(1o6M9Q9>g1S2B)W9d}J6Z1&$6{Ss9pC!#Gwcf@ThMEdN$v?> z^IDWyb^u=1Ua>uB@rr%IcR8_N#-N~=e8Sh>?c64(-f~D;uQ>)OxB0;J9Xvl(4-+j1 z3?wpC!u%-L`;I5i84gSCuz)x}>uKCXosOSrqH%fJZ0?M~v``hK@s#3W@;v`D1MMDu zTHo|@hvWo>SH~3x2-RA~-LjA!JhfM+dV$j#-o#UbLk%k?( zpAB|LVF@84kL z@~V}-80>F$YoxL<3N6867?v$^aJ*Cq=&RCuFrTT|XK9xcq_3J7eN_unne#@`2wV0y zQtdN-HS#yOo3u5bpxNIxUrlf#gu8iZ7eNEaknMXD+&tCbQYTcp@39jR@6+% z(gA#W;sj<>R5`%34(JKDb3P_)vhaRZP0uVinNCPA`hWofvpHra4MDXF&JB2sR!9y^ zsu1;V!wOBY7m@UZe z_)e?ea5O+eFGB-9#=QCfY1tGMNFamwYv8Ut4>c8khWC>W0xZ%E7hLf-{#tdao*93v z;aVj*HtVyPzx~fcj&jrf>F@rf{8h@Qq%VdapOJkbIl0L`f93D~pzKilwkCTh|89l7 z_Jju}SdYh|wT2uviVVE!L-@8@kC)XD&URA$%}C|NqLgXq-p{utbE)WJ1mlqs^|hly zRv%?&eE)Cv)Nzlvq1WhyWCBRg$GzFlxazM)v~VgLZ}DTV?H;9_;SXe8EEi4u%-xT} z26}Y3g%3z3K4t{u%5|W*-V6~$6F*@E&0)H2pZJ-$@#n9TKGBh#$n=CtR~OO5PuM}b zM|(Si3Sa*jdP+(f5>`=?onNg7P+$E5nSyod&O@C+fP2<;qUilhirBSmQ8 z2;jR8*?o6sG%wT!jxcpKw5Hs`@X)P5a!vM@ZJdbKI-EFC96#rjwOH+Qj}i-OIRS5(DHEao+Wfn zR@+3$Y755_gpaB&W0qjvD7HPPcK~;!^2usbC@lcEg6s1~jwR?oFf4)mHXJR0kSA~- z>G?XmeTOPiClpV>KkJynlJ-=xR|4ShyuYo@)MR|xcw5Uu2XlD~Tp=XkHSevCEo>$v zDCltr5u~r(z;=doykyT_I_?rZ z{;<)@L4B?ry3pfFFgYRtg21cY{Xw2cL~B6i?s#QuKnrvbtpS)id=S@wluF3ffP_1= zN9$Iw24tcq-KGFl?6Pi3HcwORwS5gp*n_~UeU?J9N=^_o1^NIK4C&R2PaocK^Z_Xt zrbfZg?oM0%om#~FiWw{cG6wX)fG^B2j6!E4SdM~Z4AWK_1NW-pubC(-Wa?xL(+zzg z@K+~ym!1I`1K;ulODH#>`oTR=_{Fp?7*Mg6L7&70{GoD*N`2N@%P>P)h8fl}GzSJs z0+~52!wi(NrzQ&E4@=A+8Ys>F7T+oL!)qC$)Q_d(4<_}4G6~Lo00&kHe{fkJn+Bin z5+Y^=Ed$6kpgO4DaQs1Y48tGJT27S!K+L#;hChSX^iW}i)KEmXAx9wSirp{xJKbDO z$7ii~x;$kdN88-$0)hxR8|_=HXAq^_25L;x;|f48U5+ak*wT=~(Qc`y5VS<4OSeU* ztbZD*U(U>l)#<&<%kxK<58ilm`sDoKubsa2_W9#?PrrQr_|f^p(|h;tUEIA$Swx)U z?(QXpLI=7U&aNWP&-qY(Nxd+Bt`y3gK0!u77d+6#D?Z!La8&q!zuWH$<{ZfUk@~&s z>QmiJw1p537c9rUhr{nqnHuFieNLnl?H;?lVO|c-oxM2$jhGlZO;>+5hrxQ<9kZ~c zZjZJ8`$y^pX&6Q&%%xsD?l%w8?FZbJd%8DI(yF7b{&|#rS(OuPv0_K+ZYno%9~5D3 zc66DNZ)O=R^1g3o2(}$+Gs_wN!Pv}@`9n6djETcGvrL0=ut+zvT+`JmKmDc_E`~BI>LTcNr zlPvA4Bg<1~NS9`sKYmVnJ~N18PKWK8}qcdlUOZ`GE%F}j9N`N$}uyp8wKkKX$B?$$7K1E=m;5ts*`%n)e)K@ z_1X-p*P43hLk6Wt&YXm82C!(|yP=NI3UlQ8^Pu><(;g;+=Rwi^l20*qiE*xu(8gx` zxx~1YgURN3PzU_$O9p-(be#}o_dMvP{yZo)dXJTdzfG~LBlO$0j?m1?8GE^fgX| zXcMK21O`z2rSE`?L@FoLU#sl%2Oqxuq5Ufj(nTUcPNERr$`*-CJawDZrCn9J+=v&6 zgfhuI+<%dv<%i+Q477Uu3tyr}RQe*cOYxc^t=`ON^#&}y!7bFldSm85afc>1-!ddT z)HN-E;?A_*1eL&flI*?}9?s8D-s8+Y#x_4A7enu(;097emk11td8}?sbzK(d1|y;q zjj82GlrMrYu#S?J8VL{**Zq#b%|UhcRDj!_`>2b*K)0Fd>>>A&e0mW83m+RO_i?WC z`F7K7PQJ4T+Idkq9gYc{>7AbL$S6eRb`8g5szCX9!xodRc;(AM2`CoLmgO$h3CIh>cOM?=kMIT zeD}@s@zL3J@ogz_qxun(swIC}rfMv58)xD6i!Y|ULw%dc*AmM`d@X5(cXdwwERFkg z!3MI!es0b0S=F&Nt2XOSkXkB%Y)LEqrb@$%6>n7;G7{^@9C-srY;p%FJv!?o_Prux ztE5|1&S}#46Bp;Dz>Vvc&j4V zZr!qufHz*7GJsb{I(pyMN(-f1m5lNx{giE0nRsF>p~>yz0GnCi_95P?WT-cBGs8Px z1JZ5n3v7_o-6dSN$OhkYV!}BR-p!5hZoomH!8ySspqvBSWpRl&X)4=$58pl?KYd5t z9F4D?Jhh{~PD(gO!nwH-&b3s#+k|DZy6(7_2oWRfyCKK~74 zjC62wqk|KT@h~BwPQw&X;HKg59t85RG+Ov5*a->eP9C<0wniQnP;irg%?b(UXDAna zzH~geNtGCcgc}TnTf=TohC5}YTOr~6EXqpXI5NzAkkD?c@IgZT6vT%=cYg2Ts}DR1 zv#II&=UOJNj+C9iZVsw;2p6B7knmkAB%Hgd9lmuQHU`^_xlpy3;$?(Z&KzL6d9K0s zqh(?Kq_f3-3xqPlruh6#3((C29&CUfdBR!Hv<~RfGl4OO8gwj*9)90a^yyi8cL%Ul zO3}mb6exOq!_b%ZU}mUQz(UL6=WRB)`^o~;@4b#MUB!dQ=upH!WVByI7To|u;nd2Q zf619T1ev+CzsBXuPytnx4N{7rPx$Xb&#BwvHT%1e2`TX7_`XhkHc{UjDb~le=ob*& zOjI}g-#7C+CY`;U5OYrT^&>uM1Kx~rro>;zePW3I44Z^8xEl6yYL;@Eq91F?B#8*7 z#Is?)?qwo5Uu4Tc=Bn>7<>8lu7I~m-IUoQM<;=rP{tP)v7AfI!kRT&L&~7EmK_-UM z(^2;sU)ofuaTP8H2_=$vw)b+t6ay}X>*owXQu;EpyXflt8=!zPtsjdQSbEcN#LPpX z1*w5+e*$5FG;pT%7AT~z0Ir9m9quW``5DR+T)0wm3!qaWW7mO<1*vJ{qaZ(`NYiFo zmyBr4fk;z;5K!L(1qg`keD5?e;hIl>+=Ge>lyAK-`PQEPusi34fHu9-&akNkKMSt7 z%?kmwGoLbV7rJ{JxFD0`p}n-jJrnStV#>&j#t$E$3mRPapyGUDvly*TQL~5eJIvVw zg@F1M%b4`o2Nm@%G27EV-{FsiW!#YvP(NbHo7HbRj-eM3!!dHT>=PlNzpI$zrKK1P zaJeQ2=^OB7AL`eAyybZ7@CE%&z>EH}w@|b3r1`Q}fET27`h4fP#WLi03U*pA0Ho7A zdmYNN_YQhTK(>Y?Pl2r-R?w=>KDLTZT-@?ayRSDNuv&zfKEd^m6mKY{B)FoLl!)V8uW$$KPQRhOC@`aDn7<3nQ0n5D-GDRE<|JAe^Aw z@r5faw^-Kw;8*kD5K{ier^aEFk^XIA^lz+Mgy|^9=(z47P>WDAK?y&ehzGa&M83TrXD&8;RL0WFLXgY+jNV$UP7pS3vu!UmN$ zIKj!~(}`os=3(7xx5C6g62fT=mM1R5d@^fkR2#jg%b)~D8YxB{2+!VS!s2QXE?fx? z-#V}U9-3@Y5#huNnSqEkD5m(j- zawc|}vDCD3A8wH-ZLieCNY7Fo7Kwy1$voVDkzmRXm!xKo50_lsx+T)= znbvt}i*fM8WQiJB&rp`6@-Bf##Y9{82(rFE04=3yy$LFT>p0ncCo`L%{KBQHH@D=r z;;8`GF0kxqdUM175G&qAFxyA7mi^O0mil?)!-K^=WFfze~K zv;#fjcFxC!P0X}SyLV5Pb5hCVmzD4ivc8fdW&|=Y$?@>GXm$FoI~JtmL?A)58IVqb zICZ#4{|?hOBte`dq)2nwy?7E7*(XhP62ztDzGbBlRm4$?O|%U%vVF<>`I>rTzRslEhDY ziR~Tr()b0A#J1ckh(Bl@hfae2#h%6^B}ckk?-UycGqsG(fB(Ar`Hp5t=D7c` z_{}+^Q|vn?Rm`mZZKO8h0YT&~xy|aR%M@OMQ-bjXWDe7H^^y9=BlVKl12Kz}`|`Nw zykLB?NmlxBCqC)w;QY3zeln`QQ(azHC7E-R3An4+U4JT||I=R{& zRGSy00-6SRv+fWdG$=)Wsnfn~(~RZU$%j=CHqJ1l26^)JE;&*??UXFKj)D48Z#XNO zWW?WxN%*0T-n4zAbA*3>@|q`|&#KLMB_y5w&RzjE8x$OB4upyd;&YG3{dDDlo@XMt z$PpT^XQBI&BR1Z%f4USwPvZ-Szdmo3=HQ=d32LBdOOu@3=SsdPkkUH|G@@ul1YK1> z7^$Z91N<$5u6)b-B;Az^;s_qUd(=k!fXy*xmHMul4PLS@jji-kB><=$L#433QeRQk zKa8r+To(oZh^S)Y3bl0Vx4_G=fNWm=~SF<2sY)JZUz|w&M)(2$iVAWleFC7W8 z5(acV_qGt1nb;!7_lnk)!VW!P=}0J*)YJZM({XiRiXvB}j*t(h^d)L{3tk}|p=rIE zidhIFkt}jW>H+ys%E=;ENDpXQZ-P?XdTLJkW|GLvz1#WH{SR*SJqJ-%K;M^Mjrl0b zk1EplnbxJF8gnReSBJt#o%4R~V_$5W-D!hnw{8&|8P+X-F8kh)L5_ zCyCmghqadUl@ms;ENhR2kzA4DrDG^M5DY`9Y4_5_5J)3e6)l_h#$N-PpfQlUP}ed; zlOK2J1Vor1MKozvaKcXy3D?ZI&@&Ug?>H-kyRaiBP)z;)-fF5g^All{A=oQ89VpQ& zK#xqcz_WTDWRymMwP$m!eV@Mf?)l@_-*|j+`o;6hdsd~I-efsfo!31pd{x~Uztq4A zvz@>qFmmbnmyvpw2ZXY}glI``E0VAFNdgz&P!rzKvli+{N9xW1+hb}n#E;eYFWX|S z%LD_!rSr%dStL^wn2owvpEktvVTuaCvoZpj!7Dl)PX4~(ug|0eMu1gTMvQk&BWYuaNLWOL{ zyMffO@bOC+S|nmw84=5Xp%pV2on$9qFx*ja-dmHp4!#1&C8w^Pu44x%P3wTca69Ki z1SO}gHkQz>&zLhMFqj2qFs;d{)6U7M(#}zogVAGW7z|Ta5S?=Yhgox9=J;DkC_?q7x#WuDSj;lr#lgM%lm6QJyn1%L5no#c z*A=Btl!3>Dgxmppv}ZCE+?J&F0FU%Nh!#XH*!GaBJHU3&W|rO0(tTNv`umZ3y)~l0 zbnoI~7@+yAkM!5nE8~|%kq&#&aDDhOTWt{55cpFF(PZvU-A(I>ZPbsC)GLFAJ1W#u zE7h8C7U;_3&XIT0-S<5vTezmFuO6wF2YEk|yA1MvI2U(!8+p(kxZ-5)%^X(K*~S)F zGjPShv<7|Hr2L#Aw7usC2kZ=8aS~oLB<8n+R-8=a9c&I25Db0zPEJyzp({?pYzBFs z?!5yN-J{ee`K*Cn2F;^3grj`z@--67nC6d%ukRtIsOcSy>a3bsBh}2hRW-xC*ZDg+ z%HmqPH!If-vW7hx=%ktnV~b=nYa^Q(GPai6P($hmUP0}QzK;mF4R;tkw*lIjRCFZZ zHb^_O#@ZRK62iv_O4?dG?aaDzv2+UxC@QwZ+-AdDk<47%DNtseQnfy%_KKESl&ZC~ z+{XNlj|Q@jx@!QctP*bHvb1KPI5`!A=9S@H+S0n9odMaEWIMjJx<*rm+uW9&dm*^t z#x$(FrMbG3mN3;j(XG#sn+o?zwq5`+1SMGW*QPwzF|YOobQ7|zhHTZI-Ow3Goe$Lo zK0{9cWPM1!8e+YtH*^Y8XX_@kK7Z*TUE~JAyvbA4Ky-+B;Kkr<_+1ZGR?&yBHOBR4Ogp#_49Ro zQc1Te43!VHpL9F(rKDNBRV6@Kw1$@6ny4(`nRptum2~uQ_FI=(wMAQ1LXKr_?!Q&l zpANni*ea0ufoo8)U~uNKX`}g5?>U*@8p-_DM&>8-uFbEkt`FXQ!K6+qKUeu{jZ}W? zR+Zn15&oXzxgt(8TWx2bOSg7XzxBGMF;6}_>HWeOBfa0+=>0@v94q~*gRNlIFU`FK z*ci~g`p-EJZaQiHT-C1)()^j$f#$ENhdy9XYVO(z|26=M*3Au7zgCDM&%hfQ^;SGe z2G5zMz@7g-g{%5Cv-zd2>etGlWb>S?+iBagPnY=C3y)Wnw`BvETm17`hwg3KE1>V1wcCEQH0UaRZS1#1Q1NR9 zGN9{`^_xy9saIq}-1EG=;P-g5>$L*6TXno`1`u`1=mD)5Q5@FC1H`#dua0 zjun4LdZ9NbpJ!|Hm$>~}@Vo?7ec z9BK4_9;sjALUma7DZxpH?5*J$W+z$epKeULQX-KDs%&ee~X~lTY~YeooyU zubaXf^aPSXicgHxpNC4t zvalUz-tp)(FzS0F#ol&6I3aFMd{*93)hO+_+nlNDY%yS0Uq9mCjDf8&DVO-`xKCiG zi(p{X-Q1dAps7Jt-yNy7WadWdSczxD$*t3rAg82x1zxbMWGK@6w9_CQe2ASUXDZ_0 z>D_7fN=Bd2L+16~X)>nCcbZINaY#sZnq0Gc{!LBbFp1Kyw4DYm!GiiM-)kSH%7WE7}1zslcqquvlcK4z%|$T zej1SPy0q(^H&!nDQ~OC_o^=Ve0@vi!^kJiXWS~6tjn2KctuosQdFpFSs!8GXt<0tW4LEG!^eY-`B97uf0%E6~dKI$pG2~<4r*`VTd!+toR8aAj8?RAS;dyO$A?{$i`c5f*xy2-V$IZV}+|p2T zzx`x(J4yEmPh)Hf#Uha@Qo;ME@EkzJA1 zUyfA1Mh#(IL{3}X8RkO|v+5DkzKLX*M7(~uk8{WBBX{?%y1NtEKN;UtH*TA6bVsEW zmwp=zB97Mun@G-31WeK7o~)~lWbdX7Hzs#xKw+Qt z{BEL7$ImqJI|CkrX`w1e<0-|%Wbnah_^G%6hDGkehK6z=jfVQA0c^X&4J{luWrk|P z`eHpWq4X{h=!QGL73BW=JOA$HXuMNs+&Fn^cjJ_^ei@kQ}Q`q2YWzu_MI zlab0lqmB>-^kBsH+B_I`(!pJ)BfoVrXD-uMF_{sc>pYn;e)*tJ=56z2PJl~OC~n&( zb0$7chK6uy_mVCLs&z6ajHS(k9UMTK3NHo*eyq-bLW_@@6Xc@w#lW_Qu8>BdC0Go@ znnqa@?!lzKiqBN+v$U%-sISuhYGW!bh5U9k3+FKRRgZ|fBy%6n@Fay)Oz}yIA4{m1 zTAdliRMS|-9EBW9ATZOnDFH%ZyWP7Bln&sBenM5;W3;O?SUEud3(^zaIm&t;?_OXw zx5>hGRxMpj@T5mKLP1UK%%?v8*5gb2LhHPtYU@5YJ3mG{$B)tK8~na-=iEF->y{_d zS$f(70x*e5$=p3gJLkzXZy2BNA%exik-$u6fbz!fHyjPnoWjt6k1?-4=J@AW0JyYL zKdQb4@5-U3;?MAY(m{YpdK3ni`HjC;ovLTXA8WW)2@d7vV&!lDy1szRQEu8l-Q9l< zvZIF{DM~kgYIrhk#rB2dv@-BZUs;wGJP z;8NemgLz;|ji&mm5iOj`MrK?>u-&6!Bkf~3X`^J~XYPI+HqgVG7Cs=E_?QvA?-M^G z=ZF*CZips+!U~$hblX1hGjT&7&XS$5^QutV4J8vlVF&FV?K|;#Ppz)bNJ&G&3X1X! zT^?u85TK-C*ZMK3J}obz!5!oZ=5B-&Qi{;d5hh5AU|Mg{X6Snd?>^*`lIAd7P5aH> z87JJkr*{J>E8uekH=YwDE0`ErLGyTuS%N-kx63iuuJuAt9wfh&Y0yym^tvju=MR`}Zuq(Rr!ArX6+ za_|TGp@vQdGg!kPAF2F9(FtpSmd3C_;H={=(UUH6;9}9qivr%D3q7s`lOqx!2)r7$ znNG%&yQ63g$lM*>8bJNn$+0U1m{A^eF%}aEQ}LvI06J2KPBtBC4M@0yz*&MdAQL@@ zUjrnYrxL9J340KDmEP0c5&5J~eEE<66QE#7rOkZ$@Q%BB!5T9w)B3S=Nezo4wZH^R zfQ*6vK{~+_y2!Y_;2{~qq*cbiy{dTp0JB20iqEb0&vwSyZv$r${AA{oQP$QarO zRLmdrT}r?oro6cndd=_%!M_%nD-4;5BO4MlVtas*p+icVSoG*Z8unG@^NdzY8zk1ij) z@#yr)`NLm3ee3P>$M2qg`TX&t^M|MR?%%t(dy%q;ILF=HO9X`ubQPUlC7qx1Vg8bO zVf9`~CE>81>bSfBprd^NzPw^35z`;;_vu(_kDd(#xwqB(#_hcbfr$ZG8h`ywfU)}O$njL9OPiQdJI_haX9kf32i3>Z35}+mBx<&N&}xb#YNquj zi5gkDB3=MpQJ96T<6$Ya)2U4X<7#k7C$X9zIzrmFP{+0a zK)XIh+G%^344wz&FE7cnp?bOcux)Q&RLedOI-5C|-1I!?4fH(dY#tNkg?k=!(Hbci zCS7Mrwzv!Usy^*dLA+=iMoIqpx6hJ*0)^oRr?{41VLZq#bL z5Yfp3iLZNE-*$tkA$L-E&T&kB0c5%TPSPa`U$rvhesj%thA(xQJeOiAG~6M2y?hFh zOQw(EOw`YZTf?k#rN;|_3`?_!>Jc&DPObkuB=40{CVw|l4XXt(5rI!^j+0efQ!q0DR9D#Ud&(@#Gk;QK{Qemd61_U$LP zuc1Pkc#%kulPH9@vPB{jPyOdYi7ZhwB;Bs%i$p@1WF9_*V<;DRxHJ`M^-{Q`FG9Oy ztSQp!nbwa7EWW`lJYz`3T_P|nX0JKn;S?$ErbcnsP?LHF?k!~Zo$zph@*b!5F}Br| zd`k^L29H5V5gkA8(T^z-(M^qrPBf-5MnQ0gAc6Lf>wfPxvO0Ua_Hb&YYrBW~ zR#Q`*-9OaV9vdk4ak{di)G1e%7PN$|FR;BK?^9&2 zM;&0+_v&e%lh`%&!;$&|CtAOJE|w$y4=xkxzrQoSd40TlaOr=%Y2=1G$)VIENsr)7 zpMJvcnU`rz8z+(${Br?TNN^~|5lYTZ^%}aX)X>C z*;bWnxFj&_wX87IPGM3>w<^qbk^G}Rr8}bFcty#e;QVGM-l}AjH|eKrtIEVvx5c=0 zrC^O`^cg4wm~^X>q29#J4T6k*j`X~DM(XYou3JQ;?>RA{c`{0q-{U>yGR^bgVCSUD-7t_}t*GQznQ>I;}jHBL><^}T2c<(!-6z;-lU*Kez~ z;awIS0%UOM{g2Oo!x$qS+|20UM2Ec?09S3(B>)04)ycF13fwdt4h7GEJS_Df9|b!h zp%&`FkkYLK6uf8|V;?Xm7hPv;+kgRi(RJ?zA>jf;Ve&!3SPtuigiDZ>o;ocN3>KFb zCnQ`(S?LQ)hRyHzps)K53kjDA9(Jke%U0R$PGC19t9I}oB-DyI+n$XS-#QN)gXL5z zBV1ZJGf3puB@S#qS{&|Vgob;6MsE#3Mz{pxvU+^#!kc@5u0MXT0ea*KXF<~})1_wu zV-7XwSQI_{zN6^N0$i>-AieVg>7f+JhyI!SYcMm^Dqx}I(2UInw-0kH=#>6mR|>J7 z5E~te7>JDaJe|(&?`5?3JghWdevQkQp#n-I8~lX-RP>y>Jzldv6`7C%KMv`6MGy47 zkz##Zi+%yYO>|?deUyszgS|{nxM-ptR9`>hlQ!VZ7-vfSb=)V0!d}wLZFy>08^q{_ zXo`NUC6go~m=e$ScL`;@a5>0a^&O_9Y76!m8WB>=GevS`^^kqh$`|qx1 zNIy3-`Z@6e8%CX+`u+wFG;nBVZzDv$2H^~8;ATbx*8+ur-h%ADlL9VLp5RR1!}bp4 zKAY#=Kx*3fD9Dc}(zMNtrcE%SF$W@D0)>G3xGq3It@_CsNVrZx2A}?DR9ly7ZX z2lR*AIUgAsCB1xKzq7Y*TGI>r8_Wq3gn&!Le3pP5k>lZU4|q^fIKxm8!llh(ba|so zhkJw%Dz0MEvvdzCF6X@U)jxtTw~RX$0&)d~Z#s^l7ZJlTa<%LeA)vpj02$z%`vl52 z;LSeNulsn*@z&uBu;#qzZ*~hc8&8^Vb_EbZ#ao7fD1E*Q+#=+73U*p209?|qJ@naV zOU!~YAKMy|JO#EoSfOWnAA1T1Cwg)v=edBi*yWqH_a44|F0DmaQB+9NH8XvJr+8XZ z!L3D@`H@s9bjndx9LUVFu7X>>iDhC-`6gC_iME1Oi;&m!UR~i2?Wqv3=U7ykGP`_J z#yG^wdiFZG5$j{u`&)}JV|`4}Sn1}Oc6jBRGJlv(b#bYyOzVZcxf)RE1|5Wn>{^7G z{)lQ3X4S; z)3M4Kezlt!GUX7PS;oX+n^~s8I9PbK2pfhYOqfG%W(?KffXxik-b3v-9T(_SN64;4 zm>XJ2VckqS(#}Vd2??6;Y1Jt3Y0tzC-7_=MP3 zshcCQooSv|r7jIcD-wa=8{(40ub`7%w1PqTKPh zD=ar}DlF%y@f^tCglNy8ku)^ObB4N^l5`Mzsjd^V#~F!12Bop=q60e-ACG zViBPgGDCk^3EsR(it9v#bNdYuiU`*&v86vW7Q&rJ#ZwRV>iK9Po&ifVPEo?0F~s^@1#3Zh36cPXa8D+$YJIxPKU_r1&v4)0Fd96)PZ)KmL(lqNW|lw)6{Q(7!|eOlk@}U2WOlEhjtd13J0e{r1y$Ue%|#8nWGGh+V8hb4T@xFtKMa zAUz+g!5^~S<&3}T?GBlGnC&iO=Fsgf(_|br^6f6ycvX$xH~i|W&XDb{!jS))*zUG! z$aa?`vWwTY=NRxI!uN3;hsA*L7NLUs^)TNgK`6M>;ZLeZ*ChMpAzi+d+L(pTIzUsW+Sz zO)}!|gCzV(sG~P+-{>si*ToGe@#1TqbUv##p zV^A!w)^RY@Ka8r+To(;B8}|6-r}QVb%LkY5w(W9Y#QfP#BCRfUj{J;2VCl#pg{xgU zBAiGCk&-VR7+@t6l)rqL{ortmNhbGJzH}tWO1N}vrAtR9whl2jN@R)(rF2^#4Olu7 zN+tF5VcbJGB2XMsN63d$`Vyt@8v0U(bc73|BYaMJy4l^NC@Z|*f;m7B*j9=Wtp&a= z;{xdc7e)`*0HySm!(Rj`JB$o#uiHX7lM7dUZvo&I*^blq={F1reIFl1xluK5C77$^ z9-^4mrK2hpM(R7L070<}+Ih!;1ayAYf7WQ+OxUVzE9tK_x$neZ#=#@{o?uMJ^OMRH~ADccyLC7EkW@4nYL2r`dc!PJkA)SUsg$JAtq zAFJfzVxlip*JFF{0wyOBjQMSqsX?!di;;@;rlax(*#l`a*A>i3tpb+c_5^Ym}V2)TOy>t9~+&QwA`Y z31Tp7E-G{kX5BeCb&Zo#b&0<{gE7Cuv<)$sX~JM!a_YKaT3+mi3<_BQ=9x14^^RjO zn&lVD zgxmppv}ZCKL4VSt5CBkp51khZ*&b4L2iWdJEP7?27U(OqvT)#2e?L;Mw?_1r?p<8e z0S$F?`mB%i*VHTHmqn2dd(l8Ce#}-Igf#^I6hbsT`H2aJ_%n5HKdB!dsaFOKcT^ji z`F$|&%Hz(F7q*n%e(SP=jnr3<)XRgsAIV(?dH;akqSm~jD^BL#%waX1BccV?3|w(A zt)X0d1uF=L(4fV5=!%o@njta29kk+PA}`&*LFRJzkQ0#Mfh$hJYzBG%U@H!n+H}&( z_zLw)B$!zm!OZaWJ$Q+l>jB#U)r^o=>{7X0F|j35%`970Gu(TfzmuaZuBB7WEV;#t z9C%oDQqB0WHI=LazB0`Z8e1{9(KpZb^etW6p9&ryVm^v=87o3`)Xr^8xh&v4texR@ z&c_H!+FCj7%!Ve6%||KRIT~caT}DXMUA8$Kmb){F}GQ} zEUh)j(jwc_KHqU0lU->BFxJj+UFJ%wo97)2x2aX)OFqc%?_2=f=xi9CAWv6a4Jded{_lCq9n`H*RJskwAFe9uMKoL zmT#6cMYW$u)kLfZQQ5P*_WRJpav7GN9H~!s8dF<>o*3q zl*U`?MY!DsBt3R5%yZ5s>;gtH@oMw%>M%&Z3)lODn{?9G-?V~~}R?YHQ zye0Erp~`>t$I}AdEPofX+xj^^du2DPcK$q6wKM{NGB<~L*3ZzY!^K-w2D`*%pd0|a zqOA(Szr$=*Iipe4Ftobxr3u&c|LeA8Z}1(Fv$+L znQg^e6^6?B-2k6f>IP@NU-epFQtT|=suG|qT0={3O;i@}OgyER8-S4e`7gT>RV$O+ z5^q%rIhMIO%%Xm-s77+2Vx44u?>U*DDL(|Tn34GnI0yviG?>&$<>wwWHPwcIMXf5o zhKfnIeC4mTt+un5Z>^lvucZ>vtuV$w@0Tjp`018mjFH}NW%PccF^-9GeT>0#rW(6R zuyNCHc=X;$^Y;rj?{bASe=Dr{YwDq^`n5)>xhp69TQ$7ccuV=JU)r}Y19RQfikAR_ z=SoSU#OdKX8 z3YDd0?)SS&<{^$=z9&A^-BRBhDfYGl;uUdo>IZU^yq94KBac%LZ^3CbQeQvfLtJ2M zOv)wxI_?wL)ThN;=^72VL@%Fbsf7CONUbF^H-h65&xRS)%S0-im29WUP^1`lUeQj& zCgIYZCTA+zX@+p{K0~}!vYjSlntZ3pG!}=1WT(kByXW6tij&gqPHrVvC*e+mv3CYo z_v(A&Wd6m?>muRPE>C^K(}v&jk0`J0m$_K)ydb;dJFVVy^i_ZTG4yrP zbb$df=Qa4kK@GS12H=siP)+XREkEp_D~%yv^i*o0X5&dSl@jd!=6vUA>nvp73wB!1 zTFI1>%zYPt*gnU?g7z8P8glOiwmQVBXFA2m`XV%f zp1&Nae2p5yx`>>%y3-$ndaRrb6WceD43mi04~I+-zBjO;>5*U)$r*}(X+cwQlEQ%9 zC%fU1jLaRW_Dv*Xl57*nG!zFxAp@Z}b#hH!!6rgJVJhB4FdYSlsZo7S9%i>lb-yM~ zGkcq06G`}J4x%15G)oh?Yqx;9$YmKvVC7c1?0<;0pBR1sqB zo@iY+HRcDG)TZ*)y?$_!Hg#>Zsf~l%AnUb`fwEqm&Mu&%3GKIny50%A@6p&>!1 z%E@|-LKVyl-8w*|O@r$r1=WREFC8ZZ4(ef1#h z9~Ig}6(Sann4$_1bt{pqcNt^TjVlwe0UW*FsFU@sL0uhmYT$doXn)lcE9*7d+Fhhv zm@iDG_2XyQ7sQ>Qx7v5VxGE>;-Poi8fR}YZ*&g`+ydD!?&}+W$7%PCFchfWx-*QM< zuQ>)OxB0;J9Xvl(^@@=L2I?9rVSbeBeaDk&h3B&#uT9kH_?afWR$$3>)AXG&m=>yn zG@eo%Ox}QEB*U*r1u$F>wiH5KQ^?Rz4y4ggPm%@C@b34mOy9yZ8t${Lxg5V(4@@Y% zOY+3sou8^z`jU^-^kBsH+B_I$>~L#IyILo6 z<}!U1lNrM->f3PpWXAYqsJ}DlBo;K^;IB-X6+^X7<^;Gjh2pktGH2rBpp&_TVaT9w zoy-YiY4ac%uKi~E)?Xnvq(Y02R`kU{pQP`uj4;L=srM=y)B3SsF$|WHgX0i_7N-z= zh|g49flVWV0Y(^oRSQ$8ym08R@veB{2IanPT=7JvK;}jmOHwh#Cn41ci6u-Hkbz8ww2uk&7_GVe z@23|}dUOLkgtElU1dtc9Bic)QT+;b5+Kna$FWTB?>9BL~54DTl$ z1gJc;LJ5>M{?2u(o*93v;m(!k1Lk5CZvP=ixoQ9OcYm603_ZA&U;s6A*uIdQ++?4> z@~_g9641g8wNG?X(qJn8ZiT(}ga;;AkH=E?{a1bL-B#=IvKo8YpXQ&e)!&R%UMxzP zhVK1+dR)K0^Nq?~7ru5>$m*l)jPL*Lo;vOkH}txhtYiQZ^ckBhC{TYjqJ>l0$c!I* zZTIL&ADgK4ENn%RiJ!Uqao9i)vs(CoWa48+FbMedF}J>ID9OZ6SV41`Zrdk*CT`N5 zu=Cs&?S_(xpRj{=kNR2JWd$G;*1)|JNE-Ox6BD+eba|ZpB?n0wHb&Aws!z*{XmAI$ zl!Xg4DMb))1SCb+7%4&vN8sK&{GAb{t8H8jtxfaA7|04zxecEq`0+%tf{l?CG>>Pj z$sorQvIKoM60iiW>pe>V-`7)@30Q*rzP_Ij=KFeJ36)!0%$F8tC|PZ@ax7uZoE{-r zZ3S6^ZkO4BWeJP3zR7C3JNZo8@oD4L zEsqY!p*EK{&lSuQ%lqdM3)z3vMaN;K4QXJHGnO2v$*lGVB<-v72fKBw`O!4z__%gB~=k zzJ7du|L&uE7w6-n@wJmqJ*{6PkTLYM{Np2)e<(Vcg3!_!HVB+`+$DO_xdCn+o&23e z3ba>sp~sbAazp|Ifmg|Hg}t+nXOw6S$lM)27;69~60$X*BX#I9MjUN}60HFVcW95+ ztzZqvM33M!a`AnSu{&+>=0=~_<9Kvq}*a-26Df6yGm z@Q1UOQ)K`Ux*yC%@Xz2iJycjBH5Acp$Pozog8qWPw$0UaeAaqx6KwkCL>CZ5$k}M$ zVy(VG$3^gJm*bj*IHYHpYuM6|!qIN2RXQxrO34k>EXee%v%j|LpGNAJGjn2ddhb$K z62E-##-r0G=MR7F^sTqgAHRG0<@3jn&L7%R?8%q-U>|Gn$#Q{7Cog%A!GEXTcvgGy?; zu`7QQW>Z>Qw&Q4-mxFUhHBLYyCf2b}CkX}9QH#nQSQnu$|3vbe{m)Zur@nurUXX_2 zo^pHIE%oAYzj=@@%7B>l9H2ZltB!#B=TY`$RZg(QiXExDTV`HImkrp=GFaq&-^>tC zJJe>DGyG~dGi2r=HnWV0!#1-_gK@A(H?v&R)ha*zrVv@pWWT}>*vv58y|NAHq@#ZZ zY-S1Y6|SyZ>1LLRt)8wa+$gIX7PX-c*vt|VEOmERAJx49Bptitv-KH}rJ=(ZA1WG+ zn)>RU{hKgMj5Kr)@MeoCLHCM8P`1o>j>!-5dk>fhj*7On3}9H8!}eHrkf5o(Ykj-9ia{8$RIhd-in9GeV+$ioY{;&ml(Hk zFxfm0TJ4ti19~2Gah4Ee`#h-rIsYg1*3I&Bq0@WUXXM)y{O3V`dtXOLw`adI_|}he z4;RvRu-3KI))QKoFf~CH2>e2Sw6Z;^JH`9Y)e~Az?5D#}Pe}iU|Ig|znpabwy#2+y z6FqJ_>MrwzwP6C}nbr6)?_32&9ct3#xIZ}>7Q!!R_1G((TyE5AybzH~>AA_=c7v%Q zcT#xH@l2ks(BJPYU1Ih1F-C@m6icDu9?>tQBaxn9fxxhz!?uQ5=Sq(k0vVQO5!EC5 z4?enIP|uBTH%2%3c}V3)<+rF5MOhn!0UvHp?BRQeW!-sGT~*sQ@T-tZzug?rG4EHFA@o5l6koQBEidf zTu*>jZwOq{7a=XvxvSfQf-BSd@qooQxP@AQJXD&96n9BJq_Xvs))S<-Gp)DeJ?g82 zV|nzE$Lq^r-lO)Ur34colHs|56wxKfs8|rlk0}z-O^k?6G^Q~}B8>sLkNOrVKuB!& zJMH0wtj^w?`?#K1>Dr#+n$|<^BfE1hHq4bgfTPLG#Rj*|j@KS)XFm1$w;o?wL7u)` zQ^VjR*vv6gjcB2#^tL(JF4E*8{Eq*1vX8oJfP736Qsixt#$Q5~fPR|u@x^qzuf6OHyT_sSY|)CTgy3Hdl6&^vw$GJ#N?=V#jvzz%)@ zkG<^gf-AKc&jfeD!tI(2zch2t?9`rmn3ciiL5lWW&pZJC?{y3*;f+D|hP+Ray$*Hg zS)Y0~%(h-G5amVwaHPJ#iPkTlA4-lWH9zSQJmJ&&jc<JVeJXdTLs zkytz0H}EXxd~TUExqAzM|Zxi~~*TUD;%y}+;< z?_rW1ZmZggw<^qb>&A4XzwvUEp}!1qy^?NKGRm8DTDDbX;^`0@OtS*{7@KQoUngQNhI=r`PY0QI6QU=H8zhR7#4sK#}aH27`0JyM3GAk1)egRpP%Gwadr(MdTIXS7u+2ziga&sTe*>7VePes&-C6g%ml00ww?H5x z)Ev8EI_kokdw{N2M>ar@JmD;8nq|85Okm8R1|5r{hu?P;eOmWfdbb7&GAViwJ;|TQ z%twr^2>D6hPaVs`D6fqDP?RomF-}y8@R>ML_X}G z6zJztlEQ~W`ZBY-=uVM-&a{3kUSLTHkoT9r!B2q(j{gljfkj>lI7J#b(|QvW0*7L9 zP|%drwDD1pA5o-fGp$QTR4fG4S1|zs;yT|u4J2Hr(|}KZ z?1PHyDayB=ns-3_BM1v_=X_+?^h!I!<`(=+xdO2u1e{{#!+njxb;$WaqZU&a#e8;Dz2xNamPYHZsEr_9mmj%h~XHyTK0($(BD;n z3~&lEe98I$4S2H;_3J*~a=dl;VwP^u08YHXEaPZ}Yz;}C0$V+-pdVfAW4`Fb#jQoSz4!3#^AA`p!c3pwrJmMQ zaBC4}ek4^2opMwa(yY`>51^T4T?MxmVde)?6I;qRu^LRY6|7o>yr%c+3Xi>0?L^P9 zs3K8z`KF9TZ=GbeN50;>E@Ysc;%Zif0)jLaE|Hev^+Yi zMpn9ejo9Uza@IyzP3rE$_^9d)b}hox^b8}3*C)r_z;{-1hdHU}2>XT4u1JA!^~@Q5 zwVPQ)84Kh<$v3l%iNiLtOoMT-@M;nA;&8v29dZw2lpyp|dcbCeX>YtwocuSiyKULE z2y;U#DXg1mN4lA1!d6^oFVn5Ut+KjmXZElVY}~@Snf}Bocj0E%GBwFuKu9ws5w z8W#aSo=JjgMhe6D5ZOrxXXrU6(>j32O+9oH!YN8ApXxN}-A9`lC|ghunIIvYA)Gut zYjoO<%uYKcmpcieQJ_Hy;mk5+^E-ZBuXhwn2KxB0~L%!rIy5@=sbp8`wFnoXP93WjA?$%8@$-Ji8bI)+H4I@7Q7ytns5>A$=Xfkn zKY`Tm)>@+V^)ZIE2$!K8fzWWr==JhRCFu7vyu|72hI4BX?!yEs(L|y`sg^YPd5o*6 z&oS&WW65tIP>N1a)bE)gk6~y(vy&@sqHto@B24U!(P+aO3j%|efR_Gv5>JI?vgZ4@ zpPb*--!H%X;L(@HM;`^6#BPJYE;E)I8PQOghr7mTS{TB(Q z{BQ;|duY=vVE7Dn*gbUN>ekJWX3wWm=)LgJR` z^KK*4wE6VNNf2i!OM7Orw1WhZJElH1Y+|P6Mef~`8PHIXF%~3HI6xJeKhRSuUEXwfH4t^|7q8@6kwS3?Y;R-2Xc&e4^T8f<%&DL3=8F>OpnReJb)T?aMdczC68u z{^-e}B#EE)65BiKrSS_KiEX)8@LoYmMwCv1{>7fgbCethGqw2bO8x$I_46Igv}%q+ zhRt#RVX1!u3Hwzl{^&Yt>fc6cBOVY$-n#bd3TqbrELItFgjE?)H|s*wt}^wz~{riFIg>;DNh6xdYWii#?6s4%+T=#-r^H zm|Wb6GB{&b0I{!`L$sIqw7-O9JSOq_KNBjJ`tZMuWD zy971M-Cmy};po{lJ_D)+`aeFSOvH}rcSknweTD>pGa~?eL4KawNpbvwRh`$tETG~` zbwzvvtoIAPNF39HQ#i)R$n|kPj2BpZ)^;@{M>9CO0 zBQOXX7Z_5vzBjz*y&zd0ZPWbq2GvG15@SfJ7mie$WcjLb-Va#VxJ+=h%b3?P;)X_b zr0kgWoGd587dHNmW#DsHPQRh#ijh&&6pzaGk|_Ridm z1L_YI6~yNrkNcM$d+8^3*W>2}Ha-+w5F78=KV6ET>v2!&W&;L3)e_V|(UxZHItD_i zm;~xhw9bJJK-yoaq4oaU~`OFrM{~s(3OTZ zjVmY{c>R@kzdmOVN`wQy7Y4~S(SP^ERH?nk+fRAbYzgi)h-<^5=i;dfdSTl zshp!({f0RsgOo2F39=Ff^j5lbWMb=zmJZQJc=^(iP%5dX53zJGMUiu$BNW~occMsr ziQ35~&i z&QbpP+~l7J`o!&=j}Dt|YNJEbf&G-@3g&_^a*3GJjQirS_R|KSUH||$S;eGjrXii}IoXW=F<+Z;ZdrRQj0BjDie+`_7>W)A!%%A4y>!|D zY2+NF+mpZg*N`TJs(!W1(B#J*I-&2;f(*~(mn4UNeJ%{lB;>yecVS0Nbb>nN+w>?C z1e=Gy1u0}RKM^+ZHx^0i6(D6&elm14OLRihD6sZyuC?#e_uf5!{Q4V@FHXOBetB`M_lWRSb!Yri10u|JqJfnB%Sb)T147wfLbRmwm8Z8;V%nm9bfoSKusx{RJCkrOsW+ zss$)nMeZPuA}mM^3m?Duq16v)Ue|a1ug z5=ec@W@OKzXB2a9muk9i}^YE6Y~FhKKJAL*~D zSH>@kA|3VuFPG%B{xMr^5Y`a*QwY&y1TEc7i#f}n7ear0q+S^`+)-_4=9i?rD~~%z z-br`gcfLS<^+>%u$opm=8{hnt&M3cpaQQCxJ6_)V{#?`FZR7+TxZ-5)&F&~g3FkV8 zMhmPNxZ+@1({nX2dsAFgu(zMoz!fLqHA7aM?VuGW6M5I8Pi=gK`UMiqnC6d%ukXQ2DV+18_Eu3~{!PRjDZnMVRhTDMl(q4fAQnk_rM8IvLRIRn+HYQbz z7b-|Qvq`v(_Lc29vpwzei?<)%UF(|US#Gmow)3vzHX3XjZgX39?uFoni;Q69trfo# zRq{!Ack6THhORa)sLGZ9ImsTc&zAw+glwxJTeW95%k+c;fK}%cjtJf$>qGL@5bHg? z0qa9&Qzfh>GZgBFBlU&s%zE(X^8E3|`TYm?F3%5Ef$ZsP*Z4))YCVDshphdgzFE=~ z)qWyX6R{pdWzX)~??V%dwO-Uuj?^bRjj1ica;Oj~eE+Vwp)?F(Ye`G5^c#a(isvo$ zqmjx>J!^&`Hi#7Y)gR5>#rxTuPy>B>(Kmlx*5>KkYJC0XdpFGgxuw2+R7l0DSsshG zWd18u`LF(XTID~X?%-bD&#iepSRHG#YUj^GRZAlPD06d|XZ>8kGFr3_Ww1+J2Fd|I z;w96bTNQ$ThuNxfMx&}>>Gfk*5UkgLA`|;>RT=Y!Y*m@&;t-K-Rk?=uRl=(A9wyo0 zHnXjGtHMyZZcK(0_X{=V?u1TH;;kwH%Az&2^wvaW0nfx!da6IbR#m!gR0m$XRVCzD z=H@Vq`h{EFXivpD$^71PGC%$N8PtLqncskeK((;UE$=?B@>_ON`MJtp`du?%k!ih2 zM~vQ2G{%+~S1k!+e4&Ci zPNx?MuyMV_P%fl8Y5rW*FI_DkIuL9AL;+)pv6&pjeZVZ8@Ne1hV&g3ps(x({MV_~Y z-44j=VrE?=fbKa{Q!34vc=PjbJ-)m<(UW42tj+KE2&pd^sOp!V+aLbi`Mrm)KJY}$ zJ!d*^c+T{eS$uX-c?n+eNe{66s$Uu|+xBdv_|^-LSCqG9bD^rc#d}bF)=Dl-!7=0S zkh!|ld8)P_{y(p`tX##frTvx&Dt^tI;`28xaW^ZnA@2ESJZMii3xJ@ty8f6VNP$0x z3Q+mCJc_?>K|5T;*^hb@8!h1Us!i#k?3y_V1v5iluIOCj`Ff|%3PlIlG)k!-u+bq+ zS7fwXMBswi?^GvbF}ab}Z}a{V7r?@@PYH&KFQJZ8{Iq)${UdjtzIlE8)cEM;==RZj zw@yCcb&=1h+v9av7nvDw&=Yu5qWHu}{pqOi!Ug>UC4%b=MC_BVO4uQcPCigSI8v{5 z8AVGb4igfE%F;6ThkYky4xQXe$@Xul?~N3D+X3;4xH+MWxRs;iy$nkzM4W$p3r?$% z`uY(c;sRS^QZDh=ai72@$1+_Q0hj3I^X#InzB^KD$;^%5xWuzz2K6$L3TGwTX)+Y) zecEXdCO*VYlQR`@@b_hNL1vGHJ59zk`A(B*EDj0DPLpf4`btKhnL}zir8^DA-sqG< z&C|neJ86D>Pxb`BJlZ} zj=pLRZ|Li!=>lVJE{?r|FC63`eFO5yP)+XREkEp_D_uhFMNg#`YBruUQ>g-dB(2jV z!PC}t$i5fsw2rmTLE?X(BdS2JydxNIKRLgvo%rPkkG?cM`sg)7O@gf<_g-MDN33&! z&(lL1kb3v=*j9a~^m4G(!yn?YWUc+Wc3n%A@Ax=gtL1oJj^lD)AOfQ9yGQDeMg420s;$u+s( zMA9#C!{^49u}FJ9(7uTfbhJ7rbBMHu!OX2pu!$smv~ktjwr(PsSYeDOmC4e>!TC>c z?VCsfMO)q3a}!}q6wl9Aq#~A2M1RW?QC%wz$QMoP$D;k15gUH4JFK-L#jFYXEY>A; z>Rw$54%%Y0tSv~CCla)WUZ?JzqlyqK_eASzB60|+P36N;Ke$Mnx-#0-#=(ujXmw@P zl{o@R)O4U-)~g{_^;i7t3-f{$Yw{7PllAIUd>Bi*b%01^k%H<%tfrQeuE@7a?On!C z9WU$Ex@`nYSB*~F6Vpx=BD}13jfv^R5mQtlqB^=*);mqG>CzG6wU2|k^g?pSruE}z*cZf_pf?_WFX**N1pqJWfU-TPJHz|U3wl@f%NP{&lKT3% zXKy*Atk)cal-qpZ`o3uCyOsk65*aFCew6Ed$CIZg&wAE2QK#c)npm6k>YHB41Hd%k zF_;#rf;65|JWO65qLV!cB{npa18FqWljK#0PL;mRSJy3aL^WZ3u^yOEdY1@v-JPGR z9QcxtIjXwbShPyj^_o6=k-%x)tQu$}p5u$(|jM!eA2Tz6@NdWY6$&6 zBY~N|)(8*^NX4XnGK4-^I)G14++(yWR5@T}DhC8A#O<7q37afzLt*V=f+sz?5gtO( zsoId4EC9x-d1;4B7T_^j{+du0n3?BdpulWuaf$gI9}V;Y3l`K|&SFxsaF5aIM`Ij9 zqTACx-$4W|V%o)fEHDGAB+_NS;b?%yXod!SjCu6|(z12sbyR1N`Wn0|hnk8%!~01G z0cIx~e|0)l&x}9TaCIu&t|x)&m{f8Xorv201{q>Vl$w!&+Q-co|0#(`W z;u!1;$;nOjxhub(#fP-71OuqCHF>J}cPs3*Cp>Y$dOVi;trH3cyy_$Rwpx#u)rija z3*DjB-;7jVEJ~S%?)@2EPM?1E!~yDSM}@3D%Fg)y-|ng79&vNh$Ddtp4SwHIe>I|o zQ`z8)AA4>0sP8~0uTYawH1RWcKgR~pErj~vgRcTx(Zt7$;6M|fbNfeY7<~egO#Flu zG>7T7ed1^0hO{i*CIKDnv0@aqO3}nm*g?BTX=k|8W0zH7>n~DkAZg%V5fa&k(&ce> z7vD9KG?>%fR<_VdNO#pd3qByYK7! z31Pmk2bRzr7bsb6trM_ptIYTc$!asq5_rjKH_ZInvjmW=mWo^xW>C~VxkpvaLpYEn z%p6NFkE&7;;?GCAYd}pfPgsIWR%1VQ<5>cf7QmnIKTUQ}J#s8TkDXx&;IjQTpfeTOPiClpV>KkJwR{hsFY{&qG~lksWe?M%QN%o#4+yF(IQ^WN&&0{vKO zk3)rsFF-ptAsw&TPR|$w32Hae^A^A^JzTR@e?C&L@}d&H37Teyxcdgh5?=Gs`mEZF zR}Ik$f4kuzuIuWMh&@a>_yY(v^wk5LbaFw=Dbybysr*CH32T6s#;`%)tm7`xlLk7t zSakAPJBu)d9#?|N5eX0kUgd2wouWy$24wDzZVf<3HuB(b@YP~#DxS0t0HzLndop~Y zOZ0L8g=`H-xP!o1f;Au$J%?Wd3Uq;-Y_c^VVGjbY`qN~eX-<~Q*d%o?j&RI z1Bzq}rgg!9inR<{-Yeh_T;F^C&{@l{Mp}k7)-rHA=K}*JfvlaDVco7}m}C9`9{-~@ z?Z6*8rG9w+pd$bqDD|V#H2?<|34hR&XirW+)(6?1;}7}9P+1Ti5W* zF8)eCTK&9u{zp7b?D8-LyINcgd^KQq%Yc&ie|JMkK?mF2a;Br*E#(uvH!fsoxx~+a z-7RC}z}+p=VjMBz-7VMlzWPsf5`@Wj(EV>aWOu{Z_psYl2I=OvD|nJ5i`EU_-4Ylq zU23=D-7OPgSG9MQ@NhM>hwN?%DHgk&aR~5=wT}B=5Uy>2a4lV$`RMq68(DU&!3D^e zGR+?kJ4A#vs6M8S1#W5R%$QddLCiRr+h*Rd&V(mK^{WXmZ%$W;uMUl$tR6V)mutH|A!;73g}`}XfwTD? z|LgUR;#xz?ggUz?MCqzj_Kxh9Xk+w+d2eDvzF)zwHT1vOw7iX#(uLNu^3lp2)?CS2 zT;z%lnWt`z%K@g>P~IBa(x7`qbYs6Sf}%q$uiV~Hy3pB^x4(FIVjjaVXg;TnW;nwc zv8>07H&gxJkYtC-H93(P-37<$X^_oSjL|<;p!wD5Y%ak8+lL%W>3PS)d$Pn-AWu;Z zdOrw#5mcUjAk#xKrqpyt^I5B?zQtUH>4n3y(scVjG=Lvi$E8@ji2CwHrZIiW1O6A) z^W$e*0YB&*x|FH^lFv&kO(f7x8K;s)X0E0kZA&<6B7#(ZebmJBfWb@|Yp{soaU=T|DJyrFI67O9Za1yJ~9G<(X{Rt5|TD2(N zyK-h<>E4BmJiy+SF>uh{m1!@I6XD*KYk99#^mnmTiP&v&kq@$2rJi{2!WekC3csV( z2j9DdKsa7dxj;DiyUgB|d}0k7N56foboHpO;o`kZNQD!Z9boU`OfO3jH)o`RFG0Y? zl=iP!U7~If;V>iN<6*?uGZ(wsq`^J)vGg{PB0N4IwyMO<8PwqEe>DPP^69u(+2j>w zFvDFt)qy}zEv~^Kq@a$Eio+m7McwAiDC%T`JWP|Qxt9PKH;spnDWEjte?m!Uva*`Q z&6!i0?fI^59Z+MCY+fH^A2BfXU5|;27=T0TjfM9QSC}6+b;(2IPD$|Y{ljIH1ix|+ z870Bjt^7PB$?Jq9Yth}FfrC#n^+vVPMyj_t14Tm6>4P6OM!0CSceXvdp!(LCgc&TT zJ4IsE1i#iQuSm4oXaL@INd?;x|DV@+hkFKbb7sFBLPa7+qvlSfN>|$0D-v}m@L_x8 z3THx7P=hA}SNfaJjhCN1dGMu2*=mU?Bo7^M#C_fGDjtwI_jNw1gXUW6u=`4X8*zs6 zAapR)CZM~yM5`~m9kNT`*yDX0~VVs}H~-m;A`0Z}XpElW3qZOON^vS)fIa z{P;s@k3Q+;h|jBM#~Z#J(L^BVr_wLq^S;PMe~1t$X+}nH*sJ#`?o8x{@WaALedCC! zii@m`p{K}Y$DJZ64Onn#?B-g5=1^xH$t#JpSz_flB78^O@b%2`ce+tsl#m+u#&(;QR%jl+*E{kyRCL&Y+x5 z|Em!ifvYFkeWxm1!Hh)hI(Pf$7QSLbf*3MU}D8y$c7vjV=!u$jQ*Zv8Jrhq#^dp?_y%KQ$3YI)@z41GpPC$kE6E1a0=aYEn} zyFy~7wHUMUvfMKbF>6R={xY2ph26A@%+n1BV7i{s$*&=mvp3Qhie2t!fV?U~rhE07 z$FGo>HlIN#@ml3>dljW&$MHmr?qIuH&h!uB?zSDWyJd_VxVvRqj3b6wA(0o2Rn=kF zz^krlLv}X|hO0ky;GHXjj~N^8kligo#nPpgTOpAZVf6*}s`jx0SRS&wCHz?I@`LPd zElR`*soG1RaZ!orO6P4RNVaF1XO_-O!^&zA>r>i)Jo7x2l-}{NveP0?kn(O~ly@!f z{5ENAoU&{MGuE|R9YAcVcIO=egmftcDzNP>*H;ssQ zBG8znPE1sX#ku0th!a~V?x*%-*gQlP-T-dup^q7u$iA5@T+D!k`08u|6%sd?JcDfZ zdMh43cS>n@YQ$BP(!O!{*`%~ryXDtgR3TA+#`*B)&hI^Z^?|2uF0XynB;?da??E&u z{gFmWbA`ms#FhT=tuu=PqS?07Bf7P+?klfHTv=6g&?6c|Z@c0>%4pD5NZd^97e}B+ zTs2iGbQKcK_csITd2+GD8O{js*{>QwutMVMEL6_pRY-(gu+(1T>^zsPb|pPQ0opIF ztO&aZD_=cGsEXM8^f{0fQj4`i^H1EW{6 zULYnqBCX|DR85RT>jgbW$u1mpIZ45le7#7Jlq{0D39*U8Gap+TXR(j%ev_%N9n05? zghC0ud=OhHe+K~dR|Id$ZO2iPDNyC5cV9kgvK10HQ>4l>tuwP;<3Wu%6XhQD z>&l}TeaRFcChdQJ)h2(*`wirTw+R^0p;u`ZwBt!rloCEQDdE5*X6Jlx*c?#Xw0*8+ zGv&(MyqZxv6TZ5!;>s#LPTKT&aKpXuyt28LR?Vov#=a8cO8*XX0>KK2t4TtVbg9u@ zUh9ghLexIrv6ZPM>5-aI3&N;)@&l_}g+$$_hNMj}Ys_)fB9iHh`JV~KH&mTQGl)CXVKEhv7FOOej zMR>VeaHF6VAUf^(b6w3x8j$o@uoEvH&e(F*ssHo3`gBi2tq$a%v2%QYp~;j)>^vNG zobAp_>MKX;wtP^~bsJqP>Rrgl8aulofp&T_?CPVApT4ce*KfXOD{=kvk$PF~gBaKi zaOZK?`QeWzC!PJ8hj*+Ij|rX*zZVC%KB9O7T;JPAb#M47IB5UNV7*w)7BHWMR_#|6x7k zp*OS?8=vGiG)cc(SO}HhR@RRR<36&n;`o8-g zLl)^zCjsh`Nl38{aD+Xy0b%rP9-_DUo(M9a@|yj zR)0a;Rc$mC8;nX)><(uJ_<~V&kotQ_$wOtonc!yaz6MOw(-UCApYW{oIdyxyma@_U zJFgq8ugu}IKq%!PKmWK-1M4%=?yjdsxn2^PA1Xb_&3E4k^qML>$zGk?wuTt|TvKQR zU0kYH33%i~yU!=%e}CuS-5iZSv;%}EPlv?_?7n*XdC5O%gwfw~lW%X}mfCgPvNK!#)BjK1+r`?J9Oq$k z&QO=bCOu}hL~(Q^ijd4d9g=g;`rrG`4Cj|aQldnL0u;pwbZ}+t7O5rCfZsUfK%iFOw(c6 zc&&xw=vkl@NPDB?nr`C#uE(+jjxs%#>*@Tc>9MR09Hr;6vqx#vqybo^36Lla{Z5oK z$1^>a&4k9!dn~zY{+5AQv?8z$qK5khbXhiSbQLUb>h~ek;O-erojPuvc}lhah-AKHdyj~H>kx?KwV_X zSY9Alg#8gOOuh&_arpU*w=W<5*n@}nbX)QD#nkf}rnt+dKVPu$;-}Tis~2w0 z0-9%;h!Hw~2c{f44O*W2WA$p z-*CIfY5UGwCC5~>gGPYPcnvho*g%tU>Ds_Ov1ZX3@BC|H?#^n!-C0p@Vlf8L0TarW zkIp#RdMz-`*aA}`+XKs2T;WRGC#(E&u0ts~ws z*hU}ioJM7C6$jBYK12g8IQr5>Qt0TAEIdkh{ha)U2mep z*vXWIXkZ%*K!Lz*bkQYy+dwoL>KP##ZbQP%E<%_F^ua`@{$rTPyDOwRQQl*qn)v-u zpS4-)r_{5nTkt~Noz+TkgWC+CZ@j()54wSPrmpNEYsqyi_90a0Sp1HJ1$uPd9#wUPtid27FW_%F+JKSE9 zxOc=9s#4<$E&kDhZX3^;^K0n4+sy18>#f&rGY)^UfVJnXlj92X z+z7cEp&w;>Zd$MBrp@iS;bwgX)!BGw#9GsH)8_Wvva8Q z>u@_246M_F&Kv%I2;d%NI&WI9^JW|!z&hDV;vETVP3KL^g@+uh(|WMZ2Al?q#xJJx zW&;j_2{+q#DomNqo6YPXSUbTw<2=s6BuE38OE-sgHs%c2hI!2ZMBXKX2G~ioX~hr@$#qybF*xiqzRzTl z9euH&;kd|S8)#>9c7SvtKbZL>9H zcIW=}<->cIx9?nBUmol#wD&yrSjI)XDUx885x)lQx5=KQ)Er?qlK;qSSQ*-#>EdtX z6akP)`uel4J+3~xlw7B)Uk;`xRLMNG5|k*M0}NkzT~H3GgQHKWKR;50>qxOi1Bl64 zN1HQ$!F55dgIr%Le0a^-Q|cSX)Jd-t@AC#Q;Sy4B5+tV&-y{gEIDiWmTw2_$B@`~T+i(l6Mw)N2??Oh`u z5Vq^&6;V&wP3E3x)vuw#>&G_f$AQ5pdeL^LCN~Q|(Yl!qvYPXrok5+)Ct5Rf$h&^5 z51?oBMC&F?4w~uRom?$=dng2B_}7nV9C{jW^|)>ZJ#nL-w7JO2m}cdu!560c-Zs+# zTY%K})16%{=v7SbT66VSH&q_N)nl9V1Fb-{e#SIiYRX1*UbGYw3mte#C(VMsLB%1} z?iwoM0TMD&GN61PMn(=6KIK0^LQX`Wv1vwiTI(2fXMkn#jm6^Q2Z0w0)2TDFnQXfw z!7{cj1RrmzU5ssc0Q^Ts>L=Ld1l~Orz*O_p)Ib>TSJPEvPrme?CmvtDV|Dz*(XHdZ z`_$=YY>VY{>Y3Fh(qf4yIUogQTrt@FNd1c=>QRpP3>4C9SPbp_4xl&$gEZv+tz-4- zGLoD?iv#Heb?>!I1kUe=VEbb0A1N3=rM`Kj_+AL8Ese3qrbF<81G2xDHz;8r*cm|}4RyS6ubs^Yd+BPyPF);|r z^zCvz9&mgR*Ml+ZVdlbao+bU$Rta4X!d!vtK}MVFg~;__<0i54g^b8Hx+kTs$A`#~ zdhKG=ryU9m-lDe*h+R(ric5^+R^)mR)gjWdm{)id-fAm#Jpm3bT3r{#Ev9<06}3&s z9cV%hgLwXVc@5U=4(xVb6Y@CviQJ#NMd9%h?Ig}WCf|9D#NF6PoO4XRkD4ya+jtyz z4;FgIN1h7`N!HB1|IH+A8@YrieV*J#`YhFs#TYIvvPJB`myqgpMB>Y8$G)V+^tM<0p> z^00vR!l&amv(xdt0kpZNzQE7VIx-%cKkdE{w-go|<@OuoyU0^JebHmk zz@o>p)Z^0+7_9cYURq69?ePyj@tAHWd+c%DD*dw6?2Q^$T|~2Ks?|R`qFRILk5T|E zV2}hz$M*yGZ(h9TC8giIcx-y^ViBhPh&(Wz!_<`h63;N_j1*+PL^pQUnER|Z#95bE z!B~0nsr~CDroyZA^(SYzQK#AQ&{nMpUeE9}fu}{0iqZ7zNaDQ6O3agH^-qu0kjw8J zZ_QpX>eL{;#Nj3loeRKO7g;MDV@=9V{qB(>Zp(@AHv?I-m(+!;i}0T2Qs7k($YO3w zOx5e~&Vr4v9!tCm#;S*{9?oHSjFBEo3A_rzl+ddnBh58|I^#g@A#BVfmb$=ZiMB-U z&AG&@_&{oMKaS>bel%tg=H?zFcq$oja1yT)K&iPVaGUK_5REysSD^yuFh$7 z-DAZA#8vwa=iW5xa1?2d|I53cH;>u%vBS=NRquUmLMLs<2m+dmPb zhs_z(bPM<3Zl5sRW#b@=4X67!7 zA7Zb+!JgMR0L1Qo?SF5wt(|O>X(y63l&RtEC8?8*hv)(ob;cb$YJYBAUw#liF z_ImN57o9fj((&Hw^%_S%k(fr_LO3p>ty+A{#asc}d#zgi*s7J&u@>zHx?zpC=~(N} zyc4&*x!tkPeVL>Oy;orl$nRLQ!t{Et#?do{ImcOIo7sEGMK`4gAeOaV+Nl0GwE=e) zl%Ol7SFQK1Ea=B83tX=nm$<1{tv{R2C{O!JKOno`V7w-#8=zNh?PDt30j6HH{tP^} z8*0^fo2MUx5-Vx0{M6&E9)I!n?GtnOdTrno;bXg-07;o1+to8g^}uS2eaWStfa7_u zTD0d4Z0Ii=ut39XBUtbWZb}5Y)Ouc1;|=wzFp+jQK7N!rr@;sSPFk7V;ptv8A>sNXwEbYP(hI)BVou&U08Eu>E|HI~@`Bq=n|fmb(- z%^bxC7vX%XL^i;O5cvEcogT}uahf&=G8+&;3HjQ(71#hod%n#!fN1}a*?<6A5UlFL zHXwSY^oGl_q74yzh(X`jv(JFMhKPP_h=@0}(k9g{NX(`deU}FNDf;rBE z32M+Z?VyHuQF&$p0o*Wv&V;Cte*>oG?pmvlN8tv4-U)J8pM}6Zs7HUw=5(#9S7*67 zoggC~BRU5=xC4LrF4jU1@#AodL~w$`a5sET-O;~%PAvoxKMJ=9r$itX-)O&iq(05m zP59zzc=p8PATb=`Rjaels#~jdj?Ti#Gk}OcOx2v-)`K&6NJLQj>`$)}! zJ9;uK)c^4I1WpWI|HC-?iP*q8BNz?LZHft+AUw}bx9VH4DX1^U^*cNZX<)ewMQ=|xQo88GM?Do!4nlYSP}wX z`JE7zf3ReIHo+6)Jj~tW!4uUSo~X>3lB#AZgt@h6z&jz<+$DR{z!O?_V<$wF6=gzv zg4+)9b(q7uYhL0l7_*??02I~f^ERRDRI{qG6FLD35e5Yg6j3_9vj~NLb=w>X5qMey zQB-jhW|_!Ogi(0Bx)UM-UJXd0cY09qb&Gqt46mYy>)#%!zlm*)>gL7u_2vERI}h&P zyt=&mw{Cvrt;>gR-~96B!~2(aZ(iKKxOaXpgevC8>P@afT(F>T;@REC^CMQ-UsTVp zo+Fi=LllhTqgMo}e|4n3xFGAC3l|Q>D1Gt5fBv}o;pLc|E?_vI#+)oJlGE^-pGCzG zIe>X%cHNJDBN`3eIC|{?XGo^JC{F<v1}@A?rhdF;IBY>%8SxoWDwyCJB_x8J*Y zh}^-v8^-!0_HG`d2k~wQlZD<58F8*LV(*5HoxE;n!L1a%c4)taRe^W&ao|PiK8!0H z4gc%{q!RBIV8c>VyA^vkM8h7HccT)QgL=0BG!`%Uk}iv}&n%GQz%$U`#>ZNN#y9l~ zM+QXpI^2e_!|ge7+%{qqtNP%cNlXp&2SpEPO6GkQTP0*URw`xPCtJNj+5y{+vCp5 zK7ZO*)6v4|1m3zbxbyh@X$F_@LZfvZfX~zOr+T`aK{eC&#w&-`(HAxN7aINT=|Uqd z3Gn>sIQj|tBx5`?+ga0$C7K>U?W^g~Bk1(g6|RXj%3Wn(Z#eL5l=Yc^JUHK7GYvn77n=)yptU0<}*v z6h>+eT^^HQ$E-W2DUEk5ejk4bb-^H>0`Zf};p$UqD3P*T5poQ3BBj_2Ko z!iB?cN={2PEEb(ExW+g<8f!fWO{Rlx)t?^WZ-X+}oOLD&SJ`>w1s5r{9Ah<1W-k!n zz?A(SiM^OXg;6HO_QD6KVsjwK&~l2N-N-1n7Xc(C)8wttULaZ(???&i>ux(Qx&El3 zk=u&^g%Ymv0qh0ybmba!{e?YA0mwuiEYoR{&ZBkO<#qjyqn|vJfjvQo=Oa(CO>04? zb+mae|Agh*Yu+12&pBba21!knd~=Yl@=jQ;&HHle*6Bxd9E*~GcDzui{1cW|s9rnX zdThs&33ZM$MY#cw59%aS0x@yd->zHqBKiDNmg}htZjV{6$BX3cg--d(;hnNvchf1$ zHOShd%jwz}YaderXl8T(3nwjn%ChoLS*~jXCyl2pV-(q4q4z%kMw$RgGN&x7jFYW< zY-K{yXBRrHxi_m_8IVL<#StVyAB~GerK1fr5J~hUKmhywO9&XaD}rujnqh;D#BqX` zEN-$^{i@ByO-6f9pe8ZNkC~$bMcippnftYnNc`00f}uPbhD3GU=h&Swm2qs55U}?b zV~+woVge1*xyRI>9jTvW>Fk#;57pTyUXy?%h>FBJKYsg2{n%n`bNW6G!zA7H@wLa9 zscNu@AqX+`d>l+qP8Jr0by-MTQc^(nb;a5{>Ap5k=B+pBb+Aij7i1G{sN+M$R>&K#qi!5tDC9X-oL(T< zlc*bI1U-q-r`={pUEC5zJI(m0vFS(jVJ-#zoupB-RU^g*T5Z*0Zzl%>*Kk^W@Oy^(3})B)K*R;_K)m zi-@M|;N(s8wh(?#qVBz$#)CbHQBKuQwb-HI#&qDj&6d2H)0wPz?d zDzpp%iX>Xy#VrG~uW%FX#N&fUI@Q_ds$M7FIQog)+`2`H?Mj#6N89Q6ppkhVa^tnr zZN_%G9MA})2T}4(S7Gg)NZgqF?>5}~?@V{xCfXgxhl;F7z3#YibdgBsI8k&>U{9ew zP$Yne+PIxDM`)e$fYaS!FD4%yGA}o5yu;m_@o=|8hq6^;I>|ZhvEBef+*UK!Q`q{L zP=;>p;h=>diL9sdlIDnNYv3T+Ba!PaK!VI8k#+7Pk(&txpZ64Um+UJB=FqYmVGi`} z{h+7N#`K`0a09x(qPym|VE*fF)ca&sylc>n#oru!!6uHqYFBlB;yC7kF28r@NhJ=~ z^b*$bGlmO=v;<%gHBmmd?m(%8TZ?4?>CnU zOlI97#Tqf_%HZ9Q$Qyg#mtkTKq^EFEd8CU2>^r0o_B>E6fU^$Po3Z2y@$ZmA_@BdU zyi@MTK+HR&kT1hcA6irgdC4zMLZ+5hDHQvLL}b!=kD++UT|&8}I;E>5c6$ zv}o5JA56Aa=WV>Uy$%2Byil#CUc{s*?INl#C^bUb)!A9J_l}P%U#lm~fct(RFn$)Y z46^tnKkD=9oqOocJG8#poLFMg62=wq)K6ztWo)S9U6dF~ zdk!|ImkBl{>chAK#5wEwJEvUHUNb)K{3W}L(~WUv8OLrjhdgJg&W{?^nU1s@jj62e zfRp>_C4NO2n78v$Xn=2~N3orW3mtE_or$yK?VW+5y+?%8^m(0$eSk-sGwwBbNjN>I zR+q75PursCJs|wn#Q#J+VFsLRk@>4u7hc1Zx9f7-6Lk~q&MGxilQd{g>`yFd*6yTD zl=`gYXevR12E_(El$BoWpEJn4E>4Wdy>K}!8=%^go*dqp*txs(cy}g3kr*7m53vEz zlPA7U5l$rr1>l{Dox1`-8Ua~Qe~?lXU8@D7s(ap@iJiL{LCU&C_~Y+1FUW#CSZ3dW zw%0MxGZ>}@2KtEgC5ex;1@lAPeyKYXgGsy;G`s*`-bNeD6rRKmx)C!tX5X0@&@qQp za||xf4(4&~aQdJRmM|Z&;%jI*00O(52J-Jf3)!4?FbY@M`Q+ueidGBxcP7qRMu4zK zYAXzJ}V42{iwIX#ZtOYJ^u(!J?~qg_vnByiia9B={}_B^A*m-G&R za>AWD_RLz0EqlCk&vW6<(WjV{6>N3Xd4K?T;tIhnUG&5_A9MOqt433=oLRc@$hgCt zxT3X6V$O(8n{Kd50)V3En!0|+-b0$2IvQ!X9cMtMrn&jH2^oY+Hoh0FK49dTn46y^r|kI>FxA zwLBW^2f$-d^5HFtox5+3cc#-#ZqO16If&5^dVc`0h);$(+PpgjJ9lnX7WNpBq)EK- zeUKxKj;ZN!V?5LHInF)6I-kT~BYO8g5~f6P+W&i`{^L}i?!I&J=IzV-SK4{cF{tp5A<{fV)ihq>;&%^%?=>IVGAG!bYm_3mH zLl`agf5@0~-6`;Y*yzdA0E_NI##^_e0g_t{|1kYwywhlnXAeOyYpb?p{x5)}3zoCM z{~_AOu)#r6bSO_pBt~WaFX*2vSb7(fKJqPS!MQ5hZ5V*lHIyX*m zQW|f#l}bm~jy2FT7@o1_!`HTvvhrI7tFdJ;2fiM1*f8C2&2&zG!mz6|15Z|2H|)7p ziFN?;&+7w$`z?Ug*a9dQ_#Dkn3mo=RYDi81X%7suWkW}|lF{ZkKCW$hDaVJ!H3)rT zcFy=rmczAW+A|&HH`-HM?a8{-()CgfKEj>Vg=%5$g^u-`Udm>6tbZ^tw|A^Rtw>%k zl;Z>^$&M~8v)i>(CfCPWFJ;>)rzUSu zB3PR~#477ia@$Thz9K&E0oYENrTKry)CF{rXsio}7sN7>)x`XNflC8=cPU^az&cEk zAl_s5s)bOWRkv1aDTFEoF2+pS-jdL}TOj0Nk*yfCQkp_A1yo2D4Ylt1?7K@M z0mrlkvGk!BrDACfr2lZ1YJu(8FUCb9NPNl&R~^8f{~V326A{MSb+&ys*?JjB+k z>!#{_^CWap=7ZchXs=S=%!LX8*K>?kynN}C>y&&$4YNK>bWEL^hcHwr>ArP!d9LmC zm+#zvV|DzoBlT~N($9cXB+Ox1@oF!-FwrHjA$5TjK`@9%Y(-{BWQ=1~U`2e~E7+=h zQ-*iZL`FR1W1=uuNfzViLHfb*g>oa;bNohIbozzWJLmSOT5}~s?Ek<($;KU zBW4xtTjXO}W2U8Z5q%D>@*5Y&3DSo}34NMBQxY{84r~DZhb3T^?Zq5ll^zo<12&BP zhdE#tw>i$X+4-kBD}C;6T3e9bG1me#6nf!QW-lg?uisHvjU9!_!@z`mCjcwPinSkJ zJEm^65@1(cIXF#<9-HUGR<<2;e5n~~(s*v(!B%UYDbNQhrz4u?VC-2#C@jO zwPUV*u!`H{XKjezqO6olc*ccFgsJQ;qX5SIKL~+yoGne)VT} z&b2*bj%V93$LE7aAlACadtgJRe*J;Bfqoyt5ANE+Z5Y-JM$P5?Ps zkZVa$r!{LsmnD?dre9d-BU5tG9D?#+RhqSd#7;pyT_6WjWw76(a@jm3F{MkNP`DYQ5~7~|z2v316vmxTo#!Y|>e zSygCp0#K$z=eI=`2hqGoX>mmEz0l$WAWhNg^TO<<%@F`=>S&h^AFTOLEm$+YYYFnv zae@1BO3DH@<6D-Ht+CA}2e9F0awZXP z{ll}hy@pG_a%llP=2}EMW%xZuK*oNnOg*;Bl*oAAbrj_bo}tyJ#RQ-YR~8nusiVy! z0?=lB{sn5&*gN7t8!mlTSNt=ybz?vq{onylR~%@w@j#n4_Y7?d&d`QiN&wo-&d_cQ zXd^pAs~&zo)!|hs_ zSxFDL>3AMR{V@bI+>KP7C=;5%8_jqOpZTm!Vn3yxUEPu=u{cH(WqizW9f0Fb4wpD@ z1LCw*l` ztCwW@4$1IPo4C5`Hy^kK6lYD~Tpp&gJSS;fhR-aL)~UZbR zyBvkD=E>6f7g`*E>ntxW0kHFn$Lggr35|A?_yizNnFJ1J zeB!sH0XSr#o*Z|9`}mk6%wWf!oCOTNTdJ5r1B`*joIGg93oroA<2UBinT=+P_sjvxQZ!)w#B!UYa&cg{km$joJFt$&Jab7p`Z?-W^- zH`txCsRE$K93X4CSLoU9oblfJ(bmnnb2b%oubVsOHBmdDCwb9eP~_dGW)SItx7NUyJn&?hsg!54T`X{0Z5m<4qCT8%`_9Gn<-u-Nd(UH!Wn6=sA_;~m@mpnn zo9tUk%@KAivEG&C@A4D@kV*PBnXf&rKD(4$r<-IBrYBTcJ+%^)D3WKMpK`t^ykl)S zNvS_SQp8o1Db_qlQ5Op4r|Xs@GJNVlhr_!kpjYp?rFK`BkNjww_% z#rwQTOt^&9qX)_9xsM(s*C4e10C-$Zg6dmGiinbuyv}e#;ZnQpv>aFIO{~Pe3&CmT zH(MdNVbr6<9!X!+$!(R`cVTRMx1kA}KOWQnTh5DbmDqP73>Nt=WW2ea2z?he^8WU- zv{kW~$pL-WR_?nZh;*8_L%KQykH%e{+;;_Nw9KHA4{#*xIigjM#s^WUM^fN&-xa{l z!lh38nFDw?&{Wtg)KsXShMi8IHGWg!I05xsLqeaJPs6%!+F{KN=qU{GRl$vBjo(u^ zj-S(0sPU(yGFm|^bE8>1-5O+9)-;Rh8`Cp}vyiEN&A@hA@)7s|jaER|es|$Gfl%0U zxH-w60=JoIv`&IK15hwMom)lFVJHk)#x_J{Z!;S&t^@U%#_KQ?YL0(~xUp@9P5QDa zu%GO0W_p1=2pXKFO|P8Ap%;D`wlxP}M{r#C=2ZrV&UzX4DFtsc+tdLVZC-|LH<_F6 z-tI(v10FjHO%eY#v%hUFGi#<^#_8qRar~1u7i<|bFKckpeA$c4y6I#t=%Lg%wLdWamHO}`L+lu*| zdSsYO^Hn)@lV;C?mGW~ePbUUZa^l@T#IPeZ~%3auednT`d!1#*flH>?HvD(vIEZ5Yrs+h zeRD-%QEz?qo~<2(dqQxoe!Syvg#D&#HOKMYx%zw+@z2#a8&fKYU-0*o-Al@~Th|`e z-9^vUg8>Z8-7S2szVpx3Z|Vu=9_NWh0**JY-MX10+?_dB-!;tfY`gV%^@YW50Vd2l zp#PM5XrNoo;|Sf>d3g@(vesw2V|&Ie#&1C=(_NMK$*idjV%vm?AwV!+uwYiN+Etwo zW)XO8jL|Xvb5O0jFF@jg4N%)9);4yRz(Tuw8>aBJ{cz2hGNLy}lceS>_C zdFshO#2eJQf$Xuzb&u7{R+~3!SalI?ritOgvd8M59Z@aB^yorJc4@|FIQEaoqt-c0 zP5ChK`g6`mK`u;mZD)gdU9 zaGuEaS_#rzhH1C@r^jl@<#&$2W-l0Z$dF#*aFfQ0z0>Lh1{uJwi--w_UVryU5%=sw z_?yA5*-N^Np-roh0dFN<1pzVUw#3xIh?kug^eSw`RpeDLRz0*eOow)bz6I3MV{B2c zf-oiYD#%E4O`y&~kb4LlGl{M)vfjyEGhWA)yQ6(*^-%VI39}J4DNDRc0Ip^)ShiO| zG$st=SQ-)7O;2XEo+iaDzS+G4|1;UlX~ z%6?mHJGRB5kT7$BPKvc8 z+;Je4TVQ?h9T2Iq-B7=Nq=={LJyZj(;tUteTt(*#7hpKbTK2w1lJ7^duV?UBl&S1} z?SnRl*w-Q0623r_VY#yRbpX@kY2-F@Un5#lZW}2OCZ-E9Ym&dO17s$16&Kvs+$%;~ z&=3};>U=;(Cl5N%*8Oem+VABV8^R>k(XB;Hey&lkjdpYKkr?}Q%y^`9%pFWq(%8+F z)7i#-#KT-E+UbRR@%C+-N84K>^8Vj3AA&LLHBKg0&dO&lKu} zzQpLaT*3!_H~OTV0C;I*4L~>COcrzLqL&{8*a`{c*#~~dMn$+}Z#VjlSI zxlg>}HF19Rg+kRghY*-3^i-Wlm~PgLs;G z%VuC_>gm-daxyR+5s1oI=~&Ie1#5SzI?u6Mups_$?*hTJAbapQbjPXlIJ6kSMzF|X z5ad9N1;VoNR4wtW$CMwr=Qmgn^2PvyPh0hTMr~F@T0POZGeYe@I8sEXYK%D#DDYf# z)I%bv$3hIJVFmTikEyeVF;3@zfnc#t7SW#5g;pwmi--HmK^E%wjuM?(sDjQP&=sue zQtJ^_SQ|ZvYygrba=?T|1f2sYh-`olDx~)x-zJUvHa~~A%$mq-Kmab}YwK2E0}$Ca@g%C~RH-0$5`kf2i z*tw7|*a>_v&M2z&9{8Y-1PQlKwZ~a@Bv(q z6;(9{K6D=V&_@r{fz;M?c=QV&sOx73>N*2H2p_8L=iozc4%GG019gB^2R^{&1<(}H zNA=g5FZHGYAG9)z;6uFbJTr(M00FueqGJ9Hn5?@ivOXRKApD692+@bG8YHEkvRPiM z>eX3pmM6%Om5%Wo5aEsr<~v;jM)c8xbT^{>FXE7irU#!>cW5x5Qv*uqUA_pxlGrM? zjOsU!)Tf!c313+aPpz1&grCXBFpO8Nc0a3bt=2i(4aro0n5s*=tp~sJun8oUSctsv zxrdTYhF?2Y#6#QxI01Q`qZ9E;6~104%jl3b+^0(|@FGaXXb#qrO)_h7kgV+TOA>Wo zGIM~WnZ_I(ev{{*6f$$*!xXXg8%^yTc34K0v%4dd%p3$jg+OVy5_5p)&_Y&V8J`L8 zN|`wbzzWIA^5Wj577k?i3-w6+oP{g=9*J)3kr3NhXCz~r6L7P1qeSO7N_4r665J%x z;uZdBmCiOwblg(5O}urDIdYQ?AFr^I9gEojyBWuqNLJh}p~H0nV8NAuTQ|@x!S5*v zfDC<_Wa=6kF(|lhezyddKC2=A;gsHXOX!@72Ld_3qW1v{F1BLE%%8VOU8kB=m7UZ{Fbf^=In2TxBoX*t3tF@bxmN;wh--=t zF#YQZ9V9d+yN*|Pu0+7Ag)KV4n#b^lg_drs{+A>5;%;xe`QX9b%ZFueqMoQH&-Xd< z`R*=~tJM?g#;x%d_T$)7a^o#Pwbkj5sgJBaV2l6CsRz06 zebMe(1_Ax*A0Df{UCJSOx=*28XjAbcXYT!S#m?~D@Gr|jE=rdy)IT{=A518zhctPD zj2x1is6~=Z0hJ*LrOQ^{l|1|D)PHfTK9b;OjwPkIpZr^nEJZsGhk;lt>|0xq1?r9a zq11BB0HD-zIFBunW0D<)(L}QF`Sm3R*&!~n(MEB1Jz72o9*`5eDg2mWG%-dv*VGh} zP2WnBHkt^@Jfn$xI2T&VXktU`ub4?{l~daAjx)h(s6bI(h#I$ z7|j52OIV3*%4j086c|kkHB0M55Ta%o%>X7#Er*D=G#4g$a;T9myho4{e#_2XoALWC zJLC8#ESKW8T#^q!{z6SU_MVpBZ_?>=n{=YL5X=HHd+A0$-H>2F6WgiNgHD~CRavNA z#|hPM*XhT0orF;5y*X*};bV_FWf$OGS&Q#LuZ6mF>|>83mIFaAze@)+U8L8<+~m{y zhsb){s-w}I34M&r6Z+8%~2QLSBxeY%_9q(>Ym*!fQarf`k#f^XR0Wt_5qwgGHnOj(f={Od^ z2sOZNrnD-O9$H`7Z$6vK9pDz0;VfsQ*9=xkD=DE^3FL>CuH;SB8Z?b#gLqJP4sQ-g zZjK-gkE+HtzBv&2eY-cuHn};1%sajU+1oRIABWv*AY+ZTY48(n9X{%yQD4k4M$NnM z(fn2CW(<1~-OC8w18Ax6HUdBF9nZ|T7pBI56wp16pLHB&&E&u|4-EcsnBG1g1Y_2i zS$8rVK%0bcEq3>CSQ(DA!L=CMB%(4DHlkz_N^ zpY>wB^WygH^M1%a`7+Wrs`>od4Q_wee`NASZSux^@q3A|2)2x9Dh=ux=lZ{e*_OMn z;k)rUOq!S=J%Ei}GI9DtHkSXO`oQXYBe6X0X2~N@Jsb{tuq;kWk zweaY7M{lC^o1{?hv)J0^hPG5;AWHv#ue%(k4vVi|-g)Dvudi-i+`hPXeh=ww*wvoT zFRsq7FYjMHcxVD&yxusL8+Y}_k69J@e)WB;lbwp7Gs4PHyvq1rAE{5GLm)(RYKtNg z>5AfC9;uHq$(!nkQW2cwErYWnVMUO?1|JNXa{iin_pNv7rkuxj%^K=|I8yIJ`S;TF z?=sB{@LQ+@N2jTo#_coj>gV$@zt%wM_YS@frF)*9#8Ol?Uk3a}2G%L7=UEy#_d z6?s0sabfU{!LP^f_4N?AObOv0k2Y)d#3dCb{^$n$j`< z_{?Bm?if10#KD??s(1;;;pFI7#fPyfj>>V~Igng|yUHN}x z1=xrB_pR2LLhv5R8^;G2M-GivC;S65LvL}8Awwr#R^|@RjgNKYD3--Lz`+@zUj54O zpLMo(W|$qE*%-XTd&?!MLtk=-XX>-F0P`@1X9jR&DLJNX85p7m4f!v*{7qANAIw3+ zdpiht9RK7yXMyRiQ6NR5wU&D0PiAE5MY4SYdO1wxeE>LHYGh@P4-FGQ4ci1j6HIcT z|6q^$jAGfLJ{Y5x!kk_5_aT1bF2s1!U&uYuqi2r-4nJTs_4ljyt-ce<)P*sF`ejI^ zJIaLmc8l(`>FphZPuST9&XiBHJD`NpY!MymSWZeoqH^JKvQ__;E;LV)hqYC*p#MGU zJGK{h-@07A_oXMFI9iPdYaTy+$L{>ehpjSvulkx4b@IRQ=DsxZv2=|<2*8T-HvY=h1wZphM^$0 zh;||?U#^`L2pwB|R6B!25Me4)JGZAXDr`)=236wlqCkH3_aB;=Sii3YE2HE1C)9Ry zEb(XsRoswPe%5#2(&|@n*JBlzpY@Fqh6Qh841C5!YJN}f+Mb4AgVXS_ft=JpwU(dB z1lK4iCpEuXyB@2xq12{CFG8YS5cP)5@y0kXKgVOQu3HpvGJ?M-VS4v@s~-Nzi1m0f z!X&OEd8~D?!Z*Fr`&@BtpLSh?dS_%r^{FYT8<+DPW5czn>VgP5_PxpH^p1SnIGu18 zjj5U+iV#bwyNdVu$GG8l&!y~|z%_MKJ~z%|)giW4Ix_^hOBJMY=GBNbbH!gZN!NM; zqWtQ0ea7Bx@4R~JwR8RZ`!wdgFsOhP>oqv^5tD!p30(Qtos54s7}w-3@nY}y+i>-} z)N1wKdAN$l_)!u3Lr2!ZV`^&gz?7wJrgMX>vUmEzA7XGD! zf9c_02KbjX{0pR@f)rGcf(lYlK?*8JK?NzOAO#hqpn?=skb)XgP(uo8NI?xLs38S4 zq@acr)R2N2Qcyz*8c0C{DQF-C4Wyuf6f}^622#*K3K~d311V@B1udjNzXa|yq=gjd z2cN;eKnhw&K?^BpAq6d@K)+|?M(ZF29i%{CO$~kxQlRf72LA#n=pY69t|;(pkOF;2 zkoy;XX)pK-q@afs^pFC5)Czn9Dd-^uJ)}VQ7P@Z+NP%vi0{;Rj7$5~2ya&GqDbNru z_!mgQ04W$C1v)Zx->e}8I&1*{0x4KS3f7Q?Hjsi1qyXB)2HM01+QjAzP|%zK3Ys%OL30KuXwCoy%^9GeIRg|lXJ8E) zXcHS~6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~ z6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~6B}p~8)y?7XcHS~6B}p~ zn-=N;w22M0i4C-g4YY|3w22M0i4C-g4YY|(2Q?Ae#HNFq2yJ2mZDP|wO@ubFfi|(} z;a-3?v4J+Rfi|(}AqCJTHqa(E&?Yv}CN|I}HUp#p+QbIh#Abj}fHtui;BJIAv4J+R zfi|&$HnD*=u~|cV0BvFeZDIp$Vgqer18rghZDIp$Vgqer18rghZDIp$Vgqer18rgh zZDIp$Vgqer3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>; zVhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTx zZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>; zVhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z3vFTxZDI>;Vhe3z2W?^p zZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#) zVh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v z2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^p zZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#) zVh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v z2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v2W?^pZDI#)Vh3$v4{c%(ZDJ2?Vh?R%4{c%( zZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2? zVh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R% z4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%( zZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{c%(ZDJ2?Vh?R%4{f48o$27S z9@@kn+Qc5(#2(tj9@@kn+Qc5(#2(tj9@@kn+Qc5(#2(tj9@@kn+Qc5(#2(tj9@@kn z+Qc5(#2(tj9@@kn+Qc5(#2(tj9@@kn+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c z!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG z0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud? z+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c z!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG z0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud?+Qb3c!~xpG0oud? z+Qb3c!~xpG0oud?+C*LHcP?jrR;e$kC%>d`tNV$Ihj%VsyLai~!9q|VHVbgOmT@5lP z&rL(jarQ6-#yD!EAc7vaI3{MR=x?2mJgLs$jbb7D9YUfpS-A?|!sSChOOj7Gj z)T;NmjsiW&zNa7@O$*iio)+r6?>@M?I#M)^-pg!>^#jub{)Ppu?}A z!>^#jub{)Ppu?}A!>_c%chrEOwt~*Tg3iB!&cA}rzk<%cg3iB!&cA}rzk<%c($3#e z1FCWr3GRFEL1R9s9>;A!C;|+!9oRtg-QntjvCM-P{F{Vf`LN?1BVI* z4iyX>Di}CaFmR|~;84N9q0)haqXv|QD;PvnFo>vN5K+M(qJlw01%rqR1`!nuA}Sq3 zSW8?xTC{<&bZPtI{tSPo0Y7iSpF8m99{hO#e_n$>>yv`U1Arm5Rk71*qvHMyNYEz5 z{Tz^>4T}3YAVHfG_j5pkHYSz?FpM@O_Sdx~aeoFRXglJ54!;E?Xd~i&2aupmi2FGp z0eq_Mhy4|8KHQ%H3EFtLp92!K>2N;>Bxu9oehx^`X2Y%m_;qbG?5}H^;rdh5H>q0{9heDeSLkL*f1mNYG}&{Tz^>jfDF-AVHf5_j5pkHV}3d zz>l?gu)nUYgZnceLE8rRb3lT&4DRQE1Z@}G&*5Bv1YjVvO|ZYNO@jL~AVC`h_j5pk zHV5wKfCOy}+|L0CI-mIDPwj5py1e(=e=0kw=W-_pWVE;)^}x%D|d6>(H_@>l0hsb zuVhn#=}gfmmX_z4wA|Jgie26W=SrvZmM78yECnxSQSjxZRQs_#EXxt9)LDc~`% zJA#c#UiP`2ok#wD_5%LH0W1ZC>fX73efjX-|J2TUi#aQ1?*-r5dAsPmq;1dp%+5A2WC!&d%dMbj^1Cr2|+B2)#!jMzOSf8jEbIC@q+{%Z9PkyuhV~R08n`mWG$|Xh2n?U<^ygvq?H) zPG0o2o%>V#xl#Uw-{;Oh_te(WFz26tZ|6^IQOebeN3b+7d6;|a+%=egZ|7%n5OMe7 z5iAXq79)NWarmOI?cCJ*BXei%fyDQAe%ynev7El(TRSfvV1T2e2s`#nOU1?Ibc6Jn)uUYAR{-<_UG&(?L#q7P{TRSfb&f9rO+n)ECoehQNXS1O)&Hd2Mi2C+s z3E$2f#8QHK<>(-otcymmw4lZTp@o#cbO1{M>X=CiP@RwuVW~jP6r>{Cw!MLCEvR)tXd&e<9l%n6I%bjr zR43#^SSnB(2dT*R?TL@=%&2!kKSp_%eV#e{+&SCb$eVrsxt%?!JR;v7EMBf}4+fY! z=S=?RpJ(zPx@J2+Zvn#rECqz#BQBEb+uIV~j((h@_Xy$ebIUjC+grJBkA7-rMg24K zDQ52l-`aUmCvE2?ZF}Blb~e;S&t`){vD%#bp`8)+?cEZ-ojHi51og_%K`>btjbdp* zjRQgpDSznzmIBl`CPj`SxJ(a(#O+z}z`!@<0DPlmE~)+xdA57!F`5AoLz_kzC*2mH2k_;~c$52#23r zzER)a$$fkDQ#&i_pOH^7doTFb&Wk!}J1=S5^FFh)p)Pth8x)Gw=G+hMjHqw#m+{M_=5`u1M#+oPY_SyBIte2Uq7!MAo^)JfZUN!yoRerRVzefzM4Z)Xl-DM7t*bP!C|MWa|+P~(8mLdst{fTaL+%p?VBv zQjzW36Cc}|QTK>`jPfq~JahKBbGEyYH~ai^J9|=jM7}*(yjHwK?}gJ0t4b*Gu?z<{*|5)GJ2^!DL-DilqfL4hSuz{G|g} z3Q)&PQh@4&drB8Tk~m_kwTjyr`45^OCkb?=w3a>Y`_}L7`Y}&i&BNi2C-;626@|h@}Me%F#hE zSr?6BX+ezxLJKK>=>V1j)G?D3pgJKR!cu|SI7mgdZ%=$|XGYy4`Z3D8?DNdo=g!&g zM&9i6&+Y6<?4;9EN{>ZI+wq;1dp%+7|o=-F&gC{~+uKeRJ` zcz>VHn-3n`)z{IJ{KS9Fj`mSxG?d=dFWM2m^QXYGY35GQ@BFF#g*)fFoi)8IaN_;Y ze{}%)H#@_3oA17QdFPFvzP>V(d|>`_e@OhF&8X^0bz^*z%RDuvKIA<>R!>iLeaBCY zzjX4>-T$3e>aMCDUcUL@;r01j_s4&^b8+|1&tBd>-%iiUUES0VpMCD?bMBw?LvyR} zXWrXu>x+h&ats}(EJ+~FW;2%(owve?WICKbP@!pr!QMMZPc*pfSfW0VR8EhQf_;S z+o$>5j!x_JMa%1;x>b|O>u!j{ayk@oIzCadcl=^b!4ND?e^)BgHO1*CI5IVlv7A0{ zCEBTGRk0;%2Vik2-lrRiOV&Pxc@TW(lyu4B%?xK!-#Ai(%&(tO?6NlMo`|B23Wi__ z_D{vEk}*u3&|mM4&S|%P+G_DLYO@+RT8s_DV)_3RvrL9JDP{I(-N{$rbizSc?EaaU zT{7}%c@lOJ!T6}$>1V8rpHt7QHW@OW4a8#m--y{J!=x^k?S1g{Suh5 zI54YZbi`s6q?uAy!4ND~#on?eBO=UO1|qSUCevHe1F+bX?^QDXVeM6zL5laPHwcSe z`CcVsAeI7yL{qB3UR;~*?oh5CmDHOyQ!H4KYoxiXWnQA(LTmUp(8nX8fv z^W!CPu3D4v5|*nXl+HpNX0DnZh$UrluG)|>6qc(-sm;n&86&Y67aOmJjH__th2&pT zenZ8`Y@qtJc)X1y%)GiF4JKjHj?%6@5&Go2%jBjX?}SA|0H z(y6>$H8l*2WwA2%WNegDW{>6Fz7;5QI0%bfajx2v@e-D+BBGf^P8@2vjl#*->OjUn zti76@W^x;agY|$j$-}*YjDc7R3=&Q0UIjz2BrDEU2Qm)Aa#gtAUe1-5tEPrwu`E|+ zGB!#nv&Zso-*PCGD048AyskQs@e-D+BBGf^PRzP$dLYi)$#d0#jG?exHA?M1n&Oaq z$~@GXp02=nk#QAnypRMgFISDc8+zbebxp=pxc5a^&GWw0036+r=c;Qmp2Bif^s!HF zu8M!mVo#i_uE`h*%T;F-?&QbJRk2T5tci2gH5oTyxvHgaW^tKmDD^puMR8qqO~y)C zu8PpOx2{yT9rGM3IWA;gi9?1p86{yML-5R{Bv*|O0@)&)?Zmn2nv9Sz7eY|F+nLPx zH#Wq9StX+*7ONo9l(GtjV6iIBRo7%hgn7$ABqNkNJ}hM<_#Ge zrIguYdADzwbybq#-5@M>#dXyU882bEDnjWja$@GH>48{mi|eWzGKRu()hM-Dbydbl zEXL)=i;SyqTUy4R+@<*qYDu1Y%15S4AJM zrmQG8hs6hliY%|IZpau4%T;HT?3S6Ct74x*7N|v1;#_q@#!XnRYU!IQ0 z-z~Zp-a|KTQ^T-$F3@K(NJ{Cm z$Meo23-mc0gvIZ-fs?bT|3}PsMMhIt+8U)KD{WaNl+)b+EIxlruFYg*g=Mb2 zd17n-NYf+=DmiD~z=RJ9`BuLn=Z}o0u+()%%1!{x)D`;_@<3()|5DBy89`yWt0iz| zcbT~>^*M`2dG1<~Q4^NCB1GZgEw6mP z=H28Skm;qJ%ve7@2#a6&e$Bhen;^AlzdpL$wI(AXEO&*xdMQ|5?wT5gCEo&lCZnX3 z|M7U48{$%X8P7jHa;MHA-$)?#dX6#d(1R zBO@%_3nTelM&ijm!ZXkbZiyErqbuACBfRc*GBbBYnb-}$lBqm*rSJL1a@PUPU70y5 z{xOR`dG1PI=#AyBGZJ?KVCJsar!3y&xhs8VH~<$JZ;K7VS+Ejb$vBC{ zD@ZV?e8u023!s1sR`88xT#L{7q zcoyO>#3}@ll9Jb5>AStL+!Ye!rC@ovYibyleC7H~U-F&OXOHKd7?`;$$#Eb{5_^*8 zuJmo+Sni6DI*Xy0xodhLmYn6eD}C)Zmb*sD-N#iNdbvki)6*4rVfq$u+=3zb+_akdmBrIOw4*L zIWFYBpl6uAyBmugHqqRb89T%Wfqap%gFF_dFZ0Gc2|?;^cQQRmY=}ein$@io5DF4c zDX(A%7O(Q$mA>X1^O`8HIbM?c;u&cQ3B4R=3o8)!%*a`mGo8M`3aNO*zM8 z_tMP_T6Y`tCCl$x^{Y0QU%r(s;FgS)c-(^QQ_L+GgC*NPkjs{gkUq(h?MUfPDlb^c z_Nrage977oSezE@TQVZz?OQMZ6z^Ma6c)#FO{Oma=V>y?LB*O3#$a*#uf>`?khoKm z(qu31N+hzcT0MVO-CC_#dX9|`S3}9Qn!XY|sns4;yFDz?>Tndu*#UAo`gU+Er}O;p zMC&Z5;?T>jN2aG!e=gT>5|!$i8tyocmkB;+mHc`2-0GPeC7(|LS63C-G7_aCY#EZ| zO*@-!%jhv6SEP>1z4AatT!dFfD9!ZB)Ceq2e@CvBmWgT<{pw_TG_5}MmW$zTx=H-}A)!{WNY8TXRMD>HI z3hp>slhG8K+amID>DOzQ*B4Z3!cJ$?Ge8|E$ZhG1#nIe0O4mNt;^wf7p&*ack{0B) z^iAW0Ekkm(%-pt$6nt6S%Dpmu$vEMa5pMZAn(vi~X_P#-rSBIhrH@OY{5H`&Qq% zqj~l#tabkTQd;XX@}9|ePlW7N?E5XD-=$Wo_d-JXD=O}lsn5u2*VHR>@3qU@x9{A4 zQ!^=di~(bm-o-F-+uEFksj4`N?Tjzz;Jh!@JZIHa`@zlAHl1Q}HXe!{K%E?6}} zJs=g)Xal7YRGV+ylza^ zSS-mFa8E`zLhgC;I$dswwsM5CuE_WYLpTX}jRCfiJo61~#z-u=i_fCe6&VqUIA=IT<6dICllZQkjP%9a`pO}dVA>6=3Q>fz=~px|G%Pqy~MwAVDF=V#&XNdor#Q zaxb#^1ouqY|H85GUDLWIv9f1)``^J1W0Q)p06%6ez$@yd)e9LG03VN~0|l!<#(D*- zK*OZVu8yxG>KEm@K*oO0=IVln%zeQs!>85Ds~1Je;0?&q2KhSBcZQd&gD)luLOor0 zu}X4%`hIadDnhz${J4M}4<<)qX@EQurf(e=ah?$mCOGFQQ9-Ux-$agQy$HQ94q#`! ziE&u+mgoBPedGeJGjjbX*GyT9`>yMTjQg-&MSI)_cGVbv&+NO7j=_>~0k>qFm&xt+ zw47U}Y{i|KbwkE=pG18*5r5v}9)*}m7TN%oTnjiQW4A0$>1d7Fc1o^*FKhY9kea@k z9P6RC^xzRSvpqgB3QM-)Qv)@9FL^e{Tke5?D964msUS{%`+QB`Ope{)htTQ;%FLVk zQp2#MEWWj{rf((B=Xra&2+vG8%TWjVK5`6ofae;5*Gi2KW@0X^u~-r>;GV?Xgxs?+ z7Y6rC*~_yF`rdLZyC5ir9+qk#Wae9p8K6EC;Oz8`<|5A7c~z3b!^@nE$5@=tYrf2u zU#S8F)0N1ajL$@z%Vj=iiYaq>R!QHCj%AgBj*j%gUCzvOhzzruIc;H1zfXp z=>*qIS<5|INA~cuJQ`BQ=2nga&atsrGB4mhs}GJrF;89#E>Q_iO5do?vT*a7?>%m3 z5V4NXX0YTgk7PPBjuUY%x9#&aeL*0HKFl$3|G`1m-pYRnvsCHsOEAn`Qu z3b5VFH0rzBFC*)p~!x7vwWiPK7cVw*;s~98t zilcJO$}oF8mJY~QfxfN1U=_Bf%dP@b7v%L~`Yv~@UL2L3hu$$u!TbSP+EB0#WE5Dk z4qI+XoVyN8eUL{*^!@O7RD@Ik{kWhQ-$~8@HKQQcr*DlHah?$mCOBluT%PsPH_78! zFG4Sj1K7EKVjL*f1LXQW83PKq&dBwnTr*`YuO#_?#Xyd$UPf#VQ|ls zy*#fP$hZs3E(nUDk7kxf86&aeUcfmSgNZn2=T%A04=-~v9%FGnulcfTD*@->NC%cV zeNVlJbGgjtoHJ!E&!y>`>#?j7l@#3N%v?G-4oltzT$AXTfNORxo#2`&Yq>{TlXr-) zJQ`N(X4X&w&atsrGB4nsjNpXa^W?>?n=_WEq;5{%bkDMI^P2BHDy`tw(+PH=ErI$_ zfRnDtI8Ma5+_ulxkf%Qdb#wamd#q{I0Y zE5_{cpsXpkq6$`ljP(jufn7Iet^!jRreSAGBh>GYt`thg;sRH_O0UO^*j>OUcIle>R+b`lgBOXj}?#q_6gz~JHzR4fY zdJ%d#aR3_wNQ@8lpx~7KhKvCPTxaC^QSO+sme)%RO07a>|scxQg7+ z?;Busb4w>4p)>2|iBVXxE#R1p#IiYN*Uh6GGi5BUAvg4!2Uramq1B7LnRWBjFf1t- z@Jym_`8>1h=5d~xau%Zw%^6uc#ZZT4!MZtXESAIzxF_Q+A@^*|g~2^j_TucKIV0mP zEW02mhCZ5^S20Fn$z9Gl{qll{b9P>p#Xa~7ve9gy36`hA9!)$?=! z`gmr2oa7&l1?5cQZUx+vah;HRxF5t+G+f>=yOid^67-;IucrZaqMm(4tiL-|C3UX7E@gSb{BJ@Jv z&Cc}`;~bc4`eg*3Ba$u;g06DH*$EamucnlbkYTDz75b?2| ziBVXxE#R1p#IiYN*Uh6GGi5BVA=B?GU^QffRxk2q*3DDHu%ukTGZ}5=^USWB$9ZPT zS≪Z!Tb{L(8d~bEL$@**Oyy;OzAK3qtPMI6H$ertIZ;75x?imRAuJLm$non=?k@ z=ukmkMZd@(;+!WbZbq40o`N%p4gD?yi}QKSmt9*4m*LJPSbM%;1xP$iyaMdHIco)&Iv}@t^!p4ctLNze^zl^Z1haBXt_z$4 z>P3Oaqu*!{axb#^#7Z({FRvKW?=)Z)V?-`EDo4qmSx;xr7c!bWZn#!*UjU+G36}B zT9}K`uRz@tKHoxyV=KUZu~+$C-6==4dS07pwpo%ZXQjT{mZ~08gH=Q z_QUGMQQ3Lu9Y=WF$G_RL7OVsPazn{FuO_zp7qjiHQ-q#()B@GjjbX*UbGbuO!p& zFkqGBZh^XabPSe^3%DiYyi9J{=glMBGG!}2g-yS=fK`)cl)@d%EQv=(V9B+BQ!;kT z;*^foN>{*_wfueahJIH8t0G&8;1N2rZk`y0C0lXbyrth($mW<`H;;1c%aV+_#C7wQ zeqRBrAtSVUkvFqGo*ITFMnxVlJ$)SQ0Pb zo{YDI+_Nzk2KP+ai_zwmev1Lis|bprk7njojFDJ!FW{Vv!9<+1^Qt80hnG2dCV|EI zyyjc{25zt-;S5k0`qZq6KyC3|t*yrth}5U&8cZq8Z(rVbR?d@`m>Sv^k&;C?OaeJ0lh&H?qJ#O9N6 zosfG`4o9pcQ}*JDaZA6`fK`kUx!|ZAlnrICxhJi;lge+aZ|S!h3RZ!APdk%ArY?x@ zCvWNZ8nAkClxh#XW1ctX56ICjdELCN$tbX79oThq?m95_K^_&+Z#dvl5mE*8;{rCm zlN^bq0R^0A#)AoZGvdMING#6fSug!o1D^FF^gFVZbWMeXMTIk-Hl!0lFk3cID>TfnNxvn7W+m{~QB zjKI;20#4~S7P2^HpEoBt<>`gIZr;-GDqvM)OD7(oGwbGwQCPB-*UcL;63gb8?u%Gr zC4AY-&tA9m`wCbM8Ig_`d2_^b$}FxY<|^TtjJEQ5X4lQ*?D+DeGS~vtfqruVLmh?% z>*lPnIEqukJsEEaxuT zr!aM9j-4DI?tS^~^ez1sgMe#xE}h_xDQkHRg?@*D<gU55pgxq?b~26=aW1#*a}Jp@msgEjvSx}^ zjiHu$l|COIXV%S`qd^&yd%j==$XHIi0_?guYXz7(P+;}Qm@Z}YJRN{OUXZPjW3l8f z=bnC}LCF20IUKPnzD^WWjOlk8u!=FFuQ)2lJWhu%>EKeL|Bp9pG6LEW5wxuIko*mZL*zf66Q$Ak164tP|AQ~~|C zfF1WGM}m?jcR~T@nekwP-i&xKITDL=d9F{t)qrQc2)!^4VCVFSaSqKj83PKq&dBwn zTr>B#yl&o+aUWJm9%6NKW=T9c21~|rZt1rdGPz}+H;-`3l&$y_c1ypvfK`)cl)@d% zk;*BvxK2+~P&e<%*e#1wo30>dDzaT{JVK{}GG-MyF*?+m5{^kcE}LWa zdGjb|JS{1xA=B?GU^QffRxk2qmc&!TKxvZEmi)YVPexn$Jo5y^?G=bGOu>7oE&b*K zhB~YltedmOVo6+%x%6bbCFGusvopBosZashOuxl| zog9ZH?*gvLm`%VnJC{yy&6Ks=qYYVi_6F2qwQedxXVz9&b3k1v=!2u*Vqp36kbAj5 z$HszOlJRSSS0ihvEDJZU`O0m3IMSgxC*wE~=e%9Yy~du{0M4Xt3(lJlBzA^XjUg#s zMbF2_sU6R(r!!}PI#I9!^xF*L6=2uRS)B5;sle*d?=z&Vo~Hv&A=$lBB>!+MC}+~@ z$vxhhjO&Eli)=ozl1$mlE5`IY4Oqn(kr$52G3(~+@mM-gunJ_XSFj4~x;b+dn7SaZ zo3F{(533hPsrJx2W__GLAWItx)`7(FO4fm0H|MScQy=8PV1llUTt7JyXc{jLI5MYeR}5jwLzo*0EC+X9ZsNGzLU zcHKP6F;m9!8Z!O90#-vtX!Rm*X5BnB3`@!dJd@E@KF{p)=5d~xa+aeG^qUJ9>aba` zZq6EuC2=wL9sT}-ko)4Z3XE0pbwZw9bo5&cSYAa?27NR$e`1WplDjyg?9Rv-OvE`m zuS#<6%a)8%3OH|8@B7@nFxx$-97S63-KGO=GyrtcI@x@|(*$`W*(WlDs)!-8?!rlzj=eBz~94ExT?W;mVgG znbC@?$sPUP0#;3)Q3`i3vu++40m|7c!W-M zx-jeMiP53V<>$>i`fY`5j@fndC}&I=i)+Xo{k{TLLq=%zB5!8hJT(lICb@&Iv;)lIZmcHYr%F05a>e0Y8H^8V`& zZeKpUdH3S&%ZKOH8K-dmBkIQa>vu1%u2wIq=U30&een9N7sP|HWM05O8F5MZ=ZWle z!A07tKbOm&tfTfBB7=_lb5{D#tLIkF}vl~tU^`S)P$w*Dc`}TA(9(gLHz9rx9B(~PF_WLLlxJ5r>cl&ednbn5B+vfv8 z*^<7!fNv5@lkm+G)9FHsvsDHDjjW~e{2Nx}7P~HNFqYiq{L`;9NcrE&U3bVoPmv02 z9sNcF&(_Tg!T030E*y#_d%1t7Uuux?USQ|vyffwg^T*P-o_??4R?5zMB|slI=Lqcd zN3BbES-rS=p5+qgIaR?Lknx>-4Ytd&$JB!Y`$xukN&Dx?1p0i5{U-G16Z|vfFRvle?>JyJL@&ReNX~=pm^D`ZL{LBEcXxL!8UK~61G|>W<(H`se^2fg z$awHMreARS;6Z#}v{wIB^~&m{9IH=^2qjNu1oD-j-*+fl3A*aG^a}a9QV=83Z#@uk z5>h1f;{<*_oE(a!1@ah?e&IpJdrpj);N6!mxdY@mG5xLskrN{X!?=K-?j{Ce$-IDX zGA@+x%@b1w77$&Wt@>jz_CvqhaE9A{FKF?1_{-MZZ`82rBIZ6a2uscd{F1R>F2CE; z1^DGDiuyyjd`WEYX@-2^PJYo!cTnA`;Yt?{!IEtOuVg%z$LsdAm{+D;<#kE=JqD~U zsYA_Z{@J`XQrIx-Cy+E3|RLUMywZ+GrPQ!!Zy-f68lHSc}e@{=>gof z)KLjZenU+bYB5NN#_QCe3OeeLr4;^xKei+ksEmo2yM z&YsBA1Nl18uR)Zo1HbOh-r@@g~vCInh-j*{)6eB7!se?%yGDydQhD?vtvMJusA zU3w*$Iw6k{>9-l?`c*kT7LXi@r3D4N=f;Q$g86m#>Sl5%7Vq-mjv;vI~>qX=o37v9_^NGpi zZ?tq9G6EBDO_$)7Vbs?F^-sh-QuLb-AKdjwz4_q5U47ggbG@}~iu#!P$m&C`DJt>} zM;A)CAtR~;H`~*EZhRS#nc+W{%Z9|vJ`lHYv*6Ryz|CSUtbx3inFq(AF~4ce)WB;lbyiO;gp#2o_-Sn zrF_dKin->$EM4=Se$QYvEfQ4a58Ji;UiCez_rtY>zs8Y^DxjH+8$z15r!1P8>-_7# zbgVvfzMrm2L-4&}a!$JDtTKb_>b84yP z)KblE7hD%syUrhb2_Q!bW+Xf zq?*%7HK&tmPAAozPO3SbRC9W%=JZm{>7|;}OEssLYECcJoL;Iqy;O60spj-j%^9Sc zGe|XOkZR5#)to`9IfGPl2C3!@Qq38pnlnf>XD!v7wN!J~Qq5UQHD@i=oV8SQ)>6$` zOEqUL)tt3db2d`V*+?~KBh{RZRC6{`&DlsbXCu{|jZ||sQq9?HPGom-S7&G9FDvnv zwfM_M{ADZtvJ-#Vi@zMiU#`Vpiglrf@up-QYnl~{)=u?|&Y9je4SREc${66;VU z)*%fn3oV#NmZe{cbx32&(mBOCq|s&RoMIi)__B0Pu?}g3SvaSLn5AEebx4EE(mBOC zq+w?1oMIi)K(lmCu?}gdSvsd!hcwtMoKvIC(l5n2r156yoMIi)h_iG~u?}g>Svsd! zhcxOeoKwTj(l5n2q=9GYoMIi)(6e+-u?}hQSvsd!hcx^wol~qs8h{qgsS#-Dmtq~# z7_@Xwu?}ezS~{m#hcpf?ol~qs8i^LpsiA1;mtq~#V6=2ju?}fCS~{m#hcqB9ol~qs z8j_aIDb^tkN(<-IsI>G;u?}fmS~{m#hcq%Rol~qs8k?5RDb^v4P7CMM@U-+xu?}f~ zS~{m#hcrYjol~qs8l;xaDb^tkQ%mQR>QF7lsp~VT4%KH;9jecyI#i!Yb*Mg*>QH?q z)uH-KszddeREO#_u?~rG>ROCb*J7Ny7UR^l7^kkqICU+?scSJ#U5jz*T8vZIVw}1b z6~I6 z(*4lVImJ4pJEEm?igif$L<{E>=Edw3J%04(Y~d>6~I6(yh_bImJ4po1>+3 zigiegQ`fpbT1c%Jr>=E}v~*6f4(T3g>6~I6(p}QhImJ4p`=o_)igD^%j8oUTRa#1| zSci18v~*6f4(WDj>6~I6(hbwnImJ38#;I%FGcBZ6j8oUTYg#&|Sci1qv~*6f4(ZNm z>6~I6(!JBdImI}2Eyk&9-99a)R;)w1fm%AJSch~AwRBFg4(TRp>6~I6664gh?xPk` zE5@m7-AOH-Q>;U}ms&cfSch~swRBFg4vG7z8!=AZh;iyhj8ivaoVpR?)QuRYZp1is zBgUy4F;3lxaq32lQ#WFqx)I~ljTonH#5i>$#;F@IPTh!c>PC!HH)5Q+5#!X27^iN; zICUe&sT(m)-H7|C8!=AZi2JD7i-PTh!c>PC!HH)5Q+5%*I!Vw}1W_ft1w zoVpSBQ#WFqx)I~ljTonH#5i>$?x$|VICUfLr*6bJbtCSlZp1isBkree#5i>$#;F@I zPW}I}_pULPW!HJwR8P-z&l%28y3&lsMU#X@ksKB!R@Xh})T2pC&1PPJ=Mv(l;4Wb6OZD^ArMgdD zs-LGW)qUzx{XBK4?o*fQ=c!9|pSo1{sY`XAx>Wb6OZD^ArMgdDs-LGW)qUzx{XBK4 z?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N# z>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}> zrS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8 zSL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lf zb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz; zQ&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(Il zK6Rz;Q&;Lfb*1i8SL!}>rS4N#>OOU)?o(IlK6Rz;Q&;Lfb*1i8*XllXt?pCT>OOU? z?o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT z>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllX zt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G z*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~ zb*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0` zQ`hP~b*=7G*XllXt?pCT>OOU??o-$5K6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o-$5 zK6S0`Q`hP~b*=7G*XllXt?pCT>OOU??o;op`_%jDKK1^C=}*71d$L%aK3+V0>)U_w z;e(xf#n!u%I}eJTqPRQ%{^@dZ=Z)uf?iX9{OrC$^hj({ARlIt9`N8$m53Y}w@6Des zmXnu?dkBZ^w>HO^U2aA*A)%@yo{^)G6-2K+& zlheoZvv)49Cq@6?+dq?mEwAT~-;Y&I>hf29cdN*a`qt&e@_O<@`#h{s-M=?k+c|sr$=T(j`PuREdhvAj*++}3>)plqqs7U|>BW=f?uVCG zC&&GtXP=J+t}owTTpVA2^mO5ieJ&Pzdbw;oczNLqeI^$A*8HQz)jOA0=ev&<*Ylw= z{l7n%pP$cX57q-*Jia_xT2eU+>M%kZ4$=N?0nf*|{H3en=l+ON0vKd>)6=2#~ zQ1*jaqEMskPom`PFGprSly6Nz*>tQy zvLDK~qoX_Gn^g&Jjl5+!GUJu>^Dd}|8I{+iDINU|LfWq&6ZxddfDkd?E4e{}W>h*ou!{R8<- z_Cxu0bd>!A>1_53h;{~){k}Y#{UWBF1!X@#M4?96pG3*ozdthjp?qr!%Km*i`y52Ul%FCf|(Q1<)sZ1#(ob{3TV z01<^6Wq%SSXaB*-?1%ELDJc67=%w zQK(V&CsA_tAC1g@DBqfbvj2$A{z$SN5oLcT7r6vwKagEJVh>!6?WHe!n_J|Y%k1EK zSc0_5a{zAv{2VpI4WJ#LvmsaloJwm*J4cyQ2`z%GVgaa~2jz3%+zlj(`97&$4_k*{ z@PmBuC4>e5T64aH^$P6X^vS+Rh_+**{BB%?9Rn}wfN)|p_}7QSM<1XE1xTwt2gt34 z!3^ouFc8nL1_9FA09S(#KmmrUK?Jq)z|{aCeOkAVK)}_Ige6u3yw=_1EL;r`(3%5R zLw6M$JS_rOLnPXc30FfW7&!*61|Yn4#~ba6e0q6#);tCjX!rW?Us`2eUB~_S!^P>7 z_pX<-mp6ZI|8etVhxicG?)+qaetbUTB~y-mAf^ zes<=(DgalruX1+wtM{hbIZF-T)$D6yXAbrk3tGx0z$n*6*V&76DFc&MC4l)6#j7$2 zTFnsA>zJosPMt*~0jn9JAJGiH6{EygGsM@r(QyurhOB0&xd3Jf@@+f@uV$YT4w$`* z3R=ws5a+c3^rM9GItwFbi;L7bEWm@=Vi;aU5awHkR#65-#%)-l&LNo@%r*p)o{WnW z>B$&|Yn!iTP%Oe75Y;vcn)hljs~^MMY7l_9wgF}FzJh8S3C6W82@dueRNGjzluMv+ zgptL2Kn5nQN&xdCidSV4glk)Oj$~_F0)T5<1n{juwM`VqwXGA!If!ao5`t?RfCTx5 zYFiqESF``v@2J{F0C8RmKtD<-ud^_6wzx>)9eFTY48yAk!hFlnD$0P!xD87bu5E+a zhCtGjagicD8N+aG^VJNBMYsc^+D1Y1UJYjTV~A>-IA;&UwG9CAzJh8S3C6W82@due zRNGjzluMv+gptL2Kn5nQN&xEva}HZD;tv(pgdf+o?iT}WZA$=fZHoZD6{xm};<&bT z;y4FUZA(IMZ3B=X-%xE!WAJMBAL<`f+Xx`eYXRs-3FUPbM$Q%&DZC>OW{Y8X6+xJ9 z8LDj=5E-{&iNdvQFxwDFdNM9jq$gt-u5G@WL6r%2Kvdf(Xx^*AtbPnpZ4>A0fw;B- zAl_F{Z6m?Bwk5&AUV~~Ii3!pP!1AOn+DC4lvT!L?2NkzuN~ zb-(swYg+<@}#iv1lomK(z@Yi}!#GOj?xy z)&~aHHt{FKsoK{4mY1z<2>`Bb5x}cuUd{f` zZAaBM0*Lck0Qylvd7Xukv&BUU@5qDMVi;aU5awHkYFh?G#%)-laBUmRHUyHMjEfZM z$ry%fo3Cb2Wx^d0)iw&6_i8Y!A462z#5sE)u5AE__Z3vzNHDH#NpP^&pxVZwrCb8l zCX6iJ12QmaRRUNa7+l-L7X+kgTlY(SwzeezxVA+A-wIUQL~&f(I&qwXsJ0~`xV8aE zkZ-8Ar7?Il`#+{1Roe(4&T9eaM+xP17Dmn%7b(0W4`z#DcojjIZyBm>84ww_VTr=E zZ7|yqNP03ZQluwi7_M!;nn9HbcR*CzC}`fR!K{7^QEe0F?18wp0U+L2P;Dc@xV9z1 z!Cr%E8;h252~?XfvUm^3z@$|PV0~b4Z4+Prk*aOodpOwImH^<|76E)KP;C>%ac%3w zaSo!|mW1Hi1|UJcq1u+l;MMHEDq~b_BY-%s1)v`#l-F4pIa^$$@Qyr~Er#J$1Yy2q zsJ3N5WZZ@&3fH#5Y(pUF$+$?7o{V9*w)tuXRVLg4QEj83d9Mbu`Y}YcO`Nj_;@Sp) zcwa%ajRfP`mIMcT4XSM{TFNC*ZNkXnJs<; z@U1|#O%%tqtrN#Nh-zCBf@>Rq1o?()TN;B`v;QKTQMHW#;=C4sew0vNXJO=QagoA1 z@?f?YhF1}U`Ie#DmI0A*8c1=ThZjB8sG9PBlywy|g_mq4`%Ba8Qd3`|;;0M-Wv*EaFhU8&mE zy+ezwZ3zIbZ4tn?0@XHA9M`r^9OodaZAl2OZ2%JF8>($-%xZQB-cAs_6iX~_TI6a5 z-yt-3eaHGMSbm^%ICj3cQk|1}#<4&l(;J*`od(6mZ&<6&$;c0ptUu#=#riYG zt!58%IU1Cla68mv3Eu|u`mxNe4FQQOPLMkrB;IIH#bME_+1E)w_j3o^j|nbj6(~Vr zgoRxpgA-RN!1<9CR%sHof}j0s;u2&*Xcc<(XbQ1{AO4FGz9s0@sj@t-NWFZY6aI^H zP(?~(FcU$TAotMuAOjNCxBmiyakY;?3Rx~7{b<1~&*Iqm;u`xSwQn%r5Xt&8u3@Y{W1LX?d`%O32&{b+mhf#buOCZT`@~s$q)__+67Myz z_K|3z_9fB5{sU_t3m3BrRF$ww2)jZCC$3U}>*I5dd9C{QsC}J0&PQ1Lk{F@(0Zfp4u=b@vtId~+pAKI@x(?fUPGB|rbjucl5quXX z|9LfgZM}cE*O|Im;a{(2lZ~`L!SL@^J1-Y6oZNiax%qo!y zUmTy$m+$X>=jLnM!ndk_%6+N&)!8TRKR>xUxpRN(;hpAd-JUwn?`#h~-F?4!@nI`m z{Ht5VFO1Pq8>QUV z$E)1*St<9wK3^QEJvaHpch)if&P|B7$LH#{HKDfcVYd%4jlT4ay zCyVd5e|Ww4vE_SLrx))Z&%b}VY**~$&Ku9|{6z7};_U47>C*qpA1jJSizlb+h3&*H z79U2~SBv|Li<5q=)cXC)Tg7ZlsK0%6d9r)8IQigl`-S<_i<2hE%Tb9qVhR7*JH>;s z+D-1>=(w#<;O$#*h#36W2Ycz=;?Be8c0O6W^!{S;^!ViTd^x%E=5sqmarfQRlQv0; zzpz!5W8Bbg`QqMX-}Fs8>+#;R-S+>=R`J#tX9X3>DJMG5T$OG|y=S)d{~ym6Up06y z+Wa+HYuI{S^tNv!0d8XuV-bk=`z>4+np%-B0&fg9o)?R|5ARH#Q|>NO2L3_sF4N@h zk|>2ccxi{6uCOiNjdK5_?k-L}DYj;zyeRr8!c9)HNEE|LK@ZN0R!zMcf$BPV&NAmlwJC?{Vs#?l`tdf^?hXza7cIe|@oR+b;Ju|FQiunYguyvaN`2 zY~v@27vEjaeRJ0Q)S&9WRNQ;`!sPk;O_OKif_ueZ9l1s~M%~b@L~|)sEk5g(!0WT0 z?3z2p@4NqtitgXtD(Z~?f|ggO-=y>ZmooToT|PN|JU@Hq@_JmVc2lig5l3U=;$nvn zOYu7+8+?E2y~S12;A{V9Tj!@(?Jo|}=C$G{9{%{`<@+u|zrS_UD|9JT6#wbS1@fKd zA(_Pw7L$9M{}|UNYHf{`ey#Y#ay{z$6#YW+{PtF-;!8W98yNND#U~%W5*Stdxvg=n zuV0y8H}3y#^Mc%)qH@do{olA#RGGXD^4=*kx$pjKd(;WcZC0Z|E`XyNE>6GJZE+x* z2HCso8r|$*H)o*v_2c(8osp(B*0i_Yb$zEh4_?}NrMUO*)A`l>rg73Vr43cMj- zEtZ!bTs0l7-F8D30{Y`__qsoO@&3I@bBOXAzxNVs=vRxEH+!rI zuSyxAKt=x8_O*A5|K81i-0-M7l<$Mi) zU$#VB{>D~usI#S^)q*vHZ_L5Z=j|MJHhrb|*~t#H2_q@X{wuv@Q^~T=#Xf&}xolGE z^1_eRerMOu7CVzSp!W-S4z0RO77N4Jd4jL3eC|mw>y)Bhw z%U~t-inX`K)Jlpmlr?`(Z%r*(gB}GCo)lg`8L{(L7ptS)V6v=#eHri-TtZH~aA~b-Dc2;%IWfam_`eJWi);uzj0>XNbA+2GzIT45i}TluuTG9e#5o$sH&A5l|F^+#kvgHD z&Ls_B6RLlzv;60auT37REMJdAnQqy}M4pAWG4MR-W>wR?Lmq-M_J20SStMDw7z0hm z6=P`t%FzGbV5rEj-dvw+BH|nDDVunHwaekXVt4X$at}Cy&;FOX7qu$7=$e+H<8M)pc zf4;M_DyEYMBdi>;zZyuyRAf$wMEE}s82r-6Z-t7@E0$3tXDSQ@Wh$~JR8RyplAMVw zG^wHeAGA7Dc?-;5z${WHrN@Pif>Nyoz5~`j9_je~z@+d10KpARTvmz&=)@*PL zC3{Z;V}n5`gDq=T0tVrqMAe6y|v+PvlQ0OcR!V}RU@+_Mlh%(%8mMukIh0d}`eMg;TiIFJN4ck~L zaxJ`#fz#Qzvn>3f<7NoTSVJ`~MV^JLF)+4W&2%-E2A~WzoMlUqW1+Kb_ajIDSELdf z+9A)lcu?Q)!f=)?MP7x@vXL%r49T-Bg8D{CZYm6C*;3?E=q%f58125~SyqBmMjFnt zrO2PqSr%wH>MSebDKiZuq7<1EA`$+l&4*`Ml0>8jp$xVZ6p=Ndf+FbJu27N}r-pdu z#)^!IGS)kYYK`@Vpo}$C&q`!TsCo{<(w$|cdmXZ>$^-ijC=6@15;+mIW=9`G+5XeO za(^aEOQsTe5tYW?K~$T@-VlBi#p-G}%T^*cLT6du=fR0=Ls`kQY;G9JXiG|qd?}aG zfze~mvZBfDAe6y|vuq`DD0G$ux<+h@Jj><>q6{~jWh;?ap|fmK-%)2-VkF9R1NEpx zu7$TTa5~$pYNe%hn=)LT6c^ z<*2i)h^Ndnkce7jPKZR5yYuVo#l_82?Yru;YEHM&gx?vmJ7r7SR z#=ybrcFwZpt~|8w1j|`=U*uW18Uu|63r^zs>{%AteKU*YEW0mqEOeGF;aQeEred%k zG08)gv+TaetI%0C(y@&pc}zu6$~eQl?7qmQ&{?+AFxq{|v#bQCj5M5O_eK7M&ayzu zQD<2ZPnl^U5h8O!B%->Fvut`0%3wo5IS^SBDky@k?FuD%acYQXZmh_dC}X{Ys6-I8 zPVk2IonWb+2O>*C)pHP*v1eI%0N)_7xER*#fyjxdH5(j5NrFExHW-95*s^9tUPPtw z$b%^RqgufyEjKcpWe-GdgwC?Q62Y-<+*vj^3>9HZN{f6cm(qdJ?Y^u?>ER%h!G?R; z1Cc|avn&Wt#HPrzYE?44z$ElQi- z-q5}i3d33UNaRN7EbC`sa3ULbmdy?08z*j5hLk=M`BE;W1EZ7dSCQ(&!F@+q?q!ce z4u#ILAUqM9BG0nNLkLZMa?OU=Rhy)59W`UaP zt(MVqhkQenG)eLIEq03B3DJoCc)S!`OjA~T}aZEzUXuG?S`%3{;H73mR` z$=Jh~2|>kUC!`+r(8S>#CQOzW!}oWRDNX>-F+R-3X~Bu=@k4yXGK;I2cazf zJxiQLHib^LAV3kXViaaydz_uE_R{Tv%PK#^{oox5v$(Bhz_O(!OhNAhRp_)DK zi3|%JZKL_yAdp8}1Z7-P{0|oEL{^2)ww;R69!#EXB{*fJWZLjyGlupoEqZUTPt!W%G&^8l{5o<}MWL!=s;E2wWvOM|j(a2dK+)h>N}dAJgHRS**6p}gk_*b3*6qi6ww;PB z37u_yYl8#YxU+3;7%Il5tQMJ5t{Meaw^3GP^>7f%V$0cfDzYhbwgmx-colhLn;(d> z+;X;^iVO>#ZIcR*I@=N>QMQ}*GLd!Rstitho0Uy-5P1m7T2oaP85gd~K-FQROa7V`gDDFwXIuHjKizn&abQK-e3Dk=})$4`1p%=Em) zxHps!6xEz<(}PeJTliopG9)UKy@RQCeR@Ms)>`guOR*<~&bEGNf&<#Pvu$n|%4$nB zl3!Myt44v~7m8rF}k>1PME@YfNWb`Q7B`Y}=_A?ZJ$iJLExEn8S3om0wDZ&bB~<^?oZIf5@NMisSn3 zEB=Au33K_K`6SJ%6Xi^rEIi>uwU`A3VZzBCbavVK#p1=|$7l29a`Lb^ zoa~=nK7RO((Nh(F&tkL4pms+n(k@gfzO&BA!(x9@jWx0z+xMo;UXed>_IjsNwO!sA zzFkt56o1DOUy(Ka92MUiT?Y}6G~M||7wY|@noRjnyJ>9eS7b@t`V9`b>ZKWsLdDmz ze&v^$bNTF@QPuhEjX~M_yM}x|5PzLl=JGi>t!?P=FLw$1wc_FAkV;rI4rRBcHp#Cx z&*pZZZ@ZQ?xjh_(ve|OtlizZVPJBT<-e?;UE%Go&j6~V~k1csFcCYryJnwB^)33L? zRR4wI>yxjJN%eRr%6ikTCO%;$>}ucyw^`x*t|pH`*=woN2Oymj|JeYhXk<189+4$dCY!q477YQ5NNIvW=d{BI4^0Tak zG*>g72oFT2g-(Rcev0>F^5iGtsW@8hhsM3a9O=2?(Yw6j~3T+_9AmHKN4lT z*aI@0qOhEOB9^ZaCfete$z!ALJE zK&dDgels(DAilSfvu>rgZuQK6-8Fq6zJ+P?1Y940>&y2S7suBhJzanZN<~KAd&z< zSt<5pcq>DI!<6$Pn^M`%MN6F=v4d8H;kIikzabpmb`3JsUEaE=lNVXpaj2M^>=yZ#!tND& zftRtH%bdbg-$ZUk)i?j@*~MI2-_Tf8+%5LY?+{noudnYz`^orQ_B#2^;?!Q(U&OoU zYxnu#NL1`iwu{_OWxKw<$F`Gk|1)>gUoDx+ZyZ0$@cPY#gWs2TvHw=b>}SO{CO^+G z`^0Ee{7pU(yI^V`$Stwk^i7x_ru|=jBYAHB3nJ$;e8v7Rjz#5wY5y0wp3?p>HWEM8 z90CO=t%`rnvJZ$owXbsfK;M|SdEmwE1Wep^-s!fBUo2*mZ;aS3;G{M`i2Tps2YENR zUDLyyEB<*)K8SqqYh&{vXv6(l7wcat-kN+<73*L?Cc{jgkl%f7@x*H9mOT;1UfOR= zeGvIQ>6j;hXL!~R(YW1#=7`yms4TE>MEULMD%(eJ#Ek7^+%0Ef`HkxMObj&kc>&!E z$c#h9++?@N3l(;c=mjM0CgW{*269@8JwNmeWcTwTbP|xdyyO$#$+7+TTYiZ$mEW;G z(%u#Oql<)HWPALEr=G-MpZSbj;%yj~U!hFpcdU=HRc3JHZ$3ry z!gyPr)St?4SVvFlcN#$w&AuEZGm5`PMNVS)<;_%nzxrsKSGQtw7*Ek=!`lX@@*CFC z+Xg{|0s{R@9V7p0aWpxQ8Tn=#=+h+_hKjVwYOzZ<-s;uPtyoRQ+R~Sh->Z)L68r?|kFGn`E!AYQUqj8|XM?FPCBJE1X+PbUC+sBSZ#j3!?_Wpf4nb$H zn6lT4#7KUO#eL3XyV#Rc*-oFivbKL5aToh>li=jvreol}yhY$nSYq*iE0cGj@~lw$yj|jqg-_8{QS4 zvsVfO+tFB5+)eh2tWIgad{Wc$o1tIAyfE!`@|)qQy)M2nK3DJO!;z@iTk3l)_SaOl z>-YKC_AtKUerLIJmfskU?wtL#8gygAemWDe9r6O2o3b$ZK;(C7AFOWGJTebV`=7}3 za{FJ92R_4B>~rE+Oh%c`ck+AXmG-O7cY^(7{4MwHwRq2p?wi4)?AtN^y*qiYOpiz9 zfyEES>Ge01e5hhQF%+TTq`re)O1ep^S zo+Q8L9`ht{e;D!v^4>l1ViMO&q_V(d`$&$Mv3&$b%#K9aZs`TcZ^XxEVxY0l3&?wS z7!G{oNLPOaN;f#ejLUAg~?`-osG7czIRW0j*PSA23me2KDvPh+6IiAyn)UQLq*zT zwaCE6TTS1)r>!PqZRyJ#h|j>GzQi>3LT1O6rCpos*HCl#*#h@>HC<0?Zc6t zdEDjq?W=5;Z{oCTA@)g_52pKP`3?N&%n76AuXbs#C(q#7@%>z|*)8%oh28X7J7ZrM zS8`7QqF&e>NcxFZbv>Ex9q^-q4wztVo0?J3T;f?vYCF#XP3e$PL; zZw6V5yLaThJ3Sti2bOzx`ThR}KghXoyZ(vt!E)Oye;WYZHYej3bot18cYZ)BBTSwU znW4oK^u0Uli7@seN4D@J`C9{+CxQFJkSEaR-t0(J7MN@w$q_RekLUqqN1|-E+`G$P zB*154ps~*j=rez2oM&&h*!NS|J)#$ow3}S#mV4+UvFC^Gp{MTV&2aCY9D|Cp!QN8- zhQdgDx!ABjicdIYWx0zk>dyXZ&_exmJl*czTODCI9RjIvc`a8@6LIb->~ zdntdX0o_4&QNhfeym!xxLdDl`?_SDZY#41beea&MIgF>s!wfgjrTpy%bOQ~v4H!9j z@17fminPgUk%5i3n!a~WTTRB=(3dD5i1)6jFHwFB_wLjjel}S8Qso1YgDLH&`|^aH zWc&^1j`D%X#n72c(Alrq3f$!%+Xj>+)eh2tWIga9DDNKh44!he^ZSU`5m>_ z4R6b}d)ja$D)uJZMQ*3EU0>g0+r#*Z+{SR{T<(dy4&6EXIT_pqhW&K%-kltcioeAN z@;4^bKA`X2DIbt|VA=oV?@Z+OzaS6%ny=XB#IdLxFxfA1J*EAs^PONn8Gpllb18ps z0^K)*ti|0s@?M!9kIDm+AH-W$gCFSMyOVw(^TBZ2T*}{_K)20Fy@M_vdGF2-NM(e{ z6CyLTc!CQd`!aggT`SXhcsi0Jg8RdeCor*NpHD1zbJN*uAIT9j%8cNM*^wA8ntB0K zkt5BGN?(%md&|P$-tG|+&$gn>Ym#4AWD)PBewsKK1+qyE{yUX8)KzGoc z!!dIw@7*(_Fd1UAS!8FUZKm(tlQwgiQdn-F-r8g~qQv&s-1)cqx zE&T$P?AH@B_&H%Zb5&wbPG!4%BXe7}hxuT$y_~%E_VT;)tJC?Tv&GvNr`M{FbkD1x= z{ai5h7%Gv+DeR`7anIOC#@kfi#GV>e-_TvTw!WjW{P>FbZn9rwbxQl?UH_;oOBMkfL za_5pqGPWgb$1Izv=e`g}M{{?yA*L=l3CyqtsfXRN5>nZJ5o$mzu z$@p9Do8|9Kp!;T!wYYmn-n-M|QF&nUgUJ63exUE&Nk5SJV7YCsMLvjbo0ED6T|V;O zoga|O2$Ls7W@zyQ{d;%T6J$=9c+yCY2<}ZooVc%MMt{&E1CSQcVYx8wkBIeJ~zr%`aScatz=v+zx|fK z4}tEWJ54cjC-2=eqfqfR*(|cN(Kgff?n#@;I9qO@RY^br^ce9Zn9tGU`qSxz6@bM8Gp;UL;j`&I(G;< z`!$<9a}gs^u{YT+@-mg}^qDJb`^OP?@prJ4?Zex0ZQO?=J^Q$eoK0oBKJKyYWZW%h zZTY(r=*&6D)EkZc)lQzZv*S=PH`y)nIECHxSvzAl8E;d46MJe@eM5KUb+Ac_elq@+8Yh2Og4*li8{^x3uZfT%-UefO=HO7|b}HNT`+V$0GVYc; zXZiaQ=+4;>L~utK_S4CGcXBjp?VEfc@;kK;=zDj{2V@?Y_CJy5<@UcI54=%sdv!jBk$e$0jZ3zctZXHg~b!}y*uj(GAAq?QT`SM=1JiGFyslw-(J+& zpSbV%*{^7kiHwvXV5884D?x15RPZ%*JdG0@oO1#~YUGmaly@tIVU-6Ahk*gc{b zkhGhOx8)vM{+Z%3fM#6ENH z&P0lB-Y^qd-ZNjx-;q$-PxoaA)5!Q6&RmuJO$l`F5OnryHWhvHk~Tk`<;+z*5P6x( zcKXbfHS6PuyZE~Y%J%rinEoB)!1m!tOa|HFE^;=N?eZ^RwP_yvB+LiP@4hSfyAtTk z38Uq&cJi#99fyj!;jCTB-;_|;O`o+hc875ldkmKPE`Lvgs&B));&Vw&bxB8a`1xR} z@A5Y#l=kcE`_N7@{+9YKe^-Lq>*5<@{XRb&$&ay=RHl7i{o!lU6p{#XDG)?ISp1#`ZA2B0sjAiREuj;4?AM*stYuFCa4x6?2o_A}>_f zJ)#$ow402#;U2n@zbAq2q4#g$-aR=66=##ZBKI3q9a}X zl}z5dCq|%RYq3@SR>UY<<#KKHK^Sk#JJ&1u`w-|3x{C^C?$R@HKv}yw&u*d)go})|S46{Ot(T zmpEYV-O0WTHI~1Q#auVpFLE%Y{d8ZRu%C>-r7tgkQv#hk1fBhwO`f@kk*L_4Y!`W% z%69t9m9_ojh`abZSjzVJW>~(B)8@l)20tH6XKnf05-QudXt6KiA0$WYh?VKyUH+~F zI&(s0{M9b8IPwgh9Y2V<<*Y4#Q$k@k8S&e3!pd~-T#4V4pz7Q3u9*J5`yjSxEI+=_ zq`p@ot5e#qukSi<`sl6_~F}~g3y9+7eZ7??5MQ*3ET|T30Yh3KzFdqub zopUALx1u{|KM=w7U)WEVc6;(Zo}9(s6Kp;Z`JLJa^u0S}YZ&imv;T=aFSq{%dEhNc z_DWgwfIpTW>t|xW{H+P4{anm$(Mh&@clmo0=)M`mY+Uz_ypN~n@$T z@7+mzx!f|{Hp}0gK)20Fy@M_vd8^D1=to}M87xmA)FLyqc!K`DJL?HDCoFvs`CAm2 zCxQFJkS8#)x%5 z0&0;LD(vRg=g#do$Mk*osusB*x`#fvg?snp7*w1s_wMpHBu3gx-@7O5AmV0;kI}zvx8fY6Za`H+&H;f-qu^OAK78%%htLc09wAEy+ zE%(myw*F5VPR8AG=9IrHfzF&T zTK;M$kCWMPsF<7V7I~b)Zu%MbjNN3sE%og{ymzJQ+wiWKzIPwkj>e+mZn9rwbxQm7 z^?hhR8Gloa6YpQCy)M2P*6;Jfk*L_4Y!|tm%69!eAKOmG-E!}KAokeMowL7IgZtL7 zpHAMdlcQ1bxA;K*#)R4j^u0Uf12PXR`=9)siQN7d*v?5BOtIIbgD1{C@5uXgdORu*EWdX@6#1XQ5A?k|>4z}xVwcNu+k7bU zL3G=k)H~?%k@xQWfK*19JRveeiznpm;g&rV=8Ea=S^gFU=1Jh*G~@|%4=6hll?4`# zcsP(#8C5~?K z-kloDkFHb!EWZb=ig9g)C`V6)FZ3&g_`nbntk#RSiwQKpi66nkc1L3cB z^4>i=4z<=zc8fetVK;s6p0S&Zx1qk*55#*{s=f{His^gzf$eB4D()uxMOLS@Utiyc z_LK29)j09~mD=m#8)N$3U5LKFcTlmn*e-utLS?&tpO3vr#@%q|T+82=KzGi5w1WHA zu%AxeyOX0)@i+NEQ?0@oiCUW~gB_)H8m_R2JSwS1JhQm1bHEcfp9ROE#UyGQf_lKvy( zZMcW7BqmyaKQkdUtX0`0`|NwR<-I zXmNF%ny4_@E8fkHw3my`E!exV+(p;&cO%eUbmVq_C6ia`i4mCmu-Pi|xly*t<=X0l zFlQ`x(6#)12y_SCMFlf=@=8533Kd_&vj+84WM`vortjU8Hiz*P>#^b9y_UZdfo`CI zwgDq2@7Hs~P?5G+Eq^g$yw&u*d)jI;*1x}1JiJ;wy}Y_U{@~*D;`I7-es=nU#mVt% zeswy3bhcO?@9n<5{O;yYZ+jB8{OyQCP$KspFc8H$5+Sadc%f8F==G8USw}N>sPmDJ(ots?^)Kp*n_L6b$=5uZx!zRO1GfD zUVL?O#4qT0AS&)A%SBeFv78IlZMkPeFb9*_@j zYLYYbOqgV*T~Gej1h?zO*T{xFe>fBse@nfWzc`_@-mvG#)`u|`&vyUp9X;QdzdP|T zx910!hhd{#dbavkyNd8u@y*FEP!%CDjh`EZ$qORq(|cjHbF0>pxnSA}Medi~34Ml3kNkBnKOmJ0CQpc* z(BcXDz2mGW$ej31ON|gY;@8O<(dR_a3Y5ykx4Sa&%f&BFW@E~PFq5Axg~b>0cPdQ2 zSnb@l&17C!n3Md?3V}I+i^h;Q@J*2HP=3V4T4AzY>=Wv&AHy3n)|0WfoS5bBQwS$! zpt8>s_+CL~AS&)A%SEoJv3xb64Un{)jQQU;^oQi{PdvbNAfPefN@v2f^Hx^}j*5dx zou~uyOoho}kq?fwcy$}bkTL$dmKcjYz^@Q7_E)r2+`ric^`NLHB^y+KXg_%hleHqp z8)xlm$7~fDTg#26{9OrjquCJ9B%>$qadX4?u@vjH#cKKM663Arg0yoht}%T-wwAv! zfu1EpK?;~U`FpzTI8?MP&koApn^4$IKSP(Xn@by0A7B3V1ngYwG57Xt?}i$TioB(F zBY%lP>jA!ZLwJD90n3R*{w@VPkqAmlg;v^}=_}6se5PBBeUYo_tmguDd)C|T?d9)O zaMlkm+u5`DAgsfosI_m(e38fLtmn_-S?jqpGTqzD->85mQmC@O>e;h+b|5P5mb1A0 zg$j-3{8>C>IT>?HwUfV3!BxBAg))C{Kd>GRM#bLb0kQw4^#B((`tb(BO}0H>{#FII z>&4g2{Jp(kn!kxq@we>x2Vx&iXT4$1j~&S+lIg};{%!@lvGxNd_sZxSYjQX$2Q2sY z2O{Uwdx5{Vr@TPsf@voddv@8KFvtae`D)I7;$T!Bm^>izKdlFJXFtIMWDZzvtL5)m zz}sq&y8X@Rqs8@ncYZQIKR#!F8ofskNacdX6Y^IrES})+?MY9NIbpf0mcMNQ@2ZoE z2i-vSu9_c_$_kS&M7C)11s6s7a-VgWjW^2QybzcZxOfb2^Y~stb|@+nEWAxLnqQ>$uy@I6WWXvtM*z)%);4Su^ zyZ1X;Tr3y6;Yw%1L|=8MW-{H|%U`V+Yq9(+Kl^Y(!pJc8rkiZ}TNUsoJ5tJD(dnk-Vq>yJ?2;vIS?!Fqg^bI;XNimWn}pZa8~E{u zi_<6XT`yfD>&-46pDF(6WV?+AmZaif(0yP2wgc3CwZrO8`u*S0`+Z;jF2iKwO+TeS z-#PtL#UGjcl<#z3jPkfiYq9$?)Oxi;X-)e2U;fK?iZ5he#1QzL)Jo9qxEfjhw2Hr5ahOk+mARuSOoIk%wyJ(e{xx6CUWH zdwS?p4=wf3N)N5|(0x7hKo33CL-pA-)o0UGpG{MJHcj=}G}UL*RG&>#eKt+?*)-K> z(^Q{Lr9PWVeKwW)Y%2BHRO+*-)Mrzv&!$qJO{G4YN_{p}`fRH7*;MJXsnTatrO&2H zpG}oMn<{-aRr+kI^x0JFv#Hi+Q?1XYTAxj|KAUQNHr4uUs`c4a>$9oWXVbnuoA&kD zw6D*meSJ3V>$7QJpH2JvY}(gn)4o2N_Vw9xpwFfQeKsBFv*|#eO$YjHI?!j+fj*lK z^x1Tv&!z)?HXZ7-=}@0dhx%+f)MwM7KAR5p*>tGSrbB%;9qO~`P@hdl`fNJVXVZ~B zn~wC^bfnLwBYid<>9grbpG`;lY&tsH*1qq2@4*8DcF%yF8nC4STN$vm0lRO&9vHBP z2CN~k_6&KoXUMBPLtgC}@@mhJS9^xM+B4+Uo*}RH40*L@$g8G7uHNgK4!Is{$g8GD zuAkSCS522(Kd&LLnm)OHUPE3ropSZOO{-jwHRM&(EZ5I#$g8GZuAkSCS53oQKd&LL znwGhKUPE3rO>_0UP1jtHHRM&(H`mW=$g8GvuAkSCS55C+Kd&LLn(n!J-ll!7#~Sjg zX`t)pHRM&(Lf6l0$g8G_uAkSCS4|sTKd&LLnnt>M-lmhT#~Sjg>80!EHRM&(P1nzB z$g8HGuAkSCS4~GM$g8HcuAkSCS50$W zJ#W)p*JBNN)%4f(^BVH1>9FhPHRM&(W7p4X$g8HyuAaAPv+J>jylNWl`gsj`)wJ66 z^BVH1X}0UnD?`71YUsC54gL11q2E3=^xLO~e*4tWZ=V|a?NdX)eQM~pPYwO{siEIK zZ5~P2?$1+0zkS*~l&+uGuwONgrR(Q4>{rc$>H2vM`<0>JK5d>&SL*qD(Rr3J5eqKXf8T#$h<{5Ui&W3*bw0VkMKd&LL zn&;T{^BVH1d6Hc}uOY9RXW7;B8v5;1L%)66JkGAy*^pPw1MT{G4SCf((ypJ^kXOw^ z?fQ8Qd1dIgPn+l3)jAvc?bGJTcKy7DylS3p*UxLntLEu;{k(>}GCbd28v5;}q2FE_ z`t7Bm-(DK}?WLjLUK;xCrJ>(m8v5;}q2FE_`t7Bm-(DK}?WLjLUK;xCrJ>(m8v5;} zq2FE_`t7Bm-(DK}?WLjLUK;xCrJ>(m8v5;};raH`&~Gmd&$pL`etT(nzP&W`+e<^g zy)^XOOGCfCG(6v48v5;};raH`&~Gmd&$pL`etT)?x0i-~duiymmxkxtOGCfCG(6v4 z8v5;};raH`&~Gmd&$pL`etT)?x0i-~duiymmxkxtOGCfCG(6v48v5;};raH`&~Gmd z{r1w(Z!Zn~_R{ctduiymmxkxtOGCfCG(6v48v5;};raH`&~Gmd{r1w(Z!Zn~_R{ct zduiymmxkxtOGCfCG(6v48v5;}q2FE_`t7Bm-(DJ?Z!Zn~_R{ctduiymmxkxtOGCfC zG(6v48v5;}q2FE_`t7Bm-(DJ?Z!Zn~_R{ctduiymmxkxtOGCfCGW6RkL%+Q;^xG># zzr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q; z^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6Rk zL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7% zGW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLM zD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuA zy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8Z_ z+bcuAy)yLMD?`7%GW6RkL%+Q;^xG>#zr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDg zzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q? z^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$T zL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=i zHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQ zYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF= zy*BjQYeT=iHuT$TL%+Q?^xJDgzr8l}+iOF=y*BjQYeT=iHuT$TL%+Q?^xJDgzr8l} z+iOF=y*BjQYeT<%-_UR0H}u>0A54GxmEDuY>h$sA;alJSlMf&4+$*-;o!og)>=eb_ z`S(wklRIxbw{ySPdS~+d8$Z0e^Qq$1)Wc|Cvpeyna%pTG9ITSb1{w=OT1*OM38|A~kHogZx#FRbQgA1o%n{iChfZ~bWN z{=LcC+S$ub&MqI#&yJVZi>I^CK3ZH|?=H?CEly5OFPWo)4Aj|NY7Q{CqxpupZ#z z@#V?lYWM60ef(f|eto^TxS4Brxj6gY?r(gsxcX@K{l(SA;%xR69`WdOzU1WJUv%UD zEHnO%d^ow3Uor~qd&5_e8z_hbe+$??g+VIw6QR|@N5T0IK zFRsoPC#Uo4MQ9EV36O71c@ruXn7m1)FXScL5qDB@kxRbJ$;#QkH#++TM5{V-^*}z8 z{ZPIg9p%|TI-C6hqMZR{zb}u%n$s%4w6mb>2eU+>M%kZ4$=ScR3$l5Wr6~KMd}|8I z{=MD6EPIjsUrI}6HwFiR9_l>JGRoc+_0*$?GgQ&9F#>Fkdr+YwRrcXE+SQ1$~^Is41e z*)Jek)lv2jH`PLMa{Ux3Kkz_j}%KlC+atX?QAS-8oH9GqRM5{W={(*cZ`=NY0I?A(wbT<11 zL^}h@eqTPwvK9UW!=KsuZK0-~J(Wxp@aX1|DOXF=Hy5K*X6_9szt_8*MQ zekk9Xg0laB&i+WU9T8=JCl|Q{Wj~OWv;S~(_6vwsb(H-B`Aqgh`F3=a{R8Q2_6vx1 z29*82Je&O@rkw?4KR`sGM%kZ4$=QE6GW(%?YYNK#Lpu8-$#z7P{heIo5|sTwR?hyT z(b+E`TGdhZ59Bl159Qm@QT7j{v)L~o+8I#x`|@n|iGn^g&Jjl5+!H<(a7wF z@~tT-`;X}Ck0jd>QTBIokxNka1KG7B_Q2KHUiz}PxkbLY%nq)HB}l702k;ia&rvhn z0NU|68-gXkskDZ)bCfxi&?3kx7J%A$P(BCF-9Vz4@005FuyyzaKgbtfLTCV>HRnrM zufXn2pX`f-Xgem#@5V*gG4P@e2q#v9e|r%#dCU1M&Q75Fo7$ za5eY<6kxa-L{K{qTnzxyr*-=X1Y8YCSYkE6Yu!!G!qor)tvPTtbXT##(;{#+M567O za5Z#-kz?R$0K#i`ywR@6r_jB;IcoxM1hGB9aX0+=6ByegBR)eI56j(Pg!)LAqV zu$m$I5zXLRF-m+jLwv0p9p~U^$ZCd~3t*NY-^OF`YW6ANfZ4mKpw&zOab62RKT0UC zvoLbDxJaGD0z8;4hT&BNVZLQ(6=gtV+=eCU9Fm#AY(pUF$+$?7o{V9*w)tuX#Uk7R zQEj83d9Mbu`Z3I{1_6j`8&DSSE2y@SU|idh;9##owT(qfxdaMF7+JgrWMI;&1Ta6M zcvU7rxVClYNVc{m0JyeA0N)B!+eC3(+d6TagQ&J8A-J{yNRV%+wxuz6HT#eKj;d`0 z5a+c3^rM9GItwFbi;EQAkq5KIFuaN&%(o1!q6~ zU(KLcggYRrZ4@-`)nHaXhN!lQbM`=7+W-*nE2y@SU|idh;9##owT(qfxdaMF7+Jgr zWMI;&1h76Z=dcAM{!n2}_;GFPelft-wgdpzwg})`fohv5j%!;dj&l&zwj>1CHUJ6o z4b`?Z2CruSq5e^|jR4}j7Jz<~P+n(Yc+SdIZh^=i20IqEjz_$X`Hc=ecwoV-9 zAgXOi2(E1a6671IZD|Z%&HhhzM%6X~i1S(i`cXo8orRIJ#YGD5$b;Ep7+ysX=39np zTLwhNZCIjkZ5zxs1d^VNixla}7=~+`uVzqX!W|IRHVT^eYA~xGLsZ+uIeQ?kZ2*Y( z6;#_uFs^M$aIn{)+Qy=#Tmsc5j4a**GB9aX0$3jyT-(GS8K!Dm_iI14wj}_#wnYHn z3RK%faa`Lvah!vwwk09BwgE_xZ>YAVF?comKgb(Z+Xx`eYXRs-3FUPbM$Q%&DZC>O zW{Y8X6+xJ98LDj=5E-{&iNdvQFxwDFdNM9jq$gt-u5G@WL6r%2Kvdf(Xx^*AtbPnp zZ4>A0fw;B-Al_F{Z6m?Bwk5&AUV~~Ii3!pP!1AOn+DC4lvT z!L?0%K|rdub-&bSYg+<@}#iv1lomK(z@Y zi}!#GOj?xy)&~aHHu3czsoK`Phl8zc2>`Bb5x}cuUd{fiGDg)l0*Lck0Qylvd7Xukv&BUU@5qDMVi;aU5awHkYFh?G#%)-laBUmR zHUyHMjEfZM$ry%fo3Cb2Wx^d0)iw&6_i8Y!A462z#5sE)u5AE__Z3vzNHDH#NpP^& zpxVZwrCb8lCX6iJ12QmaRRUNa7+l-Lm;0n@Tlbb7wzeezxVA+A-wIUQL~&f(I&qwX zsJ0~`xV8aEkZ-8Ar7?Il`!B*7Roe(4&T9eaM+xP17Dmn%7b(0W4`z#DcojjIZyBm> z84ww_VTr=EZ7|yqNP03ZQluwi7_M!;nn9HbcR*CzC}`fR!K{7^QEe0F?18wp0U+L2 zP;Dc@xV9z1!Cr%E8;h252~?XfvUm^3z@$|PV0~b4Z4+PJm8xysJG9u^mH^<|76E)K zP;C>%ac%3waSo!|mW1Hi1|UJcq1u+ltY(Mc?F7L~vBcu0MXqM>9YTZGcdWmH-L5|y}|j`X;5tZhPCRPjQkME`ZKOqtUqJiYW6Ufqe00Dw?jRa z@NF=!AIt385Rkay1i7<8;*ADX92UKreVz1kKXj~2}GERLNou3^GnJ(w@X39Aa?eCyz<%AnZz4J(*X`v&t3 zk*q)C8pirF#tF60*EF$*z}iP)3Eu|u`muzyPn@+!3bhX)@m>RKABh%fUlJYcKd|<( za51YuRSBzvuq$M6;wlBWK0fD|*NU&?bmu)-$-38Wpab7ngi!mUIebfC?GxpN+SkeB ze1x?xi4kfazy!GmYhN0)+I*?_>F@=l>#&{Y1Xi<8w`?&O!FO@;pI5Wj*87KhovE7@ z{`G1$*+}~n4F7Jm^K$XR$<2qIo9}4f%YHGt%zpU#(T#iN7su!G<@>wex%uk0@U?1h z-sg9=2Vd&GU%dG6?&Qw>t(~88AGLmU_KEw?PujnHc&GV-_g~s7zBx{x<@MF+N%M*C z*(%;LD|fs%O1-U*TfOPCQtux=UwqZz zy~!uOvySw4ZUVin$=YpuYHi;}0^G(R)A&pA&DwkW;$-ptX3ampbFbKXxBZXTiyvFQ zcXfL4{_*_#r^|K)Pwu?&+|G{`#iPZO)AgEe|K=x(R~Bbyr%#u@So4<`iw`60tHu4r z#YsO_Vxbp*b*uQraY6s~)#b_V)#BuX$L&|=PcKfIa4$zC<%lKy=k65K@%l~f-e|dP zN$2fbaZ8*8uMhUzyTzS{&+UA&cG^VU=gsGKisJ6Orzc^u6#vpzQIB&* zTMWb=@7WfFzqVEU(m00&)ykhIc$o z?l-i4c-?Om*5$lW@1NZL#wjR8Sa1wqP^|mS^fvaJscpa6ILWqAI8X20&M>#T-#o|c zH(NXMe&cU5-R4vL$6I5I&Ew09$Mfsmwja4{md)L7&mX_Ptze8XQELeLI}PpLN}iHC z7T>V78~T33(~Cd1RUD10m>VP75ng!f@?v=%Ycp{tDzQ$ABi{wn}kRHu6X)LMKeoS+ zjbWQM+p1UtOK;Am_jW!}y!h@~k((pyrv~KVrQ+Vh7bef&Z`wlt+US+LG3|z&-A2Re zvu=gFKKse8cU1ho`@gN?A8Zu|SuX~Sv`)iG7l$urNBGv|lheoZvv)49$EI#K-P$2S zREjZC`FFR9k!_ei_1@yDX&$!!v#s@)sQMQNDfC+L6AyoU^74Hbr{CYYX+IW!ee{y~ zPV*Sf;s=Y#z0H4&?LV~^$BMsJd}6sCcYQE_p?H3KYs=NEmv%lku;<5%Pd5$s(SBc5#A8AJ6ud&EnYmmIK4iePVT;Q`D;&$ zpDI4t{&{|Oc6>d5bk;O5*Z<%I%gvt;i%&ecy8NK&%q^ZSk2in2eeX}NM|1yCZvWW& zZ~F0yKT%A6b#e7%vD^Ot-}&JDe17%OPb}xF#qRe` z&lVT+^F{NW+NU2}ZxX0Mz0fk?$<4*dy~%U47tSwF7H47F2uV;}6t4tBPhS7_+5GY1 zz00$c#nqoWy?*Z(7vGzIaCW_oG`N{!LlK%TW{{@Re2b!J=UXoUX|hl!YUDtd;?Hgs zZ?yB=+)7+L`Qw+T7i*{8923@i>dx1k-^$ru>$fi-UoXy|9v@E2?^S!#>Ou9rd9|pI z4v)&m)9Ub~o>%4mL3yy-YzfztPQ?esCnGH;MX%Ae=ZbAn-J9<4a&fZ_U0+_c`$3od zl9T`G&9oOEoF6ywxjJ!m1g!b!21^?9{bGA<$<1(!lV5h3*bXmjWnMx&FXcAS=((Gs z`IBh2$tS-PPQM+ycRQiwM!0uMTpc9a|Kjg<+0?AUrZQhIE|!;9$2UvSc{Uy^#g}(} z=SN#E_kR0FTeIK#(N>dxYc4(e%=+=h^GA!5lO}VQyWze4Eq_4rv+`{1?ftEPVe)hF zgikM*r|nhJqBHs9`}ijOVCXabuIRwwA6{Oa9DBA8KGhisxMvVJIs6cf{}lV?38^T{ z(C<;QD5shMjDJv)oOzw6=((yq%xP>v%KG|c@39k^b@`A2O41scUp~xtY+wo@?e&A;YAOSO2SU3@tE#Y=h#hNX+^)2 z?DXm7oGPQlSF<rA8yDmWYX$9+665$j@yDkC5%uWPuk`w)~K7(<3s-d_(__OAjoha?JL-ip8 z!u0`y67?Y=?{q`;Aqyk5>nRuHC4*^t`BB=%u>1-c(5>tW86cd7xfLP+Ne}jvi}PSF z@ARYM8^cHs`aF&EppS4GQXU)vrIib)jmYdWKfu9nX~=-2$O65RD_4|ve#4Q?z%a7| zzj$fOGNNgMtkY^V1T#A@i_GjKVW$&fc4-V|b{)nmt%#}tIj71fam?(Z(G$!LN;*vt z6-)pGrsSW_@#z3^oh;00X+aCW4NT6xPBQqD#5~R^ZGz*sq%=Rso?kZ|{+p(B_}lk( zbAvw|O+uWmTBPuUJ8{lUn!?}JWg*ujB&(*8IUmY4KC}cLN8_)%wV#$d8{9ME=KyX2aNM$ zjAQ-iQ!&zyKEx@>`EiI8*D_!}?cOM?Najxg@aGMvo%G_4o{R_LPC-~b#7WjDp?W}6s(K^p{n!&U(vLmBsmKL7#<70%$rU^!1fB9E4`M1En)cXCCG>wxID%44B)?L3E*#L^ws2Kz!RP&bUcWbY+{tI6c)+ zyutXh25b&K$0tfV?a&630pU9%2->%xmTq$KQg8jr!Z zz1=)sX+_t!K+dT$N*v$zMx$?%2%RpVq|;<105iK%#q6>urz2u^Bp5ThBv@v4d6?4} zF*_EGcU^$?(hqH(NrY2W3B=4!oN<$!h@NFIPER!y*9U*rJhKy}opz``WI(v`Lr|hV zB;=iLs6J$2q;@^!g1lrfO)o!6yBL;VAp^Pvd5NIk+TTvY+zJ^?wFZ(N>?s%L!Cv0! zN5waWkskEv7Uw}9;WVT?I0Q;77f>6K*_9e*mjOwU1$re{t|;&Ph9jGSVP;p3V0MD6 z(+V-WBm^@%0FjxUB;%Am>yWC61Y0H2Nlq5K)7YPLq)U%WjV9e~2V42zFVNPSj>{vA3bphH-KeTx!5l&Gh5HmY*#!Ye}dX~XBJ=IWL zAN*PK%ubYc+976_0pa=pL5cd1kaxPF`jCZ@+VzwR@{++cz5FQcVpx8K4Cof*C4zoy ze>)9xD`YU$8c2Gur(B!|dwHiH72g;}deEm^oCke`(~$Du5Gbu&Ky5^3S814C1|&rm z=#^Z#qP+7Pj%)^onO!x4*$J{vE5z)Q5X|fVL}qr9u+s@KyEFzfyAI=(Rz%f+oKt0# zIA(Uy=$j-$L=8$hO-2GRv#V9iE{k$HB4$T|F|$j8WoDO$IgJssW6^lm1!yn*(B_#$ zI7O8}%Q@}sniVfhs@pj(ia2>Px4?KI4-kik@IAnC!La&aE)<(+<1d}A2tL7#4M9`q4T zL&}3gptN!UwGo+JtzmW`PDWH3!HKT5k8mR}(Qx&?WOpx@fxPQ%;^8BDbXk{;|S z7w5rV-swlhH-?cO^ywDoK_B5Xq&zqTN-Gyo8C&)Uj5VK1{FtY;?nb}FgPAA0d(iqI_I*eCZ5mf_nPL)yOnAt_6Z;}WRH7Myc z841A5?m)%tvM8q`Vs<1LGrJ^MW_Ed)(-<*37L9jZfcDZ4ZJtSlQ&b7W%ubwflbnd2 zWiU=pH5Atef7U#+6Q!MYh}mU8xIREoqCO<#oo=W;WMQOsJ>`PDWH3!HKT5k8mR}(Q zx&?WOpx@fxPQ%;^8BDbXk{;|S7w5rV-swlhH-?cO^ywDoK_B5Xq&zqTN-Gyo8`PDWH3!HKT5k8mR}(Qx&?WOpx@fxPQ%;^8BDbXk{;|S7w5rV-swlhH-?cO z^ywDoK_B5Xq&zqTN-Gyo8`PDWH3!HKT5k8mR}(Qx&?WOpx@fxPQ%;^ z8BDbXk{;|S7w5rV-swlhH-?cO^ywDoK_B5Xq&zqTN-Gyo8Hw19-KK&JoF(*LjM1Sb3P!*n(bdqn9y*xqFhjjSiVfgi}-rTy3L6 zhdAS!*lwdXRtDqrR6|$WoAnfZBDsH2b8gws$7Bt6)-rZ^Aw@=ia>gE5Ts zpx>6t?lG+geT37H^576Ctz6Iox;gX;D_4+qy5W_Z1WAzvpybLG<(+OgvKiRQk)3G2 z!(|z@v$`)v^~~(%i?B|s(U8^lm)G;XIosv-~_JKb$}xJ2YHhq z>=YZ15r{mPXI-2jCzqt1S_;r=`-`Bj)>MI@uC*zKZeft1(`KCfYWqusBAbLl*M6}o zVObc95SXBUnjgUgvoPl~$OK8Wzyy=%1QX2ToYIg9vT$KT2JjpcOhTNla6=}L0uvOG zn?!|dE(3C^>ajxg@aOH=lUzMSai<`x9vPHSJs>JoJrV$?9;_Z&9IIeY&@h7;Ow6oF9isaV-ND zlv>MzxKmG9%Sn_NVt`7oWf9=i6V`GDCosXTW5h5)LD(qu55l%zOgF~RSaskb0sUBX0g0$030g@sMW=gJHQQql> zBb$NY+urU&?-9vESGR(!(`qyX^G-0!y6_?nDG58B#$)hpZ#R!uTG2HUkaMbx634f_ z(de5bLU%?`(rGdhfSH~DA#9%6Wl>H?#Oz2gW_C%i%GK$nVmS}COHv3%V3|!%IS!h9SO$FE(w;IT^{B%M$C>y<6Re^z4SwyXA$ak7^z)PxgakY zOw-Gc(k_POSIB^FL0%&0xAwQwFt}rd2YbrJd9asv`cd(XVWbCry2W|WM>q{B z4-SFS$_3O$WOlWN*=0adWPx7El`G0Szv0McV3^r;A9^R4ognM9Ld-4+!ORXoWM(G` zJDm`-OJgvz>o8tvMN|#QIaNl9V`dkPzDXiP)S#r(WF!DHJO4x2JhRK9oQ{aukzma1 zl3d^UO|^cG@9k zmjU7W06~fRkdSw}q56=8k=pf?3-Xe|G`;*N?P6Gdg$(Ex?s%L!Cv0!N5waWkskEv7Uw}9;WVT?I0Q;7 z7f>6K*&S$@T?QmY7U-2+xuU%D8;)!ShM8UWp?8AW39?Qr#O#s~%4=ye3C7GW36_~% z9_BPg%#KClT^FFe^h29x65$k80x`1_XWS$wqGuV5(^Cz_^}(Mt&+J5LryXK;84#`y z5R|A733;a*st;Khsa;RGATJqA)60+2E{5e-$bfD^ULxqX_P5h8w?YO}t%0Nmd&{10LC%r1*^IwEFAf-$p8f@NlxhdGTAvt!YC*9B-V{m|x_L^ws2K+NpK88^v^ z=vfBi^i)G}eeh?^Gdoe*X@{6y288Pa1SRT2Lf+|y>O&SrYS&XP$V&#(^zx&$i(&Z{ zGN4to^o*>?B$()RD5F?=|P`vaUS#$PD9FrL!h*B0ksjC z-I0dbWk6D7fnLd#E6O{+;mBrSnAvq7dMB8jAnUY3%q|JR%nm?gW+w?doe;B2V=%Mp zFkWdzR1L^ERYr+pW*3dVNg_nlprq4eBw)4udi$c~rqDF1E0o!@8(JV^J@Ez zA$nuc`b)>beBsuc5okW$`{eBM(fsUqdA)eL`|FGON&EkOcYbv`e{{C^wWr%Jui5Bv zyFpC6KRD%wf4ACxbtCuT;`GUT*URnK!aqAG{Oi?EztXsFb^3Vm@U3tE$%hYi{@&f< z#gm)Q880V)`)=_!?@oVsck2)T?+-n2+a7pnJz(*#Zxuh)e&_u0<;C^Y<=OGc;_2d| zSn%sZ8`QPi)>yLyFq<`X1dtXS86Z!SQm)P$XcOE{sbFbKXcXH>!y`3)>FCITW zn=hA>PZyt>{E@TE#}B`OzQbPpTU*7K+kg1g#nqF=cc*W?w|M;i)61r?Eq5Px|9$-Q zYVp0(?>Fn_?ziWU-xpMVW3BSg*jW8e&7!zF|NiN6a_5a!_jj8A*xK117tGpLZS|symqs|89GXZQDibh_@~;me;YyuN9v>S)47d7sr=pC&wqJSB=hZZWX_| zRlL%ITrj<`dZTS`JOdOQ?=vCx#lP~qTg9hZWqy6JyuQ3@g1LLWxL96Z9be4Po6x-g zE&H7xZ51!Hgm3cOKiZo8){nOC-BW;}^cnQo=hOeb z*_=7C)ofj17HGDXjm!q$bngjNE8yni;pNqdKl#9sk${_)lUvQu3#|R`3$1H=oMouv zltnq!4B*uak^bb&>pVpbr##GQY(cMP=*`Js?w(Q3(1|dKaEdB{s~I{Wh%>HP=nVCo zG8m_)8oHWo)>HIV<@t59uioqvyFt5ZhB{Baj)An(P61lY(962vOdIdz61meX;QZPA zhcHsRzK_IB)5|;0QAWkE{0eD}TEjNp%O&~%;WShNNe`~OG&iRy&V#+Y(~t6C3?n_b zhK=!{k8m1N9vlLtl?(bct^e>M6r`PQ3Xl|8FjI2nit1dQaUa zsd{F1^Hp`H)o93S_T@GG%4xJNytuFo)DOtwc+AQXQLL0#T2cS0b5Sqn+%rmiHG4DV zn)T?LBtqS%NRPFo(_|z7Gdus&HG>%3q{R%8vn3@JG zdb(%3dwKy%K|v(Nn-1TmZCQ>mO_{R8v~^grWLh$9g-BbZq%eb`ElQ+GhD1^{Em0&z zN|sFLJF8yhQTghBgOww&3~;7Y^n{IaH`V*>^$xv>Ts+L2f%*sP{{7|d8mZxSHPJIQKcTIT7~RoAvRFZ>N=`3&l2d(r}y9M8>vc zZJ9tm1@L36lKH$QFQyec#l%>$Ewc@3@|YT4p2?RJLZ2px9bKg5h>G(IgVoHfI)OqIn1jicM--$pu-_l0kxO%Sus6G#6-8 zlN#Ta@gDBD4SM3e2m5(VAn82S^_OVqY(@m$TsLkpx;hm zFbLMs2!JE%1`SrSr_0fx=m^#b#lZlSz9JhkYvlu z=QVkZEsL$#DJELUw#+uD$zyCe(2hZZO~gop1euR(!bl`2wqcMU(MCXm+}AZ(BodSx zF-Wj!e82;bMVMKx1WXVDmWR*9&y-*=+UZx*%?1jER)N`}5t-Zh0G z3qeDb4V4nKI7}x&`+ZHHYJUA@N(s6EOeaCNMNOefg18}PxdhB9pdM|6%-1z-5skQ+ zB}Ni0m;Jt`jh@*+Gy1f*smktwJnHC{`L?F2a}yeO5~Gxkm%@*#%qGG*ZY;*_cLlVq+fJg`3qB zQ^-!+nqGB@){gp7i>KJ4CN#1kjqKPFE9Io%S)ff#XLM8AKe(fYksbSWO&#SQ0*z?@ zKsFNmgYbP#8|5E@&6Lzx$VK83G>rYe6Lo&Gu!SJFMdA{MzEa*bg&_+;LzNAc610#@ zCqestO`mFh{bouDx^PP;LAOOsp-O_dA!oS+v|*858hB8I%vA|A;%1hhOSD|}`&w@F z%m$j#$gZi%u0VFow>4E1vJ;!o$c|{jBRlTnnkEX_$!%z4*R*k5D+Sev?`l#yeNH1g zulGucD4<4uSyR%n0gde3s?~aA7i?A2NFh6AV;b3sjd^4jZdOxFAv4Dhc9-oaGYGhDCN+46+L};%1hhOSD|}`&w@F%m$j#$gZi%u0VFow>4E1 zvJ;!o$c|{jBRlTnnkEX_$!%z4*R*k5D+Sev?`l#yeNH1gulGucD4<4uSyR%n0gde3 zs?~aA7i?A2NFh6AV;b3sjd^4jZdOxFAv4Dhc9-oaGYG zhDCOh7-Sb{#LX;0muR`{_qE*UnGH0fkzG@jU4iVFZ)>V3WG6PEksZ;5M|RxDHBA(< zliSe9u4&`ARtl;S-_@ja`kY2~UhkC>Q9zCSvZkbC0~*=6Rjc*LF4(H3kwSLL#x$}M z8}rC6+^nXULU!8L^r}m=cGQnrJjE6@p^*(~WXFzJDJKQb0&QwKqnpzH!5uY>?AWhs z>L_FvXhi!5vXS5)gzszGDE|;_rlih7E)tiZVeI#vsPmhJEd;?W5|=RamGZ7B3|R;o zs%)r~poLsI3EJ;#`c(7lH&aT`g+5jV30U83c( z-`8@ZXExA`Ms`hAb_KFyzOAXEke%3sMs`FK9@%jp*ECVcPHsaZyQYofS}CYTd{>jw z>2n&{dA(OkL;*GO%bJpo4QOQNR;|_}yI`xDMhe*}8`H>6Y|JCOaI>0X3fXB})2lAg z+EG7h@f2Isghn=`ksUi?rJNKz3$&@}jBZN%2Y1vkvSYulsiTlxpb_mK$VP&H5WcTz zqx?g#nUXpSxky}shOyswqRwv?wh#ojNL<3uSIWDlFk~TUsIs9_f);Y=Bxt{{=~KI2bcvSBeqYOtp4mV%8rd~f*%io+ z`L?EtLUv*k8rcy|cx17$JGl*w?3y-?Yo(wX@m)EYPN=GrB45AKX#H$d3KGrj9~(fkw1{AR7t(LHNF=jq(q{W=iTT+rcX7$elw*6UAU!_pxdIRP$fa! zkh5F@+OWuOEe6>I8gVmA&?Q z3PToxhAJB>C1@d+PJ;IPnm*P1`puLQbm5jxf^LhNLX`w@L(XyuXu~4A^%!IqXvEDd zL6>N`?Dw_Y=$Q>Pqmf-xm0f}Cm~U&UC}bx#p^+WYghzJV$2Cn9vXk4;$gXMQxK;|P z5#QCMbo!h|c3$t55>Y^n{IaH`V}p~;M~jD+?;ITLO%I`B{Ql8&u`sN6(Z8?95&!vQ za{;%mI4Ub0x5jHPy@3x-HZQ!qcW`5SZ|mrIdS~-Oi9ELonupF0HKF)-C!3e6@4hnK zefgE+qs>R!Ki5skU!ROGnEQ6}QpH8F@pqtt-RI5Ojk1Qc`P0U{$E!+x^vX6AIo{si z**ec4V+^T*EHn$@Q^-}=9|B=D95-jcxouO!e~QO$2Gn9J=& zb9nxx>Cy4Q;q;5i$Q1wd%)$QA@#r1x&laIF_0hrM*8cYG>Cq^(@Ls(H-tGjJ;EiWq z{KCf4_&Xj#EBS8?%|9A$%qA+h$@Rkm|r2}v2z*{h{g zVt;)!{H4?87fvUC=k(yM|Gg!Fz9n$J?9lx9z+5Oc5>Fo;?Y_J}JX8KTdUsiY?Dmc6 z&d%=s%SVgtDC)@Dq>LXl7w#TSxAqQhz6zfY=N=o3&ziw?h!PFP%jV41&AsiTqv40l z2Zry1Z#JHiyB|MnE<<#DcW3)}x>W>?iXQp6&>LsW|1mI^98R{D8(uyj4^{=dMh~<% zR2$18Q)>-VPneGm#|oyZar)pb>Bid$;w>3Vvb-mZpU{k(pYrA{ZFnnVp;Jl+i?})B zDisM&ni+e{d~o>wz>F0m&?JkpGe76?lQsN&kk`XvG)@sSk*)QHtwr zVs$NoRjjFNUrkmSeKvpBo7MG(Rpj=L4_=+_Z{>$KlG3h^HLi{9i&4$@Al1BNE^M41 z&YhbpG*i+K^BaRIAX);Gt>vYL!H@cCOA=Kp?t{L2UmDLs*!Yu;;!aFDTqYG)B8V0i z=GWqwF8rR15Wg~>nUNhvJ%b`PKjY7@#Wu+Qf92qC$5ma#Vj`*1SOWmOA?QgOzZYrr z{pQlfh2iOQ-}OU-a<%X9CP$Qu20_py8pNf5cE~OWJ)Iry&?7|) z&-_z=E-mVzibXChZXd&GOpXo~%d>TV_kaU6L zt8XH++`295#Q)>lgT+@=0+*;0;Vv!qFXqC_l zn}CMUTe2Lnc#9%?rcs}DvpSq~P#;UC`F)>fi{7XdSdnMW43cF0H4Uv#nNJKq9z|=s zmP=y960kTgQUZkXbHGnpr6sC+6zGZhXTAEmW&xTY~Bic-wE1%eTr6RIxbzkb+=f;#`$oAcDdKB~YyIp@rB zpUOb*>1OU$&19H%n!6d7`Zv6#KDGERwCicqrzC+>^>o=DnnrAK=usH!iCOCn5`l$# zQ6j)PpUNS;zNErweNZYWKAT#+r&4`;u@5coseFDJtNM0CMis>6-QVk{7Tu|)+>Xv_ zy$tYTryLF33-=b@MR89k{6ySgi;k=>(JhH?>GUT)}E5n(Y{Xcs94LjO5k>K}&vS5uR$vFS%!y{48OneKuQq!6<8x|(8GYD{ z6s#bNRqIJJ)L4D8EDjN&#DoAa-xj2G=bKRMm44{#S$ZdVdV6`*tV;i$Scq1o(bAf2 zb5u~wWwP2guY3Y+h&c1>K7SU7QO{~MtF+~hRb?56J=NItiphpa(5}&d5kkb=S_IaK zTS9^5v(j0)jarrezBjik7HmBxt;(gS8F;<2H2zpKb<1XP__%WF%Kn&H6JXOq_>RVJ z`;1yhMZL)8DRD^{r^YK>;@eClk2f0`u9=l#)=7pHsRpra5mraGu|BWJ_-c0YIgJjA z6=lP(ScpYs!!LKvhF_6UmNFYYI4)xfBGB@4z*j8LqH@54G_YDPhscx(@7m#ek%R?e zGf09kgh>Q-1cI{2q+j8kpaX*y3(csof;cHO3}0EX)GIg8;Zgb#E6M<0u>g(A01wl} znuHTfQ>MPF2fJ6YuwZT$SrA4cW4oTk!?_A^rS!AmS1i<{4w8tevWOEw_?7j3X3UzE z@(f8N6Xk-hTEIsgo`{`sX0R&9*OZ4RnxP{^j3r=^V59`F@~?7;P|`=yBlIUUoDIKf zK_RtKXD5Pz|En@CRK_djv*n6c#B)i-NF) zEO_og1&8gr*Qc-;KY>O{DO-wM63eKPYKyOf|K z#Z?OisUt;}E>`zfv}dBBuJ1YL76?XgPN@27PnEk}M<<&a*)Mu;AmW-z?2A`z@vxECb?tn+I! z*y~FKoJE8kzW+MQn#Fr6)wfC7XmMYYv*eU{1W2*%h|Ii8OF!D{uUT}bnsPfjx4^F@ z);Y({FB&C+Rbo}a+*^1T#XX_$sWy{SnAlRZoKj2kOEBHc21<=wvxqLVqeZF3SlhE^ zAZ=PaBY|>JWoA5R(S!t0PT-owZ=p>cVw_cb=ioJBJI0N3PsL!NU8EQY{a?{$u8XrT zTR@BwG}?)>;nytAQ!V*rwknw%xh7wUh+XpSh>SW&O9Ec>EV@%IdUkYH_vL6_tQSB7 z_riTgnCBBgsQkTwd1&Am?)?i;>lWdumi&?k2HUU8;4UH}S*(KYm<-Rs1pqDi>lW9k z);gC6PVLuafEViph@txvgN1vMV!-;pE=TnG69Wf`!rI_JXRvN@A05CUvPQ`Te_f8` zDYN01xr&F(ybH@c&W2yND31BMRJ|_J{e)YWf0egOgP*kEa_+RGbeAp zEp($&;)z))b8#@oi8d#m>$C4U2FLC3YVjF;<}7Bc6O$~%H1(mgIBbMe^3yhtbBorx zb55yuro9HhO?EUa5zlSSFCfk>TB8!<8}gj28m!uru`-pN=q_UZL^ujZ3*OvXR7NF> z6D0AgwT#M?+2Z_&5wghRY97B9kagmhQezp3&B|}o3Jv8bJbo?uqSE6%X;*j~>u|(- zBH>_tp?h|>)P2QLuZ)2~#EJH8t_9{3sqYxte`rbKDkuG&E&I2K(QmWBmrvq*!o2^p&OEGf=FMBRWlquZ1Ue0krU#uh%#Orgk35(_EGXo0J4MV>&DRgbpNN!lC=j)L+lDH64beZQUa_iShJoz+9$X}Fw-1_{)7~W_0Qrz z-TLQJ#+$(UY%XgYz5Ws8(elHha7t9fih{5OC{qzLQ&xw16a`~TIBg`e_>c(%5y_;q zBM20#(8q?D5$V~LmWKr#3UMk_Frg4;bk+jcG8KZ$S|BXe>VZBKBGXE6Z{c7J_mtAh zP_QRf7l>V>W3)=e=ZBbpey(i2p4kUA%n+I1vxBdF`@P&iJ+h8qJN>SGP z)5yx_!y(av%zDBZzD}hIGQ*eYh}U`t@@W&9Pv8L;a+#nDMgnLfO`=} z#4>*-eE>p?153xM(#5?0aXuWJeDM;aEe%_0rLDA?i`(saF< z0zAQo5-S$Mk)a83Y8coK)iBPQ!S&(X(m=hD4v!Nt`Gjn7_PSFLgUntJQ^uQz4DYn5 z=xPcgL`2BLg1(vLK^aIEK_M}TR=xTbA(73$Vj&@UXhqD9g~J%~ugLGJXwR|4Q6r>` z^Vprj7~~<1NC9sWGEUSU!Dt5WM+z1(#!7)Q1Q|K@CtL8h+<2;t3PeXLD2*9{TNvb4%#aX6(i;>AJOMW9?p zucG2CBdi_h?*FQVl;k09mz)C54XkF`(tgWAzs#%# z4Ovurl+Ewd!3ge@8Y2%xkf0=nHhoLrzp|<~Nenh#7$*iw|7B3vn}|3I2xQBFD5{6NW!nB6cMm6`dl(W7jBr(Q;IGl z%@sJ+G;|B09iO*@-lK!j%Oj!|2hN?c8gUY!)W7#zy}e1GD3#dGd|9(dQ0EFtsd9UP zH5mj(y8?GqM)zY}fyIBi_0J`PH-Xjg0(5UyNP;|CSU8)1%?1Z!MM2qu3=E@P0cJGb z6`(-Sn#G4qAc)9tWgLNg!mJMxF(cBmE9K}A+5Aq04JH)AjLupBl-Z!)^AP;gf@l)@ z=!r}#4g@)6He$FZlpcy~^h7SF)Ujg3fuMDZ02$ar$dgPh|5u z^)wjI&l16_y$t*GZag$}k7BTaz&J5b`p-eY;53O#KyTtBviY658)PIpWs~8xbym z$mUpydy_N*>AqunW}!5nm>1VK5QA4+RT5e@x8w>WNuToCVx zQ)457cdX>l;hVsFl$QJ1{K=9%!6mc#8-)~7d!Q4Wn50eJCpwOpOhV2mw>??1fQ`&; zCrIH|Zs>~`(sHj&U+0I65CXxmjm&l9m{R6cU7efdDn?f&Iiqa;WXXauGMnEcje+CP zjWq07o6S#-#>}jM@R)4=WXWdcktyxu@VrV-)TMl-j$_%A1dc8;Buh5-)|GWiw^#aL zozw2%%!;i0$tTH@g<{mMJ6YU5KA!F$?;h;amsHUq6!a)jpxu!N2WB)wgy7%pJt&nd zSvdAF;<`?d2wwfUN*oe>qTF6+?)*G!Hiu7fcYd@a8!QVOZJR+Bls&9R%+;GHixX6_ zB5_J}P^GAG7;neA%svnEQ z@x~wnLuL76&Dat0A!J|?Vx$Zx6OaRS{mFn)a8XChKS$bcp&}K!V&{U}8!WFdQzPI! z#Y6nckXL|Q^F7E6EL>dI0jARUOZIs~kTK)u8tk)ao*I6l1NI?CkCite42vG4g+W<_ zd{*11FgTkCDj#~MBw4obk&0}(npWgfu}S;TlW3D7GE}}(!cs#k2afIq3UaXhvlaja zIeLg}hQ_7Y6U9yNZliltWHUemq40qs9LlI}Xxtdyf7@UF6AJ~Y7#K+dxAa4?4Qa}i zQ9ofwhU2htf2+UjI|VjE`=Q3T14}*>+KAY|(7nepD#_nqA=qeOln^M*mjPg30*YVe z{U=?L#9}~|0ocrO94ShAPtc3C8A~OO1vmvaRMT$z;um{4IWKnF(a4NlNLx;Ve+&7d z_@~r7yTJ|lb z3vKui>%8(iC3_=wciccUPf-b=SP)L-jYvTdRsf|pB>&%^Y&aVU0|B)BrxyFEmj5z8 zl+4LYpzucQdTqO9rajvH{pXI7)FM9B@@Gk7EdYvcNE08s{G*X0WP_Lg)Q0<__@@+K zPS5L$e@^Ws-N0vqsYQLN<-eQ-4)don;EP?{-9edI4;wpL{++@as#VXWf>-{@ICfXIrL6y$LV zYS18upn+TPp^QfJdYc;Uu>CXB9{^)Bk*fzzAciY?w1fmG3h+nie03Xmr1zJ`mRyxy$Li{|m+M}jM}?Ji>e zB&|Xvz7nS%26ZX8F)~XMyIMm*j0T%2ZvgY%_ro~BuZ_5M;+Ih4KrM_}MG&)Sn_u-c zcw*yikA*dugWXX2Lcl#`3cRFuEd;Nr5}>5A}O=&wIma=u#vPGWI>sOjMDmLjI$OAr{g1r ziG^#_o?8?$1J5UN4xKWr5FR-~?{I#(Q&EG;FAr0}8^-!rnKZ8+B!XjNacC?JOCwNU zLrW-O8c<)O^J^&7wz46u`8?EcSP7unK_R2H25$wP%wifjJE*wL2Rh+ zd(XR#=uro;0s1I~pW%9i^7J}H6fHD-5bG4ipblcmVYsC~Pz*zPHY*HCGbIIOs%0M? z#5x5qLi?e{xC7B4D1;$lpNRbsJ#vpiuyMdBArP7$D2vgT5TaLp{N##L97AOQHai+) z|5X|K#TtMel9_wS4d7USQxHQn?RI!}v4^r45)JpnFc4X^a0vcwgfNPKO6_N;oEvKs zKL3nO;B?DXi}FJIU6ct<=)42&l)05?g)dnt*H?NL!)rXDV>rvuDS{E&@FCWD<)8Hp zxe5Fn^!`MyyA02W4a%+~wm{&nbMNTJ>Bic;!!{S3RQtSY`?LXIDLR?ZByq5h)1D z3gpONte3+=igpb+$m0~npo2U_Mku}bugMWVWdeT6C0{doA)xLB9OQ8dV9_WMVCY@24TF>;&Ot4_r)G;~EI(Ld(P+8{;HLSivnG0pIx!iL< z>^8zD__jy6o%yCzT#n@RArnr$B^59z+kV|1-%{E3#H=vsJnJ%gQzqbZ<3-Fr&N+7~ zSI|e*V%j@l%@b`#J;zpMHxb-PD=5La4cT?)oKkNYx2X^5C4oojH_EnOx3G=Mwr?gz zO37K5k(x3SpBO4)-f{N1Q?G)`J|~9dSL?NAh13>h=ev(!j!Qt$qvhBUD7}yQH5W=;yt<0q-2V*8$fLYc1_I|c~>J+7*lFrG|Ic0{< zEPKkNb7_EBE(x)tZ2Q!yRMC-ftkuvNF=n5#;8y1FjQelfr%M)gQQ7uc=WKiEQRH5k zfFB%~Q3nxx`R}=>PVovV!#qd?ulw0pnMJN0BZ6OP9#FP@>eRNFK^Bxb$bq$f<;0nW zs2Bc&MyXTmg4%P7b4J5+=u(;v?OSi*kt6g3<(H>U#S1FiK1>B~7&4SoX5y;}=tmm% ztaTP?P^KYY!RT2UQMTc~ZJ#b#_(vT-5mO^Td+1G@G%t1>G0&1YqD=GDsdhmfxQN8D z#sKv$q&S@Nz(q55gnS4Y*n`(d84xA_#V+0)8ANHnpKYHm+rw9CBhStSw=IC`7WB!p zfaUso+de=pdd@v{3R_SK=7IirLy$3}GUr@3daS$&Vc5`Jv@j@(kOO&r3WKwWpz@)E z*mT(@yi!XLNi6eMwpvZ5(z;dS<(&8>UV<3)QsS;aS~>sus;DbnG0XVIr6%#E;S~(?DGd zqP`YKBSd+&I6qm9 zUgB-H^bAHXFU0a^5ua-Lx1{kFuog{UH0nWQ(SjrRx8c4h{wc+u;dX7TyY%^IYyvO; zPFV}p@?Q?bVg5ipi`eDg9h9C}g|VaM-zj9FTJ>BiSmlR`7C}!W0)W_ck)Vzi6<+?G z`WBIbAgll?Tr?oCKb{!9j*>YJ@;HSq=pYY~5lS!qX==GwM|7Cp>1mYKSSrC8W!tAt zSqnPQ!O+9&z6}1MUtPb#FT4erq(hmb$J z;L9LSc|S0tZk5c0_Hm+l>J+Y^Jw3}dukdnYPwmV@BQq)=ax|0>LU1oY3Z?mLdej!` zJ(f|)qoTL%)6`-%Z}yI%O=VtaD~)R^Q!97RnTJ zd58!m36Z00`_!pp(V1^T#i51;eTpf~20^G@^81k93CgxlozfLlwmmT!Ogc}XV1+W< zo*OShnkeTybt+WQN7YjD_iY@y3RJ34=A3gZS2ooT=Qd>5opVaPW!%;a=L|ZLQ0y7> zwtbpe*hXdBH^P@vax(e+Rhey13>7oCzu}*}W)_@L+2_Qt{Aw*@Gi53{KVm=p+K5{x zeknDU(b%l~MrrV`qWQG|jLNq6q|L$ZOirv*X4{j4F|#UQO)lF$voMQFIwwcxm3e)v zOtP2yjb%>qf+QQ?`Q2Fb|Hy>ONGmkW@gKVXhsYW>=gBLKgPk#SF5b%puj@JW%q>sei26 z;Ufm8)CCoI5b0##`NS!Bp^P$whl!YYLN9T)y;Jdm$}bNS#v8_@ADZpHOqr6drkRk2 z1$eVagE9>n@Ab@DjNPEO?bC^cf7J04A!iBNPh$Pu~zWDuqODBIqtb3tw7*}34h1#EQ-0n7E5Z66>PCzv~h zEvN+ZK!2J+XB?S~rUF4^W4 ze$_(wi2V>fOpnsQ+4fF7izp#bnje1oj(Vf3f5|bHgVQZnECy5=fX%(e$2Ti7^iyUc zAmgw@GU^~L2RIhsl(SGxyB(g@{XkWVSaU!l_rt%95JvG&s6EuR=!-{2amCn|xc2>? zAO2spC@-|%MM>RQ=L6L&BKEszP-fPHE&*p5I)yAk8$Lt@ul!=kTgBccVoS(TM~jLO zggw5E6a-}jnf8@uW{Gn>EN2J7Sr?7rA0lJ0*iW_mmnp6^&R?BG;~(2CJ%jP)j~-re zN?WLwKT8^K0dmwYc4U6~#XqI^a?GzU{yCEn-9KLb?dh*-`7dXI!~Cmj z%%sd@r`iWg&^qE0ORRbp>#0^fw>Mt-*DZRDUCD`odK4=bgj3%lQV@g{1S(wgCmYTV zf+~lz?bj?8q=P&}wkW;$LwyUu4&~%)bp(i^hSn7v=y1wf(18x7e^&Pcl`WJ>6*oSRJQ#@ zn{B_g%1n%4qH?@TnV`>5h_lX}G8R?oQEzazy;Hh^%C;vagGuLsf)&kONo8t1*LhEK{JZ{M-KkJPA60kEwued;(jl!f z=bV#5GiC;>MWi+PDwr5$2(Qf>< z{kjEURJOf0ItRPgWz40_wkHRpCr-M~&Sl#>g(#?`b8>WEndP9GGU;3zvq#2lOsym1 zlvc}0bh9(g>9qi?s~Jbx_UjgQQQ7t@-Lvi2*T>4F`C#YF{KhVkD8oE+%2rSr=0O5E z-G`3j7jk7PzLq#zh=eR`By9#+U~|Zz#)X8^`em0h4M}Iv5kuybx}f&lqUdRO4qaMT zt&B2+hoR@TtLmZr^317tLFJc+3F8ezPNh@km#Ybg5D*~^d)7LOG$_-^A_x?%sb>M6 zpltihsewTqKM_+SKzrz|ZwV+97DPHxwtePQyPytS#NK#gkTIe1z(q55gpdgt*dSb_ z3@8(jQDT2Gi1G}6wtco_fg-h$XXk?17C?WZIV(5$fnk~17vN%<^nw!1GpDcxm0%vA zgExd(A39d1pzG+P#f638)Vqil24N9U@uE+5IK26K%$= zz+u&UETf{{`&SM!r=CTW5Gc);0bs;FkraQl3qaZS*|NoeDg&??;3)GL`jRF6V#O(jv6q zMM>RQ=L6L&A~vsRP-fP{%fA1(apn}V2yOTf6|C|@MT_Q~I%0Rn4HU~Yu>_%X%gm{7 z5h)153ZTM;6c_DJHk=)VHNwxf&zwRRs^!1Ti=}aXB8LK%iD+o~w?i^2A)3FRZJ#-% zEmX^&9iF#Do%$A0{8NfApZdlw|GsQEwHN0fFaP%RSGD|?RB)I- zS&@!q6<3X79fV63tuw@`XR)4Y)pN(-m7mMN3$8|z-fHhHF{?#uuR;jI2K^!hL0N&5 znHRg(V~KN0A0u5H-$nQ*U8uKZCOEv(zFvD%+kAE|boa${{>u0zNk$J&mGnp`7!~sZc>5Rd>s_hyJ9c zRGW{^v0T|yFPu9SE4p(|s5g|Xkes;skX{malz#t$cf%rcYF1F$_KonRlpH8jAsLk@ z3T2u;K?gnYDEmBfs#8$e=S0u!Y8|Lgq0Bz#N2FO2g-h^jPg^_jOR2GQ;dNH#V$_w; z9F%RJW%h)X%C`5U&B5+08!I#F$qt#B6QER?Urlh>1iR4_@0nAGf=W6kDdUw{4yq}W z&ZRM9g+efHV`?24r?grOsfFi7TM|GJ4Py+(&$iDp3%jUn`_<0b_RyoGbJ|S&;K0n> zOMU`pm}eGzQ5ohzB6!`G!I(0`TswY0vape~8Dv43gN)Mpl?9{tvWr7UEKaElYR@gs z3j@!g;DvHu9qv71z9pN)`Q=W<3o5@nOd)F+Q29cc5Uw8N-b=&2Za0fG2-66Z!01_0 zQMU10KA@ji_(vT-5z{6?`-y`9l_yE!2>mDn&NO$bT~G%uB2Bz8%y79@c|fBfjFuT8 z0~>^klmTS|GD_@^%_!|h+4d6)6se6oD{gLEFp=}^lpFcLu=E6n&wX)%xl`DJN-z)5 z!5hN8If7Ltt?Q^JgkjNRv@j@(kg;Q*OmQ|LY%8mK82;P#*~G#}>LAvYagi{^XWLKK z8GV(oy@YSH3HmIJ_#oCPaX}qMvOTdD05vY8fS&R&GBj?4Uc=?~PfP#(kzLo%}+Hg1$??i8*F?S~rUF4^W4Ue2|P*elT^ z_b3D#2aFN|rTJOd35)AOUt(gc183W>SPZB#0Gk1hGmlqf=of1Mc1ULKMUzLd0H>UV zYTE7ayzZ|=PXYWcUMu@(UJEgm+P;ltgX?c?dz z(JR|`ro*|%24av}qBEx|E`_}EPr~btM5Z>cUX))&9W6#e5H{!+DG166 zhBVOb2<0j8JCrU)4utm`>^0lNQkhNwILCW7Wp@=pYY67_a*> z_)`ZuMp~+et73pU4R2bw!Krzf(AK+{J8}`P{BeCfY>3K*$%k^ zBHP|6TtRz!mIhYgp@KzovsUe+L&L;!Oc+1Tws-1TL<8@w*8t#a#XfGAzUV%XKjXi70Oh6ju>W+{kQEir$PmNR4pZc z-^QUUs8kU<5!zIIj%69IW6mvb>&`i)-dshpudX~!(A)OenuTptwtXXCDJ5sE_w!d} z4n9F0&OUdlQ&8FGL?7&G4fQE{j>9N(*m)ueer?396TgHS2Z~e7N_@=fjkE2Yx)fBl zy(c;jc0)A^={Q}PiBAs3=!^ifn78dSrw|2|bWV=WE3=bGr%bJv28iX75IfGccj{Ag zWSr7!?#1HZj1aUo&_SDnFUar&XWKi4DyVGxweH#W>k~_|@{#-CD9qe2$-+j`W{?GC4l|&%er3n_1AN4=Zs8iW=N6}mf#>UTq)oZ! z4v!q69yr_Hsdz!M!o0(A)Nt zB@6$k<0oQj1ZWSvX^)@8jw9w-l03>ZpE%Vnr~?;~IL;WLQ;7)exsqn=2>B2)utB&; z8Biu5N92@;P6`=BX}_OsKUuPeuhd4KrGwcPK!42C2v||^1e!8SKR_-@FrPSuEvN+Z zK!3a;a1}0QlY%laT}L$`3>&(O76xSz@-b|mOmQ|56gztMY_eqGBXtn#%D71MV6*L^ zC&nZ?!hMUtl`o~SG@`@EiBsZ&I*ep{;w?aihswjq(6|wTC3v?Hy$IeZg_kj5KfH6Q zE-4ktwx2lVEU1H6QvTf14>c^5hpj@l%!~#NSOOmC407TWt_bai8sje6W)(hA!6ITm zM2{RRAVLWCrbUzx2+fCz7BkJ1HqA~@pmKs3>pzG;u#3J7?v&Z~jL^$LnOP4mG$_k(;uNw7ZTJxDyz?KGV?FF2E6<^ zr7cv;za5^n0H|*v(I84Cc5Ost(ZV74x8c4h{t3m0>K0#h{uz5g+4d8stc7a%F9ppl z|ABfIvCF?ZC_S-ZoQaozr;vqe)pM!fl|PMM^@#psnU!pe5QGi-MGAtl0vQCx`avut zM%5J>;b+@VoI)3LkcY^Wr5FE1hJfmb4KpC47s9fG-?pD5Hoiv(I+($E-Iu|iI?y4H z+Yj$HoEO16rS3B9>xXwv)x~-LZSOs)Nn+6+4RQ$ivkN|#s22ie;>{2}IATe~@eQYN z1?}nCzIcUS*GDKqy)$Y(lN_-SoO%{fLLf9BDq5&fTR;e$e$OTZKihtiTFggf+pj6J z?V%qVgIA{3Gkr6YFW5Epaz4sBpEzYKsH}5_8dl$-l0`?7tbmm1>M}{RoCv;c7_T$m zl!~WcUA`H+K-u;ar*s9CZBGc7PUnG+BnUZHCe(A|(bFi(7`<&jaVk{MN7dc3?V&%B zW%ZG>IBY+hTj186b4tBuxLVPRSUCOm-?pEm7Pe8@_Km(tDLJW()Rd|7#85GF`x`zu zbE;EN+2_Qt{A$f%v*6{#$Rc5G2A>A)fCg4kB_Q<$BN$touq1Ay3 z6|*zWsIC-0sb<`N+kTQ+*hOXAuXoP2haSagS0>;G2WIA8S}0Mr{lqC-L1mZ+iQsi# z24l)3a_#v2$iiNCok13qImjrjUs-UbA?gJ>Vb3gFqxRh5yfE-QlOt`){djof7+v_@ zIdlKf_Q`bd>7%3Fm-n|OPQ?ozTDE z+wkAEpJevHl{$VR|JBP$JI zwD1TS*dSb_3_SH?xc5T9*7ajxqsPhHvG8O`sw>f(%Zr^E$y7|GJZTfhtfJ>^Mdh$vcC1n)MY z7r{HF@N#&sANDv^ms9~~+fU?!SW?97(l3KQWhTBbBt4;uW#50>e&Q6a2XzKTy%4FJ?I14~iOR+pky*NM#}*ZL}HCINLr@ z(<0UY?2wEBh$fF>0Zus!)wJ8;S=|p*wTPW|G;%-u+X!J4|CHL#a5&qSa2T7w>6R-t z)h@K(MVZiq&d*xSLYa`s7-G37(|Uok44pz2p$#8momYN292l|TC|6#tas&v5G7mvGd}zf;yiwfvW}z+wJCJ&V}I-5r#f^)S{$tDaNH zLbd9-RIthq6)k#>_Z7=OF?NrFutC2_K@e5|6)xT!K`^SW@PjDZe$`?@I>q4X=wN!{bzhF`sRJGIxQyOQf+cvj;k*dmDRq}&Uq8Ha zsxI=5vh7#x=`Rg(2>G)MzVqy!8lJHIGgBK{+bP4r@eQYN1?}lsvUr8(AYKf`HZfTZ z0zwoMLa+h8C?QaqzZP6-sR`z!KCv*!HQ<4m8tby=lzHdXWKg!D(ItXG3_0PuAowd`Z!wbx*vuc`4gPmkX?7q zDfO0dTQ5>!7H2U*1ifuPS+lT>%C>J*P)f;JlaZP-+nyLIV%~A~xl^5j%04HCt-rw|7=6!tEf1Xe?#-bBwJajB8Z2hID3R|UCLL=pVhh?d9?rFim1 zhpxnNY9`H{`3W>X7WypU`Qet!bs6A%q>=M`&3lILV&q(Q&djl3ELE=I-$JZdEb|_Y z6<*mUW0cm(_XxT(&FNiu#$Kv!4Lk?S*YWQ0 z>s!mibM@cx_F;XscFE_*3+D1xxkV@|J0A_tnA00`3j=fAJmTC$_~Q2A?)Htn>F4hZ zA1Lo8*x!sUyu5dCV|#Dw=y-Z(RNh~xN8ef=UoiLW9KW6yjb1u$ei-SNJ;%II5kkkQ z14RBi#6*giND&igs0=ZYAto}! zM248iaLtftRN;<5ECnSbi~98Vqyg`ff}tMCRPy>tB8qJ#KbCMVihs5ikMi%^dKfy z5fkW>HN?akVqy(3v4)sfLrkn8Ce{!WYlw+8Od(=o-FGGHh>3N?#5!VP9Wk+vm{>b^Fv-DS@M}mmVBm?C7-Ee$!98A@|jAOe5R5mpQ&WYXDV6p znM#)YnL&;%K^{Co9z0n_Odt=QAP=4(51t?oo*)mNAP=4(4~~3sg69S^O0+^VfjoGE zJa~dUc!E55f;@PFJa~czM;<&u9y~!FJV72jNs&4s51t?oo*)mNAP=4(502%9Ja~dU zc!E55f;@PFJb02Jn?N2sK^{Co9y~!F98WRw;0f~J3G(0x^56;b;0f~J$pjfT^56;b z;0f~JSSyhSPml*skOxnY2Tza(Pml*skOxmz(CUOdc!E4QHYDW16Xd}Ytkq1wa2Tzd)Pmu>tkq1wa2hThYj@G6OdGHK*@Cy6^57Zr;2HAZ8S>y6^57Zr;8^332hWfP&yWYtkO$9@2hWfP z&yWYtkO$9@2hWfP2P0{F@C)AfAtsOqhpLGFO!z_l3G(0*8nwY|Tybz}SZ<||tt-`l=@V`uvlTku$;F=w_9 zUq0GeS~xqpXwL0csrn1^qszwBKW#00KPTSLsrPdRKYzVhcyasiXnJjT|K+{ur(WB> zeP?gF_{_on(ed!U;t$BWzOQ(fXLxS&%sEJi9-eBY<#X2nhNni~U`AZ$aLeGHcAwna z?$dkQeKvdd|76~LfDk_Z)o_S^H>%Wa^j`dLrE+YP$>|$QqxW&2Y%Oa(N;Ds(nvb&P zqw&yOJi3#s_0}<59yuD$nN#DBn2W^)z51>ST%6drFHh|s)PJ1XI92^MziCi)054pZ z?H@8{t`{8)&zrN!edqK4jUO;(|L*Or^4sCu!m06##+1Vp4L8kG!}GiQckUiqwg7*v)2!v z+dqaHl&7EBJ~|#{#0=qIZIomerhc@sG&bh+wZog&^P4y0%jVqm!@K)8w~yf_&SR&f z>+FNP;j>4_dBge9_YKVXiWso?cygAbNpSfXeWydtEYeBbUJ+@Jq6s@~<~B~X1mjYY z&LEg{W=h;C%4VP>6DbkyKejdIGp4zH=G;TOP$|*I>!6upU)6ga^P{gd#$F_Iy1LZp zW6@dtz1ZvM(@tjz)mhT+j8TDBKPlBus_VyPE~EO%^pd*zE{iCJ57shk7SpTmbs@&^ zq51=7d2fA^IG@KONkE{nIZ^Zyp@(R2_}J&OC7Q;PxE=fk(Rs`-T4@ zo%z|}sm&IUIy>f+=JV#ljk~u1I&K~9e)F{M?co}hUD!MiuMY!5{z7fA+Iaf6YJt`QICu zANbVpH=i{dX8!4&9k?+&cN^bl@^p?~HXbmSlZ%%h8qA$O zbKm^=bHlUdOXf7RePMXnoGFaq2wIE>FO1Ha`J=nH^X3=L;DzBlJe>B}U}*q`)pYi# zc{mvk=H^e|2Uz~X#Y>kT96t2u+a7siF!(nH1M>

    2eP=Dwoe1v7lRdGK($bNA+P zuFSjpJG(chM^{f@-`#m_1cp(9zy)*l5p(|4>GaOl&hBln$s6z?@lZ~-+1oujek>QnhYMaF zHs^uI+i;~AOd;oRG`BDhrf6O(dW6Df!wY8qaC!^G09=9jmb(}DaRfh9cPg8Izk0rr z%~iO13g~AShA-~_!L8$(lfl)unG3MAY{BixS7E1Q|KSH<wDakGTSN55MPrh4 zxpw5LaTWGz+dJR1^;$#Vsrpy1+h3V)F=sb-w~uoDbHtO2KEFPlOR6>66+7A^=Aq<4 zq&`<4Hy83r+q!jdczgSJc;@K%5TrCZ0~_Pde(9Nw%jRPFH~9b0eD+H){)KA%&m3;= z3?DQXVOn6@d-K&hhX;4={rE4LM>Y=+ws&pHH>QNq^!97JwH>zR$~F8sSGjppTr=fPd~|=KVwIZSU%Qwy z$J70zgF{bm%{6oB+1J6v?%uoxF9+Ow1qeM`tq}|J6`{Falw{{hb$fdF@^mK`qf6+( zA=G@-%zx^&JGK7j3u3-7I%3~2=3H_nSNU^8SPBh%4L;-9eB}XcE*36mkgo!;8o^ox z|G%xzj>+dzfqy3!s2`d$?}nb6=*;umcb?wgDQxuF5;K6_UMjIj&Le<`g-3?(GWQ=G zzXF~b!ZmmE9o$i3t~Y-N`Is`elRNT-^TS8YFyB7HRu!U1u=l$Oe;mQ@hvxDQSOV-1 zZ*PM@r-utBUv5j^xbR3Wu3N*?w_ryGi;@GSkD8DG+zKv%3!q0R&{Ls++A3a-KycW3%qiF=&I zg`I(S1G=;=3X5pj)t?`j%bx~t2J(-V`&UJF=9A{!r}qvH4ximUskK?@<{vElyjib0 zc>aLL33fFq&0_w1ZW!jj8klpR+0Hc)mX8O`nOwtQa?V~a|1x~2G1RN)u7mt@w+^1T zTvqP8p06$=XjW=R(+rHT(%*clYwvsN>irOLs5&l}v;|nw>RyZA)P5g&D0{E4O$`U0 zG-p2p$oqJDh(vWCAR<5*J|HTc(D};`0(=1ve6ec&sW0E%-izwv2?w7#>j%cNVNkRSu#N$CE%$&GBmwc}X!96W z)m+ueF8g5!-t75Y^0@SovgL|WIoeMe17BY|xOIGc`!#T`XZ8+W-aRhB;{C?J4@c!c<-VHa z6=nxI*aa8&!U3Eu089d>Rw_OC{EN?TzdEg_61qQYeQHZF^I8A0Ia9*qoc0`Cft`(cI*TIQY-7cKXx1=zW<#z4$g|lAEsTme6znDDp$`qjg z0c_@uhHo?AZ?Xc3=Z~ zXg8pPTwC&~gY__C>b_$Dyd6O2Jn(RqoJVr+c=OKPts8fDUY;IX?>I9GGf$d%*c7sw z=ex#S^G=^BK}5c*f5ObaFvW(>Z6Xag<^0_n1Cik~pWmN;7R>dM#S3%SpUVRTJJUPI zuPn^Nznp`}K%p$TC0;h?4{qIRer~UUpp`kYC`EH&;rHwRx%eN9GasP_g6B8Rn){OZ z(_l&eH8`!$-#OmB4W5smksOsDlwTE!0;q>SN|%cYrPMC%1NC^LViY&xI@d z@Cc5a%=6~-i?FH9P0iiEwS~fMgaB1J(t@KUvt{OAgzd5ePKJ+|b9c6P0G7XeboD%7 zu=20%0i5m5yWljY{-_bL4a)p0=Fu1Twr@^f0ZW=5erfmkm1n27w(ss80|(W{sze%i ztJuaESK<3&7cngVFe^a%!BI3jK{35Zxd0Q(Iv{69592!8_rg23TNX{3tZF3)t z534uG`{;Q4@VJb{k$-9KJ9->-F>ug{4ubr(Eh z759LtfLKU5(Pyp~f3h|4xzk{ z0fG7VMO>~zIC(H^CkS;Q7PDv0T-!d0WDVBsZmhj%E?omKxqVn8k8|eq@%GEP>|o-W z^%myIK_I@XZ0#bU_Ruvz5;svy8U&F8%3O%kT!=1Ia*&bvthxUh>^XL(I}j3?7H3aj zJ2xOUl>Z$F1FeNgJ=_7m38A1Ob0Gim2oiDN!~Ddl{LuUl5cRnB3LN_ENv;l1nsWpL zaU&2xzP`ZU@{#hD#Qhe6Y_q=&f(lJ$8IkGmrxOQ3Wptz-)Cj@ z9!PvCKQE7B&KH;jfOP>#9vhg%T)cLCaHo!XYdXze{)R?}yP((MNxf+1r~5ng6h8oy zQM5dYo8=r+7diuHvyd~>IMRpAIT*$J<$}31-P_x}a|98P_UG^~`_r#%<>1r%mv`jL zEF4z6O8lew&N=gd%XqOW*3hPA(bwvL0PJeYtRjrQe|07Cf-$)u_Fw)Q*rkEtw)S}7 z$VzzOZ}GrB&L2(l$RN19I@&UV1j79E^V;1TtVQ_Cw7a9pwoNqdeTogJ-w4_l^qp@qnpcD?UEFoBzi3A+R=G-`;&3xNVb(R;dPght2kA0S?5_p3-ecPT=mU`O`9cm+ZsuTgueH2E z5a9f6*mmqfE-IWTTyOr)u6uBbV80@EHZQ`)rw#?xqRsseobTMedf)Z@k4DnYe(CKH ztenU;RL{Q-m`Hvq5KgnDLqACzi7(;3nls`IpBJZ!SNcB#*zm zeJja!Zmuuim<<2;PebLJKWyIRyve?3#dNs%&C|n!qdEbJdNti?2uZl*Msaw4^K6lz zG-%~Jo#)bc%zrd6?yK-aS3}03QSWzkrpi?>jB=Q#V7l+gbK{TS+19R0Lzf!!viYdP zZ774{7l&$sXJx2!2!_Xb5e9Juag1$U;LFu{jG2kEG<5G}^1s1*7fEO!hbLRHO zu^A%C{Hx~OE)ji7_}d5OZEoMN*z7D;&b)H!qrl*yf3K6)40BBaa?WA1^matWQ2{Bo zj-A|2{835j#7=c!#uro#Be*pB!s8;PL(HC5qtF#6udft7|+HtFMa{C z@;_R;`}b|W_gtRc{EjCUk8j^uoZf*A#3CHT?(RPU*WM0`Y-!iP*u-aZVjeE?vg^Ru z(PHVz%X~GTfcMk{Ji7TX6z#|hu?#j$ea7_r19!~%F<)mSPbIzF9(P#Bghy4X97&f? zkT688F=xjYn{Y9t$C`hg-jWg9w2?=mKR6^iR~xrNW>uy<%(AycT1w&j-=J0|I~%KM0y0atPh?yx*PU{sm$qHsl|LbSup z2MR>XXxDiIe#$$;X#Y6!g;>(#%Z-r|M<2uOA83V%8J_Xdv*Q{JkJi6X#+=pN>Oi>p zVimydyWhWrBkNwIedJA*8)F|8@%2#|TUXB3yM>th++%~Ty8HEY4xV!v#HR#L<`vy! z-lP&znJM73Ff=$=pO@Xr4AAPx3{U-QgVq{uILVpc4d?C#a(hRhm-|&n|O)Hs4;KSi#Xd0Eql(c&q%_)#|f?PY2IBhqFigbdpg>E84t`Kh0! zs3WM0lOL5|mgdZY*E#d2S}C7HC;hVr3Q_Q=JAuX3aVPOKKkfK}m)_V)b{w_k?lSbQ zrgYmw=E33pMd`LWNAinsv4BmXWm0QLZs!bZ=C2dE!G5xEtD~pUb?43t6+hpzPilU^ zkY9tZt~Nhd3$V(2-bm2>ZF_-zAS@|Yocqr6kdylEqPpqy_Kj(ogtchzr7Y%GOseND z`rD=a+XFM=#~eEMVn%5G0>swallk%c(BwX7?%%l7xSo8Il?$S{MU(GfejM(@`BZ%c z?#FI9l5s=Ia0FJoD*$=g#UV;-i^o@gb$au9AHS=1H4kqNt9{_y)pOtV65Pg|Kh;8? z(k^h${KbKJdL~Y8=V`yyw6hgm1x6t8q4b#-H_f>LJiBr_zftPCJQh5fAB7D2nTS6+ z-hu6XzvP>r{lSmx$p4db=5sTV?;p+iXLX| zFirCVbLNAgsjm**>O#UH{zo(EQj~ZeMwiy}y<=GO^XYMKwCAm=1!|Ck?DvO6a|ow_ z&KW{TG#ZlX@nZyG?wEJ^Bv*;NZAj4{0Q)!L_>dXW28;!WHj)XJUZlBF&(*vaK;FVq zsPBO;7I!ZyJmMJ3Ji85R1z_ffr#1$s%r6bhg3ok4!mzn>I#Q|6h!Y={Og%)w#P|ww zyzs=phRI!S5fM3)d&uGzYyJvV9;Ya;T^H9@a|fC~Xa)Hbc~*7t1>A3`Z?MiS4a_lK z=_*~?T5fAoOOAZNe;(Ao0&@$Sk5;I3{_4dm0O;O;|8it%{urE}`J!!xwHn2We}rtX zXw%Fr?(dToJ^KF(bLOQUSPAX{dQ;~X zARqJ9=I4Y>LA=(p*%RD@i5T^#r?nd#%&w72Id1QoowwN9N#bq+f@EyXFKdVgG!)12Ryb`rXE}<&R0r1Ha{|F zKC0qt(c4Gpq=91;-L+1F`7;QJ-gN{m>Bk0UtP_~GyU+>jO;P4U9vdc6*f3Y_3L7$+ zEBTcN78xcM8KT&B#zVr2)>>rVK0@4kn5ntq&tm>!;wuNoi}hXbI&erKnm0b#_LtmA z#4Yn)c9{B(xipS8R!Z!dYIcfS?)4Z9o$W;DVw*JQGB|I3jo2twBMhq>_9K3N@n)z+ z-6%L0qp%ojH5iP~71!_cUDHF={d4!CKD!PVJFmHObLP)Mik&}e^=lpFpLo|-@iXeM zHstG?eca>5Fn^V}-l??zZM*ol;Eqw+^QCROo^6}gR*nBquLIew`P?KsF)vwjNvdt_E{hM!p}$}2+c>Y*K@#2lO9ad_*|7Q1I6Uu$eET+<;`~w4{AJQn7wL_(TyWD>oNK~Mr$vMaAfbKX z6jmE^V4es|32-005GK@9&_4rgV1vKZZA4pS;g#Anwl=?^NY4Ed%wPHD>Xp~7eEi8P zuV2~O-MO-VaC{|y{NoB-gU??teEciBN7Ieht~?eV_qC0s4_|q*m1f3gmHx-wS!oJ7 za9;Syn#F80DOmjI>l(V@L2*R4`HEQyb6m}B^~X;B{^?S^6t!w*beLmEzK9#SDW}1S z=c}Uu`HC0EP*Y)}$9d%p-Tc;^S?oCB!|MIwk#6}K#Ytf)vfZ-A6}%SNd8+CBb%?RU zMd%`1@a)<7v-4wELCdGW;LQiPBQbCO#sG(^;H@@zx)62|&LN{dU4%zwbK{jJg8Wf4 z9zI4!kcrWm`0=Scfb6G}nEwfGYx;Ub^r;Ue#}Z9#yR_S(zd}UBJ2TM^BlpLKK8XBh z=gf;U&qXkVjN*CSC6&MOBb{iS7J|mHf^|d2E_5B3|7FfRJG0oD=yC5efCHc11zuz} z)i8Jvy)Xz{%5uugpC>YNx{_P&wtAV8Q9o4+Yvt<#)B8RGYV-n(D1L;8M4CPC^b7#J74#xu9 z9Qd9QU1Ml}K0@)b_P^swX}dW$9?Wel5P(=^{f`KoVBa>{M?&4+p2DH7T9tCEXg zh4$qj9Ps_gAN{_MxSJ~et{}K^bX012Zw*+&Z%3FLk!>EWdKcbXi6BpG}j&D*}l1sOoyEjX(&b;BNE)@wtaK7hy9xewm zUO=Te&`sAp*niW3La3*ZQ>a|ZbxM$vTyF?SZb3Z6m#FF?t`%G4>Jk+ljJX1Fm^a|RZo!xX92&02golfv;>gzBvgF7jJd09e zLCj=6u8zqL146;CieqX(fAhs;AmlN#7}?KA%cfkNtGz3FMJ8+l2l0n}5&nUCDJyZc-cLt{tzDYSa*`s#Po%&5uE&0OV} zEEDiN4O-7>p)i<){&*1Q?vTt!#A*cqj*EwA2)g4#8=jrO6t zrOp?NuZ_etlqINkh>$Rea*FHz;dz1S&hCk2mF7aXNPRC6tMLZP1Jy;bA)MyyeFB}cQh1Vh`MJz_>nf{*`mk>Q zCX^#Qgg3a0NPX)pNw{EKpANYuCmj#Y=g(I@T)OJ^i1ttjT{T&*&(hj0N*Zg1{aYr| zUKt)1%8z9R<`0<%TeLXZK@#b{%Oq-sYV?zOrK|YLW7s8ATiZulWihTA9m#kwp9cLD z{DJyy0e9S1()BFs>c2bigj$5=i%Q=0GU|?7sGH@tzQN&sQjnJ4v@gmMl=Z*!^HNF3 z`ScbKNHut>~Dt3XcesQs(%b~P}!>m{gSq&}sMyW>B{oX+Q`+r|BUcpUN3 z7I)2f)~@7*r(RZQdTT2iw6AeUN7m(Q3Hb%3Yc0z5^3%?(rN!-I(C|DTbMa`pcWV(! zOC7$x`0DgT65;_4C>3}`oU#h z-FLovVU@D%_S(S*^L~C{uJ=5Zt(U;=5rg`?&Yznz-_S3E!6eUnVH@McEGKT zhLF5+rpSRX|6m~h|M?cMfjZ)VXL+Cmj`CjS-M_wLT7dCcl`-l zp!weh^>dRN6}Yjxebhovtz@~U8(_F@fu3{c zdJv1hfSCVzu8VRO)8-j|m2sdIP@R#k{T^${xCc4Vwk+v_w7!`Cy~*oPv*j6hM=4tK z?m-w5+vaa~@`c`$4$QuG6t^DcQ(zx=P(I+$Zq$Ae!u-8~S=KMjo3OAv4G$fht^U%* z=T&c9R2e9pw8YN!Bj!SJpZHe(BsLUpAD({UvtN2fTAZtmJn*T4daq6+w(+qVXb0YKg)7`Gd=G5SvVjLXR6%PkYwequ_k=n9=KW{p%-8ohZ^5m5|JLA! za0ik}RuG*$T<&2S++Du;ojLRQKIPhM+VACT>9&yF^xUqYJ%~6QF8Sp-^R+$7z1*z# zJ(Hu}m&#~lW+zi&InAj6gr`-o$@bnt=iN5RIgWs>c3!)5fBQDPQ~-q+de9{wkV6GL zYDaH=T8YywuYRI`M^TOXRA31__aQ|Z&WG*v?<#S-Bh?t(ZkxZ!>kzyf^9k)bRP62C zo4~Ct`6rt#`NX8dW%ZUkXFcz^rw#ef4eF;gWe8ml+cJ2CEsg-6h)VRc0tZ?!M0w%y(JqL~43$dQ|F?df?@%&Gh7xpb#K zHs0-$(q&VwK0S8_&~LqaVP}@jxAyd0&t7}`&E!TyrOx^;c()7*myR4!7C#T0dSPf| zZnUT8ZpZTc1-Ia;({ufJ#W7efoA6$Q-r9z1PtWz-HXCu(>AAjNZ`O3D=lbqTt?#hM z>%{`pr{`)$I-H)X`z%t>RUy6(UBZ=N>&bdi?C_>adwTA&(ZcDoZ_In)6hV7tYpy*# zclq^^yXt(`?Lc>W?zRw@YZ1)%B2605hqXz2dfsBy+oX4v?wh_sX;05fF6FIyY*fnj z-P2}WdwOmMZY=}i28r0q_rNzYd3iUlK0UV_`i8IC$=K$)({uZM%jk=Mj%evUh>*5( zX)e{L=MGz*fp?Umb%{MNEV8COJug{PZs1j?=VgyoPIfM?{f$eQZ;FD-3ZOQYw#gdQ}7Y<(8hzdrr_MUi}1r!25$Bx zny^rl2eNo)de~G^r-cpjE#76$Z`$Pd0`$TL^L@M%HN*4Q;r8AUyh%B&ixQ}&X%xZq zSQ2V3&*~n{Q#rra3o+$4Z<^yXs2-#;sOCCK`(ha_>nkANH{ieSiVZOqm8=_btZH`+ zB#vVEZUh&r+TY(?beC14QEezHdM}mx|5Qd`9eP~#?lxJYtdkv8tk0_AzuH`RmqlJV z-bXrZSI{e_gc-6cP6yWskp7c5pSnem><&nQo^xrkFngjy(=47@X_f~$7;p|O7- zJ{C?o47Hv%;9}f3JPjVCT?#IL_qlpP zCpcjFgpNvq3i3uBCnG|1P=j0r*r1#W2|7!lIib@2Jg7TPU<=;{w^J@pUj^S~a;o0gh^^-`lY6G`gVOGem zxFvsAcay1pKB2FtlC)cZUshxOT1SNk3|WF|74ENbpeia=vr=hg3{ZtiZk?>kd08oX z#+}!fjQIMJdUgyXH+Nc33SeJOODAi9cHOkB{pOIKp{@bh`>Ts(#HO#qh4!F#>DH)A-zp$%NPh-yf zY{%s@?RcK$Oi;JQv*Oo^xm}1$NXe>hlYv`Y0n=s z>%&#eWg}w1E(z@7Ov0)hdPRyDzLk0dXs!Jao_n*L+1ihL5(Blw&m?#*F#dAN;tx{P9Otzc^>M`exEgm>%Bm{bF=)EQJhu^`$Vdn%X#=1924B+RY!G z4?Y>$-B#+|)V-d)(LJ((z9vRkn!Q)&^3wI6?^)N&Fxj=9do9x5@5%CN9b8XqPf+~k zH^bRWg!Z?xUW0?=d3^pw^qv?*>92Igduz2d@Ycb4gid&`NQLsB2(gX&R&YhuH^`oS zWUi-4AQNcwr+7n8^bsQRuee4?`Kw?>aOZWii$=-m@GS=>CF{oAHEPK`r$b7eH*l9B z6nH}&wiX9HFT^4dag75}gI}OT8S>_@gXfyF93UT_f@>!=l7@?Gcw-FM+xeL`NvIpT=q;{ z6hXPgo^h_=xN^oG+}RE@{aWxEiN|wqc_>iPE_57(hb}WwI<639E{fLUHY*lnqtsdF zTtWrZG16j^s@v-1-&Rs%43c>G17}wMlpbEThqoCivE&D@m*KhNgxE;g{7N_z%3>SIyU>BW_!L^kS4lk&iT7tgUO-J9II$C=DO7-(bm zZC&!?$x+h12O^M6HP3kIJfAOb-@d)wI>9e0Ay(zCrQyb(ZB~X)n2!%PqE?1tcz$W{ zH`9=iq4dsnI*TDfYeRp$^tw3w{G3rjGml?C%p51R?#lzMPCcqh_ccIToAM+*C7$vH z>_wjA1Z-D`{uEHMSp!vvU+C&>ZXeu+8kk3mc_t66i#vC3<^?ggpw8*;&FN8Z*M+|$ zLf2i=hNqtg0)scIEsPuWED1ZA??Lz7FPRG)=ks?et`zlNp(euh-5t2)4$oA;g7NBf zdS?r+93SO3-SbN)&&JNtwc~?3pE=y#*@YP!UUYvNbzVqZ;f{kW)!Tpb&j#kXX!5^& zH-C8k+1*>Wribue?sWLzHUC##X1EtQ+IbD{BmP*pcnQL;wVY434|lhL=X+>lKcbFY z+U3qCud=iL<~UdeBSKd4zhJIempPEVZFCBgP@OXG%@yBR3*>#{94nk%as|j1pAmx(b z88A&aeAURqK#_tONOb~2p(h+1r#+i|yKwh^&&*v?Uc#Em^Uk8YAmL`R!kp&^^)pDf zIm1&X$=Oxr#`Ji**l^FXyyV)Hbm11()ESf)mUg(ZGyn9?&a6Z)&oaXukL}=5gs&t0 z$nO3zye}3Y$NLad=GdEVhMafbn-FSefGxCQvU+j}l_Yy-7)YB;oEtk(7 z0Oe&WU+R;!_WS(zggO1i1-3jMs|l^p6}o)l-BG}o7W%h__AClJh9Z%DfzGbzVH&_1 znB*)$BXRMn=NKv5ZQV4B7YB(4bN$_7iP=j#W7L(8#)_2uE*k1LhnnZ zgf%L)uSL{@GssiM>9@^TnY6;eJ*Wf<`jYnv7P_tt_z;UPkH_-rhk`qbiYbrB59b~; z`U~=y4&R0?N;+Sfef|tv9*<|!i~n|%>YHpK02=6EC&3?m$)LPCytRF_bqvW82Zwdo z_y1$=-D2!Ks(f+pZPSE`lYiLZ(%NDKLV`q%qQ>nUBHK8aCJ6!JBw&YmLamOwGil&ze!Hwg`hIY@veI?{j$LgMxS65_w=yVP~9 zRjV%B@L?^rhepbk~w9sL9a-VG7v_dVsD@H>cYc)fy! z*9|>Sk)STT6S7!C0Dw_Tx%WXon1PiTXX&er4SM}3a-JeZQegYDxMx(dUJW=ts=s7B zih@p)+OPXiC21^$XJIy(p}6TmdcFUA^79FGb4K}Q$&n(9g}or@rY zXW-@9+T%E8bMq~Cob+GxM2bg)6yx}42tmfR7PFw+CIMEAEC3v{s>i&%d&a3RX6)V6 zM&iIb$Yw@)=ddS7+WQ0s5V93uBY;w09$$g6yjRxm^U=D=M;x4y9O>ym@@c4X6?1La zs$AwEm>ijZB;&Dkc60a9*&+SEM28~9>YNQ_h_MUITQL$x618q&wb@rP#Mmv{$7jT1 z9JV?azaKQ07sNKLQfroNvGxNhlu-M>n!&bi*#hl{t;*G{<8uz*aw#~PNx^9KvW!GV za%)-rc%t5fW(kH&Y9U_L2~+ZUM6J$lqf0)KWvHBuhC0u}8-vyZyl)O!)N_x+XjE6x zHro3t($M4L0k@OyERF7YUn5a@I8==Zk|%Ngb4sA5Wy>c>2*l}|mOAb~<(P-jvaaJ> z!V@=DO!rbYKb;a~blF@48{6BbhuWxFAeDJj?X_mQij_(6Y4DD_oA9H*^zm}I5?Rd~ z8n^WduZ7`Y&mR6V7xRS( z9eH2xj!GA)E(}mF8NFAJCqyD}s8tYI~z zdMH1SThyZ(XR>f{3w4Ao9>m4SuqP-jj>j^p6;3)kYD+H1QP~dqj-2aMaZiL8>_~g& zC89Ki&?P_C5JpA611KBnkc`$zgI^bYdvqYHrCUT=W{vX?DR2!|3wtlFkUK?W!HFML zmc^SMFhi)Zw*ghaEE~lwG@%+iM}3MKP8l!AO(Ko|HAHUtYsy#&38)a2&I{Is&|kaJ zR-aA1p#}vH`nXRRhJ+x5Kgcqk!5InY5-cUpVLtBUtrm3S)`fkBeIqpxCf(2+lOKlO#dvCsa>3z10w&21Be#$_vM9 zpci#eegrWs^UVC%MBOWUnT<-9#-wp#3~ZYYM5mtB4MK%Em)Dy{pUFghbjk;%Vxf+k z8cqXVu1+&Al-Rr(Pm&6zy(|*=$>=FLw>zSVfSQSexDlib6)aRgpnzTZ#Rc&*T99)b zJ&j*+^-7xEM)Uuf0uE%>@E~Mxv{&ZW-~$?gCO4`>&DkrM07o|WdXSjxE8fzRbZvE; zK=!tra6-~zmT)U)|KBg%{{Zzj8=|^h`F_OfDnWgUdhCTKX9^u1nfd zWvf^hndxm+_HR4&%$JfEpiLhfXl6fy+|fr?H}|3QRwFO2?roocWc3|SY+t&PKXFz# z@QGcHs^o<#-n>d)cy3`>ChX$!iryxq##bp2DXi=Tq;729V+DCWSc~)YBFl@Vv)9(X99UylzpbsV2DPPe#qT1{h8N`H-qXYEQFz2@XqDM?*^ECw*oMz;vReIlih?z< zl1iM8F_F7ism&Ln^=&xJ7~pWS>@!n!x~X>e_YMX*R3}puD5vI=fTL?47ER1(lIrs>vt^}`SdC(ir8mH+6+lkGSVd)Lq_Olr_Z>!LSa4t)1yt9SS z^Z-)qeaCY*D0s}~uHfnz9nvPoOam8SB{>9@T|HlvEE>}hLKGe5+l|tPFvS3CzGoKp z%z$?cxtK7LF_3~$F?-Q)-sdjt?r-kwZeN;BX9ti=jX1ZO8S#v|s=a%2SuJd=y|N^HOkPBYaXv+ly6(iu~&SFSQ)gw7{M{dIzu zZsl{Mf@i~91rVhemIkqvq43#OKxY{30lA|@iOhDvGrE3jbANN=5y1Q2+@CGY4&8t6 z-@lH#35GHz0*W=F^Rev$FTYel?PAS*d;*HKm`)dp)wA_P?OU-;cX)vZ0VpWeV$yu? zzJUzmeapLe2`JWL;*y@d$S<-&@$M+7wA?)h(8XID8h3y5e2-q%#IJ842G6Vi!&E&W z@hYX1LR7Nv?!};3TS-giVM@0=K(U&)a03>n{Vf8DwUk4>FW*{{^9P%%hUdp$6D|`_ ztPGdw$9qzyEno0fcjFg4?(kDktPHK8|1Nrf+MS_LtlV<0vlp63?rq)T&ct8)Ntsem ztQ?so&w9&M?yVMt1RL;0XL>P!Pn161V;7JP#Il~##1zPZRnwHB?b$FweLBNv^*(?o?C`Q_rPZt%)QF#vPuwAxsLi+yrw?!BvH zyLP#Ece!SMxit6gF78>J@Im3BX^A}Z(n!C%B_uVi6ZA@&dw0LZljnQq3);KfKbPj- z1TkMk>b4x#JP9(bG-T(KV6i2mwz{qnk@G&|9t4H zH_`jF7a+~OiwVidy^Fuu4NnmjFlBsF`|Q_-$>zco=iXg1yl1>;?vv8R{Prd!M=yWE zbC>4cU3x>`U6g2WKOoD!yT4!!ijiI$ZNVS@B2k7AL(<&4M{4l!&*pApb|eWG=iYlN zc~8DHG-tov$d{Mq-tEGk8E8O9bc7?IFYlt}eZDxlprpBXOJR8MMV^fPxh(f?|GjtU z>vHe*qeP)jv0pFEz1tmxJ%X8Bw8y)Fj&Wp*bMHNAZTx#IWgofK{epiJx%Yl6k&mBY zv;sia{mtwv=K4R*^jH zhkN;k_09crkM{rZ)-@E>Gy19F4myv8{^uJ>B`|`!j-L1QJ9^YHt+A&v*6T9FpYRx#jf&Jfw z`5?ReOxeHBZ}LnCgbC-OtHXXl6Vo-8Yogo>A;HQ?f1hZ5R6=F*&;>DNlR|}lNvLe5 z|5T(v-S*}p*TS+-&Lw*`FUp>5AWSg%-03{^qHoAfCyWpiv7lSBKc5&RTXZIXWwLk8 zK#SploNek~gu;eGTI(yD*%@vdYPZO@W?*kw$7eC`~?MW zj5zNB1c+}A^@OZVYHProKNx|h(^?Q)CT+hvo^G;ds%c!2+dxv}bXFSJ5F1NFA5Bi*CA-e6(MNz_WA$@&~ zHidRw-X-g&8$Sw_D?S*s;{uS2poJ!Yll@I$&?Q*hf`C{q?=QPtVdiOB2S+WY!$>eYAiK8MfGorZvtoQ}HXtjK ziDPGYsDeiwcL9O$)d_VZaxEz{784oCPQFA9Br^UC%m2whstxdA_*aBlS72J8HUIySq5!a7eZ z*sS&)F2WQm0BqNO)q&>0w|HEAU+#zyxtQNK&$2IR`2 z50j<3I�-RmCsFy#5uq{+o3562c#^C^jxaZKX6(tt?y^eMQeOOuG!34v-R#64OJO!Dy*8%X4AZZ#e|(l%8fx{H`u{5{(lf69ywr&5CNvuoDIe8s z1^Y}oOyS98s9m+#bp>2Y!uEM9c6PPvm1o|Y!W#M~#X@dei6V#wjYJPCd$lM|u8%7+ znNq=qA#9}Hm&RTVw&CZw+$t1jPkH%AXmqdNN1Hm7rzh%-!aLq5gf{`5twc%HcL%KQ zYf(HGIIZz$pSdd<^{ZN$n<|tPY;QnMs5(%GQuzf*CCb%l5?(m5&(Y6ss2P=n>w1*5 zDwwxD!%~I&LH}767nbzd$T%}|43(mCq*8gY?5Oy)eO3KDT?rO^ zmZ?j@2DggOaV7h$1;+X{tDvo5er_Rz@1FB*bSocm-pgJ zNbtECnNj$;BJ89`Nu4MXXApKSV%fZrCUIDZE6w zxImP~Da@89K9_s64EGw#aHo3p?0Au7=vJUjL-@R7DTI0r;pJ_`*{z?l4ebbRY8-wx zRl=8BKyMg@F5_ChdBD|)H4p7rUtk|Vp@h$^kikJMQF`OGZv|jPt^Unx6L>b zd%=U@-(R5X?$(B(*z#B2yfG)Nyj~o&!{4B;KmHuucKj0PiJx>$%R5`p1P>?{U@#ur z-oCg2P0V|`5ne%y@Li=cVg|@(8xohfa@y3NCh85v+Yd*SV?*ULB;16LmiDB-p+1ow z>>9mYM?R0!-Qn%A^~564yvxeeNC=1FJLx3RO6hi+D2A5mX`gzmKhnW|6CR8y&3j(yh{$TH$>?? zNySTC)!ZJ324rGs zE}uNI8iJm^S4f(V?!OS?NY@xqN=9+s?SCr$>XCzNj;ot_!5ygSg9);yhb3xDEBWVpo!I`Wg zX7ACh@F)9S*NCjXR5cGF}A@!w3Y8s)tvM@2sx@L(rGAKrt|cNTAK?rrQt zEyaaPV@OIoRE$_@<%MID@{Q|gz7-?(H`t zspfeHFCZL#NleV_&^V1%2P6Jv5m+^qLvi#eLghr;1zGkO12fn5R4CN;C#ofRb^RSB5*m7WgoooB#Xy; z;8X{$zcpA>I{xrw3L6$*baoI{CU9$zUkq{}dZCaXbabUzWlmiV&mcOtdU<>m$N^Jt zzU2-dTUG&vF%F-G5Cn_%QDYo(1^)$jvsL=>-W9Jch?A=P;f?aqzBMY2TwzR^>7W!K9gG_^Lia zAzLAAJG;62=5SL!&c|w_k-s0 zg4m`-vc|LRfRJrp&0t%%Y=QQ}R^{r}@j0z!hO;V{f}@!fj8-qpNUWLtrAbjgo~SpW zq3xlTqAkvKV^}G)@ty4u>d|3cdUcQQc zPyzVoeT_ur!BjOSNS?%jNrMlyAt*=)#E?`ib=<$&nRw!|uH#$66E{^%_fj@Lof2hq z`C2}{C>JA@c~kARMy`sLN%1Mb_PD!=miCuEUJh3xt9e7?#@!Ty8XGUPhTjST=(kF_ zNr%>wQp4hFHT#YcWWp>>&m|pT!=H7SgoDl;F=OU!6q2pBui_ROK1=JE1|7$rIvDqY z;F_3w*UWN2uQm=I?ordldE8uBmuVQI3CGMsYE|7fdxh{JrN2pc(u5BVCOo*Ldv0Hz zKCpU0`nYlnU?;VL=wu9hNqurEQL;`f!2uOAVqPqD_C!{21wW6RSk%|3a89he!9f&q z$sWTB!dMaja7ADgf|xdLe~@^F_DGKszNN<_H+EeRvR=|k5rnJ_?-YO;;5D({7S`c- zz7}SOJ$v}eT+A0DbmV=#J1SZU%(h?@$n{&B`W~B-e8|sNSt!#Tr&a zss{q(jOMCwCJPsDgM+NpOt-8@uS+qXStj-6GO5Yn*>bVQ;Wn*n4q>+$kapPP{d&jOR00P-;+VKou~H zkK#&$-=CspSH=tWI7Fh8+HEg!O&Kd8!5pGepusL7^yg2$!i=nxHJ?qrp#}J% z{s^JAN=(xX{<%IK)Kxk9_Uc;{Tx;+#7fU9XY=t961;6*<+Dhb{J_yhkhL1@N_5u&N zJ}l2GriToKry+w64`R(Zyq`%uSqF`AP=IjU*kCKJy!8Ed=F+*dPz$B%aDdcPsn_SQ zUHw)ba+gw($aJDZ_ZhWUa8pCV^~Kzy3o<~xf#-F>C99?vsZTg%>>o-P@g}iB-a)Sv z8CE!H!U{*#k=bf8tl)pQQD_0i#jJ3`@E2_tUZ^Bq5MucJlmIHj#T0}YGMF;2(uS2v zPUMFbxB!Pn5@TbhzCMJ=I~iQ#3JaI4xvt}m4rU6agM*=rAwbVF;Q}5jcbZduyIPyQ zGUHSyv3GQ$n7#C@aAF)CIRy~vfaWN`HQ3nQd>pEEX9uc@5**hrObMvw=(zTh3eIZ| zU!(^XSBaIjC$5&aIylTeKLVu)k8BR(?6IuXa_$J}iYB~(JyL}c?qb1cH@`@Dsg}j= zh6e17)LVxr*3YChje7tu&a)JZjS;N+G=;H5?uS9KVcig%X=W!$g49o_o^X1r62zLM zyl}h*deQw07hivAb8BaNci&Q%cAY*pQTNJTW~0(w15BEaPWhlzEYxvR!)d_F)oJF1 z5}P;UiIqhPSVlW*eLJN9Da=R~Vd+D}EPoHr)Ju8nT6`{B0r(H$kobJXIJxO@;b^1?y8HiWq7=i@Aht3TKHCcD45K|Dca$iRwe-am z(F7E0MCW7M1zvus4%)?<`S=7BYcZX!hu?rq+;#V_2qJbsL9rH-=6m-GAx`ew6Hu(h z#3em@kzZsL)Vrhbn3pd_CI&aW;A@~*8NTRkB%xTD-}96jpB{r^E&ggCEt!WY7Eafv zZf$4-O-~A!f?_S@Q18n>Ac+r1*#eTBKiKqxo*#crxJ*E?GF97#EV|UbP}Mip8L9!T=XKdYzfPcGP!z< z?GR*oF@R50jJ(IrdagpKL~>cqMfdJi3R#o$ zEjLxH23b3k$T2Ap&@$82hQ z;BWfm`Q+02;@rDOYVh#S=B`oZHhf8PK)W^=97gmS?y2NG`O>h7{kGU$=lwNl?%giz znSlyNyJalT?k>Xj6X)LTH^PH2@?`AKWx03z@4Z9cd-a?*%ALgGN)5XkJW8nND~fzl z`=O#Bo&C8z-VM$Q$4@2Az4v=-ox%Wnnz2U7m_ukVU`S=+|J16&UwV+cX zv)ytuvJVJr4BoQc`6&G?q_KkjW#hoA7p>2SfmMfIdeck5HN6UdG@A7(_~SlFxbano zbk8ByFdNB#*o;P#>2@_2g-H4~fk9NVpLkSTZXZ@lXS;>FVN16}Xlc(^Hy?j4prv(q zMHBeTVfb@OeGqSSjmR@lc&is-5-rbc;O~Jc{F%TX-E<&lHVi_;KzQig%${==KyCIe zUfA14G4rBdj?%$8fTDlw6K#3TaaS`I0H`&GyOAwtJAO@Hc~E^wXqAU8ZX8*xVnYZ< z#A~x1tJ=A7Mzkv@;XpU)oOF^I zr(tmy_(c7?Fdt->pDFwI`Awb)(M2GltHXXl6O+)mNq=K>8mBwYHTiI37=@K8TL<6A zDVr2R?@K~uGySI`1?sjp7ddE_eTE5g9;%D7CmRS8Og?uyPrc|HveWr*I3!!7Ar`zl|Cs7;4(xn^D`G7*$3R# z*|Uegpumk0=QTip*uhQ-s z2%N<|r2a`*Lo;s1hoxL~u$H}oGosFy$U|1|q7CRW)RANOC6T%jykNnXMIPq(;ZEYY zAk5+ep#HZ`=H?<16<>`sg^B~<%vHaafHeOjJ^@-PP_TBr`ERBbRU`x7ORU`Qam*+o z6Z&Ldm89Tusro%gTCCgFxW7}@`!cF$_ud(eEcW)#c#er&ZfD^h79`6;blaVaqJ-l@ z`ua-48{>FMq_RNeN;@#0*hC;ilze zKxuTAJ?S3y{biRc%segg;i%;Vt<{{iN2EC<6E-}4 zw2(~s$B^tm`AS(N)|?rYACg`B8w<&<6^CTkaznB`>DuiwIRwK=aiCTg4~fdO`zS2n z#RoxIk!p>Wt2^?tyU)aNH`c_&XkMV<*aLZ(|l>~b}%r`+(4 zmBWzJ>avEOPOA3`pID@KyyeRbf%`4lNU>^A9{ysp=?p)X%u{aCoM&E4^%ulFe zvpXmR^TdM9YTw}^OhHr)u#0kf-Rk#7>`tY!k-srMaB;!)-S4LAfyTPthsn+~{%ThhzYz2KSK#_@($!0d_IpLKjn4C8O5u^RTaA=( ztUxCMh^kZ)8)4hLwSwc-tAZ{gu8+2jJax1!h5v`JC@iBb;Ie>x6HseJ>yIXCJx8PX zYI&}Fx1{>tW3ygbTuiT99a7=!=z_hugxu5k!MVqKB2eT9j7N zp!cQwzk+S}c`mmKrN+^(-$$D|l&2@^jlw(LD1arC9Mi0 zYt(P?VUJj#64mP7#jcbU)S^^=!IgHBdUZM3+!u(wSsluMq%*+r(tRaLriJKhQKUDn zk*%Z}V7h$1>8H{$991lXfRlEp~&b{EvF^CGc(cg2Btp zHXlW#D}q@XlI28^ID@cr5zFR{G^w3pG>$p_av}}s`??f{n~S(>PjxfZW8+Uy5R)xW zd@lEB8SXXb{Z94j+3}+D-mO3%_^?u&r4Z^hgqOD!XScp*8|3y5u_EnQk~SmiXHzA7 zxdrrwQRp(R<(mgwomi}vb_~t89lC~3_}mH^90aDz7;G&>J1(#X=;IT0f9`EFj>Oi5 zK42HOo;arqqc(PSw|35L?~SISx|{g@1-kBTZ5WCzf91^^b0R(USa%?&$Hqg8@Li=c zVg|@(8xohfR#8&}zSW;5>J7!)4@Z>Go_i+ze@gbg+IRfC2_G%(X*c5&>A|ki+jZpg zINhB-W6s6J^@-}@xMy`W=}>a)w)U&1#DPs~?!xU2=d7KKlQT97IL3~dw*pNnGvr)e zldH@Nm`xAWI47&Nkrj@mWk0WFMh5FYSH*g7hJ79HGWHC4l8UuVRI7r^{e=69H%&TT zXZWEfAXIGCi5F#D+I9Rjdry<8eqBx7THm8pb@R@z9l;QIsaVXU_=Gid*2;9zDgBYvp96IExz*g<`)8_QIv+&uOLA#1ItqEI-E+W+t&QDo8oP(^-+;`TvIhY2 zdrM2we02YX5Eu0?ke%4%?7c#{C3_m)+?192G`LEdZaG-SaYGu20el712BH2y)zBC6 z)x1>j*?Waw4Wvc*1%)3U@cbDUuIo2gHW!2(M3~$WJ$sM*32TS?`vxx-FZ*KGv(VgB zHeelhlLvy*73Dbe8Oakb91PaDzsC+AR~jyE_`Kt6OU^M8DsFI!!x%9=dyk|w3>Uh?Q1eIEJcjTt943D5qDK-FuXK>`m0Ssdy=|q-SC~p!wAft*seJ zHP1VE0pajVVq#{8?%6qio(OJ>zY@xu2&Kxu*8}32s-SAUxk)L=Kx|*$*uMfUh{p6} z5+BR-uttIc?}iBT`<`-6*oBckvz2Pvqg&yNQ=(Xhpf0=Dz!9welJO`C8c}M)A_Ry7!cBLmgE0;8MzDwwn;kwh zbAxwX8_Qb0_D2uvA1M+a1ldT`k|Ms(hvpr;;tg5JwxZzuCEUwkpmpSwx{@ z;Yk*c_rR$RQ{R;QlP>y~DQsAL(b+**nZT_m#FzTb zx7=~keWlQt6o6rj!>1tx8P^(zT)}?h8VlRycJ`7Yyo%K7Z+*jzLFuvZrMIQBNpSZ)w%fn zpt-ytwrP>9@oYOFWZPFW*w!sup#88_xw>_H&f!}w1xGU}7_DBGkyy*2;Kvj7CN#7? z)KV0tcutsl&?9Pfb{ky}iY!AVcQj;U791ILCg5##$fBP2!KQVpNW-j*pRb}H2LS$g zUn5a@Fjb8Sk|%Ngb4q}!Wy>c>2*i+7Ep^{bpZWV zDL1{)S{TaG?`tm`Gn_VQHfH$eNH(U5J_xYk&pJ%PL1&JbG4nPG$;G;=q@0f`HJ&x< zMTl`P2(F2_cg-vZ^lIba;T|i>1z<$O@Kb$HA8r zbSC4mmY7SyK@@Vy9>WU4IN2)zToD+BAf}z=Y8@qfOOHpbV%JMp>>9uf@S0d}3xgjy zd-m{`xtK3R=*auJKKxxDWbwRVqYLWsgvf>#a(8G~>{EU3yWbTCT;G-8-mu&mw|ub` zKGM6@t60NoNcBK~oY7n{;ou>A#gDWzS-5x`BqUH;9IjYt{g{YJLq2{ zu2aQ5hX?d82MbAoXI>&oQwUx1V;HA5FZ!K}bVJ0h6<}i*eS35utEF2+T4s&&4=L;o zRttMCu8=!LWWk9aRhF$U>lKq6R2onP%;KZC(%|=}sM(eAf;|qA=%jYrOI%aNN&x1N z@lGj~rc$86E+O>SZt~SUx6iNpN zLm5MWcbc<$yIPyQGUKeaMAU&qF?;D*;lwyP$Z4%5s_Ak4!jyn&j*e?Dso=ck@I`uH zag|`tWGze4L&=&59UNw#AAwSYM>dCX>_+T?u4ICOiY6QjBvOSD?t(|Vanez2a{`IP z@r#6)YFX@VXu#e`y>*CU{Y+Xd_fs$#EZTvN$PpF=7daEHk z4Te~gloyWIKrgy~;o|ErZEo$rj5YFr2h~TX zd{8PD>bR-lG~ngxH1k4<&71LrJqH3aZW+!0X9_rwp~{1h!NFjeUxN>ru|EhkXRlxa960tIv$q~3Ci{xdF*V?kXK%|1 zCnPOq3Ab|g|NX-K50Ebf*A&q;!-x_1l|vGwXR^5l(ZxQxo9n^lA89T9j2hP^1(qsX z#k$B$Z>zF@+o@;1l)M0KGNz@Q+0P(%^igLd?cVnJM^@kQ#P+2tt3wjn4G2Nh*#|H2 zKw9YQdMX_n#cqYF@`VcDsLEGAw=gUdb^&=suW;ZKyBgB&Ay38vQa3gZl3kD|&j)K+ z5P|YaU%;+@TU%WXYD?pa-$k4a{T?+~DemKx&CB)UN%1n^7Lrkao+=EUrb2zdg7uZa ziviU7WpS)iYCdx~X>e_fC76qChz{Y63vN*PPpc#Ed;iSu#?&gC>d+ zg(SlY<-rg*At8a?(76BDxEYq-ux&q^Ve_^MojyctWV4BQ(D0_7Pl~EJsf9OfK5v-GE>=#rVgpyD+G9##Ad6 zOWQbt`s)NQ7h3rY5TzKF2C(Z{Jax?scJ>^ik{{LpNA9SOx+OsCGvjVbg7WY!d06Hu(hr1{=GQSm5+gJX7q z1T`X`UlCBO%xAq@BZ@*LJ$sQ~WEGTOty9A$u6zR>bHfW16l*bGjZcq3u@-+dke1BD z6bq;8Q?I{H?RsfPgo0u%-R6%6qtUrvAg*=;0?|WP*AK4t)c%edVpHmP$*VzIoH_>4RQl% z=l25-8VAgzP+%!2R*uY)XT9a>H9jGb>BRs(QTlj~U5Fn@ez}pq4nsgN03&i)@4$OO zGtMVM4xkiAlH5sbDI#Tpb@lU~X_-o2M@OnJ3#g}Hb6dcBYCG+d9S=3sFpio1ACR$rWZcNgT< z$N1@@+`Ig{fz)KVclqasu_@+Z(M0c)KB+kOE+!-+_b&csH#|jDz?AXHCb@Tay|jO$ zryz^s=I&>`XMF4(w>iR#apuLjcbDGqVV}(~S?=9i(tY4Z{yY9cPaZAgnLgUxNfV)o zbMGFh!NWhByJ;Y%(Hsaj7T_x-h%;cgyy!AT3 zmla5J?>%X4{Cg~AAGy{2f`1da_kJspkDp<*b8_#25?NI!9H6nJJ|L_yc*}O@qx7?o z#-#Jh#(`BYTAvRCs}8;Nrk8+gdKLa?H0x9F$2D}EvwuiC;#k9MB>!PE8cn9#)m#)J z>DvSbQOSPdQEj<>SS_9H7Vd^E-4dnzpRaB{{#;1;cQ7pz_{(AVb4q;>Z*+~wGqI}G zi!h0nXEyNnz!d&W;E!%BkTV-n_815cy_?x{&H|{--o*=h+bCvU^vh8?SO-w_kA0#w z8D~4{^*Afza&fHIT!xW_n%$?=hlCYT!xmxMMK8~1^|QHJD>k%0h8^@I>7%$=4bw1IHJs-2a0EZA6T7_T#E|p0MUcq+6a#wc0xr z>K8b+Ohn>vFW<1fxqt4_{y*NjhN5~#KQ-Jz=dsZLeB;QP6pB|As@dMLOOs*`H4gk}dl8 zz%tppW}u~nTmK>ywn1qV?fkUVtaLkN{- zH9!D=K%Kpq?W-qbq1vqh%dQooveac7*FEDxE7SGChgRrUr*@^T4u9M$?4bH5VGYf= z86TE%wGq~G?JnAYE<+tThF=n?8^H?}d|BjSjvwwMKI$n4!Yo&MCpW16w@&8fA`ul| zjWmUd1K`Y6zjq;jiTZB!kN5;=sX)Qn_2$2sR#cG;d>?gUA~$#H_c&&hkO_UVuS!yI zxm5k$m9$vl)jMUqFQa;P@0~`^F_FvdEZoC_WLb!AyK_;La9l`VUsMT%(JO`HC6USk zskC-r(BBjWT}FzA#VshbDEJ%v8m1rS$}7!9sD8pGoRDmE!G{B|XbjJc{ElyA3-m*L zzO2Roh-Vf9jaAKg5TaNyBpU|BkQ+%$&DVwb`%#M=nCW7PuiU7vJ3BzlbWsertnb>} zKMNjchN0Jf>zPaE^skhjw~%E|x`%y#+2sl|Ps=(u>H;mx4alx7HXsY}!K@hHnhnT` zj8cFHmAyJujAB6?H_#h}N zQmxT)bw^%y_n9~@po&r9)&qtNF&9*R)c@#0RSar2MApypS~mN}lPL)sMryzje#nbL zpWnQ)eQCpesthoMg>$3tHdyB;)Unwe6xMlS!9+>B7b9YO(%~XZK~!ba>-MFK0=&;xQ?=2W$nj0Pb^nD+k8hrjtKb)NQg9r( zMmz?6EXB7|?>m^pqs%`mV~lFi(wi^n3|Ld!NJ5N_9?z~!7u`8JxTel$_*IU2=R%NMjDd7RW==c?fSw+CAX4KRbLaCWq` zXB?W)5*_TeW%uekX{gGG%CKran@#fu8q>PhC8DyVb{@(iQz8Foo$%OabJJSFCK|^c z3h0l|GD}0P-ctX6Wkq_%)t{FdQOl7=jq>W5beO^u>`=REvFi%BmW1u|R_yF**DKGw zL;9G2We;9uubvbOxp5_mAR06hJ*@22qO^(zy)TWu8f?SQbG52MsozJNI+Uj;>W#uX z-YA4O0iCTxN!51;tQOazq|bp{gYv6dnVTvUY?3w>{lZzLhlE#t!L3iADp9UZ(RuC} z`yBoJhMG}HxL1#ol)m~cKI~D=jlqY;Im9!3O2Y^e3p}S`N2NA+NF$R2U9mQ8EC3fq{{CNJC}MPzL>4VlBB7x z$vuH#5vMXGRYMeT+m*Twrkh%nzu8wt;DW+2}NveWs zwNxj?z*kd@UqU|WmZU+*b94*b=zC=EO=o`K_Wq`a(&J4);GEi(*P9IHNp)g&EF**I zFE0)}X?HT!VmFA&U%`tnwT%R|<3bp+(vU1Cio_X&or_pDZ=^|Vys4Q6;X!Cld(|%| z(vZHdOJUrph`aVwH)8=yEEL3K%M+i=J(gB`DI9L?m3xhuxKq7)cD%?;bbF!g#MRg6 zk}KimZN=Gr)w2!l_8Ok%m9}O_{cNg)FSmf+FbZA9wS4n{s}pM;+DW6pK7c|ApIae= z1Kg%E23rf!jtkr|RUe-&{*AM6*ljD=F;d^WU%0)m92WSVn zA8w-FU!d#m)`p?j@>kxxF(=YfkKuKGc4rHE-+>iiIv(5JzPJIs%X@zRJG2PjRVpK9 zfPA(gahYos^~F)E{|)N;-cMBnYQJ>uBL9S@6oEddFR)T zUp=a+EQfgi*`0TyHuLja0{DQ)dhj{+=T;&XJ*SoMx&)zE|qQ7tOV)3#s zc0CIZwmbydRGJG%4%SM_-`gxn=&FvvVung9_&Qu!+Q|= z&f<;Dy^VdSrMPfujB1D{ju9)Zyl{*g-?)C*#Eqv{m!*%riTXAbFC~`ro_bv(e6_;? z9!c`-0TGdrkKiLVfx#WZ8&FRl#^INU@5IE+4&AeJ{yY)f7Jns__Z$zU%D>k`Et`ss zYQ4EhDTt14U*0gu1t#&atjJ-F1O?s=flPU+-hX89!_=X70@&#o%k;2Dg1Ycdz_AdB z9~1as#J`Rel+6ZK^A*1pSlxh$CV0UTMN(kXiC2g$;`@Iy(p}6Sy_VF9tafy->&xI=VO?q3h~dHq6f98F;x|Y4Xjt+;P%< zrO+`Oz%a()(-4A;YmGy$;J<(pWbp#PF~f~C3_HIfiao87R@1wlMy9f5slJ-Qwr<%1?T4+()ve=m4&QPqIGRbpX!Wv;#99soKc1*Jp;5h| zmZC7lbHY@49#N~a+vqA!WEm>CqtVV`mXT{&`Q#3CCg5##$fBP25k{lDv`!Ui=&|rf zq8|qU{&`;`QF$;`jR}$`abVKmLv08O5&|(KRZAUr6>!4cNz1y{Z3)lSR59Jn9I2m9 zi88u;Egzdv7bBH}Vn{i+n2pGr((Ny)CT6@qF#pmIYE0 z^Mwcwi*y>`lENtKO|%#Tr&ass{qpXf#*h;%$(SKxuKfmQk&6(pmf?xg19=&oQwUx1V~ucB^g9Wz5ny8%eS35utEF2+T4s&&4=L;oRttMCu8=!LWWk9a zRhA7m3>I`(X3~IJHj0adrKG{{Pf@cg12!6M7oF5@dvpGpGFC!@IYgyEgIz-CuifOU z&!*l`0|mpP7GDT|giu>0rfCNMT%Qi}{2=a3^(_jnHTal|B@;}x!V#l_-}^{^8FDrh zBrUXpzA$`DYOoi0$n{})UNJpnApU9Wi`8;Hz30!Qo~*;C9TXrO=`^6xl~;)1Q;r3s zo=UwwhwbXO@{qffibSRpx#&KSaD6fN=z4 zme_h>1?LUR`~NE^O#tGkIx<^L1|a+$a6t$K<6>4gVfdd)g%>I}+BC%Q`6*E-hKnf( zGh{GjUZo9)-*1goB|r0{5E@C0jh*`X5F+nnaIG1UTu?STNenFA`p=WwE=V0ed6$)**`Z zGihywq48-7V~N}kgJQ$FAvn|QwUES|_zBe$PH$C$Sd)|&j@Ljhx_{x~>o09??QHMv zTk6to$j2t?UfIiRRJv<`N%PSuAC!uPI&Nw>4S2aa&Ad=z^JYA0bz-Qo-Vsd;fSUOM zh5O1cE{LDef<#eMkFN47u3oXKpFG(AOaTWnYj_YcI2bJRYw!Wn>M({q^aSYa6-r+U1X-WRoTDo)H7d7IzXE|FGh}fJlYxLj=p_$ zb00cyH7eZd-uC%NR^Rc&_N6PULlW8z2t`${eMQu>`UXrIfpNMboRGwBh02aXg>O{l ztKhlymbSFIKrKWtybYOHY9%PZQGnEqje}(Q=ame_qVy_!o5(9BL|*9&*wt@qE5EZN z@(wkGb`fVohf4|}TZl5?7Lrkao+=EUrb2zdg7qnp#Y%0y5Un>RvdsZzO!6>z_L-?V z-Bi2#d#AljQJ|a}HSr)^TIe!&0S?>-BxdYE%97#b4!me&Qo1_i1{Whht#0gYJ`U)_ zP(9p)(kLz&)+7(~g6+h{&9L-_ZTr~_o3~YHHg4I3n|SatDP0BgNwN1G&)uNlF`K(> z6evlYF@y?$3$T(Lg37L*&wVL+sY4-#5%X;>DVULwfGGx8^F6aLZ3ehR4rT!wFd;hs zQBAXg2B{3?#446E8H7$zV;Jf^LOU%fnD@#Whk8b<|BJh`99h8u&*Y*M%#7KAEZd>} zm~|HhmCl%IrDACtM^JyApk$$y&j3-1VQCOsi7KEojP`)sfyG3K)oev{X(6<=-2@bC zwDM!y1zx`M7UFA}OV6R*Rl?!d$MiwTE82e1vp$gv?0WbPnPh)$@u{^NqOiIMDAsZj zdft5l(bV|%1QcsAaY@f!(G(ObM`p>h-q~q#Pc6@o&h%mc zpD2C2$1a3QB){CqUnh_?$z{C*A34YVxy<$7t636S4+>e6q%I86Vc1UOk5Obz@?U!| zpx#RdD;4zKTivbZ4IYc*M5F^$B4dLO;DgI%12aW^i`y}Li+)yMehi%J^iH+`GG;z-yQ0-rdi7&-n1!9076lNpH_Yr}u?r z=vwHXww#S&NBmzz4_8;pxs@B5Gc;Q+i!#iKa)YS z+`F@!8?;M%^$mDb;MYrY?+#hP0ng;3wRQ~%=1!P6_uiA%#=pl>_K{oNFZef+d+)ar z`S=+|J16%ZC=t9ZuHV|+-`sda&*t2pEzJ(yfA8PFj=KN`Y^U)8P#+M6rQWjL`6&G? zq_J=NW#hoA7p>2SfmMfIdecjwzWplv(P-AE;E(&%;T-Wp+7ZVZW+V9zo6%@8-LB@M z5J}%AFo;U_6OU?p@D8h`v)#h8=_?Pe+@fyW=vz}Zy8p+$+4I%S$Dcc!Ue_^u0)IIS ze@>|n;*G8mc_u`)dJ!hk^2`SQ9+<+P3H;HG1#)IX${qvZp?5QT&RGDp*}HgQZyUwT zi+(vu2kQWe{;^Lq%T&!zK(y72J*DYA+p$TD!l?C!gcVW47Gc^&FVCShq_tv0`-6zr zW;<53bK{I?S5CxjvrtdBVcbfgYJYg5ZkHs0jg*1oi&XCa5e8_T{kZC#C#-o9>9s2U z3mjV}67_H|->|;9f9}!#Ki;~AqIyO@HQYhxvC#i~f^y5^CrlT!bR<7#0G_Dv)h zGE4lgH}_$Oa}0MW{z_i@T5MHl>%P}xlXsYrpk?af85g=Jp`>iQTAj5VS>r$PUoo?eM5FSVT727 z1>KVU`NSaEqB8+3lf7#OT1vR}FG67(ls3`MPfN{8xAVllhqqOCK#Y_>CVZKnx$ww7 z;I_`5J^Td)Zj3mu0RrR~hS}5;vNox$0n5%sRE}?l2swF$NZg5Crg7afk8WPv9xVm^ zvefR<)!~nO5O|mApM*6u<7RwV%GE|#%eA{`1G)@#mf7oVQ@c~f(TPJgKk%)?~Mw&v!0dVH3-@A~%)JgkCd;+vopkVEK^WRJ>sz?UD zmsq*qtCRRL}0cGaA|0y2~dW!gEaIaytvp4xWLh zLzac;wmTO^3CD%>^<_N*>d4@DNu;ts zhP%4@37c?2ve5+}4#1)@JTvk;zLBlb5BWg@5YH?I8mlS_)sF8#QQcl3S|bm+*;HQ_ z_6v?nIWW`364y8VnM>!+f*+|^=xMo7pfozmo^%iU{<6yzW}cRHaMWTtj09C~Kz41h z0a=I-X2tl{Y(Q2d7cVW87m^()XTyU-^t($d7#{nvTa_P@UHcmg$*vWLWY=;-vOVeA z?TV8LtRze4|!4O^P5+;FKxIlwgHB)aBlS7 z2J8HUIySq5!a7eZ*sS&)F2WQBn?b@53qwU&$eO}cssA;l|- z%{+T?qj@VG6Bg-fq=aJynill7r;^wR+vcqm9H~xkts$j_^E%p=>OnwQ6pq>`;PPML z6#ZzT)^jw9ua+<9ckno=zs^;`dv6T3saUrXIAv!bH1?D&$w9B7XIq zbh}LS-LPstn@#fu8pEFdCTdhrI}c@%sgQrn&7gWdo13f}TbtN|C%QjA%PbAGdQ1KP zl@;k3SASk=L@m*+`s1HThbcV4M)mUrTuZ|Cc`J5yRjX2i8u};2LT+4%B8Uc!L=P)_ zwJ5EkLGMdruLj%j^IUEfiuXjM{AxTJpz8P0rVi!liF%{(jyDS7O+aTWQBs+-fYp61 zipN&?4_kJ^I1=$3@X=BnRKKc~Ib^{5ym<-B#hI&Pxr5T7S}oN^F!0qB28^jByUTLO17)nYe@%3t1# zFEyhn3LZycB6-*hiDhX>mJ>zd48qPuESopdBsSjEOcVt%r(aH_A$?z$!f5vvH?-_3U_&+31$1?MBd5r%SGcm$wyXw}fXK+MeKs zdZv z3OMVJZEs)PfR5xnzyBS2Qtm315i>wO+mN`-wI?_^ar#da^@if@ha<{o&pi|VKP6l0 z?K^(ngpZc?wAb{B^kCQM?K<*#obC>9kF6)p>4CQ!JG)yu=eGC8>sWG*;4%#XYOYgD z#}5G(!$dtL4s2R;7jADjXYFL1oUu{BG3^Mak6QA?>I&S~dseI4&I_6&KFinUBstAfk@g!_s&O*#T$ z_@Q6|Ld8~{cu~fsUB}-X>V92K-CEzHRdw^uD+wNx8)?cF1YRl@GYJV{4V|?zU35yn zr)$iM)}}S1ngVLUrdB8~Lg0h#eF4@e*$d31jR zIeV{=G#}l6A;d-f3smw;k+b&-iA(l0ycvB}6|T0Yp` z_Ff^S=B0wq-YfiSAT7c#DExSc=ih+ej62v%nx4H^Mnr$#;Kkx)U+j?lA8dIDbd*SN z2F#^ZF9&+|9?26g9PFMH!OoFBg?bv|YGk-0I(srZd|YV|yFsIVK#kBQX$?ccI7%G( zb7J-$`S00_)V@_3undm?Cl4hS2L36P?54>K;=h?(ZOVuw$qF6~c(4d90G@*KsFM! zq=@hHp?L?dctcjIZbh+nQe7f&Ign)^yzr#OB#2csQlLOku;~i_Q+h$^>o= z@{2(ZL@yNbgN`ol)!x;!m+=g|+||qDt02DAZ@%S@lkO{pj@bZ)F%F-G5M*3y9C8K! z1(YC*7XXe~8^k;vd&a3RX6)URPUfCF`R;)|IlJGNn*jJSU{)J>%a6s%`hBnQ^gy4H z9O>ym@@eRG5OZzVs+?;em>jNsB;&Dkc60a9*&+SE@Ba<+k0b1@Nyl!Gh{d7u{(PeWHj7M*u9%`dzfmG&AwO7C;lRBu8lg9SAy9qz~ zOCK+XE0NW_p>Z2^dQzv;jF-@Bu7Uvitx|4!p*19%QX|$lSO%uq%cBfO&2xd%^nGzk zI1-qIgU%c=W9DrXvXRj7Kw6SjSb7CO$8moM<6aP46LYT-*+~Kw^lIba;T|j`yq?={J2L}@#T+%(aFHav>y%~$2>`evFbY9T8@E4* zkF8y;ql9ni@u*eodP!?X7`p~A1H2~I+rl~=&)05k&7M8{WiI9m5jyg|-W?UK1XiS_ z6NuI036Tvg_JDEI9EwWKn}k1FC>od=ximsiu@heTtf0 z886r)dx=hJx4pzQWvql`0v4454R#5kzjl+aKAU<&4HOKET6`h=5khU1n5G&0bA39f zT$MQ!-=g4JgO4!?LO4ca@cRHah<}cta!wv}1$|-onABh|ChWy4riToKCsBhB4`R(Z zyq`%uSqEM?C_p&UX|NSnUhFKYBwJnNAoW!0^*L-;zm(#3du7I1%>&Khqh^pOW-mQ^SYsR= zBz9j*RQeG5-W@BcFH8xj=1B2+Nd@OMhcD6ti>riH+*{0NjH1k1`{ zTszWPRe*{n91A2;g%R#z!Du(XNO-B1#qNd%?2Xi0hbY$1q&10-Ap_Y^pQbRD$o()V zHmne}H@`xTc7%8Ago2FSFq^ zZ3@ye*<6F@Vjtbj_2BZ4w3dDrfD2fvY!&MwGrg_K{%xn8`BKsW+T?jLa@5W2XOKJk z|EzBAL+7nVg1;h-aSIjWKus(AA% zdEvSBmbSFI!)Tk38egSys1`u##-{DEyyXj)eXy1V5h$jW%P;LoP#vC0f`xVkg{ZWxdS8`*+fyIkYrd<+#grp zyGqjwwi6pS!_phJ?PoJ=-d3Txcg?^RkjfK#+Zzzwer#v=A3UJrHkyS-JoS81?0v^` zH^@9>b9cNuCT%E7>(GDU1y=->T|J-sQuKy^2myJRZ#POGLfbype9tVLG6M}ri;O^s z!Lj6@e>NBm4N@7(*@-PkO=L0%oubAt)O&<>TEZrgkQ9PF3JsW#AMyW-yR#fw!2!?Y zVpEz}1^HvvT^Lk4W2%*krEMHR{dIzu3$1(xh*AtogV@SY_-renGmQ3t+=0bJh}BwF zL=#Y~(aMi)7kK%lht4k6%*Tgg`k>?$ZI||}PZZ5?J^Th_*7)}X6l=K%z5d;W5GVKT z2`JWL;*y@d$S;y=Nf+}P4fes62E45Bo zl27DO1tt^}Ybl3%U;crnmJJ|anwv_2i)K9i`cpzdu`*m<)Z^VMW4b4}%!4POSQ%PF z|6TL|wS=HhtlV<0vlm{Z3wgBLy{@2-_Ov$YPIA$U%q|s{A7ygYn22j)kaVUO1NcP6 z$b0NUs6_J1jr?^2S(9AWJMdmeob!pay27OnqmedtHy_`gO=kzB$eQF%V#5%5^%Sxu z`OWZ%W7v@7k5Obz@?U%J)ajS6iv^&84|_cR`-KHo13qJ%m8He=g0v zy9@FJ20mY5?p?lK@1uJ!fV4jcbN%+Vx3^}~sn|{Q}Qj zc~r9KFuk&mBYv~zOrR&t#ZSyd=?_|*r5H3o0l?tGMf7Shr?Q@HCG39&mq<@8_9pzj7F2`b~P7; zNcuK`K~%DzcvM?%A683eyM<@dR~}rsC7Lq-e0B5j=fafvE|N5XzZ`}?r_=}WM%Rcu z6RTRi2$N_>X5$P69v+y&p9%cYO$Q!B;VQgwqwkv8==QreyG|_~KLoouqz9p4AUyPL zX3sebpf-CKFYIljn0e7JN9kZ4K+!+;iMG7vxT_fp^8hu6Piex>cKiaFs1FIP@~}mi zcG1goXbr)6{Sx5sRrrG{FUxBP!H9TmwqsQr>o$s*r=x$PVvBAgvrtdBSck1#yfdIa zJW;nx62L~v!0|;Y_y6EP8xiB2{kZC#C#-o9>6VpcD*g)`TP7lLxR-BO-`qd8}y3c~WKT;QKgblcEcLNvLe5|5T(v z-S*}p2hFli&LzWGby4R*JyHYjbPou8JPm2T(KeLLn@!)1tG=4UQEvJbeevu6)~L4g}1&U*j>Vxy;? zkcDcu9CStcz<%$$-xZFqle$dfx@R8Uytq9sOusDkt5Ux%^&j^ja1($IsecmI(2Se$ zVJTM|VJ+A0!h$XnSR6=x5IE&e;ZjgXj^UR?>PGN_1z#3q;* zzjZP<7m29&YNRPt8~|so`aMAYn#-9Gy81_a0<=`1VC{PI-%Km2NCv)JlEso&zsE77 zgiPp@eN~cz%cbh~AZanRhC%9`vfh_bJ-heLXk@Xse_$*qpkM{hF_FvdEZoC_WLb!A zyK_;La9l_a1TgF&w~yl`k;(#>euRyF5Ah+@T%Y#0>7%1KVm z*M<4}QHvay>0*iN%lXWub7#ThR4nwgoD3+9&ax-n!@j@la)p_vWgQ%~xwz#76YhZ zAz3IM5|wH9QCPr>4}!8H)fz2VcjRSvpNZoFs@M>i?EIGvXB=yj;yP1OU9b-fRhooW2lt}1>Z=G9(s*NXk>H|gpnMEkv> zSm(Vr5fm6BHMEVWjp}HL>iGXRRmTc+B7mq$C9x5<&08yYt7BSN3D33IYovq}{vX1k zu#C2V%dMD_VY_}bQR_Jx#aGK0G&gvh)L(B@FAp$-sc?3*v}YWU)82%<=&aY&%zq~h zRT)tkR?TO#Y2H9%xC2A33jMmwb*q9TuZ|Cc`J5yRd-rh;)`qOpA-wZ zaV3f%8Z;6;tnAgIw2B73FWvtYY{SoUwW>m?-$$D|l&2@^jlw(LD1m0yrlqFkLOTO1Sn9R2);no&u( zSC7IxVF!u-79aMA1u9Xk?p^FkSwSsIx@1Z6omCB1{ zm*(cr)0JSsXWdGape_X)+$uiH$-4YtAAapp^8D$ko0$wWRVh;C_lBKIy%1l_IfmqG za!+7b#Hmb4)er^TcBRhNqa+34R<IjQi`y0?OB$B;`qUVs?kSr&P#2JL0i&!>qq)BYNsTnu%Xf!8W^~;Ggr0?re z7fZ5qPo9VBdg7cO#<{VxyR~y}dv7!q)!oGJFVJ;&Yr{}%`73YU zm=o5b*A2mjo|L;vWyB1S&o(43bLF(DKTXsdinkw*D947%XGpjSAJcDg!yWaB^kCQM z?K<*#obJBBGxn4?uxZU*xV_<=wUcpj#zq0h(i74lMt*i@3kstFHvsPJ$F{dGZUBzJ zo(Ex}D{xn6do%3oc$eI& zAxc4>q+%@-)vDlfKjFUOO=E~jn>_^+5GuCn#EUXadL4h$L>cwC1SZr*t% z!DDnjO}WnFS1e{ye8L(!Yh}9Vlzvawm=~>0YeskiYQd&fC@(|_p`+*EYqzR5eWXUh zI;)qC>yA30F%#a5-XlA^`v5)%&B4%iyjs}nK*P^X<=~a%Mz|gb0Ml@t>Vq1l^TC#M zx;+kJU{{)y_{!JKO)d#^=|4Gpk6g|R{6flY_Z$crrpa`J=2v}m|Ai1I?`t4u?-df4 z>}hC!8hj}-XS3^PSazX4M(dR^wvXhC-iGaE_ybi#PyU`)8uL=YXYUn$HINqJ7ZiSc z!1GTFWHgoyY!*FxuZ)QPzQK#d%f8t4EHpQj+?6z_eM>GwML7Sl}VpA zjsA-VI}!Tu9)!NLcw=*KV;^cME?gSJzT$~v#7ZkK98oB5Tt943D5qDw=m zcqy@@ceQmu^Q*7Bu)DY4jHH_99lU^W_$4thvqSgnoIg(lx5Zxxre&W@J~?S-4KW=uik%j>x7Le-U*m+nWANSSR+AQcqe4B zhQJ4-mU8cdhQk_(<%lbQ)eV?vf)^|iYlOuCS=?)Q1X!Z}lJO`C8c}M)A_Ry7!cBLm z0~|671R)#Sxa_yENXIrJ8;M#{#P|8oypyj$RuKwYQI^#u0+$0J@WBgDvUt1)PIZ|2 zBk(gFfA}(m4T~>2I|wTixHZTx200MDP{ClDjWpxb7c=&5N+)sH7i90Fyv4f*_T=n- zpFq1qwgO)U%xWX=_z}0VexDCPOVw~II3qdI(}CpE(CZ-P+OSo*%t0_|Q5(Lhk7PWS z&Tj5LIy&!ASGSJOM!VAtRJjx!&7@$odRayyBe}J#emqfcLK7fEEyV&- zJSR-8=n=I#yN#|DMYcdCcQi^m%sFyHZwGZ}0^U}KeAx5#quhs^)~VWld|Aj7IspE8 zUn5a@Fjb-lOJk~arbqqflmJ!BmQRonh#{$3>NuW>sbY|ugK>43g~M6a(@B=pPp3o~ zT{aiN#`gBD&?jZT8qM9Ve)0zrf1F4Vn!l%HHxJJZ1}SdlW@?PBWBFJjY0~94(xeiZ+ipo z=*M<;|G_JHtkMHK`ldn0@uv>Py&$n7=AJp)brd9ZOF^$T4j%4N^Duxdd9RNy_Bloq zj+w{Os=96V3gKf(f0OQ{@!v^o#CS>f+`c@0VD*CDajn>Qbkf6s_mY1_GSST?9rxRN zbQ78pHg-23-=0lp2h=C05+&=z5*$z=Bj&|YXZv3&SX$8V@--@)6Dx0U5QSV?si21y z`B*Wa0B}WM6oQyGZhwd{&{4v-^mtSmyY6(@2E2pVHGmo5HL>0n*5P=*c5BN5DT(<) zgpRzgcSl7l0Zah7ert1obK?s<-%4a+SAAL-rdRjgq(qk)Ih%*gsFDUZ2Bub@pb+T}nkF(}`SkA4s^qn0s_V2B=#%#SZ1kq(`&V`HbjK7_?}8C+{dH+OP$FjFWU z91LYVtK1?hcbemQyIPyQGUIsW!R_coF?;D*;fQc_YWLmp1gkGh38?1ixb~6?&T9@| zqz4vP39GoZ{A+V!AvnxFKLVu)k8BR(+L6wkCtcBm7qCaFFv4B%XgAJ9%Qh!~PyVcK zIwTDLyWX@nOeX3V2`|;M*xk^8y^(tB5XJhLv|7>B;n{qe!dN2r!=Tu(ZV1k_4E%7cRd3(&pCA_U^uwQu4j2J~mPJ%3fxp(p>{gnvYKTpj0f> zaZ|%-z{}NX=7kcQH{(fK(`2YK)L8F`CIV_E4sJw{GE}fo{eZ%KraQHD|A20vtH@9J99`BqsZcPX}L( zM05#s_O_gGLegTEaLep2O^Et_;r<86mx61G=$c`~2>eP@Vu)6dp2_AKL>K$$ZmtKH zf26ha*K+6qOO>r+U1X-WRoTDo)H7d7UVt`zM?*9F8RU+B)#~Oxblz&@#nrv-^N*~) zOnD@MM7S@#yA7lU3zFT7(0**sV~Nyimm(SMtJh>n&|*bt`(CkP}~} z^2%O7>c+-Fvi$Q(hGJ2A6~08YUpn%N-3R3S2gl{`y1H&MJFs#nf^pH<)o*Jnzq6zA z-ru3q#1+4bI2(TEm^q@{2-+sv=4P$cpQk8T6Du}eeZYeCmB5RY+I%5eZ!WSn`kfTB zdSST%n<A z-~y~9hoG{n=W}0*UPZDHeTVrrmlX6*2OFw|DF#^cJ+rW92BPEIw#--r@MHpOc=B%oLsT0{R`^Z>Q$gF>-#%el^8Xprk)vvy;6 z+ppk&$9>2Y6e~w&$+OR`mu${;sqsW@%zxHmEX2_c4a>OC$B&Y~vO>&K}k}!*V)}mEw z2}{%;gaImTV%qrTA{W(Kw2q7V^w{=d8J;HJlMI15tO@`XX5^QPue!l27sUWH$(@bt zx|Eq`O3;o>`Q*~vdsoq!pL;J&5xUDY^UI~VcXx5m;)D;1Ro63bLTZg~F3r8m$@9JQ z1rZqK*h&9fntPWM82EgJxp(<`?R@km!tlIwY3^N4t?|=Exp(<@1F6Y!@AA(NV^ho< zxe2QoOWk|^;@rEKkc`~B_?zAE6j1?F#wWGUer=d+-b+fFdw0q3p7BmQ1dDAq#C2k4 zx(&(E;~w|@#kqHv-tb{BN;J41kmcUpU$ECL@f=ScEiR|lL<&lpd-q5UeE4kc3W5zy zZJ2g#M7j5#O5T$%4SU&dH}d7Bxp%uTZ3Z|W3UlonFd;hsk#jH3y;};ygD>)A?9XMn zcl+(1bkrhYq>^{Z0_x>ks{Cg~AAGy{2f`1da z_kJspkDp<*b8_!iIDKHl*zrDB3%UI+s4{`S9ELxq>>)g?mdw9 zwXg;ICXx(6%rV?0?90&9D`(>P5K!+yaXzB&-+lSw=I+*AJCE-Vui$FwIhy8*kM%ON#)kMKPU{ol<%1M8XXnj;dW$WPkIAxQf z3w}wcY^MKIq(I&F<|5a^vaf_)7iCX25GI&>?sT4d(KlR0qfaM{5EHSWTe3f&7$jSC zCV*wKcg;Xc3Ag@5C~Sk$CffOFI4N?LB+kO{r0Q(5F)JSvzRb^Dcw`@NTW8N6{(=HG zMx56G0pn8g-Ki&JZBknUmR&1EWvRAMhBm>_^otVfauYQkX zMhThFC;O@-1(!?J??KW957~FhdS6EM?A|+#o?{}H+gZ4W1u5h5dr#QVz^?vBdS|eCE=* zv*2+m7J6DN9bFooWluVnq`&NPg_);i9UOJRTRk@*ySCVXEW`)1Vti{hAS;rKmlnzk z$qtlT<-cYj*)=aDo1feP=f#V{x)0xj1tHnB;*jiGZb-H#y*4BZ#Y3Vp?LG<%c=16{ zR-{^^S?CvvhTtF4e$)})k*#ky13!?r9VOEXUG>fz>+w$Y6t72KQ%hj-+ za(yIBmdY@a70Qs{>N1P!y}~EfnBO_-o6)};ld{iSpAAXb=}WIWa1*dRoGr{Sda7~l zv&`pFYd%HSlOv!!<2lf~;@eE75-cc(*Y$r7c~R)|n^(3kZMZMC0fw+}ZuH#-&isTr zHoJp@Gfynoto9u)!W2YRHob23dn0nuY>t=grsKsh>glb`Me3o?EsRKc4gAwnwb7c$ zpq@Ief?v!@ai0y65?X9v*-9gkfVno&I`zJTNj%E@vogl07A?K`g3drIPBS(-Xn|M1 zo2myI>v|t1JJa~9T~+)-%&UDmG*a{=`_*sK)k}!>(?zk#6|u_LM$|^tu|#$J|C_2~ z1v(K-&qyV)5w^`+E4ZL2Jnlemu4QO_t&X;(@c$4Ng$=R=Ty9QZQm56ACTcxLqxfq1 zg6!G!pS%A50le+AbS=cv$|3EB`ByPl#;#i(QsL}qY1%k&OJR1Q2_qt_ z5ZBCqCk<5@Q5javXR~SEKx10>xuk)Cn&=cPu}a->nCym}@brtkzC)z24jEeYG_t=QSsu2m?-$$D|l&2@^jlw(LD1O^utzLViE4H4VpqxvYEde`;7U79^yMV5ZXot%btwOl&H%?t_mwD_7NW03 zk>0pQwvv(L?-4t8!)Ke;qkK>Py!4m16)DmumH&IFNlB&hV%epuQhuJU1PeaP)TLm9 zTg7KNS(hK|!>?URA3K=3naMy?l_FJsZ`iri3-QHl-_2=?z9#nshDDsplvE8-z-?FR z;@Xrn2{61^%G`kjrz$01oVhxdJ18Bh)lxO{17A%sehK-gTapGL&(SS#qwkTuH|^ZP zhqw1PtVVQu-@2iYeaL3MV6skfi?|c;~X1dgL)0&z)I!{uumWgUraJipwU-70f#Beo4Ls|$XAXIGCi5F#D+I9TRq3+k!)UEYBT2(ji z{Mr!=ftQNKOo~rfLuai_7oF1Y=^FE*wP`I!ykJu+loz6e(9v`7wOiGjK2js$%&Z<@ zCcGKFM|O7i0elXcgQ4wswXoMgjL1nY#`QS>V210JbMGH!_8z&k7afJX)b2Up#MZ`c zH;vsx_-{aFO@9M9d#{i*AKiZ;#L4>_$k}^^#3g$g+Mfokq^60$)!1}%!?KI}F`BZ2 za%@+UNs64kNB$&n9>)3xeD+@9R|9DgenH{KLp=Wmn$Ea`T@%AHJ$tW=i2lC8i^a>n z*!3(t*z!uyQKC_ICk=F$dxj2NCTH)FJn_Q84KyK&D5D()%em$t9v7g<*?T0dVJH|! zi6ehb%-$paJ!?4FJW&HV;2EOyIsGY>?54>K;=h?(y(S0497B!%iw8Rq`tTlvzO#5^ zb8llGYAG&U8pFQgiDSe{D=!>FRX465HleENRqw2DkG+ZdHWe==mh?2V1Dap$aDYdW zoG2UFW-6b@M{EKaQLI}q6uf|N_$A^yF)_14_w1ZMPXxEcUkT++gi__->!FrS`E|A4 z+@us_Ahs`W>|eRK-38K_#K*EChcyxucsE4Y8R9ADgv)L%WBI{q+QaLH9@a=u7v2dt z76S2O0w0X{*Rg`Kmq0ad@GF4T4RTl`MN(k16J?lkXnb zliswUPwmT1fDRxZvGR^Zxw3xWYs6VQk#9SM;D(4yq#_%M~KJ^g_*$UCV zvzxn*&JOASl^@z}|2Z4V5MvjZ=fnuO%R(`Jg~B~T*w8K8$7jT19JV?azaKQ07sNI# zk~PaV!usQ0i_cE=)eN?E%NA%qY*nsq9iP)`Z#bJx!O=_#vTwPLO=f>-Qq+$p>P=`= zZ>XgxO!1sB^`J-8>g+bU9u!%IO73VR^ei|s=uE)d>X1b}?}JV2RFM{67TTGYT^0cT zd0!(@c`#Lt36dvq{&Pxzs%6V3NC?D`R4sLUJYEM^syJ7a{#ou(IGny+_0uU)MwiV+ zFdn^qdZ>+>1yY$e)!sVhGm+kb#iN(00Ndm4Cj96xeY_m5L{{^L#%bBV{gbykGO}djNd~h(~!6gMfw=Yj0SiPWkT#<8CK}vM^!$A9z|Ab;G`s7riWSv-o z11e<1yjbe&iLBrsejZsaL1!`^DTC>m{ueLCD(hP63z!UK8tWVI7X=Yqz#$&mR6V7xRS(9eH2xj*3+H=Y}yalmQXk@w`TD zKd8rYl!($4LYMp)#_7$Ae&-_H5T``}Hg?gsM+dT6x<#bv8t1eGjbryi3VVaq!rqH3 z-S!gKl(7<$30PDL zG}tAC{@P8x`fTbAHBc}tYVn2eM+mi5Vwz^~&-LjTb)NziQs1KBT7!?dSTez6D;zN@ z_`MI;RwC!*LDzj@_?XmSFDC58E2f7G#6OLFv0ARD_xzdElXdvCg93yjodz_z@?vLE zCAa7z2dSr0ug_t-`mH?VE~O%o>4ZObD!F+DMF}tF9$kJ2=v3ocnTy-2cz(~{93 zUBZYri4F1&dZoy)g7;?ToT!zPCIE3%9ht2r0}%cWxFCcc7qh|%!~c{QUa0W@wL1lY zL%=84*xh^_DgtK*YKaWP6YBF*0;miZQxImzV9LBo8@PhR&h30+Z+inI{@Bj$KWM}O zqldK&L*o$&@^x(N)Ypd)c_)Kw&FJ7xjt*uDrGtZ^j3K}~%?Z3+t<7GUaROT+>Oi8H zz4WYbu04)UC5O}(rUX=TbXRN#_Iuqbw^J?Ae)3Pv!^XmlE%Ah1>MBXCF8#Gik}Vm_mep-JVt1LEa@Uhp zR;#MrRg$`mi4_JD0~m&YiCEahJG*OUVJBF#9uEvVn19ABU+TNM(9^|A(zI&t|l-rYe8bNkNBP*QAD1-Ehn9j^q3{w%Gv*~ z6)23}e+T(ea7+8wbhqmM>WL!CDqX zpuEx-u&ckWuKa@9(zxPx5w|W|w*72|P3tN&Th~}c9_Wig+7sycq}ctA=WbB&n9bev zzU(!xj)wx^0<0v5pt7mwb8m`XMY0fmhxs=5DCnkxSip=W-L{W4-!lt)ZlD4kkrSx< zEUBPDDnmINu?0^PnG8aws4)!nE}@N<@M(5%Z{(9=1$m%OK>a2j&T?c02A;`9o(JZs zL5l3e>{ERq>nsc^oiWu)<N>t^DM?Aj)^HCtUJNR#`G8zYw=eDX~`l?v2YqaH_{P<=kK|JV?_f!$$WNPAB|Y51r(q8(Ghmu>2@f0R@b(ce&wI>!u$Y!;3HeB8){227Rk+0$eQF2!W12b?L__KYtkeI5qcK)`U}Ni>e&L|&u5%*bz<`phsd=T`~SZ!jh4xI!L9?1z|%y9 z8O7ydr`$#5q8xxe1NxZUDnA}^3n-Mny+5;c{^|LZdwX}6ON85Qo|TRKb*ZE1?cH76 z^Ely!!W)uOn*4WH8x)&M-`?E?dGhMKy}Ron1j^%c>D#-zAWvWr^A*0m%hzjT^mK56 zJ&qT@pSatjV|pchdv`y_bJxb{qPKVXcLS-(-rnV(567ldG3=zw#c%IoLNeao#oug( zr!2sU_7|Be#BJ%@yUQ8x8b3U4uR>`QO%Rd0^zGfHH@w)39va*a$ll)FU$72^jn1oZ z6AuB>wkv&m_ec#c{@L8U;x;3FdvB@aBY7^|D9&j03yTAN@!NaC0P9S-hU%Bd{HFR|6KO=-v52i=-a3sH$+xx;#j;lkl}0MW4*qB~>(lT@ z8i*g!fjHJM8_B=hj7F1bb~P2HNZK}mPE@j=cvR1BP}gp@Jt&*avip;lsO#2WG#MXl z(n2Hnhhy;PbT5`;YUynASM^jT_3e10Yeb$^;FSw7iPkXNxma@Pj3YB#Djr#^p}02dlqs`*Zc-tT+R4qWXsux zUz*h)5I&uBTZCyBt(dyjkb%6u9{Bql{4o}wAO$1hwcSpy+E}+y#5|q-Gm0(BklKY> zx(BQFcZ}36k_TWbW#G80+%lb&dy?P9fi@z>Is0+dJ5N~CBGQMd_~&qJnTW*hT)t*w z>)_lI?Y}&>ilTZ(KegOJ=dsX#ee*<5+>Mdbe5rcz`lNB1)_%1SYvFEfziabSlY{U6 z<+284Q8?gRwEn{|r0BA7muR=r-L3|a6Ajgm)zJKSb{gv$NE1#)M@Rn! zP0WnfxKVpzG#aNt8#MWFVpv{Fm92sA zA}YtHg9c0X^o&H1l!i=Z-80P$hwf&HUv%_^QC}9H>;AYKVK4jYn}jtqW-~r44+s`a)W{-v7ZcncIUzRD5M< z3Ka*ynFsyeh5RK>1NA<90<=`1;O$!XUrj3pkqmrKoU4v=Ga55ee~n{C37OD7_6?F0 z94>=??@C&(Hgt+n;w^tc*e__b$bp$ImbeQT%|cJhg#x9~S^lKC*tdsW zt}yeo%!i|v6O@~fUG)>P(*e*va8Rtlfg1#S$M@V&ScGZqpvUzwlCh8bPu-UR~0k&8=SDwZvCi*M zrzR&Utnf>QePP6?Pk-o6 zj~FU+QJ)>F&Dul;_0(|{{9-Ic6Wpi8oHIhw2a8t7*O8BU|KTJaWxlPL9Mz(wH(iho z)Uss8HHV$me>GNbsjTa5n(R#FuQpZj3vsXg0ar)2xsI@|$NMkR)k}!>dquHGlX|7? zhsCY>A+fFK>1JnlI+W*4G*ZH;0-XqY+f$F&2-~K$6`ZY+uOtM#I5%n7I@^}Q|3g?5 z_PQ5v`Aguteq*FIax{vsmM>^-@HnZzE-s1Zi-n=pA(hUKmiC;3XPT*sZm?}{yj2xO zL{20s;#a?thN_IH46Ej|*)*-7F}3IF)L^(Si%g~bV{Qi3^V!^F)wn-bs}KnN@mX%E ztJPa-|65k1XG8sIsT1|<-^+{UE%VeLr_&UkV7uDY9=oo9Yf0EXZ^g#1YE`nEhATti zMDJ(CLPjc4r~o;RY@-SlCCCODqPvy7T9jJZp!cV-SA%W%?_9O2P=@cLP94f;N9v8j zGu|qMHvyfkL`l_m2dr*uQM&#ZuKt9%r*$}vPHe6IWT?z-6-o-WH=rj}9jHSY_ywp> zpp4M86{+XONqFJJK24loQ8NY-uIo|KLbojVOMKWPmZ(Ivx(%@h$_i>x27Upgok8@a zBm!3;_G)z~KbL+1j+bsLQ8GP5TZ6}G9^_*6mZ*>y0|taDIGn$Sjyav1g9z`U!1u* z=2xbq$$@IMR2xCZS6hr3Ll?7xM=D zoOzeJV{$6vIn&<@mxa>dWU9rk6P5o_FTWHebNq)jmP#aXU5)TC>!in_+)*UXAnjbl zvS}qvV)M<_c!y`4IZ>)V8c9?7zAlC7<|6LSs!`6-Xj2**#3Wmu_*`z|86LFuez&@F zvR-8G%?h-^hmCU_sk-D!czE~XY}Q|-Z>Yzl$T|GZSP5@#0li@sx{Pc2?g3XPtD>`d zyvlb5h?CL^CA@Bh3=R_0WegT(r>k*IeeX!UIrp?NBe54e82(!dblu%rHxyfb%bQl_ zguZ;A-PwkYc4#mLQ+{$jzpx1n$NRdgy*Gj6EOuxSzQ0sP%mDdpN8&O!WBY$QQV$hx zKO9kx4VBN3a1}oGw5P+2?@xovMt4`kXEWU$?jGAupVI?xH+Ococh1fC&2j8p)VV%U z4Rvr;Qd81cW21m$wQLdIT<*zL<^{~AhiaU= z25loN9ZSo8I+Pi`EhXLHM+ULpn_)kUcNu$zcai~X*{s$8F1Hiz2i$3TYSZ2hCLj#h zs*^9uaOuPNn?v274yJCk?a>-^@=ga5oRp$viB7x>Sj;~032W$jFVkeF^nH4Wd(qmo zYH3L))D-yC0m=(mLg?r@eDC()O`jMd;mlV(z)W~6dXDVu9su|pGzUZ5@k(K@gBV_u zT#V~;fX+TQg~#20#`Nqxa%nF+3hz?8=YSL2o4b1#pV~U_A^cY$6MMn($)nM#$k}^^ zr1|Ln3n5P4*FetRDszbtg_w+&ZQotS3_p@d^z1$ICkcBn^$qy! zy~3{s(jwx5!jE@&{uLCQIZ>Os^JRMq(X;o;h-l9nyjZ;Ki%rkMBQ=i%oh7mikYjk0 z+viU5mfiI1J@TVoI@pP-(>+JaI*SrM^*E_fY;qXgh zVkSpkzjOXPk=z!4C6xCRPo>Jg*RYs-@92+B`S3zw`Pp%$Aay#ww0Us(!n{G_CWkc= zpP)|&=7ADD90prRbTovu4a@YfMuNKVOu(@ah#wR9V8q{!t&Ked26GEP0yv@3!x|}) z0^gs-z25T=CY2@X?T(DiegfG@ z)RH1T&%5Rwyy6X6sc%Io>_F-Lq`FApav;k-viEF{$BP|MryETD?apD+`G=pRuwn5< zCx>BW0=EYFMJETM7Yg}7Lzk5~7c4Yw4a+$^gOuXRC37o?d3N0mCxsZYfMJZory&H{ zM*tjh1^)$6k_ zptp{uG6Kgs`Dmw$0Co#n^3iD!xBxE-#2}S|qDH+YSlY_R|?`YnCn0e%Pv9-5Nfp zwaju>Xkf<(KidP1#CUQ5@CB3qyy zel(n82vw##6Y#b=s5Ys^gEfSXc?J-eCm(4|R9<6_RTOBnEq%y6l zy}gqg#LA@jG`7bBRrt{!`gl29iL9m-of~&jbZTrmeh2w)oj|`)%1t`8mXsRy_%cXO zqTCudX?Wd%q+y37)pAKE*zju|CgGqnN6eUM9fc&oUqsA(dVjtNC-;*(yYKX?mWOaz z0q8g$4`JL3l51k_T`|uAz1lcF zvD68RW_xJH1DUG!}eKn|8}5ox-{*+wbMEw;mS|8)v` zgVn+|iYw$!5m{jIIb=~&aRaJ=S$q~(n*9C&YIbG3V2eW}veagKQ~sE;UP3Yfi+TbL zb_uCJfASTT;lq-20n5~nr0!511>K@Oz7T#1p|(m)(+vK)XB`Ap>X#_E*5G3c*49)M^o8MLQiHwNv=^_K9x@PrHTK17ca%O})DNewtlg&_6d)Yw zG@#Ly7dwk8&0i=zKXdWiStyoL>cgqqXSZE_SMJ^}r6Q5(L@v4y6eYZpyL7<|Q0w4% zU2w@7(u*W}a9Xk&q)Qm_Cb3T5L9Y~=$x1S<;D5JLY5~T@%s65Ai+bsW znh60cS2&D6>Z4-nW%5;!6SiPDJqqk1lLciCsxl>;sLgbwcuCZx@T1?T& zOrdmeFcgylyxZF9EoyagbB3+fMAU(YV)oGMg|l!RouLc~orn6dF#**anQJSlz+Q9s zA~jfCC3wY!JQ10eAIdYdT2?2A+3QE36rsuHFs?Vtp`>uxR0S1HcmaE)3I^_C!Du(1 zOGK$Xi`@-%*c++44pFS1Noy?(jUS>gmdO1uC^l>uf-`NFf|{^8sD6v;38%LzL99v2 z3&(4q7d_a!@P><9+dK2!gLL%KllR>t^`Pu#W|S_CNkc%*xY;ogTXSth8Qqoe~@aM&MVDNZ@%Un`@9=Y_q$$5nR4WYw2eJ zxPYh1RJFn7oq|UY>NU;j@ z(fp)9n_xaEcE97f8x%Zdb9YvS&q02yOdV z^F6b0$_-Q?Eph@wDwUx@DnmK3iseiOp;OcthI*IKMoU;rufSj2&e@dyZ{p!BM^<3q znOx*wY3$u&I5*T6vd+Sw(iu~&R4#4v1RAdsymTv{E!kc*bNN(R1rVhemIkqvr~*2} zX?MsSSWbjkt!70u0mT}v{N%hK%6CFvF8Mx|;`1@PN(B5A6f5&Z&-+B#3^&3rHMO3n z`ZCudtfF^Y&#%Aq!v ze~TnPAY}_ka{5Ts59&Bx&Y}tep`ch9E-&ix-g8~g;yJgsd$6^$JHI#?gO3NN?=|^$ z=_BXV6r=$uC{~8nFn$+ZKzl+^C{}JcH`ogg(uLgU`@IsigGp_qpjbIFORn|KPVE)8 z!pV@P4_Lw&O^&?z!MESX-GsWi+dg?yij;UQoO4a5kD8F%iS#z0lbpFy}RGy$@8Q0 zDzc1|5Ayh2`t~j-Fo^jI-`?fxF)?}-?hc94rEl+YYHge@dV80DH;|g_?Op!)ZfuIZ zBUh13`j#wxdlwUu@%AqMW-~lx0Zz1zV@xvH9+={{cb7A%YkU<&iLUZ>B6sQAyGw5v zyNezg+z-g!-rZlY=MfvfS7APW-%qNQlD@rrqy`uNZ0=fRzCy;MhX!D97}0IGrIL^2 zOA`n_dTO;qk7yBxyNzVISgm3e$*4iqz+kufoF2DH}B!QXY&N&xA&H`I`KXBWKTE} zH^R4*3Y-IryUyt53rpX&`iJ8nGH|&7zhu&o5>5$0;tXYg}wbbikTPv(ntsE0E+&yPqZfEY(u@~ zu(fWE)tbwIMy=U>T75w1mAfs%w2M}rU28~f#fB7&h}U*Iy=r6KMiKLLe$^Uabz1S&@8Uok5#yZwxayrJtZ5PHJ*`Y&i=q+A=dr-fFq2HrvQ0S+6fvNiC1CCU~;=zUD6Y_|VY zq(IH~rlQcIXWxw1GnZrLNLCkQS2hqPxO{7!h!viG`NusgoDmkRV1$^61HZO zz>N{-J%9jS2c6xir)6zY+Z~o&EktE$$Yj<%(+p8a-Yj*v^j!DHH3aVM@`(B-VGWJh zj1Nn>;$STk=0?0Kx;q>V-bEeIWvC;^@Jk|fBY44rFN-2f^WjF~nnQMmO@_?@9E$$g zWW0RHc$mXJxBp<6%&=IgGdIx8-b4G z`k?+A$BYs(p?&NdBq=yt2L0Xzca}T%XJqA4Mm?MRj=7E@w+r#4L%5EKLT+c_78WGS zLp0l+in4@eA$@#NcOt7-1YQ!UEKs@PgFz2r-7gB0E{39R@g5Xfl>7~T4bu;EV-rqDHoD-$0a!GKXGVUVL{sllhvlwW-swh=!?t!ALR3KV85mut= z7li$SW+?||x>({md^CX3EcCS8`$1`RmOp7O_U&PpE6hACYv8EGbo6q`O~|hL30b&H znR(poeV9lVUf<-H8&;OlHT*YEF}xT9ilSrHVX@Q@j+5nq*}&vH6|}R z`)nQ;P{nd6Lo>KGDjBK#tpD0VRZLoz-ScD3Rk1AD^o1UVymlr2NQmqmxVsRb$Nbq)AiOJ3}NBi=m#v;`5o%i;J2?K>QVDTu0U zdfmQs=inYu*X>y)o?x*HOqGUFPmfg>sW)HxOG~8s^oQ?}a1xI)-&RbHYSGf0F31LYS6he6TyIi`7xq_U^_I%I-loaU zRQ_sH6~7So`X%7{U!<#-5bgJhV!=VTWj5VJ>7r(HtUR`o>P028fRu2mKqrFf8L3BX zgl*H>3eJ|nR$?DsoSXF7I@^}Q|3g?5j@l^T@|VDheq*FIax{vsmM>^-@HnZzUa4Ll zS{+j9>}YAvIaHw~wBT9YI(+pjX{gGG%CKran@!US8pGOu71ra{&z0uu&eT-OKjvmo zJ)h0Z-sr24%@PRx@mX%EtJPa-|65k1XG8sIsT0+@(MR0xXiE0%LxT7J6N1^PKTfA9 zJi&IgtNrxu0-B#hI&PO=U`&A`{pss8&n05p;aD z#rP%UqhU!Jggi&LAdG&!?7rF09el_9U@PvgHwBb$u#c2?sXHd8G9D@Y6)NSHAe>CK z*ma`vm-q5Z3r-EqYVk~uL%E|!oI%>Th-K4Cn#AUtYO#tyR%KH%e>9S&^nG0l)6GTP zt)?U_aD(+73JPMf<%!SbHlE=@>lyA=cTU!eJVUbr?HYocFiWb&YZ4yby*QioJ>O6_ zOFXN`q|j-p-x(|6%`Koe%tDuOE#E!h>cqN-dh9F6NrOTOuUjF5gT!f{+GlsRpRruJ;H5qTH?@xovMt4`kXEWU$?jGAupVI?xH+Ococh1fC z&2c=~#y%`gZ1(Oh++JsA^<a=K#%o?pevj7-M?&9=Wua9ffzP-E+VRfMnH^6wZ4H{}srLuL97L;pPbhNs zULk2dy8l9m(=7|G>}cZ1*?Wb=C3_m)ir%UUS8J-BeL@-U+mu5u4;j9Ld&`;bf}Xuc?j%eF<0x_D&xzT4KTjmL#a{{KO{7xg-)mUhtNmcTIj$6BN9UI|4=!JTsR@usXC6M56*;VtpujsJ z25&Yx48kW!Pu)i>DcTppP17A1p~zv41a;w=fMX#LKPK?Oh`$|M8|%ymvpV|`a9K|@ z!3&lsk^7f^yM zUH~|j!>^yn*t;#A%(XlD?twiyyFZY-0G&WS!O45J@uquZ@&+5pk)BQ@KMcJNVy<;t zm2(XQlf$*|q>!x;?K``*`^4mk{!h8lHv7-nP=*+r#5^ZPz+D!K@uxDx*eu)TXT)Oc zwmKExA2gR2#5OIGRi15!glzli47N4P7HB_gRjzIgpVLQ!<*dr3U@en^>{G5`Qw{~6 z8L6*DvwB@EMd1^VumEK|yTcZao=_{3*V2Wf$bP6-APwgTxMjLC0dK2AzU4(9?6VGW z-x)?pd})XtdK7?vzNwO^JeaCMB8WFGf5d@FgKw)tQ1BoSLsHe$aby06hl$*Mckyov zOqO$5PbXP&--6U{kBKt6d@VPRHH(qTw5s-UPOjsUsWcoUCynj#Kou=*4}H8Gu0&ST ziq6gG^k@*T(lXHr^c$tz^ipf!Xqso!Ne2?u$=Eq)M(O2|l-MwW!+f1iu;JG_Ou|8D zj+imiItn?AbUc5S_Fa&^d3t}o2`BxNJG<}Hh0Ip@6o8K7@esznAh{;yo?UJj{Ygp{ z^lIba;T|>kGm6P3oU-?$74_Q5O~U(;{wCc?6IPHg>A@x4bNiaK!RiG)a}5onGUC## zoG0GX!}BI{bnYYpvsPl9{hat&IJamlstk-lEt#yVC@ss{q( zjOL1CuPA6FL6Hg^8A9<8)ET#HdFoOBldh_?b)j()J^g<;H@-hqB$ zh)Fs!FA=3Jgf96pjMJMI{mw^j9v5UTK0?5JAEh0_VIHx6OT(J6e3VVaq!ZwO4 z=IIc z^(J5aNa_yNQP3^g;|t-J5NfN$G|k|zYwIw&PaX@uM8UNNA9Jx}g2`4mVpQ;ZAFeG! z&Pn~ZH0r)Ed`xPv7r4l^X?b2TJ!ByMYV3dM-E+Cc%rkxl~|U3rP;QYCq< zKAgIJcH7l=)70>uMZ*eP6pTP88I73B$t@c$xNYia4-~O zt9M&ly+y4~ZqBgPnut2^P|O~By>Mb29WPeoc!uTYhmxPpM5-Sf6Hv{OxweuD>@|lk zQiH`+Vm6c+c+x&45;9*8$zk^T5hz7yvN??FZFVRRIBFGCG~rkvkt!Itiv^?Id@d2C z_AGWc)M0O=?m9%VekQH@LP?_3^m0pq-KZa;FqX*uFeo-`7=kk`W2?VK^@P(~4dH1p z#G0hMaJ&Y3(SyAUZ@9R%y))lE=&4IzGrxPJ9+ch8jMAksX$UwcHyg*d?|j#oPfEo? zHJj?J0WVjlnHNfIT8$@Zq!5uwkz>6hnic>x^C=4Vm0w(tKcfYSqNd)y!mqe;xmW$9 z5BeYl9LTKULCD}>u*|O^2287i2@!RwIeP`0;J~rxn7y?iG1*6a&+oLWfrvbNTTVJ5 z=`l;Vm9zg}E8Kqv`BHF95nVHk7=d3oB*F7cHrF7#*k*TgBe;B%*3w_ip$9xwwu&{G znbucj|F%)jd@1P!ZPH#W-MX3m401-FTG={)&RdNNx3WJ!|M<$gpPpa5ywW{Fy9uGF z%C)Vpx~gr!6iz{6m!k@Fp*S~hurOS=9<58OJB+ppsn}CzSZG5%w*gW&HV%^IEnl$g zgS9M(KzXGvU{`-#UHK&|eivaI?p>9wHtZ%?VvohMD#CDIt<@LD3WKK^pg!Qi+9#34 zN^QCjt+y_+4yzfL+=^NK@mM`vRlD1Br@l;4pqwFUq9I(8{HZGs4egCxNLezxw*xO5 znch7VT?JkAb@VFAONJHBopyom#Kz6A^oDKw*$kW3RcL{dFlE7Y_>HDM70f5a?sq(Q zgM!Cw?$))4G;=j@0alVjP}$V;xi>}c=}?HG!+g8a^C7hDW6k%>!k!zbKu2W4AzH_h ze`DY>&&bn#;%Kvp{`V(G)zbQr$@u8FSd06OzT~JW03%Q|KDALgZjh{QFt#`wDhdi&lO`!JAQ++MCJsZCMwJ*E*BqlT~sd00cesNAF0)byPb4h zo~GP}2;aqT@6T+Ve|mo9-rn8ia$;|A+hiI*N#EYx#XXM`UMRdFDW$KW?rI>VZc^m2 z4Rv%hIW#*WZ-kbazP-0p@{xRLj;sGR7oqz{EEf!F z$b8X7y z-61P5@JufD=3Rj%j-TSU_m;Fe@jdorABENJg6`Y94QDHb`ua^;6nD2;5DtlOiJ&F`du7oBVwGhA6LEegf%T9U0b77*nE9gpTn_b zA`-iE`I?QbgL6-`|MJ)>is~8t)N%)%$3p-0%@ea5Bk%Z9_2TtO;~lNpYR%GY*xG*A z<|V*MZakF!m&-nA@-B*D8Lj^?3@N&7+$EYZz1wwx^z7vy$8oi=1=}VP44EbV+gk^) z!zqTl6#rdKN(K|n{YSkQ#rcT7|G=dSTf5u$?>x1?vb`fXVl=P14D>Y{SKHx~qegt& zX6~T+Sz*xBFF#TJKTnT)E(FqqQ_<1Ue?b$|n-Z^sqxR@%bWr!AYd+|sL#S*Gd>^N5 zQgp$O36;(EpNbTy+1^wXTJ-EIV4b=kyRv~W!R2$-dHUramsux_5EHSWS+ccH43aIf z31FG*UenQ1!mS?_3frQziFW?5)U7l-FL5gXmm#{DpV@o-0B~C;FX;XO1#XNuuK@y# z2b1qkJuPdK+U~IIY9T61LngECnJ2bjENk*+sc)6fb$?t#;6tWw64ua|&G@jCD-PDO zm$)YSc!_sNHFOs}=rVR}OFb#X>==GYq;3Q+Sny?0glRt9NZgfSb4=PNHMxqTZ+z+x zhRNI>B%67Y_ylOFK*8I!?!TH=3?do$o*{YT57l4em{CF| zw2ys*Bn5}dpx?Wa=C5R3G_S8@>KR%7&8TN{-!V>hh!cV9m?-3S7H(ldvOGky-Ki)` zXcp4PSAR-UCi5OyKb`p~RIbzmgZ@Qf(j{2D2Za_Te}iAc^ut_vrMU>T+t`E?(&k+7 z;Q%Ze!!skl<2%^`{Se!i<$(tA%wnMRsyP=z6f1^gL#LG`-trfO{enh|9GK~1iHqT; z#nREG(OLeax!AXdU9K?mw5)-nHfNJuK-me|)x{=cAwQUv<6G?$vLd;7&q6ZgH!0bU za%-S|rd6D&ElA0(dMQ~a9$-Td;p0O=N_Mq4CA*rNl5I(6nQOC_;bVi9lm_k)m1(zG zSip-9lCmPzGM=k3dD+=#^SFR2mh+y1=4B7{^yX*%*E|)go%NTNW%vA8b5$%$cDWkX zQqHp9duTyl2h=wUuUN*Jtod%V)?g>Vwa;>&YrXpv9Z!yc@|89P~TXsmVzS2R*TwoVs}Jl~WK^+4Q=7=|r4T zi&QMqJLEldkMyXg$Eu6en=f5SZZR?P8u-tS)n;uXgL>+C=JaALMJLYsT$ytXDF&pg zBOmqt!$~~Kd|NR&szpn0x*!`!*N58sLM5YDf0f1!$dyBzCQEg7LM&!XyBaJ_8-P&a$Kt850O zgi{4xHJF}}dc;Q9HhoLMvTnI7q)>LxLBbY?eUO2H&6X#ddjDdvfdKByYlG0ayi4S|k zl98xZw;}dGSwStzz%RJcP7{4839K84y;>d0&!u00iJ z1H10CRqIhcDZgI&%l8#2(kqq!d#FiCJ>|u+>ur&8LVuX91PeaP)TLm9TgA4VtjiDf z(bO&_t<7>k!PL!62HL6=sq%Zn&ZSX^FXr^su&3x}QAomA`zw{8BZBM49I}#{`fp zcNB>;NIMs?Y+6Z^*nD#}Zsw<<)E|wcDScm;!gO;HcYCX@#sb!Gpdcn&p7>mDQ(FC? z^$d5bJ16T!o}pQRb`4>>2TLK;YX}eTUYyPP^?XA;mI$1~?~Il3<`&Q!W}(ZtmhT>L zb?ObF9{2K{0pg@|0g&*z6*4$TOqVfu5sYCC=zB-%&AF$I8Hv5%!SLTwpzH3|x}n(e zTi&!XCwgI~rZVemIyqooK~Kv4r7~g$$Y(nem$`D<)ZdQOL&e(8R=Z(}P{3yQ|@|neGmEkL{<=>4CSKJGPa`XC~WWZWBt2Kbj?S%UQcbaqt!tz5;Kp3!9CtsA|(ueWa?mbPW`qRPGt+qW{ zgHGP*K!THJmSu@fybM^(KJf`_=z1^HWT*6fdWd_`+O(RMYELcr)B(y1SwiUOIehQ- z;7y+xBH_$eJ-|$OD|(LX>>dF495e?*5#dTc5&b5$b4B{hq#Wg(uvIaeSkNl#S z9SzGVJ%oQT4kx)5fve%CNU3Kba`qm%98a2$?oX^LpAbau0wFr;cTnW)J@T_Wamk*B z`qp3^%;6e51CuK3jaYWUL=-uDuaK|yrh?DjEBtC8Eg~)`{CJ1wU%`5r6SXNMUp6W* zQsnHtG9udZ1}_#b`(o3x@JP)gL1&3-sD1{Q$=Q1(PrP(+1(k?WxoC&MvhU&36vm~` z^z1#7)-V-}qr{OvCuZ-Ff1h>8ND9^(OAV#Llfp?x)3f(zGK2hYCReLcou&^i4S2E> zp%3ptXgiBHxAr#=pq66qqDe`KCyo&-&0c1ZCa>K%W|1bx&v~}OT_;ECm#KUyv81P= z%|rOgeS5q62h~Wz3Pq(=t#AXv;g`h3Opd&M=lpphxh?)mDDNqrN|k@FVKFyfu0J;A zjhf7#QuuKnYB&Z9|ge=w& z_+ZphZl78^*1@FshO-9u_sk2}GI9dZ!x|})0^gs-JzjsRu}1wJ1i_hB`E2?O`O|q0y!FOcxw}0@+B^k|I9OyXGCd;tg4;UMtG7x=7%1Aj>|o_iT^H z>-jS(QrNKgqLah0GJ#uz{GyWs(F=wAprNb3O)*!YrmZ21XAtGCTr#(U zm>t*MaPltyMo+L(0fsRSpN0@*Lu(vz1^)$a*3&r?T8DeafZSyl?F?L&>iti7a%L`(g7Rf5lwnIX;{d5M~ znq>>LAGRu2w}#I-`kza|S|$bAr(DCP9LaxXq`nr->UFggg-<+p*m}?tYGv|Tx*il+ zhI+HnNa!%H$n~sTyg$~R33yu_vZxp9v(Gw%NQ*BA@;+7o|9n#=QF$;`l?jR-Yskjj zpC1#TYT5D$9t2`Ys+u~^If!LFJ!Hv!QZdGOcs7X7^m~GFpa(BUIF>-BYr@AXe?2(Ml6AW3K=mimO7Gjw(2w+z|!pW_mYCnWIR(8 zD{o+kLN0On?WG!2YrqwOQwVa}xc|YkX`5?AZ0Yf+R_=O9Ye$&71~3D>Cf3`+2AJo) z+uJ>ml9(?<>d5TsAwTFNn-bL}?45OMVRF z^rl6>bCGVScL>17F8a0!AO}mgh%{Z}oD`PzQR>$z>0rfCL$U0VkwXH!j}eu;u>4L;^#$pn+FaKxzK_dZ-(hMY6(8~VcV zF{!~`Y}$)gOb;1|zZ&~u%`x5|PF-2MPdg|;IMQiAqbsl3wd$xz&FT6FeIDMwe{dGY z9J5gRJ>1`!i|6#;R_ep4+h@03{VdVFT}nkF(+Pj>R8l!~QNk;^OBcKVwGN)w1(&QL zy-2bLr#>3eC5(8JSSRnGSBgw4c()y<3~Fh1mrX#dsS}fxWCFt90hfeOFfL}s3B&)C zmtGjSt3{_6J~}3V%5X6SX@(4@Ob2PhP~qmM6}SY4ey_uy#Br_#X@?B1^&WYk9&whV zlbJ&4;9w{w1$ei$z+2SnH!j29eZ9MUC~2Xj%Z&%%>>aSAKCp{)`qRikfb*aQcTJ;&^=1&PT%;&WULMC94qa?%M&k6FSks}5;G z)UOrpzk_@!IHriM8Ago2uN;!#c_y1{kX>xEySWiuzDaB8ujbGLo+?|#n#@e=tFnLF zsAs;Ebb{7gFa(MRAUxO%!4o&HY#l)Bt%ikL*`J?(eC6Fw&o5qH>0Y7Tj1Z)qyTdre46Rp;Hpa^gA)73L!KleYL<7r-Ql4PGSjIV?C>>qzEcvNSpJ<_F(?A9o^j1v%zKP%sEic1wJ6 zUXTwc*3*8qYS{pQDSNBv)aut}Zz3LUYYjq&WCGxANp>-p66fau7DJvq3-Fs`^!g}vqyL3UOc;ksH{AiaCNKF$hNBuX zRDU=|yYq{A-$wn_#W}6-H*2s@py>9PgT%M&w$&>PeYa0*NzDs>zoS`n!N&uC^=auX zzyEgGEx-Rg+NfWw$oXByoo55alU>FYyspl;%Wi5tXTk!eTs_QC^BPvAA1AS*2V9nX zY=7Fu_P429Cadz;UURQvWDfuWSa$?Xl4tu-6fMswU^&3yObOk9`ss0oRyHCFsuANOe35_D zLhsv@x0)6c=9ulGSno#-E~mYpDQHK;YEW>AMiIQ+Nq6!4syIAp>`!YC=h%ZVoe1>x zP@S`)g`e3v|MdI{zURB^Rg2)o@A>ZPp7)8W7jz-6shKZMj=RgD?T^*p^W7DB0{u9C zQ6$PEdGUL`oWx`-zYr7QtTzs`jb9q@p6@b5-$3x$E%Nt#_j|k!*@#~Do-hA!AUWZC zzWn?7QGFEyQR4S}F)2Ck`Qnc@^I2qR!;Ek05iRxxQ-z;tQ&WE?eXno{@oxEbb{byI zB~SBd@{lk>iQn^Gio+;h_E6z|LHM5U{)BZdGUCaYzK+!Jm_YoV?-3is1^FBnaHOtz z$=~x^LYoUL`JOL-1uxFW^xNQG{^jra{r5bNP&0|wFA*p42ILpy@A>@(LLVRtb^6~6 z-}Cz)c=q4M|20I98Z94KGuc5sZdMk#I)nwbp36yov)0HfFiVNw^IPibWCAe2n>hY% zSA2fF=eK)_vH>|xyvqj^=~91hDt*{j-!4sLJvwhn2L1Z|SDSfRFW;DU^RkY->iG3g zc>f&y(O}xA;g6gDCG`=lC(Tbha&` z9N&C+c7wWhvu%#qZ2rgn$xGCA>o1y&k2aZ^5&Xk3_;Xr)0PlN=DYV(Gl?yNo*W?^< zh69fejp5G-{^+g;Im02-jSZJ!wljI*Spe4Azp%GIM`s1ZKQ$s+6J?%!x>?3*o&prM zX6|WC^4W${S{%j~s_&3`>28xSbfXpORqco{=XSzs?hF^q^&pJr;E%BceaDdT+wP}V z@$B^EROVFdQZ4DTGR^LJ4(;t&L07C8R@A`tMJoC4OH;efhFmqzlo;A=mZsh^uuSK5 zZWqHkh!^e->uWZ)4$eK%{>x*lQyGz#LsOYH+P}YfLR;w8_PaJOL$ZF<6MJLiIbW(? zygq3>r!{bG<>vR7bh2OVp|Cy(u`o92vhkj1#nj!dw;WOMWqR}wh7^ac*mkqzNMnj2 z%b?$ePGjTUW@6r{Z$P^(;{_hLbYW|E`~ID$_E)xd?5Q&tVB+Vc&r1FJ6V?Co^tg9) zKzW=IDU5!R>74D1m8ns)L3F|s$3m&z zIeb*_7T)JTq;LpuhI;ztAD0>G$8f~2n6hTg)>Oq@7G8Nea!NS%<5HPJ4Ro93m2LiE zFdCV5G@F-hy}N;Fqf!4vdV4>!_xJ%|%T8X<{R4`_7Yj_I7!>%vXHcB9`19Kx;$AJ> zXK`p{7C!UD)`hw88`J1qJgSz~kHEQjgt=_$i>AJC>OZcz99G{fEz>c(@!>O99874U zy;$;;IeZ&+(4nD>{UPs6IlpyzesPmi3J)Qhg^8BMqK40n)LmIMH=X{qPJGCtn9~a8 z(aXxS%zgs7+oNUBS2U9Junz!0F{O9LC#d(2htUe2uRRay>Bcasf%%C`i{pQ-Z`4&h zcMQ<(_E;GvE;wz5eZU2z7Mp;276sX(&APb>8JA1x7Zx@IS2zV%HCaSt%*@(Vp0{W= zJe4IF%~JY6OB)Z4)=#qzYGdl5SOey)ELmY_>Q?XJtYyjL;O8*7FHW)iu{GbTzQQ~N zzj$NPJp_+mb@b42V0+d$WoWw2^QLRPH_i12Yni^ZBe`d_|4^1}KhQ;okEQN~9l~Pa zNbJ|!2C6&N`s8*|0~NAjwF-P_sz_pPA?te;KycId5d_^D{8|KKA)cQ>qhE&ppOJcd zeX)AuvlmCE`t*nHbT2y$nPc@2tCJa2bjR3w3o;m=WX5#4()?Od0Z(kxed=2er(1OL zO~xd)7B0c*qD&#D%(aqW)ZYlO4svDDCfCkZw$NZpquU%qCaZO1NYvG5d9ugw+eduO-$e+9|9=<_K*QHO@eOAM*p|5X7V zXGCY{8Itb?(HEvQ^rlznI*hGS%CBoNlaB@wKSFele2&&;24{@wKjxmJU2))2``^N< zK${dmr_zg@YR?s~Nx!_aU;U{drQwM;D&x=d?j6SpRjQ}IxRn1=y2j7YTe;Fjl06ER z7S_6qsZ<1(SyJC30GbW<5mY-=s&c8Yk-p9a!o%%@%TO7lXQ$@9QR_7M*Gdn5s}%A| zm*kUN!0ooyrF&yM>>cF#Uu+c2-H|$|3kUpOC)SdnvB6FPx!c$Hf#1<}E<-=zRz48o zR^K!pO0j)}t*IYFX$LX%j*!X9v^txM((X1GA1W}2yoY|mm3f-(%jsf?K;4_N zks5U_|3>hn9IxGu?aB2PZLN#=6t?~gs$Fau17$$pAdmmL?^dmM0njw@MZsUcuXqs$ zu>J?2<|X+m7)vnC(LWLtodw@z>RyJ~b-Fzu*hy0&l{7rd1odz8&A>Wbu^D*N;0x77 zaFQBZQ9qXvEOa{MRK6rDl>(N#!q>GgW#i$)N*O&aR?Ogz2&d{L*D`e0HP=D;l`l!9 zQ@OTnDA3JgR>w8J5dQ{cSLJ`0+*a&hg>i6jllT*p2QIu=bTKY7r z$MP!qDQi|*qomIQ0f&Qhbs2~4Mk;cMOE?=nlFmrR|eaJeAd(>T%JIb6yvjpuW zf|79FOF+mfJ<)raHXHEV1cEtJk<^;5L;c=Z9IOQthmN9YY9=kzSo&7IxtopbYj<245NJ%4M7 z&bxc-riv#UH~ahFv^Fzp%@O~pk$Rw5PG_7U^OIxlzfGl)Vz2KmgL{2_WmE%tjk@qU}qajvH=bU&haPb^9!4RXs{1u***G( zuKPWmt<&$&z#+RZLTUx#&Aw^B@SWllAN#;f>aR!Y?qU1LOXboEFdfW~-lmc+<)g#c z|LFd^FCI*;YW{0Fkbh~7GGV@cZW*$eO@s~MbW8J)yG}Y=VFjrGNqop=odi?PNj{Lt zcKc~c)n5!JaJB6l8+IHI(qdq9npDdbT_7-o!}iH=Kx{t9*=Vg=wdABJ*G_hb7K5{g z2BX9GZf_6X_K87q&U_`LM;7kv9ssr;bTLDN^h#mNg&b{@T#c%6&`^E==;1z-bG?XT zMUN~b*Y@(NY^y~LLWi_j&?5^AiStJtEQUOJdj&nRu#miDYsI7Zg{yG6eGOjYs_KaI zNKISCGKyG|F_enjs#i?Hj4UKSkT46=PC<+;Ec|RBF)DB<{CYRsz}(`hu$c)OTM08G z3(Lr8&mW>#z3h|CScA7#H3pql>QCkxt|iTBz<(qFrAfAUsbnvFOD>1J1XayrNk&2N;`bR6)k{pyzSosHA zlxgODW4X7D;i`39DR|+SU)t=-i4QA!(j`HIcVtA)I&5(9ly=0uh%RFX%3z9_NtXnD z;UPhwHxR$3Fwcm;9@|~U2uI?ZMB5#!V$4y*Ou8gC5O&Dqaj>Skb2O{>jWe!GL08P+ zEDD7@fw0sabm6te`132Z`SbuJvbLFJ2HOPLNz@-BKjypM7^1okS*tm8^{t}Uhr+&? zWwlR1iXiJgviEHNpn)A)r<+Lw$$kv=Gb9!+zUt&KEKNc0AV2AZK{aS0zi3F~V)CY_ zzCUlW_y)Pil}qNj5cBQ28&2Nk-e1T*76s_d1S%Xt60F)Loe9hs{3lQVEnfn-)~g_M zd2X4f;E`MYZCkBLAlD$qFMcad0K2xOcm~BGZMUK>#H8Cq-=p?rgfFLcX-3tnUtXdc#_xGVM%)z zAPVqG1U9Yd>9{*Yr_T1aJ&5Oe7UEZnIqIe52Cb`$1<&p69&GLG&M!{JlS9!+>W)n5 zdr|k-AwL1{I8bW6xx4igU>Qsf>Es-7Z^6WT@BaSIJ9j5b?SD_@yqQ)~%-Jx$X1~Yu zjHBL(H1s-g1;RKOywb=yc;(#4FObmr9zjid$bDV&Is!e-WRieOvvqqJyIH+%@-@=S zm|mXl13Gj_0r2__0^8N!@hms8y)dD^p;!IjSfp!>XaZZ6@?uJ$=L=*% zgtYvW`4abft6-yen z4LbY!6bHNY8uT9%D%L^_f3G^pds<7Z2W#X2Jq-yvOIPpFov_uVi8jDGAb@9#?&8F= z)5{Xp;L6hIYCR+;I5>;MX^XLoeyvfP%6{nL_E1nWfKXZXaT99}mvI?&nRV8mCgDR^ zZEWMcQVx|-2G(zMmn~E6aHhz?uK{7eFg^_|HS8ZGhk(W_w%AHCt8F%zI=ZaSkl-kh zkE_8(A#LbS3nQ!r9U>X9P5q1H1*@Z>TeZhbBCaRYW~srX!Qa=0LeXO`EV?R-01FMi z<^tmc)4&L*ncxFHwq1_6nLJNx3f5}Wzm%$|O_?QB)I&Q;T$z0F+D&oEna-3yoV?X` z-*%9V^2HXS5|!f{sgES@r`?A2eY$%UB_;v~Y63qH>P9YN5xnMJ&tA&lNvU=F37v3B z8`N}^`G&Qmt(E9x>$uo=kn-?EciR-hEoyagb2`P~f0#=%$a4YeGb42qD-nNHFDW5U zryv=Jq2#ACBI;j_siB{^sDk7~4qqs+0eqn-LB(YaWtUYrBIvS75vVLCIfC0@VnuH;Z|_`3@$P(@OP^$<~EP zBMh<#58*>k_R}?pS2eRWVQAvjw$ozNj9)heo`)5*Rq?m(p3Ux#IvkPYHHc~<$Rsz7 z-XOyR`R62l6uBW6j3E)49I2H*@k@M3dtylnl&To*Y zxBTjod4#8r^4k2Zozl2SS+A2Lkx~jX4R3vSq=%GDkY(-s4Y3WsJ0$P*MMm`RKaUyKIb4k*(QmzgS`R+ z*IJN;N&~JdU)P#^PV2n#f7~c)fvj|z*Q9d=sYbXtp2ZID6DwN>&@`+8Cs+38=O15r z_tW!>msh%1ZZ~NF>1%)VZQBWdRogWxou$RDM^zMS?3t{JVt~jxQ<++Qvp4a09b&P^ zR%wOt?-H|&S@!a!(>?;tqDYii9RJQ({e#Lf4C-6s%HxHYh?i#IOqCn&TSp#VCGs~O zRw(Mu038Bn)>b4hSaSGq*#R35@o4I>o4MPxvQ~dKR&TDWn z)E{zHWTNpR(d#Npi1!2meme$%32GtJWFgNJRS1kqrq#bAhIO(OhtLE0JeyWlY&Pdv zhjn=3Gbu~;2gI$B=WtMMna|-_9jcS`Srb&hPSki2gf`{3?j6#rjF#dGF(0q=APF-b zQ1}35WzQpIJ~eRABTZRo(3>E}vSRkDesPRPHq38AACSu7*t&6$G=VRjED9FL){AON`=>!zv^?uEHz>; zQ6+Yc7w;;viG6^7yoq9D*q*d~vXqWIN zmJg@yYy6fKNA6a_`gP+NA*@68K`qU9EJuleD#!eVax{cef= z%CyBo%3XF->(>gaoMhMj{`dA;T7KHPN506xvgmqU5@s z?T6+ytK0%Tm`fz4C`n&v?I>y$`D+|gl>F=7)u>7Sh%rUUZxRN~h3uS9Au&bC4TM&m z$H58=U`aJ^!s`&e^zv#kuTW1{OK3{{rP-*buwNKAmNgiEIhWa zA{-@t(U;R}LwebZzWl?13Ic$}>68ULD#V`6Tp?uI^oYCq(UMcJszv%ZXdmh2L{IZim>OR-^ zn~6O~lzEW9==YxueSj>~>3=VL(eHoY*?*e=0B1`<2$rY0`V|Y(14{g&-!Bs;0dhIn zTY3$n;W7g97yXvHN~r*}5+DkS+a;eLFZ%6XqI5ux7w-}RvNTAbOchcB-!3g)cy!*} zLfR2hrmFqbrUUim8`G`>^~kG^Uk|nS&%qy!>U|pixYf*DUf_tvj78geBPHl|qk(an zZB1nfm9|r%lb7%}9^tcNYUyly9n*v9Z&256w#_k{&HuPRd5OAi{Y8`U(WY`@1pjah z{+za^t3E)q<3<$P^w!D+nDNUq9C&jsm84kH_6jF%Z&g6w>0iS38!ruNI zpG6V<)QAY}NQ(ZpPq$?p%WBPK7_F$;eOl9gwqY3vNPUO&1*Y31%mZnq+x4i_R&cmv z$oOsd)2lez{S@&}r)G_U%bbc`swKT=d6}-O?;NRHMJW_3YT)`JmHhV!n2?Bf&W2ny z&y?7-jQXDMoL8Y$9g@2!ph-k!cUoVwv2}3niS}O}Tg8D)qu*K%;gf6VzrT4x8|K#b zyEZRP#y2(|O8?7c_?J;n`Hhi3eW`l!`lRuvmU3Ht;YEcDii5r*}oC0MF3E zs=L)s5O`tOifuznjx?qykQDZ9xBFPmu5Z9WAfg2xxO8D_cl-XGr}kI2ckHPn0h)j) zEIbR3s-^WKAiI%96qem~`}HTP|L5s(&!0eoaVk4Z`cG(bn*Mi9mOi>|4v%Nzs6AM+ zbgn@uG&+}~W+6`CPfK;qcF4-qsM+8&S0W`Wd$;gD2O@<-z#3`>(@wwqbdTZtNMIs_07^k9kUxBg>%ILoX)V$AQ%~`!P}^VJPn2HC<05Qa|Ab71Zr6< zYQEe^-IZel;MBhr$~7J!lJw7-7sl>*JFNcwK)KtaMONlzs0y_Q5UPiLz(qVJ6i~gN zn6xdGD!A?TJgBD|!>9)4mw3FtAyBY{d}xoAVd8?*X4nT@d5cXzJ&QwWl7u%mA>(oh zwL)t&kgJ+3J~swp?JCb(G#j4E5{zakeW0cBoeBIWl66oV<0E1ho|Sqor}|}Sp2pDB zt==PJ%aX^z&ta5qw$##`{1ajb5XnjgF&jl{BR}MO-4ZnshDFXrOAm3%W1@AlvRtpZ zB+9Z`frRZfQ0_<7$EE#@W-SM68=U||%F7kJj!w^9Ja-nNQ^n$M%e^5$d*ykTCT?ku zz-*CbYMYPAEvG3vqr1A;j4mV>^YVkMeMVO%9Pg0{hkc3^%;a=C;@RwZmpS(zD*T6z zlAP|Um(!)QLq}$Kr~{z`{ivY|a=NSKIo;LloNh}!%Xgd53|}p*&r*L*%;b*b-KKE? zKR?Lo$~4UQvBvy`C!);+16o>_+11J1Lw567wA;jRiKeZ2npU7qOz!fCyfVD!$eOEZ zvRZ#5z-&cpiD#MZ892zwK95;hsrO6oU&gVl`CgSa?=e-TeH`pNL}*jccAe)<*LrU% zJFy%oW%|;NWUkII=$8+iDBuV^<|VSvZ(W{W+&tl0a|g9pIGp?S7WMs3wLZC>M14=J zNQo>PRPXl$m~+__yPfp42u2@$ymxdBw;euk98u%-LweBQ+v|(f8{)kE64f25|l|HEbS6d>cnNH+`gJ;o`3RTAEZ^tPd%=+e-)z%JoY){iSeZ zzd2I(Ws%Dl?Q^xmMQsS4Lh7}}HT?>vlP0s_(dRusfp~$E-nwWJK|#KM4dK|TepLX; z8Eqjf-p}XMw1(cW@Nv~!dLl01_5BKJF_Vu5(YL}P{(O$|%&=!|6&2JS{dqpBD-K+0 z|65oUXp;idVyD`BMKH%cMi;j4BX|JnPX#FrPrO~#Y(FBsgmYPE!re4B3PMQ4m-X)c^f`x-y+J9-lP&`&_+ z17T>^Nvdxe52e^X!q(J}p|pb-(&Fy9@OffdrdaN6Fg{dZQ0p@E6CiUYTn*C7>T()Z zEHJ#LY{b;LV4f|gXQ4Du{TsoPa=dmsz9-jPw6!kcQ`lOzWbXSn$m756yHUN%5H9`78Gj@wIt#w*mOliO;?d#NV!KYa2LwB5N~G-F zJ86mvhL`3t(K=kQ8F&-r3)Q8UTbih!%Lo=aopLH)l9fsU%U$7%YhTht{A$Lvzazq_ zdMOZUwyyb=FKLRPHZEDX-|^P=1c-IaWdR<$+`Q6#yq}}!4Sdwh}&ZDQlggEKMhVbe?2>KsmmI8@hRtm1bW{7V-oJ2jA zM0xBOR04*@+*u~lAa9-Qn?$)bt))*^NoV!ghv%ocL?4n4(;6jx7w|PY%vP0gm{s=$ za7}0C@`Z|zJvUB*W%)zagWRL;n%q(5L7Mex7ZO~sb<8R~(tDXUYj|z~NnTf^CT;zG z+|F@X{oYtfZ*>91VH&%TaRr_NS1LTA%$o@aio~#~Cox?yD7}bZeBPfnqtGaVF4Dj=BFf&37s;gYEv3 zWU4QY)EmnQc=yw~(%}A`#~+_xobR5SPhNB;{GZYdy8S(WT8EcC`N>PX`k(+%H@c!3 zUO%&Sac2w2rj1QvP)w!nNPR?}6zzRqxWCTeDhfJ1e58PB4VJ;qHh^gWmH`~oPtNBT zHUZmU-v`u4nXWGUJ)N!7Ui{a0?Z*kRPkihHH~BXNBD1ajdZg|iwvW7AF0BC5!TjiL zD#`wj4rBkL`|rMZFuAJvujxSkW$YU6u0s~HnXn<8ZfPEJ*U3p+0G9ZW%{m#ToRfSY zlY5D#efwVwCvdfG92<5V57J^_a}Lc=Auxo)_RVlWY(B`@Xsue!(PD72L$tV=yV$zf zA)jfCVUV0Np9$%yg*&?k0ImmZ%+MjdQrL4L2je7H*Z4m$pzCA zW~LUBUu9iVd;SpB>Sdp7dME1p%`|iR%Sr{6R{#J>K)R7^Nu`o2Y%8vUhBKh?6_aD@ zN~7y_uE2ZnVlJ%hUuUKklGKK20oONx+)_d>M&&zoEOn+H&*v3Jp=jOFwBh^1yPGl}JEH;&oF^7y%B_iW{^ zfg|+^GSf<|>3Q4+bARQ&z1@9CEj0g?F@TO(0L4)9Z?<%U?Po|>FJiTroXL^b@0>qR zrN+gd358B2VdWoe`OCbUE%&xDT(yoX1uq=)OPdFm0clVJWYfbg2^zd3BP<5+bd)Vy zSjL*~h7%Y*STe&d3HriAf`D%zeocX%5q~{aROI=;rUKw+0;hh=uuF=l;4sMJAotc+ z(-zeG#u?Y8pe<%d`lHFBP|y^;jpXuv>=)9yJ;A6h>{g5lz;P7cG;6!s4ClTH{^hZgdShP2+9 zFc-n5-h4TWZx9i$Tr$@M;*kfr1YCE+Nx#QP!#5MKa0p2@4ktik@Si{hw0sHR8h4>= zS!KDZAGzhBz$3f8G$rb33d|4L415|;yRF#c|1TLe7?QgxQ=+oG>*|r%aeX zOQJX-oM_f-bgyo5)^(fBP`ifNY1d)7 ztFnn$%Qa@>1~Yk;Lj+8CdlQ}@@5(g_-;3_F#ipz3=E>`rVpC)t>ZM3y!iQjLOs&jV zgSHZlf5@s{n8Lmh4I?f-K;*AO&{6epEpd6+TV7t0y4lpGTR$7AkB=$5xNI2(FAFiw zR#jEUpLK@u$bHc9Pa)aj}77D#4VSBb4po(OMQd(H@ZJ2_Di8meVd9OUJDSDQxlXJwq z1ru}7EhOj7w31@7xw&EN!}D@UIqGAaxB_7u3=(s44qiE>{{<2|-y^6=54o>v&7+0M zKdjp;*3Ig5ldq9pvGhmj2A=S>f=QY#DL`JoL14RjMbC00GYT{8OP&RB$x?LA^Z?GN z9~_Hxtr1ONt5RM}3HW@0?0Xd*Kx3W89zE`k$j`)t1U4zLf9^BgLg*+3fdf&K>+N9!O!ZzEx%PtYAEt8NP$?(x`;FOX zW#}dQ5_R?V*1^{1FRahodE2BK%OzWg%i($Iwh>Zm8H?u!~Ut{ z#cIqYro**y@Sya;StL$dj9v6=4ck=qLkUe_Bh*z{1X^hDH5VBtfJGv}u!0Zx zpoT@_PI_(iFQqDKQ)USj_27;Y_h(ncJ<-pt+xBN18tNqwu*p8A4%R%yAA96 zboVMsj06tU1b!gYjRvwPsS-Vtq4?o?_EH8nO-9_JRwp;7lM()hD9LR=(+=YVpflt8tS`oqt<#28`lsqn1(VOijoJ8{9T`+Bt-y4%SR=MwTrsSU%?9%U z&-Swl9>WJ|3`+iD5~xOKy;;nwNIgfV`p0DJLZlG}xr2xB8H_3dZB_iOyJxezqYg(T zc@3gk2r|j-<&(mkV39-AKPT~{$PKw*JOMOM3FG9HD0m?yDzWL!hL}4T3{P5U1nvaA ztOH-v6PUbFtM3`9hlJNQqkLBYXy<#ziG)^c5@uJOSqv82v>sb}c?>fha)5fI6$`-6 z`38?`g->(%H5`PDA(&W7IAF zH0~?Hs9SiAT1{L6NJe$EEJmYl^%+IuyB?lpGLbgDTX{yUl`^WMMW#9gJ4qjPo6o3B zW-zS>jE}mFXVfjFjOu71SRmsc$*7ZG0!E$W8FgzZqdHn-93j|AGV1m(0i$l`8FgDJ zqdHm$7RdNVGV0WqfKjJ-Mx88WR7cA&MqN~|K<_jsZuU7<)0_Y3NWF|$lxNlLX;$&n zEQnVpd|uUb=rpfR@Vq)z%&U$XnWYgNRis`87xbEX^<$)7t@gY^BO1XKolvZf9D)rp z9@5MTF6mCY90yV)xnC)oSV(P^NV^yWt04khcZnssq{^{31q zIlceKSp7YcVEq!&aQ%#C`jl|yYtxjVrfHt0|H>*%i#Wdj#KoQ6CpWh)?d)$pwYC4` zWW0Q6cAdIre*XNq=JWT=Coky!VM_BE8oiF~w|~9GpT9~S zp28Rm^J#LdLYD$qQvbVoa_X$5u&??i`(Xc->XnmM^bhvPXQ?)7$hG9;TgK{#&15EI z5qZ~i*4DAQ$*?zxDnW$ew{}!l_I7=&+<_aGgW|$VQ#vHg}iH zpEjXf&f94V1)sZY1GBv_Zv5m(ExXR$^{-x?A9xn8Ww+&KuC5~!$+E;76$$;Pk$;du)Urecj=0TGBh3VG+WQxGO-8H>zQSqa>0)DkU zs9&jGG5K=hpw{20RQviP=20Ztq^JIf;%00$RNL8r;yd;GsL&g>1{HAaYfnU2*S8S; z@7|rH>6xBo(1`0WSI@=Sj{?A~FOMCQZG@a0eF#kbVCUfSl%TR#>kJ~e?W+G7^_4-3 zD|S_$w&A45@BwYBR3(bNG=Kldk9}IFxDhs~A|9cGV&pjA^Oa~pO>Xa8_YCRr!@)fB z;t6l>{o2Vz{uvZNwlC|O&(3$=`NYA#r{2)};}5fF54SD={z~hGUZ!3$xvud-{>O+Q zLB|MrLBjV!m9_tAexJIa(Kt)NHirZR|1Hch&)J$E1{7^he))%SHFxA}E_vn8*A^yv zJ>0xq<5znDW84e9cg519(7L>+Os(ymeh9Os1!@3Kq)VIU&TsASPe>c0jkx2V$rETO z+M6<@-Cosm88yu|#NHoOpjJU&=e0hO-W4ReM7-Es$**32A>?s7o;k_8QCdThdrg`Y zU0wbXf2ZMHJ0N75BRz@g;J1yWLdBKxR)p%k>biiY;07JgWPd5yZ)#`i+u&B$E#F*n z$Q$xkSi5?$noN$nXOsTn?C+V1q2n9ZLSFBP){NN^eJ**PgnfnoZ$|2BXET8sc7fp1 zdi!lp;kf4zMAuTpggURY{u5_sq1w5-p7Q+UZlAN%zHs$phzHn>s4wO z+J#SE+1)V6x1w}{`fyiy6C zBehB$7>@rQdWtz>Shx*@*G4@!SPk0SAgCJO3MmwBgSJ4F{aYF(gZ~Q>Gz^Gb_>Y(r zrJSYwu5{IAl3Vr~@|_TrGOTK^9Vr6%4QcQR`G(#eSR;^A2V-JW!O7yd532;eMmFnB zV^3q0l!n0nFV#!E1v)T>I=g?DdNKBK{fNGG{{Ee3ro`IQi0CvS-vq(7w-|%az$p*Q z#gs;H@$LDAgD1?r z%DG}6hH8P3!$;J3{m^V;F8(jvcmV{1ht@}jH4aNFp6;`grMK+PAKW{bzUUzsZo74` zckvDLgE!AF9jvcSTZQ)4pW@8y+^6}gdxj6QqiTHip{97?I(0i` zI;&K}A8){$r+xGlU!V0?wZ?YM#?@W3%4mIb2)O>j=Bp3u+~pDdpWVLn4xI)2JyJRj z=f&Bb?Y2GzV#6or^9yZZ$@s=Y3U1o3IQPW-+>;mf08tX8x^u9zb$$n4GFC2JoIk!} zW|6$=8W`vN3`l(azDM4+e)DX@Z2Q#4$)hir9NIYXx|MTJ?Cj3>=bO(y`_v|w$L5<} zdt!C-!PlO=?e@)k;UAmfMw?IVZO_lY4%RzxL+ck%7OK8)q`r#SU^^#pb9d`0tB2t> z*UUC&8#lr>b?`mf_=@I@Wd77U=G)slyYJjz>H9G?@{GSPMt17IRyPn^Yuxze78DaT zC2Gb@!tHsU)ojnpIlO>Ty^#WNj+f#)n7k00qrg?oA6uRB0-j3PJCza|Qut+RXLA`qNRs9<)sYdF5fVd174$Sy@ zei#>oj;|)SC@`;EjCtL(@zvJ6+5$k7FD;OZ;~RPP{zgaxklPcQm~=6JmX@_eJ*{t< z+JkFkDN#>ljGoE~4o!`-tXXJ>8m0GWYidx#+8JF}yGe^~O0uFqhE}bm+^AkN`TGe= zX*B{*4YCBqS!vGTR-Y#Z2R?z)gVhy*!7-=N>~>>}D;ux1#`sm?S$AP?AC7pq!1Yr0 z4m`yeU03(M_;c<5rj2e@Vsu8_q|y1zCXFs8I5fJ@uDEvQBW-goOo%m>w`wg*^0yj#gz^bEDm*4o&S~ z@5%XYGZ~mNHuiGI7-H8`v3NrSltaqpuSpEi2%I!TpWUP(#sr6kNGIRMZCycVsp@xo zJ6yANC=()SiEFGS-fk`Nb?R#-Gt?4srNM16MKgCI-ol%rcWt5~%L+lDVy>XKNKBDC z5oOoYXZJ2{QwGYNhFxF%;*Ag>3n?SKMPh_T)T9ymj3$jRCO9<0(8l1EKg_#*nYw2E z>dBF#|LC7}SQUU)VvfUGtQkfzL?9&60B>OpP;xg7Rw4pP0b$HJcq?yw73|8p(~-#) zfI_Cu!CQwK-$Uk%JiT$ZJtHxwotsy-4&dsmFQrKPYM7AWEV2V|SiD6tLE)_ui!|~l zEz;*cX^}C(NsHuyD)qO$E#7Kvu}Eg5f!@j*=o{<_d6&9la>_X&ch5oM@Z!q(t;_R^ zn`xE74@{PL#$IoGp4cnE^i%eFo5Wt!Ftyb?#Rb8 zs$3T5Hi^L+1t1O9XFqAMF~On1aw-~UasG2}u(w%*EfXkdvA3}n`=GVhyVaeO^&*Qc zO#*TE+mp`|ll@<5lRYUh*$0ft-np@EO%{WlLrk{q186E7r3SNjZFIFw2c_;UEpyw>Cb@+$xuFh9Y4KLdBx!G#SgDad zX{A2zNh^&B4y}~QS#~YCVc@EuN7h88w_8&!5GiS_x3k9jfHl@T)TznIY-8>82{+dy zJGgW86mPTMIkU*T(f>l*>?w)O-fL|3_Kn-D&0+}t?zy@@KmYj3lk>|f)y~uo&uj74 z%7m<^B-U!=PFkzab<$d6fKpmhtiT=hS~ z7%1v7`exoRuUJkA5L{kE?{VN8=Y7U%*EeprR*Q|GLph6_vg&;>4UFFI6)H0L9jJ#U zH{n`mQu9l%J}lhTLxpf6ZafRZ~mSRXT^fy{)-P};Yg|}EYx@G8LSxUvR9o zFDFVRl_Igy_R1F*6=ap7wQ-~5DyceHGVXLWl_!K=Juj}qe(?*j)AAFv=_|+#RXL-L zi*Sf74>eluboI)PNTW5e>7j)SqC=Bq&Ya~q%k(GCUAV7b!jr*-G-*ra3zx(RUt(RtN;&#j12qLJgce;f1LcxS ze$5QA(~Rpc!rGcKYh1fSY`?4kc z4onJ_wp6}gNqj#hS0-#V$q3e9P0b3SQ&-Gje~39z{jwQir{at7J?1UvwvAitTh0uv znrOi>f@NE%v!+f=R_Iq5RQQAqvhu9$WH0ArngZ49x~JB)R%Zrp6G^EI4x)(%4B z+8ko5i!@BH8i_PaBX!y^eV)^X852FsFrlr~8fIUSB$Ojt3|8)>K?YnW z4H6+bG)N}WS|*yk|9x+jK#;*Z5sQS`Mg;3+;o<&>Y zxzlphl${Wc{feEI(8)%cr(gI&>{J|Piu+Jcr20^~fD;%iG}Q&3ovWhKBdY_nsrnM2 zO*JMuG*vbr3+s{;w6mGB6KFA3iZ!4ng+*KJ3D#nT9jM9m3j$Qua7`HsVGau{wfDMG z)JGY^^$TH$or(=t+^ovVq6tG-%Zu1Z&sAKGpQ@aKmP8*HohrdgWQf0u5O)MGmWeJq)o^aqR-a zMklK92xe*}ShI}_bBIkZ(riW5NWj|`?vXZIpZm1g#zYS@TNsUG58S>Q39D~lB}!ZD zT7MT77BjF5;l`Ki6%2KVVUvzA?uSvpsQxk%vt(LD!5;mtMRwqn3$q?3f zO>GLHTURW+W$Kbv{puKEr_v)5$~`;$@KhhGD-PS#>@9vm) z)KyVk67V&3%KDtA%{C@FG~2w(u~U}>w3x6Kt4ktW8WSl;l)xI_ft09amrk5vet2~yHu!fLXnXpxs zWX?=Yp$ef%SL|HnQcQlS46#$Gh0Tg`QBm(9Yp8L}4zcM)o~)umrBx$;+E9J&(}o%o z9U5x#K#3~}zu?5IeYH|5&lE|EwpSibsUYDtSa zw$@YZQ%&jV+OJI5D#ruPOxK<-MNQ1<3VRe3HI*NlUQ1yOLxl}rp^6D6BfKUmZnE+x z4K?6CX{ZR%p`j)Zl-OFG(-jCZ`6bcm>WY=L*R`9us-rGX%~aO|3eeK#x$|55`x6#> zG`cUlyTfgoHQPk7(*na)Kex$bgV+308{(+SNb(|GOL<_0OyxdoR%o&dj9k}5MNY8Q zM5-y^JZ-Wu(V@xaWP_KLlMM!1jJ4v~U_zR-rLyQs%j~f$Y+*gILTswQKutXhq4rnI zK$+|BnqN^v>@?&0i*T=EHW;=mG|{2gU+z>DRWrd>Qzq*3o;J~#=%k6Jj@f19WP{iG zk|z8PObV5@)K%PGk|!;1#NxX%xnuT3HI*!c8CF*XdFPTFiNhb*U1vq-}If+NDZ-nJ{sd zKIR_Hnx?5*Ar$_KnI_Yvy5`r-5IdCymosrYDu+@}d`aGu%GbVbNVLDx^Ykf%))}@Nbk+#WO3+#psN`<|v$R=5H zG=(UHl3y`%WV%$>{8AWVr{XTvfI~Gy&bStb*z6)hPEi38@HAzRKG$i3jEN2nGRg`u zA!paQ>2#^C^#w>+mntew+AeS7!pg2^mbrM&gng{&u5g#YnrEWu=!%&q)9kk97rqc0 zFYjS<2Ug&SH`04C_c{u@79O5MZIz2SMYTt;)s(IJBA{(GCOWiLHX;iR*==@P>uZnD zX1AnRX`4Nn+t$*z+DMz-lDjLgT~modScIp=3rz6-|Kz=Ukfq5{A3EpsXm+;tq1BPJ z%snz_Bm)u+Qcu6TXAzq19qnqx&T6$YNJuNGeRlfn>|y&c>h4+XD8g~qJS>9|7h$f% zLk|q-b&Zf91VMzs24V0GgDyAV4;u(BTmlFg+1MZ+#=!2)?@_O>Dzoyds_#tOR}nk( z3^Vm*W#+FkE9;k4;xl;v%%aVOK_vZp$*kYlIHxXvSn*r{F;A*7-^EEF_7MOti;o1N zhQMpv00*x-=_Yua_#|;7p6;ubw?$1winJ?^UYd|}zd8_owRqLwF6>?ij%WFAiJ%rV zGW1GL3Id7_5o@2B!kdiP1rdujD+ZDDUJ2h3hTOl@B*H~aN@{OX9JK5HctaE}faZS0U|d#o57Hd!*IQ;}56dX%o9=62FqCg2aQVZFTFpsQM=cXf;WxojGy1(pNB@Q&3(JOM?X#lx;uqgP76)iBU;ZLHwxU`NuZqUpaUA z@yYJ^V7&RxckXPCpE}%p;1%-=o3DMv>dM;Y-SFdFfs4(Zy{+-~D=$BB_0aZwBxs9I z5*W3qGKi$-WsJRRV_n7AQkJ=~^-#IB@ld+W8DCw>4kB}x@s-6>T9W3+p%=<|(zcDh}3#*$oBAHL)^SY6r+06jvo*1@Wq>m7yri+Kd^Fq!btu zz^ywXsp9dH%iKwo*&l)hQ9NE^(Um~dP-L~0k7L%IbQ7~wTyZ1opBZq~bR`~KwViOd zF2=)E51)A-Oh#A9UO{|mYSERbI4#-y8AQ^!ij&)DXst~I;z_K!OddpPKLqWBgc5?< zzq05{pxh8vZS~_=btm1#YA~F-fc+Z-u$sdZEIT=N#UV>GnPO^4&c_iz2h~9$SJGP$ z&zc%?sTgY$Y!FGgh?`A_vAR2v#H~y8L1a$FEfr%~?4~7YempO>m5<}ropcko@iA7j zn|Lu++X)9Oedkhix~xBqz?IShiB!pZL40XyNTp(|&AdS*O^C5pGa(o8BvxGx5F&Ld zR;d`vB0eoi2f(pvs~^XzJLx7?gF%Ikv6}eAi?OC%amcP{c0_bQHNS))G(2~=P)s6M zvS1L;ni_Jc`moKz;Ymu?hds?n*_v3FTs9CQaVl!52+N{CfvBO(*j7D`T6fY-)Dp49 z^Ms>WAJ!Boq&|$869+4O`x3uCjC>fuE~Nnyqmt}`_|eobO69>e-3F1gO?vCmZO;0z zOZ!1&PDLpdTUlJEC24*?2OW#^*~gRnOqp35vcld?S;P8*s>;pv%l6P`{urJUHjxdrATnM~fBTQF;)_*|wR zlH*>^f_*mvvsCz8mdq8z?WYFNkKmLDkdXnD5X_hr%iEBt4mh(Dno>er6GvbwluOVmao7mYQP;0sEN?6-w0;!>t)8tZRoY#4nL3EHsp%}S z{aj|zlR(r!ItzcpTnKSYx|42VlGu-7FRy;V0H&rV@ok1fbHM`dv4ny!LK%I|75Z*=Ugs}-a^0uhu& zOj1sOM^M`YI9lCFH_=KZv+fzpGX_vKG0EH28Z;{oTH`@DZkKBCYZ1w}cQmJA{(6GGBB~vt=e3L(bc-F8B5--)_`Sk;F7mCF|OU`_QVDR^y8hxuOzM@ zsIb^~NUED--zx~#(y~pRK_qRgGz4qORHEvb%f?y)X_?4npp#ras4SVqS^`l+$hB3E zBiEgD6S?iSxCU!!Ilx*%&56TxRmWK_MG5_iA9HOlq98FUX)B25PYt64YiZdg&LEOf zv5L=V@|waEd&#}5OW;9dPDLpdUspgxq??#^a$+jf3ZtuaSu>ZsBtysdwLx*XX1|#!8mU@Z z3aKkcg~3EroIpm#VGjW-xOAsJnhM^rh8VqNdR<@C*8y?6;j=l*od((7YSKjjMWCk z0ZZS?6df@2rxCbPS|E`snJJu-U|sZ(jx>JxC1cEwT62s~kr8{ol&2aK| zwIasEVM^b`#BFp%K8rw|Mk@$A zrEp4>l~{Bp5H$o(Tk|+P-ASjx^FL|c|B7ZZdAnLMBjNa5iNFw?8U^N6O*BnMOXq+L)f&Hj$_lEbQ7CS8HtQ0 zMpx^KCMMdXL^p-EojShi0aX8uA z?eIn=5Skf4sinmqGp)@r89-UQBoH+OR@(wNSlvlC!AiwT_ssNfjpSDIlA^4NLw23a zSnpFkH?IwiXKARhp#4>`)I2fJ9|#EhDij zNh(Mvn_AHQaXjd*vN%dq?0St1$X~m0Rt4lUynu}CW~uBL1ZKjP5}E;ND&VfNNJ=1T zAmBE%kHglTbQ8AicC3GHz*du#T3ZcA?ELIKDxj{5)xmO$fNFR}`(R21c@nIWrE;Ue z`e_2JHctkT^gap5uiN-q1;~Xqfb?uw$pA;3&8)hV9YkyfBuOEN&Fm_Rrv##g;A-n1 zhpRj3CR~Y~9`-`(-x-+IJf#T3;;3B+)1@eU&Avuq{w}dTSZ$5@+VQ-oC&4SpDu{1Q zEz%NaHmf#W29cDVC9}_L%(TbqcqOB(%h^E$PK7O%9J7c@pw|#=ZO!Abbtm0~EtLwo z+gjtyW>piFd}lVOk#MXo&N2^K(6J-ZT?+jqKqXHF@uH~#^v8(kYBOaJN$-)-_3Xx) zimtB=L|28ooRQTf>>whiLX?WEESA!$v_8InwRMj})SYw_qD~?l47K7fjL524N-d3w zBXw!^ZWR|-#qwa06BnB<|CciFs2N60Is8nQS5tde^{c)h$+Q)(FPPz%yU?6of z;J-G|s#!~+I+J5o9Iq=>MjBIta=wi$2tj3#u$5dE#IL3XTPhv4*))ix@e$N$ILrcC zm(+uZ%>CM|AKheKn8jpTmDa~WYwI2dtvl%^Xq|MJ9bGk(i4|S7J#n-ecQ3`J#QHM` zFsb=@77G%llHP)N($ru|MOT|(gGd@5U9G-DE_BG~>N0&0q5b%?^6lt~FpJ)_D(#O0 z)z&@^RCm%rpoSxi8C^BKi4|QDAamc!pZbFs_HJ-;<&1$eFIr*!>h&6Gf z(&Q#~r5O2c`%p_Zz^xq>QWKH>f{YAOQ;YQ1quPwsU=Wu{i%5kiv1eUlv6+pcQ4_9s8a$f5l`1x942LX9IUq9aj?3RZi2PlUiI_3 zwghvLIBj7|9HaCdORrL=Ig4U`Fo&P!_&vyfl0)PYBtdvhvQ}<1G>K+%kdXmRb~E{c z%jb6=vmtke#3>EPdIYe^A}lG>* zVqtKm1pq5QX^%ip;fU3DvnI(Y2s*8TN{4r~h#w^ar%jJRB&Ci~?~*}vVdIPns$WS3 zRfFPbs(`8+fhvyxFj2x|41KlL{Q}fz0n2z|@o^rmb*HF3zqiOwNf{ z48A&uNy7uC2uDdoL7aSQdF)3CINDSgMABDD$SrLwD9F7$47r?JPM3>=2;1eBlSM#U zlXk}eX=@z^q&w**Ac>t8)~EU(4OD6Z65E_OE|(~VP%RD7QaXE=LiN0u8@z&v>R1}@ znxvf|el#_B5?!hwBO{k(cd0H;rArlrX2wo48dE8B#Np@~i+ZGd0Een=0UWCCq?=G} zx8XFpRADZ1!|Cf%4O$gPEPYiHvr9GjeWXZ4r?lVSw)ifkumulFoICNKLnVqQNk3`@z zdW$OAD2QK8E$9*?hc%lIgGfqkQP~6tk8O;tiWwfs&8kbxL4-C(%pGu9Y$VWa2)4HN zaoD<(PJ^w{rMjlsNIrf8X*C?N^aV>yo&x#p2q2SzRx(l$Z<-odi7wSOn-hab%I;D{ z)thrG>(X-&u~VZgb$ZI;B!PBAaJBW1!_}R15U#}Dw!Tf@=u%zNoFreDYQV5KYS*Yu zE#tveW|wL}K0mTa@Jen9;#*S#FTqP%v)M6-q-QtV>k=5WO!GvV_$;V3~ zjEW+4{R{%8pkUO-liU7ANK69usPK{qBH_Btk3k|TwRepNvu+}^So74*6~vtqV2PNz z&f+G4sDYT;&^-=VchXJ3wu`AomnzJq6;`#)aIDTp<4TW?z@3~Yaa2UYG{8}^mi{hT z5JZV--KNSQFpVlBxh;j(gbC=AKy^7gh|sCYFBMr?L?zH|2&lI9aiF@BZUQwTvi`L$ z53g&sQmD@4m=(wC0s^+Q*cwrva%3i98z3)DZNExISDQ$KAb^WEXc(@aN)Fxb;v_7( zFOdYT%jZEvP6aI$U0Ecib!mM(9kz9kgVvpN6SQR1X(HOZ17cm1m{^&xHWQ9k`UN;9 zL!WREWHF$0K*Cg#TM$p08ceC^YSU{FN#mod*&1l2y3~;h2~?Nog9x1pR4Tf%$W80g z{y0!=?c+doC*1^U6N}+etT3aiCO5I7t7%pouk<5uu~E|eHiFn-_$FZ+AUjzxtIMRA zs0OdwR2u{VEvmtuuBB|%U=a2Q>|9wSClEE15Zju^A?r@M30Wepc)H(=YVf)yIr+L& z7v~Ub;z(WQ-dTu_jUbVnW9+U~@>vkKpITg{(q9&-sYO=nF=@_fuuJSgl-p%gWwDvo zrS_{s5|KkZgHy8_tSL?Y92hk!j?@)?g$W%U!D@K2v4E9K7R0Bf7EwQr zv-{U=E)60nyGwP^ZH3GBtSX4gjHjt!rS_^U4iktPil?^Taj?3RZi3ZGfr;RALVVt+ zEolyuuS*rNC63Y6S!O3HGBScsYC6EP<#YY}$&)vFT2_iA~aXo4^%Dm+HFaD*3upF&p9FT;q1BVna_8 zosyiw@%b*dqC}VKx=oKkB&7h!?ow42i6H)hm`q?&qA~(1bq(PP;3!ZAsck(eE4~ownrJtb?Y?EUUQQ4iVq~2S0u96UTxjKmGeq>rp92s^=2wMy#9KzBO2ngF| zARz2cJPpE6_l=0F2}-{H)v$313@>mSSi@h*gI%?z&>)sf6~w)!2I5aqAhyXeh^Xvf z#->VmOr^KAg2b;&+d-u66u%Z(3CFMW0Rq3a1ql4Q6HmkMUz#CU6PA1w2+~{xbZO2K zlLJA1JOa@)Y9(z2ajL1I_9rOR+RPb5RCX^bs^F~KU6;j!$eoH?a=UABmvG!lFCcJh z+kn8WJMkuNiQO&s?CP^7cr|~?*U=gr`9fH%C-ao45vAQGqIm&}_jA`^~ZX@3I0w)P49x)X2Wx82tFPfY02 zWG2*Z1Ze52nwWR)qaQ}#N@;*bswA`^&NMZoev(3}&8a~|W%HT5%FP;UU2YE|Gdr{Q zgIbWW)?zf_Sd|_?VAa+?fmL_nO{@k3i`$C+S0-R-Ruf@b0D})hUbKgd6SE(O_%wp!_(&@mFNi};Ez*+pD4TSHh#FO9 za+(bJSRw0CF8v3QIu)^GthGo_IAWy_5Qw!cKp@tgcoVVVnA60x_!|?pn*9_iN;$?Q zfV;+hfE58}ovHcn2MFj*E~v5v9n#`m#QIykPQ5` zs1tkIrrNc_Qn@lnJ0Ru}f$} zEXsJCpQq?dr#y5fmBMmcX;G>CSf-IGSuzMxk3?oXZ}W5zQCnoj9;D{1A%nbA!YZ|| zo#*kSK-juz041C@^iN>bop=+g%_vIogUbWlQ^Kqv!=xl&{pm4kS|Z9W@Jq+&@P6!t zXGbjvG;$?B27yC|Vb$6{3TVK?J;!F~Afj670F^MayJ(mEU9@y|?2?2Kc^P}wXm;Z~ zj~z8HtxrT)TlWNN-H8WLON87LMHimxCxH1#!1`Kb5knDRU8Lgc2?|c!=2}E1Kf-B@ zN-hlINK=cha3c(kQJal}h)Thd-5Bf9tV>U(LFzJs5P>@dDUS*@FU?PY)K)$LQg`A_ zkW!d*i%xJYj0UM@LHT-RF|!dUrRh)1=U%bz`jJSk3p7k64I(PL zC06NHbBC?4P6jY?qv~>g5Lr7#D39zkFD*|X)YdzJPNf9ErHUj}+cj9RfHoIo$HNnZ(H5;}Wf#EeW{4$?}4NDNQ zQqUll92dmBrWS@t)`ZQnK}4l+%Vte@EP!lw>5`L6_d%rY6u&&4)4cQnA_UtOAn@x> zyouj-8(On#c3$(Ge1r+ob_8_S$tn|OLfhHkc{M+-8GBq zFz0p`WS&y$Ol@};cx)%-1vqXSHo$S~nRpYo#D0&{H4BrHhdREl*?@U*_%6_VCuY}d z;PZZr22W~F&?pWto~DLkit)5yb8e6*#%4U3dd53rqYKF(?Ue9K<;@E`#uEq|%9|V7 zC-Cb|yoq0OH`xrz!el(bd?a*zj3?Bl1Zd~Uu*&S3ML&$dmC^u>RLOTi+<$6FrFsDt z@-m*h%FP;UUE&X--qakKjI}(*6DT)`Ra^fAR^5p=vD$ zF&ha8k!j>g&I{sNM+3Pw>jn{(+TXHyjE)|4@1=E=OYUwLT=GwZ%nVeMT|5gs))NRD zM6RuW0=e$Qo5-bNj9XwfyLezeGRpdBiiG6|(9$m+O5x}MCQ?b$1{5S&ZpgL&$253|rhFCjtkTqnRz=I&6 zMGe{0QI)T2cEM#4ArhxzmyEVN`V?Ta<4G+@ z4Kt$zZm?!=o|4FwhtV67UmCQJF#W7M<+WNDTZyBvjn8Ub`r6*O`sKL&BFsUer@ znq9CNI*6#${*c`$yXo(9x6GBJ2GEg2(V&-D(nzQ%Pn)oPTOKrMhMpZ8{AiD&_3#j@p%09kn!O zU2YE|xgUzw_Lq$5WHjZGnn1-t%-U8UFzZgdiCJn}=59*OuGvLRYVvi>2JJ^cm!>u` zpPvm1fFPUF2#sJ#Zb6*O3J6_hz*N`lqD`+sL}fR_l1k3GWp&9ui0GX{n8$Jg9S0$7 zn}L9^JMlCKn_aVun&9N?nhhJ5!0;0Ht=aHbB7o}Iy{dG82C?M0AnrA_Fif!&7j2de zA}X5<;juvW4XsP}L8R^!zdW83s5gjT+X4iB-HE5+*X){I)I292VS=IF=?9V9%5JC~fFI!Wc9+L? z0tE+gYukXptvm51Ziy`*r)zdmvz>fhvjOuG@TK`q%&ytM=Mex0RYId!GF}iDn;MEK z#?zwBxj{r_GoDO6chEHp(lX0c?ICMzi^+bTMIPe`gbm`?);@t>cj8U_QX4&Y>uYw+ z!hB?<%Ex#@ZAyTaeglu$HH&^2fh(l}8mW@+g1GY_x!qz{5 zTzBF@$Vt2|9ToNyRgVH88SJO1dX(F_ zm<6#QF*7etMJ&ZGUgA-pln>yDZCC(DtY_j)#D-%|Ga6fO&BCN)avUm3ImX3-yL5*8 z)+~a*I#csEwTGwdp606PPvcjTV~~JwB;xH7ixw3hZ;b}XJm_^rT9h-X=E5=2BtCTbPdoE-T(@s}is!-@l{Z=ZTHCTCSLQD=9OGLro2@ zWW2RWJBX<9@z&~X%!62$J%mV|idZt<@@P{_(g%omYg>RotUK{0V#C3wiD^6DYT6Vl z-kQcGfZNTTP!WemWs+bDO~+eFpF!MfYWSsU$V)bX2NAVJ4cXH*m9J|Sq@7Z(Oh#KC zeF}sPMq69^1a{qtH?d1>W<5=KW(|2s)2IA3WW=TfWNF$IyM~N>*dI)(^x_naRLPP- z+~0~jUDBnf8uF6O(?LXSkr{iGo3n-t@=gh>WSr&kq(ImpR&D(gSam1f#A+zcdH`#7 z%`R!Ols`R2O-n#_2|-m@R$@cSFGr&y)5w+l7z7Sei?LMK?2^sUK}2PD#;!0rW68=b z#j*fdr-WKE!t&TrAZ!q|w(bekx)X1rmWZ%{uGuBcj`DTQA{HgUy3Bnt78x4BIyplE z<}^kn7Y1?ssbQ4rnq9KlIEbk1u33-TGg6R2*rNba9u*3N4T98GJ^@m9;%Oi?yJnX( z3(D6uijesx>Q%QM2e80oI*E)10yJnYc(hVXiyK7eIHaE!_Yf3<%l2@)Xbl#oeoOL^2L5H^TRTk`}m-HA7mNo@5v zU9(G?+T`n+_1cO6s6lFCcFlUffZ#DaS2y3KzWOkjDYSs0=n4! zCHJk_pv)wKmyZg;iRzjKAsGZGcGoPav5Z?*5PM2tmxAyzkL9Es0Y7zZm;r~dXW~r= zlc;FoTbNz5Fdxa2@O90Gjf-P=nSSw(`POV$f{2xZ2C?M0AQ56}VVGh~EVEcnY}SNF z%egnSATbla?Bu~8tI6)2WggE7gbjt@h6M=xx)X2Ww+T^tLu+=;!h9rteS`_pt^{=H z3!#|A2lC_DQ7d^a2x_NRe4@H$mu;pEA}YIU7FBT0Q&^CBO1P!AyURSb69^l`t!)DW zx9-H7xFxoLoUU1zlmxG@Yc^nB0=^5LP8vn=on>~-20o7fIH(dj_DaSJ0*$G~UW)Ow zY;$f9QK{+_n+;{!0Pm_6ZN?2EDx2}-)l}A4>k@wu znHhK_860G+xk0Sj`X{jJPP~cLU|@0E+)c*QGR=4*OiMs^iHx&M&Oh;KWJL(7 zf<~_7ydbW1G>~hvZV*x9k*oVFtr$jPhGIP}yW}54=2Ya8QI^Mg0%3#5we?RR*PVD1 zxnuxqZhfsTo@LE?^3fC>(M+}bulbk6Ls%gAR`U~PvQ`^~+^(dQwgNVwm z>_io=oDNJnFLs$gh}5ZwC1Wj*0<|oCfT$kH$Anw&6WG$o!vW9FEco0!r)QLUaPx-oLmt7VS zB5|iRWFCDAgbnU)ZS51-btm4$E)jY>O?XxfS<|O{U9+p$bFc(t>9_FM1!2sY{1?tp z8X%Fna#ZM2R1Fz~WYDErWX2xl=By!uyv*~SF5~P9k0+%(0FSc`{o`2mOuUKJP~`Fe z)~q4Jq~v+6KRrfGi$iwhOmK63ij1f1oEjC*ALeP~N`4Fy2BsEcsjk@-7DviT=c(|; zZm(VSx7U*US&)^9T1rF)qDgkt3XdHH!iFMjL-zz~-HA6*>r|tNfOA5aZ_UDdBw&4A zvxuPxu+qFJ=3BGKXZYu5A02+Y#Bl=K$F z*{2phsjk@-n_z>8%5H{js5R@h)MfY}(xxW3RQY#>M{NSN29arNo25n0~cbVKRQYk98 zKQ<_xALDeSmE;x#7E?no)it|f(`yh>*J4U9t}%I=c?$-`mo4B_7KObR2}R zZ3Y6u?!?m|Yd=RiP<$9mcWlwa{Zt|EIBTSdrb|*6l-F|X4xR3 zvRM-zEoa}*x^y2zYCkObi2zx0;_;k7y+Qoi79jBJPP~cVb{kr=Yj#ERoP2}{(yjz_ zSGmu6kROOZCKcIf)JmQU;#5;ZE!8!stO+6H%qU$x{T0nTSoz zi^*8aqd+MiAY!d;0Rpk^#G8l>#~cT?W*5(@=0SyuQjT#6;4bKB%WS=IqHTrJI%V_s z0EFk>Hr`5d%#n__Ha!OsH9p=NeUEw2>#~FpnN!hA##eLLGjJG`6)T;CW0m-B62Ogp1&@9|IyQ8h zLeu%Oq|YGkH8uQFHRM&Bz=Md|qK540iOSbC3(`(0S0HyouG;arTs5L)K&|e|n6XmWZ+|+;?HU z>p{|1mTbqRkt_Kzh-*y^xm4Hes?E?rL}fR_uKAl`GiApvNeGd*)BP-u9R-RFMp#?- z1Zv%hH&IK(m%v-ItC}6<>zYL@N`RGqD394Si+ncwK2~yJ5J#FCMyamZRhx~2h|2Do z^{6%@1sQ~8ropM9l}v+qR45QO2vS@51W4V9r-9V$nqAc_C|}nsW>NyBYuv6`Z0raq zQi(SmN+sn5@qMdY+Tm>!mH2i6I*7ttyfqYr;^Nq1f{8^xm4He8jH@v?wVbnPS-3*JtfRi`R*Ez)C9tY zqG`hl1ZLfdH!<68Q)+h2!h9rLeO7=Qh6vKZ2|Q8f3AtM5Q7vyS&vAK}7E|AztILoIuAx2-{{LAnZ=O31KJfQr>O7 zH4F2R5cYM=hHXY*n11n&*)}ub9kr6@f;iRGP)l{quGvf*L{xUyEUMs~r?4P1Gj&eIExFz0v7JEJAZ~3N5V&qsEu4%TDuWL47UIM=Kby3Vl+Q8=#00&h<$6m>JL7>rL&TO52lJ5kp*_<0h zR0<7jHk4@tyrY(myDsquk(d!%$=+RZ>&s(2fog;JwY5*+*PVD1zwP3#$#`0$8BeI) z2+$^66vd=Np;Gy=Oe0nDT@YuQTD+wgPir>g1`(Cbc=GBmYpivNKZwk!SS4dEkMRV` z4Pw>SKY>+u;!UhNd)#oOG2aPTqZvuF7T0C6Z}>z_cbJMkuR+eKNki)T%jApd)M`*vyagc{bgB4wYxXGSiMJr} z5vU`t^Jr7b2XMqTEPx}{Gw~*36XPvRihBrD1~k*SIB?g`aIa2vl40`$2t316jwNBN zOZp5F5T+JysjgWNl7ZhAbz)D$R=%!Tkd}$vE~D)_k3OaBPhi*9K7n0#;!W%l(KgUE z3-gh%^>xjzx9^%=r*DK}mx(cF%G@=(E?F`+8mUxf3_>!H+9EUdC^u&f8RTVRl{%qf z;8;-m+I1dJN_hZ*Ra^fAR^5p=u^I|G9>AJ4WSEo$tUo%-Dcw;qO!YY zJ$lO6H4DNX1(5QnP#|omWZY0b0aAD3X&^PbX4f?f%GWiEnUp{&cHL7nN|{gBVxrP8 zm6R96_oud9rQVudw@EjMsO+v;r6|oE&QxS_IX{T3orX~!*$I>yM5wKI0-^52n+SCX z6cika7Kv=)5pm~3d)wD{_jVxu zoF88vU7x#WbZ|KLlpGxShn|lfga26{6>k%#e1mV!>)~D5j~#pZ)z&bm>=W@7-r*PK z9v+Xj-Z%RM8)5^xPQ6$J2!*5{2E=<8hyT8u*Am6 zlQV;3XZwG$V@_rUo3VHbnd4$TNA%N%h%+%$6mJ**OVg4UASc?nwljBKI-L9Xy;w#0KA8XU~r zy|v|zR(#j%@SMrKS-h}V^T6!(x=OFJ#)jo)LQf~Eq3b2b{~=?&xtUmR@Rja$GKMb5 z)x}*kpAqKTf{*n^QHd-gK|KLNE_9tT<~1YURF5lH-x|FO=3k2Wh`2}Gbhv+QY%N6f z6o%FRWvnxBlgrMvZ5?&o)tR4z+$CmUtO^X(s6#1X*yX=Y9DiU8+n@5+TK;*t)Gc6+ zCdZu-mufXH)gD)lQ|P!wToz3`2v858m{Jcv%{eP(V618ptVXS;xg=+SIuzKtP$;B| zGkZ$hzPWep@anb0&9Y)~FgPKOpM}3B zzEM11k6HY*dHrDXj2zc>V!oC3#O-?Idneo5gX6`o!{!`@J^{2I?+{|LyET4l^YA*% zw|@p-cdm-B)F0g5J2;rTFdpr07Jmloayjp+UL$9Br_|bQ8@GVg%0kfDUe8t?QArZ- z>xt8$PZqgN`R!pt9YdzHS{1j76PL=NN;Ob|Jubcsm8;w<(JDC2SHXvR;!Dv_U7Bo< z2gi$Fhe16J0}^XB;1fSR3*K@GV0bvWx;;K9)C7qJ{^{Vh5jJoerY`Vx<-L?EIMIbx z8KJCNVwN(*M7*eBng_>MDm=gsgBkdR_*5-4?^C;u7hJryQ+?{9xlnzngxm`@1~ugR zFFko0^gPqpEYA=;oWA((p7?u>x#Q*AD6@OT+OYrA&`{S)^mk%dfb96#rXad2zRcOP-YUe|lW)uuv`90a-N(C}oUo`q6i}SI zAS*oNEQZf5GsXw@u3X%E>~LpflvV@Mi0e*HZhl*qKUH?yv;1P=c$jmgh-QJ7pYX)`yA$kzI%( zvKjGB4mwRW=R#)$35}WG-^wCbyDx(yo$7g zGKfZjlBAJ>2MqzyIrxPl^wh^~9zN%#(eC4TfF-HjsTkAnr7W5`|B==Y*kczF@MHbi-|94NEp`6B@8F;)*37>DudJ%539KUNkY&l`=I8I*M<6WSqv&$;6 z{fFY^{u$L0Rv567pqI#m$$Ic#n?z_vK!t6V@qH({ka}~?Fl0|s zYuGOL{Cy`$d7gMKWF>XEr%G?(N=ihqgTwvF)_804U>nZTTIuxZ);j(4j9B;U)F(wH z*u_Gw`-)ZWj*E>y!F<#=SaC!-e6i{1*54{lm@Yptd`aLhY|e-O%EGXP-xQUvN5@wt zyPG?sgD0RI{hZn7i`&NA+movYlLPsO$BIAP9lvuE3cAK00tI-yyCwh9b7sHX@q~f6 z7NP04*!LYI#=F|E)tno-AHln?_I3)$;{WZ5`w{JKkFFOTT=SmW zKSpo>vkP^NYGYn0Uu`5>YHR3wK`;1KH3Mp8uQ#&4))QZ!ZH_V@Zad4hWwru^1$kx^Q(+Jv=!^SuXcD$D)h@0rn!2+i zxu26->(MQdAL=J9ILcN|Jjtc>VLq?aDzS4yYC9x8($BPh={i7L4|*hr|9V#fvuF^G z#0JfzdAxH>z$sN0lPk(#9=jJ>ZxgOQ))yCYs9WWGx>Ax$J)JDBjk-!$WvLN#2kLbT zp3sg*&|+80FU$p*XNn2 zK>u{wLT$|Lf;?wPnO-w8!e+Sh-NV1krUI4iq^!cWN{3x~yL25g+c`}pOyc8dOTRH& zCmG75#o`blCw?1yajb|9r_&`DN-0_VqqMBiP!(WJZc`V@ZRZ2U7XQI=Bfqy(wkGXB z+s!a9MAZx#@$Z+j>7YCzsWumxtXpbz__l;Z#xRjLyg=rdmM|(+i%<2%>+?wJNxCvA zv@}tYvl$9)>}%VnUjJv`Yf&7ZuH7e@D~AR4SE~PN;dBOz6~(y8buYC%RbZhM%)7K; znQJ~fCF8`@YD}^!p)ezUHf`IiGG-%P{0oqb191!tHdy8l-9mw!-Z0&_@`+(bm`<(NG^Z8uV4_|tbQEE!X}Ge>NQU5 zCAe@TCUqmpuD9qixstl?N=xc&;dD|Lpbrywb0nTMDuKMKKo`lJ`eIdtOmQ-MsMNE@ zpG(Vzv@r6lF=Y$p!*o(sq{QgOm^CX>qo!s>dZ6+K2*oV2m-*o0 z_Ooe4UM+>rC$hU)N~u=~=yfy5DlmpHbv&uOQ+zlrp-|F!YE3B#g&H?Kq0mDWegu4!NVNlXAmWlSO+* z@YH#5^juxn6W=uxaZ=?^1B|<%GRJ~pe#;z=_R?vm&#Le@x`8aD3h-^w|sTJFuQpNGycANCR?uQ z+#%Kp^oH4ubTk}MJR5U(;fI&!{aB8nv%JcQ8!%lz| zO@sW2m{WYe;~H1V<#I8pv-Z_pzlaiB(r3v|^}2b}y(k+lddY=I>xyyT9rEYSo0KN} zR&A&W=Tf?J2X}k-_iMkr-aM=3TG&y1v?snHIz@mh2u5%Z_4&zTkB#@Ct<={zpQ||h z@t$~7CxYseR$SC}9V%bwi*asIy=Nxa8EiKe9d@PvG@>26X8y+9+CI2E+8*uKCC1_F z-yS`_eQF|as={{ozNPNiIcupCpX`f^(c&P01s+v(|5ite9A;jG84PuK1SWvUm|b7c z0#va29~}l8>W6o=JdU||xOerw`iw5W<4LWA6(}GD%D#*g^vn7xuk1DfA0B2O6b#)a znEqY|z6Qs}RR>8o))-DFQt?QE47UVMQ}!_A>}<08QNb|DZojgArQOqJn}q4BzuN)8 zOO8XQ?s_G6?DHm8Kt?fY|4vW5RyswA62BUA*N5CBAfdYURg>*{~|L`k+wo|R+wMfnK5r3FR!!`jY zlg)mdS?6mJY^FS4 z=HcGIp2-p-En-c4b;9c~*_M!mvPo00icHg~N%8g0e#Z>zb?xtVaATQEqQRVNpFA}Kfq9)1;TEc+?EurGbGCtjT-t~~2@1%5e8VDLI^ny|cZ|3r*>iz#w%Ics-h z@Xc`w)B@?EWBer;Kf-SdH=V8&!h1G|k*?BoKa6oTQ&ZAY;_le}t)9567>&^y?q~^C z#0r=(8qmwb@M~D0FqXICy=ewlRlKhsKRHFeNKp^F%;Mw%0FoppLGo=pI#lOabd zs31qgxQ_Pz2G2XVR7$M$1o(aQr@`;#3O`G`?s)fLJXb#i7d2Il5{B(%x+K1*A3r&R z$&scY_HCm)-)A#3kZQ2_o1VBkcBQL*9;~kJ8jDs=so?ZgADO1|a2p!p*L&hM+}WF6 zTjjpSJJj~0)X~AK)~FUY;$g$$jx@Fa<)qyA=hP#B?179|q05Cx?pR;RSPHmLAchvQ zQ}DL1@p^XgU|>(Yr_XnAL>u7X@=-)&I+NYQ@&3+uYce_<&ux{%K(@NM81K-ICsaK)t2}Or;ers$M_ z_rVgUsAIfVKyc9AEM7ek!=@kfxKxL`4-FrC7Hzp=RfV*s-Bf zp8&>2;%7^5u8IWw$goo^lXFi+PuWCLoA3fhNQu)mHcyEoRm(t-;T?ev*hmGF3KAgN zKi49}k&?)Dc;}hce+5X@d zYzpcd%T8{bfVWZMk3?jq(iB8zDwZhJu`k|Kdkj(7v4 z<{&y)$7L#TeR@nZF~yLLWx*3akN(TGx$;nj-KZdL2;?=ST4m5MMU^^ZgSa78mLP77 z7aPRQA*gGHk%Ex;$+)OyepcTroXuroBAX$S*JW;eWHTgG=+IeUOk^`gjE`*QAoUHx z!4RhLXLV~8*Wtx1gO^6Tk5|h^@zZ^9;xsrZe;U4mT{I~ENtkBsS+@)oD9h+62DK3z z(2M1=uHjn&-Hp^{v;Pcnw|`_nZwjIVIvuBNv50sd8_*5Gyr9k*(2Z}!2lVD35YQK# zjO?`dva`1hZaMj^{9mu}Tj_f0Bc5G~<@IYUONk9rS^?2j zP(4Iy$pJEy;JF~Bu?sRkiQPTc!C&FEMTYXw-aB*H~kbWTiL|N%|-tLFOd<6FvwHkdfrV zPB?xjrHGHm)f$UfR;@m|aFnjL=zP5t#fu2YMpTdut_^hN{@(aQLlYg~9g1O=q_Qd# zNdjz{20-ixYep&2E){)G99m$>uHC0dgya!Ot};a=hr|{N$x2$C%tYoPN)WF{=gdRY zm#SJxj4#_{tf=5_PhaaB3+KQPOqv{gZ!iPD4Efv!Nd`~)ZmZiV zPZjSmB{@^!`gJ8GNo*J=OcV8_waTpUgEnd~EEO+JC&{al`E>0?rSodu-ld)qm#Gs{ zrO1LsowT-RN$k~JV-tHlKv{0IG7*NVWC|m`Bkp`leg3}vF;JuT>DAY!bmFV;pT(6> zn!*fjs-=opsA6Je^ZV^$-v=14fl{l)-tPEdb9?f{*lupjRX_UfY~lr5!nr{j7P=%e zDUH<^4F#L;Tc|($(j0HFMn|~wK2{jx#R=s`vDu<}9VuZhj%8^7)Z9Wt}#hTaDm%QexNlMRB|jULandGo@F8HZXfAV_d8dcW$JG;Y|j>} zsU~k!7KT~;w~eJE`%#po%SvZ;=#o7B#i$~)62-eugCa9)ABG*z#F9dkRH|T=R$5l$ z#g~@#5cCl@Rr>=BuR>v=s(ZStYTlW>b&s&Uc@?rCeoQ2u;jo?OyC>&MC z?1YHEa7V!55}U1RwtHI*%0*Z6v0#ii;py^Lej)ynaWB8e58SC1=|=UZ!*l+rjSeC0 zJ)G?A&h1@0yn5|$?t#(u@&1E*`#W>`&Zm-L>#?UZagr%`qG;##w6`DA6aN->6A-M(=vGM3pu&9PB@m|vjGCvnT%XluUol=|wll)2#?_=7D zT>9)RsYH2#VpFNa4>F#{YblFn6}etDp29l8-qD3)-GAuiMx1)sD)*{YXkFteJhfk9 zJlNOH@ScrqX5@hK!nF5xLYK%fHP`NB@X6ddes7@$| z{xhoH-b^+G=b`4dqP9vY@_8NJFW|$u@?b-efMU5Ve^u)OgZUs)rj^fK2DG|@cou)qfrgCJjqLtVYJYeo zOU7&%)eB`t;~%{bQ-%JKh?BnXRlY*%`s3$)m0Ck%?49Py9x%I;WbH%C_?>YnZ1{szmkrIhjNA z(s1$fIn$rggPlA(vv!LtJ74OqL8m;;j0%&=1qcOO1@$Nr&nxVxk$ z8q7ZOZgMMK>q2m5E0$WzxXCTW^#adqK_$!U6o*KBHs>bSgdX^g>POA2<0>7_mtku7 z;nh}3!{QUU_7x>qs`RY<@xFrfa+(!< zl2EDHx^}r>5^qj+wxFmf6{ZTa=+AU~m7z z-qx@a%XeiTLRUA!cE#gq-|Sm2vhpS;5iQ$ z?lJ7t?K)b{Iup9|kE0R8`sRL7{FAVbs?__&>?#a>iHI;NBZa7VixXXqO74g$E zBBdVZPhtvH)P3`v#ie1F7M!Mu+{-|S(5BK0@rTF7+q3)d`Rlu*oyp~khm*t0j}LCM zemm@F>Fgl&gVeGEY17MuI9illr|Ef$1cEzUzi8loOLjlIFnVgTb8Sc8Gd0ERGPES3 zt?Zwg=^)FlH8HrU{$bcLsD~XIbDF+OxU9mG_}Gj{sUiB|`dG1L-I_dEZ(h$&p6qO` z$sasxFHR|m=vS!VzEmBk$*)kohg5I$-5HkgtEW*bWfB;*u(<##dlmAL8k8_8GC)ozUL1IE+cqwY`VkTSxkD9K+!0zW4{-+2Y{x zXnQ0t92~wme{Nf^PH^u|{dn-L8{%YdaBTMYv17;1@s`bB>5I2@rvhN(Ro9YquI3Kn zu~e_32ZTf9?{un7=y57;OM7(K;@XqpYod?-E*|b(y{|r!?9zQ(=ogZ@4i_!Rj3nc> zTTBc)93GrRE{f!JO{a_NnofvkI`vo}L{#_tb#1lbv}3ASQlEXUxcQ0k`0D1?WCyZp zxF4*AO9Pv*!!YUW0z34)f}ho)oy!WO)~wwF8(y0KuG6KXWd2Oa)!)%wy!NJ#OrrZ~ z5)$H%({6&2n4}Fk zU}I-I+AW=j+oIWy?)Kl)`tqdS3@Ht&+gMTE9uaRIyeSve&G=Xv&F$SY;-Rc2QWUq> zN9Vc4faxufb!bIqJAt4<72;lyu=9wlZmJ9qGWLtlcmj*c=@6!1+&SG6!>u zs@9o{b+AljDQz|JP|?$-H%Kg+wXDVM@w-W08%o#mCX zqMYST!;1xB>dwQWnmwE6bY1(CLo>q|xkqG46d%shXya%=)5~(uzx41HV6)Jm^w`RH zQ+HM(RtsoS%Rzw9)AZ2Ex>!T<20n*IHYEU2mig7(yGQ5XASE%=^s(ZIt-i@v(w#a4 zr36pw`spvGT|MBYH3o4kuQN(5c7sdC)m%8GOuRpLJvJLz3|>ggYX*lXmN#uFNn&|d z#?}O}&(1hccrDkLP3CE%FOT1=!aMLrx;$Qypy|X1^rG{aD#pK6yk+o>sl|AE71cLd z79Zrj2zEmJJmcvpr;fU{zjqa$kD9X$w$98Y~`s(Q6(nmieQ+0rk>l!Q!c zm*RNOwJ};`F3Swl+|<blnLaF(1#>&u}IxLL;$F`2BQ0$J7nJR4zqE+)d6yRq|C<-`^KsUwB=s)(S_q z8cim}DTC@tL*$H9Hk7cMqsr#~s3%_EUN6LnF!4 z@ZYWQ-#Pft>?xrh{yZ*zs8!*x3HJ~`e$d}zGC zbM0_c+|D+5p7Te;WtQ6;W6xcM0g3b1<8l%+;!~}bHUA{kmj^drk`oL+|4f^x_MSF$ z48L%WI%g?M7eIAI*Ej9|} zWPA#Ob-$saU{R+W%S^oHGpf~8Qb`6WG*&3dXcytSmVjRDjh3swt)p@YliWbC*w*3o ztK-s9@nfdrvv`y!BE-}6f8v8ZaXwu-Fkt-}b@;RA;M4rz^aH&%EM$*|9USgYw#HkV z2itHHW^pavUwt;!cm5$@c0fx2E$Jo9Y*zXnT* zm`ss?v^bGq%DEub=vP$>@qYZccwLeYS~ecsda*M3#lqsl4A^T!zu_sb4gH2UA~d;l zU}7S%C@;&se|KvPkwjlkF$2K4ZB_hphQ2uU*Jcp5L z?uqyH6EkY3nv!gR97&LxVB@(p{$}x~`l%Zs{p3cTU1#EY==CS|5d?n&*%$XFVX$^6 zh<@ssvBa#nY^u%@JTbRoYp30bc^NCMvVmeTRVU_Ylai&Zc~)6Q%2Cp4BC0Qc-RfwQ zm^ygIOmWzqXY#z7l$`vINyp(YCP@|9n%3&n3v*Ot7pjVEZBDxK4$-wmrMxttD*Z`r zuKEMBBAaQwNGQo-D%ND24GxIgF5%$2M))5`Lk}nrZ5*c5wYe z+pN{`wW9oAuB|DWk*5i;LGo$I&g7$$e`lm3Ip}TOenwQiW{@j(%VD#s}C1 zVSFO$jb)`)@z2_)M$QC1`<|3lxC+hX)~5xV%dfOXd}?+*A-WUa+dd*mCDaL-b(n}> zl-ur}Bw)?8Rcu2nXTi&t7e_NlOpPhWY*atO{R98CCth864v$e-dP}`jRY41)S8Fb} z-{)!_4_mF5dgAP5zc`+0(NpD48Df=t$&);7e33jA9c;A`J>5A~q-3Gu$3cvZ_O+Hd^I#VE{nDTp@w;=%W|4;Hpyx@drl)&r3Wv5Nor zqWws}wOc6ByeT49LP)dBur0upkTNBUI3|Sxt*ktZ^TTSZInsiZMmQOE$@pi(VnXCE< z<{lVbAMZa1smPqZ0-$ij`POwO>9zGPwEnGLiLD7UJ5k_>;1+dWv{|;c#N=RN3|O5sCoL#YzN8%lF!J?bl^4O*{A_CHTODEa+f z>#Ro!06$dqjnsr7{M1Uh@WA)xu4@?aH-XR$(l*mjlT;;x=n-J3)=8cEyw?*0(WY*W z35I5+%9x~1Ntm3}sgXd{t&=+akIO65t<#DYYLXF5{9*h3r4pg|p{8SeHS1ubisG$T zfM$B6?8u+6_@4I3tC}a+shDSyxk_CP+a|6`1TQR-PcHtZeX&SMB!hlJs;b5+72Mh_ z^gpleB-d_XqXH|gzPEj%?ny}--C%7)gPKIG&(>{n^F@B(ZIg@X00W)0J->Bz0$|Cl z6<&|EN*GO@BzMLAt5%l@l#yErQYD`(_EzwQD|_S?+^vtbKjv>tK>gNaZnfRzx{lI% zJydcadL_GP%l*6Tmo}MNYT-z0p01DFS;lVLJU!pS+@EZ}nbdl%C!zkpGv6s6U7%6_^@RCTMFr@&B|x zqkG4-@&0w@8C?#Lq+P)S6Kru0QyIjc7;~o=Tj;diN?R}H2r_~tSXs~w_^LFS!)(m* zZ6%9YzO_n=MLjaEYOAzb+%3@-ub74)cTO$e=FqL0n_Y)ZDBi;5e{G$)fC=G4ZHJSA z$i@bfrd-INV3$GL_ej*ssLtisy zQIrm0GDCMBr+r0JH8NB^+sF)j+dvFbBl}MNfUi~Q-Aa(ZL^2p}-?&xX>(YPvgncH3 z?|FI!!7`8?wGazbS>N50D|7$1eevdEe}hwAdE$BNv)BPzK06I_>Q?&yOw7H|;+bGG z8Bdc?3$9A-E?&B`6v^n2dg;+jJvemBVhv|27)T4)Y7N~w#vG_TcarBvjqTuz9Qw@7 z;-<6DDcB$5A$ZYWlN5O(P%MH@~qEk7=kWKM)n6}!1@iZY+ z8Wo@GixglR0=5d&DeJlM&5HMuSHm$eT6*VhNs|d81(>Z+L;sk=02Y-tHF{V+S}yq!kTG)l#dIu1pFoO_bzphL~mS)TGa5 z_i@&Y5L2 zP)F#UBaa4ja-hWWdMH&kJI#xX;)l}{h!R3hwwXnLMXs4|UDW>6GSmE}Osg}vgUmDA zw-7yi@Q(dM=dfe7O_a&7HLDhKT`0_Aol-B!TQhtqXXlAznHAbJQ}*Ht%bo-=ix#W& zFl$V)aYAWFVN3kev`o5m^C~QtKi&t2NQ0ZzkA7|nxrkQ4$)cCqbZWBUO=@Cdr#dt` zN$kEWEwQtO(}`Vtt6gG3rv&nb%PtZ+^~Lmr&K^qd%;FNdpG(V!v@r72Fl8I&!*o(! zq{Q*IghEM|IiXPFrY977sJfF3X&KXqwK9(4WvakK2%g9f!9jf*fFo(32jWzCIT+jf3Vn5wj!`0-c&7qQ$x4^@v6qnh=i8K zUj*NZV|xcY5PGG2a&P}o&fnrLAPY_Rm~)rFWBKR3HQqiP$$1~4soK6Eeq$zEm@Vx+ zc+SIx*+WnD#5hsbr6#k0rM0{FTpYQ5v*=q~u^qUAm2H4~C)?X|C68Y^bLTP9#i#AG z*xl0NyJjLztNaUhG3~+e;@4s3SP<&gjj+u0V&g14>NY7}ecWWcLfaFcI39Wmkv-e< z*LO!dlgk$mC-9c?;5O^Gp(oSWA1RlxZ0QjsUD+gyEi&QWJw>Nc$Ku5LCB7JZ>X&WC z3-I>n4%{Mr{pc@Q@1Dt)Yu2+K+*JP%dftq!P8(|vWlD{ek#20*tXcOn;@OzP3qQOb zD>hnNlPBvf*ZIklnb(0_A*r^(6qf>O1^7%q8y1|}YswjrDrME9b^|PFgeDy;Sc?6L zm~(u;<6fFr9=+s@WLXSAq7`0QI%ne}1u z(BQ$iGON9sXxpbSKybk1cu3nwW?+$;QoogE9_o*3vJGoslLo& zsnn)LudfKdHysd27aij-!3e)MjSJyD8^lOgX}S+RD~bzQZ{?*Q#9eUlTfz6HaRC%C zV>F;wlHu2|KoLO!dSCFJX?${uesPyxby#v1D_03EP!GGzB9|)lkG&qqSQID@h4hr3 zcs6c@E2rgT$k7Ta$PqEFqZckUqed1SUgN>pi z$%P-2CjGvcT#{`{1+i}&<@tW{H6h{0q;aLxJ`ajf6)hBU4=z`^C~e#mqtyNBcbo$=OWbU2=?lv>2B z#f+ZdJby@DTzaZ}yv0?_G4Th%cXQ9oVbf6&yrM_5odfYQ%sG(HF9ZYqDQ))8_d+l6 zGNvixS@J~B_yeM6qV+P;W~%=FK3ClQ#CUvlb8E5#{fsX^rWe$A2Va5>`GT`W zN>o&ljK{i_x!$NbHnvMx(g=)=#Lt3>WoY3?hH^EUdn$U$CX)1&L))fP#Sv2Cbj#en z+B_wWR4OjqIuL;l*hsOvfliT-5$|UlA<+_gHdcv>tF~8pqO(?%b)RB*wH1+>O1YR< zalk6fPCZj;3UXvJ6+^JjGZo`o6(c6btD1vAuwHPQ)=!HsJA2#UmXpuQ|MeQbm7X{w z;QI8KXkv;X8_R+xeoX8}1#v?luOV+t5I3aC62y)1VuQFjMA_h=R&RSz!h-lu#zi&r zv-)1)Y%UY*UGn&zA(Gc+ZhT}jB+L}qj1l7_n>k2ziKbjnn&rcAk_6Y`#VmuDM!Sz! zfvxzdEF_+kKMh~OE*cd7Buq@CeMebFD7JAUdF1_lQQG07H#} z)o<@Hj3Jm8)VWI!rZu8V59S~@u=Jo$hpN&;9$;eUu^!{W!qNsr38e?Aahbcs0%75jmmwU}FW8rcQMeeJ(ylGZZIz zJwX*H=dsIq;#=d3+2x65rcC8~;~}iY(0K|$Kx=hG0#DQi?{oHqy23!ESYE%zvXs~` zr4fXcaKfHw#ZN(+Iy#lNf1%u`xuM@YO1gRR4Z_* z!ooLOQ(}01G_j%r?}W#$j-@L)nMsn0iYS*3F6|n;S_&M9Bz+W+AafG_2_LLHkdfrV zj-GND4IhuIHL9hu42KmS($yAKWmKYg5dqmqPzhwhu9PX>8-HkMqQlHZl2le@B1wQv z&O^Q@4lS@`SCQ7rLvn@x$U=@Sl+H@xPi7+X5G9D$qjTmV>Pw{^qVf=Xh}5GGAZf(J z3?&3C2Jv+gwUVgvVGEK5hyZP=;k8C?hKuvV82GEH!u^H%#(v+eqfft@ck3wdq?2^6 zZp`GzyHZ(E!QGxdn(rtr14A%ra`3&u4E!?Wa~o>bTbM5CN$Q(N+Pzlu)QoJ-BlP4(YvLyCuwAjR6 z4^WmH`D<;Xc=8=_=UeLY_wA2?8of`izBZ*3zsCO0;z}q@VFowVQbjCOF_qBf_aXMl zH8Gn|Q&NaBh%>g)YfVN@Mj!L&4_z7U~Z_ zm%tmW>IheE)Vygj~rI5<)L-?W7I1_4r8I*daFS@BbeY0?9G zR|Yfi%P@oX+sOSf#sdMA^Sx`lzgMp$2PJJ{bg+2{9Ugn~fdJ|qi&iyh&J!w3NrQ_D zH}=3i@WDH&crw+iO7M*V-e_|TTZHcpVvftZb3%9CxbOEBmz1ARUtu8G&br1VCBX%5 zGx>p9QNkb=kqsGYtuFQ~3t4yjNSC_bxxzD4$`F5>VSBb%O*MI=vM|izziM(}J%~tB zh_ZBIky(vXc+6r{ky#Dl-KRm38RI@s+cKp>5lISBQmKNKKITsHllaoI5~-{x>SM_! z3lY(m_+no)uT)vDFZqd?QEBnbviyDlN)x$GSke;1+FD`YCUB!z2?kSiG-`*;nyYj27E$mFtkxWA%2QI+4|~rtK92Wq4scYqWnPo z664jrc9!?-X7ekr$e3@DI>Qy=jM9cZYmxK}*->@lO%{_skc zjM*@%7s`+`E5{43jMV9f_#ESj-L}h*TK4rpNp^LYrp)B1hPq2s>k!>O!TgV&qUwBI zRL!RZU1o&-(%{Rx(e~3`_P9D4dAoS);4Q495%y=z*%Y6wo}UUjw)oLpRl&jm3gB{y zZbgR9tqu0Y59Qo*XfbsnpI*MqRuxp5^sZZ20-rD9YxQJ|_OG&HV7L)>*)n(az2Y-cBBEg(I|D9K>(rs&%SaseC(6NkWYa+8XAmm8v1Vj-Dge z{`s5rkz>4MV{RSl$EF16!UPF4NW(hU? zm&$%$u2j|*Q>U`=ZNS-5SS6Zwed#uZRo|{E(0Sa_9yB$D{q0=Io)$~q^^s>I>su`{ zC4EX{uWM+y6RRa%X%*R$J|${LNuL@&U(%>zSN&hhqw}H;T{O>O zyAs;#qGsU`AqSJ`llZw@iII}iQ!LDr7^yGhOpNrPbz-FIu$B{IOm)F&r6THPvP%Et zXn!)=y*zFe>W{S|ihXn1_7jnG@~*MWNhYeKxGCN?cqFx?h}qW;@S(xX-CJA3i>>Q3%}GM3X6xGJf>FFV+1;949v>u? zy_&<KEcK>}?G@vHbsE>;sReePedy)Tc8yUSAlWeK0(0Sr)NY;;lrQA6soBE=R1o_!!9jIwV&LO?9912MEv1#@%HRKeE$0G zXlHWy;^E}*^5cWstlthhS~@#O{UEjMK-%;&A&!VsKOa!?0se4?8sGQ3>^RnP^6QY(}Ki5dCm{Ys{<%ooakarepVv7c1giJ$~1a4N@xxb5zY(po8)l++PGNlfJrghE*85K5cyZs(wXvSmyOH!!X%$h_e5EQO)NlP ze8;sBl#0(!9(!!OKlUFSWawNgP8h^b_QWVdc1DS;`2xfOPFyPIkRHUh&xpP5g!Z1r zVN8Oq?LF+?I&yW2BgsD97yqC;TO3>-ZI9%|gTpuH&u#0~Y1VgJ__1@mW%F11;%(ij z07#5gn^5O!?jRmZ^#<2Y>4ZJ;JDqA1+MJ4(qDO}{_;^E%a`|2afF5Ooh zYzGxxBy}AwS_RX(Zfoln6T=RN2Pa{qc~a-%nNB?x2oY5iT-R0`PCKS5%rlft(S(a@ zaykr?&MvS+&nx&@9oo69Kr*x~UE8Mci@)o1=_r{$Q*!lpbQiC^>0=I}`)TpVY1cu? z%@dNW*M4r2-n(?gf~Ggc?6y(c2-ZZz7kgc1b?UB~lm;iRL4%W3E3x>Io_II|Hg?9N z-O_ouO`7fKZvQ>4Gf(QxkkX*KA&397V?}{`M7(+Mrd$*_LkJfAO}u+XJe1W$isJVA z=sdTWgooai(z50^67OJjI;Ilu$!!Un(w5!I;rDnLwMWBWEvn&kbWYh{d|N*Y0o0gb z#T%9sC1%mBH?5@t;kgNTv@}N@SL?w;U7M1*-NRNUZLK36Sd_IJrUT}I^PkRS4(1kB ztuq(vV42EN+G^sVqNh#C38>n+u{jn{k4?E3%#FG(Ep19OcA?L7`;X2f-7M)>K}|=? zK{MBWR8yprZpQJ8TM-FmUOzLfh~&`B$awv^<5{RdTAvo*tVuu1yRRvi!eQPNt)QLd zm9nCo zw52481BQ<9m2zhsCVYtXF#miyu866xd zKMQM5e6}av%gJZu|9XwzN`IUY|Bdm=UfA1URjxpvrXRg74Od!%EP|8w z8OBrV;i)SoQ}D6E@K~V6h^iUSqov8om-;Ucm<;TPa z`-zKe>AcQ1B7^!0o>7_uYFue`*|+M1sfd863_^e_$Pk1)Qt7%Ayk|b1tDomgRD?!i zIP5A_iBI=)Ox>VNzud)=5QsP+u#i?tboaIJZu+TXG@Ivj021}~|O4hJU&#~*y{ zo9>}5VK#P9a7}51DCI=m_f6Fcun|KV>Abm057)TSxRj~Ww{rh^6krft0bt~NxvvaO z@nF3Dm~3u@*TwT`S1kti)4(H$wq&p_pVjreCwxqNpeOF(1IM12hu~%=-Nez?@!7AC z3H#sM7}i%59e?Q^r{SrBr{TX_;lFe6pZK9xiI-12Cwqvn2I3OoC%<*z^@&!ok^L*$ z?8}@~{e|NBXP*a;y_T6*4}Q26{ySIfWLp%9N;wYU@b2L1-obctv9jU)yHAey4<8!u z?_4_^6&JJ(p6C40aGB-y#@KULVL;Lg>bcK|PqkXs{F6{$9^8CMPB8osG;N~Vd)hEE z{K7fvBn@2H79PSbj9m6$T`DwbchbVkYHfjA2LPF{jSh>w*limlxT>W3f4y;`&06K3 zsItQFOiO=RD7NgQt#8_Vs@Ocp-&`-(HU*NGJFvJ@yx{DhgoV;tp*jJpe~XPmIT@dV zVBK%1C|J}f$1)SI`HX69<^_?lj<=TF!F zS%n#Dbr2uyiSy~wfdT7RsWSv8IgxWP8Sw+XHbi8PhaDX5PqxNen+Mx)8D?=a-sLBx ztbS^vUDYC#*ApLUb#THHPOB0>Q0&1GSzU{}bK?c-_CfKCDll0NP$VEN4keg!E=V;} za{5(lbF9RV9~ZAn@Le;-02^lY__XsVC?TIq!k;>i$+3}2M4H#2GW7Ij+1noaK<0h@#EjahrX*V+M-tRR zJ!&Iq*N=aypSls!uWsbobtbNdUVkzl)S&81{~vkp9&A@u)s3FD_vzEgOwx33Ua5sw zQkE8d4_b#LG>=@A^hxM+YoBzJ6KUv%KJL!mNpEvrz0XeQ@K?MID4;yPmNzZRvk(-4 zs$0Spp@1rZvK0RTEI}?5iVBtwt}jsgjyczwuf67bj1cgQVSHdreZL_hJ;SY}qt?p31%&+wzRMXTaE<}6lfWdng@sygO*o8qOc zgF$5(PKS=G33fx|KcZTjj;(`cbZN62XY*k-DZcp)6^;|W_pFL+jcjey3o$CPhaxJn z5M3iGVG=6XxmWZ9<4)*_w?|n2jpPc+RR-(}<0y?iT0U}JLTSiUh zDuR~(zcUJTW2Urm1jyDy5JhyKt|5i0nHsD^lC}Cl;QvMN;h+#aYx&8bWNUTn>|~?= zU=|ak9VVQH$Zy9t%bF5MKLRz@Vl0dSIGT0UXjR!jtcT-Mqs9ge=bjudT!m(K^oc-o znVH7pQzQiI*W)7+XF{EjY0X6YBfxQ^k^nUq%h)Nh8V(OHFHS~~s2sDnGCvxV{9{3$ z8+m%KT~!sdAbLe(xy?FP^SF%imvZC5ZlAViV$m|%mLXO(=REOlf z5Z?5BRBtQVkf8Jx@+^sntR`y}kZbhu%FiWNC8?0I3@l6)T?^t+_cazLtsq~VKq1IYLHd$6 z738NQw+Qp=s*Py@6uWrvmiSiOc2Ravihu?Rx)@`hx{MOn#G4O=lhy6)`A5WjG~G-?YKG_!f5GxUCDB4q>4EV zFY~l5f-#~5qo0LAVbV#Ehd~6DiMZRr@=LeMzya{b8!*xyx&Fv?3`WW-z4Q-P!lHmE z;qlJvQ zOnWM7)T=2skUgT8u|{uO%=u4O`W21)G#E2^S9}S<#v{-jelf*_(0}$4+`8X}i}$zp zFAc{h;m*;Lyx~vv@L+V=i7Rb=2(ACB?>GHQiuA+@sSvFjG5u*Z75YEpQz4x&nF@I> zt}j|!Dk({iAi@|)v63F~6()mdhot}d+QB@K(6@@{q(=|{KUC$7+@p+>rM)07uoK8@fozsbxfYs4Co&04s9Nipp!}&%2 zD*kvWNT5H|XpE-@^);#>9_<09v!n3s$=LBf;&WHAPmmMcXO_4augciGRghpbi|oe5 z594b^f+8ODGcr}N7GrRO((O1Pg|P0 zd@Q$nuoAbZ_u22U&6D4WKT9efsyHkDK@~WLXNc}22};&%{LK~m|A>!*Joq(_mRN0S zz!b!?TO+Fei!39>mS7~`*(Swb%UHA5$YMvpfLe(;uO%6nf?9jDIslfiGB5=-Sq*We zV+!gOrNC14NadIb2lhE0I+Mor+>Hq?Dk*e0- zknCaOENI`lG6eclmftHpGd_GDQ#L^hhp?B@u#G1hYNldksCh;RnmnAvcfe{@T-UPW zCvx=Sb<2Cj+?LMeg-{$$LB|QX@0yG4nwpV1Rlvx4Y>)ly+gXgr=*j1NuL6cqXDrRy zdS@BNp|hhzWGuIhG#Jr;6`m`~)y|DF-VhTfaI7e~PE6*6RXq*7u z_aehKO#@^2$z5K#)2wTQCOwa6k7F#ww$JK!?Q} zF2Zz(Mn(*Yig&V6#7em;#p>NE0R)(P;71s*MJg`KQ%JmN!!1thyPr*YagSy}>V1y9 zxJMDRTEwPf^fH7vO`VL(5wX{S-tCYtc08{jI+aYJ&uG%N$eiKBl}}a>7gY9g`%I3` z8J%Cdd zUZJp&+TgH|CO3#yD|BWGKJes5db5x0IkKCkHwOS8T#*0HI^`&QT9J4f(@a@|+fyHf z^Niv${IHVsp9!)t5=UbaJ$gUs@97GxYPUz}HwsAI9w&LBY)ttv00oOVY()FCHP z2o=s%W>+}e1cL)~aEAX^be#oCHTl%88u+tr^7&5l>A0uY)xgZ{$`TZ87g;9$ok@jy ziFn9ul{~+9hyNirr;StQt2?9LVNVE?w`LWO=(?<%rWKFu$zUTPbtZ#v%yzDlHkqpC zm$veXM}lWYr(#P{GsC82L~huWD@hAZOHV(Ubt)EYT;a)Sqd4$%BcRS;DChbc8-2LS z-5*E;^m=siQQy2f!|shgn3 zX#6dtZsLPXI60}CzEbW@x=&=KK9UyR_sckFnM^q+1ro#=Wh^70!#lN_0txysQy{Tu zW(p+NVK>zDbG@&aCU{Feo>jP&M9AC$Tfs8Qw}L#Qf~gsVj8n0{(JP+E%h!B9D5RPJmI@w5f8}-^II8kbx>!+_$+~yVDJpW$eZds{!E`XV$w(d&ji>Kw%HqaL0;XBuT7g{KHYS0p8n~Xh7!!i+WUK+cnn@nN~Lv` zgI5?w^rghL$))WfdoIxxy6Qx|OWr+CR%sbC<~D&@7{@Pd_BTdnRwpBP+_<+-`)i4G z7+`J-D-nY13X879x{;ZYRB2od+nHrm^g~tq`k#FB8GOolR|oe$;+ygtysbtVHKU0mic|HXt=JP6z%(@7S6(H*o6veUYwYRK`mi@o z|HMjfIK!grv}OXiN%{NK^zqlz0DG8>w%3O{EIkpm!q=nK3Xg4V?ufIilK&RAb0mxl zuNzr=m&vZ8H@7%HH+S?nf$A=jIoD4v(cW->o%)<;y*IO-X$4YN#n>kG-4Druhwj}k zT07jr>H`=0{F`s@91oza#j?@g7Y-ud9KwV8a>+)*`IwfZ;u$dvol>-fzA zwx_yp-3}MALLKKG4>SH4MIhV3W)`6S>6P32O5aWN@d*u)|K{?1HV96b^zdJ64jKLdSYDJE6$rYxs7}KnZOs>eUUkh~=-!mldJo)RFVMrb7y%Fs@(&e~ z+^u(i@CuZB(2^r&Z4CRHE#hdtop85=b4Vdg!O+|-Ff@0PlfB!sFf_^+^6)c%;hU@$ zBmse)IUbkl(QO!TGtpNu%#7!X*j;Rs_4X2PLB5H?T^%dq9E zOmex`s4G~GBfS_hO;&|_vHKwnm(pe^AW(mczIN3NT@qQry^&Jy?^^5qGgfEvgM=F# z=FQeP7xxg{7%v__Wxz~Wg*uu^7^7XJaF7dm%{+0x1{1_C4ahQ$Ye7iND@+SQ!b0&Z z2=r7~7KF@);aCKSpiRqzk1(8yBO&OJzs+bfN^csBUuko{%``y|oaFl%5cI}gC2j@V zn$iIn4I)*%N}H^AL{I|PERwssaznhSuHaPAtQsMD)CS-(wVA{M@<9|GZ`5|9a}#~m zg#JKhIw(u(gBE`{Yok#q@cJo&OIHvn2&%-R^Z+B*sE{WSZHYG<2GrkFkEF`DILTU$Rh^W!2(c^NT+tD1hPSov5p~k% zlRJ7VNp(`)bIHL~)XC6%s_Qoy;g84%3-X+>{>xyzwG9s>FX>FT%y7S{ywO!HaUlo5N1PxC@}KEu z*~o)axkIrG2|6Q{UUEa9IpEa}0_vXnWsoU%>HoXae-AfzhWy>MlzmFuko)xS>~)>w zohhh`Earqx=?qhOKv4ueBW^3?l7BDA(L~o)tGWu5NaYM;E|!mAO>^BYb;ROszlmVZg!M9od9%_$H3!I-hoK*_e`c(W{Y9VH26zK7@NKjk-n z^_4OKmArxM5${2+(I?yHylLqE=388FJGK+7*cMiRV)OUy%$pf)PKM)+;b7FC43`FM z6&j!pXX`+N2hNDLpFLmF{meM z_D;PE;wqZbIk!Gsshh5-Q%$1EsVpc_*vL&K?Suoy<4XhFOoPO)auA6OIV}7P3^EUP z67&XR#8~_N^aZTEeTq|0-ZH>0FM!_7aHE!HCW9~UWSVs2^pnRWdIn0=C)L2XX@n+* z(BaJcDoB?{lHf?d9OKf1sryy!NAA}jHKF^Hw^!v{l(!ILnisyGU=+b;BtN%)-QWFj z6KFUy<#DY#IiW_7X*2P&B`qak&bxB^T5)yODakY)S4&Ew&}DS?wnA%yLMflBU@Hj< zHNA!m%xNhJj1^Ca9x4_5I+w3e7ih960b3!*C`0M0EmaCwhTOkS~U@OZd25fZ& zcEf-lcBZbdSa?J}oD{^=ugcN8SjytNGh|RYg&?D{Kp^w_VyC_cZ6j%cOd${_kSPl$ z2QqbyddRLteZ0&gFrO*G^PTQ*-e1Kr`V%V&?&3d{KN(sIpuge8JG-?;2t^3muP_<) zgt`WQMT0j@n^iilmf)?>Wwd3);H`WrIe1U6ksG|z>j|>4Q`Hkt?9tLcHCvP4W`ucV z)DyVU@+2i8*HMJFl3q`!2{CGESTwOVm(tP`Wm_2OiLzW`dZMnt(i2M_BGf`pTj&+4 zgq&RY>Plig;ULr#Q0ynhB+@Gi(j;j0BwtvJZK)LnmswmD6@{8W5%cAfilIs&93zIL7=51`7-XaFZtB_fFj!)Y@OPetzBdZ~;k{)^+4)@WHEBTej|W~;B)rMVqawjIm=mk$g$GwgyRiKMglI{(7?b#GNiRtbed%Pk=|BGOE?4_MqNfsIK+pl z>PLba=_^3Oafn>E23IS#FHF{U=!tKqcYyrpk$t^AyB_8LQ%wDpn`V~+krHBrpaFb9 za^*w#4{<68$5H?R^75w>aCU$K&i2#7VS>?%eZssyIlbO=z3cLG!>)G$Tseo|Vzgv% zbYq-KiR1d-xyIP1jsk2iXJ_!aUKjo-gFy&29L)=0G~{sV-sH}_VTwN8lA}db3k^zYn{QsgXc$GlyD{t7mgAC0iGgHahkIHn`Gtp*r-*H9=-T%m?%O7rS4tLhpN9Tw76+)~e@pV|#d9PcC#$#}9IWiLbswegMZna17-uq}k z*jiI_{+q*eyoOFkWZz6rQjQyiE?QQH>%%jX-a_@iQ)+C}CNWz(6f%yn!;j@viFeGN17=RKL$ZCf_CLmkq~TwfkTvk+uHL+5|3FZ1FMqt}Dh-xcPwRLjno)dXmss z14Ge$^PIv+s$Ge1(tg-tOH+dIPTvV!7QwB2wnL6(yA{Z%F&#@zqD##cM+0PP<#*0- zfX_{=4`CkR&6mH;S{37({v&^%hpT;Nv#Z2V+ZY+AT?7$PGp|rAmZ7KD!zqPgv4Zgo z28CivGr(CD7W_-Zl?tLn^(m`7ES5_y56cy5U|>+gM6%%@JESp~qNIR>AlGlhI0={Qr@t?J&1yp642&2GvP z-YwfR=ui41eMqQ#^Cxe;>)4{>K?2DD@|~n(x#P}tlfd$DD0~#_GWA09B5^_f z#Xq0kAll|mX=+jZYx!K`#`+UTZYQIy&84dRzI1#4(r|ndq_ZTaEHT7y*<_Yf&A1EH z3#^{q-1i)L$lw5;Sv!En_iswUS~YnjcQK3ER2^@^ut2`-e;6W2Zq(FkIK&EXax}R3 z%<`wxH3PmN+2>327RcxP@9!mSjMt+LTQko@X?z0|+%FcfQSu0qe4rCz=w?2VKkazk z^4?wD-h8jKt2olRyr4E`x{e#!1$i$Ln|1nt&HkI?UHz!(qO(K3)gkwJPEQR{nf#NW z)0mvuI-kBv>v5m1dZE5cC<$xkc1PdfN3TUv#equ(dz*ncUAe+SV7mQr|07XxJZUsP ze1O5YciVIZWQ>ANLTwIqJdb9t|2bXx~8uT=NQ5EQs8=k{1s;A&Z~h zzmwwkNAAz~U#E+cy5Y!*&N(j>J&t7N;#Ickr&~W0vEnk#>(qG*JT4;uo{KpEJO?~K z9K^Y4I%Nith{m4IE=a)h!?Bb`J6I(z~ALPODc1`1tIncvPe(3U&6ArMK zW}Gk2VkE|jJdG9{gMfGE$mEOumt$j{j&Ss{?Sai}FlCsJhV6l-0&?tiy#?~uhC1o9 z>N+VJ3uNgKSm1k5Hd%JkWt|iKjhB#ndM^s;Zv=l;-3$F<_3V;2z>{|)mHMcmQG*S| zq|rP)wBXkv??fv2=~T*eJ-;N0Rq7KI8r@Z^r9Sa##8RKWraW7wrRaZ#RDBRst?&fB zQJ8sFc_cRIpX+aI^x-|={y>_%-vGaaRK65@5@|I`xj`%otSV8epJGL$hGV4EpG3@d zm``gY)VQpr=`>o3E}gK{LcNaT%dyER4EL}h`vEUSI1&?IjO04-&us1GSMUSqajqlimc-;9*9 z>Qu^nNQ(&^0s?Btq@{q+mO-mc{Z=&#d%QVZv3U1>p?(XF(as)|n|rmFgy zV8mH>cE5_0)=8>(wG`!`YH|`d>75|hs1;#FQCC1ZK(Nv~L1{bbomd_%y^|}l+kgB% znO+L@`_Hc+mGUJyG6!l;IZ^uLf@Gs|LM9Y+lhlTVCZ9k`f&`g{0w78fBtC$c1j#jP zY_ttAZq{28aIeWx4dWH}i~aGazjqc3S%+NCI$l}`9ldPGGY~ac+f;4w zeo;%jm%O-lcV;coLKJf|%{R?EPwa&1*X3 zE^LdbyDZ#q%+4h|fuAt;E$-lhun?r7H%1>uP;^gkQ&`TJ3ra<3W2up(QXc2!h4(IGFw*07cSgQJZL8*-M@{{-&wgI)Y&y5>Xg+WLpGU0W{O6Qt0W8S1e^eaXAK zB%>^7Hq*1~TU+D7=wf{~JwCeFmWM%)m9>6dZBqKv77nP5!a7RpP!t;LVHtt$BXFLtfmPCIW7*8X?v;kMRRA=dEav`d#Orx5-Mlb3`xb-B*TB87YsaHt$v^ zTiYkaY$^dwotaWI;cks^!A5ssw&_Zs=K5uORSu62#rXP z)k^x}W619F!{PSYV6*{=_KkBC^w(ZkY@GAjQ#uN!YEQMmw3B&{C=mNVJTrV`c#@&W7ps(n-Osq)T3fECn5XhAm+4TQvdoL`63# zBJe7b1&kU+fa@vyt6;qxQ8b&KW+TlWEiekRk_!fu*=#t`#F7Le%-Z3AhAz8O90)|y ze6NceyERl9LfPM_UHHxkqhK{nYK0`5jr{EqNJo)DeKljANi)bc5oEL3Z`0xf_5EGl zoFR>72Gq#0HX=&S^cgd;72Z#zB{Pc3acHqWn#aKMNPE6m zxwH`6tUaG2GSjPK2)G5gNdxk$1zUX&KBtJhK}Nln||<`Oy)|;B*amhcn~w# zi2Px~#0{=iRMd#@$iV4cnJZVk=yb=gP4K)$+ zqG6CB@sMMFr?I&NJl#E9s-LUO)Qo68O<0~`|DtCd#n}BbxG@p zf9!REj`@tIgl)6=vZr|aG=4=TVQ1ctpLU)Z7Jzbz7gFZ}e5v+J_W@8)JR95@uAg1v_H`J+9O|{C)@YmVvrRkPEPG4V@P_Ujd3!;YgAGRC zmH~sN<>2;8<^1AL@au3tvRu}~f98Me`U5b}|1$iy7ydg6|B**x5-bySHQ(g;-*_dr zXoL5GYZ+xtgjFD6i+6iUV=U|A$zzY)0B@WgppOC<@WWpC?Q0iaYuSg=>=mbcj*1F{@@qa9t5w);7jx`m#7P* z?mqk^z_w!I4mkQL{`l%NY$-LtI*7+o$)On4;S3Voq70fNaR2${ICrlzh`B9}-b3*A z)YD(C|0C}x$gynVKzQ|tCVPzGUp)$6R&N`+J=I%_2z0PPxppSw(O@`O+gXQcl61!0 zne&`fGeR+#wt~DTruPv>n`GpZd9Z%M^5ev$KJwp>*8ejFD6)tBf@T}xorOR$54o8b zLEQ)D$!$6MrwO>XceRrDDiwE`Z*>u1nJ~^pfMvoKgab3abQiB9yM~*CwT=GH`Ce!L zoD)3*9_hBt!4M*Y91GEf!{YTj_m4-L=ht9N(T-dRuvYgC&y6!a-* z_5Th2adY@m{daLSdF>#wJAg6B;T$v^ zv_V1mxjA+xnkoo~r92BW$}-}xD5%8;m|m?)5bVU8TY1SZyU2_YDuuaBX;vr$X?reG z544&MfiKhJoSM5pGcAUt@qIEG6Jy+@NhVv2W0cE^NO<*?CZ< zNueWrzd}#4>Z>*8v{6h&sIMN1sINjK4X=4K*@<#0hs}zAQ0uEWJ|g0Rnwho+d!6o3 zEuPcDkDsr;byHO!=!&wEh)dg?d$%ElT?2Y0mT1+oBaqMSeeKIAy~xFj|a3 z%?6u75ljTW!Avrre3A9E+rLHk-4s_Q9BSoGGuR@1bLqp1Xv}DUrGl&XKJQ_fXk2b?$Tkb{NiML3jo7_XYVG>Mw%;e@szo_f*9w2v)J!YO}W0 z{4HZVWVjob`-MhfB=kbtFPK(yY7*}gK8ZIaddt^CJ9F*`;hoG!r<#Q};{o_xjx>O} z%ID@R0wRq)Jo#DVJUj^yMrT*y%5N0-9lUaxvqv#IHrGjVwap<0q?V$|-zmt^Jj>-- zNe=PA$<}T5FAH*e+)mkxS{!#?XVW(1ElON)y$LP7EzG6fGw4tHYiGB{8~sUdp*Mf> z*1L|8Hx}eMaT^q`OlZ*(yCeYjydrf&r%RRs?3B@W>uI5a{uf@X_{~nNN(jn~nV_vO zkR_HflX*O9rFFV~t3!^YDix0-*$$?RG+eioBHUzHOXB6xVMGwvNC-2SU1q6S1Xs+t zJONcnra*SG@|7k0h6PI^g)EFF4TrlQZy-N=`)W}QWt=g zVpBKqaia!CBq?F>I6id~1m>o0Vj*C2Z0aU|mkkT&5p}A7=MKsBT^<^e$+x19j)DUE zLyenQ4CzZ$0Xx>}r?a3GOL$#$ekwKzBBGnjkrd-o8JU|30*uxGxv8ZD=&FCs3EmNy zr5~AY3xarjPtQihQjEE=3L$^H8rdy2HDv=929eiBC*p>zbLtvaWFQb*;DI0rG9=J0 z5c)!qVYDK`Z(OnR2)~l1=mIb#vfb=j}ABjsXLF zXqh_FOl|Ce{5Z1;i@;ZDuF;@um-Pg6F`q>W5&Yz>iEV57~XWBVdz;RbtTjM3Ci#JBDn8k$|h(f5NbE2 zVH;02R6xbbP{WKN86dFkup<^WHSGXVRdHQQgrCT!ofO8eXc`ZNDnbq9BLSK#%imr`-LW?5|Oc7(bBqN@Y;^s&4#fN3cFR!++ZoK zi7Bk~?M#?y&kRErjnf_yQ$54G0)@jmM^9>CTXvIck6c618VcZ0Lwb-``tspG0%T9c zpdWX)WyZm`HT*#ZA|r-G#cR+go`iChTYnJ{-l0zHM%RT&Q{>>GWvH-afj4cq#p(BN z{IUkbe}+q-S&(|u!XNi2pfxlqcj%m=3?WWahaZ=}*dZAPtJ z;INPaED)_02E1LjRv{bLc4)q{)7wY(9NA6Nn*)H~l~;lEiZ_Ahb#cf5o#4ksQ>WM9 zH`KS%a=cE{Fk`ZU&t_%7ilQ)ZM|BE7CPyl5{mWi~hmxzbX zR>|{wclaMV^U7bPn|_BraMtnbCxKzWnI4(-g05vW-z65{f&)2T(s^- z)m>pKU&g6k&@YlpjKzQ%Wk$tJ5M|U&S1s&7V5ECzcD}$Au#r|`FpnmcXwHz^#&iY2 zmuZsE;do@w1fEbvU*tKp^T2r0Y+HonJ3M03nOZ2kl*K=cQR#0!Z_|b?cF~kj-f>#!U z!IWBR;V06}6i!^2XeEXFwX76Qr%k4C={Mc-Qa3@5F+8@Ax`_{Fr*8U6xs}CFWTifm z7T)*EIB1znIVS}Y#2IBAmgTsDzyXq%0txysQy{TuW(p+Naa}j_ra&Lh${i#TGIzjM zu#EDpAkU~^!Cgn&eI*#Mn2|gP;&77(v21qoAXlodb6NOcNh;{uOzzhK3}K#ES}*p; zqyFZZp#d*rFL}fbj{9ChytI%vdfAY75APM_rqkq9@6M!hljSSS&U+j!w0rqlHi@k0 z=q*W$kjQsCo|lS(r2%@;>)t#Vl()K&RT2O-y`AQmwfI8G^qRGvfFK-?{fO2=cd=zL zk$*zt_Quwrx9gVmt*!CO)}VC1+z}%}RG!yI(boCu5-x%H$Sd@`7y~r3=uzeHW@Xdn zr`)Cn3)-)p73BNAH`0O?1Kz?s#~zv+PiD3Irbn9sS=c-p_O1uR^-2HA?RrYywq5^8 z7b~hVS@*6vMMds;uqH)+FdYmoc}&m_SJKgr9&Lw;fA-BS4QBo=qxJPA2CnB?tWD6d zXjI#=aItCN1$lKhzTaf_)pXOndHQFG>-|0VWD}&q6Xo{9Bs^y}qK_J`O)hN@*~^Bm z@bT_>&(nXwdp&+>v%fJqvpN~UQ^CD`+Fw0Sg#~|uxh)ouer$_6<-aP#iE4|?ii>}j}3FIcV2sy`TXoyZm+v|YoR9;gKx$*FNv|8b@ zt<4>Ac6~grmhD|8yNcf2;{4p)(c@~ttgwyiqBro7%(;H~h~fnM>(u8&>%E!vOe>H| ztc-1b5Sy?F^4@~DpFwhju+;}H^vA=&@zL3{!||})GVl8f@`BW~#YLxVk=hH=aTvV6 z>5u_3rN+L8t$EstI6cq%Bq~X6Q}yo(^7F{7-Z|4>?~m(@wmf-n>hgN$gFl*Pa%6Ky z!yFmRqB@mIjl}{0+f&`QZikClp^kHpBc=TQjb=O8%mUOu1>4_O`fj3^N8%bPtCOwm z6LrBGUE{(L;{>uDY~MkVRrQsAY;l0@me^~8h(%-!P_nP%D;acqs~a7yNTQrcVDfWL zE6hHI>^?snZm$hS8<20_sLag-=n;0zZ=7AQRfxhtuw}sslhC#kj0M{35LqtHB~oz-|`+Y8LODWN!V-H1^I3UB1rDofQbp}tGCIT zW>0+%b+;N8X0{udlCK4|Kt9pMh=>{nCb`^e)Dme;u}v1oVOw4^pN?9P?9#gI z)3_Fd#Js|^AS5gl&w@Ztg=Im=E~$ymMlFIcEPtEPW|Z2{z^}Ba?#)Oa7$Co&0YPu< zRpM5XAtrK7&6?5y7;-@&klCo3J>~4sYZl4fUAZCNR9A2+XjY97Jvct4t{b-N?_J#f zAc~GRYP-?7iM*Xm8yEWfnkBZya3zmMrMwFz5m68+2&%-K3I*Pw zLEv%&fr1d)8TtglD@7wu-fS37f2bZwl{s}mj&?$?aC1e+lcw|o_lSb&^T{2(m862H z?!qJ^=vU~3tFFI~L_X+yvQkvb2IH-57y`J&`n*+f8v3%98SaP58(q~>zSgn+&csio z;2Qam?>S%(wY)*y7>0DHd`bwtuy;X>pKb#49o%n=8_A_jgC&;EbFj4?t`v8F*87^W zhaOo&+~1f8xj^QRbO5gBnLa_huPNJUFgEsV(5NiNd!Es>lUN_pxss%f{`!jVHD&wv zg+(T8UpkQTsUgOY0M&We_lmN8X96v8B&?~%T?}ozfYm^4aq@yiXp1M?eBKx)w2qx>a#6E!!bHb0*18T2RpwX<8}jsB#!(3?Mb>s`lO zvw9Zjm@)Tb%~M-)yj2#u&a|EI9NqN7@zmF`-Y46yKi*ZnPxflxL)cTqG+6-n#UhaGF)TafF(@!3k=ou(cpHv^i)jTvYgbruk zS3$Zwk_1Nr<`|bAOx>?)KQYk+x<7e)WvP3GM$cP_G0h9#PcRBNBl%8-`@;-}{JOvU z<0jB>X3FDQb#g)p#5ObWGLNLCB+PkNZeJ^|&N?NTrsHZ!Nff%;PDzwcRj`!=g_>R? zHzk3w;tA11rGj7Q@-^xLO*SQ9E94kuC~eMCrEoeXk`}NPqA&usvRq=oR##v*477_W zwsb}Cj(j*Nh^b$d#XT%#@!c6RD4jx(QCT36)o2lC8Q=gd1GpnvVjxop#0g}|g2{nQ zU89;gDY>NUkw1{Ze5M4?ce=lMe-+2*Ppl-ki~m&qq_J~jkUsqlFW%Ykd&>TWP$UQZ zdrk}9G+nK0@K-c=Pt$R=1aE~dqb(!X6I4;8))UkZBQ*rTO?YPP0; z(*k$H^hNE2Bt=mUaC_C^DQ}sT^m;WG zmZ#85MTv(zX$(pAgo98|K(QaG2mW}HUQyruxXdEB>+0uLdPSioP{e%M zh0Wtyaw0ZuCQ+4H-IT~f$@ORnfZ&Y!mS%_l^GEziXc& z5y@nhcjU8+2%~N~OCXX%mJEfyd7bDGnGQBmkR1Nd1SH#}jDQ;B)rMVqaE1&UA*cETvosd~-Ny#74$zlRJH?QGi|u=vOw| zQ6hy`L!>fk#zWqaWF+9C4GjzjIy?%=kXD(|X_DzjdRrwe;Sh8fojPK|AwE>KE)tx* zzQUDBI0Oyg1ClEr!heWUK{%EI2w1ulQ~Ln*OfY(}Pnh>-oL+C8CcJ1G;CcDE6t0{Q zlT|vX$Gy={U-2Qlo4!^8(@1mJA=>ly6{IC3__^kXkGxLA%|1o~qTs@m}DcQP^w9&Cgu)&$wQ!I0iE62#&#ZD9R zcKQ3RvZ3EbUX|3A5??mTBMQ$j(r$%zOp zC2#jXa_RDiTbsk3we`{Yp?-x>t4cZU=P$amN%A@@>a^?D3UyXhu*G9=ZaFd%{HiDQ z_-?gF@ZS4qK-gNA@yxEh|& zIK~b?mX{6PzICqGg+I!C#>Y_oKL40}m)5;(INqw=#|}d`^mo=KaJgcOkIDD=2X(xC z*D<5ALjs-;2_#hWBssALhNAoCIfao_yAt1|{jkNBrUc=gz7x1?*boN^W~Vbc^cU*L zXFKFrwp)RG8q=}lB)ZgGaWo*tWYfS4!lTk7$sx=my!rCCS*v0W@c)s2^g3MaGxr>F zsKij)7#XJlH-wDy3dLd>sw#(53dLdt;~7i^3R-Fl?z2}Ku2c{us!v(vVX+)zc2<(Z zDuYnhvoS;Da(YqtA3LNmnB;IhB_~1UnY62V@e0%%klG_zjHx9hEoW7X!MTl`=shoF z435DO={u#XhmF74Pd<=zoT=+pb??M>HM=QGIKsqUvX=3ed?)Ex-nNmvuIsH~gjZ}6knBB|mCYJK1i*aU`Xu;BBR8((#e+`B;-`>rf25Ne`;7l}x;UvDj;!dM^Fq<%NalxH z-B^6RhXu*R_adf0>bwPiT#kFxH{ki9a85Q+V_>Fk08JU<*uX;cn@67Q$k zQii$0;-TO9b z5XTi!8m-2egbBCkN;eMB4hxEQD6}?`HEagcRHhZrAbF<8Am0m=bN_~Ql zMt9X}sZV?wvDBxpnOo}nGo2{kTjX*!LTqD%KCSs9s3e-%3> zfbQ^AI-^T_YUhFRq*+3}N$%61S707vD^ANa7*~XhXL!e0My%1|AO&`qT4-VTQ;5e{ z)#gBN#Eno~23MIcS~4l^4x*eH{^R5`c^}gGftg8_i?JV1&=&=Eiyw1v5|stwvaI&4 zl4ahrK32*~Q6Exj*fr)iBc-f5l`>^bzkQ3nglas(n?*zd{t#G+OcUC7H1X_A0C~YUb6U(EecXCD7?LV*o{0dSjUy>tp zp!Sp#rB5zMHYz7%LN+`vA@iq4g9m$^r*#(}Lw27Z4!73^qm3Pa1kOp#`~*@GB*+v+ zB;+JWd;l>Cl54VSMb;$9!eSC|ugOsj;}!Rd{qd;3d1e?D+2h_=yqtBsv=BOa*^p-- z_lq*&z2wEcyEDs#mhUkS(tO6eqgT9avU<2nR%zZsC^E`uftv{`!ThBEJf>g^zY~@EYcfg)*YPw%%4G zDqED^6-cO>_u(BjQ;=`Wlb2v?@9|5U{f*I?)yZgb=KkJ3?XP7!?h-%AgB%i{*6hlk z-AZPcvQwq5kQ(-EsUtL>yrM&1gsovK{Rc-I7dGT9r1w ze<<6v<+43N3VoTO9y`>Ryt_*>%7SJyJ-fcOH6DyE)@Re>ql;~M81z_K>(|vLr9W-q zfZ8amqqGi1p|M`RWlQCPng%g?>+FBrLY?sRXnoj&QuO+;L|07C&en`I0I>`ch)x4Q zXfoPfA3|{*s_nd$PTuC9dlh{?^bVFfdR#rmTPN2MsQjbLV=}$~Pr+ZLrT_jq^>7*1 zWA>s!Ee z2ELZOvP(8wquJO`vq1n_7IxXXX!0(s!}}4ehmhAiv8embnf`j8kMb*Dh`?&SqgFd5 zZhZwTgNu~SHLI3%`JTPWLG9n{0Gp)!B_VVkN7IELyjY0DBTXBXJJ! z;!HCOD#+VWuDvw&FOgsfP%XK-W~UgS zCAqW^+^oHu{VdQSa8(Qew;(rZKz_C0IZ)#qM|?ThG-HX8FCZTd8y$d*3{1YuL7Qat zd|eJsYR?wA2%>RbpU%4K-^^eXj2EgJau&8?jub0zL2d;!PTquET}8un!GnRQGH77B zP*W6!=R);{(O9leb(=3-J)<7nh&I1R4m)fS>%+~Z4f#u7GZba@d&rA=cV(8cd_MqP9;-r9!um6jN)p=xjl9L^20EDyK4Pp%z$MtLZ#rp{MA*58>( zj`{I~pDxI)@}qIz(s#fi8iqX_j9xcp3ot#%_nMni2z2}UC{8VjcV!~>nVob@0(ng) zb0uXG;+Q_m=noSnZg91tqGr?>o6pua&wzH zCXj#^4TB7chaB@ejm;(C>F(iD{aj@p7o`0~aDn_z2P0C+2~zoj1XD_x=N)EDfK?I@ zQd=p^H7ts}sY8C2>RSx@yFprQL@FF9pKlCch98%v48=BLi+sBvC!-moYJVNfZI890 zxh|;y@sGVO&@rF!l(20!U-lGlpT@7KB<#%l@zc&T10ye^sMRN8e3QH1rOKTm-FP2m;}p2UClQ+`7U0`E!yCH;95pm6Jbt}@KYLZHa?y_ z_Q(yr&aMOW-Npib*bDz1C2&X$ul+DGCeV1>+1}b2t{txAciZO0;dpZAaJ+G0(x>Cj zdN(xwseF2fzAH;FZNqxp#kocNp%_Zt-wWhR)qZmu#t6Yx%NU zP^T#y^$V=!z_^M|MY@G5e~{DfpRHMQx0(@DHd;RDE$-2 zBWPUir#;|InLed3}SAJqxS-QVS|crbcv5rGaiDA&$pJQ@rKYdh;OO_I)-J9D0sYDOUj(^io8#PmMGXp@Y5GLPy{ zSbm(C)JOjN(fWU;0Ce`UAavPVAoGx$i4oL&V4mETqko!!dwW+ad9PA&m-$u~0hS39 za0FN;Y(Y3Mhxd&Fmprt19oaS99IS2hch19eRC5kMmK<_Kjrz9D!LXXO`~-6S&i&)j z=J_=kQ?w&jf^Oq|!*ioe{x8MiW68eZ`ub>lNBw_;f7~3tRR3LEO<(ZIa(Ktd zCkakn8xG--RtUv&2_Ks$KZoWLfPuPGz=(E;CdUOGF}y8YciLD~WRKm}4L-&V5qQTY z%MnKPlJG)SiK*&&<1L^d{M;P76HOI_!&07w8D$yRO$j~{k%v2ZsdJ-FlFNOBghw6? z^{2EgkZ_9Mh{+34)z?b1N_TWfz$- zLS;CoM0ZBhJldX%)B~+1P!-%npF3+yyg>D|ka`QJLdU>_+tI(q07+O%z`wXb+h}6n zpc7r#eu=X4ph}ZMM?`^%i$ANrT4PQd#Z-j)>Y<4GDn!!onm3c3D5rARtoR4DzKY`` zA}*+zX=|{TW`fPSsnhhek~*A7GHjeWP8~AkuR_uKS9lfbRav$u>653qbl}wQ#|)=z z_ytCbF{s&KQz(Loz&Dsl=94dKFOdeB`nTx5o8roZL#^Ct23w?WE`3-LjTsy)6&yok ze(|yB4n||=so8yFAsURrk?a}=Xk@c%a73^P%gWvngOD*3+Q{)E&qfXIQOHz=?nz%1 z!7Uz|^|>>2noKSwpbI=dd1^>fwYqh7veADqhXd035;#EKAKmOSNs#UtLy(Mp5$Hp+ zjvB*3@MtFsmWsSSItOZS%5bX5RXE@}RLsL6S7we)MWaxl)Wgy7hjXCLYg0Ob9mlL$ zh6@K=TO?PfsA~8)wA?op@1e3wcUmz3$-ghi&rp9E1o&f$QoE-rMnSNuQiWV?tNB~T zct{cY!<`&$uS7yGwEco|Yk-_PCw07qvL@CW6g^hXrhia4 zY-HwCMdTXuN!`*?^|QG7fD(qiS2LO3RJ5qqLvA2@L@#2k&mgegi`@E^iTfCcmKTv| ze2P3E+86{elwS@p-8+5e65MRxhRf}@_b&~{C%48MOY)jM)xd+XbtJB>`e+kzKJfmBoEQBtGe(l^~`%ttLIc6rJ?wbjhSAz5fxP-Uu3uz7Q$B5g%a!h;>8yZ?B1~ zifQT1H;UNwMtmQBsEQV;`57-(_+B1i-Uhz5bVn?kO&@$K=ZBWqobgJ{!rs47DM_H zRltt5`spkv#S&f@ou7(Lf{5rQb0o$1R7U2ef&im6WZ%<~Fof*?aO!xW*f2(61ArR5Lf-LyH% zkjRE>lObRu$SeuE$t+ls#0SfzHSz~BclB9}7IakkWdr+Tf>!y)MCl7h0IS{;eYlz; zU;I^tRmHjscAc|gNA36{J)oe%!c$?>l?9=z`{O}!j=Z9Lu#&W^ z;K}hlGs0s8L^0cei*vB#EPY^P*N@l`y}z;y-Iq`1W*(i5kwYN zg{>%2tFZh~+pIOt1t_>y!)4`ITBeRPQyY6AKhCVeBJj1zm)sgG%*>0_Vrm`^?TK8C z2Oixf!VtrhlD4%cG9H_mldn4{T*cR(M-0=HpIJ{+Hq4}rWTvel5({j!nRIMl#4OwZ z$0Xd^Y^N#Dlh62u;X0c~8?-!F9{_p5ydl+R8g13Lt_*tqeDe2v5#0ANWfQa#2(_Ei zu#G1hDxhLzsA0yC3@NZX?1+U;O*=qTRb1B+;U{uz;dRS<#i*1HwAcb?<**Sc?o=p( zDqv(i3kNGf|6GT>kRD>JIMguejMY|K_X|x(B_dVn_%U|q}41-qai;B#M0`tHaoXV__&yq2) z&qEDJ;Q{iaNA~sh?2>mqrvA!px=SEBXLt&QniZaqXXuW|Glc_aJ(q`T-8lOf`3cT) zz|OKoEqCQJ2LdD3d$XLqqB@P#28V?dV1a10i%)LY**>!8$ZiTNJOI5wc@=0X#ewH_ zamWCj;KxN%r`O;&)VI=dyiU_F1M(hTWFVi-%77I`Vc?GH6oAOl^2O=tp+cN7ngn0x z^(PH5JlfsvHppM%ZWTg>bCuZ@4mX*2a@0&Ak_R`gM7mA+wU2zh(|lL$S#>2abGxzx z1yE*WnfQ0cV|TIm67dw;DtUhI4*yeVUJX%qM!&-z)g>?tIMXAuUa(d?vKMI@38}N) z4Up|jX-e8;s+wQg$}1jGLtu0nok|PEBk@5dq};HnvXrZ9TJh+US*K#Ls}-J{Hi`q! zDosGWi`M<9x+_fO%Q)2w`bBbyu^2F;%&3?NqKvv3mODDrPx8*}e1R!o zBdx?>9!)CIoFTW3=?a1`( zCg?GS#}-mI@xkoWO~im+wY9tyNYICw z0*OU4Qy{qxYZ_V(_+F3p@vOqFBtqs6*b0_Wz7^yd6$~;6+m)oSVzo@zO%x z=w(CRJ-k;GlTMRUy*rbNNtUlLJMVF{(C+1H*(9=}qi@+Jjqi3mFBJt#1N5TTy?HPw zZ*?K7Bmio9JIyg`@r9DRliGk~{dM$o+H@@Fw_SJOLy?Od)iR=A6_hb{KGP~3So+%vBw+z=N zm$rxOHDFixc=x>L>A&E;9>28N-x!@)os8hA;NCv%ub!vEg1^Ds7FHqz*%cODiFG5p zR?kjdc#GP3UV?e~ci^N~EAXb~#)XZgOIN*{dZ>$)Qr3I!U0eU)dDko08gvuG7SQZe z(C(i)@xCGA=u#Wq=QFYy-Sn<(7X zu`)Q~@(d*llKLveB{{Y~6aJHbDAEyNQiEAeA&VcbUA3u9A($ER-3&yK+_3=@V*zKh zYCb-k4C-z*EKHknZ0;%$2lI(8Mnu#wFv;a!qpn~%j`U*4G+7ny#cr`3?ReFq)Cttz zqLI_h(AC7dv`F(}vZnqZ;Rc6!vo+2o83Za7qW0qPQwGe0Rj8vY9^w`oxEJ!8`E=BR zWS7=upT@NyB<2;S1tDRfcoqbDDl7{^c1cZiCfN=B2phBDP{I!R+l)4&)P@Fr<*IZ$ z-_L-cH})!VE6ETOxu#}K=>QD5pb&^l_%l&M0@o~(ySs8jys56>RM4y%A$mmNQ(80* z@<9|GZ`5|9a}#+xnKmx;_ccpwi%CYMybC1}Q4lEzs>IPKh(;68MsOfeeifzCrlC-5 zr$tj!_vmr;I#)s8p+Vqs1A&4N+8O!;!7EkBlZY1Wn+?P1pIwin%AC3&2RETtxVfU^ zNptZYQ73&qxuds|R43Ivm)t+kr@8^w#AQc5=zFqKRLchAt!;STc!~9StKu~DWi2z@ zpIzSQs+RJ#j`epYG1Pv@_Z+Z?THc^;3`2T$`IHcPVec{I!h8q!+u}xYY13edrSlwY zZHFtx-JkWortG0d))4nMCPFTd`6C@b;(4Y|(8ss~V>MS2s&I@a#>Sov8kNO(&oi1{ zHp#>_VMHzYitjaL`}c)KCTw4NlJcn`#*hHjdD!=gvVCU)Epa5Qsn*aQU`=T4r185( z&C^P6R%C89tur8W&W zcZU3#w3K~H+wjX%aDsTBN|ppF!L{P4oX(NOoX{zqVJaVtqOj!8#QRjTog=lXgS-V$ zB9$`?(t_pVrxo6|^)D5k=e^|CRc)<9->AO>^BYb;YS7!WEdP?ciJF^2S5sd0qhrS0 zk2Ozi$?;ZM=sMGO!gF-f3&&GmSKlXVzE3APq5a?b45sFDV*AOfeGg$z5z}M^C^mo9 z&b*n?=43eD7!F4L$#7}FbcCixhqHA++c15sPvY_--znQqM}jYqZ~9(`eddyFGgAkP zP_0FMvKlONE!PUet3< zWg{>5J$UNVg2w45k4y9nl&DXtkC9x_(c8Af5IUTBUj^y%ND>?gm}6XeFm=DG{lr8Q z=>FvGm8I?#8a;0z#xyT{Kfx%1&q)3p|5U@Hj|gbU$NDYRHF^f|&YMS=_@?7T=vA zgVHGk8I=VBnKwj*DHKTyWD0>eflOI2IgqJqR5K?7xMWr^0`r*?Jm2a5=KWP1qd&2d z;4c1C`IE-ZjY0bKH@tXf!|y5k7ea~PKw|Kw>1thrzoNl=nvSa_cq?=nZ5g?qpo${3 zo}jLg8@$u&39_+M)e}(c(b7LPTa!F4dcAE%J%MwZCn*WJj-q2L>Ggz~5TllcMH6dW zPD@XeZDFJ*%5sV6iMm2cdIDFd5)XOO7?SD<2ce#TVm~qQB)y^_O@i(=UszxiP)tln z#1w%W&El%4DAe2%F<*9J^OP+)5mUIvQUbLys~6tlVG!Ya3b$BcGEV58DuR2enemX0JwCAt07dZt_fArdv zFZK!Z{^ay}(>1Qk&(*GR*}@)$E9dZAjFt?JZj4hYaa`ZKbmq494#4(ub_Sp8b>WXP z7=%#6(YydgLk_3zP43Jars&fxIa)+Dz8-R!!zx%NYtyj%7}J5?W)8I?SI_2LO13T} zZFKC;ydy?20Lz@?$}#eGvD3u7UH-nSZ0NU5e{jPK5 zf;*ceufw8FyKZgLgwS{l&Mikqf?xHd9^b9@2;O@i4G3GSx^3N0*N_>;{ z!xmeb5`=g9PT;arLmU$P3;Aq^9LshqkWXVemYhVFnk$Y5#5m_P@PhED^hk0D^9XOg z{B72%7}xY4`A4tA)jo62A%{u~wT;<(+C>l{yD3ZQF&e^Vczj?#`A*WYym={BTr+MGSRM|AkE)lcg_>&@iNkJA zFX%Un>Fwi!{EL4+y+Ov!ozm2z`q%Qg#*OtSklap2TboNT*=QSv8{OW&G#sASl788mJY;YH2Wtl~{ryAFPQh9=dBkm@9kdONxiHhS%qxsxQFP?Q4db{XwaxZ`wk-D znwMLj1gmJ|#+AHy&wrO|4f zNf_6x!j**_pdA(z?NDfKBx~3Vrm0LTo*nlN*uOT^Ne|U^QZyFG(jnaH^gSq>EW5>XqQCJHa!>C?A^nZukEvUr z4xxH&$;)5iJIK3{N_|xL)L?Z?nmyHxOW5CuRPfWOl<9gPcCJ{o)h-EEsZUU7bXTpG z`oyOZOMUv9@@$!Q5#rB~stamObZJlRJTRU#OQ?tZg(<-af_aRsI4#p)ToE#!;T>Zcu||tiFM+0a96p73j8$z8 z^hVqW#bt1n`JyG0((WM2Ir(T#ipl$s&JWB?s$7iyc!Ittuv`3?gOjK%5SL}OZxtr$ zeMk>2Wu>SODK+dG^P7=UR-Hg?v^8{lM9lK%Bf~T76h+bGe3cp1PL+?1wfP}NPGY>36g83VWR`?H95*+yyAYbKOXfr z&kUm?d)ym~m$QzS7D7ia8}ba~eo-&Hm%O-lcV@lN@;&B3dK`(jd-+;G;>{f?Iywl? zVg=;kE?K3iljF8ww5&jXdfl4`gYr~t;bnS@P=gsdaYR? zu*KO)%8n-}cu4J%x3(aajjcib-ucSbpln0vEy<=nWp%F(A(7W~$X(bLQ+HWr?>C;n z{l>oKJp#d76uR$ls4V&A4!H-L8{_`wx#7~zq(7c;j#%8+3{|4(Vdpe1waAIVaDCF} zoS2=bYVLSSFDybKm*hys|^? z#&!$0jMmqesv#sbx_@?iO}2%Pc1cE+&784NM%35V+loYGi_)8+({5grw=1oZetlz} zyaZc&k6+sCZ;Z~YPDYb6_xJW`e=XZ_m-tB@@~~R^ptVa|QBBj_V+==#Up- zYuHNv!O_Nr4LQr{e**XT!7hF>UGt%LZT&;pt}U1C337&>((V^K@>u%rF3Bhhn$7g= z`qtKXFuGWuO^=T*w&h{aV`Z&hfxX7Qja6oMo)-#eC{kAssf=l)CEnBL$A;~2< z^-MhZ;}+_Kr$_6<-aP%YL|07C&en`|5fZD@0RNecw%3O{z0T9Rl679lB6*vC?p5^p zP#Y{|IhY!U7#=;Y9^(rFm49@3OvV@BDfo*N|IlBj9xlUr%wCk~5qcI`jh*reRLGuu zxFGlDr!}h&TwP}TuY4f_tM!gr?Uam8aqg(*6s=SEKHniP zZcP&bH&+b;X`9FR0hsev0K~jLl>B*{tb{v9^n%`fW%!hl@_1_VZgsM?eWGqDwy3vW zmddaM`5S`ySnL$iAP$Lb)fUS3RtNPE%<^M!$q(C9RlqS+cS&28SUKYm8o^+#p8puK z`}}aYy*3zaK%#x)9KD~?URZ3L^V(B73Z`mL#;m9)c^-Rf?FVgo3G#fy+tjgc)A&6u|A}r#SrMD*6kU%7zmSb*(g+);iVn& zu7W&2-zqkS{mmAUY5$&eKaO)vAyPr54FoFfZt}w33y@H0%D3_mX%BVD9hmYFXta~5 z2`$~40g0Azb<8XvN4H8gCl_ZTKJqGLDd^}kYyo@UstKqkDhf;yfucJ&6MnlyPe+X+ z!1a{DTG-&9MI5ZS8At$ z$g=NsabvfJDnlsy8?_7HIbjs6rb(@kWOiTB(N3i^(G&VUZ^lHWW{_ifI8IYS!F45*P~ZA6ru=`&_xE4<&VgK-pWjQr6&29`(KqXrF|Mv9>xU6}hTQIj!> z5JRQ0LWiN|+SCoqTBSJh%aQ1H$bX?U#-%1T5MfujyL&szwU@^JB@zq)swG#~>=XmE zB$pO~bE(UWRhV8CL%=P_O&XA2EqD&pIM+aY&BmY{SsD2P^5L-20m#U}B+`pE$?7@4 z9GujiEru5Z^?5qv(^*&jn;DFP@j_KY&casA(cb`cpufnQkgKa2uFLfF=oEg-3RQpy z15strz;vOeC=Abq>J6u!5cGD>$)~!_7p@N1gBww#kD{hos+Q&tdt^lzhtnG*hE4Iv8(l!~050x^OHH@*FM?h1Jyg zs>k{}6F=dCr{vQGxiz*Iz`v#MfI~EHkjudh^^#@$iV4cnJZVmPYVYTXU|-0KhCznJLyq~K#^w_6boX$ney%c) z3)23?uKWxjzth2pRC0n;z97MrQs#Mw8Iy`tvRxP<1a9aI9%L6>rtOWqsY8C2>RSx@ zyFprQL@FF9pKlCch8UN0^qh@Z9PCWKU67N}j8V0}Iuz8VZ~|DYKh}!ox}HjJ%AZR-cIdjTA$%FbWKXKNI^?o(=8{*U$2s zh47R%8fW!qVZ+{%H9Va=C$x{ey&%iM2BUAwfI<5Sj;8pL<+7gXGyh}PAAp-fm*Ky? z@ZVARPrcybdi!#oJQ9;&nW$?d3h|<_XoL5GYZ+xtggHUNPa)T>RWA*%?NGqXPVe#L zu}5xzH%{3Ly9NBP7ydg+52@j`A4bLm8gDz>TRX$G!dYPY=;|W$C4DSdS~2Tf`rVab@N$h!6L6pXS^1zIyBy5BFr~+~rIu7F|N7@`Y!h zsi4_)sMD06G!>u^j78dO?eTeH)}PRW*1qLF%mS(YPcJwW;h;PcD3{_i<&fmZjSZ!L z0(sn#9#eY356TqkKKg@ST!Y-`M&?WCUoOF43Igsv{3O7(V&V=s`YHbS>NIRAHNiTF z$I>W8;d~KHqI^-@J=}l3InLed3}SAJqxbCIo_hMr^?%{s{v8E5mQ5T8uO89lC+{eb z7*5_>iwJbELAiD&`tZnpn&i6X|=V*WMzTvsi zCjXaW@v&szaD9EWy`%oW!9Q*eUpn=>*O6Vr%>n<-ITru|SCiKcBD(_^gB;#*@(Gty z*M>uQq!mK(T*AlZ$(FJzUNsxJL@!V?&@ApG1MyAw?ngu_yvg&AeJ7=GAs)qy#W3-WL$ zFLiFzNpiW5knl{kQNmd!s=*)aWS(|pM*-znndbwInxwW8f-4Fxog#1Pkmu!?Uad+H z?8KW}dC4!k$czyx!%xy&gLD<7?YT%j&}srz!Ad&-2Z8 zX(SD=c{ACGaw><-ihoe+t2jO);)0r)wg!8hMqbxcS@E2vexk0*`Be}&k!097b(}h6 z%3p<|g|hG})T^>=QPL+*bLqgT-;Wti+3*XD7GqGe!KP3I6M=6qlgw(MxvvoaTXf$| zab?1xR_-)|Ez&oaKCFnw3=Wn`2qqtk?qD=_o|@e^7NWr@9LcV6fJQdE1_#2W!b9*o zVh}QBLK``LAItOZS%5bX5RXE@}RLsL6 zS7we)MWdJW9IJDe174M<4OW5v| zpYuYF$=?^`XQ;mn0{k&Wsohf*qaawtVyn&CR`a)v@sM)Y;o%M%g^|z;ZNFez&8bPe zPxvI>&g?B;&mJr!yp#FpRI||HY$*ILM;bs~<#Tfu0g=WYp8PD5q6(}f6|O|5An?j% z&K||=*jy*c)i#G1kXlNV{!T%T=2*<95nk)Z)1FI-4H#ci?u^ zi3ai(B`!$)PC?ziu^`Wh+aL^ptG1;@OYD*W-1Cam4V^Ao3b0c~r@zxe1^q9)Sn->k zSd|cz88bm!VIWH^WhV2uQl;IU`mGK*lB!fZj$}KSGSYC}Qi^bsuC$%=wIDB#4kLoV zMnagu>@rKu#CE2Vzwo38R3(`L*~!XRmhc-EEQu7dFq$+R?t+wH@y8qZ&>p$|$aM@K z%IdoG5Ay0xY`7KV)v5yCP&BTx*q!ZH+?r8T2w{eZ6=gW>x(tQGMrKY`M6NNP)GaMl zKZ~0WC}G%pHIwO0MT>epi?T0!4-7@{k#C;SdEG&^ge2P3E+86{{!!L)J z?wvkkslxLw-QK@69G~18Z!F1c_EZB8#@3Ozw(6rz^xa}y5Q$HEXeEfLPOC}JFGVLk zI$bj9(cZ?{-|#XEWd2B}HwT=pCxz}>UVvjG{1a85$iOLRF#1BI^hSJy2_V)D>Awwq zjs*M;x3DhxMiHCdi0{J>Rna0fKjWu9-^(M++rZbBE=6Q6+~v|68Id`K*r6x{q%Hs} z#inlJ<3Sbdx!fVtgtib5lWp(HeqtQ(K`BX6Z+!+kzk--_x^E zu@qx&tU}1&u10oCh0tzulV6&VP`7I+{Cf(*$F^Pt>wU^(TF_DDmksQX z30mbF6QwU40jzpY^x<9wA&*D8=9%$- z<43N?E%)k+hn*kWHFrJKpu6E^=njGYe8!!_!si*Wh<$z-y1B%PrLIu$RPCNYf6`w&3#GgMq_@zUKY8n2#}*wg%`FI) zka+N3CB0?>D2j{-A`7d+R+Ok!Sm9BS%CWRe9ciXE_CS7|S%pR5tB^ZZF_ys0yhtsk z=Hbwu$klk@(QP6OF-$3GTYDnov6(sfx`V=1eC>I}FirXC^)zL}Oxj3h+8QFUz($)% z$M!|c!j)W|?X2W^f5<1F@eRXuHjg%Fd9XeJ@`8Cos!uoCs&8FC?Sif3_k0oD_c3J? zv=Rulo6@k2CmSlDVr8gd#*hptusiICg-uO6KvY#+*An3;@)OH-%X`JBl+NXaP{<4$ zk>XB;BB%mJ*0XT167AB42g=@piw*t~)S6@d;^%5VuZ z3sP@d_~RZ0w1!6I4!sqeA;f9w@Z<6qJ0!!P75btgGorveFb1bGE9A3eEcb3MMCS}o zp-{8J6Y>n*5qYL?0BO0rsM`2V5w3OP>|f+3ILiS$%NDiVmDn6`1}dHm7UaEI&R$WS zMrwn@LJF`zwA$q7&fs}n95O&B_;JzH={5Kb^{uoVuhTS4p$7i+fP6M916CA;fjg>G z03t`r7pJF(3US70f|ww}@Mw3t+aQ05yHyAk&Q)etINW67$x$U>;2>(VU^wBpqk^y8g_Q#|C=9!@ZFJmux z#3B5@mk=*4`~Wwx~4uZpZUd zQLr>XFM8dZ2ZQof7qUtMpr*Ie9J3Z*D4AZf))No}cWvFpwnbr|U|5iULgV(v)}XiR zmi4Wz@ygbqbidpYBf|gncn0(7yChOicr>Mw157wmU z52k~`C65VO%kkRQal6y#kbm~gEe&S=Eu;1IB?hj)>b3aQ-S~c!*;ms|_vY!JC9e1P z+>=d^%Is1TcwcNp-#uKLT-qMeM}Qp<6Bgv%^PZ>wg7826J0ji4bI0Sac=UjqF-IJ9QDwa+jZidHHwXq*p8O{^kaZTfTJFyQzn|SSe+_ z=iasT51x0uf~`R}F>C?NP6b_>_rrt+gxRdl(sR^cbg@2B9Uoo9wSJ@sRO`sq!TpbP zuxy}FM$KsAh~iYeXe;)_1uzYp@|9P~?GS1J<9X+lV%nDl#m{xa@ z%(;H~h~fnM>(u8&>%E!vOe>H|tc-0^-~FJ2_ZGzc43Z;+tv+y}KOPQ_kItSQj)(1* zdEZ}<7o?^wE;?n4)LxK|!{Gf*hYXM@HTFGh&C^!I>3QBKQAu)}s()9IpGRi(&YAvt ze_Us@<;iBb_mW!oZ#$AYVurjKL2AmbhGNDY#DY9tFBiQKV06S)GVFx~_T; z++Q!y!?qX!5DoGV6;cH7^eqKBk*e7n!~SNAIGS%K+%4f8QbNN;ur05dPe(0Cc4=Mq zXS}WYJ}*)@hJ{Q%tR98gD5)QsO?7Q zCh~SNZCvQ@YnIp+!&w`RN_iJbBBCHt5LAhyQ4ozL4~g=tD3vxe6bigUgTUnm0tF$o zGxQ09SBgfSyxB0E{u%X1s?4bia$Xa9g_|ono;0N&xJMLBpHJ@Stt1spbr&WXLBB#L zX?6WgIr2f@la-=cHW+Vh!}G>Vtj}8&r=c%vnc@D7@*XUn&?Ga;(bloLyxQ> zrq7u0zNT!a!PwZdL8G!5?|DYkPGWsX=Sq?`H~uTW*Ocww7Z#bYed$2Tr-m29!?x|k}nR1tIdj4B#({OWV$e&3| z*{8G(xljMjyiX-d0+rxe@l;Od$YM_Dl+G}f&z`^GeJa_`ky_P3-U29*${7Y}$@cL& z-obL^S*Q#7QsH^tOKx4&*2;M;kNP_>zu^R=2E9GXUy?Ubb2IP7I~X(Oeyn+FOOCh7 zLf4tL6P}}+UO1lmI@bGS`}G&!?gM9DkXQR2!k!|g$qG#)yQvTa7h>gr$-suqkJO`eufoIE6q z)wTZvbpQLwUlpF$W&`%fV^{7jO)s7sKVN%}n0>w?d7q}2I<%9Fbr2mW*EQ9p_+85j zFfympnZP@wi___E7?)KT->PC!+nd`;)gn)|u+>;CSKn?S>v zDUWN_$q6M8>sH{;xP+RKw3LK7@5=3K#noo6p`#V8SxRC`GEIj}Nz$_rjpD?VM4_wg zltlSd1zSl_sOdFwQxX^}o)A4$D)@CSU!yM2WK#mRLXJ^}*3jg-n_-Z%0t0Esx3S#P4WpNKnS$ubf3`(aEWK^h8-MF+EXN zVCjjuZeU{;flEb+hdgNvN%e$-P)|UyAAjf+Mw3xdkS0NQn=dRd3MeKfBw~udjb?Gh z6ETHA5%XmiHc#1-6ETG-j=Ir1vff-oD$zs9^=Juz;Ee@qy};1b(9#0(aI)b_D)18$ zczyLwy1rvZyCJYq=`gB{8f{BQ5~Gsberi2au8>%RBPfMTAi|g;%fzSm&9Sz?;{J4w30#BL&If4~=Qr zCMEfyyH`umgtjwsz6l1vlaJuO95@?DJ&-I$?$;YY=tJ;1du6e&GJ~9D7|T-1g}^t5 zlSXW0yf(Shry2$5g@AtL@T*EWa;%gm9G1-7uQ+c=G7@mnh6V;!#?IR&Z2pYf#Bv(F!|B&fMu+IhpmM+EkT2RjfqZj*x zd4F#u|xP3M!z?F0OEk;YmaVjN_>wA~Hwnc7$?d9wYKG*BQA7wBIp@yS* z0gQ$mPTgBf+6qs%S)F%^Ye)uAa@glx$r{`u`*E-DB*! zs{8Rd_f97EjvbN<1nbZrSXC>cq1D9WB$N>?&ctBS#!m3Kgd~vZjqkB@>v`qgYZLRA zs{s85S{_PyCMhjXQD`aTpMpw*0I7hgmi_}&g(47b+CtR_>I>TDyY@L}zwX{=Ki9c- ze_teru@mpJ_ImHNKYJ~sV;5TO47&Eb9k9$Pt{fw8=R1we+r{56%Z6@-n9~>XP}H>) z|N4)9oi!MCVG+z ztKo4yUMn_=lat;;_5U$7HonwFu|qE75IcN4EgQOZ`*g1h|H#uBA42tq+->r0TKC#w zv|YQ8afxo|kJl&gxMG{N$(M6vTPHv3TV#rx_ZT1INvPsUBv^lAP=QBJ_45L0#t4uE zP0?Nbocu_tU5TsHzTaX?Q-W}I-|<{_s)>W4B$!+wS0C^stPpE^WqO!9)dOHPE! zm(s54#YdsufOI{Ql>OC`T3qAZZ*Cz+dN1+meq-=I>INxS{kq;r541iKb)2c|R`uOU z)6;BSmO}8)Bv%ajlm7at?a^i*GV9*_v6tO(^Ww9~mFv)BXXor>y^Q+f-fnXF%4OAy z()e>FhA%EAH{l1ZFoqWk$-jfr#GmBzEhf7RDYVk*toI&1d`jyaC>X1 z;`-99{qx1>7zkxaG*>DSuer`FA#%l(ALGZ+NKwltVaL?b1*chuv8@6Wd ziK2Z26g*@bk0HqiZbA&Vna^TpvnyBkz|-&fUT0Uf(t(Bp(#g(t@XW{uk=U%#2W+li zj$2A;&}7S(&JOubhurHpJ!zsc`6o`NF*$3x`E)&65Bqd=7wW=3My#0%$9kI^y%tCn z11=ftZ3g1Br3eew8xC&bswHdK45q0}E1toB_@bK8kw=jdUWtofyzJopqR;|@Jdg*++v&#d zj>lX*=_EgNd5I1O*h`n3&re|_#*#dZ7L1r0f7v}bHq_~GM=!2Duz3xp>@hhul<=v5 zJUtY{osR#Fp-%e3x=xD50#P~y7Ptn=hO&UmZt*m~3VAiTyLYG0RfzY;G-|NEn6$8j{k=#9Kb1<6uBVqIu}XcMLZiECwbaKyjacea*A!>VxD@@* zk*W`zs-rAH*9tQ)%vVB#{^|baW*_DO_k(KfUqLEg@;&ji8l~L8mjzbkDAiBCqEW-G z$udu9Y&FvT;T>a3VvQCDDX``A>RT9o0`V9t+Z^Z(e-Vny;41S)OD2Wg zL6oyHwBb$0hmg(>rI}Q@Soz}-I#FOZ|1kw8QCT1^%WB^m;GVEjR*L!%Qp2t>zY{5C zRjCvyYy9h5tW=d#W^^m9rmFnYh^eZ&CK&Ou2G!b6p0l#Icg3!!*#EMr-?9XyOTLbj z)(NUuwG`!`YH|`d=^ZE7s1=t|ddDekC%xm#qosFZMO)Lb=W1%Zmt3*3n?lm|Q>YF3 z8d51=kRx)S_LLLlo}80xR8Gi*cx-M0_S)L$l20QgL7YrpL_$u2_y-V^AhBkRjaF}u zn;6PCPCVWY)ZR zID3+Ue%FUvgW<_yoN`+=y?U)#ATZ`Fx{I@uMso-GjJ&%Ascdc!>Url!w+DF}LjQk5 zHgD>XJFqRLzGX4`4U3-4P+PBIzGLs|6&&iDW#m^o2|TKbkO(%+4%z zCpi^R8vJ;RoEQ`vlRo1_!{=wH3Oa`}CFGI|^ThR>HaMn~I?=oAI8}|jO^s{EiM2&? zsTwnseX}#_$Ol65h7P$4+bz6wxUsQRjhd^`{r?MW;gj7IyoNbrp^T`Xt+y44@)o6M zX{n!Gj6_U(O;5fxPhO3!y@$_l^*4tn*CxZs$@_YHmEY#=xJ&#b4{}I+T9Ye(%~of{ellJ2p?7)xL*A|}=j{p7 z;wd{9LAA;IyCk73Xg1SR8{6BX!SGytHa$E%*OrGtkCnB4T?9$k6odR(3kOuKu#R$d z$P112@-15`6Vx<_(OGBzlNRcPCx#nEZ=U`$M^{Y7&en`|5pt#z0RNc`cQ%S~uX9~j z@YW3peUH2ERq**xX)I;x8FlEed`NBVH$^M|=<=A1&cZ18bF}o|-=H2Y!FtSIwA3T? zEOM&tlv}bN&&WOLY0cXGXZuk2Jv=;hsu&gKD*#J!wHWZ|n|vZ8ztk?p;-bHqt}qPk z8@gnxHJXk0X-yD-mW7?SE*j>x;!vzGKcc)tNV_~~Q1|i4{zjj*^2=X|z-skIt#(T0 z278D6oaGlg0=_r`0JsFopMZt3w!QylsYHwK~6g!Qa_p0Y>6F^8r*fQ;W8S%MJ+7fWVUCnHEy6u&z*%07;4>~f%Aqi36G6@ zg&6=ke_uvko^BPJMSrVBWZHiuy^mvDlZcd1X#);8yNkTC_X;Gyne?qDu$g?O=r&#n zJlG|-VakW2(T>F?v~+6*BwE7LF|&Y}8Fv?FBckRSN=6<+mV$~t!xpe}R*gqJQBq(9 z5qR-s0j)+6;Cj;jDp)T=6wPLX4FEFLC)hOdsSKkqE4W}lnaze1I)2I%Vb%@@ywGJA zi!S+6Gp*mn$!wOHXAuNo4JZ2gZgShM%?zlKV{Jr~lmg&4lwdB_LlyJ{)Q{00|kG zn9D&g$?87944l-SE&3M&)y6vHvq?|=^$bSBc%iBxV__@in1=0o2Xb{44c7$&15st5 zVY*OLloHQ{>J2L_m-FU}b49Rw4Y?HdVXv*uUqxZ|M*q==jP@v zz^^^<>k$0%1zyaP$3hY;6LmG;MCZG3C2!CM=YVSwWle-xLBh||x-UMHJbmSAXuPjx|O-LiGA7)@?3Mw@3RecJAI?dVUtWKpJSEh>IbS+2 z9^$Xn%;6gHjFnz#4xD9_Z`9rUp9IRSn79Lueu_V?I!(EhoM0WmV=3bhWs2nOL-dF- zI3w`>`KB;;uQI4~TMWI|?Cq(izfk}3_x6uw$mg=Ce$DDLcmqE2KfHgnFAFo8 zZTR0Tc#^qnXQB#gvfO88^yfiaSrYx&W|@ALjzVGJXrfff--A zRJm^18Wa#9M5_l1;lFBp-)OjXW*u4&jn_B(<1;X|YHsoAWN)#tG29u;e+cNRdyCV< zE%u*f^gr7wUNiM~uOz#QtpWRw<}A9>!PVsT1IX?G+93PiIJtz&sB6O^Jk|=KcrM}b zdGZTrE&&**It7eqhiGz4&;iZc{B@^|MVA}hfYIW0$B^m<%+86x9GfIZ7<89}7qW6p zRTrnfh!GgHApGJKyW^KC2#2LK3p2{n@2+Ig|KRg*CoOf>>LjV$M@V@5x6gEKs=F5X zWGC^oBW@H>j+JRX(5OlBvuE!kq3x1^ysJZAl45$bDnYR0Gq=)`-|`_dMyT|!z3|e5 zc0CuU2TDz#D!B1`?xZcT0@Zas^%h2jih=R>F8>w+Bq>t@_Qg$ljmGy4I?;vg=O{b( zsx(P-M7Wcr`f80i)rzSI_0{Eo`YJ?H|C)CxJ6=v@uvz{Ga(xxYM+7`jGt;KA*J|W) z@to#vh5{>`ND^$EI!+Zb<*q`}LYZ#~Wu82gWQ&46ahi(5n%Mz6?`S;*^H_4T8hg!MQ1hxp@oV&0h3NtjYlz%XJJopAf+j(qu zuPsD_kw21M;Q)@olP)BK(<1 zpqor8C7=r|KY6}SQnj{yYO>jXAcX_M`uw-bc||@PeA#7^Al);DAWQZIpbyPDY77U- zaH+Qj=Rg{#45yk@g#)fb#oQlqW#$-*M&=_+6V8l092|cb2kN|5xe2txy?v9L4#l+v za&?L-&BvkTzOi@@nO)p=r~I7rX$<*?jQkw+mtKIsQc-I6RK+L=R#mEys%G@GzL1-zmx)?dDfh z^NW790|M{70(C>BOOyhdKSpg&Xd$EfEngoo`9-EQy^}# z(v`*kg+-Z(a#BjIZEdNEfJKrwc0$7~CyyN^q>HO0b|?E4 zTOC&w<_aEWu4X45A3z59b1 z+v^9?o`!!b3r%nM_u+@CXc3#Av0{a5@<{1z;A>0gBQh8IA|o)T;5+1nfY=3KrO?!k zf83~neo@vDi|R)#<%^8))QuCEn!53YfX$()oA|qBM{)1Ml~5ohc$4o0A00Ub^oJTZ zvFLLzQ3dQ!tDnw-{6);SOx_ZlpYlzDi0CF$BvtY$56n$D0Y+<(!Bq{lK&!zJ=%pW+ zZgYZIe2>pY`BEiwLlr{&?W$#W(@g>wzgp8KZw^kx4O!>ZGpxWsz_-8xffr0gP)XWWeXUQMUGh7@r$lWF{10SVBe;XmDLuG2bO_lfCfslG^{{wlg&{Y3M;F$x9DuU_d7Wr@UKx0>kD$jOm%(b<9C>sYX@L-*A|6d!Wa=6D5 zRQ&yaO_e69!14~44YLpa2eJ46GP?oEw*vsy?*?OTSv=O%F-_NLw*k=mg%|%u$RfXcD#_Tf!yjMw2PXKw55kX{r zRoIFWxeBW^>7ms^2^paLJKBTBqcBU$)PZK(0x6RWTYd=NCj~TitAE>)28+y_P zCO1tNBB8*>CX)`m7cmQ$+9d5vJRcH*XVWy9e9qMj*Lr!hNy~%v0gx5UHK`sj+Nx{a z0A7*bcSUg5$D~crN+8s3a>F*BYp8(om7#_iLNY*L?O}&6Y`U}qL{-IgB@uoW!xmn- zx`($)>0DUwh0K&KQtYWv09C-qx))CKIk&SIki3&GbjT~|A*KPMZi+S+vcg(jpN zk+Ceu$ANxe+K%1L`mqrTyH(BHU@5JMiS0bI=k``W58Rny$U-~qAu-j_yvtKK<;@XK zca-}eE9=M`T=}pk0pd=@pdWX&WybQQi4)jvzdnJ;h#^rv4I0IhP_7CaaPE*mo!E}9 z^OGje!F|h6e#-)H+Hi|gofzCG<6~_ zdqRO-L7IHILlO*HqAw^iBMQs`W3Vc-M1Gl!rD9efIwyDviJGGiiZR8!6audsg9P~;INPaED)_$ty23BKO-<+7l#bc z33gmG-Sis#hCC}R#p~#X>5g(jQ9hrP0ZWQff;*~H03t{8SH-7?5^=_80tuosYw|QV zXdZ39?l#C@;BF;CxpNiS6%IE`@uaAkL?jDt4QC1`Q^^-Q`dqc;x)PYUT~UGpC^NE5 zxYaDanvXGCBQNjW<{o3_l)p?j{SGY>cHH_&pc!zYM=tMzmEsXrvI_6i5~V3<6RB!? zX)CRGL=B;&%ji^EC?4?-mO@Gmn=(t8LdO-4K9h7R=DRw|l2ffXu$-H^iWyAlbboWR z4-c*T1GXdU1y_YZpk4VAPW7CAo?K!q2F%D4DrTH0qi*^IGi&-u-kY2+lnSV&l^D#U zNhO*y6q=+~%$}<~65C6zSlLaVJ?)2=Vj1}Yjz%+v{P74}1@(6GAB~6-L);K4cx5pdOsSndN@?dkh8EhrdU@H2EbFLSZqQ`%y^dp2 zk+(EJFM8dZ27~-%He{6mKuvE)Z?hI&D4AZf))NrKV%rjE#zk>ZbC>)x8n-vM2fbY{ z-Pqn99o-(}_Q@R~B1Gl64^P?_4DFA6l%5wufMyMPR5`p^*|d3-+f-vg3yaAQTr<+V z6$55ro-PN?jV80&z3$PboOqK*!`}6v*qHP$-ma(Q-L&gJ?P5jMQr5l8j#H642G+#r z52k~`C9f28+wRCV3Qn$Ek$-W$cJ*fdONSd9OC`AelGoxlcEj(R%)Xj#x;Ib%nd5qY z$2-{wsjw4mw{_ivS!P2z0CC-;XJkg+Kkpd*=e^g%=ePQs!;@>1A&d&{?Nxs37!~II zjnZvlB|?y0VbK*?H)_^usxH;ta&zqO%uJ_fzx>;9(yOB|e{&PsEuX*S-PD6!tdz3e zbMNx{2gh75Z);FZ3|l}m(=J=IBtMF{fH0fYDSD0?4A0dks>8!`xYmypfodJOG`RnX z4wemQWz>XA9A2EN7j4C!{LzUjs*(I&#MQf5V^0h>irzf^CsumH7#3WoH5158Y7ugd z6L29q8SZQpV|ntj55VKCXtlzdx3|Xp?D|YrE!(q7c4fV}#re6pLx<&pS=Q8~ioU={ zGUs~o5yc7iH>l4E)_XJSORYdEv6gIO4#!7gAIyk-43Z^;t=)gNKPm=?ho?>zqoUn1 z?}sz;irBQpMyI$UZ7xWMJsA1h4jCX*YP|Pox;$+~oR0B6kxG(!srvUB`9)+_k5BeD z`lC9d&5zv^yS#ey!G@-p9NF94`7yI!K_D`6KZYm@DrBXz+WUE>0H2H6g_@1V%4`bj^wI6!xE>@`8eA%MMa z;wu^4_Es-CT9HIPlfV>^e#jNn#ix2Z4sTTx%(~PCiH-U8gA0 zr?sq3g{8XCOM;i*%+SL&ZvhYt^7j=IC*N6TAn(e^kyy>%Ec#n5;%L4b@wS9o_G`Evuy<3woG}0H+@H2neA@=#@mZOXyf@5b6hoyRS1p{s-_$h{&@s{eVz_cv3 zC>HW<6z-~6860srnv!`*eU0LhnlEu?#N&U=;t^m%gUg&k6hAz>QDkju50vmOtFFEG zu3o`gXk{7sUIHRW?AU;bY0(rGqwz49yH&6-v)!#H4(8KcjEJaUV4};tMqR;j9P!1F zX|gKZOS2#9D5t6xrB0y!7L1(MLzj+sX_01wU3zNj4OvIXsHN^uk#DZKPmLyw3O0_61`phEw zc2{hOU8+ks6*Q|xh#m}|VoN3J`-4_v0r@D3j%&5u;M_#KPNo+Z>U_-{+hSOloo<-u zs+N)Wp(G*_A_YN}7#ao9Xwr}zt6<Hg|HjH1?nFAWsoUz z>8A1DLYs=Malyu<#q3kqhTH9PPPj)U3j(F!TJcnL=SX6XZ!y{KeV{ng3 zZstg>>L70pl!)a_32DLd@zZCz1_qI@WRCG(V(YTDR?cg6*dIgxh9i&~^sbP86TOL= zn_O3;Cf-4tG5fIQsV!M%mHDnSZ6_>8H%&O6`a0HpvVHm-u8h3VH3)l(7@ZZM*lenu zc{9VUNio_i2E+cOSQ?Z%LWA2uJRJy|WVf&EHK#LUj(poS9rlK$rp@qJT^%fZ)uQBv z&eJ@Ilgq*yTv2Nu`RmLvZPsItEOw>t(lqhh`1#s%IAhp8&I$7R=zW?db;O?=pxv>3 z@}6eo0#@2S`Kc#u8DN(aKp&x{iOJyWTmw&C zTA-bN(zry=K#}?+`xq6H-N+;MT7%wjrhOHp%OZ()Bq$x@(gU&kmF{!g_2Xzu~_9u@h)GGo^8@Iys>PVxcx+1%rK&DM{(P zE4Qx{S7)7)Ow)0xq$CnuZKovCrz+S=ghEZPAp&z;(G}W?M??>q3ih1Kby4SOvM~W$ zBF890VRH>t3cEjw3)m7-7y(;aE;3-tE0iw`SSa?3+~ng?K}`OtDDGh?i|fshLFpud zjLLG+1DQl1P9T#Oj1FY-8dc9pVN{`k4Ei%gc)k<;t^2AtMt>?xg1gvH`A->dZVb|= zf5VA)HvBzl|6C~U#Eh4sRVC{ zKSodY<6Hx3&=z;S9 zolsE_CP8ZGf4Yl}SqgbL+HeIG_z4M2U%j2KZ|Tu)2y9e3B~?a^Zc0ZYqmtNuY&}z~ z0IA5bmPhU<&so{qyJFW!8xwkRV37j3w}P--By_g!|#Lu6$}rAee83RR6ckEt>fmvC@8 zj7}Xf;ou*tt}Y_fNL>M;_7M>yI1S)CqbncqcPo7q>4vw@1_GAOhxl3mB1ITI-zW6@ z6W!}gy*V$uSkBX%13bW$Q~1qWONMbOCXTCnZ>*`X+YbucOW7HGuGfWs^ai8>HbW4gBQH`&gT&A!JmPy()>^@}EaJQL4Ezi}HITw?yb4eQ=yN3s{ z)eJco>B!_1SB{al^PNWK?c(p-%7$(mc_`{yihupcr~riW$nI@bh-I_4g$`ILVyS8A zU0(kX?wcYCnmQ-Anz_}LJmMa5Y4eBMTg7;NV|b=epAahPSVEHc7A)$V&#isqF*vsr z8F7BqoqAla+5>p+92(%a*3@fjy_t^NqM=3hP4pxcR>R|ZyjE-!Cnvpy>i=VEY<#JU zVuxJDA$It9S~hg+_UT?1{*k9MK7{HIx!dI1wC=UVXuEbF;}YG_AFof~am6-klfP&` zM9L|2*VfvFc#IG6BvkPvlAmh~D)82+Zayf@7@3=+BVb(pocu_tU5TsHzTaX?Q-W}I z-|<`)!L5A0LvBuXE09lPI+m0~mzpb%285Vw9C(3yRC**?#yrBCAAg&*%AL<}PLTgh zKYAUk_L+JPSI?3sr`ftJ#lD|Ot{C(u{q-FT3Mr+lYdFrWuLoea0PmY}oI4m3BqwST@C9Rr~(iRMZr;x*TqrEg(Q2w3BZ zwGqeyeM-HLJZNwLH`ES5`~JZgtd-6qLGmnGQgz&#LB8r94B;g=>JA;|}BLJYT=&thk@D_8gM#_*jB z3%)1IjC>G@%_@DsX7|hSu3FZ#pMby9A@@2?PnxJq{)y9ROwO8aK3$L2!#-Wzg}NT0 zM68(#$9kI^y%tCn11=ftZ3g1x-o%zGELihy@+tR_s4$*rmme;`VA%5XI-~`46|7W& zO_n)P3*Y&Jg`kQ7PsnfcmY;irg@#*Z^0dx!-o2Lj34&m+%%o|8HC69 z;ZRD0atM@oKg}j*8{g0^F~zMFG46Rdyu3p3LDFuIGUSzO@TJ#=c^1fB%M4F8U&ZGR z1jH&Tw!n7hY*a6fOQJMdjWr1q(Z`Z*>?d%;f}$M~ty;2%&0w0!wBi|@GZzI@l^{hP zMM`)jE`ss0gZGO<3kdQ+9vp9{8^2(MGeXlze(3TN9S*RUE;*l{!bprIc^WM^1Oac) zk;#|clVd}j4tMn8+5?-{V9FkoV?zm_3do_)^%ltA80w@q)^$=e7KqXzu)sA?Hk1Wi zcEV0i^Q(|ole>F&`do#0e^uTK{bDuS$?X~5k5uZThDHrm#iTVuNx~BL_aYViR4PTf zo?eo~D)n&+jqa+|QXl^`VyRDEGqu$B=SbBDPSsJCplgMh8}pUWpntl*x!H$#!2O_F z`&W?4mwZn=twt#~@MVEjIZE}DuV~b8;Vo@hvo-N1pFzxaN}r}B)VQpr=`>o3E}U?5 zGW~Vzm;k!NQt5ylr_Ey30A7gDKom2R#R2}X~a}jT~pj1 z#Z{BPj+E93s#vuY<)CVE5;*A{C)lVJHqW@gw8y(9 zOErvD+|Tt#!~WLEA}F$l%~-sUbi8aJboknWWgz$QGT}YsRlU0s%Y+TzD?Lb$A@O#v zUJgjSxs|M=g7BJf><@Rz8cm%nuMLA`1^UzL-ZU8G$3qJ*(_4fZ%;4-v3i@3iZViSf zi*d?r)%5DMW`V$1Na`-mPGrpwdo%Lx7NoMdJ*elMAKf10Z3zAU4cWY@L+-$~nEIAw z_I~3U<~R1PUcnK(S*H39o2LG1hun?LjZuH=bg?v^^hcA~b42%~b1FBQJZXEJ7!(_m z{-wZ)DZOao#D#g{dQKZ0Q%araU3Q$RM&72zwd2IvBDqwJOCo4>u;RBB=RdyTli!*1+QVwSSTavXX|Z6qP#`vU5sQ)Z^U5>YTueCug2Ei!{@j9 zo5Pc9li}p#eZ9TPZ}WECC4Q0zIV3)<$(6rmE3tL8xEN~4;2h-j9dakOh8^ucFx))5 zDSA20?kuAxO`h)Yqh0)Dy5>Xg^7@CoU0crE6XYcl*clj`yuV8l%7SJyJ+-mDJsJ$p z)o0Vg!*gwU81z_K>(`|wr9W%ofT|VNQLYYop|M`RWlL4BX#?&l|D=UF;fdi!(VM6L z%+VE-v9mQ}U4(q<1i*hL!<~&{-0NJ|Wzp37bbfz1g`B*{-S;Z^e2CLn>d;|Y`A3(> zWONor!Jnh0|NaK`a0%99_M)X8p=Xg*+bJg&hs&OPJR|p{r!{N$pY228_wex4sbW;L zdj;@`jQmo&6pM>~so$CPwYjBl8wJ~vH+0EXYcw10)0!XvEektuT{KK}MO+-^_GM0oJ=N=zaf}UgPlSe#2~S) z+Ctvm>YyIX5&GnBKWbA|0mo3iC2d(^`HVwo0E6X9`r^~b?lVQPvpyJZLZW@`90AZ@ zdtp&K=e4JF6in5gjLV{;;JNa~)T4>S58Lz@B~)iNO1{%tbT;ORo(Ob1K9D?_)NF|z zkQ&@|t>H2ny+ti7-(3hn?--C zMP%B4B)yMgT$6~DP-)6BwBZhaaloN#cac~2UV#K%lL%^pt&s=2BE1p~@#Hk{BeOFa>0?QlRrmt8ClcwQHOzl$5Y6;v5Q+25#L z_|6HVU^Pu@g(PcsLkhlYMnMSbs|g*I^dPGv$S!exe_uDHNu!woHFB(th?3GE$xLj8 z_nVEn(MwHuWl%h@J~_Wrl$V;#<{%O>^2hTSSRP@I3N&mQDTaDLC25{ALTz9Q5z)NHLNAg7!`J;%A8%(XJ zs7q>$m(oq(MI^E0Ng66Mk*QT27!d}sxn^s-6^+|^5_5#R%5MlWt_a{9%*#&SH%A?% zF9j?NuY>t%GE)!_kd_@o1e0;nwN^VRr1Oqqf|4{(n3LdDDE7x=qM?@|v5;e))2J^2 zqq~cxdR%217li$}0~F^I`P~jiq!JyZ(gg{o6w}Y!&zJzK)C7L;>K#Si(IG!i^(_Sb zt&vt6kqk#l=UWLd!;eeT=jjIak?&^YSTJK`?XQD*)38=F*9F-3(9y07!d{@wXE-J0 zHucG#eD-Phic-Q}dOvpBd4Xr-Eh%bsMC@Sa&W9+nC={TBqOWd2BU7vfI<67mTCXkYF_vBng8*t_QT7e3-D_X z{5k}`lT14;&UfKT-k=T60oNkRnh2{v!q3yrEYBoQU%9&1*;T$* z%;1MT@avF}yOL4RL#lu6rzB$n7jNU8?QyYwu#(>`Tjz?=mDA#y08V-uV`gjAnNzxv3C(pP^r4JzM>Xp(bZ;G}Lgxr1j(Iy!ATt>k) ztIyz_`pEx0RsU}aK-bLnjAk4DHw&I*E;A;Y85Ka?N9M^bDf*`fxVLAmlJ^=Fcb;x_ z0bm(F&IN#F{1$`*Grn{euOz#Qt-<&!Z|DH@BFo=)}_8ymx&vHXXC#JaaQJ=|jdS+)rO*;ery_3z?p^7;YQ9hfy)HF*mk zG6HRo{coIH!e!L8;Se5cg-|?~@c2CW1vHlc3{;&0MzljTIVR|U=57AE)5fC9fd+51 zZQTW!3_ipS5tw6>>LjV$M@V=CAC#g;09QWQNj&X{8wHeOWttB( zY7$%}cvlqs?iSwFAumZWy;_wZ*zuWLX~}Q-kQpOX`g4jE6x6nDB#d~KJ^wxMTvpod-JZmkbesSl34Z)I?<(f)F?amsx+~kZ=U6_`w+L9QeUkx zr&=);p}x9Y)mJ$bxcSIEPa~;+&AXHxFQ+otEdK+!zKY``0v@QDY17y%qK3^c)>YQ3 zxDJO!UTGz8A_?^Ml8#jtt`x~+q><&!=36l%4<*?mb?g>=zQIg#Dc;o4ZvP&9@20644rP#xS#qEb#kB=;b&4v@$D!rEv3L)eUBb#0@8w2Hl7Gm^&xIYz zYxh*eC7$tzXcYX0US9#ZVu-?zaA_|hq zjc*f3lTO)Akp@s#`P`I6K%}9ECpU|PsBDKP4$J85%507|BQWJMWskgfESB1)5Cc++ z$WzFZ899_@xhyM5As#r{y3PJoMs5w;X*p4g;m+%9x(RuU5*N}?w{OqLi^4X@r%Wi( z61yY-@4NzaL#0cU0-8}fwEO#o`j_e?ZN;bG>V&EUr_7iMHWdbv#F8g6k2%e_w^P5} zAuF*;#o|b^gUJI8*DfW-kbp|7EsV^O*9C_WPN0?$CNMisQZvC7HTm;Tia=GODG)bV z>B{2&!lFE~a9J2l8V+|qq1_bs$W<#>mJragP+P)3{cf+`*a;1{oIG}vkS?y0*p1~^ zrjK-Xg~=)eF=ZKr!vS`82cCn~E0sc?hpIUQMpxmlEsr z^~0R|45dDqxR(Nl3B;$s17Z_{AcnHXA*Oey*DqC1lb3GopD#wownv*w;+Z|wzyqOm zB&MxuYZKvQG!{iZ5D8CuXeFpr9VO{$PAwV%!Io0NJo%O2q(`MoBt7x>AK~c@r@`n8 zkn%?m5!w*%_A~ru`#R}Ktk<#11*Ot!59Pr>pMqp0C zcgPC?u?xUTp{X1HxKRTG7~mI=!&5g-U~1~d7Xmhirf%Zzmd(O>;E7G$z7u?O11N71lOt(2f zEWXEQqkO57xuFUn{&v-}yQvr|G7iNy$eV)`aYNQQ^$aU85b!OqK;Q)#5{cyjTIgUe zs+1>Z;zgA=ri?Nq@WNGR2-pZROT1n(bC$>qX?SAh&yYV1d8=Q>Xii6!UuxJN6ST_J zCQ4m6;Nj{Af)7_y$YYUCKQlI$ zmOGX~rP`Gw)tRvhJU zfAR1CYpOI+1(tWXY?yuUKZw2mm)Yf(OM^hkH-hih?mt_M&U@aiAs9#6%^Gaazn@VV zvCkB~n~SVistT31y>ReeB|SX>jNMwm}^q~8KbSb*5xYXW*Yf@R|I!`Oxgsk1VZg5H*Dj%h6*TO8ETjzBtr^p4?BEe z)1@6Csw%E4iSV-+w(!c;J-k&)=fVO#EDd~S`_RS`_EacGu45JXLWjJP z9%33BPx?`3thP4YexV5|M`SEhv^dZYOxv-$SwA*HVYjN88!V+YF$U-~qAu-j_yvtKKwwog>ci!O2hdl`pcPa+`xT`HQhD#H#t3i*T z0+A6zqI?=OiYK956>9DtW&w3#JG#zKnmh;hEkpS&3%qH=Ely*W+-ypVdo&A@Us|xZ z*IbyQfL7nA+@Y_66NETTorud=70xerNP-!G<-4<w^csSoJS)uso-+Jxf!fp!Q*gs+Y~=Gv8L*@% zCAgzH1t4-Xe^q>XC=q9jCWr|lG>^7lcUvVW5z3vb$gXg>S&An`%_JhrKq)7mp_hJN z?C5jVURYNG6Spf$Pyl5{mWi;dSMxzjb(AHi zT5;foxv8s|!IVz-H#ht6(7GQ~cZIHe38#8aKTj?(76WGF2^BL=luf)P1IDA#kTNM8= zL{oAl_3+i~y=9=atu$&qg%NlknQ+kNHf+cRF<4k-f|yZ+0q>^{Up5Ak|C5w4mkM4* z;qu2La23?s&3`l^N(^yB9HJ_V!C*=)HEu{wKSI({IR0QlN#U^SjlZ6h!l|^06wWaS zb+Pd!T;jbSr^gr`TS(pb2a{7bb*0qTaGy>}eFQD+-Y?;xWisWI6o?aNl<~z+fjE7b zDG*;YF$EIq*e(s7Rkcqg6>bF)B6q-6u!Qn0C(o#0at0CABMvZ{3CRN|4mWw=%O)oe zVx{^yS3~m9z&kpd$vz#x5awB>^;~~6>~Ebc40xI5c&VZIEw8~lP&3y0s->)ZmmQ}fcMPnF(H~3)gG*j1Xr(UHwvIQ4 ziVpb~SKm@^=D&2fv9VNw>o0jNeq%TMzRB#X>85-0^q)Dd_jkOLjgZRhQWKb3Hl(u* z*S+xL@1J*!{`21J;qzPl&Ed(l$q+^b_x38kb&Lx0{zmDxuo5B2uCVBetQ$3J_3YGz zUt-(*6!go#4JW-i3Ntk~q22QNOWsXA*u_dI>pk}_uYYjN_42j`)x@v`)Otp1Q8=#t zK|sLu)3do$}xtw1WVmTVLH?v~gGGh!cuWC>ww_n+;LioxOGsZ+(M zFqe5fi!RRRhab+!D`L|Y8=c~cw7DQ1_F&|1J7j=Nsqx;U>GHG{aXQBPL@J3uw8lNA z{QHdjA~LJTC;J=yQJvA|$L@(;UcLEXL(@!-Z0?YsBZFR4$788!umHgJRPU|Z;UZS3 zW88CkGw%3kwu8+qK>bs&{k^&CC3?ORcA>I1+1@!)7rfCmE+BYAwu9|ED6*=4(vK|; z(A^w+O%O3xAvbc5d=p>E;I_AV(b0+|@|gst02V6M+ozG;XNqEHeK6dF#PixY3O&M( z`P$h9TZJeb1Y34xNlI{40l#{v={NB0x&$S(si>!rBb_mWl93}8=%oDxAo)^6V+?i> zXo$JPS4NJ+YW8N) z-!kB6unt(?jd)waI3$rKVQB8+See_&vEHpoSQ+UHUO1V?a=>49h<$#!LmH0VXuK z%qc|i!_$jot!h>iF6Mg)h#;|J116?T5$Z@jPdp6fZWS!dY9-#hj$q;6IP*)W|KyQVO!obACFoP>{4#or*SO^iFt)-K}c9Ao&|xPQavyn7&XyW zL%$Dnh3h5xyM${-p$!^-<&tzeKS+R}YkL*Am1u~GSW~g4cmRf2kPC!Mx>cRCL^=A* zBKme$Y=~W|OE?uYt44?(LHLvwje~p?MaQ+;Zg6fQUMJIw3w6F`j%_ha$%9cT??XvM zBt!~=Dls$)qR|NcxbOWm&K3I*oSAaGejpdb{EO7sbWR|-a+yi;#Zf5UZk zORB`K3!=3XdWD-QI-WEa@8)&Vmy_FiN2BVbs^=2>r$$5@TXo&{MBLuUM_nT;d9`dX z+TH=s)lzxSTNS6_Ue?^h{f6uEYi-q1w${!4@g#)WkGaMHJE&zB)YHw7fZpf%&J14NpIZfF?PeTo{v7rP^xwx!)+_GzR0iL*adUJFb zw`Z(oNy-28oRkrCS$o+N+1k1<3*bslz2 zQEuKDM@y{uU8-r?1FZ3_ooIj8sCgcqz^g?%`L8IiR+@yC2gh=UVi_WIM))eG^3y4< zGod4(?y+A6nKGAd8viY{sn{A9Y)o3rK80;~NS)!Z+@q2Ofl_d-cq+PcBr(UgOQ)a8 zdpC7$LTEcjB{y@VR&|g!2TH_pri8R$`S@vtZ`%4-GRJr?v2|Hn>(DpskD-6V5l9Vs zSFDQPL~o+zX4=F%XftLX);zT(%d9frb*Alv<>;mf$5UTVoKLn-o!NCSz}7PE^vD}s zgRrNF(OChC&8FI!H#6Lt6r;^zFzio?r9r79G&MSytOLwnsT*_-CEHg{c~Soo<;ooS zwre`<^-E2gnL1ebszu2Sou_#aCzrW6olbrq`RmLvZPsItEOw>t(lqhh`1#s%xVOhS zL0%udPt&9h{XifC;=7jKbkOeDK6%f$VQL^U9l-H=R#x*9khwh}Gee-ytqWJ$pzAeV z#@w>$RAV1p>BR!as6DSWQXg}G@z|##yQqM~uc|ymGUOolGk}crb`tajW586PeRKj= z+CKTICv6#EmlHrAp`|iFU~wAK&0R<%KJRs|fu}Al&`v*TT%u>7NPUuhjM%sejcrN{ z;f6Eqs~}w#NxUOL=@^$Dh~2MjKOxZsZhzADTBhz58a-{Hl4(x(euPoL8Oe3#-REaG z6KXw8QXQni+RVOEuK-^>|PUc0XB&GAN+`dv=opnkwO~<8@l1Oy5osvkOs$eS- z3N^ik2+VOsS7<995j|uo*mEw|MV+U~#sqAM9HR`ILb0bmiVN5hQ5XSRS}rnR%PW*G z3>qSHgNGs?j|yV)S4D9TOIciRh73w45oA=Biyp`%0&xPFv|w}~lh>$vPBy5CI}0Gt zpDDugo#=1fSH&^u9);hVz^OD`9`Di6cU%7NJL?z zC(?3}>505TnV#_ayk9*59u-9%@@geH8d*;`0QCeE`w@EJjwkUI1z{4jdXg@zl5MdS zg_;K4iYD~IR?0WQ0BAUd|MS0>17`!N2a?4EL=id&K4q_E?5oTmWtoy? zvH5tjlZLc|XP!=fV67mMH%H&}$wmQs!J%I%{Hjup6f0#32j`EHPPjFQye-N|z(pG_ zFq9gI<)Pc69r6%)t)kK-(vSEwC@$gPbQqmFV#2{cR9#&}sFAwD#Ys3g4d6SYDMtv`xD*kO}#lUyjX@JH8&XsxN<~Hmg%4#|6c#- zR&pvPj?3Q!UN;16!C`>yrR)qo*XzPRa&Hhq4M%eV7!5g`yf?lBw>r_$Eh$<=HNMVH zQdkAcByAdYAF^q<+f1RB=jzFvi^YZYU(e=2lnoh0fb3V8BjmO~JQe?#WRd?!fy=o8Oy>n=Q-&#|zt@UO) zZi|K%**DRXR9Fp<>+xE#QJkFg7OMY`sj=~;E{Yv;8Hd>6<7wH@t=p%2UHC_y&iD|j zKjdzcZ_~Qh7NhOjeT++VLw~$JfyWiwtWCb-5Gkk7UE44m;xRtNlTgKzG$pyFjDIQy z5on6;>gVJ~Qte7yo%a0}TbdGtv-^(cvIuVF^Br#g4NerVJ|&fh`Et?aVX=Z87}&rp@=qP2H6}SwcgcxR`BK_d zeeNjK8<4I?lCr;AQo?f9ct4z5$dTSleEQ)S9Fe+Hy1HKRz1hrWfeDX~S#sIr5;v z0W8-JK>Pl6F<2{|M`9PV=uOpe#>xfqRrg>BFS${#AP{X;=AdzLIse&s&44es_rn=X zzTlp}7qKx`k2Y-0+!ICn1}J#QHXcKg58TKYZZn_7T1H*Dx(8ZD&G$OHvX#z-1>cip zMm~tdW|cl*bNzB0Kc%X}Y6$>IauH^YZCuH%{_je&7?{n_ybYW6A99hvh=Y*n% zkzBfXU|Sf}X@OfAT3n`io!oDM#pO8YlQ!^|c#I$R;@mWy_!)%9_~B4WgK`Lzct6c1 zXB*$pEiuKd!FqAe!=aN4iVu=@dz2xsT!SyYHq5g??pjh$HXPjOx?hAN8$Nd+AXZVa z1-3h9qk3^%5~a~)=gFA6On$OCzByq#|Rf>&8*6iTclo#bx}b<%@%ofM4)qI3u>a1E3VWdWC+ z*y3q^74m9wckfQ0s}S$6%6p+-q}L-prP%i)mHMcmQG-=6Y0d8uw4~e$C3!DW!B3@9 zr0eM=Nvu*Ir_kuGS}pbQPa~H4)HPE}eSeNrec)6bWeK`gm^qlQga-Z7{msoj%meNZ zgvmP{Ao&VX`I7I6r`0Is2EHt?Do3e)@)eC5j*(J-1~Jt zu(fLb1mZDPwmHxn{vs5Y!BysqmP`t}gD7WX6<0>}j7vU*bbct!q{_w0ACJ(90=xN- zDL9GB0&!VZ`&Pjf=UE>sWu>SOAvIQG#!6Y=iIlRcREm@}{`D&%dx|LQ_RsLzj zR8?IwHC6pOQd%ddV%1WVgR03%;G}n)V53$DGq<=%d+C&5rFWdtcG5e(JX(4uR zMti(#vQ)!Z#r<4=H0*DkEP^8Y9Duca(PtfUA?bM8KCvI70-b#EFB^5da}m+37+4Q6om zBnADh54Q%xlf^jYwrYCyTC+f4%v*F9XD6~|2t-ES-GWp$w+HpS^P}5?ybYnZB%4b@ zBwRztn>yqUY>TOHS!VAyu3>&-@9GsC!JB2O@33j=uXf1Y*xVTPw@w#J<4J!snLS5z zPdXJ)nmlQHoEQ`vlm4Z^i7CBk;>3k{;(AUS98*f2=v{W4sz%xJayrDzx!gdQU9d2waRU@ivbpPxkgYFePZQ+yM6ugExW1)zh*1B7^=o0 z8_gQlA+PU{JFzwFX#auX=Gjfr%jtiCd;Dk@Kbfxi(7U|;A#c}~^Y#RJG1S+lweRne zgtDO7OiyiWZ;uAUbM@Ku@bFw)9tJ&D*7|j+N$Jm8IG}2Ub(E_^UTCbBZ`o3rpr)t5 zJL~L!(n6i^#BihN&C`G8=!(hM*_yE~LOyi@;6Ib$&PFlrb*}3Q-nxNO?{W9N3O*ky zjisblzm44rp})gr)nQ`KD%q9w<`(DY<_;aEm49@3Oh#v66#O|_`tNU050_v)W-nUm z5qcI`wViSbl*^ucJR|p{r!{N$pY228_wex4sbW-^uK*+wTYHz;dZ5vV-Ji(FFSSdt zxadcqnSyP}8@gnxHJXk0X-yD-mW7?SE*hq~aSzR4enfePkal^}pzh<7{f$0r<(Iz_ zfz|4bTJ4mKesJ!PTmf$Mq2!Al@~YM}5pZ)=BapUvj30m*Z}}y((}$A3Xp@!j=BR39 z?%!o3KOB2`w>H_{IZ|IKwy3uOyoLM?!F(F*6w)9DiEY&u^7d8-^KQ!CluHE~C*~)WY&j zW_#vZ;|4hEru5ud*npwd?HM?42$S&G$XA$x1I|6aFC#Bcw~Ec8ztti#?LU&<$1$!+ zL`tZ%0Y|0XMPAu^1rjPv`c@hu?ZGa&4O2cGjdmFnjC=4z3 z$MYCi9$}9PG;A6vhI(`<-CvHHj8TLbDvcF73^muLZeZ3b<%J6SZ{3f` z)UZz;L3#F~?O!0l5TIIOb;V9GKucn2F1T5%sVy4AQ!xbGg4m=1`PG8wK#g-6;%hef zgH0-5LOvX7HvkD4n3&5!FUjgYzzm$!o-I22;xMb40)mmxCO!4nGZ+cug{p>(g{_!l z8n){l$kkN^*VPoh%up@`!axTGqRK$SbfKmwC7uh_8%~W7bOREe=;{-$uCE(6qV?|) z%??|{`Y>~8L;lp2_zkcU-PIQhUV*=x+}XP$u>!ApS;0~AgnKlTRu>(Nws&B@(o%_P zs4h4-4(Do7mWS8f$JTGSK412$sk2pY?vE$J`MP!Rvl)4r_-NR-)EzL0hG7o}g4Ye% z0(1{@O>?sffo|Wo6=zuYgEu8IS5PJ(j`6*W{wSj322(34>XI7crF0W`5lJk0l7`Am zTwEqsu2;j&*2lC)Kkn@(&so{qyJFW8eQ$ zDV+w;0v2-2a~gG9gwfr_Qa!FRjSIs5n0xSn_5E%KVc!zA35ki}{k;8*DY-|~J38d& zslJ7vzctcoBa-1r>3l1NfEbsW&Rq=#^xcdc3ucV0{dF*J8rF*Dx`cH+lUy$(KX%%AfoJ3`DQX-MdoO@b z#k5y_dg`3p&ywGaF_Z>IfuXQ5u|MtD;JDa0wZ!b}Ac8s6YpG123=-&E_22o8mt7!_ zWMtLbVAO3HFlax?9=841YF_vBng8*t_QT7e3-D_X{5k}`UB>X%r8Gj~u`pVTX<8(hA3Y@_Yd*Ih00*6%p+D}Qw1TNmj zJKN)8{a_`(Tei*>qsi^XX!Go(PurdKuGao2zurgRrKOj4U_Hw)hI!`xy5wUauFSjz z{^8#46Ks3VsmFHlFi(cgU8?&hD0kGCt2T*KVb@YWL}0*JpuN@}pCxAf2|Z}-UF}0J zkm_H(VmZJ;IV6xTg=xz2;gg}kzx=)ZqZzq5nK%$$-J;1&-XS2-pS*Wx0q9_Za*ZdW;h-3-k2j#3B<(SG@(d%@ z^LZ|P4Ozb&SDNO{kq?C2efZHP82McG(Q8(p!8`Sl|KW8ZUlwLG+wi|x@Fa7YF|qlo zvWag~7=Vw=lUq{sPZ4l$&srt#H7f2r-Rc6sGJc#30L%C-2m`S2r3)&5)%d>AaO=!E zG#MJNZ}!J$U@X-fy(_%8I6d5Af08Xeo$M_(HikQ6`9Fg{+$vr(^>?o%yNay=`?r$JSp>?EFc#7qj5V`Z8TG-?uDC3qhTWcD{LbBf^~|)mLlGsa8xysIM+p_0?KF_7XZoQvaHFDLYJLJi zQ)+&pq{SH2nqX5Xf(gJkm`N_hn>yO<--GYnG&SRaB!MmLRsfu`3oD{9gMpT>1&i7J6NVM z)kMZyy9jiXNw9*_6M-(U{N(vQN!8l+smW&lffNo1>x)TFJ{)}6rE@>}fTJx{vM&IA zXx34qF~J$92?peMqgFxC$!e54kdPY$_T}9f6w|M;;E2Ka2x)UaKTI z>-o7;MX(HhFtSUKF-LK4n6QmTq7Q#m@NN_RbS{ z=M|_MDqW%!&`cTa{!Xc)zx*lttxl*)aLSCCU{hfrNi2CH^U`j*e!D|fVwH;ZD^7MW zd7$CirNk{Jz+c)DdGfm8Fv1Dc62b&#=SgZNxS}S1{tgyYC7J?p&`Vbq{}&eJk%h~` zXwvdqn4i$FpgnTc%9R55T;f2js`AE8Xt?F%v7>}^ah1gGWWQq9jI06$mn`AYT2UvVi3eo_Bh1!?)2bN^|*HF z*8cfobPS&QEQz=CR09u$){&UDs;y0;9)YNTNqtokk@V0?P^mgf($kF2oL_8CI~sw+ z0g0}>YIa_NhVz?_2bkQV}C7l4&QQ#bx`qXvp3#U6a>7b)&|9G<#y0#j2r zz7ViEG<6ey7w3NBJHbasP67R)#!W2x+)GpeJJjl@v!Dn?Aa4oIPx&T6M0ArWk}CO> z2j-@n0HZY|UpzoB{lIja6U5?sd^XCLDw!Lq5aMrFExUz6C^=^11;d+z6LCY4&)#P~K%JkKS-kiZL9ogrW&$Sm=C$;??I?^AtQ zCe*{1XXFn<-s+bzn$uC`mm2oR1g&zliBcC109Jh<_;58vKL4!@s|s}&<#W!89kt^R z^njd-2A=u}wi!>A@5Hobq9s3HbqeXou_7z#|+(+Z>YN|8+y_PGPCJIBot=fWYVGcB4*)I z|4n-(&%3d9!qv~HSMujv&2X)kN1L=fSRVjc!CaH-4Mtmatt&X;&L@B06~SE}lQuys zfl#~24cmCGp#sWRh8kvocqs(d9(MS`rb|0OR8?G865(et*x;3`dw8pq&V>bE$V}NH z#hwZUPz8*vd*L+SzI7G@l6Ufj4tXU##56XZ^rOyLZEd>!LK9Ms$k;T{4@}#!yIDUr zLSeV6nHwynH8EK`Q`!KeQY{qj%rIo3o%WEJ>S*5ODID9)k!4i%0(pZgANC|b+^HD! zIpmf< zJ517sN(nB3W?BUFpJ0!uNCHjIQGorv8 zFb1nKOXQcyn6S@oOqeDRofABTM9oo_kSFL4&oj9Lh|A>zOc9=Si3_k0eSb5D^_+>#Oi)P^Ja9sE)E%> z6YRKXy6H9e4S7~tir3K%Q*gs+Y~=Gv8L*@%CAgzH1t4-Xe^q>XC=q9jCXgW7AlxBb zgyzxq>u!Vm1@2ZNlsi|EUEy%E6iHg+sA0Arw2aGzOVL#B7FX2?r>F3EM#$vz>$FLv6_SHzzteduq*?zxRNZy;AFO>9Y zX(a~pXi|yh4232Irj%d6@yMVFETIg3$a8#r|7fCLTihu2QcKN$0C!>v#~)0zlEVFZQVOTiCQ`Wgmu_jP8>eShsT==b za_Xk8lv-K*bW-XgXkqt$2?s5cDW{}BoH(P5ZN6$YDG;X*GX>&{CZ<4Q9owa$v#R!~ zq{6KrLgWtE3YJj5<>VO^EcvMxU^Ekw2TmMr^1zo(P9DTcanHG6Hj{lifFaDYO6$4) zXxQI6Ss3s#%_Wbp`>*du#LEWqhOaHy?cqJVm~?_1@7*3%OltT_Y3DtL7TUdfdD&zv z>!??f4c*De_d1SAMc&c?z36ps8VvH6*^pHN05!cGz0F#9p=5f^T2DX_miZpM9a(p= zWigR|M&tJ8_Mo@xr5oGZqodn{+&;M@M1-h3udn1==c`S)1o|T%rRT*Epjm?+RSu8Y zbTbkzYg>`vt*mK!n>vn&{J=FM&08^G7Ut=4(A;RkcHgq8Eb1O@3S`mb(Xe+tC^ja2 zw(EwEwk~g-8&k6W(=Jw2EoI%i>^K#eM`NW z|I*>c#!?BczvQ*}jotA3CbO@mo9@lif9AN}-|tTHR!Uj#xp#T}gJZ6jw>79H zhAp7ksbI5z>X08rTtJx3>J&Xk4Tk6H6V>72Ib7>Uia@oF*vxTek|CeyVA+6HMoqZH z;l-(X(N^rq|7J*Sra*o#;_BV3u_uNbMQ@(|6Dz%83=6K)nh9hk<$<_@tmFh-nNEf~ z8^yTHN;s7Nt!TBvo42>d{OtNnc6#2kN_J(vxyAXpxkHELf*GCjedT)9X>)x{&;8@x zJn5{=(HHng=3GxcqBz0+2K70?dT(ZZsTD{i){<==GUO8bU`Fg?kSrl=?f$di;KG91DZ$1JZVDOivEZHEkyDK*}EG~wP_5vOCkPo$E- zrRuU~w)@F*R`&L;*d?ZJO#PO}O!D^``9)+_k5BeD`lC9d&5zv^yS#ey!G@+aI4NkM zQkMrM2RB#tmK)SZzl=aHs^hWLG*|#&d#d-=?Qjt*)G_WkAyiB=GMepRGYe4v6l{NQ z?s|!yuY_HwtWCCej?@Kjbd3w(8Du-yzJnsG>L>l!;sD*vvDX9P1H@lE`NgnB1IG3bW1gr~!I}9rLxb3$_YTI0&}v%94}^S6Nawb&B~0zFn7~gfB8kQ#(qDd?}(a20I8e#N}E`!FBRM>gYN}kv^?ubqey-|M*S5nW2Ym-U1-F zErzd<0`~W=j2wy8?9HOT)gq4OyAf|o7>6X%Bn-`697A(EIo7*12}2`&Aq_wCmmOlC zUv4?d2qHLk=5Sc5M^`Z5W`dt$m>JJj)4}JPhV;6)enbck6jI z_vtQ1L{u;^(dAyFu3$Ni_+rR3SrzU@en{V@WSiZ6ht}VMk<)tU((x{B?nk#vPfh({ z#0w7dW^0^FGze5EMD4}Hy9}5Kt58R?Nh89rEpM8SM=c0;DYxv?xE6%Oyu!2~BrFuq zfm~WSglk5j4H|ysl5{&iNPwVgdlk5qXo!heQ?aIa0ESqQ z3xrF$Rh_d%Hh5+ceY-0*#4gn(oC=y%BSa5|PpRvM7wg#AAsAe{G(5RIGAE8I-c@ubOuUAA~PFPOfZ+}1l9RWMaum|z6` z3hiIjb(Q-rYRq8zRu(!9H8uNG`oJSSr(b2DZj}KkJ;P?4U=~5cg?No&_?$ z(gC=hqx%GLPE&5C!FaLP1dU2!ykm@JGl}&fJq_ygHPtR(!Li(- zScV9l5xz=ITs?Th)P{h%$9@@P%3Qi>{I}4iVryKmF=;XT6t=;=yL*f7QOSZpDY#ZV z72P?KnB&`}(@*8So4PiDABl5Rax+J2RR?)CGxYo6MYWmcK*I@5N-a&*&# z%_rNZ4(O_;+r7~>2z!bcofV+iY^t4kGsCS(G1@E!!~UdL8k9OhgWEwo9k`lY!6#v^ zI|q;Lqm8~7bY{$vZ@Z?$9$aeL43E_nSlL%CN^a;p&4W0(EUdwG6xm1qI&(~$_1Ghe zU8%b?O*}V#zV;l>(6x_qg1kO@pQcG2j?=+n!O{fyi_K$Lm>HmNV6! zkeMOS=hlTQZP4|aE@N)lbgHoruJmGoW7MA48mW&tzWAJrKKJ*?z=+-BA;6f713^rtTFQJ#C?qX-@clgi**y zb0Rx33geXPuHv({ZV!BobY1 zrzFy+D%eVdLQSucnvy_U@rdXlQ^B5dxi0EFO*STAOXL`3cu@kjL=;BAmX?bQ*zyYH z3xmK)&c~yInEX{y+{029*P9`O(n$mvmF1!bGKoN(Kqf619mwQ0s-BbafeiXHMR>jw z{jK||I7WXeOM<)DPx((7Z*C0Ir+>o<7c~4mY5!a(f&*;@Zj|6BC%HM%JsiHu5O`?2** zu>wLW;=0bWvFj+i*oZtBs*w!*kr~QmAmR4y*M!^)zTJKIZ}Vaf2 zAw~d$;8XTm#=go7QkE%M7E>YNIRU7)0&WT%^`1&zUh;V0`!7Izf$;Br5q_% znw)UFEy_s1MH??-Tm{iSG5L9-J3m?AK%%;K~s(S*C+}{CfjA z6%)thZ_2!G$ZIHUFJ))&xn39kk$Zy>YB-t`z-Y+f3k! zn}*$oY#Q!1Q>f*+dNSu?vUM(LqhohF4$l+?u*@m093yY%JB`fS#oxExYC3Z|@=(;Z z6#x2>Q2_|$k?^(+{>I!1S?QdV(GFNCVyS8AU0(kX?wcYCnmQ*!jP()skV~6C+}l?!}h4&M}Ah_#Wu&8rBxAu+4;M`JV#Q9Zs>T$hl58%CXXn@~ZQ?IS{W;$+*h8Ed3 z(UVkI4Ug;bTCq`_ob(o||BtD$@ue<`9da3m*x}=8+0d=qr+Z!aN1o335UM}qZj*1* zy4Mz??b>~eOLRkjygq@)72B*$zNL9z$1*KpZraHZkMSX%ggQNm3g>Sg9|BF$UHzQ= zNUB|ltJA*UVoOtkaCYDETo%Ete7-|&PIfDhPh&cklth=BD~<+)m~0$)fqPVXBw5Bh z!nbu}@So{NuY=V-Q_msGIfmMX$T$wT!DXCQDCWyhRXG?_DCR2|&tUTXbW^Txy)%ge zZg8S_^(m=5%$JKU4~rG-z`zC`k$>tCtue{Z)Ln8SRKApURWCjY^#-Ktk)-UemXxrZ zHQom27ILKb5}!6WhVZ8Dl&jk!2F;ud{ll+tvwr_b5>;x&xSM39rCa4{`3Yd(058xi|SAF7aKR$pFnau8E$Va zRpaEAZtb5hM#r{Cn@gg(Qi*uYb!KsT)hE!-chd{>-L&Bm`5bxB-~e7&I{@wbFNncf z={ypNDJy}XQdfS zO5Wpc_Yk}}>cQzA4H`9Q-$8i5d=@(W-N}dmNLKZ)Lf9K9bgvET$J)JI0>V_jL zI_I2F^e~c37cbm#76`J&i_0{xllv{OxZI3`qO*Esa7Wix#U>u(hrKvAO{d5pnkk)w z#K50YkMYBylm_JxDDi%pP0lvHp<7~#n>2nH3V0|!NZRdDhP-kOzVzBK&jPt?nc>Oi ztN7f3fLKMv7TE5bjq1g5Nt8yb;o-xUXgCH=!VL?Gc1W~p$r?6;X)4o-XNc~i|0q(z zD{&EwmmR!c6k0%#2lC)}JKgw0vKyG>hb}MC;Q)K-lJof~jKo-yr_q8THTF7DxtUGA z?4BGO>U6lH7uO!xyarSDm>e5QI19+;BbEm<|3;U|Tmxl8S-@o{ zg2ZWl74m9wckfQ0s}S$6%6p+-q}L-prP%i)mHMcmQG@lxq=hBy??o#3sZ@$|-Lv3l zDM@0L`Z$F~chzdCkAE7m)Tgc~&X#daApRVw`oO6=$`bVH!px1=dSk{6)U@GeR)5Blbs5f5cPA51g2<9=i z;~QJb`$Om2D36hQA2KWpI`Gq9v2U?jXup`Djjx z$%l~652cw@xmfw*5js&|H~%pOCsA1-F3W1)>VNVlSSc$-eF&+s8nb6AXuJERxh;7o zQp&1QDN@$>*SA=yDyPioR$5I}`KJ+6Rdr1;;-qTw*OAgXK^3c(q8wCBP68*r;{+SE z!sZ$6g;Ej2vC=zEX*=m1Umh*J6D!(Y|2fmUuOXH41vw%IYEL;)?#VgHM&*P|r~$a^ zcPb&DMoNM>nY@UEoCNU?ASOX#%{Xkd$Gav=HH=l<&-F*c{?^ID5ZNsgqHXV*rtfRX zg{0$U1EIs$7Aym~kCzGWA+PG)l~^Wh_+IHjdJJHLSREys1O(z_ytBmSy&S;~M5S_O4#Rfw@_x z`VNZ#$gg(D-Pqh1^|wwJOXEp@G+`WRa9=YtSKY&^;GY&bF(@`Bea4B|IVgR{Q+m-L z6mnsnxSrDn$COehdY2uis*$&;aqT#(?cce$|(eUGirw98k5wI?B}{FErN6w`{4*oTi_{IP2_x(n6i^#BihN&C`G8 z=!(hM*_yG^eChRq+Omg zsQdV2f1}S@`Q@)fV6}RqRy!qgXV*i1&hm>L@~YM}5pZ)=BapUvj30m*Z#h8B?K=OW zO;*C2qpFd)f0vQ`aO~yX+GKm@NPVT)qTV`L%4|C$kiQ|APlKI88pI&6t=dA~-s+$p zoLO!RF8NWLstP!U>Md!@63b^CLgn)K;?v0PGexnpJ{WF7_I&MJ2=@W>*IroE&Ux)A z9R*XhC*!iHDA=mv=AEvC{IE@rQ9^ZQqvSiSMQ3B4=!rnL;{(Z)NzInn0ja@V*BUOP z(OcBQ@=a!Y=33(hrrLYB>KWWD2n@Au&%k*@n1shhzQPOuoxd+5FHg6M&7!|$K&CYx z@xr=={72IJIL0-JNC}lT;Hb2_$SZrVKtiQS-%3NIJ=i6;VakW2(T>F?v~+6*BwE7L zF|&Y}U3M2|BckR~e?}fcmV$~t!xpe}R*gqJQBq(95qR-s0j)+6;Cj;jDp)T=6j{_@ zMvCZzLL*OQ7=>BE1p~@#Hlo&XRVtFK3A)}A2NZN!1Egz4dZ2SgMcI(=cX4C4f+|BO z`x~_j-#K9vtfon=kYvriK*4v-Mjz;W-h_@ydXUu-WEacgRlRPs`1f^FnlzdjP$S3M zh$txylFYHc!mWQ-!jP-(2tVW_z_bpx|j zDKAvmf1_NC3r*4xVVAhOdj#d#i?)A(1VeyoiPaT5#Q-gdrMcj2%}-~n!t|*a0&YQU z(t!MG!E>O-IfnR}jUCyNF^=~767u0ty8%eZz{Fe*dP!FI0cPN&_H2OlGw36IhWw%3gIv?xtU{pM_ie=)cE|LlMCJ<01jKQl){$<3%&&BM zb1--7vg+;LuGIy2X-xk}#&D89is-n()QXC_q{etD-2`4l5=)+>p)wPhWyFCIVGx^Z zwzgZ*xUDBKN4Tr}hA`t|z&WRDV7$r*%*#&SH%A?%F9j?NuY>t%vWMUC>VZus^9mNDCX`V1At{_w32&ZQGOf>W|Bo=baa~ky}V03q}RFA7nUTRB zkxF!sN*5%UQcOQ@KVy=yN=*j~n>CSlbjZ(BeG5T~On&(lsY&m>P@xfgtdRJ?Ilwa?o@6yssJFuQjvTO@o z^05$CX5Iq-aBueswms+6W4m~mCqw5h+jzEUlQtD}E%EX>)CtN@`qq`efU!V(tvxKAYTg8lx6A*`h#CyKM%f4jP_o(3?X^^5&R#H}QzEHl3Hy#AuKi?GQ z?o|erZi}J!n!P>s^cU)1{@(u4jNF_|90;!-!0aaP5Rm9k-n+8^bg)6W#*@)-Pz=_` z8_-RX_Lw_)hLLJUA$rr6kq?C2efZHP82MZl)vsB72Jh5I{^zOse^Y>>=BvYuW*h!D z3!Y>y+nE?Z-ACrhEh+k^2)MUrt&;Z|6?dL)bpc=*Kh6b!W&9R|12evKseRtEH7LNV ziB=C?fUR9MzHc<#Ia)xE{(;THSPGWwrw z6|b54yH}E3#nyoRM{}0^rGu-<>j#kC0klE(zj1O2mr>V-LwKweLh)R}y9DS4X|v!t9U6o$0o@U2Hhp$g{&M?)dg!J zMqm&q2){VR?)aq&!eJ@R!i=(H%*g@{I@W^l!=1F$S*w$zavve#3E&N{8!81A^2tu( zX-C{Bpd2gHe4tU2C03xi&ZpkOs8BI5-Zd8TZy`XEG9_SN+@#lNeBYoGUD$q(vU9IW zlSGG*zW>Hk>Z>*8R4b+;)K`}S>Z=e*{cGN(?07kq!Djg%$n{km9})0C%}krdUYd#V za;cEH%{R)mbpj`n1RJN0Q-w@5mqX?k-4!ZYL{lnMz#YJQ=l z#Te9@U{ff93BWg)NoJid>ZBtMGWGAl_imCa;|{fQrwMEkzBzYcMHFT*uoO?c{C(=< z!8aJ%&SSHCZ6O+r{E_Sm2WVt#*5H8L)s~EL|3w}RLCBa1)pGp6vysL<3YkjKJ>iQ2 zxWyGtN#STWnN&(Z7g&Dse4nIhZTr+@v;RN}2ZZ%;LHo0Ud^q^B%OpX%XAD7>>}GBmpi81ECkne!;Z#u1V~k@Y(D-v$uK`TxZU$5O$OK)YN5R6Uq?acPY{U>MEa` zvIvMY^zh_nk#8v^v$d2IFb+E9GG&jvcPy6LrVs;Ci_zpKGjb@+a#>cALOgJ?b({UG zjNBTw)ACgg!=2aJbQAIxB`yd7#-MKBo{<-YZIDlyP@-kTvEOnRJx||x1?q-Mmna1^ zQ%2jw(_#eur4%Hrl>JsGR3$iN#!RrOFpwmcJdt@!sp1L&zuh4#u}a0_NV0>;0}aw*JE7s0lgEw{(#2I0yOaHjdZe=_&sndmVwVhJH4+Mk z4a}UXh+JbnsasmAe&#P9P{OeDX=bT66)mkk4}tg;ctC7o5X4aSIK=esba@G0w(r2> z_FMbsi_x*|(dLqPW=}QnKxiF_X{*}W1pk3Zc+x{FL8a=nn)LiiaMGjFC6XS;^YF#S z`wz7Q$B;U6gl5bB21zrB87F#^BCB3PGvD+^6;`1j$5s%R0LpRrM`(gP1&~vT5*IIk+ z^ht?S4RN> z{h{ugSPrR6R1P~<>)TOKgdlL@kk?1Yrvgb}5k)eGQ;be!WNaz`FlqzDra@9}OSco4 zr5_n?3xK$PPme~$QH;2;0wI698rdyWx}^iQ9O>(#193yt*?NW*=?DZ0+z|w|42fsDUTtCELq}KnN*pZJHJlHY28*Yj`+_@`ze&sO?j_eAy@sYOmk6L2KF^4G> z;)l-6iUt| ztsoNH*l1Jf*t&?BxvoiAY`D^rr2FnpQ(hz=_YK2!7LPV*da%9#@`Sm@)x$p3^~f*#Jh<;;4J2qL5OO!A0gams8BlRDaZ|+Lp=qeFV}UlPX3}0X_@Ha(h<5J7-vm1~eymK%OBxBF+>Bpz~NB zo^|8kFX9uNQ@S><7bKs^ihvt}!kRm(TL2Y_{&ID$nD^pO=%8bktsaG!%i_X@_^9FbO7oBC0i~e9Uh$)`qr~7hWC&hq3D01GwQNXgF zYQccTjNm~4hZ{VIW3z(?c~aVQF4)cF-VR`J^E}gfX*it>x6Y0Y{W5_`1()-ro5#tY zC-_SXc$1fndG+vqk#0IeP7m%$N;g@)!t{K=yM^|yUdtwt6+N{iX?tAvwVr2FQ7|_^ zE(ZPEN2Bt&uFfji0M$lEAF~#pC~2Qr%Ly>T(TgOC3H{}c*+l*tjoO>rqrsl%Y;12& zPi~J&_vVfm7NXKzyngGLt}f;hn2&st-WOvU%`9?M8NA)nY4a+#)?h*VHK8D1_w7gv z77W;hd7dqro6dI&d)=b7nk;M<4IAsxcw;`ia#*+cZNvJv`dD6-@p^F0Y07fX!kQHM z!89NV3uD|M~_%;3bdXw2!Q=|ur^xq|}_4mA! zP2ft`QO7c@Fq_c5hwCA23k<_`~d_?;Mha1%9M9aOI^h_#{Qml+@Qr`Vd2k$9}dkvDqg{?hwaX1~1 zj!({=8&AjPG_Sq0jb?9~EOgCW>XDx<$n#Uf78jhdL~4grE7CghS3UA4Ju*Uu)L8eh z6;C^1r)RxSqLAbkRsXCY-;Ye{+1cU7a9T&S<*EBqk5?-nd}*4=kS!c*YJuI5!7QrN zsnA$70D60>_txD|5zEza>WS`WzgdH52b);X^-r7aA1Hk<(aWQ81(mh=_Rfhq;f*eF zUQ!8XL8BdP+d(_4>L>l!>;Tyosdi=u%M6A$vMw z1O=TVj*iLDhku_=7>vO$0v5kqODUL6u}B?Vrf8>6*HK;ch5Jhdy4w~b0HT}xLzyHr z&7V+sYe7z=O7`Y>xMk?0v0q!_E?8elcw53L)Bvr~Lvyd_p}B{g8r+rDL(}*|UjNMR z^@w|ax#KD$fe^hj$KyghdV-;ECi*Fcnebc?`^#OD-k#zu$RDCmSM8O-VV7qpSrFIP zg8Py>q(Aw~BE17lN-&ElWb&g!fFUz?O;yc4u)0r7#wuEU5-zFwT1H2Z#Msai(vJ}V&Cbmj1%$Ou zK>jVdb6R&@dcRADFfZ$R87Qg$l<byAwu4HMx-2 zE~a-aNOCDv_I6(jLSSC?v>*g5w4Vh5o~oAxA-be0x;tVK)WhXVRp!(MIob(5!_5U9 zkDJsE(AiV{h|=ToeImE|JaYHoWKwQbwRVzb=8f3+9|? z^e;(PjUlDQ!Ku`ty$lI5BW;zWhBB?fPa6W_p893rDL3h+_1{vO##^&7Uz3)Cr_>EU zLV^-hI6fGDyMT~5hrv?XPC%m&)@K_O14v^7IomaXp~5$3~Oo0w(+7xQ5AU0 z)q#Ar@T~Wes5hmxGG42b;SAA@KX&FVy!%`c3G0lzS zj|$JW*#I25@5;r}wDH{d`Py>C;Q7jAJWZQA((es0?%2I~&xkNpkR}{J@p@O5BS-04 zX1l-jkcGgUTVG$PMb>M$jIpKdR^wh=8AJlls=c5yavyVm^0>Dm`=EfpuQCvc1UVw? z40J{Y8wvV?(ZBcTMxkDG16CfM;?|RQ43Nv)K%XF`0#VzGJf%CK-5yR};ahm>lLGDb zlSd_b2TGKurj2QJZ(HI)K|2B!&b+Szcey7C_5{o@t~{KIZ&QD~H=>3@K=mgNR~EWg z-ROA-F`{|f_Y;&N_>Sa{F@D@%{c!_mxHIKZt-3j(IAWmyEq(EmB`qXj#=A26N^(`1 zWO1Xm+h&O@9__(1WIb{rX(34)4jGa-(!A}k!&Nm;3`rVrbsds4K2^1?BnVXdjNFg} z#)>C+k0uoSIhU_c7pSr+4!Z%4k%visS}1Cw!)^eD;jkOWB|7Zp3G9VIWG3gAlAKuc zt1`KVB`m%-Lk6L10Av)FEAGe|0OB~Z#(~L>ta(N?b5iI>v?GK0ObM3n%y8?$%8${X zSW0jY|Ec^AW97!cefk^TerLn)HSiZqiHQKyoHq?u>m0n|ymhbKd|=;Sw)*o#=iP=w zI`1o=PpIi*)Mey+LX#D#`Gn>fxz0O1pHM>qw}C33fTE9<`l(r)r1|hizUlb{PHi5f zB*Z!m*#~l^geQ#?j8b|7o0L9xwe)=BE4QZouQFh;e2K+J0zoS8O3Pm-Qdr~wo{UshrBge^G`YpV(P5Va8N zrx}m?$;*@T(Xs)87Z%b(l1GyjS3)5J60B?jY+t>Hp6{5^Zg6arI}9qLMBBoVM6V># zPt9k_6A(fXUpdc3E-)fhAd+}7Y^!N_D00vCONE{A+Z*yGf9>CeC#Zd8b2EwM9r^46 z!YG@P2)4|#DGJ5s486~h>?22l?gZm3}8Ex zGavA~l{|`DL3l3(0I+g7#@2#(CMdnY6XyNN>Gh^(T#r1fnL4!3=7e_T1fSf5gIfFt zA}N*P$D7{-TJzhEEz3gN_Hr9X>HZU4uxF&8AGp{|St6OrUh)R4t zbPrluL=$1*eUMJ%S9*6bbER&N1Z}al61YF>aT??@Al` zb>!7aWhwFX6VowtC{NwiR&ArSTU(fbmBW@Amch045An7sA)%>LB0@^Z+x$x|UG{K$ zYdl-um|PeKKOv0jEqOf_bjow<9DfYPEr&;fUiG9N->dcr+IyDBrw)U#6lre4G z4uKKbx6_+cTn{C{HSR9v}9+oH6je(XDiKLcNQ6YcTBidk+r`1z(5=5SHyUNd< zgnR>1c_b_Pt2wp2CPv!aNlpx&9Wv6!V1HBuDOJN}>one@yf5iGQ|GPf0=MmHw$4k* z_$QKmqv3qGer|iZIfTf1uz2dZFSx_CqTojI)ud~={wP*FGu8ntkA$|5Vm+o_(jO8R zTkO(yRc-&mAdg4<6S&PgY^pxB@WwkP?e#JBh|2w@Qg z3e&QSH=EDlQwJPkG2I=|$E1MH(2<;!yzfpLSM zMh!*`jX&ex92;wNgrS#Z4{Tb4DhF(ijRm~rkmDoCKQ`n^Z>#g9Xe5xSLm+`~p=>M( z1n;Dto)Jtze%~hwVKJjV9RG)gLyj!M)=)Z%MeGpKc;tn7OSy8te4W4>vc5 zun%}Rk}B_a&7Va|UkaQ=SdEfy5XS5MAv>Dfcmxt>D3O#trC z_Yur%Y(;4q2II1j(G2ez%YZd%T<{}Mht4gLk0M@Un>q(_BVL5!61a+d(Sk{-chbwX z<6b8P--~pAU}{puV(iBgbfdst@ncS(M5TeaB&%(!_lhsH94%y}T_2KbJjZOW{laVyLQ~2}*oebC?Iow;Vk%*th2!{Qrv9AGwhB z$rq5qItdlemZDr#O-2GIyb}N$rGgWvaMfTBI)gy8@J>M5O?W4cM+@)diR|@Xa6qR9 zg5>i^seB2JjDflnPLw*i0NE&<&19ghvDY~n=EF!qkN{I8k&uHR@d3mjNS;~uMvIPd zHJ5LpED3nmj})vEUCwPXRqxHIc7?;gwST@MBMi4M53xjm})ouAwum0d9X{~x@0 zZI8SFTVv`f3%47A-7&>3O6!xjfjO|cPqg|L1^M9~xeuEf)8W?n@ycvIoX$B#c26yJ zOWLY#*eq#xlo*XS=EJLj5=EF#d1R6Jp3_GA6ebgcYfe+qD7%!nZj4x0AT!ZWE?K*$ zDRu|eBd_X_d$G;Jb0!-bE0=~F7q5!=Do__b*w5+LFh?wu5cRY5t~^oMA@{CCu%(UJ za4h91$R8|{mtafp@ylDo&B@ud`DA|f!NCFLk7YOJ5Wqs^xq}AU~+PHrmV{lPo07GpZR2GV>}!5Zs|+f`WvNw z(m(eq`hKVkma<&YmrbMVjva4aY6}9He{_D#rx#%r{3V+DA8t?!m!UmoE6U^uJ&Ft) zJ0)igkY6gu{rO?d+CvwIkoY}5Id^V69h*x4E`Mb&{A=X_@_~ZwibYnnO%0y8g91Ix>#X<1gjzB6;CYUK07x1uQRB{GfQ5Y)$_rw|75me@{pq3rH-5Dy`U0j;!bXC&$k0X08jK@3cqsb=3*Ecr?PeIdPcWzO;=XED^w3(_q9E(XG6JT?ktrlf!u=XVz5dHGtgIUa5qI@1C>H+rs#+t9*B{vqpq9H&|X$VOM1 z0;4TeG7_z8_mUS5o{!YJ)&R)XwnkpwCwF7=M|7i|N)2c!)P_#9jHhE}25}+oFYgZZ z?KzZ!yc(GcYWEp7gT1?I0`iFs*;G0LuRJec)F?u`UIV|Xt(SKc?GA#L2C`7pgU^dbuKs<{$0r8?K&F4{k)8 z-y??|cG&B~)ujvmOItG*8T0$div};q%$TdDQ<8~(FKlxbOmgzCX43Maqv`ez>{nW0 zT@6)%L-gU?Ak*^jy8G1n)1Ofu4$GUEWQ zRFL{7so?iiztF?*RC0n;J|jVuQs#MwDN}H-lBGq0ytzldm&#jA_qPUGtwhasqYF{q?$h`a z8v#4Be*Cs`TcGD<5H(lC-az)vp)xSr*`AHpAAsglIZ#+%qq^V^=gCi{D2jzppeTGz z?00$)oQ*fmt#JK1f}jrdSyE~gN`)CYzMYTi`R3KK9@Vq> zt=AufG5wFgKl|aIWAIOCWU@Xy_Bk z+5FZ}wJ;SYsck8j6J=~I^Q+L_LTguj&zw8y%8A?Xw zi7Q<&7OAhb#pi+9a855;2UdqL3#9s=o_{#PKzTS&uEa^oVJ;-a^39siKZ!iy=zs~m zAPl7rH6O##53WDV9w*d)xJ;b|HTU5+0fvf+I$-Eq^zr3s*ip>@>kuAEqa=pYMG%S7 zrSXCT-ap?SXYN%3F|)c!LA6t|fU3eZ)#NE<{GAFnd3Q|hBMdf4 z$eTC>Pg;F~7{EvVP0X(jC1F9Mjqu7sAee_a5zT~ZwL3W|P2RUi?#z)tg~5aUYZbrO zD7(vit&3=u38P#@vrO25aAC&hF3A7&vj?Y>tqbcgWN5a&IhYn`Nrf1tf%@m7{l}sDf}=%F{3- zFFA2yZavPW0gv|bLT4>cl1qJrfJYKZO#%mB6v+pBnYSHTQ9!v?=Gj1_BsJGf1g939 zqV(1td3KKK)uIH>PVC&u3x3%{W(-#uev{^1Ag!e*QFU(S8kL+tm2eYv?ktpeg6fu# zd<%zy^}q<+!u9CyVj4*tc!NyzX^k2M^Po%<$9(gmXuFTFxmo$uy603&rXu854@cxz z!IOsPyczFAI+eH0ioa;)S8;4a!~-=mY#Mo;O5RMK)2xLd!3sB$NL$~Ku}Z_0B$+=8 zMU!EnDU?O>>MUKPhP`U(lD9eKl60e%gsK3>Ds6DGokOhLjBEJ$doKmw3 z3>KqPv$aj3FeakE!3;8EuiVd=7U2GUblpu8GvP=w$ik}vV3d8mA__5hvs8jJ`A~EP zL&H2Zy4Mb(&M3T-U7-MYluFx;671}Ol7Vfw3x0b{hm09eEyj<$8)?*|EmIk?CvA}= zNNS`H(`0fX0a@Vj$x}mus{k4jJ?CmMb>HdJr@W?_h@@TaG;S{LjS|!M_ zkC%<>3x~e8NUFA=(rg@B?3>!}(Il6@VaZ|%$zK%Yd*UwTwRx(16gaEu=9MdLHGRt% z3n_65cXv=IjBLG7@CDV|^x z5XeeazOaP9uwYjf!3(2G2fu~+HZ3zbAlDzgPBul;z+?)mR72t#mfg2H|%|y$>gSj#riyiK6(gv0;0Yf zQQwph@K2E!L>qy?hw{fErgx`@R;nAv%3Z_DE?=jW=|#XP;4HFORGBCM8{S( z|Em96;F}Q=AN0^n5EGqFgPtFX4ti9$WY7~+@=!t_Dn5~pQ@~)fg-GFz_y}V_tQk^& zJGaj92SsdnBfbwmR7s1}_>3nje49s@wSlj#ToxfmxO=8AG9qINfuTqQq#gh##fEO; z<3GBt zNL`|G*s)sQj)GD&;q}q+sX!7~M3Kzl6r)oa8Jh|KjM@+!o7%GMFiSr&+!g?F|DGO= zilZ2DV+BI~b~UnF3WQptv!VXF=s?^Mb+(>iMLGh30(S&KEkmN)1zcZX7W5bm3;5G3 zmL8#hXg$3GfNz)w`oFS1s_xADeAevF3t3=d8$48~#WOD4?*=)XC3)eyWc}S8^mob$vXD&yi1* z4@GeL=b~?k8Vcesn!H9Z2eBm}0mr(Hwz>R%W3FoMqbZzaSi{t6#z}gyuazORGT~*&pu047UO;XA&&6fHb0rbM>88MH2VH}EFV#ZPx zD9O2$aXsg5W?|KYs-9vD$k5U>b)=fw$OG|lW)>EKu2NR8 zhV;P9yhtgg>fw-{$kVvvQFS7$W0*qHlpc{v(TEcuDyn=j*-I#Ln2D1wJ1A7em!8KA z)0Drfo~CS=NgEm5*a{-Cjg2;yj;)KBnMdJBA-%*{2Tz7U zXUw+Rs$XbKDiI0G1ud=V2e$3F&1@JMA+cNK%nh2-8JpaZepo{f)XXq=q22b7i0T>M z704V`IZ~75L6;=*DqlPt2!O1q806#LuGBcVXySJ&;2AMEDz-tRSQ1KAuKk5SoFI@V zb~g!xaZ|+Lp=qeFV}UlPX3@)3NjgO}GW{Nnf|@TaxZksTTGt+I+U%-$Yyz)>Gnm+h zPWt6E__RkgQX@^FBh_8@0?*N8kn8r0eJ>;BF+>Bpz~NB zo^|8kFX9uNE&?rLkANn=iH`pTx1qAPGF6jDE;-diKzCu9Yor zyT$KM1lDw>778t8@(<%UMwQ&lm)&{?=L{W}uO$NSBQq}ALWd2pAO;DmND$MDxQHPH zO1+|yAIXZCnSfVbnEiMHNJ5fdqV*A&@vUGX#?7xSpGNE!wYU z#SRh(89QJLSVsC*fM*o2;0VI@UI}_MGlB;J9B%L+j?E4p&9$JmdnezDa1b=A(Z}PG+uO8kn(oJW`>A^io=_bopn4S-Kx6t0z zYuUuHf~}hw%`fka*}55el`*x%S3kK}MJkJ)*P3Ii@u6e6lG_{ceNYH*J!iF!R zV?7#g%!gMFYjt3x73<&XV|i7^>%ldrDa$Og1)FSabbVFU7Cv$Jd+8wwfY6Sfu|hajn1SoooVE#HGG-*@K;i6S{zKU4(^p zX)MUQ7Co!~f^j{5d26^iIlDHWz^dTE0p*XLRbj#3U}g&|5Q69m3$DbxkzJ}gml9FC zwhKp3$Si?*`FG=lS0`aB=O&C>zRb;(^3uWL?mMT39)}d<<$bJ>vc`FEZT*91uUD`( zs49ldAV=UWC8KN3kbh1nK$z9)9KA=4CYS0P)$z$CTa;n?x!<(V!^H(xIyyNjgG z_2whmCpg@oHYZx{&7@~ifs|rpWQ%D!cuzsxYmgi+Z0(_o!|8Z*d~)vGcslMj&HJ+j zd46ix;(}9_NbL#fxE{Pe>5&mKq{h03t$5lAJ3Z@t5``qUsQPCG`F>Q~)iVZ#(HboiG@KUj!_Ext3BeonnzXx=hhdpU$E>U%!$x z5vr(usX%wzVgx{RlYc0aBr-m<$y*C@B2}_C$HT1-`)IzB@V10gr~z7|hvr_k;?rFh0wH>5j>m<1^aMlSO!QL>GvT=)_LsXPy}im%kUvDB zuG%Yu!!FNIvLLRn(Y~bG$zC}c|7DTh0VXAw#S}96(WSsViD&N)tnL$&v5HoogiETv zmeCO;F*fwXbcwZIhYIY6!NgYSh1p$fTl}@9XFuG>@Q6weOme!{C@WZoBRv_?2CM3O zY2KW=(cvOy2MA<%Rou}r-TS;j;SZF^B0z6eO3qo{BRg^|P)rk)Wn zkzC!C2=StNqfZ4*s}a11)}PX)0yvBYc|VGbYq{O%*hIcgriBZ2zh;T8F>0?#c#jHs zCrTjF=tw~jCEksKs5AiyZ+ou_`B9Wi8(J3%?4d!Rvetou;M!UD34&FsI!{Ci!drJF zZ_$U--*%`TNtHQuL5_Ap&v0`=$KxhXbzBQ8tSs<8Sb|oDsOgWOZi!M3}=4_y?N)yZ%M^BYb;XfW8<_(L)hH8lliEIa#LvwD{3m@)UV<`znhU1g!?wDp9? z=%x+Ft*e7rJZGD4l( z5j+{tsN!^OH&$C8m0zOE+7`$k`nJO!TCpJ`e06oP2o(!P4V|WC6eSM}c6yEcjpUCC z&$ihB9J%kx#nQC#-1zz0a>U^I%49rEn>v(dqt-!mplR1skK*^NE{nLUO%x4ynoi>%jh8DmS^t;W5$GKd78ReM2a$noT9HyltN z_f}*d6cG4T1|pFlM}(b$&d6XRL0>Q)^4kWXUIaE^<>4uAJ$c6fxx5YZ2~sK$O^OQ1 zD|`!2eNv#^e)6b9??8$2)U+{@D(X$h78pW>Gw-XwUG7POJpnU}D-WmQ+ti;Jg@Ec$ z91>BK*3Go{PZ%aS!ul~3JG~Ai;s8-#aP#m#MO}xY-X(0(S z-j&%`lB;^G(R^&&cZ|zT3rX5=TrD9<1Fo(^lE$a1wv_~dYM+rClE7H;1n<#=fk#30gzExuDBy>0Epwr8V4pjvgR4p%*k{|2J@K`EZ>>o)`OKFqd&2f;2!={`5nf} zje-00H@yAMhTm)8FPI`J(3SJ1;cA_OzoO2&4Tp5zj-IDwNBy>)Yar?Qgql7^T}I9) zG+B|FPiUTz>%7zR2{jaO8>sRLDEerrpPIEP;I_d1Q${|4Q=11V39(K?_JLd};Ys5J zqm*8e@T370MtIUVE-^f5p1{JBpr}OHU;>Yd5-)jD8Itk|hajJTqCYVzk)BbIDnYT$ zXVyvTqf#>pH84iCfIy5}fB?!g9}8)%2I(1v8bIOmWfeA0*pdUWwwg#E(yqw*gOZmg z=c8o<1TQS4=Mx@HR$K{%3`in^4rhS_wy)kp&v(pdH#j!R9R`(AqHW~j|1La1?JJv`Ni6TkXBQAg z*|cRp!h#dJ^E%NWG8xoDkR19@pO&psVeKtKbm)VvoNqEOTB$&AT@H*5gdRv5BQWv% zTDk~62Uq6%DmBPChLJ2a9=985#Qe|ek}G{pr2yY1TEBAWRfQZmTH0#!jSo29n4~0N zq74NMEJvSKR#9P+;YWHHlooIZIE+ReG2jp%s!A6LVx*oRjVM3c`CS0DzUteiJLKF|GO~D80ZF=Kaa(^`_aJk36ec_p{IDgm&c|dW+GL!QPEg zD#eehxC=_!0vy`5mxCF6ZqSE+mBCF2bvv53fzc4dHF1lPB3Rv$BSlo=>mippq=I9z zkOu4}JN?&NsBPvzD`NF*%B4i>g40IB9zh0Y08e|Fb4)o#+%7OpjN9e!yQ-Rgad~x8 zSxS8U#B>ZD%2R1=8%>{9vkLv^b<1aC?)$<5feBbSY^h-xTwDJTZ<`VlnmQ$7EY#cl zODV!Ti=*m7zaNgY?q*%*`U{BLFavL9qNz4xaIIj(5s%*<9pQ}L3{7gfUvV# zudTI#Iey`yBeHL&H>tQBo{;0U@y7VY%q2O_Zw;&(Ni-zvnK0oNg zzsh{V$3XpF|CoH2)V*vx-LB2YjCD5*XX|r#T(Ql^8_Z}fMsMMA|Jmkj#03+xPwsNG{*e$~GuDvl;v z@x#|J7xLIx8Fv$hQ1~srf$uA|o#g{<{ zizrZ-mfV|-xb_7YethbHL#(`F2W0o|Mh&8PgOx_DaRp%q!+b8Su%Jjs16D0q!zM5d zWje78;dArVo^|qeq=2_UMNnRT@qT7(0-?E(2glm!$}eIUf0#oz`JwYmPB_3mno+(y zhY=Vz_-WK&#L)ON{>`zmMn@QWS@yuDHK=mH=Ga)kxkL8)6)D8Z4UPPV@V)*Cs8~vY9cQo_YGbYQWFXOY7-aw#cH>cm;U5kNU1)mXjEZCd(z?p z_IDs9{8T7qxE_>tbfo$Uf|cqM5E{)@r>Q>iX~a~YdZxTvragrC9i;4ofa)X<(6z+O zt>w{Jr+dF#IZoC5+(a7PBcomw3dz}ojbNBA3=PB(xkYWTjmnl55yG=C>e)tSXc; zWKDm4ixsL0$c$#C(@<4>8ZlH=&&&-~zkn3hNvL?X6y>67G7>o9odDP<6)sofj_RcT zV}*AD(r&^#aXeahCr@;}{=11>w_f$StoK!I3dg zcfyHMCl??ag%ctn?v0b$oymuhf*=8=NFpHzLE;05L6AJN?u}N1AWKWMpg0M5*W}2C z@r?VW;dC~T95AIZ92S^%BAY|JB&2Su9je)8hMy_so3%lDWC(bt&wltw$7 z7LZ5#WQ~SSj@O3KyaM?d^lu-H%G0rlmo|!!gBe{sOPhYzCtIV*+3_r=TGc+imMmbH zD?A+Y0arG+NAr7*R+Vdsz>g{HVe<0 zY;3GltLEzN{$0aoS+@#-y70k%PQQjZVxfenpRIT0iOLSScLf5fW$it@LuU%|2aDt- z*wTCa^44&3a&~P#nV)@da6tKE*^Rlx5Aq;}#D_II^JmwRD6+ zJ-4yFJsnLh)pyh5lS^G$7<6A*%h%P$rQh!0f~qCfQA&p*(O4hfv8IMsCe^Yp`OOaU zgl8rje%t-HQqj03Nrub{FqNK!YcSnH1$8+pcXDed(2jp$q{-K88&umNk{|vrGng_AJ(ir zba4oY-{X^W=f=}=xag*J zvBLfcRzt`uo>;_vc6PWi<+@$a_zgjQEP4vD z4R49X4#9q6HLfnUuUw!|+;7IR%oc(WV5 zLn&l$GP_gPx^F<*7DbA^`3?e0t-DijF%Tx>u~8^9B?Y`qlkY6Z^YXP~b3EMYuruu+ zvfjsWsx^RYbft|%SK7Vgg@flKb)_}FmDiE>@;*#RTGd;YzQ#z2&@BxtAQ6VY80VeuYq6H*2_DJ8Y^F1 zLlBEyl1fY48Tm$mk(iZKFm##ij@s$HB!L}f-B3X3F1u0`2<)c$^FFTcR=Ua%(*8#2 z!Z%JB39B}!RVUf*@Zauulzuh3pEqNolI~>n4zk_BSJ(JJd4Fd=XGo)&0CneB7ak>N zK$02Qs^8DmthL*YTI(buzp;qX%OmwsX$@<4ilJUznE5MFgE876h6-cV9flffQ#0TY zsorjj6z>=#|B6x=my)D)gk9z4?rkW~UNrn=Cm2GjmONe2Q;eo1d9+~MRV1eLsTe}v zf<)5L`PG5tK=pIH<7;>H$Q35>>nZs(^5syg0Axg9axVuhlGSs7d2>>Cx~O(MOytn@}j{DGSlj+>6B!m-wWHEwM0JZ zU(KZDMMu-^9hf<>!nzu&0*C0sxk0Ao;dS?^_1m6S9uCW?^Hc8_&gPP0-VP4(v4T8T zel+e|Dh9ko!+^t~=yPLGfayWLZEj8`Q1nAbamGzQcx@(eC1kS2@kVWA$r4#S+8Zpu z0%*v3{oWSa?E=g^&?<)_jsjVSD8lzseh6Re!PF7hvBK@1gU&R zf-0rV^A1y{;DPa$c}L{UJ@UO&-eS7HHPC7$YPKWg(~UKl;rpfb8X+5DzEY4=(S&JA zf9=d|%UaP)m(-3Yl6|A$e7JrN)`<@1gQdaZspr1n4%dvkB|7Fa9ukJ8zS&dkK8-)I z5wJ7s$8S5g1$tfvQFBEs(q>peT=G*Xiee!YC<f@|qzO@LKR!gVUOZQY)tU1gp~9)I)(7`(_HHWu*1e)#7Y zy`+Zcei#^YD7?*fwrAt@BNhMd+`2TL&hHsdH!seIbllnC2JNrPTSMf%arDX#oX3UL z)!DO1em=&OnNbiQ9_&5C(erLScC&|DGIZ)LyLe043J+H)f7vUjGn9<<-6|(#L14jH zq`uY`p9f~cIlX8dSRKMFkm`SW{^1A%<>5fN5+^B#1!l`PYeN4d@`R%UCiH?ZlseRW z3`ak>{_ucq4zK@k8U9dEb02;aV5pd=1BSjuA77q^9n}o54&jkBN@6%&1d%9R8mq(M z{qyZ{=3XTbGh4iS&u;ChrGKRUC*0bP4tI|d+zgZC3fM0c=3xMuU| zWHcVF&o*G1B%LvL_5ugh+;|uaTS49(Q~L;mO%n1Z?$w{P`UEkJkNn#=>i=ss6xmCM zV7)p3^Drl3GpFZ>D(L(BIv=P`7bNdnBzNY>pTgk5{A2JZ|M5jgVV{@g>@K9G+W;s&MpjkH!jfX;sfLJ zlP&%?#q#6Hf$_%1WM|g=&)_e&#xHIC?sa6(cx%Lelf}(=xUL^UHU}^UIlSWJ;{=CJ z8U%-mL&>{Vt`K8E9_s{9Jd^ODMe==UCIJYj8U+k*hp2L#(GkPj!eyt6Ll5tUsyhu3 z?}edJgs?;e_Sj?@hp1W-p35pRMV$xsX%!_%2tO-F??h1r;j)ycVMbm?Ty|W=6i(xU zJle|(owYnkF7*)to(LkD^C;wlz0BK=tSF#dEAwoiQIepP5S&p66Q#HI$g^`)uNEb6 zc4FsNUhvBvGGn;Pa7tO^DXPxRT%(c`s1k0X&Ygu4Pf*Ap)e+*zrhSLV{h(ix4)0ByJ=!397zUQcvS$5 zvaeS}AqH=j3icsFEy;(XD;OH)snNZ55Oqf3o$Lw)z@t>!Zj@kwZiLryg=ggLFy zK(!b@@@}M2kG4!@$ey%Cl3HrRRIqf>X)?KxfGqI%uM);DG9oF%NIKGE=POjoPgQ zwnnK(qx}!3KpodAL5_XAY+PSB^tDA&wFQ-CMz&rk_=0NbQuB26 zd5hu~Bm+~rZr@apXT~8YwoEAA(%+auduPHMQ{bIfq->~g$y9(HGCC%nLPh@xhwP_& zu_7TLGe&~8#6T8X%1q)dAJ#h84w7#?SfXA;AjlxwHxwGwxTQZ7)q>b>_O8$hl z!$w9tJ^#F`=Xw{zXZVQ08e@~A_#Zip7u>v7~yBgUo1wz#UTaNT~(Sf)j>TEs3igW}51?~uf zT82co3%EXLjo>e;Sk{C7L48pbnlU{_fFO|tSDhfBBgiBPddVzE5{zw=a<;cpi^ zK2v)?Rn8OTx;`Gn=g23@hax!rbJ4d%Nz@tz@fS^ABbbBO5|Dsn-A3D7e!nr=<#A7^ z-x;yW%25B0q_qN5{Ri349GO&+L}w_9KbDcot^mdEc9L1DQ}!2>)9e4W;3$`SEJmf* z|7)l;Rz;?F1aG*0h`&g!|2N5%&aN*;*VZ1oIG$b(tgRs^2kdf<9%CcOK>)q*c}C1* zUl@lXmzc3s1xkRLf@ebWUd25<02E0^1eS$mVJk>F&B8*OI#Nw-T)2Yz2|n#zvb;$JRy6%oWsF-rZTr zi+aq2OU)v)s;Z_$-KAly4C$+R7wx5y7I`z{&0(pNO6Zk z5l{gk>sdHh0s1F<iiw7uszPiKw38U4hJDm7{OcdRz9AYmZ)2g?i?)5_y#`9u5RR z)>I7gac@^@99%TrD#BNk_Y4&vLoV5VF1!%d9R_8stC`zaqt)M3C?jq&ay)(cZMv2INyB)y#-ybivtG81ivntD!m4~ z(cG1mV|8@Jl+^Gl8~H?51l$l5*4$Cu0uU)$zBoNRYyfBUCe7!(UP3vQk`9k{m%9!8 zmwmSlK!tIY(G?Cg8GCZXtN|o3h{~t22R_=*konv@#>}gz6PtJl69k z)1(T%MC?RcBhMS$?ca&!6_qB61cm`;T4YuWR+2~dBu!kf3N722+?24%P&Gfbm6trC ziooD98kG)`N8*EwNx5#bNm4GQY2oQdvTnr!tCKu9)sh3Zm8Pj;2K_ic+}s?(L+jxv zLNn=9zKmPFfL{cc7>5Bd%8ZPe0LmzvoZZfh*+I9bdq;M>zyweWD=~;igGw}GD2Wvr zQhpN0B7-FGfHL|a&*|Ai)45)@xGhFI@-JVNrIE=$4AwBJC0i~e9Uh$)`qr~7hWC&hq3S1QPIJhCt%b%n(SP<9cr9wP?SZmAI8a z$k+i}z%tUe0z9LDK?GrWiR6r{grDR#yo+e> z>a}b_S;1D_jOLej#&p#U`C8AjsVJBmAQyxF?W0loTvun6Y=CNmt&drYPn5LJtmOn4 zvD`JK9i3R*H1x?|qfvWvdo`6>B;R;>E7HC!$MS=*U97ll_Wla`N${feKDrd z%pymX!Q(pZ;&|86W?4L@v>}kM`*x%S3kK}MJkJ)*P3Ii@t})atSF6dwVp7qu9*sBV zLyooOqn!)5P7QUyB>AmAmRDuG9$a&pvfQ(NvSQR`tp#0IZDlGUL%xqx=LJ(bH!IhXdvP-py zCOdaJqF8Q^bb54N{@pm?)k)Z~xe4QzFLN{1v58&wIaOO8fV{kq6;jqX53a3$@a**p z)&^C@uo=|31a-P*5&1w5 zO9!+xYDN)9B&X^_JJBcDOncpWHHC3~M(>2Pf!dVFxzHNWN`5h+^lp~eGn0++V3GbC zE4<+Zi_X)U0pu!0S0$f;g6Mp*vjKE!Y*%)^V6NW+z5y*)c*pkEOx#_c$g^VmSIM4Y zu&}(iuyE{n^YTnrd$)PKNP0&XXm%G#o$Jj<-3RHsv+v1kDF_Ehh!yP+bMtK-xYz0H2J2GI^Sv7qapHrqc?`d*@! zN8<`AYxC`$6LrEHUE;i?63&7~JJ_~^c2?C-`mxyovRh)G2`m=m@OSQ!FX0OrRC}uz z9i4EZ+!e>g!4@If4uZ`)mXeZ!sdmSD6`Xt#U$0qHLYIPi4%yQg zBdFl_>aNZu^67-Z82ln&@yoT8g6R~C)X`;%cKUP{)d^3n%Bt@kPQFy2yKON7AiBvv zlu1%Ez4*PgASY5KdviS8>adUID+zB)IE5OZHF{{23zaSMy;n5U+(S+c?#gPYX@FsC zrXj!ABkuj>j;oA3U-Zr#j|=ta35LFz=%*NF!gE3FFLy~*`z5|0e~3a|wO0m*U7n$2 zL0n&>eMzE6=1S4{FN^dJFe$+-rie~{6rDI(BGhkKM(eAi7t$)Q)CN0Qo7t^~IB)OC-d%Ld%Auz9cS`Y#j z+RuUj4-*dz14dPJcf=y7hviQ*%8XJPwEmT=((HUaqXk{VRrakUx0uK?6>Umyz>o(D zhU|{0*)K9Z`phD^x+@XlMfFCX3Yu0Uc#o+5ln#}HydOozwcKuWY$9JL)53+iU$ey4 z7%t?krW0?o3VA0=AkyeaK@cV0je@8&0c8YxuL}85luR317Ygj5L7=kMfr5}YV%;YQ zR;lVd@m3>*)RVUuhSNWz9!ZrsbwSQ+LeFq>LC53f@_ix&`aE*?;ABz?RJB|Ze@bpT zK&v01O+0B&-tSvkDY9jw>GlpZ)~&EQZj@huiD`|B__Y7*bjsoJt+q z%a9;5(pE`osC8g(lt4h-Q@;#62D{_Ffo|ch*OfP3SMbd4YIWCEF=di#qUIG)kmWhPAX{ z+4%OI&O#UR*}}8lOQPPC*2;LTPKGm>-*5s#gTcPWACi%%sp%fb#-l{bIy^dN%)P9+ zg_2`eS?D=!J>fCBX~S{r>(>2b>Z&n)HJ5*P*^T5izD3wAV02P|VDqhZW@IK?^YL_Z zJemyWDq3r)|)85L`k_o{?NA__LdbJGQw9^7mHA_VARlQ zT1HXwu=HU*uXQ8&qr$UoHULNNyK=EKZ9F%AzP21*A9kal1bJmLo~BJ5*4cBQb#+_( zp4BCohSP(WKh35h0^^R|oA-Q=b%Qx1T&J(K}G0JT-01vU&ls#Dju%1S*_)Uj^=RPZI12 zm|I`nn9~SdAPFBz3N8KJBSg@+rFQm6mUoKMh{NC5cDKJ?yvs10W{p1 z@~BqboKPIGK(nPUUg9N&B+PhMW?xCJHlqytY+OTpk{FV-;gBJT0*Rwyz3j*bL+nCF z?(EZ>>o)`OKFqd&2f;2!={`5nf}je-00H@xkFhTm)8FPI`- zc~{PxhO2cBUUA-@tbyI1Cpzyo9MXBaeX^3vaMozkoOc5*qb?)o6Pm0@%_lU^$aUW7 z`2<=2nzboLdD8O z6H>wxctVwU$&<>ElutMW`2-aGiBXC4jDl1Niful#Kq(-Zm=K6H7;aRHD;|h702Dr7 zR$=pmEjbWt0M+WcX;)kj( z90CTg0m+#UVLxQJ5p1&ofR)QJwid)QLFomaFz-)JuQ$!$eB@cpB%OW8FtjTt_~a%W z)Z#x7-gUw#mEy;n-vnBF8xf#wF9$RD+@KHtDubI4>UK151EV2^YvL9oMX)V%#o& zpQR1`F`K0<_3EUul=%9I=@>ecr*3PjwpiM&Elj}5VM`6m;M)3!c-xea(9|grAf=oa zXmo9s_y)~Y#Z%i(XgP#yawY$C^3p(#}>rj6T#w~|Of?oBc9^b3>2-285l} zn)9y@)A1@gT?leJy-CI8@Pr(%jW@<;=Yyr{|EJ{G#8Fp74+W2fx%Vq1`A}Xobl3Lz zK_C8A<`doztiX%B*FPrTC3PA9#?GhG5O6H+4u{ zWz(7$ghi!SlEau+c=O|Lvs4Mg<42hMWB%3aNQGzaJ>+nSA+~;$1$9VTsMD{5wJuHq(P7liy>c&8&Fk+>Vi2VPmN3_8t z-&Ie^Nf3F)?J8e<67mg5<&muDujZ7{i{&*j59dyDV({#cc{m0`q+&`>51WH?BY9ua zb*9c+)wL7b(`=oWmUq;yN7#**m|so0mg|pV#WQ0a!173F`>6FX^^m>*_L{~e^_8{y z0-OA;e?GlI3Uo|qXi@!b`79%1!#M=E^U3zsO0_a@<*wo7@$}U8baO>cSz;ZprA{mz zg#{_t%ByHf6X#RvCHj>5M)Gpq06gQC+5l+Ze@05Ll};lG@@~wg>iH#ue9pfZB8YF) zD#(tuDspI3@fo+2-<>WQ@C8Xe-z|KJeA2&vF99*0k2auYo{6G;0|ea97J(>v42eHb zkug*=pUmHZU$?p+-hnR;dV7kay+@WpFP=rnnnR5QCcRhT12Wey$1$F1_2HS;mwV&^ z&+RD=PGfLpt9<%ih{b)ns)hP`9}~`m>gZu@+vh>@Ek_Rw_U-uw|G%R3M=q>=@yTDFop9M#|a77eL5dIhOXMg0 zZ!4pbqZXXvXppE;_XZ-MyLc%~>Zs?&<-Azvgv@^W{!X&r@3}wj-%c0Db;FevopRn* z^f-{2i3hTIsU-=k=oOk@rg2?!-U9c__7_-mS8r(i&xW|@jfx>q{C%5D-Z8$hC~?8fiX$L_hhl>o*q-E_SH8uUK?sXzRhX7tyxDvXpE}?W zE3eoA*}c0_gDBo$rBQ379dtw#2c52!K!pWGIvTKQ!5TJ!X(-c)WoQiw@oiuf^e1me z3V0h-1m)!y?`Os)5Sj~laIBrK`~rEuzowh~(D@}N9AF>KC|{n#2#g#2G-@znX#5%f z=Ga)HBMiMPdtlQVR5@UCY%Jj1A@jmVORv2FCi!DSp7c~X?Y|h75grvR3B9|s<5FwX>kGjI}r1#5=t4a=cgpG zQhkDsMsw9^s!x0xG1aG@DX-*dWAlFpDf=LxI>`fcEirSbJR0lt&kr{@hp-QLIFc&w zw{)LHN?!_`L|BcIZV<-;sY;aWr#R6l;jX-}0=7>8+2NsdMwRyT?4jvgPoc)-s|e;bwxYBQgK=5NXoh!65555SL`NZI#B#Yo4$|R@(I;xyEzMY0qo5aF)CUDP&cllp$;S>szc)RX}DmE1ia_ z;?szss(PkW;v5A*9Dv*O2T|7&he%B{k zqsiIvET>x4KE0MKV3=ddYr*5${_e4j{6q&_+1wu0`_4~pkIF8X-VtwV>$8jWcZmae zZI8SFTVv`f3%48gn@h2KLI`hQ>Mc(PR`-eS-l9UNq3YOjW_1QtAP?K?`@By$Rmry_nbD`r!biqTyvU=M%ksrbz{W30-1@1 z!(8>0mX%%8BJ!#pxfk0kJZG}8v2tm+aZ$uqyN1uQ9!CZ0!Uy{~{Tk+og%YBEw%(N| zDm&!f6$q%7^+6mCA4~pVk-P+3dXHb;8g5R`uFWU&vkwjqD1R)wF_-v39>kFNux4le z>{{Zqj8`D8MJ|Gpb$8_2$`svx7Y0naRd@ut@)1q6;P`XJ^Vva1v-YtDeTYsa} zPx|LxMc)sV!BUnhI*fFvGaCC>$({nXRxd6r96R1jXcPo8|LFXfPcOnM_)9ePKir@e zE<=0FR+PyRdKB3jJLTu>FBRne{IF*2p^HOE{2rg2J2#$=yDb5HpddfcEyS|W?{KY- z32b^p!mIjZt23C5^)y=xK*!83J7-PagLT*+!Df37!ZF1tS~Gp0>X8?BCW+8DR}BK`n#TAJnA6s4F@@VCzuP4#;muJs$UJ#CEC+Wp}58cnCoZc!T!OU5YC7F;s6! zJLXuq%Pv$OPgy)Yi{tdAy}5Iw^Tt_WdgPUvVeRd+mQNkvIkWx7Ft zi2qZU7K63w?2eG{bmpCnaiY5eJ+v+Gt69UA_yx&gu4@TzcB6MFh3rjccj{XA4MaW* z5o$q~T6d@5VjxV$W1~=J3Y3<;$((l<Z6i~X$6#r#QqhBd9739zRxV~HIDnm&78>I{1IAJ8L+MrgQ zWV^#!yW>&%)d=KQGbSqOPFC+AyUOMLo&B64jb;MWonu{il$-%cW?-v+zuk7maou9S zv53*jBlS^f4QqFbp`72R_G1?)93S-qBh8k;AGq6h`HgEWHL^eGk-Q-_U3gc3e zw2rW=&Z9TWOnve;lxHs*{<0Gcp;b$suIMR7(~>+|FiyzEF3GFIGt;MH2z?6@NkivX z2bKfX&uJZBS3F)KpGLkMY88Nt2u$wfphdEJ4lr*{>P{EA(nR-peJty#zaGIfdc07@ zkdv?zb@W$99Nk~!&B)nRrLQZXf6JOygoS~qJkWZ&P(u{f&xL9YqkFkN+SfK*J>zga zxDjoBj~sT`VXqHYmoE4ppe>@dI~Ixe`^bw1FUU;1tHxCll3G8uQj2`lznV$Qi;kw- zJFs7Ag>^Ml1rE`NbAwFF!|U!->(4k`9uCW?^Hc8_&gKymBOfcsbLB_lzNKQoTQm$f z9Ev_yIv9WN2}}?2ZF6%nfubKeiZgB^{cAIcD`KMF_~p9c$* zWG*1i{d5QEemeKCKe`zwUuw0RLO$(iPf(WRNp*^FGiv)p$VS5;MdB{Uyr)q=0#@NX|Qk;^8Km2nH{!udTj`GAJc`PQtGC^0WdS z06R`ugjK)~`{AEs^pYB$`(a?rq3|}_*`AHpk5v4-bL-N0I=^Q;-Mly-(s5^l8??VF zZw-<6#?dP~a2~f}eJ6{5KE{-pQ4k*<>^;NL^KLzMvxi$Ubm~&sKSQabzFc*2oH85Y zYRdIv1Qv`%>T7NBd0;l2(~H)D)gjCRss5+uAC53k9uAZ%aguU)`DRVlpG2N;biia? zU{2AelpBtIaQ$KSYN7taW%z>_lUGo4AAS>HsF;gPh7w;%$= zyYX}*SbX69^X+lwUL_DSTfBSEUa?zTwtE8V1s8dqM8V1`e24FJShAa16ju zKyX^BpC}@_gAKwpn@=aB@o0Uv0n;SujJdNHIH*c5ke#drlo(8d@fOItV`?8^ut`GR z#6S9^)hCE)eB|H6R^?C<7Bt!juPg+Dd6*N?Oq3E^ku=$$C4AY7L6G|b4$z!%xU zcO&v>FE4b~@+7&`M+kVh1hsR&;Z*rxFY~q|D+(yr$~+rrlq4u6gkJu_gpRzmN1mOd zdbKElvlBbF@`7LXkQu{OhEq!9p_Rw`v^*Cn2TD$$O1OzScNR)KL3K+=zJ)`yjU+?Iso_*xrufCHW;ms07Z@x?r)F!LLSalqe}frhKKUXcuI0knCx0JZ zcWan3VNfe`nn4z6n@eA>h(ZkBEEQ}@Ip%Psj;>&6n5Rbf+CkJAg?F+m6ri5XF2NCH z4MJcP>!?a^S%))`x5sqIm;u#d{K&hJMm^dxl_7i57O7oO5wzp644o#E3kk>qk58T& z5>%~ipPO$EAI_nG)IOme;mDJpjV^YXI7o5EU?d}7ME9XtMvd+O?l?h8Vts(|hUgeb zqm8x-k!4W?B{es*Thbh~r z#rw|dXxi3!i{clAfq~J}rp8)lS1x%|L7o|hpx82@c+0@5WM~r&-4Y)u8!B8f6<|+R z&{6O-Q9=I+PgeYNFIFT3WX4F)mKexlOPNVLE>vl6r+%hKj;0C~_aoT`ri@fvH@rKt5}bz`%M%clWC&y>D_>Z`Us$k6RPe&6(r~%^HmwD8 zK(0S}9cuw)iCWhF8FU3Cuj$3QTLB)wN;C>rS>(>PD{ciTs-|E@8{w^$dj)NWjf|Ws zk6ibBQnR#D{VZNSptxc0(@Z8e6)c@T5BZ_!#DI-J;6wT2kS8hGBOz6>i6Z`s>g*Er z1;~)KKRaH6FHs>`Qh(_IrFS?Xs-dYKCL5hbH9r&`)u?dEsOCn^zATZ&qdi!{IYRx^S-YMT4DD~i~bISe-XpyTypS(Nz_S6Dj z{IS_L6>Bcod(Da*wc(GnfC35&O*z7aGp$eUKU~NS4O2Nia&-Ybh|iI4kPk)l$o*XO zHBb_@MnU{Vlh+95AhyOM;NWeS8f|m={l;jQ$32~XXT+)-LxDe%)(TAZA7mgJnKh9_ zXDEt4mXXS?096I1sxNlj^y}G6Je(ov1^!y1mCHRAqtXlfHB=g_k_-H8#Z4{nH_0V+ zE0S;Ii_yiihc1q%mjjDu2+AQ*ODUZJdf@|$n8&^_4n;08C8-KjN2S@&2@nnxvSCSu5Hs?aVc_wRdf>5PY-}X8Y&DD6-b0&G##T1W-1B$A z33@H>?l|K`0=ciJ($LP(e%I10C2IP2ywmK))hE@1Gy31U|6tul5i>HkPyUyJd}kO4 zT8N}c(2+6*8%cIGs0k_CE|Cuv)Tychap;!cT;)#n zv(;B`VNn0i1vwlGHO+4V@MR?2-(-V(^WugieHhrU3??>3PJO6^DXbSmsVBhz>eX|5 zJVm)>YdMeni{Waj%LCB0N`)9x!nP7!pb$%h%0K~lp}1fu^Hm3^Ix*)U#5`SW+ba?}FSKVM0XkBU5KFSXXtc7#(2Ym9&%X zdaVU{^4o^$tfZ|;x~P+oLQ^^R7(i>vluv0}O3TLt3&P}+F4)kDWic3WRZGF?iEW&i zD#l{)!_ho}Cicp4D?#>?d60eG>V7eut_Mv|fo;|WO_fH{QiVi@E|@WtsrSguVF8uV zHrN6it;aF&Ba8`p(K+U~L6Nf_QNPO2nG^bb9zvujfYvP?_)#NYoNo@C-3LPB=xg#+ z3D%+ttqCU(IT~SP$&x@vjU$}&p@WZ-KPp>NQiMP&!v@mHCL{q zK0oNgztVY0$trmnbQ2)ZMW)rt0s%I}AgLlu(N5JPPh=8ChImgzC~!nACWe-trE>ZE2TK|KlSuQtGCv5#{|@5!UQsNS{*)~&Qb@SC-#?~AkQA$ zVccKpKfehdJ)7<-?FA-WpzSTe`pH}>56tYXr~T!~ppDLgOW=}(Nrq+lNtF0>1r=s{ zD$^E7GxCWp-W_V;@F?kC#KikZpx`J$Xd%@Ep74cm1oWb$nRD|VZrh^0C z!?1VtS~ilh=&R)!3)#v4)+66dV={BgpUgCxKMuSe^lu-H%4gcIknP{p=W2#|sP;KE zzk#rhI6=&u>@RmfcJjYxQEzj5G}!Z;jqUB}$?Z{@$A)Yg);)Ng(8NVJ z(sWfy+Nb6?Am++yUzy)ls}fwNJ@~CS;EdtHpT;3@3di903WVsoFGAYGBY4t)T!e=1NFf9WM^YMYpzre zEeyXRE5CEc_SQ_?gPzE5X8Tvko&v_PE-ow_JKjuu6t?F~^r>fts@2aFX&=vUgX(=I zk8|t>BZ-vJpe84}rBIuw;S#8|hb|7M;l^-UM{ecn z{V|+TGsye}vdOqC91AlpFhpJ(v5|lj)c&ic9Xtx<+zL5m#B<#+Q}e?(#DET0+GhPg zX?{Ie{s(^wdTqYFbD~amqsT}kB_I~S&|q-Cyb%UT%Je}l3DOD-iuFPt!4VCpQ&k=K zXw)q`Gwn*^%(|{Ku*nW4jA9saE%L??D_U5_;0H@iNWoI8iD4G#*KqWXl|!)w?cNwq z6%#{14cgVeB`NBAC_>(6Si+QGA8IKB(*;rUjFv5NhZP>1z zKr8WxGU%l+zp`1xLl9o_JP4OJ6y!g~NayBwxYcnL{x<_VpgBbvAT=uR%0fl&(nZEI9TZx{$zem0+jl!bx{tSQELm@F#-GlnF%!K5E)JLb1J-u6y4D^-$ zSVjN0UaBWiD(F!OJ0rhy6gckc-2Qx3HOJ90}c&eE<#i_5FqSA=@6O_o6|w-)3vWL29cd({n|aV-~;kBXhosY!E;ujC7il4yHjF!AH)S zw$BZ-e1J^gx8%b}18*Eg(#IR9C^`0rv32sH^Cds)oY&p}q~=VsXcR$(NX@Y`f+1I9 zD9o)=9s*nG7^lwBxYbY}F|S7B=BL?k(s#6W_6WRCbj;3fL7+*z19%fVO2>cS`V?~x z0g{tLSCb#I=f--=W|}WtUzvP`<}`N}K-U@qjjmA!D1@p_RBaZ;p{CWEzc6V%5-hDo z$~mPqjH)#R-BY!Rs!LW|TN2gB@8kmcZ|%-=ZDYb;DW4v0UEtx8p8o79@@MT}{qJ1< zpz$#!IEz;bqah+&F3;Nd_-sBrdtpVaX3>!i%=P{sPVd+L;9&6{_-q-@440#5AHKqd z6IY|(T;vpq3J6x>lO(c1rU>Kcia04k=`7*&- z2m5J;XqSUQ{5F2G(IADWSg}!SVF@rfKPiMdsyZ^6v-V#xgJM#+;YJTe0=AIBxmd`RkQ$UM=_R z9V}8y`>ogi$Z0Tm^#9-$1pU8eEU?!ZQNcrehsjfcrz`CSi}c_6wmAXeiuya1?YoZb zIWxUTMH!ym-n>9%;rMKR#`6T1$;{gHEdB5H!h?aO^|Kqp*=!I@EwUYs+*+g6>+zh( zWOWZ4jM97LV3VYMYk2iAdAdE8Qz`T8hj)fs@HS=>_7P4t?SY-bO#7FiXUDNm!uII_ zkzlxNjQabOR%IfhT-99H(t;TaN|oL%eFEzR9m!%9ve|q(8I4ElGa5FpF#o_QYiS8m z`D*ul1%cH;jufpI4jJ8;A!+fsOGpl*(d3hT^0ZKaS9mknzgDq|8T)Y2W7a1}99%Ev z19+JC>xaQ`Tt9noI@!9gP6LYd&Ef0<4e^&BPY#SXHYPi>=3hcH28+~{|7>gg($?>S zbscYw_^-|T-AP9vFw@IJQgV?T3FTrV2;fqrheIt{wv?*~3`uolgj8-+-)%%h z3D&MmlhrbXIMU9E2+6L}YIOAf6l5io9d)!LyW^v?POf0l!~%|3J0a}|l?J$&m#wE=%_9ge*k``pgpnsxfC=Us>I+J;zisrvtA ztistMKndikh*zIMo<6wMQ&&|rt(I&G^b>TTDZy*7BN^p~`gQ^JZRBZ#BOa(t!a&|@ z2Kul7IxJxUw?0YC|0uvZL{bcQ;eNLMst9+AFn8%Gtif@|xhvo#e#e*&eRQxsZ}%v{9*bx(%^EHY4^lf!K4~ zzGds|_Gmm^*`RBD*Kb)F&gbK;`DA-*Wj5Y8xAM@%@$~YFNrENc?h^<;kK8>tX+$vj zD>g*`Uo)al7l_`Kg6QgGIJ+I1FBrA!Kb!!*N@UjsvfoeE2G2`I*7_wI!e2Hce49X6 zds*tH`Gv8CuP+i9-c3#pzApho>(Lh61a--unUVben0xal+mGu$?A>|5!($Vqq^s@u z%8s1Kl314@XYTjz4FtsRZ;1_-`>98?sNK_{`Q@DKZqc&2Hkzu zKId~zpYOsF`9Ypk)k#23>%hJ@{60(ZOX55GAI_v0+y|gq0Jgfc#4>GlJ$P^9wufDE^S?QX1zogj+km&v0PyO@No$hYsDy?*Cxdw+F&x{6D; zlm4cc2Z(f5`$qt^E|__NY_^T#fD%p}X}D)i`s3oG{g-$qO*(^+Uz~}`*hNM!bE%G9 z{y|J*wfj{|d9y3Num8QpcoUvQ z{05*0F6K(z%m>G3tCOSE{((gPF*^t3;Hb{Hx`}%W$fLP{E9KE5JAw9Xbv(nharOZ? zDrB7?YTGW_3KE1?wAJt-c>GLB5ZJ3jl*yUkQOaD1^;E)U8Fz;#D0JNISmE+E;uxC)P(fxXf;%a$Q`CxEZeU0S)A3cpK(b7ZHYL-t)t-W$40 zD;GE69xA$1TB_{3l(!EgQ^;Oje8jX@Bm=eHxVv2$+YLX4OOu`FQdWm31M=Lli?Q%L zx3xDYL12HEA+Ymb;?7VMId*YXN>jD$c-}ZeQ#tzp{Dg9$F+H|IPJwyejM$2|6GSl( zH;-Lxg~!bmBhR(>3_EV->;h2=#LZ(DSK)Cpa0bw3*&^j`01#@O0+i>DT}*}FrNM!~ zJl9H=eV5Yr0LX`#{tJvDK6wTOp^2 zcwTFn88H-ZCx~JoZl1c>3Xhw|7n0|4c7Z5`Wb5K8JZ=Wg0NN~Dq}&YvLhZE;-KA3( zQ{i`MaG(w3x%52%a-q9)>f$EcLq$&jN|o)Hc*_GsI+;TD>f$4&y&@SXJG~8PMHWv| z3fZfRk(l<1XPNvo7N8bXog~H3UF+f`x+d$}0m*ai-P1{>)B6BqLwWAZ#ZGvh>$AUc z3Qz!+Vs<{u-wUD|sOQdHJcZ}Ek&|pV&&}Nt;s8+3ow*nb&vS|A!J;j8KAXQM#08+9 zJ9BYY>TZr?W`*Z9h&RmRz=ee!D(-LIoxzrsX3L#B(aaBrFk$kLRo@?nOd3}glAhw#(V=IJO%=2a> zMZBFLih;O!=3*;6Zk{mmTz1^dSsWnHWe9+5=Ps_o<7VJ6pv|&H%H1GIF>sg8T}*}F zrNM!~JU4X|e%~^AZz#{5ySNGWP|=-0soGeXeV6j~0XPc1Gnl*hh-t4KQqf^|gNAp9 zdTZt`Mq=7)={z?OXkr6Eq|QSDckSH8NpwwaWPW)3+ewqt`v7D^dG6fBPI#W{v%hhQ zwC%Ff4V)j*y*ckvXS=SEJ_CE~HlH)Pj&Gfx85bLTF`!t-3>zF4%y&U5qkg!BNq zn_b+Mx|<`JS=7#f6FAAU>%0?c{tMmB3m0>xJX(ZG(7uzDW_UWr{?Pl}g^RcFJa@9u zJeRUMz(E1zxeFI#;dyRrZ%~5l^p?6XK$tZGP*Yu8mC{rsA1j#WT7tWHeJHkC&|@ov z+KA^(v7cXLK=}}!0E(LzF1Ets<|!l3wK)PiKIZHK@D!A-i>vUs88`!IvzA(R!pq$W zq8PYK7cQp4@6zBL7ntXg_Xa(bi<@u{72OGxD*LMBEf0v&`C|Z=TDbU#X|HgmnVufd z-HzeiA$wiA7>Q}G9n(<0FAhYHp8(S2rHhm3n%v0zqVccwPn}dcy$_($P@cPVu@j!> z`rK`t0<>LAF+0!Y?*&l}Hq0C)4!#a*epIg)|lcFxXcr__8Fx+y@9=3=gtM~hIIw(ks&#@PpwDNxT{ zx_ArEbEg~4b1ADslmU6}(#2SKp4-|RlpwIbvs`u}OxzjNR2NsJG!@DGf;=}vK{?B( z*r-03{SDC>pY-4gIS=M~Gm;_RP5|xVw;_^UbcILGGe)M%4vaaA0~A^#fu-w0D?D@t zP6L`PTd3R(k|fLT7{nFY#Z~xy8k`8sb!{ZYzEA0U0Q8FAgUEEj6Yi#>M*!t&NoD&h z-aZhypv}4{iD|QNzG=;eU%d^8MRtKm1wFD0m6$e*C!3Dk?#7$g2_hL(W*0Bfm04dB zNdGc$49AF0N}k+b{4r#^E{MVtUSG``+okQ7o$&Ja0@y3w+5$YYi>&aBH*yqG#tR(C z$(x-CXP!j-O@IX>7j5AwFL7fm?qZjA^Y;W8uf8=Dy`x7i0!!V|ku0rn&O7Uz^FnTI zd!$^b&(gI_09-cFR;0@W%jfx!mi?1A<|s5g{Qr(P0})C zr?=D{00x2q<;X=?DP={{vmo#75>RsXfyjlTtPwrRLXLs?-;5}Ww-ZD%6hn_(l!eF8 zb4K24<;zZhIlDllg3@&n79K+bhe+EkTe{qxAd5IW-AvhF>N-HjaqWM+iYSd zh-4^0^ya41m04dBNZxA+?xf7=eIRn7yw@9`j_19;GB=I^+AllZt9nJy{s%{&R1_f{_2!t-9@##r3N&U^Fs1XKddd%d~qsXIE7s1?q8=Nrp=y@Bf~ zw-zB2wDAmY##tT^rn7G-@AW3F<9Y9#koQ{O&CYKrs{=HGd9OD@9nX7Pn}ag6JAWz&>if}M^7m$lAaaJdo7_|d_Pzb^aiD;qAY}5%>QOYS-hPf2}60WH!U5{dl!tn z7dV9V7W|qLV|Q4(4yVGU3v8CQTeft$I{}OZ?^JJ|I)0}H2f4t!m%KNa_j-fWaW@s+ z3Y06rZKuZUOP94jnD>&Ws0Z628L7Q|*jN>37l5U(%~G){;1Kz;?QY9>J`5|f!>Z`Y z+(-oRSh$nYruTs)4CTGvKy^It^(EXm1}I~z((DA7zZXO{RPXgBtK)fZ~ReRV&b~c>~x} zZY@G4XyX~)jI%r-Oh+!1_j;4p@w|7j(Y%+kIzS_s_j)7M@w~USIVeMRx=URcpv*}R z%6q+e=_zGJ(zAkjucf$)?+2r-2|da}$i@6`Ml{9S2@o|HLwnQH@w|7*$a`(T!H%Ih zyFjEuwIgqwI-d6ehXC!CEmZCX0Hw};p*z)^ua4iT!HF&~?tF1g-(G%o26n^0I6x&b~j!;zXcz`L3geWSyMlSIH)0*nds~};GE9=kPK=2= zg8`*CKRu?91W8&rhi}?hDTR}vB2_o7>SkEwt29sY) z?)O-7KO$b}KbJ!;xCun)??Z&TIO;FvDAH1e4_t7 zdDM}10)%cEEq)kcf0Q{C>-iPl-?VKT(b!kxc=P2_{4i z+kDQQP2B_{6m)PdB1<~B$QeLe@8sY(8$ksB0_5OaH1>E(#TyGsE3A@*X2?+zc zfyTubEO&<#OTjyK;rgr+cWiJZFlDt0&A(&mn*bDp4$B+Qo^V*vGk~%M&f%1q@32@0 z05L6ZIy={Pk%ZOCwt$Ty+W=gJZP!ImT-!|~tCrvbMx59RA{x4by&3HcmDV=~@))?A z#-`T?=(HvypuBkDBCAAR?3+;IK%gC4IrH;h{$>CNz`WR-+@8pbBWEGSU6y=)f}Fc0 z!~`fW_J+DA@?zqNuvm5t2K|HVDzI`am#YlK% zUG0+=%~^_he;Pw`(yKv7{|V^mE{2+5M^78wGxqrJ6A$;7$>W17LL~gR5D5-@`W8pR z;qBAYZTxPV*%Ln`hW&T7P7E&q9q=E;4_NKEnVsL>Iv~gds}(oxA$P=k`wztq2`&O1 z^hdu{e2qWK{q$__mDl?BeMJ7}pzr@bo~xddY#e-D_}lI%`fC;absYV568&`={dE@o zbsqh75&d;J`1%+XJMo`K@t>>s&*S*dllafm_|LQW&-3`ti}=rRDyuk^Rh-HyPGuFR zvWin##i^{~R90~+t2mWaoXT;W%5j{^ah%F=oXT;W%5j{^ah%F=oXT;W%5j{^Nu0__ zoXSa@%1NBcNu0__oXSa@%1NBcNu0__oXTmO%4wX+X`ISwoXTmO%4wX+X`ISwoXTmO z%4wX+S)9sQoXT08%2}MsS)9sQoXT08%2}MsS)9sQoXUBe%6Xj1d7R35oXUBe%6Xj1 zd7R35oXUBe%6Xj1MV!h-oXSO<%0-;YMV!h-oXSO<%0-;YMV!h-oXTaK%4M9&Wt_@o zoXTaK%4M9&Wt_@ooXTaK%H?tp9WTDK1N=S$ey@Pv$H4Cs;P)x;`waMf4*b3Veh26s z0rZXldPe}gBY@r!K<@~kcLdNo0_Yt9^o{^}Wk3{_yNrn9zXSBjm?-}D0KGCQivK-8 zuZ)Z0e-F?rBctfw%g`wPJ3y}tj^ck0&@02E_}>Hc$^a?;_W->zM2i1CK(7pvqJJ-= zr1-_}>Hc%J?b%_W->zf{OmV458w` z1N6!uD*pEXy)ul7|2;sj45Z?J56~+^srcUm^vYl=`u8%LivJGKE90s7-vji@h${Z~ z0KGD%ivK-8uZ*gqe=ozT`0oI{GO&vOJwUGvt>S+V&?|$h_}>Hc%J3@w_i=h9B4(?* zLm}?ld-LG5zjfov@Ve-|(7$@)pY(>`B;L7OgSNY;pV@nTb=rTMc;M#!{j1miNl$!l z6TW)#_VMXi|NiwqhHn@5@1Na$ebwLExjGyKt6Ue?Zr<15y55sdukT+M;QQ}*;_cV3 z_SZk&>|GH*Ail~6ie9>N)W0hKV-O-{&7bP+0mw~J`SQX3o9k13m3Z50tJUkf`v*so z>~~$Uyb#|dzSiG+z2xr5koNWF#1B3puJ`&^hWA~$a^*3fVntP(Z}*9*8z*PYedO-xC2xtp^2*`C>$|t^?t05@ zpPk%U?H`t@qQo@`K`=kUjbjytvvseCptQfS+Ul(Viyqlj1MM4$#o(nAoWV z?u&1Zt*@E5JKDZmf{lNxhssK7u;nC7?jUv2G&g^!7nkSr2ge6T(vBa{|7Czl9ToQO z2r>bCIswR3D!o%$&r5+eJB$lnd{%s8Os3YPqby$U9;aI~B~92Z@t0#8KYRKC$=|b8 z|GN2aut5Os$uaL!aS0MU;ZOV^&F24pPYhyYytF!8y>iySR{wK=k~)AS5a+Nyy`>Uj zwfMze+#g8kNh4Z2>}^*;svo2t-CFJM9~{4Wy8X=F*=JTKAFFoNpKQOlT3?(4gjjc= zCv25zA!hwatI9;$X!uuytLL5K!J9%}*>C+-^&i)SEJ6-rtdIS+yCd~UmQBwN&h8E% zm6qmc3uPBW!p|cTs%i<5oa(G3}2N0*dW3{N_MD?+|al`8F90ZPnuW zF^+gf6!BC}JQqR0Pa?t@*TQ)hES#4P4_;Z-w)2s_eVSMvwqhBH?f!cbVrlMx8O(jP zME)0!L@J6z#>34X*$Q0z2_lS1EsRHDVNkjhQ4xKomCQ5Z>Hd=m$xyaKa+0^MIlBKT zM`mM+%qDj6;RRDA{tOY~v=(A8y1-pXF|Mks%)g&xBLV4#qD(R2?u<%Qe4Zo72}O`o zC&-H@=SQ~S*kV?T5+Cx&m~}xDDSN6M;{=#zV_Z;2OiF(s?i~krA_o5azK0Yj9H1wJa z6*POa7{pIk=yWdjq#vG%TPbBEhd+CWf8eG~ALeyk@0MlS7E$SbZQ-ZDZ9q!=-7#P&_M05UMiRLP7hNuK|0O1I zjU&-rxqjUMKQcf~9Kr4CNV$+tOrBEw5=-Ph0tH*S2SIi6=*iQky@~jHnA9~uh0i(m zy-MRm-7K9w>zI8G+}0)T&9HSJ=z;}5c4~Xst56u^3B)Y`?llS8dArU!-(0xb_7W&M|^-^h* z@GE>s4aubH!PS}@8+Ft-I2J5s_<8eXRq%l}} zC;gtEX3vy519xZ4l5f=5`cJcj>w|Exa2HOvmqeAEQwkqX1914%`<7h0SucC?J#lx; zZm)@d!4j>Hx*@!CqP-lVFz;Mxz%AvSbw1WPF5eim$~Yj?JCgvnY9|K|W*i#TTdQFx z+Ex+KHsn;fk#`0l#X1Gb9l=7|q(!*iEYPZ}ZfILU;xIV9(r6oA`}U_fovsc)z5T|W z)yduHd^!f3i1gnEr8Y!cDbDc6| z*1_%F-N~yZ&yY5EULCw5p=Xtco?SY9SRK65I2v9D_wBsF@hnTXl4lo++2_E`fmH{W z_CdMJy_UAiCJ(EFk2T7M*TEzFz977fT|dq|s+yNeuZu`)A?Uz(XLwP^NA=s@Is&VM zk2MB|*TISG%|+wbHQ?;?pd1sdI=HkUiZ*z-+`U33mcNfRLYKopU&&7O z+cE3l(p)G)_xj*6vU4JDld;VWG=Cp!Y%Ue?MNZT0-p`1L7-zsNoLL8#wm=cKhs#|i zX;>Y6tWi0<4&Iu}&bDX7*6K>Ny^{E!Nhh35u9$y^!<8ZCD+A zq7gJaQj4CZn`N^jHP#t$yJq#1OIx4_d&G~_%2DL38oReAr2e(wRB$=^qV8?@1O&@w z)xo7XP=sCao6;C;O%vthjosT5QvVuouq^AQb-8rE087@vrBzTwK1N-N6-a^GJwY5+ z2cKx9jPA{K%5dRza9`_^XGj|auMRE^gCg=tH?I(eOGr$nh*ow-GwK zP|Q9DYItf6D!m3O?SmqCYERqdX%|=>e4e8gPu~B`I=D0!iqJiD?ow$(^LNrPw<^|g(CyyO$loKncpulMaTNyF;kQ;o{ub@1xa-h&f%#w@um9rNno6A}_P^^iFBd+;(;VRi7S#^CTe zc?;?SXSgt=c-XGR z;|yV;t5WwVQm%4R@lq7oU>~i6Pl;$7a;n_OI|EQ-odTCP>pi%%2#UPDuC&V~4#U%@ z8g0Yt;NwfLgELNm-Oo97aA_12akE~GvMc2b`f&@_I2&FE2X@Q4uWfF{HUj1;N?JGx zW1D*YYnMwLmPb!XbJYS~UvM3qyg$LZS#|I!2}+x?UX8LdZ8G(Eq(+*n7Tv3xWwRqS z)){cS=0$2Gq-;w4(7|TuM#rL0N&Ra9pW!MrnsZqn-u>Q!Pe8D2RvmmwLdd2bLUzg2 zVRi5+secWScQ@ zE)9baozFb~r9@&Z>h;`=AKk!`3dFJgg2r z(TtPVcY7#v;)Z=D0O92Aq;P#v6j zG+rHCYCA==UWXdH4o*BbX1aJCe5MgMyf57zqg-5`;hmAn=mzWHQuirR_5iobBn?B` zW<<0N;d5@}odHO(PJtT$3vH7YL6P?o-O^r{{8uNat%dwY@$|Vy+weO0fm$IT(yALS#|JK9_{Ap@bhT$33++6Gzf~QSzkS1XWDeyuzvD6X|7syuWpvj4rf?r z!0nn<2bZ=$5w_RCcA2DM{cCel|60IjxC)KtT-JwozuDjul)ks6IZ%Y{A!L_K9eQt* z`quz?ceC#9SKEcv!KGDDL_X!m+)U1*KFfyw0;_}1Npslzb0urEG#l^a@{(sDSUIZ> zJ}04Pb5ve} zk1)O*-o~zjXPyRC2HsqBX)P2T@Q}BAgGj6nKGzr=UI!=gK173PcL2KvoP8c#8(4L4 zX+sok@Vemc6*95>eL+;DnvANIs_yF4dnR z+=@RJy$CVRKVfz7g~s6UI(X|qkmb)^m{-R^6)5i?JL8v(1+MOp<#gG(EtXoH8#-792b`TLSI z&uw}+b#U^w$@||uHRtuQOLL(JJ&n$}1MqwF$!tI?Elbiox9DD3_rA6EjELxR>fq8A zD8lw|xyvLCtAj5!Du>s>rxEdmD);AC~NF5%xk;be=S=r6L*|{g{47hd6vKdF-K5~Z0+k?(7lQ{ei zCv~(1tcI)3C}PU6?(VlBd`9&xVUE0g&#Im{oEfY8=o!XM%Xa>IaVC>-C-z^VLe~fyKA+7$PqIi|%F;Xe#^fo*ukdW$ht&|N zd$`(W&!0AX@!z0=*AN?CA_oPZW$zY?Vb_q^=fF+kmwB@HaW_Qv%i$%{Wj&7g2bkzb z8iK>C&yH1`-39jw`1F8%fp~DO1#B=3hb?gdq z_IYqleTk(I(vB#o-DrevYK?Yyaa_Rj`Vnd7+pI{f++GP8d2OV6n)i5|nb(hK*xks( z?k<%&G_N0##=b>`-7N;l2#+EHht7bR+zs_mjp6vEJ&^3&Ym&Q6;@(2ty~gbDiutT# z#T?1!>zcN6dhF^m;cVh(c=Dy;bR(|;ZWE`*B6{5k@I$C*HA06M&0Fk)><3NatkY)K zmx;S$)_YC-OP+LTsN86ZZt4_vQ#7jwrf_mGBF%HN?Cl(>TM=h?GNg>kE1NUjG!2d$ z&G_YV7GG(4gky0dA{K|Rni-4JQnL)ij&(%Uyw16qX;|E7o&&+L4o(z0in)(87Khi( z=a*hLXPf}L)3WO3(lAJ|n1{o$D+XohLoxS}M&aINj{|ao2_3;T5U(E#M|xlcptN-S0T~1ZgW{*3G3!kmSw! zYQiQtfXC_;srL;K`37yg`wfU!*O_;?v<#BeH`e>clz>3%<`rq4o4=(_CocRuu)$y$cUXH{DKb1&Fg8XUFD&6n>|0~Pi_dbZa&tS z9bPw|-@|ot!r7R0b7=w;dmeiYaqPM|;nbL)qSVdD8ll7M=Bhz)z+wqzHW ziAQ7B&85bZr0bz_n>smWx7Tp^bxfN7W;xh7Qm2VH!_6Vp$aYbk>87RrlLYOxvu*bL zxVX`J+L(yNA>7Q3#WUbG)+unV@apE$Do7&tI@&ImJQQ;uYb*}0n=dZCZq7IXcBf_4 z&81{m!=tnZl27k4 z$gLo8_#IB_eG5(lSFO=R4Czm=Z&!z?biW1RGcxPuV;X)oUZg$C1<{A!;iTR-K;+$Q zy!$1Gm&S1`XU6I@6m7gjH+2HvH0^lZd_o%k<_kSp(M2uYhHahPdh!fui(|ewAJee4 ziHEISI)!N6e4w6n}3%(HI;d1u|nP}a7q7gd0Zcb!B6RTqvinEV_>VdtB1+Q){ z?TDlh9%^@wP>k2jC!{%Yiz0RNROIi~QZq8&NG&y1duGX*KG6srUN>(s7_uK)62~qu6L-g~m;2saYCK809x8V^ zWZ`aF<8k<0ILqG7k@|+}=BxuU>*iAbNrGPR-HR{X#&px zt5`>5zG0cru(*lW(Y9#{Q0WU4bDwG~4zHUpFTHNgI01H7XVuN6VNi^G%6f~+t|>4$ zfWK9nY7`Eyn*-Zr-Q63qbLJ75b}lW0Bz5X!H?|Ytl84sKr=)pq!Il?XHz)5;uyJ19 zTp9&Q+8)rx#s*4KhX!s_(mc25akA{&re5qUlXV9Cn&Z{Yr9F`3?cr#bNgRHMlX~9* zR>L)EG-ykqXI|e(_d5FI4io$r8mmu9y>GyQy4iU5d+g%s=2IGq zHto{;#$+HEQXd9A0wkv*f)zc1Gjab>z%Lpv-YD zfcM^9S`0}I7PQ@m7sU*L*3G9Hp~LIuL^&|)<}`o6t~F;LmRUEKc0^JL54F2ET2Lj# z>*h1k9J%E^UpFV8kXbjECPR_?EREfH#ng<rm?tZUp?8|KUf`~oo?S+ot$m2j$gUGzdG4I+`GFvksoz0T{!=Q zxVHPs;oj+K|3&eU{tJh>J`XF@A88l*P!Ahk=S4!zIk8b{;a=Wu}c84XRFdoNNqdVGSN zqxbn-18Us+9Is%blDGMIfr}4625#%T&)iG%pxo#4QHlG!16lTV5q}T!?{kf;G5otu zEV_g<_H35C?xTFfZPQTfqS-ph)Q2$b01|(zcWl{J^yE{m_I2#(f^)2F;k9% z^XWHvj(~>P%{|2K9-;GpMD;;~cONgJ4@?}-YQ*Z@c_xV;080_pVN@N`7A5VFGiJ% z2hn_gu5rC&JQz7xRy<&3o)auF4uM%acXYor3Q{cKwborKb)Q51zDD#OPx<{uRF!Yr zFIx})LGc~^50E^3a%;@0brt$&c@CY1;?2?Na`n#E@`f8BUHmOnx*ETGnkC(Cmi%2- z!Y_#D`XA&=7}_4Q;cLwHi8&3kn|qkuCDDcY>xD+^lKxto5G7bf@8}#y#@HJbv9%;7 zK0Bv@bqf!yyG$|#iqtKrNFBoUFc?N7`gZ^Nd_;r{Qyp zaRx5WJUMFQ?hZrq?uAC@5_xxUGN_1+O5EmP{Jfid0D`6SCIK&KFx(h z$8j2cOT=+ec4t|+l{r7sVI2ZB24);5je_Ldy~epqr4GN(X?=46v*8jpYU3NNvoj9K ztg|m@=-Ptyg0;Yn4@)5tz0>;U0ETw6bbfTpJO*y-ygK`WhOsR?jO~)?L+k7djj}QP zyUvX+zRvDTJ$VSiz0Rw%FKEcyVlnLI+Y3o@5lyi_qizH(F<3 zYWyxyXKx)T%aZwJXyUGz={k&*S!b7OPZF?)*t&LOBXvtEQipIoH#*NjvRKDJ#m}s~Eon&Ia*l3l{cKi!jzzqX zfYdj=a(M9i?ybAKuO8mMwRgCCeEZ~RTZwD5)JWY^FBYANRZ zX{#F(x$?A5Kx=u1&@6GB3(eWFos5G!P!9vjY>cx zPv72>0Hl^^d}5_=(@J4Sytn^QObXBj!$Ba4>*CtY`;;Wa|HKi3(k_Art9C}%o{s+= zh!X#@C%&VefO2$ewZDIG{OalUXK$bE@2=ec-rL_h+CAES_W0oJVDIqY2js;f|L^wc z_F5r#_fB`uR>!BePx|*idHeYEZ1{F@|NhzC*H`_movR21Tf?Q5#F zE8_ljRkQzfPxPfFC@F0H2iO1hi&w5(xq9V___3aNl(tfYtTh#em=0(BdUM7HpAgr3 z{VT)!y}(f7n}lF{jKO`2E#F931bj-rP0Hr#d#ijmE{k_}Jdb!QQRI)yJwXdiho2ZLi!u zKHEDuUY+!}q|CoT+`l%Hev4qwNO}|oeT{g#T<2u9zk9rQBu%S-MM&iSJv8dJA3r@k zc=fn{t^VhL$dp003Ko&)*NJyZ8$a0JJ6qLvUw`XeR9@XVUF}ML@|tvL>faLoS5Lf~ z$g7jp-oEFbiM%R8aoSMm4*NXf*#L(P-)tRx^m)QZ`+o2$-C@VbD&w%vhZo|oBLuJ< zcI4l0U&KU*9b2is!EL;h?Tycefj1_2Il99>Z#eADIobTX;C$G=3+N8}JkMdD6CHL$ z@I!gOw0vd(k9I!H_2G1feZHOLu+NK-M{wBBi3gsSP|bMzO%qavZsyaU{`C8Oe+b8g z$Kw9=YW=J79|N=UfOEGI?9lQ?=+jT!>4Atf;ynM^=Rg0x236t}@t4%;C)=lICkOj7 zmN`8$t_;31bKx1*eA_Rc=DI{8?&tNvvB$-UFF0W5?8#yaf(SqwcaQc?U+Zt(P=Rk7xL9X4MzyXgPCu){^>v!~4dSa$KXY<${Mzo`n+K<6m7cmq zZmnKDI5xjr8GfyJ$Ldhxp{MT8);}JvKHKVJ6CK?rBkK4a>9M(FWA z`(r=I^LV;sxu#IAO!Rp5jB`kwIZ`43OUGc{SV3rvw!AIhxHI@Sx~ZD%KU zR{IC%(#1XS;%e{kse|*HzES;;F>5_Z=8=pi=a?dhFMP`*G(~t-+`nE;$JTn~!Cz}f zRBe1*%dp#x-}ywEJm&*N_pp&y82(bbBjiC zeORx^N{9?(%}vx0q%2)8EgbpH2@VhH!-YBO700T|))L+&zODb)D7A$AA8dQy#RB1D=A`3eqWfR%~NRa&#FSPaR?DlJ`<8_&!OP%Ze zR^_H*+#f|%4%-O;NVim#NEWetr-h}WTjHPg#8VwfY(S`;2@B(Wdgbfqx5CmL#>Eik z)^KWJsSk><7%8i7wJKqqr4S&nD$IO{qbLoKwNbHx+Y&SxVZCbDnV>eHD!rdBuGJk8 z_15CYx5Su3WgBc|u_(madCD64%bxR>ojDeVUsJ*pa{Snc#laQ?%i>^ny-3kx8LFD; zIBp>Vl(%awCcf?{DyUkv@yj4aV#&*(TUCvD8AOQ5xC|nzw586v3}WlG>TI30C-C*j zi-vm{G%IdNlPmB4!-87e|NC)Mk9!#el{bznJ7--65w+OlWdMcHDM@lKgP1(d5A`kHquhx~dJykA)${#h?ram1e|| zxK=M$BOrEdMy15dYLklfbPa<@WM9D`>w3d-pif|FOmf3AVT3V(UT-+lOwj$OH5268 zEJ^CRLy85KO7R(k|4IzTfTQ&K6o>0ZXDW(*LZIn&7x8uTzDeN?v*h>Wr~~#!u%CKMhk_1&ucD8qlsZek$MStjVg6HdrI5 zatGG2ELd2z@KcF1OYCoV&R9)qg2FgQXlnu;K9%01UQT-tv-V&}OC{DYD&1lvY_ABG zz^7uWdll~SRCuU;-UVKj$;AnV~SN_lygPB;xfo^wKKL9abky~+5)_{nH{5Q z2wm+=-z1AZjj+zjaos9Z=U_{XM|iHSeXrj>l`jTvA9ul6-)D7mEJpq9sBXgh00`?9 zl~uD4R@pxML?`kZDRoDp!mEgmuKtTH@kkd!bUIyB)`J49#afWh(A)g>x{&ug=lS~pE@iN;)QJfyZ%`w&sh19xo@&%c zclFk&k!tJZUX+dnNs(Ki(CZt7Jagz;0~A#pYZ`RpU`>PS9HirRUJ{-@zDAKGz`RD0 zo>$P#RD)yaMCn%iYx#h-&CvlJ`&&(C(;gXbl;n#Una)dZ-adT|KgE58g#d z)|(BeX;Dkeeoe9NQ?}UYSVNIZ4kwSx#md=f2#4q^d%q%(dI6?!b-UOabF(w3? zCIu7Ws(KLYaN&6dap&5)cOt7h{p>_AR9R0FEPkRV=55vR25GO(_O_wnD<)w^hd)t} z{pfyjY>R${`87{`m7HC2e+{9915Cs}>xrQ35M43_d*z=sNNK*>AYKpL)CX^*T^P&6 zf43F3NJi;tLtHd2407cis;}24n9sX7%!P93M-!5C$= zB}=t6tiyc$Zg1UAQ|We^%F8lIN&8FBPy;=~k4QKk7xlekr@_jgQCJV1=(A8P==H1g~D5s6mft zOkmu=1iWbg=n*7q*7`Rs0|&@NvtG_`qZ(^CK<+`L>4WYK>ck-d+ER3~FeR>~8;2?l zW3F>wn@v;Mv|DGsQcx=M4YJUd15+`TPI{@Qly`^Ap%ylW`uFh6qfT`v;T7nmx+B}2 z@$KSU``;4n&S;(=$Z}CVYUJzg)hOv^JJJ%xQ}*022ugLyUy#$%G%GUeFRDEesc*)_veH_z{RMd~ zkxK>akC@qdRpl@Npl7!7Q*i4L-LEQFs;bQKQ>e$VEHQ=P3M~ZlQ(M0ftk!0vw$zx} zq_+M`k!q&18{)S+Ecs|*m!8@h$@zudN)>i#l55w%(HPR@wz8L;o!e5mY;s$#$VqRQ z2y$LIMyOV8Q{-Ihxq2gKa$D-TjP%Wen)N~pE&U}?`Au#+Y2>!v7n)H4nA~>Za$7oU zf|osY7H3v>y)k$Ds|_UpPfxAh)bplte7 zo!iR%I67CTePg-gcDzG8=8t>H?Z^$S2}TOFwyLw+X?PWS*)3(HP;1$Yv`qY@Pl0>P zL%ZNiemg@c$qwPN65Iws?kx|EAd}(FyToVFU>{c~nL;(^A3H@5f7%n@U4tdZuiW03 zpT9WVySqBs-PztdlV842Ph7T7SBIb8e&f#S2$=X+J~R z7PI@x;oj+K|1t4s{}B&#VV@3qA0yw!tUse6$7OuD9N~?gXpdZY9$iUN!=OyPnh`&I zRXkgaMBOdDIb|7?7KiosMb(PcRINV(_UE=2g0034OeVxaijo%KHd3z|&ueOOnh%4K zP7qD~@7bwWZ0g(ijqWgDuQMoT2f#I$WJ7p~dEie$XCVG<&pd-;v#C99K9}8_vAiUU zJ6vWFC;__6%E=0R)1&Vw-_X!|Q_X(L^qw)o8<$y0a^v@uPbPR@cDT$UBzC*ZA`7;a zCHFFmEnPnfDk+DIJItE2$iK`GfHP^^%6+!#00r;92<#cZ`$Ya_7J#9d-t|29GK=We z1zk!Iewvhehs!KLRdd;0RLznUJ3G?|)wobw(ObP5q8j^sMQU;LNG&f~u;<14;B_Eq zC@R!$q#r4b{dS`xKD8xKnD;3DE2XjSIU0LCM@bnQbaw1D9+7g!@cvW~uCSu@ zJCD@jxuR|3wP}jnn4ijCmUC z)}*nA0jF0oZ-=;C;)81_C6Vb)XVuzO4Az1ReNx3>BFv5 zY1T$_^`oU~#92^k#gS|-8+(s9GfhgpY!Y|tmp^0kh|{b+n6d#C1^ikkk|s&$7fer(1kWKK{%eE=`B{l8+y|(Ly_1 zA=HpK4kXLUi^d+^2jRK<O1?_J;87k z-u?~9Gr%M}yv2y+kj1zre#MTTqD0ajy={;)k86Wc`oydyc_?NvUdG+1kQ#L}&uCf= zZh0Jb7gsq#(F${}lZYhxrJV%5{|!|qCDJTmG@+zY@`%7dW!+7-lL)}0Jg((AiHJT8 z%dtFuy<&zD-mK0EWjl!gFUsR^c_%^Z(Hom_&#NZlUM_OhEy9R}fApMTC zvFb?65~N0iFaoTU6%{?bU=L2sRyDPx`S_`d&1VQ)709wgO7SH zqYVD*w6PxkKJkWmV7wHjgKySAeI5KJvqoylzD`q5!PcPGddT-o13y`>)8;MA zr;-p^ePT-lN~-DLn`kh*`8ti-BzwC?HKeOksD}Qo5;=Hx4Sf@D4m>JVe-iY_P5gK- zFQ~1-jCptqdbzzs4|Os)7t{VYBN>dx;J8$N&qp!bYrL$RRAU$w{<^JH7y=;}s5l$s zy2pE2^c!<2KDxy#K^i_49 zeO2W_w*i|OwovSTJ~yc61Z|3=$|t^?!J0>`_|s!?(yxDqwUyl^0RRdh~CTntsP2GDqZaw zhw>j^PpbO)?Zlf;l@37+Jc+ic(O+O2x%zEt6V;}c72Ih!H%WzMot$B8FkOjTu`yki z3=HSIR<2Nm=)kUS@SB)cE>+%`sCoRBAWJ1dzm0PC))&_MM0lR{@J=QZlW9G z?AkOK*z=R?_P!)PETy)+Pj?a9XXo;be@$On3Eg74y>FtcbzXa)H}N`oIB`}3;Zp8o zAiTu^efT-&@W1Y%#j;`63=1Qn;cojH4POj%5pTCrx#-9N@!}7BBHbJQykuJ24c^CC zDqDac@w7TMXm+puvuHNBI9~23MKc7Ir)ieS_tR9I*zf%cCWsk);RCjU^N6_FUyv&} zN}s(A3N`cnPEU~PW3AmAAZ6KBnqeshwc$jd zOpP}6h~KjF6!9u+ON-h_WDaMGsS8>)Gnt0^hG2#$BPqLa2k5qCq(<-Q6M60+2?Uj* zqixc|4z4HD9R$!%hSHwrxPyoe0f7ywQWB9D~ZD$6yF zGxZ)Y)0HK_cg%>rssDait!;4}#`HG@Kd6V$Ru!f@HH0&zwsM&Uy@ik7jH&;&WX<{r zm3x^?{SQf82Csh0Sy4oILR8 zeDCniDtCHc?pw;prp4Ow3<^WL&OsFaI)&5x$|z4C>2kw*3NG40(EAsMZk@yIp3H1>4F(o8Ct{ zYaLspZy9R_1+DKqyzV+&W=Zt6q-nLUp<$8Bv%_T;;JKOJl{meRq+1tsndKKn^Dnai zJfMHL^ebb9}(j01oKjj{?njs~zs{3twcwK2s@T=V0{>-?=fb96|_i_<81hgkv z(WKkG`Ol==I1$fNmLoDdkd4WkI-IN4H@{B|Z28OWVhpw@)bvLk&K2z6fIU0(-YCAe z!On6_YUh~b;amZ|ZJ?9X=^Cs;d*&7VYfO-{9RK1FCt8}LPVftfY+;*u9L5WF_L48g zBJ!BeHoXyy$^e+94oC0yL%Bwt;1`hCa1t9d^vJ8w4lXI%Ndy2=mI|NeIf;lq0kIEe zzTq|K6;F!kC3%8hpi%8U;XDqz5VIIbxCNP%*lVhbqd+vMT@QYn`b~9H^5C~NQ`uCeR8MW2Q5VsYoOvB@ zp(N?Zke1}k?|4&d)82fxCE13BPAD0%CIsDpAXX;3>+9uebjut~LqKv9B5OY6p+BQG`Xi@a)|(I) zd#ZkiIn%iI8sKo43ALZ{b&YRGOCi5t^z1_gmo>UE^WJ!48Z3U<2R7X7H`3!px3h*d zM&ZrswbJFjY;@aX-1WjAZYR5@&FjppJJp~W9`se}ZiRpgOn4D)H_2tb({zlkhx8i| zV-eZNnMElMK&S@B0rBs8;(HU-h1++|UcYm;y}vp>&2AkWh+G&upHBPN#kHIF z^|!8X4JA;jM^ru`9_v3!X^e%B#Az%PU+jtREJ=K$aL4T zs?DIhSQub+VT}x04aHyd#4=kTjdZzD2ufnK$_=$)w6bd2hSGp`4e`fY;<0QFv{ANv zUJgpyMd;+cthg+jiA5R-)Awqp>#o^s zC?&oQANvp96#c8$-^%(9{ibih5NGam-354Yw%Ry0b2a+jn>$1dGY*Im$f zn9uf2*%p_{25x@_rcs5z;qPP(r- zyjg-Tx&?7lqA)N$+H4kp%OI}$yjCUCY_=AT`ec#?ojn*c$rID4sy=Pv6lGR7A!Gf5}xU*qH9FtO+}R@t#u=!l)nP%bNQetU^VytD-W@klM)P|UGWoR zjoN#fCEGMLmie$)_UD;cCIHhX)iw-Lb z+BKdp9j64+np2UyGEGT#WXHt{+kTd#(WJ)2j26s>2lF;5)S4(qOtO53cLEql#VRJT z?6D+U0t7YaQ4+tk#Y6Etn%mH6F7J{ru_3WXniS(;W=aN4^Yf6_g2E7g)Dx+Ztdfq_KJoxW7VV!rEWf>J3>ZC6}Maw6?58ogp2Tt@&yWh!*~ z-78>G5#8E^Z9q3}O3E#N)WLNG$Wtz-H^Ux`5eA%@qm6%bz)9v!Dv6%^jelzG{JNWa&T z7pAv#c*)4pZ^@ghmb}axHn)vYW_)~Dsbcc1`)h9nJDE1dx&PREt{!_^M2V{r=GEug ztn~Lx=G0y{+s0TmaHOf#8I()Ppw?PBE851WzTDtJNR@_=>v<2W3C+;$|W` z7ha?umRedSe5kjhn+dR#E}S2hW(Ve7tWUEz1lsUI$5eh$;}Cws%Um_QboS=<$Y@kx z>G9fDGHdG!d=Um$@w-;$s%wRTt5EYYgRA(DL+5%f)O?1LZJA@~N4rkwxQdad-*qxq zT_+vnnTxH+`Z6C|p)y~0>&(5Lj1Z5ack9SLtThN?dNS(&;bSY*gBh*xpNP)o`^S1k zD|l2rgvD0I4Qz$^L{xq><*tb$LgvuhbMzC@iR|+5u$8gKR`Q;RPTJUtej-Z#>WaZu z{MHlK&5}sM%}@_pj8zx%gcldT{fRCOO%Q##ko9XKX_H zKJjpWnVpap$q8H!gpdP2?YOV8dllQ^_BB`w{{6?YH#7jL;D+pzE}J|}dr=3{P1_0n zjI=E2Ib|^M`#tdjSDjzoJ32}OFCVsg{s;zV!j6T@URGo{K^qE@tTW!f$a=ix0XxcgG^knHoG|;q|#d?{~wLGU$IXc+A4@ zb5Uz^@EB^%T4&zs!2_K8-7acv*X_d93;*fVLU+3?sIZPjWF|f$L;f(R?sl0LdN8FC z8TB1a50RPFh>Y3<%}T8+5PrAIw1vp1kDsb;7xITPH6pW!ce_l9M>0dnB_gd5nVWf0 zt%GD1R5kHOTghj3CRHc4+qCi!jmad4#(F_A20;ZS<#l!$k0jKrCkx8Rhj)hDF4wR| zWu`l&Z{Zy1B}~~oM5;l&dk58pc#X?QkAO+FbQl&FE^2Dx4}0R{QL8#QK3km}$qzW} zovk)E;ft);spDYNy&k-B#qV{$Dt-~_0!r-?_Wv|L3`(k2m7_Tfy42_|@s%y{Od*-- z5ht5c5ujq7D+YU)4a@O3n9-Gd3423)xhLW`MT)Q#>Q|1eAo-TM6b?|i;oD7*K=~K9 zfM-!D%G>Q$dHe%LUne&YzPAwiZ7+Q;Cxtk;mwM59LwW6ZER>|Q1?}arIjsmAu*B@E zDgvmq$1E9Wuasob$z1L0NLC!c5t!LmRgCDyRIYNQ#&4mf=Kzww?5iq5UYDyXvRqq5 z@~*1b!u1rca{SM@p{#jkUsX-GxDodx-IsE%?Wd3B*g?G6MDA4;K*d=;6#7a-!huKh z=|Zk5sOX!Nb@o*iV8VH9wwhinBs@>weik?_)R^}->V~9+-hjd>k{fHW1Bb(S2j zFAs3&krDhM4t!ga6u#U`>`HGJk=1f z`n0EMf+N^~KT~Vg#)365D15wUqt53vPO@G*Psq24kp4OD*vhL`vMtLZOq37_RKK;`B#!cD^ ze#3{9-Grt1*Dt`RM?>ecFvxKL?vVWT3ve3%IZab3CxtlR+&zuWid76^j=zsB6(~b6 zI+JGiQZ|cb1C&#oBL=0H;tCMwir?;uk5Zk7KU66ImfC5hK^|lgPmhOtMBMBz;5NpH z4tpC8ztMaG>77qJc>QW$eMh5rbL*`Jkarb99D3sCdcZd!@SRi8cqszIYmolAi$B;B zPZyD*#u_%G62L6xJodW_r<4FfxB=;*xFLS0C&*Whr^pQ?ywuSHXrMhotzjlTta5fE z0%C|SY|&v^kJdK!ox`s|8O09oSr7LvC5AZajcSqnNcbYha$9hB4rj|-5K3v@O%uU| zHha^YU)oI*Bej8Smr>L62$Zqun&5v3{ZR)NSpd55>B^X1(%0a~5IS^TUPp(&(P2FqM}-MwLO zO@BQuQ5T7}>@NN8UbXfh`{H&gg5H=KS6rf9x&aCe&g$2%#+QDRu3DSZrR)JT=+RA= zvj1v)sXrmRqz!PpVdd!c+Ja9(!7IvEk0|Y&mMw}3c&!+WnYV{ zFO7S^^GpNJ)YsxS8okN$Hs-8fi%)6qt#1h3o4nNJwALg(@B0}jr;atQDs5Ncsv}l< z;#!iI9#W3(_X|GM=%%$j#MxKFsrXCj_gg4xD#ra08;SzJRFpWG^*9{WJ%P`-MGcCI z(^%#Xg)me)sqO)M2B`)-#b1AyAvGtw5?_M8qT3kv_uA^@cy$O)*lQ;(1~GlYs_8<^ z`V&}9;mx3?VcDT|Jj_^;a>|YMcM664=!_gOd2+VBI)3H${_14=P)(lP-PwkTX@H}R z*d@7L1hyQOpuC--2A}uU)P8_|<^%79m118NxM z)4@*Nio5h1P?xF!l`$f7&Jd<)vz+;J8J4m=ZjEty&1OkGfjxsgezekKJLllztJDG)iM~s0D|SZ33sUnFV^6vYPGcBCrhTf33ujCXIL(@ zc6jh!nyI!4ckaWWGxU4REytsrhHB@W6`Iqan{a2%ISum~3sswN{D9wtyJ!KTGd-+f z=2Z~V9D6~$?MT*dMmy;I8hRJ*sdzxIS6h@@eefP~21xF})scbJ$F<0Gr@!mKT@1`= zINs_5h6Y{Kay_`SF=w6EHC8O0i^7U|^qZ5jboUD$pel#pn28SWqHN#A!YKO`kG1nu8gEXqb4m6GmlTXL0xZ34(~3bh?=wOgc42Omci!gKc+C z0P(59PT!x_O21t$XsR{Y{9{%hXJXQ?i2CR)U1<2{FZBe$#JzoCKo2P;T#I3O-B}PI zdF`HXac9s~Kd{p2nv1O_5O)!GXJ-`I6F<`v@vn+Ea66xaE!2+3 z!IA`F8p$^1NSbVnxsjh?n-04Q37xsP7Jh^8G*`zM>}X|6}>}DW(0aYoe|j82N}-Lu>mt8 zVTnN7qv(jqp&>%|M`;$R1}O$+vcR?2^!f?;25_Cv=vS$Re*^trr~?Zv6TF_M5H_NN zC03HU=}9v|bv>~p*#=8cNqYTZvO|QAlTdkXnT5XIr(+}^G#aD(jWn~=`+bX&P|s%c z!93)xEofr&Au(dms+ApJxoDsxv{s~R78oh}xz?F}Hv7O!Z=9U%jusji$@^UEvWA#zieZ{c=HT%dmg-zWPV%BYG^85*!=XVXOpe!F%jRtw{?J$cx zlip7Y1i{fvC0iZTJJR?!Aasi3hNO}{u!^yAC_6g!>)D%P(Vrz%FS<85`4&gk<(M&M8GaQhZe;HP6$nXPPfUTzIe8Z3EO_9q0 zNcEYwq4PLdh9BSre2$hid@5dmHYd>Wg3p@7Np~mc#7VU$pgp9p6UF6c(<7S!dc6e= z!=<;2C}2TY%s@96XhKwX0RuCjrW__m)Q`$ER4_L=A~n^tOh8)&=$L_(rtTHcOi;}N zY-wuT0K0~)<_ixukk2jnxB->#hG8SGfS#!}>xX3qHyG&wP&LEU>JB%cp3UflRKWE+ ztQt31YMmJE5DmlP2FnI+KvgNzP*L`A*{u1vY^iYr>f^F&{^PQ0r;Quv$7QobVr+ef z!3{=!%Yf}esHmJis&NB(o;7ZOR()Ik&chBy)GyuqtephURbBB#pD_1oj(w9F9ATbu2$UryV!ERyneP1@n@ocz;C&w@&YTlr z3Hb){DVj?9-_i`*|IYS@)^~t~iuBvkK01BQvd-FGMzzXTmo}TO| zt!+;&Xis6|En`CGDh}ZLfBYzl{VEX)S2+6MjA$X=!;IZ)YKO~JCD|QElPmpmTgZM6>d)$HQE!@>I|rX zoTW(TpxP7W`ZY%lHH<2gR*tz$DAb-{#o1_27^(IIzBn^`6lCVc#~R3tw+2wSM`6_H zQ5g9>3gojHrO=N!v{INJg%P#Mz+erMk@u;NfcgWJb81Gbv4(k}#~X5r##+-Gk2k8F z8f#FSSd&I|zJWE2T3EwezuTC5?>4ZQNLZ{P+M`fWS*0i0R8nB7M`4E(MEYO@wNXJ< zhFx9XO+Z0=BU;IiMjP<&D?VzoyNRg!JKXFp!W^TSugv&Bf% zV+uB>EQ8YGa937Ifv{%aT&ri`{CPZWq3C-9|7w(5kSA94wlRdOBqstiu075`Q~!Hg z;?Z0+eEjtE;MHS$(Zs@A3Ks|E>;QWg>G%%5xG5=sXoHUUx4q~$N-1n=cjDx-doz}o zRQL{;Sp))Uk5e&#S~&@Vvvn@;GK&$uq06k>teyPJEJ9+p%Pg{BTUl~1v)IzD89XKR z?_w{rCXic_4&qhRtPUmpN#p_*`IlJ$?q+(|^W4iUqFeXOWd@}u`IlLM`sT8GSW;2b zI%hjAEC%Man>7MJdY9IEsL$tm1_f^Pn~hYz*(15}nZid1N)!_1wVs734*j{>8-}Jv z>)D9cdM1%1hFqjNtf4YJgG1wgrvGf@O`=XN>frH0O=wQWjV`p2>O#wuaT^~_6Xgc} zNXRV%BxXs&SV`q|rG*Mbf6ew`tx&2nNp_J9>QaeX-FBtrsY~ODYH9F}k z)k#NxxmZ=i)0@t*#>>TOtlJ4IFOA1MuVVt*97lh-XeHWcj;mC2T$V)Jh!IMRd%Y<0 z#|*H@%NNmoEKcN&NbiCZnb(Vz?oHD;ks6Y2rXESXFt{hGAhOS`Mid*b7ptk!3u@n) z?I^4o=#U3Uq8u_Eh2HDMMl?XTvmc%R@Kl(N!p7^xs?xwv-s{D28xzv67w7bMq8cdC z5O{OyQAM3M_J&40qzhDtN9@~XSkUP;6;^XpHLL=OU-kh2H%;fgyd+Tr9eCdVxCJ`q zV%i@wfew|p0O>WVe!u4v=$>e1-LeMXz(0~#E=teHRT|{FhnrdC8dw8=)mVz$tf=b` zd*b61Ref!h-DH?cU$AAd7DG7!>-kbdH+nM4ks4ls^!CE9Y|${PfWaZAp!+d4r6Pl} zVz6F0I|&Y0BBxV{gX8_xo4aRsUtjgNcCHct{4W!e<6yT2f);rPTTZ{)C|K~_Jl!(a zh&0l$@E3_#T0nUgmKLl+OOu0tE!6vsq{+d)z;&Evv-H}e3?SIC-AG8f)`C^iCa?iZ z6uDPb1O;h#g&CBslzo+JAk4k0VnlD`s_ND}n0r-4$m?=dMV4!;NZwTyTexl;ET?K2 zHooPu2SuES=`lc#VriRAPMb=3y@s!3vQ-Cuxd2KRjMJ5 zf$)03P?`kYAh>9}8n5IRx@z3q!>ZD28Q5~{wYSlMxBvqdR>I>(f16r9Pr_}mH%*e) z9w-}G*z@XqkS@KM zTe{5WngbUsXLzoma#>&_L**?lnDevj>Nk8i*-eZi;pJ|BjhExo^>guZw@r3}{s%sl zZqz_tx?NA8Sl}ZLOXU;!ErCAGax#mVL6AkWM$V;+EJatr=+=FxQGF`nX)2QJcaDM0 zfU-KLn^_+bH~R~6jYsLMw?Uy+z~AW!QVp%OdjrO7xNWS`q|%7?;J;Gr2;k13^GJMQ zi=L=>oNfb`Is6)w(IG4k8rs}_JeAO>dT+2~apt6sGXd1}N5wQmb=cp1`%zG>yngAW=G3|Zh##`yJdMj<)`>1C#T4A0@*0wb>5D&Ej%(nMU8t4vn-kOhWGy?<28oq9|_fc7OTVobd zU<$K{O_>kc`)q3BHTR9FAC8bSpY1!wxtlik5GR_|P{lO7TU{PGe7{ZmrefS5vD3bc zyI%aZPn>(CnQ^b`w!Q@p$CQcuAbbLIdreZ^E6og24KU%UY#F*Bd;+5@qEh5hOXqus zcUBiYH!!kkv8F!i+`z~g!v3G_Mz2Kf1JcM=s(8Krx&Po3;(D)tWq2RGv5z`xDY9=! zB4^YRX~Cdlf;2afwshSyVWOAp$uE)3vKsuTa|0v$HL$UgiWMZpQqv@yCYu%?yeZb> z4%h$D%!-qhm{@8|9kmppzKxcV-it@pp^aLK{EML-09Pl8GsUL~Movtg8yNX_+wFe2 z>}C|GO4??J%PfM2wEdI8*vc7UxwgOj%PdCthAy*it$+EKS%k!Hmsw=NwzAOX28JdW zTe`l?(hg<=*WGz|-HH6mtcmP4veqVzuzl{~!rob`UK}Sb}?M%6GqlHc_Pn#(hs$zay;Kb{L z3sW3AE^OKYCyh2ZwK@kE_OkP&BeI=djo05}GYD?7X8grjT z08-1YVeZ^Of1_ORK=^Y5<%~|#ESDM67yurBS6?o?hCukVM`+OPrU_)yZP0l9fqgl; zfv7gE;^dKFwQ~9WWnimd!|kDEg4`a;;4oNQA$cTNuxbMmrBfA2Aj=UNq>o|4Aax{I zKyMoaC3VOKtI!^h1^-$E0L3!GavaTQ9Hh<-42f*D4(4&#T@~fHfZ1ublZXHT2HY%l z@=`5?)VYCekv5^Ga#W@5yjd{8a}bQsA@nVU2^@zm*|~v} zMn|1modX|vQ`XVb>jtuA1;bci{N%rRk&fT0j(mP1EwFWxy zyg&CHq}v97IZ!$KgY@R^p=VYcqw$R;{@vwLAqs?U(8*=K_;ifk-rL_h+SRazwDce~ z^Lj>0kqdzuD7pWxC%!k)1$+C>+3R=Cw)a=Zr`e5$9vW~}f)DT9try*0H#VL2uZwFp z@9S?}-x@w=8)%;pkM$p=G|pHY^Q(BsqF4S(dS_)tMIb&tj- z)n-s$Ebay=E0~O_P2w+lVwo+F2Fg~JmLQ3iTDO_j!#~~jvE(HZYfS8N-8J!I&#rpuVqFBsE$UCnBMn8D z>n=c$v;FRQ{&g48v&}ehD2YkB_Q);T>ADNr6!Y0m&{T`b6}3eR%TCd*K03zyH_iOQ zX``)AEkAfPKc3_E+ZdpAigxD(s51C|+cYaq8_juYbq?C>fmBItHO-34#!L388A~^1 zjg+(OKwItTxU-XTqt|Y#dhN2M+^Jnr;%a;Znk5g*L$4n#%(>|>v}sqIntrr3c05&I zak4P+Be*S zv~l_Eii3&ay&^z*MX4U^M!u;x0R#ZpBQHaP&mW8bP*4Axnh79w^+BqIDwd#e z9jIwo)UDDE$d^M@^+`DTWEvKyy7cHZEV7L72lQM3S-Mg1eJ&>q4J>{BSKcdnIa$^; z@>Tp>O=J0ZFPkha>G^OEi)DYFiF4*c4&u*zYTZDD>=?%Y`;(#pHU!`&?KKPr9p2ug z-OWqLrrn_Y_+*HpfnB!TCF1=oN25s%j97so=aEjs!+9sbx)vL?Pa#KGvYdyvNdsQZ z=r{!rAAqYtCt`yJy!;sUOp4#y;$eFp&22z6mv_kra1e7MeM&*a%mYPBiiO|BB>t$E z{LnN=XscH#m(xYf3bkL8nJV-!({)6^0|S(nlVqvJ#(dY&gsMUdEhw`H^3=hbbZ{L3 zIF+f;=ee#UqFcb_h-&%8c$@PMt|P#sayea84N<0A*I|=uuRAbI7U?jQ%%1A3x6vIq zRo#J)+UG}ypUe!!X`|0?s`~sGGZd-aXr`&H@$PYIrsYkk)AFoLy}rE$ zJWo%L6*8qV;##JSzP+jH+sl-xMx1Ce_1Ya}j}sqk@}7EP8#+W5eLPObnM}LmwDER) zs@{%UoQbO$=EL zF>u@9v9@(%&v_k>%QmbekrTa*D~zC;U|Jq$8cJ`qJbGWTL6(eK@;)iE%8JL+S@2e z7d|GxLdnxPcjc58KnZK>sfI^lyI)l6+MQ=zI6o}SAQLz**<8>?zO>?lE>sy9tZL?U zZ-n?GA9R_S?u|84qE>gHOCqPWpn}zp#v{|&Sfeg8>PIjh6&l!BYScxm9rcx|gd zUuI{maZCm0ZHV(Udo|qa+737Kkk|pZcT9R#L7Wo%|o-M;@r+in?AQC9`D3`b&d6}KPz8|pB zitPMpan@j#AUb-SNVaScn!r~5wYvxp($RD4zpwE+vpFra^#i5Ig5mfO5syxg_B-eNY>(s)&2hMkK)?1_&@^T>nav(?GbYX4yGY}IX!7+FzMTVkWZ9=vkJ ze~N<+EEB&7^$(>^5OxkVKUP|f)Ueb+pB?siOi2NLNEa8Us8~>Ss7*?1uQ_766H+C~r4v3--%n^Cql_M0A!PihwQcVNM42DA4uca?^lR@hq zIPB^z)6zF@wDhUv0bC@ZN9_+qB(=qn4c17~hox|PH z9R1)jhZ zrbQtVffS5yh~Mc6@~z`3YXdEhbhDLGrvUtJ&uwa0M-MA+-wYHt!;cfhhKv- ziXD1>??quLp2u$`REy*X%ojP9+k(V%I9tqcFaoWaX5KUr>h|6Cm0-oJE5RdqWhukIo(H>$L`Vd7Ccrml~IA zvIc=7m zSD8Q?b?LCR6=kYS#?QO0a`1FW+4g*b(3co}j zHD02RD-BJnZS3JuRtdkidE7>&N9HB^i1rfwhJA^?(1p0>1i&%+?PQPNw>9P}$y1o? zh?Tc^EzJwB=7MRA_yr$^bW`FU;%uy;)^T_(JLqrtehWlR#kfCW15p6{h7u?71BbD? z=k*!)t3g;t_}7lh#B3Su%4@8|(du5;XV7YZ)s&HkPyC(fm0<)Bgo!Uf-}=>!?43Su z{Df81g_!jxu%g17K@G{W!|TdbPDAoP?*SdV@ZKpb6^hbIkWIi>#h0K5${dn(W*h&q zFlV74tBp5EvzDwgTBNLgqkr}KTi2Yilwf_3?+W=5968+bH3%R-jMp*~6YSNhe zu#f`{^vafT1TSY>c*jkM)SCgB6qK^PbrZwf8L91@;2QxOY5b`9g2VqF((b7Gp1^{NC(D9BN_`xL9;>Y+$ zbi0$0ADK#A_5EqR^xN!$W?Yl$H=rw|8gXs-nAO3VnDi^64tj^cCXk9>>Is68d#g3T zzV2rWvb1(P%r(DN7;T2Urqo};geR;0-Q&HZRe!60Mf~^{1(EXTYr}gy?hLx>2UdMt z6A0pW-DMGNk`zDF6Y(#gH-HJBgDpgl$3b`Pbtee3E5SA9ND6I?x$8+w&GY0Ma}#u- zyQI+0I(wc+}&hNw&|j~tSMHdE1&6!TG8IdjhM3{uTh3?#*Sn2P#S z+dCv#rXD032Acmuor`Gjlcm=m6^fFMpBPcfQFI;`)apD@GMdS*`GYd$do@0ALS?#T zmNyvyD%FQf46qd2Otx$c23V@W0Onh=W9r$gLYVi!Yl~X$XoXnPdX26L+e|iTA|(uk zT7x82;jhV7`sHo?(UtK;gCuGjYtyLZH=4<&0X||rCtIOicdjA8Hj^zi3fyWY^N=Of zkPkElAW<%b0lb@-@px@yOIjQ1HSv=!MIP_CVTs>H+ft|gc)qx_>C?7&PygMF^a(%| zP~OQ1%;fyz9rp;Cb!*y;f6+Or0y1_~E5o40fe0#Hbo9CB$1M76qQAtp>P$P7=A-hH zFHz%Pq0q^W`(Od-1F%n7<+~{s{aI4^qRV6XJ3ToTzu6PQ3S7#02S}g`bnu9 z4>iP4gyyyyNtlJVq04=T@3YC=Sd9AH%w#SHO^}hB%eLJsJNAug5QS<~B+e*~Dm3VH z5AN7>8e~Bl_PrBNISo>;CZD1yrT<;e@coC$GM=^?`LI~_=M0VX*Dr)G$SDSBX~bEa zzj*^-2!r}zAredG9^$9BMEqON0m9%@+J+c1m^CP(J)x8HK!!s6<|iWJi#;*VFpgHU zK?YZ8@pU0h%<}<}jM4Z2>}D);W;za`CYkUB(8L-#fs948m(+@oHpn+9`ebSo@Xw?eK#<8VIeK;^vFn`+1F zTOjl&q)-RBUV~$ez6I*ptXBBXJS6vP)In}R)C!9_%$leJ1DUB&heD4#%m+TJ=Hrf; zK^^iw@Hr1q2lK9@n$aRFp+E3h$%6U1L#NfZK!pL0HR>R-3xzuDaDqt(JIGTIk1cD( zE<1OlQ_y~iRuZIf2am!>ZAv#0Rey(@(q*6wVJL7##c_@__b7vZv0CIIf#1r^ui4n_ zvgR^ZGk{q%DRqxD_>@}g0dGes6rpD1SAW_Q-yJcDYJ2bOY;}BgaQk@sbanXY?KkeM zPVR2MwmLap9d7UK?;Y(Pg$P9Z?YhZEyyzX9^!ooZ_inwCB}d)&)b4S6$k*~k-O#EJ zfUIo~dsnSG_KfE|_A|yF3nOG9E?RB(w%?X}cYC_`l&6i51`GL;Z}f;SfB*@>@FH{c zU@#aY8-zJX!<$(v50NV~BO_x)^~?p}M&7GdL`ELsxBd}%a{2`y=P#)jrq7q+obq<1 zwmG_Gj%k}?no@jO&hV(AK}l*SX+=-hK#x)Wq8B+-k(?J$B*s*qj={&*8&-(wm`pJoUfg6S|KYDccoA>=~(OfUi zvdg{t&q?-nR7$W7)j3>uhoi8hxk#O=jWqsdmVqIq%}mauVpkr! zM?v8Wx^Iry%yNcDo0+7!giWiM%kej}jEUnmv&;jB!NT3la*rN#?~xo3#;!JEGmDXK z=Ch8ZX&<@GEWy0Y)pgI^%rdd{gtn>@^{uby{LL)kzg%|<9wX5vK*K|$hhahzwC_!N z9=te*8Z`Yc&4Zukp3e*+3A=rLJiSpuax{Cv=i(g7(e%HxO%26tfV9+2aQ$8uwoAvM zqD{Nefk3n)LvfIWglS=rp)zf8Zj@=vLAIp0>KGpeus@InVC}fLX*hgD2`SZVf1t%V zRI2HJX$N|x+J=w>_|G~%USI{J9uXHW061mYSALzIG{y_S9_RE5Q`Yd<< zAf9|37SXB=VSc@7uT<@9sbP@f>2Fl_Mf`ue9;&WS(Bj;FC?S{JQz5uiKv_Oqy~C;ARXj+SkN(ORQppVcag|I0r%BTYI@L> zg6MFw7l|r%k~6e%9l-77E)of7VukRYw@75-No+2Zwor729AlA4*b>jf;b<)vftW8` zG2r4H)#cHv8g&+A>bUgtB7K$7<&9W;gIh?(%u#t=b7kW~rw^eckp5R2O4_1x+MDL@ zRO?`6qjsdF0&{A27l%LQKr$D7pDl?I((KoiKPoq*<_bVPwzFS246u1RZ%Da1LoF|A z2S-%rd#9m|)|;yv7jxKXUH=O*BkH?3onZ&78yDsdxqc+5IyA2{+(x7bzw{3j{W(a~ z30F6EadqR?LL&1s#Hj->+uIHKQ0J-}^(}esaD+bB>kQ2;fF|MxIzwtdskbcq&`Y6V zAGxd2uv0LQ8WOJwysrKVqJYqL=a<{gy%u=@oc6M(7aO&jc8&DHoTWuh^G}%*Utk@f zgp;+~I?~bMm+ndJ#3jywZfTLdp|X=@uLB%E+8U9c*RV}-~}b{N>xEb+uCLzdw; zJN0uSOG|z>Vre=A#!LSUQy%o=5%kh9*;elpsaj$=i>f70p(V`8pQUjx=%cJeunybP zj{yL!=~bJQ-KCat&|2~o!rS|UpV*Sf-l{S@WT+Jz%iD$?l9Ls3x2l}cN7<_Ond{Gw zzg1<-8?#kq9ybgT-d2@+a4#m5niu;IXRC^duADK$9VNM8*2;Gxz*7HkpYDR}t%}EX zlNExuRb}EyY^y5e79{V+MPdS56%X(xZvHT~s=cSYb5vtTr@Zev;h?z)4=_O^vKw&_ z@F&)~mx4*COpeZaoosL(spRHHCD$Mu1jwWsrpn2{I#^Y{nA>aO7IV=3to|?$Fj3JP zy`Nne;z;$TFOXvnvBkdrAXYaCTq>vwF5B7ft{z4{;zB$DIasu))2X?k#L9^V=S~jR zvs(T1AR6RFOH^$>UporiB$4!@!Cj1kSB)+^hz2{Mx^s7_U>Db&?;H=VTq-ye(tbsG z72NVXfx@o%ylWC$>REPaK5xxILk@aiBw|AK*op?{t|5niIt>?t{pBDTY?xhMGH6B1 zz*$yrH*eD0*{cN?bNl5FN(P%^bKhfgJ!W)lKJbGxplO2r5TfPKT#m%f!|(U(e0iSc z%Yky9a&db2nF3C)FRD6K4*)$hBOn3g@bk9bd)D7hsc-s7zNrIirB34|fZk|yXj8y4 z+EaAAeqYLH_9{v*)y>gIGy{s?5k2K)e71Bid5`|(7PzNig; zJwlpk8g?pWWRf`I`aj&(_fF`14S0G8F**L4$;42)K!lu3%;$NuiGE+xD)B96k_@;d z$Ft~=`-7>JbC}CPCc$@bQf8(|NdbAwL5z)#wj5A81x9jUG&U4(IY>YeBWU-WlcL zN5X)j3N|_b@&g(v*cL{?#u(5c{lRPkmm=yUupRu6UGD=2t*34m{h4*TUi+-Kn1zUQYC$zr&KuTyLei>AX4DEAR;{)Y3D-Q%axeym5MGaixun z{nd;8DymxyWXPO&3m7A|HI#8OZ1sp@Co>FTLdtkxZ%S54ZR7U0@4bD&?J>x$)Xe;7 zMZpKp?lNteGVhS8g3clZg`~OZ@rkaXptfZqnb^{niS)okM?vZ_NPBdzq>vnkQL;a3 zkHL&_5har}D8XXKQWcVW3}&n!ax+pvdFDB^woI8nOs6{GyIit!0m3FZgjvc)?lG9N zb^z13?k4!ipP2<6GGa5!kdfi~I!4#>X|eANSoO>K_GuK%a}NBGs`@1 z7%a3NgXt8^J-Xk_jxb?Z3PEy@!HjXyW)^8Ws$ULGPBN6MOU^{%b~s`)OQ0IWv*Iwwk=kux)NZ66gXtg-fpOJ3&|^^hh(V2;hQp&{AOuT2NJqy`F1T<)uzq;W z%}gF0H@%>bm#Aobp?l@oC;Moq2afMCs2vL-@^VrCTmLhGBny>~JGtO8Dj#3Y9Z5Fj z<9?&5hYyqsE`~T+Ked-SEdh70KT}7^rI8CR(C!+EEU;Kf>3R&Bwu*WfXW{9z&U(=_ zI|v4=Hh356@v|y~m{!x>%9L)bj^w4p=wd32=Y8q%3Y zd50`W|F2sM^K@GA=`eUKPr_{QJ`2X6GCMj-a$-Z%+zl9GWOfZGY2jZG4NK)zWy`3D z&~!-aeMo(d!ymdw8=@Yi10+ABk#f#Foj0VG9?^z0+-0m&NCr5f_TTePL#wdq%;*bN za%<{`LFkwwY|o6S1L_0A)n4}T;I1&y68g@Z);XNZrg$+@#!aSs>4ms{dLbU>sSANqF!QqU`XKjzMM(Ui*%6>d9JX}^9y}77}N)o>Rr?2 zrFD4yG2iXSnu~P@jM?rooNlN&bEkC4B}Pu7IA*)c8Gp3xZk$YU%yySCbL@7PdE_u` z_}g9X;RCUvY%D(M?Jg$A86GTa9%nNnd^jDm-6brUz3%S$+g&EkM7OlUhu6Amu}yc> zc9#HVzT1U8Ro&(M_8I^!(E8C^bR`s8{jTeEXxfqi%x1)XFYwQE?dqHU-jxnW_UL9M zJX$#kV;3p*%+p)Ab4Q3)2M&F|!(~+I-mM*cZt4=|VJa%XqZ8;b%#i}GGYUNRFpm;3 zYL!iM&%|!(x zD&gyIcj(Yb8qYpC8~Xi5OH7kw zjSQJ;w2*-x5fV23bfTJ;5AX&NYL&DsNi&7TjsI}c#`%EFL%7QIUA5l3H0&AmQXY3= zA7<0(2Ux7&NUDFERIjQo8aj|Uay{*L%~L_c@UO!Br6U6p95j`CdsM_2YbCHCQv9VO z##XYkTX2}uUW|&rbR?jP8PN6W+QN*=#MTp92TGeT8eROQBVkiqPak9Hz?&|5B(+2} zg9&;nikfr=r85nkkwuz8^Ym)9Wg(1M638Vg{!5Tk(Htj%G$N)$Lf9K;#D5JENUB23 zDcx34VdT;^*3$$9i(nr;VNg*uA00sXL5(~2OM4l?PV@o|m1Nu5wsHB<5A((UV^yr3>#V=)%4FN&N<5 zYu^`dKfZYQ=7WcqXJ5Oxx@}dbY5wXG9CQ@k^f>P|^~&_62Ar3D3o<-W|20w1(J)Z< zhayG-HD367mHCn{?SVREQ9nOXuZ*yL$V$fev3iGPTg;`i8UPS!Pq$BIBY`9e0JBjy z)S`{?{He4%(0T5@&Wv1UGP$l>W@OEgVkzU6nV6@1gqE3GBOC)z$8o5gYbfIR%KC_PMPM5A5)w0r*J@o>r zSfN9gFtU+kWvL%OW@NeKMOV|clNV~1>>@~OO1zFE+EwV`Wblk~B^B8Uis-dAuq1{bhRNaC$9X| zs9?|B*W0Zh6Yh(xHk9kDZe5`$wW*<&?_)Wo)jV%!n7dm>XIeMt)sz+Pk>Q#kpl zf1Ri|TLb!=w=Xa2SSDKcu@&+3%RcU3SFcWAX2m^xCk)bSIIS4R!^%a0!2{pNj7X-BzoCLDQ;buW|Lgxx`UfYkR+)XSs1KajSJ@;cMqY-z|@mHLL(u|4u{iqcu6M52GS7DdNr)qMD2uOPDij$C< zQQpVR@P~5gDD|vnC8Ukr+FzD1r##$?79qVR*^zx)u@oEoOj(A z`zB$ls7}j#-PiuSwP=ws^nW0vkO@N@No2H2V%*SjX^qZC+G!0nv-Y$`SSPcJbTX?* zCqsQVPxGR(td-Nrtf!wK`|D6nw51j>$n(eZE219nQ-<( z-wf(xHVLh9by#caWw4&sSF?V}rS&oNCMl?TO7+`dq`oGMMXQ>qB|f80$T=!HbNfPbZwQ z|Jg)+H9N8H-oLtdczJQ>?(M6Kah=7;v0eK~taI2Wo_Y$F4dBiv@f60JRIDAeNluh6r<`s2wR5QU|*Ru6n|_uG#y zZfWrRt-JTXIX(GQ$e`)%?rZO>^s1dd4%Pm0sLJBGnGkj^J(D$>wGL%q%dowZ!+zjH z1Gqi$eQE9AINVBD=LW8RppD)YEui149lZ>;1r{|b`Wl*W-SGXH*0z_91;kF) zx7vpMp&r+scnWFz!VpK=zLnATv4+@k;;JRyXm};5hF94)@@amS9xmr%$bS3Ex1-p4|hZK43t|VPIv{bucOujTbS48)3QmA;8YMyPB)k%-_sWJQrEK4%Tk+H5|of%4Ppgne&f_>*l8CUiDPC z0Jr!?wsO%{EU)p4NTo805^ArnRFTo45SMp^b4D-HlP>2AO2xgIG?7B^;01V4ECi|S*Ci%_PYeB~3*T$`FZJ|;I$-i^>AhCe3i z&nJcZD)3*f;q%hhP$|RHeK3LQ>Zd2_b&*lDJmNSZQRoxRb3Y~%3G4N#)F>pFLj8E6 z$h!^*2O;EZcb14tnKtC%Nu}Wp_1*~`-2z*Oe4OL2nM`1(%P;^AQl4xsSiYh3 z_2yg7%pE9UIiB^08nH=&MX=)SG#Peu+?{6sq3tv|QzQB<6Q<+XH*crOn8x2}G7lSu z1ZStoJ$697r3N6`Owi4Yxzoh3JM%?V%9{HA5r6(gR|(!uli+Q}dbsE8G?}20URFzI zd(4fQ-Hy4_Bt)CzS$UUW#(V0!2GZd46iu&9*Zf6e7tM1bX|Q?vjJ^}Td} z?Ml)O&pDu^-Y_HYZ74|x+yZ@jCt_VkCFyH-{);>M9EYvfs8E+)lI{mI66%_# za|X0Q+k6A|0Mw$zc3e$$y>}V7q9vUJ?L7e3Rvz2ml(IHAwD=x?)OYh?5mlnEb%Avs zjWmt)jc^h2Hs;gRzt=wMwauUsJ<7;RseS6m1DbX{0M~IP`qrUoQ;Cknxt|SD55QeQ zrClZZ8Wknq2#<5xX7gfZ)(Vajh`dg1-uJfk61xT9pd(-CqUqB2Q+QfaP z<%cb5-JQPZLDWih(|I$9D)3cwK$Fs>!2ZmsE-=Qu_>weWZdCBb|3qtflaR{K8c!sAzb5O39bx-18A`Nv9vWG&o!>RO8+`Hv?B z#cuhxYb;g5ymWC+cxmzb{*Oup$}J|@%r<}5aLa>|toHNSg|F$6eMT$c=QN2$KC#^B zFY9&7zS5WB=;oFWug)&+fBnIoi-%|TOwY?(i?gyZ5jah9PCe?;?ZEa2qwT!@ewi#; zz*XuGPSm>5+95h^^;V$}v+UkNn zzKLW=#Blv^pXA1mT~Pu21U8YJp%EsPFz?B3A{mo-n@HwC!yqVB8x*Hb?vec_QjKEC zNwe2qrPST~cP`$!b@lk$7t@pTYm7}KW}+X3%^}tJ8sSrBD`RXT2^DQz_4cisNG4Xe zi~Nx{pFLv}Nl<93J7woXb@~MkimG4f9Q3=6gGTDtwNbya_T%7>23t_&DjkG6CGR>C zuC9%6wE=|gFv;5R%=;@&xv1Uu@|`=5cDb8Y*8s!N{jmW`)O6FLBT+ZDn@EXjp58jP zLmXPSjK{8V!WZH&-QggHCKJH9+yyNmxyuQD*cj=x3yYM7sPl2kH%Vwl2sko z)n85&VUmO2jbI=0Y2mt=Zo^gq8cVsiOHWW4(Q6AF<6_6_k~n|Eeo<0HS8U>_$B zjxII{80~TOrxTUG2|WT6(1Qb_*XBXoREMcg$J^C9nKLQ0)MS?G=h-@$W9agvCvyg{ z2v6n&vowX`zHKsR;v+qA0BQ5s2rg}7*f2Iv=7g-Yc`%%b$6Jfq<#R_=jYS74IxFb2 z#Yl~{F={NvVi?>&GXfP;(MgJvlx-q0)y9aa=p==F`-IaDnM5i;l#T0)xS{q%^h1g2 zqv-s^4`rl}+8BLQ(@+lS2|ASD@ZdzJQtjYGI~{wxk0%cjH)v-7Ix}(Cnrx!R0OskQ znY0Ils|W69)Y{<^n%3yL7n5mEK5ck}E#TX`k{*38fN z0YZlniplgcP#_ZY{D>rPZKj9bLw*h zfTlORuKp0F$e~%{FVTL|rv7;K>dq%xXX@GMryCM2#!=j)Q$FW6GvAiyHtm`&uhT>$%<55AjfJ6%_U zFZ+0*O#kmEDz6L0OpCI88EYCdczn@H8H4(xlS17dX6G>LZ}-$pB5p*cE66cKMoj(f z1Q$tVqjuODu-&6!6BRK74J1>7-aLSrjL#7Pi8q#_0-T8-lYvN%fgU2#`FJYhiu$U- zn)nGJXb#hT`^3-04H}&3HYq&C3tJ^?;wQwQ-J`M-e~Xn9`9z+Fc8`mCd3%(K4yZ}T5 z%;7YM!4Ky)5)o{Th@g2mxdcIptk7Q@iP=A?V=qh4_N*mxJ>3Fyr8lM}Q&P zcd#jMPQ*iPfsiu@iN>FtsQg{e31NVi4iN*xSu>I7X`JmK*AQ}Xwi%Opf1);w$uSTc zFuZyyWEL8*PPm|0Yd|Jr6kP+PPy=fXh*`t2W_GjM0G%d^wFV@#p*>pnj5Qz=Jt$HT znGc0M5ceu;4M>Os!>clD0FD|$KH_sOA+^FgmN{%AwZhh@6}U^Pe_4FMyd}btg>58R z*fvTQKpSjqQTBIznh&NyvaofMg>9>3VJjqA@Z%Xt7R=KbN?2G&QpiHZh?vc?px}h*@dAF=Pidx zh#jcn1^vk=WapuE3aXbFBte!tbkb@03qCPis>SrYH8C_D>4IsYD5HI2_2gkIZKB8_ ze&Ncz9912th3dBOw?~=rbKsk`|9fA{?*+F_swj{hUf_mduaxtOIj$FHa`T|DW06LqSdSw(Eb2ce+1pVm z!8TOqaNT{JnOAkO;%{ad7;^m0tk{+332tUN!=ueiE;PqAGUsn*8575CW|;>LgN3`9 zftiR21_qdb}EBt42RhD z2Qt#P5HC9?SKEQl5_Q+Q{y^HXuzghgW>A zmP3wgWPZ2r+_^P}IpnFGpxP zG@HrMe)J72dSA+A`Jqhi^i_{RcWN_Tu`sB~aRY=8#gWX@jU(|}yj)W5D9Gq1WHAyn zT;IR!iYS?eJ20q=P&U`_1)K&I_o$+(jY-1>9FVsW+FB5poxa7(2;XHJsJSJ*?Mbn1zjB zhGVf8i3BvULU_+xBr@?NKNku+Wp#xge1B2gMIvEKJP#kkF<6iV+JNfv=txOtL5Gs7 zU8Kv~8C~9p#W%QxW(+E?Yp-nFMe@3xk=Hf!MFLq1`u0vvW+aV(3b;Gfc9mL{6d(b~ zT$;1e0g@lmNHVuGk~!9p4(So<6wn_@*ZZ{V443!{kU9-*v`%M6PG`8YQnmiFy`5>a z?mA_!sPEK`JicsiMOFL0W!ZEu1doWpFnE3+Y#UVuK)^IhqgPvOo#a< zv#`ixg?cGH=z|paWzRJ>YBlW|xrTX%JZ>@jvu7RaP{PUDZ9VA#11fbEEI|!yvCDWg zwl`FEvh4MMW1V*{f4HT*wpTx!sIO9j^;;Kg>ETR$@>XAfck7@{dNh4TUBCJC^yHZn zc50TqPKzi}{a~WLBxQ7@)F>$px#N|ZnI0^XtNPRjCp21>JUyigR;6*Q5ShsiLs+!s z9L^CN{+!6tk^zWVnjjBLe&{sGoONJjl!I~lT*b8NeIiv$UWi52lBduT=H$=PxEEZQ z^5m@Q%5-xIT+gi9q*Nfal!Mlirx19(Klq8wMD|vd;UPn<*jU~+^pKpakh@jojE=S{ zsc{)6DdcWd8S}<$Rhh>PLxi_gk*KoaVs;$>j4s$4s<=X(0L zS^hvpZ*=M#h8U{ex^ts?W1aOZ_N}_b&kzMfgSu+Ao&By@{sVHbG;Zhs*op?bb0-Hg zpGEY)AOPk?3#>NR`Vw>y4Lc)>5JOn*CV7UWl@sj z$?6W(dna_h20T54m>hr2WMb%4%GLSV(DqpkkV~x+-*P6&fLn4ri+^tOhY}&jF_(i( zg74s@%!^4$0eQ#1uu?uOoQ`;%3NqcJ5Nd9aVm7n+tYxi?j!#KWa0_2bNywwZqVh zTxAX4p%Wt{3 zar#sjj0;7ajQ!P#{d944a=ZXW%Iv)y^`DAt4P~4RTOFd%ll=I$mcGtmlbGC=X@C3P z+ZRu|$6)41TjYG2ZxjWFrgD?i)GqT5sVe9!Qc!TqPBTAsntu66WPQcs6SXZ<=66#Q zTiP;_9+>DTq=zFc+mZvq(Na)*3}*ZmQL-p4sD2|cOU_VDA?b-ENzW?<$!(c3)(^QE zsh~Xb99mnZ%pa!nAYI5oPAuu(JYTDhI=L-V&e{P?x_xxHnYqFhrvSYF-UsX{%moK&Fr{6 z1~bM*n^|=wJMR9K_75~?t(ww(#AcRIU*_sc>@k>e|A)4jl`@lR2#naw5){mJ_lL8Y z?L8Ns16^IZHqxbnuKip$kA!vR@wE1HX+&AEpx?{u zI3T-q6P#Gk6xIh!JhwgLx_N86w*y6-qhdEYD-MGksol)uxr01PE~qhBJ2h?^4j)lK z2$oJ0Iy!c8!MPKH_2Xl1X7csoO)u!<1s08W^Um_)Il!Xz;PZ09O$?DKDknEhgIrLk zeB8?g*`Hg#)kaF{qyh z<$~tvw9b0bTFMp-TDfu%4C;>()=clty60Vw!ESE97(&4yxLT(=gv#|8)bB;^F_?J5 zna~u{u*cvgG@v8(7=&+N`TMr+wUh$@c+Kr52wx>n>7yQlo6uN*dAjj1e#LI34By$R z5%aG;&Zn{Aq2U12a8Ksf=s-#)wp^g2lY3H+!G0j1>Z~Xr|BK5`?{p}iTFNz|X|w;J{$ zk$@(i*LaIWCZ5s_Hq29Yxk3`hg);}{boL^VuqB>{pYS39kP@oTqa!8Tkgl6Y`n);0 zA>D8ghdc=#JfHZZGCMj-a$>`IB(s~FMsW>22B|IJys9iJBAnY+a@`!{T{JJC$~ihf z@H-|k2H=!jRsZKKdv_&2} z#s-@1?0|uAI}e68i+b*DLLn$?xAmmsIMssTt#Jl|&NTYd<*z%@)0c=|;ztbIkjG<|_0tS!G^MX9LXpQvAz zG(J*o$Xb$qPoawtfA5<56+ttta^sj`Gm{?{0{V1cXM}?=kt?eIov0mqKqB6j=B&am zpJdnsg+|@g>fcY)OKcAe2^`;-Gs$_8F7hDBmR$|<3w{0T&GdML`5ov_1AbJTX1uw?eSyXS9rnK&avSd#S|X+tcC8`YTY zE&dl%r>$%?niPN2krdJF%7Ri%9He9zT15f1Ybs-}F5e7eKNHGR#S-94k}G zW#F{$cescYd*tV^Gr7 z-lBbf>$(1#a^-z;R00_Pb?=$6ye@mWI|8UaA`GCzsxXSfiwcfk*@)?O^vLJvwQZnH3SQY(V_-~WyRnO)N9ZBg7 zs?UET&7gUDwc4@}Ml1>B0u}$YOCT4K_|H7O0Y&MI!{^E9V0filR2aE%jrA6QprFrj z@T3E3J|>VeluEfl?IJawc{+DcEq%hQ@IF)FqGg?cwu2|_e%kMNx1sgrbmrtlkX=+> z-kI`p$DGdPb@0IY;j_L;mz9$c8ypeH5U zP~ghkdp#!s4O-{~>PB--<_ z6(g7RNV}(=`KX_ts8>eVK4c|h{8+uivMuILrM=nCM$+{pU>l8}6CTWJv{ibgV>ou93@3!c|5D`+m$a zlZiQOjtfsG4u2WB%p{y;jOWo1Io28!BtRX|%A{Yc>AdB4oCK(g#4_gbv(YPha09w@ z`n(zFUFbPs0%Me`JMAL5OK0RR^tw}WhUTG5RCn6BZd4r*B+$HoDpu&KpD?nKWTi8b zl@TM$B`-SLXqNowF~D5r!^>014%Klm$cxi)^t8m>_~OY6HG_BYVJ=??LZ41|-Len$ zW}wCf@-mOe3;8kyM_zP$LecV5)BG^Vd)9MHUd+!BE&GxRLX|~AUR?R{*~2tEP2uV#*i1Hh+uHMh4clqSHq%O{SyQep)J#rli64TgI^5sd7sFBRy{Y} zu@hMa1yK4N87NGsxE--GdlIw52O8rKvd7N?y@_oP#keDEcSJ@n3bH0}PHX95GJrkx zuM_oVYe0YV_T^!bD~}y zWBm}#jB@2nIP%g!4`f6>;@ZiL`re6pd6f4D(w0%)_dBe-Yp|(5cE!nL%Z|_pPbeuw zGjheLG8%ES>+47Zq#3#5B$Q@M%qr)0Mw97~$V;(4vms?&`2~ z6MpO|!Qqb9Eb}@UP{D>h&oxS4>tQprrq)?6`TE)?bpY2asS%^~1%Y7DTTJEO!15Zr z-h|O%D*tO7V53i!!Q4D%oyEwpUHeI_S=c9@dJ1L|XbUs-gOYxz_6DhP zD8AHG&eZheLGLE^!AYSPe!F?LwaT$Vr0}h~{XK{GzsU_m(i^N=w|7JRywrrdW>`c7 zQW!T=yVLz#tMLU8$k*z9cbkq$@ z&eRWMtIC)+W~<6PZWtoGtt$86zWbM){~0?}$kn)8RTU-6*>{}Xi@jAP7@4((?s;2P zCZ6(J1YwCz5;MS*#YT_4RVBQb=Vl-2Vteww?%Kc-g(Q9NI!WI$lJqT&q;JGQAULPi ztLH{A2`KuaVuCh#pYQ0kj1+y#MnxZBNp%xkM=x_9VUTStovd#OWPSCAdEzOg?F&O3 zY5SH&+s7JW%ZaO&_#jZBo0n#83|QPW7~X$E`aajqYZ>YLmXW@%=>>hXM3uKoC-Pg? zuU@JA^p@l~8LLn~8 zXb;oZf=9T#;;L#G=63au6ZPxVH3T#VO5xKm^>lRNb6)BAyt+BvmX(f~7zICg`itsg zhl^0ApM2#L&s>{6Jw175a`WWfi2q>#eWLz+Qn;@I|K*IS5tEQIa14?7>4|z>WE3rr zI8I0u`b6{G@As5bsR2ll1oEZ#hWhbDk#}DZ$Pq11VYHYv34j<@8b&i`f_-qiq24>8 zqg!C>kdJfxHIoT!F~8;)VJuS8^*1mJrLQ;Na%S#83Cr=U{0zdpz}#sv?C7{V&Hh8% zX>z7UI}M3k$gGB?t1=XCr^%Sc-)S-r8-@gDr^!9Gr`}R7ZQEe{^knWdRoE?M)<c-Duqm~YUX##1^}8ceTE*Zk#BBD;*F!Am0z zX7AnTdCU9D1MC&5?@db2OE*>clJr$1?=??vDM?2)l>n}J5o=VEzH~Rd2>d*(!4d^No>YAr>2GlZZEF{B#f*ydn(z0FKOm)3?8G1!adnNkH%47SR zQdZ`M7GH@DK1E>8P%bRYrOM%x#%gXW(O0PA6@-P?HZLPSO}t3?@O9{brWc?Rof%f9 zA)xO8Xzm*ZXnJYUwD}ny2xBtv4T3>X`5H}L)@ST;0k>i{~PLo zWv=>Lk72I!ro06Z*A*&02Lt>U07LFVo4C)k{IEr>IDCB3L#mbPrt@YlwB>u&O6{<4gEB7CVGwQGohU>Gof% zeIIq<@hd)1w`x6ImIJl?VC13*bIXTUXBYRs{@~8V z!?Sy)=jE;LtZYmToJzcClFI5}iZsX8JJBhj>P+aS9WhYHDzNU zyI5l1Kdq~i@1GO(Hxrd_LW5X0;H9nZl=VeDV%j&642c-7AMTSJh+gcA*~#hnDOJ=I z+ucUBZz4HEBTOps)=zd5$(Y33L^2N=20=l>S)4k#M}7#KNbUmQ_Dv*aqQ_b7&G~mw zVInb+7@J5!MH^SWed{KYiIp@P2p_*B+)~KA>MLyfCX%4gR(BG1JD<-BEMfs30`+UU zu+cf_cP$59Me5g;QNObGH``BH`-F2v-|=bFNxUaN10B$uYoQ z?xxjM)5E+!P$eq8@q-`RNQt^KO4P=&txI?ZDF&4A>bmuICdzXhD$b;XQ77SDIdSIU zQ^CrJGd&n>dO;s1QRj!%+R;(;($S%7y@YoaBhs#6E%FjxeNnThgm)45fLJ+BYI;Ca zR|G%v6qN9;hp^OKeceM|K&Pt@Hzps{tt+%;4qd|g$Vzxu+vznOAhYoF*|X%#(Bv@Q z?7D}v5!Q5DIuOLWI+(0^Ezfh;@UHB4Fd*JtG==!LElAdDjRDDR9&oP?U}t?i--EC< zw7C2f)%TVnuhZiUkmEk*;oMxEO`mOobL?}B&J+Vi4sAag&n6BbFYwkE<4+|eqFf)s z$k5R6qtQ^$k0~o)!XCh2k9S4k3%^N;U#!O^l-a#;|L)b@+xPB%SA#VDKMx*dGJRCx zcm4DwAG>R{oGys53tj^2K&t+7q6m{5JS~EKL*?T2xaJ@E4sbn}jEy&Sm^iB6KcNsQ zVBZk$wRvYYHa-M~JJ0iKV^L~8p#F5C@;9MJU;=t@K=j%?hq(j<=6dh~s*j@c6F-!ZK5A|BQB6Zx7n6nU z2{y{q2jX^c0t$I)A%k`XpfeM9p{(hWc*utPxHpXfJTsxbn-`P0df@(9I_nlMgjJ{iH>q& zvbCka12GxZUw177(8FdJz^92B50H_qQ7t5xCVvQ1G`($`-#fyLNU{# zY(JAA$QUS^B>d4yp>7YebC~tFduk>TH~7AhKrfI@$P7d_qfmc4!9`Nps2#QjZ1*Vc z4F5e=#B$ce&t!aV;zzj$y2whrKFSYR6F(*cbZd^ZL{P40Rh)^R5Q63~-M3HtOx&b< zt?=9~AZ>Jxu_k^(4B9=4zEn61i_DymZldx9x)kF(jvPetg|(3{pi0y7A{s2l7Ls5H zsym?TFx(xf>qvL7HoAibdO%G&e6ko7q^(_htTi|b)4YI+2% z;CC+65KxY$6Qg#5u%XSO5H{%dbOu3i-_!eK-`qgs2?F)qTu^jTIof*T2*Q?VqbCTP zh#+)mr3y$>wdUZEiFcqJO@bhFaXH$?5(M)z#7Pec!ZslYt{knSvNulYTb8>Up_VPSb*){K|o+#|lx0TpXIiMl7GZfr4+v&)Hp4iz;2460c^a>25glqOn zS53Y~3rcu{Xc;fO?!)UjwVQ4l!V5lMh%e}VIVevLGY+uk;DmR!VEOy>Q7Eo z{;ubQFhEO(h=Jj(nMm}cYXHbKgodDf`l zk=H^2LIY!DtpPD>z;^n~WIPqJzj4jrtpN#bXph!CV-3he&+*p)7I0v#0SR$nc=buH z0YR;>0cwRbpP(~`cPw+*L}~@|^ci z?qlO53!6DDEPM{V5t1zU@r)!3=IM;_OlQ_PaTAcJKP1IU0mF;b<}}(!RKyB9XeOn(Sf=S)KQK) zZ1TDVDh9Z3se@)0hB}LmtAkmU||TKs3WTrlOtIvXO9lCQonm(hh z-!v`3Pfow!^@f;< zw!y*if@LN>9O1AkIUq@I9IUYK`k^_Gx&1eH;-LgN3`9?~R4A<|avt2s+x}}ZNq;2X~HpUvFgA!^3glW;jSlWKmE%VSRJq&WBOf!$? z4)Tx`*F6NT_|(N9?YOvUI6N~JQmVQBK$}RZwuzK#O)u!|9CTuf6r~>ehPa5VE=A(;r{zuLt@A=}%W5|H{Sfd#~N~AkAIz z*`a-vYzFLp)vP<0=8De^dc_32KhSU4{y-Zmi3?AkzRd9=&hQ(}NL^ zeZA_Dxq*8-)x-8nGm?7l{Nku@;Fv zZ;{Bv6FwK7#3E7Hd)bRb!j^a*4hLMhsfJ2wT&}!Db$N7!q0WL1^kN(7^32m`BNpG_ z7Md}ryslkjxQ*m>TO+TdS2pU*!7uLU+j}`#RN1(7cdBjCJJm7(X^Tqc=#`CrNF&Ld zc{*=Mhx7<_N~mtsDP%i1($2>o??=W{$Sr8J&ZKUp1X8*FNLwpa>o40gPxsZ0&4ICu zsvEbRt!~^>b!bj!s6F|#;W4&t(2^RU4iYUvbt4@{CVi89f23tp-KdF0UqoqshB$hl zx={;)hkVFYH!fSN8@CpV$NiDC+Ltz=-?HpOn`hWZ?y5BG7-vqMj(mXL%ve0Dzk(TY zXuI>vbeL~WtQL99(M0}bk32SNHSHRahxwTTUqB7iCqAF@&qE0(Yq#~JW5;+C)HhO_ z_oX?H1baheC(B+3IP`?aeEdW#E1kc7Hc?-t1najh*wRDp;3PlUA}&O=8b6q*FG(3q zQfkP#ApPi+h!NEXC$vN~ad1i(tV-ipAu^L4hOke_Ih>=<_&JfKB?Az#G{H1E`JvMx z4>SE%Gs?lZ#IO!Z-zQSF7650-?-sVfx|{W?P0H?)UmVH{ zN&8yzlm@v0o9rY%s<>NKhKCHbVqh1{(wXLPhxNv)9kOsD2#ODeW(}c zm#+AH3FvwKncC!onr;Ugazuv>IuAF>LiO0{>~^jphkrT^7lZwUD;Zo`(Q=RsYS-A; z2xq>!Yn)_oXTSUb$>4I;5}SLA&Go8C#^wV*I0Kp{*bgCU4bA09>^%IwW#@PqJC~h@ zpUIzY>TWjYp8%9YGXh@X?Y{)~WjI zRH_VgN-2DDG~#pqi_i1w=5$N`;$xx~{D6=^@W&HH`l2@Y^$2N3OLy&?$e_F5sgyes z^qDAx)O#oNqy>0-2r)VSn#shFn76XpzQ*raft{KLCR@%V8At*-p2c6T_kuj zPj5j%i!h+*KsedoGAah#xh`@$kdmX%@dFAJY$*_<10X-3k%Db!6l{zEwe$zG!21ha zin!5vSaY@Lrd27q1NGGO1U_LwUi7z%%BjuM1M_JP3s*te&#tA#Jv*u*%!~dOG3fzt z5gQaUxeiwvEQK#cT-r*-j~^;Sb<@P~Gd>W|`LCepuS@ujrl|J4(Wm&|Qm(n+&LZuB z=&vI_=7FUbdhIauBG;Ne5&iiT1xWsO)az7rulSJAhsJcDX*tzeXi>2-`D_CUAYb&k z;YxMWd2?=<=NU?$>Vk2hh?B9uIa$BbT?R#%uaC;1LD>XAeDkk;QMOmb#cA0lb!OU5tppY~-Jvh)c6x6m%BokZO zGLasb=qRLz!|y2Uws09cPC2qaYLCIp??seMGO!ZZl~+T=$#50$nzd$t=M=TaV8;3( zHzO64XP!fA%ar-UbgGL6Q$8^!d3kh?!JM@Nn8tPYab{lCEh9Fw3>g`&PmZ=Bz5WDa zqe&=8?lG8-dCu@?GmFT4DJPR#m`bWaa*sihiQ_i2oX-vt>07rS-O`oG`r z*ZypAj473o{Sm%q@i()W@jkH)NOD$=*vt~@%UoTFJqAg!m8J&ai;Se(@<}Bj*r+`Q zbJogiX8YsJy!vM5F(Rs~qjO>3er^{D>vrV!bE%K{V!<6Mj-wM|Cl=gA+Boy{7DC+B zBhg2J^75~!I(+AP#_gKAa`{0<#cp&~90oa3yX}nHjnqgo9b^lPs{tE41muER*V0an z0cfxOs{7`kuMs_a#~sl~XO=XxKGcg_xuCh0^qMBUvqQc3;pqV#H@%>XmsM0WzSFhP z1|;hzw*;S;3w9AiUV+MSlFJT6b_L4EQ*R_?;>B7GCl_4B<>Q?t$>wK#aMX=eP%gMk zaI&i*U(M@}^1o`L4>9nRq(KIs~m4DT`Roa-IsIRPCIS2;z zM+y5PEzh%p!R5LAVh99-EAW-19$2o&p#D%r?lG8n!kN(Y%dp4bJT#yq^%#V2VEOwp zJs1UMddlrF2!DuuB7&{<=Ap3w^DyaQe5vxjI_RjU}|B!txVVae}Dcr?iGcT8Vx0BDW7qevDNJojis1jao3Gs67UUZ<=9Q z?i4LGep13h?jjLmrz6g5VR0d~5otQ4gPWOHiv+R9VEW_9rXv>Rr7aX&2-4j>tWfMl zB9R&LJp6;A1is5Kp0Lr;Em!|_HKcrAOw>&q>Ia0Cda7ZmZq75nN zF{o>U+QCt?(?AvW7-nR93@*=M5w`x9cG@rWR@@@rQR@T4)n4|L$X;Qh9~v@yG%&o} za2b=L1|Rj1Xm$43oe0vvqC+Bgm+{J0Ia;~l(m^8qGsLL_<%X*vA6n@dWZ#?kI4d_? zo?9$FkQ-9vovz2AHji8^2_{Pr9O@S#4l(R8xCkxjNOh9or!DeW;qFQf`hbUUizvf8 z?`V}=?fI?Z2A2vN=5blME$Cy@sVmn){=DH zr;8AO@0$7*K{Ks#S=fnSsRE+B1Pt;3n4-5$$-}A@it*%%W461P9CP*>+Oj?niBRiD;3%?(X^9T_(_3h4T+m`33){jn`e|OSXw(AE*xY>yPUf`eS+SNB!5{63l=-k;#7?tx@kbX4^eqqQs-Vgg*Xx8+{x1p5P5DcfAT=OsTG`$7+3_qoa;Mn?UTbYQ+KTw zF|J}X4Y+6hA=+g<=M^f4_aer)9De2awEh{wKH$D{L?Xs@0;ye1{Hj$fd{Fnz(M}kDT@lUn$nL7n{~G{{@D?{Ajt<5l*rd*b<=+D$i1;#pwb^+~KVV&7F@5n3j& zPdgs+<5o+auS*>U8jTMv71+jm;t$qQNlQ!ei>jq5d3ws=fHzuVnj~vvs5PU74E%_Y zu<@r8)wFzoH;7PcMzthqY?9iJP)@3UIBDa2z~&)b<@&B#?_C=9jKtAx>707F($5qY zOzdM>I*BIg-zL?os!MM$|L`HB$4w)6o?>oQqT^bsP6a73DlQ^PluBdT*_SQ zbts&5BFH%`FJI10dAVm!)OYit;VMe|S*E!_{~tD}j><1A%8MY^5oub3>j==XaFA)e zkQG6$ZRMsFL9Q)iGCxDK42mE(L%y`uT&U46On!)Q`d9ii7)VJdaAoej zj+3m#s~;0%N5~-~o9U≶#KRL{EN9%qp1(=8_7iw-MC6e|) zov5gvpQu+x*gj+>WBgdX!?G>r(g_xduEb-FG0C(hgeU;aM%|E|jPblbtBCsrn*nOv zGLy-5-7+IPkEB@2xMe2hDIcL_hC*LPE;9*N84>LJG0RLQ=7gt=IAa;N%p{y;jOUNF z%b)#CGw|W6ptk9v0A6Z?<2HjvJDpRPgqDRbp8HLA*$ zpRVKb(~To9ru@_qMRW>8w_R)#^5V)**FgV;p2Qq^SwPiFkir0g5c-(!Sn{IzjUg{Y z5y9Yi3+T%Ns8-M){wDw?y3m$s$;oUifdMvu-Y2G?RnJX#?8KCTT+$@wDU+^x3e$y( z+YvjnCo!GOD~2GJJ&9k&wufTe5w?3Gv;0V&@^;qIA8?IJr^;q3z? zD$80poy>wtq&%(J#}XO|7$D^7XY(wIHscMzj9#7YGKu#Z>+cEU&3pjj1gd z?WgjulBXw1pDKg7p`2>W2JMN>TKeo#DkZ{}fb?tyutj8jD6bl0y=OMKOiKL3+@52q zpH0+Pvw7s*`&SnaFE8%gy?u2tuCo|9d}}|6H4FR1Q%^y5fwnMHKPc&kYHyG#hvG|3 z@>KPe#5n&m^G7SDf$p8eGyPup=P`*C1zg29zvuxquV zH_WP?KMqwb9biA6n~yQiNJg{Pp$u#pws&&aPio4cjlQ19R+Te4>INof>W8saWy~A1 zRb?JG3=!T|m3wgC{i_=9@{=sn8;reG#V9#KV0Ftf^jIz&iJKjJt4c63YYi>k98p=o zGx3xrH%Hj2O2oK2*6gh+;l(^Ri*43>!hy)qmLPuvl0G`>NlGi#yH3)#h$MXrBk3D) z5C}_5eHx2}2uA`Th!vIKZp5CJ9+qH~}>(8{dy>$9IxrLMUb@fXoZ+MUSA?|>- zk3PhS<}VDfmdrqh7;PVGh=-iGKE$A#mo6k~hs8~U;p0h2-{-n{b&FX*CW z6II@JPUP1uT0+lv^V&u1c++qt=?xQd2N1gfSCU3MFFpPaazEG2YvT$?f9 z?e&m~j1GmkETbJv*K1`0WsYDp|D^km6ZPxVH3YVEO5oEl^>lRNb6y7dyt+BvmSvEc zC{6Jw175a`WWfi2q>#eWLz+Qn;@I|K%D!mojh+k@)F} zdR=4`Esr=(NEG@+^W1+b<@fZa4dkmF*yVpbQRH0*1ad^nleB**Z;Qfvf_;O|k!>j- zsrOFk=oZ*IQU4{X0kP4+d8OOp&>Fdq6oS8d7;vCQVQ<+qlpy8=_J57ci z9e1bMe`q^R&eVv0%Y^B;TQGP#O~y3-PLp}qFeErTP42OSJ1?XLpl!=hyqzY7-5Gi< zWsT$}#@uNVyvq$+?`)Hr>r^Ne*F%Ig{Xt0Sx%My2@y8^jk}0X`{nM7%U7 z*kEg@u4mZl!RqoOb&!*cy~GD)$7XNC!} z`hyd-ZnSoYPFuZIsMagHH=UZ{6ei)6JR2pAQM^4()Za{0z6lLt-GG<2x>ME{RjV1B zNQOiV*AMqeZv1c-rNU2O6UiAGVbYSIM8a&s=Azm+k&H>aO(gT6VGtCm4T@7I_s9Vj zm2fO$H(_idF%y+|@8MdXRH;bqOpHw=p`wke-oAAc$;8U3gp=j=4!@2wHjxB{wz|{* zxQJ8gn}|~@E`j>Bv^b^mn^u9KO0n# z@M_q_V4^(70dXdMQY64sISKF5i8BwM3YJEk>A`5z3;HmLIzKEImX0n#`^Kt|m6Pyl z$3}=0w122S#B);610tP!bWZ9eyxZxum+#z}^>?Lg=Z=$BF614~tFNe_N?jzd)HQ9c3anIT!c+j~fV1L%8TY;5TOaFf-(EKBZcWGPXK=Cd>y7dQr+k#}h))BOZo)jktdLl8@cBT22?l*aa_vwI@}7IZ=d34xSdl zKDredpf}#IB*hzwG_R}QKcNsQVBZk$wRvYYHa-M~JD-!GF_tLdg9KCk=|tskLXW@% z^x%N#wRzAVJ_S61*{oV8b0&p;@{<|u{MkI2W9agsPG-`zDy@?_!7NRoxNn=xnfSoe z4%gEIOPF+{W52M+HBaV*th9MhZZf07vg=+{vjl3aD4ywwSwB6*U6imB)%#a%7CjKow+Bc^H~DWLcdeS4=JGKGJK+|fGFWyc_~MW5rnfa;^@ z{KOBX&S~{4&3b**(&(d_hO#auyY9zX0Wq1*X_{-Fw+npf@Ly}avp{EnG#gIXNKq)| zt~FUkjRBUXF@R?#)OYg%;p&0=fY`c#sI)NjHhMAHA|fS3anxaLIMg|Bg)fxd+R9rC zNfs88n4ckz9=K3;+YR}MspGZp?c)l4u5XmNvJ`kACIfg1oq%^O1JKyaFn~`JGaeu# zTcI*kFirjtrpTdL;xEyD(x(1hx|4#`f9KPuGxhBB(+%lUA@N6jr_-sB^M@kjrd`uz z{FNf)@##~0qF_##!M;#_++?3i^z{U8e*<)^$GT*1DpB*dP}pQATv$N505&>;V1aY` znh(>PYCBz5!!-LMXr}-76P4G6Vx~peev)0s9sW@JGWoxZ$wLO9{^+Dow};s|%=+6s zHIs;&G`c}fAmgw0DX0E+f{UcGu{gZ2wcVp(*ORdvi%-_X&t!a#2uQp!n(Nc2p5jdW zm<$Z(nC@a~MSdT@`LSL5i}2{B?vCD#pfO%2>RP$J3;8s z$NQM_46|L2Srqlw$fEr_MzSk)L7*SLKhPR@QuAT zwfan4xEyU8m!ox#AeeGA?1jA9Q90T&AqcJT8WTIpSqyFSXeE^;;YY-XZJ#P(2XamDpj5Qz=JsOcVsTEd6t-xJU!=i|p6f6;zEUY5Q!m3fSP&vo@aq!R~s{2@J zVS1lmwn`ROK(Y|6%k+f$03G~zMv?{dbjEmcse=~swo?bH=RKwo)-9|e-NGu;El{(4 z9~DuV$I9szRzOgX&J9l;mN9i`h^n4C2o-m@COBIiRWFC)j-{gxtP4Ih!ua@CE^gm@ z?XFM0^z^asO5_z==`f6%{PeG(esZE-;D%w(eElxU_2Nu!9)#12q%X7Z3#+#eLp(}-?mmu>ssXR~n^^{i9Dg$_cIA13n_15AXfu#J%XRlLHnaLYpP_hdEhItv-o*0Yb)-QvkDooyJ)apu z5*$>WQJxI1h2&_iEzmlWqnW3-$kBkLq;7)i_p-HJI(jO+cAB&`(4^HL=7{GS2-DJR zj$YOl202ouna6Vnc}R-uMuGl7x_YM_7dH)u_aIQJr5;QaX+@=`>kqV!lxpipskS1V z%>CyC7caZ00&?x-YHI+X>#p_vfmRVc-T^vUe~5P30mbgX6`!fg+WQ0TTuI;IiqGxJ zB4qP3Eo%hq5427=vb*AQhxYOsX5iZ7%eU{{TCUM6Cg}Zve#`a;T3bn6c={}N{~(@x zwJ)MQO5fR1!yv`eklmWu;_%NqW9)}Q&?jhDzkn97_IE66*UvGmo#zQ>0=R6V9?}u8 z4b5h9v>zP{i{A0BIqwkiCmpi@fZzCPlbROcFr7vdPUTq*V$!oE;aZ+5N!K3PJj{$-*X z77Jiv8IPhSriq=P)k!j%)J3VkK51cez}}%;kmt2*PvQD_rXPILxdZaY|Kpd*7u569 zn~lpPFuJf-Vk2xKDtDnh!AP?kPoH^1U;A)%_v&%WCI(;G9sdO53bmKJNMxXZ`)VgO zeyBJwMlz_0+(jbBPW9PIeT4tT~yh)c6X|+x2;|PlDRfrVYJ*SKctalZfzuUtRWrJBh)FNx=|nS z+rbgl`EDL;K%@0Ebm+{;RX1*|RIP6&Wu6|Y8>#Q+10$+#Tx;xcm}(n9dQe|_o#8qr zMF2OUfA1lYaCPG@u5MgANJPDC?|M|-xEb=HovUu#0g8wk>U4%1i^T{1k*K=Nw=Db6 zOQB&OxvSE!!zYj$Zb18S^v(YYqJYqL=a=a)-<()2^1%D+zUa2K+3NvEMSP9X4=HC9*oXSrM17SKtlzp|OAomw zBtI!8Jp92#eM!pbNU0%bYPuKFujKpSghs0p2d8wwsx*!jA~V@x7=M~iT%tSbsJr*? zT)cDZ>hZU={^0yt{lk;`IgzC$zZS7H!OS-Kq0=M>*8!JN4n`+ag5~cMsaos#FlJ>RaDaV*+RoAk&stgYqYQ@I#wxNgQ zWQE+VDrfXjwyJTGLhe?TF>lOPm3iDSM0i_O?!mp7Q0l0#=j>2^0<*WOnCKpHtIG7B z|8Tb|9@|Y;2;NqeiKi#Bsb;7|- zB(mEWk==-cKz;bU`#kI20GV7=3`0kr1_&!B8{9-HIrH=u*1Ksj!OOs+s`8DyCT;_| zpEV3IDte>K2E!0Xs<(|%y|IRPh<)oO0nwmFS?%l>aJu@B`6+_R!RP?ki3T@L4%V|; z{qz8Un_kdIOH^&XSvm^bq(Jnd!A*>Uo8HLtU{H6?oPtJKs41=$vBh=g8^?pAy7T&o zLebzhfx=oB*DD^kEmKyTd{EPEKtm4tLX(II)nlu(+gP1lcsdOigU#R|8LWB*E}+%{ zRgSvr%HL}GuPBOT$U;dzE5OhMQzMX?j=O7d1;MeH_!5>c)>5JOn z2Wm)D%8mqRCX&CHWn#`%@14;38gS+iVsiX7lZl};K!Ge!%IA5FG3r^Z65nzr$v_gw z@vPs9#pVubDQ`K*B=~MQkQp^8DIjk-h_O+0c*UG)qIszDlX6FW;b1NY2`FL&?Vht7 zWMb$EEC;2%6dPC5my)p@By5P|S?T3~c%8up)ymP4kj`ZK3^&rsnWxVf9=$w#7z!>y z2hFGMK=77+uWpz9nM>ETP4*(djS;*x$o{BuJ*)2O&ZA<$jq4(}0l5>&3tLpdMh8G{ zK+QFc!3*Z;oB_4;2eZJBHoO#3OMKeF57qT9aJFbYHJ$l5(cd;Ir#4Ry%!m4JJ}g`X zv6Pv7k6bj_#%rdNDO=tE<= z&$OIsEwtDbKm$5!F}2u>J}F$OZaQyH3Y9+91>-^yCu4uLVm}=is5{ex2`K%OIpP&C zMr>;+<7C)s5rvuL*yzZZ!#ZbC+cNEM-+TMwN%t7c{HU1JPZwp8n%ZUFAq6vMk%EHT zoF(&9r|I#DF88ChWy<_+YGO-UCei~F9fkC8G%#~^Epf)uRCYmb%anOqjgm>mT9WX} z=qac@1~b+Vxf!XTJo6k{Tc*q(rt=_O$N@i-9F06ft0Q)5Tc(_~1DM8j_lGp|$UO#A z(=%jbxIQ`BCO5`%VnF_8mNWcFo7p%)Ab&H#~Zh`0VQ?+;Y$MrXxgkR!F*)~MY`jU>}Sw!pY* zi=QV7$OUzwWIHw9HJl&4{B#S1VCgiWqhlu*+&UpxPkGJFOux{Za%;R0bq*(k z8uIGHO_W@k9)nx7yG9}lEOw%FJqEX~yM}){ttnA7&6tA0?%c|iy(GFX)MF67LC@bSjpw*M z2H_7C0DnD_VXHme+r`NSnWq~M<2%9AO@XV|BQZUW2Ig;)`|t%KG#p?W?#cX;I}+H- z3OaMIbaGGXG1!L}RGk%j3`R#T^rI42P?@|8rYAJb6-W}$MV(gURNOHD(Bo>K8UdRezVvoV${K>?V*dkHFNNAuQW06Q?hCB}o zItF!tTW&(GjOz2~NXa#%GuJJc#G&u4wk0>D8wQW%No~gLP?=r3*wADz!5H)O29yMw z{bOb*uPTd*2)C}4oPG}W=RPU(L#iA3zvC9}6o1eGk{eQUUjXoeQO=QyO@~9ulpE?n zXa`4R=R4}QL#wdq%*e?NP1AqiG*g7_m=XEid|ZFi zlTVvF9~?FT^>A2(nZW8H(eCVVIk4y#QP{hTyK`IR`0>L<6q;CspCMWX9p@ne4LXTnr@$F@M5{4zRH5;Z`)(g%o}pCB$zBfaHzZBVI$+;N5LM0>(G*pR3{mJ z+9D5Jp~fwp3_rs-;72|8bpKArI#|1{Cmnb*MX9r3iGg4v8{6Ah5*j?+Ujgg^$ClxF zlx;}4Xs`ZzqJFcYj(za<)wkciI@2BBAF<_!Gxf<^{c~;|I+;D1KBKPRe0qBF%*p8& zy)gC__0seOiZHhPN-g3<_5MWts-*FeYD3nn^guxuA^zSq^(%sATII$u!)7KwEW%+n zO26^H6SZRxNR7A2#$A$5>p&FfMfmS0>Ls=Zh6Il9%bDc7NLTCt!-S)cD@^=gQoU=s zytEFn9N+E!{3Ps_`lp*Q+g*mM3^ix&ln&tjHBH^0%-&biGLG8ra>gHRyBjA{9JAeJ z%pAMjWga;U8~%2edw4Ha^eL0%e!yyIsK#t}F*!cYZe$kjp|DZZYii7Pm#}2^y1VCZ zcbPcjd`mvkhFGA|YRq<*0A{}1kF(vi2pD%j>z8H{blUv8uGir%68o9Q&tBl4=ML%{ zzhKpOaSsoHWUo!N81g#011-4f);bZRb}U5G^H!G~IPV;k!-JFWq$J(TR8+#Ahuu6Xhu7S_ z=hNnA2%C-zFa<@7iy@?*yPEhpp!s#b>6vzU_!xk83ZaV_f7|vv+__F6{^_%tsw$VL zwJ)SSsh?XRb&xUY#;f*K!tPafkCQR(?AOT9_pqCHEo1lHzH@;2nqyEt*B8!+COPvz z1C^WZ>Fbd9qJ4j}Tz`?0fgq|oecj{F=hSYxVdBpMVXhCE&WOFHPxt$E*0N7K4)V{> z3FDn81At#7PvC`50fv4D<3mdYw(*Ym>vi-=nN~Ra)}3!~z#AK{(pI3KWi2v@njtCq->hW+^U-lbnI ztfA=l!hf4ouc|Kn+>oboJ&lHc*u4egcf`_>0SOssI>Bv{GX>Q?j#xTkY!w}fv2e=F zWsZ)$5OJC7I+Ar~#L|&~DrP|Mxl2bTwvG}0r5(36ux~%QxTQ~b-@1GMo70m|O%#9W zNZ1tD)A%%VnUrJ*ICDNBAXla?T2G2z46OeJnG>}IpbrgKQQFTkdm)Q{C^0m= z2yzjVrWI}N@gm6iT2=%(x0Rb#1i5gKN&gJdGAM%V627EKZV$&m`6fDttO#;-ZqfQc z1WEPxBTG(n2v7<@(={;2AoZ&d2kGx%9v#~C)oMSJAGg4yJ7v^&CJ$pA=drBaoCX6d z2nDXpz1MS+CHA9&TW^7Zpd%dHOfPj!#v&F(&Ozb0 zZL}mhQ9nOXuZ*yL$V$fev3iGPTg;__4Zgigk4gKyGnx>h05BVMLv}L8^QY48Ko=aB zRE=C_GP&+?mKl_gG+@|q%ZwmTksWOkQyI6+BwS@gudr7hd$1e7pPhZQW((K4SmdaEP2uV#*i1Hh+uHM zh4h8`{{IBQL}<&j}EI&~|+lk<;pou-yw*w&>`|+Ri%3{p&=%*&5K_ynT6D$1-2`q4~Ocb^0ABWMyL=#@ihBgz#XO)Hx#DE9Wk+a)4U$4M zBUhZ5(Hwt~prDrqKs9p3Nhr;j6=y$c#mPk8lU#9F+xy5BCm}VXydTmK@tKM(J)UdK zfi@=P8uW54*NS{)s*D1wF^`{(U*Ch5XhsPuWag1VX5Offp(b?>)+}J1miDJj#$mw> z%?m;bnJ~1GL`K)hj2l|6PDYF8+jTNjd4Z!fbD)!vpf#rK1x$-bC(|4tuqST3Ta%hG-e2HLD@5S-3i^1y!@= zXpPaVnZX%d74idsYv$C5(fR^UF!b%XE{m_V)XKksQbuck@H7> zKU3VTDr4T5tt#`lVTkays@#M7?qBt##Al^)qae3c#V9$y8^8nEgs^5DaR*CDG<&N` zFfwZmE!`YZS->;#lpgQFrz*M7$t#GfQOw?|5?;)6GyVFAn&$}t>}G7h#6pt3cb%kf z9!dJOrsPe{@CmaPgkRbGJbH7fc5ORAgT`ZFzTFP*+Db?#(+^EGRT zfwnK*dQ?R zwyZVC7j=U;`x-3^vhC!xGT*(-0tz~LEt*!cy;Vi(_cENF;|FViroGiD_9m(yLhKw` zJLPZYDDKV{uTw16=gT-Z2Q;00cSrF>^2orPe*#<&g}1EJjHl{ddgueTNFDl0IK4W? zvC*Lrmu0kv>3A8gl+h)^UN`z5C+gR!YY1%Tlw24&Og$Z)_?(wPKCf<0w`Cb*CPu*z zp8lfx*x@3S=_ggx66bgpH{M|Zsp+M7J57ci z9e1bMe`q^R&eVv0%f#HcTQGP#O~y3-PLp}qFeErTP42Nh^+p$F;dzp|3WK@R#IQU1 za)KI^0oqc5ceI@*!P|`WaL?IkGC}hMwvEzp9~)QZPLmLAj%TIMYI}6+@*sPGN`vWo zXU$(UtGKmVMABgM^cj2a4lY3l%>hTPi}W4vO5Z)x1R^O_SNEJLk-XPDy`?0bR!V-u zUxmDgO41kh#@iK9n|UJ)C=lwV*3z-HAJ9msyD&ms#()l)H9DZ*P`+?mBZhWdy+%Xd zM^$hcdPNJJRJ%&_g_Xx1E(BjK%ndEUAmY1u@q1K>zR<9>pJsqBqrUW&=dp2ljgE zfn~0GHXG(TZ^~PMq*>E+?0142{sLmi&?fFPEkA5g>*n-D529A8o6eg-lt=~{fTuat z1;!f6>lyp2C#)OccNEM0B3nasJ;PQHRvDKcA1F!@rVAZ%)sITI|5_pYs0)u@@j<** z>*=x_#N{6=36izK>#1ukg5*D*6coGV->$J#3G>o}0iC1%s8pcbVv@~l^LGumJSfR( zKc8LrnjYC_v=V+!aVhe(;l}jZGw)sxE4#i7M>n^8cy)Gh|LYI#Ts%CxXL?@VTAh`R ziNI-+N$OFLq`meBqqMjFewo}?z*XuGPSm>5+95h^^;V$}v+UktKPnK z6UoHNkvEYN39YZN?VCt~LR;O5>5K*kB!O*a9~cm-ex-BJ?>cxGsb3dH{mR;p!=_pv zE)POc(Wa60VG?zI zSZrOGfcA}b*Ln%>CPJhgxPhns5XVV7p&k%lvZB&1?g6o}oYeeG%hCWZdA($)^J?bLih)5^1k`WF@?2?jNGWKYjKrIWw$@cL#fjUL$((Zf8pe0`b+s zWM81g(N4VE+3#RLyxW0(Bz5n8+k#}h))H|!Kcr8s5e(<(`TEYp23YCBb)oo`v!%x#$%wgHAolE>!)1p#ktXlVG+ zXsF}IdIdbL8T^x}PKjTv$0d~6C1PBc>A&P7Yps^k1u?R~qlKD|{^djwCOP=s2=+w~ z5BYL*-3;lloxpe(B&>0$kG!sa|Aa!MfPF)}*XEts*!W0eA4nU=dn{jeiH)54(}~L8 zgdTwj=)nQeYx7{(_QqR_+SNLlGbyyxWR~jZ**ckH=yHT9n!=LPWQxKv0sI=^=E>Nh*WZ z0Xowdz>CSK@8-m0=2W@web$ghd!>qsEQ!|OU3CpI8bK2AW zbV^Rt-%fCmR5ogd7q+&0RN5KnPQ;n`nT*d70g2(0q5_kQ!bQ6^?(8&gc#r=t1Rh?%&=gi%~&Z=h|a+R0`oi5Go>|(*r-8 zk%*u(B7)}OU#+fgn)d%?E`JK|m08 zmLM$8iBrRuqwQjX&=6FIax^{ckKGQR`l$V?jV>-n+c|_ zpC<^x)uVL{)9})8&m|NRz@G?fGa6O*)yE&b{k6k89UubwEqJi>0g$i+QUdh%A{9aC z5Os=yg!pG??uA9RVvxsv*5`nonorLfb3mm}Z-FR;O1tJ=)l-EHK53Fu1u#Y8R8bNU zM0SRvyJkB*UBE9ml`w@x7@HVqs|%3*03&ID}@2kZj@7SrW+>mPWSl znXliw`|Vp_dwlDg_a1!h_Pty8A3VG~%l-dM=0SXYmt>lRH6)o!r(0M8{WQ%DLb?S% zppkB2X><$i1KNZh{$Hhf-f+m@9D8W$Igft&Rb5E5I|6s4F4ZY z(?cH>R6H@rh8%&w_3tnEd~c~1)AQDRk3ly~e>#UCRyu}VAw&BX>llOvcNk_6IgS)K z0~$ZErJ-b_-BL#(^c+9BrJRG8`j3hFjm(@_pWVK?y10LJ_rd+MM;G_LarV~R7Y`qw zee2@k{fm2Nx9{A(ymgtfh+!qW%S*sH9%}%8a{2`y)Gw(Qrq7pxn(}s~COW!-j%lJ} zT2y>l&hV(cL8%N)(z+BoCp~o3_phlp#7wm94vrTrGwI>*yOU^GHMK#werT#=ZbHr- z)hJ;cF)_&iO4yc;sHi_r|L(CWAE~0V|8>2}4eEQ;PfpYe+%W8G<@{E|_2Nu!9*720 zmVRMH=E2Kxm$U%p)PGL0ccoH-ZK&1Zx?9vo^_X*5Zp3Dm!6K#2Ol}eu`|#L33eG)r z-yE@-0Tmg8fOO_v6&^n_y4Ev{dz4uj{4tOYn+K`S-Ds@jP?sa(N4_X>;JRHj_n!Si43s= zSSb4axY*iZHfHebu|2bAVZ%W(If#pM!CT@DAO#5_0|^QoLZXO(P6ERG(J`lct^QN} zba!=iJ+;O@H;G4@dDc_a)m`16=Udgx)pg5FW|`PZ4<#Xyi29nD=msRSgamWleTZ%3 zww~hy6B81oeS74fSfof>7)9Fi-2ItBBz0T73xiQY;@~d z>4_DswGBx1SJI$b0_j@m44%3zk+i8_u|c(rq-*BsfF4g*y0=6aE)5Grh z{Vw+v&ylAwM_#nNl84Dcw|=@-HVa$j=V0>G!DMsor?a8{qUudZlxGQ1cK3cl`d$qg zv;6F^@)EswB54j(nuO~2Ev8;tnOu1KQSMSg{0z2ROB+0-`&#YPhznbP2=KW6Xl3ne zak=Xrw6vcNLHD2q;NNvOf8eOQevqN=JYP5)X7yAk0;3K+XmZ>i9SsNJDL;;p z-IqPl+^F?r$s(G0R&P$b!PHPR$=v67CJ(-I%ANg66uufu=??W;Y-5Bt<4 z=poD(>_sAj2QuI9@XkD9HliX#_977jsAMui-deHqfXhZIIb<&q336hE@RqkoWa24J z<^VTIEg+@zicT8tB9Txgo`>ZZ31YS55@_}4N9Lpo(j0_#f~#eu)iX~&8nE~Vx1fXP z7|IeT?xGz4-7-@ashsd|87c0})0?0YXlIMQz84-wC67yY+1he}YE`_CJX>3$B074> zqaV{qL}#AP8&k`XFkb{?ej6YpeT-uuB&yWRM-SSAJ)uNJCg8M(ODkR5_bp$V9_> z*c=;C<>PW`D<7BCmOw#!IE$GPwTY9*MWFPNKCP0HNaeMMw)D~AqM5@*m86 zI8R8Ct9@Lw);=yRC*NTZ09U{P8H;75+4R2lQLl-Hg5>T}!*-!Sez*h;)Y0GkK4b!+ zFV8R1`Mx<~IQTJ76E>&L_)|#$pSN1u1*C>SrVMO-i)BeZiI^a$h$2$%@7ho+ohq z#`i|*bt$7G^+sulmUBV+K_wL;eln(UtHi-6XRykRLxsqZ>@fKK4Etc5xI}l$QTHC) zy?FoZ>VtPLCgaoVbv&*hx#B2K|$yAiJSTXk*_6|v-n!_6xz`_`Dba| zr>n2P(K-3uLPD;SX}xNb&Uwi%_9cg;yp}xW5c7;=G;1BokdeNA%#k;#by`kp$W2u_ zqoY(MXDT5j6;E&6RFyGrNUF*_t{)=2RF!*hv%pYpsFGyHPE|46m3zZ}Mlr&hgJeN= zs^U@J{u)Z8yvaZNlu}iR@{(3pErhnKE2eF>0!HA20Ra0R+{&?6-g7%Hup+X?noR(iJ=5}K#e zurWw8t&(gfBQ&4xjK2Zt)-4X~TeUdc$p}~WTOg1TE}Mc~_uSkAbp0KU4A6U?a2Cw# zk>YJM?WY1`4n62d6g~XDqv*J(pXb%BNk{(nW1<%P09hA^e-rxg zNRd9S^?p5qn~`9k_8p{G;(RRSj^wbJXq41PV>)RAp6=sJj=zp%V#u$vVr~;r44S4N z9cPmCWP%*e%DV)MV8vSwGU&)*N}2daMSt*P8ZORLyyYMUNEv~jl#%sG-e4&!xv(Ivmo6nf+-r z($ATvAF&r$bOVqv1=YaOp^%dTt|ATG%4pzPpwPm(jHD{EEfd{QztGN}m)tmlsjiRrL#Ik~3#{j!n zCZtEtBKzj(B=kRKX5$`zSjOEK0xpTqdDn3aJ$8m;>gKA59Ibu+w=R6FSxx3xpkbG9~G1TQF{?)-XZ<`dJ?HB zB+X6tgo)Izm<*}yn@A?Mv~MCkFws_!dJ)ndy;oQG3??}siF%?AxyiiyW9Owtp{nOetcMx{Y+-;}YwPtZu^=9%Zv+BaqXVLH|8!=2_9a8i~WfGmO|w{OZ> z+rw&HcZVI5q`B3AACSy4jHJIlIqoKBlAJ)0pUiTGA1Ik^GyRjEL^8{mI4qfE9@q~S zS}($M3g#Z&EESZD&?MarNM`0m)|rv1jyggGasEA`%zv;`nR>w zzma+orlUMeLZ}O38T=TK(FPv^VOZ)zIz)E82-i*+w)?WOHo~v~k$E8#m6WfimL#v6 zbZ|jJxQuZ!AeAJi?a3@CE>E476p9qcP(PZO_2#r+OtV00A>rWK+O}y=OHz_9~ zT-$GmKt#A|>XxZLUF}7L`ul=`?ZljtH1w$>^&*6C;7I(c0L!ay zl5VX6VLs+|D1>hqwd9+Iz5~pCj>qzA^v&hY7A0(7^=xc-Xh*;_+%tOo(qcKo`MY6# z%qR6C+=mHNq7_Aj(dvhNDPY&oBrk&r5Xo)ElBY(X6g4D+MY$5$#7p2QRSZ8Tj`01=h*eV&I>+l-|~_Gl>2!?DDY{Gs<`TVWV?eDUpHY`ntw1=EnAxx}N zO@BOHB(x*r_B{t!2TJJ`3t|&dq!P;1NQp&aD>nUG<%erjvqy(Zu5aBs((IY1)5;e6 zezjx?8#8~h@uX}Gly}kOrl0DfUWDsNd1s#91eK_L;+)LPR4fLbw08aG)}ReM=8xb8 zR7*!kNq$TtE#2B^=?2UdT0T^;7opzbGZ0d<+t4y?Iz4g{#C24qy*5?afgUx-hAqt8 zOAy!Y-jg*zK4@$N3F0bZMk_#`&NzqC}wjJnNfjUUkR3|}PTTb4WAW~(YnoaL1h@L-&f{^rMU?DV_#W4_S zA4S-Ua2h(-k*X-ePdoSlEajF=hKCkF3N{q!x#3#Wn{OKGL|Ho<$2?YtsU-}y+uM5? zE{N@IEEw&-fI}U6);GF6Ohvive*U=jUnBJ!70K*7kFVZ+e06g7;=v=fBoXWn&^Cd` zU-x3$JL;v$3ly`^rUGC-pI%`nR!qiS2sqth!naNp~4y*Vmyrf+sm@amK}g6o$g|WENe?YD;=)y;gEEf z6tc6|-7P=eW#UY9t6rkW)fF8}o`cd|B535h{V?gSMby}#`ae2%{v+2Dv5N$N=J7`_ z@b`1AQXD61>`)b7yR5N`M11DyO#~Y7n`+#8S>r0IgYR6MyRNlC-A**u0ZCpdA*U%c zVbm&&bEL^LkLQoG6*l_BRQD0=ji~7wgG#Rnryuu1p6GW#u9kX>j;5Wkv2${@{TLa! zT0qlHZ|Hg>nqCdz+eN4E!nKbMOHJLoLD;yCA$611(vxbTVtFTQT=k0O*N#+2#qxD4 zBVl8g;A&SfzXIPc>JQaM?*xAxXeUCSZ)tDDKd@!^oogrJpMF%6S>?`ayY~>^cSlkU!qc?3khTj8;AHI;hz3rz6$0et_pj&^3cg zB~3_H?`%|2nE(b6qsF zr)uJQI;3dHctqoizjS1f!l$`(pn6EfUpiudB{{cqpKavEeLG<3NRSmXptszmBNJN~ zC=y&&N=X=Ap;TN?A7bgi)sf-24RnO`LsSB-ls-i5Pr;FnuroTsS}7*FFbCJ5 zgXp0Yst2S)Dc6^=i}Zk<(F3;hWu!{hyeu-Rne23F+uo^bJTtn}W*Aje-HPM@4GvWx1^J5zri=o9tje011y zQyUcS7MBj(;)3~A1z}_tGbic;9B$Gz=c2tCO+`BtMs^lAncwlzfKGu}7`aJEllIk4 z7^w@^F&FtC^Sv22mbLeVkyMvmIfkOg&M=gkb}wBFfi!YMefxiaG$C~Lt5t?3Kkm>; zMUt76It(lSOkN{TRm{FHFq2UK%FM!^nBbg9Zj)EyC0KyQHq#4X6PNB)njVyZaF#QW zvmHx!ir17zfwgOMt$klV|KQ@`8}B^4Jo(ne)wxxorYBi<)Oo`*zgN_q$x98)FZ)9n z|7f5F|2b06(ST6)Eh1VnfL`QIIMNR5pUBk5V|8bM?R{!8#E;ecFWX}7SlV~O1Vzgt ziK4)4)b;ggLp(3;Dr`j!DSy~9lR5PsW|=`*$rD>F<6%*N?D_G4qNi3tp*IYgr3 z)Xm%yx(#UDgoavRFq??Mz_w`U13ZHP#i^;6k|NY1;(JY5GSVwf-8cqgic_&S@=Jss zz@S~xS08@<*DlWQzjDu=F--@!J%ed2PTfEu3;YSRbCV7G6s+{YHSm~FkUJnpJ0`Qn zi&j7Vqg8BsDAgTcyPr3+V_EwOTQhb5sec)%H(De5t@F#vI-vQ656xHA%ahxz(1gDv zL}>n-tTqU12z-qQ(O?Tmj=_lgqz_@zraaN9QvLT>y*y~RdyRUCUy|~!JQ9w)ls|$a zlDO;J8IJn)Slu4v{hr)qkoSG=KNfclHZ6v(IGJ;^V>Qu940{J@i&3m&8qp6KsF>E! zZOn-6gaGwJd(bj3+{%^+6}T5&Rw_as*wPU;`J;v~#wkoV=M9XiR=!>N#7#`iAY zM1q-(5zGu<-|@X%YLu{QW)rDqHjSzoeZp~Xh}sdbO;p9TajKaOD0aXGIe-Tt)l3-M zNH()EvY8=cYq<^0IiRVSKKC-X4S5(m&=l6rY$ENNKyNnNEq%ajn&-8p|5(Osx}N@ z&oTIr*NPRPVqP+Lu z>f+($#oc@7R~N%Nkdb4%_LEpsv1hJbgP;O!b*8>o(iGL+AXO8^7owKF-pRQjrTs}a zGWE&0&^^E1JllfhP$5$I{@wnb`(X&{yPU(@+pgSHACFXC>RB@^VuJ~q(_^2u9J>j_ zepD?vzL}$nV%}0e7#C1Q&GNo@i|4;Wm;dUIAE<$M?RN{wx>f~$Q|VPZzaOgoVqdAn zb2Gh*ckO)$@#;=I#%R_$l)*0jHc$=#N=*c`bJ!C}RXL-hu3-@eJapbF-H6IgRT=Y! zq^iv0`XRzgRk;WEx`fHa+>=aIF;vc=VmY%8lLy(UDgnx@HMDfsL}kIu#1o2;)pR@I;glh1v0A2q7fcdT`PmCiSEo?rr~jleX@km7<0Q8~sBa~F=dZQxwzI3H(_Y-`xb%;= zJj{~M)v1u)FN`tN`|Tdj8{WUB5JT<6z^a;qV?5()`)3U#nB7`O|-bnm=a1 zz|vejBraeU^`P@1rF_>f?c0bVgF@l@74nQH%FC$% z%>G5|;ov)dnLdu@n*^Zmx|y7PkCq48j$f-&`z;Z4{OanLnF1&Ou2V|t7171aCP;xlhYnEryF7}&??C(Pn0$pt@mbu1dAb)1QcH7w4Q7VAT-LtE^L6l< zQ);hQqUhDNDPw(XbSTqh8SSw;UZgAKhZ0Y(`oY}I`|DJw4%=~%rgGK!W=93~_Rouzs14|zw*(J7NaJ|o{$ zKOQM^+5y2DadW9xPq;(B&@q|{Xs5j#bypva=@1v#+9%~4e;vsLwtTqcDFYTVY0?>~ zAB|MUnYlf6EXT9*F2RgvW||B|DrFjQw7|ZpN~N7`G3_ujP0rMagO{+3%qmg>-_b`q z#mh7q)A*St^RRwMa57Eqv3ve4<=?gheNVg86#@xH8zmmPS`6@Ge-`%2IwFANgq2?aIXt< z8kk~s^GRYi4NQ5m{tx?M$2>jMsZX0D1JfytJ$K*7te% zUV%(6?I5+&JDdAFOvpP(J+qM7{Em+XbYav32#D*v&!;~HpWfwqR~dq$>ptW{qOgjLeZY-mgH6&=aQ)=+!Tu+2@w5r=_i+ z`eCVjxy2-V#?9YT-14C0eEYTR%GdPBp0-QE;pqVv`2cZqa{cKKZfMz)k)cy2%$!~y83fwS?u3pEMa*9V0G+80%&XPpN-Y3(b_&lZS_{6 zl&|dGbZVx_1T-6va9@4_ua2hryOGNGs6njjk<(Up`t8)QgyD4)+mlF!N%Yqb_i=8x z2}Swv6G$RCLnB}+VcwHXA{mo-NhI^2eh?HY5QpMUadQ8@8YRl=VwmUhar|w zu_nEhr<3(gZ70aR_I;|K4iIV68~R9reIZs;-Roe_VnK(;y17BtJC9*$(^H(RS6`as ziD|lM&^voNS#Q_t5HWYeH0lshw=$CTE)s0&+C+3T;inBuC+nSpzB;MVbXo5s$6{Su zojsW~ZFQa{9}vlcUTcKA!q?muWtJU)m$g^y8??+HC+MBpFJn;93*?^lF?-J;WxeJY zq}=8M_wG$UspWuyx`s}eU!;29@nlq3U)tCJ+W4FYc~f;VdA14U<;kZ3kHNIi6{K;O z;$br07-RVUqy&=dVFL|agmxf}hI*2W-f2!4%Hggk?NK#heX$;xP;= zjNli{bvb-Lbg(kqC7JfklQ{t{O`*7Lo6MQ`IOt?9Bq@A3Y@N&rV`=js`Z97eeH)ID zK%q4Rt>}Zn{$$rhDjTEFVl0Lg8W4&)8?44#lri5oSj;5GNU$k3=K6GF*9ihs`I;d&*mH9%V^Dw zuO`@?z-(c0iTNGkyom&6%Y>A;%V^2hxjKQ_%wplbz>M01^miN$(9p}!fG-h`K0sPF zLxqSpdZ;ccS?ZG41HPFPDz8Q{Jj-IQ1hA@jH4d&@bsQR?Lhz56H3rR2pl_Dg(plFEW2(w6vFf&qw7LEWuZ=!CGkJ{SX80sSnT@CD_^Gm;g|jI5w}JS|JW{NM*WVhOtJhQSi3uJuK5LVv2-kfdf`N13CoBjv|QY~X9=KK zEd^9`>4hs++w_XnmX0M*i`8!EsSnfyt3FFGcNE(I8={t0hI>vZH{IiZJtkQ%Gj;3Gn$GW+iPEHz+b&ONJ z*~0~{5Q^}cbE{(ungd!X@)c{*c2+5AE8#2Ng7>U+l@=0MAk zh7P?F#5K#AM_LB+bif~|FXy>bR0J}2T824jWse3};18>qKQvI9{V86k^uzOqUg^im z@rS7N1GO#U59@?KxT=p$gU@#b5%YqUVU5l~KU38^jz4IQVfe#o%dQdth2`+d zui-I0bXY+(6ur}sBM|F!a}Ef^3;w*fP_xNt>v@ln^v#YgAc#=2(Vk*GgTQMWMAC1m zya5g8$kI^4(Qc`y5Ea;oX*NXMP{I)ZK2rZHGbh$3=T}!353cUL^Wfys#r@wpdGGPX z!w*j0zIgcH;{M6`-Sf+{%ald*Yuw$rq}+vkeEds3j$cwQOr9^rG3A9y9d&d!ous3V zsXP9LoZ(U914)z8EJ*L7^T*fKYhotaMhJ%smLuumz^Zs}?AqV>JRur|NiPL+Z*p$$ zO^NAE1jxtB$(*k5Tq!Rf?PXiXCpy9%f$EP3I@G3>GORGr2J?cJIk0EKTN1 zS9y#FB(t312TEq77UxAV0za8$OdOWXG7szr3pbhN9=-4K<4zcYscNhBSL)t_yBF`D zU48KG#bkVXy`D1%B(oUq64OUaoU3bqS)ThOSa8M;NM;G}Wv;GUZZgZnmS`|B6Bd?G z2PCtE1asY8pAv^iccHXx4rFQ329VAXcYV~qJ+fqY9_iA|d@0qe2oj z*FI<-Nz~@8618=U4X!E7#`f{ptNrFqr#1&Vwd!&e93muE3!@yV)#gU6#vEnKjH|Zz zX%Jlxsu>sq90R>i^>>30fqE@HH`D2{>mxLe)NAuty~Ycfs6=w^By4klMeF8W3zY_5;c##a2f!nn1A$xYXT-iWUU?GmEwt_R(K&oXt_X`}a8 z8Qe{g8cpvb^!v7t(A>)8!qbm(7Z>7Zu-#hfw|RAAY(JsJsVzkW@;v>~%G&8CWj!{O z>nAj~pAJJmA^jUI57HMrJbL`Cv*}bn$WV8lFPshFnu@wp^o@$1Q7=PJnjH5>N5esQ z{07OFJ-png^<>F{mjx1EH&vf@gQ=l*lDW_EOnxo6DJ;SFs&VP2Fnsz@k>R1ll4-a{ z^pxcjVwd(#A_aD_2x#a zH(>D%ZXp@70E)ZxRM##%)IBYM;?6w1sqPV+ThQ0H`m)7oRQEV{m$5BQ@d@Hij2?q5 zP!U~H-&p+O;>J|>Wr1!mB0AQXvbBd=YQ#WDU8~zz-PseWC zQSIYG*YoYB+kz@f3);gDGNZ*Q^#K+aEly=?ANBL%K1$sl7wvwBJpBRr*z`z|o>lwW zNBu(SoDUBFdhZvgN2u^*bd03Om++Hvdvk^_%+4g(8|pq; z_FB|oW<3TnYJ2kefxqfM+9Hdu&e6F?14|-t)fg+-y5mdrHqc$8{DBx zEl4+1bRps=V;Z+gel_I`R=IJg5IK?^hLCaO27;s6_?Sr5k^zXQT6!`SoFT~>S_fQ4 zP1q<0Q9mH^wZw83UrU}sJ31%-ERFm6=bLm%7dR;=ZrjS_^z~yy$y?t;a#BNXs>&H1rK*ygMQY_F$qyAbRb|W@lBzO~>xT$0 zRplPs6Bx=l{7I&&nC*%r53{27wS;oc=_fpPs^U@JWQE|Rs!TjRkwjFeXR4=3cB7?j>g%M0i%2-PFv7W(Zg-pdtNRcztk8(C7N=3o`NG{Vw?MB% z>eVm>{)@`s==?W~aioJ=7#$pIvu6Qt!x+O033VAJg8~CsSKrY2S%S*L669N&Jjw|P z7fv3whqn6Z00lR_p%0j-7JadDJh(|W8H9wJ7z%@ebduBdF$}8GOMAI=ekKA#9maM} z^s3S~jtoat>Gk)67ZU2LAU^+V7w7k1x#v-sElppXu39Fpj+C9iZh>~~pbH7LV$MF# zel+k;r(t8TobF_VRkwsS$jb<=oH@XBb6tad5C7NePVzFsh5Z%?WrR)f`MVaNn+rV1 z0KMl4XF=1ltXt0n<{Em?ktlljeM`}2r|E_Wz-2CB55JQ?k`o3>pI?KSp;rO#2=2bJ z;3fT1+h?}0&}{#L(V>X}%V@udfCW>^LiPqX6K#esze?rHPytnx4Sv;s7kXaZnsnsv zLMCd#4+x0{e>_s8k88bOkKkss8^*rtM>EAJ94f|<)f=BV#5CH8B_GNVk_4l3Sq5Cf#oz;Zz9 zJ;+-Q5@f^(+AU`}$iz@)DoRg>QZB{9mbn}xl!)V5dHYdI-LZFx>gVWUKz*2b{qg;? zm9D=53MljRBlZFtTtYJi)xgoAkh=n58EN2_Mg!LXh4c{sduwl->p{3Ujp_uKuGQR< zdTpK`QJ`r{&)#$tsf`ClS{>~raSE0s}k!igE z5HItT*};~W1?Gp?)==chu+_r~{L)1kTUk3UYWt?$2PtkZLT;62=0_{4J}6Z}b%x~j zO__H{l|mTPBQ6XE`zbGY=@)5axQ_k8R zR^z(+Fg~h!!!#)%^-4`m&oGky`sBEq+_1|D1gVX6X@(ytnUTbR#qaw3WR@A*VaY7> zz<#jMdJ#4ZNBz^{Atqg>TOqj@VaB*9nMuy0CzcooB(sG7GFMk(FT#xbGs!H?ABPE} z{MVxTicW2;o3mCXnQaf_khXwn%Zb|M^m-^TW2&=Yu7Y5#k>gDLH z*a--ikz8(R;n7*M6EF zBq5w>e$p->)R?b24w;}rg<*7v>?DLs+mr3*_Fq=abxyvT9D>N|tA>z?O3IhIp`Lwo zi@9DxsC^4@@^t21Qh%J8i8~45RH(SzNeHKLartcOn6kN8x9TZxLDY+IHtU1ssjD!b z0tjDyx-oiBmq7~-x^6O&7An_^aOqlb_@~qQ??sc%mA^auo**|K@w+ej+nA=K+<)?UEzTB?8gy!pG40{o-LOTMc;f~Sk63MwFIhH1k3CV57 z5_|h-lq9B2odonTLR2ka`m?YGLr`ug=aK_15(#qRd5yP7Wa5e4x}rZfODPna7-@D7y&A~92$L6K z^-@TlBELv%@fqZ>!wS{x(czNoTepfdd*bd==BG}6+ojFyg-Q2QJSb;@t6fLio1 zj)9Px-3H3EQ`H5X9ytl3KAFJiF;&{YBy@WE<-m(mr?xP2&x=;>-jfwooE9XAvzQqH zSc#5@$3^Rt58SpO+GW!*iYY%;_8#QL)K);+t1D(Rhl}*@5PbyiL7vSMQlz=;9&~{^ zNYqrN)o^B^|g1@OOqEU`r7i(iir|+5%jN08XqY+ zWKB*}Bwd*JC)d?43z})w9ES{>Bl%(RN&Z-NU^oKB{~4(bdq5)cCTmM_Ha?W#QT^Lk zy~OrFpT+Tgc_cY6>feykG!V|j!eN;ucj}W-^`7eTvfDp<^W9GHlrK9GjqpL~E<;%Q zIyARcmwahC+2W9Nmoxrg>28>GaY(w$m^n1vWggiN8-BXWJ$$dmuYTS-OhU{_7>A@g zgdx{0%rrNYoteahIVjyFRGGc*Zu#jh6K92!CTUcP*WHK?&!BXdRJrrr{%NJVaKZ8l z)&J4C^B;|D-MdBskhU2C;0yfy+)j#94;;5A2m~5c@zDvilQpg)5#P#)_*!I*^M-*A z8nvhnezme1wH3f;RnG*607)Jb7aCx#!Z=5oyp_@9vB%j0>8dq8F~LigHEHBfX#fwa zzZ<*=s6>9H%f9WSTOu$B8|N{kZfYdw zNcBvpSl$U6O-H=jckdoKQa!h%+Wd}>kh;0CuyK*#YF9B2yA+TjWheMsqwPcz0b{Y+ z<_~Od#Ig%sN2I_OY;jCyJybKHF9jo&H_23~6kYLe8EAzh8uGVmipV#lA3 zRMYwa-XMam(ax&&Wv$;`U(zkf@iP@X{%G7r{D94U%*yp$wHmxM>>1f(7Hi@kt`n6p zV6K9Lq5gGLz2>@TXixgc^|VjG9QkzVl-jUst*!pNDe z>U6ki;c%1wov^K{3?Sep%RXt+Qp>$Ca@OGUJ$kXMy)TTcsOa39t*#tH(ScwXN=>_$ zE`~rFxuz;u{{hehje*>Ox>gyQ{J29W%QQy=7{I6uO@66xP)EmH?+ZOM(Rs&N$;`rz zn9Rg$BNJoy3dMp765R}_$Yy#WY~d9gDA6l`W3Achq^qGrw2VoykiL%rE;xLZswBN9s8m5X!zqL`$SClsw~@HKhk= z)W>6WXMpW}YBI!+)%!2oVlG`l!6`?2nA#_x(mD~MC@>p!eSO*x&xe#R?H4qi3|nS0 zr{2RXGg4G#*fJB7m1wJsIXN@uk7Wm$bgRX%WhNmk1A=`!WSPmtoVYjJmlABH8@S9Q z#AS%*54Fq?%Td>;euhq7{?N&S){%h5JpO3#ijHnZy#oVx01}w=V93BPkpgCI6fiBC zAnL)`AF!jU)U|6_weDKofNEIi_$3T&Bw|?`5zBy~wG0Mbr~LsI7N_bWI|hTHj)JS{ z?ND6@T><16Oy_hRJ0fGA4j2se9PdLUDo$N*ETQW_<0j0nDlnL3%wVWTdoTF`#i{93 zV=$Oraq7}B7*m|;q5+);dKCZ+W|c4)SDZRqqmS`FW(tCL45ouZ7WfnXuc8lB?>h#g z$IdVqqOPE~&jlQ&14FD z(0s#3^Q-FR$!%6N!(S32H2RK%T7S_54#3xl5Y2&+%CWQ$)qjuG%Y%md_M=Dl-g@A9 z?I2ek2}fQ?DbjuunZ+aZ?XkK&$ooCH%OLNgdG~Ox9J=CU&drY1&`E}He4tpzG@>6e zP%*9f|JjPO9lGKqyk?Lq2dy}n$V*c-0K?)z){b+At~d#^8RY#>XT=G6nNCPA<9nBP zkzmF={%H96PVF&4jUyWHX(S z%?ue^Hn-8|5eB!R4uj`5Ks%EX2?n=`v@>0-ommk>+{Z{%wADH7Oh<)Nf!nNNZqs_N z=^T`5MVk$}!YFDgHt&^ctt_`OzeDs<#BJ6Iw{cZkb1LKV+@_$N0o9c#J@Kv8HJUQq zrq+qCaOQ&Gh8i}y>uNw5^iEXyH*n;p!o4tv!RWk|f0z6z#OyDFy`kD_NUHYi2EY4M zc!yKqGw4i+tq;XlL#+4o2EPqe*!qIT8MpptBlXSfiSpiqtBZ%17kBTSUtJ9AKt>MV z+D~HL$DX-%4GQe>D^b5!(iGL+AXO8^7oy6Un&wb?*RfB=>T5z{YFn@zDntt3zq!K@ zpo82%w09Z1sXi{X?XDRXvB7AAP^*l-nG;vYc_E<&IzH12S#PQzj0>ouW_e${m6yua zo>FMwUHfP0w62z1@BCeCrXuI~?2%pRRXe{Qs#@9ufIK%JVxCn4+!)PThceiu-v-J7 zK&f#KZ0z>Zg&aGUg3QRhh^2Lxh*Aau42j3G-joL@;?CVVuQu#f!MRHxF^6OevegG%cZE#JT=C<3;zLu_YQopXN53P4u;O9W^ z7p;SIrnl?-6~;Ky`*lX|#~NcxjD4c1`{xqeMWMMD0~@b79^43N{#@6uO{DoVPY0Sm zUcf}9xt$aKZBARv4PC$15k;PZ3q0y$4VUfnY%@jw^rLwrIWY&S7XP@*Oy{=j6-XL2Yq#&w z5?1aq(~bR>2s(bvLBVU?$-nEAl6plp#GT^_XF*d)>n2E^KWn|NE9K1JdQe41hcaE3(GIKY z1ykg6NpXmWU0-@r>AQJ6{U~VUyTYkTx6UJemNy>F-Vs(a0r|D=~%rgGK!W=93~_R zouzs14~a}_08%D_eD%Glemqj-v;)Ga%+0Z1%OP_qyAxpEZxDA=eKe*+TwrUTlym%b zBoo-_G7QvfQl89%2W(u`k4CEF%-kLv=Xh3r2BDVnGEIgel`_qd`oh^>OmMdQ-t%bk zw7PL?mne>9MtDohLd;B)Gd1GiQbs)aOp`HxTp<)8ro8E9{G?}20URD`encNp&W~NElHpjD=uIHw>zdY{TpmJe)<+?^N zuRp$jw%SB;Ve|AOcJ8LRB$$Fqe%lMyH<99ZV-&wF1?y;`c^FZDRqi}0Sl_s-ayJ0_ zU~5uv11jC67p(gcjikF9Bi&_;XrErArocLNEnsAT>n_X~^q1UckPUjTOS|5AW970x zx1SW|T9?o(usJfKI`xgNy|zy!8X5G?T}-NJs_>g8w06A$=fZXB^In~L=O8t;PTjm3 zAg-H)PrEwxIjYxFfB`b+4fw)=hVpkIj||=9 zzTEP|4!V|l?PD*Ho*qG&@{68uE!BK7Z6;iXBm~+Ts_z+@*0WYnE+DgMEn(G0uf?{8 z+Ixnr9N>TWEXq+={;`rsS*yOTU2hR7 z|M{q(;w}GnjjEQuNK2SaIRC02mdck~OtNR({5{1j4@%CrdC_#!BYS6C3CE`ATI8$C z&B^trKfIxI>k-`q^WoLW#e?5`=kCSBll!JO=GpqBY)l!UR(~#cx|OgzVW@p2m$mlK z#%k4QZ6BhxdaF>%S9WhYHL)*-d&ANM3Gk|fgR^K={oP3Ad(K+#rr z%Fc&szy^MXDq`tG^!FSQ)%q*VYmO`mGKyGMJ`Vn9umy-&(?yz&M4h^K6KPpDM$6iQ zL;*ah+u+r`+JE65m5sZib%S26YIj^t=;#J_eTr&R>2TBzu6`aE+}+b#2iJnpsz3bj zN0<5P0%rytZ5juQNY|&RSd)%Ot*lo+FLIrJIzXgNgX<#&_Jz=cvFD`vhw2Ml@NSUx zZem#4>a-m(T?loEaF_WmdL1G*mYAB~@ix@WjV|-mT_-;OYZvGDU%BUTk!usNKo!xM zNoLsT`V{S}LzneF+RJ+N^MI^(_w=J@$p=Jplxgvd6Z9@@Q9;1Va~)9DzLp1CLGSw1 zei;LT-bGVnzGsoL=@Md)a+?p_yVHc)`=K==jX+&PC(JKWy>EH4?lyucJfHKJZmLcu z&o*JYcKNO~O@qf^TIdSWxJ&Uc8J8&!-}#of?AL%j%5Z-W)WdRzJUoCfbNq4;PT#0Pi;hdp5aUfCO37jydD+6SlbWwZ-hm20b@^$C72WLucaXlbDW1!lU@K#!Cx+-0;2 znoRSCy9i}{+U=o6Phh4iK;2YBUr_yyqXC*z7#i>;;?c)!{~QaTsU&Zx@4zEu=&AS% zw4Zbk0IzM`_#@YedUo=qh9g%20bs+c^j^##igkYhSEJmNKi%1Xf!D7NO|qqvKi$4i zoZMueJM!yUjM1B+_OW(K8cgNyt+2qElJ0iT|gYTB=CaY@jWgjne z=Ksf$%8NxY)8x+mGOi$;{l7CVRP|wY_Rs(Ao;s3>H_#cI0sS@<14MyQec)qt8{maqg@thPXl)o%D7^DJQsr3LUO{9nc2 z>w#klIuHy?K)((57C^ugra)VOMYQ?|p1wmDsbh>M#6Rnp!tD4f{^&MW)5%HW(T%|l z(k(zJmQs2T7q~(w!fVd0o-HhoC3bKehCsJ7l;btq=@|o<-4X_Kwow^_)!&TNE3~MD zcY>Bd!>c}8pHrL3x*=NO^My!*?yG|$_Auq(k4LDXe$JO=%up+&z16et9t& zPp%!y_)Go8SmiH@PFMr9v=18?&N>o_p0JF|_(NpX8!6rBam|<d9Vpi0TjG7AmI+}(Yj@<0h#DYvlEaW358TUhUBdQ z34368Rd@~9;yYx2S`|_-yl?5lG*U23je>!@qy~pzb1)ntg=GxWNX9U2l`+gv|9jhf zx15Fke0uw?{*t*rT{;=Vv{lA16_PRd0gYq~=IM+9W%CDpmSXS+(A}99I>K6pX{2SC z###pI%Xuyp6@g5hmSIXgQwRPqi}?en|EPIs#~+rU^dqgkGWUMQTjN4E2lblzynWt2 z;+>XZX8D8p9pdDP_`^Km53cHCc}jeqy&QBsEod32@{e~Mf6yGm@Q2fuT_pq%XlEq= zPW~Dm(?f?9R724_4LJgVe;j|opZ6AOHaTrQ?=d#&riWcX5TRzHJ;izkf!8*O>^(Su zBDOS?aI{DPPy@pbi@n2EL#!r_ADNP0MqCC8zuxz3GUbDglY)b}Rm_TH3ukC<2nV@cYWAU7R+ z>D`?@;XYC(ws(X2$ymL>4a3`y9^HHEfmhLSy?7)y53mwgJMM|54Vu%ARR}HW*&-lqKgGCOX%!=K6o?tS|8GfK-wmtrIl3B*YVaY7>z<#iB zlUeT3)h<81bR#(Xa@~IT$t;Gu2Tf)fZu@B`vjq4uSJy2!nPp-NMZa>Xs&q1`F6{%7 zSwe!j?)K?LbY9g>_h48^miBFXHRyU8>C&b~m-ak&e`XK~VMpZjPM#6Zgd}RVeb9Oq zNz}~KnJT>~v}~pi`?q%#HF)NUY{Zc^0YF%;UMEY?*P@7C%v7 zA0d5PBnKSNSwSVFUUPkfW|4Yr7OU5IArqBI&YXm822ga>&C2%?(!PZ{w&kqRX?yr9 zd_CyWR{S}bJa;g;>3Yx`aUY>YpC~Wg^`Oh9anQ5v(v4GU}~tHWbShulV75pqe_GW_(X2oSU~rNSv$j5x=o%dv1A(V5WQYL6L517v-rDV z&8%~+#|wcBTeB4V5fLwbt^b6Oa`2x=s$sPNCPtt%A(=^op!f?Z_Sw}$MPEE9<~8;G zaSN>j_Vy)&Jg;TD5Z4Dgm7yMsxumcj)`sZuql+`W{=R+h!CRB@7XxusI-+jTk2lc6 zJvkxI!yy67c?k1`X{gIuBrb17*dcaf+9C^LJbHc0Z<(h;C*? zbgVIDYY%l55JP)Nb-#}uv^#ry?c>Z!*LL06*E7?dJ=8veGc7u%J~pD-$JyGdQ)kpZ zJ3)K6h?x<3eHYNRX)o<@5vYBn2Uns=4BEq9?c>7XBI-qbrYsYXk7YuNTYl-D7zLq?Nc63hu zSsM3gf(nR>&EAsVEu@9|w+Ovzlgw8E} zYRFAhIisUgCAGQ?lNoYTRmQv_sVeiheu(f=RqnxifuYmrh^;0W@nAh4vQyQ;*ls$N z4wJpusftH=6SH`!DicpnBvqBbFKN|gr79lkP24P>st(4l%PdMRO4 zb8fB!v#;XKt$QXo1eL)hoVR#?ZW!Z82RAo5xB)Y$#iUCWkf|=FWl-Rz;qbEr$ivd} zEgc0rA)yxP!I08V2PnAd4Sm2wwdix*D}Em^O=`FxBwWT&7|_Og3_CKs5~@mf7ZR@Y z%5q!xVQ@mir6t4WcUofTO^AnA2_AN(=_~X?!d{7mnQNCuNH|Bkb|iwq;#|{;Ir}^t zDgNm+Yz&rDu8eSH<;(!4>lO$0Jz5;@WQ243EfB~ESKtnUdSvOwn|pw+KYow_de0Ni zf@OUH^){N;Gl4OO9&{v%9)90Z^b%~XIw0Lh0VvHy>EZnhY7J(FUIn}(xV@NT*@>MW zO1ONG-(sUf69bmfo~NUi$`Uk_a%bOsO{hM!ze?rHPytnx4Sv;s$a!Afnsnq3IVSqR z4+x0{e>_s8k88bOkKkrpD}tpw(6|gSmxBZu zF@ko>Sq?HW^aPfJ5-h?7+4KNnEC&fC;&@hiIUu$moCEz_)Oxe@6*)Z{8-k?tVdnMS z;V9D2nWrByJZj01$J6Mb`4c?Qz@=|VheGZu#d)NGn;Q*W3lySC_?+Bp71asO^*LXR zIWlZ{r40y8FYG&`Dk8t2fD08@5%XD5JL-Ad11?l#E*3@==!JkQTgB*b&&uH*bD<)2 z@&k6U?vtLCyHIh}Ailgb>)66F?!FL^dWq$`j$`OW#BhvUFZ)CY=+7!Z2DkvF<>)f_ zU3jt&{p-Hmvb}ZqVtxFgKkO~ld@^l5>@hau>82o{52g2afm?(cPe!J70>ITwY~!6h zfa8-BDLag94Mm;|TOF*>GadOsAwRHoT-5eW+w=R6FSxx3kJJ~=_PzPfcD>CWO`cXa zZap;_KRsqQ_{h9ai!{)bY>}GZWqv1B3Y|o%3T`!S=BF|h@$$V0Gryaf*wVg<^uR=0 zAw3)+gGmlZBJ`_6Zbfb1l<`}{%Ovec4!sAH15$etW~}cMG*Y>F<~g+XO__g~&I8c} zE1$R+uszT`a{H#7wLPrHb@yRrUezt+#=5EL8Aj4ypB#64GRdq-JUAekO0#k5y0nH{R}V1_x9 z8|&tVmeW`_^Bit6%Y^M=qDn5S4M=7Q6lNx~?O}XWPiBrME`ZjK34L~i*mtR0L}EMh zcv_dbG!*-oxWOn?FGpv^PC&SbE9Mc|3>OKnU1nW$5qGpi4yf9)Fh0-kDCU>`w$g|(IK*v5H4&_wtY4E!d&MR zAac_i`jClA$`^A>k{65mAu@P3NC?+4P6k}NI#@^&*6C;7EMO z?l%aZR-p0KS%6XFs3o7iC*7IK+~;^Kzx-?x4)84k_9E1`oewnJGx`-F#4(C5d7#N| zp~-E=_Mr!rXhl&WR|T2;gpjO|+GZ^I{Ror_>j>qX)TgfGHe-pseKgvz#)8106`-X* zp2SnSQUxT*RwpOFTUbZY@3PchgvpPw%A}L0_^c@C1=L<@FT(ztX7GSRo>K8IB{t+P z5;1`4&qdwC!Ct-e`CFgc-zF8Ew=u70P?6YpZ`q4PA~|F&62xAFgm`))iv;U4ICqgq zC=<`a@{0sfez-t2d-U2=*SBsFY4*(1X=RK3;OUbkZ0`9Y%mq;1MKhPKr8MM8+qs-cbtnBAkZK zb)+iF@Y4={086>0kKv(=y?eS}xEA&1n}#}3R^Ii@0~ji`*DY02qT`t2h1lN4g3;jo z{q?~Pb?8~2{*n(Oxs5Ah$kl(1)NfQIvwJDw$zA=Woh?ZO`-AHmU6l2SU-v@XJL;v$ z3lt%4`TZ&iMNM05}50 z{~4(bdq5)cMnc;CLOxjsBjp! z)qASTt7#w5@!fu?xmcYIgVJ4w*p)-DS)i zn(i`>?1v3M-Q^y>8dT#3kPn&r$U?|L=`Nwl>~(j`Pj{I( zdt&LXM40QV{g8B*pk}_?4>cF-4scSfC940UbLT&DJrS3Y0MI=C=mq|Mu2qWT%|cat z?XpI5v1=gEOC#cIku?IWQ6CzoQCmlK@JrX`uKChe{oZHmB`V3Iv*<9+ktWYPoV zn>s9dQVojb)7=}LR6Akgx>qc}airQ5%U7qDUvi^MmUWf;=YQ?u{QfKVJeqSA^Xn!} zs5bg?H{U=zk<@6qu<;KZ19$C2{L_zW`m)@4ZMUBKFBMg1DR<5C+Dfd0v~i*PwAy#^ zF0FgtOB*S;DWI8@%fI z=5uN@Su^=&f%H~KI^DAYou5X*zURO+bW~vPdmQdxYG+>>`pt1o=M}n-F zOUIVGbYx-+pBu^d`Gc%0g$zAl=}0IQ*VAElFBzifg6WUjKu4G!qUdl+AENfB;7CWf zG&;gsDQ0PT+uzvXPf|2DQf9~lMx!azP4qI-zpJ^+1i;KSWfEHz57^$84)SK@;ymD$ma4^T7Q@8YWSWqsl{2+LsJufjrc2g{{bGTLs!3AWoYu_4xLQXy?R*{z z*b|c#YOGv_F7Xn&&GbUp#NR!U|4}miL}w8#N^|#x--D`J~bKQ z$Ljr;Z83K&?e}&{lFspP&{A3_LKFpNqpq({8{+wpa`~`jCLu0EJb$QVhFFfeLiICr^74mH z7PN{4H0JR~gIDz6BXsHX$7fUm(~e(E(|&M#HVT*){6amX`2%)TmAY~*t8_zu%zE!5 zb3WE@d$wMo8WuW!2}2u+Sj^*x46S7_w4)>_PSwryI1Gk53f_CDu7kl~oUUU(Hs+d` zz+kBC#Dz!~6{oIrYcBg#9|dTvg!?hHGa-ZNT8mRV;o{UTE>6`g{`L&U{0?f8H#}xA zvxLF8;?%Ce=a;+b1cfXB^UR3v^}b^;n&lVi^3IH`<^XJ{KsOKgd_IWoCYn0wm1|Aa%atGvS&tx{~z6RJOyZvNwnaK7~syo1T zCt}engRFg=eH-R}^)Dm!Mr%aBb$)qS2Q*R10h)uq;Y0gX_44F4E41M|VJyeoaX|d2 z71j{=@EW2KH%64v{P$SBJZQLky`cxW@<=%H(m)T4QA7FNxyBGt^QRW(yN;W%#L zRWn^w#kF#(nN_Q5W-X+e@ndT`S!q0AWi#gSgT|K4ZS={ry?hJy_9sP=ZexXrj@r46 zX_p1O7i(vzN(j$wqN1&})6S3^Wd-fbEao<#UaRJ%oy4mXD%EmL#k#muYcX@&hIg;& zV&3OAovYI7TJJSA)sO?i8 zccRL_fg?9~H438{SN`YnbOyIq?=J(o3Ds6ZQnh0@9bSPXyy0fSo5=c5d^N;+PjBXo z;YhcH0CovGY^#1YQs2y;DDOSEx_EebarfT&)y1$5WLVQNa@^Nm2WvR?%(ZLB047~m z-z#Z~YQK`IiQ)@U!LN);T~g>>=W1_LxHDCKGFD#`8dKYXUU;EuccBqcES+(=~q1rF@m0CPEA7Y-7jApGv8SK(;1LXjq z)PzAhhxOAD+BuAys&Yn0UBl!|O-m(`Mwq2s$WB!m^M<6V%;WkY!b?@T2lu*!$qmGl zOjR*dK49)iiLlxBWO5JO;?-M}@l#a-lv!(N>8^>&0-lMdYEzKi*5L7@i!=RQ^zC~O z-kOZR2tI4cg9Eh7+;3FC^UBCU}FG8tG^pUN=WnPx_+%A&EGoK{58Fy516Pl zckP6K>xLT}Zz|vQOZzrb|1~{Dhut0}gX>Jw@M`b+)!F(|SB5!vD7oo6)0=VEuSLR> z-F2qjx@F>Oqc4pZaE;#LL+|>f;j(?6U3~o0kLHc!!~kZ$DAaB89lttT_6j78x|y7P zkNR``Px_8uYx^w`bo}b-mzjbE?RTA0Qm@EHDarAKv!L_uS2x9UG70=Sbb!j=O5XwU&O^s3zUl>!&#POLt}J-WL?igY%Eemqj-v;zV;;^tDXo-psebS|1EBkUV%j?Sx(#&n1aZ0(bB zj=zp%0z1u^04zuiLl!Hb&Q?Dfsg5&qd+Jz@XVDiEpI*#Nlc7j`+$o!Bwx3$2$(b5u zn&>NA>HIDe%iGciTVAHgn8wdEnTPd5f|F@-kL~eqDc?SkOcP`8D0Z;Cmiss>Wp_PI zF*8j9x*6-?mXm2RL4z-=N!got8Y*a~nVBYG+Z@kIClgcyeKWv1R4z=fT-ONZ^&QfQ z23n5i{EnHh?-(z;07bzWsIn$*ECaLow}~SWayn~^42g7I&iN`lm<#$i1Z|C zMA6Fi3N$iy;52isOQ=(Cjf_oHr`}C%ow^od(%c#J&WngtZyG{t2dOv0b?Td5o%+H; zYV$ijuu&s*)C1zWO!%~`Q{T`U7u@uhD0lYDTwm`DE?vN!uika^Re$|4^mW>FfB`b+ z4t(K2fAPD3N6texxi7c;u!F8842iu!dU^yx*B3qETB`YE+Dy0%N$|9F9;)vdnbxyb zw4@~Sl^vKPv)L=?n`2u;?LEU*hgkJYZ$EdV11jTzt@>f<=3uLbA9Pvr9Upa_T1^(^ zs4M?iNr(5686>s^sYgARjyflx|iRuqa<;yK5*)wkbp5m4VCFk3` zXu9ck-`q^WoLW#e?5`=kCSBll!JO=2>@AHYP%-67QXy zOLgQ#vaS8Q!cPsTY|H-jRBERFY^+v|*7hN4tG5cJd}a5hQ!`yH!KqbZ0P+iXJ$U|Z zr1Cv#5bJv6wAGzq?7E#5BZ6A_*UDT=llC zNhA|12Tmdd?1^c5><*+oi6l_8)t&XX5}h%DpP`CaIuZRnOGHgkH$Xu~5zETQ4gLtW z05K~QYZ-2;oVwR2ZGkOD%i7Yz6Tp+|0%B0w9y&!GAv$+OYuEHJr|#`gZ7Lm(`oWE~ zspjdegIkyN+MTj1c2`i=tIOFLbacszB~+|QZwcXKy+)x5Vo5(8AkwBc^pO(vh3Hm} zlY%Qz>R}gTy~`MuqSEyK7&%$*26TvEPC%pm)f3uIhF*t=r6ZSNQ5 zYZI{n9KHTfC+qFdzB+VS?;|VgHQL%;q+FUWOy=oF&yo*_J3(*t-F+(Yg5HfSDgby{ z2b6sS|JUmi$qRZr`(+FWdN)mL*n18s>ovz9%+IGDV_8)FRL`4(7leK2D~LpzX0Lp@2J zi)`bWFeBX`2Pm&JVSTY4m{59`)QP(@|4Tk1S86et5hD^j8f!TkBQoTZ$_@3mBSo0y z;CCbJH{6QkwAI&4-k=*}#!dsnhZlFx9-LoZOvaOI>QBcMHU;eKgT6NJ91V?+H2whO z8}9D>tUx2=FGnhWje3MApa*+wug!zl$JiS_ZyXPHMe}6N9HylvvsB;D*2x^>7ulIN zc!bhA?&isy0GFmv+_p{TOne-4GM7+`=wwb9OPdD;f&eqeF9#k+Y*2+39j)ktfj$qM zt&K2dcQCMj(gy>^V(8x!UEG2ZsJ^OQW@bb%zzCzSYGEoX^sT*ZzQN`us(szK=80M$ z(>RvH6;lScC_k2wVrpX)Q%z%87np_oq@ch|Z;2TQ1r+&WehX29who}v6DKg+M4baR zrgK0Lm{DKONe?&CWwhq>zn@Ul=8i#Nwu+ev^==K1ON1|@-Dq-f`|jN%R|UOtxP*Bb z?K&YPnhESFB`A?a148@!a<_I_EZi5EQC;>sjt1zlGc@2!#G?<8mT9_!y=quHz5~z7 zp{L?6(0(C_OEmf*gx?enfUiA_AoZ3v*4G{^SkMRTg2YJ)&07dL! z%E2FxP($X#A!K5mAT(I|7h{#bC_0&f(9%9^U^weYBzn>i1+oJne>1cPy3ymBF*$ky z1cq19Y!-5mjAwXjK<4b|)&M%56EOhA6TCGbrVayGfC=oLCg@nm=>GblSZhGS9onOH z%UA<4(R27UfRj8;v)A@DAYl&-uL@6d8hz+afr5d4s7R=UdVfzJ^z*>S|J~D%+TCfp zztfJZ_7K_W;}LShV>5ox&x05XR3G)Tt} zen6p&p*uA)hV}tvYZ>&HOa_0T9?iYaWVH<4DbzCPKSlhZIWS-mh#mkJ80}hyF6Ixk zJ}f&>xsOm#`jJLf!WoKvm^^>zm40-NKSZS;m7oW3VAJCdOXDm1=EO;4wG7>9LCZkp z0q;2epgD%&52r1=$^anrdk|mzui-I0bXY+(6ur}sBM=(~CP;HRfEF+KBjQ5MCa0}O zM6;s{2qM&Mw5M3BZ_sgpzOTrBjRhehOG61qyQNm?usp?YbrrFcwafbVk@{bmIk7o8 zztWw=ukO9`;N;Q8{oguy@A1XM4^G~`c=+Jr{>k~>^UJf#ltuJw+}*hZyV&!Wd{Dlm zUYI;z3QEcgl{)I^ZaPUv9i{7zzaeLM)c8Qsq;!3!chULd>*_Tz6Kx}e!v)Kc^l)Gm zOgDDzZ^D``Ib=y69;I&%&TZ8wF�n=zNb9qFL)QM_@@i`{&^Nb!#`&PsZv6ZW!(< zx2N4)FCNLwgEV6SVkYU{JT|M2y84e%_GVQ|u=N!?Tz4O4Ue!(KC$kI|DJ3(xF)nuR z$t5hb1sp$_eJFl1%a}MUnPnc>4;F4R%RPGE<;OD^$@oiBRXt}8 zNMr*&Nz`onpzah()byV;=xZr;130N} zgKY}4Jhgp1c4|NEN5=y}ja2<%-XTI_wJ^$1t=63ywHkAj`O`cDx(ZhAu>GO$Ea4~h#JfTHW(4Sj?*F-KlCELB0w zCVM^T^3+!Rxx%=OgURN4(5i8MEs3rNU7jXH*s?bgx;4}D|bxoK&_^tLWV3P@h^Mmxl!xMk_A1LKJ5lmL+vDUpW~T4+8`~ErJ`+e=aT}uj|~qcmQ2GvqQ{FZ z<;*JVb?Z@Xn02o8cp;EsYnEa^BH{(E^`8(@4*v5MLY{{Y z;TX&X>_sAj2i*6&h{<3-E;jhMw2-?Lu#JqLfGUTVv{i#gDURYg@@BfaW^%Jy9TJFj}Cs|(AT%Ua+c;0Y0h!ZlkPue#OF8+GDSsn zbm5^N(?~>Tp3WQ7K1V{G0{STFQzQc+Y4>A4;C&jR2Xtp2LXT`8rEY4aYd^Q27^XYB zuYDxHoQn-}BnPoEcdUrygcyuyv5E zyz7|PoWUwL4izFtvcteW z4=0|`eR0&i2X`;tKfC(iUESjP^m?5a`WdZ0CQ`NJXUkNL1i+=xTJm1U30lj`Nz@OB zd@cC}7GF!ALOVJq|16FBbio!l9VfqA$kElYHmf%2oR|D!UvfywYspg@=7QXmWT&bO z8R_fC+!h0xT$0RplPs6Bx=3 z1S4m$Q&r4%1^H}$u}S%CLHHiy+KPJ>CPa*j@Yos@7I z3FoFpIM+gbr-bl#$j#kExHP00;JuZ@d#&Q8J!%9pIOZ7d`sar+j&yKSql06OaUTHJ z#~26+HAUl4V1TgdKj&u&kcX8H&FN9MEF_#ddDwn@^wR+f<^@bti$3jKzyOd{H`faZ zwQnORj9y6CE1fg^d{2gjs?zO+giFKEAsL=JGHj~StH$}2Bo-3R`gnNiO4FwRrB|PB zjJ`DAyJnzkM;Z}yA)!{x+2=tap?NwD8-p~%l@S`;ZTt;jy7rBIEAP&_=e>+@X1@gj z86o7@btmh_n|pw+S4T2H?|H&m(6kQd)-!=IhaPk!iXML7QFOe>oeS2(?+iJ#2Qx#j z0^SkaeNe)!l{#CL=z2Y?B%>tiMf+?s5jt+$`ohgw9Ze}!aEl`Lm;q!8@ zs7`R^TFuQGI>qfr0BG8Vpdde@k)~~CG;NF#?Qz>nbGFDX~@5)EVjJT&aj0AKeHwAIY9_Gipz7UYw`0-uGG4vv0I7Y6QeIf+( zXB8j=oKdep`7S)S5ZrR>Ce1V^BU-SpTrJ7Hs%?H5(yub!r>HS?e!iO49My7QF z!0AdTnU=T&W>2~S3YDU*)?xD7 zfMk~NU*_sc>_wPyeqCm;fVCFY+FT$L)GRbUvn0ZxCW*&{9T01%+_Fd{`k=SlV z?oyZf*kai#VkaP+MRK{Bk;}EPV!(K-#smRjR69O%jpb%t!&)62 z0_1P0L+AzMVU#2N+sx?SNWBQtQSPJTx`&{IP-|Qm{J3dA+)SCH!mvatSkmA6UWD`L zIw$jV0Fj&C(DfpmMJ44kT_(NzWH1L6;dSo@3E@1#$+M=WZQ5x&GCLD0E_V__qd>b2 zip%GgDVyK%>p*>m#1g_q0+wBc`K(D~rj6dwqQt%Hj}Sf zJmh*2nooYl;{e3#PZZY9x~Kh5dJ*B=enSKz!WsCWQ;$&Ji%@^MGVq-w^h$i;Y?!s! z0DBQGL!UZQFGBbRj>Ioe3d`*&2!E+ShU;F2y$F|~?*MaK>9PEV;O=DG80h)ou}IX5 za247SFb#K%UN4{iEB|L;>zj!?a+|SzctIsvQB=s)NhUubBrBx08B2aYI$O$Fl^i1n z2#0cTl-g!2vA2`fA57_ z?qe?!8LH)wr_^*pC4`W6;(|sqq^xj{<%W?srxspPT^0KAL6Ihs8=OH)MP<#E68P)93;gaiHH;*)Xb8_Fhe(@(WQ7^((=v+stq6|On;Kwp^2n~b6t)dLS z!`SLaz4=z5PL!2*9rMr{Z_KNx7o-w)LB$x`+gLCfJiWg@*rSfcr@Gt1{HgwHq<*6! znSJN+)w_?cPVVY2?QBUR*!-i?isP?)vF#o8(&Pn-*tYzB74@RhMbN)0X?&#Q04A4v zkI0&v=Fwem!aupLep%2=tL8Xl*c{0Zi*RHybcuc`SN}6o8}@)inl+Y7H?lN(PB^u#Hj+qJvrMm<*^WE+@e!_04=DoluQT-pCJO7dEi8zl0fO8`Ne1X58 z+evY%)_6AtvrrXZyR2~@iTLJ5#MdHg1UFbzjq$R^SyTrOSn_k{&2$)n2} z{WvcoO`dr?f1E9luDT|SGZZ%J5?=?H@p$9A-38mtC`VdRU7L|E;in9PZ!t@I^9Zu;()czD4=?Kl!tEHHQFtV~p zf7--CD42ul0qIc6$s!k#9&lmwfGtpJ6-E|)eJ6`FRhUL0=6vb?gDS)F5Czru(NUBi z)kxpBF#5h)fo)+_Er+6wD!imw9}5@=iaG$^acBWLojy5Xh! zX|5sN4wTIs`eE+Ti)HP7VIjt1F z!fxa&rdk$B6a{9Zu1`^hcphIV{MJ#}X#w2rIs)cyKFS0mx0Sn zLRba_`*z4OlZm+}xXf_!ypf{~SY{I9GQ{&?FD*SmvFQ*{1*)H+lb1hqvYY-k@XHcNV4@E-bWuW7mAZ@+Fy`q^_yuSw&>`^$?5HYr;aXNLLAMn2IgTPMQ4I?n zzxbim7ieDBXZ@p}+a37Lhgnu?DBTuS{SzP* z>(ISv!8zhM7SrL|hidifE+Wr2c!XULG{uw;w&a_tt~Sjs5>qza-^dc_bWp>0ASt zM%tD6brfCS)wjp$_8{+@JT`gyHC<7Db?@qfVcz%KMES{xO?@?R#mStT-BwC?h^1i7 zz!fK^HHlDd7y}k}Ssz*tU2zg#Gi34J4q9 z)MhE9m+`&Jmyuw`JpO3-`cCswFiKc8vy4MRdFrdot4X0)yxv8 zX6V2w!8t07Z6up9j~_I)Y;L2!jWf6nbr?Li5!TKuBkjyG*3M90&c{epw6%2FnI#oY z1#Z*D+=lu9ubmMp)$-h?SE|)HZo|CSbdzu!?W?_ZrorcXZd1_CfFdMxm*-onYp`v& z%`M)(7lIoqGSXdF1InOxqDnpq?@oP=+-&e=89U`OI&bCQC4b~FC)!|dsJ0rCsy({_ zGB@F^B8SwZTb*mM^`ZD`i1nV{h~I`P?8B}4*+_jeJG1URxVm_Fd2#pN`PId+4rJu; zt^Fj{RP32+*Wf`0ZFQ!;SJD*K-XK*I#TTN=nVM34`gN$EjMdkK#?-c8IaG)gzJGIv zAshkZ2BN*~%1!lgscm=7u!s#t-%x9%gg4F7)>m4{d7*zQ&$oZMseUjnz|orJeeqUa zDqDL>q04{mpQY0}O%#9x%DFWw^Q_j)-ZZOrem_*Tvb0F)Y8v^&@nNmV(cqpo3crhXc!Dr4S|RF!#LKSX$`D)-=hmoPd1Gcr_|RNPb* zL*>s-s>;X#b$sf?`0hp2sgs|o5}?dlLrZr}R2IxkJUx+ARpQFkZ3#P7CFGdrW*_Nd zIcrTuV$eW<+*CwLKi`KB) z0a@M5Ohf|cx_+5fX~x8--+b@k)!9@}ihXx6^E*C52umOA`laXg=YQ?u{QfKVJP~u( znJ%XFXZfEs>*)?Euh3h3=v}`wT(-}%k>a0zG;bs)=0MeAio49zG#oSj4wA-|u2Z$| z(egmo@k{r~+C@F+__e6tD0(wE`FE{S(yYjaxN|(=EC7Pm>iT24BM1H*IzZ*`@+fYQ zENIt@IP+AUZlr*El6(h9@%_wBD3}@Qa#{Nt&(}9%`XgYB z?mYAK^@(YdKDssj@Fo?dkJMj{3O8Kfznl`c7^IiJz)KmP?pzbRu6{aJuZoPKB@>4U zi9%;-p8Job{F^cfLxPiOa*yrZk|DEBY>(Sgj^bsS7<*?vxk_2n&lcJ| zOR+zH>vQI+sfMh^%rpt;W~_%>PNvBO%@fEph0Q#=pffW~!nQe{^#^yMWPs zpV8NMs@6qRu)cCv<*tC_4tyNnUF^xEmY@o5Np1!{W5 z0M}&ad#8bWT}r!#G^UdJb#ocXei~T0?7ql>(ot0P5`MjF%7>%v<>zO)L&Y9e0t&P)E8*|05Dbi zd=G8ym$_cAz!j~%;!X9gqpzC78~QqJI>4Bdi({{-(ipt?cOj1q-Q>R9^1}|gPLE&o zNNTC(lW8-ODo|T&1xq&t0oa=Tou{o`sJ>@pTE|)!;0gE>W7c^`CH@*N+&6?&Hd%WZnI=skLV_t z53f!x9{lDzcP}2E+&8^3&o(DzVB_5(uwHrSrnvu#Q`E}o_@s2$CyZ_zhU&Y05NN6p-&g-1eGeM?$w>(pe;tr+JZy@ zJgE+hzt?vWb%a>ED_Yl05AzN|wW)MC>IXN{rml@PwQ+FkvflpS0$HynUK~2ga~&$y zOi*a?)c7lIUf52Mdqld{PX~w;m6P@z;$^*2Ux@Y0a?QSPGg$)PvU( zQ_vwIooMvVo?h15#l&>#h$;0l-@WftBVUFuP=AikvqiEb4XdQ zIR+`W`M|w9XxLg0V=V^^)HQU%{36x+jwjF3Jw;%<$Ip38H&rK-XPYoxfos-H(`UwD zTIdSWxJ&Ucc_px2De>c~F_?D^VGV|cb|8(0dXl^rFtqe-ewd+}u)bIiOenoe1iJ3b z|B{c$m0C<@#E1lsCifr3K7`&-e>+lySq^?T!hXZ8Ncde}wsHeXi%~J=4HKljjt1&a z#}qaN?CXQRHt!q_jgP`=gM8rGnf>_D#hH$wx9>f8Ycl>K;DUokNd4tV<*!kX5C!yL zkL|U25DizK^Nxqx)jF9ohiR$FEYPWX{9~9?N>Q zI@q)=WBgSGF+<>j)!3b1e zm9UfIIoQd}){(wyZS++wOr_xx%$8?QlH$iwpKSG@ z+G|nPMlsbimJI^48}K3&U0uvUD4!@?U+H?*G0<-3ru*JeQ zCOQ`r;CPA}83bmth?y(_XhI+0aLH2mGFo$83>28nEG{v>L!S140L=TOWa%!W)tAO# z7WyAkPGF`*OuKmR3(Tn3k^GLM0UDzj8t^6J(FaJ&I#8*Oy}bGkJVJ(^ioZbnNe2P& zdp|b*$aSKgoqVa`$i>){n}bz26o(q+ru^y7{`yL(WU~P1#?Xyhi3*@27TFhylbh^w zM}9qv4{2XvGlWhjQ>U80x56HK!UYqg%VS^h!FNk_lT|hNvX2)!^Z(;W<;9|yY0`1<2_2P6&Yxs%QibbFi~L+Bz&gL(Q9sy;0*qUesr`zjcMN)g&Q zLKjI9IwM7B;Rv7_0v!ielw>Y)3>=|zJ+wONImjt8Iv^{6GT;(7F|CUCO-`P*MfQA`k)4}Rj@N9bXAJnos50C1 z_T!fNn~{2j7M1W$(6l zA>@ocSx9U` zwS5gp*aO3>{uQJggDnQ(&<+#~^v}2nl~C_H`Vc7?I-_7{cjx7f;Rc7$ED@G5bdijq zYn3s8jXe7L9zR4?A)OYi?~j?SGKLPw80Z6<$Q#0dMluHTbjE?#3( z*qjo_=&!*seG@vYpc;zaX~+?XB;%k)y0FWd+$6rd316Aw@%)BeDUyu zleaG(KDfAla(?&x^6WB&5&a@}cQE;7AIP`Wi<7UE0-5xyrM@}3XHHw+od0{Ienrsq zsP}=W!D%L>chdR4uB%^?GtzcLIAFvai4TcSz{jG4($se6#;)y-`hWB7rb%I6hGR%= z)BbmJgiO>I2`AGL8#Uvx_WwWQY#kNJ{ys9xb&4cC=tyur$c9uNl#zzFR8U|pt^Nvo z{`Ytq+vNwEf>qB9$!-}?@&vM5v7yhC%x*c;50%|Ytzry)h4aja8BL`-;%oF<& z!_IEGXYafJIG8NC6gBn(8lK%^?3eI-vz>ScK&2<>k!~z=C zs$v7gscwVo8)I%^4N|s^ld^5<*X;pr=p6#|Z0V6O0mnsgRv6|;(6%vxHtsO@xpLh_ zu-A|-{9(Z5rV;T@1PZspt|FrYOvtlxy@ocC!fg{P+?w9d#f-UC5yXs>xorThu6sB1 z8d^rI8Fq-TKg9v(mGBj!=HglioR=0jo8M_!DxzLPtAskcD@1i`^{TD@XUzw>OH%^8 zLGQ1i_Zs>?Y_Fk>mC}W$XXT@nGpsq1wOHet4wS(b=QN-#VL`%Pc{5@lPl*Er zNFMZLAVH89%SIr{)Fn2hh!P`ypVl*dRMJmt;<)d zsu;pBOn}EMlFC+p0Fi8Wa}CyH7AL`A^+VW50}kpjo6O=<5m0%WSqU>T1bEZ8cvPvb7DwpMWFe47gWTJ>0ItF0u)Lp zKaoTwbeXj4v)g*ze$(hjC9as+wqxM|(6kp~Ecr(#w_elUc52$oIXXEo$u})(gLiat zYu?n~$5&c-A4%<2B9 z-|4R7>0-(t4P?%itdm6ubjR|c9atu3G z=bn!%vy-FBs;FsHf)+*_LDZjfp|?N@9`u$3GOO0p0WHS?YF{jW@b7~n`rW|ZwP zgD2dzPz?|mea&t@N@Uv2oj^cCU0Brdp<<_G*ipAVnL6rZN-pb3)VP-bFwWA4MUA7K zW_;9m?n&I9m`<}jzN<$EsBspB)?(n%cdZj+F#rxto1g1R9DV$FOouJ&rzX4GgUFQ* z!JGFFhs7cI(SXR_A^0?xUQgmCK$3OS-AsNa(|b4>ToMv4Hhxc{?!BAGn;ohhory35 zhq~)XG;Sfylzj$qn0uo^N20z`VJ5`tn=#=04C3}=-Z_LFiQDw|cB<*7jXl_|O@R-# zFGOKj$TDj1lHREIe6GKIadGGM`{-?nArS5eFo^pm%#3Ov?rtJ+N~FtL1G|rUZ^SUl z1EJmg^8s|TmcUL}wFI2P*pnUrk^NN7JGR};2{N`V_D9A z4IDM%ZOVS^(Lj1;gKlgI^dYuxk)qsbk@=1yP1%&~x%%FnCzS+T(@QwU&lpxp4Jni( zCQF}Ncc4_lt-&(%dXIHZR^paOu%fPpe}~lW?V!hgLn1P3PPd>2LdXV2#l9gCAtJO8X(5-_Wjhdn(l7b`_&c7iM<9E?xC>A-x^sZ!)mP9uMLRtDMQ|(n$biH1~g6M{^ zepS(}tlzL8x*<#!MmJ={rNuCBNDS8^Hulv;H&Fv(Np$mpVWji0>Jqz}=vflo0;pK( zYPmNgvMkKikX6NxZV)j-L{LH~+#LQzsqohQv_KOO!b+B6vfhKCdZ&nzJ-wg$R1gyiHX;tU7_x=7z*=GhY<<^d1?W z9-?v(dRBjywHbJjecSZLW(J0fG+DHA-H2NsdG2yXvG8-HBkiUeaainU?~3 zt8Lehm|R_s8*@YAICn)m(0+pGXHs&+WBL)BdB+j-BaT@nYHmoJ_TMI;p4ToG2*WS| z_I8rufj1k(uny({w{ZI4-p$Yr0PwV;RPlN< zre^+)XCa#l%v304=M*XmBFeJZUm(y_b$=1cVpa0q{a|n+e8(0@%v;I*MSwyHr5rW;D{v<67;F>m{^NsB^JRM5#_Rr@ zMrR(7Tq;@ey7P7&ZQ|pDPNtV}?KSaDqvwE5u&jvLrt4+&PBJ#;61+7iPom=(lw=+4 z%Hu;#MyYGBD{mTIsMICS6y*kVGwLE!05NI#$0o4r${rV9gEQe_4`_Zjj*H%m}Qt3p1126Nndm`>V`0j7evP~ zHzZD-sQg2lz!gsE=nRppTc$9;4T+;}#x{Tf?}kJ-x4}l@&2Qox65T|CXs*D0furuU zHayD0uZ2Y7mqAucdy+=RFg<%h{ksG8w^9c*_qQ9~ zynTNE;#}I@_$hmA^Hufo>P6PEO?uvH{f+wA_2;Ux@AZJN9Cu~OgdufmgjE0Nn)+l- zM&~||E47`YD^N`ajDK8NGeI|``h`PvORf|Tp1ZQ<#>zri!guP=4%N$Y87u+20&yNy zT_2&Z1;AZpnc{&ixy3M<--`vJU!ZseqOZNulC0azvMl}~0L-$R&E)V^RbHUWL|Um~OX?clv&9 z?%rGHwruU%@7CqbeG|3~6zz5QO~V%`dWl^prKLcRqy_*b05-rG)6-eI0-bNHGCy2A zUXSG3biVDCG(CD{dn6$enx@cNjdw7AJ#S?(Y>%XtEyS}oxkB4P!y9l!Ka@vs z6ra4&!}LgQ{3H5fGlSaGJUI)^>ycb9k?oB+r@zU4w0+k)xQBfd>13)Ye z!$ww5^1lTq|MkTT1|@F=m9^OoScm8VxV`Z;3#C4!VoOydCeVtFXk2AYg+5S!1%U zNk6~Uvc!w{09VlF?BSbn9bTkZ*68{td5 zNPuFArESTnE2@}KEO^yMI}!tsimWSjiM1=gTdyN=8a;jJldz?^!!pK(m7L=Tv=;Kg zxB$%3*vQO}*IL+4t%W(D)!I@Rbu+fZa^s!OY|WK`Ex-(7@^E4Roq~KZ1f7@4s@3Zh zY^P4a z8h};W0EyDj?@YPN@#mQ($5Ml-Bz1tUH65nl&e}?Yu?1A(Z!>?Vs~?Z+j6=79$o+oHO&?to zVL=G6wTdDkwknmz3zIJb&m4a8?5*>MpMLQ0o^C6CbT&VE5|+58{P~Op20yA^UA>$G z45Bl!v685~rweGFXCg)@86KEYoyRmLQdDTEzc^H{mTF5mLlm#mx&ALwdGdoagcJ@cpN^oS(&LQ0#oX7eBg!8BLcHjXFKh# zN_WQ=|@SuGLPd9zP8cs8$K}eRnt|}dhIvU z@JHP4aoWH0q~u@;&_NTHjR6WnqD>Z^@fv8_sevX3o#B>nw*27p=)G z#!%h?vuh-By7gLM+NlMmM7LsyMsI27-?TPz)nLUJZ3_`VG^Q-9fLJZkx?dnh^7l5 z8gmSG!+nfH%Y4^7)510wfC6!`lD7;*qoJM&qTx0qEc_ybX}B!>A%rj76;i#Z?lDkJ z9J~0qt#v=Fo?qRP*SZq8Mpv{Bu5pKdOG34UY{2Wxm6J!R3nsP@!*uJ3DZ94Z>$@9F z|4#|`wgo2sZf#|Oi1NE{9jMP_+x6>b_wJ?fo6lOCeocL3^@_}<;Z&6XoPW>nQgF@$ z(%~UE=?bO@a5=6nMfI15>LX<;-@oQ;mkM{+376(kZTkxsIY4o$i&Nrof9X)YQl^2$ zVZ|~H)VHx{%3%|e=?eevz_#Feer37SEbiS9I7b6ZYVF z!YOt>0i07N%`)eM*k9c8BW*8Npc@omR;A7-fON_b9|Wtao0! z%{2T`1#j=<#N!I|+=P{8KC07sJvUu$&keWgGZp*RJ0sSao|`V?70q3V-+`W+Pz*tD z4GC^J<$6sw)9_`=6~j8ZRbK$>baDS=AOp8JDtF#=Ugu5ccivK2+pJ`- zfY-?P1nWE}%!2{0({YU&UWcWxCp49Q@&d466M0t*8epdvT~dmVDJUkwYRC0|OxdQx zHr!#ZqKIvvoo)O**V+lV1n%+qcW1kb_&4B=_7bs39Mj7cq&j%6_a_JHr||`L=l-Mf zhxg8J-#L49UUn54*igQv5gPLG)r@75A})4c+B@{^V82E7B&9|UyOI1yUilZ^rDiI6 z;`OW7sGI&%(}oPrSHE1Dolslm`IPQekvx-o!==g<)fL3?o`RWpNSq*RPud{gg-~DDl9UcaR;A)hC%+3zo3y&{eo#2?7W7-ZfY1TNe-NN((%}IdUcih1IiMc*FcUq7aC=y|-=HN8`G2d6o z%3M9x&pVIc>alH>OZ$pxy3~|S=)6c2GYkDm4h#AoCWlnJYpC=|!H17-p5OoCgWKm1 zZ{D2_5S}(CH_1%Nfb!{QX>WwNlLO9=`A?LPq~KWoRMt91-5Fq6eq*ut_PY|ia02z& zc1MC`Y+DE(_|5Npwu>0l^y=_Wu}v7>Jr%%I%hJ?97$37;mM^I1R@+FIC1R;y#^ZOZ z@7X>1vU=u~cRY1%_2laCse@aGJ`x#&E)LYcI-nlqh)*42(mv=;Wdzk&53{xa&QRYz zRIk<1xy#s1BmjO>21BG*^xD zC5j&O%H2?3JInxk0k=z3T_m!jN(LS}LV!wClO~QyM7CsaOx}1aTEhzX^bN;GFMC~qI_B+hTB?!89hY4j}Edx&bl@@6kF+M^i_HMX zCKrN6qxTD8?!tX~{+E`TdtQS}ob9#uPRI8epPff#nwV>&9(+_g`be~ITYEqooQ|hz z{!n_ssd;7Fs28W>2LotxPsiWTo(zDqV}NCw)A3v0BhfeP1%~~+kGH(QaAeB-%Lb^` z%X$K8*Lh_P=$`HY3?H@S--N)WyWZ|YSu-1aHvVezOBT}_RJU5^V_G=&8kOw?G&o?^ zo%)vqTno#N0z$$R#=)r7P#B+EcjjNhtwm!uwaAlXSTV7wZ;|Y#Et-)54?`udZTTy{LP6~R!ea21rh+N8YJ>6sG`?M*VS z{`sLAbNOB3t=R&jP7UT*%yAW|;y7wjY@xq(Ag}uRfg*0piSV}oS+hlIEx3+ck`j+lQ`0oudo8m4)FPX6$tbv$E@_f#ZUG%mygMm+ z-Kw|8$+8Bs0s$SbVGQHrs8+k1xsP#A)70^rtKB7rnwAXg7S@Q607#9#-b)%2=mF*9 zgaOKGyM_Dts>&Hn)uRJ6HA}nJRJ@xh`i(J_=u50icHLveBg9pQ6L396T1VR8CUA8) zsdPuh{8+IbZ*}eE+qaKQxA3?JS9kL`pP8o5>lWTD5LUhB_D{sp5};(r)--_e8PwIurbyDFj1&mLcozLW|Gp7tPG=9Y4;NzrmiPE&KB#)DVV$4i{js z&s(@}txi_Y=D~e9AUu_ljyZQVNp7|>1J8Qdgoza3phDhEUB{6$Z=Y3nCrK`=GwM8f z!YwYpLDtCnkZVC+iQr~dYV;w*PCu?G%%;9O+G z>H4JBWh-U{Aq3@K@Di!|XNN5G3UIiD12YA5RBR>T8gOXbaRsU2AYn{>>emhw@o2n< zUqC4?5Wq|+#$Erx<;(zYS%Rt+3K$A0E5YtV#TE!?OVHUp_d`r>&1s&blI>g#|N+1Ya0DX;u?7l z&T$cK)#76==3$q?Ytm2 zGs^S2(htb)7g*wrOmn(Ixn8x6kEw76uy9J<$99L4`7E%3Q#J-p5k9uN4Um-iv0XhA z(~p=dKZhAOUiPYSpHumg0Sh$DHh~2nX&p}Deq7`O zkfP8)2UTdU1s{lffDa)QL4%cqv0U-N8SU#b9}qwZ`POFFA|DVy z3j(Rui4}2V!TK3joP&l4&+hpf+7k|;^o>3H3dn1S7^a4ZcvCCwQlm}`;RJLt@QDxL z2W4Rfy{D*II@<5X4^t;Y4$QziiF4f?j2lk8gK>ko*lUO$UUG1uWpz zFrF~ihl3inKGeX)JS@}zj>UzR6&!yf0h#$&+;Flu7PmD}1NT_m1eCg7gw06+HO$B2 z^h``y8K_}|jR|lH>vp08ss}UFplRAc4e?Fog$o35!PZyczPGHnyScp4rMGc|~pyTp*}i1K7&Ti?!F+C)L&k@p~P>V@M8S<-#ig*({fFdBTODrP35rxa^sE!SV1`jHcT9OoW zE@*eKkt~u$`+`KWF3#2BKei@??f@w&jXS7AbuCOn<_>&NA|}7dfR`;wwIq=C0A4eM zlDUHbmJk^2rNkW|HdHA>NYICvCz|pgfjbDG35jHDyMs7NVQbm3QvbtSGdMAN{SVXV zM*@l7+X22#B(at0SgG6uF?vl9V`dYCPV!?^Tiz@>(RgQ0M$-f_f^O~DfKCjRnjox3 zy(WlpYJw1GbcQEl95|7HP6%Dz3E&AWd{ehYwcvzSxf4R)`VNq-(d&d@6KCLw#yes% znofw3cb&_1*9oD83GSjF8qkydy|;-^7+ibMx%cqx1WZ?mW1E^WyyO-@f_fH_sowb@NN-5AUDfy?J)~?B40U z5UN-nt2edE->??@vU+j#0%@_FS*h!L^wm9i*Z2IJ1NGU8ta}$OP`fi#1rZ|D!8JeX zzqqD;q8^j;1q@ecF-O%^;sRM!;b-0ZvwLXt`t(LJn!0h!-UH5%Vz?@EK}D=h&Aa|* z&yZIZ#|}l#{D13Ey(l)umoF~vynf$C9fAszz{ZEph~K7WrQs;IGGW zOm5lrBNq8_RZ%3Fwky6Qx*@2^V~B3#HA0Uwx?!w)(XFC7%dTrLh;9gzh0zTeacME) z=!T8Gzi!CZ*;SSDn4_DI17B@)Lk_=v$D><-4NG0^r8v4F7H0iuP5(_*DQT}BR8R-1 zAi4#hu~_6+B&MW59C!vA+^E7MtQCT@b?O%n42bM?xQ$bX+Y92nZNgHVh%&eW&2Ay} zK~o#qi(~FF8og$>F|*lC15ZiPEtmElqh`KNWZ&2{nvS;-^d%)nb!x8E_U6>=^}UT# z-Tpoqy zLZikzj5(TSIE`8-xwHat&!1}9dXRIMGm6EX>pp)vH1pBI`2^mkF}U;e{Aq?#;e|$< zbpSrk&!6h)at75*-tKtVUJewJc)vC%ZpgwP9xVVGFw7`A?$%%R!t zVjOgnAgT`o*sqU2Vj<33YO@+-5GR4mr}Z7=My{W2DuNxe?wqDH-?8|88=@a>)|x3J zfzziI?z8TWb4ppaa6WhpwQAWVM0pTJD2;BTH>DHi4DW^^FGk|`>v5lmO;ZzXLj0ox zmE$%5lOCKJz@3$kNY;~v`ubrOo(DKyb|VS}hu@T()?>I?6k06+PwR<{@Tt*V=x-h1 zZ-X+}Ty!Q1rR*YdML|Sa7W)eXI569O)o8EPr6~3nK0p-{;)uD}QS~(Tn5*6p366=0c~<+y%Ekpu3qa zlD8K+xOI@*hm=ZuUQ`WU_W2}72vhhz@Zq^1)noe29D6)=G zdouuxv;mT2PFXg&rz|%!lD@dmX~S(QzAzw(_KFiof<77-%StC7XdsfrnLq=)secUt z19wHx%}h6Ju#xuo6^o5*)UfJuv60EA3$pkyUvTh*(wU$GMcippS@^Y(Nc^qEYH1(^ zh{SB&=h&Swm2qs55U}>Eu_sBv5M`@BIZ!{%ve_@4i#r>|Ob2sT>udb21NG@@YGt?@F7y^C!GqpBNRfqe zS1Yaq!j6OD8gC3y-{699!4xtOE*OScR0rRM_8j z*vN}s#fr*!tg*|7fxnAl7rUuZ9=n8qaNsM1u?w-P#}m7#YUB!Imk<>WM12L3Df+D9 zeCEKSVbKmgJ~aFbrY~{hweW4G7Cs(~xRlKDe0y4#mZxYF9v>h!U5Oj72X8a=;N<|s z07pvmU^>j!-bwe3c`|PU?vzUpW(2YY4RvAh#)pcnlDA$*-88&V$z?r>x=}^|7-#9j ziV$>~Q6+XL+5kX|~%?*-o8i4mHlA&{~Xl0)DGcb>f64!HYGyvZ9%?dMFId)|_61VftA?Qe4XIcAK4A`zW z?f78(LKKFDHpgiGFNj9FTk&{^yCc9L?wlsyNXMWYuXm*21^-mmI@sM^k|3oZf6VVb4RbM+~cEv5YU4ot>ksVdCq~mJx z!B<7hUc&Y2z*ePu(zl|;vgkMzN!Plndg6%*Lh2b15Ws+$mDF0R(@^Lce1HjQC9KBg za&?EVFJ$;Zp=SsnC5i1`iaY~iNqEW?q;As z$)3Wkcer~y9qxAM5Vuj(n#O#Rb6#VE0fxA(X0E5O^D&{^MAh}#(jU-lGom*bZW%%OER!5rw_`$12k&FMi$;gqdH zWUk{dE(TqLWL#w5A%#pFGMROU6l=twD}#4KB5&+{Uq%wNclmBrwFSK!{2LMx_B>Fn zyf>pNr$M!?Ef%>@QL%4GMEGCAY`jzM$Uw|Hq>wMe^}yH#iPg|mm4LnzYFY5 zEOPmE2?$jTncEtVegAJb4+5bqeU$;{Dyp$wuVF!S!&twn=vLNmSP8smDpNxp@1n#|+e@%H%^18WQ6I(?AkJA@adM3Ing!j5WPSZPE5`Z8I3vcX z+sq-)002tMA;6=}8TWNo8?ZK)sbx?5qUb#!{MN+3ubwgk&c4X}Q=1E~ zVai)|Qw+=TBcGk6-p)J>iKg9x!0Q${3{a!aacCM zZco~L;GK!RyGxIEXCkD*;P^wBuK<8se4iqmih%di%)&bpdv^tbH0iXU{wSp=DyxbW z&OPtW#NOSEAZ6Vm{F}$cf;fNX-ZZtR6dd!&TR45t2OF*-&~Q{yhA31TZH40Bfflm4s56C9 zj*{n6fO=Ar`wIjJGyb9)d9^wj#s0$YWsLekq|vdW9F`6t(dsDo7Xe5mvt;g_iOYqA zSQU|Oq?GaI#rdgjrufpG`>(GK-+vV@=2Gr2g1$zflpnXhxC9slXb2RRh*PJd(#6sY z+{_(;rnctau(vd#8IFoth(SF8L!h;<1c`cJ8Jc_h3`0|MUrvuBx2fvmrreuPI9hpH zB!PQ9o&!#z4SRe5$|-m5*fX-2TK0J7o|np9qE9g`E7i_}p#8rY5UG&5_A9MPY zHKVy#&WLV0GVU-ZE`8QW%o)*n(T&zfxQ%eG-*NDerWW8;QKcA=scmn*Wl{$If}3z9 zudQF1r=Z*UG^Xiy9A=Pdnuq5o{W}!LCBD=)^J%>noXL!``Z0@}%wYSn;c;|HU38#F z1~$>_h<~mVteq)8wV(-r$D*^sTNHbD-yZKwr<>fMC5evOpos!NDdOuF9Tj*U;oT|N zyK}3ufNVgL_KMCmI&S9oi*cr+*6wqhdw_LGfx$-f>W}HKVk#`C>;Bz=`dg_!-3Myj*re;E5$7yrsSC5z%8!f0{)L&jX%PC@*`M&I{KRyM>SsHNAOUF~J@&j+Bd zDn8Xs0>Wm>viKK3(sCz!SrGpaYy0kxf0t$PFTkk_M4z!{D%+#OcpK1dSOQxA=eDt~ znXmer(Ya}YqtblC`i`{70BfLUkdLNC8YvsUWw4oA26N!+F-^T?RGDtLc0Q*+W7y4! zfhU`+8}=+3+kg&0zElE$TLFdpEr8Pm1qxrH*~!#d!Cp!Y$q69sEX6oUMw{dKxVG)3 zoE{d}AoP)0Iny&)4%Y_2Xj&%Ldgi12#(LJ)dH`UX9?tbrjy}Sj)!=Di?n=k{O)q7; zIMzQJn0q?bZ??74YtO&saT8$O=4^ipHjmPx)8?}8b!@=hCn2Q6@1^{C?P<-3UNhx1 z!BMiK3#;sQ?UYGYtSF~DNq0@L?cC%I+9@{%Qg1l-6&mFz+fF&XB0jAF*iM(oeDnt?ve;O`*!=F4>2L2zN*a0n=3Vt@Do|44LT3$_DV_X ztLN9iEbJc2_!iM0AE-P}0!Z)>+Yk#C9Wx`@4Itcto=)}6T&NImy~Jq6;!B@gr-Tt& zn2lj|g`U3y7cTAt^`{5vXTV8T7wwi}XV z3(Vp+$JsVJ|5RtA&)v z0_+N$W{u9|U@LHJK3pAq*vhtJPA@f2CA)#GrgQTSw&K!f?Zh06ovlLy3iD6gXPR9* z=EeuBxJ`Zrr|M2zJ7y@UpzV@ebEX}$n@^vbcFfI;Q%&@mVMC^V{lQ>EQBKtFLioX5TbMk+8cc7{ zmTj1sgR)z4fHjzkz;mLC*O{Y(S=>QgX@IsM3s|02tL_>V(Gyd??6}T2u^Whd-yCYO zmAH~C3PO_9DUBf}9~hoF{N&kN=MO*q;Nd;pf&A!fZh;NUURFu;8H)&hRK2=-IR_C$ z$zUTd(`{|fGf^aTrWC+ZowKSS%U>L-S4+5G0xd7uz=1Yf547pl%mFND1J2Nf{zL%UEY8qw4QL}fL#rP`e%t5H&+dNYjwwbvbB4BS znXEd{1~$vMCY&u7Jkf9NT)g>((`I^|c>-;?T?-2@=>a#OTPCJ?^@p&ucQ;b?qDp86 zZ#3hvIitwV$8GidVfFm#mc07K(VeNrbx?*oIb0IF4T#eCFf328#10kPZ*i z@e-)i64y^KJUqXBdjIU+`RZ`>#BqHQF-okzJX9YkQ~Apm7k6I2zq-EvfA)41F3qFb z_ICwRtz2gP(xG~#OaqISnlgKiE*Fy1@p1>Vpo;2K#!K_lRx?8hmTCl3KuxAKu@JpWVSoUKSj1ZF+h)Z ziY&&YcUjuD34k7RfUE-t$fDy}pl7%Ny&R&pZxGtfo__Z95?hsf}47XrU0#S|7 z^|)#_pMQ5ItcZUL_W(apJzuF~2B=?r)obeGwe&jQBy(kULT#%{RgvLtQ&Z?8BI?fl+vi_7ee~8F8lyY8mOPw&!-bPS zC!ab<&Fevm$hnw`1{4(O+M>dzVQ;{-b&wW<`o zFt)wh&?+*Ehb7;M5OtzFb|DNF#V%yLrJV?47dG<#_A^sw*IP=Lk|3tUvCBuKE1J~R z1ggooFSg<2u`58MWe$~mfFt405v#fa*GwH0NIGmJjw*5N3SeiUsB5mXv;@rRb|_j~ zsj2W6ZBt?EHx*73P|r0a>?!wY*c*T?&{G)30{$UT`!sCp_Y}67J%t*7%FvFvld79< z4VnQ;{f2<~X`M5i?Tyo~8AP@g_^2~bw%=ViO(0bE5^he43*Kg?(K-p{oTU^e#Y!E9 z_BONW;ySQD(|R3-5V=gt#9EGjhPbtDhHb{9mp7EX%}fi}gP=D#_GfYE$}huq<^b#j zj_cmM#^BIdFT>vOZZq4i12EdW4BLTT&$K4Ax!l{Gh;PduJDRRMQ-9Z7X4cNVj9~nu zE*ET>DlZqD)9jdw%=-CcF6g1uH{}^lbF9K%WHwz&XCeFjsoWm)Wu|F+%QRf>Uj!}1dSeErzgnkTj zGPgNUFEyf?G&9+DhlyouTVOu^7Hq|uMGEQY$BuuB?NQ)OR>6y%%YxMa8y~aXm@la3 zR@+E7CSuEA#^ZOZ@7cX8vwG&0cRY1%HR0tSLWJa?kLeW+fmqsbY{xFWltF4vZ+Ahsdt0P=d82w-~k%>%^`L%^OrYt0t>Fc{ zE|RWG603GJHonwtjbakNvJ--4|n8x3seTM?)0-zSJ;zWA-3I+Z1IP2Y&zV&;J+o{($<9UxQ?^yf|eb@aA2l)r=Tl?^E8@*c*l?CV;4%y^G z;7+xE*Kj*^4NFwJ#J{7<5?qF;0ZR$=%~gR#z4g_5cJF|1H^F&(*KR%C@dwt#G`>4m zpKl`mx%zf%wo2j``~eej?bd4#>+YlH>cId8=I$$>tMC1D_1pCfb5HZQ>^y2O@G``0 zyF|EqbFRJzPn|@^v+dT?)fX1K1(+~*poa#!)jUqn?K-c{fnC=6Y+TDXgkS^I_6ga>TB$4L!vu!w zbL(ykkZ^18S-s-(tMsTHRFc7^`Ud$N^W3jr^ak~-)?Bx0T=fxirDvmhq;-$k6W4Nj zr2oYM)j~{9o`PhnFc_u{0qjbi*ZKqUsC5ogb3IJF{+u&XPzw`X+gW4o>)^=d%8k{v zr+)sr^~lFn)Do1-)5R$#xKpRu@$lC)3kb?ug_$M7daRAk>x!O52-A7>&kxm@%kL6_ z%@!DS$dDFsxJc_50PeF{I$XRAU@cR9{Xh}->_qrmz^>UMwbR#;OHu-osDgkPOJ8EP z`hbnb@Ycf`+>&k1bNEq$sDiQTp{*5Gd>PUzh$;wE!l;6bw6p~3ECjiRurc>h#h{#Z zj3tN}u?CjC%aueGA6k|8_B35n{<&x*t|B8zR0+VZt>L(bkhYJk>TAcYYUbH+9kG;Hdy@NwZ;EM0@`>i_SiHt8Z7dUN$|P zlS;HLmXEADE&FY;-P9JFtK}J0O|oG*_PUC7pIiQwrQ0^WHr#YheC2JTyX!1$S8-?B zME6?|ogN?Ht64-^lkxAg=sF9sf}`e1S%a?P)<;^iMzL^frQ0pc8_E6Rc8k`)t<&um zN#>kpe^LQFvo=W~4{a{@b~{R4^p+t08_AcZtGF|l1iA5#p5<4Q?0({z z5!+8ZA8~-PGOf7q_K&3C-NVS!sR@fo&l^b5hmq%6+BiLRVL)l^fF@8nQw)~;tVwkj zqC;)ny`MkMR)ImJL4*JE7SCI&lhw0%cpia)CKVB;T~AB}ren@sY?7P9SU4K+SRs)k zpk>Bl4{>jg)s4H+B$ri-MaLgEuE0Ij7m!r+R7=0aFi+U;7xV`kklAmjyJZjOb^j1@WGCl3R??NrN&T}!7N!L_T!Ms|0y-+TlF+As zQm$MigsY3`M*Z4>BA%-EPz@->1umE=MVnb1S712GTK2j|itlRcy4Fpn+3VT|ZM-@! zCcP_Jw2V%;>-pBUTdIvER8r3Yafn14P)%U)KRLlPSfD>-sV(PXNe( zhOiKu2!Iy~*HF56n2XDe-^(*Kgh{-klSOQa0Y;+TT>N9+_UV}ENa<8N%#C{8)XkNH z#DK&!%7N`{(}H{X_HA27+gl>xVbkbH7C`7l4O#&%rZPO=6v&hk@Y9WOycc0Zetx> zA=>#!3pmOgwJrm)43!`F9p+=i^8>%zZU$MVd6-VX>uKu)e8oKQJAk9aQC4kFTR)@A ziwAzU0G)~sbl1QK8ZDdPgAa0RB-qWSdytGh#GU%DfYN&7ZjgK*b5esE#LLWEwmLge z&#pe0Q=Lh2Yn`iExM1T>Rp$j(3l_vr+g3=X@ok`}jd&copP(-I{j5!Y|h?xggLsa5F6Q5B3 z@{l@v7~^ya7zl)QRK&C$I~V$?b0J@`V(>v92@-%0Tm?82itgYCAOk5j z`+mvPxByLPdgjN04_x}Jf%pgN`rd*MLp%)!;DQhiwJesrBk4w*r&^5jw|nK!iIgm>+Zv7%?;gGNMwc zk`v!k9;fcmU_Pe?l+dd@?2d3@whGyzyz4g()JK`N313+aPpz1$Txg zpizRq(;eU#$~H>$expR6+b99L-eV5aUI8@tr&W5}DA9AbvTfn5_n0F<6d$j!nw>VY zLE$}(FVXA_Sj5C(oq^Xaq0@B%V8PXZlN;!k2yq?(U@>7(P;P#=1eZPwSonui2HP#6 zYc8Hl-3>s44_Nff0W4rq=}3xsaCKN5Nf``aA$)Lk3;+vr8fD;0-2xT^(JjG+558;w z3k`fsfJKv)c0$BLSDXfWit0;!4OaE;%9wbc#-a=`kOg@8`9)g?_NrYq*>zwYWMInp z9A@DTk_ck21ueQdq&fA}NrUhCyt;EG0$wd_kyz_7Ors80M1rRNmjm_kZf(5r;KAMV zhh=Y~uC1-UU;WtXN1UzFgHsp#=u3U_F82BF&E_{fR|vNFSF3v!36;-Ccnss(bhnTC zCx>e9mvTs+?^7sO+PwLZGxuIxu@g+!7kBVTq!nDBp#Jv*_1=V$(nwR}#L-MGa>s#$ z?f_cSYWHr}*-xkb%R}{J32v5HQi}UIzUAnWp~b)}f#Wd}0iht0(hK)TQsP(uKq+y! zkRTy5X)p|@iB#ds>q`u>lTRXf9nBRuO^gxmni>iDxIgrduo3>3;WQDFc}^4gaH+JE z)5L~YTrndNSl|-VtC!(4eLN2LR=!e_OJjnXmiNU-7+(5PhSLlXw}h9tOgT+Nmuay_ zu{Mb$q)sz{$x`AdahlPKTjsFmDt^$iL#%S-Zj<2rkh*IyuxaNVR{T@T4lATE)M;uz z2YL;ansn?vEfWX@$#n_jeZz^9>2}QLDjr>I|S$Cq0}~YNd7^r&PaP zXPDY`5=xz+t|Zmg(!Fb0bjdEjyRwAuqy@Tk_)8!F!<4<2Wx{fx4f$O<(JLbsDZ8@} z{}9<=TXi&=lObaF*yHHKUmL52^l-W}=IOKBw@-%wJoXr75x+eCgWeBVgmCqE^D;1p zq2|XPM}vV)k3D8^>0QgR34k#3*y8|T<>cs-%el9^mSu>(9KgSp<=kA$GOhldxw!F< z-a{tgQ}&%B408+1I3LFX4M!SaH&f=Q-@fWN%-bEiQ&f8uj^HR(TBctxz@Q#z0tVw; zTWYh6^?w3b+16dwaf1@DLH*)Kiaub$l^<5`UA-p?t`NvtT09yKvM}HD{O!vZ7k6I2 zf0W1_ETk8Ic`-lz5!|6NXB!?P?yy&a5sTNwRgupK0l%eJ)l?qu-~IH8dw2l^dHczU zwU{4MKeGC3gvEfNDm4rJ@_|B@6`02ZYofa-NI8+u)v}m4j^Cc5{t`tb0R2mNCeH0q zuKuF?4ABB3Up9Do@!uS%zZQBcD1ikc6PJK<_4x$G%pn$_ z7v=d>W*U2xSfrXyE{>oKkDA71d~qQ9t7`w0TaFrCYHSzILu(54}v)>Qx>32f+%Vb&?IO(QC5c&ZBQ0dn?zKHb9HS1k*MSF8cFnS zv7oT#%Gz0r=sw`ErnVxKoqPcmK~#g*7W|VhV`EOfjNF4Ou90NpRnmF@ zo>EK^Jyt&XGA>TOY~~_*nukpqC=$2J`dJHWuJy7{zD%@DX1=_30~F#>_5G)&Ueqpc zsuzEdNQ+?0$jGJgzk}75yRYH<@i|PH7{8a{$8FO8UiF^U4@Q!HqDYL2LuW6pK6J-& zq7lwihhr1QZg3DMH`SqPLq7fP=uMP0hv-@kb9&?LT~%OU=7#`~=${gC>rtD~JIp>q>G9Mf*( z)Wf0v>jU*6^h1PbZ*P%LL5KxTlk>2^e}ACf$0TpwN0h4IC{GN>eE!1?!uP6YPG_!m zhH$a^E9&jHo?ab3bz1HYx*F(`GFlC94Dy_fq?0~gTIQ> zy)4gh?vA)h4_7_tgW!t=t+bPjtABW)j&Lp)B{X1Q5S4SJR7`rI0PDcH zl$c18CRh1IcjWo_#+ku4Mt?_soOa|RPS|mmK{MDd=hyQA=AmB|m@%2r-!C7h{qmfv zK0rZ~X_?&%~aaL?#yU+x&ey%5RKoF2ihxR)L91vCL3%IVSH6(6Tvaa4~pwvi&$ zDq8oRI3z%KhzszI9C_&N8K%lhObz&y<1 znGqbGNsg&4<=zgW9l=3E;aZyd3e?{;Ed!%DXz0d2depjH_i>Y9gZB8)I0MwMO(4ho{^LFBGpS{V`e2M&3bz>){$q%rxC=3!@ddd@di3m8 zR>wbVi}81>?_0eKDaHvnC#jAQ(@_5uQt6H|p|RbfJ6)Ke0Ca?#?^CEwiVdG;cR&fH z*&;f9M)oMH-(o^h|CQcoo;45mR>=+hAGJpCgX#xX?}kPIzCZo}^Dx{yUwKA94EN+z z^B+GM0wPBNMDUclerxyVN^f$lLee}$AoBUd^|B3^Nqm4F{09a-m`Q^b-@)`+%nyWg zE)f8Pl_QqqL-m2$pUDw!I1%ZoUp!FnMwnU@{0LLdUodHlQGr?~Qkj^N-U~C4fsa|V zT}I3pw@$xO%(Eli`Lcw|>{<{T^CfPbL1h>!@)EI5MDm&HrmDXL?~B(maO(^TL4>Jn zawCMfe(Vj{CH72^pXyB|qWVuvPORV8f;*$r_(#-ojM|Oal>@YbUEC0|4yrpcf4^^O z^>=YM(=IN*>hm&;n6N))GBv*^cwXcvBvjU*k>Ejf`(s127M8GD2v|Qw^v-Y0bkw8L^p8Mws$0U-_n2dS5GU?9;9r zu-_SBSbcad>LzhMIW=6HU0qNM$G$i9oL-S{nV=KePEfazwqv&Drv<-*F?LPhn!353 zo93}vim6gphM;z7f>O@B8nFQheT>o9OwqNOfhd3XdVSLH$kGMY#hYI^)qlS)W8Mjq z3b0<2J$)<$WZU4HH@oxv?n%X74+51B_ll@+`TD^0b$r80pOlJSYk+zwZ>)g2< zh9O(9#Cv>MNeWf1unBc7)*+#`O4)KVrtC~JOmAJurvB**O1=K@!JBWK-adchV)f*w zAAIJGqql#4FWA#&yLJDV){Olx+T7qj*6<%K{6`1>(ZhcX@E;@m#|HibQqVvO8c0C{ zDQF-C4Wyuf6f}^622#*K3K~ek8d9)^6s#cyYe>NwQm}>;tRV$!NWmIXu!a=0kb)Lc z&_W7YNI?rJXdwkHq@aZqw2*=pQqVyPI!J+j3EWvo2Px1GK7;=PDd->t9i*Ux6m*aR z{hpBiW`q>z<|*)BAO$0&K!f+-&maXF!Ug{YQZPaaMo59q4Ba;y zNP$ipz<+@hY#;?2NWlhDuz?h8AO#yp!4^`mg%oTd1zSjgelW_>wuKaIAq87V!4^^g zePRoJVhep@djcqEPXGn&380`o0Ti?+fP(e}P|%(L3fdDOgBJS47W%{%`otFc#1{I* z7W%{%`otFc#1{I*7W%{%`otFc#1{I*7W%{%`otFc#1{I*7W%{%`otFc#1{I*7W%{% z`otFc#1{I*7W%{%`otFc#1{I*7W%{%`otFc#1{I*7W%{%`oy+_b^v{13w>e>ePRoJ zVhep@3w>e>ePRoJV%tMYgg&wDp(R3}*g~J!_RtccPi&!2YzMd&pigWENCEVT?Eoo& zKCy*9v4uXdg+8%`KCvAk1<)t9&?mMd)B^O0?Fd&R^ocF>i7oVrE%b>k^oi{T`UB__ zTj=o4G$6IuQi)3??A)Y-#3XJ5E`{+Tyci+>-z{q?&KzHoN;^y1O^8?cr1X|he^ zZ_IZO>gFojcCs=pwokwEwx^z6J@L7N1KwaiF?UjJ=Bq)LeM@2M}eMX-%}8d zrj_dcPAm28cOP6_oHjQfoxOfxI)C16RPf9z?|ABMt0!K$zW*ag$=gUuKC9+GU;UMz z{NmZ&vxh(V=DmAo58pb1Qyo1$4SDlZ=V$kg)DOrJ|KYD5s3-5-zw>D3>TR=(j-IwH z6GzJa2<*0jM%=)_-oU`#z`)+Xz}~>Xu5agYOQL~+y@7$f(ShAj1GYyS7~~rm82lR;{2Lhj8yNf>9sC_NU{9=pi9iDrfd(c5 z4NL?Ym zU{cWNq`*-F4Id^54NMRkm>@JTL1{jdCk0Ie07Gi8VyD$k#r+IO&@RRO9+03Niu*kvLAw+8 zdq9GACYA&+jCLjV&$TCUKLZl9A925j-vSb}6LEh7NYF0C{T`42KGpui{)u)U?q@)P zb{_8cfCTM2-0uMi+Httw0}`~`uu=ejuAPScbL}(S&wvE&G2HI~3EE${-@`Eh3EEk> zzX2qGKhd7T{)u)J?q@)Pb`$RRfCTL%-0uMi+C{kE0}`}@uu=fO*6zXnx%Lk3XF!7X z4es}V1nn8z?*R$gFSy^sxc~{kL1>>~|6IES_cI_tI|TQ8K!SD$?)QKM?F`)S0SUUE z_|WHfH*THZ`@;F{+js82esR;i%l3uyM`z}arMah=_vhxA&(X7y(UAwFxqolRUkBsA z@!;alqdO1op8}5NYocZrGQP3%biq829(?Ki{_YA7dZw(e>|CMDe)hq`+w?5VzqjMR zI`eY>iw|z=SI_RAy>Fd0ZbwllxL zo*Cs`_Ic*)Uw-raA<8*#_T}ex_N4Lbq`v`CJh z#r8mWVRB$8;4zRF+k^R+pWE4a{D-00%3nHwrGT(|B*JoIdvi1Tv7MQaJo0gl-6uY` zv!emJVbyo}pW0c`_>9bo`Fp{)c3w0{+j&Xfp7)uZ4GqzY*-)M4erRXJV_|OLUNnlO z<)c_;(?n^()Lk}=rRF6rHKY-UN3b-!l1Bq-8U|Oec%2!-Abo!3cXqxFXFgQ7*WAK586>$FrAkTW2wQt1wsvJ2jUSd4S3upX+UkGU<^wK?ztcx`LRFo zwVfOHd+66F|HAKc=U>L|MBEGLUw&`rPg{{J_7fC98Cn{cGF-;)Tp29CxAQY)kRAI2 z8H6KP8mKjZ#7}nY=T`fa{E@k#{*v-fd~bQjKhhf^zFCD;AK$tEPDv>ZR`Pj}(NFJFv$LK`u4of>})7BznBfxY3_%1Ml`m!HDWt+5K9Rfm7{}Tx-J^U z(t@6*q-><&W!pv^kbBF+2@(FFN3pAEUg>KF^$e8Jz8EnGWXJZz$985kJfa_?yvsh%oP8Ob?P}!BzWm(Io-`g&Y!8H&8{2~cmccob z|K;bI{D-00&d(EIIDn;quzMs#a$|d665G*_bL<`=9DZ*3Mq_&~kL}S`^?UUhUmp?P$|}$b3e2*qOpCb5!;!ASW3{S932GHbKF^$e8Jz8EnGWXJZz$985kJfa_?yvsh% zoP8Ob?P}!BzWm(Io-`g&Y!8H&8{2~cmccob|K;bI{D-00&d(EIIDn;quzMs#a%1~g z65G*_bL<`=9DZ*3Mq~RZkL}S`^?UUhUmp?P$|}$ zb3e2*qOpBbBepXKv6P@uIXVcY>!MLCEvR!qXd(449l%n624<21)F$LZSSnB-2dT)8 z?TL@=%xHK-KSp_%eV#e{GC14S$eVrnxt%>}Jfheh2roCb2LmjFb0+`G&olWCL$jTq zC%|w3O95f`NQmUd_DxA_M?cQ7dxUWKx#b&;?HhS)kA7-rMdLH_Ddz75-`aW6AZ_O* zeS6+#b~ZFbFJ^;EvDTdXp`8(p?b{l$ojHi51dYnkK`>nxjbdp*odZG(sekDJmI5>| zlN6vfAs@n0f%-T|MRsgYd~9b%!z21J%De3I%-NU0*{(+3?90#X>`CJh#r8mWxv@PM zU>Te<`CopX$$uD{?fg6eh67j%2)jo@BsaEiOJY0vagN<1gu~A*-)L;#%42)2i0}G5cs9-43Hq+j?=Rdr-L0(oWq~8_m;S2*$bYjle82hbXV34v{xgp* z%p~tw{@x!F|IcPr^^CecJ;`OB8dE>wJwUdepX&Oa&rg4J^tRpqoi^&OY95}y@!;X3 z(>L!=|K-lv-8+Bh{PyYP{H)x?P5toMCoVqW{*!)aZWaE{dwY$1IP8BtP+elWPh6On z*1(*r??Zom^xoxT7K^77j@Rn?=@;*wU0kf*uYPRxBX=Ks@zzVJZ@_zT^b=_R$AKCo zoM`nBy!_~rF7(L>O0>rt%jvV~gR5hdQ+!0oulf}!Uu%l5zrN%Pd9s7#YOq{=P#v${ zk8*{L=kcViPNu7t;^~K$Jo$I6z<2pb6FP1zZ|_q-y87!VZ~nOIf0J_3QJe%T$j!J5 zY^)Wu{02&&Qmy)?l$W03-;;7XP~1Mt=XMHO z$1hu6N7b*|OkQ_GTrH<#5vS7=CHufH<`fLU;`HlMoo*;jKgiLkd5q=wMXS+XwX24$ zQ9A&OOYu70Qe3juDXfFU>(m>B#qBo^)F_MVCltG^g}SG-XrY23Sgd|a%qp3~tP}R@ z-O)L(Rt~MlAGN0boZ7BNj;SMassCHdGMV0_jM*b&r(Z$P2?yawP5p_OT{83Ocp7#Y z!Stxy@yD$CUQo}iwi)VM48&slPsMDLX;Pod_C9&ykbC@bE1VCj=U2Bx!r4uy&`y555SUyH(5%zSICLHh)9RCYepWA2nsq!_1!fIcu1W zy1DugZJ5}{EcSj)%pRFS{aBj4o&L6#xtr=ks}C@l+kMJn?f(?BM&?b&G1eB2@&(IY ztD4o#Bzud`SuFk=DT`||SNZ^t#l3g^u;sB+>(xmHk9_YePZ`!^mV~7Y!JW5oAB3C# z#s^{PR2)LqWQK%=5Q5U(^7?-D%<0~voL>CQ*&F8}i7y`0HVlComxh69rELx&;=0r<+02Z6#^nOj|KdiOtm6^=+K48@wgvG9W zt&%wq%YZWrP4U8Y^C)qxx+e1=tX73e!IBzg8JZe~rCqr(letmKm_3$v>sDaQ;UFw_ z#kK01%$KlQ6;aJ1^48{mi)+=E%%QMaHA?M18)BBBjFDK3i=9_X=2f`!Lh7$P zoEUu$AOpOT*Qza05YQX01wn&SFtqtF~mWgw?7DjajuS`8|tCamvt= zSrV2q1egAMwJKGePNu_vxodoqW@YSjgWQ7_D_VX+}0 z%Sv3U_GI3K)vA_`g~w&qs?_I@71Ga&Yt^32m9SbBp)spgCBJ7eDNY%BGE2fzhTt6% zuU2K|zwtp>?21E3Pi9D12q7rltxRUE8XMw@StYY07OSp?00#$fozffSH4!s9EfGWnT4j<%H2Hjebt`K zgRoi^+L>2d<<+XGVOZJ~2cCh3YK8*w=!mc0+EB_1DRLh&I_r(@@mz{J8P^1baJG4Aag1_ z`Xa3ERxvxuQUgHEk>N;Os}5v7h1II)<7nJk75|vUp14*W$Q%l*RTmUSvnOV)iVX=_ zme;BSnKxmzs-2a>Okg7SgnfCl2xmc-$Q1|6(~;`$Ses<8G>iQ zB(-XM5Ei@Q{C6ZXBrJpwlpGEm7x z=T4^_D!FG`$ywyJ>PY54thMTunauP);KCb(BQJTnHPY56Sgi_4@M^9(yQQKkbH6n;Jk)xDF_XDb%9uTtck5PQ%;6xAEi%rEYt@m=m#|tD zQNbc}V%DnZfmm#dYt@m=p|DytN^RD@Dq|!T1nGiz0pQFi!N;#ze>=2KX$iaz#9xfN~=iw_D7S&3`a4Vgn>wd#VBo%oqG zDfTI3ffACuR^5);chRlChYt<_=Yp+#raHwthTHTU45X*o;ttvHOFa%_Y zT&v<9;4PU4VYMnWEwARPQ>&(ihoX=h^OnqwQpW7Dyz|9^022-Z*&;P8u2r{WzJ%4P zhzb^&6SG!L55!_yT&r%$915#dqts^As*I6XjEkf1mdvYg=YOD0KJ6d_36Ely?%92?@Qc_lL>7O$Y#l=2FO zVDb99a=nsi5f(B7rPzX%85-#USbP?&S27D?tyiz8l&)8A5Ej1$>y=D}ST+pmO{ooo zAy~ZrKXMx;^C7Heg(m3LUUQbq+(%6f!{WKXp2;LBWzQbZJC7`|=Wq}fzu%VYn9Q88 zsufYrB8OsDt?7YSeE+{o-rUqc?^f%?!OlDSC zH&(ksUcC}5uXarh!_u!j{?M0tr{a&t^G?G8dkzO-@hh)g>D#@r+7(gGB8OtuuIYhT ze9LRsHJMFewQH2ztlE_^5{vT!4@PEKcoas;vAo)K{La;`YcjjSqcFnjZY8q=EHwa2 zr}EmBzULdOUB{Bz75|vUpS*UZFZ9N0*9D3EJc?PnBI6g-uJoPVSnX=(Uie*R?Mi*l z(w}_amA(0c1;b#;#pq1(wBUv?Aha) zzjm1gE-992CMB<3>D#`s+7+QDnb~D0 z_nGDBOeO1;%!XJt3<^)F4TB+Ax|P?iJ&848wJVgkSI6bmuBl;IJQvtAxgVE`KOWD! zwJV4-;UFx2<+Uq)`!`m*BFb6hP|VsjJrIj;dF|Sh7!+2!M#%j3SjPSZ!$t>kE3tUtj(=19}yVCc7W3}t{%4=72hJxCa zzVI8XT^Fa)o}O0 z7jM0kKRT3z`lg&?GF!TtK@;D_zG9i*sA1LRGQab(1>BM-6OUU^eTo?YV?Z5})%x$r zbxUSQA7bg&mas2br46cEt@%o`Bd|CvShr+G#9Oys5i4Gw-Y6`N<(5oe0?xB!P=it+ zo7lp^*iZ`UkHwZelBeX_lqGxh*qrvhX6^ZTb!)X@*)ui{6k}77o#u$mk`Jc-zoJ6GtriDAs_o>z=tdYN{URXVs zW8}-BSiBebG7_aCd>K;Y%`2Plv*9_R7NP>t!RE!rbb|K`rC4wCbO*{ z$8Fm4H?u~cw`ROnC#z>8W{iE!;_z#74#^DbeGCqFTD)cXJ5kTBKFH;7_brP%dCfPH zSru0EX&;LFVCH@$`8|tE`94qH)!!JEdwQXVcUIKrgm3z2%pcNOJRhVl4d+=hs6WM)48~w_E3a+od&AM%7TO_}@o;O~ z)Hp1z3#^*VoRU`UalP}-5~~hJVR0<4ZRy*@(b^W#k1NZ*aQ^6wyA|1_M`CdI9l5-rr&9iTiasaa%5A`i%Z`Zj@Gu8!i5j! z*0$95EH34>ZQfPkI#L7`d|qu!4#48`m&G^?eOLI69S-xxg9mr@wyn4LsBMoASU~27 z)q7X(aexdq2;@us&vN?7Z0P$#`gcp~$1RobRqt8-AVDSDm#Y6(PCJPbJ+rr_)1N)R z^ZL&`x;Xw}OY^(c_pRQwqj~WsP@d$D%0HIUx;`QAnSA?9$WCHEWC{IVwOYLs63UmT z_?*Z3gxu|#du8r@;r#aPJNI9|xcRdW9^O7ZcmF=SeRl8kUgYXL_ou|CEDaR!M&1F* zQ^2C=tYpO6U5+^gn<|LDP&&hMW-dh3mIM0*?+M4s^F%_2Wfwmu>Es&e?e zEMk|@rnftn&q~CaOap%#U2lo+e(tTa5&pE~*?_FD;kh@WQ9xsoJcx3AF zSLI%x+=u$n{Vt7DgrDBsoZach#)D_o2Uo`~*aFXsf63Bb0cT|1z~b!kSjrhwX9aPB z%qoI7;UvFHR*ll=eo;)_-lr_B$$6u1yXNu6i-D+En0osQ@mizrvmTKh(P#suQRGk9 z)%TM6@alQ6`n+yT)>tgf7I05yH$v_&sU-Ngo>Y9BP-290-H`bYhHw)4ni6cId1m(x zVVVxvyKwm`Zd`=145g-G-sjIaQWRZ6TZi+CmBEWacE|T(0va zhfJM|cdXX*&Dz+GRbZoL>(ywSxnq?a2kMoy^a8HQTuZ?9<#8RZnYu2BXk;G7ifC}# z*yx+1CvkW~CRB3n>ASau+}DriA;Z2+6nHrLMsAjeb6e;}^HLukj>NH{63)pyPQ1^iAE{ME*W+9YBoT=>3y6gMUT6wE8e7gO7~H(tUvh$Xre=0iM>5 zYl`Wz#OskcUCQfuT7#LqpjsivV(GuY=aYGzko&4WpOA`gVd@tTh3}fK*CbZ<9B=(Q z*kNi?F&^O4)&qP*y|Q{K!vo;sv236~3S_QVAcf20x>E3UPR4>?lG_5A`#qm)3m&{i zm3W`AM)Fbh>gr{Yk$4k9D^YSe(07KH$bqN6S%4hhp>G`*ah{P6COG67k^JQ4n!bq~uX+)BVIIJ)dK2S7y^@|!UhC8Mkqfxa zsP&^yv$-s3i1E;I{$086h0oRYa)7N?iRobv2KiF@eR^v&d0553h0o>4P<=o6!`bSvkW zzLz|k<4f-Gv?#|+9gA=9Thlj_V>kFAw0a3K^VRRvFf1(>@Jym_`8;19SK*ndXF2LX z-$#z24)9z<@LH+Ku9=t%Yb=(=3%DmSHzD`Mmr3^8>rk7)Ew>8ywr?&{v-Zixq+ zV`Fh_r(g$(zB^sWJx^a1MO1>5(l@HJJlt~6SMJ-xkyv_{N41{J<3yY@wa$c#iG|@@ z>ZOwJ8q+teW4p$X6c1_f@p0ycFmp7P?h7P9=5k^QTynR>SrXt`Q;F9jF}9S~^K8Hw zq{!!!W3lvK5b?;oPRRYG#GMn|Gj%WDG49FS4%;zC>=j4ls7j8xlgyq6+D1Y50DW6~ zffU&L=1gvxwjkdxrtfmc_KTxbd*~f=E0{kZ)H+$|707|a@k-=y$=wp?%7JMQ@~nuy zA0E$&kX=AOFJLD;$&pwVP{4U+KA504qt;K3#Nu3D>(e*M<5e$0FU$j&`{wA?FL0cz z=9v$qP#1%Hv*EWUgS@TL&D0rB^wp^o{aaoU-@LNm(&ZRFCDjAYLmhB~Zcml|j8o3qAZX}o}YGT#z%&&FIB+%t7AuP#P1@51UTf?^n9nROLo zB$nO_I45&35$8NjakrS0;wjl%q3@(;alYL1W$Ooby|vCipx>P4KhYw0BCJnJdg zL8fo6$Er${UbxDcwRCbEmcHc(0)2bEfNORwo#2|OYk5T5kavi%A{w^W&Fq5AZk@GJ_LxFOTOT_e|Z(quPedoT+p9t}%W4 zJ+^BM*QmGC=i}qlt{HPbojD7%iGt374VlY{CBWV{XK~83fda2bVr(g|=h*;^@dBSu zjs^8h?vly5r*Fp>a$ns0Oj?DLux}FuJI3_=_}GpyLK%+AQ67M~Z_b_v+D3^KNPMn9 z3haGzCbz!M$(11AFQ)I*$M%b(RD0+hWem*ybpAx1H5AB!%mPc~z}`3K^2@Xbc~(T< z(T`_E$S$Ct7qIie_?`{phE_CO}L z?0xeHx4sTZ4<_DCZs|7+u-)VZrEujj_st_Cu=HBMDVe)vaZ0b+QVICFmY+9o>Gut= zU1X~fJVR&hnyliz;yvV+e)9m^Lq=%zGH>R-d1@Gz zmJ4_$v#ord+56^ko|$?UqYmu}xp#`84sFGKbJkcajTdlF=37GUdHUiWXkdt_U=OuD zA@eS*t|EwmF`6o{ne`T9251lR^X4u6@`8wSI%e16+_$R|&dE2WSe%Eu-BJ$2k)UqL zh$gP3+lI`~M4Z>x`BEYaor`PfmVS=`t15$4IMD@HIkS39j>FP-0oNpYCg9rF1!bLF zm?0sWf;)TLhTJ=4MYM3Y_a5uDG;u^jCX}DNY#TCz6LL?la^}ubpz4re-?mD8JDJO| zJlt~6m)#RZ@CbbdYpoafb~29>aZYPp8ts;{UFclAZ{9X!4u|a;Lrr=6oIZ-i+#hF- z#?rl90`%JqVhOPK%~=xgbxp;N0o1C30Zzn{)YP+JihRqTg`9vm#`-&(90k*-vsLsA)1H z7I2=K4<_hk@0+*S{xOsxQ|I!kcTMJlc-4z=?xGia-yG!|j>FQooNM}J1_9UX9{B{< zOkK-&lIeFCu$|<#!oGQQ43>@yxFz$vOm5lx<`Hh0x|Q!H)9)={yU7bmb5|bsigS7z z`Hl1~{l-ETr|k3QBoj=X%6F0JcNMT*WJ@QWp;Jj2b3Z*XI$X;IodPYHiDh$4V=^^3 z_I0o1d&uT9#0?x@COvE|6u1a!# z^*Sfdqp~<(?(MSoRszn$k*-?jB(^5vT(0va=S-c;t4jJk2CS+?H3e5WbKg8U4olw! zT$4GQfNORwo#2|O>w<_z?wztCTDYrwj|x28t@1!3*c{Ln3U-h>GJ_Lx&(l{$QIzz- z(eE*^Jlt|`m)S8P4f?@!1swzQn+zh(<-UE%6wiL-r>$H1T?TB|7_LQcpA%xc^onQh zr!!~q*1UY*oPL`@ECKeuIg3-K4aj{y{XRp=>v=W+V?1-;oa7&l1@%k?lLDVlztJG% zUR1*ol4R;$zGFq`{rCZFzrE(a?o!$;8_u}hwA4A>{Kr~63YVQoYU_% zh&a#42NRq#buO=Z>9-p2su!Ub#%^}4pBRUw?*gt#JWs$iPegousvjiRxk5r?iHtoVQE>;GyT#+KF`Jb=5bm0`Y%8o=rA}qJRIq&bx!7IBF@>hbdqylx8ypPSC#a83|Li((hFBP)#=Erv6JIN zJCNT_-;QL?Cg7UAZ=T?esq2D>MxHldMKo+ro4JP)aE^@ybxT^hoO}8$1|j$Icph@k z)V(~a(eE*^Jlt~6x9+|xA5kd_EybrtalC#`HT4*p4xxFC3L)?whm6W7$A~6v$k!Knm=AbEXuSwjkdx-jKN; zwqG2j+C%TCaKPM8=T8K!q~N?c{c=Nz9C+&EK7!5_N=Y9j{e}ac6(M_{eqO*vGm;}g zO_MP}o(XSBTu;P#My;RVoM%J@RWJQk177ta^upN9uIUrwu=FiQIJRUCDBwDy){k<{ z)U|vknSO@>+ez+Y`{tC3XKo%x$A-Eu=n~kHd0r;B?DOUku9&)&?1lX^->9)Uz0A?&voc zFw|kFxNpuH3u>11*#+E_`IeA-Hs-?Mo~e6rbDbxxj1U~w+n?dI+<1mX#%lkcr`CuDvm;#{utB~whDi>u0xevbjGDp4iC zRnA$P+-~gDWO4+dqu*i>aLulz6AUqREstmo`3?gsqJ_J?_qcl~0q59Q9NUraAa(Rx z3_|YZ@jT?7se5@;qu*m-dAM-joI+EiaXy>@+CzbFZ^%4O#JSwJFF9oDTzuNPqu*t~ zc8%d0^>+Gve4M#I&KwQun9Sh|BtYhJVhOPK%~=v)+CYKNCv&=#*Yj*3JH|5`sz@P$ z2}^uFnb!%q7x{cbl1$x;cZ@swod#^j7*Pd|%5m20F>`aBJE{DA^NxP2p+E}keRC#* zOj{7|n|Jhk4cLBhlxh}w$J{sPPXw(_zHi>qFE^CPfqmYb%P-R& zN4aLMZ}Gl)N58{>?IaJeeRJmicytVwjtjUY^Sn%M+56@ZZkf6j?i~kr7yWm2*nJv5>_nd*7Vol&MqsE;9YD0=A26>BKX1=DvAi6qaraI3{;gvpMEz z=eUMgE(`XM>Gu_|J!C{XUgphF&oOfoJuz1a&t$fh&og`9JkAc!S_)7H`ppFlbr>t| zo3qB^7)}ZIWWFWjo{hOMxM%8KUR}^{F<^BSK{1TcOgx=25=-v|oRc}2i1Xs}=1Gb8 zwotIcK)=ht;(WR1OW)}#wS{m7XbT0L({D0}IQKO|RlwYifxrNkVz?~WL8jkhz^Y1= z6u8QvK1jZqjq0h%O1LI-HUZc43a`VcZ%qXets~!IU`4cWx0k(d9;gx;rc zkb8MN54rbsOs$*}4@bUl%JOi_J>UAiJ)8mBLkZ_(9w*|Qjgt~K@4O{b=ki_Sj@&cF zc8wvu-cFy7k2CkpnWI4+lU83K0Wy~pOMty^&XNGr2ITwZ^!p4cujko-vs!xPGq=h~ zxxhJ~T@?6y`i%x5_jIhU$GvY81v|#{I}O;5F`^e7m1Ay+v&VzFCVhasdw_ncp+E}k zeRHN1e4Ue0knfw*?=@ii#Zjs~^p3ed&L5Cv4Fz%_v%nHLu=mZma$wqn9OaNYXXRdY?|fC8@R z_+LjF{<@c+!0zaG7_gn>vBJK2bPSe`<@@IJ+Y6c8viHp+-1<5sW4U}cnSO5p+f80j z3Rf`I>%`nokBkiUEq~v<8%Ufki&OT#Imr=Ir}AB7`dtNV7unK@XXwlg;>0LWholu3 za7<=m*&MUan@2fj>R5i>oPJ*c+e1cZ^)hefmUwCymX-^6CbO-4o_Px5_6o!Wb6J2o z&~GkasKch>zBy|wmd53X6aD^zkb64kVqz`~sraT%+}8+|b8z>~lj14aH>clY zz^Y1=6u8Qn)njrTmcHc(!bs+9030}d5iQ*9W$&8@oMU6LbS{tQ zBbmVoxu-q5G@gf2@ol0Yp40C!usmG2t1XRQ;YcjK7sPWCHxqHr#z_gAciz(1Efpo? zx0H7yc{eY%YYgf2_Dp?zoVjn#9F3)W`Mx>*HiK9K?0s{V1ei7;_j>gE3@NYY*?_ZJ z<_ir17f4N55b@|Y8id^Q^u@h~g)|W-?#qIGbNZbIY{wWO21n&6w3WG^&YlPQh7u`| zxn6-3*!$*8Zh3Z9aL$^3uL0XHj#BNRcg%fr{(w;HWK5Lrn{UW0utW~*eRHlHnD!vg zis&~S@T>^g`}Feyc0QOKiDdx=oM+~P33@Z~!Q@CR&gE4v{Z<2B^&<4Tz;us$-h2~( z#G98-6!gf`FEa?ZrfWVn@)4D)zwQh6&FOa-u$|;hg?;nr7%a<{@0-(aFJyA-%b$ur z$IK)65s?(UZ{E@GEnvIJ3nJhOX6}ziMquex-XTE0v5>_nd*7VolxG+6_su){T?K3x z+0uz;=p3zZPhZEUD&d&S#IiYN@0&-NVd_}EhfKe(fbAh8w0fC0RfaM5(^JDk8!9+& z{{OM}ZZVc6S9Vxu^`oj!vzz24yV*((Noe?LUOqDOJn~V^9+KId8MlVhY^z%w_{WJ` zm6z3t&PPq&tlMn<4J<>}*n%a08PZ7p7#Q=R55Z%=hGD>l4L>jt>>q!~kFf>$Jr>5Y zWRAy@AdF9*d(Sx$an_DlXYZ`042%Q}yQ*SE>>YbW>>Yc@J#ipLTjO=+48+#W;}c-1 zo6B!5psd5e4c5)6zFf;n9!{TgspR(;l-kq3Pb0MFD5TG`SMpm7=v*a8?D}Z(Tt)Om z*)I zw=c^62JOYTPN}_W&lhHrw7vS5)|LED1F9GUd*Q4cSs$nSqc~ug1u@n$%z|7$x!KwA zIijzdS36?thw8;q+fms&vYyTl{Q)RW8PPBekV}ZywRsx1qR$sjJDA{N4hpCI^zb zE10p$lq{~x!{~o+Ugcu!HcC^vZZ0Z8+Eo4P>q>rC0acMZIdOze*_HfJdVDB--Mq@h zNNlvmblp6v3~6I^-Mo_DS3osnAgdR7llAdbHw;(wb@NJoX<@w1^mp3hI+J$(CmY%R z<^18r<<-%n^ON(FtCQL3$&cs9N6Xpe$?V?ge6jV$;s@*he#80qN`7-;|K9xaYHNOe z|KfOlxpg}GV19YDv&|IFe<^$B=>F+!v6y@(`_|;O(~J9ezpd_!vbjNjG2+tdzntEn z{-pi?q2B&t9d&nv$y`VMMrZwRWv@+M9cz8JE6Vl;-Ni^ur~7hxW4b?nyNloS<#Z2= zzh`>ghh05xyNi*UPIrUdL)}Tc|ABtJi@CLuTJK>I{TI8{{#y3xPZ^QEMnclyl&M3R<^_O2~(CTmS??e4b`|Ew2{6+(}*M)^( z!#+Rkin6^ycQLlp>2BEPL)}Tc|Jp5WTra=Za5uHjdnG_WasJEQZvT4royoVT-9FMC zWq-pQi1D3%4wlm!odc2whW$^B^OF0YXE--3*PL&}&L|ET^cQ12t^T_GU(lblzxubg zmHdtasv&yk`?bh8ZAaEj`3@-_80JBY|19$$msoCiete!}e@^cg#CY&E(l7LU@KStl zb*|wTvhPj~$G8TeM~V^pnULRiFwMkrdebxE+g_~E=;I{$tp_1a0!30kPT=z($*w3C z7<3;SBPIkM6C)jvHC8jH=r*^lj1XYSxvstG*U1zsG>;l6A-(joisvY^oc|zIvVI zR~g3ZOxI%LI+J$R-~N){Wb6PvH0265St8T1!pFs=UbS^Ng|_j!@&&mHo64)Dw+n8tOu?3s(`%C+%h zIB)KXahXnc{>+us-M8=K&|QqtINig7F<*}v=sxV~aob&t*>t)a>>lb)+FgIvmfv=O zXHIA=ceS%;?PN!k&GmZAFFR=T=IibWy-9oP{hR!*1Lxm{rDFNckU+>@LHy*gXE-V-u_W5B~u8kjqkCR_|(CKd2=R@5|yX$Mt@_P@k z=Il~6k|>iRd$tx$4&(A7Gt7Y)-|6Roue(#ak~}c%|6-h%-2XftxISOA&x@V8wifpQ zgZ^Tyr`2D#{|ow)_Se^&<#!-py%~_UuZrHAU(L2=$FsAevng9wraPo~pq~f%H3-W* z$Y=c)9KC`i~f0)(kB#XXc>^D|l zzV04Tg|x4}7A?Q=0Bg~KsP0N;YtfM|s5PwDResrFoUVM`UDTDdtG)&;zvlpJ(4C+- za%b!AiEb$S8gv$;v++9fb@!;wq@DE@X!&giSb+w@dJ#EWcTe?0S=ykr7=bCY=Iicp ztw~$|P4yiq`OSyVHFu;QUR<2kKX+HV-n_dx^-A`o$rt*YQ*y_dK^rlmN@%m3ju(`) z&3~@9jhHk0%+NNj+mz10+nq7Klzn0Hvw|_)p4>H;88i_ipr|Ix=~#(LoBX?an~1*` zcyX0lJ9>Y9^5EU8MVp{{rJM85WM5u z|GmjmZ`J>Ib0^WZ{!Gdn3+Edt-+#DSCLr+JDx13+)X|O-mn(6F5?3m5l@hnB#O*0@ z`%2uw=AO>(+gjj`7MN>+g%((9ft41xs|D_9f%{sZmN~hWIk}cOxt2M(mN~hWIk}cO zxt2M(mN~hWIk}cOg_b#mmN|u%Ifa%vg_b#mmN|u%Ifa%vg_b#mmN})CIi;34rItCR zmN})CIi;34rItCRmN})CIi;34m6kb`mN}J{IhB?>}i>^r)AEbmN|P` z=Im*iv!`Xwo|ZZLTITF)nX|8D&c2p8`&#DgYnijJWzN2qIs01X>}#2Gpk>a1mN^Gn z<{W65bD(9;ftEQ3TIL*RnRB3J&cVT^?wi~@+uLg7jvAS(k%byrs*#l%xvNI*sge6? zq?$uJY7XtFIkcnZ(2klzJ8BN?s5!Kw=FpCsLpy2??Wj3aCzh2vSe;qcBGnwKbIaN> z)f}p`%i1y39IErn+A-A}sx!>WG3ykw7OCb?on+RIspe3fX4a0W=1`qz){d#>P@QVl zj;ZEQoorT)S!bKINHvG*e6w~;HHYeqvvy21hw7ZOc1$&g>a4SJ%sTC?MXEVeC!V!q zsyS4rp0#7DIaDW~wPUI|RHvV{W2!k+C!m#M)){CmQq7?{2dy1b&7nFAtsPU%p*jz( z9aGJrIuorNvra{8k!lXr$!P7EY7W)uXziG44%G>1?U-r~)hTK1m}(ByNonPnbyix7 zRCB1#OKZnebEwWtYsXY`sLoAm$5eBu&Q2@Gtkct4q?$u@f?7MKnnQJpT05qiLv@l` zJEodLb(&f`rj|pwI!~Q%YdMr}YdMr}YdMr}YdMr}YdMr}YdMr}YdMr}YdMr}t2w03 zQ|IbDb*|1+=juFluFg~E>O6I>&Qs^=Jaw+lQ|IbDb*|1+=juFlUSEOM?YFr)Po3A7 zptWPF_o4b4w02DOK2%?X){d#(htzrMyuJ&qlvbUm&gb$-xt(8{Iq5874c1$&g z>g&?lG1VNZFHCF4RC7q3r_Sp;(@JU8dFs5rHLV>}&7u0uC98%}0^ZFjOQd)JMIOy^=x=`n-3-x{KLY=2B)OqScou@9;dFn!apSn=zsSEXe>O!5T zF4Xs_3w55lQ0J)&b)LFV=cxOy^=x=`n-3-x{KLY=2B)OqScou@9;dFn!apSn=zsSEXe>O!5T zF4Xs_3w55lP~WF6)OqScou@9;dFn!)r!Lg@sS9Oy^=x=`n-3-x{KLY=2B)c2_ib)LFV=cx;Ip1M%ysSEXe>O!5T zF4Xs_3w55lP~WF6)OqSsou@9QbGjF4cMJQk|zR)p_btou@9QbGjF4cMJQk|zR)p_btou@9QbGjF4cMJQk|zR)p_btou@9< zdFoP~r!Lib>QbGjF4cMJQk|zR)p_btou@9QbGjF4cMJQk|zR)p_bt zou@9QbGjF4cMJQk|zR)p_btou@9QbGjF4cMJQk|zR z)p_btou@9QbGjF4cMJQk|zR)p_btou@9QbGjF4cMJ zQk|zR)p_btou@9QbGjF4cMJQk|zR)p_btou@9QbGj zF4cMJQk|zR)p_btou@9QbGjF4cMJQk|zR)p_btou@9PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1? zr>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tU zdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSk zou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>! z)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$z zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>Pnrb zuGD$zN}Z>!)OqSkou{tUdFo1?r>@j_>PnrbuGD$zN}Z>!)OqS%b)I@xou}U2&VTm# zt>gLfu>#;yW7(&dusNhlf~rL%TG-2WE*czo_zUtpPGI;d*SHf(bdC8S4WF? zXAkF#$+Ow*yH8J^yi@<C9nPUnlQ?_E4Nxj#F7^Wtif_5Z#3g#>JIHM{>_C~8co>^J^kBTM!A`o;O;YVvgR zKVj#;{gaLC>E-P7(R}h-KiN3^&7W-Cxjk7aeE8gh(~Env)1$@J{Ndr}?#(Z+w&rK| z=Euh;=MNTJ?_XRVAN7Ae{9@SP>f*im`O(z}59e;P&xg$(UM%Vvyf}9oeKu_Lz1auz z%Qr7B&$jN(uVzDG`hR~gJ3E^lZm&8xzkhK&zuY>#Mjvf&&91KI=htIxE#{{`-1^l= z^UDvm-kV>Z&rc7(!6WXS%obex_vT&yUt{{eZr}7y$={f5l>941GeVw_XSEFKMRUHG zw_~KF@-WRPB{xcMzQ(-OSk%hs*n}tNSM$rW`SHo@YVNCpO#<3Cs=NY)@>E_^=}YOB zO^?%9yI@Mb%C(i*zcZTs0-_NeP4%FCg8i_4(>vnXpmmb{0-_lKvEOZv%$kzQ!!)BH z_JdIZQzQ0AQ8N2?wg8*gEJf^x?Hg4P`**fHl_TsAS~fi*_IK@qDM9Q9ZDsc7quDPY z8qpE^2kjH=hwYo*5zhv#lk67|%?OD7ZhK_blvEz383nN)j1rg{u|JBE*`JSOKWyKq zg4mza><4XS_7|hsFCZGx5&H-26YPiWo8A%62Cb9q7ZA+|i2ZJR zkICyzEhUwQX+}Zp2crb0M(mHGWcC*$*$>+{sv!0kH2Z^=O^=BEUAtgP5c@$}nf>Kx z_6vwcbj1Ea`vm)8`=)oqvq9@5`vpWZ0%E`0-eYov{UW9r1+gC>0#hUQM^Q5S%aQDd z?Hg4P`%9YrLCdB`#Qv^bFeQlnpsmdQYBc)=L?b$4|Db(>{jh!0J7WK!b&~x8q8S0P z-)+ybU&J({Aoc@9U~0tvC`x93HIn_XeWMCue?_xDXxa3L*x$7arUbDcw3XSvJDU9h zq7fajf6zX`e%QY09kGAVI>~+k(TsrD@3v>zFJhWe5c>flFg0R-6eY8NcO?5^`$iSS z{#}~=LCdB`#Qv^bFeQlnpsmdQz0vF!5RK@F{e$)i_QUo~?}+__)=BmYh-L)Dez!f# zei74*g4hobfvFMuqbQmEdn4Ho+c&Bp_V3Z`4_Y=oBKCLff+<1l2W@5c?~i7`fM`TV z>>spGuphQ>dPnRZv`(^LKr|yD_Pgy__KTQi6vTdj2uzLGA4SRR-yg|-*uGH(v45Xt zf6%h&5wX8(7fcCaKWHnn|6nxx1wr4Spmmb{0-_lKvEOaavR}kB zqagMJL||&f{wPXj|G`N1!}g6Ti2VmN`-7HEkBI$UyI@KX`$5~)MC^dl*iO9ITUU{< zi`iazSb{Xda{y-ne4QHM8qoC5>EJB^PNWs28KX#vgcdgZS)C?iv!rYfD8;*g9vKIfvW*PdRjMEAmD0 zJ@ajs!+T!7f!Zzy*TGP;<#49@W{m~g?gldo+bs|Gz!n^h1=<0Wz~v!Y;KGR49BRi? z!N}42op;n|A#8bwHpzi*?Y7;Hr2vt3b9el9G|Fy6AUTtL_|%R^*)4=E4-XQ&1m30{ zQ4LMcsyC|3&eW_5z~$kqoSyyaz4mskrUvlx@S9^t_VyPGTH+FjC|6Fm)faOq0h3lG zfVmdMt1=2&9-=_6i#%O-+G;csuslTJM=*k`#i-`XLlm!dy<-lJhAa0*HAn09`8~UMFGXXrYtBJMv((5QbL~gt?lbRg?gc zb{o1V?Ar#T4T0oLhE9r{$qgyD*M`WqiDULa?Arhk?<>f+ zkznlGqF`^YLB5SeOI!ku!?!Hn0}?Q4RRUOVm@}Ay5r3$#3iz>a>wYo7`nCuF`?dh! zYJq&4*c|(|t~ur)@@-KF_H6*-@eTR5I0mm~_fr2T-$np2uLYoMCB*9_j2ta=Qg}xm zj26Q1DuOUqGvwP6AkuC_7lnP>V6-8SoXOBhkuw>>uy1q4^t_C}10vr>LDRG9jq2JE z`8ILP9*BJ#0OEZG`8E=aeOnal?KQ}^v1o})z}xti#d|;kCap>U>kWf_oA@)Nm~ZQT z55)Sm2mt%G0N`qYe4E%D`?jt*<{J0OIit`L;L)uV(irJEMFX0mQr(fUcDg zuahuxw9rZ69eFTX2*axg!d%UeZ%cqkyA53w_HBdFhCp&ALnlSfWC+8)%@x!0GX4&T zd>aK#&#E`7YeVGQ#4&py_H6)&_Z8&ZNHF$oQLwkyAm7HKB`yJP<69Q*0STD2DgmrF z4EAl}j|^kJt^2hf>)Rp#?Aro>s|E6HVsq@1iX!J zS-b}%VA84tu--7(w~0R?j`_Ckx4f)xivX~13jnSb$hV2jv2W{|V-6zU7KLEn1|S~a zkZ+4)@M?B{Zad1i5kSmq0q9x@@j3}3M+==4-jN5Rg)qE|Ak5Va`L+azwA;`{Vc#|w zZ3rZ1GIUbpOolM*+gvd{FXQik$hT3@^sIWLx;8|?Esnvf+5Iv7DBnf^F|P%nYbC_%B#ay_bW(Un9*h>k@G62ZS2N_> z5+KrULl=d8+hDXIketcTNs%)d!mw|1#q_+4zXKxQMnTiF>W%8!5cxK7%pQn+8vx>c z1^G4-jD1@a?Cmwkx3Or6OTgRsmc@HO0w%3W0P78deVh3CkC<=kzQe)#wg>?GwgBL2 zfqa|T9Q(GeIp!epZBYpJZ2;o&4f(b>2CruKQyHUt8v(?;7J#mm5U-OkaZBekd*C5}9tHw^Y|;>UerzODPp z4(r<@0PNcWfU5=aZDMom+q&kMgUGiiEG zPQu91LMMfHf0FxCFe7Z&|ztBw*621hC#P z*tdzF?uz-g?i*UHZ;Jr1Zwmmf7Ra}W&9QImnqv+k-xh^n-v%Ha-;i&MW0r^e;Ohk5 zN3lfbrbR9f;WvZ^pYK?`g5?@YhhxVJJ=Ga0XNX%KqR;IO=kyxqYNtW5_8a=DGt%-y zBs-s>zhdVz#4Qi+ra0=k9Dh61V~M%-#&vC(TpI!sdmN8DYb4%ikjG)s%fr{mdG5#d zwjUE*WEF5hzJ-NdA%PQDDZsgw6;^2!wS>?9)pqe%;A@3GJsLwS;lqC+!qo(QI#q6u zJyPGk=Y;#=9ORMW7#xWpjK@86K1hIs_3eIuV4UwGkOIpEq-!n6@+6KOFZ5x;UOgBu z#0jek;#}?Es!E_(`wcyq;QI#S4Uz18hCYm)&k!g0K3CGn9s+$Ig(c?J8`rfZ^nK!} zJyP&}0Ezb+==(^t;QOLzZ~uY5kA;h@0$#=U62h*Kz=^9A;ClO<;kZ`(BxifxgPyGW z>;^jUjYSB)FBrqs1o}R)z2N)0_Lz^*_eC*+?*kZ*d(ijALCf_^#Se!cAYGa5EY)Cn z_?aeJ^m=foIQq}a!xvZkNBcR`c2xM+%frdKwL8G@@0Qc&vZs%)A9k*v!@eDTG&}p3 zH?p0?qu2MYCueqkbT(VOxApe*v)jU3)t_o#s(#_{BX^#hJTXXEaz`eklMoF}K7 zgEzeIWY64f8fX9dM)s{S(k!knPmb${y%(qT%i&wi>)e8Dkx{2FwQ@gy_*^IV#uJ{{ z*aXh#qZ3$Wn{5S_Q15H5dQ!u{`0W(Vq-n|$Q$m5txNHu2_ou5MZhwP`oI zy^B127yhN}V~ck$PtM;vn*Hcx(X7qMt(Tveek{xG%^#esR%`P&pU9q{pPrsPT)55Z zzdWD6A7Ed|?#$1R`@Ujp=61I6PQAN4dCk!`&X4Cmy8c7;!QzPsd>-DN#JmW`pE;hmf4WU#wPAtu~4VBDDsF*l~-KH|ld-QJe{2OHxY z%KeM;`?IU9CZD*d7s^v_&F;UqIXVgy1F$;Kx{kk_K)c?6->m;*BN~JLOB>nV7>~GC zpc&-p*DuZ&SD`GE_VoJf;qChL`o!JfA|8U~nAiHzEYZAoM9RJega8^cX*um5B73kkb5 zMA=kWH`MXd1MWPV-M;(uq-|?S@)SMQ9Cg zo}+oa);3x8+anWvzw++!QPqOKk2W)yV`^4RkPoBHen(>b|u9Jq@ ze>!r3yj@?Cng4h`xxN06ae1OfS3&4!r=K4P{qgLhcc1r!p1zoUWU-QAb9F=fs z-Qb_2X~_{^Q#lk2nAFK5rKJ)^^ysX8F2 zgR5s&{W(rvMH#_C1@p1#O7G>YI?u9pr?;q8n4PB*TeXk*!R+#l#T(}*S0}U6lONBI zzyEMD8UEeoa}Q20?#)h*7FY9!Vg&U&8`*ArMb!hov8cO0ncX{`JEKlsgnzyHxz*L9 zvwQR7>!BE{~7q-Jhp<_RP`!)7fG%c_sVOeYBA?otCHnM%4 zE_GNfXft?Y4nChZ-S2ezM)tMI6zYVLlxF{xUb9@%?DOHAKfGAfl)5-~ZMEC!^|fp| zc?If)aFkXx_q6sZB&|LhYUNjSr#qcOU(UWb`8-r8fKa;pO}#Fqq|0C>^@253W4E{_ zt_o_o1s!L*M+@9Fj3mGngy69g%) zYvIzaHvLYg;9gcuijfL7UAqjn=(=apwZCDe*2(F@)k&JW?0?bQbzjmIy$(CD>$RZG z)PJk9UX>M-9ol+b2b7+M^?D%bNv&7+Ufoo?6unL;egBW-p+t&Wj&`;sg{cMGU+Yto z*y(~&_8%CO72_IorO9QncPDVYAOC#E#XyZ`GO-_u|0?EG5x>f~UAor8{C1;su3|1~HshEC|Gb5X@d z1XTY_H$6X>eRFbGH9gI6nPnRj<1D<5f%~BARZYhiau<}H|5t;Z#UKkiV<71`XDoF< zDf&Mc6cr<^S61hmAg1;9l(juy?dZ9aZB4!|(-VOy1^-)vf?`;;etkWrU}M9BPQ9({ z>yxjN>NQYGx&P6ioES?@W7Z2c&DS~wi)?4Il~gc*Q%e5NdL?r)e)=k>0g!gI^l4adEs8%B}{7l8aFjawvkV%`zh$O00{=uB;d_ zQOY{UP_42~7nHIF@0p8H67rq{vvem>@+d2JKq+cjvtmp{ty%9FD%?0wuVt@ucOoo7 zU@pc*6pfvOsFub~7p_ubux~ia=3;Dw&ay5GQJ_PfWmDZyN?Rx`#!D$mdrCL^GI?ed zRpw(T!&x>LV<>c%1+oU=khG4;Gj4hSmjRjKEL(_i6*|jCN?C(9F2pzsdt;z%vk+--EOkIBYBCUpQJW@zAC0(pIie=4~VoXG>S?{<> z2L1y*yiS-Iv(UH{<06X2&Oua5SEma~S;JYj6k{WFmURu|9qZO1D|wbpbwep_ILnq| zyp*D}r*yL~D=0negi_dWmMz5?3Y}$vd4gz)JhP@dq7*lrWlJ%xLTA~iyra&tL{F6J zhO=xb##(qA1E;fbXW4v9?t)U*;*G^P3wvWA>6o)DRJG}VQq*vkEyWlMon>?MH9+<( zi@}tFhO=xX##QJn8_3w0fIQ10D5ad?EL(}O6gtax5(c|3d6t#nl#+(CY$e7|=qwAQ zTR6_ca|lOsd5LDqLwu)#zfSb^^T!r;6G5- z>x5F+LSr#5qG;?KM2#sKVHyT8Q)W2JR$^>~&a$pV-id77SvJ)TWnlxQcg1)qMQKmz zfg!|Pl!Etzkt}zNcf}YAon?V}f@q37%ceV`ENwW;?uv00I?G1o9Z@YHi?`|FETPA) z7;E8e3>@#qon?!k;#qcAjI*#e2Ff<8nfAs~2Na9 ztI%0Ckg>5Ld6q>`N;$(>c2|t0&{?*VFxY(=DW~LtR~RERoMm^#_z9h5fdHeQ4?|~f~?I7C69NpE*`tGV#Gu#>l{S2$~s+8 z${NnHdt#J?yyw6yW6!d32b7|Qz<*DSiKsQ}9YeKi*6V~)*s^BDxQL>$bC%W8*y(~& z)^L{H6JsNEmUUU^9k#}uWmDZyN?Rx`#!D$mdrCL^vVzjXPAG*9b=f^JhC*jqV4fhF zBG0nvjwrGLayDvsgD2S+T z;w&5QgtDt4F5VZTCge~AS(_C~#>KHN9=oz)#6&6U97MIsI$co88qTu&Vw8lu=fEu8 zS(cHLy<;qnVp_9eOhm0&?-)u3{sTR{PM8_91c3))Ttv~>If!a$>~ujXYdFguh_Mkm z%esc~PGsZGvZ-z;r447<12JAoQQA|w*_RcR9(F=0Y&gpvh%pp8%L4NR(G+=>O?N~o zZaB*xh;bD<%SPp0MOfsSmFS65-4J>lh_M#l#=!Az+*!8zDV}8y#5fCkW1wuan#sB> zRJG}VvZ~=MdmzSG=qy{ovn=_0Sq!EWG@NA*#JCEbWdj)-6Od0c;bq=FiZJjPCZT~lmZN*RtxzK@Ox&tlgLgfx9 zO-<`ojEtyt>m5d=+6H+n7|ep#xjPe?Y2Aw95yfQZJSyp{Ww_I|r|ti3VX_z_p);*( z0`Fut?o6BN#??+Na+p{x22Lqfds;WUvx2?DPAHB4&|+sXnnI^qV4xtblDJKtcGCm6 zGm`y7gXZ$*$XjP&Ba@J=;oq z4;f{|J9lTo6eI75@e}eZf>2}5wy`c;twj5pyr>vEQQA5OQ!?@&XzO%AX>0MK^5@8r z7looWmLmAV65imo?Df!{3#8{b$)neq7@UOUv0-{^&S5+X691ooz)trKcsDkiR*OqKWwh(ht=xpoS*E@TStJ|i!p|rNJTK-sh z${TrFH)~kI>R~68#+I|K{MB-Fwgm9x+)}rdKVOc{wo!pcoo$JpDBUf4 znfx_#yqAH)+PJgr{!jC4D}Tludu5<)v!WTZus;ui6LLL@sctKOzZ{)y_u<)AinGX@ z%D7W^>N3mOR{l^qI@<=qG*%?fwg~EKAx2e}v#tCsa&)%sL=5&|@@y-?DJ?B$+ft0A z(AgG9x+;X<9#0U$WZKDE@h_%SybCOt%+cBQ0G@5hGb{#E8d}b_@`uaO*)|Ze zu_Af4MNmpR%h^``HaR-mb|MCQFnP9>;FOk@v#tD5a&)!@VvefYig-%Tzi+sEEq{}I z=idDCYHNOe|KfOlxpg}GV19YDvmHI$em8sO=>F+!v6$S=_9wfi7x(XedwegH&cAQb zS-fw)HA32Gn0(mH#D2CrDaX!4)3-YhnMq$Ue&Y1)51vz3Ch>r)R=yu7`#p<&#i;4) zlzp$oXe{vU&WgKPImvk|c0Dljm6_JB7$tG**E^r8*RR(LrL%?4^2f|MK09YoHC3HH zD1HCHz~?s7fNT# ziBJB@IXdxq{&+2I5NJt>9P*GyjNr~m_Kz(*7jLBRl05HqUxi<9bX5OV_S)ptF;ow` zall!3bwRq`0F$xpBG>~FLrL^>>KZe(%BL_ z%bzueUPh3%S*@&>iS_aL^%Z~H$LZT2NvCRkojxdiEnZdr=sEPN1KV^bSN7~H_dw}t zTEF6*bZ-56r&jiSFwocQh0@uye#NMX_T3jFCwTUCt%IT<_UxPL zhtk``Z84fka@*6p*{L<$9`-`%Y&rYxi@7X#_60WT4{i76SF25ju${$FF zXWyvA6r#&|v&(9b0t&wM6jHYz7oPFg_qQkRqAZ24o_E?Lslzx_KsQle?c=qjN4EAFw z7s;N9rG8!gO?M;5ee!$<$TRNjD|X;&BxORf|Im=Fk-vHVa+9rjcyV!BmwmlO_|-1+ z@+Y&;O+M3SUQlC7gMZ(mffymZG}NG3VqfXx|6KN&$)^SRo2HcXf88Rz80);SioA~A zpPxK<_i8cylbzI`%3hp&qLDgiLP`5?>ZQ%)ua7^s7NwcB&vv4IGMh|3=8DSCs^R8M zzAaYD);Dj??#+*nPtG4Kw%)(EJU*KDf1VxB&W_Hw0=IlUr|%EW2hBe<7>s(_rT}FPQMrq2? zAv4?+%eTc0&HAobFX+RI#mUvl#W}F36gJ3r$f#~zJKMArb27Y^%c)5l(w6^?#RrI) zm@lsCJN=~lVc*tzYu)d>!#A?8O{Ts3b#`HlbS))DLnhtCSc1}RIo0Y$+Rfw>#Ykh| zHTv0a!r0bujJp${l;tei$zSc}wBrmgRZqb(`zMAqCx4fF57bnnmor;q{;6&iemncQ z$v2%<=-u?B`l76B(qD{tl=_qAzPU+STE5$t?}#xG%7_X!yAHiu>ysH#q9@AshHP*y zf7x55`;yHDCw1pIWV5>%H&MFx1xKshhdn)hyNl73O82D{FsS(u&WX>5%uwyh3!_PV|h-&$Ym--kW9HWs;W@$d3C##OrO{d=f8$D!=k zZmD0q$mK7NA0*;>uPnbRjEoWl+i|Y7H3?z2r0{Tpx=!deoNpCSOROl^de{N&YS|CbkH&JW#z?0#P4 z-q30NKY6n|+6QH4!=2Au{>J)9eV6PnGa~x>b`;OMhC82mA;x-NhrV5iO3$X`ozFm9 zcQ2r9YtmJW=SJzeH0w&*)o@ckm%p%%Zt8bd!I3t3Q$NuQWnYWV^7qw8>%5%agwCX$ z4PVmFjY%F-sS#R}|rt(P~WHEC;0UPAt^I?7A9zo1ymPk*@e ztB>ZFAFQvgw?3J%6q~$Y1~ld6(|-Ch2Eym}7kx;r%o9HUI<8~Ka!sr}F6f$Q@X`@h&1#R1d)FUERG z{pCFS&CQF?i_B7QuEaVksy74H;_Mw`H?}Cv4r-bQ`K$AWd7$g=q>8@X#oWB5+AM#Q z9#xy8a(h`mvZT&;$S}-QcbC6iZcsi}?6V3?z2r0y-Cv=;!h4E#Adb=shAA5Y?Mp=axG3u9)*f zb?EMk9jNZk*oZAw#|ARpGuRblzmfXVb@zxWqK=_O`rykjr1iM>Xhf_HpD+)}Rx;xONodR7>4`u6#4IhZ{N?f&@lgdD2;xQLKi^$)eKk9n>`9kgQ{7OOw$$DC#0YG>)^y!H zt~JLPQ(i{?20qG5?i|3LLcS>2MCk-v$r)Su4F6Z(_(x12ln#8?cSxdfRR zpG}^{h!I>)nCh>4VqB)uoj!9Vb^I{wF2-n-?qS84uEz{?ANJ(hSMri3yUX9(SLv>| zd#F2UcgtB@{sKQbb3$vmtDQV+C;OpnZqi$m&)NySNqbxT+rIcc6Ux60YsF{m zmAQfLs4vRyCjG^zPN~1%zYp~%?Qik#`{HjLsJ(7jE!X<-VNaCpO}dNuHI?rAeLmEk zw7aF|EPurx)tue6>eWd7d^%ZoCwrspZ<+%!zEjTuU3aJEfaHN`{}bc9)c)u3!1eix z^PSii#Q}@{@>l+q`g1nZI7wlWmZo~M{GESPZw9PInXEImq)zuo@xXHbK>q%}VIJhH zE%`fma(;ZynX1k5+W@HA9F@<@@-YTve+o{IYMTi$GPKMDU3X`d^lkrG&cpIs12|3s z^+P{Spku^jPYkV0xd8e70hR6}V#I{*q}?rNV);b^d?p4GyKw=X3rO@s+1#YJ7#Aw^ z9+3-(>P_0)Qg=TPV?k7h&f8iRs=JfFcaQc#+1a3PA-|z8QeV369?_Syuc7W<$Zs*A zT67?*yOPP8bEF5#wkBQ0eBLNs>EFAHx{`J^)SwIbod#5Q?MJ6U&6^g`L!q_Y^E zjn+Y#;C`+5P7PEBYwWjOtaji*P8}bsxwpe#X zd5MClyGysF$b1bohRX)ay@O&~jKP%p(|LJ9PtyK|+;p)m#$xE)A;|3dZ1T)S^yJ$3 zahx{``2`7;?(~@}sr!duclmt@O82m0OxN89x(|DL{B{>(HkIyryN9}yb~l_k3;A6M zbmoNCa#uTf)=u_A+1#YJ7{@8}rq9|5y-9mp{QHhrccuK>uvSdh-3Pj(z9_q!^cSN# zrT%*VKGdJIzs0}H?@Ca6U05;J@AJc+DBGKK7h^k>?)rT`)Sa}uq2^rdh;bdNIlF7s zEC2cVbh7SF_D0#?G6(V-6Y4pTD{VJ9H$Fd1=NtK*iPZk*@xb-@H+a4i$8fn}(_f7B zl={mIP@HcC{R8_O>dl4x-UO;QgO!W3cVu0e?$5Qic-Aw`gBbrA=7IjbJ2?-&-Nkt@ z{NBBg-H*mddn|NDD)nY3yA7X zu5(KrT7FLg)uD?UsJln|pzLhYSB(8e>Py$%Bl?o|wbY{JcOy_OI*`>}$zDACZ&A`NUW@t;Gmzyw-HxJ+6>%Q}G;T$xF*`N1(h!$<*D+`~%e&wZ1L- z%kM}i^_N%q&CQF?3)8tnep3RSxdfZJKAW)^`z_Y=a5mk=xJ;!xedbE4<=aoEh}hdnV2ve{jX*;Kmg?H=k*+TC*2mfw{?XHIA>ceRse?PNcc%}siX zahyVLd5^!D`S5vQsymcoPL1+!!&)(2cOTds^+nm;^ce z)~LNMtQhO}`C(6#?Jc^?Z%e3j*YES8?!J9R|8A){%kN8|nzOrBy;`83PiHK~K7h^* zY5CTEQHk-LdJgE{yHo0sJh1G4@;ei${m+==+oY)t`D3kxI#8^+MKYi~?WI#Wc zr2Q@RX8FAdRBr|{W6-uK-S&Mtz=KGwZ%BeqPzV51db6C z5|4-xlRZ(oTh7Gtn-lm<3?%M-+1Zpn`zHE%{CbOVp+axE?he`B&mXzYEp=%5Jqc8Y zE^naj9_@p&vq@hu_8X}$U3ZV@OWM~`i;yBM>nbl2NG)Sa}u<*Y5gD}m0O&|2{rgaV(*717C%-E}?R8lB;b$50qNKRPdB>62094CRgsUIiMIiO@u6bnqckBku$63e%8Z)#3_K4g|N zvHa!)J`)4odpAYb-79q4!tIB$xk+y^E>!3}A{P*yKi{t6j%TSu%kN2`I&>AQyG#C` z{Qi8jZ`b~o_uUKm4T+KZa<*aLvyUjhG}WTzcOy_OIuOHMf#mPqBR#k_6uE4n!`^o0toqL6%>Bm-0Ii zr~(Zvpb%aE5S<+meHUSjtK>+V!vl+{i8 zi!qo|e>yKu=ug_;aPBDOHzm-SOOV<1*^I^5+;n<4%kQ4cZ80aO(w(lqCe`{d>@NNe zmeM_}7}NEbf$qbe7zWwwF2-ys-Su`4btmm^ICGZryAtTk39aR>cCzlC?1!?sMQ`~{ z35DMDSv#ROX>W^vli!n|{M)crOxIlox}&}*yPNbEqdKMjdjCGupR~Wlzsv7RP15rV?2WR&X%58r zPCW;7-JO~Pk_V>!PmJ?Y`=7^yvFnztJB!og3~0G`Q0|Dao>G6#W;dw6sqS9N?@ge3 zGcX6v-jVfnx<6_!F#O)Vl;4~%%mZC_C+C6VmZk1q%I{8~YI9U>FUv>P-T4kFMi}bu zWiCdBmYJaI?(9sEoUp`6@>>)*P6G8qKTe=y#AHtt3rxC?j1dzOkBAYIJyE(_&cyPY z6ZlLFBzD|1`+N7@pLX57lwX)o=shAA5S>4Aog3=zrTm@*szdL_>h5HHJ=zDwZ^D+h`uFY;eM$QoYSE?qZUm}D2eP^=nXwG}<9B%&%N>JKek)>>u5wM5EVYa3`S!Nd z-Anm>2vmdaB*KxqbYq6BK__~5nPbygjLt^uO#j|Js+(_5u@-BoK+Eq$pb9jQ){Dr= zx_hb{*P@R{YcT>FueH3KZ)7%n4w&*1^4k$8FR{ne-5E=MPN~0d8;Pkm zsQ=PbgOuNtKxZz&9InqM&s;=L43BKOi*cDsclykg)ScsyO?UBk50vg<#aR9}k(Lj` z5nMi)beG?jQ0cC>d#EF6cgtB@epdpWIR{L=me^hGWZga4k84|L&$Qg_FH13wQ|L{f zwG(=i_O{erQh)m16`?vY>Ha7lnC3x@{|xg$*WJl^Ao*aaHp}l$plWkeZZFG6*4_CIDMpxPLW~S8 zGa+aGZg_@#u9(il@>>)*P6G8qKTe>}{K=ju7Fc4$YGjOpuEIBQ+FrxGE`r#bw%H7(qGKYDfOrGGKBu5{VnGX z`ArFQ<`QIfeKvXKB6^~1Z_!+XX&8IIuc!IWp% z6Js`&?s~h2I+Av`oVDe5CD54@TE$)M9cl1Z_?ft|F$Q- zCqenQVHudNyAO0neNlEd=`ZHhl=|!a`%r(<{w5zM{@#_^>%xlh&F}NW5nN8#_IWY3 zQ|YeX=R+MyyIX$iy(i|_P|ew0t6sg;&!>~WcPD#u?JVy3mN}5$m{89FU3aJEfaHN? z|C8UDNbP?f4_u$GINynVQ5-PoFUERG{Z;$Fpg(DUOTAfsZvxeu0c&ygj;y=W{ZTwH z&4U>K8RmhmyOZ-k^1)JVmfxK~)#j+&UY3unD)Svuj4;iF7#Uh-g8scbI};=)EHR?| z76p!zK>g5<6X+aJvL}iKCf!HIhzW^D#E8kBDBUe*V)@Mpd?p4GyKw=X3rO_y*!7lQ zm{90FA{P+Vn_TCXy1V?I1gb;t-$30x+6QH4lfGi?H&S1^?jF&Xw6CQWEx#LqYSDqL z?n)->?vWlS+nRJ0rB_( zqdJpzw$$C_cOp;)8c6F!)Sn~3f{IY~fclykg z)bYcxdog+Ojl~aUmnXA(r}H-30hk2dROwh66nkcg>hFqdDc$$L)qMrW2odeB@}woXYGXE9CIwc_p7$W zx+~@1hP7h3@=DfY26KYOaQR@eXB^GNxjd96JdJay^`OTKs9G~O?ov#Kc7z4-O1iuJ4>@^`MrC!Bj(-I zb3oVKsX5>nW!nGbcP3K%pT`5&=Sva3V*eAzaJgaAUySvX`rqjOFX->{!t$dk<)ID1Fd-Rb^Zdy9L4q3&Moh(%miI`XO$#5VTzMR#)zPP=*J0kEST(xp_M5U1U?f3iQTw>k(d1#{Xybm^rjq${KAAn?-99x==_oE+)#(Em&FI{(!=tbJsP>ZhQcOy_OI*`>}$zrJM&d8=Iuu7OxN9`x{-D^ zRG=&Qod{Hc2GV*FxnyXv9-r#ov$REP`NfFwTGMs+xOSwi|7at-dpUo2ad~z0==`KF zp&c#PWwpi8&K4}P9aZw%5&QS%mseZ!^ZOUa^UJN%*$4B>qn&N0@cv8LGe`GNXN$$; zJK47;ubp1pzx!=+l$egPJ4sxL+wf1|Dk2w zi#fQGTKCr?yG!*KyH)>M_Uhz-U-e-}l-*65i&33M^X2r0G$(Ct@^)g5jq`S>NUrvF zs58p;mKl&=nb6LF!QT&OfaHM1JLaJybuAwJW4ykS?A{Y~17@t#h5!=4{%Pul)p zxTTNx<##9UruKZVNa!cp8KX^qxpNP%XWyB8i*gV0n3-u7#F$S%3(K2TkmQ1CClq79 z#fvMk9K%UYqKMvav;L`E`l$TGMshxYnetEqCbTHzv?s zG9<1Ssgu8_OZG!q+oZP`l_~V5@6aXmChcv>T*z-vz|6(YP1fzH&M3>9Wwwww7||7Ff0OoNtfteRKaVA~|1fMX#%-MTVY!*F;S97N zcJ=seFUD~??G3gMwI^+FIg870RKOD{G?=^U8FSGk?$l_RMRWOu3XSG`-9DidX>*IW zli#P{yxp+Sd`1@`2HK;}7!sLkHS#MJ+8HqT`{4|b9I$vi`K=0W*9*(dhCP4S6=i>u z_F}xJ(_YTRsiUB9R(wvFD%SG56|iFMF4;&zjV;NO-BBE{)b01gm`^_oeBGX!1>fdk z&d%fu#Mm#n6M9^5V*s{pFV2D68N~zB42bccb_R65K$uRF1D4-U%kNphYBexhUs1g` zznX2$j%Q~_XY5a-cjyi&E?8zle$~P<6MWsCoC%T>mRhy^wgs$JM+Nt?fo!ds?~!7K zX)eUPplL4nx;;A=BrhzHllB69|#uf*=4UzoeI zy)>Pe<@YUw6Eo1>b?wrx`Pk|;RM#EA<%3OgF;>)QJ|uiJ}Sk+!u|u;q6tUowO}@}yu!>+@dx@*oq>mV7CG=TN$LmAd=XWgj5p!#w8QRBn zquZTzzLb4o^0R_S?nVICgH_Z67J~z`N@NKuNG}0 z>y^$PpUr+|vf0=Jx1{V~kbPHv+X2eHoU*d}^DD0Je_KEAyYjmXll5%6l>TBj>7UO2 z#N<=%q`S?i88=BS#u7uRms3h=a<2cizj7=4Qu1Rgi>ulF_a;xhRsY}3BH6lr9pR0I z^EATuA8zg@LGXJooBKN4(T*CKtC58oS*nqh8o8@R?x~UcYUIJ@ftCr|dgzWGn(Lv3 z9$M<5l^(jQhwka2`+BIJO}U;;xt>kAo=v%)O}U;;xt>kAo=v%)O}U;;xt>jho=t_G zO@*FKg`Q1?o=t_GO@*FKg`Q1?o=t_GO{JbqrJhZto=v5mO{JbqrJhZto=v5mO{Jbq zrJhZdo=ugWO_iQam7YzNo=ugWO_iQam7YzNo=ugWO}lzF?dsXIt7p@$o=v-YHtp)! zw5w;+uAWW1dN%Fq*|eu;)1ID9dwMqQ>DjcWXVadZO?!Ga?djRHr)Sfio=y9DHtp-# zw6AB=zMf6{dN%Fr*|e``)4rZf`+7F*>)CXmXVZb6O$T~59q8F~pl8#8o=pdOHXZ2M zbf9O`!NI2Xd*?gb+Xn280h=4Jg#lX{u$2M3YryUqu=@tAfmb^QUhNoowPWDbj)7M@ z243wLc(r5T)sBHzI|g3u7r}ZOYv5I#EZ2`~;8mS2*NfE`0Tm!G_?74c}I(@Fk8hBMF(DmaQcvYv+_2U|NRVUH);~IEX zr_uG}8hBMF($(YEnRGqYz^gi!t{>OHt2&#mAJ@REI-jl|*TAbfqplvePO0m$242-k zb^W*oUe#%J{kR5R)rocexCUO;sdfFh242<4b@jM)c3qD(@T$(Q>&G?ls?M#dc-=1&l zd6jSLd6jSLd6jSLd6jSLd6jSLd6jSLd6jSLd6jP)cxA}9=Z1WHZpgRihJ1T&$hYT) ze0y%lx95g@dv3_L=Z1WHZpgRihJ1ToUrE>Q&$%Jrp4XSs_2U}$tNL2Heq6(TRbNck zk89Yk4EgrFzMHO=*^qC~>)Yx2aSgnx@2Bg>HSnswp{^g-z^nR>x_VqgzCAbO+w=OG zx?W}juj-5H`f&}ss;{c+$2IV(zO1ev*T5@7zCEw+tE*);-upG zysB@l>&G?ls=l|b9@mg>&kgzZyuP}wm)XFp`trJdTm!G_>+AY)4ZNx^uOHtNI?hep~~u>YMEPaSgnx@3O1MHRRiKL%uz)ue0lAHt?#x z(5@fXz^nR7yM9~)uj)(f`f&}sGUVIy`d+(QW<$O`uWz>N$2IV(zT2)J*TAd#cDsID z1FsDC+Y3X!y)fk43q!uWFyz|{L%zK*dKlJ)<+`K2UqaW%XDUMOx%u!A6$u_w|eO#nt5L=6}NOfBPpJ+0)C}>7)7N zw|=s5_?ti3xO025Qv2|^2d5YJW~WDstNFvj&)u6}UTw|K?#+*nPtG4Kw%)(EJU;6G zeE7w%!PUik^Yf#t4<63lW}gq6J-k@d^LTOYHu`MX=zFsd=9h0?T%K*+n_ta_!u0?C zV0LykJKSD%aDM;dcz(HcdW}BX-kM!q&CjpL+FHy{f4KFlkLH&jY`r(XJfEK)euGEc zJDDxG_V3NR{=dfbf8D<6osz#X+bH>0h-QR5Al96N#$XhQA%!<-h7RD ztFfq+(Xk0n&adW|XY=Ee+11=v2b%=6Z&Y~&3gxN1rqY+vEt?*vv39|fe3fe}vwvqa z`vpWJI-2T1`vm)8`=)oqvq9@5`vpWZ0%E`09+@>Im4|6YLF@;k1g1vpkD_Gu?`#1! zuUU%N58F4YAolNUc`8TPAGBn``z})tSPBHOfw2%KNux2HDZ4hC9^*t$$r?rQ3bI-r`aF0YV4$GYVoqKm?{n?2n>k_Ln2s58F4YAoiCu`-7HEkBI$U zyI@KX`$1co{ncpp3y4N^#Qs711p8t8rgy~tLF**@1w=CfV!zv-Wxt4NMnUWch``i{ z{ZW+6{%R!qVf#iE#QutAf6%h&5wX8(7fcCaKWHnne|I$d1wr4S zpmmb{0-_lKvEOaavR}kBqagMJL||&f{wPXj|L#ck!}g6Ti2b`X`-7HEkBI$UyI@KX z`$1co{d=R?FCZGx5&H-26YPiWo8A%o2d$Ir7ZA+|i2ZJRmi;2883nN)AOce(_D4}N z`}an&AGU8)LG0h7*&noQdPMB++67aB*bmyu?B5^FegV;lj@UnFpI|?1-}H{yKWLp~ zzkp~)KILs z?1$|eRS^3RX!Zv!n;sGSyLQ2pAohc{tBKeFrLmoOvA3=wUl+5z@~{MHgy#Ux0{A*L z!Zo1jpVPrx0-Q)INHa!}5(zDWq+lMX83*y%o7^=du=y@2UJqNDUzi8E@FkcA02+0^ zg5~n`UgycaOf+r!MEtI&2u}#7hgToq3<{7&d=8LW4TBNlt6|VQy&42a zqXS$GE&v%At_Bg*j00B#fb_I(u0X)m5QRln11#&VISW?<1T^Zv)zFn7T0gLkhMhtIA5-2CJE zW`}qQYHM~pJ3Bf%gyDmz}9u6@bgb zS2;cV)qCyjT1^e$<>5ESj_mC(7PQ1A5K*q2ZmTcmQUWHeN&s^$idSV6v^+$CUKe?~ z?zGitBw%@n!jE7CSBp{2mxm}`>w3o=91U3>qT~V?#pBy}3|`G{NH~b>TdSbeOaL*j z1)yss#OoxC94&NGXV3u;Mhjth6+xJ*8CpdN5NWrei#mhB%wV)3ketcTNs%)d!mw|1 z#q^wozXKxQMnTiF>W%8!Ftr*4AoguQSiG+w-$sJ5Z;OJxy$1O<7AxjDoOl>&}s^Z;Jr1Zwmmf7Ra}W&9QImnqv+k-xh^n-v%Ha-;i&MWAJKr zulZxhGtf!MbJAl_GyZzI9jw?)C;UW0rainNEsOVn1Wa0$0M;7@`!?}sNHO2m{T_(*Z4m(WZ2`d50{J$v zIreQ`bId{H+oBNc+W^Gl8}e;&3|`IdPj*K6HUfxwEdX6BAzmk8Wn&k#-xpDD2w?qYZ)NOomR1oXHS|eVZ$$=VkmI5cxI=nx0i}RM&>cw~1r+ zKPd|UTxKi0QJ z0NA$$09Om-+r;MBw{^`i2a#`!La=WG5RY%jx5Y7dHM>8^8|B*wAm+6IbghJVorIC2 zg-#0Z$b->B7+ysX=4ys~TLMJdZRnz~ZySs@1d=luIw^7{Lm2jLu9%*e@pnMv+bC#y zR=rVO8zSE(j@bjTZv#NQuOQz>g0XLlg1x;4`8F0UaS3=E-?DfQNWi3331Gcpux}H8 zLLBpL-EVnW-xdL2-xdH|Es$>$n`7VBHOCx8zAXyDz70S;z9HWh$KciM{@iwyZzF)1 z*8OMhjth6+xJ*8S-rj5NWrei^9HbFxn7E&SdDM$e9db*tfZ2 zdS1rg0g-Q`py^rlMs;n7e499C55&F=0P((pd>aYIzAXy&_8R2dShU0?;B9=%;yoY% zlU5~w^@hQ|P5eMW%(r#F)MtHL1b}^80C2THzD;b7eOuQYa}fEqCvzKw#WXVn|kwIT9t;+Q=U`!)c?`wH@HBpCa)DA?O;kZ)tr z5|@Ct@hyw@fCNlhl>pWo2KzSg^B*za)_sSA^=%OV_H6;c)dKl8u{ri_U31JqQ-34L;wo zdIifhln%#^7ka8QQqB;!JVc+{8_wx9&ecwXV(mBdRcEB-he&olLx082XNX%K-c51T zb2>y7K$GPyPcB=$HSch*R}(IAh*qL+uSk@MV-?QK6MxX3Etf_w`LyFvmd zu2O(=Ei0_jC~66x{j2TbvB1{~eR?#8Si*<@LWHXc`gE$?9($y|ea{K^!#T(!#W6S% zK^TvF=zNd>3G3Va0KqulM<4~33rN>mkmX4nJ6`C+guQw&UWgM`6~wvP!Bv$&vGyB! zFv0f?#v3Bp`3!v+JD(v=@O`eNkv#<_KR~)N+gYl?^6)cFwCMHVPI2^~mxnK| z_K)^+rtPTkua}3Db!&Hk;omK%&t*>^Uq9?zKZ<=j`gHd7i`hl;;p=Rhr<;Q(yYFPr+(0jXQ|+VHFC2d4&XbeoFYn&^-KVm@yperp zoIH!G%ah~!iSNZ}{f78f^G3HI-AnDD&mTV54Z87!J6lnyv)|arUN;MOv@=S)jSpMA z{Beo*51!0kHO$`RBX6%P{r0t?H#Mx?v{P&ICJ1m7o#f$5@h@c`TfBRDa{k`Y>_;by zX01+cz5K-VV_9}@{@`S_dYixbME3mr^z`K6!fjUn<@x;m0Q*9AXMTR%_Z3^xx3i6R z>iy=)Yrej5emwut^&hgox{-Z%oYCL9yg1&voF6~B-@H11a(-N!dohZXBbM~9-^%jw z@=cz)mU7cV=gk|LmXHKl_S8Ek$IaAFKbk%J-hBS>==kJpF}d~16VuK@FAa9)r?Ojj zpUD2zjjS3s8I3a#ceEH?BoE+ich(*EXea6rXm=%ej_j{(WIsP{!aQ%)iYHE>Fs?U8 zF+Y1_0zNjU@Q5 zu-iP7J$3iiM8DriY^3&-=z1|mct?43zro9SM1*JEZ`R`d)4Sia0!kJ(&!Ba`$!}u6 z$!+`1dXj7#MLW|wH`C2vzoB-JCzu^%V=C?-{YK_(Z$c=X{l^<)oz4A=^ZT=_ttKD2 zsF%%CZ_Vz%x2cF2!$7Tz#p;OdI$m!A^ZGB}tp8&px<&o^Ms_gHW3H8GhI#t+i}S@* zD2)&hPABb&`PswU^@;h3yTNHW`?V*tQX|adPAx}0vp>i;ZSH?8JTT~3xF5ddn|e!x zCU7J7jvQ-3#Iq8AXLQv5r{{|%iF14XADdrF+OVO`rpnf#rfcnPPd}19^Ug|;>to}m z2Vugq+3mYePoBI}r-lBF(JObY+I4UiNmzZZCHPB+pX^dX*&nt4Z-i_9`y1I_au&Tr zS|{P8v%^=DJ$&!t!O8vE>6;fx0(>aD%lvT{5zwQhrjaf{IX6SHvhA!oi9=J zFApg6V)luuLWZtf?;mm(LpWI&m$JqQ+qp=|Pv(wKH1pj#U z(Yw!kf=^$}KC)Ozu{zQ}ojtj^v61~N8)MUIzc{<9&G3WzGrZT1=Y~DP-@cXYCRyyQ zgI1i$o%X*rN1o!^ZEPg0HRZ@OPAg?IeW6L~jM+PF%Ienkv(x9Z+wVM_UCypkFm*1T zGISlwyt9f4>TU7*&^5cd|L!`C^YRnfk21VnU(Od7k1p$6S-s~q+qGa?Ihy)-yBpp) zeCE#WNqt0fGYz-1?`O}i{AWFa!<}T9Fcsp9>_6Mc?zBBRA%T)Wf*Tvm%cifzbr5+S zcHQEv10413?1k)^lk=0SqkQtzn-|}In0+eyX!GaU>FLqc?A~dez+C-9+pt*wc|ZHe zgUgFYb(U`aaB;N$+s)hm{HnM1KZ?zdt^TI(FZ-oz^2_tf2lK7w|Gxd`>}+=V!Q>)�iPd5?phzoKycLH#&Aff} z=xToU@Mu3Tepv40<#zeQSvju`_795tdAWaF&B|hTuh`qFcZ@4ej7Ql=11Tn1FVUtm z6`P{CEj?~YHoQ(HFXq?V*ww}5^|p}xv+H5cADtc5Q-68f`Vp`f1VNOv7rd8kt~9ys zZhrg=EfbsWg{>@gK9LlILL)p)1B{gaKBh`;rdjl>txP9Ci(<=^;olb;U=gUzbEjUhx!NvP#2Ri7jY+i585#x>0m} zQv|+K7DcpzqMs@eKzvb11g`FK{gj^IMDG>5k~Vi=+uta_*FMD7?JS~;NeK|X$^}8| z*+ai~inMQKQ^82-`d_ZVHvk9H_3cMV7sApjWN@E3xk3hrR>IT@5rE_j_V-}98SLA) z@=^8;VdM<1U~B$S6X7CS36;R^aqjc`^;zU{>vZgQjLZz57G1TV3d{P zMYl<%#;r`_G2Of8c$ys7iheZN$K-fP(P{cn(+PAVH|B!@{()A%1j7tW?^zBDU7sApj zWDss8SI7X-N|;(90+5`+KIGzNuy5bWN7*-oku&JVY1|CDh*m;s28Tdt<$};gjO-G> z#=)O#NPr~E0=c3qS8U(T8@6l$h9f)h^OwdgqcBZq+e$SWf+IUHiX7QVja!*eWEaQa z$gaaUt`&u9pj|7t#~>dD_|dmkza!80pkP=am0cd z1YzXoTvkK*vzS~kgTYqnvGdr+V&pvbfN}E};@Ej~Losq5T}Ue=H;+T4xRycW)9j7H zsuY^H;%QK#iNRR$wJZW!@r1RUzzMNdmx=O|b%!u+5*oLHjmHRqJQ!!?oZ)F&YTb&Z z04)!{4DxFIXHuLX5dJ{E6RB_5N;9tg^6)E!AZtRwuK@(2_#4_-gb)e3r}+_)U=r5O z8H@x;v=9kK(UC|njcbL5ksu2fHe>+r8HP9HD5RAYZpZ{uhy+FCny4_EOMqHY^;p4s zxZ`@Eo%9utJ{b3!w*o@%Ar7)a3El&uV%{SHXvKrxBZ*@L>;p6$!3+lK0pkP=am0cd z1ZU|5Gk~;0CKrr=WaqID(8zi00j-Ffp+g)yk8W^A&Z7%yh2-XOh!oc{h=SxuFp z1PIrRAZR^%sQ43U-^!+fk<#_i7_3zer0d&{k}ia$SI8g?gSARQKHu4)wem1^V+ z_Hh_DgMIr^GZ@0i8FT|MZU$XMD}dvod5Un?74xxx@}%K}hz<%;cF z+3?Cuz;LxUZwp9)TSlc@p=~SGXb6rw!6+-oi(*KraVyh!46gQe<2bGrm2W}2R+LfA zakV$-eN7^Cx_~WPNk#&2WLKynyCkZW5k+<+7)N$dupHT?VXed{vSZPB*9B;YerWTI zB3eO}KpfeLBd*Db!m|XXl~WDH{=pqJ9odPkTj`L0NPw{ChoFdmh_r8IL;fKNBcF2~#U%Fw_c2&R`#MaWmMrZ{?%x8^XvLbi*xf z23RzPwF`;d#9!M=SfA7$SVM$VucZgDf{B3cQl85{zol?y@}F|sQ)kzE2LSr*6@ zUAbcWcHXdM6EGavl_MfMp=~P_itM5g9N7Vg9N9^YTbWQ~7suepuERL46@_Y`T`S6{ z<~XtodS8GNQ=GdCA0Q~=A0q8r*^qxo!bs`* zkPF5ogJJsiqofOA=@l{vw_sc%$hVsBR>IT@84R@ok~7$cT-*%y?OXXM`-U)b2HkLr zn?V=RN=VJ%5Gbu&5ZZ{5U8RZa5+KR4K(6S@72CJ-hAo?b;mEEU5!nfCTd7cF7lq)+ z4nX9{PHNoBgd)2*21j-s#&NAER0Hi=QARb#kzLUHnnWm2gDqQ0MgnkTx2uZmlBiZj z6xoqr9N9&|a%7i=wGyMqjz!~L7oZ*bq0KXjXa!XQabzcsxF#nG&k~qcPBj$!2Y1wT zWGA+6r9+Wj0)+ho1V#KqqY^B`~d=YAE&(?x^X=PHf#uha$TK2>S;Jiui{}`&KsOACfRq zx<2HBamiqqzWpfaLRfl*48koKmk9E$=DU?JwL%6%t$^eV_8}KHgMIr}KFYo!jGRF? z+~Q`?MYIx9GdKiFD;I<|Vq~|ciR=;}$+AGM=*kt_xATTAn}FfSZf`_nC$w#)LXlk* zf+ITsks~{)aVryw?BW<4*>xDlwW3fBv};8f)f`85LGNo4p+F6`Y$X{9z>(d)DzZzW zS{YGfM}l!=7X{0aT^iO(j3PT0jdxvucIbyT&nTi5R0+h9ojBr}oG3g?U|KoVQ0yPv zQPYu~*t(StMRo}g_74yg@eh&qt!&6YBw?goAUMMWGsK z*NQT#Igadt-q$2Tff{VtN-`3FBfA4tWS2y>GNQ=GdCA0Q~=A0q8r*^qxo z!bs`*kPF5ogJJsiqofOA=@l{vw_sc%$hVsBR>IT@84R@ok~7$cT-*%y?OXXM`-U)b z2HkLrn?V=RN=VJ%5Gbu&5ZZ{5-GL^uOMoQH0=c3qS8U(T8@6l$h9kR!5s{tHwv`G+ zc2Nk9>;OcL?4-u6OenI8V{l~GVI0?rLN(B?6=hU&9N7iEuStXgHQ2J1WF%mD_)_zs z<%bs+r}bNb&J278|9v}7^q-f9FAvdcjaCoudgJ+9Z=%8S@VN)47x!kTM~kcZ!^7uR zA6`CcHiL=thgQ(x-z^VcShstBe)8bmtHt4q{-0YY{Oe`*H?r(!R!?NFo}_MGSU;Mb zA0I7FPwtD)Qva9Fwr@@UC(j=KN6+>zO+Rt?$N&G2XW)-#;E!kE|5r0GeJ*?Y`09g) z^U1C4+tW{HFC49R(PlqfOrFhd-+g-Wh*)k{3yxnY_o1a|$<-4=X z28noka?UtLL55<3Yu{gPY9m#pNesfARY6s@qE{BtH zb@AT({OH>L?(pkJ!+R6JW~~5YT7wCH|Kjr49egm6L4$T!uHAAIebU~2bZEu*kYy-u zl|;3o8Nka;baaT0yy7WJTcu&G#1`~&6McUfjNQkjo9J8_MYMt{fy+&F&Jagj#j=|y zbCtlfa;l-r&GmW;UPbP|P3VmH&K#vw2urV!MyVBSJ$osl4-l<{N+3CdeKv}l!M=SfA2ov^jGRF?6P4U! zni+Hvt%TGJ4uR6j1zDf#bFQ#*h1RWXc;!Yxl4SuXx^l(#t!&t`3D~j?c7pv5yJeKY z>K=*eBeUyAU|Xq1LzbIgU4>sMiB^smJC=d?P2*Ok@t9>Bh+?4}*NU=O-4ylhT2V$d zUv9nXkgkxvu&39C|Q-ioI|i6#bP#n-Y3XvGsuoWKdOR+j+^ z!ao={35{F9#$$v)9*nbc&M+pITDM{;K+DZ9gS=Yp1w6mjPz+UJ{y+BKHO7|gIuDz^ zymRMH4QDtyeCVnvisDElhfPtpyQ+YVW z#y|in4@s0{$BGf>jXVq?4v=3?;QUMs7)XG?$&WmMBubPxNF>D}>_oIB$BvU-r)t;U zXRmKPYM&Y^!5kSTIX$)ZUhnnQ-s{vt4|;BfJ%6+P2?HaYL!oQGl$F)8FqDEpg7Mq@ z03@iAd7VKdXryM4phztsL7mPs4UwRgoJq)lTtk8)iRTq2WCJCG1X;??Q6ZXBh&-$L zv`qDg_wCUWUp?5zJ%g}%C{#@KfK*cT5C}YbuzIL;Rt5t=!w80K6bObfn9>;nhQKFux6Q>$5WV2R0s~!VDGwLxA4AWyuXVqg2>x_DgNj#G_JyJ?GmH`TqVwri|v&Y1; zNX42MP|2~(5_tBQSXSs5B-o!AS&$&}uxAV+L6L?*f`CRqg51-dEr&#J9cdj-7j=D_e!e^O;SG6HL78D)>XWf_>U^mqCQ@jNne4 z_R!^z@Vw_Oi_FMfKx6Q(l5iKEAH-csrbh^YVeqb!;jh)V=TM6fvQrs6Hq}xIRE4p*{%Dd)`oeP|1wk1;_<)3E3t* zKZv`OOpg!(w;(QI_-pm;In*MAY}FDPH5ee5O@rZi&!5S@DVb4&G2F6gFsASvnlwlW zxyS|3hDCOj@>vqttD+EbmIZu?k;^{s^@c60kYQxk&vrK;JLXx>6+(6*2}X8+ghzJV z!=5LE>|`2@?0TB0v?8bmp7X2>dK@FWWcAL85Kx0pdQJusU}P81zP2MfmCEyoke!hj zBRi3pM|L`y=NKV7D>cr#fI4zNlsrWW&uAtgMt1CuopU01rqFmkvq^D%h?$|jef@cbi=QEoW*N1pl9oex@d+rdjQ;2YVfJ8!l5T5tEq57bb8MzCP3*r*8 zO?Z9~cPW`3Ap~wgT*C0z>f3XuMF`odB{XUo{3!n{)?4~)$P9fqf3-}Tvmwn#r4O>=gzQ8TjO+jjkLVP$ut<*^)yjwMNkbq=UEx_I7W8K>YWoIpa!4xoD3wu$S$6JZAW%0mFE#5J0meh zb|Nv4>~u2EF+z4$YMgZeb>x01d5RRC(M&>&?ARSU=S1*Kq49iXlj8ai@2Vp^_G!-@ zLUsxft`Cq%s1L&Po;OqaCJj|aAmNc6_ps*)Av>7{BfFj^Dy;~rf#*CcgC57oE?K>E zA_Uanlb(}-1Q^-Hv#;&QPNniZB4lSI#>h@2=8>IF<~c^l&Pt85E})Lw4<%2L!ZVsl zh>;z;W9OU*o+&h*&umg$AL3ngWXC@3xkJcKA;R?m5()J|c;54d>VryV5^WW)fm#$L`oUCxT}Rjps9)6xWA%R~^~0PkZhV zvQvm~eSkzleGs1ayrKG_k{P)RkPG4xvQ2n?5O*n=9w7v7L0rP{*XrAIs6`0bswFgP zFhDMw2E+57Ka+h^GNT4#xMkB|OyM~+X^;|fkqe*=i|pn($W9^REDQJ&BbR;N>kV60 zA;ZY7pY3iycFeP$D}?Ms5{&Es36Jc!hdoaS*~v5*+4VG0X+=;CJm*;%^f*R#$?BaG zA)p4I^qdSNz{oD1eQif}DwXFEAv+^6Ms^}GkL+|Z&oM%FR%)De0d?elD0zw$p3zJ~ zjO^GQJLg33Ori07W|QLj5bvrZJN9YM9YS^r5v~uANT?6O^PV?UA5=0UcL8!iTtc=9 z&ky1*CDS8>z%7VN82(y)dk(b-AzQVCMhynYWz%4I-t%X&Z%Ss=U<|ix8jLAChb9eD zLN0Ovv|*9mK@PH0h&amvzQo97pZ9vhmQ~0wvg>EN8;~9Itmg_LJCOtRt7zekzKNS=R^po!6!W@0|{=nKd}7J@{@~;v&H58NCO|j zfA{-{|N3V89;NOiY9}49?6;bFl?UAD-bZH_ubrG7UtKSr-2d9*TIexZBrn~T#&uU}tnzkl_gJt_S2%}dWa)!m#vSRB6k z)jx5#yYrpf#hv9i>#NCMzg_&!?ee#8Z*Bf>qk*wC@NBza@o#Ju-@lx5{@~*L`tsuJ z_~GKo;@m~^gSRGkzvTY=`t&0F930GZu>ySR1uRPpa_6`x!Z{OscL(chdxxSHJl>dAvQ7?$s~EK^}w_KV{7$uFK>O>W(LYRAd+W%oa}cD4ztcZ)lR zwl?Oz7XXhsd^FEEVzU#rf5B%JKV)_dHyjEv^^G7iSNTAD&)1o?qT7 zzP(jEza;X;jLaIXw%ut4Flg2pW57mjOIIlV+~3|RKDcD&Ym2Mvi%VzB``3%}tBcFy z^OMKUbWcOcK1j6vZ1K$D)021I{q5U3f9bnh#nVePIQh4~yLI#zzPokznMwQaj((^e z1@QQ_#lwfE=Z~(E8R{O0IoNj2^2ofDy>sf3UANzchZ0*vh9xi-WG zjawWcaHdju9@iko$WGjuM|L`y=NKV7D>cr#fI7GpM0O&DXB1^!10hCsEMe!I2%af4 zp3iJjTp!|HduYSf2li>t9jXrs5v~uANT?6O^PV?UA5=0UcL8!iTtc=9&ky1*CDS8> z;50Qt2!-cRix7s;sKEfaY#I#Dd;U!JP05TJjNz6|gE58Y(4;|1$VD!IHbUewPkY{Q zYF1trB2+tbuu_W zRj2bzuSI@?qYmsIc_AETD3W+yb184&A=k>Z`pzF#UO?asRhH4#w>_)*v^PgE?&5t{ zVIDf4UXg9(anE25)y)wMM95kf>Z(`w%&qD#kIu?qKw)~=MWCpu&0c>t2E#@zogrYN z{W=}0Zcj)&lev^uJq9dh)MFs9HewRbWG*Ee%K-T-fgcl<%;TQD z94gktU@tkASpv@+ z(y8`K7elu&=t0lTu;*`%K4D;_b0~D}m$I^27KTzVNHBhzAAkf^GOsg;1dY@T5)`Qg zB&gGQrXdovk~0YzkZVX#B=Nk$glwQ>kRVIhIVwbR3Xx|upO&c}@xDEJ;;RSyxMvVn z4~2@U9*|0^9s+@94^|JA&dOi_Xc)nejRL_i22(mizz|r65e&oPYT{G_hHTb~XVqf> zXhuB-f?;}0>8yH;VVzNrF^Ol=rbkN2#xg)bQYI0%`iJ#QIAoMpjIV&t;Vd)}~R6*7F=+mCEdKpwifWuEn14JN_3 z6YSC!UW6gJhdocjY4B}tzfV+J(KQly&a*P;aeUjGtll{hx-)`LdQJusU}P7Mw6-HV zmCEyoke!hjBRi3pM|L`y=NKV7D>cr#fI4zNlsrWW&uAtgMt1CuopU01rqFmkvq^D% zh=}Gz22~86*7$M`jPDhWXC+~xkAWJB*DlIknqTkd)V`Ykey6}kzG#{l~x4R zz;m9JL62i(m#p475dv!PNzchZ0*vh9k=Ax(r&4(y5wbH9V`L{1^T&#K?}_v2#uY&lDQZXErIW5Am)#vSXk2+#zJA5aId&iG=zfJnwl! z^+6>wau*;M#3f{#@cba|QZhY42;73WgyFB%x93oc5VBQEXw+bUTs94c=RJQW`=(?@ z4aRWGroouPb7;~aCFCL(KpPg>)j7ycA>u3x_!1+RectO0TUH^%$gUsRZa{X-vz{x2 z>_ifb>;MUm?6`+LPYBt`G#J_SG*M|qPz^liSsC;=Ms~^Sof9FT2A}kt3?#tFE*@!Z zM|LWe=Mf=0BQZvHA~BEbbTZE|LUvYaoOJcfX4GH|w`>}WDLjWJ4N^ibasjkqk=-;0*(pSv zWdUDe>ToS-o>21k~V@o|AzD7}>=mt?kH8rSd!?WM?GC$WA2Yk)2NFIY!9NN{zEF zppM)RB~OvUGnz?=ksZ5Z=bQ+hDKwtXY*Jhw;$3xQ$3E@3Lc!u0_X3H3pE-t&g) zgGy%PEqES&j_lZ{J$DG%DMYwFKq8?&2+w=oP<>FzjNApt1#t=4COkigyOd0i5CXR#E@Aj< z_3b&-B7|(!5*jraAeT*p;d#%W$-XI>QG+quvS~1;@En>nNC~;f1<-~?cKbQVP9fqf z3-}Tvmwn#r4O>=gzQ8TjO+jjkLVP$ut<*^)yjwMNkbq z=UEx_I7W8K>YWoIpa!4xoD3wu$SxjfZAW%0mFE#5J0mehb|Nv4>~u2EF+z4$YMgZe zb>x01d5RRC(M&>&?ARSU=S1*Kq49iXlj8ai@2Vp^_G!-@LUsxft`Cq%s1L&Po;Oq< zR5BxX0dhfHLbeId58^H*(<6kyEr?4P{#t!|4z&m&TeXBn4FaCJj|aAmNc6_ps*)Av>7{BfFj^Dy;~rf#*CcgC57oE?K>EA_Uanlb(}-1Q^-HBdzVo zPNniZB4lSI#>h@2=8>IF<~c^l&Pt85E})Lw4<%2L!ZVslh>;z;W9OU*o+&h*&umg$ zAL3ngWXC@3xkJcKA;R?m5()J|c;54d>VryVQ16|((%fEtEpFc;O6MLM`stWotzzCT`!&-J=Y@7<0a3N=pQ_z^lvvuFLcj+b8-6U z_3Nvn_pko5Cxw5$*?F#b`e8GwxS3F1{H3D!a59zp>h%*h$N1#@;qlek=>v8$^?&xP zpFRCopFR33&xXm<|0@lw z^t>x*H!cMqzWUWaak#tl%kL#j{lTR8Ta&|E_qI0wf1`tq4{Us3;{zKX*!aN42R1(N z_W8ihcWxJVmaouPlfQnu_?_G3Z{Ob9{NF|cV{72qcERF5-YTA3zDRuj>gx2-`Q%Rf z-`hXb4jFs=+T!8E)AL7H_g4=}x09C?;icwVcXqz7c<#-s#qrt2gE!n-lUpxu?L1R# zechcBZSCw8ca9&Nom^c_K3;ru@+0n3~z?|IMnj~72S*)j0l?St7D6-Z<&hskzBWGmkm*=}-V z`)i5as$RDCxO0JgaCtONq^d-z%HCox7ayDasA93@4mtAtcZoc8FVDw#KV0s{Ibtn3 zhbZ$wFZ2E4QW% zoclYU?PaWry~%y6jG~|Ij4SFqLsaUibIEtsQv>J8sv3&_KC$XSFDfZ}*B5Ur&X1e> z8Odo6b}Xe0uEn_Rhl$%hUp#mC?Bv$nTe{bNXRA99ZGp-0-fl18kEL!4nevYNeQ9~$ z70vk*y;o0*YO)8tTG}sD7X^_w;Fuxu-fRk<)VSYrVI$rg+iKSDhEJ;n2f&rNRM{ng*vTKaVHKc{j9Ttjd`vMKAaGQP2tQM}WL zQVM7X?hX937UMEoO3p(Kw*0%PEd%O7^B{NcpIp1!mT}U& zi%(AuqFi+T;NqccPB=SxYjJt3?#tMjzmWRYo#Iz+CHXgCBGiVe^1wRxEQ32yYhUiQ z_J!hT@=CVWI*zF(0)SdW6CeQgzlUORQ6c)o-%H~m;49RlCHh4BCPxB7##;YSZ`#il zpP3wHnKt_c@7rla1jvQjoOS8hkT?(>;VZooezJIV^7(uvbo*z-UyWtc~FpI3YLIVkog(*f=i*ew2^ROW!* zP?xJHGZzMB8BPoT;hyPNiqB0xlf`tty90ev4FsH*sR7{pGVmKuDU552p;!E0sSgAU zi26`jA830Wspr=_KiOOT)5XDLf3Vftm6-q%Ua{}!ucyD76oWEAL-Z33MR+y@rXd7= zrdP&M@yg_LnaYrZAV((!QAI$C*{T4sz~RIkp5Gp=-W0!|hD3mh-Agor&WVJd3?gXm{KaRjW$%KH9XOg|EDJ74T$-imrwe}6DPY0(3Br=o+;(GKjm$HhDTHW0pkQxP8Op- zw@L&iNy@JSx-%ucuoUO}wG8rdQ$E=^L%CG?fiN#+g1UUYm;XjHmTH)2ASu5N2+x%A zTj`mUU(5Hixm6(Eld-s31xT)|1FkdWx|j*p?WVW(30S@>I0ZjO{eA8<_++p@({7a}N_sB6u9Z()V#$fhly?!QoulbBd z-gXqp1I){i95}jslRJ*&oRf<-PY-bFfb1}&0m|o+{5*hBGZ>;;n6*s}-VxsRr#7PN>gMcZ2!Il9zMwEHC(9l9$?Pz;|6AY;uln zYvGxWenHFk)mj)V{veezAUX8OnqAzg#CiI(%*U`_=+AHui}_^6pW#BycO2mmpi&1I zH&|-m+#G-tlC9X3y)%X`Zi?ScZ5>eCC)*}j&1G;K&a}pA&bR&NdfV4UIoTax`z43P z-%ceBzzknD6CYZx!@Bj;PKn#63tIyIwKGv1zJFoP?HY@5$sRkQaG(#c) zym~-Dr40ydkhH*oW#qN4CO8Bl<^AG+O>8?0VC!WmE4OA~;PouRSD)_fyI0(we9CCw zc45Y}0;Iafb;A_jO$7~*3Vn1fapIOA&W&gC&lZk)70ip-q#mRK5lKjg1g$2h*a2Y; z6r1|H$}I zV7-2H;AQNFw{lFn;b#Get)U4J2=am1xMDFbFT!1NVK58O3>_To~-{W^g73hw;9}!Y$Q6K!TYX zpw7R)#`WoVvNAzIy5aW&C`2#ntX3KLzh7G4td*4+FB0&-p+7Kf0pAbM5Ve3SEV3a` zN!VCtbH~_1Z(|QDl61rG2c(#-3J?n%PF%~5qbi7Y!|w;Eh%QrOwJ%$f-dMb~{O~p@ z1N$;IG+w5%`(ew-elGXOMq4~R~C zqa_{I)Z1%o6Feh$n@bt1u9Y!KPvAV@H|?d42+q2_4CON3+qkWBRYcVgpk1aKfd9{O z*%NEtc#mQ_63(4d{d3Gn&3dVnK4{#Z)lTtDrGKM42^b;IxBE^+cYdmvxX8e7|$-57fdY5h*q zKl`4Q=>WR{%Av+^W^3>26Y??Ps4dUf(=Bh2AtXxbRPKP}2Frc3P37LRwy=xhGVUoA z46(1ifh>u=19U?z@qoLw7l%qLAJ7>a+`0WS%-$64MAm@M@SU^Rx(b>0&1m}ErJ&XU}sm1pK?1FDs@h9>~34LvGeL}19-d^uv=7Upp8^#SJYz-=O z@$0t|kKYUM3%0_ljIr*&7kx3#+W2bS?QPRW0BGxI1I2`VqBgRExZsi8rRfn}&4zmc zw9UZpqV_~U_J#b+m<=6*J0*gC^Sia$-FBY?Nh=4pTr5BXyCtGyz= zSbSk}w1y%&4#v?(fO>1`0|kf-{l?ZuR)mn3_PqcF;Y&M+5DDk+?VDfP#g#K`4T?oy z#fpE5#6^IFb7Nctx%rTgo9ceL4~fqgKQa02kdPpoXJ`m$A)v)vEucVD9eqInXBgMy(|02>Z&w1lLG zpY6lp)#CG$pBNqvE8CB#j({lh)dAwdjhlc2X4Qjy`FvzlH7>hwRzXx5ASFC4h~0GA zC(w7*lUL|c#*o(20(M76zanUKjhBhM0W`v29kaXhaUerR>(7q0jmMEb0LC2Yk=x5S zaU|)pN&z`ODFZx&rwGslY#ERNq49V0D{l&Ye6@Qr7S~IF`~Gg?^Z^H|Z%PRhyxVh4 zcPf8k*~odr!YntYtr%5IEYYQnY ziyn|2Ie|@jRuD~^u$ewA#zufLaw6OGWNRi7NE!wj2q%eo8UW7k22WtjBe5nLe>A1t zG{2}__4_IyKlN>RnNnIqa0S4PoWR1XAt`FW+66$0Rm^SkGtxx(fb-OsK_WaK0y5;w zc^Tw3&s8>1MSy_0sz9+&uZi?WR2EY#Xep=y9<-H$wo7)|lu7V^B}2jNydL&sd_c-J zAQ1)>XiJ1)e?AUmFqoay$kyX%U??0+*+A|uUlot00milEJ91hI0ur>9f>r};DUd;6 zc1n+TWh}228<773{JzqTgd=r!ufpuH2x=QdLAFDsoi;V!@56P|X6^>kqVd zM${h@FSKo>4)~Bc2m)QKxDd#LD&)ph`zaB_4)&Mk2T7}dD4I~g9EC6&uo2+KRA4|T z)F)XP_R|xYS0H_WgE`V8r_cDt@%Kdft#dU)dzOHnO@KWlF@vmYa{kX7%{4&P`}1?=!;_U64_hJ4lzUwPYcBoAkkgO4Rh@5KBt>V8vq=0 z(M=izm(sE>?nYBcLfbF8#eK(-Jsj<2$PS!*ojDcbD254v;(sD`4^Zyo_+ z!$ZEOwRW$YIT%MD0rah<4-_EL*Hp$9o&oM6{sV4GKRpnQ%7B+vz&ZgAQ^-G)uWPNB zR^qNXen7hUqrn*Pl1Ai!4~ba@bHGl+Z*466(r*4}Kt`?>P#~-^9m9G{WssMNnSyll zN24*|i$1#-_qs5X&+)A9^2)x9s|gsD>*kN`W-U8#N7{&5?BdkPL6Vk<~WN&@GTYoc!fTkK8`v5?6k47)fE^5;vNR0WWc(3Ai$_|Mm>T zuwMQOywcmtXby(@rqlc~9VXe$M6yDi55bNR?H8`*?C84ZhRfP4fdnx_kYm9!CS=rZ`Un z$o<#&qTguhVd?}Z1px`#N}Qvdm|Kwl0{*k*zj!%51orbw0odT} zP$*DExLM9s1&RgvgfKg$Q#COTZvY77qnGo551E4?&}PaM0vQ(CPj#3b8Q&1fz&&~t z5>41(jzXB-*$7a^Y}n5=qI)w=l)0Nwo2qz%~9&g{*5y?pLx z?P-^{9Y^v2_l8T39DR)o;8E1V1Vur*`40lRgWdc<`Al2TfsE*^-Td5!Id&%9{Lw%S z=AxT42+H03ZnQ>U*zC93;wBk_g34Z`vX{|}jSShTle-1Yqn|mWwyA@0*6fpG(#>D) z1}F!+`Fs9^)*iPA&~q@qwP7FZm0@zy+g|Plzy^EU!M!?id?myr8a{nWpFBC_U zSGXVPSog?~SF!`pML2C+Ll-C>SN^~8mS=ojg4T*_!v2l|bRXfur0-I5^&jDzj; zZN_m5@~e(An%IGU)uA)S28J8lVf-0}d6RLJ&|3#+BS5`%v;jiH&F_#LQwrd*)yM=+ z@dwGT{>r@o15vRAW|L6Z)P1Ys4UGGU4Vr8c2cFh@yg_L18@`AYDBF7Wd!t?tqc?q8NUYBZ8@r5Z^+4J zxfkFgI@yf9ciC&jeuCev#8@r6R2mOMSsv->o-&YwU5Sl=aRzc|`(`K>$R1Aj(8;F4 z18{g{2#4{YSKBv7_V1?YzYI_ioq<6fxYX~4ZLF1@cZ}zP0#jHg#Xp&*{b<03_CC}H zx03ReK^qybNO)%~qON(Mgm4;|r3B>p5da)ZVh-oz_!>R-E&~QM1wiPVChf{)`N#E1 z7ZMSHOY*&oG8(vHGJ9ByPxdnrb&9Wjxo$$(!hr$g50EcQe&qHM+#5?C&g}<+26{ML z29&40?~*2R;PZ1m^@#S*cv~iHF{b}zlJ*0p(_Z+<9=!9*$AK9yeB9nxOkMv4QA9L( zBU2F|3f$<8@r32vKPT(=?_xClR{{H((tq0@3VSk3H+Ul#!GvWQ8xM^7ATc=#js|X+ z(qCYWjQ}@vBbPtxcalhjBWz*C(ez)1kt1%E+wS>T zILu$IJ76O>{l{!Erl(miU_DdTi}&E2f0~{3!1`DeJ;h5U7Nq}h=$EMo6b14%ac*g# zO5)-`Ee-ULrwUjQAM$|4S0?$(x5VZX8)iesFNDy)M-Muxfc)@52eUcv`*UjxS9xDv zJ3~Ivn`vofEI4IJL5HybIAx|_>BTEUu^W8`d=~vlf3B!eh z>Kw)Irl*{Bzv^81pq4oD9Dz62P9Y3HeD zJI3FTaBGfZ+CC|gX%S^K83SKclgJU9@7;Q09y9)sf>TEqQyE$Tk_X^6Tyo^zbNeN| zq|_WAmULeH4$0gBwtY;>+)_(UJ^KCYL7&U-7oVDZQqARoJMwk7GMa~heiYohS6GA1 z)@~q%@rS?ssu}(&S~UF;bU6(tZiD0nZmi70SXUQ0-d+4&>cP`+!}_xB!5r*%qb~&9 z)3(4%cG&rJ@3MQv{mG|{F59w$eWFNz^IJ{>%tF8UO`NV>n|EgUP|f%=U};Y*mabj|GHv*!8yNH9gO_Kznlixg}Uvl;obIA8F?AI?bVGL+l%0fblXn@_(C1# zDigf#uW_GhY(KX$hZRM;?Wf@+Z4F(Za9HEFu;c22i5q&wFb!}Gy>pAIXW;pC-UDOn zm5II$w(a1Tv0qQGS)+*>sJ~oij1P>1q1*+v8Qb*D++pp}ZhK|EhIKy>8g9M@K8=#n zn1~Wa*3k{^wx0#~hpwN1HUjXa-3M-_Mz*ZRn%Ko_yG zZI<0pe7F5fKGic`#IhSQ_LmYdy@*u?X0V&{5n#TE)z;3CPeK$x_Hd$yE@Bn_ki(CD zyk~6C*I1<&vC&`*bP)>|Bg++LKQe2Qg1pgiq;m$ZEY zpO?X1#w#t^l(D*=_DQ>CGz3F?;Uiq|&M(8ej2Av`?-3Od&I2G{&3uvB|mU{HxQ%mK+MhcsXUyI4^%YmwvPs3n9_gC7>D`Y01T<}=cfO7Q*2oM zAc!`ZM}sd+Suf^-cm66EKR3mJjdG<1<^YNahklugKv58N=8dNaE)LXOj}$=;dFJ7^ z7$5S0_Ewzy=h4^r%qKd`>WuA$&_~U@^q^xNkRLwiVEEyEUj~2XQww?D40Qn6!)Q1K4|&)v8Eb3tevdYqN5d^}rDu8I9iBVc zQ`rOQDAWxzXwW#ON{A+2WGMl7zB2M+EMYj`S5-~B?GFOxL*4cj?m?$7yKdY?5A=-P z^UT_emkV)?=BLuG^JvHg>N;n*Vg22Wxsc$UvFlvkHA59Z?r@AZSZ?Iv@+IyVa&zu2 zxd839KL|(;b=w2YGwnPF@)4e~6QA2J$Hsq{_N_*fEAUmd)ENF4L~{h$jIH&Yd|d!# zhza70cH1j+EM%^bd#};WH;Seh=cnEF2LZOBZu?$xsU_!Nf5+G!50)Fpq;r4D5ER zSo`JSf{f(^D5W>>lcw`38cKnh&LL-M&RN;(fms zlMG-kD9zYmZe#M2YckmfetX2`Lei%gAa^zKCTv=ObX2!I!8fIA8b~g3PhkeG2SYb=Xqx9fTyX~uJaE10h)CYGf zHt+Dc7qRFXEo6s`_bALUxrnWzc@|kp0G{tATBz0#gM2J;ndl%NH?GQX5o-#7&;d=Z zZ`@j}N{B#@l>S>QaOvO8vrzM} z_cFYGk%%Y5ah48*7X`CK(L}M-RK?lPa@B7gYEX=1G z@+RXb0mvSX^D<;d?k>Z=an!-Ny5tk2+rFv-+QT3Rn1oIE^0lw|9hcpbaR4p2X}5h9 z4X(hIp5=je_!xj}UR7ux)6^*e#UFXc!tiQWS79)Q>&)BFgGYaGm$9RL~ z1}^SqSoF7;V{puPs%Tu`YNXq~ibhwUZhN3crk%$Ptmq{+w$^itkK;0==e&w0RN$-X zVcqs_J)q*C1@rB-xNU|qfaKxGZn)&gy(8Q!FLGn~o^@T3a-5oW+t=ZS z6?NMu3guvTEg!KNzg!PD!-kV^Zrh1Zo6hTSw1%3_;qJUM%UH|UbS~{Vp5W1Zih+Uy zuXYnEq;r(@1?T))c*W{0{`Zm}z*W(p3e;^s9o}uP%&Raq;Hw)mwin*3@6ZnOXlw=Q zFjtviecw&8FgBK3_sEb}@Ji`t#somg^L5p34l`Q6~i>ZHT2Fc+Khqc zQ~8?Jc*UUaoZ%<5+di6jf%?mJF8IKZ58aI2_GSjg(MC9ZTSprxG@=XFkwqUXHvWKA zi~#@8^%F3S1ns9X0yK6Xh@0m40O_`mrd^;57m)!z7-URnym_+hJ&q0nI?U7o3IQ1< zj;Dj1fS}#>(VPqPBG1}_do6I&EhMaHe79G&Wvnm2#jdCN*OH+pRW#TFHJB@Wuz}zv zTu53lHkjLXA5j_M5;j{IAR^qvi|>^(xR}sABmL}IH4E?&UBt$=S$0S9-S#v2RL^)x z%WlZnU&4&^GBO%*fi5H2&G`tB;i2&|QrmYN*~5umhV01UWlT7Z?3}BMb|>BTvj7Fr zMJ&`moBBt&h!wVEtgQz(>9&ssS7`4;eQ>W_cG6aBH^Bm5OsVXN*)iii3Uf>^VxxH$ zSxP{jF9X1_gyDQ&RW)t5oCOSM3V_f74qSly!G#o)qVm;=XC;0;ax z(Vz=c`VSr8AQYz7ZhMli!?KJDv%{bK*eV)rVM>32H9i7l;Fp^k$<7(N0qH*+_hreC z9A7^38%zG2(g4zbxcxPy|JDi|=HHhAUvB!3H)X7zll}wNGiAM)3*PxT(5Eu^wq!lP zGwrsI=38Vc0z`qEa3LM@G+D4js=0PP0Lja*#5#2rHwoO_G*{{7UmqtO+p+a4H$Y3ES}Ru~)bx&1QCPJ7Oy z2^ILNT5Nl#@9S<-g#=j5SJmRS<47Kk?1oDY+*=u0F^c3_{!O~=4+3mM-S)k|Ni8`C zGEy^s_a1DOW83eL2oB&3b)SQ6*=()Mr!aP(^E-|sZ@6t8BrkH~>CU<`<#?%-HE6ee zG?xN(+b7!QVD~}PSZ8dlhbuEC6fhP0k$c*09}S{FP3Mpy-kJBU->;X}9Z}vM2?4>w znc6_Xkynd9W@XLMvp0g;n5dxNwm%543w7JihIiW^aK~v9E|Ma0$Jn5+@=7|)>u78R z>M&PVB}I=g9z5tf=lH^}Cry=gH1PuUm+M^cfie1L?TlUOW}eA!59(+F#yZ+Sp|QpY zjFIJL0*8FtzTOS+4_!Y2Jqys@t(M`iVk8-FtVBM^cirn~+6B6B5m)1bVT~Wx8gFAP zjA07^9fU);OdX&Q*blg8Jee487RvzeD$;FV?*^cVUgTMIGp_}1J+N)ii19^U*%ljE z`Z+GsZC^)&El`8G!UrD+2SH?GYrSo&T*-k^$cO&xXx>G(GC)MQi5Gn@-k8eZVnTRz zY~27UBki`Yqrn&GA~rVHvO9|Hws$Lm4TDOIcdhJ(j7x}M`R^uYK6NzW0$oP3o3jz% zW?V=|na0aVZQpTZ4<~vVvLlC=kN3uro%4Rt?rFDuy%)~<&_yiNKbQLVM!AR;wq&fW z#ye>=ucN^g+WSx++^yKW!;f+iD@fn%Jfafhi&&Kvho@zg;eJ=US7$jjBn{_Kkc@!qd^wh3m@Tv zb$(@{MaDZIxA%yO2RY9ZbBbs27p@gWcB2xXFg#x3zEJ@_;98SI8f(s5o#V=eGO2O}-(`);xY z{)M@al4POHXCpGJK=yDOoFO}KcV(u6FSYhlsmv_xjDUdrryTNiJ7w0~m zl5OZQRCPnXZC^*DD^RyR&^*)5Q_1(X#%_CVzZ@I?VfwBfO{l5Uj`#Ks#ftt?Y?z}Tk`}cNh|G2b4u9U#oq}#rZ z=2Hw59C@{jxs2bo7j(h-yl(OI+xB%GU>EAP-yhsx-Xcq>aD+4c# z2M_wrIle%;?W2hosJ~q2f(;Be`NG&HZr&tAVktw>Zu@8g#yZ*nq2WegNO*5#O=ZPK z`fdAq8sHzgegbBbp#4-nk2Bs#iJNBl2kkVErd^;57m)!z7-URnym_+hJ&q2-AzY>o zPzcCZujA<;%lrR+dW10zm#^qWp0x${S}>jM7*8UUr5OtgkdSSq?!g*uFrNmvh#Jfl z9{E7fC(g>Ju*R-*8-G}pKpEi@Hd`4eBIY||io^N+dY(bL?WX}gqKnvA^ULlizT18( z<3(e)J-Z>cu6SNRFC(K77w9sQ-JFd8H{(Krs3J@--c)K#VZ{L1!--yo?7-ob$rj^? z$Ay8Ub$SsylP_YSin-LU%&;(C#0pztgBpHhm44g4js{m~??ZiXw_@`SF9W}f_aVIV zh)M|OfLTgFo?q+M5JP_~H8IgaK5ksk0tQ5#2*evhJ2$<)nN9O^UT$;*ez0TEZu4lI zg~{w;2j2JP1HxRRlAW=LqSAr<0rF+ZkKA59A{}o9oPtP^DS~y99bQU zk*!!De>m>Tk{>yKH0*XPMR0CE(8I{L?d$!3`b_D+wGfB-_urm*7P)03W|pS^Xpn^| z>&2__&d+t==}G3k_A_wA+3^U_pGy13E&P=I4?tX3-8({I#7&R6;o5 z%TfaHd}X4=SW4i0U-tykZ9fl~4|Ut`Tf6P&@)@78b)MOdvAGcH^xO7zG{ypToikhW z{w`nW8e7QaUB{6-9ODg^8@V_K>LhE3sbLINfo}VGKys+t9%!Cv=b1-8n>Du9bBkvf z6sV5&oJSKX@Kv?g*-nqH+@uQgRkgVEI0E)&kznpar!O)I+{>1 zNM7K^ZaBr@AN$H}7wc;6hIZQ@1OSG*?GpoVu-i?e=uK;Et%oZ!3KK9D`!RLebRG?& zKuzb6A>Nr~tYvIEm-fun1rR)(sSOkyd37C3y}p!Oiq5h)>9#)zunTqD&xd#0AIR-< z#%_CcW5)JU;iBF4(bx*qVXiX4`@VdXX6!Jx?mvz$!b#d1xG!MV_ooB)+L z(rteb;2L`87VW~o^8@*s)%eb>@0{Zcq}x84c!B!MbuRe8$oQs-T=9)9>1LiuxBWDl zfU%A?XlS^Vv+u>MKU%QL&BPMo z-OOu&TTk?VFEjF;;=tmmNr_o>w)L^dg!3V-rXGYYMu|3_!A66GoMmW05RtAcQ znYG*ABo62I!ux~acNl56{d70LM|2Szv%2h#V!Q3#N@7fkeqg+%Ww~KYpMKkZ8jZL> zmyzuHd<4kw(0CcC?Tf_|)lDv9r_p?i4B3&xN0Tka5tQ@!p)UsMx9z9VI16+U3-!;X zemBFyc-bl}%UE>1;7+^kr_taF?R}^Z?pAEp;oSs_wV1RvtO>g7c*80rANZd}^DMHI z06gDKw8(hzgDU1+ek~=FcFXBrxQI0cK6S->z)au2aguh+>0Us2+WRhP`vyKA-O^^f z+>#v_6X!%d=`x%~gDkWcKEfF9{M>i&a(M3TX8leuDF>}XtrRPuh;Sa5sR$GW;s{(F zgO4aO6Gb5Xhud6J`frl#tmnA=Nd^g=fvVjN+#*&ot`1HqC(=y!tn$mx(QV#RGc^0`@KVF#epS3b5 z&3a|PdZw%wbI3ct4Ei$8`#tv=+kRRGEQk+z zKwB(M{-u0NY(B?fHe_rsq#fFAKP?0D!v`JA=DhFAx5DPL4|(5lWDm!A8L}gH7f ziAjGP#V~O|y6vZBKzkVE0R6KGUq06}gB*5C#@bpbpxyS(HINVb z=78h@xDA&axwnklM$t-^f2ZHJpH>03p>F%$d8C${YJbPrQ4f|I+qP-<`81kNfx6ED zOMJE-eZ;J>`GiRw7m?WR!-?Y4(2 zGwv*)YkRl-G#W&Kn$96ZtTQVUDU5yS(w@1xAl>%UXg!9YY8cNi z(r)`{9Zu9xxBbE3ZhN<)1pIc4-S+CnjO|76_3gBWG8$WfI?Pojc;A(6fAc#So3UK=0h5T^M*?N7t;zg9m+^48w!qr2Xa5#0%74 zt~165#=f;p+`P#+O6apM&_+0YTSprxG~}~5;|~d}n#u_m+HGHl3s-dg1elSaeLecG zFO0WYBA>L=JeqcaE?mUb*kEw;E=Im~HQv!!*abQWhj5uX03qOpT}Wq<;|Y}I{pq*u zr_r1X^diqHmU}I5(=CiI>&nuMCka4Q@8_Tm=Fwma)L^dg$OnRq8I8U4w%v0T1(Xqv z?y{AEB0|QFV=9A-3Ds-pMeH=dM|2Sz+h%z^&Uf2Sr#r^WR(4Uw#?s4yl!a)=c{Jhz zT}HBu@eweye)ddT`|X!Koakl9jvSurzst4PNUFS{7qQa-1<^$;v<;j3W$>NN6D|I~ zAudx0XuIVsU_es7eYjGn&jQNR-ghHKb4h>3-O{d?2QprI z3A41zFdAf`z3`FMc<1N7gSVarGF}upo`E95^=+miP!z~_xZ?@L#eu4Dkp2VqGo}Bw z2^Ke8&g9P<=T?NUCC+ei_>&)7MWZcD=`U>0M}Q3ca?>B#c^vt}abK4F$noV{-?8N9 z++HfZH2p_oEllaZwE~y^-8>8Fv@17_$D1-%*Mt3+)7!CVkcBDh#ayt??YjWpVP~kLFvLPkWfv8AA(k zeh*1YKB*9mwZI1*3_rZ@%O`urrt$;ScdG(>sqBGY8bos~GGs^Y9^C?uBPi$UL+wsJ zsjwf=9tJtU?rg%Jsc-nm8AEC?!yuL^93J$1BN|+RD?Pgd@9^Boo|cYeotP1V#xYPr zIN-}t0`mNk;GQB37Y?ebX}A4;z+eZ^ChD(mzn>)1o zGF{}9&Li4V5Qa39)z9!;k}-REFiK3mHNaK`R)e#dd- z4JU4cajO7F<#WvaZ*^V}y zM}sI((>dgfcV_ue&Dc^d?U}0-DtI)XVxZu_tKEbO=|E|H!8un~y`lII(~=wws`$Wb zi_7c#PcAOb7MJ&>bnGTlNb-H6xO4pA?Bwcd^5Nn`lMkLFU{Cbs$s=nBN7Av z^&z|s>!EJm#BR^*$BG}Fyd*H&t&+jBq%9y%IMEX2L5^2tzlaQ01mlCne?sgNZY%Fm z`*7*#uv@&4tSpTD{~eRTfwCzq!uubnNv@?`R2 z`j_qJ9-Uphc5-%nb-j49opQ4CT=De7>$jdPCbxElnf3)NV+^HevQ| zap&;%u zSkD*F9X{JQ*5qAxza<}K{T0G>IMR9lD%YH=%-q-h z`zxrIdntaX_`%6GkV1Bi3}Ff?LqHc$2IMu-MUqtp&P#G+_={gJCd-U)DFD8GWf7v6?l;kfJ`Abdy zG9`bRk-zMdzs$*B4pJcqgL_1XJ^Bn0Vvh*1M}*iTLhKPC_J|OBM2I~i1YxHnLX<>^ zl2Q>NN+Lu_geZv+B@v<|LX<=ZvUx>>sE8015u&2&5FsieL`8(Ch!7PKq9Q_&t!g4f zO@ye45H%5^rYjR6Y9d5Ugs6!SH4%a=JS9R*i4apF#FPjzB|=Q;UPOo~5n@V&m=Yn# zGBYB?j0iC!Ld=K|Ga|%{2r;9(6Cq|qh#3)r5bYBo_K6VtM2LMN#6A&Xp9rx}gxIJ2 z5Fz%75af|L5n@h+m=hu9M2I;NVorpZ6CvhAh&knu2yu{>k^>^d0TJSW2ysA!I3PkC z5FrkT5C=qv10ux1K~llXr1+64QSPQfmAk1>-!91gYR9 zso*84;3cWxC8^*gso*8qoK*0VRPd5i@RC&UvLfz8DtJjMcu6XFNh)|rDmc{_so*84 z;3cWxC8^*gso-TzVuDogl2q`LRPd5iaB5;w!AnxXOH#p0Qo&17!AnxX%P9$LQo&17 z!AnxXsaKK;UXlu4k_uju3SN>5UXlu4k_ukVNOmF>yd)Kz1`?^@C8^*gso*84;3cWx zC8^*gso*84;N?C^v8002*e4adBo(|Q6}%)Byd)L8Bo(|Q6}%)Byd)L8oRd0{R`8Nk z@RC&Ul2q`LRPd5i@RC&Ul2q`LRPd5i@RC&UDyiTk!B(V#SEPbhq=HwZf>)%1SEPbh zq=HwZf>)%1SEPcInoyAnUXco3kqTas3SN;4UXco3kqTas3SN;4UXcn;%2-7zctt9B zMJjkjifBbDQbp2xMRH|D5>G{4$1+Pg3uSo^3Nd>P-1+Pg3uSo^3Nd>P-1+Pg3Cq=6!6}%=Dye1XACKbFUWwjHL2h=so*uK;5Dh>HL2h=so*uK;5Dh> z)O@6Z*QA2iq=MI^g4d*i*QA2iq=MI^g4d*i*QA0|k0TYlCKbFU6}%=Dye1XACKbFU z6}%=Dye1XACKcQT$!Z0!Nd>P-1+Pg3uSo^3Nd>P-1+Pg3uSo^3Nd>P-1)nAroP_n1 zRPZUO;8Rk;r=)_riT8=R+)#P)m-MjB4FxC9q(?k%Fm)b;x3hw6ol4sI0 zU^f(&JVS&a72M6!BveEQQo%{`bR!AVWk>~gW9!H>L8f=@{WpOOkbB^7*1D!BW(RqBMKf=@{WcRz6@%eDXHL7 zQo*OBf=@{WpOOkbB^7*1D)^LC@L5v)NJO|!%S02dpOX9~y-s&wwAvx(qA2-8dVcR> zD0w197cPR5C-&1n%+o)l_x>q`3!c?AUF;-(NI7?rlRS}f?&2nSBIVpgP4Yy_xr>?A zp2PG37cr|#xPVFikaF(AC3zy{+yzVWM9R4fmE;M+Sju@CI4(?9mvB*%{2}Gs#Ypl* z%DIb>j$qNf9CAu@oNuH zK6mUs)>ssGPA(r^9q-klc$&ewbFyn_54qsa$Vm~FF)TixqEcyu4{ju zJhlCg7u($8?pFrg@|4cFqquze_AXz&z023@FaLMmReTgEpZ;kwp?}-%+->_K^xw74 zao9GeAMS4d2=~bGp5>v^@=#@YsO}%yd9f(Y-+X+0^`vp_$*p@&?R>Q;mOsxsc3-T% znj96cOrAYGfAZ$_@#({>Nzwk#$0t{BOx|;Pb$q#S(YR=1Ve;(R$!m+V<#W$nEgm(o z+Wf=Lr%2zw`zglZ(@oF+mjAD(4xj4&=gy>fcgxQ4wfjBG)#O(3)Xpc1cehM+Kc5j7 zaKEbP7FdeV(7(61^Y!HdlV^)(%BP=g{%`x&wu)!Ie(|OAYd3@P_NymX*V{F)5dG6( z%gQo_uMT&2isJUqT|W4F^UIr^7mB-IzkKuj!O6AL=F3kt|LzM{*Ui6sm-Oqot?l2~ zDxU3_aS@--&fYle{7S^mI@&4C?0kTD$?f9S;Zxf&C)FB?sn+HsHc`C>C$*Uq;kT+= z+dozGzdnn8hxT0OM2BxV&+LBfqPX5q7lr>B+saScp6~r}xtPREzti>?qs7XIuYR#| zl{3!l@j4Y+r&_MVkOzKZ;cB!_y~>XyldhZ<|HJL#7m82CKtFoZ;=#q`!)~GNf24Tv!Nubz4g_AEUYswJJuK$WOrAP= zDn>fb>@Zf}{z~!OYj1wTLC52((?7NFtG&=!Ob$IKkuf6MA-+D1#@L6`jt0HB4 zlK!Db06&mQ(;H*^eGymvU)fDhU|`=%^mq8WL%?SXhbvzEh2&pxEBmN;f1dM{FX`H_ zI@n9M?|$sXcW-~j)%xZqwM+bbxI0<>S!*L(`UdgCPrtl1ed?7T{Nnu9OCKwWt2ZA% zKDm7B_~OarUh#qAeLuUne6+Z~{Li0j|JC8{OaJL-iuXLec<8XnP4@%atI2=xnaRKZ znc|}O;pI=Q7mr_CJbdVo{nh=SIeBYw`Q?kt$L>=N*C&pryX|iy9Wx)5pS|Bi#Nlpv z@4@Sf2X8#Na33&pA$WRydUEEx@$ztYzE!+ge1u{P7#;)&l)L@ceQbc5i?WA3>uWDuAy=07Iqx|n_C6?)~L98 zs;`-25lt-97twfhD?1+l^j3jB1Gxe2_D$ThHaQ#r zSJPKfxpYhOzP6|OT2(Q=#*k|KH0$xc1{O_qRn?5IakMH&Zb9DH$^pJMJ79cGT}0!t zS(*1Warsn#`@4*<>8og5HcRuqR;s=>Z(pRXJ*FsIw#Twc6lo@Ai0Omn)*`DxS$7IRq~pJQlM zj+zB|p98BBebo&NK|aS3>pZpSgYmijAwE}gFAeG{I+tyYWs9BKrIj}0x*FN_O!4c~`iSH2c7}&NgfgX4F z_;>kok7}#Bippj8p3nH4xIWS0@|7ReR(%)s;Rz~SOm@0Vu>}LIt1cs zEA%<>7(Vg2X%nN+=hRhHF8g-~=yT%wM2Ah>LZ8!@QTgoOA)wEJgZe~=OMC)-j-k~! zY8DiH4y>B$Yt~>9=yM#g%F_~oy!5|+XRCPc_2tRwdGo2}=t}v&_|DekKX?BU-0^<4qp-jB$6ot@VCstN3_&w|sr^#-eSlz`HlD-yM&I8)itP@z8qu zKT(i7@+TKp?h4#BctG(R(A~}43Jd@eZTq%Rzf-RB|D~-01zTKc$*|}6I=az1%A4W> z-UM8M-D_>|&uR+N!StSBD3wPS`{UOak1j=wMt!If#ByFL$dRIJyHr7>>(oEhJFW^~Yy{k=fRx7qk zr{2}1JChXyxyo0Y`2Llxq6T;{wnT^9md(S#!@MmUM0qfL1FfD(=h zmfW+&Mr+HLyXwNZ4J@3-YIEX+lN&><^3+1gZ*LW7$Y=jvmBgoit16ocfTlM=LXDSZ zH3x71nXLj1{YsWhHPfVhgwp|;#>ZZp@i&caCF{v;24!ku{OTnkuCht*D20j(Xx3|1 zbza0oKeKe=O=XkZQS=MsDqn4GxAGMrQ-mT)E-Mkf0=deUn3dqdskXSR#1gAKwOPsP zUom8gw6(mXHThS(fW}*QQERnh08J5R%NuQze+6=lukKR&U&ELyHpL1mA(gLekSYRQ z$`UGnEmk$ybcMgv*BZ2{*_14x^48*2rlAc8P7b&GirBoXq5A=K5tYaGjg33plP=EE zQ=qrMQJ0Ml08@wuR?V9@jt-Vw(_XWtaCCMbESz{=)mRo`uRV@f<*BtOgT49#IR<-G zO)Ws!3lgflv{it?UdfWFW}5tuuooax`DimggT49%+4y_MGBYP#8)6!`D$w-^F|2yO10J0 zLzIUoEz0uvfhB3N#^=Z@;O0 z2zx0@sQk6n&tTIP{!+~~6;R6=WC4}8whA)X3kc5U{>=|p;(xmUYgaMei-bPESYF#-sFFT zy#SfYN1OQ>?A0&G#@{>Uhfho%jeia#VXuQG%cDFhE}&VjS(V3L%+iTB9W)spVH|;6 z<*Uu_4E7=m2k;c3i32Vx3FInYVpf6+C)%2GS&1dqcxqO3HnjRz44q5B%fiFnu znc;uB+13FDeI;wAx@jr^LSKN)Gs4}e9kjWhL0|oXYy`ezehTcmN>HCy{OeeRzJmX7 zM-$=--h8|twRzrkm7q>9X6??O0&m>mgg~!z)@FGtXDbLSaKP_gnZ!2IBf2j|DcHH(>7tiGW&O7lI8?UR)CYvhAf>LOwlcWy zoT6T^WKE)-rY0a*2IzE7+G@c3#Ij$JkHvS=Co`DL8h=GX#AeIp6O>uS6;vBG%knjo zSv&QrvdQ|WW(xE=XKl`BvJ_c4V5>MwE&Dl4oCSKFvo?P+SqgGSG{$8qmRjek%~DoB zi;>le#+qzp^0Rmaox3(`S*?k&S=nSSlQjdq&RLtq7;F~XV+FTRY_@FqMc7PPL+7x? zwFb+saG3gB)4Z!%maL$1*YyN5*bK-n)Pv?74`DNP6`jlWO^?B5;`#)e?KQ8-2%G84 z=zNOr%2R}4fG0K#;y3{as-;&s>$->8tOr(3J#g7PjIbG$tD)ApYVFEkv;Lby95!1v z)FEsJDOGN|UOEPwC2OYIS+>z3YzF9bPTFd~V6%QjJ{I3epO#`WYy7n(5u4TAd4{-x zYQtt(9-A?16R+aVGX#2-v#LGMV5&J;Ibf?8X;fU666kf#+Wg66DY!DxSW|^imSU-O zuG%bR^|KgR#Tu)dcW#8u2xsvMDtFDY1Ns(D%4I2mUgu2AQgCIWF)mB7)H+vfma_U;jI3h(Gi|aJs^EmP zcmn!o~ zteI+O*3<-q%>bRwNm~sVY}T*H$KpHbQ^ibXjlbS2VzcFsx=>~nS5R%(EX!jvX6@9g z_PO&6fnMjV&G}50A}eETwjA(-vXnrtb0%gfxN@qo<&VrzmSU+@uDY>ERzHi8RgBG+ zKTtLKS-gVEU9;=}Kf~B;zsX)EYX*9qvo?z{*etfk3T^=vu|c&6n<;Ck9L`!?Yq0DJ zhs5U^WUE=0te|t(Vp|5A0ojFm(7fXzY^JWFbJ@P>G1yF8pJ21)=TFg{roN2Mr}(Zs zM+k^|;__$tI8O4hS@)}a=G+W&mU`gwNBKDG0lpe)RbS1LF0U}ytpC(#HUo4jC*3bo8En?C$j9P4>689UW{tmcFJiOi$5}A5iYur# z?0)mbARuP#)T@?1`9irspw~HTbH0@`jLnw6qcU+8=ylH8{K;e~B6G7c(O6T3P?ln; zb*|bhW%aWdS;g3F`2$u|!3k&a3MzNavOG3p)+SnOvX{x41F%{5OI8M(#hk6+7FrRP zKeb2LOj$$KUH7YdgJoAZB-(3`t!7!Wg3euwZ5eC^WasO_W@O1;GY&Ump|IaS&yOCxoY3w8En?iMB%Vm^P^LEZU!lJZraMgV6$XR zqMfEDAZ!NcbWYl8z+kg}MLrhaNuzIEW<_4+t74gsO@ejIn5TlKs(Ssx9gmFQK@+MI9ZEPyKqHG0mPdnOxafnMdT&7Uqyp*Jon za|F(MzM7Lp^t(WwTIH(EQf5Dkp;fH0CR^G3EM7t7uFYCzYX)!?^VVc9n>7Qy%2}Jm z95zGMY{eccz@;yVy-k~&q(kt1er*kv!xq16dn+Dkyn=$zqfEenD&3il(CoH{Y=cYBwamHsouyW#ob#rpdWIcvj<*I#y zci8NcIPzBfnP_U+hk(9N;QSU)YTPu-IUPA{_6Y+wiFT$pr*GXXJEMTjs znlHPjY@7vpl`}a@5t#utTXs&tEX7i*T(w!s>}N5uiZr&IKWy`}cm<8SW?7-;5UB&U zig|0Ym(7}iUgfOKVjMP$IRn``;%?ay3bC28hQ?vn6Kb*S3WsTwHpo`BELlP2uEn+- zHUqM=^Ik0NB*QN?6Ez?TJsG)p>t zaoDW?e0_3o=HCXIc7bh(%^;0CbSfupHQ=yWzakrp@1Rkt z#T&);5=uY0xHwx}rUT0QZ|58|&hlaINpBJMb7c*=bJa)1J-S5a3Sq0p7e0ia(>H2u z|0jzV+hl(H&Bf`X*RQW!B))(3FZVxxb#?mad^`Q~<>KVw;coe<GIFMXo;{$`!Gj=ynnd3<*9;N{<%BFyWO)7kKKP(lGE4tI-B5%RA&Iqx_B@$*g*&EVg!Jh}gK%Owx@ijP?ycgqSK z1AYR>#q-ey&YM1${$UuulexsnMB?)IZWTWm@sgP2cuykcjq;7d-JL&ktN3udGH^zq zqt*nChe`i(Pm&i9cA9r=IaO<1O{gaYuJTp=(npHtS3@d|o9d;1=Qr;b&)#$nc{TZ$ zfAj9-&$~aJrtUF`72BnrkHmii-hH_H@>Z$IzL|A>eLu{#z7GahI~VdJUYqNR&8+J$ zyQ+16*{eeLrNrCM{l!Ii-7`x%{jhI=mJa@T_q+Ynxfui?eUIA;x}GwyLl8 zmcyZ9gtLh$qJtYBJk3cLx?^Q&JxiQ-xdk04w0c*3|~I(eyCeu@uWdI{tzU+p`eI}btclJUYVs!RdBw`Bw8QJ<5tf@hDjXbbb*xh48$ zGP`!CH_uyx|Fv|nB3(nTan`;UxRWk)HUKyyFNdDH<|GOAo~vtU9JXgk?(6tQ*#n2Z z^W3x8Trs1*bCE?;J!o&4*{A>q8RURw`RPs^6)d?|9UHAJU(9#scTqQd8mo<2adHD( z#SX{v)Y_Ao`2(-NqO7)=b-e^BuPv@mBCYH-sR3Qw_Ua5l(U7Ws(`E(cqzEkPBdv!v zO&-O$uFFt?p_Kq*y{65eOihdrl%)%^&8+K(?wIE68*0p<6KYwr`>MrEVB=?u@T@(VZIKWBLwN!y-W%nM!-Qqy;?6sM7-C!@GD+YV*H?JJ%h(KIG%b3loE@Lup_RP}BJM{9W!063hAXoWn-_V$O zi}AuT_S($4PO%rz9gDp-v##an02qg$|IMuHHp|Izx|wy2u-9hRwGDf1W?cvD6;GMv zvDaqSb(58dt{CjqOr?g{OI$$1IGb757w0*@2HMQJ?tj@#4yG*UUi0>biaGa+U79SP zo_lR(T_@*W?6~Ua+-o!I`ufS6*IkD~@-ErTr-mO==tonxhj*5n*v+hKwF0$o|E{Y9 z-DSq(QPHhq-?_cZ9ZiTUc;?vV@n%_fHsQKT(61GkwL5j(&%am}!>9jxbOHI8+T+U{D{M<~h|RwZ$_*PIYw5JkjU66-v*oybhh zjK`yj*laWFdZlI(y#;LL-q3mv+swM|vJ}x5hs~Cabm)3gTtSV3W?B9nfLS}uUYl9h zAkW6STS1Sh_BOMww+aTE0Ub&;qIpe5*i2tWjgw}MwiG8OY_@D4MsK8WrgVGLsdBcN zb$zDlXfx}&WKV&b;n~i!8EgioZxd{`>{>?Hj3d_V)qgn$w; zKHF7%=DYCyOlR)gY%}ZHwYJ3FWPLW<%(^~~r>u(DY%}Y6rDhVnVQjXUb=_ezqAw1c zZDw7wex_iv&8%yH&0@b(vDxxl)aaWQ>MD8-Z{OOPb2D*$^8UV=b-lPuf5pL0tB!9m zn$JnYA6T-~s$Dm;t``s9cyi&6QK(j zn_1VJS=R&aG@J3+{GDb$)0w-|+|0V>?ld>Eu8&t{!6I54g!<@fW+77Hv4|8nb)8nqQ=$ufb!!8fulRHcPp80Qgx( z!c_v!dd`|(F2SjL;tDEv{ci!#r=Y3SEqAHbnqM;6tQqK4&e{_dZk}Q+H3nzB<~Fmg z>3PbEWjS!>HMp5|y_t1=ioqG1i`&x}Zk}TQIrs#ed5_!7x^BL=9ygqM&1_~}AMeFq zsg|(WX4ds8HY0k2*laWFx;rZ#5 zapO;`Y+~F%F@C!r7aDYYeeuTP{P_Bi5Qu|1pjUirRt0A2B z78^#CAs)8gf7yRpDOdvxNWS04Q+J?hdD>fU2uUZ{=Tf`zJVC1s605P9oyXo{Lr4aT zg{}1;Z51C+HIuaT^@i%cm=z2KAM`ddKHbUD^=oQtqpW8Yp)rbysooyOr-z7ut$-4D zQ0}q;E5ysEczeZ9dOKJj@1!7!8}k8U#4HA^KX9Yo7B(^X`opY_9O_qZ6{vkUzDXgc zX~8^QqFH&4&AX5TNij@?+t6L(0RnqDm6mjVQiH&o{=q3?;UHbu&zpQt}vv{ z6(z4`3oY$&q~1fa>5UbH-5+^JzA7vggkh!kj7(MZnM11}?EXj<|KhFUv$oI+hZxHr z+xG95M{AIK?^tV%&O^Sk4j<#Irty?D#Ol3etu@wH4bCP1>aD`)#ldUh^FgPZSBIJ_ zruTlc-XfiLzPc7C$aB_IPvbi4NY?w#dW(#p9_&3VWMe3WAeJOf?B#y?DrT0{cZ)er+10MTYE5FCoWE$Jd5x~YPhJ4g$J_S zTX!&Bo~YQW|KnDH1~w%oT8(1#3of3(TXY~*C*qRd+TOU=8BB1{N1fW z^su137;!t9@x62xM3~CkIft~9`inM0$Gk#IXNldIdN}XBt#e4T#m3fVjylM%O#bAX zi_5p}r*#*-hYWRnntaF7>J>LDti)a!Ub zx|-Pw2;o{P=+A&nfk5|0DqvnlNmwHzsZ)~5zejoo? zwE3dr*tZtL_69Nw7XVUkeN!<+^qm?^xB$^3FUVJg85dAidT+>7MV|>WRe`<2S`=G( zRM<;1^Kv&;3 zV($%Wtk8J`>L)PrFt$n>&&U?2_l`AISXU)DR~XXfikGZr3oY$&q~1fa>5UbH-5+^J zzA7vggkh!kj7(MZnM11}?EXjMc*dYXtXyshNkJJwpG^AOZeVCG|d)ij>6 zhFHC~thL7as=>L0xiA;i9iR9)-JmNQGrjkl^%m*0gZc^VoFLCxQ$3CAtRq?PJL@en zf_kv`u#gR0&8AodgD#EG`%5NurwI(%)V-01{WU-NAHuqGGEum+gs(R-+hQYVri$q64WqAyGepuosHi zp&~Z69K>vI$$?ZkB4R5tm+QTvxHkZFMJ^M0Yi3Y(;-Y>6Wlv<)9FZH#4raW!ZU$wp z$k@uv#X3-4g4ogRxI*RaoI~14jrs|+U6G};#BNMIocG?=Ii%TQV{0=<9p(L#>+8k& z_36d={pgqID@D!8j*coUq~7k?^qm&aQHN=NDN;Q_6p|pgHi2j z-9|4EI-&Q7OzKVv=nc%UH}ZyD6*Rt(B}DHDnbhm3U_CEL@1?nZ0%8e5ILHeq^$-UF#zUCrzTgm5ht^k=}PK%jdg6)-QO(3q>40KHIZWM2F%+RpyyJzU*)JkcB5XRzPL ze->@N=s5PR#jw4B%)$kL)LY+F3=w^&1`{qo^vDbHRbj>jl$G8aGF8!Mf=pFludo)y zR-V;tU@B{b(0fEC^^l4{_eS23s{(UHNZIH;Axj0lADAiv-5aTZ=_iO&EPoW{k|VnM zwh?=8SYw6GBTzqqk%zHW(s)L;K)rXYvBJ74!MVbaHdnl4HCt$Dk0bRSl1*={Ang9g zJMvXwsUQq1y=P>qqR$*!1!4C`s$lvFK~LvmHnxJH%`y-#FPucHF&1?)KnQ0OWa!7t=8 znfHJ!+D>?Uj-l*}te+=(WBUyDd&_6h=8KN4&RnvWey)hUpFkq<)<&1?B3&mgPMj;i zY9%oZ=_e5OZ!A2J<=(o3>GDLyR%I^R6H!;A7+st51m2DID>9esy`s1`0CYtz6M1W9PL9byTpP7o_*nTt5M^1R)&c1(bS-3MlnDUXZS4_5wn{^Q3DSEwU_q$~JY zp>9A_D+$!cCbfh8(0(o$wC zp|-y5YOH0X(q{9cEja2ZEniQdMQV;a@2EqQ-sA<|QHLfyF9YOVp`?MNDQYfA>uqxd zKPv>=4}oj(yv^G{SQeTyLQ-vV0&fSVfz8VXSr;g|AZ3J_36g4CTwq%k#J`so)e|t0 zV1!m>0jsvg1+3bZEQtR!%~`+*ZE}Hn2CS(guxy?S1e;NAF!4yh+fZ7Pna!8dn&Qvf zhg;tsl7Akk)5>tx80nxn62GUVw^ZqT$Y;OTi;Anyvz z>jJhZYBor^LLCVtUBS-^t&<|L2ag7trZq-r&In1hH8TRs=4FGd3pCFN@tdJ$f|Lu? zJkT^Fuxy?S1oZ@Qhgk7Z7>o|Qt8Z;$b2ixU1XD)v>IpQq;a|H%$qZ?O)at0&Oj5?L7mB5{=kVAR%l?drk)@Hz7k<8o2*9H zYMq>^{HO{p+vAyvdol0YTr4n0)Jm1fNM1cb0xVviwbn!>mR2=Zjii+-858+Y6kM*W zjuPzvcvs{Ji_FoKP@2r;)e~4?@}eqZvJzRV8?HuILMdx9KgxoOb&m7mt=M_D;}%rr zILk;)R`co!EKzxpmNHujwe@XRV=W_hinXbfMqWNn7m(t3c&hPtLCqr@ri8i}hCqs_Z9DPlb zAy02|gQj%>PtVH&c~@v&7qCrHvq91o>PR5z3Vv2-ofL^Zcr?&7tuaD#Mo6lynGskv zFB@cCpm|1!-wZVqq+FopfuPW`_+=u+=5}u|mVJiLwNb7M6xRk<^?a(t3?Eh>*|A4tZB-nLz|iQ8Po* z73#=gX$BGUd9D!D6L7ko=roac;jL9}&JLTNVagC*J%Q$S{Cn3ZnPP)kYPQ()4BK79 zA4>!m<_Wqx-svRo%0{!^oHw>T#N-`bJ%RR?@G{3n_b91ji)Cu=*!GZr;2wVV2$yV} ztJ(NF8UCWnlF*zll4_GD_)9j6Y+i=QxP+VH2TruG zLIaaE^#lp*P%3M^$jy9?w+Vi+R`PVu3lLR;o-!^6Ci^VDb8_wI(XD zw5qXcB&}4*n8=T!;BsAclxPRQyCP3mWR9kU(qt~Lp1=Z=7gZUPmB?D%a5cITN?DWn zQ5IaRbDS4%#m>7Ox1ch|Sw?EInpaO?iOP$#l-WwCt#7*;YZd>S&d4YG-p-Ion0C`s^X&`BenhVl;+g!oV3c>b6;95Ly^EMEch31TqRGXZ@ z+kt6d^RhwK1xhYR8KGu^q}mo2*p>zH@1;fc1WY6tp;cMHs;zMWtF|Qz;y+Du7BE7a zT%eu-Yw8Fro96<-W|SLDJQDCWl$K;>^QE+=`1AJR7PM#@Uueeq)6<+ zqk*PrjS-qNLQ-wbjKH#a*&yoz%`-y$W~iAU*Dg^qL)suUJ8XD@tuEn@6&i+3lqGnyur%z6q~;8f)@z(W zgnV9h$h$(z3?gWXni-O=P)80+Gl-DSbA_OufYbFvr-{4^Z>@53cG&a`Q-<*B2{gCk z-@8W16dTM^v&E)o*zOwsSR%MEPte`*PA7R+Hk$S3ys_;eChzd-3ADF_mpL}NM@b!9 zEK_sGwuk%!_wchvxMbs8&Bou!@E2W{gywvaRGU1(U$R+b^D;!%1xjj28KLHcq}mo2 z@UuX8j=^y#UW@QI+1;; zl9;2-yJSyLXEK*RaH5438knr9CrE&=MA*tEtI@SuCub@@s)Eb*c&6fB%)2%h3(OI< zQe`rdS5J@ti`QqZHBpJBRgG06X{AcWM1B+nm+PvdL^}Z96?wuUb2KHCCUbf91QwXQ zsLGhEMAqtttI?HE%9_lNvfyH!W|%vM5eecRPo z%Sffo=0{s_)KN~po(y_ho3#dB^&2zHvUeAzv!|gH0O(?+T;oTlFcHU zmm#t)P*Ow62sI}p)wZ~Rp9R8m430zbT7c>Xzt1uif8a;7V>muS8k zc`2>D>HMe*>IuqOUr&&b#2jtjC3}K8lezqX6D_RJz+_E5K>~aw!d5m}jjq)?IaB#j z6+5zyc z$P*TsqbZ>@naisuu)yR+RmNl`vQ{@-jjn`J)?|K^1sCfa=fzvG^KQp2sLXMek(#XL z)e~5v@**u|wi0UV+pfl1Mk;MKKiYz$j`H&L1X`r#xbu!WH0e!V;2m{n((^Jv-W5t3 zNSdPNg0$W?SMak!u>BCY7SG$f4TNQ(IU^+1CMWQAU>exGY>;(Sg)&4J~l{cN=_q9%jN z>jIvhmj&{!(7Y~So1$ieq$|{sK++Zbtk60s5_|AyplMoTgyxKpR9iD6uxws7$htuD zj1a#WY9>gzK+OY9GXl%zxj;})5O;_bABDl_z`Od^CN^h-4Nov-1h1YzV;lapOO(uz zHb~758=hdROZa1jhG7$B2_7vh4SOP~IYXrN8fOq8pO+o-uFx`r2%4g1hNLUhk;Bpq zBINU2A*d(dbUo2&BJaXmtK6I&Ha)|XA-sA5&F%R2u2C|@2D8*`vFRDMyM{lO2rkSM zba%YdN#2!>X1zIYY6%;gW9Xkmp0CTr>m z65uNlwzA1;bgkCOnaYo<;IciQskj&OuFb^)b40CFnT+Js6C}Xm^;v68RAOmWW7SAn zsgf~~A4S3Cy6Py=4uE$>p0LOqO$nvRTwXna1tu@5GA1jLwYuSIbS0FsCi9~#xLD^n zFW!oscROxDWsb9q)MPcUp1=~77ilT8l~7yXb~V;AQfagK(H0za{3l;ephaqqJMXAN zliuV7-cg4pJud^~U7@6bq$z4HNb7BL1wShU+YfTU=mU7R0}o7S$6lkzj;YWdWG|gGS2yJqKdIqei zBd~0q3j~`{ZZPpkz}rw-l9|ny(wgGW+lO1^{yWo^*h@5D?f+6*dDHoQU+ZK@ZXnTS zSLWS%F5u~TSs?EU&FccTDQY%IxpwKX#W%jRW+tP3>H2=SYtW`dLp)I88MBd~0q3k3B9afev(Q5cL4ysK|*Vskdw z@B~vv@ahRPw&7p9M9B;R&|7gg;hj7&cLs;L*a;uqTq5GelaiaRw3cdD$WF z3N15;pebr*NV-BDIV{Z}LO#zGf_ef@*Aty4@-Do!%FWqf(=$vN!mB6H+>U?m8YNR~ zFiXuAo1S62YxrY{;KDpXcgH)OOjwuhL!!>cFI-V$Es*ytW5b!@Rr%^lkw z@(l^@&td$W|7Uy5Lp)}sUc;AniG<0TU@};0^vCZ z$Dw#F!rzc5EHh_-l-6WC{~W^tmlu6G)0NmuG+&Lplvdt!e$)l^1pirIPmqwr9Btku zdxAQXx%`0>Ev(SMWKBIm0(>RHRyJ9UuGKm@Q~6O9T(-wE758G^wYgYej;NI?laai7 zf&^H+K5MOsN-V8vtQtuxRWc^>qbRssR~;qV0r0NK6Be1HDWNo(%d01_z~n_$#$+Y3 zRySOYu7pz7WPX$d7wa77#aprSZpSUC%yE{Hnylv46Ii11A}wXM5^C$)uEttMDs47D z+B58&c=LXV;@S!ds&@1A_I39h;4!efr?-2*uERaOJ8ROKBQO5|lOq~Pw}6)gHn>B{ z1zBsVF;8{%J7D+LDH>|m7EWdkd>$_=dCrfi7+IPKZMC~b3tdNi!9Q?PWN8-68r;_WHe zu|auUoc;~DP*GFgjHffZKu&DRgE~^=e`vmv12Pt<pV8NL(|M6XpEX6vhGkv5{om7umwDK2(C_fqk-r+l6QEoS#Qn~+a6-d6yD9U_Llq~ z+@oZSEtaWSW7|XifqVF4%W69(t@%u|C*Cn<4UN+P{{iOwvGFM;4>h%|ExZh})kR7Q z*<_rWLpDC;U$}^$MdD>0=dL&Y&PT8=v!pcVjjY_{4Z*U`BAu5h(r!>vM9vO1H)Q2D zxq+Vz;&Tv=SMgd!u$NC*XU+sUvB?2~a}Wz;UKU82ujGJ?1!@M!iES{SAN`2LP9^z+ zFk05c9|g);t1(C2dZEc=!DuROC3nvvVXhKsYg?^G*?OU*x%}upPwZ4&L!@x8E@Pd- z99e5+CPR7kpRk5RagEbHZd9!{Q;Df{?NuXctxU>Hel!JFywqEC$=1!&x2t>C?p^zL zR_-o%_p}yN=BUa@O(yehwk=V4Beay+N`$R%yBcK~skGVrXbXzycpi+m!WUeKo4OQD zlDIN+yrr}z%X#&KmYU1$&vK?KA(v>r8gnVFyy^U?i;hc5*ig@)MQe^fcYCSn0VYpy zk4rQ?fYcE23BsB8(6tb*%1G6+OvUC+U5rJY*|~U zVCg(Jh&Hm^nBtLvyBD>jGrK@eY{~=fKHehvADXY^fQ$v|cpxXX!F>K$pn0MsSDx|r z^V;S)j@Fz7@((b%Lfg85*8*NP*x(NB>jt(lYF5a)Lme4p-NDZe%@Zau36BoiW;{k| z&J0<(RWk)k=VgVo8??_9@f)IMgq$1He9$&iuymdqL>KpQ$6a*(5`VPd-nwg3o3p}} zN0>5$ThE}e7605VN`}bWq-KdNkFeP-{INsJP>Qkzj~*6BLXp;-De@1{Icp+ zCgao`vhgYZ!bSWn5-;mG_rdYkGJAXylc7u{4a(1Y>AuG4Z z4g72npM!9`iq|TFy?nwtb0)}%O%4#8gIFN*vOv;&B?n|IP%}VIY=imy=!-7+OW+&j z_@e;#o<1RsIqKF6O(qLQQ?&R(t>oU6Ntmld+S*pDQMO(vX)ZsyqKp1`k;1*2dt)zF zm?LYg%w#CHo*@AmuMb;orV>-@+N(y?TA7rY{Ah|U_Em3*b`0Fx`Gi&GsLDu9CUff< zEKqsTl`>n2u=Q3b&GIPA8v?j~B^$eD{yok%0u7q5o z`D)CiwDP9&qrNGzQ?A&0GOvLH`@ts`x_b`r@pMNM8kVy6sbz14ZBXYZc9e0^xiTvqO1n7t#1kzyUt+w=St-|En$xUcSA3 zhgT?`Id$wGczgQz4)pfy+_hi7oJk5ME|W-~wN`n%+?kwo{RaB@c>1`ca!TQpT>1aq z8+vvf0GN_anR@mc*sW{7&OW}LUY(W3KwCO{boJ=v>5|UL)hSJ!zlLn=4w_L;$>aQ? zOMjQXPQ9H{#Q7s=pi7+pG)E~{+&^muBzNV1T6OjA-L+dk&&FOAiWf{=GDqBh#0gaH zY);8~`+9nJ1w(yYQahz|a;|Io&sHTX6i@Q1>5Y7e$p5`PQub4_=^sj8D6sxcwF;RX z?OoElx%>9_a`*Itq;T&KpV;Z==F!*7_uGnNS5lco{hJp_mh9KAuwX@$I&lR%$HXh0 z&b7~K`s)iP7iLSAVz2qH3o8zdb#CVp>4JLAtXd*6b1THJwWIl^*hkqF&ghihDOFta zd^|nUl?*IiWq+r|KDkYRVQDJ=PBvzzDpF$L>^ZD!f3JR?KJG-ZQ(q`z$b5wPkVa_L3MJy)$1*73hLC^ch8rRMR2__jIQz=ZRKUXGT zD#E;`A{J9A%xI{%)omuils;CAV4V1vG>>)q!>4b5K$zsmRS}Cx^Vn!O@^tk!ggKVg zlVFnbF*%>;^e+p~ZH_Q2vZ*2#lk;!^w=3}Zu3KSVL*d1XijZ;M| zraGiRHTzn)-}HL2hG0_gF87g?jLc7kA1@-exu z>GVGf{ryESsri^3kZBo4I4wb#a4KRkIUv(UXWp?HVQ#h05)#`z)9^7ZU+MI-Jt~AF zOjrk1#9~^$G8&xQZHPpeFMIzZ7#BWf{3o40Tg81f5hn9~Rm5V(e=-`T6sr@CFzwT| zAsAOaro{uDzNo{jqzJPxy((fcEgl#RiK80-MwqrfLkK1BtTMwqZA*9ay9AJg%SPJdwjhxG`vXQ?V;F&)ns4K>q` z8jdidKE5ZIjC{=DA3FW5!!?s3Opj<)#9|KrFd7Qj9V?G8BbukY65DSw@iCpR>hxnA zeenXyqlGGBF`cg(4a2XytwoqUyNeP`W^ZN2+g2~3moQu@ym#xY*A7QSAs3I0~F4AbIa=89aggH@cBf(_n zW9%R{9xJ!LN07B-A z*@mnc(^?g=m`R(B2FG^=Y9Y*my*UUbHy^VBoVM$E5MJ4A+^33I%m#4UklTe1A`4eUTLdHI+(e|7qT#b_e`Iv%l zboz-!mJUIfV?|UEiz)cVXn6b~`*nn=vU)$k6yRgljMC}Tk9~!Q&eCgC5sO(f%4n!x zbWIh6xt-!R!4%|UjH`6|bT5BiL)JV?sft*Pah1`q_5Ot32;(~RGr<(%V+?^h{SLdH zdW0!EOck*hL!i+xXHk`@2-EeE%hlNKS(uO6c0{N5NmVZm!gP78idf9HBSypQJ!SB! zvv`vd1XF~M`DN7UyZC?TiZErHsv;Ki%LpTMngMu{zJIVW!4&0VUT=frheZWaAk6zi zs))tB-exq^PCqjQIn6iM0D>vT$5c6>(-$6i70);Rxm6L1sd4}kId-+-Za`mt1w%V*5==J|+XKq1Sqa;0^nRw5o{3WPmmF&G5S;5yrtQ zFTs@JW5VC*^mW@V!t>2nDq=C=?~H~rhVB^=X7()u!Ib7>R-e@AzrLt50Xc2QZB@i# zR-ZH)g2#M3jWC}ob|siHe9YuS5S!71#~@7V%BqOPOg>~ZEZjKhy04-^2%9;pJS#m?Cf0Ve$ScJ*)pDJQ8OKunqSxS}1-E+j^4A*13 zXGK0{2%JErozNf8H;tF5A{H|QPN25eC|(3vljTzxf~myE)QQmPs|H>kj;wi4MJ%RH zgwgP>+o3B6v%gLYf~m~M1UM$uf0{iSk7+CFsv;H>;5gafvEo)+gekMdn_#N&F)JW8 zbM7?Gg{(=lRTZ(A6%d;mt)tT;Ob*An1XGodX%43x0eM$NBFtARVlmC(l;ceIW5p0A zMb8}sqvvDRz>%s`t*_+~CUGxS#A4RKk?P>`qxvAs$crHaQ;m-)0mt1xzO}+jb-zoh zh{cqEHD=nm;sZVZp8MR z>U_-9mpc8hM3sXOrj=0@v6!hZjfS=DPVGjRUXcX}rUoC=bFWVSBK2^*lb#Tzidam~ zy+*^dzO7~;Ouy!}38p3=vmHk0LXTsPBWsqlP(>_eJB-i+wys`*FwJ&(5=<>V=05B_ zTmC$G0bz>mRz)o4KI}ca-Jg91VFtTQB$(QKOnDg7wq^{#)7l_cRm5V-!5NfBgnLE*VkS}m~1=S5KJRJ=2fr=r(g};qgF{5gjwaRidf7kSVJ$*pB1kP7HPVH6dym_v7shC8P+eM8m^y1S8J+VU}b&g=A-mj5b?teJmL6|tB-=Z%J) z2Pzaqn8S_E5==WjM&}EY^y%OA5oSeWRm5U+zD7f>%?pPh%#VO)1k;|6=^UfehaMm7 zfH0w(R1u5m9Ah-xO5=)eJpLrIyB*taI`A3%%o$g zh{X(nYoeQL``}uRG3h!IOlLmE;g?Ro_^10BgsGQa6|oqHUq(aH==1xLHIKbV5=<99 z<{Qj6_Py7&N0=Kvs))sWgZbvE-Chrb>3(A=!F1(g*4)CF@5vG9K0fOnq$K-pj)BE|pUWhPd+*J{a$@d;+&q6oxVM^~Ryu@nYwmg{wkvz`Egx_~r=Mm#hi~KmQW1+8aKdQlvAK8* z!W`LEf?#^^F=b&dR5i4D407zZbWL9&r|l0>MJ(o0w9$~QLLk0aeqCfM!Svx{X2H?;*YVvtAWV2sRm5Uu!O{4y z8P7_h*nFH7OfY@4iB95GHE2Dq=CI;Er@x;?mO)#_Q87g6YS{R6GS0MiHl) zAWXZ@s))r@JY_UA3UkEw3CY^%?#A|;{(Q`LC^bEtu`gZ){h%Ti^Bqb}mn_-19a*#I zNM3>&z{e~AYf4PdlLJ|^;HWBMF$=(&OdT@5L714#)d^-GA5-+HPM@^w^o|H~C5tLz zF-4!kmMi0b(-9`UZ&!lx;$uF-@}$@fp9To?gNj(pM_8Wtz1)bGC#mj?C73~c%%e*> zePHTi!3b09t}0?Nk1iPvZ|XIyimZ82b|t}h^D(aXpxSsyHhhe9r<^KcF|PNFh9+S% z8X`=sWk(3chmTnY*4*CK0GBW{U#^N+%sQ|py4&x22=m{M`vl|5$IO7ueY5^A_8_Mj zeySoCGXpmFF)1?R!{ltOzZ1-0K4$ZCsG2+!h4(`F+NdHHv-!EvkmI&<9%M~{gBk9{ zcF!Svj6alRg|BLdFJT%TQbjDrAIh>uKh5ikFoBuM63kFOCI}{i*=xErMNSLNtcqAn z5KIISbvxlx-ZjHo5X>+>rax@C0upbGLKxrSs))t(hb`CW-cw2=YnI;hCYa%T%rQ6- z84wWj2w^7OQbjE07@UZ-pSKpLz?d>~31$Qza~PJ6v*zA%LKwfYs))rLhNa`afwK}L zYlbY{K`k(<}u8k_xJd%Le_YH zS4Ax5G0dLl?taG=tw)3(d# z8N3nb0LBm%g5aQrqf@Y=ePr5#-~t4Eavt% zqoJH%aE<(?7gc;3KTDL%u19Tx8JC#P$qExri_DukIb(=S2~n#fw zq<*10FcN__q^BwtGzZqyw-#-!jzDL|R3M-!0?=pJjNU%*WGez)9!pg$=rim_kG5KM z0D;~bn(&~yQ47A$x&wJdf8fo83tDi$;gRy<{#?ma=E20wQb&(wDENUnkfM7bxx|!@H@vZ1bWwms#wro=gEc) zBcG*5poW1T2xyi7v<1%1M-O}d7J(M+peh!$1rE+fc+dZgK*^m`J&f%~vjw0+@E*ss z99R4h=x-bq3n~OJa@?DmXc_`t?Ou$4<_JKSzQg9UbcfvtWY>eLSkR^K@K!?EX6X=U z!-e_;G*1jpJg%X6+kApc;hVnL1IT-)($gYpP;Aa5T6nkN7?hc)BRB$e>Q_bMM% zv7qL#Xq-9!!BYhKHfah0%@=?opbBM!*RwEWQMi$+SWpC%p$xh93r~Dgo^2$c1p-h5 z=&;L9?>vt{-p{Fu1vP*U+i^*s^~fUc8fOV;p#U@qHV<}%W+g$Otu?8N1xVi1hhl|>JH0P!}B4Fk?X4Vrz#fI9oDH$I&{J7x;f{o z5YSQqXb()~Gg_SQi9j1JP!$W>15^3WfysLzi;TrO63{XM=rh!$)SR&hTQtOts#wrx zC`xJcuzzC&3YayLfR+nDO`#uUsu@@fSu}1oRk5I^N zx85e8l>(3vMut{}+bl;GrP@YSEXW8WLzd>*afeNo>~m6%P?;s1sGO zpm%T{m*Re^ZOEeF0j`f@JM3x!C<83F+*|%BjzEtGQWXoz0L!gxei^zU(3h(v2xyG} zloQTlPF2mY7J)*qQ56fy2?sJeqmnj4py-lK2xzSUG#-Yaz?uC|Akd9cRKU19Wg9Q5Q70`*=%RV=70jNUz;M}#4ZMt+z7lkG>!EFZ5a@Xv6$?ra#T{jPb?k^h|7{5- zpp61hX0XU@jQ2qVD!i4dSWsrL=+}w!3lPZY{F;CQ1fZp`KRO$d^auiVPeD~IXesQE zVpeWEi9pf)bx&gZ(Ix>X1P0YUhZ}_;(6<3p#eza$P;HT{bP@z=9h{edHVZ&bP+9!4 zv2#ZRDs+*mSdbHx7Eh|T7f*c83soneEdo#o9Bz+3p6D3@-7ZX3EGPs{w`WH0uZ=9) zKD#RcZ54nD!o;_IRqK8TGy%NZOi=j83rNH_eNC3f+AttGUn=t=?HXl#}NVw6o86c zf!dV`J)9Bf=uWC)K}D_@4Gt+%yhNbj)DH+~hX6DPK6uhK|J=q1v^@<~v7kZl#gqI0 zd89$0ssq0h&`tqpE7X8zjd>P`Kxw?FiUnpo;Kag*q!b4nY>J&QDb=s3Lq=;lPgj_?%(E zv=#)kTLAKgW3)4)cH>3s}_h)*V^&A%dz{ zkTTXy=z04n0-dWnmw@&PK$~F|v|!7{%m}o;9#yfR&9DlZ++zej-TE80gMju4K$l-6 z(I@p>Fb#oTKBX!abos?3LxUmJ-XV*c)xJVN`vo97*of5}w+p9tt~yl3g6v=;*2}y8 z7z7#{5J5l(1fW2;VoCJ*QYU0l>rGU}f&$@=W%!_-cq-qaOa3&r9~~5cDnh}70 z0y-=JO+E)xc|`HV$f7nEsfq%<%2z2h_S^_#I0L8#Y zY)+N?{s^=unyOe(3~a>G-<4C(#za7_f-9bJKd zP76RiV1M*$!GCoT=)xzeVnID%f8>2`hYo=Xw`fa1X9S>1P$k-_V#~@1^gE7<1yzDF z(OZEVToLG4&`<(8D*%;+?SJc_SEmtZ=`N~bL1kh4-)3ZQysoS7vWS4r2|y=c^qxB9 zXCwmUaiuC2bOJ{2MbT$JW;0!Mc5#1_C)pUUJp^<`u zI(_i67x*+j$ug>9K{s9-4Ik`31R{`qA~yoMBmixKFA&_=)$}TI-HkXZ7PJXIL2!L! z=qvs@+Eg_1lrMps#s9oRFe&(Zl{=!Kx=~f5KxE!YitZ z5NPo(s$xOD(CO?pR?m(=*)vQfpeq7UHkc)MJ9rjEpqMx+7L*NUNn^HgK?rnkcmM%~ z3P8DG*H$-oZWjc)G=i#FP;S_@WqIj-8G%9{o+Y5G0#M@1@Xod*3({KbDUzMs@&{i0M4_({(2w8M$k=^syeso;`D*RKY z@7ASi1Oh!?OjRtX@K2**PyO-V5GeQeoCI`306GXUGIsb}8iCUNpeh!05MmS_?$-%{ zW_GAbKsN=T9Ppa(ktVtO(jid6aa6^Erol;Xil>)vU{Lr00=g>z&ASAnckZc&5vcGps$xO&E*T9k z^m*4J(Ceyq2ICR-1$NRhx)F-R;u8i0!Zs1fb+_aMbqr{A9?Y#`aXjf|A3*(Z|FKa3xIA zo+SzBp#XICj86Y){2mPY5J$y=uAYH@bnw9!WRctXCIs|I0GbD{r5(S&{y4Iz-UX^+ zLG$3fwAEdv!UGBDu>dp#Zt#E2PKPUD4iup(7BmEI@FxV<$2TVTjWY@8i2yVj zj+9G3-MSrF6g-)#SkP!VQl43*5I+65@^UKyg$Y0@lsEdq-CiKjfmc+;f>MC%Iyw5c zK(1R~=OO_;6@aS3&N}P9J$N5Fq%KvlpsKL5u2*mJ5d@mC@ihU33qZ%=gzD4Ya<`F1 z?EWM6`+SWsdZ>)gUW;C_^P?l=N^B>?4t4e#iP znS~JOTO1V&$^#qTyS-}MM;5h@T1h~!1t0@lpFPUZU?>93c~4a=$N<-8eOoU|hd}=| zJW4=s1fc71xYzT>p)3ewXhc;k=sFzkoqf{;pB0qY@qmEd3P3*a@_GS%FMN@aWhYg! zARl;peOFMStjMDK>3T z0VpLL`?pV>stp3w_oXTploF2pCsmo=6IoR0Mp*)i6o4ATtyaAapSvQE+fAxsL5<;7 zYxJ1>uMucQ$(95ZB>>%klACpB8f-=&k5W{{f^I;`O^z?`S0m87`91{nUI5w;9ri=? z%q<9XW&u^Pp#9Kc!)LDwK%fGj<`U2c0Z6HJFY)d*9=!{GrYaVs6uWmBTdN!bHEO(* zfIbR9L!m})W3N0PkVVCsP!$Ur3N>$i;Yd8zmEJ*BET|8Z z+;n=F#SMXur;H?^PXbVCIO*+CxE9{=9Y{r0EGRXc^ggY<5wE>}4o&_#wjX^KfJQ;# z@xcF9l}E07JB+GW&?qQ8-m>z-8D!C}utEg%MF4uaTc;m#s6!Ai$3z19CICgiwq;z`bfL&~iGETQ3yOekOQDXdeNl`iby!D0 z-vyw3u;UvS{Q;j{dv&BL7PJp`eAida_d%dRr%n>k4*}>t)FiA+JQUv!PdQChEa*Pe zB)C0!UL1kaLPE{=EB~0b>$~og0 z4K-9Gpg#i8JaAp*h0{kOkau;eVnOr3bzKhDZ-Zjg#lI~9{S|;7K|i{2zA(0^%}T0b zL64vx^lEFM*k_uq5ylUVnGd{81K-G<|&ay^?NTOpz#7w zGDz>{3pU+BpwWG(iUlQu^d9oF7QQ~)6S9YZ`~;x8P?He4@bq;AI(>zzSkPUlN%#;| z0T+4&6~94169k}ui;#8O>*0dQX(gzN1qEC*8d}9P!!>eQCVwQLi2~3G`=t6T$s5@r zi{8gkv7i(7lMNr7bEZZy@_d%+ZEQc9Bmi}TV!YAmTyTu4J*O%b)D4R9#`d4%jzEd@ zZUkf$fI0#wU(zcLkVO%3R4k|?fVQuhimx~`uWvv=lLeqwaIyY0$@vKg^dXLl1+{{U z^=`qvuOo|6+w~=&DFRRl*etJhS*k~%-*Hqds03`5_a1ms27z3=OeLVH0+0)UB8I-g z#Sdw_QWXnw0Z@iqH;*Au?o$B-G)(}i4TEa=5aT>#QMS`m#e!9K?A{}W7Esx1C27npA*mw0jMFom^txt;T{O&ID)EJP(yez zvwUPfT$%a#rro>Pel$}6dI|;nv)>)qhd>!`Qxyw(3I+U2zn`y%Tz9)fE&`e*0F`?I zzn=0kFFwiNQj)4zP`MYdIN29yhd{4qRwba>0+2s^y0BA+Rg;iKw`Nfl3-X6=7k(dA z2Ja^3M|32hIRa22D7iVj>(~th>K;i|ET|Ba+;lG5$P-yqy5T4SnkxV`fctKza!1}F zP|ikF#ey2ZefR7Gkw+2e+0JDIG*1BP3b!qrv%a5$Ku3e9iUoCr+m`*U>SsrwjHwS2 z(0l=?C!CzL%kSrkK%e5MSWr(mIk~#=Ef9dx20@4I@h~j{d5)ke7L+y! zeiFj>>|JEhz6W0jXrTb4yxKGJ?5!mTbon7wu^{Euo)QjW_{tzjjkFQ59d?lblx~+! zU;ksf83^Q3ld4!yx?S*Yc)+ex$fDr&B?)M;0JI#=>++9dA zAp&{*Z9+gx1fY$0gANfLFG49J@qfriDD zC7?|LP%vDNcBoPI00Q;;N>wZ<7_LXFq&zhVfxMcxB%sX#&;eKyPs{M876LVEK~*g1 z04#~u9_%#*)iKW5>q9_W>`)zJ?R-6|)HpG`3u;xlhC%gSzXzS-PNBWjjSx zY^x3~gIld-o2ntu?fmmdtF{WvSAIQZw6F681Ug)Rs#uWn^C{C$oc)171^sps&^7^R zE9`NSWclWeK<*Q$iUnNEa=T)80wqd#Gt>^ z>kv?o08}2<&jmXqDvT_$pGj3Ls64Eni}?S&fGm3bp$7r&5`a3u4gS{t`lSf;{3BJd zpbl_@pR`xlJ_K4(ZxR9R7J$~kS%F8EwAB%4V|}V(L2KZwphG_&{33b7o$Cl_j{xKc zujg!xN_r823I|aY3-W{4b4t0z`yfN7G^Yq?uK-jR%Ay)iJ?(}-gI%bK1=WSJs1j4J zra`Xz>=Q;n`vjngP{-J%WJ_EL^WB%KSkOeMV?0v8buR+B-}ys8`vsuWu(K}I=%XXD z$oVc+v7pnivmWDD08hsQ|I79vw! zP^w}PE7p8Q7X4~QRV>IGmRs{el9fiFG5e+x&=~tWp-2-JQ*Rk5H*`1+}^Pyy#n1awvaTBb{?e|!1oP6YZBN5z7c=_VWY#%#h1 z$ixHB5zsjSNck+QOOJMC5$JCm6$?_n%eu*PcM)Wf-|goFbY1`|1uNC(gA)59(7-!X z#ezz~O7)~Mmoowlu3{e@+m9{?Ko0!c^3ivp0c$ZvXa*BAs!l$@$qkg~|{vi9X* z1R6fD69HWkfNsF;aF297@eTfAFREfeH{f>ozU%Ih2(;n;C<3}H0F~RQ)4w-3=R+1P zctBMwsN6>Qs=$t{xSV52mE{B!A^?qtPtG>YSsB;W`&OkY7Bn8dIr}#6OFv{$v*iZ~ z=!yVT30^2`Ga*?nWKoqBRK03JSp6ch=y2B<0=g;y&4G%t>8ZNlJ?zGARKjKbH=&(`MN9I8>N>zxeSkO}Fup@V; zSd2gu=QSmu8v;=MK)62Z9=Z#GuFR(@7F0jbXs9`_A|4q^{253-6Y$z~^04W6rP``DRk5HG@Y?mI3!U-)=xN9{ z0=g{#MZ%7+NQI`qkn4i4P!$V`gdJbB+uByhqI!ic5zrk0s7w@mqijUK76_E7FjcXj zGEqiD!;v+IBha8pZwTnF0Q40`2FK|Yhapf^BUQ1WuP`!DS zP}KUi*Qfjl6d6awf*hf!HK#*xBm(WPo}Ym33qWUK&2Y2KtRe^$RD-Hm&{Onvc1)xql;bUlxn-@eD zeU77IL7jHO3uQC<;c23#*LVVYBmlLATdl?J8^$3}lR;F)g4)8Z*1OGz@hjW9E2{|T zu>e#XK;Jy}o<*Qfaa1g*Hh@Z}bjHgq@509j=!pQ-5cWsK9yXtaKpl!u6$@$z`=gH& zpW>I5F3fmHKw$#Vh!CBAkJH1+2()n~Rk5HEAx1;dwyp3qG4%5f0(vR{O@`&x`5cSz zm+@MCp(++M8J1g#lO@5IO`Dr!`V!k=!v!GaCohZlXmb@=G^Qz4u^{C)FBjzOkrAc$ zx^3kM=$Qbd#AuV_$Cd~*X**T1ASFgQx0N&?&;X}a1oT`0QZ8*j2V@RFpa#jQiUlc` zwmsh*zJfr%`uGyi3jrt;&g(K8Bex>Zt-e&nf7BBeLl3 zv`_+iEdaHJin5-0^En`kj!vg47St9h${r_8T>^nTUPltp8v*DzBtu}x6a0A!|2I^{ zf{sHn44t(ZU#vf9<{T5-kKPJEyJ2yX@_8ZrzT~0iRK4SN2kae+QWXpGfU&Nz+he@0yF2X^0eujF zUc<%u-e*^F@x#mMRK)5>T`Nq}<@Q?$^s6flf80Di);N;BV^t7T22k9m)1Jw!?lBfPTRCzu&~Q z-4Mull&V9;ptf)c^0?^cczf7M z?Fi_L0OSVSmX@i;BUP~=H`ul$n&OibS=8~_FanAZfc)XZ`0!(rb&>0eKBp=c zYKT&F4)R1*qRvN3se`VsrzR05F4W<#$UjZmJT=*`&y(lRHrEEx5EGRWx_@)ediN6orEpQV7jkD+f zveHoiC8^LI*IzZ?K~*g1D1f@OO4|lml-l(?0gV@clrIw03E#E>S!ADeG=$O|RI0kL^bj1fYzt+!{af z0WJ<76G~MqC?hPl3U8W=U02yHHvvr)fJVRz&f|_B>5W{M)t#zX&hnEx36alCs zj105i-D-kBO^;9&3+f0X!{AfSZxE<$=0gNDRRDSeqj&al^KoTnyDU`2g5JRB{iE{) z{5wRBW9|~rGy%vNW>|-+qwyCBp2ShHAZM6i$2`A|UH9%@3;|6Sfc)UYHA}q$Mz3M zOA*jadsN4mEwn+Eko~7dp;lF0P5&{sj!{`{KB$vx1!~oYI4ZVP%4+lV(=X2usD)iK z(yCbk^ZUaa>kr53_aabEd#YkV{o##u-QAeq2-K~E7Xi%{fPO{8`*asp=0>0@9jS^1 z{fdV77gF59A3@!8Y!(5{5rBf=4ycrSr`ibQcbuwNP!QY!l|2-NcLeUawiD1?0Z91) zf^=si`yf!&+*HMalpi2i=6mNRa@{K9B?6i!02PI5C4cV)-w|l^WU698MWI?L`@Uj$ zGi7}BhJfY^K+4Zv*uVIIYvX&prYaVs{OrZeN-6QBZQ?qPKVv)H0s+Vi-dMjLlLAj( z@8hUgkQcnMzGYV7vdDE!cjYIbg#wTZd=WSK>HhdoqWf;DVnHtOMcjhELrx%2RLYtJ zv`7F-0;|nCxud?J7`;eERV*k8tTuOt9J5EDz(F1av{(Q-4xiwAIWG^kXtXy~v7qDd z3BG_nj;j!;`StMxv_t@E50|#3X1iWN7WKP9RV=7IT-tgc&Fqap-^;HiprrzkCtS^} z9oQBx0&`ZNDi-7kS2KnDyn_+QbLBAtS|$J~^ZtL;hgLuqwOU10EJ&I6`%HT~5P?qW z9ud%T0jL?gxc9mC@*@beBq>#~pl0ymUPfPo5rM|{_(?!31fT}ck2-jj#h12Cdr}n( zY5@JHaQAsb5oq+uOuu3~tiJ%X0G74yvW$L+ENXpybCfkL1Xr1zV_!1to>8$Auzga55ZBdX<3I z2|)eB;8$HLMg$_zQAes`LH)y^+`ZW#d^Iz@XA}Xg7l68eMF+CnJ%B)+dQlY%>H-#Z zIue85ryFs~`FCtT+8_Y+gx5tU_>9JDhE}JkiUswA*G2!jluV0a^f6Zv0@^45-2sd8 zmGKKj7QM<%RV?TZSY-F(;%NjrJ-RLd1qeW6;E;TJ)+hK5=-n8qVnJizkUahDF9#4v z_qZnkZ4!Vo!@*I8M-OrR)#o@W7L*wdj^5t7Ruh4)R5lXOW&vn9T%S#zIC(sB-Nhub$d!c+XSGDFa&KKZcL3_SI&c~SWre7f@VHU zn-5u3^!#4}+AaX)gx9Y=?5a~3fs$OHDi)LzUcZX`odQ?=8S-TR6Wd_}1)#<71(kve zmg4PU_PkWZf)>LURLZy)v_ls4oKlH^b_hU0a1`dT_RlY5k$3~RsBTKCVnI1!xiu)TQYK_k$w4;>XpaE&0Llx$?`kmzfpU0L6$^R*<%Jjf z$9pwC@K!Vd?G=EY!;;u7wFfRdK69I@SkQA=5^o6%n1U>tP$tda*nYH60GbA_>zM1} zL}XE$vQ))_rh)6K?>*N9fo3gmC!qZT&_Or~Ygn!vex)~PAyu)UgK!i!%(ZfyklW)D69jS^1 zy@RnX_;T-XWKoX&(+KF00JIp^-bvc)iXza@I4Tyj7}nm!n{9c8K+n=|CZNLtP+~}i zR%0sT0+n4EsEP$8hGdBHv7dxMK3?Yu=!gIm83UjCANmZxHqw3&Rk5JR7`Rx!ax?~k zzF&DkKt~0j(Qt#mrehsk0(vf#s#wryxWSJs=ZVXrniq9QRB-*BYN?M2K#?%kC3CQg zKrtFujH*~rB#d<-4Ikl>n-{Zl6VP!1XcL@3j#=;B8i5|np(+-%2~Hr-)vdN0xo$;_ zo`6mWKwIJESucIs3Ppj zgND0JL!ja5sEP$ugk8C({V045-PQXL0i6+mmO|ET`r)twf!g>`6$@GlS$A|wH~itp zp4aaZ&{+Yf6O0TK`*gx9)jBt*iUoCok>PmbL={ks$`t=fK<5M?Wu4Ctc#$4trhz$_B}hy3nL<$aTf$QWXoz2Fb9zQStc*)Fq-60bLM) z=0VN&%-1#BBhZjYs$xO&pk_Pk^pSHBsDG_y1QaX)t^KUi-#%FY?~f+brYaV+_A^wW z4_blWPnjOzML-t?pisD*xb=A|K926aiKqE6-9D3#tbvCo>k6u|pQs&;OQyt_VQw;7de~A4lLLnnDGr ziUqZUFA*(0>0b(g>P~jFi|t3D0?=0|FI?H|V|HXwr72X!g1$m|;RnZUc%?e_U48<( zDge!djo8LILq;M{^9ZV9K{H_^R({aF3&^4+jcO9mH34Wa+#mS|q~3x+^BYqY3mOdf zM`1amMj(*ueh&h=E&%D_IJ(!NEw~@OjiX{gdN_{mS3hkF1j>-ckAQ9nK&@c86_BIC z2V_w*7ph`Gtzfz3f5Wa10<9Umnt*N!Kzrcr!k5{$;SOsYLRBni54>I2WBPRM?!E&86`S#sfbIxDJ>aWr`E&oqYj3-mRKY&&ZrVqxVnHvURG_whi>nB9wOx4vx+eft+NRTId^CIy0{v-ERV=6y)X1Iq z-68;i1|M%lK=%cp+OQF;G_&Ft1Zs4Gs#s8M*ofuZ)*G*ablCc^8 zf$qmqv7nhyFqt8F%K^xup(E!L&_e-eITY{@8$M$_0yQ2*RV--vQ~1WhjIX#D@A2&* z0(v9>HHEXFTUFZ(K%hN$sEP$Og|ndRepk{Uiw2gyNJ<=YJ}5vcfAXNTB+6ea>e zDfQv@L2D35_l>GJASk82YAA6Ef!;SOLO@Ripsa9`pYKk*pJO=JoT^w*RyfK3+<)$5 z1lqZ)E&+uLK*~o^Iv#C}7m#yzQxywRK8n)s(T_a{RNti+0X-9d7Q!V+hM~!Vkn0M$ zQWXnY2$vwk+PsTb)Y`{LK+grBL_c);-v!#?W3*{~sfq<9`T-SXHCEwshS{gq6VM9* zDC1waOzhbXUnUMXO;s!?<6onpSoaEU$aRzRoF<@`0#Hp@Zk3Gbgk3i*FIBOiny}nz zcrM=OMsH4gNNZsWKqD+zXbGJ04g;bet@q* z$DSw|_Wq(O7F24s(aj-?MziHCgThPKM8dCZ3@x7W4+*<|tnOw+?~)bGIj;cLLBuI8!}SFyD3L zx&e8piUmD{Gu0GHbKmdATxYy&UiUr+;iLc@HWVqJ!_VdLA z6e$3Ggah{Sy4?8F2WMVT6$|8{+Sq9oux1fZhv0%4g(jd+&daKyNlv6$?^6Bi}zwm0hTg@sHCd0{VdK823cX ztL)Nr7#>C@q@(|sTgRx>^rk=84}VO!QhKUlTcy zN5z5`!U@&51C9D4(3_vr2t-mZv1`EKPVpdSK|^6qf8z;nqE=vN#S3sT-4Zg{;dw#cDmH3Ir6 z0Ns2E%i6Fp84>7i92Eulo&IWhmpH*)nB??mkDT! z05lIktA{-)jV$^aN5z8X0jTVe%omYG@5j9*ps4~-GuXoCYi++BfewwQDi+iXw(yN( zK37GcIxiiQ#rC6V0?=64OtqN(0hdxYctuq#Xe?}|PPERBL0zjAAfV|2(8c%gk=u@; zxRkn-fvQ;0#rH--veXXvko^AgngldM0BQ-frap;gTtlwAu!5>sP)n#aEw?o*t_KeM z<3T_(1)$RK3RSjqV_PAR-(RX?L8ajpDtC{q^-zq4wDTjNSprZ?I7Mu+d^4`U>eQa9 zSWrtiMO>7*VI2hWJiMBKW(z=lVB#AYy&T^GwLC&qET|7meDz*WD~~|#nT`|C90BMA z)Qtoe?TlZpbjnOsEa(K(jikv^4!=-#V&o$NnkxXEfkTOWd5-=@F$x+*RV?TX97>!! zP!JbPR=)p}faVE6wP2d)Qg>@E1giLes#s7hm?oZ=>xpOBg*7re#dg^F0#M5>aJaY0 zfLB4|Yf=>pYPkhouIw@jzx!}wQ+WbfAOO`otOnVnH=e!wfqy6Hf0w z_N@tMp#U@&UXh!5Cn^VWowoy3v7ou|irkV-4|bretKNMu0WA`MGQ;R?_*w|Nu7W33 zv7pQ_dM~>@sRRPKpO{ZTiv^%XP?K;{{|@hu(x0R%7PJUz5_S!FfY%KBa_l0YB?3?` zm?jofT8XdEmgb}?7L*I7iEd}#c_NFdPP|G$O9i0Xus=GsKtB`3sNN*1VnMZGf3z-{ z{tE)>BccdsnE>Pum$s`r588x4St6;51^L6J?figMXA!7w!xYJ5`_Xa%C^ww+eskP_ zzX_AQ5mm9E+;Gy{=I3qv`c<|aMF?nx05lX%sM;^uiFbT4aa1g5D4b9kXQZEiEGpnw zkAVCIpaZb|pYWzLe)r*f92E;X0NejsWisOX?rhzA5ztBj=oF;4@53nkMXx_`R4nKe zr1#ay*$X3!N}V+l&?*6_BXrn?TW+U8$x!YbRk5Is&|y0+33-MEb82=;Wh-apF>qF=sj4p@X>-@2;}hY zDFLk&fI?rx_kq*>#jnPviJ&SL6#5#fXL~!$L7@IE6FJBBqjdsML70z1!u6eyMfz4$ z#exdLd{lXK&5Q_iKPWo^trvh2!7B;7GK6B+t=>gdEGQAYlCb{%JTC-#=3JS8HV8oD zKIru0dMv^Rg9lPj6$={o0d9w#ujBiLGJV<;&_)611r$GQES4VM;OFj3RV?TQ6hC}% zE7B2JbSh*x0R;#^i>cdKZM4;K9 zZW7QI0Z4hF>~YXD7vws>&s4>Nlo!h89(tD&fm|DZBA~4T&}2AMJs0VZe^}~u92En(`rloTKttE3NfFy&w+TR_lO@$Z8Cc?fblrDc&)@q$a48LBW^bve$QF{kWM-4> zJwo=*9u+bwLRn3T1`Va5LOY`Dh7=7cT1u&Y*Q@ipeNWftdLMs#-0q#{>p9Nzoa>zD zKh&bt?mR`IBT811gRwUL2vV0H%ph+86aXhfSdB7%!fGq<6omrdWO)ABAp$|o*Un;) zj{vHI5y(UPFMdERYO;=}C{zU_kUge^;?lL47&8X>3ZQtHK6Cl=dJ=-1V|j`~@i2Wh zHe(qcuZ!q!&7g+@C>#!5%-0`yF7^5Vo}y4V9J&D=MjS_nF7kpWgB}T>L9h@eTx+c% zYLV(Co}$noSO_C?V7V)Tl1pM4&7LFj@r94F;Q|Y7J@@Q8C4GNWO)o>jJ2p}~m0nJR?J`zESVLU}4H7EhC z`27Yi*>2rUl|g|5XcpXp4u;QKg`lsEQWTm6x1b6=^Dn4Hx#klX^jH9Wg4h2Jdp_dd zNa-+7QRoxA{@?k2?<0aHr>ta9kO0zdqF}Il`CPm{MLm_LD5TxQO7g@-7k>)v=+L7K z3Kl@Cp+(1s{0c^gZpAR3qR?t+(NguYm8eBd_BR>yL;#(Iv7kTwQ;QK~f0d^wbQZ>f z5}VG#gSXc!Ll_hyfI7pt_q(+!c#dyp6;Dy9GmLu=xiL=EzWSe| z0cuePW1gZ=D<}?+ytfoXJ^U&e6efU%!^nxf!oE8QlJ(~)3Jr&mll9Zi-9s(v->qZo zh6^@a03Gs&jkcWEdLpP-cb=lqA%7Tw)IV_tK^n(~Gw7)RnhTwilgs;;B51*Jo}$oP z=$v#Il!?DKvia*g21N)UeV7ruux3B5%rwpBDGKSsj97fz)oQ3kwUc);C{h3ogHdAt zBcm~tJB6nxGz>yKCXTchlxR z40F@hvMJVl{l@Ed)u_pn4Q zTF`PbgJK1ca;CgNcT@Qv2pXi!QxsCpgkF&C@rMZ7vw0nZ;slT_tWUX}6SWLMbGGml zg=}GcO4Q;d_+A$fbCN;v0!S8yXv|GYS0czNmZvBr3qv$Ib7WFbiymt^Gblj--Ge#4 ziM{sW(>q+7rzms}=J<45$txnr!Xkn}i2~>wjBcg=-j;$|G}n@+DD(|Rx0Z!lS0ZSC zZU%#r1khF(7_@en+zde*@_33uTVY^ubZig&eaW#?|1s#L07`*Jx%%cS83;0(##0nZ zfk(Og$ir(&2Hl*0n&lsCBW*kB1yQOKYi7G^Gp z!UGz)L3$QCsd8vP#)o+l{rS~BaFs=t1^>_gS;zRv6E@g1pN{WW z_Le+F|Ef=((0o(aC#p$xv#0Wv%B~be!dLd5S{G(w#iRwj4sx`kB8N z^iBYo1;TTGyy|TPEtthq6fz66l$dz`ib2rl0OfWK7u|aS)By^-b)2%!ASg1Brzq3` z3cSxAeuG=Irk6H@J_w)^FBo&H+@OOX!`?hap%O3XKOe8e^K_9%r!(lI0P=%HVISr1 z;?ekb$9RfDey}La$8ghBbm%(dnlk8<05XQ2%*%Ow@qD*j9#2uo7WP1-kHN0|cXAN)o#mjm&j#MyM7qR>D1jb>cj&>umq zCiP;_7XdWf8(yaNTdhYeDs7aa&~R_~Y`WnqyeW6jyW<#?C4gGNz|nr)`B4a(?ZHzN zY6SyFKfiQ|LoGV5xQszx1&}Npx;@H6-XUnK5>HV`77ktf6^m>H4cctZplkso2UDEY z8Aov6ti=|dqL3U+ajpwLjW?`o5r3UQIRYpE9uu}%o91hDyY?@9)&>mZ!qR{7XOG!p#AAJOk zF3V+5u>ewqC#?0RY`n2amvWw>kSaW3L;SjhA;^ASvknc{(JukC0zN=cwkz)(YSEGT zJVl`u@BxAcdmgPqP;#IugMJGjWoVH}|M$3)AN-i7D5MN6>KX0jjG(g45(bqBpgu4Z z=63A{9yluN!c!FL14Chwf=oO}-sRvb29*jR>7;GHO?~mn(9E2tC?uV<{UUQm0kxW*`C>Vwh{V`hJ4{E3f+X7?RLYy;JMWEhPN10E`Y*ewBS?2s$TYl6v7y1%nhmGw6>1ngCOr z%gRj(5p+wDrzkW5rZ~Utxq&;7y-ceZR4IU3!1JizSW{e4R^2E?p%(BwIvEh!1Dy<- zF&#TLT(DIF$O8uVCQS&hM=fd}%TpBcfWbYBfwp+8ppWJV2K^O4(yt`62wJ}oL9MiS zibB$_BuqPyG#0gJnDu-HRSTeva59|m8K;S$x<)AqZG@9SJv$c{S45}4C>)`Zu@e9F^iQ^tTMWJ;&s~2%4hJQxuXe%-myZ zd<3mI$5RyA1kjg&oCyeeQn-#mrvy+D%$R(rwNymV`JX&R zp(2ZtMof%{yfOf!2 zfy*g6Hxbm-gQqC816B&WZ!r*$*HyNTWRRr*ngqjW`)|C(7i?M^o}$ns7)EP;{6i&b zQR22t23ZN9Vt9v5s!_(>t=rpqibBQk4%=EwwH`t168YxBDUqir z)Ej=IR@XN#LC|5%Zk-!0*s}si8zyc2oi>g`ztM3mo}!R8Oxli!e}$|5W?PJA&^ZB= z25X4oZcp2Qpq`dIMWHlUL!`00GcK6?lx@NwYXP(c-uMO&*p7#zy>fVpLR;XCueNmu z+}-k-u#Z9K1yCEfskTaw^_X>k92%rhi;Ujv>uS6nf zvIkF5Xu@+Si;CWf=TaZa`!ndG01ANaR(#j~I}wzwz*7_ofbP~n1Fag=q7$3nFzAv1 z(t;V2Y@hFe2r}EkQxwvI8IxWW`X>=on(&K3HUcOIb|k26)3FDFViI|ZLOHM_f$zjb z{AyXK*{Vy!b!01m7QhBUSBx)@LQsYlPf=(AY!IY<@aGiNqQJ}A46+kI9)7Tmd|XyL z1bJQIDGGV`SxO9@*5O`|;os>Dx-5XM!CSff$JPT8G^LuSD0B_p%B_|@{E1p*w|EnS zt_YxcaMqnFh;D)W!8HjT4!&T}H38(81l!0DoAn$) z-OYK5LVih>lI{H*W6_~=$;xKXbpaIp7)sX$KgXk6XTS0kg`yurIY<1PWe93FS*B~l zb>tv`q;KVO1M2pn7PXkdQxuZEm48k%!$X}Wp1m04D1hwXMQs16dR%g|@IFsb$PQk_ zI(GE5MlJf#RG&dN1du5dKkQt5<1lKGpDa&N$P|hnf=>2&i=gzC%NTT108NFezqQ`$z zFzB`b3Wo>ZylX%4s;JRgJVl{!c<{Y!y8IJrk-c>ggYF0*3s~A?(sw$3o49nIrzm6r zOM4X0o^?RbkivHiauPs#FuEnHI2%t?js3|}6w-syt!Lv~wnfmNiRBD(7C_CaU~|r{ z?Jl9;D94beDAcUVQqt{<*JcDI-)Yya;evG$K-1y%|JlmAV+gwG#8VWS4zK@7U;RuF z6kR=-L9PO*>u=~1-ydm>pvoGaqEOf0aIcHGeHlSvCbJpjCVRHKrzms>_9~HyaVSPDQapK|K^_9=YB0PXJsdI#L8_;Cib7X|EhWb`XOtpn z==V4V-4j4MO@n6xsv@Vo(yuX^zm-{x zPKKs0S24(20PTbdjdMzx#}TCRil->F6Dl;^AG!TS&{M5r4Dt~`IdC$_yr{&56L#7> zMWGxx8RA=ew?NR^OSc&0D}W-PbK>o1l!jU~+J>hn6ak$Rhi7AYBBQEbj2c1-Z^Av?D;dwM+^q8Xt zp$E`8iCOXnuN3&RZ3Ke?1W;TZe4FEapLGcOww1Pw(7DDGE8k+eGB%v}LG{(Q)W520dKLb9iPqoLP6JPI?$zV}`9JFzO+4fc zv|dK3qvmUm<0<-AnRr-ArUtcpihj^+7aR6h1qqs82rs;EtL0n}6y(ZN6e@%l-g7G6 zc$wP(SziVP3m{Fn=+-7L!7o$YllxQ@Dzo<$XZESbv?2IwWwp)<~79bo}!Q!3|E%B_nCzt$)WxX3KKvoP)WXb)Zcjs(mu>n6jFgo za^-c+yC5hzV={xn1<-6L^jf=fB0jx;XYv$dhiZ!r!)TsPf;id3RLp{U8_JX3afEu zP^17F3k52jb(dM77M=aaQxqBt1u8v0j7me$vZav>iV{HbFp=ThLJgO$&0WS*6q1LD zjP|QK+(3{dK9fPu1d#Q0*yZN)!9A!&y%TtfLe|$|La??U?mv60)iUV005XCaIm5i& zZxM7~ou?>d1T}KKo)5<3Im1tM>(y|>>um2p~^5bm2GltU{2p9#2uo z6As;yxOtff8hdpggJK1cV=0surhUJPpmx`Iib9U1P=#LK{0D;Ms;)37P5@nk>uB53 z*Uu1C+bBh$OK=@ct-d@QLBWgs85A#o#=>ZP_em}%5tOi)rzkWQM&s3HULAy>(ZO#R zlpuhP!)Sbt+o^{L(tE;F6gm#0@oOHG(3ZP3cPnX>02YyV%tMC+sF2OvV z!RQyb7;o;*R=pdpqa*=z9^TqCU*_SZ@w<2N6ot;iTbuoc9k|^6;X7>xy%a#wZwVPB zdE-NO<~>hQNct_I>Go&aqu*%eP$LGt5TRv7aXg$zPq2bCq0$DtPO4t~L)6akb6@1c3U zXC)vg@d-~+C=cF416sAah9KLnISfh_K(5dWy7(^+KgzAU@f3wzp%>(~EOi5da`(#g zX}FHw3ZTA4Flj5d827zH_wf{k`W9JAG`&WjMv&~AUJQCCfFv+nspP7zfeu|+qZEZC zFkIPrhC zKqF!0S|=qh{5+aBi>D|w5>~D~>T?S(yWag^HiNzhpeyh;p;oSm-;Wl1@f3xwz}tlJ z&SrR5f%@j#7?dS|I>B(|y`*@2_io>krzq43hAWR~Z&OEy?%(?J4Eid7+QA#2qEcoU z`i+z}@)U*I!5g3Kukyh3AoE-Wgm8vs#s>C}a!IqstfG&PRvN}CK1JiZcS)lZh05SJ`aCDK2ZEH_tY%QY z0FwB_O6q^Z@eqw_Tb`nj#NSflvaB0k8oz1RF$NU~phB1f`lex*h)!><-8@C1Lg}6` z8lyX+7S+AE#h@Po$QMQ+m-%;JhM?ROo}!R1j6nABQ^j>7s|JQKs89echdaZD>&qh% z$Uir+Rp91I_+`UhKT5=vi-|TscLf7E#9q+LNKPEEjsu)xx zfPTZCnF}l{@kYdj^*lwP->_$wup}8^cC0=$8Phgf6i`t}$NPGdY~6C{zhu;&By!@oUSg?h6?7TL2Y6FUY^p z@+NB0vmQJ}p#tay#Wb0Yf1~S%b}^_#02RU;-^E^Mt{~{#VVz%?^8}vnR{Lhd zpi%+!748gHwQ9Bq>iV6hDD)NX41Y&I!~=s(ruZ_bOaS>xcOoi}mPJr@qZEaFVfobs z{Q~?OC40YQP`LmKgICMRC)%kYDA0$eC=>>-mKyhewMM6RLGzysst`bstBsy(Oj+JI{G7kn!&V=%%3Uvd$=!m@)U)d!L-h=wq5ris4$^F zgDM5k?g|*9d6aFq0P=udP_FxfO$a)nz*7|RfL_p^ z+$A#*WV*hVL3ILX4x9|p-yawv==cVnqR<>T85U;b+(eLaOgHt03$|VWRlq3m^Uzzk zK&4A8Pf@4>Mu{^rZSgJWoaz_`9hd*_#Y)npJv}X3=b{$n_TwoENtgCKm!#lAFL{ea z3_2lzGGHFI*WFdPMy{PDPf;iX=3y@@#kE3*uGi0f3_2-*te{M{e0y+j)S{wBDGFIZ znQs5(3YSe`kc9w}z(iGI)zc^F(1pqI6on)(Q8jt!A`b*9uKmp* zO97+;eeYL$W{gD8+eRr0sX*WR{>@|f?!6(rl}5vLWF>&Mz)kg{QCVjM-G9nc6xsqe z)y{`^%|tC~-E$y=&Iq7IaC$G@yB;r*Yu$^dD6|Mp?+=~kXd%eh+=xME1<*41jixb{HS_Z$7{MfWk2zvf`6NAnPpbs$nZ|tylF@gfpc#1+FVD>+-S;z3Wi=#M3cV=S~`)ZDD)}VQZlS)oGF5`J?=8-ya0L>DQ^&9)nhe+p5Nms3O$On zl+4<<2+x)ORfuNL1p(v$o&1_65i=0fPKl=|V9*r-G#~m5&Cj-*iJ<=eJVl}T&}WD?Xpc7{9;Fh@ zAbSDy5&Cr=TVF~L)T0AWQRpM|>$IB<#3PUiJKr~#U849}w* z_wP9(NP84dQAin{M=7$7&Zv&j@Y)atImn|r#y7X*^tC!U;pTsH;{W+S>lpXml{a{> zJOgi$)&^sm}`7rM1N-|@B3uh|^-S2+rrFAIw+mG>{lRTiV>XV7f{B>jBM zowuP*2)cEKrzj--d`$IiRXprpo1e#^I|66~TnnZbH1KmjuYjj0Gy<-LcE9V+qZW0V zA+OnR(K!jA)vzMt!p_S}P>b>#r6{yox+0@hZeIlLbM40EcyRASjf6oi0!aCvyutBu1zdOS`j4k5r2G#)AERRW0ktT9*=h#43LrPA z%$%G!2H)ysSMU^t+@LZOwhzPCQSsAb4001dyI^Az*^$=Y(V@$Z;3*32f{jgHtl5dj z3i8!%Gss;4>B5#rt2TbdqaWYYd5S{1u;tOpjP~Qup|i9OW6)gz)DH&tEOo5$S-0Rk zPf@5J4DMaDv|flHjUS&GvPMH&dwG5X7(djd!f zPKH)_$!F1TBr}7jC?p3bL+`&A%Mi5WUMH=F3)WKr6~HZMN#`wH5j4<~rzlhax1cM{ zDlQ_ZUT!3V?hBwzP!`p`_54E5kdE7 z@DzolWx8?>8A+%`242mz8?K{A0%$J0we=i08_xl?_vR@I&4sr%;|Drw2x``90E7Gl zP-LpSLB{r?cc?{8TJsczB2z6Ta|U$8lT8=4!9GcDA11W%7eHwMP5dwc&zNl5&Qlah z1E};wC4QMYowA-m0Rm_slnSII|Ibj^{#2f#&_E~^_&Vgh96EGA#++tQpa5F!4_h7; z7~`92_E?^x&}x6ER3Kl>I;Wx^u^2BQ&X0PEX3LS*s zXnE?;ZRpTB#nm$Ci2&LHGx*06U*N*yi}5@~p)D|jA9Wyo9)hM1={~UG^bQd~(_jwB zFz;vqI&{N^@)U)p!5olDr!(CVG}d+ugF*$6^qX9cpJE0eXoww8QAqktuFaZp#}Ra* zbP3qy6m}^i%*9z;#rTxDA)->Id)?g$m$0YT^_(8$s*31~4c>01bq*E?~pL z2dG6WyYUo-2Etia(9HrrCRQFxVNj$1x(&-kizj|ukDwLDd5S`}VY#Sc@vd{IMU!%V zGbl;`*}#VmzwIA|=YZyZ<0%T+z=sY?=e+kp(6I@v2Q^$r&jip{SaGvmcUxc7qS+Fj zqR>}Zar5uq3tW>B;WUsz&jnCchP*-D`VYnka(Ct_3T0)$7UiSHI-(Yxl9|Dv7XoMo zTt`N^I(T!=IZb(rLNnkxa+y=;jUeSEn;8@>fcn(H2XV>`r=u1XHA+#aPYp~6>K30t z(891Y42lszOg%CjgAQGrH#rQ77eK3F z4bgRph7D>_MxzvkR>K;i8M%LOja=l&raBGRQGx*409zh8EoyLC}*m%Ng`i0C~gcmS;=bTL`jR%TpBchS9A#PK(|kXkXMJ2E7tM z(*1#--blrreA8z*t@23E`mCm^Av?DVRd1MryDL%>HhgWgWd?Br|=uKE8B4!K|RuV zib7A}H!{%I9E4irY*4|V6ai!pZ{I9ymkUhMWj~U;q7(u69+7E8HU{eLq zco=6`la=)pL6=>5ibCUIoZ->kT|bp?@&WFzfM>RiiUZ*`j;WA5Y$QS0)svXpciHG22EPX?L|-@b)KToi!xZ}wRKw|f+nAOz@U!; z=nUK$+>Yp8LePcNJVl{1aAz=%Zi;(BkFw$!^hp5KeUmqsemt@rf&#zt6ou+wU(?w? z7UDaDqJBPuJ_{gQxCPnY{Br_9GUIuQLbh-V>g6zgJvwyhcjSjOTt{gFNILGFbhHe2 zi6fnOibB$HZ{5_#xEFM|sTzaQ1(39q`qJM6D^QC*$?_D1q@~my>fEDIi|%cl$e;`X zqzj#VSJVE^2s&xXQxwvLPQJ=Mw>bzhj9<;5OaZhPZb84>2U#F!NCHn$XfND?Mt`f= zfgsyK#~JiR07<_QnbTnt-XLg;4o^`?`i)2%7vs3HY@-PN{6+pdVK*QN@7G6ViwSuQ8)EfpgbnLg|^(lYmrZFg60GYy2 zXXV5{FHwu4=kXMUOkt=~_EKkj=p6k1GAKs?Nk6=McJIRe2s-P}QxuYZcz5gmGx*i= zPxsD48!p&y0!RkdKHLqPh6iuI^x!E9$-vr&0LL|h(aF%+VkCpU3m^liV=PGz#@qI` zvE(TV89*K5u#Xz?2-=#vfI+zeNDsyt;@!4=Lx*l>9#2t755^hxzfE0@pr?kr8I&i0 zq~qRemqeK%$afM?QAj%O-G2R3KLovfV9TI<0kj%A`6~V9;$y1+ zibDCYB;lal?N9!JVl}5aMlg4c!Hltj&}zzs7L@Ac|yPLuQ{H=w)5aA3K@CAgKx$- zJcWI)BGl zh=!^^_YtcQ)JvPEC{z(`De2NkNXHU*;(ag~3f#!Cxl< zK~4UAVNjU>Y5{i!m6E{O2r6%sqEHLCGprrzJsCk=7S}PTTmZSj1#9R#V>p6-H%d{+ z4K7&qqLH--`V!QAc*6x-A%G$>q0%W^=PQB=f_aKU5t(o@EFO;6@7Hx0%b-63s69L; z9E!WSW2hrfQK&sUCKUVs>W^BKwQDhhDg}@goZf~j&*UR0ayL&=$O=yHJq1H85ET1q zKZB|S&@;F*Bwx|M0~#rxd5S{M;Lad(zw28BbsB$_L4O611bV8H^f5nCi~cl9QAh$k z)y)eZm?LP9YXF0)1<*Bu+a$!` ziK>8RJVl`;FdS_eo)C`e7<1N?Fz6qyV@x{fqVM?1^f7AwWi$St|Fe#9zdsBqPjSY} z3XHb$6#c9A`@?#=-EVPikJ+o%BN{HcT0!$`;3?@_o`dhevyyp=LN)M|%;=JeD`3+H z4q{N90J4So1y!Jv8pv#2H%2=rY-yIf7c`@vP~{hU>^e02M%~z~&M4xX|n7bDpA50h9_v9(FZEP=3EY46+nJ z)1hR&I4FM*YEhgTPf=((l&p6uS%Rl#np%u!kd**>4llfS|NZ!%shLJ83O$Dx-ZqE# zEJZC^k-3~fX9SQUR5&ZIeU9HlM}6Tb3MoQ`v)#Kj+6X$Jf0#jM1<*7Y>DA93j3LMI zJVl{tFw%Rg`@^FM`gO~ZLFWWe0zjuHt-v33eSMp!D3kz@RfLg0f~uQ5VUV=|8VQ>? zjU6o6j82AeO?iq!BViM#BMXD@gkase4-7glfR4iB*lc$@M+CiI&r=jS3XkKeLK8g9 z5+6~)pbG-%G+eMUcD+Lp6dB1=6gmwT?C@qjcs<>*J}RRcF4&6#Xe~VVzud~f!z|1C z@)U*E!gK#pXRTi7H_|;iltGsS&{60FRz!WYMJ;N6jHf7c6gq*<7dqfMpck2Q7-S=W zq~F)Lep25JK{vnf6osVU*N}O>4|h%iMsH`3tpIurzmZ?84xUTBGlr)q^csF6KhO5~ z9cSsC3k8G!r3$r{YD$)ycl#@09ikSLa(Kzxc+KN zGoGT5^)pzC=uk5qK^-?FFzAW^3WmqTOl9{X)S?a>d5S{8@R-Q=9DE)@vtQ*i$X)<# zd?;@aF>)g=rPfI1DGF_T2=}@n_wZYrL`Oll;X1l1fRf;>vzI@L%Q*%P<|zs#!CBX3 zkuH9jYH6p&plbri3#QNPe23!}Wj9Ju$P1>=ru7+e7ah7*zb7*2x&ZnGQ!{~G$8JQw zQCXuDg?_=*jE}1_{ut%FscRVIAb@hJVKv8%Yj}ie{4}1TP)@a_BzyiaJf73d`8b0d z1(5n*C|xTZejK%^tqV_4Nd2#+WY87AKj_dsslCmh8v-acSKi>smo769su^@!0J*@d)>)PIxXPkMf1aX{3(RWOcQl!TTGYa_ z^XP^P_KpCO7J7wMF2y@~@;6oqP_yOp_XEdGt|xb9|w%4$+cC&h08M}@irhvHm(fv5v4_i_2YF$lU{62%~I0dxv>e)upw2hTj7FXbr;or0You9>gHwWdA` zvl!$ffI7lR@3xd`ynR8C2~Sa|BaHN3>^KOo6u1^p#~@z;R1S;&T6(17vu;%&Pf@5G z7X8hixaI;nbW>aP7~62cJ`_M3V4NW|Db9{AdL)2S z;d%6WUoSk=>Aan%D3l7%qqv`&>Jeo2c`<|h1W=$0d_DH()%~bNE7N$2LV+%@w^E00 zc&PKt=mQM$7eFszNkWILy_yKRGKQxp^b(dNH0xWo6}718&8rLw5I|w@nCQ^l5x-6J zy~R@$3WLW)z?6lq2%1zMz@R_@G!Oc9A$s+=My`94!2e58Xdd+Ie(5IqB53fkR2F$G zfRdm~9Qa{&JUVn*%Xx}INzf&(>AbfYf?kA_Fepd>t%V(rn}q4%-zYtlrzo@*c04xv zF>^VB_VsF`*Ki#L3!n=yz`oy4G7h!qc5j}d&;=M^zrOnS0tDSZJBUF~1d!J~SOMCr z8y>FmH2*0~{OV}+xd5S`o&Io}!SsGn7T$ zG{egnJ=JB$HC#uL0%)U>g25EEH2f$(romGb+NfkD=_NBl0kvr5(LM}{59%2O2j3U&3HZFb=bXUFr08T3K`Wx!Bp^n|u}IQsYno}y3& z40V>DIcbAhw5imQLD2$e5p=hUJruQF=UfWQw7iWNXXuw?s4i)uV_GT<>!Q78zOY#-B&4M$K? zk3S5G6F`bESDw;#x-4o@Ku?~ckRr^L@9F#l_r2R5R?%;`VB-bQEVz4L)bEK)1+p5Y zC^QT1-kzW1icpK(-w$O_f&i*3ft73P4e?ob-~&%lsICN-Y$tcY$a zHT~!pJdAc=6i-n|2YRZB+THR|i}Y`7XHb#=GJ^s3cAo+dqu*%DO`f8V84R#np8ob1 zLG?{8GU%lMvV-d=|L-2$BGslmMIk%5jw~b!cr5758ZQRD5~ujMHU zJ%e}H5kpTDpceUuCNL;j0HyqqH?Vm0KNT9TVLU~lls{1JZeutULC?AuFzB@adJ7}@ zv)rwgq2EZUCr?r6t#tlq@(#Q+UZuIh_=fA~jR1NG*U@C}?Rp4mahRtl^boG27pl7g zQH#Pd)fki_fb5~CI^)LqB?wCR!c!Eoho0)*(~EGwPSwDWL8$_0G90>pOP1pY-^K|% zMWM-X=#-Pz&qOU!zPE-!Zw1gVxYx-I9o7W3=zF6Sg?_=ku3C3zHG)3NpJ32C0b~OG zx}v^!7a%B9fu|^B0{y!3i39PJ!L_w_81!BMO@dO8=~4fz5p;MRPf=(Rl!Ex2ACG62 zKfVZO&<6o@5+-~D-?-o^n%B`hMWK^0;aeG*oq`UXdcSlAeH1{|@UlGR>3+Nxud5nQ zQK%YTmY06lo{ON~R@DspB!J}MWqFv&H9TwC_6$!^NFH96R~7$jf}lTnoedf;*v|q; zx?PftxvmoWjbie7ibB%ulEO|OABmupiK7^lCV)zd?y4V8@_L0n`slL2M^?HbhXyMxLTjKPUxR>Cx>9s$(pUeaN6M3TQt@ zy^Iij?G&@!sQG1Y_<#P-I>rI+uy4<}9^UA$@=M_<`d1Bbx0JM~X={O?d&862UzH_j z{wTN>oPFPKK+wSvJVl{Va4m$Fn=eLCPun5}eHB1c;LW;!_k;-uDsPmc&=h#HUgwb5 z0zr>|x0ujy(PayuOK=-WKe`-GWIQV2DGFVJ+lbnM)5{RFbFn6aas-ewlnOX*ys;a# zXyOu{qL4F`3Yif-=L3siwA6oSDU>95oqxhYtD3lK^DjQN(+yDO$h1KuqI9+L`?qW&9S(mRc zrmLcCyN-i&_H5muvwp|+y?b=FZeO?FY|z>*JN8uPjZ{N_;DYJ@m;3Kx*ye5l`loAl zNzjo{GU6%v2R3&bKA~+wLL`E+0{62&ut0Dmc6z{+j#+#Bi1_%Jrzo`3!&0*Non9P* zB70qB&<_D*?+urQpZOuwqUXJNibD3@mXbGyALS8r|4aab3I$LdTz}sE-r+S~x6bku zh3eq?Q+*ML%Rt7Kq%i2G0FoA!e?H`mi}&?Pd5S{PqHlt7OBnC*@6yTgkS544QFey02&JS$Xn~&@dAqkf1aYyP`F3_OuI53 zL0{SoV$d%EWC%+};s*r}KrM=D%Tp9Gge4=}optbzdJlHZV9;*?R1J5OPk-!MA?WyS zo}y4S+)=vDy#Eiis9W-829*e)oIL0Ob$O3>Fv@R~qEJqrrKCvf3a(N8qH~r(r2^GKqY+CX88aq^9}s73Wo;~7*XfK;J_AF_Tu?s)cd;VBBKLI>Y; zr9lvaWSgyE&|d+R1yAO%k16=c++UujD3k?H=9?=X`XFfj*24^{7C@DL@VSho@9}1Y z!?*Dig)05v!v(`y;>FFAlN}jUBY;Za7SzIT!4dQu^?1!w6e@vR(4lR^Z=e?W4S2$! ze*#GQ^=6wcO6?JJRFkJDB>j4`@{J*~2+Ftkz@SPrQK>I9G(jMZ&hHWtq+kI&*M3Yo!JUFQHNJgb~AQDsuY1zRtGPQZHBdEai~ zp=ftQo}$nRSkD?MXM*o_EuDul=(ys4_v`$=$s6o+=&6PdU3#Mwh5TScl-(QKwM2(* zO#K`Noe)67p<3|G(68$e)V;~v|4UJ5I8+Ok^^IMLpqf?NS>&Vu8Unku4co9M3qdk# zc#1+pV7IoMUke5xC_d-{gH8#c=S>w1UVZJ4Uw1r%d5S{Mn_5X`DEz`bRjZC(3_2}< zw!^|%bCcwzs6`t)@f3x&!@}83AJg$Tkezt~gDeEl6}Z>kleNX;KpPM76osz9y)J2> z9iBTn_#vM`mICN?uDro}`@$ipMGHUj6op>r!Uqz~PNbsWC}y<6Y7#8@_u#Gg`x-a{}l{mAt{~fz`SQvY)|I6gpA` zE8X3j<1X>B`^On%Er2>gf!Owi!`dL|!2_P6P)8^bYtynN-iD_@`3{553!qE!Fj#eC zX$6AvTk#ZyF2%#Fa#;Yb+e+RW&Y%kds0nnp%HFogLC}|dJVl`<(A`onIjn_FhT6Aj z47wn7=7+q&-n)r-#6``Xrzm9e13rVcb^I>`X_ky+kgWjf1n)-` zCVgvBi>e!?DAWnwkKSw9uSd|v*~SdA6F}8aw>9w3C@Tbwox@WUs)o9)k196%5v1j_ zn?aWakUm^TdxkB>({8 z@wp#S!jrw(n;$aBUI49xrFE_M=er{)a0^dSXeBJI^PN!r7_}%p{uP6+3ZO6W8?7|n zi>I$&CGZr5zQAv^bJIcmJTe+o#Gq>es2Em()CTovj#@NNho>l1468r_S|;EUgGbga zrZ!wh*9A}l9J&_{bMbh2;(4B;Py!sf0M#ik(4mVh9Kaw40hA7PTmEIgMxz#G{NyPL zr9<7;!$T9gBPe;sR0cT;prJ6om%ge9_v>6{@)U)J!u(!(*I9EA6zZ{oK{o_Y_h0e` zhy1SpLM^hs$5RyQ{tGsY+|lYKf_5~sV9-qgB>j5xWSy8f2wEo3QxuYZy}5nR+iM8Q z-QdcgTLNew)H100%*G>-a;7{*p?y%x5VL9ko;-1gjbhMk0VMs5!{D~7-k}zG$MF<} zq@Qu9**YG-EFVzIV$dA{)EUO>!YgdG5OhkNrzq4J#_Mz(=Hrq#&2x1OauPtB;DTM5 z+j|s(hFS9zg*L$j>t%Ja68%QYe|MkOaKSnYpqa3#XL)ru2Lzdy@Dzn+!ls_ZNoDvJ z)Oyia2Du2JTaTey&}J8YE8o1Brzmvmv87~*m*O_mB9HLJ4007fu28Bo^P_?)YLW9( zo}!Q|lE ze90$H;?7CnL7t*ePk7_Yx1Eejus?qYV9;FwWC+ir=0_9oY|E#QJVhZxcpj-7b96_C z?)>;v26+e|uONAY33txQphIV2z*7|R3W5o)d4usA-yQc72Hg`tCNMB~Xhr=41ns!X zQxr0Rfx(E8OK~ymJJ~kV8?GZy0Tc~)hD2q*NCd^n@f3xk;m$C|B}p9}x+@z7G3dSk zl76dgr0z02c)MdGPfcZQKk7JrF?2aQ80qcl1RqnitPg6iSA>_d2ue zHmF7YHMcOxO8`xSMNZ`x3-2IERg0%6Gz}Ix)ikTdW2&R9&oanc07*w6RU$6;K~TT* zJVhbt2;}c}eG5^Gn*Q=&kdFX*3k6~`jWzJd$>&BX3cZB_v5UoRJ0WP}oEQfA3ZR{E z=&o<-vj(-uXf97tXeS)HqJ_(TAn2&iHwHZvK%UUaAK5Jh-|Kez@)U(Up_9Kd;4XfL z9oJmesNp(#B!E5@!WS@1udYKa(rm#~6#7&M1B1R>dZ8AjZt24yKLI2yK6=vo;|Tj&}_g)*T<`MOnjub`Hf4l^iF0GUEjOq7@0d35M18>J{@3PmwD2O2j= zEy^l#WYA*)bRC{YZC5HzN6_nHo}$oocpm90x#9(#adVz9C`bVL!#IQfo1OSxXFr#x zDC7_047U?x-B63R_H@XiHEm6d z5wzsxPzHqxply}Vfvh@m5=@f3x&Rl>5Zu_h-F)I)PFgTe%mbcklhlgeZSl{HFH zNIFC_e~(T_1eqP%&Y*As6sxFU5dH81ejbfI&QlbMRkV_5mA-6;pjqh`81z&CmBSm~ z=-YNNs71Xpc#1;h@WvN;;tjs3D(ZPLC_(^9m#B`MzXRVHvKpl*BweD~>E|I_Xn9vP zkwK9H$nl(L~k zUR&hv7fna?kKR8NW;R?$&lFMp<6A2e{mfa-aJRN|Isec9S^s$0SKh!&;X^<4gC;gg z(ZA}jucf4|-Z(pS=9n6*v%l)Op!s^x!SE`Yiig5n7V;E@^q_-bY>|2zLH+@T40<7e zUO>r>eQ+h-&gxSO^3A%iWWc*phaK3Gx3<4T}Pgx&;w{uh=qSQ z1YJFFf%x1>@DznqU=(p!P)!7a zo-OV&tKsTS6hQ4^27gye=UJ#lkCyNhh1$amzTUpluLwF9Hi|(>0w@O7;msa&QGy`5 zaGs)246MW3Q0av`812;-GU%lM(t^5?uvT60Txw5so}!Qz)QxnxadH%D(X;b=81za2 zsly%EW<+fmYLVXso}!RC+<^zLJF*f%CkyQulq`UL!Y$~MLm{sEGycg_6#5CbpuLlC zEke+v8IKtBS^!C3ct0H5gzpS#GkJMT46cGblv>8N;Dl`u-(;Yn$4Frzm6$htAd43XdXo+1he;!*!G@ zfaGBGBUrglZ`7jtMkxx(!RUwe=+H#eBKen^40DAfM&2XO~t;B*s9|TZA1U&f0jd4W3k$)*qQK%roQqsy_Wjkt7$XquDeH1`` zpX3eN2Uz1O3#)lNMIpaWFr@sdGhPbZ`QbALeG))xVDuw-=w@7#pzw&ND6|GfKkmnE zy@^_Muk}|3eHK93Fva=ʄV=ytT>DGFu76leF%!JY^jv8|p#X#!|6+!>CI%H55i zZrgc^LX+XnP}w!BHG+y0d(3ILVABQAX;_rGe)^w-2+By}DGHs2MVWrBA0I`~RV_UR zWe6Y>c=b>&*6o3yL!*<-SJJp4!D5MG7@RSc7pMs!j^HK)o2q0bPZe^X$#+!P+J;YNK(uMAp zhfc8{g33R(ncHw3eG@<@;1*7QQn?tv$z}d;ydNgL}T0R{5brcWWI_Q78)r z_ZH;F zYq*Yn3ZNbi@&@MzyWpPcpGGMP^>DD1w4NTSh0eMv{k{w;5vggfOU70M)@hB`DV zkMb<%VGQ~sfTUkEtu~&kgAQG|3r|r<`bE>Gv)2`)7TuMd$Dm39r%3|)3LOcz9Cy%Env<|8) z9&NqZ8$o@iDlTZaj%o#v2CR}_ao}|{I&|Zv@f3wLV3oXWfF0iF{*t>ogX#oOHaw4V z*DWeS(8jwwMWJkX9@!nxGDR(NXflaG^#Uji3MOA)56nZ*DH)!kP#6?UDxFtVLC~Vr zYZ-K0>Ax2%8Ns2ex&F@;LAq;rib6(k=tkRW;n6M2m=g>-A%HGIPjycB<;4hE7t2!= zx(Gehmu*t#q86=IcVf^<0i+D?urs^oYanQx22W8)8Qx*f*}uGxp#GdK#Q2fLi z23ZK8qfkVYX+6jjL5~f2ib6-Bh{#Xz3BK3;aP4B;aKTy%AZdlO=?pDP1SPxi6osS} z&J}-8Dxnr-)Q@72l>q7vYaeD+!^;2vzoWy$JNqXp-D{%ze<=!ehqVtb6~TC-%4YFG z7C9q;O3PtlUH%SU&9Qt5Pf@6}97+Z5*2kh24Gr1DptAxUE373Xm1%$Q7Eqhw#}})jHe9br|U83x&V@GqF^95Ysz5+eQuPZkZcnx zNvVa|weS1W+KH4DJ~#*C6QpeV(FFAe;>5YGohLZzO4ckU@?DNCqB!q2CAL zWh0AQ@Dzn);K4U>Z6Y4({I}^EgKh|*v(Tb)lWhDbFW<~l6gmqnx@%)^fLb&>CXhil z1&|>W13{hDd5S^>aC*0X zzG@vrTl@9;;O!}FzqVP_a2?$dK$XyESXpMY z4YlZDHcwHg68a2hr-tLoOwU<53~~}c2k*!mTziulg`lL_JVl{{ci@|c$1`W4LuV5* zlR?e`DAFCKNAK7ULeRlbo}y5syQL(dOJY5Oc6B#nkc$A4?q#cAn@VD1@_E+5%G~XF+jCKvkXC=c|1j-d{|uBEon0@^g3@8!XR$}qyz&;r_@rvBIwQxo}!Qv z3>=xJ=HUMGxBDL%2(jV{?g>vD!zj%C0ygSF9=9LWc6+qH1)7#$< zzawZ-3!b8ow99leYvDHZ8|j*LSkiF8J`_Mnuy4=$!8y3_xbs$?qEHg-+p|63N)rU7 zza7S)M*`>(Y?HXfH(3s~=;b?}qR=DQCh^>kOBo2-t~-xGegY^37PB0^-)c01R*vQ= z3Z=ke7R$gvxKv>N^&JfI7eJ~o`capyf#0kx9C(UCsxbPYto;Lz#(%50$e;iLlm@q; z-)kJ7q82^*!&4MWgImxogIBuf(2ZH-&7eR5wCat#K~TZk76?*b%u^Iv^~O>%w~d|x zg62ggGU%}Ysttgzpet;!LXc4uPf@5g0N!z&;_#*!hH5_;6eNI>;Lh-O=!S3v4OHhT z3MIjvVYK&4++|vROmS($brdXse!^J6hH(p%5j5a9Pf_S6j1{=al-xpxZgaXigPsVW zKzQ(#cIq62Aj1rvqEH|__$FS-z)*j^Nel`RK+^I;gYZ7MMgJP5C?qW}{PJoG-pj(m zZY_gC1<-6c1%ntvV|><`T;?eX&6cy0thhCzKRR?KzfUkIOaL8)>nLwbt3K$^tt#Ou z3LS;(XtjOY=?JQx;>4hE5mYa4pyM$lDo_0fR1cF)^5Jp_>g4g1K~DuxQ+R9J zsuFpJLkn1Z0=gL#TVx8SVnmfJ2HL0>K`WYBW~v=3@c zZ&bejgP`{pd5S{&pw{&GEgw9^xwl{sgI)+A-6Gf$_Fc_01eyNeDGKQpSxN@jW?e-s zDxY?lLD2$eFg#(8&CTC~AXy`xqR?P?!Un8-j7NGG_&j1zi~!1i22(R3y50!V^yMiE zVXWd(EI&0VMrG!r75dxXVPM03mT;;C|&?9t%Wbh4S9AM9Xjh5JVl|UweUPr`-bP~ z%=&0CC_w-j!>XvK+n?$oXmMYjqL4AHiVB`*>W^A9^4v5AB?_S1u(YS|g7T9H(zWI( z3f+dKJ$|Dm;sXAxqKynn52Onh)UNP4g2sBeG3b>5lGfG#9WC3IXJyrE1`pprP(ZLB|gPH)uMG@8T48J?T10~j|)fcLeSIoJVl}X zFi4&qW`_r0wW8`7^hN*`!8CM@m$n2!&7ScTg^FMr`rwZcOLXXhRC}&yxL{KR&}kUl zd+)er7=o_#<0%TAhQYmOos{r!W%+462Biw1r%+KA$j(w&6C9LHgo}$oE7)HC;JQB}5eyR>+&<6o@31)|n zc4~ze&pxc-DGFVJ+2Jc&QlbMjf71z{zdIUE!x?)?aGGh=(7N74x?L@ZbrDa z=e#OUQK&hLZq;S2U5lXQr*#;VCV*B#$xYAjXSke0+k&Skv=T~g5~@x=Ku}KhOa`S3 zAP;z(Shg&AKl+Ux=I|7SJfv?EPUG+b=RtaA49XBd*_E)maAK?B2x>8orzn(N345Vm z({e;D^1ObIL74(b9VVNaOsiKw&@l&|qL4aFHkD?W&qI)Q)jbA%5kM+%=&D|(;yIwM ze|d^RDsbp*H+XhL(8Hy%49XHf*I{(){*A+@P>Vd5@f3xw!|2vKXRl=lDhv6}psxZ* z6Q<81Zid<;C_I#>D5MF~XNMf@@O<~0j&iFSuA^)LUz_)3(Ep?BzT$=W(pL2hbKzaA6Ef4J%z9I81z{JeWqf(k?pip;L<&Q%~K-unTql9^-ki-vlJVZ(T)^ z2ZO#ypdh+*`C3cj0D4f&Qz8^ZmoC2Js>=Xfni|fauM(&ZEv)Zc;9~+%%ru@7p*pm% zUVqd(?5S?P{fa@~B+wgL<7*YWE)k#|kvt_rZ)lD0uy(6lXwk?{B@Fs5fkx4sYRsEl z9K&jL<|z>xMRTexaz0_X`_?^87ggS{*%By%24Z%DP1`_=CfV|o2u099%+t38o=9-# zb$D;VBUsO|!&$Mvnae8k=*NLAerW6?KXA zEGw|0Y;!J8iO?$Q5_|7nvIrofTJS9TO zwDh(!DET=+Dyvi%SKdbj5~z@7x4wB@{-5ztm6Qk-((Klbyb-uIG3QYi1{F%6zceaO z>)8+Y5NSN-DG~Zhqw?l2gA1TV9l9Gbs7MCUrF&wq1FL7%wR!44Ai8wX_dRgF?t`;A zgNh}P67{_`jvI7^&*;Muo)RG?>U*0^`Y{$-R4;E2gGwY&K7GBrjir8F&tAU_0Xly*ib206kUFg&IczGx zXWh*%6PlYE~(Vv}vK#RUrNr_NT`pJxf&0}#iv443P zgUTe(Uz*SeyyhAPQ1}X-5~05|p>cG@JDm5v8r#ym@`n8@ftFC=MA6>dMF4r;<0%nZ zLWL7&ojZ4e&nULTU-pn3Ee*{t}I4v?z`Pl?bx`iwFk#@zy_maQ3sDkM-AP0EWa>#uY&c}j$8(#(ln^QHLou4(MT zpxt%;d$Uq^n#os99lIP_lvyPuLfvU5|G0m@^Kj|Be1jRZM*_JQs_32B>f!*Y1$CM-wJ|{Ns-#3HlX^i7yPmBJEgCu_he3NKP$Eq_ z&D!nnH?Q%O2o1@jUXX8p zgz}ncGibjAT1+E8LxW@Z_10>^QzEpOMtqLH29&|2YqG(ZL3R?TCB0!)FPE-_7JaUg z5}}s#hP5(0*aD!GBufU_OP~%kU#Auuj$ejz$vh=O9caGpk)drvsAC-7!;V1?b)b$h z*)n&yYo9H69;5$2{>T4W$5@xvQbR9(x&zHWIEbgj|Eeynr9RL){|B1?%)^iUSB{eA zN5-n?-R)B=9H4KWJS9Srv33RrUwtkGC{rnxK~55=1+_@;-My;-mDb=X5o$p#`f)fB z&ykyCnZcj~5-9i{{iyO2&shLXT*gx(6nu~B7+coBQQN_X{}^;o0?ngBui^K`VxgDw zBc2kWc~t0izi(PV9&5~1C+T{PoJ|3vtVdLGhc z&|wJ_OoyR4RXCji=wFqT2nEw&XodGawFl_t_t^|`mOvWxVASx?#~tg(vUy5`H0Z%N zf6hz`fIg4g!k{A(NU`T}aY3Rhv?yUbPl=FX&!fBAkTC#7pK)W*Q3;fyR!eWz>Q8+D zI_||&B9x+LZ(ys}xFbNd%Wp8~m;~xa#Sc4{_6r55qDo4H`YDPZlpL|@FL!z>gIpw# z=5IRrAz>Uo8IsI+N`y3j(_BHFYrfDTukd^Zxk{ix)WPu1wHgL3vWeg+5gJ4tj5B|t zX8;t|QhjOVeRNy`eWX43rE50h*ZWl~o)V#tvYIJLFcE7_1Jg|y{ zhUej;ZHI~r3_2x&qUq^fq5l@24DJ8%ln6!B)4TuteyY%-JLd5W@{m9wk%|*kn&ak5 zmnA$ULLre-uITLL`a#|Qd8?*#NE(?hA&`{w*)e!8+Nkc zi@(sK1|xV%giPs%Z8Kp8uBA>mwSz%E5@X;Bf;m2t;{0cv=kr$k7Liiqa^O2#_I)Gj3q@|Qr{Xqv@6 zX3HmlVzhWlgtpN%OQ-jt?$9Dtho;LaZ`gAZNHHmqztnduK+29hB|?fxi6c7BQ2@>O z*q=e?CC~wyjW_Xm9S@MhC!P|a12h}27Fd9@AMFfG8FWDcwV}D|i;jJ8VZH4ro)V!p zG@9g90Q_4!w_jHrY5ri!!UEL@0;eM?()9^arRy$(cbHCD3%*P^R-? z?-qdm)Zi%*nob+aJnk=y21v^ykU^Iukd9(mg>{+zor27i{~j3QXET_aHSqjk}Gxi&Y(aEbc803(!T_L0O)0vln5Q6iK90! zmmP-|C2Ut+UU?q{NuVNH;{5*exIaMGcJP!4710vs7*E%o0KH1@!k{Y>=nq}GrUp+R z1El_*r$p!vUAh##0Q_l^F1m&cx+;PC(%DKyr%qP`sLpVn5~03ywo<$2v*tjH!cUts zC|Cl8QF&qcu;>8*ojAi&A{0jDh1vOU@yk%Bx-El3B#;-4VGn6Oz*o7M5>JVc7mZ=3 zq&R6qi;CuYGw7NG%Arb_C;Ps(h8EqQ$5SGdLzOUPX12otI(9pXLDwZv1wHHX51p(5 z(7Z^V5}^uu)|FcvQ3vQi`*a4~kU)yz|J!q3cL7?^fu}@BG5p`@T^Cm#Lv70#6e@wH z)6&elp2xibirmXnA~c>Nu^;+ zZErkPVBe_047w$O6u-%}=XrZIfcB5(DG^fqCYMK#7IE+yWnDC5P?!X=38mjp`B4wI z_WZiUQzB#&YG;sW*F6uQz}hwp3YS2wXt10W)e}3A=j!m32(_ZYa_1rbmH>6KbYW10 z1Ug4^K{Gz4eS=F^e;H4S&^ek5YOuQvj`(gq2xic236w#Pp!7=*v8U?ukf%f_gC0Q# z{$9XuT^;RY21QCB#Tm{%_F6c1T~Z|_LW(n-pYJooD)e0kau^gPfyPpw!DZTTETuL* z$WtOTmii3OS6d{*XQcO|-pb1RC|Uw7qJ{O<%}3#GcZ;7qB|?j6VLf3`=`4WSjMrw+ z9SLMj&$`*>1C!y)P+TP?Le})Gv);5-AE3>bj2RRoft=~mCz3F zVbmO;)Y{7!bXNiy(HnO1xc@m@>1-XI5+Ngc!}j_0BM_inOY9gFD}e@5-+OvNmq2LI zP79tAp@G!*{ye|@3P9cO`!VRA1o}n~)h#1hW4U{+IGz%rZ}d>@>>Gj8Xlh!q47x9Y zn$fW3qZbn*&h$nmi>!3uxwKx7vnZ(4u$Cw=n3j1iBhcXDg*NssQN83Z4?7tI>7_ zANG0S4p7$zZVY-Nf%K_H&U*5mJb+d|Z>a6qhtvbM7wkwHe8PdDkt=KN`!)F zmu}sP;W$g2>(QG*PbH96Ha%3khv5`WiYHHrkXAOGJ*#&O4@J3GYQmrt3Dl5=Ew?nr zE`!gg_8*=Sp@uYU8Q9JM_j&ouvtrOQ3ABa|3Q$`%a4tZr=JS*Yt)YVgUOl(>g3oBk zZ6^k$N}#q>vwcGOVhTVlB6&)L+EUGS>aSIJ!`5vXz@X<6=p@}oj;G3TAeL7pB|<0Z zKB~R&({^Z4C!2T%y^uijXga#Sdl&pNlvYWJ&^(%sK2-B9ZokremdT)(66gVSPHgWQ z;S^1Am6QlQpw5YRg~iQmN@KO=uLp;6!Me^ z70@j4f{gbg0ct#X0fXL1AanZG#kSkl0H9A*QX*td-@5vKMb`i_zPN)yZzWLA0u{Xh z8;foOH2D%wiBQi1n%$cD71t)})%Ijix&%@TEo;U$z=!IKDk%|C3@u&lx8ke(=+bZo zy^}zzX%VDi!Q6(>qK%e3B|@ud5k%Fj%n>eKkB6@q^j-pWr_M=dmE9Qt&3wdDBGjEa zC)JR`m07=v~a4kph`-FZc_bKJCF5P3ewoF>6*$LHbVks(CYtK zrF+KEB7J+F5}^!Q{Wm<&+YByU(U<-V`Y3_?r~`R;_rB`@C4A*65%QxB=8T45KMbi7IYtTY`@eT0hDG`dK z_tD6CK{#ybSKXOGnG&c2ohLKr+!j0ANhCSDG@sV)6QV*`Q|+VdTgs|U3njUlR%wmX!&5|I{b_R z_wtkob*7=^(OPQd0R7A8!l3UGD2--L!hUDtYa;(6Pl-?(&74g8G29jE7*+KR8I+B6 zjDH%|8*xD;AMb_78>FGZJuJe?-wha0yfu)o;OQ4GqJKm|1Pa4ac0 z1<=O_JS9Q}H1zoMbP%@agl0N}@+44VE*;{OYyBCZV;y-)gc5V@3|8yj!kxCYZOa(+ zO9DNo+4$Ev^K}7gyO*a#=sC^C&nwDb44;u{X3MpecXz%7ilg3U$l(ST0n*FjDG`dJ z-sb9FrS}2qW;B>V1rq3XvWlL4Ya6_eW{l-25xSjBXJyY&Sq6}&pBaM+B~S^Ka~Lgt zizh$i`SX+rl~6fHXZu=sEYlneo)RI&M)Cu}tCRt{9v{r05(#vVCSc7~Hz&ZC;lKl)5}|uE z0b6`L)*2w=?#T=)l|VCSeIe&;Uju;pYV(u`&7k##-}`;A|2)7che5w3&i(o>niV~KN6@etpO#QtoaGh+^;+(LUm~kXrbAN>j3#0 zXfvox0`1RH(TmJ$e*mC0qj*Y$_UF)PQQ^bK0Hon%%%Hy#D3pfqTcd360905dB|@Pz zgkPP$Xd6JcN-Y^wE`eUqsI92Sdi;7H`pr`!^nylh>E7WuLp5`v9fK+)P(`g;dV~5r z84WEmoWxTiR8h;`AbQRR>;!5A`!VRB1lmj^f>#@_<6)PD@Jl|VIU=%L}%v@<|v&v{COYS7SQ+vZ2zp+%ZQ+O4m=VfRX) zNp$J%*skmXP~)LIB|?+v(oN6Rz}?X6J#`thPXbk=ex2Q+EWEnLCRH{HS@I|-yi<50)-9vuOy zGmWQ2NQcIuE3>ybLyJ}fyD`XK0qXc?Ly`cS3%W!1fXdO?9&_n74o!x%H z2%u{z`3!QBK<()fG~>Z)cWBY=XFMfB?dcKJ`q40)jUP8aeM9AabU*^Vp^2j*^U|6D z)PEpPiO?IGIC8bFwh~$t=hB-&2PM!8ItFBP3ok!_l3aO8gkI1wAXnYId;m)NVZxw8 z5=fP1dOLnl)tzprzw^N`w@3Bb{46 z+zL?46(u0_3wHo#bL`)AD$ASaN3L4*GOX!KznQGGw8SkIzSUguP#Mu!e`{7 z%u^zCfF_O{O$OuB+ke3V2Az;VM)YME`>+vqw>B^2DG@TFFTY{wBzQ3Qi>s5x@O&ZN`$;=$4!I6 zD_GRp;aE6>+$E3>-LS1LcwqfiM;D$FAsxD5TOaceh8ETT^NK;IB#b~LmMs}_;~EmDnWy0P+x z^^`#EX${CQq6t=%WmQRuPS0=d#Db8lXpoJS9SF>C*K-RqcQ3 zMkaq}&{+v&N}ZGJIzRD-b)UjhB4kRPlfw^dK7$tNUQpdsc^~;npbHhW3H?S}9cWSQ z0G<+|3l+42zouPlfDDzj800U3X4ClS{;V-e0n(|-QzA5*#z%gs_cZ}ZSZ2tea}wwq z?JhjLemg!H9xUf65&A~E3kMyEI0Mk6By$Fxmq16U(95ZA=w@iq+GL&*p`%pjHGa}F ze}E46vSrW(2^34yXp2t7dA0xdPgHZZ9mlx06jUx zQzG<^Mto(PU%rJF+5L=S&_xNfntDMO+n#y^(8(O05~0=93kvy~iNlszCg}{iB!LuX zX6onH#m~riB2S5s;>^s(Pb{WFi*!TE7<5?z)uV4+*r(38KgIbPPl-@H`quS)U7i9^ zd6QO~D{t693AB?&d>2<`)r3o@+>ED0XeW*M{7&{X258Op!3+wLKnriuVbOi_aR&0} z4xSRBg*WN=Ye#+D`|v8=j6qiO=$ZstNp+0>4EL{vOV?)> zPl?bZ|@~O@abQy2kzgRr$p!&t?@Z|ZNtgi?2XzC3Y9?NR78}RJRgrfxV4F=L@1n!i26Jo z@DVQEp_j%Cx+#ISQs-pSUF)`R>8xJyln8C5&WYc4ze@lOAGnM`wU;Y%Hr)u& zyg@uALe;46-BQ;IJCM0Y>=+a#fm%_&&M@}vE`S~#Sv}5@;ALf>^vbWd|)fX3kS0G>jHOo;O(D09rKrjxK|u zB~SuQM_({@3jkGMt#R)$K!KWb7<5MhDYhi^v}uY{ohLf-ln5!dB-~c3 z;RjIC?kx<8kwDvMQU2iGI_;oEIeU0YgtpP5eA2|;xP<*I!;L|ACD1{7m2bYZ9A_Xe zedH+-I!Lc_$B)|8phXcQLKze*fqdu<+n`z)?sh*jlBYz-hu*OD>iS{1&X7~747w+Q zOn<8A^*d-10hg|^2TzHR=}+1zuyQ~Vv`8mEpF#H}P=j(h(7Q(?T&U_&z*8dBpxn;j z;u>2#E-8D4`nJj&HckSyrD2O}N$Z}_qIWZSN`%_du%(t`L1$=D$FSZEikCoERQ0zz zxD-cx^}~5egsiCQZ`R_^K>!VEI*~ySBv5zyp_Wa34i!O*RGaaX2z93)YH1(li@l%& zHdYLJD1nAi!Q@w~hvxvgxsj(tXc!et)^*a&h8BH&@5G=-5-2@@`gLl)^8rfvz*8cW z9zYAe!cGUHgc}j##X%_L-0~1`QYtT=hLCF%RKfSyL&ToWksb8z4 zM5sT#yq@IWz`24%*98oEDuK)iGMwg74K7{4ah?((bAn8iXRd`y*Zk)W2BkE-p#Q0*vuM#0^681za4c~fuG&eG{Hv?yAer$oq`dYcxyr`iB? z)6tYcuO*P8j`3;RS=eQ|?8H+dq^M(TGdl^t-b&xC8I&f0=FxQJ$dVf6(4yYiJS9T& zXu7iCS)Gm0qB&!n8T3X1HKLWrUX5Pj-1Q72o)V!(wDLH!*`av=HT4Z-&|3+VNF9u| zS@+|iMHy96B9urSjOhIh!vH!}{RxB8C6FC`z3W@FRtIR05>JVc9euqMUMAxjP?*Jc z2ECI&YpL+KhQSZqo)f*4r$lHi6&}}ptC9yTn*C69XXSnLUIJ~R6?`SF1$a8g(nmZc zLYrs>|7Tv_4}cOiv>5b30yU(k_rX&8UvTNXGoD#)RUIazQp;w1gPA| zmO-B+&_n9hYUKr(0hBS8r$p!>b!)vU?&7m<=VfmOeU?C)G)dm}oYEwKECYE;gfwZA zylzoYI<#nby=VqyN}w||?eFoT_HuxFsqvHuouO%e=lq=$5L%VapezX#OA8rF;Z0Ql zYPp)HL@1UPGUny{+XT?-$7KxqB7qb?w$gCbUn~JFdcsp8r1-Iw$0alH>;1HAt6h~h z>{ki2kESap_RPTwXa8H5=`OZ@!)SM=c4*wnY2tK2;#x@MfmO#EVnCjwk z4!8DrjpHd1@}0hyA6vSIhS@Pl?bSs$*Q;rIi6d$+bfm z^iu*&qXoh3E`zbCHMtH?iO@7!5Hwjc6bp}&mL@YOM*`KO8LBNs!Oh__dS=N}B2H_qrt=jI&`zTKWg;HI8 z+0(7~WjNc8r$i`}>gorZ#bHI+hke=%`Xzxng{kPZ&=|M@E?we&o)V!>VYCM19D{vt zjVxmZ=oO zPFUf3_syC-B|^#63!42`7rR^Cm&7utSOVEni~R3i!_h0jl?rr$mTOCpR$q;(>dJ8jhI5pfU;cgQmi!yvxG5p!Zc$BJ_i%!ZdS-;C(dQ zV+(`+N+4ZYK8v>P^95Se)RU(~NSBt+-aNVY04`mvGB*a5OP~t+dQXXH*ASpjRZ=2U zL0|7>JM|6#bZb#4gDND@h1;~M^{*q&^j=%cQzCTXHmwd@-p9}AcWf$y{z)KJ8cp;K zZIJ{mO1#HYBBV;AiJy&C@H1MXS-_y(_5ORak}Li0!@S%zssI^vY-t?myBnZEpL;XNRssdnGW5!={ks5U zmdR5h6imy|u0GdS0@R1tVUOqjDC593)UOjgO|-pT^F~w3<96Ldi5fI#yT@->@&1DDSQO);UTb ze|pyG7PgFpOLy0Tr$oq~o^?MyXkbOz)+d?_a*{x!sW{xRTRc|%871(P2#u!V@UY{v zw!*hAxr;u74oILSg(`Xr12)-0i+*bHln5;;v@=i%(i#nrm*WBk9h5)~X?Ckw_aE;7 zTI0l1BGiy(w~}^xY6CR-%MJz|l0cp`c^f*+7|S`jedQ?;@}$Yz5~E;SfF>GxGU%`b z(xnOZsA-$=%g}5LPl=E&O|a{pXxa%Nr!(OUa+W~WbOhgtLzkPtrL*$lDG{=!Blr$! zF2%3+vOljFbVLI6p}LXg%gzM>bgqo2M5qtdja=XNxi++D#-b7i9hE?XY3WU?&*kui>E|rB~3ad z&3lX|Os-A0X3%j7WJoR2&<_rW77cpGQzB$YExNy3XB4zZP1l)0CnQikx{t!=EyQWG zvMMPNsz>+Hqh^<}TFKolkU?$|s1Ci4-u(W7<0GGwJS9SP=zWy3!T@_g>q;Lp=%fVN zOOsAhO)Z+jrCalxr$lHkO*%PG+l$+;CQSa$Aa@Dm@SA>{WA$yEMr$#Jr$or%HVp_`yZM*ckYA5eXJgMi(Q@mJ(JlyzZ{hXi_9y_R0Y zgyh`-aZ2=$;-BR(G8jV&5` z)0{!4B~To#{tqbB!1qzhTRbH~akTn>zRlP!(4vnmZ5eb%0{PGkq~FAqPoYI`Tk(_# z`OpkxsOxn@fC6`VGssHOS<$3YkQWT zE@P0d1lmCZvHsdVg#e8`!&4%(g9c*167r4V(#@}EX;*o}o|Qn|Xs!HydaHu~_4~(D zBGir6%C#S?uDYdj=&^ZaTpVp7gmL6{jP+FCg2<@l!qq)u!`gCMe%|56~C890pyKKta^W z|5>X|SAfFpc}j$WsFQ!hAUYeM$DixjSGMSq1R6>6b%`~@umCAKlcz*zB+b|L@-ROD zP}yj023?jwOXxMxcmAD`(4sU$o)V!Y^qMHXzUL}HH_jO|C{O~G(~`lk6M0w{UoD;zp-EJ8l33#|z9!aK+A-*g1iD3+ z&N{?w8bCXj@stSNqDyD;w<$io-6H)MbX5XrE7#Inykb){H5mIa`FfC8lgL`4RpDu%L zNFeL`DtbRcE#E?mbp3fsgskt=4-mZYYYx!v8gm#FDuJ5Q@GxR@%?1F?SLP`ZYEHvL zzp+N;02MCW!l0WHXb!C}Wc{6iC*wU>#8V7<5Yl=~K1;AGhha zkg+(Dr$k7fs{J*34#16j4?Bf2C`~>;_IQoYEAP%cB|-=3cjd6Une2nm!&@BRI$k+|MH>@81;P#(Sa z4b@7z!Iz=>V0FjJ`{=d=@}e7d?&4=yNuFLMB|=_w!!BELwE`f0x84kjlt7ANyf+gz zV&BZ*Bu|Nuq8RU6%~zHHnO2|3pePA6DTp>46?>M$XJo0wQzA4eh`!zj6GH$RyV#0B z(Gn<|UdQ`xr7QrbgE>!$P&U1e+kH*{pSR>~CkEY-K;HDNyS;HmIzW>mc}j%5>07tj zWN{9(s7~7e2E|CAmo)p)uWj*7fHJD2MCc{WeoS1s&IX{TyW$yiR|2KbV9GP64h}th zck`48rO;qXS^3E}fQnyaGALF8l~${z*JpV%+@ACFB~OV^X*GL;A6jPkt?S!I*{SkA zx+j6U{-ML|b@pV? zLkTp9CMEJef5b~yS|ueygJ@FXX`0(?fL`qlXV4=FG?Nx>?bLtbfeeNFcuIt3(xUB> zK@DC3)Zxo320fNQljyXl&(R-ncv$NzPl?bZIxVV))fqK_O2?Nl=!pbcPqjV4J<74l z;^_pQ5~1}}+mqBLK@*_;mzo}^ykQe0&Agwc4p9138e9lb_#T~ zR)H3MtdbHTjem9qHe;G#2QuCgXd=Xkr$lH1b@GR=?Tr1psacO1 z^h^RBrGm+A6M9{Q77hNwQzCSf3MNOM-HYG4nWMflC{+SIqfY(_odM4QnlhTFMCciH z@-3IwzYL#|o44x0%KPZK1e!@5$mtWRW0%<7ho?knCUqbi-8@?pARQGg2EC9#b7-e+ zosN^SOI*DcPl?bR+G%@GPvajzBNiDl=%oZQqVkG5$sssSE zD#n~auOv_^O|xuD*LVfcyt_OlLa8*(vSj%R+z7L&gDrzzOQ6+M96o8B`*nbfHF!#d zR#S0!gYOm1phY{jc{3&GkWumr$k7Tw#N5<)XNT_#F6O?dMkmxQ$?BAJXc&_xT?ofBJ`aq%AVLhz#qi1 zaW7*~x&*4APJZ1cH$T8;8}^+9$}Lg+n)y|HA8je* zDG|you`}3R@nt<+x}ceZ8T4KP&85YrYm-0sgBBf`#Zw|Qmlm65pEqy>XkeHbgFZ;0 zd$eNGB41-WKuX~}B|`UT#pL?sJ&ggn)XauK84^fwu*v;3a|{5oYtB<5q&V25NzII> z0Nq*Z!k~{5NU_9u+s#}LpkwQJN`w?koOkPV!u5q=Pl6frNdno?uOwt=IyM4GJ%Oi0 z$c}y`Va>dKiO?ed?#T@LEP*;xm-s}>3Ajb>v^GzPP-p5AFDrFE0#Jj4ISk5_K#Jd$ zu6H>FAF3vYcuIs6zbjo4buDp-3B+#-r8pE!7Qx7iP$0{iiS{7$#;D1Yf zDzxa3i8h12NFXhG!}`w(i~?x)M4l2MEqcR_dh>fIKoNn)4Eid8{Ap{?)tJWkhV=;I zDG~Cgtvz1l9dWCvTK#1V`X+%C2R}~vcKtiF=v$SP2q_MJ>}vM19kl4caytfnmq7V6 z1Gz8mG8U*@T)|T!lut8|Im5?eWoErZKL%w>pq|u$RQ|IZOKy53@stSlqz>fb7JqTy zagYA74EiB~jA;13W%?#uORcTLQzB$U!~dNR>~Y?^rfUX+eoCNKv^3-3qf`K&(bFm^ z5n4q{GyV5Y#$M3=ynhVJkwA(i&b-D};{jUqi>E|LvBa4??j|1Q@M2s$=gJ#4R|476 zvrcL4C@jWPo4`{dWJ}MwKfk;-!?$k2MO_BvNgyxkZq3rDhMki(mv~Bqyr{d?MeRP; zn!41S!=PUh$e6lYyH+j7N05UGPl=E*b+?)|UxkC^)(f^UC|?5oq6+7BTTah~&**2B zlnDKz3g;)q)ve&tX~noPs6YY@RUGxdswK{;cDc(_A~cka`rqC+eJ4OUokAH@D1lzm z)7zuaGXPrjpfgX2&`WxHt8ZCf0Fd6UR0b7Epdk9zjZ_U?1yKFnJS9Rw^sSp!rz3W^ zj=jleP_YDhO7~IIK5y|2YyOs}MCd8qM-G9HT%bh_hpHc`ypKvGkR=sAygXfj@1yrs zQX*tY#Sg`6x_t!5#iKWaN+pn@oWm%k7DA^yc}j#7No8Hz_LAFaHP{z;&XR7%|=^JIN!QKlYGiO@zWrEWDz_dZ;@ zS$>)f+O78Ao0SwTYMWGD3!nr3JS9Sk7KPpX`39i%D*6oCBY`4mu>2+bA3lPNYw?r_ zMbcoouX-0;5zAk&fI+qrXawzPjcw)d4O&#Vkf%gw1np^cw_e%{TGS|d2ZQ!XAX6%( zjy|hW0#Ig^ln9wpDfO=2TktdL*2a@T`y^0xx?z=C{LhnNbX%Shq3U$QcF`@uz2R4O zhBIit1d66okiF5@LZC$syLd{3qNx<*^e}h)!6T*juNY(}ffS>O&H6*J%pjpkN`w@n z2@{`V#_$<^99qI4dkOTI_CDOR*@2arp~HAegdWr02iKXO@O@`oDG?e;Z`c&CVsE%~ z9+OQO(>#OeT8lvdSxn(65n4?13_pEdsRN`LXw9Gl66iN|PHKPar2|m?Af6JT z-_$wj9@q?rEyvWH8FWwrWl^7Dph?dW04=D`QzDc_eFlfD{@0*I?kfTrbVveir%`#- zfW!Du^W=@1E6bi* z-x=gAfx_tNt(xA~8ldQ|JS9S5^z?qv__Q^&=!T=JOXYoZL<0Sydf;9!ZE68@#fhgx z=oi%kw@nxs2#{k|7X}@bK*Q-Xdhl^R)|~A6!c!tNoIazIw=H4;iZnK4&@l<5_<5ja z_U_8iq9fyYN`w?Y4>WV+-(3LpyI{^B7YU?EJ8sqrPDcld`O3VvW0dR-uGQ#-9rdkCTo`m( z0?ne|@m#7j)DfWL7Ca?Fv*>p`t_5kkMUcBp9DG*PVfC$ySBom8&j93MCeF3{Yb>**7!c^w$hG4{u1aqeHm5^3+w^V z5G$S%q3iTz&>TM%pA1O}ehfM%ffS>*sY|19X~sK|r$k6GYOCYFIqcZwh|BJ9gtK5pt)e_v;qRu-5dQeFlRrNTA_V{GhIS0gLfs9C%8EhEwrF`wGLQ z(4xBE|1l^)0%_Ca-Yyqo8~BXWvUy5`v}tnhVz-l~0O?I`ccSu!y(odAs9;ia(rH|p z(VW6lA{0dhlP#|weFjj=8@dd-B!M>4e9rN=KLVjey+e6Qgf`NAj=-L7*8#lv>Trc#Z^)wv@zb!VDv(dB7nyB4rS042~{pw@K6MAj_j+W@t6AtG*ln5QBr}x*%F>Rqm!E3A-bVCAdqB7lGfA0JM=&v5k4q(tt3ABLfM*7ZiJ`2$No;)Q&3#e`+ zakTAsfS!89Gw7BC8go=duZh{1bpYM+4kF zpb39?N`w@r-1uDhhP6F`vy@L(-bdjQXchJAHmtiH1JH@tJS9S_s9)!P_Yt0;vL!;3 zK@k$@G|lJyTeqkKK;v)oln9-s`5c$Mi?B<4p@}|&ZcCskvjeyolt3PNv}ijsrxc)HRZ=44k!NSn!Ml?QeCw*G?qE=q1k$Bt=pM^g z;ZA}4Dk%}trDf=(85 z&-e&3?CH#)cnQ=cUqvsY*<5^joAlx-5$cjprE5)QX2PZGcqouT4uc6yTt zxl7>EX;+J8&{GL?ooY=RPg26Cck}8zB|_J!*7W$c2&_pkTAa?H6bba14vWsNtAh*c zbIo~5gg(<@(F1$g;`qoUzKlW7Bv2x4DBIrlTVME$&OYEN5lW;DWe!K5CcvfZrrpY; z@`g>7K-x4aH`}dV3{aOIJS9TfG%BCr?uF+uwsjoLpyv|kGQB4Lwai-ukhv32iO^+w zO<2bD!!uzve>Y>$3khUNUE-^KnuP#ll+9BjWJz6O+qccIU-#G0hCweSkQVh+!`qKi z1L*k}o)RG~>Zum48;F&eR;OJU^hyGimC??%7Q-$8qs9t4K(A&=E1kl%p{pXX%a|}MiaLO&Bh|4t`m7mg!E`M zVNeNCl1zOZ9fu}@h0L`hMdAayBwCG^}Wej>RfnMC8iZUCcW&j=4;VBV% zal_6aZ%+Ysx9%LVW6%c))Q4&k?E0GHy!Yv&JS9SXs3sw?%ht`%BJ&(S24zT~TQqZW zCeR6sT8HNHlnC9TnUkQ*09}CQPl#pEM+vlrpq6Epi{aAgoA8tfZ6T;`^O^WQdKr?z zpidHLWh8wWoP!VT?S=Ipi4BG=wvcd z30hS54Nr;CB^pg!pZ-@HT9iLz4uif(pz~A;5)}H=382NpcuIuMQz^(_gAceUf8x{@ z27Q%4->3tbJZoZ8fDU@_ln8yJ4&<{V>+$RT@1GlkzDb~6wC}jScTy@q->aSc|C9*r zqJ76XJxqh((giLJWs&a^$cvtJJ@!Ud11Qsyr$oq$o^{ugp8Wyn@`F?cWlJC}T9l8x z-q|0ZT@QImgtTZ;erQrBJe+eymjVX;kU&w?-MVA6xH>>aT0A8}QPkb)S9IkRw5VjK z`kBfb_NN4zL!A?cW|y%PBy|^0iO?MCoV3v`4hP8hO>YL}NTB;vbMoHH1rO<3^_HhZ z=swk)9C|jQFF;8HCo(8k0!^c}@@GA(Ux!QQJ&317Xd11RduX?E0m#F}ia~i2s6U-Z zP-kR4ECpHR%2OiLpH3v0p8vW9KyCA#81zd5ouk*p>ziZnFo&82JS9Ts=r!RGqcRqt zce4T*lrMn}U#AUa1FvCaX2NWq5~0J_X&rXo8C>wa92(D{0tvL6UggWaCE)76!%dzN zq22T<_cJlVXWjfJSqv(aKsI#g2D_->>VNm9JS9Rlbm^vFUXNu4-s_aTD(|Br36w3K z9Kx#~T)_UTQc3eq(xUC8i>Ww^_@YWmgig|;ZF1t;Lb%R8**h5YTLRU8OM8dknb(Bt zyz>W7iBSEwbZAEB?+*dWGx2259|<&oHtwnDPjLe%aUxHN&;Z)FXMZ?%9YCuB!x>a2 zfzDG;X6_&lEaw;>#8V=4o_aFp$~8R!N~`mlL4PIC2YS?Z`K*hRdnI*wN`yYpqyAp- zrF{UoEiYkExdfU*MXh}k^0Cm%ZUs+?&=e|aEoteEs3CPQdU_o1573kTJS9R6se^GX&7mVe3C^Yr z+Fk#@cQ#qkt{J1B^>9+c{|HZskQMEkX*kiIC!_ zgxB?T-2pAK>7~UWdkHj_p5Cgq9kl^k*PEwAXe>Rwhdn#|5};%I4H@Jhfo{=nH`E=w z1Q*tC+VPYK-J;)aXfw>aD?o)=<_vO_K<(*UH+B8JG-y%67oHNK_Vlf*yHW>xn_338 z404h{>NGD}eR**gfF_ONDG^eqdC@8Do1KLg^*ig$paT-<5|tNb9=&h`po%If5xPX> zg;pO-e*-kMS~Pxu@Zv@(>_eU- z*cs@ZUWvoQ;zTnB9hE>%bOhhfCXKV;Gb&BuDG_p_Blr$h>*WS5@*Zl#pkop!$yY`1 z;*2R+Q5G|dr$i{pmloEeK6VD^j*kn2TqKaeE$TpiEx|`nh%ZlxkijkbIkQ`)IP_Sk z62c%?36x5!!#*r)^%NR?`9c(pl|n<$dHPfxgm{ z;axkMApqU)!c!vjm7WZd<}+~?af6*UgHB2y#V*e4*E+=lG~S-4L`boV)ARf~T&D}n zGG>sw1S+J3_2{kV@#wIyFFYkeg|x69*ZtpWxO4>u%NTS@0_oE<%l?NCaC2qKD4r4_ zeVS&e@yZR4;LG%~V~~dg8cQ9>?QtqE;nFqq;VBUsOC3lb^*P(2MM@QZ4Dysf%2WcX z6<2{R%CC|VA!RB7&6?7E7C@FuVi|N=0%g;*fA6EKRiQ;@7Ca?F*);7RIc^ztiSOLc zV9*%}6h^P|A$3Mv2IzboPl-?%y~^{`9^$CHO{Z#pmG_aC1PY=NU*kvXv0s;0B_%>Z zG~&xr>d^`=U3J@b4Dyyh*J$FX$`6j1u8{-Q1 zEetvU?pxcVS&n=-aI8jiE(t|@uyycphZ?ktr&Dk0x5pl zzU1th8gS{B9OEewQv9@i?N9?8TJFnrV$fv?^nq3$=Ov}%XJnqoQzG<%Rvxdkj9v^a zT4@}>pg;)}TA-q5wf&$cv`B9pPl-@y0qu%vFsC^{)&1fb6eNL`6;e$?Mb>A4K37SJ z(6U0>6*Wft9zc3EvKVwl0`;TOg#J+_9JVx5<|z^CN27_(?e613M$vrbbCvheRS7hO z_7Fv_eR}{}G|FNE>{v6gZw!6nuBGi|j-u6M` zaBELYSA7PBNT6HMbRYFf#1;j0<0%ok6-`IAG_QFATD17c0tQ`^KoL}+ac9vzdIQ_}?(o5q~uDG}16V-j~x@T&>X zC*{`+x+#IK(AtDg-s}C)qJ)|}B|=wdZNfWy-fn;_mzFT-mIPWyJ=H7|-(-LWSn`wz zt)rf5VEBBTQ@t45^nB$F8zzC2=*tjS))Ys4JMZz72r1E*Ve3#G+$nHHQ-?v}5-6XZ z-Z%eh;=W^-jyxqo`SkQ2`eIWmT)JMirVNUZK&|M&x_iI78o{Mg+sjiT)anhLppqJh z3+pQHtQmA$0!2|TNZnWkw=rf{Nr_Mt^@3WIY##$HS~t>}L6H(Dhbm!us=F9Ki_-LX zN`!K#66WLF;W&nk^$TQBlmtqmXI;B-b8+d-)t{$CD2bkRx^_k@p+&!{Jz-F^1j?W% z!KV7Q;zT%&Ar1@A{0ntSc@8_xKrR>J1qv?l|W_m2z;Z$JK^4M6_8c}j$=s7t*1b}&9vLtmRS=$-_6 zP5ZnezImJlC?bugMCdi`^GXTJeg!RRG~AX!_a#sn6}9%)RL6Pm>?$b{Dx;!S-8RRu zOI&!un?Z3BXeagS^cIHCgBCq<<0%o^N&ULTCO0oZi+&YHF(_UF&7)^s;`^hG0s30P zQzA5vo^=~4<~9PzXihqV9!Q|+_o-T`S1Ohn^qb35A~gLz9mS)!XA?jU(Pa#JD1qE* z*rHRUhR2Pry~9%?^?f$sEb+X2GX_18Ks{*_`s61+a5~z$ zfTu*LCv8Iitom*PwCM6A8wMpvpe(BT>wZ-CDtzlMOy(&O%A%^j-gi6TePnpSg+YlD z$dguE{`PuFu?ir$lH4&Ey}?H@O2)^{2)RN|iv0U!4CMvS2o}=w+3Z2q}JX zzHY_}oZZqNx{N{3B~Wq@{XEdVGwJ}D4C5&gN)DnQ8(mb*5L$HhxE+IDNT6scZaI15 zm@z;JCwNMPqN%uLm-0#+TG|!+G3ccPnn*p>V*{FDck5OOPl?b(>ZwMX7US3Z(wtZZ zy^=s9=~og?>pge~EwZ1>QzA5yekEaB8#ObybP|06G%IQzEp7 z%HA9=*i45y#_6rAU8=l~((1!`jJK6%jL@2rir*ZIPW+GmvyO2_1no;bdvy-{S08MA#Xqp7{6-n!n7x9s94|NSeQZCim8M@LK`UN(Y`2p#?O#=deW;ud{ldE`#1m zpp79adft5=?*XXBFP;*ijUjdhc2}Vf zpeza0oQfarX_etZhWSmN5~1c){NU4j!W@8pwK>3`FA}H${d#Pv$g+C4t3#+6=XfO4j$9?Wv3!traJ2L2p1Zqqrpyflh^@h)AW<8z~ zp~h4Kx^{%}A%KFH>NDu41X7~8>w6VnlmT+Hp>1_DfD&5pln8C6Ikz>_uIU1F zY?CL0@+6QJ&3=6PyJ9gw>o@b12x-ym$2#Q-`~iaRDd7zIC4qeDGrHFP5H8KUdB#&B z{&c@1s!_G582-62em= zw1M77)8?M`2FSS9S_YL!AP1FNdZkx}9D^2VtMZfxIjGnh3>-S%6`*|!of%XrfdKn-!=)>Jt9qsKKKd(xW+bZUeXtpWb&R>`JS9Ri66tjJOAY1%^hjTeLFE#t zE=|C;S^N1Dv?$zwr$neOO~B65ocjQvFMfs$s*pgx2pT+d?|Xn^{CP@*ei8I6+IJ&B z{?(T-=$`}{Mx%+74VAFaYnKvFiO?__P5f(6isgk}7TGdrcZ2`ltdvAcGkLvN+Cz(q zs-#3HiI!%zlq3d2i*#bW8MH?NDOQIE4R42+Zu~u-5+TLv@Proq%mG@c5zQc538dJU zdUE!ihZd$kC)-la*bm;gZ4_GNE-g%yYBQFF5M(2o)V!*8vd*6JbM8x ziq9%z&^`&&gjOEEISnoZ=-C&Z5}_uv@_1mrRR}=y$F;gzdBg6PKwYUorJaYhB|ziH z^OOj6r2>`44>PO*T5w@7gX|=bHN9b5uC0z0Wvc>sN`$QG4ST*}M=V`isXBu}_7Z3c zJ-sJ~nqu*TL2aHAp(*tAHuM>Rb@j8XY#8Jqfkso```*8_+HmPiSMihxji$bL`I2_w zaOu83bYYOA1oEIh!$}vVL45|_C8yg0l+ZndK~54Vjrt7hCbq!NNs=~C ziBKB#8Js30JO=2(v1A4vkU-C`)2}dR&C!7t?Q`KN5qf@|w#JXkM(Ag54ucL#pwcWA zz5KX2Zvpy|$5SFynnlC^Qty+{qRysj!Ik&XAqnJ3t6KRU6L1fHtEoICLY}m$)nbk& zj$vm9YBT7t1X{0{B##|k4_Y)ah^ItoJ@px?P3krrTGU8o9D|%Cki&2Kjhn7p@qMIL zi>E}$;Wr)iuiFV%9)B%b#-Jk-$e)Vwy1qGb30hR$oTo&{pNjFe=w#rY){gh>7<5zu zO?yl~reoz$1E7X+JS9TY9@Fw!l=lj_bd$RHGw7HET12}GFWwn=2B6MbJS9SlXm_E` z>$%wX?qVCuAQuT_K~tS~hCH1L(7!4v5wf7E&H;@-PJ|W>Ps?DCs|0eRqr(b(H{s;% zus1vCqv@n6kPAV;>J@VltoVlrRqIDLW}n2=rYJn0)2nx#QUck9Q33xby(EB4&3HB(1Y$L8FXC$wSc8-bDm$6MUY(& zo}y3-Si1INrpqM+?Y6kZAa4Pr3v+yGbN*+U?yUVhMIl|7<6A3z1JB^QX2dh-h5)jF zk5UDWU3VIl$R?AgC}aU2rLr70>o_V=>iDk=@)1Be&{?*R&Bm>u8xwenLOIY`PQAEe zEP{GDw|LlCk8TPeEx5hAM>@u#64f_JQAi7J@5|dOq7k&BqCbOt1yD7#g5Jp|;}T7+ zpWh`QxqzMrE3?LS>Y!`?PFsG-4Z~) zFuGNDNM$Ixbj_de6oq_YbgTNQKJHqUbv?qM+X84dtSAfav>i`^wC>JR6q*ey%G74u ztU;Hqz~VB4?g*efXh42aAGaKpC~QAZQ78`@kTbe;z=KZTUq4}xzW^Em$5YJPtbw;N z#;5ZXg+{>f6!|}H1)viBR({8z00HzE)|!sJSTG)y==VsTqR?YlYnr+8q78zot*RJw zR{$xtlhr=GCkJmEd1lR16jE&Gpwp1!DUYC+Ib9w#R@i$2NV2Z}&$L$f_V)h3QxuY{ ztB=|?cmjfACXQoJpa6;o=v$qKmc8Y@w(B@**_5ET*Xrqx(MTS)yp^F-o%VWXBZSD zfF{B)n!84Oyd^6(+C`14iLnHrU)=E4#k(13+6e@;B{jP&r=i%f=gFo=wE=KSNFUa6fS^%LDzD`D0AG04G-ig3jKnvWlmhdZM2S2u2U3) zB08XTjLzEHQwEs!#~oh(Ui?4*=Q_rDUT}Kh_yNaI@z?g|Df(B<^MWrO_*gGR#n-g@ z!2YU8LGh(vvZ*xeBtA5wt2IwiND3yKG)kiPBWUc$S_VZ4pa!VjZgDGeQHeT#;wcI> zK<#o(rJL4zk*1yQgFz=x>u0g(^EZ==dzC4@8jC_EZMN2_SzM z+#4L6jTb+RGUX`>`NQCze&U}e2-=eRlR@zUs6kR;{e~%^63tEHDGD_}gFsM-bFZji)HIBO2EBc%E*Lpm8oE81zyA9fCfN%1fz92MLOBiP9FrscXf=&h5Gbm92`9sfoX+ECSQtZf66jFkRYA=I{_8V|8#ZoNS#Pp)&#({mrsxl>+xq{x|6Vk) zYUd-(9=(PZv4!fu~eH@6PAZ?za&~(^EUAldt6S{OI*UA}`E`Yj2{j;qY zi;q2ca-F9r)E(+y-tw8r2nuN~|D^G^eItPG!{GWJwXezOH@YLmQxv)ngX_}o4>=%c z+!hrEy%j(kU|&mCc3nJzf&r=jy0)3<3ImZ%EiSqpN8T4KNMZ*djol|jml~d9!o}y4R ztf09gTebl~2ii)9HP)jX0kj!r!*1oD!z-ibw&N)ZZHC#fPGe_JLQvL*ehm5`fVRMj z&X9NaU!&j1e)=0%#>X>xx?s#l4V?1w2KemGG?Fn&#|+AQPQX2IUDLZK$vZ z?!Fj_exoJRc#1;WP+{NxI_iO-Q#Ugi^hE&ufI|Y7H_XFRqq}{1ib6l&kbw3It$rXV zyKNZ`>fSh3n z&3EvzDF`|~kf$i*3`1!Dv<>hI8a*pB27MDim!O3kP?WY8LEWu+ib9v5g{#xM=Nwd` zk`JdCR3w1_GeJB08%f8MH7EE;SDtZnxrVC zUJNHwJ~qCEO7!+p0)xJbAQOUYD^B%n-0S`q7|NF+E9|Fi8_J{c- zhWA64uB48qC}a=&!@_n9i$^8WHj<5KtVbmR=qLpcHp9Qstlqi|DiuJMP+_;Yq~HmVS&BSG zAxo&RPrGR=pc2iuSjV6;0b~w)kM{qycSI!`v7e_XWDa|ef|{=$jG!rRtr%1;fFz$2 zoP4XF*n+&okw)44oFA#bQh(L49&BgpW6_sGTyTPc8E zz#BHPwZ$+5Eq}mM6nX(~*d3j-6%lk*UX4Lj0>~29o3u-Og@;HzI`R~SEMdKg-MfJ$ z2>NJl$e?NgR0Mk_y5AYM9hKhnGw81X8Vw~{H&pH=g7njP zibA8IMB5j+4MWhO(asF25kQiYz&ahB*?^!UV|a=}l9Rx^o<7H2%U%}(8B{BPmc+v6 zQ2N^AxudcsDGDu#g|B8EJCT7(bf7GmLH`7hWC?coo&(ybM0({sMIp%&Y_Gijc+tdw znI#OW6F}$TdlCB|o$raDiY6%vormv5D66c}LnW&AZX4BDkLm@G9SkTwF1=fepzk+$ zib8fUpjafmw*o=7Er&CxK>(>iXIbN&$9n`Vm*FW2sX=FXc*5>b1nIBSV~~~ne~;JM zK)bGuN<|rhhOXx+3fVxr?%0B<_{fCTQJWZKEr2ZH(%E(CcnCq?o1`dY374)uBmV&^ z(XBpq3_2!&2EYW@_gn{wk9VQKkEWN#W2WL02M(iXiv9yeGpWh%Tp98f>uyBJq^6+_T;2o2H6Rq{cuw9 zN2@zO5o9`1Xfg#tQ2ofDV?xuEe#waQ|prDNj-8V2Qnsu1CAks6?-3sx#=g05XRmn%VV{cv@9z z7Ee*g9ENBHrEP47pwHKrGssZ@1;9hKv)q*g^cy|&<|zsVz(X}6c3%{NLK`d?bV2|* z!INR+hkIiYl-z9p|EDPA1W$%Db-ulAxQ8IkUM3837C`1lVSm_=t9Z6^ zRd1f6koi$K&uQPknF#85%#lGZ0_X(nYk3#6$^u=w7B)OZp%bvLC3Ss19vJ*t0VJ7vYY}6C z+YByWc#1-jskbd@PkN(E=dB*fpeq8XB@Ae^lIq$6m1v&^Pf@5P3~1<2lKzOGUhbI; zx+;K_p^-n~!Q#~js%w&>kTNv#+i4C~Ly&hx1%o^VP&b%|U7Y%6DuT{b@)U)-!948C zC*S9x^^ctvC_HVfuwL?L{iET`0L_1^y{4nTN@@lF&;PmpaokgQ@2_&gkNW#fQuMDH z_Y^jh1-EWMkY%_k`>U=Aihl@3Ki;idXNZcwHG-!obO=U2>Q7w9Et%&%3>b7>0G)%; zc+D+|ZV0;7lcy+j4o2h8>9)YLS|Nwb800O0^59_3^&_=n5oBk{QxwXBgE>zwD8p;( zyxyK-&aV-x%a8fcioo z$Mvn<90Uda;VBCBg+5Ms*Q_%LvR)wjtg+Jh37|jl7>Qrswl#uc40wt{f8a4Pr{L38 z1ZCVB#GqROC>?HwEoXx(5#)KBrzn&TH$&1{2mF@2++3GIw*^pc9USU;`7%D$!(57| zD3n`gujBqW;(vDAtX{{UI|4{DtJUu7Slq0=WW-YxlFVu;AMLagT{@Q-D+c)spt~=j zXFYz|Z*=KwV|j`~cVEIB&JI4a8M{TG!eH~&3@?g^ksm~2|vVzMeKkxv0n zQ7957n;uNNk&B?YQ)?I$D1c(%W^h~+V}l?Q9iF044BQOu7q7-kZYsRGKX0tC_XW^* zxO7YVEW!K2W3TZPg}%e3yV9}(FI~IST#Z2w1d!y|gQ8hm@lt_{QanW=$*~8v^K_P? zOJ}mtkU>ELXaID0HO*F}p-Y#tiKi$u06M(+@B11fXlmka20au&ui&iBqP|805VRqQ zrzrFa&e|N9eDw-~?9`kY^hf}edc!e1_s#KdbZjC|QK;10UPpfTdwer&aSdcpumGxu z^;eTR1>%oO7`gEjh3aAbRoKO&A5e+R|0XjiL;$^pN6@UchIog8Y7I|O=rufodafV7 z5kZrel`!bB0IGlzX?!ch^K>hg^Av?DphR8Whn+&uf#9}rjrAy004;?bx_y_YI-#54 zTnJB5XesQ_9dK{H7~G&mSenabXedjxJr^cWN- zfRy3MPT6onMvRo;B; z<_-u-8E?m+2murd@&#y$#{Ruopp->nX-PC-CC4#=Vcrz$c0ENK@8iQYIc=Y4x zd7h$B7;K=?movwWlL^0L7!)Odq+n2DRn_`nRHCkbc#1+&FeuS1r5KO&c36;>4{0;)J4{U(L2oy!FzA&4dH}uhZr^|7v(27w;VB9|fL?jCRV8?# z*Vb2a7?dc0e!yKfCaGc``i)`|d5S_m;I2#d`(%JhG;E{^gOUW$co;cp8SBy;L4#Cy zibCUIFSLRHB*caSTcoK$Q)!_u;!Sp6~9Z!BZ5fY_QjfUA4Fr zL0NA33`!F~7Eq54?CSOz{YIhgJVhZ3s7Djp9GHb5$4aS}jrHiY0FoSFEW6C77D2D7 zc#1-j1B|~rmL5RR7Q=oFN*6#gVOf-l-3m_xty;oU6q*UkqWWC2x`Ciap;`=jBY=j$ zYa-_6-tGu;e8N){8Un8gCC_sF8x?gnV$fRwl=2IXh4WX&kDwFX~WSzoO z6gmVqLuMB*+-9&G6w0710pu+utG)AEJzk)4budp+$Xm)mr%-;rG%Atb$xH^l6F@!@ z@L`IQg3+i%7oB*DLOv0&)^z8m#k_m8Y^t1>7@0Ih`zYkO$RIRve^ z!&4Mm3l-MkUnHKiHEd zWFw6A+QE2brrBnmqL5@GOw_7&;Rq^>JH?<+0w@Otos9d>E<~4Z+zXzfP!0?_Ej_4Q zji7O({27!hfQ~?~{CK6V4}yA(<|zstfnIs_@XKouqi)*J1>{Z<3;rH#~xV9~tC`pn$ouuNv#o7Xh>g zcFEfpJuyPi(|J5ap+&GuzGL^5c>c)Ve-MN61&{)CmO~w)*C6Ov08dd!0XoZlrc~gO z{4g0^1{DY($w^=e5pD1yqEuO)qLAbyuvVG&Dd^Ho+P;oKUj78i?f3kAjrfvfzgW2LnR}{S-h3(EZ;rtc4$ff)?@=g$$tkKX6ks zyfX9Mt=$YN6+m|X;JYiAcfCQ-hub_wA-jLDn`1}jAM_g~wRL7tnE-Nwu4Uj2cf8f~ zayy=)kQ;O@bIo_CBB-57AcM*U&?G_`YnLc;L`nhdUz3nj;Qezh0@^CSxq~% z7?sHIQro1)dh|yC*}+6rTb&VjC5*2JPf^GYCaOBQ`1&B|e%&wzRS2LA7|`gvLLdJ| zr|Wr&LK!fiG49$DytrkTksgC81<+!c9v$COdKtQObB%e5LW^N~bm*TaD^Q6pKi$Nj zDgk5&gSYRtt#Cuo#%DZ5Aww9v?RjS|5a@yf^OIJPHwEQbpq&lCA1mLcj8qvCu@0%LeF6p&Ca*4 zaQA=6GIa*k3m`@Kjlz--4nigJSy5_>;h&Y%VXlmug{)4j735o8z2 zQxr;qF;zK@FL+w#N7wxfvg-KXo0U|d$9F?Ue-(m~y73f+R3!^0eYk<$16on*BVi|c=eAji$eaN6=0w^V3M!TV?3ak z&^pE>^>hZ=bVTbIRo8iIolhD|w8ybN>lJ)*)LNLXbK_>)IZ1YP(gmD*U3P6?pia7NjH!2kKMo6B#WqEK%* zqpW31??DLqyPzL~P79!S@H!s7cG&}T>0%6cibC(;bzJqy{sMy9_-irfi~xFH3QvY_ ziugAwZIYtU`%-%y=NY?jBhaC(5rfVOpdWD870N8eohh?+JVl`&aM#ssFwsOMI%{sp zpmPFfUKA|CzZ*CcUAkSnc#1;vqU?3j##ZA2Sew*K400Agl4nZdCCx!aj>2wHzjl|gO-NCSo|pSxeUf*_UKJVhZ5 z7_M|09fa3x+qN@ckh=gH3O(zueKm(8$hdAJV6qaQ=v6Bu+^07<#S-iP2X*$5hZk*6pmceE8W2=l&Qlbc2qV2(a{Wgk=cC&fi9zlNh2QkP~0Cj+qlDm9eF&#nn4|s|~9pI$oJ5Mj;;Ywdc zT?TmxAR{06=EFEcTw#YP@f3xOd|>IC+7(k&qR#f~7<5eljd=)1hec`P?%`kuo}$p0 zhp=XQP$@o^DzVUtLDvP4COjEF=(oeu(09J^6ooY5$)MQH9FHRE>UuKBTL8_2C7>k> z!V}PMq%fVQC^Qe2fGT#&z|&`^uSGEEh5+(}bt6CBXW~yj?!V4c6!L_1BS%K2;3c4T zt#cUUBY-3`k4IuY<5{iKZFq`8l9|T|yN2V&$u`p(2Hg}um!HE)L{sPCGs?7f@DznE zKeyKzws=AYy6cX;>Ym`ihNGS2Kfn~ z<}mDU-9iO#u3RvLrzq4MhW+QY&Ul1MWMpN?pj!gy0<;-aBLi_~S<{-QD0BhZ3~e9V z@F01Bh8=@~1d!xoNAAO<@u~0^r|=YoBp*BKVPX0Z zl_<``n?Vl+ki2B^!-D63QHd^H<|zuv!{UdqsohQ^Xu_Wu20ao$TcKU|Y(PG4ZuP0) zDGF_cc3r~~x0MLeU6jk9U;(5A{iD#R0A*C75sP_>LQ2p-TB;XTD_|0LSqR>hhXXvEX@hE~yJ}+m`69M!dn#4wD4&qs@cX>QT zq4&@vK6oUo7(pLq?`Ke$0P1v4M%z4g2tIo@PM@bJ)ajnR&Zutf@F;PeuPcMX1<-hC z*KISOCWC&XUw%AAq4ChJ8+%1l5tZno>>~z62%!70{c1zC^0Tc^IS$6c?djgdx)`X`h6bnaLnk-eo&pLzTat6WS<^MT1p#V2Sz=b5d zbWJsdrzoTVH-q%4k?qi>Yd=;#qp==E3m^k%sVa5qF#wfFb{tPp$N*ZZbL{W5M$m6p z6$Zr!pjl8Njq*1Y2#R;(DGJSk5@iMreu_2v>M0F1x8T3p5J%ZIreP$?K zLnZ1P%u^J41gn+0bdbkqE3N46$Dro|=r=S@CSMM~vn?t;c#1;5p>g6Dyxtp?=-#0? z2E_@W9kAQ|#LK~W8)JzjPf=(G>~^o3qT3rmmD%|WiWfj#Vet0%#9BPS{^C7PQK%~n z-o8Kk6t@}rsY_=z)}t2!ND&5|TD1CvPa=}h;3*0z!l09J+;rS+`FOb>gI)@t0Ok+9p70cf zq~5?#XM>SDf;M;cU{I0(Du5>Op&-YP2-57vQxqzICh?jPCAdU~%|aQJEPy)I$Y>9H zs9J}hc{_QELY-=0neOj${A5^~oXMaR0kjR;b$d2G2}jVf6rQ5cHfYyn>_2i6UAmUz zDjAe2fC6C6_QZ`IcyuePNs2-Nux7hK|6~G!6fY@cHCEU(0TdDmPllC2b5V(8Ja~#i zA(1c~-Dz`k1W8w`GU&Acx(VY9E3!tPM9`}yDGJ?$afag2g`*Mlc!>dn(gl!Y5@hI2 zr&t8NUdmGxl1zfMl^vLlpl)Gi40BwLgP01ds*Hh+T+Wg@>byj`0+QEMP{= zLVe;i1X&j+Fep<1Swq`bzPf^Gk+TQcEcYi^U#=Ig1WeFe~IGpp#I=%1c zW*9P`rzm6thjV^^I2Qj#tMAIbYph4_1W*h-RIA3l#=~es@9`9cV&I{=a6$v_ESJd) zW>B^O>I%;~H}?R4RH7JJo}y4!c-DEQCFh|_*Uwm&LGK070$6xFvwZXp1evbkDGDus zg~xqrjbG`&u*Tg8=e^PdMF} zD}_fOll$=$g}mSsPHKm{;AIA{tvwm^Q2LJeJ^5;`C-cvKuRGo&eH=;pjeX zjWtk-?ltEr3hBXcbh>dSp5xnRtj3@(0w@pON2Zin06q1F9 zs_XerhG-q*i~Y_F`ij>vuGl(AQ~!Wd2lQ9T+3^4TpJN4*4@;HZm)ec4^UEeF`d3Ll zEcM1`C_ZRw{ntSDR}~71p9{;PCb!qbm2N>HPf;irmPHL;)B=yj%j>2v=$im4gIO)h zx>zSvqUO_iib7>DtJUX-7e344j86%JiUd#&tSB=-k$@LJ7~kY63gy6xvV)%=d_}*} z>K5(ZH&(i00i+CdVWV$%Z&V_cmOMovWvC1PU;zk%j5ZBt(02i(2qkjL__Yl|GdA-S zg%qJgB~BCYxajOSJqG;{Ku2M_XjbAdBLpeN^Av@S!gkRwV_c@95-IoJ%%BnhbO7dn z(jO<{$)@rqDGD8cIUwbj*=rHxf7p&eKLt=MoT9&Bxi#J)XnTaGC=?5)=u4+e3PX_6 z2X6+I3LwdmM=iW+@s9PsO;Qw+9C_68%)o#jIA$X{@ln1<*F=5%}KO8igPqBc7trHs}#}TXnyMpxP*P z2K^C0neb4Z{(RpT1jR-36ooS3p=vg#9bR}mRdEG_Dg@B5LK*Gz>izLNomwBBqR_BH zs7F@cuAmaNwcXF4N&(b2R7ShnaU`Bg&1sUNP~T8Ee`xiY9q042eigbGC| zynkU%E>BS?3DzX6RrkZk=}1j`#Gq;cbQlK7_vWwphJK^eCMgOXhC%WdF1h$jm^!a? z2K^O4b@1YQ9&I}dLFw0cib8ep;(IdXqb#~~JL<|AR3m^6L9^C%y5cDWY1H!+g$_Zp zw(DXyX#}|%$$x09N3{Z|E%XSy^p@e{bS@e56ouMCk6@VVBQpejk5XaKKLMl;huiyk z#o;Zz<xV8~eUlW0f-3EG`sQu3LM2*t z*n~m#0_YNqiyG~y4MNbtBRoZ+OE4}P(dQgqpmHwXkwFat$QmlF@y!T11X&mG6ossz z!j9g%W(q3Nh}nJ&vg-8Tvs$?@m%4k++6)9q>+=+aa$zoYe@x1H1TDWE#~^C~G!Ysn ztxuPZMbNlAJVl|2&^T#RG#B^UoZA#I=$HWd3!SORPfLd*XjNODqR?OHOs%}~_cAKc z>dn$08|#sc0D1{e@2WPocob1(3r|t#B|N>055!+ZP*h?+2H6Ury)fqH|JK$QmB=TF zrzo@+#@x2NZ~hKJvsAPhWG8@Zprtx#pgUeYt2m0MC}aaIRmY>LBN4Rjgb{=61&|3W zYK?0#4&MwDPVy9mOkh!K|AJXF5VZZ9C4(FU&_FmktWGUP9bLM)MLb2JfpBzKMqEWH zg8XKBFzC1dk{mHHan#XE2->5^QxuXMF(GB?lZqg9?@$Ig3LtlAdk5@u!*f8gH+YId z?$GvD{niz)LVwbb$)FPgs21KwEu52Apc0*Gmi7NB3f02<=vJ{5e%1|ISIHtL1(4(f z{;)Bp@Ef-GdY+Lib6?nw$kF{2)q%d^r!)YP75Fx=(cqDG2{^{(K{=i zqL2%8TSD4y$1{&6g=P#oBY-p?!m(5*dUr>V);FG_kmf^JWl=k=4Bg&)rkrNbSpk&M zQdaxz>%HR=pr=okGDAU z4MCGzykd~E02%^qhDle}l_97{OP-?85NI>_YYxakko(3W2Du2JDtLO&_u1MUL8hB{ zib7TJ^u8&#bOVBh#>nP2)}!+RNU{WUkWMbXy{ns~C?r_|8msF&20`h41~ce_0NMbP zwkJX_;N^u;N<2lO4KQh2{-8%3f=1ctGRRc`l{|q)e!rQY(4}i=lA=(_6WFZOTqg-Z zMMdivwTI6F2K6e$4^=IhNCsUJKssJiRU6#5LKTh4WcxTU&7smJHW3VTHW{exBLX?w=w-o*62 zJVl{@unPV6j5*$@MDq@-G3crQiiBlRSDWv^{UhxoJVl{MSQeEU8;@6%MQ0l_$Ws7i z!VLa}tmXRXH#+s6rzn&OGx%K+M^8p2TA;OuL0$sL5$1rN#?CQE&=hT+qL3rZ0gcrC zIut?9p3V%qCV-wnXZgaijf~5ktQuK8Y6x^GqD5L~S1#Wz8c?Xr~&>lMm z-4Z}v04>b=TaBRUdwGgNUI1OuTKWS)i_^RrbXx#@g6@CL^CiO&r16@kDD(-s|C9C< z;Qf+&M#eJejsUs{N0jI3KEyL(i&S`uLKoqP^6uM56rvKzIOQ_PUjRvFhgV%L%R(h; zd5WhfB$*w~3em^V%Hnzk1qdK%SSqkpW^8K&P5jPN6q1Ie0>47aaXnf&Q}IhAzmMa((D1c(&(hZPNx{EGdybMoKC>Ac=s`!C;F<#z={S3Mv+!g%+(+Rmf|N6Pib7}M$#7`kbG$(1wVHf>V?7EMKyff-FnZo; zywEFPB2Q5$4yFve=FTlamu|6}3WGufkm@^l)@@jL3zbOKou??I`p#ZQqu)S#1f8py z!=T3ks3Ro%TpBU2+Q5WoU4rxRJ`1TL7@WZ3=GjMzkO;IDpBwXo}$nh z7^0DxG2je>Rt7pU=!pQjD=n+-^sF;pg+Ay$Pf_Tuw1ZCOl{^y!t#9MUpfCXx3Ny>j zd)jEA5{+xiQxpn?ndKv=+N&VQWpf;Z!Ud2qbpNws)9}plfh{~mA!F$N-*3=6f*`kN z`3#B>Ku!P+_)&+a47NSzDGE6Oq#Br^g`h11qzfABQKSG;gr|3SvtL!{(k&mzQxsB! zr?vRwMz`9?cj=5uigPseZMQ~h_U7V~k zf>N5KD6|NUOWNDkdoLVWJE8L!+rM3^d9K&>TJ%>Pna2O~f39P!g3Xl?X+d~P?+9Ipie&06OJ zhsU7nJh&xKQE0cMkMp7o_pIgC`7M51cH0QSVeS zg8cJ&ib6hcvh=5Os(2B;lfEv4(gcuAFihGu^!bV)^Eo_4A)8=(9iRBvujn^=b!#1i zUJIbPuz%tD**rWbasD<>QD`pgUuZkm+!8_OWUU#LE`THhu;*v3!fO)Nx8f-ZNd{nJ zMo-Z|Q1Dhy2E7qL9a7;;7*oBC=+d3s##0pPkZP|J(0T7a1TB9V!JxMSXg1U%hg8RC z1ZgMm6oqC(JqqoWf+sTGE9WpMLjcvngkao-QrrlP8_81?s)Y%`sD^e+QHh+-)G#Ph z0QE|QrPOC!@JO%uS)QU$uS8h&*I|Dof~J&o|JGPxvjmU<+zhwdX75FpuFX%LqL2aH z3|qd+;Mc^>scH;*CxAA0kk$S-d3P>?j_L3eg*JC^&^hvHDL&lZ`HCTfvIS5MG%$*H zw#UOPr>^o8g>s;Qp&#IL4PCl(HM<$~UI1-|5h}knIf|%6duw@$LYrZPYQuHaXav=) zaAr`B0P=x{s&8c0HUza?#ZwgWfrskY5m$R4=xJCWgFXnLQkbV}^DzmxHhsc*ibAC@ zPggNg_A-Lj3`k+nM*)-sYvdw5S~o)_S~rlVD3k6k3$DGCjSXWc>f;}a3|@4g;`@&wR0=8HS!%QQPu;V zqR=?#@LpZi+y+6dh9+e^eD-jK?MS6JxmCW`0d{jL3YVJMWOXDAvn@%GM=YP z8yw4^uL7tgoU?7F7wBL@8xKo&4o zaCB7n&!|Kx-FS*Z7BE&&UDX?}Nf>Y-ok67ns0}oU&3EsdhoEx@d5S`9ph=wKS9uRz zy3Cw%29*h*ZLs*^(`lud2zu~=rzo@y7C+3)3BX;;KU(tN8|zWI0Ga?T)dVlAwg~#7 z%~KSb04>!XoqpgsAg^mG4EiO2Hp8WxIrUKyf(~8hDGF_dOE;>U_A+$ow)~sJpx**$ z77XqUjE?@FshK*SqR=cD+#CA32Y%Lh%rjxo9|08HRaSeJ&31h0Oy~0yg@U^}=wvtB zg?Fr1-*IG6g#c>}ErOXn}|$Dm39N4bQrV z%f1dmkfv)t2Gt0lf$%DCyG|n%K^@(Aib4b7RlX#x6JDmXBu^|9p%xw!H^~>IKjMcptS^TZBg-pUCkPg$BU;sAHPv zF;t?+ouLeB5I`>Q^lm+BYXd4#pgB)b$OWF>DHE=@LQu=snGCY({NJ0EBs)OAht9z7 zqjybG6q4)!-C58Dud?tPQ^_D}0i*@f(7hi=v_~b{rOHzj(t>H|7vC1*9^as|3MGvd z_Lu<5hw-|_9g^@^(28?BMWK8cuN%?nHa;i8@4G64Yy?mkR9J&4>3AYz&kvrWP#9F$ zdcDa5(Ql-yZ@?g10rUmtyQ^#SaobyY4o^|&3(R-dIWNNl?Dub(G008;{enJh+oNy# zqTk5wHcwIL7xZCMJ%-@f;qEf08DuYjdc!J<8CKt25LD45MWNoX%HsaGs2FtVwr}=l zkb?m7fZID^d(;pF&D+9L6!L)E`}K|TX9yY)m%yOo0!Z@J%dm8Pyk@(iNs2;}uU?k5 zZ?*?P*9R9d$WZ|Ggt4I2zwPjr1gjxDMWLQB78Ja+dp3f$*~tEEtVbsVkoR}!wv02z zv%@ECd5S{b-=R0LIJ5(T*5nUj&`AMw1R9V*d$;2$1I+@SqRlk!O0GYzto_4b)rlGrTk}pqD$Q0K0 zJnwua96?2`tr>J$0O><-Lhf4xejj~n!&4N}hu*{on@_zFRK4AkL1zSzE)1i&jM<2H zQok_eDGKSrFxncA7`$^WA})eKX9Z9zSm@Pi!TK0<>B8c9ibAbmq1Vzbx8|b~Emh88 z&^ZCL6I!Z!WmkVe(EO1+MWLP0QXMwOzZyZCoNE~5EPztNU`s-IEPj3-Y7X*;xxKZPXS1S>8;4V*5NOIh$P1{3wKtr+P z9tODzpdB!OG|1)*UZ7IZBt@YeFn?rg`^*qsIuBE42Du5K;qc^3qKlx@$|(%GD1amz%ElZz zy8uCkBYBEKk_}~(uNTWBD8{CQL6-zj9;~a6F;-1P&{bQWqEH^JtDoPk0iU{d^mE&? z#(Lx-fS$ny=Yz9tW*}&19#2u|8EkN#8(L{lHg`L01Hj7y}A(e8_j#oQxtN82IR<$wZRB-9}&x-YXazHxs3L=)G4nKv{sp?DD<)% z4(I$U7lfc!_PGqYE`XHb>HV->#ajfqJMa{Rl;P?9+ITu1Il1?_o@O-^8O5wYbJp6xs>%M;m6VPDQ`bpW5XNx+#F3K-+uDwE8y)`u&fmDD(u{ z-cxoTyN;kZgZ&Kh6+o|HZ}^qSUh@&;vXG}J^cwbtYu|s0Z*S9kt_<=MKo+o#@o%YA zR|M$<@)U(EU>l=`rF;e|(Sdf47<5YjNlvA%ncVjrg7n(+6on+GQhVff9EqTxX6X#N zEr2?~5KUH%VgiCb@8l^8b%G(9m+#5 z@01~Hn{P(YNhzM9kP5t0>Sm1Dg&@T>CJeeKfD~Y?peXt900dPvNl{1v#tIx2WG^5n zIl_@afdc3^RJz+r$9^N|P9#rJ=r>fl>+wD_5tQD?k3shZ&|v8Bs)fhmu>xNuo}$oT z=|n`R?i+9 z_CG^m=W=<9Lej8$w(FcgA9U%iXh{EVtaJ|r&}~?)R3)`G9hK<#6rQ5cZCI@|=lOj% z1ikU;$Dl_7Nb???Yx=M&UVruQCQngF^B%M|wPqV2XhbV*1_cYCOE3*>P}pe>Dp75d z6ooFqG_=FcHDL%EyWNOEAp)oX=77G8>9`+3olSX)LIp4fbj7hBuCN=DEE)7z04;{j zRH8>fAcEAAd5S`dp)=()I|uKy{WaQyL7@W38OC$&jW$(A(Bm;YMImPx&k5h5oR2PD zxKk*Do(Lc%Xarh?55(QWho^XoLQ2pG^wJH+EgARHOa_Gspd$EG$naAyAJVl`* z_>FR-;_z6(=J}Nj3Ku|qL5^)ppC*J{NB&Jukfd_!bS+7LGYR| z>TiXYyFYlqQxqBmuL;dHeelzJP1msuiWETm;Ju%4ToccwYIfr(3hjgUer3ZJO>{FT z?=@ghlmK#uRp=jvs^Hyc(ffFcLe8)Xee9!;cpCb0mKlSh1<-D|y@$03G(x|T#XFv& z&~CWBZK7)ipb|-qJI$aN0W=&o`?u10i6@(qnxrT+95(y!4!f`uL2f7g85ApkRy4?H z-y9l>`=Q=WJVl`u4RGSIe%KTQwJT0w&{F{<`L48Xmm6ErrTfw(MIp&|rT^yq`-Y%$ z{UQcE6F`gL#rN@MBA#EEJeQ{^v#x;)h?mobVzd4+Wm0kTNWOSbWRtHiF*l)@4w<0CIt0 zmQx?*;iUo>_wW>jTws_b?e=BdQY}wj$DkJiC=Z_A-LeH8;8T3*Bt^5X?EbLF=p7qI*JVl|E-|TfhPkx3E2ueNa$)E%QlnURrbFp*VjY<^l z#8VVXg>T!LFEp5iN)-1qfU1VL9yd5S_)V3YEMx%UGRw0m|AgAxVM zeCWcTPw>V^S!(I?6ouwP7vB5qemoQwc&&y(Ndl-NjEl~>T!R-89l6d^6zT}$qGNn+ zWTFzasq0?ZSYeX|(7{sJ6}9IHZf^Cf=P3#uEVb7;{J2~lL1o5j3`!9|_n^%%yyGBz z9^;EOJVl{<&}Nt(9s3$Vk+FshN)|x+&*1t$gbk`*f zLCYTIGU$T+s#Qg=g`{C0IIKH1jc>?IdU)XB8`5SI~k8$KF3SIbXuT$gai{G$kzq&H$ivUW3C&SMTEAT^g zO(9QFC<&en)6~=PDs=mqj~J9MfZD)A^<_Y_Pv~Ya(c>u!wSk9f&GZ|%rF!XlI)e%X zP%zwedb6BYA!wsFPf;is?z)S%pR3Ssbfoz&27MJkgW%GY#*f8KVnZpOqR=3?bceob z<0nJsHS&KO>rtTqDu>(q#evLss6_cqQWPqO+xxqX)i_k5_(&B7eG@=|Fu)!eH@FT# zo>4qSp+Fd5??2H3zhTGpoXemh0n`hcTebBrZ4jj1i>D~m3z}OuC#1heC0cO6gh9mu zC<4a46+YSFna99`JVl`h826T+Y=>XvTk{+l^j!dT4wTU@Os(`pC7Sz%rzq4p5YC<* z7=mBr#aez0`XPYg;StneyaO+ojMe5T3dO@C=(%IcK6L3Sug5W{L;xkhtNgo5fGR3c ztv63mC<$KWO3u9k5Tw|U&!C?IXcTmo{k*eABB-=UibA8HvmABZHXT9DR!i43)}vAZ z)GQyq5t&|Qi=gjKQWRmo|gS1<-7`bTfbS!qD#CJVl||aOoCra>u7d^*v<7pkD%LH7wviIp-Jt(4oF1 zPf=(!EZ`rsDnkyH=+7HV2K^R5Jz*VVk1i#0=+b?E%TpBU3F{ay9;iQypcqvT2K^C0 z#;{BGP<>%{1l<|SQxr0WUAn<$-SMJU&-0-Sst`b4a5H>M|A>DhvkN>$AuqTY=3IO( zg-VoFnaQ9^0VLU7*irKqZf@PG;wcJAb{Dn_{f=i_!WUOEs7e5RgG;yTzNrhkbY6x$ zMWJtS>5dO-z>Sj)4-{$}D{Qp@lAPhZEKzy_g2n{#6on*bIOkL=p8!R40Ijz|C-1IqncDQO^@RMWG>ZGfa;;R*Rrp-x3&9FMu{f zb1Sy}mWc?mD&i>$ZHDHSW&LK{fHa>~#GnQNv<(I{>`a#9eaG`>^Av@)!GOl#n2OG* zL`OVj|25VlD~12wtkk20tag0T0X&(g@5NIT>e0eMr_HNP<_NMbAH*PQ0i@DRRy*Ho zFdmMc|BI(6q|(hnN2<&lcP$6b)n(8z0VG*UUFf?h2mMA3O;Qw+ETuNn9u$R2wBCCi zgKPv)9-P!VHY#;Eg2vzADGKGmNv;1Lc3FTR&z9B1+37(r+x`C%CG!4EYH*{i~HwZcw8^IuZ0pts9Z+qKw zN~lCePkD+$zR>p09QtEDg8uc%VUU9W3WNQUmX-w=dalG%6bgg=k~i;8#l49!)-?<| zE`U0~hgwFYZo(%L)HX>`r~`bcrPAjDZr6q6bgyfyu#N&K8*cBGomB7^Ij;{qMWJlC zy?y2H3`Li2tCkvrP6!|)czSQWYlug;{%P|Rg^b|oy)&~vK544ubwdW76hJQkYVV|h z8z-4fQWSat(1*?KRZ)qK{M*eSCjk@$t7qHmstiMSomU-CQ78yj&swF&;)o*=-98ruZh0lfebn=fOfzf--S&HNeJo?!BZ640dsuU z2aOno)-itWox-3q3TPeU$mGpan%@~c1YPGgL->FG&vlH7H{m3r867OpU-hU-ivCrK zH{qD^*LU&2QT(YA_E()16#qd4Eaw>7^|yohM?;^7Q(7C;7Y=XhGT%tnx^Ax}}r0PdW~mfP?& zbWngEgIokqFm$G}+qRTOko{erqEIk&rt}6?<3?b9tIZ5LFMtk}!I)b|lbr~9-2R!<5;3Q8`Xc|0$lh*&)j-Ct-MfD82D1es0xac#*ilyk%Z7=32 z3N3+gQG-{ZTM?u+SFxe7!d?A7N0!ZN_K3%HpHiBL;a2Ag3tU^B6P)uW+uK$Ws(@ih^wg3cYYY)YUzmL0$rA zAv}UQr*?jXO0@MNPf=(gJc7QhAAv`xT>q9c=$ZhU1(5QM%ohk+Q^Qjfngx*br&|2t zyKf-hY}k#=sq)ta(13sNIkWzzxYuT}kf$g#;GezD#Md>W(9KYLQ-wj^0?4SftoE^Q zXP2Q8#ryIUg^XG|=FDKD)Nw7fr0QD_;w%C`m8;1MdT{c#NP6F}w7Wwk$k z8HGDjClBxxh02>d=tRv~ABt`Ur|f(N-4Z}r(6i1?ZoplSi|=`gLR!$X-f!E^6+s=e zq?L!6UdIdg%j&fl~*aDFW^`Cjg3YO3J^fK zu$<#)wCWuM9ofWF6v~C=96^19o}x=9|ICs>cLk8-5GRWOsecF>@SLY8Bss*%dgxRu z1o;m3V9-4QBsoXUpgabT^d20-QxuY%BbSi+74L>_eLR#wfdWYKnpjt7dl!|cvm;MY zNb;K4^vessCPo!yGU&bl>Ip;2W8dz>bE&mWQWWY5L(20TMw~#G?(vLD20ai!lC6jt zEuK$AB|0{frzj-Zig>qe`bq@7zM&x1SYd+%kR=R-X@qS3j-VhPo}!Q?425le*183P zI?0Y@&_e;V6?zkOs@tX`Xj3bmqR>|8O|&nZ)d@lCw;3?#kpSufO=9=7W(yHC(1fQb z)CHQvW-U7RLr_he8H0ia(0&+myE!h~96>MQd5S{&Va)Bu&~^BIq}}f{gF*z59E|4# zw(qzCLBsm<6ouqqJmD&Q#!Wy7T#KmO-I1kIUJ#Goev$fZU`yH{9h4Z3s+nmk1z zmm2s?t=(h%hW+IwE8SR+!UT}yIGyW7@68aDagC=aBsorJ{)d-+QHgY#4`xuf0E&YJ zlUJ8%svu~j6i-no4i-#SbdNDcki&Le21N)UQ(qWGoICItf_9qn6opKE;h02U4?K#v zEM*;oA_Y(wye1kxZo+F#7p3wPg~H%9p%V}}9hK;)sx^b61kfkgHZsg91Ml-%JC>&? z^a-|&*zfFFh#<$ao(zf>Kv&@ntFNCEg)ZH~b38?%tMG>1_*m>|&C4eko z+`H~rHQqzi)tjd%WC7#e3zwh8TM}-xY}c}}9wiDO$ya;4l`O2#r7M%+DGEux+H*ob zrwsi@U$+luP?7*L4Th!E%NJ`QDA<&zC}bK8H^YMmc&O9%l^%nV1yC<&67P#0hgZ+8 zOynsF^@1j`;f`hAHq`J7uu(>3RZAjr^xrzq4LhNE?Qy}XE^L51E7N)tdop^-nenKM2IqoPTQLO-FAZxnS4 zpN#i>S}cQJ3!vm;8SSJ$^0QEhoOF4LLdnJOjhp^an^1|`dgU@GT>wpnmg>~z)A0FM zBd_rkg(gEwbx+GO=t@3@@5_YdGmclJyQ*_$Mk5xLz`kzL4&kR-CQGa^}qB9alMq@f)pDI=7S2o)NX zkWo?6_+5AB@%Wvt@AWzU_If-!@B4ng-)CIs+?fJs(Nl?Z#6qb(g1*#CQE1Uq*be;H zZUAc0>hN?1WeK1s@Zy^|#aanL#t}S4p(pU-Yoym{ErNQrD`8N!02&YT-rq8O@TIG+ zm!i;knD_25bpU=c^w`u!sqQ{{CxEs>-@9k#K76u=f*ns$Xe;!+t15NYqZYZQ4PnrG z0VF^3=i0*&{^-)JeZf-{lArk#zpgkIK{Z;_81z8^RX~-6WqR)(2&x*!QxvLzDhs_0 z84nS(^P~-fas<#dSP?T#ikXj~>8E&#Lf2qLEV_d-o{ld6c9=oA0;n2#K^q&{C?e=l zF;7vb8hSx%BXe({7M+^t&!9X3%HgOz+{+v%@P`&_C zgI!U3?kFBWEqZf-rzoTbyP_6oJd8$AXz51=eH1{KVQu2apapn&bbT35QRp(PO|;!o zfj^@Q(-oT3-AA7UPzj7Cnp`uqL@lzN!BZ3}fziaJcX@HBMdSQ?FsMKP$xA^#uFb+7 z$WGUIibC>Ikm~-{8VG`_9|jc)pm}h^9)4nryIcL5@DzpS!3}$;_&8o3eY4V>L7xRs zSGZx*A1}v!?FC!hkD?Zh{rs9i-vrPrsAF{L&`cLyx;8~TMWI(v$C%`?ITJx!4gWBxSOCd? zc~<3R-bMu381WQ^x5NF4>A z!>C2Gn-6Eu4*~QR-mp$-L+~SLWDB06&|7%J`q{6?Plni4vl#SK0L_NgmUjw?XVIm* zYRyv=nhmQh{hsItqZaLaw2nc)1W*Cou&HZj<4u2yAM+H23gCwQbEQlT)iDn1cZ@+L zxQ;qo}z!1XSkCrZ~XfOsQJT>1+l-X zRM33w0ys3Iq9N`-w>r*K6w-ztTZvWwh|8hPeM4R)=N<+30}vkDLsM^^lODa zgDM2jJa}F`P*lbDLV_hvQD`1KuU;mMS&g8%3CkGtR{&MOG|QglQ<|Vl*CUaqC{zK{ zEFNS%ry@a}vJ05%Qq826Z@?y|G0rVSQ$2~^u z`HUd{-8@C1-|#x#FzFH=!lz{2W>A#?dIKAdwhgswgrRJnqR<=IaMaj+6mHSZQP~Wt z7C;8@^}cd8Qi`BCqj`!#2JrPh)a}SNbm^Q=S2CzZ0C_@-j`bXbj~m^3hNmdx2`!TS zyo~#1x|OQU>u%Uu0W>>FBF#Cl2=CR}@{gw|G#e^mI=PSBiCSc_U?hX~cKGj|O&4K7 zFhP1B*NrS&$Ws)$2n&LnV(uv;$omoXukm34!upae=_wXAC+ZfaG%p z3y1msL@l~Efu|@WpDWnWwIT_%NOHcIK~4hb4@|(St{drwpqKSh6#4@buswV%@xdnN z%A2>SyN?bEpoQ>dXdmMB13_CVc#1*`;mfeL*jo#=D92E7a&i8Fprw&U3_2`;OkuG}`eb!q1erzg6opJ-u_@=Hw;_UNsaP@ShydCHlY28d zZ^%T@u#P-Mp*=9U*Ych;4ME*D9$?T>0W=8Su*Y2fxgw~fUW!74;0qHKNjtL-rcm$o?+gKm9=$DGHgvsO{>Fq4B6i zVIjHUSiOxwCk2oSJQ;c%{;>u@&1`szLMrfNsBE8wt1P;_IL)9_0!S|s zwq57E#Shi8dMOI&MLNlpzfQ-~{v|`B800R1;>N3||RExND8QxuXfWJI>Ge~mt) zGcM^2IxT?iz&c%4LC_@x`5oga3f+NqI?q^{6N1c&e=+Ec0Fs~7nscgOKLqJ~=P3%w zPioygru9t(Sj z8wQ;dKvl5Vq;)$J5C0!+<|ztQ!D5rens0c{t#|ri2Avl`yWxFQ=JgA&0kwU}Qxw__ z@1xO~!%b0(inRS1bU^^gj~g{;n7JE$MoUKW6our+jeeKq&qdHp&m;z26hIH|!Xrq% zv?+ofp5rMBJ-7=~VN+c3B>92{pBUsNfD~^_q|daT;qpS`hCD?f#oMs39-6QZwJ2x4 zLhHKw=#l_hPyyRU3Vc4H7TsRJQxsZI0XqeBF5N=VhL|1rA6oqP_MWsCwoe}gj-JU^L1d#mDb)&JZmLllFOP->T{Lpoa-MhOWsM)}C z4DuB~M-`N$*7pMOeUx4=MWLe#2W1(n_bDRCW?u}0`~*pUgXxe_BqEN{{ zc*EBI!q2+KIjpX2_R?qdVhcNAqTZ+tOrj~$Qi!g*E-zvK~QCb z;S358KsF%~X|DH<;Rq^i$Ws)u34xiDznexNXt>oZ1_cVBr?5rNO??k8Jg%;nqR>;= zBDZPn2>h%odbo~3K?0~b%mwW~Qih@EM?6KL<}equ{`w$%Uj5VF#~2hWfUMxBXWr+m zFhiH_MjxJ{kQMy&Om(j6OmyiEIR-K4x&YFFM^LB0L<A)lCnn^Am@hvP! zW>AO#@`eS$jh$AtN6`90o}!R9l&-yardoko)Jj^!pilwi2FuX;;VoYvsJvc^LT<1O zt)df9fS`q!mD<+bM_~d;7v}5QJ#fTRorc~#MIl|7ud{x;E)_w34f--DTmZR2ck4)! z$`jP0-3@t)LT=FAdfH@mQv@k2*Jn_K0FwWhPN&V8?-2B@UW!8UAJb{ECQTnfF87x) zC{h63fG+W>A)`7W=*R<}qRe~din*PbfQ0eR247wqJl%U)_rE(lD z6=?F7rzoTZSNLxFbK{o}^5LkM%J|0_+S~N_DrzkW8mfk{V)YKx#(7BR9 zF#^aQekGyQ^41ar^>N`T3faT2BzPG=#9dW4{`hSjx^#Qj zxG^YR0Odb`g2}xLuYvyAVtQhn_02RPGY`@p$ZP2AtY|K*>Du8v^suj<_q0dNb$pHo>381d< zWUvX}t&AXrr94HUuJB~gR&u~IC$Hmt81zs8oreu&|K1zn!U@mYJVl}Nu%WE!iw-%c zMIV&!GU$;2>Q*I@Ua;DLUwm;Lc#1;Z;0Tek(`r^DC}&d+gB}Z@NiaSNx*U*(E?tlv zPf=(RjE~H|ea1@$YhKhaC|LmYgvG>emtUMg(Cl=cqEJs*OuXp&1&_*oHPn^sZrBt7 zBwz47=koa&f({JfDGJFKe47`gCZ79Hm)3Pr&LySYPA zGJ*=f%x6%l0FobRRb%Bn7C~>n@)U*SM_O4<&BbN9a}2jJ=&1nGhjrMW9jowtw8Ds| zD5MYTu&+Li!J~e40h7a-0ptz`K)tVynS-DO zT0BJ|cQ^pbLiTeax^!*ahcM`+0FuuV=Vb@SAn1L)6ouro#1<3t%1|9+ho4g!^r{0o zk8x=Wsb0ftzdxbBD)cY^&;MD+`1Za;dfZ~kKJ-_eujDEESG|RwVHzG?j8DdsF0o;M zRfeGX4vFycYS$ZoMf)!0DGE6xI>~&~_RK~tdVc>fgI)`uF#st)Y4Z`aDE=Ah2%d|t35mz*T^YL6;$eO zx-0>-4oU_7hQIPcP+q+hh1Nl-z%83$_<42mOb-TS3!qmu66uZ^4Zb62fd@}f=v9rA zY{K2N%jh#&TRxsa?*xzrd>N|OrGG%s(h8oUkOq7i`g@1tGhqzO%o+4v0QG@KeP{pg zS_o2~&r=lY1CM(BgOT{u_1<-g{)w-eWd=0f|${L=c(0EwY zy4t7A0JW%J!fOWQ2_VJhO41&C*5SI5;(93xDKu$G_M-qg26Jvf!&l=8Sk*l|MWJIb=k}*xD?Bf{ zC3_fyJ_(>F02QX_Vra@co}$nbfNm*I!cT@Rqh~RwKmhfHsj#1mqt~NLXE=tZDAX6G z!c1ejcSe`)%$aozDilDkphaIkIh;q3jR#Lr=oPeR<>BRa2wGn2!l2IrC?f{8i{6oB zA!ts6WB;F`P)3ZCY+hk4UTjLY2x5^U0kjUfwE>!*E(p54jHf8H4!X6lPY(}0T0BW+ z&=&!;D?lQR`n|#tK}M-OMWI~*PO^_Pyhfo*_e-OQL0<(>1k7FEF>Z(Vcc%^EDGEiv z+;w61-(wN<^@NgY-F@^;03Cx%cSCsuF2;N5%2O0N2AA$`=9~lsxmWgKP_Y0SeO)5m zG|w&PHb%j~P%kgW6pfAIs#XLo!t}u(3A>FVFwWxW*90rvO zpv6#8*8A(d4hZ^CFGZonP*K)+@Cv+Qva_=tgDM2jD43yA{MxDvK?}O@6op2?3{}_6 zDfsC`;|MDG*t}ESeN-)g-oe+q{fEU{5R~_crzrFezTVGE zI^*^3DFy=>R3m_#;SuzD{rQdvnr6sT6mo_~&?421jc|*6jTlrbfaHZS|un9q-RXjx@U&Vv60R}k; zpez`OjZ6!{EBHwZd5S_=Fc4dRF(nbTXhggZgZ2p^e|Q9~x%eCRf|}pqDGK?+BdB0d zyI=(M?tYg+`vs5*d>IaGy@RVP)O+w0g-qbfP%*3&_v>u8_3s>kHEx%)(FafAvxn@`v?> zSCM}XpcWl1pU9d{1<~<1W+PW{WTsjz8AW57t(l&LWxlIr>*3ICp5+mYSX3eJ~}3VB7VRW z&7BlnMAT+5Pf;l12h210#7sghdh0NRLB|D<1sta{#bQJuYLVAIo}!Qi9H;ZmKp$_B zi+w+pK_>)If7mv1D|60o1O~n&ypXkcR+T z4<$FY%QeO$=wrPUh1NsKjnl7g_+*dWGIIu<6+k)wwY)uhE`m%a@)U)10BY3huLk;z zyf4}_$Ws9AhJ|&_Omh_k?eO9$3hjo4_4}$DT@iG&!jnPg1keE(w!F-rwE#i0{_+%s z4#2R*cw`S;jCa{2hC$~AP%x~B8H|iPilB|AJVl{kSP`oo`w&k@cZqz>pbG+MFifK< zHrj%ZR&Ex>QxqBu(`d~S&2aryQ`J8Vx+s9cV8K`I>Sw%j?RUKtg~DLLceCE(U+6Om z-mKiU?uPXeK!c$E>Vol`{pixUY~d*i4TAcsTbrY75!B@EFa})`K$BD9_frn(;Wuo9 zOrE09BECyW`K*Qm9ijZ5w<4}tt$MF<}hQsj`y)F#FUxvkJ)-lLi z0F^+W;ryRBxcEWagQqA|0)2+|X^wbQ-s86mgM0+g5BQ80jj0@nTGadxPf_Rxd`3A7 zjC!L>ch@Y4L01HjmVDC5w-fGe1tgl(c+#2`NbqzM&ej!wF`wr5}mo}!Q@RFsuP4#ta#Dm#_B)!j!| z1<>j^c&N_MH%2Y`S1(1O)p1bAcq9w2{^w`*VUWK7GKUF`7oRNfv+ijYPf^HRK8@DX zcM$qA#7xv@&@}<15&=J6aCI79!_hp#PTP=Ej$43}=}Z3*tF zng;L`g$Bc=+f~vLufukfI4~$s0EI)B_|yoCDD)YX)k{$*9J<6IW8?6%ZsjU31_cQq zKX}&lzkInCK^oRPMIk?U*7mpa1#Y2lI&| z85AafJYm=}y~UnD1P#^aDGGVQux0(`mU^f~saNJOC|m#;!|@cOdIsThwlDbd6ori8 zc#5z4^*12siJ~2YA_UM>=&81C^!x*A(G3YtQD`dkREre%OP!)Wj6Q3G|qZVBVFk(=w z0Ga{ySJKn5cpUmXkf$g#1M074KK|*8AU}x}gKi0+FVJPmDR3WvT69H;rzrFVx=fG% zY{&b6qE{SXP@Di-46}%RQ+}%=sKSz`D6|-65q+i8%}|Rn@A@$4wg8$7a|N@%e;tLO zh%%2tmF2-(^s|0GgU8k*-il?TnzRdMOG`O>~ltYa4?{1S(EB47wwL zhCz$6bGzfc!=LJWQ zgCLVB>OJf3?nD7J7@!v6U+~J~0Arq_&|rXGmFo9IQ1Mk=2Hh1vhS0%?zhZ|Ee!S<; zQxr0U4o1j^_+$h%P*}jAdje=0jM{EKf8P^*8D7^*QD_>B+71+{&qI*A#Wn`r7eJa& z8~^8iOkV^YUB*)s(uCUh`3uW&{grm&X$CzIK(SE#5Zq>G4+PcJOHn8miXSvjt@(ml zRNFO*K}iB=BaA~wbPBzVpuBE8MWKx_4(-@x5}vNyyd#}K4+YS0_>5*uFHAv@;ZB~S z&~W&S_O}~_XQ;M)_{E?{0;mS6{rk6^y%IsmIXp$78mRWy`qB>%;pgbL=~Z_hJr+Rw z@4>fjvjjiACr;ui3hloK=P~+L;nzfqz#$Au7C=94!hm;|As%{U)=N?7$4%%2PO`wg z&B|ue7?dJ_w!=8os;eVj)q38Xrzo@?#-R=8#^$4M-OLp>405Aib9{24$6|3FRVq-a&>(G`sQ?g^oXVk{zB_(g#7u?2{PuOaSS?GW6V&?eJ@2)h?c*kPa+Ed;KuS z>vZ>WKQidK09pob*ePBI@szT1K2K3-8N6W?Y%bwKuW%!U-gWm;ng9xdxq|6lSMboI zU@}ipCPkdZ|!w{aL(1bhCISDv65kY6#N*R8;O+rcI`3QPzW6q$L0%#@7#uw%E!^>ycYj}!6D`7T% z%aHi12)gvhoYuo^Av?{z{ol(VdWD9ed>3PK^X!l1}680uP?-N z*Jb^Aib63kx#!cOAPhr?Vi@#V0Hs5L%JW;JEYW9l|1eKcC>;t^j5epQMo^cc*9>|i zfRtgqyUQul00e!im!gm|taqOYc4&v7UD7`cdMkizVFiD;XR}-c&6>bd6taaC{Phb; z@rWZ z1W;F)I2wH|9v2aHy~k4&>IxG_ef3h8p-Z=O@G%Cx7eEp}*dmv%ggg0@HF%0b5gFXtND=@TtVZ0yDGt4pLDGFVIq2t8n4*{pe*3`Xqn`Krd*I^^pqH zBHtT4MWF%E3(8&OV-g*HB$Xg&SVx|sko=fLm3WVh2-4o^#h^j~ zqz7>373YNL3HLP}7D{@W z;v!E`DD)Pb9KPs(W+00d)%w-luwMmGVWdR5)O>U#f~qAvMWMn-DBzcAcSD!%=E{)_ z`X+$nljJ==+4ZKcq7d7ub~Y3DS&Q5O~R{}i3tdb{Kiuhx(PK2n=J#h(3fHK^k)qE zC4g+9&v3lS_7(_gHiM@qWD9);->5XaZDeq8F@s73Py{?wBN~}cLr}NtJVl`hc&Ikd zn6?JBNYb`N|GN9AQ~*^&d7;5^vuFfm*Go~T8p;dDe8`-Npun{Q8B`{K&cm?9@Iewj zAZW8KPf_SR3|j{FIlmi0w#h~e`YnJYP=EFE-6^~_Ax+^a3Q3^;>O$u)D-fjA--FBlzW%Ch~oWxTU@`35-ScMorbm;l130qA=RR`)IX!8DT45}4Csc;{4ll{X5Dt#PzibAPyAKmiLk44b9{L>8D+ws3Q zE8T~lYS)l5H+1Q`f8;3&-G`oP!OV9y2vQm!#UKX(Xu01A&yGWG`mUv@>Pye5yP&_=n zSA@IZ-<4jzP+?HredH{FjNn9qb_3Sn-)CR7h^HuI1Sb+mW5>NmE!q~-gF!9=C>s`h z-S!MOMW4~+Se~L#HZ1t+elxs`pphz41|1VXSf42@#bvEtDQxwXA4rED-?-2-E zxzn6M#|4n<9f|apsuQlBJ!Q{R6mq@eBzxUc09z6`@I@)U(i;4`Yt_rZk|e`{Vd=%fJp03FEvr#*28@>wlUQRoA7Al-L2#pl)c zv-rcHQvyidQ$25(bOW`h`7)lOki4gQ$2TAieMZghDi5x^VciALcz8{`T0d?cf)-lLK0Ga=Sp6cgWUr~#8mhcpX%zrt_KE;g2 zPlgdQjxp%00J;ftK{+ec@bYL)y%dFR!d%d1+W@@AR~Qh)AWs2w5}w|T8U*H|7Tpiz zDGHs0r?-nzPghjO_^L$;gU)qCb&Sc29rTR$k99!T*+_-|=l`r@Tmma5%daQk30TFB zJVpPiC9q;L!QcM{f_%0WvA^oPp!o-2l6+f~hc{~e;q5#{p#v~U?m48-5d?jIr=(GL z(_Iii=CFc4ca`cu1ZBSGDGHgx3jW=Mu677Y8rPRW7X?t6yjp2&hi(Y+(BmlzmBH|^ z&HN7j2pZqN0kin6+o|GkBLgCABO(c zOHt?*>@m6f$7lkAehyX}Qg_4p37~B-FM70jIDXXs(cmcxZG(AH_ps@k5tMggB!jLB zAOpCMk}Up}p-Y$H%2O0Ffcwbf#HJ_&o&7V1LH+`$FZ4DwTx6*Tx?avx6zU7T%^0&@ z4-lj<-;P1o1keLmUl=}cEnYGCSuaJQ2e7{2^Qod4f^OY&V^Dwqy6P{HN_JiOh+1^v zK2K5Tsz20hD~}w8Af2IM3=xH z4^`;#2BvtZGwqGB}pb!C+1B*?!mE-c! zXS5=Orzn&Ii%l`V-v2>Rf5~J9g$khRDp(LS>u8A}1tp%MP<0i2z5i{&L-^fRRtyRg zKzgv@NMiN#1%jro;wcL0!G@y;0rB{K6m|0egTe)nn!1uSPc;C~#vhI0DGI5nACx)e z#p8L=CLOOZC_(^v!HLHMeLDO_EmBtHDGGVPiO04*ugylEk-FVo21N=WO}LNtw2KTu zQ0L7&MIlYNkA|<^Is-v&X*mpv5jouQH|S`sHa1y4i6|rzQ&RREB8UtzEBaxqjVdJI%H0R#>hXvfO69-GfU@ z7N9>+XY2px{(ExWa!9mZj`^U5=npKam!f~5eJVl`j=(FE_ZruW1 zy48Mz8FWhk*}*7eacqGVx^xq-@)U*aU=-qNc^5Av_E$7!P@Ditg(aQRwBvZW;8(pA zg;HTjXKqRIUeuz+mevfqEr9Z&hih3g3;$s6mX$n3p?v7!PMY--PdM#+agagr0_g8W zsEm&2k0)-ort=hq{$7ObChwLWKrJ#E;m4pm0_ZNRv@}kv!nJk1wRws{cVVR^slx?4 z$FS+-eFh~6pq^0P+wgM`o*_0m#ZwgO3FWL0dosP@u)6!`t^l%wZ(T@#-$m%_eQO3!QOFLyb?ctZUyY!Yh;9tJ zCxC{)z83emd^|Ytj^rr{4S{_vA6Hz#d-#ra&|}bj0i+9)f-c&Y_%qt9!c!E|g-OA7 zW5XV!7RB#c#GnTPXd3KiEiTc*=VpcP<|ztIgZ->Vrrsq88k)a@K}iCrUjlS$Okd*H zOXrU~MWKEP@WXB^)!HJ+&EPD99tt3DD6}*`GyV|zG8{7GDGGT*q2*_{#kUZ&FYpF~ z9tohaP%XH^djkH9%z}7|LSvy?uyEljJc~cDWd?&D3!pgoGPGV>gy&P+x8f-Z#le?B zR=a(3X`vMWJD*q2qaB1ug?w z^B>svMme^h^M$!?N3s6RP_V^z8soQAj-p%H%V?;)TRL`2h@iE`W5QI^);E z5ef)0{m4@k(uL}bNw19Yr3;olW>A^{3WHHd#T7L*1cglGDGG(bC?w2G;UoIiSzRe) z&q@CJ-ZH%CrzdS`D z&lU$|xdXc_LQux+-VAyvfP!G(|AC*j)d=!7;VB9Q!M^`xotofhUE?qrgI)BAH(&gdP30*HmBReQ$lTeuyY=asHiO;?AbC-FyW|CUJO1qeo}!Sv zsC=lw3lr3$bxNiTdM|*ULA}XP&um+C>83W}DGEJKwwLTmhs3mC?x~p2Z@lxL%4v8c-R1!Xxf8f&zA=GAK^~4S;8zw&dVY1fAN+QxqBi z&pO?-9K2-rEAuOZ@&!;EnDmuib>D-acG)~dp*Aq-+akG1SJWaG?PfZ4_t8fIR0#|6 zdqYg#B50ouPf@567UXX=x`M~B4j%m(^hp58&&|5kEdoD+T+Z?oh2-aEX@zM|LM_@+ zV!)sR0Tcyq*tQm>8mL8!N_mPxQSgR!(0lELpog<98B{2MJfIBZa@}_Lkka<*c-XSSmZvCW2pve-!Kq6Sl=S#L zgT4x&#jx7hO>J~B`Z9PV^Av>^!)j;0pg+41^rTNUgT4u%vrtZzqNR>|K^Ob-6ot-0 zIaQGCYlT@Cf&_Dh<$)6EJG+2Zs70<- zJVl}9fl%vx(R?3*LRT9zs9XS*pOi@N-|e#;K^JX!ibCZlon*;xD()g^TBbFFDg@9t zH>j zf+h#}F{n}i`2|U&j|yMyL{O(do}!Rnkdw@Aku_eC9^UK$gZ>GiL$Jm^qv2-UuWQ|$ zrzms?*4SP4`{LrVa?5-MRSBR|@ID%8IU29=-CN016gma(BkS8r_*)l{&~S9!eN-)g za-qA`WBEV)D&LyOQxwXD?v~-^hTBm6UK z6^YcxaMnfif!?W?qJLHO6{tY}7>lJo7BOhA>VGfV{(^Ukx?Ikdd=o33WdY)aPUrRUj%jaJIkPb0w@aR+^)@;g%8_kc$KFp6a{l`ZFM%rAShAs zCWH11pjOa-ww*NJ1YJ6B2~Sa|74)A|buStrNZC4rL5>0_8omrKMwDGeP@~m6MWJZ; zGMHZOjc>Z)$z=>WAb^g&fO*kZHrEkUQ!hoKV=rI}V7Kw;xk?cRLB7mM{VP z@XZ~(*c7h8Qxs|m6R?RMQMVEF?uZtH4ho=jm?ZD36p9Ni!;bP4h0ipEPx7N3!~@C*0^q@iv>?nr~tMw-rW0aF@lyQ z6f(#~09iuah~lquC3NY=CGr%7ETL{By8FZ42#W3@(XG3WjtL-nw|46aRcY=lk&#gK`ZBCfziA!mG`KI0l^(KCz)J8RRa2+QTej zm)Y;}lyZY;o}y5Dm_>a0dJN%&ANMIo!!2W4|@ zC7TfRU~mY7&I_Ow)e`C7x6`_#7P)Eg6opn)JIT&3dgh6s_J>m$bU^^s!l*5z+cO^o zmDfvAs1`U4#uZ@7u&WBj}%F0)u=7P&?@4KN@&s27(?O;3*2VgHHZ= zHI;@4D$RM%AU^@L8E)8F&(c01=uR$AQD`&Vu#2m|;kDGIdescNDuC=h!kpX0NgEL~ zU_4Jz$PSL5zr7~E3u@7|)1CF|ZdiW-WLhPWdaRm=&rm;lhNmcGS_P|GvpSteQ0pI~ z8FWnmHK>$G+bo_k8?~tQPoAPsgG%@+{eb60THpZYSMqR?a5mDqJ|_Amq; zNw8*6gaERF`MR?EJ@M0fS|U$T$O`7`Mt6=GkD%G=hZqzofUdzq)pv0@9`W_*%2O1& z1`pNAD-CcxaOhS)21N;=<1jwb>KTGhcy-vuQxrN5qVUP+`^ANJagpz$Mkib7$poMg||Pn064 z-RXwXy8Gy+0J4BldE2&jxRd|AUW!5%Fe*`nbFLnPVg=A+n6EqAO6M(t?#<&V3O$DTx&vw2_` zWmvxXEQ4+fpcl|0#h}Ll2-4ZYQxtjuEz&xD)(b&8uWm3XUI49y%FL*f3-B9OC4;9Z zv=%Be0}F!tAgEF+gF$x$&}Aroc#~<_9TcLX0b~qMhKv3>ctuR(Cr?qx7@iFGe>JN? zQ1UD-2Hh1v=iv=&*tG-?#KLFu6ot;i8}?<=s7M5j3Yo#6djhB&`V6Kv=|<=?>KV#Y z6e@>4Lqfkoym@_1ds_zG7eL0aeq?Zb6JGB&R^}-R8N>Qf%(Bn8i0IRLX9hhGK%L>m zcdyI}?^h1lz*7|J3@<*NGu!c8kV;AbgOUW$4|p;J6+d2uF5Sm^DGL37CxgdC0Rj_y`}5Z;GLyPrzo@&W+11m9;${e-Hk9AgHi;LBOKK6GQQv&f{Md=ib9TXP{+z2 z>%$P#vaJP!o(Lcv=s*UQx5I^Ax%E;M(t!?S1KXxs5%h2M9tNcfppY_&)OGb?TxF4O z!&4LrDT8O-K=qXfYMFk4K~DwHde~jK@52+k4*R=aibCsQcj4(_<{J={J|d1m&jip| z_%if6`VbGqZff%sg~q~{!DP+#Q?-;Z5kDg0&fxE`ZA6Ga8~i4?pW7uk#dz%HcB_Ro3k|YLSzYDT7`L zAV2te|BBx77kx$xn(!2b{NU?t($oVlg6LXoWY8-CR146=wBV5lk}TsX3e^I1pz7m( z)S_oMPcSG$08Oif&Ph=3MZ7iPaA-s!S<;PdLw`~!SMg=yl?o)P+2cUp-nLSZ{l$87lKk=ePz&F0krrx zoWg5177xVEW$+Y*7XNmVWrp;{H7CI%n(5cwN0|cX0IW@18d!!mW=3oC6on4J+QhrI zzb#OU2At~8pezB@7Djw)7p%lJC)M>*6lx12K9d2pcy`OK#DGEB0%#JfP2|qcia}p* z!&082&?H!!FdTlW6Kav|G)o4(6F~Yf&*18qG!{WKr}Gqr^kJT1<)%4S2ug|C&!G1L z=wG-*TBR75i=gyqo}$pda435lQ(J|gmmRzr^g#ewz*Ogk!QWmZC|HH3C}aUsoxci= z@D$C7O$iLj5kT^R*s9J}B?z*#<0%Ts2V$+?hD||rj2oZ7XHc#xI>+(A9?NmvYy%=t z^NTb2fBw%|#8Q}mb-FX61Ny7NvUrOARi!Wi+tmK(Ap|X(RL%aXJVEna;HZ{^nyUCM zIoW`xDC7c1wPNb z`Y3>sUPB$@Ik!&e(q%|^ib6@R;U^y-dCW#F(y=yU&?f=Z0cNQ7o=M0-(A?ELMWGHb zLlqDZj8}(eJ=wyb0s(XjMg(gwuEUiu{Ze^~LbqTNIoJ+ zp0uDjx^(Uytqkh!qpt$!6C8p0{KbJ<1X-QsDGGgpBQOmse&BIvV#QzveG@=`VO7gw z<(d=(ef-N)6#5IRTAdF6!;hdg=Ee*v7C>ziV27ZCSvi8f)JsvQZGw|5QsoVvMeKgl znnB+M&{dcuzYzLUf}mgZQWUxhljQlsOz?f=(&-R`eh46Y=mai5r?3V=>pJrkh3ugd z810pSUtR+p{226806n_}z0IDbS_o>okEbZ~?3R3l=O zy8Eb90NKDY^q@S~!Kg*G^->hFfo16VsX?RBrF&K0jX`As=ou8_rJOq%h@i6-JVl{r zP>i>`+GZny`kCl4=(hlB_zg}_`6t2KbNX(8BF! z8B`&F7DvJcSRb8A1ZnQzDGDu)bdt4SXPb(k+gUdl^j83lfc=N&zH4)N@eS1H4Juax70#NPehih@rkef`)mPG3cKF>IjF=Y^a*q z3|+d$=Xi=j9pUhqtg2b_5%jX6ol)HlTP1+XV0d`k*CiZ5&;Ifhh00)f*xPw;27<27 z*J4n$02&P6x|3~37a-`&0-mDKVEER}er$IbK_la5FsMcV72Jp469^b^7(rd`@Dzm# zpg^VdUs{1}21oivOQxvj+^RMn)i^iqY zR1=Rp0a=t2hV7eLd(;GsHN=PH6O+~6q+ zO$&qL=O_F+g+8M#>XONI_mQIriiPvJr!<+0phaDI>OY`ZI75BY#2f@Y+S!{y2LzD( z=WV_BxV=QsNqe56ko@OuJ8wHW6+wlWG6p#bpgedom=#OjASfw|rzn&MPX@&cwRn;| zdb|aL4ho>gaKqZ9WGqKen+ZHcp~i5-Hnu#0zjfN)dl+;`0Ih>1PQ^86q7k&uho>mC z4wg7)ZR&=f-v1h3WYA#&Q~+bx1)A%RAgElCrzlhaW7w|+_ax{uQd}L!pd$h(4=&w_ z71f~_s+Xcr9$dPtzWrJu$RZ_^K}Q8pCwSIXmK01!P|qhkMWIgctTRd3`w>AZ8h;t& zEP&d)gkwOuIpcj^zw4za)aE5jz@D7@6hW4*9jDaYur2~<)m<3Fj%ikmT4d?QQxsZt z*GV?h)u0AJxBqA}=$HU1gNN$U*NyNVqVwfEMWHfysNQhji$9}db4(d@TmU6Pl|`za z;|tUxo4Gtip+u;%@Y%Qzw@5!~BZE!|Aahs{{2ngFGriW)JVhaMSP<;j%oLBxzjkzG zkgEVng&Q{df`u+>(FawYqEIT_usXZ`XEr`&X9$Db1ki>g_>4L(#=W2;_B=(Q4N0)7 zXaJZRq-F@UPfL23I!jx&Hxa7v~2v1RHHPj@`n`NVcE?r-kes(r2;E8J}~IK0IGl~nt$%bxJ6@!@DznAV2Z}} zju$>Cp!`HNgDwc5M3|y!8n6f-hF0RrQxr;sDH@G6?QjRObw%f?bvNuq0i<_RBAry& z0*@yC)JsuF?+FY~aY2`E!62TZ&;;n$y-4ae6hR?J ze=^8d0LgzzrmEEud@A)`XP%;v{Fh{`k|JXfG_0uAw7UDqPXJB)2v3G}SDT_1DSzQ9 z3QdHP8_)Q1JfU&gK!ZV71<(J-WtI6xssQXa&0~j1Uymcr}Br380}+J)80;LjysZ6?uw6L!o+B-!82ZLB6XF zF(^O)eS2_X3y3-A9_M!)e3wVF z5M;H4rzrFY-bY<0Hp9b~ve>f>iV#5ZGcyz8=Gh=9;TBI(NPcFfQ<&~&bm`W1zR93S z0i+J~3<^ab+8{{Qg{LT_4)qKT?_U0kpvzk`7!)Odg5lC>*R0!&piSF&ibBC~=^9^e ziHpOP-;^;ZS^zzSC&Mp|ydMavt(T(ELwGW5QuQuIE&4pR{fxRB_J#o3au3eINU!>Y zp!?%^ib7lN!O0$bwD7QHsh1XmZVI46SnzcWeQbvy-Ag=0p+Z>jHCk`K8?|WfzZndQ z5kT_eDK2%M;)|deRXjx@`SBD6F8%OA)tI@q42l&%@{>J+BV$St)OsFIQAmEWNB^WS z{PfO^a%Rvi0VLmqZfjL6MbNuwo}!R^6ME0h9q?!(v_l|+;slTyK*1GHqY$)Tg{LT_ z2GG%%Civ3nY<|q3+XCnVypL9FzKg48J8$7B3Vne0(aTXri_n)L@ogc4;swwRm>BHt z?0g5c$SsqnD0BlR24x9p_%&fLRx-2hKDr}-oaHGB?S}W!qo6k!T2dopP@(|3c3UFd5-h<}G{&_&MWJiA zon+mn&(cGmk<(%e2Hh1vp)l2{W|NG|3>KU76oo=zs`H8Z?e+*-9>0e{_XN-rn4(GQ z*$waCm)_wi3O#`-nm*Ge=%G4BxAqqqbYBhCF+SZhUvHyxi%aOQO6$V^^MBScZj~?1 ze9px4IlI+)ivCqwq0983aL{A~dGC#5f7Js)^J8Hl<9R|cKB{Gn15Z&X78Wv8GNSRE z+rXSm1|)fdlD%?yib5ygG2(1LA8!}kYWSBy4+YRmnEmirItf=Rxf<~l zgFA5T$8{)-{yzFD>CGg?`#&7j8u zXj7F$x@PP%{5m$N;VBAjs)Cw?%`#I2EnIBMpkx8G7CxiBlXCG7dCWHFDGIHH&q%w| zaa_lkm#~pRDFVnJHts2vsP9CV?m;3?QOF-Q?!6y-H4(Mwp0+E4o(Q0i_7ds08$IyK zW4I1aQRt&R{I2`^7k?16&y2HlqK#+5FDubR1 zptG=$(Y9r8B?Q^j@Dzp4!a~MO^*&k%idguKLC*w`CKTask92`Oxw976g^nOHoJyZ^`?ET@IiY&F(XRL1_ZW6N7CriC;>qo4)7F(_QKOUOEGIKf@bGgGALaD zdBSIu@H`v$pSAOOib9_78Kuvg{T4ydx{eHbDS!q+J@Apc3Qtjs?u_Lr3JrpK;H9$Z zy%DtSyf=eh37~YCMLgGND4r{rcY&uUln%3qF(K2E7(QP2de{Q@pn?x^z*?c#1+z;0>Elwgo?ev~O24=#2pK zgHGV#ltZ|-XLvkMQOFNEfnTCVW}y}tsC6-^yJ6o7p!NV27OCSxFKH*9qELH)0?(G? zN6_01V;GbvfZqI-NZsB~!u{tb8+nRCZ~i*Ts>hdhLM@u~*o;A00!S6MxmBH2P(YuN zT{2HmNENoZHO+s37cw^W-^!qD0n{A2wL_Y_HAax(0G^^ybLiIY)Ou}+T9mxcok8ye z(4rEFbl=xbhY)miKTlC;QHhhRboZkL2#Wm>!Jzj7C>BQ6+b>PQB{!#Yc#1-?FtXmZ zFuNgwMvi~Mpbr9QD6G>dCg^rWEmD*66orPuI$cn)K7Q7Po%zY29061S-?~>x^G+cs z$AhORQ~=+)Rhh?BQH$Etv@)%`k8%Z2LcB!UDb!>Qf)>^C6onGvVIkwyV%%q#vqFPG zc>?G?jM`2P8G{Q{URm-Kh2F!c?Sfe|1Jt6UPmCFqFM$3;!zrLD^Kn^}Ln=>E=ufnh ztXF(!41ydrS2O6N0D20&pwu3wxOB~EC{I!7DfEK!mt^3HBf}$y81zX1Swiu{tWQaJ zUR3QUPf^GciXWOBb;C8=zl;1BR3L!P!PmQS#0-7(WhnTh7Z=0pt$l9DmzwwLs9q#ymwKcPQsLR51siH}Za2cLseCK#DL! z)nt$+?&OCr=P3#)!VFcX%5O(diw-}~W6)Otc2_@f3v|U|#e}g#w<9 zAKP;=gT4u%9kAZ*PtZ7C{Z(o@LN?0h9x;a+9SxcvsZZdMOI!z^goZ`zgGt6{>ZUK|cggDZD0T znzh1T?<2!_ibAFEn#iiHz)yzLr!yGzQvfZ5+4%f!HuyEM_6$!^Xd%qTSLHpx=Oo-J zD`U_v0rVY46HrruM-xYW^Av@?!)PMQtCPBZ>xohor z)S{h_of-5;06mBCk?oO<^AO~b%u^J44&$SshY>mmQqc-zP`Ln-kIGeLcDQrWY8X#Z zNIoh*sB<#}L7FEYGpIrUb%o*wJ!e(C2fx)Ro}y4!D1L}`eTz?iaQIospuYm>4wQ5B zb!k+HE}hjco}$nlDCfxc4(o?n^v75-ukJpo6hJvY;P-Iv9>)_J8B=+RLODNRiF4X6 zygIzcr#FNC381V0U9eRsiy!S_P>ldO z2h-6XFO9%OM57<`6ot;gboANZrugf3?KJQ|*Cu{L$;}Rfm95aFQ*+`e3jKzXo7}Nq+|Z>}dv1aEc^d5>DOcm_{V$O9Gxzl0m+At=ee zqgmYzyI%mcE`t-joK!|3=MhjGy6$3x^%mqZDf#>06GP; z#EDV+P9SLRbDpBmDVQa;a$ku@6FY~vGU%WH@`9e~3f(t&OM|ps-A6|S zko7H?VE=V42(@VIHJ+l7^)2~NX0*b~(5Wp3FvwW|nZVHU{Qfg|Vlbj5Pf^GOhL#5} zso||XP1hST$VC7J0Tg;hQ5&@=uU?8mK>(=-9lwSy-S-qr1|1VX@)P*K-rb&sptvVI zMIreK{681e;@=@E9_+}V;{r%NlmBs1#UBK98NyQ(lF#H1?fFIxwP=)^H-k4bw;1oqh$QH5-p-^O{?47-eL@1P%t9lFTGAjbm4>1r z$|yogNJ6Ple%HJ6c>GS+_xc=vdp(|=`+biyu5;ff5Y)c8;_Ai=_PhW({$56V+3GHM z??Yje6ornzx7B$RF%-Y4Rn1U~L9PO*9SkiSyv|mk7PZ~LQxs|kL(78KllGz(DAWR$b@Xa`;#XSDQr^cPcLAgbkMdhv_iLdR^;h933Ms;) z{I_$SHEPi+8&?K-2p~xndapp~dIX)d*tg*fx*&iOCBqiED>o4& z_l~D1lnBF?D&yC8(AD%XGE0d5S`|Fkh$tbPPVdPhYNNkf#8; z5AzJyzXxANEz6eb4aZ+=!s(4yOO8RR8^#=}m5iNg(X zziwC=c;j?z;g&{W_R1=!yW6>~_EBtc*MP)lE_qlI(UjEX%;d{~|Mc23-|E zi=m3^LlupFvxky=R$Hb$>*^dx(tb(T~G!`Bc-YUw;s6}cEl^EnJ zfX+ZKXk4d>(-G8v5l>O*4D^C#PSnKpbw0NzGssT>xx<9Ugg*Q5I;>R~Pf^GnCNwfC zzYIflj5_W08RXv+)iGwb&7D;D$>1+){<7};KmX@kLD)UG7HlryJtpIO@D%;4!tUAX z>@0qdt1K>C9Abaf4MFpb;STI?xdKldne5{!3K_#4I73by?>@81xxk>C0!RV&cb}d< z6R$6=eW*nf@2W8Qeb=^)5Mo~J1Eu>n3Z?k0yf zz@9MK!Jtq9)EnNqwlvPu0YMu~d5S{4;jL?8WKe_vs)5tHaKK1)1TFZ?QxvL!(>te`xd(zg z=gAl}UPqAvs1AOk3Wo={de&t=Pf@52exn(I_wXe7wBUgZx+8!-!Bg_zjNEF}q6r~9 zMWIjdl#DOStwJq&)NbVN|hBc7sA@k?7B^})08 zW1@7B7lWb&P>TxK#o6Tu-cJ`Zn5QVzq5_^rJreO_!pSC9+8hAv|@a`3ZVg-;3Tt^zuQJnwNlQz8#LoHep$Ws)mON1rP zJ9T)va$`FU2Hh7x`fwfTb}z&u0`>MhMIn8-j=WAr97ioWx^*do5(H2X3=cD39BPhQ zWMINm6bgdjVcyT?c-sF&@-7A?3Ls-Ry`S_OD2JfsDLh3XV>rEU`@CC#TC`<=3xkpb z&>HA7cy{c8>wz~9rzmtC2D}Tu zE%ZguAGJ8^@nTE!x?Zrzm6zCqsRfFTRe-x6EYFLjkn5SVr42?wQu2lCqVZud}&%A2Gp6agJbr~EnV(8-|P-4V<`wSlTkEmykN5hkW#sf_Pn8G zxLRqdEKgBLsoYlQec30xF>|}I7K3sG&_tNri@9Zo@7@-hd5S_4VRG+7cu+gkq7`W? z8T3Q|Rm1A=wlM=L(4iarkf$hA4XeXrrR{$rNM+zY2IUH%mQW8o@~18S{-Dwzo}y4o zs0Uusf2tRP##*^D=&1nu36JuqNR?F7qQ1v?ib6l(Q9e_8%UiozN6OaMtfIk%|o zTwJXb^p>Y6B>Cjr-%;6k(#d#AE`#y}P&)Lzm#xjh<%N^Ad5S{m(Dz>6xf34#M|xK> zC|>~S!5DVQ*#3CB($j~hD5M8t*kp~zU(lh;ZzjK?@j7}gfbwC*q@mNf;q~ro>v)PncZ+OwWWK({qw80yVa6iR>#)_$d4Hu{aU z4%#!QKmg5zsm=?RXPiP%w?jNdp_wq%Y3~?>zX&tnl^=r&1yBr}-m+~jjYm++0-mB! z44mE`+gfZ#Eeg^|VbE&V=_vd4Q7uAcSK$&hLPX6*_d+t9gn-pW$8?`D}nS zYLTv@Dzn+!<_1)*TJh0WEVe~LGJ{RFT6)Cu}Enzf=ur7 z6oq`@J#qyt)KF?Dn@rP=_F%qLAeCK(nrSmLTY8 z`$Y`;EP$p!-AI*+Bkl#wmFFo6O@X?Rs9+uZDDQ8!gF#;eP~vkL?N@gv&qpn4YtB;? zN_-Bdx2v=pYLRoUBZJBXkTXn2uiGw%M|=mL@)U)fVLIAaerOVcI;sUS=&Jx40vBwE zRst^cdfOyLp&@X=N}todjG*_<=?tn6Kxd%J;%cJBX>{n~PV*Fn&OnvLr@5-V5v2a@ zBZIyPp!G1pZr5Q#3WDmJq$so=CfM6u)vZTR-F%r%jn`470CI&rM70Udc;B(YLY|_K zE9@akdJ~HakLAM#GU&Si8UqEBx?8J;pccJplA_QUD45(jcFuj&qUIfR8T3N{eV6o9 zHEx(8D6dJ1Lf@gMdU%B&t{Yju!;nEg1<)0k(5O6`gl|EbW;{iqD=?vPMX4h`8Tw~g zGN?)bje_EbtvJ@`(1kmtUc)R;6 zO?*?mUd>Y!8XgO4(ItJAWpfy!c;jA3g3LUy53weq{YpP%!c1~wpL}V7+ z&A9P`trI|V;DUYrU`~Gotq9>M3eABFw({-Ee&{-KX`{iQ-vY=3uA@)gH$F$u=C(XV zArH8Yk_>b31>1AiQU?7IKu+)*l{@^oi=b}1d5S_#@Efh@9$1Q6(#m$!Z@+cD#fjzmq&gp)@Hwoyg*}as=gm%w*6%0W=5hbvHh&UXGylpLmKw zbKqV#YVB+cO`BiNAgf;gy;(^MZb9`cHmM?L^a7rukQUs6EcFBFR=Y8c(VK?bm*M(HZjOr0Ih<4-K`hHx})D{ zOFmCgXchG9hE=rKh@c4*tr=t^fL_GF{uJkT`w%pD5>HX+MU1UZNnaiOguQ*mhe5Ui zXlMq!a3d=YAG%H6JVl|Q8St5{V>-A(W1mzagX{#*jYl%tnrFgsi*%)Vib6LY+3F~6 z%&kL*ZrFw*2H6XsJm?ainDuD~I&|HPc#1-K&?TOdIsQ3M=2|AX}n-h_Cj@x>%IFpax_K8{-KuR0}Y{*j;1e7hFW=s0(G;wcIp`3YaWTt9m# zf;xU%$sk7o^c@~vBP`ccBB-=UibCJv;dOu84!qO$^xS<6auPs4p~~XVrD>fIv~eC! zQRpXBSy*UR#-SEv-E?J;vjDP&o{Z__Bzy<<4&W&YSwl}|q`3kf9*&laV9;p+)C2Co zX+5MvQHut*;VBCBfIIN`FTS`FxOrPHgU$#b$(!+h%c?g+knwh&qLAdxc<)O5WznJg zn^DOi7XcIjhb}5K1y4#yJ?1G2MZlr+9aBCKLC1#5Z*9Ej&I%wsnEfz1U+9Edv|<=f zQAiJFKMLL};Uc0OdsPOV6F`#B)9V^n+aV~$fu|@W`8<7n{AqkL7*)(=(0Kv09A4t& zTex*7g4Dn96or<S zo}$nJxL~I_{lL>K(`af* z(@~4&O-4H-C zBwxx7vB%Z3^6PnuLNj2U&ic>I7w9*d9GlLdn*!)Myqt6Wpu68ti|U%BD0CfO&bj#M zuaO99*`tg>0Rl)GI{9_F&GAfcMUxbTq@k0aIJO7wGuR)L+17X+1qz_Sum@kGeD8GB zqFsl0ib8{75B`9xl4R5(=@$bT6eNK9!k))-*WTenm)|5sp}w%^amDA^D-pDAgf4@E z1<+<`S#6^=Z`Pp}jULHU6xu9prxT=m125WcbTnj8hya=or}u&{_uUaR)rqGlG#^gy zDf!Bps6`)2Eg5u60J%ZcUs$ucLkRN!$Ws(@gQ~y5RSpIS+BnUNL7@W34wlahU&t>- z(7fq9MIk#_KC^sl_ZC6P{;>?YEr29zshMpaUPsWy8$3lJ$y#cbl_Qne~LIAylqSlf}KldQ0Wg1UW=p_`jw%-!|9<^xf5Df-J3Lq=Esg^%>?2Vvq z$~;9OE4Zn4f71yM#FA{6GU$!~DuOYr%#lF+n7D4oQxqzKG3*n?0KDzm=+iC+-4#If zFfSUA9$JQ4bmTKnQK%l~MYp^;-T@st)mbhKiV{HcVQFUCOKbcawV%yX6q*l9Gi{?M zDXb^0T@9musg`f~Q zo}$no*cvZ&|6L4fk->&?2E_;n*_iHs0JW%PKNSYW381sEI$Qz;1_&x`lA_RASRI~p z@Ft#gias)vLGc1;5_C?E_uX?7LDolkib9j1bF%Av%mCCPmFJrnbYB4Ns)3K`bhN>x zYep}4ibA_;pol2*GakcckF#b_f&dx=os-j_((vT1-*}#)&=}~P%)7T8Z~7~~;KQIq z0b~HRrbh-(!G|vJB2Q7s0BTLEN-vy1hwk>*1O_Dupd<3K+9nE8ztL}WxPqrBbVS}x zr)KjZBLrnHDPqtA0Tg;4W*|3gABLckr94HU(EGMJ(+{@5Re$Yb>lu_RfD|I(3H!_t zk0#p0@f3v=B5ieIy2{|vwSheq%^EM*6ajP^ZmNUstaL^#`qd;wq0?|v-ML`SZ*=Ht zcWE&wRRDc}TTpb!75p2G-osNA`T(~eS=+PM5wzveN(Q9~Ag#YJ@BMl;eoSaR<|ztk z{e^+pw!gU0YyQxE40gsQ*Jqi1`f7U?;`1W;?}d%Mf7#(jpoaXdw#*3kEE{c$FK&C8xk_6&L=fQpjf3Hv?=ukqP*s@gIB~}RIDGK$2F0sNbD_?Zz z{N<%~HeN@00_Yf=b@kJ?RiGB_?Z8tMItFLm*_o|e5Oim|5`*#ukmOZ7Yr_oiVxspB zo}!TCRXlPRuGk|;HD@w|o(rHSFg}v+v_2fQsHRDZLQi0P^l0@1{37Bxs`?CiA%J>9 zPt`z9?F52)kKic^^@N`4gmcYs6?)5)hZyuy01bm_v?PNoxVC54DW0OxFqlR&8KYH= z4xL8D1qQtmK>c8RWU{Os-btrC)mkh5MCg9-#t0L+~Hja@Mn9Xf-> zJVl`Zm^taIRD-LPI)~&ls89ff!z^*^+=!6~dfy~Pp>UWbp8dDb5FNVu&eaThEr7m8 z%V>9*KIIF7zI5R!3Vn@+QtF)cYZ3J8u)?m!3-*lw8WIj~FEqTPjG(+DJVl`);qYnB zOv`Qv`twSSK}7;68;bFEO}+jPLGpz>MWJjc##6O7!H@Fu+KU+URsgMowelWb>A1GX ze=1K=XdSGTcg@LZi&|8BVF!cW37`YNWVFw$YoUc&lyi}%D0JYLtg-*eo>fbrtKcN4PSGM9`XUWeh42Ko?>ARmJ4> z=MiMrou?>t5w>6D1wA;7pn_d8yBn{gQUNp?-bi4bY7v8=;N3h$q0#V0f@|t?Y7ulI zV<3Y*3ZT?_8Esc{&G`sg_=u+{lv)o<21~b9AV^tNmqBF$Xk8{u-o~Z0M^Lj7JVl{( znXq}bw-a7Wj5uM)picrQ0M;hj7R%#$hI`gLMWFy#oA})+wj8x+P_ZS0J`12S*mvA- zjW#Yb=>CDHC{zafjtv*D?17-D>0S)_B7pWm!DNiXG<;Kan88yN+6M)bPHs8R5cJMF zmO!h;qisE)C3{VN82#dVB}1AHdS9dvGi{;EkW z3;zE<|7RWJG4o?t!Hf=h@icj zc#1-!VF=%#G^ZXxPh-37X}r3>2_RjVl=x^{im!$6IG&=AE=)=|gbW&rpoKj(7*r{M z)L}$Wam4};rY7{_DGI5>h~VTJ8$5(xaCj+$z6+qWS@06n3Zo2k=roS-6ouAi+3M^b zAJiJP$o=Im2K^8~?O|lCp&ElxoI$??kS2_*i(AjYw~-G`QWVmJk@b%E z&2de_@f5kejn`3)01AB|qaB!Y&Ih&VQz}nUDD(kzFuoPyacJK`Dh#R>K#~&BMWZT9 z5cIuCib9eS&_T~4gVD*LZ8eiYbpq%aEJK?`C>=-8>|;Ddp=YoR-Bsx%ehrA*>rD*$ zEr2A~k;2eYKLq)|;VBA9uA^Oh{qTCX=TvJ3{SiPTp%Z9x>=phR#UUM@qR>d_1io!> zz#I1xuKFU)2&iO`odA0L3AVb@C&hAdA zMY5Y!8RQ^`oh|qcKAJVt25>@=%fI;16!!JKRtp+Z8IM6 z6ou}<7OM5xxuK{<&6Q0UbV>k8NV%!iy#?$o}!R13=b34T(_bYr4{-y$VmX*kdf7%v8=l;g3iC@DGJ??vD0a-zDyrM zc^WAUauz^u;H>MewC@^%0yKGwLT})#(=FLB4ncwEiWziT042b%W!TBJ`0j0go~J03 z0K=As{g>7u=2AL00qMH z=((ytz6FIW;3*0P!t+RJ?_9iXD}b11e#rp6b67 zwWy7}K7-B)pgpi?Yj$$^Y6R6aNl|DIEZVL}^28ro`MdiNgU$<}Z7`Y`@6!R-_7v~o zDGF_a(Ztc|ZL3g=911Tm$W;KPhQVXv=5;(tZv2|3D3lsztMl-EFfLuom=MJvHvuGh zLD2GzGbW%z7dDZnC?t77Q0TZ_cvWlW<$MOY3!ro;xhdOgR)(N~UOYvibSSwgvzX+E z4&4XoY6f`-pa583$lrSqpLO?I@DzmtV11#ZYe+nTTIeerXuM!A2%u9iN&aVeCVs2H zyCx|Loq|d7fwEn2q1S0?hEzzNy70pu=S^=wC1HJulO@e>7MGW#3 zKu_V$P!=_GCV~!j=P3$3g*!usnl@L^p^MqGgF%-BkYsiEz&jOuue-LFrzj*@9qwY( z4^N|AdgaKV%K~Ur0+dBPDCvM&bfSQ#C^RYo>P8F>zD6yw*9v5imjEh*TTpMArw0+V zYywYFs0?mF(0GbM8SliR%jv%P{Ri2{IR2akh%u`!~T69jTj6qig z&@+Jg&ROGorHd!L+l?T)FRJKJVl|C(65W! z8Snx@9a07|$VUK0!t@bvpY|{SYKK zoTn%>0@ms3lY{Y1^{kU2gRTpp-7p(JP0#BQf_^&l6oq!fY`mx1o&Bgq=f7Gq$X5V) zz@*cGgtL|iI$pt36!L&cCri!Y%@AZZ(~CiV0>~BS88YsVlSa^tSv*A{SD0rwIyV$g z-Y&Zy%OHONbQ@|C^eVq?L(pJfo}$ogs7bKj)VmV3=!{eWgKh|*B)F-*?Bt3|sg0z0 zib6?nQ;lvXbs0hNo9Y;JQvgkXRjmZMS$Ix0+?b~*Gyzt%T0QSO3qe8Y-3~QgumJ*S z1>6}_jeeVpOqR`r3?xZKn`$c=(l@iH98qOpXMnFIl!Ia^u|5-3H$xiE(Qe)pk**cQ<{?=f}rHj zJVl{pFhz4}!7%)8r62_QD_a+Y`+-x0k_CNE{j2-0!Z>n?j2Y2@gm6RHatZk$tSrtW$w^J zE!wrIoI$q*P&NEU{wHe3pcc(H<|ztQ!*BGjzqSv8`XtL8ZoH1d1kibybc*x!vqey| z6rQ5cd6;y%qCQpuL0wf;7!)pm6yQ4ArI)c3L7$qWD5L<_(Q~Vr7^BVDG_=TV%Pf_R>yeMGE$kZhC4xgQ5k{Kp0I7kITD=py_EmMWKN(n%Lkoy8yMQ zYN+Co#tSw^0J+1VJG*GuW&|}G&Qla}heKD_JWLru@u#&I6f1zvCd+6q9(@&WvhY8{ zQxrOzY^zf}S{u)M@BOusL2&|T);$?*^H2ZtMVJFMJVl{d_iS}mopM-)T4b?kAA{lr z&>ont`&P9AS1Y;b@f3yjzrO)IY?PLH7mFC#VPZ3k;G&zfq4{JVl{TP!Ifk zuy-wjth+`qC_w;~-G;ZWX)EHgsAb)Fib7?#ZFOoC3L+3RdQUEc5(SW?!a06y3NBsi zzL%#cB&l#-*P7Ow z$sJYrF)^`rE`w48Pyj6JY${%ir=#WTc#1*+u&iS~arjNtBHwu?3`!M1ZQIIfyJq(3 zg$~`$`8-9Twr%Znj^=L0eU_6$lBKy6^sskZs%as(9x@DzpGz@*b*Z?nOuMOCu? z40MLd5S{eci>*9wSN+7QCBr7%f{>IkpOCdd4`vk ziMSM`x=D&c4KU9zeV#)Ig5IA}V$fp&bQubeK!PRmEoqteL? z$`n9<04m-2$q}6l7eDe8h5i5}SMwZ~8Td@qXHb>^S_jaOC2#RfHB^VED6|fsaFwHn zP>W_=KE$AG0n{BJ_dAaLv9Y*XDWrj?=wH

    P9YhC@4eC|F=4d{Z&r{&Ci6sna`E2ThI@h zzJ{kLlnH&a3$2^aL6Bc`K7(=vkR!a1;ggKZ9|T>E;VB9^!V4L!zAeK=L?3%pGw7)R z@_@C}pd5om2&(AGQxx)mwbakeyoykZG!83RHD2A%1kn9#_@biwC;aX2u19!^Lie+6 zby}`jErX!)Tr~#e37~JyWVJ8u2&zRbig?OX6#CZ8PABV(uQr0z#w}t{z5p_X$FcYB z<#>-t&+$A(Ayar9KR!6R7lMX*?O@Pz0TdJqAJN-+6;HGLX_BH)P%P|uJXVDZy`KMe zWY7x%G#6$Ojekd6L5D8n4^L5OF3cjPHUEqsUeDG9GU%lMIsw({*xdax#~QDrLII=#9gJoAnWNF6dvJiKD5L@%45z*Q zH=q{ndN+_kuLaQAa2ajYzMCo$wK`x*~!SwRIWvMgRrCZ2a(e`M(G{ zHI=6*6acgF2W$$rAn5D`Lk1NIAd5QqV#ujyc$?d{i#$aki#l5!_uM5z5wxbpl0k0; z&>JWL_0Ke!gIYAImZvE621-C@7VX8m&%UhmV$eGQGyx`#7Ikcqh@fYyc#1+3VB%=S zLwOz4qE&IR40`^3%@5#$ulQxs|s_qqvx4_hK=bI$?>6$_v)a0}`(^}Y;( zruX703Uz^7(64K)@l0>h-Z}<-5I}LT59p|V@?`{_wcsfV#lb$Hju%?~LM<}O>vp{H zf-Mn1GvEcPeOl<^-)KTUPf=(Fyg+sFm`Y3=TA4i_?*JeM0I-KPx3Q0bWJow#)0Mw%9pLa2+OaR@5cUfG% zHYN>0ubZSObQj)b(f@cBo}qG_F@f3w*VTtqjI=h_+TC#5@ zgT4u%l6n{(#(zADS~PG!Pf@6(9$we~GCdhV-Samws8Rqa!f$k6CKwm+%RJ{P3Ms;G z)Z#~<3IwgtvS!eC0Tc$0iPx2yJ5h_YC-4-7!r(EnZNq0>1j%0WVbBi&)B`HYP8E;G zMMU44q$tz_D$1PGhIU2JdFez3{S-hG;HK)aTDBgwXiE#8qR<4msRqaemmnx^eG!AI z1W*n<_~H)kACI8!8+eLBIq=|HAKm99f@UVwGpJesD6!+s%Xa4Anws5Q*=mL@;KFS-3UZ6AZ` z1dwC{?3O3H@wSoPGkA(Zk`1tW%kSX}Rx8AnLB9pi*;}v|Ptym#1yuPKPf_UXE%*k? znvQtRExbzvgZ>Dht#@IQg>tbII&|J$d5S_???RuUY}I%48}&Sp%b>pkNb-t_sS2&| zJ2QtJGyOt$==AyQR5ls6~%r>=|@i07>4Hux|A{JiApG%TpARyeA>@+G9M^ zdsoq)K_>)|J3#kt4LpHbbi6lDQOF&j?Ij)Y(DLx!6b4xfpgAyDUb8YU0701+JVl{7 zFjy}9W{Tg=Veq_|K{f(N8cJ?Vj)bZsNcja%QAip}Zmw+!|ABs^E@Px@8?Pf<0rU#C zBzT_pUW}k$O;Qwk1zQsIoh)(?bp4DHgX{#*zY-bkDHnI(79DosDGL27vDG=#>pZ^K zt*)5NAbSDy88(#ty0^Fpwdl$>o}$oa*ig2oyde^`=-g6$1~~|zj!<}f&E&Neg8D7z zDGGIj!s96O-u?)Bc>fTCP70s~Um5LZ3yW}{AuxfbDAeE!pKk0_h*yUv^uEZTQvzuD zd)O6KD32!wJN4ly3N3$ctF!u2IbNGcwuoYoqX4pkVT*RfzL)6GdFjD{cUI0bH_{b@C8?IyYUe8k$iiGjeG|LNkBaCW7I)hvV z(01rE?AHy(hpt&7Pf=()^cmKVlfw&DCkK=<$V~*n+C<^ps}<-svLDD({{g|;L}Jia zTxAhyFJs?$9k~l2=R_IpN4MH7LQt>+Pf^G@5oS)*CfT9isP^MP26+ge=g?CwcH<6osBcPj&aKG(5CyrlZTC3j*jW)L;GU(sK=JQErnIg|0&V)z#I3si;LY9)=9M zD1hqW7Bp;diV}j7FYpwF>fshNs@e$8skW=MWRRx-Y75_5{b;`vzai&KlN5#8!nal( z>!UZK7FjIwV$dZ4bPXQm7K0snpcXA#&Qlb+29NSF!`uEv(2?+123;0FBj5?UOUCIL zf;L6)6op2>6V^G#sTx6!9SRuaC4dZIwPm-S1AgJg;*LB;Ap=-#IkKn|?gho0)-mXc z02&5seD_b7&qgh}w2h}IGz`}G+(X`Iq87c*=;qLP!CnqBsxV>pnce`L0o@MJ_{mib5~p6HXi2Mr9(Xhr?0^`3N9mc*3^v zGJB37>61J~A!B&LX19D>hoA!`yBKs$09i^3y{>j0grKxio}!Q?6nafjlfo11cc;5B z=(+&v3Y#nxe6~$M(C-;MMWL>+$zt@H=ig9^JbZ64$X5Wx1N8Fb3A~5M$&aTf6c13= zo0IDhbVe$RL4E=#2@YLCgkvFUk)1S8Q78!xUDD5_r3iXuTFxMU0kjk*ordIfPC`(^ zHlCu;QkZn=oAeJ4md9tvoou|0ZU~@ZFq1!i;|Sa(9`cB%C^QUa@|PC1RYfg2Hd2K_ zHwDl!c*1Ipks6Fz6g7&cD0B>-uo_-6cvN0Hfm*bpaub6B1&|)xy?ZIu;%PLc?>t2zJ-B<<=LO=k&Sj1@gMtLmX?Tb8t>zJU z5oG;bo}$p{@3uPcf9<=6TD1G74}*dQkTTrsnrC*##Vs=ec#1;GaIcHoasU@j?3GSr zP>2BP44eMGjJv0aexrpgc#1-uVbhMDGCL_ z6peYskk9DQ$=|DIP^bV}4|j%JwOgc6i@r2TQD{Bf89MgAIt4)&yDOe*ykKt&pgI^@ zPWg5p7fu}Q!BZ5fgQ4ZA4XT9*8g@vFL16-@5{5177WHq3TJ*0;ib9nzY}v8@;WSjo zm|wV(LE*ho9pl#ik&|+4Q*jagn-cz?|Fe#9#tm3LE6zEF{;KPxJVpPi88@JeGv>`G z)co!<_OZVzLeP9uIL>LvZGdyDg0?Gw)pE zDGD8eGe@h3R2XW}YpDnZ-4Q?v(4s+;OKi}gOOWO%3MoK~cE0J1>#t;ua~X7302RUd zf?JPPxQM8zNs2;6u)Z*B=XNjDqOU2H42lv!LGavf(bo73Zc!>vQ78zW`<@k+xP#%> zSKhJlqPr)65`V#WJk{OtoLhksPf;lG7rgj!|IQVtMRiA285Avm{=o1s{gN^shh|yw z6ovl4@UZCfB-}S^l|Pq3F#_n{Klu2Ujuoz+&2N&T(7%6BvR*y8KWY(t9DzZx0w@Vy z&be#akIv{fN*~Wt6iR}ZbGGZ1jf?O@F55FGP5?FT9UePsJAz!jc#1-jy~8oSk3XOm z&1vq>pm+h)9!~FyHaUR^GLzyd3blvR`@E_Xo|NdaIfX&@1(4(uKkI6H-aycsCMgO@ zKJk+!^FLJK6L1;m8F~-ucJf(lnN)q z^O$ly1R1ID6opdZWXP_lZjYe;CzTkKB!HrzdUk+$xdnoxPVp3lqM&;A)}NGr2-^O2 zGJ_rnpk@!@>#?4txO8o51y50^*+bZ4(k5#Lf_lx@XHc>Ll7c%!|A9JqDy*VOib7Iw zXV_t&gGU6qA%_^0B7k(EoTFrCFy4bd@fJ@}NEgaES}Hi=k+oEZiwsH?KzFjC6S(^~ z-o^Q$Ns2;uvTb#SeNe%jz#8)?2Bit0u6Z)rAKtmnMJGeyPM)Gr*F1QmSKIK#=r>xD zo6n$!0_YH|(-kMIG)0i|Q=X#GAy}vLxS^+npf+073`!S3FQZ_tphe_K1pRH2qR`7I z_~e}bJbW_Pdnq_KUa%Pg$R1vDd*^DPJ%Ws`@DzpY;U%{*=}~ykV_r*j20ao$6Qbes zF{LYT)nB9xPf=(>G<aasHWwP@&8o}!Q+Y`eDYF!}-d zjrIS>dqU&azN6^7!o}y4U*nJiq(%&3G^ZGh5C`$l+2k2>x&us+tP~s^H zeFx}tw}~bQ@;MU7plku83~w)NGu7P^L8p%L6or)G?S)2f7ve5)!OL_8^z*js)p{>v*uAX-dSCqMpFJsUX0h97nibB1hU@}K{WitfXce7$pz5sdx4?ed?Elm())}5y)^aLJ!nHmPTGE>dM zi$Tu?P%g|}KQLRm8$qr2@f3w}VeZ-^^e7&}FMblspcewD0t)zxwfkov$S#+sC{zIj z{7<`-;p*9KY6T2>DS#SaXgNE6az6yEROcxQHAse*s+suEy}nS#pjQGY5_a6Y-FFlh zdZk_DDGEixj+?5kxA7$ThuUsu8ZX!a0i*)IQEkg#xVzQtH&0PW1%9JScpFc^ZaLw?pdtZedQV1sl7`V^ z1WmW*DGHh1v(@R*XBmDT{d;$dL2m_+H`I+R8FB${8+rSlrzqqNbtB`RdY(bQk@bWu z2E7wN(ls*L?N+w&KrLD@k*6pmT>~$yoBTQmL7Og>Gw8hl3Wn!V=ExB*5VZU= z`{Ic~!=XGyp{>vh`ZHD$KabWs7cuCw0GbJ_TC>l6!M~B(X`Z6cOjymSEM8A>zSDvC!6f8C++!~0WK)URYvg#b!|?>;Q= znt2I93mSNeLP_x5hju;3;klqOhWi-wO#rFF#NegiV^a}ixq+uBqzV&*s$S3Z(CPg< z)|Ek(0_ZoqeCAx?6kJ#TB#x&j^c!A2^JnX&-w0Zv7{Q?L0>}y`2CuKx!edyi-aJJi zE0`F}diMfPN9!KWWzY`+WDXrjy>|_B(4p&lgr_KE4jst6jVq&3i=19mGU%rO@`FRS z+s{lMLEB&Q6ovfY(9N>+_d$^EIQesp*HM)KdJSt6cJ~M1(L~4bJVl|`ur}d4`0FbK z`FN@_s9FF8!Fu<)4YqjR+u;&VQ78!3yO&=n!jn#+4RaavO90uzg5a`;vqR9KJN1vJ zC}ax@g4sW%@USJ<(1bxX0%$n2sAOU?UQ12bz*7_&4lPnjo$wMJI_Zb@45}4Cui~JN zahY-eYEfO26op>JLA6qfyD@_Dl>Hf0CxBKzmC=5zZHrf1ZmRGUg;qa>S>lLy?g$!S zlft0i0_Y1`Uby|#g*5u(V^@4iKi&^ z4Su7`S##GQD0jNl`Nr$$uK?Nw1^gFJ_@714of$kup3VMDVjyw5dH}VvPZoo?gS4A}}Kv46vLkzO&^WU45v|tfr@7x$e z)S}l-QWVmHMUZ)AFZLqn*no=+IwpYb!|AQ7(FPBemki`73f+g(TXul?eFUZ3MltBP z05XKr`}5Aup{PZ{c05HPLpZ%#Zyt|JZYF-nXV3`&v<1#O+W}8&5!A1Qrzo@q&N_#v z`&Ot$Po`Hh$XWnxhZQlyWEVWN44lDJ6xt3eVrPdu8jhfAz6!357p#o{nhnL_dos;% zck7=YPf=(#6o)&FvFwYWwe8dyWGjH|@50w(S7ZgFL#NrErzm89*H)+RJ#Rcda@(

    P`%D5I`MZ9ri$I@?Qk4AI(z~>HzDoKSwXd70$0*0~vHu0GUIV_(I2C z8xi!zji)GN4qf8Y1-J32Ji0cWL8k=J6zH4`-SpxBg2L-~ib7MMb27VU^AG6IxvnZ> zkfQ*41~VrH&Q)ayy11IBDD(_wPG$sqk3^8+T^YB=>&Qs}?Swmnuj0~&2-1$?DGKd` zJHsHg(fB$_?l_1+&H~6B$_r<2UD+2wE}eLaLgr9jc)7ndo)`?;uFIg)A_x|${NEIK zBIxuEp85|67OF<}F~a*(avmEp=!^hb4U@NP0}Ju<=wc>MQD`+x-qz=Cz$;>3l&u)# zB7hWOH1XK^Jf4BfP~j;GDZ*$X;6qLdIvKXvdok#&0IGt4*rTB__(jBX9C(UCRWK0i zSgVf<_(zq-GU%KD+5z3I<(56&&~McLBTrFi2XwdYA3kvjwa9pW0fWv9AQ!lM=kK(_ zMPlO?@Dzny;O>3!XhAmw?F_DCkgEVX0c(6}dFO_p7OfBADGHr{HNNCo4sxiD@m|~R z?u{3$TOU-%`0>f{$##>r;@TdKuKYj$XC33I+cMe%pU2__!LLnH^shQ~8(zxYMj9`1 zo;{?&{wjAt^UIQCw8t*^ybd*g&JmuXP+1ZTcpp8RhK}=?4@(*3A%NsUVL#oqRrv^N zR>D&hk_&||D#jPzM$q($yBTyr0JVY%*!H2VwjfA(5>HX66->Z33n@K{pflH97<5qp zwS&c`-m*0%2r~8MDGIfN#irt@dv6hRUn-PAo&v}kittYz>VX&5&q(tWg{+|n|Guf! zYXs>VWHIQH0D1*OkBpnipHPb?uj45Sy@H`f<&Y;TzXcF|F<65eZLXcb% zPf=(R^q-$~$j7&O#a?YZ8ZSC80W|q1y#H#`2t0)U*(61w$vasD-OFgM0*#q>}vD(-(=T zMcC|104)P(WT*f6!f$#lPf=(YKwnkN{-73VFHU68bpbR6 z4qeOr%kb&laS2aRXbc>>L~GNX2nq`-Vvw%@Iu3m^cg-OD8yyMeDGD8jzS)wXDNhiT z*Sdj0egdc%M%FDny5q^cWI3LqP%(_GzpIa(gdiJJ#S4uWtiJ%-0Yi@^J$B>Cy-nMA zib6YJ=&@DN!Vf`fv$Ys>LjaY)-r;R4H{)e!ogAK`PzmfE-nz*85Amj?!xKaDW({AAgiD86ou}>>3#f61Aak}p2j=|-4;Nt;qHCmPr*ra)(z3* zDGIfQyLY_rMtofM&uBm4@5@)B_hcNl|Df%uo&ceXa^Y zm0>9iiWERcUJ-NT|Y4BjsV&VFRV-X8H39?Zg%4- z3hjj#)-9PA=87QSol>5S*U?=8ln8YrVJ`BsP>c5O;wcIxLfyzXbv^ttw8+Ow42lv! zJ)zH#_1kk1g05up6oq<1pP{8}ukENs+lEbH&^-av1HSO_`>g% z(&>1Ua*DM+gQ5k{QP>src4zJi)S`1XJVl|Suq&#xjRo#*$(9^qP>cXt32UimZe348 z(2FK13ayl^rI!D}L(9tP7a0^QfM&t-$YgUpE^2)`gQqAo3!X>uu4zu_H<}p`#h^F= zBzYg>eubT@P>cEm@)U(6?_;#8+lX(f&DuU^P`m&dlp~|PZ`TAoKKjrkMWI1CP$PF# z1yB3O@2FnC^Od_f(!U-HPsoED1gG?c{Jv6(O3kfj^il`g~9Wvv)gvOPG|47h(SpL z=qwaW{^@xDuMS(e^Av^7Lc!$IhaR{^26a0a^gsYv!1L&~=lABQMMHk`6oo9{c@%xH zbPPJZe=KuiP_h6rgeUC#7Ty^M%301+6f%S-tk=EhJ_x!R9mt>*0dx-PullGic#ELY z7@nfgIjFySy!|+yMKtJ}!Jt$D^bOWhBj3J5%JkDyC)c#1+J;1*=+dTR=TO8s>ilp%m@;5w?0l4^yZ z_!~S$Ase`kjwSq#LQrd&4GelDfPw*(orhO*fktr+xJ01W`B zC?F8uR8=8rZuw0Lq3vM4ey$8H!qDw2!AK zlnr}`dRu$VMJI#wi#i772%t9_ZxXzG`qC zLFX!#GU%xQl7Z#30WU80KrK4(ji)Fi1IuTTDuMC{TDoL6gPsYX%diM?BO$*CL265R zib9uR5oB1(pw|de4|HKro&eH>_tp2_*>wSe+63_wg*4%P^)p7z{(zueGNBC07eKdR zH1TKLW;~%WQvfYXd5%APypS9dtF)ddUw>Kz^Ob%p}TOeGpo{4 zL(suX)(m}#Khr4%&fPfEAHzJqXrsLI? zoF*v>Nxl)Ox5qpY9lCkLRx#*<0D1`1(S<<|@m!GGaGs*jLzs@f;BzP+K^~6#7*ryF z_CQ5hr@pax=47W6Pf=(ORFq{^nBjSb0cEZXDiuI?;e!1U{|8T_t^34N6uJu+Y}JJl zTq9?x8^NHD0w@*swDz-8%0VZC;WVD2P%7+c9sbe)e=F*5tOVrk3r=E$O`sLHk)dqhoED&Roc(5u@-e+GRM zKuJ)4wYu4Q{LUZgNjycNB&fgAub1wE4qewvDGaI@8qjK z=&Z|F!c!Dd1}OHp8=l?r4pU;#PXTlruA?pR{}I;FB1f^M4WGpJesnhtBj=6oYC7kb-0r`i?fg(4m{-$Ws(jfDWYPmO)aejxn=1pFwrF zj&amkyU7nsGVsFs%&+`E|7RVeq@2Ut|4euES1DES6#c6t7r`(SN#?= z-xnUo)*Y1FAgET4rzqqLkK=L4_Y%-?E(}w+)_8UQ5kNnoMJMYg9YRn-I8RaNC$z}* zg5hHXIV-3$=&t}OfpMto=KCKIw5c;sQK$sQp>}=dq#|g}o<$6*7eIk<2VNj+8j7F^ zdwGgNfp7<|49nhvAoFKC7}Owu#sOpzzYjkpC+G1Lg~kEYW0Dm9oY{-9P7L}dfF{DB zQ>n7-g<2G(!BZ5P2!}2#(5@qD(F?ag23hs}@10FmFgzT#sxKb!2D$STg{ok9XsoTi z13}v=(;0M309C;3M~aH}E7YP(-+78c6)^i@`{?3m1m!O(W6*H{^Z?dUQ@^_5XG)wN zPf_RrtfdBR*%5=F!O=3;8?U1i0;qYYjCNavNqEJiR}4>4sCg*V#&>MkfS~5R1~JH5 z0Oi9N$K=?|Zs^cGZ<3-=K8$h3JZe~upv0rP46+eGt6?+Cxm`ZECgGYTPf=(!Y-TBJ z9)WANeM<}(WGjGr!4hXsL9a2WMTbjyibB0$i8Ec>4-XIZW?3=FP5^bymeH0ybE6DF z184ITg}P?LVv~1s6?EwK-}GXTy#R`V-{?T;#uo@O3g9UU#lUa0eS8Yuc0ExpjzJCr zXd~Pi>@3FN7WHq#Qxw_=cZT=3rq`nut=?3?ppyc~6FPxexoz;2a(`oC@e}@q%>}Kv^&w-|GA< z{PsedzC1;tESQaVuFZdjpuvYU7~~{?jNl`BaRnZ(=wz@u%u^IHf{*B}%lV9#IMwo& zGRRo~Wy7d#XP?D!2x^wkQxwXEQQP^fr>#(ntjF(W&}jiw26qO33%lJ2x~Rod6e@!| z!;g-=iV&{+W_c{5&_Md}6AqN^);ib9e%LD0A=9|m0zKr^6B*Xj2=0|afV<|zuzfHGYR z4^1Tm&03nspo;>?75d(@mY=+YpgzlZibAf?_r9aK98d0rL=-W|Qvf}M+4zK-X*CFP ziR38?J%!nLlfy%7P>Za)H!$du0J;t(H_vx2yNaOIJ$Q;j*P-NQf~50A=6f=hmu*Bx;!WNE=l*+Kb#vaG$?ln3 zD$$XM`^{7I54<@~W9pnjb!7y7F&f7HKySg3s0o4ZM7{CEx31?Kd5S_cA#kDJRmUCA z#|g6;4X2UWZz=eb5#LT@yf(T_C2{7vDip>|mav zkYpD~=t|QT2zqF1%b@E5$P3mbw%<3y^Ao4-c#1+^urASG&L3CMOf9_5AYTELD=VwL zsMPQ{YEj$QJVl{gSv#E?xsG#Ci;jp8~hAGdbK=7 zA;T#6yls!gc>GQMuBV-vd`-&8y1jAT%l09ptivCNMD z*ABI)u1ShQ3lre&p?xE9S6WtM0fWK?&~G>yE-&ib4?)tJJVl}3a5D7tK4OdxUFS>N z7!)CZ)B%$J*sc?T+Fs@<3aJB>?%CHIK|N$oF(^_1P4$O)zO;f<2-@0;rzkYl-&W_t zs@CNQ>b3nQgYF2RaA=XS-^CpWTDya%C=?DY+MJ}(7D4M?q%r8O04n!|7YqH3$6e`( zFL{bW<-X7&t#SByG*q*cK~Vx|7A#WD*{QV-GK?k(XWFEhL9qhJ2%bmTBWL5f zEtgw7MIj@29=+G-?udRPkM>6y6eobLz#6Ph?%*w`MF-`1ib7Xl4c7U@Ts%_L+I5LR z@dC&nIwzOr9Lq;gtKB?BA%EzcILj!$LMt;H z7k!w9pcZN`8I&M^-og7b=01OQ1+_?Cou?@D4&I;PwV@lnGmLSoVNjv~dYA|;>Ujk( zo~&`_DGEJIggGRO+uv}Desu|IykL_AkmP-+QFeB3P>cRFNl{4hKGczgw_75pcKH|v zJrF>W)y~R;YjAZ&^$MP%kYu%U=hr_T2s$6Nm_f+`Xag*s9B5xV5w*zl9#2td11z4L zSmTN(<_C1$$)FSg)C+2Y3##hsRHOctN=vo`{D7? z>)kv>q4Tf;@MZiVe8Fzb3}#T80P=x(z9$Z45vWB|vUrL@J}}SMOz|R~#&%G8%%Fz? zD5#UHcI58~(g<4KkEbXU)X7dqt;K=^=wxVR`I$lK0w@q36J>`@=OgHMlN5yl;W6<| zy&ayc3VzipxbZs55I`;9G10K&rWAsl3V4b_E#NV6Z^LmN)FQJ9Lm2c(0L_E>^1X7W zS0hM!B2Q6h9?X}|?sGy9L6go;XV7B-WcFJ|`_sAU>k+itm8U3V_8V5x`cFEKpl#oc z7?df1?!$FdpLZkj|LD5!u%5s758yJgv-iqg8JT4yMfP6VLdY&7GK!`O8KGg6WEU!F z&_)spp%RIrQlV5ze)qd`UEkC5cRuI;p4X%E+UK16oHsx-e)5zEh0%S~)@0%^fIJr* zWY9ASG=bLeWv-7n4A3rpo)V!6w1zKVHx5UakKMk&pyv{(K2;5T9sV7+h1+z8r$nee zRSlSy9KaUUX%fbu7ZPX`jh8oZjP!#RWmQRu&?p)&k6E&!I~=-e6htvn?{A8CkmtlPHB0OdU^V^E?5QY@_-pwt8FiQ2#5DG^dEt(y{3_a{IlL)!XO z-motv(B@EO?ZO!W6X4K&7|K&3v^ms9r=+|+jzDIe)L_sn3G|#^eA-jz#sC!J%u^!t zoL+pz*$eZaMOOLq8T48LIn%Ui-~bi8j|>WUN`#zgTJ>q>n!W(doU(&KNfM|!y^mg{ zG{C8E{i!@9Le=Sg^gN;ht8XGMonX)#2{fHr4{XhX7F|=m z!=SeksD!#LRuw&;19YJ_Pl-?obzAmz9UBPHku@(E^iBe$(#0@*P!n8;wrnj=iBPKI zD=50_4uI5S@)(pXfilx*_fY-P?$Dz0Dk%}lOrw=L-7d}nXk+JwzLob;iUhhqW2(#h z{IUmVL>HbCp$jyoYF@l13F;qj?d{8;)K0LEW67t^S_h`w!#+-_HUH!PtbaUAlT?Gd z?b`|eRe%joiT~AMnxr~8|M_QVe%riB?7w<1X}%YY^e%Yyt0q7tRZ=44MI*h(wxr{O zapSCY3`&zgsnnTza<>uInNFR}QzDc~ovE7P|IR~;GJR|q^g#kCW=yj7w!|L6Q(vAE zA;pZz!Lh@<0Q%VQ5`#WUpn=p6EuU+QyAcGc@RSG*q<(1oVKa{aba-tPgVH6?eR>_+ zEbWCw6RX$plnC9Y*YWES&N%w<TK4<#m%ovMltA%1aha^y6S7xvFj0C%2OibPPKK)ZLc(i7PVNpm_eBm zD4OaYH)#jrm|L4wJS9TWRR3tTBpnM|)MIxuC`$shq87z&i(dohsLKgvBq!9+~0otL&QzA5r&e8K$@76(!(u{sFC`ST0 z-KMU`C z3ws0;M+{`pcL_9s-baoBpRwyPZ6r^L&;)uPEe}!04-EY)(;1W}fs)QBYgoEqya|et8~W=)`_@Mj98y;Pz*8dh;3_R; zSnKQrEqbK$m_a`!&^f9}s2&)NKad!y%TpqBj%pIdbqT`{?reh(De~OKDlEV^jiY;4OG@1G0+${f4cIWr$nf4 zApK6%&j_4`ww$+uLB$eCv9)>frNm)y=yuQNDG^d^ZEoQ=8Gl@2O!$5Vl}I2>y1Xw$ zMdF4iLnC-fgf!{$9$PaAyN4HhoMlj{1nNgm29u5VBB4dDJ$Xum`q7i2dCF)NI7jcD zLKyT%0=1zRU&xFNSSs-57*B~%8+!3+_TPq2Z}sAL3@VdA*Xg^iPe9#c(4s*lJS9Tc z>AP;a)j90&2JI?gP`L!!qNl7qXsx9YKwi6fN`$uP+35J6j#&$b&MC8{U*!$^R|1Wv zCxdIzqFOhN`#uxJl*Jho*w}!sB6L?3)TNF#=AoOBfGSpxSFa) zeV!7bE7U)_G{p&bnbOV5Kn^=?hdBVBFh9H1|5_@ zepLUM8MPlb+i*_gDG~Cc`o}L(?OXslI3k`wmJ+D;9r~`T@B9W@WI2+jM5y*1+O+1? z61|9oaHGIIzrV-wU3A6`Y`>Xy8e~-k+lR;r_p#V&%QWZ z+50z7iIBQtG`?|UXE=1*=l5cejRaEskj{~zPD8Op3wTO|6hEYs*E0z}Aiab17-TDf z(rC>@Ny8;m0NNGIQzDc`YbLDX3i?2cN}3un$W8(|(SE2`H}A%>s5Z@cN`#zfKh)`2 zvr_ty20GgxDQzCSLre?yg+Si3cm+G9xAV&$b zo#wmG+^E1_dGsls5~1xh-)*qP`Yu5BMdb`SEP)K@p}MSw3)Wwm{^ltWGN6ZQx3DNI zT{E56E}-&;JtBd6(!_fF>;Jd9cjA1W5}}?nv0gjv#71aQJwHtb9hE>A=p329-s=SC z=xddf2wk9ablG5{AwWwT=rhPk0wvOYbZzu_{ASqHkf%f_k?x~Zdk@^KU3VZkYAQJ?208FX9%)u7QWUCYzBuBFxso)V!NG`gj{ z)q6fb5q$F!DX~Y z*TbRPReK79&PX6vDk5s{ywnV!_&Pi#LatOq6q)l4A3=5-jTq!2fog=&%wu8$+=ge# zCY}vudln9yA(_6!+W_5rpN=7s2 zf&>~zMXgO(4reYW(@L_K#KaSysU$(p+z>nJS9Sk`YWGemuP?j8@e#)q6BiJb9C(IR-A@D zrNUDpG zJG>sUZ+frthP^C-B53f|sYTCNfL1#2ln6!8;BAGsN&y_YySW1ybVUO7qTa-{*vxJK zIezCU5$Z*~i4(OAv2Nt<)ENxADuEhP5z**rt=#~+HjSr5s4*209kTSqopl~v-^8G6 z5-6X(f|Lej>jHH422Y7lK79qP4%>*+qpB*747x6XdQ&mppPxrd0jk@Gr$neX732B- z=;jUQD8Tp{gKkJ5vxn3X(=%5F$lZjeM9AzRE&3a8fxU^|uOBn$rUYsgOWQ1s@BA5{ zMoBy+LakzHIlg8MM`%%+Mizs7Bv2lGd;7opc@3a%nmi>!dGzgF;nDXUKu_Ik1y$Zh zz7j}_#u>VQQp0usX=iy#gtTa!;Y{Np;Q(1Hbz{&i2^2u{N9~5hBtVNcRp%)Y3ZVI; z$QMP`0NS-ihe5X`kTrFd6INN705o$gPl=E?LNyOe+jg{ zi1tv|=r{xpT^}=^5~1})RP~qk17GFwZ$cOpAc2Nb{ng&3olXK2^_HhZXeiZR?VKKt zlk&SqCNt=+1PY@O$j2dZ4FIwl#Zw{_MkA27JiT!@>+2pR3<{J$gM*c|e^-w-1n7__ zPl?drU>hCJaXawxK2p lX*CxJ@nBj4t>HEsg&^dC=&PzinHe^EMv-`+=;3}sM| z1e#0TmVn+#xDSotQl1i_xzughuzJu7xazvznZux933QZJQg7|AhcCX0Dk%{r zw5o9rpbia97!)Fb6pNKEIt<3Kpo}Uh5mGEx3jUF|3ZS~{ofs49< z$E*Fm0MN^6;S72xfn4eCUVF2z4YcUybe<9+SGv1Rjbcv#WOXf#L60QRJ{r&Y_T(+D z5L|wpr$lHUjpw)@^2MFN<~01vpvMx(o?ge(GsgUY7OAQ5lnB|=>sbB8s)o>_?2YY0 zD{t6X3AB}JO?#I+VD+rZW}XtEtyF8OIkP5qri`9zGU$l}(xnz{Z{a=p-Qr-GYX&xtDOvbDuG)6qdB1I zC$J_V+M1_CsP#XZrz<-+1fb5JPBJK70;$oLOo{pj94j!&;3*MOqc54IFZ~|@JHNPl=Eh_1ea)Qn?B(8oEa1e&v0XAc4~8LtXRN zfFNj5>$N;3LTU7&9==U~9zbS~)EJZ~fx6RJ!L#Bq69GE%n5RUjJB<~z-Di&#&YoSS zFzBTOT1p>*_w^s*h75LHc}j$q(nsL&Q?Y8$A}32D2ECF%kEm22&;8G0ooeOQzG<`R&#jX@vZ|cs?|)3K`9c*gXX)J zq;JJ3&X-kEBIH5y-81$r!m_CS8<#LBRRT>Us7KESSW&iQ6HkfIM1p3m-j4hBRJ<@_ z(0d8=hDH&utEyp#_hSN2iO?GwMXYPsuPYq7RCN~yrAZ(qde)6x95MpV(Vam&B|=K{ ztXuBY_836Fjs-I4g9Iv}%e&ryN4@~fI>A#SR6>_`(7YZo039rP&7hAGD4*U(m$aW_ z&Gy3IJS9T;^ggmY@3ILXCH+DMrAr_a>WEF5wgrcjzg9_!kO_6ftV90f1GLt!X;|eA z`$+=Lq=#zlE5$gZyvv`bL}(^GRPCR)dk#=w8+8U{NT6>awCmyErg$ISYs*t2^ex0j zM>|^$2ls;Z&S21I2{f9z@J=Bku(roxA5V$UXzIe7XXi#ii_+e0V$c@}^!p1{lr4&V z4_DpQWS$bC-(Tp-P^Xh7K-Qxi8I&o3cGAW0b$UG<3NstcQzEpJE{3URdSHuAd0b;q zmIT^NW7l;X9Krj@*psJ3XfusntFLW<%SHEAi)GMP31mleK%1`AG=)PqO^K&O$d2ZK zI+)w{fd8 zJ9flAR7r`DX(A0N&x-#CEvli~jX^mQXblZlK7SSU4qEiKN=k&*&~T-dVcc(kHkj!! zC|3gQ`$4}GWfFY|AiX_2B|`gtP=)goLoArQoW6oV-zCs+8d9F65{%pFT>8XQA~c+a zls8>`G6q`Id5k%O@+43i)yREmZ=wP%YN^drB9uloaydmS@Y}oX`Lhhlmp~2aebjx3 zD^6t8xxiB*)R5jsMz?c(p+&A`Aq*;zKx#Aq`}1MX3}}&MIZug@8V$gHm~9vVP~?ha z2K|sg>uJNfnzqTE0CG0qDG^#9PHXWJRSE%G8ePJmpAx7Z%{)FkG2<;jJz{uDgz70~ z9-9PV;c<_its*LK*k2N8K9#$-H++PjTQz&}lnBkIa`*jq&v0tyz4cHA6-uC+)V0jM znu#+e9c_6^glMN`(Ar zpM+Dpbg{19t-3FRN+i&88i72q?)e}%bQ^2%ln6bi5y*Or7T{GEyfU6ar4lHZo(y$Q zY2!D8<0_sKp2EG2#4-eWDbM=NT5P0emG`na}HW$8O2i~R7k}S#^)AJ1E{P= zy~xV@s7wM`(>Ft(n}2oy)TkFviI6pYGgLUXa0jTiT`vZeOQ1j0o5*T&bvr=&5Al=; z{h{8(@!_Yp0TlU7k3oMWkYYGGf7KxDTE5QaDG^c(M{jU$xgDVI6Ac+uA%VJ4XL*Kd z#z1J1+GL&*p>EV!_AhUH03eSmmJIqQfi}@s(4CE`cLCaem8V2#6MY5UnKGm&K*_Z| z8D!Do z8;3f(cH}7$TAM^rNM|fv^V{7ns`7@ll0cKG%3^>;>(OwI{+aQV2u-FcixhNncxiTXzalrp=*p?j;T&mbEK)Q;vR{ybli3@!2)!BZmC zj^-w;+W&F{$lYZpgKQ;GbGnaKzZvuoAY)ga5~1dFAJy2k1~)}9Ej`H~I|)>upoZJl zEe2@+AD$AS`UH&(HZ_12-CN|xphFU93|)2OqQma?oprQa*#ldR6XnO;}s4q`dcL>LXK2D`&7mG z8$ccRsYF-aM~)Imoi^MU>S2lB-YWZfN`%yD!wrw6%T)n#`k=<3!xE@2UEYc66=u+) z#UFV}gzD1ey{}_^U4Se{PGQgy8AQDa+ok5q0Gd6Dr~U(?-h|_T9XN~@>}l9Ck&|(_A&CnR~0a`Twmo0;wB+yv;ffu_0TKGvkqL8OVXe|A}%b87! ziUCTU@6DiN5@;WFTehqHL+INAo)V#b)NQfcoIetv55Cb1Ixc})(2K88KOZa(Prb!c zBGiIjeA%}H`UBLcK?Z|PNT9iNF=RP>s0Zh$Rzsc=p}BN1Y-wzi08ryKN->r9(Mbuk zoGu3ces?b*w3eqtXgOUBv2(qi0;Ki06N8*3PzhByM^7w4NGq17M5u%+oQv-F)dI-2 zlNN(cNg%yywY8V5RoMV7I;P4~BBWQ%R%h<4Ki2`OWwnGsrzOy7dIasM=7QapHr6~P zLZ|5wR5YwC51^2*W(+zbffA{>C3w-yKF}ifZ#*SJiB#NjJaHxN8j-2z!XOt3G>ZDL z{;_E@0rH=~QzA5q`mnLJmMBAuth@plG6 z^&|$lNgzcPO@3z0HUI_I;3*MORM8}L)5Y()uoZ<2a+g4gA4+r6N}69zH3<&&ys_j)t0hl~P&w5k)ZX^aAD~;ivKZtgfuiZX zZ}VH@0kp_+H&2OBG`;uLZx6?{TG`399#!6Smn6_cT9@kFuND4?WmF1JiO@t^m-qq`1@||hWpLimPJL@0umxixthvz_Z zH>{5Y@}v2MW5Fk}MeEdfN`(Aqe&O}HE$R@RSHmr9p|?g|Rp_<9^49L3bojM;eV+{WJR|v}ms%Pl-@R8jX*68;37GpGLk6 z@{>T{=-ceo!0(9wSvKY=5&A~oW-mAIT?H3I$d-5p`AZ9yg|wZ6<#A~b{s$!|2atq%~*6fh`60-4dF+c#~h7C@)# z@{|af(V;6nql#s^&5gnt6e@u>)AX5z;}H{p3ag|2ag1Yd<6-yET8m{)2K@TL53VobBHps-`${GE5N`zGC<7A#|q#-~% z4!3(!dBcWDpe9ralVJBI3|h4O2v3Pn6RLzMSos!5sMdbfWKg&S`cyzG1q|{p0yO9w zPl?c{0>yA;2ONMsFiD?55fZ2m_1e1lhvW2_`(&OHp+3}WJF0pD=YaHmb}}eZ0>w~e zX29F70nnmjzC0yDF;tn^Q^Opeb^Tg6Gbl;|)xAdx%D%P6)f_!q@{|bGy+`+vjmcrS z7z(%hF(_IB{iFBM?!BeGp+%o}@RSJsqxaFe+WKPw`ua41K`|1ju$;yU;;Y)^BPIdMttd(dFG}&KPxo_Gj~y2>qkW`|80y z2yLG_g+Z|r=oM9F78V}f50J$)o)V!~RGI0ve&J+jk>OP%20f8L%JjK4)XDS=KR037| zOmje!1MpS;%7CXts2Y{7J**L*1klKcXa>bgpuu#*TD$E&2ZwG?Bu|OZV7g%&dj9JP zP}9yC40ufHFHHoflZDdL%KPZK1bRqcss&oprU2CF z08fd~L;6yk@7S~>KuVuGG3bQ^a;2?W991vk+J`w`cuIs^X{(mp`qub9@*JzhpacnY zgC0TEjq<-ii}vd9lnC9RN05zLbL_)*c3Z-rLmr)fI+XAN z4xRCDGX}ksKnazC6IEcgS292ECF%>uEs4yG#{(e6I6& zN`%(afX4o99=JK@=sST7dM$xW=^PDM7KnS#p7i4>5i+H7RK6+h0vx&}&5{_DB!L=I z_g`<=RtGqAtD5ta2sNbce?`tc3xM|TEM(9d38YC+?}?or;;Y=zl&3^Ulb+rlZya#Y z>3U+*_{tmhtpqZn0+k&aMc1H3dtUOC2pLg<$}9cZZJO3Vv>GWg>)APr3^wDt!gOVjs9qJzq4xNa#JpqS#N`&fA|0waDHdeyS&)USG6bTed zHQQgOzQ%9w4qtgngkq^?duZJ_Tn?P0>&T!~3ABU8>vCT{$Kvp2<9SMicF=g;3ROR> z**3p?jY01v&}aH2zQ3~pJ{e4}@RSIBrcdIVdVR5g|DSR!gVH3B292p&yV`Vzi=jas zo)RGq8dLS{vmTc*4q2VWpbru#h@K30-A>~}wYwotiBJ$d88&JtN5SPiJFM2T%KPY} z1hS`x>T;6}hoMC};XEZm_ViHQzBi#0Kz~%bF(_RE-KT=d@fKFNruAiKo)V$^R4{pL zcc%dWtuWJJ&?gDx@LO5C--6o+joQOgBINL!I?H{G_X6}Ibp?YmB#<$UGn}-H!;V<> zG@cS6V;W~LT{XNfKia^p3A@phf#@gfb{o0%_3YJ^15B+^b}2O`Z}V4Z6JV zjDFkzpq$mo49b!~o#@3^R4@mJ(c%qxN`yMmi*K>ZjsSoXqDvU`RRRs5PvUu}w_x}G zMGQ}g&;a@*eqvz!51BASvR7`ZrpTzWX>E0nhSQ_AlS~+tD}g#wXL-pxRUEwCHkqeHs55nzHSLS8!=Za~!-+xPCD28>VGVBf zz3btfWkpvn<^RS~5Zp1^2+Jx|w z2o0in*jHO?V41?PvXe^;S4I4Kshw- zJ;$Q!aA?tGbDk2R92)n&)xNnNT;3<&r!lBR0_~$GLwMdFH-MtjcuIu!(UW14Z^#dT zymkLFs8j;2j8N9LT;TatO($gql&;bdg z=uK>Y@p36Z!!>zIgcQArjzcbN0La8eC9(28Iw*m5Q<0cq+{eKHO?Kre5!y{fVza-W zwgzZrsTzYUB~UARdN=v`0ngE_KRhKut?21}y2hYC0DaJ(!XPUN^pmc-Ew}bWK#STf z;wcgONmreZ&5S^(W4wINh(Xp}ppJ2W#z`&v-)0ZszY2-ufBc_yj4n5nwe|h=RN%il z6U9^Ff8}z6hW$O7;jdBL?qtXQD;r7kqiLn>-PB&e09mT?ln6!BO56ELQ%a#l*=F7h zvXwxs^Od!C1!v)K<&`}=B|@$9ZFD|2NjwNpy*JSevXel%w5(wDv=scnNU4$%AzfNl zkf@RS7@(A)84NlkfhJPjb*qdUI3XB4jHg6sBGp}Yyc*&Jklksemz6i2y##7bkNSQM z4&w40gEKrOLe1$>Uy+uv03g%9ofzaGfu_-K+dcd?5DUE)Rq&JuO{3qoyYZ}G4?u$# zYca@C0u|9@)51vS4RF;>S;A8yR78_a#~0PD2T)G%5(XWXKr^WRD)aCKe}L|W@RSJ6 zp!%z4+VwpE>fd}1gN{fbi(+N%L%sG^2S}v_Pl=F4F)g98Z-wuaW7}L9bW{QbP!~R` z;Q?=e)^F!25elF#eEu>gyy_+;1v1D<0;SQSum!tpp8_=S4Nr+s8Z8Q|mc9hL9-}mp z7<5bm&8K(D%vvpRFN^M)JS9T&>7CM4C(aqp(c2S+3_31>HssR9kUMcVwCKu7o)V!A zx%4}pZMK{NsQhQsSCu#H2?;cSdTqvT-Z;!svyi7mXaM!vHn!`9AA$Sksx#=M1X@Zx z0!Qr=j?kht^LR>xmQs(P{d)HpXi@E;84PlkK$ob9$fef~U4RO!q(tZv6%jq#`spV? zy*h4Y&?yPzOY4C0W=>cJkZLEM5+Pq&2Q=O4ejz~T_d7D^v;P zwcWI~bOp$N2v3QS4fWdEZ~k-$4xQ!EEC#tspjvbv8UA&j56~PZo)V#2bRYE_&=~hg zSe;kvb>)5JCV{+YW0MynUgPxHsC=FhAurn4#D3iwENa!6){Q~#63Ch$^XQjR(4ubB zc}j$=2}<022){O6Z|X4UtOROKBfYUonRp-V_2DTIYEC1){-qtUANstZ0fWv-pk%5w z?ep0g>qd^L@RSH8Q?2RVn?Fy$#c*e>IfKqiAb%=QacH(iwmSCbbZF7J^b!VLlt864 z%+kkwbbD$5jsR4C%Z3= z^#sVub0~u@Nua*;Nvt=%=1*wR%!@oFLVf9zIKruD1wiJ1<}k=x0vS|Lfr`&!ypPtE z@stP|RM_aexBoX5pvVO#47x0V@@ZD9cjg1^S-a`;lnCY1tk$mDld#9P;+7MGu1KK1 zv?{9GiQ9Q_==$I0DG};RtD@HZ9)KG%oKx{-&{YYfOfx2#dnOnG^rI0^iI6hQm;^rQ zi1XbiP2w4JO#*GC8T{=0miXsBgVDmG{vN3AB?2B~s6nW&?Czji*FtCk;xdub=r9TBLQf z7lUp}AUm3xxfz*o3ZTwTJS9SQG&R%f&WpwpjtN`#`C z+v-I4MBt=tk**E~B4vkQCTc@-e4&C`Wo(#GzfiBR~J0-UkZpg5)E>DTj1$ufP>r(RzKsKww z8FWVilM zRlfaBp&CHxo&Pe(Ujiv=dv;oC;n?-HE<7bdirSvoBkH)~M&F{{+sYd@KmsXt70{eo zG6GsO<^WHLkYZPXtn6zzt5qjclRfzsW>Am>vcE-NL0^JcRT$U6fA*G(u82r zuk6#%q7YM_5}}hcA=r6m#1Vk@y-Q$Fhy+snme6PaI`~OEA(^K{Nby@j=3hgwS8hMN zfI*=WD4bs9dp4*hLyMd=cuIuA6)Oei={Q1*hMZP;S9u@Zmq3SUysnLN7D6@8@RSH0 zrtvzH%jdClt>A|mgC0nr`7|6|abepBXi@uLJS9T&X*fFNp9O9MvVYnX28BtW>)(~N z>$`UB3DC;vJS9Tcztc`cqi4M0my@{|ZIqOYI~{Z`ii zdYRjaK@TO+Q!3yOU8avioqpeWN`#(L0e`;d#GL>sPt{`3BMGEK=V*V(oNmyfH&s$1 zq(kQ@_UO@cfXZ(yVbEg{l?WCt?TSzP^<)+^+Q=Z z-8JC^Kp9n1A~fpg9?1s&6>tjB2-AwnQGk@!=Y<< zV+Mm>NFXyB>g;Ne)CDevPgPPPWJW`su8+_404S#MW(Fllpv-8xVZ9IH&ja~3;VBWy zq`j5SzfCFwXqSm2gAygsIC`l1_x8Y=1lw&qB|_uqp=uZ!k`2(W*lP@WDS-yj4~OMn zab63D?(q|z5}`r#!(lUXdSTbnyJsweUP+*j^j&w(XcE3*Eqd{k2z^ux(bUkzZ|}F( zSqyqDffP$keeQ1egBFF^@RSHCmYOzDFFpW=uGg1Zsg?Ipk_0+WBajzevsVGMG?S-9 z=sb-;_EG7CtKC&6bz{&Q38a`2%R4bH2%wgec}j#7Gh!drSL4S?e=i*dy_G){*) zFEeLQvIHuo&ho4dyRek{^m3jOp3oy%(4tV&WCp#LKuPq?aN9L> z1+>Uv7f*>$5`8mNxBhqzAl>*92Bk@$a^>3E%f1i8L)YjTPl-^uvaL?zI=N>7^6S&; zedP`NK>{tKQjnl68*xnaU|*gRp=DGGQp=&G3qZeYhcf7+1R77T^6ABguo&-$9Z!kS zczTul=Ct$xXy%tW3`&*ILc25l1teUd;4^km3r z<7x)b_^~`CLJ9O_n74MCHMD5pMJEPjNT9(~Uik3Vk9Pny_u?rL8cgMdPj0@z?tkaM zJ`DOSfiBX(;KD#({H`mkk`kedG%$E>ohFWZ=P!(9uwy5+Pq2$$wL%i35YnJJd_7ypOUZ&{q05 zaX(U{1R#@+JS9R~>EonM`Q^n>$9UJG7lXcH9b>(gSG96C{n`)Bx3=eh{GWA3uBoAZc_2M%3*_jVsD@9qK#q(>dz*~{1Cm|Iqr zlnCijhj&V2y?p>3FV$qw4+%7ss{QN#Fv3dmjemGbgoaYJ{{X$!ZvkqxM4v%FCD5K{ z%G$?O|6z;zEafQ?+VhMGysKB>ej`4iI~nv#0x5nyc3SW>{MvkbpQl7f@$0d^d7bb+ zI@#KpL4^|Nc#g8Rk;9=ca4~FZ!&4%3JjX^yKm97Mv@O}{$Dkq!6*R!H(c5D`H;%T%KNB90zIYqg?nz_8vqn%&r>4wl;#&2 zM7DhhExPzijX|Xn=!%cBwuf?L1wdwnJS9R`d~9?kZ$ID-P>s1$81zR1h0~Dos92xF z0QH*3Qz8^jL&_^3znu(F-#{Y4r7jm8lDm(JpTW{gptYsMl8BZ(JT6I<4J2B|@X9*H%4e zunIs{Z=)GhA%U9ERcA4Kf(t+k-|>_PHKD6+c(owxht3(8!JvN>2PBXcl|{|$;XND<-Oy^P|4WIG6_rITKbsK;(4A#kEOJl+DIP)dyLMRskn?h$ z5+TJSsNbosxa->2&?OAAlt4bTY(%~E_Dg_*@AH%h`OvbF5glt_9b;shJq)svK+R~L zuIq&Q_5j(o8HvY_K*Zp zrY?N^>C%1x)&Ij&BBV@Rcs1McxTJUeB6SAYOQ1EhB;jMeCsxAjTg+1;w1$=>Ec@n( z&$@t+84PleK(7?9iJKE*p+!ldJS9S}=p0R1Q{55{-Hwi%8RRH|Jm|%DD_Q`Qy3dh8hb53BUEUWho?#_S;(neIAxFBrSB7i|0_fP6YYaLfftmy= zYcKA<6?Z&-lF3sd)FhBLpGjSSwLOi-#WLuq1X3)K+udMqcQ{8yRZ=3PSR(g+Q|~^| zq8%Pt404h{i8Q}pcJB#R{f+nJDG^Gf`2|}a8xw$Zi)&?6-bcqI&?6cbO)!f#h88s} z;VBV%MB}1q_OpusYB{$XgN{p}#HU`K`kz!`w`s5+R*px)@r2tqCof zs@Ljss-A4mE1x3I)3jgfHpmP$)i~2{cl9ljt%l->biI5ldk7gvV zxeU->9bX2Wmq4jBc)Mry;3Q~KqApK~P$~`H8XV9s07%;-ozdRHM^02d zc}j$y(lFYm_ey&J+EP82K^_ul5nT*t%sXOD!h{+;B|?knVu&lQ?E_Gkwe`MK-bbDi zXfc&iUrd|P3J%@)bvz|Pi>Z{l&60t*FmqvSF9uzdK-K9>bxU9nR%UuW;VBWSPG71W zE)-pZ79H$6fk9ppsDfrpM()*`4K3QK##16xK{F;bdX88Q&^}v323?XsF*LFMe!SOl zfcDz)lnBMp#QKe8t@Qy~m1oHyZwZtbOuun6-W*>O*7-aoLW#lj6HcX#?f_&v*^@z+ zC6J<`tp5>}1ZdHgDLf@Yii)zDGk;-kBH~6kgRV%RbQ&ers<*NwK(}x5lnABMD6wC` zlU~rGkR~4(bX5ZB(jtpzN6K*R!=a`;B|^Hi$Rb;PAMUZx*61&Tu1O%p>cV3)c48$= zPL-4hDOMNG@XF0=B6y)J=zP;a8WQrQ?dbg_?lN`!h)Z(_-!*0%ua*k=TT zZb+aJv>DnfGf!`TYWC$R5gI|8p^b>i!OyKL2lW|rQvw~OafTFS<4}MsEqO|Wj?y@T zj+-gY@fm*E$sivID&9A-b8%EZmYKbe<=}aOyAziisP|H&T*w3iv&s_#m?bwEsv%`i?*)fDG^fa z93IuK!6|4_++%MB-IG9C)De5)^=2$U7h`!!gtVw5*4XYBR%V{;8N;9;38WY$p8IXU zLx3!L@stQDMu|NK>$HOw-9DVbpkN6Uc3)Y0tyWEKfZUJpln900x6vtipMt~DLw_oL zt-OyyB+yP8>g-=scPBude({tD?WCbj_vngh(4r4>RT&g2fpY2c&eWV04^YHBo)V#4 zy1boI&G5yy*k6l5_a)Frs_n7vb`qa;0|IzTgg#Pj&zuQGcroZVS<0XX638%B#`CbxC4Sqy!pBeb{QF%-chYEL?d?ga%R{cJ`m9uK`LbE@V)Y1X7HtwzE66 z0U(zWo)RI&m}*Ap1}rlOp404G53&5wfSwa@PUNL*dZbDbHlkLkTpiVQuYMhNhYTEvU^?A~dU^ ztlU z{>T4W$EaA$V!Q1wt}$8mk*CD}O0k$_vX#$xfaXn5&aS-ao=TeE@*eF(G;0fvU5}W^ zQzF#z9ziAVcLDU^W;X`KOCUG;lKFHx6-N=he0WNP+~`ZjXF|0YfUeinWzaJTbgEiy z?Fly9@lk)E7Eg)LscN=5+PnYX-lSQp4H)!X0_D>acuj}lQ{l>aVaQV=luu9KEjL}T z)HW!_oIx)nPzH^;b?EX4F9wH)JS9RIH0JiT=^z}TQfYIJK?xG*O_SQ%9h(i?0xg=` zmZwDMO%q$4lN&N|vT3<#D1#Cu&@UPnJ@ouhCP2e?@stStqH$5DpWayfaOXubgI-FY zk-zB|e%)T-6z8!7o)V#vziB6;E&4dIeyc|*gI-CXlA5)(i>5o`_HI@^c}j#zYTD|w zJDfBOF7L|@t#T@F*w+$h1FZu}2{kcZdz zws|8!(>#WL=O`N zy_G-{XgnvxYyy_9-SFfo5t=~bIWrZhBFPeH1ARZ&9+G_&pef6EN`yAh_w$l9#z)}L9gK))P>KY)Pv6gv zr%%Vak&Tf&B|`V<`}vspWn+NWbjxK>ssu7t472>(^$l7yq&rWEkTH#Z3@A%01t{LU zUT)=m^j-p`(`1wLI@h)Ud0Ft32&L0xlhN$3bbvh4dNC+X0)^2+^4Y;Ja{=1(fu}?$ zj24o+Pf_a)&@PP$4Ei8}LVnRUAboD(QpAaxJS9RQzif0ST>a?^(94sC4EiX6Jn2Kd zQRu$9(4r`3o)RHX`cOAr72yU@$ZtyqrAwe#`Uo_LO~;GjN-0Q>wq( z72}7{n7uqDLZ(!Ib=Y_65NOe{)OO!1Z`e!;r1*6tI|G+x(4wvHc}j#7zmB9gJTMBN zqZ%U^lqG@6sjhzChuwbxGS=iN5h|y;`q35Lv2H}=ls<#LN+88A9enu|8U#>Qm6Ql6 ze(At&*3S*lqG7*xGU%HGdPHBU_DzFvm_@air$p!xeW?yDozw%MY73khlr4c2KfIf` z_$)qx-c(76km84To4wdJ9H4apehkWyKyGyC>c*_-3WsjSU7ivlH#&4HdsyQ-uV3vF z8I&u545{0q5_HK5pv?9>B|?VOZE>$v3k#2R4-_!yy97EHrmStEvJG3*<{(dr(77=B z+{$050*B7Xq}_ z0HE(xQXE=8MG`2Rp5FfHS=#^_w~nVoD4U+%6E-}wffn6=p_E^F zAN`gSfMWZN`%(XrWxvkRdAePy#5{r{gFTmXxrYJ)gvw89Nk>VQzEp0w(YHHIyDCl zU5mRe3@VdAA@sR*J1U_VAdNtt5}^?K+}gh4q8~sXTijz%xdc*lmirw_yaCYHmOLdw ziq3MgZ^e%Q+OsK%L4PGsHF|odgcutGf$s*Xp!gC%?vsyfy|!J>cST>SfFxg8c&Ik*%KR`xdSw<06KWZkwKOc zs0%&4U$`B?3TKO}JS9S1=;?hkWgvbMTUNi$AS(&9kUGn+?60K3p_^TUr$lHWb(Zyi z#Gi*2nXQavkhKJwM3c6EMs%%*&?=r1p-D7p`?qsILx6@n%3_d>1o}=-?}#{K+|f(< zF;9umcY1p3{L*_4P=8hBAC>o!tprLbp`UOXGz|wd%B!SAD5Zo7RAvu!0ch5qZVa-M zK#HGm`gqnm3J%?vy*wpCil1<5klbuFKxJ=q7<5Piy`uSU_Z*Yc0EN8eDG_=_^W8;% zt7H9DqoD>2vX?-?1hweU0S{eXm6Qkt6BIf;6Z=PfUCbHeAc4lxc-=6y`B>&jCi zG?vEeqGr#IoS`8FWknsnYE5>DwvT5o`UKr$k7VW{3Or%*CP34%%}WbX)@M zq{T}4+qXA^L-(~xN`!XOVx`5iMsI8S55}{@EDvxjD ziAzlbi=7yBQUcASj#%4zE@9B310_5qLNlo&HvCdB&UYux@@0^-1oER9G3`N?>HsyI z!&4&UM>Apx+goGz|ABuzgHB1H4>Z(i`zme}Ko@9DG^#jPlhEsy5L;-46}N_D(|B+63C8v6WX0ke?p4}@8KyCvZLO_ ztegIIphbU^dojpG0_~(n(BQs4xP?eU3Qvj9PI?5HjJkr;Is-K)FvwK`SnhaZrR&)XGN z-mo4LXg1wPx33tU1IY0OPl?cMx{rJku1<$TH(O%_gFGeBmpk+upwHG;05n3Ar$p$> z9UGk=Uz*~gza?(^47wm149a&h$V&p5(>ZdS zvGWVGX!>8C5+QRsM@^hhGzX}Wfir_HNuXO%ba_{#;$&iam6QnGilVvl6)xQX`f|^Y zLEaLmw1SocyB)yKt*{`T5~0!xT2|j~kTyVTn zW2!xxI|HQ*Q<@REjF^qpK1~g$A8y zY4~S=W+d>G2&vGZlTX0J*U+MlBlxxNlS4My^m2yLcajAo5%xE!G4 zZc`X^T>`04N!9nDzykm+b>}G&QlXNnF;5Q#0QB;Y5rb|>pb$EAc1yP6n5st^Pl-?n z9l8Zw%CXy$x4@1;Hzkl_{r-^9GuROe*5@e^Qmo%^n$|fV>KGT@^Jb7wH>hJw>DpLJ z>Gj7jX#SfB{>T4W$GGbnjr1-aHX4p|KqOCz|JAN*HaZuIdu;;9uzL*quY4uVZ%41= ziSz9H1Ekf1r$ne7y^fEjulNm6*})73-I74=G}62Bl?E1g|Fq;O5pt)I-kOs(e*)-e zj?(YSo9?y*>Q3_u(<@A|Fm-7zPl-@>nqQc6^Vnp7bS9}X=#B)sOvS@@*B9V0OZ&+@ zB|?{}c-YTi{!f6KU(#Zbp9K0y1F%PS6-B~1%BhkPp^r2G+y9FxK7psyT*@GS2{eU1 zYa8j$!8MOfYw?r_O`*@)$Ayi*LW@e)?qN`X1UgE69Q_MRGoVHJ>v&3pj#3}z$LWr< z0QwT=!l1hn=m1syX-y8p0`F%}c}j#1P}QGT@e}+u8{PXJg90T`3wo#I4{CuswD;=6 zQzF!Y-YGql2j@YH{x~Et=$-`9r5jee)4Hv2=rj)VlnCk44Qpsnb1p!2G7A|LB!SM+ z9=}2LCQSh-rAkVK&e0yfS6{BfA|j8m&5A2;*kB3tp2n`nCB7L4&=wt@5~24rcD*g@ z?pkP(=>>HLg-D=(wE2v;>XmYU3_W;Cg#OXyGZQDb!-WzmB{LWlDuFVo>tRt?(gvW9 zRZ=39NnMYq=KW%!MO)`>X3%{JG@rJx-ldlD0-%}mc}j%l(-zis-Mbb5)b+L_gC0nr zUv!S%8d|sjH1G~jiO?@PM|Y;1SOes$a-Bh866g^&oH!``$11y-l_Bx9}5fVu8yJ-QFJn>m)p2Slk zr1;%5lg*QG=5f1*a!KWV6e)pf(UT$8SZy^Nx_O#BB|^36$xsp4u^t?{)N|b!6eWRz zLg>lxST7nN-}5{rLO~()2=&BoV`%yeV z@yiVu6eEFJ(;QHWO280k(S;Q}B|@!f4(Pb1QYt_iLFNp4D1puwD{J@6E5yZ>lY)6l zgw7Y+=&1DeZvjwr%X18RB!O3g7q81EA;ILmBi~0-dFY z>OUP_oWW1o!BZl1mL972o(1m)Xmw&TgJLC6dxEOh&&EX-S}%D@gxVAIq38WZ0R2!e zWzZ7|l$c68Oir;}2Zt_o5KoCvVk-U8!J?k{&2aTltJ2CFHckS)rTu}OjZHTLWNOb- zBJ`H_2OhR01i!r(WDaG}QwcPg`Z$BO-NQqtmc>&dG@1H1*M=xZ!=YQFGnYZ}5@=s8 zjpx)@@BmsgNtdTYXkRXsuAQ2>44^)jOc?Y`0*$5Z3tR@>ISWvGZ=MpNv9x`GmBCBw z@P4Ryj6u&OPzG(=yJK4$EO(Eo#Zw}bLEHA;dTx$uAJ#1MWzY)=w3kK^AKA9v4lUAH z&Ql_^mqrnrJ@8uqht4G;o<@1EkR*mqijK zP%y0!bZc?<3P4F6c}j$WX@#JzO&LyPC?BZzr}92}DS?JRq=n?G=V3X=$0{ii8vc+< z*E*c^hZZ>{_hQg136w~4Ky?lc!oj`eDLf@Yi8Ke)eud3nfKH5XayBNOu9T6 zH)J?IlBYyy1rkJhrrigqN4X_~ z-bkP>v?}U~Rs>#k>;CeT2z8-VQ6cu#aJ^)YC7uj=D}i!p;3z&R2VdngmhzMc< zynX+$bnX28a0b1TK#JcV9AEe1WVpPyJ>V%3QvCj4%W6+gz@c+e{lK7P3FJ=Sbpflt z;)LL;&O9YT?(|*PL-{jKWF+nT%b*kq^oqV2%G-{?anbwxc}j#{(KkaAS68e_xc9zY zS>+9zDuHyUj`4?e8@%d{rty>r=};YGT;vBF_Mb3n1cTm7p!W1C*P5t*56)4C(L5zW z?derMa=iUhI7he6>oX`#0u7*t>h{tid`;M2;3*LrKo8ZwhZ>CnXl6B227Qn~iXQ=4 zHLP0!v}mLfPl=G?M?k`qp5Qr}X6VeIj}mBTxU#lW_w~3ibIlr_5}~2tGz}eYhlR(+ z(S8g{mq5-`&aq_H`~Yat(HNc*A!jP*aBtGNKODL?y%HJpNdgU_+MYe$mUjXAT_q(# zL#VdrOJG-=d3r?!p$}*^$0=J=fa0h3W6&1~r1+WIWqma9Bmc?_o)RI&&(yA6+$#j2 zw|-L?lqrD{s50|*wx)L@0qOGe>q;dIiw1X6qT0C4n~3lR>Rwyc$6Dn)8$h zZJ;MZs$m6AWPIOh$Dpqgs2`1cUyn@B0VvCur$ne9jeA!s%wGj9diBhkLEj|MJ?bAN z@9&8(zRS;fN`&rF|EO%{hFpNg42)q=wgk$fat>4F>$pmHs5(!HP#%?Y%pCc>4?tU3fn4q?Yx}mlg;2>Eo)RIKyEZz{z7N38@|RMjzm@kb$)2ngaB2Ay0`=5nXi|FQ?Dv7QsNP+k5}|17h^==GoCfD;b?c=J%9lWo=mRou{_bi3O=!bYBJ_wpAm3cuj3fCP zW_uV^Ac5A?ddU+R890pAeh*KH(0W=gIeqm@?89D4bz#sC3DkuyhLbbaxk8I9-}96R zb)k!)vBP*=+Veu|9)o^Lpw6^-w!_{oGXc6dhNncRGcBHdZ(pkeoTH|mNeucWfi&pj zL@jkZ)-islk`f^e`Z!5C)^{I3d&&wKR49QIn?+}2KK6nZEiUIN5mIax-DqOx7=WfO zZB|iv!xl*(PkIC`_;n9!P5Uk5DG~CdM^IX{z79Z%;pzcpd0(4;aW(Ji=pkB0@ z!}b=}7Q;E3Z^lz1)QdKA@NM6|Hb7=C92rz9fd6ga*<8`{6OVh5$Xh(v3kC5-5=l-Lcu(-{Bm&U*#zgN~A+q z)~W*MChjTeGU%TKa;2%a-j8?a0`#IfPl=E#O}+JeT5=Ow8dlbFEs}!r%FnM9@153b>$N7Co}w5D1$5|P!Ag2dga**C#rfL=P42D zL8Dt&$~Inr7A5{lW{{NxnnF+Sn(A$DK#Q^pc}j$)(9`?5^#Pp1R-IGIAZrO!kG|_( z+%CeauKrw}5}|tZT{o!38u{kU5>BeAQL|N7sGF_58j61J5dntdP;$$_km; z@+z{*UWIHzNW)%{hEx=xqEaD>Bxz`ph*HXKD@sL$LM6ZJ)w%t?r|WY${`R=tJI{H> z^*k>;v@A1Q%pf}f6bO4PTqtnCYvr%ad5S`Tu*bsw$69Mqiw>ljG3bN9{RO!M~6nY1jw^HuJHV8`on!+Fl0kjax4ASpsH9?SXIZsh&A(R>1 zzVRRa0`$|IG6p#cAnB()mvjE7GP8ItPf@L@Dzn~;Lyz*KeiUtF)n+xkwMP=P#xn9<(`@upEl>9zv^`^ z|Ih!dW7M~im4qptR7HQ4T^>)-ze?W*c7Ev51%Katcj9sOSGfq9e;c-?YO*j6Pm)_t z;wcKZKog(krA+4J77@rDd3i*p&|Du9N<66eez&++aY*Ou@Ug@(csr(%FL z9>RYN{==Yi0_Y7Kx(kJ_c=;^;4o^|&4IDc4H81f@Z+yERO&jj+^8zRW?uEG_>UYqg zJKdhAC=>zr!Yz}5&(JyAv||E;E(oC6uwr8MYQ$;;X_@d8g=WKwN&g2`c%5!fsy>6< z1dw!tps#uZmLX_)8c$J3x?6pDBVA6J$Hbw$u*?Q{lR54NlNCq2OwyxJ5N!lYg1e8-FI_(q87cWR&3UAA9)F&DKK}fC{tO2peH|hib7Lh z?z)9-Hw^@pEgs4sZviw1`es9J58RHRR!ez`LUW*R7Lae{f}s7ux(vE3fb!rGbaD53 zV+75)!&4N>gGZ1_rhzkpl9jeH$VUK47X+VJ-&~3yk4`*AA?bqP828)P5cJa2jzPWx zD7^?e>i4$PA}G|1rzn(OWTX9{vN1!Hyt_vVPn6A9>#C#OG7`EKxDGK?) zbma!!wFT(V&2MeUpc?{c9}IXa6~CQ8(CjumMWKB#;H@fahKGldW(OD)Ab>VU!WZk` zs<$HOlsQjPXmccdlF%s+f9!9T=Ek5v0ptLWAiH7Kc&8qAnErh=^c7bLoK?f&QlbUexGt#{^(Q$$vMAf&@BP<2A*|;lzyK?P*WG4qR<<7 z)&=^iWFe@y<_Cj<1dt>Ueo8pzOB{j{YI%x6l0YbTAD;CSLC*ReTQuCTw*^opypNOy zD&t{Gx&cp7C==dCV=a&4orr!!j9^f(0J;H1L^BLhM@_>&k(~DQ(@zG$59SjN)K>gN|=c! zw@*ihE@li*QRo*`!sHAoOGD7Pvq=oPCxB$3ZlvqM{kY`D%$27oBnx#TZ*yMadf=Hq ziWn3ofYjg|ZO=QJj9O$>#ZweggL717y-o?W$ZMIbY{PvNE`W64$xz&3Y&C}Td5S_h z@MM@*@*7Y4KMqx8P=o;Tg}(PUXH7!{dEetH3i(3cTT^4*N7SO1?WQs)QUF!JsN8he z2So&3Z_iT{s(?|s*RaBP1hp|=%b@!LC=5z&#&(Q}Ku|@a6otZ|Ko4Q7mW${0@EG<_qZEZ6!d5L_ z@tQsedTSHKpcnz938mCVA=y1qi-K%B!j+wBZ_uGDapfrrErd@JT1VauM3DI}4F)|BK&eoS*SD$|50;aD^Av?rp%`yT zOd4K3>$G?=gB}VX=_d*67uj~hEozjako1#;_Z_Fw&;Mw@iMnMcp6hPTf z&hb*?NF{2~l+8Rvp=>DUkd!QQL>EK5$0-bYEP$l%qe7SY;}JCV2~SZ-`aUu{m5!I8 z)drO@=!pR814GMcc26@9)Ip7>DAWgrmg8$~?Lsa3;@GNX!+n$_fJ&hzVa~|g9TD`# ziKi%33N;DC3_CAD(1r4W40+~1$6oocH{gv9jd3bi~c)&&mB?};5=&5RV9T1NWoqixsQOFm1s@G@A zO-6^Vw)t@er3j#z@HNNM|DvoBl+=Q!C^Qqk<|y0u8Bfvl+2q5Z=K{zJ2Ft!}^9CSD z)`+JlW01bx`U(DBTP6*OV;3*0XhY{cHEm`x?psq#9wbJFJpPf;iXI*`qtC$>W^iXSzUL74)m7`71Ue9B=cYEe@So}y4OY$1~P z!utV&wma!EC`$l&!TZQ}avwajoP3I>DC7n2qZK(@@WkNlk}V9%7C^P|mF}d}S$LZl zt5Tk#P%V6=yTIgR57eSUO*;nV2q2@ja+1ZpBk-#{VG2)C$f&KY_R{~<@KslE`6`2Q z1&|9AsJx!J)CQfS$38qoAr~l6sp+#E?~t=c<`IMP1W*<%_-51;Pejn-COk!&4wPvnd` z&Bn_*m&169LfzmKIfvKdP9jLTV=o332%w*E)%^(a#;=L8MkxyYgsbl4>n-@jw|JW- zg9-&u3{++o$=S_7hpz8-o}y3;RA!Fen26s;0Z&#k=%WCVE>yWM%DaFdizJ?+kaVHy zd)0&4=+H^}9AMBV0dx+=N6i));bOdjeR+yP=U{v^tWFt!73FT_#-Jhrlm>loxzWS% zr#*JZd5S`5(Dyd!wHw!*466uZP_Y2oe;GF1upjP&4qe|$o}$qH%P{ZV;)yvrbh8Xz zGw8DbY8wno2I^N^Bgl9;Pf@6Cu#I-barICHMMwT%P>BGN9=Zo^Ww^XB{XS1oNP6gw zmO0^B;$i=FY~66fmI|N=@FC1kZ?E;JMXh`C6on?hhcNFpj2ensbnegy29*h*Ht=N7 zI(gk0L2nQ96ouNrlfipn{8dZO?JpQGd!JVhZb*j&ggHz*H5Z&xQVs8Rq)_snds zyBI@3Yj}!6(mgX}+9ftc(4Ckf27MDiec%nde3@+!I!7U~JVl{C@P-|nn~Nu%?0U+! zX}FKR3!rcqwtVWPi`S3t^x`QBg~PC=+o|dPsE*OZN|ixBa2?~0;IW#4rfULF^Lse* z|NPH7#&|c_dN+0sE?rA*l%juCyc_%;Ze|fKB8vQ`&Hkz?LG!KQ1K76OL4N2spZ(5L z6tad7U^gz#i9jv-JAW;Mss+#*nD+O*^gsJwl`h~Z3ax=@{}lnv@xr>!y(0|zDS%X= zMV^sPxQM887*A116I5q5Ih=MWF;J6?jun^Bx_# zTjo&=`YnLUVVWh<-~nESerUl{6e@>lmL508=OZX8jaP% zOu+t^dZIUKk)9JzQAi6WU`L$qSc6)0_lpLD{s^EI@EA!+9XuC7hraR@g;v00q<5#u zeGoKm+F}OP3m_x-%x!wdvtJO@N{6Q?WCWkNy*gcEjG*m)W(@i(fc#+g1b1D23hw1_m@o>@MK7nIg2Y{ber)Mg)-pDaL3LPS6Mt=8N{HY z0!Uhf|8f00+}rdu`xgpq^dJ7-S`Yq|=rDijVNtVfBqt6p~I?E|%QKrPR|bTD5JskB$qVuu56Uonxj? zQHuuc;wcJ+Rl<+x_42^gv-K$h8DuShP61SR=l(YYr99^;3Y`KdV3qA9bm;o4&0vs? z0Qv&u9AjgO@z-a+8l@=o1%JdGEjne-Qxx)q#iqes zM{h(>-e)TY*$JRhxMBP5Z@&RSaV0!Op;EYE9g|E(Bk0v!9|oNeK*6xsG`3x?9)c3* z@f3xEVX-N2=gOT3+7S}RAbSC%_CQv0`>7vZnlZV{QxsBr0N>y*nt&_H4BO{2=%fHr zhWF8k9}dk>i-vUIDGDjW`)Jg++jCHhHkxi-H4u-iCz$aRg+k#Rtxh*R zgP=CgySHn&VI2if08Fz~%8lQFTJ*h9ib4S}&2s6c&2|K74xhjvCjs;wMr})bPQ8Yp zjw5)ALf>K3_QCQT-nVCfqdtR937~T@%@X)l?Jt6A8l@<74yIXBe;u+#EqY$Kn?a`q zQ0*V+K)$}&3PD#s@)U(?|G@SI+o!l8$WY?SAZGz&0c${KGlt{g;V>0rt}`xzb{eyw@Ipfds}8or8Z|Dg$&5i|;TLwv4AJR0xH~FKQn2MUX+bV*7^s$W;K% zf=x3<{p^d&Ir>NN6oqENrWrE3cIF`Hh|*97ofAOP1woB*?eI+R>P|dGA?bqP?gG0( z2)bsb%b@cDs0&p66(pX(H|!B}o}y3}sQPQ-p_+go&zD;mbU^^Q!0+KUzunT8AKG`>PCc7eIeu z1;4Dv{o4rId6K6n^cPm}+fH2H3qi*}CosrE0Lj3{_{zUJhakwOgr_JZ0~_PF{k#xw z)nYoUfI$}p&@Y(Wvu-AjH}sw~o2MxB3nuq!@+SJC7WKW>q(j4fbV&e7tNtD>bHwvG zI{rLGA!*g$^?Dt91l1|@Vvwf*(t0Q>N$eeh*Sm{b^Av@&9>QK0H)o_HNXuN4L0$r= zPnfJk^;^~obdEY(@DzpmguyRayl-(FL37hqGRRv1{el zwH|j)YDOGj&}9MS4uj?OC${6cpuCYhMIm<>EYCYR3{TO_JnhCH9|2Sc6B>OdIKM!L zPR*I8C{za%8U`vw5_IS)zJ@W#R{&)}-}}l;HQa$rD(5K*WkKJ2oz5A&ZSTU_uNmYg zfcnFWZ=qINC)6T)U7n&)e|Yg7DZiqMTJ*yI2ZOE%pmA^?ZOB({gP?%xJVl{#a38hT zSKN%C4Ec^78*bRE0_Yu-3KUK<#^a+Wt$2z;@1Rt`a-pXyf+Dw#V9+%I)CLx9wNCe0 ziCR>(ou??&1{Q7CzdwijbvaMwG00y4^@b7OTeC$i5LA@JQxxhABfiVEI(S2G-Tpfm zbX@>Vf-i2u_s{K)ppgT3ib9j%i<@V;^Oez|%d~M|&Mkk0!Xh$Rrs3ZV5+f3-th7jLxnS%;@6 zv>xiOKK96QK`k13<2{2y1<-$SvXV^)EmBd7ItTC+h5n1P(Vit2j3;k%<$p8io&d6e z)s_YJGNTX_-HN9uWCN=$9V9Zib5d#4rBlNV8zzAA;SIY*#vf0x-`~tr6v~G;Y_R&u zedy4IKh$7QxBx0Ig)Ky^4Z~53tP*&NLgl4!c~@5AZy%m_UCf{e0ral9oMcSED||6{ zbmJ)sy=!i(UHCf*@6`Hkj~Rm^1<+EMM*FG|EQ?wcyqBjav=pY%w!WIa0Uf&Z*QXhD zUjUtl@sUIG2z)WzeZx}}IuGNc4By=*2=W{q#GohvB%K&kd9$bpwdjZjPf`kHQRFb#AziVg*n+^y@6t-0-r_$yGc>p>pWgO za|oK2!&4Nh{tIj6!&5C0q(0GyK?wqA89Y=SIuFW6P|Ha?MWJQzP}Q~Sz6U{D}b6DFMw zYGmSzVaeh@40oSrN0BFOkKPfLs-_43%H8c%3tR26om|7 zS!etZi6(+H$GI{nSpZeR^5|^)&G_l9KAxv2R0YeUlUDD-8evkM-we-c#1+jVGNs;-w{{Qv|Ulb zpi}`g4QBF>Z15d~E{4KJDGE)4nS8%)y0~9A>z<-=!+n${fOf&!grVnR{7tJt7*A1X z7pzSj>1TiokoqbNWzY)&^a=LPa#Ws!=j)0ar6}|X_Rd<9skj%NBaLmk3`!S3^WYqL zY*E2O%P!k_ibC_?9BsVY;WL7krfp$Rh5#~$qSh5tLp{(rTJ?gbC}a*rt!Alm1_(Mm z+>Svn1yDKMu!(y{_eId<5j;hqa=2kTc`95-Q0l3x407sNNjJHz-*qVi)iG)>C}7Z={%AkOxa(IZ zpBl8p8vRxIEBJr@XC31VKn>&2jV0>}uK&op9Y9!1dT?mR^yBUnCjHXMt4 zG75V&8T4KNoq^A;kL46yLD0`eDGHr|&#p69+-ZSYbnNX)24xDM3(#d+?ht^#$XNZ3 zrzms*x=hcG{``xen6U>KlqG;tU{WHj?0@cs>Tx_pp%j>u7&qbuE{n=`b7N4p0Qv|u z33B;;@$%VScb=lqN2p22GTCH@T4eAmj6pd9s52CJyVW@16@0zlJVl|-P~iP%(i0g3 z4KaMppj-ho5=I2q`nRt_7ek9xJVl|AFe3PCKB)>pRS`cJlqY}=!S-%3D~m@WC@qqw zD0B$6cYAv2ivxljyL9Z}%2xDWNf zW0jVoi=o*Ho}y63efVmoY6U(=h6zOsDi%P+0nlYyESZR)8IO31Ld5|#+8R^Oe?ToN z?6;Xe8-fRbRv#Cn)re+0em&r=jif)$heoahP!shv<|P>BFCf)8L_Z1&<7)ip{{ z$Ot}wO*ne-8-ld*v>8+?fL6-LNuGNr;GT?5K2K3-rHrk1#^Q>@2wFE~ErZGgkOR!Q zSq%;vh7R2{2~SbT0p{E^-+Ds z%=AtjZ?1u$l13>CeSw+YA3d`%wBtyZ9t}6_HvuI5gjcocd_e> z(K*V?(qPbc0W<*WMn+C^EkjUxHcwG#0Mw1B=M5}CkcHM_2K^8~eozm*Hz^_sLCdvy zib8%+5A17ZItW1;zGe)n5Kc1pc4!kBb-Ejo0whVT@H z)Wh;4`fiS0J4PQ z2Ni`&cnv7GQHnyAQ2elF?i4&%5I=DSgX#oOGOTKS@LY&XZthRwDGDXSs@Ck$<8V#F zE%%KK`XhiW0U8`1QHc&+j0aCq$P%FUm2dFnt@6!^LG=Pi7hdJE``7CrNaZ_EQAihF z<=18{SdY%pu!TMh`YV7oLrsFa^>#x9SuNrz3T=j(gh6A?BM}rC62~CR0fLW}Ub1s993ZO!muKcuiM zrzkWw64p}3ZqPt2(w;tnLB|D0cK2Y`7Tg z22AMIvRQ}D>ccFy?;N&qr^zmhVcShMY}}^4-gu`)tNd?xr1(r`$^Xy&`~Ev17oe$_ zZ)k`9z-vu-ivEEIcoIxi9MkamMLiJoW5{g=ofJURp+BgX*tQjdUJvCd z3QdRp;N4PrJR~hU@ti>p0_Z(lt;rd4J0U2_o~J1E9`k7ckR%bl`%l9=Ixy{0k!BzevjS_H|%)< zZIUOogLG8ibDCYuW-Ws zBblg0ssRQJauYzAG4Ko99W?PgUtOaVg)(DcJFDi`Y7wN_axa721&{1Jt8FWzq zMZ_^y6Oz_7C?D0fODyCqKcpidOSs;JQ%<=n{*fV>&pH07<5?x zU4Zr5&c^HUR9XJfeyH|n##0o!_zP<5 zwBztJ_PLEG800H}`a+@QNW-zo=+HTB;wcLCg+fcM7rA)sTm0x6gZu15oZlYJ8zbVUGlfDYt_16|Tli#DqA6oooK2eQtl=M;2~jvOvz z&{Y8x52ZTueEw%barF_NqEI}P>g@e*-%12s$!glS;Xb-1fb3vmeu3K0ai~Q$**rxd zJD8Yv{e2p*NGnNtGss^6ZGeaBO_y4{SCE_*Pf=(CJXBvy^T2bletuIJbX@=$Jc953 zHE!U3-O(#NMInPnHrjUKpF5yKcdF$o2Hg-qbD$S=bap+SUCx!~DGJSjUeJ4kCU~AN za{ECB1qh%Eu*M$kQZO2wqYFEDib5A)jlDFiH=YQ&mgLT$KmoKAMiXy0|LlOEy-#_H zLQ7#ZF*Gj}Z>ADBD1t#Z1<)1vW+L!R8h(1)sPPnquD~}F%7^?s&^emr@PjDPd5kQ~c!zZfy*W&7RgKj)Uq0jGOZE#OBylYF;E)xcY2%s6T z)L=Dt7d}T0yLpO2GhnH~e6lxw1f5EEWYAp!G!2#-l(OHoK!7C4z2`2Hg`tV_L{bdOyC4d#b^9 zJVl`~Eo`;_x=hA(dD}k}Gbl^|DZs@Dzm(_7} zg$tmqa38hH_!EQ<-6CC{qR>{jkCfxwaF_V(jeZP@5J1V$_l^qRJsm+819*x;$L1l)r!gpU0IGkK>FchkIduiD%8YBr|MNfVAJ@XypHnuT#|2`y+w&CttJcET zpSk0*hoj^CZTC9%SKSvhe_^bwWOTk8p2h#Oho>mCFcy9w(KsJJfp@*LWKfg<>H&}X zcCz_X(Q(dx&r=lYAuXdeNq&b~^kOpn3asB7t#;7@Xc9c?3vK?}hoE8I60g4wuesFi+&91^b>5h)&DGK?)-EC!YcoKrr28?CU0|8_e4Ez0T$;2Q0Js-$Z z6tW729jX15eG#ZMLK0Y{vM;>w6+v3x%o&s*fGVK> z+)dIa13|OD^Av?Dp#R);QgjT09xQ-epQIPVBLVanws=_`+7kCRlNRz6g&xBeFK;a} zap~yPAow6$3MC34V;Hsl33=BHwW!x^o}!R3jN0~&d~yY~$W=9cBAER7 zzZ8Xr!^JS7sA>p;23pNzk!J$P9y)>J_FTtMb)yu8?4c9rEN@bZpc}cH7?do4h5+>B z;Al70B8xnpqR|w) zF2O{O6nZXz4#5mnzfN7P5menMMWI75L)F~4fVlp%l;;p66%qPzI2``jo+p+xw&S?i!(61o`rJT_p^O9A8$MH7`W%Qm4F zwSK}=6!M3niFY+Fc=qG@fV~WQC4i*$kL4Xslq0BcAWu<9TK~8%q@5pXQHJd~2E7(Q zo!`RmL}@5_BB;)erzq6Shl>GBkXE~pXYUlx0dM|)#VAsP3?-R$P7G*R_QK$xXJybG_Pe4$|&FT!w z6hP_lP`x)`0rQxuAXS9!9@QoNAS(x;F?`2uJKd|^tPvs{H(?J zOLw=eJVl}Punaw=Q!Sp{dzLbVL4^XSXM(IGXy2S^s6`i_^Av@8Ccy7S7>~ybf}aMi zV$eqcB!M-c;a`n$5mEIZo}!Qh)_{~225O)vcgYb60b8_7oR3v~BU|!VAXJ{chbiML;ib4r6FM4oU>+YyU7bZq9s8|4HLU+r| z=gk5H6;0wP3S~lft5)9~&t2PlzG2X30aX1ICifKIs3FMEi>D}5{S?MWhvpZe7M0aj zF{nfU>3oFO#8an01f|yT6oqtPy?gMC*iQ&bS*bL*;f5_0K+_e3aP<-U0XLrJSuOkynsPp1klhLm~_%k zYl&J^-zY_)p*67CM#kYg=+F&4WWu1Y0!SU!=`LNV^hHps!#qVHb?LXT%X@?&=xwGW zgUSVvA-pDJTz=#6(akKLqL3lHCMLeL!yD@F9v#S_3IX)`7xaQ^bOKO|rfKjLg0_JuqxBZ@L_hCU)xZ6ovM{u;sJg6?`%5 z_vy!=9|Fi8Iwy6m+V4<{uKMy6h3uhoq87XW_q{X!PGwM)0CIpwkab8v69h%cO#8nS zg&g1!^uZ|P8#+fdtJkqewE%L3k{i?iGNTamYzb$%j$6sCD%mo}$p0B>3XZ|Kk^Qjyi@a4{f+%>jY3G z%;XPT7Fdf~^tVxpLX|L+pLXIQ?gb6+IF>6j}&3?7qZ~;}DdXY|fy+0%#>Xy>~46aTr0D zQh16&E8*!a`&t8ci6aNQFvxP?zgyG<=Ic7y|4$K-%MhNTP!pK1bM?)_1^fqWZZqhp z01AP=w~>n%uG!YJq0=aS&Y)ugXgrLMy2NKKLWi#BXP%8^l%mjS=rcH!eVmG*vN?GS zIw^omVdf+@?;xIZx;K}nC}avVCzJJaaS7cwen30JVl{a&wP?ato}$opc=0{!G7&#iYabXe=#&7u0_(6h)%)Qk?AH%@ib7Xl9oF;5i0$am zjaS{vpwj~A7d*WSBuj>&7RmPGDGL39r}u~z3r-=(`PexIISZgRa4{(8MlM6p1}mPT zP#d@yF7Negiy-5|y9{y>K$oFk7i6t~i}BWd1 z!EEUm)S}R7JVhaMSmUdDJasT?(I%ft2Avf^>tTX@eR78b2r~8MDGIHJ3HG0hx*tW* z_g3wOH{3_A0_gf(=mqJO;wqYK1)ieN^}9CO=W?EXMo|0?bq2wL&_DZiw(y4a{;3#* zTI6EFQxvj=H*Dxfc|2JD@=T9G=LOJXc*Bl1*@qXZQj&R!LXY7M+iK($1Jok3LBtW@JE3f+Q*a-%?37|}PGI(6kPeX?;){3VnlnGCU zre0pSys&@HH3qp0AnEU}D0$SpLeTd{DGEt{cO~x5CKJ@6dE*ip_Sizk3t4r6hPt73zGb;i9{_bYm}l;IP`*CN;UzTsnvH0< zk1h>Fb&Tn=wrY-fr7#fvRpko&KmYSy*m)A>+&1>>jQ%R$);vZ3s+}j{Tdf7n@v4@@ zqBr}iJO#~{+=Km_>-}&g%&=WNMIp&OSTU*DvJbUr{@W=G@)AI*k6^$%A+IfJQP+1o zMIlvKv|U@U9Pi1bIC2$(yakZ-%5m@B5`XjfrcsJQ(ksW=vtlV~(IA(D47x0U_Cta9 zf(+wbs71Zc@DzpiLxJ}=dcz<;0aOI_SJ&%1;x(XOjZzdU zg8Hi-gEzHDEh<;2X3!M@R0cJ2`&+9Rp%y)A%~KRAgBrPAN9x}oNM@7L$cDT7ssM`k zCMy~9ej%QKecUKTp@?trX?)(Ys|c!nGLk{p1kh)AUadWS4KK~SPvR*GeTL^%mwoyL z2r?M5fIVRX&@-S*luResQ7o}$oVICPn|pKyzMhdpIbpa80bSwz2i7gG^b*C<7yN|;65 zW14`W<6S>9=%xT_2A}kX{b>FeK^EP3ibBoclioVZ-xJZHn{-faRKtCAO8`Z|RuXW)VLJT)l@N-%INl6fA%q!t&Xp7M<}8>)^>#6nY5D zXMRC5w;<^CPfG^f5kRWY+dPvv_8dBN*M9L7g;b%pS>?0p5rTRy_F_ zyDSEU3ZV9|3|(7tpf7@S+w&BK+QTyRnf+5M5tP5HhC%lPP&(9&uG~7pz zCr-l;)g>c%ibA8|J~FesKLIm*`+0J!g(0u_k9Im=bw`KQHi}uXpDGCjT ztM1&Bop>#Ef9P!nMG2re@KF6(U+@q?8}9KGh33FRHFuUuB06*q%Bc*B7C%4v&4d?Ux67u#5u~^03xi?=kP1xhZN2^tKfQbGJIU zqv1Y^6+jLE+4$%WL5Hs9HBV8<0iXfu%SR$ee~KD|;slUW5Y+bAcf^&M+7h0kkW&!+ zd{ftfP6&GAJ(EH40>}=QW{hfosiGDoUgjwZ*}>Ax-ZxwDJ_(jGMhto&fTTaF9P-0# zGJ;k&;VBA9e^go9Ak`4HXu%q520au&yJ0qd!h@5_2NS=Q+?g%0fQb1pg-^fAUh;^vk>Inm!~N72YvwL zv6d^IQeI}6$Dl+3R1LH7Z5&^{N6?_7JVl{un2rA#fAt|cbX(unGw87ZGOdNru6Jb= zBS`ZdPf^IU7Iq>kQpK}~FEmxgG~BRH1W*ZlYI?v>^$>yrr|=YoO5jt|M;#x}LoM3n zF_A$@0!aFUR2d6Rf)S*3k*6pm{Xwc+)Ah>{W*0OY;8*0&VQ=X#GY#4}H8T%eUEpke$WKgOAlKzzURuk871Z{i4QxuZ^ zl()pr3BQjz3~4vE;XXk4wFd`Hj~Ii8|W8T9L}9*%cFEjqY9kwLEokP}P{ zTIZSNAZYOho}!QwObm{7$-~b&gSbKly%9jtO*0Pf?u&;ltKxZzLefn$IxYN=ZC z6oux&?AAM-kP!%S%$&lY_X0?|3+S-Kt9UNRD2t~kB;5t{O4OKR2)aCJ6@xMb(9f5$ z5@SVIT*v4MPf;irE^ouNDtOZA zRgF7?vIUSlY<=MNcNuHIct@}7Lp((x1z1aMw`W5- zf_i6-WY7lzv>R%BIyGI3yIV7|d5S{2p|-~>vNxV3-Yi+bpaKCj9m;fHH+LA04&59r zo}$onDAV1ndmC?0k>hE?ph5u@1>LRSH37JDa?gvWC=>RGXgWMpFZHtaK!>i$8lIxiba<$~+B*3L zYEgXjQw9|YAayu&NwTJX2(pXeDGI5>p<7|B+#f+RyMAU+u>cwl?;}ggTa^gv+l{9v zG#uVXJ>vuM3jV3XauXWvqt60p1$+T|qNNAkux6%OZu?N4qL2oBRzD~9 zAYSnGKUu>d_^SM$qlww@ny_#V#cx;(2cDwPYsiT>$yRq1(}B)mU_nZt3w9h5X^r`2@Z#K^Maw z??SD4grAJqt;4mGlp-nYE3p+mRs z15Z(?Lye90@eMW45TrOkjX|{nC;_0yDc1K8)N~?GQ78eR(7&m;>@CD?CWGn(P+yqI zA8dRdcZoyXd5S`PVJ6?y)vPOO(U!_h4EiI0R=_Or-t4X^s6}hP@f3wtz$|g0?y+11 z)y}hKP`v<>{$lMC)p_`#D!YKEC?x&G+Uwt#&!=FWH|`cF*Bj%_Ki?_4Lg5!fdaVva$N6=4o}z!% zu286rpK5M|nt$L(9{a0~4*K`Yt1mSFruoAG2=ce&DGK$4<~z-{ZGj;7qIw1$6F{$^ z9QvqdpGyd`FXkxv9HJ3n2OHFlsCQKZ5%1;wcKXf!U8Svs$-9(C74X46+qK znJ}2Tezg^zfGx@3DGFu6V5*DqHe4g8IVO}rb^>SxY{*c*dJ(R_QW(oq6j}irGW1Lg z!gV8U&SfwNc9Z$%4I2hyoLeJT;o^tNMkxw~!5C-xZtr?@=(<%@GRR&49sMpVIg;`Y zPbsgi3;D+bm%5s?9HIF0!S+s z>PDPr;ZLY^FYy$Gv|??vZ%wMe+v(hGE@6zQWS|y(X~9zzs=95X-CunI z?x^2gw~9gM1ke+BmAA?_E=Q2{dY+=t6L^)cp6r7=C%Y04GU&VjN`du-+1&%O5w!9# zPf;iZ))$KBt~-p*(VIaY47wnI@}wICh1|y#WqxWrMWH-+9|d2o!_)r#Z6g@uCV(p8 zs?*bXhx^Y}jZzeT`yd(AqUU*U800R1{NbT$=;NA;4&9A>o}!RHJXG&_JjD~R ztEN;l$U^{?K%rN+p3U$NOU;$=6opEl(CgQ$*SNcN#Y0KbaKl~{K*wMlntH4yUJwkp z$Ws(L2IJ7U+hxnpp|h+V$)HODXakfAbnqRAe-~hO9Zykc1C$CpKfSvKK}`)7FvwE? zNf&LK?if1|ouglkQWTOd+P0E;qJW_PVoVt1C4d$@kd?ffHup4wn#b}Kg%-e<0tWdm zE(mJh(}_Xe0;n@oW4&%pyFeJ47x0Ud|{eJD}V7j)S?tC zo}!R1OtS>ekjDion+l#X$VULV!Q@^u^D#LHGAQIJ3c11Lo*NWDq86Q>{+U6(0%%n- zY;1CTW+w#gn!!^PT9phdCgo!bd&{YA{3?4yMV>R&#zUD@r zqEIt<1QiXI$MwLI64x;Zwv7DeXktFhB6c2MjjR4TKISP3&4*dUQ}Pk`YiPB;M;YWV zfU03nn3MaRG|-__RpluPRl}YzL8HoEpmS7e>&2k!0!S8~4E`IBK0r{k9Zyk67M=`o zd1EpWwESZ9$ns+}vIwNb?g|MD6|!NK@aRg0xN)V*IpQk7!15?T~Q$1YK#Sofyn?ZL3 zkaj)P1F!tE2SL^^c#1;W^-$>5Srb3&EC;7DC`14S!8b%LbvK$IXyy=}qEHZgLv-W! z+kU7;oov4_=&k@d0bB5WxjUyVf_^kgQRoD0!RKRT+O4}Li{3mv-Me>O2FTmU_TbM$Rlucio^SACxI`R#Q!>6y=iJ(jRJVl`+-(aJFitHZhJhTsj z((myUg`}Gbc72ln0YN%S4;XY`05yl#L@&2%ix4!j6Hig7IlLyO9#Fww3doq}F(^s^ zEra!L^XxVllC|I|3N3^6?()+`xE?q$rJh020;me+g48?d;kDG_=R8HBDwqp$FSs=Z z9lGU1Ri-uEurUH?IZWR63`#tK4&5kqo}$oln7ma}sSiQW!;=#k6f1yy;XWGeu*wuc zE)G0JAz!$U{7akT(Zr}C0|vzjpeoprpiSFU{GC_RVxFQ<73@ghuQF^UYSBjBy$p&M zKyP64nNQ;em7x~R)#E7&y@AbV!ulG-At>O^IR-rtKsxtf7f}1@*AWyD!c!E|xo@ML z?V0=*L6)6E8T3#9y@biz8^4nAp)=^hQxtj$lecRV2R%j5o}C#CN)SNOeYX8tcifFy zwAPfTC?wry+j`buya@6xt&%~H1ds(>-e-?`;`Q#37d%BF3%IoOM92GQ5QAid(2fkE)aVt7>V^$b5C`kZ~ffoI8 zHo<*{CM$W0LSvvs_V)s_5u|ee1cROmpp&pE%EB2%_^KNg#ZweI37evr9@WN6*d{9e z40qWKgmIs)nr(wpzR6UxfMIC`F-a z*!tkpX&GE9ppjR|pcDbr9wszCP2W2WwWw=8Pf@5nOlTy|*@#CIH>Wh4-f$m17eGtk z9No&f;DjJM2~SaI37n%BC3geRq4T)dn?b1p$Oi_?sxoKZBFOv_Pf^GR2Fr^3TkS+p za^(~Tr3s)e@^X^UPPg%9(Ko;G6otCT+iJ_aU%MAUMsrs&=!F2Pg2`LIL75mDIgh6( zR0Wf_{k7U&M^H`RK?bD@AnCTfsq-W8YvSWgo}!R++ur%}OI{(US9=czWeA|_P;0s- zQ^g#eqw+>63SEa<)2RmDczmR18o{8K0>}j}Z_{4M_+l7f##0n>fy;a2lIp3bMH#7Y z81za2O@r0{t*SlSpcZAM@f3xo!Rr6*tkTU0y41g#L9Yc+`=)Y|E=MDpBWT9}o}y6u zrncIa19syQ&==)C|cg0+eFt}!psp*wz=rzlheYZL3H)SgAqsb)?L$`n98@S3nQ*Pe=?gUxx0 zLO$@C*nQ^vc?4;14P;Q30E&j5s=?yrPZ2c1n5QTd4L#KZ2|sY>MCa*K24xGN6qrWy z>)?#P-*5knrzn&nEqA}S4{sK2s`{BhIRfY-43^C=?`n@)q}`9FDD)8q%h?8|N$Ak^ zwvn6Ja3AFgpd6U@c8$5Y4nfUqd5S_gFz>Ckb}}A_g?#GApgaL|3l3fElZrS5xfk&i zg>J#2JFD=e3u;k<-ZTc~3m_F(l<(R9FWv{^<{X})kP0lycUssVmw=wWwT?j_1W-TN zsWrs84u9>wJ&30$)DL!Qy|zwoHQJ9cs>4wR72rBXW7BS$N=v6+M9uH|AOFw)91&!n zfu2nH4P1I>J`8)*Wdhis5^x?g~Z1m~Bs6}IBy3T62 zyNd;g7u1@U$|C5)s<90EEP#@rgYoG3h6f0WTg_7xN`el? zV&_Tty{{Czgh3?&C;=WLTh@NkLQp}Y6onGtF=D*3PZVm=r*0MuDiuIwFrRZWrDqa? zest$43YEcpPNiSd-w3+1--SVC0;nCV0gWg(#vl7%Jit>FY6ojTN%M!`=hemR+YI_5 zfY!qBaK6ivEvQAia(IeDYhiep+wxf{YEkb=sSNrmfHpu!eS_jR+<%sz%u^KF03G$e zUt8d+zs+7>7*sBRx;==u*4o45}4COMb!b9PJV+&^eOnz*7`j@(VUY+meMh+G@Fb zFN5j?(0_68KDyi*Z!wYAC`F=0M|XB!#WTI=P3&H4u(kyrK=SPQnxW?&~X9O z6v_*o4`ktwE0t||ib74Hys)_4uLB6$PTxVq{~Fp>s5722W8)I#&=-q_2Wnq#2mVAX^a>4TZ;F%JB}9UN?E_ zA5gT7_Sb#Ycvb6HhmQ=h6F^5JV20|A8=hwAq{LGcIvN3WBip~?F7Xx9W^)?uqZ0y1 z`uSB!^-+8u*_rVag`}TfealYA3+tEDdNasg0PTS%LrpsuUv!Rkz2GSd?SUu5gX~^- z#22V8VbDnd6y=(9U?A{ulqR?F^JdW9`ii^WDR(LSTNdVn~ zi6iZs_IN5Rb0tqv=oU;I?P}TtPwwrCjbP9z0ptlSYW~6lPtoj;<0%SxLW?|NL-05F zM>@S_&}jkW)m%=pYV(yf=&C!}nWre^)!bIwaMe{@;Vf%b%^+t1WL6Brmb@vw5%jfD zib7_^Hrk`jo_V6Hu6df$+=d(0MF8D_+Mc$(Cd@`qdZQGD?m%tNi49|#An5RzQ4BgG zfV^OH&h6iQ3J|n%EKgC$3pVGh@eO^Cp#B~U7<5(u=|h!8>+to{5!Cb|PfzRuTm?`aK$kbw)gb7&%+CKyQK$}}VaW&ABWR_e6N{V^K<))lQD)~{hoG^m zc#1;q1vc7o1FZ59^x=LWgU$<}!O~}4mi8wEJ&xij3Jr#5-L}h%h9fAw=Q9Rf5I`Z& z3;NR19Iw;e?8Q?Q3V~iwQubJUA1yufnL%y>C<h3}*!2C?#$O29m&H>Q8Uc$<^&>3t%!$U7{tWUEK*Qi!=N&AM zKPXd@@DzoH!LzQSeZTwY9Ch=a#-NJ==o*YBmTtcE2eoMFWuBtYH5g57pW=n@BbEAf z47wzM?BJ?f^XUklysc=IqL3Y2bsf|5ajoh8)khiRDS(c_HuC)pwej$O;ToQz(6J1F zroY8oOvp!iF~~~*4T5T=OV2(yqC?mAK2K3-5L7G0SDaph&QZ^9F%0q+KrXO5N2sY% zCW7i4r6}YAyK{`Pn1pM4lJ{jX=&}GZfi6uqCPC}aeMUY+Ko;Z?2B%o+yy z2q0CM3krDA+7Ll!vv`U^sxTMixz7hzl-(HDb$-JQ>nni1K!Hk+mt!{|==gY^qRqPkP#f!F)zn3uRiU4wg4LA19 z8;h?xml~d;kP~dUu`gpD-UW2|8Vd$p6+jWN4E^fr)tRV8!`Jc@g(6@X`t;0gcws#w z)`daW1kfP3>b@MVPDhY`98Xbb5L|WZI~;C~&XHWdUwLGh--cS$H1`XGZU~^fYS{j| zSJq_&Wi?7sD6blR)K$mE4?zVJS}$n0j{*dcK71+g{ejI51VvBeDGKSsmjdtejGH1z z^}HH`0tHY3d=;fS%;6b=>KdgeQ~+N^*&4?cBIs%5Oa|Q)K>F~dfW)gKUNU(6ji)H2 z4_^utHko?^LFKwe47w$N+@MVNY`Qrfh`rb2DGIqknXc{gHn`-*DA<}oK>|n)mPg-O ze!#Uoqwnw(h2&s)RIdB#J*Y)1Tlg{Pwg4L0R!-v7`^Hgpjt0x}6op2%wbgFgyVF|) z{kP@;gMtOnYWTG0z*XlV@01je>R9=O-0gBIwiFPzFT^ zpm>1NpLE5wJrV18ibC-K`TlNGj3Bj`3`Tg+)1nKtSDGJ?!wQ>_xl^&=?%Pre2YPgT01dw!d&Y;pa5(I5K z%2O1QZq7Mrgr^~bLT5P56Iw-FIBh-Tw#hB4vb76e6qKp%AiFLPm%( zZ!=pIGD3)U6qQm@Nkc{@4QZ;B3QpG9e?_BTixzGLg{dzo|PoHtd z`6J%Z1~kPnUcGx1F%#X^tZGtqGLR4jsW-;5y8Qx2pDYZGaeS1+U6yQGk&SV#}n zCM@gLUPg6{L#AjEC_@#U$GGeLr8$8HfA*k%)t0&RfBw&jBh?4+29Do~yR|ds(G>fy zQhfkxsfp5d2$BxiO#W4woaQfuvZz&hC55Q@ts*HFS_)-R!Kcf75Tw_`l0eTmP&zC& zoev*%8$lC#(i97&!(vlh*7Dm3>VCk7Kv^6}9%d2ys(0XtqkacziiPB17E!`C+6_Tn z^Wq4U&4FgZMv2Q^7w4iD6^Nu*XeMlwko*unA3@d=iwKm%f%?I!mf^f2sC zKUmc&&UD1x+RUr11bWVaQc_{>@XD{!2#WHcDHckBLo*!9Yw`O^rCruY_;kPEKqujt z@slCfEfLfpl47BgaLl-@_Bcm$>EhRG5a=ZbQh-@RnZNnl5OjM3O|g&y%py*G`PBhI zuVM@cl*@s7!-|P{#jQ06x)n=PEYurTOtzRj4MotHL1qMc#esUl?8imp4|r0dn>arf zrzsW^oFnJg=!)07f3H?1P$36Wga>xfc~@MfTV+U7ETjkztZVQT7j)^&qIC)MmIJMZ zF4O$5rNdE+^kQg=g;qnCY2dIicxh&hoH2olIM86ILVvS(q&wISn@2vov>a^R&pb$cEjOkE#OQ!JDNFV%DMzt5r;4RCu;p!Xc8 zCzR=q(m%5Rwdk8jiiLVYneM2W*%=7(tZygK2M&~72M0f@%DN%Q=Q~ZYP<9;@wI071 zgP?iK71jzL*pD2@40hUzuQ@&qK|@#26bqTbPTSqzp5wdjT$m<-K5?M=uu~vu)@mmN z9S)}{7Mc$`1@aFy$fFh+$*dtzDF>>8a*qCGb8+Wnem|OGp(-fnnEdDpp8Ys;#GF8% zIgp^IIxc=N-sZN)oTgYv&{Li0zDE|d=>1ND(K%jCCbQZ=qQ-*5dV!R<*G{r(^VT>bd;*Mt#H+tm~sDcAY!G5~(6D9b2 zG|ro*SV#)?)2W6Z@u8FF>|kH&#LWW?QH%8BR0;Hz1Np+)N~tFL_-2^?n5J0B z7tU77YSTQ1phbfh6R3&<^@cT|70;u-q85!CLQ^c%8`gk4CO*M`qu^t^2~^F2e!$9O z#ES865OnM~O|j4qSb40M&vHdAy7|h6K;Jk}Hq7VzJ@ycP_#9r-6bog;e9n2bW#bVP zs1`_|8V*zrRTgSqPWZ!jQ=O(*s2r*+qS_YX9qYDNQVCScfrh|Z>Q#HoZ|Kq)xY860 z4S}`P9>4T(Pj%q8QUcX+AWP^o^h;Zhp{+GE#X^?QXE++(HyvHN2lG1#RL_BmTVPcF zQbG~6=#ef>u~2afOof&B;@up>0~FT_AK334CZTXQY_RRy2Q2r)Zj_-uv5_lYUV)Gp#!<^>m@M+`JARH7Mczn$bl~L3(%!o z`XQe{EgZ-cYPOdz8rO=TksoP_g-oGldqP$HKm<9>X(Ui92P%i|mW}r5-UvEAm!?># z9J*UWPV3=1#uPu94Z`QBjRVbuN|@bq9RDK7*Po_XXeLy`%r#n>h*~86SB*eFIgl+p zuo3szL7qGesLhLPQg#lOq4;;tu-{oLSCJ)Q@~U?2toIv z_7kX`17*TOhM%LD34%^W(-aG3!a~Ms>mOebG_AiQfjT(QMJUs4zIhi<-pUN1DHgg2 zWxATS$#@2G=7HM;`ptnxz!c40lRrVIMG^;TiiJkN6wM=}wsO>>hglf}`on&W-!;63{PD;XLJdKA!6Q z=L=1-&=;t1F75piFQ1KFszo3x4)h6@p3lN zFLPNux*5dx`w+;Q109BC=%tGaap$B!B*j99VHw&`M$r;M&z{B+$c6(=gd(DSO0V&V z?`8%~vCu>)A}Xp`j>|cA%NG&ImIGaqmC`EJuZ}>M?jHr3Vxdd2b~?8mqw%O*+`N@Q zb{uE{%$%rsRBuAiH<1(zEr6MmUgygfp-b2Go$SBD2iBeg^@m+iVT1Hz5L6P#W~TEu8n^o8iHE znqr|e=zDv|Dvv^tc9GO3;d6AE16jivWzUux;?aarF-@_MHJnklbkUgI2%0)wnLt-K zP&n)WoqWOD4E>FI&Y&q43Wpt_O4kG{r(DP?KPK zLst>CXmj&+0=aRZ9k4brLc9u3bxv%dDHhrRYZFH!WSkLXy3~q5?i|QTTuMvMqd%_g z(OpJUEaW6^r_;V%1^>)e&rm-CUFAUTFtofg@fRNPy%9;VkUI=5zu(wii(1sHOA>)R zIM8tD*X{KR!lMcCt~A9$!=YbS_PqliBR_lJdjffKAp1y3Enl0#_$bTa`)P`W>?3V; zR(V)|L@mn8ZYR(+4pa{tVYGkky@_t`;2fG_p?cT|V|?1Y8bK+l3Y&!wtQQA*4#Sq! z+ERE!S?)NRVxi|SZ0T!LZ-St04w?k==0GX1HW4Ddu@SXsnj=lIPztO~T(k|2LXhKU zBLewwpr}7^YDB{44+wHEqbU}O`eUopUraj!LE{&g6X-ezii5JKGs-@i29|uil9uo9(#-^5Z~v;lnpxWAIM|Z8oMU7P<=`zSeGCS`pNkUPmB* z4&(--iHiLPOA(a!l%`n74Mr1NueB>6$XQu>i|{$R$$>_}Xd<}nC7z-&RiP;s8U>?? zS$5qD5%j=Dl|TU;$OWFG<}tYgQHv~XX^Mqh;5nLftp;!U>t3Nppj#YB9eS!C!bjI5 zs6!;hLh8^{efV-9UJ;wLU^jsRIglP~k(+EHmnoU+lP;aR~0tIs*Be?5w%ZL0xEn3{2rdY@b?z+Or zQ}~3OoyMsI3gJNi!qBolSYi%>>~_!;3;hd2%hKK}o}m_5KP@HDZ4RUbH77-eQ#25? zE`z36NDFFCB%dr5M|F$}BRdHcIu6w_p3^@w*Q;CJUi7cpJC6R(|5?YV50er>xt7P! zze;{QO|kzfeVCMJnk5m9phK4ww+f%`JDlbVP8F#5bg2u1<~!3A3kgmY$jg{97(vGx zW)bKv2l@qfj@QaP_;e1BA2h{6zu?X}FfJQ+)Rhc35a=EUGJx8i{_F1FK`m+%NwJUt z)b`j#Ubu`}v?}5Rf$nplLYNAhSXQ_iK_ec}6blu?RM=-vMZBujF6~L6Fb=d9=0y_+ zH{pq+yxug$LTh1O^hEqdQHx|18wvD)1J%H)mg3cqml0GYl47A6 zSk-dY*wh0-g-2z!37?}#4padxl1-j}7eTj=(G&|+K#Q6*H1T)8^|cy-9&(^AFn1m1 z`M?-KX?ZloLS10)+OYowJR;~jO`kwf9LNctqYuXSE+VK^B*j8b@ElqDoBTzWZqe2K z1d8TB>aZZF7O09xZG${$iiOl+K`>u)j~;@)w>S_eh68Pd-ln#uiX>`LdMiz_&{pVe z?yOrr3qf{fh3@!Z2g|#G6WeL(G&|wKt)-PE{oP6$T~8EK#w?(b~&8e^K`}$ z1nE7bDHhT$haWoJfA0l?;-o7H6vu%UG)iiCvQ_(vjSRkR3{z=29&QX;+YWn58KDUxEL5}1_OwJ!>v2{Y*2W&$O0APd-}yrbs) zB-EltkrWGAz$WFrF5U9brPFwKnm|b$NF8nl+2~X}aU@YfQ!JzoH-ox~-eLqzoaIBH zWDZmewLPw{593AKzO!kHg^HoJ=ZDoFJfGth5J#XC4m24yS-3d5;6g9kTQtQ&lVOv^ zY>(}D+F!M65rLj?peHc<@%P_zo6+Csk4TDzp1|zK)^B&Zqf57aYb$|LIgp^XXR}vd zbp*}YMpG;#sO`}XNXON)gA-(pg%4~R2l9iCdgw7Re0pJ}NQ#C0prbxb_T^I4q7m{M z1WM;XXJN%euCw7FYEivNiiOU?ipkMSKgtmF_J{$2o^qf~uxNY4-3Z@xq2@HjLYrXG zw#G2h2tix2ObL|1fkL`VX&qD8I~TQRMm9~cP)K(>o%-9qR1ow`)tNw<97qM8qv*R{ zxSS(o98Iy13Oq-d%VQ2ADCfdG0zKnFP!?6{W`7m6$k&>t{sV%tD5C~dJnbJ|noXcA z4x|Kq@0(F>xEJK~nWk7s3Hsg!D${Z2~O|g&`%=AXg zRl^;~NWzE6jnv#}Mc_2a``K(9EE2drv&*q*+Dpgl)viiJF2RqOfm2^I+YoS#Ub*BodUOp=R_ z?~SWx$G)K{7TN`qjO{veuub@*GX^MqhU?4WfI2m^!bE@Qb3Ln@44&(@5*q=+%q)>})R?`#< zIl>qA=&iEZs6}CmH3{^F1FeNOgPqw>KLpt=p(z$xD|j=^-EWMblrSR#6>^{pFn2Az z$vO~0cfx6kg)YF{wf2IwXAm@9#+*QJInZC2bUN9i4!^zo^rI;j`U{gz-n(p8BIvP+ zD}jnQ&`kK0OB|?mK`pvcvVj)L( zGw2LaY(`L_N-lxkaUj9?sNi>1Kh&bku{6a(g7Hz`2L{&=RNz=gpb`#L2;-w!V-KYv zD8PxPSf~)jM;&#WaPh&bb87m{^iy&D^nqr|ruwoLRpMxvP*6p?-P$>sWgcbb#pI6~t zkoF#$VxdG>!7n=yia+K4&u;&G`2 zD&s)M;ltMvt6__vA>(O^g^t6Auf^a5zI1gLN(ofXfvnmkwRUCv+aE#6)-=UJR_$;K zsN5PnSe7XNOP~r4-A7A}Awp7JBtFegSS1VJG|G{r({aH!zM7DYV#KhR?X zfxdE}CTP)__ZL4RXjM;|VxcBzQRVbacr-E8^aO#bIFLfJq}D?Xvv>rxilkUbAsMzL zbX+w%$X^MpeV^|9bRoo?B9`8V) z9~>yN4ZcVEi?<+1K7poKD66Tq+Qi4x83bzLK%KAy^g;cmL=GK-Zy5{MF}we((C{h+9k*vT{^e!S_Eq2K!Q^&TDgaLMx)qiKs>w#M2ww9(?Xdd)b4=gUbgCLtI*?qzX_Adt#tXf`I zS64#NztJ?sLV{IGE02e`U~;>h27x*`kYMISW1Ko(GSD1IQ!FHyIT=uC(hs#r^RNMd z&W-=?%}Qx7&yah`%>uPZ`Z3_;p&O$lVhftpg_W{{M^r>__k(G&|c zrNB=K@3_$oL7yf$6X-k#a)li?K?7QGQES0unqnbW*m2XNWYKyA$zQrhpbH!*0y>c4 zEB$7nOIIh7Vxb7=Kz{D&7K|YKs%!#Tb0EPK&5akgyCUd7HBGUQV2b9BQ?Gjn($lLZ zkPQbagUQ=jYh!UK$biK(#X@B;d3!>3GhX9ceM@S;@Hw*OKucg9)^&&yu4m8;q$w6! z0_(6bs#|gG*7qLD1hV5mz2RMVBRk+ax^%5QX^Mq%(rN9}C6GM_5-d~= zjkm?+g-ZKriiHFVRmT^4;f*jdPq!1ufdje2iN}Lp^~3MF0~s{MLhf+l@$Q^Zym|JN zk`;j*IZzk4y_c%((m-?WjY z*nd@lo2|}i1%JG`a`2k>yA#_^AZHE~ z1|#cU7tTIFkk2EUVxce?SzG>EV~Q@_iJ=N6!l&DX19`)2{8E))ZxA$Jk)~M48)oCD zn#5peu7xInE_0xbFwGKhGfEjj6Hn3<3vGmHmV4U&oJK9Wl5a$yD;($)yszR%s#GAz zvw)^p=oGxK!bWG{$-RqH%n9Vmfzsh`)U?qy073swr70FlhriJ~cY8cMTyVveKyDl; z6u$di1A5{;k7}+o#X_O*-Ot>+0e_A&nw;`?(-!!nRZb z-QYlr;9S#?^d5f^^h+egLW|&B)6lzv<{)T)aw&m)IglC@s4V=hhi?YM6q;foH7HOS z8ut$GZ0b3tlR$nPs1UX!%)Wff5Vh#DNQ#9DVN1dS4>vp1qTvpT2Zax;KL_d$@48vn zJMor;?v6CYLjB=gx641_2!igm%p%ZD4zx1>j)K>U+=E)=(@Ik;v@<|(ioSXif*dz& zAW#4Ys(@9krtL@Yw*U5&T&zD+$rpj#Yh>O)B_hpPeds6`LcXo`iV zK7?uiTTKU0iz*a72^7eItY9z|t~v^Lw_=CU6bo6wU`p$X625e?$D#=o#DVs9z`HJW zIev};j?)wi?d^b5^rNb9DfPs>d;$e?pfgZcKh5rf7P@r5MN%wu2I}eu`AA!#OINSf zNT3i7G!Xg>|2&?36hYbQG{r&#q0dnFd|nHJ^4w%hh0oD#4paqq-HW4cxKtp?ou*i* z3hp{>2`^myVA!fgpimB^_5yBi^%i`4tG3Y;3#q+;K7+1VJZjN~HTnd)!+~zYyl8pJ zh85`2&00%SEOZ;@MOzm;1tO>=dOv~ga-i!_QP%38gR9W9VrYtmu0usxOK=SC*HsR1 zB+xw$bO(-@$cngWiCPpcOH(X#2acFny(g|7waCl#Hi7PQpii*abgZxr|MvGGGn!(d zPq5fDW$&*D1g*}`AW#?wIt_2{XA9RhpcYLipeYtQ4R7y9FVd;LQYj$bn8l)!+OONpA#Q4WTI(It5jKW-GM^BB;8jC4r(ikP(a~ zGP;dCjG(q&G{r(jFq*iiwr(1NH1_!rD4GL_!vov=**F0~GxpOI3yH%6d(MB%F9bzC zk0VeF2ayJ6pFByejjxw}L{OJVnqncJJX;+X!`M;;<@PZnPyz?K3J+|Dam^?MW%s2i7P<-# z?5&x%!x1#&pfiCIIgl~jbrNSL;EJ-DrZmMu#&FktkjlV21x$1A5h#fRiNjn_@1%16+5j6 z^ppdAhr7=D1xq1-(xbP~FB;aUnu zgb!>s2P%hQi*5Q@JhUv(rYROGhha;pQumJtS{bNGpd1eLDGPp6z%LQ^y$1%-6bpUI zf}U!)Vj6-biLD{fa}JbWC8?#8K5{v_bQ0n;#X|X2wmM!bhT!SwZJW#q^nwGqLPgoj zEEzofU$mK~SjZJB$`ohZ-G*8em*`5MmmH`TPCTyt*Jm$k(X}L+Vxd|%@p#mz(wPWS zRC++5TnDHH^wXoBJ0cXxiCa0=?ouBcX_>*uWIO88pw)6bp@n zBBIHrEoTsP=Y1W4UUQ)Nu;XU(_SZhBML{2EiiPIGj+=tXb+|<~bEM6M&ru!+nhVp> zduHz6jGzm1X^MsB!gO>?7k6B()a0v5pnMMW2KsfOhjyMsP@W%6vCtdn*IDX+tVWk^ z{ZBmt6>y+2us8ghw*F!SjsHbcEHno8hA*AD4%Y+6Z`e(sHylWC&UW<1Ghzt3v5}@& zNN~=!X7TMqs73klHUuiN&qw0abDHf`Zha+K&F5o5X z^7s=3`pAK7VIa1uS>ZWq(SSsnVj)`?h zq0_J;7JMwk9$h*=%V+|Xav)du!K0U3)rTYK<{6q|Ay@dpBXN8C%?SGRIiEnEInX$$ z&^V^0hYKg}meCXoje`n}bzk2EAgF9wBZ0~|P-aglt+C~IccKBcWsV7-qjDYuL(AF;26+fl@S~~!fM94jvd=ZVWS}LkPM``7bOioJdnONgk07Nk zG{r(k;BS=w`zl^5&)=+1ph^yO7HUlk4?VI&(Df}e#X@JH*0eQn)_n9g(n#7*pf4O~ z0z5|xj^9W|kZv+fvCsr~j*QZ(aXrK6k&XoV%7Fy4Ted@c;(p!GQ8dLug4wOKN-6cfp4_)cvI?7E*;e#?B)# zxWakF1}y^BaiCII9zD6o1HbF0Z=@*}Duv}yqb!RR=w=w2yqQ4t9Oxl@VRhtlaf`l) zq*&-7d|~^Pw-bgMs!eT;6=oGO}Pq3`f?uUD19`xgf56%nYB16_g{s=zNZ-XW+)B*j9PV20|- zuuG|^MRoUE3Dm@a>fv+zG1?Iy*IyJyQ!G>upW|`M6J-&!Lr(65@ab;mK13+!5$G2OdJOZTDlRMW zS=nDiQY`cs=0)v}u5LxpRoxr{wR0dBn7jTJGXZ}kEf>%f3%S7DwL#ETEd<#FR}-j% z1J%P|YKie6D|G2rhR_rX)x%(_@1}r{pLW=Vb0CE`=>bwvXr1H7J3eI zZWry2{zOpMoyr9I!-14wuHf<9WL!$!Ad+GsC73Ia`!yWzY?8>-CD30EbOdHUY!dI{ zMcaIl6bl`J*^hUN)$nxXlrhEx>f}HP@R_Jd%EOm#qB2dfPy&1=oaN$m(WP5tZ$+SU z6aIT=(;ircw$$smhW~3vBBS?FCr-xDuGq4Ys0OkAZU8vdjehHK$7sTdz@T# z8MR0vh^AOb65e%tBHL9F)X~{aAZrdJI6-BUYFaFUUW;}7e~N_!C#Za4MYZAzo0~Nzdwa56Qwg}2gp(z$B zhL>uW^7>~8>Y}oSKz1BR2L`-8x`pcz^j0LrLOL+ueV=pD1wp^gnG?vK13iP}F9gf{uKoDHiGuFICewv0YJ%cFcP~ zAV&^l058>`oZ0mV+Mq>KEMx#L)w)bYe2$z{a4vzIIM5Q93NsmP^$0-?A}JPH0#jj2 zt=;jC^&@}l2y~GHErUM8&vB!?5wxI_rdVhh^ciLsMi!$>H(;~$DdBT;i37R8T*3Tj z_PAd+Y70%VkQ>Yu1P@BsiJ<$BRSD$GfqdX*a9P<052jAW(-aH&z|CN!J@qz%tP~a# z$b|zr!vj0P)fW$2b_}K|7IKCMc4b)#F5usEayNl4bD*EFG;=4(u?+o9- zSMWb+&=d<*!Q1=2ySWPb8!dH8C6F5jTKOAJm^^Y2Lz)+9iiKAGh7wSPP<$*^SbZsh z+&PdT%utoTo`>(crL%miGXedL6qeBx3q``E zyWT1D3PCQ96i*8uSPu>~_XhlW?7Z!L5VSvzrdVk14XD2ocbtKsP`TLz^5j5X(68G% z`a~UqEC$jP3wc4m?x&IpJ}v6P{tX1W#)0G{rL@+xNtGjLwFynJkesBQ&R1g(yaQA` z^8|ssIFMiiY}mjxlMwV;B*j944Y0osE8@kb(PKObXK1`XWSZE5& z^uC-_5sUstE9{~Pga67x0aySV$jM@E`n`bp!i^#Vj*3a>22Sc6NI1z66yr<<3JPPdlc8LtUH26ccUp5 znh4*c*~4z(4y4w0eFFJ&peDHMW?VEGf}kk!NH3G7Sm-NMI5#E~<66_VqZ|noz=6I&iyD3`!e=Xe8ckCy^bJ~+@^;4+^fx+S zdz(PFIM5ily{$%1#dqCiJDOslF>rh9dQ5gl(DsT90tIrQDtM`;4o}4OSGtun#X?o^ zQXO|s`xt@@7giD|hy$s?n_>UiX1qUT_#&EOAvJh2lwZ1v-wZk-UC#)gqhJnH0KK58 zPO5n4+ML@o#X<$p3o@N@2iI)R=sl7^AslE>BK*Ka!2GA^u9NFSQ!F$n(N?GX!fQ*= zr5kLnMWEXpC>c)RFZ6K2Eou@;u~0Ic!2hUU)nf!Le!ZDMp&Te1`gMElRdI_(<51XPXMWel5$GNVS_plHMjh=)1ZCFI6bmhcKEs2Ces~c? zdT9}X?sK4F@Cx#F>Cy{9KSfe3Gz?xr_fy?(qMN}typ=#<9B4rrEP||Gdr9q$w4m29R zu*w^@{y9ASp zS%nUQGFQ?R3vEk>gNQTYO3|fjdRR@M7!K4c1CG-fxD>A+B}dT|3-!u?U!3<}hwDa~ z`gT7je2!u{kP7_Th+qHNcr;NhLsKlI0>3um=8}6IwW#cfGJzg(AXR9QliQFmbm@}J zX^MqZp+&#CC*x_f;y1bkisL|g;b`T-`R<7bN-Lx(7TOC(D?d@(ftPi>W*QUdF$W5P zcU{TQwRj^;{w$hep%8f2{a$<^7+t!*H?0U1&w-A^oNDFLeWy{2J_XPe3mu0!)mgvA zP9w-u%%4CB9OxW$@)v6^xq+ag;xxrV=b)2+S9cUHi%Q>-M4&_t^bz{ruVxP1gdo?Q zG{r(6q3^A8CLC9kNk4f{pd=3T2KGLbU;m8j826;o6brq9y${Pv(#+ANix}BKpkxm8 z2_|n3Z5gYET9iAArda3`Oy0(<%>Is`XnO@K;RBn(fy|);`6a<0uUg)6peYtIhYsYg zNr6WY)KQ~JpeGz?GQ3oqdd|^DEy}5-DHfUxFV(uO%cBwW&!RO1O65QuFrks$xabOk z8bneo)BzJ3&UO3tBIxT)a{{GtphL~D4tvobZzy{aKvOJqs2S$#67Rl5P^6d}fzmn9 zJ}3oowsrakwdlAwO|j5ECGQ(p=ots{g3-j@i6e0Jtl@E*Vj(XWO?|Ko%^R9xp$@1*pO~=M5w)oAoW%sn=0HbdU~NLiXeqiGeu$)4=x7Yo z1yzOM6|t??cM~Xw0||CTJsBN}H+t5VeH|NNhIjAqc2 zIplru1^QQcs?!wvuQG$4Ob^c)c*UgGbKcz2QJhQ{X)F35)Q; zy6$?KVxgre@YCmC4%H#(UVJox3OP^)l-$gY*oA*y(=~ynSSSNZZeFV!;g46YLIHu^ za-b>jsy`cX3~$`KH<+eaXbQaQ?;tdT%P9B2Z3j(d1KbwPili^pk-g(kq~ z_|el$Ttw88FJmoy=!!W|9Zc@2AC~Topwa@GVxc;i-0Rk`vl_Lieu_GQ-f^H>xV>eP z8x9~Sbt+A8z2aS0 z8{u>GnFB?`T~{sIpPpvG%j1S;o1me9eNls2XrL7%*6iiIqpgJDq;fhWl|es3mF1qV`sacIim zFVhjE^oOQcND0QF$I^djq86>#WJ#b(4kW0*`aM@?A%f;^rYRN@)L(^EHR9#7;3OXc zec?dD-x5kqmqE~#WSU|j;cp4)HXKJS${zTbKwmk~8z@jwdvA$P6?iy^rda3=6sY`3 zKM{bSSBHxTRK+zi1guGS)Rdb-{FtQG?iNu$# zT_nXq&tYV}=(fHUYLUt$Ia}ca`;7w~hp8}W{mFQ}TXHf@vCwgt3X|A0Xb*xEu4oXb zh6B05U@E;f1lRU-h@@D^4F*%k{>h3!(8hWL0@ZS$co4ndvJLWV}0vKn)zo z4tAe;_paZ7F5S{SG{r)8u=^|_rl$jf;-6I$=m!UKh2n4{GdcXgUe2N^7IKB+@Wq2F zx+7@x`0jSX=cthbWkcV4O-*L5)qoXo`nmAAvRH64?@%SWy z%$#V7g|eUueV^ebZPX&^7F_~0b0E_@FrRbqHU1v8h@@D^^bSDjpYd9%&q`wgwQ!*M z@P)N7Ni#+*+G{{lEHod!u$_J$#;8U4_pJ!j%7Na(bM*V|lq>|5hS3xYy@lr}fA!Zm z1oi6gPoOpq^fO&jOV>RDpEpt=l47BsuzVJluY-TV%F8^7KtDOqNvMRmzcm?`QXe=< zQ!I26Dq+$)Yu2C^8NPW>pkEv)8-7bDq3ST6fIVDDQ!JDXza`Xv+!$QPxN%wsf!aCH zA9!H@UFT?z{zj)XX^MsZzys^KaW3A{o9w1wFMMD-IM6%zO!Rv%CV?Oicba0Mckr2* zSEzm#UAnebO#=PqK(pb^(55*8m##f;qbU}e4R3~>MT7C8t(wsq0{!7Y$uOndZd{>` zS~PhLO|eiiOesri`K(7RvWPM#&|eND=$sssUvv>c%c5zDg#?`w^P(;H5cIj98-Y4G zP$xXF)BflWKu~6Xnqr|&cwn{Xd*XrE*;5Y)bWZKRH!G=xNot*YmVjTXdrs373#o*` zkw>jF@bcNJcewee%Ce6tS8WU4s-z?*c)3l z@w={V7EQ6x1$bah^=xrZb&Hp@gYY@Jz<~^*0~tFl?lSrt?e?ZA7BYkmWTpDDP;@h- z{#GTBH3y1;;_#DYv)3Z%(I1*(p%^F*7xV3`LQvMm#RRh9Ko6ir_9clb26TH5#Y6eZQkP-}A zqS~AAF^P+>b`r>m14UKAsnlto@idyM2Tie1R27`S|7GJJ@U;zj7NMa3Kj%%;Xosy4zwRG-6xAMyn_GqBu%l zE;bU#jRU=fqYr`xzgvPJy-PI3LT};dgWtC4y%BV*RmMs99JzC#n|CF(u8w(#N9Ehw zXo`hy-i5l6oKJXyy`_7dF^5H;f zunaBs*CYc$RxUKfLTRuJZI`R@6+t`dx?U7MN7p&fKIm>4Ip2#wka|5$vCux~Zv7K2 zjz8swi$)UY1_$~JrPO=xG>t}(svb?T&|fH}ekbdKS6dVUvza@eEIZ(?} z`0ekq+1*f!j7?~Yg<77%IT(IPc(vu^OCJK=U;up4=tj0~^GF zmci277sFb7$Zh;$nqr}4u=LjK^kpq-(e+Rb0tIs*CFlipT`v}hT9j~yrdUV`dO=yv zu6UkdY|m8$3gJKja6HA2B-w!o`YV!Rp#V6ZB5}XXI@F>?yUhr6n*#|qSp>z0B53s< znqncrCX4YgX4?_e^Qkj|LOIZCxEU6S+uuV_he(QrR>RF8-)$P67}Qd}PoO&-NDSud z=2`#filF`~G{r(pqJJ(#X^^$MWNZ-mLsUT zw3(1dWBVj)$y8IH`_k%b`t{{94tlh=f zCAET09#2I7s%^_?iv3qbSi=le-)nfLcV?u5i}2}=;WWQA9+sg;?!encLNj61W^v|b=i(jFU;ygQ!FF}oj|u^?fX!R()KMTPzncn1tWq}=RC)uOBb@Arda3|j0ifP zCE(jzF?TnCo^YVP?Wa!D6blJXmiAa|ISjSv)rSoP z%HlwkFX7D0El+W`Hu@t?u~6kpTb&(Ai*ZNYex3z^vN=#YOztIr&Hsch-5xEPVxe}J z+?!*#bvSC#pPQZp%Hcp;U=8TFPsniuO}|A`EVKpIfTV|d;X|A@pPb}vDQ!KO&Mr~$ypAAGU(pytNA}=`5zAjQ)g^4rqhfjVjO|j6vE_OQk>9cks zC@!LrKrcB^9~ew2mTbr6g-#D>iiP^XVCw9w|2e)OtY1G@;d7MBfds!L6nxABpW9>I zpQczy@LNJl-0f0Oi@qLHC(tVnlm)$@*V^0he9o8SG{r($&=%=IaPir?xJ=rsq* zfG_Nrf8OHfsIrKrSSSO&u$dQg@szTIh6#c4I8Z;>?EiP@*=y*o+cbrySg0Rt_V2fN z1Abr+Ty`W-J_ou8i?&bl{PrPe+ZCE(p^LC+`*7gBuISReZN5#Q0uChDe$|+J68F8c zTWE@f1lzByZU*Wh=*@;q0=?ltPm`dib@#qB)S~Q-G{r(slVFlOKEnb*a}z2FRLFs3 zVJ$WOjXs`p(@mr)7LtXv)W(YzV-XZLRKiX89KGd0Nw73Crt8}ms6~m2G{r(mur#xC z_gi-a9X&acKt&vA1B|S@uKt177iOHIDHhrQBkTLiwQ&bhy;zGt#T@7^Y`>cID-Qp@ zhWk63VxhaR{pwQC|4dgp%-lktcN|C?{zfNOxhbGacW@R>v5+?Wjgl)SPezw6vfq*I&)tPE#!O5XP|1wN>2_^dYH;Kp#1fVCPytNs}4q z(mhV5DHam!TDp$wUcV6JHjJiNC=p86ES5VRL@jc*lyet8 zu%#Tx1|Hb83%cM^ft_b)iiK?8ft~3->=uH$m1z*@GY8rTN3}evosG*2n?+JAv=NSK zX$-K(b3tMYR}rX;1HFK^caluoV07uaETSnEdI4|m1-JcFQHw?fnh~g+13iWk(9_~g zCJ0gtqA3=73?-nJQ}<0p(Ee^N1ghXbz2H;sCf&XaL8m2YiiLW?r+jB*MF@gM8s8^S zB?nqr0R<}ElfNTq&<>hnp_LW3I%`^I;%y_75_1Ujg#+DzeqC*y2`;(ml0;K1bO-u% zwX6C)K`pvJ9vh65djjWE{zR)(QVS0j>Qp~J8d z#`-|12ZAowSP`g}0~x~@c2rUj?m!-|r70FNhB534e?@%P$r||+sEz|2xDW5T>wV{= z7JU>+vCx70wmRz1dWoYJMMNYKsGb7}ekr@h?_U2P=)?n>Vj;mVWiKoqI}t&1HczXz+qXxB^gS&U_KTu8G z{D7JIPBYV^=IRGbckDD#Gv04z?)`Z38}tt>Kl}f=|1K);@%P@`H1TbCY1I21O|k#L z9)HJc-;peui=eHIkH|mp2j@y0v4fGKug70>$tN|@6bl`(gCS{qFI=#^d(~S4HF6+9 zF|2LCnzsm=xSFO|NKg!Wc&f~0)FRjWEd*-fKy7t!s)tE*A%X(KXo`i}>TGpV1}&^W z(AeIxp2ELvGY1NRH_EOnI%g5Iun$eKPzbzH?D|=5Ly&^`WCFEtp!IPu$iBX10fO2@ zQY^GS4lZ4aJYIvn@OmYIS~-vjj6y8;X5gBj&3QD%LMAW@>D_awDQeNTv8DuS<3Q13 zQd(11j{AdJ6!i~Hu~4*_olba`9kw7ilTt-tYWCC4)Y3B;?`57g8*#v6mKw?oaQp^ZqpF=jb;F zN`X#I)WDhet$#|ArdTKiIyIv#`n03FE@amj0{!7YM_^XW$0?u(x^yRY(-aFGfmty# zqfx#H8v0~Df&Ox!Kd_Qk{L=0if~KU>6bt=JrxJg6bmhYiI8WfdS66Po_ITftT|8we2;Rg{Toq> z;=0fj3q`>9=wn|OynbuEL*7gHz}j%2SSYmosFm?QlT|xuiiKjK&~iaT@L|-Vhgs7I zWXpm2KY@zQx0}157CC3r6bto#0-J6xoOpqt7-b^@*>Ru?Fc*94sV4pSv#X=Wg zF81)~Yj_&FfNjn5??I&)y8RDAbvzSV#pXtByW6QHomhq2>yK95|2#+}`&0 zS_}~sTT4?cWC6GLy#X$G7l`JPMY&n8DN9z&Dg#&$r^4=iZtTU)Zdz5L4g+4-g@7wJvhY+;S zdKZB%b0EPF$lL4cHzCNthNf6Z@B{MFX>+3yBwl4rper2cCcNt$ds!7BNU55pSm-9a z>pHwuB@tA&{1$;+Igkr<@=ZM!Jw#CJ3Yuae7wF_m&nm;SVjJ&0A&?sf>V!8#*0yT= z3evw%Q!La8Z-&Wh2jZS;zh0#Ta_2x2(1FaB3&2yPe?(F&Bmo`Be=IKJ?$*Eqe+hJz z1NDXz)jvL~zBy;j z-O~q9^G#gn|NNizk7hw|j^m#p>gZp!>@rQU|0=T}Tb<#x$3`J&Pt$SoukzwFU$D8k z$F9Y=Z)VX)Mn*({m`)aj=Z*SD1_%$@eLZ0xxGXGE!k09?y zQ3UegK<@yGP8uDJpi^-)#X|1@x|JO`20`cL^9gjF0|~y8W5#d6x99~0nqncrSF&5e zzqshw=J*c+-QYli^P5vPNNJ-MT|PlmEF?I;IaJQ-2WnBD;=b2~51lUuIuFZkV`6>q z+EIr{iiOU@vfJZ4-{T1CHdT#4ejLaj9=aoEm4BiZRf?oo$R8fM3cC&y1Wk5dMj(F< z)CW2k+hpIIM$q`HG{r)Fpo8IJbPYd8GurkM=q3kR0XyA`u6GYdklat2VxbkV)4jc7 z>=4wVxU~)h3gAG)VPw5(!t&t=a#%-GEHoTO)~7zr#sy+uqCyCCivulzj{5c~OWP5Y z6HQYrv;;cp{|@sXk6N@#E`vaU9B2^Sb(@FCBqM0fK$>EqL2%c7-PM|epa%yk2o%JD zG&&@;y8j%DH&~oGNK-7N(E)D;4SRgv&wv+QZU~>FU=HK}D=n+PyjMdll6*;1EaU(y zEfqeaWKfG9X^bFH2nW)942Ny3oQZea1WlnS7SenSr+Rok!Ka2BUY$pv+Z;$xVxabF z&LGsH(H=C#LV^;56Oq5}pcYMQ+(e*I4m1EZ{JAK`;zI&lnrMoJ2Ef)UZ{N9}5p>t^ zG=c7Lphr+!w|C?dJR)#1qA3=71hsX?!*YNM=m9RK1d8N9 zf_Gg_`&2w}R3VaLA;G&YbdI9~x*68hTq4jz4x|NZu&Q}G0#S?R)Y23SX~7z7AN3pG z5F~GKk3dlzr~nQEU-4B3m-kLyMN=$P00)6jk1!vKAnga)1d8TB|G->9+DiRds72Ev zX^MsZfw_W_wm$g?>MLDMpcoFM0<#~kVSnQhR4S5UAr+YYP@b263_22JE9L781-*sjDaA&dHFhF$8+VfudjnHgdXqCxUup(-aFu!31o|gy46m zMYq&-2^7bHy2AsjtK2yZL02Zy6bp5S2X;j2BD^3U;<24Tk2%mmn67jTJC7@)FM84x z3mt^%$^%pF`=b^e=r~89cn)L$RZe#gs0E=GS^lOe7BYY;r(QYfst8)L!H+-*97wR0 zVf3qNyz;+nBTcc8U@L>i%K3N$&DOX?0wr=F!Oz?7d#5-VwP^BVnqncr&)Z&}Bah3e zlm@;hP!b37g)Z^a_)I)6I&u(Av5+rxi9;vq;5DEevvvX{bD(~(o9gP+0(>*L9ik}~ z>Ib{2>Y}{yu?KFi3kimnpWbQ?L^nh6WK9A+;Xp^B zyY+b9ay*8O)SxLAIttyb$)}p|yyyW3BLbyzpd{E~(09TV{DswWq$w6kf*l5Z{_)0r z@5fa~2$aTw1ZzM)+hlR6Kxj2hv5;U5$glV2HR#fXE^#GLItOxwIkzck`y$cp?XZ-l zSjZXX+*aIJWo?BWElaUiLU>%U-?tH2Z7#jpgia^lr9`Q z7PTm*f~Hs~5Bdx(7Uwsj7Oj~RO`t*!)C}XJ8L_h65wvA4O|ek3V0`rT^DhK#_sA#E zTMl&oC;TFl?i+ln$2?D(VxjXtp>%Dl&07Sewf-Pb5eM4(R#Iz?OVtt7B9AtjVxg^X z;a%4}5wEr!TP<@-_#72;pwlpuzjCU%5rWnl(i97whM9cL#Txriivpt62=tBv#l3`| zAhxWqKu}-|O|ekiOIw|xnfrnev~;LGfl4?~IjjzU`~KPqLCX|riiOHyb=arG7Waa- zoZC;J_Z%o3_O*P--Jy)2$yPMQLgBElrKV*!p6cB4-hn_LIFK~#vmE`d=_-PDe4r^7 zl7@Yj?yomqMVHQF)@=fP{FTxp7hrbyW79I!Zn59k{7y@Eic9B350y<;xU--w`AkrWG!g12{YZRJRG z=_ah`8Yp~@K69X9u$!ufQX<~|D87=WSZEmRraCfV0k%YvN>eO!4c7RA#BSh|btG(rCB7!Lf%mP5HYsq5BeLezYs^DuN>$G6sQ;$ZYx8V zZmu;=vCs`DQ0eBZfp_YZeR@lvDh?zLmu~-xX&(?2SxQqZBo3EOOKC>~{XHHf5G$O>wEo^E}Bmq*ucG$2qd2NH~rKFfsT2X@52 zG{r)K@sV!VhxmcjiZvxr9S0Kp>a5-z#c8NTqaV=}3kiO8w$I+94d`a@8|+M=dJc3Q zN=E=;B*j8|V70~Q?JhjK z74D2`K!GSbkzw!{d|2c)XMI^;S zny_D4IYsq1f-YT=3Kl*`jT}g|8H&RXIt@jaF2$9mSV*wBvR5OCAYQ_+#i316K z^q|9AAMb@NsHG_u68z|aLaQRK`rBimOQ2>BG#=KE1}~e~f?A}ril$g-JggsurKFuk zEvk&%PM{VJBmswd4lS`+jG(U%X^Mp;;80KHOTn)Y^iIx-K&>3;I25<^?OU=MK~)23 ziiM6taZ8`lGkA(7{HPy++Bnb)n1OsUbf_hQE*+yO7J30QkZuWSqfv__iW3R+lLM8& z(9)>ecs$ivFOp)R5*S)83z>W#L7}tX6X+KQDu>mUr#m$9R_}dtXo`i(VYNl>SbJ{- z)m&>QP&)?-hvm_n=il*6ewi0du~0ZHkD3HsT!!ixRsYC`2p`xET*nwMdw!0#+%jwQ zubL*Q@c;kwf7UTZc*9oU(g0j?qbx;J?7u3)+g2y$^Nk_|ec!H0px>P43yyN@Ym4+-Ikd?34Nr;C z3wo#QZf~s)(6=Ee46>6z(`l|?ZFZL$0QDZqQzA5-<_fYJz1Rtm`sq&$+AV<`>01|S z+j(5dPXD@-CQ73TOgC>Ij>RZNBBJ_+pf!70k@iQ7= z*@Z!SB#;m71L|g)e-?Ez3N8wUnCN}xM5xmWYG zL4SbqE2Tu}4o&VQHEn^*(2*}LFles?+DtROp+C~G{_0>FPl?cGn(5v8V#hpa(P8~q z2JMqTPK8v?(ec|lXwkwxJS9R-g;ePEZ2>-lHn?Uo$VmcqQ_fwdCZ7h##f_&#s2jED z+Whrc;e7h{KL+iWK+589%df3w1GK!9r$k6u9A0hri^Fi~{>{?%skmVeNTB;P`>{gx z#Up^e&*mu+x=*toz4u!_2WZA^BL*FmKx65KZ7^)$d4NX5@{|aTr5m=Wy}<^AI$AKu zSpu2I&2~c$0p(*JaQd2~>wlZZ`J4g9EYSEqF?V>QKo|<9@dI;xpc0%%HOp$h%0@*g4sxEnGUa zjXWhn-bHkJXwxJ385usA%ODR4lu5&uv~xYNyH))uPl-?_4O^}rUx0IN)}3}R=$r&< zQKP2uj)hHd+sJ^`5!yjh%5`h`X#nJIlE$D566g-ii!SPr)e9is zVLT;5cW7R;UG32(0M)!y!XPgR^!2`~@#PDP_5f5^DJ4Q*?<-8wyM9>!&_5N;D-}2F zMF|u_Bfhv#)e``^Tb-vwD1=6Q0o`ih*ZZ@zK7%evpmzy$Aj7&9K>(#L;VBV%mq2sZ z)r$I(k{IMCf%NEoG-3HQ-1}fRo~J}ekKRX9 zlP5@(f1n=S$5RY&b=#eH;D0x3_0fBWrw z20%ya@{|ZEPlX?5`T+NN#c%4sAb$x|{XCug;4~>8pv#+iN`$JPr$^9}Uib)7eP+s_ z010%Ro^@|-G<^n8yXQP5Lf7e8=eIOJ7cSl2&Q=V%CV{l;)--lhao!Ekgf2WKLfUm5 zO=fm@j=ws4*1?WJffDE@6%jc-Gr}_6bqbymp_^1hq!wuP3|gdD;K88l66g)h#v7?M z!$n)Ok31zpZ)i5YVaOmH{y!NO!=M`ys2f$38Gcs7z2PCIJS9TisG=5{lAG%7u$;qUI!}pE*cXL~+4ApW0ZR5WU{I(8>Rz*^ z@uY=WiO?dqD?BAa-D^6U82ml@7@+i8(-;&cfs~83`>#g20d&1KPl=Fn(bneMReZy$ zTG=otTmmhoZ{364-^&1cS1Bbzi|JeU>YYU!Xpu$K5e7v_psb(tQ{E4D4gqLGG*5|8 z)=!1WklW#=0D0>MGAL344W!Mp(?g!(HpYu>c}j!^(&pJFPIIsqq_H!FK~WOOr z=X!mNphZQMQX*vXnSR95yWkfkV)x0B|^#6_cr~~5MO+I zdesf6xQ}8akcV1L;~dvz)1XDRdORgU9%_yz5yO6*ffmtE&N1kw1lmZO{tRcvE(Ykh z3r~sAM%wfj+U1f5K>GQk7<5Yljix?>#+U+p){QCPDG?e?eTHzK4H*DwjZ4lR&-hs2bb-=(GhOqcWZnpPIqgk7S8T3E`?V=Z7hno5^(4uZzcuItJ(Ti`YPmObM z>F%aKWza(jq)D|s9=Ga00chZBo)RHVs_k)8vBcS}KE}liN|r!cadfiBlYFf0DXo+e zA+0!t$$S3~deEXfXVe2L?xPe5)PlNO@7*@FfEKx(XN}lgpsz5~wltg4U;I;~Ta?U!D@7#?%XH=@`=; zpt;9Yu2K5HYa#WgqBfx;TPKkJj`L#Zyg4uNua7Ur`opoW&nIf z#-%(ZLRD!_HMsGm6lhWGY-0w!lt3e?%%E^|JTCZV&*3Q%8cAga)?fY@0dzQSE`wf4 zpk-0CQ()LFJS^Jf4o`{DvM4&{rR7LxfUMf@U{JaQa-n`*<7YQD;L_=L;3*Mup?;l{ zU#bG2LI-ySWk{gr)Kg7gUfmg>8jd_ALd~hCTKz#}Re%~5gfr;11k$4^nmD6(H39li zDJ4RBG)2=&YvwelV_Y{XjX`gE!Fi0PojeDxJ^B{w7$;BSfBc_yjJ4<}_z%^`U~jYW zWS$cLt6Fpvyz874Y<}0!682xcl{7z*-jdhaoyUcYzm-xVlt^#MK1TW%;X1!)s(GX0 z?#`4zCA35E(ymVSaGm3t@stRa&TYYKw^%2}h&pm!3e2i0FqKBj*RprExpB|<%@ z{>nP#0G359O`OP}ED5AWlM=!2&te_pge0C4AvKzm(0b8(5VUAU-!%-%mOy$`WwBzz zzJAc6oBeo7g!HJ&!oErE2!Nt5IWs6n0@XQ4TR-O2#lBheWu6kDItQtycb6`{j(^p< z%Ai~cw20P#nq3$(4qB8{o2Nu*5v>6gj+j3UTJ(5P5`*4LpsG|+*5OkAC4fQ~^OOix zrHZoen;ujLX#D+r2IWa0b9$%vHp=V%? z^ma(bmB+-UJS9S70~97g=)5!x&+OE&uHleo)RG!s@Wc2+z;p6 zuBBu!=!*pELPPkUh8YI|a(u*7BGiS3@V6(PUjr@D?^?#7A_??|`p<!kQ{~Emu*HRZMw1O*c*sl_(j9R4cybk9IHahZ@2$fNbs*T-@TM^Up4H#4`fiBV` z=y=lW|9RFG@RSH$q(@NBp9RC=(&>$!#-MK!=n%ciTbm~0#F5h&o)V!$^eT@T`4DFj zr}*12=(_~6p}yJ7MP25?XVf`>r$oqx`ewDZw6TX4O;9_^pdS(_ghtk@9H!zzM&nvM zB|;%IvYz^@#w>s~EeT{$i3G}^OIQAF&_!s`zNI`RLK$@FN+){aU}{2q3WI)1ptdw| zq`!aqQ-DS!@RSI(rHP|W<9uwPMYB45X3#GQq)t^9y03k)Om}EUo)RH-sgjUkD|AV$8N}xpr7p)oeM*>;U zZs=QY15}|!QI~j1ge+(`^xS!>SSs+L+<`%VCD3Nt^SEJj4J^hBucG+>ln8BB?s;79 zW&izd-r6x32pxbaW-bbxCgZf1~!1ZqP!Y_e(SaA?u|e>^2ZZRmzAAAJaa+s@L` zl|c##q}(=QV6!M4pb_(UN`#c#M)Xy8;yi;QGMGV*5@<1%+;lmq<^qsK6ib_zI?UO)(r>XDV+NlFT zJDqq+gaS{~(u_rwLipC%eozmqxR0D9kT$J%zsc2e2WVYBPl=E=t#|K_8}JvP>t;O} zv|j>Q&=P0E1*5eA@)*riB4j~JoC(MO?FUHJ+nhlMB+yc7k$Yl!V}MdBr9@~cwMe7R z$F~4Ys=k6j2PIH8ed{95G{>`*jB4*+eW685 zm+_PcDfi(2JlGZI8D7PEGsr~(&7{Sq?Oo2`XXKN>QzA5z7MpsPIcCA7JJ~LQL5C!e z^5DnE%|li}iY<#^^HWs032}IQa>k16u)_ zHkPMENcrRBJz8)74J{ge#h5|IB+!1kVZSwr$Ibp7ukw@#?WY^o@ajQav|Uoel0nBM z&Wv<6FVuGW6((nbb^L07iJI31jt|?Pl?b88n!%aTpSHh{|_Y$IwgTN zQonBHx~mfbYL?GaBD9hEbyxn?!|q^nn_RZe@8rn07@RtQzA5zp5A>8 z=Osal#(3y6=(GefqA8kH-B0TRH1HfxiI5RZ(ahVd_70#SH6}6Wj0Ezdy?C8wFUGah z>8d;>LVmOtuV|lrEr5nCT*IKV5@-z-5&fH=wG&#TwTP!gXblw+ZQHgp2O#SxX9jsl zptWVR&ui5koc9io<|z?cTSnVPMnAp+&?(KU3_2%)YSS!nSk&jK(4tjMcuIt7(=2gt zgZKUbU0$EWAWsQ2hGtHZwyaYN+!~X?myW=5F z+n)232%V+j|Je7LYoSFU{nR2W?xPD5C^MCQrq(9P4_b84fTu(#GgV=-#UT$z|MlDBBVl- zw=Y*O=>wPUffgOG zj$x3G1Tv&fzQ)|J?EuYR!c!t-NS*v`FB2vKwBk+%gM1~>L>ep`y&EyQQ~}6!k5*L04SPibMbS>` z=#6*&0kqYDr$i`{Jy`VmWMXFv9K(Ei*FepF*71MlOlY1#u0P^?XDG@5B z`MNK9(JKLJP@ExnmO!hh!nyGYn{Cjd+c`WXLaXj5Oh)aQzW^@X+2Ljk3XwpG)Mqen zwit_AcaPvH5lW;!gZ6;7Sm>qjwPsML1e!;6LEmQ|T?#E)__n@{-=+vu=XA zfNQV^P#v0Qc}BE zY6Z|LW1bSB8A%G00}st`#20zFoIz0%D37YqMqTv12G9|2o)V!vsz!5ao97KJ+N0Vk zrs9T;mO$~;3v!<35(AKVO`Z~=cVJg1b1sDfl1%XC31W4_byE z7V;MUD{n`h6920k9@N1&wA&h*uk$6C{a3M)<_`~{+4!e@Hv;slQc8q|hbT-`&s`k? zP>^{lgW@ERIaR`(tQmucaz~BhDG@TKN|-*mM~?$^*0-2JcO;PVhde$-MdR>rr5{g; zkn)E-Y$i;N2S`t?;mwMh?ydw{N29j92^rJz(p5@{&^j8mnKnFL2cXymJsA`)fd*5R z#i;?7xD_$bil;OtrBM6`N45TI5A5*U;uffRHy z-iJq3CIeK%h^ItIK_}zQQhSLjCa&(e40<4ea;eMo)Me!xfYzPnDG|!0E|cb%)9ax{ z`TtaIRoq7pCD2%EkzVW5EdaV(&Ql^ZmRjWJQT_p-t@Cvmlq`Xizx`dO-B8@jvTFfP ziIDQQzaO99nhDUpTgD7Zkw7 z|2*mO1uosUN+}V_p_=W$J(2wY`nG8YgC0vD<+HBiGu>GLrETUZ5mG+ud{j1F1nB8= zcLqI?K+41I7p6AHsjz?-JS9TP!|fvr@3jMHSlCb15#O*!Po*&^RRRUj7Al2BYh3SM>dsRl6hK?3HXrAys{bhBn?-mbV|pGzQR-N=hs>p}n;I+LeFNLe>B@A=Ag0Nsev zXV41?RG%s{``Y)z+Mb|jo)V$@RGInk&89N|HElhKL1_}G3l$zud)_l2T9jKUB|=@O z@VM2b4buT?xNQxCUP>SjnhFcQ5V;VbFO^avc%*ObOA-MytGALaFU8CvBYpF|epheL}JS9TcXu49}V}2Q2x{a<$ z49bu|kEmZaYULH2yB^`jQzG<;`gNnXTVv;B_Sbv{y_P^L=%KpgN3%cBqTaovG^uY5W%=&c0ur8zg- zL+cFyy5!4KBIHYRZu?HUV`b)*y2BZiDS=|?5GSKX1zrH1smD_y6ibIV+5Qd175umB ztr+x90__c@0dH@IIsjd<;VBW?8>%pwablz~T)O!wb_~jrK=gp21pd1M_j}8dB z^l^G`fT|qgDG{1S2LxS?pVt(iGX)t8%9TKSsgwWeiPIQ>oIdiD2<@d#e#(fcn*n-m zUdEvJ66gr^>t^k?#VO^+<9SMij!?fYGsIXEpyk0@aTPago&<`fZ{64zp2pCkaUnb< zLeccCdz3h=B|xW}8ZzjE1WKkOCL%Lq2LQCP8Bd8&G957yTGBxepnt2TF(_XG<^5AN ze%UV?wG~R=lmmebDwIIVK10vQp13p4EiL27E@VN;pNtA;L*2TVp@SELzDuBN8u6tSwf_q(@=)-U2xZfVug>KvI&kUS za$*_uLjujABPQHVXZ!|eMJ`W?&>T8q;_BknSkziP>>YzjB#;%gXx+9$KLNUH%2Og_ zMJ*avdUy%6DCJx^gMLaNEqax&)%!jJpj)0iB|=*CDi7M_@dlu*>aFfp+_1kS&^fC5 z)9dGT381E`JS9TssOsp!cdZjq5va!ouUXl~N+~Ue(d$+zU7C*X>LTX3$>=q?|cv zcJrq`v}p26o)RJD%t_ZO{y5Jtp?4~S{z;(ebb`u@Q5)|9q@~YOA~c;&P#GJ3^FDk= z`Od`*DwjY$R4_UEYUgeMJ#yhG5%Qse$zMwX@KtX2MLoXaKC>(#(nLsYtA2 zTwTOdBGi#)PNr|!G9OwLF{&qn>?BY#nvQ;%*%i0Q9Wdi55o$)$(FeZm!DXFEKIRPC zErHaiMdlf1SR6jYm#0KXjaqbM{wsZG(ZU8R7-TPjUPq}Kf83t48$P4{>O3VvucK&x z%J5ok0m@tC#GpMAD6m>hWApq@b^zU2%u^y1Sk2Mozup}EedJFQzCSo$~odw^F{&G+cuX$ zjuI$_){n-II*wylV>_M_p%_{}+J2=b77-0kS4pV2kM>HS22^;wCCIQbv}k7rPl-?i zDm=DYpQ{cn$}-Sp&^`%dNN?CXt!m*Oq6daNB|?Vuh8<@;_X$8x4;eGaNdlG9m*M`3 z*&5KI^M`p#gi7hlu)MzkmVgEq&SlVk2{ediAm{tLXacn76HkfIAew>HzVl-RwCLoh z9Sk}kfnL#;1oNJCu`J5gjHg8C6>UlA7H)|L1)TPFXV5_jG>^)n>RkS51ue4n;VBWC zM`ckaGuy^Ni@a1K800L0H0WD5{)xvufHqX;DG}13Z(aZ9wY30hG(U|&E)r-HO$^=) zd4$u^zbmCgXcJ8g*61?B2cZ5@B@8+wfgaLl6nwe!Fldo_G*5}pL;8#|>y*_6$f1em zy^0(5umtK!ukx4LI}HIcZ^~05)RSK2vric{256s+K7)=(pf}VxnXuCYXHMp9;3*M$ zL!A?&cx`;BIwenJ&`}BWh$b|~&3-ZvTC_HWr$p!xO=#5h8ki3)((b&5LB}MJ@*IqP zI{8ll`d29>LdtV6w10KMPJaEp&I~#(ft01x9*b*-0ras_N`#c9)O{Q2VNZ4Ihbs&^ zA%P-kie`PCB~<|$o6l1s6iHJwMLR#^6wTDJNept8KrYnD-;>sVBtTotc}j#_sFVNc z#JCdpjB4J_ zp8{xnW1bSBHuNfA6S!UtpvIdzFzA#7+Dm0oHqO1C1N6I6N`&@OSycUvt)2p;lWxi& zcL_8wo@Vk}zQpP1UzJiKG%%jFja+w|3((X)Rt!2VfiBS-wz+HEH)v7YzC0yDm*@?f z+vo#UW>#~tW6&81G?fl>$T-p%H=(CiN{P@^I?SQ{XmebfXz|g5L1!h9avE*Sr7Zl6 zK2=JIka8Mru+`5+TQO`XvjycZab>Q3edUAc5Y{EOAZE6g)HY zQZ!GA&>Na1E;~7PA=EMIx0uEtuij9{*u1oO(Dw3!RpGz#Ys>%mKkFFV(1Y>Nwg4Q$ zTejmV@xN+A5616ai8!U4z1xQUR~IGC52u&c24PA{)(E%y9RIY-_R z23?Xs@$_KKE<61dpdBB0N`&I+!MN1CMnkxCb;kxW=&}U5ew(H%P5*udXp}imiO}`i z)PHU{ARC~z0VxdfmOyGW?eA~ziIaP^ukn-!snN9m<7e-#17x7_nL$1hXbC;)yDplH zm##`O_mS**+7R1C-y}j6qi#Ftq4oi35ZDB~TlhBzJRN zg430jKY2=o+R!BV(6|S40D3mwi$MVrs75Jmq1tVaOEb|EcuIt7l+q5tAEx-Ij}C}s z&@~BUL*3eX%l90BOLyTKPl=EXb!+dmvc@yY66?KVP@n{gqcKi#opEOY3aZajA{0ks zoQAc-9>S$tyR4i+*CkL4jYAhSKK%?Jv*kP`LNPQB{gKo56F_V3Yd@&CVQ)yFdeow6 z*9waPnwH2@B2QQxecZIxzl>F#Ipln8yI zzS;P;eeVD?VVEm}LM4#0Kqb0i^H6{~nevnfDGO9|pXRj&sQrat28Br=9omX`zP`34 zK-Ii>N`!Q1E8@HHb+FcS&cCM&3YS1_X&2{G-5Uk~8I|*t2(_hMocr7S!IGOX^NSf2 zA%SjDZ}Y+B0(>&KF5oEZp%y66ye#u3>AQ5}_f~Q6Jx?b{0T3yUZCBErIIOL$z;p z`=bC^+VYeL)u)GQ$n|=-$>MDGat6gnpw_e%vE$>O_^sQV!&4&EnzkZpMEaFOi}H<~ z7<5wtouG&6!dHXeL5m&_4 zKL99j0Z)mL1HC5h76gxm7LAHlNv^n$;v`TgeZ4o=u-5>nQ4CLsP$+%9yZ`-wBWtUc zx(vD_fplru;(aFuw~O}G<|z@$wbymq2Dz?q1ML?-f9WZ+J?C%&6RbR$eQFJcjIGP=W;7LnSxwKTm51 zkc}}PP#Mbo&?gNwbUHLc~}Y4#GR)^NQc%^t7^W$H6Wjoa0cC% zK>O&)utq&?2DE6;Po5H?ee`5_(ZZ)cd`69?r!go|0!^e@#P31Na{)4&!BZkMk!BH( zTmQg~Fylf>7?dP|;%If)@~CDqK*K|MN`&HQb+~GM^AXUZ9a@?x6*uey38YD1h6RSf zIL)%6IZug@CVd(D_URA^Q1MoM20fHO^Qo@hsQdXR(4yPhcuIukQ(gVP3tRBT7x8Q& zgOVjsTPj@}w`&g8B)C22DG_Q*rE4h`eepAz-ggayQY6q+8i-jBdF=u%^6JM^A~cl- zV&;QA{NU2faCc_VBMFojt!g~!^dp=kAAFjpL?|zsCiispdjM4O_X>j^OQ1lS(5M^Q zaU-81zg6O{NyPhv{GmsKIWY5~0b|qUzr39HB*%st#w+a|z^bPG>9m zC*bjJ6RPo)2)UcnBWTDTJo#b3aw`VCkU)!lRgFhnS|0;nhQ=#+N`w~s(hVEsx))lM ze#eeMX%fhmRvzci8g>Do+jn_NgluW$an(#q7h&4 zKQCv1&S&zJ2o0hUU-4uGRx5q%UB;jc31p#B)41iT!~Fn?)#oV@ve0lesp=Pq!|Qxe7OC zRv%`@pgakre0uxGg{=VSL8X)kDWBe{ed^zY7G=3vGw6c^%BfM)_{fK|xY~04Bu|M@ zP7Ozsx+}Xx0o3G&1B3DFUhjDG`cDp=mVl$N2Tmx)IBuj}pk1Mtmmw2X_MKN)S(pkS&e)VwN_-F7YVM zcMK|&K(}fA=u^#Cc>YzNCOjoVw`u*z=yb_)xO7w3mow;-1R6|JG`^QVVV|L`4Nr;C zV49+7c3@X;fC`>yKdrc7KTDv+^z_cHf4(DJx@S*$N`w~E(>t>JOWa-9T7MvezDOVo zx?%kBA>jJv;rl1%rwtkb)*}OSSiRh8B%J!BZlnpvl|J61_=q z>8||T%%HCl=o9T+a~swLSC*ZA@stRCqMd7}jLu@!U*Jqv1{F&nFS?ILSsld!l^_eA z5+N_TkB<9({tYdPiwI`WHwolMpV9MIck!TrbCEnHLT>aKRc+rBcYqdZr84Nd1R6+d z6DlK8hQenQ)tsk9XdtakRIN7zU*%oa7BlFF1X8V9(|FLPAY7SeKhFFpq~;bhc=Y?y#IXw zE?tr#Pl-?tZ73V~r7zYo>baUT=$8b_RF;DD3aAB812>)$p-h@nP02TX0WAvtyPQG4 zCD1UM(Acfs)EXe4e>^2Z!)QX|VreRt86?khVo<3BGNhUlzvCB20d#CWPl=Er)tqd# zm{|lZdJy5wpfU-xlcqZBT=#bc$TgCuL}({Xb?Tie!Zp5sjS?92M*=mboc1VcYzjtU6IS6zY^#-m4fvAZiJ<4F)MjWgnm;gNRDTjc-l$5*RMDB>*{pUWl*^U8bPyLkI#1-2~dsBJS9RS zXm%@AJ#Hsly7zmH8Dy*f-)xHKHwmD_=XgqlJg8qc?OO|cl^?4T!Js`7$ctKJ6*j>Lpbph|N`$7JRl`u#lfu_=EB5TSU zoM+e;%u^yXl|~cqX0B}kb&OjzG@n=8u#Wmr$LM*#_h6HoF^l29>ZZ;A_&@6yL$9eC zU;Wv@AFgwHrIh$zgD!2nA97RpN+S zTLF6cZX$#BNubx%e_ng$pgpwcQWj5%&}-^H-`q8AB0vEqYZ&AtfwXDp;TPWi?hJp66=l`6)LvBFbS@G|K{Hg3#`VHI z_)jXOL`Xq1R54mAHK0W!H+5jpAqkXA+pe9=-4;WOG&b{;2<6hY>uQO1c^!cS?5o z*>JdYlMmT3=%@rLqv7F-?15*XMS~CXln9m4@X-I*!u9~|`{u!*V-m5i*XaQfkv$g#blQj$zPo3G|+pp^vtV!vd94Q+P^*-qSL)&JNAV0L5I-V9*H( z)Rn$nXj_QqhN`!XNL-pEe zLo9v>U!j#&al^VvphMI*i=GyZ+eJIB;wce2M18YGKZd@9OZPRwfI%lE&_60rnHKd3 zSGAtp<0%pPM+GWh?)u|^w^4^_3_2x&=Fw-gC?Nv3^k!8`iO@XyjO>s8xB@MTw6$T7 zy96?$!BpOztvJoH&yJ@=$czS4iwYfalX9c1BMdq%fwt2$%UP|vY4EK}tCSLK?}Sl1GH`+Pl?cZ}rR=bY24OrWQ-j|ZcdeEY>kC_a*EP=YwXrh&g^*(?e6!Me^b)(V5faF#9;!_z>&LD3I^oHhh z>W#O@Ik$pJDG_=@^En;cXV-=n`T1(Us<>f&B+$z{^xJl?KBhp66n;D1u8*6QZpLk9&=m<3Ps5gz8%~V}$omdYiBLQZTb`O-!<}pHI<5@5DuI-x z0;^`9y8+NGBS0hf@RSJ6q81IVaROiEi$26NC{zNSyG_4Y z{M+gwK!*7|B|_(J)35Swl5wfH8Zr!S#1_QM3 zl1fI!eH1Q%_R$#jakKQJ09jtA zdG=KXb|9Zc@RSIp{-(m?sE=o$MLNxPFeq9A715VrNomOkXpy-VPl-?weHrHGw!^RY z<=yTKijhFoeN>Hu0-W>!+F;L9B2?W+VWKj4+;wP?>)UV!-IPG3R2J26cmVD@&dlT~ z5h|s!sF_+f@XHWBAdNw{B+zvFGW4mk<2kg*$%v;!XgYlvTxNUaK#OXhEMd@X3Dlbw zG7|c%tpQM#Q#>U?y=fui-I{-^0BT;|_;tk%8!LfEMA55!x#wnpax0}oXhf94q;SO@ zTzPCgSD!&~5=eRUK_@%yW&o8`N{Nv2=!3P@*Ws{bOw>dM-H|}b?N={e&%lH6yGHYr z2r0K;tqiO27cSkvmTMSvR{{mnvrhAvIo4m*)8;7=3Z!SfpnDR?mM)#^bf+wUS`FhV5wfLAXLH?j3P88d=QHTO1S+R~-H@GGs{s0Yfu}^M zoceVKsw^4+kXjwJHx>6$q68XuUDY^o*&du=e_1IdLgTJ0OyXMST>xn3k`4??l0Xe; zH1RRH3GM*3T*^}-)PP14P2SbN1dwBjDT5wJpu6-#huiH&KY&X&{}E4#&|UhW!xLW3 z<^fdD)rvt6CD1@Ry|7cSU>qMMcH=1#8c3%Xo_OJmiy%v!?HH6SfgI@Ted1Y*KxolE z7oHL!2l{#s=xw$eS`=L1!Jrfgbd26dsk7E;19bc&Pl?bmdLPvZ&&Ltp^HDJjdL)6W z)3dJk0O!8|T`}V+5voqlx~Xj^uGZR7SJJ z=ia`Fgcfc6b%a4LB#BGM)dNLuf`IgVH3Ba=~|r)8^&S zqS-TfN`#aPzA9%2#slOSnZlr#5@;ChA#y71kqVG?6iYkVtS@{|ZE4_!ZZe>c|EUmak^pw|*;D}C$g#*VXw z7Hu=)DG}OA-?{_eZLrq#=1FS?y^%oq)Fm!BU9AkDQ>S=Jgz~9NT;+TVC-{tR{B~f_ zTM1-DgXM@-(RTqlTFO%*WJH7Im=Jf|`;a`{i$R$Zs4dl;JgV=BFPWsLQiPEZsfP|*b5r4s+>XD5~wB3Zq0nxsSrLR)zv&DLM>@_YyQ?E9F=?C*M3)V z!{$h!nKYBH{?P>c4674)N`z+8Onxnu(8h4--s=uzP_6`W_@QcSpqqvZzBk(Pln6Qe zptDNGL?E$FDy>JFD~y(3SFkRQ$0*+gpKsJy}Z%?!$uKD1yjBXM}x;yzD_P`}$$a%0gRYkR!9_hir~3AE}K z)yVy^#&^o59y}#Nt8OVwtPOLp0yg)6IfFh+psMsvX)<&OPAP{Rre_&8(fkpkfJ> zN^3x#9Y0t>i)Pm0DG^GgHJ~1T_00h~zCM>h-z3m%f7;A)W^NULmfP@@2+j7V$4J3; z9C~a@RmrZnkG@NwEUF|w+VD7z2qrz_DG|z|O7dHE`(Qc8AU$0M{g6Nnsg7~Slun1B zMK-;8N`x9x9pk)FjdsDM)7@vxpb`lbQ>CV{dTbbe8EQK5lnBLCaWrZ6xBCo$4tE_rz3YT=$8blN!{88^$dyu@*l-hB2<&Q zwW=y9xzM6nm)sfjTLSr0mucy!kf{KfUgjwg@~1A7VV3CvfTq`uU{I+9nnWd_H(gKK z0o1t;Pl?bZDgnK0p{oIqrga*F$|O)tnhFbT{O$)p)=PLwglf`M*ym<-`vTNGp@czy zB#;Mv88Wm#-3CbW9#4so2Ynevmi)v=(5UvBITbhTUkMaOqc)As7C5d|^cnO|0==gvgH`iSk#43vm7kNsAGU+qw+@fUeiL%Anm6D4q5Wm#3UahRI%uc`k$+4Q6jK*^RoB|@*M|2*@Q#w&naV)GfaM*=mWr5T;_ zY$Jf?$MKX1HKC=M_PttRpJ7!ywcLvP$Uy>qp%c9Z8!y0hy1DInN`$`9iC)ch{5nC4 z^c_1eNFjk@sm~C)`s+|=QIowqB|@>(XLyn1h9`QNe=uc`qXbf?;_yQb9Ss5Mm(Noo zq)x@*yS<7ULW{h|S}|y^1Tvu-IZKD_SaP$&oTo&{glgoxZJ)FU$luS7LHi`oE_yQ9 z?@`6wXS=TOlnCvjCqv=Nvp77QRqq^woFvc+Ss;BI<#o% ziWmm%mq5*_OZ;GW&uj1*8LZ?f5o%6d;$Y+5MgR>;%3#m|3FJwmiGrNQ#{trPz*8dR zNu!D8f9~UpZ%vOf1|5_@Pw0@_cGafhwEy6qJS9R;=#X2dQLDQ{izXbx1qKXqkw7Q1X;sVe78FWMfjiJGESHrnD`!T5&Pl?bN8Y~-L*TkOcv890wIx2zU z=;_@r|Libmk>fI+5}`PHdVk&KX8|qByPv|KV-lz~O=whE)eZaJ35h%{yx&?yOIN#moW-6!zmhwqhAB4kP9qed&do;Yef z`1w2~LM7Cq(a(D0*ZWy$EQ3x@C9L@SY9cG3bm0 zQvUKz?(ceo0Qz1jB|^$y-Z`?z26sjMSXj=Wvl3_*{qXL(#|GN~da#J6L}(ZN@NN^w zdZF+cwYjJLq2h-1kU&>xwVzuBpt4;S4Dysfe`&?!(P3wtImxo+DG~ZhD<;l6W^V^*Z1!db zotHrIw2)E#RJYmi8R_NllnBMsLdN&RU{`=#N4YZSf&_X^FTPW*d01<@%8aK(=rz6g zek5ApDK`Ty1~bS@0!^W3-GY}^>!C$$F7cEIO`&Jq?7^d2K#TJJK4s8F38Ys_Rts8qoG;9o2OT{4)bL}&w*3N*d4_kXrA zdU`YHiUf+FF|5&_)wnkC;XF@?Py~%(m#iL&m#&X$0)wtfpbR?M!|GM+bhu%w)#ND= z%Ak`y0uN5a^1}BEav9_=fyU8P=gz|OZUEi3;wcdtM^l|!b9`|zG3k~{LB)L(Ab~E? z>{iblgLVL=-sUM0xE1=Ty_B8acHE`zQ~AmvYHblm$r2cRvjcuItnKbg@cKI$8M z>)P)yW>BC6IzrnRM`>PF0Q9p`N`#KkHpcD~W-kHA=jB`mU6()$=;?i>^q2`iPOo@M zgci`#yMFI08v!yLxPw7AB#;^1NA?q@9kaMN`#82bK-da_8n-^ zXsa{^g-9UFN2uGvDK5+Tb+3X?wuoAH2=7fpl`G9o(E`7^{WhumOv%cC4Rrzb{@1yuLe(vPziO3LyJSw0D8PAi9s2!AkzvQg|b&cgI5%Q+f-OsE{z-csl z9ks%W`{*47x3Wbf~8qy*~?kLCbdW zlnCiiPj$fH)ZqYGy)k7_tOPQn`l~m0wvK@o4Svg0B4kGOSD$-cP6kNFz=}a}5@`1? zIt8@feuxs&>s)*snW!gr`KPAJt!-w0wY7 z=pVm&FzBuXQm0F&7qPu2v?#rpr$k7dE}iRFtuVNBUuVQHC|&|R3Zp{>b?;)o?$J!1 z5}`+7w0U;JoJ{~(gk>-&K>}^07U=|*?uHh14(BNm+DI*0*SYB}fHXDA7<5kp#nBDB zX@-$5K>3wYA{0kA?E9LbYXS0Fsr9MihP^L=-qM$$$KZ{Z018{hQzG=1z6_BiE3nV7 zE75>Ki4tfT^;GB055%?dDM>sfLd&S9`ee{ZykYBip2nag2^3Ffl@#u~^9@=wxeHH; zP&}PgQs-l*esJlk?zdsk0}1q;p5B?(j&%Uauapv@=k)ZR@$g|^sAH^Gc!WU@`@ngO z6SCt5T{$-uXFuNl;D7v|b&Opu(FWKZ8Cb_y|0hq0|5cYuw7KXnj;yCNUY-xvxor zXV9W+Ek85pkp${X1Kvr2p*jHhYx9%{^`!xCz@a!hfZlDd`?=z#dn|#JzZjCVw(2^7 z9`E2O5mNqQNXBICCID$>cVW;I33M)+Hv6lb9t)6K4o`{DxoCQ){OEw4z^qYb40I5#$ql>|~y->gILa{~a{md8^fq@cdp;(r}|0P0GIwpT#u66hgax}syF^a1)_ zDJ4P=>CzQ^%vcZ5g7X#(%8)=Os2Fc+Y8UJT_PD@PB6Na^@l^k`z$MNBRkkqbwFK%# zr2;+sr!ImPX;tMZ5$Z*y0@2reaKP(h>B^ut5@;>;&0@6Eah-1EJf0GvwbVBo{&Qvm zT)L3xUKNOuT$Q{4bURz7@iWL{WQ6!JEYA5fDX1vWl*LBQl72Uqiydy0PWD> zDG^eht+Z-izgYn3*cUVCodnuJkD&6M(O6~CaSu<4&<=V8wbMz%HK4~i>O~b_?<@%v zPj&TE7c9YPmg~7ZB|`C3SHEjlHZEjnne=2(wgkFCI|L6mzPuAIT|uRk2;HC^f)1Kn ztHGr^c*>kXITENdJsEUeeslzAzB^BeP-l8FOd8e%i-Ue2Ig31mp!tp#C+N&(U+ zFNFOXB}^VepE_{kTyNN177XdffnhsOkmIl31mtyzKw$x^#Z7_HcyF=DZTh^J`cj3 zwyk#MGALgHInu;Ye%pU|!%np2DG_p{iK9PT%J4p#oUZb<;yx;nK+1BC2EOqj&?46i zo)RHtImfn1?Qt~mxW6ugK1!gv^sW2TM12oH{sVYQgzD0_?oof|IQWd#dKxpRPy)q< zQf1~%J3Jt0)_I;1q1aHGlqfL8F7c@fSVUtVyu0%2Oh=pVm?x`qaa3 z-M*PS81z{Jji-9xO&1ORz@?jM!BZkMp6Y=Mbsq)5rJEJz&Y&+6D1izlzo|~&i%>XE ziBJL+On!J|k2h@R77+|8l0bFneRM**56M*04wvq7W(k9eB~VA2jsH}dKNT)r!aJT4p^h{gpO(DP03gLM z&Ekq1_L~HHMPpd4g~LAsn3k8%LJ(7MV=C& zVCvU7>vhK+H??a_V$cr>r2O%MZi}y&1Ej6WQzE4N@q*pj*|;UaW$_vYl}MnK)Mu#w z|1j?crs*pAb*#M2a!&4%ZMKe@KJa7Dl&*-i0 zRR;Z%K-TmKniN$PtI)@`<0%ocrbkeb(?~UdZrLX>=(hwaq?+xHHy+`L@6aBe5}`t> z*>-x{rZ+&{-{v!@R04gaVav{ot2V=>YoEzeBJ`DpEq`{0;7|NC8LalL;yx;qKp*H; zem}z)_op-%!c!vjfnMd|KYg{JMT5LMFzAm2dgZHXT$Wc2=S4gD@RSI>@}-{Y--9gx z`dn)`gZ@gO9yE&>wdL1!xOAblc}j$O&@AG%tO27aOpgB&oRih?|*Mr zT60&``0KqYq zCxBdYG8nX50!^T`RGWO=Qve;v) zB|;~u@7?D0nLYrio!9zaal`JBK*N>G(ESc1LW^EhN{P^LT87qg=!t8odesaWBYBbXAYLGHC5p$5pt&&U#G3LzCw#i7ffT2LINew5@+y_qRY^tG%KDGp#)mu+-~WB zQ=M_~HVkrz!f-0QnH^SFM^xi-Q?UO*8KGN@}TpgnYm+sC!o)V!=AL)0~HgvQzG=8s+A^QR>jh_zaL^5bVLFTrY}Q8-_y98BQKw) zL})O584guhi$AZaJ?$NXj!K{}r&Wyy&fA#;pV5a(DG~Z|T456JIB73jIv@XX1|5?? zGfU|V=g@36 zA)mTIivpJoWY7r-WJFzJ^|Il(Q02Our$oqzy2N^^^|2S!IN5?ht`cawvXr`TdVOfo zn@TAW+D@g^rq%1?f^YLqTNvafft-J+8n<5Ha~7boN+}U?{y`7bDQ_0Tr3>Bf%Ak`H zD2e)Yy4s^x0;D*=QzDc^{kr2Dryc?5Wo|HoPD!9S^xGWvt?pr;;qrT)5}`TY=m#d& zZvO~SJ>ygcxl5q#1lb)nwuTnHu9Om??gXtm|MVh2MQ+6mIxT^E{#G?k_Njt>?}U>) zB|<%aD@^iY3^xGO<)`}3iu>q{1PY|tt&)_Dc<^IMrIZK-((Kk??-n?xnl!m5gU(8z z`~vE38UH{id7YD&kBwVdFI z!~aEdc}j$u(y?QGI|bk)=t+bzgDy)Te_G>951)@e$8a^0r$oq~*7%0nPks+AYSYq^ zLEaMRIL$zI4IJADE?sG*ln5QC8OXwkb8&X-;kq3R@{vHvv;)+0hS?;5Jl6A+2qn`F z(Eiz{@#}r(u{(o&CD26bGn`01`wyVNCp;xW6RFP-azCXLT)J}x5e)K^K)6OvI zp#W_+cJG5Do|KR~duRsn;8CD1IIMZ7=CYXLxGYV(u`&7xUE{iHDnwOFoJT5;2bNT4$` z8z1r64+m4ll~N*fhGye!CqC^DpV5;$9T*fUfi}{?CTreAU?=eCU7ix5jdZZdmLY@d z0CcO}a0Z1*AmxG2&I8)v7OJTBJS9TP1E04Y>aYi(r4CjM3YS0!=q+jM_IWLQM*aSe zuKSMb@%#S=j)oO7-$E!mdz0cNA$x~xMYfQvyeSnzRCZEkAt@v!O`<|l5u&2(jL;sx z=d1I(eNT_ic^`j!-0q$0b&YeL=bYvk!|BuP%l& zC`JS9T*6;wFdQQrxEqb|3#N-8hd2nke+9>?RR zJ^ToVu2~3AiBK(i93N<>hj)gLT7wuADS0_I_~aHYQm!C|UwNqj|bP-P+eTB87R@8!;*7zxy#Iv8FebhZc3sPh`-236w*Ra^s07KLDg&z*8cWLyz*?{~BRQId$51 z2E|GsS6YA7cC~K@fG$nvDG_p|^;gwhG;uda#Z`^c%IoNX1o}nQ_y#Sl@Lu=GpQl9V z7gggm&s+(GLwBlHR|Y+lKr^U9b-utm7Fx8rHcyGr460DgD{GPf(8gtB7!)UgTGJ3d zVs7CTfF>{JDG_Q-L->jD1^)tcBi@oh@e-(*`V7;~8rBBLJ%Oi0sF?Z;!)vd@58_Pg z?8Kl138efQg==wx(EyF^!c!uo{2ImnH;PTrqV{{vGw6{7`bG7ee;fWB1(3mBo)V#7 zRL^n!sL=tS@|$DS z0Qb{vHQ^}{+DE^Uwo#u20KN98U{In2@~U4`Q9E?f12}Zyr+7+)yz1{Zx)=5f_dYBw z(=MyLV4q5$7BtzUKEDgDRx&K-DG_QxlTFFX)bM0DG~a+hNfOA17J6Cy_=TO5CJT8= zgnVeB*H+_or{U1e2sdNUGYK@4Mtqi+PyP#s&Lo1TL}(_B_^cY8#|gmz-7O4CmOuuv zwEiljt~o%l?RZLr3}UIzpg+GCwCIEVF$Sebpfz+0+UT}+4M2By@stRyp<7T?l;abC z8f67BC{+RlQ#~iK;vn|x@~fmoD46OwJ!1za0aVX8g+b3HkaA__smuLNH}y!7kNsA?CFB7zW?nGXi=1E!}7}OC`|$>zc}CM z=nRC;sPU8tDZe;x(9sOHB#bxj$)J}K$clcWBF}o8phYb$cuIt<=r_9X*sBm)lzVpq zgI-CXgrC%}3&|b~P-r+$iBQ5%+An#x(-D9sG+V`>*AnPT2~~Pm`pg2T@}suZaEsh!E1nXeSQ;!(FjK687A3{|GU$y2s!zS3QtS8F z_fARRDG{nqy`X0^w1xv@*yRC(-b$cZ)XCqMp@Hido9gkD2+g8S{@u}ga8^shC5J&7 z66iXu8|ggg^Dp>~Y8~V$5xP$6M&b?}zX>gJ`KwY#fBQ>W%i_E9;ln5o$7}o2YyEZ@rE-M)HP6C~v)4Qo! zzjFYoU*RbcIzgv*qg!{nA*8y1LGLAyvhSVoFbzw}HPm=Ygp_^nmP<$D>3zX`JA<+% zP$b>!HdZ)mLW^!&@RSHe(!I_oXbRp`kKgxTP>uweL$#pBZ$DoKXm2b}iO?LX1wAa; z;t7XtW|vR~eUL!PuOzH~&R2B=t01xVj9jX@tJkaFqT^JY^D0IKW6 zQzE2Xx@H$|VG2;myL<+Hl0aItUvkW{U|g;A^gU0BkQVKiZ2D&}7TEO^O{;Z1|6;i2 zX9@Iz>fROlcd-IleK=2v&N8B~lk^uLr$S=} z{gyzHRF1am(G=&8wiNM{2t`skIwW>yDnMiAn=|N-1WJ#fE^+k@*k@S2fTu(#J%Y;7 z<6N;Y*zc|#gZ@gOI`rTRcw}=H4qdNso)V!t^x$)O<2?deRNBIuLHQC$xjFML8{Qcn7?d!mNCF+8m#!Dz7=-iPR}FbegbvV4*N=DH zjBg~kc}lBBoJ4bBHVc}j#jQHv%u+qer(hVbHn3@VX8_2@TpH~fo_ z@&hG2B|`P+H=3mxI31vw(@YsuDuJre-TU+$wU_W4b(+poB2gNvp+zZ*LX^VlIeYnO&edB0WB)4>&~EZ2^2)VpzAqHacxhk22Y7l5cPtt zNA8UW$ipU(K@}3H?G<`$@34ODphdek@stR)y+Y%o+&G*bwRn@rpq&H$d$ZDq+sg0Z z7T|98K5uzSgg)G+SBTgTMd*RScLv!>pvH8syVY|K)*0Llc}j#D)4gt?C z^nPI70|VOt^w0D$404b_=4v$+I*#x0I{H*4B|_$EyN&cMZZ-m_!%a&DIZB}V^c$Vf z>^lX{y81ypB|`P-H@Z0g#8ZH-HFIK+lLWd+J=MZub$lbi&3|}GglO6yXOQ7R5Ybqv;&>jyh(pk+@B6PgwZljnMn-)Tg7Tu3#&>ji&i)Jm`MAyJ~{+x^D zDG~Zbvz9Gh7-EZBcg|waUJ0b^*X4TY&w~~fRY{4EvR^mxP5&X#qFK8u800L02Ge^@ z{q#%ldDL$YPl?cAdatQpTGD2KK7P?wt-N6ONuV8cGSqcwSOZ$5_Kl}RXa}7P;q78x z0yJ=(0fY8SAmw_76&*$%1*q#j#u*n_G-$9G|M)H&h#n7g(aRu5tp+)anu42$p2^2t;-kILVivaRz#Zw{_K$YH< zgeNBf`eozHpkoqfM=ou<-q)}zKw+DBN`!XgIvJhos*Q6%?~;5ObX)?J($q|2(~JH9 zJ$c4cB2-FKGuIMsAB7fm>-&H~CnS&;%?>wg|8xXE)%x+22zk-$aE;;J1^_f;Uk-y# zN}v)dl3QQT&=rTQ1zApHq3y`O$fsRLP z0BFr=o)V!b8ceD6@9hE5j*9sVIxT^epFq8|cXlyAmem&gUrL0OpFj-{I{6Wxz82eA z#7hEsMyo6S35%-$=*<$I5+ToMD(=eFifcMUt5 z0yNu^r$neet@fW3yLdD}2JiA2bWQ?gQ-$isy55HX>iwRlL@2v*3)PAV0G%G!v}WZ6 zdtL%9y+%_r*Hse$avaZ7BDC}xy;QKn9z%fiZ}nx+1qq~1L->w+U*a_Mm|&g~A$1zU zcUs|x>lo)Wo64Yz66gm#_;fcJ;9|VJ|L~Lu{h$Y5|G90lphZj8TQSH-0v)D%ovBVP z4m}2J;3*M0O!vBB5BBQ-B)izXgVV9;d=q(>8ij`3ORphX={@RSJY z(S%^919!Rrw729lgRV%R2UM0#IqH9fJZS&|E544qPxZ3(mTt=R74sbE#bU;l3B{6qwN8n?ZpRNLed5 zpYgrTAx5iqpx)o< zUboNA5TL$ocuItNf2TJQT^s!m9J)`NO&Js{fyUDUmCq|q;2S1yZQ&^q8cz#U0zRI= z+V%N#YX;qxKyzpaU+$3|0WCW6hNnbm4h`YMJZ~W6JJg*)ArdHuihFZBKj2xnV;E0~ zP!1LMvM+hAhC`R>7094a3Dl1I40W!5Sq3c%^X4fLYDayB26rub12nTRkwJGP(DOpt zag%I?Wq;Emo)V$wg)}OUsn-#pWplqXC`i~xGyx+{aiCD3tN?tVM7+cSXT+whbK z9jE2)=Ldw(1?cOpF${{3K-zRuZD(Pi0#KR*Pl=E=-BjHJ6@dWRf3Rdwqy$Q!&WWMw zT72>2xQ{#~LJ8D433~Y#$49=#P7I2YK!2z%INO@=2pXKEU(4NN?47x9YN@?ctTVLNF z0NFm_DG@5Ana8x8JNP2v&V5?ft-N4kB~UpH|L>g+#Wx;z?aNalR8GVHW+OCfL5u3R z8ZhXA1S+9LL=NjR@uvE=N=k%EXc5thv~+ymh)u2;gC0sCTRQ6wb=jf`hwj98o)RHj zI_mv<(JS~G1}gvxxOI#{@e=46)q>6r zTX0Qxo$T^jHF=6;biF?um&2^?b@xB9vA{ zFLA215KEnIy&Gy&UPn(PPy%&MUfr^M4A7cBJS9R2)H#`P=}T)kbemjzGAL03Y10}x zpTHqFHB-x-r$k7b*2t;-yK6Q;DR~nZ^i%@n(0brBQ-?RuqN~4oN`!J~J@CYqj=1o+ z-;`AhN|HcV=+00r|0*tvsymgZMCc0L8EoUrvF`17!=cIHD2(P2z8;2FrB-; zehi0hMt{|MmDka838Y7_KG2Pw2Jn;!>CvkXl4^XN4^Z!ux(s?Dfh?$B7e8j0 zJsi45zqs`y8<3w@5bDk0*ds=+7`Sheo z0G;Z8ltFJLP#{%Y>UH{n{Xy#iJS9SbRB_oqtR+HiTyHWcLjrxEI^XJ-?JeNY8MyJ3 z2z{VBpHGx`7PKhy_cI1%N}x70_D#CI)EuDKe|SoS+R)hd^WLHODet(kzZjGyf#Rqz z|28u6FF+5+@stR~QDNTv-9YRQe!SSALFIMyP6FlAR6~2uYD)kL_u(lK%BQJ@z3Jny zhxAAa5EeF4uX1Pv#gko)RH%8Y%XCFuM=@My?j)8I&!7s?+Jce5(!iH5M=7 zDG{npr*~M&MjwE(BUdmeM*=Co7t!3~;5j&Sm!fz|gp}Wl=(}}oKY*?`-^-v65~xds zx}tXXp}0lWON*yOs7r;DQF*I8+{)0!)`vkKCD1K8y{D=S@qrfou96a=TXcH=6F%t; zw5U_ceFlA!KrKqu689B_EheBD9;PzT2JE#d)X9Z6_J@LjoxW*)1>M!%a@NZFx$Bl!NS6?{vmNi&D~WGw7!T z8ctj5GJD*^%4Osmo)V$qw6$)Cb?iKV()C|3=$8b#M0am}9V%=5|L$sl*QOD*-wxp^ z5xPWo@8<_vYyn99++POeNgxX~9IA z+@h*7F>5WfXpcTN$4FFodv=4*+NT60vX#iIdf?roKwd5%gYW38~XukK{ z3jp1CJefg%C6FRiT`}x<16XpwGpo)V!rDmdNyeR3bPXv)@j z1{F!5ZIP z#O}6902Ne8iBK<^0C`vKWh6k=RE-%_CV^HE^znNz_BNBNq(o>1K|2>DrU0a}a500* zCD2srsM`iEI0h|JUBpu&G?hB)S|z{nHlh->lR*^{D2cjE1Jt_V)aaKgDG^GdE>pEX z0bbCe)Mj1`+BxvQixNFmYbw<4Hn{~YI`a=tiBM0~-A2nhMPMB$W5Znr*-4w}6%BL;$32z@`QP8P( zF@ttVpvKDIXwIgy(4t3ucuIsC({D7l;d32m(aOWk8&_Vi4iab%z3(T!m-zsIW**@w z5t>8q`ze^*9{1tRC>p>ZM+r0{N?nn+)N}zrI>kICLL;J_jBcEm+yPqjX!;BWIZ2>% zR4Xw0_6)D1CZ;?kLg%Pf5a2OO1t7m$8yK`(0%=i&%5zH#9cYnrFi(k)7FDR8Ti3?D zlFrTD7_>(MDc=MZG~ET)K$^2-(xUZfI^$XMo%kxeVGTfu7N$ zyyxTxI3jQw&Ql`vj2`9vw?4)7kNVzqn^ay$`z4SL&9v+|vlgeX8=v7R5z?WVme9nF zxFagBtP6tGb1neA-+GQgt`ewiJ`Lf!b;lum+BTjNp|<(dQ*C~{ zJ3y)_Q4DgEK&=S6aMfQ8TJ)w$N`zVwR4>Qx20%rh zrpew0!AoNRns}(3L5C&KAi9oz*BW>nAZ=Hk5}`qK9VH#=fLrSOd}!6Q@`61gfjX+! zR73_&!biEt-D*+zMP* zji*HD7OmS#?4gN!CX81eWzY!;G?W&$?7PsnEwt#yDxMOdp|r52mqDf}9J;ckn+!TB zfp&(`R)(kyb%16i^OOkf45O(TQ$ws>kJnFTkcR}Cb%Vx7183n&YYc|)lnBkbL9Z#j zWP=}RaXI&kL8l~87LCeN(l^wC7HvDvQzDc_qw;P>Dv#jM{npTIR(TzHN}x*t>Wcc^ zKaK+EeLbEMp-TZyM#TY{p8$Hep$CIbOQ1qJ8Rm`XxB{Tl);uLbg>*8!`5B5^fi=^{ zGssH<9ir#ag~D3b0m`qE5}`x%Jc?_%36~i3AH0%5-V(@wihEuf-Er|zeSMx1ApI8?bLlIAjkOwU>cw%pa%jCPxj%Cm}36w*Ju1%7*6F{}hcuItF=+OQCT(lEf6d#by zpz{*wSUG)lcD@&`7Sy`VQzCS%oK_1i8n6_gy$w|Usl1LZNT8{-mZ8ZJt=-TfQ%#-{ zp{cZ%p`G{p!vHN@r^BF&5=i-;`=0H8;`8x|VJ^CHAsC&2h4Dyvgk+gdKr~3FE0F_ipiBKf1UcaTjO&y?8 zXIlncl0d3-dXH{<42w=_`*=!(RO$4-zs==0LLW{t=&}T|pq^@1l@K#%(VLGvB|;X| zQ+2ue0pA$WcXSAYu1KJzv_^Hp#4^08)-mBJ5n4)XRKH#>!+M?Dg%=F+lRzhF-In95 zark1PO&57egig}Bt+U<-4#J^RQO{@4RS7hh)+9V~Ig$p4?pu|V2o0t+3FgnP;G_J2 zWs~NW7p%VoN~ie+zqL*{Sl+mdr$i{7<`-T*GWr57suS0TLDwYECz@aIUOf)ax>r?F zBJ_#o7mj6{V~d`4p2DC23AC3wkl+5k#vKN?y6}_;?WGRnIRB0Y&?0+>wG0ZBK%rmN z6+_E=4}#yw-jSz7DDzIG3bT_YD;seZV$HN);j&+JS9SHX)g8J%2&9_Y2BH42HliEH|S)zHh0$?ICK-w z@{|bOppzj-H3EAEr z^r22oh26_O_zKJ#Q+Y~+KGfN5wE5>r{Eb|$bYM`h1iC`o@t@@!H-Q$-_v0xMxAO(+DgUSNoS@s z1}NE*r$lHg6>nEn%f|81m)M;Q3Y9>E=*6q<1{oUw3V*;;A~c9zyy|dmAU00!NYKpC_pXs*8#u3k?+!BZlXL0f`m?w^MXkG~W@ zVbFaEG>K+gn)m4%0WG>)!c!tNiDp}#erRX|Em}G=mqD=-sLf|}h4b38wE>zui>E}W z&1c%m;4lbhTS`LfwyeC49!Q{6DzGQkJgf!Kj4+-Op;Rice|UBbx0}Sb(qqs=2{fG+ z9zV<3xfdWWZJrXL>9p{8+n1VHj{ahA!k{<_-P`m`HNmb(8b64Vn?^%_U2-T!2@$1Yvu5dC8F>+*3f&{Xq(L{|t9=IO3$q1ej zAzK?k_~5gMuI+vUo~_=F!~5nbjr;y&blhLCF$mIt?vb{~UoAY> zgpSZ=!GZ>V{QwG_p3I=<5~w*XYSpY2{sN$#Gk8jbn$x0IkKdO@0W|#jF9yAkKu&bQ zW?pc>8L`GUcuItv=z^VHf9*to3hQZVS6)YH66gU9|LZhz844{@YrsXzLlv{AIq+BcZsz~_Ki4r%@}&=eTwjLos<+v~Q{rDW$=Au~ zRNvMc0P3DTp8Zv?B+Y+8z0E<3j`%|JyT0Km5qd$r%||om;>@GI;YtR*mOxXfNM5ht zEH3a~GL)x8Xet%S_jK8aBX~-L zZqhSFeScYVfZTj-8I&!7>eCvzw&wHk;kDxuPl-@{S|j&)$SgyELR)$;C`SUFb5K`A z-?hFAE&9@mr$p$S1Ff=nk&h#Sp|&9m`XGT;)9KyA=1X;enr!DO5n4^B_wtoJag0;{ z%?k#7ltA-nB4fi^=fMDFR!ND_JetT@`r{K0cz+GfXV51Jw1rB_U!v}~0yKLFPl?bL zDk*<0G`|YJQKQpM+EiY!pCynr?E@-a{unnLWmQRukTvZC+S)O!DL{`)`!MK>1Tv%A zb<(+XtQA}-<0%m`quO=7IkET~&0jQyL0=`14o$32)3Nu2LpNM0I`X+%k(7o>7PG_7DEU1zap$&AebDficTR&E4yD%tM0!^TrTWB?-eb6G;);uLb z6R764WyEJchpgzxT(Q#fLgww*B&W98umj1 zt*3ik+Q3*NfWB2piO_nw*UkB98v)S69-kTXQv!XVBKe&rCAcQxL{FX)p)XVDQOr8>`gV$ct@I|)qOjW53@VdAyJ&xk-=#4};m{qf!&4%(i}t6?4(+=aAlt=m z3@VpEiB$d28rR_pK>p@DB|?c*{RnTYJqe(t_W~GHA%Vha@597jF8G9UAKOL3R>I`8;ZU zph+)))LeKQ|l;A>}|UHm4lldU$Dy z9)osCplNrgTv-@?8CrC3Do=^fv^%s+SEWZbwCHcJ346xelv%iLceA9e8JDG|~o zXph#YLV&C)${FM=fs}jkVovB-K#OKotN6c^2r2jCId=YX4WNdzwcAx*u=^y?(%N`#jFrsq-Lrn>;T8#;tR`z6pVx~cwiX*$+|itg}~2;HKa>fi?#ag{}F zEi(okkU%Chv}}LO>Ngy^=T%Z7WI{vBC1*3ILW{21Zf1~+1d60}BhPgo;ok5g+j&Za zB5B=7YIOS&fSzR@WzazhWJ06zlJMm7=nR8+N`#uwV0p1$&)onyXii|z5eama)+F@V zo`ReHW;Em}5xPoi5?qsfumgEu%}NFxl|X%HvgyS0pBte?mTP%Rg!<5AliIm0DbOOn z3l z89@ss4}{<($RMv+2Az;Vs`Mz=*O)#HTC~%fr$k7V9_7}4)$x_g4ODX&bW#G%p+onj zpD*@;UQ|hm&>T8+Q(tz%9XBzHRXS8&M;;QWCp}?3e}8EQExKUNQzF!pp0Kv1?{U66 zF-C_$rzFsA8Z7s|=8sikzk56-LbqwK?057&u0o&OPQf5g36!kdrF(TUc8O=U=P40N zrgbBYJ49}W(_7bJK7&q6Am#8sS!?+mICOujq(n$L{Qp;_-T{D)W!W;wO9E}6S0CKU z-1sj*P2cmB2yLKOAB>JKoDI;_(H;!)mO%F5v}HTyI(|uJlL=3WkbOAqcCVH-2A~U9 zLl|^M0`;OfKFd>MvAPxH&r>4Qi{|(e3VUHEKcQ|KgU(7IM_OJOXkdq1droQaln6P} z@v%OPa23?RqDvfF?X3aIt0O(z3o)RIIM!St(J2dfu zlVRHKDGa(Of%;Qn@a?u9xXEJB9-b1R{!|z|m0g5)hRjcE8RR2@JgK@h*Z%2HXi>yx zo)RHXs%~XDw8ajj?id#a`AQ%?T0Q$dE(0qk6;)Cqq(`e~e~bvmo9gn5ehj)KfeukE zXuE2dJG5w&4^N5EA*uy^Ga8-=ztMsk2@JX{fojomj!2_1;{Y0=!c!ttiS2AZ-VBI^vN=k(0l~SK!w&!tx z1{>NjC_n-^&^pFVC-SfZ*=#6JiI4-WWBjwFE52HJqNf*w0wqu>4O^P_eT=KnwNLYu z2$j;XWsSD&2WZi*zjqmQT>_OCt1EWCoPvFC(|n#1q4Hu{+>-ER9zZoGr!(k=1o}pG zhSOt);!^7ORZ=4Kjp_^v+f(=gs8KhI8FW(u#Zyl;bRs%q zHg2HRx$=Syl0Zg3Y5r(Y2b{uwTO}nzMn9d5jxR~?28V9o@_`JxC4s)@sVn-{x{ogv zG+Dt@BJ@3vw)PZvX$(-?`!g66EP+J% z0tM4{V8g-NaYih@6HkdyFl`4;-alIhq1|o_3XwoHs7t);)iW&71nuD|5voC5V(VD_ z>i`*j3t&*F1e!|A3~U>Ys}F~+Q!Y=5&{SGxa8Z529e`F(PGry>38YJ>w{;5>ENEPu z!c!uos~p4nb_;`bjMe;d85A}M)-m24?_hYr_7(nBj<@)K{?B!cwW&zH^V@4I?ga<) zl=xTGrXqQ^r)5`YeyzqDT`DiSyOQRYQBlIwV^j?|&OfW9M5v6461Ciqz5vK?tsaBI zC6MkXbw!A=>i~duSn-qy>3*U~+bM%y0aP#1gh3G!D42@m`irNv1}MKuN`!)`NS8Akt(~GA>XgBp_;)mRch87jOIx;9q0clX zPc!(fdVFsVEqXGIr$nef&EP+1HsAz6F2R`$x+j5@qc(kqrMO9X-EE!{A?2uTl9sU* zKwFztFzCJnN~P7aJD>RDs=u}W@RSIp(&|~YwohLJw0E1fUgZTFD}gT3eD|u}Z*f`F zEL)xup^G%%{kY=u3V0(8cZr$k7(8~Rs5b~}J>cy4A;oCF#}7p(QsZC(KFJIzxfG=?tN-W8?z z88qkeqYR3dK*sbqZs)NFdoo`ucuItf>2Vx!Xj4CE(S?OU3`&qd8nnW>iT<53Xwj8L zJS9RJw8Hu6$Km+x6t~D^20fBM?`Vt@V>YupK-N(_B|`6LjN>wQCvLkg*3DzkV+o{e zQIFvp9{}{a9Z!jnvPEk*8xMv<_sdSRYvpzHL;^X{>D~F##^nG-+w+tNInn8DR_=>s zmJV-vFep(1)u0{g%fhUz04k`G5}_KjW4+BCm&wqgKSL)l=&1yHL^CD}i(PTAR?aY< z5}`*lV{+81{T6^qy;m|QNdg_9QCnV@&seTZI>S>Ubbv-}dJ6)^0c26rnL*DaP$TLE z`I>03Q=E&unjQk^ z^}|>OrAVMmdX(4r?%4&PPVqb?LYed^SLq#>4$$`gISfjbKq+VG^#w~Zu}E$+fTu(# zC%Svz>vo|CTGVL{Pl?bcx_cWaIzECH z1w_tg&`Sw)Foc#x?bmGy(7#bUB|-;7=(~l%>P7%LXm4lGD+#ok`rgT#?%>I=q%}{8 z&}!;?yKL)*XPv=D4+g!KK+3|A+RU^K(4smvJS9TP!cm~*Mq_AE`m+!QrAweM@6{Ds zYOWgwkbg2yiO`q#PDW08Iyi*a>y^f!Hxg)Zwz{HlR5Dhm@~fmoXfdtKd~Ey~e

    WmC0`7Hz zX-8EMv|?$m1&tSMg8&)^a|NG2&cp>ObIf>(Lc?ILAj0)3(`6$tV<%u^JqhC(l^q>=c}u;TLx1|1eajF@d@XubOldQ$Oy(bp&v5eBgpseFa|jZAXm884ZC982SFR}@f3xm zpEGNwJ#!|4Rwx@X=%@f{2NSS&jP^JqXk<^GqEI`SfDND3b~l35_HJU3vjDP!%1oCQ zO88MeU>{FW$Olh(%1^G(^1wtiCW~p zAcR3K0>~St{atz$`5`D`Ax}}r`!0M$uex?6f(ovuFvwK^^@gEkL67322+F&`QxxhA zLrdq&F9#v0s(m4Yjtihm(1A>e{@fNp)g5?>LYJTedE!<&Zqf5?%}pDxBR2uG5N<)0 zFQ1G+P|9|mqR>LP1s$#3wF|YVZ@LPD+y#&=EM(lj*ufn^GLLzRLbkAwaVTv476f%0 zHibbR0%#FTvzQh}3_{TNCMgOnf@zjS;~gCk^vBtZL7oCA2Ckz|#;Z0U=+`lxqEHN6 zN7gEeco}+5sTG5~1W*NZAjcPvI)I>FWjsZp3g|#SADV=RmfF+(800O0N}I_`p4%6+ zMo=q#o}y4`Ge;ez;@;lq&^-%`X3z-%Bz-5$p0k@Q5afT3rzj+SCrsUQRa}|5y*8Uc zCk0S$t(>Hy;<*BXX8z_W3gy;1=#=#Fw?-|xv#5$eJ_1Mv`gNymZ5|@%D|Q3(%)qYIp_z>MkP&`3WGm-?07a&1O8$;If;iDCG9rLFZ+s61;+6ooUXX z(*no<-sn|2{QY)x=x#sbDGC|D8@+N5x(z^wE^(kMgU$$`NVtyl?pWeQTaQ6JMWINz zj>7CFn%2H5^B!_v5LIdHZTDkwnK-41T*J%v$7eM!5v1y}C zbG%De^$kx^=pHOK73K{NMNrg)FANG0K>c8==``)R#i&K0T0BLeez4Uv@_fh_1Vvty zTikdZ1qvXg069rl>(nX)oeJS83MmD^C%Hezjz>^|f*OO)37}t4N^R6~XeNSkTk#Zy zenBa`a zd**m+&&MVy3Q2bg>{qtO`z1H@v1QPC0kjdGu;c4waR+jJU!J1SMtH)G(6+5YEgEm_ z&!7te$QRDKgsXPAVDg^>JVha2IO~FyTIC>U%RYYenL>dgnR~t2%vT_5G%cNYbZJyHfiw`h1$VD?5j?XnaW z<4x#Ye0YjNvT)X2jj(uzpzf93mNZ_lp#sPdpxMfslTeEss(6Y*h5+^V*FJ!t8FMum zbXfqY0_3@<6E1h}JCCO*qzcgQZY6lCbI5gL289WrDbQz_uKT?MYEhRPJVl`?&}T4O z*k?UzQCTMo289ct8knL9&F=XZK~+jTMWGs)qFKLw+*Sk?ulHck6#-P-N?wvwK4dq7 zDmU;Hg^F7_>O}vu6^~&DCtPFDRRJ`yPEHb;rRad5zKJ|Vp^0_yA^0Z?x1$yn^?u5r zYXYb*oD9LmOO+7xtq)I8s4tuhop*nCLr~;_Vg^MBpgcGkuKVe4LXgivo}y45oD9B- z$Mg{7@T&FF#_K3j0QG{Uw`&d$)*)zK9#2uI7c9N4ULI6|pwMxH8FXC$1;E|={;{{u z5fn0>rzjKvckc_a{M=i20S<9d( z0i+Anvu(6*;-c2Mr94F;U8tUwUmlI;f>s*XGw7xOx(s*k^``$*7Nu**Qxv)kckgEk zYIyi>9~#7jhii><+RYD+i{tvC?x&H&8AaV@w@(J%O^7Ewg8fT&#T9P5WL_! zNr9&*B>kRO3n#CZ=r=0aRKTEU0W<|(^K#_fnM%~6x0`v2LQ~*1FFks1>w=*D*?$=n zBY@^zl#|#VsPBfLZ8yHvL5cHuib9v+WYB)lb^~hBj%z0v zbXNe`z{H?~&;3vY=|u1pg=}DAuyEQ0+<~0cA&Now1ke{)5i8sK3EzU|bmS=teSsCR z`E9%pqZS$L%3@H008)ZOXPIw^_u{SH%~KRofTgE`h3A$rC{X|y0(jNv4N)|w0Vc0TK!4Hr43`%&4LSJFn zGXCJBC+IhdU$}`uDFW#DW%xL9eb8>LNH$wy&WPfB(ngFu<1xp6S{&>|=DTk*hWcLfc z!fg3X7D1&Wn=fy?jvfghC78T*eQJisM-8KRib6^-c^j~1$3_f!t1u{C0G)=4!SGoY zx#&0gc7mrUbQ&rKYh;`EM|F&Wm6I9t*bKdoad_UQiKkn{tE0av>o5P$|5?Yl>Kd#L zSNFi>h2ix)MgOW**Wgv~#sBkJz^jYQ*kAQT(EO3`TGxqBF5!iY1B-czLL=d|t_ndh zxPze{YsH|a0!R%`(J+05bab4@#PJk`)Zi4gcN=#c9XiEsehkVGK++Z|UG}s>(B~#8 z3Q1cODOvssL1!$Z8I&o2(%>l>C^?I-?!vu1MWHl!O19R}(?QV9>}&=-6F}O~leut1 z3C|)P%i$>sX+uvY=8sggJvy{(xbvWqfCvqg45QoK`q)|&Qlc9 zghSUYeiU8{gB3>R|06@ZTNI!$=rAZ4gbhf6dHIN z-p)~Pim#(9lfN)1PXH}~HK3i9)_Ae$=oFr!&>~m^I{bOUSJa|+{&FiDucLecB;E7q zvoiv>=t=-jQAoPyv0eDG1qeDW!A4D9FVi6HewJVl}1u*YQ2`AfLgTO3qerVH@*z|D1c_cWQJ=( zTmfoPev=f1X2E2J;qM0A9d189nn8sEXg;jL&h9KzgrLeMDGJSpHQ2nh@A3X2(^J_D zDiS~|;WxUcuG$4b1AKXkLM!1n3RiFpMThQVLnVVg37|L8@IC6lekTz0s9DwjrzrF$ z8hTInXLd)>rWH!78!wB`0>~6vbSuDlF@h$nh3QpuH3fVxF)A>9ZBLt1gI#MEyq>H^kk4xu27MPmEulYH*}ZBCYEj@Do}y4o z=np1rS%$||GFo8_Di%OHU^3&!-2a(2>25YLVgYl?*BuKr^8$ z?daXL1|2%rJv>FBnb4I!vEmu7t&4tX!=MTQONukg5Og7%rzqqF6HfIeW%p5w z{Kok+s8Rr3gDv${Sy%Bs%MIgsibB_5OMTy)XAdIisAn95ssvDL=yl2%s#u zGc>Df-4Q{hO;Qxff;+iw0%xWKf*|iiFDO$gQhaAxQF!rzjK&mC>c^FXIa~Xo3fW{s^FU zFwbYa{TqG_PnZ@@QK%iv^F|)eE3A@p6*4wLNfcooPFsqR^RmSPj!qSc{-d8;Tj!Ab>jmkdv(bT8|eJ zr*Gsb3U&VBpmXhBKq7)VrnFwycpcd;`|t5lE-btGsunIohwgGJPf;irmfbQoR4zhL zf3?93IwXKnVKm>OzSpJ7Ht33{v0}V?|p(8bVLAcfd!C`E$fdVDB=`PQD_S+fDBpMwhTdMs}mUH zAb^~q?`<-uxdDPs*6@^c&3_Y{?)O0h9#0d!sz`@G6a(I!{q333m7H z>*d@UL8^`?800E|9>VF}vr7+Ubm)FINm1w_oZc4G>_#G}chL<79Tz~-ul*FyE5)Uw z%|7uIg`{8miP!6N3qe{_vKZthfHGh%=wYu3(@=|KB|Jr;444ZF@_LA;(VR|~G00s2 z?SfSr!{l#xzq0EYo}$n$Sf!~pv5-eC8ZXm+L*sSiA%L{tQC{_86rSB`*PN#)qy>-i z?ytV!CA+(3!x-c#fEL09TYpP$5;}Bl%Xo@H3*mzOH!cdlFmh9@0fW2*khJjsbi0CV z1bN5t6osUP|Fdpw!Sf7>JvTAPTL9U@q|^T83-NqiU@xAckS$C){hRavml$;2=ft2B z0_Y;NXwH|x<4}uUHAzwEBD6?$T(=-}GFZR3$e@z~XbPNlX;1H6LC~C+JVl`?aMsIz5?hEe8!=fO7Iv2DOd0mh5o>29DXD=;LgdWg(?j4 z6F|QApw|253%t~@(UhkscG6DTYVSSAl zbY1|Z!Wz48cgI=;`Ks^~g;HUS{qD{rRRpQHnKS5u0Lt`+-{{Cwc?2oB^Av?Ly&ZHq zZT{O8L7o1(FzBKHnsgdIS>tCo9ziPgJVl{Nr(sU@r#rr>c3Bg~pb!Cc9oBC>JWk_E zw8?9EibB_6{dT-*0e(z0q^B|Hk^u6)30>j?ukjnIiXZb7g}iS%=v>~II~W~0pRr#U z6e@s@!=%%JE(`EmTe^1) zci|N=x1~HqA?coSJ(J^W5Oh3dC4;UAAP?!VB|Zc1DPM7irzqqB!xpQ1-jfmJ(#@7Z zR|U{Wcpm8`be2VjZcBHbqR>cq9$jo7BZulA4fpvo=-M(=|JbXd+vJu0lcLaHHTE$7 z&;MEfXdMi1)C+lrtDC+`hf$2E=|E+PeWkXGdBv=#~Kb2`?l3v#2rxK|4nA6or1m%Lt#>t!aj!NH-4#-4;MbFrSnD z-Vsld$GP(qg^XZ6$Li-Sd@?wHxyGPq0i;tSCmH1VaW-nv(yu&4A)Oiroq?qq%TS9p zO@GRu7y;A+E?B$6WxEkHNuQ@E)B`S9kE;i#A}BMom_c_0kQFRLYYmgZdnIEo^Av@w zU>W+K6Xs_Sv{g}YYvXkkD}bg!-)w4OXWUVr(}t%gG!6P@?G66t(6uoi%%C^{v^N93 z*fTOj5gocV+jxpXdoy6bdtd=xSnv5uo-*d>ZBrzjK@3%hO1y5hQ#q~n$hdLV#IVdCgfz|ffpa&zM;3Yo&hkxu_(8mL84 z#U~i_Pyls+$HZ@w=D7I5>jzI!r~^DE*8VM+f}l4uq8OAWfZoDcmt6PB4z=j(OrE09 zTR7`Rd#t#LAg>Eq40KHJ z24x8#D=4L|seXzV1jn4`DGFIZDfQa*1MomhQLd0d&jnDAM{*MF#rN^B<#m%3g?c=K z{j9}ho6w<~zM=W{#_Q;X0J4C7-SG3N-O-_&wUMVNWC8uUw|(y7QMth*6$ZT&KxSF+ z*_pSea0jwqI!{r^EDLITK9%53e&<0`7?dr5q!ne0mb8mPE&A9bMImWL+0I|%ywIW3 zv@v5)jsUvY0G<4hDGv}NYs*sF?&?^Cyl?q>PmYaJFL8G*Jib7eb@Pyr`eF(KkAt0JTc>>5K37$uK zFYzp5O_LObT#_7g)>-D`_3psG*$m1TKrLYmd*yh#7HW}YJx@`nC5&Mg=Ka9Kmd%T+ z81z~IRYHMETiM|k5Tw0?rzlhj1uDi`u^-T(ON~<6(RjhW5kOfmi@12s%Wnv(zsXY+ z%7R(MwsFt!Y<$bEqZ#y80F8h*j&DqMUX7slO;QvZ0dE`+7_NZt41ISQF{nTQ4Ta~? z<#RKeBdD@Tib6x-c{Fa^ApGFd%r$4wI}rp6g7a%$E0f{5_399|h1SSZq2v()%ld#uf7vg+9SzlVMPP96G&y=Y3&Np#XY)8{RUt ze+ph5_MgvF6ncFdUQOY+R1-l9Z^-RzypDkxJ8mf93g@w1d5S{+!W7Muq&u%si-sT2Wzc5jF0aOpKv--Z?`U`>v`0x~k>fv=(B`4p_ zLy&$|9D}|KpbIdX`2OT3er3YAYM!Fd1sF{f?eI%RP}t&p1{Dh+>ts2J@{dCfs73Be zc#1;S@Sc*9UH+ONXlZgSgMJ7g`EWVO(3`TZ2r^3HDGJGlJLoiwE5nb8LBqORG+wYJ z0%-hoc*17-TtiTY;XFm5@z>$I?%Q|ZZHz-*G#OMXfO26Pt!T&0e-Sjnm8U3_3)5(M zDv~sG=!O*=GpI}e`9k%qt7R$fe+c;j4BcIVU`7hehQ!k1#*(q&{npn zMGp;lib4wt9CV!Q!g2l8ol71JDi=V~ul)>dpAdr%dp#XN?rWbis8RquhR(^TX0!2QV);6rqR?aMs9be2 zo`IZ@Qp}(#0n`U(x0b)#`xmvSLn=>Es1M9;?fW(N1!|GjKZ?5=ucK-KGzwbuv&oe$= z0Fu5}c;>j}c>C4!*E~fb>3fA!U$}dqL#H%p9fN8GP+O?XEb6U-KXdo1Ns2;kp)&LO z+>(w6iaTe|px**07+x^2v!of`R-YBbQxpn@7Yq!&?T-h`Mhd|UsuMtdFnO!wS&C=3 zj%cFO)=Nckt+asQ$&gAToRH(sy}0_YRm8S>`kHz3Hwk*6s13GNKT+`{p7q*$cIAlv2ty;Kqb{Aa#={2H6Xsj7oTidf=iR2t;=oWw=g_tY`9T7k)VX!WB#RZegZF!19zA(Eb>C+b1Gt78Bj6qHUC=u%FH|Dg*eTGhNc#1-a zP*?Afse)&>j*c>9&`|-Dp&&0=wfxUZ^c$@l%~KT0P;k__u`+N8IvIu>+r%Je0rVWs zy1rLU3lUV?Bt@a;aMtO{AICe_9KSg+=$HUn2DhNs_k&yzboVl`|DYJNrLhs;(8&8baFa<C}uC>2&)a+XcpjOrLyc$hKBb2+MG+?&;B^7eAq zK=fCc`SJh!pLL93FyP(zwC)-Dt2&+LDf(B1!GO1}l|TN{h+eG~`>VVJ&3_A%4G7w{B?I&}SRL^J4w z07`>#=)e8kGZ0kOBt@Y#7>7Q%ncV?FJ38es=%fG|4dc*H7Mf)U(oy0m3XO(wXyuU2 zUlBCVqKZL20_ZyoJt_?44j^dwE}o*$cNluuraqg2ptKiCdmFFrQvzrxtkd1-Ilu%# z{x5loLPMqNbaT$)8LIaaMl;A)0NKI}RpgM)P6)cI#Zwfrg&C^0BX6mp7Tq~x#2`Nb z)BsoatsReWZ!`2PPf@4=u5Obho3#;is?MB2rv=bNm^kv>k&eeW+y3wrg(kwpQHW|) z5rVESab?gM0W=3@*ADL+@gS6 zJVhZbC~EC>W)-d*nbPqQgZu?hT`6q4j(m-uk_Mf4ib8dzFn68t3~y#h-1CJ&0RpJU zJvqs^sx>RvO|IO&#_K3h0JVjuWZvoR_!ji$6;Dy9Ej%Ut ztGk>>keQYmgU$&c>9O&UQWTiCUCezmh@00>~Ba3@K+{;gv_{2A-mjE8H0#d5^(wq&8b@%b@cD zs2=vErWD2DJHzlLJVl{;*q3@ZZQ23UqBgPq47wnIEaA|lIec1zexufLJVhZ(ICRa% z_B(~3ooew6x+s8Ngvv?GE$i?S=d=MlMWGj=4m!#!cEuv-uYEp)LIjW#^q(&!d|QlK z^y~;vQOF7U&t_BI@EB**_gV&B5DtOt-C_hK zZ{R5kNlVv`?7oqPS~TF12ZOE%pkx?KHA_Dij3AkGo}y4P45oBSpR_?x%YhLLx+;LA z1(U~>KI0XWwu5+zLehfCY3=3l)}FOTo-*i~0Gb7ThMnOG_^i`&;3*2tf4%fb?JppHbEME&7c# zr|}ep^k4{I(_uHRviN*%FoUiOAn7ak65F-gh@dY)JVhbtEBMN{X5d1v`7LHJ=!O7F zfGOofIY;~wr+zJYib4r6rEEE)Y8^Us&a2lkC`tfrhK2PuF}`>i+I$U9QD`$PtgjD} z#S0nZAJ{YKrU2>y>kE4h@4(%yKTT2;>HzBtpYBH9K`n~t7tEkr0_Y~p^qQ6*#>>!F z{dtN)H({ptpUcnjcF~Lb6B%?{0C~b5qOrY_@V#!g6;Dyf6ZQ~kHax09En1abz@TUW z6b{=)64pM)U1H4~o}y4VY#VuNVfYV%>^18d6eED7@41hnOvIrzoVB2BQg!uzsjT`88S$iWNX}VO0LliE6xp zfBqLwQD`oV$~Qaq$KU__FnS@%XTgE&^A?ZBB(XPgL`7Gv|C4=Gx zP#ZWIY6^VuMi~DHo}y42I2rO!tixM-N)%5r=&k@d1Um%=EN?pvwJ4=bym zp#|QOu+2P*LH7hu5zNL%y|c0ylpug&;1(2@v>tDS z`Pw8!p%}OY-3Y9Bj-VmxKN*xLfO6~LRhEbS@mlJyCMgQ#*1=nD&O6UUP_<+G1C7_w zeF1b6_C9QArrHM`x;!VIqR>s)`|wD~U=f1Gd>h7~Bmr~|o<|wYBJdbi@jFjZ=o~zc zJm#(Ig`h+OLk1-ap#Ctl++$jT7X&XD@)U*o!_YF}j@@ts*<9JgpcDbL6CmY5dxFrR zTYr_OD6|uxt@_nE2=eW8ltHNis8b@mO}E?;Py26E;wcJsN_5citlEfIwZ7~OVbB8s zG!Mo{3tj6Eq87zk@DzpS!T89+D6kW1QDjC6gB}W?&#(-=cgwRU2s)CNT>$CBuqCtALR_=G@e@x`NFRnRfrG~5O%}g&r!eTT z0NMd98Z^ZTpLN5g@f3x2K#TsoGG2>XH0+!igPsVW6o6*VG~I#@ooWzIQ78o)ZHd7NX2zFi1 zQxr0U3wFBv3p`kE@i>}6nF8qST{(&KmFakjrmRVdLTB$f=uEAVUyNFmI4Flf&jgSX z)SA9(=8yNZ`VHnO3MoOY>524Fc+zR^ktzmd2_We=TICy_hN9mn(1E8YB>hI~^eG*2 z-H6`@C7Z?z_PGFB4g;~aruusjwCy8LQD`{~#4e>f#GpgBdg^Efy%0dDFj(FrKd2>w z*6HvRg;J#x?DKQ4BFNLvh(Rv}P$Y~duF3yTO@ixbo}y4Bj3(w3KX63QpDJ?(WecEU zn1NJ|ZU{pys;=fK3Khc)q_y^}o(Os~-<3f*0%#jNCN%ff;5pTHCOk!~O zLW`CbmFz{(kRqO%n!Vs71Xds4?iZ09p@UL6@Jl8E=s*X_BJQdiV-@Y|{VK z1E-zRWzZV|G!1$|#~Q4nQH##`@)U)pK`+R0<)?0_MK{VJ9}essAiljv$)~ zo}y59C{Wq9*bu+H@S=$=g9-$YBJ62h`634Ic6V98QxsB!J*}5_=UAf_$;A3I=$!yE zfd}7`XG(b9yP!#mLMHIw>lajlFId~&@eFz|fR@5p*KONodvxeb`|uQnmcm&#{C0c- zYSA&Pd)8;7(y)1#Prb9>UMv&cUO$HSS zpxdz9ecN>lV+8q};VBB;hTZPBC+|!_(1PYB4EiL1is6NI=U3e8jG(X1Ncu|Hks(Eq2x{#;m_a`TkPLK*`|mrq5OL+|nwg|grl)N6PRUO&p| z5X_*T0%%SLd5Kfj0RwdCZgu1-3eD-@sAF#AfUhH)b%_ir7eKwCdN$~M7rY|2X+2L- zs5ey4s+&k~;e^_w0tQtGpc63FsU0(>H)>I3lN5zcJaEu?JXsAdR4J*|GpJGk`9c-V z)9Kbd5LDD8MIm3PqEXu%?TzXfryc5bxbcFm!gY*>qiZJ}nEe&Mx98|F{-6J|j!`}q zmd^rq<3g`BE<8p5D*0F_xsg=jjeGh(wAf!&Eoi^Ib#unY6OrjjM~QO{>7)L<4m5SkS>hcM${InAxQnaC4+tmpwne?l3^Rlv(OJZ z{sK=?=yVyp3@y72z60-Tc9KE00w@k{j5&54aJ7=D3{O!g4sMLCY!6ODE%I6$#h~8; zXe#v0Y;p^5xAxdNo}$oH=$ow?s;+{dn)ECN)d`?4cVI-Y%F`Vky41%!MWHWuU_YI1 z6dnIKkSXwkkrrGF9B=Lb(w=qyaB%A%VQPf;iah8}+pZflJWUD)+a46J8i6?w$X31VK;CLl|TyfV#lSqwByc z+Yoeq8&6TF3#>f)`k3PhSjz_~3_2`;CjWs^o14=H1RZ?HQxux~$3bVuzNcMLi(LOH zWRSf8x(}QEhmHP(`_CIxd5S{!VYB}chi$lnk#5`kNaJ;ML;$^m>B=dqoc}>B3OmG8 z6nY2KmHn%>;m7f*uPO|35I~cz!y3?A8~oz*ci(u5LX)6?|3pV;yiRx1WD0{E1<;yv z@InTO$t~0(=LI}Pp*82=TSB#u*P`D@Il_!VP68<5Comtp$`SOTNs2-VKcUvt?G66^ z;L$eL3_2=+&ck|lh)rF!ANr#fIdAr3kh1`izL4Sfn~&NETC#f zIW%6d#|6-EX*tK+!QKeESH@Em8V=K~Jo{FzAE;Y5|>m zo}$n~=&64FehAMZo;S8-ke>i*kQP65cEZCJXA_>HPy?*s7qvQxOR4)u_%rCV0BQ$g z*bd)X;;O&jO;QwU2V+>1NwfE$L$|a|JcG^%AU(Ky%egz@+8)ieJVhZrxO?vjmkdQv z-Ija?ofSY9utjd+`vs5DZ^5j|u`3oS8zj6}ASFh|4l#;?z z6w>$$FHlW4pN(2HNVU6D;{_WafQ(^d=IaqN6A`4W##0nBhK-s10)FmCkh6y-g8~K6 zpEGijO%18I1G&?arzrI2jDwDR%o+TI{aIBuBbjX;eDD(v;$-}P9Ry`s0Q5z{xyCp(3ao>C|x-E^3|SKbS$61W-R1{vUGo z#N`|w0X#*aelYxxbyCGuf30iv85AmjenFoh+`<>Xk5TbAPf_R>^clkY^sPXL&THN} z23;0FZQ%_n-xV6d(aB&opQk9)7T%yD_hAek{>xmpXHb{`QiIJ(zpqcn-yi(hBt;=L z*sRpb`~FzeqOA783p#OqE|n)7!)ahZo>GeqF>Sp1YIrXDGJ?$@sWvBfE$9G^%pVdx&Rsv zi%lE)|IfVlju|{fq4BWTH1_c>Cj=>7vSiQ=0dxd4v* zLC0F1WKfgWosh+#TLS1KOm(_hP5+3Xe-n9%LLXtOQ@dfbB|3EB%0C%&TL9(4 zuBZY{n->TQ?a5OV%7tA~s>xaSI@-9a{jtXDC|UsN!HDlw$z=Rlz&X2lib8rY;)~he z5x*@;J7*Y!Vg%5HblAq&dUh&mk!>zdQD{OsbRd63FGh!MQ zmg9MfLP@ZbI_l`ZXAyMOYZHTF1<(?hj^35D2`{V{c=Hs6mcVrMfpP0DA?RVT6NBOe zknS(|M&#pRcp!H62TxH*_m_iCkJ>c+;5#}qghBBF$N|>)mi}q*LcfvOES{o}1FZ4A zTs=h_wWu^Cg+X@(P$jG%t(%^Mzc_#U5>HX664s9rwdW5&kXq|P2Hg`tKcN`UrLZ$z z1Zk(pQxy6M#dz~i9KDAihb_%r8n2@S0VMrOLio{5D(KJ!ZRIHnNxzcdxm#lgg5G7Q zFep&~$v`JR$Uh6uoFruO6oq7Bl08-h0Z@Dzn|;5Vw?P>T;;ZK)N5k_FHusK4rH`gR6_ z8p?QzLYJWaYMa@&=jhNKGx1|kiU4W@!8f6UGAF0pa%kI4m@E8oY;*Y<#ye8ib8YX343Jr zyM+j<+*!q-hXN=DZmRdAj^cY=yai8DC!71pbRWNc#d7dy20ao$^I=)XY3Kgw2r^RVDGJSpWt~w|LQ>G7%XT(m zP`Ut$h5D-_L1kJ9x^|4GC=?6zSB@1UE+fe8hdF~D3!qW(j)ja?*~bucx`d}FGz#9a z5L0&-mw>LF<;tKZ0!aEPw6T)&9T2q8fTt)V{S;ao*OZT_MZd0vG3coP>JJxe=}c8z z$M`marzq4PE?BLIqqycozw;voWeA{&@PQm!{lgFEg96JhZbKL zlqrBNK=tg)cBy#jt$G(vQRo6x&&GG`ig&KXJ&`-!cpW_xKm%d;zt+1quAWVJ%2N~? z2*dxnR{NKu76pt^V^Ed=@`4fHFyjaKI`SCHQxx)o5nsLHj=KmlJgLi|=K@GNlm9zd zRu7$Z|N8J0g`_k2>&>3v@#~BZ@zPsd$9M+i2%sYsu=F;g8h7%YJMk2Sj#N14e170ojOrLySmZM(cLjPMV=vtv zlghsyo{Wz3#e@7m|7RUzLV%p4b-4+il-Od!Q}nM&2yoE(l6AK;g6uxmvcKw;p!um# zH!@0TUm-fq8@})qg;JqzB)Q|77g!^)ms)Ny02vV8NQxrOxW!JxMS zNct`d(_0;J-E~&fz&OcjD@Rpkn%J@ZFPkjb6 z=%WCNh63-N^R_-gP{1jkqEIvxc#o~z@B$sWkV<_96$+qW=$n0LcKi#1%By&aLc!2C zGxu273PFYo*D&!ly{?CKB7?pN zpaD>z@^M+nWz?cU2Y8A?1E4@U8 zm`*^4Zdo8tQD`?5hfnR^`XV}XTbgMzs8|4GNR%V=ENGfRf2D0B)&Z3cR?@fgQ` z^&$rS5J3KLQ}xe1UyNF`VhvAG$RBR14IZyoqZSQLv1Cw*0Fr(K<=-{^ViBa2%2O1Q zegh@FnF@Y=!L30j8B{8OR>NRQVsN29f~p4d6opp9U@G|Qr_-oKs;*HCDic6^p*Y;5 zde}PzDIDi13hjmBaM>0v(Fp2Rmc^i-0>~UD$t_Yf(-8ElNs2<|FiGxw^Bi8mw_Wg) zLFEGIEexjW&bZ^v{%aTV6ouZxV5->R+AY+g)v@h88?U1Z0W=yqkiS$S4x$#Vi{mK@ zjfM`SfnMSl1npEF&Y(&Gv<059I&y905HzhPPf=(KJYl=W94$bQ-#$YIRSBSE7~`}z z?YaR$oA&b*g_2>66Z83BGX%ZP+r*%10i*%TXD@fP!@JKu0RscPMa`)`XE%7eSJ#Bc3LXV)_eM9=2jR>mRTF9W^0_YG-h2;j6mo-NypM%}s;EU4 zgZnY)j{tIj)7#+WC;U#B>FPX1AqP0Uqh$gIA;{Ez3WNR%AZe*Um;ASQrgy{?yYKDs+)_&wE3hj9ZAKo4O2p6dAl2|jSK>+E&Xu@T{ zJyrA@nNH;?3hBXU!uQiGIn<)zL4FLfUHRXemF~e=_xhrwJA!0`d5S{!;H~srPU7J4!yC2eLz5JROk!XHR>RB& zK_d%|7~~*;y1>bBb5(vIf_^tiQK$=?40FGlW*|sLvW-EG0_eANLgT=-0t9JIwi`LbIG3clO+5?aB zkBL+6BWT<&o}$nmc$DY3B;G?%kjWzkISZgu&kEU9382T&$se`pcshcf+~%qOqwBuox$fRSfG4HWpwiHu+A8fWBigA%OOr~wq+LX$ zP?S-MtP&DML}VqA28vJ-6_rF*GJY55@%SCr_qxyh_x*Z2J8$oC#&ynzLNB0`Kh00( z9D=&`ZsXAO99avXY^jtn3L3_MR7<5ShO@*PQj9&Oe1kLp3DGE)6p`}jh zCOn#OY~jWrI{|b8HvQ!-Da%2So&--(=mc!~Gx4^<)6w@gL^8-;01bz_`ag-6@jQdg zMxLV3aHy+)<@T{3x^!#fvKe$)0GTGjQ7zKzU!oSRjpr!}nI^)i)I$aoBS>{n4TG)- zAW!H8S$=(Y5~QPb$j%sHzR1X8BbBj8s2qM z)tUG+(YJ6igB%5rTQqcui;q_!=y$Ueh1{YoRC?bZB8ysdMn#)JP6Fs1Z2Ak`KMGF_ z>ZvUt+;nBpuk0Ti5 zDu9w7NGZ<_4+};u`tpgVD3tsF{zeDRpF%Aft)j^wHvv=wrE3AjMurHgZI+@?4V11K zISrLT&;t(x2Hg}uN>Eq-$?_(@)AHfu!m@}&U`%L(;vW76v}};M1S&r zNg`;aSv`Zi1<)So1%(B7>Wl6=MRT5_&>rXoY2FA|K+xrneVm&f*joZf9u|DZJ}KRc zpjDrEibC?R;CtfwD_l1+XW9$~-4;LP|vXL8!owd-7G~RjT$JWHhqqv+20Q{=#BsifamDap{aXNi?n|56omrdIl8(( zO%An4XOSa=dqdtb}V2tjea!3^>fK&Rn3ib{{gy`VdO zJVl|?@Em!?ck79u_%>+_@)tk@VAARIHtQCsMOWJL6om#fE@7W@>5L$cEu{>)D}Y8r z-`n$@)G-7d-^x=I8V!B#uCAMRB4~B2%+;pnC_n(o!=>BM!tXGGyyJL^Lh^9wE>0Y! zjv&85iVV6ZfJ$HtTifS1o>M(On5QUI0%O?asVn*;XzGbY3=u`0Yl4fR=w9D$UldtC^Q0wmj3E>ctWFKniYfY3!r5% zyER~xf+T_xrt=hqmci`S+xb~<(WOgwzRjQq0_Z5TsHo-DXasp&lyS&09imONI-^bENYQ?3QtkU0!l%yE{MUyf5$N=7!)pmZf{RHm!_xti2&*e#Sh_qTyVAjT6vzLP){g+Xwm&U?qK8{QDRWE04jvR)YyA= zZBUEykMb0S3SltiX!#Lujh~jej6pF1NE58n6zUBHDkJJV+at(XKZ!w41yD3pE4{D#`wc;7 z40wt{(NL}QsITu&1l7kEGAKa+MZ$uh@&LCa1Qk5xDGEixf}n}jGrTFRPyaU8o1UX* z0_YsShKd2_RkA6!y2OY9DG* z`)NEyAzj!MHpaAC9bG!9t0oLe7C`!$@B?id{P9#+WwR88^kJif-;$eA2$K2h%Agkl zs2oNFoiA_1z0EE)JVl{$7!hPv$KloCybX~IN)bTE;l$&F1UoBq>HciwDGD8j6OXr! zaFjZOrEK9T!Nrsraf;oJ+Q9@&~12*Ub%I~ljIf6QWUxk&(Y9>?aUCA@NqJO z-Uy(TP-|*8wEalbBEL^OMWK~YYnm(Xe~dAL zJ~Z$Yh4f$8OD8|gkM2_o24kU1a5}uofB%&r85Y5$)I-v zD5ymn<)OOP_&J*Okf$gV)WT9_<%S$>1WoBv%%BVb6bWyJJ3(oG(4`yLm!~Kc32%lm zJxlQZ?)a1KU7MbxOaWw)Dy7^Y?~Es4y-)EJg-lW{RQio}$v`a{^?3w?-V31L&^f7! zwVr`m)FYp#DD)dTCwhmJLJ{;)Lz6*S0;oH5x0YX@I1523nmk3J?$F)p{JfJ7g3jDA zU{JOIIsvnYX-h&6B50E@Pf_Ru%pw*p&B8s^Q|-?(C`SO*0rY6p*8&8s>A+JIssm{8 z##@f4MbXbr9`O(2vRrXDGC|E)}G2jS-cD#{vw7!9|e#tOtah^F-{pl zz9~FKAzPSc$s23?9kuB7%YK2fWWKpf68RND(TW3wPDvQMsD2BZIyO zpeQ(EV)}tKcwThWVVpdSKg2)r4(KDZi&F5RARo}$nY zcr#4PpI3*VO!=J*`YC{1V0_e7C%Y0s_j>UZg7H@Sr z9zc-l5uTz@U#PNp_HcIx1a*3Mn?b(>&?D%3AH4lwAc9(C@Dzm}LEqaZX_Y5}d?qC@ zs6+q_fWOfIi^L5GvYgCQ6dC}3BNf$y_+9tnY5{|O3m_xtZf#2GJ_$kBukjRxjG()v z{JR~lo>i)9>E84lm5Lx(d8|{L7=<9YYM%NJ2v!~gQ>9j+OBbR$kU?bv=som;tb4Tg zK#h&m z0Tcj*Uf2Du;Zsfr6|-m7u5Mq zKAx|;_3RphDh1FksF8a(f5~VB?N8(>3hja#xtQlYuAvq!8~%tvRRTx>md~EOpV|#U z(?;+Vg%n`43AL!Ww1Po@ z1<)|4zp76D7>J-}WjsZpVNidScl0%0Y|>pS=h5`Q)(D`^@P*Y^>!*dFjmvn7LY?6Y zn{<9Go`L*wPl-YQ1dtlIzK7*s2O8W%wVNR!Gwmw07?AIN@ejp zMWHO1&?x@ox(oe{dJS-6kfi`JhboI#{U+kG(94>oC}a**77~V|P9UhSX(WTr382F; zUl$@(mxcaDf10HzbQtFA_KX^X`wW{uW;4i202#t(!n$_!TLg{%#8VV9gwMq9MK|&6 z*3r2&3_34>N?>hb{yR5(>YAkrPf@4@)+Sc1?b8umI)~dmy_z1_3j*jb48;BpD8O&; zH9kB=p}#N?+jGJz0zsawrZC7_0PTih%dr8KU(uy|)S9O#v>S#kbGKFE;lJ_@Z3fv0 zpuw;!>PE|WJbByOkf$g#78AsM4D037~QAim+<+j^2+al<`O(26V37{G9nfNmJLn(sZ z+42;HX254+>&5 z8NkiZY5mLy1R1X7DGC|D%@DP_8duLsMz;5EdX6p&pvO=+(Ph-*H3<6BEJdNmP&hGr zQVQ;FN%b1Zpeq8X20jy8M@r(l`t)Wg3e~`8BKm4eyltfZs3wCP1W+4TKYEd~5#M#~ zj`I|S+Q9lzj>+jTbTg>GH(-#X0D22+<=qT6KSYM`J;PL4D2%2)qlR?e`C<{ue((m5KlTM@Tc#1+JVhaO_$jphv7PWV+7{hh23-?C$6zM^hqE(Y zhn>8drzms`X7azkN!)@iU35r2gIomA4Y<9#{fgO&T4ei>rzms-Ztr)K>+xQ^rQQ18 zYIaIWU@VQT>9?&I&y?gF!b0P~%=a%~yY|5oCXyrzq68 z7jJ}Nt81u^Q7&~ggIqVGI>tvS*|Qo-VyL37Q`SL->Eiig8ZmAYB!XpD5L_P2`d6&{$4Zie;VRt)kHK*!*_Ustz39bG!rTRcUf zWANP%8`TLH9-Bz{FzAi|`UOLH)2=h$A!vCUo}$n%7{UjQoX`%nXy%p#2Kfr06EJGK z-}H||@E917}*$@3; zl0Hq(k-q>kgl|gSVH5mj@EFch6f%TwinY|g0Mw%E=La(At^g{A@BXidB>a^;bAhKQ zR1Dw!A1lWUN6?3oISdLAKqsI@kwHrf(WMLd%~KRQ0WBIZSMLLYr044~=$-&N0L##l zn~wS-D63hDLI+?O+BT_|C4yvpPB17?0HwiaqPk!+o`5ZAmZDG^d?rk0bsK=7cxe{~ z1qmQqIPxg&Qg|F{kz0G7qL3{dc~sliC>%i++a59Kz5p5nt6Fhu0`a?U#de;e&=^?N zvbA1t4?$yMGZ^$h06l~AuXe60*GDb-+bl()XK?;ikoQBp(^hFvC4+(m&@dQ!tRJ`3 z1VQ}=^Av@K!O%lDe9K(aq6cT>?le8HAp)oxru|1{cy~mQy%|qYs2ZmI+iblt20^uX zN(_1^fKEV*K6}gHeW@+;d5S_OphbVZBXH@O-wZ7Vg$kh2Q2e0XVxbXgQQ=IUqR?n4 ze(0_w+Zwf~!DTOl!URwWR5%aUc#ro1^}NAT6e@uV=k;3$=pyKQtqp@737~zjs`WHK zMH98?dmT?vXdkRjq9om#&~|5rZBJplY}o>LbVDJ^0abJVl{u zxEWN;R;VNB$iX(gP0vxJ0BStSa%tYK0jNdmjd+SejYnB78RmHtL7~}085AXeVuIi- zi<00I2nxvIDGJ2|!NU5Z&bZdJOjV6RPXy337#~S$hv7cM3pJji&@>nym0RT~q83GX zY-Uij09pdyqrTZ8M^TH;c=8m5mcaMOz4t#{;T+h)ltD29sPR`_O7?pkLy(;WPf@7x zS6vj2RO_G?o!#ulpjZL)9;W@TJLkJ1XzdoBqR@Mo_Rlbh@j;OO<46X@380&Qpr<-_ zIPRP*jpQi`-TVW;Q!~!D3_<&P=P)Q<07b#u`-WcUL#Rc|`|uQnqTubFS)ZYVpkc>r z81z&CMZ<=&dAgl($xV5)6osN;L)qL7armxVmeJF%>48lUK+@1>h(D7x7`3QhCQngF zy0H=_BOCv))cdJZ81zg4HP$3#N~hrQQT{ZZqEKT^LX=m>7I$!*)}z|t+xuhHK?Xe+K(Am_ZW}cl*P2FE^Av?%!KnPr=NY|Gi$<)n zXHb#=@`0k(PG4PWP>cRFOHs%NidxS*yujZhyWl_uB@3XNFfZDvAOcsE9Sz|r3f+Wx zQK`2_ahdLb-me(+LIC}F46mRoJyxO?{cV<_(4WV!)wFcgEp+Jyo+xHeiU10SS5WKW zGK(e_+q^}W&csoZL9YdnHymu@uz#-;YSBC=o}!R9 z9BlHk+6s?h>#GbH^hN-!fj2|!Bpdwpepk&?6j}pshHW;J%~6YjR-I)~ssLICGmuAH z94bdGven@!3ax_~$g1ZRc(AVU zGALaD?SMYRveYa1o5s-iL;xfDAf2HH`|rX+)#_!SH8uJu|93Q|&7}-;JiY6B51Dc+rPXb6C>PDVyn2Tq(G{*B3 zh18*LM14*eUhh_PRAf-D08)Up)EB=V&Ot5e@5ECSQh>G8svz@)s73q#E@Dug05XOz z?1Bx0an1JP8lIw%F??a;i;tNh=)#tr4Eij9mcj!YK6^lW)FR)lJVl|U@W2`tJL9#9 z;CL$r^n*2%va)V7FhNizhTb4dE#Y z#lr)8vVVyLf_|S%V9*x<^dbfhi+(xVAGPSM6;Dyv>VD5G4JPrzjNL!BWMceh6NMelbJxUej~*O#spm#)HD=;spgR?{IJ zc#1-ouEH^i`@i6!<=#Dd4EipB8ux||(o@$#m(FT0Pf@6GZ@5-3Df}t_^5O)8eh8pa z7+OZ@x6MM(^Aw(*8%QTYlyKdVU7Y6+lK+bS*%@z9q0|f0+4u zAct*Q3@Q^qsZht*XJ=U=f}FSW6opcuj`3vN^!*5G_jE6V{s^D|m~_hRP>7$SjAkhc z1;C`!Mf)DGJSjy?6`#Tz4So zt40xnssxZZEchmO%#%bd`mM=R6f%be-=I;^DhRUiYZKJ;990V-O?Z3%^7rV4pqu_Y zMIlXidr$OfITk_Pq!k$SR{;6K+j~~YQoJSMd$SaUeBtfwI(B#IAQ# z`M&>-pcYLs+Sfq$@2DH2L{v2K^I2mM{ZZKKI->r3^|pMbP-4JVl{5FycF)F`@*uC}dF%gBk?T7idxO z4@taj#BMQ9QRoY_XyUL?4Ft9DuVIkMmjB+Y^cmJBqRzI*gXPc7QWW|OYZE_w$0?#Z zMuobb_nRJA(=DiuG3ismj24-BxXMC9O8)=<=l`r@d=I4pp(_3s=sN#tmZJZv_fRTO zaC~JHYW~A*Qy6qc(0l_}!4KW1Y=a=r?L0*x16aY|thhZ9K|>R@8Du7a_Q2%c&eqd$ zwbGwvDGKd@$-O?xfABJN`=Le*G8aIN`+)2+|7V)zXR{QA8utO!eNcXaTBKlM&!Dpc z=p>8?R>gnCPj_vz6opR0h@hj??^*<9e+p!fg#bz`mr|a)AQwM$_i}lPLTTkt7Ij?? z&mx|h{)$1C0w@{ox-SlGaP{n>89YUyWVq`Vp53U1TC~Brm_g?R&@AXbpPD^p0=nyz zuJROxW$#G1nRE$k7IDw}q)E7$E z@@pbH_;WwKH;N07@Ho}$o&5DOK9D$kt=+IHzIgKPv)5Nrzjr8^}R zK~{D=MWG&)sq$oytpK_RExOjm7*|BS=qau7gXFkNY$yBoLY`6ZsBkQYo> z`We-aLzhnfyCZ`f1yE|Kl(PIfm1pQ?nEZpMD3n@ip|Z5$1nvdp%?n|WlK{$qD)edn zOH>h*qQO%X%77~Lf4xtWfTt+r69&cMVM+MyJxsQY zL01LPODK1b&Xwqopt5Eu3cZAK_oyphl~IdocFBe`JxA9B(C;ERd?up)CW5kd^Av@C z7s2+Nuk-IC$RI_LK`sJl5_BMKUCoanDCi|mQD_o$AYY~>>mg{O;$j9}7eIyp?YwS@ zM{RA#@)U&(0UG;d94;c-Z?ls@Hw2Iiycss^J{yf%G{=^wDC7cfhL8LAY(p)w{bt1= zR{``7`gKE}^!kaQ1BE@4-z2^;$Xx)%L*M)Hwz;7Q zdfJkwC=?HU?>;(Ch3IeOv|aLH({tn@fObT{F(7s(VF;?*!BZ645dkGP+uNQ%(CDWF z8RRK|>|rWwu#Npr1hq`yDGJ%cRM_l$M)(k?EhCi~qV^$= z7~~^>e!&ct*+r)S1W7#PDGL3987k+CdKlWweQ%TQJVhZ{fF31$ z#ZzI=jVc-BD}b!vrD_n|5>Hop9pWhpS;0$HUd!$_x^#oz$%QsOuzmulJ>&RPt* zD}Xd%UbI6)XcdBDuJROxG+|zJpyg5AB~Gf}%b)-O)H+g1xx)QB{_q7g@DznwN5Zka z{@3hLi%i$sGU%QFa)70o={uj}Z6n%xJVhY~SeiMJxE#+j+z9t)P@n)Rg9+GsDOX;j zOJ^FvQxqzL30N(gBwUk_+AEnsK?0~NtS=l-FuH}HnBF`^p{}sLFv;nEmN-8iDPqul z0i*%z-4o|r#WTHeM|p}u8nE8IwD>8W7&Ls_Camc>dLV#SLI<*Q$uvAMIQ<<@QD`M} zAf+#y$0reWR8nA2umEZcZ|`#5V|a?@eX|sW+QQrW^#0Sk(Osu*tHz)Z0dxvR6M+l* z;O)S@F7gzGPQhqmtC_Ytg0>cJX3#?c6a&NmthNO)=w{F=;wcKn!0`XMe9tNb+08d) zP^bW!3X|mfgQg`Q==cJjqR>>BB!9Tf4A*RD-E(75m;g$HUeGP`$b1Ba2J#ez(x4ag zW^cA8YSBxXCVV3yVt>gG~j;p(J7!)pmX24o%#TUm; z2->rorzkW7)>0pw8i)5C&q=OfP=o;L2oJ1g*abW)pYwvJDAcjBK&4a1d+5@cO_YDs z^uRtAKuOR!IowSW9|N*{5>HVm2|6bSV&3C9)pJf$7!)ahJmJGPZ)vMCbm`VP^Av?V z;lo#GxY7`{NUL6(K~Vxo|1tD}mVL%G+tV9(ibDF2;m35eo_9u&&jup~JrO__Fq-JU zu$41v(b0`OMIj3qO)RXJ$DfJKkL?*0Er345yH2)+JpRI}NAeVfKEu1t)Mpp2vIra) z#Gn`fGzpfWM{2v_51+{(o}$nsScbMzRs4)D-Ha2j7!)gje!^T(r;cyiqDv=plBX#2 z6Xt>zC|O+!+&RmjL;=(l-rm9YG6o`OnJ-UKs4Kj^TWy`1grM#1ycqOc05zV# zuPb4Q4|A9;%~KTW1pT`7Z>R8j_rmxJYSC_eo}y3@6o<=Plf%n8 zFP`QyC|LlFgSnt|&7Tbjx|hIH6dDI}LH^xS4x>x=Xm|sIUI?I7Fx9E9TDb{9r$+D; zg;v2-r|aq6IS5L&=o``Wz@`YG#$U>=j?$ZipnH})MWM!D%AUXP>O%xAE1bcgmjY-C zY)LRket~cA{zW`Rp((H>Va(bi+6X$Px`sio1kfS48NNPSi{IWFYCJ`uLvS;6^p3Pa z&02lpf>{OHB8>Fuuj3heibUo zQxtj)leazVv+x?Z(4N)teu zaG-aJ=Kt*99}~$_6w-tPy)VvG`GX++ZeCRIW`V8Nr?8JGv zbnT|0?Bk~A=&bj^&>bmiVS)ufC8YqHTqa4 z-d&jdf~P1H0Nt%ct>&*nQ0B^p-G+rzq43O4qi3Eb>5*--VqF z$`nA2$6qI?e;R}?-EnK4qEO@U*Ms9SHY2F^rxk2jBaBQWagg0plkuu2bK(^%+BC9 z!)gznqEH`LGKd+xt{OpJEx#}*M*tmbENXolj6dbsk~~GBV=yr|y`|bk1exhdMm9Z1 z9|TaB)@_uxUHf2(F5QdOJVl`{tu0lw|5V^|_jO?d8T3&AWx$bE1-<{tB52Mdo}y3& z9BEZyJ~JM*$Xs5TL7xPWG(51S2L}`&=wdIPqL4H^u%(y1*C8nEs2+oI1<(xWZap3; zk9(?Dj`0+QW^!UKEh?b`o2*L1oI zgFXwO#*;k?*X!YZUcZ~ADAahehvQV$WvE4~uRUT=z5x0G{kp_XNAZD#iY`1wp&!t% z>$pqj9jarT{5OL^1-OoJmO}ZgUpjk6qJPz|mYM(mKmTVPqk0e=17ccx8{IkAC3%Ye ztJH&_rnh1TE?J*!P|2V#g60o|-sX~-k9s3$@HU>J&_L*It_iclvmZl~<)WIN?ymyK zGFD32arovV2x{|!rzm6@YoVh0JyaiEx)zEP81zj5NkFNsq_szb1 z=P3$Fz(U6Uo|^}wOP9OUmOhRj8T3;C>A*T&mx;a4pubUrEKgBL2iED@ z&l!FkK@P@63@R2t#xT>nYuaHv&EkBRrzm6$Grdn1%dJGvf_H76G(AVZ1ke}wcvW5w zFG4L+$>1ppeSwcxuZ{QdLdMGp3JfX{K*wQ3(D&M{T?opa$Ws(L4kH5h8kzg3MP--O z81!2J9fq6XT4JYU1eIOkDGD8io58qL4<9}wSGAczr2f6ULUQn~ z)3UI@m6`umnKGzM00qNc=U1tN3sgSp@DzoD;jUZ!#(y8WbW_9J81zQ~#lgo*+G#DW zHI;nCQxuAWkJsmccDRmlRF5bIl?$MbFyQSTw)ivp8%g)%DGGIj0q-M;KfO?kE+5Ka zP=x?$T&FABQ8)lW_llX7$(VY znU73BP}WTO=%xp@N&szv&6QiNhKHdR<<8I zGWZMI(w(O$^a<)lv_i7&QH$j2wHfqR04c!1Ch|7w-B62uH%n1S0S-24pL8Z2L8|MG z7*r#IV7=Q#XMqNS2I}$De?YL_T`PYLA2IPK+@3-I1khA?*JT++;^!zaf~P1n72b6l z&Knh@7Om(R#GqOMG#oZrq~sW6qZSR5=P3#ehfNk9b2n8YsLP>O45|}A17N0ig+tj? z1m!hLQD^|n^q$y0%m6_FImHaB7eH6S;dg4H)m|a!!UvwB(3Nob-Lz5KcnwH*PKTJL z=cqvd89~2JZ@)gCMI5QjQxr0Se%yg!`0j9L`n!BZ3p z0BB?Ga9ecgJf!9`$W#D1Kt-9(TWMT>b-N8uQOE%*%IeI%;u)%Y`r83CB85bhR`ZiBdXcnxcT6EFGR% zXEMl20Ih(@y_OZ<{~_pmvlN9^z~r92N*X>+XWG6s3_34>dOnm=E~@JP13}~V^Av@8 zK7>QpT{I%mrMr@Tgh3YskSgqb=owXk%Q;-$@)U(sVednM@4vAKIx@wHLDmB3CcNu9 z%6`K0Ic`&Vib6NxU8jEfg9d{1u7ohiMgaAMin5{IZcCs`H_3sgDAW@w${zn%l8T^P zWoZnu6+j1}19|-CsMZKN^@pb@bPzg_O4HBJLr}ufG6r1~Kn>6$@1uM55#+v%rzq3_ zEs8Rdz}>C#K-sva=jf6E+5r==tLH|3LeSSBo}$nWn1Fq7F$XW7eeFJ$L3RRY0}TJy zckspAMqc&cDGF_X;s4Q%$MNZfTTU%zki7tU3D42m9s5V37A-l=Qxtj$&r$BSq8ju! zQpwuMpvwZt>IWQHr==>5AmwbHqL9@OI8E1m8a`BT+tl+6x*~w4!(BH|?k+wCWY#pE zqR@1>>yFRPI)++w`jQWW90btZe^SchWLD*(7A>&jDGJT~2M_G-akv9{wO zOu%-!yAHn@On>tfh4#S&?5g9cLr{yZEdRnFCjq1jWl%(X`GpY z`*pL!CF7f(BWD3rA0nl^COI8L{UUgZLiHiAW4&gv3u;k>+#m*B6+qHZ+q3cUPQ1_S zY&V{wkTlfxl-BKPg`ibN$_%y)pi;9ihX%_9c63ZNnIDc`Cw#~MM_ z|9FZ*L*P>$ddF@r`Wvm(%Vdz70J3=oGbhE{t|Ms52A-mj%`+%GE_2+1pb-(347w?R zg5ksW=e+V;1XVOkQ79Nbd=(lWdLpQuT(_r953IWYQipfl#*5t)5tQ33MIm)~*U3(J zgdf<90}~kJA%IT9u;uu~NPN0`>_MKQ&}kU9NVIlOK`mPNR*ONN0%$sPw>li3jVGPt z-tiQLrbBnja*-si2i8{F#~?2Ov<8O%bAmJWp-VSm0#8wB4GjN{mgp`+ElPH>WstW3 z8WIDCa~3?&M$kQHo}$o@82EvSkO+L~wpRHw=#~J=hY{cHg5xy^Qm*DH3gyFy@1MgJ zCDfvn70C>`Er2v&S!dbM*bD@DtmG*QX~43MeELJ&3(5^HVvvsj8VUWn>wbgq9wPq` zo}$o5=-16&(Q+ed(crFa6Pli*I|8U3bWVC%w!uBsu5vs@p?1(YsSjV!8bMVD6&U0z zfEs_o>4VFUT6F1N8u1i`8h^rRa))tvYtQZ1Y7FudK-zG7mnu2o!Sb;;JVhaGxV>d{ z)BI73e3Z5@$X@_i!_czD=BcFLe?;}{QWusZ~E(X*_1(d1<*iPG6;B4hznFC zuJ9Cv2EvlT#Is7cfPcYXHwFa=pj`OyrKEV^Io0kpJVl{g`0!cPmM%t@u7`FMgYF5S zbm)7x*3dCUH^a|nDGH@S-@99l8Xob<1>`U&Py|&$SyaTEuL%0sEcG8ym4%AA&)CbT zMY>)8F(^m?HSPeto3atFENjT|6ondhfc|mmiMv~F2jrhMJ+Su$(E4oHNgbCp2DNC{ zL7t+}`fPZPG(K%bE&B0d3WFX9pg(0&%3(9_-A7Pj3Qtk!PZ<wbOdQ^>5$m;9EA%Yci2#tZd_7{ppjd7ibC$Nq0Cct z51y|pele0k5dz2<_7L^zyh#Z`sVO`~A!FD>q&-{WIcia)!h8lj7C`nDFoxCnh4-}H z8pcx;vaf&wem~;{2&y*Q#-K<6vjkqF#<>hR{wwQ`-kW26r^~H zLMpKOf8qPyM+kB=Y+z8V0BV31Ra`8=bwT@g@)U&{phXH>hr1)lKfdqtrUy1o0HwqC zNb0vCeg&C6;B;ggAxSLZg`G5?mB^csuj&r z6xt2X(PgiGeyEPIV7e27o^3_vF|Hceb+%!D_gM6=YOT)y=l`r@93Kondto{N58=a_ zrRcwEd@$6Do(ix)(DR!i>|d2AX#SF7DdmT@-fPfx&T!``3N0yypQo37^b|q)Ez%kE zTmb1pPiBJU`Ys51EWuM0(uJN(!13Ht1Zi(AV^ER+(uTQ$rDl$&5j0hwrzoTia|NT9 zCA~q=#01%-riU(B04;b4htHTt1t3V~8BbAY!AlF3YlnB>H*nv1Hdv_lOy%s>nq5qt8!?6stXqyL5QRq1IpPhaUl1I?>+5`r@5kQCG%uLOk zSMw2MQpZyiIt*uK9^BAL9YOuqeqm6m0BZb!w#lvM;>o?@W+@6a{y>{+qGB$B28Bu{ zH$6vb0%$&bQ>yh+qfv`$nx!Z-AHFFkFRk=M(4vlm7?du6*0pM*oY5;f7D2r_@f3yD zwX#&%f68|`g1YZeX3$#!WDPIXke_q$_MEb2DGFJ`OI3B`GyEJKdZx#qcLK-+ru{oq zPR4so<|XnJg-l@Dzrko(Z`2~UF((<6A%G&GCo}c@Q)6`L929wqLXpsuxz>8a3>>0Fr}A^1Fj>;fk`U-+78c zaxh80zi`wn1jR4RWKfm>`U9VdpTk<4p-bn#h^Hv@2R;*~ZsF4qbo@>wgR%vX66{Ou zt=hR1L7KiiMIj~FmwLi09(MvY+ID-<^uXo_plNW|4N{lAg&_HMJVl{taMw-ha|E9w zSFb;TK_3LrIw&s;o^2u55##-_24K zQin3#>8CE?mB)K^ZBm+^qXGf6A3l7#*DvFt$F+K%qR@W$@U{Hn@&mQVe2W5uz6hZ6 zk#KHL_dda>MXR>*6ot-5LNQ+ak0l7wf1<{quL39t=B`z4=sib}dNfZ_CwlL_M06GgZy)q%6tPpgeKTlEUY~xIC_|aXcMQ=`-GN@1hnZfW-?M4YM6{tDQ zQxr0T;bE`fmfsLG=c5~giUd%5Sjce7Ka6jNZl8FHLhWH8eX4ugIOpg?$FS6ncZM1P|hH+hOef$+eF9ld~W zhO(L(2K^L3I?&yE{HmoPg3|u+6oqu4yH&YxGTttlxmx~Z(*s*9fCj^dZ=CbQ7YKU3 zhNmbr7(RUZH`PX9x_hY~ zUZ-38-kw3F0%#k2%Fo=KiHDX;vv`U^+ZsRRtx|AL_36wY29*gQ6{x@ZVNfoOTI4Z{ zrzoTX^;a%$vhJXpVfyt~4EiI0O5j~*tkfj}K~rz=6opFQUH5)a&u0i~DfNp%I2 z1^7%vxfr@4s9z;dQAhzk6W6^G5)rg)g%^Wr1keF^Gq_hs-9XT=l{`hE1Mp_}9X=Cx zPFe@XGU%TGQi5r;;}7%iBPg|5ib6^-jn>As_7G~3pG+QuY6VaOlo!_RdWmkd{of;$rse3(|dS|LeF4)RDE!65o%Gv>%Om>9@u&T zG!UMngbH=s-Ku@VQxqBq&(VevQuy$hpOa@Ys6hbjih>%sSO+|8X*rdrD6}gImS#4} z3`H%{xVDBtCi?%qS!o+=`r9h&+8SNDQ7$}1p>43~Z^#e{+zV>k;wXbm1yJLv*2KE5 zxJz8wEJdNlRjucL7UIrHhs{n5IwODz;j}34rQ`5AT|u)Hg$m)csBd%cUPYH~M^p%d z%mk1$JV(~c*WjtnX-{~HLelUY_4Bdpf}m-A(ivnffbPRus>6Kqv-oe+m!~LnAJ$TT zC+xsu*nHD62Avf^$KY=iQ!@om3}&9;DGD8fztM_*dTFRduM1?~G(ATa0>~cv3e`b-_^|QQH$E|;VB9&hj-md`!RUi$dF_o23-(9y`ezGS*9=U zZgqdbQxxhA1uAN;+c%>YNsf8OAZr1X2)}VNI5~S9YS9u!o}y49{Kn1ALF0BJXou|= z2H6Op?y#`l=~ksPf;2Dk6otCO!g_Sajo}D-^<6Tx={d3$K#fZV`9>lA5ET7`rzq68 zWN>fWYTTkLT7wvLQ2@!nF5Rb{7RMvVd^t~1NCtN4R;^L%j#_jxK$$_81kfp1s5)>? z3$MeTxW`izIt2?=o5s)4MUYkd4GgjqKt6CJ?0miJ_-|y{fu|_s14qJc*FM%4L0Q{Q zGRR&4xj;SeTe-Ei=+fnF=P3%gKs|6`jxqj}k4ki5&}9J>4+jM}_DaIXQ;yA5j5 zz8-fObX@>7b|B}TJk|lVXhct*qEKT8a`n=~c;E4yL&*%fA%KeDrFxZuf+?=}pg(n*bUE z^L6WNhgu_O)J6n9P@w7SV4F9D!r z#`r5Hhe5XlkYN?Pz4uz;!=e*nd5S`YRj_J#IB^cT>vs10$DrE+C=Pl-Qx*)xD`Io{ z^Av^RpciC+_c^Yw8+Kg&ZPNqmBY;ZbU8j0=9bRo|cY>!VR0{99-|;5qs6{O^r!eS_ z04j$TITlkxh(!kAYK3J^dw)i9;3R+@=g)H9Z+C{$AouaO)5tkKOdviC0r z-4j4TP=w$AmL#6LE@_scP!JU1$0+6Fp-1_l4)2;Cxv#km9m`V`dIZ%<#`#O|uU<~G*v6m-0%$3GCB36k{~<`ylBX!N6uy#k&W@`_H$%_D zvkVFrKu$1MuyD8EH3VrD@f3xeV6Na{Xi`0b)HS>q6e55|!wl7I@Ba8J*+-M7C^Q;o zsB#wF!Q@VMF^lY=s&;h^RYLA_FM84h0>t^92Gy@3$-XJ?+Ak)3m^|D zez>cZ5{n?)&pbsT4=8?^KjSuDr>jtQVo;<2nhD>d2fgkcLs0%)o}$o9_#VwWzVRb! z(FD&B21N;=Jupe$UFl$V1T{2EQD_fLl5ev4jXQzilIaY3B7pp1OTzj>rF97MZ^cs- z@`o)65vv@|q82^aSjM1e0rU_SGCm9|#<#b{CZ3|uLs-bD@G{3Gpx%+PnN818i~t%9 zb8hLsCmcjAvW(&>3XO(2H$&&Q4^fMJ29IS>tN=O}0W0|CUGVVmB}3bk!%sWP(jT|5=0^>HVI;ssDwSP)D<7?_M8 zrB6IXp{}qXxN3cTHfmAuobwENDuC3}p-UVui)W~um3fLn>gg6LGbjG&jGzs!J`73_ zK#O2xZGON57e6Su@f3v?!N|I0#N`zTy4>m+gPsYXKCo!}v9+N(x^xFx^Av^pz@qJY zo3#^%w_yJ@?CZXuUD_S9_f5BVqW<(=Z6&*OR~&1!d*4|7eR~fXjorO>tNt#fx3@*iv9!Fdybj=XxwcnbS2g(c(8xqbHSB} zx&y0W*5}<3G=3OQQ7GyTJbxSZ_C}DUWi*451ds{zo~qyZ4MotCb38>M6X-qZ?uf;! z%F0C_8I&x5LZO^$TuRgk1kL`=Qxpn?aw^+|SshS|&M&HC&S@C<&Iv*2(jf;KGX zDGDV#vrtjAUM7j4q`SSdnr@;L0i*^kQoOvo2tkekJVhZjXpw%0m%9;^AvK*rF9nb@ zjH@b(1CAmnxD8KH$Qi~}%a$i@M9}UXx(s?HfVQN-f%+wb@uiz@$Ws*Bl47AUWpyxK z`R^HH%%Il-D5DMz37ESE*D{neOHnAJ&O+tTq>PoQMOS(|FzAf{GJ-kpLAn>dqDyDl zho>lH1ascX$D;91n%JCqz@StC6cQ$-T;@WE^~`rSPQNiOFp3O$DpX2PFccu3me{tgCZ z3ZThwGxX{H!3IHP%~BMa3^zm3h#Or|i>`J($DsEDs2zMJH2&l`A?Sb{Pf@5Hd?qHA zjmO=zQsY|;$`U~9Vfszx*Sfz53O&qI6j~3{Z++Hx!RzE1>G2H87C@%Zm9BT5lYyX! zw>(85Q|L<9w(5>c4CJTgGbl#@J%kqBYB0yu>-y7pib4;eMPU~9_*1^kNg}7|Ir<=g zbYO?U;nhQMEyDt5o}!Qr>@e72`5F%nV(a@e=%WDgegvNh_wY(|>7F$36otGW!4F^m zJU;~8bw*oeGw725dJ1c>wnK;Em6l~&d5S_$VGZ`pDF-J6{duyELAe5`FVv`31_a`9 z)$?ecqEKI`QT55|ji<5W`WWLds6S6pXbwypoUrQej9T>G>?(sk z3!ooxESySB0$wM7Y|c{@`T@tnedx933xaC%Lm8AWfO28&XsWY2E|XXJ!c!E=g|(xx zlb41dsPmk63@Q*nmT-KCSM;7Nbk_}1<|ztU!to&^hRnT$AanO}27M7g6;S_})$Pa% z1nu|WDGF6U{iE&TokEk@e#l z^i2R6K~J?^hbX-1_KY4+QOF2-s;i_X;w`FsB9}6#PyjuL?@{ga;1B51&5z>rGeHrMXSQxrN3mrmMXzbAr5o1A7)nE=X%ojNP>&Lp6}Q5RF5qEJ5U z)Y)@>04~9{%e%p#KLW@VmKsd9l?Nec+h?AlkSi=T{2l&19<^xLw8so87eFsz4C|ge z7ViT2+bl()moSF4*7={`|6F-Bi$N6vXb-#@o|ksEL@gS7ji)HI2i^>MYSd6Ahqa845}7D`l0Zy>+Fo*b!ss@MIrrA3zZw*eHS7qcGyY={S`oJA+YWw za|PFJxewsF z81zp7nL|aV(xRao(ckE2vlNBQp`!CqpYWLoa#Fj;pjrVG1+O66+`1kJI;_r96pDgZ zP)d55D}qA3QW#VxfPTVV*HTq+FM`f{^Av@C!d>T7)oCk&tXltMP`v>11!#tm8J;=W zF2z$6@&za=IdMFK#u`fJHa$lT0%&wREPaPYsiGFO-^o)H8XXTmPVR9!3_-@vhcn2; z;J@d+%i(YIepB!T1oNdH1hd4ho1E~lWnGaOgUkewY?+jD&Z|ip z=+bTf&QlbUEwfMwRFlUOgVM`A7-TMhIzND;`<}kVJ4ZLJ;3*1qegNO2&--xI!0qs8 z2Avf^XW(6TRq<6H)S?R!JVl{1@UGi&;X)kx8+9J_i9r?u=(#PND_lG<8bOkyd5S{M zZ7o!S=1jm<17qy#7-T7cM#AQd!WBAm5w!9$Pf=(jY|b$Cw!oDgY4yGHnjYA50%))Y zl=+nW#QnP12A-nOU=R3Zv%+Xx_&?fUI)khPPzU(%8I9X0g<3Rd8&6TF1AO?VI<3ZA zy|?G+GU&VjayOPzHmJW_h9JuiJVhaQV+)l5Zz6u8o8iU0!wkA0fX@5Er7OOqg`hAE zo}$osUsw^FaeWMe6nq^RWG#S7Vf{AbKQiq2gV>|N{{a5XWTXf-Zd2`hKDMqR6Uu7$3 z{%Sb&@O^i)Y}9;(Lp(*H)o|)ztMT3OL3i!n{btZb0dxu$o9r^TMIk7oS&BlZV6o}p z_ED2ii$+b9`P}r-T@paG(35$saDE$t22A293e`eS#=COJ4Fv7KGKN8R0%#|E_d9*; z;fWw^2cDwPPWbMhKd5sULCU2I8DuYjte`|U{jf1E5bILLQxvj-65XQXYkDB4=VC(! zT^2z5U}+}q+ynfxoY~D%6xs(%GbtrAtr2v``y7L=2%xcl;n!gH{l!03XLgIHC^Ysj z3_W(YSc9PO7PlGXAb>P}Nh!DWdpjHbjVvX2ib5K{EL7Y&_rPnZrt6+E$WZ{T2k6l7 zjHL)Vyq>2hv>u>yp}%nH=)kyq2000!vIr^Vi>+?qgQDf)d5S`15%4R23JdYaYsH9` z`AyG}vjCC@Xw%BcZBdIxkK`!|$piFfkK8BpH@aXofI(LUP(18=INn7n7eN;1d5S{u zu@DgWW zn`;cZA%Hf(##x&dQ-31JvMofTOP#_FF+^%iI)0G31Dj0NA0F}R$QdX`G z#eXB22|Pui^0)AN0;lruq=cc0r;-P(ph-U7%5M%Gbn zzFtFrquORE3faKOdf*;CJeZnk;K!g_0>}ZX1%tiPOb|498&6Tl0jdRaoF3x&oE=He z8FX6!^@ejC4f_Aj#@SiPJVl}2aE{}nikQ>r()}G&$RHm9R11BE{k8V^(!CnYQxvL& zK0}Y&7Wk+nDGRADP0!IC0dyGZh!-v2<$*5UyJjg09fmsM5tVwkP>T$|4q=e50O|)b zy-7!2cF#@Dzo5z!$dY)6@;9 zMXBza800U28f$yro}BA|pgSHsMWM#pp16{AM-gEv&OpxLoJMWK1{X1Eb|GaEr~`eidHPyii*<98;D)Rlg084cX3zrxlmjIUSqF6KMhtk#pfCZn0iL7SxK((Ds_Q_WqR<9-jt(ibeT<-u=YKNjkpMag zQ(;Ty+rCAYZtMk~qR>g03S0So#y14XeU<*!^c;l?APIPT_j}s|7ffn><0%SBz}s7{ z$Jf^g+A((ogCYb_91ND1+|WOPTBM=EQxuAW!E#`r44%8LzOKQb#{y_FwCG>y)8Pn8 zy}?rynhY)K@}dp?l>e&GXHcX7dJoT0FaJb5;tQ|jDGI%Z=V;X80{n#yTVc+iC;^lU zM^i}n%DhA^@>t1J6v~C8DJpMBokM@4)4?7LdLn=ZKoL>@=q{5HGo8fqRID6|-k}co$Bxb`@f3wRLBGz+IHw9hrFPR96fc10K~J?rHzOHcy4UtRMWK1nQ|;3t zG#NoR%XAs^Q~<62Kf3NaF6Zz41GtJ1NkX=4A(RmzxkI*$?2uJLw(MDnNPHB zsK_WqMMF_iib|xSQfc?Q?#|=!J6+%FbN=^wJe_y%amIDd4TI&vR!eceZvS7NqL48R zmJj%>Z;GI7BTELo7C`>c$ObA3FI7P8S0Z6d35tpf>_&8x)5hbnd5y zAgf0_MWJm_9Ih90{t$vzDF!kqN&rPv$?I+Ty9aMx-`IwyC=^izy`Y3O=MgkvOCp1! z1(39Kt%XhLeAFTZ3!b8ov~=y;d>dSuxjnjwK`{cT5Ju%1piba)FK0g+YCw&K=V`J)IRuaMtog|8hk`8+G=Rbpkx7L2xC~i z{A>&@oX1lXGK4W~kVenP2s-lc0E1ElP$~3++%2{7j+?asJVl{W=mllEjZs68R=dj# z`XqqNVM62H8ZErfOG$;NC}a*38cBxccr+2eGnhfC0>};C-s^`v8G&wwOS^cALT>Q( zJ~(rRA8JuXR5F7;3!t2q&Gn{ioP;-K-izib3gxssDXHzz=^=tN`j#;$O#o%VIW)>& z-{PgW&i#0bLRoMQO;-790|Y6ZR>^O8VABPVLn!RU8@(03f(-0fhX2L0ZsG^F_!m!6Xb%kk!;LT8L@jDD(~vL~g zhAle}HVZ*e?|T*u$`n9D;5lmYediK%*A2bTQxqBk&ynKY*7!t_LCqW(^i=>w)xf0F zpk^)zY9-H86pE^`ktEMiDMT%5YwE+GZvto}tcbOHsffR@$&FGJ8VM_6)09?xL(sAC zXa;2opaeMZMqzY*SJa{fuXu_=32@*|reE|Y1dUVs$)IcjlnbRG-7ZApAK_`$i>D}* z3#B06EGI5TP|JPt1r5(pjsVK3fK^N5m8VdPRQK}~g>ouj<|Jwut~uG0+?PS$1<-mJ zEax5@(-T4NDLh4?^)Ohj-7Kj=E$TW+pFz0-Xwq%?8%?imhM?{eo}$pC+faY?+4&%X zHeE4c&<_FB3>Fh7?Jru1pqW>BibBm`F|l;Mr#^zLYL77JrvTas5A5OYr78$AtK%sO z?Su!`VZ0XJzhAN3jX`+=ND(^uK6!)jU^#pRPf^NADGG%{t?AsHms$vF+cAwn`2xrX-rk3LrQs=>szxaa`M}$|%Dog<(X`%I z$)ExO6bnn(QAf4oQHydKr6?2&OW65p-NR6e>_4g&HaxI}0!RX{pqw>9c(r9~GEY%R z0E}0&c$=rzZ#_|^f?R;f6~*! z*ZKE5KL(Wupb!}FwqAU69JJ~s>)^iKf!K#knT;ms1!-{_17Pf^GRYUEmkRlG(mN^H7;K@|c>dQQTY>CeLv zBrDHT6q25kP`$h7Tm%g>v0_lA05XJ|!S!4BZU}0!hNmcG2scAj+*aJdcpB!xpeg}$ z9+o(#J^k<#L6^dLibCgMiF5s0Z6DO4F+IWt(&))k6gmvEh~|z9 zXCUa&(M$%_2%t4Eg!hWt7Kxzq$9RfDYhVcfyxXU}2r~Ur%b;2TG#-vGm@sH`Zv>sm z;3*1?hvN&b?c9$$81JTb`_u5i)(N0gKbYy=RO5@FPy?Q#P^uriy^jvUZ-x`MbQn}G zfSSTw!HyAwPaw$oHcwHgDa;kvbljMPF5S6iOBr<3^uKpDNvkZz^t|;8LEGecibB#V zi!kLS*AcXMy(NQ=2_Wh0N2^5>7a~Y+15Z&%I{Wdy*RE&;J$~!VpyL8a0z-J8d{rL= zS-s;a3Q1rHFTY{xNd)Eh3S`g;0c0oFTyNsFLq!M*?afmZvXeV0d1(0$AL68bFp)u4 z0;oSs?rl-F!3C3T4)GL)`orX&`iH=B)S_FNMGUeQKv$u+xo3C4J=CJeuRKMetI*rr zk~N_WK|KwWiW;6H8v*1VEU$M>X3BR2b(zLf6mk!?k+j+?(*Z$GeKi<#QUKXO@x#Mj zrnvun_8w1B$PS7h>|BpHBj`|@Sq!ojK+@{j8-f0K1%Iv*Pf;%vZn3UKuv3(7KW)0ve3eA8?iLkqUuAocjU?0gKdjTYUGgQ>3%|_4~ z2cDvk^v$5$#o`fy9_D2;$Uy*Ez@)_5YpePo==3k1qL2klN-W)?u>e6)lVyt=o})7Y zXwV-xDS7mByty)c3QtjJ&>uJ%FX;~6E;{^<8iO1K&>8rYU)Xa)5w*zfE>BVD41CHf zw3RoY7L95>kwIq#(D_JsGaT=W=X3fh@)U*6N5WZ~axOg)G=7IMgPa7A61*AYJd7Wr z7WLc7QxsByH$%*?@pu_}`G*4xIwycyz-)Zry7T`KH25P=QK$vX#;;y;62DZpX!b&cip37z47Q0=+c=v1T*M@0Fs3cq}r7?xTtm98J?n$EOa1m z#kR%I(e}b*2005L>CZQPG1|TfUApDJd5S{PpKq%2Qdoytls~JCK`sJFdbD}>6G1Wv zdOMq^C?q}F{PJ~AO$4R+sgyK4uonf;WvH@v)Fz`2K_L%#ib9v6%ED~In6oq!f;WG;AIZsfFlm|L6=!yWE4j;ZdO-y_W6szxaa^?~yk2S3`4yIbi^`!VRc0D1$*X1hP| z+k>FH&3KAJZ{XPMCf#n}J*^Mc=ribs0BQ@T!cT2$jW?7zuH`8TwS`mR+v|RLgDzc4 zkO_l41W-#@v>mQej(6#L2J;k!TEe33NXZCA1a($E%AlJ9=nj;C7XOv#pi5WWC`F+= zPy)K?Y$rTy3E%FAGAQHxe3g)+!X0F8pX?qNm< zE&iZ6h9ozvDHas53l8AurfA5^v@afS`*%RZANl*gFE~BUCuQ)cYBNT4a{TQxy96 z8;%=2_GuJ?9!o|u=&k@Vf~n3{mtWv18Y_LCqL2|xbvAeDinom{yt$A;-U6r^?mERu zr*PkUjwer1s2c9Nms79f%FNfYI~n97fGnVMvSENLK3e&?98Xcm0y-z(mN(mhE?vYb zCkFWnpfzwaY}<7dk74as^Av^Fz|F8^t~#DZ^L*mRpnC!+70Nj@PHl5XH$%izo}y4H zlyewI>!qL;IdqL<(0u{a3O4=Sik**-K3LR^rzq44HvK*Ee)ShY<@*X4i8DGx)*3+z(-q1Zo}&i>=mjj=-cr1@9zi?4@Dzn!z@qJw`);NP zvX-bb$X@{6fx&WXS@UNI+NRG_6uJY0<-Kb<)gkE0bpr-H6hKGd!`H!Fff2l!hzj6-66B_zod5S{*aC@(PHt9FIbOw{Ul{Y-FAp)o^Obi}X z^uRCG@e-b*P+OQ7Y`UXjIf8m!(P7XN0W=B@bFj%*KaXyP7FT(SLZhUIIrQ6{k07&} zB@B8hfI7oV_31?89|+Q^1!Gc3Ku|0Fj(HEd$ToyX1nneg_2;fZ1Ft63_&AnXENxO0P;?k*PHCu zV=#iI*YOmEywhP1k#3J%1eGi|XHbLy>JK&BiVH_yLr}~Lo}y5HsM$^pyHksx{HL}I ziWERoV7^XX$+SO$l0$ilLQ`PAZlAYm7J|m~yv?B30%!xQ{yUya!n-+4)p&|R8({VS zW>wu61ZD1wWY8M{q|>Uo-p)}IZBdK-ckvX3bXuL14BdSaFZkxfW-};C09}P)i{~km zcL?%}<0%SVg<;F5-H#MdizW?_t!Q|Tq6Lr~jLIDcnV2I;Wgt&cNDfBjCi?X)5fo#i z#-JDhln)Q=tp(8<2ns#PQxwXF2evr1b9V$4<>@l$tpF;4NvFLNYj8b7$}gUxPzg*r zspW;@-GyG$j2ZM!1i^DuebxICYLVr1p85|6o+H&udY-67S$7XGC{_UZz!%oL^(Q=y zHs6P*DC7fQ*uR$VE+S}9+sh1!6F?zP<@M&TXpYxmdn)r3g+iV}DfK&_i3sYmEto;^ z0!Vt!_Mq{>uISQ@-_BDMlAg1jq4l8xL8-~f3`!6{4?|!@%zi)z1O=t=6onp!*hmaU zb;4!3eMXfr=)C|s0;`tGKh@#WSCmKd6orn!s^!q3OYuV0-*YOJ4G(Oh04jue?>Wz{ z@D9*c7kG+7g)r|O_Q+@ks$+C7(PU7PDLRj_-Rj~A@v75twUTQ!|HuDX$2jbvyk4SH zQ4qS$Mm0P||5w8v!XA^$dDGAz^w(-b_P_cdXnrNsF@}tb!abR26P}__CDbu49B~$} zr5Z$7FzBNIY6Ha&dUikX3{|^Go}y42D1Nxq<`M3ww^wsuP_h6rEr3&QBsp^E(iJvJ zQOL9ah92dzxPy^*$cI5G0w^4A(Lnvyc(ebP!#qWyaJWU6m0ICZTfg*Z27MAh6)5PI^BX*vG8w~D6#4qCuPUCR zkaYK1__ft|`OI{QK7-N)kQEF)dIrD#fS?;od5S_-F!bnG@+}FqsP?f5gT4r$J8(0+ z-m?OC0<(g6ib8kbX2@Lmb|Zq`bvVkP3;`quE!whrvLR~Gi;g@+AvtJ~<-~JJ2%5Fe zjX{|L$Q$;h{#*ALPqXyh&r=lghJC5;I*i8E{!f!b8T3^EeS-&fQJnul6jubUz!VD0C5KdiCy3!)3ZE_LU6E5HuE9@uOFbOKgPa#LTGA}Fm<8-?8ZgK}f$2@t8 zLZNV+j%oJzYv|HVlH19kTmd9q$Ou1pUl&1bn(`Edqzf6>uh-*uo#PrO2K^8~lc4{+ z#rYSWp*pvgrzkWD`pr1wEy~VRsA+hP@&!;LyzA;F8a_c#pKm-xp+tDsRrN_2iJ(;z)frSEfL!5k zv}NGiod}ww$5Rw?g};&hLZb%=(snmsP@w>FgaVb!%dUqI)b%<~QOFSrRDR#MjjPZz zYS%O9w*YzpAHLOB2I2z#&viUSp%?JsQ$N+z7PUxzwH1T@2q5Xtnf<6;hc8`DqZEat zKWDaNVat06ihS+CpdtZO0cE-qeLe7`#FaNZMWG5P)9vW3`3ylReZv@3EP$H8@>$lz z>3F6$p&w6Cs0l2eStoA8L-^WLnG7lsKucjXA!*+<5Zw$-?0AYoOJOvz+`8NdwP;Xo zErb3Fpr){Gq@r(a7=kJqr6|-CwvBj=yeC1>v5DPk8y?tF0koo0Uawc%rg)uhr5;aF zXhkKQYr1p5O9UCa=`g5F0QG<&e2;r}T~LdLxbqZ+dcY9=^v_p^5R~z634_W7P*}OV zUh^dza4#sJf~P1HR&FC1^vb#lL0-!(8T3y84TM=l!%^~huHfhjo}$n|m_^jvG;lV8 zf*(0Es6qhsg~>g&v04h~(%C%bDGK$4$-MwiLq!D5XdlR+N&zHYUl==Q+Ghj}?Z8tM zlCCc-?h}NUp(D2^GN?)bZGrBVq17>5MC87Mrzo@qx?7=tYH)dBi}ytgsun;V@R_jQ zV}j=yvKpl*J{T9r_G6uNLDux7OEf8FbX_zc(vs zz|6_Ahmlu zMWGw8)pUf>1SbToRg|r7c#fZhk6JWtGJ;mM;VBBa!QA!g(0_Lk^l!TwgRBM6 zh6rhQYjHS&y6)sD3T=phGF@4BJZ#ZP)Mb#30FwT!$%gYI)?z4$rzj--S(7Q(HSj`) zM9Y{#Ck2qSr+W9r54w{1qNuGr zMWG7#9zAP%v>j^E-nW_zIwOG8e?ZN4){}GuZGXp86jJ|TBUx~7?P&xZ)iPv|qX3fr zR@C-B8ifd2H-e`qB>khoB&$>N?tEg`H=)&y3|G~ z3N3#HhtFjGHbzj{!)OMb7eI?(cFU<{`+EqA2;eCSErQvtFB2By-?sBr{K=pT0w@Be zXkx}hD-XC`G6ouMF+DJAm zs@R4uU4CL;2Du0z>8~UhwYYu^UAovLo}!TSR}$8kDs4qjTP=MCT@*m$q3_+JeTpT5 zDjTIJG#>ii+FwfcBPhk$gh7`C&@-r3s(F$$0zvm&c#1;LpjydACw?k|npPZPkgEX7 zfYp{ZlkW~eP*I~4g)(5ZWk6&So;f+W(v3lv1<-IPFWj~{bO3@jui_~R4TthV&u546 zJcA@6ltEVn(4Cj^dX*_Fn;=LhlBXzi=Os)Go;r%l-CYKzG3crQS_GYZucu#zB51EV zPf=(Qbn;sqF~?PZYaJ>XbWH#q4~O&W%P-?I$~K+hDGD7Ahw12J8LiOGU{$Cp+wj1; z384P)_8!yw0Dh^?{>@Vq>JM*k+t)+z_h|f-(F}4IK-*yvL{qn*7+t#NQ+bL)+hGx; zIy!F+YSD0BONXHMV%bIt0mi@)U)%;H5f0at7|#DL2{4pc?|n3qIx5Q;YHu z)J=w`DC7m7a*wo2?NN*Nu5n_JhX9hE7In&c4W8;;xt6CWBt0!^$<<4j5cE9Ek3lyD zP&K^kzIOUK61C`JI8RZi8s2p`E5>9ZXrFouw|0>8eJV&<#&>VPskL=YR zcOVr?d5S`F;O#wrY>VLtdSR%}pxXkd7fd>J`WF5dwdm(Oo}y4Mm~`6neJGxyu?;X_ z&>aC320zVdDLWnC-b({{ib7%V)0~R02S%Y5Ik(%upt}M{y0z!PzxQ~GW}^yEQAoPA zXIS2l2n224Y{ejN0hC@Uuh-el0S_(bY~d*irI$h#`r*nf1TBj7V33ahIs(K0J41`` z%`hO2rzmtpI{c3b$HSH$!^0TlD}b)S-zadx+(dLU$Z7Hvg|5NhXu+uIc;{M^bD0df zCxEP>&(PU;$V&t*InPrRvW7mxtM66=(4||GU(2BT0_f2{dA+>HCG`l>E#N5%J^BY9 zzPZ!z%*p53-I_K$uzmu_0Zt^4c`M%?K||;A6onk%M1t$1hvFi!(1$txMJe^m?hqN8`W9)+Naxk}9%9=aew^Y=lQ>6g|5Jm4)e= zUcqeqvw||*qP5dxcfpr2tBU zy6fu?L-5G@?rol;P#V--|L7X`5kYO_CNd~Y0G)?YfzfjLp6G9s)+j}x^H3@fTd@Ui zEBIz?%%E@qG#0j9PnjF_3PBH7@)U)}!nW(XpN`?FF#VSY81za2$-t=1J{T07h*a6A!*XP_Nu%2E7(QCeR5iUKxoGpHcb2Qxr0RPT;3-1zakS zB`IUj8v)cCmN;kB7L=l!Ay}WMDAXI4I5(z*eL*dfyQ;em}3K$&n9{1K~ZvIzRn zC`F-6I0{}?WvL;8ic2*a6fJ<_pl;;xrscR);8htmkBY;eD z<@MHGIQjs!NOl2FQOG3MMv|+L>V}{_4=fn;RsgMlat_UB;eQac*q^5;v;xXGMz59N ziKDh{9T@aZ07(yVdbP?4f6B8Or6?pl#OX@q8+?23+3CZeSOIhcHd!QB_pCxK(%Z#T z6uJSMES`SMYl1Faix1HZiW5L;&?1e{4wVS%^pU41qy{ZglBvZU((flrR|EgGJq1Oe0riXR5llwC$GdU}?pDAWduALb7V z!pqQe3i>kWy#SK_uvF)!ZZi-xw2-GLB>iEjunCWLqQB9&8Tt%L6hLiZRV!H4b`OG- zX7Ut;+QO>VTKBzY5%k^Hgh5FH$P_jyPd}^Ug`k9cJVhZ>*rdEDAQZprmbE;}pbrA* z8We}O8Mfm!f(9$_6osxqarli2)0?P8^=sW2^icrqg^e&LR(C2kLMWMa05oYb$ zzPN~}e`F|wk_FIcs5LbokYA3V@ zkJstOUtP$cGyzl(x3~J);|{1rO|J12h3et<&dPFdKrPx>x`RRK0_Yf&MQ!?acQAq` zmhlvYjzL+JnZyupvY0p9i9ufkQ2T25iJ!mM=OJjs9G;?3`)V7>xKmS%P>YIg`7tO% z0Cj~D&`%dD@nTcaZJwe~S118hHhzgO-NL4E49XNhePL^lQp?u(uIt^5rzq4Hw)Pm# zeTVlD4PRZrpsxbx2uy|bzU_uL{gpLJQRoOvg^jcr`3qgT{!bMY8lIzX0;o4YWwpEU zw0}vX6oq;N)TaN%a|n9bL7hQa0%#kQ3aC3wz%SJc9eIjE+n`in)W2-}n*wRO3>cIx zfL_6bhMTh{E&*M%ho>m?3MMp$&FQ}gwdiKbdIsePprJ5Hp1RZr*EJZZCSum6xs~mBfH>l?g%OgsAW)|0FwTI^_f#qxSZonAWu<9`UBQ= zoAtXQC|;>stA+>mmjLpC2X@fviAzz73fuA&g?!+FjmR60$FTc$=`bi?0HwmD(^V~h ze3r%G-8@C1RG4&1_i@8#l)a2w!k_{Hqz7}VDU+t9q82&C^Av^jU`{pHVqiMDbhigs zGN@1h9e~-bsBx{F5M(ourzms)X1D%%dE%R4`Uz(S{T4vdAA&!-EZiSKy{veOLed|C zPrkbePe)(+8o;1G0_Yhm+9rJ*_Xt6Ezws1>p24E+CX077(WR3apU9vh0rU;-x+b=Z z@$A;GMkxw?gS+l^a`;sQm0c-fP_Y0igMOX7lNR1Qn{}0^C{zahx;x6gEf7>+-KKTJ zb5tULvSDaBr)4K2bm>xRc#1;VFtnUJb_<@4p1oXyL4O5ME-cN=YHza~L4#KC6oqnO zX=eH0fyueU0| z_B(>S+VK>HuGc_qk5M|_4*cEJmO(4gYyNRqEOs(SXgg)?mcQzR;7$$!*f(6fNn~Q z@eJ}09m_7(nSfa+DY5_DHCNzRNFT9HWMk$R_6dDc_8umpR_zQbD zP?te90!R^NAcJ+o@%U)&Bc7s=BFsRZx_k#O>s;?(%%EBUR1f7GdAly+eO{+J@)U*Y zp`0V_-(0+>_1~TY45|}A9pMPREmx+DLYMB1B~MYPBOJju;_i-K=x=0_=*pma0W=C` zPTCGki$l;AzkR?Pg&3d z_f(HMBs1uk0Fq8Ry*Srg8?|W38J?n$bkb?j$%I(cBF&$r3_31=s-YKja)vIRIq8wd zQxvL(UXZ%#M!ZmEJxQfa!vlLl04cy+(2}9DzNkg!5}u-v0?Y-;Z`q2Er|5Z2lR;Jj zXbkij;^MAZAgH+;Pf=(L^cgn2+Pwf>x~RW%8DuShv|#(ynf62Q&EQ(fQxwvI?N`&6 z^})xDX3eo+kc|Lxg}>3`e4C4?MGWMSckUQ@rH1FoP5`Zin(eJh9sVIGaV$?!Xf4!i&v-va7qzGd?1OE9>;=$es3<#r zy&IkwtZI~^&}FD73l6A%fS~0S5(YU4AQ>1eUk@0JM|?{wd5S_ZFjy{GW3GXq*hMA` zIwOGc;AWV1&(;`Sx?78Rib8pCGyM2dgeTY^g&tv$qX5zglh?C%IdBX?H=gkng>=H8 zOWeo$7HZMtUTzFJD}c&FIqgx}n;(;W@e>fSjNMIXa`b4qZBZ51yis z6LcVbgRXlcD5c402005LcX+9;SnYss@4GTQMIm>1sTS%e;Kjt|%NH`pMF1s1;qmtu zFYvG>zfp=pNl5p=U3Pf_R-thOj6Z|;ri7)RRsG01fzs$*|D+Fz> z;wcJkh86r(J=fyd_?-(27<5Gd{f55Tv^I}LjfQGG8@0%{@Fjyh1W-8CUv1oc9)FH6{^ltPg+u*S z+VW$)5wu}?CWCGYpeL}ud${&+e15F<44$IU6WHIKzs)ceLCLpk8RRK|cEHE$(y3X^ z(WQHSho>mC13q4xw%M5@$e=~Hb`1}#mjJSXg$&)i2Y9BpTT7mzkPR$kWNLmMfuN{W zIt;ocfG*a-*|Vp{>Z2B2Sj|%ux>yGzf*Xf+BIx#uB@DVPfVx9Z=0(p+4FuV|O5OjTV zB7=McP%v}?)7qTFHQQUK@DzoDp%Zu|dorFl8h5jZLB0Yg01A&?H);Grf1~c6JVl`Z zC_FaG?vD#pRyS8tX?TwA37{1)4h@~#?*f7*x8NxXt$=ap_xi7e=+dcf(qPbi0n`ds zhmQ>@!KKt?jZze91*^mJat&$_G%R)|gZu>0^>{cYF+h6`YEg?go}$q8c-Ss_Q68`0 zy9_aB&;tSV9Uj=!zn$=znN}J+MWOHTzz!bs9=}u%I@vPFUjSu8xAxRyM||GM%yT?N zp={{Zh8T6rLzixN{w)SQ6hPAYs~p4Gcq;5b0Z&m#T7R`+=+k)!s+%3jpa20REyhb2 za-uc*8x5GtQxuXGu~p|DOC7x2Gp zD%-x{IeH|3=E0;yo$O0o4}85DPf=(dOiC;bH^-#{8mrY9^jH9Ggf6k;{N8wVxVlk_ zLK~q=ym!GHJh?air7nYl1kilAy%(Nz#ABR+VLU~l`EYyN9J_*hLGk^J85Aslq(6bG zR#!e2-3(s+d5S{PpFn+B_z^GK4z)VKpb!DH6jmN*4ba<&pdQvdMWLmz@;G^HFY~Fjb9G6uO!Y$Bh~u$Fqnol3)fs6+lN}LGZctAWhVwZTdV# zp`)-M7-D4-gj%HJp3I<70rUWRsyYKpx*+IRqZEZ6Ku>kCcRnuk@~ACk&@%ya2|j#% zrYYe!gH;_*QRotU_@15bidRh3SEzJocwnCkpbS`=2@fm7i%sQ?QWVO7rJ3AT%E{=` zwRx(^pcev2I*a&z-HAp5fQAj$AsP^bQuCgd@ZOEXP0w}gqbG`6AK6n}WgCb8+ zD7Mo{iL0sC2GpV@8!Z?VCVg&?hW_6!OaK&Rm| zQM&Z|G<4~@#qtz|PQz#7M)wr~2&x?6!=P6JDCsd&E4|Rd70ziRd5S_wkKvHp2Yd0~ zNZmD>K@kGz1$@ezEYkHyEmFSBQxtjupYr=3w;e<+QmOjEphyAK7J8~<;_~qtP`hfL zqEK7tsTyxtfIIo=tK~a3JV&nukoOZfZuE%p6x5~sT}&r^21N-VBiO8Ty8qZb1i5$SDGC|EW+i0@ z@8bw6-)+L6XaOYM`|!@^Y%>If?cpg3N%ub7*YEp3J=OO|7!)IbYGJ*5fvPrs*Bwjb zDGJrXdiN9iYTRd7sO`p}w*qJme9HftMUO`<(jCK76q*B{^7%hy@bLfJ#ZU&l6F{jj zcU>3M`3{0=FYy$GQep1;w35X|^f$Ux_L)Jk0_X|!8OG#H#&wLQ#!pb_wyIN8n+uWFsL z<0%S_fX~D}tA4meeR5(M^icq%mch@aub&c!E?se>6opdDY$V%jKXpJY@`fMjY`DFX z1&|_iPBI;?;6)JMNjya%Md+No*3F~7w!S6w`+J#4NC9`bvsr8+_OKw6tVUwgB1=-=n^|&2g=% zvjR_1Xghq5p552Mv&24|of(uPfXZQ(c%A7hFVv!dEj&e`a+oFlnV3BTT{_+O0Sx*s zfP7#vv9H1!+-I1W$Ws*ZfyKlvr<3rO?d(B`49XQirSRdKH%Yk@YEj@|o}y4GeE2Td zY<`7Wp;j;hS?uAMh+6c?YbJw! z37|)?C|`ZO{UG!=dU1=VDD(&x<*)v1S%RQFvgQoR7eLKn8m&{>{s08&%kdP2n!_~O z+ppgEjDe1=+N`$U&PB?71ulo!T# zl{`Z&@_ouv6zT-!h4$CoMs{_{HKjl^naxl0<#|*N6Vq+=Lgxd|5dG^`6;m8 zEi-QsuCho9<|zuLzIG0)%jSBYhg&EhXz(_kqEK1OlaiimX6;5T>hSIdgN~a2_s%BiQIyQX1Gpp$V4p~vR}1F!vlL-00qNK)%@DW zLkLnnz*7_ohL>vV{JHq{UixJegX{!Q8BD;w>M_P0K^hr6MWHg7fGw)e#xqoE;} zUI4X`)+7{8!#}(`asp3Ls12+~FkK>)qI3j2V<-k(A(n(xC?6ncBr zMiS_!h~ISwyE`%Hi~x$af}+;bdeI1a*n_7i6mMlCX|})1|Ey|R`7y{*0Cjr=KmN7y zW-)>;S@RTyx;=uk{`LmoX%@}gSO%RHK-=KW&_%xqF7)d8gQqC84c-hh&pA1xztNlN z1q^Z$K>OgX>r>=|-wYu$c#1;%;I2E`cE%J0jeV%lv*9^9Cx8O{VLoTRnkBk)tpa$8 zLIM6TLv^ADt_SYaR-HlT1&|rc=S+(FpJS;q8>J{@2J<;x7snN#7VR)MV9*5t6btjB zt(QjOJ&!9k@f3w(VP4d6z-|eG#=c(9AZGzI0p_liHnzsUfzt8~Pf=(B%w11f8GFQxr0ef-dp?7W)xo;aAI`%K}Jx{?(6+a{Ol4_kgD;Bt8G?iBn6wT{NUs zH?@Wb_KEW{|r8ii7n9b&G5G_HOCGQxuAW^#xns?|6sc z-24CrT^B%;p#=1qRw`b!?OwoB6q*brpuc{#az`ziG&PYyHv~`=)b@Osc&8j)x()_B zMWHCD?J;q^myMt{9z_iD5I_syUFYHPOM;;6Mkxv{fOp-PLVx_3sI6+#tKm7iDS*N& zVW;i)`bY$QuI4EUg;&~0EW0egQ_6))G#KP5fX2axZ~w`+ItYqc%2N~?2OqwiO;M}S zr91yYLQ1r zo}$oVSTWi0)dYX|s&?5j=(YgLhT0y5Ro6BlD0(+fQ79W~drtHZ$7?|5pKdYejsW@! z5A3*Az3^#K2B|znp|9}3{%bP5CHfm}(T-%$T>%sb^P=wGZ?`}#(jCK76bh7{{2+Ha z7eV{ZXEVrK0JVcE3qRR3yi;J^1)ic%JE*dFwxxGDg3ABN^lo^Ld<4*ISjdQKz7|)M z{jA_A3cZGf47Fiv@rSQ;i5i1^1yC~dy{~_?dV?-q{8FBxP%`wrM_&lU6N5=Xx(vD} zfEL22yk_dL-3ST^<|zs-gi(2q!U9|{sngz=LH7mFWq8*uZe2G8K|MS06ooFsyUsDW z{T=i-g7pOk`3az1a5K!Zs@{g6ca2gM+66a*!Skgpa__+>!th@kIL0! zd5S_2FkkoZWy(hcJslRzAb$ZQ?bl7NHXMywHY!vh;AfE3|JUFSTwnt(1{z*L^1kRtr3>*ieRcj(f^+|*>yBLUoxT0Qo|Tf|4E9q83Fq~Uv zgj!TP*oQ$O0>}g|U17~-PXsAy@Dzni;L>$cDQbqGkkio&dLn=h!b|nJ%w=2;>}1DN z6gmhm)u7C=xLPUd`ws>^6+ktxXj@p9hX>2wa(RkEHLz%#ad^~3)S?;s@_iegqfh}P z18)Z1mRDqhX`tWI+Q`-0!Vr!?6|=(c>O4|QHny+BVl(weyf5Y z%O9T^^hy9thEMs$0}ta-i;RBq6on?kr~LInb$bL&FsNiugaDF$kIuN?j6sm{G@hc6 z^n0}7QQcSssd}mQYj|KI1&|6nuy=o|<0^~sw|I&|D)7KgFYaN1pe8b-8T48JNq;|O zg4K>7)S|3LDGEt{KV`dV0{$D#ShkQsZv@a$sQQ~8|86dVRG0GxpT1+4K|yvGYww&!_@LbEDtB%j#@N*z5#>c1yBXl)w?Cw9zjr-1w2Ke3aG2^zbj`Pf^-AcGblj- z{f2_cspD>pL(tklo}$oiD3~<6se|WKcXza6(0c(S1M5dS7GA**>`YajqL2)%AEhs- z#-|vy*yF*VL;>_0-VD8Cn(sm_`rIf*q2KUkSoX3@61sHJ?_V+~NdWbR!E%q4kMJ=s z*AjV(LcL+ITr%f9J~d*BdM1ND2%u&4@bf@3b5c-?+703<3N5RLQ`Z!R;#bh0Q?(5G zD1aV9zwXck&nF0~IL%WOdIo(IDZ@VS=!MbN7fo}$q6 zU>LS2*Wsy7<3&0QN)bS-!r`nEtKO{=u*Bw1xibs4?H}e#Q;^40Ptl1V%qd7(eFeps``NJ%6$FeB=_Ff;&Qxx)t zS>i5z?%;ySv1*A7N*6%wYvuKv_rAG-ZU*&UJVl}QwKftxZB@K&WXs+n27M7g3+v?d zR=1xs5kU*~@f3v?*4aodrMAJ#I*UKH8QAa~WeA`$C~o=sLFpWVdMEP~h036~WtGYq z8C1u3caR2yGR@I>j9muJoeZS8~D~5 zb3E<;_S{VNzxpa@z8>_SBR1vkK+V5!o~J0J2mNR7pI859r)}9r27MDihEP$~Om+%h z0}3eTDGC`vMcMn9=Xm+-_&i$%WeFgE_@?yj<&B>Oqxn2VA%FO$+?N}Oj|}bc_!fh* z1&}rzk9=zDof+uTbqnGt3Teaf$ltBs;x(Xz&XEkt5kPSO$+|e=OXt~zrzjK$kZQg@ zo~}H&H=9A<1<+7fwB31f4&L*)e;-d#Xecb&{>@XIgkFpbl4R5y9=cosWCHU!hJCIN zK$mXx2cDvk3C!nMzKPaD(A-gK4EiB}LgHY%X!w{k1a%nAQxpn`g9+F%y+x?dg zehMHh=s!<(>AoC6#~gWzLR!#&?zBzQ5J6x5tYA={0Gb26P5&taMDLHPn`G1QG{7kM^A zP|OpaqR?Wf8?o|8z8X)>rt07XM@^ULk9 zAqaZl&QlbMhTi7du$}3sMcJ~33@R2t4`7MYcxaptf`a6Dib4-yiBs!S{R#wKHMU?- zi2%xhy~A5#`VB(R=9N4}p&Zyd{33q`E|@(1*q%Xu1yEDi2Q=8)46kZ!4&o^aHHCdZ z&3Z4!cb!c;9|n~QAPtyBywfTLuXoQ=;VBAfz%1euKg;vz(p7DaW>A>`3WZ-(beOYl zAi8uzEO?4Sq40}}C3_r=5wt$~2ZPE5koPZnJ*%%xafS2Y7@ne#_b(_ST2+ku43>lC z2RA%N{{+w#IKJT7y_dsLi`EX|DGFVI;|pFd^TBJW=8k<7t1#5($GU1yB%7N?dpw@)SX4e|d^RK`<%tdBG3-H`=|xgh5pT$oiwa z-mtPE_&GYfkf$hQ{SlU-Q~dGD;|KpE45}7D^Pm@G=&m#twJ7-^Pf=(d^n#|0T!DAo z__uarP>ldu4sV943G6W!gRBM6X_(v_zt0Y@4!7CBQxrN4lY5$@ z+r^+3?S2!>AR7VX4W$Cv8*^>YrAv?EDGGT*slYU?(?=1sQLTVMCk4<4*Z{j`q{9^i zY4zeM3VnbLu=&50zaePm2?dRY=g3w7r9g`;L;w6jkc|~jQ78pkn&41vBTJ(3C0fSBppusS-%yG(&Lr}zYo}$oT7+Nk= z-PIgH@4VJC$W8!BU~(_IhkgWtF5coP3Q1scZ^4WdeFV*@JHa4(0n`#YkWo?2n-Qd1 z&r=j?2_48NTP3{7B4?QggB%3VBq(>cG)==NsKhSkDGE)3a`(ZWr|`V!&d`?(IwOF* z;aA3XTWaHVx{1$tibCG-D`Waw3ud89H@R~rgB%6WVrWrJFCANS=>~V%uSO{f{eTHr?*jEH2wG~tghA&8 zPzdaoG?{r1uWAi-;3*1)z<$YDBgW!wjQf7=VbBEu)D1p-E3&`hE!+0_JVl{y@ZrdV-bTc z2_PLP=NNS{46hESne!BdbfBC=Z(YuE)S`^2Hp3d8BUb^`6G{aP;+l3um+pNuPf@5R zlnR86+mGk#Wcz3^=&}HcfWh)Mi^Y8qw6`x$Q78fi%T0^V;ML)xy)zkfMF0gfYp(ab zV?KV@MeO4#3I#MfDbXF*@h-Y_RcRX;bX5SQzLnSOu4!k6TJ%1hrzn*A7B)D`eOQN} z_2X?BbWH&Df(~Rm-!*t>Ibi}%QK*-+18Gx*U#hmQw;1FmfM&v{{IlxHeyBy}mwAdp zGvQM{*1>5EYSI0&2nM+eppJP^DqwIDzw2zvd5S_E^I)~*dSDuYPAtr3&~*VM4}<00 z7H(%zi&h))6ouqru)p4PpT>vJG#h?^L0$r=8`NL5 zm({?(xt7)_MWJp`fA!?v&%vlg`p&Klx+Q>m!{n`>N)qmSYq;p?Sd_?)Dg~<%MBY>m_<6mrg4gZbq8u1i` zqzB{wyt?8NYSD~KtEtEXrgS#@2Ew?0(pu;KVUAXd~*C&1O+IoXf`~s-U8?a zeE5u-)R`m5rX5dF=mmWEraP$4M$qAHnhf$0KqXKQd~sf9eFPb8=P3%6Kt1q*_w(`J z=uY%p2Kfr02{2f$Ih=!6mJh}76on?hU|F&Khi0fn{fAjF=$-(ogGG?ynQw7%_^jbP zMWH%a1aVaQWPqTz&h`wtFM!OR$m^Y#*@TY{%X8r=3Yk5DZ6m{vJV(&6I&TK~382dl zp-eX={uTNgt*_@P3SE9^BT0y=D@0J2RnZK3Ab_MN;i3>*c3F|xRW2XI+Q`d0%!~j|E-J5V-fVlgr_Jp28RECmcPRVNN%B@85AOb zX28seS+1`vg7!Y+DGJSinUhy9^IXu)@J6+gK~DtGC@7`w(B_aQg3fp1DGH5(QtJ5c z^YRfCxl46K!vp(N0Ck2Iwce#M3_%xn^Av?TLyIQQth|aKr}$9}3Kc+)p+(Ov=V&5m zLjq4x=rOcNX;bGp2pT8^bWpAfg7mC;ibCsPwPksi@3=xEFXt?SUI-xRA-5%+s`1M5x9>bfA?YEvino8^ zPJXPuAA?>BAVuf}?L27_iCT1TGEY%R5qd#$Ph{XZ)$MMv3EZ? zDGJ%aVq$$SIeZ>t(Z75Kg>OQ2j6bT|=-N5N;koOldaVpgI*ybnVrJ1jd zQuKc{6>58&{W{vBTeP#O!pMe)?v0MdfxvlDxg zE+WY74^L4@3zpAz+;=iXki|R?21N;=BhaGoCR&pav~E66QRoP?sL3k#vj`gT=p}=q z1<-D|>kL+J8-XD0$2>)$-Eh~%b~egEkaoLF2E_=Vdgufy%ov7idt_92ibD0!37j$^ z5^sQgv8|RtZv~JWeE07huIzo2`gibBEgs_%NG z_&9>jjdf;Ff&h911780pW75&z$VP{!DD(&hywesOUWK4L*8&*yUI3-SAx@TCKH`-} zYd4;vP%0eaKQ z05XS5cSUdWb#&>n8>J{@4wvrHue(90Mcr<<8Qt(4eGot{P)gmkEa(M-Y8s^|=d-_$i?hLEAUZWKgmIS^?Ak?RNB! zK`jb2=P3%UfNB4Cqx&>P(7+EH8I&S`6z{_1UbWX81Sx#vDGDjxg^ha&H}PGUqixHe zPXeeIR<$lxYH1-TZVXRRs2EnY);GWEj#^~ka*IK!0w^7(!m8da&_+<_i#$c4beIaW zb9#e2C+UA981z{HMZuzNJ16x=2=XfBDGEiwqV36DL-7`>+}YU-N)tfRRjtdHhT%zx zhjVy}Lef>OFY4ZSQljpTjCR9wlrDfm;idY{y8<6em3Eh>C=?1WRmU1vNzOmCE;&>J{Gr9S@4Tm%)Z(PdDE0NM&u%BCZ#@E9k4El*KsD@-YC zR1U)H3)7yhU{Iz2vMQ0+dytler~Owv=P3$VmDorQMXB1M7WL?MfI(jckR7~&j<3v} zjqbX?jZze{gICa(UPtlrncZGj27MDio1q?9_wpC~W-#B!Qxw_^^}wr^9MwiG`t>P@ zL0JMw29{=QO5*>Z78R!Q6oq79X=cv*7Yh-Tr<2T}Yyq?(9)33c<)r}#iWtXJ6xtAP zBk@`j(HTMaFO@PVM*vyF`a<3zkLC!nb>%4vS;P9m$-H@O5OlS?-I#_4_PYSm0qF0Z zmNyX;@sFn{qytcj_3C z4*~T31^kHK+{gP7RNN>6YHI(f9G>-p74BMIq^y-t`xooj_3AEN=$=5CHWa)z6s>mdXDz^>KhDGE8m&G5_n7Vg*8UjD(L0s&MG zMMSHreZokFQ>JVhY|ShT&a znuK?Al&tK_px**$0o-+#XS(3pp6RQ3ib4zEuDh~3v>vr6>!pN2e+1AaSi#r5uZ=e= zy$a(g3SEK~`~zX?xTw`u&4fWk0%#Y!z1^>9O+kMnlU_VUplr2D693@Q;oH89ibKfzKRwP<@PPf@4_W_mx^ z_nVCVMm57j8T3~G#RJqKJ_a9hTcpWT6p9Dvfx_it1dVt4%%D;MvGPkcMJ@eDG3cKFGJ~4!CQVH6VpD#j6ot&7X8Yu& z0}2S*+GZhxDg;n4%x(=|u;?GUblOThMWJBn?AFxgxIm?=#SR8l3ZUnopdQ%F6TiI+ z8l@=o{1g0=#e385s71z6XBku_fEGiS_?TKBJRLnanx`nV7`nt4ZtSW4$l&I zp2kxYN`f;M6qfltMbI|Sfefk_KsGQ~KC$!%9xPAy;wcK*z+idVBrRM@eXe>cgN|qJ|>hvhdEGhSP2s)J|7P)4~?ZM7U~TvVRp{{5P({A z?_e5%j&Pt^&}V2I`W9D-gWYI~g=RsYA>Y!o96^D#H3T}!fjnWNDq-7uS9CKtzN9G@ z@`Q=16GjK{8o5H9P6BywAXBL1&$+x04=sad(-aGtLM2~Cb0S_!-TUH5o$d$r7zZ+h zH>_!E1s;{Z>XBk0LwLiMm_EX1l??3EBG7RTG`B=lPwI;DU+B^`^+>VM+!83GNfgzg z&q&mKHGw=i&?XqeoCXg6XEt=q1FdM6EkUw zg{%`{t!Z10B!Z^O2N1}M1LeT!E0eS4E<;f7F*Lu@Lk+=~(do#a50P>!CXF?c0{%4=zgg(RUIt*v?!Lwfq} z33Q4B3F2XyZvAq71SuKN6blLBVSh^Tctv#SHs6q%-TfT-a3CGHz1`krJ0NIY1WmD! z4&2^vp2}4rNKb4sfqXg82>6VKm@4A&(ZYT-#X=+CGqRbZehEQVn->%4GzTh&H>|iu z;!xBgcT1XLp>lY`W`r!ULD1e@D+2j(AO)C|-;|s?A3?@>G{r&+Fez_Rz1#pnc9V`1 z=nMzi0G*R%)6Z2RXsr@WvCszSoD6H7j_cmZ-q#4^&w&DA3cIN7{xbx?S78VaiH&CprBC}u^T}dQ8dLu-@m|Fe>)t@P>aUM zj3v-{4kSFx!R(9`Uer2q5KXa=@GyrtNpHU+$jDxoK!F@+2rP?Q8*Ge!)aBDyY}_TE*p| z&&WodrdVhd)a$ghGH#(uSK}5)po<)+N>)toW9wnO2+51#e?q6HHSqbOKH!@UU>~i=fe;2hHhzVE^Vo<=>%r zJ9#(08CrX!Sg8CvoDck<{Q-jVOf(5}nFEc7rPRBV;trtC=*)7OVxjS{lzO<=gaiZ? z#4aUJC%23Zrda3=^v%K) zKjMLir_#oP%D^lYIg;?bXKJ_#X>PqD@Zvr60c+2qq31e zH#txOEM0TX%Ez^WL{*w%p#)gEHrs5^9n_**UIz&j$$?6sC=vSA4}aAcdD9dNl|WI# zc2imnf<&8w2^7VFGT?O_`SOzmx^x{qQY@4Kuj84QlkkeNq{S%&isnFG@O|a!uYgC^ zUQ1|-g}mVV>ghm#e1eMd%?bj=aG)WuOfHW!|4RN#esx}axd%3!)5;~IW)yW-G_49+u>>Gv+`yHiswN40IKoT!iA&F zV`z$n_CZ}V?ZDMFs6{C*_5@1cK(}BV>afB38oG2B_tO*$-GXswO5iqJW_j_{mq52U zkTsN9`br0SA?VdJnqnbqD6{Oh9faTe0W)I=bcX}QK^1XWaVp-$8L356EEES-L<4=x zVsz>D2Ne?NE(dCd_r5`{IbJGYdx54{s2$$>_I^`wHD2vk3xN_jkSM%iSHF8W3|%^j zE}CK?QFz0e)VO4#76tt&A?SXN?r|V_=meUlDaRvd%UYUZA$jNo&RG1*6G21nPa;qf z2a1G=b=OUjIS7(Tr70GQgo*WN-Gz85^}A7v2$al$a-bStvUvjjN$&fjX^Mq%pc*f? zB=9L}k(jF`fl@e-1N0fnRen=ME&9?U#X=6yXSf|y=YSynio*oD&w*yZ%;WkqmqQVx zTS-$aGy`TH-Ar6v5Oi}!D1lNrkg!6fswRUM5#66jQ!FH`P|xzgDd8K+uKHG{r&a!6+ zw@o_da34{Qbp67C@iOr3$}QcdsC z6blLW5LML$&Ot3YGFqEJnH*>&OmQk(@5JvT3wfGip^-4f>1NxRkDx~9)db4oKnvhA z8d0(iuaUdEkEU2?0enU+8x6e>q*~!Xpll8#1K$kGRz3WHZib&dQY<6`-wfI+=GPJA zH!FZZ4>?c^9FthQp$bDbIyA*XEv4{2N~;=xpye0i36#TuUO<;vTyY$p)tV4YQ!MlX zy2QHmr|^WJ=#LTt<#M3S@Li`9kd9mQx<`tIHp6$FAjlh+l%4*3PoO*wGys)6z7>LK?95LDJW77;4e%9g7K6z=6VGo^FO{rX*_7Bs-d7 zp)i=I%ji&EjG#Hstq4@eff}yEG4fv>J|ak^f~Hug;kt*wIch`-f^w!GC(u6}s2pbS z$GLpPd+a#7mQ&gD!SrOi=2IFiiNsh3>)>iyc0p| z-((S}m;(ua&HU^Bd|aW@Y@sO@68@U`!fIpu%}}(ko*VJnW-d)%vF_u7O9LNb4 zhs&B}`yr^MKTWZa6D$r-i%7tI2EkTc0zKwHHn1h(NM0Q7d%JFQX# zUY0q5p70>(ZuzA&<1uVNHckBtBpiqZmmWZ$(cpHmP%oQ#M@pOMQ6n)-KWjd(&Y1c|iXCD1btB;2z7HeDteLA5y&#A|!(!#)wHf&*oXi0Q?f z%yCDM!&RDMp=^<3f&(GR3h2^l4jjC&`+=?GK*9}Wid`R;Am~?*6blJAlqu}ol8qn- zYfS=GaiAygU1wMO2EX`L*w7RUJ%R7K+JZCV5ws^~DS@gvkPeKGf{*&(>ehl>nqnax z7$05MT`G?t8-*PNs^LHb;8k8~6S@Okx+UXjiiHNit9;fb6?~3d^ImTPz2HD1(r`1_ zzu%3ZXa|~NAra|gg81_mb*M#Nr4a%6QGTO#n@?kaZTE z)H+~75o*!5pM4i~KS%W(=o9Sb*tcnJ47zk>oixQlpI|q~dedDe5OjFW?*w|qfl8ql zq!Bkt4?&)P(i97oLN91tSgbCB%F+Y`YT!TuSbw!KKOdK)!_sMrg#@tv>en7ky!~ph z!bSo$a-e1?-kQsdtw1g6JD#Rks2Pg4g81opV*R1dK>{^#pspY|kI^_CmuTvIX^Mrq zf;eB6u~2X5sjh9SIf8BmvzrwJdc%RP!_=GV$>xEOyo1!ZTuur)Y|Wp2I4O#lt@2 zA}Fxgmq711kUC7geGo|E7vJXBG{r*dF!dI;N;(ulMe||^)Xsq_Vfb$x^m;Y=jKb&B z6bn_t@PDJK=3WFn4=g0m2M*-eM@-M!G7_(Z2??Sp7IN%!OfYEc-Z2PT)X_qqj~vJq zR#`MWTy_GrNd6m5v5+gQvM80jZ;7BGDW9ZH?Qx7r;-ExLW0rdY@wN}YE;{e>4y7PQt7=m!Uqgc-3uQ`2rB=yDrP zv5+Lph#fX|#lx25#hnED$$`efu;qNP1D+mDTtZVUG!BL>pC_f@>Q?QYk%rw5Y$pfW zeignM`f1=V)jM}-iiNgcgWX^MsRL6vya zsRaB!TC`_1fx0+QDeOBocON$xK~j5ZiiJvH-*Ms3cD$2%=RXbva8y?36X*yBQiV%*%IOQP1x+=kDHc+NOZTS!)Jz2Jh>%*`{Tv58jitysImJgeXpvnQ03FN_n?BHjJH>GZJLr|C`O|g(2{0wnH=oh@gd4Tm| z0v+Q(X)v_R`Cx)~yEpYnu}~TeEyvCi-GJ6HHe_27=(sIf$7r0OrgOJtAfDokD5C%Q zU)M2SNrCr%vN&GHSYAw1>|b>y1@@)Bhh9snm+|j6?pD!XoRkFz!aKd zp(${xfKg*7K9=h1>1zZ!!GR=Tz`J_+$$JRO@uMjgl7InkKz|=xI0}88MIbK@bjG{r(rzhJe$SL#o6>1NKaCy+M>(t|I?_da1R2$ESqQ!JzhUyK8H1mSPs z%;?{ibU$<_IgssjQ9Uh#Lfowti=`u zXo`hK=E6@Zj1R}hyZvoYP9T2{qzLoff9QAYLXeXoO|g(7%y-Y9QB{drH1p;s0tIlO z$?#pL(CC7z9}^;JiiIY_cb(nva=gMhO=7T7_XB&D1LZ+SJ#4(cIBL=P{xroxdC*bc zR_ZnywMc%eCV|dzATKB!#U7HvpON-9nqna@C>-q_l!?= zUhPj)EaVJpwmn0)obmcGW_&(S3gbo`sB-jv}+c=uV_Tbg2_!8@SL)b6pvxSnAI#H5yByF!H-qdqnqr}TFi-bP@T(lP==SP^1PbLqVsLJc zx!gj$8~W-Rnqna_IJZY@(xV6j-MSx4per2cPq^#G`M2TqR{^Os#X^6=U6)h-IuJpV zho=xIj04SuGhr4=eZu2V*%36wLUZ9vmNNwegoy~6PM~lO)B?-h^)@=hq8620peYt=f#vRZtta3X z4gO_Hpc@?M80>~VvfT(z&3x;TVxeQO8~S&{bGYyQW4%3rA~=u*Y!_X(@dECgJla50 zEMx)OMGr?DJAgi;#=E`*y2*jMI^owS%;WJJwj`0JSg5PhLm+!D7*C%q8WTgHNDfp0 zbE!4T>&ww+Q5;AP7Pa0L8}<)^+Iyr}NDdaY%9mJb zAV{OOg+S39s1}OkzJ7n;a;5Z3nqr|^D3a&6ho3=E>;j2p-Oo`B2RaH1k5jUL*rOH& zE~F_IItmMqqhe+IOzK7cb?qExC@!7*kU+7v+H(Cw1`ShF-($N|-Eybh-aYEJcAG7? zt6A9W++%U+jy?JV4^{nt?%#`om(rqiJ)8XT1jvkPnqvRJOKFqm?)eqif*{5D4df4u z<6H?%xQQ&x7UFsG8w+TPg*4$Nnxg&!53)yvi?9J1g5W5@(xO%Oa7)uiiJvG z`g*%k0A6o0OG1G_@f_#|R3D{GeQ=*$ygyB`&<&_Q)-5*2BgIQrdIU<~Kz*P;DCYeP zmko|v(-aH!f&O5ggxGKBGs=3jjzG6Lkg%RQ^{_YYJzdJDDHameQ?1T!$B)GwWmf{- z;XrawpU~a@6VLp6tI!k+$w7T$+LZ>pm7&V>0)g&wpaOt0MPA9Gn_;LIO|eh`Kpwkp z;7O`o&G!hD$bsBp#E<-WLRIjspFMIZ zUcG)Q)r3IzInZ@j(b=Z{btZ!LKAi9X@;d+Wd4>*uJoQK-D>f8?m=^mgd7IKI4P!n?u@M}V*GLk@P97r6# z>sCMC{X2qcdZbuL9KP%7C%l@4Ziawa`2FY``mM z&Ru*>pbQSQ6^d1{N=CR|?jKB3EVLDhRk5wZlu(O;e~OuQKS!AyNE!wQeG^UauEb-V zG{r*FFgS=>B0UsAS!)#ul*NG};AZG=dj`*w`Z zRqeMfJEG6%OCn9NP%kJ}oslvdgj%#q)`CC}IZy?(C|5J}DS~o_(-aF;K#M+)aKp9O zIr|S0D2D^7!Pqxy{>D`ZQgo#$7E*(;@4UbLa!`wwya*vsE(eN%;m-17YIz76UQ1If z6a&MZiB8Wg5Tv1o#oGGcnM zs{$S)DD4zYu~4+kF@fdlHJ-Rw}FKW>{ zqY-A^4{QMkdI%L46O*N82>NbJQ!MllDlSVT#qa_#ov0ZED&#=TP{y9szaPH6Wus|| zg_@y^eS7A>aMYq0nUw_khXaMcJh|2Y?IWl~(SvA;g+gGS{P#uL__T=t`#l6I;y^#( zRj%Cc?Q#U2-c3_1^aEbyD}VjMYrU=Cv=XR{14+GtBbmQElSYts3r(?*6l@8qOO5=BTBKkwU`6+H^q2z) zXFF#_UBcT<+Ipl|NI2Vh+VhDmf=a@a3G{>mO@Y4mor-Et)FR1iG{r(wpzp1hdky!3 z@i>7m!lqUN6^}XG{r*6(6956vRa8C zqsm$WRdOIF=mqtcT!>fD$W_r43pqhAC`!;!i=dUdT?DG)K#p+ms-mHT0&3BsIW)yW zj&ShmZ~9KS1Nk;|^vdoBwweQpz-0c=q{Sf!D!4*ZEF=Px`5M|8c!8L69~}bKaG*L^ zI=XUZ2wrHptuIZnP#r8CeNh!O99_E3O@9*T1qTxT6k)#W>S?G&TAOK#g@ivvm@T<} z3WA%>*Vf_!^c{v9;PW4s)mzws!qn%qf2M=vW!6W97uR@ z)cXm~a6v;*M^h{$JUB}3kHL6OBYe&W0=?ou8E`Z7t5<%GZU#p^nqr|0xEb2-SEQmA zP54`SRrhn$z<~^)16luf0bZ*ja+#)B$N)N!6N|3lb;S4js1vA>1Kog{YKX)Ny!R-q zFHNz~4XCM}P*ie9EqZKeM4%=PBwW3&V1En$0A}14nqnd0>UB56(+dzZB-4gK%^c_% ztiJKPF$nh=e)dSQ&@)(lvtQf}Z)GqYdxAi(InWsB1yv8X&OkSV+&G$Ip)t@4>KhpJ z9<}K21K|XE!-0a~RbH_{Dj7j14$>401;eY{D)}A$J?gQq9ulYpuYWX`71!0=bzB_% zRpZ;}fBx6?k8OUiDsz)QuJr1@qbc^UYV(8Mru4~c2>P_Rf&5i(InA$x=4TYjTtLmg zw}hrxs1ll=vFRf&SDuO$S>656wQ`_F7~`}_nh6kO6Gu}l)CgmoX{J}iP>aqERUl9s r2hxLDLF3-X7_uHlQ!J#H0m~9?uiB%zqm^k{{|EiAxg(Q)Ul#rcIQ)lT literal 0 HcmV?d00001 diff --git a/tmp/epout/graph.pbtxt b/tmp/epout/graph.pbtxt new file mode 100644 index 0000000..8d0c735 --- /dev/null +++ b/tmp/epout/graph.pbtxt @@ -0,0 +1,592992 @@ +node { + name: "global_step/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 0 + } + } + } +} +node { + name: "global_step" + op: "VarHandleOp" + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "shape" + value { + shape { + } + } + } + attr { + key: "shared_name" + value { + s: "global_step" + } + } +} +node { + name: "global_step/IsInitialized/VarIsInitializedOp" + op: "VarIsInitializedOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/Assign" + op: "AssignVariableOp" + input: "global_step" + input: "global_step/Initializer/zeros" + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "global_step/Read/ReadVariableOp" + op: "ReadVariableOp" + input: "global_step" + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "global_step/VarIsInitializedOp" + op: "VarIsInitializedOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/cond/Switch" + op: "Switch" + input: "global_step/VarIsInitializedOp" + input: "global_step/VarIsInitializedOp" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + shape { + } + } + } + } +} +node { + name: "global_step/cond/switch_t" + op: "Identity" + input: "global_step/cond/Switch:1" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/cond/switch_f" + op: "Identity" + input: "global_step/cond/Switch" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/cond/pred_id" + op: "Identity" + input: "global_step/VarIsInitializedOp" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/cond/Read/ReadVariableOp" + op: "ReadVariableOp" + input: "global_step/cond/Read/ReadVariableOp/Switch:1" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "global_step/cond/Read/ReadVariableOp/Switch" + op: "Switch" + input: "global_step" + input: "global_step/cond/pred_id" + attr { + key: "T" + value { + type: DT_RESOURCE + } + } + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + shape { + } + } + } + } +} +node { + name: "global_step/cond/Identity" + op: "Identity" + input: "global_step/cond/Read/ReadVariableOp" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_step/cond/Switch_1" + op: "Switch" + input: "global_step/Initializer/zeros" + input: "global_step/cond/pred_id" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@global_step" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + shape { + } + } + } + } +} +node { + name: "global_step/cond/Merge" + op: "Merge" + input: "global_step/cond/Switch_1" + input: "global_step/cond/Identity" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + shape { + } + } + } + } +} +node { + name: "global_step/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 0 + } + } + } +} +node { + name: "global_step/add" + op: "Add" + input: "global_step/cond/Merge" + input: "global_step/add/y" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Const" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./tmp/epout/train.tf_record" + } + } + } +} +node { + name: "flat_filenames/shape" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "flat_filenames" + op: "Reshape" + input: "Const" + input: "flat_filenames/shape" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "TensorSliceDataset" + op: "TensorSliceDataset" + input: "flat_filenames" + device: "/device:CPU:0" + attr { + key: "Toutput_types" + value { + list { + type: DT_STRING + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "FlatMapDataset" + op: "FlatMapDataset" + input: "TensorSliceDataset" + device: "/device:CPU:0" + attr { + key: "Targuments" + value { + list { + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "f" + value { + func { + name: "__inference_Dataset_flat_map_read_one_file_31" + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_STRING + } + } + } +} +node { + name: "count" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: -1 + } + } + } +} +node { + name: "RepeatDataset" + op: "RepeatDataset" + input: "FlatMapDataset" + input: "count" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_STRING + } + } + } +} +node { + name: "buffer_size" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 100 + } + } + } +} +node { + name: "seed" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 0 + } + } + } +} +node { + name: "seed2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 0 + } + } + } +} +node { + name: "ShuffleDataset" + op: "ShuffleDataset" + input: "RepeatDataset" + input: "buffer_size" + input: "seed" + input: "seed2" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_STRING + } + } + } + attr { + key: "reshuffle_each_iteration" + value { + b: true + } + } +} +node { + name: "batch_size" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 32 + } + } + } +} +node { + name: "num_parallel_calls" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 32 + } + } + } +} +node { + name: "drop_remainder" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_BOOL + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_BOOL + tensor_shape { + } + bool_val: true + } + } + } +} +node { + name: "ExperimentalMapAndBatchDataset" + op: "ExperimentalMapAndBatchDataset" + input: "ShuffleDataset" + input: "batch_size" + input: "num_parallel_calls" + input: "drop_remainder" + device: "/device:CPU:0" + attr { + key: "Targuments" + value { + list { + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "f" + value { + func { + name: "__inference_tf_data_experimental_map_and_batch__61" + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + } + } + } + attr { + key: "preserve_cardinality" + value { + b: true + } + } +} +node { + name: "optimizations" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 3 + } + } + string_val: "map_and_batch_fusion" + string_val: "noop_elimination" + string_val: "shuffle_and_repeat_fusion" + } + } + } +} +node { + name: "OptimizeDataset" + op: "OptimizeDataset" + input: "ExperimentalMapAndBatchDataset" + input: "optimizations" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "optimization_configs" + value { + list { + s: "map_vectorization:use_choose_fastest:false" + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + } + } + } +} +node { + name: "ModelDataset" + op: "ModelDataset" + input: "OptimizeDataset" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "cpu_budget" + value { + i: 0 + } + } + attr { + key: "output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + } + } + } +} +node { + name: "IteratorV2" + op: "IteratorV2" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "MakeIterator" + op: "MakeIterator" + input: "ModelDataset" + input: "IteratorV2" + attr { + key: "_class" + value { + list { + s: "loc:@IteratorV2" + } + } + } +} +node { + name: "IteratorToStringHandle" + op: "IteratorToStringHandle" + input: "IteratorV2" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "IteratorGetNext" + op: "IteratorGetNext" + input: "IteratorV2" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + } + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "output_types" + value { + list { + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + type: DT_INT32 + } + } + } +} +node { + name: "Cast" + op: "Cast" + input: "IteratorGetNext:2" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT32 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } +} +node { + name: "bert/embeddings/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "bert/embeddings/ExpandDims" + op: "ExpandDims" + input: "IteratorGetNext" + input: "bert/embeddings/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\210R\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal/TruncatedNormal" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/Initializer/truncated_normal" + op: "Add" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal/mul" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/word_embeddings" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/word_embeddings/Assign" + op: "Assign" + input: "bert/embeddings/word_embeddings" + input: "bert/embeddings/word_embeddings/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/word_embeddings/read" + op: "Identity" + input: "bert/embeddings/word_embeddings" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "bert/embeddings/Reshape" + op: "Reshape" + input: "bert/embeddings/ExpandDims" + input: "bert/embeddings/Reshape/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } +} +node { + name: "bert/embeddings/GatherV2/axis" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "bert/embeddings/GatherV2" + op: "GatherV2" + input: "bert/embeddings/word_embeddings/read" + input: "bert/embeddings/Reshape" + input: "bert/embeddings/GatherV2/axis" + attr { + key: "Taxis" + value { + type: DT_INT32 + } + } + attr { + key: "Tindices" + value { + type: DT_INT32 + } + } + attr { + key: "Tparams" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "batch_dims" + value { + i: 0 + } + } +} +node { + name: "bert/embeddings/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/Reshape_1" + op: "Reshape" + input: "bert/embeddings/GatherV2" + input: "bert/embeddings/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\002\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/TruncatedNormal" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal" + op: "Add" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/mul" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/Assign" + op: "Assign" + input: "bert/embeddings/token_type_embeddings" + input: "bert/embeddings/token_type_embeddings/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/read" + op: "Identity" + input: "bert/embeddings/token_type_embeddings" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "bert/embeddings/Reshape_2" + op: "Reshape" + input: "IteratorGetNext:4" + input: "bert/embeddings/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } +} +node { + name: "bert/embeddings/one_hot/on_value" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/embeddings/one_hot/off_value" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/one_hot/depth" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "bert/embeddings/one_hot" + op: "OneHot" + input: "bert/embeddings/Reshape_2" + input: "bert/embeddings/one_hot/depth" + input: "bert/embeddings/one_hot/on_value" + input: "bert/embeddings/one_hot/off_value" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "TI" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 2 + } + } + } + } + } + attr { + key: "axis" + value { + i: -1 + } + } +} +node { + name: "bert/embeddings/MatMul" + op: "MatMul" + input: "bert/embeddings/one_hot" + input: "bert/embeddings/token_type_embeddings/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/embeddings/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/Reshape_3" + op: "Reshape" + input: "bert/embeddings/MatMul" + input: "bert/embeddings/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/add" + op: "Add" + input: "bert/embeddings/Reshape_1" + input: "bert/embeddings/Reshape_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 128 + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 512 + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/LessEqual" + op: "LessEqual" + input: "bert/embeddings/assert_less_equal/x" + input: "bert/embeddings/assert_less_equal/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/All" + op: "All" + input: "bert/embeddings/assert_less_equal/LessEqual" + input: "bert/embeddings/assert_less_equal/Const" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "" + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "Condition x <= y did not hold element-wise:x (bert/embeddings/assert_less_equal/x:0) = " + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Const_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "y (bert/embeddings/assert_less_equal/y:0) = " + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Assert/data_0" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "" + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Assert/data_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "Condition x <= y did not hold element-wise:x (bert/embeddings/assert_less_equal/x:0) = " + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Assert/data_3" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "y (bert/embeddings/assert_less_equal/y:0) = " + } + } + } +} +node { + name: "bert/embeddings/assert_less_equal/Assert/Assert" + op: "Assert" + input: "bert/embeddings/assert_less_equal/All" + input: "bert/embeddings/assert_less_equal/Assert/Assert/data_0" + input: "bert/embeddings/assert_less_equal/Assert/Assert/data_1" + input: "bert/embeddings/assert_less_equal/x" + input: "bert/embeddings/assert_less_equal/Assert/Assert/data_3" + input: "bert/embeddings/assert_less_equal/y" + attr { + key: "T" + value { + list { + type: DT_STRING + type: DT_STRING + type: DT_INT32 + type: DT_STRING + type: DT_INT32 + } + } + } + attr { + key: "summarize" + value { + i: 3 + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\002\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal/TruncatedNormal" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/Initializer/truncated_normal" + op: "Add" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal/mul" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/position_embeddings" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/position_embeddings/Assign" + op: "Assign" + input: "bert/embeddings/position_embeddings" + input: "bert/embeddings/position_embeddings/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/position_embeddings/read" + op: "Identity" + input: "bert/embeddings/position_embeddings" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/Slice/begin" + op: "Const" + input: "^bert/embeddings/assert_less_equal/Assert/Assert" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\000\000\000\000\000\000\000" + } + } + } +} +node { + name: "bert/embeddings/Slice/size" + op: "Const" + input: "^bert/embeddings/assert_less_equal/Assert/Assert" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\200\000\000\000\377\377\377\377" + } + } + } +} +node { + name: "bert/embeddings/Slice" + op: "Slice" + input: "bert/embeddings/position_embeddings/read" + input: "bert/embeddings/Slice/begin" + input: "bert/embeddings/Slice/size" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/Reshape_4/shape" + op: "Const" + input: "^bert/embeddings/assert_less_equal/Assert/Assert" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: "\001\000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/Reshape_4" + op: "Reshape" + input: "bert/embeddings/Slice" + input: "bert/embeddings/Reshape_4/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/add_1" + op: "Add" + input: "bert/embeddings/add" + input: "bert/embeddings/Reshape_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta" + input: "bert/embeddings/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma" + input: "bert/embeddings/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 2 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/mean" + op: "Mean" + input: "bert/embeddings/add_1" + input: "bert/embeddings/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/embeddings/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/embeddings/add_1" + input: "bert/embeddings/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 2 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/moments/variance" + op: "Mean" + input: "bert/embeddings/LayerNorm/moments/SquaredDifference" + input: "bert/embeddings/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/embeddings/LayerNorm/moments/variance" + input: "bert/embeddings/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/embeddings/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/embeddings/LayerNorm/batchnorm/Rsqrt" + input: "bert/embeddings/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/embeddings/add_1" + input: "bert/embeddings/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/embeddings/LayerNorm/moments/mean" + input: "bert/embeddings/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/embeddings/LayerNorm/beta/read" + input: "bert/embeddings/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/embeddings/LayerNorm/batchnorm/mul_1" + input: "bert/embeddings/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/embeddings/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/embeddings/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform/sub" + op: "Sub" + input: "bert/embeddings/dropout/random_uniform/max" + input: "bert/embeddings/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform/mul" + op: "Mul" + input: "bert/embeddings/dropout/random_uniform/RandomUniform" + input: "bert/embeddings/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/random_uniform" + op: "Add" + input: "bert/embeddings/dropout/random_uniform/mul" + input: "bert/embeddings/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/embeddings/dropout/sub" + op: "Sub" + input: "bert/embeddings/dropout/sub/x" + input: "bert/embeddings/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/embeddings/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/embeddings/dropout/truediv" + op: "RealDiv" + input: "bert/embeddings/dropout/truediv/x" + input: "bert/embeddings/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/embeddings/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/embeddings/dropout/random_uniform" + input: "bert/embeddings/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/mul" + op: "Mul" + input: "bert/embeddings/LayerNorm/batchnorm/add_1" + input: "bert/embeddings/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/Cast" + op: "Cast" + input: "bert/embeddings/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/dropout/mul_1" + op: "Mul" + input: "bert/embeddings/dropout/mul" + input: "bert/embeddings/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape" + op: "Reshape" + input: "IteratorGetNext:1" + input: "bert/encoder/Reshape/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/Cast" + op: "Cast" + input: "bert/encoder/Reshape" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT32 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/ones/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "bert/encoder/ones/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/ones" + op: "Fill" + input: "bert/encoder/ones/shape_as_tensor" + input: "bert/encoder/ones/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/mul" + op: "Mul" + input: "bert/encoder/ones" + input: "bert/encoder/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\377\377\377\377\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_1" + op: "Reshape" + input: "bert/embeddings/dropout/mul_1" + input: "bert/encoder/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel" + input: "bert/encoder/layer_0/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias" + input: "bert/encoder/layer_0/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "bert/encoder/layer_0/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/attention/self/query/MatMul" + input: "bert/encoder/layer_0/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel" + input: "bert/encoder/layer_0/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias" + input: "bert/encoder/layer_0/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "bert/encoder/layer_0/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/attention/self/key/MatMul" + input: "bert/encoder/layer_0/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel" + input: "bert/encoder/layer_0/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias" + input: "bert/encoder/layer_0/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "bert/encoder/layer_0/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/attention/self/value/MatMul" + input: "bert/encoder/layer_0/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_0/attention/self/query/BiasAdd" + input: "bert/encoder/layer_0/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_0/attention/self/Reshape" + input: "bert/encoder/layer_0/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_0/attention/self/key/BiasAdd" + input: "bert/encoder/layer_0/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_0/attention/self/Reshape_1" + input: "bert/encoder/layer_0/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_0/attention/self/transpose" + input: "bert/encoder/layer_0/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/MatMul" + input: "bert/encoder/layer_0/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_0/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/sub/x" + input: "bert/encoder/layer_0/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/sub" + input: "bert/encoder/layer_0/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/add" + op: "Add" + input: "bert/encoder/layer_0/attention/self/Mul" + input: "bert/encoder/layer_0/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_0/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_0/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/dropout/sub/x" + input: "bert/encoder/layer_0/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_0/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_0/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_0/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_0/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/Softmax" + input: "bert/encoder/layer_0/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_0/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/dropout/mul" + input: "bert/encoder/layer_0/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_0/attention/self/value/BiasAdd" + input: "bert/encoder/layer_0/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_0/attention/self/Reshape_2" + input: "bert/encoder/layer_0/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_0/attention/self/dropout/mul_1" + input: "bert/encoder/layer_0/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_0/attention/self/MatMul_1" + input: "bert/encoder/layer_0/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_0/attention/self/transpose_3" + input: "bert/encoder/layer_0/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + input: "bert/encoder/layer_0/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias" + input: "bert/encoder/layer_0/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/attention/self/Reshape_3" + input: "bert/encoder/layer_0/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/attention/output/dense/MatMul" + input: "bert/encoder/layer_0/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_0/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/dropout/sub/x" + input: "bert/encoder/layer_0/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_0/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_0/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_0/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_0/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_0/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_0/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/dropout/mul" + input: "bert/encoder/layer_0/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/add" + op: "Add" + input: "bert/encoder/layer_0/attention/output/dropout/mul_1" + input: "bert/encoder/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_0/attention/output/add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_0/attention/output/add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + input: "bert/encoder/layer_0/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias" + input: "bert/encoder/layer_0/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_0/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/intermediate/dense/MatMul" + input: "bert/encoder/layer_0/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_0/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul/x" + input: "bert/encoder/layer_0/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_0/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_0/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_0/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_0/intermediate/dense/add_1/x" + input: "bert/encoder/layer_0/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_0/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_0/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel" + input: "bert/encoder/layer_0/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias" + input: "bert/encoder/layer_0/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/intermediate/dense/mul_3" + input: "bert/encoder/layer_0/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_0/output/dense/MatMul" + input: "bert/encoder/layer_0/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_0/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_0/output/dropout/random_uniform/max" + input: "bert/encoder/layer_0/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_0/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_0/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_0/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_0/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_0/output/dropout/sub/x" + input: "bert/encoder/layer_0/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_0/output/dropout/truediv/x" + input: "bert/encoder/layer_0/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_0/output/dropout/random_uniform" + input: "bert/encoder/layer_0/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_0/output/dense/BiasAdd" + input: "bert/encoder/layer_0/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_0/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/output/dropout/mul" + input: "bert/encoder/layer_0/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/add" + op: "Add" + input: "bert/encoder/layer_0/output/dropout/mul_1" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + input: "bert/encoder/layer_0/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_0/output/add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_0/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_0/output/add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_0/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_0/output/add" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_0/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_0/output/LayerNorm/beta/read" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel" + input: "bert/encoder/layer_1/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias" + input: "bert/encoder/layer_1/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_1/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/attention/self/query/MatMul" + input: "bert/encoder/layer_1/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel" + input: "bert/encoder/layer_1/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias" + input: "bert/encoder/layer_1/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_1/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/attention/self/key/MatMul" + input: "bert/encoder/layer_1/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel" + input: "bert/encoder/layer_1/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias" + input: "bert/encoder/layer_1/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_1/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/attention/self/value/MatMul" + input: "bert/encoder/layer_1/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_1/attention/self/query/BiasAdd" + input: "bert/encoder/layer_1/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_1/attention/self/Reshape" + input: "bert/encoder/layer_1/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_1/attention/self/key/BiasAdd" + input: "bert/encoder/layer_1/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_1/attention/self/Reshape_1" + input: "bert/encoder/layer_1/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_1/attention/self/transpose" + input: "bert/encoder/layer_1/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/MatMul" + input: "bert/encoder/layer_1/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_1/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/sub/x" + input: "bert/encoder/layer_1/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/sub" + input: "bert/encoder/layer_1/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/add" + op: "Add" + input: "bert/encoder/layer_1/attention/self/Mul" + input: "bert/encoder/layer_1/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_1/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_1/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/dropout/sub/x" + input: "bert/encoder/layer_1/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_1/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_1/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_1/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_1/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/Softmax" + input: "bert/encoder/layer_1/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_1/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/dropout/mul" + input: "bert/encoder/layer_1/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_1/attention/self/value/BiasAdd" + input: "bert/encoder/layer_1/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_1/attention/self/Reshape_2" + input: "bert/encoder/layer_1/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_1/attention/self/dropout/mul_1" + input: "bert/encoder/layer_1/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_1/attention/self/MatMul_1" + input: "bert/encoder/layer_1/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_1/attention/self/transpose_3" + input: "bert/encoder/layer_1/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + input: "bert/encoder/layer_1/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias" + input: "bert/encoder/layer_1/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/attention/self/Reshape_3" + input: "bert/encoder/layer_1/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/attention/output/dense/MatMul" + input: "bert/encoder/layer_1/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_1/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/dropout/sub/x" + input: "bert/encoder/layer_1/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_1/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_1/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_1/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_1/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_1/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_1/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/dropout/mul" + input: "bert/encoder/layer_1/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/add" + op: "Add" + input: "bert/encoder/layer_1/attention/output/dropout/mul_1" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_1/attention/output/add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_1/attention/output/add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + input: "bert/encoder/layer_1/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias" + input: "bert/encoder/layer_1/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_1/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/intermediate/dense/MatMul" + input: "bert/encoder/layer_1/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_1/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul/x" + input: "bert/encoder/layer_1/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_1/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_1/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_1/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_1/intermediate/dense/add_1/x" + input: "bert/encoder/layer_1/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_1/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_1/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel" + input: "bert/encoder/layer_1/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias" + input: "bert/encoder/layer_1/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/intermediate/dense/mul_3" + input: "bert/encoder/layer_1/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_1/output/dense/MatMul" + input: "bert/encoder/layer_1/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_1/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_1/output/dropout/random_uniform/max" + input: "bert/encoder/layer_1/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_1/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_1/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_1/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_1/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_1/output/dropout/sub/x" + input: "bert/encoder/layer_1/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_1/output/dropout/truediv/x" + input: "bert/encoder/layer_1/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_1/output/dropout/random_uniform" + input: "bert/encoder/layer_1/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_1/output/dense/BiasAdd" + input: "bert/encoder/layer_1/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_1/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/output/dropout/mul" + input: "bert/encoder/layer_1/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/add" + op: "Add" + input: "bert/encoder/layer_1/output/dropout/mul_1" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + input: "bert/encoder/layer_1/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_1/output/add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_1/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_1/output/add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_1/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_1/output/add" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_1/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_1/output/LayerNorm/beta/read" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel" + input: "bert/encoder/layer_2/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias" + input: "bert/encoder/layer_2/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_2/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/attention/self/query/MatMul" + input: "bert/encoder/layer_2/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel" + input: "bert/encoder/layer_2/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias" + input: "bert/encoder/layer_2/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_2/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/attention/self/key/MatMul" + input: "bert/encoder/layer_2/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel" + input: "bert/encoder/layer_2/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias" + input: "bert/encoder/layer_2/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_2/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/attention/self/value/MatMul" + input: "bert/encoder/layer_2/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_2/attention/self/query/BiasAdd" + input: "bert/encoder/layer_2/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_2/attention/self/Reshape" + input: "bert/encoder/layer_2/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_2/attention/self/key/BiasAdd" + input: "bert/encoder/layer_2/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_2/attention/self/Reshape_1" + input: "bert/encoder/layer_2/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_2/attention/self/transpose" + input: "bert/encoder/layer_2/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/MatMul" + input: "bert/encoder/layer_2/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_2/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/sub/x" + input: "bert/encoder/layer_2/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/sub" + input: "bert/encoder/layer_2/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/add" + op: "Add" + input: "bert/encoder/layer_2/attention/self/Mul" + input: "bert/encoder/layer_2/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_2/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_2/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/dropout/sub/x" + input: "bert/encoder/layer_2/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_2/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_2/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_2/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_2/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/Softmax" + input: "bert/encoder/layer_2/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_2/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/dropout/mul" + input: "bert/encoder/layer_2/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_2/attention/self/value/BiasAdd" + input: "bert/encoder/layer_2/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_2/attention/self/Reshape_2" + input: "bert/encoder/layer_2/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_2/attention/self/dropout/mul_1" + input: "bert/encoder/layer_2/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_2/attention/self/MatMul_1" + input: "bert/encoder/layer_2/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_2/attention/self/transpose_3" + input: "bert/encoder/layer_2/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + input: "bert/encoder/layer_2/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias" + input: "bert/encoder/layer_2/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/attention/self/Reshape_3" + input: "bert/encoder/layer_2/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/attention/output/dense/MatMul" + input: "bert/encoder/layer_2/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_2/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/dropout/sub/x" + input: "bert/encoder/layer_2/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_2/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_2/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_2/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_2/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_2/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_2/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/dropout/mul" + input: "bert/encoder/layer_2/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/add" + op: "Add" + input: "bert/encoder/layer_2/attention/output/dropout/mul_1" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_2/attention/output/add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_2/attention/output/add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + input: "bert/encoder/layer_2/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias" + input: "bert/encoder/layer_2/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_2/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/intermediate/dense/MatMul" + input: "bert/encoder/layer_2/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_2/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul/x" + input: "bert/encoder/layer_2/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_2/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_2/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_2/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_2/intermediate/dense/add_1/x" + input: "bert/encoder/layer_2/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_2/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_2/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel" + input: "bert/encoder/layer_2/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias" + input: "bert/encoder/layer_2/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/intermediate/dense/mul_3" + input: "bert/encoder/layer_2/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_2/output/dense/MatMul" + input: "bert/encoder/layer_2/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_2/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_2/output/dropout/random_uniform/max" + input: "bert/encoder/layer_2/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_2/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_2/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_2/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_2/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_2/output/dropout/sub/x" + input: "bert/encoder/layer_2/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_2/output/dropout/truediv/x" + input: "bert/encoder/layer_2/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_2/output/dropout/random_uniform" + input: "bert/encoder/layer_2/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_2/output/dense/BiasAdd" + input: "bert/encoder/layer_2/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_2/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/output/dropout/mul" + input: "bert/encoder/layer_2/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/add" + op: "Add" + input: "bert/encoder/layer_2/output/dropout/mul_1" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + input: "bert/encoder/layer_2/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_2/output/add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_2/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_2/output/add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_2/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_2/output/add" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_2/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_2/output/LayerNorm/beta/read" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel" + input: "bert/encoder/layer_3/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias" + input: "bert/encoder/layer_3/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_3/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/attention/self/query/MatMul" + input: "bert/encoder/layer_3/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel" + input: "bert/encoder/layer_3/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias" + input: "bert/encoder/layer_3/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_3/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/attention/self/key/MatMul" + input: "bert/encoder/layer_3/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel" + input: "bert/encoder/layer_3/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias" + input: "bert/encoder/layer_3/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_3/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/attention/self/value/MatMul" + input: "bert/encoder/layer_3/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_3/attention/self/query/BiasAdd" + input: "bert/encoder/layer_3/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_3/attention/self/Reshape" + input: "bert/encoder/layer_3/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_3/attention/self/key/BiasAdd" + input: "bert/encoder/layer_3/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_3/attention/self/Reshape_1" + input: "bert/encoder/layer_3/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_3/attention/self/transpose" + input: "bert/encoder/layer_3/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/MatMul" + input: "bert/encoder/layer_3/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_3/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/sub/x" + input: "bert/encoder/layer_3/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/sub" + input: "bert/encoder/layer_3/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/add" + op: "Add" + input: "bert/encoder/layer_3/attention/self/Mul" + input: "bert/encoder/layer_3/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_3/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_3/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/dropout/sub/x" + input: "bert/encoder/layer_3/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_3/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_3/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_3/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_3/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/Softmax" + input: "bert/encoder/layer_3/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_3/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/dropout/mul" + input: "bert/encoder/layer_3/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_3/attention/self/value/BiasAdd" + input: "bert/encoder/layer_3/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_3/attention/self/Reshape_2" + input: "bert/encoder/layer_3/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_3/attention/self/dropout/mul_1" + input: "bert/encoder/layer_3/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_3/attention/self/MatMul_1" + input: "bert/encoder/layer_3/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_3/attention/self/transpose_3" + input: "bert/encoder/layer_3/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + input: "bert/encoder/layer_3/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias" + input: "bert/encoder/layer_3/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/attention/self/Reshape_3" + input: "bert/encoder/layer_3/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/attention/output/dense/MatMul" + input: "bert/encoder/layer_3/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_3/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/dropout/sub/x" + input: "bert/encoder/layer_3/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_3/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_3/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_3/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_3/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_3/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_3/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/dropout/mul" + input: "bert/encoder/layer_3/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/add" + op: "Add" + input: "bert/encoder/layer_3/attention/output/dropout/mul_1" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_3/attention/output/add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_3/attention/output/add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + input: "bert/encoder/layer_3/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias" + input: "bert/encoder/layer_3/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_3/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/intermediate/dense/MatMul" + input: "bert/encoder/layer_3/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_3/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul/x" + input: "bert/encoder/layer_3/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_3/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_3/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_3/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_3/intermediate/dense/add_1/x" + input: "bert/encoder/layer_3/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_3/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_3/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel" + input: "bert/encoder/layer_3/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias" + input: "bert/encoder/layer_3/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/intermediate/dense/mul_3" + input: "bert/encoder/layer_3/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_3/output/dense/MatMul" + input: "bert/encoder/layer_3/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_3/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_3/output/dropout/random_uniform/max" + input: "bert/encoder/layer_3/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_3/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_3/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_3/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_3/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_3/output/dropout/sub/x" + input: "bert/encoder/layer_3/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_3/output/dropout/truediv/x" + input: "bert/encoder/layer_3/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_3/output/dropout/random_uniform" + input: "bert/encoder/layer_3/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_3/output/dense/BiasAdd" + input: "bert/encoder/layer_3/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_3/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/output/dropout/mul" + input: "bert/encoder/layer_3/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/add" + op: "Add" + input: "bert/encoder/layer_3/output/dropout/mul_1" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + input: "bert/encoder/layer_3/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_3/output/add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_3/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_3/output/add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_3/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_3/output/add" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_3/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_3/output/LayerNorm/beta/read" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel" + input: "bert/encoder/layer_4/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias" + input: "bert/encoder/layer_4/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_4/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/attention/self/query/MatMul" + input: "bert/encoder/layer_4/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel" + input: "bert/encoder/layer_4/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias" + input: "bert/encoder/layer_4/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_4/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/attention/self/key/MatMul" + input: "bert/encoder/layer_4/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel" + input: "bert/encoder/layer_4/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias" + input: "bert/encoder/layer_4/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_4/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/attention/self/value/MatMul" + input: "bert/encoder/layer_4/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_4/attention/self/query/BiasAdd" + input: "bert/encoder/layer_4/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_4/attention/self/Reshape" + input: "bert/encoder/layer_4/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_4/attention/self/key/BiasAdd" + input: "bert/encoder/layer_4/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_4/attention/self/Reshape_1" + input: "bert/encoder/layer_4/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_4/attention/self/transpose" + input: "bert/encoder/layer_4/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/MatMul" + input: "bert/encoder/layer_4/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_4/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/sub/x" + input: "bert/encoder/layer_4/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/sub" + input: "bert/encoder/layer_4/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/add" + op: "Add" + input: "bert/encoder/layer_4/attention/self/Mul" + input: "bert/encoder/layer_4/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_4/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_4/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/dropout/sub/x" + input: "bert/encoder/layer_4/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_4/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_4/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_4/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_4/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/Softmax" + input: "bert/encoder/layer_4/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_4/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/dropout/mul" + input: "bert/encoder/layer_4/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_4/attention/self/value/BiasAdd" + input: "bert/encoder/layer_4/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_4/attention/self/Reshape_2" + input: "bert/encoder/layer_4/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_4/attention/self/dropout/mul_1" + input: "bert/encoder/layer_4/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_4/attention/self/MatMul_1" + input: "bert/encoder/layer_4/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_4/attention/self/transpose_3" + input: "bert/encoder/layer_4/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + input: "bert/encoder/layer_4/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias" + input: "bert/encoder/layer_4/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/attention/self/Reshape_3" + input: "bert/encoder/layer_4/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/attention/output/dense/MatMul" + input: "bert/encoder/layer_4/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_4/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/dropout/sub/x" + input: "bert/encoder/layer_4/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_4/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_4/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_4/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_4/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_4/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_4/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/dropout/mul" + input: "bert/encoder/layer_4/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/add" + op: "Add" + input: "bert/encoder/layer_4/attention/output/dropout/mul_1" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_4/attention/output/add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_4/attention/output/add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + input: "bert/encoder/layer_4/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias" + input: "bert/encoder/layer_4/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_4/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/intermediate/dense/MatMul" + input: "bert/encoder/layer_4/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_4/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul/x" + input: "bert/encoder/layer_4/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_4/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_4/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_4/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_4/intermediate/dense/add_1/x" + input: "bert/encoder/layer_4/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_4/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_4/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel" + input: "bert/encoder/layer_4/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias" + input: "bert/encoder/layer_4/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/intermediate/dense/mul_3" + input: "bert/encoder/layer_4/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_4/output/dense/MatMul" + input: "bert/encoder/layer_4/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_4/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_4/output/dropout/random_uniform/max" + input: "bert/encoder/layer_4/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_4/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_4/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_4/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_4/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_4/output/dropout/sub/x" + input: "bert/encoder/layer_4/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_4/output/dropout/truediv/x" + input: "bert/encoder/layer_4/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_4/output/dropout/random_uniform" + input: "bert/encoder/layer_4/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_4/output/dense/BiasAdd" + input: "bert/encoder/layer_4/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_4/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/output/dropout/mul" + input: "bert/encoder/layer_4/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/add" + op: "Add" + input: "bert/encoder/layer_4/output/dropout/mul_1" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + input: "bert/encoder/layer_4/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_4/output/add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_4/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_4/output/add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_4/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_4/output/add" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_4/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_4/output/LayerNorm/beta/read" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel" + input: "bert/encoder/layer_5/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias" + input: "bert/encoder/layer_5/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_5/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/attention/self/query/MatMul" + input: "bert/encoder/layer_5/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel" + input: "bert/encoder/layer_5/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias" + input: "bert/encoder/layer_5/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_5/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/attention/self/key/MatMul" + input: "bert/encoder/layer_5/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel" + input: "bert/encoder/layer_5/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias" + input: "bert/encoder/layer_5/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_5/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/attention/self/value/MatMul" + input: "bert/encoder/layer_5/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_5/attention/self/query/BiasAdd" + input: "bert/encoder/layer_5/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_5/attention/self/Reshape" + input: "bert/encoder/layer_5/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_5/attention/self/key/BiasAdd" + input: "bert/encoder/layer_5/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_5/attention/self/Reshape_1" + input: "bert/encoder/layer_5/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_5/attention/self/transpose" + input: "bert/encoder/layer_5/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/MatMul" + input: "bert/encoder/layer_5/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_5/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/sub/x" + input: "bert/encoder/layer_5/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/sub" + input: "bert/encoder/layer_5/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/add" + op: "Add" + input: "bert/encoder/layer_5/attention/self/Mul" + input: "bert/encoder/layer_5/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_5/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_5/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/dropout/sub/x" + input: "bert/encoder/layer_5/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_5/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_5/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_5/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_5/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/Softmax" + input: "bert/encoder/layer_5/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_5/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/dropout/mul" + input: "bert/encoder/layer_5/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_5/attention/self/value/BiasAdd" + input: "bert/encoder/layer_5/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_5/attention/self/Reshape_2" + input: "bert/encoder/layer_5/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_5/attention/self/dropout/mul_1" + input: "bert/encoder/layer_5/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_5/attention/self/MatMul_1" + input: "bert/encoder/layer_5/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_5/attention/self/transpose_3" + input: "bert/encoder/layer_5/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + input: "bert/encoder/layer_5/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias" + input: "bert/encoder/layer_5/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/attention/self/Reshape_3" + input: "bert/encoder/layer_5/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/attention/output/dense/MatMul" + input: "bert/encoder/layer_5/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_5/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/dropout/sub/x" + input: "bert/encoder/layer_5/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_5/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_5/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_5/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_5/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_5/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_5/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/dropout/mul" + input: "bert/encoder/layer_5/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/add" + op: "Add" + input: "bert/encoder/layer_5/attention/output/dropout/mul_1" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_5/attention/output/add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_5/attention/output/add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + input: "bert/encoder/layer_5/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias" + input: "bert/encoder/layer_5/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_5/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/intermediate/dense/MatMul" + input: "bert/encoder/layer_5/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_5/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul/x" + input: "bert/encoder/layer_5/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_5/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_5/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_5/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_5/intermediate/dense/add_1/x" + input: "bert/encoder/layer_5/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_5/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_5/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel" + input: "bert/encoder/layer_5/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias" + input: "bert/encoder/layer_5/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/intermediate/dense/mul_3" + input: "bert/encoder/layer_5/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_5/output/dense/MatMul" + input: "bert/encoder/layer_5/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_5/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_5/output/dropout/random_uniform/max" + input: "bert/encoder/layer_5/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_5/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_5/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_5/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_5/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_5/output/dropout/sub/x" + input: "bert/encoder/layer_5/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_5/output/dropout/truediv/x" + input: "bert/encoder/layer_5/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_5/output/dropout/random_uniform" + input: "bert/encoder/layer_5/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_5/output/dense/BiasAdd" + input: "bert/encoder/layer_5/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_5/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/output/dropout/mul" + input: "bert/encoder/layer_5/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/add" + op: "Add" + input: "bert/encoder/layer_5/output/dropout/mul_1" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + input: "bert/encoder/layer_5/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_5/output/add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_5/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_5/output/add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_5/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_5/output/add" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_5/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_5/output/LayerNorm/beta/read" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel" + input: "bert/encoder/layer_6/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias" + input: "bert/encoder/layer_6/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_6/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/attention/self/query/MatMul" + input: "bert/encoder/layer_6/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel" + input: "bert/encoder/layer_6/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias" + input: "bert/encoder/layer_6/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_6/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/attention/self/key/MatMul" + input: "bert/encoder/layer_6/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel" + input: "bert/encoder/layer_6/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias" + input: "bert/encoder/layer_6/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_6/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/attention/self/value/MatMul" + input: "bert/encoder/layer_6/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_6/attention/self/query/BiasAdd" + input: "bert/encoder/layer_6/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_6/attention/self/Reshape" + input: "bert/encoder/layer_6/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_6/attention/self/key/BiasAdd" + input: "bert/encoder/layer_6/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_6/attention/self/Reshape_1" + input: "bert/encoder/layer_6/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_6/attention/self/transpose" + input: "bert/encoder/layer_6/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/MatMul" + input: "bert/encoder/layer_6/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_6/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/sub/x" + input: "bert/encoder/layer_6/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/sub" + input: "bert/encoder/layer_6/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/add" + op: "Add" + input: "bert/encoder/layer_6/attention/self/Mul" + input: "bert/encoder/layer_6/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_6/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_6/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/dropout/sub/x" + input: "bert/encoder/layer_6/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_6/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_6/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_6/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_6/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/Softmax" + input: "bert/encoder/layer_6/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_6/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/dropout/mul" + input: "bert/encoder/layer_6/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_6/attention/self/value/BiasAdd" + input: "bert/encoder/layer_6/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_6/attention/self/Reshape_2" + input: "bert/encoder/layer_6/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_6/attention/self/dropout/mul_1" + input: "bert/encoder/layer_6/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_6/attention/self/MatMul_1" + input: "bert/encoder/layer_6/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_6/attention/self/transpose_3" + input: "bert/encoder/layer_6/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + input: "bert/encoder/layer_6/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias" + input: "bert/encoder/layer_6/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/attention/self/Reshape_3" + input: "bert/encoder/layer_6/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/attention/output/dense/MatMul" + input: "bert/encoder/layer_6/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_6/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/dropout/sub/x" + input: "bert/encoder/layer_6/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_6/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_6/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_6/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_6/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_6/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_6/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/dropout/mul" + input: "bert/encoder/layer_6/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/add" + op: "Add" + input: "bert/encoder/layer_6/attention/output/dropout/mul_1" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_6/attention/output/add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_6/attention/output/add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + input: "bert/encoder/layer_6/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias" + input: "bert/encoder/layer_6/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_6/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/intermediate/dense/MatMul" + input: "bert/encoder/layer_6/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_6/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul/x" + input: "bert/encoder/layer_6/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_6/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_6/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_6/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_6/intermediate/dense/add_1/x" + input: "bert/encoder/layer_6/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_6/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_6/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel" + input: "bert/encoder/layer_6/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias" + input: "bert/encoder/layer_6/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/intermediate/dense/mul_3" + input: "bert/encoder/layer_6/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_6/output/dense/MatMul" + input: "bert/encoder/layer_6/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_6/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_6/output/dropout/random_uniform/max" + input: "bert/encoder/layer_6/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_6/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_6/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_6/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_6/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_6/output/dropout/sub/x" + input: "bert/encoder/layer_6/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_6/output/dropout/truediv/x" + input: "bert/encoder/layer_6/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_6/output/dropout/random_uniform" + input: "bert/encoder/layer_6/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_6/output/dense/BiasAdd" + input: "bert/encoder/layer_6/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_6/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/output/dropout/mul" + input: "bert/encoder/layer_6/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/add" + op: "Add" + input: "bert/encoder/layer_6/output/dropout/mul_1" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + input: "bert/encoder/layer_6/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_6/output/add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_6/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_6/output/add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_6/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_6/output/add" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_6/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_6/output/LayerNorm/beta/read" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel" + input: "bert/encoder/layer_7/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias" + input: "bert/encoder/layer_7/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_7/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/attention/self/query/MatMul" + input: "bert/encoder/layer_7/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel" + input: "bert/encoder/layer_7/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias" + input: "bert/encoder/layer_7/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_7/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/attention/self/key/MatMul" + input: "bert/encoder/layer_7/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel" + input: "bert/encoder/layer_7/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias" + input: "bert/encoder/layer_7/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_7/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/attention/self/value/MatMul" + input: "bert/encoder/layer_7/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_7/attention/self/query/BiasAdd" + input: "bert/encoder/layer_7/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_7/attention/self/Reshape" + input: "bert/encoder/layer_7/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_7/attention/self/key/BiasAdd" + input: "bert/encoder/layer_7/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_7/attention/self/Reshape_1" + input: "bert/encoder/layer_7/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_7/attention/self/transpose" + input: "bert/encoder/layer_7/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/MatMul" + input: "bert/encoder/layer_7/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_7/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/sub/x" + input: "bert/encoder/layer_7/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/sub" + input: "bert/encoder/layer_7/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/add" + op: "Add" + input: "bert/encoder/layer_7/attention/self/Mul" + input: "bert/encoder/layer_7/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_7/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_7/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/dropout/sub/x" + input: "bert/encoder/layer_7/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_7/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_7/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_7/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_7/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/Softmax" + input: "bert/encoder/layer_7/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_7/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/dropout/mul" + input: "bert/encoder/layer_7/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_7/attention/self/value/BiasAdd" + input: "bert/encoder/layer_7/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_7/attention/self/Reshape_2" + input: "bert/encoder/layer_7/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_7/attention/self/dropout/mul_1" + input: "bert/encoder/layer_7/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_7/attention/self/MatMul_1" + input: "bert/encoder/layer_7/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_7/attention/self/transpose_3" + input: "bert/encoder/layer_7/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + input: "bert/encoder/layer_7/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias" + input: "bert/encoder/layer_7/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/attention/self/Reshape_3" + input: "bert/encoder/layer_7/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/attention/output/dense/MatMul" + input: "bert/encoder/layer_7/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_7/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/dropout/sub/x" + input: "bert/encoder/layer_7/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_7/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_7/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_7/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_7/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_7/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_7/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/dropout/mul" + input: "bert/encoder/layer_7/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/add" + op: "Add" + input: "bert/encoder/layer_7/attention/output/dropout/mul_1" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_7/attention/output/add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_7/attention/output/add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + input: "bert/encoder/layer_7/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias" + input: "bert/encoder/layer_7/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_7/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/intermediate/dense/MatMul" + input: "bert/encoder/layer_7/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_7/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul/x" + input: "bert/encoder/layer_7/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_7/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_7/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_7/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_7/intermediate/dense/add_1/x" + input: "bert/encoder/layer_7/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_7/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_7/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel" + input: "bert/encoder/layer_7/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias" + input: "bert/encoder/layer_7/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/intermediate/dense/mul_3" + input: "bert/encoder/layer_7/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_7/output/dense/MatMul" + input: "bert/encoder/layer_7/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_7/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_7/output/dropout/random_uniform/max" + input: "bert/encoder/layer_7/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_7/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_7/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_7/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_7/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_7/output/dropout/sub/x" + input: "bert/encoder/layer_7/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_7/output/dropout/truediv/x" + input: "bert/encoder/layer_7/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_7/output/dropout/random_uniform" + input: "bert/encoder/layer_7/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_7/output/dense/BiasAdd" + input: "bert/encoder/layer_7/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_7/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/output/dropout/mul" + input: "bert/encoder/layer_7/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/add" + op: "Add" + input: "bert/encoder/layer_7/output/dropout/mul_1" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + input: "bert/encoder/layer_7/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_7/output/add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_7/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_7/output/add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_7/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_7/output/add" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_7/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_7/output/LayerNorm/beta/read" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel" + input: "bert/encoder/layer_8/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias" + input: "bert/encoder/layer_8/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_8/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/attention/self/query/MatMul" + input: "bert/encoder/layer_8/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel" + input: "bert/encoder/layer_8/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias" + input: "bert/encoder/layer_8/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_8/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/attention/self/key/MatMul" + input: "bert/encoder/layer_8/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel" + input: "bert/encoder/layer_8/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias" + input: "bert/encoder/layer_8/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_8/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/attention/self/value/MatMul" + input: "bert/encoder/layer_8/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_8/attention/self/query/BiasAdd" + input: "bert/encoder/layer_8/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_8/attention/self/Reshape" + input: "bert/encoder/layer_8/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_8/attention/self/key/BiasAdd" + input: "bert/encoder/layer_8/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_8/attention/self/Reshape_1" + input: "bert/encoder/layer_8/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_8/attention/self/transpose" + input: "bert/encoder/layer_8/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/MatMul" + input: "bert/encoder/layer_8/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_8/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/sub/x" + input: "bert/encoder/layer_8/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/sub" + input: "bert/encoder/layer_8/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/add" + op: "Add" + input: "bert/encoder/layer_8/attention/self/Mul" + input: "bert/encoder/layer_8/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_8/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_8/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/dropout/sub/x" + input: "bert/encoder/layer_8/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_8/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_8/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_8/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_8/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/Softmax" + input: "bert/encoder/layer_8/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_8/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/dropout/mul" + input: "bert/encoder/layer_8/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_8/attention/self/value/BiasAdd" + input: "bert/encoder/layer_8/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_8/attention/self/Reshape_2" + input: "bert/encoder/layer_8/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_8/attention/self/dropout/mul_1" + input: "bert/encoder/layer_8/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_8/attention/self/MatMul_1" + input: "bert/encoder/layer_8/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_8/attention/self/transpose_3" + input: "bert/encoder/layer_8/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + input: "bert/encoder/layer_8/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias" + input: "bert/encoder/layer_8/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/attention/self/Reshape_3" + input: "bert/encoder/layer_8/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/attention/output/dense/MatMul" + input: "bert/encoder/layer_8/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_8/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/dropout/sub/x" + input: "bert/encoder/layer_8/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_8/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_8/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_8/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_8/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_8/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_8/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/dropout/mul" + input: "bert/encoder/layer_8/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/add" + op: "Add" + input: "bert/encoder/layer_8/attention/output/dropout/mul_1" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_8/attention/output/add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_8/attention/output/add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + input: "bert/encoder/layer_8/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias" + input: "bert/encoder/layer_8/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_8/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/intermediate/dense/MatMul" + input: "bert/encoder/layer_8/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_8/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul/x" + input: "bert/encoder/layer_8/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_8/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_8/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_8/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_8/intermediate/dense/add_1/x" + input: "bert/encoder/layer_8/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_8/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_8/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel" + input: "bert/encoder/layer_8/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias" + input: "bert/encoder/layer_8/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/intermediate/dense/mul_3" + input: "bert/encoder/layer_8/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_8/output/dense/MatMul" + input: "bert/encoder/layer_8/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_8/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_8/output/dropout/random_uniform/max" + input: "bert/encoder/layer_8/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_8/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_8/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_8/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_8/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_8/output/dropout/sub/x" + input: "bert/encoder/layer_8/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_8/output/dropout/truediv/x" + input: "bert/encoder/layer_8/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_8/output/dropout/random_uniform" + input: "bert/encoder/layer_8/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_8/output/dense/BiasAdd" + input: "bert/encoder/layer_8/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_8/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/output/dropout/mul" + input: "bert/encoder/layer_8/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/add" + op: "Add" + input: "bert/encoder/layer_8/output/dropout/mul_1" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + input: "bert/encoder/layer_8/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_8/output/add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_8/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_8/output/add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_8/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_8/output/add" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_8/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_8/output/LayerNorm/beta/read" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel" + input: "bert/encoder/layer_9/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias" + input: "bert/encoder/layer_9/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_9/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/attention/self/query/MatMul" + input: "bert/encoder/layer_9/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel" + input: "bert/encoder/layer_9/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias" + input: "bert/encoder/layer_9/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_9/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/attention/self/key/MatMul" + input: "bert/encoder/layer_9/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel" + input: "bert/encoder/layer_9/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias" + input: "bert/encoder/layer_9/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_9/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/attention/self/value/MatMul" + input: "bert/encoder/layer_9/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_9/attention/self/query/BiasAdd" + input: "bert/encoder/layer_9/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_9/attention/self/Reshape" + input: "bert/encoder/layer_9/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_9/attention/self/key/BiasAdd" + input: "bert/encoder/layer_9/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_9/attention/self/Reshape_1" + input: "bert/encoder/layer_9/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_9/attention/self/transpose" + input: "bert/encoder/layer_9/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/MatMul" + input: "bert/encoder/layer_9/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_9/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/sub/x" + input: "bert/encoder/layer_9/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/sub" + input: "bert/encoder/layer_9/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/add" + op: "Add" + input: "bert/encoder/layer_9/attention/self/Mul" + input: "bert/encoder/layer_9/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_9/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_9/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/dropout/sub/x" + input: "bert/encoder/layer_9/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_9/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_9/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_9/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_9/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/Softmax" + input: "bert/encoder/layer_9/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_9/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/dropout/mul" + input: "bert/encoder/layer_9/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_9/attention/self/value/BiasAdd" + input: "bert/encoder/layer_9/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_9/attention/self/Reshape_2" + input: "bert/encoder/layer_9/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_9/attention/self/dropout/mul_1" + input: "bert/encoder/layer_9/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_9/attention/self/MatMul_1" + input: "bert/encoder/layer_9/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_9/attention/self/transpose_3" + input: "bert/encoder/layer_9/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + input: "bert/encoder/layer_9/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias" + input: "bert/encoder/layer_9/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/attention/self/Reshape_3" + input: "bert/encoder/layer_9/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/attention/output/dense/MatMul" + input: "bert/encoder/layer_9/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_9/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/dropout/sub/x" + input: "bert/encoder/layer_9/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_9/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_9/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_9/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_9/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_9/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_9/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/dropout/mul" + input: "bert/encoder/layer_9/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/add" + op: "Add" + input: "bert/encoder/layer_9/attention/output/dropout/mul_1" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_9/attention/output/add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_9/attention/output/add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + input: "bert/encoder/layer_9/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias" + input: "bert/encoder/layer_9/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_9/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/intermediate/dense/MatMul" + input: "bert/encoder/layer_9/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_9/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul/x" + input: "bert/encoder/layer_9/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_9/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_9/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_9/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_9/intermediate/dense/add_1/x" + input: "bert/encoder/layer_9/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_9/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_9/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel" + input: "bert/encoder/layer_9/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias" + input: "bert/encoder/layer_9/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/intermediate/dense/mul_3" + input: "bert/encoder/layer_9/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_9/output/dense/MatMul" + input: "bert/encoder/layer_9/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_9/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_9/output/dropout/random_uniform/max" + input: "bert/encoder/layer_9/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_9/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_9/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_9/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_9/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_9/output/dropout/sub/x" + input: "bert/encoder/layer_9/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_9/output/dropout/truediv/x" + input: "bert/encoder/layer_9/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_9/output/dropout/random_uniform" + input: "bert/encoder/layer_9/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_9/output/dense/BiasAdd" + input: "bert/encoder/layer_9/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_9/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/output/dropout/mul" + input: "bert/encoder/layer_9/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/add" + op: "Add" + input: "bert/encoder/layer_9/output/dropout/mul_1" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + input: "bert/encoder/layer_9/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_9/output/add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_9/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_9/output/add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_9/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_9/output/add" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_9/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_9/output/LayerNorm/beta/read" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel" + input: "bert/encoder/layer_10/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias" + input: "bert/encoder/layer_10/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_10/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/attention/self/query/MatMul" + input: "bert/encoder/layer_10/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel" + input: "bert/encoder/layer_10/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias" + input: "bert/encoder/layer_10/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_10/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/attention/self/key/MatMul" + input: "bert/encoder/layer_10/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel" + input: "bert/encoder/layer_10/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias" + input: "bert/encoder/layer_10/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_10/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/attention/self/value/MatMul" + input: "bert/encoder/layer_10/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_10/attention/self/query/BiasAdd" + input: "bert/encoder/layer_10/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_10/attention/self/Reshape" + input: "bert/encoder/layer_10/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_10/attention/self/key/BiasAdd" + input: "bert/encoder/layer_10/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_10/attention/self/Reshape_1" + input: "bert/encoder/layer_10/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_10/attention/self/transpose" + input: "bert/encoder/layer_10/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/MatMul" + input: "bert/encoder/layer_10/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_10/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/sub/x" + input: "bert/encoder/layer_10/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/sub" + input: "bert/encoder/layer_10/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/add" + op: "Add" + input: "bert/encoder/layer_10/attention/self/Mul" + input: "bert/encoder/layer_10/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_10/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_10/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/dropout/sub/x" + input: "bert/encoder/layer_10/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_10/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_10/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_10/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_10/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/Softmax" + input: "bert/encoder/layer_10/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_10/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/dropout/mul" + input: "bert/encoder/layer_10/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_10/attention/self/value/BiasAdd" + input: "bert/encoder/layer_10/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_10/attention/self/Reshape_2" + input: "bert/encoder/layer_10/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_10/attention/self/dropout/mul_1" + input: "bert/encoder/layer_10/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_10/attention/self/MatMul_1" + input: "bert/encoder/layer_10/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_10/attention/self/transpose_3" + input: "bert/encoder/layer_10/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + input: "bert/encoder/layer_10/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias" + input: "bert/encoder/layer_10/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/attention/self/Reshape_3" + input: "bert/encoder/layer_10/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/attention/output/dense/MatMul" + input: "bert/encoder/layer_10/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_10/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/dropout/sub/x" + input: "bert/encoder/layer_10/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_10/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_10/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_10/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_10/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_10/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_10/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/dropout/mul" + input: "bert/encoder/layer_10/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/add" + op: "Add" + input: "bert/encoder/layer_10/attention/output/dropout/mul_1" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_10/attention/output/add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_10/attention/output/add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + input: "bert/encoder/layer_10/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias" + input: "bert/encoder/layer_10/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_10/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/intermediate/dense/MatMul" + input: "bert/encoder/layer_10/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_10/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul/x" + input: "bert/encoder/layer_10/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_10/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_10/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_10/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_10/intermediate/dense/add_1/x" + input: "bert/encoder/layer_10/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_10/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_10/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel" + input: "bert/encoder/layer_10/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias" + input: "bert/encoder/layer_10/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/intermediate/dense/mul_3" + input: "bert/encoder/layer_10/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_10/output/dense/MatMul" + input: "bert/encoder/layer_10/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_10/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_10/output/dropout/random_uniform/max" + input: "bert/encoder/layer_10/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_10/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_10/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_10/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_10/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_10/output/dropout/sub/x" + input: "bert/encoder/layer_10/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_10/output/dropout/truediv/x" + input: "bert/encoder/layer_10/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_10/output/dropout/random_uniform" + input: "bert/encoder/layer_10/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_10/output/dense/BiasAdd" + input: "bert/encoder/layer_10/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_10/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/output/dropout/mul" + input: "bert/encoder/layer_10/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/add" + op: "Add" + input: "bert/encoder/layer_10/output/dropout/mul_1" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + input: "bert/encoder/layer_10/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_10/output/add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_10/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_10/output/add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_10/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_10/output/add" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_10/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_10/output/LayerNorm/beta/read" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel" + input: "bert/encoder/layer_11/attention/self/query/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias" + input: "bert/encoder/layer_11/attention/self/query/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_11/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/attention/self/query/MatMul" + input: "bert/encoder/layer_11/attention/self/query/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel" + input: "bert/encoder/layer_11/attention/self/key/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias" + input: "bert/encoder/layer_11/attention/self/key/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_11/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/attention/self/key/MatMul" + input: "bert/encoder/layer_11/attention/self/key/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel" + input: "bert/encoder/layer_11/attention/self/value/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias" + input: "bert/encoder/layer_11/attention/self/value/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/MatMul" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_11/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/attention/self/value/MatMul" + input: "bert/encoder/layer_11/attention/self/value/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape" + op: "Reshape" + input: "bert/encoder/layer_11/attention/self/query/BiasAdd" + input: "bert/encoder/layer_11/attention/self/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose" + op: "Transpose" + input: "bert/encoder/layer_11/attention/self/Reshape" + input: "bert/encoder/layer_11/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_1/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_1" + op: "Reshape" + input: "bert/encoder/layer_11/attention/self/key/BiasAdd" + input: "bert/encoder/layer_11/attention/self/Reshape_1/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_1/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_1" + op: "Transpose" + input: "bert/encoder/layer_11/attention/self/Reshape_1" + input: "bert/encoder/layer_11/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/MatMul" + op: "BatchMatMulV2" + input: "bert/encoder/layer_11/attention/self/transpose" + input: "bert/encoder/layer_11/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Mul/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.125 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/MatMul" + input: "bert/encoder/layer_11/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/ExpandDims" + op: "ExpandDims" + input: "bert/encoder/mul" + input: "bert/encoder/layer_11/attention/self/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/sub/x" + input: "bert/encoder/layer_11/attention/self/ExpandDims" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/mul_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: -10000.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/sub" + input: "bert/encoder/layer_11/attention/self/mul_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/add" + op: "Add" + input: "bert/encoder/layer_11/attention/self/Mul" + input: "bert/encoder/layer_11/attention/self/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Softmax" + op: "Softmax" + input: "bert/encoder/layer_11/attention/self/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_11/attention/self/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/max" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/mul" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/dropout/sub/x" + input: "bert/encoder/layer_11/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_11/attention/self/dropout/truediv/x" + input: "bert/encoder/layer_11/attention/self/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_11/attention/self/dropout/random_uniform" + input: "bert/encoder/layer_11/attention/self/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/Softmax" + input: "bert/encoder/layer_11/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_11/attention/self/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/dropout/mul" + input: "bert/encoder/layer_11/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_11/attention/self/value/BiasAdd" + input: "bert/encoder/layer_11/attention/self/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_2/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_2" + op: "Transpose" + input: "bert/encoder/layer_11/attention/self/Reshape_2" + input: "bert/encoder/layer_11/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_11/attention/self/dropout/mul_1" + input: "bert/encoder/layer_11/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_3/perm" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: "\000\000\000\000\002\000\000\000\001\000\000\000\003\000\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/transpose_3" + op: "Transpose" + input: "bert/encoder/layer_11/attention/self/MatMul_1" + input: "bert/encoder/layer_11/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_11/attention/self/transpose_3" + input: "bert/encoder/layer_11/attention/self/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + input: "bert/encoder/layer_11/attention/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias" + input: "bert/encoder/layer_11/attention/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_11/attention/self/Reshape_3" + input: "bert/encoder/layer_11/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/attention/output/dense/MatMul" + input: "bert/encoder/layer_11/attention/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_11/attention/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/max" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/dropout/sub/x" + input: "bert/encoder/layer_11/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_11/attention/output/dropout/truediv/x" + input: "bert/encoder/layer_11/attention/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_11/attention/output/dropout/random_uniform" + input: "bert/encoder/layer_11/attention/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/dense/BiasAdd" + input: "bert/encoder/layer_11/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_11/attention/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/dropout/mul" + input: "bert/encoder/layer_11/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/add" + op: "Add" + input: "bert/encoder/layer_11/attention/output/dropout/mul_1" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_11/attention/output/add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_11/attention/output/add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/read" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + input: "bert/encoder/layer_11/intermediate/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias" + input: "bert/encoder/layer_11/intermediate/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/layer_11/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/intermediate/dense/MatMul" + input: "bert/encoder/layer_11/intermediate/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/Pow/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 3.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/Pow" + op: "Pow" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_11/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.044714998453855515 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul/x" + input: "bert/encoder/layer_11/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/add" + op: "Add" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_11/intermediate/dense/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.7978845834732056 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul_1/x" + input: "bert/encoder/layer_11/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/Tanh" + op: "Tanh" + input: "bert/encoder/layer_11/intermediate/dense/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/add_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/add_1" + op: "Add" + input: "bert/encoder/layer_11/intermediate/dense/add_1/x" + input: "bert/encoder/layer_11/intermediate/dense/Tanh" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul_2/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.5 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul_2" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul_2/x" + input: "bert/encoder/layer_11/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/mul_3" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "bert/encoder/layer_11/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel" + input: "bert/encoder/layer_11/output/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias" + input: "bert/encoder/layer_11/output/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/MatMul" + op: "MatMul" + input: "bert/encoder/layer_11/intermediate/dense/mul_3" + input: "bert/encoder/layer_11/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/BiasAdd" + op: "BiasAdd" + input: "bert/encoder/layer_11/output/dense/MatMul" + input: "bert/encoder/layer_11/output/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "bert/encoder/layer_11/output/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform/sub" + op: "Sub" + input: "bert/encoder/layer_11/output/dropout/random_uniform/max" + input: "bert/encoder/layer_11/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform/mul" + op: "Mul" + input: "bert/encoder/layer_11/output/dropout/random_uniform/RandomUniform" + input: "bert/encoder/layer_11/output/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/random_uniform" + op: "Add" + input: "bert/encoder/layer_11/output/dropout/random_uniform/mul" + input: "bert/encoder/layer_11/output/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/sub" + op: "Sub" + input: "bert/encoder/layer_11/output/dropout/sub/x" + input: "bert/encoder/layer_11/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/truediv" + op: "RealDiv" + input: "bert/encoder/layer_11/output/dropout/truediv/x" + input: "bert/encoder/layer_11/output/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/GreaterEqual" + op: "GreaterEqual" + input: "bert/encoder/layer_11/output/dropout/random_uniform" + input: "bert/encoder/layer_11/output/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/mul" + op: "Mul" + input: "bert/encoder/layer_11/output/dense/BiasAdd" + input: "bert/encoder/layer_11/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/Cast" + op: "Cast" + input: "bert/encoder/layer_11/output/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dropout/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/output/dropout/mul" + input: "bert/encoder/layer_11/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/add" + op: "Add" + input: "bert/encoder/layer_11/output/dropout/mul_1" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + input: "bert/encoder/layer_11/output/LayerNorm/beta/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/Initializer/ones" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 1.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/Initializer/ones" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/mean/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/mean" + op: "Mean" + input: "bert/encoder/layer_11/output/add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/mean/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/StopGradient" + op: "StopGradient" + input: "bert/encoder/layer_11/output/LayerNorm/moments/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference" + op: "SquaredDifference" + input: "bert/encoder/layer_11/output/add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/StopGradient" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/variance/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/moments/variance" + op: "Mean" + input: "bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference" + input: "bert/encoder/layer_11/output/LayerNorm/moments/variance/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999960041972e-13 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add" + op: "Add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/variance" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt" + op: "Rsqrt" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul" + op: "Mul" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1" + op: "Mul" + input: "bert/encoder/layer_11/output/add" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2" + op: "Mul" + input: "bert/encoder/layer_11/output/LayerNorm/moments/mean" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/sub" + op: "Sub" + input: "bert/encoder/layer_11/output/LayerNorm/beta/read" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add_1" + op: "Add" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_2/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_2" + op: "Reshape" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_2/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_3/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_3" + op: "Reshape" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_3/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_4/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_4" + op: "Reshape" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_4/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_5/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_5" + op: "Reshape" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_5/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_6/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_6" + op: "Reshape" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_6/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_7/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_7" + op: "Reshape" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_7/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_8/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_8" + op: "Reshape" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_8/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_9/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_9" + op: "Reshape" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_9/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_10/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_10" + op: "Reshape" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_10/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_11/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_11" + op: "Reshape" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_11/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_12/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_12" + op: "Reshape" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_12/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/Reshape_13/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/Reshape_13" + op: "Reshape" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/add_1" + input: "bert/encoder/Reshape_13/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: "\000\000\000\000\000\000\000\000\000\000\000\000" + } + } + } +} +node { + name: "bert/pooler/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: "\000\000\000\000\001\000\000\000\000\000\000\000" + } + } + } +} +node { + name: "bert/pooler/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: "\001\000\000\000\001\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "bert/pooler/strided_slice" + op: "StridedSlice" + input: "bert/encoder/Reshape_13" + input: "bert/pooler/strided_slice/stack" + input: "bert/pooler/strided_slice/stack_1" + input: "bert/pooler/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 5 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 5 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "bert/pooler/Squeeze" + op: "Squeeze" + input: "bert/pooler/strided_slice" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "squeeze_dims" + value { + list { + i: 1 + } + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal/mul" + op: "Mul" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal/TruncatedNormal" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/kernel/Initializer/truncated_normal" + op: "Add" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal/mul" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/kernel" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/kernel/Assign" + op: "Assign" + input: "bert/pooler/dense/kernel" + input: "bert/pooler/dense/kernel/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/kernel/read" + op: "Identity" + input: "bert/pooler/dense/kernel" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/bias/Assign" + op: "Assign" + input: "bert/pooler/dense/bias" + input: "bert/pooler/dense/bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/bias/read" + op: "Identity" + input: "bert/pooler/dense/bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/MatMul" + op: "MatMul" + input: "bert/pooler/Squeeze" + input: "bert/pooler/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "bert/pooler/dense/BiasAdd" + op: "BiasAdd" + input: "bert/pooler/dense/MatMul" + input: "bert/pooler/dense/bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "bert/pooler/dense/Tanh" + op: "Tanh" + input: "bert/pooler/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "output_weights/Initializer/truncated_normal/shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\003\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "output_weights/Initializer/truncated_normal/mean" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "output_weights/Initializer/truncated_normal/stddev" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.019999999552965164 + } + } + } +} +node { + name: "output_weights/Initializer/truncated_normal/TruncatedNormal" + op: "TruncatedNormal" + input: "output_weights/Initializer/truncated_normal/shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "output_weights/Initializer/truncated_normal/mul" + op: "Mul" + input: "output_weights/Initializer/truncated_normal/TruncatedNormal" + input: "output_weights/Initializer/truncated_normal/stddev" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "output_weights/Initializer/truncated_normal" + op: "Add" + input: "output_weights/Initializer/truncated_normal/mul" + input: "output_weights/Initializer/truncated_normal/mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "output_weights" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_weights/Assign" + op: "Assign" + input: "output_weights" + input: "output_weights/Initializer/truncated_normal" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_weights/read" + op: "Identity" + input: "output_weights" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "output_bias/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 3 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "output_bias" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_bias/Assign" + op: "Assign" + input: "output_bias" + input: "output_bias/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_bias/read" + op: "Identity" + input: "output_bias" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "loss/dropout/rate" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "loss/dropout/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: " \000\000\000\000\003\000\000" + } + } + } +} +node { + name: "loss/dropout/random_uniform/min" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "loss/dropout/random_uniform/max" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "loss/dropout/random_uniform/RandomUniform" + op: "RandomUniform" + input: "loss/dropout/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "seed" + value { + i: 0 + } + } + attr { + key: "seed2" + value { + i: 0 + } + } +} +node { + name: "loss/dropout/random_uniform/sub" + op: "Sub" + input: "loss/dropout/random_uniform/max" + input: "loss/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "loss/dropout/random_uniform/mul" + op: "Mul" + input: "loss/dropout/random_uniform/RandomUniform" + input: "loss/dropout/random_uniform/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/dropout/random_uniform" + op: "Add" + input: "loss/dropout/random_uniform/mul" + input: "loss/dropout/random_uniform/min" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/dropout/sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "loss/dropout/sub" + op: "Sub" + input: "loss/dropout/sub/x" + input: "loss/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "loss/dropout/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "loss/dropout/truediv" + op: "RealDiv" + input: "loss/dropout/truediv/x" + input: "loss/dropout/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "loss/dropout/GreaterEqual" + op: "GreaterEqual" + input: "loss/dropout/random_uniform" + input: "loss/dropout/rate" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/dropout/mul" + op: "Mul" + input: "bert/pooler/dense/Tanh" + input: "loss/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/dropout/Cast" + op: "Cast" + input: "loss/dropout/GreaterEqual" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/dropout/mul_1" + op: "Mul" + input: "loss/dropout/mul" + input: "loss/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "loss/MatMul" + op: "MatMul" + input: "loss/dropout/mul_1" + input: "output_weights/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "loss/BiasAdd" + op: "BiasAdd" + input: "loss/MatMul" + input: "output_bias/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "loss/Softmax" + op: "Softmax" + input: "loss/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "loss/LogSoftmax" + op: "LogSoftmax" + input: "loss/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "loss/one_hot/on_value" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "loss/one_hot/off_value" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "loss/one_hot/depth" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 3 + } + } + } +} +node { + name: "loss/one_hot" + op: "OneHot" + input: "IteratorGetNext:3" + input: "loss/one_hot/depth" + input: "loss/one_hot/on_value" + input: "loss/one_hot/off_value" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "TI" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } + attr { + key: "axis" + value { + i: -1 + } + } +} +node { + name: "loss/mul" + op: "Mul" + input: "loss/one_hot" + input: "loss/LogSoftmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "loss/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "loss/Sum" + op: "Sum" + input: "loss/mul" + input: "loss/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "loss/Neg" + op: "Neg" + input: "loss/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } +} +node { + name: "loss/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "loss/Mean" + op: "Mean" + input: "loss/Neg" + input: "loss/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "checkpoint_initializer/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/embeddings/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer" + op: "RestoreV2" + input: "checkpoint_initializer/prefix" + input: "checkpoint_initializer/tensor_names" + input: "checkpoint_initializer/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta" + input: "checkpoint_initializer" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_1/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_1/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/embeddings/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_1/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_1" + op: "RestoreV2" + input: "checkpoint_initializer_1/prefix" + input: "checkpoint_initializer_1/tensor_names" + input: "checkpoint_initializer_1/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_1" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma" + input: "checkpoint_initializer_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_2/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_2/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/embeddings/position_embeddings" + } + } + } +} +node { + name: "checkpoint_initializer_2/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_2" + op: "RestoreV2" + input: "checkpoint_initializer_2/prefix" + input: "checkpoint_initializer_2/tensor_names" + input: "checkpoint_initializer_2/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_2" + op: "Assign" + input: "bert/embeddings/position_embeddings" + input: "checkpoint_initializer_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_3/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_3/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/embeddings/token_type_embeddings" + } + } + } +} +node { + name: "checkpoint_initializer_3/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_3" + op: "RestoreV2" + input: "checkpoint_initializer_3/prefix" + input: "checkpoint_initializer_3/tensor_names" + input: "checkpoint_initializer_3/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_3" + op: "Assign" + input: "bert/embeddings/token_type_embeddings" + input: "checkpoint_initializer_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_4/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_4/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/embeddings/word_embeddings" + } + } + } +} +node { + name: "checkpoint_initializer_4/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_4" + op: "RestoreV2" + input: "checkpoint_initializer_4/prefix" + input: "checkpoint_initializer_4/tensor_names" + input: "checkpoint_initializer_4/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_4" + op: "Assign" + input: "bert/embeddings/word_embeddings" + input: "checkpoint_initializer_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_5/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_5/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_5/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_5" + op: "RestoreV2" + input: "checkpoint_initializer_5/prefix" + input: "checkpoint_initializer_5/tensor_names" + input: "checkpoint_initializer_5/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_5" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_6/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_6/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_6/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_6" + op: "RestoreV2" + input: "checkpoint_initializer_6/prefix" + input: "checkpoint_initializer_6/tensor_names" + input: "checkpoint_initializer_6/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_6" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_7/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_7/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_7/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_7" + op: "RestoreV2" + input: "checkpoint_initializer_7/prefix" + input: "checkpoint_initializer_7/tensor_names" + input: "checkpoint_initializer_7/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_7" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias" + input: "checkpoint_initializer_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_8/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_8/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_8/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_8" + op: "RestoreV2" + input: "checkpoint_initializer_8/prefix" + input: "checkpoint_initializer_8/tensor_names" + input: "checkpoint_initializer_8/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_8" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + input: "checkpoint_initializer_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_9/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_9/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_9/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_9" + op: "RestoreV2" + input: "checkpoint_initializer_9/prefix" + input: "checkpoint_initializer_9/tensor_names" + input: "checkpoint_initializer_9/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_9" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias" + input: "checkpoint_initializer_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_10/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_10/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_10/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_10" + op: "RestoreV2" + input: "checkpoint_initializer_10/prefix" + input: "checkpoint_initializer_10/tensor_names" + input: "checkpoint_initializer_10/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_10" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel" + input: "checkpoint_initializer_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_11/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_11/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_11/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_11" + op: "RestoreV2" + input: "checkpoint_initializer_11/prefix" + input: "checkpoint_initializer_11/tensor_names" + input: "checkpoint_initializer_11/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_11" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias" + input: "checkpoint_initializer_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_12/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_12/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_12/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_12" + op: "RestoreV2" + input: "checkpoint_initializer_12/prefix" + input: "checkpoint_initializer_12/tensor_names" + input: "checkpoint_initializer_12/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_12" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel" + input: "checkpoint_initializer_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_13/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_13/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_13/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_13" + op: "RestoreV2" + input: "checkpoint_initializer_13/prefix" + input: "checkpoint_initializer_13/tensor_names" + input: "checkpoint_initializer_13/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_13" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias" + input: "checkpoint_initializer_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_14/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_14/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_14/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_14" + op: "RestoreV2" + input: "checkpoint_initializer_14/prefix" + input: "checkpoint_initializer_14/tensor_names" + input: "checkpoint_initializer_14/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_14" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel" + input: "checkpoint_initializer_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_15/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_15/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_15/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_15" + op: "RestoreV2" + input: "checkpoint_initializer_15/prefix" + input: "checkpoint_initializer_15/tensor_names" + input: "checkpoint_initializer_15/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_15" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias" + input: "checkpoint_initializer_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_16/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_16/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_16/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_16" + op: "RestoreV2" + input: "checkpoint_initializer_16/prefix" + input: "checkpoint_initializer_16/tensor_names" + input: "checkpoint_initializer_16/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_16" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + input: "checkpoint_initializer_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_17/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_17/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_17/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_17" + op: "RestoreV2" + input: "checkpoint_initializer_17/prefix" + input: "checkpoint_initializer_17/tensor_names" + input: "checkpoint_initializer_17/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_17" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + input: "checkpoint_initializer_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_18/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_18/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_18/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_18" + op: "RestoreV2" + input: "checkpoint_initializer_18/prefix" + input: "checkpoint_initializer_18/tensor_names" + input: "checkpoint_initializer_18/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_18" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + input: "checkpoint_initializer_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_19/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_19/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_19/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_19" + op: "RestoreV2" + input: "checkpoint_initializer_19/prefix" + input: "checkpoint_initializer_19/tensor_names" + input: "checkpoint_initializer_19/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_19" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias" + input: "checkpoint_initializer_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_20/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_20/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_0/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_20/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_20" + op: "RestoreV2" + input: "checkpoint_initializer_20/prefix" + input: "checkpoint_initializer_20/tensor_names" + input: "checkpoint_initializer_20/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_20" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel" + input: "checkpoint_initializer_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_21/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_21/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_21/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_21" + op: "RestoreV2" + input: "checkpoint_initializer_21/prefix" + input: "checkpoint_initializer_21/tensor_names" + input: "checkpoint_initializer_21/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_21" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_22/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_22/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_22/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_22" + op: "RestoreV2" + input: "checkpoint_initializer_22/prefix" + input: "checkpoint_initializer_22/tensor_names" + input: "checkpoint_initializer_22/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_22" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_23/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_23/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_23/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_23" + op: "RestoreV2" + input: "checkpoint_initializer_23/prefix" + input: "checkpoint_initializer_23/tensor_names" + input: "checkpoint_initializer_23/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_23" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias" + input: "checkpoint_initializer_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_24/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_24/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_24/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_24" + op: "RestoreV2" + input: "checkpoint_initializer_24/prefix" + input: "checkpoint_initializer_24/tensor_names" + input: "checkpoint_initializer_24/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_24" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + input: "checkpoint_initializer_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_25/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_25/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_25/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_25" + op: "RestoreV2" + input: "checkpoint_initializer_25/prefix" + input: "checkpoint_initializer_25/tensor_names" + input: "checkpoint_initializer_25/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_25" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias" + input: "checkpoint_initializer_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_26/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_26/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_26/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_26" + op: "RestoreV2" + input: "checkpoint_initializer_26/prefix" + input: "checkpoint_initializer_26/tensor_names" + input: "checkpoint_initializer_26/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_26" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel" + input: "checkpoint_initializer_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_27/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_27/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_27/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_27" + op: "RestoreV2" + input: "checkpoint_initializer_27/prefix" + input: "checkpoint_initializer_27/tensor_names" + input: "checkpoint_initializer_27/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_27" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias" + input: "checkpoint_initializer_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_28/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_28/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_28/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_28" + op: "RestoreV2" + input: "checkpoint_initializer_28/prefix" + input: "checkpoint_initializer_28/tensor_names" + input: "checkpoint_initializer_28/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_28" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel" + input: "checkpoint_initializer_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_29/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_29/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_29/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_29" + op: "RestoreV2" + input: "checkpoint_initializer_29/prefix" + input: "checkpoint_initializer_29/tensor_names" + input: "checkpoint_initializer_29/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_29" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias" + input: "checkpoint_initializer_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_30/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_30/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_30/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_30" + op: "RestoreV2" + input: "checkpoint_initializer_30/prefix" + input: "checkpoint_initializer_30/tensor_names" + input: "checkpoint_initializer_30/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_30" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel" + input: "checkpoint_initializer_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_31/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_31/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_31/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_31" + op: "RestoreV2" + input: "checkpoint_initializer_31/prefix" + input: "checkpoint_initializer_31/tensor_names" + input: "checkpoint_initializer_31/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_31" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias" + input: "checkpoint_initializer_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_32/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_32/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_32/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_32" + op: "RestoreV2" + input: "checkpoint_initializer_32/prefix" + input: "checkpoint_initializer_32/tensor_names" + input: "checkpoint_initializer_32/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_32" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + input: "checkpoint_initializer_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_33/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_33/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_33/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_33" + op: "RestoreV2" + input: "checkpoint_initializer_33/prefix" + input: "checkpoint_initializer_33/tensor_names" + input: "checkpoint_initializer_33/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_33" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + input: "checkpoint_initializer_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_34/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_34/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_34/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_34" + op: "RestoreV2" + input: "checkpoint_initializer_34/prefix" + input: "checkpoint_initializer_34/tensor_names" + input: "checkpoint_initializer_34/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_34" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + input: "checkpoint_initializer_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_35/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_35/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_35/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_35" + op: "RestoreV2" + input: "checkpoint_initializer_35/prefix" + input: "checkpoint_initializer_35/tensor_names" + input: "checkpoint_initializer_35/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_35" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias" + input: "checkpoint_initializer_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_36/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_36/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_1/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_36/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_36" + op: "RestoreV2" + input: "checkpoint_initializer_36/prefix" + input: "checkpoint_initializer_36/tensor_names" + input: "checkpoint_initializer_36/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_36" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel" + input: "checkpoint_initializer_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_37/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_37/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_37/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_37" + op: "RestoreV2" + input: "checkpoint_initializer_37/prefix" + input: "checkpoint_initializer_37/tensor_names" + input: "checkpoint_initializer_37/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_37" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_38/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_38/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_38/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_38" + op: "RestoreV2" + input: "checkpoint_initializer_38/prefix" + input: "checkpoint_initializer_38/tensor_names" + input: "checkpoint_initializer_38/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_38" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_39/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_39/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_39/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_39" + op: "RestoreV2" + input: "checkpoint_initializer_39/prefix" + input: "checkpoint_initializer_39/tensor_names" + input: "checkpoint_initializer_39/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_39" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias" + input: "checkpoint_initializer_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_40/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_40/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_40/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_40" + op: "RestoreV2" + input: "checkpoint_initializer_40/prefix" + input: "checkpoint_initializer_40/tensor_names" + input: "checkpoint_initializer_40/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_40" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + input: "checkpoint_initializer_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_41/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_41/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_41/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_41" + op: "RestoreV2" + input: "checkpoint_initializer_41/prefix" + input: "checkpoint_initializer_41/tensor_names" + input: "checkpoint_initializer_41/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_41" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias" + input: "checkpoint_initializer_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_42/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_42/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_42/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_42" + op: "RestoreV2" + input: "checkpoint_initializer_42/prefix" + input: "checkpoint_initializer_42/tensor_names" + input: "checkpoint_initializer_42/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_42" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel" + input: "checkpoint_initializer_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_43/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_43/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_43/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_43" + op: "RestoreV2" + input: "checkpoint_initializer_43/prefix" + input: "checkpoint_initializer_43/tensor_names" + input: "checkpoint_initializer_43/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_43" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias" + input: "checkpoint_initializer_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_44/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_44/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_44/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_44" + op: "RestoreV2" + input: "checkpoint_initializer_44/prefix" + input: "checkpoint_initializer_44/tensor_names" + input: "checkpoint_initializer_44/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_44" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel" + input: "checkpoint_initializer_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_45/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_45/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_45/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_45" + op: "RestoreV2" + input: "checkpoint_initializer_45/prefix" + input: "checkpoint_initializer_45/tensor_names" + input: "checkpoint_initializer_45/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_45" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias" + input: "checkpoint_initializer_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_46/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_46/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_46/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_46" + op: "RestoreV2" + input: "checkpoint_initializer_46/prefix" + input: "checkpoint_initializer_46/tensor_names" + input: "checkpoint_initializer_46/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_46" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel" + input: "checkpoint_initializer_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_47/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_47/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_47/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_47" + op: "RestoreV2" + input: "checkpoint_initializer_47/prefix" + input: "checkpoint_initializer_47/tensor_names" + input: "checkpoint_initializer_47/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_47" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias" + input: "checkpoint_initializer_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_48/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_48/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_48/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_48" + op: "RestoreV2" + input: "checkpoint_initializer_48/prefix" + input: "checkpoint_initializer_48/tensor_names" + input: "checkpoint_initializer_48/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_48" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + input: "checkpoint_initializer_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_49/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_49/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_49/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_49" + op: "RestoreV2" + input: "checkpoint_initializer_49/prefix" + input: "checkpoint_initializer_49/tensor_names" + input: "checkpoint_initializer_49/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_49" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + input: "checkpoint_initializer_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_50/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_50/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_50/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_50" + op: "RestoreV2" + input: "checkpoint_initializer_50/prefix" + input: "checkpoint_initializer_50/tensor_names" + input: "checkpoint_initializer_50/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_50" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + input: "checkpoint_initializer_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_51/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_51/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_51/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_51" + op: "RestoreV2" + input: "checkpoint_initializer_51/prefix" + input: "checkpoint_initializer_51/tensor_names" + input: "checkpoint_initializer_51/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_51" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias" + input: "checkpoint_initializer_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_52/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_52/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_10/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_52/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_52" + op: "RestoreV2" + input: "checkpoint_initializer_52/prefix" + input: "checkpoint_initializer_52/tensor_names" + input: "checkpoint_initializer_52/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_52" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel" + input: "checkpoint_initializer_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_53/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_53/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_53/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_53" + op: "RestoreV2" + input: "checkpoint_initializer_53/prefix" + input: "checkpoint_initializer_53/tensor_names" + input: "checkpoint_initializer_53/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_53" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_54/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_54/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_54/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_54" + op: "RestoreV2" + input: "checkpoint_initializer_54/prefix" + input: "checkpoint_initializer_54/tensor_names" + input: "checkpoint_initializer_54/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_54" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_55/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_55/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_55/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_55" + op: "RestoreV2" + input: "checkpoint_initializer_55/prefix" + input: "checkpoint_initializer_55/tensor_names" + input: "checkpoint_initializer_55/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_55" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias" + input: "checkpoint_initializer_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_56/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_56/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_56/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_56" + op: "RestoreV2" + input: "checkpoint_initializer_56/prefix" + input: "checkpoint_initializer_56/tensor_names" + input: "checkpoint_initializer_56/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_56" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + input: "checkpoint_initializer_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_57/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_57/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_57/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_57" + op: "RestoreV2" + input: "checkpoint_initializer_57/prefix" + input: "checkpoint_initializer_57/tensor_names" + input: "checkpoint_initializer_57/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_57" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias" + input: "checkpoint_initializer_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_58/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_58/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_58/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_58" + op: "RestoreV2" + input: "checkpoint_initializer_58/prefix" + input: "checkpoint_initializer_58/tensor_names" + input: "checkpoint_initializer_58/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_58" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel" + input: "checkpoint_initializer_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_59/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_59/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_59/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_59" + op: "RestoreV2" + input: "checkpoint_initializer_59/prefix" + input: "checkpoint_initializer_59/tensor_names" + input: "checkpoint_initializer_59/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_59" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias" + input: "checkpoint_initializer_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_60/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_60/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_60/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_60" + op: "RestoreV2" + input: "checkpoint_initializer_60/prefix" + input: "checkpoint_initializer_60/tensor_names" + input: "checkpoint_initializer_60/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_60" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel" + input: "checkpoint_initializer_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_61/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_61/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_61/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_61" + op: "RestoreV2" + input: "checkpoint_initializer_61/prefix" + input: "checkpoint_initializer_61/tensor_names" + input: "checkpoint_initializer_61/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_61" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias" + input: "checkpoint_initializer_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_62/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_62/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_62/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_62" + op: "RestoreV2" + input: "checkpoint_initializer_62/prefix" + input: "checkpoint_initializer_62/tensor_names" + input: "checkpoint_initializer_62/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_62" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel" + input: "checkpoint_initializer_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_63/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_63/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_63/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_63" + op: "RestoreV2" + input: "checkpoint_initializer_63/prefix" + input: "checkpoint_initializer_63/tensor_names" + input: "checkpoint_initializer_63/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_63" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias" + input: "checkpoint_initializer_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_64/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_64/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_64/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_64" + op: "RestoreV2" + input: "checkpoint_initializer_64/prefix" + input: "checkpoint_initializer_64/tensor_names" + input: "checkpoint_initializer_64/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_64" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + input: "checkpoint_initializer_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_65/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_65/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_65/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_65" + op: "RestoreV2" + input: "checkpoint_initializer_65/prefix" + input: "checkpoint_initializer_65/tensor_names" + input: "checkpoint_initializer_65/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_65" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + input: "checkpoint_initializer_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_66/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_66/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_66/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_66" + op: "RestoreV2" + input: "checkpoint_initializer_66/prefix" + input: "checkpoint_initializer_66/tensor_names" + input: "checkpoint_initializer_66/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_66" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + input: "checkpoint_initializer_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_67/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_67/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_67/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_67" + op: "RestoreV2" + input: "checkpoint_initializer_67/prefix" + input: "checkpoint_initializer_67/tensor_names" + input: "checkpoint_initializer_67/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_67" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias" + input: "checkpoint_initializer_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_68/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_68/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_11/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_68/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_68" + op: "RestoreV2" + input: "checkpoint_initializer_68/prefix" + input: "checkpoint_initializer_68/tensor_names" + input: "checkpoint_initializer_68/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_68" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel" + input: "checkpoint_initializer_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_69/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_69/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_69/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_69" + op: "RestoreV2" + input: "checkpoint_initializer_69/prefix" + input: "checkpoint_initializer_69/tensor_names" + input: "checkpoint_initializer_69/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_69" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_70/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_70/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_70/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_70" + op: "RestoreV2" + input: "checkpoint_initializer_70/prefix" + input: "checkpoint_initializer_70/tensor_names" + input: "checkpoint_initializer_70/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_70" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_71/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_71/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_71/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_71" + op: "RestoreV2" + input: "checkpoint_initializer_71/prefix" + input: "checkpoint_initializer_71/tensor_names" + input: "checkpoint_initializer_71/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_71" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias" + input: "checkpoint_initializer_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_72/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_72/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_72/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_72" + op: "RestoreV2" + input: "checkpoint_initializer_72/prefix" + input: "checkpoint_initializer_72/tensor_names" + input: "checkpoint_initializer_72/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_72" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + input: "checkpoint_initializer_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_73/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_73/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_73/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_73" + op: "RestoreV2" + input: "checkpoint_initializer_73/prefix" + input: "checkpoint_initializer_73/tensor_names" + input: "checkpoint_initializer_73/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_73" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias" + input: "checkpoint_initializer_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_74/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_74/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_74/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_74" + op: "RestoreV2" + input: "checkpoint_initializer_74/prefix" + input: "checkpoint_initializer_74/tensor_names" + input: "checkpoint_initializer_74/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_74" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel" + input: "checkpoint_initializer_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_75/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_75/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_75/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_75" + op: "RestoreV2" + input: "checkpoint_initializer_75/prefix" + input: "checkpoint_initializer_75/tensor_names" + input: "checkpoint_initializer_75/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_75" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias" + input: "checkpoint_initializer_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_76/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_76/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_76/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_76" + op: "RestoreV2" + input: "checkpoint_initializer_76/prefix" + input: "checkpoint_initializer_76/tensor_names" + input: "checkpoint_initializer_76/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_76" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel" + input: "checkpoint_initializer_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_77/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_77/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_77/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_77" + op: "RestoreV2" + input: "checkpoint_initializer_77/prefix" + input: "checkpoint_initializer_77/tensor_names" + input: "checkpoint_initializer_77/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_77" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias" + input: "checkpoint_initializer_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_78/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_78/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_78/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_78" + op: "RestoreV2" + input: "checkpoint_initializer_78/prefix" + input: "checkpoint_initializer_78/tensor_names" + input: "checkpoint_initializer_78/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_78" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel" + input: "checkpoint_initializer_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_79/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_79/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_79/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_79" + op: "RestoreV2" + input: "checkpoint_initializer_79/prefix" + input: "checkpoint_initializer_79/tensor_names" + input: "checkpoint_initializer_79/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_79" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias" + input: "checkpoint_initializer_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_80/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_80/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_80/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_80" + op: "RestoreV2" + input: "checkpoint_initializer_80/prefix" + input: "checkpoint_initializer_80/tensor_names" + input: "checkpoint_initializer_80/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_80" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + input: "checkpoint_initializer_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_81/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_81/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_81/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_81" + op: "RestoreV2" + input: "checkpoint_initializer_81/prefix" + input: "checkpoint_initializer_81/tensor_names" + input: "checkpoint_initializer_81/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_81" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + input: "checkpoint_initializer_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_82/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_82/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_82/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_82" + op: "RestoreV2" + input: "checkpoint_initializer_82/prefix" + input: "checkpoint_initializer_82/tensor_names" + input: "checkpoint_initializer_82/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_82" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + input: "checkpoint_initializer_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_83/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_83/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_83/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_83" + op: "RestoreV2" + input: "checkpoint_initializer_83/prefix" + input: "checkpoint_initializer_83/tensor_names" + input: "checkpoint_initializer_83/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_83" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias" + input: "checkpoint_initializer_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_84/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_84/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_2/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_84/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_84" + op: "RestoreV2" + input: "checkpoint_initializer_84/prefix" + input: "checkpoint_initializer_84/tensor_names" + input: "checkpoint_initializer_84/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_84" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel" + input: "checkpoint_initializer_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_85/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_85/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_85/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_85" + op: "RestoreV2" + input: "checkpoint_initializer_85/prefix" + input: "checkpoint_initializer_85/tensor_names" + input: "checkpoint_initializer_85/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_85" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_86/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_86/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_86/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_86" + op: "RestoreV2" + input: "checkpoint_initializer_86/prefix" + input: "checkpoint_initializer_86/tensor_names" + input: "checkpoint_initializer_86/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_86" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_87/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_87/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_87/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_87" + op: "RestoreV2" + input: "checkpoint_initializer_87/prefix" + input: "checkpoint_initializer_87/tensor_names" + input: "checkpoint_initializer_87/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_87" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias" + input: "checkpoint_initializer_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_88/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_88/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_88/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_88" + op: "RestoreV2" + input: "checkpoint_initializer_88/prefix" + input: "checkpoint_initializer_88/tensor_names" + input: "checkpoint_initializer_88/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_88" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + input: "checkpoint_initializer_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_89/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_89/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_89/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_89" + op: "RestoreV2" + input: "checkpoint_initializer_89/prefix" + input: "checkpoint_initializer_89/tensor_names" + input: "checkpoint_initializer_89/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_89" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias" + input: "checkpoint_initializer_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_90/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_90/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_90/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_90" + op: "RestoreV2" + input: "checkpoint_initializer_90/prefix" + input: "checkpoint_initializer_90/tensor_names" + input: "checkpoint_initializer_90/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_90" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel" + input: "checkpoint_initializer_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_91/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_91/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_91/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_91" + op: "RestoreV2" + input: "checkpoint_initializer_91/prefix" + input: "checkpoint_initializer_91/tensor_names" + input: "checkpoint_initializer_91/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_91" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias" + input: "checkpoint_initializer_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_92/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_92/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_92/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_92" + op: "RestoreV2" + input: "checkpoint_initializer_92/prefix" + input: "checkpoint_initializer_92/tensor_names" + input: "checkpoint_initializer_92/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_92" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel" + input: "checkpoint_initializer_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_93/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_93/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_93/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_93" + op: "RestoreV2" + input: "checkpoint_initializer_93/prefix" + input: "checkpoint_initializer_93/tensor_names" + input: "checkpoint_initializer_93/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_93" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias" + input: "checkpoint_initializer_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_94/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_94/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_94/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_94" + op: "RestoreV2" + input: "checkpoint_initializer_94/prefix" + input: "checkpoint_initializer_94/tensor_names" + input: "checkpoint_initializer_94/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_94" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel" + input: "checkpoint_initializer_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_95/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_95/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_95/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_95" + op: "RestoreV2" + input: "checkpoint_initializer_95/prefix" + input: "checkpoint_initializer_95/tensor_names" + input: "checkpoint_initializer_95/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_95" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias" + input: "checkpoint_initializer_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_96/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_96/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_96/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_96" + op: "RestoreV2" + input: "checkpoint_initializer_96/prefix" + input: "checkpoint_initializer_96/tensor_names" + input: "checkpoint_initializer_96/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_96" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + input: "checkpoint_initializer_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_97/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_97/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_97/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_97" + op: "RestoreV2" + input: "checkpoint_initializer_97/prefix" + input: "checkpoint_initializer_97/tensor_names" + input: "checkpoint_initializer_97/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_97" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + input: "checkpoint_initializer_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_98/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_98/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_98/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_98" + op: "RestoreV2" + input: "checkpoint_initializer_98/prefix" + input: "checkpoint_initializer_98/tensor_names" + input: "checkpoint_initializer_98/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_98" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + input: "checkpoint_initializer_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_99/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_99/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_99/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_99" + op: "RestoreV2" + input: "checkpoint_initializer_99/prefix" + input: "checkpoint_initializer_99/tensor_names" + input: "checkpoint_initializer_99/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_99" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias" + input: "checkpoint_initializer_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_100/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_100/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_3/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_100/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_100" + op: "RestoreV2" + input: "checkpoint_initializer_100/prefix" + input: "checkpoint_initializer_100/tensor_names" + input: "checkpoint_initializer_100/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_100" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel" + input: "checkpoint_initializer_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_101/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_101/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_101/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_101" + op: "RestoreV2" + input: "checkpoint_initializer_101/prefix" + input: "checkpoint_initializer_101/tensor_names" + input: "checkpoint_initializer_101/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_101" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_102/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_102/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_102/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_102" + op: "RestoreV2" + input: "checkpoint_initializer_102/prefix" + input: "checkpoint_initializer_102/tensor_names" + input: "checkpoint_initializer_102/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_102" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_103/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_103/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_103/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_103" + op: "RestoreV2" + input: "checkpoint_initializer_103/prefix" + input: "checkpoint_initializer_103/tensor_names" + input: "checkpoint_initializer_103/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_103" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias" + input: "checkpoint_initializer_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_104/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_104/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_104/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_104" + op: "RestoreV2" + input: "checkpoint_initializer_104/prefix" + input: "checkpoint_initializer_104/tensor_names" + input: "checkpoint_initializer_104/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_104" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + input: "checkpoint_initializer_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_105/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_105/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_105/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_105" + op: "RestoreV2" + input: "checkpoint_initializer_105/prefix" + input: "checkpoint_initializer_105/tensor_names" + input: "checkpoint_initializer_105/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_105" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias" + input: "checkpoint_initializer_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_106/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_106/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_106/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_106" + op: "RestoreV2" + input: "checkpoint_initializer_106/prefix" + input: "checkpoint_initializer_106/tensor_names" + input: "checkpoint_initializer_106/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_106" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel" + input: "checkpoint_initializer_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_107/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_107/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_107/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_107" + op: "RestoreV2" + input: "checkpoint_initializer_107/prefix" + input: "checkpoint_initializer_107/tensor_names" + input: "checkpoint_initializer_107/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_107" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias" + input: "checkpoint_initializer_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_108/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_108/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_108/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_108" + op: "RestoreV2" + input: "checkpoint_initializer_108/prefix" + input: "checkpoint_initializer_108/tensor_names" + input: "checkpoint_initializer_108/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_108" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel" + input: "checkpoint_initializer_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_109/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_109/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_109/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_109" + op: "RestoreV2" + input: "checkpoint_initializer_109/prefix" + input: "checkpoint_initializer_109/tensor_names" + input: "checkpoint_initializer_109/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_109" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias" + input: "checkpoint_initializer_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_110/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_110/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_110/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_110" + op: "RestoreV2" + input: "checkpoint_initializer_110/prefix" + input: "checkpoint_initializer_110/tensor_names" + input: "checkpoint_initializer_110/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_110" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel" + input: "checkpoint_initializer_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_111/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_111/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_111/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_111" + op: "RestoreV2" + input: "checkpoint_initializer_111/prefix" + input: "checkpoint_initializer_111/tensor_names" + input: "checkpoint_initializer_111/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_111" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias" + input: "checkpoint_initializer_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_112/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_112/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_112/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_112" + op: "RestoreV2" + input: "checkpoint_initializer_112/prefix" + input: "checkpoint_initializer_112/tensor_names" + input: "checkpoint_initializer_112/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_112" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + input: "checkpoint_initializer_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_113/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_113/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_113/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_113" + op: "RestoreV2" + input: "checkpoint_initializer_113/prefix" + input: "checkpoint_initializer_113/tensor_names" + input: "checkpoint_initializer_113/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_113" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + input: "checkpoint_initializer_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_114/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_114/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_114/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_114" + op: "RestoreV2" + input: "checkpoint_initializer_114/prefix" + input: "checkpoint_initializer_114/tensor_names" + input: "checkpoint_initializer_114/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_114" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + input: "checkpoint_initializer_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_115/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_115/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_115/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_115" + op: "RestoreV2" + input: "checkpoint_initializer_115/prefix" + input: "checkpoint_initializer_115/tensor_names" + input: "checkpoint_initializer_115/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_115" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias" + input: "checkpoint_initializer_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_116/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_116/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_4/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_116/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_116" + op: "RestoreV2" + input: "checkpoint_initializer_116/prefix" + input: "checkpoint_initializer_116/tensor_names" + input: "checkpoint_initializer_116/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_116" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel" + input: "checkpoint_initializer_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_117/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_117/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_117/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_117" + op: "RestoreV2" + input: "checkpoint_initializer_117/prefix" + input: "checkpoint_initializer_117/tensor_names" + input: "checkpoint_initializer_117/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_117" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_118/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_118/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_118/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_118" + op: "RestoreV2" + input: "checkpoint_initializer_118/prefix" + input: "checkpoint_initializer_118/tensor_names" + input: "checkpoint_initializer_118/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_118" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_119/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_119/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_119/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_119" + op: "RestoreV2" + input: "checkpoint_initializer_119/prefix" + input: "checkpoint_initializer_119/tensor_names" + input: "checkpoint_initializer_119/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_119" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias" + input: "checkpoint_initializer_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_120/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_120/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_120/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_120" + op: "RestoreV2" + input: "checkpoint_initializer_120/prefix" + input: "checkpoint_initializer_120/tensor_names" + input: "checkpoint_initializer_120/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_120" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + input: "checkpoint_initializer_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_121/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_121/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_121/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_121" + op: "RestoreV2" + input: "checkpoint_initializer_121/prefix" + input: "checkpoint_initializer_121/tensor_names" + input: "checkpoint_initializer_121/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_121" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias" + input: "checkpoint_initializer_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_122/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_122/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_122/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_122" + op: "RestoreV2" + input: "checkpoint_initializer_122/prefix" + input: "checkpoint_initializer_122/tensor_names" + input: "checkpoint_initializer_122/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_122" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel" + input: "checkpoint_initializer_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_123/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_123/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_123/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_123" + op: "RestoreV2" + input: "checkpoint_initializer_123/prefix" + input: "checkpoint_initializer_123/tensor_names" + input: "checkpoint_initializer_123/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_123" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias" + input: "checkpoint_initializer_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_124/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_124/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_124/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_124" + op: "RestoreV2" + input: "checkpoint_initializer_124/prefix" + input: "checkpoint_initializer_124/tensor_names" + input: "checkpoint_initializer_124/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_124" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel" + input: "checkpoint_initializer_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_125/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_125/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_125/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_125" + op: "RestoreV2" + input: "checkpoint_initializer_125/prefix" + input: "checkpoint_initializer_125/tensor_names" + input: "checkpoint_initializer_125/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_125" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias" + input: "checkpoint_initializer_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_126/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_126/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_126/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_126" + op: "RestoreV2" + input: "checkpoint_initializer_126/prefix" + input: "checkpoint_initializer_126/tensor_names" + input: "checkpoint_initializer_126/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_126" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel" + input: "checkpoint_initializer_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_127/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_127/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_127/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_127" + op: "RestoreV2" + input: "checkpoint_initializer_127/prefix" + input: "checkpoint_initializer_127/tensor_names" + input: "checkpoint_initializer_127/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_127" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias" + input: "checkpoint_initializer_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_128/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_128/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_128/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_128" + op: "RestoreV2" + input: "checkpoint_initializer_128/prefix" + input: "checkpoint_initializer_128/tensor_names" + input: "checkpoint_initializer_128/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_128" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + input: "checkpoint_initializer_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_129/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_129/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_129/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_129" + op: "RestoreV2" + input: "checkpoint_initializer_129/prefix" + input: "checkpoint_initializer_129/tensor_names" + input: "checkpoint_initializer_129/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_129" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + input: "checkpoint_initializer_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_130/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_130/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_130/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_130" + op: "RestoreV2" + input: "checkpoint_initializer_130/prefix" + input: "checkpoint_initializer_130/tensor_names" + input: "checkpoint_initializer_130/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_130" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + input: "checkpoint_initializer_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_131/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_131/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_131/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_131" + op: "RestoreV2" + input: "checkpoint_initializer_131/prefix" + input: "checkpoint_initializer_131/tensor_names" + input: "checkpoint_initializer_131/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_131" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias" + input: "checkpoint_initializer_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_132/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_132/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_5/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_132/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_132" + op: "RestoreV2" + input: "checkpoint_initializer_132/prefix" + input: "checkpoint_initializer_132/tensor_names" + input: "checkpoint_initializer_132/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_132" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel" + input: "checkpoint_initializer_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_133/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_133/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_133/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_133" + op: "RestoreV2" + input: "checkpoint_initializer_133/prefix" + input: "checkpoint_initializer_133/tensor_names" + input: "checkpoint_initializer_133/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_133" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_134/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_134/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_134/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_134" + op: "RestoreV2" + input: "checkpoint_initializer_134/prefix" + input: "checkpoint_initializer_134/tensor_names" + input: "checkpoint_initializer_134/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_134" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_135/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_135/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_135/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_135" + op: "RestoreV2" + input: "checkpoint_initializer_135/prefix" + input: "checkpoint_initializer_135/tensor_names" + input: "checkpoint_initializer_135/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_135" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias" + input: "checkpoint_initializer_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_136/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_136/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_136/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_136" + op: "RestoreV2" + input: "checkpoint_initializer_136/prefix" + input: "checkpoint_initializer_136/tensor_names" + input: "checkpoint_initializer_136/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_136" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + input: "checkpoint_initializer_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_137/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_137/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_137/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_137" + op: "RestoreV2" + input: "checkpoint_initializer_137/prefix" + input: "checkpoint_initializer_137/tensor_names" + input: "checkpoint_initializer_137/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_137" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias" + input: "checkpoint_initializer_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_138/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_138/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_138/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_138" + op: "RestoreV2" + input: "checkpoint_initializer_138/prefix" + input: "checkpoint_initializer_138/tensor_names" + input: "checkpoint_initializer_138/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_138" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel" + input: "checkpoint_initializer_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_139/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_139/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_139/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_139" + op: "RestoreV2" + input: "checkpoint_initializer_139/prefix" + input: "checkpoint_initializer_139/tensor_names" + input: "checkpoint_initializer_139/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_139" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias" + input: "checkpoint_initializer_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_140/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_140/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_140/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_140" + op: "RestoreV2" + input: "checkpoint_initializer_140/prefix" + input: "checkpoint_initializer_140/tensor_names" + input: "checkpoint_initializer_140/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_140" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel" + input: "checkpoint_initializer_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_141/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_141/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_141/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_141" + op: "RestoreV2" + input: "checkpoint_initializer_141/prefix" + input: "checkpoint_initializer_141/tensor_names" + input: "checkpoint_initializer_141/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_141" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias" + input: "checkpoint_initializer_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_142/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_142/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_142/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_142" + op: "RestoreV2" + input: "checkpoint_initializer_142/prefix" + input: "checkpoint_initializer_142/tensor_names" + input: "checkpoint_initializer_142/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_142" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel" + input: "checkpoint_initializer_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_143/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_143/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_143/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_143" + op: "RestoreV2" + input: "checkpoint_initializer_143/prefix" + input: "checkpoint_initializer_143/tensor_names" + input: "checkpoint_initializer_143/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_143" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias" + input: "checkpoint_initializer_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_144/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_144/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_144/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_144" + op: "RestoreV2" + input: "checkpoint_initializer_144/prefix" + input: "checkpoint_initializer_144/tensor_names" + input: "checkpoint_initializer_144/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_144" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + input: "checkpoint_initializer_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_145/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_145/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_145/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_145" + op: "RestoreV2" + input: "checkpoint_initializer_145/prefix" + input: "checkpoint_initializer_145/tensor_names" + input: "checkpoint_initializer_145/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_145" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + input: "checkpoint_initializer_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_146/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_146/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_146/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_146" + op: "RestoreV2" + input: "checkpoint_initializer_146/prefix" + input: "checkpoint_initializer_146/tensor_names" + input: "checkpoint_initializer_146/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_146" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + input: "checkpoint_initializer_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_147/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_147/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_147/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_147" + op: "RestoreV2" + input: "checkpoint_initializer_147/prefix" + input: "checkpoint_initializer_147/tensor_names" + input: "checkpoint_initializer_147/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_147" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias" + input: "checkpoint_initializer_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_148/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_148/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_6/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_148/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_148" + op: "RestoreV2" + input: "checkpoint_initializer_148/prefix" + input: "checkpoint_initializer_148/tensor_names" + input: "checkpoint_initializer_148/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_148" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel" + input: "checkpoint_initializer_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_149/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_149/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_149/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_149" + op: "RestoreV2" + input: "checkpoint_initializer_149/prefix" + input: "checkpoint_initializer_149/tensor_names" + input: "checkpoint_initializer_149/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_149" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_150/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_150/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_150/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_150" + op: "RestoreV2" + input: "checkpoint_initializer_150/prefix" + input: "checkpoint_initializer_150/tensor_names" + input: "checkpoint_initializer_150/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_150" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_151/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_151/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_151/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_151" + op: "RestoreV2" + input: "checkpoint_initializer_151/prefix" + input: "checkpoint_initializer_151/tensor_names" + input: "checkpoint_initializer_151/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_151" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias" + input: "checkpoint_initializer_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_152/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_152/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_152/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_152" + op: "RestoreV2" + input: "checkpoint_initializer_152/prefix" + input: "checkpoint_initializer_152/tensor_names" + input: "checkpoint_initializer_152/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_152" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + input: "checkpoint_initializer_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_153/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_153/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_153/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_153" + op: "RestoreV2" + input: "checkpoint_initializer_153/prefix" + input: "checkpoint_initializer_153/tensor_names" + input: "checkpoint_initializer_153/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_153" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias" + input: "checkpoint_initializer_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_154/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_154/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_154/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_154" + op: "RestoreV2" + input: "checkpoint_initializer_154/prefix" + input: "checkpoint_initializer_154/tensor_names" + input: "checkpoint_initializer_154/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_154" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel" + input: "checkpoint_initializer_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_155/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_155/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_155/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_155" + op: "RestoreV2" + input: "checkpoint_initializer_155/prefix" + input: "checkpoint_initializer_155/tensor_names" + input: "checkpoint_initializer_155/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_155" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias" + input: "checkpoint_initializer_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_156/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_156/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_156/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_156" + op: "RestoreV2" + input: "checkpoint_initializer_156/prefix" + input: "checkpoint_initializer_156/tensor_names" + input: "checkpoint_initializer_156/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_156" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel" + input: "checkpoint_initializer_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_157/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_157/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_157/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_157" + op: "RestoreV2" + input: "checkpoint_initializer_157/prefix" + input: "checkpoint_initializer_157/tensor_names" + input: "checkpoint_initializer_157/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_157" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias" + input: "checkpoint_initializer_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_158/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_158/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_158/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_158" + op: "RestoreV2" + input: "checkpoint_initializer_158/prefix" + input: "checkpoint_initializer_158/tensor_names" + input: "checkpoint_initializer_158/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_158" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel" + input: "checkpoint_initializer_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_159/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_159/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_159/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_159" + op: "RestoreV2" + input: "checkpoint_initializer_159/prefix" + input: "checkpoint_initializer_159/tensor_names" + input: "checkpoint_initializer_159/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_159" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias" + input: "checkpoint_initializer_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_160/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_160/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_160/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_160" + op: "RestoreV2" + input: "checkpoint_initializer_160/prefix" + input: "checkpoint_initializer_160/tensor_names" + input: "checkpoint_initializer_160/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_160" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + input: "checkpoint_initializer_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_161/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_161/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_161/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_161" + op: "RestoreV2" + input: "checkpoint_initializer_161/prefix" + input: "checkpoint_initializer_161/tensor_names" + input: "checkpoint_initializer_161/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_161" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + input: "checkpoint_initializer_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_162/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_162/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_162/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_162" + op: "RestoreV2" + input: "checkpoint_initializer_162/prefix" + input: "checkpoint_initializer_162/tensor_names" + input: "checkpoint_initializer_162/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_162" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + input: "checkpoint_initializer_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_163/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_163/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_163/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_163" + op: "RestoreV2" + input: "checkpoint_initializer_163/prefix" + input: "checkpoint_initializer_163/tensor_names" + input: "checkpoint_initializer_163/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_163" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias" + input: "checkpoint_initializer_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_164/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_164/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_7/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_164/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_164" + op: "RestoreV2" + input: "checkpoint_initializer_164/prefix" + input: "checkpoint_initializer_164/tensor_names" + input: "checkpoint_initializer_164/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_164" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel" + input: "checkpoint_initializer_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_165/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_165/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_165/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_165" + op: "RestoreV2" + input: "checkpoint_initializer_165/prefix" + input: "checkpoint_initializer_165/tensor_names" + input: "checkpoint_initializer_165/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_165" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_166/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_166/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_166/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_166" + op: "RestoreV2" + input: "checkpoint_initializer_166/prefix" + input: "checkpoint_initializer_166/tensor_names" + input: "checkpoint_initializer_166/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_166" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_167/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_167/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_167/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_167" + op: "RestoreV2" + input: "checkpoint_initializer_167/prefix" + input: "checkpoint_initializer_167/tensor_names" + input: "checkpoint_initializer_167/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_167" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias" + input: "checkpoint_initializer_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_168/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_168/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_168/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_168" + op: "RestoreV2" + input: "checkpoint_initializer_168/prefix" + input: "checkpoint_initializer_168/tensor_names" + input: "checkpoint_initializer_168/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_168" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + input: "checkpoint_initializer_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_169/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_169/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_169/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_169" + op: "RestoreV2" + input: "checkpoint_initializer_169/prefix" + input: "checkpoint_initializer_169/tensor_names" + input: "checkpoint_initializer_169/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_169" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias" + input: "checkpoint_initializer_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_170/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_170/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_170/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_170" + op: "RestoreV2" + input: "checkpoint_initializer_170/prefix" + input: "checkpoint_initializer_170/tensor_names" + input: "checkpoint_initializer_170/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_170" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel" + input: "checkpoint_initializer_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_171/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_171/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_171/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_171" + op: "RestoreV2" + input: "checkpoint_initializer_171/prefix" + input: "checkpoint_initializer_171/tensor_names" + input: "checkpoint_initializer_171/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_171" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias" + input: "checkpoint_initializer_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_172/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_172/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_172/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_172" + op: "RestoreV2" + input: "checkpoint_initializer_172/prefix" + input: "checkpoint_initializer_172/tensor_names" + input: "checkpoint_initializer_172/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_172" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel" + input: "checkpoint_initializer_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_173/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_173/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_173/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_173" + op: "RestoreV2" + input: "checkpoint_initializer_173/prefix" + input: "checkpoint_initializer_173/tensor_names" + input: "checkpoint_initializer_173/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_173" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias" + input: "checkpoint_initializer_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_174/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_174/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_174/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_174" + op: "RestoreV2" + input: "checkpoint_initializer_174/prefix" + input: "checkpoint_initializer_174/tensor_names" + input: "checkpoint_initializer_174/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_174" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel" + input: "checkpoint_initializer_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_175/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_175/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_175/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_175" + op: "RestoreV2" + input: "checkpoint_initializer_175/prefix" + input: "checkpoint_initializer_175/tensor_names" + input: "checkpoint_initializer_175/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_175" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias" + input: "checkpoint_initializer_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_176/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_176/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_176/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_176" + op: "RestoreV2" + input: "checkpoint_initializer_176/prefix" + input: "checkpoint_initializer_176/tensor_names" + input: "checkpoint_initializer_176/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_176" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + input: "checkpoint_initializer_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_177/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_177/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_177/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_177" + op: "RestoreV2" + input: "checkpoint_initializer_177/prefix" + input: "checkpoint_initializer_177/tensor_names" + input: "checkpoint_initializer_177/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_177" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + input: "checkpoint_initializer_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_178/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_178/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_178/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_178" + op: "RestoreV2" + input: "checkpoint_initializer_178/prefix" + input: "checkpoint_initializer_178/tensor_names" + input: "checkpoint_initializer_178/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_178" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + input: "checkpoint_initializer_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_179/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_179/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_179/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_179" + op: "RestoreV2" + input: "checkpoint_initializer_179/prefix" + input: "checkpoint_initializer_179/tensor_names" + input: "checkpoint_initializer_179/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_179" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias" + input: "checkpoint_initializer_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_180/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_180/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_8/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_180/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_180" + op: "RestoreV2" + input: "checkpoint_initializer_180/prefix" + input: "checkpoint_initializer_180/tensor_names" + input: "checkpoint_initializer_180/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_180" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel" + input: "checkpoint_initializer_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_181/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_181/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_181/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_181" + op: "RestoreV2" + input: "checkpoint_initializer_181/prefix" + input: "checkpoint_initializer_181/tensor_names" + input: "checkpoint_initializer_181/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_181" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + input: "checkpoint_initializer_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_182/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_182/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_182/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_182" + op: "RestoreV2" + input: "checkpoint_initializer_182/prefix" + input: "checkpoint_initializer_182/tensor_names" + input: "checkpoint_initializer_182/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_182" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + input: "checkpoint_initializer_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_183/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_183/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_183/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_183" + op: "RestoreV2" + input: "checkpoint_initializer_183/prefix" + input: "checkpoint_initializer_183/tensor_names" + input: "checkpoint_initializer_183/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_183" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias" + input: "checkpoint_initializer_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_184/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_184/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_184/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_184" + op: "RestoreV2" + input: "checkpoint_initializer_184/prefix" + input: "checkpoint_initializer_184/tensor_names" + input: "checkpoint_initializer_184/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_184" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + input: "checkpoint_initializer_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_185/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_185/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/key/bias" + } + } + } +} +node { + name: "checkpoint_initializer_185/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_185" + op: "RestoreV2" + input: "checkpoint_initializer_185/prefix" + input: "checkpoint_initializer_185/tensor_names" + input: "checkpoint_initializer_185/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_185" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias" + input: "checkpoint_initializer_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_186/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_186/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/key/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_186/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_186" + op: "RestoreV2" + input: "checkpoint_initializer_186/prefix" + input: "checkpoint_initializer_186/tensor_names" + input: "checkpoint_initializer_186/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_186" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel" + input: "checkpoint_initializer_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_187/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_187/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/query/bias" + } + } + } +} +node { + name: "checkpoint_initializer_187/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_187" + op: "RestoreV2" + input: "checkpoint_initializer_187/prefix" + input: "checkpoint_initializer_187/tensor_names" + input: "checkpoint_initializer_187/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_187" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias" + input: "checkpoint_initializer_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_188/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_188/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/query/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_188/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_188" + op: "RestoreV2" + input: "checkpoint_initializer_188/prefix" + input: "checkpoint_initializer_188/tensor_names" + input: "checkpoint_initializer_188/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_188" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel" + input: "checkpoint_initializer_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_189/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_189/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/value/bias" + } + } + } +} +node { + name: "checkpoint_initializer_189/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_189" + op: "RestoreV2" + input: "checkpoint_initializer_189/prefix" + input: "checkpoint_initializer_189/tensor_names" + input: "checkpoint_initializer_189/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_189" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias" + input: "checkpoint_initializer_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_190/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_190/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/attention/self/value/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_190/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_190" + op: "RestoreV2" + input: "checkpoint_initializer_190/prefix" + input: "checkpoint_initializer_190/tensor_names" + input: "checkpoint_initializer_190/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_190" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel" + input: "checkpoint_initializer_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_191/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_191/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/intermediate/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_191/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_191" + op: "RestoreV2" + input: "checkpoint_initializer_191/prefix" + input: "checkpoint_initializer_191/tensor_names" + input: "checkpoint_initializer_191/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_191" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias" + input: "checkpoint_initializer_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_192/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_192/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_192/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_192" + op: "RestoreV2" + input: "checkpoint_initializer_192/prefix" + input: "checkpoint_initializer_192/tensor_names" + input: "checkpoint_initializer_192/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_192" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + input: "checkpoint_initializer_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_193/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_193/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } +} +node { + name: "checkpoint_initializer_193/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_193" + op: "RestoreV2" + input: "checkpoint_initializer_193/prefix" + input: "checkpoint_initializer_193/tensor_names" + input: "checkpoint_initializer_193/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_193" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + input: "checkpoint_initializer_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_194/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_194/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } +} +node { + name: "checkpoint_initializer_194/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_194" + op: "RestoreV2" + input: "checkpoint_initializer_194/prefix" + input: "checkpoint_initializer_194/tensor_names" + input: "checkpoint_initializer_194/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_194" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + input: "checkpoint_initializer_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_195/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_195/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/output/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_195/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_195" + op: "RestoreV2" + input: "checkpoint_initializer_195/prefix" + input: "checkpoint_initializer_195/tensor_names" + input: "checkpoint_initializer_195/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_195" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias" + input: "checkpoint_initializer_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_196/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_196/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/encoder/layer_9/output/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_196/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_196" + op: "RestoreV2" + input: "checkpoint_initializer_196/prefix" + input: "checkpoint_initializer_196/tensor_names" + input: "checkpoint_initializer_196/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_196" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel" + input: "checkpoint_initializer_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_197/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_197/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/pooler/dense/bias" + } + } + } +} +node { + name: "checkpoint_initializer_197/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_197" + op: "RestoreV2" + input: "checkpoint_initializer_197/prefix" + input: "checkpoint_initializer_197/tensor_names" + input: "checkpoint_initializer_197/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_197" + op: "Assign" + input: "bert/pooler/dense/bias" + input: "checkpoint_initializer_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "checkpoint_initializer_198/prefix" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "./chinese_wwm_ext_L-12_H-768_A-12/bert_model.ckpt" + } + } + } +} +node { + name: "checkpoint_initializer_198/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "bert/pooler/dense/kernel" + } + } + } +} +node { + name: "checkpoint_initializer_198/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 1 + } + } + string_val: "" + } + } + } +} +node { + name: "checkpoint_initializer_198" + op: "RestoreV2" + input: "checkpoint_initializer_198/prefix" + input: "checkpoint_initializer_198/tensor_names" + input: "checkpoint_initializer_198/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + } + } + } +} +node { + name: "Assign_198" + op: "Assign" + input: "bert/pooler/dense/kernel" + input: "checkpoint_initializer_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.9999999494757503e-05 + } + } + } +} +node { + name: "PolynomialDecay/Cast/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "PolynomialDecay/Cast_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "PolynomialDecay/Cast_2/ReadVariableOp" + op: "ReadVariableOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "PolynomialDecay/Cast_2" + op: "Cast" + input: "PolynomialDecay/Cast_2/ReadVariableOp" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/Cast_3/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 14062 + } + } + } +} +node { + name: "PolynomialDecay/Cast_3" + op: "Cast" + input: "PolynomialDecay/Cast_3/x" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT32 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/Minimum/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 14062.0 + } + } + } +} +node { + name: "PolynomialDecay/Minimum" + op: "Minimum" + input: "PolynomialDecay/Cast_2" + input: "PolynomialDecay/Minimum/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/div" + op: "RealDiv" + input: "PolynomialDecay/Minimum" + input: "PolynomialDecay/Cast_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/sub" + op: "Sub" + input: "Const_1" + input: "PolynomialDecay/Cast/x" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/sub_1/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "PolynomialDecay/sub_1" + op: "Sub" + input: "PolynomialDecay/sub_1/x" + input: "PolynomialDecay/div" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/Pow" + op: "Pow" + input: "PolynomialDecay/sub_1" + input: "PolynomialDecay/Cast_1/x" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay/Mul" + op: "Mul" + input: "PolynomialDecay/sub" + input: "PolynomialDecay/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "PolynomialDecay" + op: "Add" + input: "PolynomialDecay/Mul" + input: "PolynomialDecay/Cast/x" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Cast_1/ReadVariableOp" + op: "ReadVariableOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "Cast_1" + op: "Cast" + input: "Cast_1/ReadVariableOp" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Const_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1406 + } + } + } +} +node { + name: "Cast_2" + op: "Cast" + input: "Cast_1" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT32 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Cast_3" + op: "Cast" + input: "Const_2" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_INT32 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "truediv" + op: "RealDiv" + input: "Cast_2" + input: "Cast_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.9999999494757503e-05 + } + } + } +} +node { + name: "mul" + op: "Mul" + input: "mul/x" + input: "truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Less" + op: "Less" + input: "Cast_1" + input: "Const_2" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "Cast_4" + op: "Cast" + input: "Less" + attr { + key: "DstT" + value { + type: DT_FLOAT + } + } + attr { + key: "SrcT" + value { + type: DT_BOOL + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "sub/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "sub" + op: "Sub" + input: "sub/x" + input: "Cast_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "mul_1" + op: "Mul" + input: "sub" + input: "PolynomialDecay" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "mul_2" + op: "Mul" + input: "Cast_4" + input: "mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "add" + op: "Add" + input: "mul_1" + input: "mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/grad_ys_0" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/Fill" + op: "Fill" + input: "gradients/Shape" + input: "gradients/grad_ys_0" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/loss/Mean_grad/Reshape/shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/loss/Mean_grad/Reshape" + op: "Reshape" + input: "gradients/Fill" + input: "gradients/loss/Mean_grad/Reshape/shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/loss/Mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 32 + } + } + } +} +node { + name: "gradients/loss/Mean_grad/Tile" + op: "Tile" + input: "gradients/loss/Mean_grad/Reshape" + input: "gradients/loss/Mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } +} +node { + name: "gradients/loss/Mean_grad/Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 32.0 + } + } + } +} +node { + name: "gradients/loss/Mean_grad/truediv" + op: "RealDiv" + input: "gradients/loss/Mean_grad/Tile" + input: "gradients/loss/Mean_grad/Const_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } +} +node { + name: "gradients/loss/Neg_grad/Neg" + op: "Neg" + input: "gradients/loss/Mean_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: " \000\000\000\003\000\000\000" + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/loss/Sum_grad/add" + op: "Add" + input: "loss/Sum/reduction_indices" + input: "gradients/loss/Sum_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/mod" + op: "FloorMod" + input: "gradients/loss/Sum_grad/add" + input: "gradients/loss/Sum_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/loss/Sum_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/loss/Sum_grad/range" + op: "Range" + input: "gradients/loss/Sum_grad/range/start" + input: "gradients/loss/Sum_grad/Size" + input: "gradients/loss/Sum_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Fill" + op: "Fill" + input: "gradients/loss/Sum_grad/Shape_1" + input: "gradients/loss/Sum_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/loss/Sum_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/loss/Sum_grad/range" + input: "gradients/loss/Sum_grad/mod" + input: "gradients/loss/Sum_grad/Shape" + input: "gradients/loss/Sum_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Maximum" + op: "Maximum" + input: "gradients/loss/Sum_grad/DynamicStitch" + input: "gradients/loss/Sum_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/floordiv" + op: "FloorDiv" + input: "gradients/loss/Sum_grad/Shape" + input: "gradients/loss/Sum_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/Sum_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Reshape" + op: "Reshape" + input: "gradients/loss/Neg_grad/Neg" + input: "gradients/loss/Sum_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/loss/Sum_grad/Tile" + op: "Tile" + input: "gradients/loss/Sum_grad/Reshape" + input: "gradients/loss/Sum_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/mul_grad/Mul" + op: "Mul" + input: "gradients/loss/Sum_grad/Tile" + input: "loss/LogSoftmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/mul_grad/Mul_1" + op: "Mul" + input: "gradients/loss/Sum_grad/Tile" + input: "loss/one_hot" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/LogSoftmax_grad/Exp" + op: "Exp" + input: "loss/LogSoftmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/LogSoftmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/loss/LogSoftmax_grad/Sum" + op: "Sum" + input: "gradients/loss/mul_grad/Mul_1" + input: "gradients/loss/LogSoftmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/loss/LogSoftmax_grad/mul" + op: "Mul" + input: "gradients/loss/LogSoftmax_grad/Sum" + input: "gradients/loss/LogSoftmax_grad/Exp" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/LogSoftmax_grad/sub" + op: "Sub" + input: "gradients/loss/mul_grad/Mul_1" + input: "gradients/loss/LogSoftmax_grad/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/loss/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/loss/LogSoftmax_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/loss/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/loss/LogSoftmax_grad/sub" + input: "output_weights/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/loss/MatMul_grad/MatMul_1" + op: "MatMul" + input: "gradients/loss/LogSoftmax_grad/sub" + input: "loss/dropout/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/loss/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/loss/MatMul_grad/MatMul" + input: "loss/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/loss/MatMul_grad/MatMul" + input: "loss/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: " \000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/loss/dropout/mul_grad/Shape" + input: "gradients/loss/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/loss/dropout/mul_1_grad/Mul" + input: "loss/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/loss/dropout/mul_grad/Mul" + input: "gradients/loss/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/loss/dropout/mul_grad/Sum" + input: "gradients/loss/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/pooler/dense/Tanh" + input: "gradients/loss/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/loss/dropout/mul_grad/Mul_1" + input: "gradients/loss/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/loss/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/loss/dropout/mul_grad/Sum_1" + input: "gradients/loss/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/pooler/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/pooler/dense/Tanh" + input: "gradients/loss/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/pooler/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/pooler/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/pooler/dense/Tanh_grad/TanhGrad" + input: "bert/pooler/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/pooler/Squeeze" + input: "gradients/bert/pooler/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/pooler/Squeeze_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\001\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/pooler/Squeeze_grad/Reshape" + op: "Reshape" + input: "gradients/bert/pooler/dense/MatMul_grad/MatMul" + input: "gradients/bert/pooler/Squeeze_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/pooler/strided_slice_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/pooler/strided_slice_grad/StridedSliceGrad" + op: "StridedSliceGrad" + input: "gradients/bert/pooler/strided_slice_grad/Shape" + input: "bert/pooler/strided_slice/stack" + input: "bert/pooler/strided_slice/stack_1" + input: "bert/pooler/strided_slice/stack_2" + input: "gradients/bert/pooler/Squeeze_grad/Reshape" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 5 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 5 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/Reshape_13_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/Reshape_13_grad/Reshape" + op: "Reshape" + input: "gradients/bert/pooler/strided_slice_grad/StridedSliceGrad" + input: "gradients/bert/encoder/Reshape_13_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/Reshape_13_grad/Reshape" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/Reshape_13_grad/Reshape" + input: "bert/encoder/layer_11/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/Reshape_13_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/Reshape_13_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN" + op: "AddN" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_11/output/add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_11/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_11/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_1" + op: "AddN" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_1" + input: "bert/encoder/layer_11/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_1" + input: "bert/encoder/layer_11/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_11/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_11/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_11/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_11/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_11/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_11/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_11/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_11/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_11/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_11/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_11/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_11/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_11/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_11/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_2" + op: "AddN" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_2" + input: "bert/encoder/layer_11/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_3" + op: "AddN" + input: "gradients/AddN_1" + input: "gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_3" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_3" + input: "bert/encoder/layer_11/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_3" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_3" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_4" + op: "AddN" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_4" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_11/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_5" + op: "AddN" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_5" + input: "bert/encoder/layer_11/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_5" + input: "bert/encoder/layer_11/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_11/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_11/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_11/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_11/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_11/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_11/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_11/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_11/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_11/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/Softmax" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_11/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_11/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_11/attention/self/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_11/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_11/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_11/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_11/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_11/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_11/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_11/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_6" + op: "AddN" + input: "gradients/AddN_5" + input: "gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_6" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_6" + input: "bert/encoder/layer_10/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_6" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_6" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_7" + op: "AddN" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_7" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_10/output/add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_10/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_10/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_8" + op: "AddN" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_8" + input: "bert/encoder/layer_10/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_8" + input: "bert/encoder/layer_10/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_10/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_10/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_10/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_10/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_10/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_10/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_10/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_10/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_10/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_10/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_10/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_10/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_10/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_9" + op: "AddN" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_10/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_9" + input: "bert/encoder/layer_10/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_10" + op: "AddN" + input: "gradients/AddN_8" + input: "gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_10" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_10" + input: "bert/encoder/layer_10/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_10" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_10" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_11" + op: "AddN" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_11" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_10/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_12" + op: "AddN" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_12" + input: "bert/encoder/layer_10/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_12" + input: "bert/encoder/layer_10/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_10/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_10/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_10/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_10/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_10/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_10/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_10/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_10/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_10/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/Softmax" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_10/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_10/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_10/attention/self/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_10/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_10/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_10/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_10/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_10/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_10/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_10/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_13" + op: "AddN" + input: "gradients/AddN_12" + input: "gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_13" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_13" + input: "bert/encoder/layer_9/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_13" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_13" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_14" + op: "AddN" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_14" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_9/output/add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_9/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_9/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_15" + op: "AddN" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_15" + input: "bert/encoder/layer_9/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_15" + input: "bert/encoder/layer_9/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_9/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_9/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_9/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_9/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_9/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_9/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_9/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_9/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_9/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_9/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_9/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_9/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_9/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_16" + op: "AddN" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_9/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_16" + input: "bert/encoder/layer_9/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_17" + op: "AddN" + input: "gradients/AddN_15" + input: "gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_17" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_17" + input: "bert/encoder/layer_9/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_17" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_17" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_18" + op: "AddN" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_18" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_9/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_19" + op: "AddN" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_19" + input: "bert/encoder/layer_9/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_19" + input: "bert/encoder/layer_9/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_9/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_9/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_9/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_9/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_9/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_9/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_9/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_9/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_9/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/Softmax" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_9/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_9/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_9/attention/self/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_9/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_9/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_9/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_9/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_9/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_9/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_9/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_20" + op: "AddN" + input: "gradients/AddN_19" + input: "gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_20" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_20" + input: "bert/encoder/layer_8/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_20" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_20" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_21" + op: "AddN" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_21" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_8/output/add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_8/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_8/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_22" + op: "AddN" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_22" + input: "bert/encoder/layer_8/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_22" + input: "bert/encoder/layer_8/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_8/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_8/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_8/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_8/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_8/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_8/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_8/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_8/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_8/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_8/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_8/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_8/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_8/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_23" + op: "AddN" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_8/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_23" + input: "bert/encoder/layer_8/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_24" + op: "AddN" + input: "gradients/AddN_22" + input: "gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_24" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_24" + input: "bert/encoder/layer_8/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_24" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_24" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_25" + op: "AddN" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_25" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_8/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_26" + op: "AddN" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_26" + input: "bert/encoder/layer_8/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_26" + input: "bert/encoder/layer_8/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_8/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_8/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_8/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_8/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_8/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_8/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_8/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_8/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_8/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/Softmax" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_8/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_8/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_8/attention/self/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_8/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_8/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_8/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_8/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_8/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_8/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_8/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_27" + op: "AddN" + input: "gradients/AddN_26" + input: "gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_27" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_27" + input: "bert/encoder/layer_7/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_27" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_27" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_28" + op: "AddN" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_28" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_7/output/add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_7/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_7/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_29" + op: "AddN" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_29" + input: "bert/encoder/layer_7/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_29" + input: "bert/encoder/layer_7/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_7/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_7/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_7/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_7/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_7/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_7/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_7/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_7/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_7/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_7/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_7/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_7/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_7/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_30" + op: "AddN" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_7/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_30" + input: "bert/encoder/layer_7/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_31" + op: "AddN" + input: "gradients/AddN_29" + input: "gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_31" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_31" + input: "bert/encoder/layer_7/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_31" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_31" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_32" + op: "AddN" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_32" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_7/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_33" + op: "AddN" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_33" + input: "bert/encoder/layer_7/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_33" + input: "bert/encoder/layer_7/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_7/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_7/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_7/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_7/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_7/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_7/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_7/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_7/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_7/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/Softmax" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_7/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_7/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_7/attention/self/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_7/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_7/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_7/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_7/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_7/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_7/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_7/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_34" + op: "AddN" + input: "gradients/AddN_33" + input: "gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_34" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_34" + input: "bert/encoder/layer_6/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_34" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_34" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_35" + op: "AddN" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_35" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_6/output/add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_6/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_6/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_36" + op: "AddN" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_36" + input: "bert/encoder/layer_6/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_36" + input: "bert/encoder/layer_6/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_6/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_6/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_6/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_6/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_6/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_6/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_6/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_6/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_6/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_6/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_6/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_6/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_6/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_37" + op: "AddN" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_6/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_37" + input: "bert/encoder/layer_6/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_38" + op: "AddN" + input: "gradients/AddN_36" + input: "gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_38" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_38" + input: "bert/encoder/layer_6/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_38" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_38" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_39" + op: "AddN" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_39" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_6/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_40" + op: "AddN" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_40" + input: "bert/encoder/layer_6/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_40" + input: "bert/encoder/layer_6/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_6/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_6/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_6/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_6/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_6/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_6/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_6/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_6/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_6/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/Softmax" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_6/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_6/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_6/attention/self/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_6/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_6/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_6/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_6/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_6/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_6/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_6/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_41" + op: "AddN" + input: "gradients/AddN_40" + input: "gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_41" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_41" + input: "bert/encoder/layer_5/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_41" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_41" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_42" + op: "AddN" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_42" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_5/output/add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_5/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_5/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_43" + op: "AddN" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_43" + input: "bert/encoder/layer_5/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_43" + input: "bert/encoder/layer_5/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_5/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_5/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_5/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_5/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_5/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_5/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_5/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_5/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_5/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_5/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_5/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_5/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_5/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_44" + op: "AddN" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_5/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_44" + input: "bert/encoder/layer_5/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_45" + op: "AddN" + input: "gradients/AddN_43" + input: "gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_45" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_45" + input: "bert/encoder/layer_5/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_45" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_45" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_46" + op: "AddN" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_46" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_5/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_47" + op: "AddN" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_47" + input: "bert/encoder/layer_5/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_47" + input: "bert/encoder/layer_5/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_5/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_5/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_5/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_5/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_5/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_5/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_5/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_5/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_5/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/Softmax" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_5/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_5/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_5/attention/self/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_5/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_5/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_5/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_5/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_5/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_5/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_5/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_48" + op: "AddN" + input: "gradients/AddN_47" + input: "gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_48" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_48" + input: "bert/encoder/layer_4/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_48" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_48" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_49" + op: "AddN" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_49" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_4/output/add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_4/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_4/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_50" + op: "AddN" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_50" + input: "bert/encoder/layer_4/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_50" + input: "bert/encoder/layer_4/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_4/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_4/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_4/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_4/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_4/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_4/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_4/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_4/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_4/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_4/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_4/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_4/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_4/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_51" + op: "AddN" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_4/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_51" + input: "bert/encoder/layer_4/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_52" + op: "AddN" + input: "gradients/AddN_50" + input: "gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_52" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_52" + input: "bert/encoder/layer_4/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_52" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_52" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_53" + op: "AddN" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_53" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_4/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_54" + op: "AddN" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_54" + input: "bert/encoder/layer_4/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_54" + input: "bert/encoder/layer_4/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_4/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_4/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_4/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_4/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_4/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_4/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_4/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_4/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_4/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/Softmax" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_4/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_4/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_4/attention/self/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_4/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_4/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_4/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_4/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_4/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_4/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_4/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_55" + op: "AddN" + input: "gradients/AddN_54" + input: "gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_55" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_55" + input: "bert/encoder/layer_3/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_55" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_55" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_56" + op: "AddN" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_56" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_3/output/add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_3/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_3/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_57" + op: "AddN" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_57" + input: "bert/encoder/layer_3/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_57" + input: "bert/encoder/layer_3/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_3/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_3/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_3/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_3/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_3/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_3/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_3/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_3/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_3/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_3/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_3/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_3/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_3/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_58" + op: "AddN" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_3/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_58" + input: "bert/encoder/layer_3/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_59" + op: "AddN" + input: "gradients/AddN_57" + input: "gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_59" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_59" + input: "bert/encoder/layer_3/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_59" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_59" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_60" + op: "AddN" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_60" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_3/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_61" + op: "AddN" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_61" + input: "bert/encoder/layer_3/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_61" + input: "bert/encoder/layer_3/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_3/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_3/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_3/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_3/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_3/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_3/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_3/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_3/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_3/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/Softmax" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_3/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_3/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_3/attention/self/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_3/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_3/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_3/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_3/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_3/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_3/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_3/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_62" + op: "AddN" + input: "gradients/AddN_61" + input: "gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_62" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_62" + input: "bert/encoder/layer_2/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_62" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_62" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_63" + op: "AddN" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_63" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_2/output/add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_2/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_2/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_64" + op: "AddN" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_64" + input: "bert/encoder/layer_2/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_64" + input: "bert/encoder/layer_2/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_2/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_2/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_2/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_2/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_2/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_2/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_2/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_2/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_2/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_2/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_2/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_2/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_2/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_65" + op: "AddN" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_2/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_65" + input: "bert/encoder/layer_2/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_66" + op: "AddN" + input: "gradients/AddN_64" + input: "gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_66" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_66" + input: "bert/encoder/layer_2/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_66" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_66" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_67" + op: "AddN" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_67" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_2/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_68" + op: "AddN" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_68" + input: "bert/encoder/layer_2/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_68" + input: "bert/encoder/layer_2/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_2/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_2/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_2/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_2/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_2/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_2/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_2/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_2/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_2/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/Softmax" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_2/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_2/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_2/attention/self/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_2/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_2/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_2/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_2/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_2/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_2/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_2/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_69" + op: "AddN" + input: "gradients/AddN_68" + input: "gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_69" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_69" + input: "bert/encoder/layer_1/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_69" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_69" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_70" + op: "AddN" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_70" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_1/output/add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_1/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_1/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_71" + op: "AddN" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_71" + input: "bert/encoder/layer_1/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_71" + input: "bert/encoder/layer_1/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_1/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_1/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_1/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_1/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_1/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_1/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_1/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_1/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_1/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_1/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_1/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_1/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_1/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_72" + op: "AddN" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_1/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_72" + input: "bert/encoder/layer_1/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_73" + op: "AddN" + input: "gradients/AddN_71" + input: "gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_73" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_73" + input: "bert/encoder/layer_1/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_73" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_73" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_74" + op: "AddN" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_74" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_1/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_75" + op: "AddN" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_75" + input: "bert/encoder/layer_1/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_75" + input: "bert/encoder/layer_1/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_1/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_1/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_1/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_1/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_1/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_1/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_1/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_1/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_1/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/Softmax" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_1/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_1/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_1/attention/self/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_1/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_1/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_1/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_1/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_1/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_1/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/add_1" + input: "gradients/bert/encoder/layer_1/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_76" + op: "AddN" + input: "gradients/AddN_75" + input: "gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_76" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_76" + input: "bert/encoder/layer_0/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_76" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_76" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_77" + op: "AddN" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_77" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_0/output/add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_0/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_0/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_78" + op: "AddN" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_78" + input: "bert/encoder/layer_0/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_78" + input: "bert/encoder/layer_0/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_0/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_0/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/intermediate/dense/mul_3" + input: "gradients/bert/encoder/layer_0/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_0/intermediate/dense/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul_1" + input: "bert/encoder/layer_0/intermediate/dense/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul_2/x" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_2_grad/Reshape_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Tanh_grad/TanhGrad" + op: "TanhGrad" + input: "bert/encoder/layer_0/intermediate/dense/Tanh" + input: "gradients/bert/encoder/layer_0/intermediate/dense/add_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Tanh_grad/TanhGrad" + input: "bert/encoder/layer_0/intermediate/dense/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul_1/x" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Tanh_grad/TanhGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Reshape_1" + input: "bert/encoder/layer_0/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/intermediate/dense/mul/x" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_0/intermediate/dense/Pow/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/sub/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/sub" + op: "Sub" + input: "bert/encoder/layer_0/intermediate/dense/Pow/y" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/sub/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Pow" + op: "Pow" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Greater/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Greater" + op: "Greater" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Greater/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like" + op: "Fill" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like/Shape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Select" + op: "Select" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Greater" + input: "bert/encoder/layer_0/intermediate/dense/BiasAdd" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/ones_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Log" + op: "Log" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\014\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like" + op: "Fill" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like/shape_as_tensor" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Select_1" + op: "Select" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Greater" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Log" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/zeros_like" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_2" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_grad/Reshape_1" + input: "bert/encoder/layer_0/intermediate/dense/Pow" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_3" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_2" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Select_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/mul_3" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/AddN_79" + op: "AddN" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/mul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_0/intermediate/dense/Pow_grad/Reshape" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/mul_3_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/AddN_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/AddN_79" + input: "bert/encoder/layer_0/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_1" + input: "gradients/AddN_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_80" + op: "AddN" + input: "gradients/AddN_78" + input: "gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_80" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_80" + input: "bert/encoder/layer_0/attention/output/add" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/AddN_80" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_80" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_81" + op: "AddN" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_81" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/StopGradient" + input: "^gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/encoder/layer_0/attention/output/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_82" + op: "AddN" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/AddN_82" + input: "bert/encoder/layer_0/attention/output/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/AddN_82" + input: "bert/encoder/layer_0/attention/output/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_0/attention/output/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/output/dense/BiasAdd" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_0/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/layer_0/attention/self/Reshape_3" + input: "gradients/bert/encoder/layer_0/attention/output/dropout/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\200\000\000\000\014\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_3_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_0/attention/self/transpose_3/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_3_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_3_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_3_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_3_grad/transpose" + input: "bert/encoder/layer_0/attention/self/transpose_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/MatMul_1" + op: "BatchMatMulV2" + input: "bert/encoder/layer_0/attention/self/dropout/mul_1" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_3_grad/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000@\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack_1" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -2 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1" + op: "StridedSlice" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape_1" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack_1" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/strided_slice_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/MatMul_1" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_2_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_0/attention/self/transpose_2/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_2_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_1_grad/Reshape_1" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_2_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_1_grad/Mul" + input: "bert/encoder/layer_0/attention/self/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/Softmax" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_2_grad/transpose" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/Sum/reduction_indices" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: -1 + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/mul" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/Sum/reduction_indices" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/sub" + op: "Sub" + input: "gradients/bert/encoder/layer_0/attention/self/dropout/mul_grad/Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/Sum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/mul_1" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/sub" + input: "bert/encoder/layer_0/attention/self/Softmax" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\001\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Softmax_grad/mul_1" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 4 + } + } + tensor_content: " \000\000\000\014\000\000\000\200\000\000\000\200\000\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/Mul/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Sum" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Mul" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Mul_1" + op: "Mul" + input: "bert/encoder/layer_0/attention/self/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/add_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 128 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Mul_1" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Sum_1" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_grad/MatMul" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/transpose_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: false + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/MatMul_grad/MatMul_1" + op: "BatchMatMulV2" + input: "gradients/bert/encoder/layer_0/attention/self/Mul_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/transpose" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 12 + } + dim { + size: 128 + } + dim { + size: 64 + } + } + } + } + } + attr { + key: "adj_x" + value { + b: true + } + } + attr { + key: "adj_y" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_0/attention/self/transpose/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_1_grad/InvertPermutation" + op: "InvertPermutation" + input: "bert/encoder/layer_0/attention/self/transpose_1/perm" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/transpose_1_grad/transpose" + op: "Transpose" + input: "gradients/bert/encoder/layer_0/attention/self/MatMul_grad/MatMul_1" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_1_grad/InvertPermutation" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tperm" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 12 + } + dim { + size: 64 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_grad/transpose" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/encoder/layer_0/attention/self/transpose_1_grad/transpose" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + op: "BiasAddGrad" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "data_format" + value { + s: "NHWC" + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Reshape" + input: "bert/encoder/layer_0/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/encoder/Reshape_1" + input: "gradients/bert/encoder/layer_0/attention/self/Reshape_1_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "gradients/AddN_83" + op: "AddN" + input: "gradients/AddN_82" + input: "gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul" + input: "gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul" + attr { + key: "N" + value { + i: 4 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/encoder/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/encoder/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/AddN_83" + input: "gradients/bert/encoder/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/encoder/Reshape_1_grad/Reshape" + input: "bert/embeddings/dropout/Cast" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/encoder/Reshape_1_grad/Reshape" + input: "bert/embeddings/dropout/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/dropout/mul_grad/Shape" + input: "gradients/bert/embeddings/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Mul" + op: "Mul" + input: "gradients/bert/embeddings/dropout/mul_1_grad/Mul" + input: "bert/embeddings/dropout/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/dropout/mul_grad/Mul" + input: "gradients/bert/embeddings/dropout/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/dropout/mul_grad/Sum" + input: "gradients/bert/embeddings/dropout/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Mul_1" + op: "Mul" + input: "bert/embeddings/LayerNorm/batchnorm/add_1" + input: "gradients/bert/embeddings/dropout/mul_1_grad/Mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/dropout/mul_grad/Mul_1" + input: "gradients/bert/embeddings/dropout/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/dropout/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/dropout/mul_grad/Sum_1" + input: "gradients/bert/embeddings/dropout/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul" + op: "Mul" + input: "gradients/bert/embeddings/dropout/mul_grad/Reshape" + input: "bert/embeddings/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul_1" + op: "Mul" + input: "gradients/bert/embeddings/dropout/mul_grad/Reshape" + input: "bert/embeddings/add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/dropout/mul_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/dropout/mul_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Neg" + op: "Neg" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Sum_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Neg" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Mul" + op: "Mul" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape_1" + input: "bert/embeddings/LayerNorm/batchnorm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Mul" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Mul_1" + op: "Mul" + input: "bert/embeddings/LayerNorm/moments/mean" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Mul_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Sum_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_84" + op: "AddN" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Reshape_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 768 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Mul" + op: "Mul" + input: "gradients/AddN_84" + input: "bert/embeddings/LayerNorm/gamma/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Mul" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Mul_1" + op: "Mul" + input: "bert/embeddings/LayerNorm/batchnorm/Rsqrt" + input: "gradients/AddN_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Mul_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Sum_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + op: "RsqrtGrad" + input: "bert/embeddings/LayerNorm/batchnorm/Rsqrt" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/Rsqrt_grad/RsqrtGrad" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Sum_1" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 3 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/add" + op: "Add" + input: "bert/embeddings/LayerNorm/moments/variance/reduction_indices" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/mod" + op: "FloorMod" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/add" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range" + op: "Range" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range/start" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Size" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Fill" + op: "Fill" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape_1" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/range" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/mod" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Maximum" + op: "Maximum" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/DynamicStitch" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/variance_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/add_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Tile" + op: "Tile" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/truediv" + op: "RealDiv" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Tile" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\001\000\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/scalar" + op: "Const" + input: "^gradients/bert/embeddings/LayerNorm/moments/variance_grad/truediv" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Mul" + op: "Mul" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/scalar" + input: "gradients/bert/embeddings/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/sub" + op: "Sub" + input: "bert/embeddings/add_1" + input: "bert/embeddings/LayerNorm/moments/StopGradient" + input: "^gradients/bert/embeddings/LayerNorm/moments/variance_grad/truediv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/mul_1" + op: "Mul" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Mul" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/sub" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Sum" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Sum" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Sum_1" + op: "Sum" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/mul_1" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Sum_1" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Neg" + op: "Neg" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Size" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 3 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/add" + op: "Add" + input: "bert/embeddings/LayerNorm/moments/mean/reduction_indices" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/mod" + op: "FloorMod" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/add" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Size" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape_1" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range/start" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range/delta" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range" + op: "Range" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range/start" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Size" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range/delta" + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Fill/value" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Fill" + op: "Fill" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape_1" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Fill/value" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/DynamicStitch" + op: "DynamicStitch" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/range" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/mod" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Fill" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Maximum/y" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Maximum" + op: "Maximum" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/DynamicStitch" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Maximum/y" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/floordiv" + op: "FloorDiv" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Maximum" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/moments/mean_grad/Shape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_2_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/DynamicStitch" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Tile" + op: "Tile" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/floordiv" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tmultiples" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 768.0 + } + } + } +} +node { + name: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/truediv" + op: "RealDiv" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Tile" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/AddN_85" + op: "AddN" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul" + input: "gradients/bert/embeddings/LayerNorm/moments/SquaredDifference_grad/Reshape" + input: "gradients/bert/embeddings/LayerNorm/moments/mean_grad/truediv" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/mul_1_grad/Mul" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: " \000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 3 + } + } + tensor_content: "\001\000\000\000\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/BroadcastGradientArgs" + op: "BroadcastGradientArgs" + input: "gradients/bert/embeddings/add_1_grad/Shape" + input: "gradients/bert/embeddings/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Sum" + op: "Sum" + input: "gradients/AddN_85" + input: "gradients/bert/embeddings/add_1_grad/BroadcastGradientArgs" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/add_1_grad/Sum" + input: "gradients/bert/embeddings/add_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 32 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Sum_1" + op: "Sum" + input: "gradients/AddN_85" + input: "gradients/bert/embeddings/add_1_grad/BroadcastGradientArgs:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "gradients/bert/embeddings/add_1_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/add_1_grad/Sum_1" + input: "gradients/bert/embeddings/add_1_grad/Shape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_4_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_4_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/add_1_grad/Reshape_1" + input: "gradients/bert/embeddings/Reshape_4_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_1_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_1_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/add_1_grad/Reshape" + input: "gradients/bert/embeddings/Reshape_1_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_3_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\020\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/Reshape_3_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/add_1_grad/Reshape" + input: "gradients/bert/embeddings/Reshape_3_grad/Shape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Rank" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 2 + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Shape" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\200\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/stack/1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/stack" + op: "Pack" + input: "gradients/bert/embeddings/Slice_grad/Rank" + input: "gradients/bert/embeddings/Slice_grad/stack/1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Reshape" + op: "Reshape" + input: "bert/embeddings/Slice/begin" + input: "gradients/bert/embeddings/Slice_grad/stack" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Shape_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\002\000\000\000\003\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/sub" + op: "Sub" + input: "gradients/bert/embeddings/Slice_grad/Shape_1" + input: "gradients/bert/embeddings/Slice_grad/Shape" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/sub_1" + op: "Sub" + input: "gradients/bert/embeddings/Slice_grad/sub" + input: "bert/embeddings/Slice/begin" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Reshape_1" + op: "Reshape" + input: "gradients/bert/embeddings/Slice_grad/sub_1" + input: "gradients/bert/embeddings/Slice_grad/stack" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/concat/axis" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/concat" + op: "ConcatV2" + input: "gradients/bert/embeddings/Slice_grad/Reshape" + input: "gradients/bert/embeddings/Slice_grad/Reshape_1" + input: "gradients/bert/embeddings/Slice_grad/concat/axis" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/Slice_grad/Pad" + op: "Pad" + input: "gradients/bert/embeddings/Reshape_4_grad/Reshape" + input: "gradients/bert/embeddings/Slice_grad/concat" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tpaddings" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/Shape" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\210R\000\000\000\000\000\000\000\003\000\000\000\000\000\000" + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/Cast" + op: "Cast" + input: "gradients/bert/embeddings/GatherV2_grad/Shape" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/Size" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 4096 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/ExpandDims/dim" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/ExpandDims" + op: "ExpandDims" + input: "gradients/bert/embeddings/GatherV2_grad/Size" + input: "gradients/bert/embeddings/GatherV2_grad/ExpandDims/dim" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tdim" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/strided_slice" + op: "StridedSlice" + input: "gradients/bert/embeddings/GatherV2_grad/Cast" + input: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack" + input: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack_1" + input: "gradients/bert/embeddings/GatherV2_grad/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 1 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/concat/axis" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/concat" + op: "ConcatV2" + input: "gradients/bert/embeddings/GatherV2_grad/ExpandDims" + input: "gradients/bert/embeddings/GatherV2_grad/strided_slice" + input: "gradients/bert/embeddings/GatherV2_grad/concat/axis" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/Reshape" + op: "Reshape" + input: "gradients/bert/embeddings/Reshape_1_grad/Reshape" + input: "gradients/bert/embeddings/GatherV2_grad/concat" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/GatherV2_grad/Reshape_1" + op: "Reshape" + input: "bert/embeddings/Reshape" + input: "gradients/bert/embeddings/GatherV2_grad/ExpandDims" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + } + } + } + } +} +node { + name: "gradients/bert/embeddings/MatMul_grad/MatMul" + op: "MatMul" + input: "gradients/bert/embeddings/Reshape_3_grad/Reshape" + input: "bert/embeddings/token_type_embeddings/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 2 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: false + } + } + attr { + key: "transpose_b" + value { + b: true + } + } +} +node { + name: "gradients/bert/embeddings/MatMul_grad/MatMul_1" + op: "MatMul" + input: "bert/embeddings/one_hot" + input: "gradients/bert/embeddings/Reshape_3_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "transpose_a" + value { + b: true + } + } + attr { + key: "transpose_b" + value { + b: false + } + } +} +node { + name: "global_norm/L2Loss" + op: "L2Loss" + input: "gradients/bert/embeddings/GatherV2_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/GatherV2_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_1" + op: "L2Loss" + input: "gradients/bert/embeddings/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_2" + op: "L2Loss" + input: "gradients/bert/embeddings/Slice_grad/Pad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/Slice_grad/Pad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_3" + op: "L2Loss" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_4" + op: "L2Loss" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_5" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_6" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_7" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_8" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_9" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_10" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_11" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_12" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_13" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_14" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_15" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_16" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_17" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_18" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_19" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_20" + op: "L2Loss" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_21" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_22" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_23" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_24" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_25" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_26" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_27" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_28" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_29" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_30" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_31" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_32" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_33" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_34" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_35" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_36" + op: "L2Loss" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_37" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_38" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_39" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_40" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_41" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_42" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_43" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_44" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_45" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_46" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_47" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_48" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_49" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_50" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_51" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_52" + op: "L2Loss" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_53" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_54" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_55" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_56" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_57" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_58" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_59" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_60" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_61" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_62" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_63" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_64" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_65" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_66" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_67" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_68" + op: "L2Loss" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_69" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_70" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_71" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_72" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_73" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_74" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_75" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_76" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_77" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_78" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_79" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_80" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_81" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_82" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_83" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_84" + op: "L2Loss" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_85" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_86" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_87" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_88" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_89" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_90" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_91" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_92" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_93" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_94" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_95" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_96" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_97" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_98" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_99" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_100" + op: "L2Loss" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_101" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_102" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_103" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_104" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_105" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_106" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_107" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_108" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_109" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_110" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_111" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_112" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_113" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_114" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_115" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_116" + op: "L2Loss" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_117" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_118" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_119" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_120" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_121" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_122" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_123" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_124" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_125" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_126" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_127" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_128" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_129" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_130" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_131" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_132" + op: "L2Loss" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_133" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_134" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_135" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_136" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_137" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_138" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_139" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_140" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_141" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_142" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_143" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_144" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_145" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_146" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_147" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_148" + op: "L2Loss" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_149" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_150" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_151" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_152" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_153" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_154" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_155" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_156" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_157" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_158" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_159" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_160" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_161" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_162" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_163" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_164" + op: "L2Loss" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_165" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_166" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_167" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_168" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_169" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_170" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_171" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_172" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_173" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_174" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_175" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_176" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_177" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_178" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_179" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_180" + op: "L2Loss" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_181" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_182" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_183" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_184" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_185" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_186" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_187" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_188" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_189" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_190" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_191" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_192" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_193" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_194" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_195" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_196" + op: "L2Loss" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_197" + op: "L2Loss" + input: "gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_198" + op: "L2Loss" + input: "gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_199" + op: "L2Loss" + input: "gradients/loss/MatMul_grad/MatMul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/L2Loss_200" + op: "L2Loss" + input: "gradients/loss/BiasAdd_grad/BiasAddGrad" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/stack" + op: "Pack" + input: "global_norm/L2Loss" + input: "global_norm/L2Loss_1" + input: "global_norm/L2Loss_2" + input: "global_norm/L2Loss_3" + input: "global_norm/L2Loss_4" + input: "global_norm/L2Loss_5" + input: "global_norm/L2Loss_6" + input: "global_norm/L2Loss_7" + input: "global_norm/L2Loss_8" + input: "global_norm/L2Loss_9" + input: "global_norm/L2Loss_10" + input: "global_norm/L2Loss_11" + input: "global_norm/L2Loss_12" + input: "global_norm/L2Loss_13" + input: "global_norm/L2Loss_14" + input: "global_norm/L2Loss_15" + input: "global_norm/L2Loss_16" + input: "global_norm/L2Loss_17" + input: "global_norm/L2Loss_18" + input: "global_norm/L2Loss_19" + input: "global_norm/L2Loss_20" + input: "global_norm/L2Loss_21" + input: "global_norm/L2Loss_22" + input: "global_norm/L2Loss_23" + input: "global_norm/L2Loss_24" + input: "global_norm/L2Loss_25" + input: "global_norm/L2Loss_26" + input: "global_norm/L2Loss_27" + input: "global_norm/L2Loss_28" + input: "global_norm/L2Loss_29" + input: "global_norm/L2Loss_30" + input: "global_norm/L2Loss_31" + input: "global_norm/L2Loss_32" + input: "global_norm/L2Loss_33" + input: "global_norm/L2Loss_34" + input: "global_norm/L2Loss_35" + input: "global_norm/L2Loss_36" + input: "global_norm/L2Loss_37" + input: "global_norm/L2Loss_38" + input: "global_norm/L2Loss_39" + input: "global_norm/L2Loss_40" + input: "global_norm/L2Loss_41" + input: "global_norm/L2Loss_42" + input: "global_norm/L2Loss_43" + input: "global_norm/L2Loss_44" + input: "global_norm/L2Loss_45" + input: "global_norm/L2Loss_46" + input: "global_norm/L2Loss_47" + input: "global_norm/L2Loss_48" + input: "global_norm/L2Loss_49" + input: "global_norm/L2Loss_50" + input: "global_norm/L2Loss_51" + input: "global_norm/L2Loss_52" + input: "global_norm/L2Loss_53" + input: "global_norm/L2Loss_54" + input: "global_norm/L2Loss_55" + input: "global_norm/L2Loss_56" + input: "global_norm/L2Loss_57" + input: "global_norm/L2Loss_58" + input: "global_norm/L2Loss_59" + input: "global_norm/L2Loss_60" + input: "global_norm/L2Loss_61" + input: "global_norm/L2Loss_62" + input: "global_norm/L2Loss_63" + input: "global_norm/L2Loss_64" + input: "global_norm/L2Loss_65" + input: "global_norm/L2Loss_66" + input: "global_norm/L2Loss_67" + input: "global_norm/L2Loss_68" + input: "global_norm/L2Loss_69" + input: "global_norm/L2Loss_70" + input: "global_norm/L2Loss_71" + input: "global_norm/L2Loss_72" + input: "global_norm/L2Loss_73" + input: "global_norm/L2Loss_74" + input: "global_norm/L2Loss_75" + input: "global_norm/L2Loss_76" + input: "global_norm/L2Loss_77" + input: "global_norm/L2Loss_78" + input: "global_norm/L2Loss_79" + input: "global_norm/L2Loss_80" + input: "global_norm/L2Loss_81" + input: "global_norm/L2Loss_82" + input: "global_norm/L2Loss_83" + input: "global_norm/L2Loss_84" + input: "global_norm/L2Loss_85" + input: "global_norm/L2Loss_86" + input: "global_norm/L2Loss_87" + input: "global_norm/L2Loss_88" + input: "global_norm/L2Loss_89" + input: "global_norm/L2Loss_90" + input: "global_norm/L2Loss_91" + input: "global_norm/L2Loss_92" + input: "global_norm/L2Loss_93" + input: "global_norm/L2Loss_94" + input: "global_norm/L2Loss_95" + input: "global_norm/L2Loss_96" + input: "global_norm/L2Loss_97" + input: "global_norm/L2Loss_98" + input: "global_norm/L2Loss_99" + input: "global_norm/L2Loss_100" + input: "global_norm/L2Loss_101" + input: "global_norm/L2Loss_102" + input: "global_norm/L2Loss_103" + input: "global_norm/L2Loss_104" + input: "global_norm/L2Loss_105" + input: "global_norm/L2Loss_106" + input: "global_norm/L2Loss_107" + input: "global_norm/L2Loss_108" + input: "global_norm/L2Loss_109" + input: "global_norm/L2Loss_110" + input: "global_norm/L2Loss_111" + input: "global_norm/L2Loss_112" + input: "global_norm/L2Loss_113" + input: "global_norm/L2Loss_114" + input: "global_norm/L2Loss_115" + input: "global_norm/L2Loss_116" + input: "global_norm/L2Loss_117" + input: "global_norm/L2Loss_118" + input: "global_norm/L2Loss_119" + input: "global_norm/L2Loss_120" + input: "global_norm/L2Loss_121" + input: "global_norm/L2Loss_122" + input: "global_norm/L2Loss_123" + input: "global_norm/L2Loss_124" + input: "global_norm/L2Loss_125" + input: "global_norm/L2Loss_126" + input: "global_norm/L2Loss_127" + input: "global_norm/L2Loss_128" + input: "global_norm/L2Loss_129" + input: "global_norm/L2Loss_130" + input: "global_norm/L2Loss_131" + input: "global_norm/L2Loss_132" + input: "global_norm/L2Loss_133" + input: "global_norm/L2Loss_134" + input: "global_norm/L2Loss_135" + input: "global_norm/L2Loss_136" + input: "global_norm/L2Loss_137" + input: "global_norm/L2Loss_138" + input: "global_norm/L2Loss_139" + input: "global_norm/L2Loss_140" + input: "global_norm/L2Loss_141" + input: "global_norm/L2Loss_142" + input: "global_norm/L2Loss_143" + input: "global_norm/L2Loss_144" + input: "global_norm/L2Loss_145" + input: "global_norm/L2Loss_146" + input: "global_norm/L2Loss_147" + input: "global_norm/L2Loss_148" + input: "global_norm/L2Loss_149" + input: "global_norm/L2Loss_150" + input: "global_norm/L2Loss_151" + input: "global_norm/L2Loss_152" + input: "global_norm/L2Loss_153" + input: "global_norm/L2Loss_154" + input: "global_norm/L2Loss_155" + input: "global_norm/L2Loss_156" + input: "global_norm/L2Loss_157" + input: "global_norm/L2Loss_158" + input: "global_norm/L2Loss_159" + input: "global_norm/L2Loss_160" + input: "global_norm/L2Loss_161" + input: "global_norm/L2Loss_162" + input: "global_norm/L2Loss_163" + input: "global_norm/L2Loss_164" + input: "global_norm/L2Loss_165" + input: "global_norm/L2Loss_166" + input: "global_norm/L2Loss_167" + input: "global_norm/L2Loss_168" + input: "global_norm/L2Loss_169" + input: "global_norm/L2Loss_170" + input: "global_norm/L2Loss_171" + input: "global_norm/L2Loss_172" + input: "global_norm/L2Loss_173" + input: "global_norm/L2Loss_174" + input: "global_norm/L2Loss_175" + input: "global_norm/L2Loss_176" + input: "global_norm/L2Loss_177" + input: "global_norm/L2Loss_178" + input: "global_norm/L2Loss_179" + input: "global_norm/L2Loss_180" + input: "global_norm/L2Loss_181" + input: "global_norm/L2Loss_182" + input: "global_norm/L2Loss_183" + input: "global_norm/L2Loss_184" + input: "global_norm/L2Loss_185" + input: "global_norm/L2Loss_186" + input: "global_norm/L2Loss_187" + input: "global_norm/L2Loss_188" + input: "global_norm/L2Loss_189" + input: "global_norm/L2Loss_190" + input: "global_norm/L2Loss_191" + input: "global_norm/L2Loss_192" + input: "global_norm/L2Loss_193" + input: "global_norm/L2Loss_194" + input: "global_norm/L2Loss_195" + input: "global_norm/L2Loss_196" + input: "global_norm/L2Loss_197" + input: "global_norm/L2Loss_198" + input: "global_norm/L2Loss_199" + input: "global_norm/L2Loss_200" + attr { + key: "N" + value { + i: 201 + } + } + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 201 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "global_norm/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "global_norm/Sum" + op: "Sum" + input: "global_norm/stack" + input: "global_norm/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "global_norm/Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 2.0 + } + } + } +} +node { + name: "global_norm/mul" + op: "Mul" + input: "global_norm/Sum" + input: "global_norm/Const_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "global_norm/global_norm" + op: "Sqrt" + input: "global_norm/mul" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/truediv/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "clip_by_global_norm/truediv" + op: "RealDiv" + input: "clip_by_global_norm/truediv/x" + input: "global_norm/global_norm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "clip_by_global_norm/truediv_1/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "clip_by_global_norm/truediv_1" + op: "RealDiv" + input: "clip_by_global_norm/Const" + input: "clip_by_global_norm/truediv_1/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/Minimum" + op: "Minimum" + input: "clip_by_global_norm/truediv" + input: "clip_by_global_norm/truediv_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/mul/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 1.0 + } + } + } +} +node { + name: "clip_by_global_norm/mul" + op: "Mul" + input: "clip_by_global_norm/mul/x" + input: "clip_by_global_norm/Minimum" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/IsFinite" + op: "IsFinite" + input: "global_norm/global_norm" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: nan + } + } + } +} +node { + name: "clip_by_global_norm/Select" + op: "Select" + input: "clip_by_global_norm/IsFinite" + input: "clip_by_global_norm/mul" + input: "clip_by_global_norm/Const_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_1" + op: "Mul" + input: "gradients/bert/embeddings/GatherV2_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/GatherV2_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_0" + op: "Identity" + input: "clip_by_global_norm/mul_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/GatherV2_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 4096 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_2" + op: "Mul" + input: "gradients/bert/embeddings/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_1" + op: "Identity" + input: "clip_by_global_norm/mul_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_3" + op: "Mul" + input: "gradients/bert/embeddings/Slice_grad/Pad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/Slice_grad/Pad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_2" + op: "Identity" + input: "clip_by_global_norm/mul_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/Slice_grad/Pad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_4" + op: "Mul" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_3" + op: "Identity" + input: "clip_by_global_norm/mul_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_5" + op: "Mul" + input: "gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_4" + op: "Identity" + input: "clip_by_global_norm/mul_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/embeddings/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_6" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_5" + op: "Identity" + input: "clip_by_global_norm/mul_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_7" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_6" + op: "Identity" + input: "clip_by_global_norm/mul_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_8" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_7" + op: "Identity" + input: "clip_by_global_norm/mul_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_9" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_8" + op: "Identity" + input: "clip_by_global_norm/mul_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_10" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_9" + op: "Identity" + input: "clip_by_global_norm/mul_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_11" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_10" + op: "Identity" + input: "clip_by_global_norm/mul_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_12" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_11" + op: "Identity" + input: "clip_by_global_norm/mul_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_13" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_12" + op: "Identity" + input: "clip_by_global_norm/mul_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_14" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_13" + op: "Identity" + input: "clip_by_global_norm/mul_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_15" + op: "Mul" + input: "gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_14" + op: "Identity" + input: "clip_by_global_norm/mul_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_16" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_15" + op: "Identity" + input: "clip_by_global_norm/mul_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_17" + op: "Mul" + input: "gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_16" + op: "Identity" + input: "clip_by_global_norm/mul_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_18" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_17" + op: "Identity" + input: "clip_by_global_norm/mul_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_19" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_18" + op: "Identity" + input: "clip_by_global_norm/mul_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_20" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_19" + op: "Identity" + input: "clip_by_global_norm/mul_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_21" + op: "Mul" + input: "gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_20" + op: "Identity" + input: "clip_by_global_norm/mul_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_0/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_22" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_21" + op: "Identity" + input: "clip_by_global_norm/mul_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_23" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_22" + op: "Identity" + input: "clip_by_global_norm/mul_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_24" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_23" + op: "Identity" + input: "clip_by_global_norm/mul_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_25" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_24" + op: "Identity" + input: "clip_by_global_norm/mul_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_26" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_25" + op: "Identity" + input: "clip_by_global_norm/mul_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_27" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_26" + op: "Identity" + input: "clip_by_global_norm/mul_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_28" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_27" + op: "Identity" + input: "clip_by_global_norm/mul_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_29" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_28" + op: "Identity" + input: "clip_by_global_norm/mul_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_30" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_29" + op: "Identity" + input: "clip_by_global_norm/mul_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_31" + op: "Mul" + input: "gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_30" + op: "Identity" + input: "clip_by_global_norm/mul_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_32" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_31" + op: "Identity" + input: "clip_by_global_norm/mul_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_33" + op: "Mul" + input: "gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_32" + op: "Identity" + input: "clip_by_global_norm/mul_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_34" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_33" + op: "Identity" + input: "clip_by_global_norm/mul_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_35" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_34" + op: "Identity" + input: "clip_by_global_norm/mul_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_36" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_35" + op: "Identity" + input: "clip_by_global_norm/mul_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_37" + op: "Mul" + input: "gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_36" + op: "Identity" + input: "clip_by_global_norm/mul_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_1/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_38" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_37" + op: "Identity" + input: "clip_by_global_norm/mul_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_39" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_38" + op: "Identity" + input: "clip_by_global_norm/mul_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_40" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_39" + op: "Identity" + input: "clip_by_global_norm/mul_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_41" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_40" + op: "Identity" + input: "clip_by_global_norm/mul_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_42" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_41" + op: "Identity" + input: "clip_by_global_norm/mul_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_43" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_42" + op: "Identity" + input: "clip_by_global_norm/mul_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_44" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_43" + op: "Identity" + input: "clip_by_global_norm/mul_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_45" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_44" + op: "Identity" + input: "clip_by_global_norm/mul_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_46" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_45" + op: "Identity" + input: "clip_by_global_norm/mul_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_47" + op: "Mul" + input: "gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_46" + op: "Identity" + input: "clip_by_global_norm/mul_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_48" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_47" + op: "Identity" + input: "clip_by_global_norm/mul_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_49" + op: "Mul" + input: "gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_48" + op: "Identity" + input: "clip_by_global_norm/mul_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_50" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_49" + op: "Identity" + input: "clip_by_global_norm/mul_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_51" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_50" + op: "Identity" + input: "clip_by_global_norm/mul_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_52" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_51" + op: "Identity" + input: "clip_by_global_norm/mul_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_53" + op: "Mul" + input: "gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_52" + op: "Identity" + input: "clip_by_global_norm/mul_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_2/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_54" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_53" + op: "Identity" + input: "clip_by_global_norm/mul_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_55" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_54" + op: "Identity" + input: "clip_by_global_norm/mul_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_56" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_55" + op: "Identity" + input: "clip_by_global_norm/mul_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_57" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_56" + op: "Identity" + input: "clip_by_global_norm/mul_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_58" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_57" + op: "Identity" + input: "clip_by_global_norm/mul_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_59" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_58" + op: "Identity" + input: "clip_by_global_norm/mul_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_60" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_59" + op: "Identity" + input: "clip_by_global_norm/mul_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_61" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_60" + op: "Identity" + input: "clip_by_global_norm/mul_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_62" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_61" + op: "Identity" + input: "clip_by_global_norm/mul_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_63" + op: "Mul" + input: "gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_62" + op: "Identity" + input: "clip_by_global_norm/mul_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_64" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_63" + op: "Identity" + input: "clip_by_global_norm/mul_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_65" + op: "Mul" + input: "gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_64" + op: "Identity" + input: "clip_by_global_norm/mul_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_66" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_65" + op: "Identity" + input: "clip_by_global_norm/mul_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_67" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_66" + op: "Identity" + input: "clip_by_global_norm/mul_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_68" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_67" + op: "Identity" + input: "clip_by_global_norm/mul_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_69" + op: "Mul" + input: "gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_68" + op: "Identity" + input: "clip_by_global_norm/mul_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_3/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_70" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_69" + op: "Identity" + input: "clip_by_global_norm/mul_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_71" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_70" + op: "Identity" + input: "clip_by_global_norm/mul_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_72" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_71" + op: "Identity" + input: "clip_by_global_norm/mul_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_73" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_72" + op: "Identity" + input: "clip_by_global_norm/mul_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_74" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_73" + op: "Identity" + input: "clip_by_global_norm/mul_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_75" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_74" + op: "Identity" + input: "clip_by_global_norm/mul_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_76" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_75" + op: "Identity" + input: "clip_by_global_norm/mul_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_77" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_76" + op: "Identity" + input: "clip_by_global_norm/mul_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_78" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_77" + op: "Identity" + input: "clip_by_global_norm/mul_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_79" + op: "Mul" + input: "gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_78" + op: "Identity" + input: "clip_by_global_norm/mul_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_80" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_79" + op: "Identity" + input: "clip_by_global_norm/mul_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_81" + op: "Mul" + input: "gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_80" + op: "Identity" + input: "clip_by_global_norm/mul_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_82" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_81" + op: "Identity" + input: "clip_by_global_norm/mul_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_83" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_82" + op: "Identity" + input: "clip_by_global_norm/mul_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_84" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_83" + op: "Identity" + input: "clip_by_global_norm/mul_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_85" + op: "Mul" + input: "gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_84" + op: "Identity" + input: "clip_by_global_norm/mul_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_4/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_86" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_85" + op: "Identity" + input: "clip_by_global_norm/mul_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_87" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_86" + op: "Identity" + input: "clip_by_global_norm/mul_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_88" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_87" + op: "Identity" + input: "clip_by_global_norm/mul_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_89" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_88" + op: "Identity" + input: "clip_by_global_norm/mul_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_90" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_89" + op: "Identity" + input: "clip_by_global_norm/mul_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_91" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_90" + op: "Identity" + input: "clip_by_global_norm/mul_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_92" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_91" + op: "Identity" + input: "clip_by_global_norm/mul_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_93" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_92" + op: "Identity" + input: "clip_by_global_norm/mul_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_94" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_93" + op: "Identity" + input: "clip_by_global_norm/mul_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_95" + op: "Mul" + input: "gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_94" + op: "Identity" + input: "clip_by_global_norm/mul_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_96" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_95" + op: "Identity" + input: "clip_by_global_norm/mul_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_97" + op: "Mul" + input: "gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_96" + op: "Identity" + input: "clip_by_global_norm/mul_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_98" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_97" + op: "Identity" + input: "clip_by_global_norm/mul_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_99" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_98" + op: "Identity" + input: "clip_by_global_norm/mul_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_100" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_99" + op: "Identity" + input: "clip_by_global_norm/mul_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_101" + op: "Mul" + input: "gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_100" + op: "Identity" + input: "clip_by_global_norm/mul_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_5/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_102" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_101" + op: "Identity" + input: "clip_by_global_norm/mul_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_103" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_102" + op: "Identity" + input: "clip_by_global_norm/mul_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_104" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_103" + op: "Identity" + input: "clip_by_global_norm/mul_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_105" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_104" + op: "Identity" + input: "clip_by_global_norm/mul_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_106" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_105" + op: "Identity" + input: "clip_by_global_norm/mul_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_107" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_106" + op: "Identity" + input: "clip_by_global_norm/mul_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_108" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_107" + op: "Identity" + input: "clip_by_global_norm/mul_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_109" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_108" + op: "Identity" + input: "clip_by_global_norm/mul_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_110" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_109" + op: "Identity" + input: "clip_by_global_norm/mul_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_111" + op: "Mul" + input: "gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_110" + op: "Identity" + input: "clip_by_global_norm/mul_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_112" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_111" + op: "Identity" + input: "clip_by_global_norm/mul_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_113" + op: "Mul" + input: "gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_112" + op: "Identity" + input: "clip_by_global_norm/mul_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_114" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_113" + op: "Identity" + input: "clip_by_global_norm/mul_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_115" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_114" + op: "Identity" + input: "clip_by_global_norm/mul_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_116" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_115" + op: "Identity" + input: "clip_by_global_norm/mul_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_117" + op: "Mul" + input: "gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_116" + op: "Identity" + input: "clip_by_global_norm/mul_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_6/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_118" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_117" + op: "Identity" + input: "clip_by_global_norm/mul_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_119" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_118" + op: "Identity" + input: "clip_by_global_norm/mul_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_120" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_119" + op: "Identity" + input: "clip_by_global_norm/mul_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_121" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_120" + op: "Identity" + input: "clip_by_global_norm/mul_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_122" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_121" + op: "Identity" + input: "clip_by_global_norm/mul_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_123" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_122" + op: "Identity" + input: "clip_by_global_norm/mul_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_124" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_123" + op: "Identity" + input: "clip_by_global_norm/mul_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_125" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_124" + op: "Identity" + input: "clip_by_global_norm/mul_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_126" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_125" + op: "Identity" + input: "clip_by_global_norm/mul_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_127" + op: "Mul" + input: "gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_126" + op: "Identity" + input: "clip_by_global_norm/mul_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_128" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_127" + op: "Identity" + input: "clip_by_global_norm/mul_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_129" + op: "Mul" + input: "gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_128" + op: "Identity" + input: "clip_by_global_norm/mul_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_130" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_129" + op: "Identity" + input: "clip_by_global_norm/mul_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_131" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_130" + op: "Identity" + input: "clip_by_global_norm/mul_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_132" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_131" + op: "Identity" + input: "clip_by_global_norm/mul_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_133" + op: "Mul" + input: "gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_132" + op: "Identity" + input: "clip_by_global_norm/mul_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_7/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_134" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_133" + op: "Identity" + input: "clip_by_global_norm/mul_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_135" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_134" + op: "Identity" + input: "clip_by_global_norm/mul_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_136" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_135" + op: "Identity" + input: "clip_by_global_norm/mul_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_137" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_136" + op: "Identity" + input: "clip_by_global_norm/mul_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_138" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_137" + op: "Identity" + input: "clip_by_global_norm/mul_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_139" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_138" + op: "Identity" + input: "clip_by_global_norm/mul_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_140" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_139" + op: "Identity" + input: "clip_by_global_norm/mul_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_141" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_140" + op: "Identity" + input: "clip_by_global_norm/mul_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_142" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_141" + op: "Identity" + input: "clip_by_global_norm/mul_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_143" + op: "Mul" + input: "gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_142" + op: "Identity" + input: "clip_by_global_norm/mul_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_144" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_143" + op: "Identity" + input: "clip_by_global_norm/mul_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_145" + op: "Mul" + input: "gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_144" + op: "Identity" + input: "clip_by_global_norm/mul_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_146" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_145" + op: "Identity" + input: "clip_by_global_norm/mul_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_147" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_146" + op: "Identity" + input: "clip_by_global_norm/mul_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_148" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_147" + op: "Identity" + input: "clip_by_global_norm/mul_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_149" + op: "Mul" + input: "gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_148" + op: "Identity" + input: "clip_by_global_norm/mul_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_8/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_150" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_149" + op: "Identity" + input: "clip_by_global_norm/mul_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_151" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_150" + op: "Identity" + input: "clip_by_global_norm/mul_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_152" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_151" + op: "Identity" + input: "clip_by_global_norm/mul_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_153" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_152" + op: "Identity" + input: "clip_by_global_norm/mul_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_154" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_153" + op: "Identity" + input: "clip_by_global_norm/mul_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_155" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_154" + op: "Identity" + input: "clip_by_global_norm/mul_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_156" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_155" + op: "Identity" + input: "clip_by_global_norm/mul_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_157" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_156" + op: "Identity" + input: "clip_by_global_norm/mul_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_158" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_157" + op: "Identity" + input: "clip_by_global_norm/mul_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_159" + op: "Mul" + input: "gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_158" + op: "Identity" + input: "clip_by_global_norm/mul_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_160" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_159" + op: "Identity" + input: "clip_by_global_norm/mul_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_161" + op: "Mul" + input: "gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_160" + op: "Identity" + input: "clip_by_global_norm/mul_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_162" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_161" + op: "Identity" + input: "clip_by_global_norm/mul_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_163" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_162" + op: "Identity" + input: "clip_by_global_norm/mul_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_164" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_163" + op: "Identity" + input: "clip_by_global_norm/mul_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_165" + op: "Mul" + input: "gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_164" + op: "Identity" + input: "clip_by_global_norm/mul_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_9/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_166" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_165" + op: "Identity" + input: "clip_by_global_norm/mul_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_167" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_166" + op: "Identity" + input: "clip_by_global_norm/mul_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_168" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_167" + op: "Identity" + input: "clip_by_global_norm/mul_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_169" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_168" + op: "Identity" + input: "clip_by_global_norm/mul_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_170" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_169" + op: "Identity" + input: "clip_by_global_norm/mul_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_171" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_170" + op: "Identity" + input: "clip_by_global_norm/mul_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_172" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_171" + op: "Identity" + input: "clip_by_global_norm/mul_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_173" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_172" + op: "Identity" + input: "clip_by_global_norm/mul_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_174" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_173" + op: "Identity" + input: "clip_by_global_norm/mul_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_175" + op: "Mul" + input: "gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_174" + op: "Identity" + input: "clip_by_global_norm/mul_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_176" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_175" + op: "Identity" + input: "clip_by_global_norm/mul_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_177" + op: "Mul" + input: "gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_176" + op: "Identity" + input: "clip_by_global_norm/mul_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_178" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_177" + op: "Identity" + input: "clip_by_global_norm/mul_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_179" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_178" + op: "Identity" + input: "clip_by_global_norm/mul_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_180" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_179" + op: "Identity" + input: "clip_by_global_norm/mul_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_181" + op: "Mul" + input: "gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_180" + op: "Identity" + input: "clip_by_global_norm/mul_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_10/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_182" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_181" + op: "Identity" + input: "clip_by_global_norm/mul_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_183" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_182" + op: "Identity" + input: "clip_by_global_norm/mul_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/query/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_184" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_183" + op: "Identity" + input: "clip_by_global_norm/mul_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_185" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_184" + op: "Identity" + input: "clip_by_global_norm/mul_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/key/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_186" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_185" + op: "Identity" + input: "clip_by_global_norm/mul_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_187" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_186" + op: "Identity" + input: "clip_by_global_norm/mul_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/self/value/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_188" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_187" + op: "Identity" + input: "clip_by_global_norm/mul_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_189" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_188" + op: "Identity" + input: "clip_by_global_norm/mul_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_190" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_189" + op: "Identity" + input: "clip_by_global_norm/mul_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_191" + op: "Mul" + input: "gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_190" + op: "Identity" + input: "clip_by_global_norm/mul_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/attention/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_192" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_191" + op: "Identity" + input: "clip_by_global_norm/mul_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_193" + op: "Mul" + input: "gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_192" + op: "Identity" + input: "clip_by_global_norm/mul_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/intermediate/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_194" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_193" + op: "Identity" + input: "clip_by_global_norm/mul_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_195" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_194" + op: "Identity" + input: "clip_by_global_norm/mul_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_196" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_195" + op: "Identity" + input: "clip_by_global_norm/mul_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/sub_grad/Reshape" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_197" + op: "Mul" + input: "gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_196" + op: "Identity" + input: "clip_by_global_norm/mul_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/encoder/layer_11/output/LayerNorm/batchnorm/mul_grad/Reshape_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_198" + op: "Mul" + input: "gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_197" + op: "Identity" + input: "clip_by_global_norm/mul_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_199" + op: "Mul" + input: "gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_198" + op: "Identity" + input: "clip_by_global_norm/mul_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/bert/pooler/dense/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_200" + op: "Mul" + input: "gradients/loss/MatMul_grad/MatMul_1" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_199" + op: "Identity" + input: "clip_by_global_norm/mul_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/MatMul_grad/MatMul_1" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/mul_201" + op: "Mul" + input: "gradients/loss/BiasAdd_grad/BiasAddGrad" + input: "clip_by_global_norm/Select" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "clip_by_global_norm/clip_by_global_norm/_200" + op: "Identity" + input: "clip_by_global_norm/mul_201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@gradients/loss/BiasAdd_grad/BiasAddGrad" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\210R\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m/Assign" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_m" + input: "bert/embeddings/word_embeddings/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_m/read" + op: "Identity" + input: "bert/embeddings/word_embeddings/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\210R\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v/Assign" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_v" + input: "bert/embeddings/word_embeddings/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/word_embeddings/adam_v/read" + op: "Identity" + input: "bert/embeddings/word_embeddings/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_3/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_3" + op: "Mul" + input: "Mul_3/x" + input: "bert/embeddings/word_embeddings/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_4/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_4/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "Mul_4/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "Mul_4/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "Mul_4/strided_slice" + op: "StridedSlice" + input: "gradients/bert/embeddings/GatherV2_grad/Cast" + input: "Mul_4/strided_slice/stack" + input: "Mul_4/strided_slice/stack_1" + input: "Mul_4/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 1 + } + } +} +node { + name: "Mul_4/y" + op: "UnsortedSegmentSum" + input: "clip_by_global_norm/clip_by_global_norm/_0" + input: "gradients/bert/embeddings/GatherV2_grad/Reshape_1" + input: "Mul_4/strided_slice" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tindices" + value { + type: DT_INT32 + } + } + attr { + key: "Tnumsegments" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_4" + op: "Mul" + input: "Mul_4/x" + input: "Mul_4/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_1" + op: "Add" + input: "Mul_3" + input: "Mul_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_5/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_5" + op: "Mul" + input: "Mul_5/x" + input: "bert/embeddings/word_embeddings/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square/strided_slice/stack" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "Square/strided_slice/stack_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "Square/strided_slice/stack_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "Square/strided_slice" + op: "StridedSlice" + input: "gradients/bert/embeddings/GatherV2_grad/Cast" + input: "Square/strided_slice/stack" + input: "Square/strided_slice/stack_1" + input: "Square/strided_slice/stack_2" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 1 + } + } +} +node { + name: "Square/x" + op: "UnsortedSegmentSum" + input: "clip_by_global_norm/clip_by_global_norm/_0" + input: "gradients/bert/embeddings/GatherV2_grad/Reshape_1" + input: "Square/strided_slice" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "Tindices" + value { + type: DT_INT32 + } + } + attr { + key: "Tnumsegments" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square" + op: "Square" + input: "Square/x" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_6/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_6" + op: "Mul" + input: "Mul_6/x" + input: "Square" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_2" + op: "Add" + input: "Mul_5" + input: "Mul_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt" + op: "Sqrt" + input: "add_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_3/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_3" + op: "Add" + input: "Sqrt" + input: "add_3/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_1" + op: "RealDiv" + input: "add_1" + input: "add_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_7/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_7" + op: "Mul" + input: "mul_7/x" + input: "bert/embeddings/word_embeddings/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_4" + op: "Add" + input: "truediv_1" + input: "mul_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_8" + op: "Mul" + input: "add" + input: "add_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_1" + op: "Sub" + input: "bert/embeddings/word_embeddings/read" + input: "mul_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_199" + op: "Assign" + input: "bert/embeddings/word_embeddings" + input: "sub_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_200" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_m" + input: "add_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_201" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_v" + input: "add_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\002\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m/Assign" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_m" + input: "bert/embeddings/token_type_embeddings/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_m/read" + op: "Identity" + input: "bert/embeddings/token_type_embeddings/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\002\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v/Assign" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_v" + input: "bert/embeddings/token_type_embeddings/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/token_type_embeddings/adam_v/read" + op: "Identity" + input: "bert/embeddings/token_type_embeddings/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_9/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_9" + op: "Mul" + input: "Mul_9/x" + input: "bert/embeddings/token_type_embeddings/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_10/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_10" + op: "Mul" + input: "Mul_10/x" + input: "clip_by_global_norm/clip_by_global_norm/_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_5" + op: "Add" + input: "Mul_9" + input: "Mul_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_11/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_11" + op: "Mul" + input: "Mul_11/x" + input: "bert/embeddings/token_type_embeddings/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_1" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_12/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_12" + op: "Mul" + input: "Mul_12/x" + input: "Square_1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_6" + op: "Add" + input: "Mul_11" + input: "Mul_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_1" + op: "Sqrt" + input: "add_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_7/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_7" + op: "Add" + input: "Sqrt_1" + input: "add_7/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_2" + op: "RealDiv" + input: "add_5" + input: "add_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_13/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_13" + op: "Mul" + input: "mul_13/x" + input: "bert/embeddings/token_type_embeddings/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_8" + op: "Add" + input: "truediv_2" + input: "mul_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_14" + op: "Mul" + input: "add" + input: "add_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_2" + op: "Sub" + input: "bert/embeddings/token_type_embeddings/read" + input: "mul_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_202" + op: "Assign" + input: "bert/embeddings/token_type_embeddings" + input: "sub_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_203" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_m" + input: "add_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_204" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_v" + input: "add_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\002\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m/Assign" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_m" + input: "bert/embeddings/position_embeddings/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_m/read" + op: "Identity" + input: "bert/embeddings/position_embeddings/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\002\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v/Assign" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_v" + input: "bert/embeddings/position_embeddings/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/position_embeddings/adam_v/read" + op: "Identity" + input: "bert/embeddings/position_embeddings/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_15/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_15" + op: "Mul" + input: "Mul_15/x" + input: "bert/embeddings/position_embeddings/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_16/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_16" + op: "Mul" + input: "Mul_16/x" + input: "clip_by_global_norm/clip_by_global_norm/_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_9" + op: "Add" + input: "Mul_15" + input: "Mul_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_17/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_17" + op: "Mul" + input: "Mul_17/x" + input: "bert/embeddings/position_embeddings/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_2" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_18/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_18" + op: "Mul" + input: "Mul_18/x" + input: "Square_2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_10" + op: "Add" + input: "Mul_17" + input: "Mul_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_2" + op: "Sqrt" + input: "add_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_11/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_11" + op: "Add" + input: "Sqrt_2" + input: "add_11/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_3" + op: "RealDiv" + input: "add_9" + input: "add_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_19/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_19" + op: "Mul" + input: "mul_19/x" + input: "bert/embeddings/position_embeddings/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_12" + op: "Add" + input: "truediv_3" + input: "mul_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_20" + op: "Mul" + input: "add" + input: "add_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_3" + op: "Sub" + input: "bert/embeddings/position_embeddings/read" + input: "mul_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_205" + op: "Assign" + input: "bert/embeddings/position_embeddings" + input: "sub_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_206" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_m" + input: "add_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_207" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_v" + input: "add_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_m" + input: "bert/embeddings/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_v" + input: "bert/embeddings/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_21/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_21" + op: "Mul" + input: "Mul_21/x" + input: "bert/embeddings/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_22/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_22" + op: "Mul" + input: "Mul_22/x" + input: "clip_by_global_norm/clip_by_global_norm/_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_13" + op: "Add" + input: "Mul_21" + input: "Mul_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_23/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_23" + op: "Mul" + input: "Mul_23/x" + input: "bert/embeddings/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_3" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_24/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_24" + op: "Mul" + input: "Mul_24/x" + input: "Square_3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_14" + op: "Add" + input: "Mul_23" + input: "Mul_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_3" + op: "Sqrt" + input: "add_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_15/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_15" + op: "Add" + input: "Sqrt_3" + input: "add_15/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_4" + op: "RealDiv" + input: "add_13" + input: "add_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_25" + op: "Mul" + input: "add" + input: "truediv_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_4" + op: "Sub" + input: "bert/embeddings/LayerNorm/beta/read" + input: "mul_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_208" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta" + input: "sub_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_209" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_m" + input: "add_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_210" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_v" + input: "add_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + input: "bert/embeddings/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + input: "bert/embeddings/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/embeddings/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_26/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_26" + op: "Mul" + input: "Mul_26/x" + input: "bert/embeddings/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_27/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_27" + op: "Mul" + input: "Mul_27/x" + input: "clip_by_global_norm/clip_by_global_norm/_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_16" + op: "Add" + input: "Mul_26" + input: "Mul_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_28/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_28" + op: "Mul" + input: "Mul_28/x" + input: "bert/embeddings/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_4" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_29/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_29" + op: "Mul" + input: "Mul_29/x" + input: "Square_4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_17" + op: "Add" + input: "Mul_28" + input: "Mul_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_4" + op: "Sqrt" + input: "add_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_18/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_18" + op: "Add" + input: "Sqrt_4" + input: "add_18/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_5" + op: "RealDiv" + input: "add_16" + input: "add_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_30" + op: "Mul" + input: "add" + input: "truediv_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_5" + op: "Sub" + input: "bert/embeddings/LayerNorm/gamma/read" + input: "mul_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_211" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma" + input: "sub_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_212" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + input: "add_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_213" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + input: "add_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_31/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_31" + op: "Mul" + input: "Mul_31/x" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_32/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_32" + op: "Mul" + input: "Mul_32/x" + input: "clip_by_global_norm/clip_by_global_norm/_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_19" + op: "Add" + input: "Mul_31" + input: "Mul_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_33/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_33" + op: "Mul" + input: "Mul_33/x" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_5" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_34/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_34" + op: "Mul" + input: "Mul_34/x" + input: "Square_5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_20" + op: "Add" + input: "Mul_33" + input: "Mul_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_5" + op: "Sqrt" + input: "add_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_21/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_21" + op: "Add" + input: "Sqrt_5" + input: "add_21/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_6" + op: "RealDiv" + input: "add_19" + input: "add_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_35/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_35" + op: "Mul" + input: "mul_35/x" + input: "bert/encoder/layer_0/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_22" + op: "Add" + input: "truediv_6" + input: "mul_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_36" + op: "Mul" + input: "add" + input: "add_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_6" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/query/kernel/read" + input: "mul_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_214" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel" + input: "sub_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_215" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + input: "add_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_216" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + input: "add_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_37/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_37" + op: "Mul" + input: "Mul_37/x" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_38/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_38" + op: "Mul" + input: "Mul_38/x" + input: "clip_by_global_norm/clip_by_global_norm/_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_23" + op: "Add" + input: "Mul_37" + input: "Mul_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_39/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_39" + op: "Mul" + input: "Mul_39/x" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_6" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_40/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_40" + op: "Mul" + input: "Mul_40/x" + input: "Square_6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_24" + op: "Add" + input: "Mul_39" + input: "Mul_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_6" + op: "Sqrt" + input: "add_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_25/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_25" + op: "Add" + input: "Sqrt_6" + input: "add_25/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_7" + op: "RealDiv" + input: "add_23" + input: "add_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_41" + op: "Mul" + input: "add" + input: "truediv_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_7" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/query/bias/read" + input: "mul_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_217" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias" + input: "sub_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_218" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + input: "add_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_219" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + input: "add_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_42/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_42" + op: "Mul" + input: "Mul_42/x" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_43/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_43" + op: "Mul" + input: "Mul_43/x" + input: "clip_by_global_norm/clip_by_global_norm/_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_26" + op: "Add" + input: "Mul_42" + input: "Mul_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_44/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_44" + op: "Mul" + input: "Mul_44/x" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_7" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_45/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_45" + op: "Mul" + input: "Mul_45/x" + input: "Square_7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_27" + op: "Add" + input: "Mul_44" + input: "Mul_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_7" + op: "Sqrt" + input: "add_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_28/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_28" + op: "Add" + input: "Sqrt_7" + input: "add_28/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_8" + op: "RealDiv" + input: "add_26" + input: "add_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_46/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_46" + op: "Mul" + input: "mul_46/x" + input: "bert/encoder/layer_0/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_29" + op: "Add" + input: "truediv_8" + input: "mul_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_47" + op: "Mul" + input: "add" + input: "add_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_8" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/key/kernel/read" + input: "mul_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_220" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel" + input: "sub_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_221" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + input: "add_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_222" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + input: "add_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_48/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_48" + op: "Mul" + input: "Mul_48/x" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_49/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_49" + op: "Mul" + input: "Mul_49/x" + input: "clip_by_global_norm/clip_by_global_norm/_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_30" + op: "Add" + input: "Mul_48" + input: "Mul_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_50/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_50" + op: "Mul" + input: "Mul_50/x" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_8" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_51/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_51" + op: "Mul" + input: "Mul_51/x" + input: "Square_8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_31" + op: "Add" + input: "Mul_50" + input: "Mul_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_8" + op: "Sqrt" + input: "add_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_32/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_32" + op: "Add" + input: "Sqrt_8" + input: "add_32/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_9" + op: "RealDiv" + input: "add_30" + input: "add_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_52" + op: "Mul" + input: "add" + input: "truediv_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_9" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/key/bias/read" + input: "mul_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_223" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias" + input: "sub_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_224" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + input: "add_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_225" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + input: "add_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_53/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_53" + op: "Mul" + input: "Mul_53/x" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_54/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_54" + op: "Mul" + input: "Mul_54/x" + input: "clip_by_global_norm/clip_by_global_norm/_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_33" + op: "Add" + input: "Mul_53" + input: "Mul_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_55/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_55" + op: "Mul" + input: "Mul_55/x" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_9" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_56/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_56" + op: "Mul" + input: "Mul_56/x" + input: "Square_9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_34" + op: "Add" + input: "Mul_55" + input: "Mul_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_9" + op: "Sqrt" + input: "add_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_35/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_35" + op: "Add" + input: "Sqrt_9" + input: "add_35/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_10" + op: "RealDiv" + input: "add_33" + input: "add_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_57/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_57" + op: "Mul" + input: "mul_57/x" + input: "bert/encoder/layer_0/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_36" + op: "Add" + input: "truediv_10" + input: "mul_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_58" + op: "Mul" + input: "add" + input: "add_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_10" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/value/kernel/read" + input: "mul_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_226" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel" + input: "sub_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_227" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + input: "add_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_228" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + input: "add_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_59/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_59" + op: "Mul" + input: "Mul_59/x" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_60/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_60" + op: "Mul" + input: "Mul_60/x" + input: "clip_by_global_norm/clip_by_global_norm/_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_37" + op: "Add" + input: "Mul_59" + input: "Mul_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_61/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_61" + op: "Mul" + input: "Mul_61/x" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_10" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_62/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_62" + op: "Mul" + input: "Mul_62/x" + input: "Square_10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_38" + op: "Add" + input: "Mul_61" + input: "Mul_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_10" + op: "Sqrt" + input: "add_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_39/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_39" + op: "Add" + input: "Sqrt_10" + input: "add_39/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_11" + op: "RealDiv" + input: "add_37" + input: "add_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_63" + op: "Mul" + input: "add" + input: "truediv_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_11" + op: "Sub" + input: "bert/encoder/layer_0/attention/self/value/bias/read" + input: "mul_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_229" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias" + input: "sub_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_230" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + input: "add_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_231" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + input: "add_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_64/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_64" + op: "Mul" + input: "Mul_64/x" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_65/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_65" + op: "Mul" + input: "Mul_65/x" + input: "clip_by_global_norm/clip_by_global_norm/_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_40" + op: "Add" + input: "Mul_64" + input: "Mul_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_66/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_66" + op: "Mul" + input: "Mul_66/x" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_11" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_67/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_67" + op: "Mul" + input: "Mul_67/x" + input: "Square_11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_41" + op: "Add" + input: "Mul_66" + input: "Mul_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_11" + op: "Sqrt" + input: "add_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_42/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_42" + op: "Add" + input: "Sqrt_11" + input: "add_42/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_12" + op: "RealDiv" + input: "add_40" + input: "add_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_68/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_68" + op: "Mul" + input: "mul_68/x" + input: "bert/encoder/layer_0/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_43" + op: "Add" + input: "truediv_12" + input: "mul_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_69" + op: "Mul" + input: "add" + input: "add_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_12" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/dense/kernel/read" + input: "mul_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_232" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + input: "sub_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_233" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + input: "add_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_234" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + input: "add_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_70/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_70" + op: "Mul" + input: "Mul_70/x" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_71/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_71" + op: "Mul" + input: "Mul_71/x" + input: "clip_by_global_norm/clip_by_global_norm/_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_44" + op: "Add" + input: "Mul_70" + input: "Mul_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_72/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_72" + op: "Mul" + input: "Mul_72/x" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_12" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_73/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_73" + op: "Mul" + input: "Mul_73/x" + input: "Square_12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_45" + op: "Add" + input: "Mul_72" + input: "Mul_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_12" + op: "Sqrt" + input: "add_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_46/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_46" + op: "Add" + input: "Sqrt_12" + input: "add_46/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_13" + op: "RealDiv" + input: "add_44" + input: "add_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_74" + op: "Mul" + input: "add" + input: "truediv_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_13" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/dense/bias/read" + input: "mul_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_235" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias" + input: "sub_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_236" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + input: "add_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_237" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + input: "add_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_75/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_75" + op: "Mul" + input: "Mul_75/x" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_76/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_76" + op: "Mul" + input: "Mul_76/x" + input: "clip_by_global_norm/clip_by_global_norm/_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_47" + op: "Add" + input: "Mul_75" + input: "Mul_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_77/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_77" + op: "Mul" + input: "Mul_77/x" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_13" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_78/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_78" + op: "Mul" + input: "Mul_78/x" + input: "Square_13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_48" + op: "Add" + input: "Mul_77" + input: "Mul_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_13" + op: "Sqrt" + input: "add_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_49/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_49" + op: "Add" + input: "Sqrt_13" + input: "add_49/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_14" + op: "RealDiv" + input: "add_47" + input: "add_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_79" + op: "Mul" + input: "add" + input: "truediv_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_14" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/read" + input: "mul_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_238" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + input: "sub_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_239" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + input: "add_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_240" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + input: "add_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_80/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_80" + op: "Mul" + input: "Mul_80/x" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_81/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_81" + op: "Mul" + input: "Mul_81/x" + input: "clip_by_global_norm/clip_by_global_norm/_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_50" + op: "Add" + input: "Mul_80" + input: "Mul_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_82/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_82" + op: "Mul" + input: "Mul_82/x" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_14" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_83/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_83" + op: "Mul" + input: "Mul_83/x" + input: "Square_14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_51" + op: "Add" + input: "Mul_82" + input: "Mul_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_14" + op: "Sqrt" + input: "add_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_52/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_52" + op: "Add" + input: "Sqrt_14" + input: "add_52/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_15" + op: "RealDiv" + input: "add_50" + input: "add_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_84" + op: "Mul" + input: "add" + input: "truediv_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_15" + op: "Sub" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/read" + input: "mul_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_241" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + input: "sub_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_242" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + input: "add_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_243" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + input: "add_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_85/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_85" + op: "Mul" + input: "Mul_85/x" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_86/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_86" + op: "Mul" + input: "Mul_86/x" + input: "clip_by_global_norm/clip_by_global_norm/_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_53" + op: "Add" + input: "Mul_85" + input: "Mul_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_87/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_87" + op: "Mul" + input: "Mul_87/x" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_15" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_88/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_88" + op: "Mul" + input: "Mul_88/x" + input: "Square_15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_54" + op: "Add" + input: "Mul_87" + input: "Mul_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_15" + op: "Sqrt" + input: "add_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_55/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_55" + op: "Add" + input: "Sqrt_15" + input: "add_55/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_16" + op: "RealDiv" + input: "add_53" + input: "add_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_89/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_89" + op: "Mul" + input: "mul_89/x" + input: "bert/encoder/layer_0/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_56" + op: "Add" + input: "truediv_16" + input: "mul_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_90" + op: "Mul" + input: "add" + input: "add_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_16" + op: "Sub" + input: "bert/encoder/layer_0/intermediate/dense/kernel/read" + input: "mul_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_244" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + input: "sub_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_245" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + input: "add_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_246" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + input: "add_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_91/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_91" + op: "Mul" + input: "Mul_91/x" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_92/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_92" + op: "Mul" + input: "Mul_92/x" + input: "clip_by_global_norm/clip_by_global_norm/_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_57" + op: "Add" + input: "Mul_91" + input: "Mul_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_93/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_93" + op: "Mul" + input: "Mul_93/x" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_16" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_94/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_94" + op: "Mul" + input: "Mul_94/x" + input: "Square_16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_58" + op: "Add" + input: "Mul_93" + input: "Mul_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_16" + op: "Sqrt" + input: "add_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_59/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_59" + op: "Add" + input: "Sqrt_16" + input: "add_59/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_17" + op: "RealDiv" + input: "add_57" + input: "add_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_95" + op: "Mul" + input: "add" + input: "truediv_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_17" + op: "Sub" + input: "bert/encoder/layer_0/intermediate/dense/bias/read" + input: "mul_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_247" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias" + input: "sub_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_248" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + input: "add_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_249" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + input: "add_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_96/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_96" + op: "Mul" + input: "Mul_96/x" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_97/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_97" + op: "Mul" + input: "Mul_97/x" + input: "clip_by_global_norm/clip_by_global_norm/_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_60" + op: "Add" + input: "Mul_96" + input: "Mul_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_98/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_98" + op: "Mul" + input: "Mul_98/x" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_17" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_99/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_99" + op: "Mul" + input: "Mul_99/x" + input: "Square_17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_61" + op: "Add" + input: "Mul_98" + input: "Mul_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_17" + op: "Sqrt" + input: "add_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_62/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_62" + op: "Add" + input: "Sqrt_17" + input: "add_62/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_18" + op: "RealDiv" + input: "add_60" + input: "add_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_100/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_100" + op: "Mul" + input: "mul_100/x" + input: "bert/encoder/layer_0/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_63" + op: "Add" + input: "truediv_18" + input: "mul_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_101" + op: "Mul" + input: "add" + input: "add_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_18" + op: "Sub" + input: "bert/encoder/layer_0/output/dense/kernel/read" + input: "mul_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_250" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel" + input: "sub_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_251" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + input: "add_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_252" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + input: "add_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + input: "bert/encoder/layer_0/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + input: "bert/encoder/layer_0/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_102/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_102" + op: "Mul" + input: "Mul_102/x" + input: "bert/encoder/layer_0/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_103/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_103" + op: "Mul" + input: "Mul_103/x" + input: "clip_by_global_norm/clip_by_global_norm/_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_64" + op: "Add" + input: "Mul_102" + input: "Mul_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_104/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_104" + op: "Mul" + input: "Mul_104/x" + input: "bert/encoder/layer_0/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_18" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_105/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_105" + op: "Mul" + input: "Mul_105/x" + input: "Square_18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_65" + op: "Add" + input: "Mul_104" + input: "Mul_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_18" + op: "Sqrt" + input: "add_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_66/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_66" + op: "Add" + input: "Sqrt_18" + input: "add_66/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_19" + op: "RealDiv" + input: "add_64" + input: "add_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_106" + op: "Mul" + input: "add" + input: "truediv_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_19" + op: "Sub" + input: "bert/encoder/layer_0/output/dense/bias/read" + input: "mul_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_253" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias" + input: "sub_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_254" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + input: "add_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_255" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + input: "add_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_107/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_107" + op: "Mul" + input: "Mul_107/x" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_108/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_108" + op: "Mul" + input: "Mul_108/x" + input: "clip_by_global_norm/clip_by_global_norm/_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_67" + op: "Add" + input: "Mul_107" + input: "Mul_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_109/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_109" + op: "Mul" + input: "Mul_109/x" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_19" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_110/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_110" + op: "Mul" + input: "Mul_110/x" + input: "Square_19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_68" + op: "Add" + input: "Mul_109" + input: "Mul_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_19" + op: "Sqrt" + input: "add_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_69/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_69" + op: "Add" + input: "Sqrt_19" + input: "add_69/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_20" + op: "RealDiv" + input: "add_67" + input: "add_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_111" + op: "Mul" + input: "add" + input: "truediv_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_20" + op: "Sub" + input: "bert/encoder/layer_0/output/LayerNorm/beta/read" + input: "mul_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_256" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + input: "sub_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_257" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + input: "add_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_258" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + input: "add_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_112/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_112" + op: "Mul" + input: "Mul_112/x" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_113/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_113" + op: "Mul" + input: "Mul_113/x" + input: "clip_by_global_norm/clip_by_global_norm/_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_70" + op: "Add" + input: "Mul_112" + input: "Mul_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_114/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_114" + op: "Mul" + input: "Mul_114/x" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_20" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_115/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_115" + op: "Mul" + input: "Mul_115/x" + input: "Square_20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_71" + op: "Add" + input: "Mul_114" + input: "Mul_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_20" + op: "Sqrt" + input: "add_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_72/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_72" + op: "Add" + input: "Sqrt_20" + input: "add_72/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_21" + op: "RealDiv" + input: "add_70" + input: "add_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_116" + op: "Mul" + input: "add" + input: "truediv_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_21" + op: "Sub" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/read" + input: "mul_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_259" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + input: "sub_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_260" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + input: "add_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_261" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + input: "add_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_117/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_117" + op: "Mul" + input: "Mul_117/x" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_118/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_118" + op: "Mul" + input: "Mul_118/x" + input: "clip_by_global_norm/clip_by_global_norm/_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_73" + op: "Add" + input: "Mul_117" + input: "Mul_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_119/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_119" + op: "Mul" + input: "Mul_119/x" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_21" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_120/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_120" + op: "Mul" + input: "Mul_120/x" + input: "Square_21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_74" + op: "Add" + input: "Mul_119" + input: "Mul_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_21" + op: "Sqrt" + input: "add_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_75/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_75" + op: "Add" + input: "Sqrt_21" + input: "add_75/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_22" + op: "RealDiv" + input: "add_73" + input: "add_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_121/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_121" + op: "Mul" + input: "mul_121/x" + input: "bert/encoder/layer_1/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_76" + op: "Add" + input: "truediv_22" + input: "mul_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_122" + op: "Mul" + input: "add" + input: "add_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_22" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/query/kernel/read" + input: "mul_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_262" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel" + input: "sub_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_263" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + input: "add_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_264" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + input: "add_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_123/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_123" + op: "Mul" + input: "Mul_123/x" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_124/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_124" + op: "Mul" + input: "Mul_124/x" + input: "clip_by_global_norm/clip_by_global_norm/_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_77" + op: "Add" + input: "Mul_123" + input: "Mul_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_125/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_125" + op: "Mul" + input: "Mul_125/x" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_22" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_126/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_126" + op: "Mul" + input: "Mul_126/x" + input: "Square_22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_78" + op: "Add" + input: "Mul_125" + input: "Mul_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_22" + op: "Sqrt" + input: "add_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_79/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_79" + op: "Add" + input: "Sqrt_22" + input: "add_79/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_23" + op: "RealDiv" + input: "add_77" + input: "add_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_127" + op: "Mul" + input: "add" + input: "truediv_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_23" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/query/bias/read" + input: "mul_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_265" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias" + input: "sub_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_266" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + input: "add_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_267" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + input: "add_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_128/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_128" + op: "Mul" + input: "Mul_128/x" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_129/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_129" + op: "Mul" + input: "Mul_129/x" + input: "clip_by_global_norm/clip_by_global_norm/_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_80" + op: "Add" + input: "Mul_128" + input: "Mul_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_130/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_130" + op: "Mul" + input: "Mul_130/x" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_23" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_131/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_131" + op: "Mul" + input: "Mul_131/x" + input: "Square_23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_81" + op: "Add" + input: "Mul_130" + input: "Mul_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_23" + op: "Sqrt" + input: "add_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_82/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_82" + op: "Add" + input: "Sqrt_23" + input: "add_82/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_24" + op: "RealDiv" + input: "add_80" + input: "add_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_132/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_132" + op: "Mul" + input: "mul_132/x" + input: "bert/encoder/layer_1/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_83" + op: "Add" + input: "truediv_24" + input: "mul_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_133" + op: "Mul" + input: "add" + input: "add_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_24" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/key/kernel/read" + input: "mul_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_268" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel" + input: "sub_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_269" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + input: "add_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_270" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + input: "add_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_134/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_134" + op: "Mul" + input: "Mul_134/x" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_135/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_135" + op: "Mul" + input: "Mul_135/x" + input: "clip_by_global_norm/clip_by_global_norm/_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_84" + op: "Add" + input: "Mul_134" + input: "Mul_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_136/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_136" + op: "Mul" + input: "Mul_136/x" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_24" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_137/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_137" + op: "Mul" + input: "Mul_137/x" + input: "Square_24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_85" + op: "Add" + input: "Mul_136" + input: "Mul_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_24" + op: "Sqrt" + input: "add_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_86/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_86" + op: "Add" + input: "Sqrt_24" + input: "add_86/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_25" + op: "RealDiv" + input: "add_84" + input: "add_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_138" + op: "Mul" + input: "add" + input: "truediv_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_25" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/key/bias/read" + input: "mul_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_271" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias" + input: "sub_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_272" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + input: "add_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_273" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + input: "add_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_139/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_139" + op: "Mul" + input: "Mul_139/x" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_140/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_140" + op: "Mul" + input: "Mul_140/x" + input: "clip_by_global_norm/clip_by_global_norm/_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_87" + op: "Add" + input: "Mul_139" + input: "Mul_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_141/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_141" + op: "Mul" + input: "Mul_141/x" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_25" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_142/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_142" + op: "Mul" + input: "Mul_142/x" + input: "Square_25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_88" + op: "Add" + input: "Mul_141" + input: "Mul_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_25" + op: "Sqrt" + input: "add_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_89/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_89" + op: "Add" + input: "Sqrt_25" + input: "add_89/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_26" + op: "RealDiv" + input: "add_87" + input: "add_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_143/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_143" + op: "Mul" + input: "mul_143/x" + input: "bert/encoder/layer_1/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_90" + op: "Add" + input: "truediv_26" + input: "mul_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_144" + op: "Mul" + input: "add" + input: "add_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_26" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/value/kernel/read" + input: "mul_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_274" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel" + input: "sub_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_275" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + input: "add_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_276" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + input: "add_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_145/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_145" + op: "Mul" + input: "Mul_145/x" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_146/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_146" + op: "Mul" + input: "Mul_146/x" + input: "clip_by_global_norm/clip_by_global_norm/_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_91" + op: "Add" + input: "Mul_145" + input: "Mul_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_147/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_147" + op: "Mul" + input: "Mul_147/x" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_26" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_148/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_148" + op: "Mul" + input: "Mul_148/x" + input: "Square_26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_92" + op: "Add" + input: "Mul_147" + input: "Mul_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_26" + op: "Sqrt" + input: "add_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_93/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_93" + op: "Add" + input: "Sqrt_26" + input: "add_93/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_27" + op: "RealDiv" + input: "add_91" + input: "add_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_149" + op: "Mul" + input: "add" + input: "truediv_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_27" + op: "Sub" + input: "bert/encoder/layer_1/attention/self/value/bias/read" + input: "mul_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_277" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias" + input: "sub_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_278" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + input: "add_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_279" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + input: "add_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_150/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_150" + op: "Mul" + input: "Mul_150/x" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_151/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_151" + op: "Mul" + input: "Mul_151/x" + input: "clip_by_global_norm/clip_by_global_norm/_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_94" + op: "Add" + input: "Mul_150" + input: "Mul_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_152/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_152" + op: "Mul" + input: "Mul_152/x" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_27" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_153/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_153" + op: "Mul" + input: "Mul_153/x" + input: "Square_27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_95" + op: "Add" + input: "Mul_152" + input: "Mul_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_27" + op: "Sqrt" + input: "add_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_96/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_96" + op: "Add" + input: "Sqrt_27" + input: "add_96/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_28" + op: "RealDiv" + input: "add_94" + input: "add_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_154/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_154" + op: "Mul" + input: "mul_154/x" + input: "bert/encoder/layer_1/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_97" + op: "Add" + input: "truediv_28" + input: "mul_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_155" + op: "Mul" + input: "add" + input: "add_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_28" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/dense/kernel/read" + input: "mul_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_280" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + input: "sub_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_281" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + input: "add_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_282" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + input: "add_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_156/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_156" + op: "Mul" + input: "Mul_156/x" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_157/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_157" + op: "Mul" + input: "Mul_157/x" + input: "clip_by_global_norm/clip_by_global_norm/_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_98" + op: "Add" + input: "Mul_156" + input: "Mul_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_158/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_158" + op: "Mul" + input: "Mul_158/x" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_28" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_159/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_159" + op: "Mul" + input: "Mul_159/x" + input: "Square_28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_99" + op: "Add" + input: "Mul_158" + input: "Mul_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_28" + op: "Sqrt" + input: "add_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_100/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_100" + op: "Add" + input: "Sqrt_28" + input: "add_100/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_29" + op: "RealDiv" + input: "add_98" + input: "add_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_160" + op: "Mul" + input: "add" + input: "truediv_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_29" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/dense/bias/read" + input: "mul_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_283" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias" + input: "sub_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_284" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + input: "add_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_285" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + input: "add_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_161/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_161" + op: "Mul" + input: "Mul_161/x" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_162/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_162" + op: "Mul" + input: "Mul_162/x" + input: "clip_by_global_norm/clip_by_global_norm/_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_101" + op: "Add" + input: "Mul_161" + input: "Mul_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_163/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_163" + op: "Mul" + input: "Mul_163/x" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_29" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_164/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_164" + op: "Mul" + input: "Mul_164/x" + input: "Square_29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_102" + op: "Add" + input: "Mul_163" + input: "Mul_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_29" + op: "Sqrt" + input: "add_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_103/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_103" + op: "Add" + input: "Sqrt_29" + input: "add_103/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_30" + op: "RealDiv" + input: "add_101" + input: "add_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_165" + op: "Mul" + input: "add" + input: "truediv_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_30" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/read" + input: "mul_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_286" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + input: "sub_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_287" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + input: "add_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_288" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + input: "add_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_166/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_166" + op: "Mul" + input: "Mul_166/x" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_167/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_167" + op: "Mul" + input: "Mul_167/x" + input: "clip_by_global_norm/clip_by_global_norm/_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_104" + op: "Add" + input: "Mul_166" + input: "Mul_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_168/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_168" + op: "Mul" + input: "Mul_168/x" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_30" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_169/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_169" + op: "Mul" + input: "Mul_169/x" + input: "Square_30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_105" + op: "Add" + input: "Mul_168" + input: "Mul_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_30" + op: "Sqrt" + input: "add_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_106/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_106" + op: "Add" + input: "Sqrt_30" + input: "add_106/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_31" + op: "RealDiv" + input: "add_104" + input: "add_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_170" + op: "Mul" + input: "add" + input: "truediv_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_31" + op: "Sub" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/read" + input: "mul_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_289" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + input: "sub_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_290" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + input: "add_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_291" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + input: "add_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_171/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_171" + op: "Mul" + input: "Mul_171/x" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_172/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_172" + op: "Mul" + input: "Mul_172/x" + input: "clip_by_global_norm/clip_by_global_norm/_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_107" + op: "Add" + input: "Mul_171" + input: "Mul_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_173/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_173" + op: "Mul" + input: "Mul_173/x" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_31" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_174/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_174" + op: "Mul" + input: "Mul_174/x" + input: "Square_31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_108" + op: "Add" + input: "Mul_173" + input: "Mul_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_31" + op: "Sqrt" + input: "add_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_109/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_109" + op: "Add" + input: "Sqrt_31" + input: "add_109/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_32" + op: "RealDiv" + input: "add_107" + input: "add_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_175/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_175" + op: "Mul" + input: "mul_175/x" + input: "bert/encoder/layer_1/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_110" + op: "Add" + input: "truediv_32" + input: "mul_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_176" + op: "Mul" + input: "add" + input: "add_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_32" + op: "Sub" + input: "bert/encoder/layer_1/intermediate/dense/kernel/read" + input: "mul_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_292" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + input: "sub_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_293" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + input: "add_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_294" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + input: "add_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_177/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_177" + op: "Mul" + input: "Mul_177/x" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_178/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_178" + op: "Mul" + input: "Mul_178/x" + input: "clip_by_global_norm/clip_by_global_norm/_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_111" + op: "Add" + input: "Mul_177" + input: "Mul_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_179/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_179" + op: "Mul" + input: "Mul_179/x" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_32" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_180/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_180" + op: "Mul" + input: "Mul_180/x" + input: "Square_32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_112" + op: "Add" + input: "Mul_179" + input: "Mul_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_32" + op: "Sqrt" + input: "add_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_113/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_113" + op: "Add" + input: "Sqrt_32" + input: "add_113/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_33" + op: "RealDiv" + input: "add_111" + input: "add_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_181" + op: "Mul" + input: "add" + input: "truediv_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_33" + op: "Sub" + input: "bert/encoder/layer_1/intermediate/dense/bias/read" + input: "mul_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_295" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias" + input: "sub_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_296" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + input: "add_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_297" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + input: "add_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_182/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_182" + op: "Mul" + input: "Mul_182/x" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_183/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_183" + op: "Mul" + input: "Mul_183/x" + input: "clip_by_global_norm/clip_by_global_norm/_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_114" + op: "Add" + input: "Mul_182" + input: "Mul_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_184/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_184" + op: "Mul" + input: "Mul_184/x" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_33" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_185/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_185" + op: "Mul" + input: "Mul_185/x" + input: "Square_33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_115" + op: "Add" + input: "Mul_184" + input: "Mul_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_33" + op: "Sqrt" + input: "add_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_116/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_116" + op: "Add" + input: "Sqrt_33" + input: "add_116/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_34" + op: "RealDiv" + input: "add_114" + input: "add_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_186/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_186" + op: "Mul" + input: "mul_186/x" + input: "bert/encoder/layer_1/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_117" + op: "Add" + input: "truediv_34" + input: "mul_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_187" + op: "Mul" + input: "add" + input: "add_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_34" + op: "Sub" + input: "bert/encoder/layer_1/output/dense/kernel/read" + input: "mul_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_298" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel" + input: "sub_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_299" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + input: "add_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_300" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + input: "add_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + input: "bert/encoder/layer_1/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + input: "bert/encoder/layer_1/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_188/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_188" + op: "Mul" + input: "Mul_188/x" + input: "bert/encoder/layer_1/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_189/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_189" + op: "Mul" + input: "Mul_189/x" + input: "clip_by_global_norm/clip_by_global_norm/_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_118" + op: "Add" + input: "Mul_188" + input: "Mul_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_190/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_190" + op: "Mul" + input: "Mul_190/x" + input: "bert/encoder/layer_1/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_34" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_191/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_191" + op: "Mul" + input: "Mul_191/x" + input: "Square_34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_119" + op: "Add" + input: "Mul_190" + input: "Mul_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_34" + op: "Sqrt" + input: "add_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_120/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_120" + op: "Add" + input: "Sqrt_34" + input: "add_120/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_35" + op: "RealDiv" + input: "add_118" + input: "add_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_192" + op: "Mul" + input: "add" + input: "truediv_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_35" + op: "Sub" + input: "bert/encoder/layer_1/output/dense/bias/read" + input: "mul_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_301" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias" + input: "sub_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_302" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + input: "add_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_303" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + input: "add_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_193/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_193" + op: "Mul" + input: "Mul_193/x" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_194/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_194" + op: "Mul" + input: "Mul_194/x" + input: "clip_by_global_norm/clip_by_global_norm/_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_121" + op: "Add" + input: "Mul_193" + input: "Mul_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_195/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_195" + op: "Mul" + input: "Mul_195/x" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_35" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_196/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_196" + op: "Mul" + input: "Mul_196/x" + input: "Square_35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_122" + op: "Add" + input: "Mul_195" + input: "Mul_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_35" + op: "Sqrt" + input: "add_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_123/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_123" + op: "Add" + input: "Sqrt_35" + input: "add_123/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_36" + op: "RealDiv" + input: "add_121" + input: "add_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_197" + op: "Mul" + input: "add" + input: "truediv_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_36" + op: "Sub" + input: "bert/encoder/layer_1/output/LayerNorm/beta/read" + input: "mul_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_304" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + input: "sub_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_305" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + input: "add_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_306" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + input: "add_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_198/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_198" + op: "Mul" + input: "Mul_198/x" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_199/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_199" + op: "Mul" + input: "Mul_199/x" + input: "clip_by_global_norm/clip_by_global_norm/_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_124" + op: "Add" + input: "Mul_198" + input: "Mul_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_200/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_200" + op: "Mul" + input: "Mul_200/x" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_36" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_201/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_201" + op: "Mul" + input: "Mul_201/x" + input: "Square_36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_125" + op: "Add" + input: "Mul_200" + input: "Mul_201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_36" + op: "Sqrt" + input: "add_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_126/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_126" + op: "Add" + input: "Sqrt_36" + input: "add_126/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_37" + op: "RealDiv" + input: "add_124" + input: "add_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_202" + op: "Mul" + input: "add" + input: "truediv_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_37" + op: "Sub" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/read" + input: "mul_202" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_307" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + input: "sub_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_308" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + input: "add_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_309" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + input: "add_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_203/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_203" + op: "Mul" + input: "Mul_203/x" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_204/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_204" + op: "Mul" + input: "Mul_204/x" + input: "clip_by_global_norm/clip_by_global_norm/_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_127" + op: "Add" + input: "Mul_203" + input: "Mul_204" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_205/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_205" + op: "Mul" + input: "Mul_205/x" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_37" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_206/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_206" + op: "Mul" + input: "Mul_206/x" + input: "Square_37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_128" + op: "Add" + input: "Mul_205" + input: "Mul_206" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_37" + op: "Sqrt" + input: "add_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_129/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_129" + op: "Add" + input: "Sqrt_37" + input: "add_129/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_38" + op: "RealDiv" + input: "add_127" + input: "add_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_207/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_207" + op: "Mul" + input: "mul_207/x" + input: "bert/encoder/layer_2/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_130" + op: "Add" + input: "truediv_38" + input: "mul_207" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_208" + op: "Mul" + input: "add" + input: "add_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_38" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/query/kernel/read" + input: "mul_208" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_310" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel" + input: "sub_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_311" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + input: "add_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_312" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + input: "add_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_209/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_209" + op: "Mul" + input: "Mul_209/x" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_210/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_210" + op: "Mul" + input: "Mul_210/x" + input: "clip_by_global_norm/clip_by_global_norm/_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_131" + op: "Add" + input: "Mul_209" + input: "Mul_210" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_211/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_211" + op: "Mul" + input: "Mul_211/x" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_38" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_212/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_212" + op: "Mul" + input: "Mul_212/x" + input: "Square_38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_132" + op: "Add" + input: "Mul_211" + input: "Mul_212" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_38" + op: "Sqrt" + input: "add_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_133/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_133" + op: "Add" + input: "Sqrt_38" + input: "add_133/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_39" + op: "RealDiv" + input: "add_131" + input: "add_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_213" + op: "Mul" + input: "add" + input: "truediv_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_39" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/query/bias/read" + input: "mul_213" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_313" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias" + input: "sub_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_314" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + input: "add_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_315" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + input: "add_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_214/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_214" + op: "Mul" + input: "Mul_214/x" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_215/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_215" + op: "Mul" + input: "Mul_215/x" + input: "clip_by_global_norm/clip_by_global_norm/_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_134" + op: "Add" + input: "Mul_214" + input: "Mul_215" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_216/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_216" + op: "Mul" + input: "Mul_216/x" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_39" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_217/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_217" + op: "Mul" + input: "Mul_217/x" + input: "Square_39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_135" + op: "Add" + input: "Mul_216" + input: "Mul_217" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_39" + op: "Sqrt" + input: "add_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_136/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_136" + op: "Add" + input: "Sqrt_39" + input: "add_136/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_40" + op: "RealDiv" + input: "add_134" + input: "add_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_218/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_218" + op: "Mul" + input: "mul_218/x" + input: "bert/encoder/layer_2/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_137" + op: "Add" + input: "truediv_40" + input: "mul_218" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_219" + op: "Mul" + input: "add" + input: "add_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_40" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/key/kernel/read" + input: "mul_219" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_316" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel" + input: "sub_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_317" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + input: "add_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_318" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + input: "add_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_220/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_220" + op: "Mul" + input: "Mul_220/x" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_221/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_221" + op: "Mul" + input: "Mul_221/x" + input: "clip_by_global_norm/clip_by_global_norm/_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_138" + op: "Add" + input: "Mul_220" + input: "Mul_221" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_222/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_222" + op: "Mul" + input: "Mul_222/x" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_40" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_223/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_223" + op: "Mul" + input: "Mul_223/x" + input: "Square_40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_139" + op: "Add" + input: "Mul_222" + input: "Mul_223" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_40" + op: "Sqrt" + input: "add_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_140/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_140" + op: "Add" + input: "Sqrt_40" + input: "add_140/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_41" + op: "RealDiv" + input: "add_138" + input: "add_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_224" + op: "Mul" + input: "add" + input: "truediv_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_41" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/key/bias/read" + input: "mul_224" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_319" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias" + input: "sub_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_320" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + input: "add_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_321" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + input: "add_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_225/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_225" + op: "Mul" + input: "Mul_225/x" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_226/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_226" + op: "Mul" + input: "Mul_226/x" + input: "clip_by_global_norm/clip_by_global_norm/_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_141" + op: "Add" + input: "Mul_225" + input: "Mul_226" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_227/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_227" + op: "Mul" + input: "Mul_227/x" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_41" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_228/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_228" + op: "Mul" + input: "Mul_228/x" + input: "Square_41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_142" + op: "Add" + input: "Mul_227" + input: "Mul_228" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_41" + op: "Sqrt" + input: "add_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_143/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_143" + op: "Add" + input: "Sqrt_41" + input: "add_143/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_42" + op: "RealDiv" + input: "add_141" + input: "add_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_229/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_229" + op: "Mul" + input: "mul_229/x" + input: "bert/encoder/layer_2/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_144" + op: "Add" + input: "truediv_42" + input: "mul_229" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_230" + op: "Mul" + input: "add" + input: "add_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_42" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/value/kernel/read" + input: "mul_230" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_322" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel" + input: "sub_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_323" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + input: "add_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_324" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + input: "add_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_231/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_231" + op: "Mul" + input: "Mul_231/x" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_232/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_232" + op: "Mul" + input: "Mul_232/x" + input: "clip_by_global_norm/clip_by_global_norm/_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_145" + op: "Add" + input: "Mul_231" + input: "Mul_232" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_233/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_233" + op: "Mul" + input: "Mul_233/x" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_42" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_234/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_234" + op: "Mul" + input: "Mul_234/x" + input: "Square_42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_146" + op: "Add" + input: "Mul_233" + input: "Mul_234" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_42" + op: "Sqrt" + input: "add_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_147/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_147" + op: "Add" + input: "Sqrt_42" + input: "add_147/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_43" + op: "RealDiv" + input: "add_145" + input: "add_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_235" + op: "Mul" + input: "add" + input: "truediv_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_43" + op: "Sub" + input: "bert/encoder/layer_2/attention/self/value/bias/read" + input: "mul_235" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_325" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias" + input: "sub_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_326" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + input: "add_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_327" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + input: "add_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_236/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_236" + op: "Mul" + input: "Mul_236/x" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_237/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_237" + op: "Mul" + input: "Mul_237/x" + input: "clip_by_global_norm/clip_by_global_norm/_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_148" + op: "Add" + input: "Mul_236" + input: "Mul_237" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_238/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_238" + op: "Mul" + input: "Mul_238/x" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_43" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_239/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_239" + op: "Mul" + input: "Mul_239/x" + input: "Square_43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_149" + op: "Add" + input: "Mul_238" + input: "Mul_239" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_43" + op: "Sqrt" + input: "add_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_150/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_150" + op: "Add" + input: "Sqrt_43" + input: "add_150/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_44" + op: "RealDiv" + input: "add_148" + input: "add_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_240/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_240" + op: "Mul" + input: "mul_240/x" + input: "bert/encoder/layer_2/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_151" + op: "Add" + input: "truediv_44" + input: "mul_240" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_241" + op: "Mul" + input: "add" + input: "add_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_44" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/dense/kernel/read" + input: "mul_241" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_328" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + input: "sub_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_329" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + input: "add_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_330" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + input: "add_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_242/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_242" + op: "Mul" + input: "Mul_242/x" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_243/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_243" + op: "Mul" + input: "Mul_243/x" + input: "clip_by_global_norm/clip_by_global_norm/_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_152" + op: "Add" + input: "Mul_242" + input: "Mul_243" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_244/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_244" + op: "Mul" + input: "Mul_244/x" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_44" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_245/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_245" + op: "Mul" + input: "Mul_245/x" + input: "Square_44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_153" + op: "Add" + input: "Mul_244" + input: "Mul_245" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_44" + op: "Sqrt" + input: "add_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_154/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_154" + op: "Add" + input: "Sqrt_44" + input: "add_154/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_45" + op: "RealDiv" + input: "add_152" + input: "add_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_246" + op: "Mul" + input: "add" + input: "truediv_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_45" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/dense/bias/read" + input: "mul_246" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_331" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias" + input: "sub_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_332" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + input: "add_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_333" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + input: "add_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_247/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_247" + op: "Mul" + input: "Mul_247/x" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_248/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_248" + op: "Mul" + input: "Mul_248/x" + input: "clip_by_global_norm/clip_by_global_norm/_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_155" + op: "Add" + input: "Mul_247" + input: "Mul_248" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_249/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_249" + op: "Mul" + input: "Mul_249/x" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_45" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_250/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_250" + op: "Mul" + input: "Mul_250/x" + input: "Square_45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_156" + op: "Add" + input: "Mul_249" + input: "Mul_250" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_45" + op: "Sqrt" + input: "add_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_157/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_157" + op: "Add" + input: "Sqrt_45" + input: "add_157/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_46" + op: "RealDiv" + input: "add_155" + input: "add_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_251" + op: "Mul" + input: "add" + input: "truediv_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_46" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/read" + input: "mul_251" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_334" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + input: "sub_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_335" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + input: "add_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_336" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + input: "add_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_252/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_252" + op: "Mul" + input: "Mul_252/x" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_253/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_253" + op: "Mul" + input: "Mul_253/x" + input: "clip_by_global_norm/clip_by_global_norm/_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_158" + op: "Add" + input: "Mul_252" + input: "Mul_253" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_254/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_254" + op: "Mul" + input: "Mul_254/x" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_46" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_255/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_255" + op: "Mul" + input: "Mul_255/x" + input: "Square_46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_159" + op: "Add" + input: "Mul_254" + input: "Mul_255" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_46" + op: "Sqrt" + input: "add_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_160/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_160" + op: "Add" + input: "Sqrt_46" + input: "add_160/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_47" + op: "RealDiv" + input: "add_158" + input: "add_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_256" + op: "Mul" + input: "add" + input: "truediv_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_47" + op: "Sub" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/read" + input: "mul_256" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_337" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + input: "sub_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_338" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + input: "add_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_339" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + input: "add_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_257/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_257" + op: "Mul" + input: "Mul_257/x" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_258/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_258" + op: "Mul" + input: "Mul_258/x" + input: "clip_by_global_norm/clip_by_global_norm/_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_161" + op: "Add" + input: "Mul_257" + input: "Mul_258" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_259/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_259" + op: "Mul" + input: "Mul_259/x" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_47" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_260/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_260" + op: "Mul" + input: "Mul_260/x" + input: "Square_47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_162" + op: "Add" + input: "Mul_259" + input: "Mul_260" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_47" + op: "Sqrt" + input: "add_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_163/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_163" + op: "Add" + input: "Sqrt_47" + input: "add_163/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_48" + op: "RealDiv" + input: "add_161" + input: "add_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_261/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_261" + op: "Mul" + input: "mul_261/x" + input: "bert/encoder/layer_2/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_164" + op: "Add" + input: "truediv_48" + input: "mul_261" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_262" + op: "Mul" + input: "add" + input: "add_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_48" + op: "Sub" + input: "bert/encoder/layer_2/intermediate/dense/kernel/read" + input: "mul_262" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_340" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + input: "sub_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_341" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + input: "add_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_342" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + input: "add_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_263/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_263" + op: "Mul" + input: "Mul_263/x" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_264/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_264" + op: "Mul" + input: "Mul_264/x" + input: "clip_by_global_norm/clip_by_global_norm/_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_165" + op: "Add" + input: "Mul_263" + input: "Mul_264" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_265/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_265" + op: "Mul" + input: "Mul_265/x" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_48" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_266/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_266" + op: "Mul" + input: "Mul_266/x" + input: "Square_48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_166" + op: "Add" + input: "Mul_265" + input: "Mul_266" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_48" + op: "Sqrt" + input: "add_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_167/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_167" + op: "Add" + input: "Sqrt_48" + input: "add_167/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_49" + op: "RealDiv" + input: "add_165" + input: "add_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_267" + op: "Mul" + input: "add" + input: "truediv_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_49" + op: "Sub" + input: "bert/encoder/layer_2/intermediate/dense/bias/read" + input: "mul_267" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_343" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias" + input: "sub_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_344" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + input: "add_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_345" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + input: "add_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_268/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_268" + op: "Mul" + input: "Mul_268/x" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_269/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_269" + op: "Mul" + input: "Mul_269/x" + input: "clip_by_global_norm/clip_by_global_norm/_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_168" + op: "Add" + input: "Mul_268" + input: "Mul_269" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_270/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_270" + op: "Mul" + input: "Mul_270/x" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_49" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_271/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_271" + op: "Mul" + input: "Mul_271/x" + input: "Square_49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_169" + op: "Add" + input: "Mul_270" + input: "Mul_271" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_49" + op: "Sqrt" + input: "add_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_170/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_170" + op: "Add" + input: "Sqrt_49" + input: "add_170/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_50" + op: "RealDiv" + input: "add_168" + input: "add_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_272/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_272" + op: "Mul" + input: "mul_272/x" + input: "bert/encoder/layer_2/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_171" + op: "Add" + input: "truediv_50" + input: "mul_272" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_273" + op: "Mul" + input: "add" + input: "add_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_50" + op: "Sub" + input: "bert/encoder/layer_2/output/dense/kernel/read" + input: "mul_273" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_346" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel" + input: "sub_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_347" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + input: "add_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_348" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + input: "add_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + input: "bert/encoder/layer_2/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + input: "bert/encoder/layer_2/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_274/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_274" + op: "Mul" + input: "Mul_274/x" + input: "bert/encoder/layer_2/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_275/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_275" + op: "Mul" + input: "Mul_275/x" + input: "clip_by_global_norm/clip_by_global_norm/_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_172" + op: "Add" + input: "Mul_274" + input: "Mul_275" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_276/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_276" + op: "Mul" + input: "Mul_276/x" + input: "bert/encoder/layer_2/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_50" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_277/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_277" + op: "Mul" + input: "Mul_277/x" + input: "Square_50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_173" + op: "Add" + input: "Mul_276" + input: "Mul_277" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_50" + op: "Sqrt" + input: "add_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_174/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_174" + op: "Add" + input: "Sqrt_50" + input: "add_174/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_51" + op: "RealDiv" + input: "add_172" + input: "add_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_278" + op: "Mul" + input: "add" + input: "truediv_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_51" + op: "Sub" + input: "bert/encoder/layer_2/output/dense/bias/read" + input: "mul_278" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_349" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias" + input: "sub_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_350" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + input: "add_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_351" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + input: "add_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_279/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_279" + op: "Mul" + input: "Mul_279/x" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_280/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_280" + op: "Mul" + input: "Mul_280/x" + input: "clip_by_global_norm/clip_by_global_norm/_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_175" + op: "Add" + input: "Mul_279" + input: "Mul_280" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_281/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_281" + op: "Mul" + input: "Mul_281/x" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_51" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_282/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_282" + op: "Mul" + input: "Mul_282/x" + input: "Square_51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_176" + op: "Add" + input: "Mul_281" + input: "Mul_282" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_51" + op: "Sqrt" + input: "add_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_177/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_177" + op: "Add" + input: "Sqrt_51" + input: "add_177/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_52" + op: "RealDiv" + input: "add_175" + input: "add_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_283" + op: "Mul" + input: "add" + input: "truediv_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_52" + op: "Sub" + input: "bert/encoder/layer_2/output/LayerNorm/beta/read" + input: "mul_283" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_352" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + input: "sub_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_353" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + input: "add_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_354" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + input: "add_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_284/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_284" + op: "Mul" + input: "Mul_284/x" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_285/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_285" + op: "Mul" + input: "Mul_285/x" + input: "clip_by_global_norm/clip_by_global_norm/_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_178" + op: "Add" + input: "Mul_284" + input: "Mul_285" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_286/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_286" + op: "Mul" + input: "Mul_286/x" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_52" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_287/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_287" + op: "Mul" + input: "Mul_287/x" + input: "Square_52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_179" + op: "Add" + input: "Mul_286" + input: "Mul_287" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_52" + op: "Sqrt" + input: "add_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_180/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_180" + op: "Add" + input: "Sqrt_52" + input: "add_180/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_53" + op: "RealDiv" + input: "add_178" + input: "add_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_288" + op: "Mul" + input: "add" + input: "truediv_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_53" + op: "Sub" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/read" + input: "mul_288" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_355" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + input: "sub_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_356" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + input: "add_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_357" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + input: "add_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_289/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_289" + op: "Mul" + input: "Mul_289/x" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_290/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_290" + op: "Mul" + input: "Mul_290/x" + input: "clip_by_global_norm/clip_by_global_norm/_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_181" + op: "Add" + input: "Mul_289" + input: "Mul_290" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_291/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_291" + op: "Mul" + input: "Mul_291/x" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_53" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_292/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_292" + op: "Mul" + input: "Mul_292/x" + input: "Square_53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_182" + op: "Add" + input: "Mul_291" + input: "Mul_292" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_53" + op: "Sqrt" + input: "add_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_183/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_183" + op: "Add" + input: "Sqrt_53" + input: "add_183/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_54" + op: "RealDiv" + input: "add_181" + input: "add_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_293/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_293" + op: "Mul" + input: "mul_293/x" + input: "bert/encoder/layer_3/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_184" + op: "Add" + input: "truediv_54" + input: "mul_293" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_294" + op: "Mul" + input: "add" + input: "add_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_54" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/query/kernel/read" + input: "mul_294" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_358" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel" + input: "sub_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_359" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + input: "add_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_360" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + input: "add_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_295/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_295" + op: "Mul" + input: "Mul_295/x" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_296/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_296" + op: "Mul" + input: "Mul_296/x" + input: "clip_by_global_norm/clip_by_global_norm/_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_185" + op: "Add" + input: "Mul_295" + input: "Mul_296" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_297/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_297" + op: "Mul" + input: "Mul_297/x" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_54" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_298/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_298" + op: "Mul" + input: "Mul_298/x" + input: "Square_54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_186" + op: "Add" + input: "Mul_297" + input: "Mul_298" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_54" + op: "Sqrt" + input: "add_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_187/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_187" + op: "Add" + input: "Sqrt_54" + input: "add_187/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_55" + op: "RealDiv" + input: "add_185" + input: "add_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_299" + op: "Mul" + input: "add" + input: "truediv_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_55" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/query/bias/read" + input: "mul_299" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_361" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias" + input: "sub_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_362" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + input: "add_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_363" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + input: "add_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_300/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_300" + op: "Mul" + input: "Mul_300/x" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_301/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_301" + op: "Mul" + input: "Mul_301/x" + input: "clip_by_global_norm/clip_by_global_norm/_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_188" + op: "Add" + input: "Mul_300" + input: "Mul_301" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_302/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_302" + op: "Mul" + input: "Mul_302/x" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_55" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_303/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_303" + op: "Mul" + input: "Mul_303/x" + input: "Square_55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_189" + op: "Add" + input: "Mul_302" + input: "Mul_303" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_55" + op: "Sqrt" + input: "add_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_190/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_190" + op: "Add" + input: "Sqrt_55" + input: "add_190/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_56" + op: "RealDiv" + input: "add_188" + input: "add_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_304/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_304" + op: "Mul" + input: "mul_304/x" + input: "bert/encoder/layer_3/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_191" + op: "Add" + input: "truediv_56" + input: "mul_304" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_305" + op: "Mul" + input: "add" + input: "add_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_56" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/key/kernel/read" + input: "mul_305" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_364" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel" + input: "sub_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_365" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + input: "add_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_366" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + input: "add_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_306/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_306" + op: "Mul" + input: "Mul_306/x" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_307/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_307" + op: "Mul" + input: "Mul_307/x" + input: "clip_by_global_norm/clip_by_global_norm/_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_192" + op: "Add" + input: "Mul_306" + input: "Mul_307" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_308/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_308" + op: "Mul" + input: "Mul_308/x" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_56" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_309/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_309" + op: "Mul" + input: "Mul_309/x" + input: "Square_56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_193" + op: "Add" + input: "Mul_308" + input: "Mul_309" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_56" + op: "Sqrt" + input: "add_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_194/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_194" + op: "Add" + input: "Sqrt_56" + input: "add_194/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_57" + op: "RealDiv" + input: "add_192" + input: "add_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_310" + op: "Mul" + input: "add" + input: "truediv_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_57" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/key/bias/read" + input: "mul_310" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_367" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias" + input: "sub_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_368" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + input: "add_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_369" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + input: "add_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_311/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_311" + op: "Mul" + input: "Mul_311/x" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_312/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_312" + op: "Mul" + input: "Mul_312/x" + input: "clip_by_global_norm/clip_by_global_norm/_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_195" + op: "Add" + input: "Mul_311" + input: "Mul_312" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_313/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_313" + op: "Mul" + input: "Mul_313/x" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_57" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_314/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_314" + op: "Mul" + input: "Mul_314/x" + input: "Square_57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_196" + op: "Add" + input: "Mul_313" + input: "Mul_314" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_57" + op: "Sqrt" + input: "add_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_197/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_197" + op: "Add" + input: "Sqrt_57" + input: "add_197/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_58" + op: "RealDiv" + input: "add_195" + input: "add_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_315/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_315" + op: "Mul" + input: "mul_315/x" + input: "bert/encoder/layer_3/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_198" + op: "Add" + input: "truediv_58" + input: "mul_315" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_316" + op: "Mul" + input: "add" + input: "add_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_58" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/value/kernel/read" + input: "mul_316" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_370" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel" + input: "sub_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_371" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + input: "add_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_372" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + input: "add_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_317/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_317" + op: "Mul" + input: "Mul_317/x" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_318/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_318" + op: "Mul" + input: "Mul_318/x" + input: "clip_by_global_norm/clip_by_global_norm/_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_199" + op: "Add" + input: "Mul_317" + input: "Mul_318" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_319/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_319" + op: "Mul" + input: "Mul_319/x" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_58" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_320/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_320" + op: "Mul" + input: "Mul_320/x" + input: "Square_58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_200" + op: "Add" + input: "Mul_319" + input: "Mul_320" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_58" + op: "Sqrt" + input: "add_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_201/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_201" + op: "Add" + input: "Sqrt_58" + input: "add_201/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_59" + op: "RealDiv" + input: "add_199" + input: "add_201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_321" + op: "Mul" + input: "add" + input: "truediv_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_59" + op: "Sub" + input: "bert/encoder/layer_3/attention/self/value/bias/read" + input: "mul_321" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_373" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias" + input: "sub_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_374" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + input: "add_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_375" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + input: "add_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_322/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_322" + op: "Mul" + input: "Mul_322/x" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_323/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_323" + op: "Mul" + input: "Mul_323/x" + input: "clip_by_global_norm/clip_by_global_norm/_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_202" + op: "Add" + input: "Mul_322" + input: "Mul_323" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_324/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_324" + op: "Mul" + input: "Mul_324/x" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_59" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_325/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_325" + op: "Mul" + input: "Mul_325/x" + input: "Square_59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_203" + op: "Add" + input: "Mul_324" + input: "Mul_325" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_59" + op: "Sqrt" + input: "add_203" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_204/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_204" + op: "Add" + input: "Sqrt_59" + input: "add_204/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_60" + op: "RealDiv" + input: "add_202" + input: "add_204" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_326/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_326" + op: "Mul" + input: "mul_326/x" + input: "bert/encoder/layer_3/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_205" + op: "Add" + input: "truediv_60" + input: "mul_326" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_327" + op: "Mul" + input: "add" + input: "add_205" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_60" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/dense/kernel/read" + input: "mul_327" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_376" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + input: "sub_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_377" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + input: "add_202" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_378" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + input: "add_203" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_328/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_328" + op: "Mul" + input: "Mul_328/x" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_329/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_329" + op: "Mul" + input: "Mul_329/x" + input: "clip_by_global_norm/clip_by_global_norm/_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_206" + op: "Add" + input: "Mul_328" + input: "Mul_329" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_330/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_330" + op: "Mul" + input: "Mul_330/x" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_60" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_331/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_331" + op: "Mul" + input: "Mul_331/x" + input: "Square_60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_207" + op: "Add" + input: "Mul_330" + input: "Mul_331" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_60" + op: "Sqrt" + input: "add_207" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_208/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_208" + op: "Add" + input: "Sqrt_60" + input: "add_208/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_61" + op: "RealDiv" + input: "add_206" + input: "add_208" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_332" + op: "Mul" + input: "add" + input: "truediv_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_61" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/dense/bias/read" + input: "mul_332" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_379" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias" + input: "sub_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_380" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + input: "add_206" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_381" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + input: "add_207" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_333/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_333" + op: "Mul" + input: "Mul_333/x" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_334/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_334" + op: "Mul" + input: "Mul_334/x" + input: "clip_by_global_norm/clip_by_global_norm/_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_209" + op: "Add" + input: "Mul_333" + input: "Mul_334" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_335/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_335" + op: "Mul" + input: "Mul_335/x" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_61" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_336/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_336" + op: "Mul" + input: "Mul_336/x" + input: "Square_61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_210" + op: "Add" + input: "Mul_335" + input: "Mul_336" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_61" + op: "Sqrt" + input: "add_210" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_211/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_211" + op: "Add" + input: "Sqrt_61" + input: "add_211/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_62" + op: "RealDiv" + input: "add_209" + input: "add_211" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_337" + op: "Mul" + input: "add" + input: "truediv_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_62" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/read" + input: "mul_337" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_382" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + input: "sub_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_383" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + input: "add_209" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_384" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + input: "add_210" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_338/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_338" + op: "Mul" + input: "Mul_338/x" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_339/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_339" + op: "Mul" + input: "Mul_339/x" + input: "clip_by_global_norm/clip_by_global_norm/_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_212" + op: "Add" + input: "Mul_338" + input: "Mul_339" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_340/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_340" + op: "Mul" + input: "Mul_340/x" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_62" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_341/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_341" + op: "Mul" + input: "Mul_341/x" + input: "Square_62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_213" + op: "Add" + input: "Mul_340" + input: "Mul_341" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_62" + op: "Sqrt" + input: "add_213" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_214/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_214" + op: "Add" + input: "Sqrt_62" + input: "add_214/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_63" + op: "RealDiv" + input: "add_212" + input: "add_214" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_342" + op: "Mul" + input: "add" + input: "truediv_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_63" + op: "Sub" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/read" + input: "mul_342" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_385" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + input: "sub_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_386" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + input: "add_212" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_387" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + input: "add_213" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_343/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_343" + op: "Mul" + input: "Mul_343/x" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_344/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_344" + op: "Mul" + input: "Mul_344/x" + input: "clip_by_global_norm/clip_by_global_norm/_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_215" + op: "Add" + input: "Mul_343" + input: "Mul_344" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_345/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_345" + op: "Mul" + input: "Mul_345/x" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_63" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_346/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_346" + op: "Mul" + input: "Mul_346/x" + input: "Square_63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_216" + op: "Add" + input: "Mul_345" + input: "Mul_346" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_63" + op: "Sqrt" + input: "add_216" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_217/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_217" + op: "Add" + input: "Sqrt_63" + input: "add_217/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_64" + op: "RealDiv" + input: "add_215" + input: "add_217" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_347/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_347" + op: "Mul" + input: "mul_347/x" + input: "bert/encoder/layer_3/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_218" + op: "Add" + input: "truediv_64" + input: "mul_347" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_348" + op: "Mul" + input: "add" + input: "add_218" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_64" + op: "Sub" + input: "bert/encoder/layer_3/intermediate/dense/kernel/read" + input: "mul_348" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_388" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + input: "sub_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_389" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + input: "add_215" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_390" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + input: "add_216" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_349/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_349" + op: "Mul" + input: "Mul_349/x" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_350/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_350" + op: "Mul" + input: "Mul_350/x" + input: "clip_by_global_norm/clip_by_global_norm/_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_219" + op: "Add" + input: "Mul_349" + input: "Mul_350" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_351/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_351" + op: "Mul" + input: "Mul_351/x" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_64" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_352/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_352" + op: "Mul" + input: "Mul_352/x" + input: "Square_64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_220" + op: "Add" + input: "Mul_351" + input: "Mul_352" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_64" + op: "Sqrt" + input: "add_220" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_221/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_221" + op: "Add" + input: "Sqrt_64" + input: "add_221/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_65" + op: "RealDiv" + input: "add_219" + input: "add_221" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_353" + op: "Mul" + input: "add" + input: "truediv_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_65" + op: "Sub" + input: "bert/encoder/layer_3/intermediate/dense/bias/read" + input: "mul_353" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_391" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias" + input: "sub_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_392" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + input: "add_219" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_393" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + input: "add_220" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_354/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_354" + op: "Mul" + input: "Mul_354/x" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_355/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_355" + op: "Mul" + input: "Mul_355/x" + input: "clip_by_global_norm/clip_by_global_norm/_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_222" + op: "Add" + input: "Mul_354" + input: "Mul_355" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_356/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_356" + op: "Mul" + input: "Mul_356/x" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_65" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_357/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_357" + op: "Mul" + input: "Mul_357/x" + input: "Square_65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_223" + op: "Add" + input: "Mul_356" + input: "Mul_357" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_65" + op: "Sqrt" + input: "add_223" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_224/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_224" + op: "Add" + input: "Sqrt_65" + input: "add_224/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_66" + op: "RealDiv" + input: "add_222" + input: "add_224" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_358/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_358" + op: "Mul" + input: "mul_358/x" + input: "bert/encoder/layer_3/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_225" + op: "Add" + input: "truediv_66" + input: "mul_358" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_359" + op: "Mul" + input: "add" + input: "add_225" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_66" + op: "Sub" + input: "bert/encoder/layer_3/output/dense/kernel/read" + input: "mul_359" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_394" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel" + input: "sub_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_395" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + input: "add_222" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_396" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + input: "add_223" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + input: "bert/encoder/layer_3/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + input: "bert/encoder/layer_3/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_360/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_360" + op: "Mul" + input: "Mul_360/x" + input: "bert/encoder/layer_3/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_361/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_361" + op: "Mul" + input: "Mul_361/x" + input: "clip_by_global_norm/clip_by_global_norm/_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_226" + op: "Add" + input: "Mul_360" + input: "Mul_361" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_362/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_362" + op: "Mul" + input: "Mul_362/x" + input: "bert/encoder/layer_3/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_66" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_363/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_363" + op: "Mul" + input: "Mul_363/x" + input: "Square_66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_227" + op: "Add" + input: "Mul_362" + input: "Mul_363" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_66" + op: "Sqrt" + input: "add_227" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_228/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_228" + op: "Add" + input: "Sqrt_66" + input: "add_228/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_67" + op: "RealDiv" + input: "add_226" + input: "add_228" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_364" + op: "Mul" + input: "add" + input: "truediv_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_67" + op: "Sub" + input: "bert/encoder/layer_3/output/dense/bias/read" + input: "mul_364" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_397" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias" + input: "sub_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_398" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + input: "add_226" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_399" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + input: "add_227" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_365/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_365" + op: "Mul" + input: "Mul_365/x" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_366/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_366" + op: "Mul" + input: "Mul_366/x" + input: "clip_by_global_norm/clip_by_global_norm/_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_229" + op: "Add" + input: "Mul_365" + input: "Mul_366" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_367/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_367" + op: "Mul" + input: "Mul_367/x" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_67" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_368/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_368" + op: "Mul" + input: "Mul_368/x" + input: "Square_67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_230" + op: "Add" + input: "Mul_367" + input: "Mul_368" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_67" + op: "Sqrt" + input: "add_230" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_231/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_231" + op: "Add" + input: "Sqrt_67" + input: "add_231/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_68" + op: "RealDiv" + input: "add_229" + input: "add_231" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_369" + op: "Mul" + input: "add" + input: "truediv_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_68" + op: "Sub" + input: "bert/encoder/layer_3/output/LayerNorm/beta/read" + input: "mul_369" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_400" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + input: "sub_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_401" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + input: "add_229" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_402" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + input: "add_230" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_370/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_370" + op: "Mul" + input: "Mul_370/x" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_371/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_371" + op: "Mul" + input: "Mul_371/x" + input: "clip_by_global_norm/clip_by_global_norm/_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_232" + op: "Add" + input: "Mul_370" + input: "Mul_371" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_372/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_372" + op: "Mul" + input: "Mul_372/x" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_68" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_373/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_373" + op: "Mul" + input: "Mul_373/x" + input: "Square_68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_233" + op: "Add" + input: "Mul_372" + input: "Mul_373" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_68" + op: "Sqrt" + input: "add_233" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_234/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_234" + op: "Add" + input: "Sqrt_68" + input: "add_234/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_69" + op: "RealDiv" + input: "add_232" + input: "add_234" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_374" + op: "Mul" + input: "add" + input: "truediv_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_69" + op: "Sub" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/read" + input: "mul_374" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_403" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + input: "sub_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_404" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + input: "add_232" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_405" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + input: "add_233" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_375/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_375" + op: "Mul" + input: "Mul_375/x" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_376/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_376" + op: "Mul" + input: "Mul_376/x" + input: "clip_by_global_norm/clip_by_global_norm/_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_235" + op: "Add" + input: "Mul_375" + input: "Mul_376" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_377/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_377" + op: "Mul" + input: "Mul_377/x" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_69" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_378/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_378" + op: "Mul" + input: "Mul_378/x" + input: "Square_69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_236" + op: "Add" + input: "Mul_377" + input: "Mul_378" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_69" + op: "Sqrt" + input: "add_236" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_237/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_237" + op: "Add" + input: "Sqrt_69" + input: "add_237/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_70" + op: "RealDiv" + input: "add_235" + input: "add_237" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_379/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_379" + op: "Mul" + input: "mul_379/x" + input: "bert/encoder/layer_4/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_238" + op: "Add" + input: "truediv_70" + input: "mul_379" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_380" + op: "Mul" + input: "add" + input: "add_238" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_70" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/query/kernel/read" + input: "mul_380" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_406" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel" + input: "sub_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_407" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + input: "add_235" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_408" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + input: "add_236" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_381/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_381" + op: "Mul" + input: "Mul_381/x" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_382/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_382" + op: "Mul" + input: "Mul_382/x" + input: "clip_by_global_norm/clip_by_global_norm/_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_239" + op: "Add" + input: "Mul_381" + input: "Mul_382" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_383/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_383" + op: "Mul" + input: "Mul_383/x" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_70" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_384/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_384" + op: "Mul" + input: "Mul_384/x" + input: "Square_70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_240" + op: "Add" + input: "Mul_383" + input: "Mul_384" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_70" + op: "Sqrt" + input: "add_240" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_241/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_241" + op: "Add" + input: "Sqrt_70" + input: "add_241/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_71" + op: "RealDiv" + input: "add_239" + input: "add_241" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_385" + op: "Mul" + input: "add" + input: "truediv_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_71" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/query/bias/read" + input: "mul_385" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_409" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias" + input: "sub_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_410" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + input: "add_239" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_411" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + input: "add_240" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_386/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_386" + op: "Mul" + input: "Mul_386/x" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_387/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_387" + op: "Mul" + input: "Mul_387/x" + input: "clip_by_global_norm/clip_by_global_norm/_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_242" + op: "Add" + input: "Mul_386" + input: "Mul_387" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_388/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_388" + op: "Mul" + input: "Mul_388/x" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_71" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_389/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_389" + op: "Mul" + input: "Mul_389/x" + input: "Square_71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_243" + op: "Add" + input: "Mul_388" + input: "Mul_389" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_71" + op: "Sqrt" + input: "add_243" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_244/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_244" + op: "Add" + input: "Sqrt_71" + input: "add_244/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_72" + op: "RealDiv" + input: "add_242" + input: "add_244" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_390/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_390" + op: "Mul" + input: "mul_390/x" + input: "bert/encoder/layer_4/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_245" + op: "Add" + input: "truediv_72" + input: "mul_390" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_391" + op: "Mul" + input: "add" + input: "add_245" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_72" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/key/kernel/read" + input: "mul_391" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_412" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel" + input: "sub_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_413" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + input: "add_242" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_414" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + input: "add_243" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_392/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_392" + op: "Mul" + input: "Mul_392/x" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_393/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_393" + op: "Mul" + input: "Mul_393/x" + input: "clip_by_global_norm/clip_by_global_norm/_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_246" + op: "Add" + input: "Mul_392" + input: "Mul_393" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_394/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_394" + op: "Mul" + input: "Mul_394/x" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_72" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_395/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_395" + op: "Mul" + input: "Mul_395/x" + input: "Square_72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_247" + op: "Add" + input: "Mul_394" + input: "Mul_395" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_72" + op: "Sqrt" + input: "add_247" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_248/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_248" + op: "Add" + input: "Sqrt_72" + input: "add_248/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_73" + op: "RealDiv" + input: "add_246" + input: "add_248" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_396" + op: "Mul" + input: "add" + input: "truediv_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_73" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/key/bias/read" + input: "mul_396" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_415" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias" + input: "sub_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_416" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + input: "add_246" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_417" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + input: "add_247" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_397/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_397" + op: "Mul" + input: "Mul_397/x" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_398/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_398" + op: "Mul" + input: "Mul_398/x" + input: "clip_by_global_norm/clip_by_global_norm/_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_249" + op: "Add" + input: "Mul_397" + input: "Mul_398" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_399/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_399" + op: "Mul" + input: "Mul_399/x" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_73" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_400/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_400" + op: "Mul" + input: "Mul_400/x" + input: "Square_73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_250" + op: "Add" + input: "Mul_399" + input: "Mul_400" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_73" + op: "Sqrt" + input: "add_250" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_251/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_251" + op: "Add" + input: "Sqrt_73" + input: "add_251/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_74" + op: "RealDiv" + input: "add_249" + input: "add_251" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_401/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_401" + op: "Mul" + input: "mul_401/x" + input: "bert/encoder/layer_4/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_252" + op: "Add" + input: "truediv_74" + input: "mul_401" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_402" + op: "Mul" + input: "add" + input: "add_252" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_74" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/value/kernel/read" + input: "mul_402" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_418" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel" + input: "sub_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_419" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + input: "add_249" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_420" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + input: "add_250" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_403/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_403" + op: "Mul" + input: "Mul_403/x" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_404/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_404" + op: "Mul" + input: "Mul_404/x" + input: "clip_by_global_norm/clip_by_global_norm/_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_253" + op: "Add" + input: "Mul_403" + input: "Mul_404" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_405/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_405" + op: "Mul" + input: "Mul_405/x" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_74" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_406/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_406" + op: "Mul" + input: "Mul_406/x" + input: "Square_74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_254" + op: "Add" + input: "Mul_405" + input: "Mul_406" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_74" + op: "Sqrt" + input: "add_254" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_255/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_255" + op: "Add" + input: "Sqrt_74" + input: "add_255/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_75" + op: "RealDiv" + input: "add_253" + input: "add_255" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_407" + op: "Mul" + input: "add" + input: "truediv_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_75" + op: "Sub" + input: "bert/encoder/layer_4/attention/self/value/bias/read" + input: "mul_407" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_421" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias" + input: "sub_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_422" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + input: "add_253" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_423" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + input: "add_254" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_408/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_408" + op: "Mul" + input: "Mul_408/x" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_409/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_409" + op: "Mul" + input: "Mul_409/x" + input: "clip_by_global_norm/clip_by_global_norm/_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_256" + op: "Add" + input: "Mul_408" + input: "Mul_409" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_410/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_410" + op: "Mul" + input: "Mul_410/x" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_75" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_411/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_411" + op: "Mul" + input: "Mul_411/x" + input: "Square_75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_257" + op: "Add" + input: "Mul_410" + input: "Mul_411" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_75" + op: "Sqrt" + input: "add_257" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_258/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_258" + op: "Add" + input: "Sqrt_75" + input: "add_258/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_76" + op: "RealDiv" + input: "add_256" + input: "add_258" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_412/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_412" + op: "Mul" + input: "mul_412/x" + input: "bert/encoder/layer_4/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_259" + op: "Add" + input: "truediv_76" + input: "mul_412" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_413" + op: "Mul" + input: "add" + input: "add_259" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_76" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/dense/kernel/read" + input: "mul_413" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_424" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + input: "sub_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_425" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + input: "add_256" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_426" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + input: "add_257" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_414/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_414" + op: "Mul" + input: "Mul_414/x" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_415/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_415" + op: "Mul" + input: "Mul_415/x" + input: "clip_by_global_norm/clip_by_global_norm/_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_260" + op: "Add" + input: "Mul_414" + input: "Mul_415" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_416/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_416" + op: "Mul" + input: "Mul_416/x" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_76" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_417/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_417" + op: "Mul" + input: "Mul_417/x" + input: "Square_76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_261" + op: "Add" + input: "Mul_416" + input: "Mul_417" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_76" + op: "Sqrt" + input: "add_261" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_262/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_262" + op: "Add" + input: "Sqrt_76" + input: "add_262/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_77" + op: "RealDiv" + input: "add_260" + input: "add_262" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_418" + op: "Mul" + input: "add" + input: "truediv_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_77" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/dense/bias/read" + input: "mul_418" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_427" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias" + input: "sub_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_428" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + input: "add_260" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_429" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + input: "add_261" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_419/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_419" + op: "Mul" + input: "Mul_419/x" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_420/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_420" + op: "Mul" + input: "Mul_420/x" + input: "clip_by_global_norm/clip_by_global_norm/_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_263" + op: "Add" + input: "Mul_419" + input: "Mul_420" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_421/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_421" + op: "Mul" + input: "Mul_421/x" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_77" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_422/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_422" + op: "Mul" + input: "Mul_422/x" + input: "Square_77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_264" + op: "Add" + input: "Mul_421" + input: "Mul_422" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_77" + op: "Sqrt" + input: "add_264" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_265/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_265" + op: "Add" + input: "Sqrt_77" + input: "add_265/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_78" + op: "RealDiv" + input: "add_263" + input: "add_265" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_423" + op: "Mul" + input: "add" + input: "truediv_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_78" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/read" + input: "mul_423" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_430" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + input: "sub_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_431" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + input: "add_263" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_432" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + input: "add_264" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_424/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_424" + op: "Mul" + input: "Mul_424/x" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_425/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_425" + op: "Mul" + input: "Mul_425/x" + input: "clip_by_global_norm/clip_by_global_norm/_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_266" + op: "Add" + input: "Mul_424" + input: "Mul_425" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_426/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_426" + op: "Mul" + input: "Mul_426/x" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_78" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_427/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_427" + op: "Mul" + input: "Mul_427/x" + input: "Square_78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_267" + op: "Add" + input: "Mul_426" + input: "Mul_427" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_78" + op: "Sqrt" + input: "add_267" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_268/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_268" + op: "Add" + input: "Sqrt_78" + input: "add_268/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_79" + op: "RealDiv" + input: "add_266" + input: "add_268" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_428" + op: "Mul" + input: "add" + input: "truediv_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_79" + op: "Sub" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/read" + input: "mul_428" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_433" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + input: "sub_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_434" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + input: "add_266" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_435" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + input: "add_267" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_429/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_429" + op: "Mul" + input: "Mul_429/x" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_430/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_430" + op: "Mul" + input: "Mul_430/x" + input: "clip_by_global_norm/clip_by_global_norm/_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_269" + op: "Add" + input: "Mul_429" + input: "Mul_430" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_431/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_431" + op: "Mul" + input: "Mul_431/x" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_79" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_432/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_432" + op: "Mul" + input: "Mul_432/x" + input: "Square_79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_270" + op: "Add" + input: "Mul_431" + input: "Mul_432" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_79" + op: "Sqrt" + input: "add_270" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_271/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_271" + op: "Add" + input: "Sqrt_79" + input: "add_271/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_80" + op: "RealDiv" + input: "add_269" + input: "add_271" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_433/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_433" + op: "Mul" + input: "mul_433/x" + input: "bert/encoder/layer_4/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_272" + op: "Add" + input: "truediv_80" + input: "mul_433" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_434" + op: "Mul" + input: "add" + input: "add_272" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_80" + op: "Sub" + input: "bert/encoder/layer_4/intermediate/dense/kernel/read" + input: "mul_434" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_436" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + input: "sub_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_437" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + input: "add_269" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_438" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + input: "add_270" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_435/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_435" + op: "Mul" + input: "Mul_435/x" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_436/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_436" + op: "Mul" + input: "Mul_436/x" + input: "clip_by_global_norm/clip_by_global_norm/_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_273" + op: "Add" + input: "Mul_435" + input: "Mul_436" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_437/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_437" + op: "Mul" + input: "Mul_437/x" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_80" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_438/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_438" + op: "Mul" + input: "Mul_438/x" + input: "Square_80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_274" + op: "Add" + input: "Mul_437" + input: "Mul_438" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_80" + op: "Sqrt" + input: "add_274" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_275/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_275" + op: "Add" + input: "Sqrt_80" + input: "add_275/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_81" + op: "RealDiv" + input: "add_273" + input: "add_275" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_439" + op: "Mul" + input: "add" + input: "truediv_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_81" + op: "Sub" + input: "bert/encoder/layer_4/intermediate/dense/bias/read" + input: "mul_439" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_439" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias" + input: "sub_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_440" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + input: "add_273" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_441" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + input: "add_274" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_440/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_440" + op: "Mul" + input: "Mul_440/x" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_441/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_441" + op: "Mul" + input: "Mul_441/x" + input: "clip_by_global_norm/clip_by_global_norm/_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_276" + op: "Add" + input: "Mul_440" + input: "Mul_441" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_442/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_442" + op: "Mul" + input: "Mul_442/x" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_81" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_443/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_443" + op: "Mul" + input: "Mul_443/x" + input: "Square_81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_277" + op: "Add" + input: "Mul_442" + input: "Mul_443" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_81" + op: "Sqrt" + input: "add_277" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_278/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_278" + op: "Add" + input: "Sqrt_81" + input: "add_278/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_82" + op: "RealDiv" + input: "add_276" + input: "add_278" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_444/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_444" + op: "Mul" + input: "mul_444/x" + input: "bert/encoder/layer_4/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_279" + op: "Add" + input: "truediv_82" + input: "mul_444" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_445" + op: "Mul" + input: "add" + input: "add_279" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_82" + op: "Sub" + input: "bert/encoder/layer_4/output/dense/kernel/read" + input: "mul_445" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_442" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel" + input: "sub_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_443" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + input: "add_276" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_444" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + input: "add_277" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + input: "bert/encoder/layer_4/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + input: "bert/encoder/layer_4/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_446/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_446" + op: "Mul" + input: "Mul_446/x" + input: "bert/encoder/layer_4/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_447/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_447" + op: "Mul" + input: "Mul_447/x" + input: "clip_by_global_norm/clip_by_global_norm/_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_280" + op: "Add" + input: "Mul_446" + input: "Mul_447" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_448/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_448" + op: "Mul" + input: "Mul_448/x" + input: "bert/encoder/layer_4/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_82" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_449/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_449" + op: "Mul" + input: "Mul_449/x" + input: "Square_82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_281" + op: "Add" + input: "Mul_448" + input: "Mul_449" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_82" + op: "Sqrt" + input: "add_281" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_282/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_282" + op: "Add" + input: "Sqrt_82" + input: "add_282/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_83" + op: "RealDiv" + input: "add_280" + input: "add_282" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_450" + op: "Mul" + input: "add" + input: "truediv_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_83" + op: "Sub" + input: "bert/encoder/layer_4/output/dense/bias/read" + input: "mul_450" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_445" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias" + input: "sub_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_446" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + input: "add_280" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_447" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + input: "add_281" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_451/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_451" + op: "Mul" + input: "Mul_451/x" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_452/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_452" + op: "Mul" + input: "Mul_452/x" + input: "clip_by_global_norm/clip_by_global_norm/_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_283" + op: "Add" + input: "Mul_451" + input: "Mul_452" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_453/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_453" + op: "Mul" + input: "Mul_453/x" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_83" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_454/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_454" + op: "Mul" + input: "Mul_454/x" + input: "Square_83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_284" + op: "Add" + input: "Mul_453" + input: "Mul_454" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_83" + op: "Sqrt" + input: "add_284" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_285/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_285" + op: "Add" + input: "Sqrt_83" + input: "add_285/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_84" + op: "RealDiv" + input: "add_283" + input: "add_285" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_455" + op: "Mul" + input: "add" + input: "truediv_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_84" + op: "Sub" + input: "bert/encoder/layer_4/output/LayerNorm/beta/read" + input: "mul_455" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_448" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + input: "sub_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_449" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + input: "add_283" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_450" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + input: "add_284" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_456/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_456" + op: "Mul" + input: "Mul_456/x" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_457/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_457" + op: "Mul" + input: "Mul_457/x" + input: "clip_by_global_norm/clip_by_global_norm/_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_286" + op: "Add" + input: "Mul_456" + input: "Mul_457" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_458/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_458" + op: "Mul" + input: "Mul_458/x" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_84" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_459/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_459" + op: "Mul" + input: "Mul_459/x" + input: "Square_84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_287" + op: "Add" + input: "Mul_458" + input: "Mul_459" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_84" + op: "Sqrt" + input: "add_287" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_288/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_288" + op: "Add" + input: "Sqrt_84" + input: "add_288/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_85" + op: "RealDiv" + input: "add_286" + input: "add_288" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_460" + op: "Mul" + input: "add" + input: "truediv_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_85" + op: "Sub" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/read" + input: "mul_460" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_451" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + input: "sub_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_452" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + input: "add_286" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_453" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + input: "add_287" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_461/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_461" + op: "Mul" + input: "Mul_461/x" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_462/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_462" + op: "Mul" + input: "Mul_462/x" + input: "clip_by_global_norm/clip_by_global_norm/_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_289" + op: "Add" + input: "Mul_461" + input: "Mul_462" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_463/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_463" + op: "Mul" + input: "Mul_463/x" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_85" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_464/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_464" + op: "Mul" + input: "Mul_464/x" + input: "Square_85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_290" + op: "Add" + input: "Mul_463" + input: "Mul_464" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_85" + op: "Sqrt" + input: "add_290" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_291/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_291" + op: "Add" + input: "Sqrt_85" + input: "add_291/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_86" + op: "RealDiv" + input: "add_289" + input: "add_291" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_465/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_465" + op: "Mul" + input: "mul_465/x" + input: "bert/encoder/layer_5/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_292" + op: "Add" + input: "truediv_86" + input: "mul_465" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_466" + op: "Mul" + input: "add" + input: "add_292" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_86" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/query/kernel/read" + input: "mul_466" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_454" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel" + input: "sub_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_455" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + input: "add_289" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_456" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + input: "add_290" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_467/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_467" + op: "Mul" + input: "Mul_467/x" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_468/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_468" + op: "Mul" + input: "Mul_468/x" + input: "clip_by_global_norm/clip_by_global_norm/_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_293" + op: "Add" + input: "Mul_467" + input: "Mul_468" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_469/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_469" + op: "Mul" + input: "Mul_469/x" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_86" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_470/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_470" + op: "Mul" + input: "Mul_470/x" + input: "Square_86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_294" + op: "Add" + input: "Mul_469" + input: "Mul_470" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_86" + op: "Sqrt" + input: "add_294" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_295/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_295" + op: "Add" + input: "Sqrt_86" + input: "add_295/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_87" + op: "RealDiv" + input: "add_293" + input: "add_295" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_471" + op: "Mul" + input: "add" + input: "truediv_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_87" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/query/bias/read" + input: "mul_471" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_457" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias" + input: "sub_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_458" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + input: "add_293" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_459" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + input: "add_294" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_472/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_472" + op: "Mul" + input: "Mul_472/x" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_473/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_473" + op: "Mul" + input: "Mul_473/x" + input: "clip_by_global_norm/clip_by_global_norm/_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_296" + op: "Add" + input: "Mul_472" + input: "Mul_473" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_474/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_474" + op: "Mul" + input: "Mul_474/x" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_87" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_475/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_475" + op: "Mul" + input: "Mul_475/x" + input: "Square_87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_297" + op: "Add" + input: "Mul_474" + input: "Mul_475" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_87" + op: "Sqrt" + input: "add_297" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_298/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_298" + op: "Add" + input: "Sqrt_87" + input: "add_298/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_88" + op: "RealDiv" + input: "add_296" + input: "add_298" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_476/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_476" + op: "Mul" + input: "mul_476/x" + input: "bert/encoder/layer_5/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_299" + op: "Add" + input: "truediv_88" + input: "mul_476" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_477" + op: "Mul" + input: "add" + input: "add_299" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_88" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/key/kernel/read" + input: "mul_477" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_460" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel" + input: "sub_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_461" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + input: "add_296" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_462" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + input: "add_297" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_478/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_478" + op: "Mul" + input: "Mul_478/x" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_479/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_479" + op: "Mul" + input: "Mul_479/x" + input: "clip_by_global_norm/clip_by_global_norm/_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_300" + op: "Add" + input: "Mul_478" + input: "Mul_479" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_480/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_480" + op: "Mul" + input: "Mul_480/x" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_88" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_481/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_481" + op: "Mul" + input: "Mul_481/x" + input: "Square_88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_301" + op: "Add" + input: "Mul_480" + input: "Mul_481" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_88" + op: "Sqrt" + input: "add_301" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_302/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_302" + op: "Add" + input: "Sqrt_88" + input: "add_302/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_89" + op: "RealDiv" + input: "add_300" + input: "add_302" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_482" + op: "Mul" + input: "add" + input: "truediv_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_89" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/key/bias/read" + input: "mul_482" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_463" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias" + input: "sub_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_464" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + input: "add_300" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_465" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + input: "add_301" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_483/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_483" + op: "Mul" + input: "Mul_483/x" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_484/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_484" + op: "Mul" + input: "Mul_484/x" + input: "clip_by_global_norm/clip_by_global_norm/_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_303" + op: "Add" + input: "Mul_483" + input: "Mul_484" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_485/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_485" + op: "Mul" + input: "Mul_485/x" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_89" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_486/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_486" + op: "Mul" + input: "Mul_486/x" + input: "Square_89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_304" + op: "Add" + input: "Mul_485" + input: "Mul_486" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_89" + op: "Sqrt" + input: "add_304" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_305/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_305" + op: "Add" + input: "Sqrt_89" + input: "add_305/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_90" + op: "RealDiv" + input: "add_303" + input: "add_305" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_487/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_487" + op: "Mul" + input: "mul_487/x" + input: "bert/encoder/layer_5/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_306" + op: "Add" + input: "truediv_90" + input: "mul_487" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_488" + op: "Mul" + input: "add" + input: "add_306" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_90" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/value/kernel/read" + input: "mul_488" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_466" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel" + input: "sub_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_467" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + input: "add_303" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_468" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + input: "add_304" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_489/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_489" + op: "Mul" + input: "Mul_489/x" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_490/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_490" + op: "Mul" + input: "Mul_490/x" + input: "clip_by_global_norm/clip_by_global_norm/_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_307" + op: "Add" + input: "Mul_489" + input: "Mul_490" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_491/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_491" + op: "Mul" + input: "Mul_491/x" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_90" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_492/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_492" + op: "Mul" + input: "Mul_492/x" + input: "Square_90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_308" + op: "Add" + input: "Mul_491" + input: "Mul_492" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_90" + op: "Sqrt" + input: "add_308" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_309/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_309" + op: "Add" + input: "Sqrt_90" + input: "add_309/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_91" + op: "RealDiv" + input: "add_307" + input: "add_309" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_493" + op: "Mul" + input: "add" + input: "truediv_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_91" + op: "Sub" + input: "bert/encoder/layer_5/attention/self/value/bias/read" + input: "mul_493" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_469" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias" + input: "sub_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_470" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + input: "add_307" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_471" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + input: "add_308" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_494/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_494" + op: "Mul" + input: "Mul_494/x" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_495/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_495" + op: "Mul" + input: "Mul_495/x" + input: "clip_by_global_norm/clip_by_global_norm/_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_310" + op: "Add" + input: "Mul_494" + input: "Mul_495" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_496/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_496" + op: "Mul" + input: "Mul_496/x" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_91" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_497/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_497" + op: "Mul" + input: "Mul_497/x" + input: "Square_91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_311" + op: "Add" + input: "Mul_496" + input: "Mul_497" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_91" + op: "Sqrt" + input: "add_311" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_312/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_312" + op: "Add" + input: "Sqrt_91" + input: "add_312/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_92" + op: "RealDiv" + input: "add_310" + input: "add_312" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_498/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_498" + op: "Mul" + input: "mul_498/x" + input: "bert/encoder/layer_5/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_313" + op: "Add" + input: "truediv_92" + input: "mul_498" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_499" + op: "Mul" + input: "add" + input: "add_313" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_92" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/dense/kernel/read" + input: "mul_499" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_472" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + input: "sub_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_473" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + input: "add_310" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_474" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + input: "add_311" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_500/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_500" + op: "Mul" + input: "Mul_500/x" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_501/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_501" + op: "Mul" + input: "Mul_501/x" + input: "clip_by_global_norm/clip_by_global_norm/_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_314" + op: "Add" + input: "Mul_500" + input: "Mul_501" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_502/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_502" + op: "Mul" + input: "Mul_502/x" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_92" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_503/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_503" + op: "Mul" + input: "Mul_503/x" + input: "Square_92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_315" + op: "Add" + input: "Mul_502" + input: "Mul_503" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_92" + op: "Sqrt" + input: "add_315" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_316/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_316" + op: "Add" + input: "Sqrt_92" + input: "add_316/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_93" + op: "RealDiv" + input: "add_314" + input: "add_316" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_504" + op: "Mul" + input: "add" + input: "truediv_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_93" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/dense/bias/read" + input: "mul_504" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_475" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias" + input: "sub_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_476" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + input: "add_314" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_477" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + input: "add_315" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_505/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_505" + op: "Mul" + input: "Mul_505/x" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_506/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_506" + op: "Mul" + input: "Mul_506/x" + input: "clip_by_global_norm/clip_by_global_norm/_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_317" + op: "Add" + input: "Mul_505" + input: "Mul_506" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_507/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_507" + op: "Mul" + input: "Mul_507/x" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_93" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_508/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_508" + op: "Mul" + input: "Mul_508/x" + input: "Square_93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_318" + op: "Add" + input: "Mul_507" + input: "Mul_508" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_93" + op: "Sqrt" + input: "add_318" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_319/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_319" + op: "Add" + input: "Sqrt_93" + input: "add_319/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_94" + op: "RealDiv" + input: "add_317" + input: "add_319" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_509" + op: "Mul" + input: "add" + input: "truediv_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_94" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/read" + input: "mul_509" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_478" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + input: "sub_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_479" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + input: "add_317" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_480" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + input: "add_318" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_510/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_510" + op: "Mul" + input: "Mul_510/x" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_511/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_511" + op: "Mul" + input: "Mul_511/x" + input: "clip_by_global_norm/clip_by_global_norm/_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_320" + op: "Add" + input: "Mul_510" + input: "Mul_511" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_512/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_512" + op: "Mul" + input: "Mul_512/x" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_94" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_513/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_513" + op: "Mul" + input: "Mul_513/x" + input: "Square_94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_321" + op: "Add" + input: "Mul_512" + input: "Mul_513" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_94" + op: "Sqrt" + input: "add_321" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_322/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_322" + op: "Add" + input: "Sqrt_94" + input: "add_322/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_95" + op: "RealDiv" + input: "add_320" + input: "add_322" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_514" + op: "Mul" + input: "add" + input: "truediv_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_95" + op: "Sub" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/read" + input: "mul_514" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_481" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + input: "sub_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_482" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + input: "add_320" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_483" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + input: "add_321" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_515/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_515" + op: "Mul" + input: "Mul_515/x" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_516/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_516" + op: "Mul" + input: "Mul_516/x" + input: "clip_by_global_norm/clip_by_global_norm/_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_323" + op: "Add" + input: "Mul_515" + input: "Mul_516" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_517/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_517" + op: "Mul" + input: "Mul_517/x" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_95" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_518/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_518" + op: "Mul" + input: "Mul_518/x" + input: "Square_95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_324" + op: "Add" + input: "Mul_517" + input: "Mul_518" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_95" + op: "Sqrt" + input: "add_324" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_325/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_325" + op: "Add" + input: "Sqrt_95" + input: "add_325/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_96" + op: "RealDiv" + input: "add_323" + input: "add_325" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_519/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_519" + op: "Mul" + input: "mul_519/x" + input: "bert/encoder/layer_5/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_326" + op: "Add" + input: "truediv_96" + input: "mul_519" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_520" + op: "Mul" + input: "add" + input: "add_326" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_96" + op: "Sub" + input: "bert/encoder/layer_5/intermediate/dense/kernel/read" + input: "mul_520" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_484" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + input: "sub_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_485" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + input: "add_323" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_486" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + input: "add_324" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_521/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_521" + op: "Mul" + input: "Mul_521/x" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_522/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_522" + op: "Mul" + input: "Mul_522/x" + input: "clip_by_global_norm/clip_by_global_norm/_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_327" + op: "Add" + input: "Mul_521" + input: "Mul_522" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_523/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_523" + op: "Mul" + input: "Mul_523/x" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_96" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_524/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_524" + op: "Mul" + input: "Mul_524/x" + input: "Square_96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_328" + op: "Add" + input: "Mul_523" + input: "Mul_524" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_96" + op: "Sqrt" + input: "add_328" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_329/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_329" + op: "Add" + input: "Sqrt_96" + input: "add_329/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_97" + op: "RealDiv" + input: "add_327" + input: "add_329" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_525" + op: "Mul" + input: "add" + input: "truediv_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_97" + op: "Sub" + input: "bert/encoder/layer_5/intermediate/dense/bias/read" + input: "mul_525" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_487" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias" + input: "sub_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_488" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + input: "add_327" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_489" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + input: "add_328" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_526/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_526" + op: "Mul" + input: "Mul_526/x" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_527/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_527" + op: "Mul" + input: "Mul_527/x" + input: "clip_by_global_norm/clip_by_global_norm/_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_330" + op: "Add" + input: "Mul_526" + input: "Mul_527" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_528/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_528" + op: "Mul" + input: "Mul_528/x" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_97" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_529/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_529" + op: "Mul" + input: "Mul_529/x" + input: "Square_97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_331" + op: "Add" + input: "Mul_528" + input: "Mul_529" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_97" + op: "Sqrt" + input: "add_331" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_332/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_332" + op: "Add" + input: "Sqrt_97" + input: "add_332/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_98" + op: "RealDiv" + input: "add_330" + input: "add_332" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_530/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_530" + op: "Mul" + input: "mul_530/x" + input: "bert/encoder/layer_5/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_333" + op: "Add" + input: "truediv_98" + input: "mul_530" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_531" + op: "Mul" + input: "add" + input: "add_333" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_98" + op: "Sub" + input: "bert/encoder/layer_5/output/dense/kernel/read" + input: "mul_531" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_490" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel" + input: "sub_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_491" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + input: "add_330" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_492" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + input: "add_331" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + input: "bert/encoder/layer_5/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + input: "bert/encoder/layer_5/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_532/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_532" + op: "Mul" + input: "Mul_532/x" + input: "bert/encoder/layer_5/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_533/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_533" + op: "Mul" + input: "Mul_533/x" + input: "clip_by_global_norm/clip_by_global_norm/_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_334" + op: "Add" + input: "Mul_532" + input: "Mul_533" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_534/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_534" + op: "Mul" + input: "Mul_534/x" + input: "bert/encoder/layer_5/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_98" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_535/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_535" + op: "Mul" + input: "Mul_535/x" + input: "Square_98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_335" + op: "Add" + input: "Mul_534" + input: "Mul_535" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_98" + op: "Sqrt" + input: "add_335" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_336/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_336" + op: "Add" + input: "Sqrt_98" + input: "add_336/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_99" + op: "RealDiv" + input: "add_334" + input: "add_336" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_536" + op: "Mul" + input: "add" + input: "truediv_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_99" + op: "Sub" + input: "bert/encoder/layer_5/output/dense/bias/read" + input: "mul_536" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_493" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias" + input: "sub_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_494" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + input: "add_334" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_495" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + input: "add_335" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_537/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_537" + op: "Mul" + input: "Mul_537/x" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_538/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_538" + op: "Mul" + input: "Mul_538/x" + input: "clip_by_global_norm/clip_by_global_norm/_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_337" + op: "Add" + input: "Mul_537" + input: "Mul_538" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_539/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_539" + op: "Mul" + input: "Mul_539/x" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_99" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_540/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_540" + op: "Mul" + input: "Mul_540/x" + input: "Square_99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_338" + op: "Add" + input: "Mul_539" + input: "Mul_540" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_99" + op: "Sqrt" + input: "add_338" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_339/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_339" + op: "Add" + input: "Sqrt_99" + input: "add_339/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_100" + op: "RealDiv" + input: "add_337" + input: "add_339" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_541" + op: "Mul" + input: "add" + input: "truediv_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_100" + op: "Sub" + input: "bert/encoder/layer_5/output/LayerNorm/beta/read" + input: "mul_541" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_496" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + input: "sub_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_497" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + input: "add_337" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_498" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + input: "add_338" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_542/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_542" + op: "Mul" + input: "Mul_542/x" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_543/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_543" + op: "Mul" + input: "Mul_543/x" + input: "clip_by_global_norm/clip_by_global_norm/_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_340" + op: "Add" + input: "Mul_542" + input: "Mul_543" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_544/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_544" + op: "Mul" + input: "Mul_544/x" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_100" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_545/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_545" + op: "Mul" + input: "Mul_545/x" + input: "Square_100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_341" + op: "Add" + input: "Mul_544" + input: "Mul_545" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_100" + op: "Sqrt" + input: "add_341" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_342/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_342" + op: "Add" + input: "Sqrt_100" + input: "add_342/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_101" + op: "RealDiv" + input: "add_340" + input: "add_342" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_546" + op: "Mul" + input: "add" + input: "truediv_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_101" + op: "Sub" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/read" + input: "mul_546" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_499" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + input: "sub_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_500" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + input: "add_340" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_501" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + input: "add_341" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_547/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_547" + op: "Mul" + input: "Mul_547/x" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_548/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_548" + op: "Mul" + input: "Mul_548/x" + input: "clip_by_global_norm/clip_by_global_norm/_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_343" + op: "Add" + input: "Mul_547" + input: "Mul_548" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_549/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_549" + op: "Mul" + input: "Mul_549/x" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_101" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_550/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_550" + op: "Mul" + input: "Mul_550/x" + input: "Square_101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_344" + op: "Add" + input: "Mul_549" + input: "Mul_550" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_101" + op: "Sqrt" + input: "add_344" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_345/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_345" + op: "Add" + input: "Sqrt_101" + input: "add_345/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_102" + op: "RealDiv" + input: "add_343" + input: "add_345" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_551/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_551" + op: "Mul" + input: "mul_551/x" + input: "bert/encoder/layer_6/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_346" + op: "Add" + input: "truediv_102" + input: "mul_551" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_552" + op: "Mul" + input: "add" + input: "add_346" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_102" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/query/kernel/read" + input: "mul_552" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_502" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel" + input: "sub_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_503" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + input: "add_343" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_504" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + input: "add_344" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_553/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_553" + op: "Mul" + input: "Mul_553/x" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_554/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_554" + op: "Mul" + input: "Mul_554/x" + input: "clip_by_global_norm/clip_by_global_norm/_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_347" + op: "Add" + input: "Mul_553" + input: "Mul_554" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_555/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_555" + op: "Mul" + input: "Mul_555/x" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_102" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_556/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_556" + op: "Mul" + input: "Mul_556/x" + input: "Square_102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_348" + op: "Add" + input: "Mul_555" + input: "Mul_556" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_102" + op: "Sqrt" + input: "add_348" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_349/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_349" + op: "Add" + input: "Sqrt_102" + input: "add_349/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_103" + op: "RealDiv" + input: "add_347" + input: "add_349" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_557" + op: "Mul" + input: "add" + input: "truediv_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_103" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/query/bias/read" + input: "mul_557" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_505" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias" + input: "sub_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_506" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + input: "add_347" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_507" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + input: "add_348" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_558/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_558" + op: "Mul" + input: "Mul_558/x" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_559/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_559" + op: "Mul" + input: "Mul_559/x" + input: "clip_by_global_norm/clip_by_global_norm/_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_350" + op: "Add" + input: "Mul_558" + input: "Mul_559" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_560/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_560" + op: "Mul" + input: "Mul_560/x" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_103" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_561/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_561" + op: "Mul" + input: "Mul_561/x" + input: "Square_103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_351" + op: "Add" + input: "Mul_560" + input: "Mul_561" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_103" + op: "Sqrt" + input: "add_351" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_352/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_352" + op: "Add" + input: "Sqrt_103" + input: "add_352/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_104" + op: "RealDiv" + input: "add_350" + input: "add_352" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_562/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_562" + op: "Mul" + input: "mul_562/x" + input: "bert/encoder/layer_6/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_353" + op: "Add" + input: "truediv_104" + input: "mul_562" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_563" + op: "Mul" + input: "add" + input: "add_353" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_104" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/key/kernel/read" + input: "mul_563" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_508" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel" + input: "sub_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_509" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + input: "add_350" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_510" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + input: "add_351" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_564/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_564" + op: "Mul" + input: "Mul_564/x" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_565/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_565" + op: "Mul" + input: "Mul_565/x" + input: "clip_by_global_norm/clip_by_global_norm/_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_354" + op: "Add" + input: "Mul_564" + input: "Mul_565" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_566/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_566" + op: "Mul" + input: "Mul_566/x" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_104" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_567/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_567" + op: "Mul" + input: "Mul_567/x" + input: "Square_104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_355" + op: "Add" + input: "Mul_566" + input: "Mul_567" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_104" + op: "Sqrt" + input: "add_355" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_356/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_356" + op: "Add" + input: "Sqrt_104" + input: "add_356/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_105" + op: "RealDiv" + input: "add_354" + input: "add_356" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_568" + op: "Mul" + input: "add" + input: "truediv_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_105" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/key/bias/read" + input: "mul_568" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_511" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias" + input: "sub_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_512" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + input: "add_354" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_513" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + input: "add_355" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_569/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_569" + op: "Mul" + input: "Mul_569/x" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_570/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_570" + op: "Mul" + input: "Mul_570/x" + input: "clip_by_global_norm/clip_by_global_norm/_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_357" + op: "Add" + input: "Mul_569" + input: "Mul_570" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_571/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_571" + op: "Mul" + input: "Mul_571/x" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_105" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_572/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_572" + op: "Mul" + input: "Mul_572/x" + input: "Square_105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_358" + op: "Add" + input: "Mul_571" + input: "Mul_572" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_105" + op: "Sqrt" + input: "add_358" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_359/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_359" + op: "Add" + input: "Sqrt_105" + input: "add_359/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_106" + op: "RealDiv" + input: "add_357" + input: "add_359" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_573/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_573" + op: "Mul" + input: "mul_573/x" + input: "bert/encoder/layer_6/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_360" + op: "Add" + input: "truediv_106" + input: "mul_573" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_574" + op: "Mul" + input: "add" + input: "add_360" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_106" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/value/kernel/read" + input: "mul_574" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_514" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel" + input: "sub_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_515" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + input: "add_357" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_516" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + input: "add_358" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_575/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_575" + op: "Mul" + input: "Mul_575/x" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_576/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_576" + op: "Mul" + input: "Mul_576/x" + input: "clip_by_global_norm/clip_by_global_norm/_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_361" + op: "Add" + input: "Mul_575" + input: "Mul_576" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_577/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_577" + op: "Mul" + input: "Mul_577/x" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_106" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_578/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_578" + op: "Mul" + input: "Mul_578/x" + input: "Square_106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_362" + op: "Add" + input: "Mul_577" + input: "Mul_578" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_106" + op: "Sqrt" + input: "add_362" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_363/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_363" + op: "Add" + input: "Sqrt_106" + input: "add_363/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_107" + op: "RealDiv" + input: "add_361" + input: "add_363" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_579" + op: "Mul" + input: "add" + input: "truediv_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_107" + op: "Sub" + input: "bert/encoder/layer_6/attention/self/value/bias/read" + input: "mul_579" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_517" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias" + input: "sub_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_518" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + input: "add_361" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_519" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + input: "add_362" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_580/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_580" + op: "Mul" + input: "Mul_580/x" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_581/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_581" + op: "Mul" + input: "Mul_581/x" + input: "clip_by_global_norm/clip_by_global_norm/_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_364" + op: "Add" + input: "Mul_580" + input: "Mul_581" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_582/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_582" + op: "Mul" + input: "Mul_582/x" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_107" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_583/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_583" + op: "Mul" + input: "Mul_583/x" + input: "Square_107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_365" + op: "Add" + input: "Mul_582" + input: "Mul_583" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_107" + op: "Sqrt" + input: "add_365" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_366/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_366" + op: "Add" + input: "Sqrt_107" + input: "add_366/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_108" + op: "RealDiv" + input: "add_364" + input: "add_366" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_584/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_584" + op: "Mul" + input: "mul_584/x" + input: "bert/encoder/layer_6/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_367" + op: "Add" + input: "truediv_108" + input: "mul_584" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_585" + op: "Mul" + input: "add" + input: "add_367" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_108" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/dense/kernel/read" + input: "mul_585" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_520" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + input: "sub_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_521" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + input: "add_364" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_522" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + input: "add_365" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_586/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_586" + op: "Mul" + input: "Mul_586/x" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_587/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_587" + op: "Mul" + input: "Mul_587/x" + input: "clip_by_global_norm/clip_by_global_norm/_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_368" + op: "Add" + input: "Mul_586" + input: "Mul_587" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_588/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_588" + op: "Mul" + input: "Mul_588/x" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_108" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_589/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_589" + op: "Mul" + input: "Mul_589/x" + input: "Square_108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_369" + op: "Add" + input: "Mul_588" + input: "Mul_589" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_108" + op: "Sqrt" + input: "add_369" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_370/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_370" + op: "Add" + input: "Sqrt_108" + input: "add_370/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_109" + op: "RealDiv" + input: "add_368" + input: "add_370" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_590" + op: "Mul" + input: "add" + input: "truediv_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_109" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/dense/bias/read" + input: "mul_590" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_523" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias" + input: "sub_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_524" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + input: "add_368" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_525" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + input: "add_369" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_591/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_591" + op: "Mul" + input: "Mul_591/x" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_592/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_592" + op: "Mul" + input: "Mul_592/x" + input: "clip_by_global_norm/clip_by_global_norm/_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_371" + op: "Add" + input: "Mul_591" + input: "Mul_592" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_593/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_593" + op: "Mul" + input: "Mul_593/x" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_109" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_594/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_594" + op: "Mul" + input: "Mul_594/x" + input: "Square_109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_372" + op: "Add" + input: "Mul_593" + input: "Mul_594" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_109" + op: "Sqrt" + input: "add_372" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_373/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_373" + op: "Add" + input: "Sqrt_109" + input: "add_373/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_110" + op: "RealDiv" + input: "add_371" + input: "add_373" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_595" + op: "Mul" + input: "add" + input: "truediv_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_110" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/read" + input: "mul_595" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_526" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + input: "sub_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_527" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + input: "add_371" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_528" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + input: "add_372" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_596/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_596" + op: "Mul" + input: "Mul_596/x" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_597/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_597" + op: "Mul" + input: "Mul_597/x" + input: "clip_by_global_norm/clip_by_global_norm/_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_374" + op: "Add" + input: "Mul_596" + input: "Mul_597" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_598/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_598" + op: "Mul" + input: "Mul_598/x" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_110" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_599/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_599" + op: "Mul" + input: "Mul_599/x" + input: "Square_110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_375" + op: "Add" + input: "Mul_598" + input: "Mul_599" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_110" + op: "Sqrt" + input: "add_375" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_376/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_376" + op: "Add" + input: "Sqrt_110" + input: "add_376/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_111" + op: "RealDiv" + input: "add_374" + input: "add_376" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_600" + op: "Mul" + input: "add" + input: "truediv_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_111" + op: "Sub" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/read" + input: "mul_600" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_529" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + input: "sub_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_530" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + input: "add_374" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_531" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + input: "add_375" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_601/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_601" + op: "Mul" + input: "Mul_601/x" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_602/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_602" + op: "Mul" + input: "Mul_602/x" + input: "clip_by_global_norm/clip_by_global_norm/_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_377" + op: "Add" + input: "Mul_601" + input: "Mul_602" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_603/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_603" + op: "Mul" + input: "Mul_603/x" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_111" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_604/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_604" + op: "Mul" + input: "Mul_604/x" + input: "Square_111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_378" + op: "Add" + input: "Mul_603" + input: "Mul_604" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_111" + op: "Sqrt" + input: "add_378" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_379/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_379" + op: "Add" + input: "Sqrt_111" + input: "add_379/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_112" + op: "RealDiv" + input: "add_377" + input: "add_379" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_605/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_605" + op: "Mul" + input: "mul_605/x" + input: "bert/encoder/layer_6/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_380" + op: "Add" + input: "truediv_112" + input: "mul_605" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_606" + op: "Mul" + input: "add" + input: "add_380" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_112" + op: "Sub" + input: "bert/encoder/layer_6/intermediate/dense/kernel/read" + input: "mul_606" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_532" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + input: "sub_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_533" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + input: "add_377" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_534" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + input: "add_378" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_607/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_607" + op: "Mul" + input: "Mul_607/x" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_608/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_608" + op: "Mul" + input: "Mul_608/x" + input: "clip_by_global_norm/clip_by_global_norm/_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_381" + op: "Add" + input: "Mul_607" + input: "Mul_608" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_609/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_609" + op: "Mul" + input: "Mul_609/x" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_112" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_610/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_610" + op: "Mul" + input: "Mul_610/x" + input: "Square_112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_382" + op: "Add" + input: "Mul_609" + input: "Mul_610" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_112" + op: "Sqrt" + input: "add_382" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_383/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_383" + op: "Add" + input: "Sqrt_112" + input: "add_383/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_113" + op: "RealDiv" + input: "add_381" + input: "add_383" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_611" + op: "Mul" + input: "add" + input: "truediv_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_113" + op: "Sub" + input: "bert/encoder/layer_6/intermediate/dense/bias/read" + input: "mul_611" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_535" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias" + input: "sub_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_536" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + input: "add_381" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_537" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + input: "add_382" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_612/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_612" + op: "Mul" + input: "Mul_612/x" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_613/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_613" + op: "Mul" + input: "Mul_613/x" + input: "clip_by_global_norm/clip_by_global_norm/_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_384" + op: "Add" + input: "Mul_612" + input: "Mul_613" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_614/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_614" + op: "Mul" + input: "Mul_614/x" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_113" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_615/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_615" + op: "Mul" + input: "Mul_615/x" + input: "Square_113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_385" + op: "Add" + input: "Mul_614" + input: "Mul_615" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_113" + op: "Sqrt" + input: "add_385" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_386/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_386" + op: "Add" + input: "Sqrt_113" + input: "add_386/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_114" + op: "RealDiv" + input: "add_384" + input: "add_386" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_616/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_616" + op: "Mul" + input: "mul_616/x" + input: "bert/encoder/layer_6/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_387" + op: "Add" + input: "truediv_114" + input: "mul_616" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_617" + op: "Mul" + input: "add" + input: "add_387" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_114" + op: "Sub" + input: "bert/encoder/layer_6/output/dense/kernel/read" + input: "mul_617" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_538" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel" + input: "sub_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_539" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + input: "add_384" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_540" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + input: "add_385" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + input: "bert/encoder/layer_6/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + input: "bert/encoder/layer_6/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_618/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_618" + op: "Mul" + input: "Mul_618/x" + input: "bert/encoder/layer_6/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_619/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_619" + op: "Mul" + input: "Mul_619/x" + input: "clip_by_global_norm/clip_by_global_norm/_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_388" + op: "Add" + input: "Mul_618" + input: "Mul_619" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_620/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_620" + op: "Mul" + input: "Mul_620/x" + input: "bert/encoder/layer_6/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_114" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_621/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_621" + op: "Mul" + input: "Mul_621/x" + input: "Square_114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_389" + op: "Add" + input: "Mul_620" + input: "Mul_621" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_114" + op: "Sqrt" + input: "add_389" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_390/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_390" + op: "Add" + input: "Sqrt_114" + input: "add_390/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_115" + op: "RealDiv" + input: "add_388" + input: "add_390" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_622" + op: "Mul" + input: "add" + input: "truediv_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_115" + op: "Sub" + input: "bert/encoder/layer_6/output/dense/bias/read" + input: "mul_622" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_541" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias" + input: "sub_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_542" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + input: "add_388" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_543" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + input: "add_389" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_623/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_623" + op: "Mul" + input: "Mul_623/x" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_624/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_624" + op: "Mul" + input: "Mul_624/x" + input: "clip_by_global_norm/clip_by_global_norm/_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_391" + op: "Add" + input: "Mul_623" + input: "Mul_624" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_625/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_625" + op: "Mul" + input: "Mul_625/x" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_115" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_626/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_626" + op: "Mul" + input: "Mul_626/x" + input: "Square_115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_392" + op: "Add" + input: "Mul_625" + input: "Mul_626" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_115" + op: "Sqrt" + input: "add_392" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_393/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_393" + op: "Add" + input: "Sqrt_115" + input: "add_393/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_116" + op: "RealDiv" + input: "add_391" + input: "add_393" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_627" + op: "Mul" + input: "add" + input: "truediv_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_116" + op: "Sub" + input: "bert/encoder/layer_6/output/LayerNorm/beta/read" + input: "mul_627" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_544" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + input: "sub_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_545" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + input: "add_391" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_546" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + input: "add_392" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_628/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_628" + op: "Mul" + input: "Mul_628/x" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_629/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_629" + op: "Mul" + input: "Mul_629/x" + input: "clip_by_global_norm/clip_by_global_norm/_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_394" + op: "Add" + input: "Mul_628" + input: "Mul_629" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_630/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_630" + op: "Mul" + input: "Mul_630/x" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_116" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_631/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_631" + op: "Mul" + input: "Mul_631/x" + input: "Square_116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_395" + op: "Add" + input: "Mul_630" + input: "Mul_631" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_116" + op: "Sqrt" + input: "add_395" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_396/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_396" + op: "Add" + input: "Sqrt_116" + input: "add_396/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_117" + op: "RealDiv" + input: "add_394" + input: "add_396" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_632" + op: "Mul" + input: "add" + input: "truediv_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_117" + op: "Sub" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/read" + input: "mul_632" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_547" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + input: "sub_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_548" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + input: "add_394" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_549" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + input: "add_395" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_633/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_633" + op: "Mul" + input: "Mul_633/x" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_634/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_634" + op: "Mul" + input: "Mul_634/x" + input: "clip_by_global_norm/clip_by_global_norm/_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_397" + op: "Add" + input: "Mul_633" + input: "Mul_634" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_635/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_635" + op: "Mul" + input: "Mul_635/x" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_117" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_636/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_636" + op: "Mul" + input: "Mul_636/x" + input: "Square_117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_398" + op: "Add" + input: "Mul_635" + input: "Mul_636" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_117" + op: "Sqrt" + input: "add_398" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_399/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_399" + op: "Add" + input: "Sqrt_117" + input: "add_399/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_118" + op: "RealDiv" + input: "add_397" + input: "add_399" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_637/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_637" + op: "Mul" + input: "mul_637/x" + input: "bert/encoder/layer_7/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_400" + op: "Add" + input: "truediv_118" + input: "mul_637" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_638" + op: "Mul" + input: "add" + input: "add_400" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_118" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/query/kernel/read" + input: "mul_638" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_550" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel" + input: "sub_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_551" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + input: "add_397" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_552" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + input: "add_398" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_639/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_639" + op: "Mul" + input: "Mul_639/x" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_640/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_640" + op: "Mul" + input: "Mul_640/x" + input: "clip_by_global_norm/clip_by_global_norm/_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_401" + op: "Add" + input: "Mul_639" + input: "Mul_640" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_641/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_641" + op: "Mul" + input: "Mul_641/x" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_118" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_642/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_642" + op: "Mul" + input: "Mul_642/x" + input: "Square_118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_402" + op: "Add" + input: "Mul_641" + input: "Mul_642" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_118" + op: "Sqrt" + input: "add_402" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_403/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_403" + op: "Add" + input: "Sqrt_118" + input: "add_403/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_119" + op: "RealDiv" + input: "add_401" + input: "add_403" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_643" + op: "Mul" + input: "add" + input: "truediv_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_119" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/query/bias/read" + input: "mul_643" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_553" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias" + input: "sub_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_554" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + input: "add_401" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_555" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + input: "add_402" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_644/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_644" + op: "Mul" + input: "Mul_644/x" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_645/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_645" + op: "Mul" + input: "Mul_645/x" + input: "clip_by_global_norm/clip_by_global_norm/_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_404" + op: "Add" + input: "Mul_644" + input: "Mul_645" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_646/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_646" + op: "Mul" + input: "Mul_646/x" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_119" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_647/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_647" + op: "Mul" + input: "Mul_647/x" + input: "Square_119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_405" + op: "Add" + input: "Mul_646" + input: "Mul_647" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_119" + op: "Sqrt" + input: "add_405" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_406/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_406" + op: "Add" + input: "Sqrt_119" + input: "add_406/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_120" + op: "RealDiv" + input: "add_404" + input: "add_406" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_648/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_648" + op: "Mul" + input: "mul_648/x" + input: "bert/encoder/layer_7/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_407" + op: "Add" + input: "truediv_120" + input: "mul_648" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_649" + op: "Mul" + input: "add" + input: "add_407" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_120" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/key/kernel/read" + input: "mul_649" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_556" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel" + input: "sub_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_557" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + input: "add_404" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_558" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + input: "add_405" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_650/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_650" + op: "Mul" + input: "Mul_650/x" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_651/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_651" + op: "Mul" + input: "Mul_651/x" + input: "clip_by_global_norm/clip_by_global_norm/_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_408" + op: "Add" + input: "Mul_650" + input: "Mul_651" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_652/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_652" + op: "Mul" + input: "Mul_652/x" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_120" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_653/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_653" + op: "Mul" + input: "Mul_653/x" + input: "Square_120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_409" + op: "Add" + input: "Mul_652" + input: "Mul_653" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_120" + op: "Sqrt" + input: "add_409" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_410/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_410" + op: "Add" + input: "Sqrt_120" + input: "add_410/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_121" + op: "RealDiv" + input: "add_408" + input: "add_410" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_654" + op: "Mul" + input: "add" + input: "truediv_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_121" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/key/bias/read" + input: "mul_654" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_559" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias" + input: "sub_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_560" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + input: "add_408" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_561" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + input: "add_409" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_655/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_655" + op: "Mul" + input: "Mul_655/x" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_656/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_656" + op: "Mul" + input: "Mul_656/x" + input: "clip_by_global_norm/clip_by_global_norm/_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_411" + op: "Add" + input: "Mul_655" + input: "Mul_656" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_657/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_657" + op: "Mul" + input: "Mul_657/x" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_121" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_658/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_658" + op: "Mul" + input: "Mul_658/x" + input: "Square_121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_412" + op: "Add" + input: "Mul_657" + input: "Mul_658" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_121" + op: "Sqrt" + input: "add_412" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_413/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_413" + op: "Add" + input: "Sqrt_121" + input: "add_413/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_122" + op: "RealDiv" + input: "add_411" + input: "add_413" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_659/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_659" + op: "Mul" + input: "mul_659/x" + input: "bert/encoder/layer_7/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_414" + op: "Add" + input: "truediv_122" + input: "mul_659" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_660" + op: "Mul" + input: "add" + input: "add_414" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_122" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/value/kernel/read" + input: "mul_660" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_562" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel" + input: "sub_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_563" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + input: "add_411" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_564" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + input: "add_412" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_661/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_661" + op: "Mul" + input: "Mul_661/x" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_662/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_662" + op: "Mul" + input: "Mul_662/x" + input: "clip_by_global_norm/clip_by_global_norm/_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_415" + op: "Add" + input: "Mul_661" + input: "Mul_662" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_663/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_663" + op: "Mul" + input: "Mul_663/x" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_122" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_664/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_664" + op: "Mul" + input: "Mul_664/x" + input: "Square_122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_416" + op: "Add" + input: "Mul_663" + input: "Mul_664" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_122" + op: "Sqrt" + input: "add_416" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_417/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_417" + op: "Add" + input: "Sqrt_122" + input: "add_417/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_123" + op: "RealDiv" + input: "add_415" + input: "add_417" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_665" + op: "Mul" + input: "add" + input: "truediv_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_123" + op: "Sub" + input: "bert/encoder/layer_7/attention/self/value/bias/read" + input: "mul_665" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_565" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias" + input: "sub_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_566" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + input: "add_415" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_567" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + input: "add_416" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_666/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_666" + op: "Mul" + input: "Mul_666/x" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_667/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_667" + op: "Mul" + input: "Mul_667/x" + input: "clip_by_global_norm/clip_by_global_norm/_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_418" + op: "Add" + input: "Mul_666" + input: "Mul_667" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_668/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_668" + op: "Mul" + input: "Mul_668/x" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_123" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_669/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_669" + op: "Mul" + input: "Mul_669/x" + input: "Square_123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_419" + op: "Add" + input: "Mul_668" + input: "Mul_669" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_123" + op: "Sqrt" + input: "add_419" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_420/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_420" + op: "Add" + input: "Sqrt_123" + input: "add_420/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_124" + op: "RealDiv" + input: "add_418" + input: "add_420" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_670/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_670" + op: "Mul" + input: "mul_670/x" + input: "bert/encoder/layer_7/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_421" + op: "Add" + input: "truediv_124" + input: "mul_670" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_671" + op: "Mul" + input: "add" + input: "add_421" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_124" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/dense/kernel/read" + input: "mul_671" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_568" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + input: "sub_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_569" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + input: "add_418" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_570" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + input: "add_419" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_672/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_672" + op: "Mul" + input: "Mul_672/x" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_673/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_673" + op: "Mul" + input: "Mul_673/x" + input: "clip_by_global_norm/clip_by_global_norm/_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_422" + op: "Add" + input: "Mul_672" + input: "Mul_673" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_674/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_674" + op: "Mul" + input: "Mul_674/x" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_124" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_675/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_675" + op: "Mul" + input: "Mul_675/x" + input: "Square_124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_423" + op: "Add" + input: "Mul_674" + input: "Mul_675" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_124" + op: "Sqrt" + input: "add_423" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_424/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_424" + op: "Add" + input: "Sqrt_124" + input: "add_424/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_125" + op: "RealDiv" + input: "add_422" + input: "add_424" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_676" + op: "Mul" + input: "add" + input: "truediv_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_125" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/dense/bias/read" + input: "mul_676" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_571" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias" + input: "sub_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_572" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + input: "add_422" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_573" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + input: "add_423" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_677/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_677" + op: "Mul" + input: "Mul_677/x" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_678/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_678" + op: "Mul" + input: "Mul_678/x" + input: "clip_by_global_norm/clip_by_global_norm/_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_425" + op: "Add" + input: "Mul_677" + input: "Mul_678" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_679/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_679" + op: "Mul" + input: "Mul_679/x" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_125" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_680/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_680" + op: "Mul" + input: "Mul_680/x" + input: "Square_125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_426" + op: "Add" + input: "Mul_679" + input: "Mul_680" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_125" + op: "Sqrt" + input: "add_426" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_427/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_427" + op: "Add" + input: "Sqrt_125" + input: "add_427/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_126" + op: "RealDiv" + input: "add_425" + input: "add_427" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_681" + op: "Mul" + input: "add" + input: "truediv_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_126" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/read" + input: "mul_681" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_574" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + input: "sub_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_575" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + input: "add_425" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_576" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + input: "add_426" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_682/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_682" + op: "Mul" + input: "Mul_682/x" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_683/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_683" + op: "Mul" + input: "Mul_683/x" + input: "clip_by_global_norm/clip_by_global_norm/_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_428" + op: "Add" + input: "Mul_682" + input: "Mul_683" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_684/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_684" + op: "Mul" + input: "Mul_684/x" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_126" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_685/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_685" + op: "Mul" + input: "Mul_685/x" + input: "Square_126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_429" + op: "Add" + input: "Mul_684" + input: "Mul_685" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_126" + op: "Sqrt" + input: "add_429" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_430/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_430" + op: "Add" + input: "Sqrt_126" + input: "add_430/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_127" + op: "RealDiv" + input: "add_428" + input: "add_430" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_686" + op: "Mul" + input: "add" + input: "truediv_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_127" + op: "Sub" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/read" + input: "mul_686" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_577" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + input: "sub_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_578" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + input: "add_428" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_579" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + input: "add_429" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_687/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_687" + op: "Mul" + input: "Mul_687/x" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_688/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_688" + op: "Mul" + input: "Mul_688/x" + input: "clip_by_global_norm/clip_by_global_norm/_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_431" + op: "Add" + input: "Mul_687" + input: "Mul_688" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_689/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_689" + op: "Mul" + input: "Mul_689/x" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_127" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_690/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_690" + op: "Mul" + input: "Mul_690/x" + input: "Square_127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_432" + op: "Add" + input: "Mul_689" + input: "Mul_690" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_127" + op: "Sqrt" + input: "add_432" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_433/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_433" + op: "Add" + input: "Sqrt_127" + input: "add_433/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_128" + op: "RealDiv" + input: "add_431" + input: "add_433" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_691/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_691" + op: "Mul" + input: "mul_691/x" + input: "bert/encoder/layer_7/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_434" + op: "Add" + input: "truediv_128" + input: "mul_691" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_692" + op: "Mul" + input: "add" + input: "add_434" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_128" + op: "Sub" + input: "bert/encoder/layer_7/intermediate/dense/kernel/read" + input: "mul_692" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_580" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + input: "sub_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_581" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + input: "add_431" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_582" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + input: "add_432" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_693/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_693" + op: "Mul" + input: "Mul_693/x" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_694/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_694" + op: "Mul" + input: "Mul_694/x" + input: "clip_by_global_norm/clip_by_global_norm/_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_435" + op: "Add" + input: "Mul_693" + input: "Mul_694" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_695/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_695" + op: "Mul" + input: "Mul_695/x" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_128" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_696/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_696" + op: "Mul" + input: "Mul_696/x" + input: "Square_128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_436" + op: "Add" + input: "Mul_695" + input: "Mul_696" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_128" + op: "Sqrt" + input: "add_436" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_437/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_437" + op: "Add" + input: "Sqrt_128" + input: "add_437/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_129" + op: "RealDiv" + input: "add_435" + input: "add_437" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_697" + op: "Mul" + input: "add" + input: "truediv_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_129" + op: "Sub" + input: "bert/encoder/layer_7/intermediate/dense/bias/read" + input: "mul_697" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_583" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias" + input: "sub_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_584" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + input: "add_435" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_585" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + input: "add_436" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_698/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_698" + op: "Mul" + input: "Mul_698/x" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_699/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_699" + op: "Mul" + input: "Mul_699/x" + input: "clip_by_global_norm/clip_by_global_norm/_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_438" + op: "Add" + input: "Mul_698" + input: "Mul_699" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_700/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_700" + op: "Mul" + input: "Mul_700/x" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_129" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_701/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_701" + op: "Mul" + input: "Mul_701/x" + input: "Square_129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_439" + op: "Add" + input: "Mul_700" + input: "Mul_701" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_129" + op: "Sqrt" + input: "add_439" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_440/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_440" + op: "Add" + input: "Sqrt_129" + input: "add_440/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_130" + op: "RealDiv" + input: "add_438" + input: "add_440" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_702/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_702" + op: "Mul" + input: "mul_702/x" + input: "bert/encoder/layer_7/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_441" + op: "Add" + input: "truediv_130" + input: "mul_702" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_703" + op: "Mul" + input: "add" + input: "add_441" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_130" + op: "Sub" + input: "bert/encoder/layer_7/output/dense/kernel/read" + input: "mul_703" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_586" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel" + input: "sub_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_587" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + input: "add_438" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_588" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + input: "add_439" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + input: "bert/encoder/layer_7/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + input: "bert/encoder/layer_7/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_704/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_704" + op: "Mul" + input: "Mul_704/x" + input: "bert/encoder/layer_7/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_705/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_705" + op: "Mul" + input: "Mul_705/x" + input: "clip_by_global_norm/clip_by_global_norm/_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_442" + op: "Add" + input: "Mul_704" + input: "Mul_705" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_706/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_706" + op: "Mul" + input: "Mul_706/x" + input: "bert/encoder/layer_7/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_130" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_707/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_707" + op: "Mul" + input: "Mul_707/x" + input: "Square_130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_443" + op: "Add" + input: "Mul_706" + input: "Mul_707" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_130" + op: "Sqrt" + input: "add_443" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_444/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_444" + op: "Add" + input: "Sqrt_130" + input: "add_444/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_131" + op: "RealDiv" + input: "add_442" + input: "add_444" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_708" + op: "Mul" + input: "add" + input: "truediv_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_131" + op: "Sub" + input: "bert/encoder/layer_7/output/dense/bias/read" + input: "mul_708" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_589" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias" + input: "sub_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_590" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + input: "add_442" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_591" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + input: "add_443" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_709/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_709" + op: "Mul" + input: "Mul_709/x" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_710/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_710" + op: "Mul" + input: "Mul_710/x" + input: "clip_by_global_norm/clip_by_global_norm/_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_445" + op: "Add" + input: "Mul_709" + input: "Mul_710" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_711/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_711" + op: "Mul" + input: "Mul_711/x" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_131" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_712/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_712" + op: "Mul" + input: "Mul_712/x" + input: "Square_131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_446" + op: "Add" + input: "Mul_711" + input: "Mul_712" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_131" + op: "Sqrt" + input: "add_446" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_447/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_447" + op: "Add" + input: "Sqrt_131" + input: "add_447/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_132" + op: "RealDiv" + input: "add_445" + input: "add_447" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_713" + op: "Mul" + input: "add" + input: "truediv_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_132" + op: "Sub" + input: "bert/encoder/layer_7/output/LayerNorm/beta/read" + input: "mul_713" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_592" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + input: "sub_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_593" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + input: "add_445" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_594" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + input: "add_446" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_714/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_714" + op: "Mul" + input: "Mul_714/x" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_715/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_715" + op: "Mul" + input: "Mul_715/x" + input: "clip_by_global_norm/clip_by_global_norm/_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_448" + op: "Add" + input: "Mul_714" + input: "Mul_715" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_716/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_716" + op: "Mul" + input: "Mul_716/x" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_132" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_717/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_717" + op: "Mul" + input: "Mul_717/x" + input: "Square_132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_449" + op: "Add" + input: "Mul_716" + input: "Mul_717" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_132" + op: "Sqrt" + input: "add_449" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_450/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_450" + op: "Add" + input: "Sqrt_132" + input: "add_450/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_133" + op: "RealDiv" + input: "add_448" + input: "add_450" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_718" + op: "Mul" + input: "add" + input: "truediv_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_133" + op: "Sub" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/read" + input: "mul_718" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_595" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + input: "sub_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_596" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + input: "add_448" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_597" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + input: "add_449" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_719/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_719" + op: "Mul" + input: "Mul_719/x" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_720/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_720" + op: "Mul" + input: "Mul_720/x" + input: "clip_by_global_norm/clip_by_global_norm/_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_451" + op: "Add" + input: "Mul_719" + input: "Mul_720" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_721/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_721" + op: "Mul" + input: "Mul_721/x" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_133" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_722/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_722" + op: "Mul" + input: "Mul_722/x" + input: "Square_133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_452" + op: "Add" + input: "Mul_721" + input: "Mul_722" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_133" + op: "Sqrt" + input: "add_452" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_453/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_453" + op: "Add" + input: "Sqrt_133" + input: "add_453/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_134" + op: "RealDiv" + input: "add_451" + input: "add_453" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_723/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_723" + op: "Mul" + input: "mul_723/x" + input: "bert/encoder/layer_8/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_454" + op: "Add" + input: "truediv_134" + input: "mul_723" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_724" + op: "Mul" + input: "add" + input: "add_454" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_134" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/query/kernel/read" + input: "mul_724" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_598" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel" + input: "sub_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_599" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + input: "add_451" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_600" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + input: "add_452" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_725/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_725" + op: "Mul" + input: "Mul_725/x" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_726/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_726" + op: "Mul" + input: "Mul_726/x" + input: "clip_by_global_norm/clip_by_global_norm/_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_455" + op: "Add" + input: "Mul_725" + input: "Mul_726" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_727/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_727" + op: "Mul" + input: "Mul_727/x" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_134" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_728/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_728" + op: "Mul" + input: "Mul_728/x" + input: "Square_134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_456" + op: "Add" + input: "Mul_727" + input: "Mul_728" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_134" + op: "Sqrt" + input: "add_456" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_457/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_457" + op: "Add" + input: "Sqrt_134" + input: "add_457/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_135" + op: "RealDiv" + input: "add_455" + input: "add_457" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_729" + op: "Mul" + input: "add" + input: "truediv_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_135" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/query/bias/read" + input: "mul_729" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_601" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias" + input: "sub_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_602" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + input: "add_455" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_603" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + input: "add_456" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_730/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_730" + op: "Mul" + input: "Mul_730/x" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_731/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_731" + op: "Mul" + input: "Mul_731/x" + input: "clip_by_global_norm/clip_by_global_norm/_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_458" + op: "Add" + input: "Mul_730" + input: "Mul_731" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_732/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_732" + op: "Mul" + input: "Mul_732/x" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_135" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_733/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_733" + op: "Mul" + input: "Mul_733/x" + input: "Square_135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_459" + op: "Add" + input: "Mul_732" + input: "Mul_733" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_135" + op: "Sqrt" + input: "add_459" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_460/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_460" + op: "Add" + input: "Sqrt_135" + input: "add_460/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_136" + op: "RealDiv" + input: "add_458" + input: "add_460" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_734/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_734" + op: "Mul" + input: "mul_734/x" + input: "bert/encoder/layer_8/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_461" + op: "Add" + input: "truediv_136" + input: "mul_734" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_735" + op: "Mul" + input: "add" + input: "add_461" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_136" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/key/kernel/read" + input: "mul_735" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_604" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel" + input: "sub_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_605" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + input: "add_458" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_606" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + input: "add_459" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_736/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_736" + op: "Mul" + input: "Mul_736/x" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_737/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_737" + op: "Mul" + input: "Mul_737/x" + input: "clip_by_global_norm/clip_by_global_norm/_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_462" + op: "Add" + input: "Mul_736" + input: "Mul_737" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_738/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_738" + op: "Mul" + input: "Mul_738/x" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_136" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_739/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_739" + op: "Mul" + input: "Mul_739/x" + input: "Square_136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_463" + op: "Add" + input: "Mul_738" + input: "Mul_739" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_136" + op: "Sqrt" + input: "add_463" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_464/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_464" + op: "Add" + input: "Sqrt_136" + input: "add_464/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_137" + op: "RealDiv" + input: "add_462" + input: "add_464" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_740" + op: "Mul" + input: "add" + input: "truediv_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_137" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/key/bias/read" + input: "mul_740" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_607" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias" + input: "sub_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_608" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + input: "add_462" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_609" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + input: "add_463" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_741/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_741" + op: "Mul" + input: "Mul_741/x" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_742/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_742" + op: "Mul" + input: "Mul_742/x" + input: "clip_by_global_norm/clip_by_global_norm/_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_465" + op: "Add" + input: "Mul_741" + input: "Mul_742" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_743/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_743" + op: "Mul" + input: "Mul_743/x" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_137" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_744/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_744" + op: "Mul" + input: "Mul_744/x" + input: "Square_137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_466" + op: "Add" + input: "Mul_743" + input: "Mul_744" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_137" + op: "Sqrt" + input: "add_466" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_467/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_467" + op: "Add" + input: "Sqrt_137" + input: "add_467/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_138" + op: "RealDiv" + input: "add_465" + input: "add_467" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_745/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_745" + op: "Mul" + input: "mul_745/x" + input: "bert/encoder/layer_8/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_468" + op: "Add" + input: "truediv_138" + input: "mul_745" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_746" + op: "Mul" + input: "add" + input: "add_468" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_138" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/value/kernel/read" + input: "mul_746" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_610" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel" + input: "sub_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_611" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + input: "add_465" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_612" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + input: "add_466" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_747/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_747" + op: "Mul" + input: "Mul_747/x" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_748/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_748" + op: "Mul" + input: "Mul_748/x" + input: "clip_by_global_norm/clip_by_global_norm/_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_469" + op: "Add" + input: "Mul_747" + input: "Mul_748" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_749/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_749" + op: "Mul" + input: "Mul_749/x" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_138" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_750/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_750" + op: "Mul" + input: "Mul_750/x" + input: "Square_138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_470" + op: "Add" + input: "Mul_749" + input: "Mul_750" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_138" + op: "Sqrt" + input: "add_470" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_471/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_471" + op: "Add" + input: "Sqrt_138" + input: "add_471/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_139" + op: "RealDiv" + input: "add_469" + input: "add_471" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_751" + op: "Mul" + input: "add" + input: "truediv_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_139" + op: "Sub" + input: "bert/encoder/layer_8/attention/self/value/bias/read" + input: "mul_751" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_613" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias" + input: "sub_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_614" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + input: "add_469" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_615" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + input: "add_470" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_752/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_752" + op: "Mul" + input: "Mul_752/x" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_753/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_753" + op: "Mul" + input: "Mul_753/x" + input: "clip_by_global_norm/clip_by_global_norm/_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_472" + op: "Add" + input: "Mul_752" + input: "Mul_753" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_754/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_754" + op: "Mul" + input: "Mul_754/x" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_139" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_755/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_755" + op: "Mul" + input: "Mul_755/x" + input: "Square_139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_473" + op: "Add" + input: "Mul_754" + input: "Mul_755" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_139" + op: "Sqrt" + input: "add_473" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_474/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_474" + op: "Add" + input: "Sqrt_139" + input: "add_474/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_140" + op: "RealDiv" + input: "add_472" + input: "add_474" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_756/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_756" + op: "Mul" + input: "mul_756/x" + input: "bert/encoder/layer_8/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_475" + op: "Add" + input: "truediv_140" + input: "mul_756" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_757" + op: "Mul" + input: "add" + input: "add_475" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_140" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/dense/kernel/read" + input: "mul_757" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_616" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + input: "sub_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_617" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + input: "add_472" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_618" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + input: "add_473" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_758/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_758" + op: "Mul" + input: "Mul_758/x" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_759/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_759" + op: "Mul" + input: "Mul_759/x" + input: "clip_by_global_norm/clip_by_global_norm/_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_476" + op: "Add" + input: "Mul_758" + input: "Mul_759" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_760/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_760" + op: "Mul" + input: "Mul_760/x" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_140" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_761/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_761" + op: "Mul" + input: "Mul_761/x" + input: "Square_140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_477" + op: "Add" + input: "Mul_760" + input: "Mul_761" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_140" + op: "Sqrt" + input: "add_477" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_478/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_478" + op: "Add" + input: "Sqrt_140" + input: "add_478/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_141" + op: "RealDiv" + input: "add_476" + input: "add_478" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_762" + op: "Mul" + input: "add" + input: "truediv_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_141" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/dense/bias/read" + input: "mul_762" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_619" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias" + input: "sub_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_620" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + input: "add_476" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_621" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + input: "add_477" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_763/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_763" + op: "Mul" + input: "Mul_763/x" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_764/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_764" + op: "Mul" + input: "Mul_764/x" + input: "clip_by_global_norm/clip_by_global_norm/_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_479" + op: "Add" + input: "Mul_763" + input: "Mul_764" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_765/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_765" + op: "Mul" + input: "Mul_765/x" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_141" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_766/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_766" + op: "Mul" + input: "Mul_766/x" + input: "Square_141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_480" + op: "Add" + input: "Mul_765" + input: "Mul_766" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_141" + op: "Sqrt" + input: "add_480" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_481/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_481" + op: "Add" + input: "Sqrt_141" + input: "add_481/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_142" + op: "RealDiv" + input: "add_479" + input: "add_481" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_767" + op: "Mul" + input: "add" + input: "truediv_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_142" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/read" + input: "mul_767" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_622" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + input: "sub_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_623" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + input: "add_479" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_624" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + input: "add_480" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_768/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_768" + op: "Mul" + input: "Mul_768/x" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_769/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_769" + op: "Mul" + input: "Mul_769/x" + input: "clip_by_global_norm/clip_by_global_norm/_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_482" + op: "Add" + input: "Mul_768" + input: "Mul_769" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_770/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_770" + op: "Mul" + input: "Mul_770/x" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_142" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_771/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_771" + op: "Mul" + input: "Mul_771/x" + input: "Square_142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_483" + op: "Add" + input: "Mul_770" + input: "Mul_771" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_142" + op: "Sqrt" + input: "add_483" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_484/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_484" + op: "Add" + input: "Sqrt_142" + input: "add_484/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_143" + op: "RealDiv" + input: "add_482" + input: "add_484" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_772" + op: "Mul" + input: "add" + input: "truediv_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_143" + op: "Sub" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/read" + input: "mul_772" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_625" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + input: "sub_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_626" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + input: "add_482" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_627" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + input: "add_483" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_773/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_773" + op: "Mul" + input: "Mul_773/x" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_774/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_774" + op: "Mul" + input: "Mul_774/x" + input: "clip_by_global_norm/clip_by_global_norm/_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_485" + op: "Add" + input: "Mul_773" + input: "Mul_774" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_775/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_775" + op: "Mul" + input: "Mul_775/x" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_143" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_776/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_776" + op: "Mul" + input: "Mul_776/x" + input: "Square_143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_486" + op: "Add" + input: "Mul_775" + input: "Mul_776" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_143" + op: "Sqrt" + input: "add_486" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_487/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_487" + op: "Add" + input: "Sqrt_143" + input: "add_487/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_144" + op: "RealDiv" + input: "add_485" + input: "add_487" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_777/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_777" + op: "Mul" + input: "mul_777/x" + input: "bert/encoder/layer_8/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_488" + op: "Add" + input: "truediv_144" + input: "mul_777" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_778" + op: "Mul" + input: "add" + input: "add_488" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_144" + op: "Sub" + input: "bert/encoder/layer_8/intermediate/dense/kernel/read" + input: "mul_778" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_628" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + input: "sub_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_629" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + input: "add_485" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_630" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + input: "add_486" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_779/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_779" + op: "Mul" + input: "Mul_779/x" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_780/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_780" + op: "Mul" + input: "Mul_780/x" + input: "clip_by_global_norm/clip_by_global_norm/_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_489" + op: "Add" + input: "Mul_779" + input: "Mul_780" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_781/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_781" + op: "Mul" + input: "Mul_781/x" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_144" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_782/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_782" + op: "Mul" + input: "Mul_782/x" + input: "Square_144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_490" + op: "Add" + input: "Mul_781" + input: "Mul_782" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_144" + op: "Sqrt" + input: "add_490" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_491/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_491" + op: "Add" + input: "Sqrt_144" + input: "add_491/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_145" + op: "RealDiv" + input: "add_489" + input: "add_491" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_783" + op: "Mul" + input: "add" + input: "truediv_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_145" + op: "Sub" + input: "bert/encoder/layer_8/intermediate/dense/bias/read" + input: "mul_783" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_631" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias" + input: "sub_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_632" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + input: "add_489" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_633" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + input: "add_490" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_784/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_784" + op: "Mul" + input: "Mul_784/x" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_785/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_785" + op: "Mul" + input: "Mul_785/x" + input: "clip_by_global_norm/clip_by_global_norm/_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_492" + op: "Add" + input: "Mul_784" + input: "Mul_785" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_786/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_786" + op: "Mul" + input: "Mul_786/x" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_145" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_787/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_787" + op: "Mul" + input: "Mul_787/x" + input: "Square_145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_493" + op: "Add" + input: "Mul_786" + input: "Mul_787" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_145" + op: "Sqrt" + input: "add_493" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_494/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_494" + op: "Add" + input: "Sqrt_145" + input: "add_494/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_146" + op: "RealDiv" + input: "add_492" + input: "add_494" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_788/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_788" + op: "Mul" + input: "mul_788/x" + input: "bert/encoder/layer_8/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_495" + op: "Add" + input: "truediv_146" + input: "mul_788" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_789" + op: "Mul" + input: "add" + input: "add_495" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_146" + op: "Sub" + input: "bert/encoder/layer_8/output/dense/kernel/read" + input: "mul_789" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_634" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel" + input: "sub_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_635" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + input: "add_492" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_636" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + input: "add_493" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + input: "bert/encoder/layer_8/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + input: "bert/encoder/layer_8/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_790/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_790" + op: "Mul" + input: "Mul_790/x" + input: "bert/encoder/layer_8/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_791/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_791" + op: "Mul" + input: "Mul_791/x" + input: "clip_by_global_norm/clip_by_global_norm/_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_496" + op: "Add" + input: "Mul_790" + input: "Mul_791" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_792/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_792" + op: "Mul" + input: "Mul_792/x" + input: "bert/encoder/layer_8/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_146" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_793/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_793" + op: "Mul" + input: "Mul_793/x" + input: "Square_146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_497" + op: "Add" + input: "Mul_792" + input: "Mul_793" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_146" + op: "Sqrt" + input: "add_497" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_498/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_498" + op: "Add" + input: "Sqrt_146" + input: "add_498/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_147" + op: "RealDiv" + input: "add_496" + input: "add_498" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_794" + op: "Mul" + input: "add" + input: "truediv_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_147" + op: "Sub" + input: "bert/encoder/layer_8/output/dense/bias/read" + input: "mul_794" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_637" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias" + input: "sub_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_638" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + input: "add_496" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_639" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + input: "add_497" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_795/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_795" + op: "Mul" + input: "Mul_795/x" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_796/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_796" + op: "Mul" + input: "Mul_796/x" + input: "clip_by_global_norm/clip_by_global_norm/_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_499" + op: "Add" + input: "Mul_795" + input: "Mul_796" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_797/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_797" + op: "Mul" + input: "Mul_797/x" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_147" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_798/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_798" + op: "Mul" + input: "Mul_798/x" + input: "Square_147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_500" + op: "Add" + input: "Mul_797" + input: "Mul_798" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_147" + op: "Sqrt" + input: "add_500" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_501/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_501" + op: "Add" + input: "Sqrt_147" + input: "add_501/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_148" + op: "RealDiv" + input: "add_499" + input: "add_501" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_799" + op: "Mul" + input: "add" + input: "truediv_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_148" + op: "Sub" + input: "bert/encoder/layer_8/output/LayerNorm/beta/read" + input: "mul_799" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_640" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + input: "sub_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_641" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + input: "add_499" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_642" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + input: "add_500" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_800/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_800" + op: "Mul" + input: "Mul_800/x" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_801/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_801" + op: "Mul" + input: "Mul_801/x" + input: "clip_by_global_norm/clip_by_global_norm/_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_502" + op: "Add" + input: "Mul_800" + input: "Mul_801" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_802/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_802" + op: "Mul" + input: "Mul_802/x" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_148" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_803/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_803" + op: "Mul" + input: "Mul_803/x" + input: "Square_148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_503" + op: "Add" + input: "Mul_802" + input: "Mul_803" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_148" + op: "Sqrt" + input: "add_503" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_504/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_504" + op: "Add" + input: "Sqrt_148" + input: "add_504/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_149" + op: "RealDiv" + input: "add_502" + input: "add_504" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_804" + op: "Mul" + input: "add" + input: "truediv_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_149" + op: "Sub" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/read" + input: "mul_804" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_643" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + input: "sub_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_644" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + input: "add_502" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_645" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + input: "add_503" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_805/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_805" + op: "Mul" + input: "Mul_805/x" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_806/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_806" + op: "Mul" + input: "Mul_806/x" + input: "clip_by_global_norm/clip_by_global_norm/_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_505" + op: "Add" + input: "Mul_805" + input: "Mul_806" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_807/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_807" + op: "Mul" + input: "Mul_807/x" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_149" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_808/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_808" + op: "Mul" + input: "Mul_808/x" + input: "Square_149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_506" + op: "Add" + input: "Mul_807" + input: "Mul_808" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_149" + op: "Sqrt" + input: "add_506" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_507/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_507" + op: "Add" + input: "Sqrt_149" + input: "add_507/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_150" + op: "RealDiv" + input: "add_505" + input: "add_507" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_809/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_809" + op: "Mul" + input: "mul_809/x" + input: "bert/encoder/layer_9/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_508" + op: "Add" + input: "truediv_150" + input: "mul_809" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_810" + op: "Mul" + input: "add" + input: "add_508" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_150" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/query/kernel/read" + input: "mul_810" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_646" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel" + input: "sub_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_647" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + input: "add_505" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_648" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + input: "add_506" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_811/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_811" + op: "Mul" + input: "Mul_811/x" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_812/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_812" + op: "Mul" + input: "Mul_812/x" + input: "clip_by_global_norm/clip_by_global_norm/_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_509" + op: "Add" + input: "Mul_811" + input: "Mul_812" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_813/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_813" + op: "Mul" + input: "Mul_813/x" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_150" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_814/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_814" + op: "Mul" + input: "Mul_814/x" + input: "Square_150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_510" + op: "Add" + input: "Mul_813" + input: "Mul_814" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_150" + op: "Sqrt" + input: "add_510" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_511/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_511" + op: "Add" + input: "Sqrt_150" + input: "add_511/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_151" + op: "RealDiv" + input: "add_509" + input: "add_511" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_815" + op: "Mul" + input: "add" + input: "truediv_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_151" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/query/bias/read" + input: "mul_815" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_649" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias" + input: "sub_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_650" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + input: "add_509" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_651" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + input: "add_510" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_816/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_816" + op: "Mul" + input: "Mul_816/x" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_817/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_817" + op: "Mul" + input: "Mul_817/x" + input: "clip_by_global_norm/clip_by_global_norm/_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_512" + op: "Add" + input: "Mul_816" + input: "Mul_817" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_818/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_818" + op: "Mul" + input: "Mul_818/x" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_151" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_819/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_819" + op: "Mul" + input: "Mul_819/x" + input: "Square_151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_513" + op: "Add" + input: "Mul_818" + input: "Mul_819" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_151" + op: "Sqrt" + input: "add_513" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_514/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_514" + op: "Add" + input: "Sqrt_151" + input: "add_514/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_152" + op: "RealDiv" + input: "add_512" + input: "add_514" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_820/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_820" + op: "Mul" + input: "mul_820/x" + input: "bert/encoder/layer_9/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_515" + op: "Add" + input: "truediv_152" + input: "mul_820" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_821" + op: "Mul" + input: "add" + input: "add_515" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_152" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/key/kernel/read" + input: "mul_821" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_652" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel" + input: "sub_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_653" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + input: "add_512" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_654" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + input: "add_513" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_822/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_822" + op: "Mul" + input: "Mul_822/x" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_823/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_823" + op: "Mul" + input: "Mul_823/x" + input: "clip_by_global_norm/clip_by_global_norm/_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_516" + op: "Add" + input: "Mul_822" + input: "Mul_823" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_824/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_824" + op: "Mul" + input: "Mul_824/x" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_152" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_825/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_825" + op: "Mul" + input: "Mul_825/x" + input: "Square_152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_517" + op: "Add" + input: "Mul_824" + input: "Mul_825" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_152" + op: "Sqrt" + input: "add_517" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_518/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_518" + op: "Add" + input: "Sqrt_152" + input: "add_518/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_153" + op: "RealDiv" + input: "add_516" + input: "add_518" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_826" + op: "Mul" + input: "add" + input: "truediv_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_153" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/key/bias/read" + input: "mul_826" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_655" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias" + input: "sub_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_656" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + input: "add_516" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_657" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + input: "add_517" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_827/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_827" + op: "Mul" + input: "Mul_827/x" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_828/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_828" + op: "Mul" + input: "Mul_828/x" + input: "clip_by_global_norm/clip_by_global_norm/_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_519" + op: "Add" + input: "Mul_827" + input: "Mul_828" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_829/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_829" + op: "Mul" + input: "Mul_829/x" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_153" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_830/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_830" + op: "Mul" + input: "Mul_830/x" + input: "Square_153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_520" + op: "Add" + input: "Mul_829" + input: "Mul_830" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_153" + op: "Sqrt" + input: "add_520" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_521/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_521" + op: "Add" + input: "Sqrt_153" + input: "add_521/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_154" + op: "RealDiv" + input: "add_519" + input: "add_521" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_831/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_831" + op: "Mul" + input: "mul_831/x" + input: "bert/encoder/layer_9/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_522" + op: "Add" + input: "truediv_154" + input: "mul_831" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_832" + op: "Mul" + input: "add" + input: "add_522" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_154" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/value/kernel/read" + input: "mul_832" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_658" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel" + input: "sub_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_659" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + input: "add_519" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_660" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + input: "add_520" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_833/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_833" + op: "Mul" + input: "Mul_833/x" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_834/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_834" + op: "Mul" + input: "Mul_834/x" + input: "clip_by_global_norm/clip_by_global_norm/_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_523" + op: "Add" + input: "Mul_833" + input: "Mul_834" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_835/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_835" + op: "Mul" + input: "Mul_835/x" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_154" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_836/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_836" + op: "Mul" + input: "Mul_836/x" + input: "Square_154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_524" + op: "Add" + input: "Mul_835" + input: "Mul_836" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_154" + op: "Sqrt" + input: "add_524" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_525/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_525" + op: "Add" + input: "Sqrt_154" + input: "add_525/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_155" + op: "RealDiv" + input: "add_523" + input: "add_525" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_837" + op: "Mul" + input: "add" + input: "truediv_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_155" + op: "Sub" + input: "bert/encoder/layer_9/attention/self/value/bias/read" + input: "mul_837" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_661" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias" + input: "sub_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_662" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + input: "add_523" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_663" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + input: "add_524" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_838/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_838" + op: "Mul" + input: "Mul_838/x" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_839/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_839" + op: "Mul" + input: "Mul_839/x" + input: "clip_by_global_norm/clip_by_global_norm/_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_526" + op: "Add" + input: "Mul_838" + input: "Mul_839" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_840/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_840" + op: "Mul" + input: "Mul_840/x" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_155" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_841/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_841" + op: "Mul" + input: "Mul_841/x" + input: "Square_155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_527" + op: "Add" + input: "Mul_840" + input: "Mul_841" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_155" + op: "Sqrt" + input: "add_527" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_528/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_528" + op: "Add" + input: "Sqrt_155" + input: "add_528/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_156" + op: "RealDiv" + input: "add_526" + input: "add_528" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_842/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_842" + op: "Mul" + input: "mul_842/x" + input: "bert/encoder/layer_9/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_529" + op: "Add" + input: "truediv_156" + input: "mul_842" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_843" + op: "Mul" + input: "add" + input: "add_529" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_156" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/dense/kernel/read" + input: "mul_843" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_664" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + input: "sub_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_665" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + input: "add_526" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_666" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + input: "add_527" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_844/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_844" + op: "Mul" + input: "Mul_844/x" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_845/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_845" + op: "Mul" + input: "Mul_845/x" + input: "clip_by_global_norm/clip_by_global_norm/_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_530" + op: "Add" + input: "Mul_844" + input: "Mul_845" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_846/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_846" + op: "Mul" + input: "Mul_846/x" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_156" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_847/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_847" + op: "Mul" + input: "Mul_847/x" + input: "Square_156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_531" + op: "Add" + input: "Mul_846" + input: "Mul_847" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_156" + op: "Sqrt" + input: "add_531" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_532/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_532" + op: "Add" + input: "Sqrt_156" + input: "add_532/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_157" + op: "RealDiv" + input: "add_530" + input: "add_532" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_848" + op: "Mul" + input: "add" + input: "truediv_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_157" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/dense/bias/read" + input: "mul_848" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_667" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias" + input: "sub_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_668" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + input: "add_530" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_669" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + input: "add_531" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_849/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_849" + op: "Mul" + input: "Mul_849/x" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_850/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_850" + op: "Mul" + input: "Mul_850/x" + input: "clip_by_global_norm/clip_by_global_norm/_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_533" + op: "Add" + input: "Mul_849" + input: "Mul_850" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_851/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_851" + op: "Mul" + input: "Mul_851/x" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_157" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_852/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_852" + op: "Mul" + input: "Mul_852/x" + input: "Square_157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_534" + op: "Add" + input: "Mul_851" + input: "Mul_852" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_157" + op: "Sqrt" + input: "add_534" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_535/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_535" + op: "Add" + input: "Sqrt_157" + input: "add_535/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_158" + op: "RealDiv" + input: "add_533" + input: "add_535" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_853" + op: "Mul" + input: "add" + input: "truediv_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_158" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/read" + input: "mul_853" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_670" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + input: "sub_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_671" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + input: "add_533" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_672" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + input: "add_534" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_854/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_854" + op: "Mul" + input: "Mul_854/x" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_855/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_855" + op: "Mul" + input: "Mul_855/x" + input: "clip_by_global_norm/clip_by_global_norm/_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_536" + op: "Add" + input: "Mul_854" + input: "Mul_855" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_856/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_856" + op: "Mul" + input: "Mul_856/x" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_158" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_857/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_857" + op: "Mul" + input: "Mul_857/x" + input: "Square_158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_537" + op: "Add" + input: "Mul_856" + input: "Mul_857" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_158" + op: "Sqrt" + input: "add_537" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_538/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_538" + op: "Add" + input: "Sqrt_158" + input: "add_538/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_159" + op: "RealDiv" + input: "add_536" + input: "add_538" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_858" + op: "Mul" + input: "add" + input: "truediv_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_159" + op: "Sub" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/read" + input: "mul_858" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_673" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + input: "sub_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_674" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + input: "add_536" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_675" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + input: "add_537" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_859/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_859" + op: "Mul" + input: "Mul_859/x" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_860/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_860" + op: "Mul" + input: "Mul_860/x" + input: "clip_by_global_norm/clip_by_global_norm/_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_539" + op: "Add" + input: "Mul_859" + input: "Mul_860" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_861/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_861" + op: "Mul" + input: "Mul_861/x" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_159" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_862/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_862" + op: "Mul" + input: "Mul_862/x" + input: "Square_159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_540" + op: "Add" + input: "Mul_861" + input: "Mul_862" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_159" + op: "Sqrt" + input: "add_540" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_541/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_541" + op: "Add" + input: "Sqrt_159" + input: "add_541/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_160" + op: "RealDiv" + input: "add_539" + input: "add_541" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_863/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_863" + op: "Mul" + input: "mul_863/x" + input: "bert/encoder/layer_9/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_542" + op: "Add" + input: "truediv_160" + input: "mul_863" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_864" + op: "Mul" + input: "add" + input: "add_542" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_160" + op: "Sub" + input: "bert/encoder/layer_9/intermediate/dense/kernel/read" + input: "mul_864" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_676" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + input: "sub_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_677" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + input: "add_539" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_678" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + input: "add_540" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_865/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_865" + op: "Mul" + input: "Mul_865/x" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_866/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_866" + op: "Mul" + input: "Mul_866/x" + input: "clip_by_global_norm/clip_by_global_norm/_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_543" + op: "Add" + input: "Mul_865" + input: "Mul_866" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_867/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_867" + op: "Mul" + input: "Mul_867/x" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_160" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_868/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_868" + op: "Mul" + input: "Mul_868/x" + input: "Square_160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_544" + op: "Add" + input: "Mul_867" + input: "Mul_868" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_160" + op: "Sqrt" + input: "add_544" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_545/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_545" + op: "Add" + input: "Sqrt_160" + input: "add_545/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_161" + op: "RealDiv" + input: "add_543" + input: "add_545" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_869" + op: "Mul" + input: "add" + input: "truediv_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_161" + op: "Sub" + input: "bert/encoder/layer_9/intermediate/dense/bias/read" + input: "mul_869" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_679" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias" + input: "sub_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_680" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + input: "add_543" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_681" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + input: "add_544" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_870/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_870" + op: "Mul" + input: "Mul_870/x" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_871/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_871" + op: "Mul" + input: "Mul_871/x" + input: "clip_by_global_norm/clip_by_global_norm/_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_546" + op: "Add" + input: "Mul_870" + input: "Mul_871" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_872/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_872" + op: "Mul" + input: "Mul_872/x" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_161" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_873/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_873" + op: "Mul" + input: "Mul_873/x" + input: "Square_161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_547" + op: "Add" + input: "Mul_872" + input: "Mul_873" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_161" + op: "Sqrt" + input: "add_547" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_548/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_548" + op: "Add" + input: "Sqrt_161" + input: "add_548/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_162" + op: "RealDiv" + input: "add_546" + input: "add_548" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_874/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_874" + op: "Mul" + input: "mul_874/x" + input: "bert/encoder/layer_9/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_549" + op: "Add" + input: "truediv_162" + input: "mul_874" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_875" + op: "Mul" + input: "add" + input: "add_549" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_162" + op: "Sub" + input: "bert/encoder/layer_9/output/dense/kernel/read" + input: "mul_875" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_682" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel" + input: "sub_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_683" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + input: "add_546" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_684" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + input: "add_547" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + input: "bert/encoder/layer_9/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + input: "bert/encoder/layer_9/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_876/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_876" + op: "Mul" + input: "Mul_876/x" + input: "bert/encoder/layer_9/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_877/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_877" + op: "Mul" + input: "Mul_877/x" + input: "clip_by_global_norm/clip_by_global_norm/_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_550" + op: "Add" + input: "Mul_876" + input: "Mul_877" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_878/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_878" + op: "Mul" + input: "Mul_878/x" + input: "bert/encoder/layer_9/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_162" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_879/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_879" + op: "Mul" + input: "Mul_879/x" + input: "Square_162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_551" + op: "Add" + input: "Mul_878" + input: "Mul_879" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_162" + op: "Sqrt" + input: "add_551" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_552/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_552" + op: "Add" + input: "Sqrt_162" + input: "add_552/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_163" + op: "RealDiv" + input: "add_550" + input: "add_552" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_880" + op: "Mul" + input: "add" + input: "truediv_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_163" + op: "Sub" + input: "bert/encoder/layer_9/output/dense/bias/read" + input: "mul_880" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_685" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias" + input: "sub_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_686" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + input: "add_550" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_687" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + input: "add_551" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_881/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_881" + op: "Mul" + input: "Mul_881/x" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_882/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_882" + op: "Mul" + input: "Mul_882/x" + input: "clip_by_global_norm/clip_by_global_norm/_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_553" + op: "Add" + input: "Mul_881" + input: "Mul_882" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_883/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_883" + op: "Mul" + input: "Mul_883/x" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_163" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_884/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_884" + op: "Mul" + input: "Mul_884/x" + input: "Square_163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_554" + op: "Add" + input: "Mul_883" + input: "Mul_884" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_163" + op: "Sqrt" + input: "add_554" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_555/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_555" + op: "Add" + input: "Sqrt_163" + input: "add_555/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_164" + op: "RealDiv" + input: "add_553" + input: "add_555" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_885" + op: "Mul" + input: "add" + input: "truediv_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_164" + op: "Sub" + input: "bert/encoder/layer_9/output/LayerNorm/beta/read" + input: "mul_885" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_688" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + input: "sub_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_689" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + input: "add_553" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_690" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + input: "add_554" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_886/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_886" + op: "Mul" + input: "Mul_886/x" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_887/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_887" + op: "Mul" + input: "Mul_887/x" + input: "clip_by_global_norm/clip_by_global_norm/_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_556" + op: "Add" + input: "Mul_886" + input: "Mul_887" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_888/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_888" + op: "Mul" + input: "Mul_888/x" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_164" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_889/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_889" + op: "Mul" + input: "Mul_889/x" + input: "Square_164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_557" + op: "Add" + input: "Mul_888" + input: "Mul_889" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_164" + op: "Sqrt" + input: "add_557" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_558/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_558" + op: "Add" + input: "Sqrt_164" + input: "add_558/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_165" + op: "RealDiv" + input: "add_556" + input: "add_558" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_890" + op: "Mul" + input: "add" + input: "truediv_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_165" + op: "Sub" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/read" + input: "mul_890" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_691" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + input: "sub_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_692" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + input: "add_556" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_693" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + input: "add_557" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_891/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_891" + op: "Mul" + input: "Mul_891/x" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_892/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_892" + op: "Mul" + input: "Mul_892/x" + input: "clip_by_global_norm/clip_by_global_norm/_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_559" + op: "Add" + input: "Mul_891" + input: "Mul_892" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_893/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_893" + op: "Mul" + input: "Mul_893/x" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_165" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_894/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_894" + op: "Mul" + input: "Mul_894/x" + input: "Square_165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_560" + op: "Add" + input: "Mul_893" + input: "Mul_894" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_165" + op: "Sqrt" + input: "add_560" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_561/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_561" + op: "Add" + input: "Sqrt_165" + input: "add_561/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_166" + op: "RealDiv" + input: "add_559" + input: "add_561" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_895/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_895" + op: "Mul" + input: "mul_895/x" + input: "bert/encoder/layer_10/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_562" + op: "Add" + input: "truediv_166" + input: "mul_895" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_896" + op: "Mul" + input: "add" + input: "add_562" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_166" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/query/kernel/read" + input: "mul_896" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_694" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel" + input: "sub_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_695" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + input: "add_559" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_696" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + input: "add_560" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_897/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_897" + op: "Mul" + input: "Mul_897/x" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_898/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_898" + op: "Mul" + input: "Mul_898/x" + input: "clip_by_global_norm/clip_by_global_norm/_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_563" + op: "Add" + input: "Mul_897" + input: "Mul_898" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_899/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_899" + op: "Mul" + input: "Mul_899/x" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_166" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_900/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_900" + op: "Mul" + input: "Mul_900/x" + input: "Square_166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_564" + op: "Add" + input: "Mul_899" + input: "Mul_900" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_166" + op: "Sqrt" + input: "add_564" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_565/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_565" + op: "Add" + input: "Sqrt_166" + input: "add_565/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_167" + op: "RealDiv" + input: "add_563" + input: "add_565" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_901" + op: "Mul" + input: "add" + input: "truediv_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_167" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/query/bias/read" + input: "mul_901" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_697" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias" + input: "sub_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_698" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + input: "add_563" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_699" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + input: "add_564" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_902/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_902" + op: "Mul" + input: "Mul_902/x" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_903/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_903" + op: "Mul" + input: "Mul_903/x" + input: "clip_by_global_norm/clip_by_global_norm/_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_566" + op: "Add" + input: "Mul_902" + input: "Mul_903" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_904/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_904" + op: "Mul" + input: "Mul_904/x" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_167" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_905/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_905" + op: "Mul" + input: "Mul_905/x" + input: "Square_167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_567" + op: "Add" + input: "Mul_904" + input: "Mul_905" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_167" + op: "Sqrt" + input: "add_567" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_568/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_568" + op: "Add" + input: "Sqrt_167" + input: "add_568/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_168" + op: "RealDiv" + input: "add_566" + input: "add_568" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_906/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_906" + op: "Mul" + input: "mul_906/x" + input: "bert/encoder/layer_10/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_569" + op: "Add" + input: "truediv_168" + input: "mul_906" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_907" + op: "Mul" + input: "add" + input: "add_569" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_168" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/key/kernel/read" + input: "mul_907" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_700" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel" + input: "sub_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_701" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + input: "add_566" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_702" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + input: "add_567" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_908/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_908" + op: "Mul" + input: "Mul_908/x" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_909/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_909" + op: "Mul" + input: "Mul_909/x" + input: "clip_by_global_norm/clip_by_global_norm/_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_570" + op: "Add" + input: "Mul_908" + input: "Mul_909" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_910/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_910" + op: "Mul" + input: "Mul_910/x" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_168" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_911/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_911" + op: "Mul" + input: "Mul_911/x" + input: "Square_168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_571" + op: "Add" + input: "Mul_910" + input: "Mul_911" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_168" + op: "Sqrt" + input: "add_571" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_572/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_572" + op: "Add" + input: "Sqrt_168" + input: "add_572/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_169" + op: "RealDiv" + input: "add_570" + input: "add_572" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_912" + op: "Mul" + input: "add" + input: "truediv_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_169" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/key/bias/read" + input: "mul_912" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_703" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias" + input: "sub_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_704" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + input: "add_570" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_705" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + input: "add_571" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_913/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_913" + op: "Mul" + input: "Mul_913/x" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_914/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_914" + op: "Mul" + input: "Mul_914/x" + input: "clip_by_global_norm/clip_by_global_norm/_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_573" + op: "Add" + input: "Mul_913" + input: "Mul_914" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_915/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_915" + op: "Mul" + input: "Mul_915/x" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_169" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_916/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_916" + op: "Mul" + input: "Mul_916/x" + input: "Square_169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_574" + op: "Add" + input: "Mul_915" + input: "Mul_916" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_169" + op: "Sqrt" + input: "add_574" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_575/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_575" + op: "Add" + input: "Sqrt_169" + input: "add_575/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_170" + op: "RealDiv" + input: "add_573" + input: "add_575" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_917/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_917" + op: "Mul" + input: "mul_917/x" + input: "bert/encoder/layer_10/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_576" + op: "Add" + input: "truediv_170" + input: "mul_917" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_918" + op: "Mul" + input: "add" + input: "add_576" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_170" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/value/kernel/read" + input: "mul_918" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_706" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel" + input: "sub_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_707" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + input: "add_573" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_708" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + input: "add_574" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_919/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_919" + op: "Mul" + input: "Mul_919/x" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_920/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_920" + op: "Mul" + input: "Mul_920/x" + input: "clip_by_global_norm/clip_by_global_norm/_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_577" + op: "Add" + input: "Mul_919" + input: "Mul_920" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_921/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_921" + op: "Mul" + input: "Mul_921/x" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_170" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_922/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_922" + op: "Mul" + input: "Mul_922/x" + input: "Square_170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_578" + op: "Add" + input: "Mul_921" + input: "Mul_922" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_170" + op: "Sqrt" + input: "add_578" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_579/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_579" + op: "Add" + input: "Sqrt_170" + input: "add_579/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_171" + op: "RealDiv" + input: "add_577" + input: "add_579" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_923" + op: "Mul" + input: "add" + input: "truediv_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_171" + op: "Sub" + input: "bert/encoder/layer_10/attention/self/value/bias/read" + input: "mul_923" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_709" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias" + input: "sub_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_710" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + input: "add_577" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_711" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + input: "add_578" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_924/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_924" + op: "Mul" + input: "Mul_924/x" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_925/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_925" + op: "Mul" + input: "Mul_925/x" + input: "clip_by_global_norm/clip_by_global_norm/_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_580" + op: "Add" + input: "Mul_924" + input: "Mul_925" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_926/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_926" + op: "Mul" + input: "Mul_926/x" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_171" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_927/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_927" + op: "Mul" + input: "Mul_927/x" + input: "Square_171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_581" + op: "Add" + input: "Mul_926" + input: "Mul_927" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_171" + op: "Sqrt" + input: "add_581" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_582/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_582" + op: "Add" + input: "Sqrt_171" + input: "add_582/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_172" + op: "RealDiv" + input: "add_580" + input: "add_582" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_928/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_928" + op: "Mul" + input: "mul_928/x" + input: "bert/encoder/layer_10/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_583" + op: "Add" + input: "truediv_172" + input: "mul_928" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_929" + op: "Mul" + input: "add" + input: "add_583" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_172" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/dense/kernel/read" + input: "mul_929" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_712" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + input: "sub_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_713" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + input: "add_580" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_714" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + input: "add_581" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_930/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_930" + op: "Mul" + input: "Mul_930/x" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_931/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_931" + op: "Mul" + input: "Mul_931/x" + input: "clip_by_global_norm/clip_by_global_norm/_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_584" + op: "Add" + input: "Mul_930" + input: "Mul_931" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_932/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_932" + op: "Mul" + input: "Mul_932/x" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_172" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_933/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_933" + op: "Mul" + input: "Mul_933/x" + input: "Square_172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_585" + op: "Add" + input: "Mul_932" + input: "Mul_933" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_172" + op: "Sqrt" + input: "add_585" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_586/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_586" + op: "Add" + input: "Sqrt_172" + input: "add_586/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_173" + op: "RealDiv" + input: "add_584" + input: "add_586" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_934" + op: "Mul" + input: "add" + input: "truediv_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_173" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/dense/bias/read" + input: "mul_934" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_715" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias" + input: "sub_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_716" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + input: "add_584" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_717" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + input: "add_585" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_935/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_935" + op: "Mul" + input: "Mul_935/x" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_936/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_936" + op: "Mul" + input: "Mul_936/x" + input: "clip_by_global_norm/clip_by_global_norm/_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_587" + op: "Add" + input: "Mul_935" + input: "Mul_936" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_937/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_937" + op: "Mul" + input: "Mul_937/x" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_173" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_938/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_938" + op: "Mul" + input: "Mul_938/x" + input: "Square_173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_588" + op: "Add" + input: "Mul_937" + input: "Mul_938" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_173" + op: "Sqrt" + input: "add_588" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_589/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_589" + op: "Add" + input: "Sqrt_173" + input: "add_589/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_174" + op: "RealDiv" + input: "add_587" + input: "add_589" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_939" + op: "Mul" + input: "add" + input: "truediv_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_174" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/read" + input: "mul_939" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_718" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + input: "sub_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_719" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + input: "add_587" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_720" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + input: "add_588" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_940/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_940" + op: "Mul" + input: "Mul_940/x" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_941/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_941" + op: "Mul" + input: "Mul_941/x" + input: "clip_by_global_norm/clip_by_global_norm/_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_590" + op: "Add" + input: "Mul_940" + input: "Mul_941" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_942/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_942" + op: "Mul" + input: "Mul_942/x" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_174" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_943/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_943" + op: "Mul" + input: "Mul_943/x" + input: "Square_174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_591" + op: "Add" + input: "Mul_942" + input: "Mul_943" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_174" + op: "Sqrt" + input: "add_591" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_592/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_592" + op: "Add" + input: "Sqrt_174" + input: "add_592/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_175" + op: "RealDiv" + input: "add_590" + input: "add_592" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_944" + op: "Mul" + input: "add" + input: "truediv_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_175" + op: "Sub" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/read" + input: "mul_944" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_721" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + input: "sub_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_722" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + input: "add_590" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_723" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + input: "add_591" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_945/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_945" + op: "Mul" + input: "Mul_945/x" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_946/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_946" + op: "Mul" + input: "Mul_946/x" + input: "clip_by_global_norm/clip_by_global_norm/_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_593" + op: "Add" + input: "Mul_945" + input: "Mul_946" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_947/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_947" + op: "Mul" + input: "Mul_947/x" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_175" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_948/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_948" + op: "Mul" + input: "Mul_948/x" + input: "Square_175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_594" + op: "Add" + input: "Mul_947" + input: "Mul_948" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_175" + op: "Sqrt" + input: "add_594" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_595/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_595" + op: "Add" + input: "Sqrt_175" + input: "add_595/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_176" + op: "RealDiv" + input: "add_593" + input: "add_595" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_949/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_949" + op: "Mul" + input: "mul_949/x" + input: "bert/encoder/layer_10/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_596" + op: "Add" + input: "truediv_176" + input: "mul_949" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_950" + op: "Mul" + input: "add" + input: "add_596" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_176" + op: "Sub" + input: "bert/encoder/layer_10/intermediate/dense/kernel/read" + input: "mul_950" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_724" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + input: "sub_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_725" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + input: "add_593" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_726" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + input: "add_594" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_951/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_951" + op: "Mul" + input: "Mul_951/x" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_952/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_952" + op: "Mul" + input: "Mul_952/x" + input: "clip_by_global_norm/clip_by_global_norm/_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_597" + op: "Add" + input: "Mul_951" + input: "Mul_952" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_953/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_953" + op: "Mul" + input: "Mul_953/x" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_176" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_954/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_954" + op: "Mul" + input: "Mul_954/x" + input: "Square_176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_598" + op: "Add" + input: "Mul_953" + input: "Mul_954" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_176" + op: "Sqrt" + input: "add_598" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_599/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_599" + op: "Add" + input: "Sqrt_176" + input: "add_599/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_177" + op: "RealDiv" + input: "add_597" + input: "add_599" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_955" + op: "Mul" + input: "add" + input: "truediv_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_177" + op: "Sub" + input: "bert/encoder/layer_10/intermediate/dense/bias/read" + input: "mul_955" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_727" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias" + input: "sub_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_728" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + input: "add_597" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_729" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + input: "add_598" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_956/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_956" + op: "Mul" + input: "Mul_956/x" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_957/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_957" + op: "Mul" + input: "Mul_957/x" + input: "clip_by_global_norm/clip_by_global_norm/_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_600" + op: "Add" + input: "Mul_956" + input: "Mul_957" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_958/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_958" + op: "Mul" + input: "Mul_958/x" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_177" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_959/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_959" + op: "Mul" + input: "Mul_959/x" + input: "Square_177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_601" + op: "Add" + input: "Mul_958" + input: "Mul_959" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_177" + op: "Sqrt" + input: "add_601" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_602/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_602" + op: "Add" + input: "Sqrt_177" + input: "add_602/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_178" + op: "RealDiv" + input: "add_600" + input: "add_602" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_960/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_960" + op: "Mul" + input: "mul_960/x" + input: "bert/encoder/layer_10/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_603" + op: "Add" + input: "truediv_178" + input: "mul_960" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_961" + op: "Mul" + input: "add" + input: "add_603" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_178" + op: "Sub" + input: "bert/encoder/layer_10/output/dense/kernel/read" + input: "mul_961" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_730" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel" + input: "sub_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_731" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + input: "add_600" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_732" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + input: "add_601" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + input: "bert/encoder/layer_10/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + input: "bert/encoder/layer_10/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_962/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_962" + op: "Mul" + input: "Mul_962/x" + input: "bert/encoder/layer_10/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_963/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_963" + op: "Mul" + input: "Mul_963/x" + input: "clip_by_global_norm/clip_by_global_norm/_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_604" + op: "Add" + input: "Mul_962" + input: "Mul_963" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_964/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_964" + op: "Mul" + input: "Mul_964/x" + input: "bert/encoder/layer_10/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_178" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_965/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_965" + op: "Mul" + input: "Mul_965/x" + input: "Square_178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_605" + op: "Add" + input: "Mul_964" + input: "Mul_965" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_178" + op: "Sqrt" + input: "add_605" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_606/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_606" + op: "Add" + input: "Sqrt_178" + input: "add_606/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_179" + op: "RealDiv" + input: "add_604" + input: "add_606" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_966" + op: "Mul" + input: "add" + input: "truediv_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_179" + op: "Sub" + input: "bert/encoder/layer_10/output/dense/bias/read" + input: "mul_966" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_733" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias" + input: "sub_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_734" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + input: "add_604" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_735" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + input: "add_605" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_967/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_967" + op: "Mul" + input: "Mul_967/x" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_968/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_968" + op: "Mul" + input: "Mul_968/x" + input: "clip_by_global_norm/clip_by_global_norm/_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_607" + op: "Add" + input: "Mul_967" + input: "Mul_968" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_969/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_969" + op: "Mul" + input: "Mul_969/x" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_179" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_970/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_970" + op: "Mul" + input: "Mul_970/x" + input: "Square_179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_608" + op: "Add" + input: "Mul_969" + input: "Mul_970" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_179" + op: "Sqrt" + input: "add_608" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_609/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_609" + op: "Add" + input: "Sqrt_179" + input: "add_609/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_180" + op: "RealDiv" + input: "add_607" + input: "add_609" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_971" + op: "Mul" + input: "add" + input: "truediv_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_180" + op: "Sub" + input: "bert/encoder/layer_10/output/LayerNorm/beta/read" + input: "mul_971" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_736" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + input: "sub_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_737" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + input: "add_607" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_738" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + input: "add_608" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_972/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_972" + op: "Mul" + input: "Mul_972/x" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_973/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_973" + op: "Mul" + input: "Mul_973/x" + input: "clip_by_global_norm/clip_by_global_norm/_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_610" + op: "Add" + input: "Mul_972" + input: "Mul_973" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_974/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_974" + op: "Mul" + input: "Mul_974/x" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_180" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_975/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_975" + op: "Mul" + input: "Mul_975/x" + input: "Square_180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_611" + op: "Add" + input: "Mul_974" + input: "Mul_975" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_180" + op: "Sqrt" + input: "add_611" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_612/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_612" + op: "Add" + input: "Sqrt_180" + input: "add_612/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_181" + op: "RealDiv" + input: "add_610" + input: "add_612" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_976" + op: "Mul" + input: "add" + input: "truediv_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_181" + op: "Sub" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/read" + input: "mul_976" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_739" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + input: "sub_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_740" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + input: "add_610" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_741" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + input: "add_611" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_977/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_977" + op: "Mul" + input: "Mul_977/x" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_978/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_978" + op: "Mul" + input: "Mul_978/x" + input: "clip_by_global_norm/clip_by_global_norm/_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_613" + op: "Add" + input: "Mul_977" + input: "Mul_978" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_979/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_979" + op: "Mul" + input: "Mul_979/x" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_181" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_980/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_980" + op: "Mul" + input: "Mul_980/x" + input: "Square_181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_614" + op: "Add" + input: "Mul_979" + input: "Mul_980" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_181" + op: "Sqrt" + input: "add_614" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_615/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_615" + op: "Add" + input: "Sqrt_181" + input: "add_615/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_182" + op: "RealDiv" + input: "add_613" + input: "add_615" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_981/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_981" + op: "Mul" + input: "mul_981/x" + input: "bert/encoder/layer_11/attention/self/query/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_616" + op: "Add" + input: "truediv_182" + input: "mul_981" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_982" + op: "Mul" + input: "add" + input: "add_616" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_182" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/query/kernel/read" + input: "mul_982" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_742" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel" + input: "sub_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_743" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + input: "add_613" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_744" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + input: "add_614" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/query/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_983/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_983" + op: "Mul" + input: "Mul_983/x" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_984/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_984" + op: "Mul" + input: "Mul_984/x" + input: "clip_by_global_norm/clip_by_global_norm/_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_617" + op: "Add" + input: "Mul_983" + input: "Mul_984" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_985/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_985" + op: "Mul" + input: "Mul_985/x" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_182" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_986/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_986" + op: "Mul" + input: "Mul_986/x" + input: "Square_182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_618" + op: "Add" + input: "Mul_985" + input: "Mul_986" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_182" + op: "Sqrt" + input: "add_618" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_619/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_619" + op: "Add" + input: "Sqrt_182" + input: "add_619/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_183" + op: "RealDiv" + input: "add_617" + input: "add_619" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_987" + op: "Mul" + input: "add" + input: "truediv_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_183" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/query/bias/read" + input: "mul_987" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_745" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias" + input: "sub_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_746" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + input: "add_617" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_747" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + input: "add_618" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_988/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_988" + op: "Mul" + input: "Mul_988/x" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_989/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_989" + op: "Mul" + input: "Mul_989/x" + input: "clip_by_global_norm/clip_by_global_norm/_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_620" + op: "Add" + input: "Mul_988" + input: "Mul_989" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_990/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_990" + op: "Mul" + input: "Mul_990/x" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_183" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_991/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_991" + op: "Mul" + input: "Mul_991/x" + input: "Square_183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_621" + op: "Add" + input: "Mul_990" + input: "Mul_991" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_183" + op: "Sqrt" + input: "add_621" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_622/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_622" + op: "Add" + input: "Sqrt_183" + input: "add_622/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_184" + op: "RealDiv" + input: "add_620" + input: "add_622" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_992/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_992" + op: "Mul" + input: "mul_992/x" + input: "bert/encoder/layer_11/attention/self/key/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_623" + op: "Add" + input: "truediv_184" + input: "mul_992" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_993" + op: "Mul" + input: "add" + input: "add_623" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_184" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/key/kernel/read" + input: "mul_993" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_748" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel" + input: "sub_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_749" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + input: "add_620" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_750" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + input: "add_621" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/key/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_994/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_994" + op: "Mul" + input: "Mul_994/x" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_995/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_995" + op: "Mul" + input: "Mul_995/x" + input: "clip_by_global_norm/clip_by_global_norm/_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_624" + op: "Add" + input: "Mul_994" + input: "Mul_995" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_996/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_996" + op: "Mul" + input: "Mul_996/x" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_184" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_997/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_997" + op: "Mul" + input: "Mul_997/x" + input: "Square_184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_625" + op: "Add" + input: "Mul_996" + input: "Mul_997" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_184" + op: "Sqrt" + input: "add_625" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_626/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_626" + op: "Add" + input: "Sqrt_184" + input: "add_626/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_185" + op: "RealDiv" + input: "add_624" + input: "add_626" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_998" + op: "Mul" + input: "add" + input: "truediv_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_185" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/key/bias/read" + input: "mul_998" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_751" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias" + input: "sub_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_752" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + input: "add_624" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_753" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + input: "add_625" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_999/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_999" + op: "Mul" + input: "Mul_999/x" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1000/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1000" + op: "Mul" + input: "Mul_1000/x" + input: "clip_by_global_norm/clip_by_global_norm/_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_627" + op: "Add" + input: "Mul_999" + input: "Mul_1000" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1001/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1001" + op: "Mul" + input: "Mul_1001/x" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_185" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1002/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1002" + op: "Mul" + input: "Mul_1002/x" + input: "Square_185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_628" + op: "Add" + input: "Mul_1001" + input: "Mul_1002" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_185" + op: "Sqrt" + input: "add_628" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_629/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_629" + op: "Add" + input: "Sqrt_185" + input: "add_629/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_186" + op: "RealDiv" + input: "add_627" + input: "add_629" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1003/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1003" + op: "Mul" + input: "mul_1003/x" + input: "bert/encoder/layer_11/attention/self/value/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_630" + op: "Add" + input: "truediv_186" + input: "mul_1003" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1004" + op: "Mul" + input: "add" + input: "add_630" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_186" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/value/kernel/read" + input: "mul_1004" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_754" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel" + input: "sub_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_755" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + input: "add_627" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_756" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + input: "add_628" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/self/value/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1005/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1005" + op: "Mul" + input: "Mul_1005/x" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1006/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1006" + op: "Mul" + input: "Mul_1006/x" + input: "clip_by_global_norm/clip_by_global_norm/_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_631" + op: "Add" + input: "Mul_1005" + input: "Mul_1006" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1007/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1007" + op: "Mul" + input: "Mul_1007/x" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_186" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1008/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1008" + op: "Mul" + input: "Mul_1008/x" + input: "Square_186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_632" + op: "Add" + input: "Mul_1007" + input: "Mul_1008" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_186" + op: "Sqrt" + input: "add_632" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_633/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_633" + op: "Add" + input: "Sqrt_186" + input: "add_633/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_187" + op: "RealDiv" + input: "add_631" + input: "add_633" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1009" + op: "Mul" + input: "add" + input: "truediv_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_187" + op: "Sub" + input: "bert/encoder/layer_11/attention/self/value/bias/read" + input: "mul_1009" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_757" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias" + input: "sub_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_758" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + input: "add_631" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_759" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + input: "add_632" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1010/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1010" + op: "Mul" + input: "Mul_1010/x" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1011/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1011" + op: "Mul" + input: "Mul_1011/x" + input: "clip_by_global_norm/clip_by_global_norm/_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_634" + op: "Add" + input: "Mul_1010" + input: "Mul_1011" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1012/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1012" + op: "Mul" + input: "Mul_1012/x" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_187" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1013/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1013" + op: "Mul" + input: "Mul_1013/x" + input: "Square_187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_635" + op: "Add" + input: "Mul_1012" + input: "Mul_1013" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_187" + op: "Sqrt" + input: "add_635" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_636/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_636" + op: "Add" + input: "Sqrt_187" + input: "add_636/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_188" + op: "RealDiv" + input: "add_634" + input: "add_636" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1014/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1014" + op: "Mul" + input: "mul_1014/x" + input: "bert/encoder/layer_11/attention/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_637" + op: "Add" + input: "truediv_188" + input: "mul_1014" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1015" + op: "Mul" + input: "add" + input: "add_637" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_188" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/dense/kernel/read" + input: "mul_1015" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_760" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + input: "sub_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_761" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + input: "add_634" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_762" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + input: "add_635" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1016/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1016" + op: "Mul" + input: "Mul_1016/x" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1017/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1017" + op: "Mul" + input: "Mul_1017/x" + input: "clip_by_global_norm/clip_by_global_norm/_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_638" + op: "Add" + input: "Mul_1016" + input: "Mul_1017" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1018/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1018" + op: "Mul" + input: "Mul_1018/x" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_188" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1019/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1019" + op: "Mul" + input: "Mul_1019/x" + input: "Square_188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_639" + op: "Add" + input: "Mul_1018" + input: "Mul_1019" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_188" + op: "Sqrt" + input: "add_639" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_640/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_640" + op: "Add" + input: "Sqrt_188" + input: "add_640/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_189" + op: "RealDiv" + input: "add_638" + input: "add_640" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1020" + op: "Mul" + input: "add" + input: "truediv_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_189" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/dense/bias/read" + input: "mul_1020" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_763" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias" + input: "sub_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_764" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + input: "add_638" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_765" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + input: "add_639" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1021/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1021" + op: "Mul" + input: "Mul_1021/x" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1022/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1022" + op: "Mul" + input: "Mul_1022/x" + input: "clip_by_global_norm/clip_by_global_norm/_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_641" + op: "Add" + input: "Mul_1021" + input: "Mul_1022" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1023/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1023" + op: "Mul" + input: "Mul_1023/x" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_189" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1024/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1024" + op: "Mul" + input: "Mul_1024/x" + input: "Square_189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_642" + op: "Add" + input: "Mul_1023" + input: "Mul_1024" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_189" + op: "Sqrt" + input: "add_642" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_643/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_643" + op: "Add" + input: "Sqrt_189" + input: "add_643/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_190" + op: "RealDiv" + input: "add_641" + input: "add_643" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1025" + op: "Mul" + input: "add" + input: "truediv_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_190" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/read" + input: "mul_1025" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_766" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + input: "sub_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_767" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + input: "add_641" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_768" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + input: "add_642" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1026/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1026" + op: "Mul" + input: "Mul_1026/x" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1027/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1027" + op: "Mul" + input: "Mul_1027/x" + input: "clip_by_global_norm/clip_by_global_norm/_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_644" + op: "Add" + input: "Mul_1026" + input: "Mul_1027" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1028/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1028" + op: "Mul" + input: "Mul_1028/x" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_190" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1029/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1029" + op: "Mul" + input: "Mul_1029/x" + input: "Square_190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_645" + op: "Add" + input: "Mul_1028" + input: "Mul_1029" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_190" + op: "Sqrt" + input: "add_645" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_646/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_646" + op: "Add" + input: "Sqrt_190" + input: "add_646/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_191" + op: "RealDiv" + input: "add_644" + input: "add_646" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1030" + op: "Mul" + input: "add" + input: "truediv_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_191" + op: "Sub" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/read" + input: "mul_1030" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_769" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + input: "sub_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_770" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + input: "add_644" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_771" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + input: "add_645" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\014\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1031/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1031" + op: "Mul" + input: "Mul_1031/x" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1032/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1032" + op: "Mul" + input: "Mul_1032/x" + input: "clip_by_global_norm/clip_by_global_norm/_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_647" + op: "Add" + input: "Mul_1031" + input: "Mul_1032" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1033/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1033" + op: "Mul" + input: "Mul_1033/x" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_191" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1034/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1034" + op: "Mul" + input: "Mul_1034/x" + input: "Square_191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_648" + op: "Add" + input: "Mul_1033" + input: "Mul_1034" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_191" + op: "Sqrt" + input: "add_648" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_649/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_649" + op: "Add" + input: "Sqrt_191" + input: "add_649/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_192" + op: "RealDiv" + input: "add_647" + input: "add_649" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_1035/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1035" + op: "Mul" + input: "mul_1035/x" + input: "bert/encoder/layer_11/intermediate/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_650" + op: "Add" + input: "truediv_192" + input: "mul_1035" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_1036" + op: "Mul" + input: "add" + input: "add_650" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_192" + op: "Sub" + input: "bert/encoder/layer_11/intermediate/dense/kernel/read" + input: "mul_1036" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_772" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + input: "sub_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_773" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + input: "add_647" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_774" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + input: "add_648" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 3072 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1037/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1037" + op: "Mul" + input: "Mul_1037/x" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1038/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1038" + op: "Mul" + input: "Mul_1038/x" + input: "clip_by_global_norm/clip_by_global_norm/_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_651" + op: "Add" + input: "Mul_1037" + input: "Mul_1038" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1039/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1039" + op: "Mul" + input: "Mul_1039/x" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Square_192" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Mul_1040/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1040" + op: "Mul" + input: "Mul_1040/x" + input: "Square_192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_652" + op: "Add" + input: "Mul_1039" + input: "Mul_1040" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Sqrt_192" + op: "Sqrt" + input: "add_652" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "add_653/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_653" + op: "Add" + input: "Sqrt_192" + input: "add_653/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "truediv_193" + op: "RealDiv" + input: "add_651" + input: "add_653" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "mul_1041" + op: "Mul" + input: "add" + input: "truediv_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "sub_193" + op: "Sub" + input: "bert/encoder/layer_11/intermediate/dense/bias/read" + input: "mul_1041" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } +} +node { + name: "Assign_775" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias" + input: "sub_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_776" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + input: "add_651" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_777" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + input: "add_652" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\014\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1042/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1042" + op: "Mul" + input: "Mul_1042/x" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1043/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1043" + op: "Mul" + input: "Mul_1043/x" + input: "clip_by_global_norm/clip_by_global_norm/_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_654" + op: "Add" + input: "Mul_1042" + input: "Mul_1043" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1044/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1044" + op: "Mul" + input: "Mul_1044/x" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_193" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1045/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1045" + op: "Mul" + input: "Mul_1045/x" + input: "Square_193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_655" + op: "Add" + input: "Mul_1044" + input: "Mul_1045" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_193" + op: "Sqrt" + input: "add_655" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_656/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_656" + op: "Add" + input: "Sqrt_193" + input: "add_656/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_194" + op: "RealDiv" + input: "add_654" + input: "add_656" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1046/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1046" + op: "Mul" + input: "mul_1046/x" + input: "bert/encoder/layer_11/output/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_657" + op: "Add" + input: "truediv_194" + input: "mul_1046" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1047" + op: "Mul" + input: "add" + input: "add_657" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_194" + op: "Sub" + input: "bert/encoder/layer_11/output/dense/kernel/read" + input: "mul_1047" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_778" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel" + input: "sub_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_779" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + input: "add_654" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_780" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + input: "add_655" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + input: "bert/encoder/layer_11/output/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + input: "bert/encoder/layer_11/output/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/dense/bias/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1048/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1048" + op: "Mul" + input: "Mul_1048/x" + input: "bert/encoder/layer_11/output/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1049/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1049" + op: "Mul" + input: "Mul_1049/x" + input: "clip_by_global_norm/clip_by_global_norm/_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_658" + op: "Add" + input: "Mul_1048" + input: "Mul_1049" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1050/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1050" + op: "Mul" + input: "Mul_1050/x" + input: "bert/encoder/layer_11/output/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_194" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1051/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1051" + op: "Mul" + input: "Mul_1051/x" + input: "Square_194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_659" + op: "Add" + input: "Mul_1050" + input: "Mul_1051" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_194" + op: "Sqrt" + input: "add_659" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_660/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_660" + op: "Add" + input: "Sqrt_194" + input: "add_660/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_195" + op: "RealDiv" + input: "add_658" + input: "add_660" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1052" + op: "Mul" + input: "add" + input: "truediv_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_195" + op: "Sub" + input: "bert/encoder/layer_11/output/dense/bias/read" + input: "mul_1052" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_781" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias" + input: "sub_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_782" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + input: "add_658" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_783" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + input: "add_659" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1053/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1053" + op: "Mul" + input: "Mul_1053/x" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1054/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1054" + op: "Mul" + input: "Mul_1054/x" + input: "clip_by_global_norm/clip_by_global_norm/_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_661" + op: "Add" + input: "Mul_1053" + input: "Mul_1054" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1055/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1055" + op: "Mul" + input: "Mul_1055/x" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_195" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1056/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1056" + op: "Mul" + input: "Mul_1056/x" + input: "Square_195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_662" + op: "Add" + input: "Mul_1055" + input: "Mul_1056" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_195" + op: "Sqrt" + input: "add_662" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_663/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_663" + op: "Add" + input: "Sqrt_195" + input: "add_663/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_196" + op: "RealDiv" + input: "add_661" + input: "add_663" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1057" + op: "Mul" + input: "add" + input: "truediv_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_196" + op: "Sub" + input: "bert/encoder/layer_11/output/LayerNorm/beta/read" + input: "mul_1057" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_784" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + input: "sub_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_785" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + input: "add_661" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_786" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + input: "add_662" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/Assign" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/read" + op: "Identity" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1058/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1058" + op: "Mul" + input: "Mul_1058/x" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1059/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1059" + op: "Mul" + input: "Mul_1059/x" + input: "clip_by_global_norm/clip_by_global_norm/_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_664" + op: "Add" + input: "Mul_1058" + input: "Mul_1059" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1060/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1060" + op: "Mul" + input: "Mul_1060/x" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_196" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1061/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1061" + op: "Mul" + input: "Mul_1061/x" + input: "Square_196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_665" + op: "Add" + input: "Mul_1060" + input: "Mul_1061" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_196" + op: "Sqrt" + input: "add_665" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_666/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_666" + op: "Add" + input: "Sqrt_196" + input: "add_666/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_197" + op: "RealDiv" + input: "add_664" + input: "add_666" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1062" + op: "Mul" + input: "add" + input: "truediv_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_197" + op: "Sub" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/read" + input: "mul_1062" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_787" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + input: "sub_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_788" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + input: "add_664" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_789" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + input: "add_665" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m/Initializer/zeros" + op: "Fill" + input: "bert/pooler/dense/kernel/adam_m/Initializer/zeros/shape_as_tensor" + input: "bert/pooler/dense/kernel/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m/Assign" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_m" + input: "bert/pooler/dense/kernel/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_m/read" + op: "Identity" + input: "bert/pooler/dense/kernel/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\000\003\000\000\000\003\000\000" + } + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v/Initializer/zeros" + op: "Fill" + input: "bert/pooler/dense/kernel/adam_v/Initializer/zeros/shape_as_tensor" + input: "bert/pooler/dense/kernel/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v/Assign" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_v" + input: "bert/pooler/dense/kernel/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/kernel/adam_v/read" + op: "Identity" + input: "bert/pooler/dense/kernel/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1063/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1063" + op: "Mul" + input: "Mul_1063/x" + input: "bert/pooler/dense/kernel/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1064/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1064" + op: "Mul" + input: "Mul_1064/x" + input: "clip_by_global_norm/clip_by_global_norm/_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_667" + op: "Add" + input: "Mul_1063" + input: "Mul_1064" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1065/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1065" + op: "Mul" + input: "Mul_1065/x" + input: "bert/pooler/dense/kernel/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_197" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1066/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1066" + op: "Mul" + input: "Mul_1066/x" + input: "Square_197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_668" + op: "Add" + input: "Mul_1065" + input: "Mul_1066" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_197" + op: "Sqrt" + input: "add_668" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_669/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_669" + op: "Add" + input: "Sqrt_197" + input: "add_669/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_198" + op: "RealDiv" + input: "add_667" + input: "add_669" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1067/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1067" + op: "Mul" + input: "mul_1067/x" + input: "bert/pooler/dense/kernel/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_670" + op: "Add" + input: "truediv_198" + input: "mul_1067" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1068" + op: "Mul" + input: "add" + input: "add_670" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_198" + op: "Sub" + input: "bert/pooler/dense/kernel/read" + input: "mul_1068" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_790" + op: "Assign" + input: "bert/pooler/dense/kernel" + input: "sub_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_791" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_m" + input: "add_667" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_792" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_v" + input: "add_668" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/bias/adam_m/Assign" + op: "Assign" + input: "bert/pooler/dense/bias/adam_m" + input: "bert/pooler/dense/bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/bias/adam_m/read" + op: "Identity" + input: "bert/pooler/dense/bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "bert/pooler/dense/bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 768 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "bert/pooler/dense/bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "bert/pooler/dense/bias/adam_v/Assign" + op: "Assign" + input: "bert/pooler/dense/bias/adam_v" + input: "bert/pooler/dense/bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "bert/pooler/dense/bias/adam_v/read" + op: "Identity" + input: "bert/pooler/dense/bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1069/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1069" + op: "Mul" + input: "Mul_1069/x" + input: "bert/pooler/dense/bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1070/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1070" + op: "Mul" + input: "Mul_1070/x" + input: "clip_by_global_norm/clip_by_global_norm/_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_671" + op: "Add" + input: "Mul_1069" + input: "Mul_1070" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1071/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1071" + op: "Mul" + input: "Mul_1071/x" + input: "bert/pooler/dense/bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_198" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1072/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1072" + op: "Mul" + input: "Mul_1072/x" + input: "Square_198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_672" + op: "Add" + input: "Mul_1071" + input: "Mul_1072" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_198" + op: "Sqrt" + input: "add_672" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_673/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_673" + op: "Add" + input: "Sqrt_198" + input: "add_673/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_199" + op: "RealDiv" + input: "add_671" + input: "add_673" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1073" + op: "Mul" + input: "add" + input: "truediv_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_199" + op: "Sub" + input: "bert/pooler/dense/bias/read" + input: "mul_1073" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_793" + op: "Assign" + input: "bert/pooler/dense/bias" + input: "sub_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_794" + op: "Assign" + input: "bert/pooler/dense/bias/adam_m" + input: "add_671" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_795" + op: "Assign" + input: "bert/pooler/dense/bias/adam_v" + input: "add_672" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_weights/adam_m/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\003\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "output_weights/adam_m/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "output_weights/adam_m/Initializer/zeros" + op: "Fill" + input: "output_weights/adam_m/Initializer/zeros/shape_as_tensor" + input: "output_weights/adam_m/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "output_weights/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_weights/adam_m/Assign" + op: "Assign" + input: "output_weights/adam_m" + input: "output_weights/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_weights/adam_m/read" + op: "Identity" + input: "output_weights/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "output_weights/adam_v/Initializer/zeros/shape_as_tensor" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 2 + } + } + tensor_content: "\003\000\000\000\000\003\000\000" + } + } + } +} +node { + name: "output_weights/adam_v/Initializer/zeros/Const" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0 + } + } + } +} +node { + name: "output_weights/adam_v/Initializer/zeros" + op: "Fill" + input: "output_weights/adam_v/Initializer/zeros/shape_as_tensor" + input: "output_weights/adam_v/Initializer/zeros/Const" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "index_type" + value { + type: DT_INT32 + } + } +} +node { + name: "output_weights/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_weights/adam_v/Assign" + op: "Assign" + input: "output_weights/adam_v" + input: "output_weights/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_weights/adam_v/read" + op: "Identity" + input: "output_weights/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1074/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1074" + op: "Mul" + input: "Mul_1074/x" + input: "output_weights/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1075/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1075" + op: "Mul" + input: "Mul_1075/x" + input: "clip_by_global_norm/clip_by_global_norm/_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_674" + op: "Add" + input: "Mul_1074" + input: "Mul_1075" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1076/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1076" + op: "Mul" + input: "Mul_1076/x" + input: "output_weights/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Square_199" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Mul_1077/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1077" + op: "Mul" + input: "Mul_1077/x" + input: "Square_199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_675" + op: "Add" + input: "Mul_1076" + input: "Mul_1077" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Sqrt_199" + op: "Sqrt" + input: "add_675" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_676/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_676" + op: "Add" + input: "Sqrt_199" + input: "add_676/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "truediv_200" + op: "RealDiv" + input: "add_674" + input: "add_676" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1078/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.009999999776482582 + } + } + } +} +node { + name: "mul_1078" + op: "Mul" + input: "mul_1078/x" + input: "output_weights/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "add_677" + op: "Add" + input: "truediv_200" + input: "mul_1078" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "mul_1079" + op: "Mul" + input: "add" + input: "add_677" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "sub_200" + op: "Sub" + input: "output_weights/read" + input: "mul_1079" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } +} +node { + name: "Assign_796" + op: "Assign" + input: "output_weights" + input: "sub_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_797" + op: "Assign" + input: "output_weights/adam_m" + input: "add_674" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_798" + op: "Assign" + input: "output_weights/adam_v" + input: "add_675" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_bias/adam_m/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 3 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "output_bias/adam_m" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_bias/adam_m/Assign" + op: "Assign" + input: "output_bias/adam_m" + input: "output_bias/adam_m/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_bias/adam_m/read" + op: "Identity" + input: "output_bias/adam_m" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "output_bias/adam_v/Initializer/zeros" + op: "Const" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + dim { + size: 3 + } + } + float_val: 0.0 + } + } + } +} +node { + name: "output_bias/adam_v" + op: "VariableV2" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "container" + value { + s: "" + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "shape" + value { + shape { + dim { + size: 3 + } + } + } + } + attr { + key: "shared_name" + value { + s: "" + } + } +} +node { + name: "output_bias/adam_v/Assign" + op: "Assign" + input: "output_bias/adam_v" + input: "output_bias/adam_v/Initializer/zeros" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "output_bias/adam_v/read" + op: "Identity" + input: "output_bias/adam_v" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Mul_1080/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.8999999761581421 + } + } + } +} +node { + name: "Mul_1080" + op: "Mul" + input: "Mul_1080/x" + input: "output_bias/adam_m/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Mul_1081/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.10000000149011612 + } + } + } +} +node { + name: "Mul_1081" + op: "Mul" + input: "Mul_1081/x" + input: "clip_by_global_norm/clip_by_global_norm/_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "add_678" + op: "Add" + input: "Mul_1080" + input: "Mul_1081" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Mul_1082/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.9990000128746033 + } + } + } +} +node { + name: "Mul_1082" + op: "Mul" + input: "Mul_1082/x" + input: "output_bias/adam_v/read" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Square_200" + op: "Square" + input: "clip_by_global_norm/clip_by_global_norm/_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Mul_1083/x" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 0.0010000000474974513 + } + } + } +} +node { + name: "Mul_1083" + op: "Mul" + input: "Mul_1083/x" + input: "Square_200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "add_679" + op: "Add" + input: "Mul_1082" + input: "Mul_1083" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Sqrt_200" + op: "Sqrt" + input: "add_679" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "add_680/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_FLOAT + tensor_shape { + } + float_val: 9.999999974752427e-07 + } + } + } +} +node { + name: "add_680" + op: "Add" + input: "Sqrt_200" + input: "add_680/y" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "truediv_201" + op: "RealDiv" + input: "add_678" + input: "add_680" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "mul_1084" + op: "Mul" + input: "add" + input: "truediv_201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "sub_201" + op: "Sub" + input: "output_bias/read" + input: "mul_1084" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } +} +node { + name: "Assign_799" + op: "Assign" + input: "output_bias" + input: "sub_201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_800" + op: "Assign" + input: "output_bias/adam_m" + input: "add_678" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "Assign_801" + op: "Assign" + input: "output_bias/adam_v" + input: "add_679" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: false + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "group_deps" + op: "NoOp" + input: "^Assign_199" + input: "^Assign_200" + input: "^Assign_201" + input: "^Assign_202" + input: "^Assign_203" + input: "^Assign_204" + input: "^Assign_205" + input: "^Assign_206" + input: "^Assign_207" + input: "^Assign_208" + input: "^Assign_209" + input: "^Assign_210" + input: "^Assign_211" + input: "^Assign_212" + input: "^Assign_213" + input: "^Assign_214" + input: "^Assign_215" + input: "^Assign_216" + input: "^Assign_217" + input: "^Assign_218" + input: "^Assign_219" + input: "^Assign_220" + input: "^Assign_221" + input: "^Assign_222" + input: "^Assign_223" + input: "^Assign_224" + input: "^Assign_225" + input: "^Assign_226" + input: "^Assign_227" + input: "^Assign_228" + input: "^Assign_229" + input: "^Assign_230" + input: "^Assign_231" + input: "^Assign_232" + input: "^Assign_233" + input: "^Assign_234" + input: "^Assign_235" + input: "^Assign_236" + input: "^Assign_237" + input: "^Assign_238" + input: "^Assign_239" + input: "^Assign_240" + input: "^Assign_241" + input: "^Assign_242" + input: "^Assign_243" + input: "^Assign_244" + input: "^Assign_245" + input: "^Assign_246" + input: "^Assign_247" + input: "^Assign_248" + input: "^Assign_249" + input: "^Assign_250" + input: "^Assign_251" + input: "^Assign_252" + input: "^Assign_253" + input: "^Assign_254" + input: "^Assign_255" + input: "^Assign_256" + input: "^Assign_257" + input: "^Assign_258" + input: "^Assign_259" + input: "^Assign_260" + input: "^Assign_261" + input: "^Assign_262" + input: "^Assign_263" + input: "^Assign_264" + input: "^Assign_265" + input: "^Assign_266" + input: "^Assign_267" + input: "^Assign_268" + input: "^Assign_269" + input: "^Assign_270" + input: "^Assign_271" + input: "^Assign_272" + input: "^Assign_273" + input: "^Assign_274" + input: "^Assign_275" + input: "^Assign_276" + input: "^Assign_277" + input: "^Assign_278" + input: "^Assign_279" + input: "^Assign_280" + input: "^Assign_281" + input: "^Assign_282" + input: "^Assign_283" + input: "^Assign_284" + input: "^Assign_285" + input: "^Assign_286" + input: "^Assign_287" + input: "^Assign_288" + input: "^Assign_289" + input: "^Assign_290" + input: "^Assign_291" + input: "^Assign_292" + input: "^Assign_293" + input: "^Assign_294" + input: "^Assign_295" + input: "^Assign_296" + input: "^Assign_297" + input: "^Assign_298" + input: "^Assign_299" + input: "^Assign_300" + input: "^Assign_301" + input: "^Assign_302" + input: "^Assign_303" + input: "^Assign_304" + input: "^Assign_305" + input: "^Assign_306" + input: "^Assign_307" + input: "^Assign_308" + input: "^Assign_309" + input: "^Assign_310" + input: "^Assign_311" + input: "^Assign_312" + input: "^Assign_313" + input: "^Assign_314" + input: "^Assign_315" + input: "^Assign_316" + input: "^Assign_317" + input: "^Assign_318" + input: "^Assign_319" + input: "^Assign_320" + input: "^Assign_321" + input: "^Assign_322" + input: "^Assign_323" + input: "^Assign_324" + input: "^Assign_325" + input: "^Assign_326" + input: "^Assign_327" + input: "^Assign_328" + input: "^Assign_329" + input: "^Assign_330" + input: "^Assign_331" + input: "^Assign_332" + input: "^Assign_333" + input: "^Assign_334" + input: "^Assign_335" + input: "^Assign_336" + input: "^Assign_337" + input: "^Assign_338" + input: "^Assign_339" + input: "^Assign_340" + input: "^Assign_341" + input: "^Assign_342" + input: "^Assign_343" + input: "^Assign_344" + input: "^Assign_345" + input: "^Assign_346" + input: "^Assign_347" + input: "^Assign_348" + input: "^Assign_349" + input: "^Assign_350" + input: "^Assign_351" + input: "^Assign_352" + input: "^Assign_353" + input: "^Assign_354" + input: "^Assign_355" + input: "^Assign_356" + input: "^Assign_357" + input: "^Assign_358" + input: "^Assign_359" + input: "^Assign_360" + input: "^Assign_361" + input: "^Assign_362" + input: "^Assign_363" + input: "^Assign_364" + input: "^Assign_365" + input: "^Assign_366" + input: "^Assign_367" + input: "^Assign_368" + input: "^Assign_369" + input: "^Assign_370" + input: "^Assign_371" + input: "^Assign_372" + input: "^Assign_373" + input: "^Assign_374" + input: "^Assign_375" + input: "^Assign_376" + input: "^Assign_377" + input: "^Assign_378" + input: "^Assign_379" + input: "^Assign_380" + input: "^Assign_381" + input: "^Assign_382" + input: "^Assign_383" + input: "^Assign_384" + input: "^Assign_385" + input: "^Assign_386" + input: "^Assign_387" + input: "^Assign_388" + input: "^Assign_389" + input: "^Assign_390" + input: "^Assign_391" + input: "^Assign_392" + input: "^Assign_393" + input: "^Assign_394" + input: "^Assign_395" + input: "^Assign_396" + input: "^Assign_397" + input: "^Assign_398" + input: "^Assign_399" + input: "^Assign_400" + input: "^Assign_401" + input: "^Assign_402" + input: "^Assign_403" + input: "^Assign_404" + input: "^Assign_405" + input: "^Assign_406" + input: "^Assign_407" + input: "^Assign_408" + input: "^Assign_409" + input: "^Assign_410" + input: "^Assign_411" + input: "^Assign_412" + input: "^Assign_413" + input: "^Assign_414" + input: "^Assign_415" + input: "^Assign_416" + input: "^Assign_417" + input: "^Assign_418" + input: "^Assign_419" + input: "^Assign_420" + input: "^Assign_421" + input: "^Assign_422" + input: "^Assign_423" + input: "^Assign_424" + input: "^Assign_425" + input: "^Assign_426" + input: "^Assign_427" + input: "^Assign_428" + input: "^Assign_429" + input: "^Assign_430" + input: "^Assign_431" + input: "^Assign_432" + input: "^Assign_433" + input: "^Assign_434" + input: "^Assign_435" + input: "^Assign_436" + input: "^Assign_437" + input: "^Assign_438" + input: "^Assign_439" + input: "^Assign_440" + input: "^Assign_441" + input: "^Assign_442" + input: "^Assign_443" + input: "^Assign_444" + input: "^Assign_445" + input: "^Assign_446" + input: "^Assign_447" + input: "^Assign_448" + input: "^Assign_449" + input: "^Assign_450" + input: "^Assign_451" + input: "^Assign_452" + input: "^Assign_453" + input: "^Assign_454" + input: "^Assign_455" + input: "^Assign_456" + input: "^Assign_457" + input: "^Assign_458" + input: "^Assign_459" + input: "^Assign_460" + input: "^Assign_461" + input: "^Assign_462" + input: "^Assign_463" + input: "^Assign_464" + input: "^Assign_465" + input: "^Assign_466" + input: "^Assign_467" + input: "^Assign_468" + input: "^Assign_469" + input: "^Assign_470" + input: "^Assign_471" + input: "^Assign_472" + input: "^Assign_473" + input: "^Assign_474" + input: "^Assign_475" + input: "^Assign_476" + input: "^Assign_477" + input: "^Assign_478" + input: "^Assign_479" + input: "^Assign_480" + input: "^Assign_481" + input: "^Assign_482" + input: "^Assign_483" + input: "^Assign_484" + input: "^Assign_485" + input: "^Assign_486" + input: "^Assign_487" + input: "^Assign_488" + input: "^Assign_489" + input: "^Assign_490" + input: "^Assign_491" + input: "^Assign_492" + input: "^Assign_493" + input: "^Assign_494" + input: "^Assign_495" + input: "^Assign_496" + input: "^Assign_497" + input: "^Assign_498" + input: "^Assign_499" + input: "^Assign_500" + input: "^Assign_501" + input: "^Assign_502" + input: "^Assign_503" + input: "^Assign_504" + input: "^Assign_505" + input: "^Assign_506" + input: "^Assign_507" + input: "^Assign_508" + input: "^Assign_509" + input: "^Assign_510" + input: "^Assign_511" + input: "^Assign_512" + input: "^Assign_513" + input: "^Assign_514" + input: "^Assign_515" + input: "^Assign_516" + input: "^Assign_517" + input: "^Assign_518" + input: "^Assign_519" + input: "^Assign_520" + input: "^Assign_521" + input: "^Assign_522" + input: "^Assign_523" + input: "^Assign_524" + input: "^Assign_525" + input: "^Assign_526" + input: "^Assign_527" + input: "^Assign_528" + input: "^Assign_529" + input: "^Assign_530" + input: "^Assign_531" + input: "^Assign_532" + input: "^Assign_533" + input: "^Assign_534" + input: "^Assign_535" + input: "^Assign_536" + input: "^Assign_537" + input: "^Assign_538" + input: "^Assign_539" + input: "^Assign_540" + input: "^Assign_541" + input: "^Assign_542" + input: "^Assign_543" + input: "^Assign_544" + input: "^Assign_545" + input: "^Assign_546" + input: "^Assign_547" + input: "^Assign_548" + input: "^Assign_549" + input: "^Assign_550" + input: "^Assign_551" + input: "^Assign_552" + input: "^Assign_553" + input: "^Assign_554" + input: "^Assign_555" + input: "^Assign_556" + input: "^Assign_557" + input: "^Assign_558" + input: "^Assign_559" + input: "^Assign_560" + input: "^Assign_561" + input: "^Assign_562" + input: "^Assign_563" + input: "^Assign_564" + input: "^Assign_565" + input: "^Assign_566" + input: "^Assign_567" + input: "^Assign_568" + input: "^Assign_569" + input: "^Assign_570" + input: "^Assign_571" + input: "^Assign_572" + input: "^Assign_573" + input: "^Assign_574" + input: "^Assign_575" + input: "^Assign_576" + input: "^Assign_577" + input: "^Assign_578" + input: "^Assign_579" + input: "^Assign_580" + input: "^Assign_581" + input: "^Assign_582" + input: "^Assign_583" + input: "^Assign_584" + input: "^Assign_585" + input: "^Assign_586" + input: "^Assign_587" + input: "^Assign_588" + input: "^Assign_589" + input: "^Assign_590" + input: "^Assign_591" + input: "^Assign_592" + input: "^Assign_593" + input: "^Assign_594" + input: "^Assign_595" + input: "^Assign_596" + input: "^Assign_597" + input: "^Assign_598" + input: "^Assign_599" + input: "^Assign_600" + input: "^Assign_601" + input: "^Assign_602" + input: "^Assign_603" + input: "^Assign_604" + input: "^Assign_605" + input: "^Assign_606" + input: "^Assign_607" + input: "^Assign_608" + input: "^Assign_609" + input: "^Assign_610" + input: "^Assign_611" + input: "^Assign_612" + input: "^Assign_613" + input: "^Assign_614" + input: "^Assign_615" + input: "^Assign_616" + input: "^Assign_617" + input: "^Assign_618" + input: "^Assign_619" + input: "^Assign_620" + input: "^Assign_621" + input: "^Assign_622" + input: "^Assign_623" + input: "^Assign_624" + input: "^Assign_625" + input: "^Assign_626" + input: "^Assign_627" + input: "^Assign_628" + input: "^Assign_629" + input: "^Assign_630" + input: "^Assign_631" + input: "^Assign_632" + input: "^Assign_633" + input: "^Assign_634" + input: "^Assign_635" + input: "^Assign_636" + input: "^Assign_637" + input: "^Assign_638" + input: "^Assign_639" + input: "^Assign_640" + input: "^Assign_641" + input: "^Assign_642" + input: "^Assign_643" + input: "^Assign_644" + input: "^Assign_645" + input: "^Assign_646" + input: "^Assign_647" + input: "^Assign_648" + input: "^Assign_649" + input: "^Assign_650" + input: "^Assign_651" + input: "^Assign_652" + input: "^Assign_653" + input: "^Assign_654" + input: "^Assign_655" + input: "^Assign_656" + input: "^Assign_657" + input: "^Assign_658" + input: "^Assign_659" + input: "^Assign_660" + input: "^Assign_661" + input: "^Assign_662" + input: "^Assign_663" + input: "^Assign_664" + input: "^Assign_665" + input: "^Assign_666" + input: "^Assign_667" + input: "^Assign_668" + input: "^Assign_669" + input: "^Assign_670" + input: "^Assign_671" + input: "^Assign_672" + input: "^Assign_673" + input: "^Assign_674" + input: "^Assign_675" + input: "^Assign_676" + input: "^Assign_677" + input: "^Assign_678" + input: "^Assign_679" + input: "^Assign_680" + input: "^Assign_681" + input: "^Assign_682" + input: "^Assign_683" + input: "^Assign_684" + input: "^Assign_685" + input: "^Assign_686" + input: "^Assign_687" + input: "^Assign_688" + input: "^Assign_689" + input: "^Assign_690" + input: "^Assign_691" + input: "^Assign_692" + input: "^Assign_693" + input: "^Assign_694" + input: "^Assign_695" + input: "^Assign_696" + input: "^Assign_697" + input: "^Assign_698" + input: "^Assign_699" + input: "^Assign_700" + input: "^Assign_701" + input: "^Assign_702" + input: "^Assign_703" + input: "^Assign_704" + input: "^Assign_705" + input: "^Assign_706" + input: "^Assign_707" + input: "^Assign_708" + input: "^Assign_709" + input: "^Assign_710" + input: "^Assign_711" + input: "^Assign_712" + input: "^Assign_713" + input: "^Assign_714" + input: "^Assign_715" + input: "^Assign_716" + input: "^Assign_717" + input: "^Assign_718" + input: "^Assign_719" + input: "^Assign_720" + input: "^Assign_721" + input: "^Assign_722" + input: "^Assign_723" + input: "^Assign_724" + input: "^Assign_725" + input: "^Assign_726" + input: "^Assign_727" + input: "^Assign_728" + input: "^Assign_729" + input: "^Assign_730" + input: "^Assign_731" + input: "^Assign_732" + input: "^Assign_733" + input: "^Assign_734" + input: "^Assign_735" + input: "^Assign_736" + input: "^Assign_737" + input: "^Assign_738" + input: "^Assign_739" + input: "^Assign_740" + input: "^Assign_741" + input: "^Assign_742" + input: "^Assign_743" + input: "^Assign_744" + input: "^Assign_745" + input: "^Assign_746" + input: "^Assign_747" + input: "^Assign_748" + input: "^Assign_749" + input: "^Assign_750" + input: "^Assign_751" + input: "^Assign_752" + input: "^Assign_753" + input: "^Assign_754" + input: "^Assign_755" + input: "^Assign_756" + input: "^Assign_757" + input: "^Assign_758" + input: "^Assign_759" + input: "^Assign_760" + input: "^Assign_761" + input: "^Assign_762" + input: "^Assign_763" + input: "^Assign_764" + input: "^Assign_765" + input: "^Assign_766" + input: "^Assign_767" + input: "^Assign_768" + input: "^Assign_769" + input: "^Assign_770" + input: "^Assign_771" + input: "^Assign_772" + input: "^Assign_773" + input: "^Assign_774" + input: "^Assign_775" + input: "^Assign_776" + input: "^Assign_777" + input: "^Assign_778" + input: "^Assign_779" + input: "^Assign_780" + input: "^Assign_781" + input: "^Assign_782" + input: "^Assign_783" + input: "^Assign_784" + input: "^Assign_785" + input: "^Assign_786" + input: "^Assign_787" + input: "^Assign_788" + input: "^Assign_789" + input: "^Assign_790" + input: "^Assign_791" + input: "^Assign_792" + input: "^Assign_793" + input: "^Assign_794" + input: "^Assign_795" + input: "^Assign_796" + input: "^Assign_797" + input: "^Assign_798" + input: "^Assign_799" + input: "^Assign_800" + input: "^Assign_801" +} +node { + name: "ReadVariableOp" + op: "ReadVariableOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "add_681/y" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 1 + } + } + } +} +node { + name: "add_681" + op: "Add" + input: "ReadVariableOp" + input: "add_681/y" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "AssignVariableOp" + op: "AssignVariableOp" + input: "global_step" + input: "add_681" + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "ReadVariableOp_1" + op: "ReadVariableOp" + input: "global_step" + input: "^AssignVariableOp" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "group_deps_1" + op: "NoOp" + input: "^AssignVariableOp" + input: "^group_deps" +} +node { + name: "loss_1/tags" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "loss_1" + } + } + } +} +node { + name: "loss_1" + op: "ScalarSummary" + input: "loss_1/tags" + input: "loss/Mean" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "init" + op: "NoOp" + input: "^Assign" + input: "^Assign_1" + input: "^Assign_10" + input: "^Assign_100" + input: "^Assign_101" + input: "^Assign_102" + input: "^Assign_103" + input: "^Assign_104" + input: "^Assign_105" + input: "^Assign_106" + input: "^Assign_107" + input: "^Assign_108" + input: "^Assign_109" + input: "^Assign_11" + input: "^Assign_110" + input: "^Assign_111" + input: "^Assign_112" + input: "^Assign_113" + input: "^Assign_114" + input: "^Assign_115" + input: "^Assign_116" + input: "^Assign_117" + input: "^Assign_118" + input: "^Assign_119" + input: "^Assign_12" + input: "^Assign_120" + input: "^Assign_121" + input: "^Assign_122" + input: "^Assign_123" + input: "^Assign_124" + input: "^Assign_125" + input: "^Assign_126" + input: "^Assign_127" + input: "^Assign_128" + input: "^Assign_129" + input: "^Assign_13" + input: "^Assign_130" + input: "^Assign_131" + input: "^Assign_132" + input: "^Assign_133" + input: "^Assign_134" + input: "^Assign_135" + input: "^Assign_136" + input: "^Assign_137" + input: "^Assign_138" + input: "^Assign_139" + input: "^Assign_14" + input: "^Assign_140" + input: "^Assign_141" + input: "^Assign_142" + input: "^Assign_143" + input: "^Assign_144" + input: "^Assign_145" + input: "^Assign_146" + input: "^Assign_147" + input: "^Assign_148" + input: "^Assign_149" + input: "^Assign_15" + input: "^Assign_150" + input: "^Assign_151" + input: "^Assign_152" + input: "^Assign_153" + input: "^Assign_154" + input: "^Assign_155" + input: "^Assign_156" + input: "^Assign_157" + input: "^Assign_158" + input: "^Assign_159" + input: "^Assign_16" + input: "^Assign_160" + input: "^Assign_161" + input: "^Assign_162" + input: "^Assign_163" + input: "^Assign_164" + input: "^Assign_165" + input: "^Assign_166" + input: "^Assign_167" + input: "^Assign_168" + input: "^Assign_169" + input: "^Assign_17" + input: "^Assign_170" + input: "^Assign_171" + input: "^Assign_172" + input: "^Assign_173" + input: "^Assign_174" + input: "^Assign_175" + input: "^Assign_176" + input: "^Assign_177" + input: "^Assign_178" + input: "^Assign_179" + input: "^Assign_18" + input: "^Assign_180" + input: "^Assign_181" + input: "^Assign_182" + input: "^Assign_183" + input: "^Assign_184" + input: "^Assign_185" + input: "^Assign_186" + input: "^Assign_187" + input: "^Assign_188" + input: "^Assign_189" + input: "^Assign_19" + input: "^Assign_190" + input: "^Assign_191" + input: "^Assign_192" + input: "^Assign_193" + input: "^Assign_194" + input: "^Assign_195" + input: "^Assign_196" + input: "^Assign_197" + input: "^Assign_198" + input: "^Assign_2" + input: "^Assign_20" + input: "^Assign_21" + input: "^Assign_22" + input: "^Assign_23" + input: "^Assign_24" + input: "^Assign_25" + input: "^Assign_26" + input: "^Assign_27" + input: "^Assign_28" + input: "^Assign_29" + input: "^Assign_3" + input: "^Assign_30" + input: "^Assign_31" + input: "^Assign_32" + input: "^Assign_33" + input: "^Assign_34" + input: "^Assign_35" + input: "^Assign_36" + input: "^Assign_37" + input: "^Assign_38" + input: "^Assign_39" + input: "^Assign_4" + input: "^Assign_40" + input: "^Assign_41" + input: "^Assign_42" + input: "^Assign_43" + input: "^Assign_44" + input: "^Assign_45" + input: "^Assign_46" + input: "^Assign_47" + input: "^Assign_48" + input: "^Assign_49" + input: "^Assign_5" + input: "^Assign_50" + input: "^Assign_51" + input: "^Assign_52" + input: "^Assign_53" + input: "^Assign_54" + input: "^Assign_55" + input: "^Assign_56" + input: "^Assign_57" + input: "^Assign_58" + input: "^Assign_59" + input: "^Assign_6" + input: "^Assign_60" + input: "^Assign_61" + input: "^Assign_62" + input: "^Assign_63" + input: "^Assign_64" + input: "^Assign_65" + input: "^Assign_66" + input: "^Assign_67" + input: "^Assign_68" + input: "^Assign_69" + input: "^Assign_7" + input: "^Assign_70" + input: "^Assign_71" + input: "^Assign_72" + input: "^Assign_73" + input: "^Assign_74" + input: "^Assign_75" + input: "^Assign_76" + input: "^Assign_77" + input: "^Assign_78" + input: "^Assign_79" + input: "^Assign_8" + input: "^Assign_80" + input: "^Assign_81" + input: "^Assign_82" + input: "^Assign_83" + input: "^Assign_84" + input: "^Assign_85" + input: "^Assign_86" + input: "^Assign_87" + input: "^Assign_88" + input: "^Assign_89" + input: "^Assign_9" + input: "^Assign_90" + input: "^Assign_91" + input: "^Assign_92" + input: "^Assign_93" + input: "^Assign_94" + input: "^Assign_95" + input: "^Assign_96" + input: "^Assign_97" + input: "^Assign_98" + input: "^Assign_99" + input: "^bert/embeddings/LayerNorm/beta/adam_m/Assign" + input: "^bert/embeddings/LayerNorm/beta/adam_v/Assign" + input: "^bert/embeddings/LayerNorm/gamma/adam_m/Assign" + input: "^bert/embeddings/LayerNorm/gamma/adam_v/Assign" + input: "^bert/embeddings/position_embeddings/adam_m/Assign" + input: "^bert/embeddings/position_embeddings/adam_v/Assign" + input: "^bert/embeddings/token_type_embeddings/adam_m/Assign" + input: "^bert/embeddings/token_type_embeddings/adam_v/Assign" + input: "^bert/embeddings/word_embeddings/adam_m/Assign" + input: "^bert/embeddings/word_embeddings/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_0/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_0/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_0/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_0/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_0/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_0/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_0/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_0/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_0/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_1/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_1/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_1/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_1/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_1/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_1/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_1/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_1/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_10/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_10/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_10/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_10/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_10/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_10/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_10/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_10/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_11/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_11/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_11/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_11/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_11/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_11/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_11/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_11/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_2/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_2/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_2/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_2/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_2/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_2/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_2/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_2/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_3/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_3/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_3/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_3/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_3/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_3/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_3/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_3/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_4/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_4/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_4/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_4/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_4/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_4/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_4/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_4/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_5/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_5/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_5/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_5/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_5/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_5/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_5/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_5/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_6/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_6/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_6/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_6/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_6/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_6/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_6/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_6/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_7/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_7/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_7/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_7/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_7/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_7/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_7/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_7/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_8/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_8/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_8/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_8/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_8/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_8/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_8/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_8/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/output/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/key/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/key/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/key/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/key/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/query/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/query/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/query/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/query/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/value/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/value/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/attention/self/value/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/attention/self/value/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/intermediate/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/intermediate/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/intermediate/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/intermediate/dense/kernel/adam_v/Assign" + input: "^bert/encoder/layer_9/output/LayerNorm/beta/adam_m/Assign" + input: "^bert/encoder/layer_9/output/LayerNorm/beta/adam_v/Assign" + input: "^bert/encoder/layer_9/output/LayerNorm/gamma/adam_m/Assign" + input: "^bert/encoder/layer_9/output/LayerNorm/gamma/adam_v/Assign" + input: "^bert/encoder/layer_9/output/dense/bias/adam_m/Assign" + input: "^bert/encoder/layer_9/output/dense/bias/adam_v/Assign" + input: "^bert/encoder/layer_9/output/dense/kernel/adam_m/Assign" + input: "^bert/encoder/layer_9/output/dense/kernel/adam_v/Assign" + input: "^bert/pooler/dense/bias/adam_m/Assign" + input: "^bert/pooler/dense/bias/adam_v/Assign" + input: "^bert/pooler/dense/kernel/adam_m/Assign" + input: "^bert/pooler/dense/kernel/adam_v/Assign" + input: "^global_step/Assign" + input: "^output_bias/Assign" + input: "^output_bias/adam_m/Assign" + input: "^output_bias/adam_v/Assign" + input: "^output_weights/Assign" + input: "^output_weights/adam_m/Assign" + input: "^output_weights/adam_v/Assign" +} +node { + name: "init_1" + op: "NoOp" +} +node { + name: "group_deps_2" + op: "NoOp" + input: "^init" + input: "^init_1" +} +node { + name: "report_uninitialized_variables/VarIsInitializedOp" + op: "VarIsInitializedOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_1" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_2" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_3" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_4" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_5" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_6" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_7" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_8" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_9" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_10" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_11" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_12" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_13" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_14" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_15" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_16" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_17" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_18" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_19" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_20" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_21" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_22" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_23" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_24" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_25" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_26" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_27" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_28" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_29" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_30" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_31" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_32" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_33" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_34" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_35" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_36" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_37" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_38" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_39" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_40" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_41" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_42" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_43" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_44" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_45" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_46" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_47" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_48" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_49" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_50" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_51" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_52" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_53" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_54" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_55" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_56" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_57" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_58" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_59" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_60" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_61" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_62" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_63" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_64" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_65" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_66" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_67" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_68" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_69" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_70" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_71" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_72" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_73" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_74" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_75" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_76" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_77" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_78" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_79" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_80" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_81" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_82" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_83" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_84" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_85" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_86" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_87" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_88" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_89" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_90" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_91" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_92" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_93" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_94" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_95" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_96" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_97" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_98" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_99" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_100" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_101" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_102" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_103" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_104" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_105" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_106" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_107" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_108" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_109" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_110" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_111" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_112" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_113" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_114" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_115" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_116" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_117" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_118" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_119" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_120" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_121" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_122" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_123" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_124" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_125" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_126" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_127" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_128" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_129" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_130" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_131" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_132" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_133" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_134" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_135" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_136" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_137" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_138" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_139" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_140" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_141" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_142" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_143" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_144" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_145" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_146" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_147" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_148" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_149" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_150" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_151" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_152" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_153" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_154" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_155" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_156" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_157" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_158" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_159" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_160" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_161" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_162" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_163" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_164" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_165" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_166" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_167" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_168" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_169" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_170" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_171" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_172" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_173" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_174" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_175" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_176" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_177" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_178" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_179" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_180" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_181" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_182" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_183" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_184" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_185" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_186" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_187" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_188" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_189" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_190" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_191" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_192" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_193" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_194" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_195" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_196" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_197" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_198" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_199" + op: "IsVariableInitialized" + input: "output_weights" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_200" + op: "IsVariableInitialized" + input: "output_bias" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_201" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_202" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_203" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_204" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_205" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_206" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_207" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_208" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_209" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_210" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_211" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_212" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_213" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_214" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_215" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_216" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_217" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_218" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_219" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_220" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_221" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_222" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_223" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_224" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_225" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_226" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_227" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_228" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_229" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_230" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_231" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_232" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_233" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_234" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_235" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_236" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_237" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_238" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_239" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_240" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_241" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_242" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_243" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_244" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_245" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_246" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_247" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_248" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_249" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_250" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_251" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_252" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_253" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_254" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_255" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_256" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_257" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_258" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_259" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_260" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_261" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_262" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_263" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_264" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_265" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_266" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_267" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_268" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_269" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_270" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_271" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_272" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_273" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_274" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_275" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_276" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_277" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_278" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_279" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_280" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_281" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_282" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_283" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_284" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_285" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_286" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_287" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_288" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_289" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_290" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_291" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_292" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_293" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_294" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_295" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_296" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_297" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_298" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_299" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_300" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_301" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_302" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_303" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_304" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_305" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_306" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_307" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_308" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_309" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_310" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_311" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_312" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_313" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_314" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_315" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_316" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_317" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_318" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_319" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_320" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_321" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_322" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_323" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_324" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_325" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_326" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_327" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_328" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_329" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_330" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_331" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_332" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_333" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_334" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_335" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_336" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_337" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_338" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_339" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_340" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_341" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_342" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_343" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_344" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_345" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_346" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_347" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_348" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_349" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_350" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_351" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_352" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_353" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_354" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_355" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_356" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_357" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_358" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_359" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_360" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_361" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_362" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_363" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_364" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_365" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_366" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_367" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_368" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_369" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_370" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_371" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_372" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_373" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_374" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_375" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_376" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_377" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_378" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_379" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_380" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_381" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_382" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_383" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_384" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_385" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_386" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_387" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_388" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_389" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_390" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_391" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_392" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_393" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_394" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_395" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_396" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_397" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_398" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_399" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_400" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_401" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_402" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_403" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_404" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_405" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_406" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_407" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_408" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_409" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_410" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_411" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_412" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_413" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_414" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_415" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_416" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_417" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_418" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_419" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_420" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_421" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_422" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_423" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_424" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_425" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_426" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_427" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_428" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_429" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_430" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_431" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_432" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_433" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_434" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_435" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_436" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_437" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_438" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_439" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_440" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_441" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_442" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_443" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_444" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_445" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_446" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_447" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_448" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_449" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_450" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_451" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_452" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_453" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_454" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_455" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_456" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_457" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_458" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_459" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_460" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_461" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_462" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_463" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_464" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_465" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_466" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_467" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_468" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_469" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_470" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_471" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_472" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_473" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_474" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_475" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_476" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_477" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_478" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_479" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_480" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_481" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_482" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_483" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_484" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_485" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_486" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_487" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_488" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_489" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_490" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_491" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_492" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_493" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_494" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_495" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_496" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_497" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_498" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_499" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_500" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_501" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_502" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_503" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_504" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_505" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_506" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_507" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_508" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_509" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_510" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_511" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_512" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_513" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_514" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_515" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_516" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_517" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_518" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_519" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_520" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_521" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_522" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_523" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_524" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_525" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_526" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_527" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_528" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_529" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_530" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_531" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_532" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_533" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_534" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_535" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_536" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_537" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_538" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_539" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_540" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_541" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_542" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_543" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_544" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_545" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_546" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_547" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_548" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_549" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_550" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_551" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_552" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_553" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_554" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_555" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_556" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_557" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_558" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_559" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_560" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_561" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_562" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_563" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_564" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_565" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_566" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_567" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_568" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_569" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_570" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_571" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_572" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_573" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_574" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_575" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_576" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_577" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_578" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_579" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_580" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_581" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_582" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_583" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_584" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_585" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_586" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_587" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_588" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_589" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_590" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_591" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_592" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_593" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_594" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_595" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_596" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_597" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_598" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_599" + op: "IsVariableInitialized" + input: "output_weights/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_600" + op: "IsVariableInitialized" + input: "output_weights/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_601" + op: "IsVariableInitialized" + input: "output_bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/IsVariableInitialized_602" + op: "IsVariableInitialized" + input: "output_bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables/stack" + op: "Pack" + input: "report_uninitialized_variables/VarIsInitializedOp" + input: "report_uninitialized_variables/IsVariableInitialized" + input: "report_uninitialized_variables/IsVariableInitialized_1" + input: "report_uninitialized_variables/IsVariableInitialized_2" + input: "report_uninitialized_variables/IsVariableInitialized_3" + input: "report_uninitialized_variables/IsVariableInitialized_4" + input: "report_uninitialized_variables/IsVariableInitialized_5" + input: "report_uninitialized_variables/IsVariableInitialized_6" + input: "report_uninitialized_variables/IsVariableInitialized_7" + input: "report_uninitialized_variables/IsVariableInitialized_8" + input: "report_uninitialized_variables/IsVariableInitialized_9" + input: "report_uninitialized_variables/IsVariableInitialized_10" + input: "report_uninitialized_variables/IsVariableInitialized_11" + input: "report_uninitialized_variables/IsVariableInitialized_12" + input: "report_uninitialized_variables/IsVariableInitialized_13" + input: "report_uninitialized_variables/IsVariableInitialized_14" + input: "report_uninitialized_variables/IsVariableInitialized_15" + input: "report_uninitialized_variables/IsVariableInitialized_16" + input: "report_uninitialized_variables/IsVariableInitialized_17" + input: "report_uninitialized_variables/IsVariableInitialized_18" + input: "report_uninitialized_variables/IsVariableInitialized_19" + input: "report_uninitialized_variables/IsVariableInitialized_20" + input: "report_uninitialized_variables/IsVariableInitialized_21" + input: "report_uninitialized_variables/IsVariableInitialized_22" + input: "report_uninitialized_variables/IsVariableInitialized_23" + input: "report_uninitialized_variables/IsVariableInitialized_24" + input: "report_uninitialized_variables/IsVariableInitialized_25" + input: "report_uninitialized_variables/IsVariableInitialized_26" + input: "report_uninitialized_variables/IsVariableInitialized_27" + input: "report_uninitialized_variables/IsVariableInitialized_28" + input: "report_uninitialized_variables/IsVariableInitialized_29" + input: "report_uninitialized_variables/IsVariableInitialized_30" + input: "report_uninitialized_variables/IsVariableInitialized_31" + input: "report_uninitialized_variables/IsVariableInitialized_32" + input: "report_uninitialized_variables/IsVariableInitialized_33" + input: "report_uninitialized_variables/IsVariableInitialized_34" + input: "report_uninitialized_variables/IsVariableInitialized_35" + input: "report_uninitialized_variables/IsVariableInitialized_36" + input: "report_uninitialized_variables/IsVariableInitialized_37" + input: "report_uninitialized_variables/IsVariableInitialized_38" + input: "report_uninitialized_variables/IsVariableInitialized_39" + input: "report_uninitialized_variables/IsVariableInitialized_40" + input: "report_uninitialized_variables/IsVariableInitialized_41" + input: "report_uninitialized_variables/IsVariableInitialized_42" + input: "report_uninitialized_variables/IsVariableInitialized_43" + input: "report_uninitialized_variables/IsVariableInitialized_44" + input: "report_uninitialized_variables/IsVariableInitialized_45" + input: "report_uninitialized_variables/IsVariableInitialized_46" + input: "report_uninitialized_variables/IsVariableInitialized_47" + input: "report_uninitialized_variables/IsVariableInitialized_48" + input: "report_uninitialized_variables/IsVariableInitialized_49" + input: "report_uninitialized_variables/IsVariableInitialized_50" + input: "report_uninitialized_variables/IsVariableInitialized_51" + input: "report_uninitialized_variables/IsVariableInitialized_52" + input: "report_uninitialized_variables/IsVariableInitialized_53" + input: "report_uninitialized_variables/IsVariableInitialized_54" + input: "report_uninitialized_variables/IsVariableInitialized_55" + input: "report_uninitialized_variables/IsVariableInitialized_56" + input: "report_uninitialized_variables/IsVariableInitialized_57" + input: "report_uninitialized_variables/IsVariableInitialized_58" + input: "report_uninitialized_variables/IsVariableInitialized_59" + input: "report_uninitialized_variables/IsVariableInitialized_60" + input: "report_uninitialized_variables/IsVariableInitialized_61" + input: "report_uninitialized_variables/IsVariableInitialized_62" + input: "report_uninitialized_variables/IsVariableInitialized_63" + input: "report_uninitialized_variables/IsVariableInitialized_64" + input: "report_uninitialized_variables/IsVariableInitialized_65" + input: "report_uninitialized_variables/IsVariableInitialized_66" + input: "report_uninitialized_variables/IsVariableInitialized_67" + input: "report_uninitialized_variables/IsVariableInitialized_68" + input: "report_uninitialized_variables/IsVariableInitialized_69" + input: "report_uninitialized_variables/IsVariableInitialized_70" + input: "report_uninitialized_variables/IsVariableInitialized_71" + input: "report_uninitialized_variables/IsVariableInitialized_72" + input: "report_uninitialized_variables/IsVariableInitialized_73" + input: "report_uninitialized_variables/IsVariableInitialized_74" + input: "report_uninitialized_variables/IsVariableInitialized_75" + input: "report_uninitialized_variables/IsVariableInitialized_76" + input: "report_uninitialized_variables/IsVariableInitialized_77" + input: "report_uninitialized_variables/IsVariableInitialized_78" + input: "report_uninitialized_variables/IsVariableInitialized_79" + input: "report_uninitialized_variables/IsVariableInitialized_80" + input: "report_uninitialized_variables/IsVariableInitialized_81" + input: "report_uninitialized_variables/IsVariableInitialized_82" + input: "report_uninitialized_variables/IsVariableInitialized_83" + input: "report_uninitialized_variables/IsVariableInitialized_84" + input: "report_uninitialized_variables/IsVariableInitialized_85" + input: "report_uninitialized_variables/IsVariableInitialized_86" + input: "report_uninitialized_variables/IsVariableInitialized_87" + input: "report_uninitialized_variables/IsVariableInitialized_88" + input: "report_uninitialized_variables/IsVariableInitialized_89" + input: "report_uninitialized_variables/IsVariableInitialized_90" + input: "report_uninitialized_variables/IsVariableInitialized_91" + input: "report_uninitialized_variables/IsVariableInitialized_92" + input: "report_uninitialized_variables/IsVariableInitialized_93" + input: "report_uninitialized_variables/IsVariableInitialized_94" + input: "report_uninitialized_variables/IsVariableInitialized_95" + input: "report_uninitialized_variables/IsVariableInitialized_96" + input: "report_uninitialized_variables/IsVariableInitialized_97" + input: "report_uninitialized_variables/IsVariableInitialized_98" + input: "report_uninitialized_variables/IsVariableInitialized_99" + input: "report_uninitialized_variables/IsVariableInitialized_100" + input: "report_uninitialized_variables/IsVariableInitialized_101" + input: "report_uninitialized_variables/IsVariableInitialized_102" + input: "report_uninitialized_variables/IsVariableInitialized_103" + input: "report_uninitialized_variables/IsVariableInitialized_104" + input: "report_uninitialized_variables/IsVariableInitialized_105" + input: "report_uninitialized_variables/IsVariableInitialized_106" + input: "report_uninitialized_variables/IsVariableInitialized_107" + input: "report_uninitialized_variables/IsVariableInitialized_108" + input: "report_uninitialized_variables/IsVariableInitialized_109" + input: "report_uninitialized_variables/IsVariableInitialized_110" + input: "report_uninitialized_variables/IsVariableInitialized_111" + input: "report_uninitialized_variables/IsVariableInitialized_112" + input: "report_uninitialized_variables/IsVariableInitialized_113" + input: "report_uninitialized_variables/IsVariableInitialized_114" + input: "report_uninitialized_variables/IsVariableInitialized_115" + input: "report_uninitialized_variables/IsVariableInitialized_116" + input: "report_uninitialized_variables/IsVariableInitialized_117" + input: "report_uninitialized_variables/IsVariableInitialized_118" + input: "report_uninitialized_variables/IsVariableInitialized_119" + input: "report_uninitialized_variables/IsVariableInitialized_120" + input: "report_uninitialized_variables/IsVariableInitialized_121" + input: "report_uninitialized_variables/IsVariableInitialized_122" + input: "report_uninitialized_variables/IsVariableInitialized_123" + input: "report_uninitialized_variables/IsVariableInitialized_124" + input: "report_uninitialized_variables/IsVariableInitialized_125" + input: "report_uninitialized_variables/IsVariableInitialized_126" + input: "report_uninitialized_variables/IsVariableInitialized_127" + input: "report_uninitialized_variables/IsVariableInitialized_128" + input: "report_uninitialized_variables/IsVariableInitialized_129" + input: "report_uninitialized_variables/IsVariableInitialized_130" + input: "report_uninitialized_variables/IsVariableInitialized_131" + input: "report_uninitialized_variables/IsVariableInitialized_132" + input: "report_uninitialized_variables/IsVariableInitialized_133" + input: "report_uninitialized_variables/IsVariableInitialized_134" + input: "report_uninitialized_variables/IsVariableInitialized_135" + input: "report_uninitialized_variables/IsVariableInitialized_136" + input: "report_uninitialized_variables/IsVariableInitialized_137" + input: "report_uninitialized_variables/IsVariableInitialized_138" + input: "report_uninitialized_variables/IsVariableInitialized_139" + input: "report_uninitialized_variables/IsVariableInitialized_140" + input: "report_uninitialized_variables/IsVariableInitialized_141" + input: "report_uninitialized_variables/IsVariableInitialized_142" + input: "report_uninitialized_variables/IsVariableInitialized_143" + input: "report_uninitialized_variables/IsVariableInitialized_144" + input: "report_uninitialized_variables/IsVariableInitialized_145" + input: "report_uninitialized_variables/IsVariableInitialized_146" + input: "report_uninitialized_variables/IsVariableInitialized_147" + input: "report_uninitialized_variables/IsVariableInitialized_148" + input: "report_uninitialized_variables/IsVariableInitialized_149" + input: "report_uninitialized_variables/IsVariableInitialized_150" + input: "report_uninitialized_variables/IsVariableInitialized_151" + input: "report_uninitialized_variables/IsVariableInitialized_152" + input: "report_uninitialized_variables/IsVariableInitialized_153" + input: "report_uninitialized_variables/IsVariableInitialized_154" + input: "report_uninitialized_variables/IsVariableInitialized_155" + input: "report_uninitialized_variables/IsVariableInitialized_156" + input: "report_uninitialized_variables/IsVariableInitialized_157" + input: "report_uninitialized_variables/IsVariableInitialized_158" + input: "report_uninitialized_variables/IsVariableInitialized_159" + input: "report_uninitialized_variables/IsVariableInitialized_160" + input: "report_uninitialized_variables/IsVariableInitialized_161" + input: "report_uninitialized_variables/IsVariableInitialized_162" + input: "report_uninitialized_variables/IsVariableInitialized_163" + input: "report_uninitialized_variables/IsVariableInitialized_164" + input: "report_uninitialized_variables/IsVariableInitialized_165" + input: "report_uninitialized_variables/IsVariableInitialized_166" + input: "report_uninitialized_variables/IsVariableInitialized_167" + input: "report_uninitialized_variables/IsVariableInitialized_168" + input: "report_uninitialized_variables/IsVariableInitialized_169" + input: "report_uninitialized_variables/IsVariableInitialized_170" + input: "report_uninitialized_variables/IsVariableInitialized_171" + input: "report_uninitialized_variables/IsVariableInitialized_172" + input: "report_uninitialized_variables/IsVariableInitialized_173" + input: "report_uninitialized_variables/IsVariableInitialized_174" + input: "report_uninitialized_variables/IsVariableInitialized_175" + input: "report_uninitialized_variables/IsVariableInitialized_176" + input: "report_uninitialized_variables/IsVariableInitialized_177" + input: "report_uninitialized_variables/IsVariableInitialized_178" + input: "report_uninitialized_variables/IsVariableInitialized_179" + input: "report_uninitialized_variables/IsVariableInitialized_180" + input: "report_uninitialized_variables/IsVariableInitialized_181" + input: "report_uninitialized_variables/IsVariableInitialized_182" + input: "report_uninitialized_variables/IsVariableInitialized_183" + input: "report_uninitialized_variables/IsVariableInitialized_184" + input: "report_uninitialized_variables/IsVariableInitialized_185" + input: "report_uninitialized_variables/IsVariableInitialized_186" + input: "report_uninitialized_variables/IsVariableInitialized_187" + input: "report_uninitialized_variables/IsVariableInitialized_188" + input: "report_uninitialized_variables/IsVariableInitialized_189" + input: "report_uninitialized_variables/IsVariableInitialized_190" + input: "report_uninitialized_variables/IsVariableInitialized_191" + input: "report_uninitialized_variables/IsVariableInitialized_192" + input: "report_uninitialized_variables/IsVariableInitialized_193" + input: "report_uninitialized_variables/IsVariableInitialized_194" + input: "report_uninitialized_variables/IsVariableInitialized_195" + input: "report_uninitialized_variables/IsVariableInitialized_196" + input: "report_uninitialized_variables/IsVariableInitialized_197" + input: "report_uninitialized_variables/IsVariableInitialized_198" + input: "report_uninitialized_variables/IsVariableInitialized_199" + input: "report_uninitialized_variables/IsVariableInitialized_200" + input: "report_uninitialized_variables/IsVariableInitialized_201" + input: "report_uninitialized_variables/IsVariableInitialized_202" + input: "report_uninitialized_variables/IsVariableInitialized_203" + input: "report_uninitialized_variables/IsVariableInitialized_204" + input: "report_uninitialized_variables/IsVariableInitialized_205" + input: "report_uninitialized_variables/IsVariableInitialized_206" + input: "report_uninitialized_variables/IsVariableInitialized_207" + input: "report_uninitialized_variables/IsVariableInitialized_208" + input: "report_uninitialized_variables/IsVariableInitialized_209" + input: "report_uninitialized_variables/IsVariableInitialized_210" + input: "report_uninitialized_variables/IsVariableInitialized_211" + input: "report_uninitialized_variables/IsVariableInitialized_212" + input: "report_uninitialized_variables/IsVariableInitialized_213" + input: "report_uninitialized_variables/IsVariableInitialized_214" + input: "report_uninitialized_variables/IsVariableInitialized_215" + input: "report_uninitialized_variables/IsVariableInitialized_216" + input: "report_uninitialized_variables/IsVariableInitialized_217" + input: "report_uninitialized_variables/IsVariableInitialized_218" + input: "report_uninitialized_variables/IsVariableInitialized_219" + input: "report_uninitialized_variables/IsVariableInitialized_220" + input: "report_uninitialized_variables/IsVariableInitialized_221" + input: "report_uninitialized_variables/IsVariableInitialized_222" + input: "report_uninitialized_variables/IsVariableInitialized_223" + input: "report_uninitialized_variables/IsVariableInitialized_224" + input: "report_uninitialized_variables/IsVariableInitialized_225" + input: "report_uninitialized_variables/IsVariableInitialized_226" + input: "report_uninitialized_variables/IsVariableInitialized_227" + input: "report_uninitialized_variables/IsVariableInitialized_228" + input: "report_uninitialized_variables/IsVariableInitialized_229" + input: "report_uninitialized_variables/IsVariableInitialized_230" + input: "report_uninitialized_variables/IsVariableInitialized_231" + input: "report_uninitialized_variables/IsVariableInitialized_232" + input: "report_uninitialized_variables/IsVariableInitialized_233" + input: "report_uninitialized_variables/IsVariableInitialized_234" + input: "report_uninitialized_variables/IsVariableInitialized_235" + input: "report_uninitialized_variables/IsVariableInitialized_236" + input: "report_uninitialized_variables/IsVariableInitialized_237" + input: "report_uninitialized_variables/IsVariableInitialized_238" + input: "report_uninitialized_variables/IsVariableInitialized_239" + input: "report_uninitialized_variables/IsVariableInitialized_240" + input: "report_uninitialized_variables/IsVariableInitialized_241" + input: "report_uninitialized_variables/IsVariableInitialized_242" + input: "report_uninitialized_variables/IsVariableInitialized_243" + input: "report_uninitialized_variables/IsVariableInitialized_244" + input: "report_uninitialized_variables/IsVariableInitialized_245" + input: "report_uninitialized_variables/IsVariableInitialized_246" + input: "report_uninitialized_variables/IsVariableInitialized_247" + input: "report_uninitialized_variables/IsVariableInitialized_248" + input: "report_uninitialized_variables/IsVariableInitialized_249" + input: "report_uninitialized_variables/IsVariableInitialized_250" + input: "report_uninitialized_variables/IsVariableInitialized_251" + input: "report_uninitialized_variables/IsVariableInitialized_252" + input: "report_uninitialized_variables/IsVariableInitialized_253" + input: "report_uninitialized_variables/IsVariableInitialized_254" + input: "report_uninitialized_variables/IsVariableInitialized_255" + input: "report_uninitialized_variables/IsVariableInitialized_256" + input: "report_uninitialized_variables/IsVariableInitialized_257" + input: "report_uninitialized_variables/IsVariableInitialized_258" + input: "report_uninitialized_variables/IsVariableInitialized_259" + input: "report_uninitialized_variables/IsVariableInitialized_260" + input: "report_uninitialized_variables/IsVariableInitialized_261" + input: "report_uninitialized_variables/IsVariableInitialized_262" + input: "report_uninitialized_variables/IsVariableInitialized_263" + input: "report_uninitialized_variables/IsVariableInitialized_264" + input: "report_uninitialized_variables/IsVariableInitialized_265" + input: "report_uninitialized_variables/IsVariableInitialized_266" + input: "report_uninitialized_variables/IsVariableInitialized_267" + input: "report_uninitialized_variables/IsVariableInitialized_268" + input: "report_uninitialized_variables/IsVariableInitialized_269" + input: "report_uninitialized_variables/IsVariableInitialized_270" + input: "report_uninitialized_variables/IsVariableInitialized_271" + input: "report_uninitialized_variables/IsVariableInitialized_272" + input: "report_uninitialized_variables/IsVariableInitialized_273" + input: "report_uninitialized_variables/IsVariableInitialized_274" + input: "report_uninitialized_variables/IsVariableInitialized_275" + input: "report_uninitialized_variables/IsVariableInitialized_276" + input: "report_uninitialized_variables/IsVariableInitialized_277" + input: "report_uninitialized_variables/IsVariableInitialized_278" + input: "report_uninitialized_variables/IsVariableInitialized_279" + input: "report_uninitialized_variables/IsVariableInitialized_280" + input: "report_uninitialized_variables/IsVariableInitialized_281" + input: "report_uninitialized_variables/IsVariableInitialized_282" + input: "report_uninitialized_variables/IsVariableInitialized_283" + input: "report_uninitialized_variables/IsVariableInitialized_284" + input: "report_uninitialized_variables/IsVariableInitialized_285" + input: "report_uninitialized_variables/IsVariableInitialized_286" + input: "report_uninitialized_variables/IsVariableInitialized_287" + input: "report_uninitialized_variables/IsVariableInitialized_288" + input: "report_uninitialized_variables/IsVariableInitialized_289" + input: "report_uninitialized_variables/IsVariableInitialized_290" + input: "report_uninitialized_variables/IsVariableInitialized_291" + input: "report_uninitialized_variables/IsVariableInitialized_292" + input: "report_uninitialized_variables/IsVariableInitialized_293" + input: "report_uninitialized_variables/IsVariableInitialized_294" + input: "report_uninitialized_variables/IsVariableInitialized_295" + input: "report_uninitialized_variables/IsVariableInitialized_296" + input: "report_uninitialized_variables/IsVariableInitialized_297" + input: "report_uninitialized_variables/IsVariableInitialized_298" + input: "report_uninitialized_variables/IsVariableInitialized_299" + input: "report_uninitialized_variables/IsVariableInitialized_300" + input: "report_uninitialized_variables/IsVariableInitialized_301" + input: "report_uninitialized_variables/IsVariableInitialized_302" + input: "report_uninitialized_variables/IsVariableInitialized_303" + input: "report_uninitialized_variables/IsVariableInitialized_304" + input: "report_uninitialized_variables/IsVariableInitialized_305" + input: "report_uninitialized_variables/IsVariableInitialized_306" + input: "report_uninitialized_variables/IsVariableInitialized_307" + input: "report_uninitialized_variables/IsVariableInitialized_308" + input: "report_uninitialized_variables/IsVariableInitialized_309" + input: "report_uninitialized_variables/IsVariableInitialized_310" + input: "report_uninitialized_variables/IsVariableInitialized_311" + input: "report_uninitialized_variables/IsVariableInitialized_312" + input: "report_uninitialized_variables/IsVariableInitialized_313" + input: "report_uninitialized_variables/IsVariableInitialized_314" + input: "report_uninitialized_variables/IsVariableInitialized_315" + input: "report_uninitialized_variables/IsVariableInitialized_316" + input: "report_uninitialized_variables/IsVariableInitialized_317" + input: "report_uninitialized_variables/IsVariableInitialized_318" + input: "report_uninitialized_variables/IsVariableInitialized_319" + input: "report_uninitialized_variables/IsVariableInitialized_320" + input: "report_uninitialized_variables/IsVariableInitialized_321" + input: "report_uninitialized_variables/IsVariableInitialized_322" + input: "report_uninitialized_variables/IsVariableInitialized_323" + input: "report_uninitialized_variables/IsVariableInitialized_324" + input: "report_uninitialized_variables/IsVariableInitialized_325" + input: "report_uninitialized_variables/IsVariableInitialized_326" + input: "report_uninitialized_variables/IsVariableInitialized_327" + input: "report_uninitialized_variables/IsVariableInitialized_328" + input: "report_uninitialized_variables/IsVariableInitialized_329" + input: "report_uninitialized_variables/IsVariableInitialized_330" + input: "report_uninitialized_variables/IsVariableInitialized_331" + input: "report_uninitialized_variables/IsVariableInitialized_332" + input: "report_uninitialized_variables/IsVariableInitialized_333" + input: "report_uninitialized_variables/IsVariableInitialized_334" + input: "report_uninitialized_variables/IsVariableInitialized_335" + input: "report_uninitialized_variables/IsVariableInitialized_336" + input: "report_uninitialized_variables/IsVariableInitialized_337" + input: "report_uninitialized_variables/IsVariableInitialized_338" + input: "report_uninitialized_variables/IsVariableInitialized_339" + input: "report_uninitialized_variables/IsVariableInitialized_340" + input: "report_uninitialized_variables/IsVariableInitialized_341" + input: "report_uninitialized_variables/IsVariableInitialized_342" + input: "report_uninitialized_variables/IsVariableInitialized_343" + input: "report_uninitialized_variables/IsVariableInitialized_344" + input: "report_uninitialized_variables/IsVariableInitialized_345" + input: "report_uninitialized_variables/IsVariableInitialized_346" + input: "report_uninitialized_variables/IsVariableInitialized_347" + input: "report_uninitialized_variables/IsVariableInitialized_348" + input: "report_uninitialized_variables/IsVariableInitialized_349" + input: "report_uninitialized_variables/IsVariableInitialized_350" + input: "report_uninitialized_variables/IsVariableInitialized_351" + input: "report_uninitialized_variables/IsVariableInitialized_352" + input: "report_uninitialized_variables/IsVariableInitialized_353" + input: "report_uninitialized_variables/IsVariableInitialized_354" + input: "report_uninitialized_variables/IsVariableInitialized_355" + input: "report_uninitialized_variables/IsVariableInitialized_356" + input: "report_uninitialized_variables/IsVariableInitialized_357" + input: "report_uninitialized_variables/IsVariableInitialized_358" + input: "report_uninitialized_variables/IsVariableInitialized_359" + input: "report_uninitialized_variables/IsVariableInitialized_360" + input: "report_uninitialized_variables/IsVariableInitialized_361" + input: "report_uninitialized_variables/IsVariableInitialized_362" + input: "report_uninitialized_variables/IsVariableInitialized_363" + input: "report_uninitialized_variables/IsVariableInitialized_364" + input: "report_uninitialized_variables/IsVariableInitialized_365" + input: "report_uninitialized_variables/IsVariableInitialized_366" + input: "report_uninitialized_variables/IsVariableInitialized_367" + input: "report_uninitialized_variables/IsVariableInitialized_368" + input: "report_uninitialized_variables/IsVariableInitialized_369" + input: "report_uninitialized_variables/IsVariableInitialized_370" + input: "report_uninitialized_variables/IsVariableInitialized_371" + input: "report_uninitialized_variables/IsVariableInitialized_372" + input: "report_uninitialized_variables/IsVariableInitialized_373" + input: "report_uninitialized_variables/IsVariableInitialized_374" + input: "report_uninitialized_variables/IsVariableInitialized_375" + input: "report_uninitialized_variables/IsVariableInitialized_376" + input: "report_uninitialized_variables/IsVariableInitialized_377" + input: "report_uninitialized_variables/IsVariableInitialized_378" + input: "report_uninitialized_variables/IsVariableInitialized_379" + input: "report_uninitialized_variables/IsVariableInitialized_380" + input: "report_uninitialized_variables/IsVariableInitialized_381" + input: "report_uninitialized_variables/IsVariableInitialized_382" + input: "report_uninitialized_variables/IsVariableInitialized_383" + input: "report_uninitialized_variables/IsVariableInitialized_384" + input: "report_uninitialized_variables/IsVariableInitialized_385" + input: "report_uninitialized_variables/IsVariableInitialized_386" + input: "report_uninitialized_variables/IsVariableInitialized_387" + input: "report_uninitialized_variables/IsVariableInitialized_388" + input: "report_uninitialized_variables/IsVariableInitialized_389" + input: "report_uninitialized_variables/IsVariableInitialized_390" + input: "report_uninitialized_variables/IsVariableInitialized_391" + input: "report_uninitialized_variables/IsVariableInitialized_392" + input: "report_uninitialized_variables/IsVariableInitialized_393" + input: "report_uninitialized_variables/IsVariableInitialized_394" + input: "report_uninitialized_variables/IsVariableInitialized_395" + input: "report_uninitialized_variables/IsVariableInitialized_396" + input: "report_uninitialized_variables/IsVariableInitialized_397" + input: "report_uninitialized_variables/IsVariableInitialized_398" + input: "report_uninitialized_variables/IsVariableInitialized_399" + input: "report_uninitialized_variables/IsVariableInitialized_400" + input: "report_uninitialized_variables/IsVariableInitialized_401" + input: "report_uninitialized_variables/IsVariableInitialized_402" + input: "report_uninitialized_variables/IsVariableInitialized_403" + input: "report_uninitialized_variables/IsVariableInitialized_404" + input: "report_uninitialized_variables/IsVariableInitialized_405" + input: "report_uninitialized_variables/IsVariableInitialized_406" + input: "report_uninitialized_variables/IsVariableInitialized_407" + input: "report_uninitialized_variables/IsVariableInitialized_408" + input: "report_uninitialized_variables/IsVariableInitialized_409" + input: "report_uninitialized_variables/IsVariableInitialized_410" + input: "report_uninitialized_variables/IsVariableInitialized_411" + input: "report_uninitialized_variables/IsVariableInitialized_412" + input: "report_uninitialized_variables/IsVariableInitialized_413" + input: "report_uninitialized_variables/IsVariableInitialized_414" + input: "report_uninitialized_variables/IsVariableInitialized_415" + input: "report_uninitialized_variables/IsVariableInitialized_416" + input: "report_uninitialized_variables/IsVariableInitialized_417" + input: "report_uninitialized_variables/IsVariableInitialized_418" + input: "report_uninitialized_variables/IsVariableInitialized_419" + input: "report_uninitialized_variables/IsVariableInitialized_420" + input: "report_uninitialized_variables/IsVariableInitialized_421" + input: "report_uninitialized_variables/IsVariableInitialized_422" + input: "report_uninitialized_variables/IsVariableInitialized_423" + input: "report_uninitialized_variables/IsVariableInitialized_424" + input: "report_uninitialized_variables/IsVariableInitialized_425" + input: "report_uninitialized_variables/IsVariableInitialized_426" + input: "report_uninitialized_variables/IsVariableInitialized_427" + input: "report_uninitialized_variables/IsVariableInitialized_428" + input: "report_uninitialized_variables/IsVariableInitialized_429" + input: "report_uninitialized_variables/IsVariableInitialized_430" + input: "report_uninitialized_variables/IsVariableInitialized_431" + input: "report_uninitialized_variables/IsVariableInitialized_432" + input: "report_uninitialized_variables/IsVariableInitialized_433" + input: "report_uninitialized_variables/IsVariableInitialized_434" + input: "report_uninitialized_variables/IsVariableInitialized_435" + input: "report_uninitialized_variables/IsVariableInitialized_436" + input: "report_uninitialized_variables/IsVariableInitialized_437" + input: "report_uninitialized_variables/IsVariableInitialized_438" + input: "report_uninitialized_variables/IsVariableInitialized_439" + input: "report_uninitialized_variables/IsVariableInitialized_440" + input: "report_uninitialized_variables/IsVariableInitialized_441" + input: "report_uninitialized_variables/IsVariableInitialized_442" + input: "report_uninitialized_variables/IsVariableInitialized_443" + input: "report_uninitialized_variables/IsVariableInitialized_444" + input: "report_uninitialized_variables/IsVariableInitialized_445" + input: "report_uninitialized_variables/IsVariableInitialized_446" + input: "report_uninitialized_variables/IsVariableInitialized_447" + input: "report_uninitialized_variables/IsVariableInitialized_448" + input: "report_uninitialized_variables/IsVariableInitialized_449" + input: "report_uninitialized_variables/IsVariableInitialized_450" + input: "report_uninitialized_variables/IsVariableInitialized_451" + input: "report_uninitialized_variables/IsVariableInitialized_452" + input: "report_uninitialized_variables/IsVariableInitialized_453" + input: "report_uninitialized_variables/IsVariableInitialized_454" + input: "report_uninitialized_variables/IsVariableInitialized_455" + input: "report_uninitialized_variables/IsVariableInitialized_456" + input: "report_uninitialized_variables/IsVariableInitialized_457" + input: "report_uninitialized_variables/IsVariableInitialized_458" + input: "report_uninitialized_variables/IsVariableInitialized_459" + input: "report_uninitialized_variables/IsVariableInitialized_460" + input: "report_uninitialized_variables/IsVariableInitialized_461" + input: "report_uninitialized_variables/IsVariableInitialized_462" + input: "report_uninitialized_variables/IsVariableInitialized_463" + input: "report_uninitialized_variables/IsVariableInitialized_464" + input: "report_uninitialized_variables/IsVariableInitialized_465" + input: "report_uninitialized_variables/IsVariableInitialized_466" + input: "report_uninitialized_variables/IsVariableInitialized_467" + input: "report_uninitialized_variables/IsVariableInitialized_468" + input: "report_uninitialized_variables/IsVariableInitialized_469" + input: "report_uninitialized_variables/IsVariableInitialized_470" + input: "report_uninitialized_variables/IsVariableInitialized_471" + input: "report_uninitialized_variables/IsVariableInitialized_472" + input: "report_uninitialized_variables/IsVariableInitialized_473" + input: "report_uninitialized_variables/IsVariableInitialized_474" + input: "report_uninitialized_variables/IsVariableInitialized_475" + input: "report_uninitialized_variables/IsVariableInitialized_476" + input: "report_uninitialized_variables/IsVariableInitialized_477" + input: "report_uninitialized_variables/IsVariableInitialized_478" + input: "report_uninitialized_variables/IsVariableInitialized_479" + input: "report_uninitialized_variables/IsVariableInitialized_480" + input: "report_uninitialized_variables/IsVariableInitialized_481" + input: "report_uninitialized_variables/IsVariableInitialized_482" + input: "report_uninitialized_variables/IsVariableInitialized_483" + input: "report_uninitialized_variables/IsVariableInitialized_484" + input: "report_uninitialized_variables/IsVariableInitialized_485" + input: "report_uninitialized_variables/IsVariableInitialized_486" + input: "report_uninitialized_variables/IsVariableInitialized_487" + input: "report_uninitialized_variables/IsVariableInitialized_488" + input: "report_uninitialized_variables/IsVariableInitialized_489" + input: "report_uninitialized_variables/IsVariableInitialized_490" + input: "report_uninitialized_variables/IsVariableInitialized_491" + input: "report_uninitialized_variables/IsVariableInitialized_492" + input: "report_uninitialized_variables/IsVariableInitialized_493" + input: "report_uninitialized_variables/IsVariableInitialized_494" + input: "report_uninitialized_variables/IsVariableInitialized_495" + input: "report_uninitialized_variables/IsVariableInitialized_496" + input: "report_uninitialized_variables/IsVariableInitialized_497" + input: "report_uninitialized_variables/IsVariableInitialized_498" + input: "report_uninitialized_variables/IsVariableInitialized_499" + input: "report_uninitialized_variables/IsVariableInitialized_500" + input: "report_uninitialized_variables/IsVariableInitialized_501" + input: "report_uninitialized_variables/IsVariableInitialized_502" + input: "report_uninitialized_variables/IsVariableInitialized_503" + input: "report_uninitialized_variables/IsVariableInitialized_504" + input: "report_uninitialized_variables/IsVariableInitialized_505" + input: "report_uninitialized_variables/IsVariableInitialized_506" + input: "report_uninitialized_variables/IsVariableInitialized_507" + input: "report_uninitialized_variables/IsVariableInitialized_508" + input: "report_uninitialized_variables/IsVariableInitialized_509" + input: "report_uninitialized_variables/IsVariableInitialized_510" + input: "report_uninitialized_variables/IsVariableInitialized_511" + input: "report_uninitialized_variables/IsVariableInitialized_512" + input: "report_uninitialized_variables/IsVariableInitialized_513" + input: "report_uninitialized_variables/IsVariableInitialized_514" + input: "report_uninitialized_variables/IsVariableInitialized_515" + input: "report_uninitialized_variables/IsVariableInitialized_516" + input: "report_uninitialized_variables/IsVariableInitialized_517" + input: "report_uninitialized_variables/IsVariableInitialized_518" + input: "report_uninitialized_variables/IsVariableInitialized_519" + input: "report_uninitialized_variables/IsVariableInitialized_520" + input: "report_uninitialized_variables/IsVariableInitialized_521" + input: "report_uninitialized_variables/IsVariableInitialized_522" + input: "report_uninitialized_variables/IsVariableInitialized_523" + input: "report_uninitialized_variables/IsVariableInitialized_524" + input: "report_uninitialized_variables/IsVariableInitialized_525" + input: "report_uninitialized_variables/IsVariableInitialized_526" + input: "report_uninitialized_variables/IsVariableInitialized_527" + input: "report_uninitialized_variables/IsVariableInitialized_528" + input: "report_uninitialized_variables/IsVariableInitialized_529" + input: "report_uninitialized_variables/IsVariableInitialized_530" + input: "report_uninitialized_variables/IsVariableInitialized_531" + input: "report_uninitialized_variables/IsVariableInitialized_532" + input: "report_uninitialized_variables/IsVariableInitialized_533" + input: "report_uninitialized_variables/IsVariableInitialized_534" + input: "report_uninitialized_variables/IsVariableInitialized_535" + input: "report_uninitialized_variables/IsVariableInitialized_536" + input: "report_uninitialized_variables/IsVariableInitialized_537" + input: "report_uninitialized_variables/IsVariableInitialized_538" + input: "report_uninitialized_variables/IsVariableInitialized_539" + input: "report_uninitialized_variables/IsVariableInitialized_540" + input: "report_uninitialized_variables/IsVariableInitialized_541" + input: "report_uninitialized_variables/IsVariableInitialized_542" + input: "report_uninitialized_variables/IsVariableInitialized_543" + input: "report_uninitialized_variables/IsVariableInitialized_544" + input: "report_uninitialized_variables/IsVariableInitialized_545" + input: "report_uninitialized_variables/IsVariableInitialized_546" + input: "report_uninitialized_variables/IsVariableInitialized_547" + input: "report_uninitialized_variables/IsVariableInitialized_548" + input: "report_uninitialized_variables/IsVariableInitialized_549" + input: "report_uninitialized_variables/IsVariableInitialized_550" + input: "report_uninitialized_variables/IsVariableInitialized_551" + input: "report_uninitialized_variables/IsVariableInitialized_552" + input: "report_uninitialized_variables/IsVariableInitialized_553" + input: "report_uninitialized_variables/IsVariableInitialized_554" + input: "report_uninitialized_variables/IsVariableInitialized_555" + input: "report_uninitialized_variables/IsVariableInitialized_556" + input: "report_uninitialized_variables/IsVariableInitialized_557" + input: "report_uninitialized_variables/IsVariableInitialized_558" + input: "report_uninitialized_variables/IsVariableInitialized_559" + input: "report_uninitialized_variables/IsVariableInitialized_560" + input: "report_uninitialized_variables/IsVariableInitialized_561" + input: "report_uninitialized_variables/IsVariableInitialized_562" + input: "report_uninitialized_variables/IsVariableInitialized_563" + input: "report_uninitialized_variables/IsVariableInitialized_564" + input: "report_uninitialized_variables/IsVariableInitialized_565" + input: "report_uninitialized_variables/IsVariableInitialized_566" + input: "report_uninitialized_variables/IsVariableInitialized_567" + input: "report_uninitialized_variables/IsVariableInitialized_568" + input: "report_uninitialized_variables/IsVariableInitialized_569" + input: "report_uninitialized_variables/IsVariableInitialized_570" + input: "report_uninitialized_variables/IsVariableInitialized_571" + input: "report_uninitialized_variables/IsVariableInitialized_572" + input: "report_uninitialized_variables/IsVariableInitialized_573" + input: "report_uninitialized_variables/IsVariableInitialized_574" + input: "report_uninitialized_variables/IsVariableInitialized_575" + input: "report_uninitialized_variables/IsVariableInitialized_576" + input: "report_uninitialized_variables/IsVariableInitialized_577" + input: "report_uninitialized_variables/IsVariableInitialized_578" + input: "report_uninitialized_variables/IsVariableInitialized_579" + input: "report_uninitialized_variables/IsVariableInitialized_580" + input: "report_uninitialized_variables/IsVariableInitialized_581" + input: "report_uninitialized_variables/IsVariableInitialized_582" + input: "report_uninitialized_variables/IsVariableInitialized_583" + input: "report_uninitialized_variables/IsVariableInitialized_584" + input: "report_uninitialized_variables/IsVariableInitialized_585" + input: "report_uninitialized_variables/IsVariableInitialized_586" + input: "report_uninitialized_variables/IsVariableInitialized_587" + input: "report_uninitialized_variables/IsVariableInitialized_588" + input: "report_uninitialized_variables/IsVariableInitialized_589" + input: "report_uninitialized_variables/IsVariableInitialized_590" + input: "report_uninitialized_variables/IsVariableInitialized_591" + input: "report_uninitialized_variables/IsVariableInitialized_592" + input: "report_uninitialized_variables/IsVariableInitialized_593" + input: "report_uninitialized_variables/IsVariableInitialized_594" + input: "report_uninitialized_variables/IsVariableInitialized_595" + input: "report_uninitialized_variables/IsVariableInitialized_596" + input: "report_uninitialized_variables/IsVariableInitialized_597" + input: "report_uninitialized_variables/IsVariableInitialized_598" + input: "report_uninitialized_variables/IsVariableInitialized_599" + input: "report_uninitialized_variables/IsVariableInitialized_600" + input: "report_uninitialized_variables/IsVariableInitialized_601" + input: "report_uninitialized_variables/IsVariableInitialized_602" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 604 + } + } + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables/LogicalNot" + op: "LogicalNot" + input: "report_uninitialized_variables/stack" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables/Const" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "global_step" + string_val: "bert/embeddings/word_embeddings" + string_val: "bert/embeddings/token_type_embeddings" + string_val: "bert/embeddings/position_embeddings" + string_val: "bert/embeddings/LayerNorm/beta" + string_val: "bert/embeddings/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/attention/self/query/kernel" + string_val: "bert/encoder/layer_0/attention/self/query/bias" + string_val: "bert/encoder/layer_0/attention/self/key/kernel" + string_val: "bert/encoder/layer_0/attention/self/key/bias" + string_val: "bert/encoder/layer_0/attention/self/value/kernel" + string_val: "bert/encoder/layer_0/attention/self/value/bias" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel" + string_val: "bert/encoder/layer_0/attention/output/dense/bias" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel" + string_val: "bert/encoder/layer_0/intermediate/dense/bias" + string_val: "bert/encoder/layer_0/output/dense/kernel" + string_val: "bert/encoder/layer_0/output/dense/bias" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/attention/self/query/kernel" + string_val: "bert/encoder/layer_1/attention/self/query/bias" + string_val: "bert/encoder/layer_1/attention/self/key/kernel" + string_val: "bert/encoder/layer_1/attention/self/key/bias" + string_val: "bert/encoder/layer_1/attention/self/value/kernel" + string_val: "bert/encoder/layer_1/attention/self/value/bias" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel" + string_val: "bert/encoder/layer_1/attention/output/dense/bias" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel" + string_val: "bert/encoder/layer_1/intermediate/dense/bias" + string_val: "bert/encoder/layer_1/output/dense/kernel" + string_val: "bert/encoder/layer_1/output/dense/bias" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/attention/self/query/kernel" + string_val: "bert/encoder/layer_2/attention/self/query/bias" + string_val: "bert/encoder/layer_2/attention/self/key/kernel" + string_val: "bert/encoder/layer_2/attention/self/key/bias" + string_val: "bert/encoder/layer_2/attention/self/value/kernel" + string_val: "bert/encoder/layer_2/attention/self/value/bias" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel" + string_val: "bert/encoder/layer_2/attention/output/dense/bias" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel" + string_val: "bert/encoder/layer_2/intermediate/dense/bias" + string_val: "bert/encoder/layer_2/output/dense/kernel" + string_val: "bert/encoder/layer_2/output/dense/bias" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/attention/self/query/kernel" + string_val: "bert/encoder/layer_3/attention/self/query/bias" + string_val: "bert/encoder/layer_3/attention/self/key/kernel" + string_val: "bert/encoder/layer_3/attention/self/key/bias" + string_val: "bert/encoder/layer_3/attention/self/value/kernel" + string_val: "bert/encoder/layer_3/attention/self/value/bias" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel" + string_val: "bert/encoder/layer_3/attention/output/dense/bias" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel" + string_val: "bert/encoder/layer_3/intermediate/dense/bias" + string_val: "bert/encoder/layer_3/output/dense/kernel" + string_val: "bert/encoder/layer_3/output/dense/bias" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/attention/self/query/kernel" + string_val: "bert/encoder/layer_4/attention/self/query/bias" + string_val: "bert/encoder/layer_4/attention/self/key/kernel" + string_val: "bert/encoder/layer_4/attention/self/key/bias" + string_val: "bert/encoder/layer_4/attention/self/value/kernel" + string_val: "bert/encoder/layer_4/attention/self/value/bias" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel" + string_val: "bert/encoder/layer_4/attention/output/dense/bias" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel" + string_val: "bert/encoder/layer_4/intermediate/dense/bias" + string_val: "bert/encoder/layer_4/output/dense/kernel" + string_val: "bert/encoder/layer_4/output/dense/bias" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/attention/self/query/kernel" + string_val: "bert/encoder/layer_5/attention/self/query/bias" + string_val: "bert/encoder/layer_5/attention/self/key/kernel" + string_val: "bert/encoder/layer_5/attention/self/key/bias" + string_val: "bert/encoder/layer_5/attention/self/value/kernel" + string_val: "bert/encoder/layer_5/attention/self/value/bias" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel" + string_val: "bert/encoder/layer_5/attention/output/dense/bias" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel" + string_val: "bert/encoder/layer_5/intermediate/dense/bias" + string_val: "bert/encoder/layer_5/output/dense/kernel" + string_val: "bert/encoder/layer_5/output/dense/bias" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/attention/self/query/kernel" + string_val: "bert/encoder/layer_6/attention/self/query/bias" + string_val: "bert/encoder/layer_6/attention/self/key/kernel" + string_val: "bert/encoder/layer_6/attention/self/key/bias" + string_val: "bert/encoder/layer_6/attention/self/value/kernel" + string_val: "bert/encoder/layer_6/attention/self/value/bias" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel" + string_val: "bert/encoder/layer_6/attention/output/dense/bias" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel" + string_val: "bert/encoder/layer_6/intermediate/dense/bias" + string_val: "bert/encoder/layer_6/output/dense/kernel" + string_val: "bert/encoder/layer_6/output/dense/bias" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/attention/self/query/kernel" + string_val: "bert/encoder/layer_7/attention/self/query/bias" + string_val: "bert/encoder/layer_7/attention/self/key/kernel" + string_val: "bert/encoder/layer_7/attention/self/key/bias" + string_val: "bert/encoder/layer_7/attention/self/value/kernel" + string_val: "bert/encoder/layer_7/attention/self/value/bias" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel" + string_val: "bert/encoder/layer_7/attention/output/dense/bias" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel" + string_val: "bert/encoder/layer_7/intermediate/dense/bias" + string_val: "bert/encoder/layer_7/output/dense/kernel" + string_val: "bert/encoder/layer_7/output/dense/bias" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/attention/self/query/kernel" + string_val: "bert/encoder/layer_8/attention/self/query/bias" + string_val: "bert/encoder/layer_8/attention/self/key/kernel" + string_val: "bert/encoder/layer_8/attention/self/key/bias" + string_val: "bert/encoder/layer_8/attention/self/value/kernel" + string_val: "bert/encoder/layer_8/attention/self/value/bias" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel" + string_val: "bert/encoder/layer_8/attention/output/dense/bias" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel" + string_val: "bert/encoder/layer_8/intermediate/dense/bias" + string_val: "bert/encoder/layer_8/output/dense/kernel" + string_val: "bert/encoder/layer_8/output/dense/bias" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/attention/self/query/kernel" + string_val: "bert/encoder/layer_9/attention/self/query/bias" + string_val: "bert/encoder/layer_9/attention/self/key/kernel" + string_val: "bert/encoder/layer_9/attention/self/key/bias" + string_val: "bert/encoder/layer_9/attention/self/value/kernel" + string_val: "bert/encoder/layer_9/attention/self/value/bias" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel" + string_val: "bert/encoder/layer_9/attention/output/dense/bias" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel" + string_val: "bert/encoder/layer_9/intermediate/dense/bias" + string_val: "bert/encoder/layer_9/output/dense/kernel" + string_val: "bert/encoder/layer_9/output/dense/bias" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/attention/self/query/kernel" + string_val: "bert/encoder/layer_10/attention/self/query/bias" + string_val: "bert/encoder/layer_10/attention/self/key/kernel" + string_val: "bert/encoder/layer_10/attention/self/key/bias" + string_val: "bert/encoder/layer_10/attention/self/value/kernel" + string_val: "bert/encoder/layer_10/attention/self/value/bias" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel" + string_val: "bert/encoder/layer_10/attention/output/dense/bias" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel" + string_val: "bert/encoder/layer_10/intermediate/dense/bias" + string_val: "bert/encoder/layer_10/output/dense/kernel" + string_val: "bert/encoder/layer_10/output/dense/bias" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/attention/self/query/kernel" + string_val: "bert/encoder/layer_11/attention/self/query/bias" + string_val: "bert/encoder/layer_11/attention/self/key/kernel" + string_val: "bert/encoder/layer_11/attention/self/key/bias" + string_val: "bert/encoder/layer_11/attention/self/value/kernel" + string_val: "bert/encoder/layer_11/attention/self/value/bias" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel" + string_val: "bert/encoder/layer_11/attention/output/dense/bias" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel" + string_val: "bert/encoder/layer_11/intermediate/dense/bias" + string_val: "bert/encoder/layer_11/output/dense/kernel" + string_val: "bert/encoder/layer_11/output/dense/bias" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma" + string_val: "bert/pooler/dense/kernel" + string_val: "bert/pooler/dense/bias" + string_val: "output_weights" + string_val: "output_bias" + string_val: "bert/embeddings/word_embeddings/adam_m" + string_val: "bert/embeddings/word_embeddings/adam_v" + string_val: "bert/embeddings/token_type_embeddings/adam_m" + string_val: "bert/embeddings/token_type_embeddings/adam_v" + string_val: "bert/embeddings/position_embeddings/adam_m" + string_val: "bert/embeddings/position_embeddings/adam_v" + string_val: "bert/embeddings/LayerNorm/beta/adam_m" + string_val: "bert/embeddings/LayerNorm/beta/adam_v" + string_val: "bert/embeddings/LayerNorm/gamma/adam_m" + string_val: "bert/embeddings/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + string_val: "bert/pooler/dense/kernel/adam_m" + string_val: "bert/pooler/dense/kernel/adam_v" + string_val: "bert/pooler/dense/bias/adam_m" + string_val: "bert/pooler/dense/bias/adam_v" + string_val: "output_weights/adam_m" + string_val: "output_weights/adam_v" + string_val: "output_bias/adam_m" + string_val: "output_bias/adam_v" + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Shape" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice" + op: "StridedSlice" + input: "report_uninitialized_variables/boolean_mask/Shape" + input: "report_uninitialized_variables/boolean_mask/strided_slice/stack" + input: "report_uninitialized_variables/boolean_mask/strided_slice/stack_1" + input: "report_uninitialized_variables/boolean_mask/strided_slice/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Prod/reduction_indices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Prod" + op: "Prod" + input: "report_uninitialized_variables/boolean_mask/strided_slice" + input: "report_uninitialized_variables/boolean_mask/Prod/reduction_indices" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Shape_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_1" + op: "StridedSlice" + input: "report_uninitialized_variables/boolean_mask/Shape_1" + input: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack" + input: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack_1" + input: "report_uninitialized_variables/boolean_mask/strided_slice_1/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Shape_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/strided_slice_2" + op: "StridedSlice" + input: "report_uninitialized_variables/boolean_mask/Shape_2" + input: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack" + input: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack_1" + input: "report_uninitialized_variables/boolean_mask/strided_slice_2/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 1 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/concat/values_1" + op: "Pack" + input: "report_uninitialized_variables/boolean_mask/Prod" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 1 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/concat/axis" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/concat" + op: "ConcatV2" + input: "report_uninitialized_variables/boolean_mask/strided_slice_1" + input: "report_uninitialized_variables/boolean_mask/concat/values_1" + input: "report_uninitialized_variables/boolean_mask/strided_slice_2" + input: "report_uninitialized_variables/boolean_mask/concat/axis" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Reshape" + op: "Reshape" + input: "report_uninitialized_variables/Const" + input: "report_uninitialized_variables/boolean_mask/concat" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Reshape_1/shape" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Reshape_1" + op: "Reshape" + input: "report_uninitialized_variables/LogicalNot" + input: "report_uninitialized_variables/boolean_mask/Reshape_1/shape" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Where" + op: "Where" + input: "report_uninitialized_variables/boolean_mask/Reshape_1" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/Squeeze" + op: "Squeeze" + input: "report_uninitialized_variables/boolean_mask/Where" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } + attr { + key: "squeeze_dims" + value { + list { + i: 1 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/GatherV2/axis" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables/boolean_mask/GatherV2" + op: "GatherV2" + input: "report_uninitialized_variables/boolean_mask/Reshape" + input: "report_uninitialized_variables/boolean_mask/Squeeze" + input: "report_uninitialized_variables/boolean_mask/GatherV2/axis" + device: "/device:CPU:0" + attr { + key: "Taxis" + value { + type: DT_INT32 + } + } + attr { + key: "Tindices" + value { + type: DT_INT64 + } + } + attr { + key: "Tparams" + value { + type: DT_STRING + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } + attr { + key: "batch_dims" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_resources/Const" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "concat/axis" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "concat" + op: "ConcatV2" + input: "report_uninitialized_variables/boolean_mask/GatherV2" + input: "report_uninitialized_resources/Const" + input: "concat/axis" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/VarIsInitializedOp" + op: "VarIsInitializedOp" + input: "global_step" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_1" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_2" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_3" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_4" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_5" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_6" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_7" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_8" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_9" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_10" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_11" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_12" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_13" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_14" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_15" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_16" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_17" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_18" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_19" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_20" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_21" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_22" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_23" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_24" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_25" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_26" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_27" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_28" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_29" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_30" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_31" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_32" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_33" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_34" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_35" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_36" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_37" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_38" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_39" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_40" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_41" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_42" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_43" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_44" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_45" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_46" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_47" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_48" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_49" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_50" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_51" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_52" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_53" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_54" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_55" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_56" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_57" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_58" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_59" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_60" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_61" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_62" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_63" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_64" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_65" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_66" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_67" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_68" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_69" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_70" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_71" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_72" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_73" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_74" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_75" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_76" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_77" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_78" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_79" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_80" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_81" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_82" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_83" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_84" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_85" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_86" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_87" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_88" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_89" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_90" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_91" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_92" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_93" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_94" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_95" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_96" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_97" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_98" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_99" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_100" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_101" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_102" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_103" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_104" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_105" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_106" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_107" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_108" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_109" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_110" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_111" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_112" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_113" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_114" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_115" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_116" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_117" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_118" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_119" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_120" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_121" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_122" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_123" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_124" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_125" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_126" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_127" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_128" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_129" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_130" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_131" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_132" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_133" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_134" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_135" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_136" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_137" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_138" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_139" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_140" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_141" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_142" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_143" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_144" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_145" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_146" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_147" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_148" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_149" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_150" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_151" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_152" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_153" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_154" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_155" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_156" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_157" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_158" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_159" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_160" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_161" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_162" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_163" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_164" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_165" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_166" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_167" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_168" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_169" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_170" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_171" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_172" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_173" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_174" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_175" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_176" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_177" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_178" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_179" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_180" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_181" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_182" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_183" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_184" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_185" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_186" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_187" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_188" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_189" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_190" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_191" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_192" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_193" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_194" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_195" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_196" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_197" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_198" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_199" + op: "IsVariableInitialized" + input: "output_weights" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_200" + op: "IsVariableInitialized" + input: "output_bias" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_201" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_202" + op: "IsVariableInitialized" + input: "bert/embeddings/word_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_203" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_204" + op: "IsVariableInitialized" + input: "bert/embeddings/token_type_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_205" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_206" + op: "IsVariableInitialized" + input: "bert/embeddings/position_embeddings/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_207" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_208" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_209" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_210" + op: "IsVariableInitialized" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_211" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_212" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_213" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_214" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_215" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_216" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_217" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_218" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_219" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_220" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_221" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_222" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_223" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_224" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_225" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_226" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_227" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_228" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_229" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_230" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_231" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_232" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_233" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_234" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_235" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_236" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_237" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_238" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_239" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_240" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_241" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_242" + op: "IsVariableInitialized" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_243" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_244" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_245" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_246" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_247" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_248" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_249" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_250" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_251" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_252" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_253" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_254" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_255" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_256" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_257" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_258" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_259" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_260" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_261" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_262" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_263" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_264" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_265" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_266" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_267" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_268" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_269" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_270" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_271" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_272" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_273" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_274" + op: "IsVariableInitialized" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_275" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_276" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_277" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_278" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_279" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_280" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_281" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_282" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_283" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_284" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_285" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_286" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_287" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_288" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_289" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_290" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_291" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_292" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_293" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_294" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_295" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_296" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_297" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_298" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_299" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_300" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_301" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_302" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_303" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_304" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_305" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_306" + op: "IsVariableInitialized" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_307" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_308" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_309" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_310" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_311" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_312" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_313" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_314" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_315" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_316" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_317" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_318" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_319" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_320" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_321" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_322" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_323" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_324" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_325" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_326" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_327" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_328" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_329" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_330" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_331" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_332" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_333" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_334" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_335" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_336" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_337" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_338" + op: "IsVariableInitialized" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_339" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_340" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_341" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_342" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_343" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_344" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_345" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_346" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_347" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_348" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_349" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_350" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_351" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_352" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_353" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_354" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_355" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_356" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_357" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_358" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_359" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_360" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_361" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_362" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_363" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_364" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_365" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_366" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_367" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_368" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_369" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_370" + op: "IsVariableInitialized" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_371" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_372" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_373" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_374" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_375" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_376" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_377" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_378" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_379" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_380" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_381" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_382" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_383" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_384" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_385" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_386" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_387" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_388" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_389" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_390" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_391" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_392" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_393" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_394" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_395" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_396" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_397" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_398" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_399" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_400" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_401" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_402" + op: "IsVariableInitialized" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_403" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_404" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_405" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_406" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_407" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_408" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_409" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_410" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_411" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_412" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_413" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_414" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_415" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_416" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_417" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_418" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_419" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_420" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_421" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_422" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_423" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_424" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_425" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_426" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_427" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_428" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_429" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_430" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_431" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_432" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_433" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_434" + op: "IsVariableInitialized" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_435" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_436" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_437" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_438" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_439" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_440" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_441" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_442" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_443" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_444" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_445" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_446" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_447" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_448" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_449" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_450" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_451" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_452" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_453" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_454" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_455" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_456" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_457" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_458" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_459" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_460" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_461" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_462" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_463" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_464" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_465" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_466" + op: "IsVariableInitialized" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_467" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_468" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_469" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_470" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_471" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_472" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_473" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_474" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_475" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_476" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_477" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_478" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_479" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_480" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_481" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_482" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_483" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_484" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_485" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_486" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_487" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_488" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_489" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_490" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_491" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_492" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_493" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_494" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_495" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_496" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_497" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_498" + op: "IsVariableInitialized" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_499" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_500" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_501" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_502" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_503" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_504" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_505" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_506" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_507" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_508" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_509" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_510" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_511" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_512" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_513" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_514" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_515" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_516" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_517" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_518" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_519" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_520" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_521" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_522" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_523" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_524" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_525" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_526" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_527" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_528" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_529" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_530" + op: "IsVariableInitialized" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_531" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_532" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_533" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_534" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_535" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_536" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_537" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_538" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_539" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_540" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_541" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_542" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_543" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_544" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_545" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_546" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_547" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_548" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_549" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_550" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_551" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_552" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_553" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_554" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_555" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_556" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_557" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_558" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_559" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_560" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_561" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_562" + op: "IsVariableInitialized" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_563" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_564" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_565" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_566" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_567" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_568" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_569" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_570" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_571" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_572" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_573" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_574" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_575" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_576" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_577" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_578" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_579" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_580" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_581" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_582" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_583" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_584" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_585" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_586" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_587" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_588" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_589" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_590" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_591" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_592" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_593" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_594" + op: "IsVariableInitialized" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_595" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_596" + op: "IsVariableInitialized" + input: "bert/pooler/dense/kernel/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_597" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_598" + op: "IsVariableInitialized" + input: "bert/pooler/dense/bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_599" + op: "IsVariableInitialized" + input: "output_weights/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_600" + op: "IsVariableInitialized" + input: "output_weights/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_601" + op: "IsVariableInitialized" + input: "output_bias/adam_m" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/IsVariableInitialized_602" + op: "IsVariableInitialized" + input: "output_bias/adam_v" + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_FLOAT + } + } +} +node { + name: "report_uninitialized_variables_1/stack" + op: "Pack" + input: "report_uninitialized_variables_1/VarIsInitializedOp" + input: "report_uninitialized_variables_1/IsVariableInitialized" + input: "report_uninitialized_variables_1/IsVariableInitialized_1" + input: "report_uninitialized_variables_1/IsVariableInitialized_2" + input: "report_uninitialized_variables_1/IsVariableInitialized_3" + input: "report_uninitialized_variables_1/IsVariableInitialized_4" + input: "report_uninitialized_variables_1/IsVariableInitialized_5" + input: "report_uninitialized_variables_1/IsVariableInitialized_6" + input: "report_uninitialized_variables_1/IsVariableInitialized_7" + input: "report_uninitialized_variables_1/IsVariableInitialized_8" + input: "report_uninitialized_variables_1/IsVariableInitialized_9" + input: "report_uninitialized_variables_1/IsVariableInitialized_10" + input: "report_uninitialized_variables_1/IsVariableInitialized_11" + input: "report_uninitialized_variables_1/IsVariableInitialized_12" + input: "report_uninitialized_variables_1/IsVariableInitialized_13" + input: "report_uninitialized_variables_1/IsVariableInitialized_14" + input: "report_uninitialized_variables_1/IsVariableInitialized_15" + input: "report_uninitialized_variables_1/IsVariableInitialized_16" + input: "report_uninitialized_variables_1/IsVariableInitialized_17" + input: "report_uninitialized_variables_1/IsVariableInitialized_18" + input: "report_uninitialized_variables_1/IsVariableInitialized_19" + input: "report_uninitialized_variables_1/IsVariableInitialized_20" + input: "report_uninitialized_variables_1/IsVariableInitialized_21" + input: "report_uninitialized_variables_1/IsVariableInitialized_22" + input: "report_uninitialized_variables_1/IsVariableInitialized_23" + input: "report_uninitialized_variables_1/IsVariableInitialized_24" + input: "report_uninitialized_variables_1/IsVariableInitialized_25" + input: "report_uninitialized_variables_1/IsVariableInitialized_26" + input: "report_uninitialized_variables_1/IsVariableInitialized_27" + input: "report_uninitialized_variables_1/IsVariableInitialized_28" + input: "report_uninitialized_variables_1/IsVariableInitialized_29" + input: "report_uninitialized_variables_1/IsVariableInitialized_30" + input: "report_uninitialized_variables_1/IsVariableInitialized_31" + input: "report_uninitialized_variables_1/IsVariableInitialized_32" + input: "report_uninitialized_variables_1/IsVariableInitialized_33" + input: "report_uninitialized_variables_1/IsVariableInitialized_34" + input: "report_uninitialized_variables_1/IsVariableInitialized_35" + input: "report_uninitialized_variables_1/IsVariableInitialized_36" + input: "report_uninitialized_variables_1/IsVariableInitialized_37" + input: "report_uninitialized_variables_1/IsVariableInitialized_38" + input: "report_uninitialized_variables_1/IsVariableInitialized_39" + input: "report_uninitialized_variables_1/IsVariableInitialized_40" + input: "report_uninitialized_variables_1/IsVariableInitialized_41" + input: "report_uninitialized_variables_1/IsVariableInitialized_42" + input: "report_uninitialized_variables_1/IsVariableInitialized_43" + input: "report_uninitialized_variables_1/IsVariableInitialized_44" + input: "report_uninitialized_variables_1/IsVariableInitialized_45" + input: "report_uninitialized_variables_1/IsVariableInitialized_46" + input: "report_uninitialized_variables_1/IsVariableInitialized_47" + input: "report_uninitialized_variables_1/IsVariableInitialized_48" + input: "report_uninitialized_variables_1/IsVariableInitialized_49" + input: "report_uninitialized_variables_1/IsVariableInitialized_50" + input: "report_uninitialized_variables_1/IsVariableInitialized_51" + input: "report_uninitialized_variables_1/IsVariableInitialized_52" + input: "report_uninitialized_variables_1/IsVariableInitialized_53" + input: "report_uninitialized_variables_1/IsVariableInitialized_54" + input: "report_uninitialized_variables_1/IsVariableInitialized_55" + input: "report_uninitialized_variables_1/IsVariableInitialized_56" + input: "report_uninitialized_variables_1/IsVariableInitialized_57" + input: "report_uninitialized_variables_1/IsVariableInitialized_58" + input: "report_uninitialized_variables_1/IsVariableInitialized_59" + input: "report_uninitialized_variables_1/IsVariableInitialized_60" + input: "report_uninitialized_variables_1/IsVariableInitialized_61" + input: "report_uninitialized_variables_1/IsVariableInitialized_62" + input: "report_uninitialized_variables_1/IsVariableInitialized_63" + input: "report_uninitialized_variables_1/IsVariableInitialized_64" + input: "report_uninitialized_variables_1/IsVariableInitialized_65" + input: "report_uninitialized_variables_1/IsVariableInitialized_66" + input: "report_uninitialized_variables_1/IsVariableInitialized_67" + input: "report_uninitialized_variables_1/IsVariableInitialized_68" + input: "report_uninitialized_variables_1/IsVariableInitialized_69" + input: "report_uninitialized_variables_1/IsVariableInitialized_70" + input: "report_uninitialized_variables_1/IsVariableInitialized_71" + input: "report_uninitialized_variables_1/IsVariableInitialized_72" + input: "report_uninitialized_variables_1/IsVariableInitialized_73" + input: "report_uninitialized_variables_1/IsVariableInitialized_74" + input: "report_uninitialized_variables_1/IsVariableInitialized_75" + input: "report_uninitialized_variables_1/IsVariableInitialized_76" + input: "report_uninitialized_variables_1/IsVariableInitialized_77" + input: "report_uninitialized_variables_1/IsVariableInitialized_78" + input: "report_uninitialized_variables_1/IsVariableInitialized_79" + input: "report_uninitialized_variables_1/IsVariableInitialized_80" + input: "report_uninitialized_variables_1/IsVariableInitialized_81" + input: "report_uninitialized_variables_1/IsVariableInitialized_82" + input: "report_uninitialized_variables_1/IsVariableInitialized_83" + input: "report_uninitialized_variables_1/IsVariableInitialized_84" + input: "report_uninitialized_variables_1/IsVariableInitialized_85" + input: "report_uninitialized_variables_1/IsVariableInitialized_86" + input: "report_uninitialized_variables_1/IsVariableInitialized_87" + input: "report_uninitialized_variables_1/IsVariableInitialized_88" + input: "report_uninitialized_variables_1/IsVariableInitialized_89" + input: "report_uninitialized_variables_1/IsVariableInitialized_90" + input: "report_uninitialized_variables_1/IsVariableInitialized_91" + input: "report_uninitialized_variables_1/IsVariableInitialized_92" + input: "report_uninitialized_variables_1/IsVariableInitialized_93" + input: "report_uninitialized_variables_1/IsVariableInitialized_94" + input: "report_uninitialized_variables_1/IsVariableInitialized_95" + input: "report_uninitialized_variables_1/IsVariableInitialized_96" + input: "report_uninitialized_variables_1/IsVariableInitialized_97" + input: "report_uninitialized_variables_1/IsVariableInitialized_98" + input: "report_uninitialized_variables_1/IsVariableInitialized_99" + input: "report_uninitialized_variables_1/IsVariableInitialized_100" + input: "report_uninitialized_variables_1/IsVariableInitialized_101" + input: "report_uninitialized_variables_1/IsVariableInitialized_102" + input: "report_uninitialized_variables_1/IsVariableInitialized_103" + input: "report_uninitialized_variables_1/IsVariableInitialized_104" + input: "report_uninitialized_variables_1/IsVariableInitialized_105" + input: "report_uninitialized_variables_1/IsVariableInitialized_106" + input: "report_uninitialized_variables_1/IsVariableInitialized_107" + input: "report_uninitialized_variables_1/IsVariableInitialized_108" + input: "report_uninitialized_variables_1/IsVariableInitialized_109" + input: "report_uninitialized_variables_1/IsVariableInitialized_110" + input: "report_uninitialized_variables_1/IsVariableInitialized_111" + input: "report_uninitialized_variables_1/IsVariableInitialized_112" + input: "report_uninitialized_variables_1/IsVariableInitialized_113" + input: "report_uninitialized_variables_1/IsVariableInitialized_114" + input: "report_uninitialized_variables_1/IsVariableInitialized_115" + input: "report_uninitialized_variables_1/IsVariableInitialized_116" + input: "report_uninitialized_variables_1/IsVariableInitialized_117" + input: "report_uninitialized_variables_1/IsVariableInitialized_118" + input: "report_uninitialized_variables_1/IsVariableInitialized_119" + input: "report_uninitialized_variables_1/IsVariableInitialized_120" + input: "report_uninitialized_variables_1/IsVariableInitialized_121" + input: "report_uninitialized_variables_1/IsVariableInitialized_122" + input: "report_uninitialized_variables_1/IsVariableInitialized_123" + input: "report_uninitialized_variables_1/IsVariableInitialized_124" + input: "report_uninitialized_variables_1/IsVariableInitialized_125" + input: "report_uninitialized_variables_1/IsVariableInitialized_126" + input: "report_uninitialized_variables_1/IsVariableInitialized_127" + input: "report_uninitialized_variables_1/IsVariableInitialized_128" + input: "report_uninitialized_variables_1/IsVariableInitialized_129" + input: "report_uninitialized_variables_1/IsVariableInitialized_130" + input: "report_uninitialized_variables_1/IsVariableInitialized_131" + input: "report_uninitialized_variables_1/IsVariableInitialized_132" + input: "report_uninitialized_variables_1/IsVariableInitialized_133" + input: "report_uninitialized_variables_1/IsVariableInitialized_134" + input: "report_uninitialized_variables_1/IsVariableInitialized_135" + input: "report_uninitialized_variables_1/IsVariableInitialized_136" + input: "report_uninitialized_variables_1/IsVariableInitialized_137" + input: "report_uninitialized_variables_1/IsVariableInitialized_138" + input: "report_uninitialized_variables_1/IsVariableInitialized_139" + input: "report_uninitialized_variables_1/IsVariableInitialized_140" + input: "report_uninitialized_variables_1/IsVariableInitialized_141" + input: "report_uninitialized_variables_1/IsVariableInitialized_142" + input: "report_uninitialized_variables_1/IsVariableInitialized_143" + input: "report_uninitialized_variables_1/IsVariableInitialized_144" + input: "report_uninitialized_variables_1/IsVariableInitialized_145" + input: "report_uninitialized_variables_1/IsVariableInitialized_146" + input: "report_uninitialized_variables_1/IsVariableInitialized_147" + input: "report_uninitialized_variables_1/IsVariableInitialized_148" + input: "report_uninitialized_variables_1/IsVariableInitialized_149" + input: "report_uninitialized_variables_1/IsVariableInitialized_150" + input: "report_uninitialized_variables_1/IsVariableInitialized_151" + input: "report_uninitialized_variables_1/IsVariableInitialized_152" + input: "report_uninitialized_variables_1/IsVariableInitialized_153" + input: "report_uninitialized_variables_1/IsVariableInitialized_154" + input: "report_uninitialized_variables_1/IsVariableInitialized_155" + input: "report_uninitialized_variables_1/IsVariableInitialized_156" + input: "report_uninitialized_variables_1/IsVariableInitialized_157" + input: "report_uninitialized_variables_1/IsVariableInitialized_158" + input: "report_uninitialized_variables_1/IsVariableInitialized_159" + input: "report_uninitialized_variables_1/IsVariableInitialized_160" + input: "report_uninitialized_variables_1/IsVariableInitialized_161" + input: "report_uninitialized_variables_1/IsVariableInitialized_162" + input: "report_uninitialized_variables_1/IsVariableInitialized_163" + input: "report_uninitialized_variables_1/IsVariableInitialized_164" + input: "report_uninitialized_variables_1/IsVariableInitialized_165" + input: "report_uninitialized_variables_1/IsVariableInitialized_166" + input: "report_uninitialized_variables_1/IsVariableInitialized_167" + input: "report_uninitialized_variables_1/IsVariableInitialized_168" + input: "report_uninitialized_variables_1/IsVariableInitialized_169" + input: "report_uninitialized_variables_1/IsVariableInitialized_170" + input: "report_uninitialized_variables_1/IsVariableInitialized_171" + input: "report_uninitialized_variables_1/IsVariableInitialized_172" + input: "report_uninitialized_variables_1/IsVariableInitialized_173" + input: "report_uninitialized_variables_1/IsVariableInitialized_174" + input: "report_uninitialized_variables_1/IsVariableInitialized_175" + input: "report_uninitialized_variables_1/IsVariableInitialized_176" + input: "report_uninitialized_variables_1/IsVariableInitialized_177" + input: "report_uninitialized_variables_1/IsVariableInitialized_178" + input: "report_uninitialized_variables_1/IsVariableInitialized_179" + input: "report_uninitialized_variables_1/IsVariableInitialized_180" + input: "report_uninitialized_variables_1/IsVariableInitialized_181" + input: "report_uninitialized_variables_1/IsVariableInitialized_182" + input: "report_uninitialized_variables_1/IsVariableInitialized_183" + input: "report_uninitialized_variables_1/IsVariableInitialized_184" + input: "report_uninitialized_variables_1/IsVariableInitialized_185" + input: "report_uninitialized_variables_1/IsVariableInitialized_186" + input: "report_uninitialized_variables_1/IsVariableInitialized_187" + input: "report_uninitialized_variables_1/IsVariableInitialized_188" + input: "report_uninitialized_variables_1/IsVariableInitialized_189" + input: "report_uninitialized_variables_1/IsVariableInitialized_190" + input: "report_uninitialized_variables_1/IsVariableInitialized_191" + input: "report_uninitialized_variables_1/IsVariableInitialized_192" + input: "report_uninitialized_variables_1/IsVariableInitialized_193" + input: "report_uninitialized_variables_1/IsVariableInitialized_194" + input: "report_uninitialized_variables_1/IsVariableInitialized_195" + input: "report_uninitialized_variables_1/IsVariableInitialized_196" + input: "report_uninitialized_variables_1/IsVariableInitialized_197" + input: "report_uninitialized_variables_1/IsVariableInitialized_198" + input: "report_uninitialized_variables_1/IsVariableInitialized_199" + input: "report_uninitialized_variables_1/IsVariableInitialized_200" + input: "report_uninitialized_variables_1/IsVariableInitialized_201" + input: "report_uninitialized_variables_1/IsVariableInitialized_202" + input: "report_uninitialized_variables_1/IsVariableInitialized_203" + input: "report_uninitialized_variables_1/IsVariableInitialized_204" + input: "report_uninitialized_variables_1/IsVariableInitialized_205" + input: "report_uninitialized_variables_1/IsVariableInitialized_206" + input: "report_uninitialized_variables_1/IsVariableInitialized_207" + input: "report_uninitialized_variables_1/IsVariableInitialized_208" + input: "report_uninitialized_variables_1/IsVariableInitialized_209" + input: "report_uninitialized_variables_1/IsVariableInitialized_210" + input: "report_uninitialized_variables_1/IsVariableInitialized_211" + input: "report_uninitialized_variables_1/IsVariableInitialized_212" + input: "report_uninitialized_variables_1/IsVariableInitialized_213" + input: "report_uninitialized_variables_1/IsVariableInitialized_214" + input: "report_uninitialized_variables_1/IsVariableInitialized_215" + input: "report_uninitialized_variables_1/IsVariableInitialized_216" + input: "report_uninitialized_variables_1/IsVariableInitialized_217" + input: "report_uninitialized_variables_1/IsVariableInitialized_218" + input: "report_uninitialized_variables_1/IsVariableInitialized_219" + input: "report_uninitialized_variables_1/IsVariableInitialized_220" + input: "report_uninitialized_variables_1/IsVariableInitialized_221" + input: "report_uninitialized_variables_1/IsVariableInitialized_222" + input: "report_uninitialized_variables_1/IsVariableInitialized_223" + input: "report_uninitialized_variables_1/IsVariableInitialized_224" + input: "report_uninitialized_variables_1/IsVariableInitialized_225" + input: "report_uninitialized_variables_1/IsVariableInitialized_226" + input: "report_uninitialized_variables_1/IsVariableInitialized_227" + input: "report_uninitialized_variables_1/IsVariableInitialized_228" + input: "report_uninitialized_variables_1/IsVariableInitialized_229" + input: "report_uninitialized_variables_1/IsVariableInitialized_230" + input: "report_uninitialized_variables_1/IsVariableInitialized_231" + input: "report_uninitialized_variables_1/IsVariableInitialized_232" + input: "report_uninitialized_variables_1/IsVariableInitialized_233" + input: "report_uninitialized_variables_1/IsVariableInitialized_234" + input: "report_uninitialized_variables_1/IsVariableInitialized_235" + input: "report_uninitialized_variables_1/IsVariableInitialized_236" + input: "report_uninitialized_variables_1/IsVariableInitialized_237" + input: "report_uninitialized_variables_1/IsVariableInitialized_238" + input: "report_uninitialized_variables_1/IsVariableInitialized_239" + input: "report_uninitialized_variables_1/IsVariableInitialized_240" + input: "report_uninitialized_variables_1/IsVariableInitialized_241" + input: "report_uninitialized_variables_1/IsVariableInitialized_242" + input: "report_uninitialized_variables_1/IsVariableInitialized_243" + input: "report_uninitialized_variables_1/IsVariableInitialized_244" + input: "report_uninitialized_variables_1/IsVariableInitialized_245" + input: "report_uninitialized_variables_1/IsVariableInitialized_246" + input: "report_uninitialized_variables_1/IsVariableInitialized_247" + input: "report_uninitialized_variables_1/IsVariableInitialized_248" + input: "report_uninitialized_variables_1/IsVariableInitialized_249" + input: "report_uninitialized_variables_1/IsVariableInitialized_250" + input: "report_uninitialized_variables_1/IsVariableInitialized_251" + input: "report_uninitialized_variables_1/IsVariableInitialized_252" + input: "report_uninitialized_variables_1/IsVariableInitialized_253" + input: "report_uninitialized_variables_1/IsVariableInitialized_254" + input: "report_uninitialized_variables_1/IsVariableInitialized_255" + input: "report_uninitialized_variables_1/IsVariableInitialized_256" + input: "report_uninitialized_variables_1/IsVariableInitialized_257" + input: "report_uninitialized_variables_1/IsVariableInitialized_258" + input: "report_uninitialized_variables_1/IsVariableInitialized_259" + input: "report_uninitialized_variables_1/IsVariableInitialized_260" + input: "report_uninitialized_variables_1/IsVariableInitialized_261" + input: "report_uninitialized_variables_1/IsVariableInitialized_262" + input: "report_uninitialized_variables_1/IsVariableInitialized_263" + input: "report_uninitialized_variables_1/IsVariableInitialized_264" + input: "report_uninitialized_variables_1/IsVariableInitialized_265" + input: "report_uninitialized_variables_1/IsVariableInitialized_266" + input: "report_uninitialized_variables_1/IsVariableInitialized_267" + input: "report_uninitialized_variables_1/IsVariableInitialized_268" + input: "report_uninitialized_variables_1/IsVariableInitialized_269" + input: "report_uninitialized_variables_1/IsVariableInitialized_270" + input: "report_uninitialized_variables_1/IsVariableInitialized_271" + input: "report_uninitialized_variables_1/IsVariableInitialized_272" + input: "report_uninitialized_variables_1/IsVariableInitialized_273" + input: "report_uninitialized_variables_1/IsVariableInitialized_274" + input: "report_uninitialized_variables_1/IsVariableInitialized_275" + input: "report_uninitialized_variables_1/IsVariableInitialized_276" + input: "report_uninitialized_variables_1/IsVariableInitialized_277" + input: "report_uninitialized_variables_1/IsVariableInitialized_278" + input: "report_uninitialized_variables_1/IsVariableInitialized_279" + input: "report_uninitialized_variables_1/IsVariableInitialized_280" + input: "report_uninitialized_variables_1/IsVariableInitialized_281" + input: "report_uninitialized_variables_1/IsVariableInitialized_282" + input: "report_uninitialized_variables_1/IsVariableInitialized_283" + input: "report_uninitialized_variables_1/IsVariableInitialized_284" + input: "report_uninitialized_variables_1/IsVariableInitialized_285" + input: "report_uninitialized_variables_1/IsVariableInitialized_286" + input: "report_uninitialized_variables_1/IsVariableInitialized_287" + input: "report_uninitialized_variables_1/IsVariableInitialized_288" + input: "report_uninitialized_variables_1/IsVariableInitialized_289" + input: "report_uninitialized_variables_1/IsVariableInitialized_290" + input: "report_uninitialized_variables_1/IsVariableInitialized_291" + input: "report_uninitialized_variables_1/IsVariableInitialized_292" + input: "report_uninitialized_variables_1/IsVariableInitialized_293" + input: "report_uninitialized_variables_1/IsVariableInitialized_294" + input: "report_uninitialized_variables_1/IsVariableInitialized_295" + input: "report_uninitialized_variables_1/IsVariableInitialized_296" + input: "report_uninitialized_variables_1/IsVariableInitialized_297" + input: "report_uninitialized_variables_1/IsVariableInitialized_298" + input: "report_uninitialized_variables_1/IsVariableInitialized_299" + input: "report_uninitialized_variables_1/IsVariableInitialized_300" + input: "report_uninitialized_variables_1/IsVariableInitialized_301" + input: "report_uninitialized_variables_1/IsVariableInitialized_302" + input: "report_uninitialized_variables_1/IsVariableInitialized_303" + input: "report_uninitialized_variables_1/IsVariableInitialized_304" + input: "report_uninitialized_variables_1/IsVariableInitialized_305" + input: "report_uninitialized_variables_1/IsVariableInitialized_306" + input: "report_uninitialized_variables_1/IsVariableInitialized_307" + input: "report_uninitialized_variables_1/IsVariableInitialized_308" + input: "report_uninitialized_variables_1/IsVariableInitialized_309" + input: "report_uninitialized_variables_1/IsVariableInitialized_310" + input: "report_uninitialized_variables_1/IsVariableInitialized_311" + input: "report_uninitialized_variables_1/IsVariableInitialized_312" + input: "report_uninitialized_variables_1/IsVariableInitialized_313" + input: "report_uninitialized_variables_1/IsVariableInitialized_314" + input: "report_uninitialized_variables_1/IsVariableInitialized_315" + input: "report_uninitialized_variables_1/IsVariableInitialized_316" + input: "report_uninitialized_variables_1/IsVariableInitialized_317" + input: "report_uninitialized_variables_1/IsVariableInitialized_318" + input: "report_uninitialized_variables_1/IsVariableInitialized_319" + input: "report_uninitialized_variables_1/IsVariableInitialized_320" + input: "report_uninitialized_variables_1/IsVariableInitialized_321" + input: "report_uninitialized_variables_1/IsVariableInitialized_322" + input: "report_uninitialized_variables_1/IsVariableInitialized_323" + input: "report_uninitialized_variables_1/IsVariableInitialized_324" + input: "report_uninitialized_variables_1/IsVariableInitialized_325" + input: "report_uninitialized_variables_1/IsVariableInitialized_326" + input: "report_uninitialized_variables_1/IsVariableInitialized_327" + input: "report_uninitialized_variables_1/IsVariableInitialized_328" + input: "report_uninitialized_variables_1/IsVariableInitialized_329" + input: "report_uninitialized_variables_1/IsVariableInitialized_330" + input: "report_uninitialized_variables_1/IsVariableInitialized_331" + input: "report_uninitialized_variables_1/IsVariableInitialized_332" + input: "report_uninitialized_variables_1/IsVariableInitialized_333" + input: "report_uninitialized_variables_1/IsVariableInitialized_334" + input: "report_uninitialized_variables_1/IsVariableInitialized_335" + input: "report_uninitialized_variables_1/IsVariableInitialized_336" + input: "report_uninitialized_variables_1/IsVariableInitialized_337" + input: "report_uninitialized_variables_1/IsVariableInitialized_338" + input: "report_uninitialized_variables_1/IsVariableInitialized_339" + input: "report_uninitialized_variables_1/IsVariableInitialized_340" + input: "report_uninitialized_variables_1/IsVariableInitialized_341" + input: "report_uninitialized_variables_1/IsVariableInitialized_342" + input: "report_uninitialized_variables_1/IsVariableInitialized_343" + input: "report_uninitialized_variables_1/IsVariableInitialized_344" + input: "report_uninitialized_variables_1/IsVariableInitialized_345" + input: "report_uninitialized_variables_1/IsVariableInitialized_346" + input: "report_uninitialized_variables_1/IsVariableInitialized_347" + input: "report_uninitialized_variables_1/IsVariableInitialized_348" + input: "report_uninitialized_variables_1/IsVariableInitialized_349" + input: "report_uninitialized_variables_1/IsVariableInitialized_350" + input: "report_uninitialized_variables_1/IsVariableInitialized_351" + input: "report_uninitialized_variables_1/IsVariableInitialized_352" + input: "report_uninitialized_variables_1/IsVariableInitialized_353" + input: "report_uninitialized_variables_1/IsVariableInitialized_354" + input: "report_uninitialized_variables_1/IsVariableInitialized_355" + input: "report_uninitialized_variables_1/IsVariableInitialized_356" + input: "report_uninitialized_variables_1/IsVariableInitialized_357" + input: "report_uninitialized_variables_1/IsVariableInitialized_358" + input: "report_uninitialized_variables_1/IsVariableInitialized_359" + input: "report_uninitialized_variables_1/IsVariableInitialized_360" + input: "report_uninitialized_variables_1/IsVariableInitialized_361" + input: "report_uninitialized_variables_1/IsVariableInitialized_362" + input: "report_uninitialized_variables_1/IsVariableInitialized_363" + input: "report_uninitialized_variables_1/IsVariableInitialized_364" + input: "report_uninitialized_variables_1/IsVariableInitialized_365" + input: "report_uninitialized_variables_1/IsVariableInitialized_366" + input: "report_uninitialized_variables_1/IsVariableInitialized_367" + input: "report_uninitialized_variables_1/IsVariableInitialized_368" + input: "report_uninitialized_variables_1/IsVariableInitialized_369" + input: "report_uninitialized_variables_1/IsVariableInitialized_370" + input: "report_uninitialized_variables_1/IsVariableInitialized_371" + input: "report_uninitialized_variables_1/IsVariableInitialized_372" + input: "report_uninitialized_variables_1/IsVariableInitialized_373" + input: "report_uninitialized_variables_1/IsVariableInitialized_374" + input: "report_uninitialized_variables_1/IsVariableInitialized_375" + input: "report_uninitialized_variables_1/IsVariableInitialized_376" + input: "report_uninitialized_variables_1/IsVariableInitialized_377" + input: "report_uninitialized_variables_1/IsVariableInitialized_378" + input: "report_uninitialized_variables_1/IsVariableInitialized_379" + input: "report_uninitialized_variables_1/IsVariableInitialized_380" + input: "report_uninitialized_variables_1/IsVariableInitialized_381" + input: "report_uninitialized_variables_1/IsVariableInitialized_382" + input: "report_uninitialized_variables_1/IsVariableInitialized_383" + input: "report_uninitialized_variables_1/IsVariableInitialized_384" + input: "report_uninitialized_variables_1/IsVariableInitialized_385" + input: "report_uninitialized_variables_1/IsVariableInitialized_386" + input: "report_uninitialized_variables_1/IsVariableInitialized_387" + input: "report_uninitialized_variables_1/IsVariableInitialized_388" + input: "report_uninitialized_variables_1/IsVariableInitialized_389" + input: "report_uninitialized_variables_1/IsVariableInitialized_390" + input: "report_uninitialized_variables_1/IsVariableInitialized_391" + input: "report_uninitialized_variables_1/IsVariableInitialized_392" + input: "report_uninitialized_variables_1/IsVariableInitialized_393" + input: "report_uninitialized_variables_1/IsVariableInitialized_394" + input: "report_uninitialized_variables_1/IsVariableInitialized_395" + input: "report_uninitialized_variables_1/IsVariableInitialized_396" + input: "report_uninitialized_variables_1/IsVariableInitialized_397" + input: "report_uninitialized_variables_1/IsVariableInitialized_398" + input: "report_uninitialized_variables_1/IsVariableInitialized_399" + input: "report_uninitialized_variables_1/IsVariableInitialized_400" + input: "report_uninitialized_variables_1/IsVariableInitialized_401" + input: "report_uninitialized_variables_1/IsVariableInitialized_402" + input: "report_uninitialized_variables_1/IsVariableInitialized_403" + input: "report_uninitialized_variables_1/IsVariableInitialized_404" + input: "report_uninitialized_variables_1/IsVariableInitialized_405" + input: "report_uninitialized_variables_1/IsVariableInitialized_406" + input: "report_uninitialized_variables_1/IsVariableInitialized_407" + input: "report_uninitialized_variables_1/IsVariableInitialized_408" + input: "report_uninitialized_variables_1/IsVariableInitialized_409" + input: "report_uninitialized_variables_1/IsVariableInitialized_410" + input: "report_uninitialized_variables_1/IsVariableInitialized_411" + input: "report_uninitialized_variables_1/IsVariableInitialized_412" + input: "report_uninitialized_variables_1/IsVariableInitialized_413" + input: "report_uninitialized_variables_1/IsVariableInitialized_414" + input: "report_uninitialized_variables_1/IsVariableInitialized_415" + input: "report_uninitialized_variables_1/IsVariableInitialized_416" + input: "report_uninitialized_variables_1/IsVariableInitialized_417" + input: "report_uninitialized_variables_1/IsVariableInitialized_418" + input: "report_uninitialized_variables_1/IsVariableInitialized_419" + input: "report_uninitialized_variables_1/IsVariableInitialized_420" + input: "report_uninitialized_variables_1/IsVariableInitialized_421" + input: "report_uninitialized_variables_1/IsVariableInitialized_422" + input: "report_uninitialized_variables_1/IsVariableInitialized_423" + input: "report_uninitialized_variables_1/IsVariableInitialized_424" + input: "report_uninitialized_variables_1/IsVariableInitialized_425" + input: "report_uninitialized_variables_1/IsVariableInitialized_426" + input: "report_uninitialized_variables_1/IsVariableInitialized_427" + input: "report_uninitialized_variables_1/IsVariableInitialized_428" + input: "report_uninitialized_variables_1/IsVariableInitialized_429" + input: "report_uninitialized_variables_1/IsVariableInitialized_430" + input: "report_uninitialized_variables_1/IsVariableInitialized_431" + input: "report_uninitialized_variables_1/IsVariableInitialized_432" + input: "report_uninitialized_variables_1/IsVariableInitialized_433" + input: "report_uninitialized_variables_1/IsVariableInitialized_434" + input: "report_uninitialized_variables_1/IsVariableInitialized_435" + input: "report_uninitialized_variables_1/IsVariableInitialized_436" + input: "report_uninitialized_variables_1/IsVariableInitialized_437" + input: "report_uninitialized_variables_1/IsVariableInitialized_438" + input: "report_uninitialized_variables_1/IsVariableInitialized_439" + input: "report_uninitialized_variables_1/IsVariableInitialized_440" + input: "report_uninitialized_variables_1/IsVariableInitialized_441" + input: "report_uninitialized_variables_1/IsVariableInitialized_442" + input: "report_uninitialized_variables_1/IsVariableInitialized_443" + input: "report_uninitialized_variables_1/IsVariableInitialized_444" + input: "report_uninitialized_variables_1/IsVariableInitialized_445" + input: "report_uninitialized_variables_1/IsVariableInitialized_446" + input: "report_uninitialized_variables_1/IsVariableInitialized_447" + input: "report_uninitialized_variables_1/IsVariableInitialized_448" + input: "report_uninitialized_variables_1/IsVariableInitialized_449" + input: "report_uninitialized_variables_1/IsVariableInitialized_450" + input: "report_uninitialized_variables_1/IsVariableInitialized_451" + input: "report_uninitialized_variables_1/IsVariableInitialized_452" + input: "report_uninitialized_variables_1/IsVariableInitialized_453" + input: "report_uninitialized_variables_1/IsVariableInitialized_454" + input: "report_uninitialized_variables_1/IsVariableInitialized_455" + input: "report_uninitialized_variables_1/IsVariableInitialized_456" + input: "report_uninitialized_variables_1/IsVariableInitialized_457" + input: "report_uninitialized_variables_1/IsVariableInitialized_458" + input: "report_uninitialized_variables_1/IsVariableInitialized_459" + input: "report_uninitialized_variables_1/IsVariableInitialized_460" + input: "report_uninitialized_variables_1/IsVariableInitialized_461" + input: "report_uninitialized_variables_1/IsVariableInitialized_462" + input: "report_uninitialized_variables_1/IsVariableInitialized_463" + input: "report_uninitialized_variables_1/IsVariableInitialized_464" + input: "report_uninitialized_variables_1/IsVariableInitialized_465" + input: "report_uninitialized_variables_1/IsVariableInitialized_466" + input: "report_uninitialized_variables_1/IsVariableInitialized_467" + input: "report_uninitialized_variables_1/IsVariableInitialized_468" + input: "report_uninitialized_variables_1/IsVariableInitialized_469" + input: "report_uninitialized_variables_1/IsVariableInitialized_470" + input: "report_uninitialized_variables_1/IsVariableInitialized_471" + input: "report_uninitialized_variables_1/IsVariableInitialized_472" + input: "report_uninitialized_variables_1/IsVariableInitialized_473" + input: "report_uninitialized_variables_1/IsVariableInitialized_474" + input: "report_uninitialized_variables_1/IsVariableInitialized_475" + input: "report_uninitialized_variables_1/IsVariableInitialized_476" + input: "report_uninitialized_variables_1/IsVariableInitialized_477" + input: "report_uninitialized_variables_1/IsVariableInitialized_478" + input: "report_uninitialized_variables_1/IsVariableInitialized_479" + input: "report_uninitialized_variables_1/IsVariableInitialized_480" + input: "report_uninitialized_variables_1/IsVariableInitialized_481" + input: "report_uninitialized_variables_1/IsVariableInitialized_482" + input: "report_uninitialized_variables_1/IsVariableInitialized_483" + input: "report_uninitialized_variables_1/IsVariableInitialized_484" + input: "report_uninitialized_variables_1/IsVariableInitialized_485" + input: "report_uninitialized_variables_1/IsVariableInitialized_486" + input: "report_uninitialized_variables_1/IsVariableInitialized_487" + input: "report_uninitialized_variables_1/IsVariableInitialized_488" + input: "report_uninitialized_variables_1/IsVariableInitialized_489" + input: "report_uninitialized_variables_1/IsVariableInitialized_490" + input: "report_uninitialized_variables_1/IsVariableInitialized_491" + input: "report_uninitialized_variables_1/IsVariableInitialized_492" + input: "report_uninitialized_variables_1/IsVariableInitialized_493" + input: "report_uninitialized_variables_1/IsVariableInitialized_494" + input: "report_uninitialized_variables_1/IsVariableInitialized_495" + input: "report_uninitialized_variables_1/IsVariableInitialized_496" + input: "report_uninitialized_variables_1/IsVariableInitialized_497" + input: "report_uninitialized_variables_1/IsVariableInitialized_498" + input: "report_uninitialized_variables_1/IsVariableInitialized_499" + input: "report_uninitialized_variables_1/IsVariableInitialized_500" + input: "report_uninitialized_variables_1/IsVariableInitialized_501" + input: "report_uninitialized_variables_1/IsVariableInitialized_502" + input: "report_uninitialized_variables_1/IsVariableInitialized_503" + input: "report_uninitialized_variables_1/IsVariableInitialized_504" + input: "report_uninitialized_variables_1/IsVariableInitialized_505" + input: "report_uninitialized_variables_1/IsVariableInitialized_506" + input: "report_uninitialized_variables_1/IsVariableInitialized_507" + input: "report_uninitialized_variables_1/IsVariableInitialized_508" + input: "report_uninitialized_variables_1/IsVariableInitialized_509" + input: "report_uninitialized_variables_1/IsVariableInitialized_510" + input: "report_uninitialized_variables_1/IsVariableInitialized_511" + input: "report_uninitialized_variables_1/IsVariableInitialized_512" + input: "report_uninitialized_variables_1/IsVariableInitialized_513" + input: "report_uninitialized_variables_1/IsVariableInitialized_514" + input: "report_uninitialized_variables_1/IsVariableInitialized_515" + input: "report_uninitialized_variables_1/IsVariableInitialized_516" + input: "report_uninitialized_variables_1/IsVariableInitialized_517" + input: "report_uninitialized_variables_1/IsVariableInitialized_518" + input: "report_uninitialized_variables_1/IsVariableInitialized_519" + input: "report_uninitialized_variables_1/IsVariableInitialized_520" + input: "report_uninitialized_variables_1/IsVariableInitialized_521" + input: "report_uninitialized_variables_1/IsVariableInitialized_522" + input: "report_uninitialized_variables_1/IsVariableInitialized_523" + input: "report_uninitialized_variables_1/IsVariableInitialized_524" + input: "report_uninitialized_variables_1/IsVariableInitialized_525" + input: "report_uninitialized_variables_1/IsVariableInitialized_526" + input: "report_uninitialized_variables_1/IsVariableInitialized_527" + input: "report_uninitialized_variables_1/IsVariableInitialized_528" + input: "report_uninitialized_variables_1/IsVariableInitialized_529" + input: "report_uninitialized_variables_1/IsVariableInitialized_530" + input: "report_uninitialized_variables_1/IsVariableInitialized_531" + input: "report_uninitialized_variables_1/IsVariableInitialized_532" + input: "report_uninitialized_variables_1/IsVariableInitialized_533" + input: "report_uninitialized_variables_1/IsVariableInitialized_534" + input: "report_uninitialized_variables_1/IsVariableInitialized_535" + input: "report_uninitialized_variables_1/IsVariableInitialized_536" + input: "report_uninitialized_variables_1/IsVariableInitialized_537" + input: "report_uninitialized_variables_1/IsVariableInitialized_538" + input: "report_uninitialized_variables_1/IsVariableInitialized_539" + input: "report_uninitialized_variables_1/IsVariableInitialized_540" + input: "report_uninitialized_variables_1/IsVariableInitialized_541" + input: "report_uninitialized_variables_1/IsVariableInitialized_542" + input: "report_uninitialized_variables_1/IsVariableInitialized_543" + input: "report_uninitialized_variables_1/IsVariableInitialized_544" + input: "report_uninitialized_variables_1/IsVariableInitialized_545" + input: "report_uninitialized_variables_1/IsVariableInitialized_546" + input: "report_uninitialized_variables_1/IsVariableInitialized_547" + input: "report_uninitialized_variables_1/IsVariableInitialized_548" + input: "report_uninitialized_variables_1/IsVariableInitialized_549" + input: "report_uninitialized_variables_1/IsVariableInitialized_550" + input: "report_uninitialized_variables_1/IsVariableInitialized_551" + input: "report_uninitialized_variables_1/IsVariableInitialized_552" + input: "report_uninitialized_variables_1/IsVariableInitialized_553" + input: "report_uninitialized_variables_1/IsVariableInitialized_554" + input: "report_uninitialized_variables_1/IsVariableInitialized_555" + input: "report_uninitialized_variables_1/IsVariableInitialized_556" + input: "report_uninitialized_variables_1/IsVariableInitialized_557" + input: "report_uninitialized_variables_1/IsVariableInitialized_558" + input: "report_uninitialized_variables_1/IsVariableInitialized_559" + input: "report_uninitialized_variables_1/IsVariableInitialized_560" + input: "report_uninitialized_variables_1/IsVariableInitialized_561" + input: "report_uninitialized_variables_1/IsVariableInitialized_562" + input: "report_uninitialized_variables_1/IsVariableInitialized_563" + input: "report_uninitialized_variables_1/IsVariableInitialized_564" + input: "report_uninitialized_variables_1/IsVariableInitialized_565" + input: "report_uninitialized_variables_1/IsVariableInitialized_566" + input: "report_uninitialized_variables_1/IsVariableInitialized_567" + input: "report_uninitialized_variables_1/IsVariableInitialized_568" + input: "report_uninitialized_variables_1/IsVariableInitialized_569" + input: "report_uninitialized_variables_1/IsVariableInitialized_570" + input: "report_uninitialized_variables_1/IsVariableInitialized_571" + input: "report_uninitialized_variables_1/IsVariableInitialized_572" + input: "report_uninitialized_variables_1/IsVariableInitialized_573" + input: "report_uninitialized_variables_1/IsVariableInitialized_574" + input: "report_uninitialized_variables_1/IsVariableInitialized_575" + input: "report_uninitialized_variables_1/IsVariableInitialized_576" + input: "report_uninitialized_variables_1/IsVariableInitialized_577" + input: "report_uninitialized_variables_1/IsVariableInitialized_578" + input: "report_uninitialized_variables_1/IsVariableInitialized_579" + input: "report_uninitialized_variables_1/IsVariableInitialized_580" + input: "report_uninitialized_variables_1/IsVariableInitialized_581" + input: "report_uninitialized_variables_1/IsVariableInitialized_582" + input: "report_uninitialized_variables_1/IsVariableInitialized_583" + input: "report_uninitialized_variables_1/IsVariableInitialized_584" + input: "report_uninitialized_variables_1/IsVariableInitialized_585" + input: "report_uninitialized_variables_1/IsVariableInitialized_586" + input: "report_uninitialized_variables_1/IsVariableInitialized_587" + input: "report_uninitialized_variables_1/IsVariableInitialized_588" + input: "report_uninitialized_variables_1/IsVariableInitialized_589" + input: "report_uninitialized_variables_1/IsVariableInitialized_590" + input: "report_uninitialized_variables_1/IsVariableInitialized_591" + input: "report_uninitialized_variables_1/IsVariableInitialized_592" + input: "report_uninitialized_variables_1/IsVariableInitialized_593" + input: "report_uninitialized_variables_1/IsVariableInitialized_594" + input: "report_uninitialized_variables_1/IsVariableInitialized_595" + input: "report_uninitialized_variables_1/IsVariableInitialized_596" + input: "report_uninitialized_variables_1/IsVariableInitialized_597" + input: "report_uninitialized_variables_1/IsVariableInitialized_598" + input: "report_uninitialized_variables_1/IsVariableInitialized_599" + input: "report_uninitialized_variables_1/IsVariableInitialized_600" + input: "report_uninitialized_variables_1/IsVariableInitialized_601" + input: "report_uninitialized_variables_1/IsVariableInitialized_602" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 604 + } + } + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables_1/LogicalNot" + op: "LogicalNot" + input: "report_uninitialized_variables_1/stack" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/Const" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "global_step" + string_val: "bert/embeddings/word_embeddings" + string_val: "bert/embeddings/token_type_embeddings" + string_val: "bert/embeddings/position_embeddings" + string_val: "bert/embeddings/LayerNorm/beta" + string_val: "bert/embeddings/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/attention/self/query/kernel" + string_val: "bert/encoder/layer_0/attention/self/query/bias" + string_val: "bert/encoder/layer_0/attention/self/key/kernel" + string_val: "bert/encoder/layer_0/attention/self/key/bias" + string_val: "bert/encoder/layer_0/attention/self/value/kernel" + string_val: "bert/encoder/layer_0/attention/self/value/bias" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel" + string_val: "bert/encoder/layer_0/attention/output/dense/bias" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel" + string_val: "bert/encoder/layer_0/intermediate/dense/bias" + string_val: "bert/encoder/layer_0/output/dense/kernel" + string_val: "bert/encoder/layer_0/output/dense/bias" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/attention/self/query/kernel" + string_val: "bert/encoder/layer_1/attention/self/query/bias" + string_val: "bert/encoder/layer_1/attention/self/key/kernel" + string_val: "bert/encoder/layer_1/attention/self/key/bias" + string_val: "bert/encoder/layer_1/attention/self/value/kernel" + string_val: "bert/encoder/layer_1/attention/self/value/bias" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel" + string_val: "bert/encoder/layer_1/attention/output/dense/bias" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel" + string_val: "bert/encoder/layer_1/intermediate/dense/bias" + string_val: "bert/encoder/layer_1/output/dense/kernel" + string_val: "bert/encoder/layer_1/output/dense/bias" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/attention/self/query/kernel" + string_val: "bert/encoder/layer_2/attention/self/query/bias" + string_val: "bert/encoder/layer_2/attention/self/key/kernel" + string_val: "bert/encoder/layer_2/attention/self/key/bias" + string_val: "bert/encoder/layer_2/attention/self/value/kernel" + string_val: "bert/encoder/layer_2/attention/self/value/bias" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel" + string_val: "bert/encoder/layer_2/attention/output/dense/bias" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel" + string_val: "bert/encoder/layer_2/intermediate/dense/bias" + string_val: "bert/encoder/layer_2/output/dense/kernel" + string_val: "bert/encoder/layer_2/output/dense/bias" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/attention/self/query/kernel" + string_val: "bert/encoder/layer_3/attention/self/query/bias" + string_val: "bert/encoder/layer_3/attention/self/key/kernel" + string_val: "bert/encoder/layer_3/attention/self/key/bias" + string_val: "bert/encoder/layer_3/attention/self/value/kernel" + string_val: "bert/encoder/layer_3/attention/self/value/bias" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel" + string_val: "bert/encoder/layer_3/attention/output/dense/bias" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel" + string_val: "bert/encoder/layer_3/intermediate/dense/bias" + string_val: "bert/encoder/layer_3/output/dense/kernel" + string_val: "bert/encoder/layer_3/output/dense/bias" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/attention/self/query/kernel" + string_val: "bert/encoder/layer_4/attention/self/query/bias" + string_val: "bert/encoder/layer_4/attention/self/key/kernel" + string_val: "bert/encoder/layer_4/attention/self/key/bias" + string_val: "bert/encoder/layer_4/attention/self/value/kernel" + string_val: "bert/encoder/layer_4/attention/self/value/bias" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel" + string_val: "bert/encoder/layer_4/attention/output/dense/bias" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel" + string_val: "bert/encoder/layer_4/intermediate/dense/bias" + string_val: "bert/encoder/layer_4/output/dense/kernel" + string_val: "bert/encoder/layer_4/output/dense/bias" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/attention/self/query/kernel" + string_val: "bert/encoder/layer_5/attention/self/query/bias" + string_val: "bert/encoder/layer_5/attention/self/key/kernel" + string_val: "bert/encoder/layer_5/attention/self/key/bias" + string_val: "bert/encoder/layer_5/attention/self/value/kernel" + string_val: "bert/encoder/layer_5/attention/self/value/bias" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel" + string_val: "bert/encoder/layer_5/attention/output/dense/bias" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel" + string_val: "bert/encoder/layer_5/intermediate/dense/bias" + string_val: "bert/encoder/layer_5/output/dense/kernel" + string_val: "bert/encoder/layer_5/output/dense/bias" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/attention/self/query/kernel" + string_val: "bert/encoder/layer_6/attention/self/query/bias" + string_val: "bert/encoder/layer_6/attention/self/key/kernel" + string_val: "bert/encoder/layer_6/attention/self/key/bias" + string_val: "bert/encoder/layer_6/attention/self/value/kernel" + string_val: "bert/encoder/layer_6/attention/self/value/bias" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel" + string_val: "bert/encoder/layer_6/attention/output/dense/bias" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel" + string_val: "bert/encoder/layer_6/intermediate/dense/bias" + string_val: "bert/encoder/layer_6/output/dense/kernel" + string_val: "bert/encoder/layer_6/output/dense/bias" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/attention/self/query/kernel" + string_val: "bert/encoder/layer_7/attention/self/query/bias" + string_val: "bert/encoder/layer_7/attention/self/key/kernel" + string_val: "bert/encoder/layer_7/attention/self/key/bias" + string_val: "bert/encoder/layer_7/attention/self/value/kernel" + string_val: "bert/encoder/layer_7/attention/self/value/bias" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel" + string_val: "bert/encoder/layer_7/attention/output/dense/bias" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel" + string_val: "bert/encoder/layer_7/intermediate/dense/bias" + string_val: "bert/encoder/layer_7/output/dense/kernel" + string_val: "bert/encoder/layer_7/output/dense/bias" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/attention/self/query/kernel" + string_val: "bert/encoder/layer_8/attention/self/query/bias" + string_val: "bert/encoder/layer_8/attention/self/key/kernel" + string_val: "bert/encoder/layer_8/attention/self/key/bias" + string_val: "bert/encoder/layer_8/attention/self/value/kernel" + string_val: "bert/encoder/layer_8/attention/self/value/bias" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel" + string_val: "bert/encoder/layer_8/attention/output/dense/bias" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel" + string_val: "bert/encoder/layer_8/intermediate/dense/bias" + string_val: "bert/encoder/layer_8/output/dense/kernel" + string_val: "bert/encoder/layer_8/output/dense/bias" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/attention/self/query/kernel" + string_val: "bert/encoder/layer_9/attention/self/query/bias" + string_val: "bert/encoder/layer_9/attention/self/key/kernel" + string_val: "bert/encoder/layer_9/attention/self/key/bias" + string_val: "bert/encoder/layer_9/attention/self/value/kernel" + string_val: "bert/encoder/layer_9/attention/self/value/bias" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel" + string_val: "bert/encoder/layer_9/attention/output/dense/bias" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel" + string_val: "bert/encoder/layer_9/intermediate/dense/bias" + string_val: "bert/encoder/layer_9/output/dense/kernel" + string_val: "bert/encoder/layer_9/output/dense/bias" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/attention/self/query/kernel" + string_val: "bert/encoder/layer_10/attention/self/query/bias" + string_val: "bert/encoder/layer_10/attention/self/key/kernel" + string_val: "bert/encoder/layer_10/attention/self/key/bias" + string_val: "bert/encoder/layer_10/attention/self/value/kernel" + string_val: "bert/encoder/layer_10/attention/self/value/bias" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel" + string_val: "bert/encoder/layer_10/attention/output/dense/bias" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel" + string_val: "bert/encoder/layer_10/intermediate/dense/bias" + string_val: "bert/encoder/layer_10/output/dense/kernel" + string_val: "bert/encoder/layer_10/output/dense/bias" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/attention/self/query/kernel" + string_val: "bert/encoder/layer_11/attention/self/query/bias" + string_val: "bert/encoder/layer_11/attention/self/key/kernel" + string_val: "bert/encoder/layer_11/attention/self/key/bias" + string_val: "bert/encoder/layer_11/attention/self/value/kernel" + string_val: "bert/encoder/layer_11/attention/self/value/bias" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel" + string_val: "bert/encoder/layer_11/attention/output/dense/bias" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel" + string_val: "bert/encoder/layer_11/intermediate/dense/bias" + string_val: "bert/encoder/layer_11/output/dense/kernel" + string_val: "bert/encoder/layer_11/output/dense/bias" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma" + string_val: "bert/pooler/dense/kernel" + string_val: "bert/pooler/dense/bias" + string_val: "output_weights" + string_val: "output_bias" + string_val: "bert/embeddings/word_embeddings/adam_m" + string_val: "bert/embeddings/word_embeddings/adam_v" + string_val: "bert/embeddings/token_type_embeddings/adam_m" + string_val: "bert/embeddings/token_type_embeddings/adam_v" + string_val: "bert/embeddings/position_embeddings/adam_m" + string_val: "bert/embeddings/position_embeddings/adam_v" + string_val: "bert/embeddings/LayerNorm/beta/adam_m" + string_val: "bert/embeddings/LayerNorm/beta/adam_v" + string_val: "bert/embeddings/LayerNorm/gamma/adam_m" + string_val: "bert/embeddings/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + string_val: "bert/pooler/dense/kernel/adam_m" + string_val: "bert/pooler/dense/kernel/adam_v" + string_val: "bert/pooler/dense/bias/adam_m" + string_val: "bert/pooler/dense/bias/adam_v" + string_val: "output_weights/adam_m" + string_val: "output_weights/adam_v" + string_val: "output_bias/adam_m" + string_val: "output_bias/adam_v" + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Shape" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice" + op: "StridedSlice" + input: "report_uninitialized_variables_1/boolean_mask/Shape" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack_1" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Prod/reduction_indices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Prod" + op: "Prod" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice" + input: "report_uninitialized_variables_1/boolean_mask/Prod/reduction_indices" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "keep_dims" + value { + b: false + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Shape_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_1" + op: "StridedSlice" + input: "report_uninitialized_variables_1/boolean_mask/Shape_1" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack_1" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_1/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 1 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 0 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Shape_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 604 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack_1" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack_2" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/strided_slice_2" + op: "StridedSlice" + input: "report_uninitialized_variables_1/boolean_mask/Shape_2" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack_1" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_2/stack_2" + device: "/device:CPU:0" + attr { + key: "Index" + value { + type: DT_INT32 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "begin_mask" + value { + i: 0 + } + } + attr { + key: "ellipsis_mask" + value { + i: 0 + } + } + attr { + key: "end_mask" + value { + i: 1 + } + } + attr { + key: "new_axis_mask" + value { + i: 0 + } + } + attr { + key: "shrink_axis_mask" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/concat/values_1" + op: "Pack" + input: "report_uninitialized_variables_1/boolean_mask/Prod" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 1 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/concat/axis" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/concat" + op: "ConcatV2" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_1" + input: "report_uninitialized_variables_1/boolean_mask/concat/values_1" + input: "report_uninitialized_variables_1/boolean_mask/strided_slice_2" + input: "report_uninitialized_variables_1/boolean_mask/concat/axis" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 3 + } + } + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Reshape" + op: "Reshape" + input: "report_uninitialized_variables_1/Const" + input: "report_uninitialized_variables_1/boolean_mask/concat" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Reshape_1/shape" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + dim { + size: 1 + } + } + int_val: -1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Reshape_1" + op: "Reshape" + input: "report_uninitialized_variables_1/LogicalNot" + input: "report_uninitialized_variables_1/boolean_mask/Reshape_1/shape" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "Tshape" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Where" + op: "Where" + input: "report_uninitialized_variables_1/boolean_mask/Reshape_1" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_BOOL + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + dim { + size: 1 + } + } + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/Squeeze" + op: "Squeeze" + input: "report_uninitialized_variables_1/boolean_mask/Where" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } + attr { + key: "squeeze_dims" + value { + list { + i: 1 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/GatherV2/axis" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "report_uninitialized_variables_1/boolean_mask/GatherV2" + op: "GatherV2" + input: "report_uninitialized_variables_1/boolean_mask/Reshape" + input: "report_uninitialized_variables_1/boolean_mask/Squeeze" + input: "report_uninitialized_variables_1/boolean_mask/GatherV2/axis" + device: "/device:CPU:0" + attr { + key: "Taxis" + value { + type: DT_INT32 + } + } + attr { + key: "Tindices" + value { + type: DT_INT64 + } + } + attr { + key: "Tparams" + value { + type: DT_STRING + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } + attr { + key: "batch_dims" + value { + i: 0 + } + } +} +node { + name: "report_uninitialized_resources_1/Const" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + } + } + } + } + } +} +node { + name: "concat_1/axis" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "concat_1" + op: "ConcatV2" + input: "report_uninitialized_variables_1/boolean_mask/GatherV2" + input: "report_uninitialized_resources_1/Const" + input: "concat_1/axis" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "Tidx" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: -1 + } + } + } + } + } +} +node { + name: "init_2" + op: "NoOp" +} +node { + name: "init_all_tables" + op: "NoOp" +} +node { + name: "init_3" + op: "NoOp" +} +node { + name: "group_deps_3" + op: "NoOp" + input: "^init_2" + input: "^init_3" + input: "^init_all_tables" +} +node { + name: "Merge/MergeSummary" + op: "MergeSummary" + input: "loss_1" + attr { + key: "N" + value { + i: 1 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "save/filename/input" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "model" + } + } + } +} +node { + name: "save/filename" + op: "PlaceholderWithDefault" + input: "save/filename/input" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "shape" + value { + shape { + } + } + } +} +node { + name: "save/Const" + op: "PlaceholderWithDefault" + input: "save/filename" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "shape" + value { + shape { + } + } + } +} +node { + name: "save/StringJoin/inputs_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "_temp_823f412404fa4e59893c248d5a436737/part" + } + } + } +} +node { + name: "save/StringJoin" + op: "StringJoin" + input: "save/Const" + input: "save/StringJoin/inputs_1" + attr { + key: "N" + value { + i: 2 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "separator" + value { + s: "" + } + } +} +node { + name: "save/num_shards" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 1 + } + } + } +} +node { + name: "save/ShardedFilename/shard" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT32 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT32 + tensor_shape { + } + int_val: 0 + } + } + } +} +node { + name: "save/ShardedFilename" + op: "ShardedFilename" + input: "save/StringJoin" + input: "save/ShardedFilename/shard" + input: "save/num_shards" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "save/SaveV2/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "bert/embeddings/LayerNorm/beta" + string_val: "bert/embeddings/LayerNorm/beta/adam_m" + string_val: "bert/embeddings/LayerNorm/beta/adam_v" + string_val: "bert/embeddings/LayerNorm/gamma" + string_val: "bert/embeddings/LayerNorm/gamma/adam_m" + string_val: "bert/embeddings/LayerNorm/gamma/adam_v" + string_val: "bert/embeddings/position_embeddings" + string_val: "bert/embeddings/position_embeddings/adam_m" + string_val: "bert/embeddings/position_embeddings/adam_v" + string_val: "bert/embeddings/token_type_embeddings" + string_val: "bert/embeddings/token_type_embeddings/adam_m" + string_val: "bert/embeddings/token_type_embeddings/adam_v" + string_val: "bert/embeddings/word_embeddings" + string_val: "bert/embeddings/word_embeddings/adam_m" + string_val: "bert/embeddings/word_embeddings/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/bias" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/bias" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/kernel" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/bias" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/kernel" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/bias" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/kernel" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/bias" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/output/dense/bias" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/dense/kernel" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/bias" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/bias" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/kernel" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/bias" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/kernel" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/bias" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/kernel" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/bias" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/output/dense/bias" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/dense/kernel" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/bias" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/bias" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/kernel" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/bias" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/kernel" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/bias" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/kernel" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/bias" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/output/dense/bias" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/dense/kernel" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/bias" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/bias" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/kernel" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/bias" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/kernel" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/bias" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/kernel" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/bias" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/output/dense/bias" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/dense/kernel" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/bias" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/bias" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/kernel" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/bias" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/kernel" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/bias" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/kernel" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/bias" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/output/dense/bias" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/dense/kernel" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/bias" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/bias" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/kernel" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/bias" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/kernel" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/bias" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/kernel" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/bias" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/output/dense/bias" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/dense/kernel" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/bias" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/bias" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/kernel" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/bias" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/kernel" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/bias" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/kernel" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/bias" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/output/dense/bias" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/dense/kernel" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/bias" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/bias" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/kernel" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/bias" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/kernel" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/bias" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/kernel" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/bias" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/output/dense/bias" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/dense/kernel" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/bias" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/bias" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/kernel" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/bias" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/kernel" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/bias" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/kernel" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/bias" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/output/dense/bias" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/dense/kernel" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/bias" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/bias" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/kernel" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/bias" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/kernel" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/bias" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/kernel" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/bias" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/output/dense/bias" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/dense/kernel" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/bias" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/bias" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/kernel" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/bias" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/kernel" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/bias" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/kernel" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/bias" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/output/dense/bias" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/dense/kernel" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/bias" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/bias" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/kernel" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/bias" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/kernel" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/bias" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/kernel" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/bias" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/output/dense/bias" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/dense/kernel" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_v" + string_val: "bert/pooler/dense/bias" + string_val: "bert/pooler/dense/bias/adam_m" + string_val: "bert/pooler/dense/bias/adam_v" + string_val: "bert/pooler/dense/kernel" + string_val: "bert/pooler/dense/kernel/adam_m" + string_val: "bert/pooler/dense/kernel/adam_v" + string_val: "global_step" + string_val: "output_bias" + string_val: "output_bias/adam_m" + string_val: "output_bias/adam_v" + string_val: "output_weights" + string_val: "output_weights/adam_m" + string_val: "output_weights/adam_v" + } + } + } +} +node { + name: "save/SaveV2/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + } + } + } +} +node { + name: "save/SaveV2" + op: "SaveV2" + input: "save/ShardedFilename" + input: "save/SaveV2/tensor_names" + input: "save/SaveV2/shape_and_slices" + input: "bert/embeddings/LayerNorm/beta" + input: "bert/embeddings/LayerNorm/beta/adam_m" + input: "bert/embeddings/LayerNorm/beta/adam_v" + input: "bert/embeddings/LayerNorm/gamma" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + input: "bert/embeddings/position_embeddings" + input: "bert/embeddings/position_embeddings/adam_m" + input: "bert/embeddings/position_embeddings/adam_v" + input: "bert/embeddings/token_type_embeddings" + input: "bert/embeddings/token_type_embeddings/adam_m" + input: "bert/embeddings/token_type_embeddings/adam_v" + input: "bert/embeddings/word_embeddings" + input: "bert/embeddings/word_embeddings/adam_m" + input: "bert/embeddings/word_embeddings/adam_v" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_0/attention/output/dense/bias" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/key/bias" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/key/kernel" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/query/bias" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/query/kernel" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_0/attention/self/value/bias" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_0/attention/self/value/kernel" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_0/intermediate/dense/bias" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_0/output/dense/bias" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + input: "bert/encoder/layer_0/output/dense/kernel" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_1/attention/output/dense/bias" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/key/bias" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/key/kernel" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/query/bias" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/query/kernel" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_1/attention/self/value/bias" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_1/attention/self/value/kernel" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_1/intermediate/dense/bias" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_1/output/dense/bias" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + input: "bert/encoder/layer_1/output/dense/kernel" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_10/attention/output/dense/bias" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/key/bias" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/key/kernel" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/query/bias" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/query/kernel" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_10/attention/self/value/bias" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_10/attention/self/value/kernel" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_10/intermediate/dense/bias" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_10/output/dense/bias" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + input: "bert/encoder/layer_10/output/dense/kernel" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_11/attention/output/dense/bias" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/key/bias" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/key/kernel" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/query/bias" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/query/kernel" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_11/attention/self/value/bias" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_11/attention/self/value/kernel" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_11/intermediate/dense/bias" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_11/output/dense/bias" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + input: "bert/encoder/layer_11/output/dense/kernel" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_2/attention/output/dense/bias" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/key/bias" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/key/kernel" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/query/bias" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/query/kernel" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_2/attention/self/value/bias" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_2/attention/self/value/kernel" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_2/intermediate/dense/bias" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_2/output/dense/bias" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + input: "bert/encoder/layer_2/output/dense/kernel" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_3/attention/output/dense/bias" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/key/bias" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/key/kernel" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/query/bias" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/query/kernel" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_3/attention/self/value/bias" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_3/attention/self/value/kernel" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_3/intermediate/dense/bias" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_3/output/dense/bias" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + input: "bert/encoder/layer_3/output/dense/kernel" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_4/attention/output/dense/bias" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/key/bias" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/key/kernel" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/query/bias" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/query/kernel" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_4/attention/self/value/bias" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_4/attention/self/value/kernel" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_4/intermediate/dense/bias" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_4/output/dense/bias" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + input: "bert/encoder/layer_4/output/dense/kernel" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_5/attention/output/dense/bias" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/key/bias" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/key/kernel" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/query/bias" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/query/kernel" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_5/attention/self/value/bias" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_5/attention/self/value/kernel" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_5/intermediate/dense/bias" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_5/output/dense/bias" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + input: "bert/encoder/layer_5/output/dense/kernel" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_6/attention/output/dense/bias" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/key/bias" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/key/kernel" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/query/bias" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/query/kernel" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_6/attention/self/value/bias" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_6/attention/self/value/kernel" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_6/intermediate/dense/bias" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_6/output/dense/bias" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + input: "bert/encoder/layer_6/output/dense/kernel" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_7/attention/output/dense/bias" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/key/bias" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/key/kernel" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/query/bias" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/query/kernel" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_7/attention/self/value/bias" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_7/attention/self/value/kernel" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_7/intermediate/dense/bias" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_7/output/dense/bias" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + input: "bert/encoder/layer_7/output/dense/kernel" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_8/attention/output/dense/bias" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/key/bias" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/key/kernel" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/query/bias" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/query/kernel" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_8/attention/self/value/bias" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_8/attention/self/value/kernel" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_8/intermediate/dense/bias" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_8/output/dense/bias" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + input: "bert/encoder/layer_8/output/dense/kernel" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_9/attention/output/dense/bias" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/key/bias" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/key/kernel" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/query/bias" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/query/kernel" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + input: "bert/encoder/layer_9/attention/self/value/bias" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + input: "bert/encoder/layer_9/attention/self/value/kernel" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + input: "bert/encoder/layer_9/intermediate/dense/bias" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + input: "bert/encoder/layer_9/output/dense/bias" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + input: "bert/encoder/layer_9/output/dense/kernel" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + input: "bert/pooler/dense/bias" + input: "bert/pooler/dense/bias/adam_m" + input: "bert/pooler/dense/bias/adam_v" + input: "bert/pooler/dense/kernel" + input: "bert/pooler/dense/kernel/adam_m" + input: "bert/pooler/dense/kernel/adam_v" + input: "global_step/Read/ReadVariableOp" + input: "output_bias" + input: "output_bias/adam_m" + input: "output_bias/adam_v" + input: "output_weights" + input: "output_weights/adam_m" + input: "output_weights/adam_v" + device: "/device:CPU:0" + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_INT64 + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + } + } + } +} +node { + name: "save/control_dependency" + op: "Identity" + input: "save/ShardedFilename" + input: "^save/SaveV2" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "_class" + value { + list { + s: "loc:@save/ShardedFilename" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "save/MergeV2Checkpoints/checkpoint_prefixes" + op: "Pack" + input: "save/ShardedFilename" + input: "^save/control_dependency" + device: "/device:CPU:0" + attr { + key: "N" + value { + i: 1 + } + } + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 1 + } + } + } + } + } + attr { + key: "axis" + value { + i: 0 + } + } +} +node { + name: "save/MergeV2Checkpoints" + op: "MergeV2Checkpoints" + input: "save/MergeV2Checkpoints/checkpoint_prefixes" + input: "save/Const" + device: "/device:CPU:0" + attr { + key: "delete_old_dirs" + value { + b: true + } + } +} +node { + name: "save/Identity" + op: "Identity" + input: "save/Const" + input: "^save/MergeV2Checkpoints" + input: "^save/control_dependency" + device: "/device:CPU:0" + attr { + key: "T" + value { + type: DT_STRING + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } +} +node { + name: "save/RestoreV2/tensor_names" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "bert/embeddings/LayerNorm/beta" + string_val: "bert/embeddings/LayerNorm/beta/adam_m" + string_val: "bert/embeddings/LayerNorm/beta/adam_v" + string_val: "bert/embeddings/LayerNorm/gamma" + string_val: "bert/embeddings/LayerNorm/gamma/adam_m" + string_val: "bert/embeddings/LayerNorm/gamma/adam_v" + string_val: "bert/embeddings/position_embeddings" + string_val: "bert/embeddings/position_embeddings/adam_m" + string_val: "bert/embeddings/position_embeddings/adam_v" + string_val: "bert/embeddings/token_type_embeddings" + string_val: "bert/embeddings/token_type_embeddings/adam_m" + string_val: "bert/embeddings/token_type_embeddings/adam_v" + string_val: "bert/embeddings/word_embeddings" + string_val: "bert/embeddings/word_embeddings/adam_m" + string_val: "bert/embeddings/word_embeddings/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/bias" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/bias" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/key/kernel" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/bias" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/query/kernel" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/bias" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_0/attention/self/value/kernel" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/bias" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_0/output/dense/bias" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_0/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_0/output/dense/kernel" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_0/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/bias" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/bias" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/key/kernel" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/bias" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/query/kernel" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/bias" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_1/attention/self/value/kernel" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/bias" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_1/output/dense/bias" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_1/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_1/output/dense/kernel" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_1/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/bias" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/bias" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/key/kernel" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/bias" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/query/kernel" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/bias" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_10/attention/self/value/kernel" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/bias" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_10/output/dense/bias" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_10/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_10/output/dense/kernel" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_10/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/bias" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/bias" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/key/kernel" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/bias" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/query/kernel" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/bias" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_11/attention/self/value/kernel" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/bias" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_11/output/dense/bias" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_11/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_11/output/dense/kernel" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_11/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/bias" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/bias" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/key/kernel" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/bias" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/query/kernel" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/bias" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_2/attention/self/value/kernel" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/bias" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_2/output/dense/bias" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_2/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_2/output/dense/kernel" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_2/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/bias" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/bias" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/key/kernel" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/bias" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/query/kernel" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/bias" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_3/attention/self/value/kernel" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/bias" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_3/output/dense/bias" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_3/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_3/output/dense/kernel" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_3/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/bias" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/bias" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/key/kernel" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/bias" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/query/kernel" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/bias" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_4/attention/self/value/kernel" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/bias" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_4/output/dense/bias" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_4/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_4/output/dense/kernel" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_4/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/bias" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/bias" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/key/kernel" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/bias" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/query/kernel" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/bias" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_5/attention/self/value/kernel" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/bias" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_5/output/dense/bias" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_5/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_5/output/dense/kernel" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_5/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/bias" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/bias" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/key/kernel" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/bias" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/query/kernel" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/bias" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_6/attention/self/value/kernel" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/bias" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_6/output/dense/bias" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_6/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_6/output/dense/kernel" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_6/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/bias" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/bias" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/key/kernel" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/bias" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/query/kernel" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/bias" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_7/attention/self/value/kernel" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/bias" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_7/output/dense/bias" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_7/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_7/output/dense/kernel" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_7/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/bias" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/bias" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/key/kernel" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/bias" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/query/kernel" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/bias" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_8/attention/self/value/kernel" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/bias" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_8/output/dense/bias" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_8/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_8/output/dense/kernel" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_8/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/bias" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/bias" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/key/kernel" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/bias" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/query/kernel" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/bias" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + string_val: "bert/encoder/layer_9/attention/self/value/kernel" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + string_val: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/bias" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + string_val: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + string_val: "bert/encoder/layer_9/output/dense/bias" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_m" + string_val: "bert/encoder/layer_9/output/dense/bias/adam_v" + string_val: "bert/encoder/layer_9/output/dense/kernel" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_m" + string_val: "bert/encoder/layer_9/output/dense/kernel/adam_v" + string_val: "bert/pooler/dense/bias" + string_val: "bert/pooler/dense/bias/adam_m" + string_val: "bert/pooler/dense/bias/adam_v" + string_val: "bert/pooler/dense/kernel" + string_val: "bert/pooler/dense/kernel/adam_m" + string_val: "bert/pooler/dense/kernel/adam_v" + string_val: "global_step" + string_val: "output_bias" + string_val: "output_bias/adam_m" + string_val: "output_bias/adam_v" + string_val: "output_weights" + string_val: "output_weights/adam_m" + string_val: "output_weights/adam_v" + } + } + } +} +node { + name: "save/RestoreV2/shape_and_slices" + op: "Const" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 604 + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + dim { + size: 604 + } + } + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + string_val: "" + } + } + } +} +node { + name: "save/RestoreV2" + op: "RestoreV2" + input: "save/Const" + input: "save/RestoreV2/tensor_names" + input: "save/RestoreV2/shape_and_slices" + device: "/device:CPU:0" + attr { + key: "_output_shapes" + value { + list { + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + shape { + unknown_rank: true + } + } + } + } + attr { + key: "dtypes" + value { + list { + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_INT64 + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + type: DT_FLOAT + } + } + } +} +node { + name: "save/Assign" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta" + input: "save/RestoreV2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_1" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_m" + input: "save/RestoreV2:1" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_2" + op: "Assign" + input: "bert/embeddings/LayerNorm/beta/adam_v" + input: "save/RestoreV2:2" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_3" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma" + input: "save/RestoreV2:3" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_4" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:4" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_5" + op: "Assign" + input: "bert/embeddings/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:5" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_6" + op: "Assign" + input: "bert/embeddings/position_embeddings" + input: "save/RestoreV2:6" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_7" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_m" + input: "save/RestoreV2:7" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_8" + op: "Assign" + input: "bert/embeddings/position_embeddings/adam_v" + input: "save/RestoreV2:8" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/position_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 512 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_9" + op: "Assign" + input: "bert/embeddings/token_type_embeddings" + input: "save/RestoreV2:9" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_10" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_m" + input: "save/RestoreV2:10" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_11" + op: "Assign" + input: "bert/embeddings/token_type_embeddings/adam_v" + input: "save/RestoreV2:11" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/token_type_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 2 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_12" + op: "Assign" + input: "bert/embeddings/word_embeddings" + input: "save/RestoreV2:12" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_13" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_m" + input: "save/RestoreV2:13" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_14" + op: "Assign" + input: "bert/embeddings/word_embeddings/adam_v" + input: "save/RestoreV2:14" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/embeddings/word_embeddings/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 21128 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_15" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta" + input: "save/RestoreV2:15" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_16" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:16" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_17" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:17" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_18" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:18" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_19" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:19" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_20" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:20" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_21" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias" + input: "save/RestoreV2:21" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_22" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:22" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_23" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:23" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_24" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel" + input: "save/RestoreV2:24" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_25" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:25" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_26" + op: "Assign" + input: "bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:26" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_27" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias" + input: "save/RestoreV2:27" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_28" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_m" + input: "save/RestoreV2:28" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_29" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/bias/adam_v" + input: "save/RestoreV2:29" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_30" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel" + input: "save/RestoreV2:30" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_31" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:31" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_32" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:32" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_33" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias" + input: "save/RestoreV2:33" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_34" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_m" + input: "save/RestoreV2:34" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_35" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/bias/adam_v" + input: "save/RestoreV2:35" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_36" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel" + input: "save/RestoreV2:36" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_37" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:37" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_38" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:38" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_39" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias" + input: "save/RestoreV2:39" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_40" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_m" + input: "save/RestoreV2:40" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_41" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/bias/adam_v" + input: "save/RestoreV2:41" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_42" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel" + input: "save/RestoreV2:42" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_43" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:43" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_44" + op: "Assign" + input: "bert/encoder/layer_0/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:44" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_45" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias" + input: "save/RestoreV2:45" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_46" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:46" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_47" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:47" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_48" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel" + input: "save/RestoreV2:48" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_49" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:49" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_50" + op: "Assign" + input: "bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:50" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_51" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta" + input: "save/RestoreV2:51" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_52" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:52" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_53" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:53" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_54" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma" + input: "save/RestoreV2:54" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_55" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:55" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_56" + op: "Assign" + input: "bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:56" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_57" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias" + input: "save/RestoreV2:57" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_58" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_m" + input: "save/RestoreV2:58" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_59" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/bias/adam_v" + input: "save/RestoreV2:59" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_60" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel" + input: "save/RestoreV2:60" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_61" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_m" + input: "save/RestoreV2:61" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_62" + op: "Assign" + input: "bert/encoder/layer_0/output/dense/kernel/adam_v" + input: "save/RestoreV2:62" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_0/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_63" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta" + input: "save/RestoreV2:63" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_64" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:64" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_65" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:65" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_66" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:66" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_67" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:67" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_68" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:68" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_69" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias" + input: "save/RestoreV2:69" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_70" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:70" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_71" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:71" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_72" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel" + input: "save/RestoreV2:72" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_73" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:73" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_74" + op: "Assign" + input: "bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:74" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_75" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias" + input: "save/RestoreV2:75" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_76" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_m" + input: "save/RestoreV2:76" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_77" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/bias/adam_v" + input: "save/RestoreV2:77" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_78" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel" + input: "save/RestoreV2:78" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_79" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:79" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_80" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:80" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_81" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias" + input: "save/RestoreV2:81" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_82" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_m" + input: "save/RestoreV2:82" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_83" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/bias/adam_v" + input: "save/RestoreV2:83" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_84" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel" + input: "save/RestoreV2:84" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_85" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:85" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_86" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:86" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_87" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias" + input: "save/RestoreV2:87" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_88" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_m" + input: "save/RestoreV2:88" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_89" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/bias/adam_v" + input: "save/RestoreV2:89" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_90" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel" + input: "save/RestoreV2:90" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_91" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:91" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_92" + op: "Assign" + input: "bert/encoder/layer_1/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:92" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_93" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias" + input: "save/RestoreV2:93" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_94" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:94" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_95" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:95" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_96" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel" + input: "save/RestoreV2:96" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_97" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:97" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_98" + op: "Assign" + input: "bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:98" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_99" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta" + input: "save/RestoreV2:99" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_100" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:100" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_101" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:101" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_102" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma" + input: "save/RestoreV2:102" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_103" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:103" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_104" + op: "Assign" + input: "bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:104" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_105" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias" + input: "save/RestoreV2:105" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_106" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_m" + input: "save/RestoreV2:106" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_107" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/bias/adam_v" + input: "save/RestoreV2:107" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_108" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel" + input: "save/RestoreV2:108" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_109" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_m" + input: "save/RestoreV2:109" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_110" + op: "Assign" + input: "bert/encoder/layer_1/output/dense/kernel/adam_v" + input: "save/RestoreV2:110" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_1/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_111" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta" + input: "save/RestoreV2:111" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_112" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:112" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_113" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:113" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_114" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:114" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_115" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:115" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_116" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:116" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_117" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias" + input: "save/RestoreV2:117" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_118" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:118" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_119" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:119" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_120" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel" + input: "save/RestoreV2:120" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_121" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:121" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_122" + op: "Assign" + input: "bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:122" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_123" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias" + input: "save/RestoreV2:123" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_124" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_m" + input: "save/RestoreV2:124" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_125" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/bias/adam_v" + input: "save/RestoreV2:125" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_126" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel" + input: "save/RestoreV2:126" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_127" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:127" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_128" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:128" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_129" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias" + input: "save/RestoreV2:129" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_130" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_m" + input: "save/RestoreV2:130" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_131" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/bias/adam_v" + input: "save/RestoreV2:131" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_132" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel" + input: "save/RestoreV2:132" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_133" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:133" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_134" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:134" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_135" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias" + input: "save/RestoreV2:135" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_136" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_m" + input: "save/RestoreV2:136" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_137" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/bias/adam_v" + input: "save/RestoreV2:137" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_138" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel" + input: "save/RestoreV2:138" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_139" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:139" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_140" + op: "Assign" + input: "bert/encoder/layer_10/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:140" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_141" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias" + input: "save/RestoreV2:141" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_142" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:142" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_143" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:143" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_144" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel" + input: "save/RestoreV2:144" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_145" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:145" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_146" + op: "Assign" + input: "bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:146" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_147" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta" + input: "save/RestoreV2:147" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_148" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:148" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_149" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:149" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_150" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma" + input: "save/RestoreV2:150" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_151" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:151" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_152" + op: "Assign" + input: "bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:152" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_153" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias" + input: "save/RestoreV2:153" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_154" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_m" + input: "save/RestoreV2:154" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_155" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/bias/adam_v" + input: "save/RestoreV2:155" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_156" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel" + input: "save/RestoreV2:156" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_157" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_m" + input: "save/RestoreV2:157" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_158" + op: "Assign" + input: "bert/encoder/layer_10/output/dense/kernel/adam_v" + input: "save/RestoreV2:158" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_10/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_159" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta" + input: "save/RestoreV2:159" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_160" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:160" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_161" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:161" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_162" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:162" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_163" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:163" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_164" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:164" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_165" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias" + input: "save/RestoreV2:165" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_166" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:166" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_167" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:167" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_168" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel" + input: "save/RestoreV2:168" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_169" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:169" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_170" + op: "Assign" + input: "bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:170" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_171" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias" + input: "save/RestoreV2:171" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_172" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_m" + input: "save/RestoreV2:172" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_173" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/bias/adam_v" + input: "save/RestoreV2:173" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_174" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel" + input: "save/RestoreV2:174" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_175" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:175" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_176" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:176" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_177" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias" + input: "save/RestoreV2:177" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_178" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_m" + input: "save/RestoreV2:178" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_179" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/bias/adam_v" + input: "save/RestoreV2:179" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_180" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel" + input: "save/RestoreV2:180" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_181" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:181" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_182" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:182" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_183" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias" + input: "save/RestoreV2:183" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_184" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_m" + input: "save/RestoreV2:184" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_185" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/bias/adam_v" + input: "save/RestoreV2:185" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_186" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel" + input: "save/RestoreV2:186" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_187" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:187" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_188" + op: "Assign" + input: "bert/encoder/layer_11/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:188" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_189" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias" + input: "save/RestoreV2:189" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_190" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:190" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_191" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:191" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_192" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel" + input: "save/RestoreV2:192" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_193" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:193" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_194" + op: "Assign" + input: "bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:194" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_195" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta" + input: "save/RestoreV2:195" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_196" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:196" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_197" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:197" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_198" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma" + input: "save/RestoreV2:198" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_199" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:199" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_200" + op: "Assign" + input: "bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:200" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_201" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias" + input: "save/RestoreV2:201" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_202" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_m" + input: "save/RestoreV2:202" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_203" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/bias/adam_v" + input: "save/RestoreV2:203" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_204" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel" + input: "save/RestoreV2:204" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_205" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_m" + input: "save/RestoreV2:205" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_206" + op: "Assign" + input: "bert/encoder/layer_11/output/dense/kernel/adam_v" + input: "save/RestoreV2:206" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_11/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_207" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta" + input: "save/RestoreV2:207" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_208" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:208" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_209" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:209" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_210" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:210" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_211" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:211" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_212" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:212" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_213" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias" + input: "save/RestoreV2:213" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_214" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:214" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_215" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:215" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_216" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel" + input: "save/RestoreV2:216" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_217" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:217" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_218" + op: "Assign" + input: "bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:218" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_219" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias" + input: "save/RestoreV2:219" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_220" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_m" + input: "save/RestoreV2:220" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_221" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/bias/adam_v" + input: "save/RestoreV2:221" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_222" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel" + input: "save/RestoreV2:222" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_223" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:223" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_224" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:224" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_225" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias" + input: "save/RestoreV2:225" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_226" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_m" + input: "save/RestoreV2:226" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_227" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/bias/adam_v" + input: "save/RestoreV2:227" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_228" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel" + input: "save/RestoreV2:228" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_229" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:229" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_230" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:230" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_231" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias" + input: "save/RestoreV2:231" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_232" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_m" + input: "save/RestoreV2:232" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_233" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/bias/adam_v" + input: "save/RestoreV2:233" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_234" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel" + input: "save/RestoreV2:234" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_235" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:235" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_236" + op: "Assign" + input: "bert/encoder/layer_2/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:236" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_237" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias" + input: "save/RestoreV2:237" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_238" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:238" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_239" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:239" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_240" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel" + input: "save/RestoreV2:240" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_241" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:241" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_242" + op: "Assign" + input: "bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:242" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_243" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta" + input: "save/RestoreV2:243" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_244" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:244" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_245" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:245" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_246" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma" + input: "save/RestoreV2:246" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_247" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:247" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_248" + op: "Assign" + input: "bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:248" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_249" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias" + input: "save/RestoreV2:249" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_250" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_m" + input: "save/RestoreV2:250" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_251" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/bias/adam_v" + input: "save/RestoreV2:251" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_252" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel" + input: "save/RestoreV2:252" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_253" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_m" + input: "save/RestoreV2:253" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_254" + op: "Assign" + input: "bert/encoder/layer_2/output/dense/kernel/adam_v" + input: "save/RestoreV2:254" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_2/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_255" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta" + input: "save/RestoreV2:255" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_256" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:256" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_257" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:257" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_258" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:258" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_259" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:259" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_260" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:260" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_261" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias" + input: "save/RestoreV2:261" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_262" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:262" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_263" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:263" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_264" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel" + input: "save/RestoreV2:264" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_265" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:265" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_266" + op: "Assign" + input: "bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:266" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_267" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias" + input: "save/RestoreV2:267" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_268" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_m" + input: "save/RestoreV2:268" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_269" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/bias/adam_v" + input: "save/RestoreV2:269" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_270" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel" + input: "save/RestoreV2:270" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_271" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:271" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_272" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:272" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_273" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias" + input: "save/RestoreV2:273" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_274" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_m" + input: "save/RestoreV2:274" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_275" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/bias/adam_v" + input: "save/RestoreV2:275" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_276" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel" + input: "save/RestoreV2:276" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_277" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:277" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_278" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:278" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_279" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias" + input: "save/RestoreV2:279" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_280" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_m" + input: "save/RestoreV2:280" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_281" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/bias/adam_v" + input: "save/RestoreV2:281" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_282" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel" + input: "save/RestoreV2:282" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_283" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:283" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_284" + op: "Assign" + input: "bert/encoder/layer_3/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:284" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_285" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias" + input: "save/RestoreV2:285" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_286" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:286" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_287" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:287" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_288" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel" + input: "save/RestoreV2:288" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_289" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:289" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_290" + op: "Assign" + input: "bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:290" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_291" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta" + input: "save/RestoreV2:291" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_292" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:292" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_293" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:293" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_294" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma" + input: "save/RestoreV2:294" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_295" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:295" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_296" + op: "Assign" + input: "bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:296" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_297" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias" + input: "save/RestoreV2:297" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_298" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_m" + input: "save/RestoreV2:298" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_299" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/bias/adam_v" + input: "save/RestoreV2:299" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_300" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel" + input: "save/RestoreV2:300" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_301" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_m" + input: "save/RestoreV2:301" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_302" + op: "Assign" + input: "bert/encoder/layer_3/output/dense/kernel/adam_v" + input: "save/RestoreV2:302" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_3/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_303" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta" + input: "save/RestoreV2:303" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_304" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:304" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_305" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:305" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_306" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:306" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_307" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:307" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_308" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:308" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_309" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias" + input: "save/RestoreV2:309" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_310" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:310" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_311" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:311" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_312" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel" + input: "save/RestoreV2:312" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_313" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:313" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_314" + op: "Assign" + input: "bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:314" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_315" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias" + input: "save/RestoreV2:315" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_316" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_m" + input: "save/RestoreV2:316" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_317" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/bias/adam_v" + input: "save/RestoreV2:317" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_318" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel" + input: "save/RestoreV2:318" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_319" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:319" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_320" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:320" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_321" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias" + input: "save/RestoreV2:321" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_322" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_m" + input: "save/RestoreV2:322" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_323" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/bias/adam_v" + input: "save/RestoreV2:323" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_324" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel" + input: "save/RestoreV2:324" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_325" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:325" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_326" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:326" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_327" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias" + input: "save/RestoreV2:327" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_328" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_m" + input: "save/RestoreV2:328" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_329" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/bias/adam_v" + input: "save/RestoreV2:329" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_330" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel" + input: "save/RestoreV2:330" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_331" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:331" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_332" + op: "Assign" + input: "bert/encoder/layer_4/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:332" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_333" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias" + input: "save/RestoreV2:333" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_334" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:334" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_335" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:335" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_336" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel" + input: "save/RestoreV2:336" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_337" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:337" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_338" + op: "Assign" + input: "bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:338" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_339" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta" + input: "save/RestoreV2:339" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_340" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:340" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_341" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:341" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_342" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma" + input: "save/RestoreV2:342" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_343" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:343" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_344" + op: "Assign" + input: "bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:344" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_345" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias" + input: "save/RestoreV2:345" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_346" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_m" + input: "save/RestoreV2:346" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_347" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/bias/adam_v" + input: "save/RestoreV2:347" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_348" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel" + input: "save/RestoreV2:348" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_349" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_m" + input: "save/RestoreV2:349" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_350" + op: "Assign" + input: "bert/encoder/layer_4/output/dense/kernel/adam_v" + input: "save/RestoreV2:350" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_4/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_351" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta" + input: "save/RestoreV2:351" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_352" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:352" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_353" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:353" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_354" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:354" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_355" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:355" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_356" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:356" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_357" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias" + input: "save/RestoreV2:357" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_358" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:358" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_359" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:359" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_360" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel" + input: "save/RestoreV2:360" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_361" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:361" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_362" + op: "Assign" + input: "bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:362" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_363" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias" + input: "save/RestoreV2:363" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_364" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_m" + input: "save/RestoreV2:364" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_365" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/bias/adam_v" + input: "save/RestoreV2:365" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_366" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel" + input: "save/RestoreV2:366" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_367" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:367" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_368" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:368" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_369" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias" + input: "save/RestoreV2:369" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_370" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_m" + input: "save/RestoreV2:370" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_371" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/bias/adam_v" + input: "save/RestoreV2:371" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_372" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel" + input: "save/RestoreV2:372" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_373" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:373" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_374" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:374" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_375" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias" + input: "save/RestoreV2:375" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_376" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_m" + input: "save/RestoreV2:376" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_377" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/bias/adam_v" + input: "save/RestoreV2:377" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_378" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel" + input: "save/RestoreV2:378" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_379" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:379" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_380" + op: "Assign" + input: "bert/encoder/layer_5/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:380" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_381" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias" + input: "save/RestoreV2:381" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_382" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:382" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_383" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:383" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_384" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel" + input: "save/RestoreV2:384" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_385" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:385" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_386" + op: "Assign" + input: "bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:386" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_387" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta" + input: "save/RestoreV2:387" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_388" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:388" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_389" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:389" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_390" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma" + input: "save/RestoreV2:390" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_391" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:391" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_392" + op: "Assign" + input: "bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:392" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_393" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias" + input: "save/RestoreV2:393" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_394" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_m" + input: "save/RestoreV2:394" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_395" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/bias/adam_v" + input: "save/RestoreV2:395" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_396" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel" + input: "save/RestoreV2:396" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_397" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_m" + input: "save/RestoreV2:397" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_398" + op: "Assign" + input: "bert/encoder/layer_5/output/dense/kernel/adam_v" + input: "save/RestoreV2:398" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_5/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_399" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta" + input: "save/RestoreV2:399" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_400" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:400" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_401" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:401" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_402" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:402" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_403" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:403" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_404" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:404" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_405" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias" + input: "save/RestoreV2:405" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_406" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:406" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_407" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:407" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_408" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel" + input: "save/RestoreV2:408" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_409" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:409" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_410" + op: "Assign" + input: "bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:410" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_411" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias" + input: "save/RestoreV2:411" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_412" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_m" + input: "save/RestoreV2:412" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_413" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/bias/adam_v" + input: "save/RestoreV2:413" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_414" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel" + input: "save/RestoreV2:414" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_415" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:415" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_416" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:416" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_417" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias" + input: "save/RestoreV2:417" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_418" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_m" + input: "save/RestoreV2:418" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_419" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/bias/adam_v" + input: "save/RestoreV2:419" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_420" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel" + input: "save/RestoreV2:420" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_421" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:421" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_422" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:422" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_423" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias" + input: "save/RestoreV2:423" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_424" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_m" + input: "save/RestoreV2:424" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_425" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/bias/adam_v" + input: "save/RestoreV2:425" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_426" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel" + input: "save/RestoreV2:426" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_427" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:427" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_428" + op: "Assign" + input: "bert/encoder/layer_6/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:428" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_429" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias" + input: "save/RestoreV2:429" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_430" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:430" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_431" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:431" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_432" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel" + input: "save/RestoreV2:432" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_433" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:433" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_434" + op: "Assign" + input: "bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:434" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_435" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta" + input: "save/RestoreV2:435" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_436" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:436" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_437" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:437" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_438" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma" + input: "save/RestoreV2:438" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_439" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:439" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_440" + op: "Assign" + input: "bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:440" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_441" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias" + input: "save/RestoreV2:441" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_442" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_m" + input: "save/RestoreV2:442" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_443" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/bias/adam_v" + input: "save/RestoreV2:443" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_444" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel" + input: "save/RestoreV2:444" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_445" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_m" + input: "save/RestoreV2:445" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_446" + op: "Assign" + input: "bert/encoder/layer_6/output/dense/kernel/adam_v" + input: "save/RestoreV2:446" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_6/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_447" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta" + input: "save/RestoreV2:447" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_448" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:448" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_449" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:449" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_450" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:450" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_451" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:451" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_452" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:452" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_453" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias" + input: "save/RestoreV2:453" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_454" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:454" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_455" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:455" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_456" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel" + input: "save/RestoreV2:456" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_457" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:457" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_458" + op: "Assign" + input: "bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:458" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_459" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias" + input: "save/RestoreV2:459" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_460" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_m" + input: "save/RestoreV2:460" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_461" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/bias/adam_v" + input: "save/RestoreV2:461" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_462" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel" + input: "save/RestoreV2:462" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_463" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:463" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_464" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:464" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_465" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias" + input: "save/RestoreV2:465" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_466" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_m" + input: "save/RestoreV2:466" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_467" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/bias/adam_v" + input: "save/RestoreV2:467" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_468" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel" + input: "save/RestoreV2:468" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_469" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:469" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_470" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:470" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_471" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias" + input: "save/RestoreV2:471" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_472" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_m" + input: "save/RestoreV2:472" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_473" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/bias/adam_v" + input: "save/RestoreV2:473" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_474" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel" + input: "save/RestoreV2:474" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_475" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:475" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_476" + op: "Assign" + input: "bert/encoder/layer_7/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:476" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_477" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias" + input: "save/RestoreV2:477" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_478" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:478" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_479" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:479" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_480" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel" + input: "save/RestoreV2:480" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_481" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:481" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_482" + op: "Assign" + input: "bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:482" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_483" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta" + input: "save/RestoreV2:483" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_484" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:484" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_485" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:485" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_486" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma" + input: "save/RestoreV2:486" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_487" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:487" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_488" + op: "Assign" + input: "bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:488" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_489" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias" + input: "save/RestoreV2:489" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_490" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_m" + input: "save/RestoreV2:490" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_491" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/bias/adam_v" + input: "save/RestoreV2:491" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_492" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel" + input: "save/RestoreV2:492" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_493" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_m" + input: "save/RestoreV2:493" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_494" + op: "Assign" + input: "bert/encoder/layer_7/output/dense/kernel/adam_v" + input: "save/RestoreV2:494" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_7/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_495" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta" + input: "save/RestoreV2:495" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_496" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:496" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_497" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:497" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_498" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:498" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_499" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:499" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_500" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:500" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_501" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias" + input: "save/RestoreV2:501" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_502" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:502" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_503" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:503" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_504" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel" + input: "save/RestoreV2:504" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_505" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:505" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_506" + op: "Assign" + input: "bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:506" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_507" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias" + input: "save/RestoreV2:507" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_508" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_m" + input: "save/RestoreV2:508" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_509" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/bias/adam_v" + input: "save/RestoreV2:509" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_510" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel" + input: "save/RestoreV2:510" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_511" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:511" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_512" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:512" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_513" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias" + input: "save/RestoreV2:513" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_514" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_m" + input: "save/RestoreV2:514" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_515" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/bias/adam_v" + input: "save/RestoreV2:515" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_516" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel" + input: "save/RestoreV2:516" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_517" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:517" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_518" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:518" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_519" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias" + input: "save/RestoreV2:519" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_520" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_m" + input: "save/RestoreV2:520" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_521" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/bias/adam_v" + input: "save/RestoreV2:521" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_522" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel" + input: "save/RestoreV2:522" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_523" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:523" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_524" + op: "Assign" + input: "bert/encoder/layer_8/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:524" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_525" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias" + input: "save/RestoreV2:525" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_526" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:526" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_527" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:527" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_528" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel" + input: "save/RestoreV2:528" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_529" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:529" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_530" + op: "Assign" + input: "bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:530" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_531" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta" + input: "save/RestoreV2:531" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_532" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:532" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_533" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:533" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_534" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma" + input: "save/RestoreV2:534" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_535" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:535" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_536" + op: "Assign" + input: "bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:536" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_537" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias" + input: "save/RestoreV2:537" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_538" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_m" + input: "save/RestoreV2:538" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_539" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/bias/adam_v" + input: "save/RestoreV2:539" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_540" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel" + input: "save/RestoreV2:540" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_541" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_m" + input: "save/RestoreV2:541" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_542" + op: "Assign" + input: "bert/encoder/layer_8/output/dense/kernel/adam_v" + input: "save/RestoreV2:542" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_8/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_543" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta" + input: "save/RestoreV2:543" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_544" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:544" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_545" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:545" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_546" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma" + input: "save/RestoreV2:546" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_547" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:547" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_548" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:548" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_549" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias" + input: "save/RestoreV2:549" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_550" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_m" + input: "save/RestoreV2:550" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_551" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/bias/adam_v" + input: "save/RestoreV2:551" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_552" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel" + input: "save/RestoreV2:552" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_553" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + input: "save/RestoreV2:553" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_554" + op: "Assign" + input: "bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + input: "save/RestoreV2:554" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_555" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias" + input: "save/RestoreV2:555" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_556" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_m" + input: "save/RestoreV2:556" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_557" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/bias/adam_v" + input: "save/RestoreV2:557" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_558" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel" + input: "save/RestoreV2:558" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_559" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_m" + input: "save/RestoreV2:559" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_560" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/key/kernel/adam_v" + input: "save/RestoreV2:560" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/key/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_561" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias" + input: "save/RestoreV2:561" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_562" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_m" + input: "save/RestoreV2:562" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_563" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/bias/adam_v" + input: "save/RestoreV2:563" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_564" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel" + input: "save/RestoreV2:564" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_565" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_m" + input: "save/RestoreV2:565" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_566" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/query/kernel/adam_v" + input: "save/RestoreV2:566" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/query/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_567" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias" + input: "save/RestoreV2:567" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_568" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_m" + input: "save/RestoreV2:568" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_569" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/bias/adam_v" + input: "save/RestoreV2:569" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_570" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel" + input: "save/RestoreV2:570" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_571" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_m" + input: "save/RestoreV2:571" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_572" + op: "Assign" + input: "bert/encoder/layer_9/attention/self/value/kernel/adam_v" + input: "save/RestoreV2:572" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/attention/self/value/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_573" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias" + input: "save/RestoreV2:573" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_574" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_m" + input: "save/RestoreV2:574" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_575" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/bias/adam_v" + input: "save/RestoreV2:575" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_576" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel" + input: "save/RestoreV2:576" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_577" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + input: "save/RestoreV2:577" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_578" + op: "Assign" + input: "bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + input: "save/RestoreV2:578" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/intermediate/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 3072 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_579" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta" + input: "save/RestoreV2:579" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_580" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + input: "save/RestoreV2:580" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_581" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + input: "save/RestoreV2:581" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/beta/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_582" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma" + input: "save/RestoreV2:582" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_583" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + input: "save/RestoreV2:583" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_584" + op: "Assign" + input: "bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + input: "save/RestoreV2:584" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/LayerNorm/gamma/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_585" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias" + input: "save/RestoreV2:585" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_586" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_m" + input: "save/RestoreV2:586" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_587" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/bias/adam_v" + input: "save/RestoreV2:587" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_588" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel" + input: "save/RestoreV2:588" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_589" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_m" + input: "save/RestoreV2:589" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_590" + op: "Assign" + input: "bert/encoder/layer_9/output/dense/kernel/adam_v" + input: "save/RestoreV2:590" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/encoder/layer_9/output/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3072 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_591" + op: "Assign" + input: "bert/pooler/dense/bias" + input: "save/RestoreV2:591" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_592" + op: "Assign" + input: "bert/pooler/dense/bias/adam_m" + input: "save/RestoreV2:592" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_593" + op: "Assign" + input: "bert/pooler/dense/bias/adam_v" + input: "save/RestoreV2:593" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_594" + op: "Assign" + input: "bert/pooler/dense/kernel" + input: "save/RestoreV2:594" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_595" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_m" + input: "save/RestoreV2:595" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_596" + op: "Assign" + input: "bert/pooler/dense/kernel/adam_v" + input: "save/RestoreV2:596" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@bert/pooler/dense/kernel/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 768 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Identity_1" + op: "Identity" + input: "save/RestoreV2:597" + attr { + key: "T" + value { + type: DT_INT64 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + unknown_rank: true + } + } + } + } +} +node { + name: "save/AssignVariableOp" + op: "AssignVariableOp" + input: "global_step" + input: "save/Identity_1" + attr { + key: "dtype" + value { + type: DT_INT64 + } + } +} +node { + name: "save/Assign_597" + op: "Assign" + input: "output_bias" + input: "save/RestoreV2:598" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_598" + op: "Assign" + input: "output_bias/adam_m" + input: "save/RestoreV2:599" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_599" + op: "Assign" + input: "output_bias/adam_v" + input: "save/RestoreV2:600" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_bias/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_600" + op: "Assign" + input: "output_weights" + input: "save/RestoreV2:601" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_601" + op: "Assign" + input: "output_weights/adam_m" + input: "save/RestoreV2:602" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_m" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/Assign_602" + op: "Assign" + input: "output_weights/adam_v" + input: "save/RestoreV2:603" + attr { + key: "T" + value { + type: DT_FLOAT + } + } + attr { + key: "_class" + value { + list { + s: "loc:@output_weights/adam_v" + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 3 + } + dim { + size: 768 + } + } + } + } + } + attr { + key: "use_locking" + value { + b: true + } + } + attr { + key: "validate_shape" + value { + b: true + } + } +} +node { + name: "save/restore_shard" + op: "NoOp" + input: "^save/Assign" + input: "^save/AssignVariableOp" + input: "^save/Assign_1" + input: "^save/Assign_10" + input: "^save/Assign_100" + input: "^save/Assign_101" + input: "^save/Assign_102" + input: "^save/Assign_103" + input: "^save/Assign_104" + input: "^save/Assign_105" + input: "^save/Assign_106" + input: "^save/Assign_107" + input: "^save/Assign_108" + input: "^save/Assign_109" + input: "^save/Assign_11" + input: "^save/Assign_110" + input: "^save/Assign_111" + input: "^save/Assign_112" + input: "^save/Assign_113" + input: "^save/Assign_114" + input: "^save/Assign_115" + input: "^save/Assign_116" + input: "^save/Assign_117" + input: "^save/Assign_118" + input: "^save/Assign_119" + input: "^save/Assign_12" + input: "^save/Assign_120" + input: "^save/Assign_121" + input: "^save/Assign_122" + input: "^save/Assign_123" + input: "^save/Assign_124" + input: "^save/Assign_125" + input: "^save/Assign_126" + input: "^save/Assign_127" + input: "^save/Assign_128" + input: "^save/Assign_129" + input: "^save/Assign_13" + input: "^save/Assign_130" + input: "^save/Assign_131" + input: "^save/Assign_132" + input: "^save/Assign_133" + input: "^save/Assign_134" + input: "^save/Assign_135" + input: "^save/Assign_136" + input: "^save/Assign_137" + input: "^save/Assign_138" + input: "^save/Assign_139" + input: "^save/Assign_14" + input: "^save/Assign_140" + input: "^save/Assign_141" + input: "^save/Assign_142" + input: "^save/Assign_143" + input: "^save/Assign_144" + input: "^save/Assign_145" + input: "^save/Assign_146" + input: "^save/Assign_147" + input: "^save/Assign_148" + input: "^save/Assign_149" + input: "^save/Assign_15" + input: "^save/Assign_150" + input: "^save/Assign_151" + input: "^save/Assign_152" + input: "^save/Assign_153" + input: "^save/Assign_154" + input: "^save/Assign_155" + input: "^save/Assign_156" + input: "^save/Assign_157" + input: "^save/Assign_158" + input: "^save/Assign_159" + input: "^save/Assign_16" + input: "^save/Assign_160" + input: "^save/Assign_161" + input: "^save/Assign_162" + input: "^save/Assign_163" + input: "^save/Assign_164" + input: "^save/Assign_165" + input: "^save/Assign_166" + input: "^save/Assign_167" + input: "^save/Assign_168" + input: "^save/Assign_169" + input: "^save/Assign_17" + input: "^save/Assign_170" + input: "^save/Assign_171" + input: "^save/Assign_172" + input: "^save/Assign_173" + input: "^save/Assign_174" + input: "^save/Assign_175" + input: "^save/Assign_176" + input: "^save/Assign_177" + input: "^save/Assign_178" + input: "^save/Assign_179" + input: "^save/Assign_18" + input: "^save/Assign_180" + input: "^save/Assign_181" + input: "^save/Assign_182" + input: "^save/Assign_183" + input: "^save/Assign_184" + input: "^save/Assign_185" + input: "^save/Assign_186" + input: "^save/Assign_187" + input: "^save/Assign_188" + input: "^save/Assign_189" + input: "^save/Assign_19" + input: "^save/Assign_190" + input: "^save/Assign_191" + input: "^save/Assign_192" + input: "^save/Assign_193" + input: "^save/Assign_194" + input: "^save/Assign_195" + input: "^save/Assign_196" + input: "^save/Assign_197" + input: "^save/Assign_198" + input: "^save/Assign_199" + input: "^save/Assign_2" + input: "^save/Assign_20" + input: "^save/Assign_200" + input: "^save/Assign_201" + input: "^save/Assign_202" + input: "^save/Assign_203" + input: "^save/Assign_204" + input: "^save/Assign_205" + input: "^save/Assign_206" + input: "^save/Assign_207" + input: "^save/Assign_208" + input: "^save/Assign_209" + input: "^save/Assign_21" + input: "^save/Assign_210" + input: "^save/Assign_211" + input: "^save/Assign_212" + input: "^save/Assign_213" + input: "^save/Assign_214" + input: "^save/Assign_215" + input: "^save/Assign_216" + input: "^save/Assign_217" + input: "^save/Assign_218" + input: "^save/Assign_219" + input: "^save/Assign_22" + input: "^save/Assign_220" + input: "^save/Assign_221" + input: "^save/Assign_222" + input: "^save/Assign_223" + input: "^save/Assign_224" + input: "^save/Assign_225" + input: "^save/Assign_226" + input: "^save/Assign_227" + input: "^save/Assign_228" + input: "^save/Assign_229" + input: "^save/Assign_23" + input: "^save/Assign_230" + input: "^save/Assign_231" + input: "^save/Assign_232" + input: "^save/Assign_233" + input: "^save/Assign_234" + input: "^save/Assign_235" + input: "^save/Assign_236" + input: "^save/Assign_237" + input: "^save/Assign_238" + input: "^save/Assign_239" + input: "^save/Assign_24" + input: "^save/Assign_240" + input: "^save/Assign_241" + input: "^save/Assign_242" + input: "^save/Assign_243" + input: "^save/Assign_244" + input: "^save/Assign_245" + input: "^save/Assign_246" + input: "^save/Assign_247" + input: "^save/Assign_248" + input: "^save/Assign_249" + input: "^save/Assign_25" + input: "^save/Assign_250" + input: "^save/Assign_251" + input: "^save/Assign_252" + input: "^save/Assign_253" + input: "^save/Assign_254" + input: "^save/Assign_255" + input: "^save/Assign_256" + input: "^save/Assign_257" + input: "^save/Assign_258" + input: "^save/Assign_259" + input: "^save/Assign_26" + input: "^save/Assign_260" + input: "^save/Assign_261" + input: "^save/Assign_262" + input: "^save/Assign_263" + input: "^save/Assign_264" + input: "^save/Assign_265" + input: "^save/Assign_266" + input: "^save/Assign_267" + input: "^save/Assign_268" + input: "^save/Assign_269" + input: "^save/Assign_27" + input: "^save/Assign_270" + input: "^save/Assign_271" + input: "^save/Assign_272" + input: "^save/Assign_273" + input: "^save/Assign_274" + input: "^save/Assign_275" + input: "^save/Assign_276" + input: "^save/Assign_277" + input: "^save/Assign_278" + input: "^save/Assign_279" + input: "^save/Assign_28" + input: "^save/Assign_280" + input: "^save/Assign_281" + input: "^save/Assign_282" + input: "^save/Assign_283" + input: "^save/Assign_284" + input: "^save/Assign_285" + input: "^save/Assign_286" + input: "^save/Assign_287" + input: "^save/Assign_288" + input: "^save/Assign_289" + input: "^save/Assign_29" + input: "^save/Assign_290" + input: "^save/Assign_291" + input: "^save/Assign_292" + input: "^save/Assign_293" + input: "^save/Assign_294" + input: "^save/Assign_295" + input: "^save/Assign_296" + input: "^save/Assign_297" + input: "^save/Assign_298" + input: "^save/Assign_299" + input: "^save/Assign_3" + input: "^save/Assign_30" + input: "^save/Assign_300" + input: "^save/Assign_301" + input: "^save/Assign_302" + input: "^save/Assign_303" + input: "^save/Assign_304" + input: "^save/Assign_305" + input: "^save/Assign_306" + input: "^save/Assign_307" + input: "^save/Assign_308" + input: "^save/Assign_309" + input: "^save/Assign_31" + input: "^save/Assign_310" + input: "^save/Assign_311" + input: "^save/Assign_312" + input: "^save/Assign_313" + input: "^save/Assign_314" + input: "^save/Assign_315" + input: "^save/Assign_316" + input: "^save/Assign_317" + input: "^save/Assign_318" + input: "^save/Assign_319" + input: "^save/Assign_32" + input: "^save/Assign_320" + input: "^save/Assign_321" + input: "^save/Assign_322" + input: "^save/Assign_323" + input: "^save/Assign_324" + input: "^save/Assign_325" + input: "^save/Assign_326" + input: "^save/Assign_327" + input: "^save/Assign_328" + input: "^save/Assign_329" + input: "^save/Assign_33" + input: "^save/Assign_330" + input: "^save/Assign_331" + input: "^save/Assign_332" + input: "^save/Assign_333" + input: "^save/Assign_334" + input: "^save/Assign_335" + input: "^save/Assign_336" + input: "^save/Assign_337" + input: "^save/Assign_338" + input: "^save/Assign_339" + input: "^save/Assign_34" + input: "^save/Assign_340" + input: "^save/Assign_341" + input: "^save/Assign_342" + input: "^save/Assign_343" + input: "^save/Assign_344" + input: "^save/Assign_345" + input: "^save/Assign_346" + input: "^save/Assign_347" + input: "^save/Assign_348" + input: "^save/Assign_349" + input: "^save/Assign_35" + input: "^save/Assign_350" + input: "^save/Assign_351" + input: "^save/Assign_352" + input: "^save/Assign_353" + input: "^save/Assign_354" + input: "^save/Assign_355" + input: "^save/Assign_356" + input: "^save/Assign_357" + input: "^save/Assign_358" + input: "^save/Assign_359" + input: "^save/Assign_36" + input: "^save/Assign_360" + input: "^save/Assign_361" + input: "^save/Assign_362" + input: "^save/Assign_363" + input: "^save/Assign_364" + input: "^save/Assign_365" + input: "^save/Assign_366" + input: "^save/Assign_367" + input: "^save/Assign_368" + input: "^save/Assign_369" + input: "^save/Assign_37" + input: "^save/Assign_370" + input: "^save/Assign_371" + input: "^save/Assign_372" + input: "^save/Assign_373" + input: "^save/Assign_374" + input: "^save/Assign_375" + input: "^save/Assign_376" + input: "^save/Assign_377" + input: "^save/Assign_378" + input: "^save/Assign_379" + input: "^save/Assign_38" + input: "^save/Assign_380" + input: "^save/Assign_381" + input: "^save/Assign_382" + input: "^save/Assign_383" + input: "^save/Assign_384" + input: "^save/Assign_385" + input: "^save/Assign_386" + input: "^save/Assign_387" + input: "^save/Assign_388" + input: "^save/Assign_389" + input: "^save/Assign_39" + input: "^save/Assign_390" + input: "^save/Assign_391" + input: "^save/Assign_392" + input: "^save/Assign_393" + input: "^save/Assign_394" + input: "^save/Assign_395" + input: "^save/Assign_396" + input: "^save/Assign_397" + input: "^save/Assign_398" + input: "^save/Assign_399" + input: "^save/Assign_4" + input: "^save/Assign_40" + input: "^save/Assign_400" + input: "^save/Assign_401" + input: "^save/Assign_402" + input: "^save/Assign_403" + input: "^save/Assign_404" + input: "^save/Assign_405" + input: "^save/Assign_406" + input: "^save/Assign_407" + input: "^save/Assign_408" + input: "^save/Assign_409" + input: "^save/Assign_41" + input: "^save/Assign_410" + input: "^save/Assign_411" + input: "^save/Assign_412" + input: "^save/Assign_413" + input: "^save/Assign_414" + input: "^save/Assign_415" + input: "^save/Assign_416" + input: "^save/Assign_417" + input: "^save/Assign_418" + input: "^save/Assign_419" + input: "^save/Assign_42" + input: "^save/Assign_420" + input: "^save/Assign_421" + input: "^save/Assign_422" + input: "^save/Assign_423" + input: "^save/Assign_424" + input: "^save/Assign_425" + input: "^save/Assign_426" + input: "^save/Assign_427" + input: "^save/Assign_428" + input: "^save/Assign_429" + input: "^save/Assign_43" + input: "^save/Assign_430" + input: "^save/Assign_431" + input: "^save/Assign_432" + input: "^save/Assign_433" + input: "^save/Assign_434" + input: "^save/Assign_435" + input: "^save/Assign_436" + input: "^save/Assign_437" + input: "^save/Assign_438" + input: "^save/Assign_439" + input: "^save/Assign_44" + input: "^save/Assign_440" + input: "^save/Assign_441" + input: "^save/Assign_442" + input: "^save/Assign_443" + input: "^save/Assign_444" + input: "^save/Assign_445" + input: "^save/Assign_446" + input: "^save/Assign_447" + input: "^save/Assign_448" + input: "^save/Assign_449" + input: "^save/Assign_45" + input: "^save/Assign_450" + input: "^save/Assign_451" + input: "^save/Assign_452" + input: "^save/Assign_453" + input: "^save/Assign_454" + input: "^save/Assign_455" + input: "^save/Assign_456" + input: "^save/Assign_457" + input: "^save/Assign_458" + input: "^save/Assign_459" + input: "^save/Assign_46" + input: "^save/Assign_460" + input: "^save/Assign_461" + input: "^save/Assign_462" + input: "^save/Assign_463" + input: "^save/Assign_464" + input: "^save/Assign_465" + input: "^save/Assign_466" + input: "^save/Assign_467" + input: "^save/Assign_468" + input: "^save/Assign_469" + input: "^save/Assign_47" + input: "^save/Assign_470" + input: "^save/Assign_471" + input: "^save/Assign_472" + input: "^save/Assign_473" + input: "^save/Assign_474" + input: "^save/Assign_475" + input: "^save/Assign_476" + input: "^save/Assign_477" + input: "^save/Assign_478" + input: "^save/Assign_479" + input: "^save/Assign_48" + input: "^save/Assign_480" + input: "^save/Assign_481" + input: "^save/Assign_482" + input: "^save/Assign_483" + input: "^save/Assign_484" + input: "^save/Assign_485" + input: "^save/Assign_486" + input: "^save/Assign_487" + input: "^save/Assign_488" + input: "^save/Assign_489" + input: "^save/Assign_49" + input: "^save/Assign_490" + input: "^save/Assign_491" + input: "^save/Assign_492" + input: "^save/Assign_493" + input: "^save/Assign_494" + input: "^save/Assign_495" + input: "^save/Assign_496" + input: "^save/Assign_497" + input: "^save/Assign_498" + input: "^save/Assign_499" + input: "^save/Assign_5" + input: "^save/Assign_50" + input: "^save/Assign_500" + input: "^save/Assign_501" + input: "^save/Assign_502" + input: "^save/Assign_503" + input: "^save/Assign_504" + input: "^save/Assign_505" + input: "^save/Assign_506" + input: "^save/Assign_507" + input: "^save/Assign_508" + input: "^save/Assign_509" + input: "^save/Assign_51" + input: "^save/Assign_510" + input: "^save/Assign_511" + input: "^save/Assign_512" + input: "^save/Assign_513" + input: "^save/Assign_514" + input: "^save/Assign_515" + input: "^save/Assign_516" + input: "^save/Assign_517" + input: "^save/Assign_518" + input: "^save/Assign_519" + input: "^save/Assign_52" + input: "^save/Assign_520" + input: "^save/Assign_521" + input: "^save/Assign_522" + input: "^save/Assign_523" + input: "^save/Assign_524" + input: "^save/Assign_525" + input: "^save/Assign_526" + input: "^save/Assign_527" + input: "^save/Assign_528" + input: "^save/Assign_529" + input: "^save/Assign_53" + input: "^save/Assign_530" + input: "^save/Assign_531" + input: "^save/Assign_532" + input: "^save/Assign_533" + input: "^save/Assign_534" + input: "^save/Assign_535" + input: "^save/Assign_536" + input: "^save/Assign_537" + input: "^save/Assign_538" + input: "^save/Assign_539" + input: "^save/Assign_54" + input: "^save/Assign_540" + input: "^save/Assign_541" + input: "^save/Assign_542" + input: "^save/Assign_543" + input: "^save/Assign_544" + input: "^save/Assign_545" + input: "^save/Assign_546" + input: "^save/Assign_547" + input: "^save/Assign_548" + input: "^save/Assign_549" + input: "^save/Assign_55" + input: "^save/Assign_550" + input: "^save/Assign_551" + input: "^save/Assign_552" + input: "^save/Assign_553" + input: "^save/Assign_554" + input: "^save/Assign_555" + input: "^save/Assign_556" + input: "^save/Assign_557" + input: "^save/Assign_558" + input: "^save/Assign_559" + input: "^save/Assign_56" + input: "^save/Assign_560" + input: "^save/Assign_561" + input: "^save/Assign_562" + input: "^save/Assign_563" + input: "^save/Assign_564" + input: "^save/Assign_565" + input: "^save/Assign_566" + input: "^save/Assign_567" + input: "^save/Assign_568" + input: "^save/Assign_569" + input: "^save/Assign_57" + input: "^save/Assign_570" + input: "^save/Assign_571" + input: "^save/Assign_572" + input: "^save/Assign_573" + input: "^save/Assign_574" + input: "^save/Assign_575" + input: "^save/Assign_576" + input: "^save/Assign_577" + input: "^save/Assign_578" + input: "^save/Assign_579" + input: "^save/Assign_58" + input: "^save/Assign_580" + input: "^save/Assign_581" + input: "^save/Assign_582" + input: "^save/Assign_583" + input: "^save/Assign_584" + input: "^save/Assign_585" + input: "^save/Assign_586" + input: "^save/Assign_587" + input: "^save/Assign_588" + input: "^save/Assign_589" + input: "^save/Assign_59" + input: "^save/Assign_590" + input: "^save/Assign_591" + input: "^save/Assign_592" + input: "^save/Assign_593" + input: "^save/Assign_594" + input: "^save/Assign_595" + input: "^save/Assign_596" + input: "^save/Assign_597" + input: "^save/Assign_598" + input: "^save/Assign_599" + input: "^save/Assign_6" + input: "^save/Assign_60" + input: "^save/Assign_600" + input: "^save/Assign_601" + input: "^save/Assign_602" + input: "^save/Assign_61" + input: "^save/Assign_62" + input: "^save/Assign_63" + input: "^save/Assign_64" + input: "^save/Assign_65" + input: "^save/Assign_66" + input: "^save/Assign_67" + input: "^save/Assign_68" + input: "^save/Assign_69" + input: "^save/Assign_7" + input: "^save/Assign_70" + input: "^save/Assign_71" + input: "^save/Assign_72" + input: "^save/Assign_73" + input: "^save/Assign_74" + input: "^save/Assign_75" + input: "^save/Assign_76" + input: "^save/Assign_77" + input: "^save/Assign_78" + input: "^save/Assign_79" + input: "^save/Assign_8" + input: "^save/Assign_80" + input: "^save/Assign_81" + input: "^save/Assign_82" + input: "^save/Assign_83" + input: "^save/Assign_84" + input: "^save/Assign_85" + input: "^save/Assign_86" + input: "^save/Assign_87" + input: "^save/Assign_88" + input: "^save/Assign_89" + input: "^save/Assign_9" + input: "^save/Assign_90" + input: "^save/Assign_91" + input: "^save/Assign_92" + input: "^save/Assign_93" + input: "^save/Assign_94" + input: "^save/Assign_95" + input: "^save/Assign_96" + input: "^save/Assign_97" + input: "^save/Assign_98" + input: "^save/Assign_99" +} +node { + name: "save/restore_all" + op: "NoOp" + input: "^save/restore_shard" +} +library { + function { + signature { + name: "__inference_tf_data_experimental_map_and_batch__61" + input_arg { + name: "args_0" + type: DT_STRING + } + output_arg { + name: "identity" + type: DT_INT32 + } + output_arg { + name: "identity_1" + type: DT_INT32 + } + output_arg { + name: "identity_2" + type: DT_INT32 + } + output_arg { + name: "identity_3" + type: DT_INT32 + } + output_arg { + name: "identity_4" + type: DT_INT32 + } + } + node_def { + name: "ParseSingleExample/Const" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + } + } + } + } + } + } + node_def { + name: "ParseSingleExample/Const_1" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + } + } + } + } + } + } + node_def { + name: "ParseSingleExample/Const_2" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + } + } + } + } + } + } + node_def { + name: "ParseSingleExample/Const_3" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + } + } + } + } + } + } + node_def { + name: "ParseSingleExample/Const_4" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + } + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + dim { + } + } + } + } + } + } + node_def { + name: "ParseSingleExample/ParseSingleExample" + op: "ParseSingleExample" + input: "args_0" + input: "ParseSingleExample/Const:output:0" + input: "ParseSingleExample/Const_1:output:0" + input: "ParseSingleExample/Const_2:output:0" + input: "ParseSingleExample/Const_3:output:0" + input: "ParseSingleExample/Const_4:output:0" + attr { + key: "Tdense" + value { + list { + type: DT_INT64 + type: DT_INT64 + type: DT_INT64 + type: DT_INT64 + type: DT_INT64 + } + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + shape { + dim { + size: 128 + } + } + shape { + } + shape { + } + shape { + dim { + size: 128 + } + } + } + } + } + attr { + key: "dense_keys" + value { + list { + s: "input_ids" + s: "input_mask" + s: "is_real_example" + s: "label_ids" + s: "segment_ids" + } + } + } + attr { + key: "dense_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + shape { + dim { + size: 128 + } + } + shape { + } + shape { + } + shape { + dim { + size: 128 + } + } + } + } + } + attr { + key: "num_sparse" + value { + i: 0 + } + } + attr { + key: "sparse_keys" + value { + list { + } + } + } + attr { + key: "sparse_types" + value { + list { + } + } + } + } + node_def { + name: "ToInt32" + op: "Cast" + input: "ParseSingleExample/ParseSingleExample:dense_values:0" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + node_def { + name: "ToInt32_1" + op: "Cast" + input: "ParseSingleExample/ParseSingleExample:dense_values:1" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + node_def { + name: "ToInt32_2" + op: "Cast" + input: "ParseSingleExample/ParseSingleExample:dense_values:2" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + node_def { + name: "ToInt32_3" + op: "Cast" + input: "ParseSingleExample/ParseSingleExample:dense_values:3" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + node_def { + name: "ToInt32_4" + op: "Cast" + input: "ParseSingleExample/ParseSingleExample:dense_values:4" + attr { + key: "DstT" + value { + type: DT_INT32 + } + } + attr { + key: "SrcT" + value { + type: DT_INT64 + } + } + attr { + key: "Truncate" + value { + b: false + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + node_def { + name: "Identity" + op: "Identity" + input: "ToInt32:y:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + node_def { + name: "Identity_1" + op: "Identity" + input: "ToInt32_1:y:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + node_def { + name: "Identity_2" + op: "Identity" + input: "ToInt32_2:y:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + node_def { + name: "Identity_3" + op: "Identity" + input: "ToInt32_3:y:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + node_def { + name: "Identity_4" + op: "Identity" + input: "ToInt32_4:y:0" + attr { + key: "T" + value { + type: DT_INT32 + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + dim { + size: 128 + } + } + } + } + } + } + ret { + key: "identity" + value: "Identity:output:0" + } + ret { + key: "identity_1" + value: "Identity_1:output:0" + } + ret { + key: "identity_2" + value: "Identity_2:output:0" + } + ret { + key: "identity_3" + value: "Identity_3:output:0" + } + ret { + key: "identity_4" + value: "Identity_4:output:0" + } + attr { + key: "_input_shapes" + value { + list { + shape { + } + } + } + } + arg_attr { + value { + attr { + key: "_user_specified_name" + value { + s: "args_0" + } + } + } + } + } + function { + signature { + name: "__inference_Dataset_flat_map_read_one_file_31" + input_arg { + name: "args_0" + type: DT_STRING + } + output_arg { + name: "identity" + type: DT_VARIANT + } + is_stateful: true + control_output: "TFRecordDataset" + } + node_def { + name: "compression_type" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_STRING + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_STRING + tensor_shape { + } + string_val: "" + } + } + } + } + node_def { + name: "buffer_size" + op: "Const" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + attr { + key: "dtype" + value { + type: DT_INT64 + } + } + attr { + key: "value" + value { + tensor { + dtype: DT_INT64 + tensor_shape { + } + int64_val: 262144 + } + } + } + } + node_def { + name: "TFRecordDataset" + op: "TFRecordDataset" + input: "args_0" + input: "compression_type:output:0" + input: "buffer_size:output:0" + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + node_def { + name: "Identity" + op: "Identity" + input: "TFRecordDataset:handle:0" + input: "^TFRecordDataset" + attr { + key: "T" + value { + type: DT_VARIANT + } + } + attr { + key: "_output_shapes" + value { + list { + shape { + } + } + } + } + } + ret { + key: "identity" + value: "Identity:output:0" + } + attr { + key: "_input_shapes" + value { + list { + shape { + } + } + } + } + control_ret { + key: "TFRecordDataset" + value: "TFRecordDataset" + } + arg_attr { + value { + attr { + key: "_user_specified_name" + value { + s: "args_0" + } + } + } + } + } +} +versions { + producer: 38 + min_consumer: 12 +} diff --git a/tmp/epout/model.ckpt-14062.index b/tmp/epout/model.ckpt-14062.index new file mode 100644 index 0000000000000000000000000000000000000000..03cf57b09a8ef0740e4ba55a9acaeaee9f747cc2 GIT binary patch literal 22717 zcmb81cU%)m8^;%=sb@n0i8NzD z1ER)CY#6K&1%ryl0!FYzMU9nrXZI1wX2y&6$KB_AzEgfXllhV{grcl%EgDp@wV+(< zjtPs5@(!CiCM+~`;nc6 z>oFKr>%3?&powvhbI_l=G^krxr(VN|@EH@MCWcQ7`aXMR?CNN`mh7C}Xt9MCO$Q78 z1Lq1G+cf+>abqZig^s-N{HK|PGn&6~`83ywWN)oYbsQQ*g-;Hf78Dg75oSDYHM=Ud zRlW_ovx*n(BdqsT)uD+^gYUyuW|SN=KuCv10RU$`7u>*?pfrZ)D^8eMDnUoA4{gS;W_LZernzMl~j+8-L@7H2GN< z%FmJ89n;2!hlWLZPeEHF$j3V*Dk^Lmdh*`kGovDAMv?Z68?B8CLJJmbyFlt^LP6i3&+Y^N4$n@d3xLH{XSNX{LLfds>k~O<08cMtK{m#GQ`6>8uzQ= zDwfsHp7yNn7#cQhMws`Qi6JwXC-H4kdJb8&rHa-E*SSc{PObZcx}N^VolCjOuGa3l zmewwl!y>1JP5HK+tYNJbh+#IS@K2|lS8Z)vzCRRp*HvIr<|TkI@1CjK+FQ6(Y|9Ej zj zDB?t>cNAq=+Eov76miBR4(mFtcSBqh@tMgSV3{!djDwTo#A#7skyFD$Cx%3Yd7Hk{ z1a7oAcZwHtJa2c|?q#|+nYx~<|B?07k@Cf#MT9}PomFw5Fny_SuwyCm8~uad%?u;I z$uP9iQn)fsTeCHOQ4zLbI)}{*oIKNHf3p3|THl~_k7>v;s;qtNzx<$I6K>H8qsIMtAyCi0fJhKor&ZdPn1~>08{i ze%s?(?pfW*^wq`;F9GKDnf@U@xc=n;+#Ql@?{D0{3Q8LHvVU7zyW-WRWg&Eu{1VbN zqiyA%LyX*}eIi0@!f@`ko&RiKY1O{TP5XX15cM0K&u&!HsX`n)S8JE8WS65=r>`k3 zahVJNRIQon)z&zZSj4q0InuVD7b%mBh0t8boo?kr30?nV|5Qx$hlZ0$RzbcrrDLog z<_XATtAW^KS(;V?{#j^kfej>j$^je3 zov`2ql0qrEck+H>AO#DdK)ZWcTqXxnd{_s8!aq1!;sPm_ZvZ0luhzXCoSXEU~F1kA?1rds|drG8qWZo(feQgir?SNh?)d6B03+H zAM9JT#xF49?ra9GUgklu$&O`v+786yK16^)z=9nFa!3yMLc%HOFmUJpcpt zcmM2Tym>`;xwhxJbR6etdX!3`T%~Qkdkb9G3IMXL-al?*+_f4?Cr|ff+=o9(b-=8j zo3_>y*FRmz&9DA?w<#@+`#b35>-V*IV^Fyagsnfs%z9!37!5v5kZqVGen*%V%r6?-{h?qT9UEIh96kR*W zF2{f)bzZf?Ws(BmV88#BuW=>?ly2V4lQbh`QnL`sv?sMy1Iiz3z{(rQJO!%nXF|Z` zJ0o>M%p*|0JOspEk9t{62ntk}A7=B9mt+t0#^u)mu=99+vCud_O&?*~c6X^0+Jcl{ z!a^v2-QtT|+Yr}OB38h3#d#S??J$8P(xX7cH!Qlt1i~s%B0t7Xs+K2_r_qI|#sDS>~sApE%Q<88JBr4PB@0y54Y!0&KoNcs_i&|!PVDYuq4euWX2 zbOyL*JNKP2*|Ka;*+7)md!TjvZwJyzmzW03GWtlFw`iER6A}Q%x;{$fiAoNoegg)~ zY-)O#s-S#pV1TRX4;#$;0qwW;Ew47<{b+1}QcLGi&Mt(*RGiFBz5D-WOPb=QE(73~ zn0eJ&8Ncw9P|6Ojk-r~8u~ zD1TRg7KPy^zg;pP=Ro<=J3ut=FTM5y2g>8`vdi({Z|iK+LxtJuDgX{N4V%PspcYDw z@uMyACj`^pz(Q!QTXK&_6C7B11DWSQLuw%etTpNM?PvEK290RCl7$w zdh%8r$$`rE57|IhFV@|K4Wt3!)V9UXc!Bg#>T*B1EisVDfCfN;-q^fcM-C(sJqGTY z^=Kt7kVsMl#M_Z|C!6I!k*t_$5-c3)IyT32yD}Ql0k9-9`3u2;ZOnI|YBx7r?lC8S z{0?V^lxS%Gz6((Dj?dc^D7oDaLHN%0jk~d5Pfw1O^oiSJk zDgijP?s+lKff^{SztKvDaiAW`sXlhkunrWJa$}!J=~{*7Kq-{UJ9|tbIZy#iLACL# zu?|!N;5F06lIK7jlpNH%GKmhPpP~7oWwEHS{u0@N^gaV{n++|G;2cOl`wWElsCHZM zyQ=L^T!hm5Vwk^l6bKgxa@JrkLL~sCKdqX{y9kv-DfaNNK+Hv`>RUN`u07VSS^x@f zM~&s#)c~ct^ocd3i%@YIH}&b0vTk56LS+E7uYKnW?;=zQrJ-lXS0i48YFG%JE*>ky zAITS?nv=22Gen%vyTx5_7onOfi-2hUzFXlBErx>4zi;7B zx(JoViP-qf=l$6fdvB-!VEo48fAQ`O)llkaaB?Bu8|qjH#TOk|IG%iOsN23Arkl9z zO&sptP`7^t5W^SVt3kR5)ty?&2AU8zr#3baoy7sReRa0HKoTfD95FJG7)Z`SD3Ht3 z=r80z@@uPLy31w#-(L;H(7=H)W?h8JU&b*_f`xzkT{v<N#On^&Bwoq%azP z5!%*#0qUW%L;CK_-(=(d@`4#r{R)kUPFg?J@ZZQ6q3ZPYzy(B{ooupW*`785k+s_^ z+Kh`(b^2DOfj}5p@Tn{IB2)*!|J3al^DaW^46g0)8#SXa7oid;XUv>D59?Mr07}2L zZ+UK2LFwwN)f&=8s1_J1RImzr5o!RSp`<|#-bJYRAUE|zNw%fLi%=O0p=I&9ChbYS z2$j8&!bA65NB6~Dgvt)401>e}x;|cjzCR&A*_++$s!g_xn~W8p5`cT9MM9ncHBb`P z`!@+AKs}WET^}_MD?m{hH@3YhVK+~JQYc0Di&{()paPi1o8sDI1*isK%-)1No&a@F zx^TDpA))~3w`hLoaEV@jXbo9_^pd^6-B`P~2Tp+W#(hA{*cAHd2Lhy1(%9wjbl)+^ z^hjYuA^~93myxMF0m`8?;@W&A{(xZGt5^tSn(`=m4_<(k_m6o3R2|w6`8)-X9v z(75Lh0Fk@0-E5KoRe5qYey=6|d01a)0oYVUHlF7T1C*>1*ZxlQg*YD#fZ_*U8Mcq? z3vqEea1Z;Hb;bEY{4N8C(KhXSlLRPUn#l&5apCWOv4La&Txga3krzk_r91yVZ$u2F zVIdSKMdI&F4y1`c1k-J_V(klDAWiCFAU5g54a^dt=I{}wNwDypSd(kITN#Ncj&MLk z-cT>R0Da9DphUrSrYP#;gx}xHhz1ryC++-z@fxxK4Tj?=EyGQ}*Yu>xj%9mN0+Hm` ztd1E1G#Jh?4FtlO?>thm0u+DX08#Z<|MCPVgAz5U$99YWl~A_M@H~eVpay`3o2Gm7 z1gMA7&|$UmNdgpo7G`dBWzYMpKCe8mb-2f+5)gF<-%q>poL1BbsE zg%O|x$|I{}p1}%G4nV^j>)!GNsDjes*v8vP0@MO?et4cYR)7WoR>fR(;R#TDf}6VC zpy(+?0m@hi9WFUujz4$@D7$~bi&RNrKJ zq%a~;0)Hv>t~)Aei=g7D984UGDo5FTl$C$2Z z?kWusH~PGpPZFShYXKWSzVnuHtS>~{IKcgn(#JesNTJlP^`J3CUnp1z#h9Q{_On}Mu{;4vpfvPV@)(i;<-k}CmNvl(Pz6A%w~g|60@OlD{7$`_C_n=Xp=CKP zc^OWA2WW_S4G&%W=I#fa01c<#0nzpSXy&&Ya36&c{)7MxF`w8~yHvkZ9#(+jSPp1) zp>Z)!fHElM4Ldv!BS0mTXT4OUVFjoG;J&KtBTs;OC=DrZ)tn?i(IReUH&%J}#tKjh zKxo6i>+l4qfKo>D^g%=cs#ypfE{o=Eh$FuPRImSx)`a0sl-G5~2~eH%1&AxP#&!RJ z0M!S-vddAt$`Ny^#E3)(z>BQjXL$moGr6{@g*(gf2L#hz!a``SzlyBR;RRTE|ClF0 z$+>dKHz9Ryw8?RT#=S{b<(^Sd-1110040xO7)l^?O0f>W`a%vseUF~M^L(L#l2;e| z+eBYzSqQ}svR~4i>^LGSB5d+Cs2nD+S=xR-JAX&;%n67O4aS9hmc4!$8DFHohm?c2jc@fhj zSh&RL!2r`A*BH%+vN<5&`9ycT0EOlYPzRj_mS3Ka-`~uLstagDbkc_Pvt(|aNdi{zy^wLn~~UHaGz0jie8GYtg7jDUJOumaQqP^-oK&O8Acpp+Wq?|~7Z zSj2VxYO_}?R)8`9I<+2kgC{^GlyU~l@*xRO15CZcrT8mQJpkQuYv=L=C|b_VEzska zWkdl=SqLpl>w^o@$O4q+Y=MVf@IUoSoB*ZEcK|Uje*F}@0R4YLfYO{KcGbF6+0qFs zKm`CLYYrCi1gM5mFK7Sz7y;^_{H9Ci`B(wc7rC)NYzbw4$Is~ZEF@4WO>Ww>9Fw(1=D^Gx0DCwJCDj*8bz(VM7$-Uk*i7Y_F?PRnj3|IWP-cp-2{3pJ|pEzK4TC5*WfHEir`4llsU z`^P*1Dp$%NU($e#b0)_L8aH7#5dU+?nLrYta_=5Ce#-f$C$PTI08r{`*PQ1IJ(O}U z=Z6!0A=-xqK=Dfs4%M5pyK=`EFsIOQ7imq`$t5YFzo&cp#qF(lIgb|3w=|LfHNfP}q0V9j zs0ZNaKfCAi1Snd^&28YzfdZlcr7VOF7ti{0kC6o^O}v8EgyHu0{>hssv)1N*HhEu1daRfIuMhbNBv3?pzd`p8^0(wybIPB^cxNc9#OlP=L-pxZaZzc zL-d85h1jJ^6xSpBLcZuGO!wv;FZ^dl*5?5+d)}9mBmv5k^4UO9ekGaMKq>$ly#I9> zFOU{W_K#P%5(61n2nFh#I;amhkm2Ahn68UR8jcg7;j9LTgL|*vHA{eo8wE_0U}5j< zr_)V$DG34VVwBTDiSgiczOa+Nh%fRgxoz*%}E z4Z{gglK&8hx>m8D%n+a?{t44SAUs5k55Wph4#2!S&PR9xR6%KSQfV$mfLbV@p5#>% zD?kGPx;gRjJOPR~anm}|!NZp%Kp8MLZ+f_41*im|OH%#CJOOH;Ti{?tM<9aAGNRo6lpl%P+SdPzNR7H%;!61V|TfGxIy-CdUd;0>C8K z!^|0G^bS)Fr6*g&kwgKiSO^_1>&k8BlLe@H`WCGT!%2M-&f^5A`tS~jb&Hcqe;`2B zas#^@aW97aV|t`8BGCfydE)bnJOLV@RQuxLJp2K{v=^^H1E9I?o@OP*3$XJ3F;9Tv zEgzuQExT6SG&xStxO+YV(M{lVg(N`nu}^ILrLW?cpWZRQO(g?BQ0A1x^Mw*ht#$pT z6Mdm!ArwDQ5gS4Fh34XCm~IjsF%;(u&FwEhq~;WRkOZhH`N{^$IJfy4Hjo~GHP5s7 zcefv<^yq%4F26-el8avo&f3l+}w7Lu*e_^ zP{KlJSt17M`;Y}F*|7m0dSIKLcW?rf+?N31zr9T=UVs5VAwbEFZS1PWwM!_%3Q!Kf z)2e@+=Lt{+CF}RjE*Jr7p%t zpk6Iym!oHor6Wy`6hl`rf=#fI60Hsh05++3B4+y5cf`!mr-^EXi#|yCX{xMI0 ziU&I(-`;@s!%dD8H14ZJAkqgpUMC4q5tGcuKdLRShxLUTfX=b;{0~j)ptLOUNNu7o z=vXuWia(-k@GoRv(CbowTYamA2hJCCVk!`e3yX-~2SaDj_Kl045VTq6v!vH$1-vt)fpK~*Fm(k4K9!>cQ+8v+AN4POMt3}dzdD{!ul<` z4L9Abj6`IQIKVeE#1=2W0P_VXddzk9V9>OdBtF8KgT%W(d- z$B&xqShlBwKqOmMOEg1(x?}~@Kp;%+YF7;_Kstp34lJH|gC{@^TVg)Ep<)(hN^X+{+0m`7{{v~t{QGiMo zLXSJZ>BV%i0F}ij;Gy@(&FP90pfdF|5GSJswV@oj55Yu)hfiU?;S>A4Vanq3?7CQ1 zKV9UA`>9$3K<-8xp|kNDB|Vg)d)BwJ!~W_GI+LHE0nlr(?~XkV0{q`QBkp}RaE~5( z9dhM=@Cp&jR6ry+gbAn`b;nN$9}_YqXhu|6gl#oDiXk@TSbEtCKb8-?NY!xQzM?1y zzEsST`3l#nZXy=>MOY($R89>WhwtC;#+5ffIkIt=jJg&M==L~a|1_;aQ6nrV>Yz16eXymdfEpB)Qj?-S)~2YAb`&MEr>L)v6xFppMeS=y zQB_{eJki9;uUoxD|FuB> tQoi0Prz|WZsA4($|LnhdgiknJDz~sT$Svw1=tuvI_(ZStnA!Wz{{emul3M@( literal 0 HcmV?d00001 diff --git a/tmp/epout/model.ckpt-14062.meta b/tmp/epout/model.ckpt-14062.meta new file mode 100644 index 0000000000000000000000000000000000000000..6bd51eae156f266e590541ac9ccb825c8ce87576 GIT binary patch literal 4075388 zcmd44Ymi;Xc_23T&Ts%U9t18yYJ5bAFl>qRm8c;Gv}g`R5*)rQ0;CC$O0uiHbussX z7;#>jdqDz{Y8R_l99vR3N>Wi0d-FryM4LEr9Iw6hT9z%_CB?Rsb*;SiI%_%Bx?(Gq zs&qA zxxoLu2LBfD?|b0ildom)%Z2H+-SO&Cc4Xzyky|?5g~glZ5A5&lWA9)K@a(zX5w^Io zy?bRkfhYQ%bH#o(Ke@8SA3wx8=X&!S+ta%`_jFEvXZ5n%*zuKKr#ruJ)8hUE2M--S za;$gDZLhiYdZ+UvoesMUo7&v$9bh+!_b##Ccd%o7Bi3XU}Jl_){xN~I+%P+B$x3U8-kH@=%(Z<%Kw{UT1XS3+uHFt7}iI0|m6y1B~R~OmBN_YMM zdyXwWGMS9`roF@L!1@lnn6Cf}9$*Wj;dHnHK-|K&%{DhC)4Mn#zDK}voE-odw}yKg z?->gSi|*3=55nHBiZ^b&wB0+%=J&>z01W_DkFlFx8E#$~!;c00Q0{o3|9DeorFRs)Ww;i$g|zvh^- zRy$Fxb}PGO*?VaMqSc+; z1M7=#PVtSvZ}#|a#5epm!0JjeErveF79NGy0TMx=7tZgkUnt<)YbO^Tn@qVpieHuV z=0EY@%Abz2{TKGGY_9{`88o`g_CLDAe+j$*axQpS$~iFC!tm;bl0%DDrr>u?=Jt!? z_pzH`rOC9oS+kgA)0-vR_7<;&ao15d!!8-gif#3Jokr7A&sHv?XAauO9#JIY>3~+9amu}_gy3|{^ z1WW@;%Ur9TXoBKa|?`1UtS9dThM+${04c zH`>?+1vxc!$vYIlQ@MC4@OoxGAx=C?uKq=b9eH9CwBm5L zG2Xu=xML5q{ZDM}?Cd?Z@rvSR;xT_rob&8{`GT`MemkMkMx|KHAK-$){-+Mx|KyN! zBB+mJY>{&q5NF?o@|)f*T2N2!zaT{xPU2Fn+;o9!mjWIuxkEn&LMZu{lUkj=b0=?R z`o)5PaPC)cEsYp??!z0(x0e#9&4CBmz9;v_paJ)4sJaO(BCs%AgQz@&4jefK<_ny_ zhvegrzx&E?v(Y==!uCHs0*U49ZRimnZ>J51FW@S1Jq{cfA%y_soD%7v`@t6E39mdpK zUR*K6{5UY$LC|AhyAI%h0xXwL1A4i&IR&lCSxx@3wVvdASz(LMtUR+bnQ)3u;RLcd z$EE8nM?g$h*}^m9@(Ifepz!>|>;QkZIm5$2`^?Tun*SmP?Zb@0r_b+Pnr;oR!r{8O zx%1M7@I1h7VDQ7F{1ZQ`QCNXG@PZ9E;?C_%djR28IMhm}hco}s+2PA$MM`-6zUoZ1 z7?WfA-E6Tmllk7cal!w+$T5D1?E?#vyDNMfV6#v6hTD_f9kASom2GAJ;iB~iW!u?y zHTyoc?<^?Att(p!U;JOoV{$cG$w%@o?MJDIu@QvgSN)Roo!=TYKl&O=r0g z#*IqLc{CRWk{%QmCwKYASC{4u1I@PF05gZg`_2oe>>3;*^0iC*N}F$A>8@Q{+UGnu zW1Mcdi_ZoJ~y8dwh#>b_NtK-SiK5i?G#*-;nAi@_d{{k<4ExUO% z-W-F^wzD|`2XwD;Ks%gUIO#wiHXz-HI{@1byxebi7%Jyhembu2UKw1xGI|NDL+*Wp zZ#vgrU9tz+JUE43&U24O&Up)qr4hls`nRz8bK|<8`Qp&-U?gadDY9oHCF80BH|fBdp-vm-2yrglQNYU$Riq* zYNI{>Av^p$SSdSOFKln{Ur8C0elmOH@q zt=t65f|#9QN1h)KN6OlCV_8K!mh}+Z2XEMH?DxDEkUZW6Zy+G$pn=wRu53@24&;?? z_V@VaOWp-$l5r0p1;j#1LN8tr|9WMtVDb})fbFmY@NzJ^$KaNML$408-cfd_TuNB0 z{J1;Lj*4{#z;1)dCXc&0Gz1Owb{c5UvBT%Vm>ljY%VR%Vm=0g! zxC6pzeG7ClED-mow7kfNVgF!Btk+`FfCz3-awLj4vZZAz^~#=NH=hTeV>BK?NMtPh z4lrIXLTreC4;urn1)}ba;M{~z5D)!`A16!mV7l(VBp+JZ$9@r_9_KG#xpZkWbku>B zb`d@U;U*bR0u*rl0udy>64x}2e0(=s0^h29@nAfJNL1GP7SL1YInP>p z8#Kqo@k<-qCvSS1Ta}=+7Wls9k6wX-kU{x*E0O@wk$tE)0zN>@6}g)&oKFynQ@|Gx zR@ke2btxI%I6Qlj9V;J!xFjdgcpGp5Pg!0b{48=kwbM^V^SSMI4xoJZmGKyo9mcuu z#nm0`(4_o$nTk0eY!a|Ug+X!`_fij?pYH6cs5j-M?EeKjn0x5J*KwjAV)Nteks|R? zAclBy(u~U7rWQOi*cwh=u8z%H*nU{W`Q;KjJl@>g*quNG#Qr$^WqbVUfE%CAUtY`A zEF>#lM*hhD<1RaD0xq}koP-#<-76Fhke*e~D;v8cf5!)|)KX_>%xR#Fx z2U%%exGo;Z$AQV1haBO^Q_+^9+neLr&+}IXM%npormYA3=npOm6Yu*`?v;L>wPS*+CUg+Y^aN!@Fjnn=TBuFDul<4(wHk zw8F<5o!1}eu-*lc*b!o_tgY8tp})fy)LwZC!7{e)VD=wf1LR~hzranDb#9DqTIvmg z00*|f?bz7eglsdrpuhL(J;)+dpF#|;=fLq%p&*4?e*QpyXY1rm7x+IK)761~@;e|} z&HskHk%v%K+HQ1?=RG&MG8H!44RQuP2lRV^=jEoLn8z>ibQ6~ggAE}9HYxuw*ceTW zQ2*`hkgz_=KTb{_x&S}Qk4|hXQ`{wAwEqONUW}me-RvNE8fE17*qLrzcNWLGqfDgM`ufqZ7uIUj*T1H~8*)y-wi02{?Sw z$JtHqfJBZZA&^@Jhn$Dz4xD_0_7owl-MzAM z_e;Y|EBB7p?>}|%dwc)n%l~(W{Ui3emo|4U4mSsr>3Da!Let*zd&YY^lU|uF?!TTb z4%Rn^5WYUndcC7i8nJrDc(C77+2aFi`}QB`%{||l>mO9^Zug!8{oB})fmAL+Y#?yQ zx~p^Szv-}p#*2C;heTJs+2k4E6~3f+dkiW zw!(a&7Y+Z2^~{CJ5|du7)WmvO5hh&}dnSDtoDX-HJK$k~WhXY^#0spwdfA25Vm>j}6t8>7M`?%bK#6nQ#J8HX_2$Nzk z?dhasQU6f6)zwq_RbC_x{7YOimhfT2_g=j)8?&Vv4BP^ObH{jp1eyAEzV+A-b=Y^; z)}ip_-IS*!?E#^t9C$bL-*(uW5?>0RyVkyQ|GwVb+H2pkJl)z|9`C{#zYK}kjqSHgFF{Gu`p(|S6tI21rP(X& z_$4U)8K}V6WVt-aO9RB+z#VD>x2+xDKi^yEb@z8xnaVMJvtzC~--lvP$M}HQ`>J>d zN=c8{>-JPkxR9lC=)T+y1(r}kv#)fs{42@B_xlf$roELcLJO>nBX@=jYpXtROY_A_L}B zw|W*ZXw1A=vOs0Mx&H_}=Kn8VyBn$PEnJ>3a9$uayM+Uh%Nm8Jjy6R9Gp+^)J-K=S zPPG%g`Tf^B=C}OVIw20P6UMLmk8=I`(DQF&tEPU1LwLlC^~XkaGepIAMW(SVPk!6x zaO>h|_|V|3r~2RFC{>Xvp6CDd>x%Wi1ifb#N9bUI|16yEgQ$yiUGMDL+re_+z|5~5 zw|8{u$^>#$Ye%2b^;2Qv z-2~`$I>L4fJJ!2dMA%e~~kuA9w93 z$P)~rGe`WtIHnLN>z-Fv`V7*+(60lrnECz@qZtP}*!bT>t4{ZxPQYb20X3j-&S=)df;Gp%q1 z>D3-DFfOv9yvV_S3Qe6xnjlUR_Ngv=TN~t+VYb%zFWPIGnWO4UkPeEVs%NA}`-tTv zGjWbsSh_JzGtACwAVlmFUG~NZJj*ia#(UMFzq^RRU$o*S4w%_w&YeT4CYtg8@mfPW ze`OrQ427LVq#6E=&eRa&JB?J)s;1xZp!ej^Ev5Z=Xrc{2maw8t9BY4fU_ylyXsK!@r$DgPn z5C#;iGJ!yGP_t=W(@9<$oU*y;JC*shw04^`Hs?PbIcZo2>7HOPDH=<1Qmf){3r);V@=0#X4_x!arWygc3(&EC!I`|s2Cy$EZd z_C30W$I9;}BE8E#-(gQ?fXcM>n*)U#^R}JokzDq_cG%My0I}`#-$MJTo#F#Vfj*K`0Z_F%u&Wlv|n1y*ogQ{3LMepuGX+p&B8b8nj4F*c)`8+Xji)>IgJNenqLW!MAH8riH{I~A) zNqC9`-N0id)o_oCJ)W!4CzbWLq4--XpYOCS*P&c?qCd#~>wsjRt|Z%oQonf1Emt_P zd^wsU2%P}8xTJ(~#nO=R@+%Zdx(_Kg^mMfyk|)Y8vF~!<1^ot)XB$9c&}1;?!=9F} zD*3!c=@hoMAAIq-)$Vnt=+C2#XzS=IO=O(~nj-ilR1C)fs>b?M6MhN`A}Ala=F?J& zHuLOV?9TWmALSzh73K}AXx>u;%d_z?owCeT?O$620>Q5sjCZS@?R66Kp`?5=X_11x zf(nHfBRmTVlTJkp3GBma!YODQw!_{M#}dSic~;m$mW2ggC~M^;{iT@T;NO$BB2-<= zaNlKx`N7&dMRD#sU`yr9)e~1wyzRjg*G`N!Mkls+rYHFDj}y=a&)doFesyCqUcGwa zuK2oFSMT|r6AxNtW`0n9EFF{;1&YCI-Ua#(DtPZ*TM_WS`5Fa?Yb6luRrc06#FDo1 zA5l_ukJ@BY8bk3#vz5PHDGeI089=8MD%+9L)XQFWVSm+S%ZUW<$@#?vO??&3uy9Yq z=T1?Mf~yT*ewy;~YY>Hp{w7f^xNqP5zImPxEJuOC%m-+em}md0Qx8?atTq^32wp^^ za*?;{aPpg3zjqfJPDWPn1e51@G&!kcUVKnZRV`!ysY zo(zN?M$WfgdJy@aciD5TX%P(}*W!8UnBtSh@zRi{X!AygL#eNYGWGTAUG`XOwCU(^ zb_&!3pXmq}l}*ar?KGC9fJ-TH`7e;loYd`>-c}Zp6nS*hjldk_8fy*@f;lT*jD&-m z3Q#Q=40^L{J(GFjlE9{M5lkyfe)#XZ_3$SGl^ZRlV4|p%AYm^K6<)@5L$E%jDXQ^6 zF_bUXd=3mCCrnJi)Qw=@f6`&Eb(N@?u>w{m48W46ZE?>(a6g2&zNH6QdKo6e1&ddG z77b62!+#7Muv%eCxj{aWtNH%$+S}`}B=5r{Z=d~qhu!Mprkp)h&J`=j+);!GQ3Q|> zk_TDOsRNF1;FcrGcn{})ht$cb2-v+=B(VVl;PJ%{AebXL#af+b|5YINcyASZ#0lXK zJxUk@F-Czt7l&4WlD)T1e+WH7Ktr@OV_`{u*zw*<!YJVvbFB|&pPr=719=UAtR1O*7OJ>pN7k+_vJZ6Yu|OLF_hhWKG1wn3YVk7r z?{z9xoYDY9^e77w4LMQSrx4S}*|u})MX=C>T7eqFMU5l&(?L0TX}Gmz3ku|UbVFcp z1xh(@g%h!mUy9tcA|#<}#j4KBEJ;E5zsdgM0+1l=A4Y1I9T;3$iNPEa7Mzp__iElC z)$2Pu?xzb7`yY`?W#_faj=01S(P-;Ic2;%*dq0e$cg>n=w_7{n#vl;@f=xTOuBmwF zZR)f*ymnAloOV__PA1x)qU^BW=+x`p@lEu^%>_lx01zi}(HIg^BbjJH^CVxYvUv8N z!;IrNMy-*j)&SI^)&hW|qGD`P7aU9(I8E4=m?~?Ih^T3Tq6A11;sP!_` zG4l$)@D4`X&X8OGAEac+zY)o?O>kfD-S96{hnL2#B+h5s?0fvYBTC(QtK>3BVKq6Z z9@zcjXFl>K`$X{slld`TOO;lzTML%vcdbsN`n1wviZcp2C>4v6L=5i!A-%hte_oD3 zucw5oom+5{2&Fg7K25}+j*HTFUhkx+wp<>|{CEpETjjKfvI=5o#;5#I$avBpf*fLzYM#9DfCLi~*K z2zEPIp8cbhm66TX%)L7X+ES1@k?L) z;)77)S!<$FOPm)jtE{TPgE1!{4s#v8(tFbXVR=_hm{_q znM3~rVuh$>9r9@DPC#$H^3n5hpuLt~FQOp}O;`j$QPt4nRUxtJp@6VDxw`sI8#Y+~?I`VjD$Jhx$0H0uw3C zRy%+(7~|2#D|tZuVI)@LEW`uV;+3FLH*B=V1s&rb^oY! z8)@AxQP0?#BSMozMuZHHmq|bdN{fkfoR{8yEo!2Gkn~O!luva0=eRMhr`AWTmNoGm zoZ3Uf8 zPGprGGVOv;_Hdq$Sw3DK)%J>2QmEEya!TdW>?g`SDiLzKOhS$x3=`r`ONYIW9kcdg za0j}{AU9A}X<62d&E zvzQ*W$&OmI)|$#@0lwS9{*{ocBpai{wv-BzfT%33)8lU-IwUq-Ptu)dzPT^JU^wD! z7Ib>OO0L3+N2wKjzZLeaTxuwhErJ7!3>uWqTRYq2+0MZ|%fl&fIIqWCo{TpyEyLAP zd)Jm<9`9|BH?8Cu90J?x;CEXCWKrNf4|ZTP#C& z=FGB~b4zgeqFGGmS?jLp7_aV7s5m15htAWr$STS7& zB*bfr*spd|m{X9=5qw#3Kt+(MNT+_j&Ls`%fr84RJ)+sa)pZ?Yup9wT#G;nFe*4`A zOA*=so^%#EBON?`ZP830RzSmnSxtE$PIBNpP(36@L$Lp|!%opv^BO3YM_|ywKKV_` znI~sl$O@DsE~&u2l^qoQ#FzL;Y`DC=w{Y&McRU(0kEA0HaFR^Dm&Xm<_@*3KEtbuR z2RVF>LeFYxOTE;>6{ZVs*>t|RS_WL;UOtyYl4(r^&F;jU$-e2%`@iY3zn7D4QP(}c zG-PqyGmyb@+rQCe&*sFdcG_o2Et(6N#^;p|&4A*lRm}dV%if+5@1?Vzo03I#Re6K4 z4B=^#*;KKLWu0f>gqnekJJ_kZp&k!lUnbh=6|8QdeF~{GW$szxD9W`1r zY(lGLErGWzxKww%H3Dk7qTcED3tiP7K!eT|aXEFA_HfaXGeUyZZ;dTH*cKuhk3f-} zImH{UC?{br=rmVS-WCt+cM{GxqVrAiK^=`(L1cN+EnII1twGByzZxsft~JM-taYW^ zoz52(z>S@5YW7^rzDv)Z8`>;fe$X@$Z0;-Kp_8-c<~!-Gx6j+o0vYaaTP7Q$X3y#1 z+3xp@0HHhJ)a*GOz^DsO%AV8L6L~GX$~y~oO~2Y=j!^hb=GUm%b2_xz5hrEO>F;%@ zM$4YlUte;4rCYO5Ch7u^v*%<$60+yycZw2pQiw0%m2lPI4aYV$dv1c^x6`B4<#s3t zzcq*=3t!a?cHFsV)aXxVe~iQsV?H0HBl6F2Ci)MeD{xdp0|Tua98 z4VM{e_Pm5L?5a17YgO;ggLwdv1W`2zVkEEAh?%a1baPsI0 zC{Mo*Ux(o941Ae7C*;j9a8FxKyAQ`r8vD zVb%s{n{4U|t&>TOPcTO%)%lN4L89>=)n!uao;T+w8OKu47wY>>P;P#5Gn9X^4-l(h z)Mo7Y8|GjoV`pC3!w<~%yT~(b34C^Wy>*z~Jh*sX%$$0m968le?N& zrG3+`Z}Jg0$4JhOLbsNt^zG7mY}7*h`10+#Lfd<2EpO_j&=rzza@7}zK^JFn9_uKl zS&@Rj&eJgADwSSwEJAsc9Q8~)NzBPiJ$n*X3)yYJ9}NWUHxn5iY{=56CWUI19Y~5w zIi%HGP1hg%CO1;-T`gryR%%S~;PfRrMXC`@j6rd8l0FfJeMy!iX@KX_xXhk8LfxQu>_v_d5%)(9k)9+F+LLb8UX+Qd@< z=fp}(u=J4ZO5>31N@_^91ieCKvyd!Q2ayV`%P1^NnRbG*B-T=zONcz}?5hH^D?zr> zfmz5_dE>IxrSlC~?6k#VxR2{{j_H*LIgyVvAZfn}AE!+gOQKyWhn0{A(QipLh#!*I z`XJH7QhMvOtw!;?%I1y&%9AdA{oO?Mc{(wlWe_M$c}jRILAG>asGL=HlYnTctPPL)e4x%+7LO zWIPiZFi&sJN`iOTzwfXYvy;>s5H~Ym*zbPwZ6+Gm?Qy)n+1FylfV(u75T&ua{x6^O@f& z%HC(9UcUuc|BHC_62kpf)hlRg(aK_vXtNbul>Astd?wa&>u33_gZmrsgfm4dmXSx$ zR^(>zTuK!$fh7^`%49qMbLZua?U(HG4v4nJM)V;la$8&*VEJ1hMbPuQmcmita`*<_ z3l=8%+cOo!LoY)t@K#&eQVvK-d=~Irp@uo#&MmCA0t(hV|7LueWDQEUEIt`cMFx%u zmd0nnGp-qe*o3mkSjfMQO{Y9Q8Jm^RX8}71;5QP_Oj)b9SpG{Z($lVf(Nu|Y-l#n8 z=6&xPv+V2fFoi|fTK#;IGMCtD#!7nAjf?5otI?V}1uSxDSKWi9SE3N2LC(rnElRFv z(1+sVQJrnL#RPe-_iDe7Jas5v?Xd4Bp7EL?=y!p|GHDK~8`Yu&>^T`H`?Iz**Lm|Z ztxyHUX3@Se2P1XnizkGyccT-4BdaJsKO<+f1l+1eiBtS9_4T;koOlogrHkE?+K{en z`3AG@C=T@1Xd8(GbGAAZxX~__4yrF*)^#On2(jg8rM0t|s7BG%Or9t&RHIXE5~&EY z)}wrhZeH5UR~0Euw;ydUbI@tYq7R#Y60ZbvUK^-Okyl+eljU~Y5*NFFmCXFQDuv4Y zB738n0jZ{=%_f!U7>DSW>7IaG#HdV()er@!Zd~lef^{l!5WYw^>Lmz9RZ5~R<8>)i zAHcv@3PnB@ZzGU+)y4Q!-ZH^2JafWRwhDgo6ErKyxlP$S^6*KWg?x}b(7Qh&3u&*A zDR>0_VyuTw(!yTe3Lv$J>tHeEY$$RAP<*J$1C7Y(#FZPNtjNSko|&bga7Zu)l?&bM zPdn6@z0FH*Jh>5eiK7#<0Sgr`$@F|_xV1I3GWg;qM0wryQ(7}T#NO6hZKN5*3Vzk# zmSTr^hTm62Je@0YiuEnsV5NUU>tUK8A4*(RP>)5kKj<=|Y39J|hMAjTEnPj}?37zQ zWTOFRD55>(BG`#MH$yN{L*#?^6CHLg^|aaA*@8PTC(FE&2lV3T$~wOga{$+yZmf?d zs$DqH2@DYxc{T6hwS#grNN2TUw-U1tH{f-1YrTcDu)%0Yb!EmCnK9A7qt@nrm>pa_ zz{gXZ;J3SqX5@_#G`GWbups>}kH@U|#OHckI0Ts}u@n!D9t`+%5Xi_~j)v z+g0KDmEqpT5McJp7)6fYq&+ey*BFTUib-!jTU=e>49GrBj$Bsi&fT7rvwSp;UW3s9 zV-?L}{2CNW21%8NQ{?{{N&wx)80+upQ0V^=)A-r5z=BU9kyzLf~+1ynx|2oEjxOQQrc)5VbBd4#Xn4O3Pdv zPlv+cZlhi}H?3$`;$Sn-sV$fnR6Dfn%s(<3wcYfoHWao}1#$BPh9dj`dX8*tPhq+o zbpJwManoi7!>tZ==5+hyS(M&gg*qN!A_l*v^0}OFnoSHDc3&^^;B8JcQ^0#Yo)=kG zbf?yzVbABE32^kr1@HBE4S{od6zWY2`K7_8Vn8It$owUIFh~Y6m<+$?xI`Q8`tX76<23Cv0&hOO4gPjO`c#}cdIy@Lo22&`O*x6HIU-G?Qo|XEqQL*78 zYln1fxO?3ajF36mVZY&vmwJ}8I8=;^Uw(9Fdjin{@sBN$o`?+n-Z#%Ce8bbg0>bAj zdEQe8rnm6K#^$CcxJ`bGOYZ^?rPAM5;#RBu)_il6F~~rSuMVcycE=*39?0Y^@lXZc z(eHu4Iw4wbwssh}LDat5jxq2a|1OKi_Gw+LDGxt&)-sR+27K@52rNp0$%E%1xJ z;jj3jy?I#KGnfW>Le*g6^O%qa1Y8v=1E!GZRY<}ac)810l~quW+1zsHeZ(QLFr*lt zkA?uGU28H5#5O&{ijf5XVnM3crS`-OzP>)J?s@D8#WMeh{!=)J|kU{lWwAY?0m zMi@#Rpm%@m8RFPTkdYMX@j&ufsBtA@O>UJ=b0AC#);^YCEL|9GU+yjN|9d(VeW*^^ zPy!i6V7?-w5tR1oG}IQq{xJJ&0vW}!WqgKAM!D63Y(H=;&j@XtrB*7m)wLZ2sEOSF zO9^Nb%QoPC*eadf1fJu{LSvStLU1|}0;Tnmlte;uYl;0{hrJz5fXGa0!+BK?=z7po zY`J&0zaA7>#+S3<6U>w5jX~=H-ZzIVYPrX?+=rRgX$2Z8F2sZmO!s^y6RE5@RG9&y zCvo7ez!&lW@>6@H9j%Q+8VbRRN*kPrBow~&91sB;@x}Gw+bS_ujrn8|+k(GNx zYP!UXR@419v|%v3~!Xk5mV%0$WW%E9YdhuvpAbDm@1B& zIvDnx;2IfwdGrDS8feDGr-NJ2h&WFz_Nk%?7W)O_c#Hk}waK;2?(V&bXmat@{XI8O zAp=ddZ;T5pe~l%sZrpm&KGuSeAfny^I9JfwC%Y6T6J&7&l!%C;G1c1;5tq3bqJ1HB z2Cc_H z+QJ%$=Ob`A*gpA9LB@+GbY#6D0*Hw;i$+)M@_y_dZ; zs8}_q)|8*EJX4PvE}m!95il)2k43Q+4tnsat*{iBbX2-!-ae8nA1q+#m7|PICy?K} z(S($`5W303e1uV>=gd(yo}vXl{JPP*l>^zDx{W|fta1Kr-*Al}7nWXJL+msH3nV^d zY)XDq1ry$zNFAsES>Y(I+i&gTl^^$APFY_tbQ5X)(-650A5&UOz!g^teBV$mZwk~| z7easeuD1Hgm<=T%kc(CWgm5Fow^h$|na;;8>F^Qr*M0L?otFhh1mOc%oagOnY6*0X zD`v$Bye|yD2i4gNq!!AseE0~lM*oCOW1FlMZ$>xIej;YF%2yj>e(+JI4vZ$f_|bKK zZ~X$)LbV<4^XZuNDYq-~kR_6-B%$$yeX&#=zmRafBV=?z2B=iP>d|fbhYV(6Tc_#sD#Lz=K zr0`%SQC9I`C@KW_ke1bZ*-G!81X;~FQ3n#m!9%YiPJNC}=3ro3gt!9CezxnG#axr4 zl2VSmrr?DtusBPE85dNYFPztJi!)Thd2l#remG3wifjtvDqYnUl;D{J6-{^ndyN&; zxQp<#o7R&!z83LPD-@d>lCw8rwvMM*KM~gWtOl?+FU05Be9AXt$-5tOZjCd=Cg_IX zNQ=XveT(1s^@RJk>W-(uCe}phh0km7FKSHrNr@qRyu+TNS!Rmq;xlOi<7Q}cpMInp z4ocODsx~F10WVkgBhRH+kqswtPNPYT@80ejOfwiY^JU*)*YM(m_?e%O1wK73yyEh; z>gGk6$dk8agcG6~ zvkA5;dc-3NztwR6a^y>cW6IMt6JRmGR|-gwo=L`<&Sj)}aOon=slSrKVxXy#SxiJ` zN?n!wtYFV{Dp>*AIO9Px`+4MyKD9iYLgy`?ytq6WZ(dq{_m%P9we%ZjiPszjm-AKf zLKSaXB`=s;&*z2JtmyRw!~0oKK&bd-FsU26?=gtmaKW-Yti_6!-C$^WIWV7H{m#7f zi&p&3LN-E5M~tgjG}FRL<<93}Br88{Qk4B&7ai8rf_*@PmC^dyQmvQ?*XKcZKmkiq zYiIE5U3NArcbDf5Hgc#PP0>L)ZC(j;bZ^cRH$=FpVL+1tw4QIa3VOH3R5J&W>E9!J|P$xzW^&TOH;OfYWy^SXn{{(}+c{Xw4Gs84u?Apu{dbZ9TCop0iZ!&H>Fl{GlE4_D z!lMkQ6KeDOMzMy#Y+}||Sk1v3bRpwSMc9B(6uz*>PMja;eRIeQPbl~GOxE-+YY99( z>QE1%QO7}gcS3OVeUml4?{q~Ro9*O%7@e%?{j_z+mR`DwUAIRKDx5)w90?KlCTsdZ zazr@^d(}bKS~7y(`8S9GDz0Mk_~s-RwbE#!9<#y*NKQ2xbI3NB<<#7}xwxfqT!TW~52>QbO+)+UYUo%#>NP3U+`GA87(ClM&z){J_kjRj z^V`X}ck{(z03DyNK5c6jvSAw$Tb+`IXPFm|<{PRhO0-|J9~ zmV2ka-n}*@i(p5|%&56{G9U@Lck(;MKOjX*uCL0 zL(RRHP=-DErooiz-I;uOYVN&S*ir+CkDdXvnsRB8jwk2dD+pb|CwVf}2q$i#&@k_3 zZeJ_pWmXF(kDh?K`0Mb+CtaU`FVp9lXW)oq4YM`+%gyj|5ew=H z+lScVg>sT=_n!IwowZx$Z)LZv9)nxvMF6P--yDLkGwg@)M%Nm6Iz+X+3pdd!Wa9`0 z9L{y&s{>!W>A*rL9EC>)Wx2~h?03@J&lXn~U^ff=A~b9WkAFA40~cVZ&183HGDc_S z)t*$;!8(9yAGU$6a2>)fM=VT;<`_Q10ly&d1vFtFAy!1mEy8UVB|YW6XE?4;0DiB- z7s|XrTtf(kgloN>inXzBqZ;yf^iMH1Ns;JIwLgV#{Th%k6OVuEA9ijh8=MgC9B?L7stcwe`=!B}LbixY{9aL1e?rK|MYMQJ1nrdOZCrkP!f=+!X)f{=$wz*`yao$JKP>Uwy`x? z9&P9&ru8U?&k^%M)$%=7eP0~4R0uBuAsrpn8#pi#taZv0L$DriLFeG}Q(R+k8y%Qh z!(g`Qf}bHWo9;g)F_73^L2@lB^d+E9qidIDvH>wc<%gv6%s{m-x(q%$Hov6CZXR3&hcA;nO!ZB|dbC<3l`?5U90kib06Zit(aBGDbxwvi@MdM$F$U zE;7bUH>Nmw+%z2wD2-0iC&I8V54%)irg51)j+zcoYCv{nwgFj)4<^OS%2P7CSvt7K54%lJ=|caoS|DB-*8NSP3}|<1(iu zvWA>sRbK0ZL=#IXlhd{u_2&b-zG~{D5RP!~B<6GNlho(wN`0mhC{1}vc%3v>pis-b zTLwh<1uF`@IlMOBgV1F)b570=Xn1e*<9gQl1ML3ZeZE=ekp+F!N}f9h2VrbPRWiJ0 zUwS5RF^|p;llikbaMja`*-2{MFKtF=d_F~gW|&}n{@ho)Y>*pBr=GfA1wR`}u>)ro zZUcfW`(x4QSsv8056wq=V$x+r#i(XYy<$c(P@QObuuQ1Y{;Mu~HZ!l6VX|0PM-%Es zyAaY}YSqQxFs%PYym|=%#VU$Tlx3ubHWRe9I$Ahi$^xK6tBfazWjDYdb=mz5IL#SR z6-#0xXe)9vI8wTrU2^e)+NjRv(YDz9KLkbYRT~YkoE-ykSbwL()>1esTrJQzcrd z!6v2hoU^KcO4%R(^>~=VBCM=kt!CFXz*=-{pS7aUtD;vvptGIefTyjVy-Fr>=1LSo zG)Mt0SGHXN*hPN{XVKXl#~J( zC$-GJ+F{>MJmWP(ST~>tD^X(g-43eDS`?|)@?8l2XKiV&t59NNdmVUO)`2>dmT!Ri z1Ru54X~3>`qvM4m`&52@M$TvnxK)o5)q?CV@yi~uKqbo6rHkE?R#1!5@(n=kD6ING zWLKkubse&2t3&yfcm}w>bXkd#s3FQ)6zYkqX-m11{C#A{PQEs4J<6Bp=B2%SRgpq9 zsqE*WCMA~2tJ6+1olhIFKZ#d@Ij=PUxnTwC_ebK@8WcRc`+5a z5(J|vC7qn9JSHkrqJ+>iV^FOYYa@_&)y4Qtz(?JZ_z-det+a6hZuAp0d$YQC@Zxwn ztoPTe0?Lrr-hWbOFdt+O^zKi{VA{)*15XUx$yke>B$d5_6<wq* zQg`n5q@3lWarBOj1{mXx@P&;LjG2IDe>kOI9*=hi(ELBKMomN|mO;+t;Z$ax!QJ#w zjq_luZ6t+b{j@LIBBPR}s2=&LR;ah`uy2RED3@rU801M>tfiw`EwEhPaNokFNeM;3 z4G1l^YQ&2YRC+u7YEs3)WPjcox|Om=tJTR{+&F+C@X}&2o#G?b&{bMSM5nkt-A27| zZd!>`RWkqso!WwV#od9WR&DVXAp&hEoS4;vnF&9Do+BIEQy4x6&B4%iyxg$YA=GX{ z9$x9)NT>$_hH01)w|Ra==5smWGhDI!%TqVEej@n z{LLHQyZ0Ib7wu`tFAX6TBD*T-k{=4_QFAFn`wp8%D_Zq&{ zp_a!lX!vq>o__{T54eLRrTusBH3OnNZ}4LAW^WXe{Ijb{#nlNn9D27OmnZ+-d)}8? z;b0_Hlz};Fp<%S?V9C~tagO@3?+R@%tZpb6pAzSN+;jJy_wzx;5sd0)Ksp|0j$`#J zzGSx_Oeg-Eh}9b`Az(D%!A=A|ya%D|EFKIegDKQf?ChzK6#2$6&r1E*sIt@}Ylmc6 zYG<{3-4cw@P_z#F4PU&}v!unLV$l5ZqdVJ^X*Q7Bn0H_S;p3Oc!1NZL*x20k1h>g= zap_&)p;Y?&1d1!RZ_PJH8H4EP`07BW+2=!vk0s{4tkDC3bwW_H`AbL+kA5SeXD2*- zxxTYK9d2xo_e8Uc+dH%5zpT;oPWMc}XCZhV_RI(Ke7Y`e!9$=mU*JbTAaB0pzpT** zlB4^Ru-Bd+La=B7W8d@*=S6^W8c|vVG(=d$GsFQ|*u2z16*zJBs5(q&H#i8vs~(D6 z+lXwW){;U#&t>xtUh#&kRJEdLfE!-tu*CK}b2*S@pWAsgIFEPkjF_ok6)ky+I! zX2ar(_U2(_&)gd12}uV}FBI~CfJ<|I$XO8WJxMqNFL(K>vI^>l`Ym_f_n>XQ`)~#` zjPdbl2teAk#z(H;Pe2JWc>%y!98DUd>=zRDu1hEJu`kHp6}`#32lVK@#lc`x&xjyo zEAVC*{1Tvdf9;uvY4ipeNueH$hMGSMy$)oo$*t094una;+Q)p8tq`dy(A?OPi`$~7}@W2*xS(r zh^(b(nBsXr*Mpv7%e}k(^`OWyz61`RgzkG-AMJsd;gT2`gJ*GC z;S(sPN653UbUkHs$yfv(Tk5A5^01jhl_D$mY8BRRg~}-T_-v2Iv*4pV^zm}ICbTLt zDz_VkO$eahMa8BSTB`-OK$TjvnWVH?+AK<4EDqO(pLl=`H|sD8JDoXX#1wfLGGyRL zY(j|T%49r%^ZVtE?U(GD$ToPw3;p@faop6wu;&EV$k@AKybitE`0#KG8WHC)@39t{ z1{F=XUmH@(?C##1h=vq@r@xcNE`o`)ok7p-8{-1Y7qrCH62m|sauJij7C_h3gc}22 zWS{I(m`sqx5l|u`ipEqAM!Z{KB{|l;$UBqqSWBIG9YG{wDU^m;tYow~)O_FyK`A&f zZCw8#@eJ;fp7!{b7LIbou7#+$v1=G+fY-!YTUZ0}d<5O$xy55skOMWg3n(v40idAxlF zv#}e!TRD)esoMyV+{_RZc1a$)JkHN;LMut4G~iUmBHQp0+0#dV+pWK~gIi}^iF z;faK^ANSo|SzoZ6Lqw8PvAxJKrL_cQ=8!K1>g*CifB9~{`pK9LB_WWDR`Z2$BgD5= z&uyB{$1Umb(3t(YZ(OVMGIN%UGeHQS809=~k85i}=eQt1Ul@K(sFxANdBX~p!VkLV59$EK zX?Ci&91TF&I|v0K_;oS;5HajmUF+Wg>7-buikmZyHZ9S1SRgPBC?;KNW<2=E~-f%meN-aQEtm~)~IB#MKF zUPYYx938s0L0T-@f=>|FpY3`^HP^(oq?9ACDR`j@EY1=Y_0dPES3O)>}$2h9(M zDO{0FL0mr2L&qJ1iYB~(y~YY^+{KKi-Fz+LrB*05Hza3o#B3c;v3??~+-S@(L*rAv z8B5;%1g6*o-4GmU3G*9%-`5lF-zsHdO_W~vyaxZG)1BQX_lBd5@%FS5ps2mcKHgzZ z(JV8?bT?oo&5v}$L8&@X)uyC0;N|LmygG(1_PW^a|Yn4Lgnkt#aL}aGaRmsl^_DrXe6`+j^4qT6R9yy~| zj4Np;p;KWWoNT;cX{q$ zBZpcb2WW~8%4zdT>&n?1gl(9_j2WasFs$4`l%QBiViw?Pp-9lFRt>Y%8>U;v=l)~Y z%>?0f+xC+YR^(-9Bz9Q&2w6~z$1|g(W^g|#X5WRe>l8dDV>e`sOBT>2Iytbz2Nz%^ zIUXoQJ)dbQS{2E}={tdUhmwN!5gQEE!Yu|^^L=1pOAR2Yh76$9M7&mioB$y-NF^|* zAe)n#NJJ2Ni&_AomJuqrBq-?_@QcfNkWXO7Bj(PAz;XmU5sSgJV5RrkV}t!o(pl(K zI%BLAi=_=tq5irDmou$=h7qOMEDg_AWZ`pF0iB?<<>ZbgOjJ?3;p~n@<{8CWgY)Ha z125mOG(xkpy8WiiTJ0(l4DgL&Z49UB;b+)=-Tj4jWMb#8XB2B=&}{F1Ccx2sd(SA= z#=u2Ady-#N@tt)>;W01!NRAqfJbTl4E^7cUd>fTl|BWttmf}@XA=N%THi|W{t1$i? zYSBDQ>XxNHA18>mmfOpMZxm}&43)loZb}x z?c{wJovi8ov~_})UV3Ycr2F9^J>r|J=?BU2(37xNiD_#z1OpG>NI2v2_~uO%BBx8$ zRZJVBiM1kYtTG79VO0PqF>8J~St(1eoD>7#XMoU=qDxs@i;aD9YVQ4&;pUZbdhR`R zK}yqju(_3pD{i_&^NgB%Hy5xpj%!dXo1S?VaN79h)Z9BAJli|B4lYjtKz~M9;7&Zx zACa=SbO0TnuVL<;zMe>tDfKd+2U(`8ZSs4EVC5O^KcnW}>CkFFeYRAi<=*MrPWRPw zFg>H@-pPO@NB7f z7w;JI9Ld%)#*=y^pVR=$5%5GT>ZKcJEUT>q^Z4Z4d-=XR{ymnmuesIbg5MUo_i`&W zA3s59r{vx%%85uXG%g(+4X4AwrJcR4;k38dTR8XBJ01<0InIzW_Cv%PgXhP>M=9?@ zQdQS)P#0F+zE;Q!s}@ckJpozc>+r>=S)YL~Q%i+X@C)1##~Nm9^p~6A(`3Z13Q~xq zY!i^AqCH|k-9N+@FOeGFu^~ME-SiGzfT1>%-JQu8otamAQc(x% z0IGf123jL6h-QxA`Tf^9s?Tu1F9>`Aao9(QT3K!pZo4Q6D_cYS;i-jw1E%Ng4sJb`H|fQN4i!BRbDHQ2dZqc!{90QywFnAF0eH#`lGo zjXEmyGel<7{ih@b65A_Cu0@5uCg|FwnQTBzQ2AOqeQ5$jYR}yMc}r>;Tp>n6fmpJX zPi!PxO(pBmaQHIG!&EO86jx$4ZxU0Sm%0Kraw-`DXcpwMKW;~J zeUOlfuM|u!1|cmf<&d_b6J2fm8Zm#bmNLdnH>Nm0oJr%R>0m%< zbdo+1E=hUVr4loZ%j|K~bbwL=vMY8#)^9d*lQ})p0P(@37~e`2kR{Q@N(<4LUxj2P zO5>31N@_^91ifuY7K(>RW!hyF7Usn}L0J-ODa|EBo_6->I4(>T zli^k#VV$cKY$%zf!?iNYqNBA9qW-66%3{>A>pZexN_{&N zM7f#j{cLe{0dg^21_xnmFq<5^NQc*~e$NCh5*G%ZwIeIOC<0vd^kR0BTK7xO3`hp$ z?KB|2+GT^>Kt`+JXCo;#zyXmB&7jLf5>UQ6TE{*#ADu^;E-NZVHEZe>Gm?Q&v5`CJ zDGyQDU&ZGJq|%`blbz2kL13dryAV>hDw>O`11=9v<5XmS5wBiCK(UHqL7G&qn9_xU z%&*D>ZLPEx&X=+P=+G+Tx!{s>`J{yV8}NkF+a60|BWNq0ZLn;xMjkJFjcZpRg0J?q zpNASe+7_GthoH#qb#H*>Z-Es3PKT|fkTzT`-=N>Y!X$rtrh0i8GZ+hJ*Os=FLl#(S zg56>`9ab~{oAFVVH7MPx`D8Q|890Um(pYK1nUfTD`^cUs{o#cJ+&dArbc=8^>~=VB5bXGz5&*vWBaTXgB=8zxbf#8!uP?v&T+)9?^ zXkEUu4^g|6Aa-!;W+DQ0RSK2)t;^1-UWiR*O_{hxVZTiG1mq${WlF4u$icQrb#iS= zlsf$KVk&Ya2u4*(IyqB$OjM>sq0lsAP^}iLnU{Fg#rRFYN8OV65OTdw0&jsE{RGY4 ztnMAWIGzsc{q?GV@=2Yfe2_iRyFVdGX|Iqewgm2Eti?`}%3j`zFNGX+B-GQ-hJ(bI z6d%d*KqGQGVdq9DD>89XFy536wGCoUf7+pj^le^p!_AGbTS;{`6olBC&nZl%Cw?!t z(hN^&&F~O=TW__IW)Lg*RfFSBhi6nky$12{uE;6Yw{(MLWri?{DT;h3iE9e>2VF+A z+#Gn_D0DNdrK<;=or3BiA3_`G2Vf|ox#c3*2~0ObaHS#gVIjoO>=PY!F7>ogS7P%* zAJB`VE9<;4YOt|A+E^b?l&YX~37Ww?iJ;|W9QvpMSEW>JHE(*$4W?}{s#Ne_{FX@ z+g0IN-R?d^8T&Ljuvw`)cY9LK^3gbY$3_E;#rc&ci;#19IF*?_chlF48u_*JRxZI8 z6ynu554PGyQaIL6`=Tu};&jn^%zNJbYF`3p%w0^FovmI(p_G8MWT@sWuemSv{pORo{T$z1Q%q4z)afLBp5JdHz~iX5jR|(JCs+m;86{H3OnN zZ}4LAW^WXe{GZDy1UyOz25MjJ*!~4PU&}v!r#ji9z$rkM3+w zrV1pVf!7*Tig^bX5I%m13`}p~iH*%oPjH+37MI?hP%8a>0>u^Ew?^hs#vlVRzB-s* z+Z~JPpMk{367ycx=z+jGAzE*?cEq@iDi*s3FN74j5G+kd#0<8FTL5}*&VN~>2Ripm zz-J+N9`?)!^L)B4Z9!$;8r8{3z$O1>jXsbZ-JgWLj3N+Z_-`f{MNT71+pvgdhy#L6 z^HPTf-6%p9s_79Ittlt|Ya5Y`)LK%==eca&!7JX7m8w>hhSxbPX%1q0p1B;zvd`_j zTFv9d4x%$>Yy;PG3bJ4H&4$Gn?ajl=p1C#16Os;|UMS=N0hh)|$P^M4?n@~+gOK9# z)tFUK$K@?|-sg1gzzk!2d>R6fP6FT~SMVpG1ev@5U@Qe+zmTwZT{?-6eL?oF=uO@| zpeJSbQ_r+JWGg&?e5#^%ov!`0XYAG=Z@La-B!zlBkbD+;9mrUdTcy(+2oo0#jaT+D z-()KUZ5M{ymwOBR|8$`h`%l?W0vSbMK19Y^2bY(OpG_d6ShkGMkjW^wT9EArj^!Dl zjgw@hLYq}0=Q#?{UY5Cs*)JuaO)T4h`(dkeb`yBk+#N??G6bg+Ay8T`NlC1vK=6AV z_I5O?Cu=Dhrg$FE^`NKNa_??`Jt(q_FS*0VIVO!GgU$rJtqxh#@;-uTomQaL4~tsj z#;avu_~$d3NM((w$_x-ai32wczK{nXCm~S3m^mkp+vaZ|w!b7s#^8yYR!Fx}Hecy_ z%IMO$TqSE}LzN;c_f|@-6)L0TV=&v}@hn_g9{PAWToYOq8I@b;tdnf46Jl-f%^-k& z7ZsaUXw5P-qEy2Q^4sjYUkl8ECNGN`eZpv z0)Mzv$FSrGH1ABtV=Z;&bp(-!r7W<9pkgp^g`gCim^Q9|ka!07NKbowOAAM4>{=h9 z4T9K$nVldugJA}EO{}$rH4x87Mx)+7`AtE_izjqsy$Yfm;z68g@o_DRt#HtT z^N+$(VA2tH$SnKPbg)9^f}9Qxo_P~e>O$xy55skOMWg2^bB$m&cB6MI2eLJF8-bQs zoc1x!UD1avq)Hkn?4+NfY1x}5w5pH{b8u}{aWPq|%gE6dy^l_WHt zuKK^=2A z%}(`}qcI12QF7dY!F4hH5HalcTHyuDS_=Y$R9ClzFCL#@Ki%~VD#OX-gc%Z$Qnca* zPXI86PbXBMxw9qX@JV99*s0AA?&O^WtmWEqIC2KTOd@sgVJIpD_>dOId)Z3wo&<5s zIZ+1^#lb_bB2MNxI&Hxxi0jXGJ)@dy;#yM5k=GQwPz4reiC{%+TSTSD4H*Q7gXV|B z6t2joAg+?xws3~j8?R`>XMxmML5;hZ@wA(-MZDAs#pZ_O?2VYM<0;lpgp~(W;^S(f z(-GLGd^483`w2|33A!OT(kd3aqjojgVRmr!0MF;2fcw!p?Du^=;r^|<<7u#oHBoxu z^BVk%PIq>n+#8NI;EuWqx~RR$KHgzZ(JV8?bT?oo&5v}$L8&@X)uyC0;N|Lm)YXgCa}Kx_>qiCq z$9#hW15jBbWbk3Ifv3R-OxPcUnv+-10S6v7#L3FWkpYgNA7+3EorP@Jbw>eNl{cJ%U&KeN^twy>PWAab z!eb_=+23^;HlC&h`+x>3QzEma8hQz8i*3u%Epfu=GMpTe!M@&QXR~s5dG6$=DLN>p z1-?lmTo(0_m4)>gOk&0i(jXZA+(DdIkOV>wu8CI^%gbg#cVgGg1mSht_LC7-+1JdFWr`1y>#@MLnNsDOwfD#3?#~cat-O z;rt5RVt_T@2Nw3Lfly2ssUfBgqYCV$F8{SMX9f*Y3CtEZjF`V8RXF1_U``bM!f2F>>F zX95J>xA%-eWyWXOXU!LwRWyK6 zbMGd+u3=9~G?*`-<=)LFf(j*=k;_7lhE6>>_ilme82-uFZCr$6?jj)+z+D?`97fF= zF5e&aj0{K*akX?RE|&RQ{{*iH?>?4XyT+$ z89w$AV(LV05pKIE=_y-7ax*rBU`V*u+o@O^>o%$(k4OI$W0Mq#UZ{k--;ARxe1EjV z?xo~wHI)J5vt;f+#0PEE5NGViWbZU$6^%e|%i>?dXUo(;l;`r{wc&LAa{0rHE9g|u z+M^mf=zT2yj}K09H8dQ(Yj9Nr+vN|};D?Upnf_fhy?GluzS!$U7``whfMX@@kUI44xg)qEm$^@V3S$upB_$OhXpowss33G3LQflsqB5|oR8Z3 zAHTXg+#Wr)u{Bv9ZRjKBHg%pKUOT8SpzL%^q=gRf4*ML@X{wg*vFiKcsHH-95eVt% zsNTSV5sl>>DDyywpt4h*Cn?M(#`h^^^HtwrOz4adWuGB3o9;g)F_73^L2@lB^wF_Y zG2AZAWCLP?%Gc7FfUq;Sf8J!k4*Pk0BwLNg#ge6bVk6mVp9hrD>>3G66Kwqqk=S}l z8)@ffscJ>+T+=QBfHbm%P{6nhl4X8==h77Bw)XbPZ_vSwHOyNB0_exW6NemL+$WuFEfz;X@jCfqj5jL!&lhkEL8TW-TX(q~m9FTJIvy&}FD2 z$L5#R*v*5B;P7RVhpApHC~neF9HCV!Q?Uy3}wQt|DeN6(juFf*AMl98k=Y4eUGhV+wceZh~YO9 zB_U92)f9seofYFngJg_~PIPE~jo2@!xX2hY-I(IRant0j^Rr%$orcorBz+KVn!@hkL>9r#iDgww*iZtwpc7(szIRr=`LQgLDGH|K2Dn~mPET$4l5xK zB3kCOM6!_63CU}HkZ58lWpdh9qwvzv^xE!N#76zt@A^usj{^E3%gE>2C#lcVmHJF2 zP@3|T@K%CsE6+f+rR+rq5q`mnLT?VQjrRtosUpV^HoQ0baXst&0d{}yKHseK$bvos zVYB6eJ?U@|#s;$`!)x}XX95@V=;o|sw?CT$S3SL$outPqhC`y$q9`&-_+V6+aX8`YpivU&O1I5bn2%Vh!^ncn^~lY@~)ZlUNh$>lsW+xW55U zI3ucJNo)jdMQ#QU*2rhVDqcMsmuHB!#peGZC~|w<8({f{sVpt7>~}hBErp}P)$$FR z8!Sxnx06fa!^OfdW-u1ct}X3X4xy53thqf4B5GQpR$b42Gd`-a2BljypNys=1IK6% zY|+)>=6r8qFi0N*ez)VwJ0s$U{X5@ zJ-Zs6z~zwLszdQ`6QbMeO5|7K8Q}WTWhF|YhA3-Ms3)$bEu|XeePqW@zBX$;%9rTo zrM-MrkwP`8?B}5-C6>yo)2_0@JZ!`MBwh*Tyf#pmf?eE7mgQ(&zOxUzb}24)aO-9w z0(DgimHDm9&Z%CAP39Es`O9=qKrUibro?K99Bi9ZH>*m5bt-Y9z~#kM9FrHnzfTC9^o;#C*pHvu1YOX5Sw^*-ZF#@*;AX!fRYf8cmJw1#R}Hy}Kz zlavp#2YUA>Bq{9`GR2m_os6~ENmAL%=Zi08LkJD#>I#+@2rcLpd30t*E($@IkUBdX|?eXds}a{k=hW; z`}KzFYeT&T@$jz5DVDHwLtH9t$2o`vByE)qG3*bzjA*$z@VZgxW>`yC4>&tjf{+b0 znPSqwP(*XfMX)p0r5S=}ftV1tvQKo_xzy7}U5U*LeLydcuB=ZtcD4r_+oO&3@kFVL z*4@M(Zoup2*5swwbStmOjEP*AU_(#JV@+ukxIjKxk(kJh!=@ugf7@YCH{QN`MLBk< zd;*2D;4yBKYcXa&8o$`JX1gjptJ~fEGGwq*Qg`n5 zq@3lWarBOj1{jO;D}=W%Y>Z%H1q_k_gW%=ycy|DE1SZzh3Z8R$IF*^5%i5KKtmisv zchGy=)Sw#Y!B*Qy3dj0sU$jL=oGx0A{8TH{TX)#E!(GCjAy3j`EgjWrf#vds`xZ8h z)@5@&Ucko^<5d(XSH6&;1V)b=`HVl>zmx3OCj{xg7CHwAd7 zlL?-`zX1mMJ)f0?R&T>hfTQ;{c<R%X$787(9@55~b`*S|js{^0#Lc*7Cf;i^UtF1&&S6 z-Fr=6$mjUeQIvzby?5_R_P{`Qpu;1{-OFc_k94Zpw%a87C zPo@eaOYl9_IsMjf<2E1z3kV;-Lfev3=*0uQCq-zQKUYAUIY&GyJW z${3_4##aZb{hfy@@P>a61l9>*aYP`$vV=9OwB{7|45(cLvE;w3(F2`(CM2Q8Gat;i zRQ+0qtE00k4cZpL{>vJDAUV1}348hW9TX<^P2X@{1SqEwrEOTmGsFQ|*u2z1 zhcOzor3eP8=6s0z8~yV^kd4$@Qpo4IY~H~u-jJ25R+PZ8Evm5%g@|AH!1g?IIgn+a z+j+H`$BUgkXT)%QH7~>;fA~e;Y*>8J-aM@AnOlQAA?e`hg+d+>a8)uD#c2x8z&BdH znq(E!5qry>_dUpG7~|v95P)FOGHQ&ET*04!5@hlMfU%H&=CB_7g@nE9(n);m3$k}b zZ}RQ|Jt@1NdPW2xTY)#j%xX>VIw|{W&)BU$-gF(vNDB3MAo(o#;bg4It%& z=2trGE-E&y&>Cn?EBPWNAyIIURdM|rxP_b%Itsy{}uI6gEc%Df}z_j?d7R6RL=)w6%VJR@_ zXdgV58x}azCI1@3j+7A2Gw{rtkWv>yH+dMY(<>T1XHL5D6fH2Ajos+o%7JW6-A14# z);RyRZ}vuz3rjDqA$A&p1ro1VEm#^_a;W!KTn8#ZRu#qd@BwKe(U1G?uB^I}QX6yg1en8_+R#+U+xPdatBV$v()?4M7^tWUXJp1qN|ODaidJdrxx2NJG# zgp4l80F??p%nK%1ZF-Rs6Hbe~gb{BNlk|3arMzK<2X$ECG&|K>j)oQN_ht$$z;!YG z5HajUt?)w2{m(c$!G5~y8B~Uo$q6$gAf;%<4Z(0jOBAC4zHq1s@^!)3sm%}WvkC6!Av4`@L?#*-wXyj9@2t*FI(x|lOV{66Lla_96a@;z&vrednrq@(Qp%Cn6ueLc7H5f|z_vw81Br#;aM1j4n8Fp=6vS0x*%rmbP(C~m$+}7JFgv(< zfG4d_^t$^y-1PJg(_o+S%~3c~F>Khg~crRqdgo08Iim#h1c=TfZ5 zh7*nH_{fl;#JtBd<*o+P3`Wg-**Et!yf`6#<|iaNHTCpzc*W&wbab;H^9>FRtYM9i z!H2;Ho(3N*e3Z7^{oq!o^XOgi7(Zw>lo9e-(i!`Tx(f}9GRLLwRA~U6~N`6+bXF8R5 zfHo>GnvM*b+0P?q^u5c&DRkcQ$&1UA@#dxFcV8LrU0aq(Xfq%L;YoXVNm*ZYQ`vw? zygDnmoUf8M`8rxKx3F0z!Nutny`Eq=^-8SZF3=Q;Q-H%Kbz|2-245To&#v&zLd$xN zgI?(iv#Z~kmsc++<~=6{@6;=PXCWJ=tr|?lg|48@LJTu&&Hk>74r^+`KA^$MO5oX2 zt(XbdYZaMCfCDPIBs2SZmz~YZ-Q~H1jT~x6Q*=;Hn^*D?E?M|Fieybx29uaEgER<+ zl{>I9y77gP0Av=B-6$kKDXa=-$z7m3vFm1n@VagL$p|a*GITI*nFX78@G?qW2KSR< z_FWjePQha`b~Q8KAmQPI3$T(L50s*w&$JY+OF)RjiURM>ls<%846x?=z`~XqsL5aE zEJb@fF=&uVU{0)JIT1nVEouRTT1KeglAxrwM@f@P=6{R3vmvk?0Z+tYCEghT4rVL< zCh076DxERbipA0fr%-?0gUgv#KEsGoY?g*+E42#f1f?w}cVICQ&uUdgw60geCv(Id ze%~n8+R87F8+iGKEffmAcd2980=tR?0~m6XL_SrG z6uzjIvlMGRP-Bf^^@T5H0(8yi=Yz`8tR8I^pXV-D^%J@OoJ0^cas z#(32}{Vb_P^DxQ85q#eMIIY$T>09Y#E5>bVYVht^4I(zPEGZTxxBCE%uk9+RW zNlvoJs-@iWqeQIc(c-(*ei>TZY*o9*O%7@e%?{j_Co#l7^_?1IPQf{rW%s7THn z7iR5T_4Ypvp0eYDzXl zS`P*1X`=L~`Q>D#EWPq{wgFe};Al7<4leQ3;k38dTR8XBJ02y#s+@Up?p@rimY#br zYGsuB+2ZN~R8V#sM~uy_L|k#xc*_%>~2Y z*|~Qk!s4Bknun-(f}T-x?{om89y%%aPG3)C#Z0EW;kkEnuY4nVa2GlEZoVPRKH5(w z<=*Mf-$_v-VO(3f6) z2KK;ZIppSs^b1ewk$h4EEJwf-u~>R=r}wGZ1s2a(G03H`PlvErtd3j$4?6v|s*A0cXGxkb3`qNJy6 z4av>e5P~7$T5qRfZLHgU;6fiK2JpI{@!22ZgEnf2GxlS$cN(#ZMxa-y3@H#Vud}b=vt?=^%5(Yf+Hksl zx%}b96?Cd+?NN;#^gb5<#|Njl8XAt?HMlBb?DB_e@I$8p`>vYayp0`S?Fqf9!RE~8 z8TeLP|14ZmbWN$tdaMu_<{?u(dV`hh=kU2&*n(vf2{xIf{^{Wqc35C@m+GJ8p!CTY z(nw|RL+5tAT@W9Ua@~z|>WWy**vuh+QO|bPdL}KeHZKR!_rK%ON z^9XUJZ6yRm+fyGWTIT0>E=^%>Yj2iuhcW)u+#WwNgorQmRB^*m6zfpUL^RxYL3Q`kG1%rTP7 zZ7f`l1xeBnrFlk*5~_vx@l}}|K9RxaC3%?zDp%}b(1^_is@2jz-X@SQ5Q8oXBDr`q z6q*$Lb)JUX4^!zC$0F2jV>g_5(P$1Ic7{b`^US>Ov6XC*&L0f~4Tg9oAy8{oQm9th zfuyLEL)wZ?WU$z;5&H$zQpT9+#uV=#Vf$6>iHLW=5&qa@dI9{(PJ-%#r_S%K^QRau zjZV@h!muw7yHsMPahVuLt`LR1>`H1#wgkPxd9#o#6c3S#+{-8| z%!_w|vLx10noEd0?d;QWT$n0m#A4@~!K>rC4Or~-Oj!(SHb~m9!pCWo#gb^3%3&qs z9R2ll6jNMn2FX(g16k*574JSNGM`;luL861AgUh+3ZR5l#@&fWAcG6#_r>5enbc#4- z<0=yy7|GKRq}x1%(0pJd_yiyLfPxTeM2lzBJs@5hrs2|!E-*0n4j^E-!%KHygqXGe z|NnlkwfEYOWB6c5z~{gBdhE6Ldi@ss%jVw2vs>LmJzdTD^imR!GXJco7}cVsH(k&f zh?&{}Bv88xZDc5ZJr)l%*7ZJ2cB1iDyQ=twm{)Z!My<|8TKj*Qu3kd4-zkdCJ8!b= z5~^~}ZEddaY&;2v6#7@&Xi(S}3-?Q30CeDpkO^XglyJB}C%oSFR1zCu+qAWU$7)zx zaTTwuO=}iJ+fqFU2#fq)_W~~e32fK5M`A5UqvF-_1SuEKM7S zSY0w^7n(34Xx#~F_QhAz?J@zCe${+7o2Cskh5(^R@%jAj);hQ-Pi*h}ja$6cg5eSs zIu-H{x#?BUXLFM|8mkF`&>x>=mWEor<^I2UMS9xRpOzX?^O$g2-7EUzf0hnYID#E& zSIzY90=6r8 zqFUX%*p;$^T9nE!*wRiDeIeQ0=ZU>p9m*Hd8Q^f~z7i$VLiDvL+#6TRmQ#)L7PezI ze70#l%GdeNOLuuwk-~jaxxa^+lvFA&lpR%8qtB@Lak>)B`)s5x1smMTKFi6vd~Y9a z?Na*K!PL!62I{I5uJSv>&beNQD`qX3^czxqp6>|^i&&K@sTv}W+qTpjjzWjw#a!kN zBv@4``QpsgF}E@$tz%HFmTDsy_^ONXOUOsvk~9c;j&6Y)eHZV&Y32?-y0y0v_SdTd z%DZ)v@|ZX}Ih>KCbXRB;TLO17)nYe@%3Z!)e5n~j7%?V*WI0+SPA}|S#Ik84O^U{w zn{gA5ILCro;-@3Q4C%YNf zZ5n(xJ4zwcYcMZwBhGGp$2O!-rCR|V+Y%O@ZpJn7{jp%a+&p^yD0CUu^34OTPUavq zV@;a~w;aC3d~W#+_5#yo3|<6dV%jR+Hxl>f-Zp9^_JRk)zr8@$?X3+%vH7pOX=6^* z&A!&EAfQF~j#3##2FPa{5}UbtcS$0}??>X*#oPBsluw>}!vCL=E%oLdKW)Otw7;jh zBi^3|mkr*oiqC4gJG?zM&!3V5Z`ZeXHn&f0?W*gzx?}$tc3{((JAZq_Icq25B$8wS z$C7H=p#f}f0%8RK$-sl)iLI@(>wpfh>m)t}w@652T<6UyAnkbnksaQ)#wJNyWPq?pm(`aJSW>3Hbgo>?N@uG}NyNkw-J_`Mf}XucE}b|Js=fiAy;t~EPg;auQ26ly&%c3! zN9=7^NWP*}U?j-ddu2rQ_YGbwUiQVVXW_w?SAvWZX$G`714(w%v-e1zIN@Nl8(_u0 zP}1``R( zu3qL70*?kf*on}G_aOA0#p@fp>w8d3aps%~NwFu65i8AJs8FS^UOOC&QwEP0yP%GD;QFR*fP$R+422DgFFIL*l?mJ$5SZnXW$+9_ zit`KVRS@#*+8b^;=6VSf&j7<1hfhNY(n$auas~edlpwPg0FD{ei#(0Heh|1?GWM=Z zCvn&pWbaDe?A-%^wf%FaMK&XuJLA9NG*qc7b^#M(629oKJM#Dj*DhHbacvvVDApEyiK1 z#IR>XMJ1PWbeJ-*PE9lu3cIdRayy zo!qi+x$lg`Ytg9QP)kvm;yJ48K@W=g_XbnMqsyahKunbJIgY`6=UdKA7wdGS|!jZrv z>~-d_88dC8kkKSd6?3!GD?PIXfR5u&9gKTkVuj7U%ckqltBr$)JJdW1U`x*HqlBI@IX<8q%TXdqT?k$BV;HA5E&3fG-B968m01&j zja~F@E*5T6@3J{KT>d|P+t0DH!GpX0-uwD699&(plk??dP7u^REt}lu`x*!A8 z8~C^`*kskT_4Em+44sA&M!ZREkhkf`j3fCR(_w`};^1VS3@fi$BR%wH=Om+0;*V1*qtcM~k^L2Pi9Os%B zcF5pbTpiRacvi<08O&r#2M0q@AwYg8?N)uNZx*YQmuH;nB=(MgL~-=eGlw*CU{Y+YoXkHWydb~W;?A_-!Hn+ES z_R`*mreC~wB<|t8%#_ky227gwjN?J6P^fBC!)d_F)oJGW5}P*TN%S0OF(mOs1U1$> zpoxH*aV};mI1q-JzxL^OzD#_R!hMY|&WNAUf<#eMZ=T0joL|h~=m#j^z{na7gbWS_ z8~HW(fEoLPP;>SQI>3Qr&qemugT!QC@j0djJo4;qS>c4F#Vp}g&i=nqxc>q2rQn(( zx@IC`1b*d^1nHSXn>moC~t;+swr=Iyzk^$Nz zF`91K%zgs7qyPPUV-GrSB`Vx}ckA?H^EaH|I(KnCB%$pdJBT{F;3a+e)3f>pOy;1Y z=yFsgZ}vD^cy2w^mR5ThZ4*+%TN28ZnU%eO)QydUjPlkg845+|RQN84Kt5axU|0Wj zTlt+-%1aHQUBuaNV~UZpI!rjlM&!(z2m>B&rULj6V-&2ZLVdu3^_9SjmD+S6TCYuH z8&)U9EPgf?_chh-{@%eLhw5aC0_D_v(t&VkqS$F}0}?a#AdQmYI+BU_L4KzKe6$%RFRr zw=rc&WtG4MSV;~+WmnHEMP70?+* zdqD2MVj{$9Ei0l4DAr)*C$i#7A{;g~+ic}1J09qSXgz_y3qfXo{Io`7O4 z7op?bFN8S0Zx29Z7`SIKaY@gf3`$efidsoHjR=3>QroM%BxKf?|zuc~Osdqm1F>y-BEmf?|!( z>ih4k2dE(gg*%8XXFGfTMY@njyV)yIJD5JF&`HjEk(s6Z@}o?y7JDdW2N{P63GF?{C~VeSRxH_ilw)V~Ez{;slqyuY?0H zOC4Fccl(=;#ql2$&XAPSoQb^}3d`3z_iis3C(j*}vLFKa{y8`I&L=SGp|f)D{Po)T z=uL#-cKmvInqH6k!g9VM z%O|zS@(y?=7tQf*KyWlb3IgrUxGgp z&H6a}u>+_vXMewR#Ic6iK>ovKB$`aOt0^l)(zgi=qT>C;p?Y>h(EIs(aqW@kLfdD@ z>=FFK0r+!VYtih0SU%b9P|j8t@4y>f1M+l;YJL_b(Z-n#{M|Q(KO^`fn-2KQhLk-9 z!b9(7@|=?ZYO{Ox%5|{jcq7=gj>fVzUc750 zZsw%&wUmM5i&X9(;6NJzYXR7X%Xp$=VdDX6C7J6AaS^tuUgyKJN0<~FORIE zsGh-3HFwZ?Ec9PrKRCNFu$3X_(#amFE zkKp@vUO2n4vw6q%le_cHZGFYOK;7q8tzD@HP>wngscBfT3w$L0k(m!N%TI*)=jk<$ z34t)-ly!BOFGymzTpRUQN2i0(#&lBhL0%m@C}3*9!4AGJrfghv!A~=lP4}O&6sX(Y zl;v6&_9gVp=9qd{inF{Y8we9je(ZD}zwV2?)A)t0o}Jl8ciT;wirh$qtme zxdV12)|`a=knHN8Sx9!ZI3&B88 zC8ZdB~X0Tmj`7&w*aeFwcrcYMJ_c zxvc+t*oi`)-nh7RZry&W3^0U+bEEIn2?MqGwyTMGXVVUgb;TfpU68v+;?mWs=XZ;!-Uj%ef6@&yI{4kzW;8`aAL z%wQ^<9W3n_hgc~xcEp>|o>`_ZzM6)r45;+0=Cj!}ZJ;sHB6g&8#au8=_>5W~Sr(ZJ z`G?%}s^_!0*_g7J+RFa;EVDG!>Mi&G%`4K=uKu*th)VIBkVx0zZkXi!f6U5y@w0T8 z!V&CHyK1I)7jP{J+vlv<*;T1ZG2_jwp?``kWTX-$^lW93iTea1l5(7AkRp0m*{MaL ztDN(qK~JZ#SG{exd9GGfDE0ejQ-|`yk$4UBj@Js|OhAuTqNM7(JysXhqC}pBL!9jA zwKCULC@I)pj~-ujpbn+-3wC`1bcW}wdGA)B54=}a z92Y0lYcMZwBhGI9lx=87Xj9|x{jp%a+&p^yD0CUu^34OTPEqsFj`aoh0Thb)-0~Uh z1*XdwY%D}OE<{bl`$ppa+}lQt#9r`V__r76y1lhwC^r9apjTc$oA^%q2GTbem@egF5bRBq8uA4pCREUd^EJD-Hi9A z!DWNDtKzeo?hbE{&GV<^z}xlhoz3l2Tf6EyuI|`>h8@^6=FZ>VaL(GvI5}gZfMdL) zmOQaC=W<)FG841u!_s%>P2V`YnhScxIS#+V^yqoX4u#9PH!o> zc>#HninVl9tAfk@g!_s&jiEMe?O*~z#a69&QHDvc<8L||sp%KLsHSeU@6oEdd8d^G zXT)J4@KUjuPVq5o=!})=qEqrcU1MIPHmz#mrR6NZrdB8~LP;W4k+96_ zWk?t>6JCejBilQB06quJ!O(U*FYI+-15}cW#rhloFvGqpV>lUO3DfytOFHcyhX!PZ zPXTDVamb@f|H;{VBv&0^RY^+iJKNOtFDq{8?ylB~COOohzAa@%mP6q16Uat_mK64VJ~Z#(6>rE&rWHkB@c6KX*b||v z*IN8J$r7wg;MO3&800|oLLoos=rU(2 z^3==c8Tdx?3+h!6Uh3E0a0|0@2QZ9r_%wtd?ONlIEBG&<1ev`6a4d&kpUT*~E}hIh zcl_N0dvbQa$2S4+Wx%YC$a}VSuM2?~J0m&LlY!*>zz=70ZP+TGYao~$uDzQ=wnDV; zaKf6j(7#MlMqjToJ$8x`FQ&$fiS6rZJVj}SI=%l7dZwit)4PTB7V z$>jyHO^RgIuq}sE``l9@K9|9^ZrK9uhpqC}t>beJ-*PE9lu1GME!VLr{UFdZitmhw zz~fL$QJCU6s_Q`yiuvRwx*il+hDz>8By=Bg49Z2XUhluL4*t|xD1o^`EFakq1x91p znSi&|A&WZRN09rl(>hh8VFrYo&;jtzd(jwiHaWahPGdsDB+fsL2~f3cg1m%4n7+Y- z>)GpcDmJTPjGb;_$8=X@)!Z_g0utXF6J>PS)PeEn?bE|;)XbC0w5j%L6ILgVh#$~1 zU1EFO*@PecrH_}xmB?z^(70VTZbAV4MlLt$&{|SzSbVMO*aopRu;iUJm>f7F^@cGi{@g(fC>wxA53nQfd-(9QTJX?s>sAHuo-@u0yXj4j%4M)5Ur0 zTv(TBP|<`)=x%v-SiiF7;-<;Vn6E6kAl+u;y+&lu?Mu@KR({htuEBZnW6{g*=YuXhlcTr$V7V(&4v z5dgTtFbZBw8@E4*BQIU8LxgYX@W_r`>rgJJw82(l;wMRaZGO~6z9pCeUK8tVVGWMw z?VFnhNQuoCB6Q?@y*nyW35-Y!Bh|~|$%M#;7IJ%NSLjpmp_1N|;NGy@342!8U#LX% z^VTLqB6|GdEB0ZC*irZN&EgeN!>UPjK!6&J<|&oT?k$BV;HA5E&3e?*9fq&i@vQK$ZF{pktS=L z(-KrBSbU4Z-iT^p@5L2zr-&>#@i}BsgGxQBfLSJr8=BByj1!r|_z*R_GF-67A!41> zZhO=CHKnZt2Xlx@fqJ`y(4RZ`s)jWmO}(K83Wh}uz7YNhp|(m))AatjJRM5+Nnw7O zf@}3YW@E{CldW*XDDU?!Tw8{mGtM{kh2djTy}jtLS6ne2WT5!d*cGecdUEsPBdI6r z@M#ML2uC{gXtd=e!-xvKrw#{5Jd=8T4%?M)+qB~&NVOWkioU+Hng#r z)p0-uGnvxC!BA8P(5Xy_1w3ZA>QjBQSe?8)<5VZHcVwbCdg+~k@p?&l=QW2f+yjfNgwf4f7U2>4l%TxuqYMs5pC6u5_+VK%jB7`Fj7R|$ zO?Ux&pb84^B93-L*C)u%&1ZE3Az|L%mAa#0G_vCpOX4R9FV(Qv-q3)(k$US8#rm1F z8quUMU!rIm;=>fi61krUicQoF!I_qXZP0{@?@~SC^j0~DHA#8lcn$QTduPtxeQslO zduwMeCx-Cek+_HVGE+)-88B(yGmZzPLZPZn4W|JwSErfhOKjSVC((1DwJfQRf*R`` z&_qDZI8O_HDWAln5#OY6U*n52;%BrVQPk9bKaa0CznH<%4^Y5?ku@9$85|5Y@@wz` zGxi6e=Ij-8fCI;#i|nljiOIg=b4(3*S;DQH{ePoy{{!Sp!8Jv6%|yfq z{4y$*j5f%eVm8+xy4Xi|b3M5HBQ2$$1>gdfDqF?6$V_jmvVYsDXTFqVfHo-{Ek^;( z>?e>r`W5qyJ?OlZsBrV$t<#Uq-*A5G+{O8jgth}h5OwwtKI|+fbG}n_IjWL3dmJr1 zw;pOst6kAcU3^VgfaMST8XG|B#>PQLc^fZSb}M`rL`YE4OQsb7yZW!&%FnAU4J&>Z zaW-tvEDhO$*Xz!j2*b`=i$9D}u%-(20Snew0xwo-(}ifgHj!;u9Wlwn;Mt#z#eGe+ zyT5lZ$e}u!qChz{pOg?TP4qd*WGz%ViXD59M#*q;2Te8{gc!ljJN3Z5Zsb~wFc>Ov z|FLm1QF{Hh{cMI!+bT4X4Mty3-vE} z!4*MeSI=j^6rGA>CuR_frE;<0BD;ts8fDu))_foNoMQ&Uqt;@>fnBJmdlt|jm7$zi z#d0Qt&?#yWhB}YXPD`RGMJ7N|WI@7^#MDq?JYw!_WEDS8k;z4Kyc-Z4CGB@vccEA5 zjHy;CmbQKn_16hrF0}F)AWAVT4Pq-n1$2hd9*{dql*p(;aYZx%#TwB0#8!cqU#g&H zv1UF#0mWKOr|sc4Ad}(8lkP78#ac|7>)kJeIKFRBK(Q7Rm-Oseevwg7=Z?Z-Uh0%C z6~6$?a`MG5E5uj35^?9Tlz#3)WuB@or=VDi`Ko<-42re*tDdxY9;V?|LO6Syn_F*G z(iydoPErOG6l*DmdSCtlPJDpN7LX(g#ag<>xipoRnaczeYlH;n@xCZ;yM>DRE&v6^ z8llzq-&qe(qdq7UYiv2&+4D{0LSFP{uf&T!eNLg1ob@6zOBFrq&Cf6M34u&62JnfB zG48R)T!J#zw2 z6V=baFK1tMgIAufHb(r#J~=n{-j#P=zTCULTrruDLU1@dTmx@zH_Ff$apPhTR zzYtd+?WePH@BHt2Qsd>``JeB#Ua@-AhSg~=fSY?~6Oxg8XMb}pJ$6R-oZB!|E>@z|L?A*IeukX9F5)JkTc)54` z7g2)}_1~N54ekOYMa|B=JEVFK|7`B2v*m4Vm>e2_!C?fi;hxI4Ctn)sH{WKv>(L%b zcJAFQ?3e-PE|2+8m^o;`guoaCol z_ik1x3euUMo5R~cO`@c6bMO7$+W7Za%APPJ3-$21x%Zy-z{k%p+Bvy*qXnH3#Vq}R z=A^@5cqBml9WyNT&{pT8^s|tZYik*~*j%dpvzMqo(CgNwLm%k=7hiKd_^_AYkA%HG z4u53NA=WS($bZ<3M3d=uHD!fJ`Zj?lIEKH(R%*J^LY}+@6KO^`fn-1)NbW|pC5E=%;L+@tt zoRa`*vwQZ;?iPxf7yMF52kQU|{<2H7B;#a9y&Pwyl{m&~$z?#PmFzw)-pNdz3|oY0 z7ri{6)z9W?t=JHPVewjT$EY^eZ4@w1##fbMvm&H=p`LET;X>7Z*GSyVNdRjp1IKOU zrkltq;{m!(E57&v4zv+4&f1Ty-g&~B7Li`7;y=N$Wdah1d-$)}CY{~XhgqIzfdjq; z>)!`MiY_X1iB>b+YMV!Tu=FQ!TrF(DzKH}wW(oiH#vbf&is3GWe^-)H!m78RI3L0H z@4Rq!V`uY@?I(BVo7?(|;lMS3dqvHo7>R#m=7Y@g6Jh>&dW~a3WJrm1b(k+mV!Fy6sI_4w_-#jLjvjn&K?) z$p*p%lOH>s$FKV$?{xkH4#^hic(-JKJ~2qP;O_&=c<-8l7R>`b+r-~9g{@KASUbOu zYgW3QC$>AhqU2f76pHsUU*;#yJhlh8t&?XD|9}EF2Ap>Q0^&Mo?M|HMwMlIbSavp| zGSv_p-WmbYGTUWR>z=Tij`3Ig_3)UHUeKlCk9`p0ulTnzYiQJFTv*C^Bdq1xU9_nFA3C*;6)UCndM=s4|fu`W!NDenVMO)ip#r9mCLHbxFmj5Cv$y~u!^q~O}^p) zICItS0rHnRgv#9$e~V9m<_Z+7U2p!IX+;&uz;`9X5#I)j-{6>0OeXZnzA8z<i&J4pY ze?@^*%Xte~_N05*_m^F+F!QvmgQI4PLrX9_AiL@YWa*#-6(78QcuK3}g8dv@lJUW; z7~iT1$gN?0aFq_bMq)$z z3qrE1#Ua_%+>mTfdX(onBnkNisZ3#dLIQWN%C!3^Ea1g^L0Oh+Da+LzdEVWp7*%SYaKu6HhNC@hJ1pii%M! zT6)t3oq?E59T38_Q`$x-;@4yGKx19+!(=BKf3>TMUx;}%7hD^jrllx;nXXI z$mtMiLaL(gha>SC<{hsU!kK^`twc#>(mYo8wJ3u{T5@kc+q+Hcx$S6+pV!JrWLe1zXx_qAw&Nk36wAt3&xhIs+Uo-B+SyT8O?Dg?r;_*+Mmndgo9EQoM!j*bSd; zT95K|{`1mZ-c+P;pH%Mep(Z7j$_r(e=H`#nm0;dyBXud*;8yloPS)jn`{-(ylBZ2` zKu+DvWT37};VQo~?40X`xMF5pnd*dy&+|QjVG*k`B~?S@aod)9tja|Gf4XiZDIEdB zi@D4lNU*9>^2M2}V_jv6;&)m!0;<(gofHFKbuoSk`KVix1|d%qU``9T(RcCQn`Z9d zqg#6$VSl|!Xx*)ol*h!;$>EG7rMp6-*b=ytsTR9IRPORle93vyN8NV>prsi@(g7+7 z$#S$voL<ASb@v_R~k+^~z(mNqm1Sm@hYvUOx(5 z#Tcq<7wEdZwP7eW|CKjw%n8lq(Io!D_iDei|`$#GKvh4&o(4BbNRH1-;czri?{EOD4#s{g#SM!Tk6d_e%geO zhW51A^#1f<*Wm4{_^hV8FYt^#!wzg3bLVewIA`r-oJ5i=;Fyuwivj?HtHhN@o+BXu zt_SdmQT!}I&gHgTWhQ3RheeHY?7aIJpaRlRjdQGO8(HC4TK3agW~5D{^6ACLs#x#L zu&?7CUzcQ_q+%@{)vDlfKjFUOO=GA{Tf6^MB_b_?R_x#>#ZjDfynRF)vb^Ry8f@fSRmutT<;7C4`QirQ0{FH+`^1 z!ZNFuAz{EwcpZ9=Z13y=_#89`L)-DZu-AbNP)RNp>vMq4KKrhW;be>@Oy`3w>9l(s z8ju-21!Sj_O&(c=Ac6cIYDt=l?oT{GW(y%Y=ywq0>^)LtoTxVB6L(I51_RdinR8EW zoSr9p8rq)*Wr8wh?~%kwjnEAVu6jfceOFe!BHy(NOoX1jM=l-btA45Av-b+W>Pd_6 z3kp9z;Q6QdrXHXTbe^8QS4KpC-{8gKWnb)i7Mhz%bp|pf9Fp6QU2nanOZ)~L`i%Uj z6Alg{wrHQiO78hXqRjPHl%Bmu(&~qTGg(E<-lLaW{Omn)Z6XboJoAE_y+k3XQ4SUs+8>h&uAnDHp`8c}M) zA_Ry7LQkuq4l7cpOfSo$67Gb?2SGLxw4|`_^PzbsUxA>X;L6R7y^Zz9q_4U+S)S~_ z|DL~kRqVan)n;2p?Hqy2fe`rMnWv(7y!(!KFgM9A2=N&T8x~)5vIHv=xHZTx200MD zP{3Gf!mj47}X=1@$TjFZFA0xaFAZ6cEu90T{+Od>TTKcCB&975o=ag3Mk3 zIF@FU=BW5o#@==5Bo6z6>|M#5y?bDf-dmasu}44zAzOhj17>wZ-m|rPZ{x`u&Pa~* zWFYxI=yhOoZP+Sb<{+3fvov4DyD4NVgl#7`b{?PXm;cE>wB7#oa_ODrh_MUI$Hd47 z2mNe@7`tWr_zYW&!&ay4_k-l}g4iZSvTE2ySTeK+Zl8N9#OE^D)-7A0{jgQOx^;Zc z;ae^RhcYQpRxis)WF)s%#CJyGwP==LsHG@O@f_7vo(IKzauZ$Ui7Z1UcO+yZ%rdeq zYs}vt%gzM6tqxh#@jim|fSuN3mS$C4EK1+3$kL=`@x3uoMwiV+u(7>;dbo|6c~Y4+)!sUWMCj96xeY_m5L{`&=#_h6k69VWra=Gb**8E({ASS2rAT3oS$X0c5u>c$X ztivSib>^@cGi{@gTF2|8j|b9NvN*rHwGMBCC$@L~rmJWM&~f~!gK^IbuCck-i0mYR z0(!M^@NkEkM*(cfd3|)51{F;>tRGTyans~w%!icxCfy(7zM-OF>1Ek-`_lA*l?yt@ z)j5~ZmILHSRuJtU$eACKUlt!6bCj$Ti+4bojF=Wn-Kb^-3-I%pl7W8|KTm~oLgn=i zB9lwA3KFBl=*Ttzt}u*(7t_Y=4-(Ii9_b;%w{&=9$F7~|b2*Zd))N6>26#=ZvxPM{ zp0{sq8XzS$Ux?6=^Y!kiNF^wZR4*HPo=k{rXd${|Qpk0FybP2Uhig%4g_9n|KZ>6MTaLK@=3H2gKNh*@2MiqW z%u7V63!zJX4CC~sMZdF=Ziwie#KtcAwsIh=rCUT=W{vZ=DC~`>7WQ6TA$N+%f)k&N zd|MWjnu_aD1GIxL}V%#5$?n_NMV`N?QpI<`9+6^L7cLzjl+aKAL(% z4HOKE8hj!A5khU1n5OCdb$L3}(}=tezf8fkdLOf~WW32%IAWCddl#;)M9yie41Hnv zm{e~sI_wo!Oa~bVPeXbi9>kh+ct4VQvJM(!p#b4Xr`}dbU_BFH}G*?u*s_FMaoP#Ez1%{yh&`3x7RCWm)JUC z1?ROB^`?&L0K_42a57H@Alx0qf)E16#mr*Da6jdQ7t(INohFKhIAlo>z|8|-!TQd| zlK|~G**B}BT6}y=6pGTITDbu~Szc{KD!Cu2qUR z&Ng+&3J7L0rGtZ^s1V?-`UKuARwpmdIDtu=9R4AZiQ?#`XM_{t=q&oQR&q#uVoX3a z2gkLSly_cp_`*G~xJuv$C$5iR-MrYIc?Q-p4uZqc=ZB{h{*ldLT%1UC=w`JPK}8de z1rn%&g1cBS+RaZAUaDcSy`cenBlXrHiuE&TO=C60U~IHDiVss5OXPkcC^k_y1ZSGw zNs=J(U8*OX-l`8zgCW)=<%Q!l(2MS!IeYiHjm_(t5AS8Bl8v53MTz968K5$si4Ms2Q)1JYUZ01?rVH;M*NHxB#N5) zig|p+`9;mJB;UmcDB!@z8V-aE4h9?fHTZzZZD|7J1nBG)bbtfLo{Q|Q2Z_nP;`MgZ zHb=GzWN*s~CnPOq3Ab|g|Bb@^50Ebf*A&q;6A>fuD~BXV&t!8AqKkcWH`jy9KhjeA zt7*;B6brCa*(%mWW_nwd{o77G^QGhjXw%_MXl6fw+|jqpH};_ORw6IXcehSIHh;tU zt#cRWLlW8!2tjz#1;KzI2uwD4R^NchoVyiWj;iE^D&D+GUU+U{SSHcMotm0}R=c9N z38`2L#V@p>(rtj$jg5n>AkPJBSrCEzN?*XP{_D27dQvIx{T+ZySn<1vv*CtIB1d!> zSP73fg{-2i(GqvgSXB6OBO32T{9%lOHATOWI2|gHyI84B7ozpCt6>m~4i~gI8RBPS zabHvI?(ZE8a;Q$GC{Rw#CjsbJDfXz1VkZxkg!aZBq){@Q+yPkX{-$d^UABcXNtEAH zXLj~Bws*G9$y9{I{l~`5MCtY0_Olr_ZL84HTtXAJ;YE*WZG!ov*!wQdU9aFVo4c3Y zLIH3AR+2+d+12yeFGVMH$iya!d^?sDbg$T0W>U88W6k%Gg&i{x`!<#mY(Tevw!`9N zhkOAIQW?sLRV-&R2%Vx9VW{(p=(L0jh8Fo~<0OZ~c*NY<$PWBGMJ5-`mDoUMVrnXW zmvt9!BE80`VMqeO{}DinfvEDHjPHK6l} ztpYFKssoAk(iR_Uv#UfffP!KzrqlNDOV6RT`x^hAfMP8s&Gqgr8IowQE#wz+Kn4IP zDAr=)lAb-wFEWyI?x-LR)OiBVt;=|vp?-Y>9do?;-;Tut9Iui~DMTgv?p_RvwUxAZ z9wv9oOej`U{ML7>d4By+lxCE$A3`FZO^b>6N_7dFOyF;q?@XzKhBfQ(xhDq0km3!~0 ztnlPZqY`FezAw+sy_;`4W+3G!!j12=<2E2IIP&oT0z3C^DD)3L%abub=jGnb-**mu z@6~I?ZNM+Qpe7|a&tR}~?-p6!0ng+j)3>i@5IgtYlh($+$5QrzTiq{c=iaqml(_Rq zd_d-;S$=oFmB7c(Fxok}cTb5LpppF@lg@{>Iv=H>CYF}9(gW| zPVZ2%M(_^@;LmaMXgwg7Pj)+;j9qKz{f_`7cme@5^}HXZPp z4JmsJgoobE4OMWh>^m#O$qaBP`?#Nl4PYHefh)Z_iXJhFR&EOtYpm~fdjq; z>)!`MiY_X18P@WEb-dL!k95TJlQ^yxwqRdyf+4ene|uvOb~wdwm%_g*Ny%WMdjE*G zpg14F_wT%Lc4KGrj_oIR=bPL5ig|&$&#zj$QV*aUbs|#Juws$)k@!buKFBOT5$2z# z*El8w!h}=S)nUFMiRl{4B~fq>XcbDZa@1cVQXjd>*1`A1lH_vEnT6$p*p%lOH>s$FKV$?{vZlF%}EDCHwP}x!qzBltexM-H7niD6Z;;XRGn?sW(D9fSTFMvXCB)F+}6pnhkrnU8w1WufB>}= ze|O?MuT5%mz_M$FsLXYl)Ve3^repeLDcz+@!yo$~#4PT9@m6LHjoOS0OF3_ZwOqT4 zHlWMkvt#%rfw~dAh=MP(JWTcBPU1?Vmuz8ncd71cAMhBm>`-PK@}H7r((VqnJ$SlYLc^g3G1q_aJG$ zH=wQo@kUzPuPSLl8rX_um={6;hB-&agA(&eiR%u0P)OX zpth=wEHM5;Kh4ES(a)k%heru-rc9;xPU5_6KqD~vil6Z z`BDEv3so_w*oE;}>7C?-~d@WX&B0T>C8ZdB~X0Tmj`7&w*aUX2&MRDx-!$M{tB6cB0UyH!g0S zTeqJo0}Nr|+~{eL7`Z|$AK4Fyk@4(Kje~wv9G=`l;h-lrlT#Nj;37;xRAtj^SHCwR z3C-p>$!QZlU4AGIVpCmOC98;R%ZjOsR_Tq zzo@*{C}taaOf~WJQWB3c|E#DO)uN?0UCd_LMvVj{ewnUbLipnp#Re6scu~3=Fi$H6((KyC@=BzH!v$8=o1T$MVk2yu zK2Wf1w8x#rMz6TFwGQEh`U0YDDf~Z#MLvX80hgQ8mt?Z|_DHPdh_<3yzMzl9VY6JC zUoQ`>4ykZ<@O8&HG}&%oVd6FL@mlK8u86Oup(+C^{i^wFHccC7OblqEq}L^&5dT%PbAGddvNP^NRGet3NF@qHg^=vj^V&(W71U zo~HX}=`e*O*r9gSOz$q>S`xO;S+TRLU9Ws;i@d%(#8Ye`H?BlsM1vI3!^%!AN~>tl z)9L;%ZyRo&t5p?B{XW{%q5NQLN=DG?c)jH_WQ;BkEoNRGS>{I#q4K<^ZaHk$6sRrU#_^?MT7>R0i z?_yWV3Tja*zhFx{P4tB%u&yWeW_2iENN0ehObgN1qHu3qEnCh=^0%-ZyWz7< z>ruYWe_p!Fn~D_flgj-))TE?Rd7f_a~f)TLm9TiIthS(oqaqpMv~ zuftgEYRJOC)Xhu=>Z%m3@;k%Mxn77XW+Sn*A^AMt6Brh;DpOK5L>{+osn4z!SEx9C z_eO_W75?~hyRtiz;heRTadO5+0mq^z%m_4fPzB^%Zp&5X z18Mr(m-`9#6>l0Np|rL8 zPgNqa;zb!Iy^g;*)cvBGy4Ai%tLo;RRuY_)qGpMJz)QtqI>pDVp)*#di%!Y+bd7nD z+O(=^Ne9$qg=57zgD4?%^eo-JS-t6lH4@Iu>H%iL>(F~-duI>8=b$+l+K%T7=yhP@ zQ<58r^*I2ThFwLvM9^+hvbT~LEQ!FOjfXW~kj5bTG(6jeQTK!NkjuJ=yoS3~w{{5)ohz5cia1oCHCuK^{ z-lNI%;=h?(z04;B9u0V~6QK|9LFhY+*Ee?8_n?;I%sCbIWe*i2R+_y~H7Q@cc0i+0 zjxVh^4?0L6UA^9ifz}}*a~8tRR}6i$Bk>g~UP>(KXlMsCKY!Jko!z}=Br%?Gu@?{y zzr-eHvj48_)2E5xHv22Syr+05mH)kt#YXL`_2wEOAOo>=VSVr7*{u%!Gm^<$p@%gR z6gW3T?914FP)%HL8~BYa&hZ%0!x{XiL{kX&96+oVWV<80fqgd2b_8IWSd?3F0Sl{%K5rs%6W^%hrVHo0d9m2b`k# zP%E;IZyC?kR59Jo9EtCZi88uuE`p8i?bE|;)XbC0w5j&asyG;V&JI#jPM5^?xU&gA z`b!@#hbxiQw4re`m)>I8Dg)>@a=Gb*)*QDgEk2|mzlyRT&<2*g3ouU(91-%I_i7^p zZ1}Sdld#vB!)DC1jY39)WmVkDPOl8y1JH5&se^IP3$C%bciAil^lIba;SM!joX5uK z>M{)~ns8V@q~_wL$;+4zDfvyhlg58=M55Nq0(x#=nm({{LHfAH7Qn*d6+{Q;)DMU+ ziw}-DO4f+1y7JBmm$F!zg$$ zZQTAK@eJva9wK~8hevko8m=W1eQp7%1YibuO{}wpH8`HPZ*CeOB{pA((2?`??x;v5 zFj6awR4R!uKbX2NM% zmN4Q?VuQTBUMVuH;JkLC?)fntfH))$PUgt~gu8=S5JJGXm|091?x)&?7b;0&gcv?P zCVPz42dv0${DpCr6g!(w|w1NKJh ztwR*+XVPk}krYgZe0tJ&HXo)imdO1?P;8=Z2+lM<9pbxGPdL3*4q{DGUN~L@z3ASV zvv;4{*xcUQ*)!Co-}UbuiF>D%M42dRvwK+fF_6rQ`%?(^7(F_7lh* z{g?BNJ?OlZ$cyvct<#Uq-*A5G+{O8jgth}h5OsFJOZs}IXY~!3%(+|9<)}(tsN&75 zQ?S6doZ z{4U~b#O{TD7%ZM!Zo+>jCvC@X#UI8fSW|`ifCcL-ffp;a=|Z<y;(O9 zF6%DzDxERaO2yLF52F4$!AraHS?yUWNQ)9^0T87amIkqvpaMF>Xb;F8SWJXitz|_t z0mT}u{KQs)mtT4!&0@`bd^n~La$eD9X~+5m(G1(er;7L*kk9b@DJa%*5jy?53n7B< z+Y?Z%#l$5&dzN3s)sime<=ncAw<=^YC35Rzys}R{<(7(#zS*hVhDL;?pjeCfs(pG4 zinaKwp0s!#rXe|ggbSfqBgCYIOF^+lJ{l#W_vIgGYS{n+rn#w9iacte$my5~0mT~O zvh#RvVDo~qTak^2?7hxq4X+hCYU}sR#r3M8z2Q*qwk=)UjY9B$9iGdG-{tCiyYv!1HMc58$kt ztiVFJ~L&!7I;K8@*BJ;kt5j zV{c>qF*)>MZ?ZhufB!vy^{Ut#aTr9-&AoU1$CoenZZFr&FX!go?Zq97<3A`^LsCjZ z{`M}=9z8E7FQ;U)vXJ(NRyy%Af5gh-@(J{OzQWu)e?9G^Yv;I$ z!`_1F?A$w_TJ5JVR%*Q5JAd20uik>YF-x48duJ1pk$Y!V_`*DQZtmTt*MHcv5)JkTc)54`7t#ADE!L=q-$Y`lN5Ia#JEVH4 zxoqyvnh=x>Pv}Rc92$VZVFa(?p31lNEX9vFH}`IlU@-b z7Lw}Qeu=^#yKZed46NG!;%lyls_;wjN1|CDhd;K_T>|S&jj!4-dk(RN*+BloW+a+S zx2q{DMAEkj45H%w#G%@7`+!(J*)7}+Te=~j==tK>BhQ_TujpbaBlw2{@aMRA2j1u! zkf)1B=4W9NZJgP_-+g2FGlD;|>448{7=(s_@X))NJm(~U+U%Y^v%7_2<^{i0(!n}_ zg1_t%ZFn8yu4F6#(@G9^qghUN{2IS}NxYMpIvKVI(=K{>4y_@r6&pe@EMDvF7}dtQ zjRNM$_^MKDmLbs#^>iC{7pnHVM&f2p0$58KIKD{b{s9iO5iri$kFDN$!kQM5Zg^g% z;y=N$Wdah1d-K+}2kNeK$*94OpXBtzD^y zQ;tS9Lzl!qGV?)Z`H3+9JiW#-ArK~m2KMpOkiK`mdkA|+{1!oS%_}CQ&yBvEhMimYS&t?0Pdd^K~*1x$`uz3x`uUMVg_Bj z#SJKQ+17c^san)Ig_Ple>L+Z%33=vh@L>-u8pAUqzvCL&0{tjBXaM4wrSR~E>>t&v zP_23gvZ7K-MuS$Cn2mpC_6w?|ESTwHiOb8G1vkx>jx3GNvM1fcz86cbF!Qv`g`?&Z zlpBy;U2H%W;)7W+zEu;DWy!@!3!y+sfelb0*@1Fx(r1?~B)j@&7Lr{p4#}?OhGcuv zXGM`%Ehz!TL#!hAJ_-wX@m^4trCQ2zbw{3e_vtt;po$HFSs9Dft$m{`!gX9%j};vW8M%Ca?82<`YX9lS8f)g6+tp^yw!nx6R zYOM34;_&1a3hO+vASIw|+&Q=iQ^n!g^xA#tjmX8^l6_9Ho6hFIsHaDoi`3~)Ul@@B z%G=%W5#_P`a4gna6Pc`nU(88y3631u7$99Ek=W5X@$^y>k23$PC~s=f(wi>m3^0ms zRZM5rMkwOfWAQ*^UGKwWCmMgXtBPNUc{LYY8=j`6D1Mo)UP83rDT+lJR~v0VD0hVm zqp#>0hF6y%)BWeMI9#9;-t>%A5*uOLw6%gqE8rUm0WFYaXck1QH_-60c$2@me9A3Fy&ElvF0oV>P=LCE{og zbEPd%IPvpZnL{47&zq-O2R(Xx)qy&c$}ey#Q7(;>EslwODnGxWW>gX`GvR5WE9YJO z3Lo}}1u9Xk?p^FkSwSsI#9+dZct?@;w@~)Zcv$~^(bHGKQGoUR1(J{ze^ z!3MXo&vKErWE?w~x|zy_dX(y_6t40+!_K*0h%06zu_S5Y^L$TWSj4JKN!1W}ys)J% zYg6vn-8%_A4kL!)i&V16;l*6$4kTDrDf!~e)v>NJB`F9kGXd3VsWyUvueun&gnZO3 zNrR9xWTotdbffR$y*JI=!AH0DHp2dTb#dLTla$BA(aGVAB&EATqu3I-lc^TFK~(PY zPJ9UoJ~tyX9XWE7TG5~ZK(ZVy5~mk-E@IiVktPwKr!m5j2$4RmvsR)KKOJ#H`mQed z;pQUlwx=f=DU~>d+498aax2Ra72rMEY}_hdIXO~fHoE0$yWw}wLA?g^@;2h^mT+uC zRqVMP)oC*#zCRYsmzzhgAB8UCTE2O})hT+`+c7jhOf75pn9nVr!CqjxjKR_KY|KY9 z7K!(b#QnLqjT(u);KA^3FVJ;+Yr{}%{wr_Vm=o!#$MD`Jr^Rb&3+PF?qf|zb0rJ^~ z#Aa@qHvK;R{YbpJc>Df{^2u{g`2SO~rQW>br%m{%C6j)oi}$Aoy9RGp#b-6$9o`B^5l?u^7U?7*fmcmDQ%NOp&I8{)i$!iv9#=`wannlMSQG^_0A0YI^JdM8S*3*Yw4&~1(*8?_Z4p%Lv1p; zA}1hJY}JYvWn9{I{7sv1O~3d>HFc|fk5<*qJH31aL*S)iF$ik+2?=Hmov|`qbV|Ob zYs`z(rd2Jxw47yyV}=c|6H;IsD%zv@Yg@Cyn*?&JA4IK~mf+Y6VTy;nv=f8XH6;$>g# zdKQ|SN~Opdv-imD$F8@5%jE1m@}o{T7)cd1P>vwFi}on2WSR?l_8v*A9}31%;>e#9 zv-ilqA2l4&KyU-5;}PKWIVETB(PVn@-%PHiwYudQk^_73U?)N!-h`!P=^(% zQwHbbQ3-cKk8MOY611eS@AIK~Ctra|r!`te?Hqy2fe`rMnWqdMZ|I7{?1m0ge{(Ra zG`_-TC~R1K(a935OyJfazZm2|^g9^g6J)Hf)tIa}Z2g)S9p2-4wDF!nTte zJC9HH%m3sb+HU`Px%AFNh8VlRyb+^Q>&2I;L=Y99%@AX^Y#*Oti*eX0G3=S(^@7+Y zMY3wx7Fs*Nf^xN-bYVW1!M1MM0_}&b^3|>5b6U$ZXGJarhcYQpRxis)WF)s%#CJyG zwP==LsHG@O@f_79p9jT!auZ$hi7Z1Ua3sJv09n}XW_k~QS{he?AEw;{+&2!w-~Tk5zv{!6ne&e?evTc@MB zpcOf$WZBLIiSLbxGP-;%SIwx4k;=5G_9`ii>ZA|}`S!dPfbDT-6D{p8eY_m5L{`&= z#_h6k69VWra=Gb*)*Q2&l$vz&0q6+|B-eK~o`m?{WM7@01c5d%(hic;pe>&o6OIHX zVXrfX&6sH$g^b46s%#q`TT4n!0O&aW)WNvt1=rZzyKK4+z1lccE&X2l~wV@Hf4Dgy*XA5g^Ja6CJG(bvh zz7U}!=j+{3kxGzsotO2-iJ-Q4G9j{|g}l|!t~VvPH!Qahe55yvS40h~Ce;A}YBZXw zaPc-sNT9ShT#HgGob)LEQT!Cxa&&G3`7gTsjiJ95*pbrZCFz-$h*B3qm;4yU=}n7% z$H6rMZ0w?MD+jV#x<#aA);NEQ!rq8#VeiEia;JzaIPpr=qOpMis(y!O(tueeit9L6 znZfT5QL`&OHfn8Woz!l7)A%)|tprTKQeA~)u!%~6db@6W65}vt#FJ+@AocTTZWv|uO&sJ>1GFyJ9t5Pi|g(B=uw+K5d}@;Yg<*jkdgISBooDz_fdT7GnX4XHu`vVY~9J zJmfC9BH`(TJ9jG0Ur4yVDE8=r3{Y?2aOE{TVvn*l6o5Tird%aSx-r3&U z-Z;GtA%(dUR&We{)SEh{0}zMA!O1)sfN*yZ3qr_oF|(L3+)p{-g-oA6R{MvBBkHIY zA0HEiVz`*RFhd4Yrd8Sy4L8)WpaAGc=Iij3ILk_1RiD+H#p>kc8D}*KuOk!1(M!)9*2XwGwM3;&h4{pnfNBmDua}f}UUT@uJ+Qb+ zL=S8&qY8qA5FCy^KRl)I!Lo7~XU=RbozX)J6-_u6NT3P|?qb1cH$O>usfNY&h6e17 z)LVxr*3YEXh^7uh?@uTN7qpr@W`{bWrY)x7PEv~MjmNX`HjN;50Ebf*A&q; z6A>fuD~BXV&t!8=M|X2Qxcnn6rN5e^VqmGVRjiB5^tLMdx1Dr*0&mD+S6TCYuH8&)U9EPgf? z_chh-{@%eLhw5aC0_D_va%&?xHfJ~n>aEvA9ukQedyq!SaB>Gt4qy;f^y`5K+MVES zQk9U(NCCDJ8#fcB*Kga;X4tf?LPz74P2_WgfD*d)X}%02g2- zIRuqmJ)iwjbW(>*^qt7JV@W~xij9V9VTu9Pd>>iZF#`?g2#i3y@;EmODWE|rLpiaE zlQ`901b>7&WmPAi_1D50GJgO&v@rb#zF|xb^p2BUkGu0-=2VCEtXr- zvuF85Y*s$LadyjHUtdZZxl}7vY7!+&qS3PO*JWRH5 zx;}L-rf`IYm>Kq*R0@hU^7Sa+dtd$mPJDpN7Lerh!KNRSe!RY*8`zWN5t7ks)~m|! z#Cg0g%G++CYra=NL9xa(%BwKSWEBd<8e7hG_Waw1sN_qEcp;vEQ3nB?iow^u`;76B^E z0ScpGO1Zgrd%-w)u6Mp50{Q+qH}}pb(DV5UbMO51y1JAmK6(>jI9@t8_s*wQ`{}IQ zJO8_$)Ofjf{^$L%DNR_NbTZhvcQzpzxp($AyWuGdFrxa(ZF3Q%xy{YJ+njNp@xhaO zVUBzY{bT3eZF>EOJuA^*e}I>Jw|^16kD`8f3oexg2-vxIhg9$3pUvINerCA2_nykQ zCtn(xGv99H%X4$@W?{z+#M~ySuW;i#(0vW)4lGaH*=Oh84Tb)}XL&N_=e*pz`TNeH zPkZ$ZSXkiKb93(&S>6H9_`d(ztY_gKn4aI5Rqq# zkx`k*lsyK*L+@ttoRa`*vwQZ;?iPxf7yMF5XO)x5bBUhCm@4@RAkmVs-K3V29h>9; zEEUxf?_{Pw&+Q;E&Q*KCdAJ!{W8xj!|u_+bH0)jQ%ObW*HK_ zP)~Q1-2S1+hf@r9Dg3*VloD3G1;zOYzJKS1vl}~`cWghoJKx;aSIi64HGS3E zm3qkJXk;@dcBhQQKQi+{X8DOQ|2)0MF(I-Dgmrb8FGyl!?b1m}lzk!0;>tGa?~@#5 z>)`v8vSFojmzXnRKFw4%-G9nbpl*9pmTO_y$LEq(O>vg@WCLM>$=9b7Vc7BOzGy2N zozM6O9Fi^2@ovffd}5Gn!QTg#@!mB9Et&^>wu!%I3R|PJv37nR*Q|6qPi%L1TgkJa zDHH&g!FrjWIP=&Z;I>YlJ^TX-+!%1)0SJhnRBLzQJg-e^bHK7|g{aJRnbf)`WC5nK z6!ef_>TPrV@Z!7#|1JO6hh#~-m03fhHsiul&KqGZ*Y2VX=rYuiWB4V3x)HpHf-kc? zO!eVT;_C1$*#hhCQe6R?aT2XYTk)ehnd^&$ReYsr@)ZZbnX7(pL;ezPHSxFj1Zb{6 z!P@oaznNB4kqms#5V&?_e}iL2F`3XO`>G@bmrK>}LDFoSt`5yeyph-YqSUi{?}-c*?IrvIFIAfHglrTM&|6bwaXu z3>+K~C(>#^Zey{b{RJV})#8xsYHmojCq2q#9g?K84p2PAD%0+xun=p%Bay3sM5g#+`$UFa=SSO|M=3 zZfP7D(VaEo2@+HWeHAe3>5=9lb>>SiZ2#h4Kln=fh9y}jemEBEt%>yNslzJx#hetp z`MeR2K_5-{%~8&DZSd1eNj%E@v!Y^Di6uK^|#9ksL94^oaueUvw#75XQZLQ$Z8u>4`EE(ajKgNRG`pBywmPK3*}>Nx;`5ig66LfDC~rlBf>9r#uA*=(9N(3m#bfS>TF7J^vM?{2MwpYz1_&fho{ zU;@e_Qz8G5n_l&NHaE?oZ=zl?;sx}_XPKp;R&TlgZ(fm}cJ-&FM$~YmQKP*0SvpMN z2zIDlHPgEbxR!+Nb5`u^YS$~TLe#~@^x`SDkQ-N`Frq<<=wW527UeHgx!?2FCPTU3 z_{CT38-hHx-x&=SiKo-ptKK%;JeOUCl9Rf1YhgwW_R*#e<%c8j8s;6Z6~dW-9<4-4 z)pvWW?rTx3@)32~V|6K!H`_3S`RBDVhYWb1H&4O#di40J19d2sU*J@tTpB0gg%kT! zettvEs3aUDkPe6{Z5fGQ;lm!WKqacxy^CEbE2u@O{DLj*w3=T?B5-+P?+QCq9SRnZ z^y*EhS_JWhbOtzFx<65w7NW03;oi7fww!8|x3C?%;j>NaQNGT9Ub@ShiWKgX%Kbgm zq@+@Lq3rZYBUR$Z=}Iu~vyr+KY;Y_4EGO&oy?u1GOUcuwIUr!_W+nr5RSH-6onhx( zFT@pdj`jRJ-xC-Xu_{wiHAEh_ZK=y(x~eLP)~O_=BZn7rnLChRRi)&MGgrrQ2c<)` zTB?&`;HxgiFCiawOVS|ZIl2XI^j*C7rkOkV=+@pw*k7+MuDf-T@|ZX}Ih>KCbXQ=9 zkK$1ma3@nOc7v$g<(>FaGn!fkln^Kg3CVJ_NSt2Sxrk-cMw%3jH#ZYSemqM2bi@tm zySn6un~S)sZ@5`A7DVGuP!N+XPkb)7vJ6hjG`9Ba9&JNz6|bBeDY7BmUT7n7={8xf z!MwbUIJ>VpwgHJfx7&V8dm+9*7R;BMN3S1+F5_ChdBD{vY989LzQ8_!LNT9PK7+l$ zbQyyi9-HXtR8ddyzLB^;_qI_Zv1OqT*u~BBr{pls_3fR_?NeL3%2ZT$6TiJc*X^wh zL$UdJj9Jt=pT$|y2GKHHGk%r#e8a^m#wN8;7R+xJJ5V?*ULB;16LhW51A z^!_xsZ18qfd{)!l=`&_s$FUf_tQ&h)SCbAU$98MKc!nL=H0I9V-f+&^$v8PJGVtoz(!^W>sAZPE9OFPj~$V=^<0#0nM?{w4H9fbb|WMY{@mpr=kpPao{NScf8 zzYya1eGTO7y+Y!WJq@ozUsZ*x?W);(g*^u=crQe~Oz7Erg_P=-3O;+U@T;D*2*054 z;{%?5Ju_o>OgDRPML#m=*?VO~^!E*3EME4-u4m!FmWMz_i8KTFbHm8(r`b)<-XlNi zgoA^KE!tqPV&6Wa^sbOxPXlhyv-e0^{ZKHD5=Z`=n7v2-{pdx?V;faL+<<>Lwm&_4 zk0#TL|7LRaGM^B5G~mHbgg(3nq3gIbC+=Tz92J#ma!Y4$=Ds=9jZfJUJl zUviFHG}tA+Ld8poB^?dzfad3~IlAX*dFTObfY}lAxW#Wr93kR#e+A}afKe+h-@TiNnzjTL-P(^@rJC_t`%iPoFi~K zkYyh{^Hda%ci-_2TwmdZ(D8@QP}s2eqLU?9nZT_H5xetjxq@49pnhkZfz zuH?<$J+LQd_j?3H5V95cGGJCmXCy~@GLU>9^g6J)Hf)tIa}Z1p*WOJb zTOn*axv}&3WWW4R{-N#mpR=J1F?NCZm>2<9nJLE4W{9y{wvW%S#W-wr%6>mcE-#2} zQY5Q!w(VoG?Qy|Che%LBs-8w$!@GX~uLzxsPtCwXYR&yx$&Pco#jp_}x6oo0C zqq@rTpqNi?qN_ZSWvJwioUa2wR`yLA^Y_QHGXZa_Ll$+sk7!z_inQ=?07N+?(GMyB z|Gc-6s2rH8#su+`IR7*zK-IG4<0S+_NUD}PZYy9kzPlpFhb-f{nkuH-XjS5SW1@^M zn~Pv$d;9co8#VK!GHt58#>iE%GU5mH;*;1OcQ)Zif9d1pa3!*uHZ*RRjho*ai5t1x zbV6%LEq;~YlgXUn^jf8qjKeMiZ1}Sdld#vB!)DC1jY7tfECcnxbuye*#eTE|UjXPh z{?x&^=LOf;+`DYL4!znqc(_AN7w38P&i3B+#_4UyGR(V7gNh~`)(@$nLgn=iB9lwz7*-G#6*K^@FpPp1)5h%&63>tx=^?_mba-UPu4@9- z%WibLi+l?JGr((Noh_`v@w|O=(*P;4`9g$_oUeCBMJj<2>GQ>vn;Ux@>yOE@xV_2p zWdHs5{MD;s7tR16^JGF~LkoGUpO$y}AHz7k zY0>X&q#Htr5*xed+sc8gmTnPgnKjPeqOdojTG)GWh1@A33r@T-tPy{8u%OgbT#qVX zRy>O99a<88=0nu%%5cFRhlq7jyX{Tm*OayrfH|ZuR!XI*6sWgL2>rF2eD%@P8)~3n zSk&MP;g1k%tHd--@2|_#k>t{sDY#beV>XtIH`xkDjPic(!nKviIeieIFAN`(>g`2` zz2b`LAOqoPNbkdgSPj?Hd;Uo3$vS9^g#v^loqAib<<$`T=b6;&bJ(tYD-XF#u1I(~ zk&EsF3D+0J9$kJ5Bc7i_X>dXX{{PRp`{5pNP3Vtf z@RY(ovN?=1XSQF#U|>8;7sGwA-+rXgwtE)Al4-1 zh2u5Qi|(B{d-u7G&F!t7Jw|X#xrgGtBXJM!Wu}xajY&g5&DfSWR?agL?-|E~QlU`Q zriRmim#fpv^CdQI#*^qd$WTLStam`u0-$ETN#VZ67iYxJXhEW=sfXwB73UW-IQjt! zI54t?10jQh!A5=!K48ZFAk>__f(~%t*mIG+^&m0XSA34C0gpU;TUIzBX)#N)gfpkcGA*LJ*&H(Mvj9 z3gvd*p*_nZ;x8w!pgha?dQ>fM_CQ(yvp#{mNaCyWi+bA-TT|Wp1AM}84S_CV%c4l+ z*ZTrU?SJ1`h9{K@-{0ZygoVEgIU@F(HChzd!F&)Ez-{DVr?18Tv?LyC%wVu(eNFI! zB{zMz@x})j7v<8-+ORufmU~%?|2jqy6e`Yfe+w-@u+v3Y)YwB~I1X)PBK|8r6(r7a zC!2tTAls-*5Qu~uqQ7Jh91&f2K;I(Zq%+bw_J54g!qPa!ek(uLh#da{3KbFmN%RIr zascM#-9`X-Vj|7qAK?Ws31YoM$b1gBCO=Im6v5n{VR^e!Kl=^o6f3XkC*xm=S4?is zeKN|AH%cu+=_lrhQ1X*2I~Kt?kh$rL7HkMCM^N!BzJ$laVl?_TAeTqedR>{d;0}JQ zQ*xrxAQJNho7)lPq<7YHIoX5_QG(SiEB?Q{`_Ze1&h~yg&lxRrqxfkTNW)c}3*dp( zMQcMUi2Mn)mo>jqI>(3)5FbUl6ni)Im3^8fVkE7AL2$fR;uBj1@qprUYgTI(4FH(3 z*RgIbvv%|);^4NnAhaeD0J$aJ!<-|@9!^8x zoA=r>3gf)(TkJA;o;Pvqkzz&huiRI@p|K7|=)XQ%n3)~q|}yJYPPK zgK-8ZFML)_t@v*__g)(>zQ{gGp||znp&b#cLBS;&c<_89J^1gdZ1c)V{6e@5^}c0KSJ4v}tbxJ+a_ljocSV4dBwXLh&HSpnfsm52_p z+-jHZD8D7E0sUI?{J13fWXCDsvSz+`7dQVhY!ZfU^s;(ZKO&3{XziaBD8GQ3K_0*L zevFD|hbO1J(5halCtWjd=4ftI5`z>v%@@$#jumu;ilIdfTwkP;|7UJ!*V>S+=9v;h zyUo(n8@|VE45{-1V!IgDLAda6Szoobv3Kh6{$Czho$`p(9Gdd9QUCq*gYtxKY`%US z8uM@LER_70i|{WaL-~z?b$x+&-jPXXT{Y!W*~A$X)(0UL#wJ}<))TFmy4ALp^kf7W zJ%k~}p)2;?EIHDcV#qS+x1rOR%_YrP@js&7mf-?-UO2n4vv~)gCCmXbLGi7+t^S|5 zX(hA%M3{e`UgKOH5MM+)A5-LX9r2PZ`-(_#bJX83SvuF@3M$@VI>&`QgwZcNoztDM zJT>Yzh)#Hlu~2JM2Fu>Se9wVM;Sz`))Z^EEk#|r(jU#@Alyz(NN0rTG=96b2r-Wml z;1%;$MyhQcJl1uA5a{=P+$&<0ow?Q zH=;Oc;paC8#JyIy&urI9Eqvnfjk8MWT--Xh9#aYjAv+2a&GSVSpF63m}cWY~N zbuwPrm$16lL9}y+avnXeJWK7zle@lJs=gwTr0YHa0L4TaT=8xso*viH3YM?G58CNQ z9o4{mO&wqUvWn+ug?9JXN}agiwyFDo4M@$m^-rN7d-SpHUP8*`Qu>A2JMRf6;Hk#* zh}1FD=PJ)ybQ_-X0*q=Yd7&9KBB&(L`f1idZA?29Yrvf41uGOy!|Ddknio9wehz~h zv!#~izZ?1q9*L4rr97T6XY>bn^|70S3MGo6hv<`!+d^`TO6RY z$wJC57reZ+Pnim=i1Q&UsGZkiCH4$B&>35P$nK9jgcy{pC92Y0}TN2Y< z{d0@yu9nAiSF>ZfJ@tm2Wij3VhuetIhj9V_-izw;G|V~v%p1`}c2}jFqJp{m%1G9M z;Sx-9I`cySLv?=n5}3%@jR~EX&NasoiGzmbKhUeiXC#TDtv5n?;=Dm zWN4L6nvJRY?Q6jHLgrFH({-LTJ!GsY@5ZvElxa%`k`1dJ6B#Q^8>Z#}h|nXccj17r zSU3_p#`(K+bJZ)wk;z}OnyZlAP!hz@_T-GGB<2>fhEo9qFMSt5(5}I6MHJ@ZaYI`F zWhDNpy;z;`*^47n{Obo_>BgKW|M0(z#sAQpO!dg!q_FiCbYScdnL5x*vsP?&!giKv zWZ?LT|FD#X+{`~RDzLS12~HPv6qJ-k&7_nrbp3A(tb<%x^ue`ci%IThyW0H4*xEk@ z=YP#8lR{YJ)Z^wIJfvp6nWhC*y>Q+1WkwD}p8$LZ41C{-!68)p6}rhnE94u~2V2|r zXdQklON$rvrcXA6=&%9GNltqeyd|9eQyiy%8;MtC$v0lJU(#sc=$!m~qpJP?ixI2~ z)^_Yf+YxRn1wy2ek@+qlw&6Fu^lc4}SzLbVY*Z*4NP5sY7a#l9gZWV@%);F5Sd18COU zhga>;sLG|mMogUx7Mt+W&s1rIfb7I;xfj1y3Vx+aG9~A6n_cS?dBF`pm5>pWr7^NH zMxt{kL1Vq0cyia-_{#6dI+xl{*p&}dFy^h}T8ecNwxxd5(hg!UhA^oE;V%QY*g9XlNL=rF#O3Sb*l9QeWXqEWC8SRgQxZEY2 zpcDcG*^ZVV9$gUv696oOjUdOoewSW9IF)P~MY%R@rBBfyRXg4n1gb!zxM=~8!(O<$ zjKk)9x8p+W(8?z1e4*mQ%+-@Ow%+X3+DP0cj!ll18HsLv+Ddp1yL4H(H@cB&w?@iL zG^D1fo%lb;Y-i1**f$h~j4QAdxKhpX&v4YZw3C>CfH7B0=Ju(rU1c?@`=0+piO$=58>WgU8#m4SZ`zs} zvlj(T`B#Z6k32^MhCaynrR znV%f6|82_k6pg;y3^w|jB2)ZfBwkZafIsMc^4t^t|CH{}wZzOtouElt<#T6;E){{ zhC2s;(=_e#<6&R>^gCZ>XLUy6HzRRt-9GX{xwHaItND>On95-mAFpHoL%VM{w>NoK z_g~XW{$=bM-mW!^=|ETwr+b=f-a0vvF+dWp*{l^{%DKranQUxC+P43)p1{?;V* zuF_&)a}Lc=AW*|$oieQO&8v)!)T-4SEe0c7qs3tCp}}bB_RY=eZ6BEVgkq^Nmp~dhgdSN~ zNSr(BU@_$J+bigig@xoLTPq&IA6$jYb@j|KYt$jHugK@9s@qV;dMRU2Ex9{+N(@$t zUHdKQk%i>f630TdQxGEy3qR{gjB*?bzg|AFkX$e=ZduZPZ%aMv?;oOAo!n?roYcs| z(l5%P29a0Sa2Avw&_e$u{v!b>O%maRlI>3J(r%22Fq6-9v6&d8)La_t6trj zpXQQmSs)2xJ?u`CP+fX;B^9by^;}h&99S_Bk|xZ74uNLqb_0flIgq3i(j;jhSu}Kd zHQk{}RwqExSqupf#-;(g_dlHfjQczLKmUKP3a=&UNv`_Op1!ls+2`BmB!g1X5iN6i z%q!~Ej@w#R5N^Ffg_aUbl{BJaycMEqf8NugBH)TF=8yfK@(+f z-M+ad$__5iee|NUL{vY^R$7_wI@4o$xV}_gf2@^~zbnTK5P1TM8or&CEx+*lMfO%N zX0%Y9;raLO9~^L{aq(w-r;nH-to(!3Ez?bKcQ^EHO3@+4ScdyaC0mE}W;KR=z;H67OKR-z@w>L1q&E0v%{CF6s zNtj?Xd_Y^XO`7{2fZ69-c4mY-RfXB42j4DSjPm$-a)On*!mw zL0%-FhN{Qob>`Lzx#vQCt0X@W`sq3Dy13RVc_$0QS9||~D{ZzOs5%9bmXa?P9Kwb8 zkx)0g+lN8wm*+U*UTghm8mZ>05B3ynz9ae#Vl&}&GbgATS3WGRCwXr_CaKl}3;{34 zVWU20y4hKj-a>qb*rHBZE>c?S=6=7d7Q9kj1GydUi{c!!Zi%`1!IP8yXAXx8)jvma z+l+b&%#P_fT3$n|soJRIO|Ixyq;Y1Y6|?A7^>LS=c#K3#J<0xvhq zs7vau;k%^EsN$o11CP6cg2n9xMHjEXU4Y$ULDthgXDFTIxc$k!0${@0f-bh15*V;;E2(L8G zOh=ItNUBY4h>OY9`PhIwI6%59+nFa3! zb(nHo-S=)kbmpXPIg|hAe)ebYrI%w87Ocm6(th8}8W55ixoxe!G5m|v2A*jfWu0q{ zdb;Q-cl3U4ccQxvn2Tq0?8cRcua~v(rH)nfY8lu%L~e4T#F^#JhK!x{+r_O-dB14i z+rywn(G8V(AJ==$?mo`buHadx-KP&F)i1Euhrnp8_ItS&%F~R{uVWY9$z5OE&ptNn zDMkjvWngIq{p0K-py7zs-AXdlmI9_yS2JfwUJYXpS3N|%R;@c*%#w@uYoalVR;{L@3w2>?iOhOQE7-aC9_0` zI=3AqmQ1dD8ylx*EA<1(Rc-xk>t>_&ZiDZ4rHxD+dHoWY`r+jIv=+EnPuHiS%!|NY zn9wf>W25bjMRb~b2V9iVky177Ct|`@ty9xc%QLKn0QC#4iriJE2G1QJtbhq@37QDZjX|*rD*D{@xt-=x5wjUU8z0UAW5z;sXT* zDpWOC>}u2YMU3LEG+cmnqr$)Yx{c)!v2npn;2n*ieohYC_}wm=@_c^-+cmW^Ty6u~ zb~gb*!zen1#;MSFD7+lV+G4k_9d9UQ)W2k#_VDPE;v4VH5pzi#9&_ns1A)|RrM`o`ClL(d+>66BGlE~I&0;zz znzARAtlJOK6h+!wu)6uCed;$xPtQBFfp0IERe(s(pjjqM*LNBv)z9auP%~3iwU4#3o>NU;?SHup)C^x0WnK+? zfv;%gcqNv+mzQ==scE>laB}Hna`5!hho7Guzr0kRa@$1%Dqp)#-&URQud2F65tCT@ zv=o$uJ9>>i^-EV|J=wFiHnX@Ng~IG9XBhty^Oi9*A*N2dOk+|0pOzg8TC(~#bLtCv z_Myjb4VlMNfk;7{ieCnUa9iHW!O)R=vAg<`c@@2Nr$Y?UWL7bfr|h{=^l{k^g8TQ; z^kU((d)05vsVDjk@+uEa!LjihfI8T`m{Cm%o0-L#N)=SZIf8JeMzDOX5ga+O=ttL} zrb!8<%E|6SzVs$S^mQJnv48bIbcqkYFzPw5#c!EHshI21MQ&u7&;Nm0H9|@J5@>r1 z&mKJg2gC7V6fNnu_8QV*MoaG%0w3?`K$6~gh+ZE6Ms{X|2*mqPE&zC!gUNq37vI?c z9sN0=t&QRj(V+o@SP@@F2ENKb9+m#l`ATsN7Txv&5qzw-wWL9MA8~~=SAWtjKYEPo z_WR#*^BmvWGX|Y11N&E*x6~L}0w#8wS-fUsX9W{lTh>(8raBPkZY%C8+CbIkCK&~R z>_9c6wb}AhCb)H{DsAnQi=dO?1+{SbJiQ4qr?{(Vvo*Dg0evXir4+zjMVqahTL_qr zdh*6&?kd`B^^!#Z+1|;7X`-xB^cnPD0_edl>{Ybcrfs(W(W_{)KkIo+xDgdQ=yY#y z@nG?urAw`^7T|-rP$s4~>RN@#&tn_S~JK?6?m01CVyu zffs>IE7Pn?Mpf-+?-SHK;EQ8)_ zlWu20NU=45?`#V7|4NHmg2EO)r=2o(>m==bw1rhaM8>t#|dsy5@>#vVLLr z;Q2}W6aDN%3VXcXw(#N;eS36g`tcVU&Rt>sDBEQlFFw(iE9h?Ir=_BND=$9Lm#gS) z4RcTQrje{`wX+{-vpExy^$iM~Cc#NyHu3;VmyGXE7k(YIxi)&!LJMBo0pzwnoxRM=k- zp6J`Z0s|ZF1$vngvp4(V6MaWnZ)4Ej!p#ykpX^`@Bmj^2M8EoOxq!9HH~pOGn~(Rj z?H8Zuo00h$0iBn|cJ3i4|PzvZ?&h?M?O`m4CP`vLmPK9E3nrJYyQ z5gBdk4T7MSrnnfV6iZl8sX7&Ee+hrbQGD_30cYTK>h{a8Aomf@T=NI(T=3xr(M0Fy$e&*1>A+r%U5l#l$IhZH^x_AEj@v$g{1dDDqqz- zDt!kyEiz>GZsai`9_@Y#>{k>Sb%4uitIbfA>d@r1xnYQZ%|NY*s*H`R$?A5HTYD=w zT`b1ELjZ8lp#B|mO-NunOCUSUTY7Air#-a0Ak?Y28@>h`ubm9rhgeUa9zAJm|b75UVC|1&aRCKiWpSP zm3DyukJH0WFYDPeT5P?~o-O%@o&F(nUs!C!s-dNgH%8Hmq+;Ku?qji)8pEnDj$Z;1 z4DgAllJ}6@wLqzw6-a>RUfVYvL%6Q zCH0r3L>(=Sdkg1MPZv&i7Neh=r-{fOy@3S;g>(Hjyn#CUOep*1((Cazgsg>dnNUqj zuy_feeyyY5o75tsd2!K0H&l1~fPLdJlOEKUGcRqkw)bOP_agTf8{M=fY<8+X2OUA z4|D$$YdRvw^miln&q(VuTBmw+b2GNAc zEGJCFP;1OkJmS>Oh_pJcmff;iuU;g|s#%_g8!=lf7=K#I4$xy+cWp;s0Al@RJHA)~ z=v%Yl{o7&}70)=en_a9eQEQjBj5ha9Zn;jaE4t(9R&=Sj*seY}HWgi&bDTq___a?# zfa#iUtvswPPN}9lc51ru?F{W=oKmfPJCjk<9p~3{$E|C+mG)-#LFYzIWz#>$P`}N* z$sPE+D&x}q`Ce657GZiGE8#E9h)pG!ZcF1TyXV^LuQE;#R-8!ViS%V?8~L=FFmoWxU=dTbrtxv+rzpFJolCh=junU9D}CX zuG`I~D`qy8Wh~1}E$7l&WfRlR$u1&J)z9$!)*${MvW@2Ys>LZh&i&rzM)-Tw<>4lK zBYZ-dHuU@gJIQOeT5D`~(mNFrW@oe5m*!zHu*hGGU*EbZ?#wPyPVx-_NQ4s&_R#N_R-c|6uP!mgoD0%BGs;L{YM>GhVCf z$wlfb=F>eo?IELgTvPYpXj)J(({4|eMiX{T%>TnV_5Qx&uPUx3+puz6oNkC{|5b;c zKRe6+u7D;LQN|(U2A#Mxe^=xl=dwF@_DiEb(v#bhU5XnXF6YFX_pC_rY&>GQQSX7? zKVaiDU6IGUd-xQU=#m#4My%v_=jT|uf4d|tp#HCcdZ<-kVbXq1qlIJ5;`2Sx_UU!f zWHme(+nFc&(I%K}0`|Q-(IEbHq(eac3&E9~fu25jzrCeKeZ-9k`F?C2p9oSG_2&TM zVkQp_LW_DN{`MB_+ZaggqNBgv995G87pi{>S%FPaV3a+n8ml;6`uVN>>Nf=?4adDT zYt}rGo@2T6wS>-?+pW{Q-UMq zaO~>&p0=|^#dQ&9;Re|fw#ygG_kZt6;@ULv+S481FkZxKE%yV&c}boM zh92Ztt_@t(uM0?L-gj;KgZEN=^zmvjuG8%So+K_sDvdyTytJ(mn+aD81LsBg%y2O2b@eD`Ms6w!(*bM2iPfA(Cjc^_>iiTl;P!@Kh) ziE;>jmhjv9ppLX>m8Nv#(^3Funuwf6^{8qZ{m!x$UR0`nWll*~b)Lq489UQ(88byZ zsNhV|qxmzY3StaO7h=AVy~=i`kr#W^9|5SDl+-s5)aSI$Kl-`xf=ooSVF!j_hkF&N$Umr z;m!8HjoMEtwt>@1>D8_J(}DV|d<8!Id~uQY!~0J^Jvp8nUY!hI^Q8Zu(ha(1oj>Zw zmgyyX<9_vRf+uo=C7QPN3%keryQKCM!S=)QqNthseEDj?_2}sIotYU%YrK+8-;0-?GnZ1yNlH{+vPJQ7&`g6I3`SD zwlNHNwQ&FNlrGn!HfHLOUdrsbkYC2hj)r)*iKzFXhX1jE&5P2 zTp|=L6t;qV0|NhQVWx8KV+YewPu^d_zgn27Ub4611Mvk{=5SSCMU0=(;tei#>>@-m z8#XS2R}0zk60p$y6wIrInVxCmY?%u}PvKMY$v?~~=j z2A|t27>aqbNkF~Cf7;S?0VH#!8*3J)lpGX$qx}*W-N`yBR_u7HhsXo37P8vrG@@d>7ozHa-q}6b zIVGOP(XlQWTj#cDtu|PcKi9^{wD<;1(Ykf}=H}|nxsP5{Kg(8HnNK;{t~^{{dg$oz zgh~tLKYEb0%fjy$*=xPT>I~1ncmLpkD~*dk<2#)z!pc8b*;kjCdoc8EO3@+4RYvWB0HQ@{r$2QN^}uCcFE}HCkf{I2IjZ9OV5}e z4{-;xLU-G?K2f}=_{T0erMzU&&VnWtbk`O}#+S~uST1>OFzxEq=x4m)OV-UUHUMm8H!A8ay`+~5VY0vd?Hj^d zbLskK=32OjN;GO!bY?D%iT?^Qpyfl5)!NvM4!PyuZs?lK)in|EwcF)`kd4|X@H}^! zUqqO)l8fBFxc%7u?tYOw|0B4x@*!PyHUBs@UdSz13){l=XxKyEyNNt>1<|X9JqW?qY%zx6xndP`uRY&g3TG)|%ZBdwdTBa-fGRCGm zCo~;#@Y~^_K6R6ut`@dM*pcd=dX)DwC7a7&rrucO%=TU5{`oS)Z7Q0y>TwD}K zNJU50<9)fyxqGW`Y2_zEKRw6Y7uVWH-pN9E&sMLYZinFjWzLJmODhZU3$?o0-99c- zzdXkg_gWj0X3uJ_`e0AN=6Pn+ZxEXaubVkRnZNR3aX-m>`!Pwi7GMZ?ISw23In&L~ zqVyQzJH!@s%5u&uoOE66gr-}x1Qk60B&m*>pP1t|8!AE6~GfSYD1--(Y+=3lbc2bW9RO%<$yYn4|%JPtXy5@8QUT&6Hm(*RucS)C6#Yg$> z9(Q#$KxOrX;^y_Y3$R;^=pYwVEEi%p?V@RbQoqi=;N)@j@i|efm9FuyDz(LkxjtXu zdkD@hW{N`-IjVT>j;YUtig=imTBKMu{ zfj38GPFcIdJyEoAU^uys4BG_N2kFVR8|pudn+q zFP;6xlm+YYrgG`i7pVBUF_Zs!|o>}hb{ao)vmmM%S&*<2SD-WM9OQ6B_rH<8k zCsP#vVz_!B`G$i(*oAQ3qzPN|0ZPOK%c^}t%&F((V)2=1Y`WM*y zLtr#k`@LKXN@JmG}pUMBT^z5oiRk6y48 ztSsiXQ15H@%{bn*NX%te-Ur-E4coX|3yX?&H70&k8lp|fED@s4bw`Qyv+Lfpv?gAE zAi1iozinM@)ZT49UTE~g$@OV1aIv1QPeqw0fxR%HUl7Jd+bfIcH1`g;D5E2#YT8f4 zgsobqUa6L6SPKPOnNGIail&3~DR@1RQ;b6!uEqPBvc;+z4=<(57On`WbfJE6pxy~R z;(p61N<>(GT}5qKd%(q(zc#fmnT>rH4J3qL1^ZTxN*&0)H~0pprl87{X0W1rUq(J3@eg$B3b!KcC|#t#GvR7O1v=fLc_ldtd;Q~>I}BK z+ChCS+qw`$guc0B%2)VMTNPe)8#CJy)%S=bmqA<$K}&Uu1?QFhy_ZVHKjHql$>jRZ z-u|_d;oO}AsC#EmaMR z&Tbp^XBK(DL69l-0rem%nEP{_odQ=*2+JeIB|GY0vQ2w1z_h$UY*yY5Hu}d$@ zO|HI!y(bYU=G=?JG&6!=t?0C}z2&GpSrv&}Z}u_(UPM8wTWM(OY9DK5 zJ*S$y+W&GHs2RR0%DhSJG~QGnR-fkNZv9M>_agn9bsO)+?bp|Byz_6l`3@qrew6-d zL+hO{?h|``d`$CBa_kb|^d;XTs7s^tqWdWFX9Vlbp}Otzt+doEtH^_r|H&H4dt{n_ z*?WGmr4e~USr1Cb43)}lJIjor60b*hGt{d48h1-$fE!FfM6Ke88YdP3tf6XMassJ| zLN=n-UI9a`VMDFt8miW1Q4~6jXR7!ox~5sa-*2=)Tgit7bg*=)t@YkK4K>Gl~Kx3qqI%aKQdmt zUEMM{IM}~_vVXF3ZTIB4@&Qx2tLdr9Gy8}7zr3pWvBSxWJG(F0KX$9SFge^aKlIjM zQ~etC>dUv3n`%14sd_5O!H*XAK$QnV4Xx|->b8?-kM|FsbB9nLQMcQcq?ZXMlaEY> zH`IR^@pd+|3U{_t@AVhxNy(c}st;EuQ?F9$j#}>!L8DzE8i*6!~3w*K{k1!UKQ;hc;^1&TB>Fx9X{`*5bN%<#rTU z-U3Z0P1QRWtp)b_YQv7_>Ikp(cdKh|Ye3Dn3nmIt!;81#6;hDPdsLksY4vS$s%n&K zEj?8}tM=u7tUBrxO{kVyYu!v-Z6Z#6<(#T&cWv!c6_oRBJF2dNZ#!&xg@Tz#+vu9M zq54MybIe0qNFbyW{sG2p7xoZz1^a&YGm` znT8SGLWg*KL%-04S)UqCOid%?@WP`6_0#>+mq(12CTLp-3Pt+v=v?WwxW<;sqiQ?p z7+l)6np`+jba^*8s&5^*d!K4j-1eJPBQrvM6(h2|u`APpnvz&ut&brcdsug#Ir|Am zcJ@81W*ec`1nRqp09l=^@4PtKf9BcK6DPc>_vi2`J?yc!b9i)oZTG+}fbnJ3IXX;S z*tvSJdvY>-gL>WY_Hu@D2V(XG#kL?guvLGPSk~^ZV>#_dH#i!QKt#btL46CoW8OwI zcN>r;gnaQ&&(+-BM!4jXzkRg0S%we$`D)W0ef>gUC3|B~JcU-*MP*^#tm(TLYm^ZN zW||PFn6jPv73xrF;q7Xv-qui~)z;g6zuZXxsv*8ZBLO=8}GqYhrkN))kZSLzDV8f>Ex}TJ)Qb>{$Hb2 zT|Ioc$YEE=pWfu3*Q(+0Haq_mzqt4*nlW_S_N`R!J72hc@qAbY;&a@*`lf-p$r8p1 z!!R5cQz8Xn`Dz`O3zsYToQOn{QPqEFN#=*0E%gZ7)3`4semtIll#u{dhw3(4ZZEWK zx#fZZ5!46OYYvW1PL}rSF3G;G^SG*p&UXdZcH9wL_yAtfoZ3@wat3jGmX~_Eo+C9% zKi6=4JM|PJW#Dxi5nJoe1!h_FOuXhHhFJtIO`Y%+kjeWd0MBhDlc>jeFGxjJ@I<3f;(PrBj4BXIw zf+uFhl>eL1RY#t`nstAqI#OON9yL%B0cOB2_VSB$BLZ&@@y!P*#nrQ3;FqlJ>h!|( zA^jhoWy|{tQUe%8zna!UrGLFbj|+ry{84NHSu6%jfu_2P?9+oo+MAWSe4#$1op*nw zc$LF+>Cu7u&80X&ImecA6f8Cxh3KU9roG+M-JPe4!pG@wVR-(D_kHLgTEzL?e~wJ* ziF({8kDflI$)1jXt3Qk!am+BO?_E)^F7udUX8ANl2OeViV?tb05wgA@{419D+3RGwDaZ@l{Kr-u6GfqE+= zVD(Pm&f)I0CJucuSG1Um+jo+f#m0HC{psZy$>iEolfAwD!)H#GOh0CE^NcGOzh4}v zzog#Ih*lPmc6NzmR8rJ5OnmXc$0_C17QT5~HKIEKbc3SHIbI)^V0a@nN6}O@`QUhj z9XyhlcO+HR*AN{JExuYUlUuc;7CL+4#rCYz-?+AEp5NM!O>g&e0Kflgwi-I(u^nN zNezIm=Dnx$T368hD;{)X3Fx%DS4Sv{CmiQ$u-g<&40PSXLYiS0pgJtYNskz0|<)Z9WW7 zkDi+xmd^(F4aB;qV){U&TMoG?e?Ak?mALFgKq1!1VL4?;iJ5%J3}kiJ^oa`~p=`3wK`chz@Q&}}6;9j-Po47LWI z16csB8URX`kltsRRcu;|$lej62KDr8-%Y$nY@I9*Y-~-$DHF+?0ozA7D-BViL z7K_qIi)yQ)3s|4l7dT8GkziA}CPAe3KZ{7$`Ya+tg?&V}tH1nz28(MA7PI`x0(8v) z^pPe*-mC5&Zdn=f@G4aakCzU1U!ENAq+=(ZFrn)0lI8jt!7E+U&*61lg4fT|cztmD z{synHTX~?(%iVS_8_N13kyjM3>k`1~0AK;@nx6%1sBkyH{)qwXdIQ)zhq8!WUwmH! z)ce#u!^>Gft%9Z(q&71MO0gRlvQFiaVb=6<$f3TUhwO#~vhUH5y=VJ!1KHTMa}mf^ zeE=oH(a{$iJD{3teVffIiq;JYTD9+4w7QmO(Hbh;4XwXp(7Mr}wY4uxU{+{!RR9zqxWs!(_nO5m4i}Cnaa3T zXOklpJe2^8;QA`1>7x`BX*VS()!t`O>e`+~X{fM|Qc%XM`(yo$YtbXCccq&RR5M)4 zf^`#u^*s$(?^avGjn-hT<0%f;q&W~~x6TM=zn91CmISj8YRqnKuQ!;D-QaKM)sxA= z(@W1yUS8@gr`Gp*O&_gb%eo~&tM)sKR@dq*T0@0>w8DK`0MuVMK;3G9+RBeCM7NSH z%HGk3Nw&gg8~E6YB3J$Q^b8bt7<~x?b7q7cy20f$c^L&~oDXSaUf$kpkQsYsE^58D zne8IcVB4e_B_A#8Ley+0D#e$+x~AD})4)0ZKr-yf(OHQV4)1JjmH9q~4T z$@IY{nQo}H;ff@gHWPivM2Y}=R!Beyu*;HEnh#jIyz01ZGDU!`Z!w&HHl=e>HW-&h zfYsjT0qfeH2W+Uc57>4Y^|vjo8fuUTtHz%^R+p!ORWmYkhJ$PX^Qz{O)J$jJwlmwL z(p6*Z-0sp@7)i4#N##UX8Mn{O8WL+5p}R7PP-bn9 znf)URt%gpNDk=qTF8PI#G(WqRy%sIE zswlgLBpgO$P+u0&+EI3OXh$q)wfA|@y0+&*8!GLC7G;(YuXc7dv?Eb=HU8w$x-ymQ zYDQ)z17rgTQ_W3bB57tYMcLJ*$1sxSXV=QOeP-8?sKW?7&Ftz>lUSf?|MNg~tI%D}eTV-cg!%h-qSKF;TURNPrbF-v9HiOutF(7O;frW{ynZcG`*JKyPE{BGZ zv_%=1`r{_Y*Cu=WbOjnU!yTSXdd#gMlZO$wD`Zh_b+}6`WVP>k$hwy2AsZ^~Ll)$f zi(zj4kJeh)@R#DnV8)v~Qr93-#RXqxTn42`e+W=bSYe`QW&lOSC6_b9NZMlg8!>I4 zSvBPDFd}z_DB80+Y$X<=+WS02UEA{z4VCsGno1-QoZ8i3LtBckfmx^WNL^1Bq%T4& zi|fU#Y;iP0z-rD46H_xQGkp)ps4i`Wk@S9je6zm2+8p03M#&J>nE-2dis$9;~6#K3G9MSqSs#uUh%k5SCi_lE*0KD&5Ob)u%-%0qW7ZADEF((@z$v8&lAOkB+jyMW{{ zcKI-jq+*go#QxTnL)?st5^hM$VT7J$!WFTR$hIliy4L4m8!GL?wq3%dr;(b3YuHFJ zegpIYO6s0S1$0}>b$Tn9YxQ6r`M(nO| zMO#@BCuyuBfX}k7{du^CO8amHyE(2k^-nF#8ctHIH#Olr9<`X2By3JijLu99NrOP} zYHkV>TQkeFfR{9O*)fcyVqTJM<)&+?U3)5mZAj5!1nvr3v^W;=lE|*9EbCgHhi#~| z58HNG*5W0NF)xYrCXdx#cuDNI3`&vy5TKfw!bH)`vMb;vja_aGBdM5|6g6!md&2u3 z5JW?Y4kI#J5D4I_P-YeJlE|_th`P4tAsQ;}Lo}I4LcW8=OB!Qd5^ySy)OA=z60(lK z$P8AKhJb+8>=Y)ZW|mPAU|oI;BdK)LNjr@C;e}NHdEg4ugb0_yAx)>N$Z_;4lv73A zq_OrszZ7F+V96t0q7gtRfpgm};U56G=0J zDJn6!R2fFntP)eu8YsU>31!xhv%?5I&CDtyDvh=Od7!%1=Ybk3?E^KLP(rl+k+o+v zY^7A4DR3)~S4>$F_R%miMecEIwSkAQ)if3+u4V>XlwDmS4I^oO(zJRDr70R_*O1S{ zh#XIv@|go=R}qP6ti8{J*0ntk+E8g9v?#NLc(t>uAu)-vtMMm~)~$^6H_5oB9Sjd==j>hzsqkZd&h}_ipp9QLCeHN$@rG21w%dVu5 zrcH>ltL;`6uPc{wj$O*^ZGWEueOh5~kpzV8Ss^=tYVe9nwP8|Ki)wJBC#kn;FbOM! z>=bkBiiqSyq9#gBp5=MShD!U81-UiS{bp5zNf(XQSeNSZ65~xCsrYs!VV5d9E`wmC zKLn`evoK-rO#D+U4J)wxU2z#TjHGA-*kWlMad)IrjF!J7?u;M`s=+HFHWP`Of~aeI z9-^VrK15SVCW2F|8cYgloW|F{tmAm3j%U-Q8pSf&(L%s#CJO_DneA6Wm+FekrC}r$ zcd0ISeXj--m`KzaffejkS412p5;X-@*X}%6L#2JNf_#FzR7n?&(^!`(<4Yc+_%C`6_*~v zNQwX{?qF>k!H|L>Fhh0@BkDAPDWW8iMN`0ZZO#KTRN4n-JHEcvrMhA$NwF@~h_84| zTCckaTT&yRp8*^-5rv7IDeg6$^92RER99Ro3?r$yOVzV;WLud_Kq5hBw2DPrP7wi# zL`?zGwK@;TP-!2KV9P1!Qe80wq*#|~)SWyo*BMNuWdya?MPYYqR92WRHSL56`k8Gr zL9^S6OMGD@l{dTDk6FZTwQ<%6APvLh8PaeV$)}k-McgAYa0;ug3wW%CO8Z!Cx8wYT zVKb~4?oqVaE$LPsvzUS;e3NlGDF8DwAd5ikY9Js*>03cPN z@7=z;f$y>qzRbQfdx;$Ka~P?onRZ2_BrF@ zGUDQU8=T&wE)O@`<23g~byf3Jn0zp^B&EjrJaS#r^T-XA_K^!$i+&tuZQ_EwXOm&9PMW;(>w(okt1r5XA44=tz~o>E(9@nD_K19007$T!vcaJ3!C>-Yk%tFCIk z3KLZ`TNfkDx{Mh{(hg9x+6|@1hnNkSJB-ZJ%&a21QsaI;v%1FTF&irFW42vp{T&Og zhOQ)ot8pj~SIk+;hMgI@=YsL7X0I@@YAU&YN)i2)r|c5#};4>hbdWh3f#p*)?zZ{6pVtA z7=)&E4-b3cx>(a&n8l6FX)PAj4mJrJuQ2Z)h+nmJZPd|K53z+=@lKabT=X&>O)|PqG|q`;~|6`eJr!wFh9ZnpQaktmr&;RkL82$eI~)QGM9u;m9O~_2Ec6Rc}qK zWYIHF9~M!d8t3zLvO5m>rsdYbxhngqYPd7bdEh1#BPNRcEEFK)c+k zOSEAm70184UG7r6=M}DnL$VJeZ`ZXj+P8`rPGr{9d5&v%9?Q*Mz z;S_C`V|>KJ6h9y*?Vw|ZWgv>QN9bHju$r8hpCVKhxVeR76x4*}H@8GA+aGn9gcUA& z9nD0443(tts-Qa26+(No&CR6n07S5Y2KG*O=a?+byxN-XiXo9vh zMNGQs^SgLT#G1>UVbak(`YdK;d%0mcygklzsQYpZGt-eJmmExT3b8rG)VU@iERm>* z)aeH*EfjIi(?8G&ZG<aO9t0A(Qzl=h6I2wxGVOwCeZ z;(TU^{5U|Q%adUwy-y=@ZF{9b~Z_K7BY1hX}e+)Y(LjT^du5Bg-O@!JSIb> zeN495epXGc>NhN48hVo4pFA|j5r`PH-qWCUtS$}jlA^T^qBvf`mb2CtP?FXXl%$!J zazR%s$tYOV%DYS|pxT%{=W0lJh2 zEq)+K{}%$lVpOXj>~{ERlIr zh;<#nBQ{jpM{JvIZT+f?)z!LY7)#Nv)`Vwy;I1c6cl+sTO$ZnVH^lFs0oKx*OPyhG z;LS9EwPYL7?M9StI??$mNh{pqc112)C5u=~Bx(w|uIYK?hDs-qi}t*BSL<4WwZyuU zhwH|ol$peSJPx%*q!e0SLyT(L3KRJ=D;NXT(wa-0VI-BamZEm=Grxue9!BP_C`I{I z#9AWLrcmk{pGRq^bP}bYkXq+gi?y_tU@ZZ!@?hQUm|La1CLmlK>=3h>y23=2m#tnA zrs0}Po?#^IfLW^(v6svml6V-IyJ8k)RuOZFOq;^2YkVHFq0&BPVb1Xv%*yI&T{Fz3 zD9JEDgmEYjSIl0@2BR4|4n~-H5mMI^45pbO6;zx^M!{iEUd1WwN$dlwB=L+^!63h` zi)c(u0I&x=p7~jccUe+mzLDLkj7FakA_b_?3n1de=Ldatc#{cyU-m&DZ4VCtB zo1J4x7Y$iijx~Pb0gG>C%FeN7=r|bjbF5~*ywXTTIo9RgFp_r2u}*WM6!8$NAq$9+ zy6buvd)_OY7WvWAQbJI5O8lPt&De&r#HDNxxN)E=AxZPFYt z$DS1`6zFPQcX>Dr2E5e(tPe*ToODq=X1s42XS|p#45w&UE8|QarkL9#?P_I) zWtP5BS19;V)k-o7YQoMuf4LlBCsl} zt99KFmZDv)5l8Z{#1||{8(kwq#z6-s0f^Oy{kPGS-iH^Z*hbwf{zcC|)*#zS+HflyK= zm`tVJuu)+lLNzCaiKLk!^hfx<^L3XM!$|t5J_UVX`@W{{d^OT{K3>{ArA6}T{?TFl zq9iAoL!5l+KrLQ^cCeS}IO|9#Omz3pI99+(v*qrRd)}!xGwc{$VH342)G+(g`ou6 z(Md+ZZdBZk4z1Wx17P%3nedHpbhTGCl803;J zn(Yv4M^C$zh3>|seCP@r&(qSy6F$VQCaEx~%*hB-+_kD7;FIo#h@ptR(Ci$g(1$5{a6Et!sH6wxQBV z*rL_4vd(Na3{fd|X2W`u$Lh+W@aTp8c6>z(Gbq|6Lx5_Y3KRA-1L*gG?CLUQ7)c+} z8T9h@W|KkhN@UOmppopETY1_D2|J9)Xh9&zx(a1h5le|gO+nPPJrB`P=_Ewcnf3Rr z%xYLlEuG3E)p|JY!6vJ2sI}pWB&+(d6JPcwm5d-40#*}Mn3y^ffOV-djHD>~^TBMj zB--0}bgr`zGIkh|1)zkp1}Zp-Xi6k%3b3y2dBBEBCjkp{>Tg?EH8dp=R*lbitXglz z z*ZMqAL#2J7rgMzHt5ttwq1CXKQgx=ltvp`WBYDkiVl(PY!PpFllg5Cs)m#=PuFgbu zb=fqGr1=@tYB=l#v>~a75m^9CICG%vDq=EiXz%mKIj-$_(1uF;phd|vJkGJRt6?&U zva9hYkJgRs&E;ZlHa23#fec})=`BnoIdS_+9pJbG8%EOn?CLZbItqj22baMi(}xkd zD^StCRYY&v(EjIv>RO)%YN)giR6lp7i=<9=HS{J?cD3Eg;}t&|mz!PfvE!f==nfeC z8=C0C#MR8QE2;*&R2xRp7S-TL^Hgtbt7OnKPz@H5oEqQrkaaE3LpD^}hb+jgk+xO4 z8f-{T@oF&RO&+QE<|T2Zm>HKrDbgS1)y=a)Y68_@l2JfvYO(x{n6}4iFo`=Oh=Q!T zDPl7<-e)1|*`9@HL}?$Qscl*Wr&cwX6w;KY_!^jXDvQ)j{DPc#nahsMU^Qt72w2Tz zVe-_>GV1s6$2XfICX;jpONVRqZA#A6h*m+oy2aaPpwaEU*}i3SFPPoz79re>`$ z&~yTRntN36MDnJ~n_(msH--AWwK)&WQ0Z>K{3R=C8cLGTk31&vrAb>& z!dz1C9=!4q_;l8l17khn{A zWhPy!B(!i*+ZC!PcZ#S-O$_i*bv?jCHB{P%D#{}vO06!{O+!73b*U!3$|DwEl_YFL zO^)C9GFQ`3UTNThF4aw!2*XG!?ox#&?r}RB(sCHdr-@t<7l{m<+K##|;E@|D-3_^) zu=1;+A;r2>({AOVyPkn6NM926uBK&%1tiTzVd83LNf%HEH(fpqBdM6=5HXmw<&Z9| z;o3MP<}gA_N6V=qiP%VF+Z1eF>+`S;mF@;xt4np$u#sZ?2Ix^9vG{@|A-@5RoiQaq zuxdsM6HPM%E9g?)bU87Mq^Mvj?owqjM!b_ing*MP7+x+1y|SpJX}Mi zeYk>cV%3xyE|RS-)lI`migl?boX4XUvyz03tBKJW+$IeI!TV=`m$d1!V;B^Oc}cdF zn?|PIcu6Ge8Q9i}cu6E`3bwB0dDw<`dt+jYbo0ylxdXvZM76T+GBC0(JsZ6Y> z2%uYk26#y%qoDpIPE%q6Q`Dg8A~A_OBZz{`x+UTzHQr|{PoC{rh(?t5A)3MVwRlOS zkmftYcuBxlEK;|6;U$5QnTeoNM%~iv6eds2Y&i^wge2qH0IcatL`z{3R|r_~922ih zp$iSSMBJpt`#fM>+w*`8mG%J(6PAx=W%bdJF0HU?+{9xQ^OJ-e!T_sgWCp9q1~9K` zqRJ}`Q&3{sa;Y+mq**1Vu%%E69SGErv%?529gQn7q0A~GDv@myi(k+BJWxZWeV`^2 zO1NkJBU2vUGHj((ohfiDk5^1t%BV91V>5_N8Uw;s(^#0eIuqH|CDJgGqGD723fO8b zv>^sT8}fM=k>g3TmrF;tL?osy?R`GGy0+&*8!GLC)=!|YPt?h-hQuUV2^*jBXkB5z zgwh{8$H`5|Mpy__O>SW#$&Ss%X)=IOc6I4BjHLP5)oCeo6b8_TKn;05jL={JCZReY zMt4ALiO5Y`+W$OIUF-8e4VCtRnoKA-yBcznD7)HjdF%1O&+P3+$1jTGUGBRMfyX4YCa1S_Ge=G>oRH>Nn0#&Bc_d5%cWr?6*r%*gspJBJ*!Hh3iD~#B`_cjZ;3ceBx(w*uHAXC zhD!Tj1^EPbsgf?b5*+JNWqiqF6tkCvU8>Be?Q`h@gr;V#FabX^Xo4=)EtfaLNGk49 zjTp8+S}@5f#Aeq-im)jnERm=wY`TW$u^B4e4VzY%>XzXu#ky2EXY%00Bqd>&DmQH3 z=+xvCCgx{`PSB;g<!fxwf0mFB_BVZzQ@n0hj8&jiV<=`|yKXMKtrg`#0ZEt+ExEvTJaf%yOSNm&N zg_sQ)IgI387r6g7!mPtX;xVh^0ASYj0>EsjxR2R(8_-YKa5Y?{SlepSw*b1!rafHP zq?#Npj!1Yk961TXnN~AXnBbb(%J)YRf<@FMapx+ud&?~O{h z0EAsP00@VQcZ2Xp=d6rt2uiX3)wFW~hA~Y^*o~T&AOqJJ#G0wXgxAbK{DTO@E?I^V zRm^~hSUG1W3-KG$b{MIrieHDU#N$`T0Kl*70f667@oxD213Lv9!cvR^0sRV~+dF4T zNSlE2V$^Ed3KLW_L+uYE)Vj+nY5cU2}ZGIvM?bwGZg=A zgkqOW!-y)TG1+#`Y-)x04Jkd0!~%>$rUu;lIz%QOzuNx*zpnKGenZ85{I=Wr{-zCG zOlD$z3P8KYAPNeJ%2z@H6e}x6swT8B!8FDCluLO1euPw)Q^Sbr)YVY(6%(JLCan@L z&b1-8r>o3yP~%+dFq(L*>IeW>b?pzZ8Y=E%HJMmKs_-A$fW@pP;4uJM{9K%nrw@i^ zW`v|EFmg4!g$b>*fn1kh!-$#>NUJ$f3V)2;km$q6EFF$v`zx4b9eNXwTpa-bxvu>I zazn*^-Omyf6VZvo$fSN4caM zM%1i&RM?y-g%Cz;NdIA^j;By(6$01Aa9`_?o_NIS7yyWMJpd3JD()jTnP9?P`^PqH z4f`oolnR^+fQva$85O2ra0a?bb71^xVhj^rXCvRb+#E(!DexBa6SLK%2zRw1M~IOb zE$9Wv#`)G^M)By?5dhHZ+8>}dRNO}|$}}NxoqTK9QKEcnoCRRp`<+d>DcVR2@w%L# zOU(?gDBqHh0=iUwzIECarLc$jmLxs{w5et1?j{zoO$_je^*q2MHlnzXSbq_WcuQ4b zS=srPl+v95qI_#Rmj~`LzQQRt-`ayuCEqS<`V5m0oM@gRm=&ralaK=ZI-;UP)}q)P z>E0@OW-)d{77=4yK}LmZWwg6pmeHpg{{!s0)(6-P75A|VvTdZ#&aNSoJ{q>MhS*iM zA(mx4E-%wVg9|%X4|Y#ZhHGkNxXeAK!G4^9B-9cVp>d|wEEy)4X123MHRNTNr^ATa zV!;~$VB|~|q6DpsN#3)NXJtGo61Hv{UF`Z0gsOym!ZRmDsGNl?b;j*Q5%wk7Lj#}2<2N~A2JwR=!xQ|+pdmjd6OcYS!^T=<8IJ<6uC%2K8=MLo zXc?JuFsivQOd!n+qjVz-k5QM6!-y(g2aDTjBW5nS%6J(ZGJ+U^1we!e9xj7rRH$X` zegIO}_yDA#;@v<>*TNW(8WvQnSC;b>Kq;m_37ciPapOpY^8!w#n)1Sg(#&A`T?nQw z>4p(i+%DVLwr`U0a>)5%WSuHPWn?E3Ho5q9?G6wcD&7sDbZBJ|YRFEprrC%$0XXrc zO~S_5$dJ9`Qqx+tR9SRoPv$%ZAz%>za-F3IG(7nuI;E(LouAg5%X%+mC_n_GL|Gkp%xV%e|;; zmV^|Pd&OO|YqRN^C8=kGS-9Snk(!z~0P9`X3jnjB;z`V+U8=IXW|s}ADb_Wc^euoc zrZx#1X_EplFop(z2-f5l2EjASw5V%#*`?PoqKdm_p^=NvTPVpQH52^ zsEBxg+3ezyQ%Lt=q@L=AR>pH$)-eE5u1 z3Q+7!I4BW|8Beyko%ULocS+hA;TLUv$1=uK<9~o(*ZKgzq2fM%QJM|c%{JqS^wH3b zF`igg0njePlNX3rAytZ%DhH{W@4_tNW`D`Py7 zaZ`I-&;9_bq2fMPgEew8`8tfJF=jjg*8<4md!dAUCNOjym+9n=bFAjPFrjrekn6H; z7*WMs#lXf^lcO|I<18DJe;Ap&A{S;^8S81RBLHMs*Zu&xq2fMr+htk1i)U-PRzc2wcGrXdD)Yv89FrsGFoWkD5eh?cnff%W!!*PA? zFxSc`P-Na@u5~>C5F0A)BQ}{@!^N@P#WOZMs8mrZa4rBYCPQTurGmkylW#RSh6%5k z;TNuyU3v~9D%!^8=VH^%=yamvd>gWa7@51G7v@_TJ!-5Y0OVWO{s6t9;y!v&rU`-T zYD-#WdGQV3zhhU_6m z>aHtfm~UmYsj-d$fLPZ90I{LsK4LTWwobk^v?)=(wVexqi|JFj`PLqsfo^j7f$^K5 zQOyj$u!igsco-zKC=Ew?qS`$mi+Xy-A&ZESxGQ#Hww2MRNZ90B*|k2vZm4(?yCCsg zR89@q(5K=xWX7ievY0j{E)z2YGjN3pe+&>F$C@R>1k=or3RlK1Plpk;#p*X=++J(Q zAzg@(xhqy-o|W;Wng{@S*0n#tYN)u6)l_1Mw7J_|vtvV+iZ72@*8<3HGN?++3YyiW zkJPTkI`eA|JVdVM$1tHaGvuPK*|E#eVMG;o&8|(eYnEg^3#YU)b`%Ml%&@NQ0cu0V zebj>dlIWTp8+KHzYnJgS04wH22^(XXVVQXVX%E-ID+xx_%rJ_&W=Y6CxfJ5A*@zF) zU9%+Y835^uj0)AbpHHox@p+I&6!$?IY++&C+Fi4xkET+@x@I{~@hDxv*F6clX1Q^D zhv`|NJfW`H6_<3wq*7EK7I)M(cF!RhlbmM(p)#@)37blzp4|aLL&bfBB0z>^VOrcb zU9&5O>=f&ojd%=z(|Qk2*diPGeg+65r@lCSYI+M3>{H|ryT>4?Yj#CMa1uAedY0~U zTS}772$^VGx+0@Ck+4Z*x|RpX3>Ej03ATEYw`NHxoj_w`YBI|!jZ@S$yW-Mm7*VC?EQ{jy+Vxa>Ej+Rzsb_>)RQ6qwk(x-@BxYSN0L+Gp z`Po40sRV~i0L z)HS=}GHn=9#a*+kfqUKVhNK@xZt?eeb8cv``XC)TF`v>Q|Jnq{SmV;Lj$ z&jRCV#bw+uaW7^(MeWyXt__JljLcoJ3QNl}#uFJgiB;GB0IQ+mK30>7C0r=mjHeaM zcml2kkj-U0;X5mYjRZhsj9krmVM1$W`&-0STya@9jHvm@wVE8I%T`=q4#_`^%w3TS zv#gBuM8-`b*R?-DZm77A+;&;k?&4W7tfv@FF+hOvDgZ5JKIMYX3>{CVFu6m#uAUY4 z6RJm%kOKBoRy`{0ZR`cHB=L;uQ7G50$|z7x4Db{v&jUPSBZ~Wo&B(QO7Y`|=_qC~_ zRN!15xT}~9l~I%m24|p~GzZ47CdV)d!HMQ6Zg3UKx2qyrRDQm-nj9T{!C4K=wQ#dkvG=39Gk2D(XeVEk(O3=>{6+ux!Z@~TVVVMJ|FLyk0T_185^ z(h9S!bgJNPn1R_=MxP>KliAj_KEQ6MxQ|_sXCh5_b`5#e(5GTuv+J>rw!)_XvMUY6 zLHg({I|_qMsBXx7#7NaF87BN^hE%vRc6mCCs4Z5;5v2B6LtYK(LX6B)&9gF|6bYNe zs%w9M)lhLCtEt2i0jym^UNvN?`0|)_Er2Y(3rbv$VuxmMoHPYSuI9%up*1t)qORFh zm!ZRmD(;$HpJvxA$tujST~P}&tc)E+!X{Dc+8&@bRNO}`NI8kF*;T`iignF09tB`s zW6%Ue46~xi(!xetMpBGX&4pnCe`Xj(U9+n$8;229+zcDBQ;!s6680>BR7Qm&VUr+r zjSoN?D&7sGcGv8xVL`>ZW;vGvl&%|m30q{jLHdC^hN-5!FtP9KsHRySMqRV3F6o95 zRoofdn7eP1G0Axr5Go@(k+4aGx^@Q$4HfT(P%5G|U9+o(>=f&ojd&A)bE6+j8Hfve zZzED-Txxm?6YMj?CF+`8bqO|%sN$|!&qlp&OGAbaBW>3OE-e4bs7++nBr;vg17wDZ z`^XFqY{2F$^49FCp*F?3W}{980L72p3E!HH4$2@ut?kFa)oZJo%)*4z%$B*RYnFr* z&?Cgnup6mnSUG7IQHM$Dvye??q^2egz>$sX1%TO5aUZjA$L9mud24pnkeXs$vq|3q z=wfPeW1#R~ogbYj(|L+AyMuyJlJA_qp9AnT5EW=5}{Y#&#lM zlel$#0B{>B?&CJdyI=!Ywb80iTGc30N`p7nHJflRfG_4d3A<(!!!zJcTL5i=*EHjW zfyT_Xz=-igLJAhpD2s{NP__^Fj#|hsCTVAcUzm4gjHkx`VBPFmAK*7s+{bUbylXR_ z)-dCV^&0>!z9>rgP5?VF16O1LoNqPX<&{P%Vmz(6j2lK&G2@Bo@%a@ zF`me{$z1E&A7C|9+{bD%v4jg{oAI=U8Bf5q0J6O>p1{xyXp^SE$km(|CbVXjVAM$|c}1qGr{j!rn$l<7~u+ zOdv*T>2OSO3f=~_CZj;D=@aCa-C zPm!?6x|C~ufZb4WAG;t2MVj!O8nU5J#kyuU+IP*aVcL|qhRg-iYt^j;?ft#o)5%VG zs4$$nbAZSu7h$GdKPz-8v@#|k1*bhNR>l$I_E|$Fd1rLqgYxXUj3?DZ0H0?)`}0_h zDDGo5s6?f*&_!j}kVz@MuZ=H{S=aJecKy;~QV2wfkT0%T;o=z@BUkfdm^3i6H8Sd& zT^DhrK>dhr9(Yx`lo_sXuU@!ecz!r{VXz(RzuXXX)WZEN$vPv{q71t(V@Hv&sSNAc z9-uZ<+(#|Qu!*i&(nkZfh1URBmr=6AccBXVW4TC*re2Iu&4qcTF^alo*IhOaBdWM- zHe$UVDaa)3Spcbw3Pr*uLFyVGfHYLR8%XW0*>%H$ignF$E(IvX^e15>EjMmFTf%t( z!}P3Bo>159x=XrYVCw9wQC=0d*EUWtY-F59L(UH)>ohm2>oT$v88wMe*X{tJq2k>T zN{joZYnJrUX*AX~8*vfl) zTN*Na7-`YOAwVrWZe5pAo5-w5WV)6I$P5+tkqI_{lDB3_DGkq9*KE|O0HCW3lG0Lw zimmkh+NiKNe`+!d6HYVRilVOBb(c=Vh$`-y-JDI=EJ=M9*1IxN6A7Edtm_4U*-&vG zv+Xve_FJ>-hSU`6noar^Ko?V+gk7^q0hsv#Sp*|klUo>AOfk8-jbo^5cHO1dFrtdP zW}%V$+_HvbA4YURYK1R%!Mc)+2eIb3FyS?`G>lji>n_WN5!G4Qqt&RGUJ-GJbVD1`eHf{yieDMeiOieCuj>JT z-%#;x__e!c*A34pMwo!^0?^%n2Q@;1A{@MT)M}m!6I3%pE$W(GcbPVfsN$|!*1)}P zcSF(-BX?Ka!c$lo+ldUE#I5TCfZI@UAGfVfVHcdP*>%HqignE<+za4~uZt3P%_fFt z01Wolb>HyEz{OGPn(@Mf*vwFj7*8amfJcGSt(XmE`=I2bJvVuIDPTk4LUKs_VT_9; z7GPV!b+e4|)c7B;C|v6U{DzAA_(dDNkf>-gp4KtriS-)*?G`+FfeBeihhn9QV;OFB zHxi7enQe6u_|I0dhmdedM;wvUV2_>C+0d#;X9dy}W%FK++5yha^ny zIMZtS3xn4wLe#+q?3&$h2{??XXoZii%H@cAr|IZ83l@jO(NFy06=W0 zxR2P3eXZTavtf8psiIWiKmgn^gRJz*03GLM6te=MF@7~Uh5^OQ_Ox)N?9y`>QSDbywLd^_sJM^b4*Ay5qeP2j<5d9L6`1#+rG8M_dzT0wSI{Q7*q2!97N*X2~$YG&7{am9fjyVMJ}QGLG20*E}22 zg&3L9gd)H(&a*O})P{}#kY`={1FVLM`&dow6eECjYRHBx6<;2+t_4{Zv!ui;V0LH* z$4OIQc?HOXQo#MJj2%UWO=eiv z_5ih^;y!9YhD~(MZWwk{tZSC>C;%(wMF|^YnPC~ABJD9oH5Y~nq?ut9b3b16V6rauY0X1Q^D zhv`|NJfW`H4VQGoz*O7}+t}7{hn=G)n808f4LLuItm3V8Jl{gH(}s-fL`F>_)U`W6 zXsCEMgwml^(=|)_=wsMe*KEX_0GwN9iz;l9jeI`?gl#vcn+bx`%+e?7nk69x1SfIV ztY_&yx1}WMjB+PBZrzkon;PHq+fvW+JTfDS`^W@)F-g}fDW%~V>za)^l?Ui1e(X;8 z)@*c82DM4`;WBqqlUbOFer7mDU9%*l0H@-v*{#`h&63nY%yum$q4n;jjMUV`0l=*5 z1%TO5aUZj2TNW}A?XFqUN5eJNHJkJ;fG(ys3A<*K0x&Q}7QqPC7lE1!g2vuO!> z2eIb3Fvyr$8b)2Sn=Z?S5mn5Zh*-Jr4J}D5#BY3`D&JPa6fEO8k+4bpx*h=d4HftC z8?Yu|2C}zYlt z7r=Kqd?#VIZ9;5}V$FDALTrk_b@GcsjHgYPbHj)#CNkOf;yYur3&|wyS;)IG#uEvf z#II|8fZtGYAHOKghFf2|YjzVeo>-p((Bg}tgd`|-;EaThk*fJFO!&`im5dlqn=a#q z5mn51iW;}qTpJR97@51SkzuZtF`me{Nvyi|2Urai_pzFhYi-8UCT2VV*8<4)!gvBh zGoVeH0wY&*UYO828_0E8H;kzH$hDdrot*(#!$My1W=Q^FWCkNJk5#%W6S%vTv7SiS zBywH*1LTH^`^asVW$munO~ZPM(G&yZ7_S1*ZkS!Ekdp|;VGs&c4tw}8UN!xN38gWa7@51G7v@_TJ&KH*%(t%n0eVBlee}ZQ6BMMJd~4`YqI_$-3SiseHmKbEY~+d~ z8sc^9tk9)Uz9k_AG>rUwYxgMjlW$4l8P%cC-gZkyn`&Z!&$pfjc*I5&_Ys>+FgV|m zQksV$%D1+2dEjn!y%Q=oBirKT;P}+g?**Ut|5~?8n&^n*{$|nvs>^jyjWMO*$ayc%v_|ke+_R- zvt*d?KNFQG5>k+9!8(>+?5vWUTNf_uGuY@p~HwO z?wZ}2X4fpqIwRD=3@c+tk+8`O>)IZmHdNe4ZM*&KuUqXtTZSDK>zZXe3c!jV$`f|Y zGQ-A~x@bAEfHA7MFihajEW4tv*)5li!-y*GnvIyYM+!0tdlo<{qe79eNszk62Otd< z_d(ib8N8tEuGuZaf{Jy`axMiZT{Wv(VT&v`NCqY&G)N3nO?hEr-zk?)vpkHtX183@ z4I`?!Yql|W-y~y_^DH1#Ms^}$lL&R~4iFkD?jy8a8l}a3(>1$g$WF1Y*@!m*I5Evh z*hm{0G6Sj@mzv(f1pCZziMnRDT!IZFs<>;`Gizk)nSI~&OT%NEe*~t zhGy(F|Anj@M!K#@gZWQJH`>y^24z{-(g10p;y%(qX5rds{udi=hHk_<5daR;ir(1( z?#=2=!#6U3>r{qaP>-n_9y&TaIo(;VW5Q*E4uwD9 zy$|jl@9#c!F!`M8!&}{dAKmoK!O>H@2RkRHlk4_SH?n5>@z&WY##G8&*xi(H?*PU^q{I3*{%J!~aNfabItSS9|K!o&|^V5AN-SUn|Cq2E0U> zKc`+-%z5N?cV6vbo}H+L>eKa<6CcQ$Z_qT#&vc)qbLd#lF5HSVoJ~KW>^Wy%eXtf; zXM$RRoO#gFYhHb*{;r;VYxIsgfA}GF!|Cz!6K5i-gdUHof37qU#f6-rD_^cXe_*nX zRpDZ%VP>JC?pE{ktqKg)S4UdHv>^AX^N&qvi&Or+mVa4Kbz}Yed?;M0&AeJiT=^Z# z5h};tg?R|Y#yrkCy|D~{?c${)b*uUEd-o3xhUbgFk9-|QF#(JnZ&PajaBuR`&gsjf@7yImaIdS+ zGy@(Sot!K^KG{9oDgGPH%cb5`6;*5Jm>o^q z=Zn9Ol6D#;BwnOIxIzd!-M@Y?IVml0gBnzu175&8XraLm?QI*GIU_Pl(}r)`w$VU+ zJqm>&yI_+6W7QIagM>Spcv1W6Q$5!_F?puK1N~t*Pk#eL)j~I)Lu0DLAPSy*{#rHE zllDaQMvbJ`ZVziD&AsKq9c0gWZxcNa#+y@LH&CDE>1GC`scV4=59o4X8ch(FVTNrI-cx3fAw_#=#XmNd;3=> zCnKoPVtY-vaIRr9uYL?jeC=p&c;USVM@PqxAMK5hVhOqRX#FXmzHy-51f6IqLG7)U zpGv8;`c_V?6!f2`N+|WLl zs@a_hP5rN6F16NeDaefk+UaC6pJ8YDLhx(n)Y~{qKm79H?zR1^PoD0dUVV0Wi}SA| zKP|hV$j^c_S@OSI=083(3k!EK3(c=j4CgM+!E{Ndt8eBk^7!scZe6kcoEbs36JqS9 znGPb@u&U*vdSoTwNaoW~T2W-G4z5uDbGQa}mXkfxW;KI&AUNgU#v9N?pD{o2)yEjh5DPu;}c6isM#x#bj+fQn~~y z;J{@@04(WzxJ6MP3D$%~^`U^`(&M_q1GN~vx?C|ncJ$1XM^B$#+qFuo32D^J!J6D2 zTbDra=r@<6A01A1o;^BEcxt{el)4(EI-@|P{d-=zKAfk&h59AbGK2gjwd&1j{vhPa zaGk9)?5FC5{&4%Ya-pw2ssA^S&0aqlX7q>zMV;ze0Cv|EqbyAl{utQH59)!v@wv(5 z`p(||HQGJjX<((kC9MWYTh+S!93Rm)SO@q&LtalvZbcGhuGE-KEzqA(JsYRKLqFaG zS4zNQHTQNe`nKu$R61Lz;=BHrM+oV|o2-9HT8FeTQoK&xw6}Y@yYuwX@wMI4;ll9z z6Yu-bLty%zS9E!~p8+4=J$<~Hb;>^v7pngzk!EyzT{(Plvj5DpRCzq6vt%P8(bi6B zrG9e#=w!09Tb|p#DVWf2Q8!(_v7FH7FMM>NAbFeLd8+uGVlV11@k6m`&T1q0u$;X1 zj*qUB21`XWlo#Whp$l;bi@dn=Lq^<2ef3p?V6Ev4=*{(M*Nk4{NL?KMG@}7g?JTpT_93k15&5y(LaZ@OBZ7oc^@JSeOW6$K)9S_T_0^@RaCZ8Mm0O{E}z-~X=% zYR#Jpg2A?rk9PO2k}r!1p(x-#c>K)C@O9Du9ITsyuT={~fBuR87{&ZJz~de2(=XpL zyy=2Yc%K}2|Efc7z9?+^zzhbW*`-Pk{u15($>jzLqt`kSr3;OzVA3)T)XxcIqNH18 zUWxRhJO?%sC4q@j@j3N(F>1?j4VKDIE~WYzfm~#sLN1Y&N*zWUSt!4}44rRch$;CfdQj*csgdjGAcdOarley^ z1B}G8NIV4Ctvzic?=n({^w7fb)r?U5>5*ccbV2`V#05%H*xC-psHsUs{m{VPMuc+8 zR&^~^HS#elJV-SDg@kFtg;TR^YANUFjiEs*w&pZefdlow4b&wRG@)h`@p?&kyDjUJ zberYmBa=}}h25oTwLVM}Y{T_z<6!m0>dv@EwS*N0oJG)^_0II-_Ur3IviY~%TpW@O z=&v>lp?O6Wwnf4CL3V+53*9ghE7@4aE$1x0HCUv)M!lL=C3O|3THex|pNwE9r^owy zlf9jj13F70NHNRZ_N4b;xEPIntT>_^y~z%A=ckGj zrmN45-kkV>o#ph8j7KfvX7P4)%jDo-|N6=P$=Go+>qN&o;}__e9j$% zQ1~v-W0S+HM|;#1)MU<;NYuoStIyrm4ux}ywg`UQa^nX`%n!9uE1l7WC)I9_fv)=H zfqIm&?!oTMMF-b%zsO| z4@=TErrO1|Mbq^BDBSSDMgF(Xb((M*Uj_N3FQbq8i%NJ+kB&0`{O`NwdYZpX9iVLl zeSRPD8=VQnMT2zaqEo6~T;5Oy`|*2Wb>rH!8XuoikN08jX1!-BC4HG^DwC@h_UR1Y zHG^KGdfhmZ+W8Dx>`J8xeohf%xQiow@%m5kKR2hk0NWBU3T-aQcgf7wnAO=pzZTXw z%}ez3Vs4iy!_w_>ea_HDJwG$jW;k#6kesu$*#_z#_GW@+wbP$TN7IaUl-lo*t&`OE zbesCE016gEJz23hMCcR0P2_k*GOn6U$J8ZPN_|PURikAp!kXTuuE1^QW5pK#iE<-< zbggVnx<+j`qkbW343b2h`svjTUvGW7dk9tEEh%73#}ZJCKBTA2j|r1 z^do6x(bbbdOBZztHp`&izPA0+>z~irO>aa7dSZCK_Q_x_y#;tzn7%)ir*NaApPDy?WNMJXK(! z6f8Qma4XmT{L~pIwpDdl=dD_QyxX?9vE<`=&3bDgTh+T;P6^9GIF(oMR>!SL)M%9l z>c_fO*BbGg^=6k%a`QDZ#;8l#t@%(iJ{#@x7oF;3-4_pK3#j8&g2+&nlfaWu<6Q zn22kOus;2r-8W@6P$&)>=29eZVwZZ3GpZ-IryUn{yP93^&}I5s)O}62MV%|0S=6z2ZxGk0e%dB}YlrH%m}>-z z&H~d|%rzrcMaV8e=6=c?733}E{!zEpkP$|oHFnv8#W3AzRn&+Vbu^l`q!qZ1tD;80 zUROoUSG%r?%!kSwAR1`i^c3 ziblFft*Og`qWNz31%>%k5oa`~xdFzRPsgMBwt;%ASTj|dNa2F{T3wWVp}76?Fv^YO z6KET*>0I$VU;8Pak@r|kvx`~6wmz2T-uKlW?)hGR8t+aeFEKN_efdW!VA z$gYWbQ&%f~#hKUQjTni)KTsd&N%zP#KvUO0yqv=$YcNx)>RoT=lA%{XZ_nqP-akLz zTbgyQFq%RT^*Ysz`W3LcPig9I=8@}Ozw~RN^%76$Xz%=fk?v+jcG zITwuj_lzz|O@B>Nw7GaUl=3k-P~UrAUFj(foy)8IyR7=!DU?&#sITQwaZ^@blI0ee z;)`7O9NFI`#c?>d9pry{HAKVD&#WN;cGw?rwP;^W^E#^+)P6y8Z(%Nos2>TP%TkpGFG$Wqp-b z?>?az9`*iFa&$VxjLm*&3b97V##IN&bi6T|O{QvWBrvCS>AipJ?74!|)O#8Vb~e4o zQNb|j-G62M%HsgE0QkS10>Dd-!&JldnZ9#hbYVqg6um;ur$4Em&AIDCZxWbPUB{|T zckMs03IfseR(mdMCB^w{9HIX0p#MzXTuRWst3c|gMuB8CLVfc}F5ZTIcS zIsYN2pGDiqlFcIKe15~Z`CRom!^gXEZmc0VzKy-W=HFqWdGWnHIH4iryr<;~uuvpI z#%bUG5jVaL$N9ynnxEfG57htDkA+eZQS z^Pjp9jf~C_qRmwG4iCG8B%5CKbqF?A-Ibjzr#pv)Mj5D|nC~S;I>eg%>O?ePdYeKz zludlWDl(0|vzlr+^Rxe)qIzBXyB)$@7LsVDs!=2sWy^_AsHjPl z85bYKMXMg4Yf{g2=Y@8JnXfiKX4G$VIjb}=5s+fe<`ndYT}bbNw&p>-eAsl0OIjhJFG)ptbZB z485#Cs=K$-?$X72^)2)2HdDNNkZ$y#!bNq3R+lEu^5C*6t`_Hv(>j6bK$AAfLX!-% zO&0%rQ)c~udjIf=yfUjho0;px3Q!e9(;h`!}M zg_jH}_5A~Nee5ziW;I3kPvn@lP*HlzS-T^HFU>Qc5y%uB6IWd< zvX^A2|3TgsZU;TYUg`91m zf*g?(IwtxB25GXPbNXDwefH24_j-n3#GyMmJee%j55Z-rYEaTbRrS~B@|Clg9Ni2g zzHL+%`{K*TxbECP9;gR%XSzD($?MHsW7)s^AWzAxp4r35LCU`Ch|BLFY=i~v-Sc3vwKZ?kxvwwIxIlea8+uuE%EbWz_q1EcnN`64= zNO0_td_eJH`i1R(F91BJP8Ige0!JJij9o|VtT|45y8({+)q#2_h3ur7I-hk`^imuJ z5drEsrBNiUv-&Ifs>U8B$=_ShHO|S|G@^2W6k8Codis^Io7=pRA|btF!J-RKZtIg9 zc(xHn{aCWwHi$r!TpVD#HFHI-P3mVP={v;1l+PDML#0JpP3?{@DtaHRqpHo;sN;7? z)iiS-TRj@BW=dh!g79KF^;hS*g1b&>X0h1OP{l?mH>_-^8N-NHj|F3+NlS?mUt861FomCp66h}xM)wMQH9lumf14)K=20FwWsq|FxO@bpNEMbildt=qn zaMkvzr<{^BK@&V~J4WtGrJT&GwP(q}(946FuT*?N%t{5E!Gd1dbN(tu)SefK9ZRt2 z+~4O)#Tu((#NsnX~v)Lc1ZxGgnK{s!u?WgEEyqsz9%I@K_)wEIk=@B|{8eY(U8ogUwG$?*0 zO|y>7Tb2nL&!n|&CZXJfUQCyD4c|%V?6GC?ZvPIY2U`%6(7`<&e$s|F4s|F!)PhAp z-Dg7gy^&pda4o5$Brl7BouTOG!k8W13iJ%XJ5rxBKPnqBqOMfk>x82?=^lB0u4y z;Gz{tU)V8GkVeDz9S*CEYF z)!|1@EV-X}h1MaB5Hv64*GZa5t=ToSAZd^YOE0$oQc-Tqfq$ba-Cwx$vW8Bq+w7zF z%eu{oC!Le0Z9%FVGv)cNG*(t}w?`#=>$J_WV-+z3J1Ixw4d>}^BRRK0lfgTUTij0B zD3_d`i}#p11yd3FbtR>f*z}z=O*FEsZQM$KXagcTD(xgM<uth=Sk#^7_FfkDjjp*1d-DNQZcKyaujQR@HN!u0JRxfIF`j*0N+;jz`#X6h6knL( z4YgDm6ID(n)F0oTi_;$XR>5naXqGrSoSf_&>_0aVH#fFA=lZ(d#7nk>OOrHBbe-JP zSggKiDA|19!u;u%=42nMI>MW*7hVOF4lFvM+%7g-sMiq*^W@}Ua`kk$Q2jHjFy|v< z)_`n0-6oIzSjT13V@J;n=jm^w9<1L+?vD!|2tfY+$mIB_o=Hy1n*8p`&M9?x9O(xF z(0O<(GMwZI6``cbNky7_=pOjwom4WJ>RBcE#(?Z=OP(*%cLxc-E4y>zqFGu6b5L&Tc1Q)pNDGu6THDsA|U-oxT9^Bh|q~LwtWOb8|{3~+CM)CF?;TMZ$*qogZ=nD@7 z9IcAks&>0~vvUdhoYTGC3vNUBQ^UDC2kH+5_wq;LAXK$jZdCtu^gvv--BVijp6(wV zE*(98dj0v+rN?$(o*X}MbbM{eTn|+;Y>__$#UW*Ip=j{F7%F~FeOWjQtkkoB@2ur4 zB3EdlWIkIkVLOvsx8~F@2p+`{`K_23YJ$b%`hdi@E72F2NWS;Vl9|eS2RXqH=#ViQyi79y~u8-nxDBh56z6;oOD6 z<++b8*ylR)V23(T-`$s?T>;6^>8|#9cN>ZXoh#mrxSHBl1>TCyQH`L{l^S<=ntT+033V<WYRt~FnK&G-DSo!omFBjF^FxxVxJp5O0hekaTg@J@Prq#_tqJcdU z=a89S3V1Fnjwy|X2|m;P=a@K>na)ZluHigf4`nToHw8S67fFk}j@Hzw;SBslz&*2t zFwG;MF0d_5JZy#SwJ6u61_XrhfCu~H1aCOVB3lkH&kuW#Bbr%`$~oP3{Cq@<(lpUi zr$P9mbmN6%M>^3VKN;{)a}e*Q3B^w+LLO?4B{eFgz={6TjPlPXj@{Ih_NLu%g9;yx zGg9_@oF7%ZBbkG?ED}TEV`h2c`J)yXB)}_eu-^Z8Y!KHKc^a)JEQq@DkAvh_vBHo- zY$Jg~=P^aWwBsYZ4+F>;uSF{I7OXs2VI){FKbAiuy1<}5NQC71nr=XH!ylsks76yp%$a{X;C0wCV*nicm zr#`u^r{eKIRv6*>miNuwWbw&8-Y427_mI1LcLudhfYnNyxK|!L?ZM9nx6yIhxj2Q5xYQV{JnOw16^=;O7 zY2S6XcaS_`iEJC6StU-PRFG=ab1v}0nM^zLaq@A@^oMD1BhNQg>-|lABtj z^`_qBkI>@-SR2oJf{~@jxQ2MhW#k$SFZc?o zvq$j>%tNkfdZ0VvJ}NSgt8^GI!<6O+F;BNU7w?1$llNeqE10oV=^6X+7)8D86+dR7 zEGk1p=34D%z5Yq^3N4YPs1YeS?2`2Bu@YIGOqs~0&v8RbU17@APS*bo4?ofJ7^<+1NM|RmXMm z{N7!;RYwa|%-{4l{%`N<<$(X2TPZr^bzO3jW>c;kj?v-*9qM(j9}LR1=<>`o{ZOeH zoI6F4#OtH2!RXX*H>>D1y?m`+AT+Me-NhM7#6Fa+AP=`8naypOq6{|zw#Ub}2W2ag zZ%aN7UR?~?(Z|!Ci*7k}Ylfd{W~hpfOKJuE{bq$@XI>nZghT(`&(y*OS_Z) zcybYV(sY<}IzKEMB?n3tuH;y(oma0n)^$mg`nPz9r+xlOaC-b~9q?T=+kuCs;I4^qn( zsLd}E+O5T*aGi^hQqpfSTnK$o@J@6+JKld_w0Uk*&YAi@!BxJ$+d`IIbE0>7{X^NZ zQI{{*_^6>{!SZJ#gn6rVQ-%PUgE1K=V*IY>&H1RLaucZ07sM2&IZ8E zI@fl&wG`w{0YkZ>50KuWS4WPjM}7OiMd%mci)1oBH>BPUkQDt5>g;li$m~&>GT~>J zQ$zRsAfFEugcV^k3(#x#pX)=Z_~_{L>EU?TZsXuX1?eMXXD*T@Ucf%ZJnpXD%erK{ zJ)(`hI2)3nZF!gNONV=>G=#x}9rC^H$zu0Zf1}R_4wlcu&uwe(v}l?VK6fO{ZT^oA zd478=05o1T-mG;tw-Apxd&T@8zZ3mTtK5V;PSr&Ah_NT-vyxkXYm@DrV|7ciP4$i7 zFC=v>IxWb1IO4WlnQQuX`Wi`T~V5ujWS81_M}$Mq<(_v~=E zvpyJY!s+pvIUz6dH+hqw_~FTttk-(3g5KLyMGDiKVRk#M?OtT!SBf^XIx(mw=YCVD zK*I|_hy5+4Jo*()N3GR0F?@Pl8u@D8y*$o0g-QkLHWVnfyU6o;cVMB| zl#kgEZ0d2NYnz)(exOTkLlu#LxZRqZ=hDp?;B7fK*33fU&ak^UBQe=kwji&>mV^#z zLzlAu9iD*NqoT(YYIyO*AXI|9q=NzgRd!h6hB-uOMtjYMufb!b7Gswp_I0!?V*CF*#ksJOY3K&EClMYIYU;O zQMU<8o6SxZo2DPA_V4OuO>HoW+cxh1nsxy zQJ_JRpBmn58h(a*u3@gQ#7)sC>w~;qT6`F z>$-Z_B+fN@)_$)F?|>U=v%SKNrWGAnJS}SQcauAN&&{pD>&{coucXmG>>)5kh#wDl zda6-J9gMekU<~1sE}#n7Gd&9My7F*fA)c>#Lw|STcP2d7DafxB0f2I~8 z2-lpKTZn((c7e_)O(2DBwE3#3c=k16NhO(Ay>$7s&}htPYW0EKpGYzo3$MUn_#?Uh z*|){r;l^oxvvK_}hFLUpC*dt!!z03T45EjTz8w-@?ZZedB6bj~$x-rYjG^ zn*tBQul?}r2>jAPUiO0IJb7zW;$F>*qIboga=Tw++Xe=O%qRovn9Oydc8jRio)|cWnYac!7hM9TvaN;8 zE?}LcoYb$;!kcQbhFb>!p0M>N^elE@wGVSjs{iRZ%Q3#nyip;A^4ds!S*C8GKlr}& z2N>kL{=)_MgJ{$ji`O*F;&J3LD?Mfwf@_p@1iXL5L4j??#31ndQ!Mfpb=X>J;&lj- ztxh`4-T70;|6u(q-1$FTkfZt3fdK1SQhxf5 z0FB}Fy|IWvAsaMocQPIghJ*Fp4VZ*U$Kstj%V{+OAcKi3$UCFDCSk-$Rz8u(`Nyq3 zMvNCEUl-5Ggfg+92}ndoB9P8yHyfuwRQ3d&nqY^I>Y%d0W z6UMw4_)XY~aLJ_w5yRzxD|hc3kG9UP!z85L_09h7S$KtOjy4Z--Jm*YFGTF04&s<14`(_G(0Zsm|=6_ z3e?J~%a^E=faiMhJYrNQ3Ne6{s2ZR5G98qVk7U`O=*M6jq%tkdDA4fJj{?MpT(1H;A4=z8r<`0lJb;ylb zHdreZL_hJ&SY}qtE>)uhPhCgC58j-|w7(at3|iShpqQ$TdETaYDeF*BS%%ZWU^RS0 z;dVy;EvmK2(ggm=P1?^CTL;bP(q=c#=EG`I3MJBD!YfZ%71; zuH$qwPawKRRLYqEMd^=YbH!h%6fYgS^ohm~4#P?aE=aoIhliIJCnHEyj!hj? zy|7cE;N%|)@=U5ZVTeMeEw!tvf)+%tXe_r`=V~68QT|eFB;s5O?Z${ZRn09!tZEK< z;@w6q8MfF748f~AGriR-*|Uv=cTM5$Ho*I`BqFk!tW`j+(Z?%IF^pVtd0hqJ;W~Q> zga>^$_eGh!5_52~Jto)QET#j`!u43FKKYx1T%T#fJXy&iAGq1R_5OK5jz?~mbwN4Q zeVxTgE65ioP{3``eS`j_zkYgqyxE`h7JBn1Zo1TmpVXJqdV z&o3qv7U*TOSOw7ulSKpj=jjl9TGODv!s{Au??kJGAkCN}+DZs{k|}ff$E`lCWblp- zd3v&D@d?cNzNU<|V7I6ud}dJAPao?_mQy2D%wc$$r)^VdVek={D_KE6VbV#Ehd~6D ziMZRrQZ~2Bzya`^8!*xyxpL(SiLRzqURT1RfGFYR;Mgq+`f7y%*C^Uwc`VQOF>XaE zDqJ#SkEq;f_ueS*Ha4TGf^-cnWih(?SzNimiOl}1qD*@#YFfRT0ud|r6lvoT=nlV_ zV!R1^?GjwN-+_DgxAZRz$0r~sS(38=s3IPUt~+t9txuu#9~Y7Cq(}#pkPFem5mO+o zm|#yh(vZpjH9i;836r_dRs1-(XLSLNgPq<~-ec{lkc6y95YcMZBfi395bcrlf4(j- z3u#%;=Zfg8M-Tu%RP~MAgrM|3wZa?^d~fN3Q2+4EhQMWz*i1ugQj`pmSAeCWlREKv zqb7!LJupTmBz1zq?4(Yt1gwru>f}FWBh$?>H=JMO3-RYmK?41uhGTqdP-CNt;?W*p zIz0;Co-iKrBk{?r*eA$|?lVhVj8|oB;wnfmnuUtrvXvKIhaB>?_+pWuhzI?QR8_3S z7#yt@@}E~T$!*UA2qn;uK;$*?iMk;row^Mc8ydtWYJNsvlRHG~BK%?(OBR|JyMnQ5 zlqa$M+a^!ISg^SgcE?%(jDk0;GS7A^?w6y639!g52&t0KAoP}Z!j-;q#BJJl$DhEa zHX#0}BDZ2sxlKih_J@KF3ty=?1prIhy{m_y;!!VMtR-C1eM__{ybC_5VKRXQFUTUFJS1T#KgkEF{O0be( zGvJHTB!<`m%eNIRYWcQNn4I&w9=)1*uQr8(mbhc>)y5&nopZ~#7`&Al(_mKNV%7rf z2N)*{!SKIWWu|80(4pAncp%bkBM_!cX=z)BVneiCT%rtE^24q-E8a~}syMN_ddR6S#8hSb;-gT%_F&L7~lD!yw8 z^5c0o>WbCd0NKS+m-0@rDp*ACi0u$yUUZ*?4>v`s`0L2-KB5;Jq-F z5{Y?mwJK~~U=OZ?^o!lxkEK(fa(n|#i_|L~JPHOz;f~`aaf+~|HQhkJz#0srnbA1orBK6#TBR^KbhzH6$NUfJ4jrlXbnvJ z)xiVdZ@`i7&&vxHR)J#h!=$lUY;Ym!n{inlsA;191VO6HucGmQK}D(q0+MtYSIg&T zmajbJ+YESZ0xlS_31@=v>+;GKnuTIc0}Ym8)E>(~3;LoOf^* z`#R3E*69MmQ%mDKG=o8%>2GfK;qG^T;I|vT>4Lyu!JKlA*nr46QC0(X_})1(a6k_$ zb`2G~Pft9_+wv0#Mqne`MByLJHSzpncy6iZq~yONd1p`uo@d5iA-a9{{_&*I?09O7 zksut+DwiV*g(%Wte91nV*$qXlFuE$K&NCXlrkQgV7cRz-Afj-wN)It3wQdG3FC$x{ zP? zB6v{zXHG&Ps6$RD#Jc$jgTjJI8sInf%1GSn>;GR$ z0JYFJdf||_AJ;`;>c5b+-m}vRQNVF}8p5z7-d&9q>Tt8TRDfcrvs7 zH@(|bvD@a|(DywUZcO?Yao=_X8(zHsTgtRLv2<(l=bdzaTg_?|ZlW~Xjz zw;`p*LP#@IM3??OlG3~|o7ZW2&>D=+*GI0Sqw~mq04pTbGPpz}@UtB>TWHi?b9z9c zlvS_Vj$v^jOg{$UEyaE?rH^k`+>@h?VQ-%Pj+XUsp2b&p%_MT$3TS>qu5uDuuanWv z#&DPADje}fUV~R%ykUE5R~&F3%d2bqSIM3N22IY-%^f+a7T5|~<(SrlvFy5j_=&<0 z`y15##H+uV`%E>k>avV;Qtkan&36^V^9qzBrmfw7u0I|Qj*d>B9*&3Y7KeYbAa^8( zNf)8A(`v6{$07W_)FDG`QjI+m8>p@wiTWNkN>r5G-t6BNf-aXad=#T4Ex4iY9 zQfDF;rdtZi$g-k&iRb>(_(}q)XTS&jaizpAG`+P`~JT z?A>wDkzW&loM%O5fOH817$^P)h?)8Aa@SrB7H=hX$ro@qtOIb6Wab;A79{&M4e->e zIy)FnzFcI$%cLf=JVRDV9A9i#w?cq8-W$7>U>J0>%7An< z$b={9?B1v|8)c*yJks05a!b5HuYh0h>>DGTaL`Mu`X3@r@-sNP-Y5yjr!4y13H{O0 z65Vvdt4w;Kn7k7wKT%*V7z)K9E|}J{Xa}$O+Y{uKIJG$q+F~n!nfk+1kQV|k?_e;x z0c*hs_YBg4;h`!xi?^4-y*UgBXpY{0T0KT9cPfM&MulG*=jx6qQNfeNyG1$mc5++q zL|Qpj_hyn8^fPoUS2w_f-sgS9Q&iUmSG-C^e%kw7vWHyW>uwT!O%TDb`51a({)Fchb|bx%n@cR!=UJ>B2Nidp z*Zbe{b4Ogt0eh16}3u-uJAioyhtqpEqgukG=0r+X)bcnZg0-Ny={v zGKvHU&}+T#Oxrgn;1^GqUMGl7>baO{0egbWEZkLze{9r1zpGVmt}Q-_>^(ai?yL_+ zn*iCpW={AN@<^J8W7}$E$jJ<%;D{8}k&Y{XAsnMKmHmR=Cx03Q%RRjDp>(NK!>!#R ze<&_#tI|p$wi@u5v?Ni9w-sOI5Rp6v1y39e)BWHfwU*pE;4x`CQ)-}#-8AETM@`vqd!yHyb0h;Ct`_Q%Ya4gtLzWI;{1Cd`<4+KxV+r#yq0;Cps3Xp!6pN5F_Y??(H` zfTXbU6{$QqEPM`NFvI-?y-;~bn26t+=sRGUTNOv6%yocee&l|FwuT*otf|rucwd6` z=|ba(l*uT1P)b!R)!#@p(H}OqrObfX$owpbn1`0|$iTeO(gVp`RxQbI5P~VRg_OBn zmfKe__{^0U>`6G9E1h&52Zyl z^|P`hh^8{$yGsV$Q%EwZ%*Bsv3WdTtWXUikvMDR3M>chlYKW#kgjr%A9fy}9#{=^E2d-zY~w+9yu=s)qptlhdJBqO)zPjxa%4|NS+3+P%Dy}I%G!2+3I>A=L3 zX56Q*NTu+f`fZgE8PFX~-&vI&YQl_0ja_J)T;6I0U5!97~F^jf~=hv<+HoL68_@U2^fECdX(WkfM2GZI}(^6d!5|MIw-w zqA5C;o}#IX)G3-J&-e0=yg0p>T|odS0lYKOOhMFfMcqYf@_<0IEQR|)O zYzTE!K#VM-UfZ&i)G#F(kX!|oOXQXupgHnPx|*xa?&27@lPW(+Jv~m_Yea@}+x81$ zyzuJ-5CyKSf&-{La4y6AldWB37Wq%;~$+TJ{5L~%R77iUupD2UoQM2Knzew-uGL4)L$JyeiG>`UD4CU{r$>o-h#S1_s0!5wuFP=oQI)KKGK+c_DA3s}InCu*sT$qR#T| zD2ct;YieRI7jTsu{hs&bY5f)P`D5cDK%>Xf`&yh20NyXFo`O=w+v`xzg8XAz31ym? z-sLq_363hI5<2~T@Wo!n3&SRF4tNdJJp8syJqrb)MKpXqhsw-SN-f-3(8$h!j`2yiLi3y0(F+LP?EntXqEeF8T; zw)q1Ai&%%!kSA0EDfK#)sTgCc3KEzD@102%MpNxoyl)JIt+q5J3;*n(z;zKk3=Anh zn(tvC*+%s-SxGRpTO>X3o2;;IdaEfr>t04Z?OSfCo0ThOtN)I$)x*_R^I2A+=&z8Y zpAV_`QA8xoEHaCA=qdMbQju9KVLYcnk=b?x2s@u8LRke>NTp@5UV3R+F2Ta7a&0NB z{qmI#nI5U~9rci(6wPN6UuW%s<4~Hw>agS`hPAY$1)UUAe7>99)Pr~DPpuqDh%x$R z>GEOId#)z$Njvb=m9C1v%AIlj0(nE82^Yby7R;hyBjqI2ya_GtK{p62pB#)*Q z&U8!v!fvDR7s#grUd0F!9Q6z}!UEi2l}I!hTz*%EZonrb2mPx3s{wBl zCTxutu??HE%yDUa1eDw_8nIRK<5)6+eoo~`Suc)bvmVb&h*zxc-_z~Q_d0uumCl0; zYHO$Kn1))AcVTf_ClEO8zu4Z@PogeDJM`&@Q_|=&R8bIWOxA4uQ}4r24NWn*Nb&8e zqw2i_ldx!Rj}@j%J{ya@1_aGPborPYmOz==qhYU1y1~*W|V(OaqOnJ_AWPk{6X@Bl>@zfdmiEc zDW?9&HMmR5d^pZX+3#_FT=BjP8nk7Ro(Uf_%R|o}wZI^ed$qxO{|mA~Tvz02w4Sga z>Pqeol3&FNLkh8t3=W;gAU&i&q>x=@aZhZR@&U4MWiL(F4$@O3x7(?=I>}qG5@Cg+ z;}Tx>jOc@%`YI8U=WDtF$qjGFCZBJCL#*On;pow02IGEt1|?%wglaTp$eB4PF(8R_ zbw|i=2RyMG>wbjWm-hkD+08_Wq^JtJd5HGvrxbvei;}-K6jcw^MO8c|$TB0$xbwce zn=C)M$@@g}o8& zmyv64xXzFh7})_h20@n6*?1 z8vF4WeLi8Y_%RD@QCT7~*J?-Wbx^|G30f*k(IZlFyeb`0d3L6z~j#Kq#r;_ z`UI&hC4FLjyrfSq$!;F%At$)dO8ywD{x9i~c~OfhiqmWtL>pB!8$)Z@gUprmIKFT1>g9m%n_DS5I<(Jnv4Fg;OHR^k$~CboT3nz*z3%mcLAe%PpPAky zlxha&PEkbh`e~kE^>lJCTxy+mOuW_Mmh26b)!4B%#SB#uZY9+N2o9|76X@Zh&^!!zS%=(>&W~|_ z>&$RzchVnEI8SD0f{RVBI;ZW~c3rDYUJQmClRoE##jDR!H<~<}(u^fYKFU*sB&~^Ceg{}VP=+xR|G&yx& z?|}BlvL)BqLh1*pWee2imkI6G;z;j07eyI4GK5a|zY|^0j`trJZJyhd^QQh!aFy@x zwvc7loakL%|4_DU)MbmpyeNul_mLg)?k>rxA)3wg^v3q~crZF&A6}1+&bQ`k@PlWq zWtW_!ey@!aY@-^Vlj-Pp#;HeKs@HGZWT!6Rgcs+3r;V!O$$WisEZy$RL{Q`WEOvdMi z)Y}1)qQ60%U5*i%Jt|Ws{OqzCjpfDm2#p{gCC?mUu-(SN zhYHe1$j)3OOT2*g{+lE2Iws6fFfVyomu$C3w6PawLlU$t@3MXAaQ#htnf1XA`QG+q zv3sh&(dUB)%jen_Gy-oV^@o=Dz?#KGP~U z;hIx5nLT3mN%^eg*5BG>d*@i)l5A6bBZ3qpbuBtAinOk$rnhSoWsA#$l5nROkmixE zx2mxKh^Q{XwJo)B)-g4L$66EMQDpDg;c#btFxrHy`kFa9Ca1+P**FEZsCg8h)uNnP z1yZtI2W@$M2>FjzT?a${%t+PmZ!caO(?|6Vr~{Af1|E{X$r}g74^Nh4z1DLT^xmc_ zQkdQ>vfJQthh8QjUn$zm>cp&?oC{8&0*!#Szr7&0BiP2~u)oEWC-jnL9Ng`zc{lSo z-xMkpsM}CL-R>gK>)n9`byGf;2Xp&@F1Za=L;~V=YjU1TH)nvi+cA4R1CLKf^uOFjrXOrf3v)hD&7yIm1oEsY95#VsSW`Bk$>u zM{v69QlAK{*MO1D$go_J`7P|zW2149lbA{P7;UlD!;CrHsYRpY#xpe&(_2y`pUfLQ&`fI- z;25qlaxS)GmrWFyWU0U9@;dB#EE-r0UP{CjLjx#=o3=2J7_MiWdOOiC%};ilkN7Ot z^Cofb(X;-0Re1;CNSpl?ZZ)ka0X31Zm%0|DHJuiP`@6}Vz31i@?sbnV$6k_O33!95 zQAiz(w|C$jsU=-NmUoAi%fo@Cc)sin{oRS=gkKXsQIMOXn+W`4`W`sm<0`ox+?R*v zsg}#-$X&p~F`Iw^N8ab_oPwYfx6`5ep59)L%xSuvYz;nPSL81Xa%((qR4b^?4AvKG5rS~d`Glu#WsXnBlqQhE zHrjmORJ;J2u%wdCt6sW%T4*$8G+E!u4SWw*e_mF9V3+(vlEGMb1qQ?4%l*&3E$$9C zPV>u+@VYo2clGU39ywz)drfyi|3{QZbJB3$T#y^W4M*RY0pMn@FU~K13!gmnjjLrn zd1(Hdt~>}AnjVB-`{CCS_@!Ps>5{ibC0-`-8regFG%PCNfAU)fUK42*8u>|$M|mDg z9=&oEyvM_CQ5NvSe)x4n$ZQIy2`AZzLJ^~50?oVKo$cM>`r%53H*cLEjwiPb$D8LS zeLA46cU9xB%IkvEURir-2NvXHyB>O-C-0APYvxUeFZcGIJ zc5=QHi!KGAvWM51s1mTtSSKk<*-ct_SuNIZ>j1zLw*G{k#SX0YVN^-=KRstT##fm) zDx^?e8>uhL)b#sxv2UN7nqY1_ouFDSj*$cV^(_1S_HEw$Ef#@I4H2qm>2|p ze~Lxkq7GY2O}q{vvQ=dg!-G(cQ6@>J9;{LHI`(R$_t$yETo{K3ra6D=_#do)>CXS* zf*j4K4g^?_ld=aOPQwwPL!-gbZl5P_EMkDj1`XStj7NjvV10K3hGEjlc&E;CTFn^9 zVB!k$&Zx#o7;%!7Pvo)uajTCJQv}J^#e8|^AdrvL;@Lj0L| z^4(cFsL8muf2|Vv8kKm7V0$qLoG|9aAaKG~gnKS6h?uI_&0B*ZM2l;?Jeqbvs#osb zHy&-BU561#yX%|%-LvpO*BtFa-a9-q+TwpxEIyhX7;bEgc6QbO8~o+g@C8%<_6o9R zxHaJaa?artI>`FGehAquz)0tahLumgoV+$1!;iN^sc;P;=I@YnL4IML+=AyCfQ7nq zKnZ=AhR1~+GjA?jfm&I0+3F6A#&7fyS?)mkCDmIIc(EtXBSv+i5Cd3=s_~0iZ%ac7 z`AC-iiGB>mK`PVIi~{A_Q!EA_d)y6~*LE^9>P9ssTLKtKkmQqIHV)<*`MFN+K}f#3 zfphB2^oK@$QdRa(8>k^ z$5eI9^ESmxS%-qkGMo;Es|mIv!7=i0QLRmuCh$*g(tf7cI%r0hHoI{)A6Ao6D2edA zyhcT`DzY`QwNWp`sK_o?6R|H73wlXV>J2;N!~DxTY24Em5c&Go_6qK(-!&D5Cpx z4JlO3R33x%9%(oq$^S<1;h+#aYx%aIWNU5v^klRDKo%3E9SU(rZ2PC;n`KQ2q#uDA zYcUqa036M_YK#x+DJThu^-z3j)Yzcm+>_<N6r@J`rdxGt;Kx)AYqB3(Lrl#z!R1 zggPP9nu+#Dfa69b0ctLmu~TF<93Ea?oQxn*IX30%ymudje<1%*kY`fO2}2YzZK+*V z6|^9FMPs?mI#=_!jPjRaBN2`c+KsVj8EwlDtD1A3c(?Hzct^3fdL?_gk?@A-BY~qE z0N$4+5s}qotpak5K3@5`y*;Vedmq2)Q_7vnrnY_a0 zB}*$AyrV;&o~&7X2y?!#DPt{IaKR|52%i~tMuzvglI7G$6>}I~=4o35W6Y;!0t%B( zf;is}%-ZqiBES zu^i3EOdtE)WxAq5DRba$Y(`ZD=^9$fVs!PhxN?CLnf+Hqnf6rFw0bq=ZSf@p8;?MD z_{9_xLjOCK;MV;PT)e-fe_=R2u|3{gk{A1_9v+G=J8`A0525uRW!iHU*(W+b>*<$N zQc@vWH)8tJYAWFqsNn-Ec$;WPYX7n}b_BH9U3C>H-`l>0hb+pp>LX5Mhj@ zSV@og3X?&!L(>2G`XR1YX-UuLis+cO&TZ8%O((NM*Dg(6K9<`(6d~@Pjz3E( zAF4Pj{z4TvhG&TGBMC}42DHInr?FdQEmVd4NPHCJ!LNC=RF7ni+d>`3ldiO%O#o96 z%WjRRMwS{|f{}b@n-qU6W6fS8iyZ+4>-B)_9~YE^DX6Ves{>pDn1Y(DhPcu(1@($j zV5xefax7hBe-_`OyZ_v9d?Bz!m%2Ho)HT=T<{nG-4f>P*`e`VO^(VcB-u#K1?zmym z@j~4Kfel8CVPZl2>@akDsdY|Wt(+V1?LGx0SV^!M@I_@3Lu`T7+lm&odMggijl$%d z-v>3P=M;1*O{~4zI0U(KX7v_>wfZL0OSVF{2#o%V)n#fW4i$=Bj>jS0Gy*}&lonNm zkTgf~2UI9FLYw)MuRu7A6Lb-Ia$Rx5(g*?!uj1;y$B0>z!G*4~ipVdE@l-?|Z8BOfR{Gd_GD zQ#L^hhp?B@u#G1gYNldksCh~Y@8IR{_JQGnQs;y|WDC(AiNUGL|b^T1V|a3eT0_ zFg8LBxGJ$5EXDn@UfB9ZG4Y)j?hM%l6e4it8ABG06QKKEWZ0%@%K}a1B(wBe_0H4L zOFKCMLF&A0!5H+z{oG3!tBm@gc38Yl*x99|kr6|p;+^M)zGd3sOxo`k!xx6^vcrBFtLRS1y_Al}i zG|BdLr$O&Dx9m#u5h>s1_$Wi4F9j_It!GLe0sG@Dx#ia{d%YQblj8c zYGCGeWeEzli!2klo#%_!+1ALjd$;;uXLDLGWvxm#{SGb3b^IJAZ^$Yh(J!-ZnpQlr z7ir>>)gQ}tu97yHs^*ur@`^`-XGW)DOHnh!res8J*pw^D<6~NS`q8XYv0&qIo}4y{ z15Y*r>I|lIroXw_hpXKEf!|_;5t9%3=G_^-B*U(tUnG|q$6&D-Fr&<^;LNF;x9{FRo;2GQPZmV_%>?1_ zRe2hj7Gj8oaV7Wg#RxN62&9Crda>l6vu+BnRb(bws*Qjai2H;XEUYp?3?t@wpohO@PHL%zh8h;C^oA@9TPEP8kuauj*eIzUOk+kr>U&cYpWXd@y zkRZ+|3MAKIH`IKtOFB8-x&LBT;Z_nMa|dh%%P8Lp@{9@w8H8o1 zT3BX5-kg;@2;vNdq>SW2ESsG?$d&4gU!otNsgR9{^hp*sV%HGnd8PGye?01Mof<}! zJmMzQ{cS?Lw2(J?;gEL^Z5nltC_|kjxA$&KDnnVm&g{R(QAB%JFK08*ijLN*BwzA@ zXM?2ydeQ4%KNyrZxsX*7YBWuDbC6qnw4GkF))Nqf3n|5&V0W=)agZ;gaeH%n(A#t4 z#`gC3`1Zi_pk0gzQF&f_hSvG&5-#}{-xY{p_SB~(ZqqF*3ph~!i)hZBf5kUlF?i?j zfae+ZB;I&3tKBy}*A&RY;!+9rt_Q=7N&n*QI&f6w|LkH#RVM4+Ww%q2d+rOS=ntlY z!6lCg+F?)~E&prT`pG}~p0f;Q{*9xJjU|S!=US}Q7A$tub}U?M8hAlo-i@zKn`1uR zbZ?&i?wN)X%*NXLd!Bd5j<|(JD~lsL^=#Gw}q7mL3V{jS7P1B%t)#&=&KSe$3jLoBg!D%QnJ>^tOZ2@MFd zS)HcmsKMxbeWE%#I*)7p?is7G{BJ!nJIq};AM!!pl;7ZOHOib8YZKijp45PICK%v!hSVct?-8J ztzB_;Rq|i@@c0$$)lkm$U0O5S?ahT{y?-pqQY6-ZeXW1C-4+%2$&?%gk1JKVzB{pb4Q;o#`#^y%Sv*lw8@UfzCA zYI5tMQ?^K5UNi!qNB*iq2FR2e`yRIDX)EF^TVg&}#56n{yx!n`IxJp!cTe><`r|sI zEl=E&DoMTb!Cz~O@m+IA)a14x@W_<=)$91p0=B2RZ`}?Tu|gf^o@iWoVbC<&!Dbeq z{^^z5drRL<^l~Mxp|UpF-Z@sE+tD>Hf@hHJVEYb8BsioFTO6Rfo=3yu?Eys09a2cP z^BH_4gKlqiue%jVlrsrT5iHayw2vZt&klz>>x0oIB%asIQRoqN*lCUL?oercWe~NF{43yBOqMlBTyOvzfm_cD+PY;scNEnR44+0iyYb{lAsOUPyZYN=z zS6y?L9zchf_|Dtj%G6%YdUvJGL}1HstdO zcS|^j6w(w7&7A@(a~nC)yCn-NqkJI+q8ec>6O7{rp=Q}fo1Qeo>lTIxls znq>jdkS}E*f~4axU}D;oWwTu#h=;-4t%ilM6(45H1sxwlKGMaAh#Ceax!h~i6)eY* zUJRKgtHQmsDK)!`*Dy8o_0;HVSIy8Rkrmtg?mNEw7j-?$=;~*rfqkrg1F@iFt)-K}c9Ao&|v(7UVG;7{jr+ zyn;Xuw`sZ@-$MR6qs=J2X#>B~Chf^?=PMZy^u}H#ZUx(#(g7F^A_1?`rX)htm|n9; z?(WJB@us?hQ$e$8gy>Nlz{ThGdr@?}QQM8qP4rn4`U9Qmpe(5mTKwT*cKWGUZxIPz zKSemqDu@&WRpMwAM575Pc-yfm9Wi<7M7Sk*~six9iQ%@rNNW_VlM+s&4q;x18AeKxtZcRZ=2syjJ31&e-# z&cN!D_ZH+k!}>3S@%9d2v6ghETV}XlS6<_)mh!c3=f6-&?~C+U#@s`MEX@SI0rtmaBW6^_wh zZ0y;fQCW;H6SPca6IYTnsM8-7WQDHgzueJa5xBw1A``YR-9u@5v{szD`A4}=i-4WL z+ij^k6KILl1%Nf7wG+NWx8^wk4O|04uqFSVZkCNaIF&mT%aEWmQt1OyIH4$7ri*zoLe8R)J<2^sU}h7R2Gyd zY~+TJcESPU@udN7ra|IYIfz7t92R~C2AKyt33`Ju;#Hul=?hqS`xK|1yk&r0UI4wD z;YKaBl|aKzM&!kvOp|V$e)70P&p?U#q#Br&E0)9%I-Gf51?log5*!JbV_bS5b-$|p za97|z>RwHt`;)g<Zh5ya0W zi@~I&B+PkNZeJ^|dJ+%ExZJdqWSWjkB_&blGCF%(p*2CFluuQ#l>~*FUIPO2a^tr6 zLGpx^1HFBF9^wBfrvAu%SC_z8@r3B1Qo*lt`5JYBCYutl6>^L+SmP-uS>I&?JfmV<7hHvuzq#&k#RgT`pQWoExA%oH>1R0ej_{x;mk(0DQrVxk| z$dm<>1DU!8RUY*~pzD#JEij)c!SkK$Z{1hLG5QlL3GU%PmERUx3ZVbO3m0^2jSz}- zEB-yF1#g_y4-Xx4<(Qj+u*$3w}6E2+RwNZ|F=+tM0cj4Gp}x1}SAQAut;wVo+g zkZP3N+VHjW40!-mW<~=zktz{MJQ&t~M3`xKDstQQ3u654kNJE5u6=^pS3WnBS>BP) zE+UM&DVdOzgut8Ei4Kv;zmbCE@Q22)tJMtD&@$r z((HueHAzMSF51w*un(-9i$XG_Rc3UWWcrcbR!K`Z1RX|QMoc)whpOsFf*R>7T%3eM z&;ULlx$+_WhfFtuu@pdn99ACEItKMjFnY01nD-~A*V`hy>Dt$W&s2d(1%)do#AKBY z>hbRlk8X@pDREr=rpjx}?kQ|9XJ_!aUKf5VgFy&29L)=0G~{sV-ee}?FBibmEje05 zHNMVHa##h+WNjLDAG2xbZRSuba`kM^rDW?u(niM~kgd6DhJuR(u*^BG93yWRJ59{n z<-d284gEIqvZS_@`1-N&5D?1aG;gazFAF*WD@QCfExpU@AL4yeLP4|1387|wbtP}| zKXU2vhud4j-Sv&p*`a=g&@~AmB7r17j76RIy7g4twH-q#a3e=Xf?xHd9^b9@2;O@i z4G3GSx^2zRoQ8*@{1y!(vahEnDaVaM7cFbUjp3>+k8y^VuZ7B3>WN>fagO3 z3DrD_CAT%c7&0AeU?{q8o>Le}wJY&W+7DZ7X-W{@={teTR&J#^n#=1=ggTN>cE}Cc zZUypbOvjRw=u&gV(Exv(mt0H(F9?rHk0i^OM|kt)Z?jg3UHCUl{w@FLb-3DR?m1++ z#86xRmIZebFM@GE`L#rxc3C3dS=S6pBT5V)MVQLs;TP5J8lvK4q1M#d68z zVYz}E7--=U`G*c^3?_L>JtZeW<(afaR}rf{l9m0{l8UgL#WgVo=Vo%O_pFdHI0jdx z^Ma~tHU8#3NynMGZdLbAl4T(aW^TjngkC9m^eeuA2mwheP3`>SbzS{MtogdRchRE9nLOhB3QUP5#M0pWYzj z=1ys9QT=WCOykD-6G(0+qwTGw>Xo^rTlyD<;}amAB{@2YA%4pyvvkV6ccT|7acVr$ zFI$rr863bt0s@6>z>8Hsg7SE$RGG0h6s`ywfPzjahx|f8eDv8`HXbU zfGPjB{)~c_$K^b^3tK{+r`n{YdGevqL`DA@_PtPYqF-{F9*5n4H-HRw?VehWV zC;YF|#Yx?8WJTwk7m6N7GIQ}N+cd)n$m_7;GR^DMc?MI z@ceKrrBOKqO1z(DlXr}7?3TFVW?c`5n^!14NZIXiekk$HEe_hS$O46Hnc>Oi5AeAI z0kMjTEwJ7DHfj*Z6;T?kMg|y1ednM%qC$rSMLQH)8_60rgJ~+$if51v)kBYWivF9C z5?+OiV7&a`{hhG|1oc549B`e*u^ zn|*i>xF1z(|6Qc=rPz~5t5M1gVp(8SiBkO(D;hOiy1?|5`lE>14)bY^gc_H%G@V9E z(WMiXtkk2gH<|u4c1!@>;i+^+m-hDE`^S@J3H50K863|;2<9=i;LwDYS=aA*CC~>I+ZeIRY#}WCzu+BuA_CO z1!YFJ(rT(IK8={F>T7}#XH}Cwjg;0&s(7^&<)CVE5;*CdAlRrCu2ADHlnMtyu+lq0 zX*=niSRO6ClPkJz|9SoAA0n0VB{?z&YEL;)`s9LSqjEwfWC895CP5!TN`eHLh5{f; z5+pu=m;}i+YizWQmpmC?)Q%-2;9irX#KkM_=lkPPf9urH5ZMziNKPZZ$b(tOOADc+ z7Y=y_a-XRE-9zr~-I-bYvwV+vkmfVy9UX*c5g~b~OV()WuEYg-Lkm*b+#b~Logd#G zlx+z8{|(u^qC@V$wwSuh!u^JJB5>%*47K%f9O`>ub)SIx76ti%4!H}P8{_`gnc>pz zq(7c;j?B(1o1RpE+{%_%w8)9UaAVTvoX~W`;?-xU7fp5*+ZIe@6ODM$V)rqPHeYu<7i`J>3o0V+$E7;6=C_+hq^g<4Rgjq z8Bt$bZz~d&ElTfVBwNj^^Nz4eK08mIkFC8&FKqQUN2k^%qsgiJdIz*WmhHGp{3H)@ zNPJqeD}Qz?xfrS@Kq7Vx>yVdp$n&r@?0Ely(dM~LIm>Bg?tamb1%$cB_jd7<>6#C{ z%j+M?c5S(APmnXr;~dH=jEKCwOESuWW-~p#vAsPWjLz3*)1#yFZFv~s;F z*W5iGs0s@DYcDJ|&Ux)A9R*XhCu3Gr4B$CWU8f){Uv1N4FsROKlzgYP=xodr)%he2 zh_o9YNWPdgY>6L`EbhA2a21W-q873{neCZtjT_LFchOhRkiG;SwQkSA#Xy*h%SNHX z6kGCcp1iFf&(61s&0&AbfK0O=@#dz{XRfX-`PZ!bahz)kkqRm;2thCq(6u|sb9>K0 zg03k9*(K4=%nFYj^xVtzT?%Qkr1$h~=6m;|%wt&5F z)dbWN72TwWKnHVXEGIf@6alWM?5~3LazxQ=l+6MlQ|k<_5sU?Sq`)Z5N-h{sX0zc$ zvsoI5Fl&bc8oJB|D7#pk(NQ+!Z@aj$TSJv0l>Lp`h3}j&3RcskR!Fkh$l2~&J1Sm{ zKz%i1o=G#vHW6gA*>BV0b-lJ?O5fJa8PaHGK#d%0BckL?pD`0#;r(VDa1GBk*)_Hx zzc!D7<&pNNLBpnzVyH(K=Kf05WQ-!jP-(2tVW_z_bpx|jsUTF?f1otRr6x5HVVAhO zdlSmFm&X1j5)1*VC0Ez%6a%y*mllGXwVK-fT!&s2L%=P_O&XA2EqD&pIL8rR4mQnL zV&pfF4~LBoKt={8-{qi9vU(0M2Pd^>i(CZJIIoXqUG;BfFbc*CRSh`{TQSFgrbz?R zB{)uAi(FmRa9v8}(hA@*;u(${CGcP%stg*KF4Po-;ki(~VTI)q?JjTkoP4a?eBtUl z>cNdDSJzqNysDuC)T2eL4>y-K*2HiG#>QCz=^Jg;|0W^rBjvT`OQ`I!G{ zrbb-s4dnLDDsL-=1R&W#PKSAO?c1h z0$daB^rlLiS2tcyTOji*@P8w}oG@{Ns}&VBqsG{jwt*Ls#8PHysLDjHG-+T&c!MDQ|BzScHQ z3i-UFn4m1pljg*YY5EZ0PkaJiGz>B%9&*g@G&Yxjr@Mzs^>dYZT#)uBx!~XTPj@gP zm7E}zFGw(@lzHA^#-w7EY;7->HIdhL$ahnHi$Q-kNUM!Vg(KzjjRDLMUgrC$nlRTC@dgUs3%=93A2DpG9_QS6uQtp~Rq=(e-+7BaR0*$xbo$cM> z`r%4`H*cLEjwiPb$D8LSeLC)}cU9xB$|r~DyR!7s4y?zCcAXUG$@^klnRyH1!@a#H z`S!f89=pZEJsCQ8sqUYo+@aS|Ifsfxmm*LdsFmRefd|GS?X~v!JTdD}=t1kiY9D5S zRR7a+mLnXLM*`(ioTe-Irv%$5r{?JW=j-F#z0M%!wm5pv?(M0k zf3W@)?(N@PkQ=gz1L4&VF#E|n0wji$_l6<@9c)mp-N|?~7!KBVH(;71oiTUnEGO08 z)T7OY%c{v!+W0%2Zt{+p-bWa1l95m36g+PAF=7B8`T8UEf2IIMj=j)qBfPT^NaiwU zqM1<<)V*h(+?=C-nt*%z*D86hQE`{~Ru=)53FBM@SSD;iI56W&7xe$i-TTI)t+VSe zWN3GNv%h;59>SWVNU3{=XGUB6Z;Hi7lLNzzjnU4o`hSDJ+#0@M>fc^L_6)ZM{9n#F z5Z7#$?af<*A%qHFH8E$?H4hSvjUx zs}ck|@#a=u^2;tVV}wev$Ei!pp$sHZeQp*SwVFUxa1(v*tS#{Z)wLn@7EXnZfeEg> zkbjN=k~sDaI?<&)YLuM^RhszDH_vS(xRb2$%msm7>%8$X7`PSXfO&#vTGcGh*DX*QG+9bjVrI`ZoD9GjzP$n32o%~ zk!Pa@_b6m4L-(XF65E_g%)*rlrpe?|0=mHSlWz@4s@AqoPd57xT!jO!L&ZECa%JWik49WA zrcB5JFCz~{#~;ptIJ(KCABUFvrs6$Rc1_8~i;EO~k^EgjzANrf zUc09%MnSNO#a5fOt>$kT;~`Ug8#D?dp%>bI!L*uFlX#!-8+eDJw|XVCGv`(a?_@qQ z)hx6b55Vtoqyf}bJ~w9(5NYh;$wg~TqnuZHisCH zT8buru^>nCESF~`Im81eTesOiE66QzJ7u^AjytcjX&dquB`yeunuNOjv4VVO+y=!f z6I!%%)ouKs6^g6TKr8f~SEO#}bjebH{p$sMm42E2(!TEKozIiEc4AdRP-e^oZH0j> zv6Pw2<4Tp*>H3KdSxHqY9!IhrOc`moZYf2$$)LQRyf`|H2m%`kVFt6yEH#VZiaD1j zpeo4}$WB(ivV`BTU`eEqh0&zpaQDp(2xyO7xpIX>K{NPgn5w+I6B}*?d9|v5Hx!Ml zEOul0mFa6_?y5-9V?}QI8->G0W=>T^t}&m~EiF|)i<=K9Vc2^$lj%)GORLvIAU;JN z5N!+st>KqLO!rQowgflZci?jSE&U6_@rmv6=90WSPc`sRY#oVft3KL<$D(2p>w-vp z(nBjjOm(!R$DUfGfrY<6I_c5rl1Yz-wzJ+62Zq+bHgUf^A-xea7=0m9dLur<1Q6?n z^nVR~jyf8pr8l1~V$&P(efXg&TBPP@{M6@rd4zcz_}bD1zSjs`4!w~PnNx@zib6o@ z0LxyJ)Iis{USdSTQ<;R+O%Rxyx`~B=&9SMQ{I_gaxH%zZh}y6ifKw==@Y{5=2BdnIkF2r!q1(6$BWqA-nMav-Bg=Z9x!^@9Ej7 zSc)+>Rw3lSu10oCh0ri|)}G6&q7!jL);V4v>E$~1P1Q`-&7a97(@8{9di1Cuz z907(zHe8zw0UJSPNzhGZ!4h?Q`pq(_9>Hk3Uy8Y_&tkNoqslKE*dG(L$~PuTUpN9- z^^WMn)fD;Sk1DJx)?KjcoE1B2#~o(rz%KJ6ZA&*D8=9v+ztqlGDNM36+)qjv>p~#|&WIEfT z_+uGq>>5$*Y$t`K2W4+(IivqyD~@)!*N)Kk_J(&ZTzdb%rm7*T$nuVm4Yv>R7peV! zm0juL`qSv%+WqH-;|qbkH3Z{GX{ZTy5ZGS$JR=sd&kjR3msqjX6>8D(q}+nohlL3V z;Jr$E%>+;s84*MlR)wu7QLC_x%HmwWa5(KB-P@Sj5NW11_CS7|S%pR5t5gSV(-}j0 zU}j#V7E|+ZXiwy7Jn-l?5r!D1l&IPxN+}KIq^VItlf%rMeBD9eD!%qS+?=Mod~H2V z*)WqfGP$ueL}G!BHj|F+inm?cXfG>jkKBjDfRsx}RQyRAML_-BstPC~G7?J@3>kd0&VN=r%5LFe| zwM6*wyh(J$>V7dQrSsrIC}gIfgC*RlPy|)L$a)q|dQPrUPCnHk&!vYL>)^>S>WtM^ zTlWi1NF^d;7nd8Lg@fb?D+hY}_DD48)E~KWb_u+;<94%QY=pvYRWmnON^4>YY!SH< zX4*5ukVWIPhs0FR@UB4Nu+EW-4O}0l03PI}zIsz20kWrJ(2qOYGGl0(h#b!wH=N2u zBO``H#cR+go`iBWVuu9k#BOw5m^4KW9$JP9TNZeO>J}sGP=b+}agSy}>P-tC_v{r! z3TV}0n*+$&p|^rFgg8x|jLUZV;vVh(%?`;hXobG0$eiKBl}}bE6;$?SN9PPrq3rB9 zPslSiC-O|;09w!G;aWG&{zZO*vmCIqY*EWyHNJz+KqbA{JF}d)8E|e!$s?URNWP(@@1Ur1^ptq z#8?cNQD#)k1W`ua49gvz=_h$hcD}%Dr;%1-FpnmcXwEP>rTi3*M+Qyc31##}p4)ft zA5WTXi>Pu@{^6@~-m>_IAsWV&+{2gMdj|upUFB4IKQiN>Eo|733u3Ua$^=`8B2(pK zmmgUmKbVy?l-NXm8Q#XAjIDBRG{XUYF z`bb)M-!J2!WisWQ6i5(fl(EYxh1S;cQXoMeW(p)0%}jyhIlWQ$%9-e?K&5{&E$R^z!2tnrS*J&JnCI?7YX(LVH&)XOqYZ_T0>1 za$#r4o|~~L>z6v7mx_X=0eaEvUOyO=H@T2i5&$*5o#vRe`2NE5nzf#QAY`GWZBaO? zTja}V+}_+C^!D7ivAsP$zC9@2FL%U<@c%xZ!+hl9^t>1YG_&YY0i8E2M&__pIxk|%4FTU>~<=0&x187 z`h)3UaLHqWUc7kWp`%uJ$v^t$mIgEb#?i*c5(C#?@>=}zZhXJV?5pXfd-L@764(2C z?#U)dWp=3vJXd{9AF-tK;x0=O&C>zHrIAsTXyzQp$SIz02z# zJnwo1TZ3+5*aC6^TxX|wUrT5}n9b@mJx2{j=j#*I(b0Ka>qm+}wT@iW6r~2_cgY7k zST@ioqh>U5L~*KKv=w{eDn8vV^Hx=VCZY9i*4UGyjbU$|{*IO2aE3+KY0U(3lcJ9+ zpM-|!WVEvZ9W>?T<01eX^=hIXc3n`+6|r{#CN4=*=z8&&?eT{y?-pqQY6-Xsk#x|+%UbH27XF=T0AUQ(V+WqJHQ|iv<9-r@C+54i~XP z9p|3cn+eB9vmI<^0qUQE?e8spH_^+LxQ5EwWP9gWUGPTNxbO>^1hO4$-$9X8^_6~X zae(fY*lU7_{Yj2gvd`cv8FYKA8y&4kqMS)!3SeQ_M1L4LfF1K2XBTW0qHqvw*|Cz8 zP`fFZtk)@$KgG9e21;mCQBNaBI%5W<-dA_POWi0X@*4?*G5A5i5|?W&1=lIwqd?aw ziu7qMtJAM1`OeuwwaI4+^sp^P07QfQLxm*B_aacq>kD!$RkJsT{jC;pG@nnnTf#Y{ zkfva0Y^U}+1-9lka-w%j7PdzDauBYD{7Hwn-!FF@W#q>McIIeYsz+Ba;AWz)Vwf4f z*d}M~%?ok4w@dy4g}XXd21i_;p=3c)U!%CB5I(di2=ezuIs!~;Fv}@q@uN-6$;QQe zDFYECcWl7Kw7KhKJPhV;H7v}f;g5BE4EabGBO+=TnB;P=QCF}WM|v@2nyd=kz2FMI)!1p{t2^X>&hKyVNz(T7Ua$!VM1dW^0^FG6+;EMD4}nrwo`0t58Qg zpY668#%P}@0^9P6`E=BRWS7=upT@NyB<2;S1tDRfcoqbDDl7{^c1cZiHfkXTF}VEB zUuU!#r8YG1E0?6(`AP-^y|Gt`TSS~$Ho3KTJgJ7NdppT9@?PJQm7-cU7;o>u$b}`= z=dFs<(3iE$a9=L3aaBwCS~v7}C(<|i+gtDRJqPTemN%#y!;qHCCxy@pdygR(<~z9G z7B`Yhn+8iPo#$X{EMS%QHDwPyvWA#GV*(!I4vf`YNvOgxo){Z@HfU59<2}!4dfCL4 zBn|5HhrZX8?cWy`nXrB770Pc7F@^-F&O^Rel$}7r#ulDarUX7VVBM(mH4#hGg=#2DLT;-=zT>th#-BZ5| zGUYDa^!&Hfrs3A^kUx`_vQKFn5s{mCpGuYlD#5klshrM{#hlP7ona~uC<YaRMV{au*fa12s|-o91& zhvZGv+>}=PSh8=>pY+#HZ;vKmgy-m{ z7mlaCj`cp-)r;DrX6%_e@^arp*i*zbSpka8U$rxDX0$aKjyH#cQGYUA8ZaH9snKCP z9nkjWeaY@>I`Mm<$&5Mj7rxhFpSEP%%+$dmR4o`cnmjF|IJxZg?N8(i4d4` z>%)~=biJm_m|NOTHSUKigIM5sYA zMTf>(ctoJXnfFzYE{`O^k$^eIr3X^?tJ;r0hzO3*?oZxcS?XS)(eoB!O!LC`6O01R zNWL@g)lR#R-1<>}_s313;mnlBwd&-A5{Lu1Bue0#Atk0H%z0OCUn{Q8IwhH=<5EdU z6uR0@Nt91ju$2UbnqDJ!XfKQvPlz5W75qAvuTd9hvMB*uA;%~~;aZZ)IEevUAqpd4 zE6XJYY;^^8lfV+0CAW0{^ZukDrhZix_pp@3cW20;bP7R6Wl^t8Z&y?W#PWHF%tKt~_iIoKR@SnarCLq}+C6z*} zg4ozZ+Zj3E1Owp7M{r*boDHNNNEQ>5kHbUoIeTTXuQG$2Wf;p+^KrYAMoc?imE7r5 zjRN#SK)-VMRizv`R+^n~ye7#=z(pGx7+95uG^B@4lO_GIJ~;^Jp?a+Ap4A0-9I4Yw zO*jM{MyHOLaEK38t&0RT(pSiBT%;u&f(G!N$(0ZIx0OE1bR*bj0|84HWNh*0=%Ahn zMlbdW^Zw-YdL=h5Y=O);arnV!5`Zfw#AKBY>hbRlN1ZTErNnXdn<}p@?G6gt%h?%x zuGfX%%3u&e4M+0=7!5g`y0@5!4Nte^Xc5);ddOuCt6-U|O~dYEHVwVa9BM_bp3S+G zY+Xp&=-30;5Qa>LCzEqrIY!b7-rhK|>eMRr9;WM5BDQgJmr zp~q{(jo~S{6H@(mN{vk{by4h4$T-FhKb)5h-LieA*M;B8e8$I6{VxBQe3#a}a5&zs z-N!CNH}rScCvdr9n~%v~#BwA7&xZsOs(F%}SOY`Ree;~cNUB|lZ_<9)VoOtk@J`g)0mDWC()(milYHBCJP7n+{Ahr^9XOg{B72%n0@|R{yq;^ z`^;uniJ`VJdrt#yh^U!YC>G1m)9c}sLa|uEcm{()anKpdR%uBDZU~}8^(m`7ES5_y z56cy5U|>+gL}qFK&>@Y%B!}xMISDGyq+QkLjzhfxsXdaF{ne6+u$;v;G5zLda;*2P zkm)xDSEciUvYd_KA@4~#&eV0Qy1^}PW9wJ5o3a$5hJ#J)ZR@Ma7n6?V%}cT3nsJlB z@^C186zejz;r?(z{>eX|-XP=VPHAdU{cZV7-ZC z75F7NF^M5w%Ox{_ z`UC&N5J7UIHZKuvRp!v(;*-l~q-zFzA%bh#Hg( zkcW}v105MdH}mnlA$Y~=erO2J_d0uul@81qPX;5>MXfANLTwIqJdb9t|2b zXx~A=>CNl&o%$qLMIdtHN?ttZge-mv`4(jRw@*Idf1NH)>V_jLI_JDl^f;24iwCx8 z_TcB8U^7}=rg@z@Z^0jz;~w=5cz!sDbJKLi7Rg1 zVNHW|Lje!P2PwNf&XHHXhcAOREV5PMT4s2%`2&3JKtQabVhe2dzKt5haYd9ys}Y{h zm1uY#oP-VwigqZpHj*`L2Gdlg70)0Us>dMTo0H_tNC~gPMKE4|@cz!&0)qM=501BM z8o%U~H@TxP`Ju~8UaG@hnsL57jgc5D@-$j-3u-f_Q2*fm~z0& zv9W|t1?1T4dJE()4RzA%>N+VJ3uNgKSm1k5HkJiKcG6DIh(^cr$z8qYg)};XzpCzq zevvjm0t&w5?MS6QYG~A8LosP_3Hw_R>#CAUnXc!TB(X|;f{#Xb)oQ6vd>XOTr>`lG zVcw#VP~)1x#_&u#b@#^r1JwalPVWuKOUnm3hWg>=HMhM3&dqv?OTN=Eqc%TSSc$- zeMqVC8gpP1y5oRJ)7+N44k=~Tsgx;e^IEOmKet}UMe9ln%8YKM)l^k{8ZlMX*OW$_ zma2XlDXo)K@oFi`LDl3WaMC+Luu&^qp~jun2?s&2(mO$EJL#QR9xc6-E4ps~dHv@f zB9-zbIWh-oPdQQg)Uch%J_qnG@?h5S(n9Fyg+rc!*!XszC?DQK?(W^0Sw6JT!8}Wk zBlPyJUJeMoxs{@$B^TtOE?J|gljF8ww5&jXdfn>>gYx#+!prnlpawHIcZ!03*GF4} z(W&8XPPb}$^;)w)U@nXz4_DpAnTd_Op#`aIZV&4B&W~>o$~J`Fl5A$i_8=`Fujr6F zuq~$UvT(mKyB$A{`-cOo`viiwC^R48Vhj0!4!H}P8{_`gnc>pzq(7d_o+C|0b(*n9>UyCmx(9zU#EnF@@D)098kcAa?g@t@}o;)90dyih&>TixttxZOgQ}^`_Xn!o*ahLc>9^{bt zv}RZS>{fCyR82rO>>Ab~FX@oyVQbj&{sW`UbDMIO)6CrcVl!#d-Jtb0`Fp$g$#l(! z-sSZVWxKXqwkODop}sDyeS4Q=lm*RZdU|7fdpsDOug|7ON9Wt}FzB(e)~`!VO25;> z0ku(BM`<03LSwyr%a+OoHQmhh*4clng*xHM(Z;YhPk%4b6_c~GHDjgu)JcH@WytlfXE4~5^OqtmB{Zmt>x(l(Fr12E^UXps6{=O4AnO1N`W4Kfd(GEyE*ZQiX-ws(%z zEyWh~HiEa1zaf~9#ZDm&;*i)@ZJ}&$bx;rP2n+u$aq_h`RTXdy)m_q-C05Qjghnt} ztJyz_>^(ai?yL_+n~-QZ-tUdS+NQ@~P@UN* z`A%!m*_bB=4W5pS;Ln20m8B z!em@F3KgaoYc5zs-d2!j=Uc_*u)k$MrrD2p6mX^{)$Vy?|C)6_j&n^RQbDB+1p4ew z^4#8YkkDt!w}Oyo)|SAFy5v?&`3N-HiPVIaZq0y1%eXpb7LcQlUvT~HF3v=J*K|Y`0Wxs9W{yo*HiXa!FoBOXeJ5;=3YM@mx?rdq`)Z5 zN-h{sX0zc$vspS77;Lzzb~vD+%PtlN0^7xZ+r^FD8mbJT>~GXAeCLEwu$m^dLXxSo z;b)@@u6{GrS2HFmHG^yuK{lH`HZ49--{02F8PaHGK#d%0BckLCNHP;!;r(VEjN{N^ zzc!D7<&pNNLBpnzVyH(K=Kf05WQ-!jP-(2tVW_z_bpx|jsUTF?f1otRr6x5HVVAhO zdlSmFm&X1j5)1*VC0Ez%6a%y*mllGX&1uhZ`sfHlp;yHaa0_yi2IN-@o&z<`H4tC3 zF(_A#^c%>B!$t=nBLkD~a?mDOJqMVBliIUIuCmcMua9S4^>1b{3dRdn4LJ*2F~>Y? z*K3iht7y0`crXxE1`SLXYKp?}T&UhK8q4*uZu5n!r__TR(dPHaVTUbZeYm-_A%E#> zhN7%~7kOUqj?A*U?m8u#==a!jSOIUkX?lUk3}*WG66)YflJ*3tq;_*V@KOA)j{? z6O^TS(wzM9Q8G+J2Zg+77-UF1TRtoatDTZQU6c`GBCidGs8{8dkoL=Jgbr``M>b0cSXq)b{O*`BydrQ{v zhVC4BQ$bdP4MyLV0fT0*7S1nz3rAsm<7!#Y`j;5EYIyC3 zkuibB+wRWx?r{BZCBK`u&JV|v+lJ%KbCW(DchK z%x!V>UYfV3p8mo5m+tM~T#y^Gi38!)4>0GEcLYe($$LW)fetn(*Y0FI8Vm>PyBjb~ zlFpbrb=FTRy)AA+FHX_+j+ov@7;Tb~PvjIlZuK!@QXl#HBlUl#07Z7T7c|=l?<@q8 zxy+epW>f@q@0lkz=jfj%;NJeVO5ST!+-1JiMSx|(I2Qqy30n{j%=pp;{l9YezVT@5 z>^ck?+Fjr5@1BKcspe>}@ZRB>(H8%kV)4=Bz;I(@w6m-J-{3E|hA)`s}T zmvc@K0)(BLw+2H9J95+n#9>TLUO%L}1G6UUChwrRpkt83J5D~~a_Xc*aF{sUhN<0E zh_fIMw?ZhMOZf0S`5rWv01VWf0!Fk$G&wHlh~aJFy3@v@%U0)LbOlC_k>?!5RSnfk z(06RI9AQ*12`^-on5w?G+(_&Xk|(Sj=a9R-wQWu6Z-Y7(>( z_;$l93QkpeeTO_N$MkAdf?y}!+{#OS*+piIP$@a3b=`vbM%#0ddZ5(=s)C#7b7yUd z7pSfcskd+{bPP;z<%Rrn43LDS1pJGew2dbA4LZ?T*PV6(VVP&6~+ilv6ouR{VupU&Zke5f{|Vv^Cf(#q27us#SDa<&gzeIFV%7 zICY#lWXj(5uol7`N4t6QvMgJa^vTm)I&kV2Vun*T`~suJ7}RXADHOp(;2X>&^T`)k z3GTl{_uUj%CLC(zPBYjdeRJuV5wkND$J66IJ$$;*m-Jp-&lwSqi`g<#sM1H z>>3=At-6vi$+b7fAY{yhHgf#Pvr&V46f%{ed(sz0a7!e+S_(&CnoKSwpbI=d`PPu6 zYHj=UWV8Q34hN+52|mTtn~L{P*(I%9^Im?86ZyM>d>8eXL4ZG|D7Aa4ViW|c zDpknUwwk|XjE9uF4$c;6dnFQjq3st;t2s4^_X)p&cPM(RSF(o*i7ZHJ9bc>Q0Q@dT z8bDp;b8{8}k;WdL{45GBg%mD-tHAHzmCKwxirKMPYMVn0NG&`#3!NulEXa{O%jH=~ z4)MUr)@}CB3UW)_PT7lE9Cu!4(>CNSN?d@6)B6VfNq_zH_IR^D=`Hl;Puz6J4Xy{J z7o;eEtRUYRw?XmBgcdEaO9F7uD^fRfx@0N9P8l5&|6u(~carvSKTqD;iB$5@s0R=l{n;er;({7R=c2MSgL+xDz3 zz_AhjDj~fQG#Gs$QhFmk!UPcOhV*~EF7O3uSHnMB#HKgm`|v|mv`Ed*c(KCw@(A-b z@U^82(tO-bw!WRgoGjePh|DR(4n-j#bpco@HgyvpH)>#bQWlovC8TbGz}(bLECg(h zP2J?bWfR4dBC)C@pNl>^3JT~CHEv=tq%Tng>{zRx&Vo`b;fJI1Q?W@95#3~tq!^#d z$lO#AV6+CvO)d5FK<$TF`jP3jAc)8J^lVft#h4qb5b|GFBfG7fFxv)sRdgb5$U3L4 zVMPW4u>~Fof*?Zz?E<0C1tAfa)8HZ|t%&d&R}7}l@21J)E@en$!?no}un}aI1l?p7 zEJ!y)MCl7h0IS{+eYlz;U;I&pRmHjscAc|g zNA36{J)oe%!c(WntuQJnzUo^{pUoq@1%a7)ky=d6 z!=XKqtMS02+e8>*m{Ou@kES#jHT?J{hnYF~x`V=1eC>JIFirWX^)zL}Oxnog#?}yt z1vc7DI<_xj7Ot?R#o5kEo)5tGeZn^k*V#PUq~*c-0LTmG4XHlWXsf<;CDZ)*L!k(&fRR-%9IWA>k9WyS=q1LA zLk*+Om~FLHzt9e;Kx8Zzw6v-iOxtmr*)TRjVzLv0TUpyR0fUKz)Rpm zTn*SEzajT!bgC2aH8GA)^8lepi3n{Ev>dC`R7}YYzq|WfRs^gF3IlhP zw*W+r7B5Q=4;A8!-UKl~gyzxia<}39CGJ)sR2WwoUEy$(9Z!y!DMT_y85Ml$Ewsqz zyUn?3wjncI~qD1b5}p9$3~I(V!b5q2-XT+BpUCok&X=AVh?6%}=h1eyV7T4YuW z){;l|BuyhAb+m}-5`b)Ho~EQthN?m7D+gU2l9xQ9ioobH8kH83N8*EJfQrb{ycc7i z%DNScU9Iupw2>TmPGRaQX3&pwgRQLrJhUD}QQc!=l_|b?cG58G`f|nYD!IbLMp#ONi;f;Rv6uOuh!ifhH zt%PvDnH9q6w8;>z`Gif&axPmRzh#DQf*zxLY$0?LAIuKj^p$cmi;rc6K9Uw*_sh6w znN&F^1QNs<&p6+_=Y~LnKFknEESebt$#q=M&3vYfd@?I>D~XV?1Ga!=q;CazMgfBe zLb$_}H(L~=nGrk);&6ipv21qmAXloMa|v$gVYY>L1ZFe2rvn(=JkPXV98AW8?K2|- zUdEpCh#M&Wy##-0A#eQ35w9NJEpkYw$*KPBNjW6TSD2pnIU;EH(W_Y}vQAfb$hOn? zZr3xZDEKr$F8aNjhQs1!7qUtMpz7Ocwpoj>FVxqppMNA?4zHhQruwcL} z%yVqe++@aguRYYUSeLf-X}M|k76|sPhojBefbY6x17X-u?Yu3VCV$?;@~Z5t`&XT! zNA4L|lOjKu1_qZrcF?VhtP4k^WUk1+`TCXyGyjF-&CL}CuIEyGh3amHysj5tZ!+6z zy6OHR{k_1o{+@TT38!*JgeG(d;+PI6+~D1p$$J+)qyK{Udi>J%U~7D4eKv+s!To*O zA3dYOg1^Dc7FHkx(G?b5iFqTtRL?D1r5`GbQplI?za?W)yKma;jRi6@B6dP=}^_v#NiPPm#|m#a!=m%F zW&*iN(YWF?6hvp^^P8ioI{DbyMY8HG@g}rf;jx|Vskpm7nPgBjgUu{J{Zp|0J%#Tjda)W;P+6btoIg<|ywN2tf@cuz zVA~Fgtg4>$V;={|Zh^ff9B~BDoO~T$$e`L=zUXM>B#OB>rik#v2@avqBW$1FxVvDB z5QT$apB+m{Nv`T;d%rFG8opjLP(qu6dKTH!86zkR?CAmWg@nc!{32lS%axRZ=@f(1 z(PfGvecF!dqA%QE@1VPF(E=bE8b;yZS$=(_bwp+x}d^_Q73FnYPnu4LZ zQ($OrCnx*2W?^WQFXZ88{<=%t^UEDq8A*h|&K!>m_2>!)+)VUS3^U`y3R5i}3;8Ar zcXg}`&T)B~k_DamI>jYb0d;S=MLxO?lKgWg9RVgKn8g$_`O)TPvT-rr%|Han9UCw) zZSL9tbx{+RJ!`@ebGI56X1==(#KAn)!|;e21|~V(Ym^l%!;zj0sgqUWUR>!)1((vM znIe#Xi$+d2T~`zDaz#q&4-#H*m^WMDT#`YcQXncT9^YlaOjwCJa(SQmxW6X0<@Jl{ zs0GO`t;$}r-_L-c zH})!VE6ETOxu&K~=>QD5pg55Eh?>KL5_o2jT-}u$;ze}@r-G){2;PI^Qyh$#%Sn(A zqsVw8w;LUs$k)lVaG}rFEU+~uJZ+@E@+R*=2}Be`3W6wcGzy~91XN<|SQYXcD48}i z6bj6tLEv%&fr8-L8TtglDwW6+Z#9A&ln!~Tp*j6?tCm!mT^D4(CiDz97j(SS6dYN; zOC(HRL~iS^B_&LC6DAo!ze2&mJ-=F!5Bo+|ifq|%vU48B8?UfBZ<(Bix~!##`*Vxy zUD;B;*0I5KCVAo&DDn~CIA9O8yg=P(hVJ9xujWER1&;B=SlF{cqp}$98KdcCQ(o(@ulS}Z+rKX~ zGGY7DfE3RNQHBJl&Lh4l%J!WJw8UyyQLQ08z?#t9N#oxdB`?AfxUmoalVsJHN;Kl& zRPInLLxRjmTP3*>i&+41PyI5SDL3h+@!wLKM%&X7ACs1{PpKPnpMH1c9hEEzRGe$c zQ`wy(i#efPI>SUhyZ?rFRI;5Tm8ipc3!p?QXBea<+s0c^=u$v5MIc}5c*c9lt*g>n zxxVpW3jG^SKxok4b5#Bzc@q^kC9zK?dxnGAVB_r0WNR?%FZCBs-h9U~Z<{gqu;$v9 z9J9(o*Qx6XkI_vNj%#06=aUJjq&BFGPI2>3x7)7|?<%L;z0NlXyGBft6(HDrs-1Z= z@Ib!zt>g0WzCUvCW8=&2>d-9%f z!&E_3I)LN;Sy_8)b3q(MvMzx>w?15{Mb<02jIpKdR^uLA8N>q5sJ);y;4#Oc#v7~y z2U7Q|>W??exL+PQRDbgJ%0l-Njh?pBy~DuNTRj=aTRJE`Cy1$5hOHmnociYIsvl?wix%U7riRM`}Vt&n3pL!m=fucX2WpFnos zHyWMa7>>7~JHYh|i4I#K3d3P5%OyH&bp`gqfV<+FNwOq($VZc$nEF+j+`|$U-=vMy~Tt&nL*jPMJ?gkcp-~TI#1} zZHiHz^n3#6HV;w~VjV@s*3$C{6(L3`4a*zYVwc45MA;Tbc%m$q7@nvrq=YB%s3`H0 zCzT;7pKuWJ2`Ks#qY~*E1*sBrxA}(!N&(5lgg{I=aHCp4Am;U}VX8yYGYS>AgwK~% z*gRoN4#X6$v5-Lekor?}FZ0Ml$@yprfZ&CN^pNC{WW|+K;0Giyef4&_zN1IG!LiZP zVN@9<+7^x^dL_C2)O@B~0mLY`wBak~x!46pqzpt7FNSqB4G%?b-+4uh-F+2KK)-VMRfZfnTACejJe;H?V4@8L46MvUT4qM4NroTIl#8ANu_kLsTEHRbFdB8l zfJ1z!EL|jsk-mbAThj^`ko!pR5rPJ=0m+#UVL!yJARJ2p1gu<=zQt=`XyPzI>BT;w z-=FMWZ|cqY;PX`hwfE)(xN;7^MQh1m@5VTl;>Y#9+Z?ty0|KzUoSnhv`aSrq2nHck za5OJ~(GbI_dlNhGyKXwVB}a;=#8*u&b65q-WNjLDAG2wwZRSubV)bm!r9|uEq>Y9> zf(`Bv1x6<4m~xD`UF%JKLk_#^(6Gk^Y3xQF3y0JYT3oVaXe@sPjIz z4)w=i+;YxH@T=<7<9pQ}!F%t}fUvb{ef}E_bNs?Zphfmg^d=RT!xM76KH40efj1%L zf2ZWw#8Q_<55*b#Sta%G6M50ltvl!XJ@~E2XM7CRAMm%ycS+qVN0Xh(d`ua-VKCj8 z!Q+Y@-X=fBSU7R>0pmjg3H4f%P+tR0(S7}#!bqx2iLcXs*kUWSgYb6W30$@`D+$&_ zs>tWMghi!Sl0%qRc=O|LvsA_G^S|@2UWdzl z=H5dN6&PY0v-dRMhB)KAM6pjXON&z6WG|I#J(#w3StsJi4Nh&(&(GGDv~`39u&NLKV$os{1wO5Iu)jdE@w zC;Bf4X_R9yRXQ)|>a0Ny`B2hzrpjAYwG;UoTYs9}JWIK&_}A9^$#;^j<;_R2;+b)i zz~XQyd=%?3wc#Oghy1(0KfU3Mn>(eUMftbI^NkxD%pka(jd!+JptI3=Xl``t;L>Pv z5}v@X$mU86@me&QB~LT%0re7nKy7$LzCa!_7=VM70jPif&~sC;R!ttsUCg5;RoAQf zOXSP`!4N^`Ms;2y+RDhG&c(wAi=RuE4ETa%pD(YrL>~9g-%HpS&qo`!X6}j7_y!1g z%r+iHIv=Qn7^<011-_S85klYJ6&8iO-i<s!yNBq_ zQ43D@XppF3`vxKaUAz=1Sm3#FPhJdkLOy=_{!Sjh-*bP~Kb*i43fbqjYoU7A`A3+3+AC9FoDuzJu_jNXT z+xW(AsV2B-C@V@r%aGhf|6ND{ufj!8UVia@UTgwE z9ms=Y?V8H(&$9^eq?7#6&r5bVz+RehzBr2!7%S&#)L;mWUHvcw2mQ<@U-VCojWs&L z(95z1Hm^aI1187D5?*u2vCs9E$ln?Aq=%|JDH;o8>JV7q8z>vg0&#XyhR%sX$IHoG z{g;LmI)cBd?uCAl7C%B#vF}An^-)Em3aew%><*eKNB4r;0RckyzZzN_86|1pM|J-0}YXI|r2T{59FCnEb z#hyf1jgoE<%L1zklz4zp>EfEt&zG@M2Y(WMa<45?*5pVr8J z?jw6vchl_W0a}Y8U&ZzbAUiyi&ZyF!n%+N|HB+dE{IQwf1nM=m8}T9(m%(M^ixx~uy|ZUfUF;(6xKBl53`6 zqXXVGIkI6q<9=~484tG4jG{by+>FHsv#ysGLdUNh@d)HzktV#GysUp`W}49QJ!V1r z9ErF4=+%J4TUhOMbr7CK3&vi&L?Qm--eV2Q$2IhJt=K z#@oa3nb9<-T2)`Yk}NndZeV-M^E)y-1foOU-omMD?F_4V=W9E|qK!ju>1-T~x(w$= zUf(5mU~5cOWtqR-cnY^0`;P7r2;NRdw;hfHAivfncVTm5GT1&hTA9uUliB<^(zK)t zIyY>Vv^`D?N1L<3mB5LbT-Z4A;3Dxor;YY0OeXqQouZ;qv?+1z7_qiUW}yI!yK_tMpVz%+ww$3i>Eh7ac#bs zcla_9OptFZl9yvk@9|69gRSwI_1SoK=HC83?T6+J-fNHGZ~IAR(I3muHWKS7r9+Wutd?(C zQ~5#FL5#liJNdH~@`R_yo1^|B{k=dJOwP{Mlyw;btJ47gnT^kHj$q#04L!+Q?^6Ww z4u9XP==-78SW0<*Z0%OKZ0aj}nkie;f63y|=d z9@%bfu{9cOw}?#pudMfRoNEe^3Mx&zI$LV^%R>QOyOX@6|6(NQ znnI8bwniT6k=roON1)M8rUtZ>Y6B82Tu<3w3G3yEqWNqv8)^2V4n|^Da>0Nyn-3?NSdu`5Svwri z&}Em41A%Co@AYtfw}vW1Nc$V53*R_lB&_PBmPj(Sh01(n!EX4vbUtrJN2R8dZ6e6# zv&W{!2g>__P+|XzQW%$#)Ifw?X}3My43c-EJbP*EUn0Q}pjvWu zO;0gEOLA#(aPt;Zmp6u|VhFeexk&@^s|Cw}>gPD(%fY5OpBVW9^5w8m0mz8J1-KDlxDaB(Osr_NVBHki&N z$GpNwKGPvL%a6u=OWy&9Xc+czFnZmXEkO4m-!wPpA<*rIw&INIe(?HC=1R&W#8Im) z?QgWFTOy0A-TuOIv*i)_lZ1{NT&(C(Gir=QX&ZPE=~#*^300ZM#VV~&5e#DU&DQn{ zn$Yzu=16l@+!&`^3BWm+mz}?FfjY`x3RoDQ2Md#AE+8Gh@pr2Y5qq|2d)ws$$E=c{8T<|;F-|u30D%n9Q{~$q? zQu=v^DN}ZjsJC>)*xyc36bqw3QTUkHpY?2TI@&zD!u9Jgf;v=dLG-$32Bnkf`JL%# zj?Z(A9?i1qcH)NiMon!vh!WMkXxj|JK$P+SrcJSknqzRx0ffAC$3%x zeYFqJ2HPF@VK4kTLNBS|xgSP`l%B)o`Q5U8aWt9TKALP@m44u1tys{8T<{{2RGovD?`_Lk}B}e|W1Mcr#FVB0O9(R$i zbrE2hFwRAQWx^JO3p4)dg8W}Qy>~L+zHbAX3{5w-2GjdsEY-sD6Un~O=H~ePRQ*eU zPu(**H{Ry|w+{WkZI512`?uGSU8C(G|357_WVXv4c8?m0TegQI@Kn?fV@mSILERjf zC0SQ_2hAMViCIJ|z8wCt+hp2M=pd*^Mh09JG zi?ZlZ=%m!t%WN{}9Gfgd7*$Kcb6EwZsOPn}fQ0b#bM#IWRS+&qc^YOs%jMuJjkB1- zCB8!*>E?ybjXX&%_Yne~+KbX~mhp=t`FJ<;wj(PFDA&q78)%dyC?$km{sKkmZC&z$ z9M!8t37nmnxs@0EvWLtVu2QH=O|n86NTTZ8%r$B`fimGH>fBjd;t8r7Lh>z~3WkA6 z(4+r|0g^cO4KmTAHENWd2W6W0&bKZI*nKD5_-5r-E6iylnTn8KJrt2&1y35D^JZr! z(y1IaEB->wuj1HnPhgnxuf0wGrI295HsOOGT6ea z0$`MVSP_jG94wXKOg<4^!D#F}HM(ysM4eGMl3n8fc$8AxjS?L0V*;bHcf}xN%!D># z{K&gegL@P*l_7i57D+R$p*g~tNT8ccE+rrfJU)3wNKm!Db9S~hcp!%ZQu`zaf;EbK zFuK@fIzhT;jDuwCi$EWmWz?ul*>I^hMaMu5P8n`BxdI1Vhm3hR4oibbl$DLQvv<-QS;urEzw{Px{=f!PMOqtNU zr5~6=u`~K5+>Q`gB=F8FQZ{tDWGcY%6$gH^1JqKA=`W3H@;lvFkr0#_BSBkYAnRC) zOy+T`koI=!ce`XYRj9Zh$u=-Wq~f|yDcokbyoZFhkyk~#5kX)hAk1KPktJqaOzIJO zxGMosNv1$nvhsx`{DlR(vWT-Vsx(~geyV}*?2~I(uaR(PnxG97mDhD+-K`*xUnP{n zRTjIk>`I)#bV}ebhr>ojPNhe#FrU;ct&~5Dmk%gz*!wh-$xTU%`aDE?{XkzhECjqp z)cz5DGSLy5Nc~gf1<}SJ@S*&1i0R$wvsU0``+0cWe(T`UXmS$X{H(}l_EZ87#^#Z@ zwCb%*k{*HR|44sTLgIrSnh9c}(`wN3tIc;J!mJH^ZRHZD zJEleqFES!y3b8|x2uNK3R*DVX#K(;icm;%Rg23F+O)La#jt$-9zh%wB#cf#6+l7a2 z--^CE3JT~C6>efVq%Ki9>{zX@M?n#Sz?nnd7#*LAO~Mh;P3D{w<5LkCn+gJq+5oYs zB?TB5jY2Q|$Z%T_#Ql4EG%A*2%#9TY`LD~C-BwQU^W!2j>_;yT0lXCg{Lwk3WZQn zu@?xeEC^jy9}ha`$S2AN%ag8*mursaBKf`OTcTuYwFU7P>RBV0gV+*~pku?vTc?Vv zNIC78g_Mwoqg@{Nbj>>>R#_S9|B<}bXsZ4o@ywA)70GnAMe)ZXQrR`4*xgRBU3b3z ztVl-vzY-knat|6moLxFUV&jtU~un2sWveIBsKt~=!GM4ttiW^ztsoK$Y_zF#Y+b~B zxRR^$?UlS35Q=@)*9_NLJldk^!Ri9Y6XuOmeWp=Yed|i5`2mW5==0#dk13m=nLxXqD1^aEV@)LLy|= zBE=mFMNkEdta{;K-J~Axl9$j+jJ5G(7&OoK+TB$TUM`(#hKAW!T@*M*&?h`~eCP+`jgZ&1ym zH&T$S(`u;{@t1(Sa7TFyK;&rgvh?s!A zah1^(4ma8H5_t|$5@~5Hs`9{SY-k;w<}Xn0A)r#6aP&?ZusS56xup@ zQU5moC^WBxs5_(Ip(D_`1eyV7T4YuW){;l|Bu!kf3Xtv0)0DKyP&Gfbm6trCioobH z8kH83N8*F*kaFFo%2F<_Y00BcW!;L!uGV;P+DH!ESeUws85HH*U~6jt53L7Lc~|Jl zmvO5X^o!sUV=-Vxk&!VIL>XlhBd~mTcD%p@un|^bFpmb6XvUE1M*Y>3FVZBB<5*;n z1RhXEKjb+zy?-)mmMw0KaA35j778zA@(*J<#+BU5m)&{?hlGyH*AjvEkr@|lp~Hq) z5QBwfB#7vkZJOfb(XVGk%uK*bFU)>C0i~e+Ztc(blPMHrxhWsZT$d&QcaxLb#w`em@qY!WIKL7q{-Ac7FM#FRH%6r-6DJP6`&g9ou}cJLrqs-ANRZt22i z;T@gLolD@VL~c(+IJjVvjO=f)e7K75ddNnPb4!%I& z^>DN~8}MDXeU#1J1vZlW^B$I0WoO;L>J&Y4&%l}#`N1?Wxa6^eZe4!uhN7!pZOFg* z`j!SW|Apht%@qc&zv89%b-nm{li60&P4^e+?**>)_q>x$IF-4%!ey9gIHq$4Hv(Q* zcF21dJ){4E_j>%&_F!v#W_>n>QNjIv+8;fm!h*lS%obK41kn{1U5R-kyHvMsp?GE5 zF&9$KvZVd;Z^H?%)?g~<7PMQw#LbizU|l)!Mu$Aq!wM;@J@>D!e(=oo3f2Z)#jp=( zeiqB&Ey)iP3J_+sI!o_S!|}!HMs<9A5m)<>JWwSgR|fY#*2U6+MjADvh$E6y)uOHF z6F-3H!tx&^l-|t}dwRS%>MzpYvBDe9u;@IknLw^mH0C%B1<~2~{N`xNx;lt_p%;Bd z-h`GbJhrnv6?fMs^K9APqhwd7zp%WxuyEwKnlPio(y!SlJ8f=E>Aip2UnJet1)AMO zQs-*&5yc4(HmS{tmU}bnnN%R9SQ*=-y!%h={T<>SgXC~w>-S$6Oh&`wYkF z1j!v6Ffl;|t2SBFgpVws?pDLX%y%P!IGD$J7#>l>z$B-8jk1DeIMS0Lb+Ss_i`@=s zxRf?s0fGEmG;+G>x|(>GD^gN_knn=Tyx9upk_-Zs0#RA<_$~ux!b;Rpn<5hjX%P;U zynZnqwIJE0RoUyf7KFgO#Izs;EELazKo1iS3w;|D zgr4E%f{u5ZQqAuY3DXym+xlxs2~*vKNk-7G(Ee3De;tW@*f+9LWXpz=o$~;?T48nG zGC2)(SxXQ1XBF4GvZZ{jV}t3;Po&@)`G{{Eu!mY+pl&ondRFoD5PD(nG2}vj2ls4o zBe}F_u)@N5BiP#I!R@#v`p}kmrzv~rkvYU>^9Vh~XKgz{hj9n`YAz&H;22Mgg*_WI zDvR-+F`8aBafwaqAMq96G-dnug+?Z9UmB3&86nERp*o@f=*6FqM|@M1?K=}_iPf;8 zT0?q(HKDna#=kX6UVtSC!#?~^l2v0W(TIanxkIrG2{I#X6&Lwh1nXxK0dY_LGMp(l z>8A1DQkq8F(-9w&ma+bAe`-2(%e<4{NS%$uX-ebe+1M@EF}R;kfp7 ztoda7o5R~W$?P3us7K|HBo)%G@Jmhs&`-Q%czwLOY%?9j|`>xzwnkJqb zKVMl6u?zo($*Yt1X`0lbRpG6d1XFK1Xm{+Myl31nRS=a9;JANQ*4~(zA<*a6hby(n zdPSEpwzS=9+=DBFSl}787qmuhhdfXo_f%vb6%hE9nTJG%92Ry4kdeVgg1%r>I*|DS z!2NUrR^C3vttW38V3!v_pCF~uj@V)$*9C~Y$~W-TCk2h$Pac=(9Vk(rR2w6?q6L7~ zSBp^L%=;>wF83tCo`4y~l?PJytLl&3uU{dc`jfX;7P^;c^t^=_)4cHg1f>YRBl)@Y zTmI^gn?S>zDUWOA%?ZU3`(qC&Sehawh9t~*mu6o}uBN9l(?XIu9al<7qR`cLNTPhI zgsmh9RDF%y_odKQJi&XYRPg6qzCvA~%BDDMg&gAmEyb=x{SJvoKH}XA~m0&u954!)AI?kuv6v}Q1sDK zKQ(KU-%bei%E%{hZu1}|A=Xi3Y%M*XP!VF3Ql}FUT(DlG)AI=xQ5fNg!llIUL|uV} zC+0B3$_98sm3YaM%8-;#I0*R!6#bD$?e{0?83m~lbhr731xf+Q#DqXhIdG#|ToD-s zT}6Onq6(WQY{`L`!Zj8WfDiFXixey^R3i^1=c6S6f)^IjOB|0RE3TvhKOlkWtGCnj z9X;9&j*XrUqsl1Jws0iTE6MGr<}>9AGQpn)-hxoXSI%>>3yeq^h$LPN>uMSvirl{Q ziUzT|^jVQ`!hyTEJP}s3Qg(;zMQWB0-Gw6)sP}A!q;_kevAt_CsU>!?6@Vz{;f1f>`Ignoar zd%dYQ=Y!8z6NL8OoB&r&@X0D2)Z*V0$*B}Su6|SHwI!~hu)Um}!RPut_^k*AAyjZQ zFM!by!>N0Vks=t~k|RY_;;SZ?Ijn+ZvNjF7kBQZxwwXh%h}E+>mlCaulQtUm05-V7 zbs;(w%ba7%G2(Wy)5N%4{(F`-)WdaeOd?tF+N83S`1*;-2oTB>x3$$FmUe3k9k6na zrJ|*Ob@fBMZAwUJHaQU>r96}*@AMD3blJn5?a_2&b9~=Oe?sVZ85jk3eIpii+H-3q zgtdyove{|VdGI-BB=}Wz>hZm5kKny`Xh7Inwb$0oK~Y}SrY*8>qBp6y9G;Nl_0i_& z47>>`|2rkeCYHJ^dMM6V*f_sJl27DCL$~gn>-XTdBA@XwRDZzVCf_A>uN+NwD)TXA z=!U^`V+M~ac6gioWh_GyFg_%ZP|cE7>ZGXVgf*y<4=$J2MaP_5$cg?7 zLOSLcY?jUx@}Z>bOqI8)YA5nFw*EA`d6v>*#J{%QPrj3MEpI-G70-;D1Qv%w;iKwf z>c;!w1#ft5ct6|`+#&z&?@w2`gD>1}t(PWnE3AFyd4f@^dJ1(9q(Kprm$wLMM@SMs3G`|0=6s%Q~M=r7slux~R zYGQdFO{==(%l^R-LFYzQT_W1b$f3c-=M+DeE*bCzA+&xzlE?k?_YyY7^U;Q_nR}u% zz5xOrvyDfQ&Ic+ohHBo4}ZyE>~~XgHwch}RG1 z#0m|#v^|h-b;&)R+f(DHO!i68X$;P6l~3QJ^|()$wNM}SF=5Tz>gdCM^jaiU9JplA zw>cbVAVk_R>u8C5(my0Bjwg-chYv6qw>&L}v>>k1=INCZC-2_jZ}$+rIcmY_9t{#T zY~Mh@?ah}!UJ8>s>bY@GUJP_XK7RWCP9DE^`{c9!>2z_YZn(0dbIuDzk0aT0sr54v z%P-ToPVKkA{jxnX9XHfBVEk|p=jwFI2*OS&ZPTem39k0&A>)T*DUFIDQ2c$JP2M)X zv0JJMZW^o`5_l*!NZIWgM_%~`Uj}VhM1jJz%<*RPWqj^{LoB^w3vBn!Mh)V)a+F4` zarMp%9XJUU78L1FXl(>**bJtjOe>Z_dTc#ldm6cBqs-rIOli?_+TTa^tnQ}Civ#rf zNZy4M@G4vc<>eRe=fx%v)PX!W)~>1i0Wlp<>(EJl=;tLn9AGcaIA5H_2#l5UG-@zn zX#7S0)?a zO*<#*8!sn!^HU*gGllxx7;VG6#+ICxVKD9yGM?dGV;QkVjZ+_ix)S5g z3^DmM;x$&)IglICh60ztW#o$%OiI0zUalVZy4XuTfOLOgYEs5x?8g&yqQGwPV-8ND z)IeO8Rkl?cEAL$&D`cgp4|!_XIp((_g{(T2GGtAEeTx;U3d)RTrPWYXd>S!S)z_3t zoEEBn6)CKfRPk&n%0<;=ByhqzL9kIOq?$Wyp*wX-u);e*X*=PaSRO6BlPkKKhI#Gh zSCCTqk{lTWwI`e?b#g(nQ8*zIvH*9Tv5$~x@)%MOB*+v=B;+7Sd;l>Bl54VOMN0Ku zzkXO)Oak6DIkI6q<9=~484tG4jG{byTut`Dtm~zP(D5rrJOZ&WX!%}|KD?W}tbb=_ z`p`lLvnWA706G&chhiKoQh4n)VBgTnBj#p6!g0>-X4z6 zjHWrstorJeWWj;W@kH#1L;1Z~>0c+`yU+a>)u(>fAY@ZvgOlO11Z2lZ+TFe^A z$70l5532itE(T1c6GUE?;tOS_`0O_y*-kV z7Bs8r+0C7u$#8tJx|<#!Uu?_5p!>>7zOJ-W`imAWsEx!rO6gD}8mr}7*3?M!1V7gf z`Lh=Cgr~=wqy8fOy+9XC&d%19H2|mq5!Yz|2+hXlHv#0`y`d-B`3`^ItLXcox3Scb z<7!gF+PRKq{?X54Hn{+!;4jkD|6r3^xD4$vTTv!Q=uu>??Ueu7*hf3$?)v#%?1Hz`LK)Dk7oH$<2}j& z@)N83`g?Z$i2tWk`y)4GJp%J1SPdbscw!Ov>6yXifVc81UV^}Cy`okbC1b13g(G3c z^7CEtveqOKaC2oNkhXb@?|?aP`Li9rza)RvCM)61QQ62myvs;&JhgbYKHE8eqADr2 zD7O*3h4>9YeJpwkK@Nw+wyFz7d#i(Z_$8E^y$<_MKov&`9Yf&gF$uXBjo2>^UlUN5pY+@Xh5ML z(*6k^{Xb6+>*eLWc<_?^-dz>zp~!Pajq#uDyX!fK%d=7UebRt68cQ}RuJ;c z2GArA^~i0Q=OfT)CsPAjx-|n5E#v8!`GB|(_m<~#M0R2Akk=xgf{s4JK49;xnt*(w zBES?8SOo@G!dXD8Q3SZ2vcD46%MnHM5w>ik*^fFHiCM`71IkQmTWLWz%}F<#fRhE% z4hJ-J*_F0w9c4ql*TePQ8mbH-?QfJWeB*?Xu&R?E$$OG&7(=jT z)rFbA0yP+;2r*O|OLQ1&tWC|pyhRGa$+D4;!}gK?MJbF+NopX%u5ferPLyXajr~g` z7y?vFuCD1R253nxEe>wpV(OAkcq)c~TacSHAir9$9H@S-f%uw_Lb*=m3&@wlMg<@v z0+Vw&XpyYy1I)om?dc*{ntIa?Z))%}Sx^0&5sZTILKQ>K!dBFA*^b=u*Hg$_kh7~A zu1mR}TLD}~Jj0RW1O^78otPz_~)L*Q_( zlWBQ)-F^!@=lvW3~X@gM8E6 zoQFWSueaiKL<3PZrq^dOS5hV+j_JLO{v@H}1{W)O)QlQqQQ8JxL^_rtOF~s9=7AAm z5SwqdwqMYMu4gevnycc*IOR$J&cVFw{Cx}5QT|fE!uUK`m?Zn-FZB`b@)OAJ`$nVl z8^iGyAn>nW5N3nxuI7Zui##6$!*_`YtW}n8ds08fH`tjS&vjaUZqo~ypvA><7C>BP6 zqVO@XKkM1xbhLSPh3nU01a+v^N|_gHn4zn<<9Bw*(O`wqw`IVfDOnQN@Xf{LpXSYf ze*9=rHUC-s$!iZREG#?-zxKkfBk+qn8WUidsH^xUJKx0%xkVbh1Fof)H4#>Vgv}7( z<)=4Z20fWParHVFQo}~=cHoD-@aqVHOKN!ThmkRZ!rS!x&UCbKxIDjGwl9t*v)f0L ztqZdOZFkncuJKpJGehKES$gF>tjDdGQxeEWVoaHN3*y85-KY8Xyij@D%c-)jm@P!5qpV0_HyaCcw60;tm-48h?Cw z8n%=gU>(F`X|QaltYjb$%2&!t#oWXD=bPfpz0M$Jwm5pvZp5hpf3W(cKIe0rf_HVu zv25bt@TvjKe)5h0iQ(kEy%T{BHVD^rHW?2`!;R@Cbd#h#=FZ&5Ni`!BgK6uK_r=sc z!f2C>d>VJwPdWM|(W#I8pLnDi%EAtTXd}F`5J=`BHxp}cCwm67Lq4=fZpo29?ST7x z*UR%>r^j97Yh46bCX90tV41K5;lhl6x^U;ac6#q*ynWvWG#Q$1Yz?OO!C0yVS}VL~ zbZ)%O|E9D2M6z$Rxj8;RRsV1Bm)oOP)c)-?WY=hW$p7Vnqs=7iS$oU&a0CuYHj?P! zO7g}*-5fZC6xeZ9*@#YSo2i=$+6Fni;^aGAPF))Y;n7wI6(%8`k-E7cpI9Wnh-MOi zfx1z^@OFqQ#}7K9d0V*bw6W+Rs`7c-E#6u{=qc6QW^;@dreTQa9Gfgd7*$Kcb6EwZ zsOMFbAR+wx9K91o6@<%Do`xCEl3$G@?l#VhF?pn$7dkicB)Qy22zbKpeUhD8x|n>t zn|a%j6$O-QWu6T*N|HKw%at|~_hd*SxJBu0UGjn))vH7aoSm4tl^6W7hs+qR(jq3o z7i?OdiN~5ymIn7!$GqD72BpEtR4W|y7@@JuFqAWCpvPfQ=rHhh2d7DcIPW@3#bIOKY zV6+&WnhiFEaxfA21~bXL^M$xNBs6j({~29(Q%spKsHHj0V2iZPg%2yDF{1&N3ic}G z6VVln#?Dit`^G}l8HFR+H4aeEW|!ayXA^`3A@7Po$e0Oj#Q2eSqXzdVWGX}Uq%9JW zMh!3(E>zG>CYKVB1sN2M(no=_Dno@_X{-2K z#8^nVr*L}?g~CYag|=T%t!CFGUMKu9UP0*}y%x%u3#)`zGJm9J|7~zX3ct%a4WO>H zxjBo#k;Y!0{4A136_`uvxN=Dy0#hz?_9%MCVu@`IF>q=rUiv#7awJc4c~p`^JaDpg zo&B2*xixO5tdPlZ=T$UqL*AnJ1q+@w42(E65OB4+j_eVYjE#FJ#F&78io77&7z93) zKMpayJH2rQUbdfy$L+TcE{!H9;Wf{Sd}dE2@L+5niA$^A+N8;uj*Bx(R9M^(MB;-U znh9c}(`wN3tI%JH+h6v8~EDFCE;r$f)jX=5gAj69g0Li>H@G*Z0IIFZj`|AeLLN-;A2AQ zCJ4+8-NZt`=Gf3p{#({j-0gC!=Q%7m5iUwcB7H0R>L@6nKUBDh<&e5W>9Aw9z8(cd z2m;?Sd1G{ZDmDp6L^qjpQjAYUWNa!3FlqzDrk4D9p!7p8{m5`z5XAj^dNeAQV$6*d z2>Gwemfb?7YxfungUB1A193ytS$l>R=?KIYxFZOH49OVrAWd`_AEzm|*uX^T5z3Ir zf@_l@pd-jE33|yaSmFfNXb?k%bGEl=e;o5xpZRD(M;Tu>usV44_ z`5O7+k1DJx)?BdXoFzLd!yjn@1r-*aI***XJh+{^@Km3PuH;CHs`_})IY&NGJ{WP* zzZZQ=Y-~aNg?iQq<{-8NB1>PQk42=iYeccTUCkGJ8|?@uPSWfDl~QV#d#r;>um4w6HI6DWy(7+s>xcM@ z)cU{5F2VlTb0c4iuC3jFVKliESX)C-j=ahmY%jc@5%bvhjY2nH^3U z=8aQ5Xw+5Tx_;UPTge~#Jh<;;$|h(g5OOz#VH-~~WI)Brki!%N`Dv#iu-om3g-t~} zKvb1n*8<_Ea%|x>NB4?WDP1VBK|W+SyL5iUx=1+M641sH?ocR#Dqv*Q3kNGef4oax zLN75^9BLSK#%!yt`h|8#1tMd)prwI+VA_t`%!aWM61!#2+@L9~9TVSq#~l;9p0&^} zW*BGDxa}b^)ziEy@Nih=h!;m{H$KOx>Lv0TUpyR0fUKz)VSU}kSBJd>%vY`#NeT6sIX;$H*J{3I%dhwraZq#qagLAh2QT{Kx?R1 z?$TGm8HZS>4&N_-u}d-(TA?q>GiTUv<&z~!1(m&AKa;(4hNVz;w#EbUjLnHSQy74> zSY9|HzbeABZk+v#_ylJ;U}w>yl)K`Y1Aztq_h%V17w0<7fqF3gA-I|rFp^-#G=p$mg;mU`0?ExTCxUAab;LS$cS= z5NGrzD&pH~jGTH$n@79L-OdwK2o=UvMpro8WXF>uW(twaQ9|&kLCa! zRe)?~o~EQthN}6gt-RzBRRl(t(WtbLJQ5!)15`wo=0U1Im31o?yISMHX(Ksss4#UE zGbqZr!PeFQ9$F8g@~+U8FXL7(=oi5y#$v#XA|qoah%(A1Mqv5w?0A9cP9v*%B z(Tt%8k!o(v8&W=wW064;ct9EbkmuC&{>iLawm2mBC0i~e+ zZt@^s9A)I(Hp@ndPWxlRsS}Kk8$Zuwaa5`-=giC+v zmKVASdW`O|h0slWFgtY9SCX0kw9MjTS)q@lh1dNuE?Op4&Iy48amF*2YUwaq?j#$@ z3xNcEm?4l@G&2N}>$r-BUW@k0ti-J(LdFi*0+x}!733KOOie0To+LS=2#V3n2p$A+ zxWR*1HamEbD^<_AY)rDm4KSO@JsrT{=6R;|;$Si!Y@Zn!@G|z4M_l8$?k#*R(83`1!91{6%*E6Xo_%uK+`n{Wm z!{TNavPuG=>f33yS&OeP)Yq)!1P+3`joxzGN8y5G{VN)`w|0j8T`%0+*_o{E3=8+< zj+i4v<++0ub2|^(A9;=57h`~C7C9;n-n?|$JfEr7SkS>@@_pZov|z!2S(xY8pt;Fx zUb}Bvv>M35X3?;BJsfS$27K4rt;|9}>lUWTnhyE%9+p>SXWhT*6g_g!z?u~K!89F*neU}W8DK*wTY{k=7j?**VCs9api>m+FA-{yo>gk!m z=3r7qw8hE0Q9L?Qjvx)p72L0+rva zL9~O-EI|EJu>C!S?Y^{)U+uGvm1+ z_LkdZy$iwcO%(3xSQ(t-@-!t2I`wsmOLA;~F8n9|+(}1(NeN~#g-m|5DKOc%nD1sF zg5-`3n3y(q-K?nx;$bj%8;^yVFH1V)u^xs;)G#o~>0YC(U>T0|WJsN?68B=aLtX89 z)#8B?$iGD+r<<;;iFav{=H;ed))0D@Vgva>!V3=bW-FXaG6+-(L}kU}y9}5KD^W*t zabuC#me(()qZTB)v?_ZY*Mbn3mzWlWfQ8~&5a=nfEC|shRnhr~MG%JNA2P~}QW_fg zl`GQhd_Mz%-q@?ett3NCr*BIAwRZggxSUnkSTg+5=iz}6Tp<(478o)siZ+>w;|D zgr4E%f{u5Z0dg)1-*<_W>Wj#2{k5c&s&3>YtEfGq9j^al`Z9K9UDw%(l&YpihRU34%kC2FHkp{AssHB9zrkdJ>m#ye+Tz$ zaU;33Xt2V8A1DQkq8F z(-9w&maUhR`$*rr>S{kpT{qOeu1F+CM;k zVs&4C&#oWw|8#19Vb5iI>O?G=ee4(Yrdm?y^=fQ5Rlfu!dXD!Poi zjGRwUS&*7fP}j(H-s$;-iVC<5l=%b{eYDh1&Ds>BJn8ua&TSr~B*Z%O$|%{D5}qh4 z7^U>GgeMA77~zSsTw-{lu8l~l-p!~q(s2OW-wu3b8P^>(_xqer{JvC-3ER2e1O7LFu(CAt07e5PCh zAr$eI^IYr#BT@z;i5J7ljtDgk4@GX@d4*q7gk>gg_SgPh`vkQwZEhyBye*$yL>OgL zY%wVsAu#hg(I7GzY=j^={GmQATcspFbgOF1jzBph=bNAaJopH%%Ym_h&;va6Fu( zBw(Tq1q>`lpO%@?X_DbbdKi=za0oh#MjbKW5FcVfkP>j{E6BJt4J{72j|3kfXaL)p zocVx%TgsyhH-c?85U_G7z`*d_JgxdAD81Mx^!trKzQ9(=wUr*~}Lp8#Arhu@;L zWE`hb{J6e%+lHi-2H0NC&fs(X9{g4WgAgh>nis%mh~d<|i5>XGiH>f`ks>PbRg=pc zR>3k^n}*%TY#M5tIn;_+J)3hW(YiQkqha?Sd0>bFSmuO?KoEx!w~NIR<97M)U1>va zOh;auRF)E7KQS2rLU|&st>va*Z0W8;FFUjYR?e|hwDhm8eu%eC2?@<6Cjz8Y__=rb zhg`bs;m-DGy0JOFZ=^pV#41m{5sNzSbL&ul48|?zj0C@`PCdR??Ge294h;xftJdeg zIVj4n(a<9MCVG>K%i#$*ULS3a&dmBt<$q^|yq84}#Tmyq%}?Y-L$~gn>-XTdBA@Xw zRDZzVCf_A>uN+NwD)TXA=!U^`V+M~ac6giom|C_otsF2uB#=yj_SS~p|ELW(2fi|x2SYHPWl7H!vMq`p^R$X!uM4p{?nV(yOd;?N> zBrE!>PHK5w^tQQ$oanzGq_>U1{^%Q|>d1Q9#Mq<{C0%E#yj4};wmr>mo@K&o>;1yY zknbd2%bSm4#WUk3fyLoa_^A4r`Z}^_2JJAX=Xa)~4fY&b5Usxs9?*>+G9b6hsPVP*$R%GiWhIlQS%yMkf`EGiN zzMD2YB3~d684SRUl>umc|8NS{s>!3}dGwj;dUbM1aFcx5KNuqD+^Bhwh_*6vXmIhy z;^)#O1HK^0_V189?w`MxurZ#GHf+t@6Q%JD5b&67Jc@KaPzf9G$@`JmtkVZ<_P-qO=~vC3^9KB_F1g2Zdukk&$vz1>jlr3% z^67iD9{1_87V3M1lCWkj9P42}dM%PF4qP(m+Z>KlDyn0O(UP;~K|bjp5*5djM)AW3 z7>rw0}%l8am$n7BlX<4Cocv%As;_| ze<#Qe@Z6vEPp6AJb;FevopW9&dK}5j#H(!6wZBNs90>%%=QW_OQp!oYbo4jp&W4FWwH|u#gB=Asdkh0q~j=b^>z6{#1hysOand8mo z%lO;@hgf>W7TE5cjT*#p<7%RJz!{+l*F>Pgf+8IXt&Ly}o53`cX~i-u`(`c* zzSR<|??MWA6)uAE@{9NLViO4JKpq@x*HnJV3cq89Nq*?(B|98oFU>e#oW%%?mGd-e zFk)!@MgQd3Sfe8hy)1iR^BPn+U~+6M;oKpc%OC;a_}>}wq|d7Iq-ZRVsY76aZ=h@} z3&hzipA)5!my^5tFAXV$1bBBbX=eMoReURLXEY zDDCJ-^>qkVs!vd8G*_*r`oyOZQ+@iH@@|>-5aKV9vJZl)H6EZh5;M;#R%4z1xxv=f z0OkP?hJFJloHzLrQu15{?m4e+n_$VK%K1P~)a|Kj%L!JLy3)Ub2RZ$%1Obt+}Zn*RD0D^wMf8O=(o zp{n>aVyLRGDU~=aRQ)PaSSP9C*;15?s>w*;gm;2qqg3R(ngawYyc3kR6W)pC(ZV~q zqU-gaFNI3Jf|Saaf*DlM9lK!kLY7eheuH5@d=b5^@kEK7bem$u-lk(E;z8 z9BDD0albg2j0f9iMh4Gr$DMUuZ=ULiFY;j4_0mG<_?06bf!r(dg?E#e_3zBg7h1l@ zEJ&ZD?RFo%8qjtNtDUX}w6lT!$s;|oPD3ZhYs2WX0{Q9pZW<1YQ?ZGc`W7JvGrVwy zf_^u~+r#mh(KM%8RbRc5EI2S8lJ=J8cVu=U?vS^)a4K6n!)o67+Rm_Ot8dFtS=5IHi!tKVsqk9BOx6{#WhvNXquXV{?*xZ2k3Fk&D)7fA$n?FaI zmUKbqhUKA9i<}saHfMt?ffF^kuyNwSMdEu-8|_n=O!TihMMa}%Q{vh&Vr`MkL_@oZ zT?r1XOJ36@cVe4`7mhbKS1t}VFNpZcB7k#r;!SIz>j?RHF9)w-j#wxos%PtMd7`4l z)0^Y*H2ZHje3=L)$Tt?r%dw^R_@(W^*7(f&Y&<)2Z-1Zm$D$o`i67)a42e%`cIMBn zCE`fuD02e4VHdS7d3Be(6kEd91`mw4E^NtOPV;jGJ~L>&etx)zA52$#=wDs^P_(Pd zMSBN%ImFi`weRhbjI^LxP0w!b>`aE^i`Cuq`1oR576#o{R`PYFozh>la6xS()=^4_ zBGFha-?FAgq9^#dcF3Q#kS9Dn-W>H8>F)))U~+c0rmTWB0r)70!A=8GXf{5-2~+;M zH}uFm{C%&Y?}u7rDdmZ=wKuG>qv4U`1Tz2V=P{dHfKl)lY3hHlNiAH4_L!|GlOyyf zvTQr$#o`EQARq0JyYs`E_4_XjAn|*AeD>^UGBTF{9F4T=X?_RVvr9hKA-~)%#j?c$rM&Hw5*u=qbcD91`2AE)?yp4&otvxIdyJKWtM}0mo3j zC2je{iaEQ`2nK8U{3npz_l-v9H-_Uah@P)spzl-KbBm38UVB1E!Bp*?F-s~+p2zk- z_(7W%gF$uXBjo2>^UlUN5pY)yWOV35>hEMVTjCcai@B~OTt%a|D241zW_#*d;ReDX z*Ln(s4z+Gi!9_!ujK@ZyhbcAP3z~dShrB3XE4D_1Z38mRe#Kjex|aS|*84clHHAn8 zm8N7zdbY&;@=!q6?j$eizZeO+rVwOhF-(BA@2r}De4-+o8X|B$n&3JMXf=ue*HiXa!g@KPNV5kHI_4ym z7DyWTQ3oS2E4g4mna$^@o5j*VgjqWr(9mURfN z5>|Cmn@F zzj+5nQLr)cr;8X^9;uHSG^~ykL%q5%^H-n-V-z8VN@Ix*Lyfhm8JM?71);+J7o{*R zC8>c3yTZ-gJ5ip!H1;o%Ub>S!&5N?+=ASs0r}N}vq6jph1uuQ}oB z*;V64wEjJ^>tTynAFeKKoWInTk;s_eMPAy!BQs;Jn@-6l`n|BpS4cQvW0u{Lc3GyBMBIc961{-H#oakWIU+OQOy!nn} zE->2?fi|DJ#K+nr4vW|!i90RM*j@9SNlHX2I6bqw3 zQTUkHpY?2TI@&zD!u9Jgf;v=d>3gLN@{f>R(*F@<(1MUY^3D!98mut-whS1wpXNP% ze*9=r_4HZ%$!iZREG#?-zxKkfBk)Vjnz&|PE|N!M0xT1C72jm%yLchDNP~C4we+$k z!b*^^X1tc4POjM~2MljaDNbhEe=>RE>UA)_^Z=a&+<_nV!mlIp^x%>jp8H{B$O+7c z%k#Ts`{HOayL~j-x-c8ic4z(T8h=$hGeq8%rB}|wdYnx65q~7cl$p06KHT4Z8tRH{ zz_fSjvHN(qB}3<~^N{KJRtL>qL7k>-r0-TaD?0=Rj792eW$}4nHki?i*1n?y=mk># zPcJ?cVW2!GP^`pB${}hC2E$+3cz~Iu>JOLT4;``RYVN~t z0&FWL?tr1M@yC~^VN0n2)kMLMD-yl8xNLU| z)FnJv{UYz`kYm}z!QoW{m;=FuA>;5$Irpa7|~E@o+TUm~KKhN!nxX%zd0x zGw?9jQFh4tVrm~@v`I!jjeGT{9DS1L)JOjBN7a97fFgT=60BDXWFB%eF@m}eEs|Su z1wDOz(raR12=> z>D*=XsLi}(dpH8G0?zX(f!Pb`vO zL^BD%K;0-{csoRu;|Cqlye(XI+F0~Z8>*deaOfPHEJGMoOTu$m1*WLyRg@qh{QMlf z6Gat-%Tk_(8PAeCs!sUb#{Dw;k#1h-+{lyUavve!kxbIlB;nV=k9RX~JF=pHa;?m> zfksJ!Qi4c@yYffko=o#c&H~-ry5t2ps#l2;I6E@AB zu2IVglnFOc=g!&^Pf*#sW(N(Kn+eA zZZ)|A2V94Yc{t?C%(0p`sy{rk;HStV(f)^Xpo(i#f*jw7JOa43NUqkXYS=im*f$mL zp|VSAxpuGgLu$!Cb;!@hUCJx-RQf1zR^`nrSK2E67BLo5?mFDIL7_16;-&2uRIAxF ziPs6WqJTv|0De~$1o}s>Wm63aFLy#K6oB95oCZ)=+T5H);7DUHPkt83qYBI=bzHfm z4xMtDvq#Z87E5e%h=EfJ4@d+f$#**BNS@~Ms3eDY;AEW62CJ&v>ECq7t#LbLB{z;c zucB!i@)pG}2x6i2bUfoPuSnU@>5{1cJ5YzV ziKkztzqGGA>hX)@ce=46At*COg0{p!*0B_s%;Q3pHsk$wyJR(0sJI`=HZVn`;<`_X zGX$-=^{Qw$A_!~*gc-~(vcybsMGyXhEiD02Nv1$nvhsx`{DlR(vWT-Vst4nLU-jgRyxeF0FcN6aB3R#siV~poeCHnCNIhk3FTIibO!_0_`t!c8oP)Fn!X9jo>AC@6eyLMq4`qvKPt zNjM_9$()m7d@3SiQ$c`H8`Rj;mMCnH?ivU9+3$9IufQ}%u zBHrd((C_~;Aoe7tiEvR_5X^h#!*G4cf{Fn{SbeVTK`wsm3~}bi>|HR ze_=Ga6j)nBP>z&>nqUWk?S=OyjPxHGXNAxMuZ~^%fgnF zs99KX8NQC}5$ds#-mQnGsUy|Y#vX`|GqbP=e8tbfLeIQNDW>Y-kex31)bKc4(Up9lASOxXm@ z1VZknFl^&#h771!8FH90Btr=7b~|EWQ_&6(RVCN8K=`Q~Hh9g^y`oi0_rWFYKH`AP zuofxqP$+^bU}V(`2P;5-yh~m}FEQ4}lVQ{uv#qx37uq2eh>XpviovuUx0wxNBP4dq zoVh_$T016ZXUeep5v`eFoJHfdhs0D*^RB?dVU;7tOzI_RtTaRrd5td~4kSRP2ow|Wj2I^>ra_}v63SJ1a^9^H$P>HKbz!F|V(`#3RM@h>8&tC>E{B>( z`z3-lRLU?3GzwB*TKN4Q1+<2G)_;nAFcSNG^Ei3`?O=n^zgf=)VqyP&jY9 z6%IGq@#Kh^LS)Uk)t~x$c}YIsZO&DDZj}kl+^$SP0hAf}Or&1DTnsu}Cok&X<{xzC zJ%5#M`W>1i>;{CZ&5 zX(4$eKFAI!*KMjS%}XAAD(hA(cD2TX(?)XOxrM2#m_e1!4Ysxh@X&fN^lOYzVggHU*gvuyF)812Zve3h0)CjT&whH)kL@@2Q)5E42Qr7n=QkO%rc zGUK8xbl4CJVz97`1Q8vxF8^)@^0XZJ^{j}Q33%y+*^ein6x81>el#LV41Pm~;HAc3 zFr_-RaPvXuIb@uxcrc-baP~S|T)y!)vqCtXHW|Wc0p|Y3v#uqwxZ3S60PvvdeX{H5 z5)?7vuSj0#Cg?G`#}-02@xkoSOw6`@Xl4Wt zf;imZK`fgcJjj)*=Ule;Y&=rXqYFBl$vqvw;O2Rz_2OVM9&Dc(8SpZ8f4R_$sn2~c z!CzX)8^3bItA}@sbkk{as(*V@y2a4SNDCGW zn1y+c4Vs(G=C%8#MXP}*+x4IKu)Hce>;6@z=#hH{)}+V} zrh&mFj~#T|F2G@`x@z1>{>|67G?@7>9B*!}FmU}9FU7Cx#n+q6wwi9bzes;CaIL@R zoovFX%q{#_hB1X>I?HflcIo_x4K#GQkM}NmM*jux_4uXj!PfZ9`fLoNg8TclKYB)m z1%HE?Ev!HYqAM)A67xoOsczjuiQ2VYIJ!b+X}|p2aKfuKn5nr1?UpZH@n-6w9#%+M z?YVz-^@C@wSFkqdDu#VP;`CdWMTZEIA0`wa%xZO(-lK-&i`9+l`1m5O_9J z?tiR{r2~yLYDN)9B&VuHThS*yp%qal`h$ehyIEpSk2gpCMfy8dc*7YMou@Su$W@BQ z9H*fmIvbzg98FowA@YKLnAw}qa)rlswx{Cm`edFh+k2Gk>Oh@pabe-earI(N=lEW; zQFhwgn9_Uyw7*EYs|y6OyGZI>O+KPH!NDfAIni=&W<8S%q!cS-8#g!|Kl*^YzeC(( zkQ^>-{r(Gs$!K_deD>^UGHN%?`@s%*acbD&qEnVg9jz{6g(Ua6#y@t*FCnvfdS$8QkAQ0_f+YXAXs-E;?9|y>8fxRXivEOA%hJ77h$e`L=zUXM>B#OB>rT`WO73?YW z2;1j3?k?COMByOVXU9@fLhNd{8Lxtqui@)810}R6sArKqoiTz6MvgcTst3py5*lOh zi-5&1S5gY5Qw&l^mnn+$X*;T`zffe4a2Heh1f+;uV@26ML>7RHhxZPOquH@TqWW5{DY43DT`V3O0l zMp?l!9O=oBI$0&|rA^s&r9zdf8-e^=G;+G>x|(zCipfS@<_Dsd~x5EHqkrcLPp47s2^&%fh=NE#5G9UAK~x&aA8(4g30fh)fs$!ML!rPN8U!vk5GV+VBZfXfuu9R$ zleadT)8BYQ)siaXbv`ROmU&WIB5bLElnz{$-l0*}zDp!bUqo)}uO%f+brU9gTv2VJ z9jKys7Z)By&mJKI6=K*xJqBGsx!~Mn^itAn3Qoh!)!E_d(`{X0OaljsGd4am# z3<>CcdI-I+_ZV`azk_?WxRG31G+1HbJO^9bAxUxXXT8&uJ@m*NVzYUKiqPvVTiZ?$ z?=)pQ4aUNr4H}ijc+VJ3JBigLJ+JlGSA5fy?cWD8SZZX#_N4(So)Mx94V&im-Q*G9 z6lMF)1X^M>tffn1Hyaei_b`n{?CoZz)Zq?dgb*NlV$M)QuSX z?;Vva2~?bG$y3>#Ba1nqT{^==9#9mv6s?jgc8I!nRI;5T2B1VLXBea<+s5ntTWoyc zLQlTb@r?J9TUVvEG+sx?gDLcHI02zSe~+b9MAtqHV_9! zn$(egZ-92k?#X+`4O0bC>5w-V$DL|#Oiv={bL+#ET4cSV%NSeQZZ+<~l|d}R61@W@%9Cnitjyq{ppNRM70$e`!s&8P66^_>VO)72b-$|q zNGb3o0QS@BPu^Zx=w71H^A=)E^TPKNlmhNZ;-I|@ab^g}Z~3b~ZUPN=raZ2dHzyQF z91(=n6ecYsVaB^O`&x2!-XTey4jGbYP$)+Y)$v4i0b&>0dOsl~QRr$rBvC$9!d4Ol zs=h{UNCIue6TF8?1%J-vE7S$5Y>LBH$T6Ow)is4>Z_*sLLKKF>R+dY2*y;-Gg@Lfw zK`9;b(Ih9PepM#-u!P0;X2>9P3PDC;0Y{eSzl0x5bYu#FIF3wNFxio*Ym_}F#Wr;} z8dctp~iHw{;- z9Qp^S=D~go-~#T}I9)s7H~SPf*v$Z4Qy1PpGJX+d!F5K+#7_{nV^YLN)wa zL?>z|1SzV}@*pK|8DLjRc%l$ulv2qGTQ$WMr)1<4Dk;DSPn10-h9~L@Dd7n`DoVWM zNo7dNCme))0*d}fIq(J`85sp`fMhU+z_!$kLPd;GEiQW?rbwjl`SQ7X9@mlsF@-3O zy0OIZej$1&IUg+n5FD&!g+3W3?r5MX@<_7cN-FRJ5}3YvJ6+$=qut=x=;<)3j1p}N zM-sh~+EsCzOfZkVB51`=-Hd4)d$5tf;}*9$TK(bXzZaFMF0_BXHZ-N5w z;3K#$2T}#02a?9f{rdd|41&+uEAxG&8ssd)Se6=(+l@4Y9jmeiSeb{k%#2Qx3_sG#Dro_Spu=d?5d#kKp|W(5AV&HM zxrGbJeI)n@K?B&%B+fp9Itsopr0R*gEim@Reo(W1X_6hy|WcPYgZ_WpwuSVnc z-kbne&f&LcEg9_H7^hPFxW0E=#3Igs!uE1@2A}Kq;I|?egiyiJyZ}Z+45#i*?7$zt z($OtBQbZ-bYI2#wDp)3K)3AI0UIkm|$mASqMXa99xs+&KoV3xf2e2V1h@n{K98-=F zw~L)7#_jUoyV8dKsP?r($%O37*kER=&ayseIO7;S z{6tt@);jP^#}ZI@?BE*%F$$}G9OchZWv5AX7IRThquXBW8|KrcVX8D zQVSR#5=f}mlIU=Ls|%t{(S7}#!bqx2iLcXs*kUWSgYb6W30xMztbDFZj%Aw_h^H|P zOHQCm)fGnq{C-|2HGN5x8(1I0yuwFS3`EJz?}Yw4f1iiTeP*+(z!2M*Gfo3;2(Ou! zC>G1m+w0+!M6p=Gcn5<-aSC2TYCsyUR1hVyPg&_>kSf3u%_DCs&= z<*lj$x9w?m^DN<#8~eywc}(`5q-%NeQLK1o+$69#910&*A5#k-(0D(*;I)y4_ro3X z@BaSu1}SLnl!g}N-xkj|Zfr1v;C43N*^-={3og;;(uPOmN+Xbm3iPwOeAz!3BIw+xRgj3b zGID5e@lf$|>5>6oknD3D{Q>q;o;>cKzn8Evo{u(c&D;~E@eL60m~A|YbUsiCF;p|3 z%I7~`b9C>nUVpLQ-PKv`LZ1O8i|D#|c0$G+HdtWNd!0UD^Ht@QDe+}iPrrut$hW%W z9?$J5j!t86W~+SqUWmngx~zrz9-$2NV4Ei>QbY@G zUJP_XK7I=MmM84@+@JMNr;9sv!<7}Cb6zNV9LdbY1KTux@P~7P&1im^#&v4H1@4#a zIH<)jauW1 zFAg{(G|5<~u%JkXLTe*f!)7oIWm>Tef-N;4KME~J20;UXw6zj!|{Hi4iH5?+kg;!&ROXjRi7w2rTdol#OM9I6G<6&WTdU%gJ5+mxh!=g1@Tng?^EiM?zAu z??p=WQAMK)8;VJb3)tU{m{*ll%5XhDC5e^l6MQt9t5#Ee;?sz!K7CDjB~J^{{}L(t zAgEg70eT}bbGTTIb^7N9TU!H|2Rw+%wSNgIeJS=N!fKRsgIE?=RiI=)#fnA=m)6pe z{dC9Hz4zm`GfEt&zG@M2Y(WQA?_C?K-*P5RuU&ZzbAUiyi&ZyF!n%+N|HB+dE zhN)5ziS!G*=BzdhC!3tR^>O-Cy zc8>Y2NFl3Er3_iM40B9J611wcpv-7iS`Agjrx8O{eNCyvX`$*@k-|Dj70;HUTvSa) z0w=r^1RJG7s=34G%qhGRl(rMziRID4JGmlz{ig+OUb87UPribb%9rHG7^pqrM5&Vt zl8wR%k&pm7S`fU>q&|5JDF_l|iX;+p5F|c;7zD{R)3DJ2@0uLhFrIP0IQaj_d-qto zj_Q8=?sIN#?1`Nw2MEz_)mT+4qD!kA+aYPMXyZ5rlQwpOuL~gwxjpeYc20d?Ip^5K z{N*Z;7WkHjQr@JrJVl|Uz&{0*f&i(2s+Rr(RD~iCEp4IdOMOAxeAmq0GqdN+duHup z|GqyYn`?t-uUYR|>oaS{qyE;ZVN_&~dt>op*74Fp=;-A`o`F0p%7hP+2YUBqmI*E2 zV;)3*#=N6~@GMq9p6HS_nmRdd8%E0t^rzRoeK06b#1>v=w+J!0)aU&2IAeQ?(*EkN#5FmR5rH<^?T>Xwg+V!LT^bn0t+xV{=3wHyrx55 zjBPP>mxcQcz;HW%#*gqiNSDV+P@q)QG{{g;v(@~r;Uy&OecC*oS>>vwyAOLII*@!W~wnq zH(+;0on}H!9`BI*u-(G*M;jX}l@qGb{g$L=j{MNQg=!<@qum_5hB;%QjHs`zw-t%X z7Ny6JW^+uami0j#c1Xg%KC@0 zU0W{O6Xc~(UzgUtyGt_4f@U*4y|KML9*oY{XVW92^KE$;^jKNz*X1UqKWpKDsuk8z zT8E<0STEnQrE)>dHgkRG3i2l{)Co_HHio@L`g@74n4F!h87s}FP6GUAI@;M7PI{f2 zx{|kk=*HXpbFZS$hst27M*YM(LCQG9@X!%j`A3(>bbJn;f;WDhp>_wR# zp=XiR9cr&Yh3v_P3-VxoTC?`ZxjqzrkBm;A9*&3Ywg5g-kY8$-V%g}&y7cyVmu$60 zv$3CMg8;NF?6P&yaMewN2Eh9ftcQ@-Jh7BC?nI3-T&tDX8c(Yyo@U zstKqkHWWA*5jYn)5g0X!0M~2muY&b*MA2L}VFQpg>j~~rkxvyEg;~i31IlbZoX{~N zfe5p9IG~`*E)@p?+r{7O;>Kt+}d*D$~WKdttn5d)& zSsg((pZ)3gEh-G&7(^jtk7Yoxi)nJ^H!-KRM`KZG{&VSX^61P+}*t$<=Tt3e~AP`fNIIr z6+6WMEy<;Y;O4ESj>t@}iXq?@g6lH9JA3FzvxH!U zd(j62QDvZEx=>RThUY@{h83180+8^juKvQ+&Gq0$wD~=9*kOxUA8syf$e#!3VtFXa z>i3hE^j@4cOPGW z&due4u$nqw^^X2zDmmswaF906G6`|KmR}Rzb$AJ`33qxkrA_O2Bl1TH6F0b8QBgB$j7@19co9h~WtN6E znV2%{o~vS)KE&pGTidT_!q&5xBi&W`oH*l30M5aC+4=hxsH6O)fQ9jOurN&yz_Snn zcM1lNGD!X`bcpLpL0`tn*IMnQkk31c3ChwuX-;CZYL4$`VaSVyL59Rbj`^KNeF=EF zd$>|RSDD8JX@CCh3OAWEg;8KA{F&IF z@@#N2+&I0$?du?dIn--OtT15p$#?NeZqWws1J^Rjnh2{v!a}(f z2=MZ=$kprBOT+6E3YeMnokvJphSz==8B=JyO?I{?!}Wue z{O;U3KO9f*8ICv4P5X4*S?^lyugaT4^u4k4$_}jO7-hA{6}d}36ywUwTM!@a?LNu3 z=Y93qEgtU4(7DSuUO0tJl&m+T-)YtUsj(t-Xi)Fbkym zpI&qz!a;c?P_D#j%7Nt@HC?}sJbkstbX~B4@`8Ge{@|C_;eY|Izy8AoB}PzpAAS;G zTQPA59Q_P`e03VOR5QWakH^v|Md1t*+@cI>JRJ`ApKp(I_bP*!+v4awySJyF{$l+r z+}ppSAa`UF2g0jIH2KLp1SE!&_tqi;9c)mp$#gs#3Sb>en59x){Jm{`aZ+KQn+Ld&jV#*+zJ0A&|@iZYD-h_kl%nXO8|U z0`BcutK_{##a-rGT?AMrjB^oSnXm=nz>F_l+7o%)nkoo~r92BW$}-~A z6V_FEqLY_8Yju)b?js~TJR;A|%nCj+l8<&WPdl=sfO4$N^MOW9YCbpNDjSKTG_5GO zbc(#CL!OsodbKJ+uoG`?TQTw35+C^ZROfDs$ z3p_u0c1TjSwtafK*?%;L1Je2=2g3LK{^(|xNrH6G7=mQ%i$EWmb<}7~%`=V-59B=3 zygoVy(l})})#NH1a2+b<;gBmc$9OcdF^1=|GeD^)qT>(eK%LhrNsfKIY#c5eaBY!X zouNweacH@3D&C{XE>#9Dzz)NCED5hb{-Gd07k4PH-BT5#AXrr=uUu`b`CG<#NU@8E za0iXTNa%&KUob6wYLe^|&Z27;NoREd9@ehk8S5Rs2HKnptAuwrpHiEM6n>W@4WO>_ zxjBn~NMjFAeilhl1=d6gSE5r8c;zx@k79N#mfGeJ15yiDV8KZ8oq`<7vs|8) zY~5!6svvj8?UcQ!#c}6#Hf=-RqQnIu78o&o)u2D^ubF zVK=#Q^$H5cS|V>O$aCX1C|;RRqNN|0La{U9z!bRW6{#C4U9uG5C?qZ0sV;z8N-_PV zP)&ZT6RQ$}GGiuaD-2|brOad=SE{r@?!Vn3tEo!G<4CrHDI*QnEhSe>kXW>EXG&fX z9YzF!T0)q?>@rKu!uQJ+YNqi;Kvj|{ke#f2WeLAw!Hz6K7Dkh99`1fz^S(WD&Fa+* z3(88j3b<%X8*T-8{3y|ATxGF4+poAw0R;_8)o(V10XS@A=2S)G8uLlr(n|HS zxcPt*hP_uanch^iG_Qwb|u6tJ%nos4yMxrD~2$OEE{K@dav!2PjxB(AOMXcMUk%5Ld(nZyes@ktM@1TocVHR<`4=%h!b zOC~)rEsxyvM$lmNg-Gd*_y`j~tQ%7QHS{?SXkK|A`O($Ay{mS~^YP3dxf*uKH;UNw zMtmQBsEQV;`57-(_+B1i-Uhz5a)Hxbga>PSBO@}W5IYowfYb$GrP$O>eB7vk2@G&% zdkLwVATT#|6AJ;GV^cTzZ`nj~5qMJj6W@wHItmKt4>fLLIixR91?*U>pUr~84M3<2 zd0ljVDmDosqMOW-6ysAFnVSj%jMk8SPa9_GN2c3?ARgb-vr(}WV{WWM$bVgp>=t64 zDPmjt4&=4biMS!_oVkV-83@D{cpwOZ49OgEzY|NMAnE`^!LMGi5(xdn%+)JkNMyrR zX9(B`GE0JPG7FZ_^oQ@$a>Xo|XXFoK?&`A`E$FE7OAY&Df>!y)M5zl$TwHxm^xY{57M@a(>o6&&l-aqInn&&x<+?u}BH`$dg zuCGP+)*d-G9A60RtsxkP+*(R!4Yn6P&xl3rv%}EMB~~m|h05DrJb15?o}K`TA|ryx z!m6+pC9PIrp-dfVrZ)CKew0_Vrm`^?TK8C2OiZX!VtrhlD4%c zG9H_mldn4{T*cR(Ck@k-pHoj$Hq4}rWTvel5({j!nRIMl#4KFN)%ng!Ui5?f$Y*@R zaGlMgO|UmwsC`@0xDL98m6D! zVp9ZCV0YLN3!9pDfT*gtt|Y?GwcjLsYGNfSEDr054^VHcC%q@gu-rBGdEaDYhrR@RE#6|X^~coNE0AuIn@3Dk+*=(;dziX1$&3>CI4@CMZ_`bBTM z{JBF_{Kp!HlHu65(=U*sn^%K>vnw2KGV$c7S%b)#;5KtTz^{Ge^Bw(NwOi^+ zVCHsZ2@0Uh$TC@esd(RPjl8gTxBq=Juh^U9Qlq5rnsxjfl=ch5fHOTZ>jf*tBm3>E z&a%CLqb0KRo?50fC2cZQ%`a``6_2PPFuIIRrG?^=_#hKfZrE(HG!Ih!$*fbc*wrzf zoNC2^TT0VZF@sH=>2GfK;i7eaAPmEALv$)%#;IP=FOo})#ef-QM#W4JWz|09!)CIoT1btwRCoab_-!IxoUMcy#%!nfMx~x9F9i@P2dS-^hKT%lSjr= zy=`%e-zdbc=}avYUdrMh#&V1+xrZ;i_im4brbX$!($7noA_XM>ZY!gTUq=>R_Y^Z;eEf1gOux|z>)Nhhb|NaW*Lxq~D^<__2jmQlVH zH{E zMYQ|ym247O(NS8JO+_56+P?eVegLFs0s!Z0sD^5_6dmgMw(H~3)gG(M0^d$@ron~4|{>3-9G?@9%A8l-`FmU~4uf?zK z#`l}dzM5{jw@804alOCio@|0t=9ikl1BfH~9^txpq`{S}ynE5}^k48^k6hU5Z;no_ zO-JxlaBr{jN6%AX!QWtR3o8+V>@~T1PHvic*(xy5u7rEE~|us2NQhQJkt5ZN;8=z@CLo`HHRN_Yzv~W{o{L z+8Fj0>F-$S4QE(%oz_esH!1qA>PcvbPDeW%!wGvBidO>ty=?M&v|8aE+glTHc6|n~ zitRZ}b``ya<;8`CLr0p0vBFlzq}5#{bFN=LqBz0+2K717dT(Yu(+Z>#D`T5qP{Ja} zdkf-z2FVe^)*d<69}fpdMyF2?$HR8Zyzei_i&E1T7oD<2YA;A@I6GH7C4bu?17u2# zeGgmnv=woBp7%*qlH8{1-xuT;ky$-C)!*ok>x{NM{$T3zYUhJLnr3oj3x~oS8O)+O zkxGrl0sz}n-M4Osi&&wKb1xzgLAHaC`&?!sBb_mW0^~@LB03>VOdG5A5i z5|?W&1=lGaQb*S*iu7qMt5YE+zH_!vZSwU3J#32+0MQ`-P$7k1g(Pn&$k9~I-W>L~ z3^*EBQQ3Z^(J(gT+X;6|IENafH83>y39QULlj)=B|_RFqpelurTx8Z6FTj6J3mms9<1{%e_Wj z!Ezkw#gJLDD%?w((zLsH74ri1w`k}< zSsRT?c^66|(mCR& zD}-LydxWc^^Bvr8iyO(MO@kGd&U3JJL~Y**SMt85?4d{25U=GZyP$)v%_Trai!#(AtUiZ;hH4*iiOnG~~aNyc#nJEe}rR4#hGg=#2DL zT;*p|T>th#-BZ5|GUYDa^!&Hfrs39P$e&3|*{8G(xljMjyiX-d0vo}#;;Ed@k;R|AB>`~R8)xfsbo7xYE=h$3!p?QXBeaf%g4_uylv}WDm>47$*niFwGMrw{siVX z9EH@NchzC}hvZGv+?3j!@}M7#8FN3@JhLUoTVNL&wRc4KAHAr+k!U@ zy?$BwM-TNnHx`0@Oh2?{m4h78PUMOPRro)A5nRPgIuzD8Z3$)*JC202C7u(=ufO9xQqW({-m*UV~{@m7hb%x;omj(FN6}1@Ng4L z4Bj+dt!wa?G5X-i^y(jW>WJ!veLn4UCONJ&rNQc>a|PZ~o~Jz+o86Hx3Y#w5}! z3eqI#Zu5n8jK-+cib73{(JU@`BG%AI5%XmiHc#1-6R`$SEG2*#;;*~#G(^qx$z#d& zXbFJejfM1j!V}4cE2)qPNkr1&1dYM#tM}0L9W&YufsIOsQDxLzQ(e z+&TiJBEEK>i(TMEszfC5U|9PRVW#1!$UWOHi}AZ}^!NT<`vkMEd~PPQyd$4oL>P55 zEZMN7Dg@rVPIQP&2elL=hd(r?Wt)_3r&3w8c1F%O!2o#j5!{ypX9KARlEp+s5&95( z&R$vUtIQy08OE~IeBAD&5z~&>CU^RpMge*upkF!ss#1;|E9D7?Ii2up5P3sV!T~PY z(7?c|JftB#RGMV^k)8&nB^-hdqfYYUYPJ9H?-~F|pG?kiqo~!Kq!yh-d2TJ+Py7Iz{(LzO-t{}`iFSmlu*#rIT4|yu_ufYX!qy7AZGFAeF>TxqeGue! zdXkE(;R!un8*U6wO?yk#f2Y*g#8Q{U4uy;*kG#&2d@L^;x@-GPuM5AG`HYXD`hET} z`7W({`Ea~lyN?M&H}og#Q@C8Q&Bx>~W0jQyo(~Bm)ayw?V+{;N_sw$(BdK;JzDfIG zi>=HE!aIE@aM{YOh!j9`8vX>6e6~aG$aX7`Ph&ckoJ5zJD~<;EnJ0`o5xgKgDm{`M zz&yg6FMpf0DrTSmm4Ea)SnV_S9CDz)-tUp-$^=_>z881HDjH?@?a=@w0W6YsIzu|*ng|qaDTWU|K^`hZ_omLr!=*w z{sn@&*>?SPGfRr>wNk^ z=5e2{dZE5cC<$xk!m-}qN3TUv#equ(dz*teor$!x;vd$~^t;KGt5>jz9pvNwN220* zqBTE!fWf%sX*;9^byYE-u7{tN1Y^nD{MR$0H%C1<-J?OH2JJhD=IPBF!<_0zBR8(( z#e+`B;-~lTgiHfclF#^Gr;C%i;mC^4IWH7Fj%4QIfo;l})z3S@X0*6W^Sb7|1%F(Q zd(=1J`QaeW&C-dVK?FQM97}0b4uKNyXW8T(;~TpruDJCvHSKyh6!1`dP-C~pIP%K( z@MX}3MYbwj%N$QOU&iMS1jH&Tw!n7p+o(YtZ-~-pHNx|`Qj2}=hzcDR6zyoxswHdK z45q0}E1p3zRE!TY(f1%&2 zG~;}E8Y3}o$kS-SkQ%!%V=+AbqW|UCSf?W#y=;46^BPP!;N{p@!e;{V>;P52!v2k3 zC%xsSx=xD50$Dl)7Wf{Njb(w5oz$T-q80K|a)0k7A+3<$uQvBWzgWHPr7hd^EbNR!e>2(}<-$bxnCBPfOAN9I5&s zs5-_I^x4A9EjN{`u|fY#e{-`B?*aFtYVBV_Dqo5{iL@G}+#r?(R+T8#PqCs=!=<-$ z6gWL8^(PUt9p){xgc_H%G@V9E(WQF}f5}{}_>iw+#{|$Fo=RtQX-`ZZ8Betm>Mb`( zmGOr=%wufDX_*G&ijeUP?-1 zx!Jte$-(y_ogbK)RJj=Y@hE*!V7K@&2PaWkATG;l-x}_$2v*8UQ6Exj*fr)iBc-e= zl`>^bzkQ39stU@CZl%>!ReTyTRaMuNMx2(aeibRLlT`6)Dat|B37L=$&l^-EB%eS^ zf&`hOh=iO3i4PzqL2}JBY;?fACPy`lSKQC{$D{t%sbN%PUx-3=UCcUOS_mDze8@8p z1s`Wq$cIHO@j>!H@4n1hqJ=2tX7p#wJLHKjS)-|wQ0+8cRCOkq0FyW#{@jj~ORYsZPTMKV*3 z%OYrXn(cS;c!%7F?G~Ot+Sph*-`_a*|DF7*ARq1K;5Ez{3uQ!oZN05XRJJI+IjXCE zQQnbti6oGGW0AZRTYHaO*y?YNPOVKx(^C)k_9}lY+i{opNgm{o__StM{_IvlMOxt} z>~il2xzA5A$tyeLCDdHgmmo_CIN%PIz*(G3+hU-%E7G)g~8 z+$V4I&%KI1A1Z^T8g*Cev_hEga7A^P*mIcdDtZgciwg^fj?l_Kx;&=ibMO@Wd0P7K zZ%_}HVLfIq%Jc|5iwv7)*SK-{;etGvpVq8Ba;^`B-y@^br-$RAxdq_L=Nx0tX4?bg zBL(@Tb}5#Pemh-p7~03XWUDorjr}wm1fXSMm#vG2JEb&e0K6Z;dI)*V6N|b}PW3nX ze3V~#F9NI8j#}-Mtc_BWe_S}!tPB_A^BwX)YnlkSxoQwd+dRe(z?`>Y+KwO3_!n)m z67C$)3wjTg;ZsJ+BdN{1wdwZG(YmGBqTWXE7V#!5yJ< z8WMXDeTn>_O;rUPLv@$5Wr>w@4xteYZdTHle~9cpI~?w;4@R4iM#JDK0R6QW7PWI; zdrC*aRPD)_6%_-HDZ+~=*Hp>(+w>R=sxu!Y-)Sv68}mfKT_vLdg@KUoWDQ&52PBKT zt~I=gMsHCI*`3Vx%(cc12-d0BKfFk60*_j^XW(KWOvYuSP+@wvDp*9`Rgf3vTgB$E zzhywCH7XCzfnu)F+;B!E|Cx0^j&rR+WCN8p5P-A$$i2N6Apy=B-?D+tnrDZu?c6~g z>yo=M7W+fL4D6{!+LT{D^BFx(1fPyYl;+F;A zH6IVOth1>g-|OPWZUt3_Q1&-!7rt}CC|J#sS|Q2iV*#$>edB%xeV;dDqLLnDbp+XE zr{YSEMBdfS8PaHGK#d%0BckLCNHP;!;r-^E{47Vo#>gKpVqkfsJu1+!S)>^1(S^Cc z5;Ym42r*O|D|8rYu1(zlRjCMM*j)S~`5%Mf7@#G&v=H39J)a{o)2m_#xCOaM1M;f{&w(1}IO5B}ra4QDd;$4zsC57` zGBEir2W^tobAUNGsXbfdB8bL$eLCx^zn;M~FkYx?$XVEmInwOMg52`=Q^=c;tE&pG z%LGX{bfg(utZ**^4+f&jK*MyQrYH>0h3XAwo)Gi{5^U8iId{a)CESun}T|1^_U7afeZci?@c6^3f48XN+L zbFD1P!|m?l>$mJL4}{g!`KovHCsWBWzYc!7ATN*~jr*3m0}jzJ>|uZOx-naT=|R5N z+?+z7+YcSZ88`jlHJQwnlu3wV`YfYAN|?C8)ryLmQDbaM+rW!RVkxsUw8_Nf1KLY4 z`^b;3?(JQ*>nZ-9V&;!r6}$8yHs9OYenk_up2ZyLuFB`c8CL>u4&KYo-?u;=>Y)@&yT|lrqmd%$S0wWmX9&&2+e zXM>aB#_1JqUk4G)p`-WVn8?lHZ+M=ZE9zJ;U+l zxoMw{JL_Gm{Z)B$h`u+LUfF^59KfPQuE<^Tp%_Zt-wWhR$8S z@nYE}WGdh8u8UOA?Ap^w%1?TemB0gIk@i}9e4d!~r}Ut;_i!I(fmHv~iw;CMD31in zl{if~Kz%`f@XPCuGG|}^;ez7Cmv7X};X3m4)gChk!70iv>h8l&0&FWL?tr78;g7FQ z!-C2*qI93i6(q-bWa1l94y? zsD9nyr;AB_|XR-nlgxLWGf%Am|-{YbFnmM_XrSR)!YjL&GzpExvX^2^el{jCLjs0&q2XeLu20 zfHBD79Vee~IdyF~geO}eRJeo?vv;Ja1^L(_`2{qW01Q-}0!Fk$G&wHlh~aJFy3@v@ z2OP3%$L@o?5q-xd%MnKPlJG)SiK*&5Q%xIh0R`dPa_mktRS*tKc@}1rC1*|%eztM8 zSI85cywq8%ljL$AA>oPOjk%6OKHAAV?Z}P-%CR!f2O2dAS_#30gvR{_?QXx&Nb;5r zd0vj`)v5%+PQ1C5m;ADe%ow3k>~W?q7s|l6J~s=EN==|DxQRY@)|PmI>ZXu-3#UTG zz=W4o{yhdr;@CInM3?rcQFb0wY2rKIJhzeHPO|E&HRe<+rXti=4^;Kl+UvPX=nzT6 zYu-$DqMXWMv*ItB^;H}n5phAyOq<4DY2+@mu8XUzicSxRvcL)_l1N|Q(6P$Gl_D7| zKA2wE8M4!-{g}GA>@Qc*iW+%UmMv1pUO~F#X)YZ&^#?J-DK)>qXfXyg8*BYn`)-<=i9nLU7Ty&Ar|iRuD9qqsso)qQ%#wU8x`UzZJT<%57NWr@ z9LcV503u3d?M4lDX6u3%;US57)u2D^ub?s-M(hDw(#1=uO0W8!K3f&L1o?6*3xDj_H{W`efDK$cj_Oy=bk0)D$gR#TOV z$B}FYQ$`xDTT0>QHdh#tSU2*D=rAG()DprBW|vuNCVh{Z`~`bj0;-Zsf$U`ED@*td z3wC4?vM`!79PWNx!*}+`HLF)Md?zc^(%%AA<<*_oa4X2;M~Oz`DvRCOe#LU6b1O_y z0f#vpHZpUnB65xSq;6@Y`dQq3KncU%tC>u1Dq5P?Lj*&#+Yf2?L;1v!@!kKempn~3b8{`2uNK3R*FsC#K(;q7{P!rG)a8w zCJ4+;-NZt`=GfFt{#!Ot+#Pc3yE-gO;cB5mOvtyQkB))@`a_MISPtn+Q~^8I>Swbc zZuJoeHS)UX{8VfbL_{~4BPqtGGBP(61Q@LWa#JBs%T;V=T1sc>N2c3?ARgb-vr(}W zV{WWM$bVgp?6!3hB!`q^4S8*JB5ue!XRcvI1_H4K9teUULjvsrq0a>&5jW53J9_+* z6)TU>Kd2YBC_^F}t~x`&Mvz$&bdy=IgjRm^ttGLkAb%KhSD(dbK}VHeYS%JMuH+i$6ADRk7}ZUFWRWQ9J%f4=AXx@YE?>b~>kqFndo`c!ewoUELoK zl5^x0<^2&5y5EUDB}%5&TM&QIlr@4oh;0E0I@WEx&6W2XqC*~!bo!YQtE~+E|43dd zG}V8Qc;?8Wiex(5qWEJOY3vG7>})5@OiefaLo-Nv|G!oo<#3NBsPz7SO_e69$nuVm z4Yv>R7peXKCc9##B*aU;7TsHWt(fq|O>_FMOU6i`Zv}p_@yrSgHz@ zx4n4qUL`#}0Te|>1d)YRVJk{nt-`|8hDbBDu?Os`GM)Jm8IR4($=4keu8NIemGa4xhH1(N>S@Y`nY5A2 zv^7LxfsHnkj_r$>g)6x_-&x6v0jbz$e8X^^&7)0P9;^?5ykM?L^?=bOS zhgTooBgUw7E-oo>8ai0QoeD)z1&pj`;b0}`pX-o&=^@5CcruJSW3|=R{X!E`iO5*4 zMrqgE;k6yNn+;JfJ(C zs$L?G`|{yH0%T9cpda_OWya7n32F`cT{VHoh#^t&8Z?S0p#~sHz*c9xlMrF7aoFT+n>SSC_qffh3 zL%!G{83x^;FDf!;#M8znnWOZ;7@W#HyH+k{1)_6?r)W@fj3?w7x+C&T;Q(6C<>6X4 z&i+Myg0md3vushzU5U-Uu$V^Po8{~qs-vYgI4q<93q-4(DQMrI0eSb5t5$bYMCv|h z5jHLc!SlK}WPncaVx(pFyah#CT;%ji^EC?1IqRtYL1OSu}Q6^}ldbt)FSI>wVz ztvGP7G))yV*wmT+=4Kx*TK5NhN6y;@?D2vnI+ZWuR4?ck$tA{Oz>G4ZVkU?(>Lx~F z`OfTof!U6hR$?%ZCY5N;P->DI=6ty&`5cZ%22J1zW%Na!6O%{AQ@w3*Q0~baVMzYr zt8(75_=h1H#+BT|m)(0ef!40lsPhy???+}Fw1o{DazP9hR+%7X6mfD&0M`{WBEOcE zF*60PqA>gMD71olyTy-2M2R78$P~P?7!0P=Qj@0Ubr^XmoVYNdq;P?^+)g!mh(h(3_-x)!YinOWg!L#_-re>Lxyzow}(j!Qm_6 z5**jbF8M@O>LY34eZP!@mdTWJQXoN`QN~Wk&G}O}6D5+D0txysQy{TuW(p+NVb{&f zzI)m0(LSD4xRpf6+yPs`GRn7tJfngI7nLmcxafezjO0NOhnqZzWwVn9xl-D7E_j>C z{W^dl%=1d?`Tls+-#Rrk;AQM353R@LEO~q{AzoU@8@+tUyN3^oy3=9}{*!<4%`FXP{_{s08!HT4 zf7xsCtGn_2CbO@mo9->r-%DKY@3|+NAeH&0Ch#KRh`wLA9`KArLEgRSdHOGSuSYIy z^*2YS)}|wPD!8{-`J?Blu;6bnw}q7mL3V{jS7P1BuGOvQ7NUu5PjEUyX9>*9zZ)mL zItCAOZo;_b3zxl{daR3;Qr3I!U0MI&dDko08dMX*7LbsB>$>PrLGpuy287wHPSbPL zV06AdQ5_kb$F+W>2vqCH<-z@rbg*ndE2CyKaYS*dUbGc^;sTh3rhBWZzn9Q@H*4(4 z(Z;a1NPovlZ#cuE>$GM9xk)WU&T$eNqSMjN#&FVnI5`C1@p`mc;T_vs6LEHZ2CtUw zIZSpH(4ksfSU7Z~*{7oKd|$m@P1;CA6ui zr;#I_F@pl+hyzF&dhzcI34<~CLBJB1Yb^!WDIQWs*C~qhX)UW0QAdHG`e968FVMrb z7y%Fs@(&e~^mpEZ!dnV*G*z=VhyASap}9|BXzn4$dv|4FXc}M0 z!_WL>hq&J_cN}FT5du4NBresXD;RJy(N{6dj9+e(h}|FaF{n+<6M$Kpi&`fFCIT-z)V<$I^y|kx7B2WNv8HqY zhFnkxWIn3qifnMrBDuROH^iIj4V(&^RUP$=*Y4FZ=n1PVgoh@npq zyizpshHRc+wPxy?nnYn7)wQ-8+_4FjZZcWCZ;R9d*?W z=xct^_hhB0mJP<+I{>;`VSV1JI1PPS%MACM%Nt$QQohz5{mE4F#NW35kncHQ54F5O ztqnuExqMa#y|DKfa$&xM`)zR}xwL7p!qRyTwzk8S;_lCSUsLwbBWsABNfOJ}85Hy} z?!Z{hm4qrB!L`;@jJ_vzo6_o-w_U?aFzJeAWqvX~P(r87+B0YzcK zUWXZj|EXj3>vCTeb`yyGVf}z#e(*%H5^u#dG85YtP}| zp5O#|W%52vFLk7!8(`eA`{h03hN*!x=>U$`Z)G`Zl#Vw;7WT}(83J=|eYjGKuGe%K zb4%N)#{F<*5DPp{?FFrodzk}_$Neg@j|xcqs>(wmLkBul zf7}Eb&P;h+t4>ZRf!JaWer-xjNtpAl+`dv=wU#Qq<8sqdl36-rO46Lg>zsAus0TxS zLPzf>q$CZx+D=ItpQ>Oh2?{m41_b8%9Fz9m0*n<;h#pNU_;oH{qb|^7Qv!B_9HR`K z?uPTve!o5~U^j@u2-uC~5(9Q~1$M(g9Dvo*rLB_l;iMqe{HiSOVJVC6&X7Ur8Uz`Y z<&p=o27x$%tg&EnAZxBs&772WqpFWL_|y4J37+p{f9v5Yj?tf3NpKhcsr*S}=f)s? z`Y*h2LBqdm>|Y2)a=^dm^m;;FgI9w0rKl$~>5#$O9g~~a{=2T&o*2AoI$zUev}NRa zLX!ol^@Qdcm#&^rQvs)es-A#ikCy(a*_vXEC%vA)xy_T5gj}bgW5BMI^rW$ZQA;mL zdeR^YBRy#>mzbV3S77Oh8dfY>iBKZ~mx>Y(dD0k?>IwUyo`7OM{xhK5ZF)sPngrc# zzOar_ADmiIsEIL}1tenLxEf|U85M<^KoRq07dB7Xk`u96Gm!_t+1YPi=8?ye>(LSb z!5a(dDajMbhAXMSPe|bP)qCjrjv4KSz(%FRs4{A_EgeaWN^<+D^-Q?}LMq~G=egJg zPNYgi5)X#6W*VM~+_U{M;Z^CU^RpMgg%+K)-VMRizv`R?2D+ISI!bl8gjgw4s54 zRe4CO%&0WU^h0T8#PnE5TEZddFgkU_ghPC&YF#9#k-CCZqqN2aU?~z}grEU@XL98O z{%xg?;#3fhr2qm}F2wj+P|pOT7yE>He{y<#LpiP84uD41uVpqS5nsHG09-kT-(s|6 zaCBpwN{Qp@-bK{1Ms9%ZWiSY#hNF1_jD{Spxi`5p|49#?ZpqOis`2%Z z%N$n0GFh93-N$SidYd`aid;RLb1B)nkhIaUN3g+JBB5C399NE!w~L)7=I!#|ySkcw z8+lbyTS|QW=y(VS<+1d(CP%8mD6LMIPQc0$OHE7f%KC?R-;_|$)HxBMrR44YM=o9d zaC>VwS>G6)9R^<^oQ1SH+f%Q@qRxBW`r>WPrkPzsz`5nfNbswk)Z@F=9>II>qXA)S zRkyA689H7=7FtI~WZzCtQgJmrp~q{(jp3jq4lp32@>XO)@kg?#1Kk+0V%gcuD z+CJ0k!f$0hH)dPTvV!wsI>1`jp^b$Y(p`j%>FA`81|u$w_pn zx#DO*jLE_QJ~y#GfO&*BU;Z|0Rm?vBD}SE{t9@p(tHe;-n7yX~H$>FTD-?@m=;`%f zN}*V+U_66Cq1XmW2&HOL>o-# zSM|ALP;WqLk7Q+kwWJ~}XL(Hw%DIyq?L9AKP>#V>sl1@+5*v_1K9F>rsq0pCgInIl zRhvKZ_C?^ z8|zOYxt)%-w^pj>zbbRkxOhwX zv+0@vUyw@RJB}}r&-vfqOV}8%M;o?go{6G;0~9=F8&4w12Rb2!Zss$2*X8QNdvN*}PfzB%sbw|3Q|ri;!F`BsNKL8mb} zvvodwH^kyTUG+kJmrxSc%!OmU!H-^xq>2NV4E8n$ak^eHT5@(cgo%@n`yYvlYDx0$ZT`EJ(VL?lobJ(}QG@m!M9eiWw}#QkjVpQapcAtA z>HRw)Q-9(^KI4C#E>7x(BP%-RyioKwl9`KdvQ52~;7@MI@ceKrrBOKqO1z(ClXr}7?3TFV9?)J5-$#CQb#L#gT~G1<6f=M1KB7yY zfQRCP8oNEl4<){bFM~ELvOwWl=6JICGCp@8AXZVa1-5(NMh)V4LzG6VaiwaZ4^Bdd z1w}g=v}(y3HiKy@(~4){);N2=$2U%&yaOrWZEz8cmmj>J8(TnVKFEXP?R4Xpyz&QT zc*zf4UUI?#_R@^=xFJuY1tX@$U-Z8m8|!q0qnB+DY+i#Y2fQ2`OE?e6kqM+w zY4SIQI_a%-ofM4)vUCV6@I5FS%K{-gv8m-VqS5hEa)0k7A&rjUuQvBWzet-O0R><3 zZlqElH8g6lp_sI|g#DdJ1wWNanXZRc9+z#kOSDp-pwQ^9S}pa7Pa~H4)HUVVGVLP7 zpCeTt1Xag)g02;2ZY@`1gZ`QR=4Kz>1MWxF+P{QUz7%^BX*Ei@K`aZbDp9JRVnw5d zW2DrdM9g-WPty`=T-MTb8ZAYaPT2cEyNpaqoF-qzjtQVUJeAJq(w>++GM?%s)LZ2? z{E0B;F}C8gOoMSn$asc#jAg_cEv}gYJ1gWaq!0NN;xV>qbD%fkMkp?WtIQWInUr=1 zQBJWZmrHB(9o~m@eqd%&8g=^?t1B%iGrE;lQ&sV4#8g#XGdETJDpFb}sp8dAl!L0tN#LY+f?%Uo zNHceY8h7cGV5N70(st52u{>IOCs%ac{`30JuOOB3B{?z&YEL;)`s9LSqjEwf~v2#jlTcX@sy zvp;H6khiuVmCfxz{oeVp?Lpaw&|8v?gHe}2Zsau`@?va@sk-+HqoSk<3&>$#pM>0M;RocgTI%ZsGZ(jg6J_{f%>%MSc}% z3m@&~;5Ez{3uQ!oZN05XRJJI+OAxe|^&9i9p!SVL@=|Q=J#t~Izd1U!HXTh*J>1)? z{IP7uUE(KskVE3rnqB#`TZwC`#idX~4(A}R?2wmWYuK^=qod7pn{t+uB|gvbjH4%0 zs@o(V?BXZWH6MCc)<2Z(+H%>RAm=D4yF2U%Lm}_(l8myT*-TGwY;TVTqx1FI^vLLZ zTOI~IR@V9@_;Y!Pi~Ly&2UM-Fj?y|5g~odMmMxVFDjJjgCoR+oPmVT*y+!(aiLRKO zovj%w&8JQR{AW7a*%(fGotwInw|?lx+x&B{qR)rk!BU5gG>`H25p_`cN0-NRd=8$1 zKTk{l{SE5jGOWk!MVTISy51oAfo^Rd_|q(K}K+o~;;?X3>#;lEffRrLpL zsw&_Zs=K5uORSu82n}Fx5M+hkPiZeKYUjN6l#YU_+LJLWDhidIf${o1#`oLw7!0a2 zA0^*uEjkW^N@b2RDn^Lm0U2O%#_RH zmz$ehAnkBKL6^-(7kq7D+lX@Vy)JI-R#0UKWq+e~;X5acg4Hak6_RW|^0)ieP7}*U z-{;MksH6v39YJ*8sk!vf)Qc!ar?foPu`Al z?M2(aM1mnewdCrGonnBN=3%?ugj`)!a9xdT z-3s6`;u#JdA@E=zsthzt7ix;a@LZ_gFdEDCsjmLQ)lb)h8`0+X$YF;qVtu%|v>|^U zpo`_9D4*X?UebGUX8Bxoosv!Td+a$Ze?OUg%KtQzRu>(Nw|C%ur4@#1s2UsshjXng z%fs#NwOv+5guamy{+w6G-2ym%#rS@d`_Hkao}78xC`FP&fm8{9px_tER3&% zg=sQZ5a*M1_4{uwFcE>5aq_iRJ1OMzj$(qcG*6n-3>3l(`4PMgdC@S)ka);3ztgBM z0Z(@iSL){~^SB`GFD6p`ZU-Y$$q7>Vf&^1andcp5OwDPTwqBpph~Cs8KTq{72K}v( zRvS?hj+Db49Rw4dNP_>UYe>zO`_KX%Q&g@uKS@M{nJIt0I(7d%{VUoMg- zV-hSAbv57QQd0|h zNDZ(3Ffyjlc$@5OPloFUEBW2Ib$&RW-ZLC;o}2dRxU=51+FzA7hv<7_>6IN=&(UvM z=qCAWa!-G8!wh!vZrE)*Pp5U*>$Lsl%MpjD}7)r z(q3zi&l9u$lpeJ99`3^|km`SW(SZmD<&i+S5~nE#s4wUbetG>-cn+I62Kt8!@CN~P zE0%B6tmHcK^wl1-62U3TE&}d8{3O7(V&V=s`WgQC>NISrW`eaJkEKlxu@M%WL(+=} z>k@(c&$q|9dzC@VZE^IT-P==7f3f}*?(N@EkUO%81L4&VFxy*Ki_3n}4*~fK>Ex|N z1UlHDT$AZ|G#C!nCmS$LlFpbrb(WKAMkuP^cevAnyeFpj5k{M25n!1x&P9M_ z!WM)BGrn};%y-S?;qhqe>^ck?nyhd3CuiYVss-9Bd}w%Pw8j6XSbiGWJKWe9?M#~g zH~7n~;mc9sePBeW0U0wqk2hrA*;kxb-|70ga?%g6ohZfu{+UJK{zbsS(s6loH-F^n>@Kps1$ott7yRtr0uy#Jy2=_Rl!a4 zxwE#!3sg6S)LS?e3^!K_)SyF3KYZgk z_0<}4sufca>Z=E;`f4pncL^OLX?V?>$xf70Ic!$^MYFz&<0B$2sF`Wg*h`OTqFgFv z?sAfKrIl6M37kkWY@9ky6*A?oG|@s?XbEMJyei8UC4KTVX9Y$ zYJP#yVhn0F*c6IjBJd4nlKJF|guIsPW0(9#blLSP_L88dxga zr+zHDgQ4v_HM`dqqQNK}$*ypKMmDN+oz|S{YP^+AgwP#UwHmT-XGoUGD(o`8AFha zeG%wGvyK|W0i1DCP*Osx$m^qXAdORoQ%$bI0oS2o9uB!ObBsqL^^9!emr$N3qT>(e zK%Lhrok06|**IJ{;MyX&IzyG_Xot>$kT<00j)!~Hq5S0bSo%6`GL^r=a_Px!+;tL+`W2HKemtAuwlpPFeF zTC9k`E898J0O~5Ao3jXrH1_c1XOR?@>%1f|7oA;&D}5;lymFbdM=?7VOKo$A0jWv$ zglqD53UVmVa(PygLp*S@b({UGg4`9iQ}&`3$DP;Nv<-QS5*MTZpFyq~^r!vx)7#_C z{2p+rmUk^tQEiqs92E?Ek&Q%1+c(?SLP6<)0P ztxl{;2+E9^psg^FC6+Rid0eT|?oR!7hpeV56^|p?4yKGWT(^`sL(sZguZRvKf_hrsE~!xq~UP);~KuRN3L1Dn&CTHZI=G# zN;D)AlDxVT8*T-8{3y|ATxGF4+poA)q^JSdo`2kO+`!dddRipD$$Er?=!@(_voZvnRrM+788h1kq1N@gCK_T%OR$Fr~6mn zX8R6YZojL4VK_bxNym!3W=}P6e{3CzYpXiiL^um|+mu(%o105inM^lD5|SQT31X_# zYSQy7(MgX=mrQy*U5;gIe}t+}WZ)Du7=0m9dLur<1Q6?n)PKFcKLx+T)v!yxQN*S< z;`{JJRkTRW&v>!I_woqyHt@BT3!LtV%!S^_h|DR(4n-j#bpco@HgyvpH)>#5lnt*| zllat45SW{~iG_g8v8kK_`t#!W1T^d+i*9c%ToSx~r$ zh0T!HMdzntlOQ6x$s9>BK9!NVsUW~;4Un4#VWw>D2?De9Bhzg`5RdQa*{E2GF*jBr zQ`4O!>RHLS=$Ahy5*K@em}W|$GsI@>L_#>oU264`Lo83Hzf z%#xs+%z`Df^22ZIW#2{n!Jm8tPISKe=k4tYG%>1Rf)wlehpBYCaRRR2NZnInrTlId)V z;*Vvdu`5Kevz-){+Wz-eMM!%8zt(8waE~RZ^!|TMl_sjl@{W)Vw-50bsr~;ZyVAw= zwdmg3Bj<+W3xT~g1mlofOX;k^_QK~Gv50+k7`nN{ilwShdEwcjLsYGNf*Zefl54^VHcC%q@gu-rBGdEaD zYhns4i8~%+UZOoS3|VNWJtU@jhIa)Dhjot3$P72m@Nr*097urdsTlO*zP8L5nkL@z zLzoWH$cQ0P@ftLWC!t*B*6+;ct7+H->cno4K$tW|4jx*D3R@O<(}r6tX1YXPjmV69 zGz)6pwD8A03TO?D${l(uI75iD)ZxeFFLp?VK{x1&ip&{4yz$8jrGm=dJj@D2=L}EL z*x50jkY{X8>H}1r8YP$qyP&< ztCew3?LXY0A$s?ct5$bYkl{XP6BhJ&T^uq%C-`yE)af+@!RA|O9`KamUrW@cZkPdi z4^IclXR|WkhN3WVM|BE7 z++^a(QL_e-EKoL+Pu1xBIo9Vp+PiAE-c(ltGq)>CPyl5{mWfoCmx_nZ*2oKccl#eY z^UA+TH~kKM;H=|UOF9fV(<8H9uu?p-7ir>>Rk){?DNRY6OjYwsTY1GJY6y%jqf=?2 zcqBepC8&rj%}Y-|nRO}_yE?{`)7j#{tv8jXsbU6GI@90W?88OtepKBRrt)Q+>IMBG zxx`oum{Dd_%mh(J-IV^zQ54(fOh3sxv-1UJJ6c+a!91E&qB+Cl7j!>|rVctRO{ zk>|wZk?~Y(Tikk+Q~-aB&Odxr&RZ7$Fhs++l6&~Fd+%VNwZnjx=>5oygSN0?LoSHH z!YUKQj3Oe7R`Wn7kA5vHV`d6oMPc^iQD_DAc8ed4h!R8GkcX(sVlbFeOHG=Z_Y9hs z!iftLN(yHmINn)~Nbkt6XQgl|Z8C*Rzv-5jx(RxW;jx9(O?)srbyHW8hp)80-zTzC zA4v=E`(+%oOs1TZ0tw=bGL~lP@J_9!K!QHZ6i6(ZnF7gmTuno-NBekI;Z_nMa|dh% z%P8Lp@{9@=oI%)TIsryABY6-nI)EX}^GfUa z{&>{iIyE%lWddO_uHZ>0%#rUU#7hf#qn8hP_fVnE#-ayBP3j~$(Yq(9CS@UoIfEXD z6zx8IB^y^(bjWu*o|lS(r2%@;>)t*XlrL~0t0VwwmQsDpT6}+DcFkH(KoDI2y31{I z*`Bx+E5d9?cWFNt^2vdhVa}7D;Dyfj*yuWX|==M-(U6-=IDxTJO!Q zXIg<&Vr6XegV=;j$$Japeg?@A!qy%+*B=iDM@FYl5644unb+R>jC&vC{RMeZYTDwW zQ?^L$l*VE3{4PhRB9|10N9@DzI8iX#0qtsd*1jX-5O*&*vtadKLy)ARQhhB zm#c9Nm9^>i&e6KyjjnMK!5gw2Y~MkVRrQsAY;l0@me^~8h=mIIgDCkrzLG(=x4O~M ziX_Up1f~cUHY=$5)8+s@!jAdc*#%pLC>#V^cB~{7d#zSTfqV_$t{EtyO+`J89O;Z1 z6d*^EF-elWUq~2?!4CqKxLj)~xK8nqI=W6#q)%&EorpRr*2H_@{(6BPw#5j5Xpn!X zkiv265 z{AGu@-!FF@Wh4;-J98v1)uSsIa5K?YG0cqTKB>FhChP6h#)5nkg}W+N21i_;p=3c) zUsG^Nmx{;#SfnGsqz1E`LKZ*T+)Oqu=DQh)Ai3jIOiY`s(cvQt1ar3v7G}O13BSwIJE0blGQdEeMHug=s-ZSSX$a zfgYwF7!Hh@=zP>72*dJs8ErMEvjJJ0HbhX0zyj5`;`m*L3?ziqQZ*)~l`C51MC({U?kPrEu z1NKnM8`Rn`r2XZyLgG^M|O~b9pkUx`_vQKFnG5+8CRI((n5nL;t%IO?g%n6;+ z8K&~t^EbRtCEGbtt2)SA03}j6!yqkKK7RJ5fu*uUzEpUg_mW$0YHQ`Z4v+d1nBQ;| zQiI-AjXxxBqUL7ai+3<)%>7vN%$6K)m4&V|YbQKMH@$E?^L6X{WDoT^Hx|-=Y45o$ zcy0cA@@n5h*fYfFtN_L4uiBY6GuoOC$D6~!s6QR944974;C3)u2eb`4Bii+JzW36y z@{cG78;@Xr)Au^;{uSG1M69k37NKgvxS{j3jN;@0sRO>Kbv^m(!t>f}z#e(*%H5^u z#dG85YtIq0&sQez)AUkD`ndtd9lKxNGj5m~NRtlWxc{xJJuz>Fz?@qjuGFIIHC@Ks z(srtGKU^8a0?$)>L2Kkr$OGeXzl!Xm0usN^Q*GBhovmOJ3|JgYY=2qmP;PU8U*45vc`hR zfvmYkHFHuZhKk+rX909RQ-bF^+24A&ievOARubIBe=2{{*ts!CpZ*IkT+s0E8v7SQ z@o!b|C^2}`bhWO*U((<`OULCByc={GZ5g?q&=f^#J)yZq?r@0odV*~1RP_WDd$jaV z&DJDL!*4|y^#smso}?t?It?8=mR?V&2{CHvB}q>j+rmgs8p|c7C(RXDdLkO^L3s+j zRFrtglg5x#PuLIj1Qh#;F^Tkwf;0)b+k9bxQ9v;EV&*n0T8^gpq%2b*Cv%Mk;T;xj5X8C4&;es z!OFLQ$BcGEV58DuR2enemX0JwCAt07dZt_2lt-pv6nzQBj=l706h5!?#qF*fz$)ZVuUCI@^N?wK4-5i_Elz(vkYTd zO1TjD=5W%8^`F-!clw$}0eT^zUpf4$QjQ!e8Y5IF%B|)xAqz zTXs(XwwJRr_*}0Gzm>rtgc^?K1uz`eU?z6+;&iY`q2?NiYm{q_s=eTl=yj|=x zF>jauo|O%~#vl}ACWC)fQd>%V{pffI2<5T#wkEHtLM-hg-C+V&j#z41dRNvz#QUa% zf~L+1$!V`p$lLvoT)O<>_SSH+zA-vG48B4bl_b6ni#p}Hb)#lxRRy@Ja7Fb}R^w%W zi?>ndmLnsd^duEm!xMVEHryDVn)a5e z|4yl~iKQ-y9SRvs9{KI?V|m%oUE61RUHGlcXM7CR@AHqzcWK?rhvV(qeM}g-p+8xl z!sUu>J|^FTJ!_`QUB?5M9TM<-NFbq#CowsR&{zXQ(S7rr!bqxJiEq+=*kUU)g78k? z30!t&h=b-d{FxQ`Y=_*D?N%V4#&j$>i7qu)91Vyu*);Hi@Tl}iascxPZ@&C()~c9& z{#X9d>tMCd+;hl*5<_ic_MUbTM94U=P%M_As&X);P%Kt3p246{Y$+)O_Ds@nrGh9? zeab2ii{+Bb!*T^TFwnv*@=qP24JJ8IPsvG8c_!_uUVIGd4M^>gtn9Cr6tx^Se#ReT zq>-(Oc{q2HqrK;a%)>GGER`wb14+l3x^7kXPHb1Rby+6dww7r<`A*WYT)z}6t{LkD zmIp)Oqs`0Iq6?zj3GTD{iZ9PLTqG{Yzxn6W8$_$`l%^Kd-?0% z$*tVgzc3sh-yUzS$k9m*@mlK4lB%ig+j)U{iC&<--f)4MJZ5kJ2WtnQeSd!n)=KA* z+{HWwPIbH%zeK+5e;6W2Zq(EzqOHmtG%g-2e>Pn+;0uy{UQH*T^S{5BurXebHf+s2 z6Gi(5D0s{^o%xCf{&((+b?CSOwd!1dyY6k`mG-^bylVuU|%%R2tlijQI z0h|3d$GhrL(?w^8e5*qq@|>RH=rks0w$7*T(t6ydt6r!N`p%*6xS zym^jftYTMaahc|I&3Oy{xE%MWZ@}}zL7bbV6F-9pcz!sR(x@B)CEm}n$veh3c1v7w z>tkx#^>8TQq4=Q2ZjW)~mG9xppbd*`Rk)Tpo@~C1&m9PeRa9(&?cTRhgE-z0rO|3I z9-M>@3yO9$Xw{N6YzEU*rWMa%2`t{>`s5u*32%doV7&a`{oL3BLi0f$9B-!^e?*w3 zFZrR%OHMezUYc>fJdKeUH{@xwV8qn;i~g5mW1Wt0^s?=N&1*2_fR|%q3FiShIg|Vw zL!I>Ix=xD50$Dl)7Wf{Njb(w5o%CsEM1SL@2(}<-$bxnCBPfOAN9I5&ss5-_I zbgeLRbGaHD^w0D+H~a7&aDNaJn!kipz7%^BX*Ei@K`aZbDp9JRVnw5dOK<5Y((Q9s zKR=0>?J#elCDgdArRg+UiZ0!oiuv~D(1P>itJpCCbcd(X8C}{FlSjr=y@YzRAlz?u z1oIeMaayLqxFTdc!#l<@VvQCDDX?hoJ8aHfAts+fJjOO{4)jLc2*qV^mHDD2lhW>F zl$(7xz{OtjKBV&lGm|P8V?Q3HFAD4yKjz>hDhtGAS?yb;4&I;*)Ywa|THQ?{75iwb zm|&%>6!jschFxQRGg8W`QYll`^xL;ssj8sN=vG=yRmGQ|A{I!P6; zmZBU~O-=$Qy%Pi*wZfUWNgv>Fl=}!)dM7AtC%qHPqosFpMc2?UumAiCQYl}OBXgkk zloO>-E=V>iCuBlFH%V<+Xz~f9BuJ1cib%*wkoW*%5+v75!$t?(YjRY>c*XsEe?01M zof<|(_P93|FJ>JtErgCDaF2;RMiuM!B} zqELN@{dt3e{A!2XkIjv7f9uR}WisuLr}O8C?nxI^ZrD6&dz=^yH>Q2g2?bhhp=Exa zYFTGjK`t&5-*wvPn8I|Tcf|>+8fBXr*Nziwi)5x6a}>AT6?B>zH+j56?!$Hq&mV1U ztW;AaYIOf)ss4($ARq1K;5Ez{3uQ!oZN05XRJJI+IjXCEQQoe=GSMU7SR^mS*4`r* zw)&f+Q)|=F^wh(>y~-cUcHAX?k_R~?KCRi6Kf9IiI@BCFJH_?eSoAzCN2C8J%y-!=T5?TEFHP&m6)Lx-NKg|XK zXj#~0>!RU~EDag}??c)9sz3bxW~Dz4fw` z+jc=9e?u@Ii=9Gl!y&P)+Ctgh>YyHe#I0oK586~!z%f*JNn4g!Ip+{6*cZTQ7j}h? zqG>NIYUjN6l#YU_+LJLWDoUO!(+&I*Am4A(V=$=Be3X2rwdicj69ISS0{k^wM9cQ^ z{@=+Ow!{xe7I$51coU7@q873{neCZtjT;ER5u)BCp^sX(XW(KWOvYuSP+>|-_ZD;B zRgf3vTgB$Eztti#?LV{b$8oMTh-{$J1_G6KAGx>pA|zB=<6C)%w8y&SZcO6uwITRn$J$Nk!DX77=>BM1p~@#KAg~-rGW^ub~vD*%PtiMW)O!KU=R6T7dLh* zs4|4IzfrsJofAgEYL?UrNoEfT3chPT-jTk~n=w&I53=D1vdhB`$h*2ZLmJHtsF7oB zM3kHXNoHazyq^e#^U=HIDA*YJ<3$WCkF-Yx8a9g*Lp{1M_gA7OV-z8VN@Il%L(R3R z8<@9B1);+J2c#Dz=!89;lsA|Yr*oryQ>_-J`Nq%~JuH;R~)l~)8WjPia(FrqnF9Ht+qRK$S zbfKmw49|t?4J#~{SAas+ZkH&pV-2sPa81}F~dR;-Fe=h{42l-xea|(fOUmeBiO4jTU zLv1nwc}*sBC1n!gm_Ezsj}j(saJ8bMX4Dv)(l+oSl32OYqgU?KJO?d zC`9P>Mk`V#PT_i&|tt}>4c(*C$haXxamk51%UBEQ?g zh*WZdRK6g=lv3t-hZz%Km2Bq;7yeFgl)R}!exB-E4EkFmtu~@294Vh~3}A*ImuAn? z7WR>E7vy*}W18Au2XotFt!S<*;f>cNDSsIA8BYn@rv9?0c>6Sd#YV!;ydOX9+!`2p z8AZ(}Vzn1i7XrMe@{b;Z=VZlo`hxssilJB-1%|?(iTx?h1}DRf(<|J*4kDOCy_V1b z^;uZ@hOX1ivbSUnZ|E+Nw-@Aau)(O?GGNdGuSP8{{{$cY^CO4Ldic-ck6p8GVPWAS z{MrM*4#6+-WK4o(qORtfoO~CrXIGOzkts(^jvA3AHRR7VyI! z_;pCiT|pDmLuz>KhmkRb#@l3Pdoo-}@}U@4X5NDMaBufXzCG`&$8Pa(PlnFjWtYgQv>_j&Vn0MLN1dc> zr1w@iy9xpij78dO?eTeH)}PXY*51Q?m<3Y(PcJ$U;h;PcC|BY%<-qcdnh9J-p1#^+ zCLs7hnL-)TAN=zAqu}fqYKi{gg68hSPXcT!ChmZvpW%-C2*nv}p1t^1koUy&KEh~|jC6b$;Qv=t zbFr>F{B$wJkNodb^?y`w9nChvI}3qi9&j^JrN9xB3i5$Pa%YbIDFW{8S*zr|M#Wv` zTU`WLCX90tV41K5;lPY9UFtsW+!_qQ?aMgWh4b?@lZVHnt+VSedT6q~*`J(+$Ep^V zpGNi$H#SB)ljdJSTk)acnb8*izZLZVwl#d&%)h;w>>6$j`2T6ap(`C+O<(ZI za(KtdCtOaQbO=#K?_gs=o@|9sJeTmXMe+-1E&&**It7eqhiGzK&=JGi!gZ&OMGstR zH^39CuHvQWJ2qL4Fshe?7qUuBRhO(epk?^`$d9h>?OnBtX9-UgGk>H7;oEZTPBc{z z4oi6!W|U=EuO$3zM4ss6rOsNNB$xXL2@ltx+1hXLiIIG?lX=>a9R-wQWu6Z-Y7(>( zf-4GPy@j`Q$n$bcuT~`pcH+&gyyTZ%WX1@UHiZghAc^X8v(Tv21ge6Y=yPXni5IAD z3aPhnDi{WaHyb0o%fH6}NgVqIo#@gYHOkI|DouRno96`VzKGa(PJOk;oNC2Xg!<}% zs=g|q!1e?BB8{ZsHE$+6QBLKsS@9Rm`YMi(h`69;rcGn7G;)_&*HzZ4xZbR^(i2If zuW#sBW#LMZ%pn^`7lap{Xc0{*ppjQ)*&=o96{Jg^=F)*ve-JaAQu7Op7GqGe!KP3I z6M=6qlgz}M`?TACMEBh^H4}j(gDt!(08ZJ56;YVM!BPpqFH+w(Hss)>ysP^Ki#}Px;bu=Al);DAQ}52(1&InH5ya)lX|a@&Ve*e z8BR613I|+=ig`HX%FMBuXf%6RwgIJ{h>ky;19e`jBsme-Z$~+}Lvd}9T%Dmx^Kodo zZz|rS$u4Q-s-N>hj>$h1g1KHZ8d+(7!Mil#sz<9dnFQjq3jn- zOP`v=`-DHtTb8}U*Fej1VU@@_q>^jfK$`TG?Hp+Ub(PP}Sp-BHdwBA*$d!ypR9$Gr z37+H1W0aU3i>0E$;kmts2P`omsL`y_q3at;&dtQ;cq0%Kw0cphwIzpZjLVqa^l>AmFRwV>w z#!S#w7|0S!naMmJwbJfR{dR|}rYaSWBiRn7j5J)glp@?@SWDs+(P2aos3n9M%r3Lk zjH}6Ph8}+4164_;Kz6e7l_mU!1v|0`Sr|O$RbJhR z4Yz_kew1i5uCmyTA?mI|4MYyqtYdlp7j1le0n2jF#1BI^hSJy z2_V)DssB2+&+&~SHoXzwhaakHoA|g<16>6Gd5s`E!6Wghn;+*pN>|GFC4Z7YTf@vBo0$ZMk$aYNQQa}6sp5Qr`CKoA5O5@;6)eNd_4 zx6B&e&jlT)Aka*_EqKafX+oJem8ENbaQS59NFdf63x{~z%f2}ym;oi-Er6EilYbZ_mEbHnk4z}^~yaR`T>(qbd9z3_QPEMlJ> zhHfsgVyP-r%B}=U62Nr{dc2gR*af6`(DprOXW(>)Y0=vVGSlHCG14LED zbtMsgCdU?DeRz);htj#Y6q*9VfTt`*s#F0Z<7VM_Pg0-jkbCJN#yWU1jAzDbtF8Nm zCZy7xV}_B=#0{mvgj?-gIlBa2+i|#|zbs>7I}K^5H-NWKYGQANRFo#?UnJh6|^52nC~&5ksQlHE0x1 zLb)2SL!Rdns1v)kIZ(8uUXD>`q&}wK@ z?$BGo8A6<;PR3$yB!>&DqXG@h%TKF)H$&ay=<7jegx*z9wLF7n}nNY!GLRCWr|l43BoVyAAS}xZ4Jy!nw-q3Wu8&_9nys zH;Cj%Pu14^$u`f)$mcuyyK1-AmB7sH$`TYnnUQ58tY-P8;-RxO^1|NT{)f)I@^8{j zze685>-Y_s4g=2g$gCHv6p!r5U|g~aknK!qO4?+qnqS(=D;`lpV00OsN(;p!@xdxV zMPw;g=(HPSpUgTHi(MV#$*EQxxV1D*6*HL9nf~TxA1+$=qw20Ol`rE|FX$J^CB|aF zj54EQCWtcXW_mu!6{R5W%+42>?PzHw2J>iAiRKJv3J1PSQgY~XI35`^fhUyF7kN%h z9vM&dw#BW2$bi6-fB34Lw=Di)h=y?`_wZ%+-jEVHOQT_T|KNUPW}?-!5zqoLz!HOn zRVLU%6q&sr=;9>#wXBSpDR>oy*^ft|71Y};el#LV3~@uI;FZN-Fs1fv($svOBQu2) z7baRs;eI_Ug;QyhDO~zZx4hI%&|?gbEu?PZgW0K@x>9as@e^68kEDgy)iMrRCR5Hy zfdp|z8N1xOyQC%zIx__l^kJqzV$sYLNUr0$Zszr9AI~b>N+M+LfURH|HnqNBDXZPNH| z$MaHAurxp~{x5g$0%z%Q)d$Y)9*st&hwbYJ$~6H616HnO_w;>r_xL4^MjFf3Nb<~B zeye?F`f7BUevJC|j5Xx9i-NIvE?~?f1PHr0z_MhM1$H-t#o!HO7g&d9f_W1N!34+) z0tqC+srtV9s_Of`I;YO5uWsMT^7~mDH9cRQcb)S;r>fe^4{U99-h@q7*#W33JN0YU z`2B_IH?^HeMHIn*xjS!Fz&9VY#J@`>?VY`?_VI_e_xASB?rnAOmpeEr{QsZK<^9NK zRW-&LGz07?4|s>=)BdO27_X51ap;XSvSG-#FfV5h&Fvo?7WegtRvEH@B3f#$wB!e=k~4o4}o{R$lg$PvD6B}21pOj^8Qmq z1H$iCPpdj=YkFN*RHvuc(|UiR9Z1{Ap-C81v-rNHR6bCHQ5!UI3^=7fIv0N`6i{uK z1yR+%8_|0ATkM7D_N0AO{WDd1QwWTW)A|W4T&gQ`jtg?A;b3}odopuhUPf<`yhjaq z9ur@i3@dzaZ+B*D*OwM(*&V0E@mBjt_vn!$_n&qFvletol{$f+$es1eN2)hG-c~V( zZTEiBXI7C2V$H-Bo;5HYA-=OE@Mn;aE$sZW*T(ykt<%$|pPuYb<_q(FcS}4Oo3>zd z+7>AcNb_d!es4)^C8pHd_kfzGb5UpDd7nrlX?mFQ_gms^iA6oTG~OQX>x{PZ*i*5e zFLyqQN7Dc_NZv7WOJ49h0q)5Cr2r@7T~y~{sR2Ad&i3Tzt>+`fR9jskJ=3@f4nC`F zC$+HT)V~_Gzt{<#MDLvBHB`|H&h0q^7%7ZjD<3S_pEI(EoJq8O=5tqy6soz&ly zDmLN>Z=&MY(;J!W_U0!#=Awztp=?ZH$(2Cc7l_-gOeR-1x28KX-<0>c7j7=(=uzrD zzgD}XHX$_*l3I5eBxS@>Y1;=KT>M&kzs`md=4q%;C*JA28I+tHA&r-!BTV8aBQC}$ z3ISkW)>i7yp(eK}HR&_AtS(ep1I8?Vy_J01Hdg?!hy1BZvdo0o7Js89&cs^w&Sbnh z$2pqci8xzQNXQ}0nW1@?nW1??Jl1}+(F~2VLbLgqUtJRT`{hU}V~H@cGpBi}p8A4k zZiamn!!P*mJS7e{FU<3R%i_0^NVhyIlSW;EOUX!DKVOzCY&SxEeiJvm+9y6W>TH>IWidBh10KeKwyC3*Rd~L#5|vANszFpd6oq9Tz1^p6{ipx1b{b1Ts!Cia^pZ;BtlupcxEIUG|%b~z!B zSwwetEki|9-I-HK=G6&ykJ^07JRJa&_V*^S@fx2Pde^*U>z!9p=#{*OWoj4_@_q;WZE-KTs%fy2rSrwGH9W4wJ3kwIO*uf1#Sr7~m_MYZjRYTfMiVAc7wLrw#9jUM$jTyvS)+c_Eu zhk8pV&E(}2g~|G$t^zh@A%agO!xHffO2kSg8)-=u<0~Z_IGp~ot-$kMmUY)!^Wb%A zI-be<8_vkopnbdZAJI%`X3nAARgiZ+{RQhdHLM8N(lI zuCRo>RYv_z)lZ6I)c3-1W%X3=ligi@yz8PIWNV492t9;dA*L=08OVxP?flG4cMm4} zJCm*H_+YZKmDz~e-1?1ep!P-^x})8l9sj7yz!~QIt&U_3^MlcR`d;cNb^GqF-5-BU96h-t z&-~QO3n8yN#$Vp^GOQfLrNdS*NKJJI=gkm#&uwU~6tHzomp8YBsv7*^N|FcyPwkPu z(e#Eq={)$W$RR3Z;+F@ANP+ZCoRO1}Uyt&_`G) zl;Y?)fx&vkzrL3y$wcI0bpqJRCm0KHUAm3G~!V&7_uBPDwUm zV-#gLAZ$${E+xr=cOLdj;3_8zaZl+jN2VlII&PMf#Gz~6DT%XGKDH8JP}OhPjX4f< zmDh?#>>e%^#c{4sqmJmZF&nl+j#r1^)`WpJ#!(|TY=@{68@BVg$PL^5LUzKyzcE9? z5`Qmh6LVL!;9e?Y37r}87@b3q*I4967TltUnO@vR<`9@>BXd3&y^*=!$nTsaNXc(t z>VBq($ai79`%J!%sXt|q;PK*5o%_6<8;|zY-w4h-d;Xrozfp>b5Mfb7Zr&H%l$rQIG#+gSAc|qFdheoy8gxOtL3||IaV-_;6B7CQ>J#ew$Q|vTjZJAM zqsnX1P&yL1D_Qnq@l5*#nTqskeW`oIXPvyWef#lG7ys9){AUr5mQ`Oc)Xu}ib+%}1 zL>7kSolM3f|0tJ-geUgiUvPo;f3xfV?iSw=?)`ffF=ip%K99K#g%!N=86(1Lo3xS^ z+J@De*O37+v%y*l(nLOLPiyQ_)+k7Mh-zmvSyMVdk$jN*a?+|~>XFD}ILU{62)+rg z+1{7CL6gs9B8!0wjW>E7Az|9_)zO_k*D0i3n7106$SY4dns{mBgyX|eP9j6Jr-6}b zk|D0GQl`nKAMt5WT*6`K@Y-_1gu^V9w=N>gsQiTl*hj=3VQ7#sAR75F@xxRF^jQi? zz{(9ytd-^&;q-=2dEcMCz20{R=ksrL^K|%z0_ZcndF z%0~!GgSW`}{4SnNimypUUGTUybw8HDt%;6|y!1^yp;PS)*#{pDpnZ#W&%YhAQ-|rx zPf4jOvL8^D6c58Qc6@%aJ-H;$gyjE@!Pv~F7UG9S#~eTWfo9pzqkEUz%km$c=7Q%? z{axW}@-eIX!pZ)g4j;2@x?w!qJdlSg_KIur1M$I4p_kD&Y6#DV5D5)N66#(fFGUaC z=R_k(hmz1u`?Qa(R1^`seTVqj$|Vl9pb@q@@ncKk!Ny@l=F=$yOOr%bxoa8=Fxxpn zigCjWCZej6q?fXUpI67=&zhHB{hZIHbx5z1!fZJ@jvL%CI&KCOo6jgy>Bj)Y<_q2$ zOajF*^BP8ek#nVnD1$yVf``rLqTylt1w9z(ANzxZCI0P_s9j95p>N5FF!@Z|IbM8L z;th$~BaN~@ttq4V?s;q|RKc>!2d(MtK=cYeU z4*=^YbO5NWzaBGdRhN;q$-}skY6+8&epv7q!w*A{^roVMooLH5hpH6UPjp@%Zy9n0 zE4~n}c1-+u`2D>I#tJ;z!hxj^oSi zqwS^Rt&>a7A1hUzlxGfWBJlZrnLa7zTk?`A1(ddYYfP$&!6lEs9fCU13t4E*5b+Pfk3{igQfq#M1_pVbSM88F zm}}+S>C0c2Z>W;%|npdtfuCBDmtwV^Lf44-kZ7UdM=rC3o($nl}$R zC04&We;3X~%%&**Vfb}Ap4L4jYjVj2N6~pCXCdAWt8j;`k0sleRbJ=bw@~cMaOSXt zpw_hqe+A+BVUpykblMC8>9NVgpp?y^7&_tkVJ@Yt93shnUlmjE8ec9;6~m1?9xegA zBr(Y0_G~eE6}q>W#IVU$O=xLa*#rw38y|l$y>#S;nD2^n5PR@#RFcFUReHS!OV#cn z?x?cElEgY3T5HK#Y5}WMW-gJz*jpuY3tg@+zBN(8>qt>LZ&7%^jtd~T5Avjmc6H~s zvI@GfddV-jzU&PLsXv`bzVmd7#MsfN*Mk$L#y=H)IhGrAOrW=IkJPf3u1t72mP>fC zAqOqf8gc39iTGU)PP(db(qtmA(4iDT=s{Vo2#oF&i>g^c07k7eE}O2&d&Rrk@1%51 z$OO7MMV< zDb)9KiO>f_)!8CJuK_cwos-;6|MGZeXDr_X9<#9apG^c`8lFsAodj+$pOvEOBtbvT z7rhp4Lm-0I;kmW=Si)>4ix#znI<06`I-M*S6GG_8R1%r{vw zY5fjHP%$}+?@ClZWNwn@V%Zd-qhB6`o`{fXL{-U z!SvwLGwnM||Jj*OxJ;kq$s960t&NdC*h>~+YGz0T&f=ED2baV$0$BA2b}Q2bbHc1s{YwYZkUBQ z=jN=0^H}nZq6>0B=wNzvdopV;omehhp(Xxi_}(jaeHgr0>i*L%Z~yjs8>v2_IS(yK|%2+fRc=0OqXh&fHZ~1f{rxX7c@rtcTF7c>|A}JGrzA( z^Dx2+r*`5$&C^un97BGVbj}*<9P$>*5O4{ze}TB|%4BkNb8EUI)98D%t7zu8E!OJ1 z`5B!WrkbCQGf>gWymTgwVB&v1&xnzY>Ku-dU!B|T?9CJ9+!g6{a&&l5e5%o9OGSYM zgllc#?lk%wtuQ;2IX`#Ra|6sX*2XQ=qt^3taC0F{gTqEtWg3=(7?1dlmUwG(uh^N4 zcjq{n_8%Lak1M3+5J{Ly+cHyW?-K85KbdGM&7pxdk@kUQ@pwx8m}#`fViQ`$f@cz~ z!O=0lf(j#E?jDXRV5q+(zC5uO%F}156&!r4hJYtJhFY2kT)ZyiRig>U^&I~4vEE{$ z2r+|0GDOm(*{546fLSXA&y?BWsJcFrTe(IG81WMGkpOABY*CV_OqTuOa@xJSG*y-W z_V-#>`rwoTu&R=pPm&#up5chIp$puM%V)J>w!%mFq@2kKT+YcbtnQS>0bO~;^MH@0cenc z*>^cqlT6R-=boEfjAW>|=DF7Bvi<$=j9LgVBc6Qir8xh`j* z864p9&NJM9TF3_jlj@*0)0H$u$>zC|M#FrTi_tXy(dF6;SL;1}abvRkd+ZH{bDZ@l z++`m6?@kPA{%-Nk_S+kx=H-ZKrBkG<%V8HRilvtL(eTquHC%LSfA6Y%Uuh+qYRDTL zW)9~?7M7Q%yC2(J?{#`KOuhK(gX7tOp)bse_;^da$u7!UtIR<$qT%6TmHk}~3-b1$ z&}(j}5)}L8t2p5?Lf-8MU)fN&mNGly7=M@1k3`(KQJ9rdbw-W1DGd!TCQ?hMk%hWK zEX*oy7*RgN7J6GdY-q;U8wtn8tfUU!sh&zbLWm-zzjRUlrIrz{_**fK z0tiYvO7TqWs{;(qCfiT16#msu;0`sbDKSlZRMzu`?h)}VEpdwMFlDwpGiaYz+}rb+ zr#kxHo}-`h!h7WH`_Ie&-68*Xzxa^VK&3Akw2dp9Zz_s0F zeT1Py!mq5oG5AvP;*&3u@0ez9-?ro*?vVexUlplz>?b3`x>rBf-y^%%C;JCaO!jxK z9gNj=XYCi&e$*LI?A`hF%2oNj!V7T~{~b;!^Rr+UZr^sHh#tXLk8$;AyJ5^v_(`jY zvJgtEiz!*@S3U!+6rCN1x}fB#PO=L5z!=kCb;K_cv+;o{wC+4LmUn^V|F0)|i~!|l z(`cZx!n2f~iUs4XFW7u83-7UpCqeuI1 zZHWgP6DO6YA7BoXcfTZ&PTn`S7}LQXqQfiganG)qxAj2QSp{! zAwdeT928RQb_rdsP&DK6HoEQ1qjzW^MvRMrDT9W(R1Q}7l4oZu^u_PX9;(FSYbD0oXjMK%Ky;YWYU86?S4 zf?~xT)kZUXOPg4h?01rIPC}X-I!sug$J|Evs-8KmA%F?+)gA+1l{+aN^Uich#z2gFlwB~X7{xZsm+K!$zGCx+Sy; zcC*W;LB(?lMP}k-OdtAf)Vm#!uXci_g1(lW1JxwuscM=UoRqpm%;_Uneu)*k5x~1~ zk+P48kFfWjLIOJRDs=)=&4tUkHl|c7RMlc!vfMXz-oq7FSqFsCrlJlN|EeY4pyDNI z;4d>29iH+%N^Vwps?gNhw0t|92pJRGWTTKd*izzGx>o(xq@qvw@}gzgKJ`M`vOIE9 zSRAsHYePetYW%K=HY9a<%x&@!sZwt76c&+<7Fh9QW-gYFw=h_;CFCoYO?WhS$EGsd zCd5ds#aQyETH^j@o?B!kO^BD0%OK@#Cc{EtgqlQ5KdGeHO# zXrz`-L*aEB0RjKhk~kS_RK-5hIKXt6j*Dvv+Y{_88Y8lN340qc1lAJ52Ey(%GBY~A zHcFN54i?f>qA9SQtmei-KVgw&qDB{9mv$RipIhq`x5W!j-jxk{W{9o&2Wnm7E0(z1 zts$={C7i}pBXKv5D`pQanL)sgTKV8`F%HYjobnx6&wQ$|w37eXoP0>K4F`{AW_FX) zGS}lFf1M2sKmw9`XmL2?WhL~;SQQs2D*mD_AoYLR*pTW32tTV)P)}wXrK|?d9pa{e z|2jLXDbr=Mn!9UZ<(N2na;bf!8)$t*Z?H?0znDcb1#sg``EPHo7A6&U_WL(m z-2BC?FaMCE60r%G_KD!rB3U?*tF7E9cxB-7>c|B%X)qj`N>hT77m-BvwjN%-r0}5Rti%VW)!7(Yso|&8`g%1?mraFrr)9I=06* zCi|NwRw_r8@*SD$_0=@TvP@_4YuE{>VN$9{G1){@nVdSz8ANqCS0yT*uS0M$z3u1P(ML@H9OPxq?MI(R?LmSuuyCSq?sTy&LX!y==7OW zrV^3R`;|bL9o4!(q>m&P3F)Y^5ea=n&XF8+*zAWnhw-yL+R%~bm!#RL&^=J)4>KmG z{x-YiULoK7r#m;r4HwyY&0HKh@G~Qzp#qSp-3TyC&&qEF{MSyTew^K^u@vbBI7!dU zGsvq9P5&?KGO)&i`GKqJ1mS?&<1us;W0y7AHr4ls*;`(*)75LnRDEOF;AhIZM3epx zn}{+|6D#zLMf0B>rn8rb$`jMzok3QeC6uT^gI_yksknzh{-rKMi6Xwiuc@l3D!Re1 zKWLQ|X+f^I7%lyBx+Q*%-8_5t+GPI*v3VBBT4v2!;?~Y$T4$0y`h`W=j(uf9WiAqu zEE`la_W09zx%TP_pg|E6sEh`)=9GXiYxzPQiJ0+k=uxl7Fx?aqkIaulFlz$AS!1ZN z+rnJ@n8Mz@*X+n!@0ez)>>n`?=GTgyrEK)HnO{Z8yv>8T;rfL`FpeZ*OK%?E=DF~= zuP;3ITpP=j7Syxg&K^Qe$?a_TwWp-2p2awdd02M;(J`xl!?^-kJl_(pQk$V?{I23z zrL5_%6q9CmpL&_T9Ao)&?zl&fQE;xzT^9dGOT3y!LM=tASx~>j;ijCabSX_*t0D0N zEpehs_s+%1*4A|Q@@&POv(>Rlf;xFufX8mD660cETbE6c(zHu}nYt}zIpAFe;?OYJa@5CST zD3^G5-b!GyTd&6v@p!ER9X_$!5{+wF6me{%7S}o4tHuja&(s_57mTDi z77_HXR?X{C)d_oSZ+~aSzEYqXTB}r^tnFH!XiEA69v4|FCO1xwxHIPKM+~^;+4!W) zKWT{pBZQdb$3kZuuJgVX#|N_XO2)}hcdDGTh@bIvXRT~~)}=NHYc%=B&LLDwrt?aQ zDa*xZK4pwoJ+M)0mT@>LTrGyDmsWYoT+z@*V?}%^TPJdeJ>R%XNw#c6Z5`aWI?3(_ zgK^4M_cMxY{H{}Xn5*iRo1+l-SfIWe(+=oyIk6sPb-IP-0V?aU?F1DWTqU z=KO8x$c<*FKjfLrp{`$~5HSu=o0cXn>eWm4<+bs&+r+IWZ^^MF?5Y;yaq2XAtOeJy zirNqkkvWxNG_obbPC0m^@=+HbMK5YcQeupt)`lg@$!i4MDd=NpA}ao_XDE+I-eZ%? z?Pd9ocwH*7T3ynzEICD1>y;4+_T(V3CaluIkE{l zs6|65)#sJSM0oumabomUoNfDHdjR8gN%<;jC$#gk`2u}+oQvkNwr&=gck;`clkpy4qlGsrv~XMa2+ zTkY-J%pZ$WsDzJ1Z$n!0)&)pT%THa*VZltoMd}id-PG_V;XB=pORh|lOS zL&H2%7PCkue6FpO|8~<3i=10!Zd7*t0#{)c0zNn6rUQluub!;?Gy+*7(g8zP({#Xm zC?XxOe^=p3Rv6{rNs27rz}{Q!;&yLMVIWK(*pT5h5EpcSCJnI0obVkBCPugBo2%h1 zYIC)JTOSJs(nUZn;_F-Wms@@)iT$U{!cLJqc&vJ8N%=@(9l zUGSX27TzQtZl5M&3&58%A8Z$+3AdfPHM^3smAy1%=Nces|6fbu^(rUxkNjDr(Z9IV zb$j`Nt*y@M;1aU>H|zIm(srooceK8xV2cQ}+?|JL;iRP{{_jT8+u7S{AAfjzZ*TwX z-d3lX8XC(GVRNqX74rU<+Vd_gKCv)5wGKlx6YDp7+1$7JF7c1iT~feZO&2hJ|17JVnUnt$nLzvXjaS>QAw;iL#gK z9d-i0EdD=YJKt~iPpjO0YkFO0@u#QPn|on{Tv+S&(6?q?;4h00H?q+hq(!Q??Fftk z9`uKj*?1AzS`h<8>QXNfAASqFFx{TCkE(w*+0F}rkZ>|TA%&@{{q}vX+&`S%{bObV$V6gol-AUguN2m9)W9P5`Dd?<_a|GYr%yjU*`LfGZu@r6eJ`=}S$F^k z1Zzzu6cGIJCGieU$_Gf$iyuR?E6G&Wsa>WMFWhD`EkEB9wFk%}XqjCaZ;$tN=GHm& z6h|{<1i3hY>>LXNqe#~mfsWj-E|ITYu@S+P?EU9MJLx3UWE&|YV*q_xsQJ<~!jKcL zYRvj#$A5aT^O`UOeg0tY>KP4oC$SNxBxEj-!h_}ZTMWV^vaSbFa4K)0u=XLD}S(!tWq?5u$$p)B*5xn&x5-(rd60hZ0=gwrjJLfL^Cx{c! zg+v^ZoDTeBrUQRkoNC|C?!Y@=ZPtDN-X*aSCt=fhUk|4}id;|Eoivx_7i3{b%iYwX zgyG$m_`Yaql{@Y6BJcb+C$mqsc=H_<)3&8^YDG8hs{aBLz_FZpI_jyptIy*QyXj52 zDs|H5&lRzY{ufKJJ4va7&I0Vb`flRDywbV4eBL#8hyGG)2M(RfLZU3BlY;edZ~M4K zTMQTkNxl`-eW+}0;-`qS#eSxAmz*_rWtCcV$Qw&_ep@N8GyY@C;=#C7q|^qlcEuwT z$h;#SxlyZbc;vcyCp>bOSFV!W3uG~ppI}AY+t`r!Ic9Uk8g#AuojN>)+!^fD{1|h{ zvc{9`dZ(~(UaO`?jN9!_En>t(gdFr`^{6c2MeEP$ayRm>$cS5Xw{>-Xn5DS=#*^v3J-vm?!*fQ}j(v*n)zA=ca8>PqiO zHz@V6p=`?6S*rB5#lMX%?rQz($Sp6MOlTqo?y>Se@LV`C)E7=foFTP0G$cU~3*(Hm z?OKEFaRXF6Bpz%(KnAMH;lnbitX^dk%QUfzFA`liCsgw``qNq3RMZE^F2uf}OP=z) zk4}L_EBuXq)!=_#{;zfu7~1IneH3$aq`+$BGX&dViJ)0K{-;{vUYfv*o?GpJ{SYg% zL*hdFUCf{@Br`P8)i!Ce3_1)wTJp<{1=0?rm?5rZ4@P=)g9OQtZe&LZc}*;PHy*!M z(g;BOEP7wNLx|Rl+caWGf$2Fw;|JDH#&gcNjFrfB>IEAEPk#iCm(bH_-WaA%C?a=f zS171iJVLxr9UbKND@$dYgHe5DhB9bSG`afrA;z+?*cm~tnu3~KSpkS5RhQKW zspb#h(4^Y;SnJ_RRZ}p3sk*GjOZAHQtMKM|etTM+Qa(4{y;7u0bzEf0onT!Cago1< z?FY_crR2=7K5sxlonjZ;S)Ija2jfduR?O2Zb!N+$F39XZLU+IFg6*SE$YncpWnddc z_2Vnq<-{RSZmF=M?C6?2S|TTQ?mi;q6DvqWaF(i>eC?T^?3d+()t9HcSQg@ZGgF48 ztDJ>=R_!{>JA^oSTwaH{B>&eE{~eL=%-n1y?l-t2OS~+6N!>_|BxNp>#hf5b&YE1J z1WS9_GU_36!Wzj`$^=VsWuuK`4m+4NU}PEXi=RMKFo(WkJ9p-G7>p4G+sHKcu?^bc z;JLpqew0&HnK63FVd&E$qrH*9;<!f*n2k4 z?>my{;*P&ms>lrZs)2+hRhH1K_KC5)D?nn$z862uRA-pI2`Ju^OdX0rdJ@3wSjSo}8B{dxw`xmSFC`*Slyr#iv&etD>D zb7I!*eqM3Ru~Z;dBR1?1(cUa%Km4 zPXb!HEi4&Wj4A(pC96&(qlK$sifd)zLrFy{SUoR@fW>9XW(WJzt;yErOr^~$+5S*S z*6}ucZD7bqJspHVv_U#-cH>{5qG!0kZ#AZ*SIERqNhQ0+DEaYaF{B!NNjL2~&gZJi z_7}S8WznaJIbbK|2|8@{*Lz5qH`m5zmoPn|t|!Ri+sniYXV2_Ucdu-!6l8N}JiDTb zg6@mOos;eD>D8J0Lpi>Y<@Tjt+nv0>@^@uCo$PKEztj@fM4zffCkZI5sGjyzY#GRx z)Fr71B2?B!zOIuIMMS$!oUL|Zgo9^5h#f46f7%i&RCVZV$4#N< zX77NKzLzCXD5zB?b)1n3pG;ZGQu6ZOQS2QN-`^7L%iDVw$J?8`>PgQ>dynqP9Q0WJ z?+(}DEPba;KD#bP?Lna3TJ|Vy!61MoiGE3)XeDq>Hd9fS>IV$fYhu{$2dMTb9VGg@ z{6zN*(d5$4lKWai?y6X6zdAs!Q?tt25hPS%^gikK4WXoQ5!*JeKBr>IZ~ihw?c30v z`ip+128OAKQ6d$zSG29V7hj(@T6sjAX+IRuioF-oicAeERlW`}_Pc(@hK8{>Rri+N zOM6?B{grL?jPK@&mGQyBWcOgYx4SZ%Y(Ks7?6t}MjTNtDo0i_Y4TEnLkGIcy8H`>9 zV)S48866o$AB|!3)O0+1Kn`DI($@cRM(VY@X?T2_INyG2G>_l{0H4M?#Bci_zcLBw^__<2$Hm$9TOxP{*8sQ`zwc*q-7xuPo>mnB(IW4@*9hRecx(IdNC80W z3&2-546{sM%`V=XyRRavKwAAbE%8#{27SVC%5>J821u~RXA*dAU?qrS`7*heW1ExE z4LN^lZe?q-JDd2m3s7O@(8!R3>x(}IxV6d5V`R52trphs*tqE1#oOBFc`iy@3y`k5 zV&1D`n;Bi;a_!svYdMYCcP|^>-z(nTerFW#q!l3vd|=5h0h>lWnL+}d^vB3xYGkk} z-rIh6G8uT*grxA#0XwicS7>LR?jB6`cP3j?iQHp;Zn%!vnW<|)TtX4eCS0M2=Ai)e zZ-9zM5sk6FkbMHLZ`*9ERv=Wetx|09=FdSMDTM=fRdpx0mouAUoeubHiL-bOzTJ54 zS@D+kBNWditt=o*wRu&@QuR8d*ch?3ay}n25&`5^J$i11kmHGhCI&@RM7)&%*2Sm% z7TmMB6`nV*GHP6a1{;UiwJ&3JP`5U(!t-WeEzoD% zK%LvO`4nE~dMF`fRz}s;c>przWpDrBizcf4pEvG;P;f%FN~xqW(Wg;3#% zyb&oCZvlW&8;O9txo>kTJa6tX>RjWa?0S~73cxayH}`E`h3Cxx4WQ2&ZrMuat^{!m z?WH!S!uQe|9blacP+-ig%ie6sYeNxgVDlzCLU~sLwHoA<9l2QRL%7ty=0{9_g)2>H zSjTF$xgW!;L;gCjITF)f1J!3je+{ez*br#RHcz5kvfd_;Iv1eAYPXRk2iJ%CxdWR! z;dQR-H>K4;-!&R$*SY+)0O>;g+=0!f@H*E+2`RGz=&{;n*T0dJ2%yd#*c=P5bBXJE zvMr;3cAXo)CLnhRgC5wtE3`LzDll(l1W;f#AF($R*9Ie+&ACDm&9f5d-w^?e^ZAe< zp*nY9^DVs2?aj5$rFZ-ZSZIKwJV{u%XSH@CfClAgon(gyO9O zaSQpY&5xM=3Rjvb{SsLP;uOFfhc-uI`fFVEX{FHIk8ujXz1HSQbW7H|!=cu>u7wBJ zhw9vs&7JT%7hEmuyX-obzZPWGKtFe6^C`T}^-w~pbJ=BV>_PzJ#*LslcVu%cyv`*O z=E=6$$ z4a_Z%Y`%rpx&67;xs4W*uZcksITUXth+`mc9@*Rq&zlDet#cWxL%LnFc@>^F18c?m zEL-PH3ZOc7&E{12URqmk4s|Yl4akClI(NCfPU_p&9U%0mq-tm zZLtg9_%$IZ0Cnz~&AUQ-v!_J!*0}%$Hr}zZ>49&W*S9uQJ*?ZDD-_W@T!#HSB64xo z0VE3bbJuOYh1a=*xz@RquS1*x=X2L>j)m8`m9;?)vddWNhXKZn6M&X#^Qw@gBK26p zI@d5viuwTCP;RwO&#e${vB(>dTk%$cI0o|Ob(>q^dGm0gbuMFdP`5U(!t-WeEzoC8 zX2I4icO^)-z+Sp;b1Hl$*m`G?h9 z<%K^FFlj;pROfEk+zGF9L2^LfWtX@7C4yW7b?%1Er|>%0LsFBB$Ec&ru74vbf%>@{ zHpjy2T;jT(Y%8RkF1!AXrUk;FH*DS&+M7LLux$>?G(`7ykT>$P(<@^ z3HrC;G-7Y&tOM8)ROfEkd<(C0hjXoSDPM;;1M1ujn`7a1Ze?vygTVSWO0f%J;>w_< z+Po@csYpFmt8;^VMp!Uoef41W%e;4d(vvG>IZx<~sDXG3gd{3{4`SJ7S9s<;VpO^& z60tL6&ME-Uuy$=~g{RH{5%F%zRzH&h^^Sqxplx1-@29l|!CKeomc5_S*MKzphY*`K zdBVe#mjqC+hE;Yr;;jR*3;L|hl9)aV*PC{aY3{SgDiEt+M7F6C(`N%ksL^e6-wdn- zu?!ls&6nuLtal4!e%UbG$e6?HK^Ujibt z&8+Z>*Rv$4;x+naSGKV$0^$~50*GMMW?Oj4OC-pXcL5|>3t*SL(X@bj^s3FkLVL8Q zN=sbxjvCjzP*_{F*;gp6dDsMf+?aVpSmUe%Sd#c9Agrz0%nPr2M{})uDPM-pb-(9kRIx8xpt0HvD`c!lomQ)Rn;0u+9f(~h%UY#pSqK@P_-h!B$g+4V zK`cW#^s3FW@Em&WQ0iXJDiEunc5Q}*=gzY|AT>s$U>fOx^W zw`VgeyzcdoLaKWKf~>yT^>8F5VBOoZ*%n^+5*NnuE_Qtzza}7faF6cU3@o%qdkQsg z-3yRlbAzDrygG*;(+^pLS4Ufy)(gNx=L=S&vl(jzWvo=#=`s`fl-WpTLDu75~pS9T% z(`WIj)7TXMnO%Tgt`)(1uEnb8##|}{y}7UnQtSlLWjMSJq+zJ;buLuL>t2uyQ1XUl zcHPTg3t}7U_c}ML<8`lxlqR{4HGFp68%YUR_c~Xu<8?1_;h4OOT@Odo0@l3)i*tqc zXiuT$_4SB7n)+$L&W%}tVa>UKJrveFw1j;;BCK)N0r&>%Ugzd@yzX5m)V;vsM(AC3 zVN6&Rtb3g+)$zKwvIwZdAbadSFmYwrSZ#(CGFGGvtJS?hMk9=vu|Ak(4d_`GvYaRW znjJWzGUBZSh!)JDo!is#9D0LM_X0#%KV|on7^{QYb*@v#b7){O&~M|l%UuZ&EV!pS z_p9T3YK;)E?lsJ^H&^-^5Svil>s+jkhbb=!VZE}$5pNxcU8vtnzE!oSiG!@w%5t5SDkb>)ZG> zAu&MM^w8p5p*`ADsCnyNfCL-dBKB_L+F)37E@TgdHP2GR#i^?E04Sb--G{y4Sg49j|*Ui-S63m$lRn1DqK@fQ{8+Q6XbR%ClPC8)1u_&j;Cs zvMlE!^-yJm(1XR_h-`|t62QCoePCGMxlR3%%lqT4>16j{wsLW@f3Pyyy|lMA*zc^)KEv3P z;zIjeJY(QekcNL3V%la{XP8Wfi{Ez{cApTBwI7XP7hD8l^j|}aTGZ+-5sY#LZlY)R zDI3=FgaL|-PP)$+u{EDFe8KzsI@ozn{L*`6vbt{9Hl%ZZB%Fu>W2EEwQCYk|HF2;3TbOax>QAbtf3 za5fuzAY|gDNaNM1*{d%YV}7%EQ~P0(F^4}7NKyPY#GK8*9%3;Uug7;7#vc=pw%^QW z99kMSS(|kQO;(GT5)`3J;+mN(g8;B}%T7b#ecZWH(}vJAv5=({$9;`(k}v6|GU zaT35i*tw&fVbXfXNLm-WmK^>*pj;zTur6M=nN^@J1_=uLF~2V6F9t{etc#tS+XHp6 zhZa)aW&E07|Hdu}2?45$olD&Vbup15EEnU~#qoyW#jAuyF|fLcvt53@Ojs7Ii=E5c19fp_ zm9QT9^)Gb^K!)I+?A*;BG+Cr}t2K*Fe2y_$&O#8wQ0C=a(;m#c5IV3}98)#%mV%i6 zK>*@$ZfieLfq1U&?QP3;F_K;Zi&?jP(4aXRVy!*&p*g`-#K$4hZ4NcSNDn*R8AIml z#cSIeK{CNF0s;|#3u3`yPp_~nZ12rxE8e|rWC*=abldk<2!;P1K$lUJ_%1(zy%YA% z&R13lf_>s|HrHTANTw}?)!f{@c($| z|MAHG<2C<}*Zn{CbJ_EA+4FPR^K;qrbJ_EA+4FPR^K;qrbJ_EA+4FPR_jB3zbJ_QE z+4pnV_jB3zbJ_QE+4pnV_jB3zb2;#HIq-8i@N+rvb2;#HIq-8i@N+rvb2;#HIq-8i z^m94%b2;>LIrMWm^m94%b2;>LIrMWm^m94%b2;*JIr4Kk@^d-zb2;*JIr4Kk@^d-z zb2;*JIr4M4=I3(F&*hq*%QZijYkn@*{9La2xm@#ex#s6`&Clh!pUZVWm+O8m*Zo|s z`?*~AbGh#4a^276x}VE+KbISRE;syKZuq&}@N>E0=W@f(<%XZj4L_F~el9mQI{qg> zP5?h&1%BQGe%=RuJ^+3`1b#jOe!d3$d>!~X!0#%+?<&CWD!}h5!0#%+?<&CWD!}h5 z!0#%+?<&BrOo+UCml=`&=K#MlC-VP3z^}}T{J#(IEAt}%?*shG%*gxuGBxu59Ng@dzcQcl|31L4%&7do5AZ8Fy)lj+s)ZCT9^BiV9kg%B{v{{IZpr%R?K@B1 zetfxow7qn^b+UD&d%L)0>)^)KNqgzUa<^0a`f+jWWcIt_3&jgupy>Ryo%XW)N2iXM zm4B2kFS)re-Fhxto(cGu26xw%@f_R@m2J5Gt? zt@e@b(IZEW-0yPMf(GWqd&^u&e7(JV_S$x*%4^3*0Q6PEka(;343}`{&+Vh|dK~O|P#!he_r6H`1i?MdIbi8tq%pO~%`& zr`J0;B^`+OFq2P-zvLsJWYIpgQyJV6pY8j;7UJgW%8f#dfJt@@4=*fy!cZChj^b~gM>)<@h_tJ{5Q8m$H&I`$@b*Z zLHk(#?+(gp2hxCF!kT&`Nl4V<*INE~Ahjo*C=1vtwu6)(NZq+O+1i@!UY@O-86P|| z*?+RPseWbU+$7sLI|#Fupi`!sv@S;NL8Y5STBv(ZZBxBV+;LLKE&KU<)IW}?CS=zq zdf(4BPe^-`ZPSD4!Ho`N(ylum4hz4GXsDNK=ryp0eEmq8f*vpmS`))|KcFDrI^EB% z>E}*y`^np6g0z&Y=SMi|S*56_7hKeuU~>+|Pb13d=gPSkR?hkD>7_~TJ5P+aXli+_ zQOl}WX}>z4mhuXi!yL^u@((x~=}|P&w*sFt?SFzOV~{K34X`pOQ;IYdz167bjCiR1 zU_eEbb&=vk2DzW&=&VoC*#KS8Dl|Q(;?EEz4s#{et|+h@l24dwD|7cJIY>a-p(wYT za92juFH9f_{6;FT z5Tkq=#%G^DX1#)arxlAp@rxWyuTwO=!GErVMKT(pjNHW6Ecr2t3rj_A zvAg+(22&7rFLiYQJ$P?8s`e(Axg3qW6YSQFnEN_R-5U2{Zr3?IQLb%N-h5BPsX(1U zY5Y|_V956~c@Y~)<*VqD)*j}~x%ddCag8InxSK z2o&t)Jq7{If`3pH;u|roYk&&B=h)qBO6xWAboO1xY&uZKQ@A(7*Il3sR=k6BEth~tjvoC>;qz1PEd4CSW23+6_wt<0(E!~@g$?w7}K?7>| zeb+?U$HHvvyRf6_!L31Fh|RDD7kz`R!GYgy7B;^Uvj>gJg(`Q?31naY2aN@x%KdKZ zND-AgkwS&YF_i8?ah8R8`3o%Nx*!~^ z+;t;h{!?xQc@nsi_mslqX#ftt^}a$!l=sr0H}<{xMAF31vsCM%ZiwuxYUf2a7L1Mw zEIRiz;1-I`x*ltsm(%nc5xN-!GNUsIaO*ikzw@OY*ENQ=^@wO2BFfmzI|6fJQGpAE zg|fst^!|`-L+EM#8+xD4@qmFGhRvw-%y@H~02V+|(mjbJ! zdm3TG>*$)$qdMRw?`BuYmk36PREfAS?In^npvw+1@~#ueLIAYIWrH1xPFW)eq?=F~6 zf%U=r8fC-#;2yyr65ht{A4ig^?#re3MWnY7Y@qKTZ->PPYViW=gZDKChxfsW;*CY) z*bU-ndT>Fp`ry)sDAwTMayJW`SpDAD2wkXt_w?~@sy_JY3XwuyA6&W%Md>}e4Mf!M zM9PCn&fRFIrLVEMP{!w>CTibrQol23z&xDU2baD;QFcG3eow`bF*LLf-q)xc-UqMT z%g**^qhWR*oUk%x?J<3DY5o-D4*2cwB+OX7Vtw$w#^CTic!i*5-pg(o6IaLl_86}Z zF0GxST4zFSF-&3TYF{I6q3DcsCFAB(BkENK4Y-zBQ%chIDOw((OBy$eiQx>s(Jon^ zh_)f?#b(|S$P|kTRQS9une+&XzMC8qHKEE^=GH}wGl${neT}x^eefRrUB`vt=?ofq zc)D~7in<-7jok}i(2ivvy33E@>3xl};eBv`o~Zqr?byu4QQJ3BmFeLmjIF;xTbqx% zGev;*!3P>)!|UkUicx)VM!$?*jgl#3*3l#=ZQwv@^P`;L+gPSH&w`;QAcXRPgOst8P!{!Q{6LR`4@mQCHDcf%GrDh^db;_}22((=e%Ab&bO(yE z9fWMb++q9nfHc1b=sU{0jn12I;D^=+mtH|pd6W4yrigJ5$NJy{jg--&xvqyI``||3 zjY3|K27;Hf`ry(rC@Od6*cQ$ob{>47aWuRS?iyi9;!&Q??hmt!2=&U}t^bd;S z9c*pEe4>5ufkxTzKDbBlK7b$&Z)3NJBdI}+f!7C@-a@ef=YqDIgGsCpKF}B(-Ulbj zz77@9@;r8fIGP^Z8d!aB=|dE2aB#Vsg-xsvKF|nVsDAhK@z{N{Q2pNI%ye%)eQ@b6 z6s0@pTtxj&lx;A#!TaC?(q(R5IY#aKO)@@@L?9Y44`=qlr7uvF?aaPKBzwtN!_I>b zH7bYq!Fvlk4^CJav-X%ixHNx)a)%BQ$L{W@;R@@64>bmd_rWW4K%Rr{)im^t-A^Wx z#_NMiYp1Bz!QJLXHx`WW^Wa0${cJ{jjq`Gvek0-=(l)IY4m@;K+CD|gJ${FqB~?_# zw%EkbwjmL1Lqy5UylL8wfZ|vr^15WwBPjZAhPJWjM^%sF=|hdS;eGIaqCU9Eyx52{ zg92=SjyVr5or0on2WeX{cc_PJoDJ`T1JaJ#uZ_;xPRyhNcLLUp57H|rDt9op#WIK0 z(L>U`YBgC$^}&s<*|jEleS&wh&Vvt0P}m- z7Nsvx{CW*t(xg)Se2&F8zHgJ}*J|s*U1+rCGJf5BXM-sqcs8pKF5Q8mYzHA*Fn8#_ zO`2Z=^xe$6n{RE0)(0PwaI&F;lP#7xtPegS-NWW;T$c=yeegy>Jh*h&xlLH*-HSdX zp=TopJzF%R=y~vw#?kOTIH>g~PiOas*~`=RQ{KJk(mx2{eYDCy20kzLu=C&}jk4i= za1UjT@ium6IFbmW4Y1CGOK+jrK##UI|I+kGQK09+M;e2}`{2ZtnP{BJ^VkjIXnIi9 z2v#3l`Vhq$99-^ZVH2z0N2L4Q`qVBnn&;264^F0#H^(mBg`#u^or|d7J*`pSj6pET z!TaDN(tU1TIY#aK5s8pT5YWx@RkH>9)JtPd`2pQ7dU7~GUWG-|gNWx&w35fN=esA6W` z5jYfUaiUOIT{7tr6n!^C+tP7~;^}J|ZNvNEgN66O85Cgqb4(vxIt4-9YYx)J#?#Ys zhk5v##@X;bI6zO-eoc$#>^Q(AkvHigy@H~0XMSzo^rLYMJrBM{x>v0x&O`2llh-GB zH>(dW9fG22=hP^>j!wxN)(2lB-K*xMCd#u-CeBXASTx|K&+3CqU!W-4LCO}%8g`G% z8fkv5CTF+{O{=+@?`$vy1kYym!KFJ;l= zaI(cRhxNhNNcXV08rLPmVfDdTONX7?R;aA=;A8in+rjKhqgeegAnqv3sU zQ1s0_o!t;;(}C~htn=W~KPbNLoc>%idsrWQO`~jhAKWAF#(0~_;@A!GNNR8!VD-VJ zw@_@r8Fjlkn8eP5uW1Yp?}HO%&qU*lEwCHJ(e&Wf!0LlbAEH=;gUj74Y+`-zb<%xq zgPXn&PNtC82bbu}n9n4=&B0qFe`wWB0++aE0~3*EI%*_rWW4K%PUx zWY}F{;_8^+GW+1t+9|4aaJR)Ug`unKr2E;7Y}4iC5g`m|9#&i3xRzPys=hqjOx2CSt4^7p1!WpHoOl$ zOwrqV{XHhcg!^Iswna zNf_I@gRw1?OSBKZt`Ro;UbULg54jIcUY}t9tUkDO2#R+*P#U|APRSdZsgdqg^HLM# z+3W<3wKUPMd6^msDO-1tvIX*p&aZ8d=GSThhr7_UnydNF22()rY*rt9orI8WI0)H- zxx+H^4buD?pzkQ}p40pqlSJP9n)C`nFyC0E#TI`4&SehkgKv=TVRKgu_GoE7-YDb+ zX&`txs}C+6gQ9W=JzF?`*uCf*8b`zX;Gou{Jl%A^*o|T~9jN2+&Vx(;peVkNE@@y9 z{;6>`@94eg8yaQ9`{156)8TFG7I7pss6u%6qDya~*nl(Yc5^U^^}#nZ28Z{-iL$R_ zM(7^E?h8kgff|AFO5S;J=|dE27}D169-c7J`hA0RpW84=AKdH&5jC!-GaA_|X4UV~ zT_{Q)1@jJjM!7pb?$Hy68>IW(yqa#d??+@r9&JN3VD853gG*naDBHp17Regc2j9@B z9Nq^XF02nuSQ+!Em_E2Pe~NM)B#zw&Ps0_~2j9>b9Nq`7&;fZ4x>vJ1#l%H1ceVW) z-o5D3+9|4~zB4*6))nIR@=M#(tD6^ZY+l~pyExw7+}+#XSy5@!>Iu>XZAQxJ67z@{ zhV)2zjd0#x>>_TMAZa$n@0YWLFC_Yc5J?2w5J@wWZoTjbz^kf+hYJiA5CrUUo*-{9%q#oZ9yI|FZvX7Hn! z>Q^-chj+<6B2Nd#u^Y&d)Zk_y&j)k1dm1jc>WsqO9Bh6!Y782o!yDy9@yB9y>?UzE zJ-9*qF3TXKA5l=dRqCMJ;YmTDb^R*o&bQHmd*$TygFEuw>Ld#z5UkR$yVV}Q1)h*8 zlV;#`{VM6&H!oeya)5~RE?K*rSJz8_Ao+PeCZS8klU4WgXt>vy9o{h?Ev#csSQ+zY z-7){CJb$L)bgK?d$L^V@;p+paYBfTKH_a=AMENkgb4*+m-JGp~{}Y~eX{g+)gUT(0 zzd)EK-REZHoGw$32-8TPk=Hh7hG`lcx9Y%gi)0SP;#P@R93pCFE*^nVu@)!#HV=zi zrD1Wa)F%?=Ed+sL?yDM$!~5oIiTdUYB#w>BGiYRD?$R+xx!5|tZ_XeZOF{B+0J~e^ zey&kCyl)Or7PWVqRtwwKnN*+-&+MB^&md`igFhCNk}Iopw7%|SlMA z$?J!mfm>hGfVQ3kv@Mi7bl$N?x|Yo=>}LMW&TUvU;NHzk+@wEHyt_xm(dNrJzK8cn zvu`y~!(D4yL#6o+4O2kyZ(iSAx&%qz&S}#H^M~%?J<{wOpz>xu-h2-~yuP{g43gHH z%)T*&j0Z8aZ{8!_=jPHA>`2o*y-{!r(m?QaUf*0g2TALL@X98;;=`;gJ>Y%wo<`a5 zzB#yHGk<5dmDzOQF2U=YOCKTW-a+0L%^=!0?`aGU@0)u>o(_#;HLS_18!|o8HE~{LH_HJc5 z=6#LO;eGQ8A&~#jia2(InYcRUy~hNvZ!RsKq+Kd1H!q&?Fx@BJ|7OJ7I8&#kW)Wcy zX|`6I>dY`L?VqG*2acOB{kRpQSX`fo#UavU=HjNo5`o3BsKBMd>zhljAgP>s?c2P0 z$1(T5#^Uh4`Ff(hxluS9lV(tW?bT7|&81_IG|s}|<|zUEym?=vaCqMw5O~zyO$+R7 zM_`i3?3+u^AZgt};O5N%9>&o6x=*^pttQb!?wga>C-^w8zLriw(zY`I$F8qaa)25oAgHL_;fVu3Z$2R1Be$NUZyw4An%rMbrjR)g zFI|SDcjtU+L|yM`632}P;0>7p=^i<+D5G)!vjs$?M;?(tG+++T?VC$~AnAJ;Q`e{B z$p$gBZ$8kN9o{!zZ`e0yx0MMCVhL~FF1I4!oi~>*K+>;+(=CQUf%eS@8ll7c<`qdp z{(}~A?A9@HQOs$bOWJvTb7}D;?K;zK3*j#irZpai-wS8t+&EK@h|Wm&wAv9TDw+qs zO8X}%+JWO1$R*xS8xXNLMCQy~-0aH{U=@oBRQJq&+JJ_|4I<8=qOMezEy|VMm{81p zsIfS_Z@!VJZyw6LZNTLp-Dc3pJZ~-?gW}{v2Zv*0?hNX&q(sjJpqTqmqi}fN93Uua z?~eC#CKaehF#TM521)A<0>=(xN!dgD=0nnbZZ&!ixo=KhpWp+$zPWS?lD2!{l}+U3 zU;RerA5Ywdr2E{wi$(c24P$2~J}d&61CP=lNcwh8qRy9oJaplEIBE8+CQZ0&jTT}i zm^a_!U<$%s#@xfDOOW(ENRW6mVao&$JXar*X5Rpn&(X)5?|^tyox6wAP_$u_**B(; zaX&}<=0no`Z!SH-jx_4&jBLG8Yz$J#?3+vHAo;q3t<9HyLNr7B=0lCL;eB&Z@lpP6 z!aTdJ%%%f(30~h^`Upw))Z5_Z&%HqV=0lCa;eB(D$Qz?^#sJt2Q{D^dqoLAUUIRH}EJ2Mgyxf7!S&8+LC zKalk8pmy`69}8l43$$-O(wH6IH{WR3H%GGkN{hS8?m!c06TH5;bODlnoq@P{P>sc@ zTxxjVe54UNyl-9+Uz87W1FCfm#GQ@Odb9oeeMa+7iVN*?@tTLo zz@43UOi{WAN%77A+=BW0YEGPB(Ds%HeM9a=L)V@%I=>*!wV#O9IYb2RD(rpRZ9bf}1~o--yNcHI1w>;=3*^BA+}r`Fpbn7$hR>cjBM1?(dWy zLQ=egw=JHmp|{8BdwWUj(AJ82z0vVD}cB=z^Z zi)PI1C*nI%?Q3lAZEWo~hUsXL{~2Q-Pl@-m-xX&d6k5f<-~|F2Vz)Lx&Ch75dw6_) z4{8q@ynA~Rdx*vHzQ8E`ed4M1d(tVrP7tad@hP57&>+1v2hzJ~I7RFIHI42C>wQln z(|2H+wl~C-yF>zldBNSx$;vevvbW|Sdy8cct@qb7t{2P)J>qVb55y!(3?eXZ=Z;cH zry%754yLzI?!Fj}`x?=EAQbmYDObF2Uoa8=&Eie%he;7WxHRU~x(oe_yns$a@zzij zFJIa2^mAcdd=S;H#_t|tX}6gtzr$$wG4W{o&3p|*%VR!#j5$8BM#JpZ9L#P(e4+7r zU88luc&%NCk`JSMG%a_J33m(%nODZF(ZITOXY_5(4B(cGB6aIjqz;j1JsC!IZff=z zu!=YLH(Jyyz~x|(jAm; z!JG=*+cicPz~uqiLmMgT?X(EbcNHd$Ok7^N2TAeHz}(zfz@r&jcdu)7E>L&ZWD2#p zQH?WtaQxDhOd_-HUZ=ru>(1cZ9ElG{G`!EguF6!149`s`^Lvs)V4XW!5W zU7*kIp{+4iXL3A#i#d`U+!AJOhwFC8*_4dLY~>TkxoHUc|R(@OV1tJXW!6>UZBsu+VBv& z+4K3$X2Q~#S4Z{Pr5}(~?4WpyVhruGZ)p53&}XlRILecahWY(y;_{d?bNlSl>`4lC zCgT>w7usjv&}dySUL)M?Q@uZxM-jI~h}+dYhj&q^mC>DmG77Re@#)NN3aIz;NR z*?A0D#UcY$KeO{DU4x``XBuw7{Kjx7TimsoN6XFGGvlk1S^G|L`^nqdOUIYn%g3KTCjLH8 z9VDDe@^2+RH_Cqk)qqBxURluqWV2{~VzSVaMhhpzYufkuv;Zy8?bI}JTpT-jRB1x| zpByD9>mo$3-U$a7nit{nfuuL(H=yiXoNR4PcQ4OY-oLlMwK=i>J>DAcZ0@W)vO7JP zj<=^@C^r`QzrEQ?W{{iX+2+Aycec0RzU9Ha-Pu8R!w7jryt@6WOvtn~x_?>Rl6CFc zFFDng2Cjs=^f|}B?A(zfN0yHq>E14G**ds!b<$osv5e?rN&HAlyn(i)hpmi@1kAj& zVCIff;&`ilq9x~Pl-<8lWAX*! z&Xc#dj~@TB?`&mJ%to6A#`8iUrn_5{=QaxyQ3%fG)1=rIPmcGeE2cnD|m^xONNl?*7#tO@0jY#;{R!h zR}p2kKN)X1@tG*ACKkf*4?zyt*FBLfb-=!U_yOC;R+@l)y?ZDD+rt1WV0-?4bt5JQ zY~Ppi9h_m1@pgZ`+aO?HpIgAb-d%jaz8)8_uM-2dNAX>6*j!i`zfKR>*H@wf_Vomt z&2@r&#G~TYvl6Q5uRK?zl=U$4^wUqj(M@jQyl_|Cl6|ecEdSLhzTDyLZ3Lo?YzpcJ z81?`76Q6iJjrvRCFDdE=E36Xl|X4}(C6Zv1`OIHf%z-5+_iY^cCJaxM+4_YT% zrIdPEOvKBE0!dPjH!uO6jXDwbW!=LZUNwDY=kk+@8|hzJDsi%R`vaC^F(KrD&TXIJjq(^_|#?`i{gZBamR(- zYdf=?;#npeDbNeD$dNkF{#(2`j|aM4o5cg^t9f9Md~?f~`I z|JQ&*ZUM3sF=J?B2Bh|SBz*smd{H?P3LWuu%QE9oewi5Y^38!Ig)PxJ|Liqc{b51M zA3e%}2mvTs;%V`j@G1$T6uwTnF!jSr;w3(Y8pSC@5gk-~=>InhRfPC)h!IW5&cSgl zzULMZtGv2X(c29+95?EEf|m?1$8sjy`rjSHbWDI5*#J`yBzx@-oOCyHIHh_d523ne zo&^~B<&GwiMVOADay-sN&|w}yoi1{?xDsy^N6$aIf6&hU*~y}Ek41<~I#np|%biby zvULf4Z0$osnuN~e8-ttzQ*7l#72oF=g?LQdcCdeKvNgRfL)@+BCgbhX)9X2ZtNFkB zTsC)PKDlNMfxgp-Prl-HCHBQ-aZC0!om=ZKcRsh8QC0DARe!L^mYcTHj@-1BJgrDN z`TDGHoz98Yec)yJ-CU5oEqfv%$!izY)_?xmMHPXfTv5UOT_AS{Zvbw zZpdN@LM=5SporR-`}&Tg7#F|1RBzp3Qmn(=8ct2D^-dBI+qlzMel_$InH5;dX9Z-M zc^}76LLh5n9%@kQON!oDKI4$nUD~F@T>j{i=ySNNf~_nO)vSRY9G!xo-wO7E7B!W5)b(zUehwq)Dc=l6 zl3(ud_fgAXNL%8pIIem_^06?axkzu%?TBOf=W+zZe4EiJaYhz{)IXHNAQIVEFi6&K z*yv~qB80Gt38B_k04MJbv=H?CX)Xl$ZI&jheTSedo+1}$@hM>w77^eQH#ZP+`{8Zx168qa+<5_~jI9Jipgc?i=I*yDU`R8=> zD8Ak4(n^U8Mh$7k2un!VUNJ0zPbE|XqV_tP3Wgx$@sNyI-9@z^s)KHZ2oqnBL#WaZ z6oks)SrhCW8V#=XGcNwvGf3U%kp!KisW?!#ZlJB#i2K@II&FpAFW#7)xDWKF|fDW z1Y>=@F~q)DZLfF@5ncyCSdXZxnuV~+@!?aA*ejLP4VkKEsR}ry71Te!Bwp8q5nWCv z)paLpCOed#r&F^00Ge4UNa!UgO348eGZ{9fA2Y z@&39&jhvz0dKyXxOJ^E2(o?-TYNW<`xf>-z1cnD^Zm`tG0Gd#qV_^zXm4nGSK&1x< za~{;>ARV{samd$GcoJaFP^7a8x*2M4%)z#TLB+_y8JmlggF;nBp-X)QO)dxdZH6wZ zaA}x_sliJn{E-ip&P(nE$iAKiFnv^;jh@)c#U;OO08AKs4|Ris#ZA{Dn7gjM`$RJiAYe4?b+z=< zL4`lw5^Gg!aW2xH9E?|>%U6nmnS#`NKPitV>BE#SI`sm1J;^N@mgT#n1bbsi{Nt9W zwH>_cm0+(@zJYz;R{_PdTtnMlSfFjUp3M0d-@a5Ik@WIYg}7)^=)~%~?wk&=6)zFL zUt$^&AM>D{ehF2OESiLD$cg6)rJw3_o@%pjI=cDB@^Bq`J-)3hbzSVNZP>vMY%uUli#K(PU8d^!A7fM%RmHjo`4MzI5|GHoG}mdh79^RmAu|k z0XscCijvDmo#@vmWjCpG70TC#4D+wb0x$o%krHZ3;&mDm=$9}7XHq33n#}%&Y2W}^ zXy%`D$Ef-m4v=f0tQdynjR~gS9atnldx|a=hQx2_!J(dpF|*p&ex3M{&AjrJic*zt ztqQFKFqKp3qL)ggOxSDRQ z+f$>Yooz>Jv|ry;Z5`H)-PfQ|A=E8g_(9@!#o*evzPPmAroObnp6HAjB38);XtKIYGUoc%6+kokC@-05{Seo$o^NWX=#?_x$DnsxE`&WT!?gcqyMW(_Mes@`(LfDB}!@D{%1PlJ*TVOq%$sG zfXh;Jx2vqaUBB?ktC*eaHbkpoG!xOC&7RW;u9MC$ZQVw&`ZlAq4Wyezyfme)ds3t~ z($$Ugb_a=lw6RMsZ41r0joqGV?9z6aY7ZPG-PX2pmYiMNQl(tj2ioN1*S1BIGqZDb zMy{xBsfA@x+ZG3EvJb6m?Qal`zo=~orP`KZJ6SQhp0>8tr&;n=_dwUSat$I+PHj7^ z*0$0+a*B!H(zR{h?dzJ_mRfqgFRyK7eeAtgs2aj-aocwupg=w0N1ft!)m~atyHcpK zR9@WbjS0?+rlWnkZ zTb+h2a&V`ogk3zl)Rki~o0qo7vswFo@rL&699)G>9dtfMo{drGW7OSAkEdh2G!yNx z3tr8al+}7>qfX6;4=syFlCh|#rRQfXoz&{En?XvmVhvU6N5KA*nT24RaS4<0bU9P9 zRx1*vU6l^(Z3ue5|+srbq8luA`y*W`oVl$9QQoOUbT(pK{p*_sa&G znTN$@o0;c>Rbz?W%zU4&912RQki*@~$^d6dKIXG?spu;AKliaL5x<$$z|hF(dOdbC z^VpVW|CmSY28s`CMpowz@tau(I!rm?!Bc8aL91l=|+tk@_|RlXV_D4=bcNkK&$E>g$=KK6>0q=gkJ< zEmWIm@W`mvojmHR$s>K-E<29GfrxqRQ6Q&9cP3sokabJC#GQ`E$U>P z(^LlB;v}5VoG>=+<{cX(JcWMhAdzY~E;dM%p`@BKmLWw{+I5U4nV#!6(umj~Q3H>n zO!#_Sknq?Okm!UYUAceTn0h*qM+b=-UKA(cIk1Gz`590J9dx6gK9hxy8ae8_gKm8_ z=++?eTA5(>VK_l&*4QI^eRpQ9uV&Whr_n~J=nY%N$X?Gqjn>yUmGsw%A1QDRbRvyD zvR5e5ok;7eiL@w1RuLgok#>SlX7p?@h$?g~uK{{NJbQv~pl8Ce(`oA4W`fVThY00f zPw=T@ZWaR4m%k5A5RIXj;2V|D9|u4}Nsyw+kealo)6`M0jHB$5?~w+6)akSW*hHCg zPp7R_FrS`;$a0F&Pnd<2$0;ympuz0&>9oLnRY9%T%TWy(>J+M>-&G=@IYKT8-iMLZM^~z-agMzWV)@Tcj3ChdA&!%F!G*?jq zHlXYI(y>|Z3T7jMf3ZZvp=xpipC+7_iFDGc!E7bbrAJL9-I!^L|3E~^YO9q{vKl;< zWO6Shq*kL}{0tNqO=ysY*WS}9!nYjwM?@52f3vf8@9b{t07Wz;5R{IwOU!+@l1 zt@f39_!)`ms5s<7I;xXhO_8*!gwhQG#%@(Urb}B@s$Hnrb>FdDm4`h&{x>fZiQB3? zbgooJ|I#VqFt@5QR-2B8L)-MCx2ibQwiY4cwknT3CChby5`oa#uN5$Mvs48})K(P- z-PXiRpuT_((h0IaJqONE9H}&@C-m8v-+!Ouw*z-5PJNpO`&DJ4k6Z^gT(ip?cF@r_ zI*6-h{08o1+aSAB>(ci0>gL58o0qruE{?Z1clY*pR(wB`&&C<81^Ey%I%PPNd-!@F z-(6Ff_`tcJD*UAaoIj5doa)jGUR-z81P#wLA+c^W!CDP)Q*fDA6xs-Py)2M z@t00&uW6iDozxCmy>T^nz;C?@_R2}NoNV*!;V-Mek|Dn6T@>^IzxQhZSr5jUZ_{94 zCr)PLeGI`A<9&LH*c>~TXZ&;i!m==8E5`cPu1c?&|IS?-OP6om&641o6 ztoOGav{(+z7MF#Q&~Uqaj)t$-tMCM9xLqZ>{Qh??llBaMTrsWv2A^Y0m0-sY^_6(G zr~adOE^zU+SPh_tkxEwA)uUk#QE_5-_N#V-Sb;A*Y&tlv6DQm2@Yqm&3$&n1GIQ~IJnD6k;h={0#7Nep(8>76eA1vI=x7^# zNKY0oSrzM!whh#1+j>k2l2tGm`9@fV4|GP2j_&9i(+BR1+Ca^y(MR;0cfc?fiV;0^ z<<|qn!MuUKS8d=FOXx>5GBHF)&ghtqR^!keIaA+`QKKHv!^pR4v;%r_NQVvPP(@xy z%Q?VY2QBG1j=nyA=$>C5s`JYQjzb-^)B}3zAha=s;fWl=89Ij!Fm0b;r-*b7hCn+QLPCb|>Z!t`hgh4n%tgOnc+Ow1qtZMKp|)8xtJK^3 zu8+5`O=55F>xjEH_pTjWy>_sf-3B&m9~Z|?9&IljkH4_bLr`E6&ppz!Ouq2Qvrc`2 zCvUTRD=la+FZK+nhaS%xysX>%JfEzy$hq){^ug*_4DG@r+NblWhm<0Y70(Z^ z;XUeAv?Sd&+su42%R=Yoc(y6;Jf7|C$8Ki6Pgh3xQYs{3R}=A@8G_yhecQ9KBy}+2 zc1XUx&u4}Y%x2TeyQWPY8C{9D_xT8@wF(K70$8snr73#R-v*mm4djg#H$g_G`IAr} z1yVXh=iumwZ*j%j&>e(RAE#aMRyU{WV=6%xHXNY?ZFKBcPxcMnfwrLnjP_99%rrkzv-wWyNscMhPAsa16G< zh!Y*+mccCpa7%}{H#A194si!{XX*m+gb)Y&*Ct!j>zneG$6L=$#@nZ-1z4t!a`##H zCQ4@2^xF`<#ezQFr?2lD7PnXw-=&za*Vx z0L)$$?W7yma(i><#A00{=?U+~XuX#XDZ#9PU?tvn0s7+><WkMlokQ_>VDm)%Fgi%oFrzpLqr{~TisqHwAv#n?hl=#YZ$o#ePJNqpQm|?{ z8YeSdeY^82-$75`&t%0zQ>U2^bkFq-^{s6q>Vusc z+!&z~Z**ixYjWgHys2-;X|f6nVJa0*i`#M>NY#u5B*YSXVHBwRjm0>ppPS4G%AH#w zKqL3`^GKb3w%k|`Mg>p|2^MD|v(M_sK1S4mve%1Svr#-L!=vgCdWMcVs8}FC8+J;t zG&Bzlr_?2w9q*JgOc^qvPaYPeX$VLzLKMwg9Q0?kivGx3FY`rEy_=Q9>HHk#NaNZW zz+o{H>OAFDjc-U#A)jD$BAuNT9QBEG|Aq@}*t_3IhZpV28rE>wjJ>$eeiHySFl(>p z0!K=B`N^s2;&f)zo?4(8_??Bx((VWel-XsT)3mbBFX=BpjDVkG_(dvN0124*^Oksb zU>YYM`CE|{*DQK;P=cWvq&p?7-XnH?626MlS4L_vx0{h*n4mQwzefzJZIhlBgkXgb76P z@0Y~=(GsYlZ1J+}r1XpW=INZSODl6ft`{^0@fR)ehIp}8ptpFjGa5%}-L}zo=fUEj z;lh;K4owZZ5FI+n*l4@+F~8XD4w^$0O}0A^n@zVn&nK&96TjX0KAsO627`gwj*`0A z_4?(vqRDoLpu=GfJsqdCo=M#M8f|wqyf`}UUXS1IJa%?y+Z|B4gg22!+g%Mrju-p9 zJT2$|2312x_2`K57q>L5=FaMk)T~~^tRk#SBSXWJGj&JDziODXeET6s3f-Z*ks7*V zfSe9&6V{2W=oK)g-WPeGIDkIV7srjbvr-uviH^|GQD&oEcZP1HX6T}|tD)wgP^@;Z zSiD(+FM0%#r3%Jk1x$|)o6%Li4tj!9$~w7&gLT~DgKKx5FWIR-87DRnUDFm3J_Vy^;3 z5A=cc0rz3<9o0W@?&*l+R?RAkzoxE-GgCpHlw1bH9R#XO2vZa*`B1-)$pgvX(ngiBw?;kjby}T)*0JMAj z!dRpBy?QF-OrXihO$~j%R&2D_BB4(J$_`3g{F%$Gy~`jvo6!)ye838Wty8Ai@gs>P z-v29g@Ahlib=3cQ?RMJDkd=q^LTmp4DB6kHd)?P+JHE7SLP+9>AQT}US{-`F>f_#R zySu&P4&odWAs+dUj`#-&n``}|M{MX2I{C>Nai$03^ir6f`K0qk|ydFlU-{E zc!LO}Myq@46IrS|l5|U!Qf5L%^$$mV6c5;(qEoK#nw8kKVb5a=|IV5%GJvT{y1h7| z)xS+ttM;a9XhKW5o(_Anq`6pxk61bim_UB%kP0Vr&&M=Tu~dE$Ed5KG4;x?DRrvq6P_bl{}(CX+;OpzzN; zyIB!giJ`RE6&*oK2R1;&7o8INsFV7?tSwN)XP(^yny7j+D_UE^`o)dyY_|bvKgoU| z$z$4Y7?jC@_hShqc^f0iWA3n)Enz132>@k`I^*ldN^>;q!|%A5L94y#M9In+H?Y`y zV~V{!QKG(>iwIZsISj=BlQvYXo0l=pBaQ^1B{mu!A+3Xqu}p{HC(X%8Wu7T3V_ez- zON)@^79r^$A&wc8F)mViq`~YSo`5P#)J7{~+*m@M$QZ%JG3ZksSURG|-O!Pi)-GS| zf{=03xJcF^Km~7MScGnQtqRTTw=ERG3l+GfPV{Tm%Yt8%O?Z@X2-Y{22eeL|w?ZL#0 zJV$j&%p0ENy{2wXU+!Re#UB!4AR2Q$PeVP$w}?1MAw`ikx=lFZ^&g$kA04TiBWzE3 z$QVC1Z+q;EImYsq&qVgxXHL>CSxZfg8g#LGE;~-;VHw}ud!uoXLB1?>0|YO5 z+C*n6|BubXq2R?l`*`$f9$bKK*)B&$Wh3q3P>oRYGM7ce;?Rxl z3$iK2mNftX)^Xi#N3_Y z*$VZ=TvWgk(T%Qu9G*yw3eQ#+5nE}Lj`{3eJ7>=+K#6GX9HYF6&(;hM&IO*CJHKj@W%2Yr#uLWsin6Gbn>DRraRdQez&Qt>6~+%`qzniqc@!2ynS)e1}|UsvHZGvb@~b`mfh{T){%joR`z@$l?xaQyTT-BlYU2;Z8a!k82^3<`L$84d_W!tH9p?gMn+vjPmn~GgY$LS`^~GFVz|yI$E_X3RW=q97p2#dA zks)6UYKhEBs9?&fUCce2Ce*FkWjPQT^CRI1hV9!5s$EtIkujH_8j7SZ%EX@Gf5{P< zZCUL?)djhx219m)&}g^r=C*Zsi}Zk56<;J@x`FzN!IaG*LzF7Ca=P(S-GSlmRgAxq3tl^jXXrju(%$8vh z^^*&xjM(W@tj2`zoKK&3L|)lJWvRl9r^YW-rWt>U@)5VMG6G;19y0F|`Ol0$n^Qx5ISrFS4W)6hMU zy;YTrj=FMHHV?6NvDSnOnm$q7t*T(&n60YtxHLp~TUF`7v@To+*s4k$xVT$YjEf7O zo#dv{VRkR}R+XV*)*4#7rJ=Ea7vf2NF|H*U%_Ta|G$!n=Dr3StHy^?~O+Ppx*H%cF z_kRzQQb)QuK$lnBK5{;q@nP3kYYTOG+fH5Hs^hG$^jfrsSGLd1 z0Q;;sZwuTS2)W$mz2zNgkIqQd|^l6 zx!TizK4`xi?5VX*IkCOkbJgX04a6@De@Cykptg2DJlws%`9}I8519#lCqvX2~ zDTfYA<=Zc+JC;S~Yi_2TpGX4xJRx7CQGGq7JB0*r9!g_bXB9`x@%mV)b9IG=6V|uL z=1?iive|)hT(2r+Gy7nx{&AvymAZJqrb-!DO4CY59*B^9a-zTXS?%r%>c(_i*E<%1 z7QETX&#F)CFD03N^5(PGu1uev9$lN+2TXI5MMM>< z1tC8?Qm>1QqGb(-35iNCW}f@Qh9flq8Sciu2wzvDJTJ$DUPopq%rL6W=!q#W~w`U{8}L$|ii z^!%>Hshg9RJVjcm#dOw9kt#eEGa%n}1Bx|SPvEGb1fH>rI*N){gUg&%YYY{0Lu_g& zfM=NM;3;1HEn{j8oua0GuXc;BRj!Y^IQXUy;jLOvm-P^?|5!~YtQA^MU5TM0mdPHe zKb}++xAos{@lpx%#$>SHs}&LVm}Ilp@}0ds56Uvv&lQ)wu15}8tN-vR;qW|lMLv{W z-z#sGzH~;nn0)Z)^!)yBynW~V!RbBIdGglTY2BCzj_N^A9d{iyrAHc za7ao})gK(Gb*Hr{9_{s3rMj-}-h682dl$f>5`HQtvF+gbn~5s7i9xJOm}sv%!|_PM z@V1Hln@E8+r0YlLXSbu1@wQeq!FIP{Y$7E?BNQs})=zd5DVW6DL<$c|gP>9^P@OuZ zM~ZJE1(9xh`DSb)G57o_-b8XZ%Gg9Q^4YoSUG{Dwg;=TNOwASWz?;vd>fb~%+}Z0+ z$xY;vFh3mg0SZ%+nk{eP~PD;m~Zb&tJ)9-RTU_2uy?Htk+dA3~wM7SF3 z^m9*^*MQWLIT7~o(~59+w(;WNHT~=WKhf)GZSvuAFelVr!|n{bt&NUdd!GB56hyci zh(Xu0o@ZDVx(0n3=yNS+p;m;uj{7uh1{P|5#J_#}B^4CmZWAE7w#^#W=*70F@83{A zbD_WA%+ApMHFOp3L-mY(-Y}!W(~qAcXMkh{?qH2x9{0R?ZI%8(fxA6+Z_{?J9|w)E z?O)lCT0w)mM(?yuefy4O#oK`u_u22>9Nf;-UR1sffU0hS<9bbgit2ewinnrymOy(? zp7-c(rcS5Nb)h@;_t3J?nxpdsVlnahOdzwXVc_N{z;xu(ToQXxr8AA5o|iurEb;ecm}98z1>R1Nd&VJ1R>gHZ1B-C#rm_ zcmyAy2NP=7=fN=f$6Jfq)jOFBX|kKlW8YM&uV(LLj=@TFxW-&}h?$zBlYA}dp3E6a z=?cZmzR6sOkAqI;+RlcJT=!(o=t-XkjEJ8H-_X13Ea~2q2H5zjsbd?l~i@KFYet2&6ig2`DH<-(J@M8vV@LR;4=s9ozpx zZCL1?05nfI*Qdi0Kay7TWSyX&-ZzpdEJ1zob>$URS|QTUNl*#YRUb~uI}5Ger(Y-X zOD*VtZR&3sZ=w?05#Q;q&HObWGb*u96EmH? zKFqxU@u&U}CdQ$U;m^^2(kA@7e0vQ!>B-KgO{eO)>8Cr=Ce9&Qx9RhuuvI7_?%FL~ zp5N#jjJ2U0dl0SQ0n^$km=|h?yUcUxJ>FY%tOcx7k**^x-vMD$oN(_2>4H@)+O&1A zH`I2zZiZg*@mSVQPpS=?Da$=E)1n-|ke16BCTf5w{HYkVY16)bkA@vY#yIqx#+mqqJkN0ei8q#F0i1~+^8guZBxiX6 zJg_Ey#s#{=^s;~A7vkn8JMk-9C2QhmoS@&MxHJ5*S|ln%;snGAd6r4PkMtWZakAZPH0;(acFQQ=`L{~7M0RnA-2s6;%|8&N7Yz;B$f(~XtO*veSXp71$D%iLju{#Y0 zQh|Il>CC{7rqv2qCm69n_h@ptfiAb`=LTq5QGpxG&=2VS++b!KKn)HvPY>LnHzsrd zn$A=m1a3^0E)}Tlm{D)K`7@Y{hMrC*7(0l6 zcHjps`>;Iwvp&OHsM++iGs7zb2K6B{EL6>P@1%|+0I`+CrdeSF__m!pF*Y;Q&vl#W zctT^7$loCBtP*aiouYINpKZAt% z27f%l3endYY$Y!o`cH(M(aufOpB$<3<;odXfR?6Uf#IxUk?4`~R1pz$PW`|w%o?y^ zOpZiY!0>9A`$8dvMXsrX?jOZMp6J#9`kWE54n+~B`$+!)#DpR395YfaBxH2DD)82T zj4?2r#aIIh(Ss61A)ocH54<%X;|vV1;>(sh*2=7$m!34tfHHxepl1X^y>Cgw%r7f^1f8Ledv8&JHF&#P|Mu!)K{ z=4S(IFhAmB09^^8MqxfnSi{EFZER?rpMpkVR@5lWf#w^FUf0P+zt5gEob?dtX0&_r3E6k5At^e{lc&-s$Z-w=Zs8SnOV*J`I}k6#BrNh;elzea5uBkqmv#ylJS>q zRfI{#-^^l=`@dlC$_va!Y-SnUWv;H5+|8^ITTf^+t7RsK+{`lS%XN2qN*rRl0}xn9 zZuYH-qrj2gY-aRkFL2LihL3bOD3EO+J@7?2onz4nheQal$@rG423lu;%H zEu(q5SEcd7B`R^8Icb`v&P^Nr^5VPxXum>c+;R1$FK*k&UcI?-6?+C{-onS~Wo?bK$vVzJ2zb8lPv|3OPb@gwt0cl<+nz@})PC1ukWHys2slPtz;c~#ue3@dk@cV>1Ft>yZ7Im9(`(}bVQXdmM0jQhy-~aK7?a17qAzJ0tj$l>ZFDgl>tJo zi8nGcL&_SA7&ZM=ED||PDs(I2E)p4B;(3j?NEG5JKS_bFByvI4zizM>iHs-lJS?_o z;?0fxTI7k#997_Dn!zfE*BOvbd~`+?De%nGk4G%Ns7Rq;3{cXI!ADL`IFFQcbEBlA zS1!&$&VsJjPB|7;F3uNLPB;f?Jo*@qJWz35CL3&zXe5rC8*vHLua zi{`zN16ri#k<(wjIJergK~g(6P10T0qwctfsu$-g7Z+6BnX6vZj)ZimQS0P&jZRT0 zTfMl6su$;T3x*n3;{|0#q$pgZOvqL*ZhEU1=N5s-U61BeLGQbkc<3e25RcN`X4s(< zhzjQ$V*LIU1OTC%&d<}qy~!vn=)m*-S3Kz0sMU1Q2|CQrR4@m+5c~id7)mr*yR73I z_$7H#$B-J`pqm>CH(BO7tWoP`jlD>IPy@(I%0rjxM-%l`N~ped&K4PRUpD*6hH?;Q zFSsg0cfk6Rl+lqUL(bHE3#7{nKR%){s?5PTQ?PoBLxsq(>@X1Mq1#p2$NkJ{I$2bA z@83Cp|JI|&@0?GM&aSABh{s6bS0c5|J3JFQuTJ{v5_X*A-lF_Vns6atdRoGa{mU>7|K*-&y zN=6@Ks~RUBwu4yS(?^A>xn z;t}0!h2U*fg?Qp_B2Q$C+b0;Tt%`?rGdGLvQ0Q%cN!3bVyFg7xSGQ!$SAF0lg7ZjD zH#c%RmUl1q2Z6AUw|8F;H&I1gzd&#v3F78P5Z6I?0T^nVUNk|g3kj%ULuM#_HFQr>b`j}!PUB5m8kXxlpQPvdqim!-MZaNDTX zZ{b?QEkNZO%?YShjShWuRyvGlq*YrOtr}xIE%CuD@PiE-2I}gae%RBiiClnUX*vnb z_>Xdz8ZM$*Y4h~Jd8qGZWuIGDB5B9MAs(s%%**?h5#6Eo8+2et_EN*GtwprpX=%ZZ zxzvyrAYV{{RIE~_V=pz_QWcQyQp1JC*NME3I{)6b%t8;CVHTw>^O?NQrzSw$x1e4e zp?g02I>t@tHussH)2u}m_)U2)`Lu7P7SmaG+Q-=I)jvYxLXjq8U$wHn?NaDST4NAs z3mn5d81H?Z&yx&G%g z(n?60n^SDMu#ehmiDY6=t0mF{GsOg{Zy@c_iDp7_Kr;DWo5qv-1{REq*jSc_9!&Em z=rMAurGoV-EhE*77oJ0FwN&`Se5$j10~VfM*@hX6Ipp2szJZ1JN#%ZAcl*3S)LBip zkoyMarWbffx;{C^CO0+YWPkk4tYmn!nMDAumWoOGq`c!P{$^G%aolEBcwibVw7!9~ zI2<;!X7nFo(iKF4BQ~>`?B?H81Xqhfk+#TQIbt)*STA#RCH4&@#nvGf4hR)h(9vnI?L*eTNgW73|uHV^AJWpMcp#G`g&Nm<@4x_{#EQQy$)Yr4Tt#7`RR$vaJ!{_UdjgOqYu# z+BJhN40L@1m#%Aue>$ySUNpa$(!iY+BnRNWw$i#~m$c6H4P4sKgFza&>srKI-$4Bd zf!sGR^MkVhe2i%Zj`=hkczft7N9r30-@r2W^$E8(AbgnInFl`t}G%c86{POuHS8-{#V%;A_eAD>oL%Ey0GsgZi?nRFJFq%zi>hEJ$q$ zmi>N&L?b-W`rwmj_}F66?3GS3GMlLl$$bN}ho=yfq>#;?LMtH4{;U8%voB`>J_>r2 z@hywmH!wFckq+`a?9VZRUce%ny+{=Jm4lm7iv(RIi0TW-eFM`WjZsrJ8AnNf#fCSB zRP03}gG;OszT_R^O9V5BEfy+pV zXP!_<;MKyv;*B@>PnzhrM zz|zQNFv;}7A0S7hky{##9I0419Z{~ZP#^F6VNti!K#6s9cBF@qfk~PITBk)7)=N`i z?fV8&-^~gOcdoe7j)p@-0Pmo2;e~~3NQZWSFrj~L!O(7T`q-TYasrzHpejhynkIuu zceakv3Jcc*7&1Q+Rzmc(ps;Y0FrnRSsCC-`?twnc2@97NfhWSk4e^~^-#|TK$i!@5yxzICF7&*PS)h| zP8QR_N@6%_yDOMEcDpM)G7TI4c2|0Mw=7Y1=&}t_joI#EUVPw3dcNsxu zue+E0?XD1KPi(uZT;bN3!(sFHo<+S0AMTFQBM0I6MN zb2RJOo>EMoy{WEErca$bopa|f%8}M?WwdteQMT~emGZZJrGQ*fbG3c~-8CNGho~N{ zU#_@vdbB+|UYT2*JcjOiL91@vMMdr_eKDZR-?!K{9+LXI@L=KcIZc{XEYu7;&0hLhAYX`zSFJMIIzF0 zknV9tOC~56vF)K>0^4>+`SDf)k~Z*9#2Nv7z9*td0!D@&HCn{Lj|fQ^bqmX`#RI%S zgjXdkOBrZ>IcbbHu&-1FVyph)sE_Rdn^U05^4k}oX(k?+;J7vj;~DZxaUmN3yNK2 zxRTK!Al$twE2<*S%N@57U84TE_K?vh&d$zc^p3x%p0UrHUU}utontF_+#ZLYK@UED zy;NTWtrT}DR?=>=KjZ%ZeXQmLw3u0eSzwq2^f zgZLoy;cL}*Wa&P77Do|A4lm}@Of-9y_z+1 zcPxJdp)6-|je%fsTQXV8TU(dcTJyem`|U3P_dIPoExtuWM}(`lq#c%KLF%I;b#sL6DJ2=>$L1}ReK8ld9BC(t zpcA72ABbS+ssEqOBm}RCAbLfFK-&}$GReHdR%hJeYCL=6kJb$QVhDci0s4|95T7KIJ ze%6sR#ytLb^okyQge_velhhi>TXNFSk6YG}y2U)b2e*JjE&BGlna_|HM#ZIT*Jo-? ztpar5U=0K+%sNQ}WmFj1NTFgLKW1dP&_!2T^+T5(z92cpNQfX*z0nU{)=s^#Cnalh ze~X7M-SH6>nXVU>$E`u%B{a@nR2FRPEOoJAH>dk(h*MBzI!4gru}fTMx^vitUTFHI zfL-*Zy8c#DH?16Z=`A%~L$Qiiz$~Z;5i zej;ia<$Zk3a7VfmX;+S2aSGY8BQ(OVN~bWWD^0ujF#{DdnkW;&jxXk?)^vS z4=&E{+`awid|V$fkY`-cQ;K)aA5O2St2dsW9$h=4&x_W(Va>*#)y=iO zQ~R*rdT_2BiZ3fA_jqglP@(3VepTtmN9uDzV`>|z94bUA-@04gsSQK3?DaZh5pKPV zPE^(1`*+UYzxC+xJLl7*vn%SOi7K~RY&|O?f^xx><<|Pxv7dd@XK$$Km1`edHUDuv zu4QYLe_k(O0rlZg1yHmsPlZ}M|5aLUH-9|4=iv}>QeK@&Zcgf4%&J{J4mB-JupiIO zhwx6*5A3a~09(>#P7eD$@vRDN6gFn7Dj6O1{*qc=-c!u z`ZP=OUl-dZe8G`o+d^lkp4>QD-)7z(S^smvA3)oe(>!`DWf!a^epE zgyvnExiMgI*I+neyb;p(xw}R;k-pD7-Rt{wv^0Yx46u)usOWa%M1C6}@@qa-y=s+y z+OZKk29;&)hj@Stt{crmjt-E&a;u!)t?HxI^#Z_1G40>T4gbXaVT)&>_q`{dm!dOiU)_qoo~Kdchbl`b<;tS9DEl zkT2?naMsA4=Y7tZV!s^jYf;@#^gQU&sv`A5Y&bi|56%E^NlSZdzYDQ*=;>6xg`>Kk zS-ehJyD_XxzI3DdBIR?KIsXK>9?EW6rx{Px+a22WbS6ID0ePrY@>kJF#zu!ST$a%u zrsEflN|{uXOSI$pA1CTpscQ&qSmUd>a4g0%TROP(<`x8apbwGSZv^*_rPG~mry|Dzc#f(kSyZZJK9o+(3Q$Eh|*Rf1s z9y&azwGO7^{tx&f3g44 z{Dr-D2ba)HLG``;CF%NpO=yad_kQ{}?%jRo*4G~2dh_1fU%P$p)_v1}sPzB4wV;)f zUsq1VI*UruH=3#q9z^@ag+qK8P?k_P!`aFYXe89#7@;m>KvQOo4(QjFE8JdademKs zu3e8k-ba;p*%rN!CDSlhb-S+ppYB)>0>?Z(kZS6?`LKv8(KkCA7F$$A55mH0o97Xq zo>AY!q3Ic@M9;HuI+}Lc=2=vUzL;AyZGMJ0dQjWENU8KpfBV^Mn`g92^y|I@;MRVK zi#q@=sE>KuGFNS$VXm{Tyah;_x2TvJ4Deq-3>mt_eWvG!Eoxm9G{kpxy;^Mp{E_ThU$8Ttq!c}NuC{c04tWUm2(DD z->cpAYvu5xt~VxYy>9x5-m3Li22|$~`98%(i^9aLYiCLKRQBotarmHIc4f=Rqhr0}3L2rAVE)u~f@+nFSn?2l@3Ckl6M;kSIyIVK65d6ONYNWh?f&Eg z*K<;#9uQ8#JBxcjEF32_Js>*52TFLC87y^;AZ9d~XIMHfjD&ZKw#=bRcpqBb>(+J5 z@lQW~j+`0p#k(Dj@#0-uIuOLW+9~!0dK~SoD2mf@U$~T}#U;=tDA$olt6o_}s9DhXXp3H?5`pHjbwDV{8WR9T= zfRBwO=>XHd0B2HRT>oUwFiTe`UiMAqLVQ$?4rqL4`;XxKzf`@GIU_559`v(q^sFXc zda?y-tSFvq3f=shq@gZKX9Yt}w~a(KMvcW-41*i6p$}9Dh^g}TqmvYOsjTif2bDoa zOw}PNTjIwPpNo3!kan3VNOTT#{{<9klDx1(^-*+w;)YVUq61zqPwyK_OHZ(O_v5U9 zm`vBJ^@9_t(;a-K&3L~jG!@e^}zcXwL7x`u|u^+SA#FT zm~0u75^AGO2TA6_x5>_J8_LASqhwp z$*7G*?^*_+DVJdYpC)EJKt{Ghb$(!I|1hM;i?h%y@#kniX;Xi^E8@8Cr= zC&un=PGBzQUz~*^;x!G8GUH6Nxo)ONaVhH3Fd&`kgDC#tLq#Y~H`eY!P;1VP5Y zHXBfXbX2L^!|Y76{(es#OT-Oh$;qcHNEk$%uKsp{i=>LNm|oc0?@^&SflfPrdgV;~ zLdNHafP~}GT)zW>aVCCD24t*h&t1AvCh8M`AZy}hgrGZ2FZ(BcA#M&ni&?uNXX0nX zpx>k6^ehv2*b3m2?j0&$pi42nVadTRk}sI2AEQds`XY+v5Z+h85L9>2PY-sH?qFwh z2OacaMSOcl59U!p+RnAd+I1(ysEB}05BzXOB7&U}5p)lyB?xZMK|KQrg08RWCkWJR z&l3dndpd(4xbNvh)^Bd0@u~~zyZNAq%F%W!M-YH;0G;EWAgm&SFsD^ZUNuu|U)D7j zE=QZkqdKB6 z$IH+pc3Tc{i0upocinb6vY;n+_8A2Xq=cQU^OUZde2o^A@Ltcn{=;P%ultC6UhSrv zj);WM$EX4QL~gnmpr1y=KG0oDvDSb>#^}}nI+PO`ClpHX)_|Bb(Cx_)fXo1r zqzJ4vAfpZa(R#^P0}9b|_zBJ0AX#fbMjRMkMdz&W0&y8qb)Z(rKlIa?!~2doL~4bd zQ7dqlRGLGACBl-0T_jo9bxIbfd7;lQqq>isyUKCb@rBuBx$`q2$$}kE{X8I9*gu^y zo?Pmn&*A;lftn#&>M%c()h+1fMIMWE3*Aux%RIFA2kM}If;cxkby&yLfl5C-bpXX3 z(e6msG?`90$~D26$Hg6MM;&+*j?;MEHvM^hGhh%6078vT}f_OKEj4534<`j3hFwZfE`pWc4-==}bpyKmn= zeRzKF>!09Ry?w{W~y?y8Q#jT5+K%}+nE*tqJAIC4N7pE`O;+XQIwI(;Z z!HsEh1IYw>52(`h#9x*(JZfklX>NW1q<6>p-7D(zVkX)K2ZsxmW9i`t3%2Bdbag=3 zUwL`e+IKJFOa%B8*KkMDc+7s3AR+7!*%yzW?tKL{$^HyA+^m69JbLN=0q+S zs$F^Pt~X*cD;XYbX0>!e(mKK>r}&#$!NhT!S>b_cuy8lC(xZp&Jj8 zVx%h>K-njvLv4&BHnR-#GFR72?q*hqttYgZ)u?YXrWJoP%lI$X-G`Z1ZPPsr0}`~_ zVw^9a875ZWnhZR6eg-vY`d^v{zra188A1{q)M1p69L=@`nx8>An*NuzsY8y|zUd59 zr?|F2bK9k3U(u%B=s+MSgzE5zkT5L_GE}C`&x|sSImnh2H*N7@0Q&=J0M?I-0orT+ zs{7V3KNC`_+5SNKc@Z;?lxkfs=;H-eKw#A877M`i3h$Hx412D?Yuvv27#l546oVvb*AQzUz6WUQ93c-(t`! zCg}Zvesky#G(WSFxbXDj(*1*Y^3}hH4voncw$w04@$@$;Yi2tPy!({x6Er`w9|}RA zparNhY>(XgmbL5W7}n17gfpQlq#>}?cb{GzxIvNjqhn#w`<9!1=Lf35&re?U7<8vL z(-jMY8er=oyuPQyS;{=!coM%s-(2pjUc>S2;ED_nMU+g#J($NY{FF1Ce+Jg1y3|O# z634LRO0~}r@gCFmPY9_5|7D^&77Jh^jz_azAu$=nzT=tckdzXgh4+*_XiqRQkq+`a>^E1rph5Koh5)h_i2@XG zU+tvE4;2T}XQUZY)?mcgDLUw4A#}(UfFvg5E)p4N;(3j?NEG7fi7XPe?G+ng_9Bt7 zC7y@T87;qpc!lBo4CwO4ASImz4GK4?%iBNwc*NpMbBj;>XQ;fcyRvbv(}&OzNdK!3 zC4sC34Lv6^T)@gk?K(>ZW+3lEpW`Uu43*5$D;wRAnkxYH*dG7Pet^x>c|*$88ESb^ zKR80~Cjp7NMwE{EO}_i!~RISln*Ujb)&u|4||ILE#-8E<`zH`@rllm zs`h=?vJbr!8un4TDh)e)0;wVKn!u~-uOJEtZFhd24)aZBVUfpPu2aM@kJjQ}fS>bkX6*N3=vVb8t=#L85X>#^!5lcIifI`qYspvy}NTh0+ zuUjcn}jx9Tf`t!1fO1 zmpArS6%$>-FE#m8N;jG8azK{kp1oD^*luPPZ>uWAQ(=Gb!z;E>SJH=eq_MXu9^lQ~ zEa-ubt;*9aRAZM31Z%1v^??%(&LWZB%!uqp90Wp~G;Is47nRA;S+A1~&LWlE%&6o# zWP`hQ1bQR=ZGsC}RX(5DYvSfJQ2*QE4~KRp0oUbIBj=CieDrmP|e?AAI!R1*7g7=0q-9OiORNbnt;jIs;N$wDTeqCv%54q|L1 z<7$eMFFp{X&dZk}=5mmMB1X_&a+ZTa3_XG6pthG{&SowL85`nwR&17F1_t=xkdVHG z?0q+*^W=?=gmfme&v4CUXZpWbm6&<@F?)fft&koC%~8P{9R=Cz5a;HK0bq&|ywOpR z&H!8{ikjhXzgtAbfOFSHPPgLj98jQO>tjCAqNMzQMhdpMQLr%v)Y2cCDdAE?odouS zAF}KH902vy`5d4#AA2d{+_dWlfn&<49myiUn^%Ln3d&)2%`NUBcj|YdKkdnQLSkzq5r1$&VKlh?P%^ZfC6U@&LCk_z(EoT|uahD=$hhz5n z-}0irxkcKs=x<4#hHqPXq1O&0`YW}j&qRMdMFEn(IrSnF-6uLE^r11`XL?Sx7Fx_s zUh*m4N-d_d?i8;K9qPlGH59?f+X(H@6^s$v8p=2swt7SX zCb`tvwm-+p&w5XhdkkLQzW3fax5wbI`sA&l)Bde}OSFg6YwGHar>95Pj@X@;3NO?n z10=U)D*R5a3ObDx6x_np!cP^-`ss2%YFnnl@8%}5haVZ`zCBHaWhFzWbCZ0OC+~tDp;R#Gg3i$;W@OnOoczp=Rv-R0)8gx z-tzk19CdPArjoS@Oyjzn@X6&6b92OIRv;tk`s8wRxiOX#15z92=3`zm{6L%8I6)wP zGb@-lZZj)9Fbx)3kHL=N@ad7@fr+IMBsa`07#D43IZhQdjkr5SsHMGbjo8dG>dRbR zi9H4j?*EBxW;FtQ$jvN6!CZGAX6Cg`_e>1c)zP`IZ$CGWgmrUr`?)-#F7{1{+q1^I zFc4xOj>|zGoe(>*pt&IffIg#*>mbB4|AK8>yjW19u7R5kJka%wgLi0j3^FQqqnC(> zL5|dJbNzT)kHLJ9$H)b>j~LXrYdHKgL4{y+bnN7Ub0-8FhSvzeJUX7W!^B5ah?l5n ze6D-t*(dw|Y5S?l$py7zAw*t+8&F6=x0eeph04dBTyPnck1uDAB%AW_=399$C@2@4 zr#M+Zb&xtO!40|1hnsch!U+22pu0wtP1E7I(G)7zW6-o!)Wdijo=)qm7frLhV6bUJ zaL^l8p1HJg8w#VR7M>@I3VCD&Hg6?kE;S$(m za2^`am2MQupg(*A%ir zWiN(?E4jyD`q%=?a-b6(TdDCwwS*CzF+<84j96OuQ@SSPT1ECEk%1}0~J|>*o#EQmUtdUXSDnZ5_$|SfIg3z1sU#SvLKx=+?tz4dHXEL{!8b|q{_r} zF|s_#e;dLE)}y2aklBqvNlt9Ih-7x=={+ck+5*n0%9c?Pq3Mt|_z)!YG%ujaIXXb{ zLmDaPbO$)*kf!tqH3rBHbq?2EX9_6@`a6ypEkG4E9ROr3NjbTpsUHUOz!YJ7W<(uO z9~iFoa)<|ag~@`dM)PvRc}$8r?t2~}(dP8AI}tF~idb%Fs~n#hJwT;bKL6bJ$~0Lw$z=NOh9o zr!DeWpm{L23Nrjm`J9eg8ui@Mgx?v++HF1QSfPeWooj2H1C4ChQ&C@GG|FBNIMz?Z zUe4J}{r5!udP5yM$OcdEoZo-QmLGo6YiVz)m!~gMw6yh~?GYJjI#Lxrst+gXmnDsl zR2#A;=f^6#2=Vu>s4obbX_Xs?44Y&5VF^2%?7(oui2pNDJNAG?ysZhV&B~KPG+75^ z$giwHj6uS`AE}qw9!Los-_~m5e{wb~jF@IA*&mm^pU4D?Bm{8~%1zdibEcmkmM2 zF6TE~HD+GQiAt zJ0WJ=1|{pb0BBLIA6>xsJFeH^A`<&8jM(o*{&{Y%zVR+-kE@4=sAR8S!nlYOdkdr3 z>yR)mI>tFj(xOWD#oEE=t~-~#KTrW4oj`|SjudzcqrhVibBdwc!yI}ZYNW?t(Y^At z0rLWR+87}1M2rh3PxDU?K)UM%-9^dEs0e;nvZk}q4zxeE!*?Z zbO6x9Xhn?X>BlWqRVh*9${1HxNF8L1x+SDFwC!m>DE_KsBg{7D02$+|OH%wz+c`jg z%`qsS>kDT@mst&-z?Y1I|9D$(dhGPH{CGNXoc;!7g-H}Lsx;&GJ$>C z@sJEqFDXgq~*c8{(enyM>0=3$)iY2NU zq$4SvLFr6GXJnCP&^*0aZCME;mIP84@*xQ!9Yr|_q!BUg6T-nbBmV0^QJ`cD9So;* zTSbMDOV?Nr7Wkph@q-G~e9=jejs^A?Zumis)O<^$=6iv6WVQ4Ov%*gX1w@d#%Bf!j zNfiisx1sgrnb@&<6^#=?8bvDz9#dZKMUd2Y^P%A?N{2vn7qTot$9K%B%8MYE5ouZh z_J#hv_R#>DmO%uWi(xXDRQ1s>+RDw9t=zOC$fbi!`e)EXL5LD4f?TD1X{G0oeQ*C4 z=ySaYa%s_eB7&s)`=KQ#Is_;Mp!pgIWROd$QuXfuM9@lfmq)j1KeHdVzy#L^GHO4w zAFALIG!pe`(329aP6Y~_=p~&k#^~K_!*mudD zK6B#}(<|3LxJqA7&D=^>wvo$B##KfH`{kHrrVw){(nC>PE$|tN&|xDz8?nq}oMnvX zLmD9MGi;;v5}+ljm7(*N-*yt9Wh9m{k3SyaY94e!^)3wP63JbbM()x9U8otF&#a@m z)1~W1wQMwHcN-HRZ(V7wj z+x3$dBV7sP#XLQbm+tt8%1@U&EA-KV6&C{Nx){jIDkd+~q~4PkP=1=n7lXXS<)x{g(^3xT#X%h{)*R2T0s9?w0R3Xi&8&6Qm>90?pGc@y!+;TPiaTF@>n?X80%iX zE&yaBJ(TKON9vVP-cO`0qr4xsUuiq>^V7%`r;sf>Lh}<_aZs2ucEyPq&EZ!ZS#C3S z#mOkmn3%sDwc->a?@2BnEHTW;6(=J#qr5M@H(sI(jbGnbx2S`+wwQ@R{6_xsUoyHQ?Gp zG*BI_Sql|oIi1WpF2-70T4R2OxBx(%%qF8Xt`2M6dpGH-tdrTGtC4?Ys+KAg-ZCjMf)~ES}zCs{96)*KG0)K5zq0_s{aLF%lPZsto3aa;h;Kv?n%O z#tWgIlFl$<>qB|f80$T=!HbN9O<|qtel$^EE#{GT?>{{Cxcc5wymD)yb4eyH^ZxpFAJ)Kt#Y{NzFJCid}>`kc_1+SV$E3X#gUZtgHN z3t_MO-G#p5wq8a{5!BuLch29x_2}_C=hLIJE9#?K6YkctA|j|mV30UGdqeFd4(30u z!ziB+3@&8(a#5NBMfKrP1&OvSPla0L_LZ%tRN8U3{w$x?hlrChWp9%wxsQy9QKo%8E8kaC$d$QjE=g2$(j0-*s2QVjoGRS zk4r;@w^fxMJaqqZ+(B$3;#^dEoxN4XC^_8+%#G|n-Fw$1d#lPYGHVUJLP(e-Cq+g>^r5Ib4lY8&QF`(a+6CZz2PLmX-QRz};$ z8sd}_w@m`wyfpOchsEuVXoYyvm{LUF=el_r8XDrsJU!_9c+nD7-maX;Pt#DpaI{_P zyLoLQcDx1`UbG+L0kZIwq-$3|I@~s~aRAv|N!no}ypnXAQslL}l5`EqN=YVPnA-6* zxQvIs&}3b{)MG0oG)urfO8nE07oEt&WKtcwi(Wv7a?`afYYm;eG+?){(c^A+A?ezF zIRu@&)}Xq)?c++WiVmvcpdL$MZTG%`ZE=b z-S)1|T0LAhH$4xkr^*Gm)t|SOi?(8UJrm0sY6{4vVfnE$e9kbbdN4W^;_{Ag&gi+A z(d^SKYVUz9O9`2{b0d_moqZtKFwLWqHvIr&-jiT!mn z(@)-f_S%){)6=7ClN(1dcD$ngd{Viu0)KkW*=!O?@ttY*8#yA(ekt)FahPrm-#(mO$q|VT#}Lz_3a}%x&^kTe4OL2 zW0}Ct7eOF&m72n3SdNXN($|}BIWspw;vCP?;#?}I6mO>~u%p^eBQuJZwA}>xR6Akr zG$m6b`d!N*qzotQK#I516inmqG=+zyA;H;cN{>ybx0Iv%wj9OVX=2zd_3G)?kkz!0 z+rpSTO@_A_>)|D5rzr$YNQb3--oLFBc01-ylM!u>XNC3xvb(pAoWKqH~<+6Z+S1DY~x)D%#Ou0@OexSHyEA2{e0E&Y|~ zYb%fanIUk@4K2P0AobmRSVWcRYh7S%A4)r!hD!8p%%^FJ+3V)((7u=6TDTH@9ao}n z9h#inRs_(x%c!)gL|=nX18C5>N_6vLX4WT_=sTjn-?q$Ef9o;Kb=H-)07*0X zz5fCdxzHx=Gd(|SQ7bNLeaVBUm0C<^-5{z0TG2t#r@DX}LwP-8e|5U$HQwM*<~UMg zuXI~Obv?sYr&3*=x{sP!MK36izK>!~X} zg5*D*R1~}Q-)^x~(kV^C2l-jaD3#UsY6Z$YCfUrkeAjT#gR-o)ERgPcq4y=Qt}-kP7*jfudi#9L*ltPYAu|7Bef z1!dcFxprCo!I4^bTAQNNUT;;Z_3G};r)D_hOL)7S#FE_k-ox`-dZl~o?)^8XN1vLg zznQ3V6B@+2gqQZZ6Z>X|UgbC&S zAueO6q+k+n6Dd3>4T6ee_nR|w5o*uwx*9-bEi8?=Q&TPDF&`U>CYrTYb7bDW%Sle?_ z9Sx7&88V+%^p2fhvZ7KGYx{{XNO*UalbWCL!9h3%012<|llsh8&Trp)?XKrUt`WqB zCVd_pXrko8eDc|#EpsxdZuNSYhgQP7aUFB~(~qAcXND$+>2Zt`@0!pwe)ddxrGv@7 zK+p4>cz0vJg8}ibj!I7qnloR$Gk|2%)z*OIJ`cE8NAE>X=fSrjYz-|gKSlNZK#_Hi zTzzJY)6pOxJ$c^4xtThhKGy~3Dg?RhOPwhOj2zm2be>H-LdGS@<4^iEeB5bhX!y}- zs9zWG4+7&2t+C^_p*kggsU4S4W|xU^-75XEkI1!JPUpmk1dqmAiN@J5!o;ttznmz- zBnQ76!M^cUWGPuHyc~HG!|xwah!n6d#e03;IUXAyVX+?X+j4S-g+soGPSl@HRQV?K z2uwf^CPc5#gOU!wOf`I~Yp@C_w42P)@gDopld*$}_w($X%rSI1@KhXcyXu|H8D{AU z#mm0QT!@dtbdn42+BSy8U-x9r$V#6FG5Nw(L05#_5vZ}ESxJWwd!VATf+1Us)L0v% z#$qgnG&ckzP%%{|9jm4`PExjs#8ew2rlOM+eX_8WAv@i}CbTm^#zv|k<1UolM2!J9rZIpM zlaa3@@d4rLfrlBjcDRJ5k-F-|WZIKY8y;Z`_%@*O7mH9ctp-RUe4*^dR^AR5%IeHL z{0wo*05O@qiX`D98+W1XhI)gq6O(PWB{7-4T?BJen83Fn~`JGahs5a|D2T zqtPG26gf0Y{5jfB+SGrRZ?hov-}yxAR6RHSbVs69N&GPNEuhOa}@PQN0qug%+56H@AuTPMBJRncu6Gj z4;#YYPH>S_F>2EbTl+ogUqHw!Sj3vb!0rJoWPFYYNQ@(z>vteN&cu(&z=5W-@O>Vf ziJuXI?l8UVpZJBi5eae5j+bx2q54{JCVoZ?`aMcJr^>i2B;7#1kbgp@Uzhp~M-C$S z!p6uKP^D>o5e@4gx`Me>a0_$?xgq`ZU>oTU%+q`50X6CH$zoKHwsGyTw%{8v%?m(8 zkb5Dc1l5Gt4d*rz5p0c!pnEvE1VNv(yQLT8yCi+spx@K`3BuNWPv2+#TXO@ACkWj^ z5tXBDb?-0xP~y~ZS8eDdJ0b|Qa~APASE8Ojo57OR;f2&Nnji*j!^EJxE<{-gxK zm7{H_cVsw%u!TB;AWi`NxUkn#7nUIC5HJJ*{Sch2dw?Qr7epug8BDrEd#DqPBE&yy zNy7Z(vp!*4sM++iGht&8gM9xD>_X|2Tc8P{%C37y^+W+LLz6hPvZz92XDGPqw$qaZ zyaY|cPR?0N>8i=sXh8|@^~~!(T$b^=kJjhaZo27+R``4&zM%W%pr1y0a}_O28OeaMWQEPa>9}SM9AO$1VVi{Q5(kONW=yVuln6o z%6N2dWUT>(j8SwAkU|ZtH6UgUqadS5^ir^dVyyugZRn5IOU4>dh#ujX=j?`sWr(*1 zWW<5tRdIiyYk<$WfLb9>Qt8a$eajrSky>GE)C$}sH8_N3iLhj08%Y+nostD=p5mc{ z{$Ls;3tJ~y*!D^mwm`Cwk0%{E`0SHqtF@Bi%xGR77PSTc=ytQj=*<9p*8002LmsxwY3VXyhhR+~F?xG(Znfhq)1Spg$Rf>^!ti zLG=IIQ=8w))|mWCpYeoH-j z*vX86pB^$<$));_iTbs|lvteJe)Q=4{-e8Z-#>kLe(&q2@4k2b;PL5O=MV0m-?M$< za{`gpuDfgmtl{XLnJ2&G&0Wac~D4D4-;r*Sca6NS$F^EqAR4%gj>aY);A{$^HyA+^m+ZVFX9@7O&GuIA{zIbt&_86ItBlI9XNt?DOt{$^G% zaolEBcwibV+|8`?=%jm(9rU-4=5VE`d797688;`6G4 z{ehM#M>aA)`|{q!w8iwo^at93D<<;!MDGvuo3=mD&Pw8j?nCD3#Y-3QlL(fcyb@`C{HPxVXhu6nv?Nl0ebmF~fW4_)kmt2xPhn0x`8|YsYi|2T zDhng}%{ZIE6Qv`nbfJCw_frPS^Y9@YgSo&EK-MBrfCBEToz&csk`Zzji5NSLn6@P# zCFi7cfnhHa8E9gK@Fj1ND8y5~!3Mx1wrT_msD8F^htUnFW*7AG<^lEy%hcJ5BK zT}KhxFa$s{mnXIvD5`#D@k1I(=5|Ih#~M;gkFaUzrvdDbr0ae9!4cK@K6F5%_1tM0 z8>%|Z=?r&Psx~xQ-c-v1RyXcwm-d1FNXwWMt?AQi;4H$`jYf+J zAkor6B3^alD&s@0x={-APu2aM@kJjQ}ewr6BkzZ!jF$=iD>5FTuy3L8ixvzW7%N{+l|~naP%1;5m{O`01-UE>BYEtk=p0_46W&j7qM9^{#qM@G>x1Rj!>l za6SDdJrDE#0HU|tA9U&)h8U{e7H3BF#yabb5)JCA*?#uB>X>X{ilA~ZIsmqp3NFr^ z9Lz9z{VxcByI#;o3#>NR8$S@E?7I*xhvsr5 zb{>A;vh#J@V|h*wI4c*GhZiX)^XB{$0D5RfKmy9);kMq}32(Pz*=P-tx94bf)(hT@ zjSg)JSVnt_j$c10c6cj~V>`S|s{T5aDnmV4RU-Hc{`1WX>c(_S{(NJ?2K<1KK=At$ zMf#$a{Cb2m3wffHky(~FLS>}BeMIMLz|$$jNh)r81a;3f|}_$X*W{jWY0M3w6~b(g6w+-JSkjV`aJik3V8LKSRu z0OSW0D%cihM#07yP)mP2+eByK*Ok}E>BD+I_-W`{+)ISO0rk{l3Pr8xZ*d08srA1g z^XU!?S3x<1g~k@cf7zjP!q1#=DdIXJJv&fQi;jhdJi?bEn#w`|dDaf{=%4W~kD3@= zX|PH8&dyzmxTC5pZF9lJ%p&bX^ha&x__n1NdhIauqSTr`6aD!VB}o4CYoOM9ADLe9 zA)ybA=|0nQsze*yDMxQ>_1>-^yCu4tgVn1D6EprU3 zkqXQn%)ctKHI#8OY;}l2PjZBc2>F3^&Z4$uy1aewy>o7lL2jjH;YY=!e!3`&)YPu< z4!J7mG*VDVnwx()qH8FqZJ9_W_OxXpJuuTzNDl|QYP#_H90^6(8K^x53qRbVWRkI# zW!+^hU$35m+Lozceag*91?7e3(AqK;{xF~FZQh$N^?v`uaVy99q4)>kEnlVHGhUI+#jLd{+?AZ5>6S?DF<2T}PQ%>7bGVyXA#9(}R#nSP zn%|N~Y-Sk>W^QJeq7$xZv7hIjiOo=59i0oi_H&C_B&;)!r?sEULy<2QoT1`4Iw5vq z!C9n@Gf(dk3j(s+93bW7UprVGuAORm;B4M;`U!&!#BNb}51oPZRu8f}0p31DM?&A5W4&`8YG^{@=~-t~CE0dfx5g^6`x$$)$u_W_!V4hlAd5{;r1AW zKU87xZ6Cu{dz;W$fO)#{FrFgN#1OXJp}T_fqaK5s&~SihxF_>#?nvZkXYh$e&K;NJ z_F+SK0bNrLuN=cbKaFq&liBZ+^@XOH1W5v#{X~REqXiu`v5+;&S)JyLnjPDc=x1(x zVz_jM+G8-YHwB_RuhB}(@*mFxscf}z_Uy_mY>i^v95irNa*x66MOcN<*;D*!M$ki; zFW8Gj0SdT9LTW}x2@AQ4M2wx}tZL8m_)7^keg%$2?lG7KUnVbPEfT~YgZYo=i$r#A z2=i2IdejgT=5+QVk+CJ7hvgRuTyq#`1FFxXBPH9AZZV7Wc{6fDx-@t!Pr}B``z#oP z%IxSU$%zeTk<4ypWOf~OryyU#%urrz7!?uDY%96NjC!T1#~(mBmum*-VnaWqk#cTk zlyjtF({xBJJ;Hnuh7`yRwZy0&98sO`oHj?Ru<6Xm$qnaG5w>}HU`EsdwSh4=1>He~ zI2>?gom+EV)-6&~YVR`E&x_=S+W`{oe}?#G z@Gj$B#)q_{?0XX*=j4WSi^UVUA+_P3>oKV34Y^nnOco$GoX^ma@n$L5V{j8%(vj*U z!%th}u{+^bI)+Dy9O6S39rYO8ghEi(ZtF?MVX6g#6Kq8$!xyo=oh6~^3s}HmCVWvR zg2z~PWDBDrcVtulJyE~jP{$5Umriwg*F(1a@QYqAdsDqUeUYMc z&4~Kq+>k7g)sK#t}1vAHPcZEl$VZ-0vUSG22}RnE7r$ z)Ld*kz*9C<>qn=}zcaCw?ej?NH#cIx7y0M8z52$NBXpPrBzyVJK)_4+Q?*U7lC=3*ROvo<9qi_yB)R=CKg_5Ak4~V&Fh>f!xl!P;hk2BUQI~-Bv*@k? z@je9dv^=KhK-%>@oI82iUc@*zmp}P}{;m##K4PLG__@B*);>8bGwq1+BF0^ern?#^ zdOi)x;cLgh+?7VR7=4{}BF4FC00_$AcaBe+a`>ic-V5-eWPKs(GhaEseeboqp2^vo zc>QDkFWL_`NPe&Go1>jV=sgdA%l15+yG|kg>BlWqRVh*HUq~&fw?b+!V_fUTtM*me zUhxj}eGljMYh;iy&d(_FHn#_;Zw{RU^tkO&KGzq{h)s_X{|xHvH=#k^ynlZixc-{b zRTHDVuX}jZ+k;w+4L^ zxt^Ac96W0M_RL>83LwEjQ@N3pk~{L3ju>10?^!x7M=Tu~sA2~6C3op4#FpUPs9`8< zdB9dK#a}uyHpTUHIHE~TKFlLtGdM>zgLEXNGpHf|jWmPi>CI|ObYUj0gdQ89;y*fy zauUdSB>tNl@gE&U>5Ky#13&6c0y#g63M1#PvECdI6zqY1P#370j}D;Rpe`acpLsfW zP?;h~oq+a(C+&W>Q=;P#+=kYd)0vYKK`x^5@`Wiccg(3jG+agLewHoVg)9qjjR6`O zFREvL<=qF5ZfQ?GZT^eX27=mA^zS`n+6s#xZRMtWI|rHc&k#osh#=Rqj4!R7p;Sez z6G1L4T8~ALYpTB=T5_U8fKmXOuYu-NYPrc`j>fO|llnUd5kkAZR_$l@;})3A%B%u^ znLNGs?{~r|7tP+yB$A$#Y(s%7bMN(>WG#3`slgF!hUI6JAT*3^rkA>A&Kne&9*^M7 zH6_m?jFj#N?Hh=#eP6u&`24{eZ$G#={o47X+g5d&=dUiodBbzO*VN7F%N-oA_(MW` zPzD2}HqVMyqpi7$uNxP?=b*Ybz)XfpLr>ta*ADeer_Ql+Zv^V=FQogVR zJ)F{-5TXFE7kPRB9O5_98=CojG6VJ=??<);hX zb;~|9tR$!}J$ac$46Y90?xfT6C|4c}M_#_T1~PIYK2-e^TYc+Dy)w%CiL_;u_rsw;+%@R=Y2=Di$d(@$ zX$>d@Z29Q6X%?p2dM(SYCtIn=l%#mVb=_36wrn26IC>)tC+15gUEc2R0}2 za|U2o!e{V`@+P)Elvjnir_Q%^yhaYZAZ`b}`(sp*GWZ;&g8;!91V&)&5!J#zlYOLk4mG4=70`kc_1 z+SV$E3X#gUZtgII^RAqa`@4r-S0B|n^|uU*h@c)_Ex>x`eERGSweN^;{^L4~@=3F% z^17M)sf*b~{RN=8TBz~oFV??Ul9RoM&KTUEil zF11DC?|*71=(9whLKroXzk{R#sXf5 zrzf&i)gW>+irHIL#*2AwKEy0)8|^Uh5=i>;WipfQoz#}e51gcL8AQ(YF**^tp~+%Sh2@p5CMA11zaMDy~11xsNcA;-!=IErG1B{V+eoK--rOai-xT z3~{9GTN-U2>#Ucm`r0OiD@iqTW5D9B!SMbA`o2o|n>}FPxqbTVO?7QDed^@tO!uc; zH?L)+?^{OtzOEPa(GpeOE}h74xh5oEUJvWKd1=SO8p~P7{f+x6H~?gDC28)OtZP>_ z%k(8>OG@ ztPf?esf2k`@phU5J38!6bNLh7X-cL>^t+a6%EaPjA!E%hgHpVmreGR>rzt!v4GGRp zQ+n(`y|J*Co60gP+wFUOb6MTzwAq7mytAhX{5pIy_@C|Z^{bQ_x6{huOfM` zd3q1P2Fw`^Ij?e#O466^#@iLGpiN4CHw5vqNuM-Br;40Bq z_Cs9K190VCg1(-5+cH=Et;aCeSy$cyi0cY`(!fSa{RJd)p-tRpdVbiVRvbRQNX5u5l|I!4+!)I18T+dztUIACw8l`e!OD+40cFRwhU$8Ttq!c}Nsg6?OQl3* ztsCYC^}X8dzgGJ`>cZojK2W!6JzdrVwfq4y=Qt}-ddd2jfuc%l+HGP&MXzFo>;>2jIrw2i`387+CMl_>rQJ^blU5! zO0{0yz4_FvY-%KLzxVL`mR_9Rx_kf4>CvYEHx+QZ`kRR=H=#kSOL%FoJF(A)!ytz< zl!S@>n@E8~r0a+KBscyAt5V@7u!)omjWDT%c~5o|DVW6DL<$c|gP>AvP@OuZM~ZJE z1?t(Jag0p_VWOmEO@0$8KvA(xB%`98tKMbrCQ^u%{FE-b8PrHXf{zNVvK(!qtx6oPeCP zBhG0v%_YYGd%2rdS6vVDV+)k1`KCojqJC^6CF;s3Q9H-h;?Sl)!q@@{udZ9~XQJp8 zpvmbh1QIIFq=Qi>;axd#<^V?Z(*qdo8d)DEQRj!%+R;(;($UmfFX3Iqh!nk$H+_ib zq(VI)oP>8C_kdVAP8#)qXvR?_;a#V&)LeZ%Kwj|CqWw&5@?PD#LR;q0CA<%=neW@LnP-d5jab2eWvX9oa zT2AN0Xa$cJYC8It6GfQh;CCa~mns*p#xNJ5@mEXaZ@g1d?x%kLh(e@*eJS4S^Um?u z_{ifAWM$+1)BV-ts`}H3D&K@2feGlrgy{8oFl>9{t;OO`^-ktO3N1C6rTTgHPUaZ8 z6j1!wAt*4E5lcq*WX>>4S14ZgP3A&;9CR|*wm3BCdna>7R{A_BH<^i-o~%(d79FVQ ztYDul>Psw#tib+D=Z=iUFt`C797hh;sF;dQQry+D`eqGigb`D9kQ7i;ioU(q4n+m7 zYu6W%e4nNF0?oR0fekdb-)Y|Byx`whtOmzhD~Bgj ziCVse!X`V()!jN0_V)_#xreey)a40N$! zP5eT}Cpqz@r~qf;$7DcuhKdYlP8L`bKO+R)VS3p=@e6U2pCyE6ab?59Ck1EXXT+f2 zqp}m9N)Ax@0-YRu!;ynXzOXj(1ypHTUqpjDu+ymje7**{1A3a1Q3d7cL8Ln{Pw$}z z;Bbz97mVF`i`0Yebb55dX02Pp#35uj_w zpM|7*c^2A3onRCp{#i>Bw)r+3zW$zk)+cidHJhGxCUcCVz0xN)M-!H3p~|j%NA*Nu zhZ-nz%w~LIjO+{rcinb6vY;n61`j0cWSysU)#PimpoI5&*73sYJ|dr2yXmGQBH{Ce z_=4`2gYxt+A z0_lm2Z2km^Vm4u!iLr(AO~(|1H2b}4mOxVnx7+I0xv6{lO+lywU`;#zdy zQU}d040Sl`IZOiVu%U_<^ta-Worl&bs9qvTf-H9cpyNfK(k<0|de)oLiJa}9G9BrH zX`v{ie`9szVNt7Iz~U@FqyQQta*n%*Ee%B){g!(6fZxn2^4hGk)qhOXuN9`m^7Qtj zN9Xq+-F^H1>BIATUq5~Kz4Hf;Pv1I!aR2s3kTf^n{<0pg)gExadqsU-%tYJZ;BdimEIk}p#og6?Tbpo% zmmIRJzmB>6w{+sB#Ct@-vMg9fN7TQ@niUZ&kNX0&>!jtN9sjx7{2oG;oUdy zPp=ODw|LH_c=1?n9^~5(fL?-btL)j$(NzCADc+7saZ{?!;Rfwt=2hEt{$^HyA&1}0 zs$F@W;AU1b{6L!-sbTp6HnW0><2JLx1JhvPZf2!NC*6DG)|k#NKErM?Vl#`8?oZPu zmFvdJSAt$0v6*FAt z3EHuVn#-F@`9XA=YC6p}_cFQE2v7l{HCa9{17{33x04%v$Yf}Jpz zn%GXQJ<;7!@$#a<*b(8O9KE_sVYA)e&tLT!6(4QF!80eg|i*b>jfhj0w$0#9GI zs4g#45>}bK&Vu%JT-!*OXP$mMV(}#kdtBnbMdfu_1!L71SJ}9YPGUrIi2A$s&3paY<1(7DxY)JjoOn>8y;iZ1}&)p>LAfl zxVmu}RX1v4F|=VaKSLZn&>u+)f>S=^svDOx;^UmoaBH!6+#gA+eQ7cEu4NzEJi|Up zSEXTxPjl*YY}9JH=tLgoXDYx0 zHBcYoJ?of-5>D1`>q$pLIEs}BN*(_y5X78Eg1w=#lVz_19D2e(36nTSGWDa0`YI(@ z-#TYY54l5@{iK-i@H-RrB`KpLrG}iTV?>7^AJJ%4=HQ$zSe3@1LgZL>7{WdwHxQYW zu+7BPM?{vE4M4=wQe;)pL&TLUB)ln=Q)ln5J|t4L>@QhVEqe+rVb1<6kNfr@=S#W( zcgxuUjI1v-dh0>GU zs$!!16Ww(RB$pt{XKz(Jwwtb@M7Ep#Sz)Wna*G3OM77+aIq>YQiU)Wzf5{&SdxV2q zps~wO^mNMmffEjHBaxkXJdFvae!18m1cGzg$vbzHO6&Xi!5v2Kxn^uKi;^o=`a$9RNGg;MU2(hR-7U=>Y(Dy`U2fE~9Gm z?bbz0mjcm=2DM`&6pUUfSWwXI`_3dxJzg3GAxd^2-AII25ukLXuaP{GYpD7OKx?^||0&+~Bsb{>AFoQStW z*_?j@P!7!qc!{_Bf>bNoZw@xIW8u1JMi_Z({#&k>m;$xx~`~W8Vi{T;qCFuJTMf#$a{Cb2mu^%ulO~Is;2QrkJ zNqr(msBa(9`5N$a3NbnUI+lr{6DimA3;*)D5fW0<)?>?=B#9)D<5`~wi1`K!XXbKH z07h;(po;-z#vYdf@|J@b8y#%M7myG><8iSHZ#l?75hG|XImIdWA*}z&Vqs|sNgL_K26!* zE)u-$jNq*U1UhSI0HU|cBTWbCg_?6++A6*E%i99xt6MYe`APKK=>QEc)v4j3~T zb2vvYMQzKp|7^qUF$fYATs^_9@+`bvLlbB!H%U$H3Xjgg%xR>ckhCIS;pKY_l1%Jr z%S3u$rlXJ^4u39XchV`?q3^y_ckkahfB)8_$M2j^kIt?TP%pK|VBz;7N(Sc9r2Vrc zZYCL;89URly4OQZpS_{>JskJ{L2k=bus-Exq=NFob7*au3V)bS_458m8;zvK?H~b4 zZp&1%Hi2necOS&Cwu?t>W(6{mu1}7($xWe_#t))rbl)7YnUxGb&}KGH5J+v9n}>D5 z#BrNh;elze(0UA#-gP)z2<~C5)`EUYlY0ynjEgq20^^mOTEyK6&N0HCI$|@+s4sJM zCH5FBxc`6MW_CGZGs{pg*WEtyMWuT}wYT?ivp3L4A$r>HF@8mQNEpI<|5_b1mr=O?r>_ z_2T=d2Xx#e&U13XRa7*-)3wkBBL!W zlFiSwP4ixedJHZzoa}1IS6z}aZSuw5+h}(Ux?Irq7&Ldh4pHKtPHRdOO*5`waAoDn zK`^L4O4t|aakv!>F3;>2Lm(Jj&3hi?dJO6hRpcInnJ1hHUB3)_49-FWI#Q28_y(50 zgA0YY4F%y-<(R=!CdUmRJ~S3!o^Cvh-wD2>pUQ$Kri;k%d1yGmG~AJSy>en78YDB4 zvS6yYuabKV4iN`cXVo5qT$yC{6G9R~YLCI}_oIw5!lT%?)Djt$b4iwYW(p&6`>@R3 z6o`_9v)NN<)zI0W74pYy?gT|2yA(G2-OAF0jWD@=SoUMALg?%%VF{FrE7V?UkHPd! z3oOf>qNUy(l?%rF1vQboNW{`YIjh?9oNioUMPe@!8E9fH62v-GLOkUg>;V=D*2ixh zQKSmG+{olZ3D3jwiv+GYbo6C;hU)X^NXa&&)6a`?WpYEh)URprDY_i^#2=N}(NU7^ zF}OT~GCTdRK9mG92lVZo*l-1l2(=>}6`(%HdjVC>(E*YlQmC9;o*Cubh^Z1H9d*jr zm2b(XYlHg1QMc1T6?SxZr0Z`DSW;GQxIBYJ*!o}EX~7_(`ppN1tGyg3k-fr1Ka}W8 z-(zqYlcEkE?PYtb)#+n*BFMQ~zua(Ts~oM|aOohC{u!cWP;R(N`Or$wAp71#hg-Sf z^2}oKL~clxce);f+B|ZxBv}7~;84E^@y%h6!Fgy&N2-$yKW&i*e#rp0$6)vw<{{p* zj(I2qW$m_(bgb~(ex;6w6>T9pvQdx0d1pyz@bqv6um>Dq!b=&)Ic(*pz^BxIPt>nB z)UiXKrBmJU{oy%Re)vT%iM^>_p1w$t#MXbdp(}((eq6Kh;Y9tir16nzL)PSc-KUEX zfA5O=f}ojJxpBy_IhG%mu(Qby3`dOkKNGcM4@kt@3JMNc^Rf(E2Vy~C2SeYEQUb^K<+0?vI8=l&p>@96Pt?bg=3UeEWjDoge7ARe&^i$f=TX~Tfzzd$Gj~dt zd}%p};+XBOWPG&Uk&-uAzm)exHD+U6gyDP+5VShPP&@p$wQQKW6UgW#|C$-&O2HK2j{phs$ zcWkf23?a=_9Fj0w^!e|is!bsf?23!kItQ~gmHNW6?^($ecTyfsiDn^LYL+j}kHJTHbyZ-8CTIhp0TQU&J_b^0Ylf zo|(&^Jdo~sK_4-&2!1)!ciP$~w!0x!w2h5KjOz?iyPEh_uUL4m?wf&5A*7RUepWR~NA+7n&%@=+bqeuMKW?e2N{L$k zLTX996;cNoqjrydmA0q-peQF}oY}9DP{!D`qJ7hL4$!NkLHS%?I3v2$!v759ee=*D zZ{EMZS+2jXfcNGM%a2%~V*PcGRG(M7>4u3^O-395e`uM&KJ9qOzee9)G83$U{-%Fp z<3mdYw(*|$%>=Kj@W#L|(mDXyo{S$>FhB_OF8su`psDV;$L`ER5dG*53< zTUNrzN+8im>vp97pSpMJwJbU6eS5C5%S|535A%i6`T)pUwpnZ5=d5n`&D}N)c6$Rt z7UH2)rWQ?Y*Q#z;RWG|hNR5Sf#y%HwMWDVJ?z6nRDhPaztii zWQ>^AJ}-8+ikgw)KR${I63AI5{&P?7sitl8^@?vFlx}NS7`dGJ z#(K-yiWeZs3qZ|>GhRM`ii4V|`DRYd*D9f{463J3yg>ybNShS&gD0;0BXc6_%lXVH zh#+UMynH!x<>iq%ao;V5MyMz~%reUb4*vj15#`j>b8(%KrZuP?CI3D^ruAG`1iAK= zn_dLD_K?Z_Oj`KJH-jR`&5$pxEtlKFu{F2JWe`Ey+Q~h09f%;gip0mBoY)ZH6oBIC z0%VXgz{W^d{4WR*QoFuZ?I-$i4@~InSx^p&eyBoZNF(iOFp-i};40jE11DLlS3e?S z4{F33CE?r5FLkZFLck+-@}sIXK(;>MbJ_RXHxPUKK7aS=<)gRWd31ID<;y2`y_<3V z8WJ4J(`Ro*ocFqUZTdCWI*fEJoc>i#EpdX}j0<3pL@*xMil0>xN}U*9t9`GH#j4 zc}hX`s%><8@yM_VnJ0~0W)N2y5$xAvmYG7#J;PrCUC zmQr(|mB~J3@C)r6lrg^>BtWxFEHiUrnb9jcIgPMQb5ifp4_$0e6CihSPw#;)+zc&d z)>+-@EOeuqU-Wu`RIKo!OB`8S$OhfuBr79E){~dOoJix~Z5K9;>?bc=DiO%b9OyVO zu<9o-wnPrP&fWMD$qV=0Vti!fr}LR7FLUmV9 zL}N1BQ!(xc+ar-#3dVwr5nXp(n}3>^w|WEm?Ymc3Z7lP}7|m~(&rDwxMKk;*HhLjf zJJ;ly`K==@A@DWyz>K?Djg+0cl3A zIEd1WiTUeMD^4Nup5=-o0yQI797JkHdH-mO3Jq-{tC<6BOdp=f;@yN+Oh&cymOw3w5F_+0cA$yTZKmHmSQ_v z)9S2;zP=t(Er@Hl2G|dOiD2+sOqJij^O_y)JW6dL=txz5l`=tad2BK_l~aw`pd+!N zYmAi7n<8YBS)a`*)vQj_WLTYCd-7 z5RZ5EeTg1VpEoyee`tF0{0YDPb?bGgj$@1 z_)`bJo7j&}%;%KG)VEeSR)|!-b+^1z8-{e=)f zVr~NtcMZD*=S{n7*e&y&lL``TSw0kM$^2L8+28zebO!{Pjq8m<8T_>hLvI;amh1>N+>Y$mi0TF@-l|GQXWhW`O#L9X zs)BiAwyMJ8h9M%`s!9(&bpO)xzhH-o(^I@v_)m3r=A?rh?%3Snf? z8d|$KqOpJ%;^~=eRkhrrIo9H>3h`o@n`5M_r5e}m-7gVUAnC))B)^36{UGU^XOh0T zlk|-^2qfpU$1xDT1yJ!HDssr0)w?k}fiR z-y+lZb-iHd=C#QxZx=!2w`gCzm`=f|o7XO5$D8(*q~IX|Aa9hfB;AAp(qI1SqbGMZ zI}ecE&&0{oo@cO|*Y*NY$*95DLEm$k-DURzs;>W%DqUg+et@Lvu=C$9|Idi_{r^)_cPc>UCS{se?F<)S3&=BjoWG)h>zNQ zc>aSE^M=YOdLD6{kf`*Dmbst$GMSDaXIkNhc!l}i#IScA5Xc!VCq5=BgpY24twTO8@z<$NV8FFsTuv&nf=$JKGVu}nu2Naou=@xVMs`Jn$lwr)SC)x?MYO)3PZTl zYC68aFNM-7f#;$iC=$k|J^%Z zetPHa2k(6O?t?oIT?3-h|8~c6i%=v=2D=F&)=gHDz6dwoF36^F=@6d=loaYBJYDrd z9MDXtyKq8X!GI2#H94TE2cSh){kYn?o_c&7m*g_?ikALL^o5tl{?uVcxEoqRC3=5Y zWR>WPjju#sY@->Zv9SlB_2h>f1@@BP4Y(JTa+dSaUL(9wu0-EtmFSj~rk`P22DHuF z*^o+aLM8eJT=YR28nn#|{~<2#0l1x!ZjP^P-to-Ul6R+Vo^|CdK+?P*WyN6L{{_U5 zsZHD`dw$rX)`pD~qgd*yfgxo}6Dk{PlP4T|PR0;Cf!(S)SL8$-pUByLI@v5=zyu93{gx z_A))PxAu=t%(~OsAv*2#R;606?p{1K@hS(}323;ln5Qy==dUNG+=Pa(Zoo@>-5C!- zXBvhR-zj~>^lu^s5;0sq!WXOYCRC-u&tMZN8Jb~I4fCGuCQ>j-wuuxTGz@}DwLx|2 zlpd+Ri4>?OV;$(R~^)olJ&h-cwrm!qXo z$-B&it4k+b?Es-XS)d<|sLjlJJ}kpctIMv3MSlP#s?Z;fo0>SbnG$vBl&GC!JH(;w zu}uIva*LfD#GTz;JEdry()_&AdUq-QTgI5#ur z(@%B5xiTbfo9J+3!pN!ZN9WljB4oPt#rV{!hL77plNp*Cesmfd`7zz)V!Wa9ld0vZ zFSp|o%IuIBH)Q&+#E4v*<#f)BNbqQ-rlWr~F^owLem8@CC zm}vg+ghQl&eM7w0=bh8B@j-p?+Z$NbsEi+rBK0$WHZkR!&@(UrJvbnGeI7g;ZwDD8 zVV5?&lev&WYfWaYexALPnSd^&53xFI^r#hR3t9=fCo{qj0H<=6Z@xe{z${yD} znGsp(^I*(W6yk%5$SA5=l4>kIP_bFTAzRGUSnlbkg2gbofoB8|Qwhnzo}>gx*)kJT zEuEOEgQS4sKl1H^q|6om?Q=)(z>pntZ%c{%gw#jzC1r6aZBA=n>GYtFS~`7H*H8}W z3F!C4b4 zb34yS2`!Gg8RuMtBwOVRW$pVwNHX`3#QjYCadHPRP)xQM@)1|ZyBERPUoJmUo9%SnjHKe@l`{Rmo0zgLlrk;L_Or7{YjzDg zGU>RYjSI{lpH%AhC_9H)f4`?rHR48PB?DR@6{nlOnb0DsV$=>VZ0+|bIaAY?sAHpK z;ukW$Km^Fyq4CCiWMCu{KPLn2SvLyOovAoPH1QE3=nm8C{)u0R8yy!YFCU_bkBC9P zM|I~m^9qg?kT2lxiF|VKHBSy!nS8-L{ghOi))!IUIv5ei3g`~7A^inut4w#`p58+b zz&CMfO^gL;+pHjM720F1dK2OmDI(z0gE*X-h`>F)e>gosp!^U+A6$%TT_PX|+-xrp zgcbQcy`Lbg!uRxj*6*I42ts#IWaVfpTV>@R8r?HM+Ux_c+E(Rn|Pcshwcm^1h>^X9}V*d2Rl<#^*gaKMQL<|CFovK6+xmV(O z3UGqb_>|G&cP3^dm>dJKfxxS2t5vd(S_froKp|rcYXJ3QXDSADpa#(zkh6wSkWtCr z_5vbX0}yTKkJf9!8c>Ly@&@$*uMF2lv<4vJAn>a48o*@^N@|6#2j-Be6;@8IAYD?E zL#PENK?fjNP=OAaWMS1QS>Wb%QL>QLeXOi7{g7YwN)}d1l7%>)nPkB|T`-H>Pl_GV>i#wKqI%LHi#}Zvd*kjyPaQ0~aMap^RHY!dU*NZ{O-NGS9h*3ff&}Vhirtnd+hAzV;sL?UYfpGi(}3U z)jH+aE;*=Ej?;CQf#(131)aX8gNA9c`TmJ{Ng9T)K7M@v?T2v} zm3r}1Zyw-H8K5vtKb9ABqVCQI+rLeUx1(}`ZK%$Xx|!o0%O^;)`_6=LfdZDuvmb#v5>d^1D*SL*I# z93#nYKz<(&Fl?0P9~T3KPy5HB2bGj+p+C?%Q>v{qrJ5*Sc3B1FI>^=5?duFdyu|)MtBfA++E;vn zhgg^GQS46niqBmr>DyoNxm|gL?0%+ajez}u)`%mAD?WF;CQWHEy?XD#o#h%_F(F)c zToWPl+rB^0+DqaN-G|)MPfPa?(#cojSJ6~#;q<9vxdo$Cn)t*A13&!{lg`iK+ zu4^^>x@Yb7IgYiC!^Ivv>vOyW~C@o{A`ih6gZjS5CcG7oT(SIXr#oaiqG`NW2oq zu;og%&(P1s2!6@DIKAB&!Qdy9RD%CFF&&Eq`27rzQlAq``JukEoGM#(@4v_JH3H zj=0Vb^I$_ZTJNuJ+<2+lp_!C>da7>ZzFQ28th#Y+vB!R@ZMZtLs54yWq=>6`jF1Ra zH)06euQS}`)s5=_iClFfFG`@!a5LmXyHMS@0~C>ZZme$Hcq~5XkHpnwe$}%NyA(S1 zQMxLPI>rI1;RcW`^3DHWm=UM8J3q^Z`7X2Y$b%NBy&4n7jaf|>f`qXGJV*oWi7%!6 zsf1Is+Xm9Hqc7dl+KDR^gA8$Z9!d75%1)8J5pXQ=`76N%wG2db{`%p>e2EjRU%3=Z z52fRYezHegNNP2{H8Ee%GCEUg=$VSwdA{iI$0s~mg&d6Kq+X?QtPnZX9fo*}5zi^) zk~4`u<3~)EhJGz$X-yEV3DnT5v@comUuEGp9P=F}RYN}~qH5?Vyo4G38IF6rU@L$# z^t+V}xjFF0uG*|r0R7@nUdYZ_`@8^<;(8t*T`7QMRh< z!gW!{->M4cjoGRSj~j-FY^y3gcpxU!i@9gHRpmq%?g5N%#)<%VxHc8q)mLPM(ydCy zcF}vvwyHurJ(I1fmRqn^ZP8XG1H8!1jA$JFDG2>&M1mWjv8xHU@kt-iU-Nzt4sJ4$ zoqIfw2@YTJYJU(&`*?dCBd`N9Ilf7V0ms_}H zbU*9x2Pt~vQ{Obind)uhRBt0DP)Gb5AR4qNtDpUL4MvP7AP2)!fe(O#XmAtcU6D` z?)p3Y16L6iMSojJdKMQmatLC~Idtqe$>KuyQbhZ_>{7(-0rKpBCjL12irJ-zy8+*^ zv+0nV+6R(PiN6&^f9|$M7w-enA9otQPunX#CG@E= z-6wlawH{jR&R&iw-pVYdv+fkH3?1kZATWrrr@CZZD&iFEuU_nD10xQ((nedu64cUz z&p*%Qwx%*pfvp};tf`zPnl}Ed^ zx-}JE7|8?KZJ7!W8*(#NL3!ahytYh*Kg9EZ=ivNv0+0x5hGLPjvD-40tR28Ksk<3I zius44L)bkAvFQaeGF+bmZ97K8+`j_>nfVae_ekW>zq9+-6pI;4oNtJq8QI zju)K6*&>~l^(M@}5Ds@3MT+B)D6k5LZres67qopDKk`eLcOSfd zKaw(qdb)ngF8 z!GXWq#&gm>g79G#2H&)XV#QgjJ=@zQ$OeU{n+W6S%6q+>qsHi)PRR{V4F`mV$MpoP zoO)JPjPhr6l-STUS71p%(Ni-#s`suaH>G<9 zHb1Kkcq_Yo7_xT=L}gy%m6+ilNAZNi6=a8c!%+mCq|@2$!?2M91}gI~`EqGXRdbPe zktndN06L$^y1-W*iWi9-JMovkf(TJ-yb8srXoV3k5(qSjLYUcOa5#So@$^g#%@SA3?9#us4slSrOqjw34&Wq5eGnNb&+u&S5VgQDGH-gduhJ z1pqHNihLApmJ@;(?&cHfEyKN*L8^N34BMhhz z)jAYjx${UAS=yGDqwI}oKbWh~JoS0X{9vBihxi3#O=Y<3< z16HRy5Jg^RemrU3HCysY)$|(X1gnJm7(S=oYFPNQmMn?XF(8<+N=GT?G6EEx!YCGSd6=op@SW~o}>M+ z2vbP`9-ly`Va^nIJEy=C5AzU1x6MjD4=vIYu;>}%L>@3-nkqn^hGU8kq=SgjwI~EJ zjB)3`Mgke*3?TL98=Bub zbPjM|a~#T-`XU(7)pPM5AOcPe^5*^fo9Fs-d{hGL`i+Q2UogArMu z@FSF5F;Dr(FIYdAn6Bjmyg?ncQora1c1T$@-7?HBZ3&@gtRK1k;pvU%-@j@8{-lrd z0h@<#RqDHDy?1Tcarvd>3i8_1E9+!yJZr_|=AS3ctEy|SGUTaJPsi<7)lZa1Ml2l# zkPx7$-W=g88flq>eCf!sRdy(*9DCfQgC%*CFC7R}2?M%aU0az^h1fdA+^FoQWGLay zellX|Kx|6tY5JW*ZmhVjL;>XL0%!)a`N2;b9k((zgR+U!9s4|xXYZfhthTI#kwgNy zx*)}Wd=%v+koI|*IZpi7fucZ8OD#810y%?)k=A(!reI$p%^&hP9zd;ZW4y12sJU(RPvUIe*vb*F;N6L&q>pY{ImFQM_l$|!r3~BAcdi|hfF^gkcY3_ z^diW4fK2vhm~RFyTDHQ)fe3PDNv?nIXbbt=&~tHh;nDg)1j+UHV^2g?n%0BumBZwMIh7 zWe>F%NpLE+nP2LP{C$4+-rcJ^5s!evam`bIQS?HGFgD(NzkLI-x9{_JpI$zC>zzke z=U={ja`#~470zFaJ0dwGZ$&2fx_NE-N(U1xz7j=J%)d{}3p@-|e0Ii23gkoEls5h7 zU_(DTF|UoVeaK42__2A1Wnav7Mm_j+uRUPJNolMEk|_WzM%|E|jPZQj#?G`&r{?&` zWu}nphGmBORb>!Atq*jJ60s37p0b03MLm-OTT06Gpnctk|TMDh}J9EUmQZhVR4Wx*WR7$30w)DA<45AJ0uk-RK&^3q#=I#Vt` zo#o}Hi$GrN&!jbyj1ov*mWaHB^3&Ol>2P0ux^i`~ZbciA7to%Fe5rE|H)JCN4vBa!a$B|z?r&h z$Zd7xic`pz1EEp2gHK_|<=2c{adJkJeWRoNTmmra4r;G|V^137}t(kX7u=aYBIVi@$ z0oLCvY@QcmEj_JqKNF*yoC8_2u+@;(%tIa4oGY3Nw8m-H3bY1D3rK&5M(LJfJ6hA~ ztXE030Ir#FBgX0rQWno|F;#v8&ubcLDk8By6ZFn1ze+jLaHq;-ZYrl5vq1-9GpB2e zluwZ&WRqE+%B#j$ADE3@8mMtqwqA6q`{BfVshCIJfB59`(beU>`*)vQj_WMO4&T;K zQeDM9`rLDn+}t3$ihZl5A8Ne;Ru1Ktn(CQ~HwM%-tXi`A;}i2ar7`ubRgM)Rm2chB zVMyDZ-az!XQ@Le+RO{5=GAtv41*w*vSt#AM0f*aw-GXy!(yW-gWxjJ#L82|oheEA# z`^wf+Dn0vKe}<>^F(Rpq+gnwrx9}D+F?Q81ABUQj4zQoh&ByRg(+|#Q(K=LsEyMOs z0sB$cmTJvCvRl}28#B_as$_K54XpMpUC&I$7G|VdRl&S5TUFt4!w`{eRiy_Hx_{|8 z{8?^QIZAF$-V#FAGhDKniMJ|*kwt50?dFKa0$zwGetOt?b~!z4@A`!uD((>CtqSpC znVS{8(++YZf14>u`rZ$cK3D4nFtL;LjW`IzQEHEI!lO!xKHt%6HOmxzwuY<=60h2C z9*s1iKhxaz(m9l+&VsCOw(faYB%VszzBI&{wr}RNeWD>Aa^hzAf-6aFn>7I}24JuK zW6^&=-NqWn`ZeD9dk;9dw^WG@Aa<$`D;TX%7aRBhUc|ARJ^RmzL z>u7xX>7o;vxJ;^V^bX~wb6?h)I(e-RO|a}s#kH#aVB;YkUX+a!B^g9MzwZo{il3C&Bepcq=;1 zM5<1g-)bGKYdF2V++{|mLR^v25vJ3MS1qG!6uoWq|DKp%;jSUDofEn+a+rDsIsvvv zif>QOJ{$FtFPhuaZCx)}2zT&iXFqK|yuX`f`mxtO`uvUQho&dbPi~*Qf6M&&q;g*c z{Xopwa(oU9oYd&HnsIm=afIFc;KaP4GK!u@9490yeWGRVr#=qfN-MfnR3S`ZzBe)K zT?fQxiFv1JNc5VYqPJGqAWxjdBN4LP%As?6c>r^MOD_bFDHQtswrw}f{fGB1zkcV*(|0eYCl@z_J53I|vr`IbP~i)% zR)DqlB;ihj@U~z*yq4@Vg`jx`+eYoUPmQZ^r$Izp;@M|zyrsv8>|G zYMx1h-P2FSdpFM|5nx+gB!QCpOVXYE((#u=5x;wSkIH$SRy{K0qKGvsNuT>0Z&yrh z=B?xfQmD&UImZFbgt~Jl)D;YRie)=Y&}dfp5paZ(l)PiKD~|ZCTE%R(E&{>SEA4JO7wL= z)7(n*TbZ_bgQ#?@k-H>jI&NT0M=4xlNW3IEVyah;_EluZ5 zmcM`)GPQ~OWX}(K)VhggoV^@D)5V5{LdQMV=DW4qf4#hY)P={d#n|1N^>kT}-TIH!1j$y#_1ujf zLGqtWDvI6uZ?{;ghIx!Ewz|uGw^pFsW0KEo%Xba;JP2jAWr1|pBL_Mz4TqzImHLc! z>u{5Vy$r`Tw|w;E{PN+izjN>M(fI?{^YYH>ylzYePP25D_{wqkIhT?Rug=i09Obh% z_A))PxAu=t%(~OsAv*2#R;3WL?p{1K&oo@rtshP!R=phKuP3J5godzgz)O4G*$(Nn zktn!Lu!$5%#BluxpXA2BU{wzM3^tLHp&2ICFz?xJA_bFVn@Hh7!yu?s8&s!G>5+#` zM8nU5y(3OE)4z%2Oq98NX8@eB5FBN*xi}_AX5Fx=b*n8IB2GRojdibXg`iKx$T%_3#nY?gV3PlooB+;xf8B- z^yUOqp&5-)$-BsUKFqCn^yg?D8xtd2T-E&#it0B+M6-`TWm8h?3AdTW7{F& z^~YmyA+Loz{Y$-hARF>xrO zc-KXyGXh-0y8vB^+unWGgJiqbIFQ`uf$-{GA^c4MTX(6&<;S_c_Y}DX9h)ZOkHSFc z`$B|zGjl%uR2S59nPSAq?mqKFgTm9!vq?b6iwj0T*aL2UtOB--Keto>b8cvA_|a)- z;Kz0aJgynyQ>(FD_2qV4LYW;B+{a(*!aM)52TIbeOrcpqzC_OV#+t6XJ7() za6t6>JQz2%v+>rFcJ)r?LJF-lnYH?P_D<#;x{R1+C5%6I`0(Ym-IEz%mab5|?wibo z_~`!-dh@9B|Jw9UW<*x{Jg7IBuPI@|&6srC424i*@qvoX3hX)BW*M0p%RT*6uoxyc zkSkJ1fET2giceC))v_+ja9P1Y8RVYcLsGc6PXz7IA}erRguaL{58-)%)JGvsQvGrq zhceSgEu228Ybe`dvYT<7p_t4HNczDEPbdZAp6m?JFD6@LjR6*}F+dcPao;TlM5qTo z%qV+L>fs+iTi^}@RhgzLfLJAB<@1l$O1^q;I$?@ z#7Dw~vJ0N5^JctI*1edU?}KYiNKG1u-TR&aSZwAPz$cg)50H_u-}ryP6gf3Z{2AU) z`qY1i_e_xX?qZ^KZay{rL`R|}*u9NCfu)>36(M)+nla<2D^$mxuPWI*z0YR$rSju0 z`(mQM8hymCrSjCQqoT#t2bFK3@X1cPut50&*wq05h zTG_tNYpZ7be|%D@+oS9pX8rx1I@O39a`NH81!zELAc}av{LO?GNfl#pcwuY5N7E*% zV;r?8h$em^;|oL}`yxU8@R^2fBtH;M{G1GE&QBM`L&cwp4~%5uBSO#}rq}%wzYsS^ zpWBk1QZ(@qG3fUwy$Z&owNBK;q#MW=_#c`PQ7~T%r7g74|rVf_SQ8NR3QlYRH}L=d>|7K6fu;H&rU1#qzQ z|F8u8p-5jwg0Ra80vByXf}mWEw#duTc7Y&p%h4SF`M;h#de64izrHiOxIhRY)T1pr zJbvZ4=MoAD;7>3Yqle}rkKg<9{+;#^f&3OcSo#1+SOO^lIl`HWAa#g3qd-Fbv!~&O z@m36I!lz@JwlK5ld1sm?AqqBaOGF`6+I8=$kt)!oXd0WUh%97wrlPxUJ0o4FDMF|? zBFA3CHP^;gPQK2IN_ZtAt^o;|#v3srUogArrXwQZ^C@!B@wMId%psY2nsV^RGstl4 zTtmlS=1)&d`MzgF8K9*@%ph>qsY>+VKnE;A$3>FG?mH8+5loJO;6UKjw5=%_PpvYt zHK33+s;&VfL^$>ukn@HDB2XC?EGe)(eSBTuM{a+3UjW%*sb~#A#GyZ0uLWyBA$pW! zUP2fO%a9SR0f;&XyvoE5I*}OD^`(+x;p?6|EHlM|d-|z#N#!{tSpsAWe1d|(!&R^( zlPxTrY~ho?{^0(*cfS1e&f5>(`SRTdcOJg;=<2-mzq!mK#W%Saha}S_tSQM{2HnCE z=%>l((2w-5(3(=9BWZlQlbyU@dL8gt+c`}{4ihhz9Y`b;_mR$Y8%jWrgCv=yff5`pQ=|TL@PHqE!u#@>|*z)-k3<(>|o5sI@ zukwhm=lO&C9p>c8_`__-AC{rSV+l$;$Vp^p zWjmD+2YmK;&UNBTmZp-8eoF&|uydR$W>7I+a_}<$Ix)Xmm=o*syHB26K74ZjormX- zFCYBI`B&e&eDw7EE0>QRUOqU#d++Ymoh!^DhL!ALk^1=RG=s`k&j``;$TyHYv9Hq`1!-JK4Nnht44ntU@WV3Ff*X4M`&&u}v< z8GfYA?E373*~|(ij@!%%4;%)IbTcbG`q0ZKY$l4kTA`CUV||ftW=*)Ow^|ek?+3k^ zA>b=qU9Y8^Ss}Kb(Pmc5Oxi2Dd^1BNSnBR$Y-ZPtc@~)%5TxM+6ba9&^-tfJ1Spm% z(w0t<_M-HB<`7BSmgvI7h=Di_8--YK)u>0orc9i+bmFugP6^~C?U)ShUzWaUhewS; zqP7GQwf4i}C@bmI(kN#_wWSlP2}gOzjN3+C0dQ>Ti5IOMZ9sw#S3tT}$%zX`5F}|c z|IG*0DwD3crvrK{3z@7Yav20{E3&7@Fq?cma^zXgkwI7X_83K&3~v3*5<9*J_br3o z3NYDM`30Cf3oyCs+RvNjm!YCRKLmK(ezfv$5|`fgDjmBAt^B7$&^>4g_;>rwo(Ae} zALOXJ%oo81@Xet8)AkOlL~c_LniBUXMcN z)KoMnJm*9vPhSG6cSNtT`cA`Bai!4ki0J7uOFhHor(w&iOAW>=fec%*RQnGx?>BA# zgpzXbA19_`wE!k&pj6K)#6(93h1^myD0wp9KIx%#z}}&RQ0BE_ui^G!<4-?$nFLr= zZ#c?Ol>C%zBdT=KJ;TU>oKWWBv;pgRfwWh;NEGlu_6wSbXVt38_+qN-I)v$A?Nmc(}?Gckbz3P-)J5{irX?N*-6?vbEKc`J9MBND&>s zh_RO;E3Z69K?WUX8qDV^-5eCogfY=632tJyM0IXy1*@ zgRX_tIu+K42j$M(-c-$$MbJqvg^2m##C(YptzWqm%Mqovqn}h0 zBYtaQzMy4vrryvq6)*ODA>xluc-#s(7;^@%+&ETs0KF=WgKVnif z^lKSaYm(Ul;8HXkrPMYgTk=Hn9VTBxmW%irdI~RfMt_FmUeS+bhQ=D){8E9GpSx9? zl@nk~C6En0P437yg`)ts-wO>(PU&%?qPuYn8>XKj4`8^?H^>EPVs zWn(-9!0oVwLPD!>5l~=Iwcq}^=s{8*#z(CiZwS783!Ch|{1176QU)%a> z|LC58NkYO!28CB$+D7ZNy$o#ys?u@T@yT!y5;_%{n-hTyFFYA`zmpcx_Mm1$!sQST zuR>}1YTn~IYxG`Vw8&Fp1_O}9i3njj-|pYEK$0qM3a z4*Xm7_~vuL7XK{}$Ou3^`K)>^z{3IqSLxE zDH2b=Q-So&5ODcxGBfomV4>yo^S&C~H?y!(W`gdjoq1{~z0c@bYGNQVI`DM6VCrR0 zYIp`fb(^%m!R5*<*9O0y!OiS6Ons3l zC4pM*Kxr#d8=7yP@JSo+^bluC{B^1mL!IA-K=P>N=He7{O+U7hNiraq63@on_)N=0 zDuRt{IVhz10aNO{wU#_kwjAUD=_g=0DAcfOEe8l034-=ovK$m*NIUjxWQdv(_{=g9 zE(eGbNj$6X63Y0fpIejqIVAe4YI|%lbI3+B{hWLHsd#~)Pk(|CaGlj#yQc&C(;XSUywcCGbpStHfmjj(E_3G75n2b_18z~w zHp|Ys(I5nLa!H7LmI3$huTji|fGb3L>@4zc&Rk4F6T-D;+(RMYius&(1IMsq=Qu{G zmpu{!##{x+0N2|%FUX%Ze+e@CS?XW+$)4kFz!xjLIi}3=?B$pUuFPUO>rMnK@Pb+! zj6L5)ZZS(Wo`Rh=2mr0dTIUE`;}VoTlz($uQ<0~@)(9);0oCz?}lV`dLYtPXb(qP zw>1Yu(3Ftfi?HCgjF;hBR&i&l!cUzc62=hwwLem@dl431X-Lpm<>rOw@Y**O{t(Xt zymJ6prWt5OV8(Daw|!H|+5uLRy8AdYul=2Y+lvsJUSK4{^(k;Sy)o7k2+B9JlHo_% z%*KfX<(pZ-#BrNh;eo?o;q@Zq#o<9J7;uZ?iZ4_y0mt!K2E)iRUziZ0*G5EPcWyS?QeW4pUfgqYOY@dTMUjMypvPngRPAcpuJ`fSwPZa*Wdb|@vgd6`2 z5r_!aU5lFibafCB+V2YvzO#g0$QQu|@VTZv!mtzQACUlwm-Zrr zzf_^|ZBt<{!ddD&AUx+pEWcsQQc7;9mmgQazGK$uTSRI{AT&HOdd~9sYz(iQmyJ#H zS-l7k;RTgwRZ*c-Ly3MuNmj^hGlqUY1Eq*$A`wr1n@{Ug(V&B%Sb2)xn%!m${pApi zcC4`=Fn9%M_{UK^;bjtFQN185Ls4q@*lotJkpmhk^Dz0*IPK1AE)p*i1*#R0r`F49 zi4CQTL=K?ztUB^?BVHsBauS6wvlroT{uJU#Kin%jB|YiTCO#1_5{NR%Jgl~%&@Bj~ z|07)5keWR{Tnc^bHkoG6J)Kv!I1C<7mSSp0uJAGcCzW>y9#yrJL4df)ly@7ayz8k- z1?}6(*AD_jSFz~wxP#1W6Z*|t(C`W9~#{)d^oO~!j z+>oymU)r(h0|l}3$59ZLehf?+AVS>in7;{o5zbQQI#v~B`sn~aHWfBOxeq4-fR{)M zFVm}Kz4>OTPE@q-2IjG*Zwl5rC{}m|IkdUGodu)GZHI!XkRn0Okd(?XzM?#CQ7tL(7&i@e5T~kH5qS_ zd|~1r+%UhOXr@vkseR--mFYqD{n4GSg%UrHG{^rM%<~`N*Wp{XWvfS;*nu~1*jM?rA zgk`8h3j_~L%Nv%|f9%e|t9SOjJReV=H#cv8XnOMeiRxO^(l?qh+g-`{qiuKNq>E#= zyMmcxx4Xh4hhZb%?n)0I2pBb29+`x3%yySE*1*F4xwhb+mEx|^<1#L1^PdnN$|Pz?@hb_aGMDL-Q!PRlAq`HQk)2% zn-LUo`j%Aj@d2fd&Q2Z5x6{ZJpJ@Z$q2AZEu6RFSxsjwm^~x&LEQ- zdoR?xIbdzlIA@wX_jvg@j}tapx+b8~9Cb|yycAYIu7+m|A58~gVmfVs9KFbNwsIi&8=T1HY0isiA>5RtA$q#Xp$j|au_n?S0wV)?d} znXqw-aCNAdU#}TDEEp+!X1yibi68-E)#%^x!p3cACla52+LBq7^0nUeH1tPwZ5#&D zM%$;=zf0Q1+D;GB#;yN838jr)0dGHw9iYDby5pd})EmKyuF(EI5WdY*AH8}1Mi&VG z7Bd0>g5QXc^aZn-5pXhKdB-AcEo1Jr-ln+<0Jd0 zTM=j)U*rF6<1H9?vnQy5q77H<1_nZYwGyaSvxN@40b2=aelRgz>j(Ie47w_-8#R3~ zFFmjvKpKo3Ld7@9;my6Ni%T0Yygj-y;pe&S}BuN;# z&6yMT0RcDJnse1lnq5V^LmxpHx%Ifo{Z3jNN!v1E#Mym5*DnJ&$B+?qnDj$ z?L%QCSETqPFcdp>j-j-)dw5TEtDq++J5ZTQzWv`qnvlBswJJl=j|X(JBhiVVG8Fw% z1)Pvp+LSUelT`mI+=U}Cp*fM>Nv6h2_%`zkVUfQFB1w9H^=z(T64GXi0$bPSdiy?q z_vz)Mx88Ylb^hhcCwKkLA^12NQk}OV^LyRAHhrao`4xYN4MS)Z?7vUU3%ny!@hvi1 z!cXQ3`H!}HhJ3_cHS90g{mAVPPj5W`{!R0v6Z6^#+lSO-j31l#U-rcuIp%nI1?n~J zVJPNL#_B|vqQGL*4fSbbJb$KbF}1qH^Mx6=%n(FH*Bp|S@JqeTxMe0MD`+C~-=icN z(stNz;TgHiAi^>t*ssSdGliH_o;#BLnp(_8E;ER@jPX3ZB8Sh-+76B6Wk>2~_~hkx zgDhy5325BoPe-rlJRc>SfCL8X5c=`UE>pnloC2l?zi{)~kYKw$VlEFoB-p#qvT8T) zbpxqk;p3Myw3&!y=R_eC)|&*85F1P0)vqi zr`oFzKk`eLcOSfdKYWaL9pDZO1_-4nn@waeP{@+%UuRqz@^#N(>^tol3{zJyIOjqR z1C%VJ1^)mEMe5$Pl%zPG#aOG+_fH^W`CLqCKW$!^?!+mrgvZzuW#TcZAa}$b9hl4p z2r_jKhY|==7yx&gPPV5~-4V9Oc|#Yl>bgU6-ZTF+F>m!o^xJo@uG)a+i!qwtFrS&e zDvD0rq91=E6ShTfAVyk_v5kTOx=2NUmdyP z6!PYPg!#Z!*6Y({yXUY}!1t35dr;ZV}c#NOq*Ofa)^f|>E_ds>&tWi)2oKsA#s4*af1j_uf9 z=$@Qps+nDw&a!J8fGRV@Z9>zHv z4%M0Lqt87D3C9idt(vB&^($CSlwXLdXAY+N{H|j^KB;ui?>Eo4U^!NZRK9=rzvp2X znuWx-(hi_ERHzS``O(CbrJgOrGB%hU##^O?x7=Pa>@cQhBVvd)KKCt^rXRg+_UbM7 zAGgeRPAaIPW%*FNCG%gU%YXC7kJRPA^}CfLL36B)UA4>Sq1G=Bm0B`4^SgMr-lt>6 zI-^DFPyxFP+du^Xs5N28hG4_(h-y?E zRN~8+c&kF>SmtI$@3j382y_6(>kAAajwQd0T(&R^>bsPeP_W-)1p%Fp7Yj$sL%zc#+x&K_7>FCKJU_Q%tj+k9$rh?3qf zjWN{wtuLJ3PjuKD663Z>;5t)_-6Yr;NOjsj=h^Z4LP_)IyMEc{Ws@Y+{0RdFmgd?a z2?4Y0iN4tN%lbB>$nE?1@_Be*+RE3NZhU>IFT==kOOI6Ap8pzpSs zoPVpH2l|d*>kI!a5p?|80^;5Dw)t96O4=2Jh#BoBR(pg&O{xdDV zBPLFLaRsNF?@dhk_9)t(IyWc2eHW3rmIDwlA67};GT%JmLtHRvhooHMuT!1C#tS}B zuW5O*sDvfM$^8AqY$Y>yppKPzHlA74rVkB6r7Vq6@-JonrrU1z=-xO5H9+c1P3#js zNO$>eK!xlkjwqDwt}mQ)S1_WUUQ?ckcEUP!%N!-(n(O?)tY_rCF5PwNYge9Z=fmm# zrV9IS)@iX<;EXH}2ia4M3|OabN5Dr0t(ykVNqXmbMyj_RUfM%yd*g;`uRwR7hYR@t zsplS2yWb(Vz!WN(-gz~zX*00+GSXWZw!8|bSw(b3mgR{{fM z&h`sOShPu(n7@Q*E_IXpWX}%==(@nEAdo7*9C6jkET*$=Tvb7?)Y@R|`7VLSRDCbl zX#;Du7XzWA05DSLT`aQz4$=f;~*l@C9IO{8RK223@~d$yZM!6exxQh3lX z2r3l_)u~f@k@q%L#=*_BsWYcd?HpVW zMw=$Y>0tSELWx#)*5Acb=sHrY$w#C?);sf^AUz_rPX~xp7AdeV#Cm3X9qd0z%s^X* zrkhQY^)7N)+VvD0ZwF$k)FC2V=DQtqh*$(-nstb1Tbapvmk67NHW6F$GT*|gbLC{c zGtgIuKgp2X*F5&J-dSj?6Q6$iDRzR`67&v@h=N{UQ~(6M)+_!E;$&PyyFkbHIUqNLT2bPwFBuiG?HYx2gZ0qYZSR{s3oi~ zw*wPO?~p(@YpG{2p8ubiOKo1VsUY`f4Pmnj$ zt~lBirgt(IlIbv+v&I{gGtb7pHr@BLcQWVr<;YWUyzSkS83C8BP`vJ&%!T;S3)kA7 zMFt!CAUo-v%!sk{c~H{z#!&E>=HgXK2n8^AJ;1u6H(Y{5B2J$(9q|8ah6q1uw(NY-8d8U|h zPwyMcw!kdyPnp2XRq+?9@V&TKA znqPQuGUqmTOailI&P=#BvIJbRQ@)JW-S`RxW=oGt-0!gC9T0#OA|>H6TK09WL0~rb zSa>Kf1DJ`NhyOj$01Lew4fq5z^#RhdIVnU0)8&sruDnb=6@P~JlK}#(DtSC*CO&Yd z6#i@%v$S*bsp%&=vNU7QVln$)rW)n0{WIqNTW!O}o}KIMF`SF;FuH zF@Jngsp_Nb9OnQ1o;uZto3wZ6IJQ6W#|z`c{LO?GP8Fke*g~-1quS1(_5;zxFXVoK z4QLD>nf3Gen2}8UoDm#p;^S^>qwXA%Onk%&y2JFkf8rP7Mr9$bY-;qO9qqD-Jc=eh zVh8;mRdNP3-{aa6)9*QuH1IQS#1>q8oIOLBXOaf@^ixuOT3J!4`FWCi^3M0_EQXC^C{ zJ6S>ZczTu)-y+2@O>Rej({bE@9T#N;jXPoEP^>Y=()J}5TMG%YP&(P+A6SwMzI>mqqAbQ^^hgFJBt0AlM$EA#l0Y5 z4y6SESFD-Oc^FuN4Fty$$Zx}g1rYKCpe-QB_?s~MPFhDtzSfilZ$JDy9!#5A-2ZFwC8TLAs>!9Fp(@kTKvSQh*;a8N<9+#=uQgVw#y% zh0KGDVcz&w6;k@>hJ})hAq=R!a1T`bxTgySRLmdjNnF4mxT$%V1Pf&?!y?l%xTgdD zzB4`;Fyf!K~0e@KM{Go%=9I|+&(vNUCsJmR)0k_aBXC3&1EB&AjF!8ix zgC4+v4dM@>>SNd8^F4o9RJ06S9`H`!50+y%{&3NAsssRH!4)+8Pnf2s4lAUFVvr35 z0WoHOJN+e^JlytnoqZ_@?Owe6T{@bwR(RkrSfra->Cw$DKZET?anhB#{m3^n0`4~FcH!HLs-4X3aKvVYfUj_M zy_Rlfh1eoL|LC1IYMDuM)JJS)hy+XBO*LXT;>=4N5y;Z;sUHF0YB|(5CY}s0GF{rj z>C#@5p3fX2NxOY}JR|5+lBk9DL5obHw&<0pZ8`?og!Vx@-^b&i_FDv<+5!}CbvOiw z)$k}IQ5s{EGqu{nsnvv|JY>dgqhKE)%fJNSxNAUSP${X`LLZ?;re0g*>NP+h+YX5# zla)v=f`n}WuxR^XsgKYm=g95rK{LhJp^wl``FhZuulNfvc@toA*Y%(`^Xoykh$x5a zL3g0KwCy?@DF>_!?xw)A4fPAE&bHt7eS{WXCYPRmTDrK9&cXI;Y14+d+8WAFaHd@&Ddl;bK3bh5vMz`U%mv=}H7PE0n((2rk=Arl#6S;W;NV`HkYax(45y#^Hl$ z`i_wso=Pl*hDSuti8E9ykzPyDZI$z~>y=q*Jzfc9*qWuRCYe@^Gi$o5f@TUs3Kvk}5lr9noIf+8} zTDC|O;z`lJmH?5tNRgsh}NyS~i@X+?O1d2QN^sc(cExArd+d|#r@*=ByT!hQmmLT=wxdDji z_+yF?Qyf#Nyf+LvqS*f>zX4Y9v5NT=xfhv;^JRSs%Udyl@4Q4iE09 zf^Gf+-R8Qpr`kvE%X#hL@*=B!T-tg*{~OU`WUvA4;TAF@`&{M&te2ejaCxCy`)Hq+ zYY(>vT(tik_NxK%u^W&gJF5=0kM@bo=UJcc;mOi-@?jq(K%A&pHkM5vD2QDX9R(@f zvPK=Xf&6gEefMueCXl-G{A}L=IKYn`-ZMeI`)Y(A8?%}&I^l;qXbX??iF%|8Ptm^X znTL%LwGFsNJg7*ayzEXU$=+1=DYDn24ma!dj`SKQu@qJ2hZFN9PPBgIQY=T5+Kzrw zO^o=hiTQ$-(V2Qf&s4nIA;Do4lmGaH$E}cqF=z0~jbnw#sqQeO9Y=2<1WH8nBPLZt z1CUX*>{P@;xPz|T;#CK|*BK&wHa-jTMe`jdUqinj;%n$Bywn-}8IF6rM=0<$^t+V} zxjEKu)n?@c=og2QL)N~Ap3?rfb%dEH-l_^@WT+n-OWuYa(vuoWx2lrSN7<@s3c~y& zL?tmc(ygjs-k7bb@VH@!$hNA|gS!QWTANu(UMStFa<*IE4e}6Zu4-=0J=P9}39)#q zl2Kl?LddqNLOd}RWAqvI2%A~$BdT<(lA&JYX8n!my6y0UNJ-rt;=0un{ri(ZxiaD1 z(h2WI90Zbcnv*k`L@MX_)Hg^8mzi*G>4bAV-R=M#YTMv-VatoG=6o6Mms^4sY3+wa zkV(qm5a(r#ai)V?Ivt#7j6DEuS`womyO7Y9VG1a4*KkA+0(n@L^i)m_XE-_Dqu?MU zTn2gA0ovN90~EY#Pv7K_7%*8a`f?q3aF=c}2?=*O6khebQY6DlRq6gh!sW%FDt#Bo zu&YY%h#`0(p}h*?BfoTc_rdG;BMS4S>GtP(9%_v~Fk;ZP0|xC9RC($|dc|=+wkO zWOQ6aR%&Fca%yc)JIZ~=Chc!<`7%^MRb_+Z=Zw$B4?{1S+tV%k!;lMU@at?vq50m# zus*I0exMFFN%IO>p_avnY9igD`Q{0qv;k)hai+vyr#dlorsXY-FHb4)bnQ*Hl1Va< z2}(Rme!Ne6uae`0%RwR4513Nt#kAysvgIHLNZGOfgINwRBEpv=2$utdj08b@Em;l< zF+|5cMT)eduDFf}mjgtJB%Y;R5=4AT>*#ePDJ!6#<7c`E8X23+e8GtUZT$^UK)I)% ziWk_l2vUPGlD$~HNoCdA~3fUAO4jTFTQSO!139dq`xfQq7TmTs> zplRceC_+SOL^DmBd%9#qr9wcP2MQ37bw1_uI5H`A$b#$sOAuCBz4gk~TSp-v`{iO} z`0`3W!{#3RtQJfo<0S~znIHP)uReMb@-~nlQ|@wJ+6w_^;6lZ0v*gFw|GoMBmGMmo zmos0*=<7x;dq}^-v<)^Pw6|CeNsqr!(f(2#bMnDXn7hXxE6=zGLO^?oC2v;08#snt zL>$K`^|D7oz?iE58Q=<(mXm|@moUjq{p&v2bG!}sg1#r<<(TWO%wjt0&h;v^fV9q@ z?_vtSPBor_oi+*p=`v59gN8&J=u+EcYbx>-*cxF4t?JaVwT3asapm?RyuSP3y-R5? zLal|tv4szTbt zrSMaQuN-^{Ah#D`;dil#J?)!V4@BAu?cq?3CV>Jl?o@``zNz51jF%NCm}aaY5~iZ) zV7G57SU)6ata9_hb9n8W3V(>_K^#@p@G7Rm5&%76w{I$0JHTpEcOS<`&F_!wUWC~6 z0wWo&Pl3DX%{V=QpnNkc8GfYAY@A3?zL^zF9JiSj9ykmZUN1sk9Ck|uH6s-5dm}b8 zg7!A&c1ahgX(w~<8?l)o{wrKvnY{>Ev4uZirf)Nf`&%tDX^#4c%?v?dsk<4wa6}S{ z5t+~!(Ax1i4<^L17vU-s+pXAL>TpC|?JEh=Zr>iy2uVr39G?{j0pTi>%dMPTu7?!^ zs?@e2=)q=L?f5D*mRof-mWvJn@;4065go1y2-7HM`nQ$Szp;7|;!*bKI0kv!L!cL- zWr7lZ+%+K4homr!50U>{_?F^(vi;QI%Zj_sDM92}$GQ;OS1xB+N%_h))bo$-nL|-R zXnhND@@(GoN|7mp;&Q-gCbbAF5Q#lg4idsyUR*w(1*YsS*6oO)OhPyxg5_DLFrO_t zjXuz2(1HWcHok-~^dek^798>Ey#9OHWRr>rXI{t*M1=Meg@2PC?|ja)cguJ8-w=U_ za0V!Nd;W)BgmWA24!*O5UdR{ThBf$f(Ee%Ii*TL#)UkRI!Z!#c9`tIEHXekJt)Ta& zUHG8(T-K6ro%#+4&p83h&%n~Fcbutl`0<9Pb_7Dh1EaUgXI!523>P1LOC!$hMR65ApQN1+hz>?ehZZ98H!RxBpA$V-pKDS+%{v#-XR+8SYtt8@CwlI zkE3`Ji*eDjD@!EhMLbth8q01ohF(NeCXJqw78Sh+qyE zA_q`~uPJ(l{zNPi2sw!s31%-sMm*`yh1&K?&5*)sAzma9Ws-UL7>=P_5aH6A)a)UM zR7(nD-@0|C*>g|ll`Rf~$CD+>hUl|o3{c+Ttd0+rf&kIx4?s(~cFMaRs07ZFlpV1b z;Vi2!T!(&hYwk6u0fVf8mW~e$5n@VXnrZ3W(`92iJGoPs6bQ&oM`W z^22W>0FwmKS3m{{;ymCY`#WT&0H;479}7f^ESEiqCpXOJ1qtHXbMm1Ckt_QQy$I)a z{;*3a!9oZ`h->b3)vy<#egANWBddxs{d9mI3%pe-R8gki5ghfi-hAs+Co0-^BlDnd z3f4Lm)*uKu@nA2)b!Wk7^7P^Q;D|cb^i)6vAYjzBoKZQsng5uWUu#Ha59-15dzTL% zizSJlj~d$7%q!EEI2zje&o;D%a785bkH!{3|DvYxnUX`-WW4zEg^7P~!~BAxnO@Ct z%&W%0UuE#K}6ai+R8RnimQ+2y^*mv?MN@s;}z-=3a)JZ+w%wmXEH z@S`pXc^coaln)yiR5T6<@YAR4`aUOyw}O zyT)QK46_M8w91M2dIjoP}ajaX#SK$6Gz0cO!@oHI?{+G+B{ z2kHus5QmdJ-z#H6+oCq+E@Urh~9?9pq|RJ-5lu`&}jE)lK{ z74xuvMKeO(0<;(WZ9qE_OlbJR#^3Q$z)ff;5}$tBl3A7VHL0}GDqQzI2Y}!9j`Z)+ z3og!#x3e&O$q^eL z*+1QiK+_-MGqmI2Z}tQ=P_*HS-M~NuqRwnB13#i9cKpG_bgduY4KnCDLV#FGNi`jX zKoqFJ)PEB z&6py;m&%up0#eXlID+d|MeVm?f}H;01D7?{P5{h5JqmY`tyyeKTq_DD`*V88J)NeL`S&A zWdm+;A)_M+BbOO-nsZft5jQP$g^><0K-{zpxXJ!boFwE_vI_v=T8Ddv4OxT55?04NFz55En6_n{4Jmf76XNZx>Xs9emtNP zS~Dmp7Dc}#IrQ6eVPqzD-w9RuS{k$zYrGrI~PgX6@bz~ z>KPz8y%~!FTi51#`#yj7>E)xh-g$I&{^iRjcMmpR;Y>E9I@xC{YB@N2E3&`W&1=(F zI@n+Fl_;iS{(WLz-~plHvol&!NO`nP8B!12L_Sxy8AMn{1pD=vWu_2w&u}FX?d9W^8AM#hcs^wU+ObM*aWT~d`Wc*R^2y8Z23gQ1 z6VPm&fM)cHo_xf!2}oeF!xdkYu-62Is?<%UfY~?&Ophjro211AJF80FgqBsCEpJCk z6pGZa@bOC;+DycsVPb4aT}MBI*#=$59#Fcc0|wI_B3W_j zX6Fgr7Bp@mXG&r)tDM2~7N^dYi&N)=;?z}OFs?W?p1DZdG6u6o3?>w(&Vjm)d|qHM zTPS1!n1}1KuLlNW$IdYrrmkRc&IKH1J7X>g{{RU^>fW@Jq&R`atnn5PSi|RHqWWp` z!gME2R0RygC9a9bq=MWLdvs(nEA7+1z&^_Fx$UV`cZBUh#4-nAu6Sn7ZPycyh zJ~L{#Uw!=e{@V{DuN~#eQ{~9RlN^v3^>OK+f94w}=G9T&cl+4%`8RDv`IGxko^rqA zl><2)4Afmio(x8=IEB19uo`7Mv|!E16{ksS)Xn}$T5%As8MESCk6LjGk@qZD91*e^ zx#A#ZGs^o%TXC3rnXQsuCiX7hW`Y^__|x(0d-4*wjK+*xSv9lGR5RON)eQG5cf>YX z71uW0S-I_yH5||YP|fgx6~Z|tjcq2I**e+Gn6VXe8=H>^xD9t0ViW=Gj1c}gxZ@hg zABH2-&TMn-3`mN|5RWmE6>V*Uc4o_kQ;FNGb8fSv)1^R892?v#P!^z63u9kYl2w&v zg?mjGgHo-v=Qi$l*vkNbDjURYLY3Bnd(my=HWlp*w>epCt!~kj<2JWt=UxbIxG@bo zQCY4Yq$RD+Tj-PUKI;qQ25>JyXQ}eDbvTB*=qRC^RBbh8tB&jjR3IrIGiyai)QW(Q zThHazr{b$I)<=3nmmn!SM0CFU;lzBYIJ53Qd~*5d>hj+GyH761bs%Ghd+R5u?qeT) z?m5VC=ZmrQHyNV6B zF(cioN=9d0!}QGgAhxQ4d1JP!!sCV^BHOA;4<2+0(;J9qxm7ixvYuH#VOtdf%Az&2 zcGpB>0WZYUGud)#xdm(07Hw6C9LwCSzg1l~JBi$iTClB>%UT12#mjD}g4M*_c4rus%*RMrZ zn!63czuiTTcT?A|O-7Ly;EoJd*d0)^@^z*QUtb!!er*CucGsCUUn|8>*7a*UwNM1JPAck})VcNu|gpAd!`CZT~-d^QI5i{|!p+ZEX0 z2hV=me0YC#%JgHeef0So(+^Ego}b)4c|XH`6iT0%Kc7@?xF|#$_~jNz*D?^FQ27}B z2PftYl~MF$;y58u=`1aCKlM2~E!byTql18j`QF5^w;d2p88_E@JH#D^EuFGyGNFBk zYq8BYPxufQ*g7QT5`UfQ1h!frlxaKhC6@&4=I~kkn)0&?KB07 zbljch`UBf(N~UHUTxXcZ-J2oXX$q#vcbdY(h9M!@X-bbB@NcaFK${7HDHrBfu3H52 zg+0>A$^z>|Kuuc}5r_*=r7r4j@)_38m z+#T8W4m>B~S0LTRZa@m>swg#%XeQm=Iq9xoL_74(w+y^Q*$Nm1;CkB|JK$cIG0ZyX zop)X?`%{NW;jVQ_y#l!}7xd1HtWJGrYp?xN$t^K@z$Cr%HYe3Q;q`v8I6!LUI`ze% zPJJ67b>kA{Eb0Mq-61|5>eLscP8mEe=$&`|%iOS6;4ZqAbiR2v&{zBQ$I;hWR{{fM z&O7jhgIosx67tB@P41IDKOCSdDTFU1YoKWT<;b#DW-*<0v#j#bz@6_BZB5npf}J+9 z)>ZkZ$Z^G|5Ma8_DdZ}*HPzk=Y>kMOZZx3t;x)iZM(uZNHwSx({HV*4UyHH0HS6iJ z9*gxKtBI7Yvg^4UJtE~lnN(D~_1|t$RSom-34u>kzgsI`?lH;dxaE6_dme=H+p=i7 z>yd-pR>R>quO+@d-J0Hb{{5TAwjQxfFdscRzkK-X@7%k5bpF8g#=NsRuN#vgR7@++ zG)zHxt_`M4bbNSyw&>Vf`$s2c-D&L*qV{^LQp#6%FP@r}&7?WFct-(wt->gozn++K zj~c?d0Xgk;XF4M(7(jhI_Yu>-i4>T`aQz4$=f;~*l?XqBO{8RK223@~d$yZM!6exx zQh3lX2r3l_)u~f@VBtV!8B0TYr9T-cW5nlB%E`tO)|cCX38i;Ppc``j zS7NlT&2ljo`hNCK<{ZC_jjjS|LGv*u_0;$jR*de+jDSm5C|>tX=0bcNbu!ns zILfKrlNm9VJ`d_mW>Rc+GoY2t1yE>Zn-iZ+IR0X6V&D`o_IX+M?G##q#gO{oe>3_h z8A0l+5Id>rtGvu?a{={L_TT!LialCKZcN5j5ECt7?W^@*NRqkD6C+6him5RqC5|Oj zOl>ZlVybH_ha81;1`wFpY)XJo*!>fuyU5l7e0t&qW}6GxIl%rKq$li`i!tGgh5f8r zhnOHmQ($UDfthvYhkp61kDmA&T9+MNTMxJdTt+KgAY9i?6B=qReO0cng0>mF$1?6R zS}H=~q_Wc<5P;c`lq|z#v`bK%OFh`hJr5AUW8s0o%vOL>f1cj!zXuv%IfbJEpJ1jw z?)Vp204Y%g*W z_@q+RN7*^d|NA|4su4GM?g0*<;vT_TY5rzH3#W=vJG|_*-=lF)(g_T5{}WC8LhcvX zfW}y9@d3%i&ly2>WNK}zL*S?5iuT$gn)rwnbcgA6|HLoE4H=wAUll4)K{W9ZJLvbQ z%6zBML0nr>`Yn<)AlZho?Qsqa0ZJM+7f#YZs!!{SC>*? z_^pUlI-__({lo!+@9eQS4c&8 z-Fs_f3oCr6NNK*x23%7XkRzDcnac6H?Tn0Jtzt{%fJ9FNf?t@wn3&gjQ3>w^#nmAp z-{7Cn--xmMg4staMr0x z^x!oB;07HRNftZX=y4;M90LJ@z^n1JaHeEDxz>r+fI{vV)&P}v)`AYAH6W)Bx-tTt zA;oG3*&2YjLw~eh3)X-_^c;T;kZhhtv<4vdAn>a48bD)+lzIUL1ODobC`0w*|JOZz zm}LrvnNu)Gm(=7Cay5&2x3Y|3mdO}qy)p)FsuJ-7tP0u8f{bC-D`S``$r$2*W-ey~N zsH2XnJN}}c;aTGY)|#woH$E@ARy*eRZ(%G96(-{-X0*k26I{d6#%ZH!hH9dho$lEZ?n)G<&=tGyEkZPT24#>0J%yPI}-YxQAXioX#?r{ayE8olz@D;AE*V4_b5L?GssB38m zYbPGem-=?rv|nNAY3;!$6jX7f?PJ{>CJb;Y^n{bGoz_rROt;NRs)@A&y4P zfkdsc;-6t-k#sGz58BK#iJE(QmqZQVq_z#ADa^|E@$fRFpi`T-?|K9UFoAr|mBeas zl;@dR%{^W^%0p({_7LbJWN(Xa}^UUh8_p z7&2Lj*`v@(EM0pji2VHedU2F8dL~c&rO~LOY^xMAVZti7r>FKAXiwo%-tamL9 z{e)J&6cG@3`_anV+4giCl-N&b?mr!ZenP8d&x4)@>TVz8sJqM;!3I$O*j{?P#W}p&t3hwQ-M?BMFLSKnTKh|#H>jzJX6X{YW4VV$tFVk zVytypaW{90yN;gJ3y?aH3wF?#t+Kkuxxb8U zGw&!*JB$G-qT|;+#xc!AbaN-76OF0oNR%()gohXPEo>E#Kzqn_e#u+g?jF~v!~ie?#sp4$Z8+w8(;f4=Uw=B6VyH~b7s^LTnAhPY9GhoqR>ZaHK=`D z23#bpeOw_@6lxz=T%D%%`2n7IPCo3Tw2Dzmu1}6lKYs7acV_PTO-DgWx2#b|Z6H6K zlh@_}oc!&?1lG0Y%xbz2WIQUz`RCkD9jM3Jw(kxWiuT>e zJm^|Tt!-B%Ma=9?b)O=8BkG_p?bEWia@eA0%?~H$OPpx^%B5J2D76p$WRKvG)Ng!i zV!ohdbf(_WGZioP{OkOFe8S^a$ibL1c;&{iLgZ9;7}8FmHxLrd#*dg(4Glm>)f5wE z^h0MTVYUHRP!o3QL(F%Wd=0&jh_9ii@KR^=XE^RFyQcuk1G%DFP#U{xvvLA#sRXj2 zryOIRHC^kB7Og`CGBVVUjU{hG59vt_rCU|W=%Z{^*9GCo@wcjid1JP!!sCV^BHOA; z4;}~%wN{0~Re|k~$VR$V9gXecsq_=JRmmtXGE25q72*kh4n4w_Q_C$_tF~yXlA&JY z=JpiO-^}6_gqsD>-Qg2GpZC5W#Dwn3C`%nr50uM2o(Bhqe!1G$9wgJ-;~2Qzf>h4& zsjtN~dwgoUgF%a&aIS~?0%p>-A?m`insav@n17XyXo(nn0c3D^jQRXGjd7-fbB~ve zu?N6Sx5OwgP)KNtX$2IxYdB&^k@7Gk7_5Fs2tq&U@60UM6OkPM>P+#uJ@ETO5t%Yz&rDsf=*#<;*}vXj>fk_vmr>=YlQ%TOg1Tu0cb}<}%+5YX@7&um!SfxDjWQ4 z{2cV6xjo&oKL@$60Y5-V0dvwveKj`Un;6!|wZX4vaFZ{DEMQPAcc9D?=@!j5Pxzz_ zczTF4CH^|qi6I=IAV1XdWLe5fMv-m$v6W1c0l}1bHlBFTwCpD>l`RK_R6k%!owwGK z2g;U%CP30MvOUQ=mOj~XfRK?OXs;#9K_P~o!E#XBOUX$hTMiHvuOo6LN{U81wtUQg^IjT$(j&uhNNd5E>v7EnE?@mfJ@J~heAN^ zYCHD%Hzl526>RV(*^tow&H| zo38IZc<)l$i%{A=r|_d{@;`1b!ooXXrO#P%!}Td}H@z{|69~#T zvy$OQ+RVm@1m&Aq!NhT!S>b`hVBz&5#Nl3g^ar?^7372?HnW`eW@&zNS#!&--HRSw z-n;Yg?$zb=Wcr-qDubI2Izek;?=$kv4Dny#>dNdz$cn9Jw3*d1ljf+8*vt?VmbyEE z5Y3gpZF=B|OHymcmk-7+b<0d_=N`}NQimf-E+AZzdO1ET4g$huCYN(h?_$M(DzyhD z=)u-m?f5b@mRt6=gxkZ!C6K@I(LbWeCXI5Ye{+wQj`BDOp(SAge%v)6ai);Muzm^Q z()VOL=xmqnI;RAYyWTK{Ojc69TzHba?CSKEB!rtBCj+kCj^fCam5R$lFT(YpxO@|s zvMVlczE)gp^In8o1eQaE`FhjyP;2y+`%1P1EjakI4M_`4=ta2nvb*$jUjMypNfnC- zy^tA*2<;~dZ)cBpf{1YGzac^q;kM^N_S4ltL}=g4!FQI>3;810ua1Rb@C^cq=h#i5U1pC zJ*^qlGfRFNwrM`A7vUjzpc1VrDwIkm(N8GJ3b}2@(C=rUG|MfiZ#|{7QqLs_sA6d$ zw-+I@cZfzi)>sf2yaF`*<0ziUOj6IIWhhFGFuTndHgZ5iWgd9@bwZnDWCVsoCShrO>x-nQ8Xi(|Ki!!{G5`i5$=H@ivOcmZb8I50!!dahWOa z+|#?D68HMjC_v2W3zwna+>-m8h(Ums4$B7+GOAWm#xc#bbW5kDV-<4`$5bpqv~nW> zLIOomR2aqRu{ZC2J#2W`ZG3tZB#7>YUyJs9m9~32phw&>4Y9HDg_#HC;WFHNV(<11 zvw%JnBnjd+V@4b90|GAE%+8eSN-$}F2+>{;K*ezK1?)vQOP%XjRg~$c1N@*b z(3MtErr#-p4bs%C7vU_`iHi2!$UJC6wN8ZHQA7Khd1d+%M?+ixSv66jErR|D3&^44YH^VTp_OnXb=~{IJb)&3{eIPCOuW^49*`Q^>(}y@B%Z zimu~dPRuJ}4-8qH+?S`C^8zpOK&q(g<}#OSj=%Zwq~>dpi_uM`S(Ik#8oB$TsZ;YOY-yFUW(Hk&gjNu7Etlw9fMDx zgRIeA>N~=OK`8`v~?%w1uGpD&5=W@8CrsSHlkCqv_Ba(cO9k!phwkm4uD5pvg+)SN1Mk z|Kf3uFcya-Y~1CL8q~?R$0(9&rDFLYY}^cr<#&NpXT|dE_k+C=ZI$~+e(Cb=gV*mz zG#4u7H(i?0AZ%QLb|N@ZkbtrJ?dNxVKfzUKCvvr)&90qDOJ-Hd*Lv5}aLw|@ORN*% zU)nyc{$1J;6@xEcv+S-ua*+?Djhn7Eh4!P^UfSsT6F8_Z^+vFwOTPMjP=r5Aee~x2 z8=WWowz&ZXg5QXc^aZn#iHQYb6}P_Du~ZL0rzj{&I6h!?Hcm$G(I(4 z5E~yC7*4xdDkwDFQ#nZ+3kKfo32LBd!xg)Mfd~|w*;)pEL`m%UgNf-{KfoJg(6x{* zY8s1@TqI-6^IPgurTP1lKH>*#9%5Fh@0!)%wPDAxg_lo&ST)C>3|L7||MR4I&2`mK zw_#78f75J>T^X@-AY>(6I@-0jmFO+RmTIW$3r=T7fN7de8KYnp2=l`qM>XnqpCQZSXEE`iOx~^Jf)~)KU z<%+GvXi}C3Fa`sRpXR}D#`tv>V1RiU_`$%ui~-3O3`(LF%hW=NmaL7EII^Y5JSUgP zKQkjEH!rXxM_Fz84!wnb#gR9V!zebycS(MoQen zeTOI;Bu2`j|0&Yc8&RoDSwoCeol-4`0-zc2&;+WohXzBDhCnF~4P&bEf*L`QZ3JoV z@JkR)a9#a!m7&4M4Ro>&BPeL;$Ef=ZzLWq?P~FONp~Xx#uKtx7g)L$NO3UBftR|v# zbryCq(+gn(eNRY|TYQ^g7)d=Hr4SMc1?Ha3)$)Dp=G&+DU%z+%?CPgaAKuh2hlH7| zNp)VgnBU9d#pMeHm|yfmm5`FJEyUBbBU5x19$G?A<|6q|ZS~}_74gSM;>8)VPf?RO zdd$9mS!Q$LW(Md3JC-dA)B)R`3}Zi^<7Of@$2dnlM$dFU)6E2k%N)%g%FPhVQG39j z@xpY1S{ZqETER2obA}ewdjuLa{A#vGcRE5{r%j&$1tz?bFNa@L(|(XXD+WvfC#a<+ zX%>F*s#1H?vPw4er=?w14Je0&4!^ji^++teBC*U^S~dnF^Nn&E5P21>_E2621A{TV z4rxg@w!PwYI1HxLLhbWte z!E6Ey#uTRxRJ_V!Fp$UsFc>Nt@>zqyNGwNTFhpL#n8x`GH8&7U5Lmu4;Z^#>1bzj4<#@tdW}Rb|Tyod$qu zs%x8p(|rC&T%V==3EgFu_HmzK!r*ajz0LJFk-XVpHI=7Y3#^&xaeTDK%QmsRPHlz> zbW(dQDOit~>2U(QW{%CD&+<4C%ft6jaPx9Fca%66h?yQIz-(q|-=!kecPh3M9J8Vl zU&i(>?>)jyuLv{q{X1?`#xFaeDzj#Kk2TYmvSz4xoq=rzuZpWTteL))HM0e*8P7-3 z5xHFfmt9-c$;!c3YWP`e%f@YF=E=ZqsLNn*8`;JRAsva&yWcN820B&S84E9C0TC(e^gEHuysNyp) z$PNB5CZiKo{BsPbL_Kumy_B?fC zzeR5Bp0|3iyeuwVdt!NX`AGaiic^%ELC7ZZA4E-^73!%VyjTmKNfm!~l<1ycE}m|| zawro?+<%wPbH5DPLfI?U$HnJy8iEDcOF38>%AW%?VYJ z;ie|S3std!_g%tl=2>qcfLXw?O@1c5boNEui^@ZjxSL`HrFaZ9STw8a@GxWb8;kJCR*J zs~jg%C6=G*{H3}L0idsxalYq@OTnB;0nT44gAR93fGLYl>N4c5$NAg(oWG(OnpSax zSDL#u#J_F+#YP~m?fNBk3o9@OP;*8yyImRP0G=~t-dCtcEaW2WyWtJGzSNXqHU>&o z&zTl%3$N=}8{o<2Inx1@osxt;Q~b>>`ic*|>z4$V^?2I(*wI%_AsVIt=8`>TI(`kh z>=mHzvYDKoM~TACW2Reumk2t34MWL5KWvzi(jzOxoudgOLD7e0CliN1yADwCw>*-8 z%>eCfcpw4*D)}vs zfN3p}5|oJiW*72p-oHh4BcO>;1YO?YeK&4CxbaZF0dnuY!6{y~)W~PVwdIheMn}^6SV{kj|~A_p2Q}8 z??}8-Wfe6taTt?Gbe86+--U}>3lL%w$o{wD zrHLuBA--@#2XTR{Q>2`uuT>qxKCY#kIJl5oDW8+>P!I7N3o&qZ?t~r7(QN%4glT~p zX(A}n6z;4WY0lrZND~wF0FmZAQlNS0ND~o_A88_kO-+ImX=20f@o&tmYj2|S$Q>;& z()h5qmv^HA9W1U8DAZS&ktP7T8UFB`6KNtwgE6-v366^el|ob^7hp!30Ndthw*Dr; zEZBi{C|#HiX-f#^<2y*lqYG~pU6>uaaVs$(mK9|Hl;1E>C@)x7<-t-W-6&6uUR1CS z&}VcdEUVQkSl^nba*4&nSDYG9>Mp%t-K?l8CI)Iy)Lq7kYQ8h33A+OX>=h{MFByDi zs`3r$KcMfr1lV*k$uzwJ6^$JzO+D-4>eO4JVH?XBy*l+_t?Sew857eh;5)Y-Qa#e@ z)OV(}1F1()r%p%qgU6MH;X99$I`vkA)apA#9eH)?U4T!UI`xsp4Zotgcb+LfHt3tV z$s@{bN1WH227Q&cKMH+a7bP%&<~*RJ^eb-Dd=Wx(uAAHkOFnEs*A4|4pSN&TCz|ED z7_Le{uBZ*m^PK}8UG+U9(psz)e?0l!%Y_wc;$(HT_Y7GL#42}sNR_sxbJ#-_x?#kx zr(O=$eE3<9CBJA5brAisO&e&U&yZHAD5${ohNH;-FOVx?psqGip#hw}>mIsjtCNcFN1~Q#L!6be7F$9sA zpdMgaL%hcuL?R;bf=FbbsSzY95RzLbHe~f662U+7JB|@Vd~`JKt;yvodf!Ko(;5~y zjV+=)hy?g(q1QVv4I&X+2~X(&`mCX+_|054NHKy)0E(8Xvu21upZL|s4hsbVMyz*m zq9Y!=UYQJ&K z9U?WfUd2!awj@Uf5NXj24Xt}`*jR_YL8Xuad4 zLqul~Q`I3N>&Bgf@MyjL6gC}An}`v8%y+tDr3c+(r)XatI<5DC!D0t(DR%VLhsX(H zj?g<5!Vr2#T~q)Fz0#oc46GvaI=4^w4*F&c5PHWVGk?oK%F^d3NV!Z0X6~SoPS#!& z9nc|>t`p|_soFPqGS2XP+QM{eadr8jBAAYwb}Qh~iRQY36rNHHm^^~U81v6*3HEK? z*uwMU1z%9t4x~^}izMTXPv*;>xSG$)E+oSGeBLl2^)9FrH#z?c)*^e+E;mGr1VdM9 zPWp=rLD=QsyB_Qp$+hMSHqEpo7e}Lb@4fb(9bDM1qGm1X>wW% zM~o;#i%vF^iGj=4`K}E8HOM3?hz&1ciBqpF|Zhr+G*b>YXlB2f$ zlwd}Ej^rO0G(fJMLIXZPOnm^gOrks3s#t-~1toQm4ww1r@~Ms1upt|J__|5zJ8 zlKUBKpvH`A;RBqF?_&f9+RnvL0J=R`8$ZAbip})AeB(##CJefj_jU^YMmGg#;|JJ5 zxkjt(cl?D_>swsjiBcNq7=mn%vrlq7O2eoq4Ji9G?L_Tdar3tsAxcGH;0PWSVN_Iv z5*&e=s@a^;D_0v$53P}kM_4tWv;sPgVAs>56^x2jP`sXOEJ5Z*3@m}FdaE+@em%ep zvc+nnxnJLJ2=#1@D^{aUPTNIy;uWinTZ1JG=&rDP2Ve<0cMcYk}g)0@#cFay;&@W$XFD7eQVRfF0orz< z@#C(+5~Lw0ECIb6o;W~=Cyc~-e*t9Qt?MFnjNu9XXAPz>gaH{~3eVbPtr6?x)xuQmhgP>9`aIr#BG@&S?D zv8oTK1v;>NfR8%NAOI=Do;GN!vZNtoAy_^jz#Yo1^_<}YBGz+wAHaehSUw=Y9vE7! z>;rHZ!k`bsNhJouXZPrX91mRn??zv7T`C&ut+t#bb!iNPN{`fi<~OJ$QvOs3KTuN@ zn`XkQ5UIF*nkm0qss=z~px-zLu!OJ+3TX_(iJ~!-FDM&-kT;nO{DG=_jXw-0fXfhO z*2VaP9M3ZypUXgBnI32UYrk9bn+z2jIXs!5=!Mt37f;P85MGnQ5D<*?D|xTfNZy6o&8-n^t< z4vrnwDE97QVgVi{t-$MZv;UqC)FHKa`maY)<%${;|M5sX$2G(24<6ipg({@>99 zFGY*1dix-ZSO77rsJ3_v#){RnvHi&+`m!oTO;g4W*Juy3^Rk94_`xg!iyS_fCA;@L z#$Xl`+zV#)OnP4HbAB+3h&V5pMFyN23pbd>hTeDisksh}7-qNhj9})2yMGab+4+oM z7686X?|RM+W)a(ZOu;ObndCh>KbQqbFjw8y@+%%A(XK%{j$2Hu5-siX3xfrxmIaMNPL48crZZP5xAf5d%0WFN1Ch)1Bx(9y_+5)T6-z0Z%PI zm{mw6vYK0E$W|Lp6kCnC${I5^HJ5}VD107NmWVKb<1zmgYj*D9p%Qyd_YoRSAbU;z zBvt6xYdn*IC6aPcjL85fI&U(rkI?Aj$fbt2hh)L?pdqM0C$s53Lfwh3_%nrZqXCoE z^PpMb+@iuhLfuJ#DC_4z<>&0L#IZ8Pe^{FWT3Y->Mqx9hr080p=DSD zdS=#s)H_!Wqjo)MVz@uL8XAP}QtYwpOv0lTK(=49faoZ;%Z>$zs%+PZ>S`yM=Nya4 z;~o0zr6DO`KC4+@a93i<6x>4ecq^%Tg7Z(q60?rA9#0rDY|WDFM}*y;9TOi&&3~Uv zD-^%I5Cz@>i0A>Os8kWSv!0+qR*Q3(#4jI}pmjjrDTR=ywP+XO{Cb90KQ@X~G@5ea zVCZxg%A&5((;KMa37wFq;fne|t=b&VRYP5tlZe0r=KikM_NYZIr!F&N4kFz&)swz%$bSbK~+HF`p4wd&4p(W7G9P${{}@%1ii%zw~cBfR8lAEh#%_};hf zKh%UgnOtMbU{$EJv@=qr3sR86;9Uw)f z_Hn%?j%T=sYmJjn`Y3hOaU?eV;L)dVtUD>9;vU9cS;LOnfPT0J?VPdo62A;F0oR@9 z!(|6x1AYKWuUo`oE4pPoK3Vb_s6+00j6GWKxL>Xd@%sz$eH3f`%qg3W$dxbnWHmD4FD%5zYYFYKH%Rfn zo<72epje6G&yHv*X`tW`XVA=zLz&2`?lKgB=4=lXC@u1a{T~xiwO}j{Rr8D~3)3nW zGp)d@XK4Y>CQUCwg%rO^#Mc7JS@>Eo3N3XGel4_nf+Ynm9S{@iDL(4bmDeWa1firH zkS!R6pdFhGR8s{zR7H@HDSylyv1?7X9@UT=s$xQWp{ky!@hIf8q7xY&J5)u)n-i)c z!%auwn2*6u8RPj*WsShQhyuq)1-G-_JG|UcF zJg7I&*Sis_&K<~w^6uz(?>mM}xb}#5Yel@9aS(6_O?J;Zkx=Fw9rZO-!nH@7TPxyR zLAN`I4&`MqR6>O`19-1D@ZPq-TAV@v4bD5VCt{v@E&KfDokyq3Ctnno7t0IBPwent zx5ge1ZmoE5tTomEa8@nW3gCl;5=G;nz-yKUQF>Sa`6e5@86x4@(8Kn3Byw~B1sBcG z8jM$qz8;Lh6tNO5kx=T^1BGc15;~sp9vQAwm2QZH3O|R)@Y*25nFk5?({eWC5eYX_ zczA6})7Jn>&swP3YdeD726XM9Z?Ax0C?sP}j|Yi_YIGWG48jbTMyTL!rEdVHOWo*M zmF%L|j4b*t5YPxEj-A!5TRGWya}S_P???vFCz>!404O^<1J!=YVa%=v9SKEu-!~{a zUgXXN*4=j!z3G}ZCX3GJE~u!OdFI$qsh zFSCt;FZnGcWYYc$6<>x7s3dLhReK%ujJUQO$k#z?J-`RZyg>X0@HZENbh$S9x(9AX z(8^ju23RG|3B+an!Vw*51Dc+~nH+ts>ex`Y(E(6dOXu;Of?YPM_9g>olS~jyj%IDl zJFX=^Dk;w$M4%%BQ`VVvEqWl&9rys!!Q4Rv4|@FWAb^Y*Mtjb22N4@eoEIc!W(o@NI#`G zY&;H}8odY<0&jSu(>GK=g{nC`Xk%KZc0c9#yUQ61Q@+UWF!8INI(GMyP8?WA4jSkQs>g23t=<-Uv!&(jeY&!JR z=+tSO5Oy0M^P#>DXo-N}LB-It7(_r_#b}6t8w2;K2Nmtz6lODW^8@Un36LI%s_ucR zfjFKa0&XF5a17~06pj(=Wgid$ZLR`nfE!TZi)!MFFv)fO>poa=yfyGe zM#_M7;J9nY&)Y1q6U}m6oF&GO211_iTng`MJQzsWP)sViduE>)i zs|71EJN@vorf%k_?VEO&6_2|YVdSIL=zp?gi`4usGKcW`^{U6J;FhgMJ{7qq3VIGE zPY~4hO_A@0B9^poA`KXDD{3!C9;0(HspgUeou0_;n<9Gi@Ukqt9OM7h6fHp28A)9soXp=4y}DtTv4Aatwj@!vLz z|3>OX7_PF0j_rvpFBIrS*h!8H13xZW5TnBe5W_;FqJzkWLfClEIjPYBh}`78m)#hu zK79v+ZoHE6jVzPiUD`CDZa%MFMW+|E^joZtiEFvlGmN^ zD1>bQESn1RO%at-3cVxOHE0WtfGU_oSWGX%M$_)x=(Jvh-e%*H2-TI}ghVK>D72g< zDmNs;M&BU<65$5irsXYEUR~`;g!1Z2f$t=uSD*_cL(!*UFT&3CsU!6wgflQme0$98 z8wf{>gX;NR5MH`Fq~B}F*SWp}%yW)~@o!!6z!w3aM?zg715|rCi8bWzfNgjsp;M4ws19 zi!hkR6dEn?vA{5BwbJ3o1AAhZ(g$P5H$8sU9&B8PxRTq91rwRR%)-;K3&8cZ8hQ`S zbrKO&%Rru9LV|s~niZh5M75->#mLePYb{X(PkJ|;L;yMQw8nE15qqkT-Jjn`3otAsd zJ2iR{s6>6f)X>Zn-_a2e8`E#D0Y!2&HK1HNIswtHsmG;j6qk-viaK4>DUO5+^FzWl z?voWm3<#-MZJX*~*Jvb_i!uS&a7RoXx#qG?;((1n@WFwMIsO@>2dFgwtTu=Oz` zs*zf_sH@wHP!Tni(Hjb4Yv3aJ9sJB^W%DSA-4rR>$bGcO-l%jY~!8b(B+?ThoVAN@P52#qEBeUxthcXf9CgJCFIecr#Q*W4_@0VlYBt9q&1O}9S=cZ!=0sSq5MBO?);Ay2CMZ5fQ=#mKF2@Ll~Qr6So2wYI|SO$8XJ$q z*C-NS39YfH^OxnQd3ErOX>->UwYf72X;DfZ9gDW>+i+5q|Ib=xxdaok93PbAwkb>iM)7ms=+4nSk{3Yae3HDmS)02YT00 zfqm|=a6kU)aLqG)3i8kD!x!E)Twn`tvHxwwC30J^rqa$84*t;+qXvw&u*k`7kd zo-YF*sYG`Cy@e?9A7BO^=sJU26iE&_4Bn)orYWuXjiWNe56C=)S-HB)R)eRc32m#Y zuHqRdcHfI8n}4v#K69NER0Zti<&VlMwuiSLzFijOz=}ElP;>KP8kylbB9KCT*MSln zeP9O_lkYlw01NxbVSPg}M{&VyMCU#WBwy?ch(XAxz*>TbV-*p5i6<5=T za2-TaWD9tNbn0LLS_zp%?XzHyN2o^6mSQH%Xj*WK?Oul-pnO1jt*xO&wjLj_ReZn_ zPzt0R==2TsO;uq!T4Zaw@3mAqmxZDLzb_;&(4i>1svf_uRs6nOfo)<{*~Cbh7BCH=8Bp#|u4^hshx`|aS>pSP<1+z}(Gle5;L%T4vr5*^UrKpmWQE`S)>dze$RCJrFA zw1bMut7{@fqIC~*iBUNw35P0=rYp_VFP{pBuVK3Y9wU9ChFxt=R!h( zxo2~=d>^~{_UZlC@7+JU`sve$H+5xgn8}({2h;TN>lPz?S-iM>p#UR{&Vmf1h_5Zg z(=CABh)d$Ua3)=IAl| z{$-iX;hsD2IDoxYdl{lTUGYSSq`;`vP5EhaH1C{kZPBqEFw@OMvR>n6=$o*vyeR%M z)6MwEN`7b4;hq>0Q&}%F-AsV6%rN%zIc_FmbCtsu2T=)+PtKXmbTa|sGDq`=ax+xq z1@LFW96dl-l6`qTXJ|pKN1$mHfo8Tx$D@|ba!mT%p%j>M_@(n0FlzK7_=TFJ*#z6G zN^MQcs_vvz4Je0&4!_u?l@DlMl)3)u8+HkQr${U_mX?jd$V`%f!BFe@tP!ESj&ck} zzC;0)r=E#%7z}lC)O$hX81)D<0>(Cw^xkUN5!iH*TJtUpg2%Lro(HuH=s;xuur$^$YS<8L>!Gqgx++AMU|C@S~6j$Y{vgL()%_T3#wN=S* zz~b{q;`%J@Pv|bQ4A~ues;dT#t1~@LByToYO=USy2!@V{s8?0yER>Jd(DA4UC|q4Z zdH*-p;{X$9X-B8gsRM%={(jc|H5D5J+(zn3&zv}(#cdMYnUOe-ZLKbW zZH3$9I`IV;viqD10yk9GBD=0SD1+XKDn0{)+>D$ip79|WnL~Kaa43R0UpjeRwbh(Z zZLu4?0;%$OGJ(&a7UxUvim&EKZ_%5Kf54!f%Bo2m{>1Mu#P>xr%G-Azp58w@y>k$jDvga-YOHj6Heb0?eo`O0Xx{_e}gkic^%ELC7ZZA4IJuYIwY)cOCn)qeS=o za`AKvmP46H;{LmQp3}GE>McZhm$572k5g^CbAox;pqHuV>L{-$4?pE_)zd-)u{?Ue!y=W}j-Q8evzaorcxtwZMnyGc6~Ycx5!hwg2Fd_{V9TKqq?+7li!m9~ z*r6&WwAVGPmT#EXF^n6kBI3;nRgvMQCc+CWf0IR;iLpZ!0xIVvJ7B1a z6kApfRRN&P@}cLvP!+MKFtu}mqN1OQ5{I)5Nv^O%Re&7x)O?6tlr?Y%7+0eCy<=#8 z^7%7JXHqo383%#<@>%s+k$S-LqXT^b`%jq8U-DfuAdwoqh~<}m#O;cc!8CDF-NvrV zlCymYU1z9%eZjkuSDZpB@%`)?%XaoqBZ}|GT4RkEXRX!g2R>&i%ZnJWanW)Z?6|MQ z`7_U%%5M2kfj;N2pb63y`<$t2#^wyBH^jdl5dX4j23@~K4@GWE&29(C%4TM7n`0|| zABJU^rt6n#m8L{AWtgLelGS&th2$NE_nfKR+wc9<>CHPY-?oUDdCs)WKW7RyuC$c1 zM|W-vJ^G3dz3Z0*m-TqsQtar7$1Bo#l>(Sc^kX`HsfJ@p->!1CmvyRo9wh^s$4q6P ztew?^j$iF2zhL%L;)gX;Qh8*BxW}LIz%^kcXiGjUn;<#-*>!-5zvYqqz6RQ*6USgx z@>?Fs?{E`(brN!7y1HD}xyIt_BPwmJBrgT5=bfG>v@6qP32mH+bePx@*AhAbnddEA z$hUd_78O(nO@u-iYMO)^hySXjMm{61Er&EUGGfKxgU9a^@7X<_vV7{rCof-Is#if4 z*N)z~LWR;7;%66$7cTHGM@03|iCO~Afzf~ONW4;I6*V$(7?VhJmgcG727}{T3lPHR z$hGtp@tX@l4m-e*)3`bIHRVO7ka;b+1HgP*?tDdj;fN060$Ha>IY(csI)+`%O5)%` zZtarr_-`!4z}dMIb}UD;)l`BCZ_16BktTv79X8UOziW{uCaM={Qn^zdM?gZy1uxP> zMB_)A$Y4{G;6$3(uzUR5m~f>H6T!AI+6On%_^@~0$-{%x0-ey|g3&oM(gZ*^!yle= zB2C0-s#(d`MWhL^ZH{JfYpT7y9NdZ0h3S>+5`y{o4$|@H!o8vkvtu_+B~BDT`2{z> zi&VA!ahOf)9CZ)mw?xtGAYYbw6rPZnL?nd??b>%vB)sYi` z)Dkh>^QvWUP5@G4%95KOAg=2vKHZu+b=h^(x=?gH!*}lW&D^9;y+_+C0uz4Ipsx~# zSLo}yD1iYq=N{Z}pfmp>#3Nldxeu0n*nqBU6p?%0!d0DUmg{1;Dgn8o(n%!2qOF6g zzGp;QgSEE!arh7{6Dm(|tYGi0?8Ys=6%Y6ErDb1r^8^>VQ0%FlW%`9*7}gXouS z+ECLUOOcc%nd`#E5|Z*yFA^-?^tW@UY7Ozi69S!>{OhUm~tCq>zML^i>^|M2SR-M@eD*6IDLcT{i88{^fqFdl@O{X#5T zts!iXQA>04c*c7Hy|h~V>m$(@N;?HnOSP3KK%@lK2a&85*W5G__*9e!5g#3`VJMtVW5PZ# z5QmcmkXh$IqK4foJHbI&6qmIG5=BXZ`jBL0Fznve>j<$mPqc2=&S7R(*(n5YQ<(}; z2&%A)>v2=Jikn)vxEdHWV&_*GKISXA#T;~$r#h6ZNry-ct#_+CLGB^atsEUdq(w8d zmg4n=*lsjVx^41GwYKDFy;8RxSPHs}lI(4dm{#f#VQIaCkC?6vVoH6?SNn7Ln6Knl zPiVbuZxCYiG2iKqHMHI>+E<58>wTbUy<5{(XGf=1qpv`HtX32q(4nrb6XyG=+BbOe zCILTYXAmBUT%(KQr!Bm<7FU-aDuVY^&gevQT|o*@DHcq|8)M8r%A|ngtOk8ST|1CM zK`oMuKOZz-Q2rTJBCOBn4HHuDK6>}|!`nCS-2M>>smcG{dth>Y?`VKv^_Pxcur}C> zcDW(iAeaQ!##H>`LJ)R27|nzIB7DgC(+yI%DHHONhvs27lb6K5JfdJzK)xx^SEilS zT>C(!34Ikv-sUSlj_&=tg@`|+?m-k#gA;78OoKK~Rxo^QUy0Jq97(3lX2!lVTMT+r z(BjSP!!L6UC{js1W9LRL42m~%0Js#H;(6I-j@ZXR_iHMQVa6y*H*E^XOmDBsOm8zgj#UH7n4;6n>{`mqR?b8}i?UUWsiL)<;wb1^LV}r0 zYsxF5VW}|sC95vbI>0+!rU>id{!kE$C75l!&H-E1IlvLjifuv{3u{vtj7`v}N0$J> zY~x`j9e{CiTH3)S9e9kEsU=ksZ6Bjm&&5E3*+#=9>N~`J;}Oi7DN@pz$7soq#u$Q` zBr&=AUH~#=UdOQe@E;g7K*DGW4fp^t^#Rnf0aU7^%kxt(S9U!Wzk~Kk0|J1DRz{mO zT@@c%exM+0N&pGi^usuRL7l~1_U~$xi{hur{pFKV!OcEEDdryBQkg#{vM--2PA-zq zr2KLhA5uQnNl5}z@pmh9+T$LWAUz)I`5H{I$1hvUy(WfbpS9fRM9`f7*A^m97R5x% z<^6T0T{ZXrsiQR-@dK=&*i6sMH-5x! z>ZqW~A;vis!P)o$c2KTS)ST;jg>#D`41m%Q=16pUhHQ_sOMC~9(x66Pq3qMN6Q!xc z2?D4Hp&;cr!r)O621P|E;bwr-IqHB-xte;&(cuV#>7g}%9$M%)CNUlWt$>~j2l#?& zceGzmk5-^YFJDiMC1CTn<}DzWAny?wSOPWWvnoUH*Xc|Q$(z`#%-pZ*pFAv~*e<+c zwLunO=~0Q>!#>W@H$Ij?D^|Ot_SYUus9db39>RfGLSwK5YO$J{8c6ip0822%YVxrg zljfsJQ<&bFD!LsYwE%ts{dM~hao1o8a_tnBfZh#H93aFKsE_n~0cPK>i_|fOC-|Q= zm_mz77@WjGUjD4jd2)`f9IseT zi!tC2qt=)-3upNGg?O1}mGDi_usX!#8?RWSds>XkZGq9j@iF|M;Pzei)j=V4x8>l+ zb5KJi5o_i$;%AOT{GsRoYk-nY!3Ku1R#mLWdg|j!%8lX&g+65=WTVF|V{=Rh2n?-O z*9Uge6D%JP$sMcufLfpf%Ln+VLj)+OVy1N&3p^hX;11>1dd~0x5$g#tcn1K~aSN;f z$?^dK_Q23;Et#>cZ~cr*+AA>_K5J4hkHIh~1_Rfn+R37R+(`mx4D`G^fF)!#f==)} z8pBXZW1wbt7Jl%mLIy))7ywHjTUn7nIFqkn~;#{y^2e z#UB8dAw;ei_=Cq~778``$wl$q^@P;j5DsWDtLjT)ee-eEMJejGb7|jg=fA|;O_R*t1IMIAc+GD1 zU*B7beIv{B0|oL zZjlkE7Q>Eiv9Yt=f3%yoH@+2}d#nx6jOgaWzE$Tr4MM6n^Fd&o z3&VO5xFbX=^wf2I(bywuC%~(V5dJW0K9|pKwv?5lCfZ;~hTYx+Z(`#t-7;dBLAmgg# zie_kRM!l*CY{t;s#{5f(z-C?+`gu+X#=>4hvO|2q8cC(?VA~!%AAm*ThI>_!JCQj|166k{SM~1Y)1nxJrT@Z0~9;x#t}V z-h+FuiYN9IUZ0gSa$Dky?us$Fq8l_{dQpu3Sy3A)Oh2m)mH&UF0QiA@T$0U;K5Ak7 z=f$(jr%Pdc@Qq45!~YboOG|AeFaZxv^yUWf#mUDgv`?zQ2e72j<$Z+T3e# z650O32dFB_mEZFItq!(^5%0wJ7Xc*2v*dH8zld1X0ro^Hyymn#!Mp_9Uj!%=Q_Ba@ z4Ws+(z1aZo3M%{UQJ!o_9E+Lngas)^<@n9^R;Uc)Uer^B3jep8+vZ z^{>&T@dWMg9-v6cdo5i*DaQkt&2olbCZ*6Mwi6-%rZoK6A=uHnJrQ zzFCde_)81%dNr{<-bY1Y!=oad)A;#|L})9}amZNJoR5Rq$*S^F*pPKxaWObt#;;MK zw_p)G=&dH87JTVA$gZo9Kg5@*AY3pb9tal(4T zJ`(KM6%*TwU8&5qnj0lC83j9bMFgA`yCUOF?SvV-Vk7UliWPOu2w84(v16AH1LwE% zA){NQh@BAK__2!#2nW7`8M`7@Wf!!bk51(oq@|vs!tB_^M1=!YA4EY}RTwaJJV6;iJKbQ_Y<14hMOnh*hAI36Su@93}vQlL((?C5|VMgC~C~1c`<;Sp}1MeShef${I zzvtJ;g2AhIfg&rZup>2VT=07;glyFH~6T*+idSDWo@ht-27A-Hpo4-9u>Mf=X)* z?7o$EBSuhWZf^Dx&c3Xcb);bTyk&V}yUqzbu`TwFU#wIUn|&r$+5D?yBpPtcLYT@x z7Oz&d3;@u8QAnp3i7y@zS#c9-r=TZG zWvfb-bX-lh<-EesBRa;T;-tqqNyhy>Fy$%E$=|k?4Vj+7 z2be0EY6K~o;Ra$XL#Ag4ASI6Np0hkd#FDVxS!;i!Vkx$3xSk2&g^LeL1L2k7(- zacm1K5~b=)_79RPbR4s!GL(JRxNBzix?>fZC0%>$I5j$JNvAjwDhqs*LDGg85D``T zR%@WTo2fMgAt@|pw-MH`xO-fy;%HCgb z@V;ZX#Z>jFf6z@_2io@?x4K@^l&B6G4pQGC%I0wj$0-uDHSara3*OY+p(AUJf=@Yx z)b03|fjOiXQ7}iWs~vLc1dPHpb?V?z#H~cUP1+xOG%)>TgKmt}r?t~VB_IuK=ZB{Q z*Njof)1HGhTs=47bO5$U%i)SNqP`!S3|M&9A7schp^1|^G z0n1R3dq_n-5DHn;Ly9zFz{;R?NTiLu=VesY`7CAzxrbClyd0=jRmuQZlcqMmtrWe7 zR7CzM%totnj|@cYAr<+}a6i=bSMeB;#oxo&4RAMd52={ElWJeK%JbJ;F6FFg$8(}v z1VWkmsz|o?*rHpGlsG54#YFd_8wlr%6y>4j)<}^L=R~)NkTauOWW=e(5IZD>doecl zKDy2Ce;E1NF(b^08PUxLh7S{cBFJu48=Vu~0;rhlYN;I(BWoUsZnHv9RprL>Cwhm( zn7oyu+xbCaS3!rwagDO}=wPyGohJu`Ohh$2v2|XkWyQ{OvqqV-nz4Ju>yRkx zH5pKI(TZ3l0*+ajKbNx+*Sb;MH|&vN^$?ZAn~P>>Z3Y&ykL%9ZOkZ${25iL9N1jV9 zZ%60bRw`*XY{c!Pqj~}=(?a;ECziA7M|D*@|93YRt`7oNbn98#$8fmKuPmlV zddIW9jX}y)vO}Vm_oMeJYg&T_P>7Fwe-S`ZJWHl_NStn@h*gE*&fRnHYh;5DC(c>u<{@`T48AtyH#4<- z0Dp0g5`k=mDjXYU=99=*I4{QL5FN6o^pp)S; zZafm78odN`l4%7tYrE*e64ty*##wbX1z32eCU5M_gG#_c*ll4Pq>$iy+O7JBvEl!=zz#<_7V=9sL zaF$swXAu8+A^u*jKy%-!>FTZ1yARmx#*f>w&6mXW<@2Pn&GdIOv_(ktM%gApmR*0U zs`wrUq_WUq90<6KDW-||rx(S?Y6_}(AP2Rb)d7~~xbm7XNFa6}i_aa2Yy3q?!*e8b zdHRY;WyJ(IowVKvJyQJ1k+{y+!4$BYq0Uv+{gH6_q?MU;k+Swu*>;urJwHSB6BKWT z>JKGh{(4|u{EGlEQ#PA9!&lv1Ib!2uihnWjz4(_(? zF93+Hyjb!3tgNBWi+@3F{S4LD4eSJ3qz$+Z%|w}m0Ti9gGk(%^YTS5K!i}O5&OL0{ zS?70}lM5vT(!pxOklc8Dz>VSq7Ch$zF|E9P8n)ZuRo`z+@4Zb??>!5@qU1h0L((m@ z$L`xGb{~JCr`UB~NuXn+1OPbzHo)HUpR;OYG2cQI`;!L6Hq5s@rmY^mawvAu46W68 zh54Iab0jxK4Jz#`(o(k&&mKx2RNg_uL!}aZ!;u{P68-I3gWBplc^_fjiSRg*n<=tA zm^%Fd+yam+We;zIHYTY9ad?6aIgi**p7rwb7?$Lq`PzUb zDf`Fj$ylA{)o)pn8-1GzSds&P;qx<-*P?rtq%;S`l4PsHh*88z{vA;HFB>x$l)PbN z*5)=~6P}(xUHhvRN_|FLTMk?(m5WfQl-UvnWfg=Gn1B8;R|5m(Y5IVP1T<6iT$$*V z;EV4%J@Cah_XN_XJ~rV}U8wS#O(N#sU1ZvTP#Bg^8F+-lX!V>L_zK2`^b}@FD@E#kFffnQvliMJ()bd686H#w>e|7YR@-wzd!AD2V*Z1~3xoDOLb#37KN; z((lG&B&yN#MV|>Ffr_g5TIkEY%0issHiJoHHTKa2Pjk$%{3g$ z#w&JjRI$55r>K*&=AkQFwQZWmXBv?6V|oP~%Z-moff8190I8)t9MwCHC4HL(+)LXr za?tgohGV%gaFqNG?T{s5YXGY>0TLylUp;ecUFg7;W7%jle#)_=>im&`SfnA43QY9{1x=lP`5a(QTl=J>%2jC>^?D6$N9R)+An-eGyno__3RBpWc7{-u*L4D}H!W z^IF3c*Q7t6u)yFe;-%&F5@0a;rAk1GY_IkV5iyERqaK)2T|xPzN&M-NcxjIGQ&44= zD)aAomRX+l)W_9TTaKoLyM7@)PiGfKZEsq7HcR`opjTa$Vy4H5q{9dDICNYERi$cq zHD{r+z)NLgrxd9cfG;yWP5{8nQspd<6S2J7yMq(LBaX&8!nqzN0Ayxq-%hK(+OQL| zW#2R?`z8Qjl6mOo4Xdc}=r?Nk6_q`n_U}$hG@$^4hMqPe=!{38X%vB`gi_R?zR=2M z>t6@2{fboj`={nifzq+;cC@_tpz|2`&Hbf(@w97lKZK=y(AsQAf zB!F&g8Ej){A$tT8(QE z{?P3ndSSqysjvJf>?3~iNW3_21tw&fc`INe-vO)-JuA$0K9PjlKsmLSvKE#z*ZKHh z&UA0fCjPoA`MuZPv%Gluol9R|-spU`^9kUbS*CwJ&-p~`@3Hz9&LLp7^9dlGc`I-@ z=R;)Mw190B(lY3vj_tADdUP8#{HlVtcX~oys>#P&z`0?<84BXDoSW9;+^EruP#tR4 zXEOHDs}XAr=cX;?+_aTAH*Rq~(v2E^-r{Ol$E3aJ*~MetNUA;q)@e)4s50}W^_VxU z&%B{d&KikV64n~#O-qG`>|3qd)`NA%0xR3`SH?<(U>_BGKFl~N1Zy>{qrQV$h+SQu zdDBf{ozc{QjRkC{T}@P*VBU1ZdD#~568ToJ&eP0uFu-+Ms@=#|uomZ#?Fmh$pKSmZ zPX-OJ)3F+n#>Zr!H{pT9!Ak0VCMU5g=9(3;4YV_Y4wJP{+Z-S#q6lApSKBp9egp1E zFM$z`YO92G*89VS`2P3?yM6cJ>HV|QTeojMJe}7Hjo!U2_f4z~+LIS9z*GirgZ4{F zPEu|PAsfkmrFgr<-N0RbyoS%45ovkUu{U4eTm-}&W=`0`N-L(R!PC4eQOZ*G{>n*_Pv3paR8 zBZF`Fvk|n{ znGAA*9lK&;d+p1rd6m@LtiBUT@FoO1c0~l76}uwiP3?pkyJ92L+Law5cAeLZT|WMN z$k-L35x&c@E5M$44mJ5cM#i5I1~Cd_S@)mN3{>A`E`c z5C&V1FsO!4At)JRnig{AO z?0I0VAq}?q2REIAc8d&{gP}s`b)9aJJ?3Dmn1j3%paz$-da=6%KG7;)?BpQKMeAX0 z3a|%56$UK9q)mO&*p$Z}Y<>1%(G0D{ctyaip%Au!;geOfd!kk97PcQ70fLzyMJjEN zI#+(8b<~BdrrxtNsIz*a)m~8K=jXK_>jUW7JkdH9t%DSLk3WOAheAsU+KeSFMdHwU zJpJyq9evf7A}cD*B1B@NLSb5ujk>@Vup@ijv#SMk%e= zx|B0h=sZgmT7)j&{t7r#*+Io2)vgUy;sKIma5JDh#gPX|w*$_P>93X~rxM5VRa)x= zbv3}e^Nq#g!If9FG%p3hXU!;ri~C}`l30u=iB( zbmlZQH4w&YmSy>@czQV^EX#H z(4!ErXAfhqCPe_JPYHM4c3I#uhUkg-;!y$E3#6T*>MWJ5Dj9g_{~erN&!D1+1FCT4 zjYrlVo&f1A?Ybj$HETqeU{|=ICxS8^IP{#qd!Z*LtcRIbTB>=7IWzP`gqsn1BBM?1 zf)#pVvGe#Tv@M7zB4ZYcN^k32C3~(LNFcdwoB>}g4O7(B`u&)U-4+v z5{h8ZsHxhwT0@_sq2t@8fnldvxZkkf7V0$`R~-)!Kg`5!2ZhnAj_)Kst4Af8hNDsU zKB^6{adaXU(5_q^KTfLSdjn`wtK*d?1K{j7z_Lws{D_v8gKXAR$9MW3Z`#0cOVs&W z2B?+WS^>42va|-Ur#k?{qpS3b5V&;h?LJiWWrNRdYo#EEiAjOb#_f*vDk`*XCVi0`_j6ae~z>kjUD*QSM5j~MFqr{lkXyH9{#My zpjYCl8~br??5LFwmY2n)YYIR;I)2I8nsv58knT#8^0>@<0el)_Uf!QR65CRFr+90z!ctX(X%&Z?G<5I)XRTsBQE!S} zb!!v)(n2J<g{o|qQNX8h%#R3II3M;3F67=tTZC17_X(Pn`aB_^R>B{^Vq@?5n=$TJl&xjP&%4w z*utG|RjG9+M+azX(G0Drc%3P_!I;XJKRa8?1Z?3g#8vwf(4K;{0~K4iKdIECLO)h4 z$6H>!e(ToCu!XmKaCP?`D??sf|FPm>g0RXxw|^o=4x1m#epdO=Gj(CqqrH9TY~hER zE!>;dK6doghv}(fjwd|bUYaM|pXh1@07XlS(o-l=i{T0P`X0^kgylD^_VObN>4qOX z`t*(UR(c%;KNl*&Odfr8d#Vk|F0}Zpe9_Ew@AxjV{tfm#gn6?AMvtGdfZax%EI(WZ z*dl~zyT<&@Y3;hYr$L5EgPx@gIQpWR>dzb1ZBH};XH7#!t8>4Y)ZnM*Y!~R zt0NM61tgrpfrScM&9)Nk=H*hjK>$uzVLhTKq4?Q_NIV+v;TKSg69lkOi!OjwkO{e1 zyHvifBWbqT*R$tZt21fwzV_i609R)GQv{}|poF|rDBsrsAX8?M=Y{(^VkPh5zD|U_ z@_ij(FojyIxUW&(+9eCU^nmM?Xk&340?z>?BV^hjb8?x6cw8?v`ikQkabvJqxq%pZ zl&VF?TnuBa_o!NG^diiK#s<1%Xi5EvS8>~$?v6c3R8S{v)d09xo|B@l?CxO~yO#u`jvwUdw`BwJdO4HR|Lvjj=z` zHOhKl$p>Wj8;sh1(s9)WA5)z1GWiLM1UaLQod6!qAy z+1jWB@VRPZfTZZhcIB7QVeH53W4pZuj;CBTl=F7Ujq@1;7D$*)fdwC=rbK|H*3(Fh zm&Ct;iL~qZ_?^UW8jN5}Wb){y#q!W=HYqz1A6|a2Bq>uFmjee}ZFE@}wa`GO6*DHp z>|U1S(oQRAIHX3Nm34N}LIYucqcg7QBO&unX#u5_&~SC|^)E)F3ND7*NAlXgXVPJ?|X==?e6qhON1t z-)#vsI)ECele2~)?NAXM^!^(4e{4_iLzKzY!K1OG4>eFR4+}NaEsN_ff}e~ zaSAAPT!i5yfEsjJoct2eAWT`@77`QSCy3+R(@=w)(++CDFO?@Q5Wo#vK(Rrc;LpL- z+_km(017ww-_?-AfM3}^XY;*QY?ddb`5pt#P$xPEJGg?s@(^pG2mCSEN=NW$ih=G~ zUs6~4S1zfAAaIUJ6-imAtY27&SBSO=Up!6Y5U*Ohd`4Vb4kdO8KTrvX_|4p%v)g*` z>mCvj!53-?IZY+zJl80`a3m6MVh2zJly!*8E3`O)yja|e;6(m)b*W=)43aKFxT2R-H{0C-Uw(HyTwgxA*9{)j&zG7%}+Oi(?Y$h$=RwhAg zJraZ(y#$?*hkTPC;-1SSp=+<^WNSze+fow57LXvq>uP$riCs~T1ffRftf+=3Xe#+3 zVQB+0y<@-=WaV4i0ZfQ8M8b!3OxGUS^&KEvTaO7LCx_XyH5*HhS$icWTf>Cd(%QLP za`1#yCg?@quc@i8fhX3L3QG(VV(k}}3?~|%P~Rcy$b%==Q+Q%+YD(5<^DBVyS$HDO zgc#PudA(`i3AuKP39&9ZWkP&{%HjAT%;8;|7o37IIrJNVVtxF)P4+s`EZ4=!9s`94 z4+XAPL&)=;MJVK}+omLh5orxXvBpuDI>*}t3j#GFUshKk!jRQ~6mqAdE(c?R)KNs@ zI41sNA$}mXHP%;eK74q3_u=h(cdtG;z4L=t-+FX<|Lv=vIlX`P^v=~3Wnx-ZZ*rCI zvDSQDJimN4Y0Z?crdssK_B=t09*K1%`vr;run?cDD7x2hp{l^)DT3Zh{hwSE-(62g z-GJeM7PG3pBsQKOS6!5%4h@&~?SAAN(b2;=#q2%c%*zyzI1bzj4<_9=0M^7576T%92)O(sJQ2W-9)45q~|FVscZ~4_oDj zkqs-J)dn{sx08sOUDYUHgpa77=o0bc>8QwHS7Ei;cZ+H&oG) z$GW!XU73Lt#Ej_Xd~a{lvn6$R5LOs~B$2vd?V_OK}H){F+*+Dv{k}1C2dL$s@aM z3(0O0c*;)QN~sF6cBM;e-A+W$z%bsnfR~gV)w2J#fc6%qgmhrpEwjgaQ^T{Dc?vJ* zr8)x`v&2}+DR6Q{Oz6okjaA|RcyOT-%NrUg`jewjs?i+}j%hPqG3M5g;UsFERjO5p zYiKmAeaN}s`O{sI?%s^+K7ZP;b!nllz#G;EcUI4z7Dx%7KOHs!_^h8lm7lYU$&aoq z8X^c=QSga zzGwuEjn<%YFuRKCrIB4@Be`%pgr2DkBLv#CO5-5LU#JJW9GdMK<3ukB(Dg8Y{U$ss zIq>}WWea~^6T`CS!Jk~LK|#DuSXZaXJ?~ijzCmTKcs@*l_<0-iM0e$!T+uC@54>na zITILt-e*M?LZREyPU#7AhP`2ujghFYyym}AiG=u{7NW#$03kg%HNY=tu!GcEOu^QA zdJ|tdD#G)C#8WmRQ*rT5$@vuzd99(MxAHIOqF|0FvTa@;TFAM68OjksptXIp_O}0EJ>| z`2hZcNMIaL)*rpPQl?2VkCthd$NE#FubwWzo?yV+nh#X&{YJ zCV32qiOO0tnB{;rlBc6fhQ>H(7TmsFwi+FvOU1_FRap+5uCg3J){d&vHgu91yGLUL zXeK&<)Y2YK0#%lDvS-lfNmv@AUu8LL44lNQvTOn*$y8acORFpgjigT+It^4(@dE>r zNUx}9jL}Equv4k>fd(Q0{ngQR`V|NmxE4WI3*E56Mw;UnE&efxe%Y4dAIhg!3H}a} z5%MJHKoM6>$_u|162Yr3s~v|^fC##(p2V)klqa!8LNc}c7_qUMIuf{v_``+xev-|8 z=9JBBWD5~|vl_4QmlopnYGQl5kCd~nsD$ve5pW&F=PweWtw6^i2%$M22eXq^<)t8i zfd=J1uDC1#+fjUt3cUr3;6ZOS(X!x6$3dbl%_~Ao^lC`*8Wvxsf^fl%cpzLb3^Az= zelN8D!(_<8hZEsDkKlimio^wOZDz=L}dWN_P;dYr^|iX?ALLHq`w_xW{@Lhq=0Bj^?qA|-TAqU7G`@gOHrjZQ?EfkWjo5;vL~ z4H${Cvx1&hiN*~hanv`5fRVTi0O1?EvNQC8Xft7Ie^-rOe{h_Bh3q-Vv`%aMMQq*e-ip%VY_D+@upk(+$pgFvxc zn^-bfO$Pxwiqjs&7mtXnxQVn=(37RIRV7ObT_a!}R#R}??{1w^P%mc_e3VW{2aU}0 zkfTSZ8x@_d1T>^RPnvN?sapoqaK z%sdKbZtpGPIHJ|vbacpY3P-QFTa6CTp<>g}m7IEy^#&N)q8`m|AJ7`H^)aE+Mu1s8R4Kr;xfG-!d?VTssAGM0@WCoI-o1 z2aLio!~U|i{EIOEbshCSRCL}o=m!4g;0rcG>_xlWlxK*Eq#E^b^0epjdsk1&3Alz! zxWPf3N{3h$lhBja6_j$MwO9tesFkE;oxaFDq|R^Nd34I{kcc!Y`WoDtn35X#h78E~%RUdmAjA(t0rHO`4{G10x~mSW0R zRYZ6L97H`aBf3R|oEhCBBTg-b*da08i?Oly(XFb^gOEjVa5fE^5#2Jta6FvAz3}cw zL0KkE?f)BFURM9#$UyAq7O^lCGOlWO@e8HAe<%0%36S3rOnjxK387PGdtQ`8R5F%vr5-iD`VP71VA$&vqktwrz;$-J}r_!9m(ke5cZght57PixSFwNbfGAFw7Tc%;%fAXwz#l;qRayr zz!S*uuulRrLMz4tY$*wyeB7Fm=E@n>sgiMrIZ>C^8i}b9)thc>jRXKisHCokV(%eM zO+ly0Jv`A|ecK!v=73D5hhn3vF%93b*C10Pzhn118k8m24zQB#6klrAItOkVbQt?F z@tVa=G}t~RJW@%f4-IT0_Yu9W6YQNG<c?a^vF-Ny}DvQ5Ann7M`7 z0;uz2H)V8l^SUT@u5OhVkPY*(S=T$-reF}R)_jRe3s{}z_T{1m8wP6;Y4)e8itq76 zNausfnhC>&qK$}udQp6=rl6WDa!}h@9bjo}06MO$cLsGteC|kG<1dN_*mTB=Cg7{}~%{5>fmf_SM4_W;8mRPset^PDRr#`Tnd)^@MJ&c2MxZ@h-4AS9iOQc-; zltDFq5q!NZ5yPA1G#QmTi=B?!GTJ4Zc}<#6qyNm}dCrToTvHr9HIn&>$$ z<<>{Ii&&_JOWA<3e!9;IIF=Hd)~vosS^w6++^VdqvHUuQ6fil$zuXt_TLhLtMA{fh;N` zES*xlBChrTq*E4I{*Rk`0eMNJ?giKiv4RxY`0s$G0ljMq*f6jTb4dWMpNDcn za^S+L^-gIQ6q0PaB=oKe1pioMZw&BW302SSsT+m-_V&bgO(OW>+v|g__VF-I|H_Jp z-XlmL{kw}y8xZKy^hyqI)bnfLE0qYAf3Of`o&-!Gjo9XHtfp^HS5-^^)du7(@s(1j z5J)}6XxZw^*Y>5>KD4mVm_eh+s@je3Hf+JgA1(6FfG0(r&@I{8wlSGiw4RzCdcCgT zNUj%&U=R`PHPbw@;=<#!%Nbt8$Gw8B>Sg1owqut-RTVJxD|5U^fOK)~`kd`WA{Lg_ zZ}KXqs-mMVW{wvLz%91658_3rM9ekdEzvJP(T6BpG7^2MSlU%0zL z!%fKKVgSqn{KF7zpo3YKi@Emshik<@Ea75OyTnm}|}CGq9Dan|H8P zv7MO0SZy5=P|*JcN?1`z9UXJe4Y=U43fkn1jZt)OU#fz@uZ10YoKV z{n-&o+Y{n=mX4{O50ZM`<1PTtq3Z11z%Fw6NL8%i?tz3%`T7I4#x}f&UxV<2Yg?Et zU=2QR(3UYw5dgoEQ_~uJDgnP>0tXdroAR5`h>*~ zUlA`Yub1G5(Jxg3S7bk%XNV|LbQ&JO66yL^UF>O!b37v-6MuRnUYaBQ6lj^H%KY=7 zWtJE2Y5=%YUzcb`ghmqac{&j>YI{>Y+$`;*Q*}P5w%+D?oJfj%AdiE>D|8G*^@p6Z zP(B!Qc#jj|6}^idCjez;sdARbiCEs_^f-wKj&9?*9wz{4W@-OmPK65AY*3a?0Lmn< z#s857Yc@VjM}1&L<(j8myi*i*h;@LCd}8vm#tAY(id`Ew3)rZ)ELDc4(`+?<5nw~j z`P_j@_9cTl|G7hvELH{vyAo=Fb zgGZmfu~x4$E6|3@T9|lA54fSe@$%ac&~P28I#wp6!5b+!mb!n$X1~vgXO`Fa*)Ipc zkta?E;JC`+IlRK>Jx+zm7Rd|oiDJ8cKR*&L&s*|I*P(gZT-CO} z1K3tMcoXho2QG=9JQ6R?TY(9sX5I?8b5MQD#hk!w=Mzb>O(~lfM=HJpU-nX5&OgTeRapg!g%I7p8({Uw*rTAK4h2&aB@Qa3LWNg4cygZjxd8& zoSX^{->C}qsxC9HP0Uf9^PyaY#2lFeVH3CF$&w4r74QqpHkO#Pp|)ajx~N3V0gLPt zb2f^YGi#B#K#xq389)y?6L%`C&Cc2Q?3|6q&LPjw8jD{~wz2FS* z-@5;qyNDu6CkA7VCnsvxWL3geA91PS0MCY{G ziVt1~Nl?kJm3Zrlz-Gm83-$m+we~Tz7H!9q-_?Z8lHbBT89#vmfVK6NI2wvST!`#94+9L!Eq*XrD*$X4@Su}8~-zK6NlNMe<^efQSsXKy@w`_0qk z(aA;ewS`EuqRb8EL5e}bEe#`VRCWbgf4;-a6_uGzi}vanulVv&3RTU?J|&4URhWA8 zAp7~SiFrM>B%*e%vURsusY)I3)rCl$lCrW+a6_i553-%II^oBz2u?HYY-Mo6YD92tKev8s9poJ0nZHzf=L=f|!9c4n%&rh>8|%LYnEIBNi@kckcLp5l;C z!|Lj@#wQi352T)J$k>y9i-RQ=;1q_jfL;V@JDN2_YVc=3-Cb=Db*-o0&xGIGFRA;603Ickkj4FT2lb4 zz;VgVTN@mj@-nPT!o8Pa$4vl6o0nnRu|#OO>CSc)@eO$FDBN`PZe~Abnwd44%Lv9_ zwWVN-%DhU%F^$YRU6~6wl(JKvk+ft)+sLfZ*OQ=;SqFHTc`JR}kTa!^R^&{UGOQ2< zJ?pNhek=~Zc1^7kubd>~oq_5%;b8@o)Y6|UNq&e60H|cw3F@kcd8ZzW(5GH?$*wu7 zr)nqsGhb|1Ow1G80`miR86J<==R!Vq{I|&-1$MFuz^iG(YJiQ`EaCE5@$_;;2$vCC z246gWpLoyiU76)mFFtws;_`{*(dEUpqjx+w5(DEG;^!Bsmp|}N=b-p%B8E|e4#a=; zNW5BSB{h_BAiGF4*XAvOI~`YBgb>6)u5z!4uPg*T3;}!gkoRhn4}gkkO{BM7&>mnh z;9h+3r~m>6(oV55V}^jkI5YG_ zu%}e$0Wb#|7-~s=o{Ul4&=V8ZL((gmUh1(B|Ee#M$_ZxZi3m3%^h8FR+6620#KxT< z=+zeC@r0f%a9&NGn$d_GdIG4N;~dXfp(kQG)z2#_w7f8waYIjlj6;>&dQ&jn7nW6jPZW(w((n%L&8*WaBWGpo8JIz`pUh#*N|{7j#4%0eX!} zV7RoMUtr(Zhld-qXC63JI0zhN4bvfTx3oTMxKXTO&eBftZ&Wd$u3iF`9O#?61Xgjf z^`5;u04JzLiUM{VTAC`pmv{T7%EZ9bZYJI>#`uwqH&@$E<1k zv)-V7$(r&gw#yDN<@DE59BIihyKu3DBmHj|DGG7=<2gvSihyBYqrVy^lsOfkt`ejHIHCaQX=(|n zT8QKxijBFCDtVt!b8DPOs%jWf#fMhC>^U#bVZukGPij2caiU59u9mvM^Wvxyu^9An zuF9T`_uW^FK$S32f@mIVsU7!>o51`R{n>le4 z7JP?%!Tmf)!XZOI#-As2UkH;dLjcniBL=7)0tAxev-BdczeeOhd* zXtAZMtWnjh7#2DbkX0;sZsjdY$D-A6>O*Ow^uDE*Roohy=>7_#mE!|^wP=#oWc;2M zZEHc+u4E%QR&nDatwp0)xD|A_2oF&KdpLzx+c%QCNq39Jz^$sgg*{wkEy`yVw*lsA zT3mE}iRz@eob7IuTC__LomG5pSjDQT$Ib`2@mC+FXOlZVae;O$pIDbT06wv_Dl>-? z)mT2Us+v%eKt3`0F!Cg)5)TY0Egg^or3=+yrw^X%%A1R1Avzq>_5J)V@)8(C8huS# zi;A^=c*cT&8*#Gya2W^~&*X(nFr9F&u}SF}#=_BU{5u45!6dJ6COq>*9b89~Qc*1y zP4S5b7Ak>a3yILaKrm0p_Y3&J1|&dk`z{?{wu7M2O;j9l$hY7I|K{ zuOn7+*nORdTNn-GQu)3PkeNa)qSL9k=^-G462jG+FXR8{H;bfLyCTvUy|VDq<9!na~ZS!u@rUO;Ix_DN?x9cX?Ks*H}gSWcTFDnga1 z-Us%rn2|a^OM5i+&HGK(oftEbbR3#-I8IyZBQ4Y(qy}W^D?jkt>oQ{f!0*^jHemJn-97yZl;Co1zk`MVxZlGVK)2iikS~K9Fdcf)74O&5&UTfT&9m z{|2PigXF{N2%IFOwuV}z);qziuHtB9~5X-(l=vs{OTgZ)QTGK z^Z{LlR$Wl3Bs}KhRe59iDn`;o1140NGaWNHfDiCN1w5fgKd%gX$||g+o~i#3EFTbn z3+2{&&hP;d>#1@&sqHq-1CcuiItjtXH|=0b-u7wEu) zTUejDpvGq`EE{}~B|!}Ep({2a^ah^;KA0PP4Hu|qejNBvY#@F?UDsLgf%+zg10OmM zeCVmEtpy({7u5BWg1XLt52~P!=G8p#p`U^ey(y^c8{&8le1PNyK#HI${Y?WtNMlgo z1Agy3afluO0oV(ePQ=f_WZkvM`Tz<*_#ZYPM1TC8&GK5YS)P<;c~wF^03u|b=YR-T zDp($L4H(g{>qK8hsobS77fMO$iU!LiHK2st<(1AE<@psV=`SqAD@5CbudIeuD<&&> z)!OVc;@Wa3v0321#>q5uc%Fw=zO9?S4xxsU6xS^EMDPPt8|tq(G?RVg+W2R`D}?|3-OCnddrrOITuZ*4zTEb zz=DdMSiquENs4)JwV#xv^ail-N>b2$>Plx84x*t>w@y}9SHh6h!WLcS^{AwtbMgNy#P!|Yc=O)9JE!;Oy@|T9M*Dv8 z9m{WbMw=cy)!0Wi^(kxY^S_zP@AS7qu*HAnlr#M>)YIGc{EH*8_e&*H*8LRbFPgsm zDEanYT`ul%6js!7=hD9A&i}g*-l#^b^x01CyW=uWd#E>#Xr@2%|9U(XFx98L`0f2Jqm}j9`aR^Q`l7&y(ml$NH zd>xW~YYyNvV}f{XYB-fH;GLrI4TX?s>nM{ACIfz z-6HixReUeOX$FW}#!H;%oMvP#!xO86vltFFRZ5iiKzXMbz+^df%yF9M`RM6TqrMVl z=Z?X^)Xh80_!VV`6jGeXd5t^3q7q5R_G#&Tl1^Vr(g8VfUJTQ8qt_i0bg5NunL0hF zC&>@zv{H$#;}+GY>-37QlUY;^b!BI5m)&d2B2#t--c_W0r>#mX9ov>gVL5OM_$-~G zNm>i>i^zIQ)q!YER@cI{W!d`h*I<`{;5;k_1|EA1PZ8y9S$b0pr5}6T8Vro`*yAND zf!~&82!Js2*khl6>=6tq&E;%Y2-$;|1L)nLf_5$9)U;($yMJdcZv54^CJ*82^le2Y zESBlQveo5SfN&%McJs;nK=zxmu+jXQzU}~BShk=Qd)^#BZGb^JkOBsAsqJiwjP?Hl z_|fQ^bzDg)*r0xRNzn%^xboM21RU zlEV-g2c#5ymtSX35LfJ#5yawk_)>5V=<&PLydd(Vw}1E3OYUIDhN?ss`o9+Eq`Z9M zd6V~%P*ml)%@L&GsA-(zn)X^P|ko2>t@y7*-7qDA=Nu3Pw4MI*@{StMQA|e6Y{s|+Gzt2$1+){lhH7q57m6l17=jd zZ2f+~YJ5=nQfv!;<;!+$DqptfOHe>kM9O^bw zaX7?3TZj)ue=#y??{1MkgAfa((=h(aLVPE2dioxby9BE;HCRVcuoS^P;PJPKr*3HR zb{6qq`Co}|y!H*tqstff?27+$A^u8KelJb`%Z5S$f7j0&a$kxi@bo2~W-;o$gTES` z_cT3^OShsfdbsOBI!>Q|JPpt3O#Dv^v5HGMNx|b%W|yB~WFqB-BI(-n0^N9_RbZ8u z#ieUcNUNkfo)izlI38%77n5EnHffd@n)j}G$_quNhk2pMU!N{P+zZ8ianD@HnP|BM z#U8*(c_AMUd6&IVNN-4bp#TSo?$7hM7m6%K6)%)JbL5sd>4gHUBQ7P!M6z>oxiNYp zPwxS8)8HFh|3-eRZsZXs?4&E0Gq^9O=ko#PAzu~Hn9SC{FW;*B@{(2`BRx`xj|rWQ z`QV6bH{3e}NKhT-=6yR9NbgV|4NyaxK#BQ%-yZc*%d$g#VuIQk z0w^gbyZGl2KXDB)p7IsBM|$MfkN>*O!QUtT%JMHoa_|Hgl$%L-oN$1U_}}3yT`5yE zwp(-uFXuUUp5)Eh6(~jLY!RJQSET6~)h*O(G2a}{bBX{&L|KtEpV(a)(S4*9panD?h@V)9_eF%7B>WMftZzN0 zeMu5Nrr?lC!!O3IGm@F)Q~Hav>6}w7H*TGNrihviCO6x5BgtXZI)ltGT*&i?bw)HF zI`L9lOC^MuT|})jNCYE76y{FH*;^)El37?Ogp^+(F&q(z8Bw!mw9|m;kStlbX*79&9!Iu*Me4e#G7F58$?z zsR@-33b$(hZLJzVvTD;r4-@(go8zf-V0n(Gw2O=%QwmxcK~L{KX3fK|j2Kj9gh}tp z7vFHD_qpO=pLQL_!19L@mn2Chu%aT#X6j*%~D3{@u_$+;Q<0>fzJx@#{GMb-n?<^^vws$ zCtknzi8oi@_}}dndt*J0?gzDJ?4L+;gFkHG4-NdGg+FxghaUd0g+C1N2RMT@oWUB- zU=3%mhBH{h8LZ(9)^G-EID<8u!5YqB181;-GuXfxY~Tzwa0VMVgAJU)2F_puXRv`Y zXy6PQID-bxpn)@J;0zi#g9gr^fiq~}3>r9t7S5oBGmtNVI}2&y4CI5);7{NTS~!Ci z&Y*=eXyFXxdq!@w4$h#1Gmx#R!C!+jkUfdPpTHS(a0aqh6!>d!2C_$x`xDu;7yJNc zAWQDRpTHT&QY-K$a0We`K@VpjxrOd`wr~cLJO%y)&R`2?Ai;a^*We5!gbV%z&R`2? zu!S>_Cqwr;1Dt_8Hh@2YGZ^3u1~`KO&R~Eu7~l*BID-++V1zRm;S5GN1NmT-o7)Iy zFv1y(a0Vls0rZIt^ob4hiOmU^L309T(42r7G$&vN%?X%6a{^}2oPZfLCqM@c^ob4h zi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L z4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9LO$+S+ z`osqM#0L7r2KvMX`osqM#0L7r2KvONgO&(=V$(rOgg&u>KC$VbB|@LrK%dz3a4$fg z*z|A)&?h!MoB{NS4fKf(^ob4hi4F9L%@)o8`osqM#AXXu0s6#d3wI;*i4F9L4fKf( z^ob4hiOm500rZIt^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4hi4F9L4fKf(^ob4h zi4F9LE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVr zE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k z^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF> zi7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVr zE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k z^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVrE%b>k^ocF>i7oVr9rTGE^obqx zi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb z9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE z^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqx zi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb z9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE z^obqxi5>Kb9rTGE^obqxi5>Kb9rTGE^obqxi5>KbJ@kn^^oc$6i9Pg*J@kn^^oc$6 zi9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg* zJ@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^ z^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6 zi9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kn^^oc$6i9Pg*J@kpPI@7^tJ@kqF z|Iggp#mJUj=V83VpBWxXv{W)F<|-5+ixkNsIo(y~oT~0S9CCO`Ngi9G2#I0d)aZL} zdDxh~x39aKO^@CT$8ek=4|xj`=gA0yAW!j25Cln{0>kiA5CpPgIgbBgD~e(}maG_t zp;T9QpIZBz?>qb3YgLo+gLT<+*4pdaYwh)W>TK&17uF{(tWR86pSZ9-abbPp!urI8 z^@$7X6BpJeF04;nSf9ACK5=1v;==mGh4qOG>k}8&CoZf{Tv(sDus(5Ned5CU#D(>V z3+oe?)+a8lPh47`xU@cTX?^0-`oyL6iA(Dfm)0jPtxsH9pSZL>acOl2sOCoZi|Tw0&Fv_5fZed5yk#HIC#OY0Mt z)+a8lPh47`xU@cTX?^0-`oyL6iA(Dfm)0jPtxsH9pSZL>acOl2sOCoZi|Tw0&Fv_5fZed5yk#HIC#OY0Mt)+a8l zPh47`xU@cTX?^0-`oyL6iA(Dfm)0jPtxsH9pSZL>acOl2sOCoZi|Tw0&Fv_5fZed5yk#HIC#OY0Mt)+a8lPh47` zxU@cTJ?YzAv#*G|kUOO$XrxqA=y!;fEm?ZcDz{py1U)t|V1lj`91(;$8E zwoBc;A4K_AzgYd)%`o1;9sDH$eEY)9`fV)>pn9kJv2@ptW~2M|@W*f80@yB`Fr-`A zR6nwPqhP(r%{v7{Zfm7(pIa;SBd2E<7e}+b%j0)0?%w?Q$-CYC=;I%G?bXTq9)IZO zAuGwNRFZF1w~tSL^w*vqpB|t8+V`G6KR$o&pxx7X7>g^sq{BZZm#KWrlr?K0aHR72K>@yqKXEw0UY+#?+z`p)E&-O^nY+#?+ zz&=|C_N_E#JUX*MerALG%m(?H4e~P^&uoyN*&si&L4LLl@>^-lJZ5GC{>%pa znGN_e8}Mf~;LmKppV@#vvjKl*1O99s@VC;Kq;_V5|I7yenGOCk8~kTB_|I(cpV{C) zv%!C6ga2$D{I}AWnb^!G0yCQk%xoesvx&gWCIT~?2+V9EFtdrk%q9Y}bt15p#*D;f zHYu3dq+n)~f|*SUW;Q99*`#1*lY*H|3T8Gbn5~n7tu)r~ZGtef3Bt@K2s4`?%xr=% zvkAh?CI~Z|Ak1unFk2@GTWQQtaAuQ*nN1dEHd&b2WMO8Lg_%tjW;R)v*<@j6lZDwj zS=dTr-VvDD#9?L=hnYA-k5*4=HCnR@1^w)JB{w~wa?i8+la9C7~8)a5!U`<`*+LJh_H4R+h>dj=8?6h z=pI=+itWFR2x~X7{ksui?IgB;HzKTE#P;t-gtdd{cES94?H;wBSCv=amUBdR?MufFP*#6y!uyzOAzZ((O&S3j@Bf`3# z_|mU$zPPn|{$%y+*~yD{F7~$fvOQT{9^ZXq>Gqvs?eV*x`8<4mggdxET7ACbKOgXa zd3JGfd2;sR$RN7?Ow=ue0-A1}X0XoX*}JP3N0;xtT=A8PsJgX=d-naa^Jjb&@%fJb z*Ke`BczX71ee2oj@q4TDqy4?(%k{Uiu6J(l=8g7oJBTppB>6_MB*1WYN@kUOHi`D6$+Ix5P9@+yBakuiD z^2!jqkGyd8KmEPcIYh3l-Y4%?&l=BW*Kf!ndM|@c0u8|Cx7A4VKJspLP5hhH1Nesw zIthsBp1inRoj+eaJ2}2w-Cht0&c27ZTbXPfSQ#Ssk$0;f;!~%@fO67q)kos0$lp6? zyLH>qb*XPJBkNXM3ade#mJ)U=!mvEsMQj(DbdtcRJjw*=I>e=u1$q}C3)R1uK_>wW z%v1ts6Mhz*6wpmsDT-rzhPYc9Iw?pDvHQpiSMLYs=XaT%wtAnuTRm$$uGl`%t)2txJ7-8Cka)4AH&SpiWB(yA^RPEG^taTsm2P9%VMOQCX0> zV{AHU-cqHZM&Rbq$?&*L251@q9-SOt&dMp|TXQ+8|OsBLp_rC}4|(#e8@ z_NXjK=P@>&G}v1JX{a5zIdn4MxJ_k%wh`db$$>qWm7_fNXQ;civEPH#5WkbYbp3wp z&cwZe{C)av{k#?VVn30|U=M~IIvLm*0P$0f{c5$( z_8+Wm^_SiM41I?k$M3Ev9UbgDXuEZB_y+5uIN%}fR>m>BTbcU&Lb`4}9Kn0*LERQp zc5A|({(KibeUV8g3CtfxnIMISxOB3>rXY}ont+!^}0s)%)SPbG6#)ee!PgtnuW6y_Z2Jfd-%-stfV=$-C7x@rPl$Ti?r| zlYp2m2$dkrdx*Q0$^}0s)%(G@i!Ivfee!Pgtns*F`(X2K*02z93*sTa-d$Wt!E;8vPfl+ys3DR|lOD7BTE44cdKWO#}(TL zn^zj!2Mqebxe$M!yb%8|Gv0Y@+NdlwtC=;aX z5SLCC=p2A7RR3NEodhs2Qwg9=_*ryPKp$tND30wJ;%;Ra9w9Nr?jtW;y&s&r)2OZ9 zC+}9z8jmZs4>qqfwhtKegL5JNK6xSjVQB8w*EVp-pp$^uJqVF%Y;Pm69TJz=JrLZI zcer6}Z~d`-l(btF#%Dl^{Jn#=TNeiDZe8lz%gDObV2JLmW|h)eb17lBB8=^eUBq^g zNhb-6%A-t>u0vcpS)g+OvQYhd8FUiBz)U58HsNQ{NdbMFm7+MdXNbF%VR(eZ5WA1O zaP@w0?oOk&dY`;oJ!?F!*gn|2(%3#=&=1aq`1|CA_=lmnTVLD2A%ji=V)r0KsBMR9D;5O*uX@Cb<^b{~1+>iyu{oknf-K6$r#)_7d8eXx0@v3uDWBWlQwnO3)y9a_>@(wqQ?FatYK1$lH3ga^%MgHDF+pP@1X70g+aPom-_ZHvTijPqI;`ZrF7O@O4zOVnVawFygWNQU4M?A;u(LezizHgyZa{SPyG7L2X~G(C+qfRfrrT-(_06?f9qEG_#NH%S10d$_wwRy zkx%yjp0dRM`HpM#(dt9jS90B5HCBB(xqxhbdsWv5fBpKQhp%q_&(W+pt!C$|muKgf zN8fvK{a;RwPfvbh_3Y^C_FB1%z4gUsZ(qE<{ZH$Q<|f0xC3ml}8^o%=`k-nu%e{SZ zchj0#bMi6x=ZBx_UuMy}I>BC$_JBz>omU?^dU|?%aWVN^^>dR?pPoH^^j3y(^}jr* z79J#P^$;wd=`RVdoS+~MIuNf{Uz|)K2+YykkN2SZpS-x{4A;;0ae*s4ri)JcFILmZ z=b-e6&dScV=Gy9fx~dtbpYCHyzqK-*b#OjgeP;5r5KfOoLh{$VNE(J@aDujCZ33Hg zIQldamf!JWX&ILN8?DfbP6i3;^y{6T4y)y)HF_#wBBA|9UbG8__H`ZYbG;j+jmcqwj(dzG*#E)}o2E%^DeRl%>5q4~zfyg9^2mkzhLZ%`@4RSEg`F4e z+~~2FvyheR(P{KvG)qD$!{D2@b`GNEzvE0Kow`HFoMuQU zgb+zL$4ktA$5`AmRGJ-0P)(Wn@#>>TH_r3u;=9K$R~M5HRIfaGpBhC6TVPn*EvVel za!xZM6fL1VvddKKS-MK)KdUtVkOx@*-r&6iNE3RKfeeV9)c zYuNk(b~`F}tyyR4i!o%T;dYRnxo*J1ViE-u$Tz?3m84?O1M54{f6l>Kn78Vi>A%3VdJCdN9f|BXgy9I`|-2#8D+R%&$ zMazLouH`9-mYjiwLpAD7?;D!`NM|)EGrearVIskG8*oE&AjyEYHRRN5sD_0E)m^JL zG!H_xsV zM_&N7Ec!AAl1ANg)rRI%s8)rS$*8p|CYIp2Yt@$KP^ebzNx0FUScW1}37UJZ+S0rU z)vBF*J&!Bas*GHM=&n^;nk%7N709@gT9xv(;Qc8>OS2@DG7Rn^*=4Od&OCwZ4k0bg zkWdIA5^YYVnE#Hku#jk);D<`HBMGW0D7{b(3kj;bv)j^)2t~_*O1fHAky4$j(p3Uy zmF7RvSxw4JF^wOfNSG-539h?VUC?|9)v7=>z08S~=VA?;U%+mMf3CWqITWf@ zN2MyNRRvBK;O?XEg637&d4c+?tX3VlXHELClD}46(3}cKUjVhFR)s(}3?z-ZYt;qK zr%ZZ${v(~$q|Ef5)r5%z z*MC-N4kQ_HVWBzoTKRSz|2x1-ng^j;)!JWDbCuPqIX04Z{l-jlqnt4(kT<^AVa!7& z5?pt!x}^CMs#Sq%dYMzOR?RP9x5HnnE@=*hYSmGxifUDXlLff@T6IbDD(t*K0e5?v znWklb#aY`e@YkvbG^fJR7s#_Y#SEw%l^Np%rn^=>p!pQ4RpDiyYORWiEevX~?ppPL z=1{0s?MYZtt0Gbfn!8p#pm`IjRXh259#^bY8My?}U8^3@TnW{xK*pWSRViN!-ao53 zpji@183xx}CbcSaHexC<&OCwZ&VLVRhJ->0k!W)=tvt+gsQK^6I{vZL0nLsisHV^q zqv8OCVWI3+AX?Il2t`XM=PsfpXKdjI&MM7+q_dionY%u#3G)Q5`>Y<)97ra#jmshK`Dkhc;)?KR}(i{rasyzue`V;4>h*W~+ zu2m0d-h^t^PQISU6>C*SETlB#B9ZH%kp<)#XSu|Ls^b*S7^`oSD5=t5d7dI!> zsv@fT38vANs^9kGrAZQsB1DSK;bi%~Slf-Uwi_A2N;4!0Yhkge)&v9K+jXjc=+`Sv zi%`gfGK^|fjFOyjqSL^6rCE@4UX!9yO!5bq5+)WGE#m^`m8L?H4V!w?doB$N3F|-e z+c3?CP|a%9lGI*hHEWKIggIc(G)c}yCrHolinS{vw^*os`fJx2&6-f{3WO}GT`9eU(w{rbXugDUho!mp z)O~}^I1>r4KNp_STnR-IBI@RJin;I@i~Gh(^CSuDwx(3e(g9Y(Lc;2gn*1}~C~87M zOQI%cAmQ|%*Rn5p55!E$PBGUXXCmSCpVzWac@t4X=k*6yyUuAwglbodH7UW$YS$ba zNxuPmrdd)h{v?<;8dgMAd#7|;u2-u3wd10`hSksQNwi_A2 zO7kQMt0_Fam<$VxiQ2KhcIBV-Mp1LH`PYJ$L`}{>!WlTPG!v4}tEYAyU`?1vc>S4Q zO|v1%hE3t=wPC|T!s?&9*0k4zYFDe4q>d}AU2|+C%zk_3ANkJNbAowucEw%{MZ94m z;q}+9{A=H+b_JsLawxGz%`=j4`)gPJ*>6<4j*7dJxvQ{(-H|{P=3fEF9t@OYWwqzC^BDE_fmhky&SN?%-RJ-;>+-Oj&T@k5- z&0o7VG=oC5YbRjO@3QP)tX&yfi#<@gHZ*HOwJQ*&sCK3F77aaxf!u-SODK0Z9Iai& z++m!Fgx8-7^N)F>NJ2#2oK7(p9%FIeSc~3P8VG6%PcK%(Lc;2=UHNCeQPhN3H(F&r zNqmcmGq(6E|9LI@()U1|r0n#b*My11Mg3FYywbUxbY4y2>9t|QLc;2=U0d2~Lba>a zZBoaT)vh@<66SzC(|KGj{v?<;XIBwb?VZwXxn3zScWr6rglbnHMK6aEYu7v@W%B}a zSN<7rRJ)FfTQqkSI7!F@9*kyK*n@#`tgLoD{NdHE{Bz(q3InX0lPpJD#z4}kzjo!{ z0Y|m#VWf7cM#?aDv+jcV7Ph#L)xwJRc(u=#6O{!MRGyLJNh{4UEA#oCpzwb%o- zEB}}`s$GFFH$K)%!*aW+__blSNcB(LZ-4V|dC#7#&M)^?FP@$~Tb=Kn9>2FbKbq~2 z7O&r^K6Lc-^!Vap@~AqPEKbj!K6H zqoui$hPJT!oVvGUyHNFy{JN$6rY}jly%u7lmA5+0E~;iS*EQR5kn|cjw=^Tt&TUeE zy3cLGMMCU9xBMgEnkAbW)NQ1OhlKW@yDfQ1`%ZPvl9QtDHtqd|&Yr(qJ(?Uy_Kf&Q zsQp&WKM9_->O|Vj8TMH9kc))aKab>J2S+ts5`S-H?Il&>?>z-h67qlPw{Y5(T8I|D zMLs@?eY-RMuU2nP-Y7Bt{t|Xi0yzl(ayap2phmkr*({0HeH`bVp!G-PCC#`_R>O|z}fW1CL=&8BwW`%vtI#d40)ODO$wp0ZDWBPnnEt=M7W3?!WXY=VFL z8)g%;oz%ADTqMN)9GQO(9L5zQ>*l0IRNk9hORZrARuXoCbdx{$%yB$xm#=?2+32ul}-E*7-jD&g4gK7t}ZR zr#q?tXf>Jqn3a07qwaTQ=KFNob^ET&^Czok&rV*vbFugRv-4+1tL^_jes=u)=sDeK z{tnHS1!0+JAb^d215(3wHFaaS((Yw&d)xUw{mRkpyK37_zC3$(_2TIAy_YMXJ>E$a z8CYF0TkwBPd%jO+RV6sDra_#dKHcwi&G%`JaQj}@Rvj12_&Rzg*^7pqSa!Ho~ zjI`IHW4xM%Fp7HoO~2QtbEwbUOlh`?NYlH$TX%ZUx%$_uFHWZ0IBGmErj)c7Kt}Th z3E9=ui%is6AWqP%Vh|^`JHJU*@06y*2|_Aq&5wm`_o zj=n&7r^r9ioxZoKuS~vdPG9miro>CqYydsYZk*_^rn^Hg>fZhK&U{AmAJpL_`nt}Y z?=&xVI0a6U-UG;KM&v?%rS^8S$nRh0G)IzXafPS zM{~NI*Gsf!W4y!XQ(lt(1L$d9=S1)F`2@YF`#<-J^Ip^Woc7Az(9Zw1?Qs2)%AE)J zX6FIEQ9Yi#Rp0?IKgk9HyFhcjz%E=(ce@LsE&O@EEzsQW%cZuEbmkQ0O@F1cn4hn{ zKKYuCbmb^&8dUWXY52+vC4Am4uC?mC zwHwT-KIvKZcOCfm$Whn9{;pwgTj^a47JDudFG=G8^t9*ZL@)PT1oWcr{dE=pZaJzh zh{QI+iglI1Nz!`&IqkK%kjs7cEb{x;InBo;4X_Zj$Z;J3wNxm#=u^;tro}JT;?a)fPWYG*V+TSpy^H>u8UiB z6_^+EZ@Hs+@u)ya+f5va>kK7p@XyWpH{T=sAkWR!ebDSHQ1SBb!{e+7OaarpK(6Vt zoFohI=Y#yK@h;>A`CtaQsB?eS%fBR#t6m_l%>%@_`3zmuVfje<4xpwvpaV6Z(ClV6 zQX2@&%@;K9LzCq08ySJQ`6v%b$Nstbg64UJXs^`y_z2omhqQa;pC&J8UWca1JxT4! zWADDJUo=WTAY?gPeFxqU;9rz4LaJ#*9j?<2y$a0D`Pbyp6uA=tXXs3?;uJZ5X%-$mqHC&ye|-S9UrE>v9*iEShm>neeh zVncz7a!GSA7jn6-$|Apiozr|wLf)VG%CnUL@*$^t);a%9y$d<7uDqw9zayzF1l~dC zU#>@0CDfEX+lbMp-9RHrm=^T%`k6?-Y74X3^G+
    @se~N zKu(6}szJ16^(t7|o&Es6iHBG6n6_A6mb*WdCe`?IX zevhWcR+MDYm-ffSxw*(q(tThDXfEg80eNmN?SN)YfhvZ789(Rs5*yeWq|4`1UXuR( zKA(Rb--(`QcH-Lv)GD?+oZ3X?pBVG+fnA{Q%LR5p{=K=l z3#ra&?&qHu^Ka^-dGV;!N%u~io9hfEYY6NE%>pC)ps7#&QM$TWSAi}E|3*K~iogst z%?sqpHp@w}0DnHnzuWIZUeMpkAlHm2Q0wzA`Qxe=$h(bF&(*K2&40*q^Fx{gI#3tX z`lGui>e@d^KBT=oG)bP{&65Ii^HCnsnRcVC`P~DBXyv*22wG9M?zwr*?=V2qoFbXLr}3-lwxu z)ODzpxw+U~ka#JE1_^n8wkyw82FQn;?px%5O#)VX_tT=V-FP*n*vWly;{ zH_!5s^c_G=yJrs6axI-fE$Z4I&u8@8ZBj%VPJ5qAx237?!40+tEEtIAGn&CU(fi~1 z5SXa@Kvbi-oaEt#({67(A99k;dH^}i<6Ox7zP*oJ)VX_VT+e6@ho;8%_$71Bv_CG+ z%|&jK?)^K!?`ClCfIK&sc0jZNzt89QGvvHpVgokDD?0P|e70b~=kq%noakLXp8yke z@17Xf{GJ9hF$VUEyK?N;EaI2T<#nu``QIL>=QP&~>;iv(Vz;}H+KB)6=JlNBerR4i zs-vWPC(g}vhF0sle}AB!(=0Hu5AyHL)qP0yPWQo|74bV9a8?ASfN5U9yn>u3W>>J5 z;eTVGo)_kW8MzAb!7L|>km}rD_0DNNh^t;8ugwGGT0g_*o>B9g862p!bE&>kP>;V- zaOMBKdCl)(K$GNp2XpgL9+Hmzb8~+8LLpjtZa#umvys3wncug7rpY}??FkltS3bf) z(yRXs0e;6q5mI?>PLYZ__0P@uy$WcG+{uYEba8H;;UeiafS6`t#fas(`6yyh$NsrF zzh42(kb$g8<}Lo-JjX`TasV^Uw#qTfbMtY`qMrRd7k=jg>N+$#oSRF$B#ryg^ZOT^ z=y|OzzD*$PN@^2<>Vn_Jfa)qDnT^qkw$xu~<(G>fFPfodMUm?&fs?X_5#&X8+7IOI z&wS;%xxneZk<;FqWbJ+A{G^LrgzS#=&bhyq=Jzq6suId&Pq|o2XZc9_4xpwvn*%l1 z_HK5gr|UpOqq9>fq7A3L@^23Z`bNAYod?j<49l2Y+u%Z8kPl{%i#qpLz5K2QT=fEZZ5|+3w;4W? zz5}Rf4(LF=gIa%dH&Po2%+2{d3}}+v>|ky_%0tqze{Rn2UMNH>&&@~BraGjxynmX^ z?^{6AP9n7Xzv;h{QHVi*=R2Nz!`&IqkK%kZYP! z-x$blPvF~aHNTfZLf)Sq%d?dMs3E8O);Z13T*&2GI*VMh9{;!7YJMLBsw$zp_LPgY zbe4~#Z+|V#?_zMEmVa-aK`rV!5YcF_Op0h|)-BFZ2FMXFN#}m_{4NG3dQD&IFE52H ziuiVZAA{uK`ZHg@Zy$1!^zM)6{7wcJa=&lyBiHQ5|DE=l-^+le#?}CnNnhF@7w6_8 zH%a&Yx%mOj<=i`Gd8ZMq87aQuFToYzZiz{Yro&!@a3{RezL&Fh@#T{Rp*uUS&y zm)88A1~f4SCl+_*#JRc5PqG33-28y{=K{OHTR6MjozzAGzc=UiHK2L%sP2;Poj5nw z8A{gRpPTcW8zTE4&&}0+NcDaX{T+UX1I~)Tv@gvI(3GKf_1TcK|ib0Uf9dYW-2vqOSduWPT3=nk2V7n46FCkaQeC zOY^)!wDR111g)rB|1_E3w}7U}JxT2e7U$+893;I4kkZ_(2&p_br$|Me`lrbJUIjEo z?&QQ7x;Qt_aFKNDM%?hb6^aqdbMsNeqK@4& zX`vjm{Co3pIHI22U5AF>xq!M3iyh9*C0>f*_&ZJwzkk7rUelM_XBW08&}VP>T@0wY zAi~%ft!PV~9_N?y|LVHoH!`@8%XL*2lx9B>xm$z`I!s3 zU*~<~qR!n_Wy9}dKvgBwls)BQEuG~f>D!N*-^JiS&3mZ3*^Sf&0uhbh!yrX8H0u^; zC    XUXsrJ+02ag%be)t8H#{D)$#88qz%81LGp0@nXf!IC)lmeko4~ElQuKjH*+EP z`}Y1GiaK{sjT?S11DYCJO(m1Qv_CEmghg(W?gKkO`)TeSkbiG3?SNe}_)4a}!-sSTNdQtc8iE+d4X+RTWV6V6LK7dQOA1~e}ol{)F(iF0$Ep=1q#eV|!jWFIv3srQ?!n-%Hr@H-rERs^Pi zXhXGBJ7dx1nkMdA##sBO2hTpwVh*qAPkD%3T#Q&Z4hTpe-@={ta$=4=ks@;m4m)S|Ba5v`&3Fh~&%&AP=I$^bdyCFwkXp3YC5=(%U#-L9lI z5%BH&J_gCd^=H0*-#+9d>D?dC8=A+tko$dmA9<=<>f8NO_Q%D! zxyVh@eP9P@F6Z6>e(HC*GpTI^-j(9_GvvHpVgokDJG>j^CF$Sq^Z6YOPV}xCj-XF< zOwYf6V$APpKoetdVsTeaoSV!1BpV3q0?qXTyCBca#a$3>!9OqN_cfq-@u<{E_fDLf z>kK7p2C~p03@$z;0UlZ8ZMdtS^peb@EC(h8BUd4HOhI^vh0AiYn6(g4C=A+0&9sB3z{C)*ALk6-YnKx6i zI5*F+PqZAsOtYr>JI9f;seNMlaIZ9@tx!IljA3+ ztG8dAT%H`Cp8Uq@*+U8WaN6zF`H<6n>zw9iF65e~)UPWH45Pn2SAn;&`F#wis)VB0 zQ!ZAISw51!{k8Ox=4=ks^4vUwTC-I)2#7ok-Kb|jX2IoXC&rk+> zAEFm^AMovThAMfu{>)dNn-lETXGnSvAg6hp3pvl?ywR?|x2Y}ozti37L4D1zm<;1zU%ulj`z%I~SFR%;p++5rR(H8vkVt!u(nir2s zopkTSxvr5At`~)qN1{!JifJI~;IU1ZJOUUclP8a#oaG!CFS3v&HXi za3SA8KA1tC+Cre#=XW*Wsu##*^8mTl&+xH&r5TXF*5@}fI8bwK?`Ai8y7o_!`8^D1 zl6eG(55=17VV##^ZORiG`T0KJ;CC9e1wCf*8ozQyA>go z=jIfts8jzGncu5`rpTR~I71ib;~6fJZUcyECRU7C)6R4k+j157?KQt&0nLzsc9P6n zoR8<&NLu#K%@1k!tsJvFHy_8WSxcbL&hK16U5CRR&dnuWlE&RVmzLkZ;6yL?TmrH04G< zSk7NdxBN~97jpSsiGe!%pfv08*U~M&j{#McP+oh=#p*H3N3om;YMQe-P|I`k3~Eu= zfrv(Dr&2^4PJ5rLmx>IKBVLlu1L$c6=S1(1=R@?O?gLSc&QK)}*Prdmb9173>oX+1 z`;qfI8C=NyzP-PPqR!n@t%kODG6Jww++?7*wHWi2K z>Z1Nz-rE_?^#Z#f&&@@EqAj@R#qEsderR4iDs|GmW8+eBo~|!stb4FZTd4{OxW8YW+FQ0Uf9dYW-1k;`w$@l3RWc1DYfsj?K-Po*!2qI(m9~d~q>3 zs1}pv^z7-Qx1OxdFZWh2o}N8ho$s9t8M-(i&Tvtz zBZ8P_V#SE%-b#G<##R|fcQAuR|e;QvJ19ftG+sUlLcFi znWXaoewuN4@$=UecepR9T~zU;)BabhHz#kDYTvSwbRWP@Gcg}_ zuKAtq$?o*tvfrI+*!#oZ=TkSvA*=h=JI&O5*spenJ=HJEbF1I?pLg0@Yozmjtt@*) zf3Z95Z&q(i4)xRCXC&!7fSYD;9^6;c9pM)B9f)tVcc#U+;rRFY)E&nVGfD4${QPzX zFaE1t;urPrk8%8t2F=&?2f+b9KV&88K7gI(c0TL@Ki|hL>iy5Y;?3*%eGQLtem)ri zrj7Go?!5iitFKMIDtY@6c9Q-BdqDF${~laTce)3n9R&Oz&GWMUFHxM0`HI?HPLi!y z1@O~c&x_yZ{|Piv|NdDjzsCX15R;PpR%G0^lWag7tjenfW~mL$|04UqE502=Pwk}o z3;wu3^T9X8xUjW{q{a8GE;Rf?^^M8nlF&dnDkc`#34XsrXeX|wyWWXZ_p~PT=Slpo z2a+d&abcP#=)Hg}E6EB1*h_Q7jKC#1VwRPJ-Cq;)dmTtk3LeS{9OS16XN3SB8~maN#{j^&ObS^7|OjTymXFk7n-T zEH=kR(yxCO%Wq{U$DAsHB~a6yY)s(0ur0rv0loc&6qID-;=65GK9Z*WZ~F0j8XTzk zao){tq^1<;OYl1zU|(XU%-z|`Z)(3@TEtpI1V7Eey!iFLjDR!MGjkb%xjVnl0oGhZ zWE*hRnoD41b$lP#XfQD*T*BHz#O7(v=EEM)dmp2y zcYm$T?{ge$aXbCNdI3MrZ+h@y5BT{$EK%?NnKQrd z0nVJQn;L19sWWGhopkmCdqDF${~qXbcWDnqI|%qcn&)NxUt$L~!}s`q%1p9>0DhY5 zdGY)FKfy2R-#>Tf_dvjTGceliRP@Q}@_6s~+41wE=TmjAEHjksz`qatHi*bR$iHEp ziTfbhgMaSM?}UKU=25YeE}uGW);UTR5!eZu8HRR3pS!C&A=-pLN91=!kUR;D57Vh! zycdvN!dgclM?9eYKOc79_|;S54EEF>0yQze7lPEpKrovZsB`z(UWSp?Ev*UtJ&*&M z7kY4)^a4h4r+THh{c~u3?*p7e&+g`T&sHxkR(skBSLf5C%&q>b0A8B=mEzUs?jyKF z{rYFo{Ei1Wi|&bPPqI359${g1NNw4VmEY`8hV^P1#wzO7KZEA?IlviorxcvID~ge4 z(Bo?b-ZeO+*;zSe`8Uw(Ed91lLt7{6**}5icRRodG!Qn)$kj=6j*nt6fp5(m(hSUj znx|j8*^Sf&s=w}jM~dJ1@Ttu^QZLWWPS>Ap>x=&}kRp&8I9lB;Pc5>Y4r zwO=Rn_W~cgaoI=TU!A=3-OG!+2G!>}UHo+QZ%!U=bb+6gbl{eJ!S6(Xl3z_#$;CbY zue^J{;P)9!Zg^YUP7$RNL zYdNjga$2wDv|h_;y_U1^TF%02ISa4lEWDPp@LJBoYdH(AoASU->9(Pn+lFRt8=ARoXy&$|ncIeDZX24pZD{7Up_$u; zW^Nl=H9(PDciFq9+lJQtW$&798(MdmoolXJ%w9^j4XvBZ z-ZkAev~Dwd*L2&^y3y=i(``fRR zd)IW^(7GY*UDItt>z1^4O}7oLo6^oT*Ij8ZrQ3$qeQEESZW~&6roC&rZD`$__O9u+ zp>=oKx#qe(?WJ_v(7HkGUDItt>lU?lO}7oLo7CPl-8QstQ+wC++R)tHr=IV7ZD_vl zwW0aG*M{c%UK^V4du?dG@3o=%zSoB4`(7KG@4Ict-KUbbj5 zJ$Lu1=k7lB+})?1yZh90cb|Ih?o-d#uR#0!?cCj`p08hm_O9vnL+jU|y=%Jt(E3Ga z@0xBu1|DuIaX+^*hndHQjycxw}t2 zU%wXZmDX)T>ldTFYr1V{{c5y#O}7oLUyk;!>9!$vpL)K2KiVm+yH7n|zai~i(``fR zcci^*x@~Cvmb7ZA0tVrM+vqZD{?% zw0BLn4Y~W&^YuH^PHEkJ>iPPuY44hD8(P0N?OoGtL+dxEy=%H{X#MWAb4_=jdhYI1 z&)2U{d!=>T(E0^x@0xBKTE9Z=UDItt>zAm#Yr1X7-KU+Vy}*KbmL*L2&^ z`dw=8nr<6fzfJ93(``fU_o-`lpSpJUscUzix_0-eYj>ZzcK4}kcb~d;_o-`lpSpJU zscUzix_0-eYj>ZzcK4}kcb~d;_o-`lpSpJUscUzix_0-eYj>ZzcK4}kcb~d;_o-`l zpSpIxPhGqF)V2G4>e}6>uHElb*X}-b?e0_8?ml(x?o-$9_o-`lpSpIxPhGqF)V2G4 z>e}6>uHAj=+TEwF-F@oX{XTW=?o-$9_o-`lpSpIxPhGqF)V2G4>e}6>uHAj=+TEwF z-F@oX{XTW=?o-$9_o-`lpSpIxPhGqF)U~@$UAz0#wYyJUyWgj--F@oX{XTW=?o-$9 z_o-`lpSpIxPhGqF)U~@$UAz0#wYyJUyWgj--F@oX{XTW=?o-$9_o-`lpSpJUscUzi zx_0-eYxn!qwYyJUyWgj--F@oX{XTW=?o-$9_o-`lpSpJUscUzix_0-eYxn!qwYyJU zyWgj--F@oX{XTW=?o&7JK6T^nQ#bBDb>r?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=lr?+H|{=l zr?+H|{=lr?+ zx9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fu zb?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHv zQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q! zK6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v z?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx z?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5 z>+Vyx?ml(v?o+q!K6UHvQ@8Fub?fd^x9&c5>+Vyx?ml(v?o%(^ed>j~PrcZm|FIw4 zd$zhddAfS^^>6)KkM^gpRu8^Cd1b$vR@DcN|IEq7z_`sjm`_rLaAAE^G~gX;6=tCwfzmq*`waq{Bi^5ppR zFk}8r^lz?JiDA!+kZd&bcXEW^7!eyeNjhc`Afh5pvrOm z`q_(%%gF~f|EJIVul(MF>VsFur{7ym{`22^@c1wP-hG~&z>Bg9$j3n zUOxWRlhyg<-s<_2)w5?OFW$M>`~KPav!m_*e*BsK1D9v-u3j8nzV~uvAN$$$*v9uI}t{`Yr|pFcl-ynn;s#nZEAtMk3nYwDx@ zz2nQv)r;%3_AXYZe`@bfe{XgE-rl>b^B1er$FJ)YPfm_6wCCSlb^O1m@V|b3!(Ebp zqqZXXZzwh^%ro-bSpWLaytr81t&x)|q1mj`Xsc}1w^46fN_C@jtiqEQm#g#Vt7j+2 zm#Y*GRtb211Lbp2s08w@Dg9jj$qmQp*mJ!t`GWQ=xBgkN^%IJX=xD14&u6S3KEL6P z>}>FK*7^y>W(8#Z_Iae%oKy+TW))=pV3nS#k@b&KaqFM$0c*asQe^$``3)3g{j}>FK z*7^y>W(8#Z_W8t;N35UHY*s!MUp3Yi7q1dc|tlvJbT0f=Ptb(i`Q1n!d ztbde>TYp<<{qXq>6lDFaZ2i3_Hyn}mchB{<1X(|LmRtX#*!l^@Ms#HTgXc5W51-#~ zN7g@hI&1xeVzUCWe*3&?{gh_23bKAc(Ni_D{!uD!{fk2DhtF@IAnRYq*57+_!x343 z_grsFkoALSx%DrLt)EbAL`T*?cs^tO@c9jQWc`Duv(`^2HY*_Ox6iBAPiZ!*AnOMd zJyj#?AEn~fzbv$V`1}S6vi_xP{kW(8#Z_IcI%Da~dTWc`4mr)p&Vqg34b4+^axKEHv2tp7l^{@#-t zj>!7E=XzU$tRFnft^cst`U%BGbY%U5=QGw1pWkpt)<1YUYyE^`vjVby`@CxXlxDLE zvVK6(Q#G>wQ7UfzhlSP;pWi@1)_*8lfA7f+M`ZoobGOXD@EU!dY^mM)u z#oLnIsGsZuAAAwj0D=vi&ttg~+_!zQ&kB!jcq02=ZxP-YIH&^;kDLZOJ{-RK0C$j( zY{b_Ha;IUi!uV+zJf1%dgk*yOo(4;R6bw%TCEBb5PXi#?n)T)t2zVMs$wp2C9P8fN zEIbX6U;_u9hHey_ye$GxL+{ZIPk0)-2YVX>PXl=PW;;HyIgu~V&Q8~_0VUFXeE45B z!hG~b??>NXoxJni%ZtY!x&7ap|9Jam2Ym@@@A%pA^P}gFwdbxL--6llgYQh;aV>fJ z>hW#7-SP%6THSHhpS>Es=ankxj^*HWF!b2faHqD9Z8UhtZm^^9xvR%Z;DSeUfx88q zgjbKz0VgY7+tA&5E;4bot>-OTtxtCK7&*xSwmY|dx0ZuQ)SLDAcdJQy8xo2;xlNz$ z)|2$sC%bxln6XO`ZSGcdlje6dS=Cx+ZdVE6)#EoboZHiT9qsOr28gd7f4Ow!#J@zS znJs~g@+Ros;l;L;A!DZ!kXeo5Q#nd>^%y03UFK=S=^dg%f~&_U{phV=Q4~FX^%&)A z9d~Sl#UxjcQF8&TlGs~04L;2_B^+e-cUnQGSs=uAO$e=)kX>iV#MSyi>NyO+gVp+E z_!L1hiwvEj43Vg}VTgJTlbOM4Lqc&U`$0iC>fXr$XpUP1p9NW4&QjKjR z1UR#poT&*9Zuty%O)+fWK2$ET3D7IyYM7<3|6pn3!)rN%PPWFS8 zxRZS{9NVmzNt8)FAc}1gX?|CeRjm$DY@= |%#gxIg3*d`L=*fvU>_!<=3RBC2R zAljsw#XcZI#!e+5+bRafHu^K9vDnuA9!QOCBLq0M^$08q6x-*kg-z<$hL~Xv5o%7a4fcUzxGpO+Xw-UZ9M{u0>w7^ zIF4=I(U_J(5HI1N6{_6K=Iu}vVvc1;MamXKX%$;8$AK?-~1!D@Xn ze2O5MMTTNqhDg-gFht?lHdt*)DDGrGNQpbyC&RJLikU>2)B~c}CXwcMHCff_5XCmS z=9Uo0Hb98|3W{waF^+Ac#EGv#u}!6BwgjS0s#)vqi?qr_~$2Kcw5@k{kh+>;Wn%~uARjWf3+vu8GLLA!wA@(aMwu!_z zwv7@ez6QlMm73WSh&HKau@A_Qu~P}iwu-^Ajegu`EVgxD*->NL2my|5JpzjY#Wwml zj&0rJ*alH-8zsT94Ui=EhGN?|4L;5GLpVjTO(4W}O$e=)kX>iV#MSyi3VY+j%}mFiLXJiO{HeG1forQ*IJQ9)+eS%nYy%{Ty`k7PPILA60DPSw`6w0*ZeGf($M73M zgU@%|+=68l<)c&A>qn~RqMUuYtH~`5~pcpZ$2H?q{Fw z>hYtTjV2){b%%ahVQ-UltuC`?gHYm#li1EJC3Z9@;;7VDkKYvcdAoMvenRBbsz3-z zHB7uhhK`<6KxZ{ePU$Gs6}Eqhl0T@Z!Hdg++osoywob5o!DU)+X$S zb5KMYr@@&BqDgEI)dv|Oa=z^c2+CrgKuN5eP+Dz4E6>uY>-A$8@zsO%`gG(}K{|^L zo~jI$s=r|bBe8F=-jGt=&wdP3_p?t&VxN_C#D~DxC(#Ofo2+Yf31c5!bxTQNAE3m( z2F5;-n#8_Q>cs!R*r$?Hs{&CajS|EwWa#KA1$0~eJjZ!0{UqmIy$2&%_t_0p@Rd@K z*wo82d(PNbCbNiS5DIH%@eQdsFet;Ri@>)b>30z}4eVZnDLM zgWclM|9tiMV>kZe?sKMhtHM8DJ)Yb?Z5J5+?P~gw>Vwa&FLqwvp#AFjUD^NsgKCz! z^!my5<{ZB`dVYNI?%uCm-?mNes{YBlo2ozh_(xuQ|KtOcS6+MY=#}+t-A0`EPah7h z?tZQM(4)=6)qnJ$`f7J&`uHQAY!BY2i}oKs zsJ<2y?Pyjc+=CyiaP#{T?jOIuI&|;cTB}w)!w?P!Hp3kF%Kf?}c9< zskdJ|Tm6~!34j0e)#|~wH~;u}^T!vxzjc20Z0~&a?0Zi) zH_V^Bc(zvZ1u?{tg{R8#uKt}@s{K;wCLg$#@uA;gCb1ljj$cS3dW?&Pb~N;6YWFY{SOoQ{NSL`XSGHG`nB7|Cb(A zUoY9LBqH4jM>kI~+tGlz8}aVV_5aiNS8oJ%ujl+dRvU1=ciMZn3j^2%lR*&ZyY&ZC zzgz#rz^8nu`oN=CCYuyV4>eBj!YuHI?Oo;$?^2Mr#g6Z-aQ}bxE_dQF-D2lguzNA@ zGLL(gdBnTiZcoH9-0kpe7fc4;<^Ge&2UG5Kx}4T#y48RGpe&#~J$v!=_;PR4PrO(k zln;FC`02Y3i?hIf0=5?Jo9i_C!B}kW60YmC>^dF4&bzCB_d&HRiHNrVHuN9-`q_(% z%f2j=yUO;%k6&F^w(on?uWCuQSpC5()x2Dk$!lvV)_eM$`NK$PFC*@$v5YuP$>tBe z``_RA&O&dkVYIbjbMxwN6=vYSy1Ll3U0%KY$HPx&^tLgShYsx$jIU0Ar25deZ-lrm zIDcx8EPS|n_0b0>?|<#LK2ZIqg=h2@)OETw(p>t~otc09@gMCPIMpBC{l9{+@jDNy zHnV?8%d3-b((V7}GxWc7_Rh)EkA(;1*VZq|tbSuPdG+=mWqqOz z*1pgmtA6C-hU-lU{lV(}4N7>tRs^cIT{$i0%|v)Nr|zZ?kQEMgC~@>bGAWpC4bh zLDt=HMQ_)Q*Kgm1?6tGIuCk6VpMLkYm9lQ9ty^N>zUx)Jvj5@gs=^NWe06d5z4LWP zYj5L_i9&z;&a=Jt_(QL~I=L;mw*7mWBUQg#edIPW8hcsy!ajO*J15vU)QibYcPN)A zTqjo_y5VPnyvt-rfa>kXLo2IgOh(K0SN-=&c-$ z8NWQBSbpPFYPOAx;!5wq(`aC=4pz=)*I^eHYUz|)KaLB&;pZ&<@4B5~2_x$DA z#oDCKURYfz`#xQCxV~6TC!d430P}0x^XjW+sD8SSDxK&~XPvB{tv)mPStx6dMA?Jt zuluny49nmoZ3Wv*jir;+r;*V7t{+Xy(4bcV$eY49Z$?bN-s$SFT25M{tK?SYaV8Sp z-}d8OFudr+1S0D7EPVR44(+mPCv_p(4U7B6x(s36zG22ZxO)!_Yr6;4KlJPMfMG?q z!w&R%Ezm|KU+uKoR`p~iYqeuw@vwcX0_XLR;grs+eXq`aUK1t~-v2XOl(4MjXtvJ) zO9yuQS|3w}+lUQoy9w3*697x|8uUt&wc_NRz*~R(^PMffQ5{Z}iY+5H7988e>i-Ua zrYTR(q!VdnkLt1e!AjM?2mq$p5PBIGXnC6>T)*)>{l!ktZ&q(i4h!_$Vte_o`u~QcpvokR4bC|K8>zZ{T~56(3E=-407%oS+uPU2 z0B&@6*umSYzBKuQ2v2tA)!z$%qq)>{OnW`2`DO>Ou4a?HEWjSQq``mS2RWzt(-$<5 zH~#S}9m=MfPxcE?7Wmgd5;3Ql6G|fN&jSX(G}4+wVzE2AQWXd&bDA}wfI`HOkC_@! zM%D>|r5Te1)|8uK)jq&sSV$)}5IyHKOG42TN?H;1a9EKpA`9TIEVJR=Espq8y^UWIDeQF)7MS%H%TJ>bS_nrmS<2CB2$ z+DCR*oP`875RGe^XW?iJBrS`^jDZ9+P|Mad$3nGi_ajI9E7GZ0%VJ^)V4#+*XHOE3d@aZNKP zltkF8&4;zDm_&>-kpM>m3eB2OKq0bjPN*1A##r1tSeh|OU`-+F1#4JHU<1*!p;;1& zo&&X9OIi{=IRgo3;LJ8OCz8%=QigiZY{EnW95}NL&5I;6HU+2Gj13D3Y+x?i(A)^s zvepWdimYrdn`0w^j+ivfmvSbZfZm*6he;2aNPq*iY(sM>RLcVO^s*_jmd!JgfCp;X zhUQhMmK~K>@Agl{;cb4oNN2gBxfXU~pgJq7W#@aGg{0?5G^TkLj>bUJ&1q)cNAcw; zxCI-=b^{`F*@os=sFt0huK}{|pjzW%rU}46E!)z(3e~bbSvFfO*0PAyf?(Senaj2` zmqN8{Ct=U~inS~wmp}$;*_P%{sFnp%7S*zpUV<4&B3ha`p(LW-JHEVJy|{j>ea~IX zjx&(}2lC>UW=$xd5Lq`TRLqOVSll~UnlVXWO(DucigyzXYr6@Nx0+g-C86j!P)lho z%NbZe)c;1#EX|3eGnbJWgt}oMX*DpHUC&74pY(d?p@9cLl|4g{1X&6-d^A+l~xsF)XzvAB1zG-Hy$ znu5~{*07Mk20Fn@nkAv=IZ#V!Ez225Km%uXNpm9U%qC?h>rsdW!C(^-=5`|@1C1rk zizG8Pg{aqz4GRm3b}9n3?2_h2sFt+~NGh_jS~kZ<0v$1FnlI%{Isv`$%R4mbA#=MC zk>ATMX%2;IS)iU?Hf7~vR;m7|HoICCsAZQluR^u#s3O!J&|C`DvYmuI?<>}_j9da4sAUgm{)B2-AZ1Z4OX($;fh6L9 zW=<%HXm?S|jx&(}2Wr^^nl+(-LS)^XP%$qaV{z|bX~rahHHD}btYIO64b-v+G)qF! zbD);eT9z}AfJV+N&55Klo0OYk%{l;^Fp&TU&g>!0izG8Pg(ypWz2VZfvhX}***=>+t~FFQbTu&1n{E*%p=U8wM5-)oNfadq{IERLi!JS{4&a00Xt`A=M35C#sV!DD> z5p*iXS-MIUI=3`4lFn^XjEd?107$|_*|I8hZfSZXS+c1+y_RfPNNE3g#FA-_glbx= znxrx-t7&s=B-A0RrioL|suR>3@9ePZArlGlA4K#_vnf>70tNN*DzSRaGm?=1%K&oz zd2&?NjtX2<)(V^?>=9o!qgfaBWuVk6o6#;%v3JWtLK}(7Gn#SXs0`3r{wuB4ef&`Sp#4?(BI+Gb~iv_JrAJ(Tc*l3sbTF1^lbOA3;a6 zDpcEctM`1cSlcpkWrMEo`|#_KW>ToO1!5M}wv=AN8ObJQG>1aj#B#K@RfJWqM;Kix zlq1h*{)FNRk)x!x9b=(vSqv@Bog}oTFcm9;0b0XCLK}&q{Bz_eib7F!|0{-Cx=N_F zE&C$*K!&6&6{~=8CX#L=`C!>6$%(KV?Xv8_YvH8kts=GUoMuU=wzX)J0<9o@or)hp z&a4%)YMMFaqEUi+qh*Ix51B|-9jR^SG@C-TE!g{BUM1GHc}5cQNNqc(85XK-M+Gjb zZ3Rve_K+{5Sr?AVpja!bZI?gH+LnK|97knSquRD7bNNvl% zT8?U4u=l;ZN~~@3j4VQG;gQ;wf4&^mwxa^yL-u$%Y}Ry|C%|D%Ai;)t8EW{ zn6+&~J61R<1GHsPnK4i-q>9wG{QKpowmpc{wwPE#7^!XfhssfH+Y@r5MX|O;q!PMF zZOgw!j%wRZ#GVfpYgpzMp~13_u|$%Iwkw-{Y1X2~=|k}SEfFm;>tcAe1NwxvBORNLBJPYSVn zsBJTA#jKjn@^aBALA^P{j%YMwBAw$%ZQIfw6{>B)-uLncnsJ0yy zxM*%GaFVcxauC|N!oCa?Yh|_V;SaO6<)1ajQ5m2ui^`0Fq}NDo%fDxiYTLs|ZHtK| zgpt~of4Cggwml&?S`=$rL@J?+)VBQFilwV_2TK-v(@?D>G6B3^P}1RXmR_E>O)6QPmeDyCXcFv$>Q|v z>7%#Gxk#9QH-edFP4)H;?m^7t`H$3hiFU_Acyrv3u<25`ayuTgMFU^|1 zBLLBbk3zcfqI&TmrWy3cRIMZz38zx-q7nmupJ&#Bv<4G#(L z?+5I8Nq?PJ=j=JD)^5|`U+Ap*%hjXFfn?Q)kAyp7+x*k!S=&zJ-FR7#Z4bFfm?IS* z|H?V4_>y>hD{U{)5hDG@n`U62$CjOV zo?EScxijW@wLf`%#F!Dagz;}jFw(5+vjWDACLeVO_p8?@UsMt5S#+c#T+&PnRfKDQ ziv96dI*d&O$JpaU|Tl4bl>~uZ$O%CA~y4J9Hk`)m78sL1N zPRefIym|g)_3YWni+3*ezJGT9>}a+9-^b66pC3IR`Ox3?Q9^2Wn`teEj()dNLwDsz zcXjW5$N4_()ZD%Sw^hL9*}JP3N0;xtTmcPo-zzeZ^caXT^L;w!DnWWR^1*Kjsi@Py zyJGWwnm63OE4CH%<=Mr_<;mF#pi$OI=I@Y6Y^|OnU+J9h)1Hiu@@g7FBkJ;RMq&W% z#C+yvzO&u*H0|5Fb*%?oba=h`;$*swti}stN~O|~nS9xt!sJa)iI=3^5Pq8RIPqUicZXlpf1o2epV6EMbwr8Iu2b(j?TZ~zfs>^B z5O$g+xv*cUpKQot|3T=T=1mgz?Z7dr_aUeI*E`LoT-g13?_(GB9+-B``4_^`v}<6g zcf;E|t&7vHEFVedA>1_oa^Sw2?gqE0?@)ZBxtSE-?CAMU=ic~+cu9JX;OE~EcjCWd z<$*iy5BDXtiz*W1_?N{cU$-5^-)Y|)--nze-G{K#+|GsF@8|p2Q~ff7s{ZUN?i(-W z{EOpEIp!zXL1-Un{ukJXtLbj{L9?yuFGTEt=7V1@wTGk)_bZ)2{CxHG$=6&4kua1j zBC-?w+vky;xSH;GCq$cwoX@A?JD)wY))z>+4PmAE zToKl*Fji5ofj9N%{0r;oP5n+QIMWv2)X#8{^c%sioUi zSJ&@ive=iAcu881^kw*$)}8p}xx0X0Gow&no_~KG)m%ho8*#;&OW-8wK2me>FSEO_ z%QaUP`wv3zG)I%L_b0~B)prgC*oU0%U+*+$b7A-Ey^megd!**%UwB6~C)Acb?P5)x z*o{d94p zE^?F3erOM9-_5-T^4wk81JMpb{*UH)Iscc~fsOf&`i}CFY#@Z6=6X*2JN5r$UsAgW z&D~o%XGQa7V70h=C(fH?ev%zT=I;Ef^MQSk=kDS@qhSt-@C9Ea~EgO87@|j%v>U0teqEr5&!N}S_m_**mkiOss1Ck z{LT3n@zDetXei0Z#kqTqjil+w+QtxHz9W5!C4C1D^(E$GA7n0^ zyGy(zt%vZ_9L$Mdp1TY9Mg2$WjwQ{-P~AafwlQ0*xdcv^(>_=(`q4(+z==V&d_=aZX>~{^~JZ4}EXbo!*G5j>EbK>W^-yQy*{v$E&faZ6S zuj@~Y{eFJP$?BNqwjt~^w{v0l`}schRKFB^WaiAj;*Vy|_Dm+Tz_g#vPQEyg7gw=% z5&FIR0nP8+dmzuVXL!l5auldE_{f|IQ^Ξ*ZeL9G>h)Zw6#R}_wFMc z6#Ix_rTx4jtn%EQVomi&GnK&Hz2^5cpt*Y|DbC!*S#ySqq~8!`nw=G6mVfU)idoci zU; zKh43M_%)rWKN=J^DfF&$&F_dnbq5j0#%yu!E^v}`AE-BLeuIPyyQukHZBR__{C)`u zdw*gq&tnF9A9A{Xz0;h{h5erO-ru0BNX=REdnHiK3AJTUyI5;y`A9ks^cdRAju4zl{ClmA7I@hbn%(o)f>z|C7B;^-Ozq zfw_Cl@0&pLW}q|Ny%Xo_GC#=M)(=<$m^+5Oy6As)wBizk?q;?RRL-YG2&>XrRo4bqi z=20G!o&)b0)HTih3h~OncOSu<>XGJpfp4$Z{N4yOi|)y4PqO%(`4JA1ZbMjUK39ZQ zp1V`5qFy62XnsEgnn8Dx;>=x~yJxsa`VC>G*;z4WO+)I(=tno^D)M{xn%@(FCeT2T zBqJB+?m0G+rXzFrnr2|-s6|!pWNSj-nXmcX5vVWG>~QWb@shM2>C5nYB%Ju=zKnoh zGow)5!S9qny=fw|joFIE)ZTP{xfphum$|UZHCGna4?^$scd!!n{xDdc#|*F!IVlzt z(>u-CT-bT#CAJo_dy?8isOIGNN}!q(N@h-0RbZoFv^x z`pNum2^V&MeD7n|Y$!B$=l4sXnX@&+WJZ|w)5Y1c$W79JWbWS5{LZ}xJl5}WZ&LdR z&E5Gu6FL8v*ny4t4*ySiNj4CPe=W`RocOr~>~3Fr`j5<;`F#^;-V9DI?%pYSQ!hVP z7mfVhy>4m#7uW}R?k+-1^-lAl$lSf9`5>A$kIJ2N`NZG5>kJhO3(ehIni)oRLY}*; zJ0aRcC{LpESDYt-xoMgw$T?z`lVk-E?EL--7xsc2F@s&ydt~m;@0`FjF_75i1>s&m zhQGCiP%mIX^Fjyif?mKVK2hJ1IW)gd0?nbDvAMg|1&dSa@7>213cY8*?~o|O%X2>Q z{euyRsoq1s0q6HdpjmWJCVP^_x%&tQ#Xe$KX+BqkRi3+3tf~IkS%`ddz2^5rpc!;0 zX&`eKhx6l0#W2(CtQfQWd-qW|qMjoYXns!wnm_|-lZ;#(F6Y=tCpm(e-xyJjT2qqx zG5YbXiOt>l-4UoS(e7~WF7cAI9_h>QdnBCr<+;0nUo)f743gg|f$9z-vyIte%_VS> zbRVg?_{|b7>~hVO#r}iPJN?~*gnc+I_v(Gf>HhUjb2b-tzux=UMZHIAZGNu=syPRi zdM&X%?c#hr%SY0A2sh2+9Ju9LJA+%)cPPHmo|+Wj(5zgXFAT6FUXtD;`1zd@PW=A( zKEyBTKN92ky%LhI>rag3_p6BZt?!U@AHq)iYcA}5Ki}U)QSXtNGrwN~&77^ak{M~* zPZwv)A~#9@p*^58SN9&s-DQM8B1+@0S$fu_x)awlCraqg}& zlsrRZC-@B%k)4p|?&?m6HWA5__+1n@PXhD9G*6Iw0a;Fx6@;)C=7<@I`Gjs)dy?8i zX#ULaoWM0Pz;5#bMOW%Q=PmjifLSbto92ZM+y%XW(fv#HO6PZxxjVm40?nb@vAMfA zZyx2L*h^^c&hL;Y#4G>aeFU$l-^g#k`MnWn7TuH894_@9_Yn?~ZbMjUKd%U@Ja?y9 zMZHFT`_1o%Kr`r0Qk=PqBlQdyNxuQi4ZkCz7_+7!_4n?h8x#7yd&BREKoe*nNRp9@ zbN3t@Nz;M3d$Uh7uyWL*s&}$Ap}Bj*?~XuyiNy}*?h-Fa>w&&Zvrls{Cw`uJi~TBL zUsCgm%-tJ)rv$3Gh-@}yi}i)TNz#3w=4$q7PtJv1?#pMfr~0KicnCZF-GhX^KQWf) zF$3&FPWP{OnzOmE`}N+(F6uone{T4_5~${c+OnrzoV#cFNIH+;=66asaLcuJ2Dhm1 zNPOe>Nl5W+I4kCNyL#grTEp5yDQM8A^w7KE;PM~S?sN6}HPn^5!3>6ED{NBCc zH&8@&LY}*;J0aRcBuC_TQQ$lY%n#E%LGA%%IZ0L!!cO~tF6^4N^c*p>L7_Y68h+;l zu8Dy#HkJo}@4m>4+NahQB6IhK-!S38UC;{{-9OE8B6H}5-zS0Q(2E_+-A8#?ebV|b z@SZ`#?~o|OE6?3W@QV5k%%U59Zv>h}_hhvvS<#F-oaf6#e(&D!yCRCP%5!%LL)2?z z2F>q>Kr`r0Qk=PqbN37v#X=&#cW?L|5yhD0-@A`u7WEvNK=XSd&;%Mtn`GpQiqv0z zXV!{*UB9Vm23C$*zTbTukf`fOUxMEqf%+259nRe)UW)OA`cgHWy*lyBeHj72sQ*aa z!S9qnbqA5z#%!_X5;#e^4`HW0ITv<${+h-9gU~zu9jt`CKQWf)F$3&FPWP^Nez$}R zJAc#0dlve8lG;P4=H&NEpqdkE%bs>|?w;i%={z!bZ|MBhftzc5H~Z1kcOYVuf@qLIt)idoqL}DDjS3>f2!)dwKrwuttx({Kext$9; zKlQuXlhhs}-*9M?)H>`M);2=*9lvYBiQnb_$zG;y`Whq9a`zYxM+m?LH+F6aSeIZ4Y|Wg=W$G-Uu{{ z?ulVfvZ5Jv?#`EqVWs(85mtHbPGM-a5t>2s`ytQ_x|0-V?yPzf=kXct)^cK)Y2U6G zGq2fqfjQNGXaddei9i!*pot_S7k|{9V`DW+E3?Sk1`C>jm7`9T!NgVHHW=TU*zeu> z-4UoSaiGlI6-}x4pDGK6`Z5cegE{ewsu!D70{+nbuMNLb0@YkZ92>L6noHoM*iq=W z{`_VM7j{weyV{_b-ue9!688RNSf0BN^giTt|9Yo6n+v;KYiF@*78IJh^Lr&w%?Ty8 zr(LYIvwS3-M|uqWP6-EYxz^6$7WEyOyDW>oGdz%YI4gdh{O*9JHs1IK77WGrCC%!j z_&yYeXYsp(oZ3Y!#?kq!awlCraip#@lq@2&6Eri7?1VgbS9e0RiAbKr@1nqY5}2E&d4ik|WjRS! z5W-%VBW5Hn=mlgsN!TMbF~4&H*Tg{LZ7k#8L6ds~8UA)ZLOl?E!-NBOK`&qwpLotA zb7+2_1e!x1>|pLb%ERiD=6)f(H1{jSEC1eo1h1&y$nV|xy%A^@-ILXxWbuCY5e|}W zLs)4(SAl8juOyXV+QnvTrf4`~Kgj#~b``#9=Ur!+Q2`ttnl2-KH2ROariJ16#Q zq&2KHMEX)KzemD}U+&8YI7R&j>W-G*DS>J(BD0OzV$CIRvO1==AHq)aG8cBa=E`FK zLFk?S4pzcGoR)j_KIC-&dZ#&?3%g(Mee9y%1GRR`@0CC`=fG00CAOzstf{knB%OzF z(=M9>w_IyyaEtm5#P@ce&RwPWHk>=lzjq&CN4zAxhw#&^&WYb2--q}`{fA;4oxe)H zZa6LX`uQOzN%s-#{B8*scE6wRV;A)vn7g<9ehD;lwr4e&k*580alS5clk^|j1DfBt z_duSzOM4*NLB#*@dnR)JFR=p~^Br{@dG0RmgJ=(dX>&WH`5>A$kIJ2N`B>9OoX6`6Sxbrh-o2gC z%rLSOnhK{o-j>ii_$|MS0_RC!ewgM7a*mkgWVL-SJs^JngbRB?j+nu&8B%2a-10jo za7_#(wt0cv3&`-1bROvi@Eax^xC?p#qqs$V2j! z|Mj+cqUCo-96VW_U+%45JUx51I^R1zes6VtG}~7O?|-uT(9zS=i!QR=brZ98tL5khw#7H zo%%PcHztSrsqZtA^d3S^du|@&Jh9u+j-<9wh2lM(!)noPI6Qvtk9LTer2EJY@LMLl zI}nKX!yQOBMh^J=AuCD$A@nrg^Pvy;{62b7_kZpcf40eQokp$@7n}lWZWg3pD5R?}Dbh=}x!mDijO&9TZtlm?+Pt>z>Mo zGL!5e6boqn=iLEcEFe3enN(={$nT>-)5oO#zZDv{_r$@X&QP+2$V8goN)g!!`S;K2 zPKY+~7b7u(=7?{KF=A^INh|PMU6}ZV>Kl{CC1HYal&m7M7yO=z&|X|kcfA*)T|_b` zerE;AoWP(l%^UPiNtTsl2_f{Qd1FT8lDsj?Npg+X#pV-H7AW37=lm7M2^&eKtNju{DN0?Zh(#kl5nC63} zh(}i}TtX8%f6z)ckgda@6u1cO12->&Uxg{Ei9qE*TOx$<)Pn=(2ny zZHI8vtjvL1euplDThw=?bHVSPfSrrkF6Z_VGfDHI9iVxccL(&|jj#iv4MZvuelG>A z$B5820IT(wz)I482tCc!eCRcW$>Z(p7F|^#^fYhN(D%p9`V40PY{=^Vbx-p+A9}qO z&!X4NC{&U1J1SsB3PrZ3Uad&8j3m8BYH@x;g$KD_i)WCFI*&v9n%`{!XVs&EC*45i9Ml=MzEWZpp}nBlVrVb)xxI>6w2MgQ z#P7TynG+bTrqevVPn2b4HBP4>A@rqrV@6~?+1uH!{D17dZ;U0`bsx4pJG(Pm%Ux1W zA~kNBq%Df%vR6`fzp8rOJtQUaEhR@bMKU5qBVXL;oqn8cY)|*3yZ0^EpCrd{ELld3 zfO4x?){xVb#5(X z14}{V?#%twIrrRi?)lZZ=hYqhL{_bUzTbj)V+OMiQ@FlXAUsI5?m4xBLy{|6W=^mA zL>7tGJm(afzP|#VV(+_izth#l^=fw*>6&HQ(H?1Yl8USull8!{Gp5&hf=NbeoO6;* z-&FxmvLnlf5v@JRW*(t&lB_NT`i6>ev(=y5ld~PJA(^pqPO#~FDc}iquPSEr+H-s1 zF{(Arnl;JT#?M-RZqLqIxz;)7*Yq6~@cbI6J7DVCb9?DIs@2Y#H_6Hj^VXl+^Yd1& z_itwOThVt{e4_2QdU|DiC^1}g@mTF0#UkSt4>&w6M4e3a|+ zcXQTEcIcKSHcdh)}If)?7$d#Li{CGj=Y@b@^{{)`k2{!iT#e z-0|D1v&V1TT)RTnEB$(Wy!ex|?RGuzNYy%I=)R!uc7VFCceJ`I?fzfc?Y^MzWteTO z8M^eRdZT}|_!F}q4viiTQ&~JqYm!T(TCaChS}Se+@Bhs^#izxOv0UGrJbH6>_v_7n zZX^3aSt+CH*+ z!k!(vZ->t9(9#aA?9kc{UD%-qcIcrUYWJqO-J9liZ<^b^X>RwXx!s%Qc5j;7y=iXu zrn%jl=5}u??cP+{y{WW&Q)&06((X;A-J43gH+P$f=dsAiirpoS3mED^v zyEj#KZ>sFxRN1|$vU^iy_omwJO|{*dYP&bpc5kZf-c;MYskVDlZTF_y?oGAbn-+F& zTG+j5VfUtm-J2G6Z(7*BX<_%Kh25JLc5hnPz3IU2O$T;wI_of58 zHyzl$>A>zy2X=2dw0qN`-J1^W-gId9rbD|o9ooI=(C$r#c5gbgd()xan~v<>bY%CY zBfB>p*}dt=?oCH_Z#uGj(~;eqj_lrabhK^#-ueFCUIunQ13S;amKoS816yZc7a7=t z4D4YBHp8#>GyG~l!>{%;{AxeLul6(iYCprT_A~ryKf|x~GyG~l!>^hGxp}W^D&%%- zhF>)`a=Y9NziO)FcDWgT)zr!Dax?s@sg#@LHl=bqHp8!)V!2&zhF>-1a=Y9NziJBR zcDWgT)s)Qbax?s@DVm$*HdS*wHp8!)y189$hF>+6bGzIOziMjdcDWgT)l|>Ta+~tG z9h>1-O#$65H^Z-*61rV(hF>*Bbi3RPziP_pcDWgT)fCdra+^xJ9h>1-O)cFnH^Z-* zYPwx+hF>-Hbi3RPziKM#X1PsC-Hy%htEQ-Kmz&{NO+Mb-UaQ zziNu>X1Psu-Hy%htERqgmz&{NO@-YqH^Z-*8oOO?hF>*RcC*~3%x=eK_*GMAx694& ztESX$mz&{NO|jiBH^Z-*a=Trw-LK{u_4fIm-LK|*cE6hM+5Kw1XZNf5p53qJdv?E? z@7euozGwHV`Cf)!Wz^f}8TIygM!kKWQE#7T)Z6D7_4av2y?vfhZ=Yw>+vgee_IXCV zeV$QopEoV(*8O>&QE#6&P3d;I8T(b!mTs4uv0pWf>2|pp`&CB0ectq@n|02px6hmI zbi3RPziRr^?Q%2xs_9U-%gyksrbpc@H>2J@T0n>Ka3&KZ8yG^*R>X82Xps&1E? z;a5$wx?OIDUuD$W=S{!5S?7#;`@HE`x694&tEOk&E;qxknyz)b+zh{J`qs^IGwSX0 zjC%XLX-9>vp*re$_Ov+vR5XRYtvi-t@4Wb14Od&G4(Hm)$HkquxHxsJG9Xc6PhY8Gh9?wAg{Dly}it+x0f09_A;a1US`zW%Zz$^nNe>qGwSVS zM!mhvsJE9H_4YEO-d<+Z+sll4dzn#hFEi@xWk!E{nNe>qGy2=hjCy;S(cfNX)Z5F9 zdV85sZ!a_I?PW%Pdzn#hFEjev%Zz$^nbF@~X4KoujCy;SQEx9Z>g{Dle|woxZ!a_Y z+sll4dzsPSUS`zW%Z&c^GNayJX4KoujCy;SQEx9Z`rFHldV87C-(F_a+slmp_A;a1 zUS`zW%Zz$^nNe>qGy2=hjCy;S(cfNX)Z5F9{`NAX-d<+(x0f09_A;a1US`zW%Zz$^ znbF@~X4KoujQ;jAquyR-^tYE8_4YEO-d<+Z+sll4dzsPSUS`zW%Z&c^GNayJX7sn0 z8TIxuqrbh(sJE9H_4YEO-d<+Z+slmp_A;a1US{;Sml^f;GNZq}%&50l8TIxmquyR+ z)Z43!dV7^oZ?7`y?Nvs-y~?P!R~hy8Dx=g`oVy}in)w^teU z_9~;^US-tVtBiVkl~Hf6GV1MBM!mhtsJB-c_4X>G-d<(Y+pCOvdzDdduQKZGRYtwN z%BZ(j8TIxmquyR+)Z43!dV7^oZ?7`y?Nvs-y~?P!R~hy8Dx=g`oVy}in)w^teU_9~;^US-tVtBiVkl~Hf6GV1MBM!mhtsJB-c_4X>G-d<(Y+pCOv zdzDdduQKZGRYtwN%BZ(j8TIxmquyR+)Z43!dV7^oZ?7`y?Nvs-y~?P!R~hy8Dx=g`oVy}in)w^teU_9~;^US-tVtBiVkl~Hf6GV1MBM!mhtsJB-c z_4X>G-d<(Y+pCOvdzDdduQKZGRYtwN%BZ(j8TIxmquyR+)Z43!dV8HwZ?7}z?R7@I zz0Rn&*BSNpI-}lRXVly4jCy;WQE#s^>g{z#y}iz;x7Qi<_Bx~9UT4(X>x_DPol$SE zGwSViM!mhxsJGV{_4YcW-d<g{z#y}iz;x7Qi<_Bx~9 zUT4(X>x_DPol$SEGwSViM!mhxsJGV{_4YcW-d<g{z# zy}iz;x7Qi<_Bx~9UT4(X>x_DPol$SEGwSViM!mhxsJGV{_4YcW-d<#MV8AAI-j&PR)v zk1yZ4dHUAP@%0-gPgmEo7m9lipPM~<|GRgKzrIzxdbN6bd3AIA*2USy+0EI>`PpAw zogS}GuFg(gKVMxR@9%!)^6}ZDlk=}$-pmUB&)c6Cf!8-DkKT;cW%c=6zqeJ$&wcIk z;`(OxT>C%qU@$$@yEW*{{F5wfwbrx9;DYb+cW*`1t(t^^^1C>zmcn}zmZ{B&j3Wt3%9`^L|y0PQsML6i= z@u07qytBIc>gCmw-Pcz)C#f?2?;oE$d2+Jc>mG3N=<;-RwR?VxKHl3sxw%!1Z>RDK zRBABgjVpax9@#!|hZ`4}SvHydRElr-HnHe|Iot#`~j@?Gusr_v0dy zAnylbY44wp_I?7UqKOEmq1$qBm?fucn_KC>*`*D#;koSYJwD*^zy`Mm|sw3}D#tYsL$G1;M zewK_Dy`Mm|5|H%c zA0VQoM&8e&XzyQ)^nN(LoeJ{)h1&b0k?j+a_xIx>lOXR0V`=X{814N8qE#Jve==V1 zemK5;I`aNxwCMc=qLqNWKOC?1eiGA4LEaA#(NZJtXHm5GAB^;VIKG_<^8N$0_eUe! zCnE3f$3-SV-Veso-hVjS`w2v=I`aNxyx{$CeEW3d{mE$2`w2uV0eOEoUhDlNrj>%c zA0VQoM&8e&XzxE9>HTngI~C;phidPSMz&8x-rtXlOoF^0jHSK*Xteheh*ovv{mFR2 z`{DTZ>B#$&(W3Vgh*kpf{&2k3`$Bqj;rMnc$or4f-XD!@ zpNPD_9~YSfc|RE28DbwCjqUS?z0E1|?P2!dc$k8;%Ig3@0Q|U0!Y!bEzHSbK5YVaA zL0TE5q!J1OQLzE2m4p0wVB8HPTJwccJrC=apRfac@I@>Q0JPKjEY@oqL#8WIV7G$h02Xdoc%9AGqr0JOj`8c0wp z2Sx*cc(-m_AYe4GFeVz{vF^raVKhKMI~^De{ZZ_oTLeZ!G_-vtjD~(-WEdC?U~p%S z2W>3KtqD?ZHo7=XXECWcnZsCsaq^Z^| z1u8A+>64SETDKUsULFa439^loXokvG9Y_tmQ?e=nTrWSTo2QT7+id4%ngL!fKR;IT zV1LO$Q@#WkWtVi_e6cSHFe)kmEL>3>l`LqzL`1J=o?$w5Ga3n4FA@ESB!sgVHGI89 ze64>v_QBDR^%4~qfRus1jmO|<4k=;6?A@xMXjTBRUlYJ^m5^VHFr~CONj-rHIFS~^ za1=pUI5QMQ0!X=TX^MIR$xI?G1u88WCn=>RV;IhDp_+qaG28)BZc{;Js}7`wYbZqn z0mQitD2w+Ml-rbGoZDFNV6Q>BO^c>{2_%l;%HlmhfKgEiVBw15sANGnxApIl+T6wf zaBhnL;Ve*YBZuSM)(^)%h;kbX!MP1U2L6U}8;`-!9Jc+A%54fD_GxmC(>dVjv@#PXNICk04dikO;I?vCDKx$(voqKQd%;G;oKIgIY<`69T4R<6;!tB zKx(*#D7TR^JrL(M0L1$W%56$8&TTAsu-Bm6rbSb}1QN$^W$_*$z^JGMFuyQQUF`&(D3;^f02oTN!pRmJ8~i|hT$lJuyAH5w+SHSx}_-!=e9&z3RGG$PEtxs#xR`Q zLNy0jX1D{Q+@^xcRvkzU*AV44Qlp`!2In^NXGmOb>wgcV&20<-=e7tC&I08&ayZUy{c!AqD7Uc?oZA3o;BP3m z@faM<;ZJr(O6KOFFM-haDGefye04dikO;I?vCDKx$ z(voqKQd%;G;oKIgImj}@9T4R<6;!tBKx(*#D7TR^JrL(M0L1$W%56$8&TTAsu-Bm6 zrbSb}1hUO=W$_*$z^JGMFuyQ3w~;?GvnV*og}MSyS?D7TTrac=8}V;@Ah zjfLRc1|S1}L%EH|;AjqikT)u~DS+6o31GNN$gf41Qd*p(@Q$2Fi(xp5AS|32%54Hj zxo&BS!nrMxmI9TQjFXhok}(YDwouJMmKpAVD7UGgvQ-CC!!<;?jg;wuIJW^H-d9j= zQ-X1BW5I*H2IV#_n(`%(ZH6n0_W%J#MJ0gwg~7Rv{0T9a+xp+~YI7R{z_~2~gtI`o zjU0}1TR$B8Aj)kl1m`vY8TcE@Z9E1?bNF-HQMpY4#C}Zx!&O3lEy9%2;v|K4m(qb5nA_xm-hH{$#Qm$K?qHu0Yq@_TmCF3Nev}6p!xh+(4kY$EDAj)kj zsBG1N)Nl<^ZX;!SAkJ+7i1!tg+mv9O+gR{muR*y@i>7=DWSim2;yplsQBetCeqnHK zBR~Ja<+lDi9NOH*0B~-L0O2f9ZX<`|+}01rK8SJ~3&FV!KnDJXavP7q(Hwp%V^nTa z0I^>az;Kn2UyCrMv^YuO9XXK}!*CQqSU5A3+XRqu-O?0=b6X-U1u88WCn=>RV;IhD zp_+p%Gu#1DZc{;Js}7`wYlw0iDboXSZUaEPub|wf1moPsf(LsI%57RSbpaBd6L9Auf{4v2D_3MyN5AT?Y=l-o#|9*A=r0OEZG=U{q8Bm|qy2+sIFMak;JkhL$$BF#w#~B0x9`l-tPRIJfn~u@9o$ z#zJsz1CW8gq1?t}*2_cib%Mc1vB>0RMXs0d8$!wFJGxu2!iBQowDRIi^+c&>j9V|! z=k`*Y-r~a9Sy0;bOLNr|<;tf>tv%y>rL|{_TQ46R0$~U`lT6+$bE^t6sfgmoWr#CjB!Nn z3pHi-5SaT^ScPo|a>KQRxsR0UkwoqTNW9m;+^0knxsOE;_8*x0v~bd^Kvo%M31U|e zaAcGMoL`?OnAeh@9dKiVo@%U14qf0TM#-*5HignG=X?O@9eRWJU@v&H8!tT+3>H@cPn z#_d9HTfBBV-?g@1s-CHfUbKH4*60^ik z&@c2yl0Eirap&PPJ4JE#n`fup`hT!^;my_R>GA2=lk3@?SDq>Uy{)1iXN)!(kULs_ zI>_8{kN4l(gz&evik}{5utBzR>X8QIVjAzck$X?2Ae9ufd7a7-#UDIdd?Cw%QChJ1 zZ`OwFR&je8SD5B8NqW$3_FFsV&3@-Wvgtfu+iw&`-#7)S6~>$12n*)!H}fg>n|a=TvoTWMM&T^IKh0y3{U!`_AO6~1 z=o#}av_(}8`@d;nKe0=uks8 zR{Jw2H_bA9tN9G?ZQ_|YBmCB#Vj+6*AP$^5v-|F!wnx6jxovDD%q``}GLF--HDA1> zP5tYeu4Hv*?}eS0ihJLDdUADgTVZKx@tPIAEz*6n;{?sNcw3A+xq0-)rb_eRnc_PI z-mb4!*OzZyHMO#4&uh19VrheS_Ios0+W7&uHQrx7fB)XBd5bH)R=m{Z&qhMvJ3D6j z@ZWD0_uUf%N?>##B8??>*&fcq#oN8xf4JpecXZ1Xl+ucqi|5ZS&Tfw9v%6ot{Mysv zhl>xk|9*0QetdKC`gv2p?Eb^`G)p>X_agTiTMe)+$ z(X)?y{ru$7>W$0u)790VKf8J3%d2mnymfxFjWoCwv7rb}@p9m~)KpeQ(aP&WfD$&; zN%iD`F2!HoDju|QZcimH9{<_Pvy0AXx33A^p1SjS7eeK7zX_pdmaiV)te!kQKAe}| zuJ-5EUiIyhYE>T{9+i*g)!}JX{k6>k+EjI@{)UZd^66x*b_&OPo> zYq--?imz9<+t|(JRl5`Xn;WqgZ#_9~*5~TfOLJZYH;nTb=%l^ThV? z#8zhBc$%d7LrB`}gWnj6-;T|@&1kt9?2Qs<2Wswr&vU-+fa0t zji2=!?`}Ew`t^6WmcRDyR^wtFU0!~?>tKBH`s(zwan9@A;nDoWcSL@ImewB7PkxEy zCu4z6FR#zq)1p;BClfZRZt?rix@D2}hw6h{=HxfX&K?{O5)=G(;q4xXgOC{QbTfdcl?9U6$R)=+-SEl^ zu>Laf$Z9w(qhI0mzrh%$aHrL1$a;w`jR#V?#EW`FsliUC@tFQLb8My{wW6Ow_WJm7 zPL)x^*UMKr`pRjsaiWpe^*$MUAuwikY`^yH&syLu(jv-P3o$z-7&AK-Ofx$f<}^mk zPK(C7EluKtpKi=u5THhF^ zv|z~7bQTN|PD7OiQy>+&fZCAEPWX)t{t$xzqE;5@#YQeU-q{VWtN_Ez4*aO4P0NU; z5o4WJqam2t0Vy=IqXs*j5VPYknA!E1AhjZ@2F5v6Mh(ZzE_yn{?BGbJ38I1wfM81D z+c|zafLup}IW2S0#Lxd&*`1TYUmeOgr*sw^e+Z@d&GpGmQ^db5N)LbD-fnL2x1lV= z>1st1zo{c-Zrl|9bXEX4Rqa^(>6(E}hcplVZaP$RFx)9NEi5gl_446b%H$AN#Jr`5j(F4ZWF~(`_7*a8% z9YcsyQfJ2$Nyai@KJDH}L?toYsb@ivOAKW3u}lJ-dPFP>ID%{Sm8fB{ZV7o4G1w_K z9zzg$Ag4>t@NJqJ?bI@W*2~X;x|;v^Nk9VO&)$2L{6MG8xbf@dXBCQU918x%KT>75 zp^ZflOfY<#AHf7gn6nvVf=V>O1X(o01ZA948ZtpGoY;^7+i81B*6qp zihxPTnH(?)#3?BT41v_z(GxVK9X-IQs9We5r?q29&XjfxAx=r19aAJ3 z%YX&ZOi(E2S-;K$%PUY8AIo4a!}Luov4A6(U|#^zFhOFlTUN*fSq#Ah0fu3M)M%#` zWP&`1V1hj;NWYNTfq_n&apMUl7(IXEP>|Qbp-!2x2z=Vx{0t+|w=h+xMt4y@><40*1!{i7_Xc#=JB=j35TPZ>k zQQg#(7W6z!XF)%Hlm%m$(t;rY(^)V?I1N=6Oo3G70-8gf_RguUKc^dxTtWoDvH-+J zE;-)mh9g&i;nUu^J3xxkGCH~?#yYJ=Lon|IQo6*8I3zXL=`%VQ!I;^x zV4B&reVRizDT3MhM8@c3oXE(gE0t_>| zY6P<*#yYJKvtuEc*#QX6?5M#`C&cV{3}$vcCP=M_s)2D%l~Kbnvx}a-aUw+2;7F&* zNC0MbwTan@D5oQ0c1kd2b}X1?b~4OqjF_Dkjdxvu4%`oIo-D#CY6N0tM@rl{C!%Kp z#_4H>;`|Uwm6;tm+G&TFodCl50fLzPz>Iggq5L4il-ha91$jv##*ZJRT?~^WB%xc7 zmk|1O_I4Ud5t4}NfJzH`%B8cQAMf;2t#1rdS}>$rItzvfr=iM%DUga>Ky657S6i5! z0HRhF=*31ZIo{a~udD#W%&s26?1-^WE5z(r2xfKwLNhyRu+s@KJ063XU5^P;E23&( zoKt1gaLnwYr*E7H5j8l{X)+Rkncc#~>_n8)5ivU@7&AK-Ofx$f<}^mkPK(C7E;Qyj zcGO^}6JmBe1~a=J6Qou|)xbEX%BbO(*+oy^I1wUhaHP{@BmgtJ0~50oQBFt1?37^4 z>{u|(>|~hJ7%@978t=LQ9k?IbJXwTO)Ck1Pj+D4@PDIZHjMLK$#rYwWDl+&fZCAE?!dzA1Q4~dKrc3O$??u^cx43` zW_AZ7m>n_JX@!^_3&G3|Kxk%14R$&qX2)YNv+FTIYDH8HjB~1t8jhJ=^z@ArA)*FH zI!#6bFta-}F*_0EbVSTf3C7Hh1=GwXvcX+cl9bQbjEoqnqIjbTa)hIC73!4TmzR9P?uQjrU&4aw{d zEzC{;Q7a4dVk4Iv@9c(GR)AqYH*~}WF!DHyCV~`6H!h_#O#z{%+J0`lp-V%)d7_j^ps0yK|kK;w?CvOq63a>?<|Zg^z{7-n`yBbXgA)@g;99SgzC z4nSyTM-6s5A!f&8Fth71L25-*4UBWDj2e!aUG(&g6Ct7oM>#f2dU<)? z+!YvD;A8mDogDVB*UJYfdNZT$;@yF~;nph*STA3Ee17@*$@%g1&Fbm$#qPt)$L(e? z3;x9^n*P~(`SNDm+pDw3Z`@ojKRo<*2ZevU9$qT`(d0Jx7m9liyE~+d|FS6lWOqe& zcO`Xu4g2xQ#p&_&`Pn1#TI#>|g1eRa@4T@5+b{S#sh?TC|9|gW;C&0cZ-IZUEHGln z{2PC;RlFE%nX~7=(QcjFKh*u<8z)z%tJ9x3J6~O#JXvi&2!`%9cLiNvULD{5#q~D3 z?{0s*yItVP>#Nh#vx~>qyTk9oZGVdY`%Qw-4O?#~cg=fgE&JI{?vRfLIC;8p>+Mg* z>OZ}_KD)g~DY)Q%dq*+I3D%EYr$)KCd~AuUqNF`0>tis!7E#IYQb=b+C=~Bu7sWPD3M5 zX+dAN(pk`tclxO;7{inn3~N^69@APdL^usq7EFOuxx{Fv8;)ESM6E0U zv5`xTce>$~6<}*ecB1_br)5;@>aUaX%4_ryGu37DTNq0I`uv zj(57@l@(x^+4WaG4cXc@qG`lfr`2c(W_Cad&FrYbPAA0dcnoHCJtj!4h^m2cPL)x^ zF|&)FzHuT%)Zj>`$w&afl)_pArgo6)h%l#R4tl)}FTP`CcTNWHWR-DF=`8pfM;(wJ zcp z7UFb;8?pjPFhLTzaa73W1dvnJjwP~3D7R-%boL;JI|X6(5Ku(+fG93|FaW0>%pM|6 ztDq-nn8751JYbxHF-{RM39Z8nhEQ?1af$(x2bI|xC2FMP45FMzn2&bqKh);V-i5vGsN45gS>1l@I4JMR2 zB!6hHAV)jx3?O`F1SM^(hYo+3@lH1rOsSn`WALmp(aw(_rCkh@BP5|=@T`*1ud}z) zP>PU5R0mXA(DN{z1^sxZpK5($n9_nF0n=G9L^usq7EFOuGH#nH?!{54x%y z2vch3DHr4=i5NeAly)&pj*x_IL0&@W*V)@?C`Cvjssk!5=qZ=Zf_}WyPqn@=OliT8 zZs{x-BAkXQ3#LFSasjm=nO!M-zy$WH2q0=@fnIFnlH;A-@X87>%{(xss_e6RYncR%r1KR#)%M7gCm_LBLSG%g|}qu znVpDoIwEGL1Y>5$f@x+a!<@#5*=f;u*9GXn{m|ygBAlW|AZB)?#Eo+zdM03;o@OY{ z51~|<*^#52c8J*tAe%{(xss_e6RYncR%r1KR#)%M7gCm_LBLSG%g|}qunVpDoIwEGL z1Y>5$f@x+a!<@#5*=f;u*9GXn{m|ygBAlW|AZB)?#Eo+zdM03;o@OY{51~|<*^#52 zc8J*tAen_JX@!^_3&G3|Kxk%14R$&qX2)YNv+FTIYDH8HjB~1t8jhJ=^z@Ar zA)*FHI!#6bFtZD9$<{MF5#@A5%uWf$%#H=q%ua?mjS;icqVcW^(1H7*&67nqMU6nr z>_~|l=S1{Oz&Jh4P@Eq^sWP)8M?38hvlBo#KR^(ZADHn@H<%o4caa zkHO5Y#{{VrQ8h5msWNIfW_HojH%^3z8XV~~841A5F1#gM&+J5$(-AQ{B^Wb17ECic z8Rj%b%ub8OyDmTn?uRx{7U2{%0x`2AC2pJ((K7+#^fW_peh8(?%#IxGv_s5J0O9-q zK}>#N#yj0meh^_w?L6gzyd)9h$B)u3hRG3<&@ISI2>m*HI}N1>Nknx(r3F3Z(pk`t zclxQ;H-;%K7}71B1w(|>P-VdsNJTE7HYBq+J0`lp-V%)d7_j^ps0yK|kK;w?CvOq63a>?<|Zg^z{7-n|;E$|9vM~rn^A!f%yFtY;?n%Pl%Z`DZ|8e=i|0>o zuQR@${e!#3Z{3}L_wLsF|9Rg6Q(NGL?t#UBx>fved(Zi!%Zr<<%k$&Y)zj5Q6U~p_ znce@{=6`R_Zr&j-NdLfJ2yRG^9r-iG^T&_QPp+?LA1gjO`xEDvj~;#zU9-OPk>cg! zO~hQkadPYUcZ)j@pV_%rY<+WfXYXF|Uu+exw#c7bT|HiXYyR3Bt4D7>y=)TO_3oqo zpO2qjt-gKsom=?VPaeHVsQjQ)Ih|~*zLUOF6n9U)b9Ozu^PpAftIa=b?QADH-!Gnj zcz1T^{#Nm;&Z4QhvswCQ+hgXoU9^UH?egOKCf4}F#RpGU=c}95@#XpH@#)!Bqw_0U z#V>CaFSQ`In8C6JPut#D1}He*X2g;f|H1EV6(4Pt`T5oL&E-|In7cQti|fm)^UEE@ASYNq zcAXmK=JL(e#qn)+3dL`(`}RhG!<+(RMgxPteR*{n3Qr6f4RB(0`0qgL7p>C&& zy$n^HM3hr41H4`$($7ln{1g?OWSG-92fbdRFDC=po>4B*8zGBuiW-6IC3-_3C3Y-y ziE2&)#_4H>u9us5if&atxoP&*+kIkpP_9~{%2TN0V6@ZD09r56$GV}UjrAl)?ll{5 ze$oFSOsSo(Bk9EW@y>FpNyRWZLfT1nu#NR3M^6w=LnBaW!LCYk`!=Pspdat_Q&}*E zDJ|H+##k^!I1N=6Oo3G7f_hDBKOBU_Xs4S2M6E24#6~VT-sy%{R)DP?*@^Z$oR(4T zslSrSGqc-U)ty$OA?xL<9sQCt>Jl$bEQ$I9vN#^Ic0?2_6{J>Ff9g%-$2s$i8opk> z!kMOf`o@V+^(oS$8|gF|3Bb%Qyt*b?gIl*)B6238oQ@e_%7Z;^Ft`Lr#57MAV)jxP<{|VI6puTlOLGz zPB)YvM3_=LPq`p3NyPZ^qqK`*a)cy2O^lEP;WU&Yga9fn=qZ=Zf_}WyPqn@=OliT8 zZs{x-BAkXQ3#LFSasjnrB9|ELbiN8z(|U4UTl0j06x&DXcYM zY6rQF2yVWjX3*mi*EX3(*MXq6yYj7d`#vcYR zAn=-0Qqj)0ovL>1dI{q$lslw(s6IWI+hDj;(1KboVIV@)HlZ$hb+6peem0y|LC?bE zunS1hZEkk<(<$f$bvQ-91oAdE6x}|7I3=w}tsOlVQ`*r3oQgU-#yG7VLn^k^StEov zC9Oy@mI3o=i60S_#Bisc1w}3~kj2L`32^EWODy0BuGLq7BKik;6EWB+HXcI|c_61t z&XAK+qn%m?(0ch9P*;mDu5Vooox-4jPMdM#*UQf;6xlcwI`)fI8K#A?2!aWQZ}TIV zpa^p|gG^9~CYT_LW|*Lib4o)dsD%?7GJs1=kcBv1;fAb05=@XpZX6Y|IRWHUwPT6w z5z6h^6P-QC;Z8xAJp>eyJs^t99t^;#2eXHW(<|Qbp-!2x2z=Tb-dc@`AUd!WQBKJWFlMf-!HIRX1vqQ z1XF6~*%&;lOtkakM`;(sHG?XGF5!C^e7W6z!XF)&S>8D!X7^bvf zNWgR!3=vL4l?78E6}f=sv{VlVAu-zNW&lwu3na0TOOAKC;guC&__Vjbl|3SP=;)Rh z>$DmT!Mqbl=@KvEkknwO(|8O%?d|0RsTCa~fpJciQN!_RZ}jwy6QMIBIMQh{5`dXq zcx$zu*@-BpBVu++FlKfvm}Yh|%xR36ofeIEU4RbU4{e?-!YOJ5VrEB5+&CwqX9C9Q zX@=tb5K5Jq9XZ-*hw_5}!ubJ$nEb$ucev!EaE^i!>G3{zS#q+2=*h6tyj%7Q77id;Z#NM=_G9~6PT zDguaFS)dmix#W0fH@va}3^Tj_R(1umBgQ(d5VK<;nArgc&FrYbPAA0dcnoHCJtj!4 zh^m2cPL)x^F|&)FzHuT%)Zj>`$w&ZZcHynndS)l0oQ{auDZ!Z8v0$3n$uOreVs=_I z-gN;wa6hzpvIwWB5r~-`DRJYRh@J@;r>7Z;^Ft_AW_ILgryXK;0tn{^2x9UBGv4Wj z@`DIdYUe2zGH#nH?!{54x%y2vch3DHr4=i5NeAly)&pj*x_IL0&@W*V)@? zC`Cvjssk!5=qZ=Zf_}WyPqn@=OliT8Zs{x-BAkXQ3#LFSasjm=nO$vRb^?f6S)dmi zx#W0fH@va}3^Tj_R(1umBgQ(d5VK<;nArgc&FrYbPAA0dcnoHCJtj!4h^m2cPL)x^ zF|&)FzHuT%)Zj>`$w&ZZcHynndS)l0oQ{auDZ!Z8v0$3n$uOreVs=_I-gN;wa6hzp zvIwWB5r~-`DRJYRh@J@;r>7Z;^Ft_AW_ILgryXK;0tn{^2x9UBGv4Wj@`DIdYUe2z z2|Wd#^!cKxmF3T8))by^{2$3if(0}z_o zQG=aMh}rQN%PU5R0mXA z&{Hm*1^sxZpK5($n9_nF-O^bwL^usq7EFOuoGxUMN|!pbE=FQj+tHb^oN4lT@108uLo^kO5I9PjLgS5|;wX4l`!u3&b=Sf>?Yb}R%lI{=}X9W~hLgqR(V z!OX751gRBKH89SpGHN(xcG1%}PK1aW9O*O}3Bb%QytP`->_n8)5ivU@7&AK-Ofx$f z<}^mkPK(C7E(y=>SRccG?&Pq4y@(VH1{Hysb;4YyulzG*<;}RaS7(pkxVc__c=+!Q3jcV$^J4Mb>Frg; zw|9u|d^EYR`-S4(!|o30;x~%o6Va{I?M>vzCl{y3*XL)C$eXGE-V5$x>c8{C@^8Q3 zucdxw`TqaCZ-MtM@V*89wX(p7P4jyXikF9*=In)^UtQl^Uah`0-!9s}dhPP!`eys1 z!(qG4&D7VISH~A8Pgd94(#HEEFE!b2-E6@RU;FwmJlxy)D?ga*ZdI}f(r z|Nr-G@V*bc?*s4q!23S%z7M?b1MmC5d*=f?-@98p-@ad8&;G&P;A88WsP`R`F6WljHqy&uiewj>$97)%Zp?0pNT&%IAJ@Bel35j*qOZo*Ahq}eDY5v z1)P_Q4?HYp&))yVySM-KeDOQc`3Zfk`Io{vvC-`Hr+d{$4aep9A?|bY-f2#NDx+rJ+(5pFllg-!j zm@)HRZzuER`Ap_~3HRI)*Y|&l*Vl8KB>%sCd374PZZgF=Vf8B%HYT z=kZ#4)`KpK+}l05X-;Eq*}-miwjRGdC5~J79AxLu^c&9?iqFlCf{o|m(dB8=rf`1p z&g$w|6oOZ5>?>LE@7{^_U(bn9kE*B^>D)Y7`f6{LFBi+%7tL1TSqI(E(D0|k5}rV1 zSONsV0lOJZYYD>TSvM*EFpdY$uTTRQ>l59ZToMq6tPTE+-Wz_h`26gl$s5csc)uJ+ zgr{7n2U?yz84^4d`&w^=pDtdTeaUWwjF^aRG572&#};W&%z3spu`Ntd@sFdZn0w}i zE?NbC;TzBr_BQ9)+iSfq92Ez%dW0`{nTCW>{82n-&)!fAEIVf^4CZs69RF-T&o34~ zIs3e6o_2Q!ds1ov&-08H0M5@Bwx%iN~qPxq_-x#DPc zFnZOyClmh7ywg7~z8VyNHA)6^PYu!UH-yQPArSDxuk^-P7GIqGq|q3B5Lj%In!+d7GbjZhc?gSnEv&r`vl>BHgk#lzX5CLEAhgE5L^=AH-|W)57P=YP}8oN!_O1>m1V zYhHTVhc38>YffzUB~SFe)GzlzQP0Xz%bk}7ieHSj`qJ||>8+>txK0Ji)ysXt4K4ez z-zW@kOIYF51fK4hOpy9~$)8M4H5JP1MMD>HJ-YOKPv!dlhkeQBUh*{0>iS+H;a%L0 zpSB4 z+UWU$pW%%*kx8t$9OQpF-s(MnlV0i)B}ljD%Y25HI!g3(X7JR`Xa?Z_hda|vmGW|N z#km+vw|vj@R9k-cJKx*i=%ri!(T;Un_fI9(mkVunzHxcC^6XBv>3Pj4->-O*XU+17 zV1s3ngjbEEd9f0y%U66YZ>rf-!NB=N5pVgGXLzbDzneX=)m|4WLArfu-s`~Qty5)VKQV)+dqy(=|K|yxH65fTq9Q={hPZ%JdESQ$IN%cn zaKo=^?#=oEw>CTKf)Ak#c9^7$QF)e!%NWGtNw4RLp1O>|i^Y>;jh)Wi(>%k>fukR= z^4Z0n#wV$nEsa{l`$6;PW3Lstm$nf4O zNx`!btKi7Ui3w_n+6l^~K3`8n?Igigs}%od^io@Uez)m^jmqb`7M_|Xb##-gMjA0$ zmdX8YJWtQ$(D!WiaH|sK={41A$LYjwcwg=-yAO-Q*@CX@!u8nbStPUeG;Z{)fphZ} zx+$hW?ATl)dA|7Vc&$Bad$ntlmAnkM;ncxT4}P*=pQ@P8_C~CawbKT;OYwX0Og%Bf zhs}w%wAzqvUF#2<*(Vz!7EHI#Ch5FANgF?J;MhDzQy;){HZgg2)85|w)NIrKU3-eQ zX`{uS?b}+CnoAv=q0$Yp(qpGt;Nt&@=jVwS`m9#JOZ)keu6#I+ioFl+e6imF7DY9i zi#tG~V>(;Uz(&m$IPlTGYO`Gz|5voO3s1IQ;dbTDZUntv?D+-xT)!au#qR9Kl?&26 zm@pUfYW#rehAG~SC+aB``XXE8#GO8r8!vcXwd_rBSS)7MC=+Z%S}l=I+cT_@(@y<) z!I#SNr%mm~q5~w|@C#3|P&fSk=x+E0&$5&a@}gW~-8llA^nfos(Lz1oq7J0j^CdE6 z54cwRB$n_*EW;8Y5Nd;V9F^Lc)_CjWz+mC287eEVi-KwRVzHw<6p$Z2u{Ddh1AO5L z8tMR-6(mD~EvE6G6HvB?tChn>Ol{%GT#hY3P#m)FgRJ{cYzrcf6md8F!c#qTkOWNc z>0?6S7eSp(d1}N8!0S2I#z`;ufhT%l?-dotT zZg^+Hswquy=;@xx1gXy-PI8#aEN-{WP=O#P{J0zb(DOZ&>-#3{WOF~{3x8I}_7d@` z@|#@T8hPm1oodOwbZUb?nsLqtQ;8hPj$o%BYV zbY`dCUS)1}hS%UkE9?_VYviHlZ_-O$B7$^#{$~(|H#AB#o}_I3)C`{58O;FvpQm@z znSpR}g&C49-}5}xmfy8$!sf`sj1r$UA9!`JVL(lG1o1Pay`ab^)SL>Fa2*wj+ zH{5jgBWYf&1aNu&SFTgdo^p7O>?d3Pk!N_SEx(&R!R?QDa%bJpLz%>i%e*db13&V- zPPNyCN>FZpv@h>AY{IwhhCn%NsAS7O@^sH=hSdL$c)~Z`B&i6Hy&>*~Kk~c}7jQuH zCfx9k_)4C-MnlNMtIBP750Wy*k!N|hj6ujIy`HbwsVf`2*hyyYX`W%`z|jpSxu}V` zKW*5V##rm5fOF)T9i}wE_|zu9@an253i)_qS)~X@R!j;wN1n%FUPEd{x_VHtQ>SrB z%6n=8iP{Ou0iQqApWI6ipK#{APbS^)N1orIZuouNC7!&_9`U43*$q!fBMiYZxu$RU zgM-@v_B@qCjp2l7=jyZMW5PU{Q+9^)Qm2`>XL6(Gz1~jez0ajnzBWCDQ3z6Vd*&3a zx2JBXB_3G5>%~cDJh#tS*+Wk$gB{?OvP5(B{0yHtb8XsBz#^AevfU8sb81#oFmun= z#?KtNH_z54nFis)tZ|FqO4i)7HPjg2yHwCsle#uvgi|)l15pO|x&7)w#D4Xh`I@qgj%Sh2wJ%AXeH z*UbA?XYS%(#w)(>X%~FE3O|vT+(4hz@tdeNWmJ;CE5`?9vb{ zeZSiI+5V?NKUI8T_BraOL5jy4)||D#Ur(aLleaI}EI@|`&*{G28|i0?FVB`4M%rj# zuvL-T!&AL1d!PW}Pgtk6hb4M$f9xAYh^K<^r5!{F$Mg3O46DR>^p5sbo>jE?capg9 zl<-g)7k=J!Nbu*af36RSFBLyE`@)!zNQ4{g46NYUqSXp05cm?^gjO)c!2cJ=KybAL zJdPk;i+WQKu9VOoE(mf74`3+drkDO#<6sExlXy5P7&-%x;m}+yfp2dWb!4}MpXm+~xo(2HnRe1?QIR29KCKqaJ*9-zg3wLd)~LR#K6! z`@A|!Q&BL|s<*$W9r}IgnV3f-K*<{kTHp6`l|3QbE@P+csq;{J-Q&hUR_U z%Ore&+Zn({=vkm{BlPm=I9S;C?&R@?S+WEmDV&#M>0qoPa)17)*qj{_(FW^)RN$GQ zE){fYkg0&L3|gw3?KT)nCFU(|2w?y7{7<+4g=)|tKoLPvL(X~adn!;mHAQemh1C=w z7KCdm%xNN>k~vMJg`nWR3Zf7M8clH`XfCS2%!t{jQLnKlm-^Zfp=4SL32vw$N+HCf z$OvezsW4AaUpnEn!iER4_jJ%Qd*t*y6`W{-l*4m|PYXenXMhCl0am6q|0?*{v7Y0Q z0`LloRUZ^TX$2&>kAmPnq+Yb!^R!Q&`6ba-%K$To6j0Cw`9EI|oXiYV1W0N2LE76~ zc^(LRoBI#0e&yty)zw!oubwo~f?A3yp8RRsmkD{qx-%jFeA?j~+*AQOd1)Z9)nn$hq-V$l9Z zk<#(@bGp%NeSwcS=uR30+u4#X-ds&F@uc-<GTT%Oox1*F1ZB*dNO86+bciG3HP{k!A3q;vXe*^n?xmva*j7KS$cF z`QK(}e{D&NoMy(pGB<9<$eDvP+U)6BoEF4V@qZ_4>`B>Y6@E<7@p`fI`QBUS#op{E zwBFi1kyt&3UdgTU#Xn0X>!}y~B1Gg}oocj8AM9v9`@lW;`!sc5dmB4-^M4kZ?z!oku6YiEMCH@9JMN`SQFOODzXc3Cg% znQ8ZWa3KclUY8}&;la~8ZTGrb13YDRF@QZh>C3VQ3J{(HPHhiUge2Yk3r`8*p%qvi zPlr>=zu@b3+CwW=s>KhIZvNn640uSxYCwm?f^|oBZ_u9LXf(qnO1t@k3o@)$K!Ff^ zYh}V&SsWf%yhytF7oI%A7k$#=)a%0H&~N$Lw|YScUQw}vkLpEf`+9I625es!bfd$8 z=aJg>b)9&tg;H|_S7q4E0pg;$E@RRuQvt!eM0CIB08!C9YK8A4M=^;GCg-bxCfrE-zk*omS(Vho0x~8X>!95t_n^GU-LItV05(B?K=gr^p z5`|R;X7Kewn;DS*>-+eaCF?d9%FDAXnznEr_)4NK41BM6`c81bALN?d*~=u{j1B#y zRuJ5jp<8+{pWgWSLZaD|6VcOHz*iD&7C_F=Q^3hAKm`Wt14%c3aBGJ6Hr#G0-9k_a zpt&-m|0#pz)i)7AEVCv0)4s*vz6|kYFwu+*f#&KAi+bL0mN!UBJ#0T<3ST$0ngYav zaGl146Q!a-T!}Ol9C|*eO9fp+DQ@CCoc!Z`=Ir6sOuPAm+cb0=p??@12YhwVyb&g% zv5&KLDPRF#7PMIaxqonCJlS+9Pv`c1@D7@G^B;O9s7nRiLXfF|@8mOQhfpf9@^a88 zKkPd6{7<+4g=)|tkXaXOcXc4iI&b6$sVRJM&}s@43p_D2r-IZvRDj4{k$m*>(DOs0 z5Cl40iA2EDLVY2KkVv=$vnxDE3qipR8$>CDkWNN`aLtCk6v7LhX7;}9XPG^4`sR`i zuKG98>?wz5*FGr(1-EMuvPlL$(q+Yb! z^A$in8IVLApf6=0t#lJB0zMIw449P^FY|m-+zZy>gGQR$e*@#pLjU2 z$V`00Q?TSq<-tuGu&13+hxYo#G@AOoL)!+=YnuL!A-Ha1{LGP~^A*2Ira?J5=LSX6 z&41+C9qi@@mL%GOj(A3=ZOx|y8eA#uXb&#bAP%~d20^}?U$|OBfA-DFonr3u;EoN$ z+*2oS9%#@Dq+lQ3)0k^r#NGV!Jx}FeH-GPROlwc`5LS6@H$RYNuw&BAKi~5uJJ{O} zb3bwd)&=G-}BT9wzNa>v`g3OZhr2;*6DEuPkzWS-}AM%u~SFB&6nP^-TaJM zDDRe6J^qyBe9u?j&|CL>_vGegb#Zfcc|qKM6@Njec_Lf;-S_ z^F2?;J_mlTBg&!OpJ!)Z?SI_!w@`u^GJ50%HO!ik}bXv|Egmi+}MGB)gfy^28M8Phw@h)Y6YgT zhNpTt)&QX~&HWu(YY^_wMrZN6(XamI`<@1(WC@hvSpI&=if3a+nMqlVu!9u86laRy zo(^>TF)M=(hH6K7#j)o3cq(S|0vmXSXtV(efqFt4SUj9_vEpAO-tVa*DqWH5M7Ee-&Fm=J-O`i?~zJT|Yo-QqmFV22)1YH8s@l@qi0LJj_ z(QFJ95y7uHrkpty9L!q-*}k7tHs|}EKBCHI=zs0GI@z?}Jc$d*9c5*c6iBQ;TgQ#R zQ!*DW!+=%Cut#63%g43SCVbtb2{mTFPS>+{5-ugybx2e@iY}oKQTmb z^M=t7APSnRH#pieogt_gF>L>(KiO4n|6PA5?8%&m7jIa%bN_r|u}AI$>Gl^~xS`tq zyhdaMG?#903vb;T6UhdXfcg7!pK1Qc@%h8wspe0)J?9eX_FsC|r`rCzRiL*2GWgE2 zp?ulybfE-diN(kU4w!ELrRRF8T`yFOc7FbgE!JHR$hO)B7{V9*jD|o_z#kJ&X9y|| zL?=m$Ji(BIc)56b z+>5{iFn3S=40A{B&YlAE3qFluh&V`#Ji&z;nB)NaQyYGBjYePQP?H?;`NYynk&LXH z);EI7GjOIS)gm3fxkO_UrXxwV+65TF7x+v@0G{7msiCGc!bAxd7GG(Q1N44b{B~UB ztUTx2fu}r4`z-vf=e!DPd|&9#aLQtTwo9Mkz-JQ+%uzYudCBGS^WYK<^sCD`ZhE5q zov-I9e|5=AHJAd-+f%#I^F}UCeR*Z%0`)$ry+M#Q)jIio$-MGR4*k9aJdSARsd)CL z{K*L=&$7nx?@Q*v%^3KoI*By4`M!A`2r0i}>Vs%j>@+j?#BKb{k$dyJZIYQ&TRRs) z#cw4$x2J8N=5}tbC8we;LE}_mUE`?V`*({UpS>#X{a(A|>u~enJ`D7u;69@KGsxY# zxe!D7!(aNT27l#D1?KB3ZlmT4+_|^a_B0Fq<~MS>PCD((d{IsLGhj~UgcJ8QwUHA?UR`nY zSH9Ok`+YmcG(=P+zwn>eo_3*b`*M7@ea)BClnwQwU}EjLjF5Ji*PeW#4s%fn+V^=f zrfeeD%Adv-{!}%?7APDJcBY*#;oMw33H|=LCyMD2L+$Ardgo@>g`np(&)Agjp1NP}B`<$;Svw!tjG2A&-n zZGb{x@(lYAvR7TFH86%}k7i?_h*((H>-%Pmb$SpRT%>^xVnZ9ZJp**N{erLRDbHF- zfrN3{7l)+VUbrxW6i!C~PY;#njgs7H=I$$c=pa_m0XaP1e>ctCP2?x(whu1GKnJmK z?$p-bT!xXiy(v#`8Ktm4q-T)91sKx%P#@evb*Q-r!|=|7pN^-IXI)?fUk5Z90eHS} zEyh#^qI_R;zw{$B2c83}0>F1ir`-R5FA!SY+lwUpg5B!>WnBIZF2qnRyBAMx_RX~z z9PhU-dm?$7`TGi?Y5vIV`4Zt&^QYXNUH{}OGzXsLN$ijo6?uy!BB!14Ht z!QzI?1GfK;t=MXRY?V+SuMc*@v=kg%h@sm4yaX}=noBXPTU_#VJdFuU;lZ64ruhTM zZ!W~(wqQ2Qr!pK7CTX|*p=W)n?Y~KZPs)5CY9*7ybF;bQ7K^43cj9a^Dk0 z^de8L1@&6c+-||JLdCZ(mGA9>V%R^D2Je&_6(5S_RW*D->)YhN-R6~a=<3RxZA$m z_w_t<7)dHdM*vR`m4}g%T!USbgV=K4|18%qcjWMs9K;H9CmflrZYQ5TEB8GWL^h4enu2X{AZ+ToA1EoaQ^H&6fm7$LaN!ej*C`OS?M z6SQ562BLgl^pd3AvJ5V^Pz8YR08fr@nrE2_gfeZfuew?e1?}{Vw!}j z%gX?^f6x6?+ke;03L7rVIZykPjcCdK<3-}tlz+Wx(GIs*7pUh58*NSjBwTBLPGBqMy+aAPSlrE~c9-<^Fjd5qH~{r9UmkMIO)= zi#Pw0KP6Un+gH1UKw|Bg-5~9@FN13>aG`?`PWwJj{?tVtUhXt=_vJjp+>yIeA5`Ji zgxxBpG1odR@|3=~he-~wKe^%aBu`CpNRh;9OM#zu+n2$`6*$wAI?xVJHSUS(u`918 z4U)!br7O74!ej*G`G?ju`o0)popjrmrRRL8+rCnE+c(c)XQ1r1C&Up(Www*F+rA90 zu|QqtglN*=o0}{+a-(dh=cP_FZ-10Kdfvdrg*z;!m^a06*!BL0(Wf)Y$}>6CZ4WF@ zwDU9vopxsGrxc=M>=vzTI_GpvyX}MfDMn5lcy)6_1;@GaCqTF^S3cjY_#ejf zqxQ55b=%kDyX~9%D!ADzyX{56gwu1GKkYCNuB|{F=Asg$?+Z6sC_Bux@~5$ducT$z z0)+!#Sew=sRA{h%ksPtqp01&HZgyP=dR|xTF%KIv%Ibow7Iu`hzdX3{0`-^6g6P2D zshsj_?rH__gxQq<*6`K09BZJ^U_W-k^Ph=LMFdXrZTqtJ^bZ|Bfu-T7ea#=aDvzdE zp%x$br8~t3uTR#i-7j5VpFO@fE`!@H(18o9038fGCsZEBXk<@g1J4ePHb5c3SLCL% z0pZ~MUhpqsz7*VZfnMavwV+-Lgxf8YFY14M*yfOsS(uD~Jl}c{`~5M(I&HUHcn+uv0N;C^9^Wi@>SqlAFOmoVHsq69KyaOfYT3Pb z()XKNEw~by)rS+w)6Cyj2u^pBHYkFud<3 zLJ8Z-+kVn*A6#T1z3?Sk(9TaC+KLLYL1TDPq;yW(Era_ljD|o_z@OnxHzA7phz3u$ z{{w%TtJ?m%CRp5XdBD>?>-O(O5`i#s`}d4bwf%YVbObE=7H91*4ilD_h%rnS#8diX ztKfbM)BKU+A51oF%C)&Q2Gj1>18o2P^jEe0cMC&t{)0npw>`1q-Ghn676Y1{db<64 zuBY1d!l~2FZ#WfJd||-4TLPWa?LWBR!e|H(1xZ2kE(+kD)3P?+ryrgu7sN^)JN5<*lA|&OLpUD4%|EE%8E(MLiu;nZGYrX zVNthz?{8d7&QZp*jt6?j1zgU`|WKQ{qpWzn4U@ zuegnxFLGnU4_^6$rd$(&Z`&Vv0*1QnBWqK%`)ENJpW(%d51uQ#^r4bz(|K?a1!_8n zPobUpz?T!09p;>X3C#y4?rUlzCyu;2xFR-x;>1qNwI`8&lYZO&$kQ&=ZNC`bZGXg{ zp(?xWMZttSuw?s}qb`CfxV8dyn2SnKzTZ69pdY*UE1S}_Dkr}@sDgVfGHj8CL-XLB zX_coUf?Ipew#gAgwdd&?Dm<`8#`L^-B&UM*_(7Iz@VlhHyb5l-K>g*i7Ia_)Uolqp zi>sv>rViEwSi}GJBF7piG=eKICbq^zL6UUaS9_lRq2nj8S`4*so(7h{^9E)20V~wt z7Rh(rtKhZ^bl}1&KnDZgZ>Ky#(#W312L5*zMjN0I2w38D6Q0WZ<8J$E&l5%TB2TUb z@mkP4k0Wx5V@BoMyr3BNkEFr83NE%l4d#LrIuPEQBUoi)x=!oV7{PrPW@CVeXl}f~ zy=csvqQX(8G{!nPh^_X0Qyw~q4QO)3AbH|19AXa+zG*j-E0LM(pvefx^B0yEKWJjg z_r=xb_;=E7Sp^qcr~<%ufG5W{&9nMQ(|+VWS#SS*Vxh-w^>Md-6A25Wk2O13lqCmLe!tgRoEP^>T zq}zXR(S>UJ_f6p-6f#bAt>nn-L&QL`{RdZDsJ4GEfsO#4_*tn>A{kGSKkZHD@5_Cr z`6I`FZ>V29f6DDSMgX>d&-zr`f2Re-`R9v2rtLpe3;PAw_4M_=cIOFY(9X|y;8}M) zARA9pwm&e0FZ!vr|5Ob+WQa})aUTiLPv0|1uk?Dvn74Mx!Hni?@=?_SaW8ZOS|o> z;2I0mbxw$;{XO_$g|dsB7h#wrFnRHNQ4ePo++#6%-pIvyQa6PuDEDU9JN~wP6>4#u&;CSoUwu5^-NOx6?@7an}EY+ zNoOA1STTO)z`cblE7V8uoWV{09e3MTm8Wf}+rGCi*OJp*RDsW*&8x5M!3WwU-S$;* zI|b@K2hSjP>*jt6WxF{2)QL?6=Ic*eN6ibWSFomTYG*abHs#IdSCGd=p*% z#3{dKh&AzT`>OI4HPmf?FuL2mc_N#4mEHEDU}EjLjF5Ji2iI1h4s%fn+V^=frtCh~ z%Adv-zLJ(<3lt6+ENxmt5pk2AFb7v%pm%O|T?l$ETzH`@JjjwQzVO%LU!epyUZDPR zSqm~SgqtsvE$3vzU5d0ghq1(2Df3c8|lQRqGIEBV>(oO`iG96z|wHkzRoyu zRkq@@8es=XyX}M9F3^DsD}xS(jQcK>$5I-_)Hd)%IHL_v2=FX%x@i&)&IVwSblcaS zD54j6(%jT*LCteT<%_%^7B4WzlZkKJSHZ;=sKH#&gAN3qGb$U!b+S5mL5F4&ddo)jX+t>Whb(IbHq(EZ*xjdX6Mg~`0puONJ;KAbN3ZJ!`y+xH#b{Ma0o5TopN<{-IH$n;5rL*5DRsnwtnFb3*{j# zBa&EcIk@9*+gHKG71H}qAKcxvX@@W6{KLA1;f)FFhRu{7#0K|Sn2dltf0Bb(#sxa_ zQ@$^I!as`He{ivdDgb9cxv zBH3UYFn?bmG|eBmJx}+hnm^_C?3Vw}5|8&RPkP^N_m^%J$W8yfxy{1xDx4^rb{Pg2 zSx7H@$N7r0oPQ?Aagds5^HuCl-+2k1a;_{|L#laL%zBwkRlwvv9^z6vg`z?q)ZfpmD` z1`9Q9kt9!O1b^0RG6L}Y=0*$71*S3r<@?e${~+zQ4=%Pq-S!9CZu`S>N7+$NJd#*f zCY;%C@upqp!8I1B>zq)F_VZ58_tic#y z=AO8XpE+{x;Pc6oOmiy#PP*+6J#9nX_Pq;oEjfq5w@{RQ^5D6Kg#>9b?LH4~r$F83 zKofGeZtka0c9_#ooo1T-?g{4WD{iCa3*1<^oFdm27N;(fZu=uoz)-h+v=S8UZf>LC zIG3`Q9ttB?oS|8p-s^uFpDY9yQJ|)CD4lj@_B>Vb`vPTKI;RyjP1fG{ zl3M<>D8H6h{11~A_p}Rj+aHeawm;&_Y07SUQ7~RvmZPNI_QACksKZ=Tg7$sBl%{MV z*UF#97XDNwxB|TB2e-;YCxux| z-XC|{*L(h_x9CNl)Pr~}Xr3n)#!+p>`;GGLUr-FYO!7JRI=I*ZHJA%h=s>7b$dbr$F# z7TTZM`kOl}lpSTvV~G`IDgp8}7)hgf9b8-?y$|)l-A$Wx`0xe`!^;mAGodwr5qurc zWCY;(&5agZQE)1|P`)pEN%C>ydf#(ERRH)7@Z|WWc@{f+YjAjxggdZXecWzY2iIAs zmfee|eZSJS8zzyTJl!xy&gw}mq!#YA5H&<@&zA_TZXer*a(j0D<8J$U-?KdFeYf3| zIxWaee?csm7>nGhX}@?+ELNv%(r#G?7gl=jZ92;f0TqZM8Hogs%r04S}LS z{c&S<**Zg0zMkb)VEgynPqqDby`Qk*vfe*3^B@0MVvV^F=K?9+{(`G5RNKE-jf{Zi zev9`a>nD;8wgL0^r@f~61IG_w4;UTv8C$$ecH7s%#T7Wy zlRD52&sX-;$zPH@oe_fjEKEiKo-f>JF_jUhaF9LWx8v$|={X_{tZs%zp#Hz1sz^CL{*7*06DE4FxA62t$?`gSKxT!*Y5Y396X6Bx_jh{Jk zZ@!yvl9^N6Hy1%g+-+Y6cUGWo``+KUmYk9=&?y`6fhfz`elPKAPtH*HIS@; zcAwJ{rxQ&`Ga%01NVtn&Su8G!3w|yPlOM$xWBWqK%ySa^`zd%*lS`VL0EFr_G z$iDWZ>AVguqCidOP$ANpn;R)O>Y;2p=VV%K0VeKiY9l8Oyt=ueB7bt!?rT}--t&sM z+rF+m?Lyu5N29y#ng8wgPpSiwe@dU-dTMQMQw7g-v4= zU<*&qGHijufiJ90YaS|WIG;(67%ETK&^tG~R0KV*7F-#KPio2+a#=0xJ4M`WUk5i{ zp#E}M5FHo?b~7lO&(+$gt>Fn?jx|td9PUg!DU1KJduGyYUwQh6j-P<_Ici@`&H$As zR;-?Br+IMO1v+qHpGF2lbKeEGFLCRl?atRIM#RA{MK@m5!HpM28-NfHF1t`3DJiCD ziHzji_I2%vB6^W0^`Kr0n%gZnlB|4@7sTQPmYdJh2J_%z3)EmP=s^bpRjp&#uG}@M z?cmpG4UFMScV=Uth^VaV#k!KR2VZ53b$Sq6d-{kDV#CL_4FYtxea-Vlmt33ZjEpICpaE zKa_W1c9zPcT1F{s_q5SGxVS=kAL@gOw6N7{~KM*q#z@ig+Rn;Idw&%$H`;Q7Lh z7KZmeXnu8Gr+l2t>tJ_H+bx4DEu{C|HuyS|Qk(vw(ssrt+g|u! z!hgho1~cxLBJDB^F0znb_>!s9&d<|3!wVlJ+iE0W2wx908UjTDPx+=Z1Qi;hw}9>6 zb3fJg-!-%1hRa3Je`npe{qu>%9vlA2kF6G-@u{{yuMr&q)cqDCmF*+f%%l=!!S?UV zeWv*%$G20zSYgVwt(Hu;|KM5+)%M?MNNxYkeHQ&Gu%S@d?sTC9!t2Gz26-Ue^@58m zRJ&fN80q}NjTY8j56HIK1{lH@{fveHQPA9QG2LXTIIuWE@1bdtC%EVW7kNNiEZ+PN z_+Q$pi#&utV(po20d(63*IM8r4Qlb}OOmYyO1^Dh2NzqQZu=u`xBWrzm1JdWJ@IT} zZp?m;Ix3QHyw|}s7O3l-(1!H)<|d1t-YHw_c@gi~ynPvO^t^$KH+NV}T+o`rdZ63> z&@(yIZ4VZMXy*|wtWb8_Q}PVfOuq152RBsUqiVM8JuTNZH&yh`tv;7#Wf-Q3$V>7& zgW$%B@iRy6opEKwB$lG6OVVwB=ucr$w|(z#TuaU&U!YTV+XGRC$q^pOaeSDO<8Ry7hn|3;Zu`jE6zx7_&sNEf zc)!6ZTlAqic-_;c^WY*1)N~G?O*`|Ew&|REE>@i**|IIui3j&njGQ>|>gI+D%srUe zhphAj##|TwZTwTABTu_N^7`uPX7}mk<@xGrm)nk;8!7s<``P08<45Nw*VnU86d#*? z^!)PC!!P3J5H=Z0n13^wfhSNO!;(o4ZSI@E3pG+HZ;rj*9G@3sa@ zCOaSGczg+Kta6qewfG;C_3=Gv57PCaw#y^_x1+E2cK*}FCuScb?A$$^SYsv_yGFj^ ze@>?2xy&c1snE{GlbCrwqgRSg&wiAh(MHCE6Z7>YVPeSZc+xU{VwB&>EB5on2d_`o ztKFY%wy&GBn|F@)XZPJdr>_r32Rjdnmyf#_g5&EqPM)r=XU`XRAKrPe01`ugng#kWqb&Q4xGUw!TA>_hRNY`^&U{POja^W*EA)zj@*lbsif=T2|ldAge2 z*?Xb*&lBJBtm?&$Mz~HbT$1qd-j9B$n9V-a`qBU2-q*xfb{*%jhZHruBqh2jT2xC4 z#YrSf<1o{G?|uC-E$RY@`2!fn?=jYyg>Z?=d)@@}&1qiKq{i@DaRo|(q z^Kil9&neF2K_ESH4 z{-wt|&;8UWneBY%_+LY6+j-yk^W(jp_fP-5yYoW!0smx=eWzugbnKI!eX?So?6Xe} z*e8eVlcU@S25^rVvB$r|jM!sF>@g$um=Sx-h&^V+9y4N(8NuMR%!rm5(Q+wfM9Yk5 znGr2BqGd+3%!rm5!3OV`5gjw4V@7mb4>O`;Ms&=Gjv3K0BRXaT8>(kU^vsB!8PPK% zdajun(K91@W<<}7=$R2r@roI-Vn(c(5i4fIiW#xuV=*IE%!m~;V#SPLYWA5C`^<=a zX2d=-VxJkY&y3jTqcbD+nGyTU2qtvEj5uIM955pem=Ooehy!NC0W;!&<6%Y|FeBJk z4w(^$%!or~#33`{kQs5vj5uUQ95N#gIYMT{QC>=pm=Q%9kj(TP>?mR0bURq&Qo@Rn6@?k`rsTUNnaR>50V z!CO|r+n&t{R>50V!CO|rTUNn&h*<@1Sp{!d1#ejeZ&?LzSp{!bY+|zt-m(hbvI@>w z$trluDtOB(c*`ny%PM%wDtOB(c)QQC6RY4YtKfVhu?pU@3f{5`-m(hbvI^d^3f{5` z-m(hb93DmY899jo9StKc20;2o>r9jo9StKc20;2o>r z9jo9StKh6AbgY7Rtb%u}f_JQfcdUYUtb%u}f_JQfcdUYUtb(&L*0BoSu?pU?3f{3I z+Odk%vGm@tT-mY2)3I6Hu?o&AZ^sG7qs2`Jdh!b1u?pU?3f{2_-mwbau?o(bg^pG5 zj#coERq&2g@QzjRj#coERq&2g@QzjRj#Y5(9ah0RR>3<~!8=yLJ66FvR>3<~!8=yL zJ66FvR>65dSq1M{1@Bk|?^p%zSOxD`1@Bk|?^p%zSOxD`1?L=O6})2=ykix-V->t( z6})2=ykix-V->t(6})2=oKG88!8=yLJ66FvR>3<~!8=yLJ66FvR>3<~!8=yL-Atcu zICrdqcdUYUtb%u}f_JQfcdUYUtb%u}f_JQfcdUZ|uXBE6>6})E^yk`}> zXBE6>6})E^yk`}>XBC{K)1FoEo>lOkRq&ow@Sauho>lOkRq&ow@Sauho>g#G0eV)! zdse}FR>6B#!FyK0dse}FR>6B#!FyK0dse|&(dtRAP6b++fTgj0^kk2?cs&ez~QtKdDW;61D0J*(iXHRxFd?^y-!Sq1M|1@Bn} z?^y-!Sq1M|1@Bn}?^y-szF-x+XBE6>6})E^yk`}>XBE6>6})E^yk`}>XBC`>k5%xV zRq&ow@Sauho>lOkRq&ow@Sauho>lOkRdCKYR>6B#!FyK0dse}FR>6B#!FyK0dse}F zR>6B#!QCX8R`8xx@Sauho>lOkRq&ow@Sauho>lOkRq&ow@SauhRaU{-v|h0azG4-8 z#VYuURdDx<_nEugkIJ)8^368)qu}g2`4-QLRdAMx-0wH%YFGtdu?p^fsFq1FBUlA@ zKOD-Wm=UakyI)1jBav_FuUG|Nu?p^fwk{u>Rqz$7;44@Vwm=vQEtKjZ;5||XD6RY5?M7fvM=W195U$F}A-r3Ehm=Ubd zu2|J_?+s>pm=Uakv#Pq{Q=QKnR>4=Sg0HxHz*vk8UDtb(st1z)iWzG4-8#VYuURqz$7;447zG!B?z;uUG|N zu?oIo6@0}i_=;8V6|3MYR>4=Sg70UMVKc(rw9L$O_bJ&Y`Ek0NM$;Q|ZWd)9qRpBN~_Cb!^O_%H&IdV5yvTx+bU9+2gg8|Et=d<5Ulc@?fOR^7g8IpY?NABiF_Kh6*(NXhS^TOn&=*ihxxF`Cezu-B&Li*o7*!<#)&2MitcaM)x zAHA`@SUAG(|O?h5?FlgGy=kM1A8esXp1&Erp>ojiQ~ z{>f*K-OC!A=IN7*2bagYFTZ!^x#s3+xT^p1lRNKgn)t8dz3ks@_U|tHckllFcbb=f z?BwEd{gu;456;%#|1&2KAD^x7eE$5=<<<6kHvdPu@4x+oyGZ_?%^Q2RZ+`CSo9o2F@-QD@C(pQf6YQEIge5tGXQlGwbd%Jn=@-fl+_{u#za=Cq@dE)kW zH_vSz(2EbPxQ8n4zK1UHZ^i%f#N8*t|D;DGf&%x3YcKy|^Ym9Y3bvnV-rK(C8Tx;> zUuc>~Z$CWt-`&3P@)NgztZDo(8^O;tpWS}u^wHzDuZ~aeUv4-4e?C09d~5stryt)0p=H^!~-hOoNdcs=0A@8p_xo)q2M`GF=qeKvXy?b$%0yYDz=h6iaC>itmD#K+hskF-6X_&7t^7e>n3 z5w4`H-6#`$oDlDf^>j#2SEpxeT#xkh8$BgYCs4Ou9rY0X_KV4DLqfgv(V^+s{*L&C zly4^*_C^|p3NZOfJNn9y|0iGRMqe4S|LvEW=Uo>?=`=igxS3gZKXLPk?OV;qnhzYm zeYw7H>3Dtb^o`T?{o_X`57*l_w?jE={>xL%Pc^TN8?oiCVbldDYw=2G(g z2P{rC-|G`sZ8k|LPaMyuNpSaX%>9`SIq3d*=@yyC(4R^!(AL z{={M~v18Ihym$MF&rRB>_wIbDdG__U-*Bzt@#X2yuA{aO#OCYS&pqRsuP46znumHjCa z_9bZug{1;{+FQCeY6*7hshgj8;klhpyY_<~sonI?cXzj^kF}C*`3mA=@A>4`>WR;O z38FhFJrH71_YrVlx*IbH460}3U&7NIH(kA zr^DI#yIT!*yt}MFkPshp|77Nvp&5TU!l!N3fup9ylK( zBOjp`FBZgAT;6B3xac=|{97ExxOQuC4dN;;r_=fC*S8wx#P%zLN6v2 zWg{esT~6T)c8y;*JoXi(v?_Lu6+~zvyL^(XjiOhbvy0MQ6}tv;6_-zZzqi$(pN<_z z!Y}#5Yi<(5Vk{3GH1q>niiQYJil|fTBKkXezL=oPcy>?`P$sV-b4neuKiF#aeae1x z@BF?iqGyz^kAc-gq6~>VdT!;=p&x5ib!Z5kVu$dLwi=Dh!S(@Rp85A3MD(kfMTi?E z8b+ttVH9VD>lD@(57+lk-PS|Syi^Sc9iA1Uc!VX+R*&nOQf>o^^SMf4#mP+&Rx$Z_ z{_<9X-sid--{-p7$Ke&Ar3!#v)CC1KETqlF?H9Hh^nSNYW$qbD`>(-CBThmwPCn!R z+E#;JWu9;h%noxeafea^dU>iKs`+KlX9ZUiiv|vO;!I0`D39Xa2+vX#kc{2IhHT%^ zh^gm_$)crB6*XOo3KgI4RV!jT!IyK5x>YTvOo58c_p7d}L46l6e8{y~?r&+Wh8izb ziHgy$*!~!IUPbkJJ!omEi&`&Ti;A+i-rYmL!@wBW5j;FQK$qjiHC&|A%?_dWg19o~ zbrf0Ddx^p-CLd9u_u5l>ua0T~>b;<#ip5s}q4&yE=ANPak6IHTrsD9KU+BG(LUI0e zlpj#h1-%(Z)g;h+R@j{(uYR9e9sKYss zdi$aR6dll8^_27UOC8+n(wP+|wbt(Ib&C_V! z9uhp9WocC6hSom3*ss+A4Q?}Vv2|6X_S?Yq^wjBS8)}y5>%Xd z>rl!{qOgXEG`biHb(*OadaonOPF4IGE6}i!)*1X7Hm=5Tq-<5iuR&bJ<+GO1dyRcI zX-!z~bx0j6>b(>d8a~on7t;y8oNLszYB6OBRBXO)6?!jVxSacG)rNX6Rf&qxuib>+ ztEisU16uT=-b>e_qAad_JJjPa1|AVSJUc*_Q@8I$Ivx9j6(+=$G4F^Xi+V3nSo0NW zbeMhxpE-6y@AX`~r?X4N05xAwGJ^StbU}yIIW$7|mFdiVLlppZUqCDZkxm_teC`*z zucT0%fk%`fQ1>N>s(5^E5V|j*G@orplu%LkCCRFYd?t0bN|4rG>AnX35H6HTiy9(0 znMWR-yH$d2dI`D%p9X9cM3e#w@;Wjf$y#KS4s1XsB`1b#kk^sN zeU0NWX;e}jOrG1N1{SqqiVhv2?_nz7Y|cU^}Y?g9m9L zhuSe!i;mK-=l%qrLl*VN3NgT>l`v|^bTv9ozaAFniP%z&=gH<$FrFu(yoxN`2^O;+ z(3~@HkK&7(vL1DsB(0s3`*)~d}VNtw4u9TnNH>%stKqq19CbN zUkzLzgl>Z+jq+UFqilg1GeK5I5DcFh}gW_qXdhasF1cYVI5O3v#x4;v(4ob zlv#@!RCJ_Pu9<@FoK-Ew5#@q{ypGJrQpi$F^Uzz3v()B3M-{R`UPtEhr;w!}GUhQU zONr7tDxal<3eNP7JFAJuC|gzWY^*^?=d)IAG>6`5f{n6QRWt{A9huK!LT@(q+oW&d zz1ilZ>DO| zQTmmh(3=(Y^BO_RWYnAKYIK~%wekw}Fo1~!5YH3PK~1;I#8rH_JuGHDpgCvY=JGJ= z&7fT*t@?{pI`#>@*-GopHW%tpZw4h*G~q6t(3@pCbMI`f(V@Nt$mvLYH4u8Uq*0!W zn~S)pHxp!4MBx^$(3=6RY~F2dEu-E{5>_#V8_Pmw-LLJ{r_RUl+I!eFN6MD0veqJMJ{f>GwU5$>j zxK`e;>&#07V-V93W zXnbW5db3O?^A6Pn)SCf09f_|7LT{Ed%5$-&Y=L?+K~_iP^Mufw0j+G_^|Wn(dNWB_ z$5hO$2erM~ih!ZaTGXJTBeinP6m;jTT2UNPE-1+B$b2k?-i&F+z1d32QU!S(SusmN z&CFv`mJ+3PR6a`y6`biEcUI&6XGPhnif3aDIy#@VYNLsJvlV5pDxMAUIx?Tdgx+lI zw@Kf^d$SdFwWv2!bm$0uUt1B`2|~s>>TI>hG7UO9-`fhk8L+)j4`_{tdNWmvin8}B zJ)t)%>Sr}#MayK=o9Sv)oTR&yk3kQ^nYd2?@jMabRb-@8?#&KR*T@)1@kPCvD6OJe z`M3(b*@4!Z?Nfa~y&05L(U3alBcV6TbmrdKr<#CzGa#oU@zp@+&5}lWF78vdK)snD zt0VGxLg>waRyObUDZ!%NOcK^H6*KE$ZEv=Ds0(G*q6QTmsg-kqpgU*P0mTvJf`Yt` z%*Rrw=1ep0%{I?TRUsSXb!5dX1(D?*+dO23vXm&Tq6$x!2^E~_9d}ma-fZ)LY8B7M z8dP+ob_UOe-fDu4vR73!2YDTt&tgJvHul@3ZvhpNx?0qmDLPby`@XNOi0lL*;~aIi zT4b399i8uOh29LFo!aTeFgho}SMOx!${ zkLQWKHw#bY3;SjeS8W^6P{5Odb3O? z^A2^cs5b+0Dw6PIs?eJyjq+SPplpG9GeK5I|P=`B>H>8+xm8H9w>@T7_(o*OB@BDP$>T zGihcXqbh{5lqjvE@>xo#;7sqhvl@GB^8jlV&&C>5bfi}9%>;DJYm~jJqB)~C3s15N zz1bMqq;KIBar3A>>dh1#s_(*6_Z5+yAY|U7&Q^;o)1agCy{*uj0o&zzuz4&W^=7IT z9i?CC3B6fSKdTX&hxJi!rmN9$7T3y0b-mfM;!W=C3Ywt0{h^=42~N8>Al(3@pCnRlorpxz9~=}3Gv5PGwuQJ#xOlr2zi zCdlfDe4Y?`GoY2tyUp7IP;Vv)>zMqGfy=Cjm3!m&JC|O_zqvzj2wv%1))3)I&)J4s z1g--@agcNeTE&-SZtfAFr?g7)Dl#9-YGgxmY_QwtH^v7bG;d= zW|MeK8kf|ZrI^x5(p_za0v#$s-`CbecA`C-rF4N)k1W%mp$qpyU2lf^KIA`E5TXY= zTH~RnN7tgF^ea8ro2lFm9vD!@B9UOy2I==y20Bxt;`A$dd7gkKhBJ`>3iCvi*O1ZG zEVsInvmVf#F|eoj3RzE(R#Ex5y58)&aX=<~W}e#ILqKmRaEt?z8X8jPeB^qwPgS7F zytAU3fPUvC$3k){5?>8mZ-%>1-O0pWFaI3)eW@Q@(MEShi+$9j2 zC_HI3OyLHBlv&3wfG+iBn>z$BuM{CtVJ;}hYskV)f@)+#Z#Ams zo7+=$$Od^8SvgBFn?Y~3xj6;1lqjvD@>!~yXXDOlmI*6*9aH2|}E z=&i=sD0|gKbC6e&`79>&W@BU^TaVCfZiK?#Owpks40l3nB0E8t&r<4a)yOgpDmvfW zO1&AdeaOC9Jy_9N9rk9b78Rvm=}Em=Q9r8@bV&*JX1W>`XFB^A<_YLwSjVxsMJ3D= zQC>q9?ovs;nfOlr(G1z#s1mXsz*W(yS7z8S|7A9N;hN7V<-%UngE!Y~wZ;iLF_`anMR zKi$0Gllk%2*QXEOyt;HV@ule}@dsz~kHZhWzLtG^{429p;g{syuQR(}H9a}o*!Gjn z^FbsSv_1M~9BlGg80nSq_r}A7AG39S`*-QZG!LcUv%Pb2|K#EELuY4LkZ9kGkZQjk z(1s$fY6phYoAI@3Grl&@jITuk)tT|NJIIXxoR(OF7w`}ErkI48Tlf8F(mbIMpU-Y4 z&hSIOcS^{v!#s-B9Hc2?^MnlM;pqLR^lGIc8tB)KQ|v-BQuFN?xj7=iu+QXilQF@l z%`%H(R*x}7Y>vomvv9Qj4fsXvY26+YANv(;_{A8b%5Mid!kvJD1Xwpp7quo(P^v(s zcvOCSjtMvlsQwHN*z_tx7+%YnjK3MVP8slY_4*~okPmlTIU#}}D#70w6%+2!4T!H84H*UZK zB7qf~^<^q?Xx~8>jEpc!2Pb^7o)06?um)OPKwK~xk(gkVUgCmfEEw(h&SL=+oW%kz zJ;w!H8s&sq7UbHhS-=DrxgeE$xx-Z~R z5*v)JNSVOs3e1Yybm;AwW`nA-#sniHjM8%|0v{W(!RmtQijc!dOfb11@t~?A@Uf8# ze&a@iE@h7^g5QRm>;Mm|hOn(9)&*HOvSERSz=$7FYD59Ub-RR^VKykSV}XXHF2T8? zB5h8bJk1tb+LJ^hL(KA(6+|2#vBU0)nhGL7Nz5?1B4rM(f{5cISNxM34Lb1!xntPG zpwsOkdwi~L=0fqvjzu~GL%y^G$Mw60n6kjE#Fj-mmb(V$l7DuisbujG_G~Bd^K-*B zmXEwyrX+ah#}*+$%vtCjqRtY_5_gsez@ORu>R8OhzP57^RoE0JA{c#{dqU_9A#eUKWcCFv$kv@jiw!E{eX*bc($Jeu}(F z)^0kE`k&*6rSOgMw10u_e3e9Ev}el&b8+IFG**+u1b3|Gn@SO$2{}b~ww%pW9M%7d zALq(KG@ZrhzML!&BRW$m82P=h#82Wm0tcfx*F=itEU+n(Go>sh;wb(P{Fr45%FxSA zR{-elZCNNr(;ykl{r*x!Ay61qi^&w(*}zkD4U$%qag_fj{!v8`uyM!jY{uu(VG<24 zA4|nJTSSA^U#JW8WK>jhO=eT5=VMQ?wuqX|#?cnGI>r~erAqJiAIK)0b#C;6l&}lD%7vyEJ$N-aUFdpw?DC45&+f1j}8{ntNn`G^# zL-k@T_vasTARvkW+ML%h^oDQ56o` zvk*;ZF*-Jv1!6>JN(Cd)eS*SRXwJ?xkzzRuY>MPeDT|3Xio)S~a+I_JKu6?dp%_hr zWH1-qCs4*jQMH&%k(~`ZMb{u{H5o@)I9LbH%Q|*+I<82?I9o)6)#yHfHY$p=$!rSs zeC#RK7E!a=INHKiM_c-toT2Wbt%@4a7EEaI- zIWFMROIVO=t7ZWcT;zhZ2V6A~_}It=!eSH#bD9azf>INU=9^?g{LwmG9ryZ7r`Q|d zr}=M^wVRITer+-ICy&J@jEpc!G}Y{Uku3#uzZ4kIza{y^-sY`IKs7RX=mQ1sSmi8pk$Plx9Wd#w( zN9?e>qNaifP!cnYu1J|ft03a|$Q8nU0xD^ff&)5Qo%@cpP(=nnzM6Fq*%@Zn<6<= z%3>mpqHwsL93`y)&=GlAC`QvD8O%lZ36wEWR4pb`WM>0U(KSd~O~z3c4%UJ5vW^{{ zjw@0z&KA*NHM&orjfx^|GMhp@AA5?mMbvCIj<&GX(b?}4)QHBoqpglA`QQb#)lnsn z7+`k=(ZFa*;(}RznJX|Wgyo0e+G*WJ3kY2>GQubwoPbtf6=NedSY1F|Fd31UV3c0s zf@Lho&!wgN1S}Fva25-=^c)v(=_M@4wN+<@YDP^$=XfFbHBD2`V+{>Vz;&!0?`=#s$z)bgBz;)0wj-E zV0T4zU%;UxHW*!zGJ(+*m=(1}(c3f4232K^2}VX3rRP)xJ~m>5)dkfRA%~HeU~)m? zK~+WIV_4zOsUd<0E$1T~Sj(1Sp9aMpvZFp;ZuZeB=t@J^`3|(rzL;@U9h) z>{z5DFa+HvsK#;qt|6u@Fe|ZTk&flA!MQ{@FfY10+nq#5Hr4WxH_Ma+@1Xkx^+*tN z7P^P1v&6E*on=a{!9AEgV#@}OX7iN{Zn`wV$QPq@@C0tz)UgpmtS%sGn2boAFiJ0R z0cL@?j{zJy?M3i{yet+OV3G~S<9!TeToiqq=@fed{1kbUtle}Rb>TiiH~&6CNhC%a z9kLhYg1IX^;%&qWc8Om?)|ilPR*Zfv4yi zB&{anC<_PczBaG5>DgqxH zvBBzs>WYxVNK7!fAn~B8BJi=13xxXw;}k1zg}Lm2j=t-JBO4ZI2#i4Y2`VvMw@Zi_ zW`hzt7HC-N5}Ye4(&mIE(`=!oJxMeKM~LB#P9JM6BgsUQNB#0;Y=Qs&Sqh&VoS zg>at$Og(8g5gmBfibr-V(h(Sf?h{nwxPI3VQx=$&*s@5+a@XKoA{>|(-JR`Dq9dDX z`N*4PN`iOLeS&%nzzCD-5{%pS3214py@N(MJwnqcIMQ95`6w`}Uz zh#^)N5H(CjBu*Hmm$(44K-|Xw4xRQQctKtkiwrQy2IKKQhB7XSzRh%sy#aoTyh+w> zI*z(^xld3SpCUXHa*FP3Ih(0Cs={G=X6ken zqhoVfAVzeiR4@|VCn$`C=ImS(DVDRqrby0|vY3dYC>*XQM@cIHbVObjiqSMk26NGU z0%c4TRg1|K+1bETbPbYLlW~-VgLUA%tYb%~Zp=O46wU`XkauYaltIV%oUgw!tz6K?X+&A z1%xgb8DW$TPCzTLim?$JtS%rfn2bnFFiJ0R!7>))=hD)B0u~7-IEw{bdX5XY^b!{2 z+NxQ=1Q)p=?EzO!1U@!$fv^~b!JK9Sw4l_)qWLD-5P!4|SI50R(<$}__-X!|WbLNo zxnElh{R!k`v0GaVfoP0=RWU^J!3|Y?0g^{7u)CtVFW^uT8;q_T9~-g3>VoQuki$q!Fu5S{psFJ9v5^ae`vl_@D{qCl?0}BG>x3g4 z7H9~JK=%nMF1oh?M|X2n`-&Un`KIZchG%;dL)QB3*AH1Sz=k@&N3y};2z8#v1J2Cv-wH} zH(i=wez@ORu>R8OhzP57^RoE0JA{c#{dqU_9A#eUKWcCFv$kv@jiw! zE{eX*bc($Jeu}(F)^0kEx^SOhHUBX^;%&qWc8Om?)|ilPR*Zfv4yiB&{anC<_PczisR(>*#0IMisw+YcBQe3`g2aQWionN4E)eb$ zj8m+<73Q)7I{K~?j%-+woSfF94OK`5JNShOuOtXcS_9W5B z5VL$`1rf(b?6A9{rh*7i5;KghNSQ;cAmaGQ6~cW2F!iL}M0DU?D<0XgNJn4@x=&Dz zZl zlw5;*Fnh$74IItpD;eB$X@Ze2M(N-Q+_I@-BZgRAK-4f9kvL(LUg84G0&yP$ICR>J z;01YEEHc0(8;r;M7|OUP`Zm)k_6GPV@+Mik={V}beS-b@_X$cOG1};my(kyV#ffv$ zSWObsocjcY@hQSHA*blhmb00PqbeM>XQobPF*-Jv1!6>JN(Cd)eS*SRXwJ?xkzzRu zY>MPeDT|3Xio)S~a+I_JKu6?dp%_hrWH1-qCs4*jQMH&%k(~`ZMb{u{H5o@)I9LbH z%Q|*+I<82?I9o)6)#yHfHY$p=$!rSseC#RK7E!a=INHKi$ASGmL5*mPJKE}~k`G=$ zTOC#Mhyiw25Dko`Brcfcm$?G7LRfwXuASCxw1ChBBO{E`!3k&uRxvhWgVhDZ1(Okp z2}bE9E?CBb{9IbPPrxF<1ZS~;OV4oumtMkxTw65@nBXE8q&?uOiNME3E)W)@FqqR! zfEJXRSTx@x8{&`F;p({8XFA2+06)!tldRozJojshp+A9~EOu*)ArOtxuPTN}KDeQ( zFF^8$1$I|d_XQkEVuR5YDH9l7fmu;o6umvuY*1C!m|$dtQF=~A;A0~;SY1$E5po!b z2__dL9#mBXJ~ncJaGzkDV&$zcmmSd2cb#x#!vYO~5$HZaC5G#E2{FTLP-4dd4NF~u zb45kkoUmk?Ewr>JiAIK)%#LXQ6wDI!i1|+*zjN8r*}~BerbdXf|KT;HFCxjC?Ul2T$OZO&uFC#OeZ~hRKM; z38VB97ho2M`xwBX(_REG$jf4p0VdgCJl@Ap#zoP$nNG1cz)z7k$=XfFQ5WtL9L&E@ zP!fsJMu+T0xnM3%oRh|Cl9=Y)Cn$_h5uOP-MR&HG%~Tv!;jle3bvldDvAHY|BRW$m z7>Vu^6vjexcCLvO%UNJkBxg!lOvF(X4%d^Tq!j=QILgAoI&farv7^&*MJmSGA{wkl_X)I7QKU^~Q>f=-PqDU$n$5=17PdML z?e__4L}T31R!5b5@B-TEsFFtvu)Bh2U^FFh!7RVb6_^#m@JpqQD$?eJCDUx7r9DYBGQ=!jSwY0{5j*UzsHq?Vl*A09D^ljr zDu_5fa)of808Bk;HxV6p*NR7WEYcAeg6B#1c+-9yw_Vp-zOG9}mG9?TxGWdldE`AP;iU7BFzi%~jw z0=I1H*oYxk7Z5c}MkG!crI)w>vq0R(01ln@B6vYw7K;op$p+)`K87+bioVTsioF4T zio8kIZaR*-aG&6C{(XXyNQ^c*WG~7Eb8+IFG**+uH0M4+VSI}4Ovov^v*m22;;0IT z?U||5S&WX&Wq}ycnNq<>bf2Iw7MinjO{7@P0-GW^Q_5l@j-qh5o*X5u0MHS6Stv%+ zAQ{X>_X(6SQB*A^Q)FiYPti3@sbZ`P%fmMu+*kE-5alvFnVuDe6i3^soAU~Iu?h~*`Fu_?Y;L>wk zz@?Y4AlFvS0w%b~1!)hsY9jElkqd;yC=BK_6QBj9CKk;%$%go&b+|h2^_fnwH^5Kx z-y~}{9nbyRV(3pGCyU+MVhBWI^s9;?k`HdE>I;xOVu9Ti)qMenlGtE$Mal$5S727u z7DaE*G#gZvH6|DtVU(Uz5%}1M4OSOaSA-l!VuHy9i3e2`fsc(`AlxSyr&xI_%w-33 z^j#+$*|0!EUJEXT|&$-8Pfqa=)k*HJhEevj=&IfpP(AY^}B|cvcRmw zmPI<2y9VbH;lRA;?re7w9obaNN8T(`61;=%6VxL?%vtCjqRtY_5_gs%V787w4 zg~RpaC}{yYGF`5R+U@p2(pp1#4YB8B2I~#b4u0hgjGLEuvunwG;b?oSLT#<@# zwulC+(R~7KR1|5G*%a#e*i)=6qGq#kw14$RgX;d?Zrz1~zxDe1;_A-&(Y^Eg>x(;Q zC-1B;j;my2u3KqHpv*YjIXiz?qB0MeN-!mgJBgx1N zt9Wq44;n{DtT4NQIAOCRF~Tao$PEkGkRMdj1%o;}nB+V*aPe7g;NpwekgKa_1Cw0l zhO{+YH!1k&$PNEzt3hiDo*PKrXns2`{LQMlnaM>L*n|TQ{KsK_#t{}Lgu5UmH5E~Y_qrPw8U=k~=?ns%z>JH3~>VoM_ znr4T(O2H%}GpypXDg_@MvBK&;g_LSLpyM&4M(gExe>P8}UF#q0*6h|P|~4XgMfH()l%`yjxp(_Teb z%a`RM6Kul40m43rGBRR;(R|{7#e&2DoA3hjarFOUt7&l(p0qW27C4wI6Qe#~FqkZ4 zQyC>(vlz^!NY6!`qC8*FXfBTKf7@zyDTf}!aCr|TgnaUJ1Yl1Qt;7{8>EFSjOjEpz_q9*7hPZz4m^PCcy;7!G@m$Nu^{CGoA3hj zaW1GXl>W#wU(eUoIUpNZU{?@aQP(#h1c(g_+)>{*a4?A#R(GV#V08y(M|HvUCQY+L zUBzRPkr`I;S(Sp1j#y!KLw%*lfh0!Q+>rQCS1I`D$PLoL{W$MV_FwXB0Z-lQq$4Yq zs0hq}_ZcctT)SI{A$FS*OO~iu>=v9mYEtQhEz|6wW(tQ^Q_OB4irDN(+^~u-asy_Aybl7rI_*`2wR~AFGQlPs z93bq2C?g{l7|kaRSS&~kun8|PA4gw0;NP2`DCb!Kp3|3PV$|mg29t$sI%&QpDB($& z!CZ>;T+}Jb^97CO;^;~T{aK`@vl^b*%MvlNbH#$8@IFIfG&F~2n@KU9hc`uZu9(S8 z98Kw9KY2@9F~HOLvQ&(!MKqWU?=vW)qUf5;rU=i+o}z3KHJgp2Egk9u4`v-cJeXI+ zV!TbV!E$(?K^qrE+-5q3+yFnt+$3u^9Y_67x0+Ws_sJfgpP#u`EZko|x?JCx6gFq( zCzP+mXsX`}@Hcgmj+@lT`q=uctbPXH-oI2x6V=Bks$=uhIP(AHZoH#bS*o#XSz z+s`#kbMxN$qkAV;cXx07VDt3l$=B8wm)oD+{!H`f<~jP;ozI^?x^#cte*flQUv$aq z>h$WJySqCtY(BVHUtXPGtdCF5&OY{@Pj0QA`0Pi%aCqa^hnlC)&Mz<5m)q}eo^PI| z|9$62)+dkLFm3ap2WRK6pPU_EUacP=6Bvi(_}bG5WJ(<9#Rd_nK!% zA8tR~eycfco*#YW0B(=|8Kd;-__Hs&k?zoce&Xcf^yKxk^_L#s-TmZN`^xxBAW*kH z+B|c5wZ1rUhF)&J(7dmC?u+4{KhZjHpJot^5B$jb;=$S(?Ed=| z|0nmofB)&`{SVzscFvB!764ps|Lae0|Eo_o=gr49*=7Ck_4WPxr;i?7-kJ7&&b*u7 zdwF^Kz&UW|ip*A8j>g zeOJ_;`)l7pL=Q{rN;Hg4v%~mnTg_kh6qvlTDP0(6)Oj`r^1vbUb!_^&CqXpA`LGmJ zr_bT}^{wW$Iy})+3DBVfgr2Ux1| zlB|lzDU`X^&fwbqk-#-oiH4Ch%Umm}&;56>D{xKMqM;2muiHgxzP$91s)ieJ|B~;gR6_+(4G-y)xdxQxzB-a>s3fZ*@-@`O5pk`r;M#Ms-`JIksH!D`{D{@&s^6dp`NW5`If-#NrhC9RS}Uw=ecjK~_aX3T4g#t;}Eh($YZ^)-m}GMBv=v9M0`aiw9MU zj?&LofpbOu+=tTgL06;W^fOrC+@a3718D&v$f}4)q0Bj;mHF#HT0%&|I;P?pqWH^O zVIL{=znjH9s;#OP6=hiG3!E$JXFl94`B81v)#x~jYY6a53AqoCq&0*ft0E$WGUtF+ z=C31Z4Iv5Zn9>>o?clqYFq~Xn(c#tkBXN?Z%}bi2D44pI2vgdj?$R1UeJelPqK-~X zFlpYxYo|qHH)>RzX$^t)_R%XXvJK!(e}bAPNmfHd3O(ok1~_d%T62GGdJ%swVtW1| z3Tv1`2jXfg$b@ z>Yc}HJUUT7Y>y2m z;7~}z;F$5lR{Pf%F zi+93>x6u*d9(f(*&7y)B0;#}~RySpT3%~YFP)*l<#>i-0h;9WJ#8q5=X;_PEc;Uqu z7cHi%a1G)ruC$0oWtACBii5Y}l?>X5Kp9SyjgX{7;tPE(yZi&+{yOE@SCrDK*fmxV zp^5DBNv<}E!^7xfWR&izqBw}FxP0RKy{!iQwBR@rp{pEULUq_;EUBYJ7o`;q5uOxL z-%S3Fo-ZcoGM-UWt_76IYsh>%{_TrN;?iicT9D%FH$}u*0iNOF^OYJUJS#-;7+i6- zdR*?P3+FbVIIje>;pFxG1I9_^{Iq_1Oru)eN?EBOuHq_YC8#*}mXwu5VHK0lO11nNw@f2% zZ4POx_%&9bVG9SfwQ(GprV+BuiFOsg25}8nIMn`q+;fdXG3iQJYt-E#H?Z3<&xM=NnWJ4A6U!lbjXb#@0R z&Y0KHDjfA*dy=q<$*;kM-Yad$ja#Tu4(X^Cpxz4#s#ts#5PGjnW$qcu|ETu@Vk!=w z`GwvqDHP}5yXK;|kVnVYpG&>h0kH?=(V_w^dZbnEy#&=6n+}LgX!aMxRa~A`Ld|B1 zVefT7b+HQ9AgQg2@nS8@5?Ug*8XiYKiJt9%Dk`B3krs8I3w>Q@od3BH_bQ~|Y^G6gC&Uj>EU z3m7iu{%=3HQqI1h$6*XSl=k#n)a7__EyAQz?!BPmoO#lwz9_6=I`DfuLhqFxj1hXT zL#hR+_kw~N7SiVYBlKRG%FHu|l>brh1;kVwKJyE`S5hd>zjw`pPeL9YKMqvty^bi$ zqdZzvphb_g%DtDMI%CrjWq8!%6vS0rKEDgS7gIbmPotWBBxR+7xQeTom7wCxTZd9s z5`{HPq|wE&mS5x6Y3wb^PF4IGE6}jf6}cJwg1y(~=3f=R25}Xa&ssw7H4eq3HQ|+y zcB4`6rKr&G9r~WNBBm33IoGIb)ndvNsMvhpD)e5!a5?wWstw(XRh6h1{n|~q)K*lV z-+nrzMK8M4rfX4A7FWKV+&qovEy2UHLu}nNTt^<4u)>76GUgpoWYG#!6jm|$i2e#b zb4(A4X1!Osg1t+{05xAwGJ=sG-Sl-p=)N+Yxo@ZfpzaHZMIh3d+L6!wLid#viZk$C z^C+;pRe~;6j-QT2-Pgb$O7tj|7B$51$vpCT-rXueKkzB&4tyH0Nt+V|c^#S0^0mk& z9oT?ON=^*fAg?3y+37d&jR>fjc}z-AqO^|6=cihpjhiT!&6CIWC{I=KY^*^=M`~Rk zRvXQsks4#$rNo6w4fEO{uOcI*4w<{Xgl>I}HBTCq)QKg}?NS4aS}{e3j?nk86_K4F zWSpZ#w-#BZK}Y91S>e$*U^}Y?g9m9LhuSe!i;mK-=l%pgWKh%}`>_Egt%OlSrmN9$ z`t`6lPsEmLJWn>4g7G{N93-YjX9=i->I(P|caXG3o_!A2b|syPC-s_fqTEGG13 zW6hJkg;&JQm0;AHDLPc&kz(h&ipWk7GR{$Ft3{S+(9!wcR_M)u?Q%W%_N(*u=*`rn zc>+2h&cvQ}aPd45<<&ToQin{;dO&l|z@FCdIO_@0Ix4?{7kabwZa%3u>#07V-V93W zXnbW5db3O?^A6Pn)SCf09f_|7LT{Ed%5(AE^a@5HvyR_NSnSPK(tbu!gBn9pE9U}1 zch0Jnw4YIs*OB?0FZ5&S{(3TkE^ld_a3t)uc;s+MQt&T8EMtSDQd z3eJ#?HR$Mk)~byr?#)(|y->{&u#wkwWIl@tz1djvq;KIBaYbD%>dh1#Izr#qRz!A! zka3PWTP?CogO1MkwnA?PY%kOUTH~R8GgXU93-YjX9=iOz^ds6mY(sg-*(L3hro z18F~_Ag?3yIbX<9Of&AyHb3x#vQ$A{M^?;IP;>6F%|m7=ONr7ds_?T&wLBYlR^#4m z^MGm<&&C>5bfk6$&v0*cK-sG*nuENK%x5v7Hyg)e(zk$$NL?-J%@iFf!hPS@Rz!A! zka3PWTP?CogO1MkwnA?PY%kOUTH~SKOx2>J^ea7~H!JFAz1imVr|3*mSEJ)Bu9Xi_ z2gI4Uc`P5#6MJtKp2`>Y%^QEvw1R3zcaRG~LZ8s)k8ZhED^kXgrXxi9u+^l%o;tVIoK48t>D zLI))1&RKOR?PnC^b!0y0*CNBc+2%Pal%*K5L0(7Z^QVxdpl0SVszNABiPAbMpQUPf zHtwv(z1ijgR#d?mvatpg9jTRjGeI}=8fCAlXwK-(!jr5*Z#G6Y>05Y3+&pTJdNW0b z>bvmNeMMv^2$}b&v(+NYH0bDjZ!7d>z;?MFeETO$_UPwOO7jHtFr0~-$M*3&5#`l5 zho|<%tOqpb4BR}lkF%a2t)ucQc%e5-zlkFCX7tc0+&6=gIvQUYgx)OE$-F}~0rh4; zPDkRafzX>Jjq+T4H~sXD%dCi%?$=F*pT0rOS$w(h<_-b;iIAd(7)0qg+i;7(bwKFn z{UmgORtXzree{F0C3zK@&-vBJhUUt(8VMPlGu0s*(wuvYb{f%h zfwHuU%4ex+o{d|pvBxM|)$wesK}F}YR&_Lo=4yUy*9#xrliXXdHRJp}ZI0>?NYsi7ft&PT2{`&0#*%sVTp z3Fx;Fax5gLBJtJ0^=7Es8#E3q&wj4!-SiW2Qf3{$>blgMZSD}jyi(Ml#W36=kaB^b zJ7*PTeV7Xh@*1*mlb{;e&|8hF`R4Xi9kM}QMOMyI%x2J=ZEjA%EG0^-sC<^H=GnNj z8hLE<`@?lS8*9+eky^hW0M`J_>Y=w9W25X<7tKLlMdq`Z)SHcwfowfOx497tdox9c zhA`X-t%>XeVLnT#vsEL@G^pr&Z!7g?z;>}7eETObkLiMJah`x4hIJg9TU5e45#_Zw zhr3i#ZzjHze>6ijH>!lJ2XNK&Af(dyOX|(i>+7?Pv)E?Sb%7n&n?Xs{N4_#hy;-I+ zo1gpCxx(HI$f-ztHIRCD@U;M3z^iZ^np- zuPY;QN7V<-%UngE!Y~wZ;iLF_`anMRKi$0Gllk%2*QXEOyt;HV@ule}cfR-X^7O%@ zo&4jA^~wFayX|Y)x4-zv{e4nDq@xjIU+mDa$uOD9? zx3~V$)6H}4yZ4XXf3DU)bG7~RPjCO<_cXupWV7?&?EH17_VQ}|_zrz#^Pl6*m-ycv zUpTj%IuGuAn8{s!{q*YIo5xpocR#ju`_FDP{2O0%{JT26dgtzL^BNQXva|CJ{pTl~ zBj|^Jzx4RdS2il|?lqsN`MOhEKn(B?!&P)6!Cc1@T<2BZ~xvryo6#GQY4nL%=y3fC&{9SI%$Q+ZB?CgX<(Z4mm~axNeV@XQKDYGnZSp1+j&phRjsT-Jg{h8dr(2xlgND36yG0uGG4nJV~ z2acbe_|OHtWuu9!zGZJ+dhR8O6(^HZe|VTI8RN3Qm1Nrhcn!s_rbr@XNDZx%w$STH zE3;jkA-X6vNqgN)(g%_Xk98z|)&q$%;W+@p97pT`H#>1i5MrrmH$49|&I*jCsb_vm zl7%UjQLYQ4A{!WGQ*0(hjFjQ-RFi*wc!hr+ZPfK@xCYVfi8Ugl_5?}OaT##Xra9r8 zN0;k5N#=w+Q9|J7zU_3MTW(}GIp)<)I~in-gUVhGIuTJYb=S4y&jndw^3L`H2_#Dx zOeoidNsSFmiYaynMYd9g3(8i{YvNTTT~R6Tf83xyr9UI;q(5QSKHuSUm_AI`)A6DX zp8|1(N1YU}EKy!irYj#v8uck)xm;5>hjfyHN!y|_Jd4GZ@>A^aLul}*u|d(WL+EzU zubE&gYHD9xRYtO(8Px2vK}`w8?xe^z%8(YEO*iV-DwfDDO4EbUW<4l69I{T86THm3 zG_1m-?<;9+?u7OUy>2UJB}wZ2Ok%ulo$7=R#ZIp7`m|id(#S@AExlH;k~M1a&Q5hM zo$2|(#9_zHFu`Wn&98IUUopt7uvxbL(yz!$7dBS5*Z^T8#imhYEoIoN+ZxrRdf$uV zmPU4}+v#+AqtEHc`F`0tw7$7cwNcxo+7-fWke-^E2j+I4+psV10I&0)R(ybYhI>em z%`$p-d}VEz6(86B^=1*tcVT?g+Tf#%VoNF5p&2eo>*PJ2I~5K3n_v6m&F|NKqj}CB zHoVj@=M*T7`Le@wwWdw8*sn%ZBf9F@s80w4g)`;jDSHCCqFXgxOoRTxw@T+6eC?FJ zQQuA4nHK3Hb*yZKq+HqKxgfAXpY($z-#G<$eeOGas2UV@Z1juGhhc+|AnZ^lHbrjq zvtq|IJK`c)|IDbQG8?3Aq1eq7$)OA_(i?LwDjW5mN34wbFf3oZ#0w-{Z=>(6=w7!{-}2BUkt1X7K3YcsxOjBa2pmn1+<(>GDoAR-QvG+b zXbb^&eZPQ6GQ}oQBp&tnTG4!&QM--Kfi31HdLh~8)it|DbVD95a>ASRtamwc5$=3m8tMYtCZt7jQpV1*r_q<&@0Aqb<*A9bQ+!b z<+|j+KDS5zIa3r41}6W2%;T#&Y$NGY@SLO>A@Vcz&fYKs^&YOgZ;BrsdPC_|F1%NNzsHKHp%D}HikPVBGl%z59d5A3HNuS7|U z^x@cD^2;H?@*nsd{)6>ttCx6L!T0lKGM6N|u%B+bjmFBPSR+N&Q-*j&C!I&%CFn6J z)ofz%n8U$(^-4A)lC#LGJbH~+9o@)+m76H4qmu(~c{)=LnFYsSSAE!|4$nmV!e}59 zrlap$y{}p^^71k28{}hQ0?Kt$F32q#L|viSOB89Q3^}@Vx{FC+4r??W@f-Xt�>i6brcmm5_+^E%c{jaJL|+Qhqq5MS68c5>ukAQ$Gj-(?hy$rb&8!Vtx1` zzv2)`0*JI6b(2o|_BSpno0xl~mc~0w(uxT+0aX*~!!Jd)kVbtcLwyZaT@l)4TpQ-U_4d zZTZC;WWzHfX+E%d8wx2lpCY-qrw5-@)ZJ*aLTZB#y_!Al3&3VQP(*mxq&n@l<2kO?XF)I;x}g8G#nF&>O4n(ewBa8P)ISNmgFUx*>b)DZ6)tyQ z4m?U0j3F5|&v_NaCQxJrWiSS&`nJ)P(khZ1h{*#5GV|u4YQ~U7g5V$T;L{X5?@(_L zBtu-+a%_tIk5O(`P!xXt#$QRS3-6m?o7B%+ok+1ODY6{%U-pHVm|DW+%Q7MntRv>j zvjdSsShc%nBt8nms`m?qf49%1bJbjP!jl`9>Dk2;C@{GokFIw&YBM6MHQZ4ebE9Hq zd7h=-YgGF}joMwB4jQs@D?{!?;MM0zW%xBwJS5`g(I@4pq1-i5fufbcs~ZDh`whuc z-Ycfn(Q!}SnFw7{hhXD?S%jAe{q84SF9ffcYzp+I>^b1YY(jj;;kqO{2a+ghfOfLbB;PHID4`sfUV|9W&zo6R&F2 zpC#R%SY(tR)Tp6c6FiQRCRMDAix++F+p6tUrbLzm zrLtIt#;ppDS6uQ*r{~MKwusoU0;N&AOxhIQF}-7Y4*ft4;;BZ2fN7DUqi4RXn#K{6 zDpAs;>hUb5E5f;0S-TwZR2;pMT^gAV^;JVN;f4=Hj|nv4$mw^-@Bg6s%MYy}%Ak6_ zOO6i|#dKjA^N9^qE>LU}MH(T*;E|~)9B-1=OS{A+0kN}}cZV*NVbMSe9QCTZ1#4EIW}U5ZA1jdv^f2I{4sgd|Be1a1t$^M|6g zMPWe!dL4@CP6-x1Egzc>+pYfdVhKXKy#GCkcd^D~7ugkb!7pBvJuy!9@R|+(x#)$0ul^>J%At;YE$d~x zEQPs0tKIJsU93+HxiS&Wo@?sWW80*S`pqB9`M0zl*-9zQrMRKxlBwGgn+GIxKbB%!#iZJbRe6 zqR{h;yXVMSVJLK6olJJw=(#+Kg=+syV5hLu#jO*Sg`AvOCp|;*M5&>78O)=BBpXk2 zp2a1fO1f9m&J;(tGI$}^3i&Zb!Ik0dbaVJk?*f?<<(?zbjmOG`Vdu%Y0{zBZUn^36 zdScWmlIFrT&K{e!FQC{w3dkM|3)qh*E{6=8gPJKD^n!8Alj}uK6R()kh}AYpP?G03w?=ySorA~V-pNU0rT~Ynu#_Ij{qD4r}itGN%2#+B8@c9mp z?4UMv?Ud4)$CXF{vQ$zj-xmsj*1aJ+#i)$9#ga^5aVSu%@YthF6c>sxRm$Q!gsTc< zO{(4CIQ-`mM7O^2`d94>S%iSah3#gZdYlB`I?X=i+{?=QF`J_6!tc^^8}&Q#bUV6AbwODKH3SP}sqAsTHhH7IS-wcJk9pv`(eJ>BQj$4}1M@Cr;3ad)C_fGc za3)immaMLP76rx?pWD-8{nj%aS{03=eEr8hC^npDDDbbPgFxY*z)5HLfl_w=!S|lC z3XGt)D&~@PV+hcw^m{KiQ)~_em9yr0jaXTwJRuOvGj@ua*lZm>do#l&CLE?Pg_CwF z`rK+%r)FY72X0&X+%N>RGG_fO?)bQa(QQMF$3F4-IQaB=dgeVy#c!vZ`M{X7b-cf(~{)S)v%qmV=fAGEUktSi@RTnOb{M<$=-lW)T6fseT zZg8Tt>5|0>kU4x@*%y(*6v$vNdzswyGSQF9AzPU;@b$#b$_C$ktxxPMsB**!q1tio!d1Zqrg-HuWezX>;r3x~KZi52O0>!C`i4t@+ppO9(Hh zw~G02K1gix2--mn4+z!Gh4*H>W<}e-Ty*au=Uv#^?XkhnClq^~B3CKH&FSYw2P8EC zmzl#fR@5E}IHN*jl*4!_dviJlVY)QC#Ybn?&qfxh!ss4xx<|IsMB){@f_kCqwnw}t ztWjScP$=s5+^xvh_J-DoV#5+A7e=r4-l9FIs#AA+u9sb%d{c5vnk6nb&xnKoJ>p}& z8{9W|W6pc}#7zHfQR_sX03({oA8Oy=TOR&0bVh#SxkdSb?v~`YTV!+R_PQ+$PJI4% z4eAu;_|%1A=S9244f-vzkLl{DQ~CinX)XYu0Rshsgd6qlulM}i-hdY3h;d!j&Q}gK7di+C9&PianTK|cU#4`K~ULflPi`>E;VS+f9#d?Cs7?!tLbH8TY%Yi! z9Jq#t8Xhp&4-W!B{HPgdj~HhN|Jp<=lJ5Rgy@YHMW@fvvCaSispu%sP=A^g9C>|V=co%s%A;$Y9gE&#;xAZ6W`*$Jciwnn*co;@LQv+cl zu=cK}lE1S$yxQ}U@+euOh^79R{o>{Cy#MVRKJCHf|J69>T-f>0)1*;V?32rGla~&? z5d7(jGdPSqecqv7QIvIGXWz_32u57k&f#f&av8|cQkXly{2VXZ4{>O&9->Vqm4$6) zk=p1rZ$xcfA;1{wVhg!{k7&gemFVQ>_U)4SBu&*-Q?{AmcpN8O4jTl)IU{$?>z&JdLtU&v?;E zsMU3Yo$>Y^W0fg-({DL4e+;zW z>mRI4qS$zfyiXa5ff}WE(s@xa>`}o*-s9N-v|hW_PJP1$M!#e4-N5%cb5xL?ab3Ia zPU(K-Q0Tv1cq`1)VPJD)ou*EnG_E@$9=PJGCU6fkDr3Vc<-@#}eL4UggEtyXPoHZz zuh9MJcYEKl%8K9A|9mp}#D$%0%Wdk1?o%uT8}90QxKUwK(sAiK-}^&0%#bcjl( z1ur1!^a|GU8!M%^J)znk?v~AeZmHx_w96I_36qx!p8)@%&?369`O8C#*dWT$Rni;W zzlewr&k(Hug2WW>jjD$s%UklT$}hxaiVh(C*dCPxG<~_UcBi(z{f97dTtYqE){Ts; zl6@9_`QNP)%6L=ufSht+3Dsp|W|=6qg(4RrS5u4CVo0W(HyOPrSAf1FJ*--@h%Jy+ z_;cC1MqqY6Px=odxA?)(ElMOoJrB*1J4vT9UQr#o6s%xOE`!X7u80z4oAj*w4Db=9 zFy_H}^@d3F0VC5FnaZP@!N!hLCzx#4f*IF2(~{ECW^+0Jm_y&arXA9w3o?# z7slCT8=QSivA`*Fj51i*#F4tUGGv#cNtMFfnED|UrL`)1yjr!pXCm*=O7@g`&E%pP z%f?niZB?Oo?WE=UGfF6LL>C_hBm3NTh|hznZ1LI?mL60WnJB@yW~CU9G06+VP$sDO z&WJ>yv`E{(jbpLXPfNGSPfOR!&aud{g)EC0RX{ciJkJjw`(?9!9w}7ud_C{5FjQTb znO)&;MU`&lvMRFCg;BNF231=r7N|_}C za<5xAxiP)e`yA6X8A~LQStdhLKe3yn>W$hC|JLbvR|{R|+2B_v&k|=nL5}b~a)(X~ z87~ffi|~sC;f0!AV%+-AwNJCwU84$H>W%6LdLL*y)jQ;c6#gezZU|qY{&zTzqW*U| z|M->v*}mw@Ua?XSnQ(dH)!ycsUTnlu0A-35F^RGV0jJno(dG?04vpNBJHFaxE5-TC zZ2bm7GGlb-u6xhDY1J{;XC_P{*Iii0EVbFge@?M?Dbj`HcR7%)(Co1nw(ppW3&jv@ zuZCOW-%h8g7DHd^DxgUG#Iw%7UCmw5oRxnh&x1|W4L@!ZYx2jhhCefYC z2I9PYXbDxIxDT`rFeUa>NvD@aaxES$9vpr(rbrY((G{GGlg*c{@rPdyKQ%4*7TlYy z!=Gwt$AbSZEL2BAzfK=B%&W#e8yoo1+cy8?y)Y~G9(F>0Mk3)neUl$1k#pUijh-pY zs_-=6A%N7s0~+XZkmi*hQ0~t?#2)5sHGm;~P{kpIBQnp%hwp_BQYYyQ$?fRd(cq;T zXVh^f|89VHqvry%^QTM7rKwf{r^qmgLV;4wp$*w^h{CC_9Lu#Qzj{0;)P+6IJj~Nb z1Ot#jKhI9l8Q4i7=jO;MwMR?i(852ue8i6U)!WtiraSh8c|5#ySD~kLi*)grwf}F~ ziyL29GA3WwIsLS}TC;xEB6k1yO@WKYe8ON`k(K(+i8ZPvV@7BJ9D3$9uh_^ucK&BR zK<1?Dx>bJ~oVj5rnO80>O?Wb|(&>-1Ez)G5zCmt}P5vKfGTg6`7T}ORu0&I1gmC>3 zrHmA-PQ$PBj{vC=ko38#nF9k=nTB3 z-r#o#h^_D}{0Zkc4*I7bjmFS*2lK6~~r&I2X{dxde&0(d}+;c1W#y=7lqB_mkx= z9PHU_qX5$=b}dDcu}ambz053?ZHYpL_KiN3FKwC$FMFGZnRj`oXgPESZjmNM7kG>q zOdJEpqxaphr5*ZJg1|9K^Xk_(y6i}H4bKze z4`$@DaZU(I3=?Xeh=9SIyj1jymM^>B%BAm4t5W3wr`R=hD>Gl#tgP@bNw;Z`vawfl zO??Xr20jj0<+K#H0iL_C^N&Y`QIVX+7^F6-4@niIYoy$hqb_q`74j_d9#k7=t_X=>i8NIiktgL` z+PHk($KLJmHQ8W+83R{F&YL%{)29ktJ?#Tk*a`#MkTnp`1dpRt5YR>yhyO|zVFg-d zVD=ia&4q#Xkqyu)DHif`_fQ5*kK!I?p*(CrfpvbXHMwxyydxa+@A3;o=tPhnDtD2c)SRSU?3GG8_cL}n$pRw44!D6*y9bC zl}ovr+1unpvc{bWpvPh1_=qq7O*`__PWv)!K^|mi1xY;KHT>!=!`r`q?TxwbDwj~~ zLW;yXIqLQF=fTHkjP%4HaqqOOQ6Qs1zu)gtcDH*x!}J}02)oFE%;-LA`k(O-S4ace$nz(kL(jax9XwA z8y8F<7lrMiT-YD{z(#9*MzPSt_&R0S8kHz%H&>+P(Ve7;#iUp>_XCQReIb~B3k9xK zWZBqmvi!M4etm9WuN0b#CD(YCD7VNqv8e0tbOF@SegV_nqBm%B#a+==<{K>4wS~at zyWX!N3R>o15iSx`W7%vWbj0o^TVxkR#hPwaixkf-4($uV-C0ExPI$>@qfe(A?I7g3 zPB$m}VE1)q{S+m23)wF6C}dB!@9-yC&8I^toA5uCeDiR;ne5jS!XGtGg_} z%A5(wh4Qy!q=x}cXwfO4PAv7!bq50-IiIVgwX`uX)%QYBJB_w>1#pl+g#s)xUV_0m zoFXGKOIbiHBjyt9N@#F>LVN7WaQFkZk2_7)LJm3bSHt~1Oq(=Af?eBphjh|ap?8!Sk-6^A zMy~j%1M~j!_3>5+es9Y!-XI&skPmF6MIpsP%rqB6q+LvhJJ368q0jJUy2tBMWGiz` zov-YR;AD#Snydfv(NG$Cx$eIV+JyA?GbU4P5=G)E1Mc&iRPTG4X?l@xK9B%&n%^$(M!0^B-vwpa%ncz4 zU(FS7S3G8o41ECvz~Bi+%g}#^{S1PH95wX;-J-Xu9#iD_G_u!)t)H_t(%~q@9;V1a z%5ZT?rlM|U>T|K6j6qHGxr*%NT4)H8OvGv7N1ZTN+1V4OI!x6 zD!KC{4;rk08c__@-j^Vf1-s?wI(kjyt`MX=vNT1oG(N^SM??4>sNGIsnpCaKVdX)v zf}WN-F~GX<4mk)kUY%oig7nI58QhBO&_aJX?VSK9!V7@R9rX#XSa<|$*0&{ieyoS6E9s< zE(|M$KRfC9`g-jJ*>Y8@rc!=5rb}ETIw+}{P_0=pu2GLyEDY;=5i5fiPA>Dz^u$h_ zMWXEzJnu~K2*HY-#Biz-^V+iTtr#m-*5+w4Na-_!l?EHE9HQ9$6eOb?GIi&ttR9Ol z*=pTE?=m_+D387iUcd}V)#MzXO{%MU%oI#vjOX>7aqKFW!n970 zlP#IPBV;|(IdLCEJL}Zx^jUdR_eDl=O?Fu;ulx7XS4ZhrIza0UPo17x)7+)*8 zA;;!{mQ>Iy5w$=ryGYdRfd&fJLm9Qs_tOm)BSS;L)9>#2GgI+vtFRgU;NK?_)0lyb zV%JT9SZEV>x@A|mYORT++kP`w%_I1B8+T}+`WU6B}rl%giO#6250 z9P>rUp8f|`+{L|9euo@$;l1}w8@1I;vF9jq1|wcZt)(n*C-7x%2`PmA6YRv* z%YZ}*D|%aCJ@dx*K=WBkZ(Ahy>==IlV(RI1aZI0E6LUKn8>23tQ59-wvx}YPfXRA= z8knbLdz8C^u0hd1BpS+#C#;>oHIhaG5yuJK$37qXES?mn;}!-l3|>6xh>AN-O0?aQ zSigCg6~A0o2yyO=$dxPzz}!frqvbZRwNRyq5!k26y8k^a*+A#95ByjfKwyoIPBSX) zS*l$&_S6R)7z{)PoM@N_1wn-R_5N?lf~_K`_Qh3YB-@2W&^{XxR6?;k zDYA_+G&7r?LzR;+GS29VX(45bY5++mja`(>pmg&rS*Kb$CS^jgK5k5f$I>y~(bvRr zV{Xwc5Ga`2#I%aafUkb{tXn_<`{{ERf)0k}!0pEYTi1a6Zn&{D7}yM`mHt;@vnn@p z_LsfHa~CNIsPI6!k?Y?y`!w5PnAgZWovQ*!RPsj9t5)^$@-gZg$h8#akox!v(%&A*N+B8Ow6L2Yan(Xj6K@R%C;WFr6c`FpT2pT=`x+9wN zB6Gsc$KvBtyGaxCKJ$gROMFoEG4A_K{@B>%pl0Q`KDPs~#WQ~`cRb`|2zJMvht#l% zlC4o#qGL|q!Ch9~_(f_y1UrVZ0kzdI4G#S7{B+E*L-yq3E^FuVw0+tl1F9=z5yo1W z7U><2uIMNtBpR%fp`|uQ1Ntu;^E0Y zx1?yKeBC}Sl~haeCKo75eX&xn&uy^;9X4723O7O17DzO<41ca#JZq^pCbVB=_DSw2 zx%v8!{D&Ud6a+aT;$(rpH8tGD3VGU(nB^Y4#~=> zGm@_8M3L>M251|KtwT@v)=5vzTr(?6xi@H!a%A4=K|IM2Zz# zm6l51lh8XQ-DA1FC~QQaJ_w|bNCD#y=2u4j)G9g&} z?8cdgXMX;;_K42ujrv29Oht=FQIo9u8LyusBETZqo$>Ab za|!|mHU7UoRtzO;#)a(_o~)Vd*%@>hwPgBjRTOMs~5|K3pZnh1JBUDq+Lp7OIW`ywNk)hRp$McB-|K!(hToTVZFee6Sb zaSh_tbL{0G#Kc3qMHpB!ePWi`*S=NgQp&LEA(>0U1;?j+~xjl#S86BgC5zb z4w1X`1LZ0GLecIii@{ggsK2CKt7r&6MQ1Q;{CZ}U!C9jgn;Exw)r)bPTO-*puE+<+ z6UiCT@QXypfYeWe41735*gSqaKahU*hUu&JopD@Nkd)YXL2@Y;!aNyB^NLxGHzkL> zF&8ewz0#{uzeRRLRTOrVT|cYWZMUj0vI6*BlLPCZCk}iTuTa)Yiy<$eUV3+0dhqb$ zz-9#o4+@)UhW!ISCjV)#DjT)xSl!@w>yXJj|2&V%199jXX`)CJjPa_IxgV)+@6I>x{{ zRfCq>r#PcTTA8Dez0j`iVSw;{v)4)5Ooz2VmPa2Nz5?38!_2utsPAV5uS8jaBX^&Z zMRvGw^4oD6C{$By1qCcLh9>476p8F2)zTEE79#vj;9%Y9vv1;uB#}vpER0?jW7M7t z&5t@nwu`T+w~d?cZq#DjEyTN_K@EGjF4GM(Ha0nt{vZpF5EW(~F&t;Vq4>4XoWcZ(z zmnjn13}r=l!uVo+mFk{!S#YB&Rs{qL6R*%$M0-P=tcq-wP7pS1?sV=4Vc2NQ9evoI zI+~}mKpqWbl_=GLqY^6ra_G&GSds%d>v?pu92)dut+8e6JW}f*Xbpzz<#S;91g(MV zs`hWa>$cO|s(XG^v+)dBJy6Y&>u&Q2%Y=(+~PG zen>0m)KT^h`e!7yfXAJ`Ls#tE;jvH_AC5JOr8AE!4@cgZT1a1(td?adTb-~oe&Vt@ z4w)xTGwdf9C;a;QG^_adZH2yu6bjq(xUjF~h>c(=r`R$IIGzo0p$8(%lzAbUir?N= z6+3is3r~i+duq6*AM^D_U%7>Iyd#-O6Q!?4~L?496nZ2esIL9D6+t zItQc!<0iMr^SYGlCda&xK-c!a;Q#~wx(xv){F4sDnIC+ejIt+(;^AOAHFI0k0vT#~ za7Uo`0DBEp(%2NsxzBEZf2O04bAVf&I>fK%2XJe-?km>yvonwN_Y-22xUf+RHf%qm zb``}YP-F!bIiG|g^a4e0q)}Jl0W{1PltnXd_*MKbWEnY#ZKvXRpvmX_@V)oHY1Jr) zesTU2a^RWKC@nUiXrS17iW~>en<`P31F8Z!olVj<4R^(7i)`z-{o-u;Tv$s~oTx~2 zE|kN8t25U57BNL(TzTjcXlT=^`7sk5^>RUM7VSAi67}g~X1p z(;#8^`pC=SSLl0Q$O#ii<T!pCl(@WT-eU#>9SlmvsH5_pw(w5WVxDr&w#y)9w7`M9wR2zK~eX( zO8HgKO1U$FgnY{&5FLEe4EwlgUYw78UL#L^RkE&4x7jCIx7<4+96};ULbB6mZ%7yP zHk_9|ACTQkm29pn@0mfR7AXJsq7{b=Ya*s*w_TW zECX8k9nl$eHIiadF0G^!JWD4dO)vUrmU>@Mc22{Y@IK+i^Fx8vV>o-_fH0tSocxcl z?^(^^7x#Dlgd7<|uG!d;=P33JMH(mrQpG|mU(AN=_QbqMFcVj@)!|iSDfBr+{wofP zIZJ()dM8Q}0vCzyXxbrFq@AgW!6Rsww=Jq2I{!c+r&nd5r4fu#*Ulij8@EfdNWH?_ zGH(@GISw}GZ4poj4*vN(I#FVtpws7l^}H5oJh>K?W9HXtQeoXXCfvj2mg3P<6B8v{ zWJNycB6+%_e&klgcFCT$M$d?P3z+uPvqLe8U0A*G7{%RWuNGTCnmdysNgPrYECMct zQeW&%WbVR+LnuU?s>b>uOfbm%-RsQ5#vjQ&HCEA+KK{*jNQ$tXxC?KYAp5DGdnSis zA>TZmGIaQzQ0I#dNLoeJ;*Vl#>BID=P!MuSd0<*6X{WJcM6X*yV2Qlf&E$2$tHR@g zvW4J1>A?x&f=SNCi(s5QlxfZc3C!Ko!EXKM;&{*Ts#ll#HfVts^|=fQGP*yA*eT)~ zwZPxQT!CEPzg%HT0)XfJjO92jc>kP@KA5>VFP*-q%m4=WT=A0WhoTxa-0kR0vXX6O z5=C|NBDK*Ad-0)B6fa5{TSyHPrUnh2hY)(~@8Q8z&ClLZ*mEMdY}OMGSp#yP-lDrb z*Mmo{Ql1b9UD`xQZGoH@+r`cD9G!{&TvMP(o>>{bQ&T>3q=>-=vF5|Gs85+inC;{{ zgW7D55Fvg$7mg9|;0brsO{!*jyB9cuyFJTh-cbSzh2%)|5_QiJT88xc@Ed0y&Zi-w=xplFBKkm)T>Fd^dv74$IU{X+wO72 z8Yi#(VAtK*D=$#jTso2bWv@7_3f-fu3}4}k*2kjAonkYaWXf14um3=j+%ZfHbfX%$ii&jpryj}-B;7^DI2FjvtuDCdiAN1cpm z1(#Z*e*U-&caY1#OAsGwf&!qFy|3z zznMI=QQIPI&^i&oybyE<*}QiA?Za!|ENRpEhYMn-MCh`UT~LVPcU;Vk_!)+)A( z6DMU&uY*)s3XdOo z=o{tnO}FpJtmyk~@z_6+6~a9GF6>k(u<`6~px9Kj7tGydR)@DlAh$Z##gtFCG#J7R zhe~-h-3uPsOt92W($$a+H&OykFqj4qJI)xI?lI-}qpk2*v^4A>w9OHOFBgUf(2n*C z{3KB93W_YF40lPT9Ggpl%RXN=vgMIB3+MHW8Sa%VxGR^CcGpxir;YEoNO5$W}n-*6!$91vZlAuK8T^SvhIt zBug?Ci!DM=BJKGjJ-hJ*HOi@Ke6%5NQo~5QJcP zIL5ye4!w}hTQ7&cp7$a_>*c!p@BqWjzuE7@c^a=^nxlE^dXEuGPc*b)PJB#bZ% zr~n>N-oUB!PafBAp?#;U;Lmc`E`?Jlgq-nZFrJ)qbgirll1^`K(06cq0C z*y+J-ktNRD6?GyMx&~FNO#bzd0l0C(s$fim#4GnrY?fnjNQFO#RU-A!$G5^C8)>c+ zEmD6Z=i2B>F*Y44hNhf&m1TBJi>g;+tC@HOrd{R*ifu^t~@xrE?Be!RIDCHbhp$~5up}&R6%eSWHXV}Ljg=9Clt`6p-_1GJe z>%K)~`Lj`vPN7w#Q}PM3ihROs(x02XPkCWVo32i6)E3H;p;{?EJXujo<2^&Pxg1!3 zhcRVqOFJ<0&@Rclv0&p*tjcT6-@fq@IXw`?iwh^?-nQA1U8UG76uE>g&M+meO;a`# z+csxJnhmiA>_Z03zKy=OqKh>x(%4yLGclSL=gwi>Vv97FZqrn-us7N*{tO7*G5@YG zx<}lk${m{#$sG(Wq|vCmqPe0u2@T3hedD~-#(_%@wrAk%i3bvgR?)st6kB(vr9zu5 z4ieo1TD)RCV1Jhzc~ag@@b)Iv1?53MZe`%YFx=Voc*e>y$)!l{Lvk=Qg~4xrz*aqn z9jDtPz%4l7`VJCKw%{7tr-6SLuY}QcT!^$Mn&Q29fKqeUwf_hYUl#UwFk@h+Pe}me zTbVS=iHW|WL@xlwJj`}e3RL)&1Pr{$13uv)4GM+z36pm5Z_rxN#iCt`9?yAkGw&;pfl-_u*8BWTNh|~*dFXA?4mcE06ow)iIo^v2L>Q|ktPf0rs1TGLMnO1dBiNgG4l9TH z+$ucKVTWXtdGvkB!Dy$xokMxT!}mNNpGlCh>ALUV*Y>P@)t>uOh}v=8e;MIn$~;Pz z1)&jfi#hF$*;>_}`^8+Lxi?~!q-sJFD4Y&@;VkFKi^p*UBLX2l;2h2vyqy|3!M+0f zOU3wJYNJ~=Q)~`JHc^HPvP{Ja@A=PVc+8J%ha9kWIr>c+_1nc6qIgk$)UMg>bcP6a zNZ^dg5TOkk=Y1t2L$sOR9o7yGo!XkMOqMW!KcC_5tv~hf)$(?dY3FxpQLjk;V?Dysu)We?|Gkzx}iG{xezy?5G6kR zwmebN4&kRbQMa?Zh7KM=4;R}xe?M?PUrMZ&Y>`{WF|zbawx0$O(l3g=mSU4Bl0+F! zh+DNL&Bm}c*w)@7oj#Z-Zpm|K&>D3pUz}3dM#icLrE#cZYzNHzuDf1#Bri>XnSOPkpJ<2MD47wX6soW6~F+vdnFzLULPCI_k^yYelRkhp;|Fed)xNw{?)+Wx_NwJ?% z!b_Nm&O7uDm{18$GJnajyCKL|IrfN!LdZuZy}z+rl363NlkS#oDK zv6ieBq)>olmRv}BsT5;o4bZxPLW10yyvc1euw^|wfEuQDx`)o5U87qZ+81Qht{&U$ zQ4zC1(xiI8m=ipKzCIo@^y<~wI?N6|plPGIq;Vwyb&$NdK=y#ShVyo~&wuVnSpF3B zA@qnp1sfSjpCR3^jlMnYDv;j8A-Z;3nGE+L(K2(%QjqUOg#ePVULzDs63b|g>fj9;N(&B_bSacI~NWA2;ZQLP( z4QS2;HU<_)W9Je4bWGa`J!qQ4>%>)(6TTeYZ*FiM!NR@*gb{$iv7@S)kn&ReEF5Lp7nluNv@bjqmh(qsp%lJx`(X74>Ra1t)O1i@+a zttDGu^tZxEw{lq(**K7S?83TiuMNbuQY_d$d6c1+P7SS-)=787VAJ{3&;#OLxAUQ^ zqDwU0N^~T4lDlN}*j_h`CUy8E#S{ePPwZjRSVY`zabYy%@p<26>NGn<*!RX+#?X*4 zBpeDL^bbhB z5&3+feFg$g=sv->6I#}-RV-kS=vGhI1-U|Tk@2DhY?3%llnAX*jRcy%h`L80Z|EPt z-tni3^7Fiv z@{n>xJ`YRRF-JO&F8A0uo{ODct}O{L(a5FJ%HUq~xVI(Q8oi^i8#QXpIcKnaM4v|^ z%cY1bw5?3+bd1jd@oM_mG8{edBKA$S+zH!Js2hd!=knkp(FGCK#iDCssbphFwWd-I zCwgWji`R+DLUQTtVpG)8@f>!{gmTnn%zDN$2eRj4T?;bj;(xrlLdd%DT8u+D!MFPS z>o$OC;JVR#|Da!HsI2lweoOu~N&J$@pKUhsCxc?495RhEpkJ-Wt2L~UUg4VsnloRt z!Tqc#%c~UYsZ?83*F?4S1ySd8w0~E!OTFvW1 zk}h#c!0m}fZJqlGU!*y-q{(I|ishC&x#YlfdcLec0WG5;t=3#UnN5;NQgj^+`CM7z z7U}BnbHO?07-hCDg~1Fj?lOY~k&hSwQW&hWG-|76#7l}nPG=~<^K7XgK@oh4jaL}8 zIOVz7Si-q{hPiaJl14#p)S@7VqRJGeJ>skgOIb{s1jh2i!C{;)>LwP?bQ}F_l$ngG zdQL`4HZU;jxIuKeMAi*j|GND6- zHcb^LyDDB7%oQjsb;ApwH42v=jC24k@(1d7SnP{d}kR@ipf8Z*87u#Zqy= zaSdr3Ll)auAonO1Dur&L1(M@~T_-Xa2J}UEvGqc8AuBY%pwks`WWLP&ac_uGcUZI0r%BtQH4n6C^Pxet zQQH|@0E_`y;^OEpq7o&CiDf(~dIftc1Z<6xpcW}??xQ#ppr}QA1DL{lHGAAG-PoTI z;{$khh{wBzU%h2``}eQCG56hn#3uN#GE(C|kBj_^z2C@He=Ar({Uz2vd_+~Dxngr3 z1}00C0f~H%C%y+ws&ZW?LFyu8+j5ezGvGQ25Wbn#wAZLn1+)xNl@ zjATDEMNRu`fK)=Uz$?6sG9aytiM~d-6sA)WCn`~Pd)8@A&FpZ0xS@j+7$?UTkk;xq ztj+;V*$C8Xs0z6P8Fu>Vb3?yDKKKT(KJK)97eU`yfycaggGrqj7z~0y_H=zVY%cdR zZ?E=y-U^$aUO%~>d@zt*Jb1pYQC~K* zBxnJ3LD>fE~r=Up^xWJxUiAU!y9U`p)DJ8Fqu;rtoFuz8dU6gI=M^WdeSvOXPi6GCk|??w*=j15?}Mqnd^?Y3}huusk{l}Um99Cz$)rA= z`@#2?eEneP9?ui7!_SeG1sI{st|kBr?l4!hDZD5wTX$1($h#rp>WtLMJ9NEnSrfCA z`PxGP4f>m3`|Q=OIekAoH|Lc85XqjsLYx}X&BOZe^o6os;XE8U{1XPeP!Nu&u@|m& z23c|R&h`TdBW3~H0JeR52cXCg^eMeLYm~jb~<0%8w#DBlxqS&6QzV*1*c-Q z>ZWhIy2vjXGTV{slWSKuD^04c~H^VV+hwQ{{Z@ipyM<9+kZ7`kL&I$!7nZ9n~p~pzdI7pfaMa74k2qi_-kMP^Q}YcoCLwV#f>ENY|_HdnX5CedV5j%agGa zb}O994_YQ|#F8htYuEu{S-jzZ#c;sDeh&T(&;q)4yFGD)%X(yZNFprlLop44!+Oi& zD5$0V-1mUwGSJuW4LPnnGVaK@5wb4K4_*)o?3c=kpZQTRkUM3^p%?$_={{T+HamEH zxakVqo?1F$U~qiB%tV`l|ACy51PZLR<5+PbFb0^KA1j}K{`t?oVKr3WK9%+|*)Pmc zb>WTfWgA2FV~VY#$T7;Wnwc-lhpdfUuuoee5($U?l3aQTIDGd)p9{!^uq47@4Q(Sd zhB8Z{#gcRyTWeTC2E7`yuF@9zas}RLsy&{`ikp(v;hZHEbW6y#s1_RQz0J2~>#!3$ z=H(YCTABQyhiv5*)m7kO#;T>Nya2!}NS8T|H$HDU{+Nt^VdrH!x^nm3UPf3_F6$|_!^+HU*V03 zR0klR7}|e{SF&+Hz=?^YtqibU`B=y#ilb@*_J(XFc{CJIK#y#$Gp0>vIaL|nqg**I z7Sgra>9|4Ha5^0xJj?&IVSaHmfa)4m`=?i%?XzQel2B3@%&X=)eX4xwp@AFsxh8dD zn-1+rC<##H0d)iBq>P|a5VitOdv3#bUf2q_Y~s|ip0AMaShZL2xcje>T4CDDg%co5 zHrne9#Wql+o-*W%Fk^UqAS46Ii$jox{y|WirZQv~Q%ip#x~jjXZe?1e+r<_>2=i6Y zc6&3FZXMI6!*=PtN*V{ZPTD%@m^L4>9}cRp)gfxD4};D^X+&j68VjSLw>&%Hrn}Ry zxB>52BtgEN&GbpyG7*kt09iRq*r>-_v%F$uo$A%z_sqmKjzU$~3nD~Y4REcaKMJnf zbn@QLFZ)hv@t89tU%WSD$(V&>PBJNw7`$lAwsDD)4si>lcUC>O(rW}IAv{w7Yb{`< zhJL9aVMzVk=FG`MX^G*&7BeK^^-D(2qFC_tZlnxz7Rd6^YJtut0T=WfUo;3<-W|foJt5%>~6iRRUxC2~zCjZJJV#d!&{w`^Fb@{`BZ{ zgWf!lMXh+T>czR$FJGV2_pOE(k(>t0UD9Udimy$-I=xM|)uSDh)_Z^4{NhERwZg9$ z4sj$v$UFuJ-T}T<7BDVM$fW%DALSA&d?u&<@Gx28!tlwqfzKL>T}_chY-yXqWQk20 zNIS3yh633fQ0hQLR;4=#RcuPBL9sT2$Fc zi=7k;MgLo%!|_y;f1P^=eK)vWof=#^xoC1`Fji6*$Qt!|&}|QQAIG%UBTJQ?KD}-O^AZUwdld4gFoUDpm8Plt| zJFRnKqyCbzOOpnBaWm(RU!@n2^vJSj=K`lYck4rQ55kgw}OC6;HwOaG~-k zvTlIWPsVNQxZm}7S#kFBW-^x~xv)OlZiAdmiiJ+j>nQ^i`2$TWWY~ISDXYODcQfo` z$Y5CSy;pO6S}lDrboUgKYPEY@XsZU76gN}1#&$(KWx<~O*c?1Ns8x4x?x9>JNET5$ z{FR#mtr#in`*A7BcxGf#r46EXP%NbE7Ep%8zE{+zr&|8Re(@U9fmpPIzdp86OLw7~=c*cvGNxrO5g)H%AD!`%{ccX#Uqh4)k zkbndsKotQDB0%7mIU$8POxK4mAKR^hzrlBl8FlfBZFFv+QF|q#FQSVnpN=(=RWmjK z>jc(P=g~_&&dN(?R!zwAs+JZ^YLVh~)j%47wI;35bsfsF4}~CQM4kAo90&T`upI&r z1w!j|tp7K6VgNF%9#3? zbSqQg-x7p51IVnfF(x+>Te4;Autjm&xZ{vb@Q^CRvUp9Z9#5o{TnGdp(6bVjvq_c0 zbdzIV^XB2~dqZ|wNUMsYOMSDH3xP7GK!&_!xDHqaR2gng<)&*%qwgYhED-_d$*vrvvghB*N<%0ZCbU5RC05!VXqS~26$rY&ahdVH9>!cj6 zMnARr{)|YRp~7PqS;?A9<2H<2Ib}H9)XsSY9K`t(E-U`wPgnR@vD5H+={%A;hU~U^ z==l^2={8xEp_@H4{(WF7JLH)g*rYl)8yeSG^5KxisK=|FZVs;sjfX~lMrg$Y?1gO+ zn80LGB}#USyPz6$$bvwGkANd`Y zG5aP$m7`IY64?sHC)J@9{>4D5+&&ofHUw-4K4>PKx7sqPx>#Wi=wpy(Z}c362cHz>D}p z^hP=_&yKq3T>v%gTgMdxLxS1WFG*8<2pYfa|Ficda7|w6{CD_&?wz~-@9)ms-szQDyt8%YW;!#S1#!a#1O*pV7FnVQC?Fy#t0FEH z!J?ulDkgxnSOrByh5z#;!IDTcFC^UPO#M{yZYP-Y&2ygfEZ?sj_rge;mk_ySUaj9_ zpEgOIvXY0}mnv?ZvLd{S-s-y5vlE(!+MtWXao_O?f6(kdLz|vp_ic7sKlkA4kNPaN zw%KU1IC1F}xV59q%o7#v3uM3@eOpLn0hh>TA1u7uB|fN3<|3E=!fGbB(EZE#vjq`f-aUD*M1T2Ku%(akbTjRX=ZRnegpO0Es&qpt zsXA4jXhm3>1VVgJz}?KRm*fRt3HLHp?7YvFLvymldq6;;g@Yl$Hc4;fInTSkTf&;@ ztang}a^$%c|CO%5JMC^aYyrnJT@cend&9>10@>1YXO$%dIU7Ub#A}Wf!g6DnIG$*h zh#G`feS5++WwP54ZY-9cQ13Bt6E{vrvXA8H!(n}4ZSczOFd1ul#{-?Q#Pvd)Q18Gt`*loPt0VlpVQm5ND~WB+|cbkFOl28^WwXXOv9oF= zl#ET-T{L!HC-F3=6g3bXM&78@X-3h~CBZ=73!T?U+DtfaL4is?yBx#baipGOQ%~Fsz_UVg2?VWSe&rRR+>Yi5@ zVtVqg4iDBX&DQxR&snxGKE2^_V%uSbASNx~X7CblFiM4}QF2)bbq6Sxa+o_ncFW3V zf#!z59tk^}zrd_4h0UW*{6_1@c(35{Y)ej5D`cV_P;Y>oTRpr^$WhoP;}X!)9i#gQAJY;I{lcX&a>ZQ0e&&*WR_HO0>d0h!qCdW!NpwbOY*R41Qum7jk1Sl{QauDG>{P6_%gh z1aD{c@jI_o|7n}s7nai5FG-ZniqJL&->|F(YVoCr?P`J=5K!gcqG$@y$sf`fk;@Fr zhZ-DB4#CJ=F(*#Y$|(x33ivX7*DI}pZDR$!R(>u-FG&=wpVlM9B6cmE$88ty@Hj(5 zj{@4VMkpi2xN!=!APlKWg0=ML{LgrrZtp$9{-9U^&_&PO=B1l=fdjvRL0t!!_H^=o z^+o?{Gt(6)n5kJ6-Jv+gzZ!5g0L%Jt6~yAc6S7O3TpHKOBCk~(99okTQY6493>1y9 z&Iap!aU+`&<+NItAKssx?`m?F8s4n=Gg->cU2@)yCE0G~E~QXRGDQ-pn47^{!cygW zSuJmnf1AE6M8T31>Xd*jVb?;~E=?0OVYU+uP56yi4iGv;J4I-M#mX&Le@qgbIKi~b zY@)YOOd3TtQZWxXyMX0kTfjcIp3vP3mQV*5Tot3eY4N-u%%G3@CGu+h_HY&+QtqI8 zqT;4hR%XeGmyyF})}@SMN+`0MirL|T3ZJpmM~&}{NO-nJ z@9;oj&vjpE1u0LX##%-|w#V9qc$6#1_85*y2BiV2$^P80fP15;?yc z8O*-engEtYFg5biMwHlgn{_P3eEd7F0^y?MrP=u*y*9B+a+OpaU+EtS~Z zS_O`ZCN4&_!Ox!#6^4bzI`EhqfwKy;{j=K2fB3 zb?`xkE$T402NEtQg8)N!$&y@=Ng7`2eOncS9_URcc`go14@d|zJcL&lEi!VJpCpCh zm`BH7d9?@p%8R2v@$Cj;kPf;k1kVGlY>W`%6!r9fn_|g^#%4?A zyqis8g*&Zzy5?>~E9Y#KvEdh{7&kX0c!dTx;V1Ma|gY_|87KaQwjyI>p&Ik2T`^%C%(`*; zT|z;Qz_f;aSUT-@WX(&nxz0E9indbBW`uss@WLx%q-p5jXGd+27K?V!aREI(l_9;M zXJFylBsmRB-5Mrk?(#W%l#am2nA6J^IAadtfRoocRSRV%cx1fxvtN?bmj)iCX7I?R z7zoB>P%+pSuvK|M*cyF|bjdb~5=Eu-?s-X3xnWK2M`2@iOiFAa6X~a$q>ofX z5NPWqN5k{DH;AJF!xkq`42%P?F?z_~#C&z#bQ4bW^FyzYEl%tT1IhEKYTsQHvx6eJ zRLmFeUJ)Z-NTuj8?`CK%=Ptikq=zijQF7I5t^B&^PVg!IK&Vb`dA>&d7#5!zbsxz2 z-6uCg9Yelp{IthC|3)B;o-(_>w!XlO;k{D3{=`eP>>Su>HYr;uW)lVf*io9!kY>Vd z$#w5RkE_t+wNq^1wJ-{wq$^Nn#?fECkxz{EB#b=h;Ge7}!FhK80o(rT>@1TldHdSV zH6-7OZOJh+*dCx5sQcba#T1Im#9v6u#HqoX)mfsu(+b5${jx+A)X4dk=q^L={1H+X zxMy0~G_B8Kw=b_@#AU5dqu@-`EpH$Je(muro!{fYEcFrCo({W@afz_Qp|1##%oPdty{XH z79l$hl}_x`YRy)a%@hMg7wf5* z#_2i`y1x>l!OlekQ6&(q0&kB_jtB6amn4gjIoFZjdGpD%2^(Lo0Ok)n@Q4*QUP)R} z+-xZl%f`8NV*kX7)-CqPo)KmP5fpNv4T1RHO6kb!+pdRsMd6URFwH8dL_J@v2y zPefg5tE4n zHfF%?dPdJMBi3OsR&0Lws8`1?Ouokc_dTTKrW5-bE6s|U9#YHzMeb2C5ct5twTJGl z(d$S&XRkPpWYOKySivs(Smaq~AHw!+++=j{@p`PFN>~usFFZ-#n0-mrAg<-ygx1Fs z5f>s*#AqSPX5IEa29-?xq*k#oBdAolPl<`Z0fKqtot*_A)R6&?vp1KlgDh1i|p zIpbU3CV2swsw&e>J3OlWFRAL4!+{H1qqln1sN+CZ>7=lMUbw@f#Z%`AXL0l|M}*}~ zcE)o}M24tcJmhkmB#Ii9XOvsLZp>cTB)h|l6&#A%=e96i@p9BPQEtEa^tUg1QFq2P z%u}UK(;QNzNid-pI}c5M4lTND^kA zBO09?pGJ&PXx=A|S z^!aqlGCHd_kF|t4EsZd-QecNQEv?ZV{L;U>BF^LHb8pkdbhl?AH%ZkJz5BZ-7o0UL zXagiMq+Gej8?6}R1iS|zGjJ=Y8DtMo$Zg5IV~Q$n2ef(X0XV7Kg zJHaNUE0V|s`AKqP_U?IY3?igJf@juDEA_)n!50x3G#(GS@Ad1GcgeN`SL==0t=ug8?bgdpfZLd?xRnzJ=da{lmsPrTC@OgvQjCR=;xTbOPm`zK8wu=<0&FW7V}8f;WgqjI z2?1ry5$$tw7H?IpC3(>cRTOT8(C8DzlHh&_0L4<+9v2s|JGhPMCdjvS2U?%o47^;$ zveW8p@xj16^$q)LON>1|>|rqWIQBkk-OWx_Y|>iNgTEy1jui=$V}i$`#e$`Qr$y(< z0N?z`s4se+(@SHQXBV=}mYMYwlSGjODyGRbTdbvz$SOU~D>Vb; zpyHfo27S&mU(`9dNd=rQTj$>1bNjsUtQ(Wxy;r}+l3>wk1It#JAJYRK`}7NelNT6n zs^~2Ns4h^ix*$wf)RAMLZPX;mlBWlxsmlY?6?Y@5LTdf0{ZSV57_=Jdm9P~dW7~u$XP-CWtgl-~$XFZZ{de#Bzv!*{18!BqA(#BX zDq(3v2mgrr@^n2PgkmAYa$AUuRlVN*-Zk_|QZIs?HjTmM8?(*{Yh=Z;Ls98aV3q1% z*6;}+h4t*Sd4L^Kj{WEAc$I0@Dd{PWBU*Ob9w**s*PE?JU3Yt+2G$m;}fWNa?icT z=cr$;r%tXH7YS25OSwaG)MVS>pRQ<8Z1o%(YuC$W0xj;H0M4G6Bhw3uktcTeyX=(D zW}1-mhmUoK$POpYYaKPi!hVVYevlH-AqAz0R-q0mb9?A^ag|3EXDO^ftEHPf>Q!qz z+rtYzOJx|F>nB4pXof|cgyPvIX#<@Jak^e|+Naq22vE-CsJlY(X?=J~=n5aLT1!`P zkCSGS$+7Dm*^XcsBL*^lK(I49OFlW3=V^k@KheXxS74-S^w57^MIB z6Yi@w_^W0Wwrm1OF9=THXYgfW}5Et zVru+ovF*v|+Q+~9(1eoA`wdgctru)Vw8ku%K14AODKbFCG)o&)Iz_fek+4O9Wt>Lk zW6dM}wUAq(kV%39a3~ZXc4?B83Rg!;xpWlUqe!hn^`zt3d*TOpU`XJ@VwpBcuz~%_=$v2pK`DJ0Q~) z6(re{z1n6D40ny!Rv8^27H`Dz)*mcv`TI9bwyNzLH}c3{b}O0_hjq@FLFEX=z%IXv zib3M&43U9m8rg+GGm>5_T`J8CO9;aZrUuLoNeC;ANR_8?_WRvY_C;Lx?Bilm3yVyL zU6642b7is@)@<~WYX4!E>)yKrnPF8SgC12vtnw(3?B+r`BC0ARg}cTxE$9$`Nl=}# z3M}S?AvQ4>S-RDuk6Gfz;`TKw4lK1K4{_RPj}cl21Bm2v{ zEo7||Z@CK1>_!H~Y^BI%Dn`d?pzD+)K_>{Oq*rixn>C zHi@A|;a2G7kfUS^6iJmv=;WFV`XoK*u6KYI)H5f?c33=f7`tC&1&dems^9$7vi8Mk zVPU13rCEw(n8^`+a#Z=l;v;ZEo29J~lLdNsx|*Qdhyg&M$sq2Rm>S^ z%HI-c2cPoUb7(tIMthUS*Uuh#VpmxKWy+OTJC0e}4>qL@PP`+wVr@|7Yqdwc+R$K$ zgs4DnH`-+GD$dhx$r0vhO+=*z$sGMeW3ESz-^&GXaqYdcIVY|Ya{+4W4oH(sA< z$+7FS%ZwG)+|#OuZ>7pF%GQUZD|GTII@_a7g5lapwgfmYwoS8LXV9j>SS=~9WW>!_ z6k@U+D^q*-kU}Tk)1EZ529*?ZfFk8o%%gcd)2me(^n*EAE~%531rE43D-vO+yJG4C z&y7s{)DO=_t(aQ(t+F4)Pldd)G@GoLTCGar#!vnAkjsjx1MZuAZgJwLW(3s$ph|@u zP*$BoaL0qs#|!X^URDScAm!YpVTnHW0h?|3u`m(OKC}rR&%22geyBIY7JlR3P59aH zr}y3^=U%WZ(Ool4byCcAid>~)v_K8trAEypBzQ+sb}gL}P@yOit`E@z@k?n$v8+kD z-@R3~JjAH_28H_SR*0`wDYrA|3#W1`VQ;i91T#$Ay!CJS@0O=NNAxsCaE;?O(ec#CA8fBe7YK7x`yv-D!e}ngM|vDm z?DYfc35e((2zut$+yUTwHZ(_HeFg%pgYPfzs@L&NKzsZA#=~Tl6B{ZZU>)VQq*2U9 zilk669sDf+A^<`jFg;Y!dn3z1b?!bvGga-|?H4~aU9nYCMCVN+XNZlf*vRZXW9IC9 z$vP0mD!8jVI&+T31dfxD{yfsl&I5Mh{7;fu7CY`0W|$(6uzREmT9B|6azu9$d1w$; z86X^~xkS<=m>Y~2Xg=Xw3r&?DhBnBNZFsY^Clpf0PxLA~_{&tebc+I2E30TsBoDhJ zgWbtdUjiXFj2Pp%%cE<3(iQcb%&-prZBj+w3%$qhp3^&L*k$A?qe@!_jisfjZdwUO zmxoA!t+@qmKEaTy+y&?Ukq+`CT#5 zen@fBeSJ_*C>A2_33l2{cIOXvcv@O0HoJ2t2B8&uRwizhT?XQLWX&rjhGb-#SG*uq zj_YE&0&|O1ey8~p&KI(oNd)|5kn@aYW`4QVG1{ij?QZZM(uZZiph2#+bx;&SAPqsSXB*>}t1JOro^eNv*^BQE= zAq$+ydF-R5uS3=B8Q;{%eNkQ9N%(-r@##*=2h7f&aN3S{Z}-eW%NPzDGvdS%U@H+H zRAkQLu8iEK)YJMPO~b6?Z(SoT;8#onm@U57He%c2DaRka_1XhJiOF&Z?+AZEl3uVq zc#)Zz%cPiX6iK6E^fZbHLmmm$xsxNn;e{JD|ZaA^~VP%bsgFfnFAkAuxULDX#w+q_I z{YYedK|PatUbTN~^hIv3dpqotkRIzy*v{|`{wL^C|G#4RT=>rEJ1!z9_@XSbj%Uc>A9dABuyE{-+cV?Sj`%Lv^o_)|+7WO5j4|1LBb|0@k}sV1`|uGj z)OA5N+4IuGU>nWWni`6!q{spEPoXA~w=KHYdykt28Rw8E(bO=cr%^g0T>-hApiB`8 z2e*Z2+9WIGTjZ(o9%;6?B0LwWqQJml(QS@;pHB-&Q*hQo`%zU$^7P8^W+`e5Cr28t z3(QS6Iq6k|z)#uyUU2F8103bD#0XPp7;WXMOR zN#kIoK8xEU)g+4U&ZvdOswA>bUJ>=ho5eGRWSJ5}JT`;=;?1TI{cKGZ_qIHT9^`83 zl+Drt&aKc=bq~Gqtp?yt+&m*)(WyKB$P*D?Jd@!M^~#!&US$)o>tQ0&k>-k9?H4>47Z}7OtSP-cR?*;jw)_gC0q| z-7=KLS|PY8*WBc-57E+@VQ_1k1Q>POBirQ7o>jE5EaoOpgZK7I(Zo$ z&lhxZK}}1tcg7$rZkh{Zz#Rm(;8F*;RJ+NQ08OjH@D4jU1Ee%UOD8a}8b;tR|FrIo zjgT6>mN@3M1*EL|_48w1yEKtZu@gt*tYnJ&NG8+EWCnn*Gw-VCuuncn+J6xN?aiE{ ze(|bn{u{VDS+;mz@FW@}cCfLSTL&I~BkoN6M2h^K7#voL{BdW~3>9p8Nv;~*mjgkt zM9Dr^>Nt}ue1o%UD=Zv{fH8;mnVB-jQsU$#aaybhqLe`VcAsoJXg4Aq0yL?nt4}Fv zB+&JqFG}yH`QVt4`@c)u(6={+b(@6f*L7!6eiuL@S7a@6M3bW9^Kbcu&03%!X{Cm`xOX)-fpc&?LJS(j?PM(xhn+ zZp{lTb}M$f>AB~1%>3YbW6WMm&iv!lu*ASPW^;UxSl`M_k$c_)9X$bK;!w z5i=z0qnJ_(O6+1#fbmeEPX3T?bcJ2)s_2qn6p1-WK6^uRTHPc~6eV(M{XUzsM$*kg z8r=TbS^n6!VQ@)ya&px>1B*Fr464}d5+77zQ%Ch&*w(f{Fb-XpHrVfuBd%&~s#t>L zc`J&ww_$zWI!MOKEdP#rBgj&f=4l>(CpIh9W?(9&7^q^}Ma3Kh3Gg<_YL8);HEumn za(`Mi2>hfE-!f8YJeHr0GQ{OS^*pV501gR8T0o#ExD)#2v5QB8RVffd;r9rErbSiY zSFPG3NC|L6xtpDE#2P2hd%!x57M)K0#?MX0=AFX&>EtsfHa78QjT?g$bDtu2sTfT0 z0Cym`riM_QF+P_I0j|Sbq;^}ReJ$A|!y6Jn3JjjbG40bKW+Kqx{%6qRb7h?>U4i$Z zuNo&ny1_@%LN0o$nHQyNLV6MQWO;;>LI_HJWA(g&s zJh3GMYU!Y^GTURwWk{ywWP8L4zECamT*<4VAGr^^V4A2!Vf+pB&}@P%tr^~@!WSk* zc6clV6r-?KC;Tp%(BZMd?QTRy!0kD?0*EiKa2r;hqw&pIkMrWDfk`OvR|oQuOQm@{)) zu)aP0ryqox;FlGUvw;*jvF-cBY;CNfn1dAAKhm^TwBWkD3rNU%glLmWWyzqUg%pnd6gJw}iyK>78R z(jz=gEeEwkJNSci&iq<+X?QL4!rH@kx6?vt?ub2a6|~_xEBp0I03G?1M)J!YKr|Gb zp#eAw{Nb7cItOTEE(?3WH8UcmfZOWh2*=9m)UgAS)nV*@!R^6`T`zj!Z-Wqfp7i`I zcg@;>2hge7{c8MccHDK@{Gr=6$u}yuGvHX8pEZppzZ-JMD9flj#7Vm}h(}woQQ18b4zgyI*280nVGn zoAQUvf3)1Vza;C3l`Va>Dv{G8ObIZoCnZ5iL~~nm)~CQ1+mfn640MPmWO0H)_fx^A zf>DyFghu+e1s2 zOAHAnBhx`VS>+i|It*;v!!EchuM+mli#bPKhh0FrRamFOs9=rck@Td(A)y_6!N8sn zz_D0az$x{wyLQ>Q1eJa5&Vur{+ZT-dOqHklFA+S3eIsry*2*`&ZjXI{?H>n=39x-o zJKV`?0-SdVy6}r{7QSIJClxb)w}R|)Vsmob%$yvgnEezfqhgHe4OkA>C&Oq8Vnn)j7J#L)>sh}>sdb?1RN`(@CKw>eMr~v_Q{sa+Utr`Ets*;aiCKP z^HxUoD!jZ)(9W^j9)|_2pI%SKJYxq;!w{@sJ>}Ey>y&?QLeHnRMbfv6Yrdl$|m7#@Qaloa^3gl8f^-wSGEzxAds%^Jeim zA13R|Hlu>|$P;^Ddu?l>&}4L0Zn^qnk}wsrg`-Qhw^2+QMK)3~TjpH`GQSOBWlG2( zgGVVxpHt88lIi4zFCE(YW`3z(_1yCi@vxh5f6l+p<`4GEFU)gBw9jQ$w%>WbTf+lw z!_!Ro=>Ey&KarJA?9yn>fV-JufJS&d6;nP-FWKaCU78{8rc1rkpn3``czeaEVOqMD zgSkqqs+&xUzzVeX!J|#1umkPeKVJK@pP8_6BWmsIJ{|L3M?ndZk#&>Fw)~ODtbG<)`6MW&< z>Id~ZaY1_D4*p_6k#OJ4JS-d)4suTm^TnFay!&Kb>bsJUfW=UVd->yD$GxC`w}9IV zx}*ghJ$(2un&87WbPi=I)8&T_56pn6ar5ol8uG9A?Md!$5 z9#|QsNhBm-Wtga<+V9<&Wim`}U)#BcAhKhWz!P(+?)RQkN z8Q%^M&Evpq(j-jwLMD$#o`vBbFNRhYXo|yv(G0qjgLP3xh1X7boq=}AKJc*(57yU> zF|XJGF|3@HSJt2Y%0Kv;oECM$%4)K0tbP_Jwi{4OH_A0Bq!>_w)lxB;yo#vyz=SYE zh7wE3QBopRzR#zcciN{aqyuE6G4_hAL)h}xqQFJO&$xeohR5bP!OUnj_Z8d1?g5oT&bKQR_EN_C&R8*1wXYP8_+@ znc0qW6!R%Xny8pHp51dEfzr`#FW_o-Df7&uHJ|hQWT$=7yz=KS^C^vv7Yw^p^YXaY zh1KMWczsBoD-_sC4}>k$%Z!&7;+MbQ#F6Soy%Lp)%&mWcx{5fo0vujMZ@P-`BJ$ zJ|=VoMkKt*>%%4;RlYv?NeW1r&?`)iHHFx%RICh5a}U-I6C_s-3@6?BK! zOKSZbVt(uaAX`Df=058HI&g6NY|AcPHmj5q`#V;|8}-FH_g6D@Vg*`ULD^JWB`-%&)yF)HkU>ABH3!}> z{3sP(;v5Q$6+DFIZ3iqsOTVR$?>z-rxT`KzXFP2kTf7m`gc*tTY$na8n% zVqituPQ}#FP5fq|mUCP%5ZW5uK)1fNaIL6Y+8}C;*2!~Vo2w^1v-@3ZB{zU}DA8-J z=vZXGYq98nSRZth^t+xH-kZI~%KK` zyZBoI4*Q)WDX$jGPOE|UJ7~4MUEC!t=e9;)5ntnU$!cFQWm|M1ZxxUVbvp*xIXeEzrF zyzj{j1DZx^po?r8>*J3q;Jz>4E|$V~F}|N}j`2Nffe^TP!7sAa?SYlyAL4JYgD#ez zQ1@{Q19fvwDf-E;TccY!jli~7FuyW<$R#6aziVaqF=Ynrv_((WEz%f1GL=JR=U7tc zS<%F7fjTg#6T?8~joC*a{fGR<*uR0LA-iRbZ)kuGL*76NIQM{yuAZ|k(r#rS&-}WL z>+Af?5$$u4)kHXNkfNgV_`V2}Rr;s$U74hm-8$yH|1#2SX2T9s40KOcBFj+~J;YlZ zpy}YB_Gy>o0_mo)2(u*Yk^GFGp|BkrMiQZ4Ad^$+rhhdTxLn%3S93Z7x`Qj-)_dw^ zV2YrM?veJ%zEGZg>kN%@(s+kv(4$JZis_Tp!-+NAA|a9-bpZ=en;ajW@J4@JUxp_R z^Rr$|>`S5lw~N&;I!nV=uY_3z3kJFP6MNr~{Kt_#uhmc7mPxkyumdl!vv|U{roLyn z^sw1`JF#K0vVwL2QyTW&CyKN|Is*-r2CHXKHvl))SdV#7S}_YNj_2Wy4Z@!{mOZbp&q(@Wy2+sE=Z9V)Tby`@ zU14VFc2Nv;h~!c++q@G)GTq|m^$0t{jsmG%4zLMkx>ZJQlC(zF(IsIu>XY=A;7mz^ zDnZpGZH*q}Ch}JD_A4srE8=)^l;F=peB`WiobSSB{>LzE_MCFFu~=C+O848Z=-x4* zCljn|O;L1;pxk;DP$E8*GJ=!42eabieHKrfB#4ITV@P~M9v>J#d8^)gkx za>yk;ph&nOVh^W9ik?ipWV-}q3hFrnUg?TPQX7fyKTga|5Kv=a%<_~OPHaA42dEkT zzwo!*1hMgWop;S8Ru!#VfmRji8z8%+Qw&Hx4OlhzZ^U+A;>-8Z^zT^`^Ez$yf)%RW zI6-lwb|&!ZMpZ;TBe=!u2Wi$@lM^G*62cVPze&AqU~xe z_lme!mZxrwE|%>MPU4}m*7+G*l;^>s+>q8wI>qaOMf$PN9_4a(-JC;nhFo-W7SG!~ zSNocAv@&_AUu$#+Sfm|q8}koZbw~KE*Y5gU_ihP?+B-F1B?|rs2EI86)m_Y%b|PUx*u7` z|IEECur<2G`!>A?db78{@}{Lv`)CFA+y=UQ)+y2^$7;Mayj{~Dl8tXw_-O?-be>lh zH;2BWTH=#8^9dkECcr*FCk`+B{2rF0w29vPC(CLpHcOZj2W_m>W+5e2uKHFG_MenS zblcsETa7%l9Ht>v61cX$rMfaPazJZ6K8u}5&s*;^O?_zSMdl(+;zUHme3ro@Sn zk!@y5QYmH~Mb;uSl}WF7dq720n`GJRn^nUumxPy~RueU_ClwWJYW5s6|JFv0DC-Qg ztb=e`564Pn!wR1jytVL^Z;9Uj?JMH5(XG)-c$-2RgEHl1fpt*de}q)16Fipi&IF#J zGlCLfw^X4{5gCsg=yfDcqK(LdTy$f2Ls&P+TXc+fA7t%AHl1f<8?3k!&v}ROU1sN$ z%-vaZPinFxC%*OVcCv1)A}c4}3_(bGRDH~Lih*wFtyD}(0Cc)ybK{^#{Oh?8nl2aR z&(_4wIYVc{&O04yVcO-{fk{zEp#`>t$@M-3^_x)Poe%9xjz&<;tVJO6#0hLG`OfHRDuyoiam&6+OrW z4oom-K`}>5pJi|$n?Y(Dw5%*f<2Y?AB3VIe%9LL&qAYc-oz~h|QL09HB0W8lE7D+# z^gbWVmkui{eQlmkTR-3D+-EZ^o->ZUFSCNhE3HjOB;PSvoOcfQ{X04F((HM<&Fs@f ziaAe_78EaT_DlkA1;yU=IUbcXe znHp?t#AslbOoQ6#NE)=$b3lT))F8850@DI{`vGzaq;BvyM~$^sh|?ra6OC8}wvbf$ z`jESCpXVSQ9+Ey_iefLM7m&_A)xhUE?6LzYt!#m{%}^YNHM07M4YJq1NKBBsYPdWj8F2L2PVweHc3IlB?e4-33y9x#|phiy)ty%k7hE(t#BbsJo8botngu+n$G%r%63%oC51v0r&XaRQU;+krp~JWSLKk4=%gwgjE5W z`#~N23|bG-VOTOR3-S437bLI6_dTRI>ApUwC$xjVd=8KXU>FF0P1I7wH;xB6wqX|p zbq%hQ|0npe3cQgE&mtb*bnp*vzUXQ467KW5a#KYg^A`UP~%$<~FSQR#r(#C_$~8N=ZjfxXhbGi)&= zHhibeZ?Ok5%ty4(5vxUz`YfkjXXzVRaY4{<>!AoPT`?rv@2Wvv#M}HHU>DVVrrsxY z$Y*T*6S4i%w?2j=uCV&_o!9)&&UXL4$h3TA@i$x}iB7y5g9fHi5u0?1*+P*`RE$9+ ziQeYFhg+}OCDtVJT0C#cwKH{cTw{* z#Tnrl*peDm?s1112ZS91lD>%cz;@K3;hcqY7&Sw^=RVoy89j77)f9rPhS(+)D=43a z8`{h;TsHis<};2 zre5ZjA-W8|#~0+`TQ!oC&_jfdR5vGIq^0rc1K~Y}q-3i6<>*gzUiYRUvU{#2E0LJ+ ze)hu;Vt)4iPZ$11OfxQoa^m_JE8Ihv-b|I>3~u7|a8&HP10vNN7`_oycX zm(j+>u3(um<<8VaA4~);PP`~tflK_V5Ul zC*L0AwnS@R8{{@D=!%Y`?i0M;Bg7J~22sA?HouO<8=fk56sh>c69jff$&=xHj_yAg zY_cxh-@REvcD`T@Ho_dcExT6#bjKbOz{L?SP$ylD!veC8NhMcf(a`jnSHYD3^w5}N}Y}Myy*&9 zUDDNwq#Lq=2i1k)Sh1&%_`kMJq=~?ZqkC4G z2(SYVvfV(M2sDZitI41ZWfGV|K$e2k;Lgdz5Kc8498C}!tRD9d-}F|LX~}WFcPy2Z zPbE+oH)_|=Kru%sV1bC)DM{sK_zpujtu=ZbiRWxpGzvC~;y~~)jk90T?N>wJi71uz zaT6F|7!$AcyuiUa+v}khfcdgW*gzlWtPe?7Z1CSQw>297z8rEVA~S4-oAEjd*mSzz z3{F=pm8N^NMrU#^f!ftMPduvT-KJNoCQ8-AW(r38gf_fp^^sfN%<8M`ydGh__Mdm2 ztoY1r0~82X(0!oM@Yk?DXRlN243v@=@&n68?WfaC&I@y5r5$@GrvY}+$by!x zI0XsXW_~RsT~A0#!1L1VQWk}y+y%DBz+nm|oe6m!}?d@QJ&>y6?v`pxJ3CvNjn0@@^HzUc}FeEpv_E4EJcq*UOI`bDjqhB5r3S0p(V zKesI6eWBJDD;39R3rZk#`;cxAyyXKS81UC&^*9Q>a=yJ zL`?ZZ@kI1Po!B7(I+#(b$To^eqsT@`MuX%@jYJRiS;H>acvR@6^V+3^ofs@A8KPkq zEqzGcOPX9ug_R(SFPB^iS{Lr{`DZ=xCIDE>lid%k zn~X757wY$N$X&7`BB*^OPN?Ub>2=487l&^cO^>%nM@ng4OIns z%zoeV!XcLp(~od7MQ5Qa@;0P1Fe-?gWDAjutb}<88OW_+tu;19S@ zH2d(ZxwaifBdd+s&u0xip=*N9*^QQILo4|5}FX54wJ#l8{o%Q{PBurVXf`!55==g}+AMv6(HNV1V9 zB!e!9Xi*Fh%pX7ELzxPcXdA>y{tx)a*Z|ytcz$YV<&qy%dzbRuBHSAtD&+Mj8$g2JEf6WD9F()~qiM}{rlfcx@PlOi49pRcf zUp4;>1|8Vm0HKVpXH?!&CgkR})|Dk^b6voGM4v*}D;+i}#9=G7~RfR29?x z^rzl(B2OR)$<;Vg*&g z5>Ab*HTo09lHjwTG(2)8Q`{uHAt-RKCPj`(=!^pdPc0g(X4l4>Svk4X^xn(+er1A8 z{K~{A(!7XBSU5a1KVtL_LIH#`f2Gw zk0#g0p~W1a2Qv+%uzDFRg^zWpKo5c~>Q0IH?SrT8mUy5W0j5;-kU2a1_0NM-7wHRV7M z-ro=J@3d|I|2y<<{`XDzJ3H&*6mr3dP1b!glXZh)fUWg971RHQ-WQk$f%)wcr$>6) zN7FzT5-q0*5=515I#MO<5q3+p^p0rF_0R^8a~xKlqgxf1ft9e*twy$NX0``}XN)wQ zIZ#cGT8}UU26i*tvi$qF*jv9g;0S=EcYdb)?u=T_?s+%DG)0m6Ak9i~oFpTt)vwdN zD>Oy4NrG21MLJj}V}rK@?4UpMIW%+kyk74N`fNy_>>5XN*)1>Z0=X5GuDBxBa&~ys za&&Uz@ori9EGsoJ_ONwN89$aM#xj4bY~3qs{wrhJi=LIl7FRvL8R##2#i?O8d8^$v zd0pgT_t}sOzPj%Hkh5Df4>V6vmDtL++nyiw z(_`5Nil?7*^chy5c;%hkhEQJvX`c{lpfVK}FhJ6J9dp|Cvw9#=?v9MmEW(W_YQmm^zA7 zj|eK?l47F?cCoZcpjjFTv95a`S&$&=pz}r0!U9=_f`~yQJ0g0tI(ZcxH@AS@xshHj($;B+xns}MmIfma-2DhURDbH-Z%gWT=J*4oZ*|dW^?WnXvCB*=PM>!Vp0~Lce>~UK-2k5&IgPynfni@Je za$QK8M1wtZrQz72hOBoP^aWv6NUP6MsbQcnJEy8}s1JI(c2iGWp8T6OKWFVaq%?52reJy7_Y3u~5^J`jWv9L+U(RmeU!3{5Yeel}#8Dt6wG4C6@o zA`~q+HyzL9(#`x*ze+dk_8)fH!@b05mZBEHE?2ueF2Me@Hp%TdnpM$f>9b&#fK4V7^8Y#N$E6dZG-u2&nmzuhwtvoJf8FfyJB&33t1qkb z?q8L8duvsuC2NE@^XXD?C$<$<=rOy(kGdi+xS`V@9TL=EYEj(u#6X{>d2We+xu{LD z3Ra9kI%j?@=#Y$140VOC_R!z~6c!2hi`!fiIfk-UEHKIo-x=N2_DICsIr9#2GFtwd9yA!ro{Q4?N86p^XZ~ZIWi6W*skb!qq1kN?w>v(5@N9P|RBWG5DSc6jU$&IucI*V4TN3UU*|ZUpR}*9Hvl zJ2_?HXB<cH4Qf-iZS(#b)!JO)-#L+(yN;&n*&a)h&uu0*IB|CuM$1IhWn~xLN)qx^~Fx zb04}*cJqK_BTj%Esm)SU#&cvdn$aY&HEvJroFgpSs4>|Yx4`$`TsFmo5_NX@@5o~p z;oIpR7Z6cu4iSn(k)8 z!yh!~enaA&*lGY*|0q0cqL@^QtV2B9lVTFk$kCJ@&?Kn}={G3q400d6SwJm{6QoJ) zwkx$^9-j-1C$<4!*f5}-Zea(F&*j2w%Z_=cH5FF6<98_wB#o}iqlaCLt+)u0VVC2i z*c%s()M=Bv_plupcA9Y;hhYZ{Rr60XEagF*78q6pKd`aAkGqsJB)dz7LCLv3qI&Lm zjwV^OG`O7G&y;&#lxg~8r+vy_yXjNuc4+1@A5EHMP@YY;d7fA1xoS|Kt53Ei%yDH- zwqwQ2ntXX+2_p`ze_2&;(ds*6=mn5TS#WErDZmPUo z(jzSlhx((QStkXTg}3Q7KA>C7FI7X~e*>M!x#M00ar|=0C!iLAQFtuC#M%T@^|Mz= zft3k-zUOO+9P?vq!%q0T>g>91OSxY*SaIHr6tN->+`(Tt_5T0YwP4lM8XA=x6GOC; zEH1Ezsp~yaz$#sFR(f8BvA}c%N{ZXdwmfI5*qff`3}~mzJ1PqQ=VzwP&W)(Guag_> z0%*<~8YC;s8jv1POg}~XfTIJ{LVD>lvTD_SXdK!RmM_ABM~$%oBwbO=IqIsD54vL+ zrKZ@e*liK^8O$VBzfmqSMVR#?@S7nTMVu(_l89 zZ;b3@U04pYJil#;*aQgKU*2sYYn?dq0{yb1A}<*f19QHaidpQ1?XvZp<-%>A$)eWi z#E=>~pPNVT;8e(mWCh$ck?D$Ms^LJ82=LmjzVuGjB!%gpKj-Ey2^_JUz46`7>Kjib z@k)+as?)zD{*2ZPPn#(Q_Jr%Hm`#xz{Feuokp^I3&~oBP6MqQ0&C(TkeuMv2;pQ1- zvI@~8c`=45W`(5jfnr=wHT)r|&@xf%wEJo+8SFn_m#^TpMrVm{_+$o`i3dV^=n8NK zlA<2N+1o@XzrP@hS>}@zHF*=xYPu)MXdph=8IE^6|LIqj;IJ|WgB~D*(ZV@G(mXIC zI_%OUtPxxi=7e8@&p!w9t5^+>{d|9$16Fh3yjz7~$*+GSGx;4Eul?+oB=x1)1eKbt zD!CK`MbjBnOsf2lVy~Z`2H{;`L2z9L`~?{zI5egNws~uEAXzl%j-)PqBe!BKB{n!; zTulZ&;vLp&X!vSF5{vxw7r)Z{@B2Uh;cxy&yqsc|P$c##2W89?<61jxd+0CTKX!f5 z>j|>_hk&l7XL@P$4jP%Bn)nAIw*|Kc4!PvaPhcuT25CGxN-lpJe_j^ekZCzNGxE7B z9H+GVc^`yj!cQH+7_Mwhqswof!iaGk# zE8;`^F4>@5H%B+;{+xW$!9TL#qd)fq-JY}D{o?#absSGOr&L+#R=Hrgd$+i8dV63& znDJb?BFp{2j9g%q+c_hd*Fzhh>lR-Hv4@d+Q2eo%j)Qk8S^zFj7~c2CI9@(ix8VK+ zn*_!Qq7g7X>*s{<#STP2{??@*1)4zg%iTxTl5BQMk`o7RYt0~6N-=vVQb@(@pjRqd z1F*RrxY`4*d1F1aPM*lQ6?B8X8i;3xT-xd7?t`i#U*i#8I~8yy0FTmw4)K?}SGwW3 zOVhSfWy(!6@qA0P{nAHYPQRTn_T_+fdi5V4WLH{xxooz_&g-HPE6zGnzhc9eCPh>c zRT6wcf({LGFhTo93kP*(jHUP=UH6_Wcbzpj>@@k$8|HRT*ejNGZdS_4vcs+kJLRao zi`}~)eKbkLeir`%TRupOM<(Tvl_huHuhiY|^6pwTTR3gJ#7g@_@7^pSJDu2fZ7_2rDkx?jMM|ld z0>3?iWh!(XGW=R)sDg?G@g2Y!_`v4~PooPjd>!|-nm?`gMA8Bcww?FDUTDDGz!g5@ zKB-mq$u+o$=s*WE->==fPFXE165a~M3+akvQC?U|fF1dqY{Sm8JX6bSPo{rDFW6z{ z%_)EW(br7ax&PkRe@N=4lB;G@bcSM1Q{)pW2DOrR%aV9{oQDj^kKlx8xmgfwfM)F) z8k4a(usLm$+?TaU>U=eyaIS@>%6F5Rpu%vCp5Cdb^Mxmk0;PwbyFW+0Qw*e!?g?JV zIuJWev&*#uGO>xI8;B`-=uYku(#Kx~T@j_wkJm$Qaf6Bxqz=eYpCq~JR;UFi46jo* zOXI1<0(J~H_Mbb;xg33L`*gCJJLg@b{$s(-NEZ{R3LgCQV-oAc1_^{AMx_LkC?%xPkH|_&Xmo590&}Kz4aDYpoi^8H666Mfe|vW>I9m3lAq_pJNk}MjI}CJqdu?>lNGZ zOOv_yRr3ck$>%Q^@7pS~qNPU^^ME4#R1DT287q{KJ*h{2L4r!#7w02M0dTn)sh19^ zdgwzcq}surFw9*+n_w@!f>$N1Q)!TLX?sAvpizAoA}lDkW$YLyIZ<(g7MzxSe0>EDwE}U&d?LOe6)=} zpD{tj{bBz_vhk&XN|_l{v=jrKNLfH4=UcBrfIx9#7;GWq zucNd;CI>|WjN$|7AU^S#dj{P(YYEAAaKe28NSP24p7-$XfuZG=Ws+9v3$F{yjdrw* ztPd&nzU($B&cIXu09!uEryiPw6V?s)7^DZzTV49E!>?6a2JYBI`JMM)2Bit3SD7a) zDa}&UY=au@D!SJ1vKw}GAmij@c)Av_v<2Fe;^|uaUeYz(go$VqIWZ=zNWyo?Qn+iO zh8n6?cFt%ETpMthdoQ#}mJ%@RQt03KMhA_4BvL6=&&BSaNy3V4_!zx2vY1E*9sX2w zsrp6NCc<4uCTCCNY2k{Xy{s@@T(fi{z?*Y&AYCL^SK0tER zdFm`~YSy&F|!YHZTb80kboOBA7YXHq>Xb~tj7dLjulMWqB434Y{SIpCno2$uiFRToCwdO6BjI5;d$H_*+&XJvqX0#`GI#O zI+^8hOLS$(e*gV`Tf#b_2&pwXRc<`iD&qnQJ;BC6VgFun%UnmSK&Cmj4J>0UK890{ z1J3Kg={mV7_^~lrb%AVnX@Vbn&2}((6q8MnOc3sXtY#;74U;PGjIL4lK?F8eP%Ydj zs^z44HImz8XW;+O-nYOtm7eK)#1oPaLv93<3@{)O)^?{mJDa*~cUnYIQMst#1vP*I5d;K55S6O}-USB)#Tz%#ac~qA8Aj#*zDYn5 ziRM6(jrQO9RZh+=aK7if-}}9n=UE+8=?*26tZeTpMFaaHzn{G~=#Ds9dQZ4*=1EI0 zbtVoJ%Yg*%{9{+qV?&_WY>kSMp~t8?zUykJw}&|V6=5s;q+}PEgLdd5K^f*$I-Oh( zQX&s=5x-P{+yxihiloX|&>d_O)!AEio|*X2gORY{=reLwr>4&@_^Dy*=qnNvG16#S z>3f9_%omx8uj%Hn7P^<;C@yhb?~oRxf|m4UO{{nU?v4k2!R0($CqEJ#akfS{aAJ6} zAdqyoF8aOi$A-1BHfu_ZRJ(q9K-eoDlHwv4+1>Ib=bbmt|J-qKMv+8W3@++!r)XZb z5KCYz=v3$r71QC?zuI!27@3bL@f$yI`nJv?`KUSNhveuq(ylj?e@4lmSKne0E?3~IfKrqZ>^U~zS3*A69ybW2F0r>Jd>^wHiv_dtVe=pJ|A7b4lbT|-|;Kg zCw4JpId6~vZ@kw52TXXxOdDWzh;Ge=rg0r<$p9c+Qozw4$$MEAnLYh*jKg^vwKPRjJ~vzL;=9F|E%-ko0Q(hDsFSZ%pKtT}8Ar3PN6 z906|Q{$(kmUb@qz#5vD-NSdLj2)H|a*;f*z&AKc`q9hG(YJ2?hBOXYiIqB?mZd+X&lxF!VV13aAKoleI%?`Ah}?}wDTk3G$DR%epKZ}wpyxgjIdLix|xp5d)CE%l2x zmXQo*rp0!rQ3BehV``6zDH(93=TnhAwhS!9a2h`Cw*(gVq;CdIDoU7=y zAWn7HC5Mefl265`QoN(dJ-S?gD8u7w*1+5-UNsBv(9M0!!Ovz#+ay6YYJNt*VXmLE zFH|3#jDms<&qziT6tLdE3_1;pXLZvzS-C9Kg+urG6u};Vc77*x0ciij(ps|vVcdb6 z06NCKkGXeN&j0huPjp5_Ao^)2xogAa_wjnQvr$uJk0~-lMeYv5S`n-rLFojQ91H~1 zn?d_XCbWv4u+Ixl^6aH6{i<0i#UTk~`6>dcxb<{FaH4%1KNHeCRia@#4ASj%NpxNw zlm>{Y;Ho_1f*OIcuO7VP9(srUX<+rjv+luJ73@B-N>NErj-yJnoL9lU?7L-lqqv>i z29eWjX(n`?l(=g54cp<_*ld|fqlV@^U7j^d%UQpshpSX0Q3?UKK( z$lg}b?Z6?Y{XRLo3uJqcwpKfrg%IusHlnzo?F*0qQi(aqHqm+hNr{z~WD_t_=J$!m zv2a@?P5qN8RIuSG42rDBcs^SwSrSDyP?4ySTjGjAz;gFi*Lxw#CV3`ZNwCMU-2HkG za{E|k-qMW*6J(4tfEL@uFI%&Xr<=Y$E}H&7WIZ#V*oKjjqlb(glx!P1@SxPKlA~cb zBd-$e_f90r>p}#|S+{gHvd}`a<<~UIE@LNTd4QEb!UP^#N7rJA(7W+^)v^_TTSb%g zbd_)REKS=ps+EF}n06L{@+L%%mi`MqjHsM~bg30wVCML+4od#(T|u&8JCDuAUrO{0 z!f1m|4;?2vtf=w9hA-&cy20-ChP(tg!i{2JcAM7$l=-^aY9uyn6rYi&L20F$u4eTF z#w@Y{4Ff=7L3l$?3)v#=VWzFY~&0*O>Xe$JHLEt05 zYhMviE6J0=dPz<1m=QMz{m@JoOhZ}_6l2Wv1n@EDYRhdh!pD>of4=z#!&J~KQlDTn zaa1`U=q0fVAH0r!4Pz<>RIC(H869}j@iTgfRW|-@ zgDvjHG6R+M;C1N7pVxk*gUP`UoW$fdGnm-02NJ873i_Io4N~L*%Hs9_vA`*kKC@4; zeMSmbO{@J-OdV*#Fb9(s)Dwt#nH5Qj(1F4XQIcnlgEB{Ojx>nzNH4vH!e2Bk2wm#kiVQ`g81JT%YwtJu zC3w3;vV6v{U90?&qs?Z4KKx~S151a#-=7#bPFC1(Y7wl=m^l0$7Df{ z2Fr0Vs-nc|eGI5r+^QH(Fs51=!|N>lnbFiW)idDqu%W{8E3zswqQ{)6*v{KeW9SXn zM_8DkSuQ6_JA)f#l{2j^5@srlV4AWta`f)cD?R<;WRiR~?9&)Y^7YXJ;2uX&ozf`k zfdsWIPmx=56FjRu>O5LRJDl&kTi3?eDl-1!5;S1gjVvUb-~Hml$lrbNnhB6A^kY5G(FzcdW8nZQ`Q;0% z=zCIJ(WvHBd&CP0dAFgbv_W*kns|lB%mA|E4~m8M8aEN+)S6Mpw>kRy>eeLtdkdq9 zdK#(K8-P+u25r8DAR`RLSly63g=BfWpi!L5x+c6kFP_yV>XLRr_m+lvGltXUQV41F z3Vwp;RYf94O|M)4Qo-`Obca);b3Ypz$Y9~nBv(T1u@5dQQ|WB4GU070$m(zgLd+WH zW-ve_hUGaw%{N+%2FP|NaeH?~ub~sdWZh!Bf0;2(Hc+^hp63JnG5mlv)XF_hmsw{Q zS#h4!q0mHVn-8-l-l@vzbr%nvHF+#M6-!h$9N{>kXXf%L859ZZ!S=uPz(KLuEnSdH zMez=JDq~?e=gZ(+YAfmHrouHlxXoes z6Dj+!dcZ>HFcAV)v1nt_!OXg)pIK6GSVY3ah}iHN*GS39c?id*_^t7}HFwLrJkmw_ zCB1>wqSo+eQIv<89SxQ2*M!H&b+2xBWIVSJc*YKxW}7jZp1Ff%cFcYMVxnOY$+Ih6 z8%BYVQjxvyUK9SQQ@RgY1Y@A1y-M&8g6aD4mCt!oXlzc#zPHImP*;qDE^?kY`- zmh)jrKAi;=HfrxKS!;NreS^4)vqexwd396VCEYgj7O!Cb*X&M-#j#-0z>ni< zSlU_`HxuKnb-(?~mUqQE@ATw%zju{voJK&2b4;3R7bSz2a0eB6O16DLG*3-qG!n|T z$^_l+NI}!QSeZ#zLHj;vOVoyR(;YN2`u8XvIN;J)?N=L;X=&avCUFP-5wY-KFd7Tn zwZdO*7~E~>aoDUaG2&e`a(8*&hGt3SW&UMonEO9T&LY}OF&h>}<+!(J1l6f&F4I>p z)R~Q@_g7|;vT3A6&upBaWOWodN<|{A!~b?lTf@7d`F@X6qHmhRwprKt?SYA)12M9% znswQ?)~AJ3d$f>JMJ4Aheb2Eqyq<1=7Ps!;hJf|Xxk3A-oq~SNtjR7m&I>;Gr zlAtvl#raffs$n?z$lTYK0 zhDkda4Vmq(2i9gL{7$6968|3mBa&#t0g`+@kf%{HFgGbwio57iayE#2E$5u4Pr25IJP>XbHHamNke5KT1Dj*nX4X^dx$ByzpG7pLxd{Sgua)}|wDBd}850+d_$bSo003R-EAS8oS0Z&mn z8-JHt5M;kUFAALG{27-PGBrjf0bbv%OzacGH-!=4srk)+$uZPmV&co%aE(BP9tsO6 z8O&QbROG-L-R{|}G)|KoNXdZ-x!6Oa16#qyx|eo^GTOe|Jysl#GF>9bAG2zFu?j{d zj|;-U?-DIR{)V9TKxMmhw@>tJ2myO^I1Sq^VG5uZjjP!J_M$Ihc>L?Uqh2q&{z20R zt^W{D{Ym@d&)OgV+Zv9F7}^h?CQFCJwURw<54|6O2h&crdp5fDksI?L2`gtl2;3m( zXPbMx#|*>dhJMVA%)H6Wwp{sH9oJqrXVw2*)nCX;8=kXv=~;^;O16O_@l+&I(M0na zU6tDx41|D6G&boCIb9E`0*5PyXF&*zH*%)KV!W#@w~rAlrg#jety`djh1ZI8d&xm& zEq=DUvLhGtKy;jvL7`?16{+Q5yzSrV(=7gy*8kK5G_kGUX@Icwr=8NjJdj+}eaesM zlomqqK{F8bAC1VTQ+;mxH*ojRojz%hb<6+)g5|7cF~U%LKQvwyKD$872hTXBxZ-vM zi?0dQzz>||{g93p zwTAYyi#=kdHApJ>xgl$(KM1Lkte>%J8YnbOyWqADh>xl{U6MTEs%cH(rE|JPM}W%u zfuvu$=lv+&nc(K|RB!C~13wystos9p9FduE5(7WZ(3lYmVWG!_gOk4 zH1IR=Dzb~25whKtBB|0dLi;J%K8oa0k(K0{P?HKsCiO z>Y=+LZm|2gYaD=833KMjQlOKSKk-#VfEQSElBIp#{X)%$t!_x|c+~Z%E6{ggiha9W zX-1Tp5n^DECJrgXD=^T^mDo86QwjUqd> z%$-$jyit7E=d7@U3-yyBMZp-NI{@K1IJDEH16a9QWoDdH#u@gBBFLujzfE8t;$(M$wzJR$7%0?G6*I~oQtk2BwF9OUWsMK!jvohF zrS8 zeKUsN+D1}DLsBhAPM37YtTgD1?%*ns?z2x=PnU#2RhGok&CnCT&d9uDe!z^mqjBu~ z+XdcFopjU9hgZlFvSAv5rNEeC+H^_=X~peSWX8g3Rv&%m_1Mrl@@PS?+hAaqw3?-M zsAk<1W8oWaBZW`Lf`bd{NVKSh9t^x?`J!g{Gbi-e$2=ZOZ5n^<<2Zh{>-Xb3Llh|mp>vp5RPwywK}QrB zxXroYSK@ki1{OlX{@39@$o!QGj=J`T;PEEca@SiyjjYG6*bT2z-0;Kw;LwN~0#=2u z0twY~+{>)bJSu%r1@JVhJuq3igSsY6gYPI0%`+Q!-vsb8GUJT@Qx63b|G)@8Q$CH( zSvy6CAI0v{{~?d<7~Q~z7d_kc&=yC@)>33O6{}hrOyBCZ<+mW2#J6;qY;@R| zg2`HN%4tqgzNfPsi;wpIGdamDu4=o>GjdZ89+xN?w6!%+kxL#6N*tQPYeSmD)DAbo zV_(}Ea45Jb923FK;XUDrVV})vkTivNkf_(2L5Mbml`2T&MDdD3REk@3&XT^+RDYE` z!SgHE*6LhYC45YNeGT>xIF2ag8D@p+h_jI z*M`wHn_VFriL7;Szn*(YQOe%zsFJsaUs`zHBhEKz<^U@mc%eV%r#PjEKKE)3&s1Or zXMNGP8Q|gM_17e+(A+I+#}w5>0TgWYI5F4_RuI7s;nVZzP)t2v|9qF z0QB&Xvcjj2^h+DK4crc=EBr&DN@QTF@i|K05vH*2(j&J*in=G9gF=)DGaCRWj5H%g zWs(3JynEN+Yyg`x_)ug_d5eG@{^^5Aqh~-7UP-#RZ!R14=fsgo)lC@2iE)>ixl*n z-|5WEt+0gI)LzBYZ{41}M^x`UuIODO*98vcCqgz<8Dty$mhW4z8;&xM6v| zrY<*ShFZf|snim|b$H(N2ibz65Zvg6MB@$iHPDa~pLN(6q*4ruDre%>L7_5@mFBP< zq_ZmmM(=?4#Ew?fWmm~}JGIG%?b-N%dwBTl8j#2B=6OTB6rjH zTO>=*l4AKFC(2Ek!qU=nVlc6nz7o(&<88HnRh>tcG=(=Vx(tQ5&EfYQ>*;&^G@;pM9+M4$iFU{2`nzfLHh$b=U3DA`vO8K5H1z-w3^F$98X6?DI(z-Pl72R&NlYP!xI z3Ivj+Yk9dMyJMeaz(nK|VOwt3CwFPHTK)omF!DMwPrzm>jgiKTXi-(k%jpARy4jTY-Ry zgEE7ZxDGkt0%P#D+x2pA==psr6oM?F{ zFjmbuJ11RmA9y0_>21Ey9s)`MgAgoCW983yNQ&mfF3|2*$?H5y9CovxNTYZI&}(pA zm`kUz)P9c~@yddQxBYQ{lel2vIak~=|GYO5Z0G?+`@!rWF>VEqez?t^twW8Sw_P-e znMQW&4MZ{}+encFD6RL!CgN@yL~<_jD+2N*=be|<`NRvVJ$iy~itFsHD=Nw4QJejF zKIewUAN4pBj_rm+VlEc_)mw*)`o$Z|NCvYNnC(ub(5g0ev0hBcpe#I}irm6l&chN4 zD4-x9L6;&zeSl#hXc3o+vh!UX-}7#PD{&%raZLsaR$y(X%^41ezrNCtFiVY(pQ@wl*- zQz-2Z0ajTFwubdW{~!vx^^u36<~P&E{w9V`)!1#!zMqjzvE3Oy`_tCHH{?BK;>6mp zO))~>ibW9}{934m*bvfy{bi_IQbaUV3}`0WL|6Dp(CDgc6Ad{wa$|fm>7tM>2>e;T zR-7T$pIQvFqg(%m@gMtqYRG%~Z1}~7p6GI-I7$Mmu*Hy^Yxq^naf4UvqV9mygTCt4Xd6*VvuZ z8}1{Ntc(JF(#SMko-ku!l>bp+A{`X10*RKxUds-Ub*=|UfrG|fK+YCm`x4hVDp6$} z5D|2M6noqetNkvss#)a@vEs9kO476-Li$>A#CKItu}8nOgA2hk$TNM(&xG*wIoHc9 zP_UCPGl!SW3=mIUisx>b0UpNf$T(~ADK&5X`F+DYwUGpGH{I>6lJA?b%d?K8lZ~8c zGQ^IVz0FsdD!`O}HHgyTtdH(rwig84+sSPwyrw_ofaI=b43K%t9FN|}Q<~g}7`dq# zb~76ho7J;S`)l8S$#j;+@#}$0WQz@3nge>4Mor0}k~@7wPEnIp(mR@CZ~bcKF1w z%BV-Ky{vTL0nQEC8@h+yEzDpx-NFKPYj~q`cW{wQmh2wq$UICH;Wb4r`6lNRO~%)A9UTS-O$1rt z?XSEi(}iTiM#o4O61ufWy^+HkwyWgl>qKWpoGlB^Qsa+^>A)EOdQ0tn{vgT2niooj zn+_xfR>YK3Uw}a4PSCD|ZxtCvXpDq5fP562U6S=n)imaBhV32%Hb{=L&$(7{EBMW_THn3GVLPn_LIT2jjuu?u z>;Kveaci}Q%LunqBmw`LX;|%Mq)cqhcdiMes7`6F#;HsdMvF@2HJl^fl`{_rI=JgY za@nWoTXTEqEdQj?lZu`Gh3x)WRv8VZBZ0d?gE1s^_##qbskL2rXRdd zDqk`W1{{_#tLAf*>@-DAP>~}NC*V^&4lHD)1)(=v9<1_IVl&AAE6KBpM)D|hC2B<& z0nSu3(3%rYkV2Op8t2M$@GtZVJi@{iuudtQZb&TQM5@I4;IyD*=@GwSyCZ(1yVdkM z-aTLC1IOjOOdfPhiuxQ^dZlo;I}F?5f9UNnd}$^Ee^dOCXU}2wniz+F&nN#_W+(w~ zvvZOWQFmmfy<1;W$(XC@eS8aM%_WLr;x)ooCdh4H}RaJWXCB|izK-HAsLFT ztTLappo#!(Z34={DAQ(sH5YnPLQk+_T+^|$b(gGBjOy=2Ax*HlP2|8-6$iUA6y2gK z-<~&8=`0BKvIapAv4e}$sF`#R-HkKSJ_*iMhk`Q{$0Qrb5nm{N82Kh>SSxfXWc2{O zF3xPZ>@BkyPB=KA6EHMmkYws6bLbdH&Nzl<%I2BhwR6#-sp%IbOUX87QDGadjVjf{ zUN$8Io3xvX#MHEwdRcjsh3l#|QD112yi`1FhsAIx`G%(;IRj?0vF3IquncWetQ0E; z=5B+4!38(V*33;3SKwp9_kjbl0_wNmwqKlrDL3vJUhi}91eLW zOO+T*!-M^7jSb`+$wF*0D0=S1#2h$!S1>{m!-_=CuwAw^lg@EY z5n+o5^f`UaZs21osfkq{a9pC2xA0?p%ViruV;r62{NNq-pLuL_()dKcj%lA|-z!bz zltObgen|y%T7}s8UU&M0W(ro8JQL=QmBFi4Q!!IjV`ls0s z(LhqH7`DqmRddB4wCr|C_p`3Dn!~;bTzc6T`*+XNX&n5$vG_031;|J|m1STaJEcIh z7lP+w6U&g}PM3;+^&nhwio|+>g|ry{o(s+wA7q5TDb2}pd&8uH&92dnWIJjhsDpx5 z19Q&!U*;FGmoA@y5t&OsPF>}j;2F>Ag@&PWDEqV0>Kx1gQ*h{65DR`_1dgfBel>5+ z&>5$;H(EX>Nj7Yp_Ulb?8I%lI8&p(eyEt92jnw<5LodIYH88h}1+|`~B5b_Ghz=Cd zoaSQ;533Zg{0bIZK^5#Q$TJ@HDKxoWVS+=(6edsK{t14eiI*{h$%lVE=VRDTV6!$P zRSzZ`C|Nv3)=`nTMk;YV=L+hsBe+=pT3pa%ofH$F_t?`+!^GGtth3Ds6H~<9nTdv- zX+~PQTIh$Gsed4(4XTMCQC$thf*9sd$&m)N0~)-Lt3%o8QWm;P(Gy%O#E;nDW%1E5 zZ7iO?l_Oi09uY=kVY^9dr^+2G46CVZ)^Hf9kLuR88ljtrStJbO4f_Bq6ZRfy+mNj1 zbpY%bA=c6`9dmzXhz+^gvR0sTJkq(F+e!ROmW#~Qn`Km#3^?DnQju88fQtysDD4Rv zbVR1ZZV0y|OK;72NV-_Pd{p67*4cMMuY6XJh2t=z!@~$PS{pZdttB^%gC@v->O;ex zRwmtQwrdkXK5uLqJAsm|r${Un+3AuhNDInP#LPY^0uNTDSS7dk^f%tvnu3+_uCw&M zi=MlFVA!JeY%tc*V=gzoVORie^0efYdWf5!tXo>DDtSEtIT@B8^mJ zDxD%a!P8XQqA&e8@Ij!4HxIe-y4{uSAe`2~zQ{)$#mSmLnFNY-LC*^`y~-UL#TYX~ zSngw|v9K(>N)qFTeR$bIH9h30iQhrA59Ca+M*vXSC{CG?tw7G+UU~;rK_h^bNd1Bf zE+FSD3IVZ$XkOxLa4s@q;{DO5YLX{%azH!u#kS58m@@&JFN9^_HhTv{sjO!WlkL`p z0b|(M2`QD5?V!juEsgW#S*6e}a%rL3@4k0F#P`oB`q^nSt9)+F*(SRK>7-ccT9*2? zW^S!$n=IMV!4%y{OaKQj97A)QVBGdp?fr*Kh6rKe%)ETqx#`kFR=?Kp#>mg6@CyDz zas%>D^>n&>!y7qXc(h)S1r`P~mDye|e*eufMj*#Re)G|PD*yX~|MbCnQ>Fn6hCw5ti;!rG+)UJoDGuHYK-6X4_e~DE- z8MhkqJAeG0={k?)=C7}ON@8s|IIq^5L$*;e2=i>BBJa^^Rw=uf4sksvS{xsb0(ckYt6X%aYPLLgl(cZe5gqs3|V`&m$=jVWEMsmm6YisD&N&;)W? za0iwAF64XSWH|ye6UW^SzI-%9mT4cLK)vWuoo_^<=X@`9-&zThbECW0(DklTK?=^p`-Y}GF zBs!YLx+WcRO7hH9WP5GkoMp#455W7H8*~s@hRa`5$v<~{Bt#D4ou0dc_R{M)lX_mJ z|4q!Ftm*fj%$bZ{Y1@4xxBlbtAAMJ6YyO;l=oUFXUaa4Ce}8gA&!V(avSx~$N0#lY zp@UFUgz6zcoB*+wBDcfxXkKCHmf3@jx8|Z6O@e2ML%GA2*~fj*`&}<90A9=E^!5d< z;eGVxx37V8Uo`Kk@RqAq;-(soVzYHM>vmujjjQcm`iP=TfW6rj^h2^T2xQfq4|sKO z3!D<9c@Af|$kg2vh~G6Z7w69y*A}1i&imaqI;rH3IVsE%sLbDk=gw)*J;>y}F#aCA z{=ui88x}XcA`5FHrBH35DWWv_5iW|7R4%%v5hb|@OKUZ|4@x!b1%32QVPmK|#7d?U zhmjdbBAp>>jZV+W1437Q?h*&$)zF-LStO>oD1f6Nz2{yW-ALq zw}zESkcb`1bDChuo*RO2!K0nd!*+MXNuD>@%XvzyfEaYlhM4_5$0L3>=0KuSO&53o z``M6FKNR^O*N2%>Orv8t893>K%e>v<9p~08Gc0alGWpnWG|xyWOe42Ja$R`WU)d9g zDWL|(8}oaiIwnmJ<$qXmRyZ7J7D~*~fq&5k6JWgPUgkXB6w&$7-&VN4gS*bNj!syx z!n;`VMA{&J?1f@8^>lf_**A(MgTkBQ<%(!hK_90p0xs~+)2XyYC!o>6WP%W#sTjT1 zk{j1jdA~EoDeXTw){fg z-{~;a_Rg&=QalY-rDG@iW0VX8qAIA!Y?t-ZcRS&_vw+&nDd!;J-#xm8 zUg2FL+bTG~Jb5!!upo2jdfq1J+dfvU9bA-DYGQy=k7lBl; zCorGR@@xng4n#7q78VGF3m-;+9v|zX6eWL`^X`ZzHB$IAEj3z`87a;0#bmCx-Brf> z9|m`RTL+_$np1vAj@qy(YS+W}XOyggB4?<`>vW!QhkXuAL5YYDAv!hGY87 zM)nO5zw4t>S_qWRMvty?#N*mElv#&4c@KS@RXI~x!BuEk#lV440k zjL-b%?Y~tSiU%`+itUmoCz1>Wey;PV z1D%6Z-Vn3|G=;~^UN`%R-P*7||0j0Y@&*u_wlEq@9D0L^13y}D+~L;DW*Tjx5 zmEshAnIt=?Ma!v=aAyU-f}h}-=p1XMVqDEYn}np%IcPH2V0P`M%Uwx^D$lP7HjIc> zcW|*54{7MM)GKIX_W2`&*be({D@w|0$H`<|FdNoMpZ!@^RH#HZjikQ*yWf*!W~o9O zUal1BO*)yBY!?MWjFFf8H+r3Q+w68snD~vc+Zf<~JTL1TiQmZb+4PNC(Q)~bA^Nx+ zPs;%UseSA!SlwE@u-{~yjM|WqjY&h{xi|2t42U)4Ykx)TM~)s&c2Kfy2F)?Gt! zD=3+_O*(%ga#g8xs=xATXrev3q)^mf;u_71;$aDZg;%c6j)SSg{OpyM-NVcuUA`v% zi^=#LHXJoI;&U{LlYojz+3h|^j`8yOu!;UhjSrR=SdSiI1Ouj`$l@W8kzYvn^Wl;& z3nt@p*l>B05uYPP)Gr+jY!8Htim1lD2)Hx$u}gWWf?ntq+$V?=D36h3=|RUN0WbwR zLj$p3y>pVFEUd=o?hK2wn@%PrX)v&6{w`)y{7W=o6N6l0c>RBEON=X6lMUkj;N!Zr`j1akOlM0NE*PMY}{6ovav` zlo7vl^rxB8p?ABf9^SVME2nMN|1eSuovBD!kSfp?+SmES3lgNM-u3kHMav2ouXyLo z+h^Y1@>a>aJKs(9w&)1!JywRpt#_>@KQQt`D5oC`?eH84l@7mOLei!|8OhjORskgg z)_@#ns{)?3Q~{KXCIgd0x9Fjha-;Hr?)eq)oLf9bgoYdD9wkO2W4jLChJOts&N|5asCs7$ z*<{1f^+G)?@1|t1z}ZPfwm>(<9{T*u&t@KZJsFrnH##KG?E})TQ{>j%z4ViLG1HP1 zS)Mn-aeui(F}-SDb65)vwAa(jj@D;THbMA4b02f>U<6-k_}uQ^mpxgBmZ|`VYc_=J zvtKE02&^QJgFtEOE7p+H~ zMz>A0HKXTSYh(B4`5*snrRy5R_c-_HXo5`m5K@(ADQ-Ao!R`)EBv05OPYBC^wouE@ zK2Jy3L~Llirl*%#e5>f~+M9+<<}X#VpAmNRG-+XIsVJX)T(S$6u~Z08OQ+9#BEj>EJ^H1W$qxHsJ}~t89N{*I@d`_UWTMy@g^7hgVjLUeHCcQ*yK&~* zI=s~V^70qt7_)4z?Jn3zhaS3`C>clyouwiXF+*&mYQH;A!*Q)zu-_kfu1}Hea*Y@& zDymkn6Fm3&G}uGdCvD~q`(tE>^KrQ{hnM8Jv`WBpf|6XzObY=N#nGnN^XCYZ73M>9~rw`Ei_%+G3QGuJR znpqe5Q6?*^P8x)xXf$!dM9^06J>k1h2dk#{S7wqj8#Y5NdO$fr$?7O_6jsYXpp)iu z-!+PNIJA$G3(XJ99gfqFy>9R~a8d;gK%Xc&!Xu!$4r<)VQe-gBfS-qb_RrrmT_wNi(&DLH4YIhagO1D7ILRPRh|*%( zK5z{2DK$4xl(IF#x_g{@0^@-ydqYSAjVdq78lMXK2sc5BMHG{q*(Q#mr)U~~i!p5U zE~tL|RWo8&35>L~RWB+NdZuQiUdYcF8jox8^F7JbP&rBOY?>lqxfXC+4f^ zUGB;f=fyI>Xr1(%9|3U;}BDxC#_KX8LSaXbHp6!mcThk>wh z2aT#K*tLUf7Cn-4uAsWo$IbHqpD13i#y8d(lh6}JjODPRMUWl<7<@23jXC^a>&!*E zSlhg1pLLM(S0;qpriZUHlnf-M>Z!>5H?My4&Uf!GzOwjKMAY{?7hV7E%0-Xfj9ZjI z0cX=Y8UL_M_3qc-FNwGl@#xK+Ki$80-;z^5U%BYUzqfp|KO%0?%AY@ZXZg2^7gsI9 z&n<7=iTLv8@r!zY61V89cW=FW;oa45rTi1z;Pj8Te6!(ur{BF3v0+K^n^o_{EsFj5 zlMfQ!Ij}fxan-+RPt_b*6#c=(Sm7*8$TRx5aELcnf>$BHcdJi+O{&Depb%r&OOnhJ)_H2w)P+s-up4;Jc&|kUV zuihnA+~K5L&Pxp29@HKHIY1QBJ>=aktpWBqB{ZJHMmJ=J+BMRsD#-9n*eNk-*iFOK zxPGij_n+uV&GD>Fuj~Ai_xAqsH)Mwm=K>Guc^tWvESn-(ROCgdIzbXOSmvvYO zerps-mTL{21OPOsnCJk%=sHG5h3cO{q=w|OwmK;qQx?R!liSW^!MhX*o?p3wH9#_x zqwH=G{w(q+3sa(gBp$<`vEs9&m!6mt`J!j9e|%yH(Z7c!J~2Yb)YsBy`%I>b)P|=W zBVD97>2x}ioj4oyqXy0}-B*ywLHts&;kqZY-) zn4AAl>^09&1L4_dn+?ZhtMq)6QKf?>`FSXKRC(s}yCGnvQQF2ec&cFY%zWS|zV3q2 z`~#A^^srqw-4A6&`(`w9F@~nS1c{MA+!#Z*+LaL=2F~m07VuY1wjASnsZV!~8CA1v z8lQiD^}o&ThDkyrxj0-~s%fB@=HcU4-T}^N$f*Dim9_1&$*-VUo*k2$7=O+$RV>6M z`B;%*gv@4LR3o9g{NRC*XkNAuqUM)m{ZLJP!=Zyd>a3=ZSwDGfs##zHHFKQH2(D8W z)m=L0GE(#Ui;|^en+>mXOZDcGY)S^TC%dUg6s_;zw(w(oduUAC^mA37+gw4Rw1bb%;3?I+H{CY}mRaP92*Uj-DzTrpO~Ia=*(R z?`Tffyc~8>a60?=&$hu*rAkyqZ=z1W(-ht{ZyWi-=O)?woj>e-{|~2^G=nrjKcsiB z3me3#bdgVf#5RIg*ZSUr{;wmx-5^k~PjdE+KKd%Fh%SE(?hE(pqftICit3a$M?8?U zO9vrxx6@_yv<>2->F4P$Bx|M(k_tAuzqjV*E(GEffZaK`qAqwP>z?Bfd!^gzpg~gT zvP*PGabDCDzKJ^b?V$x_^fuB2zv~mp$ZuIsZ}>hmBfW*c<+-NvU8*`qIN$Q^WZj!Ef|86c=jlj!XKN*=V>NgMH{NkrN5PGM#Lj#r#gK$945Z$C? z-4yf$MGh@UVrMG)p?L>EfVC%80XO`V$gTt=)PjB($s^nfx-Fn6WUss^Kyx5N(IzVJ z(}ehOzPm63L^0?*Stj)LoC6_^4WY=%w8!Z@4X5mM(W-gJ_!b3MdtmLR`gM@-PNh%F zvsrPnM)A|b-L!Vq)g{ei_kzf>h8wI=ToEwj2s+KaJ#UQOHe2{Oa4Cv~W0g&euiCFr zlBF=RK(hs?jc4f?L2b&0H*dcGvWJ07!p;g~-B65ZzyC>*#$Ddv+`)hLC6PW9~hS5Hj^kd)FmXM|`N&{nRdTMV|8=0^uo*DCKSEWd9q01}lvJv*Tt8BA&g<@_U#e;>SNpe@m=G z#L4e|?<(2I%&WHH;@tv0{OqD+&_A++ifm%-V-NWj_^oq*aGqiSdj@?HJac9zP_5wy z7bJNO5^cF|y3=l_kt2|icCiA;Oo7z1UX0}-^*>+Sf6tI)!)AjwMp*6>IapF#PZwx3 z>~h)r*s3{|Gxxy2?gm}-&cJQ6)KG6}g+U%2E=Hq0dX?oj7#R@i7iS)Ic<2nsW7(-# zqGD!1YC{y4CZg&`hToL?qtzqFu};N0p; zzhA5KNR}p&4tX}PmaYrB!XI>0cF6O2>5l!<96_>lz`H|U;A08cF@o2#RzOcT_mfWaW&K!-e>p z-y_F-N$-r)7UO~%XZ#p}V?uoz<9w+JyBL9k`qv-M{Q1jnHwND5cE`F5?3_n;SHri0 zbB#6%tfr%Rb@ux$bCL01KLCfdQ1hy7keA zSI82wVcIyAJ~kY8+o!i|PN!r*&a)lsBnHX;;Ci|yv^=2I;S|4$8~|=g%nxoPYgotW zyYvIcM}h{)gTTX(rBu^*=#&{}t-@nEA>5c3BIDl8IBqVo{dWcHFv7Qw*i8zUt)Xog zJ!kdMQ$@+3IN}f$iNU*ee!AjDcqOND<`WF5W-s$t$<&e4Q;}ajXdgs&pJ!omLKMB;Ll9xE z^j_(;&FwrZUC_&Jle+iCeP``?LY^@RU%$zMMc z;OWWV-T3XNC$D-l>08=UI~Ny!6HgotEaE54EL*g!eDOz5Z&^=QE?TzgP1u7sh1}Sp zW&0yOnwX8V0+7am$%}XW)9XJ9;OKzlx9uN}lU2-sWW)KX3_T!irevEal89X3$S$0E+`P{K7%v`8GoEof-*v5<4jA1F1FwIc&GLFtaRbrD2@++1^MSo<59Fr(%{F zWSfzAZU;;kyPcF#^LEcV1+zyE@2-7`D|)SFPZ>TBV5Z@C#2P&BPUbyk9LGQT@xEn- zAv~Ln^BM`)X%TRRe<(D`^PZ!!SaQs@z@bfaKSaaBuvK(BuwPQ=5zoP+Sdy)V5(ey2 znV9#;WMr7(VFVYJ&lefT!#H+r%7dFZ-}>9?yw}RImj3*51Y`u}^(fYm8}skXtEd0G ztn6pq-)xBJ{7Gwgs@I?>X=XkB$7Nsr*z!mM)J+6TOu#zPbN^7X@2wAY-f7Iv=$Yi& zcv%1&o?Ra6`Ji7?vM(rdi;6r9l6IMNS=erNvNYX&=~DcRq^7V;8Q>=gI%wQnx)cwe zoTZYZzlnLrD(`$gCL-^0l{dxNEG#=ZMe>5vK;=7A(JIrpr3V*WanOFmwa=F%dHNd4 z@IP!R!RQ|Pui%_v>q#p(ztKD_f5I>)BV>^}-Htg_YXhcI0L^A?5?^ z2&{V$wLwoAqsm1g$(im{{Hh zN)}I%byOt!GpYW#_R(&s>9n9Us}sPW9Wcq(Ow?dZ;yUA)c>i+#6~jr&W*r$&LK`#w zJ1E&UOmjuXa5VfLz2aCP?7+4Wt!xxp6)29Z2sjwh6abo3qTC?tm$k~IVo1VB&-bB`W`-qUPNk$0OYO_2QmrXbsO|-y#j-$6hP^w?KMN z?RDPy)|_=tXBDmCm)JG4_RwF7ufQD-unOsz*(<$T!+WK-rPo~=#5aI6p%3np<$2!u zyw3@*i~Mfxm;N^$Tf?sl)1*gWWj*#Ai=nRfJRQGagW!t80go5XaHauA&qFcGhZmuU z(QHH|{e57cVT_iE>9OJ2$Vl|In}+!aD>OS`W7w{c)g~L7zJ?S+=Li-RU`d^I`=uEH z=|!V&#^d+8HuwE)rT?(!1Itf9vk2tlo5LEpyFjlOST=G+se)v7<1FB;8ac9!6nPBW z9rDHi$03N64BO#)3HgKY^d&njC04))E#n#-GeTs{tv(vEpKjOz&BTz{aKWXKe(74@ zLDCCFN*VM`@`$^c-N8NRUgkJRazm=2mkEa!Oov?X<^Llnpuy8b)+h^m33qOBVlWJHVe(Hftkzs3#>A^k&CmuqsU~Bs^5pw>fprDWp*P*gR$Vqu4Om zn2h-_xRqHyGBY3H6z4C7MUO_R1@{B%^{u%HFw-Zq&ykDL0-rpGUTHmjYS97ePpN_@ z5^GNJi~{3keZDWc8MEhav!q%|-{%a(^DMK4B;ggRyGTH*eT3Q&B-9hkT50JsEP`>5@MqkpoK4Ye_xbCrl1fgX-^n zC{e(*Q#%_YaNECy_go%++I`qAMN}l&GW$3bn&WJ=uWHw@+GU#dDK zBck$Mpp&;_M%Wp$0W&y<8u4;%uEI)0}T=t6`AJp z$Q6LFUYrcv>xIDN1;RrY0=h+79+r4PBd>RkO1_FyNFVmeBQ-vEAjWcO;Ul1wyAXUP z*uq^WlYqqJ$Nf2XFb;`v!YWfQoJe$V)WM|eTi-cP)-jvAY}l&-ApCdd}#Z_mJmLX2-tl ztsJ7Fvla>wB8jKnv92s0|$&YtGR8%g|%`ILl*6k$+nawRW980;NH@6(8WVIBjp&~a0 z)Cw{AR|FJcF|)V%0u$+FRtFfIQ~|C%l*oIWN#p7#$umxdS~SCUYo|k^(0LFNf63A; zSeEQ`IW8G;?4`3=g9PZt{Hj?gq71)UNz{U2JM?!eeFs7Lp%Hj!F=_*=u0|l`z#Ak> zhaB4>=QJ5dm=ToBx2ky|3WqA%mK&BbJ-h6&VQ0ih!4h`ublIPMV?!uRPaL%$^ngLb z#&*xU(rA}5S)5GGx-PZGR5UsOOaL%?krg(K&q;Y*!IYr~YqN<#BR*+Av|4Ftmb#?q ziJar99oD~zBqj&#AIu%ub7xV!Rs0g|a56yxSOP0xK=*qwsyzDEt`F83%CR!Jfem{n zMpR`RofA3T5Hcxu7;?G}^GUXFIQag&D9e&}jX(OPL1_GIExGr-=9y~^L1E-+mW3XJ za6(hS<^{>@a=AwNw8ZtCYm%o`t+M7}nD#WS_acsrka5#IPQbx--zVA6_BXt%^HzOh z=KO@zG4ob!_b($|dfw^B>#w?T|SKFUGTwas5ibL$b>g znfJl3rTe9IKJkJs>2YB#q|8Ap1m!;}=n?@m9cL&?TvG)V{MDQ)dYkV62<@r;kd_B? zkU6H?Efe1l=3c9jD@O4DqM!HESb>JM)sY7-@eRRk<6V0VEQ`=y1ji&Xg{U-dGJ@$Q)AF~8`U zP2e~qGi19q88P1Ae%8=W2rHH7P^1yng-B~J^|=5-5QI=BN6NU5Tg zCqplaprjgfRik-3W&w#CrU|j9Sq-%;U^2NqlF39xo@;L-MSmt<&>WT@d|jyZKQv@1s7a<3(A6^WpCSWuQ-SIL zWy`GN!%qMp6CQLX7smCS_3+UF#4q))8%UN7XNZsK4S5MAE22mN71;vGx_n6P^g{}7 z$$4oq`wv6XRj*+Pb_q6?UXm@J5$il4#rl*L3y#a%1D6Bm@Yf`rJQ5Z~ta4m(U(_G6 zWT#K|!efdw{{)A0#cuX?-e$K<8Vw6xIT+YyMp?uOK*5OfsoBkBw4Zw}Bj1Vo-GWDd zG>o{tBJ+@u_*@R}aiFq;i;zn49Ja%?OIL7-UzfDpJr!z_ls)vJh&ad`=J2f7=gt^Q z&qaXw4krBm$N$CFnH4*4yJ!+K9(R`wyQ#bNOieN++encFDpKoyC?C6K0sfy zyTb2v8n(k?xE?DO3uuPQ^wBrp$eMB|%Ss#%={4>MhqyUj`2r! zQnD0^Y@s5%fgLS}2eFR5_Q}xigF=;{n5w2r#mXdh6RdFdvC|-5haH1iLDe2>tkOTg zq>6TQ?lEQLCwoqY>_~>Y4j+oR*h;c<8ab@@VDl&$uw1F3IfHk~^`V>gUz850pj%jd zbQRseedyN5X_T)DPn}f^VcS}_7&xyryezEPYllxg{ikJT7he_jNLG+uZ|LTPu*!}3 zTY=KmGJc@xujuG=Hr$x(GadJOnXxcVRvs`!_t9P7z9vKkpGU5xUUywu_yr5A=(XgoO9L1Aa|Epf; z<#v!h&JY{79+F0lv{a$@*MWs}4QCIHv({aD#R7A`1!l09+P7iSfY-ml4;RD=UiwOE z`yMwGZwL1S)*)CpBUM_DSw}ru-P!)X0vrDMj~h^eOo#6N`SJS##G)>*OO3k-{820 z`rNTOtb@Mc&`Y<8pob?}dcr<0ILZOJey#Aoi2%e3zt(f;gvarurf{D8Y2bId`R7(x z!fbMj+5BU}3!v3g$Id@rQL+Jw^p7l=HR)c+BXhE#w?`$%y4f5-lU${MzzOaKc`@H& zkA5i7)-X~dUoEbjvg|toi#=c`UZibM#rQvvV|UZ*oxhN7`9`z&OHJ>XO5Wkr$kt}0 z+eDa+hWu%Q=P?j!z-BXTx5#C3l%2>K0B-AM@p+Llkuylj!g2)Y=qXDVJ+V6so}uTTfDLI70+Vxnst8A0-1HC!31QRAjoSJoo!`x?r0zPA(flGF`6y zs0q?PH-T%XdRDnSU2$vf6AAh_C{39v!18;r{rBggm=zWwbZ}R?tq#&An=O<`HW35Q zogOTK2}Vxv)bB2t@#tmOc<=GwLY;~@{Lk&uD5xDsaXlcb_Nb@JV8z=c#(%R}ne-D7 z>Ao(E=G47T5Z57Nzaglnm-4 zcThj zg&!BBM83!alDjn4qo9KfJ37|=XQ<5(k3|&?Pvlm z7YT-e@%ma{6!gN&;AHTVTV=y`xW7#lD~@NudU@DKOM;#4)e{)cs)y)HDt*%#I#)T$ zCb5=^<%w^Uo8}9cA#hTd8W(6rcQV4<STpRZl-;g{Q zvJp&Pfl-NI5Jwh6MQwDZ%jxNI+Md&%e%*Uo>DSur%xTlEOhsSm6)43mWROkUZ~dw_$$}OqRMF%|1PcB2|1Z`kVb`1Nf7Aig6KjJn*s@MDBy|G zd22Ezx2xLZxNhYUolnN(U4h5R@({9rjTua_Zym z6q}3b>V*#_4GM!#7cY^sL!2FIITd%1V7X|w{7z(7Ksz5z2>y>Ub($@7qj!QtgVJQ0 zEz!NgHel1i_-_e!{S?^wAal>IKx8LvlP{JwfGb=<8zR6*E*6x_E)pEBdi;Y(Yyfbf zKjioqz%di!s;>Ye{Cu2)CIvSZn>hwhQK}8PAlx{vLkzYVi%TvF_r~Ir99f&f;)%NI zVL3DoE0f~6U%d9wAFQ%v*!S2XC_B{uRdlZ3eb5hFF3kn?!X<$fF-JlTK5IU|2&}=U z^k>qivdeM|(q+?+#mfZMqO}!f!$~3z#;Gr^CDK*rH=h!`wvu>Rp(gT-`A4u z>{cDmT~C2iWALg|M={kD*-yo7R`=2OC6&~vz)z)@qOIRGsdvSe`XA?HMO|0VpWm*& z6^^gAPz&Gbn4AAj<1e1fUHG#DUs*PXZhkX~Vc&yDEj(E#B|b%xL=jRxwDG zZl+2|1nPA+uBgMq!*FNa{{8k%(DxtxA?!w%ILEs8}ys6bc;adqWGNI%xz}Cv49*D%SX6fZnY*v6(5Bu(7$@eQz&!pTGVE z3ph$AbMBFvHwGM+tiVxEF()W;jEbuu6_Dhc%l%IU?uO$geFh+lx*4FS zR|rZZMm`V}_4%PVLzVEFsWtBq=cv+r6O{RO1IZHxe;`psR{<#5gOMeqUa}Uw;uK5t zi3BNKwt0YUR&$X7gO-eBPQ5qo zm5QL!egXCwpA6FaxiRfB@Naec;SLPorl0oB{7~U`kJTaj+WPsQd0EU($!EX*ki@?+ zW(SZxDDaX(F{>!D0;@2A>n#wA{xw(Sy#%PTQNP2Dj*2}1I)-d}cf0TSA7g*=C)?^L zJ7q&=79?mZM`0UegF|<*MS?}`BmO`gdz>T44r`gAH^fk6Pk$0&8&0tkwdmrPPD~%0 zCaRo;_}IZbBxq8kkG(ElrCuD83w-$%F>UfzzczVJK%3k+y-j{EY($Xy3>z#U z`|O^Pc)<#?qyO=z59Ko~UdGQqnORE8-xx3Bl$DoJOEFaxaD~L766+>l1?;2q=G3a% zX-kS`e`vaVF&IH?A_Q*0bYONkq&ler3a}V-!BfU-@Ip*Yr+BvrQIjs4&JtV@UJz!G z2L9$?lafQWC||5;@NeaJO1puhwOy9t-S36Bn|nlt2~ZY@8*C>Xcs)E^{=Q&#p)D!P z8&Xnfhg;@Kz?Fc4s5V6k)%=Ugzg#3h7560WlF)itl@M1_4KAz8r3ZN@#jSkYxXv#L z=@Fg?Kgyq9EZ80Zd)tzj`PrfZ?ooRFBIrIf9ULY)UYtW$;pN4Vhj(Ptq`muW*=t|a z>G0gu)E-AImSg49nW4Bf8t^?s4l@1qw{BGGPU}Zl+a7C)83!#=9KgHa*nx{%7iBx;^CZ#TBZLB7l zbVNV)FO5zRpfW1xQ0b>WmYgJzIlwmeV!>wkWt$W=G!6h&->_*!YLA+*_P5Ujm5qNP z`o;=f4lF{>Ut*0<`{)`X{xj;&>qm#!mRr77#>h@>Xr;(F0Z~{_ZYs6SjtlQq%N6i1Z`sDrBz1%XHb2q@i>xt2F_juQ+ zk~ym)yC*iOe^dL_74NqG=ejR1`tGWCFMe_9SG(sv`N563zd5aL<+loVia{el)eC|5 zQ*T{JdVW)6yR1>w>yPwk>2e*Xgma(2GX&d5(&gQUZ@?dwy?O=l1;6PfBjF+je=dbF|!w=H(H^^it#j@@`|bEkq$zt(G7!hI^6H1?Vp0 zXCIYedOBU62wekuS_d?_P_v7jC#aS;u%kkXm*rA}PrP6sNTC^NZ28qgRWKyD%_1b4 zRucbKYK%E6k}hvnApLv2H!Q5z0#GNTkBhfyvO_W@&1!5<%L?8efaG@Utj#OF^h2L^&O{A8mMf34Gg0$4eENRyK>n%!d%K9v zgI#e5?F`B&l~GJ7MT)4n4ARNDDL4P!1jSM5@>8%R<URzEPOIGlAm<=sy|lR!Dg$G& zgG|u*KmV(3nE4F}1lb8n=a4l#j2LEzqM9ovRZN;y6~Yc;=6!!**EH~C-Sd+Tj@dY!y z7quLUIn1nZ{a*XP3D!+$g1#qoiHisn(xe49%Q6^ff@&qY3EQBdpcJ|;^Xc`>dM4F3 z)z{RygaQ_)_|O+6TQ(CKE~VY_J>mJ@DZD}z=%NLeDf8h2^z_;YG;-zgW2#I^30H%E z;V-6}0E;7>z?(JQF_p7Gr!BN%qFLbIz#egi*E-LP?*W@*GJ4|VQ&0pU)z)P$O z1;02Qkd{whiZ)I-z-?0;11*%lU6W&wa#H|WuLV`WGLHbz3|s<5?Lc$*k8AQnVYih} zXMiWZn_CjIm}!>jC27Gxj|f9(niY#fkO}PtMA**6OW(T@J}>>;SDnl({kI<5ppeJr zOYH=Jdj9n?=-n*j9D`iL7LNAqdfy!1Ewq*-@fv3#XY)c%S=7adjY`<3yy|hd>^#f) zcG>5?Y9MxQ&Z*N0du^3rJr*5yBx|#x>SY*&-Z*X*bm*U-18V{BX0Z*%yz03+blx#g zI5dX8-k25VzSW9zonoM``wA6@TCEs=IqZ#*oE0-Q zM`hs0@Dd5eS+eOwSb38C8zb(49O>1_y`f03dfzvdmrY+)X^h~0RB_XS(FMu)Vzvk` z=$Q_A(JJtPilZuGYN4nVRvwIEH~6n&x`EF(kwTJRB%(@Vz8hF2JUx8mJQohQ-i4TF-ePTOZM^emc1bxE3GCy5*=-;I`FkxB##- zf5Se^$}Ukqc>5>Wwz0l9WKUjZ1+L8$lTDG0RNPg$QD)*S>hrDP>8FAVte==cj;R{> zZ4B1yV1c71ORyH0YBsBLxzL_nJrnD5P`}NtH5m4+nFbgJq+Ih0*cnAT{fB$!4%n1K z-z_8i-_-@*_wK5BwvfujcA4vLZ)_4iJp@YK7 z`4qF6BH2`2opPX(9>s&36p23OHcRA(Xl5|Kg?%^}4@4@NL?65w+Zn33x+wPxw>gc5 zmrv%f&)Q$FnvznLwri8eZlvt2Hbwt(S^U8@NR4AcDl;^nn?MYrdVVb@Q&P)mR&Vh; zs!U|oe13WC)^Xk3!xNKvmE%^BQ-N)ab6oqOQ; zyyo*-MNvqVuv1vb!JSwiT}sw*D=FQyLhRKf`)6oX`$CX7cD<-W+9qEeyvwI!#wk)S z#m`*DEE{u0d5@Pk#>h8JVufBOf%uYT$^I!eh@8apOSV;^eM5X>JNmS3oVu_BT%)7~ zBDuX1O}9Lm)1)v;J`Ma77O%3A?zM9q7i?yA_}6axy12)dID!o>JlN#ep_D)k^lFqz zrwf9&&AKc;?thC7bJnnD2{EDe|VOYqdLgb`|O@(l+=-? z=)=)1?p@CuN6L$qVfc8kKSD@I^FfX7qQyx&6-T zE)CfOHC4^(rRr0G^E;#^G55JUCPS-OY_qyRb&%Ud`oo&l`1dV&mn=7`29B47=){)* zGmAJiv8;RrC+5gGn1G>h;b0#H6Y?`qb${j8T?*`RLmqdbx3#}KO8P`~? zKh*#E^TQEL@n8Vi5lq4CBd{C1)#nSLSv^#t7b~|$8+@?395vhu#fUH?jYa5fP(6HR zR+&$OKel8IMD=i zey4DSq@1&M>_s0eGB1HX@cV&~S@zxz^}-vy8~t&-!Dsd7)xe=HK@m(+jbkfzKbf@(FK9uWHyJ8q(shJ2(`!?5|}*h5XK63H6?6f!q`3 z_;RB4V94#ag4|Y$DW=F4Dh_*7G+3Nq;<8Vf+6x82*JoaziACq6?8ch1VPS4sJou|UxmZq6=8VFlZR%6IXAL@ZSddyXTarE;Xy^a-N zM&+Lm{_7kIV4A+SIG8jE4tWtCvnj*~+ov_H-2Jk`zVtAo3_9n!tv z?<^aW6I{hf9J3>KH^h^X0=6(DE4WvOSOvl@PmPM{0p4n0L)cb z8q(Z23Y;OXV}&K^PaXf~oM^zg=C8X-l06tpAYVHuQK_YvEQ)M^X1gz57H_7j6xiw3 ztgcd4h_ZtJTu61%34Wgli>W_1sZH;oG-^D&JFV7p)`kve=7)IbLmhGsAVZDh+GFeh z;=TUPJE0bU9C>F=CCT$(fYez5WCz8RQ)DX@_lQ$0Xb(3|XoCbm6~9Z238PKQhtO}S zi&_=ctk!W-yq|>RLh9#9$PvrAcD4B$gHN96TJx1yWY-G&GE5VhkGmX5&cWtIE8<$v zj2gubHuI+c=%DSU-D5+=cDCW=qKxrcMWt!eToH3pjC2)kOrpTVce77f&P5gq+)n2|mCJ64wR7F6LCr6bAi(Y&978+Y_m;3j~tCU@& zm7h%;qiZ=BY<2~jTm>L2NVxz_tN=9n!q{1(Y?~6;WTiZqq}omkgQ=fxzp9I5i>ib* z^kwlTCF*KprTV7Wf~bD4Bra;vVL_iu!ZnUeDQhe_GN=o#In+dEjkc^xicYiuIBG#F}oCl2bVxVE$5&;QyRsjQY4v*Ya{n1mjvd& zo7EM<3iXD-W_1TyDvg)85FU#W`WRb}_&pMPIn{iJQjv;X9W%CmT^ z>pK4Imu@=5Pm zWOkC3fjdc=kDIrR&cXx$V>`1hea{Leo*OqX9=LbXw$I2;>rRs*g@>eVM*0XT;Wh`x z3(7+3#Pxhly*I`;t7tsYAYH{&@%5qc0@T35=+#MaeFXOMxY?jMj}w=|;)TK95S$zv z^RQ<1@VDV=d`+wQzI+mHme_zB%dnIoydz3c->j3 zsJZ}&LdXCJ8epf544#lMj5^JH8j}|Me01!h#U1r&=Am=jBImLGd>MV!s~3RmzyVfR zr+zlMV!;%Plj|i??I!te%>MX*6##cq3^-`pkkD$;Z0x7*;q0D$V|p@2?*sJFzW02q zC$)eI_MOnBP+6t(t^zgOTfD>mnH(d(a@Ou?C&*n=7=mZ3E$YFEIMgEmW&WwKZP2o{&KIc*vx3)gb42U7rT!_il6Vb@g6P8N{hVqhmM8r!LH&H}Ri6!zEO{`R3S3oMJ#dlh548&;Mk?-!9cZ5DXck}lT|3dqNj4$b^ruo;t)uh%@7on1%oO|-^C1I!4TE+Iz zogpJI)#gcHJK^l7&bd3-EqUE{$Vr|B9#b>EeUvO^=i_?t$^t^_gK|LYC}u51QlM9f zi(c$vX@BG)$r`_-bSe;>ofWprZUurQ3y8Ul7=YNq$T>JTdskL=c8_f-(;Kq5*eQI< zowa**EB_QwDWs_m`{vURW2=;%^xv!CMIF%4Z{-?z&FUiV%6V1v9$?+P8o3aX)R~gg z(jrkdeVpDd%i->d%@JLm)+RFmBZH2qq#G2?>LfuG3|h}`<(C69`Dy8=Ac75BdBJ7A zYmmm)4!z*Eym-J_%vjxb|AQ|dvh_|pHXvii-`E~dz@UUuE*EO2dFw>?A*6zhMHqb9 zr_2i8IJr`8YftP`b>Bg#M2y8t%^DmR3X=l09am}W=G)j z;$7}P!vB0@+i+@9JQxsmYEm${TqMxo4-;$QK2>QnN|3Jf(_kLj&DkrP8Fd5}1M7v& zt!{n0XJh`~ZFezjyb%wEg`GVQQrenoW6hdL*kP}M)@RAx86{MMe;d>&pz^F%Q9|K= zBVQ@!4NtffqkI?Ks zdz=mzojku|WL~lJ4c@sC_)UeyxlkYf>3)*KZg1Y$}F+GJ;mp0mz(2kcugVK8v?k_dZ_*9Emh-(V?tw3Xt6Q}6Mf$~xtX z!lB9rYHmJ**Y1EHdrLB~>mWW^g}O14;;l1z$OPmwe#E;Xi# zMnj-g+?e)|6mW~U8zlLmnD)Igshq1*0KsHH3Gb-eiJt~xk_E=WNDR8s%`YP_DH6pN zaLE4gg(k9^9XLEVh6)mUgJP)J6a%~33@Q#47*Kc$%pbBOLlGymtEUAQiQ436gRchk zdts2|y1b5;rvgG2=sg-C(GhkC895KNDTIFr>)R=lY zRb3Kz(J#^GeLdZ(x)hlVot3+#M#(bi`N()dJoEv>Dc7z^&OGfwtM}lSF8zkSA#2We zFDq1&ETCEZUQiOr_TbE&!3xdWD5iuWh1lCwJg)#M%{n+J-g``?Rh7$5k*sm4yvkV{ zn0yF?&(Dgw8u6DWyiDKu7#GJEz;f)pis_9!CoG<)Rn$myytLqJz+ZvqdLxs=T#I58 z^jrgFm!jDGd$9>}P!)VdfvuO~vD;ufenw)XS=qT!K^wpDo~<0Y$AZI-z&OehqWpIb_d-|!_ijLn z|MC#)V=4Yk>ReJouON2=>iM}`_}_2f$~Go5^zUs@Rf}~3<{?r*v%W^PJfu&VLmKJw zfXxy-x0nZ{>Xwh`BK3R}6>jp!BcKbSswRc1M&*9^Rqo04QwurE#teKM)a6Q-jlt^@ z$2=qrbiK5OS1&yR>f<;H@EZEm^Rb_`i^g9XWdq^`HL4!+K-NQ^$_i9BrfU@gzw9$~ zvqj(FWz`+82XM>^ebnE#FTL=dWiit|JTivdVdo=zZ~!-H^k5(PDaC+VY%kW*B1y)v z*iH~nGfJ}Q0$^&wS|m+7_+~pp&Oopa+n-WnN+jFH#n0~d%BCx4rOTI0>6O$fQ+ZlN znR4lj20kPKis^D%^D*aYWGm2JVh>CWUC%f28~A3Cd;QdU?{03fV0*xlDT_gf4bN!s zGq=f3zTNMI{82bCe(dcq;{;>$5on-U7qdJ>gMa0N7+b5PBH&8&XOUXP0Z48mVV+je zA#H`vSQm3eYWjuGL>MKDL-OXVn6{Ht1XjKE)a&Ee(zhOpu8AsQDkRwxkRbL8eAFsNHg9O2 z8^q+Js^*!J7|9%Ppk-A;vyYMFpRa>CQGdXZ^D7JigM0j?|!f9nH!_- z&(Mi9$jiKB%1$Nh^0r3rSJlv(a!w1x1X^iiYwjY|llr}m$u0mH)IbbTQz^a$WOn9X ztX!5&m&+=I*(QP7PoqK3UhoPA`C$Oh3Wd(Rc^Lxj1Mlk9@5thfkpNnmaO3&T}%%g+d8bZaA8BirQIrweGSp;Ke8$;_PZ?zL3|)8+(p25xoh z&pq$1xu&vMp<{7>UrV-ou>VqTHRW{_18MC2RNNiuwz=JW&9wk+@b;8Ni7_^#RkU#~`aFzXfM98-lVRzq5E{fIVL6-v z$S@|}(OX#oWYq6}@Y7)1E)kCfh@HNV8vc^7d|DT^91@aeg388gFmHL%|MV=7is&Lq zv)xnYVs}2+0K@JeZaBdX7*mSB%b90cerCM?Nh>k13+s7siB*f$Vs(OIj#1p2SrccjNj&ZM2t zsG1MKSFNI1ku_USgAhGRWF4W8%Jj4j_bs7qOu8J)BNo6Pqij=nyJ|rPr;A@YF@5ZU z1y#aiPN&c)s{jE2i2VwRfcy%?B&vk<(w%~QKM*ilus~1aqK0fX?f4e{IEM+hBf`mw z9Cr5T3(6n!##r#9Hom7J3q2S=*;e>TrI=)jtfb-+1TdR40wrC(V=~l$@bq-)6z68u z$*-a1His}VypM3|Zg!YZjv+y>dsh5}qEEg}mcoPmP+D*sB#zSM<|)ZqqKf6!7Dhj%cw2CxQ<*Zg3 z#j8?y)lPev&Y{l{d|T{_b06U3UEh1WNMT#V$p#o6yw|Z)i@IB4;2C_jN?Q5ot{Ua* zxM#=ZzEwjfkUlPodo-)NBY|!aj&~CsrxPj$4sg%Ow?gdpscet;s^CT87f25$EqGv@ z>uS`bF}R!v3?`HXXa=oGwlAGT#!GIS^WdG$5{pAByeIrUNf`s$JcA?OI*I`ehfFH& zQf!;zk?;2Ki$Nzr;_HFzAXgV&9D7u`OI$&>(U-*e+}xM~dJE@)Oc%a?YIaEGxXTeE zvf4ZIY+v$7Ty~1pg5$a0zWQH9{)gNGk@cxz=g7L(ERndw3eN=;lS>gD6_+&oYUDOq zX*3GUr_29-P5!qBG7bFr@PVvIgsyDA*AD1)%=ODtnd1BC$D%GP+74`ve=0jluDiol zm;vYvst4N(m%-R(Ap5BP(Sx(awi_8cRkWW;mdq}lbnDv6G;4Qo7{o?)ZI%5BM50xtM*bcQH{`y3|ZGD@^hUo27tYu8@k2FkZ zk!)5&q^!@^Fk!p$8k9VuMA$1b`)&Y-!+3SWXTNMnc6be&c4L|?3lAHo)q{64Wme11 zW{QD$+D0mFx%3PWCt!R8<#w;ijnR3N&F@iVtU=l>FN8YPba@H4i`M$}K^v8ku2OCu z3EnSLVRLvQ_hVLqK9j=Odf`Sb-}dW{ma9LdeG~j?$F-l>0!rH6APS z(O@MU1mQ=-!-nrWYizvCM!4|Y*b8@mY+D<^1{)q+H*Keq;G)mEsZbyQ{3{q{#X@?W zcLP-Gn*yzRI%9IX>Ij{}si%`kM|3`2{L_(v2y1{inQ&o?9j~CT1kbc=U$W&tDIz;P zIOXz*)tYgLV(KVTO~ozYVFNQXLh!fJ1;L$)bBYHt41H}Jw`!VJaa49xp;;EyNnhtK z1fr%uocmp}%?uO{XjRxJfBKyhie%9$=1wFYzCQzsjFNoX6j;Wnr#}_0l=t~<K2Ph}MMd|=b-E<4t& z*zt{Ca>5Uiw5EndQ9Ip9n;U`py>RE1O_#)A-O2v2rBOu!&>w+n+7U@3IERsEQ|Y!B z>@YHZhf(!qi>>+n>CAs82iTRHc<^F$)ygt`LNQP*c>=r0s^}dPyD{5`HIVuAx}c>| z^)ho+wpM}7vFOWm%RUt*hXG*}T9p<_n;feqVadYMYEwD4xpCI?5^-qOIm0MF8Xq)K zWm7e2v#KH*q$QkIJ{IAc12MI{a&P^_jsEw1`n|A3#SG>a$v}S@8>{6U247*F?44$T z*pgG?Q@^vI>YdU^N=F zm=)DCJwtVke(0xp>|ZZk#;X$6g;h_w7-5WF#Uuy}(Z_hEf*6cO8Tsw9_Q(b)(w?1X zxP8~4lqYXRTkZ`=tcZ5|B11S{~} zrpI-^3{!fdo^+DCVOmy zD?8X-NvJsV&n7gFJ2$kF=G_~*VS9YjM2S=rlxAAGo^bC!BL0|pd1AXFvRKrHb!mau zBVmBWC^h}pt_38+gN+iXh!1KcDx{bKisVvpX~CWJ2gpu^hCqY6Ipk?KM&nz|Jrx0A zl_tfzMW}fID`YCEfW_s=+&6&oLi&0}Xf`!BdUY(9*|wX7`3==(R{ z{V~^q6|YdQ1hV9f@yQ{NIS5(l6tjjRtEo6xQ9gJe+Y_$iX$Z^xK4IHw7!7wlF{EwTzv~8ni6FBtTggvp-blv#YRUZCtW?roj7O1L(&FYmw z8v<7|I&rUf#ms%aHDtB8O3?}GR?X_=^4s!(1_5{3m}isdtg(4EfSX=y`rEa&|6I)8 zxD<8AU{^rCAlDDghJ_dgIab46(Y=qD?$gW9qn z^49AfE$)_AOk6*0_q4n3Toz}B9++zoJ&<+FwZZk0-k7`InS!hGdz=zpxBM`{R~cbF z@~z_@l99lR!x$Otl8yjQ*kL5=!RfVcSv=Ek6#wGaWaChUbsp?zKtI_a|FeW*3Mo=R z#br!g!%dKw|87#Zn|R(I&+ZKBn|(yQ(YsJ>n3SVB$kY2YtC47I*_Z^07DS($)ot=+ zWAOaqkR_og0=+Y&!+mIS3POY6%0lj8{hR#n3gJSG#V!;S(5W%`0T^{}l46Tvn>=4t7_!v21GYrl zgLca6U;u1wUPLwnYeze+ITlqGn@`tBh8d$}528UU(<@%um>E`YCsd#Kvh5ON$I@RG zSBGg8hKR~p2A_OE8aHEdAKlbEDngsKu^W){`w+T=hY>E}*iF?MKq;Y&9xM7-=~R%r3uf)Yi3 zW18&-naz6Sx!(fP7SjLRIFS1YsZ$;`Sk8-MhWt8 zH$zW>Vbb!DcIh?{>TQqKsyLH%VQb*PZ+pt5sSgx!lTG znvFq+N#(2-Nr(7=_ec%9A`uUcj$E?xM(Zi&1VxTfaadjkDL>i8 zhvMY2qA3vO$J;>B+G|JHacV$?#+1v^yIA?^aD;P+$t{Mzv} z@Xles|v#0x)N%cm`n>iPHJ<0Qd@Yt(dBAX!5(t0|I1#npS~ zid!To+XBU{*q(|N0mzu*$iz7Cj96Ua;54%s(KVN{T6ji>{9MsyTe!$(*XzM{#!i`I zazqy&Sm899cC6yBr?a8Rv0JW5Bvo{$G$gr0BI}jr!Zmn$KFYg4&5}*G7?Af2>RM9b!3G4#X9vYR_EQY_ zxVwOjR+!7ZK+@$&{*ANhWe2#eeB>KF6jlt}wB`=hs)+dTB5^fAm5+l2SEUoN`&1Vu z9~2LCS?H%8C#Prijl~tJL5i}jwHzD`M4~`+ua?uKFiLhyi(JPQ;wnH4Am;g$4(}B! zQ!`5Y>07sL;bUi(+UY7jbcdQaVR4nNqW90Zz_~$h4@!rG2k0z+PjkyZTa*uS7%1Rn z>ML!K4ltVI6_00^`DhxXhPfIn5-am*hq-!0FAusO1K<4yDW-j~bp1@wp&)FPamDiC z>RB?)*6=3CbT_M6^4$3B(er!t+&$vK zW97qp9P7%sih$d)X7ywKI@70^kFJf_7>uLt85Uc5I+@Kv)+yP+hK%s}VK`Th7^i*JiKqso6^dK?;G(+^FZQNep zdI;edA@7VDB3FriDr(AL4`Yku{G28Zu7iDklwG@*=kyTXOqN(SE&@38}46Uj- z<(mRdsp_Hs*)=O6yJzj-JhL0r1y{4Ob)$A%G|Zf3L51&wqv>R~2czP&6)KKU%t4CO zQgPP;;sqCdc22}y_SWFL5)*wh6vn;vY4iixNipuTP4Yo{x>A)C_*B+1N2^HYn8kz& zIjB0WRhdTspUl*0ADX8EZjUAocDR0cS)3;}64YeX)0-#km4PSLAuZ(e0jCt^*xR93 zz6zR0oyVYqNE%oN+5dDBRjiOSI`dm!|L=dU1`kY0_R23}nNz=Foe%st5x&kFek6`f z=4gLx4(fsnge6~FVXGkHv56%+nlM;yvsS)Pa1NM{OQSU? z08%8f(DkDZ*{)e9#Vh?Z*X4D*HGYr9SYLze@z`^0lxeUHxmmXN?dwyKmBS=UgdyRI zh++uuLYsVfKx2fFeqag(?+WbmjTc}#(-hP}!9;9fHA2sj20PkUG0=EVs)YNz8>DPu z{xZ;UUz!is(2F~xjFksd0pTfT}p#cfl3 z`o(kK+5SU_@$nM{SoEdo^SdN|D%%y<;D7KPGn>m*Qpid2N%AQSK#~sNlfgtlp(=CS zsjxnIFUjSm%eQcnIo+~r0lJvAv0BB7Str6zLxtgv$)JvA(6gykA!2)Pipeu~GkF#K`H5%RhV z7*#6PMwm=duW!%vT3(tciP3EIeiXTr6p``(5EMQ^X4^i^PNg^4VUPT_6l^E_*^|z*c_R%nVhz2&=^J(mQ2F zzF|UT>}J(MPIBO|VPt1^hm9^f;|}-m<$rvCyluUfoi@~!U%U)~zvPG~K#P+Znnxqm zVNpnx@M!GO*z3cfA9U`Y;P3-=Zg{8fV)b?P+;7_*dFkJ6vlDC*p`IH!hvJGMQ(;G- zPU39zMW5&O?lpX<-U6Y-Qhx*Qpg4uJgS2>6L;|@EX-_DAz}g7&DNLW-nxToGks4DN za-Cbiy(q&ITE(tFqvTq^$DFH?c`Bo%irWt3B{AlkTw-Y){uZ!sw4H|jpc~cambTc! z$If#{o}Fy^lK|wDgH`>5r(Om0lE5=yn+?8Oh2~Bl+|Qt_1BO@ibbSO|R~6jAsgNxR zG&`or5laG3aj!(fiJrHO6VRnimlvv9`TIk+A{pIP(&vX^oMOQxmsD1{~x z&t9^!KAu~w5OF2+-)y@j#f10U|N42{Z-4sh`F|ACj2EFi*v8oDiCh;2e28a)D&(5S zunpIItjHW!A#3GV2~RV3#M?l|^Q13!r=9dIR^r#9gmPyC%`?!jY`n6VlIOR(?Uj8r zhyLG)SX+R2Y}CgN7#8_;Mw-dQKr*cxRHb1*T@llyXcT7CS5;ui?nGV$x=LUYlOyM4 zzt@@QkzkW|&a}A=1J4Gw4~C)otuMd5vRZA~2N|Q@$|U>7kaJcG|8a_e z$xh+*nR;*#YQoyzhE`ML;Q`sAWDYV-Zkhz!S>%Pg8+1x~f5s8|zV8#@ri4r41&}g; zni(WlEfU~FJL4EQH)$-K0=)FuR5<>E)uegudL%vao$x3Nnq>b^vx#hDx2yKte;IH@ z5B7$uDP}(fU7>M(v(va~!Hc8x^j`W{6jTPL%Wq7_ny%W(M>zdn+qjzoG$;#|H2src z1byigNNX($*NKo?7oFxdPF>g*KP03=rwyWtkk*1bY9MCa=bIJP419@@aRmu;U6fV@ z>3FDDc0SYf9FT7MB0T>9CtpSWMUekF3xZm{a-)FkV22>j-6@hrD+C>+m|BWdQE?a| zNFe!KgQz`vlk&u5+xKRn`;4%9em%ca*gL*vykXLXuv*2Dc}>y#!SHVJtLHXFH>x>Dtnpii% z$QcuA^?UKZeeln=UAZ3X_t@#PMQS<}#;8*^K#2vAju#4aQJCw7CI)#S#C=!$JedZE zE5*0u$)W}+#`TYi5AlXQsa8*Ns5ef|`cOl;^cbsEbo5s%-;+KO-BG4G+6ofMF1#kTvV&r0yOO>6!(BHbGd)M*!{aJXXo8@Gd%Q2 zEDg+8R(xkT-0vRj4Ff^mAOn*{F&ikDGthTx@HZ0?EsH|=leY1H1 z6yli0CCmZpD*BrIj#NLjnzvV#Kaw2(I6-XY+T|1ak^x!SjZq&j$UE*c;FxI6{E;LM z#i!}P)&Yb823dy<6q8PoHB=l#?obBO^mo4({`cZ*7dj6wnGegU*AFgtGrv^+x;NvZ zW;VqA8l<}d*Mu7-b;^7}61T`dQ!=7+Z-Rh>0C+W$Tzmy9doemo5fVC_W?T=Bnb>K@ z#kP}V&Ph@n@kCk|)+WC)>AJ5mG*wzLHdl-RwMB35m9?uDhZIdbOpxn&NnjE$SG+G| zk2s&38N3AoZ`q;lrH*hASwrtsT#Tl{!##fI8`~yZ)|Q?h0nO;+FPr3<0F%SI@@#@0dV6BD)nrBcNJl(E&P2c+6YFmjB zJK7w@$_jBVww+VSebjzwbp-0d#_Syu`N@x|B*@x*x42j_GqEeX^=;| zi()D$QbxsLJ{r4Fy6E+s%aFXO2{TOS=H_tk&w#20JakcbV|s5)_4JFvI^GId8>c30 zS=6l=WgN_8gER>e)?ka|)^Uf4=Hc8v=(ZfV5;*~?d5M%OzZA~m|CwnS{KAD_1HsP% z&e;Y^JT_WyC-l)ROXjrA=nug}Oc$>{qCW(UTeBLcw>L5wI$jU*mXBQ?(l>&dRV?S* zg){o{q1f5GD_>IG{ZETo`S%mlm&qq@j9CF$&p~G88pT|p$VDn{pZ6-JNrB~l1vKh-f<;7-MfY_c5s$K*zba{ocrodZkCuFY;lJ7`sI)u75BR@ z{&2hnWg)*a%qJT>I2^aj3O8FQ2I6y@sW?Oi_P`a><$xQUP;j$+R|n#B-iHNe?X2z? ztpWqHD0Pj%xIV3yKvzkQZx>zl#r0E;D&4DuHLS^h#RLxLG%JHNI`xO#T-(U)8&dOa zFU|{DA~ZIdLJ(g3YnDKRZEYz$Ode^>DzA9fm9EOmXOkt4UOt!$&-U%!|M^K4BeM9t zpd^ybZmIHI-<}w(OwKlnfojb{Xo{3Q2|yL3msEEchU8}yCnu* zPs~#_^U4iuxT2+gQ#}mgx-C^ueK{M4u zFwc-K&r~00R)gv=mgpsm5@%n9+9^GqCAuGDRwB;iv__^7op+UAK5byoywy&864*gZ z^)>j<*g~6oF?R+YgJgLD?XkVFtX8v8_5YZ2#8wN%V>e%RL|P7eL)Vb-8vP0MUoQ!* zn^j`!f!0)q?SV~s8|R|Wdi9eKtfg{@iS_xna~QAK8k}K#6sucVfn)TMRgK>ZwwR}q z{_pJ~I(7@12Zwa3tQNE~iYcW?5u~(%-3s}uo_ckO(}Je(7%>0Z&F~fyBT~prnPafOWXCU8e4!3wWPr@ zR^V~u;?IAyB*eC*&13C}onE_Q<<@9yR=hhy(=hI!xLuawT|(_stpcuRY_G#zc$%nk z7IH&&NA8&1=il#DA-urB)vHNy3K}GnMXidPu)LU9jF7(pX{5LKUEnlG&ALG7FQYP$ z3tN|51c*U6ab&t*eGNN{Q_KJRTHDH9Hk&ICwm5d`e34B&iRlO^hqY>H#8x0!+cCKl zc}2>mZsBf$kT7^uDC3qcH__EVSVnUouqDW~q2VE?(OEb%&?N%=()})>pJh8O4q<}rpf0#ROnkZJb=O!z zqAFZ5AoSY{GLD7`+hn_^x!!Yl_0#s$NoKW_jC#CmZNckaEY;23G4cN7ZrQeQP2vOtU2a2Da|CV2%W$pOUsZpm%rUwUH_E@b3MHHi_NFEi3 z?j%O>8>HAXh(sKa=gEWaz;?Qnhi$gNydj5JZ7;|QV?8d?X>`jIeKcLtObJ>S=%~du zYS*+94mm##f@P4iY5Q`>VTXN?6<9|9;ya@8?^?jJWbL=UN2M;x+st5vGZ= z>9d?Rj+w(P$-f#@6SWHL9X0r5gzbi=8_eESDZ5AtFPk<-!$>WXY!Q@BhM0MkODH{E z44M-~A?5OdsM?8I)lvofD*NG2ob5yo{hsYivYI&0-2{C9zhHMxwmEe zLTWiJ@+DJ_Nfxljz6~bb_EoU?9JX(>N^gGhV+)3Uyl>-|$@w=1L%miQx=AtZ6lpWn zSwZ3&)7u!8*&dogG}}WzRzQ{k$`y?K>PcIox&d4dm0N>%Ow>SC0}rpQfdqtBfh@3k zx=oR$+8e7Wj;bcNWp{(HsSWqk$S9*#JdkBYVZD_u3RY1xqNv}695h-5lD8T84SWq! zCFQA5+ACj`CBX76t>Ri_e`E&-*I*5HK{hGyl63jr&`vth#~c#6F+DZ5Kf0AppB`^1 z%&_CSxeaVXo8KXi4h6)1&Xeo5brEdhF&>=Kv{OZa^dp6w^eAk1L~Q9M9rkTjql%B2 z)@1A4&d|0Q4`m~&REwg>Z6ZL+dUt0^XlBFoV(nH_1(H9Bi{f}jhC+5R`S>kjq3xoVc# zj8oM4cB!rA#cRb+W=9939a#1f$ZD?1>|dc*L*1LEQ@UJ=o2ropfWy{+&BVFjTKix9 zj=ql%|8s@-u@1o;!p9xoCggTD=a1I>JG)`AxTy9d@UKo>;}?g)qN!oG4I z9Y&9h!|=+1hj{@tE>JPkvX*4ae^Nwtvh#60*zNelYArcLF?AHFrs8x=U3h1xj_H%P z^W($I12#*JsWycda#k_T>SA#{2oM^hYhus5Q_5@#Xjb2rT@^0!=_5Tc+AmjpC7IK# zE{$FnV?MHZ-unxw^K~yf4&3<{!;*I@h288Vi(cs1AT7zerT!zYS~bu^<42i!k9o z5)az3Cu{R%8rCy&>}LCG{t(=go*SeKJN}!GY;y!}$e!0uMnNxL7I=!E!Nqd zI;Hd727jk8Kja=N zg)Q{DN%J!ljr3OGx=BUBr&MhmqwFR`N>JpjkTc&XYl>X|`HhnKClpCz4X#r)wm(D8 zA;kW8t~yiwlLfi50blm5-~NH5k0H=MI_S~nQ_NNb71)=?s2cCHqw101^-@ zhutM+h}#rhq&LRB`%34&hi1SGHNCDn_J+XZs|$Grw&4?xEwHsy^N6s-sCmEF5?(=6 zqfp1IoOPYMD74?p+}#SuM*d?Xzg<=lgPLahyw`uu?e-0U<{Wrk1yTFsumbPs(Uqf@ zKDJohkEdooAfI^+;emVb=61c+`na58mQrNV0BPj{%OAJ-n)au`F?_t_es`U+kB>2b z6RsVO)S2h*)9gu|J&<)qY88hhnwayvg>xS3qu8nFOQ-h#bmcn*-!#sB{;Gg`gX|<(!JWK46O*6}PZzd9 zk{`N9yoEk0s{&<{B5n`KlpH5m>esB!lx9l%q=!U{!q-C~U}MBBARWt@R4i_j-vi43 zKHtq?kWd}HE)1m?K?)rE>95L%D1+1UK;bnkY@ z-ap=v=Wg#_cOGZ;+&$Nr|K+!TVPY(R_>bMiI#R(d?&iU6O@kG#4^RwH8yZl62^G&# z?;I8?bJ8In)*8$t7YQE9?)hkr$>L|n&o=oxyvxCwG+8U1Cezca#+kkh>0ov`fxj@( zCy~NBwKL-q1lX*YNHsylC2C9`2T>>NjIWnm_rY8SUfCisyF<+opT00|1$>MK57(#;!ArdCsVtY^a0j@{-- z(Xs#2-R%uvFTiB>Ig2T zFHCzN#oZX{^X}iz#`@gg6#j z#<_plAZ|EkA1SiKGrVorC65jB*ja;iho^B@2lx427O#TE=3_+(w>j`EF@{!*?UsYO z8Ck>Ykrz&G3DIC{TDp82kel~$A4H^z1@_ha)lXj z12#t42geX*Wb}VW{Kt1~HTUd@)d4r#xSK(mM8W+#HCVJrU0d(;_((q&XP!Sd20LelJwCH)YH9^q!ghpWDq{ zJr%3R6MyS$$fKJ+lwVb^l8OW9X;RhDnA=k0L?KLcV{7R&hS8986cQqWp;8Wv8TjpF(j=f9Z}|IcMyZb8NG zmI>y+o7A_$>t$d()F*#w{Q9+dXil2dDBp4*Fjo{Wf5N*JS&H!cu4W8JXwJfGI#z~uQTu76dq;nx!xODx9sQO|j@o?W8KgIn1^R(eoaPfsy<6v=_e zjcos{Oi6`o8Lv~`tiC?8B2cT?r(6=Wh2F;1acY3Uup|cC&a#2V=CkM=ZZoF|3d(cE zyZMm+Y`Z)0t}6_4Fc+wnh^jAr5+sJ1(PtyMy#WlwG>GKjT>b-7ixS>Rd)^iyWR$$ zo4%)Kp&EUW;QfWOH4A6M-~T^dH!EO_F1&u_>)){e;}_+>>>~~AmKqQCNItcK(N&7M zM3ELMZX^HntiuzNd7!rzKVz4Fsehlm0cx#kd5shF^lko4p2kQLy&ESOq1arLCr0 z5xpH6G|VIbFJ0xFHdW8oD(7Fk%*1{2=PRQm7K{n+34c#g*kR0r-I=Xc7}HS<(7t3+ zaiydUI%w);+h(s8%?$wrkHapHh zr_~u{HYg~{_f`P5N43<(Ckn%w}<9J*!n4`D}w*SJ_-oL@WgRa8m+k~Y~?2i z3dLZTu=r|Gcn0%QTfuTJoxzP=W5V*YUv~Llzua&xdcd@J&&?%e9_-hgv_e7+#TY2E z7w8V5pzf)xWlj724wr&r04*<@Z3oY>@_fv)o{U_ zJ30g_Y*4<2gd4WHeJ}FVd9ddSYtJCOJX7bpD&MW#r_`#tchVScqJ^prsF zQwyudlgJts9HqDX%VlxBcQ@35ej1Y&+{83RWs5gJi(H4aH}Yy^yQ~dpQtP?r+>twk zDP@JCfwjf@r=!4Nb)9__z0*gX9}cgS2fMg-c%_Wsfk0cf!6!AQTYe}8L_p2Knvt|N z=E-vqEv#nR6&LS(?}qAi*PGDOS`}(TV_yimr*}anxJl9IrtkerpSoky{L&%Za?-w0 z7^>pQ_t&EHwz@QIl7t>yfuOZobW$j06-8E1aoBf+|?J6u1JIm>!QE~+Cq8lKXS*Xe!hkT$r#EZf!CanwFA#PG% z6z+{Z0#}_9?GR)CG3wI|d|I>mXy9QLI+fQrczz987xaYp)NAqV`;sR-gOB+xR+H|~ z_te_kbNbyYr)bax_GfDdL+`npgPyPdsW!p_t>x)mWu){qYaW7jkU^Up1I6s6$ZoR= z5>{j<&rTJgwl9?MFX3T#Y!QfXcG5=`wUU7ZBeH;O@v8zohZac+?<}X^tA^e!+NV4z zPL8M!+rTV=UZUiP0{Vz(=Nyy#nz_iO9n>9kqLXB$U<1RXq|m&AME!Q z3s%1FA+-xYlb0?xMBJB@i{3vNl_Od{8^j%G2)NHWDaIUh`dAGso8|0dw%o@G1;hL- zHn-bH!4M{I^n})en*xhvd9PyZUr5qWmc@e|nIbF8vXNpk(2U1n!K~Q=G%I>d?eCfd z!G6V#iKy$MX#^^6J$=|aH3mBs-CWf6eDYX9#_evEliy7DwZNnDOJ8jy zE60#Rt0$gCF&ilOhsBw=n2?7FHO`QyDK(~<>Gz7Cfz89ZC=hXneqkWv;kw2MJM=u* zK5&Nyb-_jUANUa8QP2JIy?Z?S54PxdaTnyl+h03SzQr@e$)NjvMvAxQ%y<7k^4_lh?pUkup@BpISC5d?8$F;rwGI_*r`*Y@@Q`uf_wwSCh{`)aq>*O&3Vnb#Ik zTt?XxP(cGIi|i_jE9;;*u3%B4hzb&M85F^VL52T0Bq)ho%?$~^XuHT=zV{or-%rl> ze9!lDKIhT<>q89@@w|8*W*Z@cw$AsdGDC{_N)t8Zu!ZtOWf(F;`EAXTDp!;3X#7g; zpU84{6apv44JiGNa%WR1MG8d{shB?5gRnF?R-VVW8|EEid^Pzk&^okZk$5ZZ4-3Fp z*`s}$d0v=$!rugolA#|}l5BRca9(prpjzT6SnQw_puSy%qCe=KX3e-f1B-#K57gcP zPFgnzu_{B(W1@4F}v4Ig1-eVklD2l02F{KtQ3_ zdyRBNLlb`5#|p84jJ2zGyI|(w9}a#i$OM^LS$|F;`RpL$ybF8MUDLpNitau-gH~&y+~9{!5Wtn$FYXj%4)3bIF_&KirEnv z!>6A7?u-ZT|Jr0yZq84dNp3nZm1L!v9@LkV;vq%)z(~B^!9NnAmtExcx_8S{qI%s= zksi_|+Y@!wQ?I$FOqFN(_tICPdN^uT9^yZcT@BbhBUMx}4~;}dAm)A5ak}Lhf%kyb zR>!#=_L$t9vs-pF>fxN;8G22e>r$^+-f5+=We!A;uWMFLIYl}ovAi?%m(rVa4#WJk z{H>E7ZNjBqsDKI84&-?EqB@RT5%&|l21Hp*SMJdRTvWistC{}$R5yID&pR*7a=S@R zEAdm??M4BT74|63AHM!oOUKx0 z32qFou1>j2UZH6U?Ey+bk>C{lg}f>%Nx5uZh9uvvJl9b#nfcIKSTOhCm^p0ci1PDet07;_R|8U;4&wn|d z@>Wi~;$|psg&AOrhyRJ4i>W`A$`bB|nTsz0`T9U$np~SD$_{S}%@1E}yj=lR|FUSk z!cNO*<`J~n>@as}mpk`{e%$kc$s2z3^>2Jg>Zg#aW-HS(lme9ZPk`Q|{4u0>uW-j= z>REIm=L{4vYLX7QRDdWxhAOs+cW|rRHU%Fim@+ryf-?E|esSqC~NtC9s|?L3+ouRvHYkhfr@W=S z8p&x9l8NmK4(q<|)oGp>S}GtotvzATG-%<(3YJGFajyU)y!gjgfC--N z2P>wK``luXL@$w?pL2wibL&)%q$0A#v(xh-CxaA%s?(6DN!T*IN?pyrKCelr*F5AH zpWDPS_+)ye@Ve>C;tU&BTrVFV>yW@O$&9}I@{eJCmwC<^%b5Jl|9&+k#I%;a@!i|I zNTCz&EH#>$hy9eIl7hO{F`aam`n06uJ3E3}G*xbw#Vh7w-a(;yDQ7oX782*bo9I+s z>J@YK8o*abA*Ywu%c}}qw2rJJ74+4J$K=M`s?c8Eq6&H)!BfmN0I~!BDX)U=<&{LH zs`LE~{06c8%#Gc!W5f|_j4(aG4m;BA0V>N#9Ycr<86^9`YkU%YTC|^ZNzuD3573r# zYax_XP3!zODKosTgyQauHbYXvT`AaVUk{PtFScnQ82Ym3WmW^>ybBDiWNVLQv8xvs zNKPEsXDEP`r`|uDZq^0*l8E3wd6DSULM;}RHr9W|_$<^hT19Yk38};d^h2o8SZqfd z!T>j$5DjL8p*y=>W=3@L*(?8T36eL&pJjm0fIJU&9#8Wx2zPt0hcuBtu8!KZ;FfoH z;NFl;ssZ^vAHAkYxIVOlzaO-C^3}cG=U@iQr@LhKn}L|d)kgR*b!WHhmw$KX%Xub0 zB}@L}BC>}asm_UQ%wNpBn!}W$o+34vZB`BG2{pmPrDCo^If*p+-LE8yAd5(r8NI?? zixq{C3$rW`Dc);5PN}y<<%fg9x=xLDsB^gW^?AKfdIH+XdO*;fu(r@8&C;WEbtpcP z>VHDmG_M9qPGe~zd>9hiu#HYNLe;W;i^)bDjrNOQdTcAO@*_^(vh-Ql7#k#u>1YL13~2H5K`c;K5_v7G5s2P}0f*)ea36=YlV0h4ZWI4N zRJli{#{joX_@%Oer`N1l_*Guil%jgoG=?_*G<#dFaSZXUfB zY=IFFP~ul!!0D4eB8&8z$DCq8i)K;L-y|(8`p3Gj-CbaMWE1I&JR&$l7x_Y?7R@&5 z+}QUod&B(`!a5FqMz^p>R!idy1zSu0HKxyl=Q-#M)q&YbP@E*j&$iO!SdAhJ!(scW zb>L%0(O4n74F8`+jVC3^i6{$eidUd6P+xqfmqsC5 zq<+w;Ya>WkHv*m>Q7 z(^(%WE#<=)bhlFd4G{^7;pF4tEaMqi)7w$QKI%=saA*}IUUAzeoiGg!6Gp89xY0{-JgP^1 z!rLbn@;yxR%1_(KB9bze9*Yw@9OY({U=F1KQNw)k=t!%=t-j#>nM^$#q_&j(Nl<`s?xaHt^Om2o$|{Z+yk%{R zp$Rxt{qw$$Ez^OVc7e>0;L|SOs>V{?)j?Yo59h3g+H-m8$H0HnsV)bc=hX9CC8(f} zq)4nocPDU#d(ylt)K!qy@*NHX>j(b1rNARck3F-g7;Ckg8K8P>>3ldXOc>k=ebg7? zhct57HmR_Z50_7Z3Oe}Ha(FEHB7RYck7|i--hm1e{aOYPoRc+b0rOxYc69z@} z`}{MW+Tnll)VSixq6-76cn>2TZc5BX2~Vs<&i*WU<^~%~II*kBP@$t> zPRbnId%46(lY?H0Aa7PlBuqoxwTiY`dX0V-#=&$G%9OPd8^%e;(SR7$zjx!$-m!Ez z-VkgsxF_*bP-uB2?@?6B+Z!}V+?#X$;@cgzVoL8^&`6(>^7b9?GJ3x#Jz%5c==7v% zXXr;ENz+$N(W!bM|Iy(#g%wOjc_^bs)FFqSGlP9CnI=x_a4>L3kSx;2?WGe%4`h9E zo$8cpQ|PdasF6d0I$qnj20F*RPmWfjIJ_9DdqM2>fg6f^;UVU>Y6~RB5FhYDu6oBt zS43i2%7>9URe>Zn1gmnOfWmQwUOD7~8l{!W7S$?{9K`>Q2R^nUV-!f3566KSR_@89 z-}d~+kG^kO_llIinM9i2n04(mDWK>Yw$M78`z8aI_|-D~EdCXf~vP&Y!d0{^&Fpv@)sph#V@AW3>g z*d$%tE-n(Z3$M#7#gJblUhdHr+5|;|&H~|}H!3q)yMeP*fNfESdq7?$%$|zkg(cMD zI6;P_U$S^Hvbk50rLZPLPTlRfJ2 z-(dhPDP5RkNj-2{Q^i0tsD`Y3*g^owK{O`Fp4DQ3!B$D;j1G_SSP_PI_R`B%5A;iW zvHwkWmgnCt&HBN+Cd>1$H_v2}15Qjm(3{zkW0V4tJQ`4+8E2PvQpmZ%Gejgpp^s|e zHM*4B9ieTOB2B4Cgf|mI@fcS^K*1ImAj@FL1&NV5)u!OBlG`Df;RvQQdD zlyWf8eUUe$^&nblBx|S1+vNt?hIa>D;HR;fUWLm2xV2^bRAx1po`28T9odL8R$!X^ z&&PiFJK}1PU;xkdQMP3eSRuof53unWcF6e0 zcNR^vL+8f(n7wAxRUxIwr^q%cCYysAxP_j(I44y-5O29ET+PFZ zmq=I3Q`ZM|)7c!<8yb+;K{9+Hr#{HYsjLR0-|Z$iX^{roOc=fv*o6|*sJrW8=6{Wuk~(|fC=e#X;)4We_t zae_on=CmxIPEI3~LF*RomTd=Bw+{ZQZ(R|8d`-S{!GmG(?462iMOElJ$QabAw$hJ4 zCIy+ViM&Sl3<(;I>$BU0-E^mX!_1@LxjvEK5*oX{D#jJD@fc@SEWwLcu!Js>C09sT%Fd+F|-OLFIo(iSFHSHMm1w)B&dgZ0=F;|rgHpq(@EJ_L>1?qQ9;XYG+VQp zQa~=pK4fd+GAT_CSzTg-ypeRl8X7eEr6sT%P%SjjhQRfqhk?7VMeyMwsWRlD{7xX2 zqS6}Z^`T9!Sa_%~;9$TLXjMN08K2j}^ia(VDu)OL+?)9ozDCY}4mme%$6I#tlN|x7 zQP{BgkQr3DK|czKHUTJL*_vXqZwfhQHYP_Y#bJunQ!!Y9bf0^w==v;eak#Ne0>)zx zs7oWNxhar`H{|m0ZG#V#9-?#KsZ(R_6$}uP1PbV`utZLUtY2abqIU3&O^o#yutu&< zRY6yUYV*`vyxZj%VOuG<4U3t@#zd|}QCY-82-2o_RD*h<%`UYgMrX7=uyuHtM`sM; zfb)KrKZ>7v+_Dal(?;zWY7V79?S@v^9q5;6&(1$jW4 zgJUxc*}{Y^Y}xj2YkO9(oxC{l+fn~+GBk6S8fKHWv1TeKULid;vx?o6;ub}^sF*na zs?fg3%c6CpOLSBjE7&wG&c6;UC2kj_h3^)(%g>XVV4Z5CBri~}!Pwq@b;|T)-oXWB zbXJgFbDMX>GZ83M)xzy`k9QA9i=~n}-@fP;4dyd9(iQZ;uy+Nr{t*AGh-{>D1HkX{ zOi<{c=U6bP226>cvfboIdecCQ!4*;ApdGmtzKyGGow_{(%d}gty;){_rSSHWuET zc$Jl7hPU;UBAFryR1D6Ow>?2l#+d4^Qw_@YxuOz>ouvmEUbxw1Yf)kL&;*wskK(P|EP<1mdD6oxAb;S3)gz4LF z`;b|SkXP6U+kXLuCz%m4hRa29f8<{mVM2}2b-^~W!--J?3-wX?Z?%*HibNbR+70x@ zjk`bKn706EP@n3y=fl8+nc6HGwM1xzoRgtPdS> z8IWi2F?SJH21u;HWx5tOpD}+bS3SOMrje1e6NFww*vJRi_>&ccCMW;?^xr8>5c=2C zlg^TiHzstv&kS%ylmZxixl~M&a<_NAI5s4Mb1xtkT#z1m2j@6R@*k8HLHP*W-rVio zBwXtXb;m*cMX$L;(&V4d=v1Zq#R>XhHQ61sN46%oo3_7EdN$zWgVwWd?0&F#>Q7Hw zGJs!{UT|XHg@NbW4ja-K7S^dQhFta9FEaQnS@O{B5P!*pmQKUrvU=q0KDICKklCm_ z`RccCSPJ~ShzTcNKQYMoWQlIhNs@oQsBXdYqZSPcEWnHueUe`r6erjjmCwB~7Xrt_ zZ5?P3Mh%YD2zfrSjC$H}hs+2m8Jg5+S-I*B;Q@mSGs>qarxYa=6hH@07^_z4RM~z< z+(AYzEF0)FmpHA8wrNBDgD!)?yFzZv&4j6?P+jWw09Y~@NiN|ug(mZ|Iq2!61QxEHQ4*!WlQ-4r;XTEnqlCXuEUN6SHxutuE{Hg zx0{j_+2Um(>C=u2)7=k5rOsP1cUkmNRVp&Xy<0RWquerYDBz<~w{*XUVmv&nY2a;h zjA({w$i_AJ4jFc1r_6?AtR>J-KECm;%;djjP5<@pNZK3Yze26wQN|^YQUIefi;6*p zd9r716#Al9-bd~39`7c8-a^#xUP@hqrT11%hHnzLkdvg`E=UMY3~l1~c<+|=i_fV0 z9nSl)8C4EM6n6cM?Q>#zCY;RL^u42G1v{KLu~VCChLdzkv7REypfbl>A9~IE{@h)G z$wZ5^aI6dhJ6^J6*K%2t=Q#GOY#J&Sz_6)*TOZ23(o<=vzrw~labjRF$jH=-&I83? zo9_?e$QN8=>HbdXxGn-Ava9-*Vp*fM+@C1|IIB;|2^K9}b2n2pKt>n5B;OtKnx}sm)`~ z>cIyOL%tSj0?6vL-W{aWiQ}%v&4#>+QXHVjUMgm#@;se0oFkvD%A+gYc6vYL#0mDP zawDtUvH}+kdd7zzlJ!TH&Rgo5!Osdz=GD(DQRuGv2wz zNfI5Ov0}7sz6oB_UZ+t5ykvC?Eb!^5i&SPeX5V|GsK;yd5>uvB1Y zkS;Hio$<^PJ@S3@c6T7wvBgPTr|N{_nYHwQ`(yg?yXW{-ZlAfO^NraRhyUrz_v${V zTwtg7&G-ky4co@zWj)&G9;>l%-swAK#wWkwn~coD^BX=RpR${UocAvy+s#Z;2Bp|Q zkyI+CTXimUP?n|Wgj67;>JGWY3bI74l3sF14aH&SpAut`c0wIjb{Mx0ey!={&nN!p zswKaYjo0eDY%M0{Ws9nS?t{`ExSx0Noz?O)^gVZ|;EgPUY+gS}o6!V3zX5lkW-6PN zhl6*7Ye#nf+Jpz9K9zNHc5;rmBBW~GLjES^K^D!d;1gj5@g*6CyMH~Iw@BCVAvM$GB{!uGacq)z+9Cz1inxl@p*Fufy9KTN3kREW? zsSbPJo1MtX4mK)Htr8r6rgkwZXa2i>XnzDDa+CJWr|TRLDA2 zP`L-i9-A>SP0^ylvKsAb_zouXvglQi*kgcnQ{47HEQ7?sEIKwiHd?1j@WR$hbRc?r zK>ftCG@@Is&5m3jrPXTf@Q$B-U+ns_pY=Na16FR0^V;{I&**J$;vVwoe1N51+2Ofr z~g?9pO&y}w=T`z`CHY?LI&RJ4jhyghj!Rla7U0 zzpPDI=>T#h4!U(Hda1u^N9;WBiWx;;+&u89+~kaGND00`(w#Ucx7*B1%BK_{ZJSNS zd?`aJW09zfPKm6h3xc{Ka}1J?s%cEd}x( zR&<|uc|h9K#m4_Q#_qg+pcuonOe#3~;9;-{D7(6M#*u=t82HZXK@muM8x_RgODOvLO zVWpN<$$_`E-NQ@ayTUc{VpvjNB1M8G*9)W{7DK82_uTClez60Z%1GkW)l7!?m9Ch<~IpGvgA)Y zt?SCbh)(1_nbAwfNn(8u@#9F|a6S06_pbth>8vJOo+@gjPfsg|O!w1kovW$>=X^q+2WD>KLfwYugpy(+ZFEnD2J?37hTm3mgp?v#x_#{9xIVbwd=fW)89?ThS`J&;{)Y$c-nCEQbb=n(!Inocm56895`nNvu6?j&C&sn?C@qO%E&<&5eK@C zQzSa#+XPPXfCpaD;>N)izva@jsm5e~+yub(1@vBNmiNgBOhQWcXo8ggRaCCJm|h*V zS@lHR4fQzl)3BY8j$H&Am(9{Nd9vW{>{V2k8w4V4jYK*gN-sqW z8$a+b^A%K6R)`I%`C@mO0kn7WmL>yaO^arJs5J)tNO`F z593M+uXS?jgEQp@pP`_0bdzj-Anpt5JO(_f+>-piRBnkX1??<7*biK@gWAi~PI;cX z3BrJhoOHKAM-VL80fBj1u{R``-QzL5ph^esT)7}Pm`%xwK{P=>3+>(Hftu2j@dk=6ay4_M8%+* zc04bhhe4UMUOldm-+U;bN7AA~m!$>5FNhKJTtJ8@a;s(=cbL_XEy7E~Sn7bfd7bhi z&kRugzT&?ZAc6VQpxL{_GmFlrTSC*LTOcMh9PY_e=Xr0Xbl<@z&||t9UJL!EL^eUu zQEjbcB@dk$+=Na00Ar4Oe9MFgPkam|dk=tYQWl-(eMf3ky~1Ep7oDJp6X?NTZw3{{ zZaIo0Y4HzA^*8=Gb&?d=>K))d`3lza%cJ+t@Avw--B8AC;%vIZY{w2OZ^q3u;m<{+ z-b?bFc&hu%Y^vKsDWF_Mxp51=Re}^M^lD7n-%ykQ0zXhY4IE|A=e_E^R8T0-1JCC= zH#S;}yY{F7jd=vCr&SQJJ-AdV^6yk-bMnB1LE$yJnv2?+TcYyA?P-ndiV-tgGQJaL zjEphYP5$K3zkGj&2{#Y_Rl0&~9m{Sxv6ozHhNWGU0xCF`Kbw41=Wu8tD#?eIdBPg*BtQ={=-p+G%N# ze?DOGp1ds-pS34~Oyg`HAdL?+V?E6br1i}O>tAtZdx$ZVM8A@O}_2BI@LC+N~LWauAqce-#j{(Tq008 ziCVJq{Z*6$W)z0E%yP6Dda?gyRz_q}%cnjcTIQy(F(OX9`^k`EgNd?GZd|p~d$;ek zu*YGEp-H?RKCELNatxAYX&>wow5ZNaE9c^BDVILaIV!6VVtmN)&|(RxkwuUFF=Wx> zY3>b)B}kmsfG`aEMs8KaK`c>|%**lX;NwF2P;iDsi}kYd)OP}RLEW1Eu=X&WDnGnF zI2-aGYeB{`$VeYvcLUr(`AVqlOkXqI*hFA4G6 z`G3`r7jR9k#W@)qeu_vz6+MS zZ-M$(+Imr{`oQ!wF2w9R<=45@^rxIEA8oT|GOt@Y0MqAcX)+IOPXX@!U>lrU@s?>V zje9kf%FV8uJ@<-$gT7jx3Gq8vaSzj{fGLU>hFtFZoK)#yA)X;MNZQF35A8Y6Zu-7Y ztK>9&M|$4}%pT3nd6(H{?$-GVKzOaf&yUsusjr4<9AI_s+d@a3}eVToV*flc5lEuO_j9$qTG^Z=d{*>fjEa@v@AL+M4KOO zlvxL>ok#EU8_7a5#IJRYHVxIML;byoeFDb84DRr7$A(}VLdDu#?L{Q*}>xCt>2!t+{Aft z(c`?!eZsJ@W2C4U3++KL7FN{ZY4U!_*3c|bc6c8YYiySn_)bLd{n;p6j~6s0&mMF5 z0W)4EN474fM`$O1?|+=cO@Z{%(G0Y8lp={DYpEC`eYH~#+3TW2PQk4E{4<^yS)1VR zB)%-`-m}AB>kA*i{GP`!*Ur0QocWYL=Uo#h{`KaWOmbif(VKzd7^P^WNCOo!oHU9` zvBSy7@xc`!XMp^$5^C}A|B!kZ&r9ZkHp>YW7JOP4SThUQ8OA~^NCjLU+8MoEp_j!8 ziUcOkDsl=7=tF_2{@OvwkR)xYvA0@&BajEXg;1IdCW$iP4QV0pV=oA+Lp$Y=aOyVl znJfiLtpn5Zxqlo-F97Cc7g&MGd9C+vze``SEMxYDY!fpSIBS)lTCTB5Fj5W<2bT<2 zVNCU(cvER5xL<}thur&N{C%#Q$;5oDJ@*H)%86;RP)UAN8vRB}kw%eqBPD)6gA72- z@LG&XAQ+@K%sE4Mau(zD_?gEfEUnRa-nNy4Jz~Jd^HHpxWoOB#rITVMCTIxn34cqH zr$GMA=n=@K6i~rEgNnhdJS?-)2O4acV+Xa9Yl8QQH+v<_hV6R%-W3V zrh*sMAgl5?A2iXkiq+sA`@7+=tJmhO{yR$wy|W^iFwo`W1ijvg9PoqpM-)L3xecOr z$g$WzKP7UnD0fzIP>)xFS3kkJV@B!8{q$Nz67P!t@U{%*ePLd52VD-kC>TG}sdA{p zl3wo?r3041R>NS#x-tB;Hy&8oxJmnd@bHkOd>Dfw7iNIMmrk#06Kb!DOX$zIot#=; zpLnz66x7zPR5r=<9F&di78XD@ULKH7KxL^}S`k?tis_;?-owkAEczVj^~T)M{V;Di zU|g(R9W#u^^UL-gmr3jf#%U*?Zz^YeXc;njLl!g)k)I4+eDou3tE7zF5^ABADsR)Q z_-Jj1TZh|3Pd!#++wS0fMTZaSFRyvsL+?4T1JfzTDWz0=-F?N}9YKYhUEYoKc8?)@ zXcNEU4O-l`c6PCN%;Bf3j2Pv;u>PL~rnSuPE2jPrSvyu%pA);IMP>`H&6HvjlA~gJ zT<_7{QUj3C_Rnwhek99P_kzq|HRO5TQDSy_aZp)Ap}13C5n8O8FdJblWJabqhhxR` z0nCtT8Yo|F83lPmrV@sj3CdF<h&2nhr2xB z*%t361;+Xuv2uFJ;9CKs5hF2y8@rPx-T&aDE%QtsMwa}?MP$z#!(07}nN>MVDL|R2 zhKkw7y(Gub`ia>UksbV_s`9Vo!D3pw)4NHA4AHjGtgyKG1|XKgl+wa!C$~LzfQ-rQ zkR}Yxuu*Mu3{z>+;BvPg7BxGA`sUfc>))!RUPMG3e8 z50uzRi98Z89!tgsN6!Fd^}$O3{EA!O+3da7QnB_8F)$2Twn!Ywaqolp8tzM9eCKJv zeutJ(W*h)(Ai?;s18=zbefIB`s%;G7Xhtf|y^wn$+Cs3+#q{7@lnBDz`8p0JV8doi zBDoRKK=7N?MPq|@xsl~%;CIQin7Y*_yu^7R?i!XWLSHaV9w#^ufojzD^#`qa-&TI( z?E3cEc?3@D4eEY-pS+rG=EQ4W2D({1uh}y}Q6x(BAD(XtAbG1&iN%RBBqiJ;5hj9T z1m3<_eJQ}$0r{m}+4UwfkPFYdpW>Rl6_-GlI1%jLt(xupWdhmL<q(~R-2-2wvXS4?3-kr|BOqk8nYxa;b z$N)W|0x6@gM({K1e9J&RuYK4vsR$b2Iszcpqu?de;48{ahv#woa<0E<6uzyv}$1vcXiZwRv7Hx`|uN1g2Kk$ZGVgvC?*encf-uT zE@nuDFb-<_|ph}gM0|0ZB$XVH3v7JZTH+_N+mFu)6rm|K{>=vG4mryQnE9B}=EGk^Kp@pz<8 ztQElEk+yU1M`i?O!Sdx`z%lhHajRsFhhFoid~Ta8S8N~?2?`ib>v7?L!GB4Y3x3=& z?E}+F<qVuJEjv$e+=N^zQk>ew-ur=O?J6m0R^!9{)MH6B@XsT;&f^8{vl?3TYyY;1O!K`@t60FhNk z#heg!McjZGUPj;{{_d#1c}-b<`>>^Ewj+yDY^KO2%gH0;~KDNvn#&dMMK(DCT98c*i)jicc-sjQ4YKfr{L-<3Bc zk!^1bLh8&AQb8#|spE4DAk}}#k|ht{Huwx@vbJ#A+{z+47%@S zfNeyLIfkB{z5KJ--2LR=e1$?2YF1}n{RK&QV^C9UhMFyu0(eRrsF-Fc)=a^&6-A-~ zsGR~1F_QLIdRF;#%9~{Oc=5{1K!8N&Fvz=~^NK)PoStHmb+oAcA(z$ChVTJ8_nkTilxTa| zsIkJv%Q~|2J-1IPzHPY?{GwKo6Ppf(&EZv3HV5qsP7gRc|NoW+Zd|x8WaX4%Niw&D zqu1QvlyIs;kA6EAzX{S7qvs-v`b^)F3s#Dc_^h@-bwVo@LN&d?N zvI4e9x*b6_VTFt_Y=!ZXonZmpF|U@FPhWoZH>Qc`z0$+e$SrmgkrS_sSD8t-4p53m z6uD2uAfs+jw$C+9ewvE{X^%*SP;VL;l zypNlwPLtmW>)_|n>C<$o$FNcOr80TeE;&f0eyQvP%k&t;cdm$g=2p4kzdBVSWVQCm zA#gkYlz5YCmD^#zV>9|>$!Nv=^8~{mhNT2^N#*?N&;hPaQO-@|)Irp@fj1zpa_g6@ z^UzjvH-KnVoo|!$`aGQqKdwvN$A$2*9vD>xvy#K_Mdqms1D-zttFjiTvWlxaR10!i z`PStj*DqM|W}TLWz`%sf&>SDm(*SM~RI65Cg}(z4c6LUrG6+`L9;gGX^nb z@_qN552u;zk>Q=ND# zjTKrZ{iC~g%!eklJh@XQA(vmX6wxPU2)aWlZd2q2Mp-(-x@nYhTR#=*`_%_EOP_$^ zH)i~mMrbP{3n5rH;I7T!LClVu=?1(NgCx~I5$YUNb2E5%q;(unBv0n8p4v}p>3+%b zX)D9C!*9$zrQYd%N~}$jKjN+u6wqsVc!&txD7hWhER78*7WZ*Ko%h7Enl{L=jW!!} zA@LwKq>sCL>KVEp#4$Fz=DR}a+lt6b1UE@bBko9Z>7{~8WMz1&e-`|TT705N)G5yn zuMU0U2^!>mvdnPpa`&736Uumg5h=zF!kf3#e8IjtMRu*T}(XVEw3Btp&4M9#h7iW!wq z>ADMQvh9`JoRhE4;U@WP_1Pf7LK)SZL=LJ+UJcki!+7{ln3B`n``p99 zhxmtOJ;cC&_D#3r$XSP(SA6&F`1_1>#R@ajh1);s@i5I?KW!t6NQx8tIZ%yw6g@wO zQh+aF1ZwUE&I7=MKB`qQ;GWL?Xx+P2 zp(~ZKf~P!3fMR42Oh0YIh(#w3IAw>C9+%d5wFx7~V*a$A?0U_-n-$hvs6~1yY&8pN19I194NR4whcKeNj8CQ+|9#m3V#NR)3vpfZG;Y=G8kxi|tYG zawV@?SdZZ!4rUo(_GO|7S8`aWCr(f(9&*W|vHn6g{Yf9z+v$^E^S(b9q1!9XkgN#B zyc8_Qu-+3<*Cf>|%7ofPQG$PgFLu)`O;BWq14AP!!|N0oa=|J+&o|D8c^ix{bu0v* zVeusgW5e|C8tD%$t81{?nQ-DrB10{XgD`3Z^uBpRU(qHj4$Uo?mFAY_mO3wMhPGFn z7LHbFTafYo1Wb9xL17Dyv4_(BSHD-WR6U;kPA3Kj!+s|&-hpOS<+D<7Tdut>UmuE~ zs1Civ$?@ys4$AVNG}tP);Ywi;`+;gtS`60?CpPkjTu^@3Sm6aL=Hi`X9!_sa2<(IW zgFd;zzNy`8f%6jc0JrQ1Ais-q>n#`DFYakM@g&8Nyk-C;r}&whf z>NfWQoi2(a_vqXi_qoLpN9p3ItA0DVc<%_g$Zg`bMBJAc+kX7f|6-GoN5?aTS&gul zTa~fz+77Tu?VsM+{<<&1n_N#T3+NtKOj|7HW<%+O#@S8bpTS;uN@Np#-~Fk)H)^jG zmP-5zU-WLVynrcr0?&3-qS?aEjiQcFF)mT#%3&&I8FVq3#d7$BQ7AO%h* zeUz@1Jr(D=8!gLiW#5cVO4yNhKjoeWOPDoZMa9ZQ&MLvu`8}@Vu?|*`tSx}Dy7O@~ zeE%P6m(DgVu@|-%t|diI9LqauwrD;;DfUuiHx*-CAGfHIVGl`LsT}Y~=HaS0F3Z^#SLE$n;$;E#-ED-hx5i2)ja6g8cCO@WsVJB@#?IXq8kDZMSr_ zbXZCZjn8mp+9N@UqMhDtZuW3;Y=OhrQ^@G6{O4p7kF&_F9gUt*$Kwamm_6*e_4EI+ zENttvn@w3}2uPt6i4<7_0bq}voNL|~F@r2vkc3j_s<8y&giIplFV#95V*boToUyYZ zIcIKhELFpt7849K-&+1INO0|o>?0qY=dDttb1P-NupVwFEvhEz8K2Gbb3}M$0x@D8 zHb%bj6M+ijXRyKsl})dI+fqf{X|chen4TyyZUV&#FqmfAGAbCZ*HR<^mKC}eDBV)! zb3d|4GC>ew2M>hF#6W}Pt2%NLa5CPl7OF}-A?atr6YFj0{&UO9a`-6Y&W54yyG z+qaVAc`qg+}ch8~YNFwKoY<1M(;2iEcw{!BFbGn3j%@a9RI>U3l zW?g7Gw^XFpY!H=m2cowG4frk#`3(BiYq0MD?mhJe(RKH7SvmI}T_!w2PLLwsb#6Np zgD%CAV#zwUMT0K6?xur6&Jm*H76vy9PEK1n9e=OAeuFM$k>#90m+ruP*E8_DmPKBh zo)wq{YMN`q8D_%q!tF%>vb{1k2>t#ix6V@E!)cvd2E`5o9UE=h6U0;iovKMzA*^yk z-?kzW%68L*0f(m@hODEMX`f9mkM8i8m?*wYfsnCB$-YRDUP!2W|hf>z~rUxn&`2&_p)O8I*1D)N4BUs~27oS43_Q zZKdkI*8bJ?Q>%qXqvCjtbpPLN4?h^Nja%iH5qQsGAym`&SwHlalYwIaz{-G7laKv; z=PVOYKK`Wh5ZOM395EY@eUxGkMRrj!Mea)l4J=OtF)o%C>${=(Nx z=B0E7=N_qjV-`%8&8%b-r8r5EV^mB(sS4gADdT=rNcEFq>LcSB*1ahVXyPA;x*DGN zmQIx_S{t>^BSW%B)Fs;r5)M}*PAePfaxPlJb<~vw2V=H@Q}eu)UV}-BnUK5GE8g@A zyl39OUX$tHOKN9==uHUscQN#O#C(tpsU+$)dXMEVIR2XRrHVaR0<>632? z+66B47w+gsBY_U~xUt{W(m26h3FJxY)L$wacqk*%C)*K}!qq_zTc&DP$N=9vXg6pa857`IpyKN>*(DBUoec{zrp+G zCV%=rn}0fsd@+T5Y9a0BiB|;Nt0*# zCn>VU>-`4gMFFclvt6q_R(n2#bUmX%SnD+?zbf8Nqasw?)I<)J3`ya2J~J4Xf_q6H zx6Mtjxu{s`g>3rW-pgDYcw7Ao)n|fAc%wgS!^}9|PT3bw>1ywSD`IRLD>xW+o~#eW zXN{(QurlPn za?qtevRs}R{l%Q4^r7h{1LiEF{2VC8@n$v3&dYtA^wIZ|MJD*=@YCB#;#jHM&inh5 z60=2{j#6Y&WFr-W`30Sz;nFD|a4(5W3`Mjyf+1^?47y~7bdVC>pvw($zgG#b%I&l1 zTm2_!x@3=LOY7p8b7nN3KKA=VGfe<-5vli*JSUDje`W@TJ(L0pR+LjQxJ`v~;jX}B zAc$ktq;C3@I2Gn1)WO>qjWTwZW>;{VqyyYD^qoMgJC3?owK6Q`c-H4iG|JXtJ8d1O zRgO)K%`nq(jG1YCW5gaK&ux?W(fEE3%cq$!GDbSer2k0xZSAj2xcMY$)!XC-yEu^( zC(^Dko1XhA#RH1;QZWT|Dkl}9>8p6{T$92>RAw~99raLwyam`>D6)7_(I>}%WF4ml z2%{jFl_jbXmI*Nhzku!~htyh3ZNRoD{n#wUA}af#;B19(5o%K{T8up-j|Z?c_&m&fO*t!e5$B4)=KbB9S23wjofA^jNSy2UG+glf~R5l8>28|m%2rzQ|}=+XXDob{5rhkv}%oi zGxPhWzh{D8HZ|~Va>|KK))!_b>oTQirHCFRYl2sV*UZ}-emUx%dpl_g)#3%?aHMjM zc@@yB<(aBp`ks4@Uk1O*?Terl;Te1_iaO$r!;*SZ<<=&wC)z|#dL&W@@%bhxN^v$z z5dc^mY>PK^+$`0ruFq=<9ReXYIKdK91#~sHk#3giRJTJi!%--IwKSO*8?u%_6U=(J z=G`nU1oixiNTkEpkXi|LZ-oY+{cM}LrF+$xr7g2OXJ^~~@N?fab4>vI>0joRlAW)a z(*%{nM(HHgQHpAc9E5}y5iA>D-R%&D-!st$l>~&5z}+ zIPI1!wXkaUB$KI8Z`=DnkR)ss4MT*Yc94>HHgb8NB%DD&8Zm7R}xw zsNdvB8}`EJurX>cJnX;=%wPS1pZM&vl;C5N{qMxfEe3HuW2O}@?J;LUr@F^0gF2)Y zkswe8tijlj6_6ml#~0MhP@-tF=O8SSHq96cOOq#pE@(w0u3&(o3d<*OdPk)8I~F8y zwK;wr{034wZ6_z0cXT0Wrs160szAZM0vdIVx@_B7eU1w@@OqVJYkvRvXV`M2E$k5e z)HK<|Z;A6GSDbj^^weznxJxNuC3O=OsZpyjj$G%W^b|@58;uGYlt4hRhN5>N9hsF`KBd{H~0%fo0>V+3v)FX@;8Z)&9+# zo!$q7N+FnX4+=FORp~W#o=BU=W1XryY+#}iLyZH+21w>Z?bx&ECrd<@`M?Zmx5hMt z3aGm@IP(%`*xk|A$W&%mp^k4suq_NwV za^ipr)UO-`sXR)tl_FVKM=U{cP>nlH{hU0g=cO$am?W}rn+dHCNVgj0r%_Ql57g4C zL7p79r6wQ~-8NJi;bjdvY;DsLdrw@kJp;5Su^$VMlgU@Fb5!FEcKN0F^mO#GCK?;N8$$%-i%lE%Po z^^6`e-I&`wJ3M$X0 z7xM%AfMV=kSV4vQ#^jCvFdj8dC!S^))Ht!kdKM&F4aZAbCA$JQ2Z5|OvKVm37&igq z1SfoU&n*%)d-ib?6i7Lk5MyK;pvDG)J+CI`hW18}Sd(RIipf59o`@5NY|ok5o}-lF zFh%OA7!2}v@G;`m7l~=!S|bUe3u=5J!Rincycn*1GOLMd$vosW56YlPZD2(t5*K>B>U}U)wqJZ(jvr&301IZd0t*oIhUrz&?ZQsg zfXx8_+n{7_qh5MuBff6LefOrNwPE<0qWl<=c(Uj!pDy)%N!nB_sW#+d0Qr6-`QUMf zTZbF|ETHeuPs1LBIn>^d{K>Yw+mUA-`ID91ci!*hod1shhv_D3^S8x6`z6^tg&Z*Z znwC(CLW<;5F?&dYV!H=c8jhP<{Z4^zy4!wVZGN~;ZQOE!szrD&n}al>PWfYyDC>cR zgtkJKKg{OWHjoU-=K^CJeAXfGu>~%cUp+^_h80|#cb5OS_ryO2nBele@`fa`?KPt~ z)tOJ=hrId=n+=h0NBGRcY2%ACZ;328;(`#_c4^wqpexm3L5+8X9>p!6~ z4VHwr7nv(gi~t69UM>A8D5Ujx7s2K-6Cdvva30Qy z71VJyDldi@_)mFwZg-PmIc}`LFlu_V?8n~c%qC>YjiLu1m?oY_U;oC3q@LZX$a$w! zDET)!K6{2z01fv971QJ0#7B;;cDqL%2MbDKO@mbKj)2XoZsA?}`yiZB z={+dj>pkF}sy-|+N|j^}yilxN zMXrbo=qK*mrr&~qTn~RZ^&>lSeblm$O$seuT`Ab=gM5h{zG+jn=Y#THSGhHirJ55y z+WW!;#eHE7*c)TypjnC!B0(MhKTg1m4-_vm9&8}+Vq4}3Yb;vRzx+pBjJpX!`#<>B z8M4NSAq0#5QQL-@lwuV^C?oqABl zdYX9g^;&s_H9D;9cf8G5p=0uo%G_u_6FQdve%e}+HHE;}cGMu0QwqpyF2wI~Ke;tG z!>ti0faN5?Gf{L@a8yvjSU4JSv8_(E zU!4GjX_mS|a=);{=FsQMUlHqpdGVY2@(-}_E~|lY-nGZfg-yklo6&66)lTdbGgRI| z_JN6|05!2`3?{ZplHXc9&gsPl7+DEz8+sn!BQsZvM3%}8PODVIpwiGecPL;eA}ipU z{IqB&VEH%O7vyvMKntxqs!_C#92cg}TQPSJD2u1b5Am}ChXN*O^sOB7Q8sLRLuu{P z?5x>`dwXe1hU6P!G<0UOWiq8mpvb3GOv>A!Mijj(1RayL+&IChX;!2pTl@0aw6g|4 zZ2GfT+Y{WBw=J{B*g(R0x3>sGUYU^^j-m+z?!$Uixavs{-xCCWEUGC^kVP>Kpf$KK zSrvN?xmEqtzn|0gx-WVAAh8RzF^y3!l%!tCs}|M=qcmbK(H1~)iZr-cOdz!U#yG{1T@QlXyBHy}Dpl6(o|Cc|ifd&&hoVyMby)wtEFP zHlzpy5IR(Sk>mAASb$>GjAPBARb$WCh}CQ}+2tEQKVWHEUJUO$u_MZ0K#>1X2&LRc zxU#59SLgc_vg=V|Z(_}d)oIE0-}7TciSY(AR@y!&NenVAc4uY%IfdlEG4cEcGcTl) zQUC#a2dvqB;*}ZvVfDfm;OO6tI0G^*Ah;qa0hi?BI}mEaL)4>6iL9n?P0RGcbI^0$ z;+-KeK3ox59hyG3n!XF7uiAEbkzl8+gP%{Mid7L5ytZq=#lGg}6*ysWyKdpn=2$Wi zoOXS~z&2RxcAo6yo}GUlq&N(cwUB&T81RT7%T0TBepy778wzk@P! zz%=9qR*Obdny{&DL)(1DauA5aW!;UH-W5V^RnLFYCS*MPd;jAk&WU64*=CSkM=6phvKAmq*Gjr+ z%m=RqruA-cKzk%-l_fqC?S=rrtV64Pk!VeC?9jS0v2*etOrQ`+{w0d^I58*^&2$H2 zCo4b#VStJ$=k}9q&feLBvR$0pA+3r+ZenzCRGV@wZ$EXPZuZ>l+9}g})~f4xpwq=S zwm==P6DXBw?4(m(8iCq`5`#|>Fs{1jtr~Qbo29Fz=R)(;_OOrE89Zy2<)hkI zW6@sGrt-_cf#4C3r+a4{DVRba-C~r-vzJmpLAIS#Oh#ZXXg=cf0rFa15xd37uC2fe zGQQht8bhZbB8N)uYvc)DaPmnteGTMF4U+qj2H743Dn>$31PMpY(xrl>g4@bfg5x8^ zBYYxFzQ&H_9_{|hU-2rfz0K+i?z|z7$6Kz%TW%0I?Fx$_S!+2iYOEfz-FS%uY+#uNCH6;>yyzwB(VBn+`ub9YiC^3 zC33oj2cR|vs3n6kycQG0+T~k8bOYPtDp0Q~6puHW`--8p3MQi_h*5{H=y`m%*}>$a zhm(F~sm;S?AIW){7KA}!Xy`jv#FbHfk*7ttC!HZVrbcRxUbCIPH?Q;Fp zo$MO#HDTt5Ettj!5$m7k*8Sn4B{hK!Je+sRWGXjN8iDDj<+5ZRSexP6J7`akLZ{7D zS1J<}21$-bizk*|4|b595j{4QAvIK&4cw3V{03J_L-R9kE} zTo6|P3Gr1vpD7(4ShQ{2arA8+QR|*#H_Pm}`STxJvS=B|swmi~Q(@9LxPcGEv4U;h z?ZULFi>pH~cqe!n_i7J%?(vOvb9|lGkypWb5Jnz!@K1Jw;Izuv+a~pdTGqLCT33~! z0C#toK~l)c=GBmXcLTo?r1n#y^62>B7S#m_hBglIL2@6oGK|$gvLNP*r9#dKyJ?JT z=z-p!EI@+(PPd(In8Ug^sF%(POy)fVr?+Q%1Ec`t+gDGHHKs@Tua<}QggGlrPmb&# z^71iFKk9_lHDrquPd^9Erk_$u0Ul}r74sC*&1kz$Q$8FDk$zLFc z$>D_^tVEBccjD-!(4@#sd3Rs;!iq-yt`2^Da2{O|(M4Cy*rd^xMqKmGrMrdNTJ>R1 z9c<`T4`0WC$S^4m<05w>v?mv)a{&wJ#Mb$43Wl;p5>U<0r@N)A+#qNIV#xjp;hV@E zH@jIJo00G$HjG#7WL=o;$XE=gh=#|#ue-`I5@?aew#Gw?3^?S{A~Z^<5~^K}zN`*V zF`97;8gFG#)4E8o*luWG7(ClS!qB0eF0%v4tdEy%v=oGUt(+1D^}4k{h&|$)$ippb z?S81+u@n+qan%f#BOKSoT5ORWo*$mfTPH;_ar)>33O6=R^g16o7jqxbV5j$1 zNe90yXmilxu=Tv9oHgNQV-R|gbuZjixm>NQ*Zw}lVNPH@`88s;2{II23X zC|Qui*%X`vl*hxt$H~n(1JMs;OQ|JUUo~C^fv5NNnxP-y2layTW~coBVg>QRE9PDn zcfh{%ZQc=2oqw5d{nUEiqqhrH6%nXOc2strb60SlHvX1+O(Ew9!B%Cmt(@!fdfxG` zw!Pm6vW8s|I`xr-kBHM!EHi!-GT)Mb@LKsd3^ap2d4eKMjujgo&P@?vZE#p3;kI%= zIZkr?a3d?#AFDjq%yalA;V-NDr0cj3uHMX_5Jcw1Z(fZJrl@)CKA;{+=O zw;`t$3ZvN{AJz>&qu(-e!|p&}wiaVJ>64=JHe`&)f;e#mm%)M*hj$B8$Xd559#AN9 z2_opyj0W+699b%O$ccrjGTS-E zHaUKM+)w8mh9YK9NsTw^m~P|P*8o|GoDn1QvY&Rv)cuIc?AJYL`m85 zQ%ZYMbRBSG~I zZFabEcM;=`1+)P&V287my69?tfp$Li$D*Qj6Bq)hQ^FYEz`yXE;$#Z;z z`M#ge_w)T6@AuEX{QhZGrCav26YhAl_8&k0=Hd5;j%NpD2RQ~Oh6x-SSXg7g<{OrM z5Gl6o6k`JxCl0Y#?JL6)m#f|<#69FnM0Ln+QA$*gtVLD}B`@pVJc5jJurc4l-zcwU zIz6g5{qB{}jNAf6FP-!TN9Q0%4-=bbd{h_yPj%QfOIk21Oj~-oT-vTscZ4Sa0~>Ho zhvi6@&C-Bc+esfaDv~+yAsgYJ#^MVd;cmwrV&_siZLaa1*)x@6X%Tbct+W-cTP=Ok z&@q_6JqAJjUQ)vx_dN%j;2fri(FGJhA@_AjyKKRG8Dxj((K}23*12vkab(&5VA>4* zru4V{q``@6^zWK&Z(1q&C5p6Aahp9ZFe~Lb5yqVI$&h|op0bq`GKTF-nyj0f$lMbl ze*!RI;8rM2mKo4ZV{w#8%LR+0u(7b;vx{e7ywK8XJ?d%g)P<6{heC?z3j}xGRrKkY z-l(F8-Ju2|IgH%hk7{O4$7mEo2aQS%gD&Siinw#J&Kt1?DMW8%n#)(#(?d6un&co~ z682+Z9Y@&9jP1~iY%444IN^ZmcS(#1V}JU*_Xt_QZo}=wK1{9|BG*vzWQwe$;s&Hz zdO&Iv;4Pe15P=Os#@f5_U0rbwZN$)vrrROse*5uvzV`x`BG&0TJ7u=F{)5t1tV@4@h^2OXz8u|OuX0(F<3#4!l3A?93pABSpF*$c# zlpn3G3R~fm!|jBp@H}V#v~z;95K7$y{lTr_^SSDaf*VqOmy49o0qfcDz55bg1&;w8AVq~O_-sTO3xP9VwSu1l1 zXkG9lowg!?-IMRMe$j@%#_W901iqQ`PPLIrb|%S*{h}5#lXRStAEihg6_>+Z{Qb*N z4YMt9Er2fDcaUz38F0B8IpESgd%$Je9*E?cVw!6QtYbWk*f=aQIf?n%Nj@7@{d$`Wa# zClaoHPLhO;^g@T#PmCEaqr}6@Zo6Oj!M|8?Lb1V%6T3uK7^G5#6)`0N?XvmY%K_@6 z(hES@JTBZE&&`BQka+Gc#~*n7`MtxIA)HsL*3N3YNuxkv2&}o&2`k+UJq1JUC}S3~ zV>>T4&AwszU_0{IuD9-2ZCvxmpISDsy(XsOfSIeZgOWo?cq{ZlNISr}S7U*A0)M#& zq!=4Kt9YG)28m9%Bm&-I8P*p57lQRn1@zymtG%J{HfoC(QaHfbLd8G`@;00#JAylD z%~UwlDLW_52)YNnim3s{J^E-JkmqOFc0}3al|26eqdfA>3E%&3k;%oC{`IR(WYtLB z^iJ$}6q}(Vo04ZyB%O*&AXjAxq?djmT?J0*jTw;0R$Z12Ok3-p=HH>hpX*~fR64)4 z{vURzE_oM;?|9p9vR^P4qrt%ohO^r#>y{X`FQAJN8@5NkQk4($2ufN zsL`aZ{G#bUU+$PVu_3YIn0x`0+|8=(-c8~6>D45Ca*n(yd{D62V@q_ssFNJQ#Z;O9v zSr`8NX6V(c{zgg;9Bo8_T>^q0UzBa2bZJcMT_dDu@b;~i5TiaRn z&Yf?X+?T_@xp12td~Mv9HZ%97iIO)^ly_%J>wx`pF%#pm~_zU0{Ed5?o}e0^PtPYh=UO?sfIAajg26* z#n{+#joD|v=*>q5-}{N5306F@Y8@i`Zc8m2meU(!C` z^ozc;(Vb)UxI3?_cxB-!k5^rQ@xE_4n37Agu0mIR%4ARnS3t-Xo@+@q4Q%YPWhy*` zr`2dCv@|k&sLuf#ue#oK6Kt7qCzBFYPcM3NVW{bn!>-6^26Sci?!dPXoQ*bPbif_? zE_QJK;+OnGmdsO5>(f|anOgnsE=C*NE*<*Wq{2vDnrxvYp1*>d!|#b|msYsvM>IgV zAItByi<-uR&EMvdDYJEBD9kpqWpg7X2MX0JDh`7! z2Sq)cYI0lJHKj$|P4+4)`7$6*WQqu+!gmnz30ig*TxIpVFn*9CC{TsE)};pqF##Cd=0L;fHvNT zPkizNK7l=e^tYzOa_5p^^K2(Zpg=BB*tq6D*$CtU6A# zOd?Yrbu0XwKo1$pot#?`lhtv0$yMc!XedDSua({7F5zc--H5E`+RrqN89l>*1TV*g znBh0C`0dr7{j&)@|GYQr8|3_JgPuM!^xUN6kUeOp;ucYZUs*yu`U)nQR=Xtz_RrKt z9F9nkS4sz53OIV-<*uuh#&>nhO@Z;%R*xsd_zJJRCMehS&er-r#!P7s?X4l@rgsNB46Yu^n8M6t7;MG^^mOgd>~DXs0&&&u@~kKIwjvd z9=dqq3+B%J8ym&DF-`s`KF`)$tou46MFB(p_|f+)!!=I3kFpX+y6N2`tqNEPjrXX` z1e72};se~0NS$y`5X6QsB)2X4b0%-v7m5t8O7LWoxoWK1HC)^vS>lODusW)WUKjz4 zcu<}ToK^OR&X@r6%o^B=1Dmg71(*qIYd%e!WwI^VvVYN$9qbk|C*E(JGP6X7DEWSh zR72JqTw`dNQy9WJSjE{Q#;sN@qX7+AL*xAILX8 z6c|QU>E#}2liFonu*%Gx+aSr|&gU1(@}WaYqi73HCh9e^`TQC~Y#1ZMbI(NL-hb41 zv=i!{;m`7w%{Vi^@L|>eKACF5(Ra7}=JzCv9gdvXod)^wVY1jolpNIL3aGeGOXzGU z*UYB-V-HQ~49XSO!-}=_gH=Dx_@|3sTltk6K!{rGbpUEO<3oE$H>4DbflO8FzCEyw z_fVD?vey5$tUtCqN~0)~U5T*2c`{DVnBihR%XWMB_*BD>ETtS@lPX{O5IXLFfu;%i>UK#6T&n#qLP#0TWCz9bAJO}R3;9hBCQcs%O6-P+ zkp}wpKmFIb*dfPv!SXF+w-YAfie$OS45lf~1`?D4A?TWoIL+h3% zJM4Fafmkq=NmWpd?kC==p);V9ATdN6d|zBat%swV#pmVCs$H@Dvb&-L?sf7ILVq9c z_;=DOn0(GUO8}edWAef9L6-{AEg85SOck$toQQl` z0O8phGVOpHnZ{$14wRR$cKf56*)v zYNh{{Xw-VRBhK+}k#7_~ppm(?DRMka#H)vg5rc8djx4uyU#)nF7?dy&!QrE6c+n2E z@cr&(p@uL86q-R1TUGE5q58IK2@pY5d*>@Nm`$8|X_KNw93QIpg~)~RstWfTe04&Q z-ZxLFgSr%r;!7XkM@xQl)65(o09h(&^HAq7c_dHS#%q=JksYKYqS7tPHOsS*51oiw z`mpP0W_VOH_Uu`6m!F`#E6Mnqw=O0;8;W;fLtdQGnq>ITtI)PwL(YzWwo6b_? za}q!r37F%Va@Er4^l3HYRU2*#K+Fx6gAidcEGvLeo!^e*f7i6VdH+Dq&&W|H4xU^$ z+pwIcvIRfd5`KziEf)c8tLI64h74?!jxatB>=G7{U z=lg)C^YYhlf%`%LWhfA`>yOognmOLrtlC7N%<3hhFP?~n#G~=apQ6<^+_SSRLsvij zkWsv_j{Bzc;;+Bz-i@)~zeQf@HnfnSm^p5uamm3#-)ev8EEv~{UYMj&fMFC<%l;PY zU>IqqHk-dJ-p%CQe4;+{XOif|-c5m-ce9?7r%_}z6?ZB6##?PfD`-#!hx)aasU9}DWhaj7T1gC{dO3C> z8Mw$Wg48an9be>_X4fVFjRgs#ym#>zo<)}7lh*_iR)Q`0(6!pEDE3U3V_iWq_w#qh z)sj4&CmRs)^xE-!@q?x_Gk$7XKQy=aPsnK}UO#%w)(<@;hmPpWR2+8L<|(g}9f4^E z_KmGEmqky)^>puyeN%Ua_R2a1>aw6OpkeWn-%g?W*5vI>GRX)y5vo2&@C_09O|%jO`U;@tgs${0kJ8ddy;cXQwZg5OG+9oB9>~9UyQV}vqOU~UbYB>u#=3;t zGxzzXO+wNS@L}?mc$G$Rku091QRsy6Gat(y`WZC2uu^V`k5*Y8m1ZEFzv*5S0Yuep zyfoPff^{y|qJTCaXUGS2)XxzC$!am0u%kl4vrHvAZJ5qV+EDL%C<&C2-qkUI6id60iBLSG+4AHZmanWHEcX+@QO_sy!Y68O>(LeoaCz z02M2Qyz%F{HQ%ySkab#wSkZ(%3_*}jugX4qAkASgOtRVI2qc>m;3#&Rb1_kA+z&Ltc1$V}lt(T1F!AeRM&Y6$aVDfCn-plpjLIfzdxB~B zB##+b!w0kD>#4D_G)vC^=cBhyyg9bQ?2j&LrW?4ZRR(dD>JY4U1BH?_Sv6n+mCM!m zV!6i&<%Z~j2*aC3!D9Cs_e+6rS-1OX$RCV2$X%I0#iRY{r9ceQ3EkX5*bWaVpWuR=W3_11{}f z%lOuS%jU>F!GO!!z#XD?=%6nN&I%roZt%P;(~46Aa;e;}Zu#oYIYy-k2Lj{S)Q$!s z&+c~ItM;Zkm$@bfw%qT8lytl{Mf!`()Zrdd^7|CIN5!@AaL-(h&4OSnvgr!?vZ#|r zeIkwGmQO{Dmfr@U{_Mbdpgrt!uZXD$DiNRsY82@BJ<@*9s^E=sLkFwjT8*MYQ6j*S z|D>?3L@(Y0nvEyPA#tUAbwsn`b1*@v!ZYNa`!RWGP*PZzSG=b&EZ8hB5zNJM&|Q*z zFk7e}3H{U(jYlHY9o?bCUl9kbi0l(5bB~eY$Q*7er;ctFT$L<|NDY{qNEI@DB+GLk z*`m8U#EyO=ZRVoZh`tFeaT;Mq6N?v-Qk!|rz&>FRZ5zKxca z%W3^%E4C;}*iYxpu2m-Rk<=S|Qh`;q5Y%bQqwdcz zL_fcicVhcwg)|x$qHOw<_hDDP_{5Z4po+N&8Hr-jK3;{v!ya=Rthb)iay627bKIjqi#!n?(4L1Mr`&Hp8J4N9#Q6MV9#{=Yww zG$&4&mz&`_pOSB+NH+GrBKN|x5@ao%<+bvyK^F`L*1I-CA1oxV!K2aA8YQ@?0`zNf zE00|(*3#6YM8ycVjT*osjAhS5tN=XW#~CGYHdclwg_Q^r_I;GP7G>-c9} zHoWHFkZs?+`90(Dj@@_fw3Xq1xL7(P*knAWXa6;Y6pSR^=ft^%gJuS0CnX09vyF;N z<`zdH)!2OgmEf|_l&BT*lOfxH6bRp~32Tql30Hc7lZ!j3HrUK3@C%`#8i~SBhEzgx zN?&+uI3B+wx+F62gttmC0cA8$X`$C-c77IYL%D4$q&?AZ9pwuiXe0^~a8$>BxsT+s z+k`tYq^ivzRZ7VrWVnTjtMz;0W9X8+Kw8Ny4>fMQ+IU#%r8bh4Wr+L4xTnV5I3zbQ z+X*_N#_id@5R|ys&e;-N5{YR{9e?iJx$%>ruQyFr$MlKqiYI%b$&S>l6Mh;aj<75Z zw^D(f&)3W9NUJ2Dt|l1bXaZ{XL6-%w+hzf$j^9DDEP9#!98_DL7&hOjE#uki8tYH! z2u`f??kqdzEt+A;A@8&?87qA8_kG(HIT1_2k-Lv_0_P4=xzwhw-8&)mBaoeO*);cG!IvXVH&ppPm_GN22wqvJoxj=ELlr z^|8lth80>UMeN4kS#pa#AA)sa@6-x|m_f`7iI?|40%*`>D=7faNhgHd9-Xez(ztsW zR|CTaEn}H|w)a{`%jA#V)LU+a+0;BZF=ni6k{5B2xB>NlbGRrsTnl6hmA*Tr8|CUu zuN1D9z6P3Dm^Z|Xu}-L^_XQbz79*`?p<*CpA+rjcnAM(|w>Ci>?jk4wJP9vl)Y$Td z*V=VgGsa>WGfQO8UCo63^EUsUXIg;XJG<@xSvC<=_lDc~3`)L^BB@jyZVe54NMpbg ztM3!|=VwDM?n`l1*d|Uo}Ae^6hkjm;@_QkZ4& zE*q=jym<#=rK~sCy_?<}ylBdROGDU7kHbLT*`aJD9neUPOv@e0#EHO?=ARkRtXdGd zj!6>s0WFM{!H8R)@*Z%tBMZB+rRkiYfp*lQ#TND!=EGsn!Qa&eH(Ks-*x<#9S0O7q zB3vv`TfWSDA#>49-OQYh!6isb-(wtB>Tk_X*^CNn132jThi7jR?ogOM{qK*E#ZJ7( z0r|6Gdz`hDd<{jCskn8_e13z(Sb_wBj|Ry(pi4}XRX|-g_M>VXOr>r)Gd54P!ya|Z z-^lo$)w(xf(uG@peAT<9d+AlO3nUGgZLt-kYsyK9(d)!gMy#mJo8A6)C!Oe4L?J6! zdDKoHjrh!zOz395PGce%v^^T>9PU=>3Ajuz-NbC-Ci54#9%53WK+Dsw2ULm-T)Ifg zo$KGmOChCm`XR|PbQ&w)s_3e(;iU}{{7w2jaIk!3VPFyWzOQKrEZ;=*B8?d)f}Wjv z;TpvPXV0xMV?8i3!bX9!H~y?YIq6lW{m}!{T-QvHEh>p@@L8+^8-(Qqn>{W-*<4Np zmKtDR0DvV=xhDu$q(nDI$7OLUMhB1=O|5;-u`)z&cz?O~qNU&RO2t*IcqeVVcz%^& zK&lfKL}-)?xYg_WH7NgjR6pOnY|JFKE4u3PH|9z9+@|ctR3sNz?P4(RDff9zzs#>O(WCUqB4O8kRx23nG*Jf?!Y~bX8 zr1MEZFLR06Gi&Lr0hhheFfP;wB#>vEZ^(;^r(I_;^y@54Q(+xGYoxpcm>8AC=hBf35H`Yt<0O`#^m#VNd z*_TqhTPM_^I!PPU*DVO`rmsXMbI}dTS2lyX@JiJMw|R#DP+XuzhK^A$SwH*lzBr1d3HcPzdQ=FmIV=mE~_LZz&3)NBR!m4{|0D8DGcl*$33hpmyUlXUivtV z_|RB?CYE_p9hNP$RytL&YY=NYu=yC9aq^Y*bZ<}!S2y*fGLf?Kjg0u~*yEt>U$s3x z&Gs>@f7K(HALor>{JHpl1}sap*+gfY*RWcvI6wK`Wzk=+%077@ZB|u=?e^d8cPqS> zN%mYCxIAP%Cu_>K$RpEr^g7T4HJMe{1K;Jn&DN(;GLsfbS-gc`%~~z!(qC>@)Jh)KKlpsK9y

    Dc(7Q8X*Y0v7w@!wg5G0|(s2 z%1To27M5K8jtNaSX4c*%pT9OGn4oMvEQENDlAoc-DJpKE`wgC+)(5FENK+P+45X|p zJdcMid#fK>j?h;r47?L$5F(D}7e&EA!Jw#|a}@e>kB92%^-Kj+0H@1yglfEJ3;z<+ zEvfd;b?-6AEJ4E?7ZP#p-~#0cFA7ovHhIA3Xs5R`bwHwqLXcO(2gC7BaL&RTc$-Aq zplPy+F?POTAj**0!yc#|foCFZTlpP~L^D_py|D$Jfd>wJ=MN@qyJkn2#Va6Pi#S6%O`Q4Do>7~bp+#C8wJm2UOO@c2Sd%VhiQ z^TjSgoiJ}Ul!?NY4ND)?I2}6#t%7U7kC?A~MB}Jh8a-EZc<~a5@IX7Gc-iEIQ#N?! zxT$m8(x){@ikNDW%*Eu>uE2v_Ol4P2#zEB>_k(^#K~6*tQ^X)ZUXyM_r)^63ct=dJ zWt}h^FUxuTN028TZrGj@P9FfRKdilIm~x#~$Itu_ZBveWIR~wdMuFavc6ufEQ#-jw zHW&bwcFBej?Q==J`?OP($u2$dcya>z9$77#P7IAKGiW4H^5qmsq~frl|1cSlLLWEo z;Hzhi9|(p|nGJv#e)+h)?mD|-$5=SUoj6ovg;N}tRE=VXv21tx^ z?$n67`+5JIZOJL_v}NK}Si+IQ1LyVp?0NIhqs(z_7#Dp<%jecv=a`kzf5Yq0#dj>d zS*JyW6*s#8@*Y|@Xp`pM1BIzOIV!!t4Bd}+Yd}~f z)xV9-4A?7B~PNraw<+w=K~c?^{lK2^__5FW}){4t&P|bJZ!F>eLIeQ zWg8|O`#X#@Ix9>}`1ZwSoxcea3qP8&f@HG`-8gX^xzY?lTPZn6Ocqjcr@nzwTRHMh zLDlR^w=2Ak&{qEf*CP-&UPkmm3j=k2pt{6eH1SCEP4}y^Cma;l>Lbf#4@n)pXyR7c z8JY3uX+tde2z}c@;Ij>1FElnI!DGzFLht=#vbzZ}6<`0Lkt}y&D^h5NflNxio`UZ+ z4jmdJ9cQB3pbJLK+Mpx{NXv|!`Z^(&S3#x{qeSCnT3FBbm}kE-4hZTyjCsl;o72`fbu z4HAt4O-U7<{^5f;Cqsbz4g)5GE}DRq5&NfI4ex}q&EDvA1@in3y4-X>9kVsIMV7F!1)w z=4ki>R~SkM(_}TuB=E@5#9@C=PQ*pQ%7~pl$fj4#)N|EUbP{tlybd_ewbPSbTLi#< z4$<%)-ZJ<_UjSd*VD`d=X$&W=WA0at+oTbWN-dlGOr5{rjcF^M6}mx_!!S((fJ*@i`3QHz{p%WkW`fMRl#p{|?Q64b-fjl40!j|$hB;JR0yO}ISLZ`> zLgT3|p}S%>D*FU=v)21`0Ci`%cL~%mf4VceF)&xSTU72{Gkdph2`4L9%gLp)6c0(d zVy}b6M#lgv>zJ^^nZGRVo%FZ0+!&=Q=H@9=W2(Zc1Xt-^;1_a;DS65C*aV4}jNypW zt&5q7mRn?}%>h{1^cqWSbi#fQr2B4CAjiiGHw}*XA`N3Y>nyWcPu}SH>9wy}W(r@E zEwYuo;t^3V>7qNp)qqkrAk|aq-2N|&giTFv4g>EC2JY)w9_DTGFq(d~b1q4LZ9I%! zW*$Z%C5Kw1O;p@{_xWD;>1I{;i~*M=J_9ZnK>Q{}IpDH2YQUv5COhb!dqwm{<%b_O z(oeYeXEp*S%z%qtIN)+Z)J6J4b;7i%`TQ2yN(ZhrHrX2M&g%`CQORUl{&g|hK8``qh!Neou5=?(gQg(5Aeg^sH=VS|jZ%{9Bb|J`?S zw`Bq2E7c5ZrA%=b=dPgG3#U9ymL)y`IsZ$64L*jV>5GDr$c2&ytD@oTzCc^AW&?n6 zV=jVU-ntv|-FvPdz3N1#eNe_x4-FZmveoZibH|dCZ2BR|llKY|V^>C>g{TKM(cbaO z750bL$J`fJx*hiGiOLm@gKc01h^K3Yt#BLtb=CnfLjN1JP4?-hp(Zcn`uA>?k;2!; z3#l{nLMkZvPKuOMafv@iQn@zXm(otzj^Opvwah{2{lP9Hq=IS*D+_H_WrO}uMNBI> z#2ixIR32h(hgW%YsIc-0kGDaMKelR|lP`bsOq52EO=AZLKv8*P#x>V_+;%(fhewCB zr;sy$F?wJPI@MVR_cA8)9sG5 zh8S_rgd7ewY}C;8(o*j>UQI~H+vSnDoI_EW5SQNAJwA+;%p&O-50^& zm{CSDBK#HG=qy(qaBE`urbf|43Zdt}p0|pldlR(Lu%D^lT`$%s@L)BwDX=!U2Q(>= zt1~AeO@{B(s7wGQozhpZTj}k09`avxNh1_r#99EDi6$u1y(?%D?+q^HsLKTx7|ac- zQ5sY$+{)4M)psM4gnF?)^0QSOwy_$Q7d-M~vpX*sc#Nl5`AKg)KB_FX z>;|&ZMuW}DN2D!cZ0wwiRUx+ZXO8io^W5#WeDBZQG!6&YeeX`&IoN#mJAeJNY1tBr ze;!S`oH(bPWVS@bzaf7@k$x)fynMdba&jxYS#^YTNY@F=gV&3@$>B&;`9rca6C5hw;vxQBwvV#u$m4z;dJ)&4I?h7xTV_Wa^WrfxcqNCdhgQv zJ#!X)FE20&$f1o=gBL5;DfxFS2KoBm|p@i_&YgxJ%nYwmhpvud3X2dbdY1EJF^xK53qCgw+pZx5f-ndk~3jluKN_*{d7Rlh<_ZoZkd@v zTa4*&pLo0?K-P`v#-DxXZalnzGLUI)7^X47wc z!vX{&-0iqS)K*#%Cp)ZT_=&Q1A^aH^IX_G4D*Ox_~U6oZqbmg?F zS(U)g5H8@Z58KNu;I?=JWu|D8vWmA}oFHyi?N;DfT|l3xi_VL<_13jnUEb-_@M`4x z?Swiky~n-qM*42-0w$STPWOrG!|*JK<-p%<^iH?y?$_LN=nt!BqNH21Dm7q<{66Qg zY-licp3#_bW#h60m$8A>3+g9E*gR2Dzu0rr&xETd@?%ShhMgPa#NP2fGZdCk@+}l- zi-C8H?XK12fDg)86b7mngyskvB&h-Iik&_fkWr(ofw3lBjkUt%QEj}`fb!U#;P*fj zr_d)?nC@Q{wkI;1uJkp!JM55S({$PgUN(IctBVd+D8D8!0cL6Dm472EUmL&xK89_l zHc;{mimU^EaYzPY5-Lxr<0rZqcr9~)-8(xJ+**XvCXO15C-#$?5MzhF1JLmN6paQS z&yQmNBi!FC@S1GGN8leSKP2lX5~x!d_Qh_drPsKa4-=SJA^aHt9scUpcOz6^FiWuB|T?oZPN8ltthOU5@v;!19oq&j3b6aw!ciz9FsjX(E?XcRZzYG(#ym2Qia zt=wbWgWPUWT7+Jl%*_?queu{V3A7=x zyk(hcKPY&e^tmKX;XW4iiTeekNZqr-$dLJC|C2SQSXsddcYpNrtiP@P*VXS|2IZbz z937`!w&ho?b29nom<~y^LH&HiSN;gZcknh%mJFf3B<38{P&KRe&iNxKxVaAsWtx~N+M)NyJ9j=^){yxCQ}Dwv>MoLho?*yrZKgc-+93afXm zm-I#Lo~lvod3))UdO=c18!uDQC2J%3bc1IP=a8^_8jE=x%d9x^nP>H|JFm;La#F;5 zADO_F^RR9rxigWhFso%5pyZDz(g)EJ1F=q<+Y0V!pTn5w!wN+VoIMGD5Z*`BC{Kj7 zjYf$@JXjvp4uy>EvPXtX(qwnvNfIJWl@&x62N2v~)S?P^L{9bm@_;l%u+N`m_$E zaR|J$B`_ffX}Hz6R2_s5+`=z|>P#exM_T#Iq7rGNCrVV7M?sk%uZ}jnrYpm0{a%ks z7UzwM{NOK(CR(a9Ic=)Kifk0}m!0%kAy1p6#(pY8F$E^B^kU@18*d^r78aAmDm3;> zSy{|C+|K>yNB?0mOo>aAVo2vivfRutJ)z|N6nTh-se|sMD_|?rOHg}w6IH2D&`oZ34F$kEk!p_6CDH4433%Yqk<4 zE7S+yA9NXFen!sRi*7soP$B>w<0brCG7wPi_dt_nB(16ntB4%>trICE-36}*$gC2n z?bABxg_3PwDQClzM**9 zJ5X*9`_eU(Jd+~pskq)5-5_*z2hxN_*JqWiQ}%^GCrp9OkansDjsg>NHr*vhNi;3J zj4=ws&U5fLKdSHC_((eD%QlT_{W~8a3GKWF``iEZod(O@5}OQ#6B`#Rd#e?k5=eBW z$*}CtNbEZoCHxkVC0y(d9!2w9P4Tz7GjX=%CsyNjQZrB0ll(3)Aa zbV?5U{WVnF>6ntp!+wP#_4(PSV_L-JVqj`M9iycmyDyniPM`>)5qear{rjycQ?vfk zUU~zI*&YWI_UzYXXUql^-+tq(CTtW>WA2d}C&tDlGi=mT^5Yb!8ON1O+2;1)=II&l zK7G-ws-vIGEOWyw^`NC z*&*ttSCjOR`#w9P&&$_|@yrvS4EJW$1?JkUR+6s#Y~H5@9ORrhC9m>mKR&Wdc_J-)TueY_{-l`3&VY0qvv%dK2!&B z9^{!<*Si|u>V!J}T=k(DtVaX4?q^-Gp4&}_UD=k>dY70&2eJ`nul zmq*?7N#^cVlt3~J{O&H-ZV5&fHHr=hu3)2!Q855}>Y%g&lO|1IopNA@Y`hi|DYDUAbUUuNC}`v5dT#cPpLq|2t%0Ei?7{!PioA$#Z=|hyOs?yjmi6LR z3ayVql|uq`Rn`Wz{ZN}P-4k#&IKwSh7$17j=Lt87>GDSUyG5S&-FwMGNsm|AtZkDX zfTm%tu$^8H1xwAUzVIi3L&usNCEh#Qr|_B8ws?H(^p>5A=(MyVvbIFuADFg7iGSMn zfE;I6FYd%n$}O`M=OQIPPm#0Ooeim;p-f}5!dM((Sa~23TFTscZxu(43`fWXTRNwQ zrxSL-D{y3*pz9z_hAjv!(13T?bEgb;p?MHRV^uEhcD-?6@rcU@KE$VN)G9xouF$WuXHP)Wc<4cdZw^% zs+-#wvt4@7rxOT7fyaz{Tej|9EFIOrrhF0qBwy=R%ajN23~usF5f(*l(&iK8_O z%?dJlDR~b?x~RBM<~2x~p=zPuBPSvys)|k(VjV_bcxyQHFICYyV~zl~ULqg#*)~P2 zgpSN3bY|>v#rmLlkX8D09u`y}4POHfWxR|RBY)qZ%m2W*{T?f$v*`l5o!K9P$J+c)3*3bc| zmRZ8>gz{+Pk~QE`=$+))POhP;WCM{QFJAkXjXx3JcB4{ZjkQ3M*{b?_*y&5-V&>_};8(Z_PBBk6)jf zRZL3R`NK|}vOZ~MY-%YvM2YuOanSQXKAZO!6ctss%GUE2K;a9P2t#4U^c9|1VOlR( z6PC=~`kQpQx^A<8UW|Yz!lqZzPP>Gw9L}oiAaylfX zOg>#arB$}v19F7W`J2qeAQ2t|wq~l-)lrZ8n7`;3d|6{T#p=;^-koZ^ppE-m6JGw< zve}1Rbz+xgzS$2B5_H;E(!h*Ij+!p>0Rid-{5H|K{F=4)H-2(SQ1kowo3*&Lt}m% zY_vgaFJtvMOjtj!d9G#p(`j2-tRyt`v_7aTr~|0&0Wa7eg&o29^kx7Wf-!z(S_CGp z3>1E;0Z3b8EF%XV61{Z0AD%-_KMel$bFjXA&;>h;?Qciz1eviI%MtHo1(`RF`2F3t zEM>>6=yf$QTIlVBsDvn;NeU~IZlBhyDv7$|oy^%HzRgPt%W&9YHR8Y1Ce=UF^EUDT z4m!$e0-SdV`_#3q!;;p<3Li|d+i_mI3@Oek!ydTTOOH*f^X}mtCYo7mXKjF}YF0qG zUv|)XaemN3d9&31kK)J^$@&2uc^JDNeNFK3$B;F1Eb;L?HMtXSo~(em1B6(r=-cik z(4x0jxj*;-{{Rqystfr*kp+?}LnqZ*KP`PPs#7vPaApl=mK{dgATzfrjPO*#H>gK&mru&Zv|F&4kQ6@%DW8|gB zz)cR!v0H*mB9r<1fhDift(0kzx4Kk>Ee;+aj7T;4P)+@_haZIT@^jV~V4x`#1RHQw*$Uwq}o z-u4Gqhiu~mcLREzIT0B4xaDbNqiv?W~nHZ9R?58^$seqh{;X!Pm{CpquLE;$(oVK6P(x!Ej4pMHc|2%ifn+cSIJ$vYwB5PRq#e& zfbWT_4?D<73hU;skRRmS=U|Yx(6fv=>)l5VLcJ{z&IjD_J{);kQtj;!U(;-sM}ZQ= zjM*5Dcw-$UD>u9JTB5{hQ!rK#f!RAW8ZBbzg}yljb1d2PAu&osI2;nz5iydN`gebSIN6eV)oD?(-OO$jP;wwj z$N|z7dAWG8>UgN0w`5k`TjuYYR8@k9x2d1(_#VcNnmMo zFjPhh6h~VRRzRUX=s0x7Qj^PRfnr5vYko`(1Tfl}lRnL=guPV!3s@e16Q)Xy}vjg+dzleApvE_jUG$p9?XdE z#;`uI!_w6$vkwHB9GJTIS67fcCytivH#0ihDLJfFB~;w@Z(SCBdR2BQx+rQ(^nOxn zXzM@^rdgHCxh(^Q1fX*+1HrC(AO!F7PViVIImpQf`UL-qr*?}nf@bft+zX(aGvI=Mj|C$mOr!l^Gs56wIK&P%zk4UW;-@CqXm0VJkkbY+JA@W5T^@DpP}rHUi!ghQHF96IGV@3lDWD5 z>zJ0QtI2@N5mK%iaMANtaI=F@$~iu?%)3~0PgLyL4kaA*k|gG+?~;&R%0?Q`rZT#K z?E>St*-)KWCB)qFP328ESHrt3I!PzJ)eknIC%oSskCsQBAX{UX#;kN*6;?p^dv?Kf z{qBiS&6qvygnK%3lz9?(^6?<%qd6wkK;~qM-85S>@Mth%T1qo-`$G(=I_b9HMW8uhAk!^WW_Tr#e$L}}bi$Qh+k({@UhPmyosW55 z0tIS$p!lf?NcTAkHze?vd#LqNe2X6riwtR(mCa$hsMw4r>juNNGb{ne3Qx}4qkO^h z{clLZFAE($~% z_5Hy7AU!-TnQWIk*0abCB+r&Y)*r<3hRX2Am)|uxD!;hVm`V0bB+X_|-cd?kN0Ea- zoeMqOMv8tU=+CAf0k={U2!Zs-uxY&`yg||;UKP;ojkIot|H=iWKwFR!-7bUP2}Qqq zeC!&}%K_+^ER|)-&ri!nB&`6fQHa&|V zlsznVkV(ncQvjy8HeNk%7e_lCnP1Ah_0VNhCu|i#fs@NYXqwh2@_pK8YW;Nl{UNMy z$_6{nUCR>W&)x5+1A8B-554N$WjWl67*tKcbO3Clu!Ke3M=D|tl5F7n$1~V}s>ZBj z9ytO+ zX1UURGQ2K>43wVNA=T14zTx+Y4^07bqAI#Cs!mw$r{3ySpsXf$Bv7swfOhn{1pX;o z_#5RoqL$tzc}=>@v=yJ!Tkf1*6A#(SJ`3Sc&#Mw>6uF$G(RzB&C6}z`Vrq2UnLNOh z*ldj)Wp=ELyX1rNd6rd}&sWYkZ$yAtsm#9VOu7C z#2X!a(7La+X=dW1X(soj_ZO1IWaDe&zSNkxFJ+W`D@BT^INTcbyJLkNl=v5kclz{u zT#2ZGbT^bx>*CtIJ~xoNRT@?TN6Ox!pItO`vJ|)y}yj-HG#)yUbuvNXbFpdJ`45g+@<>EhCu`MvZZ`bLzn4&SwzQtJfd~4m9j8H&B9SqUVuxvhb+952ftno4YW(U4zhY#;Nf#)sD z@YwjOPVB8%DcaN1HFUrGJ@>)j_*mdV*uu|Xj=i^JW-_<_J#~xNpz$~qa!TeR;}5#A zg}_0BZmomC#TrvjmmdcL3WmPvJ$rYYX~lUM_7^E>W9OzgF-qdiI9<9a`CW?KrsDP* zQlyIs#&wG$D};Jq>`iG97z{JU$5zB7 zg{=wN7@$#_IH3ALh9sG*ZsRQq84SdIJ9^9g?w3X7AYZIbF?Lko)Pp*|kb^v+E^@OE`QTDo(FQg6FhF1k$LmbH5O>?#))gHCa{I3LoPx9y}?G zPiJ@`iA*hnAAk|)F8Zmw*yus`6n*x6VZ!plvU5LWS;fiWcD!E;v``u1&-qY}&6I^Ea5Dq0x$hPsap3itb@W>QdZ=k><6Q{p zCir?l+QZ59&x0UEJ#Vf5HPvHDVn~zf-q){v?=;leKs=3r>iXACtAJ}1&Mb{?QZ4`a z-oHPsvX3Efz*lbO1|ZUs z0BMFw;Us_k@e@?QC~O37jDS4g2{b7_dLb3OLjm#mz-F@?hSD zQtfyUJ6RBlk|)io9ir3zr9$i_s|;IBI;4>5;e08%F{8<|-SaGv^cONY$|PYfXQk^Y zdM&3AN;ty@$hMv3b6G4awt7nEkprkfP%G#O@- zP!Ry-NRZ#T393H^U2w0yAT(Xht{{U2{G1i@AFt}UU|DA6w0^CX!YO0lr22B;HO_p# zPPlJ+-gMZ|A{!;v?^L>Biu+VF)@iCS16~Go!^Yk2`Pq&#aYM)$^|TKS;GmzZOx+v) z-&h&rX0j-ssL%YFB(k$8PF#xzDTHCp@Ony~Mv>K2943Wyd~kwWiFSILtc}D^O=1i! z{0jAPMGXieRnHm^*Z1l4vjfu8+Z_Hv`pe%luX+M$z0YdWpg1AQ5%r0Br2A*pK%izF zQzu*x{j&`UJhEMMT7@3&nJKqr&pg__!Bv7z!GKG>6n{F>l(3uqVSX#VbRY&;nGq^* z$|t`Yi%6CehkmVyWT7-#i?~yEPHd#519O6I2$GOtPrrI;?0^*Os5LEXv`gL_MOh^S_x%cg3`|@gjTKqcMTa0(+*n zc7foUsQ$TAO{>s%xBTY!B+H36(0k0*l_E+GF}VWhrIp?B-s^gZd&j#VxRI{pVg+WO zsEu@!(~{#*7rb`*3g{02%}Ih}Aa}gmcxOqiPXUl>Utm(WyI_Ncnt7GndQhlzn86t< z0PHf`BMidM+NipIk!ZOxu}McdF{_M~LWnN}ZFEvt0=JIt6JLjjZcYT0xkF2SQB(pq zmlMy|^G^EIK-czaa)`;NpNQ~y9i0md|EcqC%wmntPv1luR|My|83!)Os3^O z8h$;U+-B$CI`PW1%q%JSn38`W$>xP&F}yQ8sdNxdL6N7yLX#5n{?p!xI{P_U`vKO{HZ2VG82%NJ$S zXBoY$j=Pa*eV`nyjc%7!33_Gt zZ_y}98B{TCVw|=nc-6f*8B!DcvZwyPKan(c@N;5Q1##nH`>T9P4(;XHRNSRmiTrD- zzn1%L6gM&}xFrGGexCEygdg0Wb8ya0@6Mo)`@S(Rfx_aPF7aab?6AcTP0Cozl)X&SNaJjf%GST?68gY2%k<>-X>AFvoMhEvxU} zd7VRj?5=Nnnf#8Hx6gh=QePXtqs$E7d6Ybtf}WGO&C`pfHAokER{AcQm>yg?dH&?Z z6Pr~Fm|K%S?2Ed>y9rzGtKoHY4$~@IEWI}C7+K?G46sx)mA(%3IRIm9&~6rIc@CfU zJ#Z-fYL_K*;`7Z$P8{U7!lsDEu0`B6Q;j*n`!k`_5ba%yc-PcrjC!B4MNsU2;%)UV z&e_PNvK-fPzZPKZMV1EjW{*a?#gGHYaUgB{sNrDGpxX#YvTpne6jRD3n(T>6|B;%^ zXJ=2G*HXfsW%xcOg_0*xWI1f3VN==~rcrL?TogdrZo4ccsvcTa9hOEOVLD8QY($Io z0UUIudPZ|PZNfzF4+D;o1a_EkVzUBV&cpV$t10;^imbrgVJ5#k3ahf!r(JV}O$w+x z825akaW-uPg(Fi$Bf_e0<>3!36%d_vqhdvxF+F4##3JgYYl776euw=uimkCNazg{* z#_055BwxKauC1%K&b4867;#n{bI|&7F+!8s8;fOsN%A+Di-jLeSwXU$ILT9KW&pNQ za?k@Pq~emf+n{lvS+!jXsU05Bz3^JY=R@+Q7NYPqAj_K`vVz;8Y;bL#*}>EE3Kh*N z+~96hmcm_6;JPIt?Xqfa0p}`kLFsw(T^*%U#R@TI4(HQ14uK3S#84ai0&o0nQ6wbG zO9Z%}sIy#ePTAz4osQg`ATiP|Ly3`b-lDVFwKxPV)&^asp5tPP3oD!x==MMM?4*y* zDHk-$)lCdabX2;XorBEj&5kPy9Qjvg@mP*LfP;>*dMusSm9LuJS8Ew0eSSCV#3>ak zfm35!9#Xqk`tO5GaHD5qaDj54ppP^~Iy!Rk9O7)n@N+jh_<(f`FaGH}-=A!9JOckv z`5{@)ZlB`3fG0c6+?36f96Z{MRNQ?jcB0mbPse}@s{=L6vxM}Eh$A%jh2i4$*W z_nJYan3C%#pfZhX6r>v|C|ktkM5Cwx0ZAeO?8e8&E-;^!3xK`B9r9npN3!p)!wt7~RH038Q@a2z@8Kp0_9Zq|afp6*$5V z37b{tNT2&b(k;3q-Yu$S@V;ZDcSbf{7nB)WM=zO($CHHjr`0mhqZHZimpD@$fEvcZ&U&wcnUNiOQWR| zK)=Qo(~(Qc;p)$ ze3X@ym>~A)Q@(9ldMc)Wyo79V;-%+vv!!P*C5Lvu9aLPs;4EG3T_SDt+#iA!K~0{I zNSABQTZPbIeARmecR-31CJ>6{bsAb)55rmGJ>BG_&pqJMUgLU}yN8n_8h#z7dA|_k zy7z!w1v@TKec!90Ok_fj^sirSBCFW#iJW*7U2KN6Y)TH+ zD4mK+lRc28NYVo@kqq%IaSpeZxfPD}_@KooItCR&zfz9e0n=tgEi+Ifa9EAx*s$;+0M{=FmtF1P}MJXjOp~x00 zZl9<#C|7uX_73vwm61~CEZGrw|E;SK6iJhvC0c{9W~G~+S3|Fz(nXM?bZ>M%w83;J zx6>`qMYwooBmJd#FTB$8#y;oIfA4ld!Sln8`G|EZGg7?dpO#&i`Hv>d{O~yS4RVT| z<#FOA>z*0Hu2OP{;ax-~!wyadsN6ugBJgB;bU?pX541=n@*!v|gmx+j%LN{xGh;O# z{T@3aF9){qkZ}@Yx%!|%7rpPH5Yz-w-wD@%`c0bb5e)+>tNrJ$idhwNblTBrYF)4{ z_>R{dFO8xLw55;9Gh^2WW0ViIaCE|Y2?|YMf7aZ&I$@XV=iCI3J3%-&=2X?$^ngoM z7!*4A_j{=Cncke0CU~gxOm9wcyRFLI+Xa(W4U`Qhtf05{A^mT16Uy8l^`0l|oY*q$ zFvE?Ol0#EKE*00x-0|*`)<U3 zu|2Y)NseR)cfozWGxfgDcBq%Dkzeso5JQqEnQP>u#jZ^(o2;N0h8S{3hoU;9sN2;p zJ4u?ryVe57txk9sML_ZG1;Jg?D9EP2P!xOig_H=?Si_;^cL&{(pjL>HtrqV}lQl>> z1=Rv9n*kETQFFbI5qU2+E29Kqs_)Eyn`5#%-~IpCdl$H-(ldYDBb<=D7;+<+69Gjc z2r?srp#l!-wB2cEr~PmHYqz`m?{?b$q?g&V1$)vn)c|o!fz=Ja5@$F*UhBNj^Wg3 z1gE2v9Q^GH2;>E6`R&WEiTAiBa4wL8Z`UqE{hkuwU#5HScCB6Bp?jY#*zH=eycJ$| zNBn8!-#!!J$)A1W*{R2tx@%a0h#Hx;YLmzSqD=nQ4wCTN_}jqp9TOr_Q}R@bY^Ne0 z$o6^dq|)56nn-nPR_y#i-}w2eHgJxHh!#HU07WtAp;iOMA{m^0UIU=?GjWhuK8q8A zljXxt_{=|DP(Ly)7J8nCXUAL_GbKSt#@*r#{T~i5iQ}pE@(TRgB;B$^r>T!BTRQU- zp27)UXx@;Nem8Qu#9%apcZI(piQ`3&?HCgf+ZkgTGAMZ(MN+6pJ#_)A4Z#F-x!+q1 zMTdz_{dCjJI`OCjCLlVYpm1x@em~ViPO%|m^4;@h!;UkLW+-+U?)N+@o%?w^{80?Mn9lM3T)Bo5%^O^3{u68n9Sv%>P0O(7 zL@!A2Jh{$Rz?mpsp5n)11h6~NJYE#^xkS_AWxJ)sm?>250YuuOWY4YXp9hO&4@SI$5v=C%+bU_!ETb- zo^2i!;{}L0!hhzQs${SU$k_3!WJc`c3*lf$1Kqg{q?Yb4QymB!mO|AF?3Itfez<{N zMSZc%#=tS2P&P0!zS6eu?!+~enHH;>sR=y_im)5KJ0W6@kp-Qqk*-jyl=JbDP7qZ& z96U8X|C1ZS=9W$>3mFwDjpu+wGGFAe43e5v6Bu^of%_BM1&$IfgRZw)A6a^n~0glsBnF8NSg!&{EmhlSdd9{Dafesee)okA3qa=wNZirzAX(T zI}knfS)aUMY2++F{a>7C2buC2*ti~coC`9;en1v~F8v6aH>>)X?eZ^#8<;+R{7hA& zaC0E$YoKExgiIhwoaTM9ed^fxR63Tpl4GL+Oh~h;W1_VNJyecni<7bkX zd4e7MX*nP)ICF6C4<4D8H?zTv{jRFe(|K$VC7F_MqR2)n@^hE-VTD2IqF!LyX_&{% z`7mHMT;XIX|69aWEXRcflWE0M7XSXAgI@K9F2{M5ysgky44rAUBE}a>&l;WXLjD*X z>zD@%ZNj}b8#@fngV~s@@co(lRqw4%=;nZ?RBI7<7mEVt7f|m zikNA>-2C2k_tn1TAwVHoQZCfB`$1E%*By&cRoF+jNs4=}VV8^`SSr%#jZ89g)Gui! zat~~bl!>9l>TASvNMVJNX@UK-66YJXHx44@L6U99dEjG4i^hIR4hoPZK!RI5?0P+@ zhRzeW0m)WX%4w5#Leb#iV4!RxC!o@~iC^x8I7z411gP-MI*(jszCf+qJfq!N-;AS4 zdel!(Q9-5h4hCCzqibSyJ=BTgW~{Z*zkY7=CvkusZv0FCso!)7V&j_Fu{&#~16is2 zSus+x)rwBnCTKP$UTHdUfCnn z02R81?qI5CsG)WcsnMej$DOnikcY${C9<<`T6^iN7_Ae!4>YV@07_5ldYW z7P+Yv2@vi;JCaS8hhzj{b;l7UezFEi5>;DW8+E1Z`OdLaoJV`$-4I)9G_i!hcxL5A zGdbbUSlNo*tH@vivSDn=Zl1*M^90;+_WO4R!~ zO|-!eLXShw!W=rrvDj~?--s;DZLMRhV4qiyIKz3wVJ|N$v|NO`r|Gaq?;wqVCERU` z4=N{XqO5?r@sK#_;4HnUkGI@_qs1S7)=4Vu*zmL&A>#}shxU~XRAdA2rPYcN*(Xj- z(kx{*NUx;RNg*S$-bGr;m3ejIBg@XOIJu&Q(+#_2ZE&jd73htN^%@^-uUD0KPa z9-XJES=K}HK$PR?1lb2RL&!7SOvpB|vP;wd@%@_#rot_3_Rw}5-Zvxjg5BZ}O@>}^ zP0}My@$w4)v``4$wJ7pMxn8;S-H=|Xs!>)h9Aesq^_(?6Ds-vU3JlxqkPkt)_)Cv| z_bOLl-`0uOOF!s`Emu@PzPOXdkBvv+MT4iyNr>QoF-R`f==x|`qmRpNk{aXE0bp*R_^!xTh2T5I9mZ-%S zW@`DDpdR_8lNO%#DN}B?5FXDKkV(56mVt;JJW^6ePI()QM&am>4v-8-Qf2gP?xp0Q zo0LyQ#_*8b0X3SnCmn$<2FM)x=DW>ggB`Ck(1Cl*I+IGtw^IO* z$dZ}abd3L5Wxl9`n*y2zz$4i%xwWVp91+MPB51qmq~%|L7W~xp^RW=xz7x=~OQ8~tb zTH&QRU|+vZIW98bEYmw{D=D^PEA*)ma;hjf)B}Bj6=Km)xDhQVC#oj5@{lICT)K{j z8V1ONL*;@hVHfOKP^3L(ewi=+U&Y10M;$aw1%H@}t=orPu#_xc^ax7dF>~FhOGHBB zcUl$amJ@{G!}5^FuXg8epT`vg8uy6jKtu zKCIq9(Jhm^)e|?_>l{D!X?Koww8i$zDmY>T5DU*WkE7$lSwp7$OM7+G%rPpF3#C)i zC>NrrM!HCq3=ItPe9}dmm#08HPi2d#HhbsUV8HAkHu%6i2F9V*O?$6&Z~J6)+3h%R zXGWJ@SK)*#?LzZ8+9?H+#$x4zmUZId3 z`@1DZUP~q=2aZcB6?p{wT776A04PE;f^jx!) z9VCyV+TdERyulphZE}4i)~oC48Z&1pkNT-41&XO?J!F|2^K;d+S-7OHw6|MmGLgL@v);q z^)E~bw`Q2CsDRZPk>Z<0=a5eRar)s~2j9?2dg&a9Ibekqe8#EwALL|k;ucpd?x!`J zUSTsG&6|STZPDF*?vbWF0J{K+O$}ijp49k@5D?QGj`s_w1l2s z5bOK_Do{ch07&7ghC%ceW4l-<0-sC)ymdfjjd9z*ssYWmyA4$ z)0F%KMe3=@yCDx39i~fV=SY$?#%1Rw$rv6jTCj6w9q-nxU6R&_oipp9;oM`8 zt==t0B3o<30p~p&yyBc77btE?(hWfcp}EebvOEr+*FYzdi=Y9ta2MYx&wF?ZRGHNZM8}u}xRe7yRojzxyv3#0DR+y8HIKbT|I% zo|UGGTxO)Kl7YHfIkzBmI6TcghGc^p-a+oF@HD?G{w=|6fpwFn9kZOI*8&*!yRQE% zt;^Fi_hLpg0t+K<@iqq{3mvIP==ABzt!mvgQzXDU20O0>F6`G}H2j-W)~oJeT!d&> z9%r579_sY6n=3XsW_-Kn|2wd3lVhx7srxRd9qpcCTV^?ArFfNf?){LmN#{4KP@$X#KA|t9&Hh;0g>m_h{ckAjuIcI25)I*e&pNHOvEUte`FM5ww!YU9gA7 zG4Sh-I^e6zOV>eB0qS`q2i#zunZOBS)N-@N=FF!Ej$e|$`+WoI-n-^BK(5&F@-}R= zyxpecHz{%h>V3dV*XiST>WY}LvHO=N&l7p}u@lf!M1;C;mgw-(4KX3NcB|y=II>la~=(aJeEDq(-^crH80iMSMEP42{+v0Q3|!V_vW`a)*~jsZDdxcDoXq z$tt&aRC=^YGDNBy-19oX^}Fu{xu<8(`k1?8+n1JM!nWUGywTZV;>gV-A*LMKYNIFp zY3tL!eE2sMK|(t{z}+v?hU$tgau-Ae)bYAxYHy@2r@1FHkKDA9B2dikLDD#4&bLW4Gl)e>!>p`iQ}i zZ~w02yX2vR-F{aBeZa@;$~I8)Sc=3@ksDm~f0AcFCB}b)^*|K89Hx`N+9WPz2UB^N z|Mw=toV8Plk^1iPt9TWgSxcfT#0~8gavO#^cQFdM27oa+&+~Kr=If!V9=ADL$ z96Sw0(wjjGs(Q(B(jM5z?UtSMP+>hhimBr!C~=l9vw@UatFAD!axSKDtojF{n9uYRMO3~<*#alV${4S|-+OCO0VeEKAaS^Ls8zFV@>xlcDuJVv%y*$vC*@^sna#a z@yV;Tjx7=ABVbDv-s`FMZVBuW56#ptwV^rOn1bzZvrRt^=c|!zZVh~5cBm%H-)JjYlU}6C)()BW5 zswd4Yt+zq93U0=NXmW6QdNnKDO#hFzmH*{waAdSg7hNN%cI?O;GIC@JC^@jMv#H2V zehv1X1W`cXM}`);UGzKanaRzY*9Lo;ZYA`j1fq4NbC0{$qnyUq*TvgJ&Ga7nAV(w1 zt)olt9O(49Jd=VusMK6AnZKG)ckG1gaJ2{5I2fApNf@hZc3wf}0C%3I&s?V}%Tp zYi;@VGv05C3o2vN0ldOo1{ZA&8r0v>W9iP?myx5~LxudfW|B)zH|(E-f@CNW+}u_rFM8 z&B}mc4RS00esjaESLFA;>hZNs($5?9$>43Cae+VLu!)q0_Ab1?u#4_p2=#p-x5b;t z+QnM>zPNYcZGHpvPQEhxL1+#A7$R#8AXi=L{+JgpzaYWm9n7$cS}_oU>TUmH>Kq6~ zjZ+^+rADtiZ(4_HrnECbgrz@w1I~IDGsT|zKJ3G!W7dGIlBepC=5dhAdIf|S8~H;n z*rXlFCIiq>;}~evV^Rj5;MAawz^md7%r0dD2VH|9N%h%+i`#!FTd?l4APH9y4Mhj)y^3rZOmoYLLR(nV+NQcZZ0GekzHjA zLo)_A6Nm4ZL2U4n6?~^9y|t%rGKvWH>z}~}d2F=pso?(kFufZNjS5jky(k#nnk-!_ zOdBKS`N_5vL(gAn+aniUXUChy*x2~Cc8m%$5xFPHAiWS5@~*Mj#%^vhb8A-Ye4BUD zHJwgNfnimmf$5hR@BNmH)u!gcZiBC848Tzk4u%>EtgF}6)MMkqUZ{-dBsO0&p1Y$K z&+~Iv+V%jeMPxev(z*XM6*jP2RG3j*=m@{TE0XW^Kn_WxaC4xlJ|rG`C~xsbcVem+ zhSvfX(>fM@qwA~Hz_P}_S-39fj&#SoZe_MIRd8!orCVw^XoRMQmxo}qWaRGx2!H7a99-?5$F>1W8*EJ*)b63KaOay4~EZs!3?4lQ)|QjzFo>Km#bbcvn6 z-E-987`KW>N8^6Tl{eD-RL#QO+A?oz2%VPgCVO8S;G8uA zP7Nimq{v|^5@I-km}?kzL6JXQ_tWx_s=#C1V_YoH>vWEyAhb5G)~j9E$;ZTAt6T*w zG-$LzShIJ)13{Ll?00W+E8%7U2QEd>?hJ^+yfYRjY3a3rNX#$w#St(`I4oRg1vFU| zLSq0nSqQo2IREqm1BA-G&dNylYXhMequS*Ml>8n=?oyE_or;3-e@OLZ1feKux9lXy zcn!M%P>y@8fqpB)bk@>Z<(Ys>-`mJVDjbxOCi04c_3;;E zt@1wLal}up#9Fm{QIlJ|9KNeVZdQ!wmX)WCh&@9 z!f-ME*>sJ|VP#QpA~{Q{=MIsNP^Mvh=oz>`g?|m{jz>|bBX?h>yFN=<6=-yz)cvfu zD)3A|EUBWqm|xfmd1F90p_l(!wBrY+RUT}%)%F`~BxcG_p-w(HUfwHRFFm8EfuM1l z#JJLYvd^;R*IR@M%ZIbtnZ@@cn^qXC(Ial=k4cpslOwMfEm3DF`Du!rz*IP<7O*eV ze!rZc9**7tR%I*u7N>Bog3xljTvZ^r%1IDil8icF6?zY39E#@Na_!-4^giGWwCu&I z9?mw=Izct*;b6&n52PTP;8ovZecPx}hliX3!FBglg7T0~eh;T&@fIjxPvEF(AobM4 z>7pOFf;H*kXqXY%M%QBR3=U+~S?1L(g2;p({O2!XgCBOA@qf75av;C}sGjfMDkFJz z45$VpKpmpwK(;8SBF}_s6iw3W!D{8@r8N+m!G_I-;vpnV)Inlv)B#0a`a`#Ku|4Jn zX-QZ!y$NMrl9^6OW_>D5k?eCrYLr^h5_msEg?+^!H!wRvgxz!1ism_;Aemt`m)Y9b zVS|qsoy!V4(~jMAc2XL!b1L$$n@QO#ri9yQgr}pF97;zjK)l5-+PlbY51kmcn%O$H z5wh7kh(=aCbC^EOY2?@Ndxd?nLhiQt_5K1NoawMJxc3c<6+v5)wj72k)pp#5Yry0=f~)Y`z(o1&+~Pxys- zU};CcTWnh3{7SVMGgHPKE#EFVPV|4HAlU=5buJ#UnhmjhE-V9u<>OiFO!J^PwP?>T zOxaR)8zMHte>wuYl;dPXHsX*g==Xh6IyB;NMVJ*_Jwxq20?IsT?pez2DU26ehJ>g7 zhqWg9rQ?`K!Z^&A>7DOXim+X=h&s^n%J)u(tYu|zWI9DxC;kQJTyj%12 zHOo=I5Sx^=3p3~T(j`B#vYBBCoTmtS{>4kN!0H56z@Z!>*Z=;j6%M>3-Qtaq!9WKR zWAa59IPT;3(jP}06{Uc(VJ|(*fT~x$e<6L*FURS${7~>^`2d`V2?!PT+{KO{Re>#v z3!Zyt!I^lxns?5l$*r2;75BvhA&0`Mi7F{10p!S%84T=98h+LToG}Dhp=+FzV+Q~6 zYo-PBc58dgl)&qwbx;5@Y)FlKc|{kF*|NXE$8WZ)w(UW{*)F)52q}w3^aY>X5_g)x$tw(15o9#$rrR% zFL-Kas}-G!H9o2wl0zUYgNZHFFuU)vO3=a7!UpOdumj!KC^$zWD*5-j14o1f>_7B&KtZ( z99o@0m&5gQ&+Z5f)8RgqSa>q07GdJa$y50Ai}e>qO)+7&eu|k{&I!VFP&(c3Uh3Q^ zRbe3l>MH*?C<0@nfr(=DKu(vwxEFwb9<$#YS%_(nY$LM7x#IMI&_E!XbnWpr>?<}rL*vKot~oU zU^fePE9Uvju)n-JnccA+FC%8iGO+9UYo+Lj+vXWXOKvVqlmnjyS|&9M8>Q>v=N|^4 z>&O5z_JmY#)O9(-PnflV)N(3;ofjiOeW*@yg%h)2Yk0Tv1YHdkZBK^Oc=ZY|1lPln z4O%!UmO;Y8!K`>%W9zgI)8dfl^E=t94BS!f3q`Ep;rRzn%$v@?6X2G)Q*FF7ghMn}e=GR6+_O^8k zpLipc5w)c770Vu+>fG}DECt9{oxZkr#?OJ+_eGN3o@2^pwrxa-nV&pdc7Iu)=Q*hc?mFgHzF2WhygFbT^s-G3!ZDDFR3JX$w^h>YcNcm{4ZB6sDu3_Rd7d~4Yf(_(8^TK8 zQNZcm530$i^`v2Xq&uCop{gO5S{{Ck@uUH*+X zI{BT>Lm?Vj#@u?zIoQsW(}f|>vYi*}w1@8Xy~xuaKch&Qsc~tIxG6jCwOa7x=tWLH z{izVIiFIm?D5p1(O6SVJO=Jykn47@E(KU)V-b3z}Z|Fy8nghdR9=wuK#@pmM6}bvup3LtMP-cU&DMw_w-gERB*ZoJRZ%VHJUHcFh%Z{v zof{uK69(}3!&&VpI{?dPRV;bc=~!7xbc=>vV&=yxhh0v3#qsi;RTa?D?gqEP7r8LK z^jW`VC|6J`(8U^c*c@07sz~QVG0->B2SlsOG9w$1FBo=dl%@vhxK;X=j`$WT zGP!qMCbbq$U=}R~(FDex%q2U2`p@ZJhIQw${PcRFX6GE+?}`e=1!GsKVoDAg-+QRY zWAq&{>Zw(MAB7_Ed+2`g7VlfU<8IY++XF`&;==cajyT-(YY)V;I=y=W_R#A2Jy1xn z$tg#P$C{m@C6&(T%Qi}HEU^)iEC$RtRz`nlGO#dz?GMY|{I?(jU}S$%rI8Xxa>nRR zS5tDh^@pg)bYEmc7CGOM9#JOpbgW0nX9B?p`;FlyHCk}g^%!<@=yOU4OyKkiSMkmS z^wQM}pgye%G?0sh_nqrxN0dkX(nTn!49f%teKUe?fGT${jV08(I3F)jbudE&x66l> z@ld!r=`MJazxB466LtpQZd0GhAMdJt)q75K2dDWZ&QdEr{1D9m241?{i$fY@>jB+S zZuOpxGEi~EDmW~7D+xG-LJ2(h_RrIyV2p8FL(mTfpJliRGGDS8~k)`h0;vuGURs1j$oUE^usUTvygFhd1)?vP}-o=_yP8_ci9px zNR&#_AslC@Wj;EuP0}CQqR3HJPgNBmtNW_E#kRSRrlYcYK<#&%nA!0=IoII0I`}$7 zlNd*`+vt%?rsSI_vXP3+Qr2>kxv}9GrpES`x;A9xLTn-iu`OmQA2yf_%W+{kbGCiY z3K#Zkch~>h?k%Qde={ukK9c8ywSv2#KcT9Iw`J}@38+?Xvu!D5jv;)16PC1_<1{7Ns`z_N9^}sMoNrKUnV67&Rr@MnIYE*@6Le8 zhf;$Z z{GtDy_n|PLLI~37^3Y;owjhoRvYJptfIv}Y&h6x*Ku{it#n&v(p;Z{V!7e*T6bXUF zLY-86l~AKd7hUnX2~>&rpp2lEdj)$1uw(j+`@u?7F^ZYL8JZ(OZRDl^t>ma5{)QHX zDo=1nx@$RB+^SMshuyH)cGLk!FOi*nGhVJys7jqnonN+MPZV2EK{nABFTvw)gX*#v z1Gcs$`ClSi?KmsA&j>9!lpMUP42S}X_W0~3gPf!Dx)$~GvgjkB&WTsW z$4gKtQNNV~s(7@oc1|+0d3lNibjfCB)9E5;$%n$7>7snmVPCii8-QgglR_G4AWi8T zg&96Uw;zwEi&Up(Y9)P6P13{#iDV7rk#Lv+CLIFRs%~dhz9?JS3N)(25cO7dKntcO z>3U{8gI^?>$x&V)NMeRBqu+JH27?*-_VxF^SiWO<-Zx{ZSn9b-tBE6XqSnaVb;g+f z^Y4ECt*=%Hrb$}JIV(OZ9(FC2og=uq=w@3>DcQs)bc1U!a|Q9`Xef{^E~b4$~K3Z>vvP!E}1j@(&(NrrWq3Z@$fR8;|1_ z3)>_+BGjPVfwCNB?!%BzjPW1!=^#0DzfNzp92&xAh;WxPRV&WR(KObi!E0N=$7I~q zOE-tiHu${XDfr*NCOhqTb9C6qS`<=p5VFgmBG=E^FHR73i?&0C@H0AIe$}Uj*7#`_ zcaT#+p2SbHK@l7!S$1CZT!(ErT$ zQ7cNgHNe`*^RcygnJ}tM%z(AuI(BV(bDqJh4E|WUmh2j@RML(Yml`8eQ%1=l)LTeJ zZlYqAL*xZ=Sh@ltY!T4qes8e~J8b0(Zg2+}JVVvQZ^a1ptssU=Xzy|^0;_p*Ya+DBmZz{%j5+K zEts?kVZ?%QCwq!{j9e~l>@ZC`+)I(ET+`%FJtltr2SiDtY~M z{OkfqAl!3_RoZ%?89Sq_P&4*I8^1nV^z{v<*m&N4*zuxc1_9MV+|lTAK~`7NYLi?&rg5d{tY#_xh#IjB`65HW*SEbP9SJbpT0L7dXeF z<+8BzVV`+za?N)JI`;>d6GDPToO2@YbC>gBAAGRYwGk?Q8tHz1kr$j0HN#p~gdIM{ zaX3s5y?EL!{?t4^#!>B_7IbWh<5w$2kEP0oMA|u6mfQ>5mxJ6@zF$D@CC}%Muj&kR zI6CTA$+H$nESWV{K(b^&E1h#Wa7CY~IWhCpX7CO{hnH1?7#?P^fFg+v>b=byPB|vp!e1{c%x}HDTN_=js_l~Y;sN%j}dX^;e?#`&^#>`*i(;9&i?oCom1=i42!Vy^~-7i%u6K7S< zZH;JSGz>0wRz zK9KG6`VyR3wNj^=s&jU0BnXLI==6ylJ~hmMESA^7)$yPPxLUYCRU^^PsR{&*Z_san zYG|Y@*{qrMWU!()j@CrTv$4EBPk)m#VCAvv^ktO8_z6b+$gHzOv)*O&dO#>8o(muNQ=B&@b5Y*`LH)A<+RDOl(<c9F14Zhn$Sp3tzHz(^ z!6B;rjUqSWcP$Z>fuGF%#48otl~zdZBs2Q=NPv4l>bQM$Lil=!yY++q7@l{KdtREs z>tzzclSGaDKKD9qf@qUlKDW}jj*H#$uw+#OE!3bJweKLk$FCqXokqHw7Wg)kcw#2| zOc*^N&%K%W6Z6xcvIiUkn0|Y5@B~?H$KjPsBY144bVwZ z0$N$R2q;yOa@iy%Y=XDc!g-$H=o7z;mF=UJJX&61sz*l!lkerIO1SNy7FQJ*%^Tp}nzhyi+JOVZ z5rr{;)Y`()^%91@Q|qDlw94=PqEQDd3c-SQ{3Mr-^N9DhYB{QfV0z9Rm^(0W_!wVl z9!wuDSasdB7sM+C95ekPke}H!ZvdFCIYCYHR6{d!p)~&(R-QWJg zu*CePZI1`(dc`OOtBgcZ2Pk9wY+o60WQ)}R^AW~hCUGVEM7U{u(CU#OV%CG?6*>_z-Cw_;caqN zLSUS86(?y?1HE!r*eXG)qV|oIDvg&02>N+G7d#(`n&}<>D%4A@;WyK&BR-Ej@XM$b zXE@0MWN`0+jlm<2TzZhGj(a8W@XO()rQE?#E%z1FLUs7NdsUK!+90-uva0&yWtJB;sGfaf{v+%LGf>B+Zoa-%NLzpJ|*PHiLOjftc9(9 znzusZ<%zcb=dNB;`7gW0h#5hzb+eo3BYxHHC^R`7p5~Y8-SXys((ci}u$eB9>~ur% zsm>*7pe0_l_zGtYFCp+Qh_JRqY~yqbx4P7j7CBVrh+E_(OY1$-IOij5w4zNEFwa~W z+dzgDV5mP(9yf$0!OGX~^X-a;ABK1^ck@Lz-?#+c6yiR@%oiVo| zypS|H-C$7SxHaM$^cYBH);T6~4+JFf%DqruIo7d~H{yonVH>?$BPwPOaH9l+A!_eN zkVO4du{xmq?Q7y(&N}E*lK~7EJl;o_(x*wi=;6W}Q1jBW7~13TngWlw6fG$aIVU|4 zn&(!=0ff(;Ffu2Py=N#K@6|++S@q8c(oFl&zb2F@Go9)ipx061vMwkH%;cW+yusfm zzQXxfRuDSjsh#Mpx0DGp@$n~mWm@L<|KU|{7XYQ8Q1yugsJXz@SG(R?q|2Z8Kr78H z$`-{IW#yb$!Kj0V8Ij?!Hi_2t?xIhHm>thufT8Mbf_|dq7tK5d)XE+(6PSM-bwEYA z9!@TeS5!J*cTWfH@L}#5MIl$GE!QL+cJ1L{{5yfOQMyLDZn0WV(4n+5g%wMdr(q0XXniLH;($@*j4nfYbQT}b2xMr2bzP@ ziJ~}MZLZ1w!N3X4|%y^dPCC79aWVOOgFZ-S{-js^R#wOWuQpyZT z4|&{45cpB)I5eSk;vLwAxJ}Y7+~Tde7<@!INT&+ooTrdJ!xl*sH8pIn`^V32Z!?V* zzb4o)6In)n{}ylTIe;ihhK44qeW8M#T;xpsie(`x*iP!C&yU+W%Wb8>Ctg+^9z;6q zID;^1aE3v1q07BMaP*tmD>nY!CAcQ~?X2ahJIl6xD}MQ%WpV$2GV&UxL{=+NieyufP11bPDTpl}1Wuvq zo=iIj$uX(nCluu}XtG2f3WNScjkI1v1LKMK5#};LYzXFi)B)YDTskK>FW7bkvXe&8 z*lD-oYx4-&rTG2hWd;lNo%$c{B^9sDuKltRjy|R2(3kEw71;$DQILd?4T&2-Ahi*; z;GLwOhawV}INj1*`itP>q+43bxxBQ<6ZED+Q^Q*$>f8pP3hprvyX+kDY6~3Z9}3;2 zR4W?!8s-{b3nfoQ-0hMI{%UAH*9D?zmw(hx+Ju)lXN0X0#WPnq9rfGp+fR-Mx5?Ft zKKCk4wp*!hns@oc%b_L>w3k?^iDGZ-FW-OLG{rE2OhI7hJ*bb z3k0hKUw9;N`rP$xj__}_0!y^(WsR~2pyL2Og%Q z+$@?MXX*gk?HXjp*-eJ7o|Q}c>4V^`eALDK`M#_rVl4=3eiCp#B1Lji(8<3yZxsji zUn;)U@{V>{Hl3rqxe#qeI{m3a!yJ&U4!~>w*2VnwzKy6}lb?Uv6j8ETov82ZJ8{7@ z7tY4G*fC?&Oy)eFt6_?l6@-4mIZA4IUBZLhmf${6Nz3CLC!3cSgdQa=!AJe^vvQB3 z*-0}i=i=u!G9pF~?}QL#G=gp3bp68#(N}8$Dsyk*V>{NtkXsPnQ56RdhU7z50;Q+0 zbWg+Fb*WymGf0IbzcvXL`>EE4kI2>us-aVHf(WC6*b8~o0nbXdkpR(z;4;=dv91Ql zGwhB2Icl-N;r;n%ONvM-JBQbf{grb@4sR_bucF8iz3#JGk*XL9K_Syo2mFI6gSv(N zfCm)EzX<&pVZSWv2yB$0-07%8dte4o$K$;-M0f={xLv@|Lg5)~NsQ#IOMs#xq9Hj> z<=iJ(B^Z)`%Uq72^H2c;1(h(ude;q5o?9lu5n3Z^ z_zj>F+Yil?^0`?5QbRYx>ppiiw;*`GsNO#+;L+rt8n#cv^Co5dBa?kZ_FtxdY3kUR z*^*+0XVjsGGwfPTZn#|)s5Hz8aXl}}IhMlsIv#@9CLIUuPvDLs^-?dg-44Ok40lxz zM=Qw&xolh&*16fLbCMPA^JH)7xYx6?I`+H#*W4}*HC>eKc4wV$WJh*Va**ZSN=2@$ z;g`9$Mx@i}0qAeUIA)3KIcpu&3JvF|f0np}UeBPHi;~kTAHOjbyBCXV-bQ3lYa)X)hIi0Iuw)54clZ!>N*t?Z#bzK?^Yr*HB0up(>d+3z{yk!&u(2=Ga7w6 zzJ)}m8ve=?q1)!8kmkzPh)rIxoG3w+Fd;DCZ5T8?3j8%*%}zaw%Q!W3jG)iGPgW&_ zs-k&qK7}DC=rX8`yZ3iEN*MeHHW2KD;r zI<<97tzi4$e1Yvmhs^)WExYsPuea4FtnR%1ZU%nz_aEMHGWfI&U#s~mS!2ii?_EY- z>K00#M3Dq4vd_Pd)-1lUq%t(ob3KEFjh*~PN#(5FVvP+U^xAB7_8Ng`7bORhMD5%0}s9-;duwDz5d{nx|9tvW>s=)&wMZt9hnUbp_KAofbbj}p|Pb?5_&CS(IACb6CKMOOp z8FssEF=Ncq{o-8*!nZ?>^OY6DE?33fvR>cTh(x#J!JmF}7rl`?wYH2!YUZW0W7QMb z8MB$Hw=bKPF54|I%oHQ94#*JoaH;}P>L*IjE9~JUff8jhgF4E%p~5m@QUj zz=H9tbqXtNOdIxT zEbm=3tm|(##p{P*MlAq@t>=E+@!0|5J->JEzv={#P_;apE|Q-nHy2J7SS*0YRPR4^aA5@xD!o1?&-Y2VYsIERB+HIh zlA}f|$v#S6N|9nJa<^E`UCUE**MSap62vEZfg;qxJtANIMwH;ZRLft(*%LOxtaaQu zuiOi3CRV?Jx?lP;+5_)~JUIgrxAUeq zpZHOBuvv6+g{zAJHiy3XZZp~N+JwWPm-85RC6$tIr{GbKT)CRJ+b1Es(KUlp2og?b zJ=Z`Ktco8iNDyU-wFKMCtjty>GuJ|IxO~-3khfNlXBzH!UeCgg*&k+t8p%Ov?p@3$ z;jyGq@<}+VtNiK1O>cKB8*$jpeRAXvfBJCqyI*wxm^UUN;eTrr84O1ze`^OxU}re& zH*o}Qna3IqH6>4_$aakMrV60j8tR*X3Xg`tDs#008#SlXAc~A>M17SOwl+B&Tsg;f zS|?*y2iES(6h*~Kd`cyzVGlFWg%(aS2-P2V(gOXnMWGWs!%o&n`|JdvzC3UQ+5>Z@ zqzuwB)&zo|!S{N9=m#7M$NHLv;} zxQ70i?D0%m_~d6xM3wLXxdAPLKqEhCVZIN3*dS6X>UgzYR-M%s%mnKRf5BMR{D_sk zqkiYU{WSN9S@CssNBBKI&g&IbM4ZU?84m0O=@@qMF7Onh@ zzDKLB2d79@P8Eb;+S#GVnmBp!Sof~|a_LtK7szbOQljgm+PdU(7Y$Q6Czs9#xmeXU zQNCzuGc~Mm@#2ZJ@~6GCix|`8#7ya0fnT8{T5$MH6?9(?(+fP8hhRNPpHqxSi-%gN z6Tj33_tFK5a@+QpeRdPrJ#SOw_rup?Wwq^>5^;Zakh3X+kc~Wz9dB{X@C|VftGdP8 z99ZU?NOa00D(K{Lzyk`um8w$bQtNT(pMluyD^}<>JBsx`aGTq3(X<}MOcRtmWwYNQ zI?Wvm6OQ^lkgWDq9rf!F-eA<;@j-BBGlDRIo~?{=(<|db<)a38Wbx2WCzmdj<#-x|qqkl{pe5pNlz?ZR72;l;CW^GEyw7Ihl49Qx1J$H~8X0p_1Pl5E^l>hg+$>mCM#{rKFp^CZQN z5%!4@!tyCOe&QtU`B{R@*l9$52=Cf~Jl%RT66jergeO35U z!&LC2K_om1Vy44xszc&iOakc%)-XL%BoEkZ*_td^UOw#;JA<8_{I9~2_uLF%`CqkZ z7s+Nv0<4WO8TQ?j9NKH8K~Tz|LWinzTa;~q34v`h`}oKM#6Qu3Pv`W4h8()OdBJ6z zA<2_ZrXpW9!O57|9cFPAyUDQI#Lle$yYZ9X7*O)b`>s-Qi(LfPewSmi-l*H|LrOkG zk^9&hBbiCz;6hgzk}pDi;PT*Z06uiY7?C9kG~yWldcSm0M$o=_m7xP%txQ$K#V}Di z{mB0qcV&g*v?3K$e}Yh_&w@)n|+UQQ+)5jHyd?mfIPMS zHKtWT4pREqN7X0;adgnSeqsY51=TL>%rUI!(ILTfn??iwP=^BNWUZ&B*rBRw2s zw_SHPexv`&zZ=kZGi>90a?_3%v~@;!`jV0lQe*(}lpq?WvzFFEV`iPUYZI^~WBjve z1l5S_6Q>KF7d-dQ+B<7-;d-WWS-NOMcFJGv4P=$&cj+`g03K%3o1`VOL|#$wh%Aff z8}p1k^^@f){jXTh=oPn~g&0(MoFTHB?1J50RbVGye_A(PPFDu1kQLP=?O+kL;k$Lk0jPrB1f4h@FS@^I2t=GN39ACLXtis-{}9kRRXB1sT_ZV5 zYL}wPpP)&eyg&8g5m-haI`%SN5(D;xcZI(piR>(&9R~}Gjh6WgN}fiM6e@CyH*}8T zt`Y$AWleamRMo`S$jbr>gU&h~WqOejv)8#BKwR&sIwiXT@vQ14DvTo7ewCjvpjfQe z6FHTC`-^F&A_dR0>FgK?W)u~)5EsI*gl>xU8ZO$YBDg2F!n||${ogrV z`kBFMJi1dXAy@5q#d>692kubvTNLR*0qP>|W&t`VY3@-0`C@2J3frn2Xh+y486sy~ z^Eerx;L}bLcnO?tdL8F1K}DV-=R_WK&w!52*c>d+r$fBeN8bUpJGfog>zbw9&h4c$ zm6gG`_SDFZ__Rr|>-;U=Ex4{#zDt729S3KB3?9oCZ`3@E5)}BYb*c604(^d2hmGKu zk{D=tS4h?S7t*y}k3I4Q8(cRtc}y<1at<c)xP)05@|kx?X3+38ET$uOD7@U0ln{rm?X? zudHWrYXnx%wgguQG1S-Sa!_&)3TP{N-OAR84$>G{!o_Pc-ajFOMhq^slSkQvVPu4< z=if{gSU+gJIy2N@ZnovhGROf((qifOCYRp%=*7KwvsxEUn938(hN8;Ao+<0*KR#keom2a9vzTBPmcHFO3#d z`$KEwQ3qsNu8kOVfQEq!Fb?+ctsLO$U-Lew9Ca9(oaIU~boY^0m6O_oyU9e4r;~nEd z2O4D~FwoTsw7-}|YLw;D4IB`Q$zAZsWt#|p!JS1wKh>WPFOr!v3aw1U-!YP3MZ49}_K?REvv37sk}zgwnOZ08*EOQ)e-hX~tUrOo{S z-KdT73^BgNpSmu~8@zA(;{p6>8}Y%H5O5NM1v{VHr#9~yKnLS$_W`&CBEnAK$#D?v(>~CLh zAsb(t=v|RfR3M#_gCDhnibQtR4dFSD9)c;ZMrjSN$Sp-O8l3La8d0ho3vp2>{9tBk6^uQ(W!&w1ie zhiupgX(e~0NhF#Y;73ua1{$+QS;`)=-zS-gnZG$OdBz92|0Wygod6)7LE}ZQjHZ|U z8&(EpntOx8@MO{h_UoCOOAlbw@Rl306d=+8uX z9Bg}ZMR)}uhjKVbC7mp6UNWH*N7YLPp5({xz3YADRd-x_rNc~{MC)1}pgItC7NRXc zfx`x;YVRE$x%4BKB%dwbHlALiSztBSMpxSQ4J!*Wy*fMLUtJ9@%Fo-$N|I#9oT(C{ z1t^n}1BWh^itLwW5@b*AR1`|E^tagsmLROO)-aEm{m{A$PXJjU7dV*A6w8|38bSSX zDs&rIVdUAWVq1W~=&Ud@ZR`L4?mMPbOE$Gpc1(3KL&7YPoqZFi)yS}g07;7?OL=qQ z6mu2y7_baLHpk0zzaYG&HqOUjHhx=Dw~_3AZFV-bMne& zR^S=uNTU?T7~rSRixpT@BKh3?fAIm!Q1i^*n0fW$UowGD->Uh(lWBGKE0y~(Q%bJi zreLr2D8Vfr3QJ(c#uo23$vL{k*)|bX^S9sf**8CiEj}^7==@>v!QcJTU`>vH=Z-hg zu(Kw1yp4%ADonaZ$?sC+4piEb%D@a@#i_r&e#N%=(LCff={(vFa?ACkQ&DiV0Nj-` ziVP>Df2$Sk&c(r_4oCwZB1q4!oHHUBm8f(qYw6&^A)+s9Iyn;!OAfv5YYDt=**OnY zCa9}s(+50|Kc#AP+AWTmzu!+MN)t1Ghd(TFSmb-sX%JSobwRywfEsaYSF4_Y`2Dy zUPvBcM_`poH8q>534p^E#_72+|JpO7;GYdPY3oOS_jPjNwXsR}j69qhl)Q@~*Qv-1 z9u}Zwa(i{&Nhc}fwt}yNYnWPbmy^37_FvkUsjz+o4OKGJswf2mc5`7-u&P(OUJ4w8 z-Q1xN)Y=?!x$V;A*2(XcX48`mD1oJCL1v2&0*O*1 z?{u|&bZ^2q89S*qc+K%#^KY+utJTdjZZqq6*Tk(6X@U~a+{opWa8@?Zn`fkYKq8u) zQP|c5n9d4|sW!dTrteuTHI!p);Ex%@QnM||{}S2iNcI`o=p0H85wQ#^QeU^xL2|tk zc{}8lbHG-MQWZP7*lZzB*(Ci`@Ts6YWRq{UTee%Tt_8yr!T^+Zz-|Ik`nH@Q7W47E zrMBUWu$mA1jW$rnhXPH>Z*1~I_WLgb|6{CAyP1+FQY2ol){;(t<~|}z3f1Kr5yWc6 z1Fr3tk>}uODY&1z(8dQ^+tx2I)$lPReiIdt98?sXsm$TtBt24X@Bl2?$AepCsZ9U8 z4o(hz31~wHB(Z{RkW@ZpEZo76CMEAmAhpb9;ep%Ru5w;`yFhrauV zw@UtC@IJoZb&imM@l-VI7uzJsDD$&=x;$z+d5qiHGIy+&ZHnK(&~0zs#*LHEa;B6uP?h5Fc9)b6tzQz!gGF@ny%qL z8$zx4*sE36G3yKAU0pM) zp)lkt6wc3;sP4&5DR(&4kpjAuR-r1;lp$LOO&0LB8I;Exc+N+)=O)t`$&S5XGo6th zaAO4zxG_+O-r&E*8(1iN-La3)R#%%!$B%dK;h4p+Fmt`F4*x+I^fOaf*lk|e%tUv& z-&?#+PyhrbY)*)`|%Td5w<$t)Qf%@4>fdhc^y;|wtepo{9Rr~}>l4rQZk zwJ!`bbDvk1ascFrb)3}m^2()pDgIA69gCs`w|Q$^RcnG$gU-&Z4=E2h>UV19Nht17 zWfA;7ZBUHPD#XriK1_FtY~P!J;RSEaKf~kEfb5p@^KhH*e@YFoI`#MOT_u~?u~F>U z;6T;Kn07q7D0wCSYz z=!oJCL5T_~7v&;%0Z~!8de>3G!4XBga1k9Rj)EeC3g7c2!6A{OIgqe%y1!}RM8*z)IJ9|&$%>OtUqT+yIlHNMXZZ8NLW7I%cMhzd?b2h z%B*6(ZaaLE*XRbIqsRt=67$IpENJ&XLz2O6tdyS)(dCO0XRey*Xm4_~pRNPrO~%>z zl30>6fgc`KnJtO-)GrQ_%#p(VPHZn~EbK)o#XzpJfQspJ@15Hs)4S!$5_pAF6Au?p zjHvbouAK>71{hd9CvK3V)Mq?}7twm?m36!hm((h1aVu|K{BPu zo|l3%qysLSn9V-vft|u}+$M2@&C^+R_`z$B+yvIE4wt=<6GvNF7S9UOqs-$SqxEyk zAq|HC7hq`w5$t}CL;>ZOKT)^= z^Lz1veZf6S9f-Ebk%9e={1nNYjyxOt=b&0$f#{n0f}#obZ|~aKwu}K97C5oDpT)lH z#Jc*v@%NR-sUcnAyQL-`NtE?~{dfZC2IbPp{u$8eV`oIX=o@c4VqP?oeXsxJ*Se~I zG#is&mHt~lX`Dnpv-o~qr!1tXk}io2?({`p6D82oR2Zah_cXpsiC3c#xgd8yAtScg zp=uFE)|QHZp?KOdXbgPOT-!c5R^Q&SzI?%4vVWals({N!|2*N(e_y4xg^~Sy#3}ZL z=8Lecqdd}7H2xwez5o%UxNIZ1M=|o)9{KeAgbyTU%abYGc%39n0xw{A{3?xNwoqg< z71IQi;V9&U{hIwA_)6C#*egR?;B-13gxjk5Z4lJT1uB3Vh`WwM*vj}%xABu5ZY0L` zk=gk4m&9y5{;ButZf1bItNZv*B>qbir_Hgz`DTjIQ)Imy%G)C zUOAk&H|)eA4V=g_pC||6gdAf>k^{lOY3_&HIODNBW_jc>I9a!eakU2ueqlBu-?{4A zLoR=5j7YzQ5&4v2Zc*e06|(`PwfmrrxL=820!kb~A-`X#yXL*k`+np)w?2XmDrvxz z-3I=6?z9#L64AInLekm_QY##E!N#~fJ~aZM;<-%nMQc1ONHR3HBt`ZJk+*V%XQJSY z`kcC#m+U|2((iFOXw%I7n#Z2Nsj+G1adj3c_UfKd=6xq{!{k=CT44k14scwQ3&Gn4 zT`&u~+ZAFw?irF*^FU##Mb;p@;fA{7Lwe&VE1m{KLVY0dF%rhd2F_>DI=T}h;B)`H z-^{UQ9dug5#KI$pF?($AfFd{0nfJ@35BXf0HPR%E@b9`3{D z3T(G#*>_q)!NQx50imMs0_g$uQXtJ)0v%vFRM|eGc`QZF+7{Vr4;{*Tqj;{VXAErp zI2aOJ(O_xy9g74UogED$gmWf-aO*o#YrDZE7v#i=LKcgX0aAeTcs@<0RB4Vqd}vU`zdCAH3{QMh&fRnc+!u%KIh)(4$R)Z0DllLeX`n123H zbz^oVU8PD0shgVaV}DfHh#E`7^Xx0T!zlaPdwXB@{jxr+T)jt~6P_1-FtBENmD^UY zo6@@^FQD9~NP32Bb=~TT{r4-THEVMvC49XyA*9LkqWBoyAT)?`_(;WM|M?z1QDd0< z;ltSdnA565=>#{gzwE?1sCU|-98||fKNXeJE4Qc$)W%8z$F_NnFnwHrF~Z{xJH-hY z6SB{3|IpLC@VrsC?E=}rEs)~Gr75Krfs`E-lR*J5U`z+pg%2q9xdNGfgRCj!CL}O( zXW@F&!E1#=@iSBRN>;lyLPFDc6V{DiCwaWqkcM#}RN;ati}l4}Xfh3)^7h>2lgud5 zcD}76OPpAl1B#=ERZAyPOaet#e-VYh?U^rX6qft7$RG?d#j%UW$kSreERX!CL(g)u z922wu{arWfM4!`cby+fyCb=-IU}yoocZiIkiW7E)bcN=G)d%%M{lmvo?}VnQ9=Pe1 z4;)^aHZ;370mS0_KRzK4P5_zkgM;&XqRmDl;Dgl#WS@4Tn0R&QspAPH(`Dj;IgBHKTw6JBC+yT}cqC+M1B zed80Qu0z=sn!syPb*Pf(SBIoS33YiiRz&s)Q)Uc?4!8`R!Js^zsq#hLu=V0<2+5Di zRE+{kmcX*%)evmC{N@Iw#pM6Z3Nk~f%uu?NBONCK+~`_X)nb9d0{CPv(p1A zsRH&de5ZKP6=|0uhu;$keWo4>ypM_g9~u^-f=da%mQMHI;nDueW#j8?(GHKeSFXZw zleT@~RcZ0UcI92ruspAA*5Y%$;>XgfixPN$85(d@uwj`>yP}EZJw|hy9ZjF{)*k-4 z8956pgcwzg{0wvWHbATfI&>OCWg{8>r5WhuM8NOYEgy^t3;TqTP(eI&) z4bBIY;MrQb4VI}Me&%$1rpT)th=JNX%5W{iyYXUdo(6JV<|Yc$lpT>>3xf&R z?q4~7rCe8`xC@aH1E})Y)7X#r4>g_~#VoK7X2&bsX2EGYnS=9p|4nSRFDtiP`!!iR z3EKCDI}qC_CY2(aQ12KC@@@-s2jE+|HY{IKq5xx#f+bjzy*5mTMITshjW?n$$+!Sy z6!;jI4GbrIIPc2;-oTrwubJ`j@EhO!C8__?_%zonV1ABb&QjzHpGZT{p?%z(UwrijSO0UF? z8L~ip1`59EbTb1_G((jhmQ7_trW@k|J^bB%8z#32D->z64bwlF6HhI!P=G;}_3%x@ zu4TAc-O+qEp6|On=PC!=YWS!4C1c4zI&mPFRYmg{OmL;&XFe6^5Z{Cjv|Uk`r$Iy7 z8c(cc%l8M_D0<`6^`s`qAU0~R<^Z=SNNk!y!zFx_X5FLM&ghjZ+zOz5CIg%u6FCVt zOuTrS?1-8dzZr)+9>fNxg_%u};=gp;NQo1>!lx~aOAW;wq(~(dvs=+8{VXz9jYg(f zdrjI+SAOrR^!;{as_cZQS&Q|M*XWzlbGfYjMS57C-+oe9QEmb7@ zpJFcYZ$$KwRGvNz@5}ehq1OU4X_N9V+$9Tp+*HX&9xbP1cagF)%ild@dD`}~6NC*OWQO;tPk;P(wn{q34bdSp}B55q^46PioLT3 zU2c1tR3PlV4h@^<(ZT4crE!?U|AQr$PueIIQ?Eu@Kp+Qmt)}ZqZ4fT~{9v@te!;jEL?;w6tkIvS%_(KYjeXUt;-~LRz?_pA@9iR)u%}B+{Me> zZwY}s=L52xP7q|%wX*Rllje>Nd*fi_@xCPJ2+sO?<1fwVczmZ=My_z%WIOSi^VkC9 zcPQpIMQ&0tt38fru)sSV7&h)JK9;3{RGlsl7|u}Y8{5{Fh(1?j)AEC_ohm(@T?> zc)ny0$O7vyu+{FpVKPWEqFnS&hy$!;GQv!)+15)S_wUUYYnGjlEWSWjG!Lt+ogAmF zr(khg4DxuGup9hpdIwdd0us){(wwkE{3g%t85zF1c+p+gWKA7!RaBj?Zq?)tKcsZP zIHmFCHX;7P+hauyid~NCC>ZUT`y;^smLMr<<+QvbU$MA?Z%M@2;}2HzgDU z)i`-n%u=dd3G%74Q`7@4-E@oh2O#0HBJi$q-<*=*0hc1T%&#^m(gO>Bp1^Ap#!dQE zVVGJ@cMB4DH{9R42I0DF{##2)k=s>ioA`Rj@=4irY2ZWR2zPAkhyk-#;;;jcY z-421z7ZYYa8V#PF8N>l6<}F=6(b^w&+69QkQ7`lEbI;PW2|I;P1egwZ_>G&=KCk}h zZQjS}bDl`UeU$%@>TL5nKFIg+0o(%pRXi&!wVRQKj^@5-D=A8ni*YA*2t8>U#K%@@4X7(7ekR z=_eOb7)MBlJxrIpSiF~?;d&F2kjT}Lrt0Ajoxcu8Es#`1ZZ+gPGwotBCU62fgw_KIre!AF51GVL^0@>_ostvC}H%dH=Dr$KMRY!oi>H zAsJ5W3s+nC!n-I2oR@qAmZ`=VGt$U(2|MrWBqwZ#$6gsUSA(1##2%3|2z?lg3hN5g zg}~(i;Yd@u+EkE-DQm39>QI{QjkiCmkWH4$hHbLF?HMP_HF2YFbpH>{kjbF>|ACxw zV#s`I0hucl(?XGEDkjw{&40zTq{wPI8!|^}suHg~vzk<$;aGrv*L9zJvqIN9r_;Av zst?0jhs{&59(}LmsNYdJxoLo_%yIB&PtM``8UwUK8U5F-ay@7FOxm9 z(p0Ba36cxE1W?J^9N0}T-Ko38^e9Wba_MX!XB%|c<8#(0+5aNxb6+-RTcBQv?+@-a zmK~vrqlqyu-Yx%?7wypO1uz>hXnD&E5XlK(6EcJTvyL_ctoKL$r^wPtpiMu#glj#; ztfR;pti&``reMMbZFy}_*BtCEc+m&-g1>h(mu>s(M>F!+p5gY{J8i<@cSrC3i?v*q z({8X>q_J9L*GVpLIGL+aanqbGxh%dc)-}wl@ioZXR8?+i9vAFW;b0SLYy<7n@48K6 z+T$ippq==wpS9*$(}FlHHdu&0hC4CHG^I?%+iKvIllb5qs&QmqO3Vs-(eC3tfi)Mgna1v4avKwU7MJ{Guj#<4(Xl9LR%`$vMVgIt)T zMTwfU@CO~@cgl|XB?vx{rMMa!dYm@i!oGxk{_wl6d7Giv^vgYqNeZ`spc7Yl?zPyw z<@zfT=e?og;i6UwL=&0gqCX_ zM;oR#DDDZ6W~5r33X#85ra|sN+hwzvdg>t$KX8nR!cpVyhreO15BqH0uM=k-SyW(6 z=EtbHIFwavlp7;4D1?JTCH?LWobG2PjMMx-^R(kmuv?K?*Gu29t_bJi*Elg|SZc+w zZVH%(%mkaaJy%2K@|}0y!6uX{H{`0p=A;%G(u(VlNIpw*QD&lPb~KVWp=W#yhY^y$ z`~FuNqs%MKrd&k^*~4w6abowU$zr8BN->8hQcK0eiVWiN$R3FBXN1+vgG%LmNf(_l zc`;@Ou=2Y}g(_LsjOm_i=vpeKlSv)#g1AeOBsm{qkXP|xL*6LICKnuPh}PsV!IHVd zCX#^b1g+O$vOgS?PD@2bN}SWd2KqREE3;}|#^lk%ZPY*-4l4HJQQN6CX@VcW>=cUV zxwgoROU`w2+oM$+AAxBA6Bg1XJEcQ?X?G(wF}f}{FfH$-1#Oh)L9O0rWC?W+njmu_ zKk2Ac_}J%u)YJRi2icdPk?Pos>X~n7%qHh#%-=SUvXS;xPVB-qSQwUiim9Q2X~B=R7Lmdw;(2&wCbrKyBn_Fe{^%nND=ViCV^VqJa8yoA81A712u3swvx| z475HhBY5=o#5l2I2Ns(}c`j~Vc#$1g>+)RpEHZ;A;+>D$NR<;O9wE;-jI`)9#hj$b zF)F53HlSz_XGrye4w6b*m2LD@Wo*Qis1F3&eUAt_!;68F*K~|q+di_HEb+P_&5+i} zYNIkG50&fTy6JZcuhTb=hhyYSEktLj`AqY+I$Ig zeMjDuAEPs*ebUiF$Z^;)W9o${8Tkfw?2KTPs7--4|Jm1!C(YWGHDsF;;|Z8&hT*A* zV)7}HOT~OT`=i;%BC+H|ue=hTIKM-2Kpi(VO?5xAI;33PAV<};X01MKx3tc)L)jDB z@4k!Qp=|UxOBaz2h|XM$JQkTxnzY9vZ%fZ>M~jD%XXA^rW5&xDL)rC+8(t{EFB`4X zOfS{0nzgZZC%^R zq9T(odJ+nH??z`6KV=hd=Km)6o=zG2r#yoFcHU6MH(&p1v~?G))8?arOmW!4l0z{& zD5$oH!FWl!;EtyWm z6b2N{YoilFT1kWJTg|GXd2cnz-@4{?ZF)Zq$IaSpGZJ_Sk`KkFAe} zDKnr*H=tKs5m4dgNL-BpRSxM=8bT-~TUljIYUZ>Cg@tV`OJmZm|5O1R<ylh9QGX}4BP<2H zx%6GvIG+E0YyScdnmGET_mW^b0=^)KpOwX%S?u?S*uaN z(Rz8~_}VrU4|LV{tIP|P_S7#9l1wMwDnn>)m`77eF~t;6K*j*6kk@w2?cfIC#;IMB z0Y$TReQ+jlK4M4yS|&DlyS5aCO5zCQXP+;4{a=MF?QKBZ)Re$n(GiLVv!b47OJ27TfTGY)vqL@C4 zJb=0x5q7y|gkdQ@W)yFGo|VLbudB<4c6_~ZHxIw-)pUg)(l+A_gDz-G(lmWUhXpq1 zAsg%Xm+{v5r>UBJs(f;deRMr!sbIiGAAC?V;8H8Ppze(V!eBwltR|neOs5Rl9R^$q zrlw6BaM|iP5TrY;Izb=-l_1I1WP2|It(7bdO0lGA_AAUyEGEXq4y9gjcUptIpRdDY zRi0`fXuo2mJWZ9rOO~TFt?604@-zs^BHwTUT?(#!tfK&+ALnGJtU~@47XzIC}s^sR#GuTwskUYE)UW5BS z&#-PP;Q|RK_M2H+P|_v0=B`r~{wj|LT9Cr3koBYxvIo_$0ed3Y6P4-ySh&Z{F-#zA zlVJ-so-Pk;gE-*)FaG5A*$>PMNcU`# z0Gg8ICe&YnY5|^D;c4Q0-XB;Pt~&-ae;OfN&%S`?T`{YJ#@Qj>tZOCb(7e>*u$~?gTdSy0JKYde!t@p3T}LL9OEb zQQ7oapDkgvGL+BQB*8l?0?t7+aMgUgt4Zs)Oyz{J0_pj182#0`zdP_9z8Nkqfi6o) z+(_ogiG37ED-1J7dWu<3k#(jHUbG748mxA2>{3C>hj%bLW*BGhmc#6iekw=tB)b`z z<2p%cO+3WKg>m8qjD@DC6P6~uvK;EJ9=KK0Eg?5%-{4<`bqb`G4wGH84wHup=*~BV zk;lv1<$xR0nllOvjOLN={rOU*o7rl-tNZv*Bz_WD;^AM!%@m`j$a*T~9Nh=iLo1a< z(oAzg7 zU#e=BbPF6mm8!)AHE@2@x*Bn@*jV*{)U{LgKS2p$BY+);RF;w10i2X_|8?vAFv~`A zx46T#D?ClJgsP_XR83H-+(0irLO}H+cstW?++Uz}-gp^c_@~6mt2OUTwSoxxOj+zU4SY}S^GH~OrD-%HtqDU_OrxU}1rCbt4Wt)D{{KEul^P*$)MZY4?jr{Y}1O9QvI2t?ioY;tp z=gy!bE|!vYUNg*oXYZTx>FTS?|^vew;3p6@$1rsI|H` zYpX*x%boCsF((M zd1R+BdEQd$CfNn88N**)fFS5WVY;+Yh39umo3#);k2)slribp^H1oI`G|_~cWjMwz z_40^jE&j~q_0BB|hR`VQ$&|9O_Jmn}+wEgP>>ux7hsg*IH8t<2h3{KSLwrdb6c!fo-iz1m+Ojr1B zX-{P4^fuuM*P`&Fe*Nw`APa-cT^C5w=#^)o33`tYUK(_1;;ob8l@(B_gprd*c_A;u zwK3eGjX-0FBTztl zCF<7fqjag{hIqO7GGI(n|}_glv-R*PMgSRxNy>=nQYx z)(JBCU9TIynf~h9e{9mGF8b4-9FvzD6F7$aTdRq+$4ho_jI<6F4!n2CI=1#DS%+9c zgxDMp+aF#buUx$c9G)iu26~4yf%nM$JkzY*%&!X1rgw&o=cbtxlI#Hy8^n#k-Zi*?TNHG_e%>ZVbi@?)=hVO z9&niD?BePcr0yv!w}f4H7vRctJlZm$U@!OS6UY~Yw<;|__6ew_s}E73*UJ+!G?_! zcf_z58g8dwh#z)aGm=a5de-m$ve3M2efP-!*+mXIu@4N@kHdW64=LsZMUGN2Lp{tz zUgZ#rMMm5wK5;&UP||3W3IW3QSslE3&~DHxu~#Bpk`ZQrUXKb=D};g*6MNAqg4Z7eMjq+4px+GtM3F)&w0|Km|ycAT5MJx5{rKFnvIJ`VE zmp~wav2& z%nE=GBn z2Kfon?!CAZdVJ18kmjyy38@Y2k@mY^mG+TFsIx}#b;ku|MvIG~1!**MZp$<5xENuD znlNG6)<U{QY6D7@ z9ne4c!h(8R=EBB%I6}$BQOAB;eZ%+v%eti6X$=WW`SdB^@X{+2c@19BpK$3A;pM&M>NR$Ylb{r8nILM?IR8%&Zc3qI(%nG8xZf(xQf zkVdEo-Z=Fz|1kf8cu!QRvMMBHdO}E@=aT@;CDlWgNv}Lc7kjmP7yD+YYUv z6=?WxE@0ty@9zh%7rdB?-FeIU2n!W^Qsg6gZ5Zx{I;FJ%EizOI0lJnap$|ekWl50@ zbiQa*TNQ4f?iY7+ng1`Ia{LjyX^H;UZ+|$Jf=MUF3~)IO+e;@?OcF&BsF*VEP8!uc zu!Sp4mF$1dUAGiuHk)M`lgE82vOUG@c#K%QLrMGY_g{7@l_3F`<-nH>MJa9-n$Kl% z(FVFTBu!P#>v!*2aMmYf#$ae6)fI~Q<7?qLnl`sJg7ZE{q4Tep&hgi^hYp6KjY$;k z1&+D9be&(kBdEQ&n2q5JhUKxv1lr{VyV)3F8#A%WBW{MZN-vi{zZ07h7ERv@vX<$B zmgKddmfG!(c&iP{0~WY?QX2%l9k|0m%6@E3zQONNTqmc=IU4K@@}eNGN$=cpO*Sx_ zAZd^8wkNJ$*tl%}CCfelvp@U6Hzwts2r~!M6#u2$MoPG~7d!FJ_q4_8Q$sOO-cU)! z6sVE#D_K%GAM%GHXaU4n>rK~G`A+(bYXgK^dXyNw9drR|Cq)IhHM^IW6ILJ80y(1( zRjFP%VTpo7!47pPDE#6b$dHp0c1}F>oGvyvAL283DbDALw2oKob!mY;@MI?h+1+9t zRebiXDP{x({72PWWHUDeIdOIO9t#BRq!@65w^K3KfCi&jQO7LfmD2csR8mayFBWIG zR(q7v%LU7LM_tPo>f&B$TKLYVudL&(18VKfq;%msaj*QoS$oetAv&4gDZJpgpzT

    eohd-o9hkTU( zezl3L;|2^T_ECx~fRRoyTPcz#h(iMMmU?sKq&9 znBZF?N}r5!y1FX?jWfVr$aGNl1oy9Ts}ghqdwRa40h;np2QP=Bxmf!LSH=p0Av9Rc z;h0~tJ0c^5Rwft{H}*!F*Pu7{7G#ieC$3>?v{+RRQ4F-HR#7o1&m1?iBzSST_%bh6 zG~jY;?qE1bF_d~YCF5OGgqRWC&t>K9v0Xh-aGO*lqsId4z zAAAFd){d$ZfDW!iab`{_Y z%HFwcAl#ZRi3eh3T@GD2|43w&Z@Cy!P8iTT7N}RAR9xmERq=NJ%K17JB|-YzQsqgF z`JO6Yz4AI~gb?(g3-Ui5b%hwv9-otfL6=%tr?0M`&RS3hHGVgg^@2vZUfCjR54{fQ zb6uuq7QD4yiNsO|Bdmjj4h5(MT3`6W0?wM!Q&BIwdYMT)=7`^3S&zaHu1lU@9bypQ zn3pfg7o8#~tpCOm0odR``)cJ!hndDknKR;F!D?*A%>zT9o zxW@3p4hR^-yR$#a$wW<9?&m^Vch@_uZ^Y6`pW(V;dIB^>m+(7162KTdmi2;hXlAg! zsVOvBzALKRR;2~@kCn?6F3{jI?pW{T1P!W8n)=?K&8yds+RqZwM4lV3bDDPSYC5Y>~N z^mfnnp4t2c;pL#i5Xs%@nnq$pCz%Fj!{k!>2{3nDlddL51gH7>>G1)ZB!~GGQM>$VJo8Fy1<@A6{b$GcpJ+MJ& z5GO<@NfMyb|KPNQkR%WpN#NauG{tW3UZnwcd0bHV!bQe@H#~>WVFXW{B9s4e+Pa&` zX$=%he^Mvp4D!JrGNv7K13MVw!Kdnh2;;#zpK>uCposKMI&o&=OvfZ8t<1-0@nJQL z-CuIDQWFxs{;f#bY^8dCq5MaXu=;!pB}uG3zL@hKhMBThnIj&u=3o(c9E{ zGm^l;#?BMR2L!F>+C~UjKj!e)J6a}w|791MX_WUU`#m~TIbpFP>@C6op6(3Ut?CoR zPSq8JrOd!`hdZEx*QL>YaZ4MtBy}ip4~FtCcjP%hn+)^~@}AHtH*8rt#P0wX>8doB z*P$%r>vr(bICa3br{7&aIaikKUjf?Cx>{kGH(IO)#q|()*(f)qxq%cu0EgD)gms3m zWm?5;!dy@|A9O*cc&w(teI0JA@OHcr-*>5K8wj^yPZSQn)1yOKP1i*|mO@nfgWzsR zt{w|)f~c6NHPF{;SMvlEnAKzbA=b8j}Q?S;I10859G0G^r?} zl}YLZ_0k66QG&fn4dPB=hvKZHXcmfO>6I1aBDf|urHwwT=9B}g$YDX9-?&(-(E!Tf z)(PI=)pOk+y!1V|(An6m8Hngm8u`i|iI>USXp9)*&RvJ`m~=WlI6YX0MDDjeVY7}l z0JB%Qv?awSpEzx)6 zxwBR(x2V!5L)6g&dMYta(<0lVx&vxj_@wTfcn2R^Gb8rQ!q`v(uO$TR%@)<^;8pWp z5SJVcVr;V_FTCryz~kl>trv=-tr?40NPv+3$Drz2uqq%;mEpTBu-`q8_dwC3Y?9Z~ z4Zv9Q#ck!-PEd10Cx{P7~t1 zUgn{rKVC4FdM+o5 z>opwSXg3y^u+Fg^E=>7B$4{(7$}ExFY6w~3il3!H{{WIM4}DD?%!a`0(H+Xh`Ix40 z+!yWHe=MiZ^x4N9c81epW(!3&Q!!Z@ z*hqY;xI6W9)M=i9zQJDt`n1Sai3GTLygpFZZVWjcUgeuFF(sN`;P2aMifjkS7ux_k zz4dd4`H}$S)F#j0SQiUFTT1S{UN&JV2{r^~N|K3QnZth+imA&AP^W3u4hT>ddJVtW zakY9QgT{6gkNmu&&a$I8=_iW$-#1&1KX(1`2Kms5la!xWn6%3jbCDt+Q!zC``=l5U z0sjYNHs$GblPXKo3{jGrc~E%emjD5@!-8dDTW4$u%hJ@Vd-w+Wo_GtybuaJ|f}oA9 zOOfb0=yKi_t>6Lm8EGz!G2u0yc#}yGtp{75PLNV>yrm|nOOZ<_NIHdG3Q!^6=8Y1a z9m)d=^jPt4NmJop(`MB2#K3Nsre^w}OR84~H1EdS$`N8al)f;CV?DA3e(X>hA^SMt z#{nq|rkmll?Cq&*Ncxv1lv`y1#uAEw1vih1i4QuUdK)-=C5@{8hsP+Z2~^M^3v4y|KhY&3@pA%qukUydwza#IIc}dulfkmJs*k_c(`DU^JLpP)ea1ZjR+J@3(Q18 z#MzzJ9M3GY$y-!q!N%`biwpvndgXt&E6bs^BOi-Z>tVBFQjLLLvNM{;Qmleq?J?-G zYH~UhXb-Bf40ea6T6kh!{ft&I@>aG$%yYj&mn$jZLtZ6hWAuPM8_;Z@9a|Rt5kKfr zo#Zw<#96v*0 zcU@~ixn`X|QtWo{8WiIeFt!mh!|dDtG-hs=&i|dYI?1#6aN;Nmi(<*j@1V31Cat%4 zVKLk#*A@KTiUH_Kx)1vnl$ZdTUAml3p0_-rRI$bXQt;)VowVs{H@Q5!S=*v~2o<*7 zg4}6%kAplJ?tmGF9pj;60K{%&ItSLw84FpN6GMrGtgKd6OW%*UHTxlWRnTH2DTqo_ zAqV2ExrhIEH!}pk`EQ*2%dgE*k)z+e<4-!cMMa!=1zKv+M{u8F?os3p71Jn3-2f<% z2}_5rYv_$MHLvyZa8uMBb%yA0`7zZ)LZ}V>>w`hdAw*|%%Iju8O{TH-5^JB}`j)U} zvd+J5hOSzj${bPLG*MT*S52=|8ouD?#o_eIU2eUkT$n$1@g`m$xlFpCM4=RFI9uKB z$UBr)9QD`)5xx?B1x`c7beg;#hB~ zMX+QC#XvRpHVBpg2gqOBmCf3v#305M3#9l2$D?5FktX|h(*uG!|78E8(2w}&XWa|` zip(Ge`Vf6bx+1K~F0@iZU*VB{k9yum8qAL8*w?2KLblYLJ^%DDz>JbVlpJ11b~rIg z>MT%FPBFVEQbfh%DYNM{o|S&NqAIuM$XrpmB9pJ{^z9?M)WCZ{<^8!L8#(|m0JDOZ z?0=Y)lf&dN)bHWFzy@$6sF21Y%6?_4aEC`Juid*vc4>ibuk5fd25R21tA8_U)Qn&h z>+3g5?G;hISL0@P>5KtS!!-?cRUw4>5i`sMp-Kj9GCyO`A@->UE zd2jO`dXMtRIv$?XnbZQBg0_YYD3YR}>Ry#7hzrVwU{Jq%0^Z+9OgZ;cW{_~Q3eM|s)bJl~_=*{LE#JD8 zLv~Lh=PazkVTu7WQBB2^f9H?{ zT_8o$0hg4Z2H}89ouJVZ*&l$S5ME&-XTasY41eC47ArD6i3`zyOQ)hv&=2PyM;oT* z@s5%<5KC^VC1~}8IQ`XzH&2+PX5Dpu8{R3spL5C*MPvn zH>|T{JH@0^WGfYO6P(z6vLpJZxxm?P28m0c%Tf~O#I6xbUEoeA!#0-WVktv1 zu=%#gJ`=B2mIYx$+PKC3hUUw53=BOnKEK&9FoMfG@oe4K3q)pIyj?!|FJz4q<09X} zP;8}`6tv1Q)%}{kWNDroN*GK8PMA>Q-YrWG8K3514&Q zZHLE@8^k`AJ+E+s#}dOEuUQi^vM?E<_Th2GF*LO6rbt+P(WH4B8 zbZhQz#S_7KEp%*sU6=Ho1fFBWp{+kG?h^qMVm*ihUUQlV=k-1wE2p2dF8O!bbRx6~*kQ$X+UDjbN#NUHIn;gR)Jykslw_DeR}4dE3+p zf}O4xfWNikm7H%>x$TtY@ehg>nYvj!L>UaV zcl@ifTbK)i3Mt-u&8uYoC!y8i<5BhkBsa7TEmzO^Z6_T5;Km=XSo^G8n$Dbf%fsT~ zRx-s<4bkVmMRY5y5m*|keA6X5)FLo=)%%WnafI!-u+xOwI80jGdv4ZFiPJ7QEZ)*( z(g8LC`KOM$-*exs=#$=&V^B9ivS&U9Z*ur|y)gxlt-)wrT~yLIr66se^QYf`8)t27 zw{co|sJ`v5#EeDj#EC7(X^WNTsn!VwNs0*6La4Sxn|v1(nLh8DDXHV-cs(&n;3UAl z2p#b>xTjjZok5M+EDcJxU0RS!ZS%Fho;zKce3~OQTnI(W^)~T{C@$G*(m% z_6bXT?ZrVp3cNi99?QYlfM$L=@J|;OS(iUJt&L$RXy~M~G-y*GXl5+8n~FL}`H~jp z|3smk%@b{x>9c(hPKITo>eGF-lguuyw)1TrS>nXu5(uLXi>oG4Oaet#V;d!k02_q` zu>A*10&3=^sqE!V+Wyho$oa%!Ks(*a4HLoPf`T{Am^eTC(gbpm+cw9E+ddvzAm|pw z+@MH16@#s6nEFIoXI&OluB539@_wL=xgy=nG`Q}N-h_bc&G2*Llpx40V(#;(I+t$u zHUw{H@}$sD2AOQ*vj+J}D58O?m=0xE`2IO7<+^Tpn*Z9c0heq#E(nX+jRo&G!d@Bh zmp95$!YdZ&0*sGiMio1Kkw+j_q*r3CyRKYuTW}j{Um&M>B(j=*`2BRcGaOU2X!*>4 z68^RLx1Ls2tGkKr<8W*6+ia~^)*!o`wnFs22GxzpW~|-(_ty8xN+-5IxfbY2p%_>| zH&8J{1l~KtfMO1oC+zG;{>yiUUx0!J;0w2&7OPLgUN5=Nw!Pl@`Qvc1Llf#&NDjSf z#zxT$-aS(5#MroEfsF==0l)Sn6?11wlGic;mP5fN7|g(NMYVXqB|iXvo}Yh#P7k~j zXmn%gCqS}$KcduoJ4kOeKn}Y=ge1M4bcXAIOIhGH&rZmNKL~A5qz9hVoRQw(p9djy zz4Fc!4DB^*aa)y3C-V|rfjinOlV26mr9cLGly@temE^VCuMO9k$YBqrp)U-s1`_JLJbYl@(w80@z2jkrK7WC9Fc1~ zYSF8(J2+Vf(tw!XwKpA;s&ZO+*MXI*y+N*uFYyCW{HKuZzDKQ34a+N|BpH{H1dM6nEVyr-LV@OFCKpK+Wn<2lH3aKO+P z#=_#rHcW86iycGj{$Hrdx+3MJ!V624O4fpM?=!(^{uyo`2pSaq5zX4_kTpJ+gA5?7 zcUyL54yvSe${-y?e*m=wy=0l(L6ESTs!<@sY6Qo;WJgFyP-ykb&e~I}$a3!#=73~S zhWC9R^jc#@x$093?v1a+qH*eMM1W(a`8lJqBVdHE2X$i+|NV((4`kVj;3H(vh0_2| z9QFWppke;XI*M6Ck(E?TCvB#CN2S*m*~YL6_4t?!t2wg;KUR;7%gaT7nl#tjjEJUR z?paJyCP88G@Q*T&VxWwD2Ne?=+$i7ZmN-S1GV4?IP5P`xmoK^&_Mx~Ys93C5E)nI> z*P+q}7~<|LI(;$Wr8`Dr&@v%pO=P;{iu9yc4&6rb-Qp=n&44W*vc{lY;XIj8TK z^VR~cU%z_0(u|c&NkJFM#z|zi1y*t>27+Z7fR%_MuX3Q4$OZ9H00NSfKvOVO0xm}B zHvQzqi$T|Yr+O9HtuT?7=&+S4&A&m`15FV3XW&n``G0{DBc&M^LQx}FbLR!h(Vy@x zTgOA3wwRG6!cqVAQwvdzeT)C`e@c1n&q>slu=eO!kyIs3 z)gN7}!5bX~T(f~1y`giEg&X?2xa3f%@|%CWXrmN^o?oTHSW>dTu0UNV!v-m1mFXUz zRM`oUvAPh6$a^9y=i@?OAiBZ>6XW1n;1YDVTxlFT3-EVwnFr@}>U>`HP^!!ej`q|q z4wB4~^8QX7J*}}=9ZM+&L@f)bn5yXoQP-6#U<><@l#7pn0C$h_v+$iBe|ey|6m&Ug z2X*_^T1|y#K~%H0O_*Vn4eO7trmLnO6!*@}oPIDM4>$;m7ZxjUpjtY~>mf-EEL`XS zxR2h@3ixk~4kSE#Kkh@93VVW}268D*gR`KvX_(^2;p6apj9EPAIP;IT4AZ3t|U-Hq79 z=*ngt^X{V4C5AvKY!>6+j6z4g!To3cd24yS;H`D`O+kQCUymeEADb(*CI#c-k(4_51V_D3`aQb;@ z{TEU5->k#>PV3#U1nLiHP?S!W6qx`r;2?Pm9*VualIMMdPp{cF$4_q`w*zi#((8*m z&W=S)%z0x1ghfm&ftM7uENm@sZteE#fjq=r*Yd~?Xc6ppZ<1$7FcSS)I7rZ8xh4jZ zYGn;_z0yFphM?GXPFS5ET&$)`d8_8N(RrY2*YEC_N+sLy^#ViT0Ms37KJ)80%&V2e z=RYQp1}82~x?{2aT%(v)id>>%^hzL42LxcRXpgdue~))97?nC&-E^_RW!_z(jdDZi zMUZmEsm>O4@C?8%fFT%YyhxR;b2H%T0~suQB287P$q5^DsZiX7HLNqdR(K6iao_!k z5>}M!Bo$DjSN4%}(3XJ;nhmlWZn|`SI{#+G&4?}#*vzGq0=CINA$#01ArErQ`$D9y z6$)T>yJmZ~^1Hp$RG&mt3u}ejnQflk(ieG5b^*-`Fg4=k7XXc$QTm0~n^&!y*`KXi zabmY8-2!$`8{C^zow7!GzsFrKor%;UH830US1+2>7ksv(xojIwMl^bJd&Cc&<8;4x zp6ql7yYUzy{!cL{zV@li?A?m*iGNQLo!GmDB;l}9whW4aEpsXrgKgxugSUj4LTw4) zgknD2w58Eq@y><-ZUV2=yG@uArpptV1XQ=UIW&N4noEul8L|s@0s{LmoFFps5Bnw; zTB{>CZ4l+01w@|8HdI9|^|<$yD!26{UfDf&#k4e4H|Zs7m~8J_sG`MiScafJDBu4U z=o+32xd`+V7(7dmbSX@Hp>4u;ut4v%D=#j5pg2co2#OWkd8G@t(;43D_)WlmlTF`@ zsN+?IH*4`N@+H+FJ&HBVE?&7JVa(O}qlD2AoY;S}BgCdFoM1#1wpXt3Fk6~`Z6$A! zByLe(Cw5(cv|yNB%A}Yyifo}`b}qUKM6oyAOC@RkDDK;($VejDe0^B3-G9@} zrGj$5xJl`fT)$djhvKnt`K0l1f=7iYn=mo9oBW%<_B-qDbC$O4+ky=tEwF1y57uD~ z6c$o$RT;#0LLUgzj8e+F@%SL0{&Tr}qE8kmE%YnZOz17yQh6(G4XPGIK(bJMyr4j6~G zxHsv|_pJrzz9jglvshf7%G2eGI)OqIIbqi_eFRgAedM%9r)-}`E_9r(@mw{p(Q_@+ zMJM>@1bznXET~F$P}5ENJutT4&Z|{`rWr)@JrY7HM7^%JeadD*xIeIZz5|j=+)*>K z&th*Z?Q>GUa*IlU{iLR#c`UzRw(enxw!fzA+J5L8m}L2FTHihoAa zp{#>SxWmet&JjKly2do2+@xsr8+g@`w z!=>`IZ=WNpojCpq^F0gzTPS8T1)pn7GH};a(pjL13=1D3p+kx7tw!1PD#k$XiaNo| zngJD&Nc=fYHiG+nk8vVxe4N_?cx%d%e{wOSqww?Jyhmc47#$Gw8itNUidjpMRa8t; z&pXU7&u*f@)Jy^RwjCj5K(d+*X_Nc8^5{}fs3#E{6afW&%=Sx1pI zz?CFPQJ;naq-9V}R6nCtoFwSr#my`U9*2~S^~~7<3RnfKA9MI?PN10h^Uc5ig>^-> z)23)xYoxR3%K1q1`_?M$N=??3Mqvl@7B~z--2o_j#`+K==kIo=(-&JccDtrxmA&a+ zh=t|SX&QW47Hp7j5$SNS4(I?X3mzIvmjx~eas^|9PZuc`pQG;t;`kkm?xf-}57~vg zqu^aodI!Z5T4cR}YiH`PF<)2iS)v$pLE70A^{Uwgv$4nxW6pKFd=ZXmtft1|O01Z! z)r=5Cu@PxsxEu~TvW6}vq)k}-%PZX)v-vq0^S2G8%!$oUgN6C2rx<8EJxImGOZQL!M^tI7?}l`1|} z7YfF%qu?+QS)+pEYfd1d!d*!VV+N7?quxtolM{oe!U9CO6axvVOem=h%aEFQZI98X zNvG^`(E3PJkHcK@Q9%_{CS_|L(Rj31-A%U%xBS@1_;wKT#~EPGs)gG;9qI~?1t}aS zpVRH!kkUC}ZvV^f0Ao`a#szg%Zv7tBA!#aPesSpBvrLFhkYIV%k*7I9V&acoTy@r6 zg-+}1u=EKcbpgr(q8id}Nd<)3+o1U(cUCTqXT}>FHsJ!j5l4bxgt0l{f>Qcj6wEa* zDsNY5>qy}w5V9G*B|S(nl@!@S#nh>Ll-vC~fkxm6Z&`p2d1>Ni*7)kOMQiw3^hXk1 zlgdEf;9u9|OH4ey`O=&)5YR#o5p}P0*SxoR=YfZi6LwCV55kVe>BE8=vR<&yPuC2N zAvRe&5cJ79{B}XH2WIN*u(Ppwmbm$iBYf`MPH~zg=MCvkd^g}%)_iYHYgSl@=&(B- z^SpV|#fvd0_t$E=9ct0;3Jvt)#Rt@9q)0@()@zr#4@6da6lZu-io(JP9r3jj(04 z_0t>6eYPNCEaQ*%Sx$(Uc<$j8`ekP@?}Swg$#u|*Jry;hKt*(zkA(o3~$k-W( zq1fA=dzdO{J072H(J1iY05m*1q+tE)P1elzEM)N;r@saNYto+A>XnxkoTN8S{}hCQ z)9DMmkHziK+K?<+;d4ry1C;R>#0NdIC7txnG507u`v*UJW)vUoxa>JO8Fc5Z`FrCp zVgL5B$I1?ZPIWq+L+fB4-6?B=vi=;phK$cQ-vY3D*2+ zz&){*iiw|u;wO#bTu|4-d@NG09RohBtzNh0e!w?~8x;2-C6mfjdG>`~hvKwsQ87@Q zywx+q@wUY#;8=Q^_-f3+c+mjtz)5^m_R2eEkX-u5xCnCpOS5NLKXG_Oc-cf|$wV?h z#k2{HH2bP{?B|E$pvW#ux{Pp-*aEZtMb)d4Qcn<%yg0y^RhHpD+0vr zsj_}0Mp@qN=Rf{R$HIaDUA?*{Xfx9-PxD_Jh5_ANx@X3%Fx^Gc=bok6<5RCbMOFsr zGI+V7TwrC|DcUeuw|D;H1VI;#{I|MQip9;UZeh__n06csiWk`?_Ur6cYXo|O3E5>^ zwpb^HoHkO+lC44uq*pfcRs^PMPcgbHrqe{ z5$+rdR7M!iZl^dwh4Oq$JxgSU%G>3W|3cO{F;qZ=a#*SMR*Ff%utv;P<-m-5zpAK> zJ`IXy?NwzNZ-eB5xJzp~i#}49Hs5bGuPUlws$HyvEpx>c9=42Wk9)YmBlTYce5@Ok zSvq@7^#(8^qZ@9T7SFTck}~A?+sDpl0Mc#LZj9#ojgdh%<0=nKg9kvPA@h-Z}&;#Iyr0}qCF;L_!!(K$7#jYudKLo z%+qXC-l*Glfo$Nm5p&``!%_Thi;BV**U}$AyLz=CR%8(8($yhdG}M3Ri|mf?;g}jr@q)!T_I~lvdw+V@ z$BdFU&rUc`Qn{hTd5dz$J`0rOQ%o*JvZ$Clq3!f#aVa0`w9@H#qU%uB%!4eFuuGvs z7W8!bqyl;X@+Ga}dT?~G(gB;$PBEo{x-8Ev(K)0%5`)VI_|HL@8M|e)H%v_93BKY* z1)ITg^Lw|;NZv>RKqodF$1EUGNipE}mQyjNUC#b+XiC-tl{HQ|HX-iimrvCpT|UC% zy>!Wd!W2enQk@fn$w21OYJLsLo6!=WLk9h9O#?7oLsAk1)3E1Rue`3zlb#eTUJL_> zfloiEYfcVSQ`td7ZR`jcc{Lp?ZrIn(qUD{}erv|gd+Xxokel2LjT574xkX&Ck7A%p zshf(yj@1p5fhK#02X3g_y)P{QfZg^$c%2RES00HRbjgr_{16Btrl}sd70Qxd*~{N8 z*4=q?9Y2$AfNpsmme*l59xBeXD|1|HrsMH8Z)4o&W6u-ws@yKThH)a@ae6aT9{CtH zYS60$Ax57X!C9ZykOqE@cfS&tx(^Eq`Eam8wB0v7(AZvk?rYhae&urarJ_73Jas{x z<9b6`FSseK3b{-=l=zlazJo5;ymMUB;Vr8}8s+f>|BM0Xe$jz2BlKW52>0^)+?SFY z{A${1V}VoO|Hq8KFZQ^*q%d-2)E0mIT^M;|-oe0{>CIXQ0fu%ccY5rDe0~DtI&#AE z!jG$W1+FHAVR)y5Rx=BmGAhGmF@j@WT>a6+$JR2uTv$1r*bib+6OI!gK~J^#z}G;e zp%b=&neO{%q|6w^8k$bps1<^(l14!ytnkG}Fk6GHgTVDfSMd$Box*)1Pw1$C@^n!f zc}#npEetqy!+Klgv{7M}jZ%Z_S)ZjMqwK;J>0R%dpslJi(21HJTn@#Xb-V^wY(Fze zBbt6L6)n+z=w7?bR#vBwc93DK9$22cRu%d~7*aSr+DPz9#Xy)ne0%ZAHOxG&ofi94Pp)Dy^fqz1lwQTzTAsHixCxi5FuEAx2gg>}qj9@cp^Yj^Tml>P3B4m&j1HE%qwUWkk31=w2QaL4Ey^SXc=WYpgR2LRvVtHBNKU);?O{9pI};{M;4 z#7v{AOVdUQ7w%FwX=~|^L2l+6{gF3T&$p3U+A#Hl;5xbjXbxMYU7BWXtY}$q9#_(ZVYK|67~py>+`*zM>)_rx!^t~36U?7Ugp zC+qy4_;iJ@_RwKiu#Lt_n^D_{u}p(4h#1TGqkZ<&?|xEg-LB%aTWpp-n*VNBRsvm1 z4!xX|EF6F|d`eJ*ut8`LmqwRF*V28mO_CCRZBz?$479=zOG~0pD#rx_w#>Q3A^|T( zJC9tRV_#E7aHgmOoBqk??`xg8Ko>+EP}fXP;2i_!J5!P$a8y)DJ^JcKev#*K9`Zh286uD2uWX;;`SD?-WZV_Ws zQQz}^`H&DgTFkiN2zP~M&(UxoGFJ&@_zS0m%aQ3 zMY;GU0aecMwM;iTsL{0vw?*XuPdP>mv0A8Bh&O;5jjS^~O|@%go9Yu)=Jacz_NwQv zR6bB_^Xvir&igY^CibmdeoIIh-ATWN@nj^>%cfi1io>yj16s#1*z}MX#1F~xFk_J^ zE|e{@W}dDC3YJcKeeQEufT!{$El}GETTgrr1qz%t1$FT8q`TI{LthdnhlQpnm%is# zGtY>#+n%}8mdT-_LT}^oW6MIHS$Q}ZrDvXX?Y%-3|$1?qm$yC!9qxS(u$l_!*A%eMJ~o(pDZP}8Nudud=BByd3x^7mLdMEqi$Z0B!+t!dne6wEO|7Y(_;G)XT^kJ`f zi>iZA8wFLjDHSO}5L*_Dh?vsuB$K4GOx8)hOlC4kq_a-8NyVPdB!ejKA}FAO7N9J$ zDxfGTn~mZw(5NV)f-EkLt>8kV!uP&ag;E7q>rz$xOEW(cx!XCk_dNGK=e*1FAPtqF zzYfrE?t$phgRoWLC`PZXK|UHgM)Bu1&>2B&AivosbCNwV(o`_>kqGZym9&U5AOPJR zs_k*ypfaZgi{s86t2VA}HJ6We*tIr`RP<8ea#;i2N01J1{8Y47vCT7CYPeGY zxf25cg`qi;ZBjIDgHF5LukhA0cZ2PTIE#=qVi|3`U`L2W{^^5D=F}rBs$?B>f*7Zg zkCZ!nnp8-F94*`mHT&8sX~(n<8f!{E6}8c4R8`V()V|^VV(s`ghjq>ElXckW!VTwZ zoe{F-IwKgo+##MGLA0)ykotiHeWuBz_K1HdZU2DS;h@ z2W-$Gb*|{mZPJ_~E2l;UcF5c17yXZUuM;d2ZUrf=XCRwZB)&)rJn90L3DaO_UM_zi zy*4`y;t|!#F7-}+=ez@pK!6;2c&1EtE1SR9_sXKyg;ynAa=dDlz2S~~HtRC-`Y;%C zK4i!JZ!h>`#$T@MrP^nvy^SM-n^nOp!`v=3%f^RhyK z)F$L%a7u`_nuiKGNW`gwIDJW2@%$`MaO@4XyLuM4fgf%V#yqALHDosyqj-%cX8w4? z`{quVktsBeyH_JDt`=0Lcp8B92?v~unH-2C?qGDPqf?Gf+2hiswrZBJXcSBzvKT9- z&Fp^0j+K(S9q;-W*C_deKPV+>4xFr4VKP!%T z$KQ}OEK9C=ul!+fGB4h9ueWxE0HOfJfhVO}>?pAj99X*PxOzk-Hd>c-*Q8bw%gwqtaI#*$iFMIYObSJ|P?0_KM?m+R z;&We6A?g;DyA*o%0z=hq{++-&`D0QdTF*oU7I>u5bwDw{__ps2*!s{4I1{KlxunIa=tfMLe4!}MsSBT}heWX;0 zkL;H3pKUF2#sUhHxrFV+N^o#m!AzKWJ~GQ(hT38G%q+svRhV5Rj}vszH@ zK&&^+(YBjnU_Q#BBCEn7jYL!?zV?-k)Q4+b9);;tt>o66E|9lqSL_Gc`BLW2TnvOa zYC2r+Kzi?h6H21oQQnorkpZVW%F7E% zR$lxt?%M*X5iz3sqTi7?ZisQJy68$sfT?k+QxsRRn-c_7d{Y^}1NVt~k@ zh>EQDh!Z3#`iNfCdKZm3fTx(%;6iJ2ujelnw!1aZu>vh#N>s$mgqqHCq)yfiS#@xT zQgBy6eajznh++MVXBbbztNHZQp?`v$aeO=!s)AZ+z{K~ z&l(cHWdy~l-yeQR&Tum)4(xfmV`5HPDCRN+rL&Qh3$IBEL9)CCs+FL$HLQokOhPWa zc)?T2IpvB;S-bIMo{@VW@Z;zQ3KalHcdvB*N-A@33Fl`K^!yY1(%nGEzJ5VCqE zxaZYMs??|8sB_*f{dqR;(Hxx$pL`~;^oz6XGZ|)6tUa{w`}+O6?*y3VusE#Og(WKm z>CE(GW_rm`$SF??(Z-59pzH%{GPD;$aUg9qdsy&ZIgYgjJ=*3GZYI(7?|rvl_eAV^ zpsd?Tm8%lrYknagYw0&@o40Q0#Uf1D@)ugz7bi?ixak(Y-h9u@C0gjfZWWdtc3Mb_ zXnlw_jj58a^Zdjg6QZsID^QYq9WeB2b9iY218wUQ8YqS?1Yryj@3p|fOuI#TC%B8w z`u5E)-ds57+lid6cA_Y=2wkt5Q7l9Zm&YFc9e4+ftC!930hdV98?$<0A+f_W+Zz56 z*Sv6~rArfZL42^>C3`^$zc08$n#12dqjXlrf?EAEkcdE5`Br;3X|7B5`JNW#Ln*DD zZPVC+!e!XGKb!Z(?z|fnBwr8$8TQ zl0Y%>6p5uGmkQHV>F&$j@2g{ZSILUWMP7A~ZM4}hHsGAt&L92AA+U6yN4B)@6;7)W zs;=Ry0p?-Bk?FGzyaQ$l{`K=g6oW3E|3Dcd*uhjQtLS|2$tJtOo|dU_@1bLqy-<*W z`&mfC_eplWh%&f7>zS3#%cpdE<9u?3*Ez41Y*imzxMV4n#qa0a1!0z==w%BJd*F#3 zMUuWRrv@1f$?C+OJ*2>a4aq4JLsCvL2Pv|jifrWN&pggUl7mXGLE502m&V&iklsC2 z{RkBOc1-JDuu_?=#%EHe;WH?tRxWrHmO-Kdi@oq(%9NF$)`y;0+!?n-+%jFzPSS18?sDc(QE7FUvq7g z9px25lG8Sy*jX*2YHxi)R5rbJMh9I*BP~a{OOvWalp`4o?RSe34mu%O2Qs!laN9?) z(Op^gtS7-OBD_=Nb;08SR99Bf)&BKBC7+`@JipF4ZQeE?+XWvT6>JIoZ%y+ynxALP zsnta1z~<+$iTNp{7^tG%MMY+Nmne(G{dAAKS`saIG-v2Mefq6A;8A&x=$Ae_JA`tMO7S_AJz{O*F91N&*l3s^RtzKF|!BOL%^&<8=Z2qH(&q$ zfBZ9J1dQ90N0-TFZu@5k-WCD(>#${VHpOI6B%O*xDmvT-X*Vfy6j?xCsol)`3}jAl ztAupl8{G|chPhy7`W4Bpn9q^`rX$r4RVga?59h3Mu@j{92(SuQHp0+o-At?fZuIAj zNNN7ctt_&45@|4b#WfTI8f_K8+d_cqjXd_h$K>#ig*M683cKVxq{-3^E(1={&bQvH zdhZz7yeRXZ+j;e_CEh1Ik&I2q6_zd6_Hp|R@yn=BwqPeyyET&{8Zt{t{I zzOy;#SCfq}y7|kNzme7442}bDwm@uan8DdfF%ah3L`4GWAuN#FT-FKJLGZ836_jd# zyX!2`d8Mg!s!!ziraTVcJFetVoFL8xqgFoK@Wr4L!A5MP`(z}MLT=b_;K>N|1%_dx zf?^;VZ~${E6BYF`B#PSZg=tegf~#Vzg$8OWk>1^pnf!*tOQfr-Bncqm)~LeW)~b-l z;e$>m)jGA|zM+$Dx<%g!Dj4IwYqerOy^jK#MMJ{)nt`;N|Zw6G$Xe%w~!tP?4SUPOI91M;rfcg5X9Ly%fv@j%7RDIB&cE ztvP$;$hd9wL}&RM=QiAy_iOEw+@?g&wp-WbMsTD~{q65b;%mmsQepy_Oo{20r?Oj+&-;g+9BVkk}~n-1=(~K z4cb?j$X_Fgjew~K{Cuf&6XfSM2R)Tk@UfN%rI4_xb}#P+53803eQ|4qm$AR|o^{fG z1+vd+D@1ssSa!;_wwfu;ey7SVr;TP5k6%fB2obG6@$So&)DFun4St zxCA6XK~1d}L>LY7NZXvZd1#UREH>zd$8Mn1Z3gnotCExQzb<*gPf(YHJ>eV9wLPvq zuKmh>NUHjuC0H|k*7=d>Z%Z27iUYGjT%bhp_a(bM>tWG+-M=^xst-`N*!n7Iy8sQ5 zX9=2K@=V^tpZ=5iCfi|c56kX1licbuF>EH&K2Smr9I($2?7Mx;o~LWuK6vl zP%>|CuXD|v?|PVXLOHCBVc~zu6Q)EAI^{@yJ;d)0#T{7bf%0Cf1P21gZIJbY#9^fE zHK?QZ%gJa=I1n*#)8Dv1nw9#$@g(at+aJ}K?2k$*1~yE4sK^b<3uODW_PK+;*A%fq zg`RhU(W6{TFMe@j(04J0xd(k81M6T9z2D;kIU3LrahBwS_CQiZAIXzclJ%bbZmlr! zEY?3=1hJ=KJ=l(%g9DgGEOke_Fe);}aFHEMKW}+QWS+;(WwvqP^)3tjb~SHEU3}1K zolBPAZMXCDQ^!Ba3P#Z)w2Wc2t!=U_FsWH-G$ubjH{m=ZChX?700C}lAd z$ncG|rzm&^%hDY_g`v2k9mhhUp%JnWCqs{q+fR0!ELxTDzCZk%0y)QiqNg4e^j_~?#W-wWx#8i(>|*VBzp6bu9!8l|Cj9h!agTkC82SC~ znhj(pw_vgZn-<8j85T?~p_n}sDZpTI()2p0Jno@4z*;0hebA>#)5tqZ21PxjjciiZ zxt6LMj7uw5$TfcFCfVVmp8Z?-~(bdlE%>2*muBrnvt+I|Hy zR*bzsn|)R=QPckIzrSQe&EtRi@=wVT2evHNO>odaG3O|98ZAq&;_zGp```NTln8Ar z5cr^D{`@;FB$kKOgLTg9T*?K-{x=}ut&;SAY00sLhAP7>`usZ;K^b&3>37>2&=&^2 zv<~p991I1)^lUmuofK9cs;w5_kJG%?;9c&y{EDD6ftBPDG{FXgPA5o_mm$PdsyzQr zs$UOX9@-5-rYPYm!qs}&2qtEz8TIs4*2m;NH&Yb&w_AVU4Vj`7kB*aN+)R-JZ;;bX zAh?NQ;wZA7ii{Qa(9zV#z|Ng2t^_JmsF#lx7W>6<27t+1AKQtg7#dseV{f%`#lb&Y z$i_0{?F5n#pn4qmJZ*27U-j>?@NK@T8=K$f_C-2udgR$}U!Cxl(Ny^UXZd2X zWfCbh`Nnop46sz}pdwdJy2P$2J49RlwZWBHZY1SNvEA73-}rM=)`MisLQ=6OR;aHLqFa;&ydle+o;5 zYmDFe5YUg}Yc~ZQmBh^48lYXrFBCOPQPh*`%7P~mnGeQbiXqZD(5B9&BR z$&^E&-rOC0B;ckj0R}P0H^(=b(YYfxwlc-YB-}S-al?|b+o-AwT3p~#=XzDLI0I6%OGNduTyZt8 z4LE*M={3Ax*UMTKU6bUB5qnu71uhBv4M4}-L2IFyNP^T*@nKIuEu=)0sE7}%qdQ!0 zPpMY)Nq0`;GVyYzT<%=1LnMdoEb4QTR@80>hWWJfur2uGS;Isx7#Idrsb(7$klw`#IoT) z_H)dL@20W3YMCE?Wf;mm=G-A2*Ef%^&eFGUx7W-oXhSqT(e)oswB2y#tRvdWfS%jzVc;LeE zv`DAY3mbOPr(vCWj*k~EkW(|(Lj~FR!oWfR7>vRg26C)T*r&tko>XI?ac_L{n|F#y z&LmQ6vIQ-p7;yHN00ox{CBAy0bbGgWJqi{qT=Lf}3jSBL@a|Xc%If%+L|yO=YAgJA zFtFnkSVIS1q---mNi4;zrN}BOa%Xsndh^WV{@OU_9=dXROVIy=RO{CgBp2SF zlkvV05Wj2Qr6f1FWhgoD^s;o~aD&oIG509a4f>am)(8&y$wfekj)m))Ub;nB7HXRo z#e&XDiA38YPYc}ZeZ_EX1K)79NrRQh%R!6^Q{B$Jqdl{*df^8C$uD1B_}N#J{=HT& z=^k6|sr*QpEnO?G@b4jI;Mvvr)Cl?&yYvqqTDatI%Sg1( z0I<04otNrcC_+KK5`TkU3{rG#qW*IAyB!M=)=ipS?t^oC`L*=nMb{TT6=WCT>rl0fAf6z^Q#4C)p3H1ckTrChJBDK9+0Dzw8Ur)TX&7mA^dmUipl?Pi4k@3 z|NTNE*}x5T4s4(bObk>i#el`yN=2r^njTZPP+qHc_KDByR8M86(1(x9_Mw|Nxq+&P zWDnn7Y7Qgjn6V&aL??E*z&p};*L*kau*nQ8J7AJ zuf%e#Fpgj9SgeETB*O6*TKG68z)-itaunv@fj>AJNb>09X{Gd zNJ{B<>xLrKG@#WTblU4Zpg`52Y`+I?;4C~W$_&9=f=i-qMV510uwC}mmLS>GWQ;ts z2f#TYdE$rvy6!)$S+F|rHi>1yy7B9cPz%jIiI zTS#og7Lo*{{8&8}8*$AJwPK@<`xx*vS~oVoHrr=l-|vlEnm^S4Y!>;%f%D(im~4%o zQp{%*c?4Ty*IIq1WmI6Uq(*`oz=?`nF~l3x#ZU|eQNPN0kb(jaweyYyd<2`6L`9{0 zo%*4?7ZxQkQlZUBURz>Q-F{(qd1S7rZ|P zK#jCecDTaLK;3!&*7sg_+Kr80xF*@kE7c(pg|lC&FmzpmMnQQJDO*xcwm?~oH%sJH&BStT~ws&v<+ zRuVfJ;gjPo-$=fRlSM}{DHPcP`MY#^P=*?}-K$)3#i-hV^xBy5jcUDZbPlk@^}C^A zXmKfo(p0-Fc%F~R^8@andEV5n2kfBvad-C*%^8!q?9m+g3JRK1BODshe_ha5k8%-+|3@Q2g|q0Q1`1a?Q5*ODu=&zH|SuoglEKY!Lf>~Vt~ zDqp?vrC)d&=a%NTFaD7vyk?8kVv`vpgJQs<=%`2}azIKI;Dy6_#16=+ZdW|}MTVpU zmJ^-xv?U7@-!Yur6X~gKD}pfNFA*ri>Rrq2lxaOQv>agae5NoS40U0bpX@OC@INDt zzios`*{ly%l00rp5C`_moiG8=A&LRI%6(L1h3H5?G?O%|hL4&0aURD))21JDLspOh zr=wvh!gRNTGx~yW$+t-1Z~tUjlLo}t2>S7Q1o>g;b~#0kiZiG70Xa}fSc&EW=@M^| zKJ)D(x$0c?eMPdg1boE%NRV$~nh_}K^{Xq%_DKYkyoPNN z3MeL*BH2`Ampsn-is*`{YI2NHiw6dlie7#txgL5w6daQkptBvvDa|do_GO-mbU}xdE6Npi{*tt0vni0%?mca@2Qd=85vHP8*xp`VSQIF;&siuBNf-@FEfCYrjTC|vju z^f?n1wRC|b9hktbdtwWn>Ql)@2q>W{;+?=6{(kALIa^hSKvevp`YbTW)j)dh%7~2~ zO`1n((+5QYBgSTJ{q}LP!hu6hpus!LSSC@- zCW^#Skx{~W@n3SNJCZ!=FH6fmu_#p0a~r6st%b zy3+g|nGqN#zxMU3WFxmVu>-qrf$($~jCN2A5D9Gq7|q=@xy2jyro{}N zEayT&vg?vNf=xkXs@R})qTRxE{9YMk_#A{CST|h&K>&MG3vq#zsjaXpq9#a-7rp;# zw~GTWeOcTtX}p8@(AUv`*q|KY7XKnITe@4uYWd79q5ho3ijCzmC(B1It3R4&t^(q)*kMupC=nI$ zcgs;ISG%A8DbSVfmqrQo;U6sY(PqM0)WBYN-(|d_LEOPHmZ9d}h7&j@esSp+37<9s%$+hlB z)ksd6P2)9fm2{~vN0CVvE-asxCP3UC4LhVB^wpxbc&YFjf43squVY%N5;-Ka*TcK% z@`YGwc)+=WufLSTOB1XPj}@;DN3KJ|&HembaUA4jg4=Vu*BKznOj?9(2Ax(4?~K$m z;f~Ky&6W)=nqoNmqt&J6?A%-uq8xY|RceB`=Op7djLy6d(U-Sve&@V$mrwj}%?6+J zq^%&xe185#x?9w#sTC{}8gzS8e6Uoum%rEh!3_NS6uC9Wmh+nFfLK}EYTlU*&pQrB z#hKGEbAgHj8z=C!40FE3Q%o#H)1sC*VQUZlRsX)Nr zK=0?{UIi&qVuKC`KuWd@8CYN!f*YPTkYGIF+)b}?vHKcjtSGS-1+Qw)2_+MLzGUSc zYgpMGIHJhH$_~_vfEuWkh7bof#F}W_t13jDzb+s%OmNBug=6DwTFgPuwQPm(a z%wvySuYp2HMRo2q}~YfhPqk(hjM0 zYAo8<`Rpd;g3@`tl4p{3MF!o^Z-Cf!sd7_Lmb#K?dt3{=6CwW#|HOwiL4$(OHgMIP zBE6vg^=tIrG^u*!DIwZq**Q8{Rz5AuMSGLiBI%#JTCrM@sMyR~DjfPUrb2?lGXbVc9ToS?=%(Afk9nU7yez&h27Xg!7ky6rK-{D`Ez0Lz zU@kCuLi^#!V!W7s;4jW>1CJ5CaKZ~Uk$jGN-NQ|*CZ+fkdmRnfIs-(nG>x)-zD+~rGA zkeHckX1_RU)KrKblTOI@t)&h3a(xXyYsFdMuS4TB9z~GcF!kKd*bPfV9k2b%?A-4g zSGcDZ27QY(a9iOz@D>S(Q-}He+bHH5MXrDjtw%Pf>>)J_E@cZvy$TdP%N1vdSBA8! zx@Z*0)7R1?4;5Cq)yb-02XxV=ozeD$cY$bqm3*CNt{68|Du$kv=!1meF^3->pl21uB}-La;!)@-6LU zrdQD`$2FBs1&VBp7PbCvcxKb>s#U_`Suc;cj}c_g(PquT5THi;`OiNt`i+|rWMyA^ zuYs&}VBa28%nyT23dL-p;Jrrn1-Awd&h7Eao>eZWrK7*Rf%nX}OWy7G0n$yai)al# z{^e_uUda~eI8Ieu$9uG~vBx!T6V(LGj6c5aq^bMJrl2A(NFnW!ZSug}0=?YDPWK|- z4KHAILehki&@NlKODq^Oj?l26N83EY2^tf>|C{I$q^XBZzRyC6$)m_FDzeQr$*ni+ z*6hbnkDTkf+9QEfd-q24hBaw=74gh!MH$s6%Hb#R)4(B-HNAU!{q$UMi99x-%{5DX zft+_e3RzD1(6&50T2RLHhS_Po$#GmpvB+h4XYMVp5}|**DAx8Wodg> z{Zw*{ATL8n7-stpIw3dI4X+}vB3_(xvj|Jd2A%SRJD3VSou^7F0%aKWR-87u&;$fIVSX2*ta#0^jLQYkcLaAptvNL? zHE-2p&fMK*@T65uvxlPaL^A}YyApETEix%EA;Y$ zK_{G4(gf${Cq*QCULf_n^H6MrG630hr;D~1G+7UVpTpj363b9E%F5c#b#&>Cljf<6 zrbBb;Cx=P;8?)-DG%++q6azVQc~m4O)TRQF@dfEws5UEbi4)wOG7td-?-<3659?L5 zx#xhm`-+fW2>V_nS9q(ye+KMPDAZan>vyjR(y7v?XTxP}hoY9`Lo7JgXSvJ{g(p^E z#uP|4woPw+ucXbKdytD0-+^rlD^nqA#^3QU)HF6rd&prmR^v8u2>aV=5G+K6t$v-w z{cQW??frbCC2{g|iY76m(X=@5*aO|uu!Tt?#cZU=1}ZYm`6fBxdkF|galgCPU3<;< ztQ$I9afxE@X=B9Ow{Y~~`ybKNt{2!>CtvxXAjsTWIBfm^%lB_288l>2v)gsoE-t9Y*{A$h&TD4%rmQ55|}XrbvOVC$g^net&nBMYuHmp0F9da5J#|@h*>;JOB8fHodo^@E5DX|C?ZA`NiW}HomrNSzD1<9i02qq%p zv9TKZCjF^ouXnCE8+@zv&bdBqt{bRi=_(-f(jsXtGNht|7NzpE=u5*|RX9xK*UvyQ zhE%!-bRu@zh6>ddAA!jiVWU~>O5@pIOg64eZvL|6Z)CLtyGOE3Rv}v{2Ik65RAf4d z7d-MS^wF2U!(h{WBJH4mn)ov*?nf|VdKca35icIU*@)w4+svfBsv~YDPyWlV{>6NC z!nz2!H3vj>urfP7tOB%g_VZg@_Ic+DpNyyQq~V`$@ymbJ+p_U}PA}hat@6W<=jZww zEynM6*K8m=U$ebTwTTrep%@_PDWD>^Fv}U74jO^VOHZ4fEIk1mTg(l>5X>HzF15BO zyp3eiCGu{%6Evc$Qj|{7jfJOHpq1h-y3hCm_&+fq#|qHy#_*Mc@iWkXrRx& zeUx{DS0&G*JLp~XijZ5gFD)F8G{#}W%LO#Un%VtOscv}3oaBoOG#uAd5*GR|B&Y)d zFrZigbr&Qn!9p3V>C$G>2d1~t8$1eT4#?xf8&%pOmm-%zry(5|`@2-QfW*X@*cm3m zs|7P=4tdbT>is0sf!C=YnOKB<6ayL!yQ#=Wq={+Ll+k(N_d!+guxmdb3A`FL$$pRM zec~5a&TD`pUwmuMK5>uYs-&MEFMt{g)v`dOGzRZ{c#b-0RvA=rHEFu!t>jW@hcp%1 z+j@07W+a%v0#0m<4E8x@6kiIpaY0*$xp8q=4;zcQxDwDq_Ve21Is9|VK}k!{NuWfk zC*=^5>2xaunzy}Pn}V7&*ClnLcuyp)PWC)YZU-0p+Zqw04CL5AGD=4_zQheA-#HQe z%$z}x%M9eW&LP6WH<&E#S40bwWhh=)zp$VG^mDDEPDPI_n?9)?bV>$=ODrJmS7>!U zTa~DB*dT%hZkG#a$s{VW=|XvfN1f|(9)$i96?mXiZJ)kewvHbyTs2cCjpbp`csj(y{<4{3Mc^-8pfPV)nbxlfV1ROAWL17o^l+Ce#xpFkdBvNWHs-R+4* z;}t=v^l8;aPs~he@-Cm2AvrOxf$k$ZpW<0}W!1cP1vXCb!L2kBGVk$jROJhh%vPs5 zsP5(O>J?eGPN@QBk)HevSm1zXGuOh=|uJMd>QwEdJ6PMRSkFK@ljw#%vU-tZ^Ah;E zkdR$K+6gX$qJ_vSYISNHXAE($Brn6rtM0L5=v(5yy!SaHhMK=}D~s&qmJ{c|=mI{K zVL2`}6a$Xp3Mw*25KjQ4zG}t>sg;d!)JOJvgc0!NIxctS5SN zD(I)510>DWW)MU5Dn1)WF7mV}?Og9sNwS=udUVsfA?dP)JoSk8glSFZgVR)&UNJ;Z z%2GpmT)V8R2(c%?tl8QBQ~NLJFe8k)q?V&OQOrJkVC%OL4i2EfN{V2|D@Z)7lldM8aOWz^v^G8FVuATNYj|HT11brSqkE{-1_!nep;6HXilH&3g#3Ci`SCn!}Aas(0;23om3zj9`)f?-v@$ z1_$1e6_}tnm14G0WGm3!0_{PfqDPV#lFY;4BghTW)!sUlUTKwXRK*GI!@|XIek-&N z_}A@-*I}R<19XSAw%;Kq=uY&QTVHF=KKNQ~o>({rkz@UGU^A1$zY|=|t5Tl=wjRUz zJ+FIm9sdjnjYJ8PX2H%X0?6~>?8VbMRf=c4U_jBN*(uTS+guHIAaEsJ?HeV0HaCl| z0%^rMSuNeFh!UP4v0|NV5~ge#HBX^{ysZ89!rw-Yw(m|>iD9{pIhg6weMX$O|7!xm4)6a&@C`N%POAEE)M z!)B-$LyZmO#zDcuX6bgI(HL|(q$(7ZgdKwAX(l{{YDFbsrOI;z1VSLj@ZmKHip^_t zR09E}N=y@LkyyQZ9&P+Af&u$#^!c69`nKj8yU`eB5jFA2-L%8zlB=)POvEC?y3u7Z z{I5x~Qt+5;;KMxaRSST*7m!=4wkteMNBqN*BuEA z2X?Zsr0}+qMwL#5RK(N`92;14>X#{_foDh1;}qy1t`Ibp-rISw!W zGe#V-V?((4pJ{V!IO=8$1nh^;G|yAHpXcL8&8-lfhnD!9Cs1Kw9yzE+!D1-uy$jg^WpoBOHgm;E5vNEc<^yD?bHG8F zNn@Z-+XUZite}g|RA+kGj9PpI;-6pOI23l9%b))T84oxOh2z@hf0=dRpn18w!`cs) z0`^_f0*~^b{rrcr7e|en<+3Q@IzdWErzU|~_s+w2^S`k6|2wZi%38yn9zlLsjw-?J z+3aK9u&EW4gzcDC57TJ=l){ufpu8Su(*aH zc@oTBs`#+$?ucdvBzViG>C`>+74WcS(n(Wvs>2?AVY?%GynqJ*S11^zX_j7gZv%h* z4KJNf9n_Cuc+rltlC@y50WL=Ef)iLKs@B$g@{VyU{FI;mV^ZP3Tj47v3zV}IbBZFz zsmPOZGzrfH{rp-zM|!*WGhdx*!~FX20jH1oG0K9e82x(&8M|fh_Zijo@MUhNRfD0a zz7-3z_{eUU;?p-}i?ZBhr7}$tCG1jf7MIaUZqEc)0+KzEd%YvHNt5bw+a+nr_c@dnLOgZiS4=*x@xqSixb4 z0O$uHwV?;rXM;{j?sasfAaU|43jg3*J6f0newbCjemo`^y=*jX1+ka?j+Tfb1JGGqh~I$=kKJMm#TzW56u2&v~7 z%cd)WPzwNiz(31oT>^UDM8!=gh0&>MXTQ|MS%fl{1&+0sUxq^2NoAmUBCo^FQY=Zk z+4MH&rBgv_szsvRG9{D7HOY|~DUhIf&w0y~9=~1*xRbVlR6?@!rDOE6_r&oY=i%nq z|5zXAG^b4H|C-+hdMKv&9iLMq+JQGxX(mQxBgJf>$U0PRUdJyKt#c_C6#L(xu_E@k zd-T`9oPs3H;6Co3eHVK9S{IC%e30yFcECc{@cOS#dVr_}$-|0xOO`CbWGBqLJ@xsX zfbkjJ7lUK*Fu!OrE=Qcq#)Ou0tFD{7Y8`ez!{U53eDL+MHrMi?@}ON(EHwi!3EGAW zK^@Ye%TSIMbTAZ42($*Fx;EFCfJ_?G=d@9C4NV(FJNc<}6QjktX*;OCSPYt1u1#!< zB~GB3Sbp_GvAK2wm$}A){rfCR4=4kVWjzhEw)$(AksUOqSB$sC!6CN@K(@t(MZKL@ zdOA%X^2h#h$~uzjz`LAslU>ekipi%)4i#A%Uf@wOIb%Trf7RQD<9?8x9R2EWIn+(l zkC0qnL%SDZsMyv%2(btQFE=0-jVafMI2jRYY0cI{Up6A<*w3$gLTcU^e{Guyz8Wb8 zY7fs+k-jpT@by{s9kMj-a>vPW$oc8C>Y zAM2Et0~%#V0xky>(G5XM1?{{9pBL>pBh>~Vd^tpoce#S_M~l6GVqQ$oWo~j{4hEJo z`}^O@{(jqg2fwi_;<&0^Ur1RckC}zR#3oIjyv_BbTBk;t_)PT^I@kBO>bhhXeNA$2 zZY;0dr9!m9=OJAKX%Y9Po)Gjwo_8|u9z0bcYSOIsc%&~cI^M<+FEUR>E6x$-OLAHiQe4I{+ zxTLVf_`c-5kA3K0(y0w@aht;=zuYuOXave1OD2Cn*1a)vSgr}Iwo%Mhw2YvCh;^*} zijS1dOph#Csspw=Ef#Sj5fQAA@f7m$5*1Ac}r7k``c?I{LX zk3px~qGb0(#R+G)UqdR%P@h^ux;}crFgDu;JK-)m7plXo9iK70d5dO!Ycps$ea4RK z`ux$mzxvh$qkU2D-2VZ2_Qqtnq?p*3^%S#)A}gs#WYn!Bxx!pwRtTt}XtmZN)Pk>c zz6s)nxx#+84%j7sCPx~=Trm_TLPeBbFG;TygFamY@wPM3_ko`vN<+(E%uQ9>L+J%wHf5GS{5yhwHe%?6A{PPahq!g{ZP^y zh7xI6kgAA!X!%sG_!LQ>ULprJH^@orhAW0YJ&-h5>RLr_3n+xD9~Ay*aA{W^@a}@= zy1^FwPkz|gKlibX!Z7px^258l|pCcf@j{kXj?;;os76=;_i z=y$ir{A0zqR2}-MFX_lp_qpM&tR!q>U^d+$&Gkr@YI|e{{gyg+gsx}0>5?$(cU&yc zu^z@O9b0*s8;sY^=T9}ycpABaap3SFODUtl!XalfY|I;hq^Vw(7Pcw85~_b-L;r+E zum5g-i>O|f&x;av17XY786azf`j9(8jiiJw4~iD1iS5G7VjN(r0c0gm*qXBg=-q$+ zv((c#Pc^@N@sA|ojoA!=F3B*TZU)7Gyqk`SeE8)x@AfP-{QZZWM5jLFy9?q$4wS<~CO?J_?KZc(d`K)A_m}rzpS$da<7=Ypw0TDM8?p;x$=^N+ zOoL6D?bQ8+<%@Kx-STvPx_iFBz!8|ty9jrYO6;)csYIvR`0v#VchFk{Ht_L&oB-Rt zxLfzHLwB1ryF!l$`aM<(>)fgZwzV#nVCv=T?&o)H3a*wi-T`xm{zzNmz+N`i=))cu z?Tc`;7CBbBA*}2tM<+lwkUSI)udS__!Nj4a+Mu^O_~ge&K(L+ z!uuhMp(bRyyp;P zJFvi%r3${Y>K}{gRPXzRQbx#%!a~)`c_`la|70m-873?kL>oUc`yHqMF)vwtttKs& zGS*I*vU=d#MnyE_FWv9ElSmvqNI zA-x}V`|Ff(LVD)@>?|2@a#&v?0kZXB+u~S?Sxb>sRODIpQ%SRQt-H2^Zlg2Qx8~Sn zaq*JZIr=RN27mO1cD~9!_@jA}QwqPn`wK=u6wKh=Cshud`*zg?5OoxDh9W1Sh8Z$v zbgEK$!t`8mHk}3x1loF8799uEdpG^pr9zY>!e{P60@Dd$FaN$o+cB+=taU%CzAHOL z3PJ~+GSwwviHhC)?GSm?qJT`jES|q5sFmdKQvC*xGh1&|stCRY9Aia<_~n(+6NKdkaGLMngo2c;x! zG<_=v-Zg=o)iArWhhiX5ol8ZQC<`QYpmmqUKknK+8x$XSDL$$6V{)9Xh21tD?G_l0 zGvK&N+A-~%Fx#a`bJ!K`01-Ey9fOSU#{^q<&MWb+v(3K9d^l|ZBlA9Oan25-cYO+e z^hG0N7M8hok!o)HBnJ)=Tr&YvJ;j`*$SEokmCc%^-HJ<6?M3>r>>_w@PlM3nBT<5Q zr#mikb*gM$YY@;NqNJ5J!Kc_Oj@cB{O<$O@9ij?*<}aJAErAXD9YGZh^c_e+Rw>Bh z-w&nVIG#N-UK?nQAojK(hpitd+d-X+TxV+W-`A%?jMI zUSigGH_4Hxw2gl8+@Qbtb0zX57`s-lYVY$J&Zgg@f0YehwMD;&>qG1x_9#mZ$2I)F zJ&@B&Y0VeEEi%qbt5dH3 zf~9yo~s0{Aua0$yNv6KkYXe@f?cDqR37vG716# zf5kXJ7C%{vv7d{)I$5^M=680yt5a3bvAkoUO>pWReVTU$)tUKYY zOTRM0BH%Y~y+t}5c#2s*ak%f~5ykXU3 zK-G~I(oXQHHX45qcCw@qm{4y7*V46+Hry3#*v!T9niaUYHgpVr!V95wvbCT#yHtq% zXm>$9-8oUaYK11AcUIpu`unhj7XM37r#4Gn;9SmKg;c|?F_mtfTIm&| zH17K>htlwQ%AUstEOIiG6Bezl)CL-Xck`Qfib>9BjywnUy+GFbFh^b)#gtN{go-@j znLo3azn(!)Z7d{U8aTxCV)K~3+eL4hd;}Q1wXFcSl^_qkihu{EUX~5J|Lj>1){emc zSCYd4z?oZ1Bk2Nqml40o(o>|xzlW@OD}QFL4cXjnjwc(LCU!i%`~IK(%-P3SC`fzg zruVKwaIZ^;A<3%lH8f6uR#gXRA1(*H8wEOE)i zbKIpHVcA+Wst!nUK_%p#G=Vfr&*{Cq{qBvbq3a!Vj38IoBirPGGujVp(T(IJ}ZbmlIy;Gv|N?c=rs++3cZM#>Lup#Kq zTjyo>=u>2aPn+vzpDoG@<7yQ-VZ)ACFz@>386oeP+YyIlF=Vj}4el3x5~gRVO<097W$)n3;vjgv*_L(eLvnd7yeAB7ON93BM zaN!L_K4k001}){SRm2I71Xc^~xulApPRk2#(sa6%h}X=?3*SGxL%M}J_qCJ%68-Lp zpYEb}QP+PM{e!kI*>^2v1yXE=iqSgahLqsUUuM1T+^AT)*dgC3$tCTIEPgJ3Det7Z zNrMDE?OxG-ZLWJIRdk=U&#g&QV8|WMrin%%a9T-7%CI~u0F`yEB zn2K!Y9SHz#QpFlMdVpKW!H|#CA4|6~%L5^SX&(?k;N}ObJM7CFx@(4CU?8f{l_EM)6+)7jqH^!>1I*`O0vR3oQqQpi64j)?53T67d|nzDvu zyY2PGuB=0!HSjXQ+Q<{D(3cgec-K5)jZP&ABTkn3zJVGzs^6JWb&^`8Z|lJ ztONfrX5bwm)dIZJ<|Hy27v;%3E%RnQr;Vw6x}=GrCC1wszUtxa}Mr4I*G z^EP?hkYm#<`2E?C3f|@l3#2gQ?;L~48XKTY&}G4iMHn)9h!dbD{OR%!{yfvTqHz+d z_mfNq_OyItvPRiQF~GsGn~EF)R6m`upjwS;A@#srrNxZuA&C$y7myx|ypv5Q_*4?4zJiNIE=4ZdCZM$&13pX!W!Q;raWkxAD-Sv0 zX5u&KLmSQ0qTUcsKTCF3GVgxiKA(rdI#qp8`Ly*+qpC@R{QQZEJR#n}2*5*-WlRlx z5LRJ#CYPbX=8UGHhIanxlR9hXTSmL$`=8~D$rf%qB?orCKt95-1chA`lSz>ssQugx z*uXtahQ2aE&p~$no!+pr`6vld&BNUdvP8jzgt;gA(yPAZDy_k2=|v-g?uX^taq-L< zDyGwu-2n46x;ltIGzIxNMUxl@o`Qh5f7swBQp`q*Y@i~K23!f)7>2^~NW7lRWPmF- zn|=i8QkO*aV_98T^|mb@X{&w<_r{#27;1%Z`NdZ|)H?14L#z=rZ^_ z7dFmViwfSjb2QYBZSuJ0qs^qt=C>=-r@`vBUGdm2W0n?W&q|fZ683*t+aLV&KmYLC zND0k25sFfV9Q*pKOIID(^~O@VdR|uNy2}q4R$`TpL0-sk-sy(Q>?q``O;qgi!_)nW zZBPsNIJ}C!!b_j7g>3Wbn|TjrRP(m-+7-AMF$fJUm(|;cbi-nEV>+RjopCoOuWG?= zZluM(_>R~(<)phNwUXF3W_1EGX~QNc9mN2N|CXUh0Ip7&r7<&4O7*J~OkIi==HN~z zoop1`d27{7Z8iwZ7kTZ4^d}qcd00nkOweKN&2~3*$v>H6o(VZJLg>KhN-Sxc^R725&JZGj&)GEx1O8*@5Tu zq?ukj^hA`}I_I4djLy_KYcXXS8Eq?q8o||3!S6yU4wz6D-dva>>{o0G0 zJjSiK&2-Q$!H;Q+@j3g~ul;}stZEl(%x>8W8#9Q|K9$&m2OWy3zhiLryqOl#1IDMIpyH&gsSsNuau17`pMZY6) z4(tLcG?`h_C5hfLZr zg`tux=@)-#jl!B1Em?uO(g+_I$7se$@%POP|Iww_4~ zfvQtTWU8F3^Md_Kw9r5-`HyW0zrTkZ<^~uCHYPxeKFoP?nqp2+S@r!lAv;(9U4 z%Y0-HMRGYPJ6{%D)M)=Iy`t9Rn z1-FH<15Zm(+CFSyoJ293C=y3SCIIhpsj}bg9F1iD+njGtf&F%-8uJUZ#eT?UU9?zG`<(=AT!#V2EaQZYIcgq=j>&KhijL<0i(t8bLEjOFt zxc@SeV}kG$irGQ|m?9IzW%J8j_Vc46%IFN({_LG`2BMyZJ=1=EZ&-#J-E#KY0F8;@=`Ybm14SN!nbusgh({)pfFGdPTyhz$foslXQf(xmHOB zD3=Z*ia+eb_WvL@^MB&Z8pRT%&|2>0alW5>g>L9zNi7&*eDphF3o*5LSp%{VGJ~ znJW1@&xatojEX{_GDzA&i+OD>H-JO3LewgHCc(mXlzizUH(^nnKK*ilJ?7Q1B5WAg z?BdAnYy0wzP$OnU&I@;vJ=`$kz>8H7Z5f7{DvALmk%LrZJ#1qdRdIrDTJMQfJqvyq zd{>55pqc6l`uzM(7w!4^kAN!Usaup#pC=nF>}*z?JMXGn9cUeb$pC4VbvFwIsu z;Dj)i*8qJqX;OT8Xp~FDKzEZSF+|62bIn+AZFWXTxm%h9=g_g5D?8%J0IV^h>iz$c zf5ki!>9COime^wZTy&tc(5<3GMUhuUP>tl)?33y){|8|W-m#kUpfal8?K37_+{~m2 zD#=#=)ETMb#(A+3XTZOrQ{ms;-m$!G{s#iesEwa*mydOfUoiyML*Ny^V2ro9t$Pk1 zKOHsAXn@q~S67nlZ;SytWMY5{C?=NzD$+<4d`AV461^1oIb}P89=kii$HTKgTQh^s z0*a06l7s3B*DU?jRKH77!%+puCB}GM<{O>{PIc!Zdx4AVcIx?T^!#9q=y_GMwc!_% zrW*0`jl7@#ifkK=?#qE!HU~{mkxwzO$IgNxA38yuqUf1{k)TX^b-1>|e+PpL9c`4b zTDcwM8gNp|gsG)Rrt?`J(iL1u&H&c|gp|`llEL4Efg*ilb(?!V-#+Yd)|(u=tb-@8 zo1DcP_Z|7iO6aeWFK3`k9d%ieDIeW3cA9`*{;67mm6!KmpS!=#DLM4Ia~WfH(`+Xkj{I?JawObbCE|5!2Ry%-d8 z?PDfj;4B&a=TpUK2lg5KmspC?%7JEoILp(V%b0~SxNd&Ce4XHrrQyqF6$hqC{$x0B z)STBODlYPRXxt(U-F))-J?d(86$EV2rkBp&B8?YxxIPou^E(A_u*P;gWVduhA6AsS8Xk^o{6MUI zRCw%Kgd#&}vvc_WKYL#S*JOU??>qR0_!Q>r4kqB})au_OL(V6MA+wHO2W4F88 z?#_Q&X*+JG?X=B0otd_X2jT&!fPxx8E;$6bR8%fS(Q#B@Q1KQd;xRY~iiisT=SiZH zNHi}=*y#2%pMm!d=J$P{-}8Hp@3&4;Df*nV$LF%!CGlq8PWi{G()s@rO~KE2kiKlB zkCPk1c+R*0ggrdQ2Oy^5@#TMi{(-CBdi+WKsci}0dxxQxK}&%@>2coxa;~t zE;!`_Gj=}EiSAOIny%p|cwL z%nD$fUXcDIsb4xEdYuY;9&X@&g(CI zY|OFqid65LVBfjzluA+ww8tZ1@lx+ZeSOpaMsUqcP?#P)8=YD4_3I*IGDtR{u;V&7 z6V#C%POEt*-48+74aq_8OO{3K_r=ck4yO*_ySBzm!Dgt;#Kriyvm^fSsDPfEbM&)H zy8OQiNTnV7uU{C<69*|~KSgRm5$uhO;AFYxodjT50`CDx_!D+wmiimdMn4`63BjY*l#dA-M7;QG2tUTNa zZ+Ht;#`eX;i-~GE8rg=h^L_)9F_{l~ITtziB}fup;561}E zKSurPgQ(yB>D~7i{87A|VwO-OdYr;1rvJGKq3_iERBvB=GTZw}8#}Mnj^~|z18a1R zVl))#c+3cibdT{+wN%9Ig%0?&Vf%b)r`_@aQv)ImZA^RMrr`SN>q6Iss){G~L9lkC zTtMw<#DMc!vT}~;7t#DT=DB8R>?cj2Ewy^X0%QH3XSpTqIGArj6R6lL z#lK0oR9XlnZV=K<;v{j}NhfEM$8DwQK{D|jS$x>+!MFI6*ge3sx%^Yv)!oKzV-Sb9jmB=Ihn28)h8# z?cn3d9C|Hjgty$59=U#DgFxFYu&Ka{e%W8N=UTiLeM&2_X+j;AoSHA2*xW^QCX7_>~)Lq&2SrxKJ zQ0`p9y&Z52dOVf}V3@WfJYh;-=#WD;v@qlEt@L_PUuZeq%PWGtCTSfHkA5jH4a48P zyg2xc=WJk7%!R>dhUDon>jQ(8C856kOKy!NoCtR8r0g{CdY`DYpx%>OF(ljVcutZg z8IY(@c?`KRS`|&mjRDQJHT*v}6Sszf3 zpZLSeADj%-FOcGrdv}oSugn7Jr~y=}DF(`GcTrLKoD5!Wuv)2Gx~Q&#s^I{+6Pg>W zs^PW*SCiVSNupi(Af-wzEJQM=b?`Cmm@Yzm4LaurS7U8MvU~@pU7P|5rXIS(W!RxB z;DY$cqcCrDC$!0Rna?G$RlH=&@blC(8E!pB!PtL)YkkXQqQ_5}L&h-@YsUdRP;?qm zsxs{^@6gcZ0?OTx=lq)I?YxLdH*AqWq7OFMIvU=fVeMpX)}lLDGO6OyGDx zyi5=m))txWRy(OHpbc_B72+PZ0@u$#?k6Ysq8wM#Mb5F#wjcPhjl1o`wyaDEwWaU1 z75}%InakWX5DYy7>7#uLnXt&CRBN(fhn;bU*$s_uPcy)nb(78NRHx4$GeYEs@X7}xxe>JNY+kV%X zNA~_>)_3)8?e*F9Uy&m(SygtIfvbClVop)y3n~hi#5r__@E}xUCwZZwcC}Nse}fow zk{iKf^vH|B_0{$lCd)s*;nwX_4xK2i?$z!a8CCzRs_;+I&e!>%dV-1Z zc+L{C6;z9A0zsr`tw(jp_Bq8u^VUt4z6d7m&GSx}@jjS-5i+LaJ$l(Q0)M?CKcTIP z>u@^FYnRsX2gn8gKEGD~A!#EgSDDEzpSM`h$SLODBZ#U_sN!>G>wW;G6IxXwY*lTw?dyq{dGMld&#U&lYtf5PcgvvvrpF>qRk4Y z`&{D`aaOsM`IY$%$qGd1yy)6PuqmV)2$i#3(3#2e&rvqfSAo$*tyqW`)Xv!)tX6c% zwn~uPC0ek5YCLD*4dHEce{b*NvD|2 z6iKC`mIwF2QVC>-P=6O3nXuF7?{ommDK^9x2r|H`s0M=CgPyYSo9BjDI7(~~^3;l0 z_@QRpe!M314P(w*6J)WvczUwDSvm-XKv4Zi-wbGy#f$Dtt?-0(Y3-ab$r_mPH;#S@ zn{RaVC)nnwX+u4V2VlP`{kOmW^LKuzHy!d-KL1EgurnQYJgamYn2t*nbDkn+kYDZ- ztaVGLH6ghWG3f|OB&rMu7359nrkm#>t44|JVCY_D_2U)PdD02C2Jbmav=NxcyE}oJ zy(}DJFnBIoSt5f6Ky#b!=0!Qm4qyV_;{z4Kx+`$4RV2|1I;VrM4KAMaVWDj0T7s?- zl+yL`w|j1KT9^xy9rj)~$ks4$e<;?qK9^FM&5QEX^jK@l^h-8-!Hg@Gw9VV^yJLzv z7xiMkg7r4Kv?`1xmAU~(2dCSm4gm4FB5g8OQ)*4=xXek~p<#^<(*tVNGh06Ybc1o; z(L{PL4QgJuz&EKA*~p>sjZk4vdIP~)RBX!lyEh1f^NP$&-%f7%;s34mPPuTG^5*L+ zW~lbiP;(%;>a-;+H+VpTMG45VfTD}zXE5+BXJs_3Y8IaVF#p+8{$R}bWw*7sCPcYg z6*+V*sCpy^T$Mvs2OaBiwO)_EVROd#@wMamo1GP#?A!FW1$sAjUTs7$X&)`ZX2-6? zuz|CBn__NJC!}^X6J^KMQWp$N)Cr3DoFYf4s9nHy`|%af zT16&ojjWzm58N-i0{TKT1aS=jSC!t6`8j_)I zS`J+f)j93*UA}9X7?*s`#|u8qqj$;5>8>$w=30uf7aA4a1*{Qh8f828yy^0PYh|4# zIU>kb>T=wj(7%-LjWRX*OgHvs_dqpF$>e7KIQq$Kal;w_W4s~r$LnT=>Fv#iJXr?W zX~z}4tp>K`5XBszNF5c`iPa3;4KoKLib6qt2J(T?g1uo!$@N)kMKK&y)rr+k{nGx#Vo2VWtVe%f<)_ z-EF2#$|w`pYyT2hQ)@{9upNhpO%woQYe5J7D7XW3$4eZKDf-EB=UZ=pnaSpsIBpSL zaZ6L|n|4H5;+RD>D3*vi!@d&shaPuK7g+`ubH;RxkTU1d7G0CCxh-~WU} z+p)D#8(5n}idjREl}I**bqlyk!G?iWD5D(bhJnW;V9p>vzIHr+v$GrDxRKxXEj=U- z{o?$Wq+v9N*N&}2hXG7lDW;hs$Em0b!aZR*w2oAPljk2j8xuE!&Zy!VAF)=nRXH4z z?wvOIQlJVI;?=PBMzWU-&dq>LK$=kPv`o+nY1ND34$>e-HTq$P=zx8*b&V^j(}#m{ zln)@Alq^pQuL-Q@?UD>T#5yhIRXd@K7JfiHCtkD$^4q(#u1~%=o~KK4r^B+Jwcle7 z)L2YQ&$*NxP)$SaBfsCTbkW1==NHHVvi2o|Rfz$tGARbk(IzTtr~6W=+696-O*9s3 z>N1W6B1}#0(DJ)q<+c$Zx!LKS1ReUDg4+XQphWq(DM5B$xyh@?h!K;|v(^dI7zuRw z_3uS`jAZgt+DW_}&qhTCcu-T!CW>st*(lGyM0PJMk=HKYszgEP0>Nq|;{ICmv@>$m zLBAptJSz^t7J+e04U7d0)*A5VJ;yW{M)7GUcnTukeA)YzkagjvYXQ)?KrUIWL~)p2 z>3Y{aKFRW7-|DF=oECQQkvmAk?~-AIbgXMPKy>JvSDVXy=YmpJu&- ztWfgWa@TzvV_nymiVCRAz_esjOeRItXj=9GNU$1Z$RU%g)A|qV13Q9L<)P;!`=<^& zoRjp19aHo;Vu@}V2U+W|eHM9x{tlg}_kQzk)JN}q@cZ9HiD|}xP}IIRH~!qX5r9o) zt{tbPOb`iaAemG>wH(^*RTsrW&{fqXOLT0W+yFiDC@Qb(x5u3I4(@qqk!@9|szYl1 z`ham458k(VB74l{2W^n`tj#k|dM`5TwlTe--KLC8@Gjy!GvJI3NGe#PclQoEM z@^WBJg9Y2R4}lnbtLBfi!EH9*VKvcAxciF*izn&*)oZ`L@D~zi$Np-b!K9K(F@WIp zR8-N7DwiZu@7WiD&czDUMj3+$I$A)d+dSkG$!UnZC zb!w%VrQ4VcxiT61_Vjfn*lmo{JQT)<9~T!0#`D`}RX$|=71?y3)%S0|lk>m5d+NGS zZ#3dIUH%nW!>)AOj#JHr1}0_;#iUVW0~M7VTp@0qv(2R{yuv%0S~T$HMRAubm)jTG z7I_1*&_f{S9M3sL8tGm0+o7OgoY2n7%#6n(NB{CPl>O{S#zdxU+)_KnhY5;P2-_>u zCbubSr*+Q)5eKL#SU$TgQX@<7>Yg=DV0db{%!S3%JH}^gb?Q6$#-R@riazafj4s@l zb<6kolt8mW91x)+BQR8L2unn3L`S^O2-_m>@K5kBi;JLGOfyxjhz=<9P*1V_w{7m% zt9{{~{;>IX%xZG6-&gE1SN8tPA3?31qsrx8C*=^n06u?=W9;=kE*|!UHXfwoFO}sY z)rw|nJ?D{ote{*;B4Z z=BO1noKJ>hm?K)C!fMt$deEhV-^|z0eG#WzPr1gEz+)I4z3HNUQ52M zR;&R1fo(45IdLF!mJNdMP;tB3Dc=o$mdH+q56V@zQtp5Nf0EZZ$<;Z?>~48rZHyN2 z3BpWnSO`7S7SZYb_q~isabFP+rOv>EEu$F7`V~=8o4oG3RXde%KNt2$`a=`A1vAo^ zCg>R)Kx7`pg3CR>aKl2kW#V1VZIK5hP11Z$jG#0O1*rxkFvAL}r=15Hx)Sbo zr(=AzqFPZEZUeJr9XMHtol$Q8r$d}dOGt6;cg`g{4M6jR6gQ1?U|unu?u(#ki_D*X z&U?RKl2;GWz~F26#UY10PfOM?jdYsNK7P9Iac;f)5V&l>XG&`5QE8mdzJDH;!il^yztRe` za?XJ6;T=f|_tMNxSq5;!Rf=!OJ7uSRGeuZjiT38cw3NOm*$pI@nyCXMOIgD&_TB7j zJHS|i7MP=~4-@?pUk{r1pD%k5wh-!AJ7?dUeKm5wOW)kXVx2gBbigLB6sSbP(OwyO zSZuqoeRgnIjrOy*+w6#GuVj>Z0m_H^R)jHgrQIfOjvLs9I*O^L$Q~*R)y;CmDcova zs;{bASXyyE2`>7x3 zJ=Y9s@H^x15zZ?GIfO zwkcvEhIr8CVi52sN$`4d0-9pM7#DGn47eVa-NElJ`{(bd)wc-FbldIaRjOSou7}ECSD0AGdkxK7&`3|nm>VX{W zqCYR?TyRq1#rPC_z+Yd0i|3CHp0yQnUi`_=zv8b=RlAG@Q# z{qu**>2eNanN+^c{A*u4o>QuY0QL-Gp6ae|G_YaDNkT{B!anr2~TY<`JF za4~IdcF$}GF^=on?G$7pQoESCA}`ibug8#bVFQRa$2wLs#a!eI?50n+-=CQ}4RvVL zit{sBL(Mb=PJn?AhoDy8=3+^L+t+smsNBftX`Gumv9_@CU z9b@%U+HZVf<3{G8ap9AR4ozG$s1>?~mjVGa=61?KkW!K9c2hp=fP2YY(Ah>gl^Wp{ zAP+uii!}YS-zCWuOkfsViu zDry^@&p8FMsQo~)Rpz$EZK?O>fXnnXvY1m3r7Hb3^R7ZYBP;ktmCPjd?_oK`JUnQcN%5=5s3NtFF9xabBEvTjUua zlSCS)CVr2bn!DSv4nk|yVOd|h35mwed8%Xc@Uqs(MAta)q-iDd7nRKW?a9a#|2}S8 zWFAmYb_+Y{IG=T-bq;FfwRt|2cEJPp@dvprbT-8H>Xi5)yJq&dn<4yfu{=X@TP;@=kenKC6@_v~oF@a%X| zx3nwrpr7s!yvjXE@8lnHJ;*uiyY>y;z4+;QTrA+yeKM+`pHTkNYJeRHUMp;WJ$3$y zpX(vGY)i~Ea>H3OdAaQUVSK2i+#`N3eNum2)Fob`UABCRcABb1n8j`O z=->l$2VWH{s3j;fh*zlSl<&`mdJqGyfpqNV(IZr{4(oJ!7U@z}-*gA4wlrZHG zDFH+D$Q`Cee!`R?hu(k%2~)PwcybrFg4`50(z!slT2A9Fx{GvY`^Be5oYfs$9D&!T zL%-8HlB8^;m$_%UuMpf64>@dbR&#Db30HL>p1B;R{}|A(^*P3W{2|qWPfi^od&6?Q zw$TRz;(Q+e6)&2%|JNXC0KjK=9_|0`5@QVkHf6+iT%>M73IXGcTG`muPSw0~5^aPL z7aWJlVcE9eCg_7}cvJPF;e7Us9P<%&$Qbj-u)gsZ_vieIuQxyQPj5ItR@(7|2Nj$n zCcHF?*+7vbDk=jSu7LkiQmMSgDWr#HFPnI4Ugj^>%)d47VtPl31Y znnn>SBw^KPzPHWc#miGk9J$K=kUEw!OX|gN_+llQt--nwTBHRp*<9{iy{?NRGmcETwfj< zEjUg#hDXb~BAfhnGTlBK!!<4%{sC?-9XqiwIM4r>BE#{KcL!f1N%7BgPM?lTtrdRR zb5opSeCj!8+|s8)Pm^E10M~I%{DbxMN*rH)uXA_kxsJ>hw<6qXe4; z3Pf7zdM#tF3MHYn$~24$Yv$HZ>J}mc^TLHMs;V^kO`CIKOelG5AI$lHFY zR-7B2J3_4S@!Y3b?<6aHP&Mm)zU8I&QHq9tvXf-k@zS@(!0XsSF(CU~Kt-i^#D#sX z=ng9)Pq*44PkDSP-{^iepjn|i*MfaXk&U8_?ni}{;tE+m=Q`Qvvsl>{+2DRn+QHwz z(Or$tsIX1%8*9isHYX!~TMsBtpTrI_u2+J8_%l6ZPR=|#ft;~p$lNo4%yo*nN|7s6 z)KxA@gkqJQP9PL17_cULmCKNpniBH#IO800)JQTMn?d4tE$J2l>Gl>ktnR)i-{w*~ z?bAN_1!27;n{)*=OTP%qU}KMg%B>Qp+`_dnXa~0ougWt(F(A*s zXx|~1)?6vq}$lXc`_=-8|I+L_@43DulT{)B`WfEV7R8Q*)+2VWNuN9^$Ojx%? zuJKwDfCY$;&u{j6d>*dl<2BaDj>V|aLCWgbVTBrM&eG(dmt8hkPxznYrDJ(r8qqx- zlx5EjoH>Yi_Ac8WVTFj-P6p1tXdG%W5z*-7Z}$X`ZM{!-P@f#T#|Fsyz!u?#8Harf z#jT1M9}U!YT!ma*AGZK1J{n1?a=%EWyUY47q~RB7&R~pc)D0hF^*vUDVZSB8X-TVM z#roMs_P4iNNy2DNOLkmE0ITT{ndNkf*-VjCDhdk+G*By`lNuU!K(T4;OX{XC1P(aW z2!X~!fr3=nBzahTkT?2z#OkOqw`E`?qOJc&jPgX&aQ?6P(tk8Y2%A*$%ZHsSZg*y< zi*m%;K3OcOM5lJp<#KQxy*9Lo2Ce7-6|TEMY6LWkp#5!?%cbww+77~)jm;E}GT2r= zH1t2eb~SD{vD-YZNd}#Df_CL}bs*4C^@03hQK*K-!q~1Loq%FI z2h#voqyx^G$_AdQPNJE5F-!x|Hw|6tlS6mVnUFlf(@9?iEZR>~!y8wz#QOYeQ2>xhJ-%>~v4$G|?rp<8Nqdy2oJ>0y{GXbl9JD znvj;cvCdd)({9mWLRV8On+lvg$oPXot4$DrD|bEQ3mUDG0qB1!)iz=GbNV74@ow^t z@lQ&TXLUSKG6sx{OHU^&3qt+$Cl~*pQa^)z7WLQlq|AI`GX) z;=f#xw?*#tP4mg72j$CO&k~n{#b4%%A-62978;LB3 z*_c4|;#tUIW5f6X-}l|H0m(%{d1MFw5VwIF&&i&fCRi=d zNV1`(VUyRRpv$nxP7^GNXp0==#)B{YFrwQ7YUQR4Ikd}fbLwc6lFpv{keAOnMB3%i zf?Q>}8}3_o45yJ!Z zk^|gC9x@kZi26e@OV}yD6MDy*1+y<22R5TUE`Eys4>oNeMq6+@{$^6&pY-m^FB*S5 zkK7(Dm}|emXp&&C2w(gfGenUGI_|M0KI?sAgE5(0?Ud#dL(=IbK2_nd!EGQd-$O5n zz~r-5&==Zc3q+-KtLs|PN>^mu9CkR%^g(fEwbRYnJ2?6CZ@BfzintqQsPdJ$-2J>f zrnv-P(K44S#hrY z@>@Q-&lq$`<7ns{8VhXVMW=|}c3Ydq6UX_bZm7;}ik&N8#rLTXt71FB3P+W&gb>!yl;Gg#GBc-B2 zcq^(yGDRocQ@EY7yih3Z>)@kY-8$0CJ1PAN1alsIJ%w8_^Dy-PH-HIxeA`Jdhv-Fh z!D_!nVzJLO=9gi_dokLkh3~?Il+`Z373k->56dTgL{^VB*V%FMsldPu-b^v62<)f= z&?4Lug5`A8Qz6jT1RLq1J|JDf#6+^Zl2mv?oLSd*GH!VZ_8_oyc+I=tv_*9OTE+hu zS5dtpM#e8QjJ)pJuPs}VJ-@K}nyBwr-_9@WjjGM7x?|+$REG)ii|L(M}j4P__He+C-)C<=% z85~@X;Cd!mj*FKA+-_+e>}PP!lVtf9t~K;Ml0(-?x*?F;Eo>(m39gSU_9R~T{g2Tq zyzse}USN8b8O`s0Z+qLnyo%x8BFQbcMmmvXh&zG@ge`Oq==I%`XK|OiEamla_s!lJ zeq1~V(vLvr3s=VS5`?<5Lo!So49Vg-%UmA?#ZHV1%L8n+MP8BS(ea`cwrG7DBan>d z7)xRE?ER*J^zpxc$XGJ**zuak#Kd!ZQfc^p@lxJuUXoXtTPKqzKEcmW3_+f%6C@~N zC&n---l@t?#wGxHc4)LmSmMUSQ>=gZ{J~72u|nIk86-P)K1`^#X=E7tNs|o9QHW-f zV!!CT--a38G`5|gKG+`R_7LnX!+q2mZjzmmC&*Ft8ntY$%3}oi;%<6pIP|*($GYzI zEt|7`Qc>_j*WR!$CDz;&h;-8(lzl<~Zz+%sHCacofF)3j49n=d)_{#^1GC|`yT1E| zez{!b|4}R{wBty@DW zVby`vPA5dwK&M&4O%gTnt>)}l8!6U62|mL#QbzGC<&nwI_|`-tF9-_l^l@{5 z2m<+~wsGP)otzw~-b)fyh3hgBZJz7hhaGhPSqp2{hMKh~)>UszcQ9siV6&*UW8cOE z_d};Vc47vT;eJBYC0#KQEY2GmI#t4jgD68K0>Dgxf`1v=i@nJ1n^MdCzes5Uu zb^X-zv+FIJ$eve*ajMN=BKwSD8Y$91MLqh5`gidEa#_4zHp` zc%N*3$udWK3}914F|fqWrJ}C5t&Tu}zieeq;C)#-y;x8n(lJ$OYY>o|3dM4;J3yt; z_B3IRNv)`uKIjSpkkE>aTGvIC?!PhE1an4rED$m@66Smg`)f=?D7p8nx1$~|jbE$# zD_LsC7Gs5yK9x3JpMI$Bf%f zDo_a|>7sZ}y=Mc_3X!Ew$NUsFH>q&F-6Xrs>y+1rrMn#q$pbg2z-a}gp|h2>etSYx z=n<9Ayr*qv!zpZSXukKR;P$|hh;vZ>2h|d}!JSYOIOx*id4xaga31Ka%4MC3yH1M* zJKZ&Ot71r&0F{Ex9_e&0jqa4HRiUG^YUS;c)NpfybLgww7=nMZ26AH$ofE_3IdBoU zU1NaX2qfBW^Vqd-Ci^^|wccD;MY8R9*4l3{YgJGTBrl7pDAXV;4$<&2%scE*!%gri z^z9YbNs{HQu7$o3%#{vub-j(CbBzZu#+xPH30itew8 zSnFFSs{z5tj##?OYF3-@?bgaDYg`y*o-~_Ss67G9zFQ=_nh>T(Z{M& zeD3eD!< z$MQR6^Jt5%Wo6gxw+vR{b$!9>dT4lmP`!X`v|~SIrvYZSQ4B~JZKa|v$eZ{#glR6# z!X|!OGiQMm)#bo2)QQsg$3kyzDeH*gmvn8Wz>%bE0RrxwlI9Dk?o&Bg@)-xlxEpp`94t{mmGWYA!3fX{oNS4RRh4eEL zGcR-19dC<72$qMolVz@Lk+BnVl*@so^QwG54^!3g;Jai)z2_P~B(I6*SeBeL8(xM6 z$GklY;Aa|M**|s}o}-6Sy8OQiNF}?Vjs4EE6k9i$j=cx$PsgQ0tsYUN(|HD0ZWlnwG&>8TWZBNx24i+(&HY-J>sOmS`;hT9Na-?&($?9 zZuBj6Q>~wL!Kp(>m4@Kz05+srPL{U{3%%+&D2|ZKT@iuBU6w4WY>F@o+6(sG{Lk2z zJ!g=_N#f)|)N9Z=I-<~5g^^z@@l7P{K9&sG*xp;S2YPIeAM;(NP01)Zg9#tBeClZ| z*~n)3Y{%xsgutVYz2OO|14JM z^U&iPZT$NOf2fjJqV#IV4vq<>*K5*zP9JFsEOkS%3eBv;zQ-UC*9c6lu~Tljp|J4y zI>aoHVPh;9KHj@|_x_1`uOw3Qp^7YK=atxTP&VDbD_Ki1@f2C5i~LtRp%QSl6Lz0< z%G2rNuH%aa-58jK3B%zUmp#Vo^~w+Qn0RyhK@W1{mBGXcgY@Dc#ef{n9V)6#kw*jb zr2;gSGt-T4f(;pY=Z)~Y)C{tA^gzsfIu}b zx59J%BqY!*o6{fqg=;sxa%LN(3f24-f;CKs@E~U)GJvQmh4+Cbx=FU)XUGBlpCN}U z%y#BHXy?Rp;-rbZ8_J#1h0W6Q1X7E^`60T;sug{*l7NMsoc(@C`PnIdKwk@ZGz(a6 zn!(-;kUYAcQ^W0|bq_~f3>`=H!a9f%MF-$t0xtMvw^oJSCKvPG$$Qh-k1^o@Xu$Db zOY{8cRm7#}`mc9!0XLC@h9s>U+hhX-4XWlLU>6%o|(j(}DI#jK`C92JH6X0Y!!U6Ixf8KuU5F|8eVbMRvu zggpPLTNBD6UUpjEt8=~(t`EgP>p;+T*-hRdSLDQ8&0Xfx<8c~@d4?P?*xKTQqDF^E zzTY)B+_I6OF!;L78TDr@g~1bSv%|`2P~V{!@LqNf#P0c*#0k_#In)Nx1*b~Sd!$s> zNH2IevF=CR6K^lP8o74C^?B7lPnf@8JYb<6RJPc#4HBMxDm!evTeW3|vDn@#GNV)( z%qWEv13q^S6_qZk75?=LkqSF}kT-9&YmRc0Ktrc-(tJMN;+q?s$V>D2%NAd7z(H_0 z9LnU}%aJEM^h(fhkP0j1Z~=vP;FM&W>4+`1p0PoA;8|!`d}Pl*9nPGscUFJ?#hmS= z)Q+9i69(g7Pca}ZyqAi?B^F4(VlD-A>!HJAyA#N>7D97$CoGHA3T*sP4LDbY*Kl#o zRYMp0*7+>dCVmvV#pfKgcb}6q`VC0HqxII`?)5E&BHU&^EUq!xj zObx%<>FnGSBCAb()&`ozc1Qn!>B~Rg^s+mz;bl+Y%{*2|FdnhUXK_fj+wmyL zoVI9_ym$I%fG@RTd&r_id35UZF6AQpgmij~TlLhHoJEJpXUZ;Rnk2yq&#RT|=br|i zGrVubTX-Z{zKdG{O)B__i_EY%*0v=k=P!d3FSa=BCapy-xBHD*%IucQ&V&F(ANLel z5^!tE9e2=`4^59~q`%@Xf9Cr>Z?~w$DrTO8_f$N!;Z8=S55A|F$bKT6=k} z@)Q&q>uy>7RssiSrmghHsmky|K9pPnxPisc_&5+C2X{6 z`mpeD7_v_FkS)GXk7RMH!Z%O8J`0cRR(z$%b6U%LD7F?Bj0eWzNy_+vR@yg>jnP~m z*N@-oG+tRf%X|0ot)KD*9sE>3=uGr#4J?A<-u&?R@GF8#g5?wI=+20JQ?k9gg|Uts zC!h7pb83sshXnBo!9ZwPKrT5-4w2HZWLTB2^Z82HF2@zRPJ1iytxMuOV7SY2iIbjm z?6Iy`j>T{qVREd5B<*8N!zoMg<0)pg9OZK|cm*PK9LwB5oKwD4Iq1>_9T}~PRz-P8 z7hTFZXVdPCXTA_N-}W;%+wg$tZ+n!N*M1Y!@0O_k^jrNbwEJBbDY-e3#2M6_JfN5X zirl55(7=Cu7g#Fy`|c3z5VS%yh$@kXM;qo{BY3D+s`gqFoDqyQ_DEz2K~{{x>ME$? zCJg|Qa$Hy~!Cr9{QZFUTv8TNfMD?(>RIR{f^mLyZ->slKqh0S{xBE?3BvgRULQ&Ru zQI4{P#vH*-*G^E3!6+RxIO`vc4@L^73}}H)2;ad$(G*opV6$rtzn0`W4m&L6pl)n& zNPlRdpv23Bt*f~o;7#TI%Ow?NdxUo%;S%yoJGn;jnn*=^AYEA-iK$(!KwxC%y( zKIaGhXUT?%z->7)Rh&mLAd!{{fqlAPR_}Gf9Ra0Q?xD{yOFcmFC=ztXeeie(zfKa% zix0dK)JR{J=W&L3HNtIlEY#hjz|r;@HG#HrFLXwRjhz}X0pO+$trW{b8vFGg&Sd!> zdf5wUnx)uObSh8-Dt^hp*NgNVsytB%NOg7a(RQi5n!uiA@QNqrg8kDng%pT3$u*!v zT|ITsxgGSLtjm)54Du{Mu@=3Y zI(jbYVbn7SW=VcNN-4~%&;H4FweR9;nEU*<*mZf&h zR@3x9f3xl11N4Bo_Whe>B%d8%?03Q@!0tMtYh>LI_kaJ<&nn-$ zHt%F4bOVFl`!fGdc@_BZG2|-u3z1r}np-TirmtrW>}V%PD}(48^PLycdhGQ7(Ca9P zVTT<%HZ~as*jYz02^3jPMeXxj4zi+c3hnHLdI%3)bw0&VkRiedzclk`sMEb z?pNtSaWrfikM!EH>yv2U`oz4(3{&JG6@@9xPqEkkEVFnr%G=(it3_GdCJ;RBam(ZM z&x#h*Lpc{FHM1a!fXZ&4LEpAov20=!e?RYV*q}U(DV6ODhr$WJJpXI6PmnHIC$k*L z$Lr{o1fNpCt#)eSBQL-PQ6&(PT;d-JiJejux>%VDMM6#dZefbYb@xiJCRo^!;;~z? z&80AG$RU^8BrFQOCT-N8!TluVj<6Ju)}W$Lyy76YlpcA-MUM*EkVA?`5qFzQqMz>0 z6ptLb%QxG*HE0<4XL||MI_f^=)o_ET_wVP|zUP?kTW_v$rV>f@A$-&1-BJUiu z=%^K&M0Y|}P0~8JBGG>U1nEmko#xg)eW&o-x*D%+Br_@BD z;~0Ty3{*?@M@433-#_t<} zS3ucOk4HvuE@bWTs$S_P54=LH7<4(|-WCbuC&EU$-?}-oXGX;Wq&)L1YaaOWYq1-7 zdZXl@`h&w{`72|TAdWs_L6$->>nV~*MJ0K4(;GxvmAW(th$;7RF)O33IlJkKK_%Ny zXd_0D4KPOBX7kPOy|vfZlGT^u%kIZQXxzzi*{qc?oL6Edz@+*?wdH&z7ayFr&8 z@RkoboRAjr(tYA1(jb`JDZdjME!Ze&wrxd#nS(#7(_l*wu=94kzVrUKjajhRxFB}y z?V8ZpKx%QMcGPF`F_kgwfYh~}@+7Zuol^h>W`oFpYNo>>kI&_E%IVH;=W_c(ZGjrIEyk!5kN)6jZ@h)aOiw%Qv*fFc z*DfZU{zBhkSj8iaPl4c8(1pO1;7VU?c-^P$g*JeIec{;prDBEOEG zq?Z^ox!P@iiwSPl);W)OyP(u}lL7=}+;8^K%)@*HeY6P z>y(%sEDW>A9I*Y~*_aQu&$i@xcIM-J*qqzO&|qUe?05-m0!T&NT3Fre^{t+JFH8k( zZ@fXj(_RJMLp-Q|kzfJ6X8dH7wByVMNVcs@HiX@F7sqe@OuyVYIrHoUa)w>l-Hw;R zK-f7V?0%hMu2SR*5S2_l=!ZH66|x3*)Job;I!TQ%iClHg6V-Suc4~F)gJ!1NlS;!c ziPu8uW$eUk@zRL6upW8`C=p=Ydn+fF>4iGXbkGpYhU%w%{Cgx?@X#Mr3zV@;ty?!7 z%LIi!-7(~x#HS6AZm6uNb?c=!NMa$)hi7$Jy+P-;$WGEbJ9c6&H=nf7>y)UrQ08_U z0yg-Gu@hsNhw?AwSK+!+#|0U@yyr_Iu)^Ai8Ol1FhWHw}SMYBOwi(O1J=;iU$HvBl zcpEn6)cW;9!bVl?guFEAv<5o1RG${h@}}IM*Ul}Qr`qOno`Ygi8I#7rMZqt9EFc&; zShm_WJ#$r(s;$PWXuEY_Of2u)BdTC>*{diP9sp{^?K7~wYR{Ad?of(bAV~5Wr?G!B zb}T^fvYBKJ%cP>XA5%n9bv{lF)(*BeXXmbtG7FkJH5rBt1hAnOg% z^A!uP(ObQ+Z!<=q)yhV)witk!hTo;Mg=w-9W-NinsQ0qM%!D=pNBJGSi~LFcp6lfB zE8`-AM?AtsZljo!6ltNNx_xeoGnfxgy0?X;dOYI&LAReQ|AJ28=0hkj#kqA(TjXgF z9!(Bz;$tJh#gO~VYG@c349IlOfLy^Qug5p%hUACr^H~fo?GeaW_tNTrxBykgZIM;s z72Y2&y7A^kahD7;32l+c&USY0L-FE&?e=MlyZ|{V=-DId-?}^xSWaN8U-(db+3oT5 zF;R;k?r5e-vl0+l7`9Py_sqf5cTepV>7kX$PiZIduMD(`44|c^m`xPf2*lIeIw*e| zrZ0$#LeD^OcgW$S^s4ib?_%J$k%LFKK@eDXwD3K%2l$z9mgNEV0Acw7W4yd|W^ZD+ z9w5$l4=0md>;PiNizzTQBWzD2#WYZ)o{B=Ufn<4daI2y^42yQLygS|Ju&j34Vc+(k z8lkFBwpOsyeZ!2yzG_8{P_4WmJVEYq56fB=x^~^_kk6pc@|0r@T_a5bo!(XjHdU^0 z8g^(^G;_*)63GJ{$3abCc_^|cbpYcmHqJHCUkO)1^^IjSA$ufQ-Kva=wV>r+-8C+Q zv0Kl_L}3d$#X2o7DA7i}k^-O$R&~(Hvl1Q8S+(_S)UVL+p!oxBc%QBJSUESZoqKQA z+<(*?nDWUF-zANsl{?t68G!=-5oY8x#k5kS8Cfen2M%Ive~poB1h(m0L65{BXS2m^ zwbQMj@=zV$G=ivxj+0h0Snq$3bJ!Ov_tgrNq8xTO=oiDB^0+fq*A0MS!W>wTsg(yg ziGs2K>;}jYG|8ap8x_Rz{2w}Z@bR;1{dR)`t^s1&M#t4ox<{xLm9ichny%Ff{LFms zSk7_6%FdXNHWuc@6C7*{? zDz9;RJWfbkK@hf1GCaGR#yxBV!ryiDxxli3dQO|7)UDQS7+i>hen@t6jMVV!Jud)1 zRV%ck#Ru+H*6oOYnUspsgP+X_Op^_(?81fhHG zK|ef>38@XD(Vzb!A9Q?&Zp4_rU=&Rc-6%mA`;GH|>{D6vve#Ik5G9g)4Le}?81vmY zy8#nPH(fI=2vWVo$}+ZC&Sy^Z-=VNg-e4w{FndwqdbLc+=z83D`205|$H^LJSR{I&l) zIsezjw4!X1q;_0mVS;cp9yKn3t!clX&H#Sp{*^lj!zfocX)A{d1zVY#!2Idwydj*Z z4JOtSWa00N*z4FT-S2lSpn9ri7P{6jv8L0z1271+g4fPf>DtA0e)TztmJJ2nHjoE+ zk*zdiaBYJ&Gd}nQm;Cg-nBZyp`EKcl9;->Z9XEhi8%)F{6a&lkd@Aa;{F40M)b@xp zNf8i*Ze$8X`~0uY1hTny`2e{fF7V&!u7<89RYvIgFi2^nPEVG{IyFl(mAw#-hT+}j z(gON=V93A$k~fkz_odS6sg{@)6W_VDFks^GV?NBv;!ODE{kQ+%uSd*SBpdA5D?em_ zuP2fXU_2nFT|Ko>p<3%wrE2~1eUJhSx98lP)-Axs-A+5qfs+_~3*MK=g0CO#os1P*wn$u;<;WUH_n;c3$iI<(K53 z9Xl#l3?`D36w^YH6I4`n*v$Y9w@$WtauRod=yMbr)kH5mP`YyvYuiRDYAx&N(RcGZedBp z?SNs2HfW<7f}dF~9WFLl8NYBe%`z-Drp6xakL>+4XZ||J9d8lKYcY1uR^Cd zN0|-RHA(J;0mZNq{-Y0(WfOC$9tNKZgnH6~G3ghYiIOomA0uuV4F)eb#%lVZ%KrO) zxp9xb-7bYpbnRni8-w{6O|6|Xc@cNuwpb?KkBG>Zc>U*KCS`iAN3ID^!5?(g{C0^2@{D)IinTj<@pY zcy3=%gZOjB4&Q30Gx8l!G?Pfu>E*NAB(}_s|H9P7iF4<9Wp^qxjZyk{?*swMMxLl9(XsXRoX*$xa2CY&%*1gosN)# zxx)?_ycGn=4YQSde9$-gD>4m#BIWC>A!YS^QuEccO5-M}m#T-(MB@}>hC(r;gUl%B zB*==Q+8KUae8^EF*%FouA>i>v$cusTWR|iu77U(m8WoW`y@O%^0QkGu6s1tdm;Ke^~VaGmhjlqcTpqT9xDL^sYEx-(a-mf(z zk!j+#hG2(DKIfAwZXe&_7KW_m)q5s*fE2Re7y%(9x(wK~2P2k5G|oB7v_&rAoC(oL zc7|nfas`7CHMH(2c+LjMU=4yNlg!8s>qCK^N%_ttq<2go<7(K=MyR1MmVcLh6wiz>4nV@+O$QsF*N$CQ6BZEVmotKq z8Z<_*#ZA@aq&h-M-IC?4+U2lT*)JnF4pKs(P73Y(v+;k_@5}eL^>k!5qDDF>xNT!h zE%;}#>&xzgw7Vq*Az{anIh9T=bf#M$cfo@4$cl&<<+a&s!y3pX?`t7LvWwn#LbH^a z)AFD}`UG@44TP?c+j`W;L>@E_qt0^e`_a5<(ZAC>s=sgkxj*@m-P+fV3k_BpMEf67 z%zX;9kD@yGN8ii=!Pi7yKCn@B2h~f;Jyo4@kY{wmhD3B=LF|2Uy+oA|fhp!M+)5%& zg%3MGy*0N18XvppHJ|{G_d;K7ldN1=>(UG%_>Jys<;&c`^z39h=yQ?=UWbb+n{0FG zpnJpeIj1J?^T(b>P%#KpD~>xh1r9rGb3wJ`d;#`L_Ckkchs$lADh9Z1fWcFgBY>vK z{x>@1mA+^Fvq`dRnxvb?{>m?e_=Ix0I#AUW6wm1m+sIwYN$|pl*GLaS4xqsK>P)p_ zKyt`=*a5pPF<4}`k^cC?kGg-Ohu`(^gc;hr;JR~?K@h&^ z;O_{oDe*#vC?fNGt5 z;h;-_S9_qUj^C)4T*kj7k4|y1TQsrrp{_4p)Zwki-tS5p5=ho)8E!i^aPL#=?BXkY6&d!%LL|YJsYQp}nM^Y?u+l+u=Fnx6J*V zJXyYFDs*mp9B^N_2hgil?vVlBY65G?5J*<4j!({^Q3hd%r=D3IvdDUK>zI8fV_!5&tM461BKPb+Uy6J z2E`JO4+DJlro=z}Z)-`e9ZzGxAu}Q;u#;lSDYAo#Dv~sC*M;hSqSyt9?U4-Tj-=ZK zd503*dw^!ZSfRGTJ)TqVIpknU2e;mR9ntU~0U2xmEbAE!?SPpFlF`kC;Zd{4->eHw zgJhJV1S*>P*AQc_Jrk66I74S+$9s&A2KYAiDJnc+S|4;t^2()?xz)h)bA`S+dk4(* zvJ>u2!u5fw;>o?!vx1Q&zSOP4Q+p15%|R~42}l0-_n~>#g4^%Bf1uX!aH!tc{AgEU z1}V2=W7BM4Yz|ON9Yv~9Y<6h+GNSvbGdAVID&bK`N?i$Q;$IZ6f4z>hMsDELyZ7>% z73tHvg*U|&;wvF_G*+ zSLYN9+ah~;TO$XYPf3RyP7Cv#EXTU_ERTU5GRwL=?;>`Vr~C(Btg)oSz>X_FO>`J6 zn^@;~W@-z)d}0T`PBJ7*mgg&bXQOhU<(PQhxQuqrObm^7t4)usQ6K!^UA=Aj?j^@Q za(*IINRIRuzNDDz6uC-8Vd;^UAr%|ecX2zl>des*y4k5>IyChh^edvVOh3*2oQJB% zaesKqY_+nPqu~!bpwKfi%B^xasMLzN;wlrR*N_;Yn{J#p=-kKGMpy=%tHL|w$mLoG z`7C6-Ia9-Tva;e+$vW zP3Sl;i|nxDsSNs8M)=Kj6axeyd#EVPYVC9%bj}c|3Pi{@+$n!Rw~=CDE+G6fB`|^L zs9xF>51qFG!{VSz8s}y}ldKZT$}1)>_2`o?m8$SQ4BsS@rNQN)z0%q_aG!cXue3>e z=O0hb*Ews=Qng|+DzBsOx5DC?%x2hr7Ot$0u*0_V=70I4ryfF&n4@t-ZO0JWX8@sM zih=s9ZI27k57RYTsXM5(fXa(*=w<0A{j(s9j2q_yk+Qlf9OVY+{oDz2gTBF*6jOP4T1gECI}U0b{Kb>*sbX>(N&^>d8M4FkQ(u(>mcN(#t)q% z;bMld7q_Az6N#SZ^$3e|x)?B(Te#+wmeA%7#aHyXh1IcuJ+BzLanHIuea_ zfS%Jf`l9%j+X2UWkQa;>WzS9V_)@fu#w*l{eIcvlwVWjGknIaGwAe5MPfyIAeq&rP zm2{|Ykz9|E4Qu_*kd#*jAr%G)$)y$mLmmFzFW6+ z3Ga?%B{bn8YtT`0eHOIf^EU!&kN|PKB)N!?kmsoGbX;uwSiafucn4XTbo-5!+$EiB z-1L^D^}U@7Nva($baxrphJ1ZIX z#U;Gm;;SH<6Cb#G@+~OET1k4`n)prp9=CnrJDnOp6=j2{f%~PkGl>Qm91Dn_Oe_XZ zvhEHm5L3T%2-)p2?9^htjk1HRF|37ATgQ45eJ+=GI3W=X}*_Hq5QCE!YPvG z#0q7Pi9#u$SlE;27@V)&^uu5yjaZ8Zbdr@Dn!bVCY`r_s<%%>&(w(X_u#W8y*b;%1 zK$v}}S;k*F`>Ysy8DQAKL5rk~V0ofc9F-@|2`m$${xIdg-#3Eg?T-e&Pma7YcI-71 zXq}_jGZbk>AKMM-Ct>lj3;~3e`Ddf5g4UADq4!C*_tm-evNkyG=clL}2>#qg$A~UU zdwp+6A&LcU+HeG(-7kD7t(H7JyO!Vx{Mkr|T}H>c-VTcqee8KT z;u^0>vgWmI^bzk^QImdro)6xKcVtLD_S{BqVdISFIly-8P#l5E7k;!}!?lzzmYb7& zJFQ>E0&Vw6I+KqbD75NB((~f|z}DU1y~RIaY6Yur6oA3b9(s8|y??oPw+y3_S3I%W z8Y7dr^is0Z)lhw1F4-l;s~BC)mmcA34wAjE>)V?OmV4QCZL=>JtQYo+?&PMIy1smB z-anQH=F-Qhz2e=n zcrRoX>z;m7u7hZE4_zQAq80*@pd=($o$uRCABm`#*CPK28rULrCN@>7@WKoSxUE2f zw+*^ALt@SyV8qj^&Om<(sLp#^t6H>Z(JjGF_yZp}7$g>DYKtxu_+WurlK)NuRsJjdP+n32>=6|NtPMfsiB&%tEu`z;aQ0cgNAtFZ1%e$ zE#Un*FS{SE#iLr7zz!0yF|eB?{8}E6OV_&f(tFj=dkE!T;gvqE+8xs0YGpBj1=2l$ zjva3e2xA2c9jhcbiQRO0u%C=anaQ&4F-i@{gqVFNI?Zc9oObtXp*SUzzXO;{FNo37 zz~)s=JRmyVAwU`$;5@CLRzD50bRR_S5EOvHfR%)op#{~|x%#bfiyZ5qb0C{-zgpiN zEm6`f>kh@0W+&5bZ;q)kV8#?dcBXIOgqewJ$lsDC8JC=>?%!+3Vkh?90-e-|RDxuR zO{Ba%>-rT`7*H2$AsmBwiAYbr&^FMjho&Hw!9*FX8=FQqFe zb}2<-%;fDTN9-tnA>%1-nDCbU_eq5j6B)1l>bE5Il`%u0LU9Cwaw!&?pJ!0fSTsnnZe+W;I=4CjL4<8(R*g>k`9C? ziMIF;C^mzuen8nTUCPwRvUpFgoCKy1M=Ba;l|Cn(Lap!<5%c}7g^gmW@+svKs&-9_ z+|XR!kxzN_m%++sKl)$?p5^rHop)DwzVVma%yZhD)?l@erG_yU{c3=LmUD0Rv4#eU zc{6%v=tL`gD!gj=OPX}ukwJJmzZiisw{Ms@Hk{yl(E_w_(2s7Li z-k+xW@y$~pxj0vyIGfr+CMVPZ3b6)tm3Nz>)bo*|P`b|_`FC{mXQ6FTtxS~Q4@S_J?T)$`p&=0SlOecGtctL72fy5Dth3MWvD*Adz z3v*GrlW7VVf)HFnNPz@$OG0+3GhKIi7mI5GY632ZFNkZwd$v0=p1)>Vfe1Fgc=Hbb zx|tf}i#|$~i;Y z=r-h54+K%Is#ZOXT(Qp}Ic|GvMn}NI1S>Y$K%ADV34i@e_bu~6QVaE(NMFzyvWkGe zvyr|CY>}}9xdKy_tD@3XP|%G`j;*Rex0ax0I+n7_EBNB~gS!-7Jf}VG;3fs9W$r$e zcWc%cjJl@%E7$W#`75(nL4WZP?zn>#3rvL7m@%pKX_nm#E>l1FLbYmzdzm7S#(JOq zVP#AWxggG(fm^8z$s@%cSFP$SELqj`;;BRKZBa@5oHrBLs-Ph{cjk7FYoQr*azwTG z6k98+3D^S}l6tE;s9BsmrB;zF$>yD6JCywqwNU1q0nQ$5V?O40hWp!Ks&OnGjrEN< ze*Nk^^ZuOArY||MWn-ad=hDDT33_6?l$ZTW12ZIBW+pNT?8Ygd2e-(J#;s<3SUOfq ze|W$+{kE(Nw;g)>x_Mfg(>h@+q_FMt-^oi=ruZbV1|FX=zr?ml<;B9cZQiktvMl^j z!pIZve=GN^?-VWDBe`I*-ZW7x?3NBw(J2f>ih+D;e^_RS=E{^_ z`pOi&<2eoj$~wBw6x?kz%uP zI6&8~q2Ks81H)QaI5rUQEQ-SZ`juw$mNZUlkXUFkQ>@tGpUvChpG2;PU4NqsDk-vg zSLc=>!%`k^01CSXgS(V}tERic9{aXMb%?izW~nRb@kqS4Ug{=m!n0zs^BGPy3bo^N z)m`&&veRP3LY!M0SsIw5?jZRL61Jf>z|xEH&382l(K=ArS>jeq#|ab@tG+htujUMX z`%Tv8Cq#Jz!6_t>XZUHL4PN8Z5AJU0#!kmFW$=P_`|S$tBr&gbLS1Gpsh_sot933C z33n^{`Q=hPe#l2|^HcNA(7lf8RA`-COkPy811lb~4YzQTi}Sh^T;6F8GiMEbsazr! zm_$o};i4@n&F{}2{|aPiludEQ4V*Z_1?0U;K$V4W!co!~#F% z61ExpfqAb66tsul1|?9>>+@`j%Ah|FD;8IH)ylN$GmcsVN0k^SY>YaCV@`6y29^4L z4aHY=IO>{3})x0%6h4x7rvErF?mbVpi+T#{Z`eVX| zqO%jG8qwkV>#7B0qZ6Z}!UP@LC>BaWw^GqJJcet=5fbPWZ(&YB>U{=XH5Iu|Ft&Rd zD!?AQltRfk-StO=NvY($HTTt#`gFkj&pF* z5)+g^S8p;$#4DmDEZ_%?6TMNQqc=}UnBzeDzUK?h8vHze2FF~LJ`Pj(7@huq>r zE-acShRQUgyun)*RS9;H1Qz-@s8)nDEApj@{B~K4Xc4xzvirhg4P;Dg5eomCFOlpp-{*n8eAK z25y2Pu>y4gBz`=AzNeTOlIL18L!+CGL>Msw$0Y9B1upi+3e1=Jn|}{B?-b{>4l)ZZ z?0RUVHmUGR@wv*|;#nZ+6M}j`PCZQSIpCk?s;TDf^M{V9bk3V~!05P|6op`l$W3`B zq}^8wYseY62{$r-xuPs%Zi|Oc!8D!1A-64_S6$b-=F$E1--fRzuuy-F8VD7RR&`C@ z1xfrn{L8`>0}!iEa6-~VxBvd&?*b!|{;WCuM-t1;G;-d!D3q&? z^yF=%Sl~5TOGPKm+s*5wYlWqr{q)+ew?Tuj6aKLty^XI?#3yI+d4JUT&YfYe$M6n4JSw%%RyJkyIih+2iC)D7)*W9Jo z*x%-TvCOQ4#Ea*#_Z{!aR(@dafV1GaLIu6nFOA7(GWi7(!(%q3Hk(==B(#cQe|G7)K!8^rd!!aVgt^xl~fg*CFxRNhTbEW z&Hj4=Dyc5T1FtGKJX0lT4Q=qQqz*AlXXCNGvBLo!&)>e-4unxhT2_9G*LS$H!+V?i zch8Xxqv6JhU2?ljaHFHx9ExO7(Rm@rBaO{^&{Yc&(yB&UN-kw0Ms+L|Z!fS>M zaOA1)Qr@TUGlOpSUQ5hn!CmskCB}RrJ0Dxt7^4{=tAG2gbyJM%P5bL-eos=I*p~(- z$cXHu9E#1Nh}J;;`RQA~0pF?n8gPtm^4~IZjj}BYDL^}g+vnDKZJ4#2H{^B*5+Lyo z%#2MUUEYZyYk+|G|H0%}KZ^d-Pk#NoUqwr4)`d{iXfWyewf}#Zlb7!>k;Ybvg}v}* zDjMIxa_>jVCa*L<-MhJ-SfjDob+gBUPyHW-EkKuKt-9W$N{~Ed3ICcLlhq;8#&1$r z@f|IS3AVnnc78*jnU32l>$H2}(y+3KFBx^o{7PXDsT<99b>cnGB@;xqP%IQ@9HpYy z@IFzNORIuv)ergkOv$_^Ngnhm#j_uK4)F>+n&oSu!l)8#61*cDdb$;bYgKhnwDHKL zpO2rHd!?xHT3ZyJ#ZT?Z8~g^3UC_AFF#B49XE#E08js_YUDK8Z#)#HWT@j{LRZTr1 zt`gw+Jl-MUc0r2zR^V}=EpLmhFTnFRM`l-_f%l6~SbhO^2QAWlV8oPiwcpoB(K(b^A?e$ByGQG-L2BUp}XVwQ4O7-EnLQxunT@LbSycrAp3m}sbtqeTQ zf!*%55bVLMPlUy9$6Tg%S@N{1&{cEwS&kUwD-`w6CA>`L zew_ujOR@I?gDJq1lrwpLhf%OzK5YX43LGM?UdnfU4x#4E&cKB^KttvedmdoAm zK%0dw<<9v%YUHAV&u&vgMTtCj&d} zh+X-1iUs0J9TmMUs9uU>DVjmI6v&sl5tgq)qJ#nuG@%BZZveYoTqfflonG#5+7URT(4Fd6xp!O6ug)$qOBk6>b+Kz0g#$9znx*c0HW?KXZ(D zd8;Eoo)mDyixV$FhfMIYmtuEQq@0T00l6q`QIC|_{ujI;VUBL#wTAWxbLpd`%VU}E z$FO(GX9^>(&fOn2#5+oNhZaV3ku<+s^g}7$n~B_-K}e{E zfqaxiNvS7p%&r7K4o3bqt*TXpvLs06*U`5in)onyCxaxh&9X8`XKM{Q9fXY5 zIRtWQeNTsu15~(z#PIst4-A%O;sg?E!}oSfHg62YCA#Ru>klk9BkYxYiiK3kZ1C~{ zdr+M6fn>Gs5~y%*jmr8`2dqfQ50veGesUx52`!6UEb5|HhCBjF{ao>%d+xwjpbHd2xV|VuaegZ@;rB!f2~(&?z!Wg%ih+T1}+iK(UbM zQj2`m-NFlRU6O8dPh?J$q_=cmx->r#6bbV1WcXye*Z%YNcknn@l?uM>9hCKxCJjcD^7&jR__p#aaA}1L*Z$4u3dxA~o zX3MiK7w7%^6APwHJg-;`oW;ZR8kjk(dFZ>uy*jAb^$~xK57ZCUF~=j0Gnq`gtO?3J zn!qpfNRi63B6KiD9C1VR2;DKp?|sp5%ZNU7KKCnPqXvmfzx;EOFo_hJXpk)wn?{ih zRP;Iepx62t#Sn|ZM1W=T9fB@Z4y|Lb57YV?XBe!DNMw*Ity5U+z*XJ!AItfXn+|u_ zk>y7|$`-?UjlT|2;xElp_?>p|W1+Ud5a&!))+&~sJ z_!K9wPW;l;&#s!M@H;IgEF|S0C0Z4hhi(tmw}4W%fa6T7YGb#`vOQx$9N{>9o(QeO z1+Ev*o5lXuIN@T#2Ydca^0IGhHu)b9%$WITPNM{W764K8@xVKQxgG<&M~VSO4-nj+ zdvmjY>HMXEUGwvJLp;0=kw0iYmG1Lb$Uvb!S7u0#2bTFa%iE&X29 z1oz1xIZX0kisOOh^j50wgN}Eq1jhsG*fmt#Kb3_e&*as)wryvIMQ=TjMi?K~eH{J@ zn|^Zw5_SKh6TV+J8YS1m>OLcfop^uTV`43xrP$LH`Iw3>0bj@x>WUJBf*J$kUo5yu zP{$ODf$$6Z^cd2!u_-R3w=lrdRPB$gBSr!-MX0fE**jT5#eeDeJXmqU!$Qv9!G3n`gIj+==z^{BN3vt=8+6ZxsiBkE6s@i-UAhg-q;Z3A8cBpT=hV*7r2 zBT4aTWAtNGTLKr0(u6>s?V1I`y}@5C=oGdwI4@o=We)J|bU z?*Na+;>e=#)~LUyMV^XkASJv(veK_L>MwX$Hh251eC7tfHfU^XoF$am0rr-r9@F{R zA5HXn-KcRYXa8v>DRko1?68R{+()r{D6$LLH#&t6e%doX4cg3AcwJN{@o^F9bZ_Co z;*;--BpaByna>OoTyQcadFn=bwE-$3)`#rdjXa zOz&04QZ4cf`mDG*Xs>#qeG(V!fQM(-r7@&vHE52y4wh>Ew_2)uD8QK zl{O3eJdtR>EebnNT%G$!ksOgIsim8R1IkOnB#$g!PpDSCS+y0wlnaY!F__qRwOLrj zM;4BYyaNzT($6S7>5lu5wEK zSn4Gl0;VVMvK>{Nz(n;j(@S3L5_aM&JPR)22T;R!mH{()w|_tMm3bI^Cm8e+X5h7f zF^t`rfhTjw?W8!HhxvD9vfYpq(Z$|ptTujamXOWBvH2UV8)z9g^K(x9d8To_y8A=< z3bJ*ysuU*%Nv+8mR!XtZu&M}lWRZL3^+Gsc$PKq)#stk4dFj+tWrbI(YPVcZ(%q`+ zQt4i6gc|+3ey2%+2#Ltb!iU`ZNiN+@?}4g?6t!lX`$+>8Vv8J^ob6r|XESJ?Xq4d( z)`7v=yDWodDDRgmUUolmm0$t3fjL9pljHwd&ojU@w$4G;>T@5tb@G4i6b`v)S^lF0 zH6|onKkz5>UWZQWC$`YNaKRucWzGd=N{^^72&%vxxSf~hS}W@oF87Kbi#=hJc&vj3 zy!tlHZu=lku%JF`n%r*Q9p)9ul(NuEX3*`pur)NzFNZ!Y%=OsmSIBIY_0yQY8v`D~ zviWxuN9pTtIA&Kox>T)360ReoPvMZWe`|byE8l2)%--}bhsX+US=>&X%>cA6BeLk$ zQ|vm5BvH}#cre)$pP-7Xy4A6bJS;s<>)wA$qaAmi(@HW?^Zox_t~P3m zBhi0bM@lD=W|Pz^)3|8CH=ZBb{KetI`^lioh}`WpkvJJJTSfo+nM zOHazK$xqVB5n9#tHxik}qFe0MInDAl0&wC-LHdB~kpinEZce=xYPf$lbi0LcfC}$y zi#n!S8rb*tB`F>y1{HYJkVblk9XelIC#zwD-|F9E79uBEIj_@uRgX)=LI;2Y#7;;dZDj*dF@Wr5dPob-=frNUnuKVAQ8ryo|Y^AIW)@bI@Dy%R1CTa~eKQ@q+oe}pa?h)+ii`M}W{ymC)zFTFjp%4 zx9ow3J#3g!rE4`-o-8irA4Z*a%qRJO`~8p1d$&GYk>tdPs4!V@p7e7aZiUnUwJ(sy znF1w9H|0l3){Ft(auDkUev2`9D?hd)m1W1KEpim!n*^d-^T%qEvJ*X_a zGytg_K$jGVE{7xr4JZyN5}Ex0S~c?NEnMei=v4Idr>%yFC7E$?5u2+OlVgitVB-bd zuRl?k_a<`MaHEB8LnHf!=S!=mYceD?;$m@L2+ADWMI09PN7!AW&w1dSG{N)$wMaQrhsN)OyGVXMLi`i_E{;~3@T%%GL|_^?vr>vI<^l%N^d`Zv%gkl zm=8TJmqYK9)yz8Awy1Tk&BEPtjuS}W3y%@u9P#{e+c}^)!O28hSa)D6%s9b`N?(+} z%zQ88v~aSpyXkrR5;#UOnQ9?4lO`u#Z;M*N-ze*)Ka=bH*Nk&d8V+GcTLo<6IBvM8 z{Z42QaAa`|oikIfPC><;MV!@&lKmKu6)f|Xhvs^?@tUi$)3pDS{X?N|2` zgDHa;=oH|e@i1eY6$h5}itrmQr_9^YJMG$Hp~3v!w`$(*p08taW;W98^OwK7?49wD z^|LRn71Dop&^Vm2Ed5c8SE@gHQ>m}ff&bf*gKJ3kBxr~`GLBM4v5*+Qor+#8s(9n~ zTA$XSBK{wiPr{??Zye;WnA8fST&?m2%c(!z|9<(~weMxVwUSzJ3UV4U=%Vl%umi9X zXK+dqWW}SwSj6viH_T!%57T{aL=R3WQ`l9CRtU#Bn-A`?r_8ts$7y>@JV;7k|FS1! z8@j6YghFex1OC-e#qls0E6VH&L+hYp9aycK%bs^}0_#No;zOw~dlDUb->U?>p!c`F ze*}8qyOgb}lVZ)d!@@dH*jGDQ10HUm2vz&6`p~$%d~l_Skb%+4eVjO_AbH}*wQ%V~ zHfADuNJTGb0(QdXlbVGG_=h5D_<6h{QpD@0o8=FEW0}vqDtHY9JHvDfA41YWzh|1) zE_O%6L9fKfWJ!jk4BR2WkvDV2q?18s#cfg6jW`6gXXgU)LsOMkqwMqT2WFlHzAa=E zNn~^aocDQulhve)!N2>@tVO2P3#^@Pw2aJUr-OcM&SdDcX0HW`#S}^9ynLpC#JX0_ zE1x;!u@`)1nihE)G#J_9c~RXQG+Z-1ewvCS5UkNIcDctg5Js_pC#s@dYN2iF zy2%#=+ozPNGec69SYW)1tP9FhpC%f-FreHYmP=z_)f|Ye0goNjfTGtOo$JZr z)z;IY*!CPa3dH2Sv+WzL8^ui^PHXl4q~hdf(}q*$-VDA((w&%Vq0(d%S4gp-S8}Q7 zMCPU+Fl^8Vq?HkMd`Ne6U69Sw&Bol@rGYx{1sQ&gl5F?3s04Q9YyEU0kj*4Y7U*VI z@fW}r53WPgMqZb4m!vKt+g+e|<9+7>~fTb=c%^7bh4~;(I^&8Zn}!)Ftyc zS?R=R$u&XC28vCgNHP`O%4E3#xqJAP;6^$p{Dw4-w~a2~9U~t@)iN;ut82!{D8}L8 z>u=?N46aj;>3(j=_=U&phi@8@(fYk*xulX?(xemTu$(c$(P4^hpvXs5bPsTJ!8Cl_N;-?wZcY7YB(gt4)@50hHYlC>9p>(Wxzj;ywv-{kcAp65wP^lR&B)r zc_UO1hmSVmPabjk^BG1gE&Khnc#`47TBgbb9VHZ7M3Hz|t4kc%GB?nK1iWxrcc>8ZNoEP8>h7(7dJ_x>(`cufa_H zGDc6DSfbcHC-sdXH%+QC-s@aIw*nKP&Wm^Q>w*scuVcEPH$jX@tGeu4K69MJ`&25{3i|1Y@%h9_X-qtL^Wj&`ox5C=h7+4C znI?S_vWBzre86S#5`=8)k7j)fTFFVB2Q_oH88gP&YA2KD<~Q;8 zdPsdh#k@-CQJ@D*4{3|455R_WhAXwejA|IERiT(tm4;a#zkciW*hyo5*{yu^9PB*( ztu{XJr^7hCz0T`i|C?lxXP&XbMXfmRzYGHABLfE+6uX%so2Y1PgtkYO4CKDZ29+$? z|5|Z)r|>g~0^srf*K~p+S-0XwV5h!c;{MkL#>pVD0q9TB=)g}`iVvRu7vugWlNx%D zoN(g(&8H?Zzeus|6ltTP4@~cdT8shZ!l%p0W0xdRFP#gO#+pQ?S6mjJ8M1!bfb?^Z zQ-UI@5%>9@pOYDq#2=7i6@4{b5;!Q+T!T^o{P4g{(;>bhFALYr zKF`!KNqlRTIU69uRZos~{|mv!NjFY7l{@9XU-rcQ2LD8mDK>%D6w(iYg|bVD&J_)o z6L#uJ39({h9Q#gigvJX+YOfo(K_lejZ5`&m6)vlj6YrEQcw!8+rWk2PUjGCBm}J`u zB&Se>*eNs|oTM}PE%J1zFHC`+S6bD9nHchFm&HL48RC&h)q$**25M$UW#n?i*Jx_Q zA(<0DJ~1aW*tl9{`Q)r8#ZJ6h9Wz<2swox-+V@b=r{)!hVp=tlsX+6yX4h=#UB84` z&_{l4P?5i;jX5<>M{k@Lqksa{Ps2Wi;tx%%s4V;;Hj-$Oqe;I#G>N~J$(k`3uF=t( zW{Rg-P}eH(Zy4297R@8(Yr_jdsCiFI^Rm2JT^IxVJUbKM8+`n zn1))0$WQv-{gpXmq0=U=Sl}SU(o>-CR}3kSvi0GOG@7#g@VX9cMVa2n0`6 zz?fvpd9Y;=jHUp7JR{x9oNL`_fnkAHJ&pMY$Vu(9=q&h+Sv=@UU$tt;73^PK5m;K&QLNFxuXt$ zop;tuZLnj#xIz%o6X<2lV#E9-f_1I?FV;8I`!yOQkt7uht z9k(|+8k@+h5836f86=hSTI84&zCXYk4~Z?ZM)F|IPa>L9{l^sq%l96;^ubX+q(-g`(W-`&usOK? zhUUsQW5Oyz^>oEGPWqJ$Yt?l9ooQ2! zs0WEATiOLwf;vHrXj$Oekc0?*ue;2UerYNscN)HK!>mL8Mo&s+NQ(!gJHedTe+>hS zhQCK1tXhTAmoO!IG^4_4cNO#h_~m~w*Huny?O4!0Kmey#F2J**tmB`UmIYpuUiw1N z61G{4VF0aa`K%+dL}rbT=CU+SeuBWlrDyQb`ZUAoL`fgL6W*zl(>vWO)$v{qTECoC z`*1RRCZ7GH`_Y%(Ixd4o=exfOQA{n-%Gv*XwdsMV7(Fc!7hng*vqo2RwDcK-}u{~ZZM{|>c)2=y3 zToaOAjPNM>{1<;AF;48t)tcZoiDDBdvI@DlG19j>Amx@9#?|L2c4#1u?nlE zL&oKZ6ILg9PW1S`xi^l>e%gtXpe*?A(j!;6#e{V$@HYg9!5(R2q4!;U!f|!7rb-YW*{iHjY-RI-e0}^mYxTmg1`(E| z=E(EpPjkOH7FFTIh_Ik4_Nl>HA}NW);tNQ%P8Xy@s^4<2BhZHC7=(vUhmP;MV+}Bj zo)bsn;F-CVhgSEz?ERX1LV+9~{>Sc=3%fj*_@zQ6Y`Hi1)v(SrL$Ws%@&(XDMy95P z^#MhCrfY1LS|Gwr?!qgct*ZS2D?-Yla|brT?4^sSw*u|IzMLpA_u47uG(x~z; zpNi6|vZZ(U^-Qhes#`yO`TaUzC)*11K`O3#ApA1SH5lAWr-s0+FxxeG9`B)VwfK~M zaulqDm8HdI)br1Si?!D*!|KKgJQ2yq??M~ z3K6E#z$3C8$XNp;28)1}h_a?1P-@zPN&^kW@Y(L|tQP#RB_zQ^A81PP*{VK8H<7i# zh|mM^tE1$Im!=hbFX_`ZK?rfJ`vYm^yhOBdpnFIdAtd*KNrdf7X zkSHmp^L(_bE5VP!rMt+$blOXILN+*F9+Y*uFEV7*UFG#c6LGM%**d0)G(*}QY|Qha z5sje<8X#Vk3Ax?fP=Lu=G z_zJDhJ!qFf_QAsOuw85uw3$>KaA;ZKj1rZcSl;-a`Yodp3S0DXC#ia6ln|7Mj7XL| zPO(QQ(g-c4^wb8Zdt%&6{UC~7Ai=;6?v7BsKv}%oAw{ZkBU9;L6|~8}fdyrA0J-Dir$k6jLvou23h#%R%P;Yp7?x(517As@pIpZ>trUJ!)GZCveS z%>!ANs&+EQ9AiW)d?42$fh~}fDYgJ@w@tYYwxRDuI>Y=rcX&Qta4^%nL%7pg%NBZm zXGn6TTc>~Sg{~G9U&Gg0|2glUFqh?mV3#*GdoEK9%0SDE_|JS9;Qr*EH)0IZe%`cW zzQ#@0{NVjJ9+>x=de&xk;$5_bu2Qv_rsA8#*PM6FR~ASxmT7RU;3w>4x5>XW5EA(Z z-CBYMgFAw6csx?%O2C~{AAUDHiLDY`^16m7_8vuUV>q#R%8g(wF6^Pp=&GO$`mDG*=wa|q2F*`{ zMOz(&+1^9$_%o5g4paw7ZNQ>jb*^Vqz>x9?v;HYMpmm)D3=v$6$Cr(Rx-1uyrP4 zWz`(*oGQVpIgV0KqbUbYDr>X}9CXgVub2Ltol-u?I+x87a= zmI{Qt;wGJ%H26-(cMHD0`YWq|Ca2~bb^m!Ps*%1ibt79Ju!7$OW|CFVv<+ONt1WBK zJ`YsG*9!mipOeH!aK)ux{y9nD1}-PoKZPdX+Cs5V9J>L@Fc0`6T_vndnUJ-6gTD?8 zQN2Oj3Vru5U7{=;ZPx;iG+6>0KQ66|tPXmO6#cLW9s2u^5R(;Qq^?b#AbYl0sz*_F%=mR4jD60-Fb@#nt!q+b>AWgH0B?2|w{1 zP++@w=<2GQjXl!ZW&IH~GcE)Tx-|>2-RqGlYyGgf3ErFw@dTuh(YvZ|26xMFCDhTK zK05k>k4>Jw5%PDSmKpIa+&})37r1sT{B=&OWi0U5-3W^nRk`7oIG(5TPMD(E=6;?* zEo0Sc!Tkw8XM~lYc+L!txiq)(JNYkrnFyv?WzZRc8ho{J)oAceD;9g1CqX0JD zA+TjH1ZBDHomn#jD@D*xg9U63$AOh%AGxCFKRZU)PiN(}Kl;kAe%Ue8xc=Pzp?n3| z>cn*buz-v(XGu}p~x=a%9S0Qj^wxo`Y=sA&)}^`k3B{wk%!-4sJ6@Wb@cHb zdm>lODRD20iV52jc`E=v4e`=s$RzgIrA1igjSR4-rOM5H5$?v8eB8zypZw8r$PC0hmBoO@xn=N_54~SF`6k^!u1^_k=r)eiB}wG5;4N4 z(Nb(WMK&5J#m@2i-O!z4=>H84m1_U3>|*L;$q?+GQ4O`Kx@DhveFk>%4#D_9ggZdS zgE(lygs*<`aI*Oh_}R6`iFJgfeJ;4q^j0nQV8fIs>?2bdv12@FqRie(D@AWM%JGMm zmy~6%t^EUK)EW=nAD$qOTrA%_TJ*?i%eM*8ZI93zi4?n^|1Z+H4_k*q5 zgN(?DOYPr5wmY$(h#1jI(KY!Er6vmk zZc9W5rfZPR7#XW!Z@EOY_jQdnvNTXfFA?qZ1V-ZqR zPZU>6FBg64zanISSMGg#ral*CQ-D^ropceLJ2||Kg}=rC1bCsrCa}U_*@iyIQsFh^ zp2XLeXe3Iy>25!4NQXPzo5AgXjJ%Ko0XZRy1szJ{Nv;km3+cqF;D!&|S<8^mPAS^!#QM3-_|$R=V>{ciFKSvas778b@k z>NF=F z{0DfwG^W|6O@T(|c*Wq-9Uq;?5e0|^$BN%LyJbzc;-}y2G_AIR&6%a6PR&c@Hw9=~Rh>da2J#T<6Zn~2`X=-ifuzS9?zuGNu8fCuY8yiA4v1$aw2YAL z@uho}M!m8jIp8c=|H`a4;K3c?+WvaQL|nE(2z8CgGo%A3UPhq& zuxY|6k~RtYfsK5^1r!U_c{x<{8M0m4^hLwZ)~IUnJ#rOX=WS7`%BC+C_%?m9QG(fG z`MzzC*0D?4Pvf0cf(GyPye>tR;8R|KZPVDaQHbLlH{Uyb46yLo2+x72Co63hWrOxAC?? zI?@WCVsVQcGv@a~jm85Ht?Ia{R<=0m*mNxV+2WbRHkbBfC}hxz|Dd7A7vDl8$vv@hO?trs#$eyz6i5 zo!U)j(H%i;BvEo6*aW-D`G7`xL6+~zs5-hnU<IK^k-U1mUu{LJGW^Fau5fcIDt%W3?=cMFeBHJT;9zph$9 zHoh|3=n50-Y8%DoQe-O?ja-zubO(@%;bv}&XSVwT&wkGpK5KlIG3zE5N)OE3>{`OR zCdXr~%5be^T9Yn7|^LViPH{nu^x5=<1mbkVOT1mtoz;xNxbM3DQbTn2mD$ z;U8`%rJK9&xu}Yl&#~X7%=b+Us(Vwbij$uXG9+Uc0JQ;r%7ax#tHP3IBw#C^a_&ug zwSRcAB&|fq$c5y^liJ~q^uponc7C%pcXveJ7$(l~ciUV3g{We*zIt(ZF1g64%3E6la zyBK!X+uVDCuQ8aX-Yz@9y9!>tB!1S6PGO52d7iM%(PI~El9WMLP1n(QUx|CVyPnIZ zi#Y%p<_DDhGa++bX(JK$NJ-lPMHUp#$Zz5P`3EseJ74w|HqhA_n${R;1JTgMW#@D# za^Q|n>FJ2aZnd08j2VxQoScLc!+z@ZU#j69)6pO8Cs~t7jfoE5NwL84SV%>qHyx-0 z&z+Koi_j00$uy9z}sMq zWJwyld&PY~{`QI64P~k_jaeew%`5QfQdCk;@A!Si`vqoXe`q&JqoFq z<(|v8C+J#<4>%2i%WBvzq52Ig7Z)RyUq<6``*P{KV2kaC4zsJG5}9D(I{UAdXV(A&S%^1L2TzkW|jA481yg zDcMRV@-fK{+-lXJG0JE53Gt#q2AX}Y&=EcsAzO;w5l}iGD9io zPQlJ`A}Kd}9@ac~vmxVh$nC*9Er-&7L?o{^w|%(;C!KhEWWiG2AzmlC65L4d5GRRJ z$Px8E|D8MxC{+j5kXqRpaa$DLv5cQfchmXI8lN-JthHuFy3a~AUaJ;DbU@ri4-&(D z`OLt~Ca)ukxJcxu(?U*RBW$>^u_DgWDeM*3LN~@@yOb?i^C}E4Ggkk$!^@)7)Lfo! z)GF$PxEiwMC97)OXQBePQ)~f6bX4>b@USc)8T2J-YgF@Fm%;Ud^)H!BzDFr@RIF2P zkQ@Y}s8Z%!KrT?ew5ZymbaWG4$FEn^zIj6W0Jhi%p|xHeKQZ7O)8#M%kujsf4hR@+ z3d;)ep4V5~78>zV=6ymzdPc*G^SU3&a+BV9_bB!@MQ$NKa;Na~`M8h1DbExv4r!66 zNGj*$(iksp4%!CQBF)f*0h>Ei)4L`QO-3qQO(XD8Ef1*o-yBlUEKCl^XwWsL6R3*f z$)50DI##d{ha@t+V*Tycn0Sv|b!Sw`N}dmF-ojmXRL$yrZ|op_sd%tR|yQ!J>d zJydkF;ucge?w?m7=v9J~Pq<&0A#G8edvl9_BGc#D$-Wa^NtF5mk#jJTC}NdM5;IgFyH;KkT#)`IDfV^GE}_l42p(w#<-m zkVlHe*}w)@uYeBk!r>~iUhuVT@$9Az)nx;`tQp7~rd4&ix9}D&L`LmKW>8kg$BMGM zuXj(!S)hB+>}FSN4YNdqvt{#&!pp+bpdcFF;{PY2AN^v%AEh+wLMZ25HXeLwG4q}g zJAcXFdz~C|V&!wy1k`O5dx|0_spxLt&Fq4aijWG6a>G(_5LPftiv6{=IYybg#It zFRI)qyAiCx<*8Y*)UQqbC@kB(UA9}MNe}4&vYIYsoO~Bs#n%*y?!rn4&5P%L9+oJ9 zR!R-7(CrAyJFR>v?5~!d(NpFptkO}O)Y8O%eQQ(M%iiT?Cy=i~dYxo{0BSXiUld4I z0>R@|UNvw*Y!1QXQEWTWF2mBfBU4(Yp=;1)PoGs%Fy3SvNVt51ZTE5l$%NqFt>`rG zOyRU%K?_|dI)$48nuS`QW?4nxUF9~qz_(glEv{o0hi8ZMk!E45yj%DXv=H90lP;dJ zm_ap^De6*eB7@<5u4$0B)uFt_47#O3ynrk6yWs*3C{LUEam|6l;A9={)p3JVbiscjR3v4Kg+(K?9h-IcjL5iNj$iniZ$K zlBX3x)EihXs{|Eu2c)=Hg~KX+=K!q=61<@RdttS_TL!ygpM;PE=vNP{z}>Q{P&}*( z)ohRyNV-%hKCtzP9^P;L_l3WbI48E5Ep)ZN z7|~6i3Ec+{6-}w%F|cq_l|#}(f6e{?gPEh*1AD^^Iz|0Lu)RR$R$t}}TB~Os+l`z+ zJHhSk%XiI5M7V@5oEP<$xKMyQ+`AQ0a&Q-wI@#JkFH&J+lbK;OXGEb6w8s~}V2;z8 zsb%IB-Qb=X5%7Fc3Vr=w(a(QJanLJHdC)6|UPAT@(c@n-{|Yb_X7f79khGQ*k)4Xc z;9GFD2qHkK!1~Y|fiB!8*)4t%apnio8#Fqv9P;pprTXe=FshsZjKzHoYA) zLKht~uljY`-L?gSPAKb!){Yu%WM3<6q%qyfu=rqU6LNGORP=?dhM3XvfLyrR4cBcG zBF86ei_AHGg-y3}(u5Ogx6XL@kJW;p!YgBPCNR!uZUiTBF3$SZ4%T#&MXvRcRxd)9oXYzOM;^$ zkvTD~1ya$Grry?7!UhR&#fD8^VbdV(zs)~7&1k7~zZraqq`xwj3Z#~gu+0l87Mhvl zVrE6RtVN!uj1^rAyANEN=tRN_{%!8n(7Pgm#oC8f`Eac+hM#n_R*fBrk!S+b$uXS~ zN_Op1u`s&OxM~tJ<74zm?Qr%Bg>A2Y!HAajuUyX~<=oKXycRnIBuApcTZUZnNH~%xS?kdn+7^}WQ|j3k_0cG_SY@p+7>7g`OsIFQ_2)+!Ul5g z%?m+&t`8u!xAkWSBxj`KvcchMd*EyCC?T89%V~>4dHrFQuX!V_SHz07&|)jaXYChb zM0=#GrY@T{py(4;P1jsjKJe(Jd&L9cT2;3HMQQc?1c?SM7-UTwA2?a3K%OoZ;{~U` zf7sG8W!N@({mk!4$}3}=l$zKkITV{k5iJ#+$2$bsN_WE7_#EWdd7lyMpqDIG3?v5K z=JkM`awmKj?>xZgIWqP>7bX(LV=_6o@NMueEFh0j_YH|@27ie&q zcdYk@{i-$GoKy(jv=N)gbrhRKkp$@cOs<7u@D)l3^BL^Ipt-6%0R4lJ!pGsSmf644 z>c?p|&e4aQmUrh3poXN+`JaC*f^W1rw5rw2{;+-i+lfXu+mL#jDY+b4Dq9gYZm9(Z zfLaaj@iJSD@6F{a|Fz0I9Q=wbBo?CDsUZi1=b59VUa~AQ9{MsizXR!x)D`6=h}sN_ zuexb{T7x>F2j7ZGdjbzZ!|QzYN&ma_c0s%-k8a={2J5AQdY>LgKpb9+tz+0E3m*$ks+3o=jRsM~@%|2B#3L~n5 z)_66DusOzM>BWdbC^*^%_bn$|{PCWi(DMPUp^jfYZA6J90ivHfdh517e%*+Y>tS`D zk;5-pDPxZb7S2-aX^MP|)&KnwMIn0vHLXE;J{b~CX<)uIPMNOMG`nU)LM8I>wL;MT zZy5CNRDA+TZC5;7p@TqCc=}tId!3`+0o<6GA#sqYl+D9%pP|xl2=cC*ylUn`zGZ=E znOy6xo879~El29!O*1haz>o&IlQ{#lF*lSt^~Je{vt{9!9$Fk)FiW$|{Up?h4D)|j zYrENqpl5UrhZ9zz&$7i;+%)Bn=DQxJ%^T)_pv z3xraQRCKB57SGGRrbUC>F80zS8k z()Hn*ZW$7L)R6VzyCla6)*WPquZuABG_Vaa*2!~Bz_RWuY`W_|J|6h9l@6EKuU(S1 zMGXWm7A^DG229%PXCQk@x8jClcNC-kv#flaqmFgtNpA0{)6xh{``(vUn-fbpt*v5# z@d`=CmOyt#eM$X(Rk=4%WxTim$o}AXC7?Q?{q&Fi>S8oeia!6vUr3A-n<&s~c7(?` ziDDBdvdX}Lmg1v%2nmS=q9L~u*G2o(mwtwMsjMnBTZOx)vHi+^0`Sn}f8f zn|@kVsb3PoMm0cqkgL8=<4uNsHOr%80uO>?7s^wiMNPNt^t2sQ;imqX_kkk_t(G(5 z9t9rN(Su@53Gaf~kf4ne`2)OT;w1inbSKl`dJ&i!G+og0q*q?WyW**XLQ5Qp)9O#( z3cSx~t}z&kEaBBc-rRkrSgf1D%N=~xQ4@<;@6S&IAho55Lb z1Ik^b2CDbFY0Wjc)A9nn^YCYP%&EYgc9F6`?me8-sBc5DoA;Txlg5@d**r~#q()pU z#shuQJ_-qFb&)HfeCCjcBTZGkmatAmjo8*0KfXxCjP*Jv6*cjr6M_X_GODQgmBJoU z=ftblC6iUFgQxP*B-eHDC+r-T($ud0#@&!ez_z~V3Vg+ zg<|48kUz~Cl2(g%LK4D|dlEla)JQ&898qQ&t$BiB*r@#>Y37VVoYl^ZKmbKG>0MPPHN>{9pa#r_sN9N_6JLNlg|gz3;e{ zDo=|~(#61|3(S&-{rGbkpZO!spGId2f5xf$N_g2IGqmoqd{B5i3?wA6g%>%YPiOGMX5z!Ov*p z9DVMCJ;y72jhmX^mK1hN z;$vHg8a^Cg+@^`wrNWa%;rd21pj6aDJmCAna|hpExUfK!B+7M%o3ld>_@kz&rgeU| z-RxQzXblcbHOTXh?1v6cdSrsnf}@E+M%2Wm_U|Cuxq0iH*giRGg0U)!g{Hx~spwO% zsePo_Mps2ZYiZsVPmET5s#q2YJnm#0-6Jn$3wSw@c%3Gz;V(F%-WYU^EO@Bg9;#z% zX0%0BD9Tlv7^F761&qtbR7(S^esoE?*=GegCI5`h@IDHBF_{d`tMlARS3_q2`z&bo zMVtNXol8Rh+2#Zo&swog?9;VCim`6;1;Jgvgjre@bd2!S#Duj7%e=9#ZzmAEZ}jUY zpl!O8rOZX9m+1i7b#%YlN7B#N1UBK|`BU2Es%5S7UVO^gv1DmGv4+Vu(J-kLyOttr zsOXas8`))So1&A(F7>_iMp-U+{~sxIOoR6s*?6F2q)4qs%*a9G@VU5TliN3qE6?wG zes`4|du3LhPfT!eo?@Yw+G#4fP_&ip6yBlF2OyKq=fS{W;FHMQgp8$HVVQrXYqfiQ zfT7!KAIbMUq*y%V7Gzdn##0r$HYACUET%|>+bcc}ZOIM$sJ5sZ{00x40Z6&ugy%s zNRY$w(s8Z%;3RBoC1r>EEU!MJ?42ik|5EYW9~w2!2UnU18F1pAQnE?RdFezpX5#32d(NQe9= zY4WI!G+x;dTcKDZIqM5$<}sCir5*+0YXsFX%`vF9bvNuNS>_E*OtE^t@%Q`?%nuBV zAMiQ>uyH+{4RYo8duqM|1x65Rqmh@VTF67*zdkPwkSlWUKiM+ zSfnyaK8$^Q&lY$)ZtHCO^YOdi#l7NbTChsK{;e})#S7MdlxyY?Z={%yDfq4-+=m>; zM#Tg(z=Oj`+U<%%Mgr} zsKy1-?}<$MMgP3=jSI0~N(-esfv|i)**dFP*sAD>%;l$&hms4DHer?Qp`cxPW{h^L zoWFGEc7;BfDC=MP2!5sO1}fIPbLFJvx?;s3@8We|U z=re(((<&52npOV0rXW3FX~b$M0N?EMrMyC+r~9A|3VWl+WaY6L>4VD-{B#1ahYfl5 zo1!b0jX*A2)@FqgMh|t^C=OFdS_x|`tz8BJWB(J3Fg7z051R-lAK+%u(pRVKiZ(f< z8}y0{vUenh)P>DMy_rLLlwv-iNCg#F6%_yK7TH;O>nyC(&!@3J@TgzBXt8ImwoKDH zD+k!^vD366ww`xa(WAV>^fQn%)S|@UHL}$E6zK&mXy{A@?$934Uu}=V;@g~vj2S!l zc%~t?BLrB#B#S{eQ;q|@0x50wdCQ1fP-;4zKao94o{Np{S{aT&$B zZ5UyTHGb}ehF|^0wD7!EST%**bm7qbQnS$g0L9#+NH2u$m5EG}1bOO^u)J8&4`dl^ zv#`kL44olBQCqBXD~`mz^*m9dOh0v@oJC1!V_8(THXW$0v$eGlHz|@N@fOOLQUgGo zuqLucnM0TIugj|{`!I8HT0yZ;A63E z=ne!0*2_-wdid$|XFh8~3p60BhXq$U8S<_}hb)DWgJZJzm~4YEFtm)Q*Lh#@XWJ*X z(LG9?dLKVoQtHv~nJmeso8gjeyd++u%s}cA(ysh!MC`}?-@{Yi`^Q?WP>i9}qm8~h zD6d%Q_fWNqKB=ydA+0Ucz!*fJ^-~8x(j5OcN(aUU!(p2p8|KQ`F!JawiFA`q6`3s4 z`^6LgLRNAsXK>-zeV&=Kyoq8`(X)r>ZHlr^-l}Yc0)<2-2l`g8DN7@cL6*9Wo>w7T z0xc}iKm^!16tr64d|()-k25?*H|sXv{rgw@MKcHCQ5suG3MLR>$Qe=udx&C6C?GP4 zTM@9%KpM>&1<=od97<1Q$?R5T@2vJHo%&(e-7xkRNM<8>)?>Fr+HQh^*^k|-BI|@m zxCN{wno{2sWs;>2f=;021&(3`q& zc*P36DQrhzt=u447@nlsMxzPI)F%6vX<(6zNMzWKuq0k4xf*_I%6b9tCR9SDez{L) zSmpH9!OoYXvk_Cn7D8tS`lt77O0tZza9Lbj*mboMmD#JpGD>VDW4SVnI`%QQ2@;O| zNSj7`ZQA|J*G=x~qIKVYi0YiBVSeuYy z7G<0;jKXmUwoV$&21Uc|}byY8d7S9bBfrH;GH26K0qK|PaVz>aJLDP{jo z`oF;yRiZ>xn(R|imykv6j_a_rT#twsvDh^PAw9?fv3N;~&wXC2(gBH`K`dDPX*YPS zj_8;Z+`#+emba6L$?6ojWt@b1;9KPrmS{(@geb;rmlol5}n{SQp;zl$u$V9TWp$lU!rN{BFMg-yfp5vQ!JHNEG|j<#q4Ne3lg&b8alsZsU;%?qI0-_#DA8O9sljW?c+mpS`{X=t zU%)kzs@+2+f8BT|p1D1F=o$Uj@QjV%auj5SVaV!%ks)Gvl$%NVLyz?L@0kF0WA>_- z$PI2L$#orivcxRC{UycpQKSb=QbFV=Q@8kIHoIAfov!t2ofb_Ii_C-#pz)5icemx3 z=!Og;(h>IXcl-5DHzu`nB61^&cx;KJ&3C2ub;vL2)bTMTvvulE$@N*CAovYooID9| zoYMN>G+vezxW)nAo1=5WYiYJSy2Btgktac^kq&LW8q2is*TJkiHW#Gv+Q@1_mJ}&Y z%lH}L>=yr$*@xhDi5SU%Nuzqi>;+5HLj-DurR{Ym|Y+oPe5a$%EXh4Z~zniAeYr)qM+ z_OKmLxZwO)?}(FT3sy$F(a8sQ|MMcHugOqo6PJ~f&D<6#7d9aW%}huk#XvEG9#qf_ ztaG>mrPC+W$^NbCJP6!?W+*NZ`I=VceHz(c*a150l}hbFZDB;C>?nDps8g4dI(a$4 zvIb+R78dwA%j(#I7pvZk@q|;5^PdmmXv=;gm)(t9X>{7b-!$nu^nsvU&?fIkL0NRM zPt7aa=^XzJ(f1s`0ea3n*gVtE8OPD5t*;^@?Z3y3|Gz(Z*JMY6em!|IY5&UDk%eZ? zZXd<;P^62BYXiv$$WuTrYarB^NX!%<%&t*N`ikRP( zPc;GWyZOI(muzz3Z9LG>4JrF7pcv?m&Y|KK2|DOJNk#N79#rMx<^oHAQHAP*f;j`8 z^F^J&jF_d}tHS6GTOz$>h~FFvS{%MTq7SYFfHl*2SWlOQqQcb&&hUh7wo}hqmK<%Q zU_4HCYFtnJin4E;;PPSafg7a8h2hd}2A2kkIZKf{cMGn21 z7)!_57C8LZ7UkdWDsIaSXek8GWs7zx!PFIt4oiC!+zx1aAcF@SFToTm?zpqG84^s$ zLMpaSzFn~e0*WchkBqTGupq`uQ^?#ggStIVUu91}%C9I1m^;k1$w_g_Y zyRm3r5h0DlOZ1^;aZu0?9Ws{+lT#jUpS2X`eRwmN&Agvsmy}Jmsi&>h#B9rM_EY zK_#eBhGpR!UaHabdbW&NP>lODIJ>gV_+WE-NL@FU@l&s)Z&_AHd_~Iitkg{)=h=WX zF&uVS1D;swh5gv*UVJRO6^4Fp2R%j>#ss7kija4I`+N3JZ~x|T#avc?FSOaP5(q-Oh;a*kSudfDQMlf8HI|ZQsN1-#q%F6Qq_0?U$mz0j_Di z8uu;Nv@AB&&i#WpId9&xR^VVT7LL6^A|DW4a$C>PjcI&>skcVzi z{x~#hkw`JiDYAr$`zw1(ur72P-RoQIhs5Y(YLB0MJvMypPp%!y7o5I!*UhW_@XVY! zmWo7Q5zhm3M}`FBwonYnvTXvYZm5~UeeWhsdSEw=GSP-@&S%~mf|iQbkjs8|g4h$h zYhl;{TPsZBb@|kWe*%=;W3c710qY~IEY9EZ{Kc=epE4PZhh7g)kwcjhJqxC zVpdRODHVr-^#QM|;i#nhKSCyDG-b~R#8{fVr=RuK7stO;m;jMJ^}XMbl&?%SsMrh; z*%SknD(O_59!iqdg(is3_(IrIH5ixQ@7bVel$B{fwVB2`rgamKGwF2Kq&&%unW*2p zS9LG&9tcD@d+B^WPDZpFPPZI|+eIfWSqNP=DrSXOuugtZb4uDlW7-d6>5xx@)ZfDJ zU3^UNEfL*k6Q#QF1WcoOZI^rw%o^ImaV8orfya=vAEGZzQMg}1&| zI18b9PFn}uD;bc1>Ej(1?tEpfN0&!BBv_nnCWgVZu}1v_XJf&} znIFDxICbiS%Qe@_vW>E~RFgh%nK22jGk)i6oQHCxFdtIK(>nDlejh+CN zp)!6pRM(>igLTv!c=JWspw7?%WhYMG)!GJ=;bz9!zG-t*#-A)3%3OAZvC<-j0t!aH z14D6ZmwVnzNX(TWYT&Uuy?cF86ATKCBZXW@}QI2;bcX2wo@3>(MM2RPw_ zdL{L(WJ_Pf3Q5VQ?u8NTEfB+5D#}$i`Ll_WmWqx$BX{+@Z^X85{&}M~{ye8|-gQ^( z>)OpaOKvxpH5pde+KvcXVO2;`qB7xtC$iOI3k7y1xA^qV>h*jGU3}PuQwe++dlg$4 z=bH-y+-yVc(DlRdC>?1;PRJd06VKt4LWF+c209_YQ6OE`hqFAS1>-9j& zE}Mwd31Hyd+hhxZF90W%(>FynYlRGh^>BPT^vH^U_3l03A%%3X9zq1V!j{n@}jmQlX zjRO=XlWYx3A0F_J(0 zyf1rn=WTJw(SisB|!V;CGjQLf*a+f*mnL| z`l`=-(U-Cu2gy4;{gR*g%4TH1mgSiPIrwDAgZ*Ka217~&@4fT$xcAPW?+sEOQ!esv%$pawFyDZwJ zO(yN0nbUgs4eCR)oC*?K0WmDx<>bZ0y8Y0eS9L_0ppvfq&pfh++vdcD7o5+`KvYFB zhbdA{#UUYovZQ2oiici?!hq}%QV+H71D>c#l>vJ36*O=r`ZdZ@gKo=f`4_|;bPf~| zAa6urL_aaronS3&sYfPRD=3aMP~)ICtBcz0m#<;*=tZVSxhFtRXVZ&bN(}-NW#Hm6 zhGs{zu8js^BcaaXnB5lY&vzF*{0|e3{^Q`LH_62bq~B};Zc{?}%yjtA+2il1^pVZC;hUBhpO%j+lF%c&N;aEub-fQvzwA zy()chvoIa1I9QbO)2Ww;+Gnt+r-UKyMkUf170^q<=ZkV^Br&R2pAmP`$AC!_n2Rv^ z!&cL+>g)VY{z`tSN3Q6QunSm*2FI)lx~bhO2Vsqp*+Z_=sjtgd2j@U_bX8Eltl6K< z6(xwShm63)=j2(km^ABeEvM7E9T=&H&Sj;<=1-qJ$5Ol7Wz)@8bh_(RrK(%r1E8YK zK8V@r(M)4_K5Y_^v|@SvE!7z@mP+(P`*5i*nwY)J!Wgzmx*cRo%R$5h`z6fEqR5u%A4I1UO?fN?4~%)3*{IyK!34^ zkL*Y~bt0U@9Xyc80cz9fg2;uQH-!D3^LZ7Dh0HwTC!H5}z$fXv&TBeAP`-e=>C>Y? zmcnLr3g51X;240*V~PO+hP6nQmc~m9#4ijNqk0epQkME+ zzjwC}TO-Yv<;!qW>{Msf{CVQ`otoe4%y*m~Xx9yYEfOzUH^pRW+Fl8}LN<=HC39hK z8iZYk;5eURfO#t$lhD{dj7Z+>(?KIeI@VfaWlg?jdqe|JuVMFCyy(k_X7%N$$8N}O z-tSqf!15VT$^uS)ye{5jDdjw@bA$FUZ^~)twhrx3)@QY)60;Qv<2G<=FG1lNNLi=w zOF_l_3-K-QO1dA)JJ|FU-t-wT@L1<||@4ev(^W>TmKx3Ld|;mt=6;@wl+*C^Pd=c2Z0sMe?Y)TKAIKw*<&8 zjFkw6_z&)PwlMXPr5>NrC-^0^Yu!()AVk-yO!mLyVc36T>35s3LxIBcMS-hHC+xx@ zVXf0*bm_1rPxLU%uAR%Xo_fy6ALgl;4m|5a>^RrCU31s6xZhe0{kN`wsN~Y1O5Uxo z3c4Y-k5sB_g{9ujz^BT<&)fgES?%%8(Uy~iTaIz_`O91dw z&0kkodaGRA5!e08z-}NZ!cd7RKKM9UFKLLye;+BXkuM@o zlE(p$+_ulEmv1IbqA|iaP_=TPc0@~I2Vm{GC8P0&nU>NlE*q`2BDS(WrqZv1J~bUP z7GF%4&E+4??2Z$px}yt5f%`;X83tGakF!K)?U2%qW7O7?dxIXy)ejYq1S-B@&r+j zq5FF))D79;bYv|s8g$?CKp!m_c!XvDsC>U z3okd|u?`8crBe*hwxt?7J<6xu4a4AlapX3TKq#Gd9mErK>hrv$pp)tX{-vOX*e>@J z<@zZlvmxb7oMT@s=h_ZDSq|Zp_bu`-Uiz8|A#+NE9i(ysxnlO!)=|tUiX4Z02nbCK z(7mcm?GZsG5B<^yF`ZNNyuO)-4cZh>&vnyy?GDhY*bZ713qtVKS$Nef#JgbkEIm%& znAs4E_viAKOc)$%=>(nn9;u@bllq8;*dFCdrbe9ra*=o)eyZN18+Ik7k^wqXTZvRj zGc%{zGktbU#t8QRIfv|+E=~`(>pq*`sW|x?%b1(XdNEd_XsEr@sw|qF1EeUh&BZt6 zBq#iY*w8GPJi9FPVc0I7bA#~~2;z!bi`$*{mJ@2nXDELY|BuDRql+mD$-}(mHE0rA zE7%U{1}x0nMx&??Zu2o#s|Pyk28Kn3NWA_F@d{<)r1lw34<>R2)l;MAG&~swKAEyY zIGW&_>sm)*CHPj)`&_x#c3byq8kF}Xh|(o&z7z{6L0oXJYO8j6lrwY2=78ah z4)fcc_Di!dB5)>+2F#>>3huf>w1dGqDUqcx0NbY=YUQ@f+!7=((Hg2t_at}?aGYs zy|bS(py=t_h12WD; z%9F9JL+0`$_0rc(hGfyY@4rRLUD%K`nHiGP6mxvtTjqaS% zOh5|^%JK8*UG!G3hhY`kG{9soZ>@WdcJ7mptT%+7>tk1V+@88D`fThYHyq=U8%AS* zZ5@6Qoq9g+p{kp-0vl=|jE#z{k?tCYEEDeZFiM7EU3~?8 zJ4~m>F_w8}OOc_ePF1I>3fdVCiDrFplB8ALBwjk}vD-GUGc-t;W4&9WOiv@}Uk4`^ zbZ#>;trrOgw?>EU+W@Yk;<{Yq!K8RKNOT(&!tSlPN^<;-bGFL27u?UmHW7Jl#Y zoQtu4?+nD1sywI#4sHMDH8}J)vQ1be+wR}0TtqIu;?!Nd)ngz}GRK zXg`y&S0ih^8r0>}SX|cU(IxQ5;gh80!r&py@XGtHZb}MxJ1ot-IO5Qda8ae5(d732*ZWKp<5!WL)g z8TjLHij&_&&FovU+H%ox*>V>v%Lq+oy3E$BW{QFO(o0m_YQeJaT#;^z*sZuCeIPB)P+Z?kZlSB<7$oXcO& z&tOWZURj}cc6bT3G-!Q5D@1MD6uTA6WA;erX8x8~8;sy6Avp}AwjCG&gBFKw z%Ke)y<%nF?99hvxs)tg!eezHEWd;ToHYv37m663mEdF+fzqM_W9saS;9gvgV8FwJE zg|X~dbXiDPX{FTDcf)GMMYD1FS?>i^l=|R3nw1_`r3I=DzyXDHJJ+FFJbubTj|Se# ziT|6dB%*i!`1T*--g)osxWByfo8P||XJo{sMo2e}|M>s!n5Z-@OdHk&T_Ecxklkh$ zJ%?hpQY3?lI}M4svrt)(rM)qeO_yX#yJkV2co(mcDd%;1fw%BQ1kS}u4@?naf-o`s zbWo`$()p|qoeug^>D-yI?SZvoYoGhr*?$422iA292|oYMj!#}O0j2(}h1sNpn+M~% ze;GM%hWAe?<|suzq2jOuAQy`2=2iN&DzD3ry0gfzfdX5|ufd`^T8gqTVzr>394DF6 z@Ife?0yzxyjXwk3p?S+ErbFpB@+YP!S9n!v_4Hv)semn}OG8T}ZN9ksItQAaP{1eN zq%q_Z;lVkGb)7m{n=DJ)Cc~1RGJbIch-%s3mkh(`s2&L`Ps*^v!#?%-2Npj3p4TiB zmVW+OR3X{r!ZDfCX6UG-7%1>QNW~c^6u5@^J#Puth2mr`Q*MMh-D}g&PJisC7j*|B zB3gV-@$JZW4bd5Y7}ZcD+Qgj&i}IX?Cd@W^2CceX;ipExr(?~BRR!S<3k7Q?r3J=w<*OS7PKGT$&pyBnN1W^rRrv2&f3;Ls zv?5E09cGEl8&eJ@O=ng5zVrnTVhGkNMDY znhe*qa?ul~6RiAV^J_c7R-Be6V9FxVz;B{Fvi}z!A8B6_GY3Oo3GAS7fK7Ex)@8q;$yD+B)$YFz*gH%aWPboB@irTZUePyJ)JMIYu9A! z?2HO1wtk2m*ZuU^+v@{OHmCZvwIw9Wh5a2+q8O6#-c2#EhHj_gwt1Dz-W_mu&Ic#v zT$Xkya`-)w4YAw2l6dFH-M6rWdfp5F-?J248{X;S!igv= zvT7N;9{wswYA*755OW&ns4x?Lo_AK2%ikB!+_GiC%&aX`y+MsnuAYO z+_C&c&nwlsZa^7}v9>0&&&}Y0CO=weW_U_PT|9`vQsGPfsT0CKlz+w|V zoVaLM0}Cg7PyvTd-1vsc2RZtSOP`Y?FBr*xvzcE~Pcd~AIYq@C)S^&Ur+h%!6Pc{V zD1NJQ2e=$M^`_|=?xh~bw29%JVU;o#Blj8Mc@U0ULpsGd5xEhSeh*cp5pW{g2k+C< zC!u|V3%?x*(-dKaxNXM;1&HLYD(!Oqj)x9^r@eVo6Z9DC{fdX>X3~*;C2J#Lk zKq5?tNWcsb5-D!>sS~yV--97VfGsUgLQC27VuD0N3Dhb+Fqh}R+dN6U;SF4l`|+i~ z{*0+9yu~2FcAuZG*{izGhv3tppn7#Dq!w~e>bzCi&M%NHkKG7d5_LkzP9%uB+;{mj zd9cR8QAG)xY4F5RciYY4Si%jkITEr)Jocuj_LRy5s-K-6cb24$WUE{_f4ScbI(ZbM zr=Vy&Za!5+Is|6`Knm7aKr{F@ zuS+~keWoZe)roQ%I<>R*(W8gNh%-9sgVrt3_iwi(TKdAncS$a6XROp+9ST|;jx=Qn zAd&))JW4_)@oay#NBxaI8Nen;Jh^)8URa03_p-fcOJXXQt+BE~HKlvG>=g`~VJf0t zoicG=v)BHR{;;H=Qja8FrQd$f(^?cHucY-;on;G#O{Gmp80v%!8^JjTtRrEBozi&s z?LY9~n^qUMK(_?4h?~pe!s`x{g$;?_rBKXTimamIPH7*&R(pjPcpc3KK(R>@4A5c2 zQjN4qFt%2O1B#pj7b~O2e7S@kja<14FEJTrxL8Xut0=OPin~Ov1fY|-*(XnOnpDX; z0UVeG9a}sYro%Q+SeZE|9-bWjORpEbS#5nlsmC>OruIukr!t3sgQTiE!`g&5h39$Q z^b%F6$8KqT>;@jbdMG#>yF`^8v3OQPY&~pdJC#l11JQU-o3PZkn!X}j?`gamU*Xj| zX7|#IXJyl8=yOvRsS3g$NFN6{2tKw!V(4=7yw{Gw38&SD>dEeX-IBq>WqlDVoE#ZI z6=C39QP;}zB+bHZAPnC|?|psAtF63U^oEyeG&*(e^y9=q6K~5jaRY}fBRb|9Zs1rE zQ?Hq5az(EHSJOvi88=tNg>yf!;~nCPq*4rkejOE8qs|bNMlVw$EoK+g5g^SEk}nvU z>X8&^x3ohzhCt$y{oxS87(JR!=J|Xv-Gq(4pUD@KEfYwE*_T>GF*_(y0G(a`^1+Td zD_<@3Fub@sXM^O!4?aqvZoi>ZuaEi6=b>N#z{LybsmSq3iUUG7a+bZM28=GZjba(eDrk^5rsknGiLF51*>(z=Qn}v77?h3Q% zTy6I(2)v^r{9e^Ud2J|*g7K^53*{g4OCw4>76nz40saymbc-Abf9!@e+Zde3Kz*5} zSi&xdsZnQX!TFIxtRK5)VK7t=XolfhmAAucfIcyYpD()1s|5o2YsxFojipoXq7Msd z=tB_nWt*U@Hbq$mzS3r3$9wE%2>rvafYIJ&;d*M*t2!Xl!d5Lk-5&L%Vh;<;Yk_no zoi3yCcc=vk`@z@jb@>4Z681;#@%yEXSberhb;8-ebMM^#D#m9~(Hi6>LXiO;MLSms zuK16o-}1!3*fQ}?E*`@d){WG(Uq&}uR%*NKDrKdFJdKCj>GPVUyhjQ>G`kiBu26O< zPisGxWd{PsswqIK#d z-aS$t)d@*_wB}77+anT0xl(-Z{0H_cF&D1u0qp!%TkVV9xOhuCFzcYEJ*rb#5%gG` zJH0WWAvV*q8~8M9g#$s`eIKfB%xw0mgQlPDlY02Zo$_({V$Z>Ir5;s^;}DR+_vihu z`DA)l2p4%ju%zFuC`5Xa#JHb7I?HO{&0cMDsp^4d_8#J1P zkn27j#7aU)PdAB^{n1J`$}R`oR_W=FWw*khRZ)y}Esj72mr1l4MO+4Tz_qL9{q!9# z(*pGKCNh_-apAIkhy)FZw`Ecclof2G;)=yJioTbzxD9HMa8HA@y?Zo!U>DNlvC{)d zdx5}L+!J|Y=9*~S5W#LjcGPnWAYk*X4>1j65rE@lJH|)5`F_VT;KHQ`$b}=gR>CtN zq4%9iD6j-EG@q4%M~XwU_j;uJjcM?I@Y%KvLW5V2-A~Phh7W&iG8{iVNxexvb78}A z$INiFP|Q_|G*NMAG*Xlw&(1MkLP za6rjId`7%BB-y`FhMFl>%$RMsF|#`kels!_xtX4C_kOT~NJkwmJ$pP1LvK8Y7ey8}; zFHN@Mw~gCWq{W3-r1@q+{T_+|Ld#An4%dgn;J0G;`WE0s#>&VfSa6agOR3d@PI)>F zlD`-kPi1!SPfHd-dwRa+bIIatn;8@Bg4JSD?KnhknM>qnSCs}6Uyhy z=uS^0-Yt&A3)zhDWl&GQ$vwwCl`N$OyinS)EOfwYUL7z;)p;i}^{OKtnM5)@0qPoHz=l^A}zoR z&g)Sgk`{WOp|^{7&G`&eQ<6f9czZpn=@leHm?=3Y?ormq8egR-H}Ecq?3JWrFi!JhSCh94AFt4=qWh0Xr1ufsM7|2UnzjZvjB>hXEU6k!_)$ zAh6bo9;j9e5@ZhrSlIccGSjm@cB!&Ba+}v7(5PsLh3Jv2O?bwuH}brHLu@xa@VzV2 zENL4#?|%Xa&6n`5NK4@3p0S$#&mz;h;Y_($k?U^4TJ`^__#0X1!WM3e8G4c_W(`G> zsJJrmS?Fckt$3uwmgq7t2)ndNkXyhb&zKVHU&?6C*gDap`3!sRp*!KA9lb?IuMdF?83FK{1;tl19b#lSHOVv6`PR%U89F*H62rObecSbz-t4U$uT3&|LFMw8@eT zpnfyHuMWRX))*w{>_In|du)9|%@%boA7JOT+@4UEt)&`Q{--jj2_lNWzfn(Cx$ufo zXolT%irGYwR4T5Czl_(a?4{R|Hu|)xQ(Wq?%Hyi=x*T^_Ns>!J?Xi0li&O_acPT*W zabHX#ld7^y>60r`z8cVV-+n*Va&Z~ndidh?3$@WZ6tJVB+2f=Tw&Kt@gsGxt`j$6_ z;+||$xK6sA*K@#%ou6ReZPv}4a{15qSw`kuHVR`U)`l}~Ad71Xy%UMe$m_%RXvTW# z4HIMA{M$E&qq)oa{Etx0GG6|Y`mSXJ#!8?ABr&E{$xch+Lk`W_AxV^A*Z0+c9?1Wn z5Z@8D&uA7l0&ukhc>6W_ho4rMw;9soD^lVcdk4~T7 zEqvsLSDbS17PR@K8+y-yoj|rrxh%#w!XvldSxD-at3K-26KTAs)T1=?%8X3UhjQb4 z>=`-%Lc7Iu$!r{NExA3pA0il6W*imiHO6s}jTEmb4eb^z6yR^%N#9Wwc&?u^d=q2a zkG&{H6~r(~~Z1iL4|XQ^MQmOtMyhX(vMpJ6TdPJ8{zfsl(AU zY#JSWz(%w>_*0)h94EAn3#$9whwqyl+l+hF6Uc2Bc5GLg1%n?^%zcXVQ*jl-Y*j1N zSC?thm@3&8NDV-wwa5n>oG$?zDP%5rC7O1AZb0j-#hyAaXr;b@ogMr(;AFlky#vhf zho|GI$8JcRV&q0m5WyH&N|+XSNVp)db=J*UkAV5Kk6sX1E8fa4^*9>Qpui(~{(`{s znj9c7+sRKi6dP`%jl&?}a{+(DOM2jIX6txofs!R9{BunvAAdEYM)H7XE^mFnQNKrS z8O(y<^uT`4TEpK`P2cB1Uv@VXeeL7N!>>}Kz|TnH={*vejBs|LypDF+B5!U0?4?4# zD>4D_{o;v#AuC@newRG6-PR_GNyQ}~ZaFU(+S&|q69xw9D*>%))P%iHS59Y9djd=T zQGOeU9fTc@b&W-0=4HquPv|=^T=GVy$HoleDtFc@ z`&`$3{K_Z$hgtU8yX;oTO85LlrU$C;^ufvgkKI!EJE1b}kpjf`ZcSeLl1^Q(?g5d; z^5Df|QN23|A#R`@yvzANxq)`+q`eB7pO28wkS{_ z*)C0tT^pDfUO|_MG0J-@WMeqHPj%WaM2mzr-QL&9Y_r^c?wItD=R`#$)2r>go(y~>toLX6U$^;i*1!A_mLAuQ z!X@YZ@?R{K8eMkjv7)_*N{gtIhc#Ydtl&t1?X+vI^nxgfZ73n=Nf@sfB* zLiX4V?@VKKUPXbc344UJ&*-L8-8YKt4f~A({-7OCo>fQm(m<&uEMZdfH+B!fPMT~l-Da)K?uMut_LdLit+ z9Jal3Jl1T6(=#vpQ|6bfz=WRQ)F0H4#V#EB%`(HodWu;`k=0aOqs$OHDG%B&MUMq$ z7X_!#K5Yx{C%v96W81*UWvZV!Hx53@$*he3@%a3AEn{w0!dNNct@L~#yeI}jsYT3b zQ6=w@+h5D*RsLV9ob9I!|B7t>_J`l<)FYg}eb@C|ZbqE@!HZr4RR`seP|Jb^{AK*Z znmRdJ4G@EdTK*2YkCZ{Cn$6%DY5lWlgC~NnioI@z&9Jf=x)!d6%^zUrg`A)}{!g=x zeDH6k73CMZemOwuzB2yE7iR0vHHx`Rkwz*mTU296WRO$e_&Z>$EAqdg?(NtMcj4rmZa? z2DD-cUL9fQgxjG@U#Cv?ubbm^(_}FnHsiwLK4-t3{~tv&zHYKT{}|#k*nUn?YrVIWfP58w_v!DIiK>0#&AP{WX#_0XSBNb_VDuW+O#Dh9VStIn-6}fs#2i z>e<@e(;JxWV$|VT?^O$F{(3dmEny?PJ~(aCsj+Il;Qq~W`-bejnv-MUx;DF|ShvO3 z1Q2cFvU0N7h4;`0&5&D2F_47TQ*o&BTraH?VmEDrsyGs*d~qYK-tD&(Nf>lM^H8B! z$guT-eX?YK%r_YwWb9U56IlyvG7$bVs;3_(nbSa@#0fq4+<|QV;19mD7qHgu;|Aoo z`wv=Pc(y~3WE_mO6-QR`_D3!CNKqn1N1ow$a@5|aWdBc8o2KSZNvA*aK^D_I35xAG z-&e5{CDta_VaGY4WPIoDKmYMHlT~@Z&-@uFP}yNhb?`rcFt1~>Z?A33e;tDKKRHk zeJwlTH)xw|{K*}afu#%nX1Ppp*)+K@psXxvEfKeU=63ll4$!IhNcTvp=`1=&TOWH% zu)+iUTtNYOs}^ztJ2h9OC3LbjNwUXL zGA5QYH+sBSj_8>8w@-ez&r)sNWs{Uvl*JomUE(^GPJKHpoxUAr;QUCR*d*Q_u@&|% zSkdLM1dnasIhVP%ZA8ah#c638SDkfnLWpT;y8gXeMI_gS9o}lQ#ioQ}_EMynip$^? z@Jb+*f^09XN>uDc37cD!?}V(Lyn1q$=l+lyb)~FUc82Z;g^Xsn<}f)l^{jiBdx7AQ zTaNa|Or3wNY^V3xN%eA6qsHs8_xm`xF|!lA;R0Dc{b-gWogFUNdF~?O7}yv_;e;LP z8xuCZZ7IdbWfgMWZ6L8C+=x^}{p7aXuq{$v_4&-FRoS3w3We1L32BD{ z2S>u?nKR54OV1q4fhQlle`SJY&92L?IaZ2^^EKNeu+A=5YS<6Q$DkNMuV*I7^Ur~* zmgTXT1QdMFK$Sm)fu=5$kIkfWopG=r4xCr(LvgYm<9vVr?gdNh@j}&QTCox7J|`pY zsB}ss51V_BEQ@aO&lUlJSQ)+5i#?%Ep2i+gWP~4401u|3n*QHDp?ClI_8;QjdGGDG zzr6FC-@g|pr5QIuQP;kH%5SR4FZpi%FWx1aT-Yx;VCI(;Pz*pLhl10JG{fWw8Q;K}BY6<(pKT%x_cG`ks zhcU76NUR%#(VXFCOx!=esItr=4mTz)yzRA;D=hLsm$lvdvM5Wtjs7$wm;cCZo%_B^PiIJycr9VobdgV9K#jUpc}rCktOK>!VxDu|U021`*7D=cFPU5uH;Hy1$sVc0 z-Gw)|pP0ETdnl%eB0Er6uTz{PND8fd#UQ13I;hmMR^FpLszD;&jtHIh!&Q8?N0|$3 zctvx1g!V0=k2eH7`T;gp9@b7zBnU=l=$5VIok`g~4T|z6n1*tQW@#^fh@Vyw@!`)ZS#hH z1Uj*7d`OXF+PD5ZGhrWdVLvc})^t3iI@x@>S+H(vkTvIN{^{U9r~ z=?f&4eu*KMfFWEDHIxfO3aG7K_uX53^z_A7oVyZ@;F8*Eenz;}sfVnapAnp%adZE5 zdyQppEElK8g%>$1U9;De*njq`!@UIwdMaM!q5ju6%FDj>(js@wey(n=kCWepphc7>>u%zndVoO|j0kXn4j#b0uQJcb> z!FfLr2&;(Ek=6&cGLT1PiX&5$cv`1UnfTakQ+N_@yEtDq7Fh!OAToSP9ew0qev)#< z(si}sAwJ>W8l*Hf?sT9Jc+7HhPyc$Gzt5+)jL9ob9-`~6)#ra-u-xCYn=W|#)?SkF zf<^kDYRP-67<@~ZU?l55m{P2ePbqz#9^DKEtJQ5 zu#5e-i;E-i>J;TY1$4{1;mLIqJH>2*NKa!SWf2cdOao-94N7W``{$SAY{s{RGwS5G zRt4}@^cV5#>>oxJB>d3p&p|9&>yS|(B zdeiI6UPHc`42bNk6b!@^M;foe-BG8wk6xwRthxBgW)0qLEZIFHulHW;&`PL5ljjgL z4G!kOpPM5O7h48x=!U9SwKAqcTOW}g*y2+kv2RA3>@ca1sFt30c*?*b zxIvTTK+ND!Qjc%XpJ}o+@0Z3_k^&d5UpQuFJq}S!2}Sl&afRMXgVyuUYifh`@rJ&B z%5s$;EAh0(b5#K)xr01c89qIcfGGY$L4EACH`3p%|HheL?0aJ^_4_^lTKUGEZ{3-5 zHr7!QJcj{daGi1xFsv_2BLwECAMU?%%5wMYvhi9gQIo2m0vhG#N`SR{Z}8#SOQDx^ zr{)f*Jf7zz1z{Pk9w-{vTASTwbZ*T^^u4b~>lZj@LX&_TyJp=8{z zKiKqmlF7(4zI^_DlKho%XdopyL?t+jVlpYBqvGbm|Iqw^N+NKnXVd4&BZZz<1wGHG zbwBrzu!}$U48+}_W+tjlmc&~sYJIsWxIL;>c{lQ&-{NWe{hWv_yFq0zH4Z}12*bRx z;v8eyE#$Js#7h4Ubi0Qx3^=GsjM>Nk405BmvMi20tJtVHuSw;d=aon*=tZIq(7>z@ zozJTx&=Le!;HzyAnZ=)YWg4_P4)D{R)&4X3J<$lW=ZHg`%nJ2ydOshz$*dgv_IECm zwJz+t>@+hJTPOxJk2X8Q^2TX)&)D=qbCU zDd8Q8)$VNH%Wb~cPh#YxWSi*&ft7x(%Ior@?n@m?k|JK4;($LmrXcqhA94gpHucgx{|bVn%coT}(N%)0 zV!g*BH&l$rpF0FK(d|)>-0=G{ zs%$tl429Nh(nxWcf(dWdzUWPj9rS7e_`*O_xZCdzoeA}xSQ9F3;InnYOwuO1<$Z2S zu@d!C65KcYq&V9hHqD3~_!u^q&N#pcALA=toe}hpm0~)m(P^vcPUR*|34H}>UlvYT z0!+;fu^@B`1u(5|UXd2eX;B`Wb62s;XNxRPat2sA24aeM86oXa)%0c2a#;zr=H=4R z`q=y5T}OSe_n)u5*5kl%@DNDYfv^z_%8&sa^zt|9g^XpZ<5wibWTi1O1N`F86{Qgy z{ZoT#g^OqP${wgH6qUSEk4r(_L`TY}=dSpTM{k%SrfD$8hy(kgb3 zCgwWYSs3979&*V3so}INxNeo&$;DAeEc<3%){C~%Bb)5sBHyV2&QouEwT)g9xJPn^ zUK5SkCg&6@JPS591NK>iIN=;8z*5J464&KrGBQ7JB6Gr z_w)hpX0JN_yuO(a<<2g* zv>78qOqKDa6M!*fNKP0Te{Is1b*jOI==8X=B#m3l#&xYC0dblk*~dJJ0nNTFDy~y` zOL^4YfSxsxW%MTRlJhsDpl_NKwL;2h{Y^>HBm(y$c+t$6@SsDDuyfl=>ck!hUEYLU!(lVRyr@Y_mKn zPtpNhBX!dDW~jm^}Dz15k5gUl9~*;>H{6m+Mq@Di#BR=SVSBfTBDkvUAtT zk^a9d)Xw~b1Mu+#DAry)3kTNDaWXyQ{LX(n(=z#FCC3vV0$uJrWNTWbTBJhyx83|b z0Xrw(^<1QK&Le2;PlV&Pe*4cD!YQXXefzFk(zyKZ-&|!W1IuM6?ZR0EE26PUl9e8% zf+9@|q_klfs#Poq?2djQJ*2ImS>w4@CGMM@?a@3%rL7PjG#*dybndHNVuLG{a-e4VYn$Y?xx_&ZU*nIa#rB zuU>g|>a4*n&Sz1DWS0wjB&W>)QAsh-J#vtW+sdyF#PCjWWTs~i@Im07217TG-PTX3 zh*&=j&(+FvwY@XbfF5^1nI`P_Gal1L9f@ia9+3j~zsE|!S=CX$;z&HBr;{X)-O{1< zAAQvUufrNp3N{3SU}WB54R!@M0(CZEXwX8~{?iIBxE$t&q1c%}{@hahmCJ(Tx=Sgs zA{9FT{Umt-rOau``?cZ{&pP!Vo*qe276h#QC!PAVq*wBk;(aaD>fMpHD)Iu5+#Si@ zjW<4}c}FGWT5+4ORdHHU7z{MtnoLP5IWzT;6yIYvNdyS)5>n*ch`n*;>>ES~n5+nJy81@cpo5z*%clFC6e% z9yQ=~!oL%?%}G3G#j1w=$=s)5*bq)Rz-by>H>~;VLw`H)Ytstz;i{!Clj~oZfbC*4 z9<_doxl565Xr`XBY|6Umb&v{b;GLb#o)K^5>(tm`J>Yqbckz|8ASJ$0ohDnPN)JKT zd8bE*)}-r+Z`pX!P0w`6O;6O&Dgza*B;HNW^@4@+a#8>^Fc8lBX>VkyZ{N(!X(&%V z7|X+)DKa}O_1NZhiI?ZW7Ax9;@2p1(tTFf^8}w?mcRXPmhssqmfC5 ztqRJQt^*VJ_cB`NpQ>WuPwu z$Yp8NUfxx?n>InO?}z@!NdhEw}3X9)!{cf`N;UxSs(q*1d89+{XCd_?!r6pVtMyGPMQUr?^|24&Bj@U33^WNZ?z(r>F*Bl9re5>HQ`mczI> zcp|f8YB$}aTt2%lM5i`>N`~OUoH8MMglwPG!v}2?-r*qpAf391xlh-Hmij`But}K% zHqm%}nP!iq)*Ua!Kjb_R7j2R0)Q6@XfFOW@*+{$CKvrZUpoV!#PCoDq>W2|(Imb^) zeEGj$H^Fbwy6?Y5%13HdaN#X_lbPi?O))1Z@+lS9N_y#5#Z73-Ul7ucZTr&>iAOpH zOOD~nX5p=nMkbg4MMUYeOF{J!X@P5BFQf4;sNz$0(|s^Te8{bkP5z6>{%QRrS9?a2 z7WiQqeaEvKw5jiaNG;x`QyXu|rVojMNHlOhb)HwIS68TMRn{W~Aw$#z{2JoHG`(jS^kj&eodx5#N1Uhz83%*z#u0j?-s~9u2Qvu6zx&Jp2%I2FJ~+zmmxNYmSccE&O625 z1Mhn!xx8XJf%;&{YdQR$$VYAmHJO@x-f2lOy_S^Bo=<&nTzD9mo$dv@=8Ph z{=w&li=RdXYkt-lh=PBc+$#mhUs4*{4L!iEbbZubk9*|#Pmev&cff-7+^2yP^r^D= zxN#yA!rm{Q_!qL0o1^Wz&km4d4K2#qL@}wjHbYVj5}%E-dkQuw6zQ;#IO{lB919J- zi(^ZD=cUs~qSYvi7p2DH2G0pv#>3~(J}?a*$I(Y_-u<`hJ|=k7zq5B9Np)cZw9jn9 zb15c=B3p67QiO?|RTCE}`nYA4rqnSVL@?W((8eGr-T~ z#Z!O&JNps*-|*r?c=7i?e+bN25kqfkh^-3JkxFR$t$+*{$joXlOIv)PCC00SN+#f3 zgUh^9ltbT~vPZfsrjO3m9tDu5YbtGQG9MM7>|3Os3nsjP6QHP`{mgGCn~YDuZ%XHq zk6qaK>@_n!+b9NPX}6&FlS+DJnUWs~jC88n(2bQg{*Td6@` zC5RX8h*=Za&Yus}=BPJY9#udUP}qr8sVoIG^gPKqs19ybmii`$P&yFLo|3ex(*u_& zuSie8eR-lzXND6-1uaJtF%H38gY@w?=9!F2vEL~L>2P7AvdApne2-#!DRPI3ONN}R z9vZl_AR*NP1>s#nC>igDAn-C}8(k881cJLQuybBEv7hMt6T{Ib-m6LiU7yEpX*^^* zS^~;U8KU+4Rz=q5#o)Ewa-FuEu&L_KunhMyuYMq&D1$gu3siz(HY|Qto%~YJ zhL;|@mGO<29w95ejUNmj_GFkI`C(0+ob3p!lVh;icuPh&`rwOHot`PmTY|$H{2kFt zFQfAa{1$5Vye0=q-LcUV>w=z^tBwv!BQ0{HJ2nU>PG)Kx|71g<(1g5K&u{pIEdR>v z%D0&mMsA z-yeBoW|_ExKB(EnZ$)7`St_nrdrCauIbW2aUJ$ZHt0Va8b7g((y4SBk=CgqM>}yT%^K5Jv zu*&K|z9+|hVA2_JtsIMzFOgQ|;N9u8p4Sk&gP*}$2vR?}yads1#d&eRXJ^OD$On6~z^aXjxbaNFgeG=Oj@{;AtKt z%(Hr?eVngWW z*Y=SH<%2oJ!krNHDe+l6nrZr8c6=R=e`D>qdss2(6*hK6D+6&Kr-nCs=ge z@-?}k#{$d3qvjSbfO?z?D7T3jp@m{>Zg8%oi$MtG(?uVLt(_MSbWAKZH|W%^!ub)Y zVMl$^d~rTAku~x57|J1G0ec{3gaIFO)1SZHdELouk9IwL|1Ttpn?2(33Lj(&hGdy1 zQcOHW)|WB-MvqTG@B`(0|L-*ih4h<86V^Sq#oE_c5o+bb*c{(Nc<{i*Px zbQM!ZCxia=b-L3(eP+Eh$C1&$=L5(cPTC*w;)J0{_Kc7~odMYd>xNeX5YO>r zwx-|hqafVFSV6j7*Ur;51I2F#V`q)X(cs3?Fg^b?WZ1`icqSr#Ju$l?rA`?iku^LH z)WCM-5a(hu#UxWCiPBz&OrLhuBf2XB#;#CY3erco7}lZe8wOT7Xb1V1EKi@)ShMuu zh-Pk+PU6&mekd{{W7gLHdxEUyF*33(kglA|j}miPghzBD zIrRFtC4reolLIguG5)N5astK#@r;xLYefe2xxJ`CkvFbUH4b2pzP7RU z&Q!C}Ff0tYO47KkYfHOlTRVVz$z$*(&83@E*CrzSlPh7^<=%*vfH7Jf=!`N zr{%KbPy{Blbt|i9=<58NWp_NFo;tRk^=y#m&?kTtpF>BxFT*UBPM1o-N93wxnf=$r zoN)Bq&#*pXwI8EBF{QZTo3Hw!w?O<6=#D1{Iwb|9JG@HIJfgn6UehzqCVcnE2$A?@fQzF%k836TdQwdZ08h zF55`dwDGeJ&xgb&Ogz7Q+>SV5f~tFuWc!)z#P5r1)|0GBq}t+jETb5xs8)o_fpmKF zYb$`VaoMHH_2NR@gx9NgiQ3%r1;@#8s9dC1V{zaDze>q~(@~!+_W`Gmg{AZL>JMOR z6%(Cn1zR=6?)PB-{{vyYcWvM;*%`9W?_8MuhtHE>GJoAGuph9!fnioWJr4r%yNL65$CRb~i7m0t{l9tm5gIwy~3S|af% zeL)@+O0d82VtMME1{`b|OE-3J=4R6x#JTTV*B9VoI(WRsWvNDh<*bc0-FiW<3dLGl z<@qwGHJUzozj6oUq>USZj3dthMRrHZi#pu>Z#!cdvm@g6sP`h-{Dp~q?6dHv@+b!4 zt(lY-CB0XO^JNhHzz8b}xn{Yy85KK^`aB3hUBf}AQvcJiKyFl&g>}#d+TdSA+JRY( ztjfZ$jsDMB(Jy+mqddm}DdXTdJ^zoGpI-LnPs~X1&%EACTD~wy>9s)0Es8Nvqz$5E zF1k7z`9J1AjSAeIOCYzxVyLB6u61-fly2MZg@Vs|by-;TjKcv%WLdKSt6S(A6_+Jg zAhK0nC+L%>OkQ?agR~+8~g! z3uKZUx>RsQ(4Z^>V*VAsGq56PoC%rcG02GQcV05-4485gnL(*9a4fk$(Pc$8cmfxQyX9jCAYAZ-3{;m&}%__3hiaWG|0{ zZcP@priNl**-D^&w4;*BHGay8&Ya3o_Dx+uFv-MV|29zy9P0V{wIeY;{;%p*ukh%6(+- z41=*moT-o;0Kz}fxmB`@#y}I!Y*3087%R9*l%q)#8$b~jw~q?MH-+7HQ6Og@%`xDL zy;D1!+4<QZmMyuSDQvxug(Y^;vCrFIoZ~QO`~jES9ZOMll7Q5!Fl0| z-g2RdtUva?+tEFQaE%BDXf{Aim+RP7#H0>L8B|?xTboZ0j2h*--1GIx1hSIfxay3 zpBzIy{a?B4j>^Qht8hsP?GnWpuX!LVfL!bkNh2)rKpZw-kT$g<>{wu&Xs^1CP8LEo zTxD>ZYlAay^UahuetpM!?&59qk7YK>b-oGf9L!5aI@}?;l{o{BPE`07A)Y~HUes>b ztE}<9H?7UR7IHSah26?}?@ytY7vyI^Ayvm(FPrSt2p$YH+0`ytcgD$1QHMHHuUV&Y zu;f2r>L#ih3_7iIuZ2_@JaX(Kw*F5uf41exv)t;?W=>C@zwY7O+`q()B{t08|1wZ1 z8%l9Vp_mO6Sx;#zgBxJsoF(j_610xWUD-4c`c`#&-8HmCXw9~%d^uv@lnoy2o2d(b?F z3k$^-m0s4lBMIt;%D8FJEySR(Q6|A8Xn0=GDnIT#erCgFgJJSU9E_Fc-BaKB=?|@w zieD*%!ICI*ArwO=2i1|tGAyg z8@a8Td2CHeEL@K)ihXNGvZjpEVqX@ z(GTA+{o5RQHWJiMC68ydzuppgi}cJ%gUW4DLAS_x%~41vJ+FEAMoVN3xkWapqU1*d z69kS%6DI@>oqrq)pb4A5nY!jxm)pJ|flBxEPU@ES^TkH#tizf_5onm~_fD1H5p4|C zeWdOOA;LCSEL>k8z3oyZ$NvZOa?&%8KA_OGhZVhss@o1*;f{LrpCQ39FkrVIBamb# zL=8l~Ze8v13u03aTlgeJ6az%eJWAUj+%#)Ixk+?0@={1vU;+q zEmu>jTsoWQ|lGauUkDyukKgWO7DmqWn;#SF2{gtf^zWde;SJq z%i}qS#fQz&Jfc4YL7W}3HrMo-MbmQV4+R^8O)DIba}VivyEipKuxW0aD{>GzX|z%8 zDwHC06gHj@?I=<5yaNXviSO)P<7{5()P1Syze)5+MXq?fo&hcGA=zTd6q7`e1WJqY z*(s`hWMi;io#kF4xB+~)iwl9dgs!dQhudyeX?21We?T35OK}KjbxU%5OXnHr zC3y7?kE;M%VIAq188Gm#LHYv6BQUY2DgPgWurOFt7LgQZGP7sy!~qZ^SC{K z-sYO=PMxZ-j=AubQNR*k!RS1`&lrb2M^*z{wtq5I_1#66Noq*Cw1)JKYw8#$VK-}7 zIr1|n$WB=Q^=~x?nWwB<-@a2y@_9UE)meBgH z=S)r4Tx?78ecC$U^z>+(@AKm;bcM^)_C0ig@Q74G;=I+^Q zn)SerVB&-w$}jl6nKCnWlz)4xnXI1#InF~J%XEs_Mv<-PRkW)xIBKN9LuEOfdZHlp z5B611nDYj*k2sw|UEU*t56eGas=lP6NT;b%uD~yKh)KB)pw> zSUjcE66C5?(nYgsL`5#0suCdF>(w!&DL7hGPu4q_H9ZU+wwsP&w>q?e6Jn_>9rI)B z1vHni9DgAX<;;d!z-LQqB;cb&HCVhZ#VLYytD4o#=3@5toLg)eWt%}{OJ{bx=KUpZ z|2mc(ARfblWe2F=ZM#>SE2bfuh|HLwdd+LQS6O%m{kH+9D5_i4u6pQQ1Gzo8h|Zg~ zd+~Z1!N9Q@C*wRp%2ucURblo#PJZ)SSI7n)d##}SG9=cvgJM9cX&a?2 zBxIq1n-=82+B`mVi=aF}ua0xs?tcRQ?)OHTQafb7J9x

    )Uf;s9wP> z!5OkweIVe-T)nyg5=%M0^`m<5ql3e!zKYXL>~I(vUG_Dvb9N2>a6if5w$9-(WvS9)owJ8xV0)o}(zdy# zdgem5pI&`dab8(PpPio{cwbd1JVO#i`(3J_9$;-?mM|lzN)Fm69(96ybMMJpB9ozr z>8|jT(@(md7bFB4NOn++Gm1W4kl-~AnW{KIhIPF%!ibN#>4yc$Tdb!S7Q1YK!q^bA zM4un26^B4{=Fzw~ch>)rwmnPhyBzz8RJ0Vpxsx7aYR7I~6 zKUCbFQ!mbQsTPz4><_5-UOfrxj#kl4!51VAZhdg=D)EVc>mm2u9AYahkZK#)huI8^ zJ{<8Ot-4fbO(fysgz$K-VWFM$!d6W>z0NzEuAAKkseeiC=`&lDM}73_6QCnfE8ih^ zJQ{|MBnM;+z0Kjrk8^JQW`@}ddBU8ECHhI^h{g21n_>znvXjyx?|i#g2T%+1{R-(4 zh;~EJ@-XO`A?f%yJa3SUITvh@gL2xmXa#(z3`V&;qzR;>qRVIWM+A%G;I3>5eN0qH zJ47{_!;>ENB0g`h|AXf=;qZ4yEH~&^9e-cznWK{3z_(l>PM1O?%|KTH@GSdb);2Wot}R7f-Lr2AD%&j;i> z_qaV#>ADpl+^1USzVr~?6W$=d>Cz`}b9Z!*b@V3C{$_7zk0pDAQmJ{e$#mUxog{o= zmb|+x=1e`sq)}uGr7d1?QuR3$`dCuI9G}r3S+ZtYRM0Nr6(~BZS647)vtt6&LR%u+ zRZC0?I7^!7XuC2Nhd%uw$D;$tkPaMl*tR6=zI9zMmYUHyn$!j9(%W;^L+O>{0SVIe zs&xN4x_HsDm5cAZb@QElZ*PAm`JJna(*4KJI&*xg4tj>eyXEA0@YmH}diU@{f!TWe z{=?oAWF?Q;BbgTVU?asOQY4+bD`9Hkw_H0O}$Iqx%SO>%@}#}gYF-blV6yn$}J11U!<7x6ltckQ9=2B_mtNn(*2S| zk7X--F9+uf`bbOU32CD%X7IK_v6WtueZ_}o`i7l=STaFSiCSce;E1} zy!d!e?bn;9n8D@u+lnP*3y(cnNY)#|gWgFo5OLo@X&V(CFn~yIH+*&3Uv!y1ZLVeE zMN<<+TYm%w5lLokbkWomP~-xg6HHWw*x9j}mmw_&d1l6HD(OH+bC{_d`{2KL5J!8K zlYyM@cIk|%k!GkELZ_vY!#sw{1q-O0qL>pDsWnymY60#c>__jU^9AQX7@)}+2#4;g z=VFzWUXYA$RP?*&(GNmWjHKrdC6Za2J(8ql;g!J$<{KpdNRMli2NYK&mjo3H@$XLe z$^hMU6&7!LqA2k?HWPT#8-k94DZ%Uadn2`N&?!gL!W<>Jp$7V{B0)N8w2d82FS8?~ z!xJYoP5h7P_)G;K}_yg)Yb7)^UE(3DFtAS9hZX`9uZ^plVWA-V$bfP4_Jgo_;% z?=#cruN!OX^bw@D<8@fv0E!NpVzeM6;W&$99KkUGWpo&@=`L&~|~zI}o| zG7ZWk_%@8kF7oIQWfLgIbkW}?b_rd~#^gVB=l(=ucx+5^EsV()ibi%yIBpY zd`VDh1ebB)Xf$q#j?LdKX^`y(eH>%0(s4=NG7kmE!5-59a=Dq4`IzvVdmWwD&=?+f z;jximflH3~syI$)l9(=5o`F;vOa|;$WlYB0DoCygyFb3%XD)EzGTdXmmK`o%@7)$= zO-15uW(^Cu$V{$8!OvB%6$j~ZffsWopeZ1pNeNvQIOvCUNI5K1$NI2sjpUm=&@q3T z?H#bAV5C@3Vd1aBtO*3ejmE1-6DajEKxO0f$;gHubXq$(oh}g^cSfagq(KD47@5)g=o+z+U>UMxS)rN@p!s=v3>AIe<-S7P6Ryh& zsFQviwsxMe=K2b8lo%7eM;#G)`EYWC%2N~b?C;CJVRTod*PRRbmU-cFdv5h-V2k<*&@&JkM-zPofni#chkAP8KM;;h_Y)^J&Wh} z%()^xPO=w9!TPm$KKvEmXIdgtR9V6bpX;id5FWWbr+w~ml00wq-1F|cUh5%vzBVvD ze1`^#lX|c6j*Y;3DqPljU!(6c(NH8IK~zUu-+NJ6T0TDq$d7ww<D5?>3uEugjs&K|Vh5Y9d8N}pT#8!oF>qJf znPVhhmPhZ1gobR8F>hCO%SwYvgSw%hEHErmR8P+@Lkat4HM%{J*Vqi%Cwg`7obvhj z4Q=j?LY;}7@rh=?Vi}eg#&**VWs7&TVz(#>WTtVyvNAY93g7m$TU1SHXD&;iwlJ`# zuv^?vST@o$H5yC}ftP&_M*t~1rbbw2PuSu)O*_$yv*QlHqqbzJ(nnvjRWZV7+!?S#)f4jtOj6)(8tD zYUK@(+mJ7aUC^#9@H-%=gUpyd$k(oLIqTi&-5<0GR;>+^9(m35Iy%O=B2cf+k*{_u zi)e6IaT4PrwB_luy@-w+1NNss!pp~BKj4j^rjM*kkMcHr%Ti3VRxmjCjOMYdRld$0 zW*97!(X2KWio_%@-A?yUB&qHPVMlqSIlv(p;R;%Z{D=`c^tU(MYYcV z8>D^`>9nxbS1INaMJ`ZU2vxWt$HPG9gYH;*P>QNwn2KexP#NB!G|*kLl+bonC4JDX zMVYVO=Df|rMCL$=GQE1}?d4L#^b}Qvi%9{^K!5HJ1#xIYn9+N{{k-R03_9WWwaWL* z(%tYfxL)uq3d1bELMRE^=K55~vy(=~K))MsluR{qax_RFfEmT*T5#FL7&tBU zN{M*&Nsg0df%`bbZ>L0J~o zL2o4$!5PL5jMO&l6#4c?Jre?}XkC`9U9dtN?QRULp+xW@$ci3|@o*@NEP%o7!xscb zJO1h6>Nm_V`swYaG;)X=xtG5$2x+l+8%|OT6kn)83SNUURu!vCQI+~%65O4VtxgC3{m;~d;t~~eA<_&9@`u--~1q zXSXYl&nTxJxSc1-J_!r-YK$gs@jM+-7;%bZ`m{u5`V1)3eDT}jm~8=<1e;$w;(yeK zI|fIIDSP{4!(hh^Q#*fuF>a|DQ)M&GC`rdiJ6Jpp;jFT#V)r@4^ibqJrQPI{DeZw} zKp9;T43fl}o$ehW34;BeeR5r%v{<;xKRy(vo^3PBX6TZA(J|jPQ?IV`ZP%ncgz?HTs8X#70r*Hs4sb_%N1HNq}BN?Z|q8)jrYi@)Z;Y=gX$-ZU$YYuwdg@EK-~xFLw$INkom z4-~W+SiRrzJw;aV*wxCg0MQ1DSx=ETO1qOT@+^ZgF`${7?3?UcAjLiP0`a|Z@VBjp z!A1aB-{$bo>;M=cls<7$koM<{85C}hdM}dAlgK`cXP-wgAZ46MY0rQ%PlM~{@_bkW z*9tO1>*!2a7B#3pmsbb0tNyE!u9s}{i1urV%p$qM@)-k8_X7_u`0$MZr!v^}sF&u- z5*(X*Z{vUrd%v*3k)y(CK=^Cd+rM3R^^3^SSgv^@+UZ_Cf3IK3H?PP#l%}QE8QBFv zg5Vq}Sp+I(kjLHU7SG`4-Nw&j^*wBY+i37?Q^#Yxh#k+HR!$I2FdGw1)`35hCr<1a zk6?{>961ERvLT_Nc#2s|ku{X|WW>qoC~X0YBcFbUE7!3N+{=-|cBf<8zoPivMr(l? zF05l7FNav9XK*RpqAtP3GLS_JT%ZDM0j!4cf8$~^hK}dUQDK@2hu$+D$6GEiTdeCL z@^+J)rNUt!iDzmcwH||SN2c$WZwzi%WiLqbxEg$ltQYQq38N#ZU6mlnm97ar;8!C$ zElG5)7U0?R8Aql&8Va_K{4iI@!NbpKGN3L9e)%g;^K8=k`uRVSWNyI}9?vGF7PCn< z#eg}{Q`(J_FT1P|=gT_j75;|l#{xQyYS8z9>bGoXmdS9ieUfVSUD3U zLel9rbrftz{k?+fOM@X7UvQ*=y{US2 zzk9v6qepn?+t0~_48768$DB;a#C0nhddFfy_`4b-EG7iwX_tfhAPdFx7qeHIJ@yGo zK!xh~j0+Nsw;mAU6&+!E4Yo7|-`^}1Dfrm9%m+J2@369;&{QC7bg+DW6{p_G}-;8;4|C?uxQh6oM+Se|*_N{9_I~xhL zTc!B@FEDSTHB3Wn)EExg&09{a>|54VG`}DV=yVIPB~eTQMb=T;0%^K`yBf==AeVJ% zr|JSMpvON^4u-@wOt6ogqs}b|YIvhp@_d z!7J*Z$7$OrJ9K|w#Q!q(32214=|q9k=ICkWN$K6fU%gMZO(KUZUa(yhlTVRcO1mdu zV=$1nQ&g*7y9}kQ+f}XdGqAq6A*fO|DDT6bdnV}&(P34*Ruz_SX;35xP&C`DQkf#G z0IkBjS-GJ==S}u?kZWM$sJzhTj`&!|*vM^Ecq`w#+#~oO{=W3=@mAJXZH>0J?C___C_epeE*9a@aJK7|T8tP+&^2=@~|F-Yu zt3J(CeftX3li0QBhU#w*lnKG-rq&2koei%$dPqljw6=}%2sb*oiIYL2rY!x}v;S6^>zPZEsQS{AQ|bYTN+@Yh9&cZK&08x>{YDJsn9uMAG7AI{w^!HA8qJjXuIkW$QGd3iC+DI zcas|^Dq&L%X_ePP!pn7%H!DSz<_iowu51D z)c82!cz>e!S30waIu+p~B3(S57!xgw&Wef5AVnTi+LTaKZ0;pckdJ)gRwGMMZS^@x z_tWLnr^3A1ss1^_UGi%A+4(mlh4jGOmdIuS?pj@gi7cHic29F(AKWACCCkO9$n80` zfu-|$!|%J-PQ`HE(0-ytco(LG7Ly8*UK1sTa9{8Nzr&gV(TRvu&ogAzYteq^eQUg{ zeOn?ch1H@Qqn>##$km@CaiJUi4-0nE{R@g*a)pVi2KkAI6UtslS*VUk_wNUzQ{`Il z8pOFlt0b3hiCi^ly=wWC3*J|KS5Hb7d_)ZLgWuXgZw*9q8td1gXi(G#0BA=GtD$DX z{>SjUxWxLX)&G%v=2cI%ixO`Oxb~ILL4vJV7z@tS%Eg~U#JRz>N^^>C3?7u^xg1n~ zrpEa((IeK+;r7OCD>&oB5jJLoKDpuJvu@eGWo9FF@q4T0k{)h;6pwcK11D)RX%l1i0O&-aAFr z=9)`4NmBhY>D{np+8miI>`pqWw$qRh#|I_Nng84do=8$4=BSi=p02F(nif z!Zs?r;Zj_OyBjE;b5zjjuG_&FrsKQ&IqCAc25)K#1l=)7(g384Re?El2ffkba|Le3 zUtDMmLg0!%9h6sklvoZ3_v%aiDyvclcWlA=-NP3m;6B361g+r_M9d< zDI(3+Kvo7p!Y#Cpk0Z?;Pvf)i`tFB|m&`OHr1z(a)no@Zgz(rSsj@&wDaAk~*IlsI z2Ssax66X~NgJ)~nRlSmA>CIV?t&le@S^){z8z*{4l89X42?(vMmmoiV#pI1X=OmkaGSqz#U%uyZ&t;QO zy(Hc7k-(>e$rTjYcg z&N3`1vzXV08`4N8?2|XEV<;paV&R@~G!AB<*$7vM!IG_GZg7oFJ^C+;&6CR9WgoYZ zijgK2{-T<+T1+Y*QOrq-)KS`EWuge`4*;WJjrfjAhtd6=^oPLk!btqy875laN9ue~ zCqogkPCCYADY9hp=*Mp7rk1I6`URyys9csHMUk-_W00*|h2>fF>MYqFA+Q6nGy-N> z=FLKCV7^~|Sg9aZ0h`_h;!FbP6GVxxWzEGH?r1|XHe&3lzcSjtY*1otXUCW-o|$T0 zS8jOeK_0JyS*pjSs7k!{&d@bDXUTvSx;qT%4akVzNnu|uQpFB4LO&88+crE3o<`ogS&PgxkD3W@aJ@$CMHYo1#J>ml$`>V<8XQMukXn@;z=Pv-#7 z;O4w^-{s;#jL)B?*K(^vP6&LVQa*3K#KchalhLRdcE*I)Uy4UZnx`X^U|^70;HFA zX2%QGwoTvlZL@E3@W*~XC8v1oo7}b->{g0_QYdFJ(*)T_dGvL%-n}_kSL%OST_8QM z;P#wPmAB_?ojDj%qB>4G6%|6*8U|6KustrIrT`WM69aT;doD|mvC@XsC@(HNGxey? zgAmNBL}7PO_{OSJcok|nV~+#M%HRXAU##njFaWU$c~CuWPk^6^D)X3ffN59;*No6u zFlwE&F+uaPET8FC;uoJ#F88#-U~$3AaKkg&mX`yIli`_|wM0f+%UW?UJUkA^uoRx# z>2_6iLG_Wk3c@gPE>Ip7Cn|*9?l6eJVQ{8ZegJrLHbCIzgJIJZ9Fb&GpZ2(@;{Ju- zTBrLD56JO%RmY}HfZNO%bj46gt$YWR%gheGBslBYAWu=;369K-hMmJb0ddTL zQ=eO{9Pc+yzj8z4mnB~)JnT%It(QoK{3pp*aH9~^}%jlH2)S!OU)Birt7Bb zB!R~h)h>&PN>4GMyt{?c?hd~q#szO5DRO~o+01?KE8;pjD*UsMB9KfRn2G0GB1@E? z6Fk>AUIY~GU>IrW?BB)?hLOTQnTPuR=w}AV?~7~JldMUk+F~5aC}s~uiqK&|Dr&b1 zx0|t?S(dPTzV67}OCS?=JRk~)?I)Dof%gM50?WelW*Ov(B9j1a!@Q%@2b~TF=s~;x zstl8yFH2(P>aZE*lVQ47*+0?WofOKb?l!dXx^wuQ6>Av zGY!tQZ#AN?*epn=Ps0kdKnz}Dq6m|bv!;DA_ns>_eYiu{@4n4B*BL$li2>@$m@n*+ z#4|k-%w%33d3yS?98ISrS9lvf8p~Lk5~1<#YC)wi*>}tY2fN|N8dhV(6DR!ecboU) z&22xhE|$dGAWXf5Ieb?B=z&|OcjmNRZXFWL3HzMX(Mz7Z-S+0b?u+Vgb@;8C@7(=vU&!^4v3oZT0{0l9%{AxMPch|pzUG;0 z*WeHLlZ=t54m|dRp^Da!J%c?I0|c)EC^;!T;=fnfCrgm#z8&+fUi~<5Kvfx#%H%@4 zw>lyTLNd*QonfiQ(pe`pdd+Eds%Ntz2@JH@&d75XE2=Cnd<=^E4``^^T*g-RV zbnOD8;oXk8h22I9`oHz3-#rc6{K8#9?(o;ZE_! z$=L+606Ia3?|^U@qamc}s&!CeZd zNUnj29Ph&-;)hNlFm zx-Sf20`vukn6fmA*+PMJ&~~e8!Tw-f)ANFaz?R4jTx;y5Xz@o$PMdto<<+qiZp z9bunFHY6Oa*2tYWpjG<=ukYFDQg8QYSGUTGgd2go)+cXsN2Q<~jRAOVCWW7S(`q0R z(&&*W>QwX*q*R*Ze8BnE6-fKtlfv`;a33*;-nFRPb?rP{)L|3_Dy!)Rosdzt-K!&{ zQGrxDWXsuSdB*XaI7iogzx}<${~PlYmjk~N3RTma?9RG)BHd<=C$B(jM)jw!#f^h3nwT~Y&bJ}l65hj;h|*y z1|fi8Gj!I;p_m;MNvE`zNF^A&GP*+m3UukwdftV>_H zvIr~u;c>6zv8?@_Wh6q7@d9h9~~S>U(cca0RbfhX(sz@G7KWtQyJYx;%Yo?uni z^pI%9akm1o>0jL@_Z5)^;%?zXSWjT;VUr|Fm`5K_IBr-*KO#mq#o{U2^8=?5;cxiw zYT^Hh{`*RxuKWI`uRU0F_3h}z+kd?BYdLfq!HvpJdh=_j*MJp!jtcIBrfZ2ZSK~NX zTS1X4%B^m9*vHq_{_@-Z&umF1KKap1V&JhWyV4>|^MGQyDDoMl%>c@}0kQ#lNwlKj z%U5Mxq8LGu%XW`$;occNGx1-&dex+g;4`mpQ>QJ6kI3_1J*kSmrN{_fGwDuHgX>O^ zDr^99P06C&bb7#L*|w>j;LV98|(>Hq)+089L!(&(Z zw8cbRNil~hau92(H@F@E>CSZjDrGeDk@|+0?rz9sDF1*tb026vmg}l#9QA2*z304> zKB26X^b7UOH5uspJ`$uY=#zBO>s?Mks`sE%vad;*TCc`Lm-W6#=;~I)dt&!`Ukpd! z7ho{K9@fnvaI=EWXg<1C{;AqJ!Z&!$&$CPj(Ydsn!u(Ub^r8iSNt*cm|m zsR5@7mn4ztsMPtK(6o%jN%^xvKX zQ{lK`j{|TyVQ=DO*~DgRC5kTyU71C<;+)4y|ARsC%%=fKp2lh)OhzcOm6LtBSRM|> z;02f3h@lts;Fwdp4$nFEsuvO3AZn7}h5~M(B?ZNZv8u(m3Ku;a8k+12BBz|B&8 zJGM36nvMKQ`Nu2_<=Vic@Jjl;po@MK)}bkt#=x$_`f1IfSzpcjLDF|Gd>NuYUtP7h zjqY+gLq4PL*nz@j(+JxH8jEScrf$c45j$wUdOhbYYfUuX@&Z}(&oY6hHK6>&?d*Ky zqk@Zp;?B6887(LlV_lU?5>#ik&q~;~vET><+d8$wRqQ|*!N;BWu1jF%60?uH=;(q_ za-CavhQIM}0%cf+thVk`%w3Ayp|lx5f+_(=xGJJVFzAFVE>s7<;ngWQPjB)&6lN@A zhUBT6LZkPKekp8=chTw6bl6p|zqkX2B5D(4vor~U{lHTTSI%yUJT2Kt8+w8_PE$*PP03LDErht=qms6Lu>sK-Y%wC6iF_7 z-;$kHU#4Tlxj`+FJ;Lh(J?Oq&Q{E5Hf;7U-(%Qi6@#9P2JIrOUIluE`xPje88DS|! z6<7Xt$=A)s<@m2IeM)LxvD~RP3lq{zF^v>C12veY*9KycuF0=mnE~4Z8P4aXCK617 zN|Sz0P>!U@Z%tr)WRHBEAS&n_)E0!CsYuh+m6BtD(IVtN+z6>NYOs|F@a|&aWeMJg zJ>oSheB(qt@^b2cpj%Z=ncABpOW)3+>VELq;<{hnS+xBphklUy&e_QFh-xrM*IYjd z!C%1H*l>Z`izcL(o#To+cBuU@^^qGf!>ZIN<0G<$$GcEKMI5r`-b^vc6iK4A$-cPh zluPR{=$oxUhXk~Q+r3u!YV>f?)kP%)hqB692h|q%u-@#z6aRH} z^1FXu6wo}np5E)#7_rG`rKobDA#AtsX2@Z|O3?|(=-uYgL?`e0*|5d$7sKY>wV>WPoJjE_M-PYzNg3v9?v8h78YOw#jK}D+|!w)lJ0lAp(>CX z^)0STZmKGnaU0!dPu<30es=wM9dMchCeY`@7Flc6@ODANBBHZrN`dcVsBRtidfe*+ zPU&>T^sNEyVTqyyD1qJ*c}0^biVE82_nG3Js*98fb6_K}TlulNC35I}2Km;R*m@VF zRZ1f+!ZkVKI0ux>un{#w?ZZgNb_WwD)KKkfnNOU}rljsmRsT(*d2C7`gKda2kxVg3 z6iJ}8E5uzOkCUxg6<8b4;C9^krs|joS2NWMcXJ}U*<~nfg@s+dj6)}!U@>80)=z$J ztpm&=w_7D2P~HmL0P~GrCd55bY3!j zP!=;CA`jIGF zM=D%y$vRx2_JJr-l)bP=zIXO*F}5^EKK)ds%yzq_fUGIkv(V?wd3itX^tTTzZAaS( z7INe!yP)u$x^u%ek6$Tr0m~L~RM2O}x?E8U3Z(6-qw}_f>aLT9VB{QB2II$V>TaRE za)E3c362o3ty4Q(#cpuce|Fu+%j~u`|8oCQvX#efEA0Lbaa;2#2DWpuQ2!^MNl|ss z#{!N@c1u2V?o{Q58WkWxo-R+CyJ_|guPxppgib&ie&nKT^=kaB^34vO3FBhi8T$j0EMrrIf(#Tp%mh0m9wgPx;#M0dFs0rECZ z?v&^T1A62slShlRF#~Bhw5(c2g2~IVb%eK*87M||(x`W~E8ISwCYOP-5fogyw9B=eKru1;k=sV=S`{os9?HXR&ccl_ zU3r+vdS`1|)CPI-{L0`*bX{PdJXMjZh#$}BxQ#<#HpF&$x;AyfZ3uX~HR>0U{GYST zZpnZCrTqxm$&JFsV_Nlb3n%6v#X#t=6t{dzre#B2okUSNsF>bRftYE<^bbjwtSlTE z@ZivF49=mGKoUXMspyCD!+2OStkFipz#Hy> zuHv9hag>WCb>1dKv0cRN4Y2Y^M)=6~JzU!}o|Hw-G()HNr;62N2RG>OIJ8n_0i9Bc zDW=FSN{c#%=LJtdoTXE;+cOb1g|YPCdHNWLz@UK4DRO&GAw4j6)ubFM9khg2Pild% z%bH2;ssUxbAfCCdx;HHsfetqyIvtoZi|>vha4 zWhPkoCPv@7K$@b$R0>mq1zL@Q@!s5l!LxDb9XiY-b2iY)Z6teE*^E2hnWSG;;qs|!^BWDy`y%XR9RNnJ#=DZIU@ZrGcq^mBgcU@rKiUBl z?Dk}YFgB zw>fX~z)~HiRjv-}czz(-9j422zu^U}d3lS#9w2z}c#RSwFYeS1XE{M+!jXSC^~BQ* zk=EDG|B)n5g1Ev^`<+cOph~2twA-hzS7nHMpwIb zkBk3U_J}ld^I`dmchY5H1KTO)Iz_Hf+8dzR2h$Ji>Sc#6ha#{i=*r+V)1sjI$9a+o z%6@uvuQ(uKGMsmV=>Muot6UC5te!OB^z<5FZ2I-hEDLB*bkcbD z^WM{>(S9wFh2ki2%#`Kg0jFcK`{B#Q_)SmmZHasUckYZ^Wm+8d9TN-U)PLRpF`)&D+} zMJfwJegX(e!f}e~SO605k)Hq+Yj8D%N|Sivum%+npqwe}H-a?Api_%lZ)9K^Yncx^ zWoxPh_7*ZQ51ai^^D-Q8JO8tvyr1S|Hcq=9zW)~zHB!nkkFlY*FhPkF6Hk$~AffGk z3F_c|Cg}@FfLwIg-NX&%m%iKFo}Eos+cfYt^<>|R*j=tSj{WDB05c?JrT;C7O+UMfz zgP-FTKUws&rO&M;vNt?h&_4H5*Yg2Q9*H1pyV~t(Pd(w60y8z2X4@G_Hi5*Z>0`V= zGSWiYXZjC+WSwXJN=0y4a?g<^dB+1M64*H26^`p@6J6yjbHX)Unh`p_spqzaagBJf zFCJdp8@owqR%Tt9ZZ;{J_}HUlJC98Y)LQoqBc7f^=QlSjKP*Ae=cZS$MFBs7Rbp7U49?yomjXA0XUG~9Py9ORdn76-0~S*~@&;E^NpjO2 zcAlJ`kNH@^^}Hwh9{EiZ9=0ygA*#iq6RP&2fw82I*1ea-DZm1zFp3xnjksxYU;ohRIHgFdx z()P(K0xJUTsUylA2QTt$*+!LnJEtjV;v3yC%bTa5SG90&nr|2VctPjXIvQ)2od~G$ z{%CQ(+f^A>5o>VWEX4wMcY@MFqeBzCYh;;1ytXA0|4k6A6g*TG%5Kn~yH~;8jZ?b_ z%40ibYOw=ixWmDQJ2)YR3i;mZx8E`&rf8<%KB@Y`I4xH!oR$WPfsO2wz*QhCgRaW9 z`dyXf3D$@@RhU(}7S_v;W%+_Nld^r^NMI7|m0sGS} zk!`Lqfk{k^M^AWy=VxIRKD)wC(Dh<;ek$fgD~`-Jfa7x@po{Ja&(du3Eep65*c^6Z zO8)}P?B7LWYEqAPbJ#sP2h>fAg9^mGb25EKg-|;H@I^Lk)GM*zNBat1^>x!udhOz? zvQ+tbFgw``Gnj4a6{34ym4cnl=V3A37lI7iO1jDU$lMM^OC%n*s}8H1{40DqTrog{ zEjJ_wgqfNeQHSfOhS!!?J%u9V{onmT_iw-W?JxiIYuRdwSxJ$omyYVQ3vBp~(OkmK z1YOq6$+MnSFVBC#g23kpE$ z(9=bg{QsHAzD;pB2chwW-ExBJ#7P6|3)N;sM4tM^5t7N{*xXSIM3hhr@SzGR?KQ7{ z;o~o@RkjOGDBE1KfIfFtvD^3F)Otk~$^Z>H+Z?&)ozE6QJmq`)7d`lT&mzM&?tb%Z zWHXe;NKrk%uGwQJ^yCPGHwyFiY>Ihy@WRm_-APWEp}w^GzrH^fq6?2H4J<^Lt?C@8 zD|mZOy|g7VQ4~k&1M8)C6*gJ60nVeIL^k}X$O=^5x2tXq1I#=mZH2a+vJA=2M6ZofOu26`?Ox$hp@|CDt?|y&7+@#L)X8YWd^UpVne7?H!KmOyCVqRNb)iBoFiGVO5^N7Z1 z=mX_$$tCg`tTVTJ-4E=DeA>P)v`>!p(c9#$s%nAd^-YTVf%Y92+a9@1FP~i>JU739 zw@ZotFm&!$n2~exo8P)ZHt^WV-D3gi9TWowV;iNt{jDpDuF5bvTNn{FH;3K?fAwmN zf8mBR>Sb*AikgeXwoz0?udeVZjJV;|JRYjmqeHJrX_s`N5r=44BS$O^yWdd_ zcf|X4S|^%uaY%T)$;^_L_H4ruLbSkg(zUBl2Vu|&tA=6d`IE2!mSV@XoH08ihS zbtt-pMPY3;>JSx(Zwi;gHa0G|P-i6@1T_#~U8EiO-0R$HHMzpu1YPE6x7)6RPCMP3 zu$=04FTJ`fta?VDyv-d&3%3MR2E5Q~a};oFLD-AH#c3Lw_})*MgTFWXI)7+pPpA*ZAa zLVhI?ZDf-RqU0_VB&=2LK}rOk=ujj@V2HR{fJyfa@|!M*y+NnES-9<}t8%S_VumRy z%+4!QRw)bokU^m<3(uQX;g$=4>UZl#eIPV>m|wa1ohUKZ@HNl_e(mZqWv=EZP_$a* z%}}9g(CL;UAAVk($J4926nCeTDRqYFedLN?w=f=7BIuf7?7B>eg?!Vc-Oz0XRK){^ z#y|g0!tX^8AfB>@!p5^6N z&uCZOQXF@Fp-FksYt8K(cXV*;f3#O~L&x-YJ!`EiH*@i#c$o}Ed_~j7|2C|*W;_hfA>i@;0SkLP zgDFuJ$x4O0J&(^AdfaT3Y2O;y0EHqyOFOdj1t&XyyuJ@0iuSmIShFL&M`Zdb>u)`(Cy28@bVD%lW}6n;Z6;MAZz?saPR)!741C8~lj zJi6x$wI9enza?wQAv&AB=DOsb~tbwmjHCqtF!jP*ZjmN9# zatoVKNHHM7ofTsnU^^Bx56{@3Nu%4o)kP}B$sYZDJW`ECDLcB^?D z9A??AHcGZb#t8<3LGC3`54Az^$3akPxB;-qf(*aS;VnzoMC_kqwjh73h^!{Ncx*vV zT3C=H6ayw@Kc&4XzA5g|VBo4jStco7(8}og+!|aH1a0anItpfyJa93vfX5c^O>Vk8 zxgiWgW?*7m`{lZIi~9f)Ik4_c5Fj`T#0BEq&~&Wg{dm!UQxu#&Ew~5Q_7b) zYC2&bCmT`0-hqS8a6-j|o?kyKvM$igTU0>e{1E>3Hj3GbuCn&3s$I1m%E0XN%hBZf zCBBgmoF~r|-klaDjuZ9M6)x4HIMF7*5@|o3F@89)5gyO1g@e#AtXod-pxoE^|Ke3I zih0KaQ&u5^IR$W1G5t{3pc7Us>v!*f73va8Y@_(Cmyj z`8%y3t9a~;WLema6pGnEk@b|e!X@83Q@YjX0caki(=CiKgL#XvA`Ewcb;q5@t8SEO zR5{>c%*+8UIO2wj?|uB7ArO#}l)4>7EbUQS&kd6~M z^ws4uBG2r&`GCt4=6Wrs5#etzsp%UNVy(CChHuThdK+}D@`uU-DU`qj+FK>PPk9fd z1a(-6u;0BW+@P!!;+o~UX3)v>KZip2IkdQg`T4z$JYom4=TBr;t)&Ec>zcC21{$CM z9TFmSsGEIeu?}S-@4JlCkPjU;+lGJW9piJ%KKvuCVYlRZFL5!?F8`@J_a_qbim6LN z!Q3G$m@O0oY2q6w?bewOfWMO}uk%OYmE$uCe4{~})~F(iu1O_rpbh@x>5kY2m#357 z_&^1}g&i(`%@@wLUa4@|bm8$llWhT-O%$_{B8ik1xn`B11AkZ1L4!Jv(ae;~?x?V$ z@@~(I3+)vBvUPB|K*rWS?Qz}y&@~IJ-3%_k;BhFOg&}lDh*7c_;3ewS`LZV=x_k6x z2`+!}uP%0SAIYQd(%YPKoq>tA&GUgmHy#rb7koT5EaS1zX$qMYJC?0|9^->$Yach{ zKx8zaC=7_H>)cH>Ebl53MLOgs>eZJe)iWx9Nme!^Up5$qF%#JEJFF=R%X8V_{#ceW z2jeM@bzVmWmS?De1AADG5&C#9`TN~UvjdVo?Ki(CDLf8)fOPQ?XCsGVfB}?FX&aog zWWE0Juj!6xK!nTxqzZ+&boW$y7~MvX&xN|J0i{9sDc-X*Xu#>_yfpPKFW{=?8QIQw ztwSf4v5ou)LuRKYcF$u3M|Hx7QTe}IWX8$6wcp=Ej=W-y>O~9eoTiu$DRP|BHY##M znBjXnZ3#jXY$Y4nW>EKrIW6BpeZgFzf{$VDcCeWL<0*+X+QEle(blmMH)GPqJw zEhtgegX(AjRKt}fil7{-x>2F4aJeEoI{9OFl-bAZ5WO1rd9ZVkAuN+Lt5JCnY16Ip zl%Sg`&Xd>JAU4#>y!e9@(w=A*QSVe<-tNhGVUGMAvpS>@E zYcjp|_7&fdJQ%VOOx^)TB!U4P5eQdginDauez%>r)AcTG=e8YgZ=E~s*I%1X+qrE~ zQBjda!38vcED=PIMHH3I5pfw698?rhK_a-|D4;T^@ST%{k;FvvhCt)oZ~ToU@A44y z{NMAO=bZofAKgrp9z{B$mOkLNN}l0}nl5&8V>@EljXU0}<4=4uz{_Z6T7F)+m~5U# z_Lw}m1r!5&&KxSn(77T$>f@KOIHWRi#O03U4%C=j;uL@=z!GK&GvWdzE)m%HqKMj} z|F(C5np3a*yrlBg(#U)eS1X#(?xMDT$rc{F@qUxfxZnC?X~FV4@00fE+w&{T=_71x zh!gv{R>&rp4c5`8g*jX;jT*Q8+zMa4#ut*;@c&Exx4qRZiqk?^%CxjL2nAHiIF~r> z!hKi?4=Lq1X)~?AWrrY(hn>57q*- zEIb&-L(jmsZKpo^pnm3m8%@pig-LVB^=V|KNwMAM6my><{ZtHwz02XxUiVfgoLedw zh=5oYme2Jc;zlv^p=1}s8-Z^A)x0&=*Y2w|ue&JOW@sgvktej>#D zai1zY95glvp0(ZS&g;EM=@1$K%LSX6-uK_N+X073+uksSzgT0IQ|Vc{7PWaN~aXn$b3e zfB0!9sdnOBd7FuUc7kGH^=hPIP=}>0BvY@)fFDzh3&* zvA6cV^ZuFW*62nWuXzx8ke}>%7&?RE-MG9B%!clGiY#6!4UI)$Ndn0YPVFot<>U33 zMd;-B`rLk5$A1ig9xSa#?c?L*`n=`SO5b|;)vK@F{KblCkL61}?0Ym=1Ks+HW$lbz zPXarjC+^SpFlS#oZRXJmpZctimZJ-64oQ-Cl2X}j|K^Y@p`FCOS+(mc|KcW_F!9CJ zllOuhCLUy8%${PjF{+$Be6xo#%vW0Add19;-)bajXt z(?^{&p53ofXM5V#z;WT2Kv| z8HRUw*`)>Xf*ndjyYLa0rGk}$n;?EMe%w75LDtreXCF5~gFSCpA!y3O;Y~mJnbD$j zefRCRNgKPGQs?#835c$Y$voVom>!DgpneR+_?cp1O~}XGJXI524kC#d;DtRnh@V89 z4e3(snls{Zk)X0uCLK=-MOu2u1K(kA3b$Ig@y5n98~n3HS7vm{HqGc_)TryyENxL3 zc5!M%Db|N{hxLGK1Ut2UQ1@v=9uC}yL+R)``UJmDiJwKI==9hdY=}`W6>MZ0V3(>N zAI)bdP9&b5o18%V{R7*>sI33ET=&x2%FDFuK8ID;4SI!J#9|(x5B4S z{&DJC7$jUO$PyW1@VMR13`T7xoP!gJM3I41ZTTy(@Y~0Ki0$_}{@-ZZ2&|rY=Zy|l zUdsK~VPn>`{oB+2nLKo2_Xf&7#+a8i6cbO8<_=bycn`&!N~dfpoU5@XxsKYi76 zVEXh8lXQd?Os5R}$E;UgH7-ppKaR^IRZhI2K4*f4qZD(PA|F#RseY&3TR9)|(s)~= zA^KRr$$~~@M`V}9D3SA7M0wbLW%aB2vgNC?^TI|NYsz#q-ZkO^8d^*?S;qrOI9@y_ zL$ODkO%B5ChAp|ehFi^~h2rj9qj)^Ol%qz8ZIl?pAY_+3kyAXQpL;u?nX@)BBWiGw zy=Z(6mduSl`}cNYj~$jid0kgwUU&F3&D)8c87sAjSg4GO(6K%la&;G{mu?TqRiWE4 ziHqXs1lo#}(d#Gcr*)((UiGtJzR|+Gc5dUxWHmeI*Lj1H1hf*ytWN0^vymdHR80J| zhQRGKx~f`EIXB0n7<|*pg{Qn9L>>-9Q_|$x&wb2GoY;~U(}}kgHU`ULdhbNOYu~jj z(wzDKB}r9UsS_{|U9|LJSDoZUa3HWX(?L8({zNl>N;$v20@T|l^ulWIH` zm%5ey18RkAMC%0N((=%(&^k5AHSj0IXfVoLvQChdPP~PIg4i)G!$yiprAP`Dv;XbO z;-W=Yp|fc^=h^~oL>*ltOy-^PZsK?Fa>RMAH9~v07M8!^=N+{X8EDC#Hd z5}*85m=PI5*F`yGmlGr7qzN+WD5jbs`>2>pVW(ZMxfl3ebMNFY`|4$JF`(q5d$KHLC4GAs^3OJemLxK{SV|t~sfY28U;&Wd(08H#4 zq~80dPhU4e=n*&jC!~hm?!|fIfaIbH;7(G^af%#4`e>m5sss#OVGOJ(lpwy#D;Bi7 z0h6muQs6l#L01ctM;Zl2K7f^l8pJ!p3OLxL0k{ZI{W{`;7wuNY1?zcQTU?97dZCw6 zGVjP ztWR|qCX#mfVW$Um4-_e2&o1oVg+C31nevEcpH6-ztqX%F;SI8tJuuO*B?D7(P|H$CBtzW@1kX+<%}F*;2tLGvncX zN(3?xot7zha9Zt(4~zhOeP@FYxxo&AP8^SkHv!-f#XO+M02OmbQpv|~-PRfV6`HyA zpeLrKOMhW;+@Xz@{x>y5?B$#ycX{2>ed1KV4mnEP zrOA(}G%Ea5a9noEy8+mX*OeJF?nb2WHuI|8Y8O5v_^KOvnm?rRIzzZ|qv&(*-O5ct z*cvj)uip1_?;MYM-{P>Fyu6T3{vPfa7jd$zp9OF zBP%gkAhRh3LOK~#43;UO7@Z+8ii=a5q@3W-SV3Gcrm*vA3>n|_+z`53sZrpW$q7Q( zHS13%XTn~wa=E7dz2V**xe+59Qv%PE4bup;W*oC^%BL78Cd#5>I{8<;E{kKStMc>F zOW<#yL~;(xDrEbWo%D7ZGdi(?8et!)<-Iq0;(T-;FIG?|@0HgG&oJ+ed1f5{&$nwn zjL(nbKh2X7SzoaQykRobl}G3hdQ8Q$(E_K12VP@JML?& zLndeO@z-V;*PFIi&i#(0TB_i7UNcC_Ocuvnih(?~hKk93wLjvZxQXA%Kd`8n&QvZ7 zF4zAomNi6Fhg2?XjlLwmJa@VKRoQ{CHTuY7L*QoSk$6zp#lF&-?flpQ6aB)-_V=*8 z-a1TfA5Z({?~O1KN&Y>O++nwGa^j79qRGA~b}I9jB12Tn5Vu%}`-~c4yRiCIERs~0 z1>n+@Om0ilU6&9-6MsmxB~dsTYcTmvUAw=+=|75IO3si1m?+7f;HnWtTBxYgz2L-nuce zl(h|&1bk*Lna74p;lvJ*6*0XdVx7DZIsijBIl(O}6bpt?8|}l|z(fuPV?YVPqfRo> zPV$(EWeKD)cTVifn~R%#J~Xaf7k;ojoD6(v)~r-yA19!>}yf&`h=_q*;n8u!g?Hm7bNJJK@i< z8Kb4~Nf3g#4ZVz$c{(~hK>rzm%#tR@;&?oYMi4*GzE?B~!)H1~5<~9HSgRkQdS*%l zRO`U{RpAP?65jYJa0RV+yJwNpu3x8DJ^IO)aU-^0c?u%_59a{`DL5k(y`ng#Ri=J| zIjp5X25Gr$P%@zrr!i)Pr7fDYM85S!jZ?LWXP&NSvw%7A7R^d>AF@u+1*bx`Js_WM zjjoKmz-*M9cRlOV3w?3+MeYnM7iYTU$g5juPRlW z|0!=!@Pi`oYw~o?DQ_(DuJ9OePZT9`hUJ;cdI{z_Kofy`4ia9+WVZqxf4!8qHs&mmodn~r#(IPO zuITE3RE+|p*beb@lC?9=dylx}s&M(*ufmi1@;BU;m3lRb`e@Wz!{h_j(ZMaDWed0a zCwL;)NbozAk=Q2a1 zaW%FdA8|njPbZzL!nWj-UI}A6E*vr2wvUni+s|L~HCm3x%<)x3UdDKh5HW-h^d*+2E zc$Ua2pr@ZkvF5ekr22JnGJK9jmWEyQpBPhNI|RO1K+Ip*KCdjUwC>9&o05_!|}KzKu7KesiyN7^Coknd4qKcPfU zc|^79Ku|A@*Ekk$4#3$c?pUDc`Ip9tcF!w_^!3dTS~#d^{kKm= zh9;r4!eSxpws(OnV1k=QkrCSGwN!9A2~^7I)eC)8;f7fzl(-F9EG zc173c|GS)7PyzApFSbi6m3&2d##9Sc)f;hl4{*rqYs2%TM#Q40!IB`im2j6 zH@tcS9BVSH%#}R|u`-lH53l{7UfC?;8Wi_Cp9GT0ZVhr`O9Dj{V~ArV6jMZz0xAak z{f;#jZIWDb%cHl4R=XYG)OhHb#X4B7FiML$SpDwC4`Q>H&7%xTVWlMTdFsrWcCCX{AF`zr28>Z|=t3#>w-l;kGCs zhpbT7D7P`_kYt5o;YupE`qiD#%zE`p)yyEt;p$}lvOE0hnQD|>gv4W5$IJ-~z_CO2 z)0X175Vi9H>&SjL{BW+sxC{yJ3V%bAr;%b4`<+EGTPczbLWQF3kR_>;Ef1=lwbr9q z+9t_zy-(Dg`f9aSMt=f!kT4}_azlep%*&4}MPl1O*UoR0nEv; zYzUFHK>f>*`w>~EM>98{#w+SMZIU$}iJ}tjE>h##8g0`)d-u$oOv7s67K0KDD{uVx0I-7^%;8fx8ei-cEnyI~Bj(z``b=+-P9^1#~H zO78-3yxT-DH64T*i}OC7gRC$!rN7|qWhx_Pj>Y_KJt=cy%ru!`<{-rYhz=lTHqY)> zr9?cW4=VqfB>!6y(CQ!vBDlmksO;xz6pPCOK2dB5O%WyX5;-4!I9ejPR8Sy>2uYh{ z-z)e(sQ>1vVx!6<(5Y!rB=hp<*r;q^HIxfK{1ExR*;Ft6;fEh$xf4moIPMScUmzQuc*WUmvf^kd2AZH}Q!)MW1L5eWpsrsR2b5=! zg_P%cCA2Sck5qjia0O`oJ*0K~Zs@wv}EpL>vBDl1nu(i=JVAS&PG``G{byb2ItN#v}Y zhF8{bll+F{=}i2zVIV#=anB3OWm_eEP{OgB^vips^`2Nxg5pA{3s!`mP!w_R(51j* z!P1c~2;&u~YLqR}XQEU64y$^kxG9Iy4Y(4B3DcI5ZlaTIaNnhjpN7};NU=icG`&l? zB7E4TD}2O%aQ=1KcDgJ4L3n0BYjk2rPxv1HWx?^BCaK0Rb-@n0kuLKp;~eBQ1RnNH z<@c0Aqongj~*R8bwRi4KW$vao&w-RADChtaTa1n)P?YWITm3=;rI9K$Rg!V zjKXGD{6$@_X}An9VO$QA_=<%xd@Ez%33R^hBCJ_Au?G zakL$gIdB(_vU++Z(C<*yYW!wyTq~S)ELpTgY!0$R(fdD|$}z9*cUs9(m(Ea{AY|XTLrdss4|LH|9 z!@om6fDN;!vJ+lAS)?2Nlv`sSW}jP-m! zdpL)kWo1dGUi>_3!HX_OeNy#EoG#UX=;n||Wc;CnJ*lIG3%I0=yBg;uH5RhuB;U;)-5yE&RvY;{gs|b znQQh)j@0q>%-RIE>X~O2o}<;>^RaZbT(*(PhZaHfeN)sZ_H(?JIGB85rWdI z;a4;X%NFW?g`s5#l7IKQ-j>#Lu19srJ3V^C)GOWMc#ys719PKpYMZQ@OH@+FHpRFEf71s?QtfTR5qG({Ix?Q*rIGdR3Op~v4tA`{dILDub z4N7Wfr7leHc8uk*VkX98kgN>pki+b7^4)83mFC>5r-^z_yvSK0igkk&3zpU+7ZmcuF3f z7(7-=9?N`o%MSvD<^tH65tlOyi^E!BLk+4upi%_VRRvxbNrGE450}3i?&^f;1pu=f z?q?QWq9!^jN3=Y-uQ#jko>ao2aCvECpIKWCK=8pQ<-6kJfvc9I|JfSP+Qc= zzp?1D_)_FLa(ZDCzgM^;GMQKHzK2`S%N9L|yfJ@RcEw*eGhbc-gx@RBxY$rf@Nj;+ zFiozTcOYQ5>PlFgKOC3uA+6C!E5?2%Sp9fUc2bNLN!4!m=Vi}ALh*7{spzn8E-Cjs ztURH}f)U%~b*cu>?Q|BWJ!Ffr6&g)#V=#Yp6+Y8nuF3}@a#X#-F4*3mUU(s}dR9H> z_&Y;U*&<9_43bX%Ik@bjtMcuD((W)koy25`uP8UTmj>Wyp8!p;)^JHrcs*xXaJ%=H zV@E|ve!el2XT)I8=68;e6;6ynAmxl%%r;QWdWs}dF{|}i_A(Aqw>EiS2uA8w!R$8w ztHKiA5xQ)`?TeW)u-)94-8jK-SuJ8zW!%k|gN?xG`S#5+Qoyc?(uw`!MiUrSQA{NT z3mdeE>YQH=4Vnrdj2$pj%zh~c34ka~Gw9&~Ikk>x^Z6e@6D z4BR*e{o`h7KXhXUFB#Q^i-rBN8?PW&;||^FF*JR#79`RNMXTgl4pYIm9cj<)XwPHCEyu7r1@yC9(&=G*Mp>54mskdPq=o5H~0zF6V^Z{7exl?w$?NN%B-h z5gpKw*?}8U3;)v5Lw3!?=*<)Olif@ZXN{$1{OSBYb7I8P@jNHqH$k2L7|*4EVn9qh zhl)WuajB?$?j8u8bnS^ zNw*UB(?pd88%i%%`Ji-Wm%Msrrf6RnmMnCTRN2SeI-WKtO@2aAr}uj%C+ufO+;6;7 zZ`=LW0TS}PM;9Yh-honu{_!}{o~S%=*xY0}3|fo&FI>>ZQy+ByI% z1b(TJo}oMcVchqGFVW2;Ndv^zFfcO+!MqH=J_xL>CW)LoUWMFhwb_Ut%G^246Bph)kn4bny~c>&ZwO1zv|&WXQ8; z^>cf|)-pX|*zp#-Q?y3or_<6v{t%CTZU|6vtG|t1nDA3u0>5qOw6H)q@{k3*y{y&KS19*3yS$seZk522^dQ$!iy2vBC}k_CG>jcqqTB4g=8-XM|=-wI)Ar%_$jTLNKsGW)H);@Xw*ictQz0zTqZs7*^VV4d5 zmqfXA9RvglRm*~l122p5x8Ue!qD>!c_H+4URvrAx1QFKf($a$Cv>SHOFL&pdmvpj8 zfjY5gZKc>#FA1DUSI;aIt>%olV8N$0s4+0r4^&A^p=UDR6Y|Dpa8Jl@w?1g&>F=I< z(Yd2K2)rQq3W%H-o5`F$IKN|llN5_s)`uKe1Wt_aHc32H=vM=cyIMT&fNGbHK1+8`Uq?>xR{Yteug} z!@D(#RtRH*>PeTPP_$jTUo_%UE<8j)&I0?aR)=*083;91@Dr$Uebq{uWc}Q`U`|Wr z$5a_#67J}fg!DPSFM7d@mcBPXO^ygb275MH$19JxAFy_gMuF)Fb-yeHgbFYbJ~>fu z{gZC9u`w>2*3E-s&#+oar!YU-_&;-x%>LTAK!gF=}jc#w7j;28g@zfYCff<_Wtn7k1i}aI-_WgR6#ErkJ%9S&ed@wF}g!nugS>91q+l>n1CdYx1SqG|47+OwJot_G0SM0};uy zjgYwi6X^=F?Mst<0vV$*$)_@kDWOOasPmDt!c6~kNv1MOpJvKb?%|w$ooV6r ze(l@ly1!2Atmc@YB#mO$Q6vfZyhRZ@`i^Uwya_Uj72;EZPGJM5DB=ixYm(g3v2$iK zF2>$839nh_mzZUt8dLbh}t|VEOf@V$}^{6qy$}Wn5olYSYlMDfK%r|T4Q_w1@ z3j|6~`2ZVH`x=7ugL&x_aJ@caJZscH8TIm$#V*cBaWy zOZRgN1$*6FWfM_}Z8MB4E+=0cbU3iA!|456hd(oKdBJ8^dL+#+Qz4wPxcq*gEw#4(G7TtlN^jg}?pi`8gXWUVg0PY>>!ppggQ+&vJV zUmgzY%*HuKl>Hz%Ju*KL3kMpYZ0%xT`V5U7wjaZ?9k)H?#d|>^wd)fIK%s?7%W8U z4c^9_3LKK}1w*9nmk)W=aJwNToWQ6L5i~rN-dOdzf)^+4LpF>E4OZ%cT)HhpuXk1f z{Dd42G*z(K?xZWdH+VGBgENOfOa%Yrs;+wBZ_p^v?NRRoJ>hQpu{mv-LsX@d=Yg&B2)C?E;rof+=b>!TU?jCTq)ZI?Cs28^(v^i#5!4QdNFA! zVfNF_?$NjRYU@jpr7X|dcb088m!Yzv-_*lh_EJl94msl;A9yBOqqrP+0-mBc*cD{> zjRcmHyMe{fwX=@C4O0Gw*H?qfeXHG$&FhtY>YFlSDGwcq<77E>=PMR=OAER8NrME0 z{T0=-wlSajUUYC`h5ZDLvEHUH?VSsY_pzEF=gp$k{NdGqGpE>E;ih7%ynMQrK8?liHlvFd4R@!s66T;6YZXa9*eIm9N*nZr&S`O?<_9@lLFClFXyi zC4=G}ocknJuu*h~yJ6O~@XPWipWmlPWM%B8Zl3jLUZ@ctvOlZ0l9DgYCg_9-nrbQL07a@WS(GNv z1h!*~;YWx^8_ahT(>Sz3L5e_PVUcp zJpbN&jp7vd5KkYe%@oyuuhk?g7eYE(c869Q6rbB9X~EazMU*1~x%S}A+`d?N|1^MD z9*L+e)5aJP^dRKVGScb976w{(jS1EEQOsu)xdrl2P?3Rj*j=7bP0=h}Dp)DF2~$1d z0=iMgKpcuPtnoO&zl{B64QRj!9@d5wK)W1d&4Tc%_hG?xlEXbWyKqiHV2@YztX-bk zH?J(}A-nwzNA&{Gd2}4pDr=Kqadw*g-0aK+nNS=85uN(kkHjcKq{b>)+=n+W%<)hU zMBI<)q^sSKWv8W2C^iKc`VeA-P@_P72W+Q+ZMW6eTNA z`MpqIg!Qu^kyJ!uG01%~EE^Ce^F~|@m^?VUT{wCSfZD>T<3mLNv{LTzN>#yqC<>va z%VmQSHO6Tq6RDr}VL@F+!x43we2BZ;6J735_ z$xtn&QLK0)iF=APNmsn_;a`)uBQ9xj^#Oj5*XjjME5i6cALdW_-$e)?fR13xg13Ji z8y*^@825!`iJKLg1u2|5a!ODFH{p8#*!1FF3s(1b{ zRJQfbA8V1u7=dxVbbxIe=O?gyyRz2%_KVKwYz7msxF!Tc$Qs2S?x183tWy&+jh4T} z(W5p^^615r_>NfA$T-FSOSt(d!*nWH%o6zQ3Keas`BQu zZ4*?U;ij06=cxCrpfaUK^CPZ#If<33jYLr&jZf%W24w#m;SFX2GwEJw;*1QC1xt@U z4%%V(2Y1i7=}GmQlt%6)6zdHYt|`rZBfg=7*+DA92C@>t<=8$iM{D$C=XP z&C*mq^kfYOaobfa)6&JV)8Kw*(&)(|85uWXHW2~sc^aN!rEEO)tmnXR&id2vgO|y* z7u|u;0`V%(!4s|;1!{aZyF+plXyqkVu$%r}vor|ky_C@}GWAX$QtROMf_v#2EP+_)`qtuE{2o?x`L@rEt8x+Tc?L5I&2+w+}Cvs{hP|yXkH#O$5#=JrO=!c zFE~J&8N-+;rWhb;ZKq;V7pw^HmQ^imjlL*6{uM(9$bpEO5ImmC6f03?4s~9Rsm@AJ zj2k!9!!Fe!>j+-4gmVn4;*LOH*$Zx!(~qgTRTG31mawp3CdQ2bivw8qgcbQU|7os6 z=d@cPtIBf7Q5aQI+Oi6e>OdKEZj8I+IRv15$3GsGhTFVLtRA z8FAUdKeG_QkO}hG6Xa*GoOuh@W8&r=np?2hk4$;9RQ!8$#;r)QoGyX)o%%J);4u-UX&0N9g5PYpZn0~^*&Eq>X!;=H}%m%k=kUa(!xK9hB& zh+<&tn@`28Bb%AsoQq%4DE3Cx``+Xo_RJKmfObQ%KFhtaYxyBwz3;F~&+G@U3`5Wa z&m@24c0jco(rEDtjbe-6E%)02c-@3qp7DZW+_P_GfM-~pwvQ{6Mz{oJ{yl}{JFzQ! z$OIad6a!1sE-EHn5)VoYdTIVaQne6CRfbaOA@}&GwUJPcD@paMcGFAM8A#WgXZJ`4 zXg#&xz@lslX%E5s3}pbwh^zLi5ytwoKq>FXs(3d?B4)fGd15`_m6Zv%=kPN?^0Wrf zl=bg?@Kf`~= zyWdOKN$MnN)6u9HF4fWZ#N`n!ilz{3?t!I91L2_IzSQ-Btdnn8%(@*&ZE+}e9EcuP z*qKT%|5`LL8l5titP^CV6B`{M*p7+1ZKRk~ilk66S-f1j!nYs#&ZqFYIHyC}m`3`- zt3``?16rf+2+QE7BW=d>pJx`xc>Z7xyn+=nsCPBXeq&BZdD{3mv0H0Jgc$>H%Xlf$ zOfRUq;jVBmiE3je7aztcwFM>)oPC>zvJMmFZ#(u*Ga8F%-S5<73A+%A69@F6cgz^m zl0q?w6j?*X9C!n(d$Ks$+&a~{896UMfR3KI|A=<6wzXokd5#@-VODv`b-57}8&d+$ zlMPP1fK-^waz4d?bZ!=Og48oiaLZaMS_;|!R8ayGPod)qT2G#-A`PTvplA?_U9o#O zXj67VyYS_Kt=`pK4YaQR{{fVDWB&B>5B~3e$B1dhg;12GNWCEAUqb{&p!}|U`kzR` zG*F%$`_yfrn9XP+V%j8FJdEY&SewZ?7`WW~5>)c)m6vjqOQHE!icfXeuq*{O6eBL` z$+6zHm}`^$fCGKHXFaag&9#MDa9;M;!C%ECz395^BY_u~0&b&d^k++Sjqm}v7XALY zuWtO_-G6=`xzq*RLjQb-$74LL5jsXU()%Nm{0b(NWrXmt1HJ|{KMQ*fIL8XU)Z9n+ z@4jXs!tnjfI@6O^~mgCtg{47D2?7T47{Y>;G z$z%UMx@u7|ovB1pLMHuKj7NnsSDuv%{0~+Ed`Pjm1+T#b_y#Yv8$I9?%%4@ATLkT%bCw%<Es_i(hnpnVw8E7cQiQQk6pR=cn&Y?}(hyc6d2j2|p^qTxBWwZ(od z@O>j#K1zyzh4eVFtGmKPFk_Hn?op(Via{SXnb#{VU9?1Shrf1)`eNV^G$LwuD-J6Z z?GW52p8zZ60Vx-PlvKJ0s`)2l+A!T?$OB)FxD0u~?X)4Brcw0E5|~|_G8k;1>XGNp zFrx;{#lVeoI)vBc>MOE4|I{W;T@de?7pksag!IXSN^QV^yL!WG_(r4BR~w@1T?6?@ zLm93?E#r$ZYPk1Qbr4rheyMJ{IuT0LbYURT8pmk|LBDip&E73bA-(u{L0LW&rC#m^ zpWqut8uD;jy$n|C-jw-2NYgw&7~{m&$VxB<pcOCpw{?uoOF`wLiA6) zZ8qmQ24VtETZc)>SMT+hx2Unw31)|28`Hyqri6<=6%}yw-Y0OGp`UjFr!KJj6?+$T zJO8=1ebuOBAQ@>|AL>cQ8({&RBh1zc@Qw->Ke?rA9 z5wwAGip2Ysyk2=?$RlX++{9l^>iDaq$-Ldd1a6nSk=Ga55ZD!#%*zF5cUZPVTmV<$ zowZ(r#Blw6PO9H-pv)ZzFXx<8o>rnO^WlddE&-|dW}lrc_*R~TZp9Jeqfa>pm zdD&cTgAMz|i6g&OL_@aBuAbf(a5cDo?vq!e`+3~Q*Uv5VH=J(-mCbE0O(2!QWESnf zj>-KK_2z%p?`oc5G&^mtockR~bzwk9AcVOk#+Ren61BjWu5o&_+Q(6v@=SyhEvPCe1B| zJ=c<#a^2KxBQe@Np)eYSf<15=y>}vie(S-1N4#wW$m<{X|0_8*jr5p&vll4l97S5- zn-yL4N@1?KKl0Wnj!CK%>qCx{!g(Vu=Y=cWu~P#EERyG}o}+Ic_ObVcAbmL=)@cD9 zQJBL`lb`iOZy+ml6>tYq{q9HLJ4`9`NxKx~-kmTlyM00X6nDq5;Y*=fw4bk0;4i@q zYk@cs_QSE^`0KPm)jX7b!?WkS+vFHctDf2Dt%W=hN<-Q@rS{Cq7f-o8FKnDjb`Lq^ zaEROQV}A8P%y0hi?yukby?6!1ETc&5Bs=s^|Jz*tb!oME2LWgOtWX&;Hh{ODVv;Gc zmWnwhDGP>lWS2aRTdh#%(c43@qjZ-%PO9;+6=qL=Gxi;^8LsvXZ}(lS;5yaw)~pP3 z{+tz>+AaCUujpi3J(o`#eYIj*3q;(}BkbhglOYKS&m?lnm6b}v5ySiK9v1`qNIj<| z8sFo)GtooxBJNht*60ks(a{EG_j4;8N{Y34ifpkSc2CIDH?o={=iNT^Tzxcqx^cPc z`LFiBl2z>d8Rz|%5y%&eNl0y`7_gMQ1kvL&TZf*RxpT6itt2Gelzy@e(^4;Pa@{hSXNr$}B8`l~GL1QxK z879j1sh+S|wKy;( z>+?p!mJ03*)t!nPP=#0%0(FH@Afmqiif*R9u_i|54At>S^6mzqor|1U0q(?cQ|}o3 zi~}cQF*Ek{b1`7c1=ew5DGXN<{VvZudFZs*v69tH3$2sox_8njx^$gnEZC~dWrpO5 zQMcwj7EeA$+h$zYGMNs%h7~TTA0GM5*6}d>owyvt3d6ryiiL%Hq!89eA&nM=p?uLX z#)8#6+J+!jLptQ*1wW*x`y02ozb!eGL~<-8l%06y>rFPYtVpY*he937VzCVyV;7ZYymkmG41-v=E+4J?L-Pltar~VPKIDcI=L6$S z$Z|sz^ME1)ptLVp$wTc?l*s@K(G9v8x1~tj9&*15+8TAx9O@c(ivp@amSP8wg-xL_ z_se<9c_p%B-bH=m@H?)Bb1)B4AkKF0mqRkZ@2y!^3P@c-0WGM{9AKBox9h!fm0{D<~~I4B&S3JlK8-` zZ?r7BEbipT2ZFFExYioQ@@YFgv*;$-A8*Ztzk0&fktW$n71TaL==#vNwtPMJ#~p8! zf9uMR9>3P{#%Di0_KOPmne*1iZ)CmG@y0l$+9w)H_L?zPP@?>A$3~m)$DP(wwzBsw zP#pHHBPXFvYz=3~eUIeOOqAs)i8>}v3K`(GLjT`P2=g7Fll)qvhujl6CrKmlorc^m zh&M^{0un>cdrho6G!CTL3j&+&V~3QpJHPfdiP7i??+SlIlAYM-6q^{GEQ;Amk#r31 z7x*_wwn=cw0zapFetzh|KyANJ5(3&k{3q7du*!HrH#&ady4tcELnCgCnR;5*4c(so{&!DfO zb#IMs2*h#l4vlKRe;wT;jpLwV(CL}?BU+;m5xk=nc%X*sHH!1hx{y=A?5yLLc_oTI z<>r!&bB-uA3UsULB3r@bVGU0EpvKmrVb5;14z+K8>qL)_5mvfa=U*lnPHcIqOdzy_ zVzyHxkBT|%`2b_R9vL1(z=s-&O68Wx2Dn>Xn}oRYYK=z6bx3i6n=?BHD)LLXb#(uH zT&qgGI{AIVbwG)@EzYBR6=y?OPq@tnV4TRX`S-BB&^j_KILB0dd+GPbqwVg**(WR7 z?#EPXL$T%!n_Z2#APo<@TwUYhz6$p0XZ39uvGL!C%URE(3p$}o2}GV#{d8e#87(ko(c#Gu~Am zh9lJw#K{&Fcop%EL>_p#-+}4?j)3|Y%i;i#8-}nB_3(FOx277=6Sp$BfsD9VTGpJF zbV^`rHHJu)L@@~zSw+R%f35a)-6Aa~cXlJ)w&>hz8{e1^+xFZUvH=m#9b+Ocu#Ski z-`KiyiV+d2oIQUck6o-^Yym6IYbgoT42;3bT8ddsk(E?Tr8mZ*mkMG9NWF4sYJtVn zJs&A726N!0oBs4}l$#L`e^HCW=X;$U0~YAT8mx3x_}wQ^&8J zRqeKubP*lB9|ADLvNp*B_DWB@RQlftbKck}b;%dh<-77Lr^eWnMZEVo2od5aaetjrD0dx{l7_7DbdP zuhJP&x1j6zYR{w)b*rKjuC0T6ksI6`QauwS-bLlIN>PqSaTpSZT=C%9Gf8Ux~Z{jxHzPLF1;x{-8x6wm0Jtxkkvp9tfmwy4D01jQ%T z$WgDJ1H>s)*8S>N3(N}~oz~!3DOp_Vw^mXO`2wSD}6=YOB|AvKv&on)rhehvp?lrGwc12?bUaHBES%Irh_GdtwJZ;R1y`UZmU9Npi%yxY)l)UFg@P=yOl?n*d{DHJ_0xNWH+kzr+9bP`gV122 zmu^*D@?Q~hC9-S63(@Sxj=i9;@jmO=k#VO*nvbCsRmcX#Ls4Zx8byUn1FV@APKI9}jrXST9*J9|hkXxwHStk5C`A+(49+nRue-LO zSN^%QVtlKy-5@hDP?q1=h$oAe@owB~4mAc|WPes~B_&Rrcs^kQj#`R=MXQR6`5d}~ zod4BjaZB`-=>w8ndMUXNt>w=|pYm?vm&y91_1=59*fP90a-*bB^vGN5mdrzkE5W;J zMr-t0pIh=r-iNsPs%&l_uLtBI`gq5M!!D;p)ghSyIzBQXy5y^)`(%f>Ye)mHIPw;2 zxhRWYJ&RIE%kysGsO4Y8@~CygV_`2&UC;lH|MSV^PHb|lkjonav9#&^40(=69%=D@ zKaO|W^R%+b{n`RBH2ggax2f{THu3#=1+a}e$4LxqkTe1R9I3un{B>ci(b%`KP}C%C zgk*J>9OXq)X5W!bzzB==Ur|cN$D;>`nk$zcaFvUZ9OS_!2-TbCbsAx z#ne*d02NaJrA*Cyh{Qy#59yo_wQu!M8QCM%()WoPL&4Q<%jc|zI(r@X%ScWr5NnY% zj?H3+U7GpjzDosJB4ByfNpv%7r)#`Aq(vY}lgLqD3q!(ite{W2-t&R9SDq4C9+u5j z=c@Jv)cfMEI1ChrapB5gAabx>vJRs1h^b}fc||sUkQ0Xitz;Z4y;J?F=p`U!2ILmh z(Y1pe_q>PJmh-k?#g=g#eU0b8%PDhrh7BT|xCY9K^V7quhDD{y&}pMQqId?1JD>Gg z4LNo#y?J)G%Fr4GdQP_}vS6o$5h;CF4y?Jw;t^B;x!_U6YmyFe)ml*rh-%@&*CF2= zs8I~d(zt3A!|I;jEWPiBa?u@Rzf3cCT$ToB-4+B{FeJ}CG+smvZU5PCDdWQQ*yHg@ zpn5oOfN6`#B9urmYbX*=#o%gUAjNI)Xrf!0563-Qc73ZhOtD?#j`t?(3yKB*b_#c2 z(HV0M80h*wCda>lV%Ae6nTolu7fQMP|%P#-+sngX;IlCQEx9 zCw4h1P0ZYOih=f>+o+iRsysm>I2{Fob;3;8tW?h|bt`q#LXTIF>Hr6#UDzh+BL`G_ zUS7>vJ|`}?dB!0zQn1@32UL9|Pmts{Avu62Gsqq)qjr#eO~;ep`utbcp|Wev#u0OM zZl_HZS<&0Z-D{eB=K{UBB|09>(q>QWRFWo7c?A!qO^>G@d19EqmyUIdrv?{Iis`Ti z4bPv;@n2yb4VK)HIrEceyy!G2EYK*T+9aRL3PnR6Es7S<{;`EAGXC${d<2^?VSdvj zyk~`pslK8ToriIO`9(W5!ua9r9MTdXYZmWRVyW#bHt*);DqA?nF8Gy2M=vBz^!Xk1;` zH;fpW+W()oNP{IO#EHGst0qV|Louf*a*B#E=I$iJvg`9IWJe;AJu?uoNs{e87?BXQ zFQALb6Wx_#S!-)_5#1EFjnggLCdm_Z@*n%3@;(xI4Z6$MLC)|KAk-av?XvjqSLL^* z`Ks&ljx(#KqejCXWrknctW%=((@W`{3$D*A;OJzXq+G1|#)sQqPvRtTk2Cpj{}pi` zjrT95uF9{7v$=Q+Ue0c5vlBg|C}Q)I6_~8sqs?>odI^oVT9t9-S7hxAhN~SWxY|lF z=@i*W#jN@2WpR4IN7bT7QN2DIzb46G#2wdkK>=s^v}Dgsf=wP9d{#``Gq*#2mvc*5 z#A7$jtiJoP7qiF5*t;k0wRL>lseHHaRU5n4BZIE#B+{P%wdXr3^oQj z0%(I${YG4xTyw-#{s(6+*8hh=(_&eB(0V94&7*6HdN(u{fHKK>XGtF9`fjKSq29f9 zW~&U+Gc_SAJ@<;+Tx-eN86z&mGprVDC>gAs`#UF0x*6%;XRg!PgWWRi@JesWnMi7gz)aecbx zJF}!lJ9F%t-?~KBIkBAq*}5@yW*fyoy~q|Srhazj?8c}flK%5s->!NcJCF1V3smI` zjsR;PkApY{h(ECF`qWx%F2LHfBw(DP}&HR```0B zPL{IUC^@kK$ufb|I*Li6NCFkJF|b&8fj0cz8r=`AqZ669s8tSPF4;IqhRbY&5UU{_ za&g7-SDTeaPy}WEJ%!{uvE?{qGR>6~1612xR1DIMP6a;vNv@*6q;R`eJoEHv3fqoEH)?~`dTfeu~+_hyxGH_yRV#N_|lP(Kh1)JQzUWI0y zw_hIc-Wq(#|6*XTbUVEfcER{g<5w)q=Uyk(ZpVI_`DXU7Q0KCj2J)LcSFe7lQM`Ab z(;>VnMJ=*azu4eHu||QiA;-ynRkd6Cj1n0Ru_0m?2-kf+e{JMF`eNYnz-qS=*>SRE zc9!dVaXu?U?Z%>Gk@qw(?Zu)z_)j`>8E-3U*@NzSUG*JGtkE-w;b4><9`>bz)zOnp z1Y+@p+WXzubsxJ&*=du@zrGw?@`iD7TE70fKPI(Kyg0R+EKa8=22`AmQZYz_=$9V| z&nD}582VRNMV!xTyi?w1eQIavE0Q!&T5mX~t`gsr z*YgTErLs~lJ7U2sp#HRBU~z;k-dRC?>ML<&ub5jJHmjBs=c%n2OvGG^;y$T?+IvII zbQ&}%*{>`GnYzhOJK2mPwlEmy^}m|c{6ljo7N<=;TM<{;A;^eQ1CiQe$UO@byzTZ9 z#`je|e}>05zP)#|vJ+Fnx$j3VHm*G7z9(g*+lg16AcIKsl1j0bl4(Jqf86h+KEvqqC`#)Y<7CWfUM8S@$aPqq8Nz{O7v6gNMNhCeshE>Q;Db*KAQ^@^!RSVZ}?s%y$iI+(< ziUtYXi?Zy8paw~u&Z&^8wLx8SlmKcMqW^~7ZE;9%)Ai7&zo}FWxyz>wAP$ z2I(c7ll(sSW}g-?vn0(%ud~O1t zn-l}OFFNEmY*Y1fdqKXmN7^IZ7n&?s6?s@$18uCKqt=6#8N!eUdqLpaJWm3 za;zZr8}u<3ef2!rO6X|9&^8u%&v|VLiLB5zWkb!6zVW6JZ98Xk?vgroA!R4#-dr-l zW)sCg>B%u*x4(KvewxnW8Cp`H_CptRXw!GnK_b^Y0T^|JB^zlOEsE2^&mumbpHD*` zRZCx$9aPo^=5ue)?~kZj2zU*4s{lIBq6CCS(ZX5d(V!}XtahdMepQoH zN9((Z%DSLN;vs!cRxCvG_dvGY<1~$AuL*MJ9RUI>^Ej@ClQFA6U@M7i6-9DXSfGj->i-<#uUM9lZ=x1J;G*{xen zoV+MC!B-B&KyGv^6|>U&9Gwk<5G~QEZ(ISrXRK#QomCXk$J@_&JolWw^~9a z^6WRU3_B^NkRn?rc)v}gRaB%x%uK)ShEVmBViNA#<34CV>Uo+fnekHtnfnJK6h}D zxwhGtB>iAT^$ZiWn1axOJ9{S_n)N~a2l;%NqO4F%iRFjlUI({781=zK*>2e#=u~Ow z+%?fLY`alqBRcH8Y=QgjUShN>WiDAK$Vw-+D^L$PhR?T=Vp1uRLd9VA(j-COqS&D< zlH74klOIrEKgUT~XP(Reo4I~+*TlbKWkIHT_E)brFQ;|dM3t4&Sxku|GIF8NWW)ty zcgXnG(n#beeOW!|j&C8T4=Q93n}hDL5KivoXRG$B4y!=0S%N3x-Nqg8wLZ0MpfYBO z8GUJe;BlT|2bKRj^3J#ZU^F7XZ2Z9@a@SIEiW6tFlT7kjOQtfz6nRL+bn+89w*x9< zhrIJucX@{Xdg_#jhoIJ?4Z^K4o`(Ij7-d+qW*v6FB5rc8*B?D5y9jOgDuhGwdhl^@ z2yE=PQG^vzvBA2Tb*cu>#E>nng?{;<+B)L0k#i5mxZ`z%#%oKxFs*Sr;K{FTq;bRD z#p#D+`Wnv$&k>hSJ~RiUGnMK%{~BShFcGq97%0|-t$g_~sdUR$b;!Gb?bQ)7EFW>% ztjZQO!7oVitM|o0F-M;mn&gMXcQqcwAqqv!oHO((qLaLU8Lm@r-yR4sVzBY`byXyn z-A3Dq(Rt8h*{h%!*m)OIG21;Za@I0gp<23jf%>YKE)1oG3YZPB;2HW}ARS}K9czAz zC=B|Qin2x9C2h>+mo$nd*|LB(Nn>DsD9}(F0*mMafh(Yob_=NMwMpWlIvp(oJ5ED4 zYVhFaI4zsm?S21Vx8Gb&(P>vXE7FPB%Cj;O=QerafTWfb!$aK~eMzudQbir-^ph@` zPEx?xC^CfXGNEhnI>C_dK_&JohTYWcHvek3c42%t5Z7g?3mfS&FboZWpUgb$;drmu zK{zm*#qqyowct^|y?FD6c@+$s-IWtZ^{mv)6iHH|?uQ%xZjIjLeIdA6aUnQMJSbl0 zR!0(nM2y0&xOOFS>V1b@@>SQ|`?&|g9gm===Gs0;JvEHO59BJ^XPTGnI<0B3Ql6U@ zitQ@4dsHo~@LxiGxc-exAT5*6?T6OFHz2k<3@$|?=PYTYJHLH#Q6gs&W*eu8{fs+& zaxD(X&WM~T3I9W|(PQcP_RTU<@Pc{GjV69c6~(|VtlS`n*CxS~5-_nU-5!9X-&xW_ zl6gJiEVp5O&8-@f*;{-*3mb9iVjgm^f9&!($0Trl6{1HnpEuw?{LF;%rI`4hPWSgLBJ;fwbWGxkw;HEE-3NKpJ51aB#kXc-OSlKUIEhv&q;Qrk5oY{;Q%ePP1 z(FM7Aul&x4imV5X)5z^F4Jr~$cId+t^N=EgR7|X3w_6@vCs|x8EQSpBe&tTEHVMow zPPJPNeUVcj-Xh8I$fPs>AA9ct*JOI8jr)pkNFEG15y<-vs7QbTb>v`DLBHtqyzO+i z+jiUOcH3@S>F(NZyPdC_o&9+mEynRE&IGGJmCIxpl%yJpC>`Gfi~!m zzL-b+-zyL-lc@WRu^ZpxPO#JFl=Cc3P!ZjN+^3SQx^ZzTCdP4S7|Mf`G!(`0uA&=A@uu*>q!Nhubm7qz3 zSLXYbIb5&6dQR=ci4*-A>(4X%>b=7q-lM6rTz0=?gC@(|_k_vC6b9pRkXOKtJM90) zEY+amh&yB*%A>L3&J=b#u`oeUD#jcn$Py$=%A@ln2I!g^@HC&mMao=|fvv)N zckF6dM`6sac*%h0W(Wuyp=dZ!5W5s#ZCGfk-b|1fX=5-}EbBmivLbbNwH8%afFJw z$t3gpNo(Z&nJ6eZKo^G8`c;ZgN5OUqT$j_llL{oM8Fa&h_ao06qV*trT`A5{<+xXh z+vr=;%8+8ZU4uR4NSwBEZXaykW8v8bQc53@KAekJUtPRA1Utbi#WkTEK=QKbay)=u z_HOo{;{+h8>t~{WceeoO)R(IMLRPsjfc97cD4AlDD3UkAk=>jxL*b~(( zgJPPA>6qC6t$%){-LUxlj>&n#2^LeRZ?|6gxdj&Q+;s0D*IXDD16HuON3nM(avL+8 z6=C`6NAqh!;-k_QA&JC#CXG21btbS{S<5%@F3RuBuLaKp)hqh`=BD((k~St)R;s8M zpY+#}Vy1IOAMX?Y2LAD&E?Kdn6qpp%%CgNF@trBOyM7Oq!T%Wnyf1EJ$H?pMT+_>`3RLSKu=(YH!vMNvh-=8yEF=pl8> zd->ZXEu>9Vl+qcjkv<)zQ(hAH!>;LW_(6r9ZWTTXuM{_FYDl#JufU_2iyZJwVeT>8Wn~_h z;VdS-SBokj&s;%C=Vm7(SEr#gAyY**Q zYv2Qly+@HdR7^kFM?dA|L{y4zdu~`5KeI9<4$_(U4A-H3^a{}tWgPF4Y}=y2H`_IL z=Ux);CEF!BRUwV-c}*HTzioDI$UYkHIm|DK*f4K(M3W}dw^p@+&5O+C;c-3h0?-%N zh34`KVRd}U+pE41{WMC?%Z80w*5Z{w^pr&~5P3Hwr+<5^@*#v1y)P!e=egIA{oR7BM)4*Te%d-!8}?~cLeco#v_ePg0@$+NV8Pz&Np zI*V5jy$kzMbygTUOtA+kQf!tY+WM_V*j4G2Yo`@6iO~i{zGsEUu8?)p+UI^k7kC>K zoqi9YU<=JtKiMAEq`Bxc-@gx?SCC~-mv;g8#`T_t7?$r@4Czn zqdfJBmat6Eqk$keCo#@K>2CDSv6L%OaAM)Eh~-%-1B3O5t0K~%Tl{+1L#T8GekW3| zc>n#wvf5cYWOu{02?A`saTI%(((wi( z{YcsaeB4lH=3M}q^y6yja6JCT_Jv+*rp7p#el{fUeT)Bd^#`jLkltxzi&eUN)l@ca zDjB3=R!`ePR!2Ni8_1ezgKn#*_0n6&3a`Qd^Q*L(-3vOy&Cd&no|hi7nSU>2%`}iz zr#r(9%-U(21liCzcTkY|p;yQ@7k0zJ>mHFN$)Q-_B+Q^lnQ{&Le|a4HaiQL$NY-GxS(w zD6{Bn(F)mVc@Fa!;$jBdQ8pZCQ;N!T*|jC--S~g`I}18~ z*Z#ZP=Edis9& zr5QzU07qO3gKTs51i3ller$)A`_WhaJ~-iiN~G zB9luKi<(@Pw*WplqEj7JXmgc=p=ZhU0IeQ$iBp*M!Xb}j_;gt+V`QVBcI8}MfO*); zuuUMe_0+o{Qj59a_V5OI55GGQ%M&mUty4Y;U(aY0B;D{%ST&3FQ7-HMul?bw1CbW= ziQSj%B}Fc5+|F4UwlZFN z8@Og#H*^UW`<@O#hNoN}rax8-k{6!!TxMXhJeR@0j{=)Chx{wv7<5YyTpx9TJ|$_= zJfP`naNa6PpZsG2rMn^H+Up+ZY(5qIK%& zwYQf4-m>CIWIu@_U0)b~C&9|!iJi(q(tZGWPMbi1Bmp`F^>cdoMyP+r|FVEe<4{nm zun!uLF9{oF=u`uqH>Kut=Bnw28Rm1C&_H@R3?Vmo_weJ{y$tffJ*B8}Qn&aKP&$(9s_LJZ7V*ryFKu zt6bLUi9G(d4`*5+_?`W~`Zuy`8u$?-_xt%23ju%}DrT=&F|W)MOamm%=q9kt0ri&@KdI!T=yR4X+KHv^B?3Q{VL$!A`S)Nxhh9>V2qw!j({1%*o_|1F|WA+rsvGr!qKcfT$1oE zT%>4&;|E&?Pt#@6R*y}BUD3zIi6VfBXDL6&Bwhm>YuK>=sf{s4o`)Wv)}Iy9IRS}m zzvsuo#>iy@bg$ntL27UoeMwjovW_Vfubr8zx)5@jt`KC?rBRMHR};sW1$K^s*5{-A zwNJWSHXNvVDE=);{KB}+g;qu>gJO46Bo&33pw}lEe9Ls?dBftdRwb(AV5^3gmd`aynUs zC!zKWatBDLkun|jDp*l}Da+J(Op9_O2lyGt;SJrS7{O@*@dzL#fW{wf zzpPY!9VB#?V>G!zUaC&=stiE=uR~;O7~rWL=nc)e4Fl+pBttC`o+yvp9|9^|VFlhg z)y;szDo{&K52}Z#Pzs~%mZ!?fm4#w0Mx^l*0=uiCfIZ^b1T7!wbaPk=(*|uLcoxFa zdm~(SkHl$BoAThP+wVp*@xE3uKpX75TbVeB_FN*}zBv&W#HK>AW?8T2Ghn4mmtaG5 zrP$F6(CA;a)oU)Rlhut*eqkFe8~$ua|7VL)S`cTPN3MTi)-7Oh8Bxd6NwJW8ZAa}R zECoPtptRK^DdeRfd{Rm845@e{?hR9$&j}!7NeI))j+pYTn6tw84yKEc@fm~!A&jD4 z>zxSoil*!Fc4URF@IW$fGhuXvN0kgK_LI~-Obf}AZM3SU#GkhzkJ-}Zh zz6LY=EWDW4yWkX^6It%4rYcS|Vb?5k8qqNqbF&n_->U!Yb#GK?kekKQRjcvU)G?q^Gt@)*vs6coJFaJE*{C*bb`QHUr})JIPAXPU6%(vt2;3vHx%k zB*q?v6Hunk{`ni#>L?`NgB6(vkTW-f%^G~ee-^b-A9pZER*PgUb!bwp_vEz z=SY%Ls>c0Hncsj{BG86mu_)|rNV_7@9c5u#>5X1{SmP`d4aVA>;eyh9YV=-p?D-ha z(TQLSAvd1)wC5h#dll}Z(el!-88?mHKtfWbSjB?4GR8dN{}BmL`dSu!H#`rv|2L%l zvbsN1y`_dpf^3$s8<92p9RLuEj_1LRDz4h9ZD?}GCz$?&gw zyigy$-&v36SN`HluRHN1>Q&wHt%PXU6JE1yb3iSGPOtGYXLr&_HHms3n*%=c(8RTaclpH68*(%M*DWmYPnnL%JRL+;9*Dc#`(=A(o&8^F ze%E+ftQha2pYYC*UeB{Ya%=W)l4evy_N#&Dwo&=~0P?hJF|_~u5Qh6Fw!At6Miz)o z2u++|G-bsAP=Q!rB(~H&xNPyAjY@`cPlFwc!*Q~28aZk8gFZsB zhbU44nH6~zY(i5Q+&@5JK??Kdd@7cT6M-(vZGN1nlfExMOQ5mqHd7?-a%%5qIOXRWQg_PNNzLN3G z(96~V^SSYS66J zyI$FgR=Wms;6~g*#bcRdv-%u@5P7pNgrXHxMIz zNTLIuBsb_*NG@!O`aSd2ZN95RQ9lL?5t^A>A*QM(tYbPXS)(Zd>e~Gw+H*d2plPM2 zHGPE*tg!FsRy!PBMm(deyz2 zlj7k#|9bJ6eN%6J&t$x^ZJ!L#5LQWz|G!Vc`=+9*7%Q(j%1DanMg&n!A$eFaMF&=Q!cfbstO;JVd8 z3l?m;f&agqt`x?zxLq=-Kx5G-4v&WBmJ}VkW?7yHn)Wj!4bUB@pScPM+4N9|r8SpP zZ4IlFB6|RijzeOt4a7MIro&aH?*D!$urctidw`ROrl$dFa8Qi|#Y@J*t$o09guQsl z7-JySHn7I1hnd2xywv&~3)tTNpy$Wr!!OLj*I@p4Os402w^; zPyjZ<-1Rw@Wq9$6aMTeF6YZ$Cv}+u0Hs^2rf7$A7u{qzb-hG*DJGS9a4eS{ty#P6qt3inq|ny1b!`E(JaGlU7iOg z{VTiXYV<`MFlD_EjULFsul2pUrBf{^S+#a<9T{?SS-vosIU~CMURaEwm6Dh`Y;T^V(VoYQ{OM&nVMr6>J z&4S%b;hXqsQUPI_4?9s|WA2=HI5|tWda-2%Y5dWuY*OyR8)?{_jPM-JP%NY~KA>VU z$x~^q$L{IpMS72F9%^Fl^68LXV}WY~jt;uvzaX*+d)0o}MVBmG564h!#lYnA@my|T zbI_pM6Ot}jKG@<0#EC2ohFTWH(e59w_qK!(TsL zr^K@p!IYio7&aT+9wz`EPHPWUAHK`$buaA2wFLZ;SkZNlL;hIRmEqnHX%#1(z}Ly? zR{Jo)X-J1$%n1`yKPbNOU;kvWFBj)sokFf~+sC-@s`J>&8r`AT+Z1U7zH-Q9@Agd; zpbHYumPVSoV-NX#sA^a2k*(rgaNo<C3F zW9-MuDp`hlYZUN>ra*QAc|L1G;RPlI@Z;&{JoRi9B17jc_b1^xB?^TgjX*k`L$9CS zPcYp(hFP-@ncR#N_p4t7bII)G18-Qc7WD5G%g7FHSaacalQJu;?W0(5hWAi0o&Ig| zXVO^FW~LVcZpWbw!F*%@dm-sw&u(&!*Qham`Rd#HcRyPCaLJ}0=)Ze%$>zULSb9-| z5}eq^FvbMG_|=a0$V8wr-uJZUzMFsf{jJfQ5MPjOk_`rl?Yx`PwU}5?YS%%VHQJjV z^#T5Mv7SCBg%uBXO&J2L_Bcy2XD9LXkUi7|47qj7bpi1aCrF1hLpp2|s_1^1_Bv^h z_lnPw^+DQQ?m6x|iS}b9&{AMnw*oZoV(0GRY2(>SVYOhx8#(T#QZA&aSVt=9Hb`PT z|F1z_OESF-KxJ$i0C}=6cwF!};{Wm7PP$=9j}YrpE9pI8DU6e2)De=CXiH{XgP?wS5GJ9jdB`G~8vGDePp8_XVf!#1+cbQyGavsr z|5m2R0+9Dgr~fzE&cj%;Q+g04UT3~2Rc$}Hz+yap z`N_fpQap{ExB87%Q*0$gj!`j}-@gC#tgqf!GW^)2sU?p|`B(n4V#(GYURwg%fj_JI zwf?JvZfn23^M@_?8jN=8q9U)&AZ*(hwwXx}K1KiK{VQ)bX+YwXE?9Uhjyx99q-l=K44IIH=lJDY{i0v>Vj`2n z>4)sPzr)||7+m@-3x-O4p8lMia^VHYXa(-e6x&FV1}f$@a3`X}3&I7kQM&VkkCxv1 zTa4bdiAzKD9%sp&`ADFb?Ym_u_PX2;?gk(D**|Q0`?I$bsn6c|;Hym<0b;0F0T9x0z^LDtfi+)X5J+E@iQuA9R= zZH2a@6bn>E2dS7O(n**2Zk>MJ;~b!BmFFEH#&uv>^GWr&8P=$I!rL!u(roiQC~eYgges>d z4TwpJp`~tmwduNx8Us@&=!RmNXF&&pc0wX6pTGUhThbPOhUA=kJ&aI8>ceoXGwvA? z7d4#hip_bjoNy#eUiDsF&fA8`7zTw}^!M|R2UzS*{?L!gNQMiqXyBcV$Z{1?Yym~` zP$xD|ba38d;coACRSoE;-qtL`kOv0wuKTW(C4^Rj6jZF}i1Y$&01Tnf63U1_{gU4Q zBPh{q2&?49vl$Uh8ocZ3qWVRZ%DYZsX>$@Yng_Re{F%xZ=3h|kv12G z3@8bYfJ_g?c2T60iupLILH<;#Q>_jLMnvDskYq?fI^YA)x|=7tD>&($L1)kf(x)FQDNLz2UD63EGDB`iF^jBhjWC9`nwJ+~P@0;}YC`ZRNSMJ8jUx%z9D=&h zp3nqd4}+2u>ZblQ_pg6H+k(1P@6Fgm(!VglK@cSyfzbmL3(>+{DrUHvR|oN;0ngpO zSk-INrAuL2pbi~N)ByomQe??N;$Uo#$&+mIK!K4@JugO^bQO){eCU0-e$FNj^FfC$ zf##WX48Dd@;kd`&PLIiCEr{`a(sz|?AIqh7-OVO}4%892&{J#{MKYy!GxauhDU38{yj(^(%hNjbn;!m-#`BF2-|Ngr+u!&YuBhwQWl#U_Z#^1K!S)4-E89?(FIF6n#>Nt` zAkceaMA(}D*B}*yfP0bqkcSi zH#;oOMEjH}8zMr=e_&a0epUSI0n*^Yp78^#)un}EZ&0Keiq0iw`m`IWM8Va?2Z6_N z1$%d~)*P(KRV8}IMHptZtISMm+B5D|{4=}{pxZZjVY_N)$Tcxo!v-J(ei9z%18pvq zVXgj~R6}mV*I^*2Uf3y415%Tk&|U7kyw5{{GRULxHnH^zxB(Ji4e|l6MrEVY{Fd0^ zfZEb1#GtJVLrsT=o+oAxxm|`hW*pVVR4sX=*vxb?59wn|UW&W%qh5CP!uC>qo3aZ0p+NtQc&1l3uDxT0)IJ!%=qRmUhLsa zkQfyO5CC|73%A~1hv1`UjqD9iTPgP~mC;U>gGkSI> ziYqS0X!Lu(>@_!BshhX{OlZN?(#zXFARAm5S3nm!Vk4DGvD+z%z+T&{oz@O`4R~QO&e7mb8qRQ<;g@~OuQ40z^>kPDLI&wnB_!WA78c`q*aaze+{ zWxozz{0)ogIQgq<_ed?bNS+HvY+9`hO(VrNP~;pHQ|i}GI^>NDbLcpg`FU^PGiaaL z=6RUEnjLi8Le@w1d*ZW-uzb^z4`;O!e4QcKHX9%NiTSaEzU+sux6SSp5BL>|@p&KJ zqsGUCc}5{VrU#yr;A6|IPSg4AQRc@M_O$7IA8j71RD+M4EgM%5a#^oNUdsi5chwv=t3FP|4 z;`QU1^zpn8*Lyfw9;!68aG%s-c@%&BN+a3)nni>m-#NmONvGId6xm6|U?ol8{O&-^ zVCDurWfFLIz$Mm$TZ@{!SH#u4L}2bN;8ziB?gRP~hl`+&J#XVc&DaAu_^55veARv7 z-;So#a#_#C27Ok&`(EklH=z%qN}!)}SQabNmPMbFA|b@2n2M2eWiJ3m-ZL4`Z3AGe z!1SNKOa;!emA6`MJ%m$UU;M*cr>9%in2vwF{%5kD+osrs;i0#J$4-g`hiWSogFQ|u z^RJ6Ly>v>fa6(EN45_Bd8+qo=C~T6N98{R6loK|bf`YLJ&8ge{fxv=|IXnOUG+Dz9 z8!j9$fr^$98=7qtyOkn|RLu3IH=*RWga_4l{2r3Q9+aLJ9S{I*Z}_0wLFw&DMFq@> z{lFM|0-bgAOKCp4`4(Wf1-q>vtGNNgg&St~S^*=4Vz*FaGZlkW(v@@z)2prohRc=n ziul@tflxu&?R&zvLEbq@?UKXuX+JKUr7t!7Nvt1F6#bA6v*2QVN_P<{7%R@h;g&v1vD@Y#K&5`t?wWxuSIHlNP)dIDWckn}MsaviZ{ z=75stPRA>sS_6!CMyml+XF^wPe2wZ+OpGVdWUb9-*LeGNuINyRtmH z=@eK>ni-?w0Dl#*GZ%%NqBp$z?3;I?7xR{M)1sE}R$(6$k^r4114&+Zmf^cI6t6!8 zqANJUhIj8QY1bfcZ@Z#AI{(dLNlSPf^qj1>jCi*hPC|3Qj!zT3+vtRkPQ_9;(OI zqTmXmQP*?8`b)gpQG+_*W%73ifBw2xAys%Z$g`oE@kHnm>HTo9W6B2cst7&Z2PE)~ zVa;Kc;zU6wL{cZI48Sxi9Ov5sGil!t+-%#9-~Z+BMt6s~tn*=`6Aar`F&=U_dXEZw zM+-cQqE@qwn)^!{ei6TPCv|aY?AIH=k?{6v6RF{WMTP~f!bhIR#c}RkGg`zoq#?|S zUMEZ%N2H9j4)#s?IIrS_6zV_U`RLwgU0NIRuc$v{mW4Lb#$Zdfqd{2e49S#^*A!;JtIW?Rdq9B1bk2JDIg$g7>bXJPzJ0#h(V83pcrc=2uSDM_+KSR&5d_2V|?Z2A>42 z>s%f#Y~bL0eGh$i?v;h@bf$B;cr7z)Cq}FdoKsL?J*fA-q@x-OMvM_NQ^_$G4#!@x zn)S033&c=$RE$|A?J*E!lr#BInh3NOSHzw4A3@yD+;3?1W%6ZsVU}tOG?kXq>&SBL zUBT9{R(ge~BsdYMq?RLji-AeuHw0qseGh1xWJD})VURz!G7L&({Z`45qkH+*>78P1 z&PH~`qpECQXdSGl>m&E@@YVtEwW5kh&P^TT!iP0l?0aQB{jYq*ZGluTR=hKj@T?14 z92?=;D&d|ts=)=@$Q%@Z=7&Aa`$ZLDT{Akv(aNlywtH6Ltn-p})6(hmxxk92s3nJj zpnR5pSYtlpTqk(~Uh*rEF`)}0s?Nz?_bd={rW}KGU!Dl7*oJwrVb3h@C&=clEmszb zQo=xC{YHvgpL>>d;tc+QWekn{BaE@wmT9CySMHTSy*05}F zR3VX@9^4JW@u;*gAVn=-^ec7Bg2?vp1mNgL85sv!fAhECxUYL$->S`hzEo<*tq6sFui7ilK-^kMn-FC_qD zMx)XQ+Fe;HthmdkoT@Ku53t5uNK6Nvhpe@K{>x8B>ymTnA(ecoh>O`=Ew z6@v=Xs5Y_lI;RFf%!9<73~H*$)f*#>FKinAm(8J=1{6Tx^ z_hcP6*Vct$kz)mm9Tb~Fku3m=fKG6b5(USF7}3>k_A<`OlORrp+_2*m3ECa*7`=41 z?1u)Jh?kD!*pq3Cl3xh&zgBx;8=?HPh?X#n%W4aJO2L=btMqh{Zwj+RmK|IdYJTjH zCoHOm4xa=l$AV@rBe3^3Dmh19IZp>CB#c-@IKL6c8*DozsVy06M)SJapigTR_Jj|3 zB02tPNCY1asugrd`+Wx8HoUvy2aWQyh>Ot&#Rm5(x=B;2PL8e&OQ&;#^mMN3LP#BN zXn{_NSKJH93alZg6;3nW8|K+I_k085m;+klc2jIBMYcnuhi^YH~>4^H5i#9fT%H*`N7vTU0Eq2+s1 z$Wym593~gu#KLMZqSbs8#jdBw8lWNY1zi^ODh?>lz2*Fk$=JW&S5Nxb_c{BhZD+ao zvje9?Eq+ExXYmT6=jLa)Fi`5OfO42(p{%4BTF-ekp({jsx|FnI|L{?|SA9g?7^eMz zZ-gkw2mA}-JV_5<`)9N%=$WGFK=)9U#o`cha`$>7UPKhTm z;#$eih*;-1>ikz@YS{I#@6l0MdnBp*$>_rS7lR=#j1L>d_&22YgtvTS7a5p80zv)4 zT@gcW_d;qy(-`9{ocgx#Cgq2U6SK95p`6Iv$P(U=n~~0mz<7F>`*Gz#f1OhUBNM}e zqh|lbactX|himTFkLF#oL2rh2^a+wAuVV3O=_)Th4WxJYvsxj_kZ6m=2Shh#Uw|He zC>T_iN0&>Cv#yc*;hpqter}NYOs&AcgDeNG1ErDe&f@f)*c6z5rZ0Yh6F*{m3SO%v z-mq$5?|)de`SdvM)4A z1YG*7AZVDvZ1d15cL7TQgWe)oN27wii<*T`_%F`k~+cg7<;_8c*BRN{pja=*eek8k%TZL$$$uZ6a4 zhyOrHa~4^!6|{2WesYAHX>(zB3P3huUv`>e!T+uSQCGog!A7q-=xgkPO3QdqS!XihD83|uWpS6LI%0JZ-;%wvU-PMKb(E{N27V8VI;(oZSVhtKSf z(9cQrIY&~aUn6;YoL_JX?hZ)O-|*z4%uqM%3U;kl72Uv99~Kc^7^^!vZDa z;_#a!jT@9)c$F--g3^A9g|JUHbR3x`_IT);2%WN7mMFN)gPs!uvpJwT@B;l178mTx zL{7>I4@6a|8VkU2`v%KsKP;laf{Kpe3F0)zkYh}6%ynt3xt{p3gZ~lLVP}1%IS_wS zm=~DKyX<#Uc#?b+v|-*U`VeV?pz99VDz;wU!!+?WKq>3T#3*Z5B)T_gTInibiEq99 z6x~WWeMMtF_ywG>I#nrsPdpkDT#^ee?7`Ro5XL01rA+$?zhrJckaj%v{Di;!e;~wcAAlya zAj*uzkHVh5K`d6H#4Y1vvet#IM7EWc*iNy@6iK3Dnq_N6cbOr#+x&~LJ4H9Nh2%u+ z^)iK@Vi%y-IjJVY&bc;u5DvEpFBy=Nv6yo9*PniQbSM%So7(4f4SGaS zs~!2Rf!71vN@TunmUWX8v;QA705Y7QG4%&qzH;baEJowxckTxeg9{st6;@%pUW$E4 zk^58(h{*BcS=@s_`-TcT+_P2n^pUv@%3IQWdcUZ|_q=4aM5ny?c0=IuJju~uGp9^L zV5z#z7y0r_$#wB^XhVYN9&%3}^lu2Pr`J)-FNpW@vu3CIb<)Q|iuudUN*)7Vn>?~r z`Lc^|>y+mqyaFR$j4p(%1LV82R7Znb`P~ql!%uNt+$vll!uN(w`{lR;ZxvS0xF*gv zwLd+fA1Uyiv`K378^?(Zo)_o@!SdzH_4F>E_<2ZVQ)vADHh^@1}Y|%w@0;0c~c6tyOQ(#OwWFG z0WlD5mFi<+E}GHvT9`ZY9s?8E@C9oG&w_MHo$6Q+6r(B(ka0nI@RR-(vrdu(UbR2K z&A=do0}gO3Wc7@5^vbz~bWQ+12dy*NZS1b|z7d4VoZa%X1pklPoWq3}C`bd0XSmGb z`0Ks1;p4?eq%dWEK>9cbI#n=Etn*s4AjQN9eawCPm#}x+hMoPNhvwKT=J)S?rG0eo zw2gM;1l~R20YxT%4Tb;i^GaV-4QoyU?-H@`H;?^}4moP)*XfY=;xY)Q1;%wZ)tQ>D zH%6nvWt}1$aK*?it~8igGVQG)-ezD@{VHbtKe!G}tN0PZM zOstI{nBi@^Q|gaZ_HH8 zCUsTjprY0fS0JaO%uXBlf=@F%`ObRG^_-bMno5~V_{D{z{x%q)%~Aim-g%INhB%&q z>5OcEr1A-Ip}1K#`D%z9fZ;3*j5#jz_QP|pdl2R}=r$ObD#a;M^|#u`U&f#Q4icw& zCNO(KZEwFHNB!36KmF%lFNe7XL-71QXFub#vQP(SF8KuM*lpJ4FzQkYSOp{SGo2?8iC7Pc$2*TsdAd%X-m z3#d~f_aM+iNhT-b_iBJx{pw9GoC%EKvAF>xHTlV}Myur7ke9`x(FQrJ=HhkCHqU1; z1E$~~N?Z3ZjYGadkLh;U! zkLP~iSLD&I!nTV%(e|iL8bibv=zemEbo(~Tu#S1M(7~@Y(F9fZ+SR~`4k?o|zqJuPd~`75*&#QqDrxhr z7p{WrN@ZB7qF?sJS?wy;Hi@%oo1`U%8+LntRmFSV!yPT;u2&UZFHGQ}7jiE|+a8Yk zl?lAnJ`W<3`8z=qWYR1eVuTAatTE->SKN@X=j{!9-?SJHH;L*n$(}|&u=0%v#k?wL&J4vT12dw%iPd63Ts~Fi zhn*F9K&pVxay+!xr32G-GUQl4K&pa!AaVa#Q4rZEuL&)hQxJ*G8+jg%*QawB_{YZZ zxR18~>35q)uR~l=8Y~ME z)f)FKkLSjPzaP9XdXMu3L57W;d|Ci#x-1w9N>}eu9gE7QW9RC~{-`G9fyIaU=Ojgo zjlw2Pm9Q?)iXp4NUCZJM#?-8r%;@Pj3APCv1#olE--ubyEGK(b#a=aJ!A!KETi5p*(snPn#!6 zVLtV|2n9l0C?IsuL#?cy#v;K!UUI!KXmq%&XJTW%%>{*^$1d8$H(;y}N$#`P# z=i5;e`igP5`6kC^hD3*07NI%me@!AeE}XZiwXz0f6kAG>BC{@>7CZYYJRXSq)TL_d z@H-C6Kym;wykVm8I$xVI_ck<(w)q+(TZQ}SB=QKtV3lEUf}_DhZV99#Qm1?!YBV8ET2cdX# z=Yh78UoFU{kzxiL5)OhWta-UVC@zRR82DJB#nw&q37~xwGRN9Ha15%UODTmx*%Kq} zl(G(oOw0JceVcc3nikjf37h8sKq9vonh(GI&JD8V3*+sA#@z@Oi#-&ZPLW+`mUafF zFj%E@h*Sy9L`eku|)@=;tz{4KY z|M}@?B@Qm0tP6+cZ3v1Om_bE@yuu?LBo1qY*LdB;044P1qrim}P%pa}eHr9L%m%W{ zw=SX@jx1Z&AlJ`1uIiR{$PKW?>4Q$t=CG2vPJ+((;4*UAav(-V47KB+q5Jn1nMteZgB7^yWJT`@7@*KK|}e%tfHP!2u73BOZ&ns>|@T^{!ZS=ek8(-ki% zq|-x6pgl~_I~IaD=Bm&kag*jr_;v9qdc|}-FB44P+5ltFvSl5SMj$UZr1(_mqy=o` z%#ROHBS&z~^ZGCU=35gfOK@G+ABZ4FI>?_BIAyDlzW=2Al_i_`ka233rAvAk?ba~E zqE;GSK<}7iH{P$BH%`cS)nE=hfBw@LuhIN%F1yCqU}VGIUkun~tG2(n0$Vm<4Vi== z`-`(+&m_Nim-CNsTTi}qyRUupM(MRuayM+ZH;DFVe*j0+2y4knS_H|02$nJ#V$PLiw z=gt6>Z3zAPz3YQZEgPi;?|;%tDqJ|q+ibO3d`z()Qsfj==YuP;f@)*ps7H#=6xpyR zPNyF)xHJD&IH}V zJ4jaG4k~l$fp1rTi@yfXsVX-r-{T@zV&5ND6&ub_s$*VG;Yt~Mm?QfnASu5Hz zt4Xs~l;+dz^Mv5p!yqqH&2~qQ*LK-senw-3)9Y~p%Qkk#V!vH<`)BgeMAk05BH5s( z#(ZIuW@kvZB3D(qunH=hw#{gdJSffQZvr_({1@r3^z^ofCJok^w+cT51pcXwIlAPL z;sTu^$dh2|2`?Nk7+yJ76Gh7_-(Z{G~=tpHF!-cI+rIqz5qSyk8 z-5bqZr(%ISRB)1@i=3T!_}VS2@#bUD4-w;)n)A*ZVrZ2s|NZcNyU4FB#y&J9NueQ1Aa`fWOXTIc6ac zvrlLYfo4S1!k$|@KxZwkRwY6F*z{4M6F_2wq}w@y#eUiwyavd(KY`DY5dry8<}=w< zeof$WugXDSC3J1UpA&ifqL zgI;w-P?IJ}9cPlc8on>?jruTL_A2Sy{)OUr_5q3a$>2E@TO6SCGFxCe=SVnlf(}(y z_Vov&^W83MRBU9h`^Zgc;gU9{3OwE>O|Ga_xR;OZuFXCx7MxmmmTing62P*hIsEI2 z-63thjgeJ!f~3@W{_ykZ93M!Y-|wtv+(7c?4+(#C5zQAQEnq_}GGksY|Foh~{ES)Q zwR)Ptcc-e$qrlssSTjwh#C`iy!8Q<{YmBV$$nfp;!zrE2gyew5^SmM{XB`H96D=_#cT;rOCc66LBi1jm1(+zY@D0tGyXfs?Jh}Y?qV-t%_ z7&{|iG~TCUVUCk!ni`(wD;g~W!NoGUa80=l(TE3MN|0oW_JU#^_Nttiy)P2P$8ZmK zC}h=KONYYbu^GV(as$#>Q|_E&?|icAD@!a!=sR`)P(+Tquo1dyWrRMW*fSJ4iP|v( zp2!+@KOA-G^t5S?Gd+7i(Df9M2 zpXLTO2W^IW3B0|`FF627C4d?qG#-X;`}`(7C`@Fq8SI^TI^{-Qfr-uk>f)-<@vURr z05UFHGA;yhvNTh+i)9+~#Z67|#=pZ=%OB z3tX`=j<*=sEu641^%te(IxmZ<`Q>%8j3kYv>+8am4+pHwPbS4edOQtfF+oTMiv6%L z&0Obx_J4JU0?lP76hvT=p}|x)3>;J_7lV>7=6VR{!-p$wELLS)Sm1J#ZQMxSij~f{ zVC20DO*P4P;pGTYnIlTYk5eopr^+z9fQt~Qo+e1rB^aGuE2tD-3*G+a(3cu!8yGC~ zd>_*e+j()K13=xfkC%YbR`d$_MlUV)QEc&gD&5H*;MaJ4Ah5)Av9LTA2-vTC9P-x% zY@M#{mZn6UAv&*gI>#ZLxZ{M%E7l<{pcwmF+fW%Jl{w`bKRWW)*FE@qKez!DgLGci zGHr+NY5J5j4XVU1(D#EOz<(Cx7|zXTP&(gqJf96t$asF6)1Prd#?&9{e{pd%8;r|F z@NDqAREDL1&>3c{vlyc+g~5fVi7kY^vk~fS^d@mJfJi}PElHJ67DPC~1I>lgfnk0F zPVkuWoqzt`v(c5`Tz1GV+^b@vPW(djCXbAWH4C=!GXy8dTK6Gwmg*r>Da1O8deCg| z;h!LhlC|DVni9$J;40am0t+Nkm}~C2&R62(45~@;T!MUhe-dnQPfoqPrJQ8BuzONt z<(?d**kXzlnmgIrRXzNJ{z%Gz-6;dG!!_xwHq7XtwbLxn^PHv&#D(E@Ju%vc#fqX8!r&X^h+FQdAM1ZrN8s&`>_Q! z+y4CZuaHYF44Ymn*xaVrHj3P$VroK)L`5R(xhh+zZ4j3FBX5U+DO-3<32Y7J(OC1a zgI6BXqb`qbg`OqT2_zv}x2OeHsAu8DJoBkSpG1M>GmPqIu@@TR+JLa>Fq zTdoU;kBE=j5&;zwxA{#Z9(q#FkWaijX)S)BjEL=$JP8y;G@5#h@y6r;{CK+} zk|b#}Dk34=i~X3meJ@rN``W6~aVXHnMBgh9jtNh0K%08@jQf9BfcCqFU(O--To`B@ zteX3tQtW3G>7!x}hjc?_6_Sc?mt+V&@x{Wu{gIVIopRZLXRFslE*x_D9D147m*|Zbc1v-jqV>qk|j@iWMD;+Auwx?k4B` zHu(%PMi76+QR)?sqx#;;06p{lqAx~c-_%_`?;fTs@O$R<>>L?%>W}TT?fTbcxDg1d#-T?o=NptCu-7cAuH$Jl0id>pH7(-StT=H)jj__ zjD+%+d-=B)mc-mz(i!ZS%`>-288JV$pE_jbY){G<{>!9?SaR%MfN7j}RzTg~zriiM)j52%=|$ZhUc9PmTKnyM@;dgnQ z4~?C9+OJ*Gq`@pW3dMisS1U+iQe_X+hn0H+n>48YiGt_5LSiXx($Wcl%1(6HHghjM zGk&0OGkiOizPERD$oLBq6SNUvPEtd6K-hZ0Grv;BNpiwB7J7Rx_@fLgRuk)#`yyBI zs(A@vN&FoC1mMAK0=a^T+vwxJ`fog*_Ka@iw$U$*RcuDOIQXGQsXB$(I-{4)m_BLW z-^jnvp6_zx^^^6S(+|^iD+J=D5#Jq6VZ_B^xb9zOi-`$^0~?izf&@v4@5)6*3$a=g zYA}7vqcPy8y)Vc9C!G@Le0c?t)&3al>nC}Vti=#kR^6F@RE6Ru7-wk}VxcHj=(UI& z!k~OO0z-aqW>GRL(Zb-C*gCL0k7N?q>s#{Y1oN90U6LN!^*gdi#w*x3VJN&K( z_}9g)P||@nxNg9ve?aklivuL_`}ZlNo|{wb!VtV~;U{fQVK*0CeKb`(Qlm_c*JW%hR$zLC(Q`V5GP$&mQ8b7?@1He!3f@chbnGH;) z=g~l&GJ)3!JNUzLxD{ne>lIzlB3f&TP=6+?1V&r}T zf%i&bu!b4lXBKenV*wB*Sq{%L%2Gw|AC}e5($l+q;^*NnV>$fK=0aePv3{weG{KX$ zjQ`DljQnPp1!C({x{FAG3;Wn-tw2*jvA|Vvn2IU!&GyB@aO9YVN}Xe1hmbT2`Ldu0 z2Z>Buk+cmn2RYEic`oEa$S(H_-dYUg4uw?ALQ8`JWyo`WDilh)usYOSetsDorQNfO z_^tHOU{k{qoK0bt;kBpfXMVSr zb-&{*i}3l>juRwKFaF~Mk}2Q*>kq$0TUMRE?**JCE4ZyXE}TKguv&GtQ0!)kY@%YY z6ul-C_b-)nmI^%FyRhFyepsg?=5y!_Kd}=+!KND zYG|HO>;Od`V+(>#xizd#ia`)mSIHACuMW&rCQHhrwTbTCbf#)(L50UI_k$2L(Mt~S zQ3NoZzRD)ORRa6gT4|x!blL8xoQOe%fp?be3?Fja%g14{67Ycv9h zk!kzhGS0fqzewNmO{e1^O@*gQ)i`b?{jm~B%};}580@Ga%tB-TZLhl5v%;fC(FAp}0HNI{ACKawU7jEM#+UEIQdpaNk_416xlHx!9i-9R^dIx zNmUhyL0y2DoE9HY1t<$zs4zr`8d2x@wF28Q6GvJJI}MbT1MJB6w>8#XRnEN8uFZ=@ zHLjbt2bG_Z+3n9IHg*O+Ci&{y^P1V;Jp5I~(#Eeo{pxSDze%C`-!U%vO?q&t`mm2a zx|2?ot)00!V2f8vcpOCh&dtc=9|}3*EU^U39E=AF7F^hxMTZ=>?e6@%?)L9kixu&F z(sz|?=VnD*7(T#+G@>{{PqA4P$wX?T@^9aiHfegKC;e~8vMJpUjZ1U^t+4JG{O`

    E0pj_sIvuqy`^Y@;C~~lh6r5p^bq>yM253nX{ei1Gy@Oq6bqid0UY3{p#J9$yOKk zcc6t~#N6+tSnzXqQ!z>E(kP&$2FJEvfNhcx-LDKeBsovEc!BqzpM&j_b+QLQ=m>#Y zO#0k@s~$?qQb0Mx=^CcvpLd6t2%Cf0wjEi!YHY;Ldju@ zJxGya^fR8&x99bfZn`(Hns=7eL5Dwz^4t!z5L|a*NA~xxpdc3V2uTefGo&-SNlKn z!z`h5*PIcv%N0x`ZguL@m76~M$6x%@&BNbyN zhHVvMuTnd0gmiH^UEq-+FzKK|_OTqc&v_Com4M}{E;OIUbY!zk$2$~a>S1Y+W7B`3 zI8V|R-U3-}9Osr4+#k5FpJ2WqHTWSUEwxB6aNnd+H{^CpnosAwc~g37;UG|C$wNJoD z95W`(6>TpVEQcI;5x8gGJo4+;T|XB^6HZH0n7ebE*<5}LQ{(lRBs1q_I^~1NirLTn z9{QGsl!aV?CaOs@E*$$09J?I7ILptEa9gTkH%?zDu-KwGJOBPPS;Ng1x$ve5_DCaa z(Kd?RN|8h=26^*aBX{^!gf(eeg)7+FS*sT2t8=|ecr~OX;`YMHLy&3C>^B?Fr+6}+ zy_3A}(EAo(T>Zi71*F%7z27ZXG2vBH*|@1>kcw%8^6uTYEW#dFQ=yvfrX^r5# zNbixQ-ucF(s0@%s*cI}SUI9F@z%U^-67+d@2A-mC06|Twuus~fE>$4weVXKE(4Bd= z7p{dmy@7ceZBCr~7T~iz=YH82Yru?>SW%4>$HezUzx|658m1n2v2D>7 z7+eR+cE0 zVz*Ny8OTxmGbLRMbkk3%pFcxdbXZ2<3?4^D+1w!KBk1#K<_4wDZ*%&SAKzcU?i~v_ z?kuQ&K+d?ZPtj zls*gD!%L^HKu)v&|0C~B;F?U&wQ*nZ4atKc8-e5%R3O5Ds4RvNanR{3?U}ZxJzY-E z_qFHr-*)I?PkTD;m(J_TaqFmHWA$d%3RA^zhNKR`HOPWSDoXWz{UA6WuZ!}C4|W}(2p(#VXFB)N z*>2TyO2xwtXP4|24m*6dutHJcbV?HIbIC1LFk}PzEi0jh1s=@1l^tqMQ*SRfjx;}? z7`Nkqp^12Mv;gZ{dg#HRc>i2)q*u~FbHey%+1Lv6=D@^VY$0P}7^|N)4VjFOH{BBI zjY{nH8^0iNuZ=fTU|@pMC?=I6Td2rbxrWmhu#sB;OI6IYeeS)^(Jrf}#fu93c6hhT zQ-b&Vfzqm5sd5)x&uKI2B|t^VaR7|`5#Hh5?f4oq8!arQt{o_srhmvJGD z-8LJTD2-`UoKRE)VeA!Vh1(IIYoLg6N>=Fw`J9a47GYd4bhzpi#ggUmfk%9~~OW%(n0`7b%9`r~b%>h+&76KYi8RPHMxH2*^=mg znWVg8fn6xD8IxYhrI;*=WKfaG0jD{~J#PvRaxXy^`?zcI(jn)e@ZFM2bUD92G#NO# z@;JxIE$&eG0GJv)&K~($;dF20r;S~biG#;-fV^k~)8KJ=yioj))hBt5U}Ha4Nu1%d z24QPGHqPCLWK=F)1Obu`C_uoz1}w>HmOPM+R7|wXH@jZgHU_M4o`%{OQf-ADR`QT%}%$ zA}^gl^pK-Wb3YuC4ps^OkGm`orQ<+W8}?k-W{J&}#hheI+2kf}Zk}Ok{7)~&$;+)rK;X#B~62^IhYaK_0 z>KW`P=9mNQ z@oj+a?BoMG>a=2^M2{8WJ>hRj0=renetmV&@fnMv42l7(m`X)HknR>>mUglmFW7_SXa}F%}Udd5@MjRdddh>N!?Mp?TOb z2dn0dWZk=1VRh>Jedi9p>a>PB(p=;j!#~K1*}w3ALtenlovn4J+pN*pVd5*l{xHT{ zzq}Ma`N08_@rrq?T7$K4FU1s5B%g}xcit({`lT`_=pihXc77ocvWo>-erp{=9 z9*+Ya8ID~9ug5WJ@03NKIpYpEU34R?9SUf*bf+{iEMJ6oZg2uhW#}WciRm%Nw2Z&( zJv(a_E?C}V8b)q$J_{}3B(vkKj|omPEJ0`yf9pK`b#Sw!X7>Na{@qNtu$*km zZ~y!kcmCP9vXM<}-hLe^uqI=@?Hv@e4gax`DKm@1x}=SZMt?R%Bm)!k3W%Gb^=PLX z7p3{#Ua(J=KrV4?VmZc}G&AAx)O3tDhAqzJmSy}fj42E`bIs=mB5FCXYHaI;1Rym+_!2@87b3s+~e33RV zN!lNpFjNMf>CdH1| z855k{?VMU@s(^AN?|4qVv{}+F-$N>RlV&`|Ot)DG8gu_-eH?u2->1&e+mV4ENmrAd z(@2%U*R_OVfY7)Q*|z$=egOEjuFgRVoXA5yEw%S!k7RkBFvIz>@S5bZupX!(GoAb7 zcn5a3T%D6BO7g-<=#y3P)7%F_vFW@?am}X+?mQto4!ywtWBvIs>uno-H^(`&`XsRW zzS{2!9`n|PWyT>iHipEGbA2WvfopiV!Ph-BKuQ@6^eAooG}i*Z8}7(?ny2ht2mM$xCMT@L$D~Ma?w4f<>08y+P&-uxTzg99OjEw&%6+b zoC&!jl((pWu2!Jn3?_9@gBq71)T6<)?s3O^PK6Zyu`U8|+N^{Et0G|J-^I#Q9PBJ< zpQm4-mbK<>Ao=VvI(BSpP8h7ShbaaUh5M+;ESF-aQR;TRvIOb_)XK83OHcvO9Z&}% z;_dQ^*}7Es$fb7oXwaL{-GJSv8XBGf!aBGUjSQ-Esgt zW=V^<$$Hw=R)X4?CD=Ns`!Konq)ZQ<^jZJ;UnF@NDK&WOITW*#BI#5lGG}2aJn%^F z;bw3;Wo3&ioGwZrfCWvtnpvx)Bb?N|(o{(|FDDRFI~@>AzX3G-Sb1P;HgyYLuyx*~ ziBVWNH1_*d)Z8u)G-g?`+uV)`MwJrx)3ins$K-P>oYb(~#JD~VYKk}-+%79~P625u zJZ_iua8kI(J=1B#H`dRb5#3%;L%M@BPHM#!|GbcEJiI=7o%o z1Hx9y9{Q^Prcf=XJLm)yP3F)$6lv}WuCW#}Bv=6PF?M6zk0!ut-f314r*3)HZ++E~ z6cvy!Moz(Wx{au|a?1P$$X&V(>fH3F+CiJ~anSvsjNl#`9oo1sAo)j%LM^cWBdCyB zS*?UqLoUhkF3^!}<8PkcBTNK^;xmd=H*H{1$T5Eu&1>Ti(b-F?m3T(gCC%g9kydls z*Gn-%Jb7qhg(WL*q+bNd z@BWjVXH45|k{ufxtFW**TF@C(FI^+(mLt1s2faCPl~4AP9g2Jb_IwT~HS}f9s(=Gd zPy9X&L#o&Uzr#olTOg>GeX1x~e2;$*2m`AEHq39J%N9RzNDOUKw16L>wQeY$oRbEY;t0in=;D6h735$QJ<>R7 z=>d}CUi~h>R5Le`TMp4@P$J4vR&(I`h<1w2A{otZS+m| zdhkd_hEp7#kB(0FLj3!QL%mcRh&t9UQ1WF1A#1|gn($@^>+8o-e)+2VD~+5CMUOCh zNx!5ow2hzYyM9@P)A_fDxR2h$dPGdb{%I z&Ta|0X2*8rk%3*gLowYH>4IW?ZYl@;jf~)qkhM;#R_3n!y#EIO^OEiIw%~O{b>4qS zu2!HBnwCa!B9uW}GqYp%3PBx2-q7uR;(*i&CBA1Md)+5%B~7l`yeAG-k}l2$*n20- zn>ZI1Z(SS(EF2dWUxGd$)fr(mr^c^~qdSS0@#dt(+Q83dRyZ|>mV(Yr6@6hb-1?dG z1qd={`89Em`5y)0fva<}TvVMwU7S1Kcf1P(S3{pTTv>8qak}SeNPk9?FG6(He$P8g zEC4rZX>9THIqb__UZ1{uSnu$5wxDY6NQBY?Yk_tKld zEtu+dLtY({E#Ei0gaZno;-lQhy#EENZZ^<(ZvJd|f)zBVwLPd^ZERy6UVVUX(x%&wog(`a+W+n~7;6wW68X358YZ zbggfy8)mp`ee(r(9h2p?5DYvNP_HQPE0n0RxWjK;O!dgYIQ z%g^BI5#!+N5KUt2c!>dKo-s>IGR177$VOm3rjsDpddWrWt-9e3Zi`;MwMrH|Vx}iLX%l*m)%r4Z!LJ_088@riZ6?V?W9nYl9w6~5l=ROo1tx@i(RO+7oZR4T zq;k;?&5*n4+~tF zHJ{eHi5)K1Jox$Vrt8<1u3xtNmBh01M(nq6kZ256qHPodwsA8Rxr^3vG!U;-D;~XB zF$0tx0{eo7oFSqtY3FRR_d{FD)Mie zNr@f%9d!n?T|+U@2vAN%ZgSnhzp(g#Y$u~j1!6!|r#m;~vc|Q>@2c=B&@-tOc;9*d z3TG9PkKxH4#l^s3Sv4m;p3#nA@=2LuCbvah4vFr=CY3t23+&%+#d?>hFs6Ad>z4okp_JjzI+Nhhv=Abc)$Q zk!?_ozU&%ANSY#&7TpEDgMMy7NQ*~v$ftB6RG4f04tX93%l8C*U`bh6owSdrT{e46 zj`^4XLRii8L~mw=5b7^k|M7o~o84cNFo}uQ_$JvGA)O(8+zsJ9@`K|2;T3cC6Cv1U3C75vAT(aI&A*OcC%eaMmltt+pd2@TMB7Ch z*r=FHjFbZ0}h5C zkU;-ZKx|kl2kZ8AoJ*-d^Kvl!_5zJlpX|Ixg>ob4_Eht7=+P&z;;kd(*1RI1cEK99 zEHVILl@iW1&VEwG*9Ml#(j>Xc63}3L$SHu$^F4V2Z=3Rj;xvO=h4`~T93O~fEf=J< z2n&~J`I*ig{Hm~9^V;|iLpqf?W|mg4!pu|fSoFax4$zwLZ|bG_qGb7bZo2~61s^S|PynlfLZ$WF$&8*ra9}~YY8MpW zpnp3VwrE0lG7Axw-L!rApH^7X+-kqY>Cl}p)&owYn0Sh;ry{Wv=^*zDA@(7ujyrZp zN7Bcd-2e23u?}SYd(FUkHlD`pAJ_pELELb zTDMe7rv=gt}_+_ z0v>0#?gT|OhKlqE%H*fCl6ZEZI6GcjfzIaS|=Miv=hdKjje=fT#Uo&GiCly_s+0{ zD9nys6ca>Y*g2-5%bj|rtB!ld^BNTmLapC^kt$z=uFH_ic@I@w*da$~;ufs8O;yr3 zCSU?!8fO&ipQ7&k`szPcBp*38a-mE+jzPMFy#c#t+R%o)Xu`~a#EYJ{)wL?BA;wd` zsd>ztjwR~XW>W)u<}sU^-4p{dxw|k_k}Ti63?gM%u6WZ8v4M?)$pN?Is#MO1lHo{u zGq(3YydTT;QHBEzO|#@2spY5;O4tCcBYVIbY~Z|_1DKcZn5_?+mL!JQioNs@Ouo)Hv2YZyE~Y>}n|WjbHCHRz0;_ctGp%81s=&F;M21O-06dq&jc( z-sFPJ7j1NWV3Kz)9nZPr`-R{VP-n)2Fv|&#l$jm;$Gi$BO#k91wc>Nx-I>Keq|z=& zxfU$Mw;^=cg2m8kSQ{eG#$yMT@Bi)RX?(pyyX?}IPsloU4y_%pOuG!Al}a&ND3U}) zHhT2X1%9{Yg7n#l6Wb@ldV^%%-C3|PR>IkfA*$W72G=@%QHTxU^K8NO(y z8v-v2p*Rw{elYfWob2J2h1H6(gRhAiBQSD*|ILKA+DJ67UUVU1*r9XQgPGC1YrwRv z_`x=-x2n~8`C4{pJO zyth7JU6(;zgeCy=-`+8r-ju^ktkBF zye`IQ9FjM6fV-hqAR%0{WWCERc@3{wQsk6?#7~O9Aj?5%sLmlq-^5bG#2R4Fp!VWV z76P!@Y5I?heUayFj2*8nCOnR7o*fIWxD<6~+3Dd@9 zhfnLr#=9N6jhmR*zibh15q`9K;l1hU?(N~fzXW9i`GQ)~6fr1lmlwO&)3MT~2rZrE za+iC?1x2EINw0gSvVTthoLh^UBF>4=iBE_xeXTvDd?E5!Tn;E%&=i4n22l$Sghfj* z@e?nqP&7rHCL5u!Ezc*<2k32G2H`iHCv{T12LitE8z+u8Gja1``(XO&gwbOfH{-?N zrdLPz8i(~?6Tj9(z#kW-E8Je^HvWC*6`+)=IwjN6N#5-WWXIHnTAnyGdQ^BcOZNL+ znft^6>kmKxIDD^Dvjpi!pEzjdUJu*hhY7-V`9u2R%ogjh?2nyx>mY3Gh}QX>HT)Y_ zsouFc_4RLDBb(ScH+H;3-fQ68?4%f|%ilpoW`M@a)#>?yWO>VI7FPKK$>_*eKTnKo7!zMVUi9)zTw^oWCWf%dX;w?l)PQ<%_Nz{BaFry> zH@J35VWaGn%U!)t?Og|oKZOu3&2ZGZT=%bZMa|rf5S13PLAX$yQmU1BQyg=5R!7J| z?x5tcN8-#r?(GF?#d+>gSYR~tLvD-kFsLo4(3Bw44(x)!z+kNtQtIc>s~4_yvUtO4#QhB5X?5a=={_GeNvKl9CLNvdF5v_hr=-g1aT5kD^cO;{@Jn6 zGpH-gj~4WjawU*k18aqQ^HS{_T-qOMEl#N`f4KF14E2=pIoDNCtyw#9}#G%{hHG(FEs@tPSniy6_I;H8; zL1__){T~4Dk6KZ(te>m^Eg|G(KrvwSp4M@)p=`5W`tipf|J)orjcZG6e3G5RQ+6cM zah4u^-z@mqugDH|OPL*Sx(^#HWrY+23Ng8`lm+&=l`K*#w8A3tFsOqbcF+nhaqv@t zxXGge)-MPV3NJ9d;ws4Ff=x-O}Jm&Cp%y}mLE&OWh(>L z)bs8d7d>2l+Cn}giPOdl|Jkof5a>~l@ozFI1~{Fzqd+y*qtx-^1MkcFJEkXg{C-Pm!knV$}I>>^6ubQI3YZ+xOXzvu?Zo;R?|ObT+?Re-t7g6#vz*5B=&1U ztE|*@j|{>%DCB(C@w_NT1d?>PJI8E+s!ZH1FP7yhCj%_{`C*5dF_UWJkH73(_VIGP z^(gf`CnKHXSr0pQabpbJp9d7vPmy~_d=SSBl15O;&4<2@82=pbZcyb=l@W|g2p3&5 zf@5KCv`*H}0n+0H9^SFW^;jr4C`iha>We}`7hQD^j-0M@?1McvXbmmy=G}EXP1e1k ziWXGQL4~bU-+LpnL!+bOFT-c*93jMm0VHfX0BK)%ve#ENb|54rQ*B|o7pm?$ri-x9 z?hBwz#T!SOA(Q1XywkJs_kt`4W+$*ZM^y^Q#@@(0Ik006I-?@>E9mtf_ z-|;?8qV3qphQQ32HEI*ZY^2Br|7s)EiRHr|Vy*u>WOX*}aW{Y#Z6&}Hjvt=eqjWtsEhO+0GT)9hZn-MWW0j*4QQ z-b~Eh_Ae*M>S@r8G4|V-LNS{ul0Zc^OGO9RaFXsI(8(TJBbUv z*#)j6drzAp!D#T;1QjWN-fS`n5Y_$WiPMf7|8)EI$GUVYq+&{UaVpL_9<0T+oFQ!{Ldlp|8 z7eJmmK0G$C!b{_X$Vl=k2rjOX=2Ub<6;M-@(`^bnMQpZ92rXqQI?MyX7f0?=QbICXZtilCxtIV}e#r?R`j54a-n1FrfbBro0#!P(O9u?^!2(?9&uc z$;;(-@IT|X@v+!ppSU70#Th!O18*&=6ZR~~c)QE0>&`ecJ#kcxGnf@m`6ev(d>tDD zMy0)zgIk>Tdk%U&_Eimf);VcBzp(X8KR-F<&i(UG+V+fTi}KCCyXO1oo!kG~b95ug zvg6Zsi(@(x%@&0xB5x+M6-nVs%%#9JTupa&Srd-S2=Xn9TNMfg}mFUOhac($e zOy5YYm{Ch7&hL{a^4d9gCC6RuT{8FX%=6H<@_^LRqjyCK289Lw)&ZW?;PwoJRsr7D z_bRR!=bG)-$2E~xUI`KUI$;@gecmO%v7Z|wIwAgo1#*`-9kPnRJnrC(su>ktnaZO) zjgxie^~IAgp;>?NkXE_M&d8Mia`7eOxZG>v*qVsqwRq%*#W1?RF@~2c$7su~MfXFi zNG>dEiJ_X%O$6zq(&-r6(PSIVHET2(J#VYSoI8J5U|gwdqQtdDelFmue=mt~PG(2SqR)yz^=IN@&r;rNl?(Z7-i9>)pZ56xO_T;=ZvnW4F*S;hsbZdSpA_x;MXF zUIuBNl7Lf^EpL@VpU9~NJu;DsZx!^a<2{r5p* z_Lt{rJM1`CYl7ni+vf8{b$+8ib*X9WjqjORAU>>Y08)h&zKtQ-OHdr-7KniqF1;yV zLEV?N@$by)_rN4J;;O_Cy|Yf)z3IhJIu|S7D9>Kf0L5!f5qr2eN(Qfuk7GU#8sc7A zvQ3%x{*dK~1J=+oxpo*{cp+}gxM&($LF~4JmkS9GjB{j?EE@0eaQ_ zROAuAZ17Kz?5Bzzlx%tPvam-wBu|zfSiGNX4jhuJR+61iSf>qa;}6ly+oxErp^+W)Fa&4nLegoFC&}~KJiSNQ&#&X|bWP-0tBK79 zxH1DZFTZ7(rs1BS9zAo@|DH&4$E)GBW9Rp%0Y(l`3}le@Qjzui4tk$w&72nDO^CVe z7Vh)Ja&BZuMS5;ky|fJ)nXs@+SB42(9-JqBQQ@FNlTMcbJypKwurJhr-w3U&|Z+-cFrp->r4}BB4df+(tIz*G0X(Y>FE|Mu`6Gb*s zk@kIbs{}%ln=D%vQhqTI#reF0q;duKk zs$VwaPyZgK$Hf+nEQ1`dJ~BTJh(P|D<~qSeQ@pJ6)Ei zhWGon%aMw1Ep?IltA#UBoQECjOic>H%}NL7eHYAODT5 zoiQKbhCT}GBbK;lh43lAe1ERmSewRfw@)V2W_I~(S7dRzE@pFP~& zZYvyn=nmP#pcQV=lO{bT>Y?Kthvi5i(-hIj!PdmW&{aGQeKV*Pni^NQHPD6JtNvNG z>a?~TFf5T^=IyKiL;c@xFUd8oim_Y3n5c85uBi#Z6SNbD)MJ?xw104aJa&v{6oefBqp>zqNW#8_y_2)fi|HG z_)*ZVpH(z+l0;>Tu?iY_QlcOvALAbtJ{Z&*w2O}Mf8w)MG8lwm_7)+CEK4ga4`rBs z4HkTl;ia1%*U~GjzQ^{vC;jf{IorJTE^y(K9~>YVuh^biYv2~|rI;d$s=FgprqHRNZ{f54*CHp z4gG{$xd^t`eezy$vYf4d`%;)$hVN%~<)y!f6o6h7GRevp`Kf87T~ z-?|1^6{-VwyXFhjlOGO%DKrxpE#g2 zz=6f*!~?!@ZvZWjq+_-UBKiyp#sJAK_Pd`AShol|@cogY4*%SA?_Zb^x;e~tfg&^K z|CwDd{$%-7k2`{Cb1_atXv#ZnziV{UuR6{1FMULk>^SgJVz9zwQw&hRtEtF6vI~oc zfJ`-m7w=yxixQ;r)r!L4M)3uYPu))`&%s)8_4_NywU83-N86}7@BDscK<#^DkFCCv zjGY%tU@~?bYky`16YARA_3v4-(%5l6(!@#=O|t14|2^Dz&N(qCa0cJxoSt`mUWSeL z%Hz(mc@P@+QJbEemGWk&SZ_OIe}AWuY_wzB0nX@{mVk7M*+G$Q$fxm;?g^@vl!gMo ze}I0U1*CCJ5ixJn%y}fPaB3w77iYPYhHB^%(OHFcyHQO~k9l)$u%wtC%!*gonT<^m zfB8+g9w9y}*B6jO)5t}Gi8w(q$0$-mMOMu2=Qb*`K=TwybB0_xX$+!k=%cbuy4*j3 z*D4=`uz6JYLBC$Qj*GN^D5s4Hw`BS5WgEFilpQpX)I$EY2WrnALTzS>OC4O-(6PWs znC}E?Fu+d=;&qF)f!iWdpzh?e`6nHd}rGiES-FFLSf0$I(N z{cexvRyM{@(?cnu<0BPWF^vF${+RD{BE`g0WIYv$-eR;M4+<1PUF^7Hjt zuKkia{56^-Fbb*rP^->L84WX0mkWa8J z3|LLPp&_x-6pud)D+@9uJ*v;eSXP9M1+nApZ=Hb;RYNgA*HTVJA_H!2z*WdP;bK!N zshABs2{fL)9nwp;@y~#}oEwlsmjbD&S~&!S*!$d5ffW}%a}P?QASSDktcP7xjuQE{ zb$7S%Px@RIX1NSI><#OdWVrx`E);_wlstlR-Ll0PvCFdt-;@QAGX$W)(fBZ8bN2dg zjyXH%J=nr8fAwb)^@=5UfF*g1yPZHWaTHlcMgGsQ!+HK?;i{SEl$U`-TSLdZamAK- zNR#i}tcf@IwCygKwl%c&)*@r7!PmsDm|#pCBymjXoHA!5tX&~EqSzd`!@JJ8nzu5b zed%ZJJ-)VGjh>rhvmi5K2cCP>rYF{wFqe$=AlLxIe)sXjgn9%;WAy?Hs7i2!h z8?u%49uH)yY$fFJQg^xsg0UKed*#M&^t6A zl}`U}vcZny|M>>a%MOa!hE6%Kw)cR>!({R@g_8O-`O1JToE3tY&}hew+4tnBioRfI13&Kh*^(9)_>jPfj=)m- z3)9z9_~)F{il834Rd`FD%}sN++IN)o?KW#?crJ6($G4`25=%k{MDt6k6I1`L@ON8&w_{z zH_cD2sF`!q9aUm?yIon*L+@0UEFI#i2IeDCE(U#28mlmL2$}`1EU_@3H6AXlh5hly zw&YyjKQ}xy-dfvj&6A03^w`Eiy>LvW?IZl-ymhlCd7S5LeTtQVo7z_+cQjso>=qIx zmc>Wkx-N$3E&s}rN8;_?WlM9pdqlh4Vn_mS>oWWd(%{gR8qe7xDsevzZ9kWVwRDj< z6C^cnxa||S@wfl$b#as+ZFU}K5CVB6!5cjKLpORi(498WAuE7VZK zo4$M~(yu<5{FGJ_KVFQ^j>EZy25XF(VzyIcD;1d>e1U1>BWuF(f8R~#dPfPg!bG9GBKmXAmqrVD0Wpuv>9;W)1zil;l>4g(I0nof~JZoJv4Vqq$$3OnZa!(F;GS{PvP$VbV(=dQn0l$Ildgli(E))wa4U_it(A*Z$n%b}Rr|s| zbyq9yFG%ECL2%q~!Y2TyaZJ$2DZ_y+Jo;gKQ0_lg0(u~{8InMIojM_qk}ud21_G-M z!a>h2cO+0az^n>ymK=AS92=vXJ66!JaJtz%{(riQs*PL3*o4~bcOwU6xMR~r`4pp} zKu0OEgU*BaZ1o(qqTC;={GpIfmf%?Lc2tbzdl+oNV{Gv#C2`&>1F&vkzfS{QOCO%U z%cTL%M+=aG!-nW#1(xS52%G6=cW?ZIFTXSQ?o}a<%B++h4uPN!>Ls?z zE0^u0ONFY7fgKX8P=WR?&`it_sq#grs@SMN!jDc`%a3F7MW;!omnsV+B33Q~Zj}t) zmrJ9<)e3xUL`%96D3g2XW0IZLBA>Gi49}qhPOzJFQ+J#8QN{_zP5Jhv+Fu#V+`T4t z#)Mj*hE8#5V)Q#{NbI;|`PtB9n3^T?`ov5>ZM!Sq-9kqg_nX?SF)`6w3O&!DwtAk% zuy1LozQ!$8G9GT;z8(+Khz*5x>cMRrtFWk#{BrS47<`ul*ru(siQ5>llGDz?&&i-yl@1Jr!?Nq*`lX*j>?VKSCQ?PBNvUx4!p22Wg3I#9 z9@QQ(OgyJ4BKCvp;!JL*@<9u!P7kUy0$9nisalzVkv66n3djP&@_0^< zaQ~7FPF2XUP?S&0fz+l(naUZUHJpx+rie4pY+o<@T;2qr#dNr4?x&JIak0!Yb7(^( zS!52kZ~5k&1$rcXR1r~43hg+1eaZkE8t7luGCYq?Sf}x zs5(Q^K=~L?ScU}sblCzB79nBu-^Ioso+^#VO?pQUpQ1UOd!)*a;d9LZK6MmxmLjL9 z$QT}0yr>oDxK+Nmr2_W3+u!J23gr$Mu&b9A0=#xQUWD9%hQ3EvDUZ3TP&zALfE&1@ zq(#`{0`O`Ap54-r>Ji$^dc~mRC`k?&l&Fpp4e+df>b-7}s!5SAc;cXARn?62^_7Nn z&?^Ml&dor;ii-C=Lg1G7O^2Tx&R<$%QJ- zWI2ZVY*4lE!WYkCre8RYwJ)+W9YKo7B4ZXqHp`hEhmlQi9$w|_^FglFXu*wlRHxh$ zc=>|fuq5v%4ykTA^v1W3z_K++R4u|bel%RzKf)dS#37xom|Yr*WNEfFBP>v5yv|q~ zNCxB8{`^-1v-L}iGA_1~q}j380)l#D`b~-`25LApR3v)1m5bte5HLoAkrA9MAMj1& zbEQnM`|csCwh_h&e2Cm| zPjgR)A$z3xqBd`EclUt9iwxz%^DcS7VMWMYT2%@H)<+3sI&jPhBF!2|dI~8MA0~n) zc1ZG_=P7vAY009dkE9cT&rrD0JINc`RGv6=gs2t3O(@J4j5Jr*1!_Rgs$GuVCC5EW z0^p9Qg<4@1JKpF*b*k|v2bPm< z?6yL7?9UuDSk>|<2736js7MXHZB8c!feCfy8{~y?t5sDju>I|nfC|Qnz6)Da!(3fNz-^N~%<@-eOM@Eko=Z6DV$d+lK z_cGRX&`=DtBW6;O_vmP{n@;yj@oAUgOKXZ~CwPM;uXuvL(yYf=eR=IS%k%5LYj-VSv9n{V zVuHo4KCohTv!p|&TC=c^+e2eK1&DTMSgkn1EcEDZG85Y@MzqeYGavEi8H3_^LA6(J z>i5g;LaiM{&q7p*Zu>lS0wYMLy4*I)!wY7^%&E6hVayBuAx4)>o3)jzy8a~9uPLX# z{*7y76FdLIer+JgT#a2CcTxp@cX-Ukh2 z{c8h~JOhwyr zJt{`$lO;MvAG7&SUiP}-g?e=S^wUK}q?jE#?6{Hfya76@DF({3j!=;a%x+IBMY$#crSq+QJmQQR#v~+I!BbsD&%%B|y?}6|+0!tm3BoZtv~#1af#j zv<}2E$?{rg$=*iV_z9$aVT?xtQ^X)yh$>#xOIIzylRyA3uO)??{?LlqJKW;tO)#!9 z1_3NFWMpVwa5(07Oypx#T5G+R9z&}>nz@0bv%`=bdo2|P7}`TI(7BWc%aBv6dx}>V zuz7#DJy>-j__O)R@;b+Oh}drP?+3~StV_ukU};0Xs6f1b2}m4qYH2lptsvP+Bf(CU z3LfU9p~U=*_chPk3umX@d z8)Old?tgdwcRtYDo>I3*KPI))NQc3f_7cT_6koz6ds>!DJa1TfS-yIjhBL4rX4+k*omo9CUZjzzgDafM z{fb~=*y>gYdS&~4J}d(Y_R>&%_xR?sT;d!XfWD{HuTF|h?oAONw(&Edg&+<}2MR&f zSFMN=oFtDtaO4l;MaVEW{KhFLh1%nCmyHUZ4X0$Bc`|>NOz&65DfzeGZ?ed(!q=Ey z$ZnHLCJ2L?CB5$Jfh$JTF}udEUV0j6&C)@%>j5YgsdRKvrJ?)gJdi=E2!pi9Y^9-d zL-ImOIQdRCS&_!T$O8P18O8b^S>boe8yUCH8h1_FElf-h@$Oi>j`T}%Xbtze80kBZ zO)NezI{2on!KGbZF1X>xsSvGhX~ehqmZBRxjRu_Y+Hh?0X~r2@zv!%6CO1k@ zF=G>1<6F(!DnWMME8-2|iM*smd5W!)9N(shEu#If7(tXj=q_DII)iHbP&K$t`q&42 zR=PXrL|EgBek!CP?E>$*w}#(MXsh zTX|33sAv#AaVT~wc7ldhtVuiW*da-$bAz8aApe1NUh?RFl*ym$XTju9R=vP#7VJ0Z zb$ScG&^ROTN~MWR#Qi_v!merNC*1Y&gVGVIR4lK9OgQuclzQHj0pBW^65udJ;)a+v zw)v~hk$s9mUerumJF(_<&W>wEtcu^~2Q!W6CGY*@&yM<4sP@gOzmZjTybA3!ScQ@( zCXpiX$j^`r5jNPyAn|TbPzUgL+y(Mow05?tEP4JFkDf8xsEs}_Nk2_PV4R)!RBG4H zzxOpgET;DT{Bv^bwF%?iGyu^BifN$8c`6chd$pj!wZj$k=)=nvUsq2^aEa3ZJEjSAHr*hFV~0W%21bRpaiN?jZ*cq$F)q_N%T$`X(r5FX>; z=MhH410MKGN;uazeRH5(s>Jt<0;RL@an%l{BBWl5g`e>v{MD->y@^3dDUf6cAJ^zLaPKK`WS_E*h5Q z2CJ1oCWEbH>2#h?qRRtWtkm`n45Gq(K#j2zHiy%cDc1)7WqF`}O^QwK-Ajt>*i|`g zu*y_W%wdWg#F~}!+;+tArX5!9(FoNx!MJKYEFZDMGj99`HBV2;uN0Du(+ zlSA14jxpO*$!Vf?xdhRLS$tnrdxJ_fsK01kWN^oPh_{3i0Z|L@1Y`tT3p)fBUm zB2klGOK1Ea_Ol;3r@yDi$AkC3`V&&a&gZq?^cA^b0P_ZlfqIxTRAiN8Shk0Ilv@Ra zW?H(1qT%lWyL*`Efpj6X8K)|J@kRD81~QF=SqTEc{^eLt~?3uU`Pu)(~x>jj!GycNStap)>HJqx<=Ct_fjn3Tk zzbBGhJ9cV--e*h}-~h!yr}JJaa`W^SCtbKjg+j&{cZ1%t)1)W_Im_DUbIYDMU=>SN za3@^?Qb{_I03AoKN|)`zBB}iZLwM!>tAkZY(VI@EgBm2t2_#87X)Bh1vj)zKt%B8% z!wxu0fAIeAh#oj44jE_38aoC~jsbACP)rg<5~;|P(0-`2M38j5t_Kk;D1)n*jj4!% z1&?8~gZYWcvL++5$Oai!n|3n>VuuW`qWO!Ar6pdIf;zPUj-Ki{ehB}|;-U@-6j)XI zr$&^|y-1Gv_f5K{quBs5W74g8%bN#7))-gizEY_m6ScS4ahfh#Bj}c^VtA>NlY&pU z&639+Io_%S9_lvci>f(&+=q0!C@yS!IBbj6%7M@!XVqQDdTEX_&*#dLDte=LI=x%A z*;g|+$-9i+DDRP5-{zQ&i%D8)MzLkuuzX83YruGUc^)x#9Cb6XYXx4yM_Zh9wXKAs!K>L*Nt zr10HM-!f+HdER{3F-S}>`E2n&FM8sm`x|S-A)f0|$Ja@YCoNjNu!XJ)>jJP`;`GvZ zrIhaF6^5pZ8W$Z6UMWae{CHkd#KFZomDwSgq7<7E!%Yk?W7Zz4BSkl$X}qYrG#?tP z7};%@#DroI3Ip5%4bFT~`P}Te<#YG?!G0!LUN3Fq>xd*V;NJ?Gzz7$Ns914N`+^CZ z2pC>4oYgLw2E#Z72~%RGht>UKbqTeKkNLOp)7;Bpi=5%!?{ORgwO9QcLmGjlsR}ka zWsBoQXkse73c~tASIR4#x&xY-Jjfnx;8lBWoH;D}oPU;dNOF}GUQH3}$uS<@xnW`X z!a8mv1Yf&dTIC1*3PRfW6@iDm4tbv9eJ;yT^a!(;0D*Yufw{RB5{FC#cVkS@i#}R} z!LSFPCK`LTY_`X*p3jTnxFl~?hGQ4WRzhU1*u9NEz=1O-6dAk`+HCrZkjXJdBQwVm z85Yepn}^?i<8MjE?Hg>6VZXKJ#6+`4tn^}FqXIRJ4nQBO+C^0*>6m@P9qXVjlD$r) zq1d802@dYlSz-l^r%&1b!Zfwi#X#_FkYq}&xWz?ZrWESq(gRc)(H^cYd9FG)GlhrT zl{$K`eNe}yq04&>6!azMpV52K)t9Hm%af)Gb_K6!nG#*+l;MzWMMc^h<7+NQ-rSbR# ztqoKwvc$!72mi5+IR5nCP)pviEx+Wfd{O(|wchEO)$Tczb;0|03NLR-m8aQzH4DF&ioy=g5xTW>6*>lh!Vvm|Yaf(WSMq2oK5|1S4ec9Wp!w zIMxLw%h9s7Kwoq#NXMqrxwNWTk{obLu0nsheU4gzu9ZqhjEScz7um$YwhFdh29yb0 zAG+B(#eXzA-BTP*G}Fj2gV%hJV)jv_7}bJyG9SiF!yl`sHAd9D_wmOcV_+?vE)4As zfZ&>UodkkrP+t8YD4B<6+C3iwE$s#97~JUnrAJf5c5e`y3Mr%Ng#D6jGZIDRvu=nJ zdF`Mx)JtD@|GGGbPOxH5$11P<+4p(^FL(S$Sbd%CcMbcQ1HJ84FPLhb-8H8vbnT+i zp9tnWpFY`;{2s5DRC_i>+~aA5J0-RBj&GLz(`4Q( z69Y0||JvZI-e9CF^1lm;(zyKw#{zf0gU#+?Qo*R`FJf@s`Tdv^i9oA)+wtPa?+40sH_9tUDEU6T;g(68*B-Ykm1scjCw@)Av z7SAbjN#Uq&KzrHd=@+DEIY#<|uyxfYdy(g608YFf;E4VH-dtnCZ8koK9q)Zju&^KT zD~7HR^e~FKDs(fnkcEsA^av-*syqXenGkyRz$ATI^ZTF=?Q3`XQ_L}#t1PJ4$n5&Pe0if7p9k+zC#{MjSJ>p6YF83lnzP!hvchgZ1azm zs`|u_#n`;6p%44sT?k>;TF2(Va-S&Jzr+P&3J8mMkGK{w8u~KF#^9O2f~1F!ZNW2T zN}*&~m2ttG-MXhHisJ@Jk9>vG3A&Bm7~U>F5HKJ=uiOk$?FE7s`8sGc+%IgAHs z`qfAF_jek}Ms^8lJ6@NH4A!M|irGPtZB(R|ujRM#H*z(cD8WEzDsUm=+SC-006Oxq zat){0eW&D@q)w{k--R}^${BHN_;W4zdQSjP7W_8zA3ghybGNG=A>XY|yF@m#Lx>%h zEEF3cB#UCei%p{&JYc~ z7cNXfL}daf88J8J&-cXcRBS%@FaM@TiMm_xm|S>mP||0Bk`9V#qsVnCa+ULKC^1?7 zdj9zNL~-`iOSaL6PVtgo8TUEAQVjtndY_{NL(V6F%;l^i!x5SfcT#6R_z;U1P+&8T zSu?Ab&R*8PwBwtx?`{50Q^YZUtcJpZrZYla-5nCuUH55+oz-5>x`>j1bWz`&Tk^&T z9O#Q>olC97c&fJ!)`Z8z49@sZnRb3=Jga!E9x_BMHx`u>pYDK1d zH(bC@^$wt<=#hdJH+Hfg=T-1J_>~0s-dahw3)1Y=$yWMa6JHZUYz@MR@aw14DAxN@j{kdKcGmRtjN_ZyujA0^ z6IHT6kEM?)BC1KD9S3<&8GyW;VxUar02R3(su$7G!TLo##FOP{eefjOm!mQ~T1P<{ z#Rf{P=%@Q>EoY3QrFB#=W7c=ihbgp0V}~ zRTK!5Te>Z`rr9z0F1!r5FS*VR*Z=V^{;F4?>d62mYW zbJM+lA;xAv%{O0w6}4DxgWN9%MPLB)IGY3Ujv;7`#Hw1Ls7sa~fvUN3pSzM&X~pcP z99ekpZnrB-EC&zFp>6Q<5^$JtnH7Gf{!^OkdmO#_`LB}$C&)@WHb0pL=4Ug-Bv2%d zirnd)CB{UPfaN=N?%CZ=^05%^a41F7FTh z1bQ1bFi6&^!lLc31Y1x8n+9 zi}aeLCQR5bY%4)paHy-njQ9vsm5%)(UohDg%uM@5O*S*)Hr|6zDZNE{?DF^=8FpZ2 z((IPv5oCkM<~-vmW<5pLQjv#1M{jwuyg#(xxtfP5%}FbeFj$Su0ffN=lXUR$&afZL z^ngg8^`HMmlI_@W0ZzIxY3&?}0o8zXD)P+R*TuW&)e*ae>j)#IlYiqK1%}x?47cQ#B#yJmrs*9GjFi*89jLU&=%dn>U&Qu zije&3RWH6g;Bktly5@5S__%R_h=%yy$R-DcH}SwK`x*1+gbgsvAJy_ZSpj28;gz%Z zjDhid^w)j^M<@L(uPi@pAs>>&X=IPVyUwH-VEx+;oZP{e{4zzlzjf3* zH@t2+U-jPWG&3oAN zYab_SYrh2n9<==H#u)&+EsZgeny4h##clj5I!2)OZiH^@5+*megFX~+LQzFyV{0Gj z@qOaZ=ClDItc&VCaoFJ0t~fzAxT3gB2mLrmg}H|&MXF?QL`4s~?%|PxhO_jS3tpo|8LqM$8F;d<@=S_R%~@KYtf1kWXEn_?>_K zsovhiq(ys?>vrtPJ~l9fUr@~F6uFJ!1j+JL-+Oc#(722eOTl(F)i)hZVaaY0(7YV> z#UP5R7vd>}OIHZ6`(ITrJxcJNHaUWcraUuv6nK%ZehOD@(#LJnhK(UkQl=TgwY}kkY8&K%vP3$*7&I< zn_V-ekHnYp$y?m0@ML*xL>oVy?tq@`<@)~!_$Ri?2S@&88Pk_9f4cRuHR4`|$mg!> z!N&@5Q&xY^eJZ0zob$thD`X41@SgqJWwPG@pBjn*FDa9X%;V(JpF@f}k-6b@ajLE< zx0gQsZtTyhK`d0m>5^h|F6^yB(){8%7aR+DjS)NE-}<8~Z(n)$Ld2FgqHW7TVM#LK zdGlk%Grz8M_2TN0;^6BLO=89?D7RxL8S2N!?7WjHW)np=QjyJIB!C*`p7TBDHa-fX zWx(X`2U-%`Qcd!9icKoXic@RGP^PWQ`5(SFXe<=WCP-_?h%q56i~(6}WWzo;m5wPW z9r#v<=6^ZA9yA_>NwK>wEp9m!ZMI^Wfu9%Ma%ksUUm6z++pPy}qWrhU zuR~r-uuw;Z47{D-cYw2kq%C-u3{#0e(m{P=kIQjGq2d%^O?cgbCqH)Q{*BQdB6i@MpUS-mZQvZ$8xP? zZr&1AVrYvv$|+l^mH^KXw?dJ(2 zIQEb{mA?<%UEE?pg;>7m{sP2a1vI#PfjmOHAaxnXWCWiWQ4dR}D`t-rO60iYpu$De z!W=rrzfZ2y%X!R;=RI-28j53~@%~qqRLnNV2-(ccj78}4pI?eg(~x^RYY8z^R45BHf@Xu^Qs4VVOz^-tPxYRJ|aZq?y$&KOsH_yZk2($fZ{Z%h^Vc)xq5$byMko zNQ`xflVG?ej-Z!U#k>wljeiDnB<#GXO18mmHLnujH9B;b_=peG zv?>}$DAl%d%KTKZ^6kJ>z19h89?}%04dgB91U#6}vYC_Hp-l z_Ro2`D}Qc&Cyu4(P0U2jJogXMpD=BW4*yGAvLy|fcDyAw(U953ZxOCth^*1;rdNBO zpzG;V%5&m$-vQD~S9?}^;h9G19xjOc!JbSPJJ0jE<2wn_^NGR7)TCSQ{732aDVFs7 z+wrc;M9)8#^{EwAJk4C#Iia|zs*BE~bGQj_OacrHC(c|@7(QkDqoKzM-&5(=r_*6x zJko2&;o}5@>bvNv%oB<{!tgO>4w{t13$vhp^Ppr8w>~fqT+*6*xvQFmOAk#jJZR`;{utZn?-1+;X7HetXw@HH?gXOLPSlTG&Iz_Hhktn89 z4;BZLqe-O4qZn##qXb%EFa7W<*TtaB^zBQ~6ON3aw-&YW@zd>j4+!$rbcs9p5DerF zkYe8i{~j83H1PT@;Y#0+Qoi2#AE%dAe{kr7(@V3c74L0*|I|-+eWmp)m;Sxzt6zS* z^(((`iol-owXQdVcF}FZ{o=8|yOu8bTI1XG-|1gEyzG;IZHlO+(Ra_G(Nf`WP%Eyv zUtnGeP0K-M8h$Uj@-k?$0vVO={)UUE9%PT1)3HR&F2`lZ+xv0@NbaT>P>kP&i>Q_^ z<}{BMvFD479^OP4i0gQe9fARt(Dy zIYR7Gk@Jj){$*3c=DKy@#pXAfckl0%Z{}IzT-z~VOgPu5hO8I4?Er!l;5O%WKtjTt zaG!av)cm{51cmtnTItRt&%f?AZV|WJEQN`7@KTTa(z?Z&qT}?lPtSDX4unPX?#t>G zx()U|apNLbI3O-KD2(=g3|a6bD)XoD%PvIho7XPeMvBA-xmBbwBKcdJ-n;rv&4&l9 z-3&g_kO)wHcL?aAV;~{`&dBIyciF)q6zkUt>ob1ESs_9m}zFl z1C#EYbs_b1w4WbZTJEfOR?6P{pnmF*O>m4aD_`~-5LrJ=anP7gTd zb!tw!=q{~_7N{Z2{H4b|pijcYR=%KdGFKzE=y*DPHvaJMr>620dUQmb{&6|UwB!9! zrNMj`Q_OCP6i|`b$^#w{@#pl4lYw9l``y~zqd|FT!~Z$`UED!KWCpSb8L)kBmuEN* zN}~RM_PzwJ$@E&^Py9miV#vmld;`i7!61$-h5>bgGwpP4_xASoF1@|&Z7aREcAd8A z^v-m_9TgE&TtEYefFPhOE~o@>P;dta6~zrmM90BFaA8p4e@+q{5{cx4gnyiA=T}L- zyZ_j?@D^ubJtZT}DbctD@Uz9axHF*A|E9v9U$Lu*Sjx zh+wt;O#DUquLPFe)!Ae>oj9LhrSJL&P)xO$Y6mfTy>BM7Y3BZ!J+gM$6NnNnn{+z# zbZ9QOSh#%Bf^7k(IXQyts0HBW%)c&~C*mG1wiX+DNCzUv|@Nzt0_MBzEQ` zKRM#DC1cnM)%K3pC&JaTRJYx*m4tXG65di}E8KPlEfQQ6Zv6gmunwl@~53!!x9qA7n={&m$dEUB@H2!GMvzid+b@(#pUHzV= zbcECHn5~FG^l(?kK&MAhuSAWV-IBtf^{6Y*4^Six$O({gQg;e_qKwT>upJv>85(jG zi26Kq@(yrjv^2UWm3|N$=57eo$<;gNLCjaA^C>|Z11vaRZS)#OFF8gATsAATV61YK z4fI)Bja3K6W^*|{_}NC$2+SoJy(aaEm`{|p}$lHvAqHT z#`CZNn3l#7K6XzRsAuz;P|UC#|bOZ}W=p~`0@eDl@8ZPbiEuOPdI zqv|+upzgSt^L&V6AXT-Oit7yTlXOV8gKJ#r*Uc~TFQhP~`Iys8(2*{rumYx;Kn0Ab zhSLCQ0hs2@A~zz`o#9AIhUV{}gjB;>?Safcxc$!+OX-#kSdzk%QD zeo4|vKcq2nnWOyBr(c>9lgVl0Uk*)B71RALSmfLnTIgQsw$Zahv`U4=N9BwvpZRMUuz)Shnf)bWrcC z^C=e|4bjQY&ncL`+dGTfF>jQ3+DlNDx zHj%NjNb9q2bL2*gUIri&08r2Pan^kYDT9EPO>pwpko1-j@!j7uQiOBFd9cip-z;rftPRHn=PPXGq zeeOWa+0DHwG`ib;?q!hC?whQ>D(;)SZ%(3DneYVcI3d|UR|HiA6?qkT0Z~T>gsld; z&;43ZI;uxj`lU>%j6|Z3@iUO;gUg#ye~X3hV~3}}%vinn&9jGZ#240|g511g^>3Jf z^@N-CL!uoHSWX=FyJ7~cGZb@@BK1@pTFi{7?MjV&k=GAEf}-PwwDj#dxPo&(DQ2uuY*CN4X%51E59(&u|x5 z2b8bq9nH%gFVu$RD>Z?oGwyok&Ct1)3xO+REYL*bgC$Zr8D=v#1hOy56sdXR>P{`Hx@x$WlYcX+32t;y&%+%@AcdCxn2FaS;J(6tqa*GHoC{k8UM- zGs>bm!*d~w1CpP>uQRq+PqkxLZ>tl=5+kO?#p-Z&JHZMg6YCex4z)BOR=Y^{lFG0H z0SEjW+=~r1d8_yBS$*7lymINz;AT|{@2s+qyDn_}9#fWz4R-sAM;wUN960ZCU-!Y! z{?Af)!f6eN6_E*Jcm;h+HI@jh0ON5a#2{Fn=9%WHbx8KwIzCt!4eNtA;0!CkP8|5L zVz*@?icOHtc@b|-x+XQCj=PqdK|lKmoS8@DfwV`4dL^C0W>vjBlb7#R%E?!jMJKx^ z@*ap2LbiZsdwUjGM~@T-TSXSQvfSl308O6_`Pcsp{oMr9hDVvb|Ig&9i_`Af@aOnzK@WbHBhK*Ph2#8_o#iN~eX!w+r7I3SgCGfM&=k*d zKR9j3A~gOK?-4OHXdA{ELvK?LzwTzT67Q)m{DCB}V_rIO(kIux*FhFGd@X>qdyA59~gJEjE zbH@uRt%0Zmu@DVVHIbzA4}&p?yNn3!yOw1_PP;L&Quu?-ky<1m#R*|i5P+!;SrO3& zQ$6n5(D0qw`0x*Y`{VYLoo~Us_4s=N(<<^#+2lWyL?>ow7nrRJTPbD>Zi?a>q+8@# z(Bnm#kYUe?X&RJ8Yzo^r7c(j7LtJ8dC3nLc9YTy9ZfGQSA5>{6DUB=bkizU zja*VZ+|-Wcn60oF%V&({URJQ6ULgl(PdCBBMWozMa-0|zADF>nFU6Enpo$-dVu#HJ z+Co*xRj-W!4Sb#aeiWwP%B8iwCDOxlnpK&cLe5R@dAdh>*7uw+o15;P!rP$C4b=0W z^5CqfUfSh@{n4>WmQgzc^`owaHH=;v+{)zHikr{uiIrDY#vkorcDQ-;%O8bCns6g@ zjm;)|CXq8{(^E|`Kr=Z=#bwZ^zK!! zM=_`iR~BX9u?`l98um849K4?Q=hfsd5C%!oA+ zm2Q{kouIqH#Yup7{VvB{k8(1C)8{T_npK7L8IsIfDa+#|Pix^Un^erp^wUe%L@b}A zk)zT^gXE^0-2pQia16I(v6*;VuC=~ko!y%Dnu`fGg`fW7eG>1)P7QSN8DcAvDP|2t zR#9;nHZ2fr2n0dQ#;}DKypBh2_d6(cTuGAEZ`{V|w|c*$PFSD*VaOm8d(S0=S}rzD z8#1x7e3bc?hg5(f9d(_>#Xu!wDHYcXg5LQ8 ztP#5|t_nFzUkprBR)`jZU~UH_X%r+Z{oC)Ez_`PaK7NWxQpH;zC6#ah#D1i%z^amoju{h4m$bmU+J4@ zg3i(vp+}*j&Uy2^;EN7faMn^xB1KkGaS!KLy`hiI|l4^_C?Y(F*CW-A5m}n|n%l zliMT11|L|!(n@k8vOV&}kH;-k1KoU{8-_mZdVIw+k zfDkmBhP5c=KG+fjmzhS-vM8O*kU}ntLazK?MQ(&)c`|BdV-C?VYtHi+wHfB0EkVx@ z;(#MHVgD_)WEWYXfWBKlT0p$O8<_UEMgadcVg3#VP-l~C?<&_tEo7Y%CCw>PSFNY#hj$DwY>6a z|34@vHj+;4jawVW`?0>U{ohQOxcq~%$E0aEH`Iw|zT0f(brjP|k!w_3q2%7=d)Q!8 z3L=k3puuVeogcI=xEg%UOQ8FSjFSPEOb#~e#mHJNh))-}t(EPZr_P{Hf{;xuD4Mi} zWzgvnJy9rEQW)`wqh1%RQ5sqZ?;suFNgPZrXK_10FERs~5!C>Ft<7Vlr$Mp@bw$;; zJoACtQVXrk+V~5WhVGt|B7>F>{n9*6Pn6pD3*tc7{tCPJb<(9EgHY zCsoh}C;{H;n(JEWmdm|Owt1CCFMR*qJG`7QY#EJoR*+^g7IRkml~3#QTG%5^5v04R z4+b}Smb<3SS-4QkMHQ!HUXHREW9-R5M*+pk{M&rk zI#7o5#sA}asoKi~p6}IcyGYi5Zq_Zx4-VNxXHyK&C$|kYr%9D<2t?zgksp<8m!Nzt zw0P~L@6n4%{M;R4{ZM@^@ZV}c$@Buf$IZb>qcLF0W_*n1!opwOs-JH{$?SJOY$KKI zP~yB>KGI@_lG7A(f+BTP-09Fx+E^yMDq4;DYsTt7b#6p1T}4WRPr)*m%v&CDB&5>K z*uE}R*2b@ym&8#Y4p|SF(mmR#GHXYKgjd7zf;%DR*DI-p`?Y~Q+Gy12sg~>!?18LRfuN4)ICXT- zw3-oF6ShK(&4OgpwIjgt?eFcM(r*$syZWt3Un7rQhH;>s7)VeDGQ|H~O)&`+Sw_X_ z!1HbJ+$ZX!7vlfMD@$bdZQC;EW;cw}8|!_PyH^zLsG`?lBvt2x#0K?xM0Af2~q^@^dk@G{Nju6 zGw6ExVsd-d;Se3D>a>ANi)?fd7)7rZ3Fl4nTxAsjmlxG*+dK}0w$ayR7?8oJjk?`u zmjK!7+OTR#F;`vSznb%K);b_$r9^$`(;zJqf~YcHqdrAWkvjSasp0hrdZuaQXr+&aixpbk9Ny6#johC#b4T8%NOk0I5rQFhbU#7Cz--}*Qg7jqI*LPlaGSyBGW4G5fJ`x)`VAaZ_aUDsO9hR44ZECV02WB6c5JtId zL}X7?Rdf=s4CoSP#0_DFkRPSrbc-nRLk!a9#U3?6r$nz>$|0~x1is12K}k@lCh~K z+<-^AJv-@ES*BxY~W_ zzs>`fzczlBa$t5ft;S!Y84MdPeJccaWa>ih@#uc(M(CZkC%RY&;UR4p{z6DJMSng6 z{PF_9P3|JW=SIh$kQ;CO)&#lVHU4x4`Iy~i&56CLRc2n*6N>qiA`dY)YV0kYD!b!V z0Q({Ei*VP4)sm^Qy`({U-|w_{4g?b66dpu!A9_`=6;q@!L)d;9b9>$IM^*YYc^(PD z&g0lRKqK#fmq^vM1n}lqpk_d52<)=?ujRUME!i<|#jwcGKzdB+dz+ zE}wIBCeNY!LR%I`YyPz(SIF4^3_16*mvHwIoh*s-nAAk3aQpmABbrqh#=&W02`4REq*&CD{Dm>E-uJqoZ3z2vhT1bO@uJNwhWXHT>zfw*nr-WOrJV|%i zo$)vpB$y$@&ahzRXHy^Li=~$OxNQ7vCtl&KsOheXJ{N?v!5l;UuNk7$dGs+?y#zXs zGPe#2@|jm5*%H)Ygmd zz1rov_LX8NPTV)+CbWHr{-?gx^!8V3{F^)*naj-Dz?;$UUmv%!5F4Mm&*s3&H)j9C ztmeRZz5fs9UyT2bX`UMo)JR?5{Iqclng-;&1lktDpQd?pHtl#ezQ=#gd#C*ThOsd+Z~1@KwF!10|#%O#tmI*VUHX)I# zJ|0~uFA(kXJx;pC$n46DS|5W9QjHv6Ie_Oi2DF%i$j;bFKP1ny+%kVoRvIhav{Gdy zK3Cn;HU3Y1F&U0Z-~%pD&<)J4PEXju<@x8)_o6yMHN>&@W5`6Z1;>zE9s2UU2G>7Y zg2QRIG*+g&Hlj8niQI~;i0p+_@J-nfo*D@+$Q}4-N7apcw_dgbJmUOBK$9&o|ZgI-?zl<_EE88xcx10JiR#CR{y zvy7y%@kyN6gS8TO)6*?b0htkc&rMw=EEZO}p}b5UjWH5jtQ=rbtbX}$!ggG|XcYUO zdpqn#gk>a+4KAEFH(@Qtw$c5=n0{%qYLn=uJ5&-uspC?wTR!(C{nL85HzF&Q``l0{ z(|80W?zLe%Y2B+@{{n8gZ~$aAQFe1DkkAUF`k)PT(!qQwF0VmrrGjv6=cx`{k zOvpJFchmZN6bQ7!c93#U z2NwmR^{KLgAg!=dh$Tu11gin=hIaz34f%&xg_ZNvd7|CGrA!D}%|7crHc5xf zz*s~v&^alOiaSW}mpysS_%9^QK~53df@34r2L96^d>;?c#AEd_X~FN8a_VC+yM8me z6P#o?Ex=QordLb%Dvr!_SQp^2qiaNny4CcK>Z`ZunTbxv%K_Lw1HJH>#s z)mChkRVG{vWn$`TX`;7Aem&qE=c=HAXAoZPj(H+(<_x%C$LIJ^z=6v$-ZEYmcV!Gp zH;il*M)jqBcGGC5{j(7q^NO8Osql1hbgp>33pH=dETSNurq56iL9H52kW} z-ndIq49VlF=zmns4uuvgKukOsLjTCy1c;VbFTO)III-1$&R9e0sCQ6I7DY73sX@K5 z15jGJS)h|+Yph-%)uhkGO z-)ZFa@*2rWCVg&;B-JfX0C0EcSx^zBheJhH4eRL zdc6X+<;ETdOahTutB|uMTOlD?Iv%Uc0>$1hrmR zLth}eJqx8~I)y!;Lazmvc|)KkaK+R;3cLkso5|f^ptAUZl}eTEQFQU4)E<}(bl!{< zi2gymaF+npuZ-FuS{TS)II{^mHkgEGm)Hu95w1%)|D9isrKs77u>(*UvTzM4YOa`* zIJXr_3kzf0W$!(#@Fl*2yaU^5}=e zUbq-$dd3EoVFt9*q5EHf9D&Kqd@-cyPm-cYq(ZIsK>Bv%!>Bf3|J%&3@aLnNvFtW+#_%Eso~rKSEHIP;8xSi1m)6N-@T$H zRWFDUen9t2@z{A)vG<+uYh>I?+SwZuyJ6sk1F-{0LQwlZz3j%M0;*|~AV_g7Cfp0m_WK3pk`EET~?!?AqhnX?iNHOavl1#;w1}8&a z2&74yeSwn#+YpdPFy@2GqOcsgD%xQn7;c`}0AslO9d(QyFuwdR|66G3zkW`bb5^|F z%CI6gO>`Q6vm__Dk>BOotlA~sGObniDTuP|0rA|d06YW2w{T6WY{5>>f@alLMR9aD zY>1288ay`!^t&vUz`xQ?nWF&523zLl#`MlX;F@%F%VmMdjJ#7e`OhTra}#I-4$}~C zcPqtgK?YV_pR76yld?TZIPE?yAkd}8@Q48^mn40Fl$4kXdIv}(ILbr} zJCSUFG3)>iI{5vctz2)JXtt7}?W1?ht5Is?N2X!=)RaL>3q~Ervd9{LXd$GLSAYTw z_yMT!nlH|y``j0kMS{DYyCIUPM%`jakOkkvwhGwjY_Z^4%u-OGRWAi_=9+fozmrIJ^%q7#SG{adDW zD2?-`s#i-bF

    ^1^T7!{pDjOH2GqXmL{j;H57KPR?f;`#d=eI)&G z%Q4zyZ65w};7PLBi6d~(;C0Amcpb&8rAQ(bcSKPY@=$;tIQA+75B!doPWA|bat`!G ztjvpT)9lD>&hWpO_2p(5UiQRK-%Itdk(Zqraa9PdGa4b{t&!_wAH8Z|>!iwn81|1k z;K8|I_3bH{8JaN5k8<`1`lp?tHwJB%Uz#@q z=wV?!qzgQ?8vlE|E%I!Lmz)u&%eMP%oRS08vT%Fj2bDj4ugWfJf z9h?mmlR}XsD(>Q(O!4yB#(&{i;C>8&+}`k{5RH8K>~Z&TY(=+in6%$j9Z}!-l_e~m zkE^_Vldl!74XoiTfjrMix{4PsSnmF*Z!NFVt=*?7@*k6L!jE;+wm$Rm#&4OHqXYjD z^h0vei8na6&7S`jifN|Ec`B|tq~q08nKn!>Eu%Z6=#XHL4OsCS=)Zv{It_9gYV0w0 zN_a3hRn{-N24$!yV{GzCu(jA}c{k`qmMYqjP_KbAsj^%2CvXV6MkjHQA(uxl7NFy9 zxS%WwYb=*W=;^-6>f{;v$Tf^HuYrybYFM`Mhvp~@+iU!|2exdmZ8ak{u7EaPHnO>~ zj-=ssN|I00eq^b8?6hHjD@w+VphkN`+!YF1#@N;eCwJWK3eJlyh!}bP?0*G2L{wIb z{*z;}H@`jpu%0Y&VtWIX?n449>nSFgB5SBPleRV1SQz{DYUY-Po|)1~JCaImW#Vi> z1wQ;%Zg<#WcBsg?HCz9-2^GhFcI9JI`?(3~x0zw6iDDWla)ye#ChnIuc$TSda?e9H zJ7F#g?i>v%q)}IApAUuY8ND5ru@9t45xKPSigj*D z1b2>2oJ{5xw_Q-JJPIu?izD!NVr|}R1)+_rkk!lBqIF(V)3+ohSPAb5e?yX;*w_@C z0VR`SU_DBs;y!@tU+mtYZXjh!Z0SDWlI5iX4a#^yh4_vP1Ng>qj-|qxc-Qx3IE26CUSn|8y*M$xcjuvQn3f!nURr453mM*56^(Vt z>NARS3Tz*4+_^Z=)wOB%t*!))b7)wZkBLumfBg1WO)E^uvYG!%j!hyUG%>_$yg)G@ zQlx>3yEl0qxhCEVS`jF!S23qNs8tpZEe-Hkv+5$Jo5thUL>hTz*g8@e)vVI<8=#=~ z^1RBhW=^xJGE5&?8CDpzaZ2ITW>pzXYd5_wrYGvG=&Yz{##zzB*N>6Q^NxttPOS{< z5aAV-VO_$liq&uCa(iI3b)=hjGkUAy!W+i-_#13h{8hbV{xxxj6pyi5HO4xBFPPfr zPO*8gS0w<51CA(de|@K;f(zg4!vx9P)k3=rY>S z8%hU47eLZajp$uW+zfaI4QfP9b%U&II|SO1BF464@~*`{{%-!w1r0 zVH>TJJtWOcn@1zp!A4{hI5A!ydS$599E(Q_=q%25Ba@fpoDW(`;^zQ`mF6aV%YxW~< zN>oqOUawA&(AUXsa<};uh}MOrNx&$c@YYMPEMyyRbM*S)erZ|MhQI^p!##cFL}Bj>QZ++STL+^;+{C4NdCILx5?s*~M$C1H9VeE}4VvGKOL zS)MAxzNP9U@}Y+g*euPU6|1L5Ek8WcT(Uxn`C4IoH_``;@;ob~OnA~8ym2gT#+)}8 z2m)S1L&~WXvyLKbskoz(2G4v>`OJ;(b#y%tdO=|*|grdOgLsG)I2&(HS-gDJ$uRngufD1@V0u82Kx?|c#_mbE` zsx)>j*ZBD;Qmr=p&IVJqS#l>lUeGDTKOm3RhG7e=9LqWYn;|ru5wi2Kp>Shoh&KQ9 z#N02II!`w*fn$&{7at1BpVks|%k%jo#vB(OFBg9-D~Wjuz28%f;^(Hj`ecqYT4Ti4 zP&nB2^^NVn__5`(<+M>-D+`f{M|)k+B!3tFcS6$02f8crW*hma`IX~!V;rT=SSE_Ofjhx zSx3b+$u-J+pTx<~>cZMKd5Br-I5o#<0I;KbxA(=7*7z_rn(VdSDo2%SR){z7bc7VDTbB;b5qK7%o z_E_VpWA<=Bz6+9oXO#_})pP?t2{5%^QtJmQPF@eBOFajm47!1TFeWkdL3kUTBzhq2 zk#$Y2^*a}lZ`T?`yP{z<({AM+R(6OYx`v0oCOh=QvlGsfGZAA1tXAeKv$#d#O84fNJO#+vDt5)R@hig) z1RU^h;BO2_hTw0!;1NmwB6e?C^9OC{FHK;1e{I66q#Xt;vSO=5?D#9`MJuYqT7lz&!Zt&U)twUCfx>}P;P{=z!ju@ zGF^Kmm2MyMx_!!`PNYP@)UQGK)s1`U7RLKk!c~}ohB(&s4>yKOAPmC(D(SIv3*>_ohva+tsR#ZV4RruAi0t z`gb!;(3{_ypGXRv*p;p~vpfeW2Ks=M8_AZvWXHT52-B_rm}Gz-4aLikg7-zMKkz-_ z*9Nn$Zg9;O^Svi7@i`;DC;miON8f$p zdhB-kX;6o>Gq_pxC}!mxtRSePOJcEpUj5`s-sipcLqB=ZZle#u$l7FT+0e3 zRN4P6OMcn)<(lLexHEM1$fLW!R5ZgLNS*JNZ;Q^sXue|W_t**w>mxYk3@cDfh|Td(%!@z0Q6_e$O^PaNrJNFIHL7|$;CTnyvkQ9C})w$9P8Yw~jk?KfYn=E!;X zAX^11ay(34%TKS81tewCFmXaBreE$h^JKCp2I^STsW@~}9)h6!V_JPJx(IvR?+SV% zYm9|vGFOE==}X=hIi<1h>SEt*jLitfOtq18gF5qe1n+Qg8$FmCp(M6<$-KNTk(R0=eJv9`grN|-ZARp2<^G;}b#65b`^u;_q$oK#s3%5lbbmrun z+#cD9n9RvKPJ`V(F{JRK}*^5aclfzBKPVIQ044N}zDA)9+4ff+@j#O$N z+uV<7dS7_6b;~l$QZw_G;c|_Lx3va*P$who?5u2SV%Ri+K-cz4H(vXTR@} z;p;kj*`&eq%O~Lpjr@ksqZqt4nXKjRr162oE64W0z4$MO3>EpsZ+@r#m0>t^H=V;>3~Of|kPbIaNpt_b*#kE=qh;mE8Rg-j&y@8?q)=A! z7t8)=PMgfK((&w$7rGn}0nVFoHLQW#BI$GQl~h9zRvXyip}s2I5SZ=JMyJl&52*+H z#Oa29w^m-Whk})X?0ncd6oy;BCLEgWJ=J8CzPamXzaU#Dk%MOIS0Tkfv3o8Rmjqkv z9TB?(zuf=N**`6RWB1o?d}G`EUpA?-{F+tE-@F?8r}xjRnpAh*Jg)-Xk7(3#UJ7lm zHi>crIs=ktRLrU2Ws4g;>*y@M(oj5R_F&oTOJF}M&p&FMW&znTd&76N2^pJGLN1Yw?6%#`i+M;j4b5fdQVbA^IGBP z3TqayvM_<{p~B{l^Eb#0OMwP9_4iI3pRuAdQR1^pfW0HOdas$Okz>bJjU1yb=@Cux zHOx+rssJpnP2wbR@@7s3n&{IB`+aj{isCTDf%>d;CiOQe%-3@8@%yY#y}e!YU~^sR)> zpgDA!s7bZ%t(<>Z^3}&bDEU&(mu~&|=?`+>IIrq=>GVw&WTDEMdu3RbsDj!;K4t1T zIf8uuCe_0)XMA_X4@-ZxbbjXi^Qua>PiL0RtPHE5w$8~C-Iwg=>ZE^42-%`MCRid! z;xwrm-r5(t@}JLs`QkTIzkA^u=fD1^y9+jbGxJO5ZBmZ5Cxu}Ws~|`#>=ZVI z={WZQK}WeIG?I}u@($@^at>%Z&8kdJAqN$HGu_L$JzV4c8hL6!XK+UF8F53{)8GSh zalB?#KIb&a<77Z_W*havR&8NezzSC}5ULF*&yToc`HfzSPs`}sKoE{k3Dex2cZ z32I>B#Tt3O460;slnT<}`IuW7)*ql=>!J5cbW7%G0-M5mqK-m&Orxij@kW=dUGt`RZ*15(0p(Iu1r!k2|9fbeRe;fm}-ht zg2qcop`?L-#jn(-Us_KWN-(=~-m^HcTw3dkdy;0=4e@>cIj=2>T4>U}L#!ti^zzxY z&;$}%f5@uo0`4`ECb{L4$UP*l;b6l?NI23rpr#;`*Y8pplORi;f`ae2eDuC$bT<#e zb#$(B`RqN4u{%7Y`1(hFT~ znpuIw77J(x8)*2wtXhk|U%mGFYo-ON<)3cnl5%zd8YeEhJ7>1e)>2FjMYP7$eIA4f zuZC6A{XBIo>GMbu9iDxK15m*?=~GU`VBjW6gy%K#tr0aGb*YaY7Fn#F(#W&8D`WHk z8N91kvfVG&_4X_*5{JUQ!Ao#-On&X8abSd7kE}85z8~H;CZst4V9#JU7CZGp#rglS zRHU~e{C*_*nEw{?POcF);3r4b0=>&Tf`yQJ}_~NzvF)J72Avn)-Tmc(gO-%Omuj1Bg}9faOc$hbE9F(fV(1{5Dl@gcZ-wrK zU^Za4b6uRTLS@ekQG%N~Ux4}`59tp@U|J032v#hItU9L27?YWK;e?Luq;)g%zY%U; zW9c*N#E}}P&mEFZ?TcOvA|j{_vLO(uG~?IJ&0-F10gJ^gKf+|Nf;Wvet=-;zeeYoIx>LDYAu%YxG2lN~s9-0AL8q$Q*;f zk6w}twMtq5LWAd7J7g7XnH*c@e28;_FKroi%-2{q5X1N()N1cv{>pM&%*K&%;*F1$ z=8&if-y%7rDB%KK4r})_L9|G{EF_2S0L}Pz2)8{DC%d)?``kO}cOALH44G;6SQ>J( z17FskJ}}8rIGhbSoOnrr>VY9k%CmkKyTpeSU3>t>zAPSxz`nT~vI%}C0w31J2;vkcH3rFvFI<%ACB-*Y-7U3FZ7v1a&Ri3bk z%m!`#@%&m@e5vDPvW;IbJHzd?ogL4K{wSMh@;R2S2t7&$T%7iICELubz*>q)q{vDt z4g((;UsV@HW+~ffd^w&7efGfF0@Y_1kK+qgP@P!&(RY-V)`HC{;>13N6{FW8(MwxE z2fD)-q&0C@Vk#9tPd1;^OLVepK^V6mKLL7MkTld=9WOAw_f7v>;U;U56}V#~DR$x= z$|*Bzp{1BZ6gfb}r3K#uQH2J6r!Zbn$g7V@56zr?Ahdw1gLF-N=nd~q-!3=cK^fF` z`(17X+;s1EIYvKzz25~RVf`*BIg$~K_a7seXVl26IN3qzp-1?&uu$%fZ6gKTqp=-) z9QW|l9($2b*g|3`JcfU5AJDKeF4Vu@D*F4&-lGE77h|d()D^H22sNKD@iNAo{b=UH zR#c2;?6KU#3KbLge5dA*FMHs_SQ>`KVCDXxPvpAD8+rzG(-(<8*w7Uq$I#{sgpTP- z$klYwZ64zT7h4fxW+TQj^)}wi3Lz8TobYd1mi+0LDy(jWC2jOHG;&m`x*LwNaizh@ z#1P+D;Z+r|GcsGehr2`BBCDZ$h2sMlm_PQIv2oJJc$sy~{G|I271KZ9dr^gJ+JB&(0k4Ulg@cw$`AX=79YXd(fDf9MhXIB4n7`t%GJx-;MA7zyunB zVEuQ<#g_~;&>vt3XgVlHN0C-4u8m)1>MwDUM#^C|y-d*S-UXo>C=Uh3e-0Frw$s~{ z*ThX;J=~kB1&yAC9BmjD4RjM7r+|AYvRSp&yOy_E@-!%)GkC0yzQpu`)Pi2R+#Q>n z6mmCBZ&u-559mZ5-ldVJNy?;HjDQU6!TYKL)_7duWCbMiutcHX1+_f-T^fYhT=-Xd zkCzdgBf*+`_Y&8_kk8GFFDUV#p`j_R>104s3e* zm$yuqdi0gIe?pEp@m~Cf8Gg=D3=o}8QE^v=2V>A0KPOAH$78MR64Ol@JT>y;u9*=xRmWX>!*6+_WWqh3 zR=Ad1Fdfx)(>SfN6|h79)aR0_Szh9!k?$qP1vUP~k(X51D;vrJRK?zlpoC!GoHOG4 z&>=-5$J(bCnxEk(ZhVZ=@Spvnqt*d5T!Qib{MhJ8CQzyL@2JTlb~Jw{P9bHOK`@14 zk|?s8ii0*cq3R}(<1nNbJ(}dXbQS5OM}dGB&4MlGY(tPOBiZ*FYVFcnzc67U}5c-0nWH#f~w zod$PygeP%&q-TAfNX`)TUwUK^J;M7o1a{JYT_m^*%^4a!J3V`%23#6F55?YsK1rWS zw{voQ)VcI>awxXa4-1@%B8}tsNYf*#p-Ke*!x*)ei;0gmejmMKUX8K?YBnpwuwDuO zGh*2`cnz~=Iqr}Fr`8SBcl19ieBA`RAKpHfP7Z!520?_ zau1YvZTH00vjqTG$1!5EGUhQU1D)VB??f-;=zOfu$k+c%y;Lhip?f`j-yQI-T*C~a zhP5eiJj_`CN`@P9zJK)Xu~286*uGhz&gkh!1jM2tBDvq^1Mb4d!lj;hbcs*ATLP65 zRZFgl8-PIrm5=a#+*Uib=osBxv%hQa<)7cTY`c_c_KdEhn6(s1q~f|gyM1<(O5S2Y zGY6j0JbK)Vc%J=vHh!DWt{<-#)=l~_KBx0OSgU_f{|6t_`qT82eGADJC*EN}d%q#E zpnQsfUfS7ITrZO%yBV$XDWSXRdy~s&E>~5|{L{XeEwAJ$K8dUh&vs_y}abIWrc{- zRvB4QiGm)pG0S{!dzMRk#r@LjvQp0NS>@9DnEIHT(N(-$ZoBXZ^cCoL+38WQ=oJ?R z>FL4ayXciMX9W(02~6e~kTNzXjBqzQq-;7Kbj8bLRlZlV?IKyv&Z~9eBu0su4a%k% zSVFf^aaTAAt~%LbYDs8eY$8(>oh~m8-KnSrDe*3n`}$5A&p#xk(524UQ(Y%pr78_Q zFRk=*yk9a(OpSnuv7he!O0h zP>my1)*Id$j+);YB0QtkNc56M&*S92`+C^D76@)acVP!&V3d<@T8hR#$<{_-Wl$zo z?F!OLOa?`GPxu>>%+8=VF=4OR3|N^I0}RkKV-w?|$XLZGW z^jg_~OO5|iUnpx8s4qlfwaKV=wAN;6nB$`a)G#C2_fU@KuQpp&C9>gkI#DGgzgX}{3d(b9m|kGw^g*~6gJcCF6a#uBEsDJ4okKtN za=aehD1m~O$G*t07;VG*zCYrdY{~qKn?4|`oOsIvg1$qxJZTiOi6R@QxO2>Y?tUm{ z?~Lj4ucUb2o|P87e*Pne)~gRY zhqhXQVQ+TeL3XeZt$ODxmJNoT)>Q>s)exKUteNmlAj}lUw#%SYl~mK!q+HxhZiwr^ zzgVQiBjvDJi5K)lm5WQHDWC$5uN^HhWt?DJF=7@UaUeDZhh~S7oo@$oV@;;xn@7LD zha6_-qdKtza@h<~rzqw)MUGK%nLO+wjLL(itxujLi<`yWIAg$NpKqT>G7qVr`+S?^ z=Y-f^_5euyZ%|_9dB6p|Ssmv{2v#?)5a6-xe%(I(E`1(nebw2D#nbEP54bwc(GdI> zw3r2$7QzM;cypsC-j+vS=61VlxnoRcj|zNdNF3?SAe`9!{L-Do2bSS_r!_fN@(4+y zLGrIwu@~BooPs>vW`%Z~>fz14Z`-EX?3QtP!^*}?cz0Ut^`Dt+Op)8{>Ey=e#>VuU z*_bYh>7>Xg9!I|!$@$gd4RX4{oq((jGW;0~}H|&hkvUSSp0FzN# z`p%R@lEKa>IdQbV(#$CBrWk1DmruoEI=tC;DG;7g!JWpMNnoHWR{|sWg5(0|Ug~5E zE|F?wCipB1s+CI&di8}gw!qdXPbf9Y9FQD1rfil}ORAN?uX3$cCUG3GRUb9JhME!k z!)wCRjTP7ZO^A8QoLosX>=5I`;oQS!h$*HRsEXQ2#ih*7UJyQ~5>8_gy> zef++qZ~J^vwiEj!Ry^1i$*0l+5ws`+UlcjfxElh)V<0fJNV=vT=V4>30T(1LbSSF= zt^r%BjV^=Y+H|>T3F?VL_oPLJoPUi+b%^DL)-Fi054`R2tC$U99o~<3q z&P@y6L1!szd6$IOpf(jY9Q;zBp;zY3c;XI%#Sr<(oBB!Q9z`whVbodS9*<^~-tTBg zr*A!{lv~Mbk-+4EPl2oy92rN-Q7-k|<92&y23@8&Eq_4kI2!^%cRer@8j>5w(aAv1 zVMc>{4X??wm76JQR%MfioKwO=P9D9J?sdudb4aPUqmM$JBRID*}J;c2s}d-lfq z91cU_bzT$oA9sBJXi4GyoDd(Z5Po|=vZ{{042V|@q!hCTb_4h>J411 zbK0dSj%<-Ys1D6q3kW#(NSi~JghKaf+*sv7SwII}=&+)IF+=_tNbUP#arB4xEQy&` z7>#;^$Q+b`VfqHME?QwS558H^mT3dw#Yy+n=4%w%2y(bDa*0@2|a@pHjD ziU*?RViG?$o4dw47r0e$Xopf8hV^9+X5~OV?ZQi>*BwKDRnbYjGT}m_k%DfN5aq*x zO{PlAPY^&FfH7ta7Gil@8y}Mgctxx1h@Vk*YhfREC49KC?v0Ho-)J9=7p|XHPY(YW z40km5hZl~$?8F)@V?}5@1pOKHBW{_0fv8;yK&a#NO41!!KQi1GYWqYFcemq?v9p0I z?)~-$mXY9>YJX)V-iyz*F^#xev__=fEA66d6pj2=?jwZ;LjI_qhGomxO`?_D@mRsz z=F{tLxD5Rj_4dGuIb|TYcUaj?R)bK;wwT-qN35#0gUG50a0pw|I*2NCcmL0_FV^#b zao)8R__)jBWxR8zC_}SMr`d79CVDz6cnmernV3V;r%;4jSM4C!bGdnykkB*8D_LFr0#v^G1 z=H3XXp!SMS@|FmiR9mQ;e;;_W;Rj3SpI5cfC?khdrY8AWUnp$m+?U)4*vJJr!_Y(? zHnHw^!KT&q9G$e%?KbmtW_x%Y^jE=+QDGE{_jJ>H$!*9q=zwfzN5HWI)C)E?hErBh zJBq{B0VUu1=Ovcqtxn5Zuu`z9k6ZzzXV`Z25lJRFbdgt+G)q)LEeMkOB2iw{uCk6MH4+%xp<5 z#nezli(ZM5(S#XD>@Q%DoLj~_90Ia)bZKm#d;DA@Pg+l>NsbFJ@~SpA6vqor&WE>| zJp54>bry6nJE5MW%pd;`F0Vp;iqjfCSPgPQs-^XiuGWT~n6}0nvlpL<#Ic?z;|-&M z(MYys@cbx$CBtp3U^L-C>EpjzvbouW_MLbEv(l~cXh<7BNBN;orz|CAKerhgxSo&A z5cNxI$>I5Z+$XPTHf%o4i9zhNiT$ zWHHHON40d~(A!Znv$UULpb>Bh6=%%UU<#-Ph!{75%BQK3Vq58kB^@up6|finU8HPQ^^y{4XG{bCX;1^dz&}H=#-j(g{THFI_HXU75}*wZp2h}B zjeH2r%*ERsVNmP>0=rH9X9KZXj;Lv8cYX6^-%xLz)&j!q_41{>JG^6FC(cvxkub5LHh!+5BG8L%BX~5u9>**q<3A@L1;CaNxG05eoUcaG06HxOhLd4=LP6cfG_G2YNGI_7J`oSN zR4WJR1?p5;6ZFYACB$CBYTUQ?$nJWkkh8IB)WyZ4C9yqx^)c7f$uEo@v%%MJ3(gpi zS;Z(Td`>W7sKl z&_PznnE21%(7iR?gbWvvazDvow}p1%rR4)N9POnTC|}%7#TEEp^XcJMk&mTaQw&r` z1Xio8mte1ni;-8|9?UvPj)s7{5}AW~;u_^DsSdcphE9M_+%feu;L=OhFnH}DsE{@g zEU_dX3mi8D+BaFr24y2UD|STO?`OHlJWaSc@%3+9BkLxSJ!W6{c8Y->FI&O0gu;&W zz;5V<(LhQ${j!VFPv~kxLZ(`pDTju7ki*NP^^!cgOu9|d32B!MIzjO9tZ{gnNB8a8 z@oTg)6su?0dE=b3N|*oLvUj=D=8CLzBflH_;n((mc@tIkuaAGKk6rinwJ-I(_3qv; zfA}-?-Y;iRmwvwP?RPg(SzpPYKW+>b)}XZlGJ79p1=@-C3NzMQ?rEGB8&-BH=R=mO zbhPnblRsf36Z@+d%$B`+ihnYfn*iCOdd2{PeGu~?XO7hp=ZC2e7H+a@P8z7ro zC(7mCCe5mR&M`tp=rP6)N~gDFm%i*< zpE6+<(L>BFVNR3$o|`d2+$7ga+T*f!b-9;K#^n zizJEH@2N(qw+xWXK`KB&Ae5-*!hr%&J`m*l+*f}5(+j#5j1tmvPeDjR^uKfw(OwVI))z43QL2`8+2Y>}OoXS(Kq9E-6|YgkQf z7V~QHG~4&BB|=!utMevxe{)cL#y+J*?WFr<-Q%$N8ztB+v1#}Bf%ltF zS+*{+(jlnC=Zsh*-xhj%7V<)pd_Pv`WIMq7L~->63;Mo&C$>p-Lb)OEysBB%@zwkA zZd>T`+2>W2ZWV7gsrqCc{L%Dgw*4<2nmB=6X$*EE-%nd{$i;qQ!+;rJRiTxCKiFhZr~!9rwMCHp+q#AHC?5*kc&(&wODI&a27|^!(oIw zZ?3<%)6%jyZE(hl@(T8nF<1o9cXy8$YzVyPj^20y^zwZ~mPV9C?ej%VDX2tCm0^cP zjXX2rtZ$_olH&4!uG`EcP?=HdV_wKF7~y+-@#(SJ3P)KBR?FjrZ`U5yTW+w~c(qQP zL9&vG)YGt^l2lFB_!x*SH)Th7CD_>N5a?^K_E4|$cp@GLVVTdc#STZG-RJ06M~n43 zj%m&L?eT~8WXb`GVMrX7%{@YPtflSP*V zqBf@63v4*^LFU*GZlK;pU&3 z>TSMI)EWDY%*j&?&h!wh1j=FR0VtI z1!7(HKDKBR7KI||%Kv{Jsd-$#)8-ZVq7x(R_dxG+#f;SpbJ0J^+=|e4HSu|B2lWPoqu@d zPVW@oi;Mo}+@jc!TFBS0b|epYw||G6OEsX8!Uw zB$-_p#C|syNU0HuvMB~67t^T70Xl189b_f+ij!f@3mPDbHdx};`=xWzK$aIc{~+tF zS(@bCp{SqTCfx6VF*Mz>d7I*sz;!l^edtEha#Wc(QZ`1}=6kzWno0-Uu>a}l8@?4@HBufU_w!=Ak@9bujCqq9zGi0zvKe+zADde$({dyWHMsq|A#jK^s zDk?G;qE}0o3Gk0@deZV3O`iVPa4@-Hl0LG6!<3)NqL07mRiYc-x(d3wjZk^BGN6;M zQ8m(+7T$;5dE{TJHk_SQi!nAdzVr?z2^G4pVFeB9LTXMV&)|CoZ2j&rvdWIvBw0q5 zViUz|q(}l4x&A#Y)9Pp9ojOCS{f2^^fhs%>mF_+ePFT|td;{~GGPJQWNdsz zYC6-`Y#I@@*T9^)B99_Ya>xPk&H0&RvnV+vQ(VU>mX-QxRC~CW{BxZ$xrJe!e$B4; zJg-O&(!HSe@rZkiw@I?yCQQL5V{JJ|zO*8fv5I8}lK1!hAkNf!utAv}Z?l=PHTo?! z%x>KL>M{_FQ8!BwN2Os%Q9n6R{*g#x0i%((+5C9fy9?aH4AaW@%L~#;DZ7w{9cOkl z8%;uoDW;YpRaB%#aU=l$Yf*K=zbInX4O&54$OFO2*}LZFaZY_Z@trcCGxRyg@u;OY zkTv8As12g;(!edIFOVkiX|((;WHaXsUBhg4UMcG36^2zg)%dr$ZXktWD3c1wyL|k- zTU@S6qmRI(w?$GKHgT_q1&T&mqp^-BjItJI<*lU#6qS0Nl95h3c1~7K9T_lvNHO;) za+iukmWfmaN^UnxQNaCtkan0p6l1B`${p?*t{nk>~dKebQ*h$kQ5Gqze?&LXl=rM}xRLGKMZAcjfSCI!a@S*} zM=d}9RV)fX4YLA4OUP=EcK#Ni49udhid$4U+<0y}uLNp^z*-67fb0#`)t2lM-B;qx zXXso{jVfA@Ll1C|`VD`<5{)Vggq+dt9rCLdb_)**mx0K0na`3w`8D_f8WrAqEqnkt z&_&)6?tKL8s+B~0a-WNE%^fmrT z93MPpMyu!etCm`{;Rb<8r3#=@9b_mCv2nW zaX`O3|J6Z<>%45oR7Je^7akp)JrTCXk`4vq0hIOO^K6d^ZyxD!HcV0#Z@+tntY_zB z+wtlv&uB{9Mlo9`7zH4}ZIzS~^$AH-1a1vM4t*?hNLlZf7*4uTBH^2D$vx;MOBIw`Nc@e60! zFdN@98J}3qDpP;(FAmP%G?;R>`G>>RxGz1j}j$B9g${O`{+CC-ZiwAl1NquM^k>T{{=KrlWSc8X0s>Ic*5JF1#tnPSoikxJ1LsRS5LlksgsfZHkMx^JJ|wmp5p9U7vx^QI382v0-tZw1GuvN~P z;zK@qCJkI+-41NxcgQN7H~HL?=V2UL>rYmArhah8>3hF1 zz~;Z3|1*GmX2-BuXB0SmOfg?j<&D9qNv-J4vRkgvUk6Au$ybxhnqm-*@P zxv(vA@aE7}K3(Kxi5|YU>k;=wCvxAv;GFb=?5x)YmmFH7Is{afb(~UJ4xPuj&P^8` z@T_nykGM6rEWC=RQ8oCVlMc!r`ekqjc$1tEriUir5Z&6go-N`bV<`gnx%h@`#~w#pX#9VUN=G}MZsu%02y<9C)|J)K&DLF zm-)XhdX{Iaq*Df4Op@floAUh&PKmKv)h0G!tg*2K5@YpYi|ec)G1YxZnwmc#9X-SJOhw`ve{pD~ZPkuG0+ zc~RVC6~kJBCgZiaEwJ-!pR@n?+Kd;y)OmmS^;z9BwEQPJZM*?zZD_M&s&@}vCYyZ2 zVc|GGI}kQL!^&zhop&a&5f7$(tF>yDvqXZ$g6?T5=E!}4;8~;V~2`W zQSWhHbioEKjVor1@z%^&W0g=lz20TWL9dwFrxS?5Vj?YYFG0CUHGi+up1{-pSD63% z_kS1plV83c`NyC9>c4*(DW(|*LQ$jCuTJ^#H}k)1x}(I#->}~>BS|;%H#SgAJVj!t zNCTk?M&h1>fTYhMfx^D5+$#WzP&(uOc16|7DQJ9oJ+07_!(aNM>8zLG9-gl9b+jbn$SZfr6 zeMB++6uA${o17l^CRw$xkGs$7-fXN^&ZX-(`)5Z5=0idqZt$&f($Yx(hjY%yisgar zP_Bl28M!pXLV!L2gRr~q0xCeSoPB1MNB%c}8*<9|;r8VmFpFKTS` zX&Bn9&^Prze&z7h;Nf_hch)9SWXI_H*l6~trkFz%IY34B%s4G==OZ6`KPjK11r?Gg z!2{U*$qB7k5EFPm7}BG>uPwg4D2I!pcmokdvY3c;r&6zWe(sxB#OTo$7Gk3(NvSQvMQ_(pkWA1>`ztzqFQ_Y`2*9P zM>fum9p~Sf*{9U|$<=hGE@Ly^yIOcG9BZW-ILG|SA$wO%N!rd zFjZ${i#kq~FoE<52Z6agHy~NwEY;t+BYZc}6}mWQi?+#YeYV+xgUmW`vW|5b*-Jx_ zFnJC4FDeRKM?HCpZjHaf1iUD=BEGiP$IN3C2k@U<==0ZMTGPg?6t~(LK zG)S^$4>{lp4J1)O3FU!r?5jf#6?94j@=VqHHP7w?k2=qNROu0`4U)xR8GQ|9wclad z(Wlp=Z~Gg-Qqo<#jO5z!DhU*lxfqbl(&^T9TQ4)}k(rcw3MrEgvZ;i@N4VzSGnFbthq?4ucSL(9lM@t1u?2@VkJ znwAM-VI)FKhjT)o&ATU~Ij}|l`ew3ebmv(vR6F)_%|v{5Ek-Vp4*6cc9^s~lV|1f< z&|#NLV$eP=>gM6m1}KQ#G^5H1HPf)&=Yk7JfT&CA{6czbZuguXde8x{q>^}fs&`A! z@YTz6GL-thhWo#?aED)|Z_TVYPOj56Nwxo4`ETQR56Ml{HA$Lit^9WyRX+CtxY191 z_Ae|BPxWqsppE{UG^&-nmAt*K-So<7$Gz@(u9}tr^jUjf-RHB`r&Fp?B>*w>cE2=w zIoU~f&xq&lk!3*L<)h%W@^sD-aa=@`w3auFYbzphBs#}`Y{0R3M7F>CtpOi{-TKX* zBk?5jrJ3aFjI3NK#ejhAZeYWjndpKw4xLDLBm`FXSO%#PF&A{*94PmuqXoUryF=69 z&jg-&+X9WxAnhw|iG+wkR{GY|z-;>XA!hZrvVkI&3VUzS-Xa z0~P=P<^*%YWV`d`0WjL@@6g&$Of`VbNbTlv{&t^SnO`c45;QIiuayL*DykG`y(VukEE|KTQ?uKZ4{ zIv72Gv7e+-^GW-_%2e2I=zHx?8t0qlS(?%Fx$V(U^jYyk4wpImpn3}ZxNgW@n|x@D zjSAi5F==QR{-T!z6`g-PD78>Xb$4h2uba+r?vv|1I;=Ry0`+DoCc0w0Gu3vi<&kQfahWEut7G>ez)GF4+(-Yn2Q*H_q7!^o1ZAF5Ck&-?;&w13}b*h&nHI zwz3bD%=1N2^U~-vaG@LLG==D;$klNoNI!t~`#mrL>X^8JXCosy0dzkFx-H;s9^F4L z`zOUL17KGCc18?IV;2;&U#}bZHAja2_fQN}Am#zX9bYR+;MD@_T)%TPrw}%B9*`B% zexgxfy?idGhIbc~R`1ejzJ)$o`nvRjUxzH6cV3cBZ<=w}xn|a`&=s7cq=)HX@KX=n zYK#420D%>-WD6)wscL=wRD=OBKFik^kOS-xW5>n^sA5KJ79FD)*gdMJBAaBjz%GJv ztEgCZiRqJgsN=PTsW$;z;%(QY zuye>yf@CF$Gpa%dWr^YH>Y!SQMv*Vt9umhX5*9$B-jG8R>7lX4u|d{FgH&4Dbd7Rw zUZx8$o$U80ll9PZ_c(A7U`g>o0(d(D(P$|hq3Or)pd4oc<DXs%(^F7$NH zvWwl~#*UZ6M~x7EfMUuhQbI+3tVrM;e&?$AFE`|Uki@i%$q?NOyCY5E98h$5mCE`c z0Vsioscw17)1q1^4_Y?|R?Kt;-6lTAoa7w=Ca2^OJWd9B_A*jS=W}c6l>$_&C=<1b z3&Z+@pZHjFCSzk&Mp~FrkFEGBRAjIIrn><;EkD`6lx$@O9XnnZmm5JRk7B@DW@1q8 zFfg~_ULS1fp(TV&5+K{)cgiq8mqu^&Jt9t3>L}ZW{La&D!h_+xVx*IM1Pflcq`%Gk9vWXDTmNT5V@M1XBK z#pF{Y7Yzw+TIslxF)ttU6N}yV0XrpdNhZs&s5wNV<-~0da`m`5aAznX`Ja(?iX=P_e?AJ>tKz2LA zj@+i0E{bR&r!^F*52`^8z8km{(nBEQx!$kec>}l{eY5MlK(i__ED0DC_Q(==gR%pT z$ox~_SL<6ER___>1~L3J(RNjeYMWqV5R~^oIgsIWwz3J9&pDxap(PN7Yl5uV^)3xc zb@QSDXXKGfRjis<8kQRnB*mzoj(WE@yi=J=qim91)az5{#Na3=U~gEu zG?z|=AEO~T+)ia|h*}G)cO+W?vxmP~Pb2dT_ASTes0PT2CygP$^H-*%%b( zkEFIG?y*dhw8zngCTwDmFdjy0ell;azV@}cp9dODQ`fg|myo<^q|xY|uArFx6e*=5 zkIAClZu1&AL-SG<*W9aVtX64})+!rZGhOOu-w00f?xo|so21>rbs%&&qP6mnPn8q4 z$W&JIyXA=>`;7Z;<=if51Lu$uTOu#U=fLitWK{;Zi4BLz3u?vjFh3llZN1p7^!wfB z-@ft-19m=&U-KI2dcjg$RvFO-^;67!iu7V$LMQAK4ay)>_7O=W-NGWbK5nxVg)I`p zFtw#RNGk&xI`ojMgU$#~$R5ylJk+%kKsB_60IzQ^RCJ_=U?#a*>)EgC1d3Iqz0y=~ zm|u`rs9qg^$=zv-Aila=R9W{Z$8La&_OAOZQ6e03W*= z&mJL0--7OWt)yAn?3%#S{q}yMFRaanLbUU49f}T}`c5_e&@N6kOoYkuO!vBlP>L;l z0Yx8|El}?=4IJC;cBGkzcDi#ZcSm?m$Zl?TsQMUyDp`)a>v=wReDQRPYoh2GpTmkq z=-IAVnagKyx7it1=5p#TU+s;r8LZOOz7IYl4R%}ZD3Ckdehw=}4L5Yl6?R;FF>u!2Lcaa=0 zT?s^unw?$bjugue)yP@aEY)-7t@JMqt76iCI}g+H&>@E^L^NF0emiyH-=Xf_PIZ z;LiFRpVoqxrLeKI-S1w#byR9FL?^!S%`0RByH&p(hnAtJdBm!JC&i>uWIIF;W!W^Y z%R5PD;Bvuc(H?GLNCFSj*;lAQdc?JavxnCW)$I>Bd$?P;z4R%Pz&k(Vw#_AFVZ_D< z5=)1A{OvZL)qZ4ZRqWQ0F=Hn%!cssF&Z`9i=cBIczkBA5BR@OzT_8~a7k0Uzf=cz+ zsvLA!_U4g)N`33XS2Dku_0FI}kNiF_fqPicOJb(i^Q)YeO}_;#Z!xi|G+&D+7Iu$) zd=MBp0kQi&OIx#=hN!Q;+wh@jA|9J*$d0XwnY6u~%B$j7rk4!)Us{+ne~Y9)xQ=&9 zsZlNI_FYmcYyj2Br1>lTSMy5Uo1_gAjjDxHO84?mv?x_^9ZGJ}=p~=JEkVt!n{zhW z2wW4#mTd~Fy>h27H{IT}Tk~RO!?K-U!aSneU7EQlWp&BFo`>3MxVC;Uw|90UeVnr~ z=(OOzwfIjhd2QJZwZ>PBpBNSF8?fAVUszyWK%{7>#{A#l%o#Evk&@X*KwzUR(XQGB~T)bS%vcfhA*X z?p|A-XXTVk&FWsMbTuI2`*k}mkd1a6z6FKo5!N7+Vu0mo2WC;FDt5b-DzhQ;Az6L_ zrc=MkC$-=A1ex6#BCsF3SS}H-|xCGci2gJ7%&>g2nK7>TQgQ zGOM3t2ao^x)4Dir~%uFGN*lpt3as1<=5tvR=%rS~IP?6=_eZpnFTF=X*Lw@!9 zm~h+6-|eMF?XwPf(i^ZDoDN%jUvgSC(rMGNEk?_b$yA-8yft@FR>wI>&Ug+vLU)z#0WE46a$oasi5KDe3TsFUJ#!A&*N{d z_#iH*SlP};l;nEFarTEVbKI&tO5#MPJ!C>mIvY!^L<`mL@iU)zK;2OAfODQt^6Lq}o7pYwfL)w+NI(2^Mvv1v zvLxa4qrz;`HA&ky;{XGTAsv5H^I#Y)B{=tL!953q9Vz_czdj;SFU_n38#Nr`&_gU%i7ZiWYH5Ib(((Kg-s6Fy`pv-#&#|cTzb(% z?&xQtavacsv^f=pPClWE0bmwE)h1wkey6OUI`!R~^M5BXm=57x;jc-89WQ~4jBHvu z#lWsi3Kf|~f2=?*^V;c{jh5+G?0%l?1;)R1s`|ia)tXl!S2bA<>CL2{`9& zN#|W!SgR{wLgFKRxn(ss&Li7>Q&?%J8t+OB!~8!?2aXHUmrL!D!KLfuwu`=q<&a;p z{FXQyZ*TlZu>nr9zrEE$;_cY0g(+o(FO^0y+bOb@ibMu#RD@Gk3vbYC<=f=NveTrS z#)=v!-V2I#Q5UBbf*U0@NzPBYM&*_ zuw&1}j0dhKXhNO!LWseqiO}9*>3W`zdX;Jwuwtmuf;GDx5iiYjDdya8zqAlIP;3sH z@qxzn+sEcX^ZvK)FMZEoLe9;(G=*Gz!M1B28kv<_6mydzHy}TapESR4Q5V_kw_9B1 z(J^nxCsh$0v4zRwWYb!H#e#M|o=cIOP@bVz^ZJA1fGN8nY#Z?5fWr&zQ>j0sQ!T1Y z((KqV?}F;FYa4w|g~y#B&YU7SF4*qY1^jZLAHVW39E%qtjH&`&naTzY9RvIvzQNJ=OQl6kqnQ?9`{4Cl{0p^)N!zC2Zue!yDq$Y?w~_1)68_wy*aJ`V!6Mr~qD6tG1U(N_t1;BGUV}*?=ueLU~n;Hx=?`1k?Ig~Y}^KJ*WGI`ug z>ik=2q9*?SfJuAeU;d>nd-h+ppR6NR&)$CL?@Nb28Zcchn^~1sh2EEDDf@%32kS}r z)-5PbWV)xx&@^`2GJkZD0(8kFiC3GmLcTJw-b&g1tc#CLezs& z(TmQx_2{ct#iw4|O_q@xklGLzl&)MJSmB%@iWhw@tqM98ewTkoJgK&g4Q3|=tcEXP zhuQMzf4TUg6CV!?j4bVzu9ad<-%^aA;GS}-qRy+$X;NXqhOcHaVZp9UYw8>~^>kAa zfj^YNUcpgEtz>d>!h+do zQgQOUg1&~fuZB%Tx6+OyvgeKP@>I9-Abrm{i5u??5iSUU>{MpV%H?!`f(s= zLa&6TD&jbGUPBI;HQmR3NFaUIGiqKsCr(rqR1G`Ht3tML2DrI&iX>G58dCh>9&sI% zs<<_$g9AGN$Zd^@^{|~R+3uFk(O*Os1Kb{)2r$`9Q$`M!B_AdLOyQsZ*Yg#_GS}x< zRZGa0mnQ9CzY&ynQA{>Pc2bd9z^Q)Y&6+o27Hbx17oB)Z^TzGKSW&q=H{gnR3zOlWyq%@UGX)Un^}6!nrKTy8}qnG3;M9bKw6^s)4re|Mj1R2B@svcI{^*){Y}0 zptU_>KX3=dq)=ot6?vN5p;*SOalQWf9iSsd#u1FRcgR;rQ(aJ7K3{~qBS;0>CA~MH z7oN>K$=W!u>3jlhS^2P2S2-@dYHCW%yo0&)G66b`*jNy z^ZMF-45s8adyd4DOglCubw;M7lwyFqb2k;aPPKpfHNWiewfy6+k34QsRSO@Io2qjv zkdu?^ujl&=w`>%_9v~<$XjB8-GI1}RwJ?^cf;5^crwsB~Tnx_2fDO#fMBr>oWV3=Z z6@KUk9j1zg5a@RLedN!675VEwy#HTIelK1{G0Q0uH92vq=e)>}WHoXEda6;_;Ggl& z%YWV_Z4~8jZ;_!l@BZX-!3FxTGIs%<-$h>*?smN<-oKzgu*0=e*(Ggp-RrvF^-otp z&v5m$3ObHc%WtGtPAmV;okfR!x@ub6j9YWtX6ytq#{%&ozi7^R$=|B!8*(e!Mk_(b zfTHI-F_xFuL8nOZ)gQm;q&Rwla@fFurAVp*mnO;bHZk&z<5Y=Dlzf4f#>!*;vIQ%T z<9LHo8&8$GaT_N}SdKaS=B!XMCCGK^H>Me;C6)HKY7#wKx{4jIo`FSrgx8x$F>w@G zPeqo^`6Lik$dct}oGSx>Pft&LZ$&F%>Bv}y6H9xuUOV$Rnb}tSXd-#}_UnWZGkN*@ zyebwX@RH@pbIWJ012;TNnFmgKviwf)4#$JCyHJjd(g+%66|Wf1sypderVMI>@MD$d zli(qTPNxs?ACwErR-(4TlU_Vky^X@6a4Fe5ELT9HjE z*u_-rcaM`?FfuU5DF)n}!&KyF^3AVfUeXcnrWr@LN62+Dz-^LN&_%NJz=H4shf;de z3_NO4p-f9JeP7lm-!$V3>2=;YRh3g!knz=3s^M$MMO{JTo!7*^lV$nUL~$}&PmTmYN5cbk=`G^+w~+Tj?*RG zRe@6#PH~469SS5AY?W*bUqiJhVwryD%Fx|VEu9@oy8h32Vs4>75{Z zhE*@SL)SQs;dGA+c%xXB@tk@dF3kf^v!Oiap9~=TY4HaGq}h%`HlG^->N>?-p-Ag+ z)f29i6G<)#+1x53+l^P~h5gWcOVku~sxsEp^Y*8dQ7W?jz0bLV#NEl#Fqgxaj zaK!t`Coaj;^EmyWvCz&>6ZK0Firw_mcDh1*TaI;BxdF(rUgF+Ek6l{u_rlZju!Psl z<7pH_Hl^&>KRNlT0b>p#WjV=y!3^19Bixly4A4F8p(0VPFkeu|$))Q!aY1`s*HGvc z;~sdEtW=1JnXQs19@*aN7O*(E^oj^|f*{)kd64m6ymo7Dl%VF74*AEhli8`PnYG4A zT@Aq?tYd0Xtg&utg*CA>7CzWqH;<+BTdEmT4WMVUB<$F&HlqvlfM5;CHjstgKI14E z5L^~q4bSo!gv6^p*=dh*NwIQyK(VMpwp?&Tkmj1ZAe~dlL0+(Ad4b>xr_AGovPr1* z#6V*YT@QrGHgc&Z4j48tu)M8#z&z;ktTYv6c%cAOYh-9jCzfTJvK>3%W|G%) z=^n=$oLbqr1s7`}h>j4#KZ_LKp7IHo4PH?GG=a)d{4D9Z!3;c>gK?NvqQUEhPgJM-MjM`az zTWP6j+RRIW5Hm?OxE1x8-)SW(G^jJd_XPoNP|D5;srOfB!ejU4un2_T_YQt4%vG*vhF z7ec6FxTFI~P;L95pZ-1BKk=ux+VsHu6CWkIWxsFXaY^JQo?!+G*!taLWEHy^$c`5_ zATBo|Az~B7Y@|p6P%Lt*6gbyZh_s#IB&QZ)6ikya|&S@p?{bWdV2UwwQ^SXtng$Dxf z(#r_u-(mQ^S=jG%UYJAo%j$TrPatTdS9{!@o*NeJq*2|So+4NoqNN+@TFD>~ZZyf- zLoNm0o!2EDaMP&v$m%5R{FT#6=>dKw_hIa^Rns)8B4-eYpwGJB<-~@t#HHo9GImbbe`1Hn(!aFLdeN&L>-}Q5MM88?V3C7| zAY(dng;ZVdS4|)GSmn)fsMtNgXIiru>1Xb>^9lK40l+i7jlMO z=B#pVl|%{jY2F}UO1Apn^T0HRWcgKLo!35bCx|mqba7FVm_rvClv{7ECJ~u7#70?YC9k4Y|*`LQ8a<{ zTpVY++iFOg)6sF(@v>pNxSPh^;B2ro8>hE9Jqf-ad`H^Ey{NN2kO+ZF%KZeJ0qf6# zo80BTYD`Jce??j#6xD8>oeZ^X$@0pXcY;sIYGyUTdZ|NRMr!$L+<&a)^?{FnYc4Fb zI^@Ytxso_eiz0`vft~52Ua+ddee(2>JV7BxD~{nF3r|&G&N2Q6xfE5NJ&s-C^k;S; z+2f7DfV@;2F!XI{9^8?6CTsR1w1k7XE8TN?=$rGcZ!RrF*jQf11abcC?HRl7ef>pG`A(aC$3x@1 z(|h<&q#nHEamlAedQFrz8-3Y2zn$J1Rfl{~)=roCG}6~ZJ11q$Jg!lHc95)ohLv5R zlHYEgY06IdtYd1&@i#M!mHQM$!W>8a|65eof7t!z_Gv@^dTXghS5A#(BO9m3O%fZP zeu!D&#pz?K{M=Nyj7@He z9UB-kGG{2O8y8dy>dFPY7(uZ+gzDw{6#WEAH*a#!(ZFa3;k$gV^}6e){MNV*TC^KAh@^&M(0OKhph9YaJNK8;0^htCZeDyL6I4;Ql z4vQQZCQHWMdZTH1Pv$M0({B<|O@&g~7!CVfz?;*1MTPZ5H;t^BuZxlH>67NTUs)B@h~QZq%I|tw-ye zt1M~laWpJ`8~>s9uVe+g#jhP3m7PXLC5d7ZDH2CTmdY>}X#8^RY!L>3oRLOiYBCf5R4VzRv0 zeV><>os6H&P&F#nEQi`x*UqUjP1m>EAikLd{Kw8wf@&BJEH6eolH~odu*or3a5`+z zp+4}YbIjazucl82$+FJS9e&;Ni;~GuFSE$g9cAz=AG&e2H1Fw-vX-CfS$(9DF<2ev zM-MKM&Frj>9XmB;Mph@6Vqo_;gNj`5(xAL4ujg%Wc`SzLP)x-2;1bY2OMrq>b+zA4 z<}`fB7adZ@GWv?;a|>=d8IqySEhq}b2?@C1ZKsbnVM{Vj+>CYGzZKp4V2J@Ki)tc5 z$aQu|vE$&%kP%YuQp_ES+@>NQM|97)CcZ204}RivMRGuVG_;ke^+M`-nKo`@Gg0+xmbxn+|hI{#KBi${p zps$L5H{74@Od6g^RkVRPXgBOs-H~1=aX~kN`+}eSeq%3c!k8Py(2ety5e(mGUi;&( z3}dc>-%0B1I6>;F5sI2A<`hLvQju5YTo&FJ?xJ_nsgC({jv$TJhL(gJ2YO(QDoSu` zZac{kJ(NcalDVhIey5wkhd5aGu2Eg_&kR|{+>>;P8-dZIhpweF!!)Ylwx5S3dF!vH zD!_mNCYyv8h28XCpLB7RQ)PHMWccc@Bg1Ab?<6lSD8}s$^wI#iyvxV6dYcrK7>&PW zA6U(J)F1fkj+tsu*sVuoMsMPj^09cU5>n;plT4Wpe$o{Sm4;XhI8~){z1F z6v$CGg7rchhD2#Kb2ESGhdWT$GG-O&d?`jmwM@cOEd?Wvy5l65g4)*oa3CV z#=LFBPcb_D#n;dL+;mgZZjI1>BLF>Bo~YwR2{J@cpdWFLY36I`z6b~f`9})|(E{}`u~yzlu-@b{P>Ep#<~DC*ka`anDWS5IlM`MZ z7eFi;;)yk41(2ySt@wMUS=}$m29}w8?^ekdvNKR0kT1gEAw=-vjEiPQt@Me}+OLnBf@Xg2K#J5jwtUqV$k9VLN}BAWyusS(-~9^t~Cp zmlTCH%GS!&=vAbN24%UNKY?2TN@{S&aBEn@;9;Lk<9}Eg9RVLRxrJfKqpfS7s<_}% z4O+4Kj~``%7DYF`J*1Aa6#2Z{2!3%By@pytCCl~ShRjkcLN5Ds(`%fTLaX9M>s>wx z?56u*xddC6i8HD~@gUv*j5{{1<|T$bq6b`XIUdJ>O!?{{-NLGxUjre+7+1sZXn}g# z4=lvov(t@@&#-d9r@Sux&R>6M2={*P6Y3pub{e^B^uFJqm}?YiqassD2NanPI_wrV z^N*4m|I1#tysiV`d8O+Y!CJ}I*&4+sib02BU|3lPLLX=7!kKm6rQB3SjsIrR5iW4g z0?$f1a}qe=kpH-y-^s__yiR2wEVgPTSH<1ZGp<0=FDUeRAiN2xXf^(81VzFYRhPTg zqm(|Ux=gBr(&!rCw@nPo7Cr(#1ziZzzm1$$+1KIE2j1(KaESVGov*ivdvRrbP zw>BV~z8FyD6z_e){bGPdb)D4l$|YO4C%kKUAOfJ;;FzOa9k4e5KUFy;^Ll0;h{$la zy_Ymju#XAWi8lC4Sothd)bnIapg$TibiAn&0~m<{Xhbk1#$b@Oq*u*x%8zKmBSJ1V>Bywz43 z0HeRtY=AU+54Jfg*&X$VX$EKv3q7V8J=3S!iCX$F_$DX;uf}xBI!-lgpLWu=DtLU( z-`B#2`J5hXbBWbkX}|GK_wz@C{@Z|wHEZJ*kZyM7!j8)l)=wR|Fn>ZZkm2xM>4RFy(U2ZF1x0zrv;>70jT54T*9 zsyINe2Dcg0_EHtcWYKQw&fsS0S}CygLPx2JUibAh+V0wG8`zbLt9dfAUm$>W1 zmcrs7Bce2{j*}j;!#gwNpmKL;@^sxgu~~{G6DW3C;=Vhy&MTL0=i}uw^fI2hKw!7) zv2FJy(NY8cPJH8=SI7oCPTd0u^btjsJ1GXN=XNSGodZPPoCkByd8|`4@wwWI@b>QChi)4Eq?&k&rN)-ugyK0@rHeEV=+he%F23DalU=DDRU-7 z-A9Z?GR175NIWFth8MeMbFYe9R9WAygNzasd5QDymDczcdKQvNc;X}fmKGjw~@8DN->uyauJdn1AFQ8kT}jA-)dnugjVxrxB8SY zz*#4|%j+QO8j>LTB6t^_7^Xi}s|$R7tVnVDL{SL-@d`#Gt8%*I(G;Xkhj|_8ewrN7 zFSsf{>XbDH4Z-(xNt+x~m8pszI)*#sfZ8B?xY%_Xts(e1 z&j+RK!6(3=Yl_<`l8f=!WrDheaU3i*8QY3kK9E)ow~h$|Rs(6juH}0Tx4s=@K-k*k z-o0eE9Xl;2ji6seF$XD94n^DGVb=J^d!G{T4N)Ubu~2uwSug#W?Oie-GJ2w>t7`#z z8qP-7VyOK;8@>lNtcM(`g9cnuIEY!T1j=-Q67`t40esKc5J=2x6YBHy@N}E-An>NH zkTwIjErIQr__2IKWrORnzjn+P?0`M3pyE5T4ER~`+Ziz=&5rR?WrUwS6ayJDc_?R- zEU)*=_qyV)lPhVKUI|Zdtn}9tsj6}Gw>haxg;=^$;R6M@dw9+Ke9;z?MmPGZkzh$j z8_~h3hZL~^m$RX#eX2k}cnvkSaT{|mj2c+Y6dLt9@ek|}v+3%-TGP-C8z0Vo0}#Ya z1PW!3Zn)o@2N5W=Oh_CM>sAtIyMjuyLux64%ywX-3;Aa+rJHVuza*G2vwf}wfvkrz zD8NPW@(%d{XN|1i`DVavUN%t8pzPjG1}pQI2>>+uu+2^eiHI?6HHFjf+{DjOo(F<& zn5eRqP$m*kGpj8KqAwlp%Y5-FX0U2$oEek6%SaEnZUug@&0Rion~wY&z8Iyi@r&|B ziD4QBb&5-ch*F@F%TnUO1L2T^-nsz`juz_ZvXmY2$|Y4!d5{1S2b1VM&r&b-&FRo> z&3t|3@PpU1o~@FDp(AHVOEJf0nwvn|rwG>k`cLmD4Kvq?$iHqRC3ZY>0e!@X5J5e~ z)KH|7iqvm4-4-7d#yIwIOE_6{kNYvYnclPDs<@utPBxGMZax1PT>|w<*Tl7wS|y&_ zCm!S-5a)2~ICY$x^HDjrm%jRytK#K>*+i9`%w`){JRRL}AF!ItsJs2Ir<^vo zk#GFCpFPVPErbxuUMJps`$4(DwAAfc#yvYub=YrYZJw6DwQvq{N>%lNTY-S_hI<0% zGwvpCnqRDoMzv2|tE;C>qo0VANS#+-#4ad$ieb9vb}I9rx;s_zkdzAdJ3X{**wR8k zj5>QveAvzRVBV&bp8TH!FFND==fVeJx%6Xa15a8meQDv}!x~t)LtuWCY<9)`@Q=bz zGiB4vwEf`*zhrZ+ES>DPW51-u$Wk1on8OsQr6RZbHvt1Y`k?u+$*9wwL?;%P$hmZG z07hnV>FdJgAPCAd@sWTE?@)KhAJL_BRZz0LIw(b&BE@2C^e!QOjJL+PVyjI~P10yV z7F1H7C%x`N4tOOlsBB?XP>ujOjXRY!vnCj$nTQ=3)wmd)96!$30qMJ8Pe!`1!UD7Mv=eV4gE=m{|IS89^95Yzl@v1q`#j}N9NS>7w% z=$fUh4tfZu)HmN)_9NK6=#bxmd~{oFX^k5zPp!;2fMCHbKfE$M+}VJYinrfAL)Npy ziX8{;@{F*ujbgS?Fba`+=CdSk^)><4B7Ek(N0CjhgiLcN5XOmZyW@7(Zu-jnHG;`= zcvz#uf|0dGlmGqE6vkAd`X!ld%qVSliA(7YdUZ&OFAB`&avwt0Z$H6YnL3XC>~_U9 zxblFu@v>*+@Uex*$a`&h%s-I$bJLXXXV=7bY$D90f!F$XzJ~d1)pRju-F#Hl0Omx9 zhl1nUAUoq}2>I`c_!PF%kx?LxE)B>3%ACraZayQrXyp-L3x=l?h>epX%=@tm_qM$^bKYTe{qrKH2@?WI2)q=~FWi8OXsoc@u>JOO|86%HSKLJM$PAZBZN_ee~dW z4ZcS@HSi5`%8nPspBZ@*S16{HA{r_(1$Z*k=)$lHAO=eDt>hPm0fWdHVU}{cTkM>A zzsh+@;EEzgP$9^-7E1GAVfUO9V0Rk)*Hq;R*<}ut|H& zfuJ6@qcK~&G%O$Lc9W4t8vjA=`c*3R4L78tf8ZPYsu~+2Ykj_;$MLYo8Rz{TZDMs7 zWX<3v9 zLBcD;oQU`YF@Q1Pl1Agk8!jBw*U1sGArP2~C2CCY$9QSNE6-mpSmiG<-e+r-FMm!a ztlnJv4R$?DDXfthOiJ3!U;c(9kLKFi@9!t2MusDsVnFUKjfyOB*V4H1(M4!*y))>5 zYn;-s+phT#!bh8OYwl2JkvHT(_Q@JKQG(;Mha7MTjCb0|PCSLvctH&Y<7pqZI4}?D zchBBDV7dgd+i0Dc<&JJL7>?56hM$fp0OKrC^R)bnZo5PGdteE`6J8uo-4D!0M}Wh* zn?55vA=62AZIRp#Mn*LBT{;b?AtiU8mqw9GZ*pm(HRie6*^kx)FfwvpF&wl1esg!2 z$H*uS_mp)7oPT)DU|%W%{VLqcxU%%u6N=2iOE%qkP%^?q_MOe;H0 zTA{^GiAGf{EAwfjKmOV2@2_%R@mjXWfV^X#^;U4!MUVBY4lDFf;pYP_Ne4=2Yc`eBfBH)baMG?%lqb;ttMH?h$UDVjD#DU=vDdJ=~ak z>n%*VGqpVCyR+JrGl!PAnC2>&+2k*v*M`>+Eqz~%eBL><8g0M6?xDmtfv9^NONGU- zpuif5$#|sr3|POYR(o*I?}*ilYQOK9`?Rlb`-Nft`6zzPYou#5|JHtgKUrnO2h&e6 z_bJj#MGizHahs(BF4aM?u*nQiM*pJ@_Jk6{njlhJ$&ZJyJ+70nG`pJ4C*2@Dk>a-3 zFV|^U9}u)e8~h9Dd%{DoC3GBEC*r+Vz;@6X=L!Xs1*Op+!*aS=ie@KSp2^k9aOdbQ z-3@Fn5X#AstB)yw9v^Z#QxyYr*1|fty4nfbp%5wVF%`M>&Q4XVae~$ZE+;@9ccp)W zUn#J>AP;9Lc;>LP;9U~1O>xH?ODOSYR{TK!{rF4n%ep}zb=P9Ui^%#tF*QY2|Ezz1 z%)ck%8y`6Y8lczp?b{_JZyISd@-QkW2ExImR3z33c82EDouT?Ixgm#Sd6{g5G*!|h z#n>4>`bz(OUKt>rmrLUUrdf(gudNbf0K!^){ak@=s=lTSxLtJZS|$DD6W~6ztOEb} zCSl}Ns}9G=jwg6*emF++PVIKbX6TileqC(9j_hx5wUBsrkv=#L3v-ItvbU-h{+w^uqO)nA&l!h`VueS#Ra==L`i zUPl}a2x+P{(&FAYKbGKpdmDH4au>`BTy!DX@@s5eSdRtt~Gs)ZF0*xkon z6ERud+S8%26c$f!n52W#XU_fcw+2|GKWv;v?z}Xxh%xen1}WwdMf#~or2eh|)vZBU zVmNTdwM8^@KA({m2mwa}bKomMI^>Wt6BzV)CQ~NI|_HN}Sh0 z(LlU+r|c0FTd$R;!`@XN%)yD_%b>J2Tigq02jq3YWz!avz-tn=yoMeN7Q3mnp8cSz z92c~Q+rUYHWZ^@8yGgQBt^^1qJ_dffmGCeAm~3TXSX5vGxMWC(hu7jn>71y*Y7oD@ z18!B5^yYlL^>dHUJuc9lN~i|MuW5(OvRb+wxGfWS5N^`14DFU0?`KDP(oM_U*dzzr zu`3RhCnM}q0>#8qWE~Y*AizzhM*0c|3Y`eblv zn{oXC(Xg{g_Pc9O9vDHQi(<4CX{REWIo|U;>h;(o*C{>Zw1-CZ=Ns}WC|26-v`+?w z1b$xxcaWC(%YcR`+oNONXR^Ks(8Ugl@z9@LJMEf0owL{VoZGr-O_C$xPL=+AhI22C z`edJneGv?yNA5*o68F|zY=;&9n?=|*np5vr3yFvNFYi<#g`>uBKWZiCfx>81VCB3f z;Z6u&;&qAEh1sNeapGIs&?pQ=3M^Sn9L5~x{y!8P3BbKvf&4+PSB5B$TRF3j`@FyUbH082mn?mr7-N|Y>Wxwy~k9wx3 z^9>eg@rBKY$vQjUh1g|eQ&K2qGewf9$PT(?ei^5jKJHWMev&kZG^z&fN~YiW5n1Jo z(uH_#hw{{<#DWcAvJ|!Y7k7LJ9M4_sQ0q9qw9C|HJev<}$17?x1`cyIH-+634$4q{ z*1(IN=~B$O;eKi1WLr3kp)#KFp1?ikp%UbB;DRZ6ncePqo1r7al7gtfd~uQ6rG<5V zJH0`ZWb(^WbfaU@_)lOQ$9t>!@gF5#I5qsM&fod5VIKMMq3!RGOQY>3*s-bTH=0{M zrI=2NbWoA`0*n^NIAIidoA-W?R8E;+3^2{dafTG>AsAEED2m;YoFNaEx1gehA8!O} z6fsbYf~WG_b*Dd3q&wCDWsLsRr_R07BRm~9Fs4lu$V2$`T#1!%yGffcUb1w-oyr&A3?vz+n&5R)3CyQu6>@ez{|vDcJ1R0GR<+l6R`I zf?fdxn`*3x+Tji-6c{&2jVK9J_t*{MIS}c{^J(#@ z@<>%+zG15*n}$13YhaDwBtg=;b*fzYbK!cq?n0IlYWB00*#>y&&S84bI!G8*=L-g$ zu`4{WV-`eZsyuQ3sg6@TGlvG|5m<=TfUXm?+c8VmqFD-3W}*AUI+DoWgb|gS+?MLdCN^p71)Gr?AP}vu*EhaZ1a@js=@z?PqP&D4Ir-P zoaeL3c>~Z}wbQ%k)zj9xo}vHL1tLU8Nk8e9Uni%@T^=?_Rg_EOAOksLRzCNH(yHYx z#-j(HCIqZ!db8e%w-3Dgs;R%s1}b(O;WOh&_sP(CLiVAD1hPfTxls{7eqRuBMD}-x z^;r%J>kYXjJ?+ZLZwxm0Mj75&n@EuzudzQivI*4`1N8?7sK}l-unH^5bwxzx?0R0A zD3yDT8FcvabDOv@tXgP%u1T25O?Un@7-{iFwq5I*DaN+f#2MV*_L2nu-+kH7CI3t@ zUOg5*nyX^T(>7{jIV3E1Y>d8OoF7Y>E%29SXx)3~}t3VexBbD2$dfkcaSz+)L6ET3nfCA)z))0&k< znWD}3r=f|E;n}f8F++w|u_&I~3E2=DAlgj##3l7@R|9*|X-IkA;hyS^)w@aF`GV85 zI)R7jg{5I>^i}bIGpxoj1F=A%7V_?fi{-3kC&qsCtjBxoF0Imnl_i-v(6eT~hrw%o z@AQ;&B*l)sR;Xth5emttm|TivQIRX0Z$TD!8hy;KJVNt2mhR_pZxMATU9IEjMNZ69 z1Sg$dcaRLrTDaS_n!c!9dLD9%)kycTBBG9ls9Mb{xpb1ZZJZR=U>RiwvI;s(I$;Ht zDOLZ;U2a;CZMUYxOr>}n=RCrq|E2o%u-bjchx+NDwDg{?~E{0 z+zav)P5gSOD){RSc`@X^*lgN(Iw0eO$kY36b-)TD)U_Ems25$O4OM`36|>0EpoL^z zBplo5s@7MAwDU8Rhm^JSX=M{c^KtgUX#T<1)QO7+oTDPtH{7=dY!q!~62toB2VX;% zR#&G|D!dGewb&Hs(T0n*Iz7NIZWE&1@Pq+2PW+5wXpB!y6j!Y9Gc`suG{uxi<|SDx zn-MZ@m3&EE0orYsVHE@jLA|3QZ(uoVfg*xI8SeHpa@rN@LDvr0OoHm9(6v%4fadvw zJAOz#zu$>o{RQujOMh>0T11k63nx8x?6kxg<$Xs@Wu8!EfQnqZ7-O|qfVP2j%zGGK z&qw~ldts%1#j-@6Ms<>CLz|>ee1=;*_AiCtCs#*V49NmY}S<>Z!4Nf(|bVjU`Kk>jE&RlIGCgRh@2K4;U0o|%8CWG^g4oW zQ4_y+68xd3MuMFcdU~7956n+&b05C)UDGA^3k41{%kYndg`92P74%ku*5xS4=58g| zg&p%s{hFj2RT`bo&G}h12rFF}HHF1C1Huj+)(O^G@;5P$4(%%!-koWfy}niO)Bhsd zrxB0`935>1 z&5yt+86@^=sjt4`_;pj^zGsbz9Y=J{$n#-7LX&Kx@7?g#9(y8MnGUa`UKpVLaQGiv z;q_Tuu|)8*9oXoMc?AF0s;`rY$xu!`!OyR7I9DPn-+H5;nvXOAwiu!T-nJ zx4<=(o_Tx3b4Xqcxe-jxfTBbg5Jw1v4(P<0PH)@o&i1z3{kGk&Ut8(6ww-RbO>Jk| zU1Yq=RmBTvK;-kO=Ca2#Clie$SJHkwk(yA)(RT_BW7odxLrZ zdEV!JF8{xa5Cp?|Bu?R+4Q#`_j7RTITtqsV0{?hZIpu-%#a-)mYuh7m}@&n zbk1$IJ(i(-?xI}pUs&_wxotkl2IE~iuN<~scf8L5@7)fcOdstfwo8$x%BQ<{Nx;LF zqb?Kz5t2_fa7q>XT?Mjm)Bi^SE*vm2=)3P691J%KCyb1FcwRkt)_kM9sldibHbkYM?E9p*t zQ8d=wRM2ROT>1^*f_Q85;!4~%&R=@M*K9?;SGC~+S?R{BQ?Z44+f1<#ncG0cbwM~e zU2%!o!0eVZz-ppXHYql#TcVx?=`~ocYr3LS<}*Dr@NBvw<9$5$Y*>mF=X~l8&bxEo ze3(%FdsIl&bt5oB8P=^gy5BBaYa zA8vrGKb}29s%8Pf7o2(Kagp?To1V}qKMU!PYNl^Mm1^1uIyva!-~?HJM~%nBeBPWdahf3aU5aMpEX=jeBytJfx$nP>*vK(90`2d^h+a|; zoN`MQnaWxkH4RK4N5MBphlMOw<*A&E<6!~D5#+;;<;;tFlmGj+b!y0M_s=%cK;6;h z{H3x^Fk9&gpmfIFbDp znngoxeFcqjss}>KXF-#2n{2--S$ybiEizwcGtfYR?G;EKxs9}tLa%vPh40EKvU4(q zTu5A*<>xdR?wghM`t0uGcg!xyk8U?)lKpP%xauu@r4tkjJml3>+zxpvA6cS$$z|zc z;a!DM523?zq3=lo+!ul_*qj{qtmGYMj_@(?X40W(7Q_ozj&BuUQzEdn!!{mQm79Lr zz218H0(N^O%2#NM=-ojbiqAsYNF@-+ucX?@-8XjfGNdN{a&3vcS+Gyp1tc8TB*%l$ z4|DLVFd~!>jeV=Lju|FSC?At_x91Q4Q;T-nW#v5WL3Jltz!U}@q3=oh`6+=%NwM&* zuwAkI&E4V)fjfmeh4EqKEK+|vh!YO8=pdvxyn|!6{zzPY z#*iW?nLQAa8JJH!oPA$-fj&uY&v5+&5Bm7I%=MthT>g@i`JjHBT9iD|ycQW|MO-JD zuT2tqpT)|xm13b}A(x8lkMokhMA9hZohnU-$ zw{C`HsygJRNZWS#nm2(FR{{M?$xILQySfZ-T(D)e<`@rM1+w?n{xO^}qvf~9dybJ$ zh9a?Z9B~TL?>4T+@>{%mmpz_rr|gic7Zb zN=vtY>8sv%c1Lj(V(^uM8#WJ_)-92}Qjo2O@{7Was>kl2ra3`6@L$FD!qfCKP+t z`n-?H5Bb$Ptwh^ytPEu&hC1gY;Mm5>P^AdzZ_GSl&C%kvp%fdeD~Cf2idxY81U^e) zwyIHyonnPTmn{&(%#&T180L9bUE+j^v0wkj2VZ{AywW`3ZTu0b7;2?)<80p*iQ2CEaS+n% z0rsHVNA2Zry8ok7V@8BQH7SD}7|I=T<2CA{#k`-S*kcrwVZ1#`hPUFcl{egNd4VR8_}}I)7O#Af6*z=Nb15lv*H9n0I|t1#Q1HEE$2$!a%FS3)J%pl0qoirq|+Y$^_|ONAhb)go72iy|o|b0*MX zUI|_EW|QcqOc$7>g2OF7=|*QYnaP--&8N|^HC4bYq458%1MmoWIqXLy*E=~OcuYp$ z8p%lX8{OD#v7z6H;#pTk?SLE5Nf)+EY*qiS-(rEWyFZxGpfuYK&B-4hBpcn>b{w*> z9mNz2G!g+1mH8SfKiqjRSDYAdkZgKi} zBkvfBlO(-yI;_2@SwGl&o^mfkrVEcft zhZ@SM=Q!C7s&oCgf2YiA%~PMJXGp)t5Pr5B?~&JAtUt*VyPN_>4M<1%*43GZ!-#AmCp8UJK|0=-KTetx4+>ncFc%pi?^C2E80j*4Ad|W4mEu z=Z$S;;&rcjS^oj`wn#)k0@Cq=ZxL?d&~>)AY0oEW=QP_hlvD4qZ8U~hGRFS+$)qEJ z=9Q)B*$?)RY&YJ;R9LJcJ1KS>MGC37zur_dL&@Y(vI~Cdl%IR2gk!ZzEm(|vsZLcU zZT>p=;=QcwemAZ0Ue&uA6oLG!{cdACc% z&h3QE(B@#^tW&{a8jKTUs7c2ozCO(inU6khIY_o}OEunQDZT9p69twJ1U~SH&n@)AY4M`t5Z=c-owyJ=1!x0+bD}dmuFx;)XJ+VZ^NM7dwlirk^%FsoDURWHbe3V|$oL6~*{V-P^&*;irXo33bn|Cz^}ZE}3|S#a6J@I)2(v8`D*+CuQ~8DBMBnC! zRPyQM6+XqlraWh>q$c_R6x~(w=V+IFE%9A#g0NGtZE`#P%mZ~Q^1W90Jdq$hS)uTm zN1-@Zjc0KbQ$ngu)}lDXpxo<*2?vz;6cp2tjow`fBm{q+wlXEF1~0D81=be`bM(a* zw$Qc(%lN9tg;(9{!E%u%g`O#9unMo^AGPT7qVwxSlkvqzIs6Md0Wka^F1TtN07H2_ zkCx?DSr>cTsGr{X_I43|wM!RY4taUxLG`M@PP#Qxr`|rhg1)JE zpulD_opQhArfk$eWgnED8w;1v6wh2T&Dy7&`*rm{B^RxNAq!&1|(}5eX8x{{N zg*0Z5O@u<}3XIq$Gbw?3<)JCs%^|hoXUceAHx3?>NTh1gq^%-}c}(>VkM z*Ck|`Mr-;7tB?0NdLPh*p;!TAqhX6Bssc`l)jo9~yf^5?&KP_QhtKw!uZkD8i<$){ z2AEvBLWHGWZA_m}MfhMgBd%C_jw%*UWs?bBv{u1kVn6oxDP{KQdcl zI!b0bpmrY5={0xxM}04l0>6HbI?_QeQDp^RVba1!!{|7Kks<7Zvrlot2=(2){lEC1 zRhwY??%(xYttt%x%VEkV{jc*g2!`KSyJh;{n^&xl3-;e8$A*g6xv|r7%VOW7r`U58 zX`td(c_;F9K088nfd%gjevL_kr%`$Dq`s3Eg4d2IcZQhmDhvV#hgS*Fy#nZ5siphq zhJb2G&BTTNsqg4DeY{f8_FO^Mg?9UIlWg-%i>`tuS$x$ZsPcd6p8*`ZSD0IGH<5Mf zmG91*Pv!CYczVsJY=uv=pi5Fomx&h|Z%PS#?0=Kq?2}I~<-}CwkO^{hg&G8qm%X%Y zg7&Dt68kMPv~EXL-6zN0*cocIfYC*YJx`HOsW|j7vQF$c9f38l8v~C`9 z+kl>puBM$qmxL*SQMGi&giUI_X44xRc~z4kbI#;>KAdrKR{gA6V^{|{X6*H0%y<|f$gDv( z%rQ3r+6LLuExC`awIkft{MZn+Fo~d*$~$NStvwvFRnjcDBDx}KSDp{g=I5ynhh#H} zUYz?-xqZSGcXNY*#iK4gunmJDJle71@C`eH%)6!E?x;>7o85S~RAsSSDxp}AsNaUk zM(lonUR?qJ!M4O|vfQVdV4=#uE(#kNjitR9FNP|XNi7Q0I?^V3B{S`!Y!E>@^fvB* zAYy}V@m0|hVC&#$9bPiG+)y)&ZLoWX^O8Tve3=_+_AdYT0P8Aow>2n-EKu`BiyBtW zZSXt7Gtf^Xo_Qn)v!ho+{kIky+8gL~P%#Din~;HC?(@t8@WD@JvQ?|6V3k!q-9-Al z4Umm!0n})_6iIBK&njM{vQddZb1q(RdYptA`+&gVE>4&kb7(`5#+uEAOJ{)_?~rVe z(p+Re3)2Is^L_nXaM~Imof8YnUEU&Q$1T5_d{46nJ#< zQ6%q~w$WSf%4VjKL5l-cINtjE|2In|HZMyX1*@A#@;Fjt@jcd2Y$iq4QgMy{i2Kha zY}w2+Bxly+=`HdW=Bk%Ync{Vl=rt&VxQO4vJd7^oJ@Ky!*%avf@8^*JHtwJ7ki$6r zB&RQ}`+nzo_xD|qngJquAo`G`jw8@|Jg62an__`DCxeRHrMT+Nt{&$VWt*Ksf zm{jo>svjr{gQ}RtD!m4H(j#wqX9VAXjcpeF$WeN9ZnMQ5B2HfPZTz{f2}C}aaANwq zW{5=1`?Q&qj{`y7!4Nr3u_q`}OU0qT*Bg)_Efix#+CUMkwoJUiZ>g+Is!e9@cq6%O zX~?B%$lsQ&dKOY1aEz`9Pgm4=Zk8knfpKS5pkAIJtXFE!hW06Th9n3}L$qnpm)Jkg z`>*z3ozisHM3K>AEKpb#2=&pvy;7`X8R$UkB~=V?%K_I0chA!iJPqq%+PY{1H?~f5 z!jt=UqCfc;-Tm=qJhlDe>R-uHZdIJ_YfPbHbMV(<9mRq(zmke8{s5QO1h(m&dO?d{ zG2Ka@QnttB@ty{q_1`VuN-vG65S?R3X+brjixb3c;n4->r#>lpY`q+Dac*C zid*Oe;jZ9RP#@Rki+4a)DaF4;1pX_?3Y?48DNnL`e=O#zfXx(8ehAj9Mng@)!QGMz zAYnb^gRSEi^UdaC+Pd!?Ba7TP(vo9g5LQ#{N{XZ+txFwb2X#=Ghp!5Nb`G2DA#KyQ14yy&?e_Fir>TB&T6=7{PJ_=o z34`aZ@vd6;9Jy^U&PFqoiOaM_ft&WtKH8nU7R6yw8CuL7^&DN30&`;TJUeOn>>h@b zRT(?>ud+02_S4tIs@UK~EntdNl$3|>ow9dw8Sk8|A4CEh zCg(wX|AgZ9jGE9jF^?cXglhB9f*ns`4>quUhc_uM%{uFUET++Si~2Ze`Py~KvYDl@ z(rxC|ffU=7$Ss53X1Bawq1O~nS`~OHv|mvZx-a$!ZHSz&au~DrVsEe+;quyUuvy($ zCcXSMPqXt<`=!eNA`6D9*L2^#8`)&B60M@xG>Rl6PY~u6fk4cZL&{ZS9i29nzYS#d zmist7jaHwYorm^MzMTU)={DPjXUg7|D-ts-%pB6Yh200^H6bsx@>!`T$ z@K*UE(G_N)@6(88LC<>=LM?L<5osTpta-4I$P1Re`r>?3TC7KJ~jQNt`)4u&^Hu zUl;%<&G+!b*ha$;`DNIkU`Z5EtY#Rp>5tX{LxyOX~3ZIqrx5eB_x;kajh zDT+noU{@sVQ*_Df<;>L@Pjn-J> zl2h!rd$M*j&mh+>_c^210^O8SI|`Vwm|wdAWbu^i4>EuM)0xlPAmbh2?G-|b&drE-IT-l9=FV>quFSxuB$9rv?leDGp=LZjXrSGlSM{>rI z!xkU?Zi)ph@$Fb6gmEiuABF0mDt#MD)GKp?PY0; z;^^e^0QmT?h;kxPbG%TzDz+lL*lWEnc4({!LwV&^y3ni0xkTq9nu$S^f-gCK*u&m# zdnSfh`rS9AUHw5|i1mu&wgD3xYe>7|B9LgGg?$U~6;?2J6>YLZ($Oq`8~}=aSRJ@~ z^nP=~>e#WrS-SoYW=rz3+V9UM58QYolVZ_9KYuLSPm#|-CXZZ~5vfpa z<6Q_Wg`h?Y=wBr>+an7Dn*~+OePCF5K<|#p@opF00t%vLf$8phe#huEaUPwm?iAv) zaF^m4Jg3*330dNc_q29q7U?2&xe2eEOh|wVuhJ+$~*3)=$a`J2P=3biFo$ zTAxmNUQ{V>3V#>@nzGY+80=H|RIF2?Mpj95qL)FsRFMhL!aEO6%2rkK%b_-}En9yrkhuH4He z^d+Awm&@F+!DZx;-OC9ZV^zzuGNR2U=zC>bvq^~?o1l6N6Lf@PD=AV=#dRq*_&t>G zor#s7{jmJqRFqHYQq)P;h?@nKbRFo#?hm{WlHs*c93NKg|CHb0oj#=!(ga1(WwD#o z1HXWI=~#b zO9p5R$QEns{pd~sf+*@B2XXKyAu-Pm0@V__32DYl#<`>D7}dKu~Ps}np9u2)v`cFTLA z)Bx*i`xF^$o1}z3<8?QpR9s5#2x{qO(Xo&e9yZw@`X_9$ZwlXyqfGeX|(D^K;;-0;CU6La)062O;EbFXrvudq0H+ZiknN4HX`s+1o z0Bi-!;=sbGE5lo6!XDha#CtREn7mkCA*f-F0ioFi|1;i4$^6*W5avmo*%Wa)tUS^n zZzd&@2lOXo6YUgueKF-u#Mu{PI^#w6NB;l$4Ks4acK!S_Qo{{7ZtUgWv_Q@|ify3C zSt{-ro|GwCK;ntuh3LSL4W(q+<8ps{UNwRUui5Szx91uJK`hRaN z`Z-W3Yi!!8)@y3ifGM`X^KiDlyTleSKr z16jwEttW`|q~6(cur*Vie~ea9SMPx3XY^U$EHPH+I_?o(*poRY7v;?I7%Gh}_0& zk8F*omo+N$fQ46!0bk4vVDrdgeue08XuS-=zK}3?HEt1c4q?~%P<>%UPS_pePybo@ zsuy0Nbvqi847xS>(`xCI*p1%%r$V1eXtgAXJ?o#T%2b_;g=BsegY73Nfjh}FkB|Nr zAwT}lxS#y`$8mr9$#4GS*Krb>^&k{ARDS+DO3AH2vjg+0Dzt44S)?&Uw)tu61#Nz9 zes|uw^A?h97x9k2ovzTU`=d7dKxb|VSQ4NjgCtKLt*zqc(g#Bqsf>E|CL?rN+6h%4 zXnN4CK*kuSqsSqnGK4Gff>jvd32sIu$=eT@OWXXT)PYyL|Wykp8X!!-ul0& zFvE)*Dh|z9UTDqW;a#;a|Dt*T4aC73aV2XjfGB$hll*uvf4)A%gs}R z&$`DyYdn&PjRIWkH5-=8hqH~ui{o(u$e3~S4ZpFj!jlmGuYdB>xL^PHmveq^@?6|^ z4+eR^!AYJ}icO-(GAd3FeE?W(I`Cg3_#>kLiXo;Wog(CyAO45-4?q6RPk#R^$s&sV zgd*`4sCe-y*PpQs6G2|cEScFOJ@NP7x<*#GaXcTm2M6W&Hc@OAMb=Ys=w6_8xE6HJ zAFP8RGKt0RvI2JXWUvu^-tod6^0Mi$YssIMtwMf;qUdXqPoxb{BIjz9IA%8hFD}M} zZ9ojQZHblV=&U35Zo9D9h{50V+bXOVK~ZM(lAvP0+tXVV3&-`)o0%HbqH(u0b9M&m zg76TlFDqxQ3GDz*s$F5Nns!-(Z$59%c@T=&7g`)z8M4}2@7<~?3pE5BB3J2d=uXN} ztqFa=htSE?IdkSX4Nwjv$HLe=KX4YHI3b5hxi#tc)F|jt?Ad@77|x1%)}Xr1YoN#K54qc!Ij7^Ks{ep9C+mz z3b_Gl-aBoRr@O6q z3Y@D`@0ZWrPj^SB`BVg^MSl+D+Yg~QHXp<_AIe6}i{T6o$L5;zQ=GstrvBq!=lqA+ zt}I!a9822W*sd(MXj*$pv3(SIM8z2uJwBms%ah5;DMrx`lMYA==@GTajVCIFo+T z0j=I?(RGj}EDtCTIO7R2cRDpHm}41k#)LBjBEVQ#Gw;dlKJOD>jt7{WkOF^0ym0&M z3@H@e9QA}fm}|_JGqA$h5ZUnnC#;MO_=WC{FjVfo0mK-LP1O`%94a!Mn3v=|R; zhPAKX1Jy^G>Dx1$>gL!nQFd3f*fo|jZ}MvIZ5#>nr5h&)Y%pJ52EqvJRy5E>6c*)` zd&SS(7n=*5bopR>^eog$UKL$dpsg`o$W>R+X>aWcD+0yFA`017bu)K`?P1HkwC%E` zvW_S%hJvoe9E_@#v?vdRoRMsRQidYRsV$JsFhSPKkuz}gX2;+8Sxbu8(Bi0s@_DQ! z1@@0*F@7iW={isJJ&mn5s0NRpZ5EvsrbRtfq_>SX65Zly^j(Yn3WXg}S@h>lE^mJR zIgI#Iem;~_UvV?^(C0+FejoSmzlrjZKAI?|^rSP|=?U8R{jUV@_zy?t-!HV@vWu{2m=C@OHmVgul zdIQ~p2H!0DgnYhGi)3=U8Jwa%Z_pb{S2PQfNgHS?m&kjiPR|&R=)<-!-S&Qr_TBfI z*p{dU;otqtyc62y74;^$@!Ir)^jmnh9TeM6kvrIBZxB2IELYH->dXm+;^Ns|0&Qks zCs>+O%4BgLoinQn=sL1JcQM*c@22puQzTcN>$eF?dGQG>v|Kk8sY0+P+VmRxIPja0 z3;$i3RT#9?cL7s1S(`<#3aGgIsm?0<>FN?_z(00ks`eUV*76Xg;hb$o#be0&2 z(@NwUq}ooIE-(q|`kF+yVDElPiHd&7u$h}Xx6^OA43yOif);q+7BthjwCS%kGP0rk zo5i=!dEXq1uaScX&%agTO}{Vyi5YTF?roQnYuq5`#=i3t3&`E0*t-;I!_=qAcg_-* z1O11#PF^C4pO~pwt|*g1Wkri3e^PC5O5k>A5ZX!?(>;nzl@|CKeCk3D1|OPI?zJf( z57sqw;IQcT&>KlU*MoLU=%i6nv=Wr;81y@Wi6y)`*-iS955P*#-jn4|YfoU# zf)Bi_Sbmlu+>rt4vHz!&SAe$Q3%fyg0o-$_nO^N*dw%At9*Vv~S`>Xe9c#qPf{BY& z17-+k1tQR1S@a@rEFgbjR;)h_ryX;gZl}HP%Ld}KGLJb|T=Gj_vo-o&)rJdXAsKM>?xuL3r%{cUlRs!A1{m67h57b6g(9IF4 z1Un!NV7&Ly^MwZ>i=D)3t0q77&*R-DI7B{-w@}>{ru{s!oj%4}&n}9BFmEDMPV5cK zpVq`no1jgbSPs`(6jd=g<!;9%Hr&=8Px;57_QLhf3>T0LGHNb_hiPcrAi4kd z2fiL=HXm&t+}%YAxNV=@7(TTYhGiec?x9Et71t~PG6qOw-r`lz+v%(j%xW9tmGU}Z zz{Jf-pDbROEWR6dnJqBW(9}t9d6fgEuq3RDw~MJ0te&D%_RT!4tO>m;+8>ikXHG~1 zdvroxL0<{&B&(;?Ifq)tj0*=*H3(pXUOEwL->!JK>Q!gZ#E3|Xq7`_g76tB$-4xpe zv5_;LS0$T$`a`!v6tbQE%Oa>NcD?ENVqUn-_7{)2{Do~BGL*k`M>2)yVk?91@|Syk40i+J6Fz4QKy3D_n$;w_DZ{)K?q*@r>uZW} z)ktr^o)>AvD#8=}yQ2;AlVn>YJijvhnMb$W5P-FLNo*mg(5I_TiMb@Jxy|38Z>D1} z?Vfi|^XI-h)IaaZsr;H5SRbCt_%1oZZNYM5PZ+492kpH-rPw-(oTlPh6zKWtK@7tn z)nX?u#??Ay)j*ty84~=7VLzGpoA(55Te!a8HN&^yCU#jFwI<+`^I1=PvXqV?=E zkJhks#hr;&Kvz`lKVJs1UJU59MdG-H;8raAuqXa~` zm+5HP^itpn;AR;|@ylWN7Xxqn;n5FTt;vSCBt_l#`$BARCe;f{CPT&zqsPG3ij@H8 z>5}nHq&()C$H7n?6b-IYZzfIH(3{HC(L3a4-omOst`p8_7ECugiV+9iiqj((zQ0d4 z#%y`io6G)0o_e@#k{?*H2HTP(id{yL#mJP^1ez~hAgBWcAoFKB63^lEnGJoj9Wy-i zgDyVHX^9-W=X`0jbs>d~3WQo3c@$7HXV5h^IkodF7Gyo-I+WPTflk^$FTB9YgyjLk;G{bDXBbeK(?G$Im}&UGn*2 z9L9}T5F3S^riB5!F2R-{|3jm)B>MJ@QCJ~fz?hu~e&IP+pRkSKSvzhNSeNW^aX8!; z?!-pr;o@=0O!w3Wbl1!!<9a|oCz+{}cTcTiDn!R(uymnS&?eJs%0ThjAm7HTlRx0& z*-E-s+V6qw0s}|f98OL%Y$uF{Hij-_2Z&{kRR6Qt^Z4#*>TBdvZhIa#wk;1V{FIv% zdxIiZvBqKL_+qFoFbmbCk@5gzQCXq5UWOr>B=%l}NvgOJXjcn@t}tmp>1m)1!N#&x zB#^J5H? z(mEMxH$C&{;_aT?Jrx50i;V?7mlZ7vgk_hc49L|0%+W`9_hiVDhDEix>U+V+L5iEE z;brhfhB6!Av-k8c--mBqJNIAKaux1c#A-``;;jD#QBvS1Va%;=oDR95Y{SotuYsl_EnC*EQZ7aZ7LrG|*? zQHS;}diPal3dr_sBS=nDCN);TeCoF|I5kenbf_TB!kY_X=IdC|jl z>W}wgd-$RaRh$Lg}+YP@3s*>my#b+UhLwleuBOewMyr&%) zdhvG)&~R(lOHW7(L$bWI7svuxsnQStWYpNHlI3?gGSPR#6sN+@Q3Esd;_)(!*nizN zni6LHa*H)P9v3&*jWK0|OV8-WoKUax1V?0nswet>BsAh|pKxi`-LOnmivq=VcPk%w zYjcGLMJ+fc=uf!N7Q!m?c`#llk9qn@KMr!UNk zU6cI6Q?5T@+dd58fQ(J}@)fOh+>MLPaO17BjqoA%*BAt?^31^FVb_EQgXe>mC<-K@ z#$ypv9+|7^_c$I_Cz#XkQ68Beay%?6xEVV4FEY&n6uxmW)a*jlkV}kfku~!#JU>tK zuDSM0mH$N+aN9Mz@!GPz=WzT`o%`m)jK*!Dci-%) z|J8BUD~sDk%51C^4`n+;P<_0L-w==?xaoIFxnlwz8`-H`{;jvdh~0R#dd}5{-0*tn zOC{T^!}M+o3mbv=MVeN1wZ!zZQL|4{JuOppOQZeO7V643=D+v(SN6>Q)j5q`ldhOA ztdpY_(IeQ!g5fk``cv z&iBe8(7NPnpt0l=DaN6uQ@Dw}%P$nR)7nB1OsZmVFy}l4$JoL%??k;}Mn}+Z%IA=^uUI?T9t*^5p;+iI z+eF2ki@gBL(mfe8%qNad<+nlfq*r=h&=vj6W2sLq)9-OjvMaXV<5a-Pn0}AtJ{RbD z^X9e5fVT|k6(BD>8RIPle$Fu@Lrp73_`J9%4K=6}&e?{Kwwmm*4$azN=05kHG4`1` z^MM38cTALUFIlaf@VCD>gk7KVi;ubdjO|Z(s0gUrZj7cb(WDE_*5h|2^l5o<8IGLU@L=rX5ilcB%9!~*{75DfY&K#dZQvdmH)|Cu1oI8`ettM z>y)et95uk(h4Mib#O3fdzrlx^oB#Y9Gd$kgRx^>@abwqGi3L$_AH_bRNH=z4=#}fH z<_6S-w=wZTB(u6mo(37{M?`y#c`QBP*`i!Q(tPerX@FBpsC2~^sB~L7{_Gpb=ySmH zk+J6J5$W{XK4GDm1ZCaSN92j5&}#{W4NhIrIqLYZ(x9ryYm!bVTY^$;NCaxzn5&Y) zX^FmVvKH7M;V>o9rzc>O(*9s{aPj}=rRSybVcDKffJo?mq)yrHRl%Gl*F_EV~A9aEbf6aS>gXg>CvmBSj}zA_O{3F=PQwD=VQ%2zA+9=fMnP zMeIY+Fax;g(&}4PSF9x)U#Y^CjTSS%l44USl0?N-j&FyS$3E}tl6D|iJv{ZQ=&tV* zAZs4Y81%rD**EtttezeCL#%$=wjK}We6RLZXL!n|6NFbqs{)PnZ5M$Xrc2QseV**} z>XjavQXwi7&RwLclV6tR`ixo+kextyzGPTUn5zzL17Qf?d(6?(^J`URA4dLX?FO>L zjeVHY7Utm)#U7-{J}M5g!r1Y5*B2;+?uOOLo3Xq=ZKP1HmmvpOh3F<-?qw=Vs+0H3 zSf^g-e@*htW5LW+&umo|=ttMf(iL@{z1{~SF%}1juRb5-N61xwPD48gK51fjJHl&W zOwW$o{pF0H=U5u<)rf(`2|;7W=l^-^|E&6zc4niv#J`N`mOg*gsOgqsYT+E@61t^z zg5A>-rWf%3Sgbhfzfs&5mJLCD3<2+!G`>A@;0!y%j-zVWVMl)PyF_Bm-^0aWb>sRK z8_YgOJq8teC=X@ z;7mFOl96NJWg`^!cm7+Dc{%#+j_MS$nOnrgjdwLw7K>L2#R6)!A>(cmo37~dX{GN; zQh=m#?hbh>e_LcJNn)=+W6wQFVrUYJDkO7FLK6j&x%u=_PpqLUW|qA@ckZs>Q;@6N z7MTzIA#2&W$Tp3qHv1(r3BshncG`IYCXS(In6>I&iB9H_wEJ$Ya33HcQU8$i;-ZUN1)smeN=UByc=cC(IxJuyxcXL#5-y(7JC-oLke zXx*`FqX`%s$mKpr84UT>3Q;N_Y24RMeHQUI2s7CGX^!;M;rcu`@6c^y`*V)}A;CIU zVk7*3dy6W%S7hvUz=}e{%=}5v__Cec5slV&Yhae_Ma5_bAYKzxtoUtH;bgP_@svHe zl<3@eI}E$uK?FI+VV<;Tx~3bs{&7n>O31f4@UI`Yv&8+gUBa(8-Ig!(B8d@b5UT& zXtsfpdSlPG-ZEQ}ebfH5m~3@pD{|bziX5O=AiOB0;CXfPNo9~hx&><2?h}KL7Ehj}clwn?rz;Lkt5ZGTgFGfMv7HLkPyTpb zv(MhJXCA8plc%3n;r3;Ls!+V#a|Os!;~8gInUR~6A#9BuNO;*TwvjSaNVXy;Gu4{B z&21|pZP2kH%Vh;!!(^+nL8}E=p*+`0E9E(`zfEQyk*g$IxXt$ySdB86ectnDUY?Lg zAM$(VQeCQnIdTXi0}s3KHz$lxULA{ytc4le79%!97_hXpjky_7OB)^QwCE<1Kdl~W ztTDcxAk;J3a<4W9C8QBc$K*ZH8$z%cJA1tAS0RT`g2uyz2*IH_p=8YeJ@v*GYb9SU zAy+q!rrFT_tp`Gbc6uM+1!;7urj{!~&ZB}xRxe|)1P`|Oq${!mngz>ZfFTF7Jk4C%&#jn8U~#Sejgfu%B;VwP=djXIb#OiHRi3jeVAK3!h~N#R5A) z0n`ad`{mF^9J*J6RzuF`cq*VCA|G85x~VW$4{EHbJw- z4GBs;VWoAdIMIKRXA@lQm9Fq{hF9|9gpC*jF0wP6tP8bc`k(*dJu_mS@HYO4R18&H z?Z!2bS1eF;hGL<;;}{ipR$M~YD@{N3nq9%AK`Y+TYc4RIK$5Zy3e`IN3VG#Tb%GMQ zDEgTH>WTdxi!@dKd2|Q7TEe8T@xqPjDt{dit#tA~^~2_(Wuyc$(oew~Zc=pe7YElX zOQ1)!URlNTMs(5(C$0c>`5pXv<%W>e-n%6&3Ouu0((kb&_L_LN#6b)?0b~Ov$Hwm$ zfo;zj+aNQ1yH>+91Ii=VBY`Y*V-vL50zByyyMiJqR9yYUPO{ktWkR*90(*%8golNo9>NbLc;HXZG z^h)(IZLwFemxGpIBy(sVf=2QVjy?M?fBI^_bv1$8Zh&mm2b}j^%XSGe*t6<~5jtgk zczJ+c)8~CmzM8kq_XLzzHw!B1RK7u;AiO~GnFBEmA;{)9D(W5fO}q8F;Mi1hd5V+a zpelTQuUbb^xa@!2*mtoJcZnD7o?0YL0gx0$Z;jDwo+u9aZKOXTxxDqtKJR3vRfAW{ z)d#2PHT%`;J#*+%$$2swcEuq?ae;<|_tqWplUX|*F9z`4cqeSb5$W}=H#ep=LaTZ; zc$FqU#H3<58Z$n4LhPRX!4HkXYuol@>55;@uBpap3-YEy(F8d5WtcnD4(L& z6fjUlE6|aL>N5n-Bs2Gdb;@G(9Z^kaxu{E2OJ@e^HJ`(d=w2}Hn!CMMGcLuI-5|1F zm*CiySQ0P+XQ!F>EgyZ{a*%8pM?mg$P;PcF#gObih)s45j+buAiO|cs&l0n5C3|*wUEG+=>F_iLEWC-vCHu~j5D7Mc#TYO5nNYzY3 zvM&^>s<3M9gnGeD=T#-zM3tl1wK<@ZE^xx`nEPI5fA~!^L`o(+`7x<=V~7|mAaa3X z8!1vx#XbI73U6zWPPrB;S+R`u3S@fr2WQb)VN1LkAt1XGYO7ND7nnNH%5bb>yf0d= zz~5QoTMqTy zlWYf2ln2BMwb*oqLSgyzrCF{^3YjK(1gPSK60SF|yL%|oTHM8LV=Xr1T8hQ%rX&dS z6xdsyt2RJa2B;5Kgr_U&Wk@orQ)e?fcu0z3q&csbl|)yB@29(?(|jrdQGUaem39{O z(vUMSg6TJ!ks$|l!bwg>Wz4%rmtInsm#Z~tp%=+&H(suGTP#<36bqZ_YzX4}?w-1y zDGfR-X`oSCx#hj9Ahw~`EC4O#M7}BebC%S}Ovk8?4A~sxGbS}$1tzHuhb$f6i!zhtktg_vL!S7y zL&<>w($e7h-u1mSYa4K2xOyRhIdv2xPK3jdLvg~$*oB=BeyK4V6N73}207ry#^j=f zF*!-G$0$-m#bG~}0d_*U5Yca?^CxY87c`+kQKQ#;Uu+g#CPos#ThoyJ5zC^k@2fiUD7!s*4KH zs0nnGw9|O+Ce{#8&2M2gsL!a2DCi+`&@wrSrkA_719uDwmY1I7hNkZ@m2K8&;u6wx z<2<4b>}?GxlHz_#55+9d+JucPCU)m6aki>PiUhev&dPXUQFI><%XY7d&dIKdK4(D1 zqh5wJREzl)qQjwjfMkLYH%egWa0sl^yA*nXSrwQHWeul9mHhl^SgMB&Q^p~YrDOYq z`LIp9ER7c)lhp7Xw0f?`o@3F$Prv;S8tbSZ7c{x??#V{*5|Lz-phKCUHosMBpeF@@ zVQ8nB*9lv0i2vvyF1aVU>pDsf;ZCEs7=lCZ2Z3gn9v>ByD1OBsMQT z^XT*52#R?S=E9WBm?}9UqbWi@$Z$j|+6%H=aps#0Zy7?@%|nxtb5Iy_c@| zXx_)=;62xb7SX4a85357*U9$LOTg913dRQt#aTcdw?@1P#NI1_C#+vy5weNRX6k)$ zu`JB%2iF_HKAMR0^!w(w`W|Pn6I8Q=WI_hiXXT>@^rwa&9xgw|4QIz?7cacQJobdqd~yewVrqeZ=jW>Xv6 z6;V!Pvk75^!lx0=+RD6q+PK2yo3WM}9rIO#R59y4k zSN`X~aifuN?W}*Dpn<hg5xFU%mj>LdoqUj?z-<1-}rB1KCy0pebKXZW20iD zCALfcFrriOP`-s;>;+N{47(iJH@4r!;McgTTkL%K{8JMl~ArRVZDZs(ERuh`PlV6pU6Q!IG46;#~O$;td4 zh`gez(5HU=^0XK&mR%hUNfsCTT?NUhZIRl;;tP@m6Bnz{fk}Nss|#G`-w?7P3=bg? zTs8Tr{~7fY3Dj%(ba?8Nu!NyF_@U6XAU3oz{4#9B%L9yJMX;%c&2$>C({E9@cJb5} z2qYVP_QW_2HpgM~5*um0JS0Z!IB6S38?&qbXzjt+a0>c9z%+4N)GnrzzXbU7D?|?< ze7@$*N_pahevdV89;0`;wnWzYAIjy^wtm)?$F@K1S85$n{O-d~e`vNa!5f=;h<+#w z+p@5Rz~;q-ZCs9pO0 z!YI=1#;NcX7Gfg{#36EdX7folfeD;d0f6OlKLb5T@Jr$W;F45^A~7Aziz z@1|4UWVf-E#!*doHY?M4{T`RYpFl@8!mZ!KAh-Y$TU#O*FkQTpP=aqzW=NZrZL+#> zTt!Xe=ruK=i7~l!GenyE<*WHW|Jy@uXo%W|+)$?YJHoQB2b*D7``(IuB!^pG!hQD? zer4nfPc6zl!E%J?^E}mXft%gt{4FLqFc4`<6inBmAv~_a)?B#b4 ziGON_&B6@{Q^<8Uwq8#yfN-B;KcmQPDlP*G);E$Hz*?5BNQ=g#gHHKS4(y8X4NnkW z34+Rqd|Id8OnR7RL0WVjeFZAs@iwTXsFW|92EQsq=UC$}B<8&`9XLk>?IJ9@xGe3C zZVbl-Yfnr8$k1N(+8)^~*f=GhITV`Aw2KZ#<ihZ zbSVsUYec&!ndy!;$nmB|WwJO+d`Pq%hQqhVbYX==&hVAD^GbU~k(=#jKuxjy!+xDxvAKt9-lkz_#e z-g$8>t0{PX=omsDU7(g%dx?+1^wd7GmIy8M&@UvAq^<~c%IqqUxBX)p8l}}fMmwTO=adHN# z7Jw5Kp3G#?`O-r1I*lQ+DD>L2BzDO-R{+j2gjtyv7r@N-X2O~`&6x2JtIJ3(H_W(k zRPu-gW=bh`7e#hZaanZy%sP3A2!9`$v?&lmkswwLTGTtqsQ}m!?u*5QXcqlBjS(l@ z;6YiTsZA?ajSCa*=UOA`W$Qf~d=q5LeR`$Mux~_~^;06?(JmA^6F*MD%>ay?h@U0@ z&|mth*Sn(}Qk!f)baxg7^@uPCfT^x@$hg$ec;=Z$i(f);ouJt#pY9=A?4epLygM3& z7>>=l6Q|e4P@LAIF^7Nse?!NcZA;rPuKtxQ9Y^vlzLe`I7Mdhh0?RiPS7NbkNwiMg z52fH(6; zbSR3Vm-_CNUiP$z?4Z({WE~> ze4V-u=v^RLA}CUo!~Q8%yx12R{nkp)f(GRSpqT|~oq*wiI|938X{ERySz*ia&^FSV z-B(&$#91?C_Bzz;XD^ zI6N44ZP#C3^(xY3RLQ%IvlFRXnyoQ=nO(D5q89t6^6NdD1&g9qsZryg=v$kpE&rxd ze(Kv21v#RCUeX$IoAfCKQB%C`hp*3zu)~8*FeG%3|QF-4l`^2 zt+tKFP-!Q&^x}WMjxsy2G{vAeF^1?DuwPh*7lZQlf3S|;X(vz|9p2&RYy+jDXitT8 zVGfrzIyd%#ZPXVTj5WrlHRYKH>Vp;vON?sh+B^u#T$X0YMi)AEf@TypNtXE9+22;` zZFY6{ef66SWcfG{2pGJt$fVe{6#V&dOFTD8Zz`I}NBd$w+8F|dT98Y(L_GuAJ*+h8 zjYtl~exjx0mj;bqM-iuw6+c|df4$3&Y=0MrGTLq%N(uN;BJ-+q)a-w2|64lcEnfUg zZMEcZNHX}GIbaXE_u)xsKHq}cNm`IL&ogw0}gF|R8c z6m$G~Nu~;jqj?SV4LX5f0YZOh{6uXb(D7jv1nTTyCDv)>CTP)J0FcuYEH^mg+sp&q zo$>&*pr%n!<{FE;T7Y1!O!&atSX>+r;~bMW3$WEUCGfF-B?$H81LZm7gz}-ZtKCQ2 zOy7DNX(TbhU~I~&mm$Z=erUO}7r-My;vi6Rd1)7bFK+ql|Gi_z){kyCWRm@EToR+V zK+6e=t))mc71u0SADh6>580$?)U2bLznS*Ui{HOJ`Dp8@t-pe1h>V4-J)s9jX{-G}dIe7AFU)i<->-|*et z@9LE85!WRx1V?Kk`_yTZJ_*a?rSP)(M`nK#c3}2)IxA!ZTPN4T!!9HYc@rl?MIAl4 zn)<*DroI0s=ttz_Yh$QxTR`tJ#R3-3nG6-C6CRNgQKPcc&zMQ@%9kcH+E(ZbPhwA! zjZ;uMu1nD=S*$uVG0!cVq+L4sFCa@iOp7_y+XTnRkS|_Ssx*b zy(+c>>avg+XMfBxc?)A|5H-;AjQ_)eFH?>iD@*a-t#mP68M+v}@ni;vM5AX?YJYi; zsB5t}j0GDKfzMCg-TRL<|6oSfdq=wd4>|GLpsURSU6&~K0!12;j3R#KNwQ3%1)fM_ zlB&b=pmO(Q%u+#Y4+}bxYvQu>EDhSI1q^b0nQHy8I1iaDa@BekmqI9wM{PNLqC!?E z&ZqZ--l_Hq!5Hv0$r;JP(1W2^s|V$LNo=2w$v@L>oYE|4R3@?MihLRgaX>azye~FE zxc@DDy%ZR5f#b^)MO2HKWp;}4{!PI3978r=<=**Mg*Cei>aPoh~{PtIq$68nWz9vS*hIkb8xImpxx?-(#ojUEUZmG5jcK=Wq4pc`u z1aXx1R&g+7$dM{1UGdPX0$lAwQ-*1Ya(t?+pR^xRmM)3)IL%GyQc@|vf)0J&#IUnq zF4A`8syFzxC=N00K9Bt&^wbiWOXG`@$W|KJ{BcL`fOT^ePL|+$*%5cJDHlgx7n$L- zH1o#4lO#9Zz;Cqx$OekdpvW3i6(MG-FNj*@usJFXSrDd0*VGWX*w|veUin0_enK0R zDb!9s{ti&cIxv6SKUqg+ocp_Qe}dDRH1_vzJl_6&Gd_M+{PR9i&&}MpaeC^$1)7>D z_8LVlQ*o!2ec;5Fl9jMy);>^VNLPuq#mr^)rr#2#gw9c8D_!mEKJOHsk#3+5j0>n7 zLaojMte8MOcvz{RqY4iT#mGW{+Z`k;Fcko5S1Jk=d2}k#W&}R~%8NXD5h)ZOP$#jl zr>d0B7v`w*=y>7MSfJlV#phgg9gSs!1rTJa3~ha@SpeoQx|J@W%fsP`Vvy7{#>cZg z>mi&tTtLwgM83@8aJX(0sx#_*n|^P`m{|J57}DX!HYnMmD>!~E3ypStR2=SnA#tI& z3k{VEpnn%uA$bR^KF>VRr9oGwl+?*#&z!EvR3@oVUFewy{wx&ZPoSM$q&gg$qpk>l z8u2jVnFszX6ys00grS#{tls}AzYkI#xI51XP9QfRaE)(rEz%Is#oIVV7YNOFfeir* z{WAox9^R9IR-Uqf&gNq?d7n3qg@ekPR^%(q+?MkB;Cm!Cx0``Opl(y+8Jmk-KgT z!KD@u{G4KYDe@3x*+49@UA7cdKaA9HnAl0B=y@O-jjS#epqqtUKl6oYJY67c4x0sM2JF<{Arj9xk*0)LWTk+ z8xZ;Qdys`;u6p4BHx>AgSrv!_px2KfrCc@stD3>_E0DY<=$rmHFx5IB@Jca3*$5I~ zvvCC^*7_7$4D{rvn@9npEe1t^n-hxOtaN_oxXmR7FK}4?h8@S*7i~LaNnii1&RV|H zZ39JnEZnIr6bmKrn^1*ww$m7thP7>#x>31GeSy~I z!QKrv`#Lp(#k9x67H_m>F4T+W0E;1=9cRN~U_9Hf7@|yNZ1nz*izVg-O#a{RHVaoQ<2N_cajYUvV$DPOnuMKgFjOMt zEe4G8>6;OHl5|W+n$Qk0LaarbTLGC#+_WJpVZAJiz9<46J*d{f|6CnB;({s5 z^~klzDoJOTy}TpOjq{i`I`Rzk)(}%ar%Cw-!o)}#ch?tdEO5^;TtvkMA32x#cfs4+ z9MZN;X`dvDV9X24Z;$sJBcE_{J={1%xY1%6T1l~~6iK4uu(KE1(?wU9&5{#xtpT#2 z7~6a5pU8K$d3a%d?8eIrPrCNR`ljl=*2zq_-MiRGPhy@3YpbwM#URD9mLdvMLCbvem$&Pv!QAlI}F%fxFSPvT28$nxjeCyet$QE{OVRTf5h(J;-ceVA}P2CUzK8zu_rDP`6T z-mi%}YNPc#(f16mSK2`;=ze)VWQ|bHqfY)wFjk;rj0C-q7RBelKm=kP#uxMxX%c~h z4$sY!8XizIGZi77M0;%7)=21^n7)?Dnz~!wE74X1eX~y4=68&)m1G5H0cS1dr_j;a zN-vE;plUI%T_|45Yp0P$uTp9Vz*`!9^63J}vbVKqkTGcV!E8YVB+fhNX2?X<%hDCw z-~IpWeFAAN@Jcs1Pkc~ic1OcAGo2cWisx-Zu5# znQ1SgxP#z;3rGOj1Q8SvL|IkD(HR9rMR7$UIu0X(G7Ktw&yz$Yk!TJi+-SdFe;v*_ zIp+n>`_J<}@3Z{>>v{V@f*SOxAWzmheU;EUV*a)0m3&atgyd8}50F~b0Bi9-@N#w} z*yVv`cc#;NF;@C5cc_ZPQ|1;(hFvfPGqUCu><1PEqpsWp%<>i4gQqh^hO!fhPBQHyRh~L2EEbNQOXZq(J z5a3ZvaKYCuORy5*I`1@%k{vbTd+ybuOktL~fk8#p17th2xEzf<@s-Px6R!QFV!>UX z9?~q>=9$D#39Ats@3br0IKk;T3yVELW*tsb?*HsJE+*&{eg5{lB*ux|UY!|C5-B#G zB5RC$U_ITc?1b(BB){6~6+g$W(`xHf*fRaLj%CmLtWW=FJE7`1htB`Z1c{&R-}+T@ zo?8&biSsv5oiKt5>tl+AuFXy=`c`oB<94}gU)u1k&Hr@j^}CCX|FHCp%ZeSsQtzE! zW$NVt{gUgNl~XdjKMK!PUsaqGZ3zMg*ZpC@Mjn3FrYRKf3rta*C3SH9V|Aq(Y0EcD za|3IEQy;{=7v_owM4gI8CQ+E}(WXJsnG{6>n+2ps=M>0=k2L1@NhPmYc6fdx_&t@p zv);$u+uYkU1*b)f zK=U~UXDp<<$NVm0{uAdg)LtEbcpeP8)uI|YKgZsY0?49q_mZ6 zE><(POEbg9|+61R07)a0cUF(Vi4#^Pae{j$d%b_-ZZRgx$|MtuD^{ zeO2AAD|IzlqNdmC|4x>_G?r+anI%f5*d&T1U{j_}Y3zf%H4g%Y?b6$#wTjMJgP!os z-(gn{X&W$SJGyNc%#ODx#0%fEWWI6QZMPLl9^ITm(jh?A9=%*QN2feUe=4~ss#%B? zxow&TCQp5H?hQ$c&xFKROmk!Vd>c%<9cT2}LvgZZlzPqUU$8`nl}We)+tzjbLAqX^ z<6h&oTYQ%6^u%Le=Vde5t0u=0uQ+}wdGrcij!AG@UZjveCM~z5AAh=-)`|B^R=Cm| zgRqJ})3w&4RSUT|WOV>(#_^DdJi@2VKRE%}=26FgqjhVN z_niwBmXxkfW5apfB~obSRBWZ#EeK?gI>ee4bS*I0+b-P=f=ULjxm^lQHJuS`P%%K+ zL)@$Y>BmG^J5KPhwuKl!8lD=#Iy`<>)3VCfwO?-ycXI- zXEH0pS_D`$P^PL-Wk4Kkv)6}_t;|Z{j___s4&F~XA zV}x6Rngw|G&AEqQn^emiAo0LBpUD)8PY^vY>NUad*fPs*s+0DI#5l2FdYqdfl1**s z6+Z%yVB=MiFa^rcMuy#W6q`no&8UMl$V0v*WAGDN^{~Pu2lk@FF4*M^3y0e}UZGoU z2>6=NK{z-;8>2DIgMD7~b4KAb56)Z3`upJp3oSbgoi?CirJE2##83LcAUKZ%vCv~L zM6$aHDY_hq725$j@+rq3a|7)Eto!VzubB)-``2&gk#Z+qc}|*Hv_llzK#^K1`l4iW zU_X6XvJ%z{eRvCfh*v5vpS60<2GXk8#C$frm%rJkCGxapRnSIA>!tbk(y{EufaIBF zK9!Nhyeh9lyhQ#MNE}}B-ROmPHU`y!9QG0M14TED=NlnexJ?~Pw$ayQmn1#jZJGgc zW8SjhbpK4rcs&v>O{pUtj^`ea7igSdM9u6!^PPV+!RYjy^OMLqZjP`M@3=oR1KmxE zy-ty9RP<`aYDJ+K-QX0(QE5yR?!9sSgpfWauKFUX7GP!F8ciAqD;H5k6c*QEi@i>X zp_%chhv7rWAF%^yin}0BS*1KGEf%70 zmOj%spcM?;Zs48^#xdS7I)+naOWquhbg2_#?VuUf9!r<*0T(z|jeHJoB``hS&GK3i zwz%G%^_EVV#C9?%Q&&;Y#D;9wsBpU8BSyI115&Bjpi;4*)H|EBN{3zG*=G4$DO0sK zK(yrslz(PRPz6^XUg2|nPB;B@R(bT-|JTpk~<#)m6nE_$K$#_o0-nLvUs0Zr%~P4hSlW={)~L*pt*thvay* zLr^WKlol}W{q>!O=@~xVw2m}DIc-hIU1_EwUA5h#nb#T)a$W-C(N#sIs(^X-uYb>{ zE9E#wry?hO9lt|R7FiEQWT&JiWWTsdJP7NUm2c*A|Gjpf0?RM(bN_dh2PG%!c#Qf90W9UUNMSI389v!U#I!VSgue;zJ=(HZi$ zE?6eKq^OU)81ez>m6)HkY6^bvaM92|U$9qtFQgaU(I-QJtN&#~{FPjLDmg%`k(CJN#M^ zjx`RjjSozg_RsQs?D*maCIK^L?mW{P_U5ViJIU_RYR;Y5!#QEL;5JZfEk*WI(b-gu z5~-`UumuvRgGo`u1tHm|PB{?KB;NE&EI%ot83{bId3)X45W&qk~V=9muXZ(15P2swYec8kAsy4J|Xmroy1)r_-a*O9my zJF1@5hxzUom=MeSxYMp)R&wFH+`3>(Qz+aNsXgd(5K3qa$+vq6d!?XW!!ETpQ-OMK5~795;$XNA#}((X64CO7q1gf-q8XcN>|}PBA`htOEK_1)99i({StcRq zfGUeTkgR|(T$cMlL<7X!@Hm0Vck7ku{Nuwi!>(&O6$wnSS0ArUljWZ7-$yQ~KNj`U znZmw^%X2nBL{F!D>-ekH^G?!t+~2|=Qc{GmfcWH$El_#!eqeWKAJDv(&Mp)ugrs>T z0V^x`#Nt!l$X(T@xgtq`Fcd}xchKD;9kWNZW*REE_cMAaYMPxBon^X#)fIm;M6^Hw zet^6jq^eYIb=H?&bU#-QT9=9Z0n(``#}0uUb%hVkAt)`iOZUC1MJ1*fK4X#SYg5=7c(voggY^4wL>k!e-mFjqff_#)H!ITt7F*eOj?-?JHU=&Fh3BiS5%` zpi6rDw6(y>j$)M`D$62}UK4MwpQ^_chBrg@}T#SO6G3!kEB@kE_S>`^kdycq=k%!Q8i_A6?# zI!m1JFv12{9nel6SjWRCjZBmF^G|-=vZl^ycRg0BZTn4_Ch51 zE`q47K?TVOkXT_&c!%<1C90oewh*zTr_pUeaqE+!qw-;wTs2bTbvfit)${tMN@_nt_|g=)AN6N>UU_5d{I-uFwSL0znCZ4DkT? zH@Prq2Pgq%d9Mx12M$*pc-RH+LUljXUmt`5b0f3r7=07FgUK9PaApY;+s-+GX|l)8 zbq55G_SI)M9VBbG?W>)*>}9*zvX)A*n<$b@MI$5CV$^rs2eiFkyV)$rsZW{`yDrH=VY$J8kOWOL7xdGG_el4ivua;QNwpFpnYFn`){XWY>7L(!D=FM5LrMpXsKf?kQ*>=e7Oz|8*aeDH zRh>Dc9Dt<}L+<%FgR;~pw;}~zk9mKO&*k^&TG2IV{k_C&r!R;ui1zp#g`%Gw9eWPyuI1j(_ThfDkPn6!{hTCHv#RI%Svq z<)~V7-VJPB-B+^yU@}vGY_)w6e2rf;XB#<#~6f8{wc;Er2+RJAuejwi@*eS)9?)G0H4 zjSSxicwWiG?Y$2H{4~23L;}thoG3gs<NJyK3HW`@TV%5nhgd4ijL3G1g~{JWMc1%(GcPGR z1XUiL@&=z8x1FAQBn>{On5R=d^u4tYJ7Kwzl)4LsC@#?H5rn>cR+kr(y8bX)c7vRhguDXTaU4##&LKtj%*z@ul&-t zk~SHW{vY@sBeA1-DNekYWSSY64HUbcBI~H=JbDGers!_k$ZJ|R|CubNj+_A7K;g*0 zVDz!AgW`kIwdN^=xroEY}f^x77ZSTKa|CxMsXLd9rS2b=D`Z2^Dh>a`?V@tl{e?2 z%%;7qh;8_IYK?jN!sydH=Bw86G0IXiDeg+`JWEcfr7$3*4K(q_%f2l5|d z3(_s%DujY6D32O;$@eQ1o+bLg68Z60vCgG~?hC~xc`cT4;BUnTv&j1)t3CS_HE!6> z1nk|a*jL~oShg_H{7Bj@S8e=T@VZUNSEeytc^1?jeP$hUv0FTnE#vl18 zPyk9?<-^c>p`RM+2B6WXRjE^-gs2%dDI;})f!7p?P#c-dFiiCr2~cgE)JHFIEI-Ez zh_U=GF3(s8#Ar6?8!a;y2AFKnA4(3bC)rNCJb{eKh~=q_VgWTfjrOKR_Nbvzr^E)w zT0w&!6pG zd20K^XsoDt%6@Qxgbnv{Le1oTsVo13GNI<7$HU`f$i-sT8gZqqOlU! zU_Dx)sxFVNBQ`AqHoQu1x7Yv(ZUfrsZcc!hbmWH*#lJ8C;@g*9`^W`uwYg5bD;hL| z&nFanlOortXe>yB=6V_WA3CLO<`z|HFof7r6rjw9hgc)G1iS8m;s>1N1f~l-6xd1| z>e!aV@v#O0n-@D3@3pf!|3dMHK+S!E2com>(zPCjU>zv=@i5Gbc}Xn%h0pS7EEvV| zagfg86+#n&&Oe{t=z*l){mc?9YJerOIRyRkq=>$VB?TaPynj~C>@*>E;bQl~5-bEY zp5Pz0q)?3B7mDQQQSf%jmO%JDrtg9c;C#w3Irf4P-3vC;hhKEcLSx}JHnBog5-PKG zK!+UWn=wv{QQTXeYdsv@WMpP6xXrfNqYgdcW{ST2^%cV}IyOqZhXZgo(;?VG=@%uu z-r`!hU^t*&3Z*lOUjE8J=oOha959g!7AI^t6brN|~KIxXyK$aP7)rX;F6TziPDb?Y>2l6$>tJ%C3FyBLiD9boK;5!UdF zK$BB1CAT&KO*Pw;~?8h0;59aSz9ha;WZu8Zi@yrE3x55YeRZ!vr4~c1fREOZQAaA90&l8us#(O3r+UBsOVmGKD4bRF8J6zZpxG66;oO@>%Ut3t(%Juf3xUc zPeU;a^1W{|@x>>Jt7lxE^ZA@XsHi*aS4AZ*sAO))K9?nlaeLD&ZPWD9*S~sMl1nEl z?CkS!0L~}|XXGV2q6UT{=LBmfn}D-?WpE=IcHw4{oHrbkZU&t76uXWhtEp(*i#A4_ zUI1RnpgV?SDjnUvI{J*+f{4+dbnMmh8^1JYi3q2S=3AL#Tq1y&G7sqIa?}kF5t-<9 z!ZbzBI>(&8;*;NB^?(0jSs&}Ph_F)e+D|S^-dh&@w`+A5Xto|yQY?@m z?53g<#JzNj3>gsenRxco&}?3cVi{?I)XpVH>ekESW_QWANtYOJs}mKA@TL}7ZO8#t z0r2?v$VT0QJLUzYPY_W9#9U(Q>V`o`o|zZYE!uTgr+?<$EAUL_e~ zP?Vwzn8azP{2Ha(qD@(kzBD`H zV`g@;hGO?oK%W?Wi+?MsScJ?gCwx-C%~|JJwV;GF16gIC;QIV^pfQ#~U*})vm&pr- zy-}!KhSV6z>TXq&Pc9wj)26BACHwaiT+cG-b!^pw0%b3Bo@Ia~EFp!$HjN&tnv(rL zu5D|P8>G@P>hJ16j3aY&gD_j zrQQSN80nuIH>HfeBJNjQ5x@1;8cmw=wCpw%H6g$0iYY@baZ@&YAtQLh7j8zq_0|cx z!tFz0m+*x0mSV+}jKDH_r)0nC`uqa#6=65$*b4_{(`FA)m@&b6ddW!t$01ACyjnh>9^L+sYs*>mZ%8utD7gzH+;Cy+5djv`dj*dDsOj z5|CXPS)@`Fcj#^zB&K1wh=V+x`ZH1{0!r^>|D4%}B%eWvz_5#+P8Js^pIpJXUzRvHTbuvY>98CcnHEwoLU7hVvqO@*yNg^$xR_jvgJ&uf-D^cSilz{;+@ z!LL(sLw49zTP8xbrx?(d=~Si!Usq_8#Z?PB=*k&u7)PfNN16_9c4_33jy<++mr{EI z6J1QpSkdQize{4Ac!v(sA|sZpM2d~4$QlqkGf;0C=>Jh~#whITkag!dvt-lsKW7X_ zd~ALCN2|nL{NdM+|6{TG<37F%)&nfYtrf(xSksNs0xW|{_QygfzToAUM73d{_5F3k>r@ikfN|v#NrjAYCsej<32GK3MD@V_ z=_curCN9{b?Bg{_+B7)}^S!ggNKF|Nu>qXqPQ?Mzz;~RL87o>$xEXDt?Q*So(ih%) z(ItHAf>uRc6#>yL<`zG7TZ~n;IweZsf$|}ce;U`5y7|K{>v+YYM>`?M&GqAn8QY{k ze%xUPoSc%$=U*-TrDeFqX>*KLLMxeHyDZrruq?Rn>z5_>+}kwCy!Ak`w08PtpBCkM zHdAtbVY;dw_@ALf1dtexztz{tL@{J``w*|p^9@NJ!s>dUkL07NVUFOaa4_z;Rk{+ z59Tmfj4%UZLdb~c9d*bFA(Jw^&c5ww@=*S{A^j|Q--$QAP*6L<=g6j5pjt_%qKm?6 z*n(N@0iVh`fh+K8$dF=jGw+-T6QNxYeaoSn1=TZJXe50D^2<8XO@9)e!fOrF(;M9# zjg&E=1kH$};9(k<8%j=g{LRNw7i6#5WE4_tK1FhRN2#q2%>9v~BZiLYMzk+NQT5)@G5!_puD zhjFbP^lrLekrviOFPn$)=~mC}OwO!U&F)|P^~8@q`Iq$9+cZbj$0enKdjgv1JnG&T zPiyS`%^n|Wo_GfK{${@<1J%Ohu=J?R0y{ ziwW)vZK79Bxjw%e0uf0b9VBK(pW?Cr!w}o)axgx}NF(e$%H{V$vppUvVm($(Y15SO zR`4J4&WHv;>mx<6a!LV%&($%>ye7JjcZ}5dwrTPMPI~6?E=hKKx4>R?gU16&%T!{Y5ErfWK^^cGnh7IE_N19U<#1qWRG?3I6ue9x=mhQ{58I1>oS*W1t;z!18052R%#2yLRjE^ zW3%Fwh@qf3p91AjP`eb9$T{lTkmIV8s$_pm5TO>3V-1D*q;f)%`J)a!)83Qce9<|N zA4GLe&m#9<(c5QK6rvBD3TtD#9m2lYqV4v)RTU2?2PX_uOaevBNb*LkN61ctG~LH7=iX4Pq* zL;P{ol0SEr$Azosjz0Dq<|Q6})3V(E=>jDu#)FmKgh9oMX=qlA;oDf@a`&4c47S+l zwje}O6v_TNbvKX^9`xBe0~?sZT|s?TRIP~>uAUC~LEV~-ZY5EJATk#dh5z5^rrqdf zpwu07M^as++rbA+SttB0i?IbI(Q|3o;>Ad}JhqiuKy+!c?;s z6|8P~?hDr`mtgVgRYjq&i@y1#j2|}rxI^&ZP0S2nog5gI#-JiPN1f=}E_%eVWzYA{ z4FaQ>1;ZtK{zb;ItQ(HeD&elr8=Pj@JM@xlgFx4L#5O2}VmDA^Jr#XX)Zn+BzUy1& zmq9NN9(L)I8_6h7(s$e)>zRx;*EXYQw1*vdt@S)jD%A^mA>c&kJxa7?G}$qloG_XbON*0^`vX3 z;_f_*ySI8Ki8~n}TEP#nsQ#Mw1i+XBc1Jj#V@0Ix&vC=<&2LZo>O9kC<;_#`caq&s zyjeM6=1n$GY%N9hQ_-2gUf#$IcwwDdo8~^VL%7vdA9=;Qmd@k#09o5#)hF0a2oB|%FOER@m~Qqi$&hIgii z@xQHLMbMJuER1JsLIy!Xtb-h(PXdYPkW0P$@-X~~+p}`sx~ZK^fp@p8P17v9KED^b zlZ{Utm&P%fj`6?j0hzG_!}^olAk+ET#)Gp=km>(d*$T343h?rboS9OJEulyeurhk@ zb!`%VJpGWQT2vB+XKk8lLA7X*w_35B%I1MmkK&$twde?48&VQ=9|STskdmkxw_bh; z6vTbV8~F)dt8qK-ia+gP^DSP7J?6#{-Z8hiaoR0&^MASfd7D7f`m2h?WQ!BqnMyM> z7f>wdlVzjc3y>7Z&V&?-dt9+j|DjKu&pml}D3sqKl!2unKwz+cXErb_mz6oKg?P{FMQR_$#M0yB&0^aZ7yz0{Mz2Q1MG#l&ii^ zV+~x3tO)oj^x;)%9l$G#y0vKK>q%ef|I(wQ8!n!{H~;*iSGoQDn&O}V3|myOK!4J}>~Z^)Gz<1c;`ZqT)5ojw>Yh-NT27Na z)|nf_bDYcy^=`}wjb)SAOOoEP(o%-Ksah3927qvUFY>_6FVt2W=ly4 z#THSdfQoJmDp`;zzAV8;v3hzyia+z+64(=>ldjjGkpROhy^;6f3_Jz{9&nsp+ivG(Q@p2VdcJK!Owezq zO(WO2xhhUvf3ke?$Sqwz#r9F8hl)nZrCqORlcv{%56P1vK|!hy_*1dYtz9}46f@nx z%ai5ar@-d7d^(Yj)ox{UO^Bfc0lVM^+<_i}0dg;;w#E(p9G$XR(BY8-&T=OV6$31~ z{U9&~dSWMGfCi>h5i5*=@N$W~a*npx>qnR-elLUSO(<@ub zbgAJ_N??{U4}FH@gYJd?PD}B%@r&QPH{S%j48_k2$sTUNbKbT(a>~rGHBu}H&eT!S zgMy0Loe(QdfwH(-dWWRXt8_l-QPJtDn{&(M4HgUnf1q-r#l5 z?NX03J~z%D>h0xWh|?4M6#2T`o>05oHM;_i4K zXv}D^HjQ?=Zhi3+O5E{YiB zC5T&~xqiJ{@2n+P;e!-~UXbeEEWg7qkZhH%a$C|OYgJ+~cNM)X5Ct}CVVyx@$#GhT zr>1Yr0C{2;j7Lw6JiZIo0Ww-aU8`Tpk1gBjx$L-{c#*Qg3UJZqEPp$_%d-bO4{#~G zPkNqIonN?zY116`OW+@)>-gPt+>~p-*b$lj#%8MbYxzIg5xHWDPWd^s0>rad6ev5? zA!>A__Qd$0H9q(}>**h@nR?_!&qP(qH(@7sWV&jp?EFH^Zr6q+`?o6w-F3 zCfV)1P1-DLf~)%?(}UNNLHA_!g`h0;T5>D2S(fkcZpJZV>RDF74xf4bX8v2_;jnaK z=hq5{WgOq=r8ET^4gW0^=g&?=f}n$Nl1C>JP){Ok@5p*W^L2dsZYD!XFdws8U~7o?!EMV zW&l>8a#4q%F=(aRInhz|a@a3t6R1G-sF&&m$a`K(AND)!-2?Z=vlm0|MBtU(imU29 zWVs`{iN*tiCr+;YfxyYbrAkkIagn7u9+wr!d8?U$aBSoX^jI&^NM+gVs=ea9RQS2H z*>yst;-Ad7E%c0;@G-_K?3b-~j(HshPH$Y2e3^Hq9MTU(|lpg*hD_Ej-{02=AG3$k%cav1P)Ze6byP zWPt^4UwEhW81MYE#~&>VY+e#eWThMjl}6g7*}SR+P=KRTV`L{rn836uftFi^%6y-O zYFlL~iq+FUQ)SS3w6<54Wzaub$F76zc~>aj)2g(8@a8eV#tZq!1KIiD%x~V@nlv`e||oCyvco**hO2`=Lzglh7Nai9Yk? zR?U$7xcaoF`x||WS~bJr83CF6E=VIKMe4&pfNUdPIjw2cbVFl;zoFO?#!SI3>0%>@g`O(-+nDY zjJH7o0lYI5iO_DHewZZkuPfdU(e919=%(!!Rf#hKdk6@P027P@@H0lhSh!Udvu%e{ z>wrl}SoLRL)1sq}UtLGir$E!tNNc^5Vhbsf4+=lxn{(GI&VT*Nq9;eK8a$8te({&j zFT!?^3_3NcUKK0c9hs%RF%KF+49_5M`Vq+G;PZzh#-9#&`-T@a?kTXt>mF?g>r*gF zj>dV-__r%Jt)FYc%b#jA4W!74i?NQH0lAuDK{ljEWViN#pPdY5KF zx2V_qN*G@05PUArrI%7e@_teV(f%~$#_5;C;z&E&MB|ffntn1QNAA{oI!=gB?Gc;b zRZl;V;GK4Z`M7cIAS1&K*K2sErIxuzr%hN}$tGrq%T$^CDhLuBCYX-KWKn7$w7W4i zGj~L`YPy2+XFqU<;2qp|RdI*OD>LjjuFKB|LDpmS{uuzP(;e+l*f-NZPhOC7aIYttNxgMX& z4v{Xnv6e3P>r-gAv3lQ5#n!N47p$rIL^U+8Ly!jn;ClWtVV-M8Ch#LX4(QzDu=YUYfraJ(^m2sqj-74yr|m zx{!2>QWT{s-FVBpAJ3WH5%PHC@f}#VPs3kyak1PeaY-CGF{-R=tLlOt`rPr>eyolS z-Uf}d+A_MApUpexoyl~|07Na)26Mk+z@ergcI8jD||C8o{4-t$m|V3EgiZb*jJDs zKFZft2c=5mSyOL;$jI(}eERX}G4MYyD@G5dZhiv4Otg4$KD}jDmpXym2;b<@Ods`f zfNqJ6ka@m~VrSe;`_FIxZGp-Cb-#Bcg;Y9mX!o?4L)%2LhbYnj0?-ln=`{a6q$XtV z%)Nd`y?ea%f*bSh@Ee1gq4h{F#~`n<4YDlaxN4PWH}sm|osdxqyBbpCc1DCMuU8~k zvD-~!0p?E6OpuPg>tiIf8sxQVc6x3J%2D6qANK52?3C1mWCZRZpUi0z>je+_8mwWh0k5AjV%n!V+pj*t~jj3m&n8WE)1NU`rzB$0|nCVymC%=OSF z`^U4le5&UmZTGIoY{j+(j%Y4=_WW=F#!SGCRZGEs~$LtL${N1bx?UMA3O;eaDBA$A7KGlu)g7q3TIk~~=9 zQZT0c6Zyjdur1iH-Wi_nb_i;w52$qh@nPDJL7BNlR;VtKchm1J6>5>GUk^@!R!{du zlmX2`nnEwl00p`Xy2cl}*^v+R?!21t8oQTMb9>cK8|=qdp7Gko@)5VMtkafO&b;u` zBhyWbjDBJ0Ws)|9Kz?_GE#E=0+bNPuMQ@xLD;$9G+pWq|;!h;Ic?Ha%dqF^#e48{6 zy2??Ov!2HF9Ai}lOo4E{e+5t=pgduWu!qLlg|sMO;h27NuH(L`<3!6?fMC-zoY3OD zbqlYr&%0q6?SDyD9xJ&4)4F5W9qZ};t^Ki$xoDf9Xy2QT2=*jf>WVxaopoMB1NY%b zWA{WoL{Tv_-S-ZEfE-(x?KE z_MdJd)4H-c?aFURJhv>66C(!Hi$*Mu=@gqvkxf+e9iOB02f+zqER@tK>t@D#pf)w0 zVL@L(fRSLnBDi01(apYmh0mH^o00R#PkPp1cKmGZFW>u#Wg)iH=8dfsU4wj`Fy9?m zq6&E(q20inktWn4{ak9`9XLd3E6nVr%CtIYI@+&Jfb0&`Uc2CuGPV>QhxgBU-XG1s zgLf;|EygGzkjXvS(`Ni*S&^lIx9Mz;a=DIof!Xv5T8?G`o*e}q)Oz(XIu|M@KbOH3 zB_;_cY8~Q?CJsm)&#lMmihf`zKElNxb>jGZvDqq`mOcY{pggig2FblEufX3tE-iS@;1#p1vGA&s%^&n|+;@J9aGgEMt zVlPpoor*?j-Qw{62&8{26c$JzHm6f2tGiYC^d-qrx#6^bZh;a@i*_ir{d18Ptb|;j z-2t^qcu^l7&qDC;nwQ}~wpxbrss<`c?J#tmta43OBe!H#cp=bMW`;qQuf?ZRxkdv# zlKrF>Q~)#RYu?&Ed0H5Zk{y=L+ySFg9askymxaPY5Xjl;si#w>YImrNe<%`Zu5yFF z3)6))zPn~yTfS_>(_@##?gv)DX&q0aI4_e1zJ2hBhY4eEUL+1k`v zd9&Ovj};@_e-pPSxM5`bKkdIZ9*t)w-X>eoc<$wAf?U>>nY($Niq1v<0o8B9Y>!&i zAcU5&{B4==knjjOE_674p3EvYBt3b~(MMPQvwiD7R+=vd)|+JiLN`b#bt-ezS4jb* zt&-eQ>`)bl*YXO5HNLAH&h#~MRyiSKV1K0JxohZ4y2s4TV2b+3qi$jjHv+ADH0UU8uT1?LDaPQEQekn zhZ_=3AQ|DQI2a?7&M!al7up1p{vY@sBe5?HNHWbpvVmgPQ)C?#-7Yoqq;v!GiGiy9 zk#dp4G4Rw}+6aoLo^$ktTfRBUKocmM-q=t@a-7&F1#a{aK503{LL9ajgx!743tJ+G zB}XE4$~}@j;u^Qj(%itqerKTtxl_K5pA^vmq1TU;EkOE|8MazcML(1r5oEGh6#oe{ zT~tIu6(g0i@G#w~DgFA*MY^xW{N}U<=+>Ew_Fr9W1i}crF#4grfx*d5r3!zbrYxD% zxy0F=c)_v4v5tFbBMB4K9bsn}^eQhvF7248dO9hx6S^&8grE9ou@DxwKssf7WLgvk z_dixwIDFX|2deB1B(w4U{fB=@vdrpm!HE-ZXRPFa>ghEEg%}EXHEzv<98%m+V@I90|>We5QZy0YL|pz?6DpsZ2M0m~a0=DYJ<%oQx4APY-10D-|nO3t7`5 zZJlQmwE`t+eVUO9uluxlx{hCEZ$A|XN_EGl~%Lv$&@*A-BTjSU{FgvUq&NCVZXp!w&vXsu%9G+=u|3eEFu zA9O-@D(c7~*iqpC@Q6;058d@`V~x?{Rl-AJEy|$-#e!(iy*{Yx)d5n=>!P(-@V`T{ zU)AE8?4PU7^}gkEj5w{>kah5>_gwt)7fs-cd;j}CCUv8M&xu`|i)Lo(1jT+xkq@Zo z9l~ww1&|}g|L;~DhKf!o_%#N44b^T4K@_aQ@C!&)VYP1}(+vJlS%e`X(?KIPhE*|9 znc^6z=tN>v91}WFL$Ed*#{=SJ-+FbUXFEU}+M(e$U5Z=|`Xab7gWpvUfXp3TprB{dTe)g(te$^n9Z`SSer}UxdA8FQz0{bY=dlJKrkT%@Tfu2zHEvlm z^BL?)2R#t+U0>)Nm6pI-Wn*-0!fCj!|+*lW+GWpS3K2aazP!DU7=zY4K?c8W42?5wjiy@%nh>ez@1} zhZ=G05K9S5h)M+t_j}Ae=0FfFgCf_Ohb{48Wg1EY_XMn* zl0PSZ&K8otAW7H-ddovDPp%;M_F=Meit&^GKMh3q_Md<8pV7bi-7lj5>sP=3t5DX=s!k*rp9(+T30{94h0AS6s5^ypL+d!1hh;a*4@#R&1% z39xlzn{}>p$o+m!tIgy)fB9*wz_cTJ{p_ZLWQ`MN51?snMD`$+VmDDFnTjsr4GB7F z<9|_bHIH5j1(VpbhRUf%!BrHLm;g{bnl5hOc=VLxe{uo`_1>57owwwfbXr%n!VDab zIVG#zu=1eC^&qbkTv44eCb*G-0yd2OU+2Y=n5eP{pi~O1QW-xk4crO4X9(APs7&=* zIZvnVhOKd{Bl%OfLFLKPVLWmiWH>=((k9)n?+ZcHDPbw>j-izI(oJvM;Ot zP<@YlG=WKR1(N*-Qf&Q6_wEQS3x~BP zpU$7LT=pE|K2B3M+ShaJ*Pc18)~9Tgpo{Zb`BLvy@3@-q)%05Z-^p@r>z5O!l(w0D zO_C`#i6RM9^ck`u1j7G`uDx`N*C9!IWL}tdzv{T8I;c8m0!lE=L1jC-&7X7h*-G)0 zB1=B6mn0Tvh5hRj`Bf2`u?|7aw53s4`m)ROv{y!8OmIy&s=|-o{opib&I$}sM*qKC zR@Yf6OH1Wt;x!=rbU2)!E{uh6^=5VNKUCHv#*+`1Htt-y^FC$Qy zFk+KkOR=DywU>$>7H5VHkeGRwB9Duja2b%>fAOm-(~@p9FBXY04rV z1|SFeF;IRk@H@+7D&jy0Z=Gua=*u>{H}LmHt&Pa#ZB<^AZTGke9O=t|gdNY?G#~=O ztYS9s@*`q}Cq#BKNMrsUSbQy>oN~wfXiscfr0uc{(Y+)qkd=Vm(M6Y`C$)3l3irx6 zdTBSU2jYZG{>s3Ch$iu7b~}BHJr4xzEnm5@D3c!}%=88LKoS)1Dt*07oatMp+T~cQ zIUcBtHpTY3=An;dEfwva-ak9BMX@5kw>ohBjJ}9Jb&w6DT(W^Yc|!I9`aM&lOpY8F)XHSJg6`k4T}-& zb83Ad75;g~H-2DRn-psUze!FwaY(n@Y_+*eu@@+Ej*9LOl)!3mK4^e+E7~;2Nim)0 zzh4zMJ)bUB^m;c-ue-r&BG?xhBRnH|f9gfIa!Ho}>)1|uo>W~4Zv(gM{K5y4Sl5>D z>{-j9z@?9O2GRo$=bVKaa2)42$Y|~ZLd0I#U0A_td8mAD{Irr+Ex*|ry}nY zv8AxH6-DH$3goG5;IHFv4A?Cmq!awJeCy_>D0=DJU%xDAl^y!#Wl2Ivb4cO@?m#$B zf=w^b9yf8D1gGUO-KTMX?Jbl2SyUGlMy_)6jh#1|M25|*)g6lcl!9<&^gByhfk*oT zCTEuRilk=Y{m>LBZ0MbZhN+2eQ|=6ZcgaD|ROxXh%nFaz$l`0s~ckvO(9)wto)$k6eSpTPgyl69Vgii1Aou8Dl(bBS!R2qKs| zketb&E8SPUvU$3;S(Y@dH2CjJYI&O?d%btL0YwN7f?|Aa?Efr5Re_-|Z+h*VNV`Jw ziTh{&(nNpYx-&e6TCzi!OP8s8=v_=^*axoKI`3>ARFl&=3meqgVKYkcEbAe=FbRN<4R{3zHc=Q!;0=zoJz$viFoY82oNj7sOJ6|5|+oiNw7~2`0Pa2pmr~L zewVr^VmqBfe?~X>EoaWM+oTQ5-C0KJ@h8_$LW6E1KUq=ZmIKA0t(qg>x-2;j;_17* z%OZ9Mb}BXrx2oDS7)mvclNVkPzF$>p7Xbm=QDK`SZMe&(d4Kjz+}ynPU(Hx(nLX!X zTb$UNvXWO%iYOE(@UbT^*Bct>umi5ft!NQ0WqQFmubtr)KIjcippD^%q-W2*JwN?? zIiW1mSXS~s?NV$j+$e1nr%E9=9p{rzH!v3^+OF_)6|_A1PjG6G%Nw=l8++bZoW7gR zn;bqY%FnbETXfom$BKL+cDI%Ic0w+#Szar`E+y?UvW-rFNKL&I3O#Wr)ggdp$%uoz zhoSdEu_-b|v4L!yj$KW6XQBEgL>?CIah!rM7Zi`S#yk&=WBr^HiYG}{?M!wv?VaAy zp7}F0BROqsf#jL(z&BHD3Pm;;NvRH#A-Ohp_U+e7gO5rfMC_UkP1!f+wlXM=Q{y|a z7LF4%#yZbqc#ab^CZ9;~Y_eQnoYsz5S@|-G2Vji)yrV910>-57=9Y>ulg*I-w>F)WyfpEakIXDg9mQ5tAj1@$L4OEDR++BG zfS68s(XBWfss;x=uyGzL%tKpbD}}e9#cA>4#p&Le!o_+zHc;=|Mb`@A*+Ma{Rv zsApaq4;DY3f8neDSe1FWpGo5B{8I%L3ywVZpt#-{X&`quPl>BUU=9-&ryg1RBp3KF6m&ki5GDv);Sh+NG=UQl}zDxL1-9 ze0=)yh~|)sA@42eWSZ#Xl3^&as#oWkyda%2FDwrTKpX@6aRJ`=C3p^W$M$jH!>D6X zCaV&#Z0!!R*NL|{XUxpq5sGc3$N?()d~m0HBd=5b=x2(eBJw=DNvczBbZ?#CropnA zl(1$&O-L`lC}PX(Ewd|`7A5x8K6;`al0}121Cllf*6;De0*UvuAHQ*FQ8yUJlrU6# z2C7!U<^@;}yfJF%6-XR1mC^@r|HaoKOV6Ak?`V_1IA-M70UV>*$;a2m1`W{j;}TA# z{_>TFKXfrcsp#{!-z72JJYVMx8WGSL7*Y9|NU`x0Swlr*LfWXds5j6}?NG*hjTr=E zog{qD766TB{INd!TEH{Ys+I1BD=a8HBnj-8520cpyZDA>X*72b1$ zeN2i1d##{dyAQOXmkLpO1U^JQJ3UF{XNB#3j>yO|?BY8glqqJ1b}2FUH#62$mnv+KsNJ-}#(H-W8qrm+|sJ{~_c zJ^()+#_k8406eMdn_1udi)j`6x2AttMDC2nvFW@xC+p1$rIt=+hbVHNivCcU$WH

    b&|ck*(94_M2SW9NT)J9NDIRh;zi=lz;@csHz3!0so`Td;OB6OPBvV*lUy>y@YfoC_! zG1_C1vgymo)ezV;s2y&y{=!G8?snd&NXFp74RcHu>h-Gy>qwy!``Jg#{Oo-c3sKB+ zDmo>sU{-s;4H-7gW%FugB2`F?@TL#!;-}yA!OzeZ^?L7hJp{z`wW1*xyuRLlgGa5X z#%;ZSgP&edGxLt&lbb$$61;0W{i!r_W=0@#cQylG^_2jq)U^X_Oha%*^|QTcJN{jq z_U&IhJ^j{;UJ?<<$8B1-Op80Oe7eUqMS=ZsTH_`xAtYIjWxR&;5=u;>xMZBqWvK1Z zqSJ)3fVF}isH;j*j5DcZeGf*3tL9B-#wggG5MRHe(CHK6D~CRKWWKQ zC&mTHwT%epB~fewMb=W$IqK`6-UnF*yhgZZ$1rpG;boCe@# zHj=*ogIjNymXZg&tbZl7PP~*rQ}&3Z;;q!Ls8R#<9sS|(% zr(2Z3U@u1(-83gAxKM(Ef1tAik^?AFl1HNhivl^h>Sp?`s!%oPj-6snbDG2OZggC0 zeEZ}zeC#b-&bG-@rs`hD^ntPLw?&40Z-X8yJ^K`&2R6+yKAAk$O^@w>dTc#79&p0Z z~bjVP}szUbvVQ0@dP^dfEzs2_k)vPbi(j0Gmp)zq6Q*59MgE#0hAWKaI$3ujMFYOR%%+%f53$Xf@djidTG1# zqNI)I$Q+@yS+Z^Vtv&6y>(-}#l$62b-Z#Fxl9-&?QkTq+Ajw8@&AyVGC^nfQNmMjS z(KYx1Baa-lShGE}NnYiC4;42+tI(w1TOgURO(VQN+a|#Dy>s0AerjY+6X3j-x$1Yn zUHhVMe^YqPsw>dGUM;u=WK@mvzNj`022_iHxi0ztSMHmj%L%a5swH##EOEglYUadc zlvd*LC}5Hn72~EwX;@>ck2XfwZD8d^(SeCHdYsJOQ>Mw07v6Lc2)#@zN&Bm3|3s2s zn)2FGvjs7SVzVeHxs9%;n+0{EVo|RIXg0#|462&~;$efEt4@Q`$v*kL&~5%z3(#e` zIk#C}7SZgA9!ui13D^R2S{+9?r;daOj$NA2{9O+dOy0am-Xcj;NQv3|&7xTF{nCtk z0PF+SqW*^V=G>&n;c(;{9d_BFDh}_U4Z1PdiMK;lBrW#e_bPlVx2iV3es%1X?l%7G z$NF*|D*+C+;pH&Ihmy{}_M#{23WY5|1*exj(S$~}w~1XUjA*tAtTqDQp0{xW>*v1j zRwzt%V^dPlIkIuIB1b27F3Qb}Tpq=Ow~kU}V^u7b6imsw-RlRWb zk@6nhGvjlgTHZ0z>DkHHJ4kCgavm*FR)5+RNzq^Xwtz7q=MNwBA0f-85C}4jc;)X? zY$8SCjZu-fSGELckz%=l9uR4Fi%Nur!n&Zn6PkX(f4|MI)2=V;`GavRqu4ZxY^I_yoZ71Frc(nm=o((3*r@xgZG}$2)WAE2 zOS?f6^e{OG>={U$J~5jen{htET09@68Pq#-B~deth@D9c=bIbk}lUV`W2tvdl>l z_o3FaA7q?$V4nLVt;z}?RA9G{COW5owoPx;U=cVCJMNn~*$U?kJm&fhg;+8^I&BQi zin=n&ej9j(cZOr$5P8#4Agn29!Wq;7RW{>w+z8}004KaozV*Ryh{)u9tWLY~8xrrt zjSxG`+@f@fg<$$7%;B#fC30xZrM0*CxgiJ*s4UD;1B()?mmVgq%0f47rvlv~Trwu8 z*u!RgjPP=}&Ggvr=7x{fuYMzJvI!r{R|YqdVQ!fiC*HoKo1tkv#jc~sYIL#-g~Kkp znKkqCcwK7kHSaa!qW~V;EZJsX#ya+Rp8LJK@2CHdy?23&>OAwuJ;FK6yg1B_fjK9j z0wV~Lgp0!pCS%&9>9%Rp-FCO#-L~7C>~8Hf?N04)((8x$V8_0Y4`Fn9aGjks99o8OXjmXao|X*1$OnqFUWDlT2RKq6knmV$~8~0U2+Z@ z`c;E;<5pjL7BY-Eb$oC%<^Xm&=<$BmOXiW%=L8uR0=~U;3FI69a+N&@>Be@^N1o?> zT0%C=2R+!y7xT0h6?U3@i^E2If8EFCj4?i_aNNBYu_y*%tb#cY<#bJol&~)P)0aE# zR_tW)$y?`aQ1(qMkNMY?sN(MVj<}!iOkYdV9e8_EW#UklQXB-x3#jM<@kVK#z8-rG zDF!{nOP~+>V1Cc^BlHlv$A1a4l$5w$lEsKuxZRX3^GEsTW;#RI!)D7m7*v$oAYT?7 z2NHTqsH^NPMY{jFAp4|ht+fjyp2irjW6t435AyOlsC?y$*uSsR2DBkNyxK+8lB*!M zS1G%}=Fh1QY*JLs%$gaeun$%&&#BcYvOI)c&J6$iJK5$NS%-~(S=gFlW9oc*iEv+l z7Q3w_NaMpMdwTVAW6d1fWfGq0tc~HbXcz;kFC*lD(t}p zQV;PU{ZTl%hFJ*|9r72&F=*2dHgsa4jc(-FoD}{ ziUSStG6>1bH-hh)0W}dvBYSA{O!|QbcF}8d*Z^oQ=cs|j0LL=ImdzQYt6U$5a;3<2 z)~U`&y1?>W32D_7i93`_{VHXB5u3wWG?~;#Utjj-f!9xIs$6dbZ?^TaXFHfZLWoV? z97+7_EwvFiN2A~SfRsD1Q&E*h^wCYbhAX8 z?1zKp&q1#!Nw~(fgFPG(J6pTWy%~h6+T_>|aDwZV?O^e^DVhjAcv+t*AC`IfL>WQk zcJD|E+2g<RPZ6YZk$(m1Bht#T(@UW;F zT6E)>nQr@LrOp}-MTMh2wh)SdfiqRdYzbQ@?PkVFjj|__jMy3%pm-r^D*NKAk45yK zDWM_@ar-7ku}=n6@U^J)hNVe?R5&EUfR)Kdn!M`Hq>_wyp=Dh2KE9hQqeZdoTdC%4 zPM(v!l!c}!STfQ75~dITk|@Zh_k}}4(NE)#UHSIBB*2Lm!Y9qbqM z-ucTp?0LL~?V^)}6I_~}GJ(7hWWG2(W$cNZ_L5#WXRd?yoZw`e$;y>RaT_U;j4Rim zVD&uwwztq42qn9I@EEY6Q9$#=wWWxE}`(FdD|1;a@s5nX0 zdq}PWJ1QTVO!qE|1A&oJ2=+61bb4T&Jk~i$oa?Sr6?pY1RygksKB8)djuVZdQe_Kl zZc+owgZp3Zopl&`MqZLVkX>QVy@F?weOoo_UMde>E;K#c${5btW>fv;obW}0xutmY z`O6_~^3L)?Ge)sD_0pQG<9{^bMk@bdB)RQO5=`0=#Z2WMQe=>d1{S<Bvxc3HD$g^x^F%$}dVGLCtW;178<;h&eW9SCE9rnLs?zxny*r4ug{c!}3ObnjASB+ecYgp$pQ^|Tb(L#6 z4a2H?!L__3x&u1h$$rbjxryCPl*Bm&)V_{wEClw32Nm(G)0+t1?n43<5LgL-E^H|v`^~nJ%1(-pBoM(RD6x9Z;Yq7)f8E2;85ajd}f6Y z?w%$Q1EZ(z>3A4@_2j*8+5E;|{f&6|U1>ui$##Yq$LI%HL2+diDWRed2A$)Ucpi}4 zdo@LX-4suHLpx~6zQ|k}_1%|U6{y^$uoojg zRaX0!D|;2~BxZ(ARU)_}+~&SrupVNlwx5%yKki5F!js;5Y6Ar6b@aG3#+%zZ6 zSzw!Ggk6w;%Fu4_tD&u+ZKy>JD?<~{*2ymOw|=u>^Lc;rfV@!QxQ@p>(Ls-Ay{xA~ za-LZ#PM2ycBq$GIo2f3N-DpZw-$fBcnf1;s6+ zNX+9F>&e5|{477b-h6xOPif;4)Bjz+V`S-blX1*2St-|3Tp~r*QqhO$cPpdbEejm! z+S&}eyF9vM?y&!4Bgkr#XLF1^Wx$p!9awqv9d-1lFl?=?&8>=q@E*(ZSY5<_SIp_WrW3 zrhPyE^~EvNWm%?Rp|HLBM5-yFih1 z&?E`y2S|77V-L)2fEY~jyh;^joV5dPl~F*U)3-D4^-6Tp76n$S^nn!&y9|2V77aUf zL7)bEBn`S@YaD$WPo1jH=b?8xeO!f%?JI2ljO7ZQYS<~8s#4*117mOBg2l;S(PE@0UD|`uomzS4%xu{XdSAdv?|d5f(2zzbU{HqV zDG7p!gXZxi&DtYR!V|A`YHHKYjyW$IA@)v{rj8UpV~U_hO<;F`;wmY!n~KIhpO_(n zP+YYhf7(rggAx1HRpDAxnoeP3gNFTU=__oKOPizzh%eeR^to4%hzB@Z;&T)t#|cbD zMCz&r&`p(4n(Ym))@&5YEol^y91r2QE~$kb%ZH zYXSqUm0DCCAlcCgOx_D2d0}nxLAQ*cYUb0I4+CqZE3}(2WD?NVX;Iy#zYt^!+Cm=4 z7Ejn-JaHn&hZFq@^u%!|deSmZ{_@?RKy&F7KDlWJMvw&sn2V8vY<(c`GwIBk<-s~t zkvNM>VKD(&C@mD*uj`hDIkX-!76!1>;h%=y`v-ISuEUm&SRmwHocD2XPap;Y4Yamg zcc8Up2vesTM+jPA+1iK|{B-TXG( z4s20YnB><7DDECbdeQGinIo zuuBp`ru?q5b7l%#4OgIV+`(@5j3ui;vTiuA3Xa1bX-J8_boGM8?f4#Fx2t0Y?Yg78?ozkYZUg?{6{l)=d20poabl9fcB7 ze1AO=aasqatymDJZDW&Wub!=g=-pwaQMg3WOR$?)nxa*692TP#Sg%$Jm%7|nek5Ar z(V@y-*U z1~rOSd3J`U(_NwjX%=%M7#qFe(F0MtU0XGq{d(y-VUtIUxMX&E1RihI;57vv1M-BB z{pw}GhO;^q9^Ef$_Nime2leX5jps&2Y}M3?%DwQ4%BTmpjYZA=2w z?ePI92m*&U>%GL^wh+@QheRV4A z|In={V5&oUB-fxotuwqubxW}a*bI|p$gCzxjjVN6&)5VnClX+0n9uVN9>L| zA)0x&I0BhdDB&>Zc2o}OsLH52a1cx5S|9|E)p*6;=uGFT z3;c0_%}`GbeQ;oaoh~UX-dGu)rAFl$4A$cn;I?mt%|C|xV`ne-hs}ZF6L&3+cw+jA z>7AlD#R-xmyzR9MbRl-aA5<3H8)-d_#x~G0!Pyf(jDeyt4~6|wcX+N5T0j3}UJ2RZ zOir52Ngc&iQ)E9C4G%5iO6XDb30bmVo^OxuIw49<=E7r7VRr>y1C|jUJrK~UF}&u| z$G8r5gYbb2j=)uS=k~+7zKF`G3S|$y(Yr;lXyC~MXGRij2gUPlp=?U$Nh*LCxh{|4>7XJYc zWKKmyLI^g_-bJaFP1WGXID1YLGtpFx-$8W$B50hst0Bv%eWfX6Ut*e(Ct%&4~Qx%>2a?0(BeWQ@=<^~+C`pUp5DmTwpR;+JHLGudbIi!7!% z20PZsey`me z>?gp@I%q#(06U#pwC7@;IVLQ$>A*-Ss+j8hwo5wc0iwmCugS~zjCuf$`Y?If3C9gx zeMkA)XD=9+760Ef*+%d5>WA~Kn%LPpy-+E>j#(kB5hmD*OE(b! z*%Ma$05q6)+{x8wYyMAr`j2E4zrC^JhJ(QC9vvv$L~$Ugv7U;?ApcgcF8YH2hy}7x zYCY%<%HXQ)7H{5dtg7eWk&4$~KzI;K%^80tpy{O*%aUjz{t*odY3 zB`|IBe4hsKc8N9zMC6aF`UEv(8|XOdAS;A*NJ&zi7i5LtUH?m+GY`|{UTyMrzwThj z{-m%=T{>st{aAQ{WAS&}c%fu$`(oWuo{-{YZl-*(_v}BLS4#6?wK#670@_WCHaCxI zsIlSVcIaVp7P^-}?By3%UeEm2jkmAG{-5M;8rP%>&j&vthxx5Z4h)h`6Of#txE6{uQPG!0E9S0YJCuD98RA0! zRY4^lgKlFTOJSScszPo<*Q5<@J0bbG*{^tZZdi#&h75K14g@Sww`$HXy&&h?D5?tS z6&6RdYCaNgl7r0aH}w51_UTc$VfrqQbFVy*?S8EX@}`%)F3B#+Zi;Fni>eZwebiW)1&2K^=Ezn7~LxO2C^)Y#)gU1X{=$`U~?4a*$x|zPhU`;D3KVOk;@Q+jIi&*jc4)*xW0r_oE>#PHo zOF9TPXm5A`HgeUNN^b%(-VyL}_&Y66o(TZ{!-a7KF&?9{K{r{1af)6BAr0v;F*Ao)TWjusn>-&+tOV z)LQM^jpoI$4vP#6<*mEEeo3|z`b6t{?}e2!7_!5>=8?$Nqyl=79Em&tI72jm41>bsslIcjCr8RfZ9{7-ush(bt~E|g*h2}wf+RSVQ0 zL}f2*F9n`poBl(uAP?vOK!iiN3iKE;N>Ve=*319Mqu@U)kCy$BHzPyY^p)4FIPT8i z?N|Qv)0d39>B>2OUP+3^5~p+EJ@ZEfvtHj4hy9jkz+CWa|p*J>v zJs4COeqinz@diPa>y}w*uo9i{JQ#rml__kV;3^bb4k>e`rStCuLzOh#sZ%{s z&uDXshkvk}WH@jJ8)^_nm8z9dTnR-Aspw_FcjbS9nr+SJU%v=3T+B&+7QR;4MCZ?F z7xmL8rOk9My-s*mk|1n@&CYuAfq$YDqG_`FSr7eJfodvVQ{~z|Z!3fEpk34aGq%b- zj=)9#Y7;D*jA)w=ys%6q=8A>{M#wk?I4vcy4h)%W6Ud}c+M;C3+uQty zZ!^Y~y(qRiVg3r%-!Mu&?!Psoy@q5DBpj*g(N!gh6F+xqc$Yz6t{&UsZ?~KN17Z%|A`)FnWQ-lveWVb zk~AlsPVv(=$hJ!grB^{m0kZ!X{BMW8)03sb&kqfg2MWt*7%ME7?vj#Vu3F%*89WPe z3XSsZl9G9!zC7}p!nSxIab&ya5|=91gUpe@52o#)@rqjkO$?Zlh=!L@kR19YM&Z#~ zdJ{9qrrASFXXHmdF0eduQ%6Rz{VOl9P!(}|e_$T>dQK>57NTU>V-u_ka+Gi>LI+jH z>6pCewE#{6b4)8C*ut#Yua<)<4+Ev*Y>+=kx1w zTp#a~dB3ju`(o1Hcp*z(;Ge?&YAdyP{lXRB`25G8F8Ece=CJfWNns20TH0qwqF<>Vn0LqA zF7=AJN7Ws3?LGHnPm>jh82d&$AJq=Wc}@fojvM;5ga`w-UAtDW-PpX+UOc9!%$ya7 z_*-=4m{NW^qQ{(0=CHdJ3+gIcJmX)y^kPkr4w@#R2wV!g+dDoiBkV@-<}hr;lESvh zdn1biw}&GdRzsLW2MX$+g(JItsW?WAMSvG056o@yY_K0=)fxbdtP5kkJ_TRnKFw>t zGF44}|2^|~sl&p>f|ABk?!K&>0|FS(Qn8)Isi-rw&FaE=&4X2#;5DLcF6M;^YGAu= zwRuY?K5m8sFDe$AJz;}neY2;L#>ZMTL4N~R!zuv%4c3*_o&Wt?^F*u#c0@jn&#v4H z``vW~X9r%T%cJgy^bd6K#WR8D=$asu>72xeYu!T~vC+2q#aS~7uV1C(rYmFSQ9d(` zp$&C>Tn*XEFA?j&=3>7|grtHF|N1(;+#p z?9#VCP4VzkzqZ)xum7~Ut%&=3bAOnIocpW-6OT~k3cJp8P!F~k(qhGB7G5>3R zwluO}x)Cp(uY_JAn;kggQfY#KB8r1tTrL&8#(C|uwa#mtH>>d9wbSqw3DXJ8ZLdP$ zsyB+R&0ELn{2Jx!gcUx8pK*_ot#B@)hvs%GN|kkTtn%!qPkA4qdldy5qVsDBX(t2lXXsRyp%Kw-4tvCX;?*2gER1zd z5@)Ce$O`8nr&#BbH}}33_eRCE(%>OsqTdSVZFH>j7S$zLr?N}D!nwj{mAFMw;Zq?! z3zcH&^aEL=!eKkv@&ln(dhnFF@pVi3MTd8?WTNlE{82=#+QT8{8Y|?(lf7 zJN&NOLsr}FzBjm?ZkpZ=6`)$^UMycKE}Oj=3(Hzmi|uE$j5cJvw#!NrH15kR!zM4V zDb>qpr&@ltdl5->U^}(PWK}DmI1ty$2ERsG=x+c*{+tfBW?q7@LSm@yDRk8y3@WB~ zk>$c%2)5MB+X8`@Hu<2N)^HC-T-Fh6lMGEzr`Zo4ev0%_(O3|Mjart?zaUvPH`~2Uz6VyTY*_~b zdRpzG=Ez#*T2Bo3^+f}bk)zWgopLR-o$&v}OQ+fvw%`9F@t(j6sBg=3+czuQU8lm9n_B$RT4l9l8PiXO zosiOdPnqbANAVViUHBThlm0pp+N=Zb)-2Fwad~S6y%UTTLsQNk_N*K|sp1GIiJnB( zwcdpAuwPFzIW|3C>mqfYh?0`yCe|$}DIE+#<&z!kmcUGDyQs>yC=ylL&oBu=X^J$( zNzw=nT5I77o!zs|UrrWF6+b&uwDFqRQHSURiaA0Lv58bCcqTK4GT|Z|4YS2+x6kD>c zbxji5wLF98#E*{A*cAr++?qM*FQ=P#>~Yv!l7;R(x$3m}>C#VMPGoheLxBm>L^eI( zF#W+-7vEnH_w|+q=>hihZO8f%^8m(Jx7+O)FJMqV{_*Ts^H!&P5=sudx>#t8S{zZq zEL^BlZFDVq^$@uTU7m_wO<@oDBnV@~6|hHY)#R&NR88_Caa+hIK9htg9$>Qfe#r!w zx$M_I2{tY--EZG0Cj}0?yc{-JUMeYWH$^I_XoJmYe64Xo0VtVg`ql@vaf8qt<7((I zh?L}tv)wz{!zA%*Rd1yH;@E=SuJ;MhnxJ|E-?h^q%NBZ=u5v9Edi$X%y2wgd`)G?*?sCZ9~^b7CL(nHK6hTe4>xvEw9F?;*MTw$cu~o%zrN zX}c&6JeyJ~8jCn^E3@dTSC(e|tDW@SS-`w@`ckf?pWF)Qq_f?3Fjt`|*tu6QOamJo zln!lWR?CN+iUNz}?KI}2KQKpC7#LzWzciDd& zoNLb0w?J^e{*_Czbr8|srCvI%SFtakT9gDQiA`KKA*q&Hupa7w`8LJuXb+=PdP%ap(ZDmeHwaNPg<-x;FXFRiK zTmb`Nw;c6Y7_=JUHZ%ug{T=4VI)1eI4bHgKxP8%oj%;*bH)WT};*n2ru;0m~qIY<; zi<13}{SuIc&?c{+36fnK!gd8}Z-rg}7S!qKWiLR&2m2~$&k1_nxq48rd*6$z5vmQ|?_gKjl$xx>hg74UxcG&3{DG3J^Q}kpfd2QiSAZXXf z+GCjDWnrdnynz= z%LLVuJiiX83_#m(-R%ZD0ekS+>{vc94%1bAKX z-ISkSAPdPler}2b(-2Bc+>}g;1LFT?OafKYjmi%KFdYm|FVG8Gxs++>?B>xZkzF2e zTy>a6RV4%MC4rf&I)HJZ$HI);^6)oadvU-#G5k35&4C?O=+`~UVckq|n<#kJ(Q(iX zsa3N>X%Mr&JFAW{2-@oe6@jHf6tN$2Dixxv{Un$U8BUQ$51KK+iIAe$(EJ09s^IDGMk^YjDax-Zd13e|HZ%j zeG%NSh2(DUG|wbqsj?*es5*6)HkXbKN~c>?+JumO>S_@X49*8_oZUljp|*P6Qe?O; z2?NnCzxXh`4bSA$hMNY!t}z6hl5uJnLdPwK2Lb&{}8hE=ury*cU> z*6^1>_teG^of>j-yQ=-Qh*zEDryOn7PQ!!^6GAVF_-fgE#J&do|D0iegRLCYXyeTTvzz%@I`e# z9c;Jakc+KSz&xhRN=)z=(KeSZIlZS%ZL}9hqu=|0lsj8eOy2~*v-?c-M4zIlWqyz8?-}ON!^w2VC!d{mGlV?%t#A(uQ&qB6oIBh>(rb5 zdg%-%IjTAY>(*kOJA6KUd6~-!XXvo*amo{P7ewV=k8W8OoFT0aNfOq;*WFU&&qvJw`fRu?PZP7ocuoMMWrUwv&NR22tabSzL+&}U5cI}@0yhU$y0UMLZ`JL-fC zJ<(#SC~&EG-%CAoF_i~2+k6mS&Q<4nftS2qTn+q+9%#f;7LIL$mXR#z%f2!UPlL>0 z54~*8S;-y_Q%*!X0b7@uT{TSt)u zD!M)p>2LZzw)iF^%-F!;ba*;lJ=u!e)=%^VXO8FXyv)f|Ho0q8pb-{x(%)Z4^7ut> z9XMv!U;?7u6jwo!GAcTQL4)!!>6D*RoYW_MR(;d3rhN!~&@-63f@?4K(C?RUdz6DL zRyo(omq0FPgYPaM6vW1@FfM=%`ioJy;RHDhWn>4K3PF;vc)?X0X9=u_*;7|5ohf!5X89S#&vR#5> zVmr-77u0IcrxT#vict}|x#kldHd;iO+Y8<`MrRUL@cX-EW~{`tC9tu2$Kbgehs8gfPJ5V_yClBZBJ;08pWgvebz9WRS%@8{u`wYpy85E zC;CkaVRu{xY7zp!Z|>;2Zl0BVJOkjsX2?ROvPrSdz1k1uPjl$wopmc3d3dgr(5U&jWr?IILxs`0%SnhTSXPFDnO_B+R&iW+jl z`yR-uT(blF;WdqGg!Vi*o=pQ5hKI>#9xQJ(Y(FuU=V6SDvEyz=gSYRltL@aaf3GD%hW2Ojo3wuD@ z6Fucl8a)w%Aae!LO9bJ*x`XWt?F$`r8+6mDzF@XXwAh-xMTHzIXj~!a0omrb=@p=w zf({P4OsTV=?jWR=E`&}Z7eadU%0hZ+GU#blMIaAC-5cJFWpu`s(l);$ z%fGMjoWb8@)HUhaH;nLGP$}vnb^PGxxIZj%$;9+FQQQfN9HXL_1Puff`e(gR=hH1~ z@*WiIlC_Jn1>5O7zw6T*z@!e+%`+RIjNqZL6_oH2{gwmuVY_F|%#*KTVKx4XBP?e+ zMFmVIc_15dx+FRoR4Tv{$wpDD<`B8y-6$Gzij(5MR|7NrlGTR*s4JZZB68@qkX6n& zHdg%~0oS_#{u2B=9>YB;m>P%SE{Y9GGvDAlENja`NGFr57aC*8XpnY8B0D1tDhdYJ z8hJ0>rPkv1Xs1sWFuJhc*`)Zx{E7=kEW$V@koPUT*5awJ|4*^Ac?R5J@nIpSyXS@8r^T!U|23>JNs`*oZ5IJqatw$O+tVSJOLXE%UH> zQ7?U2bW~9*e<*Cx-2LZ2J^a@TKi>GpDNXat4rQHpf!{7!kM|B~EosqIeE-_pW&e`( zl|QBY$K3^g8vIW4FKi<3U?ti{jE|MC6UU=dNn#;w#83Zs{f?2P4xGfyFhRt6ic6%( zT0jJFD8c7dZl1rw4M~&tl{yu^pM0BQWt+Q+PaEG6co~%`a+z-F3r0}*|E6jo+2p{W z0G*dng#_CuE|(%%pqVPq((K?exMNI#PdDjPp7idezkse)*!I0h+`~c*l57YVH>bW@ z@-`-&?zr{RtxOG*EJzR73e~Tl_^h`t$9y6df$#M5xt6wBAW&nh27p|C#4ek~K_a8N6QGRk^S7;XKDlZ&%8hLjGQcEXfjUjp#e+s)U zq8c`=*P!!6m+}f6gcN26J2KGn`EfIbo$~2qKN~@8B8FsaKf^Y#;f0zhH-7N%ZxADD z%AGPkCMz8nHL%2tT5~p1Trx%0QPEwYh?(;d=z8`+5Kh}9%JLg>YE)K8kkgPFG)ddt zk@@3;jFEdLT^&zy>U?aimdOWaV_Td@M{NqfdUL} zXrtqnIP&{_{6@KCxp_F(Le$jIrA$wQ0j>~yc0Hp$L}!Gu_6nrvfk_>98q(K{~1 zC%iaRQTUC zuL4rp+5%Vnub8?$qd`~;nQdg&VA2RqB`|e9o%8c&K*tj9HrWCU;>M5!Cg11d;Kq=T zgVW^p_-$AR{YRdRr8j&~ZPm43X{1ZcJBK)I2nIy|N9~+GptvN8#2fn1fciFM|B*Qb z?W|H~bt~@Di%|w~vfVGM=h>>+vE4=dOx*Q*xZ^}4KYhbm|>6))GT4Cxt2acmrWj6)Vu&Q}&Mr4j|&i?R(7n*{GA!s>G{V!yg79MwJov$!l~ z^*m7cR_=A#5OGkQ@47pxQ;tmYtH%4VdZqv6$|&^MGQu#(tUV0hc1_SHT0O6wu2yu) zW9DWsF&^4NX?ev8C4XxmdmAa0qm7!G27hJ z=eHx>)c8#;%mvB5$Q)9t?4fZC?Vx*SIB>T7w>~nSwnN+_=uPa65x9fpiIHD7BJR*H z&fg>r4!jp@H^EQ~#Whjn1QmVhjY~2tQ>VFy$}_X>AG>!S4}e?~>%Ciyi@DV#e6 z#ci#cbHV3>&T&gT2i%r0YsigI5LOj#V+sY!868ZmPOwwi06Vwcf&0{bf+Fu7kwZ?{ zu6D?&-8EBMJY&cyOOvI+_d`y7B!f#2h!NM)8{Bq+V%BEAM$v6A!ze>eJC#FD4T3b! z1N_^fOcqf18ZKU|A64`7Gw00Z(Hzz%vLJ zdfXOmb3f0F=+#Z)v*q{09QQNgeIvggro*nJmCKwbv_=jPqWWS2nPJXrFTX}D0yj}j4*f-MYDo$G|?VS0c zAX9*YpceGIiAH zT%aAfznxA6?L_DyX&l3#rkKKhQq3Gw6Q@;rSpdZZOx{7(qvnSr4eiQq@iqj?vD=->7I4YFKTFK%Q~1YWtMy!ofB3h z?h@@*_R0AKui+c9_HGIe~|fuMM< zHl&6?r-%;r#PkO;{hK|4is{2pesD*4nmm}lW;S%JNRxjgt`Obu=wQoSwH2ak^ML;# z?UWCacuyPf9L9^2QIN6lc4F|cY<%`L-MnGmJeH4J>%iEu&}ueGSQ0)!w8?%wH1;T2 zN(SAQh>`D-KPNA&3M3X6Hz_VcX=Ix`Lz?WDB+RBD?1x#%M7D=*2ee!hZ1u9^iZ(F} zO$;Yby89Q-3GY1XJwLmG^XHtHzCl>ydm-eqEK{BAw_JDy2$%Z9F4?l!*~^1t#9c})0L$xws~%ywURm;7I#-;)l)2uGfS~bAkTgTjre<29 zc$Mq0Q-kdK^d&REg+YgB2Xh9}>0RN6KooV@sm|x2H&B8DtFUMB8uwUF?RxhIvUd80 z$AqPX_#w`Mmtx`YL_q2L>xaP#BZiK?^{tC!{WIp`LP7PY{Z1CeL5<56D!R!7@^y+f zIW~-dtRmJKVT}~FN*+9zX018@xj)@|_Z4zuyty9^1ks3D?CNj@f+-uqR zy!)WL+`Q27nNn7@P?d>YNeZD=KoLDOx0u5J@};E$LxH`$QSjJ=SPP5CZnX0O%dj}S zD8ZXG+K!Fug3gi+W7!S|US*&hdX(+Rqc~9D&7h(?At8*dYLI-T_gr$+`_(%Q zzye)1!=;<-kso(ogp0{$YRhXSUu|7*cfq=d0{=^qCtg}2E(2AcE99z|PBm$Q%)F6c z*ZKc)aLY&-YlHJm7kS&WzQI`*yu%BF9hYP+nmupUK@&6-ioWYs&eW6CSvz5mGr8*t z4`h5fZIkeh7c!_#0TDu6KwVY5y08=>XWHST-oPX_Ov2TDvA6FVb?7fPnK ze(Q#pQ$O&){m_tdU1Se^OVQ*3-K-_` zfo0+{F+^-jq3N~GuhjFQESJ`<6X^WXM1|7Tky~ee91I(AjQ3&;Si9CEorb1) z0w3}3%coaS$e6^gB<0{p<PJ zGnX=aPU8E!P3+kt;X@R$%cmc>ot|F7E-IAXcIj1Ql0mj=R;s2qpqE|ap6u5w8B*d= zs40T>G|>D}nZ+$St8W9?54}uw`rNgr=FPZ`n%>0O&$6A2EX~mv7qiZ=8EIs{121N0 zO(y>s#T}u@K`Q!)tR`p^^hAOZ)r_z>7TaHBzYQO-J~S8l*f%M96v(IE=3Xd>p%$O` zh8C@WPK7=0;q#^MEk4K$&qYS|4UaV7Uu$u=krDvxTYtl24XD3p(K;B|)9?YStMln> zb*8Ttxz@GvMwdLnr=(DTBd>DR^7O1Gh7Z$q@X4R}VTsqGH|4eKXTx2MX!<|x=|7TH z&Lq!d(l$|C3Psja(FveVygO>+>?1(9#fFFlOs)E$2Qung6vcwIk@u8K?JGrU{v7zs zfcd@l{LE_x9G3#}_30N|%{zoT>~7ma&(AEbb7r3UH0gv6^=$t!Ny`RzXysZ?MMv;wFbnZW1lKtf> z`+ry>ze012xe;8>U^)9v<$&y=|4Hv7k!b-3L{(CJe^Q)3Nj0nEz>MvH@r02qW9IdL z>i+4oUjNhKwS(>j358Y6=HQIE)siET{VyMwfo+0#x)D!!g4Uk05&_l@XS3@pBVdg1 z9i{$#T!7f{Zj9z2L8ph_wEJNo0e=T1;(+CfB zd|VCL`i#M2zX?1_C=T$FPenI?fOkHFeGwDP zS;mWR&Gg6GP&=&mYQf>S3i1EP~Q%UQe1->47K3 z5B)p1;;<@il2#u6STLC~yM3^Xg|U{KYI)rjFQYwa`PuG8B-Me#D0@t7M*+pb7CoB+ zGCnr%O^}v*L4$hG#Qlf>IiEeQO7krAK%#W6cgI|vYEU0j-4E3K%d)i|x#}-KdAE$Y zugHPuXb!zZY){zm$I2+PVK=0JxC@`)Dzzs2y?_NqtbF_6zm|~${0MRm?6jOSLDq4K z`-mcksOUyzwWNC1HMf!17RW-jYRUt4`ERBcCsKPf$9p4ig$6dUB(T>X{I{e2uKa&|vjq?51TS%e< zFH0pR#v`5LwooJ$;v{P9DuU6nN)`4g!q6M=w!cKH7U|pVbT6s#H8xnZH`hrID4x(Y z{_MwrMV`0tgYJoUmkmjc2+0&}xIz;6Ek=%uedwG%I_{>UxXl#VL`7q<*r5AmqdEd0 z3A5Gv!i@lM3BNA=Ze8=;>HChUf(dji~oDo4;AL&R|^{q{2ih= zV~;yL!4B0`ue)=X&+k!yUQuUwJcl1sg0CU(8CN09_SiDmZSnt|qdr)Ezf*@5i7N=M zKVseskB>X-z_o-HI_F_+kWSsJ7v3JBv`mJDcnsmzAp5bKCgH%_>o;7T$eQB7kg%|( zpo#{h`rP_}e1L6tag9M09!zU*jm(3!20Lr!(b`*z15m;~DgFsev^4-R+m1{EJf=+9 ztWNu}c~xnS$!fiU;y$2A66*2|`=_v(zC+4e^yzS5GyB~UJ`|=w7IjZx)y!f4N&ewk z|3oL{K|cBtUW^BCtu;*JGWJ{u+_B= zXxQoWCD~Gk4^UIQcp+Pe%u%c#qui(ZEJIqEKn9dT_vhRtk){^KLa z+58PZ4BuSili+LIxGaBX`dX6iOsY(5NGZjExh{b4(`y$Oe7c}jgVnvaCFfqb3c7x! z-|StmPM9T!UdU0ql?BXE`c%}-z?@ka*6b%k%4EO$Zx)Abn17JjuLgP&JLqoymC<#GIhGV-z_;MaK$}CEUU8 zSg=H_UF(_`u-p4OXp6qV;A}pGT|Qfeb6~)aTHi35 zoCTGlE>g!Yy5_jpCznhtXA{MppvW;QI!0U*l&?NUbn1A{z;G<_D4UnU-WCC8`aCHV z?^9QcYQW5O(b_8@Xx%_M<>>(^nhlorvaFcC3>9lzpAy_g7v&jYbzoSh?YRlymqLdZw#pq#L;+y zbR1}V%rIJafFCGtl}Cn}Ya{RpxHxdc)`Dt+q4HUKEA#>oW-*MIEvSsr9#Hp^mBMOC zeNaEBO=<^7C8&lin}0-gLGlF*hdzl(+D1nPum|Dh8Y_NJa|v_cy9RgMup$6 zjO)_7iK|{B-42`=Tw&6&VSwW9QKT1rbW~F@^z}Iq)S@bso(}1h7X=oBY`#8=gB3c- z(#oht@Xd2z#i*I5Z6J6JmMNl_y^b7E?eyscg4}NJb`kdbXaOa-N+7Wqln49RLxBk( z4~17j{rsXu*px1xPG*|j_NY%xkC~ni4I{M2xXtr#ys&Dvw$u}QEw-_{y-#}MK#OwJ z9c<0K1Yv~)HV%3TkMrNGa{c(NnE6_)SGxd=hZJ_HxIhNiR?2R$S0vZin7K`g^FbSD zYnv3+A=NarhV=$74C-GsM z=XWrTN}gWA7{9)8{OGNjfib>`-yzFx>KKeR$6Ww+%!&0dFRXUh7~x72+x0oc^;4wJ zpyhI3sqGVF`nJhqKBabaVc{#hp;os27Mb7$W5F_S0$e z6Ip0e#MB3E3%vD0I_T7CH-ue~3^~=(%jOI@eXiOAi2oqqQV8_a9E29HF~Nm0=x~qM zr7`iEcuk-3K1mYZC)eFJvlqkRU}@CnTwFlTD>~Kn=|l7Jt|VbiP&2(zx-3|yN?;QG z&bePF+Cx4G!W8yHL6R^*8Y4zM-U8+jX$FM=yr|A+$muqX4RRia#tFd#M_$&6>Cn`~ zCrRkArPBO<6#sAe4;gc9AwF)E<3`?z1=XPo-l^)HUYXL7*A~sK1s{Di^R=GW4FtSQ z@<4XMyC)!1njd&qzCNP!KR;Vg3wSUT@VD5FLT)sDg-go6Knj685q7A-Rp*6r7=CtCJCe8B%Xf9m!0xqX=gD>(m4+dqt zSRiO1h0<-t<2qHcI?wN<;Ie;wol(v)21=#3ni3bv~$A$OAyClnJdO@QZCBEB5wREFs1=QhIxyA+^49ul*fz_$*ljUKp z8f-(|rA%^Z)vWcXrB{;#=C)U%pvtvTltagev7vB_1}pG}oO0<6(}$d(l@7=y=~Nhc z0v%}Y3Q2+5(1I4@4NrxW#c8!O&MbcK(@y;RwD+CNSduPG;XY6TD&cb1Jl6`MaQ3k2t z7s7U6mj@b|j}lt^km717NFhgGm1O#zkqyYO=i;?_sp>oQ7U)cH3Hptmso>;f`A`eM-?yH`BFHONaMdmh~y~=Tr%6$Y#%V!a9%bf|%gDAUzTzz7HxF_mq8X zg;%BkM@6HuifpITL7?xHqER%Y%x5MJQEM@0c{v*M%!IM{{+~*g`594EJp6;*Bx5Xp z$Z- z=`ABA{0xr+=Q)66FbYgn6t|Bed#LDAWs>kzRBZs}%}@aN7>V=mU{6JD2LYl&ai_8* z{HR*zH$b+CPDLTtaIGsU7>#fY3@NgT@H(H5RkeC8kzpr{d1Ji010n(!z%RNd!z0xp zNy25XP2AWRCen7uGMXL}e2o7?Z)y5}mCL_yUHc%)HLp)q_uCdZ@6P(+s~ zSPfjWM8Qt47EOsOz8bE$;2lFH*kA$D&zH?1VI%Pw%f@Cb&*zstfzIZNMbF9h%7VU8 zK3yGxx zE(?`ENy1dMp)vXzcZ^wMogPvzYUW{H3-fK=;WfhU@IvvTMTTV7<4?`@N5)}>Z2w1$ zF=EkoI}r|Z<4)OE1_+!ZX`AUfnz9C2;lGPco{0cgxK-1r*c`U7%s+!|rVGW;i8h5jDE~}ygmi&r zIv;f2uHqyk(=jfDjNEL;*9Wuy@5Psl=A?4YpI4G12j={MZpf$|@_vc~i?NG}#(Z;U zz-GTaU>q>wkFIF*SND7eGSB~Bp_CXvtTTsqo z9=!vc+lJThxK4HN<#lkTojw7%@E-bORk0w+<4`zun7td(5@NeGYFv02Z8J>Y+YBN6 z7NcBk+CKA|&d0g#4(z!?DbOfHJyHYcr2E)gKp8@2V5Zx?SrCfTsSVo1+rsM@lwgeK zGQx^v`vX4N@sxhukEbF6Q22a!0@EuYl1;7{RzhJZo zI2`jfDe(V-*$%9U6CUjX$Rt1yY+yCUhR`_u0!h#RR7@T|i~PKB0HMo=UX1G`+jk>xgoR zXP>esuur*5xtCrEqT+Ylwa1w&p(*S~E+yf^P6bQ`u%+=9U;*mGcZMdDSUF=4KD>

    0%6`J+q4TfHFLSoaD_xMc?} zojo~Ak=Zl0W{%BnnUs(GjL458WV_9oFMN!M1ADy|crvj<^lqSW$!>34iCVCo2`)T_ zGKoosk~JVQEGUzLokw31pDFmyzb|DYTvyWR&Ef+he@3Ilks0S?+@eZ~i5 z)pLH#nKLpOemk!D^?c#-m1n(&Xn}0l2}5U>UE8M(J0(C1z0Pn$e?MTg%|D{F z7c?#5iU(vZ4Plq#ydXQp^R2imUpE?(L%%qGlQcXxxt(?sd)Pv8K(jnSMYoe$dZ!GV z+wAaa4p}1XRh(3R0^O^JoMOF)ryn78%t2`_Jph6{rGi=-&x{;d3t}uwn4{_>m$kwg zrcv0c*(f~=I?soJb=gj0W-MoR&LKBCGHo_i&E-Y)45uRN#$KN!&|>VPX+ zHD~3`l1A8z?*yLaMrn<(#&^9-J=yQ^0e`L?4-lFdOTHiYBK3R!8?aw#gjM>CU;mDz zJU6haFo9Jr#er*^PDP&-od?E#`E=y`_j;XYn(1|s?V#M#7dq(nv8o>`k(y^#fl9z3 zpKSM5%~h`=@o`BpY?;o_!Xx(fbjOVoz7x;$`r>~L**4vXlg?K{FOkjsaN@X2E~zxZ zNfE_uqe!lSsEg9(2ZM^~EiPL;44v1C0=I`(xneOk3ZEli1(!PP6PxUp5tcD61?b*T z1@6)*FIAp+Nv8r9rX9tA$Bh<)S>cHR>n~s6^e3sa5iOd|ceG^5bAy(26SSqn-ppAUtU-|r$sUo5^!}Ydz*_fro=i-jB%r#5B>Ay+Nu9w#6;Uaews&ijz#Eq zU^~%lGV=`-S4RQKKRPe01GL(%g_dFTz8O>nkXBtB0g7(d;%1Rw8;nE3I#{i*`PAA%^m5KiGH7Z7m3Tg%7S|%Vbm6=_TEKufhL#0 z;8KLAwMAhK1Z$O6O(dse9a#7}AtTq1A5oTl->4#q#NyzjJ(gE10@ z)aA3)GylGFJGts}Tpx6mPGPSwtHE_Go!`dk)y*E*0z46+V?w zh^dN*e0t-npM%yA4%P`3jRR1PkG;RvdNe6|y^ez%Dm2PSfle$juKG9mf_%Y;dWgli zuA(q;P$X_rTw?bJ!dFAElee#AGfa(zvr)&aN0Z^o(k+^LxzW>++>!j2Bsnl|VY|r^ zl|gZkKun{edje8?uep_j+H^X7Tvg7j6Y5^vI2)R;)8!z)a9joTU~gQK)rxj18@zXs zT2ZyAQN()!`94^@ems*tp6}s%55JYFdE+-`m}^n9t z^9nlvwawXr^USdSLw|_C>CYud>zHZ=BXG5m8~pq9!rBkK@2Km^HX84$5BdVCpA(pa z%z)iX6%xQxHAEp?LuW%=rut-i=y0h%Rh^lkOE&$K^r z4y?n(6Nj+NNw?bfXP9S;a!u^{Mv6=5GVL~P#6|yi{f?2P z4(ys{m;iP?#U)Z?Efsy4?I*YCE{KK!$5r3iF%R^=b3h@(=DXIX!Rfg`>(m3k-ud4@ zHiF{E`?tJF&N(nB224P4o#MJE(m_Q-^c}*kTfMsI+tZWDHMcDuU2}$%`bhlI$PRXk zyv%jLOqxp}5muZ~x$-JMtJnMbW^927N zd+!1l)p_QRdxSI0yg1B_fjK9DA_HU)BNvAVno*mkY14MwZQ5<~Y4@LPD%)M#q}%RJ z)9co|pmI^c3+OOjU<45ryrF`CB1WTv0UgCFGJ>%f1Vus=|IfpWh8dYS2N-@wxA7C1 za~(YI_dM_OK9}#eKI#~~YHA048tCNaP}i9`)IC)`NV#2w$VkuBZhzz!?4a?y|ABIa ze=hT$Kl8-Q3dKEDBV$orga32r`_CpAbbQjTZx0PCu73~LGdW>}QX6)?9b+?0sz2)2 zlkGehv#sI0)=|IpZK(6oO&}^u;T8NC0q=;lctA(%z(9qU2Pwx~jnNlS# zi3RRK___i~*U))chK3;LHis)>9M9*3pi!pl9XB02;-9M~L??akb>e2WT5&ZdE#!y@ zin5!TQ+^EzM+u~mpuBFES0+=ZY-YZo@A)oM!eNV|j9n13Q+QihDX35#k`_tv#JoPH zc(*zDLj{oIzz^59a{#rs{OmnPxv!_e!e8g!nKl{nm}UX5X@&L#-Nr1BH>i{Ra%0T8 zBLBHwO=2i^qn8RA8Du&&s8ONJtXmV<=9?SaL9b!&un$FL-f3eOFc$GcQLXY}7@lzl zxTU(Io0&DjGO#(P11~EU_*DZr?{Z+8YKd+SC>J2lN169!#XdO(^u9drl=lh{CK%Lz zLj{y`oAmg|pnk2dP&3%QBL1#pu9{1fm%s4^;v}<;ViZl{a%lsb7h{1s2XvEn2B2g{ zJ}}BRK?u*Jg^uVET&QtG&EOy~h^sEWOJbk*=GJm)T1X*ulA>Nf>Qt22y(_*1&8Cll z@vB@|*+;!x7ibZ$3u+7AE;I_P=U9&$eM*F#!W=q3tRM^%()po{VJmze$ZzOU{Y(7! zPP8Zr!tz6Fl^a=ux&?Yxdp*|pH->>yQ}h=4Xsj9NISjt%G|7IsP)C9C0cGCS``|qW zVPH7I((MoX=JR6+rxQclObEo5Q@?q%x8{-0JRlY;fxB-0Pzhdzc~>RAR2S z6Ee(o%H6Pxhq;WtCkKnVIM}{i_W2Ugg_eFLZi>!%%lRpYJnt) zs#6w+f2{8Dt@Aq@b8JF~|Dk}{b;=4wuL_go>7vi%*P^fu$4pXUPzVSMK5r}0szKfx z)Pk`1a}a2If&;h0?#Z`s|M2VLw-z}Pj=dl=mlJw2RBCEwE{A19k4mPsq6S=;!k7ns zNff}uFNMVlhbD;lK<}fqRSR43A9>)Pp+LZu!5K76T)FuW&Uv%=>W$ZI_U30d&t#MR z1J!i$*kd-^S(W1yTSJi}RN^UFa{Qdg8{fDhKS8QxwMvT$xc;YL(mFe&gT~g9{q$WZ zfJyhN*QLt3eQ?i3GS6#o!uEvIvOM9U$z|SG1LjIDg3weBy$=)xETDpl`X>*jwgq8J zOS)H+bhmt)a19(^#^!55;sU<&C`p<)XDrBG_?`k`j75`&fU7}6#-%B3@R)Hvp^f>S zPuLG^U}BbrzeP?D1STF&QEhf$xQZr(0GFo6RlS@=K`%13y{;!5a3e$ z1iLf5%)9PuAU_v!Sewjj)s)8O(uW`qr^oh#&Dzq~W(K4x`cBl(^*6@!-+??InBUzQ zaLN0T2kKCl(wiYJ1Yx7IF=u1&OZEh22j#&HKO)r;jUY{q0cQNusmcbL5WQ6e71jzPbLV`@Y^aX~XzVVTHC$ znJTML6>EAu3W6^OY@vHYGoT(XTXr{Y3!SWNAN$N9f7YXS@i=GXxOm^8p7u)I?$@4o zqL1U+wO)BKmBC5gdW^wf^T0?AtKH*wWsL1E7@-a4)?t7J)##sAC#MN)bJA}<>N-Z| zKWB5&2D>?FCB>#wB!x=Eb&H9lng1t@$^%5ZI4TsaP3yZdKM>4 zP%FQE_um}b>v@X_Ck^vu-c3;4-|d4@FKipK6}Fh@)DXR4qIDlmB6FaYMSx^A*49^j zhdeXJx+rm{MR5ry&K48HiVXXkal*;y+Jx|{6K#IRjBEMJNCA&49gf-g6#FT5F9l+- zi51WSjg+U>kgQc^9pttPWY|#F*9?TA+U>$i((SSoS%D;{?|)z~%As>WG#H6gt(2J{Xn5WoioK1KNgbjv5kp1_YHo2nM^w2|S}pHAd4= z*ywm1QgOmYhb?EwYl_Oz_oA%VNZOSXp!etIR?t+Nt&6qSa;4!{fF0O zd^|QN{V#Uo=f#tLcGcH5=e&1`%p%J@$#y%_zJX#56v?I%TR=6gn=Fn&vcEjKhbfX; zt7ntEOF`9TPatMOP<9!$drY8R*ePrcI5z~vTi5^9oF6VeWY=#$=rcK4aQ>2By#3kO z(wA%)346b67Fok%zhsXcc(+h2^g?Z<5*G)q3QLofd4H1wvoLhzd5M|R=sC~ z%A)F^9I7_ery(5It|f;zU~B0^J;K zOdl;Wu6mIl?2Pod3jEN=a&~>qNf=yPAJC|nr(Eg_nL1EgJ)kL@_Q+#@cqVf|TcyF` zsRPqLfxgVN5RiKgtqj&@Da!j)0&Yca1vfVb?uV6Du!O>EuyAUdur0QU#;x|TABpe| zg9O}=(PT42U-iT7Xd3d*Hb>Dr$*8~pwK z2`tX~(Tk;1Yzjq|P>C2F$xpuUt@ z*G>xKx}$StF#kaWwb$c;U{f4^8d=b|f)?96XNNL(Z#cIVt>eWdj`1@tHPSrBjFT{6 zl4qk=4e07-36g<3J=d>O*e>dsYK*y|TkL6qvinS?$G1GHPFW}@2u2;UB+1f+5?uy?*{D_6!b6xS2~794QlJ1sv|N&QrHDSi(__^MLt_-3-pj8l}55(*Kkcc zLk5sNoQ4fioB%}4pA;VL7&qf&toW;r=_LGTp&O%8qYg!6>COQ6QVxBV)B(-D9;u&l zwI=9ZKLfm@O}>|**SW}NNQC=bEChEnak=~8&mCLa^8QeoLwUULanc-`3`MbK-TIJ| zkryLY`OJ}YkgIa+20SDALR=}@AjBhPT_%H#!xh@|u=JUAP`ow%#*|rS08Q?iu52U7 z9VU;h%-|v5%9SV0K7SF}=Aj8||M?hMgG5gUr050usy}%9Vcip=6;>i8}FHd@l-eX*}ITAA}6^ z$}yL`H$!L#d4Et(-d5dxk)TJ`0ZY475h_aGoK!=13Gg!O?N5<((Z6$gkse#65%}4d z_p3s=4S}k^yeWa=ITBnvgTuivS_&tWHteo+?}l@1{7I^N zOKh_qGtR9(OqOt)zj$m{K>Dy>MmdXOS5qVd?Mek$7MyjgPTDB~jiVLf8rs_HVG7O# zin;lst`W=`ae|H8A~R1z@Nmt$2PHOe{I+DwAIP#71{?);;8;hoYjGu)*r>3Y4sdAD z*=Q0k_Orr5zs^5pVwL8cz(Uu+WP|0BBNg-=giOrSL+Oq^;~)L+d*L>C)XZE_N;Y$Y z2Yl%)I_0w=$DClTO~{wSg0C>ns-$|y*F-;1bb+U9 zZA^dQlgvOQxK7+P$wZsgUkLS$ijTY=dCVE3uK^GD9?+aT^4Q2;SMArP_zx$WAV;h6 zELU~d5VURgf9I%eu+g1(_aNCYki*Jja2&D&$4-g`-Kniq;#N(eVl%iOdMgg#LxC)l zIT*1hW}XD9SGG!)`RyjPs#%LBU)JW(E!tGUtjf@|khBns7*DEI&T0}bnw(2thKv)Y znaaF#WwU_zVnz5FdhM{fw!@2zVFSiju#aAg=xKATT;XkvwUa89-EZ8PUN_^~^v6Fl znGLU{zxEJf`Z(HY%<7j%(pBZ8J~n(q?82R#JF5d^h^Q!~=y%>WYw{=k+5aXqE z3MXic_I`Ar)e#ymNbbi8kQTmd6zq`ffELlU6O1t+;?We-44J`IL2Z47v$HYgbx%2h zNa6$}ccq*MK;PQ8zeyZxV?ivk-+D2PVwX`QnMy>8`9=i>?xB|@k2Ev8{y{?3|A}&Y zCwZGrrPjRn#yHzNqD@IYMAma#%o{K#Q3E zU>&~Ce+`~;!m-MY%jCjiycF8u#Xzyy6j?(hUh@Wy&t*`qvT^*5@!jJx$3RvySB@Ky z?J1C4Bc@<|lXsK1{{FaYQ>}$ipN&8eZIA#NJ|sA~Q_-6T11Xz9dFcD_6jY-V$ZYg` z6H_U6DMgZ~#65xOg4#X?yPblg5E^jPtK{SteX{%6Px8sBBl>~U48z~JK~L>*onwm@ z7mLB)f0?u9t*v37y=qX``+fE*R=rtOqyC>@``O4I!Kiw_cSo@Ht+75f8&UekTW82p z9xq<=?TlA8#jc^?xh6J6Z;M4b`K<6e^jzrE!}I)xP~eO{$6x0E#dmg;I$Du1xpnACK`{n zNpDO6A%de`7DY*H3;mBI{=2_R{Q0lnP5i^p-~X>)C(3EogHY4}v9Hk^UM(LVXEPsf z@7-!7C7z_t?kzn`u@w|4qY_(%J0n_2lla>AuE;y(waO#HZm9da!#oVzL@;N*ck0ba zSWax#RYznpr!;3YS&9?nq`DAlcGgE;4y%a2F|j+k1GKV0j0EoS=j#lv8Fs~g5`dSj zj50>fkt~QT*DPVpx@|9K`QMy~(Qhoz7_!CZ{5IRY+RitI`lfSwoB3NP`rhpic1^RP zsQ$+ba!IKtIcJBWk0|yiMGjMm2ejK_8x;_sT*-XllLu?Xrs#Ciry!C?F9Lnp6c#Lh zmY_qFBmnM2dR0&kObMn5$NkSsA?wxG25~glBF+Wo!?Up`W3j>twQOc%Z^SyUGk(PZ z7zTah(V^@SW-^!K)4e)CWI%5%@M)vVh0TKN(dIs)yg_=&?!r=klnn+~oUp{-wd2G8 zosjORa>~WWZ3r)mzV@y zlYk|QsTH?Lb0tO6&tl8QZ;kI1*2kikl`3mceiWGx7UX(BeQX8YGj)By62VD#CLdhH zij$Y$eW>AvmA7lJk9BPMdb-?|$1{zS#x83iS|($a21G?aaE+`rjJR~FA{7z(5-I!{ zzk2*XT9Tb(M!9b0=pdWFYnc{(g=BkzxOxA;>{g0};?Z1`7TFZ`D9rkIy{-YeqRt6A z=rakm;zhn^gE~Yl+Ot8si75`xSlah@XPmX1qj~b9Fte^k)gd<+Uhs45{Nm!#@;GYiq=D@(b0mvG^j&dBHL0?e*rQ&M^}#Nk z(l89^A;(3=h_E3)#ycl`KfY)Lettaq-|v6JW=1|}$oc^}>`AWKnUPN@ww8i*Lt-ZA zNAC%2X6nQxkRpSRnC@94DAfR^Rk{cwVN)PBJ`KtRTcE7Cjs6frsNDp(Su+_}HIH3@ zYhbJ{G=cJTpT)OP3LK-$fvOYT2MDZ*;qGzOakSreWkt5x9_v_Y?Lt;!-k1UWj? zD%Rhdv>Ye_@#IbrBHrd-sj(jT&{YG)X*isC3Y>PkIFy{s4Sy&9tnW`{g*H@qgnP^* z3wfLo*klLyOp0AWku)mtBhftHzG!X^4Po7u=vJVfEDpfNiZo`MeAS3iPFj(}0kAGY zZFo-O23XzuW3MnaT&!Lmah|N=R-VG+X=s-nit;EHU}c1eHGSN_PT3TVbQyaEg)e_0 zy%q(WD_Rsy$z;BewkoS)npypRZMpamN%Ed2nXCJF+(l9#DHpatEsMLkx;Vhf(cBFW zew^Ipt#5S(I`-`FHfiLf3&*VP3SSPZ8*Jl68?sqzR__V4h|l}y(zRX&bpgc7=97)k zt=r;mLV_=!GS>nB@=Eu9;$%lgc?(Z`INoMQ7W_778Oh;hNBBE)LYAxl8g4tqLP|29 zN-Txq`Z-V$YgPHs<7N&Jl+OdrCh%-gHu8}NGP&hy%V~7QT7WvzdIh*@D2v@Fg+eET z8rO2C{FBFkFJ_{Tdb#Q^xjgt>@#Sb$9&fw+g)W}h@#e!k$AB!C1rU!B<3vt!WmJ>+ zq`C)0s`XYmm}UkYb`*sf_ECSGYkd1Sjg@G~=b! zP`TP#8g&Zvc`-w$IOAQ!mJU_ADOyX|JMPbs#ABF$9dTz_NCJl6c${P1MoT5+;( zdw_n)nC-q5bgeiwYMyAX;$g(4DE&ofDy)tu7xsFT3#%hG(>urR5Lv{vvAF@WQK#_Z zK4si;D8p)I76x7h+|H3am;kxPa{7p1j%4T9RKe_2S#bc!6<#DJ=pM;rN>q82K3AI` z27F<3r?6T86CbjuGz+RC9!_khF+O2d?;M-wWuYI2!Qtg`b--eGE4nzqLgPvLP9c_+ z?+EK;s|6t9-72UQG%}sS`(!r!*WZ@9YUUnKa&GZl^?J0E{IKP+J&GLAnd%Xyib^!K%JS*ubd#baphdMc zzRddwY&B25KN+{oy7Zv*pj8uBO*{?@xVe&AX>xq4yhUu*W%)y0bZD)(%sW-k9$-)} zd1*a&HIR`4Pd{9^^wGJ&eg6mXiycdWpH5Nn*it!BYd{K0o1blMnL?+!|9kGza~p<} zT!O?fd=J-KIEO^dh40n;#%5{$tM0uC z=_T(@VT%?c)fQ+l!g%?as54Q=RmY)&@y3*z@Hvuu(fW(ffQNHdIlcO|&jW#rA*wi1 zkDK_R0;pY@LA0@p)GKd?Bkkqg&`Lp?u;69AA;73jm9>%$!iP{_mL*sjRW0A{zcECg zE?VxtD;R!Y2@hsd^(DYLhGnVHjZeDCTFv~474#+V9o~ARYTW__ZDy!r#fsJ(+7f%o zJ1^#u$ExugWOo%S!_g(h`ygHT!j~6B)vIVx7f^8zGl05V$d1Eyz&McW!|ONXw4$ zYL~p~s0+bcSU6F7XcAxjZsqhVa+3gj6?&ttR#`(A(Rm@c^jxnF`4ZjQ37cQ)0z&^J z5J${~7dS0sfnepBkpPUV7_swqhYA1}uIGdi>Tj!Bf9aUAdfJ)c@%-c@jfG_Gm!xgJ zdSs_OB!Ef~k2#WjZLN51M0r%kKRK*mTzITQI)xKfhV)I`E_RNU##3H99cli#1P*z; zs@h|>s(PZ%o2NYx_J)}1!?F6}xfSk0#oj-x=5Pmyv>jmf1XB z4;`}u#D0q1OOahvqA})@GzI)a^cby`Ga$qW;-^WH6U18G(HgKS7+NDPfm6CXuADwO zCRvOpBl~Tm0>wohdEE2PWN?2rFom5Q(+$<_ud1#>s)P-X$&SSGA@4o!g| zHx>hBGMjvN3a$r0F;P8G=57SB5eNpxWlNfwavGHMK=-qeDUcw6gFF43Az^4hR1VKh zdSl1vf3_jTNcDW1oZ^NQ9?wXBw?oQhifyEbnMyPVw@&Vyc3=4TXT5F})j9o8V$1a3 zANfw>_ZoD_^nWX|R%{5VkJ=Y`Oug8%CH6Q|CTbshPL(6Fyciim9O&d*1Lmug(;fsgpd-x(>RD%o)4>r^Pc*% zEaAuw$eg9vPbgALC1OXy5z)qwJmF2U%nx~$my4Pql!AFR5Ud$!x9EDNR!_elU#m@H+L(i4%-j`D?^PX{ zeqLvo(If15a=UK-5^$Na!pq>udAMSgVtyo^Sq$6T;pegGJwiOkC4@W~tr3LElkarp zgY)@pfORx%^IcI3VdUBg#|8VhJ^JY2{ zSvg^V!ggMh_X^(!a`ZWopdoXN_5Cax*W=R$3BD=~o`Ze59rDbWEXhZrRbgusc@wij z4C?cw6r#5l8snwzBq$u(gLdD-h0`0%-|)w#=mTn{ZO%IW?eAPBD+XG7@!0j*VKvkuJ(uz7oSl6j&IvTRJ1ILXuMnB0#)rdsH^2z;H?sDly9 zMk4SxI7AP-rFiOw&Jq2yjLMCURGLr6_;?HwCxn@M0VIF9&Xo3H49owQ?0U&kCUpH+Y2Y``aD+Sa|DoIO$BWjH`|)mK_Xl4>}DUzF0SuKB1U7 zLDE2q5TkI#>>c1+!?TWq7m8;gCtDSWvG(|&U?HBfDyZAXB7O+XQngBxo5aM20RzK5 z{C)T1G%->Ss(zOAUE4Z$=HbqF$Z;OWW?Svnv==BA2-)hX#MG$e(KWPjV!kjt(m1(I zdJx(wYNg1&Y1SD74#a0i-VrTE?JhNqb+4r`5tz|l91M&r*iC6j5Dyn7Im^*^yvV&tk4Z5p95dlL; ztzx0)hJafPn4pM7o;CEwamh-vuK)d(LLuM5v7KBxe1`#C&$2!Jj=PRSZs5}07@zOR zIm5-d;c;r)2^&odWTWfV&3(O?*k6Dq$*P`;vP_h1$UlX(xHPDmR9SNWWRdFUfa2R(0$2}T?D z9g0}x0~$zh#H>5;O{@bh@H@`E%ZTI1JU*Q>avopk=AZ1Ds`ng4p?SN2av}h|MhgUN z(mZ-$Od0zHbRD)seaYpC^T=kpRaCCYi_MS4{Uhaj;DB)&Z8wi_|Jysa69WWqski?7 zKmX=f(aptM;juUCq}Cf7fA$3S_!wo^NM6h)-$x$U$7+I-I%K3DNd$31QU4(xfzNpn zv%fF5xhbl@y-s*2@=n|%kZC|S<(J$M@Tq%mKh<<>OCqbbB7Xm z{SNOIpuot9v?G=n{$FpvJG%5NM?AL%{wSAsiNDi|r(&1%{l~HUZPh{KQ*P4=j~BB^ zcFplU6nmc{cR^K4+#b1`tO`0J+yMCQJPMX5poKr(R-iOt}GFBaMm+ z(|SFyrT!S%uPTE*P3lk8W!_!kk09j#iMlnQ*P}F{Fb=gXsx+yB96CR&QL*slkAYSR zNz+Ox{8JrK=6!JT0(M8rq8u^=*e8<_79I=^sK1w&T(Xbc0!%JVfG7{H$e*r1(LsK}ATN7CtWuDt;i6**g^RuAu~l*saJfjN;cAoesPlyW%w?T}qi`1uyTNii8WO zYy$BBRk7e0U7;$5(v}<=GXk~J&nKo!P;V~ZuULR*Rf{U1Ub|k48)j+a%T*t7_4l4x|3x;Jgl~@nCokI~h z^yia-|530DM(kk19`8134G{Q1FQOW^+UcEitFqRs6BKYvu|2_{pLLX!i+6a}1e^md z$xme08Fx;oLmYlCBg*MqZgc0i=H8|qBe}ed@;iymVvi{D_2$c=aUMRcjYw8*5~fC_ zzr1O(yIH=U9vYW%@${u`-NOkNqmD|?XF4|h@D>+N8hntS#YCG0eNw|n=M3rKb}y8T zJUBTorZV`E2MU*ObE9e2XK;FBkuo z$H81DRK@co25pG|aD;sGJ-$dG+@|bdN@bRDnapxwN=#YofxbNk^|sjJ$l1s;jI6*| znt&%-NG=f3H8V)%(KnNY7`^67$|CB-C)nBZBpu>y!C=F@XCqTXuGY)~+3IGkiQXzX zM0N)rhv?Nh^&!#%i+TI+g5m}fJwLL8#?s*&x<~lnRs2;-;w;d3nJy~uo;~}XcXQtz zYrTNJ3aUwGqh{(+jR`^o<-#VQQC=AFD69pvVS0p1reNZwFG@COoI2umIJe;)=osZ` zcI*}4t$A`%{ht?_N8gzEIlEP|B%u|&R+C>FJpO}|VJydvDcOGwu};Ew7}DtYknbtT9HvC*gBNo>zrx0fb&tVoWraHv)LY#5-X=AplPQ5Z>284_07)iH+6L3fU zjs3=79gE|hD~b#!m3;Yv;z;035$#P#(ruGxOM5*^=zC*p$udz7UFN+`)1`v04rocu z&@K>VG8sNc$;ene?x-1={Eu_g^n6g%F0;8E8^o)wkyLIuA^z595Rh8y=b{=YHk%@A zsKmq5(?E<7WSJoKZ&9F_Flq=R18OsqM6Sq@V*~YY+r>@Z_-AC0!>vlY&*HexKk0wx z*QSP^a->q>vWDUBBlB@(!gNij&1`i2Ou2|`9H=>r$75e%H}=I83q<{eRN`F)=vN6A$JB}Q zAo5uffJE7c1x1kVnOL4$Igf1Y-guruyx@MRe`C#QR52OW>^B9)+eoh87@ zCQn&xHA1==4n%|sGn&P;S;jOpoWQAp-uMwkH zzMzIS11HeQn0wKtu{Gi7rhXQSdt0<4=Av4MjtdNif3Th3(K!qT=v@OkO&@KmV$Ii{NHU#`i5xFvA$NpHQh$freDudIrA ze13LJT^~zzeptP-$0tARK-g#U?&xdL*)i9k#~iP|9lpqC5bngVjNO-`=;RakA>=f6 z{B;Yw4}LV4vVrQM??cc5^x!Rb8SshrcVAN}b}2=Ys6^EH!vBHhVD8i$h{6v7X50br z)DZc#r0I}~~UnRN4K z58NmiV_WFA9#vIQ)VM&46OABdk0Cba$Y$nrp z^PE)!CA5Vw)OE-xZM?+vRGrC3@74D35BPn`4*y z9wHSAi`YWnlVj;#GlS}Uz?2GQ{4{SR?Kq)t)c;Ddf8-e6<+AqXaciGwDcJvQeW#e_K=ke@i>W z5K=tB>YAB#C@#~C>J4G2t_iIz!fU9e2g<*h%z1)cLD?aUuY zX$d<7DBOuBER(wuzgK492h^=Qq51f3pY`f5; z*&wv;+2Z?|`k`nq7tT4?2E=yOknzIl+gv|+w%62ekF~i=-`)C)|0L^p>@MxM^PdYT zHlHGSR3i2z?opMAI%o_At_vyAd=^_a{$fOZ>_NZff>p|$vL(Pt*v2f6-#gKw_-i?h zhmH!52#*pJoiPU74aBXo@s)yfSKoivpD@n>z{$U%WS^`MBsnp+)JF^fn zP|eKcu$z;v#%C$ay5(`TbY={mT&vn4NRng-FmMEk@tKR?O8&Rb>3_H}^U~Mv%uIVT zL$-1_3F3!rY3ybu&Nz2^LEul4*Eq7%IU#)e$ZK2d?4*cfKZE*`cTwyr-&%;_nkHZu zPYR1v8Ckj$uqH|JDh0CRIYGsd`+Ur*3~igH%sY?9eVZrbs|(|LglD9igeH2mcL&kq zVg@hBdI?IaL)(% z=#2?b&u@%@_5TbPT~;Sp7~F1`kUh9t59zA!{`c6GLy__1F(jPGcw+x+1_)_w)*@Gy z9u4K5z>JU{pVNR9WQHk;+k+gJopERxvzd-iBne7k`(j9Ns2MzCcOu9oq}W-rej7u= z!dG8>=$mAfV^iY`VmzF*Kw2A+a%9Gs+mQdnGVfe%IgJFn-Q=XN-kl&AU?^P2#Q+z( zaaUB;r5wiQ;;vpEah|N=@rr1dozt2}u`u*TD$zpMDX#;^`yA+J>>vlV#xZ^Ejg5-4 zK_$@dd@UR-px4C|e?vumca%?K%Z554^HleNZ1DeBhvws_XLs$@5{OI>t7MH(=>xrD!TB9%C)BXA9w%O;w?IH!a%;VYTf!*wLn__QKq?Jm%J@#(k zIZclb)^_I!7fnu&sMlqE@1uWj{7Ea+Yg}PE=q@lMYZMQ|b_3hX?=v9=XV%@J=Rqm& z^7yqRDPo&nlNfJeP~V|pr9Qb9mhlgx*K2O*)g zS+7Nr>0F^nfJ`4*;a6#s09C`$B6axR72-7uf)>Y^b@;b&{8678`tFQ8s1CdukR!t^ z0`4!@Y-DTbS%~i}|5M~kda-_^KSaPiHMI?DN59|a@kdU8qk@#bkvsB|y&$ug6MD7d z@oi8Fm?}Vt8H4(?@;uWGobF9fOy15kc%4@M!w>F5lJYM<()!sf(c6{l&ykfpwnRJa zEYT*4HBw|fl~_aP(wFEPQ_b*NUn52S7pC2$OM@2(Op>lRa9_m50Y?ZVsbSxeiDum? zd6T#zxCPRa8((ri2nK0B9>3*o`^n9GgfIDSlw*aklS14p^7hF&vRug_a0L1^T{1#a z*!sY&8VF@IDpo{V6h~r?xP!IW*`wk%0?zI-{QEhL0Ds$g(%!i!_?pdtEL{2hACp5o zHXxVm49IDU{g@&jApw1l&pCRx_N3Yf#HqXGD_+|iyGnRad^avr(j;E><`sFT5DS>D zF*hgPCtK;>$%{SrMP^DaNz0}sdDiRpe!b+423RMdJKQrhot$SDdtMQrj@T|-{z`G= zr<1cI&og-5$KSX;E9dJi)6=NKU)xBvy}ocpgDy+42psGD;2ZE87}Sg87FDmxBp94w z;s*8M@|?nd{&r9-^idX2iStMoh=-R**ZG&J ztpbm0*j%VsX=d<9Io+b_6qk8pN(@6iC)E{-4xopz?l-6}hn*AD>*jjlS$o5dh|WOM zT!U_K2ztg?4hO)V@fCwQJVpLNqsj)EoUyEAwK;r(0w(iuY@(LPx)8Jnbn!h7D1|D2dVO1=GPiDN%8fhCa@l!5NVXc?9$YY=$TmE*kt+(?%(a$m^VVcINe?gdEYhsg^s3hTE%HRwzK7ANg56;YJ@W;vs9#1( zq`SQrjj2;E^2~%VNDke}tYsg_*GBB^)A&pBbhEGO3MjsGJ^FugA0Y6~&OtGNO`vXF zX;|{SH@ucek!Ty_>Y711=rd9#8tK|2>wvlf2=@;Cz<1{ah*6Vv9}bSU%_XZ%DkItB zN$Ty)^HGXDOpyw#W&a!`v<>PCdI@O{z)Wg|svyh)B-3U{%dVC!AuFIO>rUiJ@p)+- zT|?i{waQEAUXNCJSyZpb{#YaM>fEJQK#%pRur&&kuug5(b&%a*i#_p5+)nbuo~;ez zZ%%rA@2%iJ+!8JIZX$DrN?7d5c=dm6UH?4rIJml((|GZ>>gV-Czx8?E7d1W1qL;3U zQ)3QAqzcNs@lzdpO0&=tKWz}EcrEhmCgsBQ0iD7PTjJds_yj?B!`0cB9q@H_0N{lM( zu>ORU4hv;F=+D_4+4h8e>gB*cw9vCcF^4G(E(~7e**x{&*qo4Kpz*PJ+OnV{ur?|P zUMeuMJ#?Y_jiGQj8#I>W(X-~I2xirb?%ISQ0< z0S%9*8z*vRJyX{Qq{P=XgjS!wMCE>esFp#Y0%{fcSHGYoqCt5TfS3OLt8H#-dT6(xOrl5h7~|B zSLpvW%!ZraY(J7lHgUrZkE;VJ?Ql~QtQuX{xKha%d5kK$XCu zh>eQ%ep`Y5&?XfTMb{Q3jI+^>x|K6==^2HdQF>Yo2 z@FE6>d6-Z$tcbyz&8YwBj~#1Dp5|!h@dV_gawHe#k7aS2nM09Qu|)kbwNZ49siAwB z%MhZ)L5E}^;-f@*Q2=eFR?a$9&})K1{}!!V!+1kN!?0uE%Nsg((ih*2{hq(gx-`CW z?zbd^#|4lut@N8`Hd8EcZ5yaWY&T!#*Q?5-?@nF|)**Z90`_t|YQr5C=4lTRJyI-S z6LF7k3xxagV)E!ixU37f*F6ZB)%Lyw|sIAi2)+|qyDB!4}RY3Oss6Uk34Qi z8`Qg1?a)%(sWM0G6@Df~Y~*XR{5QFSQQP_Gae;>OyA1ykZqP{o&WyCL+bqcRQgI8Z z^dwj8Uer2@Jw=fdRN|$;ow6R<`gbOCb^Hw72--T%pMGPRURAN+@s&HL@00J7 zBke(V_?oD4aSspxo}-OWQIRUcJ@=KfJ|eqRM*>evOy0{y59G59>RNe`6p2VtDK8bs zNGs!)i^{zBsq4A30lH{DxI0M>S8%m5&H*&QLWo*EzU0@AT%=AYKXGM&tu5Pqvwig- zu?{p>nckL3x2Bz;Z-)a5)&;hcG116rlOsb`n|d{>oZgB|3p_@$WAr9yoJ#k~ky)4; zvOYe~O--l6Fe+RhxJ&1A8WsMAI6YI({r-8URmo&Jp;-%w+aTLxj4{#p5y{junK_XK za$G#us_sDkuq3ENnB-kZ6;jQDs)$Af?k|@v47@DIeQtDkhX!Jo@HnWGIDvSys%6Tr z9YeM+NH)=FFt?MjQvTf%-4d0=VREpUDI0g!=OF`2C8)9(hX&L%;ek*31dAei8V%JbG z3W*D*&Xsh_o9V?dS>edRa|e{H%(`w;MJh#=psQ!ydlmM&7ZLqOh$A}sU+UJ+W&g4G zNU#kZ57`sR#K7Z#{z1F(E}~cv@YsUXfSt_2$vE#6NMO!ENtZHjBm``Npe=h=~gkx=P_Vv>;O|rv3n>|LM2uxD(FV$3R5lU6kZmXUrP-}67*Er?Xk)6 zf1ik6NCRsTe-hBlY!kKtDc}|X?kSHtPdyK7wY}lEQ^S9+BkH^?KVuzhQ z6boL-4!{mDEP?U@v=7#a+r8>_$b{dZgX#l0HklfLgy6FLWA)M>qET5k^%A}GhnQ2o z8&@7>)*+8UuGY-nR$|W~USI3AGoUII8WemEjV}XE_hFCI3?FcY1{BAuxdA8V&G*xT zY{0Qhi@rj#c?_IVJK$`k*ew*v1p!E)m+X*d!h$H1$^565>AQlHyrE6`fOdHt$n;3B z(`%t+5{cYCk+v)9qDNzhYP0zT4ea|WZdPq3>hts!#qseoX|3A zV)eWR$Dj*uBY94uL5?HT%n#7GW+QJ{j=7i+}`%;E5 zOR;_&_9yKd*QVU|a-rf3bUv+%IYCy*u%zh(xg54{oV(C6lqFB!R5VQJcjB1}Lb)VLl1i6n05!R33~m3JL|rm>asqp1Cp$eOjCs+Y(tL%YfRO?V>c{9{HDh zMgT5cK=OEP?IwF+hc|)j;_n@aRJeeI$2GW4Xjd>JWff(ppnFC20+go&nthetTJUmA zo-e`W6dpY5ch!HSjL9GXVmFZP+BoGmE%ZCKwa(4Sm7kH1JV}e)lHxqYo~6hqRAN?m z4!u60%)6BS1S*D>35%rlHm%KOKS374&IYt#}0H z8iWvc6SrkNkEnPEHD9I%s43t0$a zLmA_1=$&}Ug=rNa8NWl5F@CKUZnJCp2g8Iz?l^mvSsXIN`fhN!*#GN8skP+sqR0uC zLV*Mu0`z+V>(ytZ@S~Y24mSlOx56+Dwd(`pDkxmP>SMUtq*Exw1w&y~ytTKYWf>U687<5&DkpZ(mw;$k-*;`H44 zYhPEr+E_o7unB+NZO9k)Hz`keW01nQK-MI-(D&r^>c`4sP~!&Ty9j&~$8cfuLIN41fKu=MyIUnq*j zC2#?V$-4c)4e$V5dX@uR=k8rEh}|f?v8LL~=3LagTJgVR0gs&v5HjfJTx3w}a*Cu< ziPoZTWG$$m^W{B05NEN_o3&Wg*yHC;X~BUo<`Q15um3=oyKNsgyw?8Y?Wh?xSWN!# zlO|H;NgC~*{z-~GPLUcau~Ri`)*1RbJ@vTqo|8s%4`>&4#TiRcF1;NL2y}o#+19VtJ3q(0+9qA zH_f^{C_BX4T7>}Xb!G1&uL+)=9G@!ajNTcL6a&PQ==$7c3 zNm5L{U$Nkte1rB3oyJ&Hs0p>u6A6)e5mS5ETa5}m3!Wfpc(fz%hxue-$ihMyO8U;;Il2i?)NnZ#%$adkD@M;;>&4S8u zXk*y5=((c#RFc>x_a+QFr33>F5tX{19`~n94dyEQ%A*N!J1=6bf() z8}GlMai2krmwb8Tq`FXu)c;3gQEW4ZE{y9D778y$q>CP3;kdV)P8Ap<4C+HK)d^dm zNVt%;e#^hRjzL!@GQD-sMi4JVQDiKATOWW!w^Ma6!K^zI zbza&=pL?&IM$XENkhH)OY1gDh?q=r)J2A*{A0`Xq{5ITtvmW8>%L<#NlHQSiK+<0r zOI2j&0vRb5WI(d0#1rIzc3))KIOLRD8``4Hd~;z$wI&y=3+Oy}V_7a9Es5K%0?i?y zqR9>_9~m;B!2rrR42JK>YlhyL<82UG@Y|qeB!|a~@iMzX-%hcRYR;z;=S)Uz)KmfT zW#}`R3ULjcBfBHX7Z{;iwN^Dh990K%>3ryC)#Cw!x>^NgET0BdGDpZBwk|=R;b#OM zp>D+yQXOnkz+502+>-9A$7L`W7+}iy>Mt3@QBFo>RL|7gyPtQ&WK1a_cf>a)Hiya^~*W*n574fFfO1c|D1ewh0i527; z^SLTZae{0LJtI6rKNJnuYRG>0h5<45HyoxdPKcrIl?_D4{vL<32hb=)1UtSQY zHwByMJ8^BwM;zKk0>enahkH^DXy`P@5}N%nG^hIkzQZLph$ zYA6<3rYot$I%pchs)ih>D%W?)3!$O`x#>=@kD|-GcaxP-=)vF`D|3u~H$g>jter!? zt1}6(oNA&E3t>sk-19D{)5$qOTF3&B6Vc;J+Zy@9vmvu@&=$ycPQM}S6gSa0x9Lme zw;0oejF8iC^RjV+caFH?t{((l`^Xt_)(htSt3~&L4N3dn^;MEvo&!*h@wgf$*)A&l z1;ut#1b49KeIa}+Lg@PIyAzdlJpuq}hc?d9xQwDWU@WII- zVefW~xrW52&T?+sUfrilbX09D&f|odH*l+mO75Vn*NT!%8Tr~NJuvYmnY>)8F zxJS}ketFs>LbI+bVwJF1aFc8in7nrhcd1T;*3q)KC8Stzq;JdQT`53zME6aZL&VV_ zEzz?t8|%I9#1xyY%3=P$0wDzSyeJX3>zyH}DVlQ}JHR<|o| z!OOclxJZhE0QycQg+-d?q=;lcuol;tjnF!^AZ+ur%d&QQhv2xsL0zW=`FSWn19bu4 z7W$rQ8&nS03(7ScS=5HBQ(8N?n#5Mxq5WDsemgpkUKP_c#CY-$k@1ve=(pq9!~Yd$ zIJYnZOlG5I?X3KGq76qguH`Qy1w6*lF*_XXr`Wv|*+nI8*5Z5vc&P&pI*a1soA;*c z_x>x+Cg5WiL7h<{7?~153nUlx86oK+?4R2cm=d2oRewtRK)zXfYKC!){_zxL2Ia;c z-^wSyXx}_gSA)jPmtZiIGbe3Ones@g{E+Ee9_tVo^*pb6!y?+N&5MSX>yl zLFPeSx%1G@Y5MT5=NyGfo+~o&JUg4ShGH`*vVuySM-Gaykg@4ISHQqDfWQU%F_)yv z{Le}uK_x#%w-41uaC`7vEDE0@Bj)kGf9cL zMBs7->SZmP^vSSxTeyUUe%|RyhFY*G@q2dxNPGAtNvnm5t0)HV8DifQ6x-5_tFckH*;W(fS{k z{z{U0Y=TU7BfgemVVYfOCDF7%fHhx|<=?E%1>$E5qzho-jKNw=3}8UTef*dM3<9^& z9^eLU+QrFsjIx}ZmH8_hHvXKJ^a^R^F*X+2#YU)IKv8kf@&rS1UV2%fKRCGtvrgh`%*~1SA*PobQar^% zHw)UNJEWV~D}ksf(5N^Yv;%VGSoddjo;u_D=pM@A7sM2dEsW{gxEyvnut|C?s&4~- z13k`n)v}II{cg>^NP~Ktf2Br0XL7D&nXp?i7m5Te;vF&fz+p-YftVl=Go7J(OgN zH<228uEa9VFtIY&tg~j4mWeV#=0sY?<%Mjf=K$4#Wn5)&y>8=6mBGCpM@0BH)C&eT z=nU%Ps*8epCH5egb*n*J>43(dMP3Snx>B@BzJ*5e=zKvpX@<;Guc$I4BjhN$Pq2-@ z8MmpAn^EiM)NudE$^3CLYE=F&PL?`02s`P+O_fzd+@aTNP=)50yjs*ZpDd2ao;pur zlyQ4@oc=)rp6v3|AMk4T?d0_I`Rh&2dYHTAd9T-50MpnlKlEuKT`t9T@f=w_bXRSL z?iZXWthyJNdLB9Ox(r*Zk#HYQ9q!Bxwz9IezY^PYRP13qNfvtodwc)o$7+hrpvZF6 z9kD*CtwMl^{A^%_B`=CIJBxE=t)OmU?N02heOfSNV-!y;GSpz33f4~h)vdwj~J3x;h`xe;1C|EX_PxSo{XPc zKTsEUfB*lM;J~)nGH+u{GDP>V=hj*-tTzI205shKp9D1YJO1~7(P)glr_6rcW;C{i z3ht1K7shB@wlf;F6nl~)$EieYQ|e@}YX#LyQ0ExkY~&Zmswsnd<(QpxBfDF+RenIT zQ*{~?m8=oFM#ZV14Ui#P<9`D>?6A$Zuf7UWpW&_Gtz+%4wW}#huoE~OOx`7N842Z4 zsC$A2$$Asut&yU*D2g@VEwBR*zkD#}iMy$B?&df@!rdx;e_hLu9cg@do8fjs%$v>@ zNe`$I0_|cfvFpqJ7_@~#ciM=7AGM3aJn+SC+{?+29ToqM?0+2B#JmNClf^o=-0n@N zje!t)vpO>pL+BVp(IZ>rR>^kA8yjN=^CAy+kX%HD?cE;iw{X9X6Ea5MIQEAR9alKK zMTV0#ORjeQ)a$`*!Jh{n5jBZV2PQ>i0~=#%6o&DCl}T0o@}sZS{o;%1zrtqMQ*y{X z3hstJ6hV6jG$%pyO|wUXSGv)*aD$5DY``4?u3o_jD%8Ju{H*PH@0_*JMH9A@>#9DQ z&I&rkKZ(ql>c)|GIub5H!s*rSJHQDNqkQlE;N3Wz|N8dctwvJ9O!UvRVN?sLfpSRHMK~+<7 zeKwJLXo3AdG_{@;l&9>)P?W9e#~(W;!>!~Jiu(Ft8xo3Q zG5KR{IO!{bYE)Ed%0!?5mC0aJfQgfvmk$mKAuiv2@F)If_OBZpIe~e*Y;wZ#YoU?op-q`fTj!jclOS!F z$!ygWDjF60Um2MJ{&EO74E2{6jld^PzRc+GrT>1-vFV7nz;Mz=gxNl%xC4qIjGoTM zXze=xl!+r5U9k?VtC$!87}$LUH%xT@dY+G?K8_P%nJ?bB0&IgwRIu1{+mt1w7I?t! zh9-ez!4;-cxGQ*}XA_w-)(}#wIIS!cHIqf2IpH@aHVa0;7P~qY_H(Y?V{^*}r?KF# z|M$i3-hIijIOXX^Rvs^PoN%(@bb>MS*VEffcguV;)q78; zYo@BF3Q^oaP;dbaK?MX61>8{fLctvr6-88#h)baeiWDw<=Oj@{B$yWxCfen%p@!;HE$W%jd?)sTVMWGu1y%J5aN3n;dGIXU+0uOWJm%>i7!vSv} zyv@OmpB#;G>+t@&KesiEC0)pcy;LjdLRiKAI(q@HKw#JxGFCX=>+@8S8eK(l-9dRO zk4CZQD)PXiFDlpF;UF1taP2_Oi2XRa@yyc6_bl<@vaX1g5q})J%OinzKMLZt#-U9X zprUF457V5X-t}<>u19YBbc>HFQ5@=Td&I5ViV^ZQCj9o)*H*q|TIDWU5^@Ok+KJ2J zW0TpsbrmIBNs(n#oH4^ak46`@i$-;p@j-y)yJ~|Cu)JklJ~KnapU;0Sz;adlTKOPW zR=U;n?T|*vZofV$iaAxyhRAC6bQHQZ9Bc?G;EgM4Yyiu1zhYcaV)_#6V0m}ZkAG}w zDwss*UD(yIVpyu_)1lRL#tcoLB;6xVRKXdqxy53@ZNZAgE#vYrMHuz|FHDoi<2ywn za+z6()O8nPKqr)dRE z5DB{A?xVA}sk_uCLb3#~+z-NZz_Z-Lv(NZl3HqFMBIGUv1~L^TAuf@_O@OMGOL7|RNGaMw)g z;x;Q#EZ10V4N@7g=|F^`x+rFStO?Dbp%s_aTUCY$?Y1|#!>uCB7t^5Pa|dLTg1Tug zR8i`bA`udGPSe>-W9^RNIUN@x4Z00&)+enIeBQUV`2Oy{f=y=Y<4-!v$u?$Y%Z1}F zht15^eoD57A|;^tVo=7tMnCjQ^ltIFGbMHMN?|rsMZ=*#Kk2E%9A}bQTFS56J;hw<}!4birXFJQGbE9K{W&?t3 zykzgLN5A{72|Dk7-uKVs=tP(nhG#}xq-5tQ(oDr2S7b~^0}GkmGuDLmNi_pu?P17* z2iMb>P@|{gV^b&Z_N(w%8G9D=v<(%Dur}Txdzn;IJEM({fw7&Elka4K2=|a1hU8Hl zW6;Ye7zyNPHk2@WD%tOj2{~H-=Fmb8CXeHixH~M5J~tI5j#FaR#}x5T^NRTU{Zcr6 z!ZfVhXOJ~s6kwzf*gABM9Tl7A4V%V=+3WAJ>)Q|h=i`sZl0D?Y?v$16p`##c_*a7z zRW5sv5Jjl6=|t}>>_+bWs0!akNsZq|_l@32(HzlDu?ZXuFJg*uhTYqJdHrV{C#5FK zvo1C00$JG`q7X3=C;0%av6g_ICui(*LF*`d8LLr83as+{)GXvdsJ<@k&$_$Dt>y$-%oa10 z`^E{_yWYE*teW_UYz zad+ncRTG8Gp8Y+)tx1gQwp1_B zr;l7GOAxOBCvQO1;Z;T6427r(f9PEC_m%_*m$ge)m>D`qRK(2H(x_HZ{M1Y-$fjx) zTUDhnp2q4TFk=amU;{g(Uy7xzb>VuTQRxKZV6B)~s23BLP+ z&l?hBwo@{)l0T1`Lt`!K$_j)dI2*y zay-)h5_nAiw-paYb;*?IgHj9~nu5N^`L{VMz^6rDCN=62jlo5&VrlrI@FBN;@>Eg- z5>hQb&7m6P{||aqhVJ(6;iChKmvl)`AV^Q^1GlOmM+x>J99-AfG0=I2Bga7~`BC-_ zOQ8iO{)!84C0LPVK#`4B(PPheo~D0RSq!eC>V2|(w~o8Mrp-!ujvH+odph$kKa97G z2J$w%=cv}=$3!)xmj-R^qNw=!7mMm7ddg|5=# z*ABWqD9xc|MqWMC3{dpy&K!8fg+B-CWG0{0^HKjrvTh=Q&D&uQG>?)2wJn>9)B11l z!Q!~cbjAnxVTP!Zz8$%XRUoL6H-bERY7{b|yTyYLozQc;B*~zJu7h>v7oO?T$KF~H zl(w4zV9URl!6SY)Bft1*%NN*r4YL>3Ww!|z{_C#A0uxfQxog`=@<Z29ZLM9n0#O@Sdw20Xnlp;$S{J|R0ot4M<7 zN~?QH%oXWDWocBRw~p5-(SgU(C|MY&RSmMc`6q>%WI=9tZ4gG<5_s{jR7vx{=ZO(` zta!9lg%P)WSRmE(cn0ZFo*5Zk-sG31sT-VFuW7D7pD=}K=9RTU>eS*|3nrLecvPHuZ)NT>^BG9mlv z7&)NA3bIowMTCJyIFk}N{L*Nrujc439{Yt+rY=w$@IVoJ|!rnn-mAV z^HkNazxX!|1ncuyoUhxeu_PS0@KVW2!aAdLKKNExf~7Dn6RYCF^NbZ4U1Zf6NL0o0 z4leEmYC5=s^wv2ws?vxWKM2v4N(RO=iIy3T9EFls_w&0%P4r~bO1kx(^M531E=)Km zF$=3~rDWL@(Nb|pQCbJQNh~?I#AlaR9(@YLEKm7d7PY9cw&GJsx1q*jkGw}x8ug_p zG4@N7SjA%Bv?J@uDFd41unbA z$&)*{dTtJVk9$IKO|HTCeXBGtxI_s#BtiFroDL1cP}+eHM?qlJei-57t;V!!#pDsI`lE5f3=*Q6D5A3&t`{)`@ZzUn-ymwp`E zJ@qPGs@@_f_1On(gDOrBea~~7S78+7ze3hhx)f)l^*#@$CbOaqMkh3m4bM={lNKeT+C~}30(}nl3jb%Qvr-M-% z5cNfM0{39hbD8)gfh@&zDFkden2m^v4mGT*>20AyZoq5-m8l#RLg6Xrv{(9!MyUMJ zNcTqPvkJp2JeE>J%5(GsTIar$s)RaKS@f5YMf|6bXIazbYyOO>^aBEGgil_Rx z%OMG&gImLC@&?8O@?M%1=lwM+V>c?7f@i)_d7WKJUxVKY^CvOfmoClrN(k>3_e(*n zDs}S5BccXJ?0*|MBcaCddogl-ANI7bnQOw?cMtt=7b$0E+FW==e$fn-$0^wniX5im zy68%7l1vX2-8*8G5W!4qD7td~08B~m#VO6hHEQoCl-8}`7|*$saJ^320{Ajy<$1x1!paj18d1=SXx70+_yC^Hy^R|A4UXAZn{ z_SIh%vrK^a{gM6#ve1QHAXuvp8{#z-D5{erDlVI><~$G9^hwJjG2^X`pXUF+urcuJ zaW?w14je+8x?=X0`yCp%5N$%mx;%-F>}BS#xNxwq*$hL6DcR=~sixv^<({EFCQFw# z2=&r>f&@;@4RJoJNoBkKf>I_4Ay{W|Ex9O>>9&@E4&T?YYa12 zkl@`as#LZ?h)DC|WU^ofL=N#J1`&||jV@0Ijh|L{jm0wA3OlVB8L+@k3gvE$B79tj%kKvlN~<{Mm%u9%^vE3u znU_a(EFgNhH^;T4cf9|9|7fx^x`*`>$sHGVV^*5+^PW<&M-&;L;vPkBnSOryRrx0u z)PKDqT);XMl?6~}Q6J|QvT9hBTnH(Hs+67!5_Q~@?vLFs3z`I2GI-O>{C956csRA1 z&Q;xcS6JwqlL(h>)>XS$HJRFh_MRK z)R1ggJK;_F0Xo=aHQu<017tqxnsi}EQP36PC)cE1@cjwA%WzvYgdMBtMs9=8+Ns7H zjo)%r4vIAw)cnBj>($bg+n}6roigJBEBKRsZQ7}xedm~~BVlq5kf2z)8BPq1BC9S9pE z6~4HaihNLHWHdtDMyojHTOEj7`6vj9v7BSR6$+55uB1=+9`oHH(-#yxOpWN`RoMf_IH~zitf5HZn6g5ZHU0o zmq73{k<%$J2CIPH52B#Lx5BsCJ5|sLzP3}RGK&FWgbj;(o%ykU)!`GCyX&tN1=Pxp zyN;F5!pzt&BsXV0mF$!}jIQt~jlg9WYWcN!Eap53=%zOWHG)f&ML!Y7hXCb=q>&oG zKIfMqWdw#{mcf}oz`m`|Nq5%X`9N+mA$io9TNttq zBDi15O4w!5n_+c55Z3Ft0kW(yV2jIfbU`|3-0fV)dLYp%jMpyUHA*@_Irt0MyF4-X z%3Nm}5__>?fQ=IYV;w91`kf#C+LFrkvPp5_NyQ3b3-^fz2yQo3cpL^-xq-WrJXC%W z+Xj`av*;F?R^dz>ygZZ)kn!@(PHeG`jE>*${;Fl+YOUFenL^2uD6*W2L%pt2Ngimf zmd^d?vm{tJX#Goq$M*pno{?=&*@=(SO4rt(@(Ak(*L~T(``P0_%L<=XN@K1dx5G|{ zUf?&9m7GC(w`4$A$L^cl$F3IPqV-Tv8m0Z(IzeI7_GoPFo|Q8)Tg1xqw7=OlgbbU; z$0xaGl-Ya)gt%@n>#G?{?^|vmS=p({V3o*mKfXoXBd=p0hC)Os-qq2d6CnH3gV=i^Q%W zH|!SGHt&7EYwnWp%-Dy~gTVPq$4rMdVZ6syRb^-&`#Sq9d#=loYdQdtADE5be^mkX`$n~^C>AdL(djU1{y+}sJQLC z&9M1cEvcH41~J>Jp#4G(3fnh%8)fhxv5H_r8^f!~EUbiqn)M~h5~bGbywd4Tm%UhE zG_ZDDY~55&Ui9l)g$X6_E$_Zt>l_c$H<|q{}A9S;>T~!xD_-}|ixo^fYhcTMn*Ke|LJLp!CH zqZff>k(*9CG{QBlLXAR4c*5NtZr~ z(5jH^pj9+T`lB@YunRvReh#>R@! zyoef~eQXWn*U;@m>vbEHKhmZ69Vwh{dbLN7$0efmKvM7mULDEy`hxRjY%Nasefe~n zXPV3YSUTyy$Vz5&nG5?u1!i;EMoP8;C*C-8etQTMh?}WZREAdg8kFnNdv27J^6O_d zhJ$NVOR9smc(sZgn5a9tnPFndj8oO<{WDHnRQZm8Yta` z?5mP>P!A9V7;{w#B#*sM)C?r79Mx0FCDsG(%8&1%Rzms2XD+k;j~cRonVE56 z&S#bx!cr+&GDTKEl?*77j#g#R^O_U{r7#l(r#8oM4H%7+jc8%irz0+|JJ|bfqzMzh zJTrR-DRyC6)hV-4uAyX=6gf!6A<<-UXc1orHBTz(6;R;~_X`gMLi_}|CtFnqruE8) z+)jjaL!orwO>uhVSTr3qdslj24(OHVc%w#<+_3yxihLITRmIUtKMsda%~F(uELJTs`uLA_%y#!+yYDRqmoK` zy(lM87qom*7QNAZH$Rugna0tnM`nXNRtTAO=D-iuAv9RrpW*e)} zX~z0yzUoQX{n%gC@fgorf9xZ@xh|XQb#uOAX^aUYo(~$*$v$STiVGW?vu04Qr(|^$ zsiEQ)00H;Y2a<~+b?g$!TG$uR@sdQ@;O$-&qW4>-y3anK+DfV+;B#M82bunbpefK6 zS}N(LH^vA*LRg!^hP997o%qkv^w_VPkTkc9+eK<7LQ?&3CQ=h6 zJ4umaxV_WOe=4njJmuujf@z=xqV8wuMa9xv(zWdJ$UHjJzd)$xmP$53y&Bxx(JGz< zbipj3V>KxHA!3gVe&pyUvoOD?!lR$`(1onzP^?DM2dw@=Sm|sLcCs!C&a-e_lS4NK zt%fXCsD#eisKBbY1p$zX&rc35R%HcX6`cE`My@@h;>^K#J}!2Q3M1>}y6^D!4v0Sv zH(^Th-=ri9Zc(F%NP2h^kl}hcC>nxs{x4@P@j0(N z&w0YJ2Qn6;=^O-Ee2E<&-w*^%%>G@5rB!m-od_%D^6s~;K$21po#Edq&tY$g>J^?B zWwZB#X{+I6vkRxKlkEl*8P9tl>66w98@UIg>)03PtQG9{y%e-Pxjg^Nkc8_4T>BZuhS~BRYQ@vFr_t$>tPB1kQnb=F6jgod5ii>@y%Jr43)FuyEIOU zczp@0)-LQj2kg%crSZV&V88I6J#R;ttPZ~We^Svilk{3 zYeIWH(S#ie`XaL27i%qS3ON|Cjn0kejan&<@&1$AVTsFGjFZG?o!9D40kjMtw6IiZTs zMn33-N0$kflT)zr$!F>5Q$Fz_cX`emAA1I6m|-9L)3?$Cqd{?9PucHV%k+O+JDb~i zhgiCB^hgd&OY}KF-;dIiN;(yLeL!8d9X7VhRj4q5dA}%G>^xwb;fguR&2D$vhj+is z7g-W=UXC2Q@b5{1Q&LZ=S}WWNug*v{WNl=2ks=aif+)EyfUwJ@RG)9=#68? z1OQGO?GZY*Zsxw#eB!WWopdJY)~;(4iIuAA>Fh@3JvxD8(N~1c>Vglh2+^%Zy%0Qe znZ7422skfa5{|2+T-9~Yt@K9!viBd7aR3q%_`H}MM;LDhH@&s+tMh_P&PYjjF;D^~ zl0#wSi_Pac~0H<}53Sft7fE5GW zyUNfmTGOH!fZ_)lEZNm3?gpOSZ7+0CBKxB1NIGYeQ_IPX3DXWmv3Y0KF@0u}*HX*O zCzo9iTS>`UCO#L67R#uZr-52(!Og(9Lfu}+nk@5&I^ri=5isJ-PHeP}fDx?PgtxSp zKYHDT?vdzH2~y)3lM_icdy#mf|D~X2sHO6dWW^p=q;T3HRFXxn<=he1&bYuIFQC|p zkLR-XTA08H?JB3lr_EsJ&pI%~+;KljffT$!SAw7s!C8=Addv1o}{= z{P`=q=6|?xa?;FKX)+Ig>|;1)j3k8A7{1TeTYa_`^;nXnTy{pWN{ree!~K>)NOAo9 z-77+UL_NJXI$LnqkQRP9pw1xNfx;dw>OtuZQcuTI`ni8;doO!li~6n*8}_}o=UcsB zKk?10PE+o%3jojUiu1t0%#_rryQLqPRye;Z{$hOP&NY@fXG7@m1kE!9q@k-(hlj z=KRvA!thG&Qr3Q93F|xriBqGVN}#zw@Q4)ntze;Dsg)JcEzud1yT#k0biy9G79=>a z<&vscQo_>1|Ja_+=_I@5P0?$oFXWtq=}yO~V_%=$2`bs^eI9u3V|PO4`psF!9$r1^ zjTvLZ@n-w+D{T!kj9vRbKiy@SaBC%pwcMjXkViL$=8})0oKj??r~nE{U>5XN5GK^z z3~yrfL@xH+u7uKS*`!dkOW6^o!5V$J5&01y$p7jm`pSQr^$W%_7}k3o`x!=q;kr|Q zMeuKHEtf(}0&*_AmbS8R!dz^~8h|?0v4dVf&f;l|OQ~{Iuj6V5+K4Zk!EIxIj=JnS zr00TVPRYv=eAh)a$ugU>o@b6=PW~1)aJohDu}}yw@$C&kny!e$?#?w~BaMo!BR|r; z&TO-8*8VuZ`_KPm@>v8vznwrD-x%`l9W%6CrDSasxk$y`lP1Y-2)nq|^gviAFlsia z4$kck)GFq$=Uob(-#}-^E{NF=xdS^qPH^#x>ba%L6O)dCD(@ryB4EUKJ?VtN<%sas+Y{>%!*_JOdQu%~Xk;+6ywD&^%d zSmP@{VDl9C?mc`gq>=_QX<5o7FP#i}-Uju^){7d&^Os3=!Ul3XWEpGzsB^Nd;IrL# zvHi-?zjPF@EWUs%*k?WR|P zrpKzO-Lr~Whu!PvZl>={TZNQ<_9LXsNb5jrbtk(HQr|Fw*F@tzqrTW<9Ux{2iao<1wvn8474e@EMaI!G}qn|7iS4_#Fd&t`GJxX2pDIZO`v`<|Rn1hNWtHCe-i30;9d+O}?N@#mNiC;;`-gLU0d_hldO6!fr> zG&Uqy8;O_NnM3BKE}YU}9Wn`SfBE^3O;(1-Tl0_PybD{I2WD2Li;{Ixq@9Yx*=1oM z{=GjWBd~m~R)Hx#rSdC5njE?;CP&pG(}iaPXSSc9_9d#DX;L|z;Oy#HSaZ0Edq||w zi=eoye{<-d^j73bK+ThYORT+;&9aZ0L)XUU&>i$j?;O>$eyd2*vpZvG9|4!vQ8mJnYJ%U?@^33lNV7+5(S_rGRc5$+F1DD@(s9m(>P4u_ zs)w<@z`qevG__B%%R^J|{)lYxS|-hu!bF!x4|-v#lQMzccRQ~{-p+3dJvUV^(HtYm ztdmolKqk#V$9h(z)bZxbVF()mC1Yul3UhvKxvS)|%Wx~(U0VN3GYWm73VJZG_n|U1 z@ISHgD1_Q%o8$&J2Z>dA^uxftv!O(F0&iEu0HkdXx#6FY%)wY5sKveZkGn-j4*tl@ zvfTQ8u+owaVTE;oQp;Eb78e<);$-w4jdg}Thgb*}2+yNo8*)&X6y7Qt^wexr46;wt z{cztt_FbsLXh^|8=^wQ6&KVy2N9DPFckp-^yJI#gE~^Xj?dc;+XCyV#}EckwGcodZ=%oY?_g%PySq znGtj;KW)}qAtqPl#y{LHA^FULG%oC_)SI~~Wt0qp$EB#Hi3M>`B56noN}7#)tWJzl zubL`;5&tZ2yB~^`WI>kvVp1SzD>xr7;1vi^hSd=+G98UcIM{X59f*}gHQ2C( z)yoj5hhy!~wAs7CB9(J@cwifr1zi?c>Nh^t$253MpfTq0X3&AkD}|*rnU&zI&r(S) zGqd8tv0u=i8W#WCOUX(pNKuPR4y^$-h-ZI4`#HSL)?@SQj6HQb+W%q#oKb8DoMM0q z8!o-Fx2e=}H}Va!DOUC_Pr;t$$JeBvJ&-i3GXmqOhtX%%8KNT~Mz%ksK~XHJrx!$R z3EvB4E;?ED^u0;<_!S^X$S7N}-u0(NeNZima#Nw6P^2x4rjGTMqhXuwkH874Ul87s-kujMZQdyrbFX zb&0iAl`h4)IeF6-iN}RyWfu(W99kx~S@(B^tl!+FO#tctf&Vd*Fp)rQnPJv@H6>d` zk(E?j0at5>D_N^K>@xz?hq)|Y0qQ)9#E@J~?~zmy;P0#JS>U3AIu#Ibf$3l9b1ABW zYgAt>qI3QGqMSHU7?EfreqVAboQ-2f_@#E=sQ=;{Cfo7w{crw))VlCy(N!~}cAAob z^7nCYNkKvrC4uYd%dB;Ro$S3H4LfQ%LCMHfS{l??Cqgzu)qIoOOXzjwLMm3H+1f;{4Pm*XtOs~o54?d?h_7Hp2qFQ_RtpfZO&nD zY|m6Ag;#jsQSsCy2g^N-I8ye3WW=r<+QxHXg7JO(_QK!^eJVh#I5s6B_OK##vNK)@^88w0u_zTF_epxLT|0C4y+$e{wE z%So7Xa668f2S#a_mKn?8sX_ zCqAut-BlNKtPXD7zg+?8?-JHtkE`5VHpo%>?T2*NcIgsMzIO@B$nyp^9JG!XR&2t@ z@OR#^FtFMiBYaRzC!-EprlPs5$*_`kwjuZiebT#w4J5b@uCZXzB5{+WW2IIf_^=bU zVPe<~<8;Uf6BASk|J^P%`5x<1gD#M@uUSY9#LI?FExD9z3k4w`18ojDXONRR5SAQ- z)O`)|XFKS9A_FUF!;D?gi9YGlB_7*(J0_Q^4EuW*M2n^_2}ezi4tj-`6RnGF;2CaF zh95f*N!G#h`KQ!x=b3;K`O%pUQZW)xTzG@8)eI;nDA`ep)KhW&@92F|`xr5BJ+e#H z!A<-{g)h(!Sf@S!0rh~~5L_(Rz_e5>uZ3(gJey;x!)LrU2hO5mGA^eRc*RgA6j*Rd zt^X707h!F}1>P;{Q=sH%~vuo0wP*$i#9nX6l+$fo!L#OkutLK->>(~v-ic951R{EbmzXulA={|Z$kv>58^RtB&z@WMeOY1|TeL#k( z^gm20m0I>`@5NJ|x}8$BhpzJPp-)3`;}-RkfGy+z-Q?ZtRRxkUZSwMf3jb}&7p964 zkgV7j2R1)Pr*-2q!u|#I%RdQkx|?8E|CQ=b$)eY6g4=9ncG4(WDn*hZ_1>VCTlDQq zCSjps$^IantRea_>rU8+qw>vcP9`N=Pr)d}fs!<^EpbQvRM4Qb z!ebwMH-7_JN*RZF7bsoDgDe*Q8RL*M7zT#uV9@i}E=?`C)@sRFaM`GgRWe09PZJ+o zM|1+K($vepz%sse}Sd zU6PFjSzA z=AsUgcF_LPU_$i*-cCshXGa8PhhV8Q)L}j{yHVJ~#RO=O<;3f&$N|-3?;R02sz!Q& z`na#A$-9ayCzUgnO&TEQmAUR3oyUUkCNXx92YHa|*%c9nW()~p!!phVK8Bf5u&SOO z^zPwmw(~a2E|JIH&%O|IQSd$pOoJyKIOur_?nG5_O$WD0T&7BjJ`MGUbsigLEDX^J zvY-NCC4EggOkzbnLGS|dqU1lrUBcL)UF={@4`96~G&Z@KF zozC(*Mv7y4;e9_XLcZr0(siOVR$lO9pUij92(*f`K?i&r;ax9c_d;dB>tP2WJl731 zMAr%G*x8`We1q%>C;{=aWR`=mWG8S&Fa?eU49PL+ydPaNFp=0+J%ET zu0__I|JaKn>PhB9&+n!S=i(==!f*kWpTL7w(d{Wq-!}dY{>HiRbKwi7VwK{9iXxB%NPVXy zq;&e4z|)@lRB##_D?Aqamq+HSGA7^iY*t{Nic$8t#~mg!31mjvr|iH493$8jo0{x_ z5hD{bL3*-#<~t@Rl?DIT60+Tep>)IyN(U*~ev0e?8o9tI?t{hy*WbXubRQ6D(mnK| zK6W>KRap|E1xdTleYOR+$V%x_+MxT5`AMy!UEuowe-pO0L(@Ib;$T!5bU)q(kGc}* z;Vp`Ek2_)@h#NZ7bc#0># z@~_e7Nu#6`7!G<`ALwA8#PU&(@YzOGjBuxOo0y>@uXr$iENS;Hyjr)Cc3(;Vx>$aWpFXK72=X5#n|vYUmo7!qG@fv>5lF^K$hp`t>;3ETYyOX=kS3GuNf(as zoi>}Cp383Tgqotb+gQo%752~Solz#dEqyGll2_A>++?UodO8Mwwx|yUrA9rXwchvH zCGzWG55j6eBI_;B@-8muguaajQ1~{*sQ)i zvEc77esAHg^WQtG)++V{+#q8{93vRnezMc9=ej>5IsQJIn|PtlWSNFbcv(a(Pl5N#Rq2K zu(;WYA0P9CJEKhC_@{l_b)=LTI9zzu+++rh&nX#fEL1>#l5_zN)qp_HN33Ze8%Vc! zP`csAtJoS4_kNaxuj!O01g1+{SW5!&wC1QPDf}4OI78=NAOMlIYLVU-8OJz<-S!>? zcL)UVE|IjD%FrAYMsPKK5D~^?enSyJtV>%uOuEf(Q^O2J3~Y?)QhR}Rqm%XMUri{A zT%eyx+FrAbg-2#s>8512DAGm6b%b@(Sl|oXk}Sc1sKU3KuHj^RVT=Wd=R=pnZz z@n(>YLg7Q)qJj#KA6^zdmEe*cuhlA^O7^=Oidvv__yS%=;8yq>^`x{4xVBQcS%z6s zH^c{ECv&Hyn{E;p&h6tahmw-`Jry9tYW&V!EVvvu4SFGS?yv1Ct$$Lu=1@=}>?d4< z0w~F>>(W{dpw46CWKA2v%W9Pv2*90)WR@l?u$%s50T24PBR;4~;lKgzn+$rrhFb|+ zI=+)?$+~&1w4tqVYm9o1cf$0dwO)-NN`?jI`lGd8x$FkWyiS*%qSr;o^Bxk13kZ$! z;xn{4qx9t1Q`VPRAv+Q>E$ibj@ta+EPP0-?=coiFWK8+Cx+!Hb3A~LnASlx)y6(|N z-{S(=BBUH-B*sNr#xTqbQR86UDeFiJOw@krYnrK4Nz1CpCKnFj9WHC4og)asW>1 z>1*<8Ndag(Z&f*s#>Er?FMBK2M{z7?tlN!|wj)3Ok)Ob_3IP*a;=EAd3336 zdq0i^pLV1Bf@vrv-zqXygETNdp<&4F*jIal7O}=HPQx5bV}Xe6_vTNj*<H=`xff-9LO5pVYF=ohZyYj$0 zTBvlBN7qTpgYdIffiig+kb02986bGZkg|J2o(6Jxpk^y6Ho#24%G*F&GCQ=)?IeeFRi#VlU-o z(ItvDcA}3~(Fqzxup$S`l_AQB*6d@KPOYH_pce-lF~`AXY%sCB`@c_r@7rlTX1kRP?T-^PxOVi~1?2A-Y$bB*SP4 zvMlk;Wx6EzwAV`ZdH>C05Bpm__z^zag^>~ZadguuPS&3-70h0%Fp3pbvob@_4&`)l zet>uXE+0!j8s-#~y6Z4bFSw5ClvenrPOX5fCCvai;Z^NywGq3ZV3=7PJ^1GRS$Ax< zeIw(X$YewzND(@V z#9A*+0j~gN8Vn5ADpTIx5{XWYgQpeq(J_6p3p^wCSbBPN_=B>;nXWe_CXksZ}Pp3~>>y;M_W(M+Smr^$(niV%d10JJoAC|~h zf!^w2P6Z^SJtDeE;}NE30H|RLTjwL{{-fm?mQ~4L_DNlMbz-IV_~n_mX6Yk}*ez

    `K1Fs$`ew*qt{(_Hs!2IohLqPR3<_A%hnr1K0ST4K_wxV;E1LU@LX+B^yi{2zp z3_7GNjnLBvl@)V_+>p^$>kBbgs0|GKe{QKgWOMORodOBb~UC{&nJf;rWw@twfKRBVXHpq64Dy-K`DjLjhD z5OOVSw|hp+kXr|R0LXcFNfw>&d3|;ky$|xGm-ytVaO0qcmBGUI5%t3>*eUYU|_{*?!U0Y172#05X^9m?JuZ>Nc zUde7}>jVo!KC5(gRBreae|~HeB%WVAeh1beaX?J}@=T#BfqaPt0&c)$ni<|GN%Xl4DQg!*MUyYiX%^Q8p;UAN z?;@Baq$nf<0kdj)8>E+X(`%yM z!qg|TPSXW%U*NZ>JGhI*UGJQx+xRD^+TnAJ(Hz)hdmLS@(T3mIy6qVO71WQ83nDE= z>0h3rTsWv>MW{XzXbRn8^np@1=UK=EKTMuV#vdgUCfI}%8^+bqmsrQioEaCZN9`duXIi+RjvND&y)XecQ zd=#B`%m^QpSHlNuUU#h^q_h=9^haqGwKGfvesn_7uQiZv^B@M)4J9iuV|sOWJ&fg2}0qZ_a9%f{yKw8+y)p^v20ajx{sPf%CF)dC}rH zATa{YgsA?(k1bii7k zZ8etRGuV^+{DUg2c!NybB$*M)nqpR~tb^_)<)9OVMX3$4h1->dqDpp}e*q6xtf^6u zGa@bEIg#692Z~>q4#UwXP#nW4W+?vh=RdBtbW)gPh`4ZsZo3(@Hc_$+imU_o*Y{A+ z!|!!)kMj%Bxs6SZouBl*9p9LrHa$5ump(@Hz@|Gwl38~GLDWuscbtNEcA#kt7_t4< zpOk#3@jpz)BSSo$B(E0LLx4Yw8B#+hNvoyRi^kFnM-=XtaPr{aQH0MO;M2EafCmjmPU}>NV3a6LGsNmA@ z0f`1L93Xf_r*vs}C#zDKAsb|CA*BiaaUJxa=iR6`BPRD&?YGwd^IQ|~zI*70yGS`R z;->4)&*Y++VLDF9K!X7&Fu1f;f1a9`n zWXk)I1-uV2Du9A{7!4?-u=0gnJRUpjn8mDU!$;fA*Yhi-bG}{ZXM)oc*|B9r%M4B~ zya8Hn2B#uQ1{2>lD$cmP>QWgiULY$t8N&EaKyOGFj=U{H*r&;m=9Hsa?B6Tj<|LW~0EjK9N5XZ>M*5wXw z9lb18=Ue94OP~CISA>n?C;uxYk8KJIJ8qzD?`duA%AEI2Hsp6_Gk!omf6a7a+RZG* z8A=8fFHT@U7z-%f3DYXNR8`YK4Ibp2HJGlJJw10r8x+SN53@E%7sSx!IK-UTFag9hv5XfWftRb&VO)sb2yIu`h?s?8|;iwud4m#=^;HUp~GDEY>BeDl!;b8r2*& z#NV$d58NYYpzDPfgKD8*Xa{#Ed$;iV>=M>KV76~m?hmR_U7Yh+h^wD?UIXi}1n+85 zANEV;q)R)vC%Dx>#JNj92y<+vB~ze{IzKrGWZ2C=BqUC?%w}NX1H14V+)ADU?mlgj zmwV@icd1GRTLRXD_XAs1mC8K=tnFpAO6z0dC7P~?Pgj8yJuZue+**M4+$-#+Q=?i4 zQcYp@A?;3CF2Sgb9W4%o$gr+{QdaoTG9TMzCm}1j*Dc^e7$%)(#jjDG0a$bCzO%)Eilv?-1bD57j-}?yv{4_3YQFfe6KpRf7oBu^VrTanvEvhxx4ohiD@>< zn)aLDlXMr3CzqPdFFBM9axJr{xGI=iTGZ(fer@ykeA;IJE7A?X8`r;p+ zD|3h8T*H(d9t~4k)aO~fbVgw9-1APC?X?>xV}rpf?`DJ(sw(9-|2q2_PNY9CAbZ{z zujPyxst!>y5R$2);_gga7JYF}wqPNzVtOt+L)0Aca!Usn+m6vUK@B|1zleV^sF;3s zPjf^M2!?ck0{tL;-)H!l6{LoBLu|YY&t(fMxy`T+H|muS1>l`yb*k+~-Kf3{oO~s& zM!z8I%RIl=J(YVW`?7omYs2(b*#^}Bu&%e!N5wkNY(b@RkT=$G8~y2xaMsp=9^qD} zHZe0V*Dt01yTW8%)W`m%oMgW-<^_t@4r5dlQ!>cy-j2IVz`|nn&+4WVCuauji`^^g zla@zfV6KDP8JR;5gmp@LJ=Y5MiaOm-zB3Tk3mJWQRj)W53hSUYcqMyH=w8?fz_K~W z3p^U+9M-U{2piRMI00+RFEuTmV!}+oZ!6}L^)BqX>@`EdHcFO5k8JS4fQZ1y?#_5e&Q?t)qb z4zJ&5K2l>DeDj_h`-z!(+26+V5t^VP`Rr@WWYt8n!)!#dDA`7eY@p(>3_Vt|8vcL2 z5dUaadUS%$V7m0mq$`tdd16UCkhTo3k~@2)VR&+O4TNK#`G~zS8VJ|jwcXeF|6=}N zLdeo($uXqcg^kGyGhWRTN;XK5hg95Fs`W>E$SqIJVn}U!>efOQdp>nL?F;m~LiUl_ zxTmC53`EtjVEu1WQJdyvq`x-`iva2)2oA)KqZ0u zp4aGV$s*Vrg5m{Q)jHM{A!gWa5q3gAW)G?6;>5MdyE+I8r(Y4GqSZxquczjOLg#Lz zQ9xl*&tw6nthI_Dln0tfLvEK0mD($f+@Be>Gursr%R|yD*Bp<2MB^72XhAbdC7Yo5 zwy~ojxAmgiVI{Pl{z7?*SHpTk=rtRA9@PKTguk;hFHRsAM%v4C;k~7YX6E!JCA&_M zYvAU5>x%GDR4?h^UJOc^QZc=j`w$AxRzu;79J-Gz4XA-4fJJn!|D|b9VE1Sb>67Aa zR6odVE}6I^I@{~9@Al}W6ES|bC!j^$%*yodrF-R#obrIx5E!}}ffd^i1+Dbn9{rFM zOX_L7=M3k%=l19p^!9B^;FU=1#uKm@JN3Dim>@t zwcl=sSV6w(v~nwbMYwOKUW7HD;(5&xhoRihXkNWthIv$9iNVPcO|r4kx?xWLb(5dv zZ1{$Rk*!SXE7?W-+uo>^WHch0d{w7nQ9xn9Wr1ex)IMPuy&3l0QR1;#(QPQj-LCp* zv_T)`xidLsA9k30gMHV2>p%bNBVw{fC2qPCWQhyUhTusLTVSoDWN8#hMcu`CNRq0L zN_=~l2h4Itf#@`U%@%fP)Nc2r$xfuAkND6TjrE84zkw`vVY9N;3=V54*&2$ZVB7|47k5fg zx)rr-P`C9%OpMFnt@3YH6jBZwMVp>FgQ2#mKSy1}$f~?G?UVU`KiLF~)_2bTk)*jW zFiOmTv6YfRbz?06gNLLTtzx@qyT~9EeK)E!Vt^c1WK71K{uGXm)h*6}TuMlY^1%Gd zCNbt=ZkW*pSyRpi4Fi~%+XF}8!Ad{OVDgUm$yH1H@rF3AR;&#s??PT^GfPkJV}HCb zq-I`xY^QXM_n>!$M}|-9ROchdY~Yx{$m~WZe^^-i4~y9*`{5SomOv6+*nVs=vmfb{ zY&AtzQE?~%gu{(8K%G*}4LY07VW+$ee3w~^J@sSW5M($4wvRN!j>mX8BZD#F;y*q( zVVRlD#9+8?ypmW+FfZjN@UjFzhtZ3gyh|nVyvwtjffH_2!RmlD?_*D}Ck28oda1{_ z84ugR^lb7OBT8(%{j1l%QSEIq8~;?Z={#BU#w=!__RX+G@fJ#^qd==ZZVS7E#81E|Z9(fN(*kn(hJ8J(umDlQAlV)WcDNd^Qn zaBHbe*uvK2(mNownx^WM?i4*nl^N3odBBQl72zcxwv)VI-0AsiCH)XeZb4oNd)?cb zI@kd>UbIj2VJche-zvHrp=nU)IH|yXKu)mMU$b3!P?_f64QdphRMM-Y-LOM$$duIj z_eWudE&dvJfmdbr_SyVuJe!J~_;dK%X8qNOR7H5c?rMm7$!@<>q!CJuqp+`mtNRRN zW}`Fwi8$-nRs=Zf`>pkB9RVW*V&8hd@q^E%n{3dcKTcgqvY7cUE^M7D%&gN+N>)gb zd@61U=V`j*i zZQVk1cf>se~Xn~~G0$yoUWoR?RYTH#?oLYuyOd~(0u^s(p zW|rmP+H2tglV!=~u5BmD%q)uwlN7*^3|kdxDOn~()>Cm;q)%rQ_*Fz7CuPbO^%ZFe z>zwxx|0dZ9M1(H2@!W&yUFrhA6|9Qr<(|&58!wF^b2z=!g;N@?E&9v6zpbcB2RE7A zh)4{=l-C?pBk1~|1De6r0_T6tq&5$YF;A^iniQQGJ5K0eo_FRHFudF;yXII|tjT@^ zEL^dj?4L-^nZ0-olwUXxbNeIV%wuM_r@&8ot=n&NpV zcTmUP9(|Z}MQBwE$wRWy9m<-p%LG007Ihn)8=~W!6X_+Ll7fKMAsrqoIj4NqvO6U$ z>O(>Cyo1qo>_waoZa;Y-K}pNAKG!{$#kQ!8KTp#qxRuJ~vM~%8^H+3~S^uR|FF#vb zTNqh<*EQJRxE!)3zyzh=@2p!zw!AS@QH|LfQA)`m19ArymmjpszjRWa&$FM+>a*$= zb&)4<=0xjQ;Oelp0QDQWX`s6>C3(tqcyI99B1k6CDlnk{RE7Ln)Hi@HH^0qC@B74O zezSOfpQM4n(HM|E0`eRP{tn3!} zE#r2P8W%>#6*F`+QL>X1IYz~0`tPEfqFSQLgTIuYi@NXGOD;i{X|4)$Q`ZTM=T4D& zC|7?bsz35v)GA7=xDol-J5R8a-N79U+7(<4QCUo@zYoi0t>Osj6{6@NGS?oluSvRT zV;@+I4trhk_D0bv|9bk1nP+CMcsrkUm|zOQjmW2RsE#f~UiSc59BB^_nPTb%x)>ad zf*ot87-5Ry^v(LgSd@HSc<R3i&!kL05!zWSwAf zM039O`Q9oO8hi|F;^IBEPvZ4lr$vNohmISV6)hbu8&WXGEMnPo*;dZNG zQ)s0}3g-y9AFJsVriY`$tW_*^YSiCg6YK?!9R`KbE!KhaUU=cpxh6|9@BF&Y$?}m% z46Yj+0!`)N)+d9Kt)oa96;}rpxj|{J2qTHyZl47*-3*jkF(&&Rg(Sa9_9i*DIKW)k zI1~)QVPmgjzsx!~=(3ZCEo+uoDGid~e_s?I0`V!>h}E%bK+_xx<}QqF0DZO%Gls$z zv7SOT-V(Wv)hT@{#mq|W>>l?;tX061`a;q(wN$YZC?kvK{myyg!SMN8+V-(OXe&Kf zfBqvZQeF2=oZQ=gX)O7quKO>uCK7R6k-|v;U9pcJNOI`K;Gk5~1?+g<4t6y;#e3|N z%u40#g6)Y0!DDHcwy_KeiZkxXe@$sE*!nJ zB1m^sa*4IYD=+x^%mMPr6`|&nkNyf92Vk}|N28H)4g74-7Rsa7dNo4CPpgRM?eQzZ z^7fv$k|m?CTxD`OjUY`p47+G>ua zwj~oU$Au}mAhbIy(w9cbQYn&5#Z`+M-M0vLvGsJCe<4*!HE|mySA{(xcsyJYeJ(?m2b2}=rZW{MLb~ZZb^;mn?Bgz<4UAk=%X1zsY#n~9*1$=4U^MSc_%7#z zO_to1H^hcync0w3N|sEK6;xaS?-Z@$6bJ^rFo>(iiX(=6qj9$%wh}6xgcIxCe&yDm zy+o!}5&s_lcO+$`NUaNd9H4kRZ1&JmGDx4ypyJNHi~ENOJdoBrs077LpIe@1W1ejj z;}-GRcX#^@xncb&yl^X!4#p>C*)CIJef%7DW)xHMC#}z)Oc0UgNO|P$M3Q8t5;;W4 z9#fF`2XpokP8YD4QRuWlkS;}Ija3K@iGR4ohQg*pZUf|mSC!m&TndqNH1RvXedZf| z!nLtjdZQ~WnRSe`!1x&dV74A=nWRQxaVTu71}XHjF~|lm?!sUjX5i$g;(50`D||bp zS$=0hk}`XGk2{`g5@R3R>0&yWb=@GZdWo#`LC3R~#D`#!di<3&@Sic?Cr9ySbRDwRWS8-jJ=Ic!u^!9UMRF5X}jz9Bl1uKSWt_(OA)rDPnFiOY5WWyFfWoPqcN zTqh{V3DsQbNZu{JthEf1dhX9jd zAMX3wFmhwMWR%Hj1TMnEr^Vm5tOx9}AQ`2?FGfdNWH_%r@k#V4jEIkn4>`oT1l;@+ zihN-LFHhA?9||g@FOg2^UkBNzA+Q1gS-Qh3KJxaQBB*eV>F>>oJUW-Ia!859m&edv zzzhS=%bhx)g%MyT{OqJpt|fiJN|1EfJgg6tFzI&472#uUdnm@kE9v^Eo1iGXon98b zQ*!d14a)i`r@yK%9ul)LcyXVzj~I=C>+Z_U>uCS!Jd-;d`O%pUQsKh;NUdh8qZ5?u zC`IZ~XcDC*)1_5G`@!`^pF)F&=wGKxm#Z_u2iJ6P6N7T-Ba>4&7C=q*?HgvG*I&ok>~G-eHY-Y$P1B*hJg`}TDrA~LPn`gl z1qFg@@;tg(u}=gpH>Mh3{^Q4tUK3_OA7*0gIArp2Far8pm)538EF-%vJ3Cp4)#B32 z$nW2xOaNKAUha+v<1%S{7S(2>V=FAoZW@OVj9@XrYttXLeZypC9=`w0Uyxd6>4`49 z=DBKSlulDJs2zQrifdM!=2x-?K|}6I*d1{d)Bx5hn!Hmudb&IawO!(42gn_9vHLRd z65eg8Mo(AMD9eHA(aEem0R@8p&)%27HFanEd&L`)9}U?ECRb3A2xYNlF^q_f?JS-B zz1iN(o0-nco0;2ch2{G&w)^g@3jf82yUv((?5Fw6zPm~ zO;l|-R#zuQ?f_}so@jg;HIt!E1!_*RgHr`%aVYh0MUeTFK1Y;IR}1%qzN0CYWz*?? z>0J8|y9&Wk7$S#9xZ(D-t|{C8pAA%3=dOO5T;&ET9%qr3+C`*3p_l=R^izq8k=d<; z#cYVqpI5I%OhF_18oDtqH5}XCTeFJo49d1t`wWneLKi`&M=W>C^D7l}2DB^A&FiIK zIsbLcR_1Fv#GR8%p015Azu3ziw5QLaDZ4FGKL22z(

    F+Km(OF!GmAINE;(+qE-baONM-5z_f@==&e$%X;)Op zp>GL})?@1eH8iIM-v!vtM<~GNcS`Tm5cx<*R?9A4u zZMNRj-}KETW%H8~kOc#>-LiglrdnS>9|VE3>)yt>nL^7Q6{@X_6&$rQhWv)T-qP<0 zC#l2+UL*2}=bda&Dg0Jn6G^$C=f>wxBh=865X&AH20lZUlcCoo(>LjUc(#1KUzi z2cHhC9e8Bes}&Ms=yKQ4<+Nd%L32=xzf#g{d>Hj5VE_mx=&$)>upYzT#WD2|jva3fYuqSL9kIxcx7S`hw?EQ`w=4g0Z4W8p@s_U14qKHJ z19qv5N<73E<35h9o^d|BEq3H_o38#FDK9L1vG9cqYT<8+f4Kg|BZ){3y=EbXVg>}; z72^}o?cVmhNrz#L94Los0yk;oXb(-cb*HTNa-X3 zTjLS*hI)#rp~yih@tSCDltrL0m#h(O4=YhvtEFq_=*y&caqrMCUMA_FlO<3Q4@vhu zvZ~N7_1)+7sMWqUv^TmR^5%!-nZhF2l_-%QaUQzpofMyy!roYXMOhWTCElP(iLu1+ zaP!;)RLe3cPWM>dRJTl(2fo_eri$=2LihDUN3S}zG2!d)N?00`G%GIzCKKqKs5|uI zOLEkx!Yd8(hOivLMfpEGl)8HGoqg!1>xg~mQIB5Wgjg!I@#tT_WHT;1W(sbQI&OXn zk2mHQ?QCQ-#k@L4Gf26!AJx-Q~;Ge;`>;z;I zVKsN3(h!^yvnD1bK2_GE>Q*<1dSZ*@R?(VFwRKgi_QBVR2}q5=hOJ9|S=z0}r_$pf zya?oV^rjJcbktpA_Mg z9ckepw?M!L4m_Dmf7G0~h>1rH;gQ!~`r%hL(Ka(P@V(&UB>4#o-9Rss5f0ZHidjvO zRj7|YRap`e)=$sg=CBv{L5D+rex0INE8T7phGN!{`RszI&S-F3|MRKnA zg;x5bP=f}uEH4*&-}Ht0#zi1^svi*a%lAa)30LUg-S-LegqtvfWTZ=kJD#qI(jSH0 z5f<3mp3xqJl2>?UZD@~nAarloeXo?@g^`Ux<@EcpcSx%=Pk53vDu%rVh>3YOa`DsE zqCH{Q=o9{J1=^wzOe^s=gY!ATUD2%{z4eq0cUS)F!vB!vPuLP=v@@L>DF!xpYaxFw z$W)u>^s0x$`eXE{DSm;pC_fP_mEDahi7t!V9hecE`P8_sSFWy9RtI%7rf_u6umkzE zp^X)eNlGqEa~|iUAWbl0!TmIS`Dx&&G47xi`$IMga-bVEq=}T)k_wc%!sOA3se2+J zAC+YJQRF`ecK9T)xatK4#qJ0XF3#LQdmhUd>xAtcpGV9DKqG#Hzkx! zJqTGtls?N7TC^4`<4u9>*q1BnMx5SLjclBM!kp&X`4Aqt%5E^6cw7@(|4Pyv+b2ED zbl`E~howoG$Au zeCmv#L%c)2OjjMXhV-bm#@!oTIOMT2%n3z~2E&7E5j%=RO`-5Ah+1{f zmr0xdF5wczPImipSz@OgFHD!ij02Zv0*NQ%AD!!={DDUI+_;xJ6ka zC=Do<)I?E)U3T4amYoE_jEb{@Rbg8SSpYUFyOOGoG(GU8zL{VaXe^dV@==f3K&;EuUr`nbK2a=AK$9{SQ!;Ym zxnb>bR>bo9$l*tQ{ML!+zdP=Sc)L?(+52S1EDm4hU*z8lgve&`E&6=;;&8KSeSD60 zjUO77Jb9Pkq{Qe`FDMBy(>Hw2g&QMGx&d-uWmcq#n#B#Gqw#out*B7gDcz}R^&kE6 z@u-P#0|p*7{HR~!WN)a0Z+t}?WwSTSGy8Uv9X#F-9k;U})f97pBKxRBoO_dQ*W8P; za=B4as9hagA9TT|DC9`&(rKvV(BgNJHtCw^Mqw$9d!#&J<=hifb3vE$IBAqY$AZZV zyl>9N3WhaZ&*f+82bmg;b+mS zi%VXz;pfz>a}&r}ZtE3~V>7qy&~}w#x+&60C4MOUXmWqtLbXA2o0LH$Z*NEweM@;p zG$5^{PR!~JtdieS=E|(ct-jV{eCE|k`QeC7G2Oa9{TaY2tfODHKK1gefBG|~%y5Kt zGd@)BAVyGLTsPy@lYzD2f407hNw-MZM8917RI%ig@?!wv%lExrzAGz`l?POV?E1?+ zx|ctDYwLig!mhyP-||>Ws4)ne-hOtY_4mRXJ0M->a4T=KVs~ZNCjN^JYx{p3@)9}D zV_S9A4n6G@W1`3@tOwr`079>-G|>(q>t0r`go+R7J8Px=9@M5Vz$mR#+*jQr+4R!c zMUag((%G^>g+bA*gx)EBP|oe6uMwv(*-+7qxz=H?B0-J_le77Bm3V=~QslNJ09g|T z%^JVEa;zY?h}>B3I}%hZIjFsykS6MtE{{J=z|Gnc&=Z{$Su97fiWYxhqGr=Z-(3HE z_|;NeX#nJQOJ;A^SkCONma{WXmTcnB{KN}>YeQG~@1{&4-A`Dua*Up1=z3(GW}Am>6v!L3UUl9i!;?9Jkzhb71kXF zWmd5%G58CB_m9=@1zIE*WXr0=AZ3=&&ge^hN_{Mk+!Qwpd277b!Pt0vX4qI-%<*{3 z$I@c%y!5=(S`u6o0+LsAjA6!Z(AaN9~L9wd873ezFR0?ht6APHpz z55j_aM_3wgLhi00fNiq5!-i!+gUjy3Pfo(`AJQA;Z7hIATeN`;koA((DD<%xNFY(M zhSY`^3D8+FXu5)6yB~ESDlO(l+@1h<@rbuskxdu6;)*@oa|MLVRmvO&sC-vNGfGKd7M z7hHtgxXp7wcOqX~EE$K+ayWIP0nOv5eC&6ygJv*c-cElTK)zqUcYY?t z@++6X4Jq_44LC}-(A7~Tab-a!-KOAW;l~9>f-0#T$&sLo>Q9vEGEd^-v*{QGLOy%g zgGbyDvSWHwvty$TF2rE|mSGT=|;! znFq^bNG!1n+6_|7EsFF}iF?V_SyufaELk}iRw@K<<(gk=R8{zrn1*ot+o0JLVt1puVhz)0|FRO7^y)YDAEd5c{ zg2@kxd-dyP6bM&6tN%dSqI@@U#he~Xu(}zx@~} zTY=Dk;qBV+0m&L^hSn-`mz8icv_9x#=}PrV$uh}tQtfl|*>1@Ec>ZJN`0+ll`JVYq zM~uqJrQmM>bn4V!Wk2cOFa}pKqx4?DeN}qAC9l&#+lsic*+NAiUN4|i7_1i^_QGAZ z=Yc!AXp36|up{6)@{kjBC-g4QP=wo7nLWLw$;8Oxy>+AAy0MR9pn##&${LAGsE}DQ2l?OV}3hC{TwnE&iZ-)6^TYLiF?$8j_za5e=rx=tndW|Is0Qd=7gMy)03{Qbu3`u?M|De zSivH^1?!_C3G~i}d<$a$2w949|F)=Garehz`LmlX*I{Azn8#1J!J<$&{0&Fhkk85X znnh4#5wkNEd!4l_phaCiCukIDr7(VE<(uup>a6tl-)1W+E$rkiGN>9 z_V9R>X|`Ks8Yl)@Ay-j}mD2Kn2Kiw!AgxlB(eEpA^9%uF#*vG#cxYjH+=a0shm&QYUZTIg#8G?ZiK>`m z5z4{dD>c$iU|l9f=6DY(ntkg<5T+}Z-0*p2%u~;Cid;qt$8pDcJ3CUo{kyF{cU0}= ztsBCk!@DuO8B)fmX@?QlU6I9-4^>y6Z3zEF(dUb#1Ejf)D-v9W0a%E!PA!+`*~rQP+WsE5hG>JzKL$q~I3Ixlb+4WGVx$<5X3eD%JH*&n9+O9D9=R=W)8v`7V z{ggS*_G8C$*A3j(m(0H}HBYlyk#BGR^>4}M&y5vRluaR(xKGm_wPIG2 z?7`1AU5)fM>DK-3##eG*Dt}?{SHDBOkk)64Lkhv2Mfuz&sDEpM=>fjZ#`qgd9 zY}vZ$*VS96)O$)Ngd@1@jfv&VU2w7=XIad>KRxNaNv``>g>9Nz4gP2^{Zf&9mQHrg?-vh2D9oMq4y?G}r^geojY!#>v!7IF=Q7%Q4-@#g*l8GMFW6nIp)B zRB(l=TGU5Z%6p)n&Gz_;`2M&qD8;>`wqD<@?F?9~z3YS9>}`q%79CFJg?jzop~?swaNc~YtBP#n=E?AQ zr+mZ?X8S1yHb{G@M9d4;(U;WS!Yu(!bPm)m-xRm{>#c`JMAaf-^jlA^$TNsRV+?8x zU#>h1escG$vw_22m@ce=Zl$NA%nGbMw3HgE@6cNWYrOS$=>F&rv=tz`>At9q`+(+N zw8hH*@RGy^omb1A`U_dbW19n=kVd$*n<-`^E{utd;nh+5XY3DJABFZJRnVr~;hzJx z<7Qy0U<1jFX`6FT1XB$#`0E?V)F0e8sgJVjqnPkV4d=GkxLFq6qK$99YJ*44?WRfO zBOb$Jl^t2~KE*)y$3ZCfQ7;GnZ0Gz-@?F?+1av(hO^glFo9K%IloG1;F+)$|3`qSg z2Q{JNWQk7S6PXhO)gUSqNN7cNfiWT{rZBcBb};&K^xCLKAa*xHWcPx!Lu{nGqiaFV zxdcdzZJJdHr=nWKdK|VKibPh-#TM?vUOkbSYHZFm5L!GfUA8C^iMkyS!Knm?yAy4gV8>NCRK1*4f1GE;C}~n&g#SZ)&1(K*;v+`CfTe=k7t#!BJ6b+QRf51pu8aYwo|q34jyceDSG-KiUBJr^ko22@EiV5n89|44~h#R5ww_kQ(7`9cL^ zcG&B(dQoHr=&22dRZji|}q$F-gxIdveCLdP7vbcIt_4G{|xB&;roY}0=_e7%vBz73RB)>kp zBxtt;zbZZ2^tc+pSt>~L4SN-6as9;2;vKSGFvqB!a*pllO+=N=akHshRIc-RT}#mp8YW zEVr$ew%?1(UVXy(>ixIkZJ1eaROOHg9%H7>4l_q6<`6~dsKlLLdi~B9q48U6l@#me zb3`WHZS^Wq59v}t;Zi`eszbV0en8u*FzK$p@cN+_theIHC3({eb=o}85`HgewXh=o zsGySC7r9NbamuTvpT7IOV?X?Jnka=CBwMs6v{)^7KD=I2M_*QN6Ql~RPeTvAdWO*l z>*>ber@0MCPM5#Yql?qpNA>?8Dt59BNbS#@d6i`F7)X2UfRs-$K$kFp_!_uffA3Up zo4ZG7(C!45Oq(D_u#z;24nr5gopfO^^tsg5&M;`w`H9J7N-6~$CiqylxXOrS`9Po3w7!T(p)%x5ASeUx znOjH?DpHt3!s_X(gjOYpGAOGe zV<8fkRfNYi*eqpBElTWUT%fJAZmSQ5*;*PPAGq45MOhZt8Qlbdy^p2I5`(snw$wIZ z#ePvp0bK_X9h4!`Z%|x|-X(lT+7O1%;fTG`Zdjd+5k0hfx7e;56n6Jy#4C<)8TI+U zy}$5^BQKl_a(L{2urS8ak8M|M42N)?Re2K|PJR@+#Mhu1Z{y>#^%?c_j`G2`X52pi z&qeXvCA=ZNAZenHk_!Et6bUF7zD;J0djc_R3AFfoCAGDeI%_XI}Hi z&2%CC^3B;FOs)30C|{J|QS^+(evA?XEa2NmxNR;|6>4kAuoo&~HiT8XO_Bc5voQ*IJbFMkz51m)vA>vT+s77fVagMAvDd<*$Yb6mA3u+DLN!?G{H zbiB0*5hiHS|JV1cQ<#2XEy1!dgXX?!zjqt30`xUuTbVWuCf1e_P+&6BsIU*a(enga z)kL6*Kb+JED*~&+4Y*6Sj@$;G(EskKv0&q^;6pstn*lVHUVby3Ox$biwmIV!U#2{$#6tqG`t_*g!zfAD733vcTGW_<_i=?z+6E658Ocs6%Ni+>@sio##{o*Z9n zi@m4Jk~{?PG2(KhQ61Bj9+I)0#IJRdbq z#`wOEd;Ha}-N|s=JMnhyV%aCPE4tMG5G8{krIr8mzo5!s`Iudq{5f1E;Fl}2$U7n3l4g({{j-{pq*&MD&a{eCumdlN7sp) zXza6tQhMDR!Z%rh6bAbkV;KU* z|FdPqvOalcOlnjXv@y~j32%wL55kWhMc2;Arq4=iqKxzh+K&^ksav@yd-^LG!TmwU z{h{%o=wx6DlQbL0Kze6=kd;tzorX6-aieB$tOZYDEbUIQj(C-Y$cOLSrN*|(s8eC~ zyAchyLPH%_z&fTCX><=y*p+;$>Pe^8AqzNDjm@s`bH|2OpdAvE8VJyOaM4PBNrzcJ zH}Cwk4!`}8!(OyNn=5r^GYB`E=CZJh!v_=?4runRXv3;4yC( z$^u60tM^h2@V$3Yi5H{?K)czb>kwllTLU?&Dgbw8Rp4cHlH>^Zoapa%&_&Q;cF4C| zcb;U-t%({ErixbiL*^*VI3Y{-r3I4GfMk*n6=`j{!(YDuQNvTZB&t)rYyK^elgJC{C&jWo zg8n(1Ku#c^ZdP{CXQWmBS%Lxj3=kv^2bRR;$Ob%GO)w^8J%FbNe*R1PPXsd_WWQgX zx{72^BGqc$IWSm&KON z-@%mpYt|QEzx_p%u0XpN1=}UP>QxDsqxwb9n(i``yzYb zlb*osOAK}E#OJld-~2xD5C8p(-@g2sdeO{lXW#vC&(iyY+f zs&v*4%Eu|@C`B5vTU&B)kuXitDf{fG%@XBZLpCxuK^O0&tP_;2aMP3Fn=9*56$Kjn z%s{t7YSpP3U8Gw!Q41>;++Vm61J7p#A}DztXSD7Q)OfVtzcWq^xBFvl$|85Uhc+TDbJeD;SySQa7tE zM^*c5(xk*Rt5Rh*r@!)`7Z>(3LU-4oc_k$#C#HnL|E~zv`8CsZ@uh-Zy4YtKb^NE+ z>z}AL&MekePe*|+OFMJC1xF~BTnESM6=MW8Jx0B)Lg@2k&}7VA75EB99!sI}yLw6i zeMOlAt)F^mOxdOiF!Wau*e$G~SIo(L3SxoSJSInlK_l!n(w~pw z?TL9m{kGp&N?3Tjb7v`GfvP5`W5^L40+9=>dV3#4EKp3g4w72tu@uxk)+{)?!DSQX z>|qaH``)*2f5DM)&e!D_#9v2lyw*}o8bwm5#1`c)RT>0{QO4-D3dlrV>II@p!O*~9 zcf=@HoKb(`e4S@bOVxzokl_y}*p@17-oC$(dtSVC(g-xN9A2H zf_}H~JV-gzKR@0;up#L2co;R*#`GXJJT#ok8FZ9~<*k)q5qw=N>QiPZP*ny5UMlA% zhnlAkhxO7s=O-l$dnJRi3rZU80Uu=8%c>8PudN7NF{PT(4}Aq46eHae4b|AHK?UC8 zAvirAF7|%w!)Iec1?$7y;Bv=&wm_URCx7c7nd{xamfo4u=MpBEI(YzL7KBb_4yiTZvux|#<ww9efW)BAth7aO_)Hofw% zPdY0Po8?^g-@qhGy5~Tgqgav=>>=;s(O*edzWR?I!85ONTSbm+zPj15i4se%A}l#> z5Scr6!$G@V?qWcK9C&8cIk_sP?Iuc_gSLuF8LhJV)H6>KFD(o9F>= zj~IKGKkE_p#PQG*pFAg1!r$zC+B1q1j;jn8ca+C_6_&*k3A&4XuZq!I(PIHzksRdW zqiO^6&2Vi;Y>u~)Hh3n>?Hq^P5bk^oPh7qK)w{PTn>D%Tckcwb&*RwXW;-*NN---b zvW!YZFVrH(gNlOK*X9k67^%m7{H{##zs_(1oD3Rzl^1gG>_d5mNwF##m?Fl zm!&OH7W#cQL|!37Z252e(Buk)Myz|DK#S8cZzO*h>=^9gZA6_Ve1)Q2)(|PC{qwb` zBWYP0+7-xx)wlR}K}<9=vfJIAYMg$emWv#xd)&5AoGiw~BZ{G7$4-q;6q%l-d*jlW zE}&x|Z2F!vC2EcD(%czkmK>tj)cUde6@zyLp^j zIbr8s)KLuZEe}AgI}kIjFe{)2@O|i!XHnv}E*S&Qn5TD$7l#+XTG2~yrYwS#ZGsEZ z^8z!CHG4?^*(fm5X9b9(VK3|RR*B>`z)rXDvKCeOcVl%T$|GCf#%#HH?6Bf$UPi1e zk2++xXD+_N4i_h+>4`ZV3uuuILJ=Fkdz37h}kq73jQR(pAU5FTbI#6YmM=Vp5~bu?u3xZZ6p7ivu#8nGLpK$9$a~ z8BXj2)&8xs6Mb!Hsr!NfszEHu`` zkjD)%c)8OuBdeedti}87uq5t0haUTB=8Pr#$31rTCF$7^#|YUI<-@YY!9@LX45WiN zv2sYgm(+=YXPBur1aAzqoRkK1h?n}0*X+aeaRY$I=KG<6*^No>&iMau^gnpJ_hs=l zD(Pl%xoS{u?LU7(dR5ixod;y0n=|VL<*IF=;|Ku!eDL?S-Tm2EXg@sO z#IVqQDrIKqXP7Okk**HTfR0c=W;X&2@$}T)Q^tD(!@;aLpFNLVJ@E~{8$mYL^B!}2 zIWc^0T+b>y*K;SuY^TUJEWAQBNrWvVsdQZ zTs?iU_DZBa4Vqb|Fj>khsCQZrbPR%SgG65}E0%bK*y55Maki(iyXJn}oJ+{s=eSql zt;@l(`#2WgOLx+n5-j1;J%WpZ22uwr#u0C;AQJY~xERtBXrYje*BZikzTH2yPDkYA zcTBiA?N#YmWb^sUQvlA&$erw`ZNe8%x)gL%xcKR7k+;9PWKuSLFl_VWe&HZFLDEFM z&~iB`@;d2(QmT8=x1z@jOB|e7tF>@C^B%YuwC}6^PkK{9>)LQD`j)tWMwVkYOl+0r zxbNGm63#-_1$DSnqq6Dxc{RfNdD(P@KWJ_HmryOrd~KI}$gfM&qQv{~kt7Ob`6>dD z5n__~8M!8Gq1Z@7mqz}dRkts>4(+<^~n)@6I?FR~mXA)k|74NHvV zfscX?%Q7EpMRf_ae?|d_LUyTpXuL9RRs|=V3y|Ca5Ld_e&h-C$C=&8Hfd)(5q*u9I zY4Kw}me$V6^Se14h(_3eu3m7$A2`;ei3~2l7&o#2pyx*u3*T|%4Dr@qWnlps z=_;s!vr!FQRMH?CQ7HefM!%NG7g5+b?uX&vCI9Ja$uAl6j>G zyAn2P4#v0A2jZ_!Gso^z9G<=_;h6gTG?Ok>b~JF^RI~V=D2r)U-i%9lT zQbn2SA>VuQoavRmC&-Zh0s3Z~=RoHgwwya7aMgWaav15qzbXEc%oZcb6|cKYQn|%Q zcpN$dk**QxIRnLj1i%K2k>rv!qDucmB7NcsE=G!F%jktfj4sTBW&#*t(FtPs^eoAs}5?O91Aa;jh1}SKL zlZd=N)JZ6et(=R&r5?JK&JndM?kk2B7CHqIxOT}>WYs=cq-M}G%*<3(_OYUU!X;q6j?0C_9q3B)xoG;wl+$C znPjT*_q^$Q4;&4+zggN#SBY!r4`%O_7R#3US57`Ggv-T}4sm1L=R`q_w`HAA#eMQ+ zFB|$wKKboiB#E0v;<2M+u(L>M6q7=c6;vX&%0Te9Y0^X+!>XsBBpb(~KtFmj;fg^T z{ph2A@^$*P8pr(L6Gc8^$pU5xvO>WYiHl`9vR)eX=(8c3`(A*)MOXp}!Xe+Aaai8F zEXG92;*4~vOy3jTqAalpEVgMt^$=UEZw%Y3#SSc3+XaHJ$*)P{eq80~>)4m<;NrDwa<`sV4qunG^7cEv^cF_peTwUudCoQ^eU#+7iu z4K(h6gKJ~>i_(@kR)_O86vtAw-lBy1inyizMmkM&Cae&|=*{9D)vy5pT99DubIHPC#0*W}aS2CaVKsA|FF-C>p)@whE2D!K&J>gY|X3>(qr!>3;-ip z#rooS1YE2W%N0so;4qeZhu63MW~R+u9e7E(gly$;kwu-I1GWlo;I0+^$?bkTX1AilxzGr6C5SxFp$1CSuvXq;D!s8uJft`P{o?TqnB4`{luCojuw96Cyh)MCpOWZ73v*Bv5V~B*euwV&#G=K4u zeAoO=b%Uk#-a^q4VHtgtu2U@xd0*T*cfD>|NR#D!91-F&I4mojCt%=wOitjS!cYC| zi%)uia7J)1{Zx zGD2^Aq&2li?TK{lu0DFj^RG8NkKN7|PqKex$)q}ZMZ$4n(hWR&Pnj8mZ*y6Fg0w`{ z2=J|r5NM?Z0vGs(C8FE$b#&DwZl77%wG!Kyt-G5aa9 z7aQTA#-kakkMH^}etK`19+Nq_!sYSURIyFd1{?WIbzS(RMi*OKUVqv9qg&HwWc zl74GLPx$YqOd;K$8}ux)3P|R_P90fx}ZvkzkRB{FMsC((7f+MN| zNw@FagvEh+Y?B0X{B&Jpk#K!HMi;Z`Yg4kZVV9+ePJwnyfR)yeO)m-V3*8vD!>=d+ zGC!H>YM&0V^}+3QC6i)%xOzIONLmJHoQko^3LibbGs3JuYhuy~zV9-hQIH_qZkt%# zAg(5hlP5lb4%r&)n8jvE7B3dH_}7S2n2cc5PP4Y6&}W6-pchDRUxYGR=|X6~r>!N1 zU_B~T7$9EU0n`*D%rnqvQRJdnwjIB0Kug)c6wlP-VsSvbBi3CDhV8?0-`sHh^dEgc zx4GdhQ}=Jtn-Ug5KesOEl;CE3J{l#K zS6zg#3nRJr7Ip})NE`m*KUePHs2usXmUwbD{f@>O4l=6{I!L<2|HlB zj~?wy+h&m$ZC=YCPGxb6De^9rSg6ev-VpT1-2<4u9Wk-PF!$l{CJM z_1@jVJEhBI%O)8#Ns>$9-Lp7jj0;X!yb_l#KN=Mezshb~oFucUU(Z?iKaTYxyiGl@ zRDfLYc^Aw^s}$(k$ggZy)YF4f&hz>3C&~GWOOW6^>ewIP1c`~)!oR*iV>2|mw{j>{c%R&dvk&JI zlWkCGf9A}qB!kE1Wse<(^C<>4Jq9WB?sjHROcgc!Y|ci+%QjNgGu(Z|My3d z$t^m}e3AnH(*uJe@)JKluF@92vS?QjPY|{URkd$letIr@bZbvUfUCVb2qC6a0#j42SBwM1V5xtFx34}o%w2NVz3K5>^2FmkKQHjtaIvzvni*vXVGh9UxiK z=?~2ns;5^34k|kQP?*T`YYAsOJivyV=ihNSc04$VpicPh!f!o4mW>LJEep#=<(Ohm zBvv+|z@W{-J{YRlK0#q1;NDk7URtXn}0tC~Z&YRt#`M|?U8ly%P zqRSBqZajbb-zS!-ZMI~6diYtgj@yQo$2cjsvpj_qlSh#pDiNiOp+uRkk@kYlH7q+o zD}q>SMW88gkDAndq7}7Tw#Pi$F#*McSPA{Y z8Sj5C9~F6RCQXq>pA>n1)-8FmYbfGIn|4=JV?a8UE9L=0PCYNdwm5@2uZX60?tB>?W>JD^=Sq^=jBwqP6Fa9OM0 z7uH92sZnyE%D*s1KSbxxtJfYUJ7U`vEB!pTeQ{~}A2=b-etKyBIXNT~v)7tuJZ}Tb z>^I))Ak~vdyWO0;OEJeN(nKX9cd!yjfG`V7g0lqM154xaQ`Q$~@pOUY<3P)4uC_Y1 zEN%m;I2xhJCq98+S)9v70<##Nu7T4Hf#)J^hxnxB@Q|=eV^%fB0cZ8Bg|LPfj*^>k z2j>9cuuU^aGPEc*iA^*z)yQcp3gP_KAM><1n@XqGxXmxOY2z(X?d=~aXE;V!AC9K+ z*gav1qb5rZKYL6IFtzeqjS+PUGu=b~FRVMR0hk;3+;KNMfaz6#?_-S!%X!;R)Axep0`9rmXJIKL2y%Jq&LsBOn0U_O@sv< zP4w=7Y`R!jEQ2n$+P{sT-QYOeW1dA<2Vh@XoP}UN%>TwR$6i3Z#Rf}vpX0N3PCi50 z<`fB#dDtj88g_W1hD zSwU`rEEOJm4m}*v;%FOOoqW#saY76Ad+Hl!|JAk<{i)|q-Q*n}2U9+B@Z^J04|b6 zyx>FO8M=~MC2ICIE0BNCFWeJp()C2A3b4zek?t23Xp2Jd7wA%BQ{MMO%+p)x8!pP4 zjS*gsaC7C{C8%+O*Y95Y`qv!Q(^*9cU&LdywA!s*hbg9>A~hIvwrXBv zst;;`(FP*vkQR_(5g^w4q3}SC08@vU|Fhny&k;OeyXhB5K;Br@AvV+3ME!AkWGb{P zy62!r#_$pU~txB<@m|$D749VOfkxx6Iq*dl96L&duu=?(oUbn#GNZ*6;R) z0wr0zPk3HX9bY{xW}_qu0WP7|CNZ z#NvEq(?h})bMS9Scp2&abB4cQ(4cxsXTW}66V!_pOFWlpbOt9*2zNGshpxT)ORw#Y z?Ux^p_4612Id<#N8j4v>kySuX@!k>C8E|v<0}?fA;-en^|G6!n0*T$YeKS0M@bT&u zJO%}edPNHaR8NyW5XeE#B1y80fW%R$PpOZ7chGnP!)1f;>3ogP0Xr(@j@ac7-rk4v_@~eQ!9P*H#--TC(+V3(q&K90AZe4H zsNbJJjroX#PY%Uwp-2{$h|Pac>~qj}Q;bP>Fl_Nu?DJ#N<#^ZlS^FqAix0%FeF{p> zRf{9=?)6WPi%l5~o;?m`cmsAkA7T6^J63RUaYuQ~Mq#b(vi1bCRCnbFkY8n$!h!T5 z>P;3)+Qv1X%n38&0T$bPoG-1*f6}$mtLN>2eoCnK*5?257mW1#v4=(am41y{qt6<@ zyK*e(#Mv(%$D&S7)9q27ygioKO@xyrW5PhzM4uV9m1WVZQ&*8}9;bGy?N*Up6a!6U ziXc2juzxoSm7wxw_Z)*}t*ioMq_;~pYu=Of03pd3(LJY2{lW9^fgDl~2!)=~nRIP} zBnfI~;1w*x=mK`;SqQA*vz}RFI3mX0RkEJhPqmxQI3b4m`u5i)j+;8ghW`0?33ooE*!^y&TMbujOZOSgP&W~A~f z-)JQ(KQ}S?B0ErQqL>X7%mS3^WCXWLkp-?lC+zb@3E*L`j9|zKYwm~d4LcX1&z9{C zD+z0y>Pdcq{rtL!j?WB>{o@`xJKXX~>R1Z?cpMaADfl~2tkt6!66qEuCFmQ%T0oZ; zbN(1GwN`eRX^WycE)(>iJa$HavN&Q(TS75lC)})`F8KGEw%`xaY^x zV}r+Lg{6}iR;FXi^=5AnCbvZ4>B_m73tk+$Hp(MNdBiNc45CMj;fa?%uIu=pi4V3k zOCuXWxPcR{c^!C;w4R}kTe z8GqcE-+AuGKX8nVJ)Do@@w~G{z)nEBlyt`Ww@G(h-V&b*yv5T~KU9~4^jWH)ew-69 zuFSt}`FPxvx%GASuYXoY%JCqZt+>Y=Urr2@NR{1}Y$wHRr^q%`FdU3d4Q~ZiIn*#b z^}_1s4Vpsv65lKG9G|^G#UOdQIDEeYHfTt zzn!sfcyN6g8-hOtoZAmLA$Y(k&bC4#ekG`8I`y(yjffwRT6hmzL#tY?xVZs z^r{=71ka#ZG0QAYjY^VK&MgnA3jasANtYjGgssfLv-@P1#9hAl+OC9qN?dC%tM`TB zQ@u2+%>o zEPLHSa+sWupO9yVWGNctWpQhQmIP0By9x{EPkZj_x5P)@{Ep3} zEc;{CE%NRpa?S3$c!6ThQREDj_^u{Z1{yJfhHy}lQI$#SME&w=pS`N3)Z)TVJiV|!4#=-844$zOqv*Di*9 z_-qPuH0*F7K5HE%Q;k=l-;%27nL)@3?S-OHgXXQj{4I|Ry>Kn^?3~fyZ*+j-Xc9*O zFLpppF-$z+$ccRzP&_8Au@HFMUL5jGO1RUqN(6m6Z(~5or(<~HfE^{%^UBVQC8^9~{IDdIQ78vhPI6HB z=jNwy5Gb^3RRvuSgzeh+hp?hJq!4rCk4;F&g!wHNAW?1!tfItB!J zlpQyz>Tby6ONe=l29^?HGq}oVEG%IcghmIHb&e+i@>5{AHuIm}K5hr>w&Uud)Ps)d zpik5a!=f0P6NB;<2emi8x{(z7JnAe>v?I3MJ5znlf2H5M&@cIhZ|#f|+PiYxd1ZOB zqBh_ayzJDL(b%C@0Y*M4uS zhIzmGw#p1zxx-#aXX;XeN|y$!di3W>rWy%Qr>ACXfeHl#Y|Jd>fcr8=odM^jVQk2M zV1}HJJ(jB-e=hu?%m$fU@w&?-l^bMu9M1(({s@wafnqjMWCN9mOt_+u4!;&*D_scN z;T%D~>SSQnzhfL_M-f1m>z@C%ay4D zeTUz2Z+WzmR*}7;ijA5!j9q#w3MeR#0Rqm6#E1j!j|O6`7IU>Ki`L^TPU{2VE#< zQ{_5RoIUNiYfE##`+G+QByUM7ES$!MnHQu+x+ze9DDW7#2SdKS{~J2vu5sl(Hw_>2bHv-?Lhcg+s~{vj z-paMx?EOxBrxasgcfgIbHorjLUAw|KZIN==t4)v@vk1y6>f%cU9dxo}r5{Agu!BNF zSOygC80j?dWNYY7h_+vrb}B!jiv+ixJ?jl!9ZXD%GGAK{(hAm53+RsM9kHXeGp-v9 zhb6_;Q+B*O@)d4i`bEKC%EN46ipYLDofJBAO?d2~9<~Em1;s$XbvKn*9iOlMK=Nvr z;H)}T@Mr5;n+`pY+oW4}_*>=AANdwAKcRNCU${Eh46EC9MRqWq`BwY){@Vte zUpM_=KDohTE3|szNY7)zMCLw4?of#Z+Jb;nbL&Dg)h)^naV@9{quxrHpez=32&@X` z){XGFd8g*${h=9jpe zhpK7cQ*-h7Ki!EpSM8ZA_p#&6$x1juyY+8BaE#UP);D5_zU+SCk{ru`I%3yN9rUeG zRf|lzOX?2swa9j+NVqw8OW67Fw%Emid(}NaV`!r56qlf`@RIzl{IdFnV0*mL$E15- zUgTe=FlaMlw#RRe?;cfZ`LPDV9gunKn~dTHPLQF#|J%+V3vFi0E7U8QEdJcYYM|}l zi0xG-#jK&oYAUfY9A%|W$2N-AKDAL`o}Q^*I;rO;AAkAa56ttIO>*lh=ECf`j3O7t za_22)qm8=9^Avczxn|)hAUklS;Er%*Or76-FQB=2=c@~X^~baZ?FK<**y`X<0#C?? zqXz`3g3jpS=o_LMOk`Y;W(0dAgljitE@Q!Q<**yY{U_`eW$KKseU1t%pA)?d7Cn|d zbCdm6%PM_yea+$?sr}g*K_6X5uOO*{>jA~WYM*9tePExcMRQh~BQKW&^@%jmpFL3s zo(hiP8H+P&EPaM8_X7z={P|PgQGOnkBCXQY5Wa;lZ-?LhNIfc9V-<5@jAthJ5wqj! z+&^LnkG#ZgM?MUu#g2A_w~i}|WjGp<9gKfWx;)|D5Iou`#l{*YT?a7Jk4B^kt!GVi zu5TZGG@?NA!LzG_8^bZTxF{wy5S}A#{s@^H-+{9C;5~B%AJ6gANChqb!}~P*NkC9#2(3jzIDVXq*lDWwLjI zqlg!8jR}h^7dAMs&S)VNsha7NfeoZC-2A_hCU9+Jz8f(~KPu10POS_fpkWz*S_`r)wC^ogl=$J+B^G_&d&hCE7@b^2@n|g37=1aqOFbO+Uc7$0WGiFswSq6fQe^Ti z(q>J*0mSTXAyn0Cszue)uTNeR)h#>)k=#!d=IQXJ*wO_%vn*nE#u_xqk^|4+pJA^S zCH^9qu6hfzRjzKgU4_;|PR!^pIE9kq#ect(w=2{Jt)>^&RFeEjq`_`Z$|(kd6g#QJ zo#B^YcbFsCMq^j>b&6)`T@W}KmMv44Yf_jr(TD2Si>U3?>+7FeMlQ9IDLCDtR2o94xZH@R`&pz-B3c;bL>gxA-Y$$Me+BL8`JTwlcoWMFUGRa$| zvzZce%(N_WkjJS-=twyt>VBMJj#8u%8IorDmb^`q28GhLHeSe$JrHXZ7b>9B3D6dH z#Da8-pjMzisNE4;KrfF!PjFE{CN9=eLnIbC53ighgTBCDKv=;QkR;S&m(+~8HBr~S z)0uV9b`lj|_6kbrUTL)t%I-IWzp{C9%B&Xh3S?dJma*RfyN-S>U7ky@Q(sK0F+10pvx$FrRG$0!)=| zA6UN%w5S1+O~0e*5cf(e0&8KRs+?P=xUWL$%8*~0s9tbgtiL8&3sTAbAo`>4peq88 zle_-Cz@WO|ljO6Evb2}lqxr`nzYnL0q^pzrMp<|MnC2xze1358$ zp*v%fB}EXz$b)j=+K2-BzADe(NM|YZ;AKL6a6bqS>))4`&{%tZkJJXhK57&zJ-2_` zs6SqAZ^QX+P9~nep^V_hIoBO+%)}`xQ>8w9H#2s_C2=1$lfX+1%r-8u>=baF_vb~Y@XQ^nkq|)srKof zV?8R2L0trh@tLO|0|kCFy~+nKxDN~Vb1`OQ*bnEnGj?P+OB1~MUgQ}^KbN=R7Z!pW z_6F1AZugbi@K3<;81h4(MZYhs5300RE62BuEvJ4AZku!$2I7Q^iD@TGN*uMqc#8`b zU9d|b7)hz2hkQHL`hIzTTrsrI{DH~hmJU5WWry6y*t*i={lDbBab~8mfrDLD$-{u6zL#+5MIGzkR(ZsbV$8q z%KE3f0!kqOvu{?j_+H3^S9Xvtvi|9snccoUIuuZC4qx*eK4HDytSSz{*Y0}_`qW6R zcUfP_Al<(1!r>^<=G=q9t8PNyoAU~`IyOT6oZyG0Md~0)VNAMo!G)+}@xAmAZJbp! z%b+<6{eQ1b*{VPue+k_aT_j1Gov&U#*Q)>IE)+a+(nbT3M-J$w7uO%BTfbm)X4}7Z zwUCrMt4`zZRt_@IBeMgCDW;wxHB_RRKB#@5v7zr$ul2tKbn7HZ?Gyt9>p>cs*0)OY zpx7u=oh8^Md`AkrS=H^(%2K87ZW)$Sl)I)na zp@|gBTBX+^3uUGcL5{m!Q6xVm(r-{)i%u0`ce63TYBXSBHztm+j2=1ctJki+`-3mq zVDyJmS>GdvxJB)Fyi#4Z1K4{M(@c?fsl-0IVfyaizDQ&@+=|A;Kvw9RRnmR4un*P7 zFa7~IR;u;S-Id?>di#x|&;O-l{t&P=@6JCXz}tH*JvFUI#lqp}Zr##JizgkLc2RQE z-#{9|&xhA*mP~3_7izQVvVacpVW^C}8(1Rh@T;SblSbLqS+_}h)SgJJqO1{b@&7os zNU+s^&zus`?RbL*sT7>h=UKej(WwuGT&F|{72SYGpc{^#6LzWY zf_!|Rs3dM-B&Mazssq87)hiUGP*j`dy?C-gvp{f6+o-^GFjZDZcleYD*FSxPSQ?4P zu7qlCgjScjP@V*`XluL+wO!DL- zQD&+SYFm_MbrVWc39Jp$9y8fozPm2p>k@)o9?g9>ZqEoTaMYaQ?FOAiMG6b7A9z8S z7h31nOq&81DPebB31!Nauy-Gx4Q@>+Bs*$=PW z@Pph4+v@b@TU}LT8#kYa$EA)(?AEgV6tkBid#J=i(Bvq4=Ence-kZQhm0kJcKJgV* z7mL~`ta&->>+j zs6&Wz?X^(Pe?(mX1|-KX$4`r=@$k{NPcAvi?-ruSL9_F68J0i-BfPUtbxcqp(Bk1P z#Zi7CjdL=cO6OJ$)K$n3fTb!x@N7Q}+5$8)+>PCRA$ZuFoc}Z4dh?8#=NF(j3=j)h zG=>bSL}jh|Bhg19ePs!5;Ebyo(nzztvgzLN9=UdzXql)_-Kp9y+ApeA=ZNkGYmssR zHC>CtYj}5q4OECo4XS8%$3GLAPmR|#_JhWFt=oRgvJJYwf7|kRjrOM4>x_bQv$HoG zc2ZWEu!lUN7;vV04eTLEx7!M9RY1T2`JZuL(v~WAF#rtA8t>(4d*tinsq|A%ZT8&k zS@%NklN`TndVs9*yd;eUYQ%(vF`il+1MevA&(tFQ(dQILkYUjUseWl82CAY$UK_Iq z7(zC=U0A%yt(CXQt;r3z^;%%51HZgoT;l~RUx~^m0qgvSkbA>@%vbX3u~xctfG!gX1){|<{B%v70OvYK z16G1wfKIi2M*f^!QJiQ{wuep$xf*mWILQ~!pH?-YVg(#A%#Dsqj|$Rw`8-_YXmRi7 z*LbzNRJfG7Jc5<+J#>Z3C65ifT>b#T0S}yN?_;hH={ksFBb7`^#3vr9bSsn0TTaCZ z@?rLNO^NN;Fxc78bd4R{%(vN_Fw3TEyg66Y(3|tk%Nf|rxi~yJTByoM_K&7YybkhV zNkR;~;r@Ru@hbA~_SDwX!$1m=1wb5DUJ`ZDUH)*WUzzK-$>)arj#`U}9O?XZ5vyGx zV(AAZn$-b0zB#-ik9s;wt;Gwo>8m80?zZPrynR8&9!!?*_{W{UTy9==$JuF_g+jVw zahohvq8$(qh?hk!i-ZI-l$^+-t3Z6)276%XQCW4|>VCGWHTzw}%9u=^17zYpgb^nW7Vg-(=t6e^MU{3J*O3k(`kiMdR!m2qd1NAJ=`(3&LACrcR zvHfEWO2%G4S=*LTvfA&`|C??!7dQU1{V!zQG?H!dRwhvlj4E5Gs3Yn^Slz2utYqGg zB{!hHvcaoj)dPDaC$XZe~Z zO>ox1u#g5(Ojp3{v`LNyaVzCe+LoxSq-%V0U}~yuh1u+K(J^P3mL~J=lWs-X!eWp` z!g;adtp{uD8i4FL&>CB2h25#MF5iBg7|lnKQ`%{=hQr=QrioqKMllH#iKn747J{ZL zKKwdAU-n>ju@~h3b}BJ3xF71kPY2e_aXcKx^eY==jJeI>o9vMB&s&7wHFq^QYdkEt z6`xB-pvYph2sv;k6$ztVY7HVrubs?Kc8K`-AEr8)&rH~aIXIjdV__as_suIyE=xbW zrmR@P$Z6*&@)<9{D%QR_ddiZD0d0xhWCs>GRNEnWGvGsNtOg!1HJI(sI{1 zP1L;emU&VF8*9Shps9rvhAxPN8U4^r)}DlaTPt}d0@3o5ptVT`?de)GA0we0{at4DR5O+9HI72&M&Y9U;s zS7}h*282Lp=8@tNXs3-hWz#whN}|QM0CB1kBW3pLcq_Vv9X70J&`y`I!p79M+~UNv0_8sAwVee`lB#VRlO0isw=^o3=%9b>Z6}5>Y2H1W(}`Wd_3YP zugZJdoYu%o5xYs8dwEz-aBaYDX(ucTql5%rIpi7-SFMH;rDUdvcfqdgXzZ}_ycsb$ zVkg{SGpw*foqD+7LvvV4osR0I3~VE?wo8E85007p@^ zU+TFrBrQZ+<9o+7L6W1==Be{#9fk#}tI7xAm5{iUs4V2IU(hD&BURobP6Mx@J`F;( z)w4GoGTL;SGc%5ccY;HGlDF2+RvTs=@=S&!wmUZAvhd|d%vBw`Pa(d`H z^FEn($SVanya%8AgQ=i#-6Duo2?Q$*rS7)P@seirtL$Z>c zmEmrTi0m@4G6@tDPmx$8YDQ+UA!UmS)qu0NCM8C(?Qn=()n=$j6i{=;k1hZz?9^-Dd>aM)2e0FlM{|I{Bx~+ zsD`)G)Cb`AVWN>&XXRS#B{t(4HqR0b?#E~LM zXx_K5R`_V&=L@nx%Y$UbxR!s+>ll<%<%h3}I2PIpg~EBF zJBpk0d*l`T*2pY=J|DL#T=HljM24-&7y@j1xX+$}!?wO42I`**Kl_t8Gu;1pHBp**ooc5Y^|(l|S6lNA#^cNu0#t#OYN zEtWTd|ETFB zh4fXZvdRk48rGk)^=ZnG&E7+=CCfy+VW>z{=D|P!3v`e*3F$`qbV(G++@L%NBFgqQ zI`(+h`bTf)2hPgCaCd6K_t)_?_QY$SI-#S~B^mx@{&kUPhq7=V;uH|KQ% zDTV=#P??N}*2xR_J(%bV)Ck#hg=m}Pl(-WXo$(x&;WUHzp7yjV&G`nehIepYiE2;~ zJ=2aw%cIb<6P8Dtv)d;tEKgm#`>(Uj{T9wb#)3nI`8I|OS-sZK4(D4lce_Esss3=T zXv7KgY*Qs`M7NdNR%Wlb#2ITls~n>eZqko7V#K5kN9?e|$dt(LKl+*H$UR@6!QttX zg-kf4Y1gS@d9jj=klXXt1NBoOFNwLZ7=<|pV9n^DqEXnc>|1Q_QSb#l$__9uXvof| zECXiq_mn$bjkE5uuf5$&HcTTqCZ;`wVzyK84+GLSoeKTF49!)N!_&qI3PRgt8{FQ1 z|EhAArZ~JqxDvK+$a8q5p2)p5DUY4q12?#Rc7C?wHfQ;Pk5jF|-6fSxrEl~JjOdt~ z^moU}Y7XZ+WthNrE5&S~NE{Wl$E_LY<1*+jAqsRpq))4kIB$C$91!3nME1c+!Ra%Q zSdZ9e_e{nXE5uSIF+b{g&j^gPNA=UlT@C{y#w7eUOfiosGDt-YElS{P6J4@pNM%tC zVuy*!BLD5Yw2)0c{rqL3yYfs?3>2p!i;7Ow32GlBPA$ChfMg(<+Zh3qvWuisnI^aj zYKg4`*9noF2QMz>8*WEMac%O9Ea1mLhRZ@F5YMR!#rs6a5p|zlv1r5zk0Y7fF4;%? z22X8{rkk{eW~!NwMBqnT1r0yH&sYyWok>}-R>_L!Q%Iuig(C%4uhrC z1T1?f21?9!QBis7z4}-jBwpftsQJq=^RK7@IT)hf>48FhnVP%)t3k^(Ppwyh)yIR% z-Lpcffi8E%sl%lMfa1W!nJ)%ZLRSc{ntME7kRY!G|EQY29oVQC;_DcFKo^%bQt8KZ zqoP+3!yI=|A&cb~o&}67H{yUZLj`*tD2+B`TYSK$WGlPnN$xuNq{PJHWKj$Vm84No zr~`|T(5VJps=T}C_Y-{DWP`xfrBfkkQ%`V>EQ>}`IV5DNhEc3rQ|0|Wt{0%Aol3XZ zJ^pOctt^1@+}LE&kezofy>;fbuNZ-{YRmV3M5?Bdc9SREKryE%P??Itbg^u@QFh%$ z?@+dYWYk)c9}e##IBH#VgC}~OWf3(pWK&O6rjyNrG{GJ9a9I3|60fQSJ3<;XjdE?1 zu-FS61zvv02~dwYrAnSrtjGkm8*MtQ)uhsGa-bh=lcz(pHBRs@W>?4ZvR&4QE(GW^ zZ6R;peWypeOEF}~l}NjxxGkQr4Yy9DE1v6{7am~OSFF4Q?yd{`=K9r7#YRM}OS$qN zWFv>&7pSKkwu+unEChp>!>@dMhEX8b~$XY6@ zc6JxouYM&=g`UVZlB|$-@xMuq*KjwP&N4_|KhXE)Y$G5v8`o8log6k8M@%}ufMRkf zU<-=snte2&L)afYg_ll?giNtUi zF944*1I!MJNy4S8sCMblbmW^|r%Y51V?R-=sl2tk7<7JF3Zs z5?4Gv;)F_`7w0BHM0a&qiVtKf5FnkTQvBq0Saw|sd@ug1j38<{k)iI|6BuZX_&=ox1W8UK)7ki`l;Q$Jh(tr~OwY6~>c(Sdc++hO~8*E}AC z4Y_L1Xr4$jG`HP1z|(caSpC9>VeP|zVIvMW$?oBEwzTMVsd(sBUo5Ws>Zjjc|L^&8 z3MKfN2zDXUqm{SDU#Dsm$C87}?72O1(DH~(@xg2Q6i~?wp9lTwplIhNK|PNOmv}*g zBe_|uMX>{|#klVPD-=`r*TwzX&o~+W-M+d_WcM@zs;6V-oW&HgpCScRR1=KKyLQBtKYPJM}+|s5$nXyFR3o-SQxJ z|7B#kN&3_Sis_}uT`CF_50;7I1Ryw-?zYk`i{4Fba?x%H+BWC7dc+AJ&Ob5ZQ%HaU zwpLExuBjnnK=rUakb{>A>Mh z_V0yn(q__!yxU|e-AX;tx6;Y*L^PkD-DzTT&r!@L6ltWQ>gj>`v6AC-m7q)b z6ruqsk_X|fklPIi>PO2#=lk+SmFI(em}n@v^DZ*K(*{4 zHJ*&>ZRJkrMCk=U9g*fTv6w~jMpeU9C0sn0|E9spB&x>a#T<)MEvda z*h^>mM2pVKN7U!&d!ZHH*F0K6j|S`sEm3?d>;u-j(L32Wk9IT5V=T0VO;$GgzYbjS zebuR!abQN`b-m18qp%>j zJfC_dLvr1@S&wFdD0Il9nM3?~NjATd&Ia={#E%suDz8VzgqOKxhYveF{_={syCMx+ zTlppP3`aLcUiHRfRe{I6;ufY$vcq$D$Kcnjmvt(#!`ayl0KjCK4z^fh1uN=d?$^q` zY=qU1Z#Jip!|X;d4m(b*CLlaPF~G=LM@8L`cS0;1dCj&=FQn655BnZbS1iU#rs8mn z=%CbCkq6K&fa^0P-58RisrMgo>Iwdohb+~|i&x>YI;1uLGxhrUC0-AFQ|YvTqL4kI zx5T9`I#rLj8hkBW=ZY5fh%br{Ly|ChOz4zV%x)JSj=;3y3E4u&LXR!ro?Y*pT-^Ms z6W)yqQ(@*q7tpGfdbX&V#jLHvvxj2ymQ}rMbL+O>$jZP>z5b{FX*RE_{(`JtTByP9 zR8}pB6%>mzG$ZOB`Z_;)Q630M@ATOnau`yrdceH&z$)gD>rOB-hV=1EnmECTI(m^# z)#28}2Qs{ner0(Wu3;cEC(g(mf9#Vh!l$vMzKN2jCV@4BCsOpkoO_4VMUM9{@}0&G5{f#yV%Av zmQ7FMeU>&)n&~0e)#8gh3@;5kr3s!& zpGwodR3m7fT`kC-o9cfwyw9~Y@``jnzf(CFUN0#C8mn4mkAI5l7+KH5&s2yO<@h0O zQ)gIFU^fs_*9NQ9C_xUHmcNZ++Y&b~w4e2N53mt66I_{*uYDE%Q%&Y&Lyy%R&S-n@$>~7E~x4#uL z7?}y%G8=8x%Wi83{qKLDNJR^WXI>U6T99)FXI@a(0^${r5ZJF+DX*rR+&3~9E7De~ zcQM_;28Sn>7cF|=Tj8>P!5L9iz#(aytY7g-fTJ*?@2iObKu1KE`Hwd&=Ss}=Ynky+5Hp7b8eiwL2FOuxJX(2m2cFsQJJfW$m zmH#jspRJXx+kTm4hhiL$le=%+NPFAs1ze@gCRtxMas)I*QH)^IOq~kFn2)JrCA-}UA(ijK;&*Gj zxxz`cYv zhTsN%;B%5J=kN-3nn{RuGsSG8NDLKqPrN~SO`$FIOr4HQK%g3a*gwsC$E3`W%q+OI z-E4N7!*|(PvmL!pvc7Eu#m%sqd*owwaaj(}S-MRi^eM%hqsS*z6y^bAloSQVb*c=_ z&G`n6Mm%0CIWLV48+5(n+DyksbSf+9R7o$S>$RvlRR$&sjPB=3N$zR_)UP}8k*`kmt920@WWP#O zrueK4$XHw}Z4~NMm`Yc~OHzw_)yEVusJ#=k8JvVGsZxCBQm9D6WQeA~TS5eJH&S{$-+MUysvGxj(O7}*aa zRsrCYE7xHxB8##GV3DytPil<=5 zOW3`aoSheozESeCw~cN{@r)-wA;8DmWrBS>15 zS3HzSmq@Wz?5-;cE6=D?`(9#3eM`%K~|&l7Np)gIQDz;qN3Wi|&yZhq-TI-?+Wfrm7el1!3p44tXJRz3 zQp_cav;k2#WcTY-cV|V5Vgy5Unrc{ylbk`A{Wbuz{N^_=OGnSuL80wazaHB3jx7Ft zk1l^qrAt#?6DRT3@H&+b!Y)9{HzuDXcwdh^$U8yj%u7-A(tGJl(N^!=Z(Nqf2|f!R z23@&jGj_UnL#p@DH+IpZ-xJHrqbYrF(Z4reP>n;@x(Vu1c8pNhI0T&!tVKA;D~ zA;9OCEPEh(KtJ$BiXffpJ{`l?-r%9Smx0z@TggugFK$3X~A=g4yYXT4$z7M|VhxGl?A7;ka$zmiw~ySWY$XYpY{^XLd@ zl$Svc-Y?+K)`4Y}9zv{#jbqyt1p_bKde4=nZ~TX0xx$j&ENfn$o( z)9|`ioifufY$PgotCxcYUzPX6V9?xalIQa<5igd9oIyYZT`@bsr`NR@Mxaimc4bIP z$Z}DoctF`JfOB{mGWBKw7aP(IS>r>n?KZY}G4UNnci4WZ71y)Fi|FX?|9Rc$p|pMD zW)>;=!g#^WCcaA@#Q^(BB^C8#>7^yTOALR<%otVzVFL5s?_1>0j$p{(z274>O(k9E zf0AtSS;>2{G?8DS8g?24Qt~0?uv5UhiIFnd1$3j}GyAjxTNW>gY;W@T@6pLr{STLr-Y*P{ zHknk7ESt&uEln^$XOXQT2h=^mePIPI&3;?GJA~(X!-@p& zL(h1KYyo3MbIB}sPnQs>W?dxJ9rHeohlp7wMO0u)oG3ccUx=Z z&G|?CuF$*aey9b{;WY)73uEWCM)nDgsh@fl!}-RbI00viM*hAub?S5@25b2e*Za%BCxWx`VM`w#4_ou!p|EE1J7?W;OkpXD3K0?BuNy^-kB# zusf-e#rK+Rb^Z$)vhyi+!`%5TNxB)M!`y4<-y;d^>=TExCyGpvltD4T6R3k=51p~N zMpLV9k{#RZVRaD$fvLdEe*VHrE}B$G{_B*lxgg)W6IFl`@ls{*n| zM^`dTU>LJAj7*ywcrU%vd^J1EMQAgATE@f^TnMQH^jMm7l-b3_#?@JN8=@p{o%;DK<{ceOE z^ji^lX5J-PFRaERYu)5nIjeABI;nAJ|Bp6q*}`3Z=Ua2lGkMu$dUM!lSjhYRoD>oB zHdf6^3eck9V+Z~J!_uR*a2f5!$@-L3hFx_xnvK7&*?FF9nMMjt-o)J$1CMtn6@`k1 z2S|xmo6I0Z2Wq{@VS%{}C0;;n60ka?3v_X}x-~+Ys@71R+9o^bevUZaQd|3YOwJ>H3;9$AK{w$3GGu1%@I?h!$3fg? zXpdsuK@hhL?YqK%+2U1g7D#|b{ z*1v|b6ZaKakl|#gWF2w3_j;@Z*~uZC;k_rcg@=noI=9tf+7@|jK&%ASUJq-^L8lZ? ztW(}n_9{_{J{Cx#unEI#J}acgYs9HZ9vxN{kOefnyQRlMi^H?&!~U(XMhOguS3D}| zYs!-ZWKH4JASn*-lGW3huxxM*>{^<0f=?33(4^8^{Ij7xMW?C=O#o4hXPFT>@B`Xp zXUHa(lSGU1K)cCS0sdfZp`wB4+|ENqOPTbx5(%NP5JjhY=bAE+&)EnpD+EscTEmxr zB{3o}UAXlsiDfsV=J3P{4BD9Ju8v|-D6*Z3>I@t48wpMI8x(I38u7dHy(~VOte%K^ zNry|iB!5=Ig4JOMAcUW&EQcke5kE&*uW<|ydxJJkBMvz9mz=ozm%Bh5o+DXsfl`C^ zim|Rf5oD?Rfap7MI;O<*3Nge#tT^P{J@<$T>1Jwpc>W9|T;)?Eew|PPmn*sqpB?mV z_r$>KU}78x5cZfD^?mf84nTvUNy{dtcI{v0n6G48SORW>I^&Aj@otqqxTu>5=H9?~ z0lp{-v>C`~JCz$FQWinAp9mQ+^u+ry?s$E%A`8?wA);90t z>mi_~%oPo}b}AEm45vn%TH&HL+4bQ4taYhPp5_F5#-y8`+cEhjb~~1PW*ql}yYrUS zf0%#f>&8*(w`Y@oKtB4yj7nEcMx{?Erja72A;|ER%fM#X2Quwg*|-rSo+2uI$E@MMw2&${QmBbpM5jytH-~d_vYYtZZGL0pvE4u z3538fXJR+M3&aV|E<8`?(ZgXosqSxge4`~28*Qd@JkCb8Ms6o%lK2@n=O6#lF@7B= z78JO&M($g5XknVi1JCOVk~9@>z~VS$C9}3~4t+Vsye;frjVV)P&Hq)UFrq4T)~|m@ z5~q=3lkUi*m|YamH%0Y>Cz>lc<`pYI%AFR~g@A5hAxxQ(%@(3R`Qf-RK<@hEXEl8l z%+(2~HGTJ>tZw#>85`$0c5!$j518#kV{`o%#QJ=2GV(XCdhRcdbJ;Ymej-70zdi-H zi`G^x*b$Nn1|-{6+ZWLd!z|K}fe!j6pFWRCkVCf_xNJ~p!?o;y`-gk^qCle^x$)iG zMI>h$0XF_IlZ`TpDW$-+M5T#36-nf(vX@sU*yU5~l`MMxh@oh%k8WRjS&9`OFlp7_ zH5h*1Ia!n}TCoVJX#0R&JLxMG5|eXKBENpoNxNoN+T2BB4D;ukHgtuhTUd1wcW0%# z3s-k7F=FRCNB?0zDd%v+=Yk2SPf-j=lpX_6^R5NgL$Oj*(1^Ig#ZXe9(;V?@jl?>V z4k5Z9y@9)AEs^`)j(MZyt>J$<8`(@B7Lq>~LP{xA9_-5l-t8`}UIp4#VlB37je@<}gJLQc*h=Hb}Dg$AzPxiAq>n z6K2v|XO=)M0zv}MwniQxk4UVi;acFYg242H*UmwK@ZK5C{=H1k{B8}lusW>e^D9?} zosC@m^&3z=T<8%i*%63mZ34?U9cFj@jPo6~9Abx`;jXQ({cX`hWN7pZBe)>0XY_9M z?1iH74$!c`)L{_bu3BI(h*;BotYBnKgLb+`TmS8)uR6D8n+%mytHgW7z#cR3TG#9t z{$6o9B!wdeMuLmQ+7@1n3*>&wt_8O`FpOu?C9FDJKb)A{?ywzJc4O+#7QK1Tyuyi1 zP7QZMPvnTnC{;i)xfIEUR8#&vahvRt-(62FtTgcrbE`C1?J|&(z56b%C?Vf_n=&nY zON6#fS>;(yzl%`I6Sb&XRF|apJ?2l zeZ}aftlIMZACan8ERmw!1QiVwbBZD#8+2=s=T#nynJ3-NG!P# ziV<-vjyj?)S^z4teNc#ilX;ABwEL)`?|Dhzz$^@=vzqJAH-I@2vyhC0Yu1^KO$NiQ_!-15L>w39H!`@lXbV-w5 z2^QT>yx0NnsL4aH`N5mCV*ee>4}P451ny3^>wkJD!hH3Tvt`;ARw|q1M*TF1MDNmM z^EwImEs0iXclkRiUCU+| zKQj;Z$1t?W3dU1QL)?CCu6SfYDrmW&Gi+FKM$@#!kUx;(bCV1yvqE!1ix%fYSzy=f zqX8X4Sa}4F1;c`JYoyMnGYm6U+QA9!)1=J6D;;QGCe|ko=?;CPtonqzrscptcK+17 z@PMES42UV* zY=W+76Asy91>LE$zU+3}&FJ+0tew0=;>UA(Ih;BJJVRp^HPR^tij-0?l_ zOZ(qH`OQ!N@GI$RidjLC=x2kBr7ee`QBIEGVHBip#r2N_@sqkAVyyV+fC!*`Vmbm4|Z?Jt2~+@&)jj) zu*sM_1In{rwDXo_P>kbZPpR3mPi0>EXrX3M@A;+rpg>*Xe1p-R9J6iyfUJ7*<~L2k zKC36sUHj}>Fq}NqIKHf06L=KrjmAL^cYi+=SB)KCHc?CrMb=SK*QZ~f4mP7|0seUx zXE<53V-sq|$DCF0bwm-t_&?|Kqn6*(A;#_qZi0OWiwF8{Cd+Ff|rjzFbLnFz%-MH_0K{GDn$7?@)g##wwsH zavA`Ktfp~-wPbZzCfymfSB!iaAjqH(ZPlvrRxSPojD#-vygJCR-N+=*fVMi7_vxdW z6SMU+V;PzwnkIP_$#RC|Y^Rg|`tWP*OY$k))2X6Gd0urSKD?1$?N!@BZyBVRhZO0fqOK|v_=(B} zey0ivVNxZ>LOa~@A?%+ZDO)%gd^NZVSf#XEl*>fOMbW9Rk3$LA@__OHc(v}zF*Z|1 z9}3Ks-16vG6!3dO;qV@}N_DQJjD{tsR9gG7^SiyY_LR@NShw3BUbZkTq#NHDP&&ah zt`1?kce@z*)~n$Q9@EQ2-HOMqXF}e6%-ak34_a7Ipi`8WLUpP^aW9?COQLr0x4c;o zv$;coO{#8rf#0XR4q=~QEw4nnD;y-X-D7!uq`(j6bdtmV@M}85j`;n#l3q(vCBq8N z*0p@)n@?*d!XCrnanJ&L3>F@AK~`xhog#wVy~?1Xg<15VY$Kz^^-7&8E9Cv`xhFI_ zO|<9`WJ)wDRzPpI2VD`y0>|f|%yAwy+PDcAS_>x9K8Jl;H%~W*iG}A|O{aO+)5t8Z zQ&q!*-7>w9eky2&^l3;oQ49vh!W$edLQeTD)4L%7C5N}n`TopUUU6uRmt!C91fH*b zo}%SeSPf*{joy{Nea6jP{(`gfZVS2`4FN;SvkUizR@2Qv)pWA2-jRF!<*l$dojrH5 zbxv3q1k2!Z#2HqAo$~eH{VH*>(ZOl?(aKCx#?HavF5<~~lZn8`6a%tcA5l>Q&Ml-R z^7hX!OFxY~yz~m_YhZE63PF>&Z*g|`^#$kY?C@A#Yvgj0$^S&s&+iepMwYt|IOm4e z(}$pDCB|jI`Iho-@TzIWibQ^DV0YjV_3CM@k-hYC(o25=F2!z=Ku&0ms2>KmM&5E= zr_9z20;NqoozJhP%Y|2jXmKZEPbV-Gz1;nu03@Xm-`W4F$M&vzJ`7Klru%dV*M0l4 zv?VeD>Y$@3#K;|WI=`4krxN?JH8NUMEY1X0l>YESMZCbUmF{ST0{v>mt*nd= zHIT3+ZX(Jk9G+EK5I4ufypPEP*PRi1!T?DHWaj-}jB+fnemg7p59T^dUyuwt3vyDJ zp&2Vd*)|A6dFDvB z&9t7r0`l)>p|XzsaAFe}EQ}L)cw*3eb~h_mcdGMmK09fi6T?~PSjc}#v!H`O-O5Gk}6Ac-L;qA;ZwZi(aV>vo;^~Vw`{_Q&EIDI zjjS+2{eI!Ke>We0IExVrE(gk4BYn^@GDy&G-He1Zn5UoQyVb3gSxYS!T_v~OaItjA z?XV_s`a}XCEINXHf|T`*PvA;c0GXna6@UNx#vt9=-<3Zkr`ZMRIPAgRHVM+TQw-E# zpQoZQH!}lh9AFvfB*B?azp}&^c67W0>J}9WRp?ZgeGY20CB8%MI#rjjTY-{ZZL(rm zo=l&eAwi%;2i}6{UWd@2N{~r+)0h%o;ex{0&}=$zOru;u0W5ETnj|NCt#keB<>aemc+5M zLmXZ#*lS{k(kKRUwUVi*4&fD#olr~P$VZ*g*2psYJnuMtRvs@o1XXBQ+>jq~LQyHs zb^))Ne^$P7(LJdZhG61dXA7Vv{;e->Vg3g{`%rF#$cb-#`x4najqEphw!0_>f}}fu zdKY9^V}p{&#u-?A))u7Afshs^XpK1GTUyD-oUDzE;SELGt&MhD>3S0quC$JR#I`(L zy)l0I1DenktX@~{j`|zU`Zb%2X0eI&ayYfcg2YyZOLwS&-2*7|G5@fT*T!U!Oid#l z=iDaNu9Mg4OJMCceSR6v*h2he?b+*mYx2LnVNOP7LGKsKkk4qINUwRE7q1F1RH$!& zDaMd%nnYVGIWIk~S~;^>3~~_XLea!b&ckKa>*2kq9agr3yKmT%pUP*NGq`Zp zSIsaP`Lp%SAdD7|kB@pSU0Ky9Z=wu>pn{tfR*_Q=G6stm%yo z-eZN>DgEE@keX+*SV#dG;vax%A+iuei;hI-R0sJ_=ba}gepwMT?9`<=>@)1N!RHR> zUUoqm+$Nvv3;KYa>45v6wS`N2{8t42MgNJ@&ZpoL4!IhxXqP@!te`GtbEN9@y^3`P5 zc$Kc)#W$%o8JLPF215V&RMb1Myz_3lFBv~NgrEo@ZBi9_SA%gZ;5RDpGtWOqQ%zF+ zt4Vj{1JFuoQXN#?^Q;Eaxl5unXS`yKXYq8r;+X3}{vF^1Kyz!Wp|jYlwt&Xm$l<=l zFR<~1Wzd{E>AKEJozJwui?HDK#Irn(yKN!i!lZYCS1fjH6FV!;%O*-ioGqpmN=2`GBa z@OMjOrT7s+LZ5oN=(SwGqSxvrS;*QAlbxK&67qoVVfik;s6*{~?hAsB;Q#)gPt3z5 zoSpJm2!Ld2k{6{)ZqDBXj>j=|tYnii)xVy;@XznxdZXdnO>b>@tMD78-%NFgbe083 zu>-AzMjUXM9cWu?OIMkDsurBTa+kCaBrHzjbt-{(El;$I+466to+n9?CPPyL2V;3v zB%ePh%hYty(Jpo$vB!T_?7G1bBesR##U>!h-B9a-(miiE8%LJ<*Q-AyE64LxIXsgA zeTp%hED017Pmx$E3aNS^kXqrJ6xruL0C@&OuG%))2Di$fHsurdNy6?Za;(Sj=pB>u zY1x!~clRHE=wpOM{>YC?N!k}?C9l#1F#9P6xS(=zC9hX`a9%q9nCpGIRsd>^?icvU zy!$kAN*2PAJHmR&9!-O1YvhH1-O?J>KEZLiUYajELGNCO^zWIPJ#@9~faEiIgJ(8f zA?Q@*^Bn>+>%lbMbjEPZ5%Aiz^!^TW_6^Q@Ar_c6P`lWmh*di~#YKC~BTwXzK785p zx2BI@)`la_TQ(TuOsA$?Smyi4eAdEd)a34b%+lm5UGhk|ZQeG?N3Uf$H_?yexju1% zG-Y8#Yh;DXg~cmC#+FAMAW zsV>(SbSsX^S|hJ7SWdnFKUw^JAgp>{yo`GPFNW<4i~sWe2ZqfUCe`<}Fw57$0}vY= zjN-v@PQ){CSsrS>y>;-9=IPLE3xIXtT zs<*U@6MWvS4){!7NDo6X-vEImM7l!Mt8hbf1i^WbYSda(l{T za$89;px$waiu&-H^6-*}3Z(4W<91wq%xi#G9##w_76%udleL4yYFS8QP@{0G_lnn& z0uo*;qe>&zG7n&!LaDJvVXfeJ#PtPfy!FmyR6i-4l|;^ow@kktJOl;6nB$NR>Z-@c zanh|o%~ez=w$GqtgMl#}qMqNgi`n`0*ZWoG`~(&X`x8OLur;zu)+qc`TqZf7=%hcP zTP9UeY4l%g)x#XUcoIJ?5A@^ADY@(Sy}sl^?3ay3<;ORhQ^?`*0;$~n{lI}aHjsLP zVnD*Pj*40}?RM~e`XT&TJ*`?;Pw$v{j37l=iK5Q4!sW1H9TasAEAG=Ni&iqNkstve zjRo>0L*`V1{skp>wH>*4g{10bdXc3!5PC~jc!Pbfx&OTx&}B7VxaE4 zMiwtP6nI=!9dtyMA3nryk=F+7=I4i_*3bkjidD#e-c(q5#P(OkGV*U8|3naB1k<)` zMH(sPu#91g32=^4%tsXDk3C#!NOO4X&42dKm#~r zjV29>h%+^(lAj>|(27GL9eUNT9HS33p?eXHN{Kyy}9v0|aL&=|0V}Zs7{>HyXw9 zGakYWXRl!Wykq1Pzg&DGVm%Nf?Ur^*tS{WmfHD;olWMY;+%CpEEIoZ zJw$9!&Eg*VvR@DV!7@>qq-%DrqnnI5=QZyXZP|tpP_9{=z+rRC;!42dSZn>Ua7!49qP~bi{dBz!#6=~ri z^WrMc^Su4)4k0#^O*hbkWTjsmtOcdaTqAyBx6BfBc{d~JYRNb?6c{LjT-# z%gA991m;r=6bNKdQK)ydLeL#}I}jy=U?)?9v^l6$bDdv_iW`Nz^$RX69&x%V%@v_M zcOsnL>Uw|_g1R!WRse#AT#*e9|98*~YU1E}IZ$l5;zPk(bIt3xER;sAg!S0w8JTpB zY8Z-iD`vOz+xa?GK6TqYG4Q$)p8dMWwR7~6*vmwpDKhEXo>g-V#GBc3Z^J7&;*{W{ z)8LSyEmiDd3?0!XTc=D>+Iz&9_|qPJg>8AJmhWNrDif}{3#xzksMFhMV^+R5Glryc zc-*Qmu^Rg*22v1msHi4+GH)X@nscDjU}9t!B>p$jwX>nFnTA|y6hzgjP<&lm3ZlM- zE81kSf?+wjI{k{25Xfp8B&XfFAvv~-h8_D`4?98GGA3<2!wSk%UQ?f%`5Pl*-Y7Ua zlid2kj7n=wRHTL|<`G36P*MG$dzPrY!0&TK@ymURGp~W$)9JC?KT(OIONX2xpf4Pf zq2q%z#Bl=T4~!2_rMC*2FmJA-4Fz+GBIw26ay2vJSqzL z{Z6Z@^@{LF0~0O6f_JPKF-Y`&`o`rY==vb;%0ON$S5h5O2=05G0y+OG`PC5>vt#&) z({Zbi*CfXtu5?In*)HbIMicg;+v?(?<*zy+;AvH?q>t_rR?Mz&LH$OZ>O3ztV5I=3 zYDl`@1uK_Nr6-wfbB?QXco!BwkpfAR;yetC1M~YFot0_3wRlmt+2I_kfsKmY{nd5m zi89Y8^K#gNSjg%bo^uhFB(+&eH1Rqpk4&78IhpqNZ(iDC)>(*`He>e-EaOD|lc;^> z%SvprGdVm3x3D^fEa~X8j5w7^Z!5E9S(@8{jfw-L1h`@+9WvGdX3~}lkXj)D0}n#7Q}fLO_WdOy6| zXz%m+t#JJnX+AiLJAL-?V}lwzAM9C|QW~)v9AaH3}V9RQG`yx_;@k(RsI2Puf)D_0+S2-r%Mk zAouBf9_sMo8f>n}FotO>`9rQlK=yP{k}GKgZ`d*E5_Um7J3w1?wa(dxyy}TkEuIyg z2hN%QT#ICnZ?{(xJD;qcU>hN!W z?_?aB@<03K2PB%q0X?0`m>EYg8!57$iYi@r2EzM?7sLo2^Sgk?9!K|F=i4x?FClXB@AQ22mb2ZFE?4U@k~zqo{h0bR@J*7A<<$0pZ#gbf-1A z74fn!J(QC?aCqDiXFo<`(kteXIF(9s=oCnbK-lhu3%YW8Itj+zntwA7xxj*vKIhE0{|$Km?aTMPY*W4f)3;4(eAiX*gC=CEhF6sYaYoYYvkU z^pltEvfIHi?xhir=OZI+qZ}Cwo1K@-KJ)DK)T!_m$2l{$LdysRiSJN?~)38WvXwRMozNVO2zm7Cl*GPgZ%&2mf}Z2V4TiJS2oalp}#b9dE3qh(ooe%nW6 zJv+aI!@<-&CN?LTVzyBvfr>f;vVWM7i=2^KOmfN-9p$Bkq|&K@m`tK~N(|Wo_Ou%> z>^hd#^>5a;p%<7|8!8+Vkf?J>#psexA z6YZy)!doK~X0-SO zzN>Bg0vm5*XVmgit6q}{e*E}ACX)i2pSX5ft* zWb4u_et-Ciz?iS#=jyOeJhC)5=dTWXL{2E~s6n_;ol2LwJRr^|((cwCRf=-hem>{ES@`92j4J5>ZhPTtHuAP(#koSC^g49~Yl_QKzL=nUhv zm!z9!p?D&;fcIYUJ`;%d7Sp@jFEMdn?h~#KYl*z#oypJQuMRsKX^s116Yk6{4BLGx zU}_n6QU84L?TN%)IPCgbh`XE;pOK!CCirY*THKK*Y^C3Bx7NsH-cI+b>2ZSCp!=^c zC&l7|iA~t7=#w#*u!GH*8y&oP{(4f1yU_ssea+7EWDAGmE`=t>W;ey8QDi5oZZrvj zymk>*ZY3(?NKs&kmo}E%2*uswush;}EDKOT53FubAgj6^m=cl_(%?A>BNe-Te`XgP z-0f2=n~QPcFjKeXuUcfTD)<*6AXE}E&&%14O(_+GNCh%nEN`~B8R9Si+ zY$(|55Vim{B~l^`D=?=8_j~9x!F4_!)2Z&y>Tt`V�T#Xrtf;oeGI3JLtOk>3nDr zD6H@cp$7VlbjU5qw_8)G-o>!aonwQMXQsx6t+4>GVw;t*neyq{r-?yEM6FBg+fNEO zjHr_)h^n9%$g4j{MO{_y(%cMb;vMtaEk#mWEEtB&yrFNc66iElM7z~5U2xlK-)$pTGRuXEInRvqCr`ecZQ}F#$jh!(56u4v zSotyIDMvKPD~e)~U9yMdC?0!&D z$A^VX;STkHqQ-O0thJ)Dh+P_;s$5zq$qP6uZ$W_T1Q@j>oHQt z&x;sxUnA;udjM;uZA#2~8f~}2<;2`VNxEyQf4{97?AC+F26sh&3ro}(FSiVyapp&G(bN-<8NG07&CP66H=LmjbgD73-+um`RUUt+^8?0~z>WlDv=(Y2|6V{;kF z;IL~`YvS6JPz(fm_fkUr7*NOW=hq1iE*eta2`iu7 zplOu%(5II^B3EYgc-{v^nJSMF8vWYQ{n(&FO|fDpu(KtG^tvvaQ8C-0zSl&cV!5Ly zeufoPsM81AzH3hT#@Xx!3nXsac+0@XtP04a_3>I>hY(e%lGOc+l4a$yVy4&dh7~c> zb*c@#0dck$`bRB6TNky-Zp|!ZdR=g7u{+*KY&)6Di+XS(nEBk$ zyyy#TJjV(%Q#}vu+xZ=%d3ocb{(mMX#!GkLa2)1_iOIP@G3P1L0^weHy5y?(x-y%- zIbW;uDVkgA`BZ+(HCb5di8S2W!%zr-qZ6F1B6#t6b(V8}c$RZoNM%sg(thFk1$X7y zbB{|`x@`%nq;nS634j+&i~kLRtKTF?&!|#ZMYaU#&pl88M?A81BeNxbUz;q6Xpw2P zNsa?ue0T@BFWaPiw4{AW6}c?ED{uH-zi`CqAiqR>S6(suXh5FyCHYr&0o>=vW3}-@ zC|U-b^AFa2`l^dXln0b}y_*UPOt79Ejs$TPE=B$qdBx$y;hjMJwZVBalo{bH(1ASK z&pm$DPg>hmtQ?mc={oUs9V6!a1;WU_B6RWg?Vv_L3LS94_WPPs)YykOC zooa}F;90-mB-0X_3C>R})2ui+uS=Fn4~F;B1(Uay%@$0KfzI=-`d_DhU{3YLS$D<) zdDn4$0!-tAs@W~_ILR9Ke(8P^`3XDTUEnO zc+U-wJef}QiJ;FT)j!qWVIDAB0GJOQ&v%`Djr-PH~;0!tK0T zucvw#INEn<&e9)g)`35h$ltFm?AlmZ_|DW|$2E8sxF`GX1d9Elg*u;FNzQ8r zNR_w4MlMUQxwV~d>3&BYw>+Z7RR#_AWO~(3 zF&8Luo{Cz|bSnCQTPjx)C+HThix_l`@oWVD`oxSb*?zxXdM_Olafd8Daor5T!N$mdAo)sj6>NLTE+j~);g zg=7GyNIhr`oacQiy98ROoysm*oM7*a3yY5i49Yh708{lWysK4J=m9K@^4rQp<&fLv zpv%Gp$yTea;-n+SO6zCYn2j^pr@zz>dRQcW$Lu={XenPXpR(reY1sZ;;=a?Ffm7|6a%zqNmLZ3e;P-OV(~>D z7VoA?wDB`ays$QRK>R2IMm=QhHSpox58F05vf8#G#OjW=`9gLG`Sa{QZdMu*vMoN~ zQ?ixA2q`f^NEXHHrbrqUwZmsSgMUtF&PZF_vjQ(d-jBX4G9o^rMGpJCOu7t|h}U>F z(A7alR5f0$kwdQaboRn!qQT&+!FD$urhR7{P|(#dZO7hwtbjsYx&G^aFgGP^0z({L z{IXzpt}AQxt~F?%m3qd4zV((MZJ%7{voWM9Ae%m+(Vl=c3p|rNLz@gDXK60WLT*fg z)U}OxvNwF5Ywta;yx*GFCs;`AH?ZWc1)<$i&#i8aC?hwye9ogk-kN7QdhKL>vU)eU z8*@E)cD~Xa5^RhHcmHL?0jzjyW9Dt1)p;iYaD4MGB~>QxGjS=-uN2LbT{CaI3}!CHQpmaE&2RaDP@3a5-R! zTsH3pY5smIu(=wxLlSPnVpHY?0Rnq^PjW|Efol{b`S^LCs5p zAP(?cj`30g=NeJ-=MOr|$sTs7;js5{%mg)uDCPh~ihyHNT?lo%4}3u*Rh-UW7f}s? zI8-LbAY7cFM?S3R0*0DSWr^>QyS6{PY$2u<=gacckhP_`4C_=F8N>ivT1bWlL;qUL zp2vg0c4TA@w`=aPy`W=&*H}|FvF)orN&cES@iS)^z%7tG$MSB@hh=VA#cab&c;L0J z*-i2c4T52n0CfjC=(}=^tsYm^$XY;lHBZ{9#L#LcqzGf)4X#EVbMBI50ofJq=Zb*l z2n6@A47USfzAdU2yDBu=6%7*ufXUtL9INRwW1W`WVNg{E;7(k~9v7 zPAW}Cnf(+4lipk^>fRfdrTI&)Deo_N2o;eR0xt%fBsb@$(&5IH3VJA~h zXDnVdBU_!M$>Hbtw)x$FTHFtl$pLAqdZSx05X$3C1L5`bXG?a$t&TtjTM&%vq&NT% ztn3N(^PsNwsYb}GToZT{mfBb~z+ro`(*$6fC?!_%7_dOokr@@8198jh2f>h!E zn+}s;TsuLML$W>y>mpUYf~KtCAdHre;4YqTO*GKTo%`OT>?&m2B3GY6heFj@?Oq4 zk~V%~-EV8>zWhE`c85xgUK(yLjPyMJ5Qlvf3-U)@`V?F}KXjo7kZv})V}gMe11fa} zd_WEq*7IlcIwou3ialap-rM%L!ZKnSo%h}fHI7PR=Ow$zes-f0hl93fOh%<@im9N; zVJfPNZgP(k3wzuu=P3@iOf(y8Upr8U8K!M02 zD!2^FJ}Qom3W^%R4Fth)Py|H=75?u_;z%OFoRIL(c<*n1BPVAIobNgB_kQp4JgSeU zSCNCtNA6f3xxzhXW}Mf9fLwZ|ODiAgT{~c*q)!-JQ6}+nmHGY$B!{PKhvRi^@;rL0 zCr0y70`fdL;b-*>G;;9SgKLy2$7I*Y`&+wxOZH!??;9W_^7{QGQqK-T4s3dE8$swA z#Q=$C6D*ja2p?J`Hj%sAo4d{klP6@v=}r_m1E%)0ul2G`hXuM5Sk@; zCu9p%V?l1;Z1mF$={-HA&>?ROh`v zmG6}uhI*A6A7K6kot$27d4x@@#~4p{(j+U#eaw5>;-L9?YqY}K0>OW5ax)mBUtT7Q zNU{Sj$#)rH~7lE1!GAwU}Gs(il1< zObo#waJ6KgDBZbz_BHVXX^#8LA4&pvYb-2AkEchrk|34j}_R> zgjEhY_jA%Ar(W6p&_q$JAd!Q!P$sBZ51(2J&2G(-J6?CZF3{L~i5G00u^R+gZn+l- zw$6YF$#$7)E6tHrQwtk2^_S4_VADXb!%jnHLc3V1-A?ByktQ5^3-?1m(HS9Z+?->7 z6JqkLEgfc)%k2J<)vPo5O{cIUroIWA`Z)()L7DMnu+uA5jtxQjwGwg>4L5`Ba62E; zF3DA1;KT|(COzWRuuG&iuqx=PIFYl~V+XXa+8tw-V#?H3*c(*$-%(xvmSNtx8BueG zeB{8sNr%x)a*<*nz5!`7x}M$1?PiG*soJzO@C@wJ-?b>Pl_V9 z(P^Z9-dbK7i4{DCqNaM`PWJ|&_ia$#t53gL9}R3Gu-zwoB);bP zv2twoC|dRTn62xzP<5Eb~hx32&1s(^t$d--tCFArdMU4C1dF8K&zefo8B z3b%vKqd|WOihJ#DZ$2MRmQecqI(u&v|6<1)Q$>pxWO_2Af1yud-goJZ#7^t`q4)V& zVaVJWACcjk&d*VPBfl6g(pciha*H!gfNhwy zquEp(oeh2*IJ##>N*xO~u?AO57YZJ}*2uwP2uw5Fov19rh!L|86(g>gn9X$yb`+W%r;kI$=U}fJ* z$^)uqAQuGwY@lnw3riL)58lO%3mWpzrO&vZ2}C9dj02Q-YJBS<5MArGTsjDSFPDXn zL+^)zJRwsr-4d28(y2h6-u>yX2?7DcVR;Ee5dRq>~8=CESlhSNgq}a=JOE zj@~3n1a6Kx`nDu%27XQqS^d&-@A1w?qgW9OP0@2ErcrJ@&Ry($WsmO`|Kf|Ti>!BZ zF>H5&$(XOttHjz5d|nlyI&9NsFC%rH2R6N{EaT0MEai z6+|Ze<+PW|G|Kmacsyoeg2!c-03mG(EacOB&vij5!0BQKEAqU-vJ67cyNm6QADI^w z_C+VBM@P0-ULH{(!q)gSV0Gc;0#Vbj>kt$lrv6xw?y=ts8K#ik8B*9tGzFW@l2kcL z0}XHFGaG&Eg$v6De~cN%_TxT$M|Sq3accMvO$$98HcJ4Jy%7cApOHcqZ5Fn+Wt zkia6&OG{&g`}I&{>g<)BAglasmW!}G(rjbJNceXzn6d`FAlNa(F$BEz@)Qvy5un|= zn0x{i;26y7l_6FJouFTmf6-%S_yGv&b((2EfR|5ZmP{st2ml=VW(oG$mn(Zn91lr9u$<<_-|w(dV0Py8XjoLH z(g4HbG5^>=N*ox5^+tB6mSR9jtXyx0a4C=M^X(ESGe@2fb(`nXfK;Fg`2e(++c-l` z4bm*25jm--2AP%2V09{R#p}%%P~ig^bwI+)>QdnWXy`oetwydIeWm(gQmDXY2sQRN zbaIkm4%cVe>*U z)Gcu#^%sp|A7>A&>vgI+`o`Q%vmQj%^5c0`AOn3=J}|$8v%@Vp#I`^gFFHn;8S9}i z%9Z9#&%gf5f?m_g3x|y(nyID0deC+nWdSgNjDhDgzfS2rI%}%+lgKC!(V|_JBEX`X z+3+Ou2pFY-VbV+g{PTCm6B2Y_J7Xp!_@4UxzhwSm#ll3S6qs39<-0I5LD?kqULQ=UhvZ>4*TsR z#c=r%NxgKNcpKD2Wcn8iZDevRMcN3P@|;cEz|Bv(^0o0$8#!<|*$lN2Hdr2&uZQyE zO5vx0xKy*z5VIT*3lPEbj8 z02W3q<@L?U=j1^lG46?XeX?(Y@nHcl45q<018(VEbbrv!HB~HiSm(ox2I+tsYK(4! z5^)r6yDjYiHNYaOK!E?^&wg^!t%ELbK^?UT8+OYFTtDuX4%m3lW9Nl5&NwsOv>f7D zQV|FCHTD}>haD7?OOb6_ZOY4 zsO|Nuz@YiDFvGWMMg?+DF)bmb1|BWDeGLI3xTq4$+>x#{0@uNn>b4Z$QX?!5IJPH$^mwx5(m^f zz!vIp)H`R)UrAw9cMs>Na|<_1^xpqAY&CRK*F6nrY5c|hcS)K9NAsZ@e?+BKKE>ow zWIGj8Acz$p**9o0#tIITQesfbYL;|_gYX(IZ?PGr!lf{88M4|r3$)xVz8k$i6*a2b zNdrGEaCcy*w8zi(4z7jh80pH6zHMWWm^Uz@BmG;x=sxWx=ClZTQ!89BN%Q0rj(wZ_ z7H>elHF$R*rgTscXJGzX=D_@$b9(4L@_^t~dW_&XcC;qr1BQ*!FyeORfiap-`=jXS ziz0)UlFi@PLJ}R=ODQt)QZy8kL6OZ=OcS%jBW_ATcx_PC)LX*%DH=srXitPrRX-1w zLcRm8XF$OkbAO)%SIe@TTP1b`#fWjRg12ns)%M$F2as>pddf_#$+N)}2X;WrQ2cg4 zoo$=^tmmLC84`e?NdB<`>4wxBq+ zrp9;)lsHC05Fcne)XQdG1pbaxdAR~u!uPoBae4O`T^XV;h{SnDeLHjsluGU-Z3HU0 zJ3T-T3NrIPjXSlkOpTK0& zDZXW*c5ouQ;ayg1>15tnXlv(|MmBP_P(cs8?&*>wULV~Jfv+Jay~6XQpiM7lg+21B zq_>L?`F|ir#-}czv}on0f*@K#RL1-QQN1wPajOM|qPKj8Mry+HKHeWi2(Gi>Kh1%5X`QW2e{>suMb1kwHO!hpaGUm}-~7uY8JI9RT`%bVW!j-7h)on>tlp z8MuwXEA)R9GC;4AuabYr9r7=NKlQ~ZnRZ;aGvsP|ld0Od!`dy7avKr-&7_zu6iK6E z`sP&At5qezeRF#J66beAp;GpA{ij_#{5 z`i*S7mz8;?Mi|V~FVD{1N%lDK%KNktLTV@m{GCHoOe_EPv{g|$6}mP{ILDRU zK=89qq){yQUJ=$$uzNbry%$*Jw{!8lf%%oR#`iGqn#X-9*9@b~$&9-=K1_ttvfVQ)0#(n%18wd2}P(3Tu&leSPy8hZaRhx~7QUg7kDAJhs>AH6ay1cX0}I1*GS zR=3iJWsgI3Ub#vgqftB{^tOd|Lb8lgN^aIk&>~!x5t?|nD%@XW%U&&;Hu`d&Kdscb&(7DR3p0R>sqDS%x z8VU9`eEiSYk~^>)2e4|y<x?Fs;}iqc{fD3Q(5h3oShe542g-)|hd^c$S8n+n;9ca&ZF8%S~A!P#0S?XesEM9>KKj2Ezn%9W^J29#wGxpBUA!7N4a-Y+0 z=NQ2BrwUaKDPjjE2M+xmHv&^R#XvFDek$e^#1Kk6OC$G+k8rZWP;3hcEaQT@J(HZ1 zc_4^HuYA4q7mNG90qRQLr=>a-QvH51XPs{^w^=^ms!_n9U2Yj}`h!^0dH1W`-+-Ni zPMWWM{N>?|+lvmqSsx9XnE{yr*1*Md0PO-F(=V~&v*z5r$)N`LWc%i9B*pCDbR%U`F}cbdD3Zt( zR0-2L`rdeyCoK@6d;v1Wteu+g2{q%AL0MwRK7XhgU&GP*XcQ*|9|}qXd&nZJEw_mc zF$2PJ%}r+RX^Y*TME&gS8wRAjpMUTsIW~pdFnYu~iaAG-GgORTl=DL_5?cf71*+8O zMQOHrqV!R-q?bk}?7Q?*(np^a^;-Knipgpe z4blV#Xg2g_T}vbNOM_I6a*VgZ^~4m?QbDYDE2wLA3md%DE##2%rY}LSA-_YG1%|(c zY;{u)K&KJx0_Ck7TH`k4g#TGKlH(ea=Ue{K&ai06SQ#P5bq=H@uYP?zM#zDeU}lUE zLcVG)RB%ZLa1o^^e!1@1MHdKeaJz)q2#HvaY>_89Hw&ToB1d^hiTXbM)_zj=nN8I4`a zfz1h!YmUHg2E}Zq;Ps2i6k*Ur-QaN;in=;vNzQ5v`XYB-vt+NI8UfwytS{=ym?&W; z_GZCu=ry{rc_<5)T&_0FUOQ?s#9ThQB4np~UsOg|M%Zbuk=uXzuR1#0wFe5a^vBQ8 zc*X=-xGx?QMycPUWb_L{53>A)@lv)KnUxfZNv235niX7wXN5IGmZ>Wo zUDf4MjT<@{EL{Yn7tqQNmV(I2+gp8E+V|?-4l=-^q-)PoqWyx+DMyS3{{Y1RWBeW} z2ARi8Wt(7Cf~iG-0{? z5=IL5&pQP)_*HW!%#yIjhsi+N>7#jkyz|bVzGqsq%O-~6xN$^crotE3VVE$yF&C>6 zkiY{i%>6lWUPuG4Cv#BefxI6sfU#q4g3!YrJGRLZJ`ZwV-`bIB8f9}>kJW74tT$4h z7(fMKXhF#cYY92!oz0E)?&YeliQD9D%tOw;@T>lqVpivcog@iRr*g&rwimRV3}3Ou zBVkr1%{p-Lyr=v5achR(WIFl$UbfsYZ#qWtUnl>|l9X*K!*tUgHZ7eTnnErbS@07S zbBrQKsF?jxeIUKi4v!RcfWUHN`3LGYoby^41U5Wp0aig~icSol1o7`wIflkDn*>Ej zNYQg(+G>s(T>&jLtAKn=m^wBXJ5~w=LrzPbajO!#gbD+(`v`8>C0!-OSz=f%qFIt5 z$1{*UJ5PBYn6a&T%WW`dCSy3-NA|ii{&&Pz#ad3G_%vV~TbeMoGNjnxercwj9MO)e=<$J=Jvc4B zf)$7-09pQ&_B}41sd*%JmsT+G_OsfWxUg(d!n79OPu}g zmC)k9jE8Cfm^$jBmoll`l>o&no@+QunS8h|F{Gc=1~rD%(PeIl92;Sh72HOfDWBPA z8wg(c`6qp*L0dM9BnJ-mn`t?^6`T>Kh57^ad6E%!)F+XrrR$~Y?c6F4NFCA;`EP_Q zUM>_NAQy;Lu9FobV+A~{_!jn;eEBblKbvltdXx#Ps>oJ$L0bp5FHn~|VxrnfF@P5> z6|>&`zU-Xz4pjSh$#tp{PE&Xm6pSVX71CN4jbdQdp}=~6D<3pRxOvKp$~E4bz0OG+ zxZ6m!+Zp;`R3W`8Dg$~X8$8ZSHbVx%M%Lq*fwKZH#s=h>JKAQ)>wip`Iop7jvYCHf zK?+|OynJMYmqQc-3-Wzb%uXm!0lop9>cZRk-^>1K#y7ClwTxaT?*iZWg0Mr@6R?+l zK=%0`a=*&^%EH-faxcIEdP%UbBo@Oj!zHvp2BQ z0rL=!ep{MC878L3ZjVoqK_`ce4H4k$7~z^EQp_5Pti*oC54jqJKI>z7O}EV-_45bc z=3i;i4b2b!XyM%7PW{auVgN*mQ`Skc;tQ5N1QCf5AlXDQDHKVjVp7AfSqurhhlTA? zb-ag^E7hPe0TtF2)3I)LqKFi5*q)D;k?B}1Vg4$)9qVz{jdnNPk%kaF+fC1Mo zAqsC|TU=LYZP1|8r-6eJwx4U*A!H<4jJ9opijg-lkC0L933WM3kzvYg$0m^CxDFLD z!-H2MDi2uekuIo*{}Px2AV5T2;Fa#zLy=pxlI{k6v>U1i@)}+e=R)9GUM`(3s^+%{ z8$$MxYW|H`+1vxNE@+f)i*$5Y> zDF&kbA3X1Dss~h)k=r0Y9@WHb0_DtVQVPsr7dSn1YFItLl#;2F9i?vUkidn0wK8xXol ztmS~9l!-qcMBk?uQCnvqXc*m<&nK%#7rZMBvVWxMb=O;9drWoUL_D5T$igKxp&bZIKtQXrcG2a_J~Jr;V6%|Vq!K~ zIWCi29%no>#l#mXUSb9(SXP5coIL09h*Iu$sJku@OrZASh@rL!CnL_Es7>=Y`CHY2 zH%z6`9Cit0M(nJWpGZ2x#NF3V|3n&96h56^5lP1p@o{BNicJ>?y%6U@?!3QI2h6_y7I-5Ler0pn{}S2tWtmnz4?6*V$DWYVP^JgZp(x@mp0+ovCwWe9LbqXL^h zF_VmfYM`Npg|~>f3_-?xKvu={Pn7$m!$**|xlP^&wJ$g4fMO#wELXT}=9C3Ksl+_$ zj4HDA&P9|Jc(VbR#)jJPQTBLa1*XY=xR83uw5s1sp?tZ2KRMz4fw+@o@y<))RU5r^ zs(j8Ga)bv&$E4e{o}cbf9&lf_nIz7?%D3W)vj1D*3f9A5)wi&}h}AGS?$W`%_#3-I z3@+@d)SlgBC%Z+F16TE*FfvUQ6myUw2dJ28iC&;{XE;n?I;L1y;exB>VT}MreQsWH zI8uM9Tcg(jrA&bc`ErYez&o#$2_T-HJ#kz8f1$;u;eIxNa9mSK%p?aEWkJ0pb{mv(;-)k$ zs0vysh@a9WJ?-Dh-{D^lZV(nkED;>wo)OkdJ5)=_U7&VI;^Du^QTd#FWlw}gaYV9B z(#nsUvQxZLh5T3XQ+m03ZBZU!Jo1*H#uCe6eAn@R`ZvzvA`inf)I9sbpGeAR=|cy` zMv2j6lS463cAx<&4@eue^1DT66&l}qNiwh0rvOsr<@`)>te{c|<#51>*iY)Ebu`M? zKjIAp7rEk*YS~&3yL(vHf?@o_J@y^UyHAsl?`35O0Hs}`U%R_M)=&NYt z9G4+uXtSg`sM#O*@Nwl~_hBg9VZOk9sTF z^};x^Cyi`N6~zEGYZ*2+Bb*aO4M5O@#rMsUSb@5T+e4oqM?@X6Z4lni_iOgo_q|~^ z8)%I`x#sC9D5aFA#5%ub3HCYaFH)aWWI5|kuHh7WG%dg{sT5wfMQs$j7r~O*2+tY+ zYx3oikXq9oGB%!z122Wl2-qVJrbe-D-fm7eKfwc5FGqrkATaZ>qLn{MeEL*}S1; zL6<4>*RvUB$Mw{S8GbKpX2;+fv@b*Ra_UrMR>@J;a4NYSws1Sy;mA@9v%Bl~&o+-? z_pj$(GEFkGabF#HRxy*3Zl|juY`@c^R2ld3zQAm5w?wDPRc6S0{Hi6`ar%(kO@k~N zT{J_Z*f^s@SPd(qb-sfjUh$c9LqKiNRrwmuPO^)8)Q;tGmg49czg53u{D3iYhp+aX zF8Z=zcG>C2xl5`Y*nz!bH0{+>46KNcQ!zx z8Sd4omd#${yU}&8pH79JG0e7(`zh2WoS$7M((l_iBRO(qP~1z8LeF{CidP7#$i^8L zpd@Y`cavnZL!5sEc~}meqeyqulD?UfHXKTr0y6&s)b?zW6e}dm9SWdS)H?V zsp?qNrk9uUfM#D`@B&;y4ZsxFE=lCn_*A$Y_gFt%-hyo*um)Z|wauqRej{`@*#@j( z$&vL?bX(<_$T_iKrC>Pf(kxH#RpYTTx=4~3Qpn4Lz$7wjSIx`^N}*0^K1cthNx-6j z{MlVXb*vzfQyG%N)hLfC)N9?b&0fz3V)5r`z*#NGY_|cZ##bcwTzIzJ(8>YWat`Pw!L(CiK1Q z@)B6qmqy}AyWn=GZBd|QAHoNf-9TR%`OENUBj*%ZI<<1a zN6wYPoxo3qVtPB=kkv9({t?jYTQwzF_$-Z1V>{qR7u)dzD-$!BlBfnHhG|TASNNYK z>4k~T6dO%?SroICBI#62-YXkk>XZ&}PkNo9m-`RYxbi(&AZ7x5$gh zCy>Y1scJZVzz$U9IW#c~GPZ!i8OM$|`ux^aN4*Dp4G<|B`q6%p^};{|lU{T<0Z%9?uFFsJc|<&jwOP5X#c23I(S{yXZlu_=ptcpi`+=re6nt z(5Y!bwrH84Vg_z^%d^1Qx#ABR)h_=n=i_X{4SLD<$#|O5 z9C&qZrYWsNmItxibzVr&5(iF9dO*&waNA*6rQMi2VUv*B4Js_4UlGO7v z!jN=Er|O~Czj{@?gt{)@6@6cZbTszHjm0>b2%|9qK=H-@B9AoSbLN4?O_J6R!h()o><4Buc@7GBu&AzAIfF}oc` z_AZ@bHc=#nidiLJB|pKe0^bxn6mRjCf-VRsx$crB0h#yC@XIbJ=iLVk3KNGDOTaP0 zLfIJ-PjAl-9HNhES5Gy7qw_y5|DCL2=bJci3Q}tXr8J5GE4YD*Sr%ETERy8NKVxzQ zH(b^EkPc|4Q%DV`Rd|FG$LodaxOUlucD=Gg($kT)_ttkq=D!|ez{Z1+|B;bacGz&- zIg-Rq9=QbWp%};&d`iXK39p!eygfM!?n!oXTltxgp8hO!cUU!jJ+zL#p<2|#Tgllg zzQl=}xlgoDbOU61Z_e2(+Zve2JE^Fa^gyz*o{u-lRvre!)FN+MkT^eCR2!tl6c)9=%% z`Z(>9C--gT;}+1Yy5-SYx`k+GCW`Rt6Qq;elK9O`iG8t$lsm)q3efw# z8Wl;r?cYu#D35{VnCi@cHEu_gT9<2vQ*KL7kn9<1?BQ;cV_8rZ2<{cQtfW?2(;NHD z$QX9m7C7d#umbXkU|QyV|qGR1Q2_nQ2OkSKa3~&ei{`Q(&3IRGW%9TB&X+br; z&bM*mGINGM(BcPe=7HF2xA_M>N}Y-0PT~tU{WZ{(Yr$coD`rR_)^dx5NxYrmD5{CP zgRz2@M4iRa_c}uX45pQl)}xiyyCoB|rnVTkPaRNuVb$oq?11}g#fJY88Qj-w{>B!P z$j*IrU@xS|$oOd}CW9iIkzzR)NS50a*Q4v9%mgIZV&}KX2i&{)>QZ58xTX;jKO<+y&)MhN!9CTpMuCD1E@)cm zHEt8^n(E(Z%P?W|_D{%%+Uc&}a5Wf>vaf#g3|Yg@XgDw?@{KT&K{1;t7=@U9WUr4# zQ7DE&6mC&u2fa-NW}-lp2?3FM87M%?KLDDiy?!~MCp0m$4~(;A=$PQy2YxTkA3EDj zC0Ab%i(y8fwOO*ptw4mq{hi^97bCM6Ml(`3u|WzB&1d#1M_7c6t}R*dECvwqSV*n~__#w=N69Klu9pE&|*PUu9`!NK8b*dJ!%{>j8 zXwO?oIAI$^jYQWo+g8D?@txF;NruTux&6Rj$z!K4)QXLOn>u3Ao&TfR4mJaRIB@+e4g;1y1~ z5mMGt%qoh+Q!%&-E)?93&iek*H~&Au<_Z{z9tX?X|4PO7^ zwM9Xuoft2O_iCnZ<1|-G@AcCtE^(515KSxMrr*TU;hZM}Nwi{4OZNZvkKkzjtdB%m@1*RzmwIXL?u6w!Ceb+Y)7Vo-QT zrQBX9Vp>K9Tw~`$u?aAg0(Tj3YP^f0pj*nJ}!rH={9+%6xp2A9ic1dw+KtaJ7|o%wd1-- zlEf<)lup$swm5HbA96Yzb~tRz35_*sOikrDK92<;Rda79y=tzy zap9HdUTy-D&&lGo$lGCJf>COgZ0DBE8&2+Ib644m4e6OboAnqxa|iqEU^fO1s~mXa z0{3H63YHgSelo)(gPINi<^k8s!W3VP;tYsA;8vz+m3yvu@fMGSS!3D+Tce2$zHR;8 zW8Ke-bL`KiRRa#2=`&Lyfb+u68MoXr>5Azf8r2{z5Pb*~wK(|2GH6+U2m+~ET067Z z!%#s2Q&VKE-(}%-&qSOvpz|vkx=jy;12^155D88u_r;jg$yL_V2g7X(6E-+_&a^Vt zu(CeJ&dy~rp6`!`V9;?LdEiwa8Ar*d7+_`Hjue#-A(GMs6zd0lukst+FM^M&@ofw_ z0pxIv(gD{-X_lyoS>mz9V>yToAlgvMyPcmH@-XzW@MA~{B9FRlv1B~R<3g74t!(<~ z$TW|4O}mgC*20+SE8a%!3*0<&udtQhL$}W<@!mm~2JR0$z}2aaD^H1AA;Q-0x?x$$qOhEPge9s925@3NPmh6Zg<6Ezq z^|$!(#NHej9cE&0XjzJ*Hu@*Y;{z&qTRa|#mT|j;8n*$rd`?YNGB4Zpj-nct^G7)S z_UGPzG8io3F(LF8_`A{EDc`0InZ3U6`JNz4rx4(B9g!?sPciE#vX+Xu<=!Sw0SdM| z!i{130@cTrY0hip2c#1hzkKp6Eq>XaUNAvhtPHo~R_q+~+rGv$SjHwE;=t)WGcmu4 z=}5Yk;yd6n;Ic*08bX5!`(mJ`YZ0UW-jyR!3J zrc4$N>yDb?roifERPAcx)CPUXZHIvGMz?xe#|$~?^SrybS)5D}9tAf-wjp51sg-|V z{(wA}#%k&J?2H1obH&q0h<#DO%DPS3y6njRvVz8ZU~SyytIq`8b-5DN4z=?aKz;5Z zjr48WX~^oBN!JASPSpAa;J^|~_65+0yNfetwf)fGcbte0`3^b5uH4mur&0#(*leR1D5c**#q>xI26RB^9tD4P zNV>cZaPEQN4`grFx@3iIm|Dxr2I_)d$RU>p#fM?q84BzRL`Y`)++iE_nX%Yu?Xa`> z3ad_KH%}c)a{IPvj~^Rn*Kyr=Vy0EG!e!~yHUiQ<@@^=VuB4Iv7;-b6S6Eor#eVCGoL9> z3w5eH!rsts(x6CS(j^}#$^*8G+c-6xTcKFoTT2>3;@lsHR=A+f;3B-uW(gbu$&RAP zO73CzUTztD(ID?wltxh$Il#RwzQD=nq zfZ|Z8@}a=k`RbuSgF`0dFsvVn*G-G6eWeSflo$E4TG^Gi(~Su?(23P?b%o|+b< z#%{~Fmv;rLQB4-3R(pMV|UYjMC#X9bUFNrN*;*3Dq zL~I7sR&6o$Gt5lqaqg>l+uhpv>8{7<1{#I>h62yivB3>eJaSE(%+*e-nla#pM=kQ< zZSnW9dJ|zjx5al_au4%w;%MsthwY*GVdY0YZyHv>uYTD3V{)8bLdSvYj5>|XMia$c zphzPXvxb}|+m)64qt3Sj;$2($W%C}AMA5adL;iQf{h=+P)gV=PS@@81pL-0J$i`rAgf`9iF% z1h-B8+jgH;|KDm9_`6yBj`fBI zBv52EHd!F6&Hj0(#7CXmB~_$R@$STXCx|(V@M~*uT5vz}`29`K`FFl=fP~2F_mfEd zXpnGV&-JzuB(71+Wr{RWF&In3nk02&NOMS=&ru)f(bBk82f>f}kaqr%Q@u<_hEM2r zPK7QE6pO?zuC1K)&h7k7&x*>w)wh^T_qmE~GJ+4LKc`UE*N1lp1dU za-`fN=Hq6vlR zhn(6WumfGSI6Y<3d123qCRhY9tEQaiUo;ld%tP$zvhUCSqL;-DkI;2btbTja_Ln*% zpZlR#We8Lir1-YMq9`@^284pJU}O)cnJM@0mK=lm>4op~de``-N zkBL#HoyjS`N_fN1V2b*FDvc-G95_=1LQ^9u!AmG+7e$Jwm;pLxL5=dL&mPXFuP$Cn zK{oKZX9-gZbYv9ZtDl}sO8=(!qix>wA>q&H9q8KbFmN#lK`L0 zKEZ(ean(z${Nm^$>R|MpXxlno;{c7RrLYzsqh7}jG}Djg{>;Syn)lS_|4df0TU}Zq*cK|I zj??EmSpn(utL(p#`IY8)Q))IdY-h)TVCkRNWF5P?#DN#ipkX{>F3F@AkQPkS zuaZ{DGoao0jPN+DX|&v>D7dB`JQV|zVtRxhpNzH@_-6)}v#nz%mfYa*K*csaAr$UbN( z26D*RQ1}LZ5h~x5f%dRYg>%YIajVNw!BHQL@75Vd!#@oj@{bFu1AXg!&MomVeyrdF zaUFd`axJXhxnA1secBFMIU`{=4hWCDjm@`NL6|Dt@MijW<{ig%xrmv0r)j}y(K_+! zSxr$6Ocyj9Uv1^La!ZA&Fkz~X`Xuu9dmZPg8x>kQkFFvUScgsZW(+9Y?d5f1Dg5rT`%pGzdO#wkrno0qz5(D_PAieYW|tjTKAf_Y3GK+E{x6e zV${*tmwESVX=IlNm<44OL19n*e-krzj2JLWVKL&83E5-^i#J+(w59|o4hsu2%q*B; z%TZ!#4c1&xRsyjBp`_la|I(iM4I1FDE@5cSIkJ{ zlt!lcbtvr&2X-cHq{nGr5U?_7ld6AmX30VWOd=MaZ6y`#LhlY70)ZCN5ux{!6my&+ zbyUnLlEmBSy`EeV9g1q_V@;4wwb8q2L7VJ=|6yKdL>+xIFkPxsAz^H1M7~(3N)$cz z!WB>pw^VYJG|TJgP0SGp%k|L3LHA|3bUf*!2W7XVnf~b#{hf8HKB(fvBTI+HyxK`6 z7u&}04s}qOa%TErv5wQ@xywpXF18cuNc35_Z6!$4WbSW0HYAu9Rh)Z+3^;)r>&Ssv zO))DdvYd)3{O+>XZ@+bYp`G8;vHp;;-{G;YwC^!?@36xb;6LQ0-!m;NGE=`(;@R(d z-Y1Kb?{mX7Ro=r(3o3`8c2m$EI$L>|S52daVi{cwRqKUO51{(|+`J1=j8VjGR9z(f zu9@GK%b{Vmjyq2A93ve2tK61ykS8d=UcWf%o z3d2t~=OEKSte^|ro63+@J_g@%>5gG~HobR?dyp0@SE*w(3M8ekcLw1es6Xlqx9Nh~ zHn@I@e_NSIR=A!N-#hes(`g!ZlfkZJJ*b;PZnN`r9e4Fl)=nO2g$60+Aw~MBm;zC^b36TjEGqE5 z$mx>q4_=fBgre$px;>y*ULK)eRkea?X_gd+_d|+j$O#oOupUCA!1D*ViM;JmrBK#* z&2x#M70As~=3?nSmR3U+2%7Mn^`hxm3YEf5m1EiL?a)lf4naXhKRHHUg2perQzA$I zjpqs84NpszA-=^57RCh}AIZ)T_q6Ja9)l)O!byy*|F$*nf zXLut>bF|3&!73tk2euC27vv~)(7l#4uaEqi4P2jEEBkM~(Rux@X^_TYEt8qB%Pn_3 zV<@)$7#b%gm?6brF)SM7aakAZE@o$w{*k@z2c|@S4hso0)O~I8HlHM3eaPhybyH9W zbo1PB#b&U^2~?5%e5fo!1-$8>Uob%%>`?JX*Q)$khM8(%Oa59?;J_a6F(VK7AjLpg z-d^ks%=d@Rz+uwXBHq3Otkkx~yXjWdP4(x5uZ* z;Apz&4qTT41+pWSREZR`hJutb0BWlzA_FJ<6@pvx#WossmSbcANLb#{Hm92h$%FTD z7MT{gJrI74T<4LQ~LJoaj(nv%BV^vU&Tcm?a&uTO#7x;*zIzF$3nGLSqFQn*RO`|Y zrLm2(AJBVbP}m5^DuL?pSQH+aIJ?0T31g0)Ef1PU!YJLLRL}l^?|jka01E?)B*Q$h zIz>BdbUT*=e;kxF_C$3k_YvGIjeHVeTS@iGPJ0cwT^n12+BO5i5+Vk}VVmzB>n^My zGRf=GvQyIyK8G@4RTbIlzz*vnBmZJ2#QcJnywgkz@7Xgjwk5{<@jNl7qRlIx!B;b-V9 z9+I^sN9OylR%L**+D^BSRC$2_C3;ZZICUzJR98*ac&f3WA$I<7H{t(bB}Pm$PK2UH znSLgp-r+y@e;0{Y-Z#?U`BRHH-8X@`D=(x6BmOIH3hqNO$rkY{>BbrLd=!;Lny7rw zF6om~HH!3re-=shk$4_nrBm&a)kxx|?EcPQQJV5$=qh>qltHJsDJj&7Uw-nIep~Bv zhR*lYZ|A&q@tYOD-0|AtQo2L7+`m0MehRi19VK|T6z%Dbs*0%X^F9hL_UPqydv??4_Ucsa z;Y-5Og6jDw1634RAlgIk^K1;65TsiH^a)7>{NeR82Tj?Fzfb`y~ZOol7o%X`y zO^WYvNduHeC4{LHMJsrf!rGt&CM&Fl(=5LuTw|GAX^Y{r95;H$(-z-p$vs+J=KuV+ zhM3Cx>sHPtoi9weO1x3INnEuHqi1(JL9%Bw z%aP4ZUw{I&C7S~)1AFMlq4z^kELn|pA@$sg!V}?{^jdddannMN>OtjBFh|vrB?2tS zE{d+0{^(12G&Bzs!{&ymmtA_fcg5;FIQD+U^d#Pq$V18@fA!xgrVqFlhziAr0$ceF z(k^LlWEZW*gGdyYNq-FGCi-6s>S!d(K#utqZ$pSiQR|l~$2yjhfXXP~C+zWT^tv%K zHw5{@)cvky^PXPiu!Sn;zLo8yFhEar;^*ZgdkU#Cikj`Am|}_)QZWYpt_q;3SSRlS znhKo?XaDk<=gD^e4}y;=j?rh_V+9&gMkUQV>T?B#_n~WY6du{`Uri^9^0>#Ko>4EV zQWCJmPouzWjUD_2w(t4q%S=n&Tz2P}pC3kvhB&UX_+jLKzh_DZXog4cGuc(3xh{{$ z5?2b>`RY{tBu`o6wVt%mE5gzPhUt#1Zi`^|>#UNQcp1E!V75qYm9y^QrZ$O1D;i zHGM!7=YNrg%(H{e1CL`y69YGC0!!F7N%2G^GiHifi!!W zYn7))(I-qJX^K6v)5>~j9q2S%5$K{CRgZu8==B}1Wxjq6#StZ0oJxKYuQ5D@JmRmD zKazLRO9UHdtdpbowodirJCmIEyX=78+A4w#l_j!!TsYtBns~2YH-8^U7Nz@@MzNfu zCk{Rq;>K{Ac_XXw&7Wo(`FlauXJ!JJg&<~!EBahottTfe6r=}a(wH9~a>`MHgN<0_s!Z9KTYb zMuDm^2?+uinXu6s5u7z(VIj&<&hpmLfuy+Kh$Qsz~LWYQOfgR*2$ABC&M=jQE~ z?U5h1OM|l&5BpiOZimNh{xu!NO?KR3w!2Gxz7}DatbTcR?oP7Dfh#3W8%<0#6jMo& zLsZPUd8odeNoNMXi|cS52k_|CCReX@S?iL>`NTCx*~;x9>LWo#G#ZmuU@-;D(E&`S zT~H7_;Cf>&LODm=PEH36$aCqkd6+VNC5zI3{=iiy`_wfB`bbJ6%V*x0 z1x!>d=NwZU+63Is!X5UQeiOUTtiU~~x?sU9sR2XB-}u%QvfhC)wA%} zc<#tp4U*^Hd3-0Z0?FicKRY?gRG#V!m2@zprG<@0$UoQ%-DZiRYS83{j_{Cl0SfXS zpSLjpnv%3LQLMW(vJFH*^a4dFP*&@A8KmOD5MgtQonlP-0k#@I2Fo%&n3%uXJb-qk zMZf(!gGISHZ{5q}rUP4)6-LqH&nTvkBE1+rPUP&0zAi6~#!_z#D7F$cMvM=;canN( zHLsswU1yIpT@n`*3*p*E2p^;PYT`H4+L^sS&Hl>bqtU}$Wm;N06Hm*26;Iob?pHu* zbP}&5dVu@rOSn*q4L;1vrQ?EzoQj~ayPv2Vq&?CW;SIUQ7t$H|f+FCQ*3PW;y9kN@ z0eLz^ zEvlclmiOdmqiW;Zt6#hHy_?bJR5~b=%O;P+)pVR-edL+%B)qocogJ@LLwhrl8*g-N zoNI67ScsEnoyz~ii6u|qKTnm+_BPC1?F&M$k_>jSWe3io1Cz*zKDa`P0k+3nDh6kO z)f|{EkQ@!Gh+Fx6|X=y_`JYVbM|v^z{a;CduTOUJ#e#<)WqFWuv6 z$fKaWVsNWvUG^f0w;J@HzXq0!J^OQ(`6g{Kjo3Wv6g#d}G?#^K$iemaU_%$C1?nIshS>PRzS!m++xUO}_X%(;4 zaO-&$yu->~E?#91(>Q))jC-*?Ds0S8n>StF4l#`aJ8Z9lnYeEGwA-OIlH;ABw^j3U&C5-siHtYUGU_un-DwpTOmAX`&*rf(n*Z83 zS@)x8!wmH1kst0R9njby198uT{X=AjiRi@zD*0%KxH*E-QyHl z>2f)+2kN@A!t``9Ru4I=A+Z7Rh+^Sg`DSh_l$vCUy6BZI6GpSa>an7= zw&D?4a|x^0({Y!RGS{+uzcJX2px?gq61g#2m5k&5{$#mP*td^jdMVOP#WYC&u2)IJ z##=Q;buJ2%NN?mV-WnJDwEflLometcBRMy310)FK+mx;Re%Bh#?l6sFKwblR_j?fo zuIk&;6zRRl>Y$F<9m6FyaKCNJWea%9wJsNsRrp6Fy_1V`r$X`8Gz?{ z|5z$HFom2mvh;NnbC@DEAh1k#!rystT92o?Qra$gmwm4Rh7*Y(otu%V9P1y6DPSZ22v0#&45EIrU_=9k&0ZE@J|v_=q5qZo)8Y@lM+a*Kr>bR{3@f%E;E{gJX}iD0b34&(gAs?td)<2G^LT3NSpiycZ)(pUGpsT zA98wMy-IpM1ZTJ=w<6wfkf)!V1d6P`sgX)L#do9R#@yY3_%)avx7Lwnhayv}X2EBx zFa*udtO_iy3y%$7D-9&{9>noFIpkD+12YSfyGG z|L*4WhxS0PK`Q8_bkWGqw@X$LzQr%g8Lt6$gwTp`EQT0#svFK@A!ovdsg0_GZ=6%% z_OU?Zea6~|bdU4oNKk)hVi53zsxYQ>Mp!S!yXo(wbZDYsntE zmRBB7HKW3%Ubvl$Tkw1Fv@-#&;{lZgw#v#=SXwLU8{Y`8ax*~imzT*Rl01d%GJ4Ik zDF%3)GpHEU5p^Ok!SI~!kvb}I4$A!Q6#nwIa0F{Se}!vu0Q|@Yae9Kv*%kouE#&VCp_lWFUW-CJX28NZZJ6Cui1Km zY#7bpIB*<#kCDOIPB9>Wyp@WMzm<=9 zi4TG+rXx3KHju{ZRK46D`e^uyX;;KOa~h?!LHgCoJ}905My=`nBrzmidBFc7$sE28 zud$yAv>Gk?d1!o)7~^X5XgRUuyJ4>wY|2k>p2;8w*@bN!*xKlftj%$XsiVkY6ho+_ zdqF@61rfCLQRjcI19?&OKQcvkUE1W>9Qq_&Sk52do&h%|7LLJbtLJuv?-eIQWbtqM zYcvYH5q8VuD$Du$+iDbdgndAFgSTG_j?z&eXekrllH*!Yqez5$A2eRMbhX5KJ@;|K z?D-d~1NcX7%&i3`lHSNcr*2`v zLfnENU>Kf{>T);SuLw4Kd>mC6wcGt1gaNMr4b;8qr9O!vHG$dOc76j)dHS=nXAH_R zeDV9%GFt*KaW=h-)1q|~pWUsEmGvl9S$BUWHy~Fh?IN*9a;6v_dFPKSGDCq}v=HU8b6&IE@oK#dx zYUqBUbt7u=n{EBjTl^JPooL>l8Kuma`o!n`-+a*}7FzjjoL+EL9@5y4sZnf}f`D5S!$0PFo>Jyk3Z`(SOI48AA^7Q1Ji8MvrBFVjs%-YeLdmBAC}r^N7XBO!aHXi!!l$_qoD68j*s3yt0J~c;mg@W~R$@o3=BZF4K-oTRXj7({_5hpn{4D zf`SWb09gVkC?F^*n~J!zh>E)`5toZ1Aac11|K~}fl1MZM5`NMC+t1aLv%KJWzj@y0 zeU|U{qYn*6W&e*{rS!cwEO-}hlnDQnUJ{*EQiCaKTo*iIt?^`0+ul$bh~~=Z$T2q8%@3yNs4J zHNkEr=*<|oNr%8qDsyAr<-kv5-ApGI3*U1~@fifA>keq*On__?p2Vuw0>NaMA+#Xl zFhw>nJTDJv-V}}DBTxO_sg#wb0^3|dH+Hg?* zR8XbMIbq((#@p;MrqdEo*v9o8Q9e+N-ZQ8K|WrfMfG~h-^c4@c7+ciKd%Uv z)AfPF4w$(<9svbxvg&~25$S%pyaFB;q74S$k)5T7+-&{IEquX-ed(ybM!gJM9C9+- z_8Sz{ysMO%2Ish}C-(a<1BLd{#^ZS?Z!o;ry~4GLIu9in*kSpIyc!S}fQ3>Rb6Lwa zc*OE=DHq?SZinu7JuPfio)(^f(#$$)aE4ZLJ9Le!Z4p0iyxU}d{o3~rvP9WORO8$8NW z-9xa21YJNxfBIhatR&Cnf<)?x_=IB|X#YHk>h>Ipsv^@QrLq`76@}SowH(!$^>?5u z!d`KYJGKYx2!9mbAsuvq3I)kU@(@)l?SjaQO;9#Lfa$Pi3zYeO=gKV8{jS}b7c)Cp zjC1H{S+T97i%ekCA!ufkS$xe5STxD68N(KrIa!gZ|Fver7aa{&r2aECf1_jVI7SBi zjic;B3c)55bRrRr>qYf~dv3$B91*soV2Av}z%$O<+#gbtjtq-n0pD<2O#Q3I1K9~f4d9~|8FMf$QGuQHJ08)Z7~5^NVicS3OiY`(XHcv2#LJ2Z#X z^*)w{mwU9xFLIGw22*dUir2QJf_-kSY#nUTfP5NtLwIo4NSRd{-XB&^rb%0sbpge^6%awG@=s$+!=drtsXr`3^!PJHp-8k) zChHGd!c)&q z?g=x_1XjYjx`VtSu9GxE$tL~=83Ej_*F;b?2GGo%;*7gvU-e2f?VL8#jNB*NWLO6JZ{QDHS`KuW!^pNOhaY?<5vb2oneFb&qmDzv<2J`eVWTj4Myv83|Crk`??is1^eHqc#R>5E za=>B!PEggwMweU4)!$C|{_RB%zgqdV>hC`NQmgW@{}J9TW!;xz{u{Mta>>Eq3bIIc zg5K@AGvdzsL%c+OJ&s$0`Zi|~88=SjFo3pC~dx@Yg5YZUo+wi5BsO5p3ATX#Q8-rjM za0aTVG*mNON`=UDze^u46ZUH#zmK=w|03!TLiA9P12o4&E?N=^j`i`&>xv{kh7Tde ztt%i(g>Hbo-uNRhvD(KV&n!NbF9!7mP|Sp2C-(m9ySlaHDgv1VkcAJKXOW|C6@dnE zShn9$|H(oyS_59jV0+!CSApuqwO{?V@;jy}axTF`JBF8;gnFs)Duv|zBMt+?0pU{R zQsrcL+ArSt^CL?j@%-jVJUH?f$57Ku*INP!GkLM0s1%;6(g}&wpOdJklLfpkAfK)x zb>hp|XbIA}dhW*pL7Ei*OP#(1Wg(CYEtNru<|KY$sD%cBFFct3^nXn$tn4)mMVq0kisLtdn51eY#3|FsTvD|=s*<0r z06~R64?BD`?C_T!7Bwy~QyC-B@r=g=7KpElhh-@~`pbx$HU_jc5D%J}n0vRj{jeG` zUOU&^ZZMCGF*52?ql;Yr)wF`#ZqZ_a>vFz%&VRo2`OqO)69F(eK>s)*^3)6o|F$%!_H~fGp!RNB`-?Kvq7g z`QguF{SCg#-1NUE)49`tD{gf0OF6-Ui1r>L`VKIxlu&~%hxtwNy|TmnGSRICt;!qB zVLFez>tE&nNOmu9fUcDfJ3NwYa9m2YdJa2iLs!qkPe}gX4iW%m;tZ!@hwBjJeSw{_N~dl9>3&vz`4fEOqh`Unubk6&*?tT3cP_deX{sq`w~LP%mB2WDi@e<9 z7B42MKvX9y1Qi{qrotEqmSN=suk)l5*m3Kqw=O7OJsKy-PWf&Ao!OMZr2Oi{z;Svx zx9FT5Z;UdHEaL`(O(N(7A{uM$fC?sXmD5iDT@w7HFC87fG|sE%<5gEJG<&Za!Y0>- z|Nen|ngJC`?Qc}{QadjHOgEb6WP(j3=ru%iw=`a=qXg2Bn}NBoOQ!0QB{|yCZ7??# z_BzX)EXCBBPvU+))qsiEm4S6Ykq@-kqn(d+1iO}?VU-hRD`kmW}MClY$MmkLqe}~DH!Tm`5>U%eQ5=JxQsZ3WjPz$FLGOD*zcep=8)8TL(vl2OdS!10``RH;w}`i-BG z4`{6`#Msmd$Q?^=Ic<-~4Ha7XM-T?^A|tS*?&Hel@BcC zSZmxSyv1aMWHDLHe&!mlKm?KeR**kl@278k8*xAiL(~UArobk*K3)Qw1p?Rm;Xk!J z26zj1@lGsAWm28uJg|~kPwb!?m|qcjMN&I^#Nni@ibTg6AFrKVI~!@qnxu#QuL`dU zQ+%*uVHfByVR1o$2uV;hWTNOKU$ukPdSyVXe>Yh!dJ>=w-QwTu1A^HdGxmx~y>WDm z?Jb?T6B0}#4&@#fygNcvPklG?`#_@(7TAemazYmB6%U)orZqAV%o2w!t=O}|< zVP}{|L}PhUq9|Kj=%(kqOy+53b~E>w5r?Z#i7?`@Qg9C}4}Rs-r|U!hljKfY43Saa z{?|>fHLm~L9}51LZ-B_$E#Ej!uNZ6HvSTk98rep9$r}lF13@Q&3Ndw$pR!;@Si5r{ zGvb?1m6Dg}!QdP73f*=nOUaSRrY^L6`mHgkmXBe6K{=H4bjDl*F#h~e=OH?eo9(b; z$Q(0*%zlD}?#L1%dim#Xh*OD`@2~iv+$oP56y9=)rHaV{w;^7mv}jSc5EY7Bl}O*X z+^JMK?2t`<{9cy0fW8UJ0(f_pxEQuX`>00#QxTqzZ*u%Rx2D zsF*pl95S!JlmCr_($5*7Q#hM?NY~hLF!#C<7#a!o3_+hHqIW}O+90hObotl?g(AUs ztOC|DP$Geb{*_R{0(oK(G6HXS&DXZPpZ|qYVJkc0a7|e2okun~mwM~T%T((=*yE4E z{vlqvFH|(8GI`|gsJqg91uC{_p(S#}0n1%FnMyB1xl1b3MHdM*B+9lmD6q>B>U4Ey zpgVT~e&PrrYl6fKb_}Oq4~)@wn72G*6thhE@Bg~_ZxfNwvSVvxMnWrtzwiB<&~((Q z%wQUwV*+#h77xts5)Oh1$_`oUm@e;R%0mYw9MLd(2Dlz`LlW1~$A5p}x!?P}jcr6_ zbPBJ=xt=^Z;}eggd=(0NrjuLsRdI%W@)3{Ab1%;Y%M!1QSE7M&+AX^xNt4n{mMwYGyiL*m_$%j3cS3d>$}_XK$(+^hyd^*_j|DFOn6L~AG^D2wf_`T!i=El3 zcIP27dtt4PvI=iYA<7rEe9p#5(9O!Y;9$voPVgS{a2RWeId?z%Tc#A=T(W zcJodPYslmXywk?e8W*Z`wq}B`dClWp-@LQH)Rb_+i5({)%^02|JXm=|Jx*%5u3t4r zkqbfA7bmVjMCBTGoMByViD->$AMfGh)JBaDBCnZWD_p+*_m1D0CMC^e^R6>lzDbeC zAWyd^B8zDzGx+VGFs4@MYTq+J78eD8Ns>d6?NfAaD>B`%TLV>k)g1U!$2TdXodizD7Ec$q)xd7~J-#j6aM`qe&Tc&! zGZk?dg;ll4kxw(9st3lo7{NNH-q}|q=^`61DlenOQnTRk5dqR1VlFS8Jm-#NFWCyDcK-jNVnCpIsALw_HMkfeaTQWae58m_zNDIxQyLs1yNOqbW-U!iH=q^RVn*zbp;9>~UF2Sy!SU!^LrU#}6 zE}4`yXgo7+?b#h`JZ`2WXVv$AVOoIuM(t|M6y>gQty5f)T;>hSR9av@Izi7_zDgd1vMJ1GYZ}vl?xEMh$6y}dOIJg-b;YRJ`eBGuzwz^im>bnfM^qOcnzo$MCs?}riv z{-H3uFHX?o-XJ&(Gz%R<8=0mFTP7QIDVeuSV*rM@ek(5PO($pBykLW$ULhQ)*oj>P z?vVD*9&%Hmf!IY%j-&H@P2qrx=g&;W0ViBc{ouz7eNDIAW_Gq{lz|m1xSqru(55-d z_>0@gzTi|Q-(y>3rl^fMO2r8DRnVaK(L%M8os9ne)568qeeVOkexm4tutwJE-R$*5 zbjWS-P4O;r@m=>m**c$I_r=!y{&JWcBQIji4_2HtPLpH53I6cfbAi7#EI&*RBcLCQ zRm5$_!Q2ERe`thYpAhsX*y^a&#a)W2V~#4+g-vB1$)LpFaW$R6?}n^oqw_xAUfEHP zwNbq;ur~4X{M3qrzN%(n7B9u;w)8+qm2iN`a*YjXeOKeM7fLW+oYRt8za=;tetU4m zOQVq46pa}2?Z!%*NaWAm17gZCg7eZMAGPa6aH;U8bqm4Q!iJ>V?plvhxUGmx_h0XJ zUOFr*5Vb&$6MA0gAz{`QhlVoIPVcAi;}b=LE_-G0Tgor7gOW4!r!=-e;s2X}Iw0A# z$_Ia=H_$0^4nHDsHX!i*i<^$ntEYiaF*;(KMzEU*I)#XCl$3g79YGl#6IelBq>@F4 zW}cK)&pH;`DjNoevsHdgnBiDuW8H5fBf`x*89i##4{wQiijNOXwvfm{N7EG(c1XV0 zTP@EQYk-^Bn(L6ovtDj#w!TZl1Z)d59e=uvv9^0b%guz&5sZW1mpjj}cREdjc% zy|A&qAKh8g)+{hADCx`(3+OUC-cg@3n)5n>JxtKmL^Sf6R(?z0oqPY8Of)P(J7*e$PGbt;(>RyUYMWG0CL>+kkvHgtAGr3ZXM{Y<&cBn%Rx}$ zn$zP5sJ6qfGB%DJcvr{6u9=%Q<17uh%B0ux%OlT%?W|rf5%i238Dql5F1Ho&MHjqbiM1x0_&h5p+HgjV5Nu^=3#7WFRmx5fc^%VoAB+ z%chf`lBHy2kbYe~Ekr(#q8V`j8gON)Edf=+E?Gq+g#6$h>{?pM-|k-{i)AN*BJ-1K z1$>zwd;(V_%j>;O+eyp_z2qqR7%hpCtq1*!dCd@_1yBEqWMfDf>=dpEKc!UrBuBkA zOYQ~l5$qA{qJRqlqNSBQ+lNNR`*J7n-Yj~Q)0fwNSNR)>o<*i?3mHaV$Mpoej-b~P z(P*a89D`g&i(DI;!1g$zcdNB-E2w2JKb4^ApMUsk)$f1w%OC&l7vdEJyPN>&Uz7h? zG?q1QdP_i&wCJ5=Q^R4mmAPgt8j|>9`1TgB+Yf>qpk@sLn6BU^DVFSxIK=sNhju~= zXrFc{~K-t_2}u6b7RY`VTu4foRn!I-bepRnJm&f<{Ox(;eI!mO1Mt++-VX z;`yYF6DME;pATV^18z7`6kW-A*Wk8R%>C0!dM7ux)qZ2u^a-PN@*u&2#7r3xJ>-Is z(qV@_Sq(WT8Fn~94>Q*V!w!vfsp27j*kN5{9aK=^bu>Nce%0?JxTflYxBatgq09!I zS*)=-=6zAps$4vXTu{_;vWZ*_8dV@z^(l#0)bi?gj=5P2g5$@U$&&l*qm_`?K~)x- za(CFRl`+G{fg9nz|2gaoEMZWoc^o9~+MSz#mVYf@Enhva3)=-rQw%sr>)?qnVV6beMi;kS50QL93L2eFH6C)Whn ze5rbQmMz85%TL<&%(s92i?5q151A48Sw63eZt*~#KMX`*GotDqe?w3X_>st)vd$+K zzNFoZ3X@&fq}a`1Kl_LSDwRXTGoKoyHPDGz1yW+&q>VPpmcHg^W_gUiZSuZ3`3Uy= zO$(g!=|_ygdD)a4aEac?&3Uoo4r!1&8`Xu9ORykZltDzJNL8X}ShgW(Ft{x^URucO z4u!}qegjIV>r|%8f!VN|!eEg;>Z>}W=%v?1;(2IRl4A9YRo5aDbWRBGCSzIi<9|-= z_{L`qK#AM%_3zTvb{yotW;FE;1bdnQUpBf$p2XLZ37`%F6}<}!dB7CaF#~Gjq?ulz zmkCace%`TC4sRK9xd9#7Dm_()hFlKP;%nWGI=4G(opt2~a|=OD6|LvcbXAou3ahJ{ zgdk&@&oaSvad$9OPTq%N0j;D)p&tce{;PoBu0Yi22t-Y_*icI)|?F(#ERTjVM9KB`kT;ye0SUyOO2IwsRPrvMcgR+n^!jrg`I zFH*f^ic6)-MqZ~(6K+dO0&_Qhpm3Xbu6J_+3UU6^&NDL%rsvAL7k@*iyftf7iIJho zCRlKM)kO3slHE{U^E^>;FK}xJ^sa^Kf%rZ;7yWS!3n7bsW@y9%!OuJE%ndc%qp{k&4} zR30zdA&CjR8G8GJ3-9aCRXRO8xBqj`&Q*~4HlY3SKp>RoM=4 z{#53)JZ;7%(J}8!uyAGjCQ=EGcZK?K@ZL+pEZ2JS*^_z|NBz~9K&=<@GwgR;&8CK| zoYo?5qVRcq3cD%wKMQ|O37hYG(D#Wrhindt3GW70hm}qstOWNS2gNo0Xli*v#MW1K zs7!(tGPX&hKfMBn>t@XQp2C2&lhJ?MK$qBY;;GRHJ%CbW#Hqc`a@$!s3|>NP}82d0<`WScD3TZsG); zsbxu@EOIm0pKsM}zesPeWBXHVWPh><7CIfa1M{YDQ^Z5(Tl3nTcX|{@>=i%dmx3O# zTE3fXif9mNotr@>pjEk^K0>batOc@(UjDwwrGi?vVG=l=$B7b)W_ltwaY6~P|0}H4}C+>u++7;u;h3lPC!H0#~U2@z9T>O6-8 z2D|Z_(&>Mq*K+fb?RfW8V1%D-1iJ+t`e@ueVL)5QaMlHNHU|}lq5I$#)IX>)s4Eg| zR#9Pp71rKZb)|X599saU(T*q{zhX4&U7G)^mamz%)!VJXFw-divCBSQHZacZU(oBm z1P^)~TfNd|U?3(rJl}VP^1N(Gt6W3&I2OBKgX~bFW0sg}toMtkPMj%ws=Po&W=q=8EOU$n;!GCCq>{nOcE8 zZ>Q)&pzuAbcp|yUgTP|4RLzTg}j0eURK6Ad|#=n||>Fm~$mk-16fB9<;n~-q>9rhn1loWYP9FkIWm0^eY17zrrk~H&IHvshv$j0o zWJ{)eGJO5Z9tMDXzhTOGI*r?Q#*S?XG_{U`Qvtz(%x^XkU9_l!X;n508>MI?P*J>0 zTqbH&=DF?)9}Ir#+Y45uk0~WrQLW1DUZvu=fIV~Cf|pWjnDnSEilu~V`TOgDsbw9{ zb}$Ayg*KoQkQL|>1%GjBb;(K9>b zy68AL%se3P1;^2MJzJHvkoN(7acAUNz7|v}JIdQ0c-{B9dq&`9#VO%!I>{}QH$Z_h zpuA;fyK@8YCY1dTfR5ko(DihUq)ZeefUcuCyP(naruZ_i*QHLB29gw*VcO_<6e9ZR zWWM4s*-Li{UH@8GQ^^!C-=Kh(HV}nq45@^C{R1J-MNz) zP_(&W^L&?1RkTCa3^~Af>4hM?-Y3&vZFgFzsp8wFaO!qA*Z!uGCeApkde%`3isJiD?01okp17@Z> znYcL{f{?`(^Tq{&ZgLVC;W?wjsu(aoSfXcLnkl)$8>MGshC-p3y2AF-!1Z$wI84{L zo>i!}ITnR(a~uettcd)JcW&|Azj$O)j?Aw=ezAdDOl^kU#@ftS0i<{scEIUJdClqt zBMxbj8b1|!too1JW+coj4egr5mIKaJ*pmVS_#x(zFosgd)VVbg2ZIcrNPKGV9=ed5 zCt}B?{ilpPkxGI+K+yY$XcQkm>|Y%)pws=`>%Pt-Zq8o!mhk5Hk^}K9X8SPhhZG^h zvT~0?5dJI>AzcX8?bZPagIZqYzk|&c_eMORVOu5X2Kv+O?&(k%R~6I|e3@6~T?EBa z<1t%g=wA{xk@aL6l>t?OL%bs4K3+Bb zlt089aw+x4l0&r||Ge+FViGz0Ilfjay|mX4!^!B|?|1HBPwtQLGgy=Q57t-E+1#v& z9q)^Qvwc*=b}zv~-CGe_lOf*f1^8!SsA`ZtDBr}Y;`q4o?00M@x6JMWmDfW7TGokWl50Py>VRzbJgGP=)s^#a^3L4%2W(8>~y9MS*krRepb zw)l?fjGwGDEjuuiug`O>h}`R2vFMNOi>`~iWVyV)a14(pQWxm-f4;FO5#pD4JRp1` z-tyT${wq}iVDpdL?235+A5ac@m~yS#?e^FV34EhV zEx#N31&3U)tUg}a=xPX-+0s^xJy#Y1z-qgXv44lL&YB0nSi9mYzo=~sF!&)Q-Nnmj zjUDG~>Wn;#eFO`{U&X*2rwg=efu!M@(<%DSeBAf8D*NaRP>INsBt`a1`ocHzDu`|V z`vbPTQ{xOoaKaIBjdQ`AVTYB1YC2JLKJp?NOJPe!v)5YSUb!vJr)m^+ZU+>7yo=;v zE0;G}h@4kD75b}V0btzcI3b6Kz8(4n)BUjBniDfS+#E%|?{=>>-Zw$yC!W~~6$Mx6 zPDtA1GszM6rL|Ol*cv7=Xhe7+2w9PP$ZNtJXj6U!deQB$sRe;UC6+q&hxJC_J+`tz z6UNK95%KB=x#8tbljl!OUDuag5<8Bvnenr8fD;2nW3eTpNs7_4kAU_K3$ByHtH>m( zk-X)2N#0D}bvZ*_kGKu}T&l}#AA`Mv?as9eS1(us|FaczILB^Hjn1nPF#fCNvGX`s z`uXXG73ucRuKkscA4`MPj$KzMP#U%W-9oTn6*myk?arI1H9j8)G|3YKuJJO&sr*i0 zg=Yi6tz zC=K?_TumLNw>jSRoJ=4*pGZp(@%-jVJTQ+4xhV4F*A3QU)vpdcqR-f||9a2JGF&6r z%LILqi0=053tukSLjNY;quV1%&?(!Xqv$LtpIdUT}A4<>S*jw}T9JQ$-jn0)p zyTY%Be!{Nh7mJ1hGhH%ZW4zU8^bonB`JspWmMU76Cj^@%y=3e^--H_LRwc%T{`|-C z_n*akT9q3;%iJz6d?G<!_WCND%$O&`|TYUWABlx<9)+`QQKektK|SmSD+rwXwqI z*P_xqmd4m05zCxD`F@lEU%t!N?4P zkK7ogKQ5Z_n8uRI8X)SxCysvu@2?X9D-f?FaA1DstV^L=%w zPN|CJUxn_l)oe3)dEwTG5eMWOD-d9L+^1xhtic0WqIOYOad}O6AOZ;&!B0lkcZAO+ zNhJ&Aba}4T5}D=Wie;+=Pux&}vw?od z2Z|^lfS-HV|B>vP=UJaVpq4rWoV>OCO0S)s>X|AWqI7!cycoenGZYK|k4%{{LdQUs zc~ka($_JlyFqo}fpZxqII){CO1Q2$No*Vk)bFi*dl`72^`A*bVIy6 z@cb8Vh})f?&btW8n?3G_BFmlB(997hXb@j5H%*B z-Cw|~bZT_2;}yeQyZO7qZ%s%b+#D7yMVRqAVe-k`;1f6>Cdtb%t!%MdOJb(v1{52Z z=X4lJ{K8P=k?e3^8`TzEpy-Ackfj1up6gW#i!LW?E6#boCoqRx9_40HVov+4Gfh3) zEjrAknQH>_XCuS6%D{_q%s)}o3B9S4g@hNgV<9G9JTOT|+%Vy^O?H}ZSY_sJ`Nna2 zg&jv}vW?b_jRd=app%H`D&cvyDrha#x_%AYG9 z4E<_g`mhZ@a|6cSEsKBlraPsF0&u;6?yVy{twi6tql%Xf(1*g#(zZ+Nj6H>xqhjpm zZFQC#Di*)<{$o>N54&AL%t&=~IiVO~m2(1{PPRaeVgsEjxWs_wTpyFbrZPjWU1X}@ z%zN11w*-^Od%cm{tw5l&ChZ__hhY^}u~V^b92^TPsWGjfrph@<*%U zhZ|^q-1qdLw_%NW%ASg+)o;yC=8(~fR79{l2|5pmBVE=5bAP4N&WLvB!w|E_D5{2h z2!)+k)uN_rc~=8kLXYu|@`jjNIEyf=oP~D}k^LSmp*>^{nGN+0&myok@@^&$FRP;J z`C!3+$ea46le~3HmIUUwZx!T_gTRu0A*jc3wX!>K z2Wz8we*R+NV0?2vdIZ}Xb2BNtZZ_^scUv>4jF4N(mD8G}H(=4J3M%qZE4q}Yg-_`u zev{M+9Ms$<%5q?E8+W32a{>mj;AZYp(^A^Eq&wYA;q9h?e0eS~{N+!RMYMp3z6+_@A~M$} zQ*o4EOZC!~vpU>cLXFRY@QUGlFNth`Rf~|o4UcNzs237^`^7PWt5n?FzNk~=o%uFd zZY%@A%MR-3Gu)^6^-plKBaedP7D^0|5k3@tNGI7bWFUM#s;(`AV4;{P4a7IzxxyZl zed5tg7K)E~X95S;O>w(VFDxx9Lzxo|mayAs2Vm zj+4A*#0@*iTMK#@?hfoybOhG}oyCB7H}9G-m8lMBV)NN*_d)u;bhAV)$Hgimq?2iO zYxd70D?o~+%;O{!OeXj&56qsC;fEz|H=MQ3vAzRztTaurSz>+2S%N2HGc)EVt3zto zr)G;Osp!jw#E!9HhB|Z~wajOWXQynF=qPVr;Ffn{1W);+&wUR`hj^9ll%35Iqq!mWZ%EI=?J}g8iXiE zJB(=ryNRGvi0BH?XKQzUR#MsydxrMlA;C>?udbAI7ZAIC^hj3aKQxKG5FFpP)!(bR z0nNYfF=~12_nVmc-~Q=u`Zgy~3jFIeTI>Tn-&y%|yi z?ZoM#N~kH`>0cB&nT169S>yl=D^C0ChvEi|2eWSeGTgARd~4s%47zk0-DI?|93fc9 z!B!H{$Y+LGQ@xtCs*^e#pd;0!(w#DAVUE5E%15sOVcBJ}Ox!6OacGe@3kT_E!AvOj z8FI-75~egNK_>`V6VUapo|hig&o%y;yd)qWXaeQomY_%B%|J7)IvR2`L_gXDbFv&J zYfSI&6(?UmBd@rF+hl>z6z%l8=%4;F`orJ<=;w=nCtg9Y%LzKhc4X$S_|->Wni2Jb zv!6GhXi(xT1f_fkWQW5C|4DTD%tH3C=8>kkg9aTZ{2ozJ$TN@=cFYQ zMg3mfJZBqzHB7%XqapL!OWKSH#*jigSR`3P?AEh$poE9M5CT|hkJim z7Ox^wmCCH-ZwGSfDq#m1%eKn6P19@N*K2REc%olBl67z3W&_KAD@IeCQKPK~EGHX5gYf{bHIqA8|4{Q=A`u z`RP{-E76^>+6VNBH;gi)%V>4FM6eeK`Wz9Bd82kZW@fD$Q0@3{l`o7u{grYX(wp?C?{zBI45Wv zY1jeI`}XrzO;gKDsKcD%fHuXODHu7R&OF|}Hv8<*y9R7k%>C0!dZ!(GIVX&8bdX?S z#VSMXu{N)+P^fzfZSxulMY5+_YJk!xN|>{r$SBtWdB}S5v=BRi6N84JEgd&{eX^^* z8pU}f<{@(wt!y@3OFj1AHhYKexmIPaqAp;7Y8IXZ@pc^Pv46T113p$p3I{B_`d0Hu z87n0Dn~vdPQ=MhIUEjKWI;feN;Ai8!s(EX>QHj z>RH!0jH8w=lp?xKL%8?^l+{l# z?rS{pmg2_L(;M@%@z4|c#0fX{8^BoNqu$9FmYYq<0hj2FcI?Xnx9zBM%UpuZBIpbv zTHhG9Jpx-$Q>O#rO$yb<>j?v{FSojYU1S?CLo^Khq!7$&k~RjZA|W8EUUd2A4(15dRs zjwN!?I*qJ!+8uS0*(Z4dHQT$RfLq-k_twWf_V_m_8X@R^NqErri5Pdxw!zIr0AnzF zubKMsUFHNBV*2+y!YAU$+ON$t=g6i+Rt3e;+x>BOl>GjKMXAgg_)}FC)b1P?Fyi~{ zQX8X{#R4z4ZNT%=)N)7DOGDV?M0D54{_h!99)V!P*XfIPyz)FUT6ww%7DSlai0CFM zG8T7*?iIJllYs6R=t^7U-B1eG4UW^(;77q2oS|1vqjQXwhfM^VLeR-X zG!~EO+4&Ab$=iBC6(3iI<@63n%TxtTs$i+^jac?db=KOJZ}6qDIejVZ*DovzIylR; z1=?;S`(~P&u{H6Gd`L$GXCUsm=QjC;ng|gK(fSgaY<+?gTBnMMi{wm$x%leN|Nc3> z&5qq%5Xl`CVA(~mKoyt^YC_=LqMO<7+#^W=)ms!?hO#zvC4uq_|HAsxgNj6I#NmqM zuA*EDerqSQOfclq$rMw^9HD;%s;WjD07#HCvaYA`Y`QIk%d33RXG2-_Z%%NTazPyV z&;Mh9%ZFF;Jm`+G>K5$R2GUEXj)qVl!S)bzHxUhyIi^vP2|?sT^1a?mmL$@*Lr<~C z-%-mK7rS3`+UBT@Se)fwFN$%i^Mt%17F}ZPaZD5fDLE)jQXs-zifR|{#DY`?Q;v8! zlkSJg2d#6LOjQBJFnX9}0*wcj#^jJX#_d!n>WqoP?zNLLyyd#@R=wPeerwetU$Ug# zZS{h7x6?k|@{Hm+wO*~ zxN7<81*%qHwCaecnopjU%HK%N+E9Y|m^ zLY{XTR4;(#>^N-rHd7T)N)EkoKnar1Qx&`K6lkGi=C(8=#Aaq_d{}w0#JmJPE1kII z<@i&c0V@uE4$J5`ZdkG7G8afnj&hUUy2Q*~Q&tPWptKv5W4 z6_oCWE7MZJ5{xS)f_nVbfFV~LYDtq6WnT+KEpi-v2hgWtazfQBJ>Yi$2;IBLY;qkl zkzx&tLG;3&a5x%2#J;fqz&oaan771NF%z~aao0jjv3h~(n4?yL+w&0z{r@N1uGnwb z=HI7C0;SI^%a^}$onAkU-edHY+(EEFe!q=~zQ)`O{LNB9Y*en|mdkJP6VhI^Docfp zvelq{aa!ITae3j!@C?v|YgNWGA2WMpM?Dg$T5p@b?(u%*#{Xrq>`88$3A=5IUGM(w z$Up-~ZvX4O5;~vT@?*ztNxc!)D+sonpi7D9GJ2g45F2_w`#Y7vPq{udWE;~;ZTBB` zXk*Ig69W9Z#4iZ~IO*hlKfDLF*5Ll)bL)I+y=x>jWPe!Nj2%dm=&ir2RaxR!=37QT zeAk-ai}8L1#$0+VL$^hD_5I)Vthom4{Q0BKLv-FW`k2v6-%qe*1YH8kTr=za)batp zgb)?VHKj6`s9(viA=mjpOo`X*l^UU1>xM;_*i($L(|)frGgMlyK^isM$`)efW)P+K_$UC!5H_0gLatBnPYIn*>83H(wxXcg|6qbH@_TJ`j)Aj|9@v zf&$+{@iL!Er;GkMAy?Q!w;`A1g5%;kelB^L*9&5aF@XhQeF1c`kWg#x^>Uo4v7LOG@uD zvNvji-Ad4#iRb~)zN~;nwu4dqS+xv^t3Fa8gU#)*yJ6U&mIAq-zo@DMK=4ea_JNYk z{_s*b*y3TmC3MW0w-ipJzWav5F^99-$ph}EKlV2OWN!N3lj&SL&RQQen(}gjEhS)Q z7k$H5t*C$$-dUGHT9x5=o4!gHLSpa=tHQ1Tt)wEfQ-0RvzO<9Ui>jpp=p9IYKSprJ z;~p=GYQ$E8n6UaqYI(WGQ+^^}zn@0XG)XmcH@Y^G)%;b2_3$x%D7*{~IAb*^aI!2@ z8k$qfK4$<<;cV(5U1P_&W->V<$8V!c1^E=DUDT}w# zZ&g%-urg?)ON$bO>Up|=LU@{}z^BNk-?>3pO5&ALQhz0%+7r>c@F;IxP(N?8M=3cj zoLL!CR>K)5ND)uZ9qf!SKx&glmO+B#G1KPyej7;ep~QT{z3mZ&taJcO16-Z zsDE~WZy#^3cRx^{X-Om=GEmTze>Em*T*k+;Ntzf|Ec+Sbx$8?`(>8h24iSA3vW?F)NKmjjhdea1-MNPB3Plk#wR|Y5ifj}Y2^*vh9<55; z8OI3n73C3`qFpop*5Uq9KW}Gfg9ui<_cU{D1%Zjpzcu|FW0Y`PqP~4%&#Ye<(DKo` zRqxWb?HDa9j4Jd72^N~0dWq=Dpe|XX>>&gpRf*IUR_!|EdO>nQaz(23%I4jYsg}bs zahdI7Vgwg`MjYye<&k;b1!7gZB+vWEjK}m^elxj;*AOt`&?qYv?tAB{$KZ@2^1%Yt z7UkB!n_gRH*8@|PO6_`#E|1u*)H=t^Y9@g z2(EbN2p=e~QUeO+3_G;UIp%$t?UX%nen>9&t0DITq1HL~Tk}-QfEp{G!u#vP_E5F6 zvs|}}E(F&F0G$p0q-^ow0M&)y!vWXbHwEVaIhJN_qqLP)odueb)2?q~ncmBFSyNKJ z^qV46$v!T*PCH)IK!0ge=rEmNw-NLfD7t~tQ6#a&QkuQ9p+c)PJjb<3T1{^ZFN4JS z>USzxRaOXgTC{q^N>g~13y)2TD#o}>nX%Qe$)~U1C%-<`V4PxC2G-Fd4xFr;9sAE9 zw=@bN>j-u&LB~USgV&v}ZoV;FJllij%5hhPe&PQqc+Lj*}5b=Y9+rCI& zo<x@lvj{A2q-DI&W z7c@fhXD9R3ii5s2|1K=WC^tW~ZFfm(+d} z&;Qx{_!nUYu>IuRfsqom0lXC#s&Q>pi0%(PW3UW)cMHh921C?H7ew;Q)!_m1tec%h`a9Hex)>!(BE@Q z2ay8oRl!yiwF0*}$WT_jVAli#c{aq5S$tS>ffI)8*U*1uW4#+|z>v$Q1DEMd(}2Wq zwC&drY&Jn>649x=M*oN8GEjj=MzMl_Lq4&GM&Aa2cXt3X z0zFb=-fi z`*O#0Q3j}H7;4F}+9M+*l|ja>X0nrlP+ziZf~%oX;fe&y(Dfg%s(imo1JFZYRhlY~ z-0bn#GfA-Av5vx3Ty=%12;HIRA+JGGo>kyJ5m=jBNXHB1(a&*$_0*=%3V!%igPr-{ zNZ)_ZC%J{1?e}{@-!?LrmkIVFLAMamX?_pML&9FN;eB1@LnTx`3`GsjKoPiNC>bty zdQ8`oanml6-B5AVx3IyZhIcw*#kAj?`pK&A+*=g)FTIO?ok?6+-21P$7ghYz)qmUn zpTBK)R{Jzc_Ik%n>sxq=P7-wcuK-ryGFcBWQmpg&JEmXV48RCal2wx;xSnjXwOET2LN!7QQ@=p-Raff3{*yiDF%#S`(3 zFE;@j`8i3SI8MBLCe)j2$R{o!s}2?EYh0nVv&jt~%T{zUm}E_XSkklqTg4DPfBja3 z!SgY#alpwUO*#6rn}_@ipsD|0eFdFu$DlcE1e(1BTTIYJMD&xNbS>&4!5!ye+d@0! zA+X?ayL$!v*G(qO02ccNF;NHo&x%oLr3a*uu{vz4U_($1lr+{rD=@n3@yu4H!yS`e zx1iv0+iYw)zaq6RW?+7;1siKXW%L-_VB=K3Ibga0Hn)Fv?XPq^w}s4pw*s`r2xMCb z7V6nHKzkK0iAI7f9rHkoT&>X6_aJ>arZl@0`^iU6ec@(!u13FTs~vMhlGT3h>jsMb z#upBn*7k6T7~3&6%v1v5-fz2mk*wLR(YY{G)y-UzbVXyERPMi>_@I;79atb3n4dXo z#39z@j`KZ8Bp`Fe5+ni0;Gdt-9jf`m6zeQWH;}$aKyjTtAcsRoT4*lsXCzT zv&OmzkjTz-du92Vvs#0)P!qg#Su01 zu*0A*&O;N9f7Z~G1fJ)UZZV3UKQTE6wKJ>ko8rQ54Trh*?H^%ULPK`Rk{q|sf=x|< z08pJpZ5~SAn0Kg0{Kcwtgh! z#SC)<%P7NvhcCvl_7Nwr5dQxg|Janc_2t#Zj(4|aSZ1-cNiFY|#!H{aVll$jF02DV zF15T#uAW4@0#nQdE!IZFtHv{rmXFSUEH?E+?ADB!@i`#m1! zHI(DX06cevP10n~%1~^tj+;}-N!rsIlZ0bfWKpd3!#Iqy=3zqCPYjTV+rDfTeS;e$ z?09GM)Cdv}2=-%wz5`NAp(VhDv&Fw8@Vs*I;=#bx^M)O^_@4=^bh;ea!z90pS9eG1 zuO`2{dR~iiZ)Ebj=anmzC4q|<*Rw^!VTWeH3MDkadK{}6-?7ROCn6><==%bm>P#r^Z+ z<}Gejwg$9;D8d85;>D$8d_*(37|ymk7e>In>pU(BcTy{+bqB-!W$bb1K3+OK2%|1_ zEMxCUZZV4!sr!Cy%qOw|%5H1Jzddqdvzq~N->Thyk=|g(=|IT-jmjZs5iF2bZ6~7B z$ouTs?;gcm2v}0l&5}lh|aWQlmH9fD3lZvEYPv;B%)W-@s9e2v{a@+ zbk#S`qeKMg=!3drEs2ylCnzX@!ZL#7a7+l}7JBi)*)1W-Aa1x=b((4mPGzz}5-AN3 zu!!4P=7zbWF5!O4?*_!}n#z2zk=;Qy)2{ThZs(CccG|cjrWChIxWm0?w zL9FsZkU{1kM}f)WR0b=eRsbEuE<*o*+Zr6B>E(tbqepG}A?D_PWVjgYNyQhw)J(6j zV|xOU1*7cAR)XD3zUo zad_ptaGP&17Ox!5dKWoaj;R$7LudbEv2EVtZwNxs(Iv?<_Ih9N)MCS<;(}-Cq_g;1 zxHXzI;JkpKziREQKNw8LH#-_=x^EhtY}9zVd@36=m3~S@w=w-m32ZXI*X5S-3Uhhk zcK<|ZSGq78??cfC+z*sTe&Uhks+J$|y6=ZYsCexmAD?VhA{R`yqK(G$F@cA|UR=AS zyaZhYRpf5novF+@;tyLrPe$Nl%ZVJteLwxLx`&5VYZqtdhNk-#LGwwLYflNHPt)tMw`kg;F?OA6_3Xa~3M4q>FaXxWquA z?%~-T!KdiLP~b!=3~eW`NY=O>R3rgih_(D2&WqX@PO#Cb7e-}5N1Px%bxz$e-=7=o zSHLgcd56Ab$1YE-kzpMm*gk^pA)=8r=DcI3=mD6qQtt}*4{&lJNGm~mmrp@lIF;!l ztDF4A`A-tnFl-Wq|9YvxH*+#IzW?}Yx91|I&W zmOqKw;{`NO;~vohM7fHRLjZp7H#VtdHrWP?H_AYA2NCc;+I(s_^R-MoDQ;Lmk!CV_qe#+4#Z{w-ci6msVy z=BptvWx0K{x&<^chcA7w*eZVXrOYW#259|Rb>VmPDm%7mAh0rO8@HKYA&In}h(0{~ znozfm!(KNu3|DDf>`K7J$_68be;=}uEc@8+nbTyCJ%mlpa$D~H`d7(3Q%C77iT>0X z0cjt>?j`7ANN$0ITrCByC8D&bejwv6B`bqqwbRUP;DN5Tt^%W$%2j0YlK7Y=?F$D| z&~(wDq!N@|?|C!}2Wh;ymcj^9zI-^~vVY+mRU1SPHhV&Kc;t@ARiLA5xy^Vj1~}pD zB^zOS;fKVyct<<|eiy`S^J@Yk1rcIC7e$E&8HxB~R@|NGhx)bcdH zWyCJYBVZy*qAyL-etI0Q*5cPQksCREh3$6}mbPSPttkP-%XK$)?6;U;wgO&?%I~vd!?gss*-y$X`{a$Z(vvSs4f5O%2G{pC^KLc-B1N zfArCHzG(#djlz(bsPr1A`tZ0pPz74xc2B3&l*C7$SLgS7RCS($sv~up<4p(OB;EY zgxGzIIy_bWJ6P4BfE6KXh#g?3`$O2Ks^y2hRGBV)3~DBwmahmIalrjU8NC;Raz}Vt z$R>?AY?j;$)__oG7m1JK6|1Bo$2U#S=WFbP2~H6{&S{$Lw=O$nsPe!U4fy%+eA+kZ zBX%5CX)^-$If4aM$}=d(gBns;D04*uKrQ#ca++f57&x1!g?-^EirZ4Wm`;95<^rzE z{PzczhC>aGBSe(TJupjl87_}F=*P)dbk9#w9Prx)tb7=t+bK{z`_SQ@L4kD=wgoi- z!P|%fs-|z1kHFUnV|%xF%jiDm)eBOm15wD5CD{smwne@rU>AqFQ=7nj>@NYHwfKv` z3GP#Vx-@e3n@(hf;ROw_zB=QUE4tCFIa#m#k95Hr-!QL3ptrOBS)fuI67OR^F z$g?XJ17*w`xuNH94)v=!2Akv{QtYF1?ARuO=)tH^ZyCXs5cF;$`jC9DH?aFc{5>>3 zv;^p<@Pul_|C%gAv;z!}?j!ECF9#HZ%uBz-aO%$dPUg;h;CVO%B?X7%cjl{*F}Rai z`Tw!^C2&op=iVOS9FiA9HUh~RP%#l?abz)6)QK~<%e`H0dvAM}-hTbs+jh8Z=eC{p z&P~_bDefB}D4>F}$R>!$s-m)~h|8#;s8L*zh>nAUpva)Y_dH2(NFd@z=d|mffuKbm(G+DTbC5KI=xK*|U^WjN7ZlliQwN;5X07bdJ0^Sb zgqQPuQs+)Cn?1ItC>EUNqoA|}tePc@kX19A*$#DeIpQ3zPI(QL<2C5j$2%uGLO*`1 zDCD+Z?ToIdt-)GgX^sd^UR!)B|g%l(+UUzu!zc<(ETisO!F$*v)U~G#8`v(n0$0 zIJ)z>jKqkAAIB@_vElSQyKZ2k;E_D@)nAqywOO^>lr{6U$Z&XT-jM*9c`&{;IeT(E4l@y z;wD^9CxU8Ru^9Rzuudn*sUz_$j>+CIgbA%u6Gs?&Kr0C$ElT z_fupqNJRqgS2N_<9+AF#`mo+KSTRCb@Sx>R+C=!pwb$5NRUJBxL zSg)y4-hRfhU75EE8!WYd*h z>W49zj!rw;h?|!iH`_zCHUWzpZf^eh=yO#tE^ADzRJ{E8rs97d%J0b#6cAs_qSK|v zga(R{Hl{L)%sqn?0yf)E`>|2Isz;@Kb_lCK7?=Tya}pk5NsmO#tF^AZ>JO5|s z{VA4(QCtGHE?jnJrDQ5QdPB^6wVpeDD=2akLz>ZGiSpE_4R`_y_DdI) z%HOlcXxKh-<1sI`-+laNZv67s{5(ucpyn?$d`OnLa2jZvnFpIfvB?xkq+*+C%tg1# zu)G4hjj(t7G6B}^MV~I3a7$q8K#zaO#&;#Rd8F|lR}{IMV4>0fMoSjIHlejFGY}jWeOM=gNMC>Dz8imBKE_A~apXghtBYzkc;+T+>diF-D(AZEPqVN8|e zX<$E{7T6-_@oS(_JMc2IM&6+AmZ1Vt7IYv2FHvNR;Avn7iJNwrSssc7+D<`@t)^ux z2XtiL^TTTY9_yy1@SmH0Vd;)q@#BzWW4Aw=a1?T^@Ov6j&N$^)eA&O(mS6d0_i)B> z>%a0CPoe8NZRhsx+Z|xCB2U>9t4SuenZh(3ULDE(;hdjLiwYplu8kW0>f>sc4n;$r5~#RssJk5%$b=Sc1sL z@#26C8}I0t)3{BF)tkTe!7oj7&pSJtrjn1{$tp8uj6sThM3DznY_Sju3wKIV$WD^L zb_fQd56>x@*&c|51D%TH{#PX?!Bwnz?tmxsq^bc5^tI_smrO5M#_aU$2Fjl!0lD6N zyc0y5D8A{3+ylmggI?DpX9HS!+5~oxmkDI1Edspe&YW^udsM#Iw>;#ENE^@A&wv}x zLDDE=Abp2$8+#eZHU_=hz(>lB)_Go)47s(6bZ?b{*6UOMx8=BZN^ ziTpOhiWuUmxG=`AnJ7LOjID#Uo_FT#;UzM+ytGThvc+}buLnDZ>jp5F^~W5Oo4}Vf z?HBKp)Yr!GsWfwZ@+lSqHCZ6I2Z4P3^a@35c$d6pKG5h2KlQ&r(x!Hax6_NK=>;u< zZqe3YJO`=9pYrbuZq4qcOQxj(U)oa+<^1C_B^*qexw|vIkF)uzwx9msU)3g@9FP5Q z1F3K)(1I~)<$91}ffDrq_GIj!hd_?3$j=bAK#kP~X{&6P9F-!BhfpV?Rfb}$7$57V zu_XGv_fnKNvF!EI$uoo*!cGb#Z;xC65#K z6FYs-asT?}V==}%n}6srPvkbC5Lfjp{y={v&7OW&*Ihqa>vz4y+v*36A&>yV`4MQ$ zDxvwV6T0V-n`hXqUD_y{pq|GOTycZ=$W_jN4Oa9nooX^L`h^kKNXA%kJlCB_NsXB` z+(EJ1DUwgcVzW=X6gUsrp?_Os4{OoM1B+{H~u2$&uVFF05G6 zE!(Dq&Y}|V=^BH#DR)4(G?Fj?>sYbycXnF1F$aEUSTV*E zx3kqzCPO0fSd>f3xOucLJjt9gGcgSmTT4Mrc5JJxLy)iTq6_B(pW)QI{x$P~u`t~? ziN7)b{Yorj zHP5rQxbR$LC4O@XSWvA(T!B#bfjOfKkQX9!JG&u=QH{GP&SW6ri zP9+YTf#5#H-lND};5L2xn&f|ODry!zlrIh24}a@4$bV45JXz2^6}e+h`=1sTdNwk+ zcdK6`b2s8DNtfz0#p2J;sEH|KvgqaRxCcoS?#!uza2x)8Yu*KzpO!J}l=b}6{wv(m zgL3G2w$>A`xc}T0PXkZ;H$w64$oU}AHK&EQKk!otUYQ41o(aDQ_q!5)KN^d}b(+T{ zm)8Pyx^co&lGDOQrqp+_$1+(<_!w~IY^KI`PVC(KyrHsgrskdbd&yfS%zgjPnGCXz zn@MtEH0sRIc${LJDRP*KZ4gxWHA74hYyA)T9}nrDlQUyGtz%B9hh()-0B#)gyoGUL zo7EtkuqgiR-&TF~@uGEKe*Dd~@2v2`((sMqYSO2;!1U6$=az~FygutZG}VD;=w^DQJHEpsNDDgn<6Yl9rx^o3&yU=gdz=ijF+bCm zoe{+iKd~&!v{LHSgNL{YsC8LNL$;fKrK)02LJK@GPZon=g+bE_k8|Rr|z-- z*pA^{xbC`?{)!dI+XJsFn$C_d7@$p=Z)~R!Q`PfZk@LkD7v<3USlT|Mab>m9KhB zBCd()JzNrO_=;xED&a_Xn5rpxHD6h)qGR96{O*lIX%M=_O`it4xZ`pvWdFwugYMD(^h`L@*EmMTkDVGphqK z6umQzdv|)Ks54cYr1?RO0Y}I{#GQ%C{=OE*j0!sVZt<&L8epVX!c{cVl4FER+X>-` zhLEC}sMoQK)yYvbHH+??xGq?(=$IIDz;P)(Idp;Ym?d?A%NCAUAtS*0Q(QNJuQP3# z45C>AvW}FrDTWZ?SNI?2hS@ zm&|Grq|N9MBuDo}Exhh8p=L!hLUQugAnK zXj6f_My04yWb93J3NA*E6Y!`5v+q^v}V7IZB%$tJ-4zHKjI3yd+~_bgoDp#ijl1s-_!v|t-B}xu{R*Wo~8*GPQ_YjnkZu8 z`6clI4wCJl8FdCigm|4!Q{ji_=aObvsH=svC5#N>u&EWHGWIHM^60 zW7WgEuw95ZV`&+p*vAwZ0PcG4S#QexFFqYCDW;> zgEf*iF+4eGQyjDo;U#!F9&hpJ10JYE@(MDw9KC<`Ykys;amH?R)Q$U`(yM;n zu-jc9bd4u^*0R~Y+GZ9VcGmEon%cKbhH38a&UTPm7tZyxo6V*tDfT!;nz0Q6JF`%O zsalllv7O!)Qzxoc47r^VR{3SBGL>!NCH#A`L2uwsQB^}|B~!I)!Kcx{P=OZ>kXGpZ zT_VKdRxREMTLcruS>g+TlA^$(Kx`ZA3mEjypel~Or|u-t zOTp6+H~RPTYdwcW9iBPjRuF*6@J;3qDV#HTjR#&9W``?sY&t(=*45RPu8Yf>6e|+8 z8^V!aB7p^;wO4>Vl1s@p+8El$u)lK#)atKm`$${8g;Nf2G9;7g8veNKS0>kG?+?9Y z=JpMXTT zji4NqsjesO(pw->HY87u>5EPj<1N5|%~~=m-lx*nAQ68utQNR&`yvLt@T4}zc*-vM zp4rHztHrl)7`{4a4}qevfm(Ish6Z^dJOjleYCUla(;P0l)cVukFG+phguY+5{y3C; z^4g$pjT!o$Q0!+E>8D~*Ey~DjXh0wGbYVyk3nZXO6yz{(!fG60WK3J8DzG(qm;BV5 zO(Di@BODSUwYz{llSuE@k;7QjdNWR=IeGpw2 zZXi$vMlr<=1qREG1k_H)Y)D1Wmf)VbNgl&)1-z>gz2IJiftsO7vLW2iUeGM=g(Ulf zxdV#gS*QnjFGzbgtVLcay37>P`+@+L_?6qjYxxLa=-~yqHDo7i2P!<*`oAw* zLAJd%Kxr@oN(IG2VMPfQdo^&k^b;?j;9v^a19SCcl}D;iAFqHN@WK>U0XwKD<8>;o z1{SbqeFgzIIKADLbt-mCGiRrXE5s;16Bl-vKS1`WuS-hkEgoll(xsKoC=ne$Xhws` z@xItm#fM9yO`!SCo*mhw^0fg?s~KnxQ7oXS7BpypWn8DpQ1;SU-o12J@PN-s_iD*b zpS?4V1VC&DtjQ75ra}4Pn||^9A$f~n4{#b5sk3L`(Q}&pKFh)|J`f-FFrttyoTk%^ zoYyV7B`W3Zhm^cSwgpU^Cr!m;9Xnx%oT({kMy$ zol$62T4WHEj2u*~@Pfo1Oj{|+o1nv*<@BZAxWQoeaKHEl*4K9F7ytZSOPFv82)S^j zpA`Th2e_WjA?x|=(jv)@;L<33#Xxf^uU2)NNhWm$o^Y)7LSocpg4I6zqT=kPp?vWe z8IFz@KsMaku2bY!8<(>tYvT2|{}S2AZBBAwm{gnDnL>)qqeu=F+ZWlXNSo3(f5@!@ zv`6xQUwe6Us^SJ1lp(i$>Z&iK%&!ZtT=dj`f8em&TJL-6TZ(qWB1Ve~ur>Urk*QPO zga4f7_>LDTBQuR%aCpJRnfkl`*WZMc@1LG@j-;;5Fp-{GaT|1U5Ueo>T%XieRZc2++We7GSr{qL&5%=bV?*q>60R+Z`F}v0h&= zg!UY=KGTdv%=JNRD!Wf6o>(;<9N^PNJC7%SCSf;r^Z;T zc7O0GNf+pGXL}g?27npVvoy--kF;%l&d*N}&U%!e-PiiBKE_1hy0v-wif_GVS$F!H zte>ou&&JKoqR)tOd42K{p0+8Zgzf-w$1GSZq3(=I$Cirn)SZevTI+mlo4~|%+IogT zCvW_ERYbUDD?b-(xUhF)rSbnGAUs|jcvRjlUF*FWLT*^PvyBFtW~M1*y+=QDfxfh` zgs1b*S7*`nZzN2V=m=MUOcV!QmHza6dq{&^cvWqMGzgofu82~U<;)R}%4l?gwP;H~ z{ne*w>Vz2#u6D!8cEa-Zst+ZK2$Q>;qs&VqrDM6fE^R2Ok#M37N`@cGQ+e5{gA!b!q>wb;px5VT9R|96LVz^_8-7ia6?_AX09GSZ_??oRB1!xmF}LPjc3=FTkKZf#n@m*+@4E~QgjwU|WNYxS+nP6O z*>3vtdjhp*&WzK7GB8tk^KKgNjhEm}h9eUB!){pRbepLQAM`EX9%3|?n(&Y;> zzB~YVs5$|D%mq=G9KWKWH6mZFV~fS;A8qj{@PO_Hyh9c)?9pFz zY1m$4{ozwnuEH&7*E7#;wons)IB!@HpAk%E1kQ;IxPOrQN=$;npRQr zKWMj-T`ru6JZZLKs;5}!W~rfK?<`m||IoBjuj;@{3zsp-<_}W=tbS0@6&f0DN3HeN z9^&bx#y3S02*8~YB`S5%1JG}aHNrR#Kyi07oy4@DxcSuDC_{C)Dqmd`2vzKb^yZnj zWe;d=wY&#{(+$!-xh};1jw+kcGS<52Mf>aolXbL=HS_q*ap(J(5cHkKtry4!Zbc6+ zJSUZ#;VqY9vnjHbicJNDwNB``ZYH0+b7hhC@mCU=Hq9ooSpAg0-8+#vPs-l@-5-8+ zP6N&2(mY6#=XpKl54m-{1DshTW1e$oC_6A^Cu&}N`HG_VL;kwh^^l0>KsO{1s-qO%QX^Jvmebz4_vdyO`DqC1JJ03imYWXtR$8VliCxuqPWVV8T zTAeFNWU6Ae0*OtN^m?RokYfj8Mr_cFhkl$#LpT58%&X4PiAxilQ`X2knOtv>j!0D? zUxaf8@`b;%ZKM3cTR7z)x4Gt4Id7khsb>7?M{>w1;&@h@qU@$`L4jJaxCh!Bwab_* zp4({*vlh~ug0!iMV;(!B?4Q$L{v+8o5+je+v)7m3*v>P!jReTe63-7}fBMVVU;XzF zfBxvfaCWhCRBM z)_-G*BjZD~#Z8>>);gkpy67icziI+W$nx3$LXL3rSzUOU+-U}%^AvlQA}v&GfqJ?B zklUVVcm4ZGr60z?`s5EIb_lOB0}+WKmC+a>yCTY&q1`hrRe7K8C0USs8k9gMCaFNP5RnwQpbrX_mNw>6hix2`m@mJ3h!gGp1ZOYzF85cm%Dsq}_X^qAXU( z^|pKO3+e=mg9~76Jv|tVJ=ymnU?QDxXU5jCw|USUbfR^06LRWKv?b%rYf|E3g`Z}S zd_48;8&wRp^Jj^NF;;^M5uU0rwi@Y!w0T~6UPhN4k7IHtQ`JLvcwZ592nuP8>SC{m z;Xj9>WgMUwg%9)17Y{SHqt*?|SBFBAEVDhX>X1Xf^Jp&lXYw~4f(!2qfBC}O`SePl z+&~9HyPkTyXhS&o0-Bo<$=_&Pl=ybSm)gE~9fEBid!2JY7pg5k|2V*4%WdsG$2u6s z2+B?li`^GUO!lP0E&C){`HI<-d^6B&q}UXSBvY~2I+$6eSlg((XawRyS_{$Ju-b>u8zAm)zi+kP*81ACHn1VL!I7-YUjbb-Y zBnc=L6zyygv&FYX-XW@%l=+4sWKK6hBonAWk6*W zdWNor0+oruknPZU=6X01Ay)6Te&w}H%S4~cIvrM0ciZV~;Z8~3!al`;x%-26lFLvY z^k8nbSX(7O?0=u%ASj--L^$ZBT}5!!0vuq-yE^E7D)>xr_7o=%|2W3o1OTf;=8yx{ zO~x3W$0XmAM`ZSx5f`3$teBC`xx;}v%@I-)(=0}vM#u#FX3=}ZwVwCk@2i2yF~GE< z=pdIC_R`%TjN)V%jf^>WIE`HC+*ie$yadY}j?0=6D}9VFdZTnVjW!+I3dH%eb6VSGL4fBp(^~SXit-m`bX{g56CKR z;dvL1lN6YFubU_~l_Kk@SO|*)xi`dDNrT`rsBIu*w8_y}n1qSU!k|tTm@*~=LdK_` z8<3o9F-G6Lu%6DesOf>+GNnrkew zE`V;pNbh?Xmh;_Ll2Up^3UWu zB)4XRQ=6x*5$mU)_w5kmgJx$J{rH_#)N$YSK~4aR<X^BA-000@_VKYyJgW^{R$!9m!dQ*ii{EYgvcV* zP)H+lJL<0g?I@7@EEeDQ&Yt3&)$#b?;&f9p=?68!A%V$5dHce~Lu8E$mrZXs^EuKf zb|XbnsMy{Z)b~o_=c`+cw8FHuoM~r~PJ3Sz7>#n~rr#3brG*Y?4z=}pEBMM65;n zZRA)tsMst4>2jUwis&%doWjsf$)Hn74!FX~%zoI5oflkRcK%RqjLC{@ER<)H-7aiJ z+RUuT5sE!Tkp?O@J?Ma7X%J-bfh)Gw6Un&nU7WBadJxoza7NL7q)PWqXVQc642SC9 zj$5#DE(TghP9AtCZh@&AZ=-mdx^odWyBn^|R~tDfp`Q=paD{Yr@Bs9x54&~IdRSK{ zGFf7LZ;#tz7c|UpGRB)14E@Uvak4m*{$u{F(|=tHW@IKs@dYeKf?iOB!*c(tKrnb` z4saUWP!tOj`CWD)<7Fdcahw~FEG~B9+jrN9_lQh3W_8B(pOOS_u@o0h>+CSINLwj3 zogy27C)=wv9L1%zBhnAxp@IsDq5ONXuv77v-pC&2gTg~LQyGoY5cb#)n@5ftK8(i1 z=9}AbVqnvMhY3yi_)X=M-;;P3o`Z_bMtlp!ZpKJYY(LQBFA*la0ZgJuu=QRc9S^!Q zC`A8WVL-hf>|X2Lq(;ZKe%^cKp5>ks5lI`*fKNm)meb5KIsb2ZfB&j4bTFKnAq1jY zq{pe6Ip_l}YGEkUj3oP31w7`J@jDd>F>tt<8IVqhEf^0mtsRdzhf6tuW74Fz=e1bY zlDsBC6Dw6F=yDiJ*taQjC5B06&|4ps4^`}SKF&t1}_o=%d1wBWE z?y@;PE7a!e{MQhqmK}L*gX|tiDkv^6z4Vg>-BT;+r~E8&ZFt-?%<$co;khWBF%SXv z2H;QEz$=;r6OFp4Du5BodHh+^@7K9 zPLmedEnXd6t~jp7<2uc4S;~B)_8fY$x8@=GFcVZMJ*j#~cLF~ScI-GyDB2Eujlk_i zFd6r;#AV;u8D(-!491yXbmZ1Jg|(Fsy>7@%+SNftv#G449;e^t7jzMxEG zTEM^QfHtiLX*R!(Nn}cRkAVZj;M)`npGG(!MDGAljDW-Jb=+t$UC7DkO#bL%`NxzA zLQlP)o+3kTW5jV?I1IbR3|5I0yOtuWsMt3DTK{(G-MQKp=`GJ{&(sO$zcG%64Ir_1 zyvKFDb&&Y9?;iZ?;-|hqTI3Dsj4In~X*Ox0NTzlGi4H26Q%~O6k{A>^h3XQ?Z|tP6I8{T_02% z-R*xe>?m&wyT-}PHDeBrZJ-$QQm5`b@X_)fwFwl*V?W$LD!75dg$tBg%s_FFV(Te# zfQoIGEe*qN3Y{kD%NLk+;8(QyY-b+DY-aS+)A-9|dclR@Gd>*x%-dWPwSwbvHekKC zPLr<&e*}0(llVC^I^ea;Pp1LTglU&%i#tL7VqZ+YfgT03&>#vGF(|zZz7LW%W~%lF z*DiFxBFZ)pjWRAaZQ75oKR15rt}%f~A2}_Z?0;=QbkPh%Cn)wPMVhGChq4^;RYiy3 z+BdF&fOg+kuSt?)26>r46k0nYQCuDPL~0HgRQ1HY4EA%I>;#bPcbi$3dz zBzIaMPnDK=zSkv7j=^UayXbI-7T2f76YJ=kiUERDuIbW2*lLoWAcj*;kaSp}mU(um zoe@xX%nv>8-_z=a*4@ZgY7sFn-!s=zRm^2$bXHUZ=t8R&A4_QsMZHi05*sM!6%<-AI|HXO3jpGmVLwV6zHaDkJ^K3|9~+mQXjTR7!l z@MVwxYe{CoCC%)@i)|}37ZrX3UhRrruTI5hi?-1{9>ufb!VdEj#W$v-No?R3hT3n( zxD1tT$S~M6F1NM&+?&%{{>##beNB*I#p6w9D*Sr+UGjDQSWVr=U`b00FqaQRw+Nnu z_CRn2#Rkz~Sx;)!I?YPYW&Dle#PIv-T(#3N!vQ8{^Pk-~c*za2w=Z0#9GO#w|8+OX zer@Iy@TNws@XII`2-%7a{iO2@Wsj9WmlG!(cI#BsPTxeT0&3<%;o}t%3Z>M61k__7 z;{q}6O3^@sUYaFF;kjYAVsSeZ&taW)+tk|W#-5Nq`EA)gCrnD?g^U?~oZ!@cKX&1I zOB&hN#Ijf+n7yXh4;`b8u*6yJ-p*#T%YzakSGZ@YR)X?QRY0GjFIuN*14d7zOQ{M# zJ?_NtRs7>IPPZ!Ksg^@{yu>It;P8$2Bg?*SvMYZo+INQ>btkvX9!MRf`rxwNkwEy^pM+jZ{*}L zZViwdeB;)J@LsQy=*uC!j!Aoy5nGZjM)< zBA>2|PF1wXfn79FoUhJ+J^<|cSSh><(IiI={HVcj3e=l=)&QZN;4GYQ!J<}XF+%cit12lTYHy}+@hG{s-i#k8}#WBK*S|e{Xnjt zo~kI67O1NNj4i9UdaC0kd9;garx!c9hRP1?I0z+U-^Mz2%)7v}hIdjOjF=M#b}C&WYOR{57@Z@Yo6%%j=x|{(a{An)xR4;q~xnD%s;s z!1Wzv*PAK!Fhv@v*j8B=l!sts9+cRCYeJD6T}LOsSrL{ZE`s;awS{r~2ciTJ_r$7b zEmn?TQ9y4Do{wiiOBcpDw7A_a1rmxK!5yGc3LDs=(Jlp%w^muZG}*gOi8Ztd?8<5H z?U5dyj>J0P9)wp$b?@U+j*R-tBX$HykNYo? zjjxUCYo*oXZrEYSNcZwLha6U40giMO1G_P!+yAz#8N}N2q8>=o0v|}9_(SEC7zv;< zRqMT-ky6OW|UkqNLLdqg=_$q2#X?*%G+e6QTTd-tW%yQ_nm7vOvA_> z3E(n069OaO%jzoTc!2y1_7FO`u>G(?=#;w+zIU|bL~Io_Y&#qnEzEF5gHG$vM2 zHHK5}!a%WNW~wEnz(txOhP+}5uK*2I0uDwKvxkPPhWI-gbb8LoM zZ(QOzpUrms7;!uL3EOV=D^UN-FJ}@3X=IxN|4@UO{&l&V?rdws1sylNqE1D>-opDf5>sck~ z@hf4v)Z6{+dNjswFkT+gvFB;vgmG%~uK5ptW`fe>r{A4P^xUE+E<6P-ojlq~KBU-Q zihK&4X;QC3}z7$3y4%BEpH%e|IGSfTb&s?byc|3S{2NFb;CsA0i_Qv5QKFsJ zej4=za#X`^cv26WE+jV#CmDAPQ5 zEx@k*EE)yxI=owXP`T3p-}RIP@0Tb@6pV7GoQcwx%!C9YBMYIm|_Pg^3d2bS7Y$Sw71w&(E-6uPh%%+Bs zU;|h?_$R|QvAFg=t|){xQNDVFR2dVB;0WK7r6^4|Du(PdKJ%zWk4*~5QwY-{X@`Bo zZh7iL`lkFAv@e|?*c_p~At@H7@USP(*t&O5mcSYg43L`nb-VNjutVV&28DQR9h25ZfLF*Q|zXSGWW z=M@WkJWh+3i+4t~Li4pDQ{FCJ=5?A_#;`LY`*Hi;q$m@PnG^+3o9SfnE*(6AeJ2t1hHX$Milct^@6R7^-l z!2Dmp2^Eu4OAck(!@KRm87C{e+ZoDU`jCG)Q|q}&njf@`)yXqe`Si}IIt^x_QX`ix zI11sQ78w?3E_ZUp!sjDrBbr8B%L&63`*8MG?P)M^VXxFmgGnEBXeaX^5nZxqv9MT- zRhGDx(I)c>VGRxYdt4q1RzI%+KBQ zD`K)CiWO%pram{nG^&HnQx^*{uG%4p_l);!WZL23G>DhJG3<73e!4V?zeP33%U9>~ zmOuiqc6z1%4N@4I4jh#B>28etT$~@z&mUClTO&8OJIxrdIG9oYCv|Vmvk~K`1QbW zxMu!xNv@=mx#?7X4m%8rlTC_sSU>6fmc}xE`${3jN|bh$dx`Y6PnBO{_!C7ZtJAE~ ztmEB>RW0h;ZSgH6nX00w3T7f#Kb%aS9VU$HrPi_W)~(W2(WYtWJ9~CylS*zmB^UNp zTFqwDLlg`8thH2ZjjB|Xs;HU?!LWRFs|>l&o=8s-L)^^RY;^7&Y&Sw3%|3{{cGC%h z4$rHBCxqIPXPei7I4D)|>_AgUs-jhPYxYI|A-4|imf&^skI&3k4=yP3gIaZ+Dw~Oe zh;nxHh8R0zES&IVeG(n>2v7-sdfh3uDEKUyBjcH18IAV@#*2GP=j3 zQ}HCEX=>BdOjUFA@>$4^x7+`){9tgss8p1qY?Y!zkR>(1J5n@o*jZP3@q>DR{8beox)YZQB#A{UK`rIA7(P=AIN#jD`N8~bAl)V+S^ zNrG>Ix}G30Bj&TQH)iYfb&AU92AfPe~oUiq!iJ zyP@>apm#kf@W5l2NwKhw*9Cq)ij4Cc6~@Y7#4tvFIs_lfQ`Glljo|JNyH$B$P706J zMrA6Yi5&jeG+T}jdt+K1n|Im*)9|n(zx(SVnA#Cs8ih0;pmqj}!PGg(>TJ^x+VXR9 zImZ@AaJiqO&b>LM`-mkZTsEq2g+&aX&|ADdMJ_ZbE+^$MzpSGE{~(V-i1j&VtWv?6 zGU)|N&M_-wU8lwCN&j1&ip!#^!0Te2rc-f&Br^9PiYJZ}58rB1~2%NrNQg%vf3{pw8B zW}-!1B2?Ek2sI~1>)2IJkII|>^6<`9P&{{`lXqAL#en{UJHaMn5t{XvWK!_jEN7sX ze^k%HZi72ZHKnET45@A+pgW#jKfE z>xl&P7+dM06W9W15Yb+h49QaYa-MR(uWCM_243j&`8`imUu)z}It zTb3Y>MA;v2esJ2J^sx(vQLLno5A%ljSVWtty6%@Dtc>ndV9$W|7)eqb3^rERBAg27 z7CA0r9f|_$BgY*d*6-!WBWE7|XVA}0h9%@b-*|)Eawp5oDp&g{wvQq`RP1s-(rFjc zjUeX$Ph^TR6FMu30yhK=yWz8$5B(aueDmouUb_X_ROKyR0TL}???tLo8}D<*D^;1q z-wiV&E`g&q$+P;jRxKc2eWzs)}jJ`d{H3qub* zi&Aut4q?U9!u zFCFi5n?c5y7LSYts8=%Vwpi1us-KYylo@5gx!#4$QF#F`UAjH;9#Dy;&FE5dF(lk= zfi1I!)pKk%7uMa6u~xPqm`*%Di2dm=V}JGEKm7T-A4pbE>{5!vIlGqOrT?7DZx_Em zYh+<1R_`JCuZ^#C$m{{Brq~LK>_Q@>{lTXs=?oUVK)ST%< zZn?o3sy10}@EQ$}*o9UoK#JNgi@qSqQWY_qXSB??BW&?#rW@!1DT-8g2#SR9Q;Pih z$q>-cCW`OOzB+G}&MM{k% z-`FyQUG_VIOQaaT!DM8fBt;#!AXRZUEF)5@Cn?~{;oC%UujH)Xt=Y)^(IQ7@3tK_f zDVIUb@45LWSS@zo9#Hp^IuTy!U@$la2sl~VI>fe}$7%AM{Oe1qvi~;0<|obnw20hy zVc4uQTMsOr%nniHF%{b(SSFf(2gS<(7LNw&oN3z)= zw>6qd%_cbSw73!+v_}8+aJ)AB)SFpBOGvA#O#_k?{!if!AE_3rt0d<%_rp#K_CWt* zwnv@!EyY28Ien1drb%HMWA+3d=8x<>r}<4MBpkaHOT&)-?KR2&+*F(cGOPdU(!}#q zlz7tekfz}M!Ix*Bn}31+RCs~LHYJ^=XXaI8#N8Hs_00=kiZW_s$JwNM&9?1nth zM#dpLYpY)3hAIt*Fb9m%3olwl#vGl zLQ>hIb|N`B(qq+5hvKi00jYgBb2zp}gUx0nJ%6HgY|YK?Gv3IkSc&^8C*(P~I$G((IkQ&|xT5lmju!X5$a9w6A9kD|ME zv>s}^@PHP>sHV&CF6g~45f%aG2XKTS{SDS#r78{p!#&zLM@@_l2i5cV7}uViEx+jF zo1hcA`5%vx72KfX!mIc^Gw7sI>;{S?QLzK07Yh2;d&UXtL@0H!aaNH}1+#;$V0Q5v zCom!C?F$m0JND39(AZW&{n_nWcBJvT@CQq%6f)1y7 zFtM!h$)i63W;FvZW2>yv^F91;0u0Ld0OJTqy!am00rSCHwckIRz;R+x_&<;{+~yq@ zhE0zdY&t0R8bvN+KDbSl#7B~|r64$#EY6557Gq|;O;teGLqAlG-(3YbA=#0S->UUo z8al{JQF#HDnBvUx$rf^r%!MV7St^aV^r40>lrb}|P;ZUd*xBf3v=hEAHZRI`e@ z&BXf@NLC5g`ImU5&8U=j2p)R&MPps+($ELe9Ip;RF0Xb*vhO339+U$dT4f;FgdR^x zV83XU=PC-nlQE0Y3VgEo=cNE1AN*cyuWUK`V(?9x^zD`UV=r&6bKRt`H3~F6Xl@h> zp+@iubha3h`zV)=$*KgP8b3k~gnT5@KJ<>4y>#SW_6yp2oGSusy|2CJSvNrU|C3{W=3Zly>%6`P=38de1AwI@|25H8*nbeo(cg-kW9kzjev zmN7a_gkCJl1bcyD%k{th+tM#sYU?~-TI#|;up$Ih z#M?x6N~-bS*VQ&X)%cxxWI*N^_$Z9IKNb(-$ zehf5s+mto)abqGpVau_x4J(#cIsg5J54RTjn-KB+)057TbQfM;L!MyN7#CA4H2vpM zu_Yjh^XG^1j_|%nAeXtRD20sTV#f650K6%nR}gL5)NGH{Q?5yGDcbpko|rArqM!i) zLOdkbwn%$C^2JEHYnMr|Hej|5|BH=+wR_vMpWD-~nRNQv7ffJj|GPT{q{io1vQ!%f69%Nse z#rH$@F39!9@W&SACtfIFob(p{y-aZ*GS1EP-UT|%y_F{GYXcUo^pJ=S=24X?Sg1 z-m7L$Lkq=1MDREj+al-&G10Bl2O{ps+cXW|e!A$KW~2DDV7>R60LW1qR>#>ht_G$< z1%*yiI%~iym6tR>O`M@L?yaF$29=5K`KO6<;CLH@Yh&aXP7^=i8SkLeAiu1pC; z_wm9?uZ`j}hJ0xf9|e<912I!oK&A55u;Y44ob_akep<%!lnli5MpzPaTH(JOA;sMn z1vTVIgbS?7N*jU;-?%D#&r+%O`DK&qet%*`kJc!7 zP(qbZsBMgmd8mC_EW|zIdZ_KekZl-tbk*L=zp-@;7k3rD^M+~m3H)X4yJV9)*==UF zw^JR zl(h>Ve_>G3LF4h-g?&7yj)<+l{%G?$viZ63jpT&kj6{zLv&^$NpcI`;`lBd$sG%x*8d;SBmircEl zg=2FCX4C5?icO`+df=>8Jf^XuvyeVNwGyUGV~@8MN%gQjA#Y(@R2pw$64%C&;|dMC z#+$>%Z%+Pjm1Wr0WlO%T1Y+~l*^$>IeUkMa74x+9Kuh+&Kp=rXxh-SNVQ3t7Dkqm| z(*3M7?{Aun#om7l{60D1PVSgJW$hHJqsTcbwjdHK%!cGx28*gxb#zrgpTf}XgJrGS z4)48Th0;VOhnE(ry)7$(R>ZA-_aLp`Whit_V0X?q^`@cm8(XbMI&V=c09$TTltq#~ zaP=|}BWjX%MK zqQF%s4DFewk6J|_J_0%8Nt#hMN2)Pss9V%O?*zM3vkicex+XD3rTYT$h4DvGB?Mxz*?Qshz8L?Lm zQ_{=$*ld@o=nymo?2*>QG%^l%PoBrvc)|YsogBM|P-iyIF`?u)wVFm!;tsLS(Le72 zimjnwaUJ^!@5YSd^c`VhI4-GA(1-b_rkf?3uM4Mso3K&-E@awe~ey` z58PPSqXr=6j;?EmKrbkuFZ$f`&kMjZU!A5ya5b=j-m?JT)p~Xamisr*JyWki9xN-U zhZF_wo?p&e;RV%q>cjjl`8Mgx=RmG-auA%ke(X!n=LYA1KQ%^M26bK56}A$<1*&gh zBXchzAyTI)1~rZ$`CgAg2pygdEOoLj!5D*KD=fyi)Tz6;!D9G(Z=d%v8IzhXf9(ue zJ64Ui3;QKSW(FpMVmDDRGOX|VgH~?i@ZHuC^IuO7{Ze8=hx{*JY$NNq`LnJ&iNYi_dSc3=*ew*#A{@0?fo))BH5wkHNYT(JWf zayU-Lx?#}vU;XGCCWPFX+xRIt=EBZMm)V$Kq}cNmIcw}PdgytD6p~WmA%E=+MQ_Yq z@GsDAKpFO3|1+YGr)f7#1@0HsP>50Hsdx^mjt{$`1|&8q^(ipo-wdsacm218UJtzz znhz~TWih!5Xv)TZ6Ku*JoB}j*gWgH}GxPw-Q`d#3C=wnSz+j1{{-J* zHzb=w8Lwh-pA7Y6`&sM&uzAfi4z!FIopHX7g^lO;S+_J}q_-wLNqdwiHeoAAkamM4 za91k6?Ep0@3t@>S0W zb&DG4Q<7c?dgMyVcu1-7zXAwb;(QZ=0Lu^IhKQfIe_i`ylZjZqb;%5JjhmO@!fTwT zW(MU`iv5IwWEaHjj|O)Ls^kMu5Z*y|D!Rh107d-`a$QjfoGh0@3PW`oJSSDLW=a+u z9|FfVA>>TJ4PG3-O?EkI6LgBUF_Q;Xk_;A4)RY63qcDNdefwLBPy8P zo<&SW*b+W=!<&`SB zwNk6nD$7(|7F`zQ(TjyeK$>1GJ|k)YUfw=f8CR>jj4XB&;GsSnHyhz%dcvPw=JXBg zaQWl>zguX@!{M@_dn=q9?f!TCJ89$JI!!9b5iACM6m(`g>9%=WgYV3#l4JAf*p}QW zKmMx+gfW67)Kk&)?*@#laM%;8NhUY<#f5_;zyUOBJyc4upn$QRicJ)^$TA|Ygx`;b z*cm7Oqo-SA|Lp5SV^ScoDkp9CKaLB54y)hEG@ zqzE%dDK~hH=}?&8&_VA$JR7ye()*AT;m<$#;eW;c=odeX{r!)A`RiZAN@&)NP}CTR zbK#(g71Gq*{wVU7qQr7pq*LhxNe(QWoIqPMGIq9sV&v?Jd;Q`Qzn}e!$%=gYB=r?? zmYWrEVNl#Rvm!Ss_ButbP_fA6gEe_Tl|ho(l&@raeH?*HjZcCP!SY6@9-0f3msqrf zmIA21)lEJ~JeUXqD-fea4Qrqfn>id&I~{9ikdUlqKFD;G(3R1tit|35s)mr2P(WA^ zjO-i`ylG^*B5pHnFy%c45+FR-Cr26&G)r0}UfmSbOHen-0up| zL}djHc;Ram)4;=5)WyKtmHICB&@tuA1z5-LjJo29QU`V6TRa{|?Be07&jh{yRg76d zDMHV>8y&3Qu}5e^-)}0X{GP>~z&{$Df>zNFO=9(lMvXq#M z1gxO7+U<{e*Ea|hEvvK>G{JF)!V7GK0}ok;!WeGNr0v_b&;6YV9+&=U`CRgV+Z5=+ zalrLvW^37GcJX8~M8)QTJOM`V+BCcP_2m&0k1RO2_acPU{S^c8auBRL5n~i zR3~`gHS%rn{g7LOpk8{Pf0F2>qxZMT;)GaK_>`Z>^!V)$?u)oV3e-8g4iYyF_hKc- zCxSv+FLhZ>gWMnXe(hB+sAyw41QqOK8fBFVfSPrOuz$`fa9@s*UJ15c54o*X^wL|^ zr__4s=9%?xlq!$Pk9Z8ZwaF62E9V~O<2&QoSwUT(v!l~&X3BVY$9kp(8iz|@F;^0u zsa!F)A}9$MP4I4YqMP)xsr`n%)ues~)JF>{{NP-uKF2G&1ihq9hTmv2)2ZkOO+PNy zk-LerGDX?4n}Z4DWa6k_|JhAsDbCGh>F2_pnH7!kGM~p@6^iZNSOJ*ID^R@mnRNMr zR-Xjl5(;Tw*UU?u!UYB#MrQ0&x!`E*J2+?Wz}K@=eNB^J`JCpp+Loe<9wj& zMI4yq16pPEL@!@M^z=hXn+lHI^uuKwR+__`e02+51Pgt9^D6O&q=~-ne z7WQ?~N&I3V1Pyxq)&lP?F68y}N6d|gEgn?>Fk`1b3O9peoz+!uxA>8w+Xsx1t#+C4intnLL!}^Fi^av+-Om6zsN5mer z3m2ZJtf*ZeY4}#Zg2;4fa}e;2LwpHS8?7?!5@FIC6EPQ=k2VJwnP2GScNfC`@()VW z%v3V`ue(V$HwVUrSMPOZ^H>?hLTzX<6}!s4LvZ#VE8gi^giTTTu*|}yx&$^ZtTcEF zBjmn%BQY;vk94aCc48!#FD5indBDV~>$0*NL$Ga_Q1$FwzPOLt)6DiIBiQ%5x z2bM)!k|erp0frU!##F_icSFdHh9?m$-%@cXM zTGID&=|2~am{^hi+po@4W6BbHYqeWvl)(xZtNdzI>C*G0GWsCUplVZ;d7+0T52E$* zM?v~1JbTy;GopQxr;=1f87R0xO9F70h2{`#wL~Y+HzY-KrfV^Sm1+pSWkU52EU7k3 z=@#ww-xW3>)jonGSc4Q*-jd08I9hG7Tcg!3hyc)asn4cYAfl|*^LWJMT1_$Zk3Q$$7&6vdQN8s>AZ!FYTRcO zxYeT7F`hbJL*ob_z0lgc2rVx-#tBH1KK;V`zpyL?b6MMyZ3d+E6uXWh@xUY^ucL8j zGu8V80U5A9E~J;p)_RY3WMA;(zU+`~pm^C`9Dke}DE`#Ev(8f3f=iOfg;&yh%^>tl z2BR88ZjQ?fJg_i5Q9R(4F0BNLq{|__q;u|Y;O)7ndegk14}=t&qgO%K%Z4crCC#8o zdl@QZ;)E-_kiWQr9&iHn%rh+51}$qA<=Jf)W3u3|b{ngoUhAo+kw>5|_+dmAy(|p*OX}yn zXZR~ciMj7g)kFCy$(=dPAfCM@01KYaOfL+;Ps~&;3)~-U)>vU5h^&EYocl*Xk&}Iz zd^%v`*DX_(uT-Oqm4u|8#_Bo@&pGHdQ?(u% z+I#t@pg2mW=^+DPC055c6$sCbAO~bTcdavbSVzVuEz{VyO>@@V-<|CsweF2glL&5RJ6z z<*aVQ!gqrnk7iGj!!<%1dy;48?pHPe!T6FqTL$a%OLFW@*eW_Z7uMn^IMVLIJ%SOi z#s=RbT<)?%fAibEPputlIlH{J;W@iXbyIbO-yngcIGfi;vcg+PMQB#|?r_L!p=lsp z(8gaZ2Gvd@Wu`9FM0s#DP_Y>aHiP~os8LC!_ z@r@JoTK+l7CTah?qR3RU!FYHE`UVW3vV*sbHy51JJgPbJCq>lyL z0vlc{XZq>mzKzqeTxb;NCs3*TBL5H*L_iBHZRYs zI-*7SNVGn(hDM=z1Fv`1sJ{b_#EvI9>~kKBHkz=*cSH9t^Cjl|WYw1I|BuA8RCA_vnf2@z8aI z^~()(yE4&z)JU;sL>)nsJ@*~$IsUcpyluU@fw_3#Rb z&ScbjGt6Q?W3<2E^R8$8p+4_sPJE5kSCqS}!)^Z%u*O>LjI&da4P~=ko(-@GjA)G& zR!+MghKeF-^aTkT-tO|OlwXm=&K{EhFa$3)m>s%iEPk@XY{7d9$;-}eik;9p^;l4| zXf2a2+bnLq=VTh%vs*JOCcdeQ|1x}zXl zzfW2^ca*SW1ElAqBO|;tI=Ha{5*1#(FkNHLPfkSt?*>xHEvDVxetpu@DXUzo@LsV)x357`Q8;^IS6cCS%;#P zKEiL;qF>uXI>+~_H)s?o56I%YDkaNdpK;+%*}@J*r`Hz>;NXF`L6f#bUZ6?mVITd% zRsyfRqL5qS`_Z>~W)S`A^kHq z^v3zhuT6j14gKmBCKmUmg(z!gG`Jy%nS=B1&qwXCZsiuxR~Y(}Ei0djP8BzRKnF;G zu6hH=qk-PSRJ5~=gPxP2GoLSX=h#Q?Bpf*Q=Fho`RSX(;<@<`_@0(+gSiIq%$axOW zO!qCKN;}20P~-*`T@f1Rg&Tp5AfK2#v6oaxa>16ZBD6Q4H(m^O5kDlqRDniuf8IQ5B~yaklnP=X}O)Q5J3y9crElGrOH1RV}4?M;hHQhuyc0j z?47b^Fm`I2)6M@db>6Pv#lbxRK)0)RPx8x=LVLAO+Z>~nbG_oIa%u1d2{!Jd1_ZV$ zKx$2@MBvS&=!vWfYJa_t)bTG!VujCvsFRRpSNIv}BwJu(hqSua!!B6oKx}rMI9%wn zkujP0zscrCNfhgB^tvFaQZ5_C>|<*zUUb|}487KfGz66{pNZGcsZYoXYl zHnmUQ6gg&=scRo>uZL(w*%T15No` zspxhfI$>&D=e}hgs{O{y#;N6U(54v%I!0{N`rR@QTi6l=0}371mkl&p>5KuNk{zU; z21A$P32xp+H^m8^_X^~F(rZs!&6%i$Q zd2KaMcSf_q!{nXb2&-{8fB~ppROGuZvW`aev4pX(>9Ggt&xQq8VB7MZAuYeFd&fK@ zJ$UzTeoPK?xbAY@BAPW&%o&QDqN34}y2Uqc;$hKUMLdI1ZAkJ3=YS>RE>WIu2Y3T6 z5g!C?%2<$^uY;KYW)>R#2v8F0C_ycSbP)P@NaKPIDtNGZGtjG4ho|QNmAfLeNQrB@ z4jOeDF3C};!DO0&K&pdwSCbaQTGd+UtU9DQ99S=MNRG%cm^=%d=O3|zfIZJx1;lY* zdE@wDz8O?*0d7mkat?!PhXtsTDP}!I)>6?sJPUX?y)V9@!=f8DCnEz&hfS?m)UTZL z!I(~%Y;a-EcslDGR=61d=YU^jTd$s9k@?8RS{oZ7bD*89FY@TD3Q30qIzlSw!&4Il zDNH$NPqpwm#)gADG~QgrQ4bX~)4Dr^4Kg^qFtX80Vu~*qNe12W19AFjn=_^%g$*Qz z&K=92v1{JmVNJMvMId2=H;TnKV>#JCr%Z2$MQygYL(vb;qPM)y`{5!t1-6Esvp?hf z(P0xV#-<^}^l$v|-@l$S#hhTYz7ckfY++YN;V>Dh)FK_(O)+3Wnn^{MN7XauUf<$Z z=)FSl*ni>SsH6OaU?dV;s45RyCOG6?=#B0%hoj0vzfd3LH_+#yv#30@OK_B5=slpU zp=(L37+t3u=zUUGg&B5e86ib=K7fsP4DVTeWTqJ{3vOh`ksJ=Eqd-s@#(FwPF(ni! zrlN;J0x}EAg_S`EWTw`1(3gxv2LO|8mJZ8XyF4%Rwnb%yXMue3P7Tg5Q075?LVH+; zq-{=t@(e*2!0xbqsIJwychV0$AM$j^-CN~pbfXjF%c$XCjSwfngI&(LDXm&E%M701 zlxwR=?gRpI8pFn@jAGzRyPt}_K>9pRzOmJPt5?p9T|lPJ_P+vk^2hUkf02q?0AyEK zM(WjjrZ)NAnq4ig672yJZEpC15b*DjJmlx{H_d3$?&a^6C4}AtYbi9kYn@dVaMtg< z9}KwnAS`P!dE8D;MOBG%W;p9h{7Q@xcjI|M{>N$7hQ^%TFtlM&yv47JZhXJ#n+ep1 zC}&wYtri63%m*;89O2i{{rp;`>Co1<+5yHW;bVGUHoshGdX)b*dPkg4)+(I^UW46c4#>DGWjH}v(y?Z7-?qCj^DOkr+% zo659nJeN;0ITE2LhaotX2m3aW=sQE8eKtJ*T^YIo{@$`PLLAM7} zZ_GP9{fgv{*BX#;0^&+P|It@q4;M?TxL=eEGC79VmML`TBXujH&iz(Ij@Y@R0jG?O zIfZc2H`#a_c4NcYy`kVE2?-%)(A@fmJB1{h-4cq!e1#ed?3Yr^eu@-P(QW**^mXq~ zz2n1i8fsCUc7I4vpbTW5Zb<4wA<3Gy!xLT8I^EIOx`G}EikYHUm%U~X=E@KFZVB1p zt5>fUqm*Iqtg_bz-Lk?vfc}8nNrk+tpq{!{Su)e9amtxr0cS4HcOLkq;AOY9$`fQk zMvWS&C&@->R=_er4Yby0`{zdPj@l5~<#F8ixHtYh7R1h5b5x;*?SPN!%~)K_D&|q@ z?+?;1d(La(np{)iN_#24H)`*^dm%uV>m!%sr^#C3NwO$Sx=+~&kVz1D>9zG#MpRg#7H6zt;Wk{a^g}xBn$u zK`~1y5@Wq`e$jIlP_XY*R)85d%fdklARV@#@!_BwP9bGME8r0}614(33HBI+KM!>3WHK~x zY-7yG*crN;^M0~}Oo?Q$LnzT zzie_=NQ%5n)C14)XZhr2fa~R~S$+c&Mhz|Jp~>P-XQOuI zjfR`c-SQH#!Oh*}Svt4a|Lpv$vUX(_AG@S?c_#4Ak*a@4;5BNmEVw2^AuK%7D|saQ zD1rAsLua}Q8&3wrQII^jfr}m}Y8UldM-TGyFpv1a?PLS$8nHVdsS0tJ;xNA4r-AO3YzfB-UXB?05HmH~yh~LB%F6^B z0q*#mB+GpU-L`t2BsnuKsm?{@Qn^$Xn7vl;T~GLT7*wN0%?qxVyJdKKnTss{T)pi) z*)V|=SbV-aC5dS0pJxIWs_-7&=5-BN}}geT>}I z#wpbTut@Mp0n=^l;f!TNUVp(aJo3w2{die@XSqueN{rpFec6qx$~0Xhj=wx0D;zvy zP=}?Ow@Q)6?}%{0qIYDFJ~?VHTAy4nMh{q_b=(ivUAat|L&sy!$EU#>gVhbV3wPKP zAHFw{Krw455{ncEbRept>wr4Fk5@3Zs{2nd@M8EpJ$Ec0b?LwPz2BGtk^Z1&0=YY3 z1Y-#f4}0AB;ln;aF%K!y_jF4NrPMR&cDj`|PJMmy?nvBcLQ4LCd=QH6gKp`9#o}yL zCder0)y-jvM7N1&I#3~k-Z4pz)FfJ0L2pqPh8*x{Qr>|cFLZp(3{3zAf3v7wp;zCP zWx{Uaicr*O{6cX>qC+A-8oWNFQ?#uSdi7QRF7H9NE=8&wyBm>M)2(WdoKWuIqn&G- zbXj--Ux$VfpxY3!Ef7772Hmg+p&i;G)1eyJsK)Lpz4~!j0nezvP!<#eEd|(Udr6)q z1$S5j{nHo2>}ke3mXB-)nQsMM`pI9H!#I)~HU}1XpCn)g5+`g?CxQ@1GgNYom;uD_ znn<~@9yG*SK=P%UcWNpur{bB!U@Ys!3ClD|ehzhk*c(dH$ST zO50%Wm+)IfC4pTu__oryRE{`R-Y4H8%%x()dz1}cxVUk7rNrWcXED|m*SY$M?RYWL z?3Cc15gllr#Ddd4T1T?j?K*R~5_i-htJzO6MHE!$qW2MKd5&7*m*=%kQKY>K>VscI z(*?N2jnr+E4HBEkeC|QK9!25k*=cg!_Ls>mDro zhTqV+Fk}I<7tKmeMc#jWTL127QR-X+?7z*fLBa%LH5Ft)-BCyjm- zWDq}^mk%mWPXwyBdDkhb08?l?d?+xF=#W@Ex>cJKLG%pBMuVy{zyB<;)_CXayktYQ z-RM&G^shUU%e}VD1FZy)+alcwB~rl+G;+Cb@jeMzJ*VA8#5wCxaA?G>Z{mXItY)Wi zHT!b@QAY=#-)bl_=Ox@^F$M7_}i*CKkg>GIToop5%Oy1;vxq8`>lPp@-eGH__QNlZ;BE8e8Qz^GX9@GBGtSBIN}YSkSd(i5O-X z6gF&Bf z3}IOY*Jh?anVKdB3}}6>%9@~9W1x>J%RSEr>iRuVBcfGRLs0lK4a!Ne!bSKW>*(n? z_Ffb(EnjB~&x;$5zb7(JWDCx3Izm=+xZ@2P9EWvNrBci$iX>6dar`|JtgV$s><+{* zg2JWAs+f7Db8&+#Mr^W=cL_zx*2m3mxNY6V`Nvs>n(=pLt@~H&P05$aOWei|YTeha z$@b2_sX8|QKJ+JLP;0*Rcs_KY@H&iL4r@GHB)=DyM#JOwV3~5651QoaNJEoy2mstYRZ!MTd}>kd)~Spfgw$i7CP& z(D-2<$znoc#7(}%v(Nh>jbNGLlLUhdV>7CPJ0iFFw42UrAPaVaNTTziX2d$ARLC#`&jkVh5C&O=qSB zn}PDj4_nH}E)E0bxCKxSP|QAx6jIU0f>PfqqVK~V&ZpSU)nj4*)qB4vb2U3$niQcnKp--~7DfTH~9twyqz!vQ7FBA}#E3@~(4 zsOU4&^3Wk|{=M@qG4%>m<&X0!n$$rT2n&R$)`DyUq`qNAyuUt+K(udr!JD$e~oy3(>lXeGx51r+ALxsx! z?WF6g<-%)0EvjtwK#=~mf}rvs*h6L#XVpLf@3=1<_eb7k4OmGOdKY@Pim=w(0o1B= zX^Q%c^oAs!DfEUa4W!T6!2Y?ZN)(idZbIUH&ml7jwg`MS+}Lr%0fgrC-0T#XBTjJ2 zH*3iX4o94w77=G7#cZHRA{E^wh~*{nk^~#4T;`nxDTKZAKb4{nQHA6tUF_Yf8nZBC z`uy!5Y6r$^6qm3HHk413d#bfwyA7dp6IV!g$16Fc4ajitpa{ODWBjS(2V>tSKZ=jg z!An_v^4wiVep0pi%O9CzMaw^Z;~V4>hhxQ;7O~t9vns0VAgl4`C&EFz`}nBCig90Z)EjL!aWky@xblyEuHGy2U7}Pq7`Gf+g=ICZo7XLQM2FhY4y*y~p?O{YESv`5# z6kXBG2V^h5M0^xp%mfVsAch^(Jn}k9@NWYg7<5Y%AZfRT&W9F1ybG|&f{q223or0) zK`S9XhIPx6JEMvQf&&N6VB_KCX6{Ah%cdG<50JTf*TXZfD? zJYzL-++DVQIQ8TvYgI!w2{d;@Gs&^|W^ARH6wLOcF}`7x!eFLb<9SSjs=SxvYd{6F zQnE*~Q{q2S*jj2jXX(7!G0Tzd|H#c#zUx12<2%^?$VaHBa`%f@ z@MzLrcXO&z^OdUqCCfOR7y~C|7)d3GV%AY4fr{QZrDR?)jY32)Q$X4&>G0HHNp`0O z4Hq5AWkv%b`|&!OcVT}mE4+^Xdf?kTtwjkr%V4n~GuR+r7Q9!PC|Dle3Vshe{j=zF z;o|UgAgy3`MS6IfcIAY6g@M+CkVi8Vbw7Lk0XY`5dkhHIcCPlV*D&k{o$VE@KtlcJ z#9#1UGsli^?EcBm$W{)=j)NAlBbQ=e4vR+U* zXKOfk4ph!b5EMy1z@;`Wp&ohQR)DVe_13A$swLD~e!mB9FX~dj(;J&xU17axe@wB4 z%<$)I!SLeqS;1x8y}94MKfw$aZR>AzWbsR;5tL>Dp>-6KK#?_6^cKj5aFKFJzJp&n z7q^FMWyhRJrhoB3n9r~aCwAY>c_-S&$&M+0O6&B4&9auei50PttgHg!({hr*KO-*T zU!Rue^+iPaw5@*I0xt(OMvQT)VsF*ineGhsf$El4`Dg1;!3Gu_u6Ec6B3RNuMZz>X zEd*HxU}5tYAv4*F3H!#;nP+X0{QWg+(h?g?a5z-hpg8rsdrejV_Ly1x$Ntx4*Hi_{ z2c(A1fR^Mcx?27~XMaHUxtL8mG@5ozGs82FF$y4ePrb3yn(W91 zA>6J05*su~tdAOK@PS9IgVl3%o8@=H40KV%SShOvAp(Eejf&wHjltu8`5*6g>)l|^ zE@f?OQ)ki}r?k@B{b1{CM|hF4PmV+E>Mo0s#>T!u9_dQg9;mIq~m_(_Y}%y>ZEu0GNiSs1cJ95Y4N46T9&utOZX4Aed(J()#Z3!h8qsHVb z%kq;oeZ<2)0ara__Z{V|Hz6r67XEcLfmX=}l1^y?+w z=5kl#kM}PkDX$EL9$ept?ZaeKOcq6Um5IjM@4_;ads}QUW~$QEl}-<5pNZpojazAN#@bC6BMo*dz<&2_stJlW|#wr zn^aRmGT8+X4i`L*Sb$<5#elZoUMjkX^vE|v_J!RBZT#b2{T?OKdiR#u?V6hIrEM0tKsv$66a+|rYXi;AAIpdQVKGLy9En3Ba z!mJ`j94A=fmpHf8rygYXaDpk8HO_JNm=7C~(GCBw8#(zQt2CRce2f3pdFCMFzjV!R za^RH-GUqIUOfAJ6qsU<@x?5E)ZH(xoyXZ5p74y+e)gEX&Um5}rDzN`tloC`Ysusiv zHu|H;dAjWJ*8$F6;eUkHnRr5-q-dZh42mRZ}{xQ(3VLIB+Kg*niPW2}3 zby>;GR=I&He>Y*lxrk!|bT~{E_eIvwMf11QDM4%bqX6dUfcKm@WPHhSG`@AVt-w0& zJj+4kaP7o~Zy{2olU4PKcxJn%jekvMs->VJxq)7ySOX^QsK2!csrl&3k_xR@c%WZp zECf{Wo8bi{Y^P5g#h^cbSVjwp=ilA&S8d0llWXp*eA#WdQS|e1*!?h&)8>8gl>{kP zdKQ@k@s|X&PTfV9sf&HPm0RVD5JJcjSwhY)GVe2k119Oj7OI0|fBX&W3P)NIG)Hh$G$E%|6eD-k1johUylZqk72j86-b%kQEO zc;$cs@6vz+A=$izOM@>+7Vh$_4b11asPNSU>0Zw*p8X)DhilQJ{7$-D&>N{&;~jpp z@1MVz|9Jj?H-uh*Hkbibxd0pVoH$HQ8Xu<*6RYvz?qcJiSpE8F>M-2>mk}Fv7=tjK zDVBA58+g65s(GL-ZG5g*-%wqAqZ3s5`vLRN9t^LVUfIFF%)|2^`=9YC7amX#N{VI0 zN+%-xnE)^<$Uk#8=N%ZoW86nyF=xY<{`PhjDdsSvvB4t0JxVdv6sbUY&}{#91_Wylef#8v_~|0O@h>DqN%w(D=qLUc;4X|@ywn{%b`&%CgiHz5L`Z5 z+S{mr@!ZsQbRc0D3%};?@dxYuMb2(H+t@Kg-{eCYl=w3mbMMsb_c8e=iTHpZ2GJ*(TAG|i^e%i-zB2F3cro2pz`S#NN!6UBxn@CsC= zvc5?DoKE*H8lRui?4h@+vFkYA6W8Cp)3<9^%x=_f;Q8uDKrADr%cq zr*PG*#132|L>e}TVR7z_@p0d`X0Nb;3x}DIHn=X>=BZbo);yBk@xsM5Hl4d_)So*L z_CVsfH*nF5?y0|=D>o-8l6#V0l0mnc0yLb>n{F=Z|u9ME^>Cjy)4TvsdGc(x(k!NCsi(oPQwx@Re&jxE+ zy)8aE$TOh)gI?V;y_(kyNutS^CO&G*yfOdU+5348|F@3CV|G(KXQ?YC9*q;Mt#4nE z$;5_HE)oyW69Lm#EZbmUKu@qnbv*NzVA$IWa0nV>f$Jyl@#l=!zGb~)dA4@X;hDq6 zD#pxqM7c&?BWQqT(K>UtUn>at6%ieB5KJeHQLtly0Q1c8WRP6?z;=lH&YYAtUpMC) zrNRHRlI&)eZ*X|W=VOa};}FHb7SKK_dKaDPbHt~bcap62`xsPYb<22b{iRau* zCqh@`QSNcX=eiF{Dy7k<)IeVP(@mA`e{ZU?-0^rRZ2sUgT=X=HF1~RZ^m`so-8NOX z9~^QM1p{y2m+4kdE(bk}#wpc;6amh(qfb}Z}(Q`6tnt*h@ zGU!PhKb+dMWIQuSkw;YYCf;Ik6}^JoRV4YP27rH{@1PrsV438#uJCooeG>%9s=f0z z3eofHfV&R+ikF6f<(GRJ{Scgz(?M`_U1SrJBfi4N|4~i$7QIop959#U*C2U%H7OHL}jqMVRysuE=hhJbeX8g7hgmC=19?RFrG;X z+5)VJLAM4vUC^Ss?swe}9qjZP%nXYn+xTZ`)cr!yslEIwk_^&J<1zGwKxO4YHzfA; zd;ZjAARHN*b|3w1t@fTZXZNL=-fY;jWNKRJdc~JYlsV|2PZEv z&pjfUcetNnP0BHOj=Yi;>yWs^xZI%c%ZI^M#^n0HT1{Dm~a>pE>4GSLt>=8 z=)PM2Dc0HY%U0h<`t43ljyQedWiS@J9(8u^e&u%Hsg%K_bS9|WBR}wxyn4Ee;b8Gi zWrYokD_#499X1kz3O}%(YB-Ax8`H{}png?;;4)Yt?`Q7J2Caf1R76h>DbswICLK^_ z^FWZ2cLV79E#O;Rrdg$E;+>s~sYn&{qMn`WV$iU{iA9RQA)JiQmiFlb%z1Imo9j!- z&Q~TcK5Vh7DyA5)7~4Zdqn-}p2yGNVnmk#B!YO*SS!U;<(Fhl-yQnpb`Ue}s_XX+_#{u%bqbSJrJ6XO21yDCq&rs(&~cDiAkOr}iMTu7@c9p{ zNtT=qFE$93OQ@TwsFdDvKCX zNHKdUl1oL$l4~*p2wLls{94qfruNAP-AeqA3R1|Sz|*pBMk*;2`j=wn+Z9CEe2HkFeod&9A>(yBLFez&%!`9g{ z-W4I%lyTLy6NCl3Rb{jB4AZ6>>&$4BamW%U&w0icGpyp!xc@!zy(;TC#92t$h%EJr z3aHGX4)G=VC+c+fl6i50wNv!!BDy@ZUa?%TT!2LwQ$bI!K0F=QS4JIvAYke>&=Mg1 z)M#v&WAyiEQ6WG&=onY@1ULsYY$5UD^VxwW?3a0UFS`V3D~-iqYzffih&$cOf{J}x zY2BTWI%P2s7SKz|06Gx`P(a0+T`Rp)lOLQPjD=X%k>bT4zvUT@g2eJJ&VR-ZB)-?r z#lLCJa31k9en2WXOxV3_k+Ynpn3EK#rJ_4{ae{ujMY(ZGjBNPhWYv0-!t_9e9bakE zp67MZcV^~!t@UXIwnn!Jd(9X=rx}#=l}|E#Ss$rTK>@ts9(4Lx{aLTFrYrO0MuoQxj>O~RCJX*Z(7l`jwz=}qM%UO z1HFp4%haUBzL;H}IneWR=<9jR{m4GwetwO#3jlPKUoOPI4@I3-?1r3X@B9kTnA;td z1;WYb!F!Zo>3AQnfWAexgYX04Oq&Ij=AG2z`M>`N&qeXmW05sdye}-`;C)dy z#B@!8_a@jMZ<(1E>g^SiZoKuxTA=>jH$7~u7!6PiaX`9J)pjfMn||qs>3uan^yxfW;O4oH@=9) zqDyipEx}HQ(IpA<_>5_&+6?wKQeBZGMtpl#slN{QR@*g|)0%0c@D=_lDBFy0Rlr+q zS4rV707bjN_^GCS>Mw@Cm3M3k`HaJ{E7oH3yq{toP^1T4UvS3~bo`N}g*H;>rd)?c zJRBxny;o~QD{NM0h4;yuJrmtmDG~)IC50dwvRk~JsR&IEH~Cq$hb8blxmKa>90xzv9fZKg#N0=@j9D_3f8~q-7^}?Y$ zEgW=f?()&H(rA%1zp`^GW7%$y0rFgbp-6wCA zXY;Pg3g;TZjoqJmb)B*ZfLX~)3F7RgIjhJv{@rOcnbwlMoL%GD5aBJG-7MNFZ=q}G zt@0a^0k;9SMCnDw!F@e|V?fV<}=78iD;I@P;p8z^h!>5U4 zidj#QwN&&@%{5iFS0V4DR}a5KG2qtC&+~7YeVg>F(gH39jr9=N_3>lA;Mk|=@Kvln ze(rV(ov15ZV$DP5Y^<=sEN=5%74?PqJbgjZMVHLKCd-|F6YL;{1Vl3F#4+}6wHvZy z3|e-OE&C3Y{?{epj4VN9>*Yv0MNpS6a!67yTR!hwi@mTHiYI#ab2<@w0^o?U8>$AX;BqV@AuN>MPwL9zDZr+ zy_sJg0YIpLx^mjQK9B9Z3=JGWN2?-G`%Z*n(atxjHSNX?^F_{Yj$;u?o>?+m92got zC;emo-X&&{@d>y=*Na+d;}i4`F;P1~km_(C z+KNEc1l*$KxH*MtMaim#`+b%v^q#P&H8uXV(rA5U+QU5kdb8-husbr=>om!jx>MTW z6q$Lag%l^p7k03&TToVHy(7h@eU7^cIk91(S12tDSSmg$E`_bnRf=W^iFl-Tdnju@lB1a5@Gl`y-yCXKFSbxH8$CtN8z- zd?DG);ap|EMXs`oVqmeeor*5uB}J}_JSW{RJV1BzS8MOc7oH(?iZjyBXMQ@fN!u22 zGyE##B1=5G!`6|9UI*qi&87GhK8OcAlAt{wYVS9_TO)c2R$1=|GDEu~ z(Los65r9Y4tum?IHABJQbUGCJu?aIj*xV3>jSHF3p;-!7$GU?+-X(d)R5+!;yL9eO z4QfQ6p3*AMASD_dYD9o`RX85Tw`c{eqeB{9n|q78UQrORR|zYr3aINt1vpuS1!V)> z?^zLQD%j%oI#PC$RU5n>0CgF}Y^3#G`N3FP?viciEy7xEJB>O@`{#DalKk!l;IEdz z(;*}1)hO4FyXbyuJY&Y5$HU%>WD~oMC=Ry*@3RPM28sc1@eJHqs8`?y62NMeB7uip zZ)Sa?{XRKfx-MCAIP!I3h0u;yD!ZdX7FQ-z_bWoX`Pd`loHk9{L1a`h&4opsqyO=! zz`C&iQuWT+DE6P8bSzTum*Qh|Z$;7`j{aX_&1WYQ)HZTl^@{D;?6WgJUNh4iJr>-^ zjw3k}fXOnvuz!$ZK%KCdie9DKFT5sO7uBTw)YA}|#gEn2DL?frhtJ#e?3`pzL_a@4 zfW`lN6E8?MdXx*h1RFhec{W0-nJ{U~r1hjraF4zvI?69r#)vzCl5kZV7mN*$c<;j? zXqOJMoK8$PM)^&&Sfe=ei)nj2Mrcyw?qcJ+S%p7VnZdJZUC;%xk;7?Eu?4ELCDuIyLu3c}4mQT{=}Hg#_??xaaY}1fao#V<7xDia0qcR9QvE zXP-I_JKLz5929Z>W#2fsB>z&0B&r-~9ISXuoefAjN07iW2TOSCJX_Q&!&d^c35hYy zVMP%gknARaXi1eZE}=3+1-*kRR|9ti4C+kEN27tlh=b)ce2jQ^#}C@Z#|ZQ0`1BdS zilNL==dtJG(`3Mn)eX2i@sh0;@hE{})=(rC2*}`r0YocOI$!tk{JQzhZKSk+*zACh z{Wo*sndSdh@jEXwApW^}+j+8K0)dX!VPwZ06a!mJ+o9XK7|0|M%QKN{=_WWz@_;{ZU%Elx1 z+}-hWSbe#;JK2Zd5&p|~GfI}N3^@k!I-K1G1E>y%?^=q9qsS^M8Z|Ulc`3wxmI*qujr6)UKY|MRtguvf8gf>8_Fmc&=*F?((AHZ zx;>)VKS^_w(VgPmfSqM5vL})b#kJ5alRCg>La&YyW7kIl?}iFhs9?n-jhBKQj_Jk2 z((|$=*qvS?P6>ik7W3O$b-G{&e-EAIhaD#d-&217+YGDbBw%af*}5&NuIYO9zQBvY zJD`Y!@3VdCLFFS^G5wjhfv(mRf{fabEFW~I&}&tux7zyd?(g9Z2Hh$`&qbucHbc4~ zA?WO+PWM=`Zs~&UpuLte17sxTeWLDH9iDzgk|RDGST0xueZp|ot5NhHU-f#=WovdF zXXy_a7IV}FibCD}W>X z4kI119e+kx@K8zH7tDRn9FvMBJ$j$ia5yHlTA<=Q#WYf+9;?ist7-#}`V1)hB3DI~ zsFy*jTB0CDlQFLcIvY_DPC+Svc6uFo5Oqx1%F=z*PGs3g*jXJO2|zZFu|rn(&p%ve&B0?snBi`=4>jQ7 z8T1od;}%tpv2&<$PLV`+gBXI7RcHdISN~sd8p2f5;CExKCm|cL9<@esUp775)Kzu; z?IN&XYo%j_|MzhYeVnYmk=*@4?HRnf|99p|=qEMbo=@&^coJG`kr*!?&kRuHp@|D~ z3IN*YvBF~)ePqFQdC%+``V)0M4>@zF2GgQirP=~nF>ZJ^Y5V6T@o$)qB;-$C&@->s;FIif%JLomp>5fqL+I0 ztJXp1_ErCzQ9ZNqz1yHVj5`g$$pV&)UY)GkB)t+^q&&*c=A{cxyXT0}>kxC~)XB+$ z2jB?z2{AxVZO`muyjl=pd%*wH*K|Bnlj)6*c4}^r7G9^Ie_pvTRldS6RfKoIui_X^ zSN6|s_ zS9_1HB|Baje1LV}uzGe8#Q-5VkBa^RH2Gpct?!TbmB;1}D4VpKeJ+M1@H)M3$@(Vg z{hA5NXZ20$<|jy-2_8ETQme)_8SXf35#zH)<29|yO;J|@yNu_Y)#4RSpTQkL%}Bw- z*$7(VF?H5_GittZ>^u3SjKeeAMGG99q8MmGI8H@lqY`RwB5_mK4d%sQh#pWFg8fX! z?xjcbmiwebuRt2K&0JFGR`?xK>s1?P4!T_q+BD;aqzx)6H^Z9LAM>z*2{)*hd)*>6 z^RLZMRw1NV z=wUTUjr(&nGy4NGoGyNQ`E1g|VK}X~NR*b1XBLkq15|XsXS^q9rcE&YTrnXTrXSEO z^6pUN%2&-@>z4?c!`a^VMJLJaS$cJ9z&$!wUMJcTzMCGH-K1R#iBbmt032Ij2DCh)qvD?FRD={&S6H1Nwr6_O0i zv7i-#7T`HwQ0NxIPEwvf?~s$PrxtEiK&@hC@`@{@@kY4DcPUD``<5n zAsgAih2B6zua4LWX*e#|Q+dm0Eb|AMA|{?m4?h@y>&`f@9C3#N$nDLdE1`PE;M)w9 z=LFsk&rEtS@EqN$)T{NL=L3z){c_>9DDb8OgJty3NLFEg*49anq{-p%HZ|@KNf*6d zn6DXd%ht34V;&Bc3z5K%sy6%Q_Q@-!RY*_}u91iL!So%8=f)G$dAU5G@jQ8VAjGTi zD#|D z2f<(0|3@(SoLwfyVfOlJi;($gn&vox0T1FF(Em?oz3vd0=3gq zAV!0@g9cdqp;zM#8=)5&HiOXz2H(D!en<-F8*s0~LAksm%2YY%SBh6f=}t*DNn^w@ zVswJe3P-sn)2$BqKa#<(Q$QaU?9^NWmRdP1H(Qku<)P=|N+`YsBKG#ID!DFMg}>%% zC5BOxp3f~FXTjQ9clLy}eg&JACWql?L+t_^Zq|6}AdwRQB5*4YJ4-7hH|cb7Ezr}F zKsecvoXmlA$#@B?T+Y!qap1Z#_rz~mbA{L-gS!PfTa+hLPt_3+_QU=O-6m-t$Qu}V zwRAmQE*yHKSEo!L3@e{%lJ~zPKQ|?R@)!Vy6-u6@@UGvm`^VhQpIaB2IJ*yPqja=k z;uRsv*H+FMbj$V47n?YxgKpWdi}vy4?aUgV9O)%_yE4%o{lUg6Hn1LX2fpKuT*7LN zH~xdZh&QYqo;iyT8=lQ(=S}SJTn@wv-9BX}y~7hFk4xmJauZx3Nffk-DyI#&p_y}9 z07i(h019h}m;!?%qt5CYb`Y6)GkAryQyH6yhQqb$LW>&IPKtruK|K|X{kS(&t)QIU zt*VY%2kGt=$!+B(X#&4H?6GQz=LwBoljUBZY*fSwPDIrymj=W$3wt7u@DW}MT^xWM z1QkGw6-HQ}+*kMYOlu=>&PEa&4&|m$0X7cO<-u|MLxJUQ=pGVW1z~rK)0-Gx@DbZT z&@Z@=lV4^RJ=V_qNET&YKW{Q9)5(4g7ZHKsGi-0-IK>>HNEH=*N?jCjO1&{shgIla zNs=^<*)2;5O_f85Iz4<9ELiacJe0so<)NSeN^t3r-7nxY` zNdjwKQym%#CR@V0mFvVhlh_va?&)DkUFfa51BwHuB+G+pNRpxmmd*}F7W;uTEHP#E z&@qIu0@Aq1itm5hdK%(ve6cyPRM0r77+a$@gf@rm4$PGoL61hCDrW{LNT7a1!vDFK|-K|uHdJK#|m?)W?{3dnJA9K#IfPG_bOH}86S7~_D*ZH3C_aAhJJ)eOd6Fy zj>*@1CaX{-qykJ9pfHS#cs*}lVD|J&{-`*8NuDj!-B+&lx#_*u=RPz5rNdIF-^1Zp zVb`EIjuh8sA*_%x?&!q91AgY2DR=O@`$_ss*7I9oF zj`0Qizqb38&TaNboslJCeBJ>SlSd#WYv6VBku$g*6wc3uwy4Ud8t7V{0rc@vZReAa zH4E1AvEAIkndkXv8WoJp^V9R+!Le6u!|38xdb7w3nFZ%J9U-eDLYP4btz?3f6ZyZ@L5`BMl&1*o<@!R~(!4 zC*eQ8?4`D~s|1f4uxc!=_`^B&oP@VaBSGb4~NInfeUTNkJWdYfAg|))b~fVE82mh zzFB@o(-3-R@+aylIkxZuISO)1)2k>(kV#)A$*Rx6gn8wJ&sFV`SRhfM6Bv>Ru%ok4 zyMv?)&U>HLBDa0^wL|^Mc$3tn{PJF?+C#2 zQP8WGUP5-zjoQ2Ke(=AhqZi+(RW2c!yhiQTz`gT7xC{<;cycCPIu}h>?*X@KJ4o6z zf$i80vRPh3_tHD)F5sj#i;$6+p*f_fiL9Af5eh{4+2z8WZ`@Vfp4F}`6Xgl>n3h*W zw~-p0-T%SME-%~SyGOEwX_J5w2Jnl|`<(*Bl+ErVIeuAyuS{{fwrx(PcnMP{xed77 z=52ahC2rI%71u}9iF$m_YF!$t*o3AhX^jilK0{H=c4!(QwV}p|{?sEh&treoo%=0` zWtZA;n2V5QF~e=9m}H8qr=mgcJSfY!G&bhMEvh6xsLo)*)92AUt8}i(Yr&<^85%0n zPj*1Eyw;V^5`GYNX@(gZZcebTAIkm1(wTh}19jZJRCIbcN(bm-#Fdj*OgS6c z0WJNy{XU?qCeGt`L>L7Ln&}=>Q z^9E31j#Cdr#jCObWi!Y+BBK(}W^zt&Wds=&P@D)1>szq{O3crg*FDSuru3cn8pxWL zYz>ud5p=dt%w`HkCc2iSkXmv6rz39ctNF2^TY)DF%3$b-Jkt_qY|9&6rY(fhDSm9v!P}+rw z;TMe2W80Wmc>R~xtg9|;2u&PDi;a3rT0n~$|A#H+7S&aM?4G*-%zm_w?IZc(nt-td zj$M#3B_d-47Sr4O%e$Zd)>=e_vwO!jWI?djs#oL4LdVZ>_g2{cYgFryR4}$<44&M!C%3LCjZx3URZWdIN;2fiQLvve^bd7f#o8p)A1NT6w} zo}UY?u}HGM==^r+k;m6wbUa;zwA7zD(aczq}{eyLW!C7?J$R(2eb_Z4kUGmpq zr5^WpO}^-3&3D9aCKOl5&wlx9-S6K2#gBjcU$PYxvy>t+Pa*IZm=q9StS~Wtyyu^0 zyk%aIWuWUV8HEG z#CoWaA1A0s79&0cLNK>v<)H&^2P4wq+#Y&85bRq;P1?17ZF9QljZ^k{W&Kk}z|xRE z>S~}U4^c?;82Y1;go0-@i_S>*`CON#P0s{M3O??Yq%&2Z2!isAyJcJap#4>5kZ1c} zX3{+zn8NIYlr_XyK4%|i+6PM3)FGl4pQg|%c=xv)-D!e1Ae?Z0K-#wlkNy&xjg3Ov|ekT%8=SuI#2 zC9A5Gw|E;p>*zZ7E%OH5>J=q{T}Fk6dPSVDTu?*fDhM@&u?r|i3Z47#X4xRaiKK;0 z%{K2P?j>_g1vf+RBahD6n;5;OgTDMd_T{&+0Tp17C9YV3YJ6X4=Ihq$pqI*S$;N6Y zHfR+nAf#$fGW&hNu_qiTEXbimafb@Xf7lN<%Tcg8odsB|u5|5F+hF-!<=_7B_vUEx zt(#{E>E>{>S!dD3zH~ejGoC!AqI2b4v#{WLGi)#{QP3qwVJ^P$NLJ0e8r3Q)pL#6l z47mr>&;@W29eOQUwRyoKkac0Sz-{t@610Bd>!2rn{Q1$mXrkwrd{AdLwJ-4E}j#m!x1mzDLh& z6#B&7?`F?>{%Q9jkT=$A{`5%J4sU;5c9B^;$@E6}3*xQtHaAoT&499!q%fbpjW@aP zVLFaS3f{ycRie{mH+P%aZ6j`^O~vNQ9U|*ZMb4(yHns^HKmfK$yL>{)yru4)!d3HD zOh7>xtiR+&CPfzT^?r472a+1c0|Sd-#CGy={Fj0&*+HZI@sU4Uue#VQ>N#A&v#~~N zkJuKt%;PqAZ|f@f_rr3)Wa;)S$Hb{dHBR=f(MI_u4qWa%{y%+2qlduZHdPxIFy#^5 zB+m>9eHiL-#(zYWJP30G>aZ6V|iz!tGs$u*e!BPfxd^`B-X#$*YzQX zO`I9F)^cFn*XJOfRtN0mun`8QT@X94q}sNr&m;~S(^9{Q2Bk~n8`Q5zEHJNR3~SpV0d z{6#h02RtR*A1aDyr*&f+%2ZI<#@J$^717XKyxPh$OUbm zq|GBu)CrFV9i>-rbzTFG`?S|Ku zDRhT4DX_%|+HO1PGSOmj16{%gr%|s~*q}CULS{@oEz1$-Qn{2%VdNl!jPuP4VK9ml z*1y}AAT#4c`O#aAWGy?Ka5uXGeedDC+BAv*d(adrx(K>KZqF+9H*QTMT(Iz#HgTCk z?>XpZdVWb>?^EvsWXQCMjp|tO*yJG&A!HOIZ||kvGy3OC%?Qa5ZM;DeMxr`$c)rQC zz=57(woqg<6@5Ze5Y?{fnANU%`lv|@9u@RrV8Y_^wuNq-vn}xUtYT#%InQ%xcz8M( z>|yZqT30@@J=I*f`FA^B_L3<&i0jpd0t@&cb9~3MENBIoO=rtGK(<)dqIyVwrrEC; zn=oJ=Lv}E5<-l9s+cpeFsHTtq+Sy=rK%%A{9%t@KIFP8vqajf4LFZ<+zs}ZwYC`#iTe7y)5PFyN2WdyZhqNSWKX}|3HAzQ^u3ww%I)$NRWF?lR%koOV)4D118&QG@b7I?@$YM* zdo#<0r5^WYj!7cQ4kD&dFyfdx`5-HZP>JqeFCNWQ!{Jqpt*J(qHuX4IzHDL&n5@8M zf;)mW!ZvLaFC*{~m?2d1>IB;-SNkSR+WpnVOpZVJbO5Pxxayy&t7!HHQMsYq{yp-2Ft7|$sW$cz60*9(@ zeql(nXMschQ^P(=2Oipq3J2~yx=S8b%sOJtS7d_<4i{2wFd0$Qqz82EFG+MqNsbu$ zHtH4CywiaVfyw;Xd1zlUCNt2`u-OR=V}uyGW-NYQ|FfX&uL~I2>N~S_Hz&2I?|3DL zv;kXMSKxg`IOtaG+oHOqIPN~iw#8fpM#=A2Cdu15+lKdSpneXi%9T$nQymZmx7*j_(>+rk+&{ zC_BQch>%rM-Vnpc$!L#^vJo9 zgI%`j`*)Ln8Egj5A3tm86 zs z0fw~{6GxF%sNWYe?=Z=h8R+d=-Iwy^^A2j99474h!0j1h`|jl2vseX_afc6)?Ef?) zV*KOpO((4!o`sfLv{T%tm@bNZfu?}*OtPv=aWLSJ23;a`E$R$U-G__CcNK{OUA-3= z?9i7~YyG;sb$uS$>VtGoWWQIZNcWI`j@ljaY^8S!zvV7%Yyo^3(F`boA zOfdy!g3))pP!$8P)2Qwvt@1>WqwV)tD+Cjk%4zq*cF~>mLtX_~#-p7A7z?5leu6Y# zoG)%vT;$_VwCF(9E>L!p?&T**OXp?+!6{8zDKSPQRM&7OX^92SMoOHmkG%jr_P@OJ zyN@?mm*t-2J8~EwHi~a1bHQeQBB>C?GkWzMA4I|xS<_f)VGJ7rJI8+56=U=Hqjh!oxNyM&O`OO_#S%8%>u5ha-!PE(;{sqqbv=xPq4jJSLMa6Sfas zf-T05h6~dnlqEf%*ji)1{2Ul$d1bjZ;&wmV;|4w zfr)`dGarz>{F1Q{8g?vhj4;|xKJ-K1e|p!tn#Kkq98O+s)ZR>G6Rhl8W6F1FnCWu-yZM$C!}o z5k=E;q}jYe?+cO^*mB-4d_ZdGl@oT-|5GNqMNB8ePH58pm*M?$3pRi4zaEf%{-{MB z&y>oJ`<@rz1un9RWHbIFB&x5RITPwP_aJS9Yt8 zgUM4$5Xcya@Vt$|aZb@ecWfTA1wQj}Ctz?exhI;YGBvDl=?eyNO z%-}x+-O+VX528>ZbeSSwICQ*ORG=)H=!%u{$V0{w9V6ew1xG)r`S};b9AFCF(mw%D zHHh%TyKgs9OcF)b;U<5wDn|^0Qe|qi03LMPBJGK6;T;ar9f)iUHBev7PgpQ^A1KQR z?}Crc)-72@cS zD*VR4q4p_uxEdI2qd3`<-!Ynn4R;|=Y%FY+hz)eBs9ggK8|0Db)f*;uNTA#Ju&42_ zu@ydN1DK~UbN!<$o`adcHP0kc`M*VxPIjp(hhdRm!EuTi&pf6`KNXEspe{u}-#{Cn z7b8KsVd52`UX4d^!=_49L)R-VGTo~EK6{mk0(@1s#`74Mf#RNVvZ{i<>R+nf1F}l^ zr(OZO%7GiAP%zA}N{sJ>3$S{{QP_=aRbBuy2h>Z@orHsE)Q#d@ z6*Mpx-m7{q?;ZU0Hb7J~{fb!rX$946&eENJ`-jhUzc54a&Aek%$Y&$zU2qs^u@*q< zr7k;LE>{NwabE^w7ln@w$S%1jNZW!<262zW3V1p^1UBntwi>8d<2a`a7l-2u zh$@uO*Q?utjc49fq{`2c^l%*WcDjkz$+QIl5B@^3MwK607^1^ttF?H{ zKtDMMTA~t^Ovj(e6Lo2Hvl=8ugB#V|{A88sZj11JkNYMFu9KA_d>g!dvuKI9AS#PK zq1vI+trRW;X3ZMU94WpLf)oE~IpRK|<1EB%PmAmC^o;-OnikRiW2dH>-$iuT9F5S^ z-43tk;rihJ>V4)J_m>e#bqaG#I`MaZf0e9%$zoEz#cFUn#em}QRx0}AIpyk)Nf%PS zryf*7l(-~M@+*f4=woFMA74a|Yec?UIl5jYp+h>?c5MdTrIIg;(z zksMeO^0F6&mrv*j+0NWoZu43(VVl=F=|2B9ZLzG=8>gL{VKsC!xW5c2W5fqS(gL#h z6}*q#J5=pqY0|wB=O%SR`8Zj1#;4B%H=11s5J!>WylULY7_*Ixk&43GI<-;OH5)b+ zJr3_I*(g0FLz8V@}Xi$h^+U3G6^I8A}MZl0U(3$wU0f=zP zDoi+bLWbK4E{0{ZL2ndRJa{qw*X3R5U9WI1^Pk=D51(R2cs3PPm)yklYR16+WQcIw^kUkl(s4L3(;h z2LI`$)zA-nNge}4san}FfeUh7XCcgq;4;U<&Cnp)2a2Jy$L{BMgWA4j4GJ5dW-sZ4DgaF6_+7da zvL$>G9x>7p$NB@c@6&{UJ$4_Y@ts%&aqgya=MsJzG~2xH{f|%Q}Z3ufnrQ zyH2_+#PqzuGt)Tlob@~Jhn(6|Q8&H!KoYa?Dh~^=u)UxximV8|_(qd9CghPSX<9Q~ zB`Wkr8!|8q3vt4?!*0Njkfhi?H2QCcZNP_xeD`;A%^32yUzV&|ek-p$`fDyNzGd8TB@;F}{p9JCGi-S>ox z94B|X24lm}=OEn9EQn6G?i6`N@)jF?CD?zq(PIz@cNktCd0nAfW)FfuA$G8LN8TV; z=zF3Rl1i!t7qs1x)q?!MMn$Q|8ow0qaYzX)P%RO(D67Ob#QC5|2ajhvApngucB2Q( zNOv9mu~moePqnt{uwkpU-v_%u8|ZW>h9mPeO`7f1C)cHVqIf9UuIVzSHH#Y6&`*5R zyWjJAcoh`N`#s=kj`(qyi#}KLCoqbScl4j(bM9a}=8sVD;%*rK-22P_vi23^Y`w*X zQ(zi>j&ylrBM=}GEJJ)-h&bJ7y~tFrs#laUf% z4sSx=u}Dp>P|QV&oTs9*H4T9n#M=1XszK>hm0>0-8o)kt1&FCMs&DZyH>q<^B)Th- zW>I@sy`o!r#^;gzG+9&!YdpOgWpbf}7k(o&Z&DHXG*k(?WvIb|o?+#l=L0|Z;Dbg% ziQM!aO9PCzNmez&wsRI;NsQaWipAn=US$v#UJ$fZ%D71hf&`wfb!s2M?dV3IQn7Bi z&s7=z3I_V(8_6%Cc-n`#VF`%S9zGwswlR0+|Ficca7|@d`abc6E&Z_P+$_(C8#~PS2ZN;j+Pfy-(h3`_4kc|5TPQeE6ewJ+}P9 zN0^ZuciE~tvf+>BdqWO8BU#w(!6GvBopx{__X&lda0P?)&RDT(SP4xWUIBD$AZu5Q znsv?!vXj5}){Nf^jPug_rBnYz)=eRKCZ>KH#cV}nN7So7Y*2T4)bcaw>*6TVMc=tK`0r4)Ia`c(~fhMPHMnP|+O`f*j39C_XT*Kk^UB5TK zYJ|t{8j}8%9AOu0abSn0*#zYmD5jPoXQ&9w#XPHeEIvHzDljRm1qrcKI**Hrmw)Sl ztwhM+lmI&=y%Ic&INr{XAs}2&x-o3OMn7y<6v>UK+8iHnnQU_4YDXad8W!=*rkD(hq)`!idat1Bg&m;nvYpvF zs|_N+6)x$n_Y_sKR(S!$-gNY`IhTd#$yRev8y9t`Iw1L1t61~es2(WWS8CpV>_czf zlWosg*~Lj`fBMj4i4iEJ-e+Z`&4Gck!X!=i0mbxEq>GA3;~nHL0~%bt5+Of|+29?| z917S6Gv-R(IaM`XsfgvQhv5G~KH8U_&{U8pTQPeL!S-53`~2?tnmkdH1UZ0G>1?GY z$*+%I=9egs()W@C`veb2vHTQJRMtvwy>=XoRfS6}{~+X5p~z4N^nus4ssq08Xb$)D zNACugstTulvan3PO^Qu%Eq2ax$f-vj%Q>^?y7;d*<)xf;E?Sl0le7xl0taeM%&5p0 z6memNg&}5x5YFe;Y3#@wqxBi>F(ZK4A;c|X5O7L)iXq{ze)q*IKRBFW$v zDdOcz1=oT5QzzRP1SxA%KmV_VkGxF>=YJ|QT*C!$0SBV?KCF%T7`}&N8j3M|4%ES+rvm$Padokyf>a6&t zM-lg9eLvVvt0?9m*L)+VQGJEv%t`RM=$s*L<~D*t(n-z{-g3@Wml&q()$vtYj|t?X z*^H<6*&=&bdv+k-b82If+K45ca(WUu;J{eAY=WiJ6mxmx&)9O7dJX1wC z#07M!=x9IzokOpcAC)CQ8B&Kdi$)Lkh-V_VTb!sw9+0cz47XgxV1QO}L{KP2< zboyS|b|!~Lf@a`7^E&GFIDD5&B{#t*!Dp3nl``6;+WUzoh6vk2*1AD1AG+08NpfIO zC440Ai^4}?8sJz=tX?yZoMCfGYbF_192 zhl?V5CIf2=#z>|L)z#uL2H<8FdW(g zYtMbi7%rU1`{?k?9xXiKc?>oaPC#1O7D<+0tGtwMlbxqmhYdPC@G6)dDQKHuDFPN_ zJ%$k*`y;IE3Y8UAQ)^Ce=CE;O3zTICpzsI-JQ&>36v}ooQG)KUjX|)@xyQW-X0Th_ z6c7F0e0*pnz0b7~THKV`HdB1$;bH-irxOYO9JwxI8PTy zDuqQ*3EQa1aKktqw%y@|NmxVvX`&9ggn2}x0s-Vu;jK(4u+5=+xKO5$C^tNNUU&+8 z;#7LQ&yZ6K-4l#yTlqk$(i4h|dZ{7+u`nM#b>09pn9YoP93 zk|;0sKr$Vub3xU zrlih_23+ap?VPg``1X<{D`8o5U)E2G!f(&Zp&!lDDi$B*#0hr1jR_lzvs9@x5{%c% zDus9bVtEj-QsRvikL%*HS9Oy0@?JX0meifBfHtbJu^s#p7Pjan8D}t6M%kap6DM}q zb6laWWRv+WhGN!GWEB-*P_EG6?)pjJ%fhj&n{Dv=N5_K=K4c7C94q_urg^PCn=D}m zUQAo4;;;7J8c@tx1KFy6sdNgYm?H`FiU7T&jR7k|PUu2|Q_>O~N8)%zvYjEC!bKQ; zu`_r^2Z<$1F@}+6@raA!hrTnO8jYCMhe(P8yS)cY%)m~HfsBtFEMmbOs3b`bB!~6R zj}&MsT&@VO2vccYNIf{DgR*_jP#w@s4yuNnFtWGZr+h&Rjcq^m8VuL7eHc#j1ok0= z*N?~85%4)OSPZPyP1(&6C(i#;W?p9Nuq$Z`Dgry`I)14)rfo(-b@my>7C((4hci)r z;GM{DNY&gQc7?Rc_k9gcBEwI}a%e2khU8ErhRIYlNm4}ssKYacoG>c)*d1!Lm73O& zowLF9wW%Gk1!kV2$7(=W(U28pCfzvFIQbuIBD&$Ln*l0oZk57HVU`M2PP7WtxY2aO z(m79rSvkh_xzU9glwT8QFd6Xk{b(zOVJuC?4ymzxn)T1z`UPjQk8vU@7L3r;i1;g~a%p+s-B*rPruOZ7-kb7+TwfAK(A3yT-`V*9)7VDZPI znW|Temc;kJD;AS2BbBT;@akr-iFMgQF_2ll9ap;pt|(B_!jA<5Qo!w|4+^ybX~4dz z1ARWwS5xULN6si>McuGwp^np_-@w+YK3Ql-P+1F~;Wp&+W81^#X^)oS zGx_~rM!6WF@)ym;Kal8`3@X_sP}xE;i4@sDMc|x*v3L}I)ogRlcHT;W6Lqz(rkgY| zkAjhr6{aM)9kKBA-5YZD*ETiwo+s+#=p@*>Z(jKV%p5)?&sos@>=Vb1G?vKu+@4r>G-ha7M$SkT zRspdgC(RRE9M)B_kgBrDtA>kWjk%mnzG-4Ss}wdc$LSM*Cnycz9h`rhzUb8rlxCR= zOGVj%yCEQ->DmbRsNpsM9rLI22jqiJDS?~3%IQ4cW_d3VgfyzNor|F?r#kF2U?Evb zLH2GX)aQ&HE@tCtYxJ0PKju?*=;_>Y$;Et{ViU=A+*DLzVFGIkO~j?5SsyIw2@>Bq|y4b)&0Givhz+)m(34<^C5{G$%Ap= zawf<{AGVZ^qnH?qta%#K#RNE zcCN9YNsoj6U`BKJLElbE67XGC%mB8L-uYNRJ7Hh>r=w*R0>ImN+J2&bT1LPK9?0bX zxV0zNJn8>=0UQUODJ&EM6mbWDUaU{}SbU#*LiYZ?H?F*S>{sny-}95-Z=Rh>?*`_C zHPj!|C`}r*j(Wn24=aPxpJ>v+KPnl3^4~!ygM9zbnm_*@DA0h=tYXFu@d1CW>Y%Dh zaz3alu#5Nph1XkudGO5@-)LX-d#$3-qHSMiSq6mhQ#i@5lab^MjCNJ|t=a$do1 zS>w}Md|BtvT5iP_mzMFh{ZH=+xyDJy$?Z7C5BFSOXQ6j}`pvZbE;ua*S zq(PzDafsZX^UUQ|1Oe%{@I)B!egJC+>mg@56lU$MMvU2W_CEc((CM0aS{oZb)q!Jj z7RViP=(FN3UZFtK;G8bn?o&x0^@5XjX)fnp$YWQ`#~eHR+*n4-YGk097|REY@tMVI z_AC<_Ek-(jQ!|Nm;N6`96C0qV7)U?bLPeaPo+jR=I4g(^N>Y`J@;%eU_3AWnjb{sg zGiQtHw6IQ*?^ztSTYM;Blj{>+gL;Cj0;?g{&kkc8Lza(Oo<2r!PpDPr^Ha?k>KxW- zwLqPR*#RILQz`_xU_q(~6HYWYJZd3xaoOC>Os-n1coKX+IKgNCD_1zB!pr0c_l9o| z7ya5ifxXu@>{0U6kzhIOESs}J$>i_!&7K=%G$X4MyLXYDUl_;Zw29HFpqTv>DWf7z z&{eW0;!IbaPHUil70y!sR?Km%CL_;EJ_E+njFzMeoylK+~Y^m0a>KhpgIM zPO5*5TbgqrXQ=>tf`G2Ehce$zj~FJ7_q3l1mtrQ3v6aahJKCP2&g>^kR9R{s=DeYD zJhU(l?4Veng-PR;FDM8d;399!Ue9_3uus)GchHI43Pl#=DWucMf;+ENxF?5ePK8}0 z$;t|sR>-8rCk!fiSTR!J(x^bKibr&=c$);W6e0yESKS@v3VjD|GVQf{qta`+%1Z(m@dV6{|}rHU0*1*}pwNihD)3UD?@l^Mu}pISoDNY{U> z(S5^w?y@jN4_(2#uW!k8&sv-+suZsGIR@W>Ty+|69e0Iax44DhA-wh4nXo>RP1gn0D;5Kjk^sKAf-xp=U^29S#m5NU@wH)doSpHRNbO4t-S!#Ri7@ zZF}O31*(QS%7$Y*B5O_GrWkX-$zhF$1y86>8YRFH1bUyJMIB#9Yl^sd z$;dWL;{&_BQ#{Vo@nN_P)~a|Y(aq2*t^{nJJ>+DlkTY}&ck;_2Jr`hX?>FH6x8RC6Xh-Z-fuJm zxIonjs2g#2s{qK+K6_)+oVeHXLoSEtB(>muTnVVwW5J$=haDxgneDNo;ep?a&E^%Q z_+Vy<`RS-8K6+~fruFa`x2eeO-rfNRAJ4D6r=w%zNKOb3+Bq-6S}ea~1S zWAfvsroWj_VHRczeEIc?+90hD)PaCuZx^Sz0FUb41(iW+%ATY53UOLdSbHoa;er$gI%Z0M246Yk36Vn9B{+4 zA*VaMjX_!BOw&alJk!ZyIY|P;vDT+nvWKkX<$7F}Ckg7+E2iW?B&t-FB3U`*_l@cs z;$)xUmyjp~yBPkMP3Al{2Fz?~vv_I?)5gkxQ6Z(X8_W}aEM&^0igfbh1Zgr2q(C4G zhUL-OU{Wf9j@~9O)%q7;{p+{BPg}HP_4*xm*Xe)4yf#cVT9)=-U;QgtJp~9EhG$T1 zr5NymHc}Bdp`9k1L_@yIdENBAH?E6wI5n!PUdcY0zE@t|1H`>a(+>JgBs{Fg27D83 zG3ZwI!3rCW8v|y!;BTujnmHaz1aTBk9hNcVV)KC1%Zf#9J2g?F4whk z!@u?F26dylRCt z9^lpN4vP~dPPO+MoIM(zxvC=#qRp;Zo){a1XCus+AfCsGp)`P9YEx9hTs~iDEWVB#w%xlXeJGMg5#Y;T`uwQ;R&{`&i@*dNM$Uy8>j7%l~PD zEV1!by?pO_grr`_{k~=V+iP!|Q-nBdsCKUjVxQ50WWkC$0rZzzfKqC}^(62(>v;|{ zRUpxwC#r!^9omkSz|5ROYgH?KFs_JvL&)CTB+HvkLABDN@FhK59ep^UM|L4V1GXCun@Nh>#CUyH z48!NP+ITb(Bf)MM9QIoneeieA<~90kW)%ls@>;0qzu>;|wdz@@OQQK(ihqG}U*H}R z!$4-!=`VE$*wG6a36$1@d!(oBckSItg^SE{bRE_$wUBlDIk#y)fi2@Rf#`)+(gUvR z{cwPlj2r$C0*uQ16$8IE*WAdms={R4EoIf9FXh|xc11FOY7*~ z&|WBW>YLZZK=DW=jX4Q*(l%pJ1nPI6CKW)%izIkSeuax7!_i4u4Ems=L8mGj|4nx- z6G7G{q|eszyJvMeZ)H$53k`3ExRuk%KL!crt#YK~Tsi+n$XVx4KsN{FOf{)`9=&?$ z;Q;(SVk(>&e&d)VGBiXRPm+<)@KS2i_#EGnq<{ zP)rp?Dhve&C_Hn=y;a_(LM1d!8fZZr0zsLaIo;Ad3oC;Xd{8X*5YZ~*MQI!;x=G|# zbATpsuVB#p#a<+%Hr1cV!h zMG*5S26ntNsR)!)-|v@2w?G+Cr^gASo^_*o1Hmf^+>h?4P!r*q@fiexztz)1lx0dbZJcIEhqcBx?$&K@a~v?xC0x3M9`Xjs_II zd0niNMXPkOQgt@>1izcUF|$_KF)#U_3&kjG-X#B6ku+@!L7EoB9iUZ)%IB5yFL*Sn zk)$lm3(Qb&@mS7%_aacQI zAyD*8{Ij0ZElrXP%1*-q8p*$bDi7-G2c4Eq*MfFzMcB@eCE4=b&i@xE`pqAH^81LN z{q`pjfBM;fzxUe+G0iv;iWEv zw1j(24u#A7466IBDp`sD*=}-4C;?Z7TpV}n0frQ z(@cGcRuM(AxUoR}@PNeG!AG%?{j;(4dr=8U@q@WD&6!sn)}&ZqS=G@^$XJB(6L;KE z*1k@vo#0|t!-!cm{5Bul(K}={{Y~z85d5F>jNZxn73ykIFw)Y-fn&F)OkAmh6jM%- zy;KC2swVmEcq1|#3kKt6uYrh2DqZFpJv)c4=d{9h@_=ip@Vx5^&vbGWVrM9_mZ@r$ zZ*;%zqmS8Pu?kkBU=ek`0Lw@YLe;tk>V>Y< zb2mrFixn*_Ln`URli&HC5kBv=y?2XzJcWE}G9p(f<}yVtQW5Ra{mLvs-#ksTw1uA~ zh~;!hKc15XML;?9T2i2dS6Kse12tr;ABaBBs}Wbv0>{9`5P#RpG9&p6lIxa8vai8SF;>86-_6#0ybh@PUG(N7xHQA~xbhE($7c>t>w~M6LF?Q<7M5}` zNDpb$+}p2WEDfJ;RJR2ijvLi`7V0EDz&B)g<|N63kH%}Q@HKvteeUvK$3zMG{GtS#zf>BiL7;0Dho^Q4 z*8%6^k|l}qwqPv1-Qraeh;(^v!S%2Wx!otvC7Oz+>}RpYiQ(t4XR8aWFgz(B>!VXD zqkTFR@wbhndYfT)u!xRIev$a+d#Q3 zTX`M?j4ulZ+%ANyc`Z>M4QN3TS{+nR*DKD0swlFn)k7ZmWnnd^7S`uYkOf)=B06cD zTvf9i`5v_jtotqv#ZNl)8uE0UrXP*Jq-31HvS=K~3oy&T`fc<6@4W2MzXOUx9_xL2 zxar*U^q_oQV59nD#Wv=E=p5%fJs>~k-l#@x*#^26i0-N;tjv}*7REcsRz7bT3nRGf zRDfXX3Cd_>p13_ZO9q`rXy(A{_H8CcCYEB>QY0Eu3&99v(U_r|;L`;OEcTmhUN|hH zgXD#e+wG=hki1>kb=X|7+F@gRyG#auJH@0@WE&NcN*BrEcsoOIuRurdol_cmct#A= z6KOVirMOoCDb#sk26w<6m}haFAIIwvV=Y=Gzmxvp13RyzMhAq&Ny4s$*p>k?k~whh z@b-V(Xf;0hkJ3k^-hnfwx=ldRN-@n8xkg1K`($uy`MYInNf8g%Zi7x~yvN}kUbpA% zmfd>olCa~|Es`}rJbnmv1x^!P2$n44Ow&iMi;EWBlou?zFFQ@&o`=_qxjjHliAN_u zdK8IY6XoUNJ93B*f@{($&zXY^GqIfGG=73s(FN&%J=_xIfcFlle|>s&aQ5a`Px@XK z_5dFZ@^oZ7-=9~--$JS&s@OSgLy%U{JFA*=!Fz+-sMc&eFtZH6(eI3Auvo#2O8$@b z>R=FkwOQC*~ca@tE89%6y(c9EG`$Pi%xTnd1GPxV$9;iB&ihDSygNhNJnz= zd@F?=!geXH#v9d-IQaq~26jys<$UY97=y%G#qC#HIV+~51?-^vLJNfzyh@;=J2dlt z@X9GzMqLG2p?w_O$InthoPaD>jtY*^1_|%m13=b%?r3e;0qD5P#l7Znzb{CxxdpCP zq@}`Y*#sX=iQmTx($5K%>O~u+lPPJzNkT+9g-V@lvJ#7 zW-U%epM753f-Sb#nHu`%F#`(YT%|t!lY=CE3W0dXu)!;)7+_Dy$9l6Iy3-j80D9y{ zIUT{8COMu)3OebI;320@y2Y&{SSM)#o`lD)o0NO}LAe53V}jN;uVW;NY5dYMzoP;9 zzv*c?b7LvMP34;7RrNSwL^ zWfAZ)?#yd22N`#6v2C4nz%5JF0gg0ovSXeYZqvtd?(mMe*MUG?KYX1~s2y3^AU0_6 zdt)Q`zt}9j;I5^w1v~gJ_4taHITeJ%&UY3#Dqy4YwIziieWCiK9o(s}4(khjAX(7sgd2IN zFc`AOe_tT#Q5pCof%HF(QyO~FYu79PHvllG>;}MLeUAV96Pf!h;|%j*?*7~41Us=T z2lhE`naoFx6w^SF^BCN%l5FPSR#BRB31>SQfOID04A z*W}O#LD&e{$Xg&LGw6h*A9z$63XVLewLYRzaYKW#6kLsK^qI%jebLyMDAPG>%x6{^ zweuhUt;d{%*e$IrTvv+6aD)M)HOK2mSCyT-~WPX0>hpMT6Z+Md@hZ9YQQd|_VM}*YUS7Su{-lpp@kb zjAW8sbCzVO8X5FS+Je#TD3$d~Rs@tnmbRT?V?9jF-3Rl5we2}On5g~4-!SJvgqJ^T zG&WL997STNh)rKZvBDO9iAR#;5bq#ARoonYU%V`E^Y}B`S-$}*Uq0)8$NK=Qm+!ds zLE$}b-Uu-o525p-46@6CjmKFNxK>e21x5B7=%5mnx96>gbxXB(VMr2FE3I(pl&_cH z^{azeSpf|z^_j<9lcsG~b<ZjUli7O)rQiJBpQyLbL^`l6R^ zdZQZiDz%D4mm*mrw^oVd-xp`~haMNqUi zV_^xrg13T)wLY2z*;R3i;+_8DH$$&aqigwm60gSh`aon zdJZynV-!fMz%&~4Xf;Oz)=UN62xyz8! z%Aho55BDO0o_4rlHr-Nzrkzs_Sp#R>j?>pTm|_~&E5K(DXuS}3 z`kSMJj}^2gFFtl{hTLeKHpd5CCYz>^-6lpLn_?hpnnpzwg`bii@Q17rzg|g!>mac7 zbbAz$ZT`pUm>nVP)%jc`8VbXTI^lf!CwH*}N!qLKXX1 zWn|j|h*tN(4zZyRFVvg!#IV^xcHCq>Vu3*>mHt@K!pA}*48&>ToSTF-WD7T41)14+ z55lPsSs8L_CRwT!kLA*SS52?*DZANT&jeBV0+_2Xvkn=Xfpym;8FezCLDa#E%+I{@rU-s?h7Ji!mWX8D-(({rkx>3DS z-TlgDCShhl=zdNw2#=fsTHS15V$>?`%WAmDzWtf(Fdq}M*K#5Qd%0~Mczv@dHoPZv ztKBH8#*F~WgE+1`?99!Z8{)<0q?r!u3tM3Gtdk~jQt2egNkN{-AS}FIo-zZoS|Pa> z%YJ&~=onuNYDHTDTLF+2E-K<0NQ|_KwQfgMEwE{`)~!i$j10;&CrPnKqP$1HYkn*KS^4`orFj)Bn`z2%Ju_mVX0|V^lk5?1eWh96!+iiGHvPfP zke=NWoG9P*_55!&FRK3Lwjb83FMJh+J6q_hAaMU)B-KvdOSAuG`&c1m4gp)A7fy}* zvA=OOG$-}%@g(~TvpZB{vZ&ijF{Km$2P#)mKpyZk*%|UI2c}MXa8WzSg<|_Q2n}n> z;iQUQN%BOvv>S420=l8XkkwL6;ym0i)+$h~B2kV)>D#8oyb4bkY!bcz@%l{scQIujuR@wyRzwVM&P^O_BGm= zcZ+J)lMDxr2UnX6eksMkT(gskxJ>$8Tlh8KG}O(XgSz<&mqMtZ|I2;Zr(UhH)#2q- z5#2-5p}4SLn9JS7O#|}HUdc}PEM++r9ndP<Jcp9Ffj$>8)ljjQu#6bzrAccH`i%Np+L{$7V5r>+ePa_Id)Ko=>1cs+|JXeT7aAiLr&fN zoxq}yDDQX0Cw4)eLtki=AXnT$N4sb`=XcIerH{DxyJDE-n16?VuA*F|o3SA%30Zls zkUT+x4_?lkvqzj2lxk2viWF#7*Tvwx;F)%_P4rI$7Ba zJCeZX3AvS+dDk8GSPYoZs_GztMkhJK)k#qDEzZVkBpc#pO!%<$5G&kFUh+G)*+E7d zv^ueS7uh)yzo7%C@18cXOBEEepCVFfR5i;D&hdl9an`d{3rn;X35deogpoBk`(xyOGA^S9|Nloayv(nc9IRN>y}_B zbJg*mz=A1Z(*ZRaL@m4o<`?qG5@ealU z@oI~HE4+O^+)MGeD_!Lt$HQAWF3>Jz#sDnlwTgYyTWABneX6M2quvuXD^def1EGbb z$<8hi`W#H7A6bBvWeAPLZ8lihz2DsktM6B*Tp}CUMRgo_ab9c!!VHQ55}_1>#=%pk z2u8VHfyD&KqcG%z{~lE22`T_GJu;-;2b-1%H2Iy*I{xKpV|9Ei4Acmh$Bd~F`m*Jv zyWiXOB{w5VetwlKCh=bwloXkuB%NYFK_!`rFt{|8!kbXz9WSZ^F^laCP>H{`O?fpa z$4`S+EDEpHUN!2A5r6^>DtvX+Wy<&^l}1a}20#*)KNz&PLgr&v+O=nR$? zzEk74Y~3i2;{(R{C@c45a`Q^3i!XcC$-1yh;#`l*@)c9MLQ7=_=l4qtT$af`2dA%_ zI?gXM>rt={7w~Oo-ShY!V1)~+a+{w(Y_t`!zrR^e)=z=h`|z)1D#ZX<_Ez{x>Y3;6 zh2(;k(U8+#&-l<ZDWoD+s&j$rve=_nl1X0| z*Q?LGeg!C-8`P(z@v1yukacuzRPQ7uoDC$J6Ui$G%?NE&_i&HH_TwJXFWx$LgKuwIKhespO_XF%nX` z9tzke20^gUM$oXi#cfvM|en{=1P=>L@Etj z6RqH3Ugsv4qHx?ks+h6MrA>uWb9tgXfnBZ4*s+3u8EwSJen09w=Y-mb6`gW=5;?#Q zD-P_wfPCpNn{t|BPEh116|tS(NxJALP?e}x-~3js-gQa!K21K8Z~oGG`n;r3Jt(^m zrEUe>WJNXocus~dq{?t(obi66Iv)(oJz;E6z58R&$oWZ}I$^gsT0ZC$#pDY%GpT-s zvhrz$mdIYVgErh>F0BRlb(5n3ASLKopseJBks3#8$N11>YKulV8#A2V+VTJXWX}I@ zfjIxBM=zuTz_wbOEY7)ZT8&X&6OySXdOd3Od_U&bGR6lj`-mg#qd2b7Uc9hg=k?UY zEDXL*k{vjz1}fCUj7C1i0B=hc6_MnZN~d^y&hVqBW{M3;16C)TY%n`_wX{Z`fs6iY zxjwg6(?X*^Qo^~$8F269>o<_G_Q*~EAHPu;BTh$oj2(<`{CwIfPa|4>d~Q+$N&3P} zNc&8%l}9l-6v?C_P)YH_c9O#B61VW%=Xd#^2gR&hPQq)|ymL^7^!R(2`BcCyg|xse z3vy;)2@}HUfbV1R71ApIi{VkNqGZwU?1U_H5E<+3#t&Xw2d`}GRsdePS$RSGMGC_l7y)=;p; zRDiL*Knu^YHzgK2vhihBrexBu)~J7I?v02E@3%kudBksj@|(qf5Yvnkp&WPOgtGqO zlhFo>Sx=F5kcb&x5{OKJ^$N{4=TE)%%20^`%h2rg^|w5VRszLx2X?w+87OT>w|>*S zAlhLo6fKlEpC;)u4~s6rbdjtm35=UlBF~1@p*89TnI=;OQXO)v-q7H>*Pvp)U7f+z zOM7%k2jngMD**#;tEcu8O}98viFvz^-F317NO*hfj+?uc+*R%dCPS@asrus?Ho}#a zP_a0LjC8{m=&=ly5xne4zkf8>&B-|96g>RRha}R0XB!inG4~zA&0w?yWS_?A1lkiQKbqf zqyWmI@t%>pw%8+2VDNH~Wp{Vj1t?66s z#8i5lV!-uWzcG!*HjZh8zkVeLtr;P4E4JHcaz zj+Gb~V>8Rp`TpXcf6ZJ*#$kOLi~2SNp{?@t8JQwI?RIJCscAWM4>yJxa>AJDWA~jQ znBGq?qSETsh8fnDGo#1&hM#|ItXO!yC)-}JjD?Y$mi*jUYV?JW1Dg>G3n8p3Tk>Hg zeG_J_y`FKrJkYem6_H(caz-9VYk@NI)AqY+87T7}{dKCj+hTzo&=8zh5s(Gdc->IM zV3@Hr$XblKO_xCh^)RUlDCQh<)*N|#yChk#a(=96k1r}DSIj_ilveq2X&tTU0Uhp4 zv5si4r&6Gj?50b?io?1B_rPAizTWE!6s>o_%?hptQ(HTroC3WP+^s`Cz#*q2udkes z{cFk=Ab-IpOphE*!Xe%{&&a?^p>_U&wWBp=Z{Ya{tr}v>qczgk+F`3*-|g=whh|<4D&y+-yS?-2LEsFl zbmGm1Z;~qublqUyl4yae|^2c_!isETy+ zlBN}KyLeTkOXsh&|qCI-9gp=-xp!@}yS1d?;YyARSX^58Td7Bjw!NMO70Edu0WdM}c{*Y& zVD$8n3HiCXvM1hLg4bb9hb2*6@KNu$1Nj?Po#Vj-6&RThkTqj${>((2voaNvoi{6w znNKDbZ1xR)t|GxTU9`#dIyn5-e{@~^x0~{%)O}#*Pjii1kTT<(2vbk1=-8kOk{}pV{Y@^_ktWkY5V4eH!uoVAGvU~J-`UIUEesNkHFU6xWNXNfq zlOn6JAjlk1&z;)f}Z>Fe1|3(l~-u%eBstfqCMU zQz~3mOhILci(WTES5Eni4_^l$K!_AnxaedRLAl&ZoG9LjR|djuX+v!abQ=KtX&07( z9yzpfnK=!H!+I9l#I#~mA>$OI`X#C^w4zGdm_^?PprLanMn#Mb; zrabXn&WWDVN$;|WNraKHDOhYc>skiO2>S|D|920iNsPuv*d_cOiF4p^ZJ~*wNuwC> zfs?3+&p=74a8aaSal&i0;zHr3nTs*ax`;bCD}h_ZYk)Gb+pjjNV}s6y6w>RVHm52q zK~&7OrQvw`MK$|5Hi(#YVuw?95czBQ?2G2QAO}oFVkgDqQ6z_oDDmirg77wyz-{Lb zI;D9Ohkdp%Q+1h~2KxI{DAjCKXDHWsc9P>F{PON0cfhx~A?^d^gj>#M2wwk8RtMQa zCuDu|D(p5wV;h!X9*iA^1iNH6EDpOP_+Rh*<=$6~CgpbxN&iZYd||3OnoVGTfnsVY zaP}j*>3Ggn&qj4Oz3i>);`Kh=G|oYd>W8xHkhiOy**^0VQKhhf$%EqNR5+?vCn_75 zf$%=M0`l$}nBxSMAI{TpygPxld~8)kR|*@|{UnaJHK3Ni(K(8<(K${0#24qZ5?Qt~ zL%eb7LBB&Y(RYduOXFRXY#`OVG%?=KcFtC6eQG5;lm~p-8@X|RIfi{tUU10zD|9XO z?~l!?5**gku|R%+>I}smyS=M;Idd}H^t?^DJyye(f z*ELp#aPr)rF8TI9mY&Qem&MKU0oS~Fiw_4Jr#H=MRAXlUS^x9?tETJ-S?-$x*);Ja zhmI%FQ;v~UQdT%}pDwb0ef6(o^+=0s z2Tn$T`oUrO{aYypjNe8o0*nZhkfqVv{PzW5zI7}|pJ#o|H!f^YRR2FH6(~)EVhBbLe{^=jkrj zi9?Hi)UAh%y+LBy|5vxpA2crj`hsi}S|}Gf0`ZnZ#PBbq8v#*@{IX=p1iK|T2-ZTv zbinL?HcR-sVDoqh8%Q{A7#ir_hR-n1qAUC7Zt}wa8r2^wI>MHRFBSB|GGiyVio}B6 z{6(+N0z7o?l6m!|r4(}(`w@b-;r z-Hf9FZQtu!Waw#!+5J}e?N`qzwyU%X>?>zZ1M!sj@uoc$P275ug+pRu0yC zR(NW#jIh)BHobj;jo41C9U(Tzu(nTIv}1>i`n1YsE0;IdIMMD)fug3sS4Zzt+>o>b zxmmyK2A@WC75|#YTF!9@S2hQ%mM0M0?b;*5){%iaG46nARm-@oBvYw_{Et(rO!;zU z-FT%)EWIl><1(`4w$$_ZuCp2!$DQdbuF>x~8|Rth|5WuivYg$1u;VV=2ryRUD~8&(?9(!;bos; z66axly#^^kk&H{D=V?QQ++)ri^g8zmQ|Wp-AXbmH-i$qc%KkrDS+dFT?Mad5dv$DL ztPX52EbQVP@b7~uaND$)S2f5!J1C0}YXf2gEigl3>5k^MvJR9OiN&~c6;~8E1~o}6}XFIn4L4v&W&~%%gn?QB_lge zqhDohpjg?PNkzX*hz&DhYIBY(jqG(`Ox2rU>KMfwp-2@l-7~rwjp{=`ye|HWzN(~A zohyh9Iy@^%*#Qo8q+pvr>a5(9_mgZ@I_wps(jCHTpHl9LFsvZSQawFGrzKaA=e?Fw z$?p)R@h)?;5GrYg=kWPK-iffqi;qE4+D`X+^^;#*e0|H8PcLdv8+satXJcO=vBuA{ z2Ur1#3QNoIGT$$D*c>$rJHS9T3lIS1Cdp~RX)qP7uoX3qfqnKSSsz#f3D%w3-~m>U znEbnh@4v+{+Lm`u_MIThUb4KSbQ62Ckz(Q~5<^AwyVmk?rqLYsKO`&^BE281ad6%m z%RDl^VX+1&6YTKn{~*gq`Po2-hxwA&VGV_aMQ%snb|5WBZeiTcLB7zN9<|OoNuJ>U z(nX%(*)jx13c3b6|K#C6RwnD6FJBiIF1jf%U(`+-)h#a9U){uQl|AN127VHFC7{*& ziEpF22HJpAa}JuU3Ev~`pl=D!dD?$Uvf8Sc`mfK1-d+d4YW(1>5aWy@bY7G}b~&)O z3JacL0kSHJsi4SyDqJuT2GOtKfGeYz z-4rRNA~sX+B~xerZTpW(zlC&cN8Y$D{=0FVl)!`)J2zk z>W^22wNf-ct*}jDXrWaU3Xw~_QN0^VPKR}=RSbleaIgbgnI)U!Y@GOd`glR_4}SWN z5jm~@{GIQRCI>b?4^6Oehho|(qN5^oG$x`nfuo`qMKF})==16@j@u+z>!vB@G%+po zAx;O7kYYmJ5*=jQ8T14}h@+m9D9;xk@HY^mV+Ey?u!lN<>NYgEW?E5a*69P;M}*zCR9;nYe9+2NC4+@*i~@*lVuP0(L7 z7yke?POMPnz|_|ua5~Ie+Cnjj6xl#U6mwU)Y83`LgKXy=$|~7%U`oFmjFJm`Jq^af z+yK1z9gg#^mC(@JhjD%stH(1QtoT@9gvR@&Q~yNPv4e&K`!!HsKMWe%C}u1EvJnU8 zuXKF~xo}#=WslSu8N40p*163jXHKqbCUWsxI^6AHi*3tv6DIVaFqV$SE%~S*b zGo)W{a;t@nLA%=m>xTS!57ORa&dybM=ewJg<8eS8*kV{oI8 zsdq@R>4Qb>i!^xeqx+Dr4dOmWLl3BMvE3H378bgA=dbr=-O?lprfO?yh49ZA#eiPy z0~M7|l5Vjk7veB`J)sm1z&TFe$dK}lm2;i^7c0Y9iLU1@ zlr65j0H>O-4ZLGsjBbI|_XuQ)p!^!jtM~audTAoPvPJjj@A63S>2l2h31Pcid6U0` z#Tzww&fX{NOuoZbRZaJpQ@q4zc1pd^%1E07hu&699==?8Kry`(>7pW5@QPqD0-gy9 z9sG5v02AFbeRFj)l0f^XVuog|+fh}H%LX?L!$Eki1&Uz@T+@_soLVq0I@n`IxVP|8 zod{EzFaU=X$3sqeA{~u#h3MlnD)xHTD|)!bz#e8-?V7WLy92&0npXKn_v=2pXAQV( zu=)PEAfmcby;7~;z3vT#A|uc9u-4w?(x%cx&aacE(z=iqzK(x+S{B^~{7hN}`dd0k zpQ)sKB@m^lSLh^l(tb&4=vBz70bcD`(jL;L7jYPH{UwX9ge}a9)qSx0O8k5J!2A!4 z$ou36E9R442S(lo6XY$Q%q*Qu20?k7cg#JH+egY?)kzM6E<~0xGbCNq1rn4=`p|5( zccrjHxHqg+T~0mXwzy;|b3H3u_J%DJ>~udSD5tK4#BuUqH{fX>pDr@*V~_j_$I!=C zRXMfIKga)=ccU6V@8)xDmotqrYty#Q57`eSL@U zA(P{OEd+m;%#diPynp(e48dN;NbSow_oSP+<x-HQ~jhi`q$sybKl7Qv-|ox?2Py%s#-9`XqVKw_cdgx1KXuk6W2PPVqz(> zmWoKFfmOo~D$^92Rj}s52@$uiC)RVb60A7Aj0`AyUS$W?s(1g#$(*p?VM96=lmvjj zw!|Zvv&ppw9Fi(!vNBD1OTBoXqTRbE6jwS2_~&6}=@1&uiWa1aV{Juh{KA18-JFf! zfV{9r+uUS@5bDg##ShH$XI`qpc?&r;xG;|*H)z~`?k62`Aj1Y9qm^^jV+C6Lctiv<(irg4OsSG+i0~J|&LSGz3vBi6S)(2Z4JbTd22bO^_QWEFG=KXyy zyH|omXIMjoG{ozDG)h26~Qi|C{ zk)2dT25Ex`ojZ@@HkU{yz+;LO6#u`pOKojxtiZu=|H1OHjp4w`j!>QQ+F!ox1!h+O zGZ>TA*UDo&HeVM8 zd%A^R5tQc9pgs#q8;9MF@Rka?J!&PlTpHC|Ij!aaG(0_ErlkQu5p z)T!$98+n7Wea?A;U2_trLgDIw8}3LpNzTC@VZF+}yBiH08mqxDY-X^2NXOpNG8jg(2%UqMKJ+jej{2YN zT|%}x@JhJM#7N{)43unSppm!)d;F1x?6)j>CHecH?8u@4ZlgMt-pQbZS4!YD-v-Zy z1$D1CFY2Cu`_*m!#}_>ySLtR*XVJ+9n`L%0$416;`OB9{E=D*yvk?)Ai^$S1s58_}1K_zIu9X#H4E>~OQ1MJC-X{cm6ML^3uO z!-4rFE#$*t#d@6-b4LsbM*j=b2V!Y-_&kTW3Ayn7+y6cFWp_a_C;4DNl0*aa26I3% zNRc@wPqZN@NrH75@u8V>kg=eKDi%1}$r zo93(q4jWy!z(Y_g>*6KKD}^^{odgpMkyIAxxlXCHs!vr%IY$GQ3fBAV;!8f+7*vYXE0X*^Y*tz2|FFf@=Qd{?N0uS=><%kz zP5wjYl)sr1!8xqyu|VzCNrrq={ZjlZn9IV`%qQF%9)0w6ae?xb>>8coah%>mmh&3b z#q?obUuZ6;S-xGB3HPswKLf4JPar~D&oSKVktKsP*a}h=b~s=QH;UIyV~^O6ZKp3A zBIntt+v0!~a;P1hp|6gI8Nq?q&=#2ykcrDE$bgm#nc7xMF9z8b9@Y${l@PL~E1S0I zJaxWkmT{`;`>}K-**?-9umcBNt4yY^5{iLjf&waH-M6lbqevIf=Wmr9guv@2={;ek zB0-b}q&oSMdPi>e|hn`;N; zSP6|$CLx=mLf@HX>#``r?cQrf(4@-W$s>Cl*uS}80+6E=bC@EBsE9VY)H98a`%-04 z8{Pj()hkCl>!pS(M?AMV?*^x4c>pllE+gB))xbveo-O=H!9^uD+&=gCt4BP0LQw=n zuK=wDW{hfHhPcc1&`jig8{TZKORGFm&=NsO@?W(OR~t(*t4 zR(TE=8)(|4xtuIjlW&wC`kjgLPJWZ5JM2mDYDgcb=hVR*qLXB)s=^vM9Z+!b`ELaD z>S3R}qG=0m(x;cj0+q^R*Bc?(su);Q*GVz2K2g5Z`;?nurLMW^b=9kV{v(p%Z}`n1 zDibT}rZBZVx5hu0u}y=kA{iso5Z3r%8A1 zk?-XnbI$;Z3p>nNmW>1(%VfD1+uvbjnI_-+>A4l=Y7cDon;bY~Y(X0$Q>-H=U#pnW z%Uvx`mgkDM!L|fkXq9h++}a665RARmAhJD9UfAQ+-w#gvfpNn6Me#2ONZk~2&&0oI zrkHCKX#zn>$cr!czUcw;J+CBhBHB5Q!UdGL++7;^cd4@_cT4c7R)wCb}+$QqV>c zqwEQdCMSTXXF1fKW5(m1z*?CmXZl5Q^lP^loqsFl+k;LiUw` z%|Rs|vEUHv{Pu(-%2#?7hiSSY&Qd3h346haTRk`=ro4%V)4gKKS8wuZ8_w#}96pvNPiiJSYH1nt=ZRRo8 zIQb^g3kKElYh{0jJwVuh##o+U1rWzw^12_dx%*#6BlO|==((hwU9pG*&rK^$od14` zxlfU9Dk6akwoZxcUz&pfD}hNrCM-D|y!Sj&I@!R(jqsxZU}LXBVrC*pjMT^~+%EZR zuww0$Br<$I-5Yw=xg<~%KNq9V@pErLIJ(pMhX0UL88CLdOi=Z=gq=y4;?>EsX=-%BycVbMOHuH zK+B=G&T3OV2)h&5I}5p6bLQZlO1_}PqmgrjhrhB`fhj>boI^7;C4L_(BIjc^*b(?k z9!qyl%jY<(-@oF$yJhAfM28JNTL=oqIoC}?{@o@C&dAro24!(`;>dF41R2NShRZrb zIsA!2y1 z56+>>B%1#PFu>4Sfrp6!fZ3<8!o#HZVpn8YNr$QpF9~dttl^xZHG9O3G7R1fIiVgD z=n_r`m4pc<)w1IKKXqTt-jIDyS-pS94Y!29wd!XxjfP_R`(Eovsso3OLG^5yp(vsl z@WFGbh&4b&*{B{6r@*dt>--_#Es`w1b}1xj%9DMtVztb7DW_P}D%%aMZa|uMbMR7L zp8Kh={?Oy1C|>LQgjWWgvS)V8tQ0K`hx#<%@&)!WH7ua~6w>BD?Sg4mP^Nyi{ihGj zrC1%-->@LNI^f#OS;b57-|F875ona2LdjL6)V0sZzi8O3M8%7`u**GGsF=LqOUgop z(T=E3|KuP^XSV=y+@&3CObxd>#S~LWk$fs*2fa(!D({sf${)yttBC~kutQFwZl z;KGYyJYybZD-mXHYfR6_{OOh-lw4f%FXmAdhqWXY!aa~u1^HF2$b@|~pn%Stvk}DR zazGkHFRxh{bX1lEvy#3R9+#bmcuS(G39vC^8*3I)`G(HDf^$X14VP9x_+8*rfHgc>3x#mz#c2q&JFWe zGbsif5-k;xslv2MgPfF63A4eoR(Ul@3woJ(BGewLq#K#Ese1WB6ya@?A=MLxyzG|; zZX9WAVPZNz;F*1)!U`snTYt6d?rbAWoJ6WJlEuy;ao~vT5ff1Dp%_3}5pwfy3_=nx z%@(f~{vjw{!mPS2UO6;WGvz7<17Mm`T=g+VLXXq91N=NoNq1sprIs z5@+Uek9cA+%2k&=-uNI=GT0P(V@KKZxU(hl{w3q1JKr(l=8YpgKO(0_O1^VobbvbV zFq88M#ayCDJrz;sjzqLt1@iUi1wtYPoi4H5p5TgwFzARKy;ex3syw7sK0s$KtX3t6 z)-jqS$^U2XP2if!()8iF;tk2ckd0t+1&bwuEGmnk1)W&c-BmqPHPid@@837mtISMS zFE!mg>FVz4uEqrw1q1~b&;TMtHWhJ0Wp%-&N~gSk1mhYD6Rv-I-m|`z`aY_=vf@3 z9+2+Xe3yTcDECGmpOdc0Ps6p(7aouB5Pr_A#L%BRIV zY*lHW@Nxv01>5HTh0|sC8?4X&2nBbOj-6WZpPnXz^PQS47sql_H^gQ1G9HqxJq*kQN~Kb05K4+z9C(-387SCL(*#?ID%_57mJ5-H&GEjb zZIkib56bRezU6?dCc}B75cHMaKfLUerJJWF&1!_SEi5Hi^uNaKG&C$Z9+{Siu+eZl z|Cqziu)^x3H#083^Rm+kqTMpmmO{kjZ18~mDA7W0KFwc^;$)rR7Z?-w7(1zVO_%E`VsF+;!9b$|2};3?ToE6euHNU}6@H0hl(%y{ zGAzu1tw6DO%;AUDjmfbOwOpae<4DN-^uI{bE3-8&GIN}^P)r6zHc+wqrDwe}p++e` zqBdBa8<8yP^Vtg?lKMKgiT>0JYc655zkeH1VKHZpKV*f9 z*WS&}X`O9yGs;8$xPlbDGH%96GdJT9#Q^!wUZV56Fu_kG-A`c{}=c_z8zp3t3D99F74Y&ky2&Gwcv@bart4%N{PncAJBX%q64$nJ!(v&ygrX)))aBz}IPC$gME^i_!^v6bAuD7~x`JmWTH zq1#GOBCPUj4eQ`QCsP^lSpqr^wF~ddHv4}p-8oBzcUF6#Dz_`gV}+AFJ;ZFB$B>#Ddcs)qBKcxk|j$%_SAp97&QC+k!N zoUEWEk33~Jz26>UA!D^-qr1f8w`{CHM9n+3`1;FUWnkP!sWYS*(kJ2h5%8j5Msq-h z-32XQJU2FC;>82m`3_c?nDq175B|Z*;f(8`k>dfmhrl1g=28RyZiBR)9`MkDfZ5YJ zxDFHPklIKfWuVYh_R#Aic8k<^g{$S~qw|$j1Vo9_<<-D=h}-&9dlqurh_cZQ%@*qS|6UzV>w!k`DYZ;NQvVkt%JSw{@1>G(%T2#po!~M8)c%EVja@GqQeB5_;Xr59}H%PJ4 zvVc=Zw~{t_KU_IM=W~(8!Wc!yqTH2avEYvMB;3Y%Mn5V` zwBE1Cx56z!eofvbza-Ja$Lip;+FlmuSj0UKQ7Zs@?zDBpd1a=pu@rP>!xrYmxU(Yv zT)@fZoDMBY+**u8$JDAvqYmD;J%kkZbX)g^giu$DluId$BM&lc5Fhxupd3U z(Z_zbb!&0;{iz!)gTq!7kXL`z7}`kT?FFhY6sy#!>gbBF9^Nr-DfBs13h_@&?QED? zIpexJ?pxRRPsDf6_G>rHz}UY{o1e7)wU1C`;=Co;f9wAFL(8%`EA?jW+!TWLSKT6P z5gIk-O1NE-nQpj1W2pYIXPh9J(@yt7#X-tMY0t68hs{$wt{Yk5!+GuVUf+*Cd)dq8 z^boqoj1dG&G)kmSD^pLk-?1Sz z_!3z^nd~uJ%JV4(${4e$Sj?PcN2HQ2AyNn&3r?3C!%=0z7EYNEwwMOhTD;b%xG&3$ zD56UZCGwSFShAEp#liDq`+N*r+$Ug_VKpDlYe3Fi$*uJ_0i}582m4626NgqR%@Dkc zVoE4dNX6!Iug}}L@PjyNDd__0lN0nkW(oDdhsL9FdPSgCa*w$_@52v1!c(29?H=92 z$Kt{_@aCIyI_7pNyMn0qaWzYjjv2%C1*fp<~i+m$&D>X+p>&9C~9vIJVzGdvokjfxbI*u#@rx;ALlJlrF9 zh9^X8$O&!5*1E2;TM0-4}0|{Hzc%N8gdEP-_WO0KcKbTt03E37S+y2?0^yt+-8*1 z!{|SrI}~(A+&4E)&?tTkmAGfgQC=yhnQ4#~iZI@L1{xQ;3}r0wyq&TXUN-MKcd?*0 zs8CQxEp>q7%`Gq2cRdsXdMIU4iQbiTNu*9u z%4ro2x!@c3$@D&W1{&TPOs9$oC(?e$} zG=4Wgnh48+b*fBx4;Y#qoTK6-29pOn=uNH~MWd?vCu{yW>5V%JAAh+?rSZB*54zL~ zv!&bVr2*>$9Jd^f(*m(KEjB*EY5}C?_O*1r>>W#JsgRvgjjBbMd&FYCeo2)dW*E^* zISvFP%?S%&jP$s}&ancKOS8!iIqxQtSka-*jsO_>=FxXa;Y?uVKNx+P zLDtiBxf*L-vPHnZ(;?qDt8P}So4Rfm^c--?fs#Lmi|fmB)u7*nh!Y`dXuDQqN3C;E z-8lPMYm&T63N-rxS>Vog3}EA#Etv9KI#nK`{`}-cH5t)#P?~g`1YMZANL}=a7zm6s`xp8LW9g-SP@IJYF^bG#QX*K+lTt z9MU<}3v(SI(FjxRtbob)2;Vrv1em7+o=#gdWkswr1@aRqSaw}UKbf^wUK^YtS>yI( z);Ss~=_-6yarz}&ea-=~E$r9p=nS_5;*C)Mtrb1~lGb-8w_JQQq=a;kb%9;sh8yxD zvb{RUIeK7zwbJgpcsszrWKZmCKG-cuTO|!QEF0de^p)i)ajSrq8T;9jIjPY#zL^o~ zUfKQdy&%Jk9>Y#IRD=U7s=8eQ39{=uV zK}Vo%C5tWt!L_ECGU3KR)Fgi}N2h9vIS_iB+!fz|N~+7u;;E?8vwnV>Up*Ih*^cpr z#}8%u#)zN(AXq_}`cZt;ZzZNxNqA5A8S4zrIJ@%U$EIQQVEJ@XJvi zT$QKG&(rI<1>9s#w--VM zz-Lhvr=HhF z+QMLGEx#~5&A;xgt#mnric98|MfOSR`A50gp*qzPM)yxk12u}(oNU)!3;HDa+y|T` z%nk0f&4^Ck1xdqyk~WnpJ*K4cBK<7d9-HqC^nfL~WE zA{$>BMC~&})OL!=qsUe&wl%Cww%mQ$>^vZV=$LDM)Tr7>Jz3bKg1$7mo}V^v`CN^n zBr-?&^zM$id0ywhlV(5vEGA|=P+{>ltZ)5Bt>m99S#zJKzjR{9R$~T;6pBfr$SPwF z(~uKrfn881XsPyrwO5JEKpVepF~e7*+#EPWbaQmf@7=Erk<1 zBUZE#8zp*4J0JVUvgj>dXO$1b7fLx|tbav3l)?KdH~RbZec~hq20K;pTFCg21sNKEQ#{=p{8c4w!Bgp~$nqbgpSf1w)`nsfmBY1J5_ z7Iv7KR|64$>@Kf}F!?94KdHBnos-EYW(#{2#X$W>ITf2My5`kIU+4D9Gb7G=>uGcs zaqTry=%Aiy)jX^c09Q6QTEO3h(05z@Qu@_lOOHOlQC=M3rS zeKBXmI10Fs6#i=;C(@RQdg1D8euEw019d5E!r#v82k!h{-0v-`C)xbXoOring&n?2 zo*G>Z!ZVjc%e;2RR6!75%Q*oAbd|z)B*)@SH$haF-WoZo*pXcTVmZh?=9~AvZ(57K{R#Cp zInQpd?8N1^_so{mYZUV-MOvxYYDpH2twzQr)FivvFX@dc6m4R(ututljV4C6)Qmuk z&UDgHZr4V1ihbS#9?u@d^|gkBo9uIRff%;F`@AzGw}D4KZhoN%7v65y!_b|*EySQ_ zam`ClUz(an7ccDPV|AEDu}xVS2G136`bk-o7O3lt?~L+V$h)e3+B<^pvy8I_#5z+w zDo69h?SQgjfMfg85?jxXLw0lzr~J7g9CCt{r$yCQ$NkR7WQLSU2~}hZJ2T|G zt{ORHW`;^ArjR090kFMv9h9lth{;!k6Ez6;-)DiWF)1kq_5Vd>&+<&-x z$^s2{G|9NW_m2`qoaQp_V~8(`yc&X*k5J!TXx=`qTk`ki;2m|>~_iigsfm^WSrR7fogY&_KH*YO$acxZP8udgT+eN|`Xn0^saXI*upW zivIL1fe9JEEt~R3k~kSCGlqYqn<-`!?wVtPpcpK17uXSv0=m_Mk=e~4_|P{OTse%6 zp)Uy$7hp`7usAx4@6}|V$MI#cc$AglaNeI+Xwoc!WnI%tm4dcXy;K)7^iQ`Way z$MxU-zmC6JS`w#?_FFM=n6d4olbA!G0@Clk);|xNA9ZEWRz^?nnK9rAo)7#}C_wf4 zeo4EB!^y1-(=#S0u%1}<4_Jr7Ncq7Vfl>d>m=+rMM-MKO4eVSHCl1{1HCr#W6a$gm z93W2fE%YrCXnl7GZ%A=tlP*66m}`b@0D^i+#_C&xm176(n*sibaP5t7VJw z>Ey6@&nLj|eT2JeS_R`s8#5+cJ+(6I;bQX>+@v*2!%axZ@!z(d>~LZ_4JfZ3=6+UE z3~YK2P_e~P16-WP+Td&9b@Xnp&!=w?)l2K>N3&5#=+^W*P-|Vy-|e+g(jPu3(|IqG zUh-TwwKGzqET_8y&rH`S&P;y{_n^{ZpKMq3E-!p?^XqH;b*iJ#+?pS%QPjCshZXww za*+YvPB`$9#ebgISnHuVR*XM07J7?Vk`Ux*LVudorGa(?&uZ8F(l?$PE zzLB$Bh4uLtJkz+@ymr3EznvTiJ)wx73=vaKvM4E{SkfX~MV3#lrf*0)IK5z5R(j?| z91C8dyc}#Nl$cGGT_9sNqJ0k7LFUIRZe?5Y=(0&GI`7t>SfK;Pz?mAW=lb1mgUeeK zRBq6I-AN$&-U!NQ=;IiPK1(87fxF2O*m-`Y$AG5i2XVle`MW#hv@xEj9CRaw#U8B3r1~Lf>xrSy4V$ry7L5v2FAca*ys*UGYk0P6(5sQ2XiG zEI7M(T1w2um}9|d(6%}er1N92A?UJ<~gbc)GkcD3>c1IG67krVLOX`6jMf#T~zGTtQqFw>*<5DQNu^QJMd#jmUX)Vt?^>Y zm|=(dbUj48fg4M7Rs7f+>)?%PoJLNe0O{J)$AXLLG|-y@wQrv?U>?Q1kWSGH6yA2U z=d6HZwsRQiV019BxylMS&KsXi4GK*9i3vD!7wcz{Rwv#ogBI{GM0ZilEsAtPoCMS+ zFfVycesW#~!WbHW)HmiD6@oSXNfGTdZh7>Q6kdmXpR6^ko~usu?B;6}c>ijTW6Bg> z4{s%Bk9hs`{nMX7p;NZFYtCP8kT}6EkHUZikhM5K){-aU3vae8ye2$H_Y-xON2xqv z3JO@^8=MsE59*bj_3n|LLe`Cf0KC*HZ0BofZOC>`GN;1hVvxF5et7PukO@5q?MTRR zsK&Rhh^%nKAu1xVb^4RgF33n{gJkS;DN_6X5JviO*m%z3SHkXxuw+J->wlW?_M7JI z-sIY3ve=B%wwGd{Ri}%JT{$^>7E%vvB<;YSjlEs%eAEDgP}h3*&t>b~Z}IaCTy1#U zL+Kun|4N9?h+Z){TX2GN)O+RRM%CT_Y62OX&*4#2doJN_l$;B@J?%Ucqi7WC-HY9y z-n-uY>D@Y2>pN|r^9#CDFzU@Y4T{f+-n)R?0Hmdj(&Z87eDHOXs#La7vd`^vh@`BT zjOQLGjbm)5*1y&C)x~c%sdTEPp*dfzhp|oy&I#+K1M|D&#qPc0T=AjjDNdbbaxLbT zUD2=eMp%>1M<2ITF?mIH=b*7M%qFH&%vy@9p<<2o?L*}0pICcgs8@e|nbcmD>Xu-! zerz+mtq7|7o&>-M-6t*L4tH8IFs!6tjjRiKycU8li@+haC5V zGt{SKRlwEkuo|dQC(TAsj5>xRF21#9+p+&Lf#UeT-SHk zy;l5@*tAT^{`QqdvWDF<<-EZx0;*fXeA+CE*-Vj5RBS%?CJ6UogNHgURPTfBB)ClA zW=>Dv7Z17g%8`nBC8vxA+Tc{dK5vuBcgSo#8A9(UNQRk^AD<3WfMuP9a! zqNkfB2WJ~NV`ngXzJx06<5++e`tAc)$l`{5-v3@uiD>~k>iADbF=b;#Z`Q6?Mm zo&81Gq--*2F#GWxp%`cms6h7o%Rpw(L#OiUrELT|0vqS;j@iVka9gZuiP{7i;P&}C znNGEf+rp`LJ?ot=-yp(gTE**qb*d+x_0pce^YCgr9q-*R{T2^O5v1tq;c+4NiKkA5 zBQJiv92SR*KrMLQclY!bPKs-r9l8p~^aFkRL$&_rxPQ2;48HS5SH9Z3_UddCipoR& zxPlZpaSZvS8N3fs3`mRY#eK1%X$aeju$*qdqcY5p7{SiN`(oTe3;-LQ8Z}F?v89I` zR;sbOp+mlestepJgUt&*mmPt_G%D1XdQ0KKV9G@9v8Y4JC4vkoJf9V)hwJ^4HXl29 zxsHMLxKYh6XgF!gWL3Cjg}BrDxckh|^97gX#~`_OmVZfD5xgbzNnizt@$*i##l!SCfJMQ6@e z>kP72&m|hiJbShy$GT57{K}X=h{c1}k#lS3*M8$=LQdmP_q|6pIkD%m-wY)M6a%$C zxya0o9155oT^ZaVYZvYn*OD6E1IRnpy;-fy4fqr~1P12cnu~QGK;|FTCG4P);=eYu zhS%+u&n@?@R#rQdXLnqI8mr*iLc{bcEF{L>sa}R47UVVvWc1ceyo75~iy+MC&7ym~G0Nb5^<`M?5li zYZL}kgZKAH%jt~y{SrHPzUC+x7rurW9P?4fe>AJ~gG@`$mJMQ@*pac~2N#02+Ix+P zbHd|6S};kUco<33hFo&od*yl14w*$?AU$4q2Q*2#U@m8$cTz-lgu2G}iFdnU4>st? zO7ZbQW&Dt^HbO?vjP75V7NVOmYi5$0FPRIo!Ys4+kYXNCq?d|4EjSI)**HOK2&NK~ z!y2S<0*u%0=iQp8ZjhEwL)Gcz&}{-Z=oB{rvF!>$6?70(`R$h*cy{BVZ2YmL3ml$@ zoFSJ6>C*6WpkxcF`SgJ{G^XD1FL%=l!8A^SnR2T7-|a)8~P$CRfx7pG6;uZ+%j^ z1t~`qpMR>zucaNK$84v^1*^sN6d*`xnBwE1mv`ISj7^_#iTwkUc&-Fg5sduR22}}n2s`Pj-~(dZUE%f|@=wU=ognRoSarG_ zMe4O7?flI%HqW@uT@2Q3P?j34K28RqBJqMJK6OIeAWiht(}$F>lRMUkRFY| zeH_++Vq&Yp?GTh)91u6Vs*#vzt;b_A)@v0ChFmU@es|=~z&C>T2cHT)0V2B$R!?jU zh8m0_L7(gbwAJj8o%6)|M%vZc2E$}5`3+~ zN1Qw1WlGSP_T1*7QFJP+{M5~oHU39=z;Dt;-xaQwcgU}WKOhNG^{3)u`XaqdT1EPN zIu!M;4sFO7-`qbxXZAW{-2{!W2b(k)_+6KmogyO9Ejz+c-4uld-zmJhg#+$OArEP2 zh(j%M_%%Ouk|#|6hCMv&1%T1VVF!;d|6=*C{$N_QzJ2X1A-(KYE$2sPg0h@x_>UmyKx^Y=A=qWwV^RlVFMRi||L=fpGk=@OX-ffs^|pR|Gh zq#QdD$5|vuG58Jj?A+?3~&F;b*W)Xtz4# zNYWWUb5k&$215m8e$qlwp=X-N?opkQ3P)oNfi^QjAVk=Aem#|>#q!QLWU?s2g z8Q^Ldd`P1>LaKvXOoCwZd(k=j9ky&iPo()Y$i0BN@t(&rSZp{etlX zq@CC#SqVsQ@k`|5f0&ejd}n3YfX5Qo{V^X4lb9aqW2uI$@kfZXL3#oKgVllq?w^FN zn%7BRUa((oJnKj^H7=-(Jj-@D`=IE(tuMPqm!AJ%PQ3SaTF=jiBEtdhUNN%NUlI3A zw3;N$h>i1Xm$_pz2rJKps_?!MZ^^7?g}@9MDBAfLLR$fqEk?xvbv6eJ>k38Kq%-95 zsn-^-+MsL>R&+N=v*-#Rr1G!f^h?hBt1%w;sTWca=s~No48#JOCD+1r3OygQM^M?@ z?b;wc9GpeBK^VWAE_Exli#)xJ^Tf_tvzZVp_p+KN=XC)3$_HkShX%=scj#7VkcuUB zvV7MnFD%d~7i|?cOf8c1Mdd1Qh_l6eWxY{4Rm-f6k|FVliBdfPfUyL*6-e6ubf(9R ze9LxRrws&KX|TnLtVFj8-ZDlfGgJY`35J5sM`uqNDI#zGuW18g7ZM%_lEDW;9i+Fq>iqD&X|)V z&!Zp`H>au5&>_QldXG+pWUzQXE;JuVWD1}Q2d|~e3%TcDgae-ao_MTN6^e=k)sp+N zxX^4@<9+y)Hlz=vy}G&WlFG1K!ehbPIsI-o=iFCbQ7)7AyX#cRqI*CNc4m5lG*NnF zG_j-+kYE$C*tt#C?jGUYt)F6rEb51wy%%|ykoB`xvWTRz3%xk8f3njIhdC4j8n&5K z?25_ld_+JBFPnEmdQXn63f**#?>YKlG$4X=C$Lb2)JA85&IMIVDghsNAXzmLwrfm? zvJM2hUFn_N@xXFx{EC$I<(Oe)9mS+jBnb(&G|KhUw|MU4whGhdWQjD2rkEX(|Hm3K z98zCeo9lSO)oJ~9s^1dwzO z;^Ym|botS5W_sp>vH*su+TGeAT>120e56f!7PizY1t^S_9JU&2P&(ulKJEUj!j{AT zI+m};^kcMUzzydu-(hUWtjxRf?u)xK7Jkc8mDXu@K2~&Q5pkOW`=T@o9kPvL4@!B= zM&{FiUM^OVH!3D#(RhX;7KnZJ&;-0=kXNjYsved%V4{syvjQ^-kt!-I7Ik zzCUV>JXd@;vO``IjTLgpIE1yt%iX)DU!?oPZ_4r^s^1daLr~~Kr#ek?{L&-v&Qh;- zevYzb7SJp%YzG#{XHP(RFlds({R{5Py5w~?_=shLlb~iNhG-^ddBj`}%*xeE+7@ zWQ9(>`>$8XT6Wta=UsS#5qY@H+Db7{O}-g4M(O1wpH6}QmP4B*YF2heuJ(8jOZKo6 zvry1Q?~`qXbW+v4A(x{OMI$f2J?cRh8=K!iwOX;qx6o}Rh5t4xl9)zC6)03=R2bV2Oui8& zx3XP#bFfmo!fhMf5{!wmJ5tPqX%rt61?`ALWm;&x+v;B)(=53?ZRN}l>L4*@4;N~T z_!)hP8XdAWU6l;3-a(o0^TgxH88Ya?Y8i84BLnAhTa z4mTnCp`_>UL45~PV*?;;nm~Ks^n>#^b1YRcoOV-WMd!jOpNt(FM&5nov@|XidB#qX zQ_;Zq4WzudbH_i`*amXkH<`EE!yZ5O(dNX4c^~;lKiZ&nxW?{R|G0L)&JrbToL46f z*;>I=Hk7rU^{(gF^Sgxo)2gRoH$U>rXx(%wE$6`;yoMYAUG#p~-Wc!Da_jkxMgmkj z%#igV#)v8F1KR1z)aFZn__k?rk*^MXhn#ideRP-Ea&eVnS}1Y}IjYdz#G=MDQLlJQ zREfAmtd8g1k!m=5gRs5mjuiL6YCU~3q5+s5;=MJJ#jk6Wa3XGghx`#;FKzNpoqi13 zX@^{l;vKzW!?|FBbW9(Hj39LOuOZh77_*g56$_0N2u3>@%t<8zW{$woNS0js`DQ zcIdUXqgQ8HN~BxS7Dr7UwMJPMy+a6nf!rM+)T0f-1l=0X4n?|~7B&|XS}nk80!I7B z?0k*Y1URqz7w>yLW}eCRWXXS2NcOVZ7&-9@1?Ahr9NS|QQ%jL5Di+&gi#b((r-8eB zV_*;In^D6#K{wEA{2zLE%C<`4C%@>9<�H>W!=7rJ>v5|2@2OF?(eF;#Kqaii5wOLwka=!=_SpQ<|gpI14-J#HQYfLY*uOj7a9v9a8He@x3H@w$Y68k z&1MPmH>$TO2Rv4ILVW`m*-lsiTS&S*6I6kWmyZP(iHjnTUbsPeLX3AHIV^j+0EdCc z5@@3#WrS%^*}nG21DUN5vMsvF|$@L}}m}CC*mQO6fFgZ z8pCrK`W^3D!*&X71`*qV$9eygX{R4rDiu86)9A#LZ$-Vps4bTz8k7}^u<-zjEJ&{$ zc0e8SWm1iYdh^tzSs2rqI2_oF0ONu&&Uv-{aqDJ;{_EFX`d%U<3ul7#LKv1g#0k`Y zjR);NOauL&4S0$X>j)TOA(`~`zj(O+V=W};+3F=o@nLLh8gju-X^o=F?|%5i*$>lf z*=k0fPVPkgS@^fFZuw`EH#({JKR+YKoY)(^X6B7HQOr4toPC-XY?fe^zdFf(X(%!$ zJffGm-{fOca)WfG=e2P4$3mn7+8mK8Qa4I#gOQ#WPh`06pS6NlDXa}{lvZ*E+_k=i zg4$r%=PD}v)cML|!I^?y?#DtSLwz!*+p7$ePb$MOd-FG2shJdlra>zyR(@*e7GrMxZglH>`4I?ZjdyKcT!pNlD zwd;QUveQdp7`~dW=P#L}<*eX!$#tqGe)9YkVOY7+&c{Gp9(_YxM_&!NIg5=I7~=#O zA7cxTalILfhgrdcdh|yR_g|WprVrPwoK0@9i{mh|GvVmt3wMcqyy=l04KKMCf{Mr(1<9=YFC9o-Lw&T9TsYKO#3c z`=|EISRazi$)a}!Y82TKNfEvBRE4=*cy&8`5a@|D6=vg&KvdEvBO1T*^ckDPLY)cd+em? zo(rmP&64!c>i~018mMzUN;%iOIwLPHP;d7*2QmF-31(dTNaDfIBsPJ+d?1Z z5xHr-+mIT3iL7V0-Z?Ms2?(qV%UkAC49KWwQ?cptPH3?@q^u3XY`jL%JEO)|UE+7! z0}s_&U&9~!rFu!8@)51R0V#tv@%n(QfEyevbkG75i9@47>|r(B9*zSyFL?CS?~2P_ z_C0&Iq;9T8vD+)pE5*On4dvUAZEVP86{zhb$@+m>IbEI}x!HC7lwKeM+eTj%8;(4d zs&=~KtFzuKsZ>dq5E;lY1-X)HkQQ*Vc~5|!d4GV8S2N?N0!KZm{A?Pqkd=$$r)m^W z=%M*tz!tpORSi5h{KlF2u0sKsT7VNcSXTH7Pvbo-aWy`q3`fn2U%vnxW|!F+pt`9s z6~8eVpaVblkdoUkS;9NP%&mSzF+e_kpNa*o11MG~_SwZf4MF5|`4#c{D0T6|#R8*Z zU~iPVoxeV6kE} zHs~(r)p0-;DvvHeA*c>HCa-#=9f~bJ8<-3U49is)`)u(k6vfT$P-OYQIC@DXcZ=5r zXfidvQIB+sPX^y`=N+jwq}1ySWW~4mfR+a+x+i+#O^MjCEoM|WJ zF$Z5qN%l(6<-HdKeSkXn3pPuf*0Z*b$Tv^_#p4?$^!-og|6C^@kAyxaF3!1Sw#c)`h;)2jxvFoeBx`uu$--ct00OyzUBn!#m^| zf%gCh)%0@MHh{is)Fb+KXuWJpbbf?3q=NU@T}$WD`4M}`P3Y9i2U13zDg|`!I^?&3 z^Q(k=3$!#GT`wkI7g{b^>i;(3^z(2O80xJKT(_{z!R7s zp;0zV2HZjEqEXr#zD>D4qz8BaK`cnrCAHZq*fK#j1L4J6LE zES+LozV(sr84E&28pqz}tgkI2S>QSX2fdC{|LVD(NbsHa2I00+8_ny6QJLmC)I+JET=S zFeTFsw*B}{iJlEojpEq|pG%Xb;`To~0{cKw(f1zsPQd7bDz0bYGsg4cd5^>*DT|xqhe_CifOd}tF6{De1BFEgxY=4aZ3weh|Jp|laSVC4C+$F8HN8^Ws0A!AaF&wc2(3YM{Z@la)+Jo78pKKAD z0gcyc4-ho`M9~9b#R2y&x(s;sioMlqeOp0nuwJ^@7u+S_9#5ZQSLepcER6{vR>rdD zX;uiCG=(?xpD%k!N3~>7UK_kYw4ZZ|E|`_6xCh2ybI^g{)F`9|?hG#wR2q)4Vkpah zJAfxC_;d1q7#T@7V+5~<`tz>>erc&K$HuXC;+(w|B|2l~YZIdb(b5L#DH#%|W92ex zgsT&zDLfRZt8hc_1>3PO>w6AzxRC6S?UN0;pk}W_;WZ}PV{%^XX=^OUQhjYxy`?IT z6`>ofhCuzrZ4kPzng?b*9aR0WXMgyA*${=TnwQKg6FT(keD>F1|HVFg){(E+8Fk{U zPNKvm{r6F({nK~$7iE*Ok;1`Fob!ZcyJ5kaBNS6jkqQh4Zvo%r7(`oYgDQBS>9tsZ zLdHfFQzP{e6cp0il%N97ZQ&S8mKzly@KiPJa4_y+4hO4O@XEZfIDVV54O%{jTzaF* zW3XZT5&d~kmI!%oE)p&0tSH`ltH0s*vPzW+bEO>S!I?bdQ347De{1d&7wb6d?EU(D5ERF)Hqkh`X1yJ!LLWKB$&mp-ym*h{iG>Ne68bCK?pmB;M! z9&nEn*sAyE+G=CP9I@MtMa64mU9>DtJW=R3m zB4^Q?!NZN4Z;+|QoYWG6hw1VHuS-y#Q6B+GAq?%|3$&nFw9de4s;7ISUcxld2wsQt zIt@Q={ab;hJLIeemsoLd;zDl$t%)93&TGlx*9Y8-qt3?0d^654yxp?hp@nL(;9-GQY_tB&qbjRg=X| zj3i)x9OfjYQcN;MR#UObq7$-8A*u~(6nDZ~!+;>Np0Zeio#a>pmMqfDZt%P1-V~!ztae?ioNyf31R^jYj)Tj{ zBXh1k-1i-msStVp`ZZGT#HQkonW^}cVp=J3nTovut}}{`9t0l0dTBMZZ(KaYHHRhbM=*!Fr+2X0Yg0Ob56SS2zid^dBcWV~biY|)FBd>=vdKL$% zSCTa4?qGeSI$MNKq;fU>8GK;uGQNjKQ39ohPyel>aa(W$_5(Tw%5p;RO>yv@kTHYw zl(Zyr8-cw6oL@=35QMgZ=0!^n+v_7k<=I(Qmglt~=FpELO%M~hF3csnUYVliGiHFS zrkDzf9He41=br$pgS+v9fXg1~9=L%Ua%q-41WJQMw>#ko4r4>I8Yyvb+j$R^PY-+M zN2syZ4}^J5{BlDs81>2_YayuB!if`Pi2ziFjb)EGJpy3!;C?6;g^U-RzY~68`mSkqF%~bFlQDwi1*4At z@-KR>{b#hvmTb_>HNHrCk01SPwEhcfVeq=yZKy9n`iqy@iw06 zc&lb1o*O4bUNlwDuCoQxvnL(YO1VhyW2^+-R!Cy}d>U@7T<912A%B6o{Q+}3$M_Qcd@J5l3M+mwYstK$G1u{9?nFG}_|Epr;4 zOqZDullA(q-Y4no5aYzFO_>>D@+byqA+xC1z2f5n{I5~9{`+lT#-68ic@B9j-nO9K zt;F-D?6_}<=h~=jL5FPH0&P%^r)UGLsF&jHXdo76Gm#B3D%A9O75+h|D( z`HFZYRtO{8$QBQcVr^74Af-6ANmV3H4lCj%0#RUrU|B#4D05=(o1^hCd?s08WB7v( ze*N+9fBo;4oF>mxhB~pOutLC6{q-8iPwP}yXRYUMn)ZjD1)Kcxxdpz9jzYvtqgZrB zuu8H))a|xOIp~t}`hM=V$Yy5HWr=@|=(ul(tUY4WG(5TzxoKLl`=HByPPzEFZ;t5c zm1?F`R4xW115Z`BJ%GxJ9MMw205?rk8&o7Ik}SG0do?swEh_iu0F$Is9h1~cyJ*u# z!8gqZArKbmV+Ed(S05HWy?5TXI zERB^uOK6AEsV>un0T+W(1gQAa^H0ZpkJI_w3`t{Tqw3D0=d?64-EnZMaBYa+;Jxo>4v)J6^Bui^;YLO2bUa z7Y{o4McSp+=OiZALwHa48%MxabZA$XPls7nohMk zqK#*KwT8Q2ZhX}#Tkn3HvgeTf7;b4)W%~C($5|V1J3Yv& z0W#2`pbH?wgVe;=-FHB{S(Voio#xj!cT@0D?&Sq}hVECL>H!&)U5vtNgC4H~uI)7H z5bor@8f^j28_M{X!@H6=CIr5J^1%tRbTWa;k73);brh3Akt8Y>s}|3=W{H3yA)rnpz0#B@exNRaq($fYP^(~K+{>G-e}t~0XE3(pn`x?PjQ z4hL_E-b$Z^U_Vw=>J9BoAd8x&%;gqGl|>~@1qnMCuwzCF?+h7$tP!+a6>w^2ToNXS zC36l3Yn1Qdn>9$WGB`ds51MsTc+ka#bhS5JcLLe}9qE8ab_8m}SMrmXow75tpMIuv zx$!;W{m@W0Nc$!29-Ei~@yK*qw*Fp*xz2WfIsA9>^mXE2<|r)1d0&w%ffXs=YPwsp zL->aRD*5YwxbmZY3zt&HGs}!8^l+kEQcdp=7Eym_gy79iZkez=1|`FnNgL=odd1{! z3DP`j{Hy6)kG{Yi!j+SCs-t9iLNL|ex>C*G6Jbx$G}50Rx%@de#AM3;{9(so zvVAf+Ze{}xPz>~&?WSVeN!q;p|9A0?l+VeX%8+e=t ztX4ti^^HC{e(8P79h=j(Oj_AW?sCKCTv$}_e`)^e+uj;Qt5|1W9NGjI8^LPRSoXY$ z6<8;Itud+95)*6|Ea%;eLaD=WxBrU=IYK&qHsdq3Ad8e?t(! z^`tAXMWBeXmaYtH6$0&e0w`>26dRbMydjr!bT+S@Z!D_={&Qv8n+9%)YhHaIK2so0 z0vhTemkbV)=i3`U;{(kwEZGe%#(`LY=Cu#o22YNs@5qVw<5v2P9?U6FbO23av*aM% z2ejhpa?~!W48vDB9%=r^gOWv+VUImaA`vw=fOzvpB$h|45Zrd84Kht$Iz`AYe)B)- z|NZf=fBO4hikDN&Qi{Y4L(9l7?Tn%4hWl3dr3bzy+wz1toj^2BY+DYS*_Is?1C(3a zsaQRI1VY3(%?7s>TNb-r)8&=IHU4-(jr>B$A%mncyO^WpU^9^#8}*k6kYZ7t z&Fd#R<=|Xn`;h~D6+79L=caJHC)nAQnm_$=;AQW+MmF6$QY7}>A8?d+NtntTkhk;k zb<3<$x|lqAvy;x@9#Xcxl@w7C3Hlm9krdTQ_wY-AcvHPi*#S}PbWnD#q;JpN1!_K3 zUITJFqm7mYb8{CqQJ&J?1 zUaSBlJ)(Dhz4S@IAmm|p^0s@N2pS~00effV39h>@<*eq!g&MjTd8Z`HLYD!#&?zV# z?xVArW86|YjWmVTaJuBF(femrE9>ZE+%&RT5hqy1sqh)qX0V+CYopNa&VT(`+1|SAdwWZOIkT^Ad;Lg^OX)#%zg@l>|#;hJBi7TfJ8rZLDtbb zzk(f)1su#40AIcz&5op9#$Z8B+@SKlzkqpQKACNfV&3H_R5jGz7pK- zHhBS58+E#Mf)rkws63`dUIQEX-J!)1#_~UOz_rlX-9oY>F#mz?u+O_ufud0M!ik+x zVz)W~>jr5cTrv9%6Mp7?aK4>XyfVh8*$i2qP|PWc)Ik%rWU-(?o(GjZ1)QzX3GzPo zI(h)8Ke5hdkE~y;{BN1``g$pB@IdSOEJ4RI$Q>W8NG zjya!Uccgt`noKWPS)pe)yKSBOHOmSNE2Rt7{M}yA`3`~DN2CYpHP`x{;8!b8(^b5D z?ls{?$?b5Rsz+Qc$rSZPb$g+!aW^W>KXHDwvQxHIvW&CFA1C2}Yqc^N6yP3k@Y!a0 z9W)bM5%-WPSS}hlI67va>%;EJlf&*tX@lU-IXyF4fY2kKt`Oan-JGLS<+%6I$)ZC6 zcfo*Y>5Qpoh)%U5D&22Y^q|XnzY`K9h^dH9_e+qs3-^i-Dto#3UN0~j4xb^*8T;b+ z*=&sT%DOEW$vJUaa_bw@?gq~{8Loh(tBc5iSH^H%Ff&{yDCQVNYN^;w!ALK%E4nzU zo!=}km0>;QuISC4=Y7}0ZhlKpG7nw$hm!dD)xn*!Q+PW)GCkIW^>LF$N4ec}fA~!q zY95|dLV`qbMUWUW=+Z`x^0aOm#b(!fA&zzeswf6s5bgM0+i7FEq@I6{PK}28P40%7 zFTngrczbrvk-vy%@527o_cvre{-p^@AFf$Bo7`XrB_|F7EH^94?WdUg6zKs5^@tqz z0lMhlY8E0F00^F9Tf{Q=oBV3To9A=G+E++g0u zDDJ9J82?82vG#d4NR9GFPtO(#oVFS_tYGzAOFjSdI)w0B?#Sy!KeJJkre(_Rnr4Y}I>j!J&PiaiQYa7QN z&uy>BO7}CbMSr#|b!Jn#;l$?2O2xJw1r<@n3}s7>@ODHROu$iDg~!F9i2(F)(Af;` z;SWu~>(sw~Ltt6Z<=@ z;w@O2ut_VwoAk4{O^(cOnlipkjyQ21wasjqJ5Mq76!`@6s2T+>0B6)gR;YSy-%{^H%7m&X9#77YK^Im7IC1IVW-|;XQ_O0L ztfXQQe}gWEfO@(N6tiy5VKwQae)Fs@z8F5NhO^tv?{!S_v=rKSMI4eQGeCSH%YhMQ zJ->mU1-h6(AWa(;;nkO!UO6U4@i5n;jkJZS8=zgLihL&Pp057PwO#QZK7~oteKBXm z*!}>t)%^7ltzp;?WBj*Bu$95GkzSy=fkNlqfhh!2_lP@nW!M+z4+a8^SKn;DYUQ_MPwq)@SG{eqoh7<33AO$*XdtX2? z`4q`TQ9G2qghT@-y}-m23f^m$Jk5NmJHZJ--um~nGzP}w1YN>T$b0q54Xt^PJ(0Dq zm%GQeZ-z!OMjP?MsbYZ@E7s(NZ!xAD`1f;eSTgomp}Nb4t|pyoH|KI_OYm7`erT>3 zDy^cBmbX8=E&Q7Bn(zd@dqPEWSbeEu`g+Fp5GxPQc`Fb;%hbN}vbQ*(^MUHO1)OYN z8qgA?&fXo0c`8hYs~^wbe+Vy=aP4bkc}nNd3Ti1!g_eCdD^!EV<$K?}rK z9k#ZxLh%U3VQd(E@l&i&{MsM01=}Z?7M*1)LXVN5S7xIFOgY0Aoi!AbNRb387K{F% zny(6C_gSK34z`gy3?rk>mM!=h?RiICyKvy$Z@QZhQU2CD=g4X&4yY8E&2c8hY^30O zjct~syJ;oKkOn^rZ7;=97ds`-!W77+U zeX-8mkPavmJc4V5B0QZXuA4QEi676b&oF!D`BcHD-eTg2!SHOGO3~G;7riDphiVL#MFTB!)pS1bR|@WIL2g*K4mr;l+lN9 z$km7+Tv=vWNbr11lM_$8m2!d83vNR-P@$kjuHGctKdlz3hzbSOvSrNQVft;Fd$1i2 zncz_dwQf9q^hbFUZE`&xeB1vtS;8*V>cn`-HgihWQp_5PBmxZpzh3sKxI0u`3Jf17Y?;jD1^_2SvfU-r0J%{zSyfBUE3VYxSk%i2hNSebBRpiY$wDVJgn zb`9N-W`p#^YU)q-xhG`0uQq;d<(tXWpA$XtzFPXm+{d1E^nF>E+@768!+)fePm>)2@?+?ppc ztu#6O^=%~im04*(qitAxRYNhE6xm3{*2~HuxOQQ>UZ%!9PP62S81&bWc34X<5j;)) zRzgK@fof0md9RfdQ~Nd+=!`ts_C5Rde=Q6Vn*bvF+gBRNnpXxOC1wE0q8KPW+C;@- zDLWFlLiO6=fNm&AL(%gJpDfX3zj}X&%-2hJ+0o!Z58hG#-4^k!ISV( zh|#O7`E`)kN8*?!rZOxMGI;7j0cQ5v=|WLzl#W4a-r67?X8>4b+T=RUUQ*3Zm+zF7 zM6Q<&cA7%$vh_i}Um@{~Fdm~8iG;36Ri z%4TQEduBizuJD=^KR13D{7?X5r|rJ5zEq9i=DhZe+^Ik0ny}{*=&}UD(oWkQ2Nfm5 z7o>EGSxb>MR4lGS*nQsysu<`L-;GR{V*@k-Y+@|S@D;;mD>M9|33$y87gLsf_xj6T ziGSZCQ3^`K*blFUY_PIActynDg&^O?O8kjfYs`Vh4inlb6-z8#T`N8!7JPxOBIHHN zqM9YxT9+VImwN5-Ro~{V3tR1RObG&dK@~tpi2rV$nl#H%1zu)3^GE;YWEeXUs zv4pRkZkF^#LD}>a`)t3*^eY(ti+KLE*{^J@2Ti{!|D}%!KFXwoDze3ieUU?EfGDAu zLW*dq*j=7|K4reR&ScT~v?1ZUPqvl86kw5{K?=U^8(Zma*TR|5D~w&rtEOZ@cBfE) z9%`-pp1bjRZ3r~?NP4+RjD2R@az^Zc0Lvlma?HvOzUKc{+Z0PZQ8w8?Cl20OQ5wBR zS4`^;kQ4 zgDffG*14I?%%9a4eot18RF>krt1N-q`eAF)Mv4Ij+_hA!1}fmrd7h=!^}<|knXF%; zMx&B3btUYywIQ`Zxo@U&HU1OgXBoe2mH>D*Lk@mr-O6OY_k+QgowFAm>}E;)%uT@t zdS3KS^nvX(_G;~aeH8~oB9$~sh;{-4h7Kq)@XBJ|!9a~7JK{ZK$}lNHz1pL~Cm)oQ z993_?5**LF9C*e7K(P*v5yB^vVlG6rzwF@?ELcM4beqz!Q$7lO@D&Q98>doi?}rb$d;tF-Z#=Rf+S~ImpTluj8uNkdd=!%zu+Goqhb;T;^hA>7 z#M_+;GoNE8#X#VvfQm%}p-~tTJ`nZCjvaL~c&Es`ltn+HJ2rT+rdRo4)EBz8iB zm0{R{uOq0>b_L{5n1NM>&aHL^(2jAQYn&5#bQ3vAWo#Gdkt)#*ql{0PhBQF9FQfN z;F-30q8$mN(C3hh4+0)s3MW8xsWi?nd#pYm%E%`&i{VrJKysE-zT?eWZ}N# zZ&c~>c>M(R&vJP^FCx(X{A+g5fvckEf%liLzU+}y+?-+^8E#H1{LXsEh;q~kN)WAD z`MO@FGfC5%=xRElh%J>F@Y*x08S*g6k){H*e5gdjeKF>FYeI}dl$2lL?F<`ZKLozm z%06@CAUF;mhLa(g@Q&f(iLvlhxNyXbg{Ptg%t*RqB@m>BeVHc@)VUOdomB-OcDNj-4I2C<37^GI`KbE2ZgkIG^Kl*YGKTXyW zzD)3d#0pTcB`>fDx-L?~uKBhAujg{XlRylNA*aOvKMAxJHwX06u>l8W)rsy$9tZC7 zCSfJi$ZnePfIJ4)r#umy;kSg}0L$LlII8Rm-hHw78$h7-Z^w?*kx&2 z&>i}TG@Gh@tAT%LRQeotpyU~&!uiA~u*J#BP?7}|EiXIwNypbOOFP4%Q-}8nlkR_< z#E8y>bibxo9VLm9CgFMrrey@z@KCSC8LQ36Q^XM+BM;()v)s_}aKZ;(Kd@Mw3~JyV za>|7*&Zkxu=Mtp^X&?g?jfL@X0u+Vrk?WEK=YUW@mb%82gP5jX(*xnMLAN$pM#Q?v zLAMmXk>5*F0)T5_g3ra_!`oj26Q!-+iae%YFUWjt&k z9CX9fj#)_>3+orHo7^K_q|2T8*fUXr$sfIH=+=Xpc;*p&bXVx+Db?Wz&??ed3+gpk zrhrxWFAUdl1Q=(_!}`DrK*dH$Y-)#v7YDdL zg--;>K}ziI+=n0#emVI1>?3p~|CsWrTMqqD^3-jEXSpODnw5>hdx}GTJ3dM=1aO1+_AZ9TDFV|C%I?RKn)M^^HYV zwk3m7!X7D&if#$*QWW$13A88A-Z}*(n`;8LOzBW&1@?$L<|c-0l_drnv>kKHL9M6P z=Pv(pK;`(@n$f^!^kA_6Np{%8e=xtumKWs}iQ}^{rZme}^NYk-IGyXCOW%nA|EIyd z6Wog~r)LA7nJ~NY^30RYJ#nwV|Lw7aYhBo+uw-`%`Mva_O=QsnvO5suy#o@Iu>3&S z9}7?p%s4N0-cmEtOxleLr_?r%^cR`Wk=Z}M7^mFAN#F$(c+6x`Ky(rqCfAmy}|z8e9Iqb;gxk2_E*+i7Ai9&|HrZEgT1 z{*CY*yhPp#Q8zQ{-W?8zVco+#{{lNIo_B?P<=Hb)u>NJQJV}Kfu2cpa7ES!qUpg;_ zmaPiE4U#tHuGyP_i!1K+Qhu?bUY4V&CD09y{Pnodj6=KeXaVuO$#lduc0i1fH=Ho= z#=AfIH;ch(oqb^fIq$+d;Rja6#v`38s6)e)h|4Qr=oeE)shT;#J=PA=@) z?6<;S5v44k$PU;o@r=Sc8S3LMikY=JU`f<%_eEJ6qj00_wB`(52V&qSpzQkQyG;-S z{7imE+yrXVC&a1D&36r^Gv&~{+9a$8lH)xjo34t|t42n_h*MzyNk@Rl{!g?2PIeHv z^@mQgB|m_Q#9>WW$&@mYA|FxFs6>zwaK}@J>Lq8O;91wfG=;Q>IOUss?!V2>ulc!S zIO8Jw*F3_y;JVdg9>48fZQE?>vH?1l_R$i}e(y?Y&a5L5_oOHmpAwKRT1Akf2W7aL zd9CzSW}D_Z|FRVCG&`_cr}eFj?Y0nH4k(Fn_-xxVOZFhbo^c#~4<|%S`1dz{9cf$Ez*0T1 zXZ{(wPf_Bb+bTN)RG-~mgKj9`nJ7UQ@rHaa(0pbGl+D8izIgd5GrQHf@O_#ga{2@% zDt+ouDx+DhSEcY$Kz6i~eiFKaZlqD?9P6adK({cIDA9>L)UEwU^n}5#9lfegQ6`D^ z8uZc?`mKxXQ(W^MbbCN@K>?wC6rwx!Ox+9LWP5JlIJexU&Sjage)_AM6K(T~T!OtW z?EA80AzS_OpbTc4_ge3Ab%i>S*Qi+HgA|mwM0G6O2525%6U71G93eWk+wcAPm#0*g zeYW=ad;3Y|BvNbT%9c_}$i@^<(Mf_vRfgaQUE&LEdAQ>obZeG3(lsG?d_t^Ooz^V( z9CX7F*=~YSv$CKqGxRF_=1bIrZYb-SJF|`1G6U6&+Zg0t8&p?MbHY4u__Cj4GsgDC zj`TDqYvQ`DZbj1KwpkX;-1u&H8Od|u*vt_tFjrB^3JR!?qpQAQ_$F$P>Qy)A9HF~k z+a|l_dv|UwUC%@I(H&t~o-OiJrbj#|T@&!o&!DXnmihLGug}A?wZVI&csJtMz&pXi zb8Ebc0`_mR*!%5$7RO?4;z+e^=}oPM8Ui=t0upIN9s|~Rjn)uY-60*~U4b5>YLsLyD_fn*iipJ1w6)?&| z9;gN)vj?Y^`=a-}^CuU+VmXC;D}!z=;CkZ`jHS|RXHILg;kf4L8R#Qn*D*A>{4X-z< zS4{1tmq5w*Ven!wlJq#ZoGzZ<3Qn?7c#XF*xG!RpCPTVSvwq4}%_&WjXQj-b-K0qp z6wg1Ux#F8Hd8)YlW)8nELa)O9mR7$?*<hFO*+ z#h-OsNijDXT-V(hl9N`8SS_W5)MOPEy-%|n!k zVVTi1O1Xg|DL@Sck)zGvZ`6e7YC?Kwv!1&VT;$EFnvg-a3cvMUO;9K?o+~ED>2}Ce z+A|2J<#K|@_on~cww!^bLf}5oN*Bs%#Kp4;1MiEgJu*G}7_742r*2fI(`!i=82EwN zb#$rk#o!IE8?+bry`Dz#ZfQB)E2x$An(jbBX`}j(-x&Qxw(o-DybOEWh7Q?A z9@}|N^WwTO^!Yyw|K-10tcggp?)&7N3tN+WR@S78Qg%?}D##nkPeZ$O7OX6v!4YV_ z>o(_Avkv;%(>ltB4*qko{!m0r;RGLX;o_&wT{t67e3S;{D z%=^OW6_{{1^wEc*7NlhcWUVJf+&mu_E-*e}h07XB2?X$asOSUgW_gLoq|kf+jm~L< z;pu@_BC#8;dRndqYDQBTth`$~s~2j6tEa8zXSm-Wi@|Zh6cCEN< zxdcleiawxh=53*_PjT9BwMUZR6VKMDLaGx`P&>~7Gxy@}aQGee&hzkR#^aciS($kI zIojms4p3`0hc9h$7rMkLsy=_^S%dT83^=8??I&+r9iqa)uD%T3S zfdsKS{1~_&C#S`MX5%42S=2S(lmILOe8TMZA7HXHz~PL1-hC0E@#O<SOXCla+0=Lj#@u8HFeCA&r)`%R)h-|AjyS+u@$_=hJ77_^%M?kVocdxA108Pc5ayzu=&OEd;87O?k+_R{824BD!wU4hMSUIo7KE(NA7 z4BBJR9h1mAOIjD|4!>p4;^BSqW9b#&p<6b-`TmEaEimK^r;#kq^UsY6TbJXEMYgpV zuZXQ-sococWcy-JCr*$iJ1ROG(kbirYz{$}1&SF#Vi>m)xuB7g1>m{{{J^(`xNc>D zAzEcx!La++?C4-A*q|F|?D(Db%W;Pz6Wb#O1+Fp&rXf7DeI2#+b2xrRg_dU@I+xuS zV%syt1wStAL$kDgHA?T38ZpjtA%)w&EX~b1Y>kx9eqW9}V;3eI`GIzh)mQsZ)XUDc zhpq?6NrOracq6&I{)lyg>IIKNEBs&!Z|=27s&)`(Ji7likv!s8Wgfzb)EUZ7OrxuSuAv+C&K}dMm?q>_7+=N@T=rZ38ug3~7Bm&x6k{as^ zXOQ}sJ_h=9$Bl*F9xx3zD2^$Qc`+h4zj-RiR;u1*qj4amHf+N6lyWmgHX?UXc|>AJ zG1(dc(YSc|*^twcI0>dCV8+{kgtgZfk_Hy8Az*@Nmo9_y@#p6eUi`(q_zuof{`vgB zcJ5g&Oa_<(s@I`cgAF5oU{XSfc_ez*D>h1r@eNF(J#w|Wy4 zIG95vx@*48a-%R2x*N*FJL$OK!9chPvaHyskS6PwVS_1Tp(L0T#hlTg8>*`ui>!!h z7vpn1;(O9AQ(Oi$=Jtt!&jQO1Xj}%cyAFQtz5!%8EOGK^xg1*aMxBpLFioy2mp& z+UA5@*0W*B@}$ne<{2o)yFGUU2m&J=Sq@zze(Kf}+C6J?m<~xU@D3*tQjIi0_DuJ1 zGceK^&N#>EMW6VS8)fRT?2262t6|v{8HKpk+@jZc)X`hidxEQ|Em3*=yRyejt#1z~ zfj(8NpN^fmm&K1~Iy1~-HTSLxrgu(DVr>SS1w7CSh0 z!xb^ZO_9y96%xn#%kPVmJ{gNaabd);7!;E#7x>0dAMW2zPI>`hW+xQ*-4fT%+Y%NF zqHeG{mHJ|BAgU8P9tk7P*y!Lf;!qAfxp;4L18vzD-T#Mx<0O{bF4~2Yei>Gt=~_y; zh9aw}XiH67DQs_Q#Mfx(EWp~iGozry8uLWh3yLv+5XS+cf1&iSwHYq!p|V&3luarG z+K*!Bfi-A5g}W3xypn=M8n(|m$4emzylyP%PEz@r| zh@1wOkq55*bvPAeSxP+bHKvkEZcB*^Z(UlgU~z;}9->G+By@Nef_v%guuRVZy8Nvk zCOHx`{i15UZ_B#pcJNAR?8Voskl7K6l)__OuEfvEm%3pphfP9v$B2& z-6n>l(~rPeT`APdj?mkv^sj#W-R_0&=e-;E_WN<)-}#Le?OE@;)2`7Eyztoy4`87o z%LJ|3v7paxBn{akEKgdcOyC#@y|yhHB+)z3ZB>+{x6Z};mXE!uvvR1oXcKbIPGY&(w+}blZf6pU6U?(U>eAbyK*+Efo_Iu}Fc4?dqKP~x| zL3N5=7Fh4ED+o&oCBANSQ@X)H^U=Quzc?3fHdJaAjD2U zJYxf#2?iXS9S|c)Aaot|Z}qTPj=$*6{(-CC!)OC++8Z~2kpar&GaAR(o&q|S_J(K6oEEl*ucOVB}WQU7k!KvMXZBpJb%Bq=zZdTZ{`lLgBy`Ngk)tNVZN{pY{`;g{0olyWIWVx9#7+iW}Od~;JEiMCjw2O)n} z5aUbc*Th&=*7Z`#TNL>esi{(#18-iIVi#c-jch)+ZPwu_Jru-&8DJjK8SXV+b+XIS z12fviSax@Te*m)2W?JZGIR^G{L-ncR_7vDuVY(SB<8o&<0D)T@q@J;-5J%4R+&d$i zh63P1UJbvN9Q8?(v@toFbh;dhW$-hgb&;PDhU1_f17fl~0^M`ms^e;!FKP$|9UtsR zLBd8oaAf6+@FJ25V`G9YJ?Olwi@v`Q8=h=^cDteVEZ7`<#tNkD_Q_54z3^XKK)2^T zFFEOU;iWI$3Sf^YWj{skQPJko{&adJ_)MMjX}=C2Oel{?HR<#qt&CoEknZqlksk)Br)WHI$Z+r-ZTU6?_-w9+X7>Q%s71? zBTzt5NY)BI6}}>soBom`@7OA7aLIMK@Jh&{yYZB{Br2eL#O)zX{C4r)$k?!J3T%(o ztM;j@7wqr?{&I42_JQCdG>Xk!WrprQ1NINAEVlDOUl}tW8+ldgkq7DmVl7q=%E~O* z^^0R|nq41h*G)^m|C2ppi{*;neC6jP!G$eXftAy~l~SfrWCIm_I2bXI4N9af-W}dO zP$GVx9MJS?TBFWE)i!dDUS*EZ_r;HZ!@==s3AV49^Nh0%mwlP3 zQt#wvgKBMycI($KLgiMgwuYa{ZwhG^Us!Nf49#tt{DAYkE^!?Y2;Unw?NZM8a5NBD z#^Z*Mzx>6c&BwCLw7qfecO=Dy-4H0A7}iOfMJa)zRgaytkaLKH(rsufLR!{-pK=B4 zTvm8oC3(En-gsKC+6#0*+q|3QIpB%b0FQJwC>3P;Iy|EM%%pIJ$C3*(bkE-H*D}xi-&E*$Z4#cAl+(?yyWZiqOIq(&4jf}S^b%-~M1-`A7n?74 zfH)QiY@=}k2vzCz9?#ZYeMPpyERO6MIz!L}AX>&7fOy+h-wx&p(;5Y&mRo)Eq}9`| zgRpv!c=0I$ZAyykA&dJIM}oWPbP49eoZk4x9aH0Db>0a(s<74Ka@hnYi%3|TES*lD zrNanUBi#(ENI6{{f?@y=cYQ=7hx>SJRlay@InDJKk2(E_ll_>G{F}-*Z2eR&_QQoY z!Ysj)VhEsNCIPhp+hk=yNd!2<3nK>j7$?K{FLIDUq9K+Hn3MsWzx_o>a6;2KZGZ9a zIyAk0!dTW97p@;?SzmeuU5cGv=b5-kLr2RenKMy;xyF=2>KYSL=<4cKxTYKuYzr`G z^ZEN1CWJX1Dx*M&2`f00&iub^8?;H&EV~&^LVPXRI#STgg|PstBg1w$J1HeJFy&Iw zm9XnI3NOs+m0wWXVQE=J>ZpVK z1L}VNCJ#&?VsCSw!Vv7NS$@v@YaalfGk}vWu>)X)klBQrKGUUhEJoya)!KSeIFb=@ zVTa?Gl@ZxbDXS<_K}Ban2>?*+kQ?FIw3)m(m(HCEwfiOy7ApF&Q68zIu&M=QvdqNE zCdwdFmw=J(_Quk1lcdCb9@cgy309KJQZsilL|mHXH6g1(>+!+@P-$_DhNFFu`rLxp zb>Ik;u|w)RS=Dk|wGo%KDJ=RS$H~s{Q<}%pHrYpBx_bX7eqG@Q)F@Pi4lAbcF)w3M z|0ow5O|qRJ=TI)L)SR7C@|uOzzXSBpo1hBC%<^<8GRb{^u(|s_$CL_&9n=x}AXFAK zgc}s7n`Wlw9;4ZE242jd8XIcZ@iKx5nz-uPoV@~z37Wa-2aRO83rA(Mtjx`NO1X|A zNmO(xb9dSr2~zg2pHeyHiQvMl9`O@l%e)xiVqYWa6^wDBxlV`uz~DOa*zV^9465_j=GETPCwY;3I&CMZ$pRoCMgFGN}6 zH356(fuq_#tuMkzuK@v%>S=pG&cmSUmo>R>2L{s^5IBPZM7;*9cAMO@L9+oI)@Jz~ z&)VrdAbSQ~s}+8Gf!CwQu^h_w;=msN)R92oWJ)F+skstsOFsCDm=YGM!V}`dAo2qK ztQc#jQv#ahSaF*Y@G+cQ3++l3ifu|HEFDK@g&UL{36|%N6|>>sWm}*06#)l}2aC`Y|Xajw32yZa4tr`7`O@t94(Fkc@?E+l3=3EL_{!8f=cw zm!wFlqEIQbL-~l8$7>+KCn&(<5`G6$BfifxkalrO00Kx?%+KHf+Y4Mo@x%P79|a}z z8=Y9I!~rZLtw|1s$QQ#sHXJYavN$F`Z6}LJGB+L|7xq?*t=yPQN(uX)%~W))xCjbe zkwl?IRW2^&>k=ghg(mstuw8*AuUC5{O7Ixe9;UZL=&gfklpcmi%Q)DU(P89^xpdZ1 z$>odR{DTD}64{R;$!!;QUK6dFdSWIjpHO4~`dNd!-pJ!M(!J7Z!Dn-i(07E#mDu-S z&>jxP{;!mPmiG)lNTja4qca~I`-A3qPTnXlA+SFge|1N{3D4Wvk5KC0*CPQ4YLLvNE|zpVMr zo!>QRv0N@C;KG6q#pWrs1mw;A+C$7&W4>nj$yMdVRd^I|I zs2@K4?YGBLj^M&(ilrO@Yu=g^@lcy|cSb!V70-)nLOPffevSq^$)FNn(c_~l^X;UM z(ACfYHV)`L26blQ5OP zQl#fM$#Xr*y&Ke)EOH}tPMU)GB zzbxc+$dQYAL)}MShc&r$S11x^>r$d>LQ?qKc-g+6$#Z8y%mxopQ=~<;Gd!10lOyI6ltNNH}IB- z@|bK8aE=o+dnXBcpi*Hs>0=Tln1#kF$z*=+)E?%%Om`Jng<4cef|8jfGqDCHX1;FM zi~(=x`T!B<&+BqQkp!zBP?)HL`PlcVA~{mG-m8huoNm5gZZ|YjCic)Psouyg4Jy(P zc&+v}H{xFvUz3+dbc81fkeaBIUN<>8qAw!G!`?O$JHf?jenxS@F39be@Q2rB7I2C0 zh<{CzUKwx|S%E8qQf{S48mM(j@>C@=ckp*WYcjCBcQA;TLAOlJJ`Lhxo2GE#5|M6u zSd6Gg+!K197e9Gtc$X97HJ?w@=pgd><4!v`boRibf3-kFBwF`216-;%pF-(V-xG2a^RHExGi7R~Ox(O5sDU$SA~0m<$0n zkHK9nJdFHemvXl6()mW=JrCUh^;Pk8XkqR#QNWnS!4qw=U4ffG`v`mBo8`KUuy_WG zuW+<}QqMb~O7q#PsSy?_w(%~CbyuXl@*es+5MJp662fBUZkT(UHhLk)0%|8gqBBW= zv1oL0fFqk&Pc^CQ@$=*ZrddCiX<9XT1tl2vn&mU zV1T0MqWGdXM)ZiTfx6BN-e;i>^@vB9Ae&QMaAfnSQ-96>(uZ%qX0btm|5?3=Y;<88 zRAFU{Ps;0p~Wx zies^;;XbJm<2G5B=A+jnD6x|AI^PKVURkh?)IygblwBfo26Ayb-&~9uBu;~ed02wjDu6AJ8 z$f&V1g0;a@j)D{?>`-^|*DSN8aduhkuuwK*1)W*e7YiI72Oso1<5}W?&_LDQ8a~<^ z=PfiN%o`VIjPSVAPH}?9gd4x{-t)4rJ;@Q};&e%;tVn?MxdUDY`8{-&rc7<3oD>^X zp8yqVO~@duLWR^&K)8_l^MEt;!!6Ux1th~vhSR_?G%z=iY^iSkooxo`rOFPlq`SaEO60w{!T}hxbXHk-3l7XlroVb zAE5vzR!Uvx_kaihP$j3+1){r(vrgRZvd)fu^UwOY(=M>j|46~KpXL8GTW!I_%=8bF zNsbF+0)+F1r3Wi0WjRGksA%j{FNWZ_e)EK};6O*sDFBgq8zy zQ=H!=sSz9%+%)G&b=%eD;;JZUaD}=%yiJ!YLrx83YboXLRv^np?Yw7su*Wuc_5tS$ zJMrA3vHdAlPjdp!gpJ?*ed$=T&@Nm7#gc_a|HN#57Iy-(f6{BZQeK>KaLQawQiZNF!$=|sa1 zO(({4k8D3Oz^=S#IvsfZ*j+=zJj+6p&ipu^?4CqEwzB7kDJ5tI)l$*7!dDA=#2eKQ zB5wn=t6p_lyu~|Fvevs8G%QhV#VFhq96zNAt`tYzi`Y%h(5>yzH)7ldOFJw~)aa#?s*(j1-(-CrB#=F;(?vpL|sLbOSF z$LnsySSO3q1U%Qa+gu#uWNoH% zJ}?#{tyQ4Z^0#Zu2e{$97~D&mggd+nBaYLxf<Bp!aFBkbXd|F;@tTd?PNIHwsV>c*EJ@`#DeIFmPKjVidl!rpc`@7AUzZ&4cq;! zp_HpB5>G`RN5fv+aWF7;V649_IwriT&QmKYH2s!n{#Q%iy}NJ?)&KVWuR7M|*##UMUx?+`(cpvi z8BV~Vc3s=(8EG*k8*&vHq{4**DJ@o}q=8aGDSsUmolCCVfME<1N5#3~epfeDBWf0F+gN)OZK42R;RrqVQ~kd)-*0cn$=T4Zc`0V%G&uK$@`m zht+GP79c4;{AvqXd z@h|`OTZLWmcX z7~eNSnC_N1o2~!?${3)jM8W7mwOXJ7#`3hU@Xew?yd2rVL_J;c&>#^}1} zGvc(MJM)(Y=J{h;)<55(bdR1Wwqq7H~cv_ zAf-t)JkET1O{GBCEIF#G$#`8+jv=Bbs2fTnWu_sQK6{Pm^7e zmP2~;y0QrV0WcUoR7={Wi2|4Hr#`w!5I@C&zs?1rm&s;s_;cM&BLTY9VH>dmN(tma z*(g!nBtJqwls{#jDq6Ie-@0UyA~I+*c$cK-#n4p%b6UnT&fCg6C+(M>lf?rJH%Jcg zTi)1Bn#1?0cLXhgKI&bH)5Lj3ZgjTob8pC@i{9|T>Ti8@lWn!!v(+vxj0u(sy1vi> z&voui!oA_Q6}tlm=$+)WZO-y#+ka#+Bte;1?-K(zL*l}XjrXk#$qh=`Nf9HqyCMfrsqZfTm6R?;lTB~u zr!qs$XvneK#~`y1h+0J*g2u+gkgd~>Pg}0VJav{PW^S|mi*;tu?FR3VA_KT{Q<<|q z=Y7ulpu<%Isw6<=OyYv$C$FS%-Rl&d_d(9*Sh9wXguF&rCUNgqGIOQJ;@aSi${Rc* zP+C>WfXBC!Zjxtf45m}~X~#()KS8+|2SkErz3MoL56pjF&*K=a(C6l8vxkDbZ$&NZ{`H|`{QVlER$g;CfKIVY3Kj&X#0Loi}|Vg+PBY; zk6yAGJ1D6e=B;d|lp87duA?&qcjs^RNfsb4LkFYR>rR z?tX2#=lEg4P7Dma_D-8~+kx2^r|e%gOtvL+XW>k5mS+g6Uje9>j8N(3 zx^BR^O+_ng1AZ)VHM6FiUQ_DZ?oktRpU&m?34ygBThj&Xq*zOh3JeEfFNu8lh=g{> zs4To_Iyk<#CRn^^Oh;Vjgb)|5NSnARb+%;l@=^=2=ntfa^;D*8M` zGx7ra-sxW0B`kwx_eOF;8TWcw&@TTLZT0saz4^!Px3B!mlW%svi5ytF16TdaAMuw~ zEt1yN(V3cyv)4_A`pw|OYLpf!`>_G&<;r|75?sH~AZ~cG?cb9mp9s?&w_c3~)z7nd zjzBby!3ovWJF`!H@V*5+7ydDBF1hQ%;8|-G+FLeJxnv@FN<}v+&hhHi&BBe!7C=Wk z>EPiR_)Xp;Kgi#xY}KYhomWbLiDH$f4=9_t(*xti2ssmg1i3&2|5;Eo?=Y`Xt)H3( zu)0bTLoj5s#2o>3AlN7bdREi3*)%#rPo&3{Pu z8-@5;_?JC(!)Mk7RRTSTLEEKBaz8{mm!Ub&j}Gka&D^LhOuqJ0GZdYUgbU1BVFRpo9E)Ab#5AzizVbIJr5e zQE@*aJt#d$H!cq9XpDiQUijyY!)X>=*9n^R$)uhiSm2RC4ZK56xv-V`)C#wkC}kT( z3>ddThaE#LZL)Hr>+p>YN@Y%In&f+?FJ4>#?bDcMFp0x93(NhDmOIZ*DF;Q^LvQI- ztV^k5?sb1|i|oVgoFD_Y}tPOVVCVhM+`U7?Big&=!Wp zQZ@7*X#@G7hTfqrh|K5zbsK#Z>PQFpS@iO`XJu7BPBWQf1EnMvxCzdw$) ztpaSP70Ke3;&kD>ICOLk+v4q_ zl+c7xge+&hbgQxv@{WgPPuWi^g>WiznP)DJKmjAQ@S5d^6ia8G7lVNt@JxjEalOX8 zvjZN^LK^?R0=1bB_I*^5!_5^c#vxO=smcqpH+zG6==YD3Qs21X@^Eu92TQtYgs(<4 zXF}H8pL9&OKyTUarmQCEuZ+d2wz62ol(LW_c~mqqZ$1jeR#AvaO>a~H6^TipWeJta z+>m!LXF}3uG%C`RO^{805ZMcLdqpN|bvPJnG}^^qJP}KouQ}XKne2-LjFqG9lxe)b zcKz#sZL8HQQd_~&m{p)Ilx1nmn&^8$P_lZOb7uTS(`5%DUNnSLuCJX~{O)foHsrU> zKb=WFZ>;n6JQ<*bS0}!11RQ+_W?nz+Qf^x-HSr06F zxBzDQg8T)l?I6+3*6AH{p-F;YB);uoI+>@bqie+F5fy$A_Jb%HMR(Ge z>@N#SBJm7NPn+zP_#>uLu3Hy!Sc;Q4z6OWGOgqJ(uno#>Gj`zZqQAOZaB<{Ib$=(z zTo@PItZ&%;J1KI1+bN4 zb>miFtT?X%>y*lr`EK>SAukWVPd*ER`u2#c40=CT8I-iZy7p#alj4f>_ACsSBlQ-( zVmK5=AlHYri+8J+&C<1~I#j3vf7bg5XlS12)r4S=)*%oE!hC3(Y>Q9AZ2U7X@OwSi zh2*%m$d4&%r=OhF2uwoQkqS9e(b;fxy+&9!$ME9VzDQ&E{n>X-ziwe_s0DGE0a@!w zkqa+ZC#*13Ln-%BWDgZ>6y68ZyX(5EO->jPRskdz?LyNq|`o3}*g=uX9k zNF(X@EMsm(JPhp#Jph@`66ml3uFx9b3hfNX=Bv0#L$_}A?$Ruu)JXUH;4p}T17rtX zD0mQA;@#??^rsUb@$CGL;tVHSN?r9@k<3^Cv>`e8JX!C;0NP~*pd3mGj&TMRjh(|q zKwPn?K(tMXq+u5pv?|LZn&eLck-+n>2wZV1RbL27_au;&TkUdh%W|TVRE5c zbc7dv6kj~sBVzyx+mH@9&kZQ|1b>s-ZkS$?2oB5k${f1^nps?Ky1fTHAwG;*uO6>L z3jgO^)Ywij2Nc;p>eSyu9(Q>D+G0TVyyqn+-6M6#xNx*4-YQ!2h*I`bTumfk}pC^l|i>-3VaBID4zjMEqWkb8rTx7@J3Tdyeu6X`qX=QL_Fl0AUnJn@# zzh>%Fw?2i@tAP~rv1%-lXY?xZ7<8-lFo$Q-1EA(rb5BtuM&4YP4c4$*D&8;Yuh@1^ zxNKb)OIt@Jf0uX@!P379zx2uZqKq&A49+0JDIM?9jFB{^XR)ro0|;}H)W6=uePh?d=ge6KC`pF=F`SA1%33EAnw z4)Ad+2e_J2f`mdP6^*t7!$*zK+nvfF9O~(Q@qkxNNVReW=-f7h8x)zIeV~>L^0bVJ zFdd4NA1d;A83HV+#|S^FS7J%9xvmHE@3p~ufGr)j&3a8&Xcvu(8t%RwWqU@4l`l-q zGe-`xJ?vN+q26qwp>xB(*s5^2>3um+aT*uK700 zQCzNa{%XORDd6B}@+cIb{CQ=4XUmfl87q$UZC6Aw$rBXCc2As-W&b zZA`klUIsJOqS{G12+CKZ%%8c+A0uDrXFBSo9sVzeaTu7QeAj)0y&Oh#eUh864F!YRP=Fzij=n@%4*6lA18Qds=>nQtK-1mo&j=C zt6r;CVw*bJ7abNOV~Q1- zigk17t-dH^Y8E5JSPK>{W&3V%|71#%68~GF*anrsP)UJGKi%HQu;Bp9!W}4|Bjz~= zw$sl2cXptRutsrOjlvewMy7Ti zvX`at7Uhdp&C3=ox-oY@v>cRA?xh!{dh|y?SKQU$cqV&1%*iML!hOa$KEVkf6aTbC z{^wu|K+1YbV@d8SQwwp}3Qm=j67rIzRCHW$Hf;>W4uhTH-QFL;226+e((`t!+dXo^ z>S$dtzg?V8pHLZvg?xkGQw2s?%pB;tY*XUk1XOHhPB#j(G)CcWwdt8tWS6R65Hkgu z#KR7S_0W!6&qj$eRwKj_POP`DPFb?)zn3%H;advK`hUF)Dhw^!hu_M1d-YdawK@Eo zWV`f#oY^6qoc_wmZ`f8UydrxXmb!+u9-o2zxTHxH^h5PZc37rlgA5BgxBAxexsvpS55;zuIF9RK;AGUdr#3VSioX{L8|GfTgpp`JN*kiYm{g zqAIE)vP-j6bVGhlb{$5|($s_uy!!lFUyk|X-)*JWx!4&OF5+YnW$&biq@!9?x>|9O zxY;`i$~ATMymb03>4muR8Lnkw=mb z9t(1_n3%$czxaT}a5FJ3j1j$+iAkcA2^3j{;y|zu?iQ>IIU|PcHP(_`ozXhO=>T~4 z>$h*!dGhTWZ!9)$NF5MJbA z$?f%dJq%Xwn#Z{>#pfLj7Wr( zfM8mKd#9|6M%ASMF&4vcb?1q{;*S=`;~%b_CZx}W9gk$IV(X<7l`#{^6Dm4U(hCW@ zyuibSSC;v zWzd=lIsStE9h(CT)IENq;v=t9noc3!f~36r0#liz?){|Ky4=}Cpq+Aj<^jft%5MzI9=XHrBm4yw~#k>4mM@BlK#&YLAUUH;DP3+Y6V;U&5m65%NJ) z>$e9!vyUjx1&=-*sfR!evyj!CTen`52 zJY*5;f%jbN) zCUD2{NV_7(S7md;@`SX+E}B>%Q|6X&lB{rH$YfhVW&@>6p-3_nok`XT8vGLl2d1(8 zZ(QA4`wNxEkKNORT5iaA@Afa=n_xkPW_#tI$P+hi1GsFxCxOC*VS$+hO1X+6E2wCb zmhxNZXv9Lp$A5u7mwA?6ND%wE2^!af)lt1_Tlq+rEx~7zifnx=m0ztYk2p31_P+Tb zUArP|^ZfD%qYzJL)4G&^-0|H6b6Z5lZ@#&`>4Pz}-#09#qTy#3KP88_1?pWml47(n z6)lvqnIfmC=(3;_pz(M}aZ;7x-T^!t*eQ_6%K*Bt#i)tgBwQ-(gf0S^d}r8lud+~_ z?$coTyIwWG6p0f-Vi6Qyka2U^Q z26B|`njH55(1_@l4S3;kz6AfPGXif=O2A4FU5$F1vcfMV06RrN86|W2KJ_VS%&bF# zLjC}wJL|0vSTXhJygtQgzcd+~z%qf|aB`F1XY!kK2HlX0x74>IJe!UMS(+HpQFkbz zt?)}wb^<$mEjg#oaUTrS>4BaoUxdFP$12uO-Uc0NXlhQ5sG*1c1&0LiN52~TS=W8< zw|w{Eo(PLY`(b541}UFJ;LRWAH5{arbrh+F`bK&U6F<41ta@$Pl=28Di{{l%$EqR} zBade`&FEkfA~(;+Fsl(71Mkr%Ra<=WKrHrvx{u#G!(3X6&u;b3@W|Bk`Xb{XhF4aE ztq4P^4fIF4=p=#p&UKU9J^GmabF(xz=ZxK}b<&$}=ljim8@J`hWgX4F4cC%vl?q(e zWnz&~=oRjS&ehYr2kJpElh{?X*z_Ox3GMX2XmAa-=8H}LfuCcCBs(ZRzcL!se*Ul% z54fSWVDgW@@!v~x+bp1jjU%a`F}5@OsiMLUJJKK>h5aO0(c4e3`k_UhsTrJ$WtpZn z1LXKd^72k7-vm;QKIpE7^fs7yqYgXb0=CcH4v%C)$Z}hK5 z1o~)z$<1?m@!ccgyR)Y16P0B((;k0sKgo3AAWyB;GFwV1izre+MK2e0@KVFhGfyEM zu~e``w2B=@>&C(f%X-!Y zj#q)K01{x(UXL3#Omt$)0MaTOO@z=bs_`YB;s_Xf=X=ceu>h?-F!`Sv z)+u}mCF`eu%3#;Bli%af zf5RhvFOECM?ZtK3P4wLD#jCz$abAPtX8#juaAD`Q!^%dSrIa62qzSlFzkXS|%j>S^ zg#~rN{luWnib(d_$*&8}Q|VO=set7Y&Klu3d*@~OH| zkS-}#PfDkc z(@pXR@R^Ar*Fu3FQFdB>olEjc+A*@FAGpeDg;vjzuD8lWf??&iWd#njqCdcLMQsh%b zc|<+$ws6pG=r+CTsbY^il%^|kVSU4g`pG7O9N>CQEV<6V1HtWrz|Z9Myi@8k>Kuqr z;@~<=R-`gU;T7p|lJR}>AZ^SM+VqJicR-jv@&h$C}k2w5~%3BsKlsr z`V-HMQAXrQ`;;}Z-VZXxWoOT(WTk7G`DX<~$AcDF4Z6$|Urc9P^? zAS&}UXk$c$5d-|~(h@$940^967*IPUJtd8GU*eNP8WoV0oZqNOP~M-@3vn*J>P$$t z!zc#XzCwGaEJ_uI@b|Ms$1?4ayphHt&9NodZWroyhT zj$S|IVsJe$e%JGIJ(?UZM*F;3-~x^3jpOKR+@MkQ_Pwv!dLA!TbeF{&srT;yrO*cV z)!tYDbC$%*&H4gP4Y9CHrJYD&dD?lW*^x3~u=MPYE&gc7_rLQU(&obcD9AkyE63=j zls72SNkt=HD{glZC7mYPx?MAF3tl$1LX5!i(ZS$m zM6be14IOHV9w2$rTzZ>Uw@G<~#@O9rtiie;R0NqO4CJLlZ~iWK;O3o~=5q=}jj%`u zg&-5j4JiBSgp#alzAJg`jUw)Ld#MAkEYmE%uXqm&(4BZP`4{N!}+Bin8TF8!CT>zNUj-iF=M zG_un>PgTZW@7Xf%iQsodo|k8LC^Dp33vzukC3WQYws*8>v7c#EaFfvX)+HJYcj71< zgE!VwHu7(d0ocf6kL9f2Pv6~a+nMRI1vMw|liTlv=E%F1w>n8VzSNl9=Zt%?Gr_+V9cvwg?!crN$FAKy9u8Ob8adeby@Aqti zzh*bD6l!Sdd28ljX@*{PDX>VqMD!_>I@vBGWfveFyolL@(XtI0*T%w?=fYJIEX2>~ z?cDlOqGY>QrA+r(XnTmMA_~QAQTy1g$+YK-weKGLVPW5Rj=JmD@4u00OZ@zbI4~@f z(96De733iQat#u(8LA%f0DadFxoRHI@0l{dujiHeV#9dP6x0M9H;mYi4zuNO)a<`# zoJ+S_{%pI&X4$o3iNHEeeuo5$bV=?; z;W@Ly>uFGUEeL$f;05eAfbH_hSv@prq3JThaOZkLtXHAe$_#j8t14`p%lyZu&*blu(ufl{oHLPfKN@HGKhAl!;uie@=- zZWYJ~qHK$MIe(>xBWI7}{5;>hIB>c+&5!GD48HT*&HrJWgmzi8!jc`w zlFU=?2mN%Z9>vgVf_=I*^oG}CA^2O&KOtVmYxl?p8MXbIUeYez4)U#a^mea2)ze^H zN}ZlD7(IX;h}IE?;s)S@$M?rR!eyyt7$b+N|z}&HkxOIep5zI>gA=ed61vxH!9KdeXVSa55JYPR)C8#kQs2WtSBet^o|crZNLw`!omCW^EBuN)I~}u{#?`9Z2V# zVp1LvM=mS?DG_9c*b7;~rL((;DFd0x)IjUng$1{jz$B0WM0@=Ny{2Ay8uTfi4e6A% z0S`cfyDnv3l?s=wGTtQ~h6GIR*duz6r$N!;*WyF^VS-oHGf7y-OxgTj5yQ2g z6wA5h>Y35AzH@q#Z5G;PEfPy!xj>X5xFIjs7_?`_b&5t1BRA_cZkjO|YS6BjglB8S z_nFV+sfq52!l#|EOBXrHePkiyr zO82U$C6xKNPg)r9As#w%2hTEZqg!oQ26NzPZWy_{_?;kIa&ec%2n$^}YFg`6RZ-mn z2s~qR>{GY1Q@Z`iBidwUF4=KF!>|eD0E}TzIscFoFed(3>Hp}TED^bXI7xk-eC)zC zc6Y2oh1V$M6^gVE@yPc@R7L4k1?r;kR3^F%mz{tc?eRNJ7?Ci4v6KB-#W!{z8yHD&mW4D^`L$g`mMj( zl0tI{>bh?Fov=_yW2)e`JJwqR!4#4&mO~FJm0y@`_juwz`F=&S_?0`?!b_zb9c`(yhK18QoQq z%3Sfi;)@*}so?XjpWGKUZeZewCCBRzu-KauD_Uq<4muX9w{%HQ&<7O@b*tX|#J5ST zO9qmRDv;yNmsI!_`Si{JDKjimI2>F^<30o1ZZ`6&)OGX?c}AFCmCVnb+E0+07f2}s z&d{F)=~YeQ?fg1H?o3C|0@p8_bz8vlrH%G9r{&&tSL>LSPuJO&P;#*-F5Dr+QVF_G z)e@G+EAxG-==aG3$?2F`+4SMy9B@e-4c!+*%T92=cpwMg5i~ol+P3eV3o2X}@1)Sm zW6@JeP^#ESMW+NLD4X0@&sp_)E_f~AqF@TLpI{^6b-q!h%bae6){Lac#9*kt7C%3} z;lRnT6ChSb&w(G$I@Fi`&%qvBCzT68T(>HVu(-7M-{_n+L_+te`=?M5tZU&Rne9`> z5&AsACGaD!bK_}4xM72JDjj}d``0wTeaEta`5SW> z^eJ)+E{?tP?j^F;g<~cqRv6euDWM8-3l+WVo2FvNYAAN>9Ib~0YughI{CLhmF~-xS}cm_iGHQZ9bE*v}orPyKAKk1Zm z3q>|j(Vf%!$w7WLy++>6Bzb4@k(=O%v%tq_*gh z515KhszAxH3hJ$)M%zS- zfAJGJbGE_re$Vqh&+|Ud_bXfz(QY#z)%>ZC{$w3KzO*Tmwcz&^{ab(foylbUzVVlH z$fpiGxvVu~SAR&cKrYw^4h4-$9X)g_1UAd5Rz{aDx)F2~2nIV~$6knZz}?(7s}oXX zII+|Lm)~VU#>~CKBnFdd`@D|_Y*yrH@E$`>Ws9R0V)0qJbEVss+4cS_7S^lF{iltID{k%v{3 zSU!m)hVaK#hwVv~!bfx0&ToLgDcK_%)(i09W|#hHW%8!brl8FWD_o9;r$A9$yPvLC zl@jt;GL|=>MqtZ+n@|$ND}|r)`p(OiGnT{FUvtbs4bn|Px7@MR8b%${`?eAN9Bo6M zw#O<+JnbfSIL;3eZ(n);MazK5VIjd0;;doTFxaBE(yJg;2OMjvBxa+iSBTQbx@gy% z;*|kIzW5V)Nbx87#P)P=hX>UvgjiaK;fw9~lNUm!$o$G)d)AY1I1>#o0=1}%{BO*G z^&MQE*%okFx_>V6-e!3wF^x);$i4w((df(!Iq3~8V7J3=ho#ZWMDfl$eNe zaG56uNd3y^6arfUvU=FGczRsWv4qm&&$h!7FX+5*+3l@mmi<0_CM*Zews3R@wTOql ze-->7r0HKK0YW%I9pqK+lf$8VK~wSKBo!;-t4Lv^6AuDmiJElPIUyGKw6aA}^9SuauAs*Tb`ni|T!(Qu?9bMzFEh zsnzpt*yp4|WGY}E^xX%#-MYcqmj$I#ol!`>mA>zm9b(*tin*20?r&^~0aJg={UYgS z3>{g>KY)!`wirh1-^<4~{OKDL9U{1%-zYh3J;=qi--z%v!RL2-kHwK){4#S6Jnz() zy#=Kd3v1H_RAhlmnRH`Bt$Y{A!RWQ?JuZYKlhs~jZcEbTNuo9oy1OnfRwc^P1-afA zVM@yNx))wevW0qWgX^J%>%!CM6{Jjr?(5T{2J z4zbNtXw^U&PU3+g>u+jBG}|=&;=od}?YWt44w}t2xfGjCkzJtN%50u_ zMv_XBWmh8}3R=V~L>t4)!xA-vD!q0?1b*rOf|aXtPlG~Lyr@@x&b!;N7+EzxaY1EJ zY|uGPzk8ldi7zdOkq1^rJ64uAv*G8ZpSFKtS(f6kfe!}_J{H>#I<1_&Env_ogKQMx zoRcRxMlcw%xwG)JXK(F5KkYuYKF(`udEv#=l4Ea~OvudzakI%y2VNIiZMH7-1;zGJ zq!){=2USThPn?;4$5q!L&6Av&er9^R_zcue*9mk9;k{4`oeNdZX+E(6T_%YIoee_~ zG*d@X+^>h?d=e|`p>w=j+;q8;6aKe=wnA4G)*{A=6}U`RNN4->hQsR8iaD6v9dg1` z+ZdGJZC2C?;I5d_t5zVH(^XBpsE3q;Tw*LJ^Xk&*2f|Bkb%H3#PInaWy(6llH%-4w z>r$p4R~_{%ks%>(o}?8z>+&R5H4Q=Cbm`m#w;YWwU2xmEl0GZl?jG%%M(bH!$$~>b z502+O9~5vew0Orp_=*X8&ELA2P4+o3_Rg72NXIA^;-N|^(n#cPtj0$DF{En88U-l$ zZ&vhpu2TZvAP@{nTYonE>WkLAgR{O>A2G#r>yFh6YQb4M~7w$mI(nb9MchNL{N^?Jiok=04Dfkn zq0cV2Ht$Eig~D9l)l*UZCsuIV^GZ;u>X;B0!w=0U6y}HTiHMonq%HB=?~8lRYwa+H z+6H3dhZ5J7zxt}U{WTM43T6rJkZOLQao`fFt7f2SpxDzCq%b3~69^l#(?d{(D#>N5 zBFhigBtKIc6e+O0`xMETsl&+;37m$U(nVzpl3X%npkZIOAX9d5;UlGPyZHQqZETdp zAR`|Wk{=Q;ijv%h^-i1&hnz6~nMS7tc7vyemtx-%ejTo9A!~ga{Sh?NCPe5eVIF-H zTr5ZfAq$?cvkjzdZ^0UFZd(>+g48cRnqNSQ9e6rBYX*`Uimjx`5h@a=q$W2|Fp(BY zv*AXubB$Wa>?P|52w7Q-N=8qu%_}dvU_j$&*sz3k)mAb;76d0tWT% z*(X1I)=5r>-Ph}$m2|0R{UTH)K_NRx78`5g@?@B$!Kmw)s9SkO8UtOelalM>ixK1G zF<}Ti^#;5!@`AWT^DmaMJfAeN1N$@_kw5mn;>6RAp z&lU+ESCzJiy9D`yY~fAvz`Y<89iCVayg%uIy-)aCApK3RFpmKH1L+Mv>CvFf(&(`i z47oS)pg~W%SHc-bvGR)d8I^$BU4a9%-+P{~|2^sZ_Qr9(RP7$!~%7XTHE; zBY%(lzkFuN4dbvA7YEx5R_W!?Sw1E6u|vKzJc(&m+y`V$iV3o?1y))li>vJak(Wi9 z^25On4@-B3k419eg)0u{2iH9F7lTG07J_4%rpCK7v`cvcVx#|!{(LK8g!7sGu`}Wu zrv4|CKq z^W771NN`=gI;2s&4O-&xY@UJdi05a`CXcbV42G?_{kYHJWkRT^1E&wVo1*F;)$F)P zHt|c;IB-Qxu~`hhi(=C$vV)2ooViU?;GVgt0m!4Wh4o^~&rRC%+9akT=+MkVP*Ro# zVfFBt_Qewp_&6~&Zmd{);q{-tpY*JgdSq!5B3cD;vM5OFgU8B71+JUE)vHlG=(N@A zQedMRIXtj;x<#Saw$S^1>*$Rn2F|ztsE>2ltsT7apU2OXtncf4)5By+nqR*7Hc52g z1YL=lamb`ta9UHT$o$#GvO%XUY&JdUlu5@igHCs(?TQZ9GY|1|LFC|kq{;gC+i#?db2T}WsdTS)J@ z^oC=j7#2(pxaLV71w+hmPl=1Vb{v_fKYdGIaO7wEjR5!4f2Q8O_l_+8iscl;VabzT zxh4mV1XUf;|4di>O01wsyOk`{JQD48ixpfTRj<7B-fz!q2c5Pl%Egs*rt2fopi{Yc zImr=1VI)bJZ%2@f+s`v##eN{+*(Tt)u76?CcSM%id_H**$NiTP4uP{4@hNgXyg|9l z8<&3%Lj6Ck0^~@}LYD0$gwfr~UGfd=9*?$Jee-hF6)qdt@_;To2gpXCj>{t26XT|P z`}~{2Z6dU}4(y^e#6+1GGUBKm~1aNUG4RD4IR_33scF!0-r6(ocSF z(f(Pbxf2lza^Q43hfq+pqQt+It`GtdGV0bLcZ+GcuFbF~eCDsYiu1ic7sw&ooy)6l#zk!yTBDSvfm8(5^A|=n-_((dRa9 z_xoGEZJF7=&4wHM8bgDQ;f?cUd0JqZXk`S9x6Me8P zRd-(50If*f|1Zt{)*F%U{^B>ke=E{x7pVTKcbosg>Ofxl92(1R1*U{l1ssAVvdWC&5sp9z_{D( zvLK7jp$%fx7*0ar7IgKYN4i&a=H(c>cKVv^H;+dFu$QQE1x^&+1B*Fh%?hL%hE<{4ZbY!>SfR)QB49kTa3QLpKa zehvQGLiPLr#6Smx`{g?oX?|&bLr#AibTVQayT95My~+(9LrzCDv4Ya@175KLY$fX_ zr`@W60LEI#+6srGk;%EoD#$MOxy4wv4L(!;E=P0VN)gt^L(?)AEkfnt7*4Pz<=mi;4wn=VO-XinQAezPpWS4YnRHj|Xq^fb;k( zt=iz8jd%M+sQHVdc=dxaNau5-yz#0px}6*CPa3j0;JLB#Gd*)+X}{_;>a z6jIym7mM>~?EA`fEQ+%2f?TRw0m~CdiP$t;7Kyo63Z& zlPq})pQ(8b*f>ViyggT1;|!Fp|t z=pYO=~)R_j4}jj%cf*Ke)nWqx(G-&MEC;{EWnx`01kR za~iu^_1Y`aoAc`F3Q@bbQ*}qA*Cq=4NrUo{@1;PTbWTs-2!*DX0(+p$>-2Q|J6pZg zD-C!)9*J@VCrEGbfa@{g0DRJGPm$vwu!i^9=ea4U!sWJS;gcpSTLEh%$gJPy#uzuS z#xhY;HGhemX#!jH+tb&RwC5(Mt1yG+UWx_b@mwnMinIoJgtDNKtW8iGfH~(RpugLy z*upj{k>2^NuuZ*rX0hkqKqH3-(gb1g=`B%KP#WE&UhdaE1NntoXk#zwgt&sZhmlWm zkqb6Yxcm4IZ6cdL__YZ!0dKzeBI%e)qRlA8`Y5)SBDbl?9PcJ|s!xLgS-X$}qea{a zeETRNv%`I}z$(IvG_YU5W*rN?_Nh5Ee>x*n7Os_ZkpAbt&7Ja|;qd(6vSU-aCJ- zrV9}&<8o&80feaIbYh|*J1=a>=1`{%wvmhUv#S8P)*BVZ| z8)UL+Ym$13NWn9vYJSRWuB)Kf!xTA4MH&-`y{db`C|ZDx)3}I(`N-1n&QNsKizG!7 zT;kDXN!AOS8KdwC{v8$8rbY&$G~mo@aMs;bb6d&ktp39A?*5p%47ZZNo(_Q!@iQ(r(Ij+$qzz{ql1l&roNfg-# zyl0}EIi=xk$mxkrld<_G)n}Vm&YVP_0oRWCSdh-kl5jqD>kjY&jdi!Q$ysYYzToqc z2^!zq^Q&Kz?avJ~z{ZS#Mn1*nQUI68CN(m%U`859T9DQ(i7EEH;xgm}&YyCP(op=E z=+mZdp>Mex&Ovq-tcGk>V0DmguWSQ!>!!(CXyih(LrMV0l^0w%{A-)w!{R2NpJU&9 z!33A+)q%&!kP~#Tk9^K?6uX`xYml|O0#bAjp$r067@{P1pk8J&U%1D=xHXUR@yV08 z!ELU`2!tKi0vGQd{I6%71E`wZC)j$wUtZvzKf6s;1$laE+y;0iDNBm&j= zRlMLzouqHQY1x$Eu)z>Vn*sWxh9Y6C1~fXQI60s{QREUO>4u#0q{72WWEf_H$%F?l zWV|r_{ot*Zz9@$iw{rHTpkAg|bpu+$GDW&v&1TnuV3bX~0f9gc(4h2(PoBmzK2bg+ zFqv;CzY%a)9j*WV<%fT;Y<_fDqruV6SmN0a@>icSTNGDG9i1fH=o=?Xm*v8~V%2iV z8M*P(WZI4A0<_dn%v&`tn;g%wIq$9}d-%mE4%|Hbftmk&gkld-WIq-8q4Wl`TGSuh z0|mv+Y=Qd;N%y=aZORWSULX4Qd97ZH?O3()2Jy)mYv&?8WvY6W=mSYC6ceV=ZK^Nk zeh^qjb^ARKX2_bf8UNDyt==E^EIzMIRez$12F?3-@_go{AR|0j}NT2q@8nEm~aruLh!dA&`bgLhe`@p1NGJT5^k7y3~gDW?VoR;b@p|at|!uBfJdDlqvVU ze=Ih^=9PGyW-)R+0b_Bs{XTL7X4Ol7O_^>2NBg4SYb0eV0ZH-^D;#?$76u@TitGg* zQ7jZn5>}{D<_C;7ZsIH1tSAQNCEfAxRCQC(=7kk5*!bM6$dwy3$FV#TIO+%7kb=I} zA755`8FHQjVE%#{KiOpzxCda2ti0pCiZ2STAF`|%a9B^4qdK5P+#h@+xN7DGw=5vr z{yaR<=eFWwwqH@eCbJt@Jci{WsBmT6U+JIBrqLg&V}n+@KMJm+52$kH-Bgpc(YC*> zBg^=uTO9Yz0JfWvMl*q8V=1zZitG$M3`92F%02$N4A)+!4y3D4D9u)|;4zBUA&AEf zZFK7A^_{QDOqj?JZ@xid9oVadV%iZ?V=Bd_P-H6=Sv2jzSM?F8)9S(>29yfgn3@Po zowbPjB0ibku88x2l5aRx;c_c%ms`t1{bbSQS~uh)?ShTQ^eMODXEfgX;_n3t6F{VQ zq`xKc{EUVJ=V%Mf;FM0WQ1p;YMW)M62d5u zR*(INYbIlp+n7W(tiSpvli~PP-H#WOJ5x!VS6n*aNHz19ax18c^F# zcP@blC7X^XEzmcdrD-9X{JMjCp)9{-&XCh4=AJ5xUIlf=1)+G>DuQb!M)ek~ZvJ3? zg}Q9PR(3btD?b7`i)%rx9$Q%ut;Cuo=(E8(%e%2^lvvq1w0xh)d~%bahMivs*z2UWGCI$+2t zi@qYgEVw6+0kuu&vPP!825G;dRJ8>d!YdZxFFm|Jih4fE_Z`iR^xk0)qs8bd1th?p@fp0a0&N`&FGE8;%T$e#bZS8_WDH^g19*`l1CMSEfpP7&G zxS(Tx3%h)JaA@m!%ZirKM#X{sUXCi4KByYVU);+4%{am8wVk9tJWbvre&l=Bt3Co# z{)M3%Lu_p!o-juI5b%T>+4vB*@p!)}`i{TJ>6w%E_XLvtjAi?enZ-m0C>FSdi>OGX z9_n|kRz({KHd;MF?X(`0<*^9^+NJ}CoQ!{R=&gP|Bq6+##%@LYe;uh2oD2FGW~x){ zC84hG<@Omzg}opd-vW)6#h$gZkY^bG*c4X$fHT%~^W{EUVuu?zb1ns~u$0H%D%ZEhyJ}{}%gr31o36`rvR>Pc$63qw0V z_xHW-A*_p;>%_aKb18sNBQyyFjT-~EHh)G%3$gnT}h zsd3!jpKLG-28Ss2Aw|BRA~WR|MMu;bvINPkz>(Wc+5-^TIfn(T(dA$ zJwO+bv(R${8jwHzY;pV#pTSU2(^sALkEMOn0WOkWn+}B%Io|1XzH6eU+y6jdA9E=1 zP+-0K&7^+jk+-nR_SVm@y*rUs zM90l75RO*FL?0B4tE5{QU9YOhKi#>DM1{qZWs)NQRCTZFktkMVIIuFHUbW2QxL_!3 z8@m^H_df9bSPF_4jm{9RdldP@IVr$IZDYr1c#OvhmoL2{*c!fCfwTcCBBUQ2F!q7S`HkRRUU-YYDSr$bh9^;CokG84s2-A|%L zee*WD>9vRHyWzVvACSGkg;yRHEeQc1lg<1Vwj9_p?K1;mHpPNwLOK-*+H7E%KJvsi`6jgvv*oa|A&gp>WFS_| zAQ+cH_9r3y0gZJ<#th;CFSJ6-7A8%zlPww^zKK9+_#XE7sVtjy*|NFqIY|g|^ynGe z5wOb4NakT6kZq@P)osjv|1?lZiB!}$@R}$` zMa?mStDk$_O0~y;HLFsJQJ9_W0>=PYg$a&p?YaB+AG}*^sp$7irH4U0Y($k!BE=?9 zB$kSN0CZ~TfTXJ9gj>BHd1VN@LN`J?BQq#CZR~p($Pp098f2*~R+CBjR_rvCX)MKRa9M@Tk zJ}i3Q)8vLW{o=qQ58^T64sHxi z_y|7`uXthGTrU$={{8Hf^CX#H$l}1sU}%pXG3n${EG$xHQjvACoiiX4D?J>M#GsTL zAVv@+(Q9)B(SeQ1e5hdPS9As5^IEA(s zF_JFgOsVwWsU3N*fMnY~qUD;T-sD}y>|P>aO1>MPPjpJKrYXWe<{4B4vb=YtY` z5`88`d2D=evK?r+@8t6Njl?-7EArl7+seuAspN#&%X)}n_fw>Vid@57m_7(A%`qXd zf}_IKQ>$DO++wBKNJqZa|c`#efsE&z@*Y9xZ=|4*+YogY!R%TR833$5nO_*f&%#21L_b%FJ{ikBYQ=81bAwl@c3TQ5&x>0`#`Z`m;8wDA$$3#d?8?8 z+x8`h_*pr1^w4h0N^XbkedVayJ{NJ3{sJmPP#0j8Ha&dBd@O>i@$Y5Y1l>TmzCvUKrL?tJZtqo=#QS0?J1c`P4MYy(A3Q<258&B85gKhPp48ImRi9;F&ol&KA&fc88chNd9phC?l-PM16{7BM1VDI%lxV}#`y@(#AHk*u&Kw3 zvNR~lhRL_epf)B!_&IdnXSntQGeZx3ieOFLXCSZDBCZjo!`#*mYe_BAL0UM5~+R>b>&VnLkk9u=9(WYNn> z#f$-0q@s%_I#iNfCdrdwAKH*ple!zE-!aTz8Sp3=h~x)c(?gKK-#}K=s4Vu|s~S|D z@Qzj{3bAL>P^*&SkswNrSOe8h7u@$6tWlY%K{;CN)s1HhrESc|q#x9iJ5@Us zIuIUZ@?>~3{L)7Ecu|!oTi7biT$Dwx4p|)nc^}lP9ljQA9p-^zVL~5Z@`XD5p7G;0ce6BWr`34h1Cz8bVuycY5cEu<$k$;_RCp#mAVko#(7nOSAt?p^xmS&# zYUY6315o6$MFejr1IE{0jXIx~0h@Bwc2(CtVE#cF)4~57amwQ zWjDP>c>$yx`&}Ev`^i3FG~eT&D@R{S*Xa`P+$@FODV>h@d^c*GICe}FRvfpT@a{ME z|DUh8nruzk*S~X)taD&n1EMk`RwGj=b}I$XH4;6iT=x<|jaNbhu6Nh_>$O)S&Ph`j zUXUJAl&E@?jnYq-Yf`TN7umpXW!r(* zp&(f}0!uq6Hkl$@K&=+$tTv_%3MpZ>%BT0LuyKD-b;wzVM<7Y0X$}Iht63$A4IVfV z+09^>A%Q2F(0ew!-d-O%oa0g^vX^<715d1a>W1=i)NkH03zSGzUqR13M%~%_fok6bpI+d#OmwysTk9h3VqD zCk8>K;iuKriaZH)cdU2+P#P6hK-x&PD$lo_ZiN$CJ#|sjmM+C4b{%~cBw^FRn8b^2 zdnQE0`FsIcz%=>+eVdM-MpwNk zRg)$^^YTVg%b(f$u7t-r-rj z2=CNQ*9eZukv!`eOofb5Gwirox0gFqkpU(MmUI^{Cpivmx{jNfu7eZ{-D$;CWV&dX z#OUrmj#c44JBBoXRP%4p8?^+ujI*+6_8=k|} zUFp^B3jPqROQ0gVu0?!AT`0|>YaqjjFIo#=zc_v;P&*J^9G%g1-EJRhj3wI z&?zb~#ivTN!@Wt_E8OS(NLb_DPf&Wenb{>ozbjjFTsCo3jj!yLgS(#p`QsJw1(nB;P!J zzx$s5==MuhpOxK(BNqidbh|3sCrZ*NJ1Z-N`T)H)N`m*P2#RsG1#(6YvJGt1=$qPL z&)+{gl4r>w?6Ac<9IU{|GHP^*bx7&Z0`p7NeBH``Gny8IM>OPg!M#|}3?%mHAt-o{ zhf~$|u0n^w8RwweU|64g3vL54hD~_kR`Sv`OAcW^M#XWRHNwFzd_+?aikMIH5?or(^yi2YeA*q=`*(jaj=0| zL1Lc?;K_qx)hIvt7Irzx&r*2(=eCTASPBPD_io&d5I�oBu{-3;)_sULSyVGl1J+o*$ z)Zkhp`pY}Tuj1axf{B&j_%di2If8cF4tw-J3oO^3`GCWLoePeY>PBTL6y0^x+o?)1 zWKTMwArgwE^x-*@B9~50o&+Spj*&+7+W99y-*J-Q!5bZuv$Qy+dSUBsl_@8Z_R?Fw zBS{XNO#?dK5v$2r6boz}X|S3+<2dx1qzf_$N@ul*KM5}JA5@(obzl)74uRY(<~lG| zZ4e)h$N-yA{bI$;CdeiaFg13!5Bx{oWFqWn6Y$o&91m*6Tkl6&O2j@>#4a2{u%CbP zn*6V=%;xE}WVL*N=@btv+~K)Z(V#pl{#d#raLBjfMZGp>&gaq`|1TEbC!eZrNGHW& zj0ZJ#28`kM+^C6+OFQ_ihbwxNlmLpSR5kvMrLnlqG${jaF?d#uzr{6^@sWG{$4YGc z;H9n~NldfL+ojXqCF}WBHaW2Gk!Lm+ZKv36=tW1O-+{c!t&lC(VJ+kUlPt!i9V}mg zoY_P7Zn{U69*CupYhRlrWLX8KFCA0ZnhYqxcijEcuPj|ths|(sIGcK}XeAa`?DJS8u=-`l! zK-n}*YwB`AmrIA#PtX$t>f^)vhn!kH8<%0iSOs~6NbvQjF~xtJEAmLb8Ap^*rX}Ln$a2u zLae#%gwNrJkg0L6=ULKJIxIpU$Q=>&KBlBh54j@U5E3IvgBqkJbuK7X0X z(QA>1109VmvP|H)UGBDJ_HtO1-RHSQRvtDWta@>STY@Y@czgO;Wi{QTHJ-EQYiV?Y zs6vz|$eB|jIN*{7rTNE%hvZM8%jfqB^T);*j>zL(YqeLbgSnYBldeH#gAwBDbig^PIOT&BQ{lPu-Nw;EIm$C)e653T8$yG^)ti$z$ z>p@|Y;0wjdg+rlfzIeupz;xLci*E(?hTC$+;)%4!MrOt@(FN3l@z`XLp$S9nE=>>5Z~*{docF%lp@I49jD#Cm)b=giVPp!RCoaRveFMQu3;oa)8UV!c5l0xK4hTln%b@58HD6D9Y=mHH zzv?!fPeH{s?Af~@fyt+^wE%DWS#Yg?AE}hCcgdCH3tFID=^~uT8*erR)tE>dpWTzc zBmIkIWvj!+L>#rM)rtY8V}7YBn{E&7m|FmBFM91>(ExKuj(RT#7Y=~xW{dckd_eqA z-W67&&}(sDwPH|}D~NyPp}aY0&}o2)kvt%92XUSAB^B@Cwdysu0S2u49=<2>o5H3a z|IJUGOlBzmi(mhhL`@~BX76u2#crU;S}O9gG&3|iBn^r)cK90wR+`lId9+XYCtLI6 zpK=Sk9Om`p9oK%pzeW4gXMGXyD72iV(K)USFufr47yh5Bwm}ySZ{S!1)=3V94gUsS zV4bpeQ)?kHnTZmo^wVVZSkbfNVwz-`*|aSbn@EuaD)J6f zS);OzX;wr*GOk{J+_r?xPeqS)XfZekPra8dPvAz&pQ)c~|FI+;(gm&Hq;@IOL8N-G zBo>-quZypGZkvg*k6xRlS+Nk0*U|l?5Nys8Jh&3nGnb@FK87_R?Ch(Wc`rD__3$ja z&7M49G^Y1yOX!TeSI)sr*_LGNe3G*c958cm%I|`pC)pL`b_MmSdeyz)YJf;*V7)5Y zwdoE0KgUw!U(-c(wDG=8O8kju z*$he(g^nejMh^c(wqNpk_42Ex8R(%f$W zwr#*rQ7(vwwsB~k45{(H1yVuSAf?0O4@_i$H|HJH=sJU=0`oxZv>zyUp^SiDrEa0~ z#?dcf>yfY=8PjLpgTg%*jFBUAT$i9LIit(n1gIa??6^oaO@+4Zk;`a_FVDr#4XPS%RM5WS0y01TJ!F?#k$VEc^Und+Z1!4mj?y?$ z!g)Vi9pHr$$F=JlFNc2UVM0ms%NO4!i4I&_4V)??R+%#?7W#2gsmL^WMG$0EWJ?}z zG-)fP_eqELyta;BvLw&faI_BCc8-Z^ys;_RsFQfYyN#*w&h@n?ipPmxqw%s8SjV^z zPyD+4wBcV(FiE{7c|^`TFid*QVA4*pEfl#%MYf1D7j0lGg3ya?V>*PuM@udWwh3Z` zy5x;s*go8$?NnCEu8Cnx6K|F?I6X zJqF77mhk^;&D&>L6V6;8^j#ub92jSysx-nJg;k&56KX&i9q8twkfdH zfNF&=T<$Ld#!;Y|H`2CUIi{W!Xuh^kSRH2u%w@hn)8K z7lzi(x*|1{{x=CQl5AEaF`dD&0)sMEiQjTjhp=KsuHcg3fi1*rticb@nFmgvIJExK zD*Ux&WgSOJTC(6=L<2m(MA=2J+*!@OX>^Z#xAQKc4p(IDFKub@Z{_hoEv~fh$LGWz z>pUZ?7Mkpe?}~MM$f2p^g4qjMPq9FvR!c=LXZ2rQ;nx}roa2F=uwHwDTn)XexE#=} z+~pY+qSvC{L5sL0xHPnxiIS`W8HB|7+kMglQE9PQRYyM{cNGod)l&}(GGxW`&uX#* z^FWOeO4+E6@9Teg_dDmcxCb@H_1Z4wR>d7UJ#ZKKSem2hA}2`GeEd0~Ie`b1c)y|( z0UIwUIc~K5=9FbqENi+sN~ineP+wUD+jb5`AZA}4V}dcM=VILiueK8s5Z?B8s!1=T;~0Rp$Ai#&IW8 z!l6k&D5{T0Rih)-?icNw=e-+pb;pFu1ly+|OC1VB69k^btGcK7UhlVbV z-XpoIx&T9uABUWZU5Z^MVeom4`QwSg{$tMVf{e8Q{#t43pk;;ab7DI<3WYN@ZOj?> z^9yuovMOmIWXljWlYo~|BX1ReMqOq99jETO(-RQ`a$ui>OALr!$?m2%yY2%G)&mg* z?xzGPAvc&5<|e3#=YUpt5~C+)y(SYQR$*hzNr zfEZJ+ZC2FL4NC0aG9Eo6KNr-X%yFr3sir%{MG zmkv5-?I7LMr0x(x2XeEbCjteYP-za&(xr!_3)Z+bh}XE?_JkB{qu@NfRuU!YhCXM* zvF++!6_&rjF%&pU2ycKRLr(ZTpk+uNknVY@>Q$l-ppRr?-tRatWNtXQKTQZwogaPw zPZOb!bYLsPK_6MHDqnz&12^aGgEh%?`968~yj$`FkzRWp*bvr{TfrcpC+KlKY=Yn`Yf0B_?Q?6h6l%@pZ+@yhJdVEbx7VTEDytQtwD^;_ooOJ zBPD}MBFLXgR=W3v$N5AH_V}*~LYC`fvHiiq7bPZ}F(Cw4b8)e@;X{!r!kM$}pXS3VSKNXQ~fX|n=k+(PXShLHoq z0~jMOv-@`B-MGD$-KP#qcyY*VsPiZml4Y6Lu9m7ca*5{3fe{5YB@043gdiumMX*fr znG!?ccIaI*w4qh{?_`k&LzksNDbs2?0oL-O7S^lE=h%^+mdEgOxDob20m$SO4VPl&@JFd_H9oIA~ zYW*ujy4+Xt{&VFkSPqiz+!(6sR<2RzNzP~<%GXWSY{=gzJaJHMc90i3UU>8C|Mq_w zQyjA;A>d=OnO_{^z*7x~qK}AUvMCl?L(-|p1765<-^LtJ?PLu)>A;VoJv%IizFl!f zt~>SB9`LM;HHSSBotj2vBCKH|x8L?zF;onHc0L}1%O6%FW*r{i!}rvgw|;a+W&%ov zc=HVsI~L!P11CT7&ESwqu_+YUN=5dFVgxbFRe4izr>aH#ek-$_x~HmMJgBO0*}M3j z>b*hLfG}2citLw{0VVoL&5p_8q#X|o_zvlK&gZr+EB-2Uvt-h7SZ9QT8wcrFFq!_D z5+yU1E=7sIr7_enM<|q??FA0UDv~E_R&;sp3xgdMGf=ns6JWy8b$NDy#?#WH!uHS& zY_lSVF7jTw6ctAQc9~!)bZC~wuu%(_Br*75!XC_cU^2p9aBhwV8Y8dahRGNOX;Xgl z(>Y2@{n^po6b_s+16GR>Q_Uudji<;4DiS$f^`uZ*9%XzMv@CP3+JRd zy*gD5bi#C6;E+?B04bm9Rh80}URUK%^D0{nT{gOy*{Q%cjD?GtQ1ou=HVDJuS&0r# zdtp2<5gd$tEjK#GvLiqLn%{RULmdu*t?Qww8dSc~9Th^ha(W7*JK%+j?q{Yyl*|nG!dL}^csx)&}_Yz-yk?_R_%rMZ(Xuvk$q0GYH4QkPXfipQe+(!+3(sI zgmiOFU}JQc&_Cdns!pH0CEhQAXq|oe^8Cr#+EN`@VM(U-OnDw01Y4-Uljwu%G;Pcw zSG}spWx#Di#D<7E8jV~$TPVd8>+x_bp6gZ?db9yTF8CKp^IYs3-5SQ#Ixvi$YwQgQ zZae1`HV_hE3VutviD4Z;a~E%4ZYaGfS`=5ooUMYqC= z1PH~mBu&;$mjdl2>L=K?j&{758jUg=r*D_O^52%Bjl)L89AQj1^Rc2DL@7r9nQ;8d z`<1qGoO$1Oq8D)6jxik77d%QYOD9re6F#Z>j7z-rCz5D>P_ zljSJdL%Zo65!oRH!tPKgA*iHR&fXM+I>yCrNP`Pq=aEXohHtQ~P!bZ}7SpsIBmUF35DP7Nah-^A29Hpo(t3c2r?HPh>FPb3Y`{{PwK3C)8^P{)2H2$>%%g5tc|fOPZTyaXZ2CJZ^n%5Y!D{036hDTYjW_tAIOShc{t`Mo^*AryMT#7F z>N#U(@Tw^m#QzRIrtFP(ZkBc{2bc!uUGm$cQ&qMAG$LP$Rxa%)@pC?PE%!bAvM$%X zUvYBQkWCRoG+!spXosMX_>0EV+WY8(q=X2>Mzd@(6(2hBSP9M2n@K5tg^BZzH zGrdE6iWG$ElKt+0NMxQ2tE+7Qt%*X&+$eteX`+bY2ccsRU-+FRBe26dwHzG3*ei@=`K6!Y@(lz~yWon;EQA6kAV`>I1Jf4vEMyu*^ln9)`+9SykcgNJC@Xg!m+SorVc6T_ehEt>q-_wF%vjCY4dwX ze|V8QR39*9bgu6WU|k%~81j9*;~I0mUowvKdCgZ-G<&{&%u>_9VG+WiQc$km=M5fm zo@Br^iK&_&D=L*v1^`BljWviEbV`cPrHh#qiFhwPIM znJ$pD|F9n_gF$2^!=>r*L&p~A7y~e0?iwddc-fv8t`GfWiJu8eb+2wLBbg42rCKvA z?W0&&IV_|iFXAE*)ykZQm?zdN#bdkkhY|N9YQ;_30jQrnuhna}$Z`eS{m=ejC-BC_ z2{UH)L&Mku2%fU(9&&l^xdjd4O|lOo_6v&Xt3N-rcr$fiaq_o%zNYs&VfzksTft_W z=-|AJ7i_4WM#)7>*zhTgbKs^A4v-xRtXKEYtpK(6TA2z!+MnSkM%lSc7G1k|1y!>6 z&&T0M{o)?FS5@Sn>eCdoKVV-#3uty+T6AH$al7B6SE$#XAU*W;Z@qJ3aiPaDXT0&< z@K2`O(t)xS$PC7a6Ege{ykETN?rf8_agu2clC0;((>ZEpoAy&Il-cd2B3Fl{1m5z@ zqWk8REofu*%^Gmm9S_K-*8pv364NWpBZI0#&UrF)Xoj4+1o?tIS)+TFCfB#lbBE^@ zmkn%>e5asMnJ#g}BGxGJZMSU!^vLm>C{{+^%a+^E30A&6=hfd@E>b=xQxwONW}9a* zT`BIC=E(}&jfSbf)tpU91ZvHD3OEBx2}V=4M1`e1036xhM&9xg0$K?H4y z7`Jmhf$_2mKohwCcrR7XEcN)R$)WxCo99xg$BF7-#2J8oMz`WEi z$dbg$jQpZaY8?18B+lg&F(^h}T2u`sFUy_VXXJS2(09VSl{#c0#fbw5B2ie;E(^u3#X^@` zHWk?~Gv$d|sPLu-9#` zFiF@D@|oW%*JkFna}u+KI<>env|Sa)*c8tr(KJzPkGz*Hx4E(X#?Pkhe$51zf?0w) zr24s8P`YXcqy~yTO_7sSBxZ0+y+882Bhsad`bkejuHe&o$sR*aSl^Hx(xyhLLafN! zBWZ)Oi7q8pImC#HRT(p{ha||NA-&YBh!qVgjd{!=r&^fv;+P%&r`SuvK^5MmNqyGK zU{8C42V7$XrQw$rJ(OPveT1btNlc?MQ-eL2Sd4x{Qwa?b*&*YZ>m~{|D`!65IQ*|w zO?NFPuID6G!!iG^6d~2ZpbD)H7FEU1+2pS4CmReL?O1Y!qA$8$@962s+~2N(%0Db1 zUM{J1D+KXwLrnGoDTbbza^L%^2XqT8fOjbmhv;(VG(pj@ZAF)D3LD;<{rLD>K)j5?l7}0dWwY=kaJX|(OpLIT!V@(zU{^ZqXE~u;dennO}A3iq)r!= zEht{xA&gpR=+x^cWeZZ(DGOJ*=E;&3A4u>_00ru+r7WyhCoxMBUa6#)pi6yw`dOv! zx*SmiJ3sULa%Yoz1B<2w8@vbH4uaFBs}_GK?NPR=m$bOF3YMg*dsJOOq8KatP}dAePaZfNyCjN` zU;p;^x<9?~n_v9#&BqD=Pl3l%?>o^GJTLqA{=sB`m@Gy@K6hXPwBd!3>tsU|`;a1E zP?2}&9R!P~R)wYbr1?TppM2t*sQFM`9uGkpSNfvygyScDWuA1fMzhNQY2*GojZsyvjc%I1!ynbt;>pDcs^amo4E&(P>(&!Gd8p|p zO>YkSM{zMPlP-SXaNLvyS!HI-`YHAvMS3t5iRCzhs`7vx5yp})9V+smdv#h}4b?Z0 zbnGWFBtLvW2;seKr>1dn)WTHt-5I6cu;%#yh@&sDO~3(sVfv6$sWeJ*b6%NmqY{@q zAL1HllN!rRuE}$?NGFXvX}Sj0CO`a8BUmoUl_L+Sp~_&v897!2Ek&n1Rh{g2Wbq++ zt$0bB;JorS+r}=@)zg(hNeprhM+M@HiWu%??x~Ims{#szkC^3dhI)EDp&Q(LQzKv( z5Vz&fwf>*_4~FS-H4baVKK%R8e98p9M{bYKkU=MhEh-@(@;ZWCFP380Q6vUCX0Roq zYCdE|n-z8>$9T#+ZiRJu${p-;nBPRzvg)HZ-AsV^&Paxh}U3?|6&^~Hu6e7P^srPFSZiU{3uN!0XiupwurX9%AXUfE(Stw7Wm zevB=?`N{X#@Ii>ua(&ccSI%LOM;MKGirqkwwNzy8H?PU#e0u0yK}q0p(SeA5X~E=3 zQCL1&YX;f!n#uUg%WP1`XMFvL}1X zSc<6?nZ1fTDHe#fw^NarMom>W1#MoqJLHIOerOk|aOsm~&cgQT?ckK@dR1q=wyBRq z;33GJnNA!}Vtw@Ud-5J@p6>sK3Fh^3I&OHda{9UgOFCZ;*4!fZZn}&vQyVm55`9X6 zd>s<+nVL-?i#pjap4F&$GPrDh^f^Jr$KNJSmN{LAbuT!wwEN|qnld-z-%Z+h>WlBj zeDfBpUt=aDgFKY?h&ojp!-w}*8=}fedb4YjowYX+b5_IG+_Kpmu9i3C1z+m*ORg)P zb(CPRx_M@rYz-3?mL;i&FeJyj#m&$HgC%&zR6~t_FQdb%DRaVbdmFuRqGVjmK$b{& z(q~>MdEpmxZ%0ovg|l7%edT}1nyEmtF)||9MzLUAHc^qSp6#Kq$|k9l8WPj0XtpwF zmk?bOWO6z2%{z;CyCnqK`>~Jwr}G+ukr&!|o8J&PY(8mj-QG@-2^({^{c}B8#V^(2 zxICkvIzJNPn<;h^MdGPQV|r}JX;V;3FeW*SDUMsBO+hPXM}uP4WP%$oYCXQ~3}}nn zudn>pPb~8re1cgA1_nnqWOK-V|GK#iA@L$4q^n??nLNoPgTXik) zH7jl3ly40V%-z$mKiL`_yrw>C`|H!)E!pEaNZO65qzk|>se{!kY;H9zYM}TJcJ&yu zBq<;{mh1bOANGL_zyiPhU(#`ogsmU^sQd963CFd_Oxo{$_p-^bl+Af}HQDow6?K1L zW^9g7Ebubyhgw|UGTF9(25~>pYflTysQvO&qB6G<0q$#n-q#(0SABN`E|ct+rK+#` zCPW~Yc(URYSpju%Algay2j2)j1SR41^f`Cj?xqh!ocD?e_@6F_l`}LwVLN72`CSvo zG4FQ~TfGG1g!L^L0_^h~n22$Yd6<(2U zn{`ixy-3Z>zz-_FQ!!(`aL5UBg9C2YL#hqe*$N)U!LuPq#)k%utGGcjR^24?@wbC5 z``I|!#_FIStTPxr1FY}LnNu`-pYLgPzVL{D#Uj0Sy>po~FSNpCTRYYdS3C?PkPsfy-F&pD$=NJ$xcJmrD$1MTVE;*fN*-6bO zUUJ-j8R6)=##Z~rpxtzyZ>6+Fj08eR)M+FXYGalO?oV>_EDwNN0ZZGR@|%zgt(FZV z9GwPV(03Ank-H9bU^=9eo$C7lO!rLajI<1zvJG3tTvc-~_e42xZ( z?mqqE&W^Qg!VG`-AW<+|V`<^y3i zR5~a4-;aoKL7(Oe0fu&)-9CNM$Yj)_IY(bpfI7Vr#o9iZfiy>#nKJn;2JewWw=h?O z8)90U@^sv{%G@I;U0!D;t$7t4_E5yye8tyI_|%W%a<(;?WEI1 zyFzM3Uns6i6O|wvj4beHy<%l|XP|WOBwHIk0Jk}cpK!y=-?(P zsb1TltfN;3tO`TL0NdL-(`fKP#Ar)rvozc&Y;c>r@YpLqb~epCbziA|pF}&b1H02~ zN=&5K1d7CBTgC7avCh!x44x?TSezmI)cv9;n^Llmda$E+SPy~G*V=OrHw4C-_3oG7 z^{|wRb=bi!|FafvMapT`iBQxSP@$@R^u-UQCJVAA<@$e- z4g88z95=WJcYEYIzz&K{rv88S-UY6y?Ajmqh$kc)^Cpm-0TqcLh%*R=iaO{_JJWXh znEs}{z4zZ=?{Dt3(tGQjPVd~gscomdMSP(mf`STY0C@=l3h{*sf-igvh>GG92>2Kr z1Z5Bv{%a*cNlY*Y5^l72{EVEl&pA7I_BU&T4J{6&(?bFll;50% zV!C!dEPf7?Cmy-53VP%Xd>R|1cc-DtGD2R7?S|+1%;QLd)8F^)fb+k6)uqz9-bw~` z+O#0`a*r<122LCK3?#$bv}KI@-}^}ZWHq^rDhkyt9S1^{qIa)_&=pH2g1MIA61MqXAjKa! z(f9Mev0J9dpyO#HM4tN#BV;dFL4BMOvVrA;cQtk4tgB?Y18)4~c;Om0G zF49Z@Q^N>nxUzu7pzyNyMOLsF-L&lS?I7cHa_h%;OGqBOb%_(l$ABeyKz>pw#lQ+! z3u&{S3A_}@U1<#hkutv?ZVFIrZR2jAd~;gg^iHlud6wMgRdTL{`3dTE7mqWV8r|$^*GcMYV`@rGOwCq`0TQij zD)Q>+Wn(r=YTw?YyyX)!<|6PSqZY76ne4qZ7}N66a8v*rsYdyY{3$O*a0xhUzX87W z3V#0NBZ0fzz9EL^W5zfPMl6bU9BocaL0FLx`9t$+E~oWZtkh>6lb{k-Dn6j(T4(h4{_hWdX}+WDw7{^EZ+Lk& z=;?u09+D>ZxnKwsX{JFa4cbO982W4^Ss>%tGb3@p4jFq&XGTsjPB4G|vUxw*;>4j5 z(1su2o$jHSofIjdBI7yL^xdGtoMhEs_VX_>>cxWF{1~_S@ZW{^PtHmEU@P!_?(oru zB@<}6HunJMJ}*kRNAftdhpXQBZkbyWwe&qaL4o)U{v8Ck^5olqX%E|w_FTP}eh_BR zplFlzD6r#~F$)hq&@7;1aRM2vaqv$b{3*=L_c~pCOy0=99CRh{YVZ|KH%aE`CTw)g_q{!J-a2jrQ_GuI z7tujo`M1mec0K}q5{6Hps~hOfVO%(VK3Xhj(CDG~4YKLsh+smhqZW_c=9d{lDcf|rI= z(epZ}tMDtp?^5WIPT%KM(d&O5JNJBqwMo;FxEu_#&#qa4=keY2w|tC9d&-<#PBczz zr1qO&r-));uWt*a63BC0clfBQ1(iHx`+Q6r*fx9R+2Tg`JwaQAOCvN>_ejdz9x<(< zXnp4G0dlrN_j!8jvtttO0-lgx;b@enLp98D8N_oO5xub!1gs3?(3{zrrjv=G|Ie*| zR#RX2g2Xs6U_b+QfU(&?F~A_c28_*HT0JFBry?K3$d0<+cf(K~#{BD)`EKR1Vz*+u z=JUUFY}i2KrJdRF23F7*9r3pI%=<=Ye0DzNBXY=zp>fj$8s{jcnj)vE$VPs)Dpgnu zDT~1OiuCr0+-}ke*#TK}s&I*S%(y&N`oztsy|Im3L1)wJ#>GL=sX-YlRaoMGO4&Ls z4v58{NcRWUDfK$@S78E4QZ({g-n=AjRNhg}ze$oHpBFv_YEjTBs*%M)XbB zR8*pd#5njYT$f>iHcm{?RDJXQ zcRwKq+4(|Fyd{0Z1cS8{bCx2fsK}>4^{kcOo~jLNR%CIvxi*0Sr#hQnt!Pu!cwm7C z0Be!Bk&h_ro^fSb4wUnELPhHBsgLB3v>bSm;gMn~si5Fh3S1uASmz^Q%yZU>yQ&)jv zeGE6js)u87u$>ixy%)B)nDa-nsVQ{gJZvj8m^GmLoG)x2f0(z=^WpONG;H8a21IbIf9rg{1x?8(mbyo-To=`|^ zmp3luL84IyEb_a&G0uYx2BOOm7~^My!D_IdJ>~cVR+G}GEuYmyGsa0t@VM(TS#OC~ z!-=OJ2=5H=jB_alV#XOjmaEV~S|lvO-4=$_$OukeIzXp zZq*CSw?UR<|AiPgtdA|=I5=;T6?)8E8AoEsI(q(Le=lub*YTRnG*&7;@`ZK4u>w+c zneN5l2_wzD4)z=>>26Vv-~>nk;g-WD&wa`^8dD{Zh7xl#uStqI#)q1`@G-|(rQx6m zK;!5@AD3|VWuTuh-hp}yh`Ip&1!(lkE z(D9!;uYI5}PB`lm0xywuPQ13>VKTYoP)sI8(y7QRppBXy67Sj|+U1=uEDRmE+j!0A znh(mv@A7V^4+JJDR(tH9UL@*sdHyFClyWwLj%w<-ve1L}r-Wm}7il)Jn4DS|yJ0q2 zpIa>KM~+g4`du*B<#pQq$ZIA*d8V|xnXdH4|C1CqCKiKTsPum7T_16g(?KVNRP#G{ z1!2jmLXX^l8fAS%6_9jP&%}U#vMNJd2?TaUa=g3JyH%A9>Zv_~r;;V!WxSKXF_;fI zG}xZWZJBZ5og8i_ov(&{HNYsVK=QXa6$4RC|=DF)b^v#7|; zz^H-BTWP{5sDmy z4W$W30!zGl&;^-aBRA+xPzYh%&c4PPHgE;Grx-TuLPhVtJ+1F-~s41F7NZH6$9g z8S(kzpI!K2@yF*QYC&ByiyTy38NX@VVcszO-{FDdWyZ%=D7<(dJG-;vqu+e;sS!%I zXRe-1ZaXoQmYGNqKBAZ|igZ$usP30Msa2%|<>hQuJ)=?9$bjxW0F?qU6AKtVRMqre zSOw|mXn);=ELD?NHn36Y_?JAh;rf2%4j)`SVd-L@OQJYch?^2?f>S_;5QBIZNRbF) zI#IJw+POh`RZ`DYx4YtDkKnpQqr5NaloW=l;~_w-qpy%AuQd3(O%~@FB`li&spmvYB$MdH|IHV=;x4$(c4KeB^234 zMJ9v+4&OA0Lg?r$RlabsOyhnI=oO1OkLE0>2R-qlKDjii^!B;b$m*1M0b@nQa~2?> zW?XP3=z=4jb@W3?J%j)nqz5PTxs-}^6DnuNafX}OV`@$8i!#%`hjaE`LC0^+apSb^ zw-qEH+Xh+?48kDuU3v&)L}*kAVMz+qJq1~1-br2ol?3|Ez^3Y3B+}9MWaZpdOshVi zqocRU&j65gbT!ZzA(iybfW<*-B-ZQlUp3j@3wk>uVK@W-!n3UG&FJ^b|7Ei|)h?UR zj}!aYhfVCwGh*I4$UpDkB?mr$#ajEgvylC#QI^c;3PZoUiY|e@$}9m;UWRT8FdT28 zi|G4f>mqu;*Zl4C5gO%J^Y(rSnPAYv4xgQVRW!EueE9gsE8js!>T7$;cG(YRFCb-D z4Z-c{AN_sB43k2DJH~UNYc8EGKF*7FhXg-a%Zv(elJ*4@b5Y)?-xV3wi?#h>HqfxW z*Wuqe{%_|S%y}+fsgivw+?SY8qtAVicF58oEkcbu*XVFG>)qaXvN+_7$011kObt2f zjUtcWw?a(0N6cNGCqauW9Rld@icVFWEwh2F~=x!go29y5E1Hsy?D;R=1>kS_f_Dqdo%TiGNdH{@L?* zVhdD|192nSJ6HW+?u|M15qQ!idLp^u)qhhI6*ts)k$rGx9f3pL@~YIm^}|=aHfCPA zvQ5-9*bOL1lrTO)YY@s)sv1VfvhI_Adn_^*?f{Jzh^S#fE zc4p7V?lN+hot<&w<@IutK+t1~>88j-LuLhrlcR+i$mWcLLK$G)eK7ruYx{ULuBxtu z_sX?l_XUl7%+bA#kn%QiagBHfCfBvR2MgZ%*>_&6Z+xp7jJzxtLe zk8@8}2??UPG-id4s3Xj7w1;bNY=54WWg6vH`@v4R(K4lt|Md4HiJfI~;>=JG#2K*J zl}$0gYm!DqR*Nb?3l^feY4j;&9%mJ!4MRX7Pj4FCsnExN3Ox^TOCj8xEyOHz!x{@k zW9uAF-^&hFCYHi>hz#M*pEaab35_`UTs!uUB-V*HC-O~3ej~*sqoW>a7~q@SZ7|m$ z?^T~mmI}p=s%IRS&>~qCkvpXflD_YHw20c>hF6x!;svuFa29_BqJyuI$tEyVs z;)T%2FK+sF584%vSJs$}8ww(=pc!OaJgE3m6U)SEeX?>Dfy_KH##~2koah zMqK^=c-Jwq=(WLBh6%3LQcOHWR#A~C*wp!hW0E7hTsj)mnP6i)Ux<5}tj2S|D;&n5 zt*98rM-1uFiHbgc}qXU;{NS^$Dj@>lUogp zYdzeB!W_B^syy+oT*wQI3yd3I99{t#gH_^rxnV5q0qbG06)F~O1M5$-{eka(8B=Xu zrQkhZ&` zzXw!QTG4J5ru|0?FO0P8gY9ph{RKVI-ROb*>MEH>64-elPCVCa zGx0z&C`Ln(RHP5kh3=k}4pe&Soa5q6UPvs{qbLi+RKD~HtC$sUEFNb#Pf{#%tMbbC zJ;U8CcBq9h0&z%_%&lVt%VM1%AjAc4S;n+~qd z-w(C;E!?PCCd0$n+$67- zl|olZW1%@)RTv8O4@x-PHS>%FY_05^yZ#FDie8j|4Zlws1dP>dM0?@Zg%PJ94>d`# z*B|rwURVtc2}u@xm*F=K#F%xX`RjYN-!{)lXA_KZVmoxf1bxpk*vo=D!=inQ#b?FO zkCGGxerw(aMq%*-@$-FcExm>mgx!}k%k%xBgh&Fauc#5F_&3YXiVLSy(C6qiZ`Y3B zFTpd!XF*sGw?1OsxSbxCXFri_4ES$#><`-faZCH2evo?7pD;LfhhFx?9B$x@10M3E@RMT~6x zl?f={3<&tgQp|FSETtmB4D|oEOM+%be-eu!c!C{jj6Mon)4nSf6I3dpH7urcIG;vl?^TGSfl z-Z$H)KLyKt-0vn-N#=|9hBops$^Oo`RL>Mxs58Qr^r%dyWO<>X2o{5 zR@KOXge6qijI$l?wGNfVuitqs@z0VQ{}OD(#qy+% z?WE9&-HekafGwk#y%gC^MWSzu3(ym)Gm=V9tVfeqUFcT<37$G>e_%Pqb2Fr5pHjy2 zFxG(q4vlh0_&F~uP}}X4(#715Mw+CU_ z1?MS$3V8|a!(Da+rOS)|?4gXri@n&th22OvZOB31c1I*OnwSj! zx|<}kqAMtLhh*cwjD3z#W}6py2}J0&obRmO6xID3uSWCX4Sez zOEBqIlPZp;R)3q>GHUPrHQ;u(91&C4aeYzn!&!l@}U2Ym>vDdMHP26-z60ezbhh>XP z+YEa^jwI5xIs�vem~Oc7c^W8?D~(#VzwXK`Z5WSYZj&wSqnuoJty`OC~1=8geHF z{_Rk&&+MhyHUuN!^_l-(Y%(vLH<-5dNW{ZQMsv~iOW9Jg`L)SygZ=6O#-)T}Ala<| z2*Nyf_%tYDTpxs?bCl_>QD#b-fgvH0lL*Bbbsm?XGO$(N%v8|z5gDpPU%abLcFt?P zxLU5o5_-PYr`M(Htp+B|AHOa+s@rd&=$iw+6s)js)v+=It{qO>Il>}yGI*z5{aT@S z$4p&AeP}9_461KqWRdB(%H61Z)2iGMuA;J1Wy3yy}?%hgCO z*UUWTYzLTRkBH0pY}D8wNp*L-ioQ0nR(5y z&pR(euv^^jRnD0Q#eHry{|gI59V=K+TYPp;f7K<*wcLE+t{L4#U%t7?3%8~*d{(DS zkscSW_TA$^Q4mXCa~Af((&G+mV+D*+GdQ;(@ z?6hnuGS;n(mpW^yAX=C%hV2@3N>XQag*5_Wpt>`xbf!j$Byn0GH2nsC*3$R5|xEqVxuQrCE_Vhq(lPqK;5eNGmAB~YWpvjY2VMc5lUHsQjCo36Zf zqTbx;eI3$Z<9UUl4bqkFu=9lbO%;L5Ldt_YSr=sWYcS7QTCZwltPLD7N=Z{@$d4kQ)UBI0a^@$62yFzhlmz#^1?Ea z7%E?g@qa@xV7?e-Ct9a^(R~ZuOs+z}zY8+k9&@$|TO!b3 zY=Fc;Yy&nmeybdrv<{L@;ZJ6@jaw3UBp^yy%*8t3{=Q-v`@VWRyz>6RVEa}Fz2Ewm z*Fy5L^Ojl$*NLXnMoZK6k?%=_Ema$`M!<(=g&C+DK@~Me@+|Z=Ya9E z(;+MKM%|KK-eaC;z{ZL=uYClEb>Iw{Ofm3fSW873ijBHRxd=tjy2Dxo%V*_;9O9

    W1I!Ld)_Spjdq&&+JLbYbRupy0xtqM=g1-vL;MRV4Ra7>+ zAZ%+m)V_GtM;w;ip8A*T64V#eC=YYn>3soZv&-D>D^@}2dE2xeX?;W<=P1dg`=%cZ zsNkFio<)uFLU6TveFS!zu6j)BJ(E*;OqC0JC@VcuzyIu4|QLXZ#?r3>5iI*S<3P2!d*vmq4Q0aD{p zZuSNDi`W`6++XaX=j?VPH6~0`1La#bS>(S=k#)-bnS;fwDr1Xa4MKlEx94afXBhyN5Ij56Zow)xKTR6T}sCxAYd->9=f5 z%ZzMcG>H;!kakat8RO{r!faS=2aDN*4nDOGmMc$IFMidHMuT+s_!Zy_L!u*!%;DZ_ zYRJuCl+`f^E@PN)L}MVf;{ZKo2M_*$?ycj%g3%cL3HL*3kTIBg>&JIXNS+fHTU41C zh*FB#MG@_D7W-&no&;tT1EDPdWq;t(5RI}C7!}ikKvU&0y%lOjG0((sp+Q=tC>q!2 zvSMr+a1K0}p5~Pf^4dv|abVyELjfgR6zdh8kcz{8@M4FZL04Rcivz}gIESpTGm4*6 z6JgG4@3hW{6{dO=*v4(VT2VjECrW7l&lKy}h>GD1@Q|Kmg^JPN?>N8Q!)Q@{TDj>W zS^J9d$QGMenM{g-&81DC8yA+SN(Xy#!|S2^qUQxq{8=2H#O+TB(O0B`Q2TtGmkK?# zs-r$}yffp{ht2tC!z;o*t!{H%<&DAF)R!%BYW~q7vo!sLe_qh~#-Uy2l zwt@Cue|N2PVQ6`9z4v07dYRi|9@0Yg2u=qb<8EYDu-$eU>Oj3PayB%u89D0%Wyxu9 z+ECl)lOEnN523J$-#c%Wo3-FdwQHsXXDsw_K#EzgWb)#mYS+@4cX^nYhHg!(Jl@q1 zkgIl$cdeFH@czDOQf_W8TlA`z7vqwrMphn-$!X<~&e*9)b$Hm+qB*g}o>_d*!Hul! z*{GxEPMj4OC!^B$KBy%tUz^;uJd-&!m0~tfFlLci5+qg9_%0iNd`8B^Mn2}dB8#ya z*fsPiuJOEwZn4}0B*%5c!Yj65Y2dWyU4yIi;nHf$|%yEhw zr6QMkSBq-GFLFBQ4c?c7YLrLm;_w|3E#1LOo^qe5o6+J^B#IheA8}vMK|g%^(c4R> z=W(=jugfN)jlknIkb`ue-!iq4Z@9ledRMZ9!RIkwX^D4=ZzqVl)r248Kay{iobaiF z0vsI(zhz9^AxfKsJ@&dR6t3}WCuu|9|FIb$1LqSK$1jY$jSZ}<4D}w69>c)c-$p=BFUb+;LmZKKla~L7jb9Kb&37RRjs^)p$~?Y z4(-&;ZZMp7&HJ4n|2f>8%HlPdi>we|G%KQl@+BJOQl3r@x=!wD6cJ3ESsrX?-tT{4 z;(>_|r$KNB*GW4G)LTrL2Nhoe^<#n;@wtP6i-W4^gX5zys^st;H0Y}40bdtiaAGJ zp->MOH5r0VV2{nEv7qy7vSBJLZ}K=_ll;k5^f6kE4)Ol!S2$4Y71RR62Mtmrjk2#R zGkhHMooUJO5Zl!yt*T`alCRlxbW>KgiYaTPI@AB5dmQRu* zA*`B4zA!Z=rP&t=W?#hq)1d&Q|5-==WCsxP!v&w1*K)lkQ;-!}(``U!3gj3FoCayF z=!9I2LgVmEPW58II>mQUcGiRM?=BcGt*<~z)s*3+}QPXsBuCPxz5QX+u23l zoH$^A#$@`cpqMg>?4=^ta~2Ah@RslpK@e?2OVd5!Y+#Qvn=W(}(=NRu=#}G~ z2J36Z`I#`)Vc9tzVv+e?|6CeL&o57`_RWVX)7EJuk zu8Iw@GX+Q)z zv^@fSz8h}smD8E-BjS{pj^S&8iP4=O<;)vS5rz{_G**f*a;H2IEFG&0&Ed8~{d2D( zmbX~eH%*Pp9Gpa!cr52>lxN1R zgiNtMYLxqmF@EMN=YfNCXeC^mcPOr-b&r{)TBKzZH7}yvH81tbW_PO<0sP|Urm*zgmi>Ps8 z8)C&T$pMX>G`ehpdYgRr)Vm&0KxSh9;kIF1Y(az#-PrvW>xh_fe4eYn(Fdt|XH6-| zVz;ny;xey;CLYTUih(_eA}SJdjJ^r)pqJ0RF8Q)a5k)LY=u6^ETI=5^42Yf@l0c7~lT zfEs8^tRLGKQ}2B$|D(AfVG~JlV%S(QF-Q z-R%lb^SxxJt<|uy$$i7CN)pF8Ca=d`^y_KyA%gcl?Wg_FmYHz_Z63e!yx3oR{n2*5~KF z&G{Xj)__=HY+TL#N>N1jPFn5$2$sOd0up$bIZzAbBlQtG+#d+q=n7g#Z{ge)omOg; zmjinhXj*y|mEeg+39INHSps)a5b{~veyhe|D!Ca}mp&{nbH5k>VfrKmZnZ#NP_v?J zLYjD)@U8-Bq6?_<;5uauAH+sU6)E5tiig_hq_Kvp_BV>dhKvD4b_WhC$c*wF_oquT zBVv7Yznv*NPg8ra>d z^x5dT(IcK*gg{CXq$WL;TqGw2_ovoJRMBxvBJ2R36kK*R>HW|#!D@(E-)tQdmX^$4 zE!nfzoM-kmiP&0UrM(BWgDc)xJZ>u|g@1+szu#Ls{CZth$eTGEL-HhLZp%Wp2H+8F zBTHM{>La$$t9|eD@QLz(%}kuQ33xEETzLo7*cZ+D4CzHRN_@cnlh)87W@2bqL+E7> z_>=$dSFgIHU?acKBb~06SI}C~Wv2c+Y44w$qfsuF)c|R15>Ps(PqhDd4)$JIjP+pM z*!>zibF#GO*R|7(&P#@G);dze&b@Ws01`Q6;?|T?3`A-60Iv#g-{NXn>!X7Ob*ChU zzTuSyk{a5uW<|aC{Vyb+*a&&a31k)+? z4{1dS;eDYE3Ygu*8s$y0$z9#SFZ4+otInn?q;_CHU;#H1Tk^7#;o|Sf)PVF?Jx9=+B7e9H3hR6Q^(u3iH`h&c z(oXJHVGdm$@Nk+#ERj7BFx=6$^gJtrGisavcjL^}G@RDY1tz8eE97S~AD6ixf`A^a zMu|}sH7RTl0@?SMs zjYSyyz&aF81|_$PocRsX#e&=Xkwu2Buz=ZNq!6M199FOxwP8MW*F5&-w6L%eO)Ct& z>V|uit-wrvRdfX?MWGxl(HF}uFlw8n0yRiA>ZB7`Q@u_I5fV=-s_peJoUbSGm;c(k0MpkF$o<@q{|suvtg6GjsXsK9rTxX z#ts&_iVuqT##PPSi|Y@O6=Mhp1P_?YQz&LVMG~pVUSNfY3rE&K(mL(DcboS*ZVq@Tl>wW5>m#~2 zI%%6CAv_(-V>9QfTPHVTqJhU?{nQ4~6GCDHN0>=l5%%0qv2f%7xQ&)C`(ijWeol-T zD>QzYl6`&<1Jet?waPC9ZSj_^`^Z!uf^7xIpig(IkV$r2q7z=7p1 zsgC67LZe0bTs!uUB-V+uOF&_GfOE8wVv-S@k&EG5-YuvB&H;#WG)UX0XG`?CNo8(p z=3buzB*)@zf<@oEyir!g5#N0a@0R6&gDtSoi36HgO*o@mmuo`I*S1!ckho4Pb3>hB z6cy0XyS$P68;S5hWuP7mZ>n(VbUlYiKKJz78YSL85@W%BAVy{^*bRi!5^eoLlN84o z4T#`z*JZN)70WBxX<}4zDFy_0GLTE^AtZ&+Y9u>PQsDJy zA!sz+laBXI8@I^=qhqkw7Pb=77Zbu-xCNlzwKV{C)*9(3 z_nVXEwb3zdeJ&}|I!T(iT85(cS*m>DVvt5i4@pvB5BA2&FcIP~tXM~gg-nuBaa7Gu zrAGTAx-a^S#5=Kr0&A854oW)3KzU!vb15EN3#(ISb%hyN)f%MDibUUj6?Sz+;Ife2 zGf#C9e*(TuE52)snr zjUhWszO6YFlSz?uDpD)Qtl_)f2c{&-4@`lKQdGGDGY;7lz@~Y5b}Y1uu7s zI@2Q?xBzyG4@?1}*6=eVliUQVC3~Cl{_h3;9P)UeK_A$WosZd#htvAA;#uEOnbRb( z$y{{eR3j^-N(R}C9(NSZK=YDIWA%BeqCT#d* zI`)8&BXGqELZdl7XIyhJ# z*z59#JgSo3O8p^bOq#z|q)}GTy?!}#tNglOjH7efhK?ym7^CNX?k}9kf7MltPm-mA zW{@wep0WBtd5NEaTV$MT( zn<7hPXoPAB`TDojSU=p~w9*~qFuDoE-3UI`3dybudg!d#m7omxmvdg&VIv5m$8cge z*yUzbF$kkCSYRi3u|m#sGcm}KeL%$uIivsYLeC=eloO{#j+G=5*vnC1QUOZG?jFB_ zqc+52^*IK(*RwUCK)Ux$Bm#XrbVV+fgTopfEc+ko?d;I;n-300*_ig4q$u{<0rhP8 z!cKXQ2gbrLgyK3f+VvZGk-XN!@Wm{k?DXxk<=@Y43~cGqKKHO21E&qYZv5Sz5OdRE zMw2E%e4Bq!!EG0 zAkG_mJfn%7_^NYEln3Kl7o(+VbPf#@U_n?9SEJk)pbf(idL|r?C~$3qtl>-VFnn6* zm=$D4CH$!7d*;@HO|;B;XH8r^qTQ(x3vF=-Y;qQ-Afiqc~noU-k=UwxZ&N z2OZqN4i$Cev zt-KnX6Jp<_J6myMlL@gSmRR9tbc)O2O|LrdK(?wA^6GCYmP|e&1v~)DI7VMFOV++4 zZqp#y3JaS$v*S&yU@>aud*3NGXOw^47jv;9jrKNE2Jz~8ao2$#jgJC4+7pl80Vam zbAo?DYMpq_X*Zd38Y$*FMXpehapV>_iyELj!V*zb4n~^#R-l}RWVKE%3O;73kjlQ# zrIXu6G32@g;?YQHZs@MhrC@TVdn3?5cFN!a${N=xF?RiQ+QVswNsl{n z?Eu^k!`37Ij&Nw^;hK!mFkoW=Ni3U8eiPFZc9y8|){+I35oAx@pPm6VG*%rrafS-8 zln*Ecildm76p5iCD}9m!i#aQ#P^XN3jk@~7!A9Bi-P-ZY7Wmll<81EAYm&6$RcAd( zQdIN11&6o?1OBFtqRxra#0NxB3#!}|KH@`d%P`}2-{$eL(~YdY9EzchmRl=hQ`@65#oW5w|Bey0$=Kgiv7eF?%BZH8aF$ zCFFlnZz9_)5fM1?z@Ij;1qUc*KSfHZ$SQCJw{VdCITsm?To<}Q{L^RsERb3(qu(bZ;FC&?qwlTD>#v#ExVOPVDD*nm}RD6YFzNK_`V& z^UHz@!U{bKCC&00a8TBeW!{xMJg$$pFSsSI^;`3Hp$GO=7QBX(&Mb4g7*Obe{hk$T z=?-4$Oi;!UchJ?M6#r)VS#jZ%3cCN3ZcN-2o*~9ItZ&B+%KC_P<92#np8Z6!G2p+| zr{+}8Y4f+yfH?q4FG1d*2zqwWI+QFh8Bp%u`uVL_9bQ*NS43GV;5@-yrEFEIP)Bd} z)#*w8)IIL$0xkD|s0h4Wl;0`jwuRzGwc$2!WqRD>dDuZ6kjtXH;8CAAo}OMMDyRdx zYXgNH>`;A$lm{P{q;opC2M9!imw7Cmo=rEyI=WerI&OP-^K^B-u-!FI90wan&C?s{ z6zOr{av6RcEUb6QH#42w^K^|2Wf$t@Hbeu6~X(U zyl_`|XIQg5Ly#d(pLj%iR(wQSE$fvhd33_2!RfIlNhL_Er-v+m;XaKL@61zl!0to; zC)j9(tc;SO*&!Gj^b>8n$O=PL!=EM`W0ELl4MkQ1 zza$U0DYkWxBZQd zY!HOq_tI(8wZEk~;pJ@v#xKu~I!{uZc+%Npg4leD$)y0rdL_yA;%nix^pU_-z*D%* zeJyw+&73=4o$|{}h5(jD+_H(?VTa!;<{Wi}Eoj&7HP6~9hof3 z0>u^_TRRv4D_yWd@!wK^zUUnzSe|e*eo4xm7%W#zz;cFSPEzC;6?saOq_`zXmvo0+ zR8>psf{$>Pik<{&In5x4jkH0H{MAglB;Mm_=mu%`G>tOFKZ|Y?<*BqYs-^1$;FIu5 z;k<_%$4gRRrtzOK(YP4AWPOMHz2SJ<8d4*=D{Xu0A?U^(fSdl>r1�?Kn7BFV1$qHx&Yi0+iR`=WI&D|9V;eGnauOdJF__y=EkkE_^pa$2X=$WVMV=6{_M1A?q$C3V9c?r5W-S-dRh2rzOv?F>qQ zB5;5owVGmLDYBf3Y>=WFBxXRV_jBUIj|6TV(ieTOFPlx*Y_SwJb!6w8tdG8h*vqH| zCpLZIWkf{XC%fj8{#k1sFCdk(Lx87cJDNO8ny;%VcK@vMfz2hBiCHX+U@9q8EtG zIj>)wRh)Pu!3sflnOmV-x?tJ(GG4hTMVJNxkts~I4B0UZEvTm6e^w1?-zkpf)G@yE zA>d%ZXgTiCx`o*=@u&05Ni$wVh!bx+TOrbDW*&2*h3R65z$St8Ia&Z6KRu*E4vg1z z%1rlYVV=Z>c&hz|gdH^O_qi|Y$vQMbWWk@AC-Ae$t#o2%)k?ZQrsG!5jhfXAB(?Xv z=fi*N+^fM~#knMZgnd;OczM|n+x7aK_4S+0tJ+^RLX1P)W z2;?DKonhe4+`w2oX9czaH~;a;2DOd*feq4*@H%C_crO%GWT~3ri;QI!Nn^n-P@JnRE8`$4O}rz?AX|MTyQC4Hgr7+oQ6^WLtikU!;_p4CP4x}1-w zj{w?R;0TL_0@rfda0lLNM~l<{`pte05hZ55*#Gpx2VS>_Q!e7fQ8_E+BB(9a?RH?I zfkL6r1y|XQAAbuYXA2}OJm&DGpZxC3Km2_`0vT;_1(Zg|5R8xfFC@%B)-dPjNprdk zHnDr>b<4?S69*)bViG742Lk|Pt^%-+2I z)D}x&a8)fm?YOo}J1_i`*yxvL@YmfWaqRq3C!Rb&8fri+PD3%N6xjfjVIF#JE*h{AtnHY>^|q>PMOnC$?0zM^p)ES~o)`t9gCbLK!R zlzRzb7)DwstXJx!t<&N--O_Wu!-x+({~cH#+4M45_d5IsSgjeI_p250?&jQZqtRHO zD^Dl8oH!6tXJQ79QVei-RX`QGo_cAyB2|&3*st8-qi$9lm=QHg|3fQMZw)94MTvKf zvPOmh)J}P7h_TAcz!q2uS+UuwJikWw8ZJCkHUZNYfqNy7lMeeo8W2>_K~2vt?`QYK zi46}TYf6JO-L(V)ChFFq?jhTR@aL9d2uDLhJ1c~bYQ8AnEi?k@bM4qal2}V+&CdHp zAo(WX+DI|U7^sOvhMNZIS@>HWKcyLJL{BO2dSFmq1FE(-@gSt~g@!^Ld&z6OY{1w< z=4CzF=el*s9E>_~&c_Ikr_9OaMB~JE1`65-xEnkXcCfx*HYba(lMPjK)l*>NK zE8{^;k6thvMYBgjID6n28INs)kc9^w-1NP?gKp+lgbgU1IP`AC=FHdeQv~zBAv^u% z*Un50ncp+3Te5`uW9iS9QMEIx`ABOJ55>qR5m(J`6V*o?4Tuu<2#(3)I2S+&w{|98 zsgF3JO7u+)xfzTtsSeN>vkfqo?2Jj9-Qn|*V%_h}*Thbn&}L@n9bM5* zg+A}qKpJ?G>{FpE*j3p6LqfhX*vwad>3~_~g`pYJkiXE2{ToQ8-{(n&P2T>9NFOhXlytvt6;(g{&3|NqK zDiRkZNP-{FLuML1GMWu*8G2igo`<-}YpW2G=V3)GN)OQr$^@s!9`areStsrENwV74 zp_9&VfHHI#y!=UaK=HckUNan}V<(=6tSB9qiaQ|@1Q>FdDwuzc1SW;(05I1iU-A1{ zwyNlgAGi6}MP!nFbFa*~8?;S4GK7IcLe$FvG3+}c?*H5O%u~~>WPq(zEEFCBA|a$^ z#_#XJfRm)_EpZEIQNnW3F?ktp{p9+HC6KgsVcZJHn=Th&;pqCw znCZCM7w>lHRMiiuB#MyVe)b>g|N8iMpM3G}lBE>0h$2zu2sS;p7oG>{=UZ3atB@O= zm9+7n{+=W`v9qEzaaOV^1`-6)sK~FT@1Jl=xhUv5jQ(Gm6y*Uc7@&A`ZSroTHwKi_ zo9SHcBC_0V|AbvqpmP-LknEoOSso{weyED)UH81?<=F8bhLIa&+=hRNm64-<^x(f{ z{Cy=Jt$wF^PM^!4vgt$I3@!#$3g&!0$FWecJg7DS$nv8OZC$hO+zj(z37hb%6UXGO zgi)|!rgD|j>!p8PQ1E@nt&A-Vq)oW!pHM7y;Gi}|<#nEU zVu{m6S*>KeG)VUa6muas(yS=qT;pKz574(JI(AiSaXf4sWsA=`vXRyFG5S;Y?;r9v zx*1jPtSKc~mhu6dctLy6#P!%gF~t-qq9UW2?yx7{YUCe!=elH3&$kUF@}4x>vOy$j@&`GP#!n50Wi( z4SkS@WjA)8Yixv4zsa!r)84SM4yAYRe{a#N9!g7_b% zHz^Yw45*PMGiAKTq(-*UbJf)Jkk-(AxH|A}yTibA_=XM&(|+uIW*rn3Vy>fowiR%G zV6-n^=Iy;rjbxG78A=;Pccw6T#Ev^7hwO-knEOAV;yfA-Rpw1SZPr9TP|Ad z3oLvU(q0hd(kL;GpQJ#tatsc2lX9+}L<>oTbwFv3kzSnY@M1RHo-foZpkne{HeDe- zLn>!$lz3C0i;hOZrE0lWgw9ZpKtDxdl1mL=>jU~^_>0tY47JCGU&2=W4jqVA4b|Wm zY(#9_%>UNB>I~(L{I0owNTI%-bN7v_xT-dl8dDOdJ5-AU>9d!z>;x0+#d+yc?sMe zZob$qffdJ~!p5N=lNnOl}_stTT!D-_3fayaZ!Wg@@$OO!~U@pqxW~l z9fmrLJ+5)VP1CCM-~VO~sN{M*@V+D}6_-u0_4_}_%YNCVp@HOOy$=66JCLMv{_`Dk zDIli>i51zOy>F&Z?eN;})*!twPA88F(oImW;qCG1m8>n7s$yeNxb&)EueLQ zGmbv18a2cA_}tWA$aIJGH1b)%{)KzDbh)g6!h~AG^aFQylYDWaZ;Q`5s1d|x+UWGL zFQ}@sAGx6+Xz=+3ud@G5OrKzdpwa*IRoh<1=miNLcU>my$B>;S?n5rc00Vvo6AOMIa!9AorO{KZ zpqGa1@NwLCdzb@a`WEeY#%e&E*Q=Z{k^Ib@dx=f#&Uyc3#0parhH!G}Cb?e2cB^Dx zK&SkycZz4Mo5pul$Ts;&vT;(nD|{1bWGHg04NGt@_f3MW(ImwZ@4`^W0bP&Hp`+c7 z(6-Q3@uMC2#Ar}TT+&aIWlp?1lxBL(MG~k;9j%jMWd_oKw$TMt)O4&0g8ET! zT(&&r#o2E^+rk@sv5Cc4zxc2NTi98)>w$q6UQRdc#36nwbi+Ej$xGu~7hJ$=P%N2z zLfZAFZbFK*6AHW1#C04@#&3|K_BN2~s2ZeLRq9~VU>tE4sOk4g9Eg?YU0X+u=wi@^ z<_KXk#k_jhsg^JDFY(_v>ENVh#hG#IAf>WZbzmYC0UZp$0-v?xc1(XTsS==*Mkn*O z1!<;^(0R=izu8GZc*2VSO_QI1HDTrw9Jys~&wz@Z90kAU!w%mQG2E zRAr|m;Y}z+Mk<1YH#;S%N;s*5lr zD6y|>TD>pej3k}cO-iAbP@}{|?i9$-)?a`MWzNn3%x*eKGA0fWIYUH6{|q$P&zFw; zU^iHwbu4TSjLKYpeErXjUhGGwsrSjbG3369v)Dv2Hz;zIip=Edu~wc-$Mb-E9#m!X#g!b4kzepQKI5)O90jA6;U4T{8gTPE?eBMaI#borcd!tqc?da`r`bu)2B!@*o@-9!5nO$ z`>$CYPgaBJyk5ad)x!s``dTT`w@n|GYvfmsyQ1H%%oonbM&v*(o1jtcmCqlZMX`zk zix*fy)?yEiX=DZ2QR?X%kDIG2IBj6pieAE1QJXs^z^F6bi#bhRV0sP&UUl2)^FKky zWiu)qGx%0}Wrd2-9b0|<_{Ldj?#1nW0m%q%(@+0h%MFnyxW z(lb`wG!}fH8rebEv+m&zpz<0aD`PX1`e(6`Ss0D_sp|^9)CdmIebHwmo*g)xIFqW# z1RUuU15$D+RAf3+>~%M&le>R<2aUTC#i7Sx8Pq4yGZ%|0Acs4iyH$Ejct}`5CwSg| z3neN?1SISsGW=~Ht0rc^1)-{Qr}?bHCcfgtA$%)ykdA-JvoN$q?@ap0^o#DaC8=(#_+47g+L~EUR!m$$RPYUS+ zL?l7VZqb}3Mai6p@-J_Z?Ez)M>!fEn=e$y+CsgUYD%#kn3%xz9&(Gf2u{7jE!jq+r z?A+$2UNFgAo|esgw@Jdd9#;+CZ5w5&(E%+-yO_3lEJzgUjjU%Sz*1 z6LIq6{UpOuQoj@X83#<}pJIvu-*hV#nIb(7LZhV+SOcA}DGA)%@#v5yzpqiY$$Auh zE-9Wl^eU!>yUFVoi1OSDt>^T)+*EXvvs?_~Y9L&K_dp`8@3Qgf6A(HpIBgDbR&7DX zz$wP!u>+7|eeRi}_+FQ}2seCn`uoEcnn`6~`2lm72V;xU_;U4+aP1uE#>&UVp3{s^pvrs10a^ zyv=fY{p4uKFt|u`{v%{OY{k>SNoG(W*y-Xg>k4x}HM$|UXRe-1ZnKM(IkAJXZ1h0t z*+&%9MUhS_a_!iwZd=?n%2iBy2(nJ2ZwecUxl?k-e+^Oy59MbhT^xwj*Kv{*25xGN z@<`Y|)xk+uMYxZXCe|0U;-fXPsGxkw0aE0tQ5FKV*Co;|KxSJcI!3RjN6@6m7whvS zx%iMiR~Q|nBt@3wY#?k=`#$x)BT(xmU}ko#8+t(M!pSL~%}O0P8`ub&o=x&5c`d*e ze#6U5C-k<<`>c1Y8{CKg4&W98yhC~woff3VY%8oQ((5|Zalmz&a$w@sVD)c?g1vz_ zbkwom(dqT(wa%}JJ7c9XS{s(6(C?9>4z9W=bh%9XKcvZGBPQ%*Wo+-}?#|A}K%+e> zX)BHkZO*->N+9W7DD4T%rIiWhFQs0v0akdIKGkKhyq?n@(#S+C;At>qcn7d+f z8*RUxXB&}Yv(c~}8-w3@;YagJ|Gq*j$0TvQIF7y;)xeoQSOUx7FK<{^Z9&AaK46Fs zk9ouWlUF^dqVlf!NI(r~j#k|HVODetDDo}-g4hg8Q3X@%-w;K(kGY#D3gSAU43 zHnhiBt&6BD>pNDx>cm~qzOlS!d9fm$X@;yjQ(DNIip#UhCREY=#E>_Pw2-Oq+!!HR zNESFTohgPk{5C(u4kxGnwCd->ks5X4^)pbg3=rnXq!^G=*hEFfK=5{ zMXlgSAa3Sk#mgqIC}9q_l5-bG0(Jr!r$&ii%}ri0Z#}3p#)P1sRK1m)!cAMDdJc@xmv`i}U zp#e&=P?{FS2QKy{RGkL1zxPA-&l z6?50k%9sfJTH>y^b@YQt*{V%mz}q#wQhGEX4l2da-Pq~Z3R7u2Y#48$3#6N-K6+cD zw4>!2KBfl(++K)=R7+NlXxd-Z?&ivqPP>b2MP;(m=c%_wiJZd?(v|M&E1a^~4N~0X zQETb4;JduF!L=gT7c`V&T%K)53=A@ALx9I1z1i;)J9xA#|0Gsqv@FY0Z~QA+#V$DS z#LHojr5vyv-b67e6j@J2CWS2X&?uKdDPlZNofOg`!=l3uSuUL-g@ROrj2OZyjgAuL z3Pvja>a1w$Uht0+b9^|h`LHtLwRAc!UkpKY)k5LE03E-Eh8i1Whl=c1he1X#jWRW)p5T5R<_@fLuV>UKBh##SOzZfqa;#$C zEmODB<@{AlzECfdoTb|3jdZtaln%=U8{-g{da8a#V zet>LZ2Nx$!tKVk=mqLmGXC{}5Y=IJsXsStZN>xP1aVmYHd0lT=YO9Y(QX~SUWSsXF z_x|f`iiGg~=QK)u(ov>q*m!w1!uCB#a{U|MoR`gMryna!YnUY6sn|a~OCLE^-w~Xj z`CvL!e4;_2+yB%>=CO$a9)ou z#CvD7`DA0orl&*EZMD2yw3OG(so|Wbk3nYg8G_rXiGl~xN?s|I>vN1#O7-x2T`HJ0 zq*aw6K1$#jX^We~;ldUs{hps=LBcX{<-DF=_P*^}^Q<(d^+|wCV1RX7O);?)Sx!a9 zi4Ta1MCGC;dY(bU(%}<`WpLR7Rm)F0ww0Ad8~yt&S?RAjo&DB;0%^M|Qo%xA`qGG{ z5%8ly`i&f7VyK21C2SLEAoUbkyU+uvXNurFhQdFNHfXkhq~C^_{B$fv*ahN8efP8L zzkk&uX?4o>e@>(--|L)nYmP>_^xHty=C_P2Lq%3_hPkCGw%=gAH3=tHK?eQM(5CKjr-E+7shFUQXg0wQP?2S zH(y*HtWlMMAkRG+S`^1r09zQ@*8?7TcYl53UE|Va$=aX%BRSy2TfkRMypFRJbBZFz zQKM-+XQ9xT$AH22Eu1*cGT~iV??z3}rb}mPlyRIcQqMr41JqXaxhz*KS0pLUzoX6~ z|37He_ki-xWWphyb=HFDS_h|@STo0&;w=gjQv?Cj3wpBdTRWbT>m zog5fMbkU<4Eq75?|_Zj^2*(*@mZ#o3wN zp{ly7x~=c2_j|whc%H1VROZ0~12#mYGWRqaAR`RX)+R_J=Ytv#wgC1rSm7|_f@&Gq za?t_&pGVN9QXGkA2lL1#O)n>L#_%O!DTU)r8eZv7qO4go@b($(z4iVcx!@ zY0>Y5PT)nN;)o8-y2$I~I=TAt_JF05Yjev)y)(MyxZk2L0lMVjPmNlTFO_8Yo3AR4 zxcc$|$-e1}CF@n4;X6PSwMCz+!*<}s5_91%ilN$mLyI1N!8=!&d{QQQd~5!16wfz< z#|&k4>PXPrwY~RCTQjFvAa(7(F1$llj7FIw+*lA~_(sN8Hc%`?g4R+oC^QKb^HOXS z?{@E@PpLlxp~x2ff#40njnZ62H&ahesriqiT@Q8ClXxLu)H^umAh&J&r`@kE|Fs1h zyWj9olN$o|cvo046WpiReu~_oVv1x(*^H>S(B(zHsZjBLn7L>g)<_-F^=x-g_Ou+u z1Ex!DxZ-_4bSNm*Ek}_qJ?-1+)1mrU)vFwEhoap~MThi;cN>TaruZX=L5C_u91k>q zH@r*8S@3D|>5gfsZq-1%zDpGgg-uAu^j5LBhrS)r7hzZ`Y4XgaPlt42O>H~ zU6UH>n`Y(%p+JFWmm2??84pT5+qECb^QYD*uFdVDyOB>)41&(Uf{i!W{(=Du9n8gd z$j)BqnNLpp-tg(vB)J;;$;qHTkzlzAxREwmVaxe-cj9eXGz1&v<>19=7u5$Bh%)Jd z0OZ9_5bcBx)7V+%x=zg|l>xPui$ym>&Ab--K`s}fdfE#R)mw-&%`JMnXdyJD-={A{ z9)Gb-ipMg1Ydn%>qWh00BAHJcEcbUQ@c$D+iX>1$_p!Ka-k{6A>6<+UT`tR;LO{C^ z7Y?YHS09uYRv(-;ZD{Voig_#bn@NUtp;6mS*XlCqebeuR4!N{w`&9_DN?Z3r`56V-{ZpI*4U82qp|O&bz99#Syj zrr4l>YUF&mo?=revWALD1ng{&JSC%b4Wg~u)X%n<$(`}Pa966pt$_}Q1n4%p8GBcaNgfG|eH{10nZU3DUu+Q(E6zr0{ z^hX)}w#GuR6BWnnkMDbaU^@#LCkzLIlhL6{7cC9W1y`nLE*#CL(ZiTHf7vd~K3@7v z09wso!3(pKbYK706x)IZ!9s(h(gF#%wx|lk_r>j+4WyE4*JQg7&;>rbfITunaa>hl zV#H5Sbg7|SO0)-Pu=D68QSp*2W@C`k@xl`(j{xIryzqhv^~S83U$LE2c%`lcYa6+k zjFEK|NkoH$*qA=i*0FkIOJr}bYGC~i8MFP;F`e1@(D5FfSW#s z3ZE4JZg;1@ha-MGuUB%!b5`+7bmtL#wE}tjx6Pe7y&n)Zni>9D#fm$NFo)pEG&z~>bfQ=Fkx5tWKJoR zWkClS5IXcSXOPF(xSS` zcJI_MY%M^qDie6QQW-3f$)CDJJ&}2I`$9I4424s0FzQYJU3&g)Ka0_L$R1rmjL(eG zsIoE|+b9;AnYU0e*h*SNm(uG+l|hCYah)WWZfEa9RmFhV$fEjgUf52}E#P)Tsa*)R z7enA2G1qxJ2}XRZ$s^p@Y>YbM<2?A=U%h_n^O=_fJeP1V zFChn87kvRp3i4?4FmLu~1w2gX6k;A5`-Xm;<2Z)PxgA))*H;|!wOEmk`JtCchJe>F zl~#74h+<(D$%V=kSZJGD$$@jc&%`)B%)mF~a&Fcc1(asdW;JX`kX0&gK>c1#h`H7h zb(~gt6e=()J%;K>yT`&lWY}1Nk-K((6TcI=kuh2#c=->%p|LHKe3HCPz$S#FrgPsE zBQPqNV(UPp(e+n;R$JUHZ}zTS|Mu2G%)9>eu)zmvzl4 z(4~82Y5Pg6tW#G`A0#)FSKSY4I(04j>X6EL4Zu&4Hoctgn$w~$hx+U#;G*s(Do_JeEN0oZta3vQH*Voj)zZ~DLV zvSsecWd6O7>=f|a^Rd<3bAV#&DYB1>iJfJxSlmYE`NzsmL0>v5kLJ>g!rIvk|6AlZ z**6{cPRR1z>O{@@Poo>Ydt<@D=uduF`SXhCQ~DyhAp8^gzUeJ`^BE|0lHuP+-zB;9 zp~w?HXXNEUjp`PCqdMO$!+#*EK{{5a3xXfYKp9+VoJQ4q^+u@F95+_t}I?`t^^qGQyj}b(XJ;ymK%@r8eUIc zxD)+B<@vMvVYfDHYC75b964(RiK7&Ih$08D{kMnC5*5RHn-vC?&B|;=lIToOn%F4b z0Of^M^D01wsFvb`;cyj zqO67ZE^h`J6O5)WEECm(B-cJsqhw*TzDV{NeT-DQT~rRbbVwoAt}T*mlw<~;(vJ;` zcA?4o4duHxCS37ClW;4W6}x{>Vw(bd((M&+w2dQQiGdWO7FC2 zeK0Pn>q%|!*zhEHvcTh2AG9{$v$7WK)Mel9$M9_hLQ zFNYYgm9?G01BN!bGZf>mhl$Yx%Hx5l_=e{mkWes%V9H53woVSXBV|*Dn?n*l_QHYJ z+&2yw3}4D^KSo&=O+Fjx>uF2i^^Dk*ax19iQY;vmOmk~li@rXxI;07lr5&1Lvefgo zdmmY%?)55rxnF)!SxeH~i^a{*R@NW-FaWo^=5*@zNp8{3uc#R_s)mQ(Nq^sx#Js7p z7#VrD{MRIjUjj(L{t9TVk1$DD6bohD>DWQEiAA>BF5foN3vPHjQ!LK+yW*WDzTk5k zn2QT$v<7t09f8mppTI(ckT}ubNjV;yAsrJ$Z0_NV-Sdl=o&EbVCAG5aP)TtAOKDId zn?1b^3Qj)uYZ99@YBeW(K5+fOvq07b&4;I5yXd=f+vo0AwrjIQ`v`Dg&(5FiYzJmc zIKkP<8Q>UkPxTb=+rsw%rXg< z6BttGKz|jq@0Um1@H_{WWymE#(J^P4dpmId=QqZ|zptLgmS|uz=S!9DqplDVHySCc0De?&)q%J2|gWibsrEJaRH* z39NONgf&45pg3YcyoFBm$p|~91$q_;3boN#m{6enz`6Qx;{k+YAz*zsZs+zl1$$nI z8n!BnR~3=n{H%(AL3G;6svM$NNG#P;F>TU&pz4xM)=2u8qvQZ2o-weJ3-irA023b1 zg$B?4qFC8cvVz&KH20mhG6U|Lrrq<|?9rlcWnyIzZ`2*|ycmF`lplEQhrBe_%N!Xu;I5yyZj~B zHpI_oO(ft6iX&5jOxUSYK#?G?!?LhOpd2bt4!L|Jx*UqW%2r5V z9E8f+1elkQ4Xjy<=M1@^X63T5Ug#Y!lnuG`kwMKJjrqt@pL60iD1$jS-+Vld?NBax zv6o2zWlLOTtq;6LYM{G)dssfOh$V>{>Gi56*3he|n3u{d30Mt%=M8}ef{Wls!~r)} zf+$DVCM^+XiO`|S3cI7hV_*rVe-c(B$6`*r#WKkW-DZ!18CPh-7EzP>-rTi5Mj$!E za1pxgX7L&GNw9>o3xl6$f1L37dvRYqY)jc9*xCmUdX{EwNBC;skt_Ay;b&6OD-qX% z(LuI@CTY1J)>7Vhe)$XInpXH31b(B<@807%o7>YFEsVEeQ`ygKiEy8hDQmlx_p^y& zp{(kI;fOJiw?K@>zd6Jl1s-xiij|t6az&i1W6rUez=DY%7E9VZ+L>$WQq8A96Huzj z4^odU#dzRi&$XAM(-t>=IzFfJkSe;??3OiXNcBDg{bomMr!{wG*Sigd*=fYlfiIfewtg>G}zb>^szVU)(p91299u??1tat z3CM|Ck8#ZnzhJ-e=K?KIYxwe-N|G}g)C3%;fhx@rnWP;QTS}2_R7{erH{voqK=;tQ zU;m$mubl^G$p(6rC~bP9v_=1scJVCpF<|G`o)IC{2TE%`)$P zvUpbes~6zP;hTL5=ze#IXh4GgP684pUVZsgz{Zj`L5$e$^(pk(HT?;Rq`1nk}9 zSRo>fV%Jh6iHbP_(Txt6SlYE_@w#reRB<&Vrjf7@`;weGn%(MmEcswsaHX1R1awflH;840r$hR z&WU^6*D1=p&oTpk^+7w8y%GC;PDW-cCI%|@Bjl0m;Y?sWx;HO`P$wtlrv27}kmV~9 z=8>*vCiuI`D){@5Vh1R4mx}qsd!u)G1TunNgJN|QAVRHoqppWWPZ^>rg|Zs&`k+ef zHhFJEt^0kt);(EV=Lac-J8tP7sZ5XevXC|!^Ve`_KUoh-y4= z#GzOOjP(-j%&8eYUPkRD|6Pz{0_l?KY2XQ)-0Z`$T|xLu@YLA3MRX-7OF*VvnIOR% zZJY)gpAjfv(W@+uXb0Jf3`k8K4Qll@;6;0MTf^gI$-dR$gHT3@KU}dV5dL6aW*)uK z`zYB8pIixT(OwA^Y~WWX_WICPyH~J*Uk){1OedJvHppR{PiIa65fKR8LF!=p3;VoR zd%>X81npFqg&hs$v+t4;@h+K3+W=Sd<1FCBs&K8~@52zf5piyol`{BRL^R`TD%|ESpi)?sirZp%k zAF-k;pjeabcGwV(@SnF_;R*iDxLo3;WE8BhfJoj?A0-3y0;ZUQ-b zuQ+@9KJNhr7}k$@pnN<=hYA%K6LNy}fiW5nRxAc+l#w3)Ayv~qTn(^bi~-=evAk;V#{*;~%*L9ydPq@jqXiJMgmi z!VQQeej;0H%Wo=J$AyE;JH@|Ui)9Rk!vSZZpa40e9MV2wGvc4SYZp{M_b!e#(sc?HUTBkM%xZ()`l_j^va>)mhB7oL%5clIKy(_!7FyVDck?-UrXq+M zx`AxNC&(=ew*A zvY42eneQc&JOP`S{Z=NXf?@&D+o_lbo|$x)e3i!m-Ey~D`iKrwjrnw%M2Od36?&Eiz%jE|u_A|WJQm7;*P&<3_dyh^>^!%(1n z2$C<3bfz3>U>xa8@k8_*>wOl4T7Z-7pOZ$m@mmiG*AFA00y|@g1b>u4u#(+a%ZZ`!KiQbwA$eNp;f`F+1ikOphRuB1JpNdbYc`#qu+8( z53Gr92rLjG59c*?V@Nw`(cGE2#Mk@{b)wHAuS9{-t2{0PP>}I`Kz%^bCN)Qv&d<&U^1S<^BH1C)L5)$nV|Kgdpx-rRr*uGk2V@7D zMg3$oTOLspw30nPyVMEx%wB`_)CJc6dNj%Vx4*hLp6(~%t_-+%+Tvt;UW8ISunI*% z#VRwaYoW|R&I+HI$36h@xuavw{kdOd=|amK7QHVjj9h+ZJmVoN&$ydnZ&Kt26$5=l z;tfG36i8at8h%@`LwOJ?Y&tZVGt#AlntahM(x~Z{+zedi-l8uM;hDJKKp&zz^=HWG zc_s6VFCS3!%)q~RzWE+T?H%_n`5k2*$adoS)o$kfT+4o^ex2+fRP}wz98eTPM`zEB zE*g*A^1h|1c3ZE@7d2|?{5ti^^!=b}UFU;m;V`y)^k@x(#JD$D_b>edmLzKeb@)!Od+da#6s3(*r9M-K5wX6uCyl zG(ie!!;DVdHQ@GZliq;^dc0d!n7OkKs?-mYQtu{pDq}tfO;M(q?lKU@W6^pwF!&Wm zW5L%+C_o-^>0)kxYjkeD0aKmWLWkm67`p<7d61hm`%kr!UM0r7ptowk0}Ge)=>1T9 zfjUv(pcROKLCJK*G7y9{YE5TiWpb0c!Y41xsKdTJXw1Shdo}5zJh~?LzFT-F8`%t#11 zt83Sql#OwzkEJ!oJ=??voPcmWjVp0eV@pOvmV2_u(CG$#^){&OaxV6HVBOYJ83Tgt zzvR~|{xUco!cGAbNdTALh=}qdg+HV#Ldi9BJdR5CORB>UDf`GR8X@Cc42&`=_QGP6 zOPzWs<)gQs@H7M)m*Ma#_Q|)&i$&!TINf1rFjto$`(KC=!#LSB%1#~6(QnRgqVxOY zmtS?aEO#njdF>=w^&E&Dj(o>5D0V#s|Jayp-Cl^6qLK`BzYlnP0#bH0A+a(;ia$F1 z8@;gxGDnBS*N)BdjsDLxAGK=`M%vz|-OPcX@f!redK)e^zxRD{7#%bJ^gb!#7l0A4 zw{h4C-!63DgHOr ztJzK(YC2N<3l#`{?18IxEA_tM26g6;F*giB>+X;XYP1=ah%UG#&8(c)4qXD};J{SP zygJ)aQcS#Gf{`x-*R}&*h5=8{2evEQFWP*QpfFR#x2mdiJ;YE396+ceKH)xvo%78T zu?PnY=opHiRE6h{dUbD{RW8ay6kg=?7)+Gm@C29QDHUge?c37=#@+cMr>a(br>gh$1K;WRt1`V%j z_iV+PAS8O}QXitrG^L6&WUFV3KG6-A^F@)n#Mr8Ql1^kwBxRanRli&J6bG3IE(q8I ztI;h5*Bu>m-0l3sfBd=yCMIRN;AbaQRFxk^S);n{+S=*Alxc8~7(PWgu;sblaCoZlG!^aOB;~Q^2SMU1$A4^`dOh@Hj58faR{H7xT z`z^qbI3l@shGLs3(nQ5{OoO1?&-28$ye}%-2-3H-=yz!Pm0QAhXby&d8oD^@kaqBe z0rx={Ob%lWc#9roN%4xF$O2KG?)sdS&tDB45F0_ZIAvZSWN^0q)vT8RZ{KoUE0{2ZnNpA@&2;c0% z&oknOtk0Qmp8hc(U`<;2PxFHR*@CQdKQ6ydn!jL#c0E@3X{XrB6uC&nUAY3YUx4AN%>|m5)NGmF>3K7>{H0mwKFOHLoQo9 zPSV(@+YbN51svYbL_`h=17HWa;*I>|`Sc>mp1>iO4iH;O@*H4}kzD0TdX4;Pc5ec~ zG#b=60mi);TlW3?7i-uzSNHX3?rLugy*3Yz0vDe_XlR3EUB!PyEpbc0uCN z*%PU(wmmg%T?=$|V{mVhgc=*#cX(eO+te=!>3PgBzUF31y z5pZRkf)G1FFxnLG=-5+7bBrHy+l*1lHmNA(ubXVS8u-`|0Z%#{loK8FJ@ zFZSpPeHaREN}(*2pB)H}XA9rPwtTNyN1%HfB}3 z6^OFHB4^U(N-?M)lU3^KCcHbw-bX#rWkuYNdX!mTxRZ8!`k(J=Ex6DheWQwG^P6S_ zysX)01@BUdg-VwqDyE!HQ=Jff>}}MZo4=JsLLdyDEll(+hfL$1$N>hr;??U_o0WS5 z&@Da?ks<33y{^Iig$?vw(nF_0)x&#Nm@-%xrHHT3Nr!XStNKG7ZiVcI{-a4`^bJQ_ zocl9>-4bEJ%~vPd~U%m~;uf#l){*W^Qrt))mc6@%JXjp{{yZQ8$?RQiB!BUe{0 zKQBV!j5s7q%?rcmSgUHO?%FIvyk`epAG8PPx`$i_UO}Smt4zJb#2!;=${^;_DMWfmfauTl@U6brgz5#n$J4J?9y*72myX zTW0x%a#J|kKZX;*$l#5MVdRJUzxI!jFyNj#y*!hCJ=nbK@x~c8!!;NG`Kt* z1Z!AxAp;r?B=sLIK}hj8G`gY&Y)eGGY`a^hZgWs62wOR7#&Nib7&+SG*yaYxC_dJt zH+tse+ZvQFS}lLBu7<978BGWV>Y?7GhF%m^C7rl(gGZxfFD4#cFhLu(!U5rtS%9(JztgnvHznIgscpls$C32g*67 z_M*e$G;rR>(I4-xy*kS6V?zH7)1<&~*v?zJS3UIa~)XnNY6+?@Vs^m)PwQ1$3ir*4Fuc<$?O~`PKo;@((y4IP;xnc70+w)Rv ztp~?T+9$u}-8;W(*6`nEeU1DsxvIzJnEm7~!Q-(nE_<;>-vH}|H6)QK48Z*c#V|d7 zLBRjLq<{M9uYUW>7$wcR5Q-XQi6GqhI$89CAKMlZ@W~1b*sgGt9bo+|^4NYFv^U^n zNcXJE`6FwMyHC6Frvrw+C1tb!0(QO|2bWKQ`^e9R6$56~h?=e>NI|>}! zuz^p;fE?~gX%w?KDLy{oRogx@!LE`yI>#!5%4heIIR9+L_Wmt?nJyd)#cJd>s} z=6&-SS-^}vywX|f>?+8y17!@O&*KtaK%v4Wef`J(V6i#TmC{a9$8W(b+&n0`WM!e6 zDE2r-j#4qr+ToTmqZXw^1~rC{M3+J6uEpo9+{Eq#hUO#rdjn#Vkx6CV$_uOWyBUZr zV5v;CTbs5E6kLnN>s~Nwhg@pJs7~Ij{Rm`ZP36@!u*$~i>#!saOwwt3I~1)n1TKaG z>uVyULq&51M)8^`ehbj%r0;5R?j zwq1|56|CoDON6^X1o_aBl?jjJ?Xl|+l~+%PKU9}_7tw=r5iF>)XVg|rJps`^JY$Ju zsprGcX7zIEUFXK~9zkFXczSfLGdGS=kuTXM2{{t^^+CC^c*wa9xio_+_bzpEMA3Z5 zYft|ikcFQ%N?g&qkw2M;w+n~k(stiDW zs8QX*{=M=L^MmmF_g|X!MW^V&GBk5#5-cn_!#4&cJwL|a$M~#>T}Ew8!H;qNE~V0EhiCX_Dr)4tQN7|Fen-jmIW77m_}^Y;v$!NWk2u9Seae8^ ziJ}R|QJeqbowib|U#RLv4*6Ac$h8n=nW2|zg3?(-j&4sNQn}RxU4SNV)LFe4kQ{l# zGZu`DApsUFHFTzbqCXPl7(NSs;Et>+LoRuAs zxu_ZW!BcR^^CRwQXZhFiS=)*wJ_bd=AvTWsrZ#Dlw1WnN60z5{i#CV&4b9q}GIHDSIX_W2~qjJ{{kTEoBkLb<-g~de8K78RZmK4rdFf#pT z$7f3|@K|tW-G^kQfLAhGtOh)tV%Je5g^G!rvUd7u`6AhQd7}7|vRS|3D~&)zfQ1mH znu?GtWdmIiQW}{bF;Qa;)!ltHRTf zzVAaCdEGG8ToZCU=wy(&Qr{Wa;s>IUZpVo2vEk&YS8;==u`}dHwh~{04ZuUODABDYDBME_FX4$ZmwM(s8_53>J%b;e+GK5;%M;Jk?w zKHx^dXr^RC6LUarF(oT9KKW0wTEO!RScwtH*hsPI6j?{boG1A+_bWTXafQ#GPUgBMl5AmM8!s#m}U!M8H|eULoQhLhGkAh?Gkkptb?6zd&U4L z#1$`ajOFC0_Nc#1wV04^75)58^1*XtkJX!6OtG+H%%fs(HBDI`VVdtPG!!nSf`TjaC*hkz(U1vI41+dKq&SEaY%n zRVaJy_`+v9hRZHQ*l{06pVH=W`-09gxg3dkcP%MK79AG6W31b93OOluYLa@V+tk#$3qgpg2MsP(odw zGvv~w-WzmWX|CQk)$Zp4G?vSd{@~P%#t`IlYx8IlJHg9!6lBJMj!`fD@pI9WZJk=d zMoc)o?Ji(9!pKMx8xIUL&Cr+8qOX;HrtT$e()T*0SG=3Udn0<(+2UBLhOP#l^HgSu zI#w1hDT=J6&6gOp%VpO{nxe~TB81}xz^KDF{)7Cy?e~84+=CgGX{ztv)XT^w0f$k5 z328(arHo=rC{j$t+!qzeuIdx1|Jg#7%Lk$w=tE?e;$lE?*jD9L_ZIy-XGDiInp zQObgKFPJ|*=+`3ZkhTGj`S586f>)Y8O&W2r!`H^+290Iz89z*{@4ycpH=a`1~=jZR`FIZ+8 z7lm#Y$rbQS^P$yDvy)=WC{kix;$X1;0Z4yc71z*)55?KyU3y4);`?W zy_#_te0bqz($~tjO>(v1rs1X9cgYd~<7Sf;Zc->VnIZ{POe#|!bj3RX#uwHvSt8^- zX$IBvEFhVlc*0U@`fBCiRxv7O4q#is2|bJEm=w_cAMI8?Yj=P<9oH zO~7&t^Y*ntH#(Y`joyYEp3r^naA$PXVX+$>kL|{&_i^+=e&~2N;^u#NTZ~5YPb=Oc z8w6}LfV6SMvZjDy^C%$3j#2;)*qvDFkLxnn@JVCoX}pvx!9V*0ZDD$}ak>XsxK zbb(4}b+0&kdX4|;NTYU8v)lC`8PuR!n>mh(dxw-w>I)%_VzV*^YPO-ZK~pp6vN#}7 zZ6=W2EIto3Zkv@z=oY75dTs8aAC0FYSireKj*j4b zx=0ou1u+s7(=oUFRq1-j`;r=YeXvoxLVX%qEYJbQFg}RcLVMtZI5%U#j?FMTW%upi zH7(DgTnZ*_s@*VO(`@3bZ&jtvx(b=V2^nnbk+N^-t*@DgJ#MqHbZsAFE2z$)MBXc2 zJFhPyUxz`3CN<_+7s+s`33{D<1mhJK+;;j6x-4_QMhty(D(AH`izJu)4cljK5pRL+ zasy5RC0@&As4Hz!XVe)I{g;UDXu2ScdlgFGOQNqZen<8F^!?z(5H;WPOv#;DppbC)g^K!V5Z=RC6APGf!dJnJS_ri&)zchhHmrwd{Frb~gob-2 z^^01%PHjQUk#GLvMY3izv1j7e6o;OJXTgv#LY2njDnfQDcQt)zq7y z#BxDirvmM!OcYiJShT9znMAjv1ntK}c*xgopmTmK#q_8*P8(p%3woq3xjf_445PM33WO-BOq;YtTR9KKX$xh|uyBXP zcq#*I=HVt5>uVwmz2(o0_uS}x0cIPcZZ*5yEp>{aNu3JaVXN7_IxJc+$Lh=#4Gx-& z_yNMc)x0Cn!pq`Kn%8>j-)*}KIok5hghNIAa66l!K zWVMfCA+S=3EavUn939p>K?u59n@g9=k+TbQYX@Jr?G9v1U9bQ;t;(en6j!{n!d5Vb zRHhW>rsZy?Mv%Ta`;|2z4?WM%%%nf|#-%kZrLn%Q6_!#TdmH+G0J9#J(82jAX1SlV z0@);lWyD!w<9cg&z6kbBSeCQK^R_4xug|s3Dn7~XCg8vdNA3{Cu912(OtD!M*+|7Sslj4QzYx+uqvlRKgVW%Edx9uIVQv&NA3-LShoPk+ z@LoakSeL-2$ksV4(E50+#)t@~J0{$S7-j7~=|A30{>c}8A#~akCyZ6P1WAkbFa%Wx zH7Wi}eXe-r())|fs3wH3u{P@z(Y$!7u-Mz z!xb`}t*XE2D%5bikwzZLrNKi1ZFDj@?7B7FG3I8h@OkPC^BA3D-r@(J_rH7ZiK_l* zB*Mbc15oL{OPn0>yNbxIp6!}CNtW{W<|AfW5TyRt5Zowz{Bh3wO80}3cFhLzKj&su zhaB|F4~m_&-v^mO%o3-VAu466*=EnP@(!tk`9U0svN1zylneh$=2>O$0$SO_KnWVt zqCZT|DEpPkzE|C`=XbdXWNRc37(8}Hp6_;$tPm4k`Er$lO>S>vZA(y>D zOMNfMyF3iYW`LhNW;(GGMf@;p=S>}O8ZXRF9{A-s)|P`-sBs`JR@qGYBQ2_TRYdm5 z%))@kosmB?N4L`y!*7Aipiz7KwTu69G*}MJvw{hHUiGL2fuEaOcXx9nnW6(BR8#V z_9qm3fg@HiulV|$Pr^^DRiQk479WE%&xI{>hgmBhxUvxua{f6{>rsUo9}$l{gs{in*q>KD0o=khZk5TzdLn}zpWM+AMZuD780tl zM?&k7a@g^&UsCS(t@5c8ol#BU5Ym>%-D{CM)cyD3rSG~UITwn2SS&|$m&a2jzgOED( zi(Mv3k}f(uuu592DpWM8yXcDn)sP)Dn2Y$@Oqqc42n^>!a;Qn&s!H_l@qP=5G%zFN zly2v70N}Bmaq!^d?RXiMN#doGpSO*Ixdlrr6Aw$B zqyrQ>`ap*=pH7m*5#ZTqlWx*=G3Mg^-L9vgB`z+oW7;h#Sf5-SuCalTAFJs0YpOlY z1Q{BoOGI14+r8rg+aRZ8ZnJ~>gEWsu8c!~gm3bE^-+BwLNpfwKSNIglwuT#SOB=;l zd)Em$js8=Gqcrb1+C(d6s!Y&|yhbyyV>c z0r3%?QP)M+2EQLRmgs@g2BoM$p08)giJR!`e3E{~`U*{j8MhoMG(Qht)| zCl{52ni9>f*=;a%sA%o@p|fvfK6mQ2HUbXV<2N#bt-nsa^-tml7GQrgGxH9)=OWnO z3Sv+r?Z^s>T}F|`R1EU!n6$CMH+=jb$3u}DHM~H|eH({=!V9F6!i{UBFIoJ~)^A+T zBRd4VtT}0gh5Zy;N0AyT=F&f+j9kjg%jTY+ovo`??NlVP>wR!PPkf!+j##hik61~D z0)_%Q)h+rC&3YAXHNe7Rld2@Dm2J^y_}9|Ok?$r)RzKef)K8W4`+@HyM_yERs>_2K z;hGx7A|TIBdAUH_qF({mR?gcI@ZkdZT+WN=3)Cp`b1(hTQjjHyuYnSFA*Q8A`D7=@kF0urAsufw$2{ zmp7J1yNMHyWw*awIL~5a-0vJpCA$P{WKLNbnFfjlvGqDC2Kg#dnLC<{up+t?$!Z5Y zI+0b9-m6Pc;NcbOt>Uh5Gh?vuw)b}FrJH^7Ai`O-P_`gjgh9rp6xN>hL@e0H2Kd5 zwtvy-(}(qXOfh|I)M@ccnmP7?9piJy%k0C%IKE(vm-52IRxrWOO*3~12WGa)cLV>70(!xTwcyfo3(SG4xI-$3=lI?}Obct7+6uU_}q@9}6UgaA6;z7+v7=4CvGeL2C zZaFc;1zz#S>8~<6T~z}=px*KWIDH}ekYOd6f$S{B!53#1Q`cPI zGXEdv>o%T$w4KlSsqf)54{q4R*u14 z7@HDUdZ1!t4lrwg4Wcs@BTvMWU_BEj%csHagqMjNzhdP*IGonSu|vdfa0F{p0+z9HmPg^kt?W8i63rmJpI4y5)uBkxj_K zoTDotd87**s3DgEw*t2&(828h;iw6khsHMOBZtR1+q|sY z*mQwt;lqE*{W)x_bzOXjLaP#?Oud=}NB15C*a)_Z<)8ca> z#8Gdyonv7?0_?n@gHGc{z-UZPAAh_3CEHaGpSZn%`^GrdNPnYSBD2AcH0#s8(+Do^ zDSa)yNtHu?;1BK@ILCQ9)PJwl?GDaj4@E9g^^(7HW|*|_+luNyLp!`o@TpDJ>p!z`12s9ZcZ%{NxoZK{s?^eel* z{hLix+iSI7UGZfss>*Z479q$ZzII9Z?iFSaFv1_vwa;zQBLTo)45i=f3T)Bel9xvR z@m<_HH-FOt{0rX%cgD)B?s2iu*>9FcxUKeXj>AN~yTRuj+u%MQLnGh>AxA1;v-rlW z9gvxV2pbl|6a{SyZ&lrzYbckW2O{my!XLPU)RtLtwV0Zq$%{Wa#xdkPE=~B4o@U$N zzF;FO99f6Qjak>}GGFvpu_p}c__0h0IRQWPIXBa&MH27yz;4j;z*2)QQxrB8mbSQ} zDSLUvly=`vT`dIiu=@+-r=Tam@!;T(gHcTcj}f--R9?=@Wa*|?gU3^JAmBwZN6`UR z!VO9i{Y^D+X1fw6i}iKfvev#~!n5pzko`Aw(3wsBUEVKP+!YsvZWqbr=dK7C6ddG| zoBS*1#mNekhAZBGO%xZajM^>V&C%6_z!8jlG;8ze)4toEha5mw*n&V9^LS5r>~;+CQ?7R2N4#d2Nk7j0Y2V+M4$xNe zDYKR215rp7~yu zDW>(sGruD#{H7QIuZYX6R>V0J3z0Y@@Ho-+(hnh_J*l=J zU@&m-)3!;s(Ru#evyMr3i5t}ou!PA3X@w5WR*;2G5H)HF18U?>AM}x<#RmW*uXFY` z_k43~Fg$HM={*61h9kXKAS#;`C+P%r#LG~b(k9;teTnI!wM+rj@gGL9T+`=`((RG^ z#YfZy;!ibI(RU&Ax=q|Bt&^1dbV&C@HG=WgkH5X*hexBg&~0ux^V>D2)Mqrc^giz) zhvoGwnBW5`E2Hwnj$=W#SJ80W)=Pax!e|^mEsE4#Rj$_Gn3WM`Fv-$QsBIY@K6?N$ zeC|a4?f>D%UwzTrzM9k+2g}uEAn{{AIp&d17bzXL;^Z+L_KdQ_QuTa(7oT^0(!yS! zS3E5K>CY~Zg(R7ue=6V*NQsqynoY4l{F4F2BOv~i3&o9xG>g2DCMz!tl)zB4r9*Yv z^Y*mL*KW++E&`=W{F5%aiSz{EWRowb&(%fN;eEjlOQaW0l zh=7rB!U_p>6bsZYd#M;~6vWVnSrT)Te=4(5pJD3Df|@d5ff{nj)m;^D^lp=yON=_{ z4!S&|SDdSBVU604{hGu>E@t5<^XJ!q?(JH5(hDnJ=%NeVP;hg=qeh%0d3<%3gKAI4 zh?FOwV|8GZk>`7e+l#&RyTTi`wMK#snQ_z=_0XM~51)O?4%mLobpUux9nr z>*g7zP_a|ffm)K;)7zw|Pq|?R(l_ptOaxf$N5}|6FuuXJ>uO#InKWmH=%THH9EW0c zfiB%^Ik~8)b}Of2ryhal4%HV^vHk)mx1dx9xx`UNtJ*C-VykjR6n@gI&ji5(tY}XQ zC{+y5S;{PBr|uNc(tj%bEaKe!^O`uur}Xv_c?n z=X%T4u!-%M_J5*K7WCi$_G+(}Eskv_^Y4XZC%?sya6wN#wsLt7P;5O#_5oe2cb;oo zXgqr#Sa=#}WVgQ{zpkkcsF&QLTjih|C|M+W;JGEdUb07dMSU>5MROpyE%af4QG3GY zqUs8}6$;r>{0l_eG$%bX{84JGhqQ~f&_~GuMR`y~P@!V@9(e7_sGSgq>(Dex_e7!r z9k0oBw55DxIqk59Pdg>$-`%`^Yr)a-6$$f57e5>ccZ6xX#;4X_DGhCfaBWe-yG5#(F}Z-?F?2r?Vfc=Yd}q+o1x{pYwA7F!;_`$ zlp`TJ)FBkft^!51c^!2k_#TtWGzD+B@D3C(m3tjNLk=1TU>wh;A)V@#3Eu>L5)kjJ8^lIp0gvD5hxo;I((O~J8 z3ydxJtJW!uK%kAk-KgC#vyTk9bWFp}$U4z!U;N({4E&fC>eC_r;WYd`0^-;<@QnD-nO1mDrq~3EtfFG}29=6C=#D^>_iXMT%L8UHC&Bk=D61*%JgzgtSnu6#k&+#+aNTH3mel6fi#ooOSD0`LFTM@w|po~T+L#^8}eJc zGw6a-0bXmUh0;X>UY|q~fu_5i0YkPYFqPRX&YxKbTrP#O8laRLa(Q>q1t~DH!p_Za zqmix$N+j{T;a1}zm-hx;nql4%tdqdY6baW?c z5Lsa1$w%6BFWfFvT%Ti_xrRR&@VI9p`@=DA2L21DzVG(E*`vFJ&UvTJcjbi?;Z9nj zunp(N)2}Gpe;MKES6my?7P{}{55f!&JiDN0WNUbiE|qC#kOR}wC6BmX`Fu5tMc##q zi;+pL$;7D`w;(E&BIM0q|JLx|Z~W?~fB1!R8O1K9NbKm)est8C7&qVh#MSc!wmRQ@ z{2BoVfC{brnhz*;1OC_;T#xSwJmHhd^lAn?ajQjNPiliN$g`*K4fsr%%4`qX8_+1< z3!_y5V&;^H*CRQoPj^323_Hd((?=lYy}vYN#j-W#({ z(#cQ&(WuUM15pPts@d!Z7Uc#yUQ!29xO!le!_ORQKJ3TBqv?Zl!;ui-6=y@O$|26M3 z{?JASMAy=I@pcI8q84HOj8tYoTnSFfUY#Yn+NxTl|7e;+K{H~Y$Bmy6_j2SWH-4U9 zr+CwrLGBsxO*r`GJ_DKNx34gzK!X4U-@5XM#k1Nq52W#uGeOx9qtA9f9bPBdq3Ks{ zfpVLb1V{_L3&quLjZ!?eUbP5xn?F%L@afiVpTCmC%KAjk0n@ml3;FrkZQL~MXqcg^vfql)X*u4 z%6W$5h@0vQknzEw-#A$=9RLo_d22kl;kGlQ1fnO0SHHJtiUms7{_Da!WW_V%_~u&y zZv(}G=eE}D_#V?<@y4d49e$g|kVO2%`)=f?(srl~?jw5x9w?VfE{{8d&134|ha?_1 z9_Oi3QhNUoV!_6W)Sm5RD?e-qc+Gs&3LDiF3kn3gsF)4GjnaBj7IyXJ^T7?uI?;{L zLv({OXMPr4No~-UGSwlu^g6{lARgSNx#hi0yu*8K$QF8+Xsc(MI7!q=9{40gB`Xp{ zCj-*NmF^pKmDHu#w^i8+Jemjb$V}bE@a?+&5C^~FecNF@ko@rT6r(fz(KvB4eEH6( zlrY<5H6Oo5z+pL#G&Xj@f+Dy{>a5KBW*~H~)T#zG<|KA1V|;msUr~^$K-7FFhnz5p z(yW{fhnq!fr(AH$3R}VCkZvX`%tX&nsk{N|HeK?3C|~GwP)+&p_&5|d!#8!n-`~0@ zBBopH&XV8x#*@rv#_m*G*_{%KEv85Th^5W1P@SaHrf;E7sn5xlK4-nZ>^ZDtO=U8u zhtZ9iZsmHaSBxA|x8`OlvclHA5GQHTw~Ok7GnJ>IH?l=v;?*Sm`*MgEwjTfHtYI(a z+c(Ncfq)(21}h9yQfvi9%Bh$VaV-tw$efDaMCW|ox_UCde4d3jEFtSpnRo(3Tq4EE)P)dv;G;@u$a zi7g2OQw#;N9346yxI_ZhWNC-{(}A-Ln5N#$w7WGAE8({JIFPrtJM z`G5PW1rl3lh`LEFKS&68adgoN5{(plj3P&0`;!J{P(?TH}gg;H6TuqG<){Uz=Xpsx(@sp zyh)25FH2=E`ZepX#+=d*y4-WSPVbl0!G78|KmE_1Xzb-#IeT}M`Ph=5EPM6X>v>;3 zrRNXIG25i!^^tvldBbk;pF^+O*4_zrCgP~3D+($Gj!U331sc+$RgZ1#C<>cyK$W9SCk!WiTDnTJ#l>)$UcEd;Hs^ zA4a40Oh4&jZmLg0Wx@efqk7_iV=rPXufOx5V)+Jc#4NsXr^?o1aJ-oP?oAMDPX%S{ zK}{|s`$10yD<&uI={|Z~_C4lDr%%`h_hTLkd3f@ zMSZ#t#VuOo`I&gU+6Px^!eBc5MZDlT`QPGy_n~buEuUDdfLAaaHNmYa$o?q}#;CGc zZFH+@*VL6FLw{&OFcwDkL}Kc3$fXjh7L5L7Ga*Vs7e{od2N+ZlPVsL8c5bYtbSB#D z#MS7&<}e7H@b@1ZrvHBOaP;;4SqGp?ju+Gf?2Bx)TG%8~Y&=C)P%(WA_IG{3LNt!c1Eq^ck@oT;3@d3|Pf9i*O=36Mn(1RrNPr zh1$^W9cN-X#r6?AJdw)ocA&%BRyYw6BX50E-8so%u@^@pr%OoBXq9fl^_NJJRV!26 zWOj%m52zR{UdyLb{Ifux5Q)@}NFb1ER)+2;IXXl}ivN%cHl>z$Etj=3yP{Hb*lUbL zVyUnfFG=(@YL|#ExMjEvxga44mdbUI6sUtc5j?=$n42@R+kFL-0p)Y)KC9df=1b~C z!@N^?X)Lfy4T!e~b-O1?;@$EBjoR%&<#78Gpg~uA{0e28oggI@<1zV^GHOY~MijDr8!}dTiAJWw@Udr$Yt(^`LUNjr?t**{SS|(SmUrX*MQed)mVT zpV^jgKN&m`uxrav(v7wUU*JKP0j5qg=#n_IFCyJ*lWqyq{bI2QE9uM`9`n9YyHj(E zHXwOIqo!PLQXj(p#r*{P-cCV-CNMc*^c7867yaSXP5yX}1AQRgQ$LK`$p16W@P72T+?i`|o`6K|$t zF!e@bV`aWyKHUuqB|r?44H>n&r*91EjQ};POd5$=K>%8m>;FKBml?ItJBw$v(Y4Yt z;8yI=)W|O>dn0xPYzN-N1W^x^ca9rnj>XnvsErq8XCvYb->@KPa_@h9N)GTtkbsv_ zSF8|piegVvtwNIiq10xkL;x(^f^ zw?$uJYN4LIO*-;k=<2kQfvA2jzCd#pt{x-Ivc(fVYFKnb|$-e6Z2G045l| zkIw#o(KsKn{a?iMOzw$clu)p6-_*q?f4}or78~^Ctp|L`-vnG1wcKhMHbAj=DRP^N zNf&j}HPY2A63yV-yhw)90EWGhdm~Q^Y>U%DdE%QAU5*t-HZ%|XA>J=9) z=+&I?$<|$;bCSLZ3nv4TJ*BJKVH#T+92>P+TpUp>UhM^|I=o|nr~#RwH5d*wr1+E5xkruuxS62|w=FucJ{4M*-bXAVF^Wp?I5x=3lpk?hoy8ZwCIg_op7BKYi^gQ!n`fF26>3D<;YQ zwItnkN%nk-0+t1N2ciK^I z`&rqb9iUIFMXGR5-?W<1t6i=&mGx}eSs(9IeJtO52rhLO?lB>)0j*}(LOAvwM zaCh3}sHHx!bB|3q>XS*guvOCdsfJ@yj9QE^V-VP2TB@T9Vl>GY6e?~>O%6@2t_@<= zD`my1E_%6}p+|jLew8u$uMTUY(?JU32*jyzagSkb!+DZFGb^lOM%I+Qo;&>va0D0c z1(dk zLDg$eff59sHE(#*$?@|OE<;k(~D=iIY=Pv}L_s!#)+ zYqK(ozN)IEQ&ih{7oo-|9x^OCq`PG3Ld06y2Imebx_!n)e2oTB&xe9-?*4Fd;~p=g z{Q2?8*G`iRZk`+mjxFpqkyb?%3pKMjR1DTqrz%p!xK)B6XpbyI+D~F87$S~^ zejPOO?N+(839ziBclN%(eSt61ASN%4)5%y@pVNnN7|c!g{GsW+fEh-?^1+Ag`$z#d z!Q#NWr@=(9?50>q(kiE73_Ok47Uif1NRV3pQW44;W3mNMbAzPm4X=zslEz-3ct*~K z+a49upv#gFYn!TJXH_av_eAEpVZ1`4d>FPSype{O#TPR_a^L#uckH-$@|SbW`J7*o zc((;w>UPCruf<*^u;!S$r)T;&<9=D9V^*;!snp14)+$Bk%%Ib)5R7=X(YK%p@lLNq z;Z84|*ao4d=Lv$bNTlb@JJx%un?c|Gl2OLAe(gpcshCJ0Tr(C zV_LXjGotf2$7Dnw1NN_O*)mxhebXJ%Z^0~Fu)uKjjQf#@i%`Oos)!fHPDTUZG=0WB zpU#8s0LayUT!bko7f8DTzoD^{ffXE{z`5`#+<5>;c(!XVkZ$D9IOC3wE$9)9T4jvo z{Xb(DjC5g?3&Bkz0B<94`F+gKe;xCiKm6nu^MC)8K68|xZnMWSCrpSmr*d%A-<4P( zXXtyQV=6K%^t+#m@F!vr zwaTZC*%W|1ahk(!(~Z35u&W9z!-xDo(XvPz^CmuTiMfL5d2`wFUUvBVeZ;7p%AK?) z$kLH=M;+MC%{5V_>nS#cB5SA^RDD2$sHl}Euwbg}^ujhZAZPM{kvCOQ$fkMQ*&H!j zBu?vZ_N+ZmEen%C$$!s~8yd5Pik78Zzi~;Rd3QEIKcA!0onk6~6^scCFIn*T=@;QK)=ON*A!KAyT~47hHhpqz(J1cgK9bv%zI zP?03tEyWk}+LI%ZdtANLB>^gD&WY*+v`kv~T~W0F+)qQlhy^Fq-xFs>jE)F5`Q&S1 z=HyfsSe?*`hl!Nx4bqz-NGque>->9kI(yBkcRAw!*1i`lZ%9VSRdQV0F*oe5>v%@d z`N#WrOwA@0&fRp`+yPh67rpY# zP<6GeLyFdhE?kY-l1A zEbZp}*Z%Ee>4)UN0>nZ;Bs5E&dRaDlU|;HVa$knMh4-q)S$tTh9uX^?mwapw6dn7C zuX!pam$0(~@2D*#ZXO8D7Wa}}%pzer`E9q?c_!ZXG5>eNucj(Kq0@MU5QARKUZ2?# zeU|B=Q$rgCIw*8&bIBGL`7IJ^B`wjJzC(}ag%yVF^N&-tL?01ukoM5s-?|L z{=Bea?j=d5Jdf8K-V&WQd58P?IgcbuzS`~I5`FP)>Tbff9J+?s~}0 zLxLST(*Hi>nXr)7K*HwhZN-G#AhG{@Tdtd%9$a=l4(wC0V5k@>grFHCSv zLppYL&a?|8o1j-Mk*KlM!yt0d@mArDIolmrafa1t59zby0)vmMt3@4QtDrO_pYD_= zJ1?}cD6$4&W-`aZsTHu{FyudfyQudc%ZY-`MK<{Nkn_9}?-K7u=R-3ghd-)1SY0H1 z9AZxo+eoWoJuF6g+J4t+0@ByG7-uE_nU)!_Wahoi_W)^cv3rg&NN_s#8-3 zosN(rIq??LaJiXi>DFO%FPI_TTSzX4o zvTgJUWvw*XxmBbt4k!(4Lz4lkxb{K1GwwlRgF2>S20grlP(!=leSu4$78#2{AhD#1 zMVlgzUJ_a#s7_T>(|@@v$qPLtN`aPk8s+>r;Vs#Gc=ZEr2O*)L8#*E0l6@FE8`;Ki za=I-e zH(+svWR`OOWJn=+$UyQ7hKd&XYpCM+4!d03%vXmkED6~xTP!f5?(MVd_mkzXjKnK2 zQHB{5yPhH`DDei|mb$k(Cr2S#v4h`j&qiQI#Ga0Krx`jsq9tu(BiUTl3_o{NbvWVN z*0b*!!IAdYuf9vpIIzF>uE}y=r&yp#yn?n5rZJ(6GE!-_O4}8yU5=3iSrZLqt3^Vj z3_B{vQ94l48nb9rE*Dlm8>O$ey4)t_@BwIiVEEN!moJ73IDqAqs@?5zR439 zIOp6IKat~2yp;vVA1wS3Z(72_u-89^!!aOg44?YqM?LoK^POxIMjdlPE|biOKwvt2 zzf(-HAoFvfp+w|uvO`iywaXs~$`p%)z2VPZX^!4N6?{AYhc~~hefQzqq95IWur4%){b=#53dknAHu6rU*UtGui-u-aJ&vw1+F@~qvqRn1aUY%i}L@;eM zgSQxLc&tFuiZxQ`Hk(g(Ohr9*l6TX~zM{eKB^?!VU;OzD-H^bL)p$}2+;ZkuE#b!`sJr$E0S_B0tZGsf0U2&UW(`szLR0!n(IdgRMEzj~d z4vDK=G8KD8cr+eJ!X@X7rZMBhx(Uu*^ZA~6s^%+Vl2}OM+{P>Nz9hL%KbfB8yMy$= ze*JEEEh%&_XYzRG16!ieOS4pUKJbbpL%M{IBiBv3JFO+UTh!p1C61Tc88MuJ^3Sv8p4w_RlSYSY?2~1+o5FHfRK-gH(tW1QxhkE1W27)(EN1cVOK&QM@UKv#x z1<~Yo#WtRn*%_3?B#CiAuOf+QAlrED;Oo*TABWrvF_Nx;VmLStQ*JG96Gro&8~qD^ z`sTkl7|kzW;Tdjn-eEm2^jq|%Um6wAhpQ9cAYBf;xn5%8i0r4>dlc!xf+K?^f_}wj zJtPf0P{&1v5}{OH6{OAd@_K-mcQ9nYEsGv-I}_Xi6m`9dl`d6Kjg$)E@vFdqi3LT5 z(P`mW_?AUC>-j*io)+ga+$$0$x-`kIM>GZxhJZs+RU~YZ<)|>ig)?quugQzV)jTYN z>V+?u;Tf!B8k&7^U@rW|_{j-uiyXTyo#%n~Q?h6ZlqoG(P{{O?PTD}awi)*1hW`>0 ziwp%vZ2Fz};@dHXBg2=3(|6*vKm3t;Y)+E*Y#D+BTR0Yp^Qs{muRJKz_nK$6xRN>& z)XF|_N_p!PzmoclSLs&e(j{BMKM9Gyn6%RwwkhCd(7njYH}}4|$XTOAW{`vYZo4)- z9us~n$U|FPwv3;$dz#;ymoizX0qPIG?z(LD`hQvcQ_a^}=6?M3bOO}u2gA3x?UF(O zI-LxCiEQx;zjzKVulmKa*=~g8S3N>5u;bQoXuezU>ky;*%J$7kBc&rz$2f36^tj0` zshVQE8yB=&t!-=oddp~M zKm1jOIXi^ICI(yJd%&(fT@2PdHwGUUwM4IWKgZxtB;TuJ>I1r%9N1#7mfal3Xc9*x z*kFyVo5C^@Mv!~d=0D5+pE+@v!=j_eM3ii#*mSg=V!C8VfH~lLSdLaMTB%y*R)~7Q ztuq7uKi5nUqzX8&{3YmE4rm!)6 z`MEBrH9-18K5(6V^y4%C<3YpEjVh<%+qZm4`$Q6N;ugD0v3DqPi;DSR;YN>DGxMA` zx^Iqxw0$JHfok@^Ol6l}`fe7|h~c z&Zu_;zzF2m!3_Uq_nV5_&eiX3j#?95=%=5f-80_*xCl#-vR(H~+vr~7bIX6wX^UGK zOaPr<@mIik2c14#*y&Xnb%RzTWgYHG^c^BJ%KB-B(v+>!&j+Y)dYuyCuYez(%F|5I ze=_#qyY0Nw>!#O5Nei!LhMrc+;NrtyD^(F2^lBvNIj%i$|A0%7xsKuzhjQHi8L`m4 zt0u5dh>m47I%5A%qdgm2Qpl`ary1>JjJMBWZ_90rx&&X#(l8fG=fq{C}J4*t&+8Pu`fEFoxN>cY=Jo{9M<1vVV$wm z6*(lK@TE9FU;ly~A}YK){tgRm?K+Rad-3cdT5-j){usdvJK;Nfa=yto?s;aU|I1;r ziiK>o3Myv*SCM6MExF~<=CakjQP3^I;`su)kSPy}=U5{56I|Ln97z;K9_wAY#uk{+ejYW65UL9NZSN`+@F}&ZanMia9|IZg$j)w>@kwgM-P1}aQK543~98r;%;dA3lW1y zt`PDy^uXJz+UTA+wwUY)uhqh((V)P>R5rbASq=T>(+i)QyTV=(m1Dt4_CS7F^1-5@ zGHB+}K5GY5ukV|cs;H3nOU}?+Iv)#4d{IdNc_n|D3x=m-vI&MnHfn8x`CQ4`v#cC;tx5V@+R=Uj~vS$Z;E)Y!=m*)Q~@JwSIL zZITrU+kCr%7Sz(4r8k2zLt};aBYVh#UD5+ncZW9z=c-zu7csKK*eQ*Bw97xt-WYlP z#(34Tni;`k_8M=O>W+B=i3L`!GS5^+lMK@WF%1}_7`3!MF>9IZKxi{_N}j@7?fbOy zWQ_j{YaZvg0@h4vlQG;@z+nr%+P@~_`x@m)$>6{3Bw7cSBh@ByWIM$I6MPXBbDB*O z=hN3hT9oSZerK7D9!>6@@`q$^z{isPydywMIp~B;lo+)KFFeNM_6AhV(I~eIw+oRn z_oxRBg2crOq6?x!%%|8g-Y#}5M=TE$A|{76!SC=%xMBI9-}~XjuNxr~xOnCdNrMAJ zrriWGXDId)iZoI&SWcnuoT^01vIX(N&X6Q#wQmpf8(V&0O()cPoD)?C<0xyjnvH;Eg0knI42O^9 zV{X8@yQMfGzzD1vS$|tY@F0TZz8q*&M(ZKq<&J%>8l=__~6iOQHZdY=ju zYc0K;RJm;MIv;eFbolr3G^$(vO;CWN3(uz;NV%tefJTW=UbEVJEQ5pV9m`;fTjuloqcCo_F|Hfems!+Y(*Kq^pwotK8S{wKFb> zi>43oPIxuT3TVUG31yOaEi{X2hg4UaFQ1jG^L+UYcWu2^+*Ze7E#gE`Ry2A<+ z3pM9yq2=N&kvdg3T`SFJ?oO)%opPHL`r(m&bN%t;H2whsWnOf)&qdJ|w*_d|l!Rje z&la~QPMBDLW{kD=2X`oq6~x>>(Yhx%f%vtpfs@`LMiEo)q&-2FzGN0lu8BNcPq8T! zS@Q)qH{u}qb$yQ+;Jbh_uLfu2$!LZ!zh>S

    >rbA2i_!IFhRk`6nl&!hpCuTw3ZL)MXFSvLf(F$cSGi%_0u6N2&U*hRS&@y%7$VQ z+yK=?9Ovx=`LWSCZVHxy)Y4i0i7r^JgS-~oy`Kaw_QG0@9$Bv>#W!Jgva>cc3%H$; zK?qAsYUwR*A3I+}M((Ng0Z@+$caT$P-Mfv>)pR|n4IHH=9jmDM0$7%gMh7L!;2FW1 znLy5c`RwaPK>7W;YChTEz<}Cm0;mFt1v#^sifNQD^4F)YU|(J=BEqQR#b9lyu`5xp zq7b0cD&6yjx;kJp!~}F;Z9$N6pQ=@aJfxQ*yJbxt2PC5e!i$$_tWbIJCyZ_cH>gk- z2G&ZAP!Zh`{e~nvFjSxc-mvZvT8f1NoD9%F^v2mc$RT=JWI;6OA3>-5qG(TKE<_BH zf_L~Gq6?x|y8>G(dm8TE6`T}Q2V}5zgT?Sg8x2H;KX0#NP7tBw>1XHus}UkUEBo&O z(&WGp`P>8|S1I-qMOvZXXZXin@ot8Ho1-y0bUnK4D}UVjlcjH`Q;WWZ%(*~*;oC#E zk=QA1WG$(HR2MuxA6P~wzh322tLm1uMCVA7-E$;8aCB1C<9e1!3Oqz7yW^e9ikg;eaeUmipuVZ(aA!@-L2TiH;50 z3OQD>)HOv%m|>EB(c19iZtKOxkuQenXfS3;H*vWBLTGZsSk||{w9(uIdPNc*EEr5} z0xeVH*-u)cuXxri8#=eV`)o2wf5tO|{?qS4Z&+8=%|MB0c09_q3a&B=MV0Lpw) zVAyKAOD9;@*sZUHMPjBwCoDe%eim=T@lG!s@nWGe;wv4+MULng#V3sJ5%;>2zGFd$ z11E1 z?0(Ph$K=>Zo*D<15I0PWpjL`)p~z_}rUQ1=$Jk|xRK)?wg=q&QuwTMDq5X3g#!;&w zK?(zy^}x4-%m{`Nm2($rfr}tE-{6dr467HcN!j8k zVK`Vg&mHa{?Z8TCqX|xu zD0T%!5~-MmfFv=dRUvKDRuA1c8lxr)^J0*g3?H`>%OF|)ZrZ<@#~NRe9gT&^q&75D z35ia*&|(bptgy#jmKYnqLeD#|&jO$aq!L;XK-tOwBxzxFmA@c7z}W9BuATwvkA zJ`xMniw`89i105sUqyM6-4FrT8C({b1ZAx`bfX~i&DCD_$)gz$B!>e|(WQ3mtY^B6 z*5kvdYfK-v+u2XI{k3?8(Lnv+!}fiofLraE1LLE?#B|+Fu{$VIPQ|ndu7Hmti(Vak zmv@`q4!u@dqN{>Ce zV4SWHo(y~duCxczj;Y{CI~`RT1{<5GW8}otL*(tw>$cl>T3$S7vIzH_sj6mkyY66=19M`pkZ!@8sM8Rd(nr#ZB}U-f3AelQ6k2yiRCqgYOx!qgdhp4RVqjb{tqmb(>)4GQ~oe`5YBf3x-OM=atYM;vPx1+f`ov)B-w> zSH!!nsOR4x+u-l@hVo;w(&7Ybbih@(#WskKO{Z;jUEucmTdB08Q3OBWIp%W$xCys zkZ1!7q0t}yE;^W{L-3ftmXw6&(wF=oaRTN!Kj+iQ$YZz+ux%X%ZH}<#XgV%tRGWAF zAIyy)ht*^j%$h@VBGLM67ePn4=qC2lS@`!VX$(H?wPx}s-fKMK`OjYIj80MYOB2|~ z{3L!py~-C?vIVXsum>kEi7tW`X3%aE?q^RmoLRTZR(|bm2?i@a(yC#cw%t^j&$)k| zWfV?7{Ul;5DSKsv(@7KIR7bH;jJubLInTyX;t-~lCw-azwb>*&+;mSEElMC@~os6&CaEZcvs0IufN_W zn3q*gn!{S6`+)F2HMCKXquMnS6VD+JSa5;CDaH;yTOw|l&l_~5Y4^YX#)zTa zKXH?ho39LpmYH-Wcu2AR6uC#md_M28B%fIdDNMJfd^QsaV%in!ztltDnzAgi!F3l; zqul7+D5{VbK$ElU5#5R{ZVd$c&1`Y26_k4xiFffz!f`|cIqCHf)ZH3>TL@OOv@3Ry z1pX7J!{ASBAR3Pw@-1!}tgQcRXgcCMec(Rt`I@cEYLa?g@*yx&9Ld`FMr5AMl3 z!Y+z#$UCEt!*AI@w!^;zyd?he**hc|{+aXsbYezs4p51tfrHxEbwtJ_PQjo{VvdEn0cP~8O1&Ia=$S1x^0Iwm@L$2*95k|k?qD2+tEC^!{Y{NDN0sc}5LOQ@$nPsw z@*72!RAS&UrVi3vl9(djJX$x`Fu}(Grvh-~*cI0Bm9ZHVY#gxQW-j@J4JL{E)(%_o?_2Zq?wArm^YHMVs#v3_Vg=?rr(s@h5G#yLA}Uml&dQAgO1Xn zNdDi#mWl?Q8s$3C-B~~z(JDPNS*LiqjmONOL8tnF8d0gJS$S1~XUivLfg|x#)lT4w zs8uZq?NtmqA!SJ_xCt<|DqeU>lpcsKr@CpMx%Mg+PhS_TM&_hm-YOWB;@dC97NCRr zicV`Kk`2LStr8lJZHvd8z%`*W^z@I-lP$Q!IviM&SxCd!2SCVziaK?d=ULwY*8x}C z${dU9Vs^s{_?SI0KBxZ~(fMn!QQBk+(yo#u2bMM^CelVjv6&Rv08}FEStgzIhG&Zx z36mII7zoi}JA(xRAuc3KN`w>~;`UDUp*JEil-1p`(r`7@GWJQkgk#Gv;7 ztFnD81uG6LF)S3U6wr6*E;2CtdPIgl?wFo9W%%ELpXi6Kpg;92;jN!O=(H&^FLa~w zDib?f-78-s+~ko*4hB}cEx_u{rNWzv)Jf_#!Q#MNS_fOM4F4RMAdB7zJ!TjA?+${H zNw30C&D#NmMgtu8b7mycOlSPEj0{3qk-Quc(mfeu#dFr zaOL%%2acsA!GQr~p(MeOJYFo>8=z5jBcoN0s!h;I1J_kjaNQg@$6TK3;FVr`pq=Bm zh^EnOwXIs*F%SCny5)gm>0{!+kqircOgg3>g#A(mX%waLu1M}E(5Y7|-z{kfYh|y6 zJk&cz#{62XpTK&tVf~!87}>k|yT=z84HcdD$8ys7lGUm#Hu3Y_r`Wp`xkJTN1A};N zU=M#&5Y~iL2cz)Mi0EaOKGcjpv!GZ-iVwnQGbI>d_#rHadP7OC9REqM&FuoKR z<}>{F2{w5eZW>DC)-!RyA-zqW%-SjWuslMo4aJsQ$UTeO zHa#0olS~)B#;w{HejQynex6=%ts!4TZPi#ulO(<>x+ZI*=ehp*<{ae?>&daeI-cs&=hG{xVM@ccMJ@Hkpyx?d z8M8>S2jU3{Odpsibx@#v81R<@_wuKF9+pe*5+gHGO4Ld(d+Q|YWsaFWVhh3YoH@8z z%2Rk*o#v{`g3LE`xJ!^}+?r$qZq3SC`U(Gx`xOslYDJUJ-Y?wj|7`vHH+#^|r<~ru zk0QOm7c0({sM724 zWG!+eQUb15g$L#~FTDy}MkmN?*(NOj$` zD&;BB7Kol=NHj&fG7@L)4Ohou7!}GYG`>rxs83Dp3qL471}tyAyc)n;guHYQZWNY1i5Wh_P$(44?_kqB~V*NT~?A zzZD5nv28+ia6Uk;k;W#0ZFIZhI5c?gRMq%kIZdg@mGDByZtbLzixwG%@dK8>UIm;2 zHW0zqi&ZnV_T1C4%Uo_^b@TsRS!J%)xVVfRxN5?Jb~X^VsTT#U2W5PWWc$=aYz&1E z18`fnL89V~H*J6o4y(ukMjU1x`_MTHe!1))t8OAoeB;g4ir#;^1l>8V&+1eqLXzRF zS%>Hk7uLMp@J_-zm*-ylVM}zDU%xcP=Meqpg`2-J9!TMUk1qtv*q8r>GZg*q?Fe(h z;jsHA3!(sb!+L)5y{aSzbWT2<4lGw#R%NKeOPw?xEBy3Tb6M)A4~@eCCyg@UlmGeJ z@4SsR$z%4|GNO59(rLh88AjY!O0iIZRR9)@po4C7ZE#K}$^2ePo8X?y3714_L9VKR zX_DpBYy8mBlB>!TLQq_>)CF5%-|{G-N+>MWyDGY_K-0QDpqkc(p7g@&_CXyNczq#C zOpfdVwm8MlK2R6|vwltBIg<9u0H(qOFnJUUO^vlw%mvt>0LQH2jQicG?M!_j6h;I; zl0Y|>HR09ZMau_%v`y}d{4oWmo9>b{0TJy5kLB*_d?^sjbj{MKY+}G#T+A^6#p0}E zzi}-p*X?C*+=CJVow4X00>5t+yEP1TO%?R?T7b^sN}vE%V^6qt1$LrJ2dcY1QO-2= zIY`wr6wAO9B>nG)#o862`=BF3Qe6r@1Q8L?1luS0o0taL!6?2%*oQ-~z0p2Zt zaMioyOOqHiVgmE!AY1!7Z-DpcOZ}usd5#^#D;-L1>VeTCGDgDYC>`$q_@>$ zKX0j5J~&d4DKB0eFFq_^=dVr+-43&0@cALVhBOLRkeeX}n%-9F`8iv8hJ?-ffF>D+ zk|7UQu2D7zw+63b?}m5Co0aFJ(E1d@)=l)?@K|7~#g89!!VktI&qmoH`Zj$LVvx|P z10`CL7*91ybl&BVVqrPhSy^;KP!ae5_0nuq@OxH0+4Kscn;ZOGu0Hs-Ip6OqV(VC- zExzfFz2Pp3x}ky=oi82sWVv~sxUmL6&wJip_dclYEtq7KNL~MP@vmeVH?zipvw8DP z%$jtH1y#I`is_*-RT8;IbWrKv5`AM{n`nn*chF&Fn)G_aq9Dx9+-H9qX4rySgN>(> zW7u={K2-ha+S@!MHk|yO;>ltM)>hz^ALc1ZrP#F;Sxv>HD&nQ>ibJ4jilwDcDt}CN zT$ay^b14-eNxa=HRkJ0sdZo>tv-c?{HA2lx{^Psv8ga388gqx#a+4Vj?2x@=f}%!> zg@mW0R7@wmip3^?XM#JVNby<|kva)my=j;o0nJKGKtWP`%zwf(YNeh__kBh&2Jip4_+6|+ zh`EliD>(F%b3@3nUw-+r%xJJ2{pPnWk+lwN$Zj_=K{it?bUE2b#WXW*0_-KK@okjW z(k;=Sho?Z{*RI(b)hft4*w0*!tRa;^;-XQmVAi|js#@S`x9p_dD}!J;UI~y;scv2$ zf71w(4-5C)Acq|oB-c$q(n7JPDRK%x5{;H4veWI1IxXuI6zTh$oKPl-dqb|0mF#vP z5XWlPyV3(d5tKr$dS~(5XWl(OCr#WMeOk6XQaA02B%iKhY6Zx|nCiM2N-lJf9fE^_ z52TA{7rOULkC8O-6S;2M8PX-oBOPHm;>&QiS&<)wH5oufCf?_doH6(mjv5Th^6L;( zMq$6LIO=B5J;F)PjSgDo^2oY#E>eq={Gs03;&<6R&z6hd&4D*e7P4}iWUUH}BHVa) z=iICx9QrqnvcemZqFTHfojV|!@Nk@W-W*`W1qEXsyD~a9KF}z0%Dcro-7pw{gEWB85yx_CA+~y=wCitBGqo0ov zISId=ypm+SGCr6p6P#_MScp^>QZe^j_R8#MsWSNQf={I;5T8+6DRrT-78o1|Qpj zJ}%M4o+&COubcC7n>=m56K>*U!dIr;k{ZDwx+D4xNp@gx00HVSSHG5GH&G;mirMj% zJ9CfzhvDB;MfO|CZ#@PQXN(Uc$zhiGiPH&Xt+Zb8#Obu)AQ-l*73*!*cHp_|$QdG_ zcb+?;jc#&+2-WcKz6;F>o*mZBVu7+5ed-O~n*w5myCrEpr{pEP9!MT&6kG{SdL0>1 z4@Defo5M6pgBQQS`^l_9CoR99v;i$<6%eQ70GI|`ufhK2iF{3@x=GL|zaqIEq{ghL zJ@ip1LB+jR>)a}rk0Hj0F@R$F$*hBcSA&5~3@ey7u)TunfDE=BO4a7ymX?IhS9i!8 zUF-O8xJ3J+4^$jQqUNp9lIKmwKF^b5V3!hF9vC zc14%$u;d_r&+Ku7tEQ&HRUoPu4f6r7C-hDDL&U^JKcj&9-)#q0lPx1L zG&pcHy3Rys=_nTbJX@)l`A8xZ55ee`=qLPYQQ@>!zb@G^2pDxk_T3?Rem$uT{A|YV z*{40Sd41s;h&3G)w?q%lOXk-KMi-9XF4S151}O6D_rcVzm_61WhWTR37ya=CGV)dKH%iNC>rCg8Z2%D>A6} zJKj0>ou0WZU#|N~_q&^^tar=)t9Nb#?*Py4Fc~FE?7lDi(eWSs=$A&6e7HLC4btVn z8}TJ38RGpEdygVLRLpkR`}FGj24?%@GaEf>d=Bv2nJlq3bn(Ot1`)A}T{002UOBYS znRPF0QBXQGNh+nY0`a^3Y43x*jVTr7&|40d?v3Hm(qE@`E(y?4Zc28qdW|I zwPJ`iwFye5uNNNY?GtH**(%gqTK;|4q~JV~r^=%5vE`BFp)JwarycF)4CWI$jaSG^ z<}Z&>=}ojxJ&v(fh|>QA}$8332lOu zbkuLH!Bu<{fQ}YS!}XD^j)zMu%cjludz40mzV8> zB;eagJ?54BCNYL_MdCfs>**A&g-uxtbkpgTKboxrR&pG>OqLA2G6DBxPiWIJdEKH-ncc( zKf+^nzhrs&M^IK1{sVoeuPED!>0p<*tIR)uOq^2H^T?$!BvvvPS;IBTeUNxzXi^a`j1^*!RQpL7YmoIEGskWS zJAJx7?R}$RlejD?igY`$O})azRC!FX0~EQBrV3W%uL*xh*9V-RGXUm_I#pj<5F1nt zg^PxINOiNUK2RHqut1WscEw7UR9+R33*n!h$WBNoZC7B}9IqP6DXUyMXyoLD%{yd% z<#zUiyTL}Wu`IP`JK4%jY&fvP@R*6%sG`_C6xl_^^eQ%a zCMY&CC&gz-Rd9!3L)0;{+82u7n4?oG{j_3@a$8i|*!FzY)je`F02n zKvaBN)W>g}pRlR{!Ii4Pht{C-_EpKNeo%Xp`W&?unTT#9U_ zV$hX_$>o^jg?~E)`E-#G3E?*?d%?eg6`0r&3EKz&DL5}vSn82ar$Ew7t8~B>yD-&6 z*h&#}o7XTntPjRwRYD0zP`5=wmcWLq*=D$abaf>D6u0euBSW zE!{8OLWZv2@G70r9DVBt58pfa<+yoTD*fFD^G-(}gdR)Rd zc?o&54tZ8g!`P=SRHU5{^aT_~c{E|qFWtUv6k7B4_r6DtIA}LrSjhmuT0YXy5l~GkL$WpURkqP`X&0dBZOQOqzvauc4!N3ywif0xb z3lVF~D9;jS!G1d<>NF5MHpwuijYYxG$xB)*#irTVkqe9V(16CSqSKEfnnpm*+?&>+r+LN_uVP2mN_tX^`Q- z$rTnFWVA|8>YJlycw7!%qiBNM^bTyCi`fm>?G2s5_(xslQs$zrpLS7{3h{+rD9x*Z zuvHQCS+RwzvAmRHf|ccI$9!kY*Z%Y;f1?ySG;eJs$#GyQRBs}MDkv5*0ZXZv4*7jr zGjq2@%OCU`e%u^=IvQ8G8j`$vXdRshWLLHHz3|6;XhS?F!?%yt(ub8T(T5Ox7vX7ep6C zNMK=DbhL}t1-FH?EA;lvEm^a&lD?`c5+9UrWpVw7ZIG;dyBvp&79<=$hV$^jbCVS2 z{IL$}7PP99hd?vzidvc zZFw4--{Gc{7JZ=l!&q{(99V!@$k93-RT_p}v$KMd#2wO(FpN!Dw+RZ}q)#*p@#KP5qCPhr9dS>|fQ=6L#I!pfij`BQTO(3dI? z*a9i&c{a}D%;e7$r-IJ``*1$*qDvu@N1qivAPwY_~dsT&J>3I!tX8P_UOkqPGyokFIh!zi^-PhD8(M4 z$N`LER?>@RWBE54sVhNFERRg(rSdKYFLUoxtoCh`)qSHS8cXc(I5sHLH``UCL`JtN zm!+Xd_IHAGNOj>@Th%D51xGR(tEkh;6i2}UGyroIGYOL+T>cHn4s_WD%;(5pd{I<2 z=c-~?fG&KguoxH)#(=m{gUMXXjCekJoLL6blSl9Wr}NNW`%ASSkpymG3I|qDn@wPr zLa}Qol0?N^p9DFU1_EMro%dreB&0497P+VogdPYT)9R06eK-JS6tj>1k-WF=1e;fC zI;>4=q4x49WIPmr({{z=4PF}39GpcT6ShQW``n+rmf-ne`jmSTb6?RwpPIVW4J1qO z643>wS&=#pVN4F|K3o)xenG8#(z2i!p#p*O@lo}gr)^qjLIA$_HzA=JgDvffs;L<= zJu}%(FKj(i!W$fiBbH`l{cR1&b6_kTFu~GJiiIuFcI?1{VV!DGc|ZlwSOGJzs7j*7 z!V)xew|Z2DsrU1Aum|dtpY*x~O}<*CX`wp;py}K`RS#JToJjS7gHGp2zpJ`fxS54! z>&pGSl&I?9*q|QQ@_=hFw@uw|9HirL{{g*2YAaP@ZYUrsgr<47%g<*7X}*$5*Kc~ zX=5;qv=GCtalp#3r|f@d87usf$iSC98kFtRCKzx{@tt4t=0Ry*!2DJ06-7QXKgavV zn{d_~z4|M;|C$F`lBc7`3k@7_U=R`Zhk*VQEaPAVnL)ijvt`N~Mq9RW#-EmwVs6eh z2R2rYm`Ibo6uX-uJFsRunUBdXz!&SgbV@$GMbgfwm%3k79hFzpivtJ!uta*PI}R&n zE-~2n=WIZp^G4@9XN^+hy9(%J@sFYOJ2lkMNN&&xvpKJdt}B|A$v{neA#nZlCfkLK zIAhIJDLr#$D=7XUd(y-pBYL{NbF-WjP9)$tAGUd|q}ZJl(LvvGNfOg7(uL{m!=<5^ zcOYw!E95nN5l|J^$^@YCoRa3F0b)71Q20G~PgXM5)Dvra8xUa=zN$ALQ z(+xML_2RQYBi#+pfdA%^>tXt_mu+gzZ2i)46>Rb1Ieoz$SKH70S@aKx5j*8h+7o1{ z1Mi}tsBIXi*Hdf?Mb=O;NcUmb(tsloja=*w)~wWb2J2KA`hty@K0;`5f~+mDXyqK7 zAp6>HHhk$L7b7x0P@nldN#v%J92fHhNMnZKXamKjQe-U^)9AeN&E?ad$Xlh@7#oaQ z)H+ym@{or&G|KzD?c)#v_+m936-c44-I`=x@5m+l&4J6>EL2CXefP5Dqic#}{@t$* zQIeN4yOo2Y!;x1+S3|Dbw@UU4L>P}a0OOfw?RjD1+fhG%+54#MRc#GxAW!7#oBVab z`pOMdCmGTs&?-j_DN3qp-YWKSNTN##HC{@E!*UypOna{Ju(>%YlnMX-`{cSALki`? z_I;#)n?iBmO%6nchAEWY6uW~W9NxC|24{&_nUq?gM2np z`kMnwk>w`zc8@6b0Y&!}#@2PG2l2cgM> zm9ncmwg(uP3^74lqpS&B`>!|W)&v3tIoufsF@!>aI~J@HC1V|B84Dx$cqeqO{A%&9 zjrf?i_26Xk=__OYB$^1W0gAm(k-JpPdL~X-OD`c=;`D%=IeYb~EqXDQs=zukuxxsv z$5@g$3Ci!0%mIX0ldMQwAAp%Io1wN0B|(;dwO|!nxB>mXGR{5mP(LM1d&757SV z={SCew2RKBb9u?IL04mZGDWUYbqEXcWeB+m%H?go9oIBm`)ZA(1G?zgm>5 zLKo=Wo!p^n2N`8PCE)zg+4?i)x{=F{{pHh|xH=f))_b}&s>?uwr{x!kOFep~6{<3T z^52ioM9CH>?g>83YQ1FkiJ-7*k}so?JJnbs@Er*H-^ z>QM`$E>mYP59GOYJXJL{nO_KmdU0?B4)o*7?Ey<>CAd9}Q-DFI$O*or3uygIsHWUIBb;=OWPZYw6V(GYIN zch{a1v2PoV?^z2!X(Lq*yg6$%Q7I=V_9zAM{Frz)jkik*EJJh=Z-eTXs7ZNJxP@28 zBMIgxOQ}Kjnmp*W$~6JKfc&dMdy2jZ(<-HBmSt3HKZJy5X0zP|ZgNB?b6L+A zk@CMs`VNyt4vds+6Qr!8*kp>VqGE7gqsDYmEcIP8xhkwUK(A)BJ{z6K+vKns6}DS& zt7hVaiV5G}T=$#lMpQV7Rl7*81EXTU2`Y9_Y&k`?Q8AsseTli?kKDWI>%3|jiXO9h zAFCcoK3LfLp1R!sxH4Y2RCp5thI+Q%^H3^@UR|tqOOf3TuMO;cLwz)ICluRa4=+p? zuMS@6Wf*|o)K2i6+74Y-<7P;LJpZRvn6PFLC)~XDRbA<8FMHl^hkP}7Q@6Tn>HB1d z-!E7G=$FaP*crKphUPX15V&r;b${_1a6-n5r?c4zPRO9XHs>d5^R2Cg9prfyT8qF} zZm$Zp|5Zu0syyhZG=D~4c>1`$c>SBQ_U+s3wOTn5r?>C8akdi;Q7it%sA#m*z&FXq z4yC65El4EQtLUJyTmT)dP%Mz*TO80T)e$`fIF{e1DtdT#uz@7@;ln}6 zuj@-((D`cI6sJ#XXX+sVVu!q)xjkz=d(

    4WonmT)?o!`s!9`hosg|jfL28R0*Y< z`-wO&v_23lADrd(EPQA^tH%D1SYxwwdWjT13%*9`MIgGaDpqqRpo6x$ zS$RrS^xVjhwa9j0Q>%~VlB~)l)mM_k-#Ps(Cw$#T8dQR zZnzt=8&aVq%K0FzB+d0PAL25>eQ%US=F(4`$|7sX31!n{$_j;xc{V9&<4Nc^+FQ;`cci1TF)_>Ro<8)|w{YS{W zdd+G6ob%?9!Dq9F9k^bpz(lfSQ0#h&q);(BAcDu10Ljj+B5Z}zM&oYhFkKy-s<;`H z8M@M?ieDSJ)CO|E>SeVKFhi=w>S@P3&@y259egh^#He<%eRI-C=}2nFaqTj4+yore z6uXxqyQvr~W5Mnp7Syw4$B zYV59DZU7nr(u;qN6+xB(wD$1kZ@=sv)SF~|Q1OIi;oWq(X9rjsee{htSI$nDd{ny3 z?+&!HQ6nSOLvqJ!qwj#nez1@Sop4AlZzEZuI4rM~qO-qA)+hk3GQVn3_nZqNq=(ls zp9QS(z(g0cytf3zPft~(&-g5$e(Fh;8f{**fboHN=Qt1bfMUt(nOJ6p?a(W|bLQmJ zCzK6j(5VfwrFz5Baj11k55?aJoj>^?>2m2>>BppQ4!99h70V+AT$e{+FI+v@(Lu1% zJ5_h3c(NcDx-fUiw4n{o9nx4~x9A9Q*lan=SQ|;ZT~WAf7)x%Q1B*5bxpf-f=HL=O z3L-4egMC#q(?RFYsGkZUCAMwMy;OG2+lY!6k}K-)G*$VGY6Ay1Im(i?2m%2*EztERKr zpn#eOZ<&h@hfP1RmbB0`dFA1_I@u*MqNK!R#;a`)O zKMf6Oj{Xz~fbCCFWc{j}nJ+e5-;pQcq+X~)9T)swym@k9C1PRo)CrXwS(9|(JH4t! z$^1blG=94zT9=aW0bUzqz974eElX=hUJH|}+yG-lRUNQHyl~+{?2eP+kCSO#;zIEQ;3BrvX@mkslzMTmWBKBM zTaQf3#JVI>9CzMh$C=fFiWB5YyLMKZbD}xyhRFiU+2-ix2>jau?Pl}nIM6D%VC{5s z#HN57Z&t{Y$ysJxLrq*3o6~BIaj<2fGLpG7uYT~~&CMMyE*1w4KUpx2+5{c+Jqa?c zg04cQkv0L^F4+DYtq`P-I2F|>DDo>)-12XN!N%T7je=dg0{RrVcCUFtD+BZaf8se9 zmgTnq16M6xr=w7VLA0V(Q6z+zLYDY~=%Y`>i39`}7RnBNuF=dw2?T5=5D3$^B+c^g zpdlZL-z`$#3aJm!GAmuwS_Ts|3{5*U%IC@=E?~4;##lS>Ie@)h8kIN2sK>rt{Ig$> zjokE@1LsZbG0|fs6boeqc~I^qjrGn2n)>GG;%{BiJ8YViX9KQyXrVf7Eh!1Fk!DEm z@cZGchmU7nlC`8?(iwf%H7PI$z)H zr3!glNFdKuJryHmo)zBeGNzRs>OgI`MYgaKRyr#WhIA>WM_<(1CDGAM$_LO2u0oO=rIl1j zDkIZenr};kU;zR;7=smqo26FifGb?W?b1%K{XEQZM!pqr&ZH`g{ZN`@ zkEFdov+zhd>eYS0+aL4dzhM5+-nJ}dMzD@w>;A?6yz|m;oP4*&8)kKP9 zQS3&FqyyUzugLq7r6)7EuI04Ep+KP zF?eu7h(RliV!gRP!U-YNGS{j&b62m!dR;8Ib&o-IT)cr<(cq(1m++26q$Wu;>!%PV7%Wv%yoM7_WymNb3nn$X+ zcwiiOi)T}Vh&8zD9fNt`f+92bBjl;T?;G6 zaf6bTliFa|cUuDwnrA>ctPHV`(}cZ18@!YF`u04n&_NdaHx9U+^irei%Z~D4r1x+A zLXY&c{jPD6853?r?O!;-XnZ6r4LS(9rX!fI4(yxVWMYu4rr4DfSvJ(9xXcr4F=0>h zcff~Xg{?t}{jgwpX3HpvnR8@`c{(>2E#kmfSz$uc`yTrCMM4Bx9tg7UXG;XT1xw?`68>s=5# zOUr0Omnq_eSjwQz@V^6%%6eq0SbSntBy#U!xeMMj{I1w{du!M!x}T_*hMtZ(9aS23 zn{>)6qp(H>{|^0^Fp%l6$=^4Buw4EnqeN=`+KoI?F_D}y`H~-?*gA?-Q!!UUcY9a5 z8vbpLKJmTguV4D%>FAc|lU|vj``2ei6NwizueYzerDxjd?&+GIv^`bT)u6bbvZ$Z}5oDMbCwA)zFZXdWa?w5R-)yu9TGbN@N_+_U`uL?=u4uA|NG#k;D! z9t5qMGT!CJ8p|-k$+Z5^@|WJKE0!(8C<}w*x{Oyp`O5Fi`ygMb1b_v0vLQj9D}nby zjZ*0Ny+)>o@xsw6Y?$0+C4fzynTQvb0jxYX=Z3j#mj%76Xpp(;hXh)3@OieCc}VK$ zhrjx0DHxQB$X#Ds`Q@^oPVDDxb2(*Pa~``rK0}^3 z5eRIJs4j}9*bNj}2NjX@oqz^KyyU(or0x$wC)EqH>0f+Mwmz^15Vn3#YsTXQ2*-65 zE0@0MWL|CHu)Ado^#doNX5cS3nRg#3S|YY9F9oCqM2qt1eIC7`rBg5a!^W9D!N2U^ z9D)oB`^7g%k30*yT~R>?=<0m2CJOJk>CqgLtFG{}lfAPB$1iq$tU)FBdpO~k%3JdF zz2<#HuavW4p||LmAGDCo#I{&yAw%J0+(&Cr1if0eS8;yf08`>%wHm)(559_9oyWDCI9Js@1qshMd3B^95NIw?h zVwXTXB#S#k()|qeX(*qC1^*hn)WH^sbJWe!`?A}>{g2&=c-bIq-XPu*Xee;&@hA)J z@u>2^W~8T1xarOY+f)V)Fm&&?$~J)f_Ljh0(D8yKc7X(|>2TyLb573Nha70aPDL_v z&;ttUFrCmKE(>iC>u9`RbHUy4SQ~wDS_Yj9tqKN(!5a*SpNQLp2b5I;2rd>#3^L(` zvd=(-;2hD1#Ie<~E5HI^_*GYm?uQzV^Q9G@4!g1b(z+XI8Y2Ktg?R}`4>tfh?p}{1 zngDR+WcDdV2C2xw@UH2Qu2$d%zY;e4HM}aXN>Hq=53Aud$RGRk&>w)Ck0mZA6_vE% z8s1klE!sKXcyApv?qB3-Ri7)53tA$sOON_ic|E%m&mw3134(fYH9Vb5eb~0yHM}Oe zn?5G!j95Q)Cn)d4vG|TpnEa3{)8hlX$+p>ePr9H?))G-8$f0}a;w68sTv92%Chb+6 zRaUEe`FaHq14s_?)54&5 zw?k)20zY9Qb4^1_nAIzxeHJ{6qgq}GN81|y*QTHSfi~iz@B7}T$VzU;$$?kLOcNAs zrq~3EY@{MFk%d}YPn@teyOS>Uy9uV|e}ZP-%JjFznYP;n?>P(JUtk192J^E5vY#6$ z9C*b#X99{^iakb=DyaE}QpqwXm24C0fGXy`4C!QI+?(8UVUO12S~$Bmv_#S-jFF=v zE>?ZxDrR_fe^w2zfM4O#AZ&@iu1z#d7|!kwY7<_YyUzav$q3yEy)MWv+`IsJ0*(V` zY!h8aFT=}*N0*JyZ@1bMy?94A+SPGkEap4?LFV-{&u`Bh7%diRYPQSzgLX;Vp^z8L zY_T<@B|@*b$M(=2;4q}hPO7n+`yrfxZaDXoVR-L{q8|DtTpiy`TQMDD1%nkskLwo8 z>r!FT3XaauxwJ&!W-3~A-AC_vmNdw-JUn} z6@vFCO4ho!sB|+cTy99Wg1$SJkaB|KSfMqVVR_j(E?Bb+ElJv66V2OEUX!G{g+`SU zk1Ieh(!yKodqte&{nV+P=~e8IosmvB_Yh2=m3ZMY)i$|Td!>7kxzlUGTPy<(1Wfg5 zhPr-oq1r85tp!2hP8(1Ta{RYg`DspYCJJWx_4q>lu5WqWo5oeBe9oWOlHIS3(+e$0 zBT{NdDE1IV_EV9|AN@<#PuEcYes|f9rT<<)^(`?w(XnLj!c56EX{=ARI#mQR#NLM2 ziSDsJaqRl=V}949gHl|#p!7pt2w|EhiX0%a$x1|;^c>pWunJwIL7i`T%I{K{qO(yTjR2Gdtt(Ca+}+- z?aard-!+5I_d$l%r%q7;Y2JB518e%;)O$9o8hJfxWS2fc zpsVgOsSLa?Y;uR(6Y~0A^{J*S-bf3BcQ2RTVK7Ipn=|Oz4qGc6aXYUlM6*rusp9rL z{7q|lwd!0DP(TKnd(7vu%0Lvbz2xt(Nr8X(`p@>7YcahhCJLHtN4PxOC^m^Ao2ba! z;?t^KfrHW~PERG7{(WzzuoZ$2Njx|$9pVd+o~*K0y@dsnz`-h67{NZLi&y;9ugtqv zEHsD22;u`D`t6vN4{5Pz=QGoK=C4*J!WZ+Z&!#ZDTXpk)y|tgG`E&Mu&B=9e+?Bub z&6~~Ug`5uSKr@0#>kKRg~wblGGbr*!ZB@ekrPHH&t2v3XC8Iq;%zvvr)wd2 zh_zPOTn+54cLZAkK~7E7s~B+q-%#u`@Ckod>}O8Gzy%ZzTz_DJvH?puZ;1P4T2+oZ z-)ARqdkk$vkjFzW&5IZ!d1(QfOyrHbFt=5<+P6iud;*Hy9o5K0pWpE2t4{Tu`Smth z?_27zW7_R`hC=VoknZ`1-`wJ>^}Z@?Bg^9c>FQhOmyWXx;Muq8xCI}MFA3{`kB1!}&u$j65idsJt$FZ5g zP`i!(>%{DGID%=yb~ytT`&OCsh86J$YJB9grGc^Q6jZn3_ZEZ*M|(F^K_ zYvvd52SA|{dxi=_%j|_Ds{u12E@FCU19XfsM9stZ~TWJEA zVv2>0ZUGg!jTbG#KCD*RS^jaTd#obRz1YmdnrNh?)v5~l+h#*&7Zfx;bwZEihO|I* zKnXYLXw1VkiQDKJULOy4#VD?+RiPoW!?HAf$k@kA@kseuqN@=ywFk$;Sf7ytfFfplo-d{d4uWoi&OjxK_F7@blZHdTpJ0@sU#RvwK_tYCm zt>B14=Q(i}Vt7`pL&b!q;f8)SPN<+hID2QmIV!m9VI0`?uz;Z$*V{t$OtAy~fbtX9 z^9#$s#^~l8R!>0Od>A~&pH^aM!a)?Z|F3s`ZS*dF-s&NQ^p_6VJRm$kHIU*tnr`Q0CT?1yKu`CE_Q<1WXw{k| zNxllp&Y}g^yrFLFGE|*at83j4gMf%8OI;B%^b5(?x#zntuN2;QMV-A?*;>&Z2E^w< zBPd7RP2)NIH4KUXtNkna=^kYvH8bIT4Hq(6bmH7IaDNaJ?Q8}0d5#?G8)XrC$dN2BRS;SqDYk$@(UzcuD#4| z=@GwfA$$^;jc;e3`kf2Y)XW-`t`tRyfZAF|AEvK}?UiA*_RDR$!!&tiW-e^=lG6g= zxc2?89zXuttM21MaD7IMe8{`4q5H){1tdu1^S|EAFw{ALb+U6nO~QEcI1U_nvyjJW z5O#?(*%ErcM9VMp)6-o+Pn`BkIu_J~99M1*t*4vZYj}5pu1imm8}^_Ox;_X%SpBa2Tg!$U_{VlAH|kXq=88%p>qiPB=ex($mj~Qi0nK8U2{-E2e)=~3XssBG)2Hva4#=;55g-*Ctx0U!^?xOCQwXF(Mto`0DK?cN z+YChaSSz|Wyep(b8W*9#GU5w#i|UH_13r+HV&SN!i}_fFJbi@%TYn8L%!#%582LrC zFqBPCbHaz?=GL3Yr#t_)N?0>J@3^mre!ZtoSj>vG^XT4A5SmRPXdSjqpP7hPO$L$%OCQND1W&)||3?_HI4DsuVAX_O2+IPts-FN&lu!q{e)C<3&uLI$+&ESj58E_iX!z?Opyol?(lbBhx7@zzbzs0?=cY&C{VZj77t|62TvfSQx>*G(kXfw#fOO(1fB zV#_E}OhtCPY?EDr3ZiJ2A~DieogxX)p?6JM21Lef-?;(YY0VL9|7FM5?k(B4tm_-M zm!RTIt^a|A*Z+RxwLRY&z81Z3D{#C%_3MHvrAj*2);fl^gM!ISj{5~CSDSM0VgAG2 z2%I0(q+cXkxLG0xUZWtmGGgPli(;WmD;K=lG+4(-my`Sz; z4U3mXH)n&}88zO~MPCW5_Fo%pc*fr8#Ia*)+^~7!0KfU8Yp;2imcterSm?W|r_;jI zm{?{gG`}@?zn>;v(nAlp6-`@Bq5_W2&vh9a$#aa$#0x)6vx{S!bF)G0 z@RE}qnY{DYzxh{lU)Ny~V!?ae^27A7+~Ch;J+4C?dA&?L57iZPbhZC5<@^NT;n`Gj zLB_ME?0>`!8G~!+vh@XzIu9wO9sUr`2T(3`E(!SI`kiVk0+2_Q%H@80WYK2Vv6jcA~Qgw zs7TyNchTi^8|)Seq)FZxaCAhiXX*$ZVXumzuNGS@@$+m4PVdJ-4}y9_GlP?vhYGE# zO$Zi06t#zMg_L;Um=B5VKXF@p+%OF`|HNMtBa=Vv*q3J+QPcMm`5Lm*fg^3zCU7pL zSfD;C#L8H7Jh1K*Bk~!(D6D;!Mp)jJgA+2VQJx9wY_&;r}wR;y@`XV8VRLZEB3J(#&R?^v)OHU~2|((6e}a{llQCnIPIKmX0U zB+7w7qcs6dBE`m0WIYwxtinKQVdwx~b3l33=Zs&IU$e>%Gyc-KvdX}{bQIg)z|Fu# z{L8{~kBnA!ddBugq~8fh9Ei&wBmebxk-z`1pZ{jrAEj$3b`?dUY;TJ>{^Ri%8?moh zN3p9ZvJ(2aL}}qGhX4CtC_MXAIYD94&q6MKIGz$$2liVnl(=GXKkR$M>s14MEo)Fz zUygfTWY$GNt~UAxfkZN?_Aj(`-sagXnH^bWS4J){X3yK}HR;-VM6vx8c|b*O2-bTJiPY|!*$X?UqyB4P_oSnfyes*A^B&7F5RX+RCA?Nv z(*_hr*ymTt&k$`?wlW5$EG#vEOL$?WNJrldYW3-&lczp)LOKdy7Jz+MHtdU`QLB7m zvkFCJOMO356-X*vj;i*})I$eh#?*4Uh1H-^pjK7tTNwr$>8Nmcw|R%(#8S9Q{yyl% z%277}Z*01U#_)JDlfsrta68=u@7xW)r*2N8st8ipyZHDMfAe}(jqEaMXE4=R!Ea|Y z$YK7j!**|3M%pM+l9S%*{KZ$_Frv@rcNNRXc5ZQ92QF$kU}C}YC>AnEJE=(A>bFGf z738_q`(dA74}FZ>391k4o7cdvri&C!?#L}3>r<~dt?DCB6nEt}NFrzo7V#Cd{r*=Vb9G?<aZ^wCR8Xr98yTx+X)&6 zPHexvjTkXf(Vy}?BTh1@fp3#DQ^=<#AMjO*ZKX&Hi0y?n$Zjj{dVHYFSUmC=Ri+=$ z&t06jIM**@ajmRg)v;i8c*f!a5Ww3D-%@PoXo(o$@0a$_n@NiQedQkM9`{Y5HOdp@ z9KFZAU09=htV{)tg3G*)1vf~FI%;VyFdMZ*oMS89+Td!2v{$-%=#DzcJ^w@DyWCnL zkm=+my~|BU$3ofL4btFTM;{7L@=jcQf;0x64oO^W<@bIZGxN+H;dC_O+r=`Jh zExTapH}9GE0Xgh0+(Ls=s|;ngG9)=ta3|J@lDt}Fmlh56pCG53l{kCBRCBsyJzhBt zJMNn~;dSzEm%qGcZg(6O78VTRBi?RNJ97`#m9o%e=8SZ+8yY#Ssz{vYb{Xnq%W16B zKE+<*=S$H&K{5NQ(ocmaoOLw5x+T{G2N&Lp&(%N?=6%Va^#1%7=mr@c4YWlAa2&&P zT2;REX^2)G1#-A?-m4%9Rl`I5QY_Ql%Qw6eZtQh)YGBpdPvTg^@1j*z21Z${0A)SE zI9a_H+-x28I01~xYeP4!Qpe2!a@-a=qBmKeFH&qXMH-<7g5N{hRY~5O zF8Kg+Omb10OLr>DC3;WXWgTOdH@a5}@NI4|_(n*pe2i(7 zCn*e}r~GbqKk*MSGn!uAOF%uJrj#a2!z4y}XW&|O!*_s3rdjmTPhaoY!EJvU@I zuxkxGAtPMtREph3kt9PwFOs>m%79g@832LRz1a{1sPIe_>>|Y;klx-JTq>EsrZukM z;4r`Ax;5jUX8m;TFgR}HZzKg%pgCjYr+S!T4^m_w6^T`DdV2Y$Y5PDg8bc*zp`9U^ zE_~{QpE1T|kiD%lD4y37c2KSva9ayI+r9iu9$v!WO&kjqOTkZ_?)@+-Jk|FeeS6*$ zr&Ql;_<49n7Qs|HEKRngYT^2+zrf}XT^iR%J>p?bOM~Nn|5l$0csS1pp0_I^YDghB z%i_4Z0CLjgJ8+a@furaE+$ozG4#Q6w)$l4h_bjS3ML(^3G ze7&$PG?#xH+^00}7spWgcFfhA@C zr}=B=BQgc?v-HAdRWBdAmTE&g6>Y*TQ%mR^@OyD{Rl+;rj9acY;eLWClt!<*P`nF0 z-x@&{AImr%(HmyBL~IhARrm5QkYblsCf@srQ!9gcw)nu*X=jMxIZg;30}x-f1Y5k$ z1-APRmPVKt48JCIp%&_%lY^@jF>K1Lr$H5-cZ043Vxi}S>BVnofJun16yLOS=G1(0 zH~_=^QMK5wqC8N!0{pRglvhTIAIZY-u=QzbeWosoBnY~Eor-ul4jq=%1GLyY1 zMO`CkSJXhd6QkWN5f^zN+YO#ikDw6P9n!p0eFx}Gp^pX7j}UUyH(CvudSxdxNOmlU z{yzo=2#k>9w&qr6$12rdS|r)>4sWbb@lN@~U(t(ebf^f6bS#NwISuB5=}+vMxcc=rqyG4)Ri6 zQUO}={;lFHpR&+Y$r(wi_c-mv%cpXzpnCbp_<-mCYFb^Z5xmzeUks@)p zE80)=bZuxF#&V%b$`iLUWucf9u)8@hpC@aPV*Z@HUt31X<-c>I{EddAw5tS+?rXEY z)SBSsAjR&ZNC_2L9=22csiJ#&DFmbM2u`>k6_zi&$?KCoamrrQ8`>bg%Ix8nh2rb2 z-#+=(T~za1yWi^i#-*h>zkK49My<%CR-{l>KdAYNe#sxQsFEsh`$Z_1bHCbm;X*xi$ z4=HjV>W`uSw~ur~)Fel=k&hhR(0Cm^Enk!w44ry#3JUZjmE!YkZz!e@w>oRk4C&?^ z^#>J!Tk`xN#wvUaYdMiQ0xMJT7_uY6qv9U&Fie9RZ^%mFPTZikh$n`4!5f8TGgHV( zHPVKz7479Cap*2Ls9SAk4l)%29Pf!!v~TasxGPr*2HDQiSRTppx+Z-brE)U@v`ye!EP z_iF+hE zsiGn`FRloLKQ614`dYM=3$HCg-I@#hmd0-yaWuL2 z*PoJFZa8w_Joil#9G$0F2os;BB8!1rv7SCUL#tXnC4*iwKPG%{_?jtM{AzWo?~$by zlK#-_MX912%spPApv|pp<`XA8&hS1u1E@KDbNH*mi>5t<=*=D&XMn!1)Ot6_qC|!# z@Ob3?%@KGnMu4tM56;!g@R`BzuIYGRe`vR};i+=EQ?j0<1>;$+7SNu!GJ%^lSQ>H9 zFAgoJd)0$FiHq<0YE_#8PA$r%w}j`&Qh7VVPk}ioQgltb$cy*CAj_c-LY2cqzYIwi zT`OylFFUn}>jaMY%&a~f)1+HH`q*#ewuJTEyCQzo$tAG*4h!ypU@^}vnYpQ0<$Ibm zLh~E``B+@#^|82-eiRf7gjWgPHM5>V*11NYgR z7ez9u$aY16&sKinY@Od$*mggbo?`2uv#!=ZIam*JeOO+;b^3nyON$1jb#%#F4NJOw zn}Rz-tJx)iv62- z)Dm%AS>=UebS&7eyr;B>0`1s=WMXz~1fOTaSO(H|&54#!qjmYg!QGjp%z>>-qltC- zfMNkL70ADcjn_H!IrpCVcLW<;8WhF6ZL{nBpz;!qGU$u|6y+?Z%RD+Fa@1Ityk61H zyRX3HcanEGy^%fOk_W`VO=4_j!4{T$-bo;;uK|8>>{8IGaOCa2w_taVN6+pazg@Bs zVocxzFP=^P7>_I?X0*L|nernK^M+rCEs?O$!ix&SdliWSO)|5KuUC|S5Uhbf>lBIg z!G;y2$+|C&V|S@5JhQy|L*v*a?*d7ID06C?OPjs5!p4r1F#^EK$2j3+@|w~lA9D&v zhsB8nQbuG~LuIgJ24#ffp`x>h&h-Yq*9-|F1iIaX1G0EYJhT%>iLl`b%#f_Z1-ew4 zf!7sv{4^PG_!x?d*UlejjK$(4jU6suSQ3k4kLLn6xNIEIoi|r6c;2IN;J}^*je@N3 z>Oj~&^Kgl(R;-@W7#QcKL2jgK#UUxg`3iYAgX$n2_lQQLl+56Qbi{evmyn!f{7cS@ z=Woo&dGWC0zG@jcU*A0UvX!cS2F-Bf7wVXMhzBau=q?zvA6AMoB^nHaLeB;>WBAC&{c&0|faQ;T6d%BGzlA>-_>ljq*Ka0-w=>;78T5HYt$)fa@J~Ro1gS>A zY8>-#l_h!i%QV&M7Scwqo?R`|f}mrLl#- zY6;JrFg*EW{DYvkjk}@DhtRZr*MWCK8%^pSo>1%~iu6;FDgF(@BYvx=SBksDZNgYk z?psOSWSZSen035*x)HFjY>V?c|8~VbML+NbbbEAzK35~%K66uGOT>q~ErCe*T1j_% zV54NDyF*Nz1(5q-QU(SU2cBA!<*LiJ?mnhCcK4SR&YT!z?6}V`Sn;&ApOo2#GYxeu~%Uay4K{%n(Zzr4Pvi!6v6X+>wc*q5~R&)ot zo9ylGjG2MhlEJandCR!zX}kWWIYu0I6J-GpN!|vM>clyTBqrFPie4Gm6$E{khEvH* zlY0~RV#RKE<+?cy@@nW^vFG~r#k7rWA^YNSyB%1@!zj+!WOk2d3nI=19P+SnzTt;Ffs>}SndLn7>9k6ne}sb z`(=^MU`C)qrV=*Qt6A)*g;l65w2|&rd`Q-X>xIaGQApWcHpUu6jdYK{IJPsSS|`=* zQX2i=pPZd^o}_ZCjd9#fAUR}`J1C%79YwOK$O2IZKU#DbT2>(Pi_y9yU~)yF$RW`b zB#7NVvq%j4E1+IGMYJ>Ej|v%!ppy51a$8uU03*W%BHWAF%QlS}EW@a=6KHA!avsb# zg5{Uz7VIGFF~?ofr)pJ$;l(b?mygdTjT=M5SQ_Q$c;Hp~^`x^Jmob2?Dy5eQjaqvQE(nuNJlXn_MELzCeXsE=jWA^!{NAv26f-t zaW29LI(_hrRC2_DqdXs*n4(h@3xr;^RAfaU@Pa^!;I3~YWT4vIYGsJXWG2nK2O2Q@ zm{|8dCRbhNaoaBya@zU43$o8blf1KocZ>y(;v4t*J@R#;!qAw2KHf!d z+;vxxlvx)7A!+4&R@o**rCb9~|3R=(h5Q~mI^uF@r5NPcp=afUx`lK7<}+B?8RpO4 zh~p7Am~Z>_8+TuI0aR3i>XH|TP{u3+$n!_BNgO}JG2Q-hK#XbFaX-Wf5R-09+cDG2 zXkKF8p1zS}IIwxCFflKCDHd9t@{#(!P53FZZE;IPjNpjh8R;pK;Eg3Q-6Y=!IV397 z89sPDQBX}c2%|(tK*pnKVS}(G;sNw0WccoJNm64m76f#U&gl^>>!yo&_5jt(5o2L- z8+m%%_;`K*PKcqR!?#X))k{44NtZY&d>xbTv-6GB%2?le#TsBCen>JTmlPS&mWZf; zTLh2FB^^MvektJI{LJ8vh-F%Jk$7EjBX76=GJD|Ti@Or zIj&CzeJUd>hrifzRSVlyDPAWbQq#e7L(MQA#|K7>>Y-MmhIee%&7i&fTwo4v6L0nD zouyTs2x$cpiVwh4m3s8f+6{#iAB!Oh*+0lwEzQ%%^-ucA!8?ZcK+(k{GZ&#J^8K5PeXP#dz-6+?*r0s; zd5E6-WAq7b+M>e1PR7`k)d-ylL=*d*s71!PSxO* z^t)7fX}ZOk@^~H;>|vw@@{d?;6%XbC6a{C*$Q!5HZ3*Iv6f>h>7qfr}n`3va?y-Z%o<`$*B9y^e}ut#=5 z)c(w$hqxNUw&h>@RukF4EuHPav6Xz2*h(73Zl_=tA}_e>WvQZga8&bxD|`~kMF`kk zmtL1%0is}(2dnVuA=jnp-s$cW=FD&{$FvI)MjP~JKmNlJMG~V~+_s zGAMQjMN+89mF!Nh^fwKEL-jX^zHOMkU3n5{gW>@v==|QE`;g@O-}0#N%2FQ}=85M_Kbe;RTI*W30XKUgf1IetZ@@56ZTj<#Z@e?tybIPsW9daw zE#7aHUgmRvE`$3A`7I_hxxrw*&i* z`DZZ3kJo${o4@8{R2K<~5 zUo2|C0BWY7QQj&Wa4mz#JxCB5Iz}+<*9C?H0=;XM)`*%5jusi&M#l8qu~NZ z^mj~)n|sg2F1BKfNFKRF+Qn9ILh@wM?>>u^8v%3j+uylLHajq2icJ8slVUR{vV)3z zw0PNldX-E0!Xmm$+!C>NMpVFgd8Xi&G*fCg|3I3e>IfPVi0PIyPZZe2 zh?2!v^FLvF7BGLDv1B&q_Ii2gm-F6TY&0Q1x!sgT4m+?3X)!S&Cn>g$BF8}l6zE*^ z&|}sJ<&QYs+j-@F$NU=QFy*TSsj?46&8khIHM|RfdReRND?cuPa6etO zz))`qSpbCD7>yEVc$lq3FMl~U0C6%klaKt{0+*i|!PNfUzyEvEI)!|0GG})v_7(*K zUJiP)xW=;?9&N%5zZzND%+*t_N_*z5 znNm#`O^XIzpL#{~v<_%BD3lG4*e1j<$O+ejFuC|nCA>=@wYF+$0th^Idz3-dTNXrN ztLZF~q(0#av?7dd&L$}PfArSy9hPU@{*(H`X}37zoF>(A-Q^_d>D8}#gKVp;ok5ig z4bmq(BJsR&*bs~JZOsL%uxiafHok`wRwuVh-n(GFn!P3kEf&_kM!A8I?3iDvhaQSq z9tR<<*#r4Tel|3yVe@0BqI_W~ zJA#{V3)V!p@Y4M>C>Fn3d5I6L{8C(F+k^vO>!kJ2*dNEH^6oKL)eV7K;C~qNKFuoV zz0lIV?BduknHuo5fL|L|v){MoDalQ4`2xp{UXo~&{=8m_?WRZ<6`AUbZL>$zanRM6 z95(nRKWgEC_DA|)QJqcda#_&v+2QJZ~L_vWhnTZLH z6{gI|4Axv?dzl6x6M?o~9lcuwq9Pe|AA2Kw-Rxygow`8?vwUIH{B^uC_ywwE)iR7R zBr`+AH*7jEexf7-{yHt-wN=s|d71*${-}EyDo0!&0nm9rcKJv1BCF?XBOG^CCKiO* z+k~0y4(DC`44)!86WGMhLM3=6o$n7SNunY#_EYxDisxMP{zzS9r>yme4|({ID+q?4 zv-eNSt6t<=|GCcG#pbfYIWTZ6>=d*MqG4r1;o>@a&C)L4O6nSri?>94B#vHq$Q9y< zVfmiNq`CBVuJ1d82Akyt7cYoA~b952~7vzK||R4dy#jL6zOw@_Qy}a$!v5`}SV-g+QIVOVt^6+KmIb?JEz_#{ zpb)EI_KBGnWw`6=q4mNFFC;S33p=4bdsza%V0K5uUI;nW(1W13eSXQGN|*fpe1uM| zQ+GmmD_W%YT!zKQhRa1uRzy*7*><=a56DIVpGDAcz3}yz*=;UH*u0~;@JF)Ffuqs6 zCP3IuvB?zKOhxw5#eNOJ)ljTiwIE(lCe|GDTPwg^emw66Y4g4nfcqiaP36l!VosC& z^5JZD2RDnd!*io6zhOrlj`jw^1z&x_(tMP_g;RAH

    3&s$v&wRb-t!|>DrkJg1Yk04fJqgell4M2y)vo9dx)Ib8af4Ya*bniRUHmL^Uj7+mY2(~_!&zMQf&@$WKzxbn-*U+w!Y zN`&H-HlZPKj3O#gqVsZCV1T0hPC%Um-=o(5D2;WW>EEbX@(`rdR|>Mi@%(ZA?l2Tt z9tQ}m1JmKg2adcRIwrU@V0eVRZryIS1emqhvC4vs9V*W~$q75u zzCX!2&2vZ&yF6LQ7xgmnJQN(z4i$Gn10+^-j>fLx9GroWD5=Z zACeM}WadNXY6!*lhjzHehHLs=%NO>&namhoYE^dxTLQuQToE4*TmMGP^kgPW^0_QU zU9+f29OHUhXjhbRf|c2lWJ44^H`MKObky&S2ISDs-Q?uXlu?pc4h)A_llqrO6x&ac z2S^T%ls_%J65ei~K_*eKYR<48SKnsH8y7^Mg zQQtiJs4v8LC8^W?vXMp7>C|aC^t&5b&BvtQ6@KEWsDJ`#qx*&hMdA$KUD9Iy>41um zm7;Z|k{V$vy6KovCH!KkK6Unr_*7Wua%8q78aV zp>B)O%eKg$Ivr8(3F%;t0cQaGM%ik;4BszR);|A&d$W6mCng_Kr|ne$cR(iZb47(G z@(a9zqDyXT-;NJ7qAxbNr_L`t1 zi(<7DNdrD{w^j13UT36dr2BnVvfTlliU#3X|5a|-Rs}Iz?4tG)+11+Gn9@O6#4XjGDinYzf4&%LB@@%9|d` zGoS!;Bfmg$K&kc4oVq+p)Fs9;2)of+nav(|+?Wkurf9>kjjxr}!&+{nH+VhlU`!I^TJNbdrz{;-~{gQ`Y|Hd!KF$dmO zUpHApnkW{?n?Ir=J4p_ym8}ZMaR2N}wX$Y;PT1+Ws7vtxLijBa(cx=-I~E}6ZWbFg zJ;kpyAdx*c^W!-co-GlbBymobXPOFQbDgA~lrM}A+%f9_d&>h~H^@5{^pn1M~~`8CVm`%@GZHGOj+7hqZq zs!@Fnjwd;wkQ1mV?{6``H0SVr-dSMXv{~j&uhz|$(r|lsi)EEsN-U*NIJLOkOM>Kh>4Bb zL$MGl)=`ldk!cbG6U_#5o`+W^3#Wa5>nt5PK(D_Klm}t`dHtL z!*iDH$0(i*C4Btr@0c$-7VLPt>WVl%u#N}fLp1Qc^b=?g93nx%;JBN}x!wQysrmNUVU37|4R0TLqKFdh;a?MWGr7*& zhk~V$1(5Ct#j;8amQ=X+(AmHSJ~5C*P;rFq*m215Nlvh(cKkT?d*3j+Bqx4#`BQS- zfj2^JCQe2(#Wqsp3>Dd|GO&suDb*q8MtO-x9^ELf^3UZ%S#DT{&wyyCWjez<2Xt)W zdAa;}3G##tF@(_Fbeo%2bTw#o!*B@|mkk=;~emwbS^0{X<7YJqOfF1}`fzd@bk zm94%3*(ePZKzKLE9xHQzTOM_dkr{fJ!^q&gB2+MVr=ySf!65BGa&*YKhIfiIc-q=< z*$hY40>sMuEaRyE%(X9>FI2BdO54J+hMQ`naWHJ0VV9IkC#fGnep^THi~w}p@<>vz z(9zd}OFZ_?Tw!a3j6Pvwz|H6b+4!v2KMKU=dm%39a9~?wVdsRjiWzin@KB#4XjTYT z`5p>D!JGCs*NN88E0btYTGX?7i_6U6dW*}QVRMAik1qQ>XGU$wep%@90;>F0dZiN*7>2YmQ+TA{kaoX)Z z@0OR8Q8vtz*rw$?nb9>7-xL3yBs#Fi3tq>Fin>gSg^vDIDl%TuLk}=(-e_mAY{bZc z*DCu+l?ENqcZ15*z5F;%)GbaFDpeQVfX@NCkp9)*H2xws&?jp;*;rvz!p`sMANc#_0s>f zBER_U&m;f*i{Jg>w~&^o0S(?_l@RYJgghgMt5;C9+MUxKHNIT zh>UmMZ9hu#92glVOptMiVnITql!`>o)qD|17%?@xqceK>8InE)GKY1^@#npvpV(wt z1JWdI;CN=xAE~;6^s>sp2IoY<+9}7SXXUwcr*PdAt!h7s0Yh{`P$4)vqd;`s0|U#g zfu#X9H8dQc^V0B$jZiZ*3@7MNr?Pw7%%$utD8{!yEWcNg77Q$bC_AO6*HM@!OApHp zF5v+k3Nj7elHW%igYfyAa^O_l#4X=UI>Wi^yl zRUp|zM-8>cSpLv1_|HXqBDuem`uU`F*@<7pUi;7V)@4I>BV z>SZ@Tg|AL}k*8H@`Gd-Bk}|&r`JI3UVJh$I7U*3pO6gYGLX z`)`|#&wk8a5hpW9caiNA6@X8yFbs7-QQ@*>YNrtGNe11*0#*F*$VqBE&VZO}W^k|c zF(`rcN;M76Wgt4x3?bO$fDT4;&i#yQfoMge9Q#Y4cX2$%V8j&K{E(Y5a9GN^U2iP? z?W?|?!6u4VEqelz)i;99LdLNatdCxZYLAymWneo~Ngso{r2H7(!!=|I)rR`KPYuP;a5?gG(rLu5yTO4nM|aO>e_w?XMjGQfxH zDD9=u(``TvJ^3!Lo8H25A%2#wT3i0$zI_7d*oU^H;>kO-SVn~4rG*T&{P3E zU@foBO;atb8zQaM)2M@l6-7DpRp6k(xJ`@0t~n8Z zZaix)d-IwUEkGaZ2#b?Qv2hex4^2r>?sh~S4=%3;````V0y{ZOFT|Zy$bR7n_PN;d zpYx;430++6Zzb>b+i+uoKVbsazXHOLx z-HoLew|zj?JFvTvXY%DwrPys0NqQ#kcazrzv63TxC6GAUujr;V_*dVX84^vMe>^Wy zQak}QO2&`bDl;%*eX}q-Hb2d6flIus$(U*c$E|<6@)r`zt#Qn86By7WHgX+Fp;$;? zY(YY0G!965jHLb`r(^^!F`mh~OURuB8r{6Z3i+A%6-zN1A z?2&eu7=RXvJx`G)V72v$W4m2G4~h%hBdvg5uJde#OFdmB*y_3u%4Q!Zj`(22Q!m^z z8`!vbz`7-0MRMq!u)f6eY9&3hx@au0Z3Arv6dg{P^$^IgTOyt)&cL~KOLOQvy2p1P z{eV6W862p@5th-N>INZ*ngK(9FInePGxKIp$}EGdSPn3Lr2vgv7ij9;lHX(6k-uMg zl^gk`wO}z{aK`eQ3$`q;I-^xj_HI6VXSxwu`b9z4NSXsf3lu&^L^^g;EbM-As7Ulg z(KAhz1LYvhsvMPC`(69y8N0zy&lP1ZZ_j(`^oi^Fg*s4t)pUUt`Ov`$zj){c`w(=l zpE_lRZV9t_Sbb>nY=IUoD6sZkPH365_~g7_{-Y5sf9U)}2l>c>(ejB2S}s%U#}v5$ z3Lnf7KO{)bQD5X?*&PyEXjQl8w)3FR3zX{Vo-jB7lN40pi3IJST+SZ=lY{m9P#DEa zW(+;9n9w<@?jfjz9?#R<^teC2)Hjhpv*~W>5kJh>9P?|GU-N3>^@iV)YrF9hm`Ry|F*^v+P@D&7qaHE=d#o~SGo8&;dGDSxpAyto^tmxGKGye9wa zdp|EWkH|S}4V{G;8>-(W@sa>0JDhj8Yc_?Vj?4*XbaP60trJZWW<4%mK9{z_xb$Olk+yqO~=Di0>E13M}^O&paZirqw!cq$SCX`W~J#URCu`MWk~1ZrnAcige>RI6HU zA2-$HBOm$VwSdK>MgQ3E{I?~rdJ#yl-@D0F4sl>3vvpx;-@Kk#c)Hp@f7VH5T6i~I z3{Gb~)B!imjPu_tuJn%Qp$i)ynBabJZc9WvFH=_SpWwYsa+;(~Z;+q$OO>4rs}$>f zZM|r50NIP3)se`x!ExfJ0a1&L)};AIF*&51Te#MNEA}s%*q#q57W&XWfN(7-3`SwH zCyHobP;UeDT;uO${u?Pvx6Q8i>u^r+?&UXm9+$T8dOY?LD5&-CA#G%Dcymy?J9Np* zwn33imF%u>9b^PnlB2>~WaQn~r5nR<1oe1yDQbAt!rSwX_}x~PFI=tsWcvO2mloCf z8{{<)jz^{(X=4W!Xp8>UV;D=nIIx_gykdMj2Tc%@PqCoryNim9n%^ian~6M< zS|F9opkrW>0tFDo5awr$cIc4T()P> zcbteA`;@=rNz0ziH0*r~=LURJfq}zSdJI%$aEb2e?GMT~A)Gd-I*b@R&Snx&l#pK#bIF--j(0AoP zfur~ACynwRIrI)fwim^q@F%YMNVQfTlEq)`-y3v;JPOjP)<<;A*%X=*p$}Q!C@+v4 z^SdGLU=00<8a%`}JG27@54!7U=t53fSQ!YWo?P}HaBFZzy)g~)xEVyeF$h2C4pI4B z2D=iI8GPGR-Ys4sZ&N6e^WYK2Xpj1K(eQRv1P4IAr$N_7`7Y`D1+dx1dZ$KtlK0i6 zmwD^o0QnbAlRVOd+xeK=Bs*-i_OAcfy3yU}`23(I{UX`oz~P$`6IW&z#e$xAI@Zl4 zGo8XL^>INpq?^-2TNbZZR4X=4t8mGtACmK+*Ko*7tEv?1T;c<(7VHR)77Xy~XgwLU zKYbWi#@u1hJ~MK5I^qOhYTfO%XJ2)c)Y0g-4wHXtVYSlHiGpJOWeCaOhHAhyj;-)) z6b?)+5OoH{^SbCOfvpT?ByAmhd45XAXyl$B$i`I44)@w*JnQK# z3xxuEF!Q%UwuEU~yes`+7bEPUv%Czq7=(3ig@D%@f=)1eqiaKtsI}@NKGhJ*z#TFs zhuWd+T9XWQE@z|$-j6kb@qy8TeY1ubzhqHRe0zZuiH)%mJ$?AijyYwaczc~vHzQSf z5#n$;>U57Q;x^%+>sJj1U%r7r`*eLsj1p zbiscImb>661LA%e5M;Hm|4ZCq4+}SP%pWXU@$XBG_?y4tTpOw27BqC=WJRlqr+u1Y zPg10gij1E25fFT|L?El?^^gs+9=h*q*Py60o_B$4ljP8=$*O=TN~8Og=FD3yOX7k4 zF`|*t~Ib<&3^0moe{PEJ#^YnuR(3;@NlLB0XgTrA8K0PY92Arv$r!0 zRPxsEx!X89*fTTpHns(J>1pauzcS)!kIVeo-nmU{P@7b^!?DiWWaqn{@2UIv2e3{yK1lE%$r~WF&2Wh z8v|1Xk_7J_54;T-0Xx-50g=pXb!(vaiOUo@>PK{=5J!d^kW?7WEr^JU^k~zfE|G>F5EsGDW(q-tV|_FuZGe3s@K|8I2L#7dE-?Ra6S==_gD}#C}CR=~Ucc)Mu@3xp48@MQxhCEeqTW+goxABenZSX~Y^;|1=ivVLGBD-&Pl?S+&>@;r2^Ge0%)29Jpu5jj`{V`36SXG zxK@mB+D2OwES=*vubigMaXZ*9GukI&qw$GNyZ)~v&Vh~3ZWH5^PO+&J*#;|^XRSX< z-W#?98`)wWirn8|G9(x|s`tAQ)EAoP2A(>&=HtkBa>EDrs3XVm$9g*_;8M;}-%tPB z>Hyb+D}tN2R@p-Z{=Z_y3QUL$Yvrt1(d~@sw;@hPjfSH&8`GqX$pnt+owz+zKBFz~ z`$pTrZ1DXSIs4l9%HSH0=)Jl|v6m_GF%?IuuOMhamrAA8BFt&nyCt3KXM-z8k3Y@J@t`$q6;)(B;G2;Hg>4dv^1*^kGv!SME1ZMV8ETqbc zu&^04lz0i=pDQ(}>Q}?VO-#H5`#YaHUH6C`#Hpt`sSsm4Pt@9o!&Kq21S|8%Rp=hgcr{&$E%zct;s&Ow%+$M8m2iJf89iK z9oVTkZUUPF6kA4-Vk$D0i3RiU?i#2Z%qE$T19@kSHvnj(GInY|LD|S`=$Xr)KXJ`+ z&+~|5%iK@+4Z80MdFs>?X4q|G9T~27= z9V!b&5ta#({TpEitcB2sM<(R(XW#Mmnr#GylSF-xz!rsh?JjNfJDzqwhhszK>LbgMwK!U`qlAV~6vjpj*5cpX^0t zp_tV_D98Je#T6?}vG5f9rDJ}&83~N7OBb!hjTL^0P4Tq5^}65R*4D`|;h!rCBpY2a zgTYfZh!FP^Z2i72#grX}gKW9(yky3>VdEu(+3@nph92ivoprBb$(?{WNtvQeSSdNk z*Q7Ce*-B9cR%>*IXw~m_`ECur2>TNqePQ}U%{aH|27_S3DSgpU+`N!UK5rkGY+P_+ z)&|s)r%t0)GdpZOBh&+qTvie&b|Xb%!KH=9&YK>EOV)q2GO$pIuEs>+!gOw|K#A!C z6ZFb5N)G6?tE)*rKpmq|NQwj^4l8Gs7- z3)73=FccRWCC1je>xG?)T<0n>?s|AD;9`YqG1@KT!^f;^xMf|0;>+7bl-Db_&U)ka$+XTJH+Oy$Z^>10_mQP-Ywb> zvVvOG0X8OlDB--5)q7?@9nbm&`1hU=n_D4UA!dB|c>ZZ_h{^e{-3wfeMyLF1-)bTo z9N6esP`b~d*O2$0C`RwtT}@iqIy#$9_rqRZo1@_Q&);x{D>^KVW9##r&_T(+b1Kut zh>mwO7yd}ry<(lc(8xO?AhMlelPR*9ip*Buo~Oas*#p;T(dX{H{4Gk&I#CY&(DxxI zC%_g4)VXZEgMHyla)ZkYhqCEK%WzS8gzkRT=}3^Cv1;mGkL6Hrs>+aTW-EY`q*d0= z#L7Sp{O+D@9SFGt&p9nnq=#vWYI9AYB4(z!|cEc7)5bp<^nSrkg7gaG?`- z;9dX=S_N2k+u&d8Q^UI-+8l_4SKV|8)V}ow4Y(T4*YG-AYuwwJB+}tpDZa>i;uPn- zDxeM4*Glo{@@j>BU70Z;Wq8SaHm!Cz%J|0i{#U)Oq?y+yTvjArGyeiKd|a5mT3IB$ zKsJT#@Y}RBm%kT+A8Y2H^Js~v7Ot9qE99)?rlg79ufV43dh!S;s@msws#_xP7>`io z;t|~&bZOC`v}0~QH*Rn%CTJW$7zTvlrxowTWBA-!_y^0;q zxy~u1!exhZ89y^PmeHIdu|D^h7G*56W$G1|t-<^K9+Le&L*3liN;K5DiLKoyNZ-8U z(tAvK2+~M*%+d5JuyXMfTjC2h=h1MF3j4dy&D`vUPpeRBr&U(RgFZ{lD>TZJ83U^T z^eQb#P@*^jgq@&9M5p|;N zboRN+YvZ-Zs*=iXWCf+#T5}0TtEZJ5g0-;- ziiisT=Se~(k!TJi{G#)pex{rqdfxAO-sgRm@0Y_*Xjek^a@};7?IxNh+PIuDc_u#LF>}L^O$;+i#^e3XqFqgGRI}Y!Su?`1wS(dr5 zMY6$VT^o$V62_?|ZemshCLyb4YqEJOqT6W4^9Vo3l-OgJoZ?$Q=kH4GI5aSt@;DcU zg$->U8qom-&2iNx554-T zq$KDL|07WmmS6+3eF3wL!M$eobI~d`DJ*^BDdN1f`6yxRK;T+FiW7_}dTN}1fZ4bd z4g6>y$#UVkmP!k&x0_PzphzJ!A}SW17Prz}@9wlrdbdvw=>pFBis;?+J!n0|E0e+=2j-9jp9?&_`ZS0S8h~@@l2d{@qlApp z920Jk`E!4Nv^7Ns7Z=urGYd8dJq~~EiUcV-mbvfozUx)&yH&b2Dvnw1-U~=72SSg7 z!G|J?cpF}Aq0^ZH<{YeJ7A_~9zRSsZ(ZY^dr%64%@Vuzg_p0Osz0gSpY4p&tv?7k$ zJ9)@==|r=AX@2eeyCjv{g4=~1TWAj&QZtoHDM02{OT}!L^!h)TpDFH_1ILIqpaIZe zs`XCvEbuK9A}deYgaJ4FC($!c_^Fp3L=#L3mCg^=j2I@4!h?Xc;H+?z9x@GFB?Dtdv|`mfj763|~tdjF$4 zI23{t=KbxKb&spdCb4XEt(qFm(pX&VK}bGUXb=syBEqhd=9{4}@yRxqVSaL#^S3xT zIO7&iI`Ect$NbYJS1xRFY;@5hgPmbU5~&yG)A5Rh{cby>7G?$|fy+`4`OYmq>*42p z_|luDtGrXE?&Ouo_e{GnyFsK^-{ozO>?eDH8SxQY<-2}r61d98**)^>zWBTcvOV&C zc{95@z@cE{0;cC#1V=*5HkjUiKjt%QYr@4dcHN)?v0;0VlqGhOX1nkLugeR=d72Vm zq%C~pQjPtj6|X$m+6#-IIN|4hRUxIK4q`f<&$|gu&o=L(D7;p$PK($m%?sKHB(Bxp zK^E3Kf*UisjCY;(+Yy)#y-v#&+q{!NwQ%XICiU8d7v@*|OU<|Tzj^FO9rMqstLJx+ zCiMoU!t=7^F(3cEbk@y(?I35>BZQ*8fO?+ov3-7IFgt&$^=&&#$Dfc6Abrjy5%!B| zbfC|d#&$3_6nW4a_Tj%a$>QJYj#@UWn}?+@9gL2?9zFn4EK4H#EuFZL{khM$YZPzz z8BZPkDccr)m_jJmU7^MwlFzrMBy(Bk)&|KK(&D1bSEH<)ErN;Vakx~~7_ zNYX-|R8~uCgDdz0ZoOW0()ticZa$%u9&lSZ0m<`q^yfeaouMoV#k|Gt>Bf9rvlMrl zl&DjsS*botpQQ(0DNwf1XC~_C)e|v%^peeMy6O+p-3C(;4}SdtiG69nq_qH3GNo8e zk(E#+XcVm6EAON+Gy4g<&S%7f`ixK7j#KtnBRd}B@Jr^rpZII*zFL=^!EE%k!^c5{a=9+-sC+A=^|ky`8QUfX7eZU<){s%)w04AUhqLuPXJ` z(OKd#oMHywu6?6Cc=g!*gTE<3YZXdh!m2TdiDOfTC(+@V+ zqOMG7)GePv;aH=^KB$b{d@~3}+feySetfmHT7(TzhF;Re@0wRGYLM4T(p9BV2EXEH zz50^q=1e@c#_tYauigMXiN*3$!hO+W_3G_@<%1S?@SQmMr|nlhO!&=p!{eX(|75l` zt>sJNgKV>yc+x4w28yIoF?YUxasIven+USe#Zs{K!D4rk}2LGY|| z*4a7d=_Y=eRI4iD4Y;*?;v;fAJ0K#9cOH>c^GoH}ZS84fMm;@h4L4-iDxUC5S8;*gmgT7Hi^7?M2V6iQopALEGq%pF;rh^d={> zIPOO?CS3l`EwBz-vq9<;~MguYeMR zr4h9uTICu)&;fw`0XLJpA1WT9geNSinv}hygF!-?Hu+UiQyA*pzzraO)9;2dI9Db0 zGkJ`YVtm z0iv8`S4RyNP7tvF;$_7b&s&Gzo({FRaGjiukla1b?kHoIyiYQ}E-FJ6OjAZFfYh^?iYXBv3c_q}E6gpX zRNyV2mBJ$E4KY@*Y=b7XtO#W8z)7S{zG?1&TRkL(O!LJVk9a9GM|VRtK$CL0=YU(I z3~OJGNGgGRs5W@Fa5t#z8E8k^3cFBZ=WOA~43TY=3=0QOS=v zAE-_I-#FAuy#Vok@Lzk|DM5`!MX|r^!sq^i9spd(2@W0vVKe>IEE*m_Vx@__?>s79e<)Nt{ z%t+H_46k8G4XcK$W)HZTDgv{~T0Ro&C%=O3Akw(C$+7Sl`-qLjp{+n8cqi%~Z+ZBV zfMwBHL+p$0b7ayVDYHCE7)-3d8U_O#I=a(C55X_zyPj57$zI!Lb(_<7Il*;Y!P%Nm zXP7F&(l4(7vyo8Q(7eBX<#mnZ zf}5s&9%-PRo&Fm+a2yAc;htn0I1#^I{?HmEE?Z7-0}!>rMN^8vZfKR=(`u$@a3?FG zPzWoQp5TxbiqJ#$6$yKBjE_78)5di8mxi{8NC!A{@1(xtPu(N`l*i@W$sXK9sj zOn$%}FTEPi?vdcCxu1V5dWYvpfs_ z#UxWoDZt3=pkj)6AI?r=Q=;O81#BMOAhl(3bwX3E2Gph~X<_MMyL$sE-d>F_n$=yBj2Ja95H$>wIe4F8Dlkm-7#N_qT8PTg#letZQOJ(kxDR3;4H=(HUZ# zIpTx^l4?3Z2)Z)#P0?1mX37|kw_QVRGzQKe-l4yNlck`3a;59%KQPZPzbyV$A8B+4 zeT|`MkZY9UGDVt!>K+(Cmi;7q!S3KDHSn3yMNxO94T3o4xN3LsS#>L$ODBor`K_~B z1j`i_UfJS4uf;-W`=7UI>KQsE`e1NIK%dvqDUWB|mG--}cpeq@yA{0(OVcU+Zi)P2 z_NKfX&i1< z$i0#sKK*WGp_^XO1z(QrckA@rq^j^tg^Sh)9-Wd60;_hw!v3KgH13Z0Idss5-{v-y zE*rd^t#;pLohP=Djztcw^T9B1;ugWa=)IE9W!HVnX0%C<@{aJXO0KbG!JkC7Nv}vA zNm^$aWIO5n06b&I5D))?IUF4kMmRiwq?f)V7`d|jO6Q9{L7C`)_W8AkH_D7eDVT(S zq#rh@W4@tK=9`o}f3nC8%#evLW`J6q0@4t8Y^~;iWc|NZ@ zO0Ql&X|E589p-+|5Cm=8YZT=gBiI^mb^{WXc!TcpjMJy?x?gJZl-ivdGp0^NPvw&?7Y@QCTL@Lfk5>#(IOV)(l1-{e4BdNSZpTn=6^Jx*Z2~az-&A%6#8*}OAu=3fN zL5ciy!9xO1q0^wJckrGr8fl(;$uZCzee|UYsPONWZ9Ns zlGTbMq|dt@h$*_+&C`I44G*V#$EiQ^?~oNRc{EBEY7VMa1AF!fl09`B4UaTFPOHQq zoDTZI^ojvDq*FQY+7VJ9xh5$T=Bkhk@!X_RRue1C6z>wR7GL+>OR&f?699EipgAtD zr0)l3ij914<#Sr-`$5kUU$YpoL1Do|&X>x|8DBK_iW&NKZNcRF|D;*Qq8zv!|^dA#dv zJWP!DZ5g&e=1i zn;jGLFm$->1JRf<7Jl)k&HUXecWbK3mjo6dJ33@C$)FS)DHw$qkO9JFd^;Eu?0@bQ zATvj*3Kg%QpB|2G_Q?0p(e3hNk1<1uz33S9N4^+7CvPmS`C+9sO_U9OC2SWk_6?jM zJ0g)-36)wjN#YA3#q_$cRud6pxuu?^vq!Mr|e1rV3@J)r@9 z$P|%JDRL-K6N<^6s$;Tz@A_$#d*ny_4g}=GY6HCpth7sG^Pq>haZ(4}FE>EP{1IL! zUE#TY>SrK1bp-BM!Z^K{#$Hei&K6I8IuZg;Kg%{KhNyrcDy9>vdaxifnP~pf0n9iU?$sQI99SCreg`?I+5~O|Qp**JQ61!7 zIJy@DZka)?EH(yIk52H;inL6MDw|qVqRVqtR}4;pL1C?bboHnF`M&$V9Y2X)luT=;$7z zxk%F3bzv=_i=?C1dpAH{LGvLg_BHV_=;)&W9w$3pBL~&Uz+&%Rul#a(*F3Y4Y5LKU zJW}q$M&`VQk@=8P9Hq!%Dy9kIU6nL;+M-%`6}ja3I2?~O%8HmmUJn^OqgfSUi0q_W zz{YgQ_Q`5PK<}XE!`jdczi%%Ztgqu_a>jq-R}BZp@qVqcl`(+IvzGCjdu8f z7;^Nf=a2LNr$KVvGKgglPR+9Ri=SRCx$fT&9kj#x$}J>zGRi~k^3L!!dBmqkS9EO% z6gdl$!cx2|1GL_K-no)`554-HDw!Y4PxR|0Nntl;8`%BI6ldrDMs9jWidXyYvyJNS zOq#Ic4fB+huzyg30N18RBCrFNW`ov6e7yak=Iwya~&+#^HYp#*} z2=n%P945VyV{M4tw0qM&oZ(e(Rlm_Z#f+%OiW4h{)`gv*Llzj>NhzShU>g;K6hJ^U zhIu~hgDZ`y4C`T_u@-CIv7%-5#C+h@Mz`pk&((=oEt4}@bIvCZxV#I+rjt-HgJnA( zDbpubk<}9oG?%fp^KCl_0(Sm3`!BO?ZiZQ!k7Iggy<;tH#w97>!ZAM^LTE^NYK9oF zp0dSEv!aq$#TdLR{8OR|#m#J?(D_aKLx++NO0rfb80hA+9D>WlY=gdlTR#g za)MF-gG?P2vzWh&cZ5l2I_a9=tpQ8A?~<*!>09ZxFhE0X2HYN)Y2BWeI`hi>^X zdBUs5rLD{*PrbT9wpCF>S5DKb>q0sSy2Y(*5zq8IBjQ9^ zkX+a&ex}R;2JnuomdIXe2XsmR%ZKQb{gv)G4pd#zQ%MpsxFU zy`MY3Bgi}t)xW;BoaDOjJXC8j5AC58#T3~|#gzJGi%Zq-uIHWSm+%^ue=zMgsn4p@ zScvHg22WsNWF@^vg6&uJu&B)vUiaOgc+787@BO~+U(UaAR-Gk0^;(5jHNhi?zj5_z z`{tijk9=+!7CBt-JnWS6#lc}8KjMR&@JwyGaDC~EUXg90%iSeUhICOHuL5#Iy6LcD ze&n+wvX|h2TvZyxMT`gAJvF;{SX!H_Dij{0aSeoxDA`^IrfHS9IC3yb28VIHx zV}Ei&%lKWtlBSNPw$g>y%Qk8&FN$&_P3*54r1)=UA)6t+<5@)EKTg7m%@MX6L^hw` zoD}*dG!*2^`hikdH2nN@HV%22tFUO=d%)sU_7RIo>+D zOlsuuZe=%0Z-wW0mq|0}3V#%PZ425warF#mfyi#Kd@&DMZ8!%jlNe7xKQPk{fRtw<2E_DEVB(Lhz!NWYD%$^BFm_lb!?mvy_Ox3 z8Z2!XYap1%@%eag(ttnbv9&SFxdB~SvKq(;noURot`1~o%dJOp@l zd$y~#c_uMVuYh}cQ0>Lg(|0=e$d5fdS!-0-c%A6aZB^&+G`LW#nypn<6niOBLdEou8s(X= zPoRVjx5{`W%w?uQrt>^SlKH(}Yojg;6Q}n>&w35NM%hW$Mzyj%WIa&MjsfTRRT4B(`y0nZ7G# zni)DhZ^@RDtuE}6R9nELgi-*RVi6S+H!*8!#ngn5@`*a>rjQ0f4zEGbA?cl%&$~(9 zHT>)kRUmcu_v`Fm@5#6Lv?Ea!P-k{WRZH{fVqeVRRSNo~HPQt4-JxC4TIDLAI-jNP zEuuQ+w5q{@HJZlZbR0k&hz@v^ZGa3@V?zD$e%b%JnL$(Z;MX6JSa$-dfmqmWW2hgTGp0>+~UNc}45WS^svSka!& z)P?jgcqE@!$>7yZ>fPVp@Xu$}9kWi8dO9~S$$Yf<9XtvgmEtbRtwzjb4e~f+LmA$#ResFx^3@!PI1-LY9O6*AasS#Pe+p@c3n@6*vJug`W~aTym-Y*vVy%HUJn(tpUKyTma~l> z7iSdH+nCSf*F+i0tC5-XXOVg}te<5%IvJYB7Be*r_B=kCYnW6gy&B=DsdSiOTj)~6p0jBLB&*fHp=?EuSVcRQy|f) zfPL0MZ-?7s>=o>~<%>&pu=wQB5f-C{UM+DPXer z$Yp$yGziLgSnYMmb1(1;A0wN{G4hGjM0tZe!l?lRZg?fm;b-1}h)WOAE0B9p8<6Eu z;aL`V1S-ljsR0K11}_f!Vl-&*joeSje0+i+)5QtWGZS0&9FTXhW2Ax|9%Ma55;>frGR9XmWnY6ilCEi+A&wP&#y^&Mx<$@W9ODe)ycv9VY*ollt1x8 zOgX~=t+Y+rC~Jdyg(j%F&4(UT>{E8iJU1#JnN8(0@G#<2>ZV`Bz362su<#8vh;Gbo zrx$J)US_f)szf_Qw`Gf`CweZ-fvy`3D*5KA$^j>Bbk)?=-Y1nn#*+ma0VZmHY*z1P z3&nQHsEz1Nk9zb+^jn{QlWj9HOe$#HXI}}+vu2reS;{yY^p^+$lc{Ky*3i8oqfs#W zL`X+q($su3q2I0FIc8wEA>j&+;hy1)W1PS--haWI|C(U-a6A6%(qG96chGqnI^%Dm z6o8g>CXN7bQk9^#tyLw2)CIQ*J{4f+oQ~cpKjUKn3ZuHHq!6dr;o)Btw;>q*8O}J$ zZ3tX;jhV1i`JdJ`(=KZ`Y*az_Kx{0Iy3W3TPu>Rc4$R#2(M@4-!YL$4P92BN}DbSQCL-)r2Dc83=Wcp?Mhc*6eukL#`gXO-?V~ zbr*@ePfq{T8WAr^LdOQMF@FPcu#qqwnqu)zT7Wzcigo~-dfqdfasF4Y7o=MAvAOKr zVS{_^BPfg745D!c`aG!$DHUDzZF*b(-6SfWf}Wc1w0`BYch0JpQh&;${#4;i zNgey$u5X=y#$PzIiTaZh&7D0%Z#!D&j6pX&lao!O-gvj}9qSB@jnu|5NfJMaX@`Y? zCYf)L^^ztaDAQDUVz9MM&<>gdT38`w(iuV^Im5*w5KMy@g=TwjkraanI(oYh$z2PD zPWxC#Hzqh*&KVa@W8%88q_`6v@h`gaPOn#E7$#^l>BoK=Trf3D$5agC3Bc?F>yzup zYKs$C$FE&@Yrk|bM{`H?JCf|e<X#&4ch}SuBZW$bE`ygUH%M9M_yq}yPGVX^*QXgBl zz<8Cj8EbnsuUf7AB zEc(YvxuA#;vbB~>y2u>&J1|*uOqs>sE;$&C3c8u}(ty)sXXK*PIhA20;#J;vysD-y zDkKI`lT61{O|7Sq5I*q>kSVUBpiFn1>lFjtCg$3C=lWd(!zd?ai;e z=srlHxEyraw8|U2B2WS?g~gMmnofycC2WO?_tK~urdfL6HLY@cutu+}Va|aEdyS`; zkHKuh5e{ZM;2a1HeX4CZ43~vFvHsCrYi3s)1gg7uy`)7@H2*qV0!n%(KtLX33Ve&9 zD*n)96f;0wXKYibexq%E%2$yNvz{(ykHgqiiq<$?L~CTj0i!bX-8o^K({Q-1NB8YT z^ZxrE%U4ZfZ%@AKw}Z5jJUTt-Z?VENVYgnPLBP|@YK?B>j| zayaZAac8@i))?Ys&c@w*yLg56|$1~s}i|cspf+t&g;Qr00WSNpB6nZxbBJDdJJa@d8#?bj^A?H^N$28x`dVy?gbUOd$g`NDiv zS5Qf4xBSA?h5*fGk54B7rB$9V5oVc8aj$egjWRZx^$Xx?ytc$Q&-ZFDXmJ61UcdaH za;HeE?DZ-G@zMRW2fkD$uJGLNrHA7e8JxflP0tL3H8jTDk`;LA(y*PPCS|MgzFdR4 zT_->+2bD}Qr+GkCDS&I6rOSl3Sj_>|j!3)9<2mER%93#2wG&I+rl_y|KB4UY%&7Wh z{SW4oJHycTtQkMFW-@NPqMsrUshCFSuEQo??A2YopiBx9qMkVt=+reGJ~wG4 zi4|6gaCL85y(ja#C21^1TJ->$E$|X#1>03=ym*oe zQaerRmhb4nA3m$zq}-q^k`xB!t8RGd)#pWvL%XLHMd>_yy>gS$N@u z7OFVIFV~uhz-8@+4Q_$5;4YY&QUjWlry^P+aK7r3td79;s!aMLCF-ol3sD#nA~)|l z4Gvbr$r&A1_d9Es6FSCEefQ^@MP?7Dbjm3?X?G_H7EZ%`N^zGWcc_?D@IP;d;|#CC z-pW#WI*_+F!P}1&E)#ahz(Y)ls-&^1Eo}no*5g0u3+O=C9-FGL^ZUsejIVvfTnpdg zvpu*JYT!}>@>B(ZcsKaATNPNwdW|iRfCm7&)!T!g+=#yDlf5$O5&yM5x55{}{~BeL zVaYym5xCvWu2!IzfX!O*Lag;R`5Cum#zMVZRiC#h!?;U+>a{{o4gNWUfArP}+nT@?H7 znW8z(YgK*(D>3}EaPeh6i!xxaqoa>{*uCr;p;6&%%f6WQM)(Wuyp6g!uJc8YAESEd*R7pU(R4FQrGRe5%(q*hBF5v7;qX_x+@BT;ge{cWxKmYf?B}*yA zVv59mF{oA>b`I9)FXJ|K{ofDHc3R5=K3)0Zx+zs4pdD%^w3Gs*5jG-)!S>Mmz$>@R zJ;QIs+zztbU8~Fw?V6>VY@j!Yfer`q>cC&j-wu?sWr3YzC6&MtG=QPe1Hzss*al6@ zD+edoDG#_^2pX{2WG#OPiF}3UAt=-bwi>$6Yb?J`_>boDuI+h8j=IM7yMCcuk1fCW zmBiod^r}{`s3KcjI9Pnp!bs-V{z2Z$O=z{ce8Yf*G&PGO6llL(_oe9V9c*FiEd&5gdY} z+n|c1`Sqhy?x?atw@247T?EPLZ_dmLZDmjR?3Z+qez#&aOVKJvqkoi~nsHfrN3m&9Cqx2grajO0?)zyEr?D%hD%!cKD*q`O3Ww`v33u7VH zB2?5xDL$vjXTV-fUzn#==Be_0o5BY~+r0ORHOL7Hg6ZLXUYA4{nR}87Pvq)IfjWNZ zR8C_dIumt`P4;LOY1#oJxK3*DsHYQo2|@cIa@0Kyif)iDHjB|sf8>SnoB_90-gR(i z^*pEq4%j}=xN<(FdIS`sX>8ieRo+_VqV%9UbeynIY!Vj3cp>IJ%ayHch5s&cQ>N+l zDxY(abOBv_CS4}RJn4<;18$fpI;qU^F#TtsO9&9HVN#+R=)&oxQVVC$5i{HuQ{AXH zp3HH!XSj%mint$}L8rYXd`!;1V4&-^n0nhNMGHl)P%-V$kX9R<^ZM1VG|VrLyfIx< z%rpam;VmB{s&GuRB37`Ncr1V}#|nR_ERS3!+#ZRr_=NO&p@k}sJR@plyL_AcG+mOk z2!v8>Xdy%)6Z|g2Lg0}X&>>^yv(X0;gN)#D%qGZ~uK`wLxTi9BtHN|ht4viHf>E`- zMS!g=<&k+1%((%|?Q#5WUM8KWxDrsN(qKi;jp;Sh9>K#e%(DR2lc6+j{XfQ;)nD^u z*EY@sUp^aQUElTe`rL($n~f^2?ZT~!4yI)GiV2wK-aPHH1Z~Z_370%!I5AJ1B<^4^ z--+q+)POs%c*Kiig-gA1B*k##l4os*GsRB1tsjT6IZsB{!CSV$vi&R5l3(;J6_Wb& z&{(>6b4DEVfD}b-nphmv$2&nfNuyG)E?_oIt0p-l)#=R;Plk=7*?4lJlecUmVK_Jc zgCFkO`utFz3!4&~p+2-1rbJX}l&MMr!~Reuf%D8xNOQIKa|Dgq zf;bs5@OayxdGL*nl|g2orvCM{!g!+mDqTO^4x$L`Ls#o_1mQU$- z!|PI6EC$>xz3iLJ+b(r*$YErF8fIp4a7=m5lBD0;nf*=#Inj+wd9aBh8$VlPnl4^g zAi*jGO@n-05bkJ>%kgARaQv(FK&6r3cL2`cp*2{#X5vyPlVMYlrbVzWtV}lGmiy!! zlFXLL;uZQYF9xRK2HB0-JIKL6O*2I4%4Fw7?U129B+#Tqw9y^10})!~VbH&-hWZ#y zWmp5m@z6&#{t2et0cETYNN~^^^hNr3T17Iy!hfg8$X|+ojVT0pH>$$FMR3nklOKV< z0r@PG7>EWsZpelk#*rVhZ3BjxI9xYJ{Ikmw{AQaCRwnziLbBI|L(m^vn4mgJ0XC_M zioxnly&9J=3mXInqW4T$xQS^4Fjb?$S@2#~%zH=Y=hNA$n=|!l6yCfls!-_Fw?Z=_ zKAc$eDz24MCf+BD89cXAo-q;RIGI-E-KY$Y$!hGylANiczE07D@j5~zZ)Jx zF~&^zJ}_NA3OnUofRB7Ksr947g??wCne<;*xF7hfND37in*YOn~2ET z@M=82J4&lu61X}-uhs<@i~GRRWds~yVqr?VF=eNyL$;KE8gO`3lET9jsc!nUu)Vx( zG~Tf$pk~S_Ao83+93_ID^YoFQwvC{cJ14&SirMM}{3hQ!$tdPn1EJSGZqthHhmv@xrpnE78)m!^*pxP35ICnd03{ zCFCR8B8^tB)c4rDT-7ZfQ~nBp?8N%KkwVzni*jxdUJ?BM>)~dkWC!G~Cp)>V&t2D% zCm&hhs*+Mb!Nz_nW*r+RG^L%4Q>L*3u#GMQh0$DcfW8}bf~3zmLE=L-)zWrBfuzF! zvHuvYp4T=Ra=`FsORVb=h5ZFn#cx$?RSIIUh0GW6Mbw*IDbotzCqD6h) z$^1ewo*(UWJ4(3OUe}G%kZx|CyI5!j(7UA*{zMYF0myYHTBsTsYM3@ticRRQ$23c? z5#5wU$monkI>>f~F+nKTfmXSXK!OWJ576hTVGW}mc6&#MGd%2l@F=hQeaV;C$;|LL z{_XEvCToYwX}a(-XtxCjwo-~rifpE0$`)KP3PE?xYY=qt+vPf5)9kj$oxE1210)nr z3QFWnv)4mMRFY!J>_S)!#fsK@ofe!P&AQm$k>CskqiwMF=_9|+_QZ54^52W2%^=yJ zlV_2AF6^!}SpecFr8rEHYAWWcZzjE)F-`}#SxIAgxu!I#5*B6G0|l;0*Fslzie(Ab_;o<1`&;MfMuaErFWueLS()(oWAchpTH zpSm!RmRpqQ^ihiY6uCPnl!xoJc;V404fNvaJEnkZ(ddIqwdCoj@_2LRB~gnYjeP{G zxK9}D`O*fIOkb@y%tNh5&3V7vc}3Lw?}1_(1ZW!MrNCnXLX%APR2}3OG>Jagg}s+_ zv+<;lcU-k(u2#7^qFCN0#r-m$b3}vj@IvuEzteun(~Wus+l5smd3p)6-mfYoR=9nN z>Cq{#U1Ul{h0r4L!FxHp`k*FBX;d0})-Pq!?qKs@(ZOJcdMn$>wz7};eLRe6r%hN+ z;XfOtx59f#`J9&|r2RGJnk}!J!S{%l{T8WoVennFuwthu#R-ZWqhdD9SV`)mDqdL} zx`EjQPR;%3TEShgVa0T(UD@jd(8XTH9zR#+l3* zm`V+35j4uQ%5y)w|7ISI1m**7yW~3}`{h_6h}u9S1I7p@A+A^&!LJz6EnzncK9QKs zjr{LlX(DS}*xW!3{*V|?CZ*U+kxf(#@=3>s#)s~koyt7+yE(IqZUI}NS62&6M>}9@ z>kZc`>*sdQZK2o9?eiL&VdpyL;O&fP47uLK&3-(3<@<3SW`xv#srqkZi3=lSs|7+* zDMboJlBk${(bs*!yM^S^t(ylkCxTtsd}gWzJw*kEDE7}*PpA-9alCO268@P{*r zb(0mBX~l&X>o!`l4tit+R4AY!OZAY>@!ko#ez=IzO~?L+7C~%ioBuI+zuR#+{=1c} znu-&06F zxAcJvPew;9AhwTEfVJ5TfkMH0p26>~bnquuh^Da!ne~1Vn52cw63Bu%q{PYYJk zX%UUeQhF^EtwX~qb)BPc8A)kr)#__W84UKlwBW{5V&8lL&FKfC2-22Rm~ zUw=Si-3hR{3|VC)Q;O9TSxLnpr;EAZ0V}ypY+GY1g}a)&rb$j?*XWx>1Gl zi+?HkgBcQR9xEiDbF(Eb?5M1^h!qb|iboWAK*h8&^}xng#><#g5xHmDRwxnfqjTrg zsCK+%%2FqZGwD2E?10)XEb&ExuJ_+NJEt0IH1*8cIs5&AsX&klWS)Dbot@JyzZY(z z9cU5sPArw5opT;C*{#q3n#O7a&NJzPZuZcuRu;Q1S4U^e#jd_WaVv}4C&-S-!EO9) zc;(>!q7UA~)H#U0!(8~tXGi1>ww8Z9df%J}-^_pW*w?O!Zm`|F-MsU``5r4@SuzP* z0dpc&N9U;;AQ^8mbf*7+%S_|MBNC`ceZ^%1#{KDc)+)^iRG)bJ5Xp9jdfcH#shCoL zrr&ldW|jNaH;(@Mv0wgisd^=wDc&znbZ=0tVGXQ~_xpSDLMoqLO8(L>UmW`Pe))!< zXaAf24Z~OVz4`miKmF)upU*F%x_pmh1Wn-~6*ry`5w@n=LgQH%l|cn)S0k^=hHl zMXv)kGadc0_vsLgc5X>1_S%<*9*59Jm+S~%2Z`mu=RT`R-p*0A3qQ}EV}?Cz=dZVU z87KU>t{a*4C#B~UGkzW`POKo>m&QRpWbqm7q!b|Du?_k33*ZB2lOiKM77cHwkvPAN zZVQ6u2`tXMNRR_p=hY}Eol@b6928n5x}XDYCl_En4pLRN zZGv|cCxZWT{yb}Omlvvu$%cf>GQS$SY4*E|=dSnKBl&}A&r-sr70z!7zBzM`eC$MA zMu37*jII5*+O~eLpZ{{Ybt=h54(804*Uc?eoBrLTt`yxTH`M=q=@%Q{*zxt#Uy;!ea6R=0rMQ13JyfT9y$PKK67dhs) zF@hnmo^b4cuNDG*d)R#l(g_ z{Q4D%j=^XhMBpJVU&vd+Xn9zuzeF|=p3JWdJ2&yr^j@z9K?A>n4~1^h8=wTJn|_kC zN%zS6f$gJOP~uxBPM02+@0y#fI>OhhQzjmdE@0MAUBTAVeNgw79+V4DJ`k;AdSs0946vT>fo4*?3+JG+EResJQmmm!B6M$0ULpe$KCFW> zyDF&xT{P0+JIDEW20-ja>ocC=j8mM@I^Jht-w|t)8XLSX$K{pu1HU6+O|G#Q=3RtR zq!!2r7O;8rO?j#Oxc?S^P)(Xs5}L+aspy$`XtJGO{b=8T-Q)0uEg0=@wC{dS-H}5U{Oc*jI*KGyF(yKyYbp&gIVO`O5m<_i+>UL~e6YzQBVvr8 zY6cDm#M%cn2VKPpvE#39^!Bx$PM%)>xbA|V*qCB4@KfP=*S{24)RRD%O@m1tl!6&s zkl>iw2^0qnzx|J~4HQN9_d5RD>|=beX8CKR!<{U($j|jsihC640=iYRX6!+rnbxZf z;UJP1SrMK3N`m4H@bG4eD@C72JqXvU??E?dCVijuDozQjLiPi#WvS0mC_G7vz`esN z-JCSuX>m2FCZ~lhqB^E1DxcoQq(z)nRLl2EQUZ7TVhd3TbD3$AZ;0+=(;-b@dU9O| z{?;mDeZOA4i`OO3r%QwP1>?Ens?WlUAl;VFtL5uFi+x)Jt&yh#4+9})R!F+EbyjTX z8rf!_gDRB$t0MjUV-OWQO%JpUl@RyxttY|#S}5yCF#rxpB}ah$U-n7-_2`^DDXYya~AXpYNs7irUn@34<}xpVmh)U z;*JFFg!_7Cc1=UkFCDFOpj&L;7};$`95#HM#>jO&-|wD(`MYL)j#0QE2K z$YNRqhk3@U+CUV!|2yJG1Kvwm-rb z%7Z#ktU7RD~o-#$`(PTC=MDVkys`p;u_FHIOAL3jC+fZz+rREQ{O`TZ!WC0 z4#K=7Q;d!9Ob?S9fMk3|ZmJybV&=MUv-H_26i>Z12OY2z6b|}r_CMy=Usio}s(ID3 zQ3?q17eEb&4=THAmHqPl9_`_cvZG`duOjkNScNC@axGev0oj#q zHi_A)G62JhE^@B~#MvMOqn~od|B?8RPlYECXF>%t6oz<|FyIfi2-;~Ue1LYo>fye+ zA;;{0h4x;_Y2I<&Sm=yo(ZF0YVgeVh+D`VnlXDh`IYue!C{jzsl!y-nU5`-J@1cuq8bI4*^41HkZvjqE|V6kvX};s zZe9cLx_mSUa}1P*plax*y|@|XI!;h>-Bt7I`HSDNX6xlLBf0K>nJv=cTpG*Y%H%86 zS3^7HnbU5}t_snsxB4gY>gmlMIlOp9t*T2^;hCkl&Ta=+C&zoAr(TUg5O4V;$L`(S`fuN}wmL4GTC`yftEE>YE7#mc*0?* z1*f@43WXSv!FtO=aXXxU4^`y~m|k)NDt>g+@qrIVIPH!G978SNGqy*B8cybL{HtV~ z+S(6(I#%bx%Oe}!ZY#S~t?@PRJC%bbJgkLejifzwW=ZE?y84;&PND4Oc_iwlT82PUl(Vezo zF}{=c2iENDZb%9FgsgXm1kli=P(G!|p`bn>rf+U4uTb0)fLeKnLbgO>q5L%!hPbs# ztPm)V*85!w(;%ldk`iJEb$$fUVpPKfa48HGbN(SGcH8fMm}s2q8s%30v|#H=#fF{U z=ZDFUOQHkO&^8Q_C?spE6kQ+ToIE1H58hm^GAc8W@)j!!etE%ZLWUTKRr77svl4YI3g?2Sp8lYv|o5W}>E z7@T$LV7ZxxA#x;G!n;sx@in~DwmjarcF=*+_KGKk0ec?(t7=0 zF3v|noGf2MP*><0_AU*PStfVhW65rB+-p)_`PP;BXVquul<3i6z>a|$DNO8m^r$Z3 z1QXYdLq&?}CrmKIq~pIX{gte6;X*f^1%5YC3YaU`Q86aExa+c3_Jj92ft*3Ft_v|8 ztE3I00m%|Rp1Hlis)TY>!~g7?IfM4uk8{{rZlGoUrTtYqQ$K})yH(}XWns3sMm2Q1 zNv(hL@SAHFuKV$W|2Xt>y^(<{lRnRE3TXf~oE}l`v`SI8e2=_Wf^A;45bCV(#I_KF zlk^)n56sikjKc(k|amc z%0LX?6KmDXOh3&+;i+CNm7_Q7a+Pziz?_kORf$NstUxHHT@H zTGhe8YAD6Y2g%TMX$^A&WCk2M#X8^v_~MhW4-@uXb$O$14%Dk&ccRqF_>~9GjHR_-mC3e#<>I+q^F_w;@U2EX6-Vt6a2m;?<~Fp+RI6 zTR%7H7Cdt?*$OI!m0_s?M?hJkH>w_j#{*y5Lsv}Q4d*p3%MQlLnvVa=ohzaLVK!R- zad7k3$OUdj%XQZPq}Re|eM%|XDbh;CoK*I)*c)Jw?e;;L6JXyxL9(aD3va!${8b1A z7m5>oPX4T2P{Iy`C-I9wV;i(g8s035NaokiH(rV>N37+=8y0Dqy}9sC~ruh08{3Tf#ySrL_CRRl}kt7)k0o#e?}s!xbUQ^!;x4G_Iia|~>-Wv}(gU?& z@%(~7(?l05?DGQFZ{u;i4lQOuU;)gWCU*M_oKuFKEf&gvu&=foL<)cHM42@ z-P!c-k;5*$I=g0JY(AzG4HP*E{1U>=!RYQ+L&_Q_MC4oCIe#%#GQUZk@=ZK)Lw(S= ze^%WOEi;R!7mIg;=yL0L6gh|@ zaA|C#4C2{fajvn;gxjI|6#2^%rYG{+fO{{USuRZGmx@cHS_GRXmduU~Erhx4@rNb?xyD{ zkZ?epy>)Sut$Ex z&mih|>-CCb&O6OV9vNPSm>14Suz>6{Ea)^@)Pr*$H!PD=>P+4-=EHP)nVsTM}HY7&%kx-YS;@i z{(YJm8iQgu)fUkBLUYF0e-gExcXWzYc}Y|;uQV7Z=0Y)sEe!M_K_%Te`2<<(ZCa@! zOG}Gj$HY>V>8Po%B^kOW5Azyi?NnWej=n01=WldISL&!gMB_I(dXQK@ffHz4H=2L3 z*S-5imvqb&XF@o(SqjnKDVke8rYNqFa{6=8D)t09;}3CYlrnCijW#ijE%v?S>3q94 zLg-o}&Efl;pgVqiMoN}-Wfm6?#dQs)E%Baed7dzd`IuJ%^doBm9{2$>VxiFaaCv&1 z?E=u#w>kaDpW=4?F2X$1$p5U_LUy@u*yN9gs|I!;vvWjPt-v*dAtzoSVzsLX%{Aj_@Q=hjuJJTjLNvM?w?Pv zuF2wJR9tsLwN+a7kl8Nm`^J`UUY`GW{)cZM7g{x)OhNKfsbg}d>DAe9UY?)Eu8(@m z$GygX?wPq!nio_vPI;?&%J%5pnC3uM|r`(Yn>j}I-o9_3%8-{*&UK+|0=$;kS=k$MLzC6BqVTY()Jt>5r79QNRcVHOQv3_jFCCN1D-30kZ?r!E@^UbdwEq z?sNzUX;oN#fOYSM;sZgbD2(Pv2PO))oeLUHFtK!9oC=doKMkEKHG8z8JEGr_WEb{m zLCtSSA~1_m0Bu-0CUDy5Y!D)dorV4>CU5i#O}y~HYXfdLw8%x5t1<@#aS}p$cV~~c zanMFK=uRDP)8EBqiWuiPyZxfj43T$BC;W*da?1$1u*+I#0jSNCViW$cF$=4pVBm;f zO~69*RC?Hq2rRUNgpPc1P&L#O)G$XR$B0haLArU_Q=Mx#zCZ(aFnw{WQ#W#cz4lkX zHN)fe9d(n)r`+J-!V}GM3#RElN^zeecd3{SOq}q6JdZ}QBLlt4yH0hBuj!JdOHI7+ z`Lw2-ErP1atcW&a1C~~a3uH|y4@h4ldcAJ(*M({DUrWQv#clp+EV{{BrRj(Uv(ZZe zbLX{qH}Zh?5N}BmNOedecG~iol0lr6+;sHf^sF*5pMFKh*{N)U4>71Q&!!vOLwD5`Z=#)Sm{RtG5 z8z7vs#=Ftm6tscD&O}cG-Al0A*QCyk+Qb8HDA$(7PxRCRDRh?jI&+@Zw15YN{ zPgNv6`+5uAC2LjU0j;u2-a#O`jM1Dt8Xtwxos_5sx>C>y<)NNBx|^30RS>vDcGL@= z)eREIt;{E^0jdG7L&&IV>ZLI3FVoRIWVz>n+b5*g3!hRWJv#^c3tQ;BvkD}dGVvB) zgmL4f4j?g+;Zs}|ezEV!SN@k7e&Zkicp7PYY2dfm0)F=>MHfXr$LJv@%uLXNGIwOp z%cQsXnoHaZg{c8hI@}_-?7Iqhx@v<9r(E{^;^H>4+#?G(t5-4_>`QHBkI&pgub!9> z;?`Q_r(S30fDtv&WpfkxYbHS{4v1B2st9UTSNNl4g`(j!_MFd$v(wm4=p5422Agc4 zHehvx@nR(YGSc`x4qVT#4SwWfti*4YuJT3!PN+K0Q*H4*0wRP5l{-afuMYyjP9Ckn zU3@Oq>g|m$#Xm#e2E}#b zL}rjJi!k*vLs%d!^aksQrgZLJnB`*5#@?tsG)`fo@e9BD=5MThD_z z9A797BpYEx6erx`i}Y$R(cp}R8@cq-0Bp9ZntccSl`&f9nLxt@#zSG|>}}gH?&Ej= zTcFt&`u(mWYe(vVqf4Fac{2RFoR{wL| z4>RA)|JLpKx92x~v-zKz-hQ~C^xGZt-z!*9@ngJr|F@gps`y6B4{{f5rCPpq?e)#y z%=udJ_iGm3e5+&rrUmzZnf;ZMKi)V0Jx9Zb3qownI;O)8hKy~54AY1(zV+@~Gp)(9 zI7O#iwh`JE2NEdNJthz%`dd zAQ1(=&M6318{i*@V`&HkMhB}Q4|4FPL7aDJG}SmR9A>h?s)XD=*w45KW7gP~o5XLY zM_ooSf?=@_5+iu}h{T&mzMQfqf%xF{=#!AO|1iG`Eb~xqgy|b;-EP;=@KQ z`>u%~l^swr9l{h?(^$qu~$k*5cUK5=k&@TV~>B0w5 z#A%wNJjean-`9+brvKcxh-`8vAR{~E^%qbI@DOvTm@Hu6fmCdC1tik?yfjFrf&{Nw zAdOkAs`JU13k!z>(=z66B-muESrbqLEIl`8YLy@!5e1?60Q^*)TJJWJk8dnKL{3BE zy5WLz-~ZKQFEdukzxu5+WEHoZv+L#-KtOCL>`i( z!1(gYdR}?-R@M4R<#Y1rH31E>F_~TP8|Mo0LA&7WpBGZo|L03)bTohCW%g}CTx7$wJ>})SPix(6fxhbfiCZIq75fxBG zxepibf`iJ4S0th`I0#-CRQSIyiAo~T97y;_yW`LC9Bp(jFaZx(xvU)N8KabZyQeo@2t?CvaGQI*B`tud0 zUA#xS0Z2J|X=H3blZmtps3w*kR4&}&gSQ$VhPVP_q~&|g_PDUa$@)=V1+y<$8W}5Z zCz3nm(nUUZ2~sz0BzHx1B$w_Z8((Vzio;Ua*T*m9r-nBP$M;~l{Pr9Nfy=GObuPC- zaN5Q17YlzkM`LWiZCz(!cWU9=J%x;Q>D)Q}+>Vx99c44;4xN<3mYLek<6D$!j z?mEl%!ICAF*~T|0&bw<;BxUo9ARL=HElZ-o1x*KovgF&9#bF=d1(%sTuqT)~ws(l0 zBbc9IM8<{Pr!M9M=1C>r*t*k_N!n?Hu~yVgV_nNsI~U|bfl&f)g=;F4M$QOs%gU&< znJp2!VL61&2S#Bwog^c?RoxD>zF4%B!Nablc43WRjbMEs=v>mkVJ90DBnCLfB-#Nn zPmQ)cc5p%r^{08?)5=X&=I~#A`vO_(N+3-?Vo|h}Vs#YROhxN4kg9`D4@{V}$@cp8 zuv>u8wo3qo*_b@l(+zG0*+*^Au#CD^sZ5;|)iQd7jt^dqc`WSlz@KpsPGjM`tD$pW zlqgvfNIdS}JxeyYj#g&iv{NbBV}@uw#b!|?6MSHZqNOrrQFTyji#3u^(};achT*zv z<6H#{&Rt&T#oeGKHcS&;s5t5Mm2Z+@D=*tegF5I*uh@GkdV;4h!`d`GJ-L&8x2(hZ z9`A$P$tI*MUm03YhFqK%70}o=!p>7%!qN35`SRS6>`x*a%8&JDo^|E7Tkmn4F zEu;UOTR7yf_0b>2C`?jbK3F}T9FFr^MbH`?iJ!F;n@Ev0RP?97^^Mi;*!QY2(#2Oy zyA^yTN;BawvIRnDD;$E8XWaCUKPsrQ?3;F4Kefngm>;V-HOr0NIZH*YN=+9Cr^kto z5TId^uJ;EosbU%ig0)i{LNvHPk6rJ-9voNArSS98PuZkbY)mje;~75lkgWqh+B)W$ zHy-CfZnncC`Q5v&kCwoO4>e?o6EA_mNggqqBvWhxMLwaTu^M5!Y`YAXuNvH#c=iO- znAwJjaSrepF6V@a$-nvWCD#|d-Y70)RTTb(s2y{UoSdJ`D-k1EX{&m@-&)e7h+`&B zLTh-yY=ef;Sd3wW$9xGVXi&dgH?;08leyS2gLjKmb4%Jdak|=C+I_O?afm#8Kv!#XH z_FOJs7gjIDj?{cnCw+3*WoedQ@>wFvV}R9Pc{&UuD>>mfpOx^g@Df7G=l0SlrHtQt zGiZeeRwQ)JN#LOtW}9rO{4>unIXPSaINF?L%f7LokY)-Mmb45{cj7p4w$BRr!WptV zs%74XXGk^qt#P|moeF9)OQ4hvlYbp_%e*bY*Jqa~lE{QGXyf2;2f0T-tAlIOY3$4aw=&QI5-HR9##JyVO0T8}LwUG}#D4pLF{y{Na zJY1NhqqmE0D^APcROlnULf?aW&PS}#`I3B*{ey(u;bg^%a7WBDuh_e#YW|BJBLU70 z2m$AdK<>DP*GZ>Cw0=x>|CtZUhJ&_@u?@Ge`zh8<#wan8$t(6fh!B_@mzimQbC9fX z;&l#$PDiYB)>G^{iX>6dSf#yG6eC(L*Lp$}6??*Kcx(9il0JF7w?j*zMjbp`fMI&w zN4?%rJ2?Sk(*FJ0%a$c;T;>-ic1)~P-gPj*tKwTM&Iv!^yNcPw#IrG>)pV1vfW9Ve z_C+QPEH&waBFHLPCV#E_sriF0iN5&%rBhA;qjC{s0dEF5(iU+**3T61)xQBb>12Ap1|9YUx0o`+Ke7x z3cE#LlnzK8Ip)K|Hr9c6;FueKE>0HV8~IACW32o@>sV%{3<|7TSjn zQPCi1NY|5&xfSl^A!igFb9F%t{tZ4IbMMV9{h5cQK2 zlTMLzSF+D+;PNRJB;~U}L0+^PDmkjASFmN&4zEuBZQ0X%Fc6jswCQJfU7=0FtpVNQ zR<)jYcXkKwyn6=|>$<{K=ScfLzA+iodDkD0 zzbycIeDVZt{BE^jmSw>!7kAi+t4pku$6}9#u?SYvzO=VKQe#56bs8fBNGzj6m0bLl1viVqsGLbOg zf`W-GL$i*8(E{}QAO7Qep(cm)+IPB&NFKM!0w<2cHk!GwWfZ%YBE?j+j_wrhr!lkF zqHYqF@)HA=^Kv~K+;TwG6(*QP&9F*ukzbwH&9tbqApX`O$9BO}YRLZu$NJr$y&oo* z`*Z>iS=3REDxO_AwZi?D;Cxu8VymnNgwc?p+MY<{gtR9U%;;Osupe9x{nKloZhqB- z9T$nJlw>>cV&+pbH0+^RD1O{YMPnal#q@IVazP&Ne+Kv`={zZxl>jX*h~4dGN@jIn z2cn`o_;~PFvl>KYR0i9jST^(cqCuBqrb|x(J9?h#WN;aP1r@qmA|A@qfz~$3tJ~fU z#r8u2*Rg)~`J8MH^{)vr2WOjb^I?U$h7@qKIZo`_9yUW+IR()`vKJ(!nHW*3yIzLf zLk3}|0&5Bmk~C7`t_?`@{Zf24GDf6HBDIoD3{u%I73E1`Exqg2cI6GFCYP7M>y66e zJ&+xTrh{x%D|B-VFooh|g7PA}dFbozkAU-A7bo(KN;P(g#@G*RBP`Lgab(lc4)#(IP{S*a{WR4GmpAhDzUYf5Z-^tM34rW@J$r0;I})wbs^eVo{CSXug@&wC;8 zrmyCPPoDyXGqg$^eN8ca+^Zt2ov&45rT$05|Jl>yusS@p0AR%1KD=r5W_CPaeFFMh zeb#H1ywY6aC{A2DYlVf{0MS{9%%a}7nZGNVS?-Sno=^3hCQCdhTEMi*Km!wG4bRWX{IRQERblHHYhwfCP@{92d zt+HNrDg3mb0T$vlf^2#x|AMr~r;m?`$E2A>5T%I;Jwx(%@ld{UN72V*D3^)~#fMeL zrT8t-DI&QP-biOCdlZ8%r@Xtx!0<$R6uO{;!F9eRQCouXPPHM}qSP(~Rt<$#iFdg@ zqlR}v)g(^{+Y{JGuXK%b#p(?xvVr}X(=UZTgRo|XRgC-ukU@{eYnl6gCBnS2TSoyr_Bpkk<-*qEnKWso{0edSZvtXLsG!)g(*fn ztn{^%+=ZDYfwzs1;&6BA0X_)KRjaz`qvC3rVH%_hwaR;>SlkZ(*3Q6wudSYjX*#-2 zprdo>6=XQkP%Q3?ygoZQ>J#A5*GX>p)iA@nv^hgwnOB$$7b(CikpzFM3`cu zQMM?7SK+P;Xav^N%`$^=xry=XkL*KOr@|^q@Z$MB=55^Nz4B5d-{(f^GaJVFSO8zdg!c#*M4(dhTrWJfmde zaZd1|rsS2REi_GREk9bGP0B{2?r~zT>5SQAcZgyeDDo*49V04;9N_N?UNJY7X%`xV zW%IAkE?9`Y0q~hd)(t$~E>*p}-3F zdw$)shXT>JL7GLa67P!l%mKnHyeD3B2IwsZm`;$y%wRV`bUI~TuIEtTc(7)7e2oO4 zjaVA(H0$_UQ}lPqmih6QWFcoIn{K4zKf7?4kpD6%oez9K@$A`!r~OWO_qg>b;=St~ z-*+%#kT_vx#C49|oVDC@wI$`c6=w6KnOQT_WoyGUv4ZlDtNfS+28g;{Q8dZRpp4e> zu^x*jVfzEMxX{5Lt-pk$gnpbiV*hWA7YfGHAML!q8L`qI-7H_jFBIz}m10OW=1Ycx zvsIf!;B*^kj2agOEenFOymMs0vvW=cZ@^<{PO(^13>7<&*~Byk%5!%x-4Pn(U}#bP zDQ1D@GOXIce->cM)2{D_?bshVvsl}<-wW~9*XO-q*~rFansH*#SfP@_^me~TbJz;0 zK{2fMZHS0xd)#i(-H?0GB=hxCwE;*(ht=dd$si=gH!;04v;lV^Qut-Wc$^R$Ms4gc zvEk<9IOUBy>*q^No~`14-)bR=uAnYDvNu^rv70H91~e&5VMIaXrhsx`Uf`bKjWdg3 zk$6>I9<*`hpi8o(JZN9&pv(T(T0M$`_XanRO~Ge^9sEp9zxEg&?KmEo?U!Xm4T6PiGN=tk) z*oVwg&oWh>s&{US(k`VewgH4AXl=Wx9nRte+Q~P)zF%m`z0YNmapHux6$t^iAF&YBtUqZE6HA`MhD2D4M;@W_c_K7I)KG^b%1&Ip$gA(@lTy0mi>TGR$+0SGEeTYy$=stuoW6i-AkF zg&UYgQUGc$T2J_B%gfGr)U2%%xa_xKPT(|ZPP{JojQnGS2~J|SMcc@(7c5A7+zk8G z6bsF3`>AMbJ;4ee%sFBcz&V0CyX_>;=X2)ZD@i1qjwQXoo4b@;k~Ya3A!(f+bc;sf zkAVfaH;H@>>O-9o*CNuUAEpQBGU^Myepb5>udVUwVDQ|m;p_2Ct+Gje!akn6Bk*nm zoS%D&b-eGp`MVq6HbLn6+?p@RAt&~AJI&zoImMo)$Y+ld*mtB~0Gs77T`uzT@8+$Z z*GtM(kV{12%T{HwXJtStv(kOH&w2MDmu7iY$Xy_mOP!v_|J3)q`yT(Tz#MZGIAk)W zXwrFmef#*C)3qwCYQN$T>6H$-v;fyfH=P=8yr7>T$q*hLazO=v7A59GD}%SP74FS) z9o@<%gj5FLS0S0EQH_Q}5XKQWdX@=&E`UrYSjW+5;l(9?nAZ5Bx8v(+{Zt6TL6zo= z0pMrXk0iOU#qcx+ZrwRxPa-oY^kIGLNt4?mu}csvbFP8_k( znqer3V&f^Ynu=cPhVn#udKGgztdIbKOlayv&s4^Nw*^3sfylG(`kMDIfBT|WZC1@_ zjeI{9h@*-^waQXut2#jv>${8J!5mXogW1q3&n`@vdOmF8A~c+%a?c(kJ09g^FR0Zg zzt}v_G>hn%f6OO)oOl)i|6)Y+zn)@iDN;p6?~#_wD)TI(a%E+n4gMX9N1xa8`yx9O z10Ktu)%ww~7WIvPxcP3|cVga4c=y`d@o%6cY~pZzc$1kXVp_;b|JyR;*h>p(5}pl7 zdL>45E~G|zNw{L}Crny!AGFVv23H2xhKzTfvc+uJ=slR9Hcm`&!;$D?kExdA%+KIJ z#$`n{1lxccdwXzUM2}lJ^g3yQ77qW6_la`*#W_5GZZ{t1*_@ug^G5B$SFd~5-(+AO zvPV`C?PyUu=l%IfxtY-^q}c5g$)%!=t&Wg#jEa3_fL|YZNs9DiZL(fcHC>|*>R``@ zY1wvYpM2o4iYaEGBY%1LUZ7w;!q$fr&>0eqo-SAQl1z{NGf=I!QtTK7Ru1Skw;m%u zjT6bEU2YxS_uu=)r$b5pZ7fe7t%8truPVPCpZSg;vSyAcF}-kk@<~ME6JsZvsL32zIuC_0icA z_b(@~PX7DG0WVLJ%keMg$@?VPiPtnc&D@JjiiIgbVD)cjqlZp-$W*~)gO`gZw>>WRAB8pOM!^jhpX4-N%Sh!*2_k~-oS=Qn^t(Re? zN#L4SodD@HVnjnO54nRV@(j?a(l)x zr7m&;ac(=tanf-#vY*`lyS@MZx0Z#FPqQUEacTuZ8Y2R$TPQZ2A{(e^VJQ*v;ZCST;vubL7Z%J}-z=P|kIw&10&YLzt$3&kI#hA#&#qdtX>zU9%u zw7dq4P^nDwyaV&fgDTwjd9*6>ln^!z?j`5lHS4CFch9AFco(ZqinsCg^u{2~Mp@S! zBiG9Y$xi-hzpCkd{3@cU4K9wT@vjZZ<1OLodG^UQ*bP1-!DI(0rM!{*>d#H<;QwsP z^(9xFcpbdNjQ6RJVtXjkO+_Q83Ub}G$zoqk3rY4a@kf@*+8O7?U0%uj^I~K&Fmi@1 z6L( zpe-yLNXL+Od#Px+=ux@O>tcYhnzCB9Qh@PXt#VMY&n-V72fhvY5BK^aZwt^oE1P6j z-E%$9!Ow&pN!T)SNm>B3x^0jNz{EhV=SjNDOFtFa5}!6=4#Ck#Oxd>m6!NT3Y@-Eh z|KXJxKeLR;y(C6UXJ)jLDK>#3pCCrN=5_&t7<4lCxNY#$s$zYcft>m20e z2(tILKL395)44Iv7l?Oa2ib~1Jn)~;ccd*!t!l`{7?Q)t%X-fyus_vNyLr$!VpQ+G z5A+TB6I7OD8wkv7!ShGZ4rf?5BWt2OHI{9LP8*D~(m2=^n&_KEkhJw)a3<+v4k|MQ z`I2)4g8R^so~^=2Z2@&>cJc%UhBzX@4r6TFEvzG9vn!W2nJUu~%aQ>rnfXFGC*U5rJ_|a;v1~MrMJH_{o~q@qZqrk?yl4VGTYt)< zy_-%;r@Xq_@RlVDm(xxuR(Q2g493U-UPa=eH}x*r{NDXV|Fv@w>;x8^u2;6oCKS0g zLDu9~OaM|Je+4JVQr@SlODvJ`lB`dx;HbhKS>?4#jHg@y@dxbY(ij)5ef*+O6p{Hq z908hG|EXVDsxDg*H%7_GZuvu^qgT$=teLU3^RC{mxzy6YJc7-PHk-^LY ziiI^o4u~5|@|2jL-!)T%0jabZsmvqUh9c>-o1r2h@R{jn zqWcG(yU-zpha3c(vBTvNcplqfyKZoTi}O1Fo|`{DW$Bu}B(seb5BpQTL3Y`{cABPx z*Da3qy&y#eGMo4_?1~>-3sUU-r8eBIEv59gw8?_p{a(NkvUIeHc;~f%1hi^KWQ*5Q zY$8S0P|@eaSrNN^O8hZIeqP$j*BE2q4zpa(c+@r;wP)PHF^8-hwKch-?^@O!IqiDI zN{KKPG`;e4SwBI;VDKys9dbdjpPk-mWca{N#i{uN%y5}L z(zs&f$&d?jc-Bgud~TeTM2dnseb0+i0_<;EIDqDn`!o8#RtA5pySahp*ihnaOG-yB zPKOg0t63p{Eb>7L`;NIof$hRGe|=H<@fGF-Cx z%BZHPqjjMgY{~sn{0S599W!yVUysJerg5HVFBN_`PKIK#*U$TXEpzx*Qsw9g_Jb3b z1lf)9kZ#aA)nozpoIzGa;@mYI%zem!ra;}+LWc(AjOw86Qj~*BWpEo<`0!(syh~6BY8j92V&4ggGx{ID z22X9+^?gVc&*0B7!6q{xYdtA+;tJNIW;)nIk$*@atzH8G-nQq3jdfncOH1_BAJ zaW=9PcfuW>d}^^nMS~q7F%l!=h6g8uIBEV5Y7Lg*6HflG(^^z((mI!XDrpp!MWyxS z@}W8PvbCN-hsbNB>t#*ym9C%=$Q$rH;R_Z>#ac&Ab%84@3#zD3Z#BBUB4yVP;n{nKd%gAY$$W};}+J(uHeUYu|3|`)h zw!l6J#144ej;svJ53Cd;aj$*nkLNyI+lTtOx3cRAoU94;QM5;`pJ_TWESz(ZY;xl1 zsLX6S+D@_1=b24KVEqdi753feX`-9NQ z)h_I%PkWcoyhn~Hk4~$fu2pJ12Rv7^UqZaDJ|v0U;~!OK^4ItpWa+Xd&oSW64xlna z*ytY~LB;leIf07$(RZT4855{BB!_%X*1t3hsS+~?=1?r888fKpCbuL(dQcXvPI%=o=}v zjv_Tw^iF;?jk3a<0#5i|@o5s5QHSLX{%Zt#q-E4Gei@Y@$Y47ZXBTEL74F?Z?e3+# zOi3BlA~VS1Le2yy0qy4wuP^0G1=SyDy}JF^`x(N=OYZ4FlfxYlBd*|Pavnr%8&~YciL;4TVmH~}yeKW=ttQ3voyb%2 zI?!8L3YDI@!P}!s6m`A@kq7A_UJ3tfXf<62qUulfwW{;q!7K6nTc-0*z-)Oov{zs} z^Mw6-!0ZLG5%{pnX8(Khe`lE{c%j;MtZ~+hwz4?MBuCz-FNbMWJE03%tGp)7aCBks zD5GNI*pG6hV|Q6MF#HIg9n(zHk}7^x71`p%(-H*mMuc{EP%N~8>Or-D9-MoGT%WB~ z4lLN=UNrA+WEv<|bc2W!DvuS&u;QRvHmFFDHPBqKgL2qM<8F-rXtb>knx7h5*}Ic+S)N~0fEa3j~iFj)=?Ojb?b z6cFPcM-|QQm*s@(=$vqk*0VYU|7uESl_>iKF@6(JF*`46Wvc0ZS+3_kJB0k7jKEV{V}#1EHF@f0cHF@U zD%9`)o9cMo1eKp$Ke>tQcVeitnnC3-#Wqr;j*7>(xoeGH_~ z&ykdXf<*%!iG1iZYE?G4rB3&iF9Rq_@_0Ar6sz_qdgw;_Ag#Ge z1{GN<4DeNolLQzq{#=&8yX8>`^)Q{Ps_7N(8R7BJ-oDB!Tl!Ihn?bfyas0J0S*h`W zX~g{W?9DNNj1!n9XH1E!d&2~#2mH*RkV+>8(>XIR9jDkM6gh|__LVe}yg{%HR8`kW z_PXoni_&~a9+24&0Hd1*>9nD1DL_Z3FWAi-hCcO=uaRWFHUQ5DF4|Nki&Tas39wdb zD6opOF%P^n$azy4wv|`3D3PyK;-QL&A(vyJtLN!y(B{ro~^<>C}Y%v?RM2cNQ zkyS8mus}uKl6K)n-?O1>N!khcEjc0P}jiE!Q*=}PTaedB5MaN_!G zE3F=FGStnf5*LcMMCE`#YAm1{F5%r8uyX+t{p|%^;t9n2b0Kc#soELG&$!*y?b{|> zGVwhffpua8SmE14#V}mUXq6A9JgSGB2<3yPAGJ**|MX=J-(%gl)jw2ix1=I=+IX53 za^VbK4$!F;`xJWj@pW{bs+R5-_tLFir{vo_waWOYO$%DpSW=M6RE8aWEd$zsVnPkm zassZ&@6HC9b7h6#ZsaEBp*$uuAC$K?N%qr&E>*m15f$$GnH6@A$J&GuGYetmgPl=h z9V4T~VMCiWH=du(abjCyC7XjWSz~DvS`_@-EI+#tIEN>40x^xOjYzRG4xWEI>qz-y z@~Sy6Bk{eS-ZyXPYsKm~(Q>v`f#uqj!RN&qq#kaQHHM@FXu1{G=HmX;(5sAtD@+~_ zkd62lX+9ob>&Ol3`1r;Bul>fd1kOsOSrrJiSMs;`)JYQ-^hcJ>DqWDHDi`;YmWWpM zC26~GV8M;KN0nD+{rToA59Bpo$mHBDZ;5y)-vhd#ebNU13$m(^j@Me%Mc>T)Hr@bF z{+~OlLXV5HIuseoZZTd6hb84f31PV10yT~8&}C5$nf?an62aM%9K$$`95b7@xsIvzaT?9yJ8k8=W=_taWv7|bx-eE+;FiCGL>WUa z)zH{+YJQg&s1vq=P+^bTPJRaOPapk3M?>F-p57mXEzJ{TU&aDkt`Ho{`J7-oxopv# z$(CA?PKypJB9}a+kwMx?du7=_c`08u{?a^ zHQb;v_V3|(B`0W5zdBa_;q%wS&Kv6?R@TGYJa_Z1_~^luXl~9qAEu|1`2+k3G8*P1 zW*bh-ubGHX)^QT@_FL4iO!3q=cGUZk8!t^gb)1=^#sJ0MqsVP4x>=4v;d0?pa+MEh z<2}GWm=QPV#in^F(p=qpUX^>&sr@kj4 z|Gdk+Qw5sWQ5&LQw(4HFy@bQ2yLDSlaA&+3mztZytL{o51&o364v5 zq7yH-Gt6=lk2&IYiFS!#^@J)t4?JKaOQoOLBg=5poaXLkkJE7242p4n57*0=O?ju# z#{`O&f8P56N#kaboHw0GO3gr#N3r1lY@?zxcjGq7Kr{aJ0d{Vp^;EiqCdYtm}hmD_pB#-w8^0ZuTyM>m28ULN`e3LSP!a^u9;C7fzcTa(uoWzkR+kP zy_DYv4MDm<%{j6|kRieDAPvqiCrFb#MY2f*GmFN7e)G6+;=p|JD?k5MlO?+(mx;!S zabkswQYSIcS*qm7CVGDaW(a{+reaz`STA(?Cz)EG{Lk`4OvsCDz|KD3U3@1*oa9*GP;o89KG7~Jsx5WQRlAO38w$Kc( z859eZmg!XVdVa~QHT+AF@oYb7RhNe9XO+-v!{ej!L_Ku1uvLu>$ax}NV3Non?4C0q z+s0pM|Bk-VY|L0p+GzK&_dZV8rNkYJ;%!veKnliTZa$+5!hmNbR4FG%<^^iXr>>b& z8=|L|_#gI!s6*O}dqh*=-Y%|(NNESt$Io=M!rEBI!Z@ke`{|qCkJDHCao$@YGAlq+?`b(&h z{#?E6>ldY`)IcF2L76ojy=*~(BvEoO3Y!`^Cg10yv}G+3=)f|m zXq_ZmYLKLbsx7Xx<%pAH^$`JD87=;%1n8m|M&LUB8N)T>>n zqYV)NDe%&}9qOEOM#^I|2pbN=$)ri&7yj~UmbU%9nQb%iFFJ+GV8DwcQ@Tv^KY1wh_0}AMer;TUZYr!`z z{EWoXqis83>Pz3m1WR@%r!79P!nf4I!xr^cU(~2iWm?rId{;#!O8S^YNk1g~kSneT zLTP%sQPNAJC#6;13Oy#Qa4!#9qIxJu_psL(TN#qE551Ke+G#g8SL&_zzkTedCPVYn z1Dn4=&b~B;rr*ra+@RP_igZxX^&wf(l|jHZ8oDZ~k-qKGOSS@0MlK!ew>xS}uyKFY zEAgI5^Y4;IFhRE>ixp>h2_dcOPk|f!&sUfp2sLVzO+d%9o8LDhb9$$+C}@kGh=fO>S!zw9K)N@3BMkkr^7- zzgB_#-8Bh?meGGJVe?qQx+$GPOxqh4Eaad@Cse|bk7;PkFL zulE=i`K=RCCL^>#ugD;Kxfvnn4Iz*gGb7YMvCv9V4dH0uZAuBi>XUp)vbUb@Y?5*LhT#?p?qi02)g?2jf$_ zeNKkulWy_0@EB3I7|Ty8q4g4vb^|xma*yjm)ZQJpPtD*=2En0I2o*Gko#DI)-sgI%wb@UVzStGwj11lYiIbj7p{uP%iiBm7fMqio_fpPGee zsoG@NWw%MhnU9I%(2W^8j+{&3cd(9~_8;{JkEcf0i4kO_M)&UGTZ=X??qlLvOgUWx zxuN=SBn+u-% z@D;6Uq!u)8*GY~4j==-5?P-q>0FHxzbu%(r)nDX=<&$4@jhUssw**=AjnGVw+Icm+ z1WCWLo<5;!lI!U;f+pEj_f~avR4Z>MKSrdJRJd0Lr?VJp#eL_+6$?`W($!jJVnFQ7 zl<5uw%kz1up6$8X2FWp;#SJZAJNBmr%Wl^f%3-n60c)iA8Ls~Vg+%&_&+WN733C1H zh{#lxxTP|vDepLNJUwoljLXxPJ8X}2xNQA>*A#D4P_FH@Gar%^CoabWIbc}aSafh3ae9%jvyV*lMin}d;= z9k(5h4o;95VND!{1=EeKLt?b~Ce}swHA@0Hr(G9Yp@cI=V*6!iFNR!D;mpWYTScmT z@|2fn=0ne+VABx+HYLDsp9x~?dh)!=_h?Qp{}JieR|VzeF(#Kn_ab3zEU&{e+7vhvbO z$s)FW=;<0BM(c-MI^8uj{=0dl3wHQvj`A+g!R#55@$B_Q@vaGbS-FqGlPBfKEl#80 zywQX=(yso(#>xW&H0fbXb66y~1{nQ-ETD#WWlpOS)hZ_p5T-%pfD1>$;Hi6a!iDoz z9$fwgxoOE8$E9G~iR&!4n|ZQZC^nrU8^DwGx$PSxIuM2QVry7~Fl8Rl`p6EeFhY_V z{!maSjc2=Q?0`@5YM3xC1P(Btz$s|bq#u0}wc{^~_`EP;mAp-sCxy7G;^efYqCDxi zt3{qN9yaZ;1rj#h$j&Ea3a`y4rsZsrOU5y>l3R_56Ppas02)zavVme#D3VM?uTb|Y z*3T>j&5a$TgItvAfu*NeJm`|hA9Pv847!~5n;1{^(LmaQ4AV0A(Ipe|Z*<=6dbJ4| zhVZHB?tBMfX4S}uS4@v8KBm5tC^um3FxMg86jV^ z>($gL&`{YdM`|d%{?XUyY5mMAKANu-X)`)iopfBtDlaYcA&t#9FFy_gk3)K>MYF%@E=~B z5&k~1I%%%))_XS5wZTSyF}U@#c#sY^(3>FZ{65l$;owSSUc)MMXCV z48DDlrTh=q^Uny1_|3}SKHAfw#%$pU-zMRsLs;ckNtZ}l=7BE^JtYT}N%J2HTGXl3 zm9OW#VOVrZy+?X+)_Q(5X;Byd(6Cth)-Bj)*Qw8*kg^FYHl9@U_UtxGemJL%&01l6 zI|Stb(52t11pe$HWCC$c5 zvUavY%CXBtHdF{5K&|Qv%oU$^h1hQPG8plU_~29TA##AeTBv^fQs}wH0V$b&_Hbww&k?3I|}=X~=E6 z4L4vI{%}kjZ34! z9@pus6&*aJPCgDxvYM!50aR4>lVhIepp^iB&b~Jf%^NP-l^Ld$E?6T-^wTP_IiSLQ zYe3n2Or;rQ8Q~VIa}E$4F=;twb0l~R7S8{VZ<;_ix$oy+ka{P!Mpw+N(J6{ONs$v& z^oh4n#XEzSGCg5I&(s{%17Gw@ii7tC=kfMNHHg*-mJ2Qk4Z{A&yU>+f$M2zYJqsce zBx}7&6#Z~zeMlyGAg$wfDvpzC*&r;mb3Lo3cZKd!?eHq2J_#@gPttqczmi{{{eJRm zi45N0`s_Nv`>pC@!V3$FE=42>s)Z}TNY#h52i$enDib7~id9h? zBxte*=pOenDvj)xT%46OAL?l4BiHN7@Md9?I76Nw=>}ybsNfYA`P7G$&N9fLH5wA0 zUq+k@$>Cp+E(vOhNVj7P(s;3EF`bPOSx;Z4|EI<~|7b!_$9La*kF+`QB=(gV9=a&@ z8bu69u-y!a$YI8)QvR1>Sb!OT4FW^8u!*tS1CWqvr$6x=R)24Xn(kH1CZ(p4M$WNq z`~r}4D2}WQO9?=k_rtQ71)3)xTa<=~cs7lFAk~QmVEMj17~;$w3^I761msJQbt=vs z0>iCrt|tbPLEJG=vt3l=gF^8qd^M})UjViEn{xfk>Zl~a-ubtE;phb(?&zG89GMY< z@n}7Ldv^cq_uto?UvNKUnQEEJ7;hdq`Pmsi7I5PDaRiJw*{VtR4ZoR8nb7vo^Wkwa z=rT&Y$a#xE$YwJHCQ$4r6o~_Id&o-wS+Kk2u;LsKLjWU(Zp71a@TLtQaqv$t`t7ZQ zWH9e_L68X$jc=?iBU!Gb*6bOVP%ID{6h3;!#-_h!c@C^L@3u8P~Y#@&C<1e-TL?fWq~{DNgXVd5GQ96C#}XDA?miQWJV^sxeL zkM5=ov?kvxU4}tL9Sv>MDUv+k3f9pvqCVj|uZsa7MpVPQr!c^hAB6XCn!{q$O1jGn z--_6fp+F6$FKc+Kru6ZRAttzGf@Bj+Otm3PMUc>lK++gkwU>MC^}=ZplkK-aqY}^;b&D|?N;aJSZrWegvOJls2umYn)T!wmiVw8` zEnx+M-^RLrI`x8RJF6tpKW*0t03$rWwt6|Tmv?1IQ%x)k!KI~QGHc0>NO zT^Rc}>9T%^NvDN031dT6%q_5+;{8~$_GF^5*cv~~hVL+7xG+^|a&Xi~epXI0xdnBc zcnMWy=KAcWSV*C7r=l-S|dKNmkr1$5IN)c8O2Z3lhFv;0BD2Pla&6ueiS z3d=1Z;3Ubw<6AagQ^ei}CC0M(#o|)calaTpJvc?0%7A8Gxw6FH#v(17N6vnjvH2@) zI&nr^-q&CB?%BRMU%lQVy(sOOI|y|X*a+9f?BrL=uKQk}eIU4MdaF9g?SN`02xcr1 zed&vJSluAjw41IJpK>%u9gwS}QiS~SKYye7{m*{&&%gV3=?aQnMv)i`3-I(un?`RO zCvlpXsD#5?g5EHhmIwUIpO8vda?b3XJWjDdy?v01z7%Mfc1MM|ywuqPOpgNpUkDkS z#IW1I0+vN50Fz|gf|$^9V9(e}jz;Cu#`D&RPC;<>YQ!?Ke4dUzLQc~AqkvHxxY}<( zhA4mTDc@FgzGRKp2~s{cU04f5d0nBO0V5RNu~m&lzqo~>Tix_-A7uCbV(zK=wIO5F zj~?r|Jw1|E_l#-uxQ(0BirX(~`XcBzCP2llN|;Z&oj8R2iCH@RA;k_*>`sN2LmmBlhd!}3@XW2f;TX!%w zJWkEm991R>a1oDh4hk<}_Uq9}`I5c9d8%~pZ2G1gk731oeAK1|LoUs7EpJaSRxTqE zq*nQG&Yd~KD#sm60uRgDaf#o-6!{oW$Da^dvAj={TOJ(@F7t5z@CADM9Kq8o;r$tM zx&L|!X{YxGVOZe2`yPLcRm1Oy)h2kA(~1mpdNC$-KIHe0M=Qg5!(7(1GO(^gPh)fk zX#$|C0}~m@C;(NfcU5Pn!=4UCJH2W87htT1Pub0DW3Tu;+J*es_s7DkXrq}Mr@^)H zTezS0`!D}xw$Nm$7N6PhDOvr}c-6UPGk7}1ZlFjCsH#9Y5H`?cfkoIPDpMX%;cRPQ z8VN=KAkPJ~wNsbNZ3*OYf83*A!4Vvz-^2lXxBT0Bon5R@fp?1! zG)>W23u~YYDoIc>V*vV~Ak2$G!yOEMM1>S2x@u7lxuh}&RG60ped`%OSd20}`&AFT z@9WOtomd2rbe?qBhkBN zlNaN@OFlXYuI1v{)3RPc9B)~uZbrAbOA6e;^Kt*CfIQx5nO6BHJndfa<{8UE|A?D* z1eX0kg?O`{=~?btvfI*eaoRw#71=J--AE700w#=0vZ|1MFsufcZiZZ7*kNCBq|u8W za>*oXVS$HU=Vg)yt|ze(I```THJK1ln0Qu`?KfLmD#*Xjeqen50aT#IsAKlG< z|8g1^=e6C)=#{HiLl#fjJC2h5PXgJOY>Nl!&53G%+v1bKjuuP_MY6mlx8 z213MZ7VHn&E(Oip=}^VR=;<#bAQYuS0kAe^z@s6eKhjuS)kBso*eJ^m+9o{^2o)bv z2Y?u1H+~@754?K;WVM(Icc#5p1qL+AQgkxE6A4o2O2Z?ebo%S!F z8TqSN*|zD)Te;SeyE#31=Uq=sug~6RNi6@8G;La;oX5a3gsI(uYcFiCdoHh7u|<|3 zX<^d?GD5e5XyApwL?~8AiO60-r((Tdx<`@XHeV+(_~LN~*_X#iwF&Mc=EldHkMH1^ zjh}w^NOYiSs`>TKxdc4oJ#?oiRVK!5^#4(_U#& z*M0x~-=2B%l=>vJinppU6r8&FqI3y$h4~cH*h|?=3AV)SX3mj|Aav3d+Q#Pbw^4`R z>f@i1|G86m#izyoELvkm&13i&H|&j_AGduY zenycdD!N%d;F$s96*=^YkO2l4(Hh)1%N9gpzAOEe8g?tJgEc$7b5vC%&aE0!pqTF7 zG518sR}4s`ba{OST{J^3mHd23Dzls|jL1>}J&L=jz6-ngpsp)R)h2`D_7)ycR+w(D zRqhGi6?}piKpdlDTFneN>u%7lV7&TNNUWffj)#PRk#WcN<(R#IWq~v6A8kZn+G~F- znrs5r@|B_WWXQ#7GiM;nHWDg{6uX8ZtElK)&jj9S(*4&DzS*Ka4<^LU+#xnTNt=9I zo9|`M6FGTU)auyFI$skaesW^cDU$9=ppj@q{cS$Q0!?!k`nId!aU;1alqmVE_b8EG z^0^|vApJtP6Bg8sbj3TG8*f}#lopZ*`R_|Ul~8>!BzlKoZGzQ#L2|1W|&W**kp<%P|>O32HFsD2&#im zLW1}*bD7EXK(VE&d7Hx>hT@UP8M$pMx<_6!5!o ztt0=mgW%Zl=VsS3RVTf9CiC&@Lw667WlkJ-1HpkDnX4<}Si{^nnOsj)1%b6V?Rr514L4P=VY`W~jY zPQOo9h|lp#y@8R_uhVyXRHM(C$Xgx>ArF0%{VqUKr`GfOY;3OngclQ9Crwy@XLrx3 z4Z+Y!2ZU@hB*rsu&BDsTdf#53UE*uonQa@Ut2D6xWwk1c=c>$NpvU>i^2yM zRI4;t%hXG%94NG9F*YMV+Jc{v=TFr3+h2?MuSKSb=e_zL>>}kajo*6KY??Ysu|Nya zKt(qxZV51#44nZBvn85kJp9n?70)JmHpBWEt0}8O)_d%Q9`k;W9aAgZ6L>v7*QA|7 zt#Vn=J`mmM;MIn-$xzjJ$OYxPk;W*MN$2$`k_5<8HrgF4r;QDOA7=ji2mfV))Az2NAf%5QoSZi^ zN|MbgfR{~Xfnn<*6@BVk7p4Du1>*m`0{zraL9@77c|ugg+w0yRnH$tdV~s-{zeG{z zTM&7WE`l<3jPGVc@e3Zu{_`XOUa93Hc~UOjM~Z3V_Ay~~aICAIpGCKL_0Gs}Tj9E# zmk5D89gQNkU($ExXqC&^R>*MToo*{oMtiyZN<_9~splb*AThsItGqtDk?vHZNb>23 zCV7D@kDnM&;l7kz%m3k7YI#L{X9q^|;fxjFNfxZ(@b5;oSt5rE}x#5*zYoNC!H@uH2pdL|~WU2NC>FKI@ z_U>kSwte7YpPqdu`yP$@&FlYTDFJ6iX)RlozQ8Ess#P`wCrA<*9lX!$-~DL$KMlQI z{?6WS#s1Z?w{$^c^e4yiYjJs~_N+I1-p|U-2X)#I9(m{T?C+T7u6;ia{0TYY#B1>D zX3k9;#kNx96p)?JD@eg2n1mJ$K%{y7ydFpxb^|Zt;V9(V(<;wPmw2_RbAy%uQA8aO z%5L%Ll4S^LcwG><&IQhg+7PVHJSG1UCY@y9iRhZMjeib0s`Ty^?xp-qQI9^gst-pM z`Sg2asXCY}_OAF~)YW+xr1id43w89i$U&C_QS1Hc_;vhR!5+!gdAcCz+u~oIbBkU| zutTes)Nts;hJ!&^HmeNTqLSelI_Ik`qT_vdngt}OLqc+ z8u-Q!=j`Tbs`+afSad^9@~BZ!;StaIkW00&kwy<0Tm{9tpj@EX(kd@X^VHj=xPsS7 zK=&tN$fd~VW{^&j5nit9CGEmfY^6BS?y{A3#LeR|x5ox2+)VoP{*u2N4+pvv`!ZHI z&`X1Rfw$(Q7}CxCvUX3*J>T|t$bfRk9>`zk14#{JskN#Dfhd)H^|eQ26po*W#|S8U zVCI?k$gKZ?-ZIJXl06)Dn^%!W>}r~V)P@f=WQh|`K{~T3D4Ajt zDDnvvJvL4{mZdz0xZSIMG*4=_jZFg|tN(;n)+4ebPsvb2^pjQ)~*al+ z&+ev;{c(6zJKZka>|H+dGE*9i*C1PY|MfcYOS(OR!DA$HT59{G$NyesrRb~4*CZR4o}rfb_C23OYTR+_QqG!4<_+0b3>C>q`9(=(k&g;=uf$Zj3vx{6G`H1sv zM8$|JCT5cpDkwqoN4H;eRHTNNNXsNWKDvNx(G{QjAYpoO@kcS#GOBUW*lhne2FgZM z*s&?qwjY)oD(*j+c`MxHoGAXF*+O=@5|DTtQG8NGvCz+0Mn$)SpmZveAVDhbW4u&R zD&#G?nJ%WmZK>$&LRd-;EZE|+GJK8uAgY;kK#lWP5E?rPd|w;qYn7O-(;QPCA|-C8 z!wj;c;$_VK2(1z`#xXz^+!>(-Ny($)WI+cn*7qzw8{CQu(q*!b#y_sM4%8^~_Oo`` zHX6@3i_<_kZ}H-<7A*R?W%T4}at7!9`GLTCB-}G77U~?*sp!Nvkj7^NJ1oJY>6Ig! zomP2Syq|}hn;GF)gsEwk_eJ9FgNh~I#gTh`Cqg4?jCg(q676%yI-bLRSMc_W9+)+< zlBYs|6;;7NV${En!3Ik`T_1^-P^&DUvQ)svG4;qW3x_$gM>gF_@8oCj)=lXa$Aq4n zg}wN9q@Bnl-6q2Y6YAfN-_$)G$T4EXioJV$+i`DL$H-`o?DoN#xt8q#FO+Gq(#y~+ z#})G@p7kp1JThVm!!1DilMsf?;uu21>~hx3Sl{Bv@|iHXRsyY5DwFA%67Udoz4ogr zg;h}PVn@iBkL`E?WPUgMeXA z;Fu6gSswU03QFTelA|1G-sKy7OD_4Rjf)fKcC7Hr8wF1=h>J1YHmK?a9=h$Kt>9*N zL)G}8A~_1H=USl32??bfHS89zJa-gW+`vH}t=p47Mt`v971PQn@K+V@lZ`J;o!wqD zk2jZMq2y^R6}@c1{Z|qs`k?KihrWX@#ZxYcH_PslRy8Db=-#Ljw_VU{(5k-6!+)>N zyFsq1SGsm8(q{C7wK*@oO!NWm4rDF_0B{G<6Nkn@_!++K;G#>L|FVkVkR$56y9S9Z zF;JfpFyN694qgg~34)+kaU?4EOps+69sy3sa3mBg?p<2=)1!Qo5pfA}SxREL84)L* zezutzkyMIZOOZq>dV6qzY&e7XSaLhp^CU$2ja-%!Hs6@WfdeiqM$WMxFK2#q=P%1y z-2~l>Mh~Q$-Y9z**e>jmHHT$-?4OZM=cumFE>YwL87r%@|~PybOV!*3Y~O zq{~_*X8VvS0VbN6j*TYWK}c1$?{%w?bX;Q9nc8}kDAvZpd_GEfIiT0HblAHN%8Bs-x8s9Lr} zlmm6Rz4CNfY)H(4VHLJIQVmxoE#Ca@kPGrfbwuoV^<2n?S1*gNGFY~MXLge z4tORBj31ih)dG#dvkK(W&Vf|{Ih&Adx;UbfE~YEO)_G-4*B}gG@`CDq_@NBb`UxII zE%!d5HXzM+$Yu2_6>RtLPoK)%6{kpUG5ABF7o^Fevt|@W4r{;-FzLbyb}g{8?e=ln zo(}74p;3Z!lX+`aetSG6aZa2Oct%NFvzJyC&mzkYGWny1O)Hx^MT5@Hpv(5)!U*Jh zNA7MAciSlInuAl`;qVkW!p5<7%T^;ba<(#UJIFj4Z~M(PpMUXz$w(FZA6Jl0CpJ>C zW=85B#ongKEh^d=%wOZ4B*12}R0j1@D(Tpe`rv(`2bF!1Ao!Rhz=(boS?M*L{CxjD z$&y}*(A*~maw+0{O(lPePZfcD5psMZmq$Oe2J|tSWUvqG1(*?y3C)+bDlbbjRhm2| z8>*c2^it6(W;O6?W=u=v=g@F*i?TzBfAb}%Qn5u{C`S9#8rcd_K(J5Ak%Ni`{{#s0 zVKiA6&`30B2%E8g8>(#ZOV@dQ25OPXkp+!Ng7X!)BL?L^x<9!!ifm*FnE%x9duA?l~cvVZ1bn&ryge)lB##EECQ zJhRZvCW_ri!6-x<=zTC3XzD;zsT=1eB%9jni%Eh`C z#R(mg0=+kc1(?uLF!ZCnB*Tf(QE7&bT@<^6BKcG_>X9Hd<|f7f0lxFnR=%c`-vm6E zt-iM+hkc$+%J}f|kQL!WAR*8i(hR+j+xYjQpunM+43l2olrJS$`F+CG^RRHDO@=sf zKtPZ)WE^QlMu3nLWG3HDIr+o!@W?pv^2!R2%-;FOrJwq05EUTgCfWf#j1y&LMvRhe zSthd!Cg{VCgvIM^B!Wi)56$C9d~I+gt)DtD#`cJD{dTs!i{tu=FD~F>$UPp?YcwT8Tn^$Tn7UYJ@!Rt^ILmlvHQaNKY75A;o#TAQBdz}nB5V$T3 zYJefc2CU1tv2v=WR&qv~Lw71RF}*WxG1xkkBp5zp?dvh3eNtF6)5}3$I7v{%yTI#U z3ix;sR?lvgo|`}u(a&57rFxn} zWVa7aOnQ1hl&Nc~WrK9x2fdYs3`o`wOh9-+o4kOif?YxsrYHW|`y+Ou}|*+a$^8qdC){f;slmi#%) z1dUaxy}QT`R{~PiBN|C6D0V+ZN~vf}u3(A4us$kOj3@^c`-YoB^gxd?Oqq|Zq!{6B zlXWVR+%%Vgk{}flf7p4_!ThO`J`FL_R_POk^*3u%%F?nkzKXc+a#kgT4E ziKRAXiC}-w_y>sVgMX%dA8~NJL%_+{J8vHN<`0+u+|reKNrJUj9G@0m8V~ziF%*w9 ztjPF=p}6&eGTm2FiQ(0FF0xHbgOJSb@w;@L5x15x)#ua6$8 zD~zmhBd>yPY}ex+eUa(1W1+`A`aK>jh;y~Wq~(}*VZ3j`q)9)%*8Lyvnv77f_k*92 z#uu#B-e6`i&roa&MVf(g3hY>EU>$h7E2kP2&v*Oei#j5*=}KX@Sko?CF*lpW4m-4_ zSx`*_#xa$N6QL>zmd&6%EVwO?Zn4)_3+3Ti1pDUO>14hhDsZZZHUQFCDR2NetMVm} zcEff3mv<+L< zZE0m!?M#6QSu@lA<{(+&3aS(%r?>SKyN)7B$l71ypBJcAcD=Gywwz}WwhJ@K1Lj!W>6rD6Me(4L(vWE>W9)JY$K>?)rWvi1yJ{**vfjpJzi%QK9DxjrHd~w>JDmA zZzG4keQ8lw=uXK*NaffLb~6j%Ac#M?ZS0z_4ox+oA%0DEWld=2zM#AzXp-MyR>&YE1wqTX}UB`a~C#+JEjPKvJQD?ecXcX7&;Qc@H-UKeH z^V}cL5zb-e#bGuEXHLL~jNkx9wxJNcV@%rIHf?X0+uQcv_O?wM`nxx`?fvDpQ+u1X z!F^>@P(d9)HbGD}aRtE*qfx=2Vh|N%(HIPZBEg0K^UMs&3>=*UGyF!|egbpebIv<> z-tT$dXZ=2ySly*ciiXyhU|tM{LZV{uaS=xG_Q;Nr8r;XrIQv2BXcuZ(W==Ht@^EzY>qfSLeiv z35Tz)T7J>@4&5$KopD}qIygacNVAo06ZJ#FQ=;2yMf8`yTmh_C8IW zV!3N9_)%84Ho&J;sDfSL3atJLgD0~3Uy_`icG$mD~QFNf@TR`o1!}BbMCe6zZ@I1_+Vr-8EC`Um|(NkG)04NjV`2{YJ$a|~mH2#hd{_x$TE`D;kQ z3lqdSZV_eJN3pvpQclGr3b7gxxqa@DICcjxh8ffc19ytAg+iG%y%lPL4CK6Axs*u7vaje6w!)^O~94Qdz(P{G3mrdvM znSQ83YKJ6)WI+R%QY8$ggH=vTnq5w3&)u!95_Hmd*Ox#))d{%^)$(c}O&ox^xuvtc z@$*!0jXcQ-H(oOIrR=X39H-1^d+7P_G!$_^?dd^ z7W`j<%-U1S2QE(RlH>5}8uPY2uY!9mDYA|(5bN`3gBDv`EeAtt!2&xl%=1_qKCs|1 z?YJ-9@!e&UX!Ewx_kF*UWYKe$ISIK*!?J1CQfvZ6)=)9K1z$)8JnETqKtVU)(ZJ-< z4?NJK=2jEN{V2I6ux;tN?snWb_jdEG&r^PB9YS_m_bW$G_!G}Kwl?yvEJgL%yeqOY z=p>b?Jp%1m4Ql+I=Z@zl6b!b(z_b%MZqY~1%ngIl@-YKiF6CNNv-8-YYy43?3mRPogvME)cY3T#wZai^DqUE*LdZN)0pJQL!Jh8 z?_3N}s)qUFNJ!Xgq~Rww|JRycg%4PqSm)p%W5LY&hu}&nGh_hga+OMgc+o(eq9yGUsk$7pLAj$LnzV8EVSUj^Yx=^rT=KBd>i+iK|n=Sv} z{~ROvS~%|i#i#TCr&WZrt`uHVqQ|&8u=CYpWT&8nIp$^yv~9!U=wfBeqc%Wz-cSG9 z>ut7%8s4dRkEA-WOBWpa!z`bCiUmzL3o4hVA^%rA)9qcZ?3ZE1I-;S(8yc+2YND?L zCt#y)mvX0|pFp0lTZ!_jEIuSplPN&fr!zj-CuDK(2F8|EwBdn{ddav}f7BW6GssI4 zQOv%(%VwLg@_wbRh7>w6R!&%8+&2iP>GsDoq;OqDq}{58^j# zW^V$=f17F-S!w*-pu70`nl}r+f9aibx+LmP$<+EcD&3FI2T{WuBfCg%VC8hYr9roq zN`G_jH%i`G|Hir3G02qf-y3*45YO2!oqSyI8BKcF2R$$GgU<{m=uK<(jgf6=oLB>{ zuz=L#W+J9GG#DMq2y27PmI04=aK2!^Qv*}4z)oqcB8)mA4>~hR=+9yT0;|6{S8YJQjYObSgJU<3o_byCk4B6q$GT(pO{#^%=_|GU<7-A?Pv;~?PvD5_EoTu=~IJ~cTL`pFlF zaik-tJsis06)P1uqBn(_SIJ8lbo|$g%6v_3e+P>b+j!$f+ji_@A#)>zTcC{M*n26r z_cw2Qn$?+KHIsKqk}KJ1A>uPB7Sg3QQ88Iz$OO>nvqYgsKgYX^*hi52@$OYw#cTD7 zLS~_F>I{szRMDw{Ur3TkhsV$r<4;)T{?{7MjQu5i#>C@1&2K53Hm*PAN57E&lNlpF zyK#0C**99S$cattMhifnpx9c99HC3%g#N7OF= z<*qATmy0`~cWVhm5>~k;MMKsJ{m2FH+QpQBTzaG|VYYdu&si=$NG`v=R&ZB&PPD?c zQHQC%cz2zo#BIN3>C`I8Dp#!JuJ@{f?3fh#srEM9Zlhunh_rNc*8yvv~dAr=y%(X!cZ9|;O2 z4UW-vD7W6y!9SX%&#!BLFpu1JB?%UN8x~GtA5r8!71OJVW6ucM)FxvDQ|s@MN_S&S z7zk}lKK2F%Op(w;l1ry*`^nMB{n|YD(w8&B&{1D2?V;C3-0gs*pMwcy^*s zkzxivV;5e$wV&LHY*$_cGUIwhj{@6+SGr@@nO#07wB~*@Chke>AKsyArHAG>&)o=R zJebEebP;yLdMI27&Nz%CEEJ-ux`H|ySV)(8X9(*R_sP(eYvx>(;9pP{3EgL)qeQ#t zGf`P&tLVk(;NZMgZ*9ki9o9nY1#v%fh`2@K%d*>Gb~WhM{qTwm6RdjXUXYgo>(~bA z5kXnx7wU(K3|V&A%~|nb?cqBFACa|Qk5mtWOy_Tq?77L1Ong*4fPOtORBfoAQ-R*Xi$$;#tLz+sV9-Cq1(s)&h z>c(v3@IIlu#F&|EHb{|RtPSFw4%*oNS=iPl}fSj60Dz_2^=)$b551{B2pZhT5Pq7Oqxw6FcO}Jh(ft9RKYdLvD^Sk zOt3Ox(52EH>Fo{b8*f(3OaHgpA1(Pw#k^DMYu@^SImv<))w;3-InPv`(pS^*y04$#sX<00|j4gsxLuXuXDq%j&)!0l=1?r z_2z{iODk(kM^<0YdB`n^x=$>xv2K^;v|AKM1FQ_@3d9w%=U$nW9aia)A@PtR#GU8XVP?-6y7)cz$+WxGQ53uemdP0;#e6o4peZis69?_ zqhb^bjAGQ0`Ev9AB)Kj5JxLtR|Kq&cJK15e3uRC&v?NKRVwTQa7lM)n#l<3Vz6>f; zuF4AOBnU}e_rC5OH#2S~RPASLzfg23fVttCH%f(UZNALW1uSESjXl7@JqI^zV)Ruz ztcz^<7)Va6!f@0Io4UkeWl689oi=sMa;z}nSQlHs!f~rZhu>TMd!LzRso^5k?j|{| zNfeEk>3_p6U=An*H&v+s*W>t zK!Saa?TFeQ_!(QJP8KBi)zSlVN*NsM)ZyI=DdX+?y>=>5$EcLHD--OS{%^|zw{=gq z=L&GYoudhLr;RpbA3GrunQ`;_g$;+vDkn~`*=7OgG>Y9okz^_cdt2qb0)es))gj?p z&j&IHRGuV7;it6^WYFGvsKB-r0+CT)a^GQcd%_Qy4Qi{`;f8<76>E zQ`w1?luQdKt)tjPio{VdwMOozZlgs42#m5;q*iy=gH#Ph@mGpOwgyyyI+4l z(2R=Gu98J0?}b6dQ43V;rr2@{h`M7MwT~k5=<`#dd;lW}ZOoxSWJUyuQ>R=PU~EU8 zqiqDjAv}{Si)T#TBkGiQMLBf+R0D)a>cE$4P~Rihi9SnH3mG@i&{|RHzG=>JaP1{{ z)_dBsP}@f2*t%`nrN{um`AKr+zgU;YI;|1RQSiFQcaLwCWD__zHifl>Tmwc_)Jm)9 z^5{<{mTU|ksI3c)*G}aBcx^M(t5;rcn{O5<4+NP%BUP^ClEs^Nnqq+x@HiE7Q@0s5?ULk)-*5s+?c-o920qx(oADzIp$f4f7tkr2M?=JB!|Kfltfds(F3& zH?VMh=tf+&v^m9|2@pAH2uV<-2h`FZF4zSbwHKHkx+n_L_Wyms0hq2RDpT{Ph17yW z@Oh_T;OoubU+}-TMM%%z6b}92bxq-!8lJm@t?Op<8_Ibhzg6eAZceMzHcx+95p4$4 zhCF2ksct$GS!7+*j>!_NFL4z_29Gi_~2gHHFABy~snA?_4*9M%GoB$@lI_03t z6;W$gKQsWqKE3*+DdoPIbF%4jw*w+z5&Kv@lvyz7QWeqw!am<0qu`KA2fetmwMey6 z=Y~vIE2sg__HgI{;}_2ymtjMJR?j`bJh5Z3q7DU;#W(Rg-~}Y-RTF;@8DHXR29mD* zeLY#|#MV)|1;UdkHl8A@sh9#W5}6ts!ewjYL;6TJz0~Vr=y*`a4-1d4whgS0FQ2dv z{J^?B>{ojvW=yQyboKv`I48ygWKa#0dRr(qjUpSUm^R2AHP%R8^S%?&Ea~#p=R#Ym zY;6q_#~R60@#IOe(%;0-*QS~fwAh9Z%dKuAK*9?jlfK;d>P>6^Js;<+6K8gEc=q$5 ze`LCJfZnAUp!bD$gEKKjRq0VP{Q@~8s}t2p)eYSvf?p#f3?7V%7nVVG9})*Ga=WL@ zqn85RY~rV4kK9?nnzbPoh0>DyrNN8OXKeXfBac1&4u?84cL5thYIdwE+8&C99F$TjCX>4Lox5)=eN|OBwq}9J&soU| zaazD_dZGABWe!~+?G+sMu^YXMapi4+j->|s)FZp0i5GNU`om}U))6ywN?kHek>xLp zi$BK#Xd5UtnIcJ4jIp{A3#zd)VX3g3);BPw!iIXNd_1D8^t`~=$v3XG=%-AJx@)cTlb}mjP9$XtigtW*dED_SNum*+b#7h+>UFy-3&wL7_gN`Nr{~ zt4!njH3lsT9&84P`c{jiE#jzrt9+yUjJ8pCUjATC*VIA(PUU@|fk2xm6&gx!r)%k} zbeV3@zd>O-zh=%sVXAsS)ulKK-*-on8GK{*BKBxte<);8(U9vMkxCkLDPj0KkM0j` zi}=hRYX`1+e?bI$Zo+B;oj)}P&`q@ z9wYbW70N$WKlIOdxzhcELb=nHxvc)L-9MF?rIYf*HycO-Kk4Me=2MY{bV{dKAbC!u zVw&X0?SPKu?g*oo8YLAb(&f>uVXFwpsuWd&8Y{K0f{ArlxC2T9CPH)KC8dToI(`CY zOji$n#qF=weWRUL8|LT)-T3xpsD%CSnyO?T=4WAv&=EnYvV6w*sYe9&$p$Eu-RQHH z#ELJ?YScA{EEKO`vY~|NEPYDk@TDKqy2JX9X)$@+#7oPePA${?`bDJW#grA)XD}&0 zbw-K`I|yJht;1n4a;0no%E<-#^;6$37&G&A;vU^_(+tKlV|QCY^Vip4D{m&r{3_ zNYkqq-Y3bfV5|<0v}98(WSAMKm}beoXam&G+?iJ9i|i^@A?Z@&YZ`RP(dKwB_383J zvh>TcN8tBdB+m1#p1jn@phkk73E08M1QTZy9kr1M?xgZxL(zP zh)l9p*bQV+dtb}eBD=-}pmT&2u@SB#?sN12KdI*WUmqs^-Yh)!{K8W~ZaT4BZlwkF z!+nbFr^sEPzof5u*DD5~?*kH(mH{_HcwmtaT!XvUW4+%WxqZBclW9FoPadT;Tr)*gk<)5!X;Nxe{+-#t-XGsLbfwkQ8{4tapWuWy80wT#AKes~Nb+hXuG$Z7OW&t&fc;@oob@Lm$zn zF&AOUtDvCof$XrLh_n&CDJK@WcreYrmb5FEI5=)+qcpHmavZPg@LLb3&7o`hu>0%Q z-GKoY!?Hd2PQaAtlhx_7mnT- zEarbfa544UL4UATYB+R<7IG8luTxa9!3DB(?^EGReJ;xuK^w!G=!r!HE`yDzuvoE) zICw!&@uLI3+cM3(t!U#`9wb|wSgF}>A;yX*wtxcajF_xfK8ry0U{`dd5REY~%)ki4 z|D!U4!mSU~Ui?4O7$mwiLDCBT-=r3oMc(v61AnL)g`*&DBalbm_HfiesOAUcExQj& z#zXPsyhb14pm^$4K1{Xjl7`fM(CHPOtyP%{J7yY1Ux&#?9Z#-KVc1+kxbrB99`sq zMTU62piP;_WJ;fWYS5*B|Kb}!UO6X2*5!H3_p+)QqK2q`ulHz( zGF7XjgjJI=x<^bkvYTJ<*oj5VISaj}mST@kq=t$) z{m)lqCEi=*b&7}nhPU>}s^{(y9D@ENYrTL%SMde?NZAD?a7P8*-iw7vs>IhS-O+$g z4LS}iPsi1L;EX;liVwNy{m7+GR6YY-({!V*PS7u}rRzjp(5~33Yi*VCUiH? z5iE;1B1oriLo&nAca852vi*e^+IwvNJ+3$9gDtz)S%(ju*1gNo!Lt(_b3L+5Pd!E; zDtx-pAs5u3b~v5;C=cGoWgg{D#}4t6d%ExS$68l^jMP({H?+mY-ZYG8vVp-WNVJEl z1WCd+rd?h$1Jn8S^;3=jiVCE8bPY4$)h)`fzj``yscnRfwMZB_k)21lVRQ1sju1cd z4pcb!qY9G2ZwGQ>+o#H62P&pm$Y&~`Vodad1~u3+I|bFWkts-kAVRw6*g55H2cz`J z1!g)?>7Ev_h*SzrlSJV@`VLgdUh~>X^vKYRC>$c>IN^Zy;$sAp!@-q%nOBRDeBluv-cuV)s_kPdleepIx8bwn`Y48KDcuenJ?qv!=+MmJF(v&h*$f<7i+SG>60n#RhewE*bb?@SE#Iy9MR4PEl#p z%E{bd9T(KBB@_Q+PB_jz^!FLoh7^Yda^#IGvPRvt@Z_kYQN4mv=CW+zoHNp_8Ha?s zqz1tm@lIho-35u#AJ4xc!}BG|Bw;@|IZW5xo>oFP=e}JMq3p zE<1%?v%2B!Z`5t^z9K6O8~TQFS-I?t_%!^-{dhV5eZx*Ta0!Ne3;&PYI>sm_x%0l- z-`{HZ&MRi2_3iDy{!g-*pU`sR-EE(R&?=^WYgK8ST6}4bSVu;gx8@MJi;AflvV>`Uc|7eP zoY)%YXa{iuRL@JQ0%?lsI1O>c(Bzq20qwId1nG-?5}Cz1)Mt-iYM6kL7eC0 z^I6JsM5&n=h}JsmF9#bW~uU1k5SOXEKDHa7jBP$GvAF8C^l~ zXm~L8e+0j0J7gb>{rhaa^92zfMW43)&YB_Y1@X*q@QGoPiyryyn&dTfzXZ5klKu20 z?j??DI>ikrTcOQ8gJaJ8@B6RygqRhMl_@>Nq{x*3edREPqmp95GT4ndw#Wi#dYa9W zWWPcx9=hUnyKSD)rMl{0KBF(%;CWck#u(Hmr=9($LI$&O@hU9%#Ae=w^A?H+Jd-9L z)|!gdP3hB%#F^8ZC56=Ww|eLGf3=XpmfxGF#LcpoonY&OACe}|{%xOy;3ZR>*J`Zz z@xn%HE*CyB#feoC4(1nAmDEw?*=cz}mBK|(HIP1$`o|OW*@EmRS5MrhdA+~=@Odv9 zxd%CJhXwc(9QJW6sGrc;bFRN+ryx~8C(Y%?18{#VFoQaRNI(lcg)d5YVKO~#E=*A+L#=ebN1Z61 z*~skGd>ULv4~UKiV$}(rsfHYe;b-*QJkQeg3KXqRzF@m-zXrPwchRP)GquTna3fH| zCPn7^FP*tt+!bEzu1^Qca0gVL6bZV3r#BHiB`j7?-pdNb`v5`^mrc9IPYtyDU z!WGN0sC>b}F^%1a`Ry+oug(3!x}KgFx}4U#$VChAmq(9_Nc2aU!3c~1q$z;3u2s}C zKL-ft<*{fq`RKddQdIFFC#Ka+GiX8SOp-1M9`fSFhdi9N zO6SO@Jg#T#fRM`2Q|9;LI_-Nm_)cF_xY-3GahsP#ie1TR3l~f^#a2>eAG8z!s2F`( zCkfgUkCaf!CB!@;h`{y&-O_-U{)p_f*SXgW>fEqi6-4IN2zE$rdLbhhzy|&3*d+>0 zRz13U>lGKtZL>uKQ?wA+CA3xahM9{1f_Sy5RnQ8sS;nVAj1xi@Z>`nA5E{nxn-_$f z*9#N$^(_xyGDAq4wfoQHp^MYPoouqeb{xg7qR4XOjf2uMfLy*h&$rw0v|t=?c_vUf z?sN1oFQC4(@r>84XIG#(vG0we0u2);cTUOG9who*UYMfRO5@O=ON**1q{1BwxZ9xD zw^8{8jTiGxKn4Q6F0{8f6ugy6c(Y+1o-B|i`{mL0Wz!vofM}@f5q7A}dgUz}kBbfKv_pn-V#&N_Fim zv+Vf&xwP+-!%obn&|)DwKB3q;ikzZia;L8mF7nHimdoDX>yaof^2(6?&y%qf)fw?_ z!PV(SK(%D-p9mSk*my4`tWH!JvX&G^4&BtK8<+zJq{aNEJ;(08?$wOw>>aQTe0be#XYjk<2P-q5aq({!c#09{1RNMj*}cPqf4 z=KZ9({-q9 zl#2t9p9)NKQ_94|#+rG8);@@|VHRGXeJO9+Nq6h4r;$~zPHYQteSFP&Kpoghqi-KMcS4c>P%xQWddd=Xwby)%4^=2Boo_|loHi1B*Yo57vN zqh+q+a~m1~!BY&4uBwa}9Y5A?%*y+PwP4{B33~oSPQ3z?VX(F%MOEUxMvxkhMVGtGMO;5 z#Gpoc+g8!iz#OP4F=*527A@kSKsw-w{i5&6T2uobJ&GK`IxplYxELpa!j7nlkyUQFF(XUd(t%llRM?A#rR%&IoB?kAbIyiF>me%{}wu zHwylbn;B=dU#b2(S>j5zS}bQW#U@cCo{GV4Klw~1c_1s1R)!4F-QL|GyYEKV()Y+S zp4HR;;<22y_34jg?(y8i>*+i1?k^0Hwg#GEQQB3qh~zo3_&I8U*WDBg^a>?Z3FLsQ(5N5}Du_Di9FSk-zT=H&4#S$QkhWEX44ZmfA9+FgrEh;F zTWPJ#IIRWEVG)%utK>Dlse$@p_fD-bOpYZkA(jOzhEmxIX#MFv*UmU?4$O}na=6f${1t)dk7Qv}^I2bs!gptKO|fKnB5hP2DO z-EcgeZe(MFvmM)&*j7V(G(qDSI1SAj^VcVIzDT~CUt4^r?xT^$@ z?iTL`b*t#I?`1HA1|Ux>hsLs7{Uyl4$_T@2u&V!h_(9?I@I%v#WsFrJ10KD~M#$f7 z1Ct9sZ5MBazUV!)X%^(=D2p8M+^IbMD)?`V6JbIZ_B=`UGkrZ@=<Ee2&LasV%XJh>$$52kpgSGzsVk${Vu}<2uS0l!$ltEW8gv;zsBuMxw90jo zyV2WV5xb#`=a%=~=t}nu%-!hpfOw`+w?X^icy~PYiq4w@2p!yp<0+A1<0!I< zidi4FOM~UQs0%mAcgthN6Vumk#FE(pmJv5iz`&V5V^C z|0eaxmd+&;u%IKqf?FPpvSqwK054mQsZ=9ppvlKZ) z#q3gcLzoBD*q~x?ncbMZWqPq7QMfj;8d`evLnrxbp@X6AO0<#F$%yN$Yok{yyl2C3h+Og{#F3mz%=xHanbFvZ$R_aa%FGCcriEo3$T8%?~l z9n>52Oq<3Ubqx##4|25k>2_s2khWuQk*(z8=y;}vcYk1;aK%qw9sOz!{BR9Nt^4ed zwIknY{cao)2J9f5IlWU+KZph@4!n*YT= z))rLYztSsX`ervh_7kqwmW5-5^Z%1Q$TkEy1SC(Lv_f5+LWY7nE*YoDa#xaLA-gwF zY%)cXkg5lHAaGZQUZfZUF$P2Of~o_nfYhr_aR!=`PUP-x+aom#)TcfK`<~?YQaWw= zqvW^W-k>leoasx+%uR5B8Lssq;xcF%g? zvq$}fI)hytbXl7sY-2kF7nnX$8g>X2kAvQ^!Y>s!W={a#@E70=pqz?rG%QtLW_`6Moe?^L_C3K;% zZ#EBsJ2nJ8h7#`&qoHNAF>YzO$M-B_9RPYkEavSNsM|uZX%yK&#XzYvatmUxut~K` zdq3Eiih%X2%b~R<0MIX<(PxCEPe!9D8>Tdqa#C>popV8#sH+!YZ zbUk*RPi2<}n67|QxOs`f!g;$i8-xD5Uf_uRXe=NYAv5fa00bu=i9kKNNd0f5dAs`8 z-Pfn)msRO!xURhkxDA2T2$)YsKc$JUU5p@BRd+H;MXo{m-h;)Gm|I@Z{j1&XZlGEv);1~@PjRX zo8;K?H_Nkx!8;vMPHiuY*;`osW!)EN#b(*_e%a(dlQsP8DJMn^v=1L~yRP;=P z5{90IJnW4Fj`@zPyb}J8<($JD zw^d=g!#ZSfIz2KpV^%&7i-zMtj$zm~OpIaTao& zVg{MCcTC9(Xk(FybCGz!qt2~xUZ?h86#g{zPf1Z>6>}Xl90U?D%~m$qFNqrezW$7V zr)|&I0sH#1{w_Ydw(~kM*6we9+1iKew1yCeb9ulc-+vdW79j~8_(}Xe@x2(-;@t>A z<4n&i&l=$!M>TfmfCvxBa?IeEVSbwG-pcd;`i5EOSi0`JKOzTRNwb9jIYY4@Q{*Em zX32M2ATg{_ml9U2Sxc4&?ABCH8}#2spJWyXtyvXFJWf8@i z9#z^m8G&oGOF?+eoYO`HN=u9K)n-KLPX6Km z$$Vk7Sg2_prp1;}EKp-_$FNtEyo~5;d^^LhD~`Dp7^&6`T77Ipy#jr}NP^xADK6b^ zkY-yZ!E(cNDfsZP(g=ALD#K3zb#iC$Aq}d$4QgP{zE7W?njyBks$)Osj|t(NHy!j2 zk9gj5;n3>6>2*Nc0}L?l6^I8^6P$ryez>+QG3Pos@B}aEHtEz)lc!pT7M)gM;RqHc z3U7O$8{Wu|ihdE)t2@C-k9>jmlu#;&*(#ge%E8J-^A~y3aoFx}{m50d1tg>GbR%c7 z?+7oDyj1t+Zw*d0?=KGQqDaT;B{H^E~>2JdWgsL9dmdHC}jIyF5o*NB1h> zf>t`mvnLvJTvrhb0$7$8sGp<|SV$87l?~nS9lf|RuFF3L56}&&gl-6|= zn@Ev3XbP{rF&pVrn%!Gz%)V|=W5O0XVeN{}p?CGEziou@Q|EBRNp1+cW;aZK-;9ck zd$q3Q)(d0gteG?%FOMkpK1KSe7~Gjdk13~Fo?)kPn1-!l1pm_gm_z9W1I(=N)^ zVki)77ujH_KA$;8@PYvq{!9}UgRsPcApKQYA&pmJXc7}%;%1?^Y*6U5sr*gFkX=S}oCESL8F;@elu;WWRp-po=MvkO$Gq9=cqH$ubXpPC&;3 zB(B3LjM5rgWTE)J6;RJ?Ru}j;c&!CShjzL>z?f!}V*ESvd<(>f0`(1w6ctuw|HejW zJ`tf4^r_@lgIa0^V5Vq&3yJ3kK1GrUWpj|s|N#?+Qh(HiaD&}-fe z;ng(O6ah&x9S<$5`p8|-=b;E*hp8{nZPWMMlsb8%F3YnCX#W?H0l_JCJ>wW1i4S5p z_8EHu*phz+ha|;p#>)yCxjCb7sC*&`?Ph!O< z={PrZ=U&jPVB(oXuj**W;NbCtAO3+KIELTp*w3rNTAO5MaiaY2%?6S%nt|lR`bLoj z1kx!M8a<>^F%9Z^QJF99Bk$FtH%8GUUoO6>dM}U0TqUC>yAPU}zSl}uDca@xwE8k- zn|g)2!@a)za4<|vI2Hy^$+Jw^{lgQ_HldwZNN||YUBIH>sM|fOU5-U#mC|zWis0H; zk1}cE2^0(FRkaZn4#vR9>A6kldA;BAwI&VZv#B_-MB$(qL~^ek;eC-!@_dK4b0id%~K#~7^%k}o2THwdqcNZy04tvPq5QrStOF4 zq8W_;A=My~mD&w_Heza+hy ziJ&s1M|j$;GGyoLre+we&`hHu8tAr220TMkV-34KL+*d!*q$!;eD*rS{l<<$TH(Ax zn<@X}S7w;ilD?lQmXWRetQ;q{e<90%n9f>Cu{$Ynnm}=4xdgqsMiF94GPv%2!4En# z7HgIUwyBT>A#O^sMsM)zi)@5Ue;mLLy*&zpUtCz)6#RLTY@U(frjHFiBEy8i0T2C| zzyTI#Jrov}Wa$?YM=r9KZ_sBX!J83=g-=eaMy*@GS z2x_l=*ZkQFWiu+CJ(XRrAirOsvv#F9Z88Ffb1ZXus|d4;vbB(Ubep~++oI8@2gHgi zr$Lslpffas$(@*5LG#Mkyw>>A+jTXcJ{hmwZ_>*L=3o0C^CqLpQHjYH{5BaU)?wo; z!a#!*`+y=}QZdL?ifSuXpe9Au(E}dnXMGrYH}saGPH<6zOxIPQ=MK7H?s$Q_ev!B{ zG@dz4dc3hfIi3C(isA<#4rBVQ$Ac~?Wvp~J_?1HA__Qf`P*sMP9u34WQH9UZ$o*RE zsF^0hS#kCw^HJ@Mm&TLp;m}(?Ub-}}Ec_FBuWUKA0R%oNw|@D1c-??%hXiX&@*rG* zksyP*CmP~U*+9~~U0e@AqzoZ4UpA>$$xC26#LF>vuR+lp+NJ`wAI4bX1{0+?Zyd<) zoeMvC-b-9tNIsM77suAQtqbT1?S(K7VxfbDj0rqdrfe0p)16SX*c4JNj}NTzEnN9+_p4^|E=h9YR68(- zhS^7%6blrko2Zyp_bap3L;b-6QJttqX?fH`-+beW?8#W0??eB7@@-;AxfQ|FF2<&sF(Q1g(2Iaoe?%BFpaud*%wdTU|KJ|G-I|IIPX{LYDnQ| zz;R;R`h*2=4p1y4`&Ljf$pYwK1r@7Zz{yi6I1sojY^l#>*nv=6hE2+psD#w)rnZ~Q z!g^ILq*{B_YtRK7R8{z3+pC*^DIhj%VCogPhgFMG!rJMhftaj^dUBq$WqO__Q);)| z2K&Kh2vhcf5jT9=HvCNDZidfa^ymLTRygtAlxqQ%jTDdd~cI7Xh?8;&~|`q@JFwQZXrEZ4rb1w>?@#+vtVNQF0y_ zebOVhYGcJWgOf>(Z||H(E{|MNH66a!Wu2l&F1uuDTbYpTEARc@nu=srk`TSQ$`?VWd5d4yC!`}i&1C&`W3SeaHBY4EGFbF`3a zg=|K_6LG*bm0icVccCAy{xEYqV#A4%!yz_OpiV9yN;hu4yj6}RBCVp+vo6bTOELm4 zOD9@t*vwODS2Wl#87nXO-9J`$pxuHmj1+Nttc%9-GrN~X+VI%%5}Ra*jdR)avSYt$ z7#I(Opc6}q8VlX@aRahVemyk0((=8#6h-by>~RRmbwr&~SGt$cCDK@Nx9WO>C%_RKj(jtWZXGEIeM8@+?h^)rwJKRntfzX`=1t#kIwsiu>m zJnW`$wR{iL5;dbND#d0y-<8^$|-Vo8lL;h@9m8nJE6wz`4G9+ zbO$Dz?cnSG(fl{Eay0*|6HAml3yG3Sv5=j!4m>CHm064Y>ecn?PHoap>gE+|D&3QQ zlJ>?W_|z&o=CNC_kcs!P_bNZ_rRIfzr%hz%bKDxM{!GxGNoH&;Ssr{8=

    yv^j10 zKG|Y{jRcBaLy?tK4Ek7&4p-DfFxzb+-DksA&lW@syL`evawB5Xyu(YZi@`?PxK13c z;VStG`(VLew|Lh8^K27ztK6yFJqzf%5FW_URT{8brN3ATX4O1GF0fN+Cj z;mM-Uttol<_{E%fli?unIW_BSRMw1E8VS>T=+5x7QOQwPrO;Cuy2|aGb6dDuicI`{ zv#&~zNE>xXA#+94=27Y1E4u;;t3kcRJHhY5jK;{_YR7FvW5CgHSUC(+@7$~Xg|)lP zX%lNXTx037)$VHpOg|fRRg!z8Lw8PBOLwdC1qJ?iu2z~0>x>ln`D0|P zjOh+;$uWvnL0$7*c<&#pL@Y~iARsuh{;G-zuu@dG5Rru7oK@F<}s3j|> zVHxX!1qodt(;bZd$ zYeG!D8Yj{K;)xe~K*6RtxbZ?Y9gY;5?cku)?;R)0oH&yt+rs8uPqFJLl1Rne3T>C8 zJ%gmA*fOfx*se1#sL^MM0$EAzu_4&>m`xYi4gwyN+Tl`e5R9Uuy!2tadinESfYB@H z)b^8`f*Z32=Il^D@M$Ia-b-EceD_koZ5g!8b;GPwb-QxChXHb3ANt&wef;HtIrkuS z#F)4Ecx*9~t34vfrZ@Wmb6cQ8z+^jQxO5AB1Jgs^=IUci`i^|hV5pZ9ddLq<}R)W93!TChW^BqflnyieOJ zfW8qw$$j(Aw@%K(_H@l5N0knd5Th;)4|Jg17OW;M$@!QYN|X|2VvElC0K*Xg;F^;cFkaeQ@XikF#$Zls>_Q{$VZ~WV6dO;G z)l>}Xp!(x{^N>P%*LMPC>g8c=~&y^O;g<+2m4bY9Mr5SM~~8f#}&`YdZQXZ~ID*e!rte zd5Mup-z?A6SQFwn?bgCUaRW^oCL=+(L5-ZzDXN~mrovPEPGnnUUY z5I4!zBISgBUqFhgla7~GKod;p)CCMB&@1y{=%z4zS!707f!N5gRUEz9Yva`2QiB>R zGxO+0;+3jRYJIXG*{@9v;g~c9lU& zwM_-{d-AVSA0*FOKRrEELzC&=1M#3886UP4OdOVLrT|1|MJ02)-a_EHly`a$y zhQH(|CFiSd-V&f>Xii)XMARF-(@7%CDNj20}V6zPSiwPT6&=vsMFpb?WIeBC&u@3V( ztt86UokkuM_1^ zbjVX{S0;k!z##E;MXLzdN2)@0zlNTZc;Y z2X0h7^UR4Kl{Eg=45S;cCCnr@MysiC;@xGr1^?rh6x&CUUff-h1)UJs!8M0gTA<9N z1>A;Du;v+I&62n= zQ;Ndh+b-T}WV=`-9`I~XW5xm~KLeiizyNqrQo|ra&OR+%*D6YyQtq2W>+2v(Y9ra^ zx%ahlC^gN7OsBK-ZCbxm(83f(4hXUYDXPowFzt!uiwJV_$duO;o`}caFt-2Xn!>o+{ z^yNm~z#Pyrw3z}NaoAQf-WU9tZ|WE?;C}q0WQjO$TKkIMZaQqOA3Low#G&!#hPA1? zBQ~(cVvQt0vt-c4&bE+^7_kYfHqK?w%XndR($7Eo;EvKPKGI+L?SGRLC-&|EMc*(Z zD2HN!hajDb!Q9y*k_WXsgP~{X4v$9Nm&)z3hS2jNiNblBqKv*7~Nv!_gUs9+j1u8de41rT-``oVHRO@tXKzxpJqh?MX{it}!tN7su2OuTJ5>ob4OH`umJefGng-xcOk6V?-3upepVw? z9bKb$dUku-LvD=^L!${T7_$S8=3G9Q@VrMNKZa7$Nvgn>**vw&w@Gzf zW{;KQ_mFK|XMU5~;c9++km#SjQDLpCydd?a9GXok)%E@R^GrV*bUR*O@b`OfHR|5o z@bw+9-+Sv_On1SwmL6gD+}_9mI!}W{OQxikbFW>XF@xKYShEpTql*bUpoeRUKa>cF zbvl&OnmHT^PSx@j)pB8cNYRus*j*ZRHNrOK0a9KD{hJTsvRzRZ;I^719>6(C#I#+x9A1}d)pIBMAblzU;bf_0ZtZ*bx?G#kg z`{z~Bwe(TuxTamXls!hW!QXo)vP`qo;RccQYiqL@vA)s4$8*+Qn*5Ko-4uH!saBB7 z(xKnDxgfjNA!!+14JAO;B0SR&)h@?^pcECJ%Mx6cwy4&U+YU%iuv}o9abmsUSWaX2 zqw8M9U-_EZ;JNqKKfgl`y)XvPRSSdXEX9Ha^f48KmFm^Ji!eLGpPn8 z=R*el*9P~h4n!th76yF9*s|slM z&&w}MP2a&^+Zi*ctU>nlW8R=kKElkk&YN;t9fpJC5BnNi^TrlmyU4l_gSs{l8ggMs z$kg57dZaEK4 zWT);TF&*KQ)pTK`Jwaq#U3(#kH&mW+12>Xp-`4h9=g7Pu3NS}T%{jIsszj5JFdm-0(T!bzW2@S}orw&k3+QG@i0TyioI$dF*}a_;*$&SyKW!?LNdo z|7UW^H_4B=>5(31t1e40kxTn{No-l;#}`lF2X4tQN_2bD^02G|YwyZ)wa0LHTrRQ) z0uD|;BhGWLVOpU!agEPK*5Ehjze?H0nMX_LH?vsF=eokm^4}=`%Z-qA_D79^fkT@eOl?8UaLKgWMAteiy4eEVlhd9*zqE1 zjNow;W**P`iRfF`tR)H`{lke9>o{c1o!LdKDP8WyY-4giX4w^jN3L7cMT5;e=rYvO zblv1G$u8v*#Yxf)Es^6`s96G6uXZrvm=>}zT*>F3VtyJY5+~Mi{pHLX!Ie@xxF@Eh@)C@L6k)DI7A)y|0X>n*4O&&%gIF z?=}Bgv*iL==SoT}gmf0gW>7#A9n&h>FMSYn3)-t~r}wfZsx&CZ`O@czP``_mFrUxY zUm{f@*XF|uw_n=@CS;j!A8BL_>beatEyvD z%qVGk^}_oknV&}D#O{+)3$SNXEL2GwsF;P+BbVhqd31m1MagmXNw$Q!!1M|-1?Au$ ziH+PSsEICle0((U66sK-LBrBc(be!e#{~}`qY9d0lK=eu|JDET7r%Sw5C0`wMzM=2 z5^DwQk%vR~JL>muG+E!Ut_~S#=Qy!?%25N-2(%a`63At`B<8flpsrJ{3n-)tsa6pt z)&fg4^dGIJ^&6yH=_2s$cR{}Y&N-=Wk6e<34`h#A(AX)9?4p~&e;X@4_%hzw8EPV| z?N!A~i%1(WDZNF4u5g1o9?Ypc-#q$uWS0W39iQ+Vhj1K;AQWo_aB+% z&8I&pyH6UNIK{TdLV~nWYzsv$Q!y*VC&_yEe$phbV3*9%-}ExICIvlJ(9L2W;}%lN zEK!`Eax`-1oYUG$_s^i7S-%^M=9P*b*-l8)&Vw!Vq68{m)rmrVLr5Q!?01d6=Xo^n zB4ltMjohy_sP9GZoU>A7Hl48k=zz8dzV&1ENunP*%q4KpBtIYi9%f-84_zL;U5r^0 z2K6CLrBI*pleDjAyk7UMRo{e;LaIZ;Hq}M>jwzqceTqWmKuB2vjN^z4cifN~P5HU6 z{qum0$`97uT!%OecJFJJtcW-i2u4UA_&gfa$eqy+Pt>moX;YO+u*dqX(B=>nI!Vx4d&A5{!ac#q)h()}GgkpM z5Xy}-zk`8$Wbx8=`ixJh<{Hd`&6O+NuZ0eV>f@!Q%w@*3?+smWC@_xA2+LqvMfk;7 z2Z>*S308L^t3q1HP0x0^n#QkQZmnfww5zD6l%| zt8|^BQ*tx%q|ZHan7&5G!N&WsI?cL&*_dAk4p}qyVaqp&-|OJP9Qd1cD@>=2B5*Xf z+$+1O8t}-4#l8O42I`|X*ZwmitDQbZ&ds+Ay}soGe)^Hkuz313c0bDt7L(Gtf4V?p z7DhqoA10C97se7hVxdP>P%LEl7E>|t?B(cwa)iW(^ayYGlq>sXrA&o;E&XM{hM8#r zSLr>nJ+h9dazU48hV1;*BZ4o~2ZfEgEYd?mN8Y2*g7=Qc^@v2T%ILk*R;n6xr3@C} z9uZ*es_EA8Dc9C`uo-607Aw|MkL$Q{v0IIGeE=Viu@hTuV6hLYMp{R)i4=*WVvhO7 zLp6vVf9=(21pc$ll`7z?b!8^51Y(ZR1CXiLO$O9<_D`O^Ct7q^Wkao$c5i7<+P4W zve`mb#8d2QimaewP{lkL1xQ%!Ss>l7O?12uZ`6gf0T83!=jg#3d)J>>FyxQ=^>jn`!!3aR!NeicSD!CK5|*=x{z6|x-q+! z-bUkD$b=v3VXJa7+WPp-zPs{Hxs5id0e%3bnBsnU5a>t{Y8({2_ zs#2J#ZDUO1`1j+tR7p(7_}Ad^!RgN-7se=V5$AQ*FIszG*LcDZ&g)|!9ASt7?7``h0OlSlX1W+9e)0L`dK;=(0=M4QUW70$ZVK8p#j$0?*75a=^bE_Or@qj^EF0 zkXEBBxegXr*XjD#pZ92Y+?3K_+>MN0+#+xz7f53xcClya3}L=3j~?*oa$LsLXv@QA ztw+1xQDeOHl9vL$6@SaRMFbxs$azI0;pl0VP1pG3MVfv>H4@N%^pj@6BV~VRd>}Xi z?Vrp2^o6z^Mo*u|-e-RvZiupu1URj?jUzv!#~V|j^hg$l>AUT8t7x&XUtaFr#x%QE z)2Gz&QuN5SsqWMHvMxxNN@FtV9KjV?hU2we+t=N4$FTQ0ThGT&d^G3Y>$8zK?DDL0 z8vGACEI?_Nv`n(rcdaips3fPwyS#cM|CCQ{r~Y)y8%Bm5K|xpGWqC#PX14*=>G0Fy z%d>8IJIYlt)+KqoK-SOFPRG2K#CZjB@!yxOv2HrUXRC2ye;Y@Onnl!t1M@NkT?(TJ zYLXZF=m!LSkq;bR&97&LtvGos^2YOl2ex>Lz)8QiHi zY@bD%UCDim#p|Nj&nVJK#TfijeNt3sC0PQ@g4P#BVc8!tsrP$)?vVpDpU{p0QkNQi z_INx}7WiRb_E>S5?-jq@TId?nphn#yJ|w}j-m{B-?~bGjnE4BU2ltTXIs@tXYb6VH zA5Xi^6iAZ=t6p9*wMu|Aq^~I2egcYed*|>;O31^u5ozh##DqJcaNu#-TX2coL4}Sa~49UmSQ2b zp@xdV3KkR(;L|G*T$FUtf9qB04{1=c;HfsImd=v3F^5Psus*bdfxc6jC@cy`(ueJu z9pO##e3`x-1k!!M-q)JseF1SZ_1Lkim2Q$32MxL;L8jIH6 z6mO^s!R_b19l}%NLW(7Bh9B7ihNoV`ErM=!rhYHnEN&!j^Rh_sXkAU5*tK`sLa0E;qb4O)1X+Tg%o&I2dsLz zk_B&#V68Bf=ySA{(+b4z8Ou-HvC72UNis!+UWAOWxUdu2R?k%J!Rf%pDy?RANERT! z;;9+q+VUI(AtRigijilWzC&&LarPGy}W$vWZx8P9sGr^kSuv3?b= z`@Zbl_>Q%@`+`IkIFw^74Y}#nrdpyfVDl_2M`@??!S8}3AdujTeFnbLPg$JznAE$M)aO^b~=AW2<|Cd#mH z_~bXBJo~M|m=-(+S6tehV$TG~?r#W5P^Aad(jP832WOCswJ2)8=F(S-Y%Sdsz8@M4G=*o5SL@{mis2f`vp$ap5-+{Pz1NJE{P+6b{^d8; zDKZ?nC8!A?GY`CysAOhqQBF6g-+Qk|mICgE8?$day!3JSli~IJhu`Sn&%B<$^Xi6c zb1xKG`|F-7`-sCUhn>sgflB5;;Ic4-dbihk0>1uc$+ECR;Dx&r(GdmlFtS4u$9^g} zG{d0oo1Ln~eF#%7n&m(aKpau+xO$G(|DbBm*Pmw-R!p6 zYyH$lP0H*taJw2UU=jkNFmEu#`4SW?84vydVY;N74?en2G)vl4XUJ{gpi2W28*wJE(qpR}x@*CO zND_|Ty$YW*KG(blU9Kxyz@4YZxHi&?Gyqp#fwWp)MW;)*OfMF!a36HZ*0!l&a~u3h zIT&EOG(atG{GW}E8a{X%Wjh+fkZlnB@%(?9X)BV?zRu5{%br(o%azeKq00yUa9d?|$8AUoyhzqNacr^N z0u{Lw3(agYsF+=lm51$)fKIi-XTUR2m>70Da*Y=h11Jv*APz8q?*i9UdghONWKKw447m|Z&Jaf=!US>T7Ak7dnO?T-w%hIOcDwy;yWMT&zg^pH zJN-3lce`zo5fv2#1r^i;?TkR!8!!!lSUGm9Qwl`Y?jTK*(JYE>q;aU zGjg>Fs#a+m6hR@DKIX%nt7k*?Q|h#eh^pY-;tqa1=@YJ=yX3xis-S~^Yfe5dXC}t* z5`7;k5AZKTFTyJ7fmegv7}c&ksi>qY7eP%G+ykoS`Sj-syfxi?V-ejARTE>_D~$(Y z&x6kK8G8u5?6qTv{kFTf!+te-rJP>}E7WQJGU&C}Rn5w$TcjKPVN-nD7t0*n$I0t> z4xC1X<00H~#W5-_{i68Hzgvw=@GstclXSddMrOH9)A&P*>7z&w$WaUHpj#;+IG08n zlMoE#X_#$ZF2Ce|PgyF_A;s^Lz>R|840cnv?U33|RUQn{ zeMru!chKk52QSUgn^mBcOLHX(9A==RS$D=m-M zp*2XdAQW^b2yIb`sqIXmpQ?{jRGn&y6%GSdo#x-z8r^rF{YgAz=*8!y&1ND-af?rm_h;7dj*)DB z1&Ci%@@|IRAg9H}Ab3&2JLc(zjTlSNIPhS`jgjXDa15Fak85|>J2kjyIhMef;z(+sS%CF7%8XR~W_k6R)6}Mpk$?a8MvoGX!Nr zr;W1tHHk+1^q!Cw3ok#00YV!1>4za>}2 zSH&3LUxMudHFQ5P&OAGEBWRwVT{bX!{%QA}x!?8=2_9A}QvCV9|AEAD*ox?FtVjyQ zBvE8F@(b1Re+_Wvq3NLJfpU}x2o*ri=dHCLm{K?K%Of5L%|G&#tV99o)S`T#2 zwZMRxMsA4BN=|73m3)+xouIw)Q$#oeBddS$^qGk`xH4Hi&t89qO#qF{%lSvrv(&gs)II!PJDF&S7CAYc5?8mPL#USr^;_Yx#W)AQdV0@-&8_iBt?)o zrJp1N#Zy*zy7F)^Ho6ZecZ(ssR;xiqvL(&3bpgG8u8)F@{5rFk_vdyUBgb*|F}qde z-3^!j)5q#){IrEECaLVUy&Nu+gVgYdh}(9G0Y7C66?+mgDtdJey-kx4d{l##ZE>PJ zjWOyRfeMk+z#6t?K2%r}D8(%cFAI#PZYqtuA}Hi`8B3VM2_HM691iJ{3-4UbuVbtj z@&4lB6|#xL&S;s)7X>L$cV^&phIP=50kv`PaU6 zo~+?8QVML4vV~$cQ!op$$n2!+@htbht1^m9_^_zxwMCj@AXM3+*0l!SjIQLzfoMaa zq);?r%>L}r;gT=@O9x@K)=Yj+|E}Ow?|DaF73}TNYdhv*wW92Vs4WbrQ!6z)7(-Au z-4}L_z9TmTr3mhMSIuY(1G1I1^KXO=Kh_{mR}T0z_?80u8?wk%D+&WmbS}L!9L0mp zqjZHBg39T2%5LQy`Jte6CHgxqsA^*mq367oQBS-SBQLpu6@(`J%Zb2m+Y5#Oo7RYR zeiOx{Q6!a$?NdIAt`*kHp9J>Obx=Lg6jMUi3JQ2#QX^eAyN*6Sft7;z8L*gumwaTd zz07I?xLdWetZD9*x2;Q$heUgTbD|nvl21IdXWr%6*Q8aD2Cfiyg`DNB=QYLLpT25-fha{_xi>+T7`<9` zU|uST3rS)URC@L9srB+YdR>649T*mX9@TO@J?UCJvt-d&87}?en^xQeC(Ql_a_qI) z0JYno>=MO5BCwu{#fp=xa0{<89`%dc!p}oyPOmPfQ}{(PT_$hW+*@Y+Ud zT2WE@%BZQA!|K)v2_$S_O?Z1l=sthFx`vkz>~?xt6|AHfD#i!%v1pEA167y7@F^Z& zaKHkp6_Nk4{Z$X)BOgL9shXkNFr~!{Yn7_zPIT1O@`*b&$t)(vRkyNQJSYEl*Uo>S zt+wXj`vIp&JUd&%-PISGS4TQw>nUa(Mb=WW8^7CXE?})zO=R33kg__~oNcpW9CUO~P*7s~$@^MJz0&W_?5xwEyjnco=1F;Omwr z%8?GeP&DLW5qF*_li-{k7~uoBZH8?JX1A0aJXUkc-n(Fv^y6^c-a$FRpkluw+vk7` z=nmSIdlYBi`CL{RI#Hk*IaN*pXygMEbj&EGwc1BSAMa^r*sJ z&~R${N0^W24&4L?R`!FU2Aah7MFX6zgmzE{P^&QnSVWg$bD71%hIAdOXZ%IBDbB;f zP4n+ySXQu@T=Krhj!>&JnwZ|Zm+WHajB~XTt)Q4g6gfb}mPqbKbq5X5(EJPC zdHAYFSx)sRFAFyE+k_9_ZGrvE3c+UJ;Ii=7cF8Y^s}+{Ji|EAYf|<>MMpdJti1!#` zZPkiaS*5tbtHkd(v>6`=N#UIf?u)WqX`m}zR+#TR+MY8%qn~hq9=4~~)`!oZZpDu_ zDY1%d<8T!ABO65QqL@O8>-KY@+F!cHK!!8$agKVKy{7y_4b zfk(Gpi9DgmMW9zNmp9PqQ*?PiM>^mGb!%#4)P7Z?I>kMjy{88l1RBXF?yAMz(6-rv8Q#`R&5d-2|RQu;z zd=wo*qfxzQ`u@47&IvE^ho#gFXie7ocFPMTo0uaY1~lX`qe?BK(b$)OE$y$t#_=9i+OLJqJ_?O`vyEDw~cZjk(0kX4pr z>>VE-@;Esigh6NC_C*f4ArJgr(*inx{RuTwQ*=Ici`_wn!(A34vz~_+m6iEy`LJk5lbvjgv!G#Y*UFdComM+skU$AGbwv1282`!sn`bjHBd#tyjvxW@=K_j zQzh0F$%hT}(RI<%L3+=oMKwb!z-OBf8{np;mgyx4=nozho5GU+zrs zm}1|Q&L))0UAz;AE$rB{fq2ku4(f%n0p#1N6yK!N=B-ufhIN_UvGh2YJtv`I^-&xQ z;)c_SCh7NnV_jobB__`!-EWXJHpGliC}w~nk06c#bY%^p=(07c>pjzzxI}!UsaUWR zVyWu_`ei2BA37D~^FXBZ{99Q2VYweUuBxG%ch}U9fjn=(yAJxjYlF@HWV3R>8<=0^ z^*%f43}C5-84?c2%(GJ|UKempx(y;6xY=z4hoKRQIqJO|r=c|A7f=KV^l5(ivmQo) z!U8JWUXvO$ZNfN_fd&aWcqECr6NX`Y+&mxiG|4cC0^t&<0ny>x;?_A``Qhvvyc_fT z$=>i|5o?0c2ZquRk_7$7K=g=zMj!tb$b@ip%lr0wWq0fg7M|y}NSObK55y~)9@#R%HJ8`Oz1$`^X?k8hkc;nd>~gJI-D-NnYLsG3ALz(3 z4sVkTHXeB@#UxW?4Ha9fK^<=_NP~<)y7DiSGr%cxZvI_7xlRn{c>YJRV3T{Rqp~Mi zF`>;q@O$#agVg{WHX%UdHKK?miDFh$WF=NnV!2`y15q2K5S~D1W#r_&96clNp15OC z1aF0p{elS41gOq%*Mo|z6jdg)43#y+O4rk`Fs^-sEXWjURutru?y z&8H9155>C6Q!h`|E8%|#Z{q)M__DJ9r0qX>eC_fNnq4tGcfWPKoQ}G0TLE(`vgR&1 z@!HsvP8-~RLNS*pa)F9%gI!dC2-)qECE%rCNDy<1$-Hd-;Yd@&sV^6(YXg?S)~gaE znR5blZT@lLy0zpWB%k`d%Xqo4Upp$h9{wN#r~@??wtfS>3RDi)_zYmDGe#J1D0gVl z2QttJzEGVPbwgR}(=SVg&A0(-^*}`g-*3RDb^1NQ*=gzC-E>^!DrT4OE?@lS;XC&* z+x+xuC%;av`HGD8wQ+RLxv$gKLmhdqx<8u}vEBQaf4Z`Szry>hwqnK#?-a-$IXN-T zeN0Z?<tGznWaiQ9L$-)3BDPEJLz?i}abwJV z*_QB*n2SKAe_vKF+$H^pw^lGfX90-I=rxp3@Dce-V=bBG8t5~(FMFi zKiI!euf8kJ6kZCw9`!+TXa>}`l?B$(2j744J-xb&^eN9nYxQrNV)E&;q(O60wPh;u zoVN-46sttV;GJCXU+9+7&_ zY^l+lI;`<;SN>%ZhfW6RCdL%mybC+KbX)(PZ=}^O{q(}zU8H0*^&jqT0?9d>6|9D0 zpx)vmV06-Ecs4>W@x!1DPhDxy8YR9c4Fc+oR;lGhtMsu@w^~#z!z=Zk?L1v~#9l9a z(Wht%UfQRy@aA#9p(uXSrq zlWaUrwzKQqD6Y-RPuU=a-3!OK(q)M5+w+BS*5h%&09FVIACyu(@m*Am>V|S{lkf?w zaaG^Dw(t?j3}}*V*B%k%d)1N}D80+oemMW8Ut`30=Goa#e@u9AcJOiB=NJ#8Bm<~z zU;oqB>^oXHixCGss^AsBwH!zbu=sJfixEkmnxXn?!f2nK&Y072V|9I(R0Bx?7N zkHWhC7~cBjuSq(G{bHldT9ZpL(1c{b-ep`PbOs4F9$h7sp;-HnBG~a}jenfzqI|$} z7ewnCghqdWimd0YBLb+FJq#>PSm}WE4eqrfPUGbH)x)iTVTF^)UoOd>X5X8}CiueP zf+z=_Y}R^kED3j@D?)oBbEUc_W``^zxFK{G=6y|J|3Y zXlnk(tvs@i!)Q8hgQg=CQ$s=fQ*3&;N!S9696NoId6&p%i_U#L87M0E@T+MWx5vVH34_JvDN;u##Wj;>v}^z_1CjS)po)?$tcKbD8=VC(ElRxqM8Kk$kjkDtup z?UMsP`M-516<2w~Jx0*<_}TmxX`k|N+%I1?t%8;zZrzxR63= zqZ+zmw|agNhK@cFF>)4OnNc60 z!|)$oN(W~L>L&^%eTpvmAy7utCCV?EuUMmnmHbb{SS$09wpN4wkM0vt>P5b)Wg;Yo z*CAc}pDLygK!sh77W<--mm?9q}xC5tn}=(jRIYv$V4xR7j?|7 z4;>1F*|oUSZur@R!o^|ITE5BxPRfboHy$sE36901`T61gm zBcfOD@~sdwMzkw;(|G}STsQX=(-_eh^Z30}KkfNu&9}0?QvKJrzH{S;x?HO7yBA}Q zkqm7Hu{>l5ax27G<7&hb=1yF-^&cp8vTYWXU392&Yi1?I3=MKNA!ACBsFepL(A{){ zypsPU3>nTFz^t2Ox!MxYOs<%IC_;DLFAfB7jp8iHphtP2&hmIWbg*VgPzg6JARj2C zQlsi==&TKDln!}x$qRV6I3NxBO~3n*ha!eNu(aC{)F8+AvPi+>IGAQaww2T0R~ty^lB6!S?haRj87Qo8(Lij*mnaEv2T*p%c~(kY@)AAhx==xE`sQ=-}(-8 z<$}+{%bLMi^Uum4aM#UHNB*BAd+J^XYD0VYsnMr+t;&_*4Z;eqO5T8S#hb~zz2S)h ztWMbhq1v;Wqru6ckCkryCfj@)S$?f;o^|I7R_NevO#idAwc8|CA7{I8;|-F`&b{XD zUq*^;JR&{CK=b5gDz;b97M}5DZuor}{!&3UdCLxOvzHVD&|t7Eo6 zhocVT#?7(q$F7hf5C5PmPnx(e03yS|*{a5xlG zYct#XD5iuWyQ$dQ?_HDbQ=B9H1WFge_54g;T*#Jy<8%XH-~=h;qurn!vZ_cXJSz&W3dWL}XvU6TIY*JEeg2(ZbnIVciw&z`}lcu`Mey65E~&* z{6ttI!GE6})krL79Q|v#)6rGu&$E3;&c+e`zX9L9_PkXNsGv~U%pCi%ECXgzx@K95 zXY;sS1lIQ0Sp&d&;Bh{{$^dXTNBFJppZSTvYMEwbeD63}!QllBRLe)?7B*7M28yIm zvFN-MX?90gVcHX^GeUn5tn9|5dsCa=9-4?$ArPEa8~b*#Dyan@n35pUVfPmBh}UU_Zs@NkLpUHt9{BzhY1 z$mXAv9pJ~!x0J06dE|paXEN!6t0)9~QFTPnM&lNy8w3QFSj3(O6$d@<$d4{u8c*r4 z%=ZXMKz6AdZK0?^o=11fKUa*(xi;px8Zm9IIkcT;Ru+J}zU~$NufL+S`nsD^gFhh~ zIqd81v+;HFCQKUdy zd@)r_A%`y3$3f{^3zVR>36D#&Krsf%HpM_aMS*Rm7ePv&r7k2t+vS)||CLb~VLRP~;F5do*~f_82h< zkIIf{Tc&0CCJ-ZZ)%EeW&$u6nZ1hdDOv0v^IML!>VX2>iz8`r&RxhYm+!Nf9pCS3Y zGLi-*e`);l(RdD85wvY!xaw%=mx5uQUD+r*ulXl66?)_JtKV%O-3re$$Q90V%zUInb^pmi~rBF2B4c^mCR4-@^G$)tzAiR*wYnE-+ zRBGI2_MXnu81eA*d3PN+#=|K5A(I{+Y?ay<$ULt)z+p4vpbTc&)Uv=V-Zf>lOt0M^ zeOXf?DFP{|&XC*6dvrXxNAHHF*-H@4>lAO24GPvxlo$bf(6|dGw&OYm&EGfk4E7su z&IV;1Y^6=+QvXKvc@$~06mpi+bwTFV0c&(OSpqVC>ZDnhh^{lN0YY?*Y83F(tB(jO z1n%{2*<)hFQer!eE`8+~6QdZC$v7!z? z0~a^hC{uBopNuSIc!+#5P3qP_R5F2#Vve?eR|i$b^>mB0iC^fYLnnP#1oW?BCwHLNfBTGX8oPFrqPNTk8Ohz^zNqqf?ZEZi zscnY38NE7FbwF#C5?mLsoL5b^2?vySq`CCTse1Lts!SDhIHW7Fu@`?cy}DTjoyBW? z@iAl6Z2{0g!hjGUz;+t^8`@y|fJM?n9+qDsIh3x03av z`O4h&^~qivUwH?`fRxHsD%KE$v7k!GSbsqmN+1L!*L|wp4|XbDc_8qJAcN%7`LlBA zgkaqkQQiC}s)tcq)D;mJnsU4QGghPJQVTW8V2**){z+8rJS#}N9~@672VR?EiHkPy zsiT-96se(N<3v?&9tJTl1D!{k8HaP};YJ1Qm$Hh0?Vg$fhsaET#ze-_UFiv8A{vo4W?ZFqtpQ+7ghm7rR2fz~8x z6PEj@5p1Vx5%So<7Gh2&iL(;n?T{ryua`s$A)J# zZ|Rq-mXUw58l#}?Hy)BE4tq=B9F3@t=%N@CMcSy?`_uCHHL4pBHQudw0*$$*h?Q^k z(k-fb;W6(60Hi0gYgAYYSfe^chOd-(-6s9grkEPl6IB}j3yA&n3ZB4Srs@5pD-vJh zs~gIG(k5&Tye}+|-l1($ZRcCw%ko?{zhiFvU+wu?Q_K@pCb%~H0*woo{&5H0D(zI{ z@%wOAgOJVQ*>7Asw?_2>M7jxTFSbcfuR2B> zE2vQ|Qr?Q-?f0 zAp<^Vz1Ev=xdGhPrE?RfT#w#3!|4A412pmz9mnw)K{Aeabnelf%^|8ETP@R%Kic|r za)q5`;xI`1ZI;YWDW;Pm?XaN(BQPw=jh(|f+_+bGVNbL%3M~;T)(?5$zTIM8kOa9# zfM;=6*E%~xqg&kLS1&va0>5|UxuV4&0})u}1s%xqQm2_@pT6^m40#lYHU#B?0jgIR z#h3hxM7QSLhh;TEWeBO_Wdsj;JW=I>qJ6cnjz-DwLhWwxW|^7A0$=6=F$ca61PMqk z4ZmW89QFl1RU|da=)XmMPO>^C0lKyuLEddt`c*!Q=uec9rLexf?=HU~NT~0B0f{{ElshT_F(e4)?R8D_Luy_cg*h3ze6>E_1pm)Ib zdB7({phLFYZu*ZnK2pQ2Q+6wlkpEc*b@;#|2bD*#`pE{ARel3X%*8D^t2BTv4I;UtjXb3 z4!AEy*sOIFvz8)>RBSd?p)Q2fzd=&u+vtnEU9Pt^&KL11Xc*e3o82ftkD6&otd4 ziq`LY&$mlZsd=C{!bc*7R8lY0tJkS^ijF`gq+fGc_C#d}LM8EMqd2r;Y{2&vg)TgB z1YXDBi#V#Rv9C~M6H4YV1Rc~a?t%^Zo`pA+A1{0W3{5Bkk~VimWRa>Qx=ivYx>ne( ztclL!-y$V`J7gDp%O)U#Y@Rjt&~Y@`u08+HyT76PrqvP!C(Ql_a*V@1bGwbzxkND+ zC{jHbkdSSyBdhf_61lR_ycdgVgz@X-0@DR!ldi5~{?zfroO)18Ih&Rv`S$Q4ew!~Pph*$OL^FC zW_7~zXVq$;aWM2=RGil;3Kz)pbe14pc~;Vnv_1K=4r%d%L$hKxqwJwFP7`DI{*GQc zwlbr6vpb?%g!Uu`Y%C9l^Aipz588pK9m*d)yC6eVF&$f#i>aMHS9zA&dpA?GKR;>C z!|3^E-FYV9ujebjZM8VJB5UrF6QlV)9NrB1wlHLY#eC*w<0$RQ- z_(&^5VY$EWBLDl-|EYp zS{VAb(QsTV(z4SMvEEFf69Pp^kj4n<_aool9T zRtX=R1$4~qr8iD1^Zj&DchH9+f6z6-PblpHk^82Y z`p{l_MP!;5pD+nm&n|`c%?wC`{^xU&0pZ5!r7x#G9SizL`XXaJe%beg75b@v{A_BT zJ+qSo0;SKse@$AVcm#?J22GOBd2bL*^xEljT$2+~B6%P$r&g1cdF_e`u{_q#fYlVd z>~nMBb#_z0SvSUK(sg3L%jT>b<6z&6%VxYtukEF)W;{UYp?2jLv_WD3eR}j&9F)qr z5DT_Lu|TZtgKmDs3bB)Re)ar^|6yI3-raR{8u|1MvdSjqJU}szC~}{Qwe(+Fe3bLz zlL`=v+2)7gEhDf^$B8->u*Ix_T&|_BvRu7ejQ1|*wSWxjf%yYUlw)m{p%|;q5P+&- z&9d^q8}rTSXg*NPf=7xKQ_ww}uFUjFnraf}(#=zQgfP>{N6-W@vSwMOrbUcAPBpx3 zGxX{P`C4Bb`m{P1!fjZN**vvPh!4W=V%~!J*6~sVc#E#Ys|0va56DjYV2XS76a(Cy zPv4Eew>vQZ)*RHJi5G3132&Fm|Kp&?pDO9sqjQtHZqeW(!C$GZ;EOr+lfz^?yYd|l z2Y;$;oTw6t0ilQ@q&qeH_vrA$VQ3E7H*>((3~-G9q$~5GP6NME_|37@X2gqn-Z}#G zL}2YIrepkC*9lq?dlYBifspaZsU{hYpgTRim~tvsm$c^ooTAi)@gwjbbt>vWbdC5?nKBKQ42~Gle}b$Z*#OdWaUrIm^FwynY$v z7beCz8v_cQn8Wek;&_q!+u+dO*@yNx>*+X%Ihs@#=hR6o|2D={iXV}Vn2RwWss?pc zTLO+wOMvS}O+ECe*GV8nF6pJW&NNE$=;N?N;!}lgW~LkiDu!*u_^18_BREFIC^1J$ z`m^&$wIi`&&^G77!= zxMH`sE3!hpofs`4?0kBSa*g&T2>)Rq8yUU2>Alh0rt7MMn!KB3Mgr~R#Zm3b{ffKd zeSrpvZqql5G>w7-vR0`P(hVhk=R>Q4%lr$;hADbAI(GR`Bv>u8~-|c!H~uAb|Xm z_OWj_6l9wH@1*d&prOd^{0gs&^ES@MBYnV%$Xfa;bW8#5a)IBV0x#^<>;d;1dYGb& z(9PM$FY&uLub4u1v3A%O-A6**mhbGwuj zykZUwSfUYm&T@*`PZ1-!y2BdwIa)LS6FR-+5k(qkx07K9Fgmklzt-aD#)*=7$EEG+ zo~h=2WY5%nGj&*tVQDZ%PG7hg6|J#d;wJEM0w*tS@ya94i(TUlxfV|rqcjHiJokuW zJpJy>lS4hm>fF5ua}q8ar2=myj_j>K%&*X6IFLL{GmIKf#D}6 z=BHx}jN;l%ZZ`hcD*H_mXK~?RLzE@J;;eKf)J$mxL|H(elgw)b2e+TBf;~~X63_M% zJTsBnHqPV2>Xb|jJSMTi$E5PD2d2I1;b4m#m33`Fap`JVz2~0!rAnk#N@5x(pt*A7 z+_C`1$j97%%nBHjJ%96ih0yAZEV{huW3rmVZtPAQ$0L(sHc=#vicMg4i?`3X%=DO>wz(MW2nA_}W z1xMYNs{cfmvkTX9xJZ154V2O-CY2(|RO}PM0pQlEqmL^<$Q*#wCGDT84`>7N#iEJT zrNSw97DTRthjSx+^Si3$S}QD0#{OvoDP;!>4oC6pZD3JLF%Z^2j9VKN3BbhBVf8-$ z*642NR|HW9Gogq9Xe8@`;^rdsfda+#gGpx zkm9vwnL&aoi}`eqXSxzA%JGgI$OvG9exHB5r~v2#Twt-B1<}hxDkJZ34k$0WE9_oE zqP@Ncn=)_?gUEpzNP0L%(17=E(SF4a23c*JWo6SLL;2Qfo-SXqhu$$T)!Vf>FnR>CM^So9tn=HpQR+`yWUghdmU%jr~ZW zm?Vm7LU7{5T$;1b?3Pxnse=pVdOlRWyPZavG&FPS8-3|VB)p3cU0a5q5? zakP=KwOon;N5y~us+yZV1@nPu7BE`~{B_!b!22_6cq_cOPOX~J4z@*?EiI192wpv} zS-EkRf!-T5oJMpr)SN=q$mQedv0Fhk>G;1azWJWjbR=x}+aHjs*Tzw4vBAkXiaAY@ z6I5*Xw0-`%HUj_D2n_VV{I&CqKoF5aK2hp!Dp&LJp&T3;`TJ!MOKn#^UQ`}^IRbJ( z?aCs0FcMc8T@E~Tao%ORMpiF7A_bl@Y#G5}q59ya9}bPtmw3{_?-S}kNKKqiSBR>E zJKr($AukhsrrhUmk&n)&Yc&-DV4?ul2U(u>;=3$RICexmg%*ng=TZ2d!c9N3-w|;( z&gNi`b8pIL(p>&E;EOHgbxDnMj( zs8Es>Zc=QDPFF701KaAk;N$B3Q&K|~v-sW*fBY}}=`!Q+!gpbRkriep{ZZwYHPyN} zHNSKD1CsUzDYdccITW*a2N#}QYeWXnh3@UUeb%9l5$)Sb5X%Zc7 zMzN`IBUC%K11B!d(Ph;&;n;HYEdKbh&jyP(8N;pYW`nbZ+`|$1_b*KCeM&^sn&RGXFS~yCzF|rT3u69M|?XI@c!`;Gye{?$jnrs^)e98K?8@ z=VO?I;YY@Xk>RJ^_w#kZ$l^Qm?Ikc@DYirhdYupFl+QDU7O8FWt2H<5JAFcz zM>c5PMKOgG$){pB%Wj9|193wqki0{_1+a^InjsI1klX`BcbE>lu}u+442mydif70} zFI*+EG&JF@Sbo+F%5rCX+*X0_ERaWCGb|_Ff@pR??&%zSM{323;?G}cBGJ1q|^4m=ZbT)4DZhRtnw&(eIl zlfFT21BP7de2hBpW5d&^L%H|mfd%wfMQjUqzgts$-bv6F&Ap3aC(>3LuZ2uf* zK1sG?y5fFT-%Rd)v3eSZ*4YQm9K@#Dgj;$T0G^V6(61kvsmuQG^%)0Wd{>q~g6W!BRP0-zqg6l#)L%E(;rw>X2D7g+LVms|mq z5o5e2m^t?!C;RQ|7};e0xa)8c2Tkyp<%2*hG&$1`q=$nK2Nz0i5OW@RtLHwDiJag( zG8=r4fS=nBe^v3NPwbw1Du) z#h9iTJoixiSbAnrGVk}pH@n*pVFQzAHV=;63>1&|a&|7Xx~bne`u)A+@EhccjemEV zVop%x7!_OUXQDBhit!}NYOxek9%Z1zIpFiGB^%G7s@-`S^{26^8}GpAdz;U}5Mxww zXoI>6MURllqBJOlKwmdr-9_4gOm1c5gNSSh$u2Dt?O@s=|FV{!07N&{iiBBt;eAR= z`lVgDH9S?abS!f==6SH0l$X31!I`;PyF^0&h233Dou=sCPgK5T>Y-4p|T z7(1!hYGCN-1i#9{NI9f|h8)zI)kC%kWf;1ZI*Y9O5J(s2(^pCHLb#&@|2Z0*t}Hdj zWp&x%*z{+F0Bmh=n*4s$Sg`tH7`X_^FMg9TTxEykA2i+)d0PQf{dj-mIvpx%=7rY=uLcnj*Spt_CWtjU z91Y~+^X$;k_MLC+on}RcX+hXEvW3IwD7QgJ5ye1vMlKb5g7ivnD9=jM0zj+S@17t- zPzSOuXJ=lB`ZT05CM}>B6j=w9X>Vl+F2*#O6Mn-Il=d^>0n_*Bc?RA-VKN4BNwuNMc+>fjS7L~MsrkF;aAt*(F=km0- z=By0g7BZ+PlwjFJ&$Q|VyZK#^NURIKsLJ-qr}u<3#UN=^n{eZ_>*4n!3#5Yz2o^_F zN8@-q$m!^Dqj8*|vxVS@SK~$D#p@@d-mtZfNfIt+q6g6%!vAr~IC*eS<(w zWklEiR*K0$0}{K->kR0oAiJ9}%A$X1R`*6V!&*&)TsPqJIq-!b3D#KoLdGyBPQhcu zcgXGgPV3+Bn!sbhJM&u>SmDw5gXOuToWp*~WgFa{pqOJ6`Iw4DK|JJUMrj7DSKbK? zqR1+#L!En+>Sz3CJ!S%DS~8PjXyGEW_V(A-;=NsUNxN>RS#^J6Sz{c z1IO4VdMr<|1IL~ZwTcLiW=huRiEuucVPX$0Vx^-S34Z2}aNh#!zF6bO4CC zaLc`VZlS16xP1l~vcP4cWujic7U@R+2Hzy6Xxjc+(81w~UQ*{B zmo!V-PxkrT_bwx4ZYBV<`@Gp$7rRH@`qD8BMk&6cc7*^|I znKE_+?3o`gLaiW3NR$ShR^RknJ^NfBrZJ&*(S^D=*c!aHqJjQId^4;`-3UzK5PuAN5T;8Dz)>-4UcfU!(4w9m@V1Q3 zS+?(Y?*@DM4ezi1b^j8Q!QnOQfQ_jspcq)hvXL^QPkA`fqJ3kgz((yX+;D9OJt{>R z5tEq@L$9_737dfX4LRUaeJTSitk_t$uXiV>6I;vk)L}6nj4%~0K4i6sP1^D9@4oc5 z6)PSR?E#X@VXS;?gOxIhf!c;WRO|*77Ux1GE?pXQGZq$t70sM~q9!FND10v&5t@4Ci^n-M7bi{j=V zcSPNN)%k1p@K1T0gl$6fgbk8jLC%6bb8q`y0gCKX-lt|>S+G@;%fB76!wX)|%!Ry@ zfzF&CC(6*QQ#QqH6x^Df_{Mrrecm6E%ri(5mFtviqgs`R-^mqqctbUZ@RaxJH(ZMy zmm$Z+ZV)SrL;W!5Zx6ldJE}hDPsYL?OZ9)V?4~c+7`lq&(;&a#iF8bEU3#AX=Cgsu z^H00)3@d0%+WGKz%k1e9*zBmdn-L)n$Q00{K@SEOFhfYQtS7QpnjUUp^m487aAJ;~ z1(6d&x%8GVQXc=OJ%tk+Sa5hR<$&ez$HJ>an)^_QB#)03?`;N>erj)KHGcbnI@+3bMV zQOsJ3BvP@K_Ec!=zzypSAUV(WahFZ7|M+ZPJ2l5PPfW}URt95ILgLrfzhgy2`K;fs zBt;yKi9ol(h`O4OCO4tv|IaA>9QoaJoWnJ zg`WqiEM z6rB0sUp#_)uCn{!Icv3_RQ>6cz1%xz>D?UY&|`y97Y{X62Yf0+E&CI_Iw@S28r20h z<5RzU`jY?6aJ{;dUgd>kISuk<^Wle^@$3w8M>h}dLG19T?2!Are_LM^XkBoAz30d} zlFcqz$l+3KsJtGrJQ*owFGY4yv0I>!eEc$@b zS$+SDT89!2>&YX zmP2R0xp?(#942+@P2Z-NWpA0F&UjrwjsJjm6O-+e9J+62kFY6bUBG&uYC-<2FXkN5 zj@4F;-3X0tnw~p5Y*kqPWMzb?zq=NZZ(q*BSvWZ;P}%8|Dlqyd^NM8#O*I`C@+7Py zW-*pM-B4bR(yNd8?^oyp8bgZ(E+-XrI5lj+;cy@qUtkB0M>qb>U{9IOS#UTYLBHyM z%fAi+YY%z7OapUe!7hGNOra!MQXt9-@219mz04Y(?hrU91U&<-N2;a zTAxJuSdD|T)5aDjPmP4LZ!pID*x@8#OWy|jSet_&Rq4Eo^DG3w*Wu%ac;+7dX>pF$ zTrzasZ<*+%_A~Kz=)8iW_lesWjbRqX__?xsjMXe~H(+^9@taVY)zKB-6aR{&yg@+d zeuRgZMKQ2h&ZJ`77WN5O&(+mK6&ZFtG|Q0u0OFRYU~J|Q#L5liU9;W@Jn^&A0><18 z;rw)r0V3w_s`KwKwmVrtgu5xRum5ny@9l+s94O-SMfH1UPQ3=ANt>7>f~BY!jAcJ* z0ZV%#H!cK8Xw0JGOuB-~)%+zX7kYH)8Cc(FxUb?g^WJC1QF;Qc@e3Hfr zl&6;6#bp}9zwv7Vl-(aTEVS1_=B!KNK=b&@f@&H&$l@X&l1lMMVC%4)Rfl~TaUq4$ z1Damw0>fAJj+19u6aO96#BCIwbALFYWg_Y*K>UHJ^@3_Yp#wyd(xa zt~^h@)b zZux*xAMh}0i?SDt2}mW=Nzq-92X%$PU^zL?fr90rF1WokynX-It)A^`eSu;?m+}l1yMNwU$OYyCsl9>D_0tuJkaX#W_@*2Sjj+Rb&(uq#U0E@G*W0>G zAHA$c2yH&?%DvGicsVn1Uxr=D6-<-*1kX~XwqKPLy=6WuFi4h<3>rwTX{0Mcb*L=T z26~j(Oaa$XUBsy9rc?R7ek(%vM(_2*yV`_!TeYlBu^38S(7;tJG)M*&nX&<;?h4Zf-Lcrhk7fUOvknU< z3MDP%*(eQiya3+-)AWAQATN}lM`md#svd$~wUa(YINKzYVG8*EtDcnXk4jW#s4V|tl($#VAU4vgc$u;^NxyWvs-M~Dou%#; zC#Z7i$oif227JDJP5E3$8-R z63AvlasZ9cdY{MA3Q)+}6>(PkSZajtGp?CfCWmfOgS?oBU4q7l8mR9t6tzm*h`S)g z9z%93+codiCCr1nrOq(!IjIy_&{Ft2OYBN{94^ zpt40H>EUoAnne?3C*8`c3O)$#nu$IGt=1s4c~=}SY7OiLWr-mVOGU<2Xp0>1*{wJa zRh^e<%i98-W|x^vbNHqKpG+VT!BJ6d-0YxXBP-sPL+gETG)qNCXIO(A``^Ip`vTZ! z%!_n%if5rDe-^g940*IGIlG{>$y={ZvEr}m=PiFEiLVX*@@@S242pp`*9IzfuUDrw zUIZ42)(y}(3u?3$BW7-wMO>g>I6CO=vQf|536@st`2pN?=0BGjBVKi9C!JpbUNa^+ zAjFpr`}NXUIJwZEj@vlAiaG|yDBP&^Pc21CG3;rddLlXEldcO*TC+Ibz zO4I~=z$5-D7N7Nqr`&(&__H4EE6UmO-kiz#)4ywVJbu&pn+|fCo#Vmb>djAW9FMCM z^9e;RQL(*(Op@i7%+ss$Adz778{ppq>8duNap7(n3$L1DFcA2N)X|Us>DaqffA@ju zJ=3=j{P4GWb-nO|Ge0bN7mfzzC(rMQ*-9n;!-KEfT-3R6CtV~fl5OzX#q^V=nB%Gv z$*SNQRgY({;6PwGwB&D=Jqb*cY?j@hS;K3JIU0Oe+833~`{2&PJ>RPN%89T3_Kf-{ z_|`k=XCvy>mfy0UYzlSa3&-MHVz<>A&&!v8%N{R@X!wa0Uf;gv*+;IvL16zeqGsYV zin&FR4k|WHn-`v@O_Xnw>ySn=Be-6O8;G4?-YrKMUryk zU?;un?PA$d^M4(n5wF))5naz)Z9tLSCWI(4E;aeHj!TaUaIpJPCdEqMN?|oLq+Ag6 z`VD#HMBR_FP`?iN)O!|+EZd)C-pJJDUuK>h&GEDOV`a^_8#ak~Kl{hCtZ@79KeQhv zJKq2~laX*cNHJv;XpI9$Pc@{WdpwaTmsF7U+qcFh#Fz8@NKKD#&qQaFHB3ASHxBb!eZT`E)Tc}v9PN$ z^kjINTWiwC19x|_Z2o@mr#I{)HJmN`a}brmbaYj4y0Xe|K$)X$S0d#Hp39{-grZ>d zfcNEq8-BMV_wp*dmV4h4)+_e%CS*++%cQ%)hP`YIRyJ-@_-tyLedf|ZGVo~B77&)$ zJ9oF(98|>^O@q+rkCB?xsDi)_Xa>P}>IL66;l5eB#dzLau~6Za`c7#`t{-ll^y;lM zfgu~_Z?Bu$x5qLqb_>zVKH)K4XE!aJ4d9%rP0g2Ew>&5R`rEB!J-Yx7cRg|f-L@kF z1v@AP9NDc@tbsnQE`Vy#W?7bJL&R|MXv35iFI{Gkfv(kD6<-x6F_;+dq_@r3%Qsga zXE1Jdu@Hn~#rl{4%;F(wm9oNqU;LW*Cl0pv$c&t>Y*O!vi1S)Sr9q}(qfb4JwSqVO zQs$L_-OJI|PG53_^h6#ZiGfR+AU%b9dXOdcvOEW;`enV*X#q=W7WD~pv<1?9U{5xu zys>i!6`(*|d+x~S4vQIi?$)tF$Ieuac=_C0jOoO?Z%qyuPkTRiop$1&y&u~YFz->T zN#-@nj!ydoXkrjJLmpVefSp`xln3NP9<9+P`XfkN=g$HP#F;nbZNkHmxzO-1a#>Cue*f!zQYmz6U|)rkE5@%qW!d(gIe* zoSrf0k<6Tuxf?@IC!Z~to<8X2gZ`<%TW%kebr8+O!XYTB2I)PhD8jmZtR9%?WVQL@ zaPGr&^CvgT@T|U>++DVNfBcWPiFLzM>XCJZtQ<`ji@U!+2nCPa)@-7fG>W8BvBQPC z7$LLLT%sp}T_!jEK;soimM6>}AbOmGvBlpU1#NCSZ!x*kta`quEc8v2=(71H&`Q>=@j0r^_tLAe1iuYp zpwQop5$}%{y5WO-?jsn}C;Qweu0P@Ui#ke$6nE`4FZyS@cvh?O>k|);lMgxEJh|P* zPHdo<6pAEKvFYK+G6u0;&#k@}{2Tbmp--U7p--{nFOcnZ60WYt<4Zk~Erw%3)`t$i|3lqnJ#JY@%XQ1UDC6lQzX<^Xr8h z7FE($7HoqES&!d2)$wS(`V_s*Hz9Z^aI>sG>^wLa69oizC;%|HaHd~0lw)&pdF3}Z z2Uzh@JoJP8B#WJ!%HfPnr44fSQp_%j6hg^ubf*^UIg@$F5MwecoaJgU`BBZe!p+1aET6Y8aI5CO zb)PS6m39X;#SHN4=v+FFE~gK;Jq-VBDp_LV*@JFA`up~uFZjT^++;nfdxPA0ZI+w0 zHm%Kr6!Qf|`l;9&|0gO7fmp4k0@yZm26~xj6)BXYfS)?#(WgveEM3P207<=YIS&+h zQNGtg=m|GLHB7S%YnOEKAjpe-iw_|#zfaumnE}K?_&oM4SIY(!g_2rLRdBJt4qxDN z_!iK@NP6X&G86*FF)a`33MFkqOiSm|Kq>tm=2`+zN*#f!sRB?qBg ztWd$~>eb&sZ%5Y=@ww`V0H++5FePl1Fg4+-x&ddb2RqZ`%!sbJr?@>dNncAdu^;o>dFuj7reWjQ72boWYsy}_;IFPp56m{n9w_>hTc=jd zXyI`7w@CkL1A?F z)^gT05EvfC@pfsHU?Da1R`xY=|f(9 zaI5UYp!k3~`XSk;=%Ev%^XOLThe16qW*v0h1i5yn%Wi@=TQ2$4lD8%=vYMR8r5D;r z1v_tu!;4X~%@^Pd#hj!_9Tf}hlaQj==ijHy1BwN#&%U9pJ&`Cb zj7s;pu+c~7sa+W_y0V}$bggRaC73S!T8wVgh7A$RpT%cA3$<1}os9j{22#onPaNKP z0q^37^4nU9siMeXDmE@8lZVUJDUt|V?_^%Kw3pO_d%BO8%)1k0no~OM7dxne@2y=_ z^R=tr_(fApx|eCr?FA{GN3<=|GG!-3i-C;i&Bcu|x8K~ydr0DamWL;+`uOFbqq1C{ zP4f7evcr)^-o=;-FC4Fs{}5DBK3BB~4YVm@oQt9RK*SO?cDG9|)wlRqLG;rWvY4c@ z0}+SAJCLUy0ix{`1ERrOu=`dI%$dmZY$o|YW~m0>6oCl{rD3<$Xs*9BMwvN0dpsY3 zoiaCn=hJ)r-7U*C9*o#ct4m*A{uw2wtr&Uue!wXb|Jq>00kLnc)PEqA@Zn16emp|+Zdm{UJFOt!N#6WsmF2)G|3mXZ>Rfd<7Qln*|v?Iq~B zo>3=Dke)Z0w=Fs^9D1-TL%VqGAkc>dQ8oTmGcsllh3PCLW6d)3di83&Gi%C_M^<=? z7z82Dl6_D(nk7k};s$AHXP`g2b836m)nA61^rOG9w<4S^)_0&rd)pVaQ(&>cc2sOm z?)iV_Z6r?EnK#9(Q`5&o?akp`jRR`$>fp|IQUvWvq;IM4ii6^oE@B9}8r(~=LW-h~ z(-nfFl1y2Gs(ju>In-M!N|furT<+7Z$k7&uCqZi==sgPB73EYslO@?CY4mOP)vMR| zoYh{G#|vWe3fK+~pz(B7e`&B1MuM|A9 z;Vwk%R;(2ExG%Lm{`ns{Ypy$HXQ(>XPP!9nMN49O?_RQt!)Q5WgO&>F|7Y)A;F`*= z{c(?Y4#~!l7r`U~K8RotM;?aZVWZP&d#9b=-ac<%GwtoY(ucR5cJADC+Uc(ye4wa^ zpr9feKm-I4-&`-qR0U%@;2GI;Gn1$LMgjFPk^TWczf-bjD&5CR z^*v5+hMo;)yQK;IRg>E1o)gAQs^V5Do_Z#DA=%h9z8uG>Dk*sxMGis6=bZBxE!e4Q<((i8T`OTW z+PNT;E)USEvN_w8aH&)Tva@qmM8tBns_n`Fr=C}I(~V4tP=0^G{o(Xml0!jSMWx@W zpeTNpa-IBBr9P&E5x?qLErK{gS==T=az$GtbJ+JJVo1_bB$*)GU}+Qp4w zibaMqT1Am^i#S8n#h?I|}J z9M{i#E|bkKjW-0OuS0@|*_0gE&(o>Mipe_{>~Zf2JLgyU%FyEmb-Ca%X;(L@o7A=R z?yubPD10T?2V2z&VH*Orqsr+WQ?aPZE&zEpqGLeHvp(!`F!6_{_cey@6Q{TZ(kpYdin8EX?j{ay zNkmUc7Zx&Y(sWn*`YK-x9&3DjajRXAmY=^~W=;_DLJ1{VpaDsfFQ0?L*APOPfy?X) zLB61s*9KAsaS^9nPtDRQ+L!{+J4jE$-P%K=_&SjI!#l2keDY!CeL9wN zS5n2)NtVvA1727|W7vt&_9(GDN9KPr`zKS4(~i#{$`+F?!g$&l0((X z4$PatoC%Z&J)%nV?xI(5%LUbbH8XU)gpf3*T8SbjTE+SBlpqu`X;5hJM=PBcgc*C7 za)Ax(fRJZnJ0_?+>&5m5mZ9>|zx1vcOBtpEgUCV|roojqfT)Y4LNK?9xhC4K)G7cl zMUdW^G%pJLZhOMSY=GIJ#_VQ0e$1KQuwyKVSq|)*SV+vmU7Z|e>y)!3ihsqgN|CSJ zro8Q$?z%fHoB1U0F{!7!AwbupF6TUfil?O^)uf7RXnB76k@*ckSk=H>A@S}XKt7j8 zt`Kq-nM0}$mfJ&CjVL?J3_ByiNvpn!RiocskG=n*v#6d7)`>5Y{bbGD!>TxL^d!^I zD<>t%mQBuPVuIqB6h*0UEojlh&S69keIwwiI3~#6hGWzt&uSdmH?};-3|9{KYwq=n z=CTqm$u7195eg)}YxKs;4Jw?fnQ{)&GiS?^m z?<=$NrZ*{&=VQR>vb*` zjladqR}sq!O%rCnCM@(YqN(9mB}+)M1EcAH37T>!IZ!2LP?0N`7T|vEb*&HBr59j2 zqsS30I>}(SfIyLK9@{q~yTltOZAf z^D+1rnfzEzJExpZ6(ZwUtovy}gTk1Zn@V32ln2yHTA^rOgGG$3Jcy|9bNL9P_w5TB zRxtU(p|#_DVn7Nzn1tl~{-8Peg2NgR3v>!SKxR}<-;_0}4O@IV663sIm>05K@rien zI*&%)jh=bwGrKt`WR>Bzik}kH0m-urLSN=m+*B2lJj zF#q)Ig}2`UD~wPvH)j59EFx$Q9K5$6f>tOz;@c+Es#bE(sZzdq_D5N-pOSY5rUcyx zI3Mu9U#rLxt_s2ceJ*Dpu%2n8QTsYsh3tnHmFce8%+6Vf;pG9y!dWM6gBjYl!q#!&R&xIOGj1P-Pdz_4m__K?- z^>A3f#DZc@nM)3n#syvmkjX6YXmM$ib#u}=8gxQddp!X`)hDi+T>gFMY7!0Ha9E%^ zE{2Wekn1N2dknF8pRCNtkTB13u zjP^o;2Fy3e^uQ1<)DBdH+ROe}kF23C>5yYPVJl_Zp}+gIOMIb~v)! zkFBp%pAnw;zZ9tv z^FreNl9gT3V}6zNP0;!N6v7v!)XoLX@)~aP)K=aB|1|!eWBi)sh2jQv*PH7X-CbP% z*5!qb>Qlf~`!FPm-!-Le{(X8s+#2J47836B7rwvhYg$Ful!u%$kn23Bs*`*&f5V)! zu7~J{A!kXSdy^VHo-EaENS8eyaA#_k^F!9v8tiZ~WNKl1$P!5G?9E*9&L54%d2!$r zCJW9>CX_2;E`7P6QQpldo&k2ImhOlA$CLT*n}gK7$Dl z)_Zq(EHR`v*(syRCdgum8f%)jf3WROAw~=Ig9E$MND;gEi35ja>rL#{N0b~&f6A!H z;<=T)R+oq3QtGD1XLECXmh-=Sra`^_>y3*}ey#fTecx^R>BS#uir=^<`rxMQ(n6r^ z=B)R|n)XzBWzZe&^_d6#n$!=)b-e3JEvHy|)U#7y$fQi;BJnoz0lnJ^sh08F9$@15 zpRxH(HbK$ww#pi?SZwonR_qjKBZ_KYt^5aB!45?ZY;ks&peT`&Cr~7gimVo1gbL^C zSr~tbp1ohGIlw6vqy;6(*8J?=xJVWc9Sf^rG4!T!I6m^~oZo~QVX-+|nnp?-*u4R6 zuOZ=;8cO~VMJlPt`?72%9X60l#2IdSRcaJls4_u%h

    7)q zN^v;_J2+q@K;{)9{=Fz$=TMWyb>IZ+!;206)#4{*4)DQ|Hlj)A zfc6#$+Uu#lF@+XcW44Z);Iv^#Ss-SN7+`n%t8cUwn$u3Q3D-IBlE?x{T|92&GlLB{wT0cT0=ft*yoz#vzK#PfGh2)8BOYH zvI{7K>R-D7MW%JULt*8D9$6K0ClCn;(_D)VKG5DqAS=-A~rZ(fNyt*a;*P$02G$_|6)+kdCeH^UHROyUkF$j-RW^$= zIHlp53&DAm%~4=soY;fILPs_}%?ccjJEO0zUez`hU&V1@57j?I*WJ&2m7_rQU$p?m z>J5JHfD?GB~!FuTeZo{UrfmhD3XubhDlKF zSrK{zjDVr^vqg~TZSZ&=xs=l=#abl5HEf?NmaId42P}HAWdpEI9*|%Gdp!>)#iJyf z-W0T7iE1L~FD2SOW#4MsFr2eF!wNPNLZmyilZ>!Y>)zFnr7u`KCDjB<36wmJ zB5PsI%WZW_r5io_TsOMcDC2}rT(5C%*;sJex{jQKI)p zKzo92be(f9A7dmNz7jtbTWRhF)Vc;6){o|Wnz z!#~2iNLTY#a`lAf$}0OO+m03{_JGFFK3Tzp(v+z?U5tk11I>l^NDRBss{=1*fX`rv z8QMb0p?YrvnxS>_91!vV3Ja)G#wE>eWu915D%4FsD$=Xp`E>pN1S-qj*4uWkv$`dpH87fH*ObWmK;ya>;A$8I^X}Mr4f3aO$isuaE zOMwdNZCRVnTs>aPhfY&+kP@q*A}@HY0pi$lQ030!R`dE?w|iU-8$4Go zSnplp-y^H$A!%%*`~c_3v~EtB%YBmL(75`I+ZYNVRU|q0&c{n>{BNNLacg0}?IB)L_y|gCd@bW&Rn;gApr0c|O_io&Yjp za)qCGWYZUB+zjdfvV?VV{Hti|d*f*cw5&hG-;2%6HSrS1 zDZ!k`%V7im78u5EM3jPDQQ7o;b3mjB*;jS~C4M%Y`F{|dvl0}~H3;^&$O;q_-Z*!) z!F-qS%M7)SyUS^5FR;v|Be)n|W2|mp@4bR~B<+;lao0&Ve+4bWsArZ@4w#i#v7;w5 z{4Z`9D=SydU1m;#&c@m}u;)``;`wA!a>&5eayI@u`oI3{by(B0pOZDv`cs8C~wOa-A;J8_hAoB@iumd~&+!%F4>S=Uo z-a0p-k!)q>Av>^T0V&-fk(yje4r&}b4de;NWd6spd`Pp+RON-}AW3qa{ApkT5ZGcW z|B@@P4T<48jO8rj=ZHWo>^_~yDe>%ci|1B>lar~kAMM7d!7>OQJAtL~J97@1$8a6i z1F;a5z2lDVE>v=P>($>{fjy@pRI^QS#y?3`;kSQAK4^E3>jcAYum<*-J_g5Xm*Bv&JS6Q-=}m)@r{l(oVE#ivTmr%Ii8 zr>a?80-V>{@#3e$kzt1o%VTNlE3B|Fp(DhHuEUzn_u=N$g%K z@}X2WJ=(99VEiI2sFOY|z^NTFh1}|3iJ?gd$rUCFpU6%Na``aV-(v6s=;gw{niHhJ z1EUpb&e#DIlGjV{{E`|agd2d5z6s(IosvG84L5q&4bOq$l>$|V~JXup>m{- z*rE2E>hxXnVjzbFhlTQz7ad7d z9sZlYvcR42F;;ghsS{W(`I{ipA^fAT%%hYrh|h=(Yo|sx63a z<0@(i23v|k?zm(2Er#iE$KZBQJ%i}Vq)WrGGPNZjC87gpjmK$?H5xow1c41Ny!Fi| zynil&Fl|Jkf13>L2xR_iRXbH@7Y#UVbN<9D!!3Ib!e^WiF*t~9kTH1Ac>G-Ooudas zjXrIv>=(IY|8QB^4jiBT*u()pM#&-Lzk-Tf8c?dN@W55{S!IlS9J4eaQy3S}B3mQv zlI#ZA(p&S7`Neu3<7UiX9nqw&6!%NpMQNgbrxH$@=%U~=vQ2T8zlC20Z0R4LwW6{|Cr=cU=H9w)3^8zs#=D!RdJnt9jbpns{LlXsEnqL1)WMfHju^r~saBD~%P zIvU8eb%3)*+QBQLW1O4R7re5W{WCg2jdla+;(jXo90qp_(eX;di+u_fUKMpnvp9Q& zV}11I&%VuA0cBlrRcitQz(;e9eq zCfz9kpLvb6jae(ZJ+II=o@g>eZ4B}&V?yV4)mfG1zM$17RnaWa)B{m&y`qfUCaYK6 zhqy7u6xPbu$_u7xvS^fh&!i8|yf4-CL3Mmrm`>d9gd%nop<2~|Qv&G}w9BfwdtBnV znr&0o%^juNY_s>lvY{H)oAKY%mcL`Z1?jNsC<~j4z2a84Y&sWq43H36(>5c@KNVtm z15SnEh2c70zk8cPM*~AUIJ*N*>%Bko)W8aygQABCA$9aIFZ*fEyrLONvYQKacpA$IaXGSaa*Io$?-p2)JciQ{ExE(3ao5pLIk!kWQwLoE zQCE)7$200V)k;IGb{Xm8RJ*N^Bu&;t`xOP(^K_8GzfOKbc3M#%VE87aZNP7fafzWY zv>P@5WB1K|G5iE_b;bj~HW~KZEz1vn%8ld3Pz{O(MJ~V36?Yji2B|0gfxBU)4z-QI z`G6T%aSRp`m>su%);BlC-xvt(&VlX@4uL$!R-MP{8m${$;0(>)~1I}W4}{FZJ`t>}MnG`qMlOf{(w z*dCV8v3@Mz;f&{SSUnc_d$Dn49rw+i^NW9dZ=tfQ~$z?au zIqqLZK%jicrtcw24y@tDR3tJeZue;A88$Ld??kVng!FfJ+}FZOp)sVfA69BrWeYwJ z#69{u?oA5KZ`aIt6i{ebY&!a~{J#MOwTyIZQ7pf-HeV@1lqj*ycC1U1Io)Bm7v?+Y4jW#!AeuE;VW21U#AGgG zIyyS2XH`!`n7(5pKf!@SRYMz_OM4kNcPmRvTroVsVb#h@6d1&Hnv{Uk%6uFLz zV_X^-WPp%nGE-Fto66|rEy+}E_ehf6fFw{1|KmCWPLLQ&)IFco(Z@K{-p>$ZO{;Bi$iIpefNN z8*H{@qx+RP?VK#SUz!k>E7W2S^paF4W;&xtbN=GD)OUox+cN$xre1=2M$pc}+JoJA zK3g!@@L+h@u^LaujU0z{Kj2I^0!_7Ubp_e(z-ywzCU3wVN}fxRY$|d$e}~+lqkfCI zLi*eq0O~hXkPqEKgO`n0``q?Q&%UBLyJNI66srjD)`JXko(oi&);k3MSM! zcWt0 zO{It8tHMfx{}zifIXmcd?+0`S-K6des}*9pQXTJ8=?y_*Xospnz4dkc+^E(n5<^$c zZ<{}0WAWP2;%u0Sl_ip%-@Im#=9}g@Uu+gkjyu&_D(f*!T@8v(x}6ghp=lAc@{kWW zi$?cJAM8z%RV!146#`J#QE2L<_d=TBpc%<){58cu7y@-YL0FU3>UPjSA>?yR1;7@# zw~zVC)zdBbR?i=a;jh{dH#S{s<<*OKyBjUe5306bBpcaT90$e?uuBgK1!qulU?$m4 zMGjIjK9&NTOq=W@&$X`bW1~ANFkSOLpwg>t+@DfBORI>AC=GvK|G!e6#4w z{>c!*>4y@B990&--)WC~vR}P>zf-zvn>5Y4lse$Qhdh~+>YEN8PT9rwfv`;5{NOuD!NxN+|Ug3RBU<5x{$}pmmR{T@s|w2r zE@ZMew@HIK%3n7N#j!O2?4+H{3G+%IBwjum7lrg#;uG0(tc6KZi_Z`VR1 zDeO!`YH`ubosg}KY010!xAX;?H<@$7ey6iu?YvZ99pqDGadJ6JnUksv(LPRtJ{4KV zOAPII%7U|P^BdJgA-#cZGQ4+P&@EB#6g&7fEFkur3DXQ8Hu%^v`G=2RbaRra%9mzv zR=IXbTIqzaty5y%kCRQVyTcBv(m6*2y`&-74g>m}XU}4+pVN!&FR?NwRQ#b`tuH#c zFHR>IWi_gnV{(HbpJ99ffkSB}C~QCKBlO4$6cg`$m=!EDMpQN@_+28KhNA*-;OsZp z-5C-c%%bFw4wptnVz@n*AL{~QTb&?y+~>O1EmL1!ZeTRh?4452YlRC3rmbN#DM2>^ zTJ^QG7{$*QYzG;0n^zu~44akU8Ghz4>B+VqEQ9ASqW{WsHUg#g)yjX66%L$;0Cc@W zK#@qv6DSf#MW%T7ka7XmL2GcJfpGw;CFA+_AMW7>$!Z7;ce5SOS%$zci|vWm_P_g= zAR{cKf7NU!d)Y-~9d}(u&X|Cxf|4JhNGTO**bdCrn~M&}&qDNQ9)#3`xoPK=!(LxS z=r$-X)GUtTgCIS917fY&-kNN0{J4VKEiIB>R6%@Iycl%?@ua3D;6X?&{ZM>Y(h;2I zhXr;}DZGo`PwJ#v1(G&ilC?njz!;}EtMN9x(Rt3wb7@su4jxe<)i_woD== zCXap>C5Pn39Z;LCx;;-*J9p(AH0YXWZoPXvR|n;sbnhe3F*`-myMo%RP_n3cGjOVYh!is^J8D%?Oo+rMwPO z_0kxdYbJ=Vr{uu8cZ!Ovmtfw3A!XFS4Ei+iQ6Lx!IENafWiHX4F@5ClVmYqRc-{-Nr6HSo3_cY+q9syvICNmuq8|C zC1|kl3oiud#7mZB1Q{|{u}bM9&IPZ1Q=vi=Ga6D9>%huc$_o85T*vxqJjebFd+E!+ z!SH7uy!Yn!-!fvVaqi^_=QfO+e3>W=g!b4veNGs>BPyiHU+i1!C4T^G4$8U!**pZhIadcbYiSGSrHMd7kGLY z62)H^mJ*Q(e)d|oGX&+FR`?J8E-3G0xNo(rdRD&ly!&&^@_3`i6jH+@2hXxHtkehp zJ#n*nY{o*`exfic0_p5>L<7Og_)$n0T&~<9FXur$K7SSemgJ6XOeC0=|D@rsupEHl zJF&xARs-O;&gLb~xwqs-E9Cll&tk$|GTMv0&wlqe62)$2ap2Ie z*2KERQ}T5bSwlslF1B8H2U!(Bc;}8gT9t8!kF`4lcE!rDou3sne)nfaNbva^ze_F- z7xQ;qX9+Y$hfXUUl)Q~1H^8L`Y5`H&dP$x0c|i}I!cU37JVso{q8{Y-=}6U*0+|Z2 zE@j+XbUK~Mi3h@!Dsf}@iimrnG8aR{C?(>!0zbDw%&5;b8`jQEa^$Pe_HC0ReIADQ z8s!BZnBiYN9Vz`V1eHy9iIRf4ImI)YfT-gvtf4z;6fDK>sst|lHd%wBo`;{bioB3M zP7_oRpfu@m`cow`@YT`Lo>xWB2dT9NX{aw^qvK(5G6v@(YBmbStm_jn%)}Hs$ub?!Z1*pjwpf@->jAwX3VR z2gFzWHiVq>yee+xq5N&XQ=K4#)9;kwR=r>;B=BZ&?t6eduFC-*{F&<8AuSEpDzG;o zZkuuhVw#sgRqm*$_LWY-farvCzf(WA%DX^%B)rTef6CxA5Vkh_vN2< zN7|Ra(yBA%oP+T%x=<75+Zwp&vHqtQD1k9|Dk@@QKx-h{9h|k|Y|d!?yvNIGw4c|J zjjvg@E5qzWO?a)R@UA&~)Ju{rWr1rdKK$~6EY8aLgFkEO$6gyqy#ix$#iBcq&|l+v zfD<2nC^S{M&aFFQx&P$_hh1+9@g7Lr=d_2GEzII{!(zLZ#@0EityA#cO|DO2mcrRG zR(_mr5p~#bsm%&&BTrs7LtOXrnK5J8$8_MC%ECV8z1OdabRddQ7Pg+VhZHiEbS?*D z!NsC7$h~{YJ;&b*n`@nc{Z9BV$R#|xevW@ubPmc*46Qd#J`%iNxL?>FT5P=k3X{Oa zx#%1}+og&*E2`%11^(4SX0)@B=^)z}L5MNaFBs*gKa>=lHBYZ%V}Bgj--dd|Az_>i zlzcr!VyVc(s&#Tq#KHo{;#s&#Kj0cOdyAKC#+lVHSO*GLUD~Aa{I_5I-b|wdyy9K2 zSd!|%(Yi8|Vctv0^C)1@i`?Pe@;S=%80F3m65?Ce=F* z_8{}=Ebk_D4nL2%PZHc}l&jsZi_Q|fckyZdY5ro!y%N`P<02C5;@((?ff42un~b_? zC#>8UDrD*ZylyUl{F0av3v$V+bRVaH!GHGz8w=`_XWa|V6`oxLJTEJl7MFXXTVx!K z3OgW~4z@i%wG156T3LKn@uO=%e12LMe&Ll`VI6Pr;%?F<>38ZTIh-20-ziz~U~<{i8j>}$b7ryVg5Wsa zFWblM2&`6O&o&zuc29kT*VtB*It(Yf*VkclXWHNX%VzUxR)^)XOEsB^5-52bMb=W0 z(cE}$uCP1+InXrA_z5B95M|SxB8zPXLTg`NYr$%5f3|AJ3ak_EyvyN@W!Kq(T~-Ub z&PV)KLe>)U?P;=tP;V(mRUkX5N)@IBVccamA5y2*c`x@b3Q2G)_pYTeFl)%P!j!0P z&SUb=nHT*} z-PTDr`exHdrDbjxl^P7)8#VwAO}p!9ck^PcIWeF0Y-Khq^~27bC?A!4`@2Sv#QgEd zBXY)pL2}0gBrTNuGDR+8!etI<$Dq`DqUxqRNjC5*AEmkII~H|1TtE?%E!pC`0$7Tx~ z3aNsuut=lDmpR_pezt_G!Sn3!^}`<|-8QdPVzV@I;252SDyBs5R+moSWRfn_PAl;1 zlclQ$=Ep@Gr=RdPacan3P79swqE&nnSg1U~@1lFYc}?^IYWw7YzCfL%5;VT-Yq`b% zt3d?V4?ZT{7-0CURqS)=071n{@ha{m6bukV`(@HP`jH5?gcFFKGPgV+-nm(j zO2-h*R<|VC5fEKN+4BxMA*2(w>ok?X>4hx2O^RM}#!b`bdd7cED4GM5N(YYwcaCp* z;B5w9#EKVt=;Is1Xs~^+Sw&3Qz2pwNfbxH5=IbxI3$&S8sxO~KM%_~+eP)|%KnawF zX+W1&x)2UZJg~VkrAs5^IxXh5_o(SHKR(?9jXP|9yR? zd5FegSJ@Vbi<-rlFV#EazE@{prYeOG?1ks)G;S*oO&%0qt3bf(w4%?ggVsx57K=*5 zGwBZC+e1-F4HgG9DB`)WY62nUz;CNiL77LJ|(vm-cus!b9-?PB_D z047%8nfTq$ALYF07G{?lDs$H_GAMK7-tsQDIOYN={+BqWLEQpOjg4v`k%Kcmf+wQ? z7svc%u%*2XjI{;(UotT^z`=97U%s{I$7eM4`!Yc_$?YqC%}9x4r~ITxmfN~p%`FUkwtz;ckVhb%{MB%}R~DIY*S8Ir8n*EU-ZlkuJ7S;53{>n*OHty8N8_W&aQ zv5^!`B6TJ(six!=6gfggCVJ!EN|Nl3doEwkx{xEv2GOIt5{$p!gaGD!9$>^!S$;#% z4krqI;~`X<7Nmt3NzuZk{5#wy?wv9XYF%FO+dMnQZ?$^`fq>~wRf%U^fL685Im;OX zv8&uSPVN*O^TVTDf?<|1Vr&c``Vy*4PmIm*umX|eI(bV=^Us=ZIXbL|Vqp`pUht`G zr_WJWRHR9zuZfDJeS#Ct252{_w=zfP+?1{2W&$(e(K(r%gZ{urD2?Y9co@!>E7!=! zh|v*{GzJKd(DT?Xvw|@7mtX(&L-SmJHqm~^HKd@4H8d8TOv&L5*hobddg$oEYDYa! zO||ltq)m2ydiD&IH-It$SuM$9Y!WfCZX{VE$+}K$(29BRFTI||spze96B^0Z;VRf1 zc)th4qK9PE=2CKi#7-bMT2RG}X&pbtW@|f--`+b17y+C##?3{2XeMK-{Sx4-jIJ+k|+SZnz z_%P);j{7by5nK-aiP4s5@9>|J#+Rl>vD?JT=qPzBMXphim2|5hFJ!auW?&Q0MwKXa z(#>Hj1y`6l=Uo27kn(`fJdbkwo$#n3_&y{j+?wAzB}3H4H0diBe?2}@xqK{5jRWYYXn|rMPqd;t{Rn*L=@n~Zby+`t$*uO4j zHfppVM|h8}zt__IM3+#M1pM19n$Q8KQmB_hjx)m-X9Mv}opU#* zBk&gU>;E%AjQ~Ve_IEX&M))=Ss$>aCc3}7&Ffm*?lpIP2GN8B`R2H|0GksIM ztAP74ijT(!MI~W3W|mTkuH7Mgr2P)biPk7w7H~=(O!gWM=*}4HcxSn6(I>*O2&-`)k(e)(b1zVa| znan*!lpNB@_fU~FGxoSI2WpQ^oO{7#Gj};(r1L_CJJ+E8?~fO)f4%gpx`j)>e)~83 zzQ2QteJy>_(yvD^`fpSlZuDFOn!%Vs_(eaqo{_9X(g^+qT(&WkEF-Caw`RKe7L>!< z5(|4{nX1p}-oV?^N|0zv6>1eXNDgy}=@Qg0jDw}q4OtXW524a@*=x7RUb-G=T5k}b z0#UA66lbH-_~#7QXd&^OE^TtjG9-ozA{@E&Ns)OdhfPe_aib8HRBoH*?EcC&Wt*(+ zTg&}(zEh-ZQt$psr}E+aMxJ&W7IB`GmeXtIHO;9EI6T&!FYCvXEtstDd2B6tOTYC! zbFMoP;r{+t{~7uFU;b|Kdm>uyL@3Ags9=YE=(776qg^ilgmcnl$i6r#y&>u5RFGVk z{|g7jN*85pV>qh2En5-k_M#JFqdOvaAb5wwJMgkM-(>jHDEW4ZY^5UO z9HhiO4(q8TS)%tk`Tnqd9@oRpPhT~yhU|mm<#VGwdtDDr>*mm)Zl zK4&x;ZUxLuk8Jw#f~3jzUGR=(Fa|w98$W-O?pO_mQizgh(-_+ReeJDn?Y_9~Y})Y7Zfb0T#(QCRBGEn8jD#m86q2-e#>JqoQ-$b+{?z_rvAZvSBTAu#&HcM zv9O1fsnXFmJ?{F}!=9dw*7+7eIVEVypyG9>>=ID*KAhj@3L;p-N78Oto%8{>%;i(* zeJ|W<)GD?xrAoaJMw!cAP8w6ki(@`j#w)c7Y=SyBO>zxV%7Z*qowP^*lQZ7e1Jn*r zT%-7!GM8T0I?#_j{Nq&Ln;xjG(JYSQ4>+Ng?0)4zf7Ft^7p%GIv70}jyZ|Kzb-Z{k z{vMmWviu=E5&9Z%!W!^4*$x#JdK%OOlVtdNJ(8vb;jeLlqy)j82H z3~V49ee;6$D=*La#OpC1cBi3W4V?V@^Eb;A+#cI_hJ7xK*f~QZbY;`aPCw;$n+uVA z*&%k|h?@m*lI!B6nK_)ch%R9E*feuFwwa>3ZZ6-pi!Cc3x;0~L zWmh(BX4xhTvnHFk>*GgD%*zp95<6j`s(`_c;TP?W|aMJ^a>@lbiqBeJ~i+#iTW`8-W0WiDQo*!8a zfa5L--~HZSUNzqfVlyW>a5&CF0w+fJZ!vvt$ZiNK(m*r_`is|t(CFNIHBN}^s98bP zeq}D_-H;_y(*_%l$e`K>2FKpc4=b<_wJ`QXI`myJtLyh7;|wMJ@bw0=-hpSRJti|$ zDka}Wk!0A0(6c!kIXY4B5nGwpT{B&B;sNr!bH=O9HF z+breRy8~Yx>?0V&?GesR@_Yf1QzXf{Jl0B*CbPFvqd?$L+hhGPOWc0-y_Gko8LiI0 z?*6aek!=oabq<+Wojgho+s|24WD}&W+zr$!a{2p}DK7f5A9Sf6aP#?ScyK2O_h7n! zV+J|rU^8fX8;oux)CpCPEa9r4!!V#omst^-Yd3>5W&p4pSZkZJ^RfC~ef-ORE~2pN z@O*(GodFBc2B}7l%*jz4nX?n?zcPYa0v?3epUd)#0c8oLFK)E+(c25XJ~sDrzZ|@O z@x0kIZYjN#Zy=dm>AhcmiX?M2-O_%KZt0q@gZEml+#xTK?$8?=J?CzU;uI+g=>dy! z$=*rWEU_?T9!2`kkWq(S8FyIHzULDHNtT4>%e8oG9x3e{6I|>i0@%&wL%-O5Y&a zbDn)ix6H8PSOGilw-#zD#j!tqw6FtWI#GOV*J}K22Aux(KK{B%vNhCI(Z+ySs(6N; zXdWXIF%)|6X+gNc!y-IQVt9D~5-MfV(SEDklQ<5Wl|aqgu+htiJKci7Yb52R!Ck3| zU$mQ&1Lu=NA*Q4Z?Zfy47Z=-u-kc8ooc+^F`AQuzjTw3JHwSE zI`Cw*+GMi&oRaramWKL-~2L4kN!0L{(5k#-C zIa(Mbe@}PqmSxh{J&yS8q>u2TB65Y@ob;Kgbe=ee$%NPAs*nE875+U88l*$?!;r_y zXwLzsddYPk-SpMaaVOMrV`8wOV_b)K0)+#1eo}e`csdTcHc59Y6NQ%;@Vuk=6;p9G z+UkbE#Mp?&0AOG^0||A7%m@5BMaFFthn=!4 zzbf-G=eC)r2D6DJIWR6QWH_UJMI0Cem$K+W-z|)uC)BVwI}g$}hFK1skv_NUL0CS4 zGu?m_Iyd!_XwR#nHd$4e#?V6y!Jtxu!kCB-&1@Qq9F6rIGNYLsvNW*IZT?AE;$O!? zfak!eR~88Hj(+Q!sBdn)AlWZR9nVentK)qZxDM)M_X&0}kC`0zLMG2gXJ<_t(;;Dn z4bxlg`pG~2^?&g*PHBam1kCoS6n0+8W6}bd z*Qh>vlyi~Z9o8ON;L*y%RBHoGRT{HGG7xCU+{PGm9WTeDNgeO}Ik_sl=XuxTq9EHP zihoXcU6?=RPT)#z8>-Jn*wm<4%jRHWJuG~%DU!CI*r9bcBB%D%%72g*4je_^VS

    ucLlwBpfQYa*=Z%BD>$dnCg- z@7q7W`M!D5ro$32SV+S>2?jt1+7NB9D1PF)dG>LoK2h(y;N?LPbv(GKb zCQcH$$UQDyI^~FO#mq{&SKOdJ`Mn1V8`Wd-ghz|uzh?enFg$cqH{!8j58w5+$r>Eyai$kEfqgUEV^sex0 zdu46tT0XLYWI^6%6u(n|w8%(^+ySBfa!krQ!FyOz~s&BO4%K0Z%dxMKY&XRa8CS|62XXtl%vwo@a(HTV#`9FY#6_|ht!(JnAxnEY8 zpr)kvzG?2WI&93%f){!+tVn=jkg396SIthbW@|_0d0yVgmJ5qvD7=Rf{BBr-)h$rsvtd zk*t=8eQqb50I6p?b&V zlz{qUpEp6qQA%D#!KyCuDEKu;rHS5U(|g5-Rk{2(q?481;90~NeD z;~&087}dVEp_J@&;JK&T1Q^AX9Fm*zsYrd=Jb#6MCJb-9dzYYF8po`ME#rrDv}ccV zTtI=x6&U)}vTDElDcziPVcEV-YBVu~cM>AkkZc1aO*i!A)9| zKjrd*>p^i-vGQhI`OIK&tpbI9CKv}K4EYLnpdb^z9sSS6W@(^d(O^YwS`aRXtGO)# zsH@Et8Ym9O>`4q6i;)klN74B10P!PaUo)r4X0xtwVDHref%ah)GKd}EBnmGO z)#t8h2q_AA>IpuH9CDM!6&WmmF|Nq?_q*%b&9w|2*0F$4*^pSsHcFn1`yP=x8maM{ z#rn`Za+4ZXByF+_f=X$tTe%V{Is?wr-O{QVC9JdsRZMTXmCr5iinz-<=B43azJTj+!F7~`C zs77xX*CI_6f%zkBq;;sZqcuOH^POE4SeWG7zW2T&*jw2OLC&^AL z_9-=n!vh`|2CkQ6(+O_n-Ww;kxZpbLgtNZpxm0uqe%BnIYwhfA7La(kAs* z-+$bnzcJt-L}cM`W571=YF;0w6v!vz0@78OsNcmc3AelB!eZ1--hCSm@UUI6n!BiX zof_tvYb83YYhytdaT&L5{$^o`%VSbYw@Eua4>OxMiJ>d!pL%1ZYlC_h_2{e7i*|nV z)Ei^0vl#YiTRO=S9hSPW`FVEe*qu>#Vl44|$IZAQ7UKC)e2B4&O2P(zHmNf>h^q$m zMRB{VB>W-xCynX?*?v(h_hUd?DV6J!Exj^7{8qhjw?$*7qVD-)U#7CMx(_NbS7r$`5(rH z!~P%pZ!^z5_;S^$1B1mvK0=`f?kFRfMY3`kDH2pMSIOfr{9G*U1zMauP8`$j+V51y zt6~iNgf-+o+3x;uR+IX=s3Y)!XO`-mFpINQ+%s>3+ZYGm;t-7)6&8CQ+Xc&}sIH8= z`9+_Un#HJ=*==0vteKS%hC(v&vz_d9V7K;+iCbGi$&XN^l!{Dax;g8;v7B_Byey!P&IykTPlJ-L zt4Nz8EQiH!R!|p>pVyEpeyMb$yh&Y2o1b4pD(NP*R@F-irCC9Dr#97j49t%MwA)pHvK2ZGtIQ2%^$fUH3Adj~v$NwKVqty}=Q`PWp(S zroyMb7D-=V6QsVV22LU|$G^j@+{AM)D? zDGN!mi@_}}_OU}Y!HZ20V%iBSyiEA%Pr0+r@xo?f;KlQ7di)O2kG%~4>LsmcEPh9k zJfGSr z%gzSqGKGbWK4t$|@^3Ht423aL6dFnkLgRv|vcoBSq=N8$vlw}MFmb6$1*{*Co{ij2 zkEJ^CGCr`6%HYo3$~~ z?5s6bXt1+5&FYR{&olZs@0O{nNZurJ(!`ZGLdk)kx`c`}JZ=z5R~B$G+;C49q~;(u zG1gtv9g?Uz4q@yhP_V019dp-Ix?c6ErH|A4E7v_zVOF?Fr$iKm)Hv(J7*F3e<%Cp| z#{F!59+NCD<6=Zr&q>=sual#uM=F$>k9mmTQQBpc#yUj!d$BR=j+@6Csu4Q9=;h0m zemS!~6L+~ji|^aVLm3e?)C)vW9Svk7aAnA*_$I4D^&}`dxb^E_-oNQwLc;pc| z^U~~J-!YjpTPXQuid;m_r;gxa5t5l}RV{+58Tv}JJMOuB+#T+g^|>XYfGq z1AC?M+&#g)dTQ%RaR)EgeXF=Zp#c_kfnoQ#DD*LGh~MYo-6$vf5xql|AUo&}o)6}4 z_qkOopZsUX!Uiv-aeTYz7? zgUR%;%J#9t*B3||+tJ7XY$B4Mm=no4>~h-z!(A)y%Cx1^(q`7UW`VTJJ;7R8QNV6R zCJYn|CePsfV)nxgSq?0V3^at}Vp>Ks=pzr=*juogo+kY4$IjQL8cmSTAIcV!Ee;&h zD={%DyC^wS4edag&Qvw18O}?3IX$dNTE<5kqlpXXLh3<#KxX)HWeKF=JF&yr-+ZP2x8_OTFVx<& zh171`Me282#jON!=&Yb}-g>XnuvB5c6CUAC()f+Me$cERemjiubGo!_M>>C&b=*7z z$HrzjZgj{}B(X0dky9S9e9k2ShWQLWEyNX}S`~H2HGf;m&lT#$qx6-i`ZW`XH~Y84 z9``Z7JQ`CkwTeB#RWp!mvO%E>UMDvep6sGG`gR6(%Qm}o0kdT*{c!%}1y(I-o-=aB zOV$Hxs4MgQv*-5u^mhv?N%ChO>G-t5~IXke9w)K$> z8NMQ;Ns)f|dIMSi(!}Y24sS@&QYt0iMv-JHGDo%kwYr7x|5dBN8K@}arbjlNiSaWa zto1IOlP^8*Ui%xo4D}n>;tH^j)i1oa(B42{j}YU8GuolI@;=oA=Lz37BINe`s(a+b zOM{RO6NFr%_>aCO2LJ~}?TO4!Hzk3#hcJqNAd8_yY zUz5f_xohxtC9+5z4cZ)jC=^cB%!uRmxpe^j<v=BuNe&cPTPKOC}`;d9u{Oy>eUz zC&{wuicl~dnCGA;rURx89i2`#a8bOW8pHIt{65#H2&6}S7*Z$J+?g6P+pY`dF^&Al z132mytemR}^>b(J`R9tY#JFJ9;5olCARAl8#|4q5Uzz|?hn;krY>$^#fjo!yo|1-+ zE-N4nz0uCcslTDq%r~JOb|c7w@aavDc#@&4b54mkuFz*iH7FVs9ny22X~OHl&Eh`S zLh;J^RvCI*J#1E^J&twrtjxy5uCLY!%!?IYlBJJ@(#B>ndLo&sOpj71jMfw|&GMTb zdpT)L0_haQJ2wk7WiI%!lrB=9bx);r0Iy`f`f;!%>=5NhDR6H{jrU!Px#Pd=)H~jrfqPdjEOJYVW2!HHT zX3r_-^!S!dFLP-REdwbjJ8>V6IF?qSWyD@=-(cyZP0Wj)G21MTp1s09M}_NKWMHWu zf8u@pz*sTHwjV$%dSvC0PPjJj=kJZhA$8nc24cY>O%tUnpPmnbWXufW_{S>WAZC`t?wISVwZmyJ)b;e z7jJc7jBGGrep)e6zH}n#ry^@1n!0!4P1z;=W@R0(Q*c~)BXGa)6xlC|o|MPw0&eU& zp6<0v@h`=&`eBGpdOqNRf0KH- zf1$_u|Bt;lfotka_s2cr3CW8g8-bitpej)s#L8l*fQ`1(PH$)LEO+kGxxF(Zvvj62 z?aWP`Y5Si+Ma2z4!38yd2nd1-xS_HtLbYn3pi#6cvS=+Xph!{S|2#=-Ng$d72{-os zW@rehs}uQ0$ApQ>(y22RyzY zu2qzJTxRyEOvhX48_B!3vAqEbr&Ayp@ zy)ZR!D;>uxr15IhGl450$9nXYi_+f=3xF_gueg5j_)((x!##L;%C`g0d>vbut z0t`!1gn^y6Ky={wQH`TKG7$F7@}sLq?@w;am&cDwejH*pO-oZccaZ{arpbj%-i})Y z_e&}E07dqhNa>OTbm|-OmF&aNLMW+y`%z&%7SxxM?Q{v-PRDz^jqzA4j>{9>a!+9j zeJ^;V@^xyYp3~+9Jc!Pr>%C9Wpu5^E#9Nkx>C|of%|Ol$0U69x0z7^rwTfW(99sg{2tE>dRF=+I{TfRCdkgkzB;F%8juOIP)9&*%?pXBsfyRO4< z=%aUHea!Zz>G?D7kz^N64;EQikgXI8;)ptI3tAME1w;xu-;sm#F(Yuo@0hf|9M)UIgiGwj zbzcBt15{VQHZGv9FwI7fd6&A)vrw`&x?SBY-XkqhR|*!5+ZcRAurn%&Sv)Q^@HTx& zd@r&V$PUlWz8C3SM8TvE7gUeIgG0UwDOfW}Vzwq3!nK!4!Z@gu99Scyqu7lUSx?0z z2XxRSuoJxkuO$-0LSNoxaT|l(b!Bub=x7)VRnfg{l5LrAK;fXe(Ois^>5Gx!_Q3P^X)LZrevIogoNf<{&wLoree7I0>Pv3W zJ^N|Uo*j8iuXmlWBImLDX$nus^`E<%~rQsFx zsxrGo*iI_HF?-^R7K>vY#h#?daVn-lHeZxYFDA!Hb7&&bZuXro`cjcaU!qGVqzm#S z*w?N-PD~9RRg*7>mwRcKd$r1P)yGK}G^rK1(z1J^i>au4_XQ4DizZ?3+ zsCCvuv=zb}P#0Jd*2`#{g{XGc6;Tq^!hBE!EKqGwWH(=MF`!v^MZAw+A?s6gL9tAO zvWw3RnRa1{>o$4l>p5XDc;c9u{$@p2U z_zcgdF#r6+Ch8CGpHZC=Yz=Q>@tmW_HtczA7sOC#dc9+nHv@^056W`qUVmuMsI4ro)Nny`RmI^?TC}L7$c8-|GKqD z;oJn3O5{L;cC2h77XZ6&WFsQvkqs(Kan zAs6}H2rQyo>3K`#N5bYg2TOLLWxxV5#Gz9V#0f29WiKpP>}U2!-aIv?k))0z`z;p! zJc@-{_^q&+0IDhwP!B&kF_&*z?kni^P-oaIoV!6)6g*ewbw*X{G56A=<9a$N62HAO zxsq2)*U_aOs{&Sgm5K4ZbEah2@yxI1#wk<6jaSljW8A+P+ZG*T2FtyWzbi=dH)aDJ zYfWMIi*jdaG;WOb z;x!>i#fmlzg3_Oeb;?|Is?P;!o&*)icKT!iTQ|_dEf5s}lXnNCrGYA{g~9891a9IP zUP2I(C*(}c4?800cSA*R)8C{S6u`ly4oeSP%7Cg_>4aifJWu-<0}DeguZL_^8?dXcpx9E19H3&BdsVBkJGp9tk#Dk3(MqEtYm;m{ zSQ>4zBv-k6G90upZT#5i?W$(_mXEnfwahn#G5SboZgQq((LK;7k|(+jS?g?CPiqgV z69frR4P)$v-sXkZl<^O!jV7l_1}YT3BkF1$4utHs4kLgUHwdl!?6?0An<2DxNleU4+U zh}-0h2rJxviqq904Qt)4Yhbt}uw57{wn}*r9~9=DE`AR)PTgd?eC z7<56bO#rFHDnX7%pBrjJ04vLkSW&ICdX&5t?kE_YA}QNGi|3DKUA4vs7X#wLK8p?H z+yfELR@gf?$znyFzS;}o1CdR#Tz2im)ATM6EmHdT5M;!~eA4d8&>-NXFr?)xJq+ZG z3Up%Iw`xbSz__qwv9Z7yi@MV!b9Vp*dkgeVKq{vDtCP~o3GYXY; z`tI^rAc_?MEz22R4&6!0Cyk1&7+^Wane9{Kd++y#es#ICALdQ>RyQZ7K9W10nMPV>3r-Z6y zznpsRuFz~~=A2n~m@IQ)LjzTl0~VB2id{#MWGcqQHJL7GgAz*cH87~AUC)aH0^4qa za?X#UU!&q@nC6lzI2?+A#|*{^9AjSM@&0`zn;RF7j@a1TV1sP}@0M>KztjVCc;yeL zAUQ#4WNX;PNUVD);#DX*fj|KZb4O3#fh(FuWoH8@`>A?Xg0<+_H)Qi`L!u0A%iHk3 z_;2-0EcNJ>#<6Lzz4?^CI&fteoJOUx4(2*=gq@oe>j5V|CWXG((rT<0@O|5PdtuHH zvGQUbV>3qd)z+J@zG!ViT-G14VH8f2jbYf^xGyR@43of+1D?9W=iDq&BHKe+-d~peB`#o+}!v#@~qka`2O}^|DJ5(w*7VC_y{m& z4DeR+DHh}zvZ+St`YLlaQ7It_H|?kG;nbjqVp@}xzMW84huX6~N2U4=y1+RBOT z;aBBH-TU3p%|U`nV;QeeOj=p#-NHBpZNmVUl|gY%?Bu2Xh@hvvPH7K03k7I<6u0PD zQMM|Dxe;2d?xxT1vP79)sXU;5ib@oe2cavAMV+Tes@f@2Jm~Eo#+(lt!x<+y85HVe z+17vj*=$hWt^3tf^2Ja^CN69?mRS@@KA_kyDbh{FAbA$HFQ+hIB`!+iMP(sNytCAL z8iT4jsGMwtGRMxy4&VmEwW^hF4ctAk3mlId0ejR{K}Lo7I{A5$Gc`RdOO54}pnpr( zi;pSrGVQ|PPUwHeh*R>;fmy_96C9wEX3eKQXIXb4+GyXy>Ubm0Yr^kvad3Y}t#NlVh8zU~@ zw=zHjKRWzwvGWo$i1LP!Bob<)}!HgRtXp97zw|kDX=4PQ%aRvPj7{20Ldgu=6p+ z9;L`(Dh34)D<_vK5gN;c_!5cc~7b}W0n z``z*+W&CEKEv-`3$xT<|76*I9x)dA3KnfhWgYtRB5$Rzju(FK`RK5a_{g~n8uZ>ZK-}?=123Gfv1CEmg<4C&2{H>Lp9)G2M&;IdaIOAJR&)#)?<(DqL^XhwMP^90h8Aomm6$Er) z-)4nH5U`J8?^C1)ED-65X!BeeeMY<}sC3r-P-qZJVGPO(9$kvEfPS|IUZ*@&wp~@C z+#&8$4Ih&JKY}jaJQRu>l=7k>dn5m@}Vv( zDA+TY1?_jMlb1-ZGuULELzjh=@h?bC9n-gH?Ol&luV!JTAd9|FFPn)Rj2Rb6pxRel7t$&;(Z=JwYsO#jfOZ2^-%H@N7~{+s z0WIWVuImHsPRUBiHhL3P^*R#ZRY!CZqcq!T?>iIXco#f$%4LccUa@cOEDW9ES2|@$ zcqi$1`%v3Gy(CJftee?3^`>|^b7*SYG%c>l*ky+Vw#b|n4~r&VfYEDQaX&KNglF*b zl|DeYSIMiE>E%x$_FjB*DbHE~^cxZ=u%Q=<><&8R72i8E5Nuz{uT5V-Hv8U?W1m}@ z>TW3T9SN!tbxorpW0+Suq? z;0wXS(VS#_#&%D%jR|qqY1>QF5P6d^o^J)NooZgLKKQWZ5ZT6Ur|-hCMcCL6sKVG! zv3n?Bk&k&etpqsY-`mG)^~;ddDgR>nj8VoE=9Hga_~;9h0@rS7!%Tyw1`?c2Y+_Va z*dA$ysJSHEIx9nRc~UOFf;4FEyz&0Me_!z3#yO5;TC^Kz_IW=J z!=1G1hidEo6_>RrHab$$CZ`4z`a*pt4Sus5e%eaw<@I!;Q+uL^o&r1PdFY#+c$Skv zq2}B?D6*zHN;PGI>ik}8$S%X|Fj#IR;b1)^xZvCi{*UAJq*_F zVdb8q@(1=$fZeFDe@KU1$q5x>1tI@ml{FvhU`AipjZT3K(mJ~qcE_!L+ zKN5Ig?5Xfmp+%LJrB&73q)xWjA(T+DB4r@H2yx)^q=ktM&XZo2Lp1raG*5(KyPT<)<(SVzz0eJl+tv7! zS7$zWwH(hov zRYrIW^B0D9n_~6U{|qxDL-BX*X0r1egNzdv$S9}SLlg*{Kwk?s?gNc@1MeV^tLAvD z_Ie=Qz~)M|XkD6xX%I2#irAxGJmcz&`(%lC4&8#Rn;(gV z=2VhNHM()KKEyOE*1VqrqP%{$IL~XsVHHVOOsTUhk;RV;<3rF(o8JGwrQ(=De+r#- zHMAb8`YQc81?_6RI2lwwyLpG7U&U_^-{iA$S~}GI%}tPO3p_U~HE?c&MlV05D(CGC zcA_g|n!Dk{WEgYL;g&zY`?|n-GsGp^^Ym?#F!p~hGLw&)<$d6^U>0C>7ajBT*aHa; zGwHBPoBiHdWGzndRPB##2v+RmUl4;Tg;y(`6j8}rfu(j6S9ouOp4@xTw6Awx?|d{&6H9kOA)+|H^dE5U}!2V;5lxE!sMg3yEN5)9# zQeD`GwLzDR{*O9w z$r;h?JE!m32n9}iV=8IvuV+~Ajcsfy)25`#w$dGeP)j|&IAVoI4y_M?twFl`Rbo&= z%tCr>l=w%@i03i^gI^Noo#8YAu4|Wn-_=t5l6i0ZfS>UuDH})5TWl&%QtWYx)KW1G z%0$SS^m^ur9)?1xJ@#IdK@C~I+m?Wms6DE4Q?-ZHIZ<8w4(O#%;6d(IQV!~7-4N~C zL1zaXB|Wn8NniRM5x}OjSkW53lI;M6)hZQU)}YklJ$M~-+Iv1^w$WXRBvJSDj7hDs z?V=*zv!L<<&P1;l?l`|xJi2Q*SsN~ZvT4_!`~_#^aRQXUhc zGv{2C_Pnqe!fwkn8==w~!v8HyyL2}cTOX4Z1?!aa<~d1~&tk5IXWJ}>aLNa6_-Jg% z-u<$9Ohbo-`rF|^juWD&=KZ<9ooGhXL-zPmqH|$HK?r|9mE;bJ1uC>{RE()@1baik zyqJi)*Lvx}1%rN##Q{6SUCf~wNJ@jP9CYa0AB6|MM^J zJ~VTB&Mj^HF6ng}A{p(v`#rMW0xZiYb_qq|sF?QfDkvgorIAq68D_zbNqjOu>=?ue zm)Hi#Z*PnCKJ5i8$b5)}1RZkVJlx5zfYl?7UL0`U?>NbcS{!gti9Lz9>S#@;iX%Sv z>*C|)9Vi_@h!M8w{cevM2{AUCGqqWGoD_hbnZYkkUMV;wOoQHhT)q0;9z^f*Knv6) zyAxqv!xAJ&@Q^3Uq4Qv%;X&`o$gy@gIDZ}%u52S`D4X&2j=O)c?g#pYWDjk04?+JT zb}v2>KF^)9U6l{oj98pITqq1{o{Try4Wi-y#IWu>`1&v0z0I!3PbxQ`A#1o*ueffg zh!k3^MVS-}WC5G0m|oB9ur~f>HJGD1IyN|CQomaZbDu7TXyTR7RvEH-l!X+41Y^Hj zGXI!5RsFB=F!ovkUtrHe`Z<>IY;JG4fWZ zC>{?LO4UL5p@~7&nhttir~E)5!U2|}2BSFnWpdpMqSJNWE7@d4`BW{^xFjo@h2%R<5W6x<^ne94R|1c^KRtS-@u4K(PkS5TEUb#o#M# zuSxnc?I)AWh9>Sk{}m*So1t-GaDn9XfGzb-ip{4;E){d-*E#gH=_YmfB)%>nRhYuu z0-7NpTG;H1GV?i8lLNG=fww{3HBVAbmX7Zs54<7mL7T3|+FoRFtx{d(K}v$|h1|#R zJea}rH1Lc=urzQ$2OVFYD>>n5Moh`8->E0dU3ixajNbz=u#sY6CzL|P6hH+&($KA9 z)1})Z``k_ebr60=U5V|HRisjILbxg5d{BdfC5|391sv!5aYH%uJKMHn2!ED(wMll- zx*BfqlH|hfiH+8vM=fJ-x6|>E9pQYr>{;;=rYi|K>7K5~0I!{k5tLEAC=PaebxSJ zsPDW(Tc?X{WMvDaYxye#DG;6juJn-4WC!SD5XVOH(QeRdNatbpEe*K)dp%D<%|R_) zp(s*~%GB5~6N8*MCp`E;Lvor4*QJfH;?V|p-}ZD;JdV^^yqkw9wt^z1RLlwg z)#Q9oCqY4|fyWJ+>%n?OnwHp`)Gx>KFNrG&CxuB3a2D(CNf#-q>=`wL6J93~62StNmHTc9JRB%1< z4M9<3?{n{rB#h#^?}x20N;(i#=1y*!QVZMTdyy4%vrsQQ5P2wkIY9lOYLUO*vu$Rn z2Y%kI*cPZ$B3(0{9}O+FVS=42b~#_)a^%qWtR}= zVOB4us9N zLa#E`*`H9>fk_$pO859#8U{{GtKQ>>9r3rEa6P1fsspyvRuBA%AMLRcun;_P97kVZ8z~8r2ePMKgENI+s>J+0_L66VxGz>0 zfy@igPy^*nE8P`gqL|GS^?If-$SbQ;j!KQ1ooE=~8Q`I#=YH^bw$U(@-SGO&hkn+T ztlyBTA{%wD$VrTJOnd!w%JeX;asLa%1RxiLnVaJ9cBkZM2hWrF3>O&L9K;D<7=G}= zAFS1=Y{)pJgz5NIDx{7%LGF=1WT=k_{(x*$>m+!t)T5eLtHSt^j(?Qs_z(O}@h?qz zC`b^b%2u)lO?A-R^l7>LxpDGc9xcktKVC5>g~3SFDZkQ?76lFl^8^?4@^?6_BAv@a zw|aqx{^+0y;Pmdg?qdGkf4;B#zhzXAcGi^n16Cyy@m8K>sT@0E%g5_{M=OZ-2!mGQ0#Jw#6yBNu!U`stqt3wnrE8akrjds z_qp3M#lszYB=>PLVPi$7K6=Yq5{ira;kq`^ma;UWTM5tm-RHPd#5@(*3)kM z%*lL=dHxO1kTWAi)B2v4%;#2<>cZ7!X%@IjqSyqAEQdBv*tJ}c8XyHz@Z345DMdZY z@(rEnfMw(+7a%fOg#!=Af#Ouifj{Dq9(r4#iAHTHr%Li^EDi zl6-bgG*P7GN}W?xJb0FdVKW8~;)FBYY=(4x;hh;~oAIkpXBLn`7hYXXSy+KeiY=qa zK`JI)pjYhhNratoEq#jYB&ot=UjK|P{*Kuhb94baAvxPD)CDYgcD{S3G*!5duNSWg zY5ZBK$4B!%%9yh^>QZFdlr6GLufBij)vGfecwc{}ScTPbNKAbo5Ub>Jrz{zVW4+(^ zgZOdt-E{#yp_fB-%C)jnyb+q3VPI(JMQG63kpYVnhNuq$lednf7s7?zRvWz#sGf3k zV(A1_*X||-qy=g!I~bi3KkbZeVn2mS_-0`@NuwW1*UPd3vuJ|`+eI%kxo}kq(=0qV zdk3ASDpafof&!Go9U+ixOt%B@Wn<)#o$gwr^svs)aM|d%u%}`pbAurfV|KT}uhg%S z?h&Kw(oCcC;+&~}%mmqsipLD$l&^R>zdjpjUW?Y{DALK^aiqb* z_H3Q&DU38*EYPa;;oa*#1%A(&mt{bH{?<}rO5#X&3)c>5sqL7`fGL>^|0e4su3R}$~Si%PZy9@bMVaAB^dVz6oi zllv8PpCD&qo~SBFTk3%yw$p`lr!*z3LD}H+n&p8L#|Z`lZ*=mpZ5RwuHZpeSU;eGt zTDW|0xtR-ZT5QOZcQHEvE}#2XGI<{PR6eyOU}rRP7w1#RP*ezYB8#3I-Acu$4=-i& zMQzpse@_)QY{>YbOk0(zJg8gL4kbiJ=8#+wrg75fwBWR0ZJm6#yD1~A-8^aYB%>>f zHEOzm4SvZ$T)l&QKGI#(<7@yYpbT7KjK_}~d1T+&HY`6sx@__^^ZGL9a_$O}$1Qu~ z!cjC({~NHyIY6#CIJJ~DLy(e(+ib~Ptx;!cJgsM-=?V$nyxFH zT?70^trKb$4Vn+MA8J1Wmf%KBJ-r7+m(Hsnja5UhifWJ-`Zj1xXBt6J(sXocwga4+ zk-=&Z&deu91}VIINA1G@_?a0z$7hH9h}65Vb8^c94$TyMkpde17<5sAhX|BQkX7$f z-#$fBBr+8#^&u?~HGLR*H?)F2%-1Qqryr5FhUY55#$-%_(7tl38l@eRB66o>Ou7`N z#gn+nxdgO&=lE5VwaEc3O1uIiI|+~`xaxZ*0E%XfHfdpSC%r3jksp2!?lGFESkd+= zH($WNMCk-ni`?f6^m{tq9u;~=rjpx!w@j}>-WA_-v$PqWT?{g_JuXG<9N5@`)1dP< z@nzc@4dv#PythO8wAWA9$uBdwBZA6#-wt3xqcgqoA%eIesNXGD1RT>qW?$-2C)^US z5Og6?OyP0UYV)~r|6(`@;r=V!@8E`z9X~sfHj)5~3%kfR0xT%Vu|vFh67uWe|52F( z#c4a`#~~+_6zPC9hs&hd45(E>o0Lh__%Q7&4yYN7A7nJqSk#;`Y>3< za_AL)S9$5~8`RyAyZQ0a+J(XClCqF=_jnI2u2IDidUxa;hXn{3`VybZoPp&INSfv> z2-%O7K?{oM#K>VW{DmW_f^=ac0+gKt9O2azyNV(!sF(}l<7h=bLX^^Jejh|no8gsVnB*5K#@hUKmeRZ#jJ?F9C~E-Rf6=; z+GaR7s7?f?-z;_fw-R}JdaojRQp)&v>g@_ypJGLHakx&|r`YeF9@Z)|DDy-soGfQ; zKTZZ252s-S2M#K&tg_}L;erzvHXb$@Oj3Ed>dqN?l2qBpq7ronoljj5pXasGYeF(Y z+QSo{$*0mJt#n~vGlc6-Ps^vAwVYT@D|dic4dA4Y%J_*d3Lk}8&a693mbvgwd7Fi4 zNTt|y6iKFHa@F;!5};XaP}ajLp=}RM;60G#&}VsA0o$O=rptlTt<{OH!{E8$INgH> zaMBrW`k!G{7K3wP>rrk2r9z4Y;=t`x4ARG&IzShQa;52lk|-#8 znS49A*td3;Nfc*~xIAdTbPI%R-|iv!bqnu;M+##)4JR9Ty6B?dB_f0J?G(l|9MUN{ zyz7`F!UionU!QO+)VjMT{K;A_^r;H)+t4;OHt|7R3!}3@@Ei`Ue5rwW*epaIq;|UX z|Fs4ql5ATTN#;NEt*wd6@}t?X51rz4={EXYaG%@i7yI0A4A|{Y-!JLe)NvnLwTv7+Mj&x z!tQJQ*n#1WUW&a(kvpih2EM84G>x3ZnO;!f*y~yT%t|lJz+4Eq9bq~#>24^Vs26tv zg-K~-v09rT!8+X*(7ivQtbluTN~~JhFu5OkF%o#t7}Ua?n|0D(r^Himbq0TFB!07! z1!9XF`hh2o)$jJDY*BP|M5pIY{wLxS%6vYOQiFy;Yj}2;_J%xXYJyL(50XEch*bLB z8hH8qPFfdG2SnE8@VCP2acxYOXj|#K(yoY1esM$!Q|DLbR~@tq`jQT+N@taNAV1&@ zAPP0P;f^s**61AZ=f^BwLx68g@ZaYB&bqCHDc*+ER)IGRJ*X zOd8b9mP8qVhAWqMOs!KF_!O(t=wemel!K~OAdGTbP$tGg2A#59aS%GoZb}z`AM^!i z(_Rie5ZKEcoN`mTlD+;+srOF)wdplUe5&H1|Af}Oh5FjqwL852n-iamd0klfDStR*Dx)jAt=>i-P-_iJ9h62b0J^~QywX|yZU^04lz5R&xs;hN(k78IS|^MLvC$@`Q&<3m zlPQd;(4mXJd*YqRsE)kcf4v`SNIWWmsPk?XcF4OTuoyIjLA?i^vO*EZ;(eUZ=NMdB zqRiGOE|Bxkzq{erW?U6`%y^Dma^XE#p9NBGQEUfAKBrOwj+q@PY1P2-9I+^N_%t<+8{X+p@qCk9`;{bfowEE zWD8u6&JP5c?X_;Go!3Xdy?M>LCcSj`r;sexiaspupX zvy-3BOOnI|r@V|+yvdWY=?Vzs>>(Ax7Eoh}`?gLA(Zwhv;n+r}yo^E%eQrAxB|g3U z*yzRnjUj1)22Bo~E-6$b2jJCEQY?lIm~e&n)|VImq=i>QcX}oV;FYODgC-Y#VWZrytyeC_o0q*;vpx9{6V6_iwk1WLS|Ktie+fe)K_;JB+m|>Mp z^?sL}bm8Ug3kwK+MzKv4F;Fqb6c}WLN^fbp2dpzWQ_Ck^07>-|;LhA5#bRx$>^Qj~ zZiJ%kI(e^Gde~LoU5~B^JkvqvdBjrR$}2 z)dJGXUn05=w>}wLqXi}_7(NkOw!t*iVpjA!--f51_p(mjB@xOfx~xdGx6)!$!y!EAAw{^eRW+5L^#sKLH$Kn(E+#a2?JjEY&wcF@U^TxFi* zmJmy}jw#CM%E`IP;)vK8tssq@3XS^tR1UpEQSV(4o+#MD%Yv4LRwyh_VY1X^bT-{C zL<>_4#N9ozW-rXKLHPx`))>Eo8orub z^4&^bl-?(J)qAV>Uod|G=FM>2(it?Dq-Vu7^hxRcky4r2BOPKd4lJ)5Au71#xG2ry z6n|?f2pjAQrtNG;XeGR?NJO-UR|qag8huQ$QIsF9O^zx9rj{0_Ca4EUzM!_dNrr{q z_N?^| z;pI|f?81RX8*;*?L`BierQ^*)5m;$jp|18G5Z6gzx|lf6YeFa#!(>O3tZIVuyBCJL z7CjNFBfOyOpWk}zX|L1RL%Kj#Cd0c0Y+4xjEo~I>{7)umnhnY&3F>>7Rwq|?|{HRv1$f5O+0>R%%0p*)}Bmqg-o~wNB961L` zkLpPf->#6QNz#G6M5kOCX8b)w;1P10SQZ!oaAzLKk0VDcXmNOIu{94rv@n&C z+7yqB3FRanJ0^w;r@^+=k!K!PWH{6Y>+i2E*IM(DaIsG=95S-ORI+3C73PppK@p_Q zgpjY;1dnIYLGON_b$u6KM`uJ5(*PWm%xwLfr$fIx2|$y8l` zW;d_V{g(S7?^;+-u&Lqy^rUlg&8fe-|K#^?o44A<6Hh8gy9;l%7g>~{-lNz%6uC{s ze9-5Hyz!VHZ}w_|Ky3}d+&PAe>0ZwrK2QyMVoq$dsSfm*LJy(gE$SpV)9?0SpIZ&o z3T|XtWk{)pe7@RZy1=_MB-R^fefh=WT7}khF-C2)C6d+&iGmKg8ScIzPY=W6uGdY% zzaQLH+>z;pE8TYlH2UT8vBNWk!SdO=c7J1 z{OMT#1J7$$cO0_La zc6gn9{hWF8DJVfy7V?@{hc{qZuQ9{Nll4fB{TyQOmZ{9PX=>U(lSuYBa>U}%?xk30 zy52>_6e>=s@h!S8IO<&qQPN(fWri-`vnfe}BZ6M$C_Mdhvy;EG*i#qK>6@!g^}&4M zB`IFH&1+}$p}<^T&a*ee3umu;?y|UL#%EJHefP@FLD6?I5anU`$H6Pg@SnK#yXV*^ z&gl(yT?g{HF@KtGP1VQ6w771-3fR^LrXjK^7Feg#O(fIWEYS)+Cb6`b&qQt6qQFGh zS9i!WrTuQ(MO)aMsb&1?h%9i6S9tIDG$?im@+3`6p2(TxBu0Ra1n-VTxaDZ8n|`&* zx(~u-9b_9l4@d+1cI*sXQ_|?TUiNwtE8945fowMw?AGC8MPF8JfobT%|c$!z8 zl2^Y|PnM4(xfXBTMv7fe!9O-;71QIn3drcLO^1^33m!YdjYM=9V>`?@C`n3rRwHUShLS^%#f5&TW1}e_$l3@h{LZcUS0?%>V)`lJN15p=IUFYy|nD*W&d^T zmDV|pnv0;6l|^4+j?=hSE*W<{xQWf>CrB`0(x$2AC5pB5Mne2f)G@u~^{bE~ z0E!zj)ZIEIj_V*R=6JX{0BbPBhi_=UZ!OT~vIfY8KBdXOEz*mEt z+g|Tmj+#6UkJ`?8<}jjTF6Cra#@=G*23q?nE{h5qeo3sT{25?tkax%vr)q&<^9qzW zU=8M|?XR*&rz3k-W%<~Ufk zWC^5yI_0Os=KGg=q)h4vT<)jUyLa(dLIdM2k9M_DIcL1~bI-;oY}HO-nq;w}&pn~c z<+T5G*c{!bKNfaIItxhlqUACE$k)(w;1MUZjQw=df~nRfgiDOoh0`}S49<4dj&M-c zN|uyF?V75^{cB1)DqJ;^m zqu7%aIZnl3rEf<#gi(+AtPa%XNOtlYl!=0LWtDU}@8cj7IWGkEVnyje3A_`+7UiL5 zPXV_ca`r+k9B{W{Tdm$5Z$R~F#0TO8>)x<0Ha%=5s|9g;$$Fss&EaWt)t8wTsP)g8 z8XLSZAU3#JsO@9)f*SYv5S>8^dTa)`EL}SV2jJ$c+KrOOSQn_|ew?6*+>t;(4v}oGDlB3*$B`%!yZLzQ`rz!SRis-2r>`B!Ndpu7{ z^-#u*x=9G`?6As-+FV{bDA)IR#*0o$aREb$JH7jten$kTYVtuDy&(t-ObqM}alhLY z-z;^qIth}+cz+rl4@MRlfX&iM*sNA8PLp(ijuK>&Rr#_jzOCUvUIXLzzPdI7IqQ=m z?$Zb0&sI8D+8>Sunz2AOlJ^o9L&xF$c=VnO|IZ=b16!wFD0oHitQlC}-~Q|0lT9w{ z)&RHJfJVi9iiPB5HbTWz|CYj-8b2Uo*(5_{lCK&+QW$Jpg!A3gv41fkL?_!0-H4#C zb=~^_eFy0rJ0qd{3F&9BRnZAvuY&+!YaWMvo|}RE*0H^!@0;QB(&4UuC&#$iAr}tX zUbO(%If^|)kp?QJ!#{auE-y9kkT8dLYew?SmKpbEBzkA?kr1th-WXuete0I6?WEhF z335Jlg{g)L{}W`Re;2b$WzcL4*gze6ZI=oQd+#c~RAkZ35boU#E&5&jL&E)&a1n2r zk;A(Lq!C!HxyilW8}HmhS{VEV2=Rt3^2V$7PulE@9r{Pr2UVXlRgyIkAQ}i{GN)y$ z_@6VyT&lqx1VuaTll{n?{BI?H6q>;`=P8vvKsJzO;hH%sB|7C@&qAiie-FrsHiT$D z|4ypUY0{-gcfU&dXJqiXEFShh!lAC@INd`H;goaSEYI@U@4aFj!h5Q6&^E$)CSJ`# z-m;kql2)1N(~j`&2yigKHA$6S;O~qsp|?W6{eu_sB$d1t=5zm@(HQX4e!{yPngTUC z^%GiUPL-8L2V<*<4SkkznV?Sk55qFCZ7&PJLJM%@okIM=nBPNqR0i* zAFWdL@UkF&qQyU;-VrFq4hb(xH>>*HfOs(?0hR;|`re6X2uU#N{uS^$z;{jeEapSk zz{<#65ppwBFdw@Eo0_s~_v!n*9b#>+ z2pYttc~gNc4`h7G#9IQk1Q@R~cz4j38LY}&4s?1zD+_d!dY#D?Rr0WQGv4Fz zQJ#^aY!K|YoOldGZKLdO4Z^iQHKXj;g}><~buR2Y-?Tu~Ws3cbB284xA^`^W8k9&m zv{ju8#8VjlMF=60I5@`VX5~Xe2X4S1(x6kXhm>^2q#WLz$(6h;wH~@qawhb8c1CKq zhc(mbVc?MXqme>bog&a-a#fzoLk7lnBaFI$La5F=QP4vU1?>uGP~HFuN*E1@u&@>f z-EpKJ0CcxNXfR!}D&R{$NRUoV4oCzAm|c-M6O6p+7{^4qy~n%h|C@Mv44fWgwT38@ z7`x?**Wb48lYOezJsaJyxVoXv`Wnezj|HMllVLY@Lyp4AC18Xw;R$;=_%wQM634Qd z$IY-~zq4(B-jDjpCF^J*mqe8duYWdzp0^0{nqyICf@G^2X@rcL3JXNBB4iI5MYP}? z+izhUi-;$UCVEzI)OsWJ4Z(#CwK;Qj`g~EI1j;#~3sSq>t6B~89;#A-F`wB?o8a>0B1Ke8CV!(SkDosq zdQ)OW`4PRqjCntN>#Pi~JV`wbOwM2;U=xz1&YiLaIv@MpFl>jEK|rtqef8L5R6+0X zneX1l&xeNT*l3_=F|dj5o%B@*_vfl>NQJr>C|Vr1#MlW2dv`GogNAK?V+j9r%xjse zvY#_U$4#Q%N3ys@gk9JNIcx#6Jrr9+k)4=s=!!5U9I)6*+x>j&g!3dD%Cb;;aADxv zNNkV08EkDwoh6-e_mo}F!A@#fWD2CE_A#5}+HzhdR6QIDET2>+-WXP^x*S?2E>m?v zmI}Xiq-Am}${Yr=G4k@tFWk+zsd=H|A0&<&Zd^B4OF$cYK$ z=#aXTRM2{1xp)V}((u~1QCF^sdEl*=*GkWe52+n#BfgqF`%&?jRq3nI9Cu)QN%~vb z@5RrU@iOI|(z#?kH#6eG%gt^C9aH4;GtM0v_u>AnEIw5PmH$j zx$``r4o8W?CfVAsJt}QyWU<)b*BEu$`_6{>X4mu& zzRl@pY)}7C+Zun#r!PM38pm1ccHyxAL%=p*;%*PusXx$Gzm&i`NfHH1<$G9zW`(Gn zbkMQiMXu6Juf3zgEyp2f`jwC2bZp=caKZ<5IO2`he`q!t@3z0&Mn2)@cerpN)aMok z<~+rorO0V2rkTV(TcKDqH9w-0zWwcs(tloIPC>0xHh+(_0y4p?0*;d{GNg0Y;>War zRRPPM|4bS$s-gFgI4ESoVx?9{+$D;tWm#UILe~J^0K%f8RRM7VV5x#{^%IVhKDXnf zZDtOQi0vVX(Z|U-wuU~(ht7f`-+DS-pf&zYj6aEc@9iK#c$_3nfnq;YH>m}x7#Ny8 zXIF)dUB>Dw96IR04mkHoW^DVv)q-To59iCEo4`YxK4I@vY zz^!m26e6|<0#leZlUk%-UvQaO0>$Esg0g5FK#L+IWpt`69x_GAf*krRuVn^gi;yb* zwy;^K3n-eJE@)8Ji$@xdu^U8OjmijtHF?2LI-Yi6qr53);(O%UH0@FMR#_UI9i|=W zYz^z7+B3z&8hC^^aq?uwZhoe}F5uDTr~l`BNxBPfe#$I-lU)=Ge583)%nH$7kNRi7 zWD2}ffJin!q5z7qHw5X0JH0k~6;H%6>`kiuK_yf=RYEOb7czyC9X?;mlOzVsvEXCD z1>VT`lH;2p$qvsA@AWy%&lhwtPJxs?ka?0p7$G`_yqyzdsN3JWF0`&m;<6CAaQN0n zwNjJpV|VR27PTg@z7k6)vD+RS?=a1CTTmv}%X{r|Ifs1y+)r{t68G_kdxs(B+pDdK z*|}iDg*QevXxu-6YRO&v7Uo`HrWYugD~f`3$|C>sJf!Ej6Hx};YZrq$N#(>sNwH?* zOHhg{I6rbBD|@Dg8$?WI$DVOVdd-^3H^*3SZk{%ZmJ`7pbm43)%0CoT4*_?mTFr#I z=~&S`r>q-~d*t>`^W(;F#&?|7lrcT-2OF&O!8Q`9NIst$cpC`g+v#}CF<>YF`g%;h zcQGl>S+*bZcFK*}7VO;qFFURC!CaOT7j`vlWQy1E=8F~y770+ZN{a>2 zs5*rLyboR|_w8kn1qyqYw~N|gBa3ax{ccb(uZEX^sUZ#Cr4wL7L~GGAxe%gv8sDki zu($2aKIWww{#SoG@b^E<&AXXn-~HZaWc4^8(I1$S-$JoaC%B1pb|HdSQhkZ;D=lXfKC2OQyVk8GH^lVJGYpt<>yF)NS| z(&bY$Gn20uCh%f`qTv+uf1MKLc=WkFI@jZQi^emi^E&0Zpen)8Z`JT#1lu-e7~c%} zZg6`M{(Jm=-qW5;H26W7RagOw(%j}>#J@Cm8@-eNiP&`J#E%;^O9UT_ZVT>{?ubUs z31v4KdOnAIm?Jnj41A2>*X+NQ8$SLge4)U4L&U|&ab5Qf1Uv?=y{jp96-8EnwzzP0 z;Cin;vyvjAE_(@d?{tXk>7CIxBSr@{11Hi>2o1btRF2z!p7>wZeHt#iOSaK@Q6|1Z zSMzFBdc|#l*1*3l=#gy!nw`;~WIL-4MrC3xeuR^u7#sAf%nWPw9~-hcUog484Vp^7 z)uc|@34W|jdCj&G;{v-M*jEeQs4!(|;EMQC|;f|4DK4o3f<)*OFPVFpdwvw!?ENpazA5s2drsI$+a zSO^xTQ86u~BIxMEPWsfGi_-4tos-)%nbLz(_DgRwE7^5GpVBO>SKW(TG;WiuTGpmH z9dapRF(@;X@gGPxDBGP*x<2fgckoPG568{w{p{aPPWZo7bKe`i*%z6eQo`b(S5Uhy zvPF5?|3NgcRFqCAW*@>AAh#q!@SabpgsLp67Rw*O}52Kvz}s8D6*Q0**&o&DwlVZ zAYD#zgpq&%O9Qakd^=r8W4UYjsCM<)4Gnzfor?yG(K$h54AVTVz&cjOB^KhkJ~`QL z0i?|on@W*&R7`bHyW;$`)h~8M8#GPgRUSEhwIJb}?;h`cQQQd;-njQEHQL;Fk&}3XiEGFG)AMV|a(^=OZ*(lmc2|G9Iq#8Ap3YqwzSf)ks{@%z8$qE7o zo1ag5eyxAIq*eIPM(EiY{{oH=)fs<rhY!)?pNj&qES4O>zjt-Lxj&`D@d)$OhnK_i&gG*!-Y}TqPWGlG9S- zx{0lhf-l0X>ylh{V{4<*=vZV2eaAnWU+K4%ZVI{Mr&F#5J|xWPZ{Xci)Qj^(U5Z}s zd>V^eSGw>UtgXYmS7sLl5v*_*6ZhprTI*y$dI~y60owwED1Y~Ma;tRgkVwGIB zHLm+65F65#s3~S_NJn{wH6hg?8D;#K^K6{l`Rj|Nt+E?KmbW?mlgG(*A6w`1xS+y? zU0EAx$@Ac`ZgSt`ok&WewD~l;t|`p<$t6(72n<;_#mkvPQ`@F#_2Nwa#mIW7s7{mg zyWve0>S}Lf$#JfeYd26HIV1!BbRt}K>p!1j-K50@C@$>0+Gy$OViI5clr8h>q}Td) zNk5;QHaUxKSKNjLs6pPsZ1!J79{S;@YtxSi4k#A+^my-&PLnM5y!U+L^c#T&%@@o) z&<^SX>Fs9WTK`3U-P22=oMLuH1Q{o~TYoCP5b0&!@BHclnM;yfIK~P&+yU9V42lJo zlZ}wgV|2=m(K&%FY^euw8s$Jvwo|Iz7`;{9!j#WU5CGTFEwWpRS2fdn{f@IsnTyim zQE0Cj=Bzu>M3{!=&4nACBwbcRyAy@4#MhiQvE=h990&w!{^+r zQRfY?v&d)^cCl74;{=MaKl#?hPU{G~%Ywp2O#P9n0oHjSTe#8lM2d+XqD94l6n?2P z2ZRL*eNReze2NuLSsAQmfeT;wS8m0iF`RLN6F$bg@>=CTCYcwOxcB^5kTh351?|f0I2huZE%xD(gkY>?`f*O>i6KbM$ z$}@CFc$eb*l-S@{=+Wt*O;KAs$`kc^UKE}VbCx*gFc+n-zy=4-fY+F+U-`B8nGy5m zsWFWtl^bGQ*n8b?ftWmsg?5hOH3d-o@nRf*{XBsqR#xv!Uj+3tFQjc3q9J|JuBCaC?i&Jo79Q23p&ZuY2 zVDY@)eU7Z-1{N1KFF;g2z{|~{*sT;qPh!&9Wd14;y}35+9yIkpeG$fqO>T`*;R4xU z3p_f(7tW>^6Qrpr_B&3p=wx-p_=VGw)eW5Gv$@UQh>*hVr#M}mEWW+t1v65b{^eRW z+06|pE*w3nw?N7fimjwb8FIgNPe;+EIL~XstGs*=R>~8_$#q`oVfYR9e6|Wr|IHJ{ zQ)MA~_q+0DS`Wobx7^b};WpkQo~o0hTz3wAMA7SYDXfLL>rqCh1tjpWHZnFi5B5?W z@=vB_(U;}0a0$Do_q*)}u2}~ZuZ;*s1J@|4V@{YE;$p*1Bpeh(T-Ha)p zPH9TmXiJI=_{-b(FbCW)N{5_WN!K;iDfvJ0wq}HVswxR=u*BlZVk}i_;H5B!fk~}N zmMa~d&5tDj?ZU6+Eu-?uHhv?gEt?c%UTumx3Kx)^airS9Oz)#uhR6v~de=ajy3NZ9Gyr%Ly&ROav(d*;ZND~^)-ADx?>`OCFb*X$K@cK@{UTQx5q zW)J)QO&3rqsPpe6xwDQ4);(7XjSB6rcf6MNt?W5V=Ac;L7G8$hp>t-}WLQ~UTOS=c zOSUIt2+Cg9-6UL*zw_)5%*fgQ?^E6+$GQ0_uKO<|*DPLuCWiT#tW^|?)9B;$V!wu%74v;|!r&{}G}Q(1jlen~<{DcSX{t&Y zvM*uDOIAS9%w}ONl<3z_D4K~^EE<5>*cXJZ`ZpN zTK~KD5Hq@#rgZKi1uh)tKW>4SQi?r5k$os@Yih#RyOh!~%Tcuaa&dKebA=#<7rpE@~K%Wd{;P@?{9pJKmzic#6t0Q4~SS&b)5 z)zDv$3|K>t>4>v8etPj%u-VjH{lWDjk~@?*j|(qXH5PyEBj2{pt&wM>fI>6Nq70AO#!)Wk46o8 zdR3|ker%f_&A!kE z>r~@6q=LalV)B-+UU(8HaT+v-zl|7KEjq_PpguGT$>~9JWglDy-RATYXRDbR{$_*H zIxF-d+33Q)ONoW?*iNxft)E53U@Q^fJQ!Dq>zUFg15ZLc3XoJp~7{;kZy*B^)TlXl%> z+guE_!Sfce53FmmxtI_a_EcKB2?LhkKtW(KkkKte|ogve#=Wvu5;UOyKY#L zEFC*AgLI!_dnj@j_fIQXow7oP`qxW9i4?)1#ZZySX>9?tmu_cDL{MyPRvB(PU&M(=vyaGnf^#9H#WF+f=-F`6WVUzc*6o-{7Q@4t*z=xUKzg~ z_%w9tG`fjp(GR?3ClXB&N%jeaePKEbMhy^&{B8=lEg0p|+yrJ?|M^t06G*rKR{ zsYu1*#QoABGr%tO%%(M#-1bi@UmY|0q_pKV_Fv@bOCMe{|+$=Zlr zy!OyK`NPnwyiX?N(4S0j^gC|H?AQkiI|g#_CEP&qyWEc{>yD6bh-Ye}f8-W~-*(Qv z0_@d$$YMdujC(VdjNAOi2XX&#?zLnpeq4j5QPbzPWZc0Y;BmX=7~N6B9+1wZ?qHJuw53*y4Astxit zWcNA&k(D$$Em*sgUjYg~pMu`{r@nV4EF1MLkKM3v%H-P~^x(6fHdv>SU3Ob*BPpDv z1_5G-Lp_ioiATR%7n2KHk`$(GW=$Ye8Eg+j0v_COY?@pXsKw|OdadZgCI{U11kcrE zI-F-B?4EqPUWQpG? z8oz~-CSmf-{wW|t1LAS5!jdRxYv!LO+ocI1dEUq1ZfuZ0rq-$3X0)rb)ftXSHIEQ* zL&9)`STo+fB&Ff|#0;S#xAYTa$x~K^kYxdsbrhRSkt8bSs45ov!)g_6o;&y}+1TJl zdAgv{FBgh*i$T(KbcyQlL35aYM@*~TW7sx0r}v!x#5(luvg?qIU^?=Lt`F>lTDKhD zxuA9-C|3877UrsamB(?C2=zzEIy+jy5aXP1zyZATc2DDoFSLz=A(F>qm6{o}HRY@g zX4iejmc3?rX=J1P7JX?-YdDA~0U=4XOfRVSu2)rg-5y2GNO(?MW?;mRmBRp>X25k< z?stp+-ts&1f-?5uzfC5sE?kAO&>{nTmts39a+8Y5g?8k0-dX73>LzLQIcYZkKGefP zcJ-uZJ$)&19yTmDDD@%u?WSkySnUcAJUmZ6eo?2)^<6Q3ULJ7ZCoz@GUC$#EwA-d! z;D5sJ70<)*43MwwQzS)#u$Q<`oHV}YdF^&lk?+2!E$Xf02*Kk{b&h)8K2;|K_2-LB ze;0iSF#N34tP^RMd!3_1^AbIJD5d3iH&yNo@fYL_06!A-<_r?f6r`^ z((l!bBR5>wB(1P8NqrOxl7&4~3|3OFgvM0N^#MU9_%+7jkrv+n&)&CyHFaidyTu)n z4?}JQv$vok5gMeGi{Te>qMb|Ue*e$RKXc|ja?TmgOnW+WGM=_GXAl*!MFbTsC;>!> zAYQmAqJW@ysaLS5QM`d5SWAnbNU6%dRuWqh*_urfp4fAqK81Y`?(fa|*0yKHMm4KkyTMm$~?xw;RR(glYozc^|Uu3KB(^!U_=0~UKM3o zuB8fk0zmqGW+B{Hsq9cBKzZ#K!#0i6#!s%8V+CLi^Q&UA+k*l4zRiN_FvT9E$N?(8 zgWj+0P;Ak*D2_lOkRC&MZM3wf55wj$k?+JH%^_8~2?LcIAKFilZaHzx*L-m_XWWhO3!kn<#RciZ2gb z&JKmVzHUy^>_1z!QM&!zz?z6AbsuSzR>-de*7(#BJl77bNi&#a!F5p2NSbX>=g<%3 z+0)u*V~`{#G+T3&t`_Hn;@(hnuP9ki6EPIB<$KrW7Yh4g@YrcN7>JMq3!1G*`k0}R zBjl7iC#C?XbGK$a6s5<02xu?__XgY%Jrunkz;htoJbJ^%*0E80d=;bBPGO2@;aU2BMN^**;k98^dG@Ra zKIthBPF`}5@wRC91g;A%p@zIH)ZCL&D{YpixqAF=o%rpb#dG2lMtAGaJ~&YoWu1Xk zZ|Jkh&I#m{%?xyaVyh`qg*56~F%qeJLy<(iB1&K51GD}NocSM$QbmifyI>b^sIkRY zbx4n5Bg{KBz>r=QrO)^8rh7x*BZtGzkhgW<0$>ZU3&b#|EZoANUUb$wQPCInD6A|z znZo~!^t+1Nk>$~yw9`U`4!-(flZJzvqkH3@-NkFN@!Z_-i${{`?8C}@%!=p!%Lqph zdMmw!Ud;@CrEz|@auE0t8BB}ff#`6I9*a>{_#5do(V4J8W(s=;*W;X@e?0~I;CS?Ld z61$YQnZ*tOkRNayw2RNY?Af^EwvGEq4;5;K)X|sMRFN$sLCS-DTUdP$lew#)Sm;l) z1A4v(FAdLy*>gZ~V0x{zR(jL-USy-HjowO^1#A~~N4GO25I}4e>=X_VkRp@Rf-kc= zxIU;>nl9Tb&(ysqDx}L**@|vD&-Y_>ooeUYYfuh<#hK*0k z&%>X05>A`=1IWGo_YI~rV6!4iF)#@z4^R}=CjM&?mHNHEUj5ac`Aew}nX_}3S#~>M zcMa(kot?Xl`s<##4ZMogfnWVyfo0Gx;+>;h;)> z8+?&^LAh^@Z&$!B>5v!x=iUj%iII15Y@dER_slYqHOc>W?~O@Tf28H53$K%P9_+G~ z+4vq?C>C1e87#7}po3}Br-|}~jpEB;HPh<_x$M%J>wT|79rICoo$v%H zkNsF`(iID#>y>!l{A|AY$EUyh5zdYpCL=t{m4|r2gj%|9-TVKu8tOHH_u8Q-(0&x! zsM-@=9+U_ia!5baMhyBPl=f>%gj!_{tjmDPlXE(^ z4)E|7%XhGc-|Nn8em3O%xQpMjFHw4~xCI=ANEcqter5SrEk8}V26CUAd-<>H7pz~< zIv>b1ziRv5p>LYzLpk&-`z=R+IhPf;%WpMISj`cunVit0^gkBwgKLqNvwhBf&EB6Z zZh6_H+dd~x^Vf|pFW$Vsq}wSR{Kl0ZEq>#U_yl{DZi(KnG3i>O^E6IDkT0x0^IIL* za0YhVu#&8nr`vnAd@_a}9Ng#dfic(9$MgjB09~#Tgt|?-LEnDCy5QA;jVcW4Zw&5* zK>xypJ*t~?x|w1jw(i7YC+s`&QRKpf8>5_p$TNn~DFi*^+0L9k7X8aR-?e(RFCXgr z896?IblEtMmnilEMVc+nFPxV^dM8NS*9HD ztqf@pZJd-Y8e-CA%fb`F8$@OEO6J|3yjfwXQbbCwNrycN@JSnF_{@^1T5*j}ia&nh zcK@q9x=A=SDh_;W>ePiUoR9HWk0- zOP$&kQ!PJH+ry+QhYo!wfJleMwRhDND-m;$T_`(Q0oV3AsEyeDq$X+VG zO$?5#V9?jXSULNH17FTwP`9A%>xp0P`IpV#JN#4ZB~Ub@C1j=A6#NLBw_K=P!eXad zdeHAo;Azd_uoGk@(gfKv$#gHMxn!O)s! zx#)Uy0gVi$OR@{VRcca}2v)HlMe2_PrHV2b)aqyyw==s1rSv1OLg7J?(FYqCrJ65l zQM5D1$`H5!0WKMF`-kwdgyZi^zOnV+t$2BBb@EH3i=V;qTvJMtYziv}DE1yj zdZ~DvHdiVSOiqqVkf3^52J>!Ue#}S8G8qJZ>u79I*+$<8D3Tl!b^wG zhmT^GGgpH;W*I~$dCjv~7iGq!0!I37!E9{@?uC_pWCgBlKE z87_#_NG}Uom|pSrAU%40EliR8xJX~1t?{W=R|Iv?$78!`q*wq27>aTyVnKy|6P5E- zm&|4^FlA57X!d4lH2d79>@Tky-?GnCJy%g?XmvSk>MNw!e2V0PahjbWTsP+$1fS~Y zVpv-*ml%ZS*n7g1Nwv~rvHLa4r&Na?n7)lJ1v_;E)?1Y?*3P);i~iB-;9BW+;u1J; z57p>_!u?EFoZ$vaVCIHz+h@Kx63n>D!v76w%El2<6-J1#Ie)A(l`#jf{|9pAe%z4(isQIVM z3P=?{H^_r)A0Yrf%+kC^u}3I!h>FMNIK{FY`lK4F1!gailuj|IYvenmdXSs&$HtNR z%LH5CCm8LW~1-TImjbUocJMphft|ce*_wNthzdpynsK%YwP5<}Rw#|W7ko@Az zxN{`y3j>n9HXtdcSSXX+Ld6#fYXz<33VlahC@dB(5gga%%XeYPlJXrvLU@|U0-Tnq z*=UF&jM1B8?!~}LNBl`-s}!?s2Sf*^cPr02mkfXU7&&KAp8jlS&Q>1vO1G!ma41S) zrzzCZZBjNv;Mfq{sHh0i@7EsG9yc`$Lo!;{e|5E7; zSHg2$GH5M3+;J(O*v%BlL0zI21!~Ug6C``VSI6Ls*_F5+kV9V(;PMEf8Rs zE^k}pz>9Tkm>qbKi!b44)g~W3AM1=Xg9m%69MTL$iiW^M(obK5%yp)^PE6T#nenu;?lp_^GnE=){np^0R zvhehmFv*<39D{8)hhSllMhS&c4KJUYxF?^V{c^&hSnFzcgHe@B_V8QndN6#NZM@VY z6nlswbyWN%wvQ>6A=`022jWUECfe+-CKB4#|BM3UA_@u{t3Qu|n&umhnWM=gOY`bJq_tCw| z<$;w`yJ;s;I7*1w*qxDg_yFq3YyV^8>&Xt1r`kbvY0wudcns=es#@u)*o|@edgz9i zDoB;#fx9t{{-~TXX51;>U)(4E{q8x!{ly(2i0`=*`+Ie-T=cS9nUYWb@D@q%VD5{- z#%`oh>?(@9O~spM7P2v}Jf_jxp#3QFN<{8hgh`*ol_M75$-CTtgcnxFy~oZRuxD|5 z?6Q}GmyM0wTNG)NTgh#|;@}5yScqr@QEk*&ZWE`HE^t57fI^T%?-HQuhH)xnUes$R zUh#2q5^!KJ9YM(m^MC_ScJ2~hC>dWmyY9Est%hUS>r+xl&KJf#fi7CZvVS`%wuB-@ z7(}b1(YQb;h28GI-Tx^4sX$*Ns+F#tQUOJ+1Bx{7)xmwjUF0MPbPgyCewJGMtFdX` z)nPewby%M=hhC?JRA9nH%PIFPfJTWgTgW*Fr4jEoH2vvvfE6*#zumKlWP0#aQ)z>N zB8r9FK|W&Oab(mSPF){!SBiRrNPfFJ`yD^MQMOaqCPq2`?a_JKJl{U1$-7{xL48f0 z3kp%O#+j{TUvvfxzuU}=v!g%a(7E6sKkIV;D?q$t+DcO3!2p5f?J$7srdVK8?WE#&3itTjCo85VgeL~Kuto@v z6$O-tmQSb#8o-&rJXz8Nldb@y6Y_j}=|Q@6`ugdI1-Vgm^f~cOwr<`{--F^trP0SY zy(S_txG&}?Br219q31L(UZKV{$klM-j~m+=<+D8(1Be%HsF#O!U$7@ZdaQHHL3qR- zRULF6Qw?4RCg~C-MNpTk2iHQ=6|m|B;Lmo0r;+d97dhy=P17J-HK`D}Mm^qlUP_;6 z(B{x4*45w`KBi6qW%z?!yv++J<77cTGwtOFITZc|eAfgcv9TbqeKsikb(?ALz+-Ns z@u+UyW4y0Yy?!2VP4ESicP0TDix0@Ky+?f87$w} zJh?vTinL}P7IL_m-aXQ5c66Xey4Rh@c$vI$$3IT;wvWK^u`C|UG2n>T<;z6&nCn+6&R%7-BqE%pM5MUCAlyWr4}JBlegBjM1q zow;at(%=4UPltW3vdA2ySB#!41BTi#*S3cgg&dfZDk=y4N>sEQq_@ng*B*~rr>xU< z0FMAyNf_9*93PC!4%nyN?ndqHQ9|M|LYx`(r{2qLW|D#mH@swBFIUa{uN7q51oEDZ zLElfYFd)0ZphHtkOpP}A*NP926;ZI90M%-Xh{7W5omt@D8?ftz4ux@QvLr#WI&cvt z_zmi$h#k^gC7oLyiQTB)s7{|!6Ky$+mn_;T%!Nd*<$1XiUGx2QazGvr z%<$h&ebJ7dVES_j92|Sdw6~HbYaZK-oPVT~LVg)!&y5U`1{-U#hhm{F zV;5!*v3q8g;yn=UHwQJE)y?W{ni7R2j*C6-Fs+y|(X3d2nZ`C)SPyw+2y37)sX($i zxJ9vtt?>a-gcMNE*y@J_2JAJS!PEz}GhIs8CXsjnXSnM$(zaumaOkSn){9S0wF2k6 z+kXECvdM#UKl^N;P(raq6e*yB}bgE2`#f+eK0?;}REe`QYeGVMI5n$~RlEGX6Efs?rB~mOb6zo5baBEgK zM${?l6a_T4NqqdkkCd3B?gV{zP+~2BH2Rw2|sRK=idj3+c92 z+zSaQ2S^Ss0N6;j=9$Ny1wUQ@8+Us0nsR#%B?naz0~H6t_R5?6FG8=uI@L1Pfw%Om zzjid|9eMMkd9Jhf@Otx}>j5MMFkSNazdA)uz$u8AjHQcc9W_Tl zF{Uo!%;K;ECwI%|vvIVsBF91{L3=%wVd;H<)q( zh6p!O8@`epw$eXKcLKCbgg2Nzd9kcV46aO*5|=ubRNiJbgW09A@DGvVINcs_0yNwT^h4rd~n8P(hi?#d2jMBQlo2xIi~(ltO>cq;-qB0 z^eBxbAGR9Fmaq?5Tit2y>1c;*!2 z#04YUP+pdCT-%>SP0xF;JWK@O0w_z+LEi(1H%&A^Rl0=-iOYkG^t+*sbn$j-c=%()YA&4m z)Mww!3nSwOB2NCoUKf|oOzgq2I}R1yx5?o!BYj1R$iis_X}Q?3vjQY2k+X#c2MI1$ zq;291#zKR!{OqRp1b!5W_p~#pUs+DqM7Ya?8#THf&q>b)%BZdg)Bk?IeH(d?WwUU! zjL#Jqq=and5{ad|CLX%4SsvJ;*iJrxMD$jw<(s)tE13cu@|P>5?cVxU z(xgNsxIyM4CEf)3Zi5WohiBhdeEj8mE|Su=jE~bmcnk>qJ*pue`6hm5X3tRd>-Lmf zK4!*q9e%<=;Ozhfzq3;Cl2M5TW+sc$4!SgpEDQ#V#_|LIcl-`xy%3g|WGO)kFbV?0 zYy2L_5BePvxvvD*Sy+t>3g?d=?D(@*V)a>GFQ52dB*lY$7La8h#+KVev6-0tjL)ZP z)Ry7Y-xQvsF{WD@a0MjT@!+CGt<0r)ApHnlOlPz~`}f5yKXcb4GJH5jHQ>V^;HGVU zb|iD^>tFppR(Q00<5nT5@L+hHwSh+i#nw}#mWs#tu6b!Q=r`zxfmvMtoDsIGaV>-^ zAANg3Vko*GM1#-Q`Kv_h4R09z99}9~#D#4b~;)B6})fMMt z=g_N}ZXm(5FtrhTz){X%whPa|f0o>LuQJOlhz=6$J7yNrohqoUW*#XoGWsULIq;p< zMzlp*z9ITPX3(!#hJ8kGU9YbbAB9K{#&Oooh5DsdutZJ%eZo~NvTaafT|b~GkGTR( zvh;Xln|#0KhH6cO9@VD$RA&Nlt{n0@rp^^@(HiN@8J$|=%oRQ!+c}2cm!jh5sUO%^ zpm3CKz#D{~-~WAsX=ku4Y_mV|zxz~sRO_Z2)OYAiwmP(xoFIL|bAdPKfM&)kMol*) zaWY`v?Vf~#7U~JJJ*w|3kDGW6hUcP04Y2QWOb?a6)BXITYYjhS6% z3qJp&O&9xadyE$eo$GWdd-WqLdEyf{!3(o2Sa7I4OEO2 z3&jO_SSX9kgMN~zxTnmM-H%05o=Nw3FH3Mepq73pN{_7!IUc)1idzQt%{dR`Hv&Et z9Hmdo`(xY&67zUEffN_uf%_4>kV4(-`f<)Ltw{Oh{!OotOCwdHcyP{Ozy>`ZP;4hf z+7Ugdty?6>6`{h!5~`0WkFlf=FhBz=Plynj>mNUi?1jbhT7O{R$Wr{bi_r0Ys}C5Q zV5V~Djgs=1PC6gFT62ngfF#VuXM$u5pmBh}g_e+X1@tI_%Td;9PhgRJ&^K51_8dLN z6jw-YGx}cTT7S%1toLiA&p>x0WZEE+Ek(Ekj9r`f9IejZq0%NIAxlJ zGR*&$C$jKP0MKq8b8^=Xi`)P_{afnqVOFq}b(ba*;{cz%hDboIzbR_fv9Pbi}_vlJ8fiT0ZlU*P!IGN7Wz5k~_f7^Q z4w7S-81YuO9^{1y&o%v%Vpmg=H+Pf6^w8C2}2utaZL{D@bi0&JVIB-P=v}6;DsrUwd$B zfujn_C`k%#R%7D8yjI4z+yKFYxE6l^R4k-Npq4KbKVgIo5)5#17?T)pMbZ_9z&RPc zwr`&I9xn)u?0NA1Oh<0uhETrZ$h-E$Jx`5t6Amigp4d{^mU-=}`oPb(nsw)N8BAta zqqrubBnIlTXEsXf1F;A)h25cg7--VnoU=&-{Ssxpba6;>*mnOx-}=Difg4_Mm&tJ& z8;%V0D6Zkg#zX0m@yKpeV z5b=Y?2oG@6AzsiJ|L>1>Z?dm#^H^wb6y0HoK$>@p6!@GS(W}@NMW-rltkiaG2GJo% z*xWHTJNzIqvHRDJ_WeXW77|-++|(?J-9V9bRJ?`qVy3$oX@eGPm5}FcY2I8&o7gKTXXOLCLAT)SnQJn%6B!>AY?u$e0vW5+#dbE zoUG5`#SMTF79`_dHN~y6*TwyU%px2rx;TGSt55i@4egH9TZ%xo`e9fE9oI3NIIQF4 zkQG5!0UdW%1E%9K`aP%X% zGx43MvwjzXO}boBgCvzL5+3&(P~4rC9B0zCFgqso32P&^(s)(HjH`3*#+(zo0S+S! zgfTjVxDjU$7K`%u(uGdAANWtFXv?Zn!0c8 z7FcJY*WT+tMwU(hde87|*jkEBqsS^W-#Ng1Sk3g(CttoMFPVRXL9#Q}P>jttZur}B z@Z}G`WlXkty?oC##^O!wJA_tLc!hc;lI0$pyv?_PYX-%xp~z|~9z$S>lI1TfgZ9WJ z)NTLU&|s)rxe_Xm56l_+W)2PlcFes{UVt6H;<8*3^4MDJY*@i7#dyWZM3JUI3S+;-pP7KYd0cy7Xc>#OOp_SWPJ;72AFT7=6 zm*cUnDo4Q(mP{R^Ya`mkYbTb1_t_lL1`NhDs1ZW*w>SKl8Wm2@M&MSKpojiIh4M}L z@^*pyA;4$chgmY0kF)&)d6_lp4_$vv7;l|(maPasOoqHXH?5LwGa0U?*c6Jaq~eds zJD8)yEO~(SbqD6SYf)nVs_ph2L4@5cW3tN&5##@^`o_|it%!Ia%=-nY^h8)G*UBy3J8QkRPC?-w6Ab z!U`zq9qFqi&4U58!v;{f6bm*di;BNU27FCAq>~N!E)kSKOQL2d05<8mqSwHLbc>t` ztnn!nHcB_Va7?o+IxBD=bUJkZ6a+&z8it#jTm8kh_-C!Ei9_5?4OjNg}>Jk>s&=CLj0E_A2Qf9wxIAOiMS$bR)H^#5e?P{GpC4Cl%O&L=Joz-)VC4|SLN>UD zif^A?DIAFG_1!H`4J#5<(aGKiWmW7Q(Ic;`@~XL)RH?GvA$ZQDyDlGytfSLocSj%B zUXOl9(WyJG?bL3OcM!a502EM;L7C>mz$*5U*ZaaPq{{b^SE8tW_7ZPHFdn-a(i^ao z+}8{;IidK8&f~`Ah_L|5Y20ulW&|Dzl{`4_2hY2t;PMGeCvWv$?o%0?G+|JfND85^ z;Z9X@RLfT`&mWAd)9n=2LF{-ZY^O%th&&1+234{33X`st*)+qTPL4t){=e-F#{)Nk z*OnXw3#kRT4~qS#W{mDeDG3xraD}N^|ClI5}cRrJ1!RJ{^#Xn>c{m`%MP;K{lC@g;5D6Bx+4QpPg ziqfp|Tc&JsLqMPNo5gPkI3MHk6TF7NbBjKkHdFsH(~6BZ-)gTWTRnK9I${Iby%Y;l z7iCm@nXEc=#e_odyT0|oX9ANaKyRw396BdFVRqN-2mW~MsH~qffz#70NN16mbXSo! zcTM;*5gMGGLXhAJ?uF!6vwyR^Miv&A%-$VuW>j6tZ`l3yo5Z(YmK)mXD^uxgZ z7*L#h^9FMfXc;JgI^>1dW-xH50S*mLs!44+Y&%g^y=hsZzwO~+)p9c>=DQJzv?&T1f z{7NQS8&Rl5tzY~z@Dn^%Pwu%VklCNo1KxVg40tB(9Lyp83l(1nisN0nDHF?kWpOJx9XpoXIruIyZ7g8C#7E) ztbly-un5h&#lJw?#}v;8N<&QlA3o7nwJ`OD%bO$ehs5@GJ6bauice@jof84udAs z=Gh5!lj{WsreDFjH0JaaXDykt7#aaqmP4b!&G1_$ulo6w|5-(c3E^3S1LT?B4zbQ;v2uOnDpBho){0c3!8LdmtSJ#mFuL_j8>+!wrfN@&n^H&Aa#w`!1_r5Ss2b z8|O8fVzVf+fr>ATE{Z-QYmV9#dj=vPRZ#GrD=O4pesST#8lSAtJ2Y;!iSdeT?P=|9 z`L5Wpus5SZle189;QqIc><)CM_RxbT9*)|>c3{0g|HCiunO^`c6biH@F<4%f?4$2f z-HR*{wfW)tw>jc`#2(dV=+xX9T`U<3fOssW>Nl^vH6zx#{M}$w<&r%f9R6>%u?a^g z_7FwtsQ6s(rr0Gu2S7f!dS*_@4P{5@5WUA|Y50I|lCBQ2jo8Nww^7vZhH^mK7S%1q zZKHOxc0=ran4FGi%EB|l3M3|7SpfcCBh6M=cn|sXQpujcKBfnf2cHr=Rvh#Wgnw7Y zwo21vDS^l2hn=kq7?trDd0w(R>VSoK$|by}BI@tw9&NW@<~-@idaftVC9i|p8YttI zIb#UIa3ly6K0Jr9g@)mMl{m#%eu=UIMv0W7D zpyHQLxEPQ#?VR$aZ_{^<{*;}%Lw|33EYwv zMWrG^Vy?Y}-cUKvHS_%dtPS``X`K1M_XJrFBo_?ko7h`X7~RJwm~?pk5$_J@N3~-T z_Ng-I9{&k{`}`xfNj9@c{Y#YGoP&=v($>e#?2oZ@eJzacoo%59)(cW2s;3>`WYIW_ zpC=K;^T=n2vEAS{L?bMssJ8G`@A^H?09+vpN%|Ki0{|i}!!iJQ6blNK8>#r_z?#6b z^gUl_C=UdmMT?-u#0L|=XmX%8%q(P>W9()TBC18v&0JJug}zl1^O4UYcdp4{5S>7j zEd*S=m(=L@mH)6}Bza}(T+%%f9oU1tn|Evo4G$@HkRtb~__MHlZWCW5Idr#ldDu`$ zv$|N)P3O@MM7N;${Xif>vgLw2ARE+9J!L=)8m~q0sylXtM(j$L!al9Xkab zMDQ=UDu1N9Ij5Wcl-8F|-$J)UtoN;uCx&)RJ0A8(RV0U+bYG9mS|7Vbn?9umu#rQT zD5|1w&TNKGM?G|^EF-Eu2zv=*qv@5(hA4}+8E*BGp|I4zMit66yUN9K!hVfB^|{E7 z&G^-SQ_s7MatBOBXW;)`Gxf!?9&gmAfGNt&6VgcU-@!p1=~lNM<2D~7xIN>%PrW9v z?=$YPNd}J2;oG7+W|@LfqVDzk-(2>iUEkO<|IasdC)HV#Hpngnub6u9MN?p&ut8KG zbXia-bk&#q8H4B`JcfPWpYcE!o#q7(>QmkQ}CaM zA-9Q_hZWH8O?q!qc}zP4Ey;|~gm+!=7(s)-gX^NPf97L&fn(gEuf<<{-WxWc1l@<4 z^hwxau9qB`uE&jPZ7*quz)ha-z8Q6@W$c*Y0tSee-FW7Jy6ysA&=~)%tJS};XEc04 zJXH>^LytG=Zts=foP!G)%O0+z-AJb799#!6#CgBFPE^I}_XS$l&%Zb`?i|Vb!o*ni z+Qe9jDHh1yTd4S4(EzB&bkmrk2GZ6+&4I8&XqtE+tRx6JQI-fYXD*K`kHKb_s5;ay z)nn8T7b*IEp&876&GIQYFD3aV`?w894uWNvqhi159=QDCjjt`XpIx4-R1!yvAWS16 zMcbl}mo=#nC^Fa_2*Yr-(i`UEV>#jQnsMAS)FVHYd&XJzr%7SFgSO*q9`nozWtP*JL%m=RwBh7p!RVl4&bR z0YCfW!TLi$%^Q|_-c7NfaJCat&w`!84snOLQItpqg(kruG)*1!%?`Myy{f$ftx{Kc zuZ}Q*zEGiXwJ_z&*Fcc>k=I8rH=BQ7H~$_KriyZcPLKo%UT@N!(Kaek8E?>kh0Cd& zE<@R9;ly!Y`p?_{Oj&XB(C^`C^2lohU&w=#ew%F2kxH@eP$Zd(-xJuZhQh;Mc}CO* z^I}QwHV1~|&6*E8GVD*g-(Bapjmy`6{_8(Fqu1@hP~p()-Vxmo?r@`bKIDciMI&bZ zK^y(RH^FTUL^uc55yZI5CE@&|;Kbge_f|gd-R+>QjQCv^9nfCoM}do~Lk~=EXRzV= z*xLpNhxutE;+AvVhGB$hVchsrKhW9>qjKtxTHD@(Uze{f^ak}2AmUJ0zk2W6Z+{ti zto4d+`lfJ+q&Z?eICVYZRg+*vf#1y*MUmXi1LWXn*!Xp0@g?Lp8Xl{Re*d4pJjLG0 z;F!0UP&b$!(5uN4b}KhipU!vlz;k_vo(%}DJKcNy3j&CbfAP}m_L4hK#^pUYCc+`Z zgQZ|77p6yV;*bbPhwJ?sfv4XFi`zlJr3+Sr6xf-tL1yew{cIrH7&Z5zci7Xpf&9I_ zDN(bmE0Wi1boHcU0y%Eu{!q-mMpiVyI)QX`s_=Vqjd z3bc)Mg?H9mgBqrqdQlgt@o5!yk+a^Yvz5WDjd%|l4xvyD#v|UFQ=@KDHV`9SFDUlc z?*rBAHhPKfSGzd+c2E>fe4g!@iK& zW25#QrO8)i*c$7M?5?a$d`z`TT`25}8SqW??wHm}8X>50U0y<;^*^iWrjMzvx!X7T z3{Sj)``MlD-QWdW>e>&&X4|V&@?n*D@TA0{X?cQd2R8E-dWqyCCB|s=_+zIq7iv#C zATV1j0i%g(O=uUfL0gOBzG}#;ovE4DQ?3x)nP@W zauvH`YA?O=|6P+eDj$Xz)Sq7i9YoySUw8CF@3K)I@j?7-%s>BF_hWmYJn7zgaLAoQ ziDfIjeo9AFaR}0&Ev1GT%=#%egDgiZ^HULYzuO;MK;N2m9hwmJ3)Tgnl$L<0sTMzU zH~ZlE1JB}BKY8(acUI2G>I0kSyxtL*E3YGew(Oy#dLg|;i|OB5x}9#GldZWmt3s7V zE{s9zJNyQZ0t+L4@m+n8pCM74Eq>%ov6crrD;&jIn^|0(B`mlmHwyZM6;L{;$7-D% znSL9n6DR3lzeM=noD8Nwn@84w{&6|fALqy*RM@O;SDT>Jv_-L8+Z7IakwDQ`e*mgb z8O%{xFD$@Ym2K*)AP9v=*7zIgi-H1ej;vhjg2nhK(Drnn#YK)wWnF%sJ;#8Lv+KcA z6bB=rk7H3+yqGSQw z3&;PqlzJ0;-kgz@0|l>Em>qC8CPi{Y*f7dEA@`uk25_S@YiX^QpZAun`#~^$fS{d# z#f`W;2zt%LnAkqmnZT6TUSiHSm;-niC_de=zf9g}6 zKgny(8n@)%mTj_ke)-rO51z9)+~Nws4M7H|KtXw!;(n~Y1TtKAr?IqPcHkivHlHM`v%$Lq74~h-ydeJ6`rko$&C{1UR!%`!PeQ&;KP~*9qb3PPq z_itAn5f;!*q)l;_z7HxqM}c9BG{?h1McQqOG+_z&8@Yl5vroU&5xt7NF25gJ3nAbf z5U8jVoPyR>sj@Zx_%wWiDaB*7IA z_H-}UB){IHScvl;qT&aIyX9E?(I`%k?2k35560F+Jo0)c_`|SAUe)3pU_0KW*8rVw zUv#cKC$vLs(yb1@Ij0vqP-u}LzjlXrN95>?obHmiXVzhI&^&S8dN>)5I^!t zCZ(~1;$~%4NKJ4qi{&NQ@V-EHhuIhmeWT?@MN!CQHh0uL_S}N6;h-BO*0}NY{okGa z@ei$7%B2RsO-@fBAKDDhRf=t)h>41CT#yjAH2fp@c{k>+FRoQFf}S0q|KR1N;cal( zl7>Vw5@=5t=}K6yVpg&(^1z%;Q8=UEJ-D4rVe~ycrV#{Z3Zo8&8PpFW??vkI&Q!sU zNto%Zg~?}Al)g8%Q`@AhqqBe!olI3K8laz0N+2dPhrIHwdL^yWo6O~~4KFk)^|?No zGqBLZazR5>L)5bqoNgbp;o~=K`&mQDJ!a2U_Hw}+@rUedo;iwzG3)^q7X^~d3bV3O zPVi>Mx?nuo%PF(=thYGY^A4k8v>!0a`@SHkINf6yw5QoUSNQ}EPT(#e(_L8-p8_|j>g@^Mr zSUHU8VOGk#`Lu_+;2bxmkEy=1LS~(v^2F;~Nh-e@5D%tqL&@2&8W01;ZluV&RJ=8N zoxz~;)A*Z6UE$ z!mU{b^}>@cU1Vyd{fxnH(C?8~ebA2RMiqAWMmH9>&VoGbVOcvwQ(Gb)$QMFo4g>dW zbdO*XFSOVjaO=K1fb#`G3svyyU+fJDpBS$PueLetA&NvMgjdh5o_!g}47*7ktfzAY zIrK>=SH&Fn9p4=HAejq=d=Sp{40jyg6fv%Hvel4;{HbOkd6%CP;=$0^V`C4uQfvW5 zHdFEK4Eh|9_@h%r8$me~SLh%3o|~H}=~0v`9(moG)gVfg+?kRCQrpdtC^)V{2-S%C zW43@iZJnZ(G)1^|CglSadt2xZh&+B~-;a*Z_?{Iidw&-43v$APE8uV041Np6nkaIP zicj(BrVF6~GPKzv6ckx-2e?mRNl_@5fZ2lE;pPicfs%l;Ws)yY z80d0Zj}f2Uq#~$T0(!&k0g0h_Q=K>!pn`2L>!FeOr*x5glVZs0=A6Soh{#g(F+hqa zftF|+A)WaM>gfB)PT@%@enJQ29BPHDgSTnStycJVj*yG3H{l}nq z%5Q#?=aTPcFMH97t_7_{DWsU6A@blnB#18zbIkTpEcEiNpyDmbDD3!v1}TF%BWzcn zgx*&7NiP&hKys>C(4@pzV#)kL-&U<2340b*j+4@ski^hIzy4X7z}Y%RZ;Zozi--qE zU1E+4%PMw~yzo9bDloXpUb!44RvW3F_F{?a#JamrX0$k3=H;mD5Ubvv%D#`f0y<_UJz6FP?*`&No2K{oQAdNlu zT|t$`piYmer5`A_PihipFw4V^l6=7t-@}63sE;5@SOrD62c`|g40&x1MkRIZ(Q8m+ zJb5=s0)Cv{A~#s4!V{g5sRF1egElO=-mY^rN6U!fT)(&sTD-tB?honLw}n|JnzHWF zL}Gl-*er)_CZ9@*1yW@x72h%Ka9FOOPHd!+54mGkcfc zYT(SY`87di_?&kpJK*~%c^H@htfWix`lSzGdOE2-2_cbF6CVZ^&1(PyYd=w~}(zfTBFQ z)5TWfnPciWik|r>mtMdw5WjCP*un=F9vn>LkaEHHsTOfC{mE&I7AeIo#a*!|?AP^+K|`iU)yK4Dq7E8W2n z{yF?#@Z%Tl-W3Ng6Yja+v@P|2Z?ktmz92S)!(G9gPr0g1+&;T6CdEgOLdKT~W{ln( zj7ug(0T&s4jk;K7q_2ZO$B-9_-6cwjTpC~>0p{jIqsBVd16_1{-M*%S|Ho=fTE20s zkW}!?+IVpK_^gdhY@pbBiqul^$78$cjln(SiqGz#Vp#dCp1YbEh)jhfazXI+=sJ2p zF#zkJHo8IzKw|F4)`62!OXJlB@N&!MY@@5#W@S3LCqEp9`W*&!f+WeeJh~og8wX&$ zv^(S|y+!MV3C+3--`n)o#Aj0d(ss^yLaRpz%E)4tZg*{B`ag1)1&*e zQE+DDpY~j5&+(fqkM6mk*I3QX%$$FulR^((6gSuykUbO&z}ZE`8)Z9%knAO=wOe%` z%&zgltvV>`>6zUhb4N4~*&1o-@{ulE7QR02(F+Fk4b_^6BLo<+AbNgMy)?W)`zc)# zvkQ8(UuPcpAEnKzt4t-dQLfcwxtcNS3~+31g|kqxy`L9ws1sG~KXRtZ*@NdPjww=MCR1=XZ z>k<^pa_Eav9?6hE{M3ez{UNHM}@mZ>tVGaJ=_x$fI-`&}QtH<}^h z_3ifDHPm<0e=`2rl|Y{R_Y;mvAY>)o1mP5rBFs`?pkL`w!!L*~FF`DbRF+GvI`JsK2_Akg%+hF4arBUoE zioA{eKH8ye(4gOS2;>?daAyj0bFheOz#IV!*S+pMKGbmeoA!xAj$B-#_mW8{+IMiq z6<`mV*AzOOGx{P2eee1{gy>yG?E63qXcpkPR>lO$+-xezcQu3Ay6UNCNTV`S(-~bT zUk%h~lP(h^0e1>7vIB~@gHBDm_2L5;*QYx2%7@P%PuX_q2ylNtM@qeTtfKOGQ*_)^ z>l*31+kXECvdM$jNc(KeLkYzeQKS%>)v2z|*+0edvsrfz9M&9=*g)}wkE3g*=V=>6 zMjw+dZ~7ipji?vAp-sV;!J)k(OcsI0*H0v@o1B!DH z>dIhxl;5P^!YfQfM#t&QUE5|%GFvn2~ zW+BA4GuV*3A_%H5aEZJ8g|!jZ&6N?`r}vFvj|ZM&PwqnKlh5Ua+3{q<4%PDxjoXte zg0}kALMF3j+7>Oc?`wt2U$`X9p=+iUgV-CEQ;(f3gAw6l&Ym1V_Z{H}4PQ3p-2bes z(AvrMFXqg!{A||MO7D}K;I^1_4dgyqJGmtyX9h;yRtHx|d+AS94+Y6#IRLsG`lt-= zL#?Mz0^7{WQ0peIpOWQ^>P^)_C4yFBoOzT!Hgj)WzF)I)!2gbaCCGW-(quv2s5_`W z=qSC{`EF=@K=%}9{E1V}!tq-XldE5|x+ee8^t+kl!wKZ=@xxt{2Ne4WMf$1ugz!|- z72OtD9*u3W^|wW9L0k(vDV)|mkbfw)bhf?qV!NPGwKnu&U_r2cwf{u{YU=L|ubHM# zmEmTMPpY6EyqJfPrLcTz4>|(djS*KOAc%)yw^HT^!OMUgE65KyOScP#yt<%IvT^D% z<-ss$kJf9B&E26lsq^SWa-B|yI}75as3$cDic+X2RW-L-oDiqS2HvO%XQV%zl{IO| z3wzp~AghB>AM9-%|h~t)!nkQhh8nYD^1K1$({QXS*63HbdeJ zP;4IL;vIfKS@5WBs<+j^9Q{9a|4o)n0D*$x^V>R#O{YjI6_17Eme`(!URNx^hV6-x zeJ_o%gNiTA(YNp7m;21e@p}87OMSk)cIa*U(67hZ5004hTQ``$qe|S4Nr{TPisUeC z=!Z)CIU!k!uIM$OmYXX&ujrm~o>U9o12GQpeW`n@&HkC%OW`*H)(aD+cIpm?Wrg)aK__zhvZ5bqqoy-|Nn!Q;+Fn0akO-{M$W?Nah!YybB5|!;DK2#X@X2pNdbM zxH1-`x0x()0bL$*a!NZ>B5H-yXbydD# z>9Li<)_@bFdiIc8BZ7`$g<~3WBUt|NA3x2tm-gplKs=a0$I%ENS3N*dBL?NGbw+x? z_Z(!oDpU=DC?(gd1nn?^c}sUTXcY*!_kpzc3duF;Bd;qyw`LjX6`_y3a@E%-T8=@* z408vP?&c>7>a=+7pgc9=5%XZ?RY}Us)QF4PLHPj533UO*8!+G$hB%<*iW>b(f5Bp|8jrjR==RyEf^1j~Z<u~Rk z#S2$Y*o;xb=9xEEedXVW?2QN?m)C=7-5gArWl{NlgHX$0q}NC7)|gw?^b;MYFcJUZ)e1Bfs*MhBxC>i?c9+K<9q1RfQG2TV7+bL2^ z#g_$O>9SGQsA>|otF~ybNb{z5)60|zk`z%I3!RwhBJB}yCXr{`tIQMT2HXwAwz3PNq(*R z9y!I&KlR|L<+hF4xk9m*D01Pk4mLU$tD-i>mB)M(X#nz53$#h=3eX=8D={BN&6zgn zEdP+vBhvsWyiE}`)7r#X4EY|p>x-*pgF0{e^(gG;I_Nv(h3YkU8&*>Sg{w(fBDkkI zEyrri@|esSogkg9zf6p@L2F?MAJo)~iUi#>5<(A$ZIcgqVR0q)0PcX|1MEAU;bT;! z0_)2WY&(cXJFDd6t^@E@-L6sFuYWw&>Ts-Ru({eE&DE$rGs?Os%4!=o+70v9}e*nKh!N>JNfVy3|SK zlkmwPsw%n~^2jxn1WrMS!-|X+HpB1Z1tF?u$4|d!&*1%nETK7x$XP(JY*gN09zu|1 zfE3HlM|3Eb2VN2axi{7XlShHJo5`G9AZcORnWB(AfeHTSr|g<+P-6=>{l)P6^Z^hD zzzg)9(ucDC7p*%5^u3rSwlAhhb9ZX5wnLTTTdTP{3yZJJ*Dg~U{4$uGbJLm1sc%EB zsXS(1Xo62NwME;`q{njBvT zzZ5vts z`}+9?b(MODG=stFY)pjX{ctb(V0Z$BeUnKR5E1njk?7p0m8gJW4hxmX?12h~-2^|a zL)@t{__V{*@k#g;eMj_`8SP9hy({)T!3j3SAD?IWT>Q<4i657?{_q3MX5J$!EB-gfb9%Kbq+&v4JCl^L)8Om2ma>K|We zCaWh9h|&xrwd7FjCW>T2u{1dV!Fuy52gJ0e6-%1o-4$pz!m*Zg(`wSLMm!TIqS;a8dE)PGqH?G7oDbSYO%NbyM%B~LJ@S4>EcLD{_I z2^q{PHX;1+t&L%aLq5C7brxed+C~Rt8z+4D<6O4Q3&_+DE8+s}**6}$@aG_PRs>be zMg1T>F74Z3l7Y<@1vW-<&&WSzs2l~wQy$@#Q@lViK0EL~_d1i~^k7;aM~?G!HF4CE;DqAf*tz5~ZcB2&_}x1D7&5Pgk;m5j z5{`(or7C66uhZNV7Anm1rniYJHN7DJ*saW+s4w*?^)WB2v6CyxY+?IWC@XJLu3~ee zw#oNv9=NrIBOeUeI3|`Im*8cjhkPsh^mY3NB_8W^aI_pjS@@Gu4Cd)`MHpqd!I(#P zfK<&zHTW?#7JSIvBMpe#j&WOy5mr^>c7Hee*FUo^Ue^Epm9LQV{IZ-L?04L?nRaec z>DL9I8#=BhmHWeANR z2$3nA9lm$ENw+Z=pVKAJm2XtvH6O>!MP=N1MP^u|xI1tOn>z!q#-mX0AAacNYw`rh ziIhT4WH2rp`f*jpL9glPoRCeLzOP?cP$=w+u^hv)mra`L`8l#js%N0wNdq*3O>*?m z(-EDzsYhq8EdQhz9vWt-Tdr;6suNasSJ7N(^fIw!n5#z-F$K8(BPXm-@`JPN3nL&vxblZX+=C1m$$6s!mm=G3l1X<>cT)|G|7J7sOebvbs$`5>f_ zZfDX+3Y-2?S@`=icexxBPY%Z@QS#)2TzrHdO8)z-Z?e4`%;9ZX`hisYbcpqTTl|JW zoi*t^F{$?iVv4L!xQZQ<(EVpGi`VP_>@N3j@Ou58JG;+}i#-)+wI07XGwvM8;^zi? z?!S!ewXrqD6bp^}wmc^L8T^uFpPy3B=rK!OBtSp&D7`H1WBDo;iRO?nY*wE&FS#&f zXrz&?*CswB#8fi`e3$4>PIF(s9A22Wor)}bZi2t*$Ev?etYCS)eBytR6nZT+T5N!?I5uK^cp|ds3&}q0>a!zMZUkltmbxb9P z`NG3ZXn5kWA%(vYIMegfgCn^dJwK6aP~(Gb1h6(Ghc>a$O~s(a+z^@#lyh}0jo^uQ zZhwFuHtg?q-~Mj|VZQ%nH67z0{=XTd!-Kulr8Y#UdlcJCk&mhPovJhutZYC^p z(U4b`0B^mTpj zrLTFReYLN7P1}iS(l)r^3WAD?=m5$hiy-2LvMTNy5Kt6xK?XF2DBwa+;s5*143QZb zGZ$w3-`LlDs(0=^_sn?D_x{fB{Lc6H{dqh9c5uBiQP!uZnv8vI6?TOg+G6{B+2ybW zeo(t@x1vOYOqNpt4FOQv2Yaq1#%mx88rG=37G|Uz8!nxM>TlT^ELgmB8vOP~buCGR z(4jso-{+#|pwkxj23p%kPP%IsNp3K3^lT8I#)Y_fzhJ3w$7Q2{$5WOSp}wNXGmwVR z6m^!)aVZt9h(13ZLTQp7x+&`7)7@1UNm_$yUL5-S zk#zWBy5_}vKfHDrBD++hn_$6kWa2-@@%aR9tUJt?ZXil ze!uAcZ!GKWJtN52Zni|uqFBht)lso&zCEBYwH4ABZ$kcchMI- zC}tsA4AGO@;@ygD$!2$Lo&XoS#+n$7!7ahs)xNluTuGO|;#7%(K7}@mU{Iz_98X}e8+lD|##;}m+;`jxDLX)GObB`$Q>HV+ z?Z)ViJxb>bCz}wo;9c+KB%K?Ac)SRPB->$e+hU3>q@Z|QY_;OUYfu>gh$;aI?pbb7 zLI8@J>{RI^vApCNMfZeEzZ=YX8SXf)1(nrf8jyAlUC!>1wuf&ESmm)<(;Sm0=w$}| z4$}wd0o4}wK|ed-%k7~Z{|y8Lx8LX3UMK^Mf=7790Gt5hKe&LM(D#4r)oRid@Ox~i)+hcOjt?w*yV-W zeXY`>u+0$QH{`dqD0)@RqpNg2rZKS*6l401(cIkj(SMwpi4m}P&>GvS3nC(KqxFQZYNBs$>L0F}26Xp%&j37IstEXu+9bv6QJL~mm> z0I>C*Z=Wc;Zy7S-Ef3pD5X3;bKBwqsj%gA?fjICoq_j2*7WpHO7DDa@O^?}6Ynp%u z&uy%29jURmZx7Aq%mj+%KY}egvwCbOtMmxC*Pph0m=ipbR6VGqoe^=|I`}EwaHNJ z4|_E#2w@d8tlg(gB8}?Rva+ZI-}ymkYmT~Hqnp)Qlw`oKZKF@gbLn%^E|B<06eO^n zusMC7&i36t4bwxgKkZHx%2dtuIp#j-)EU$dvYX2^+Biv<6uV0jWzit2r_`%br=J=c3xo3JZuS6M6m@F z*+Rt{;REt1%zja4*v5eUlV;B+n}DcY8T^R2jXurnY@jOKaveH}5q`sFF};cI6)&Bp zSD^eywd#`)hnO~QHf+Rc%OTMh>o#oJYvaD_WCBd#=fC@q#Bp02@i-o%GlNnR#U@Z> zDG(nf>JD>NrVKR8!On*=As;oE7wdxheCh z47=^+VEEVql#y+TV?nd_9rxEPB{R4zmv{^ms{$ME5B#8M=G=d6{f|Wy25NVbc#z7B z3#y1LqTkD-l3}`fZ@_8If!S1gk|xZs7yZi|&v5cEj$VY&RlT#2?s(86YE&>z*Ng1}HeRL#Q7 z=0@OJ!pf^ox+=ILEN7;6o#IAFC%wiy6L<;^wJT#mi6hW4)aN&2Hl2OVcgw zO}Eh~>}GbuuBmb1SU63mHr|#3VsB?A_JzzY7M_q|mnJCP@M@4^thYsx64b$HH;9{} z+L?+;aT4u%MZ3Dkv(d}&{mWf*A-7)V+Ze2aqSH-LU9|1~?^uBH$e`GMGz5e%``fxx zJ3?mJ_^6dD)!rsp>8FQWB56E^Rhb#A3Mh69MY18R|NXe=%HXn@D})f!PhrwLmQG7y zFsi$l!Z4jt#Rs8;=k%%t%GnjJhg6?Ts#A4~j2N2-l9f93g}}8h*gpPWHV)?3$Bq{B z{_ORM6I#YKe7$wnD<-roT=m1Bk*a5AS=4HVtloa1TrBw=wry`63>zTvknld9)i}4_19qo!qgHE`s zeM!F9Ra+*%!7O1e$&q7@i!bd7X}%2S!o8YaVLg0hWAL^B*q=%g4XALdDxL0S%6x8m zou9sgqxAjQ@n^H8;snjxCp!BBOo*DA{!uc?;j#UxHN(g*iUq>uHY#?$-&(~|V*Il) zrYWX5rc`=}&Vxb)*GPu@5%GXii|0;QGvzVYNWAj2tWD4W*VcKR52{z60zrj7`8u!d zk<}y->Vcl|?qP0uE)vxNm0`e1PoMVclG@qdu^$vTjr5l`=46Vf*Js}Qp=EI#-bR$I zl)LGepuI`9g4Ot3(c{()evjkQ4WIb*Z9&8nJ8;lNPKX%qbAI;n=S-%=?++DklXcEy zm)RS;g<@gXXA>3Mt~%ti)uYUP=p5wT{F&3m%%Cr%^6-$2IS#BX$2W<>MWcv#7CemnnP zI)9S&=8YL|r!Zym+No)wNn}Cj+xe0-?@IT#b?O^IX`$UB+>>gKX%jDWzAWAsfU&#s z!1&O^^cugWsK%I`vj%^-?iKsg10DfrV?nVI-6M2rhf~~8-1L{oMA49olKP_~B%Yho z%j5VAlx`fht+a+>S5YL1ie2rCrA@ZFYOAJg2*tIL7VCrly;GAGjdV;r>rm@wF~po; zWO1(}KmN;4fAx#85WIQ3No1w6eHmLMZBbzD2iQgKRK-bpJhi&$?Lql3Gz+z*0jq+0 zA|J?6F&6?Y^`Nu9)%Bzp6@f}4P;DKPQ?os>568)iaKg$+GpK_A^Y@HJYD!Lo+77^rEF!KW`DV)ofz`m z6MM4nMNZHd7xJ6V=0KAVyS=Mq9?AWR5ugs58I;`=3!H@#Di(9bw3r6e#oS=#NzPA) zplYMCHzb8=BXQ4HxW-BLMe5WT=tlw`T&$!M1-LPU+X1Cvoj{cPTu~XUZ5QkhJRmg6 zSK^j{;pFGAk#Gz$813k-a6*&yY3GY8_QI5PGzlxGuCPp(w~}X`%07^HglvO|H#}P? z1mCGL#K392KMHWwOfQngg(s5@E)6cNu2@UzpkVtnomvO7a}Py{B1hZ#Q9pWnJ!32V zSby{*?AP&kEji<4{5LGsHmsCXZd0~&^(}gXD!Dd)k z+~??zoM6G!R_x8(76Hlgi%@vZ!EqKD;7$Siz#9_e`9mS?mWE~Z|FF>cc$5cwY zgVOyI1^sRtJu5^vVfQB+*eg5aTiiFxYXrMr)GK<(X)nF_tiaKB&jeJPac=jy<;ic# zd+&Ut9E+F&e=Q=^up17LFT@m(kB}UBRke)StwiZ;J#ZFo0gZCboH_ZDO3!_oM4-2% zOjr>$pz5ScBNABD7F*_iQ1ig;l;(llv3TJQlc$51!G|v`d){P6zF+WPzatyC*%2PY zWUm=a3Mn>^f~qo*tlG|`2%^J9zv@8VlPnG18rmJ;wQtqmL%Q6fbRwRvn6$xjFaRqzC(4@GBjU8^1whc> zI`M!oi?k@tguxBjnmj)|y({XJJR8{jKx`8>!g_v;z_SfRU$`9{4#oyDZXl`)`_m%J zxCw6!ipeHIZKQ+(exwb;>S!Eu}4xq|R)8ZPTplGqm$0&4P_UF^14MB>Sm!zD3eU#Rc1$ zbBdyg=cn6~mT21m*=*G9H9q4w4G@1QXv#Qh6Lt902?H%?>=n4Y0jGKC;u&qC6u>hE@z?Mrf-vAr!oo zN)XM!(hWmW{-BM&Dma_g`KEgH5zP8RY^5;Z)Ov4wPNz%r@(2C+vm!zwryp0PR<*wxN|UEPAuef!fQS z8$C-S`rVg%oRDHUAgBstJXK@Rc1ouglYH52af3%LokLdAj76_5Y zsMsvvK3`=rMD@y(LAr0vnMa{^Bod-GQC&Z3`IiO%*8SRrpJsjg-psXc7ynl(F;wq< z_!~%7{z=V@?$It{Zc!ZdI|rc}gIFO_<&en2C2^H2q->jSthKTo zhMXWZep*-A&%WxZ%a}cClyB1NVXrTR!Cf6JCv}C{q^TuFfb}wWRB53LLQ62e8H>~* zd18%w9hRc%k*9>V8I%ae9Cx!ePGd)lwVuazo*P>J*nh9rGSL5wIKWo61<=7Y$|~ee zUF|zhf<>RYo^l;|;|SY?ps|5rRCoX2C$D^D>DqF^1dqunR-E3Iu%~#E=oR<8<0R$& zRg?4mDq;TFEZ^r>K$p;!Q##x#rDp}5F?XSM(C!!a1uXQgl&&Vmvkv!oY=X%Mv&^W* zp2JNyH`YA&Ba`ovLG^!^oZ_+X@~N5ca*<+NDAG*DUMDNbHsJ-qLB%FT7HNmoZk*&M zeIukf<~%!h;zH*;Aw4h0JLd@wk$Ds9qHYB)bRL|1EDZAnb?UB=o^Wu5CtveOo3=!B zPKtdUraMIOZgVC2Nf)N1PdOL?kr{D?x;f_VOYKar@RE21aGNUB6|Oa2E1`8sWGTJE z>o{maT@t5HStG(8Iz;v2^Md>E3mOIOp7Z>BRSnW(D(nec*$B4LLS_UM*?7t}+;Ezk zsBmh|ILZXDhwcwgkpU-89q>0CNj8|lF_B`IQDiZaTTqK-mueHMJ>Ye~3$!(C9B|tn zzpbrofsd_y*|VL9`bYD06CfIYwji67Ig_(yPx2_m9-_zrD)v}p7rn^;miK^DmlFR> zBK;8XPxontA_&PM;I)ejBC|C61Fp&ABpWApNo%3{aTmSNd9!F?;NRk%Z$~6Uj<8OR zNyE)Cx4cg)(|oaujh;nxt{_*jOx2@24V4l)#oL(yr$j-G&w!J0Pa!q%E^{bs(esCA zyn~&Le*A111RGz@k(!NqTsXag{B;HLf`0sSsR>I<(k}m&B=C6l0$JN(oZ5{PyPhIz zsaT{;#(Ad!0&_xf$=$2ErrZq~V%s502m`xVomJZc0%D&lxUPjq4G*r;+y-op;oEG# zl@oBsX;yxmAu!?Ne~;chLgw=rA5e~L7(P}}Y!XEhsMs~pt0#B3_Q-d6t(wp!Ma3nI zBH;?4MP>nsJ3c^l5PB590i=&jP#zJpE+y)MczRqX!@ zCBZgBVc)^Ftv7CvaOeDK;}JnG_3%N~2;6r<6sl)?8)V zA(Q=#xKdmQ>FB$K?M$cYpdw@P$36{isb0-7)$~HiEk%#8+wCx2I3??;WH8xiWX$~` z+dwgYG?bl?N`L0J`a49LGDOK41!SS$1BK8hFN4r(W-$ zP8TY$T>(81{xr~KGopqYk5S{q_8YsjU;mz)X?iL9=64#%G9LS(kfJlp4^5-kbrd|; z*ln(z2I@3wS9eJh1sR~8w8gcHMk><{uP#U|$F06buSQ2};}3i69LIauosK=^WI@KA zKa=fZNvD2BLiJY2*9}102X{lzuF4Y@(K)^quAfb8bKUBZCs;{#McwvVIw2imv#42| z!t4yY-;^~jbu7F6SAf2Z_SwaJRa9RF20$>gyWfmDrQ zR-}Ytizrfnn!{UNYo^?ubc9ZehJsVhsP0$lwbf&>b92l#kEV%3*Y}ggu>FCnqbh?h z`DN3$T)=S=mw-a$ZsF~h+r;VgcBssO_f^x|JZ?{N1R_R_f=76G%7uMK;-(DD-P`{I z6K3}O!tdY7@sY$)c)WbMZib^4ifyJy6Ugm5Ckk$P-cfDzhyj^AAkYb+YME+_{~=NzTrnx#|Acz|#0-~8_dYk&;<`QMPRM-0QIZ7P#y~I^;zB1; zc8lHy`N`dGRg)Je>w-b+I3h6^`)wB-rgu!dIuZLu*_9otKKUVH{FdFq!iWK;LUfvR ziz-~}gj@V|>I~TtRl9emrUBx4wqw&Sh#QW+Cm-8vt*p;xtn(3_(uA&K-~HZ2vXa}f znZMc&pnW~uer=*yU?OawVzKC5qM$S)BQ#rcfoy=%?zdd(q}wz*UR(~d6MojN0|mJ4 zp7qK@K6-H*jWm}G0F~`9uz6f88jcb^7QNX2^Fs>cm-eW9S;U<8%?sQaz$iA?&78pO)Pt^>o$bxdLpTOYc_M*J#*` z=O@n_c=d_+|y49RfKaJdHM3ogrhfEu4J8&LI;W05)ZZbj{+a_JtI;%K~X z-$5}xEV~R1*XM3Sdu~I+TT@e8@wJk#`i>9o6VH=C?ig+gABTMP62a1GT$s@|47W{9 zY{RNeUD>q}C#;T32wFK_Y?^&$o?Ux@EPZBNSEw8{Z0Wt8V%Jh+H5H4LdP7B@eUS2j z5&uJEKsD%eC?HXW%xp+>Z&9F1ZMUlRcA%^k`h^l zH$=K(spH5)qJItOWQ*0tdmQk@NW!xa$EbRBV1!7$2Bdzqf z5n{_2+VD6zHL(d3oO-lRV@{YDKc^^p{)9(2-akPdfA-tX213J`n!p;lbA_%OY}r6{b6Nnx_! z5=!`_FxLcI+$)3k2h=M!xZDrH4ngsuMX_xv`~jjZATMS=r8t|Q#?{i;0d3^8@{Xkz zDQ^vm6?M@#$r|t8$U;Hx)D{JnYS)X`3sB&xLV8h_BwFKr-Rm+WZk!~^!kb}xqHBDf zlBwZ7<5(o*hDNMJUHUrqQiwWkNf>y!Q98IPW z_ms4sc=oEcg)R{-@yM2g7|bU2ngA7?I_ddj6PS@i5e8%nDg4hd_~8zh$IOKdgU`JQ zC%BAT^1?;+SXeDQj_g|@#38tL0VygY6g5eU!cZozM+}k)S)d7)JGBuKSApW-o+P>` zUlp7uM+$DP=Bbb_M~5ix2;q9i>u$gNl_k6SarF!yBf$#eT2F6rzaU6pGeS)jx37Al z2raIS9b+<{@ZfD7`zLhakW1FBNm=INH!NwAUnx{rp-$>)oJT-$Ot3QmjX{e7g>qKQ z%A%f%5ilRcuO2MEcIbCMwG7+y*8i}wpRk3l7nk~Momi@ggGI_o^*Vnoe!$Au``(O; z*0xIbMqYpf)IG1khq&k?s$4ozmP?yzWMh0ST?T6s)Vjt(*bh8(YEW@%58L{JPHo>v zuC4fZ%=H-IggNrY_Q|^07-8#j+~ot!<0qI_#Ml12_3vZ}k9P=j&DOZ76btBHMa5=o zaHx;NBD77M%=8NfoVNJ&3oBeP(J(y%)pyPc5(A&+yut|^V;gQxcXPr9m0f@Brkrs%u!ONvAH3Zeg`3ZqW$BZoR@32GfZLRJkBSYSeOTXBH@R z?v92GCzuR38=MB6dt**88AreKe!*A0HsyA3KQR)Yr74R#KyM3M@0Adl7QL6g9Wmgv zLsjclD_ZknE?p8G=% z*J!kzbe8-+$m(u`q_LIMEpe6{C%V0wtzp_cpi5>zsY{(kn=iw5mEbJ(GunztsZ-MF z&6)&uwQm+AwD!B5rnM;LbDT6x(k=+n`I`E!C9R}h85dqf7I~D&<|&p;0FiFvwRHM~ zy1}(wT^ZcY3|-QxQ8g-we5R~1{(yPlp5ac`b^!#9Q6;m+`vORO>w&zdZD%U!IBG5^ zBvwsM^Tpr$8wUW@9JZtU6n3TWj^!Npm0C(!nfH2Cr@bDAlnCm*d&sBaPefbjMXn$P zm%^+Upsa&VeTC^GD4>x|Yq6l=UBPPKY*>CFzy_V>*vC4JU>FRi#sCIuUE1Zmb*nXN z)zzPU)n(RA5`Z8Ww>{u0eND7205@;!C_!uc7}y8^+dH&VOX~m_VeK@2W9Q!UzNTrb zaPVilNCvlQi^pNRO0$*LHj0Ic%=uJom$DzEpYkP}-Ae>lJ+Vs6P#Dh`s==+oKC(|e zPqO!=F1lV>8li{vc%PyoumZep{iHs!+V?c^P~Lq}4h=yvIg>ec7a8 zCs;k|rhmTXHOshnBs1}OqJ(5Iv8aQ=T)4*3rfvRK`>T1YkY-g%f`~0?q z8m1-0gmGfJ`!cM}hG3P2%Qi={(#HmrM+0l`qlV6Rwa=Msi{BqA-X`m~*%ltB8|^Z) zEn6ryn zKYYf!YH}A1*TO-JU7aqwCeLszt8Oe%u{S)1_cwdoLh^MJRQ~;XLmJuZ43d4rmmS9_ z_Ao_iF|}r2Knc*{7y2iJ?o{2PwdH~Fpml>$4|q9^DWQV_6_fBa#TWFLqF$^ml@96K zmokG+xUR)2*jT2gGwhJd?6`1baqn?yR40kBJ<5ThN102{lcb}@tZ{p!r}EfQv67#<#T|Lr?ZbH3TilUho&0K|pfRTSb<9m`j=?i`r4QsEB?{gj zVr)Cyo^XTDqv>!|nBZ_z{08qL%at;3cap5EedE5_AD!oUj@c|kq5Z1BJWoTJpDg*& zX}i4Egk!ujA9kY4{a5>9m6y1g{cc!I8EQSKk>_*-)%Op%Knh4MT_EX%yu~bf9^~QL z9}}Ym!6P^rjyMdjZB4D)kP(6_c~HrBVwG4kQbM&siXJ4q`l=*Ntdg4_~lPlj|?9_h0if$0Sc%9pW#?(W7a_Y9|i3kQZC{ z+0>yG%ZKgdWQ#Ka#q(j0bPvVuq{wzE*08ZD$%9N6{I8#he{Qq91Xv+Ed~Q#Hpf!!x zmMIla654HW$TY|m0lb}et6RVz%|%x_MI;xY2XI z-(_2CW7rE7w&@omI_>w3w)Mmbuhf73b>4K#dSP55V?2(Il$jaE0*Zz7_H4*L1x^aq zheHx=@=H|{DnwdT6V#; zJ6u0;i*r~gZOpdNHmGoJ^^1?@*I6p>JR@kaqSI3mLDa1>IA3x=m`+zrf;b%B&_-j2 z6~N}U?>c$Ruz56_a)!o}Zus|~M*iz~liRXjQP4p$$gPwQkB!VmGb6KtVwY272^Cu< zEs|oE2gqb^7GmO{y@&cqk9*iPTM#j<`BOXGtW^GMFY!my`mrXCo3i$B|UN= zA8z#20n@fpTtz-s^vWBga06NzV{`3oX%#sekmCYpTNGHuP+RA+8Q7Pm5}54_PGQ{v zS0fEPNuBTA2-E;Fv|i?(FI(*kRh{?x#|7ofGc`RxdI4=_TLUt-fa;MYvB!~ZxcuRN z?C}_jFT>-Yy%k^P{?}8<{%J?(dU}Cih3_I!7nDk^)8q?xgtm!O1M9@S%q6i-UFTLK z8gN?g-%W4yM2XYg!X_mawL?+Tb%IP`LsXO93bsE!Qe#BR;~m=R%s=AZzG}H3;^NEj z7%f(oPbZ|8L(eMO)hm2=xf*49PDz`A+FwogDeUWr*f2!4!or5G?AkuCMJ`aW1p$Q$jMSjQ0We@NUNhj-8?hR4#y0tuMPb7FfYbJ% zV~Q>Gg3xAC8quOyMhZi<-EQ4(2Q=;Kv!q0r8aR}xI_QKxueLX$EJ~+d5Sk96$0^K; zSNh%hqX(V-2GzS)fQN&@eNb3{eP@`C@+E^#rP2#Pro{R}deBJ*qP=gNMBAv=hb6G} z>SH3?+v`8+*(;kN066X`EKw!6wStvf%oQTiCkXG@qKU=Ql1^+vWqICn8g3_kMwLa)=pbN^SG$23^h zAN2^^wEUgg|0Vj3&pb`DQ{%69%_gbb)=m8Vml06*8n%y-N3pnT?Xl}8aEqT17{|P?&Tgw`!&xnI-r6Ot{Nr19+NU@HS)XPrG2Ao>L-fjs# zMCQNLMlQGR4AHDdvqf zsvFgbf-Y&Q=4eooQ zM<*3$6PSi(kWv0U8q9G5(>T#jKRxr$74=7@Cxi^Bh@!^&4Uc?>?GQe3_-w(?6FYFw z#lKEp_`=sr-fPS2*K^1YXVPHi_SI5s4Mi%4IP59RJ<{XZqQKCB0HYAObdMYqEtH4p zYSjXEwvk+Tf#goDRJsjo;~HgTejU3_MISb(;7VJbc*ay_le@X#~7q>;Oj6ZL$x?q9&1@X(@Nb*gO~)p z!{@4JIiv#OhQ2`(nk0Q6J z*mCJ&kF67N(Niwf1!k~NgVnWJ-73v@T}-7guoS|~Y7|BS(MoK4P5D`fPK`-LM_oXx z5?AKAQ@^-|Y1KnlFN1C}N^GFG&mz~WA=>)jLFxX$gYqJP`d17>(yLRZ7o=D!RJH0no7W7>c_&&_rdz~EN@fKx= znT2Vl*d_{m{@7loU9eLySJXf+f`ZOio&iaisZ{E#I`yuo&Y2hGZDc2j2T`Y1CEL^Q+AA#-#NCW*GJL@f1Fb1J zqEJyQZwy`&4Pv(-k@7TnaHAXwE>Cq-Z8;4Ee`kn{P#OKY$^4A({mrN3Adk(@RWtL` zM6nQWIE^KU(t(E8rOcNV$oB;Rs}}--NYY#{sHSl@X@9^~74BINIw9E`5so=A?M!b- zzgsT7&cBDWlUC`VQ@$)N9H_mrjL>?RO^`8Rj3*|-?53x)>7Arsm`m4+^x|G&=42hn z{?{v;)jBmQv*MK4?_MdydK80B8Lp|Jagr|SYUS++uIsH)0^ZWB*zlPf?zsW}=K~e1 zENu@LH=Dl;2x7&A&hqr2`XG=Ub=HqQ);c*$Y-S}|+=#51JAb&p%Ks?Y6Hw+670nBcIMtp)N4Y@%S%rCY!>ZDyg2d1nX;kAS-m zB8@iX&!ux` zs$&9W(gKX4b=q=G4_O4E^8JC0%Kd>=fqUfH^cGOt+E1#1F^KCpWBr18V9dH*9qjIn z+DySz$JzF9xZ&q_S4R4n%+y2n_!6RXCgobxtVA z4}08>7<6hA*MrJ?Z#1rB@dYmrY*eSyuo`X@G(t0oz8wzBNF#Tn-z`Jd&fKT>lRQXd*%V_mCl@2Tl;wftf#|3swHPLx0!h1|Sas?p z?K;JckWPB*tae_K%a?fw&{eG~!H!{Bs@9HyrIR+Z3q&!VV5Q zu+`3uy501>HI~&Izfw$dD-|8rC>jFt1ZNaj^8}aY7yZuAxJid`r%$|(dtudqT>63A z&VU6$IiZ+!(9Y!ffgWDiSy-A6wd@veoYWrfFmR03c8q$^zvM>ilV(6w)xAHO=Ap$) z5~D~Lk7GN_%vMYfDYl;?_o&z=uXa_Ps-5Wl3nUAji|8ERLvHKc=DEi^4>&D!MuAwJ zdR1_Hbi8xAe^*FPxK6#u`Hu1|*)GKYFLr)6Yvz`pub5e3`twiCF*P3TOo3#PGj6l5 zaa-)%7=z!D7Ml0X8#5ATy_f&95dS>HoX zKWK`H3%~5g+clGQtc^sxKmJgfqcGu5+AV#DB=H!3MP~C<2E{@$_UY2Uaw_EKwfN2RH}Xp%Lr$l`sSP77S3Pfs zr?}=S5)Dyi2ZLo)I5GG%jzWmBUrsm~pZceTzLpNM70uVUZ-^&kHD zKe2L}bs`isl40lV$~*E{--54tdC?mls5O$UDO42D1K`g^lF;JsTVQ?b=p$SS`5qW+^VM~-+?5Ra@D?FD5 zQg^2+$)(S&S6Dx>R&GbI*$HJP$k}gFTk%W(ILqu!-nuSU@-dTMG0c_XHPdb}t6iES zSNZLc>oi5Mmwg(_A2r8Ru(uT#C2K&Y=BQ*rP=%<6J|JutAjzgp`cQdQPz@QG_5YOl zn4D}36}qUY%W?~WOMH*VnZ8zbLK1@)c<96WKt245d|GG;)v8TQCgvgD9U zV~p`{n5Td+wn(CX`CwqHcdK^}T^-pRb3#2oXsalP{=nGCQEMtY7#I6llwmz>`%`&Z zglSUx@$P~QQtC|V&8+zWimj$d1r=K{snMfISR`ENJaiQAoWeYa?(%LE&x3qx;I62W zy)QhETf_sZE?{@2(}fZEsW~RicfNm}cnOm%D}Nr|_Q-V42A3`VbAiOSX7Yot4LJ2h zY!&8*4LB7l>cqPhka{ax9J=kj|A&@G-a zq3EIe!&78{$5P}Q%;23!vCAm3n2NpSl1-yl&BKs8AsMhWJ$BFyd(CV={Boq)-j_XF zazg6(1v4uy2ABXbHT|PxlH&|Dw}x-f?4no*b8MqxuP}>56@h6IV3+{EHWMh~=Y>6P zAA7Wkb3kIQQnV0EMOIU>C)KyZd+4>0Z+ZRY9jb@m`VJ$EY z9(5JO|1v=JYSo0&2;7vahSSeAlKwgMWh+OYmO|P5WJMnyAb4+3AcEN2$w`hmFF=i*}7Ts1`Us49v)FZHFW$a8ZNDe)5i` zx-f6ejupM&o3G9}HRD{&u2*5^*e+N#p_r~FM@Y7?aa!XvPD5(@L$==1c34>N!x2}w z!QwXCJjcUibADX2@hn-zZRd!`QQZYIGbpl=irwzn=hhtaFno>d0;v(yDLPzB zRcmC6o^Om^g(DKq2Y2pF05 zxo`I?%PbSTzfz=4D@ot!^n*a1`a$%<7YsRX6|Q>mxrx~{CN<&y-~mxlSf^K6)O8y9 z-~DdL(j0WUEH9*Qcp1yNr@+=>tGrnVS^RyZQivIpH3DP+SJJJnJAnd@T5ga_3CX(ciSaz7Ie}1tES=0XcSj)|oAWP62_Q2x zar$5e6YC$W8>M&hH~#0JOa_R_W+db$Hv`0D+IWH)9dwXlA5i3TD)uhjFMRN|3{kNP zB`0A28I=GbrLqyEka{EPm95fyPc;a^Wpl_|b!R_KS?LN)cu(6 z7B{Mr$XVFRx<%(fu3m=_e@zZu@!UzWH7rYWQL=n$F1-&*P0a`QU#%@Q%-=)P1CG- z{`yIInzIoXecI^0iKVJsU=$~?bLPy+m*gE4{yE zLdrW$>)#*;c)Vu6Vm9_4Q*0eYPExTAQT@_QtbWoiFU*uK6c~y(Vfqq^N7Sk2`=dYv zW_;r1?NK>Uoh5-iN_6fR@6qA``a77{psiQtKn``Or_Q}P5I!Jz7z0A+`@qLq$O_df zl0^7%5MTn7b*~WGqIYyoiXS^^4yYPCPM!#&$1d6k z6K;EG<8LRXCCW{xQhoGlBU#~0wweuII>o|fb1D_PUeGN{pOVJ(Ov&)v?6b$GG$sY| ziqd8AZpH4UG0idM;DlZf#JimkTnS0leAwmxloraP83JoarrF|&sk zDE1sh&QP(&B}bb&LzGT073k?39w@DB$X&@0Aqgkhy+nG5lt^w!P{{<#A7%qgv?Z=3 zuCtqAQFYh94|JAxc{Qrfc!TnKy)rjo6$ss+^7B0?x4LlZa^VJbzMqlLo~420U+56K z%hue73W^|Aa-UvDaMkm-eqjTo1rIu5-^eo28-$xo%RS7XK{-}ipg9J3JvL);IPs} zD3V0QqC5v?btHj~Q#M_#7$mb%s%}2n1OeE}U|2Cc)o2En=Rca~INEzSJ%9e%;&(p2 z`S8_p<>y^|kQRL<>5??nU%Y)z*onnsV=b&u^D42pV*{ zLiV^{fqE36w&a7JNX4Xd*-lNTyB6blc9~VkH7!73+Mma1fcN` ze?EoipBXo0zS$nb9g6Lt$W1EtEYPq1c9l6eN5REY!g& zDqnU>48AXD5`zn@^VMQt2JLFfgxR=5bx}T`LN_%7s&i{UQzkq%LhT}PC;a*(@`+m~_)*C(?eW1bK*sw5 z;qA`C=D+Xp`Kog+Fya5iq=S=J_|{C(=04vjkCSw2swU6LrE&4{bmv(vqDGHw#Zx%q zWRAxD&#GIO#*(ee-JfHweSzmh(A;2POSx#lmiQ7+giAB3*+@s8wZF>QX)%uv@>^w;vRCIn6-s!Sy{-#|gJv*W= zrS79*Z__lUpB{3Fq;Z?Zc+Ax)GxM4XC>EIR*;H(jC_`}N%}eqsQbZ3xo^F;~wyc|J zkZuNb3@GdX9@KIn6v}cv6|mT2e_(^OPKmt3wX%GlWpM2i%>_Z7wEM*#db?n7k<*IkH`;-E|BG3 zhv`AOR+3HkiS?k}m*mmT#s&39H^=BF^)krD!40l#$#P+VybA8h5bO)kX|RWMfAro9 zLo3+SDV>@;NjFnvJHo~Y7~9=!Q%iGLSpURn5~D;By9%96#wbSru9nQ@W{h|oSxz@I zM#&VLNRee!tVwSe3n})Hs>#~Zns~R=!1~~)8fq{MpN$BzKGY7n@;zBjw`DHX<1|4Y zkGz%qsvDwJ!7Yjg=`j^@t~tQpYqbE>j0I(RA5<-?{49 zwd9abqNvq(_8Rv(x>9_LPMI*u^1F{c<&jTt#l#ojz=M&yw98rRm>6NpecZrHzw`X( zqT}80Fz`RCq(xHHCbD1jj_c^z1|Y7zj=lYV{^cF%%O+pr-HMnRQpn=~@G&#*p`2o2 zZL^DtT{WTA6{EMk5gDP!*@F@1r{9#`jo2|w=bP$fsCtnoz{_nAMQdkFr*Zv@``Mt& zh^>%-iwyJmQ}4X2&6i+}i_erBeK&{h1&*&4-4irRmBAgtxsrTWdrU>P0m96VeEEm5 zQpF1>=p7eSDb*&*dcgMIr}6deQ&df^l;ZtZ!Q`a6P*^~h(3R4w5QR+? zXtC;B3WKCLENKfB0R>vDRfgRmDHp3qk0EmJkO3Puq@x`>PE!DX9mE{A=w!GFNpF2T zZ7V6^w&>w;1m?8a3sFO{l@!?rbeb7gnL^k_$YZi;jJfE9xvsf_yD#SndR1K^_eh?m zE9A6SwWe1XACoEjY;wB)tr=&bX4yi?=gR9ZpO)mhHpW!bpDRyGnt+U0DNKWGPW*&- z6q{O}lJ6G{wLB%qmd~>7iog2SHcRH770Mgx@InQ6uZvI&yCzb5K!{XpV@^>UZ8+5q zL5hWDD^V=ek*)&!=+LE6a+y=Z)!$sL~Of{ld0ivDE;4F-0LAUnHnd5r+H){ zH&eske;L_qW@=I>b|pntP_fw>Ts$QT9)@6s)MCo`-&M~7$*%7ff6E>w!8QOHTiFb6 zE}hz;Z8-BYS634*KG2@|Gg<6B!YYc#6HSiU=&z&L6pE~*VwXp(5bDE>e?mC}m@G=A z#ey=ZuDg=tMpP?GA~|V=jQvU53;@S2o*(;%ShsJ#eYfg|mdhkAoNyk4#>%1!g@3oJ zGKF)giph1#!@vX^0QH)sBvG)yZ;h-=iJ9KDq*k67UCf+e+QjFlZwx36*g9>ePh(h} zurchmupn$DSxWHUL&93vJGd<0{h}VyTgs*V^fHnJv+aJ;7wQBO6kU^((I$yYbM!y@Tu;$CkVFwWN0I)b1$(K24Jd5EmP?IT!y{?7S74k<33V#3zle^!s^4Bs#jePG5kPEQfBZ({sohL%Ke20q>+m-YJrOrKtIVGC@tK(zP@7 z;!ni=ZX01EJO#dCmr9F2HcL}WsysjUviOa_-8&dqNMq=`N2&8o2tDPxGStW)<}HKQ zIu8FF^YM908EP(zE&c|%t*L1vjeJ10$K`_bg4C#o1l7`IT4SvrtlKl_gi+!y^_k!d zQHAU0s%%&RB(S9sH3DErpmtKC3>9MVew{ik^h$8<)C%zZ+Zm)|*2Sk{ixsiCJ#PeTwa)$Q_XDfFz4_`h@z7&(+W^pjDPmr%5hC z+9{UhTlm7NU~P}`inN1Ch)k4qs@mv`&|DfbnTv(h@B%*+C5jB2oAt_mcT{u2yV^m; ztluryM^9tjKirpG?23zTeB-fq>=oAKSm8Z|shRSro2j@`y|~mTg*m2(i`Etd6onZQ zBCmqJ?|=&L#NUJ^wTr^?CD^|fiwxrBRl$29@1w-U@HACV!S~&mOF!rY^_A4vnmdDm&$@F}W^$H}^fja4f*mavnzl^UZAF28vC^ zm{Kf~b&OR6p$wHGUeqqYGMCzUY@MnYM6;5fo88W=bbla+brxn*KAj}F(G0lNQL^I& zPF8OGTcSU_XPH*;xOdFs$exwtfjZ?wzuP~!H#0+Ym@a=k^VN)*sWDwok1jrJJZ-|s zu=Y>u&^k_zklgK-CWFgTo5yIeVqg|KLnXZ6Ti#tEWz>5AOga;Mta#^+m_^PTTsC;j zn{d|!gxx$2(|hFILH3_Itch;Q|l-y5+=5K8h(5Nf*66 zs5E*h?2|SM@KlnhGbB-dE`1AjqnEiK)ZCs_r-Wi&Q+q1uCDF)43lvt5>E(YP2 zB56@XNz@we!w^5rnts}4Ky)OsPqA(4ZehOPfM^e-)VGO0n{`P(7s~gw09MvoLnCa!TUed*KK^xKzAYumD^?=K<+&g7`s5^^8K?xDy|B$HIr zNbfEb_Rvc_?lPS!)TYdw+y?KZ9xBlnM)V6yBifmsNIg)ZAVh1Rf^T+jVIkQh$)lgq zsSPX4L>X#G89xjBXI@cINF<6AwnbJ+%99 zJECGz3Ul*oi84G5GuaiOU15q=l4!f&q9C2_c57#hmr$q(b|4{$j8~1t4UlYw=Z#@y z+`NQvHSV8a=B{3mMeE!vh4?23UU3i^9|l|dzZ2$fh~d+)@3Zyr`kSJkS?)4&SrqX& zC~IZ^G1%Ca=>B~<67`8&e{Iz!|3JjfFpEMgjzL0Av>b%SBxqqRCjC$E% zl>4D{V~qx74myGIQLf$rJIS!U*R1Y<0z=x%p7-6*mBGSRRnrnhrBbc2QdXNdTLU*h zQ-_Vi=$`(gv2fHlSbzE>tW)`Giv$04(;Jqe&R?l$y%nM7Jij7o^2D2R(4=GMPDC-y ze)_f-s;Fsm{TbhVn3^>SX^M-2&z-1%k;E%yZ@XFD6hU0 zszQRoe?Dka>ghbSF+1w%SOypxTSt3ncC*)Lepx@(jl&3Y-T04^kFFhW!p4F{K?lj8 z6SodvgPyqlIrg`Ih<)cTzxv(VfBqu*<;#Ajqpw5p+2ICb1;s9>$Py~HC`>OlQ0Ind z1$GPAaQhe8f*)=@+NX7If78wXH6h}U^>0ljpYqsREH&e7KA_moDRLLofBX-ES&R*reSyth?5jKJ52W7|x!_xiAASn|I7K2uVKCDyyG*CM%;A-UF=NEcm1{>VduEzgki9Za=Ec62JeRWi2@Or_D zsk_8iJ<{o`9)nIjk)0YyUx`4DX9{!Jb0aLb7gM+aSpX$<<`JV-H>9-K0;nNl^fdqG zWKYIvUwu7#EJS}EuckrsXqX4Hi(*SDvW<#eNxfI_qboq??4j=xOug-3%c5!o38B^C z+T11hsYkxr_eS_OK{BMK9n*Xo(j1dJ^_Z&1a1C3+UIMU)>x5m9GP2K4r(WT^%XO=3 z7c3SNBGaPnJ0;o?X7({yre5A_UGZm2Did$#CM&chlQWy8X%laZ%n8*ZoekYv`;MS3 zkKM*mwxuh(w&7$<#=q7d=Iv*i%l@~xb_LnYZ8^kaV^U*gOiC#hVs%@o*zTZ{;v1qv z9vwdKz-3b{i{#=Lf|}*1F{gwV+g@Oswt15xHJ?2SSjG4|zim0b}wZrqe_eV8sYp(azj<_byVu?JLWh65eN zrcq=aE{jdMAhP6$X4c@*?DE>rS9k&Nl zNv15%>$I#frsD0D-+b_`&KW2lbbk8XS*#y`Y%=2}V6I@Z+)xA-Q_t?G>Qq|X z3~u(S5U!eV93(c6LA@=^N&TXFBgl6s&X8`mlcW@aI%S|-Q!y!(fKf`FlEN%tK zmKU%aoW=s3k!B!!A;>y(4kw)do~42}mpx@3J1?;C8MYG2rr1psNvC2D3XTMz?q!?= zG@xM5aJk1-PmKDnm-GQYXMaFJYUrFuk^o)K{?1_L`uOLLwY;PjqR761vSo!Ny~+C5)eCZvztAz@-QEr z<~ZDsZglXGb%QZdeD{wNzAv}bFXZj2*os!6UcMaGD64(5HE9wQY}Pgcmo|q+z342> z4Uc?D7L_^k=>dad=-XikO?VgJyH(-YW z|8t-X;slk){8h){WtH>4iY>$SR^r@x_o@jg6M&|)%M0t78dn0HkVcirW`OW#(<~G* zKsgHBu0d+{2A6!GTBk61Ak=?hO08U5;#%T58$W&G))a-?cZYlqc|wON45kO?Ner}7 zd!_%5?TJ4&1aQJWqrlC`^LBf7-H$GQK9-GV9@`iz8_!sg6YKASc-TpIlx%gV!LSZC zTYzo-ZpU5m-jVKM$z8HS`nmu0HNxWPcBWF?6s1k4S8K31@gm8lNe;BYR{r6(JWad3 z`&NF#NT)fO3I4uJU(4L%?PhXP%f9)Y2C~eVDSqkSui0qKoKi@g{a( zK)my9dXaOcXX=y;5uR&~Sqqsvr^tRNjQ#XbVk0_+W{Z&q(jn(Kp@YAHr0Fxi)@q}S zS5;uEG$RyrvGtIrfUX<5ZQAQE4~AX!ER|+y;w0K+?~1U30CeOU1y`6JC_$RS;693B zg@AF*)G31@SUu>Jd@w}YsN5geCSDkrO&^_D_(HxccPdJmYcXhnwTTwKa9-Fb*yV*h zll`PRaJBCtD5QQwb;!l&D`KGq_&wJIt#WK~i>?tY0D$(pWysbDyXl9X-J-Koj7{s6 zy-ay{i=s%?;E^$TgNybW=yaFKZ!q|kCrJy_?*`RU;R60XaEEk2h3jI@DV`BV2yb1O zU%&6Gw4AVb8(Xn5lVX%Uom%sXAza`3%HR*R$TQjD)B6;Sq{+hBYAYRL5TDQ~CtEpg zg4=)n+;YO=Ehwzaq&nXW7Zk3=1({a6+piW%;+^&~NZB41^eK!QTD78Fx)u`L-@Z-1 zjr=4BPpY<}V`wUU^!F2n)()4rp+k6WIf*d2Ge6#4kU>hh#oYKi6_a{1uj~NDg06oB z6`S#zF~z+(2D8R+4>>0+FAVE>pJ#iS)v%w_LuWsq@7pG>rt2UZ`;Mv?>Wf#3>%e0> zM0!L;s^kdF)~<)^QP_bq5+Nq_|i=HVA z9;Jm|Q{DvyOQ^O6=bB?um?ApIHycLwbEq)W&a4nNLB786S{r>$Uhjpid*~yoO6h=8 zjZf9&0jCwdH9iAQb0v@@;9m(!{$n5i;R9tyEDY~<%*UMU5B0(4wNCd%cJ-ygF z-g!)Wv6A&ZR5#YBcc8_UVGZ2}kT|BWueq^dIBZ3Q4PDu_$?xZ$i~sVhcOH+um8^Fh z>@}g=WiaaP%uZ6L8gPma#Xq-_cCu4-iym+)o>(tV_lE|B5eAuV)Y(p_b2t>n8r?|P z3Wdks{?UD%&39R!oCaR|czl>?B9gky*i5!@n}~P}pHpTNQ8mR@P-HLG+k^^MP+|zF zOU)jJT5M3?3P}w2NU5})=?X&?c%!^GWHl~x&zI$d8Y`?Jj5bV6cyZ7PJIJQH=}YpB z0a_$NK|l)XnnUmkV?rImCLqtG(`QJJC*EtHA`hbm*JEfiJ7F`3tTnu=gL23w2px<4 z$11XYBx)s(J(fB%pzNpEDvFc?wetBsh0YUpG}Y<;32c|43_6DM3goB3DOnttHC<0P zdv!+)I#mj;DYv>pewpzkXj}_>nakpR0o96X`lPyAagK$qpr$|$)IY;r)mmaql0~8~ zP&82u3$0o`l&ry0Yh#|m90Zjy@0?ny)H0xQ_sNBZ;zA6aF82Uji1@ zmF1r&zDLz1MQs#Ry?~08AV?CnBBG`=Nq0J(-nU85f4ZkTJu)+Guj!uZ^mI>mBZ}gN zpnwX>qAaq@;)1fOh|!2(P*Fq$DbW~=f+C{A|D38KQYEGJiYmT}GoPRG>b`q#;k&s)G_Js}$G? zMGJIyKIw%CCRc(7h^9ceo!v642hv#+Ro#k{WQW^Dx;?x_2JRDZfasm#i9T3WYpGCI z!zhNoc^WcHT@D`((ET=f%He<{{;la)%2AW1N{3 zRCL2V3PMIq-bX859HZUKb{~vi!EXqQVoiMtTtcrLST8c;fhh1VgroEW$aUSW*5xT- z(j&$Ko2;1ebF#}P3qfI*?|#~({%)%G!wh2R7oW^6BBfKwNu&3-o?>e#a)63F&YtJx zN%a4=LNVsuzIff_oF^G}f{eh=fb=g<0y%+gkg{{n{}2yTchHEo+3y-@RWJKd^I~l2 z64btwIP<7~XK)>u;F=_8Y(w+y@@WI`+L!yI8x!;j-v&7nU!S6FVg0x8;d6cQp7vZV zy2oiNb6op;ar-NUd;^B&Z28_%vciFF##u(-PNvw66p5!I56-(rnw-O9_%>oQ@r%NA1PUMcAYH$;MOYiN=JNc04lmUP>#08*27lRAOzHYyH?vP2Hr3Q#YY zN0e$sM~29No>i%r|AVX>?-1^|Vv!UWiJEN`n?jLIR3sktx|t4_0?Ag7e%^|Zenp%5 zvgA0q{_3*8p3pQ^`NA4-SN4jkT`*17b_jU-l5xhz)2Fl9UGw<3;q|xM?-=kgrT^bQ zBMr~Z3F?Xwa!yn1DGIc*NUT(kb*c*db*XH6Pr-=Oiojw3%O`WTL#R7Cmde93=T|sm z5O6kK3016yT=BS+G^<&u$>3McJH%TvyFz@{Czq~L^^woXVSc^$P{6J^5IkHtuUmn< zLHZlmIO3EArzxBx#%nv90f)C1_cgOi+%@Q8*UzAzEspt|V~PuB#91Pj%=O`b$I!1v zB$}!-T+T@jd@?tq{A`6-=PRCf!r<6jGQUZsPoEfZ!T_#5RjS8iS)4}KhyyT2hfK=P z&zSXfKQ#!PpB~uyHaYLWdM;F$9g_$BDaH0sq#N?olVEW3?S=$hUE}xqv#JFa;+5K^3-#)pJh^& zcV3NSMtr*j`dM@6IB^D3$Q*F*l|cl~tmU$rApG1NS~NQiYMDoc_eB87)QmyWDX9?Q zOf^hKNWTAef1Evzxf0w6o`#m7V*2#F+x)69B%pfSep0;L7h9jQEpmX5IcYJ0cJoLy zzmtt)xTGHbf}j0muT?U{+b!v%H;^p>cf69nk^JL&uSZ@fu5D^8a@0w-sQ!_|8v^C| z6Sl^!OVBUbD~{3KUhi|?>k7TYEf<0wp0dv7pR}cSCV~`8zr&h-+>r8b(ObVV%OI8p zej-~zwofGxAUEdiFQeEJiWE|jN1z@_vg`hc#JMZN#$KZ_2S2ZIYf|C8R+;`&;#`cE z!sxlps^g*-H!eo3}IH_G#VBkUJaEI8eAsK{!UA-Bswe3+&@CECG5qYjW)p+=3qg>?D>-6JSd zXz*SQ_$1=_n2D=7;vWb0pUu)6!R4VnG^AA2LsgO!z`#4xyPH;%=kH;F zOZzKlKOl+E4P45M;F3kL5TBI>g(wuO)mXV4|1<;?f)C+}d`K1rgwXmwP=00|dzH6T zI7srv5MlwRtr4e;@Qo3*o=Gp-K^_{do7FHen#J}Xe)C%HAyeK5hs{th!{mTYBVDHc z3iS##_$jgC9u~q#Q=<0>J@rlh#|Q$@;j|IHFl+9S_e@J-IV>{F6uClIVvoEts7+lU zyziG9GVHWDVgtdSoubr`biZYRo0wCgV{$Ff3si`@m38!4vfe8-1ggRDK6TwR`>21b zT0a6@*YcbYY(XM|ZUv6g_i9tO!ih?Mua3srar;%bq0;o87psAo;LN;o@oDu0q>I&p z563MVj8$~Qc2dKz}4*7x@;AzQ|?3_GwLX0OqnlSi>YPqaf%BGb>S@V+bW zRUmbDk^&2sXppA6MRv*eT4*mFOD)0O2*dGm>CqTaO&QzavfF$0+#3)Jeu&pA)2C5= z_~8VQGX7fH3P0n|ZhfnH{QUHfOZue-tcY%leoNvV*l1I1go_M{1-^$ADiV@XfSp?l z^e`A522}326Jhl{q~L}vs$4!OT{h>OC`mCq`vNoUo8VzL1=VE`9QOER9uDJ} zTBZap`R#X1m0O41o6M-ICA zI4lgh(wlnMx?7h#V(HRZ2NV59W9fHT(~lcWw#=9{z%xh^CqJiX663&hC$@QsZrBU&a`xHAsk$x%?*#&`Wl4#J|(&xEGhSfdjE*i6;vsG>EA>N47 zH9C{d=EXx6877d$1T^?J%a{3S&V;U+kMaF!N~|NYo0Ni?;tGAzOM`DdZVR<9t#d1x z(anrFA?X8lfSYiTfm?~ZwJw7sDy&8n1@%3m!nCApj$w2(TMApt7WGH`Lg^|&lyE)7 zYHQN{()~^-uXt)=ye>-agl+R2aRTa09@H*YdJ&$(7q zEQ0vuR^?&-cAs)@?DKy-5%$8Ue%`Ooo4rG5z}^StFaDXVol5eJUj3~Uy9Hw;A|EgX z!ZvlnTNfAVum1Ak-wYRj{<|K5i9SeMn8VvHUB`9?@3$}Cw*L3$AOXglJM8;Radu7; zz;Q!RepuV`U#4t}U#bACJw{mmf=#|B^iKGCUjFRSpGk^V_uZse^tt4y97AYw-)fch zyaB~(pbz*O-?~Qds*zsYrtW#;iw`zG9X-aSb+{gd_~UtOePtfkZs$(EWlEa&Bvc$Y z2HXra-cgBG(MYhul_rCkR3V5J_Jd}@IB9DUa2J4LdD??U7{Wc|(xu26u=gxWPx+NghH21?Ut7ZvwMn%UL zKkf3H7*Bd-=dBP^Yrh$z`bvI{=wJXoye!rAS2c0qG0Ak@KdTpCOLV#I@v=pnzwxj4 zv`?HK>C+~%@jdfz<2Z&3$2D-izwU1{+ze{zFFMF#lEBSf=(rjcfk28ekrkN~n?{jT zDzXW{T79BUdM- z_!=jukBSc)!+N~wQeG-~>kR`&9`Z7OLTa8Hj9f6nNE5{#r^rz%vR~dBbe=g5X#x9r z$*zN>$^VppAK)cTc~ShiSCsHD#6k_cqMcDD&!K_(0}3`ipp#vpSUmzw#GP%TCd8fVBki@<5!gSM-r^uYbS%q*s}IGr{?G z@G$tQLE0=Uaz8frw9HEV*D64b0?VSGh!d8sDfU>(>{mT>?+S)e1bVwK+JuJX`EBKH@=KO&``H)W*@`vt4#YM`XfjHQqpt!a^Ufv0lFt(NRp6@dd&Qw_WmKpO$v z6WN}IFWwC^U!99=5;tDbg=Hb{K8!^o+62-oI!=m0hvgd7xmV4K5_Ty+1x51Av|TT) zgYsWSkysN1oJG>f@=lM@`R%r4v;TMH6hsMYUs^6pbIVfgMR#Kf5K za#ej=-AHHp_AByz^*tKJIiMY3;p#jab1sg%Zk#~!g%}ujDw{U`w^uS_8pFT^Bo4e? znTdF4mmZv%&xF(e{Nd&=O-aujHip3rjTy$*tqVyFNt{;`wA+8T{{@J)*5H8%-J`WLJA&4@ zw4j!P=Ncq8ixcLX?=8=xH0HdON9=-} z)XS1fiahr+CQVi7tA&%xu+vexQ8GEeuoMkrHjv2x1IO@)e{a4TXnOc{Slz@-ihKi8 zOe1MWw<2b0^ou3)TLden4m-t6U97)MpNwFcF-yw<7-L4+|I)me7{`A_wQkBg8Zs(6 z{#&z+lsq?PxD!T#rj}x>C{js9V*1ckU{Xa=jp`Y7&I1szk5&@RWnb-#eqQEGZ0?tr zkyzM~vOosKG4CLD-%^f>;d(P6T-1L(^k>)U0VCwJRaBfdQ8F`1A_kY*X}B^(}6diBSx6m zN3pvpQbtAUHH{_Ns(vU57AMXY?r=LQY2YIZ<&x2W(VH@~*<$=>FWnKOj{~g=yTEja zx&+a}4Dk|7QN@v{uof~!GjqKkchZ!Hqvzdbc7B_K;qiJ-=!%o@nNl-tref0N|B}?>Xe4F9JqDABaB8w)p zQKubSzr9jnfW@YSfO9078!Q|+>Itgjj0s)Hp;$-~$Ux=yGld+rPD4%oPCEbNb2ROBi}0>ra5`6uz(*`>k^WtU*& z+ntc;-V=P5xk)E5l_G3yQ{VW;B}pE$j!p4Pr_VwG;|%_O)mezh#tMws!;VnkzWifl z!bBGM>0>$VkJS7#yN_DP43?y*p=0^`0{azU2hRtVn=(%=#O?G+>ZCQiexBi0hiHEU z)@sd$HxxS{K?)ina-Eyq(;;RDoNOnQ8O}K&2LsZSn|a54s`)lCKNvlfiLK3;buyZD z;vYDzjpI7YM*g<<&!*>Mhqcz4Iq9NFrNSFZU7=&gP*4xE;=7>S#~L>~%8l_*WI0~P z8*RTfPI#Si_^l6ArUX7*^b*GfJTXJ}ldXC{_sdHHu}F=k#T`nCFpV@anS#;PWA)q? zRu_8GFaIQ9#3>!hrnR{D^RDu)zuGJ>hlK3U{F`M4D792{MR3t_*p*YguIVVfiCD;~`17w7{a$gwBQs5sg-2V!M$EGl)o|W@3rW>PP zFb(Y^pykE_#deK8m>5#50)a(?e)RS2sixlL=L9Kce9=hQkQ#E*D@_^awbM^{q>m zVDaB5A!xW**$@i}W0Y){pxHSM5`;Fscn~~>IlSu-C3^(IB$E_(Wj4zfvkD7ORZh*V zF1L1W=c4+Kg)ydFln#4bHp4sFB|sC~HqWin%OOZl)u!I+-#N3A>Xr|PFlj7>S?;@z zJx2z?)YIyoE5$+$4*(1Y{8N3F3I~`5rUCF}Px&z+tZ-X?PQ5u{g^F1G=Fm)olv)14 zw6!Fin{U~HRTU^iKIYg{Lb0H`u5HAN@pnvZXXr)8EvV}uDU__&LdCC*WAv zIYC>wd?MZ;2S$t;Z%{ukD)7PU=M?L~SALSdJaYptE(i|!IN zF!q!$j>8<(3#06~&crKhOHcDSW2%KZv{#!%kliwfboRec4{>ctDMC~m>o4l8Km zxB(GF=Dicg{CRZ3C)fXp2XAqMO_5{qw?4b$!JqrL|JJbE?faRVjNEYG-ENf;1MmZi z1xa+5ij3jofl3p@kAaB&UK-C`%lvjiC3K|w0h(WIEadgkd%p!mfy1F;wX59Cq4CACqN@_21tYv7Es+WD3^g3x`~_wAK%)XUfA5@ea-aKL&)ehnGV|_rsFs zg9CWVADZ7i?_@w<;r(C!l|)U2IzeMkJ@FK~jv}k6$QxdFUVZ3#jIXJh6XUhRzfpX~ zV=&m3Z|Jk%6@3qj-egNBvtd6@>WSkz`+ms(jq4HvB+9*;O!B3)+jn8i#0c8`t;f_M&LwfrA+kM z;=J2;7yZc$Ej^l%z1v@trP4}Ul{l$pyU$54WD-hJT=vZKEr&vc*bl8%@Y`Tne!m~29)xOi9`Y>u0%*l(z|N)7gl%tGYLVO=PxG~*j2v$& zZgR8djbj%~LBd7TcHod5GmzTj1!N7X-2`I4S{Za`R0XU6t{8np<1q*u#aQDu)o@q~ za~kAME|feyGp$@sAzYQqx32Fn5ik!LFrxjHun`%?JY;3kU$be$S;5U zd(D6S>^JZK@mG=+6uXQfQRB&uv28Y6UUL(1k|b1+%&TICLD5mITUAT8adTWaup~KP zq{0_bEZC)UfvZ)qj%}CrD$YsudAZ0lSmVE)#U$*N!W$qmFlq9Jd%wI5b|y`<@Dtuu z$vQO@Xmc5H>V&GYp8(Yz%AE~L_FwUg z6ZgWrS9!gnn1H(yP?F-vwa_$Wry}0D9U=uS-O(lvpfOmo?9w7+nHNvOABu8KQ|lad zzcQ07hl$v|GGvuXQgkY`@{@vHj0RDK3=o)IkpW6eZ2&fBIA0^V^ldQ5AYZQr;nyqb z4+oQte3yS+&;g}ZS`T*XCVx%l%r3#lBHZQj#m79(*)&|##xSu60+w_#50i1|MyCAX zufLi4ia{z>&iV67QsBVH^^c4s&;g2tWUAd%B<2YZ5|kb>UXQ#!b&VEA2{pYg886~r z*og2zed5J_uqb4%!M1VbNOw>ah z#jc@9EES2&VJ!lP(Z!_25_!`rlM$c$ARYpCH!%kzs)LGLmIap6=mIQ%d87MM;d-AI z`Y*^6$bJ^ zK@5MK_c8?gTz~nF>mx(j3fJC(d!`$pR@9{s;+mqTxU@{RL9g$O2yW5S4% z$G>}AP#q+zUiQQS9Cu#5{<6b5z5a9E+^up04AZCo`gbI0Dk(P-Z`l+JIo|1z-4Jx( zJ2lFC!HA6Tb7Dh^rLgcYiA~;ABZ?*ltlj62XtRV^^6OoTYxOb(D4ltUESj)-!VE(#4<5q5@TxhDE(b()@z#Hm%cJLZe)`Cv|>qXMz|C;pqJ z?2C9P8I{GH!Dmi=vt7FmwVc2=mYzKx#nb8iHMX^v=Aq_&w z<~2!@L(0T>&qGbE&*2^|#) zc3Tk3VPYyw?Di|^CJy}3Uxxm7Wq3guQb8VbhmN0D=-&t@9Ef(OyXWubRl?_CrvcY; zHU7ND`{?yT|4EZ2T;X9~P&}SJh5BO^&mcvd{G6gmj00;nz$ZM$9g;+`8z{0KHJc7m zqK_sOXm62#Ywz^cjHX$%F{no#@2tVt%t>{Ra9Zid3t{&wZc2yi?pr?z8a)vu_3a`> z4vY&Ze>?^k)fBs*B71=oNqTq&(hoqfP}B_3B#ApmHjP1iX#LP2GZOA9-3nyH*FZWj zuuns8M$@f82Sk&kIY{pr$MED9Ss_#myQo5zbB#W+p(Ci7*8?SxaS~)Tw1GU~gdfwr z#O&vZ!;2e!M*j29Z@gu|Pth#qHmRFRE*dRUGsQv?i(^#ewizd8oEEH792F!fPC#l$ z7M-Ga;0lKVWPr>FNm5jTKe>{IrmZ0f;-nXsU^^aCJ{BDbZIkQS%#Jwc(xViv5WsSt z*XO#{{bSLI8S6YCDtnI?T9TKM?Q|T!hlWxJZE(uL@bfq^GNVNc_ju)nmNA21Le^X$ zvGbvhDE&$3>bcwmhkaIV{3YeGo}5$4g2|%nP|NaU;>gl{f>|lUXyb+$Tw`n zn7>4y@^Cb$*cmY#VKPRJO&khOo54IxlK-&0MPQI1i_UI3L{^U{K^)gxOLiJbkQ9pD zM3F=)vN@o^H`DcV*V!JAHD1=mn6AN9o?&FU^01KBx<^y0Wo$_xPfEr zNc&!!2S-p?2i6RLUn%&~X!%mA6 z6c?uJt|ti$+<<^W!oC0)o!vYT#tH4D>i@aGykHFC)gdf z#M2w>fBlPTWS+ySBW6Mdfmj}b5-=Fy>N_o8z4g|nx34T(8nXJO5z$#uH@jkLn|f(T zkNl!^pJc6Sw=e8CynGc(W<%&(kxh+e!(0r7+`J0K)DO}U+X907bsH)Y6MULTRb zKO?G>M+K$=fxxh0Egxf^;G`7V&F+HeVZ4u3w);pe5EQqm>jdZ7ekg;6$0EJ;=b~`! z9BxvEBjAnY2DeX7!}k;SxB)&^mG}KGJIM`Zx~*|*P$6;bak36; z8~lwyI{R$5;-YV#=Nu1~%gA8MmC%UM|M`J22Ytx2ZK(=zEaq zf}Qfk*IylR0v;SvsTh_c@Q!%Lf#n!^)phxLb<#Y=_Ia4Gr^$5fSKOCp(-<1tA21w% zePC;Zcn!>+K^h;24YA{dzA3Mjl>gk{Z~~u`{*MHbGnE`RI@Io=ScuTu1td4_7(;RZ zcDCzu*)(+Q`8Ju?#hw0{x-ixggy3no%FimiK@pH z&lVY4elt9g^$hQwnYPP2O*!DgC06)}H*UNJ+ewG{CJKz#&x^urlNzsSAKaF4U2;R5#ppQ7>^&mN1jID&k-O#x=A{ z(fF69w25=+DL9${fS)q!#BXr}u=L?~rkPrTpOauIGd5y$aPNoX*nQ<*MVfNy`~yl& zRG@Z7J%cwed{4LIeJnysl8X@+RSf}c0h)e!A@38>p!|O;sE$ad7MQYc4o zlpvF^{!cE7C5U)(l>J-IInA84z<`L>_m*dq%JCdX4s6gnYqahkQ7kY>9HJtZ`Q7xq z8JrQ)Ebj;^7ir4;dc4;A9Oi4HzgNF-$hC%sw#@`@ZxU~Z$eCDp6LLb3o&=Thy|NLf z_1{ehuc0fv`xQxwC}D~e%6#5jgdOxvrHUg2ny@`xjN_pSZ?O+vLJ&?aaOh!x?;bCt zps;dBvIH^Q&frh>wgzA31;$VJ{PK4T4OseC!;f~6DsJ(+4!i}OH$vIR6bob%N2thC zzD4ep)TNj2NQ>MLOV+Wy^vbFIiVIA?q8O-Uv~tvqn}a%nV!<$Un50No@O$aKOt(Cb zhj&&^P4&T^N9h&-jCNW)AK9|;Q|wfr@+VdE+SJt$^M`%A1ZAEXaMnibA^~6*eVhQ{ zziNPB;Lk~tqOw}7*&ArqfDFg%n||`&r;oy^;NN~pwvH#792hhZvo|J)uaIJaBp`>1 z%n(+(wJ2MZJ*t*}D*47Ph}>=Yr-ZNJ&kfZ1H!m;TC;<}et9aI=@^`bv(Q6>tCXU%Wzee=?1%3rUz`5NHzyOP7}$qomO z?}S1QW1vw^v7qA?QIVNI39&L{C2$!HN;kOaa>et7NQjlBKoXh??}TtoCxm69xe)Dd zJ;5z9ViAyV`~SZJ?kjayf6d1$PQrUe+W4ioHF6bcr(QkPN}Tf$|c2r=4?tcAlk ziV1;?gud^Q$)$2HCSIA>bTIo=Tv^ z+?ZE9hhibtKZA-qri)I9V`CMauh;YIK}pyjdY{dZbc%XKZR*PsNI;zbX;_@NWPT@P zwATn~rKjlAvO2nfUnbAwRnMI0BYehx%laq$jEN_Bo6}qExH`eTs0D+x0Vo6C_c>0Y zxv9twd=Sep!utk_T~Cp*nRzqu=dFOyOt5;VLY{bcsM~~YIP-f+XoFGlAhmtND^9XXMF9*w=Jv1aM z0bq;=8!tXi$4M=k^2a~?<{J~?UvyxZVupWFAJEdk^ag7V12+dox1@_RLNq(wE-)D( z>mVcr-O=cywr@4wu}jYxOJfJy@p{vL{e0E04Z7*|dFyAB>)dox2UbH?8WG6cr`Q3C z^iz=-<(s6a4bWBSX=M-$NaB|ssM+eC6nFsw8X%YoqJql_@W#}D5vGF1_?&CNX^sL( z7xNS?yamM3n#-QsJTHf?2m`8@h+k>oNBRDOhR1k^`!*j@=bCDXF;9XWN62e{jFfex5V_SNPG_YLCJY^C)+~T%l4JrTngO?04$@e$a zi^=AxWRKB7b0@`SQ$V{DS*J>!kD1d2!i!8D#Mr^xhPKb;vNX35r)K#wzb?8Yu;k@j zdb@hW319VAWg_VmRlD4hCc@UB?`O|NvV291)dt-sJFLX$l$=(?afSR^b@GR-0V=K! z2F{aB&&~F=+X(Tw6q`knOz?5@YXkN{F#7ogbpiLKeXe@WvUD24!i9J)LQ1iI#f`xA zUPnnsP?T_`s+5Oux!F)Yu3K?AwBH77#-kt_@8<*nKi+gUYy0Cnvybo%NSU+cdq>HN z=LRVNj4?*%WQyHLk$5Vyz`szsN|6fI`A7UXzDASaR^yY))bgtWG9mB*|JYOTJ+W|H zfbqm&J2s@h{%)gb!BU4kPMfKuin?Ggl$!j+b6B1bF-WTB1NMi*5|d5qjy8BW8gW8rbgQxzDl6k490Em`aH2B5cv3wh z$|aV{QYC~p16XsRFaktmWmzA@qC=1+kQ+z0vZQH32(3yu|L&^>5PeXst|x`uK;*!7 zmt#gCs-oCR3KVoGx&vuHa1zQuo|iWxg6)JuyWoG(Cuxuec1y( z6T*#Q@YbQizni8QIqbG%Cie-;B1D0)qDoaqpA(&BVrHM`#X48cs|4G~8n+makLIpZ zUkQdV7pO^m#(m}7YL`?+->j=_<-8r!KZV)}J@V~ReZ^u-!^WI6%!We2)EANB(L`*# z#n}!DwgAB6IXG=zQ;Lgtw@vN64htMJ7F@kH12wCgo%?y17}*w{D{Pmpnx@h6P6`Ie zE$1^KN9Vv{e-cBxxufIp3fl9E8#>+XagK29?x&sHyRU7RyB@|mokvu9A z!mUZ17^|$KSs$Irtx&soF9aT;#a5uX3l)rkSa1Fo=bbc?9l{nAnhuq-p#&USadYQD z@fPV`8RWrU3d*HX8EKkkkHZjm#Eh$y*c?2Z5c9&F-!}jF%WkDjr`NLa{8d2HerMJr z$gw|1^ewmPXi_M(@iy|z^|BZj&z#A&9XR2Fx?6bhy;ls9WB(%(Aq?rAkTaI)z3Ag$oA#5Opy7(L?( zlW`ku0wZqdDIVJPFB3^hb6k_OxumoLArz)fc59{&lcH&*43xm8Yy$|6t7i#*I8JGk z?!iA>vj0tkJc-%(-SxAU?CBcj7dk2y)PA5-KbeUVx`3c-OFgoNt)s~X&wYG{;5 zS@br~eDN&-SYIv#Y?|FE0tfO2w}%p4R3(;=%cb|LhMjWhhwhpTm!%=sm^7btA)0GU zRA9cOmf)BXr&^NZ3^@VaVb!7e;)kw1IvEEu?gLW|wnJz{C%FNq>1KE-`tQAZlE>V^yQ&V4MDjcAlvp(P2+QdyFvhg@Y5KR3LjbTb1W~#Oofp`&vRU zGCl+U3sx}3(rZw)Qqknz#XCiJGn!J*0%7$_*`U*&WweS$a!rOMvhAVUez6J@Tl_Dk zvH1n3?dXNjUmv(`3YF(1Bg71_Qhoa9p3vm^7`AW)vJAVKWA9~BRd1d6%A@Z`LAC5+ zj~LRXUjNr%1a{vxujMkk?q3rDmd(*LX2ShTfBq}ek^)?y z;=ucqnF0hmsRG}JaH7^Ueyd;bVGREOlnzLT>I0B@c0aUBd2z0$RB=poKvfi)=u_Zd z=T|J!QCIcJQl$7n*YzH+h|Q*3-s}Q_wHNAQeTaAZ!}B3`x=_05>jnOOvP?)u(sNdQ zSnggj;i&M5;OFTZ(-<&Vad}s;=zVvCp7VqHZD+|wZhDU6=DCnkqdjZ~#X=^)HY#$r zWW(##u9@N--ciBW>sGZULkVJohzcfyiUlzPg7!lyNVcCJ*!b-#maLT zvU0=LI@5Yb4y(K58?F0Rirs=BjZEYvDH=i5+%3OF-_#M9X^xZ3ndlQjMHB7#$P8EP zQRNKPEbP6~hNsVc!MQA(^?kYND@RLT`%17uh-CU?C6i(YHsiu+XH0xb4aFXy$UdmV z&07X9d91K7ywm+0oQKOr7d;m*hSwg5HQHrR?n;G+BWz7l)XyrHB&vG-(PO-LajSaq zmVmYVJAyt^!!!#PpH_n(+P_&=7(N7PtJ$hj7?Hu$NiRvT9u9JcHw#c_xkt)H6NVXg z06l#-8NF*gbo3)R0f_qiWj7gP0MPF~8aPUpIWT~rZr&IGZKT+Eimam|M?yOld;PP3 z$1)m9;$-udk^XSVp7q}?p70`bUJMIS!Fk?^4&a1}7iRxIwZK%YaM5iYIA+0&5YmUS zBth^bguI|#MN8)kZ#koBc9BL1%#WHM7Zl^QXQOb)@k}#~ z$)n<$1xP@>Vl`WS3oG98} z2$z5Q6Vt3#hgDn5UM%{_ z2~;n9>hAL^PXnl0-`}%@Y;j;v?KJ{b9>qcpwH;8|Q5QjZa2AryodNDJNPxwfen;ReC;q>u1p(&FB)=u~LsP%TA#bxxWRhk(l}TU9*+ zd932ZapYQPc_@PXs^TCYeiQgTK71OYCkDy!q!fp?r+_uS?r)K%2_|MTGm=7Lyn0>x z!xQ1?vd^_mt~oIyMH=HZNa~qx1(NPR?f?IjU%|>-9G<9^-(cM^PA|FRn$>^5*5!cG zAbr%we^x~@9eAgLbbv8O?NW*@rbq!5c@_w)qNyti@CYgTc_*burJwqrRQ1x!!kR>F zqK5F(kRPyuf1HLACDLJe22<+UC1@7xkBAmt^d0eSanAR_Fq%dhb)Gb3dr(;*j2k)H4UkPCCQ$LO`3kNRsK> zC0GUYaZ3SD&??MeuXqm8=NEL#OZ{?SOqJl0q>pU%M>_9%raJUvrZaepYSY)Y!wj)7 zL-ovDmrIhySC%p77i>|je^cLXD=gUrKVy;gR8&}glN)}nd@$`jXM@aXe68;9WVr*& zob5(3Cy`r`a<%iZ$TULRah zRL{(q*&sbB{)4{dctrZFKEVggKi;nQ;0FXbmA`vQ5*LK$;-$i7`GtT!F!;o{pXWW1%WA8IePX4pXc!9)P7CX} zfok8oagj$1h%0l-I6+o&Q-d9N_X1k1G4`8H6q`tq1S&FG_z4#1MBl-B(1AK6ZM>ux zHA|;=E87*=u+GD-ZiKNI(p_@`_!Fb;*YKl198!B5knxB;zKW!cSAN8S{Q*@*>t9T< z1r*syMGnzf3+h#R>zfv$6^cTUKvmCuh_z(2ZgGqTt4%Bw=8NO`8VumgQsMTe%MvLM zDo|#sMxfkm9o+@V_nUe4QFu7P!kjF#13Jvhz4I@u{Lw^2iQ@`MVn&pF41sjW>yqFR z<24Z82qxd$>3w1PPsq}-<^Q}FIt2dwg}g;?8$`)(Pp5pJ92!rQIPm&+8i|rmD7Kj* zC#c9?zvO9a`OAb`$m6Rd#V2$!FAtc!Dx|e*YJdzy8@T7ynU0vrdF^U)erCPZL z4WGExm+Px_=#8I~`m`;6)bKXh=b?^2tNBMgPC+Th)dznl*yUqN_hN=sCDV1av|FCc zd&GZy)v7)XnK>JsV*|S3xVj;HFOw<0Eg1?-=4D87=%r}ifDnCT)F|Oyf3-r6RqGn* z(ed^H5*NgnieSU%@t1spt(*|!xZD0OTTgbl7}S=(XwLkR#7>3!g<~bjW{ORs$Od%K zr71TCH9$_fE>O0W!FaX~s9Jf3-sPhCSpMm&n3QZE6pXuA+@LY;6n1Ik1P#aCyD!zQ zQUBfmjo4Lj^GL4)+jiF&vHw4!*ddDCqasUq9ipN)`ax6Gf*Vjr=#RDW`XK6{N_0$G zEUfoVg%S}^En<*V&cl9{5DAwL_S`#R$6xOY7;#GR!`L`<1zh%ovaG`L@TjnE<#s5` z3VdYU;G51DL&WY@*Q5d(e z#UedDqW<3k;da(xD*$kNF_v^+`k`r1%yZ&&G800hg)D0@CoGz?Um%-I3n!VWPoxc9$wwzpw*qqzt7mF3p!R}}on^$S zdS<)yMwljDTq`OTZJU{;y5XJ=nT!xUi?l3_0lVgCc1zlTPHM!dThS|9WvB8)*5@vd zdA1XNsN+n)>8*BL-MQz-*S{HR*zQEm3wMxR4!qqp8L4P>6boAZespE#`__BsLFF{e zvpuGQvNm9{E8*n@eB!|f&AI^KKH%41O<{>&k@+93TJlA zY?a}^;ZEEK0XW~q=_`g)v8Z(*wxB&q>mGdFq&Jf5GIK;af+$`7B&b$lkM3@v=4u7JB zL#~i1*&*5=VG~3=bE&L=%rob-MF;bcaeb%g?$-3HWZv2p4M)IhYVxcf~ z6Z#?h1dz-|YwF}Rq9k6mG6^zlYXb&dy99|ogRWrk6I4wHMK>Lkl6-NOAR#<~q)GLD ztA3Kfta2YvoD%KeA>C^o-3)Uv3=qxd(gnhp0Hnuy097V&G|t=%p#k{ov<4;5=Y&)%05@jm<#x#hh@VGlx0_0x z=R{97Baig)eRzV>g65m1jH2*_)gQj!uM^U;F|kgU+l?7Uc{ZZqeD;EHhR3s~v`LTj z<=^^znnCJl7X)7-sSd2yKxKh3M(hHL-ARFCGP1#k1j!)PuZMo8tYEvgdl@Ba6O+1LTOMZt&^e%K$PGKDqm=e6VJHo!g>9lkO<>W zJK;w2`_Xu&ZN)!c`_$AP=&<3yW(<`X%&8f<-YD^#<+s6D4eT(=UeOSJ$E!wE4?76d zY;@ry`6s#tT%a-0@8U9^6Evu9;mLPQV;{a$+O3%g2b2qVlCPbsK{Fp7>u&@$%j3!N z5Y3q&Etp@@11A$MT+lL^PBQaoaf|!)H)k2Pn1P?jR*>yeNu7}-FQZt<*eIkT%RED9Fqeu!BxhJAJXqm8*x~SNtI_uv(ziRrPh%#OxZ@mwCEZfwrQmwR4 zwgaPm-7md-S5h0WnVpn3)*6UBYk6&PpA$qJ*E#w6qhFMJ7$DOA%GnP{q64e0Wk%S| zqFA6YOQRxlLiZ~$4Xr`kH>;P<0QI984xcez7bTj@lG};~app{HOmiD@D`k#2Yqol< zdl?&YnSodB?oG7@zk0!D3zV9lgA;yVSOKw63Ij|wB?O!!$qo#Y3L}{0P%OmlWKfZ( z)wKbQWQDqTMuzJp1y%qTmXgo+Bc89YZ{~Ju!kEn@tC`TdJbJxS~}Jl`9oUf4u)fVE?7Nk{I|1d z(%QK?7vN1lEMuQ89t_v&N&zFMJoX!L(%VY96kp1yPo2zaOSjf^h{6Yw6-$!d=2@mvr$ke?3&SUnMKz^@eg&!K_D>v36zA zC-c(gi94Lc&yg8gy}P zxwoOx8n?xK;wf6!U>PQ?n|mUoxMAY=OQ$k#7%=gWm-!P?!_Cz1xH=ioO^roR6U9O> z!%-?SHlT3!)_KuBm9CGk9>21L2xG)kefBaL>~{4deokm! zXqDem)o9%K;WsbAeGC`>>xX|E4gGFYUp2&sw=BFoZgBYF?|#^_h+Ay90UMUV&mw{I zluq1%IIuSBZ>Bm97bCv|1Imm>)DUokD$rko!RtK*nq@o05F zCn}TQ;T=^TX9uAy$+nrjq5;yz-k0U_D(6+ZR0%LK|HOh~(K@!7r|)~jc`{D?R;{|_ zi)VFDPP-GeH1=P=YPvf)EJn=iH&xU7c`K(5e|^~xPQ4X76$s(MhTBAzzW6z4W(OT3WJmi`o&K6>O9#h6H73R(v0g@{0n=LFB<@j_e z&M&B*p&z|eI7sxPH~-UVbt8?KY*S?}{-vJO++ z02SGu8T(~tXN~sL&gceXbb-I-qHiXFz{+b(wh99)+oeOUSTGIa`>JQG6y5;VUOb^g zfi<9#iRWhnSEi;apu)RSJmhLerFo)Qu?`8(9>obOFSzV_GgD+xjaQ{!{tvQ_n`-R1 z5rn{;I94^@MzJXr*+fNROg0oSme$U!fs{8qChN(FAcC}(v??LEx7#z?FG+F0zm#|V zWxE^w3=3)nnw|{>8}yL1{AQoTfDidU-fAW5zhsHDz$!E5h@DQcTPd=Iip>AvB}w5z zt-yQI@Gbacuvqcjzv`!}-0_%9x#Qt?u3KlcCU96?$odRetsq z=B3KFy*{{Ms=k=f5+EuFBE&FJX){yl-!APCCHiP|+)j$BuoUJNy(M5kQNqvn#h9>| zIS@LJ)-i2pJWvy3O{#9F7p%d2k$p3d(q&9TKxa_4O2?o86Spgp6#HgwgDN$*q3X;a znBZ{wgQPCtNNAgU#A$%sf~1p-5dEb3FSl-cu;N{PQW1Do%d?l{i5}+HXdG9h;|lKk z&98j@%N`=NO0ixM?Y3U_*O;kaT&|c3-gTcSVUy@c=x4!w^sYDbG_2*2gum4Tvz;(- z>ejTAP=4tkP_$xO%~vYC?In074C5yNXqyI3Fs5z<@0|N5!v^*JD<=u*=cd3ou1lUI z7-i8do5Ds-0dB6yOtOJr;l0f>T38#P*~ausZc9d-Iz;u}J87gnixQqtwn;~zEt{^B z_R@RiHA$YAzYd#AcPaOG?xs<{4ts_-6UVz$ctKPL>d(c9X3+-Ulj2O?XP_SC`!v$Y z;`Khoq5>(L)_Iw{1m8Mq&)3`q1Xm%&{Tr+s!A(dFyA1tNU=UJ^&TcwHR&x_lj=Na{ zztvd%I)!34Q6!Oytn`j!PSIE`r_x)KC&W!lAbM#MYgQ7d0KiOgq? z3MKbn{J8Ti12~FiF}F$Ga|4cxM&M|s*b@{vMn!fqwE_DidPN+nlFY*G32MOY_B};6 z(s=GZ;-5=n-Zi+#AqKQrnn^H@9DE77fm-^cSC?*IgTuu1)r_W2s+|FD_7+*E0)66X zO5|26fpcBD@4*FW$`v7yABDv$@p}dp9U@H@cq}q!_VapajBnJseF`Oiwm^D0S5=EE zq@Jbpo4wCbNcAYT{yy9QsUPeu$ROq1AmzYH5u`GWInN)WSO{&crXpKpn4=%#6+5i~ z!nN>-i3PjbSl#h50gA~bg60ij=dKwcOIlG_SimU>`1e}Y-OgH4~w(1UJlejw>J$efeVbp)J zKD_y;X@zNU4Hv(+1IraNp*XEd%yvwAal|Q?#$uE$x@)fcge}&`0!&OUOPAq4CH&8d z4kv?@D7^p6zmlk_5NSL1^@^w1bre}mMSlL)B}unuDcwgWOA}QMyc*HbNlH&PSpZ9( zrOA*<`N&DUI&Kh&yY^1!y9P*DUIPU+9XpNA4mSS5e(n3WRNm@g1$@TxX zLYbx3(A1Et>Z3o}_myF%qd!W1d)O)e$Ctjfe$l0GwW+ffVBcY?jljs z`}l?77bTmZWj8s;-12Nw5Ak+O`e?1ZkKZsY#%tKAi}dk_oibh9B8Hu?H8*q>lr=sK z5m|W*ejW@zDc0XVLdM<)FFAdIBNLb$YFV;$B<`L6)8Z%)ldwGT7DeOtr zov=KK21+LKD!lK^N>e6yRCwpksfI!{(ZW0l=yrGlzy#G>&Pd3UrnJ9oxANg~dQjGM z<|KU_cPsze)c~#OhQvilabUI8%)WX+b8PdHjx$^JNg5@Qx2`^jN;Z%G#| zJn_HlJghgar2ApK%t_r}4VLmN4@bHo%r&^80NIB*3h939*i-Z%Nfcu+9}?`<(0h4N z^OF=Cf>NYu$|dolav8_!mLt z;mNPO!~q!-SZD4-cb)oAv7bTVZG2-xCCM5uO2&aVmV-uWZ3V@aQjq=&3}#w5s-5!h zrHAEcvQ4LVL$p-M{B~B8>eEM`6Cu+)MtSCh91PH))h%jTs$z{|uME%1eRRLPjJXJC ziJ69!KtK&+=(;T;CxpAT^%?8J0UVFT!MIn?_#%$C&8t;Z?2_%KoSF{nH8jJLna5=C zdq7Au1LdNYMnxJK9FF+M>G&A)+y`L`K?=+ss861)N)O!PU;PsHLQ2ehVYz6pYMsxr zz;0j`K+h1&A?T4~h~AJJvc#+lsZ+&)fwsbXt^33ct&bPZc33e0=<(DOx-yTIaYA@r zxcWA=&$M6;mkr5rb$C)`B#^!+rIR~n)r*59$|aVHrB=EZgg_mo90kqKY5V0gv2x1s%rEr-j~MA@0GQQW`}+%5D4)@Bm(%xWQGh)Z~N`>*_S_+aU4mifs z*!&YWki0&#K=f;a{cLjmx$zXOG>YE6Pq70O>8B!7m>k}EUIsslUPcCJs8r#4 z3DQp5)B~?*XWaDYaw&2z_o|tm$CU8mm}4_4eXD(TLS4vx&IjENL8>)oIyTE1=|29R zc}IYIL@Rj2Pl!lUE){BLXd^zFg>*E_AV0fQ92-y%MB!Wflc!-&!&2s$YV7QNaHdLh zJwkTUV_P>s7*sg~R^hB|>T_h!^(@I{vQ*uQWnub$RiSGXqx0eKpcT~7$|Kv#I9{Z3!(1dGvDct?fl+acD{EB!^JeCU_&nb$?*(FjZF47v8w)ic)ejzN58 znzEHiQWW?XO40m-b*PdQdI#$suUh^|uWGlm{)s-lvUK{2hXyO&UGuC6(`1BijL=8W z?N=3rcF{}KAJ4GPbzm8uCMH0}N6gb(`~P)dbBF;?^8eOsBPCOzj_26um|BVjV@D+{ z1cbUmro2b9v|Wl}ts_ntv)T}zrtAU-4ss>0Qk|_u7eU_l z!j2FN?YreYP=E*}cGdKg5U+_@2gqXa~iD zI?laZJf=>1Ns<(@BJ2~WLHwEEvgcj-QSyLvhduIYQ=f6i!Y4)S zHqS;no}A?v`(({pTZ$=@4al0W;xuOa!on@4oULZ4Q8DE-in^j$7Pu3x{(-6SWe@|O z&#MxoOInmCh2{Qx{0qYi!w>RL3Ugg5cpZ|$@Z0ol*W0dF=h|C?wDS4dJs_6bJL|A} zOz5jMYxo9XGH1*8j*=DJgoy)twzG_cNixN5q)0p!xz``O(~wit3-0S&Z;XA>;6;f? zi$HhrKQW8!jp?y3GADeG9cbt4h?*|)W%p1+l_Sw<_Bb3->g4&pz@^&CESC+srb+jK zZ|K7hKg6t`qR>GoX`*$jeet%7CMs-nKssg+H3SsWrA#{;@7xZ$bxlx|@JuKMNe$6i z3+f>V<^UXBnp6)ZP&a9*FqY(t1Z%T*O+u78{@?@0? z2NoY@EGby(W}ow?uBYcA2~`xF&GdgJQ)UtV%mmt7)dj+G`i$-~ zqy<15660atB*kcbWbB80H)rOx&1s}lq-V%O$(=AH#mV5W=bhHgkvy}Se(TdhZOnZLf()9h<#>)&cI7jT%)sQ zErFyszEzCv6G4x~fSt@bYw#c92H3%$zq#Htq`_hB^=5(;+LV0}dCZX8WnfE8QtX|- zmcM>lnsPbQ;Q|HUT7&u}kAS_Uncn}3uKwFDfBdl9U(=^b^+*xmhjek43V%Y0&|SO> z3=T?F^v%+9a%8FYcmd5SItd?wbUX*-o7t-4WZ>0D!FPipF1H&nnWPwdyv}vOku?G+ zk5cMaLzl>lLgPu6swU_VuPUsZc@TV!WX%E^m|^*#v_Cvw44Eo{R~4FkVZKl+Xp!|m zMeOPzoC3#;j?mNyVukA>Iz`aaVTr}KVeeJt)ZeiNndJIl;5^y%+?=d-8%eNSiUq@7 zCKXx8>!a)WErKpV2LIgmE=iUJX7JB{_mTvIVk+c^z3zJ*hESGVdH~EnXGtlqo9Tth zI=#X5(!2BtWzwu>J8#RD%d{udnu&nX%^40fEdt2}6ppK`nX6?P6A(2YH>M;-1vFndT6>s;be;&RmmSgJAPDO!j`68h}tNwNi7i4nVN&wFNH z_cP^;cUTR$$VjjFf*-!jGY;a}Zb%b-n&okfripx}*d#4u_PA&+L1lnGI-3_wE=h8k zopdL2BXC%LH~6${Jd>!cUr4Lg+}6LuIP-DSQ;+_BN1Z8W{FlltWQO3rN)RQyN3OqI z;oYyOVTN46l|1B{1N9r@K== z>V7?`>aN~!-vB|t1vCL!B8Vv9hO&#eD^RFWTtgymxPS|y!aFC4C5c3HAz`BZ%3n!x zmydYAf6jNlv;6M>`=|!psk&k=fy6ch-~IP)lH~ysK;hrG0AxYX__x;N1iR3ZP{`g&Vmyq*Jy@a*#|E zCUJYXtAhGvT~J)GYSw2WEaL8^F8is$f2ko0xj+UaUlsH~w$d-jqs?{Uc5zHxqY9$g5LftJ$zfkqOTtzZ-LD5a=A*qD1^VK2AQBB-N?vimeih;-! zQKw87kJ0YCRAP7o)drkAsR58 zlD6Zi<@NGIAo`#s)yj^@fyh>JDK7(dKGl=@PMuH(2eWU}zP0I3-?jB^?YNM& zpxq6Mh~f(WTv4;Mj#s2ChZP`8a#)ZRf~1kTGAESp!Be)PQMo4O7$WAma#BUpIy; z&a;B**wyDesYw=}RhhiJmfGmXrA_;+oawCu1M+5CBEcvt6b}TO|89#v1$z95+)E%!4*gW1Cvtz@5K{)95_ zN;)HGCv{k`lWLYV%lZPLxoDr_g!n;N2heBYpf!pIkvchs{0;=}mQ=>pb5xjl-XQPc zLMiDT;YxZ#RGP>{xBC3kV6ha!R!bH}7e$vtMgS64sED2d#GESp1Q-IE*X3UpTRS;V zjG`OW5kS#0a_kKOM<*O&Wvj+a|LgbrZ85pMjJFSX zq|-%ol5d|+O+=EHT3+OLmag`C2HXl2vO;Ao1=(I~^w}lbWVt-a%d9fyqS6s(%!@0m z0;_p(-`eLQ)2?wV8{>vyk9GGaRYGoUEndHG6Q$6Vy)3}KB}PWKQa zxaJT$Bd1ME+HR`|;I`nf!|}f_c3*6r@Q zR19kQ1G16#`K_I@%31l!-J*@KUP0x{A=$&26P`ID5Ql}(uNpk#9f2xzfwyxub5b}O zKk&}vcS$`^x1uwTpQ)IO_1}x9E(+cp-3y&>DG_M_)zRR5r$k&6xA-66^iIg6F}jGI zZnwl&Ajmb~u^Bw~eNaYzTwcO8lE#qZ@(w>$CKU5*hAL14RS0RU^boi8qCX87^LN{* zpWB)yJK;#&yXavT_CVYd(WXZB`}4KD1B z;fmv|a4}Xi?xXK~$Fgwk`El@1siWhl>sBGQc7o9n)F~pNUD5?bYN-%qfQ<#eOZY4} z|Gi_f0$__Yfckb0*y5ZhjHo>!aqnEDyh7)C_W-Lw_JsZZc%&i*uRKg02)zbd@oUlh z{jr3lQ;FM?Ob`Ugqz`*SR78rwT|J4uYLP$>D$l?pN#b4YbUkmse-}tQcL6EEWBARu zrLTpWsEU|OdcS|W1U1Cf^8GAI4cJVg6Ji|>&o`?{bl*OV>%YEt&sK=-vVe?NdHtN*0b7QEp?3`;i#@8b&d|f9Ps1@oU z68hpYxWL51xkPu9<&dP+a+)ZE>1UEJ#+M(+^PQS&)hz0T6DWLhkaf?VPIxy_)f^Y^|u^C_evzoocF0h0uU zH^MzA#9|MxE<&N>HrfshiPRbPs475JeOKoNtVix7Tg>!OrfqtR|VgdAnnzasFm7?*3*i8L?anRIO^)-Gca>KY#%)5d9$W^wM(%@Jzo3)ew*1<$td4NSWv zlTP!g3992CBsJh%0BoavzN@g8@klxG&R3|BoZff~i+=P0--5aK&#c`~t#IS~tDTsm z5qj;fPgmhq4cTV4aW4QNQJ;6BV3pTVAf|>*zdV`QI5~-H&mh=7u93fI351dVLI+2% z4}>A=NX8zIk39B~1suO?`Oh%wmK%d(g;lKRDZxA@sE0(t+Ao+KB$97Dn(@AoTa;ju;{`)YB?qc^^jJ5>b-dCIfy(!GEve4 z+yV`uXa={CWnhj@%{l`gy$f2i@~Bi$%yrvH5i9tO>DnE6(Uy1GZR1LIn52Ofgv|EM z;1)*BLw*y~^*SA}3E1$o!I*M}5X8mSll zjf2K!xR{IcZ$y5H7V)Iqkr#*uF z$E9goeY-*ovKQCh`|A3Hm2*zP=57_pIX|S3!B`iMbWG+~XQ)bXV^|6hG?|5-sv}Bb znz?6%c|Jz|eeRa%F0xd)K#&WG z+3|qY6xK%;0PQ%YKLmWW z2g!%PP>c1IYhrfDZYlDo93*difAN=fvz@RPa1DrtKn2S~<8vpM+Y+a|QIStO^e6|Y zJfCzl11F&VDtHdqC(SBR6`a8J7J<<`}~K6gx8Kb@(b*H zQXl{B*_Zt-hUTXy$DE?F*clo(_L}xs8Jc2((GpZXRA~buu(Jo<7F8~_n%6El9Z*3- zY9HbT`N{(^25^<0kS9fp{ZlygP(-jPNVUuVF4^LX4pmhs=x{<&C-SSl9QPUdOqoU) z7K{e3EKd3yw!<9zhPHJib&PJjD%q)HToGRpzs|pa)B}$?va4JawTSv4DWZ$N$!qd8 z`#6b$74h}{7ez%7AYTzb021F9MVb)vl?O$2WWPtcXC zDCW}pv~X1h&Ox$8s+Q++m&d{$y+sNc54go;Ho-DG!!`}}X4gr_*?}Z$dH*=V;-ozD zd3J*8_h2=E8>h-PSUD*v1hbN$mJtb9-}2`g@?>ZqT@iDaKN|FpL(QxM@EPhhmmh1~ zeA!@2I{ZfMZ0t}AFAA>a=0n4CTYT@F^N?#tQI_V=i;}skzj#bp3cFgNC}bt;IY+6j(evQH9$B^v z{gM#@!weLcz{5T;hHys4{G#y>O}3PNY$CmGyv}H>)|oVdNg=3}L;|*HLJ8{v!38li zUsTT0$@JvIkmRYZG$KP!w1Ws5`eqj%typ@_$F?#3Ey0AHHumRj$>t(>;ANV<3j9~P zlB|@UpVA0@3MP3%BUu23Xg~K61*c*U#GIPdD82`mKyP0rjTQ?lYfO}(2Ko~hz{x?p z*e*gg@0|~vpDyM%*~UoRcEPa|^*AD3FRvuF`ZftR#hYIzM>YuzvggOzz%4;~gLC2S zOcC^?XvjxYCACM~?{P4$gsKf)$5eYEhu9;^g@P_41%UIX-aI0!C5QMd)A-nyQ0rw+ z){UdAcBBt5TU|u9-DR|9HItY~4jN<&$_JQ*3;<+E$vYEdn zYmOc6^K0T9uWBzW(CqL-A$5$k6$4$hmc;5%l+oGXIe((+Y20?;DaOeCC?sN4^`*vc8 zJ4I_Ox<$2D<*j_5r``$D? zzc_FuaGCY;k9yy5LYJexr})xwR?|HimcjA~?Hh<8JP%^|#PNT6)64Xem<@`Z)D9u` zimagS`Ze==z2|dEVs!Djlgqd#-t8v!R5!2J`w+EiVl)3)@J?B$FoUWOJQVAC^S}-z zpKnQCLYuP>utQ1w$(>V!EdFW9zz=s(Id1Hq)>!$c+X)7?uf;?Hwz^)6UQJ!&O=1BIUVnp-RV5$E?1wH3S1m zjy*&I+HqiG@;@S}_G;$uhM2XUzb&pE;?#?zP2R}fVct)oYA>|%6+pS)Rqh$irl2li zCfy>%3LlN%aS**wVI?3Gg+U|}-e5ZoQm<-&(ym8VEl{D@+SQp|WOKwoq!U&Ihv&)a z8~eI9BcDB1^W?t4>j~dz{JL$f=H;!h`+k0V?3*o8GoLS_)SS&LA=)KtqChVlNj5R0 zdegT8+H;D%N(e2q;p@qru~&SxV3As+DWYQ_S6w?^8(2c%A1AqWP$d-lSX=^cWa|mP;FX^V%f_8r#z6A*C`N*8=U> zPUVyEF4ARX<(6kN{0Beshx@7#>_7M+b{p>drTpim--xs=R&-l$#ZLL5*=U%J2v$J5 z$_Bg&4sh=ENPW!!E;`K$sbhHmTH)>Se8=9q z)3;cl%JIBL?Utyp3C7iyYKZB3JtR!t&k%RI_7E#bjP1L<@Fnd>JA{;4a>In=Y0JD+ zNIY@V_Ze_5srq~_N7X@XUpw9Bts$!?weXXf)Tnzg>)$C5jLuip!D$%miMsarTSjL! z4end6S6TOXp~_-1j>b;rQQdBAGE%JwR~C$ApsV~Tk#JXX%D-e%mSj1!K;(-H$o2>W zeb0Nxls;cf?QWB8jqaE2@!aKql**kv;BlVPh9!n~MDFm+nuYB%10MCf)=B1;nWsg*F0K+Z@HBb!GW z@W9tE?lpfhEKooh4kTBAPSat0I|hGKT=tsXw1bnt7ZE_1%VoMQlc{d(kV)XjP9wb-_Z zi4A1jxO&A-$&=crC;Tp*_U`kagZV5@M->O-xzT2`fLUb^5NAVS=<$+OKbiif*FWmy z4X|^o2rM8gXBp^I6YhF06YNtaGrbh5x?wNz=q@=d=6dK!anVs$tIe2``zI3K7O%DP ztN+wWtsF-cS$VD51hbBSe{2FoV0o80b83mQN(KsSuv$obUtzH7!}C?566-+R#hl6b_J>YDL?) z&^_weB1J~zP0F?iEMK@f6FJ^D#ui4YfERG;imA9IF8002Tk%dclMl=#RTEpF9wsf; zKtCk4+{C!Fh$q}V;&#})8$iS`F|KmfZRrzkmZV5&5_0RIlbJrBy}*oNN|53`3#e6G zls+^@r~94{Q?+nQ$O1S&9~k97eIPpv2`5!PdCerKa9f(ngPpgufCLg%V>EIZz9k## zd-J2-ebZHsk)ORyjvU)afs3iE(Xx!z7RdoSX??>ey<=x5`+w>2cY(zm{k~%SAE_jE z>|5>|!GtoY!I7d31hXEUa?nVI$PciTYRD`J@SyA*e~Ch+(;ZVz&BD%g)v8(BIg1#b zOvl;J8>Jo}_w{o1xht-I*8-1aA5}e|j=weVxM>BC3j}kHpw18pH@V0jn;x(#NXzYt zJWp-#EC!{Wzjj5ckdt(79ni9MMQ)|$&aL3C3hIjNRL*_i)f0n)xyy)(X}f0K^~~X2 ziMS@siQ4_{N>CNd4p`5qefoS3ghEhKUJ7F^BdUBKNVQ>FQXAG8iE3?$;l<))W;w4z z_P`6cd!Xk8k7m;6r&VzANG7a_X;W3#I3>Ujen@!;j!k+PF> z-KVUD8b085gI)kowCRCz!L*njdT;a-$XBGrv~haqOuElUJq?>g)T6aGWRDs4^FN$B z>|^G=wXGjYEcQisPxxCZmEHQ~#$i0jdksp2w#X_j2|YL;q2 zxuI90_tC{`$YY=ku|Uu$OY%jf&^&HQOiok}?21Okfo2aL*HnJn0_6IO>Si7q5z7DhX;Ak$_rGnRJ5!Gxx|i0kMw-0u)j~x~!#M zP`A{^!E-617U{$3-BY?iS)(xuioG$@f-#h3Uh^;vm`S(E&DC!yB6JKv7g&agY%JU~ zS8U!kX(l)&CCz%49Pt62Flnr2$$cv^|8PG4?{7M{Xg@XIYZ>%@eu0Z9?{TVr;-NB8#evh-l&lo&23X96{(CY+vzGetk=q2XMga0eU zVq*@@SzSrxyYU{V!OHmUAQ&h=FC`K>{E%%4tIhko!EfCjas)K&!v|aLRCy*_cUOS+ZF8IKE?x$~h$u~w}wIt{%HU*|qd*7Xh4po!u zG!N`Qc?T!i{s1mI!^&Wc`GzJ>@}@%qJDN;nZjUA1-g)BfT%*%h2gl~hAxJnjlvA%_ z1&Og^)v2d#sYTtkSkewzD2B@mlpFo@eB}JBhz6!V4Y?7teY?m+qK8I(uNrdg)CCYh zoqO4DzF@bh&U)_Yuk=c4C)WteNw@(d738VBBGBE%9tT%=t`6a4NEc(!L8l_cJQzE? zr2n6|s0kKVwf)^Qzo#<+3FtHxT+3#4vgt)o+ z9lq!D9nBZ&{zffg7j<>t4L!BV3RD>clSWV}pdF{&<=-Z66{@fx25S>CryJ-+QZ?l1 zoV|N_{nR=PqSp`Q)GLzToBEqKz4Qr*S02cE=w^N*@1(yjzK~qNOAft0T_Xpc!-(ZP zAP4kIkQzLqJRwFhn5Xm&`2F3`_mTqaQg^veX#>+5n8N75(OB+Nr^b=e<=3( zzhKZB=rEiHtyHdi`2vybY?5g62O3)n7dG2vH?~Q32w%)>$Js>69N-hdP7)wGZBTTP zdECGCaZOC8P6;8un2=Y#Vuzwv4&~I7;qNhj47RLQA8l(~Mdh(utK2xZTyM2jRS?W} zg4znLazB1ByH28(qjCotht&5!NhE+#AKhnMy3A6{h1}~@w+!3kROmuuWoMr#p0+kuBlK~&o7Kxuc&@!{yY6L z71pC;CUNfExz3%`7mv~A75mw?8NfwvS*w4T5QHM=IC-t?Z!5v z%F1Sx5)2e7ZHDR+xjLv>vT^d=fRoaSn47)^8aF%x9+>dfl06WNLwcutC5kT8z~1`W z3=N!FCb$jtj(UE2NLOSJv|oeX&rRQ)nDkgz?Ty$1W{8F3rO9^20d|1-&@e$V5)lSB zc3JI+FjNT3xa|@&KK&jj8>T1QB`baUO`yW1s12fYO@U27e78OJFy|<iR%C~0AbH-}fI!rK41hpR<)^Pul#W^B7&nX~pN2<~TGP!Fe zG)wNw(5x&AE*F6?u6GYz$6Lp#RHz!M4SoZP8i`5S%N8t|G2nqsu?4`gnZ-Fl z?SQ!6dEjZmozeQxX8u6vZO;8k6@iapVY}2XLlVPGsIA-)cz=qXES+%fy)LpXGJp83 zO3uU5&=6w4HJqrgVZyC8%T~p)~qo4Dtc3XYlXF zIi@W%wK-~E@F?w)20eWTQ8DO_(RrDdaP6*bUXe|(&5eU9cCwIFp-ZCLC4Jt>@b{4k zt3xnWa+ER1ABSW|ZT4w}=$o0tWx!*LU=!1zs1dFTYLg>l(>lq`$W9WgN9#EoqS8dQ z!cv!HAaUj$#LLho!ubo)=V1JoU+^QNELKGJC)Gx3+goErj$2ugT7rR{RV9%Cj9=s~ z|KimJFFC!?QS_-tfX7|Sl25Q6z8WLNd z17X-Nj&A$fQ(KGUwm}y=HYAB#D!f7?WpYQPI<7n%N+Qv1DGxs)X_xGtxPqhF3HsW4 zpK`hmD8O?#W(shGf=P+FK=}lmT`X~^h^dKab5-Ke83-8mG{53Tc6)2B*GKnlEsopb z#f}YG5})H!Bg_i+^1&R7JcE=$MtQg}UHi}&EQyQdi?aIO9;cx4giqWBgc-ah(i9~aLooss`u zZ_Lh-%ey-RK|`65SAfHzQ>+j~6n$$++%GI^*4Sr1nnD@btyylo_g-j4`|*Hax(VtI zQt)<>SGhIhD!Pld6guEilQ|w&s=J<$xRz%GbwKB{QIZzZOcse%m{$h24-nzi%QLws zZLrL1zysUuJ7rMvttHdFfU+`Hm9N|z(kv{Wg=atGZi~C=n;Cu;d1jFSI+Mmjst4Zl z1#7%8+}JMJ9;iy@C-Z^ClY2o}ELzM%u8{PIHD1^#xh<|9(%>5M3Ax8xC)@4?zecSz z$}V^%3Toe}h(X>FAhX3jDg3`!ggZc7S8;RrEV7JaStVVC(xH&+*fBGRqazL-q%F8x zn-+SOTEh-^Zrswi!^&4GBpAp}T^^IhqJf#3QL*zD2xGkrGkL%YqEHCZbe` zLMczkolJSSmb}Hk0ac#m;iqI+(Fq;WDG|sCi$m0Kk5UZ^tepXOxl(+S)`l&cs+Opg zwUi2X^w=Az!jLmc!Dayg&}O*sDAfk5XDWrofol0*=R+C;F98XQ0wrFjYLQ}HMYnWs z$Wvcvdr9SCe@hJ)M+U(1(_)H6NYIeVThA;4T?MdneLgi@Js-QaABanZm^Q&4zc#r+ zUTbC>89tgEGa|#g3bu}{_V1J88ZVC|#OTIG%1(%}Yg*-`ZgR!@m&CA@2hg=ju%J}W z@8&Jz+@i~3pYb-6b+VI!Zpr{9SP#i-(i-6@X0zuRV0EaI)XB;fx_E5$&*2ntF!1<@ zY5{_pS_oUC=_?Ao=2t!0d^IY;9RxP4W`2^FG1h!$OYCMJ9AhLqJeMMmh265fw5#Il zU;M|ylAhlHI6^MgJ28)eAUsF#z|=AfdTwC3&|oa}*D5Pww9qbm*Yhlt=bYi(o|q^| z;`NaQf>m@K6!cXRx!(6B7lFrG$2_8*N#_d+NnJdKnTiA`zk{QoduAS~M|RR5b&ZP! zc)hu`^KHaZ`jxefQ!H!Z!=Fl*P@CQw=c&%hc`7FuP$@1U640jLtfz~vQ_Y@6NOtMs zyUFG+K@og3O2GBTl=(dX3i>VNL6J^YI~i32FNwFwa=EGRtIFQL8GaThYx0#wSj4)> zO7hW+d%~;U1^$K^BY}tQSKR#GNd{=>!AEe^|M0QCk>ugscrmk+htG^^=HH_)(GApQ z@));Dm?Owxp7PdkbCr#}dH{%?#IqO_DCb<{WC@%Q6mV#MhJg@=#vSQJ)4$*;j%{alZhC~tIc}E}iI4D9o5@Of zI<#)9&fZo7P?xp~IK zgMxZ0wKL7aDVb455{YCN2!KeoP2NTRIWbNzzX1`O?Z6L?;Lau8$@<*-1mOcVwuJ2XeR#pBPw$oRblnzE+80aSke#))v>qeWxhGQ zTXY)lsaO}E1GHk;?g)f3+!AOoxh=%MH%LnanvfE~8O}A>yWshQD*wAs*>C>Uf|KtbCqAG)bK}+No)v&Q2<9?DT_6&cjYIm)tdPGpavS*9 zsm+`<+|^%fXA%W#MJT(pbet}BeQ2&IgVX1;%(o-bOq{ePEQxm@bd^tO5C(oysWY(i z)<*XQVy0(_34dVjOknu?u{*g)GR1y`)%Xqi>pJ`^81oTZdOS8 z`xSnrz&gDWxYSA@5Pr&ZRq2FO-hD|k|6*_>jDn-<;`_XD%!kt}XWb<)ygr%l9Q(p% zK{H?bx-SVk_+OZ_bL@XvP}cPA+d-6J9JRnIs&}7YV0Ld42@Q%IPGOWd6{=e|5LO33Otx5p4IHSZ$ZwI^3xxX7T7)fe$Y{B0lOW&8>iuNtW4o* zf>}jSNsxw{g<^48&|F*>it@-<2H;AWfziau=FzS{YgK34;iu;sc9uw3<5SrmMgknoN_&mT)UPpq*qnvw* zz9CEUQY*Lp>(lS@Z1wED$6Ll4G;uOds^xj0$D5 z6!Iam@(^UWT|ecT^giVC-xAG^s~>(h#+KsRZ7Y@R&{^ljR7YH-(NWjMYXk2~>Uo%Y zKro(`82NSNnt&!H$O-duxjErmVux>V*jWVD=84_#!@ioGc}o1tn)TnYn5R8I4*n^1 zl$`_QzOEg0-D+iQCm2{NPhstBv+PPlvkVjqo4()YeK16oN#}84&(J8=`QM9Cq2!Rk zat&$<<$CX(f(8ZThcKR$C;;CddwvEK)sr&mEs7G!0>R#QRjFa<=i?Rak`nypw;Unq2LMWdvVBWFBC$8+9Pv)_v<;x<{Q?2^uX^v$`t_(eQqY0iwoBRTZO5vA5&`G^Of91aWy-|wo|_D#~P8{Y{3 z$I@0^47Epx1>0nh5Cs__Gs_A@4l?PZRO=+Qyp*~j9Pn61K_x!jOErkjh_Sf8hnqWD zLnbjvOtyTxB8f2-1FeW{5Dj>=OO^_5x2hEaN9-)aoOwLkVYUHN0%dPO$r zFE@_l*r{%;QC#6*L3IbjB=f{g(a2d@CR!lbmr(j`j9jY9MO7g!@;g+4Cr~D!#3EjS zIzY&RidX1K(D@M7PBuzZO;Fqc?hAB-+~d`76X60ar;MYA5+SVJI3`r#^)0?=wTeYW z%2q|AuUev7AXq$geJEI??YK>$|&iQc8j~DyJnpbYbP#}LWOu^l({1tjH0a77hNDnk@~V3a_NyV ztLA>=h>yjI{?7~4Tq@0t_xMoQGpI2&mtfQcl}#jI(0V_&L6Ha|c2nTC^-y&` zzrQ8g;%*2%XXjDd*~RSK*A!AGtX9N2f~h8`y+p#|aixk4ihkKf-zDQ9kq^XS4WYYY zJCzFsl|+tU4c8#U(sn(6;fzb-C*DUP+@>KH#$|eGCGB*dcL`VLua-adZv=tN1EB@P zllPZ@DTDZA+dsAcd=ak{HrHKbZ{(Fo^iwzb_RFe)jS|Cd&iKNx`zAQ(r8?p35&Zp` z$3iU-D(@;=Kxy9^2pzD3&~Ac(zQ8gf;l#8gUKw3UY!$aNnt8--sv z^ud$a7e9C~yZXQGynk-CF8)+JHV~fSV4cpL=<3N>#;oqO+d7}N=`rhniO{H6;YV*kRY9rWKG$UxUpq?w`-m-!V*^!z-Ip4uTkO65-O zg0j&0g7Yq~UHp7P*iQH7H@f%*t0j{#)Ys(yV=>Cz@LDqyay`+>Os{uKgo#V;e$2xd zGlLwal3Jt%%A#O%)VI~MY6i5W*9SiH|D0lOv(I^miR46OLdD?y8BLxT5-j7anCSxQ zf3_I2zZ^LibalbSbKbWsMl(M7Oixw2armU&YI!+MFd*o8kVq&NJTIgwAwX)XSYVcK z$cchPA^cg(xg<^$+=IFT^B;`HJs0Z4XxQqIF0zCu21aFVAao{+)-qY~B?M~S^-nGp z_Qdr0T#l-Sw6{)PPxcDuVGf|$3t0|S1;FrvwHX*oGX?poy>0~5(cT|^`Ri84Q#!E__`zaR~PIfy%; zNQ|orshzC#(uQeCZCHciLPRq+b9{l)2qNEE6S6of-%pH7<)wcqF|G?#^Dx2KMY<@> z2$Sw0VqRfVhGSD+kG9-jP1zJ?LCmLrHdIlY-FRIBs?tFi*h4To391|^w~=)xF>aSX z*3e6a%=AF1 z{8C>diT5N!eV(~4r9g>kgW;?+cAvfz!0N(i>Iyve^nw zSp>6|pfZSrJ72y8{a=aU1yFR?>)p)Hj;awJirqc!q<=HNSym&gnR3;;M(A?VoWb+Q z4lsl7bnR>Vfce?(uT_6q9iUiy?q`PzVz;<8DQuBlQjc2IzaS*f`42e8~f!xbRA^|zfu#pIB zM&=p(7KCKd=ib|>xTV+^4E2JE0^D>i<<$rqLI*q+haNDU0HrTzNpn4P96`zF)_L#? zXF_6#u~?zReI32bpJcDF745d8cD)-Yst!&|g*G-*Oak(AG!%h!k`0R5(2~FmsHsO@Mg*`fkm`WEMVucR(trJ?CwD!H378|0W|Gk*n$!>$=zW*}nGb>YgfM6O3 zs+LGF0(S(KPg_8aq>Y($zpOTNkNB=6gDNmJssPgUeCQI1X$M}369Ktg%Tq+hHna5P&ON`Iy*nb@c zTg=zJ9{%^_dyOm7Yzb>#_FdiAN`gw=;Du8|FdGPJJybY}t}<9^iS=pyP&aD;Zjt5E z0gv>EVqYv(GwBd1ahu$v0R4sVUdiZ)R1FIU!=GM<@3e0gf9LX?W14_2Ybv{xrf>u zY~$bMpE#qB#@70sRC@1rHz}vArA3q4*27IFOQeh_4N*d&?vzqy7ypu|}6;GKDBDEqOwDPvu z%QRBW7nPD(J8U`!@_&Q-%_dLs8TD}ActT!5>ytEVD7=whV{$RVvoyHM7fjod;?uQ z>9I`J7SSHjE&&cHd=GKY5KQT+<)^9bqmf&2Y$j|if~%hYi*Z@ag!{UR6)T@b&9*?~ z+Xw#FcB+cqmdSnpWz;z-jrmjsB%%VNe|`hS;)^6V(oU*ao^N_<4i8T8?q37S1UWH&KVk>^My@j)32i zZlzX<32to7ubP32DR$dqMKH#!u)Wo+Cq8WdYIxrY+i!{e_b+hH?$x&gKiFtnnCZ5D zv7J&&G=*!Tb}CePT%)Ya?}C>KQ$Q_zji`&Po&;9#3FLsZRO#cpgs7dk%O5>4P?4WkNN)72 z=hSdBxy$0LJSKKe9K zKD)y|G5yys)52#<4J)Uc6h>IoADSy?vN&g`h2d*Cum?%ul#*$&SLkcr>qF0SA3^#2 zR-Yojv-DZ|ng4;o7rY%bA!EOK)+6Ul-_bAq@|_Q^&BhpA9{T`d)Z%s&_cTsS& zY&&P|)I+1wg#YJk?%Ecoo;cFeqSA5PnEniE)rDk4a&Jz5lkgP z?IIGeFCY6{d#R032fWMQECP%z_;}`ccmr$?E=Q@bWb?kThQH#SMFD@SA@jrr&I6e? ztcQySY80E80^r!kM*cqU4k({Nnpy*02s}dr9=j!tlWHQ8{83zFo^yhXEWl>p@3B4O z1T=oU{VR`bYiQWCWxKJ5ZKvV}JZn>%b+s3kdt$IEQLsUAmAimSj4Kvh5bF8pKNXU- z6v{^FefEU!@-6e%LTn1wsVC$K@jzrAEN`_lH$-*#b@*u@Pch)}gtJA^%WI3gM&(b* zqz`^DY}a{+DRnAzh8TaihnlZ$I{l`1RqvYlxjYb-cLPrlZn5^v1V4G#%pS3UUKg-i z1iU{*oD5kq5U34!U_Cf?^MPG(l4%(F`HyhchaS`kM_H{-WBz$v@zJT4)yYGo+)WjX z8zL<2#s$6mt=6EO1XE5>+XxUX)AO%SGjJAABf<*tB{Oop?+P1d zfAY7k5N875!Q^T`cCSs@y7gZv)303^a5QKO9(2eCr9ruN@H-0WjTey{Q0 zZMVvH_Lx_I+zoV-nR}#`vsHZ6dm*QrJPE383y2Q-?+;`s00{b|QD-C{7ZF>CkFQTd zu0hTBYQLoU^4_o9pN$kedE7(KuOqhtH;k!nE&p%d~+s?9#gRn{&^6aXzN<`Yl~O4Y*O6qF-Usew2; z2ewFU@>IB|k-|G__+3I^A#0Z$r3QRdn7vK%LaMVTWDf5!M4FSMppdT<$~-Vn0cnSg zlOdI{i7`&XPhhTr?IMTqC2_$wVu3p>GK7(9<_XdKub1xnqs$PjnpVH zX?G+2P!Xq>GRhvwjHEH7kDDbq%IN$%$utUBGV7JFI;DW_H-18Fq=h#<^hUpxJ_nSR zw`Q)(wAic(c;I>QqIF8#VFV4ah&cVwAh5G?z{dP-tv*O#!Q1aE#{ZE@V&~_$aWt>k z3Y{AWW<7q`gzX^0Qb?{69*on+J&JAMFX7cfcD&@DF8u`6xDkI)^U7IAJpSjKvxnh@ z44MfRFWG{L#n2IZyRqa3vjZR-utfoEw@@f$qQM0g5pZeu zzB_7=FwGkaT#RU*+1`A0)4xx*rHXgkSX_;jA=pkZ(34V3B-E2lfd<)qNrv}A!Lm^6 zcWv?25ql=@3CxbV3Y23yARRNm-xa&nR~rrcc>ia-OVhUc9+6&})&gDlXf$?++g!u2 z$pQp+7Z$b??8x=3RvO}8>aJYyvaB?ps!sobS~gUm&W$%ag;p?HM=+TLwVFui7H0wl zL+wmef#@Es=Im5}Hi-d>q0(YIB6V^UEzv-Nc@)hCJGeM9(ayQXK3wwVPm$QR(z01) z+}H`R(-C|x^orjA-7sxc(AAm!Q;+yH0YA_cXqve~S1E44SAk(5~R|WhLujz#qSU;nXz2DlhMUDOfyc`1!+PwA*T01 zu0sz?wAe08&M_%NjM7YLXW443v9Tp?ya3rzWP2{)fSh;N{EYq;&;x+=xSxqS=y_Vv zHfV*@^D%w2M|>~5Uj{|Du0zF(2^khXUfk}oSE;{DYaB@poExt-c52`*Kq^ZYzirk< z(GKx4-&$T5na6GBui|VX&xPiR^SF2YkP6ex$b11!S}5K1KC=>o}FpE#@5-bZiZb<-Id;lz%y8g)KGa%f0w+ z3=lhXnK_(G^dYjEd&uuF`enVhFC)L>T0`J}IJX`CXZj|BP1%jY^mJ zMt+&BY8E6##n5?6Z>0>B0q6}Ek9!cAA9Zuu3Qn`|tZ+3|M{Xvoy)t}rM2BQ|{qD+^ zjO+KnYZ3*ulbdDtcys6O;68}NS9PF$gIeBM$|d78!z_kzCTg4sCi`yhj?6pDY*PSk zyNR|F)UWn3k`F-f0GmE;`l6N*6wV%#X9QvE_-L#%gTLSoLtyYtqwv~31cu0okDdPS ztN*Xe;_qrE{`w;-(~VOA6;>W>0l@%HG(^Ju35kO9(arq5vBx-9B6ab#egmL_aY%;M z*Lwb^o5)n&pFjEACOFm#tv#p&Py&_N*lmrh4;RIE{alG1hjB8L9dIFDSmA`IYdyZ# zwoT|Q@pSFbgxr><^0r1p)wWuRTp?I?P%JtPqG&gLQCR_o$ZIH#$qITvTAaLlhC3ZIHk|H~(TN#XqQ`9*E$c|MBhpGC&(&&0lwO1K^M2WAT z1!y3ErswPA7la1RDQUlKi{K)sWkRFae61en88Eq%C_vf-*Fk3ZXfc@#mqCUdS~jJ9 z>)&h}=I!*QpAS8vJS$A&fi!Bnq|duYhBTNrXKd%FdSoa4wN&?v0gpP#QeIQwB5Ai& z9h4l^ptu5j1JzUW`28(dUD0ovKDECkiRI)j_2~ zB;tQ8SsZ7eu{RWVp7;}I_qJ3UhL;a`s1!U*5L01KCe6$7{b5WV?}OW#!zq z44JaC?MANmwNn@H?vmR%25M*U+&c1Hc!sht=p^@Y@LboKsTa^0^vZE`54vd-UjAU$ z#A&v&jc&Vk+0m0UWtLG=CX;?JJ(q&;%BfkU(I!`SMSOl3yl}89w<2qNFs@w0Dfc@_-iRz$Tomgmh#SLiG=B!Qypgxs z=dpiA&^;Jo8y6E7_ri|@McnO>Bg%ouI@#i|eh;8Fn%OM9Iw z0`#X7E&NqjE^{|-W> z!zH>KlF{oXoCAI7XOYJ^y)pfhHB;6D7v|&HecpxKngIB~9dyd~rCG|PRK zL$Dg*-rjsWJ7}?j712N8n_rVZH#J8;tfF$+dCu+|MgSFy!G@`fV1S``3z1+J?ZYw` zq*={Zn)YOE@>b|_JtD&`$tV)HZgQabMSrI&OuTT7sz}s7fMX zZjLfXIky$48nA%MAXoe401wQ0PVQt>YfItW^}{1rR#hWh6=YU1SNmN8TKqKuo$+w@ z+(s~MbI(T4-5qrbDs7$!7fH=m>G?f$wFqC=na&#}Yr>MiJmR%R$w8_tY9z377J`N} zW9&dm&b4n2|L)tLjs5$)mcU*l& z@^v4<)Vv-KBN%)Hw-ziqy7Nt!pXi~BgKxgGNiq9(l=JMyELjl|yhts4=ny=@D3_fD+>SfJ496`z9|v0$bZ3h9pl_EeWccQ9x$ME(Pr+ z@5hV=1{^UPFKrt>pBmS_U}rX{m^9&|aOzM%P^gS=*Hd7_|Ae1B3!pxdSk|9tLji7l0w9S*KO z@5A9hqXW(9mVi|;SpQqH>(et_*c58#OX}!1-_GqWJIwN%@2)Gs*9v$`M^gUe#yem; zRot-$X+16OFad6#urLtr5jIw7? zXmT;ILe#40j8dUmXzk1tUWHIKdgW2iap5p3p5HJ^Kfjxm_|AJ4|E2ARiwda9x5j@t zW91MZCKzal-cKY{DAFNmlS!lNdq=nuNk_4Gw>-Q^jI5@r?hvD-UD6#3ye(Kq+%BmJ zJjpltAh>$g0^jlzGKm?WmwA;CTC!1AAwwm;3K_PN;1M;{V)k)Y%*4O9M#I*shLjq%qNFBCSO<)N@WyJLJ8*L2hkwKHE@bC534hn1 z8*8yQiRJO&+~7I>oJ11oLu*Lv2SKBOJt5kOi=-;O@Bs7{T@db##qBFH z+^UdxWt67hkqNh65L`Udmf=|$j4_grZp{C+#b8{YwQ4GL{jG6OmRbc^9uv$%g6bg> zbn-qQEPGQO^Gp<=!>~BCA{xt{Rh1&+qr~`zq(Y;kQ&b|hH0QfSMx0W*> zVl3;YbSnG2b42q65T$t*S;i?O@lNzWE(T}I!4b}+dt@mb)9u$dC8n5Q4)jf#huN&m zQ{qvzyftDGPlW+ul>e+DYbR$1sfq#W&CpSs%w$6r3O@_q^4~>crEj_tkq+o6 zJN#-V=ZSBJY*%PGMqn<*$8KU%Zj&3MlNsbORcXT-6e&;^hiu7a0qGKA*dD5*hJ?L| zdTy{B{SCvrj+KcT`(LHrrM4vpc5?OY5=^nE^dO*ccSfq`D6Isx58?HX-S&>rdC%^# zyKR@=yU9QLuQ#0yx7@Fn%HbS=W}R)^E&gS(TW4OM{*WpI6_DD=cS&HW^vLzT14Rhi zW?d53bGoLPkEM@?9g(n1NS&3-e1u>Q5>z9R&_89((4P|yb!dyB>+TR0D?eDXT)>_gEw8_eg?tO%6z=8Q|EisGGnYwouUTk;z>h zTm(mUfWGf3<}pty z>vRV6?1OUnMKmkF)cm7=EKb8jL3PA+f3#7zCoYfe<6fV>*{3At3ST^PP2)ZZhSckrlC<+&|}%xKFwc6ux%C_0{k6dY|x4 ziP+#-?A$b&Goj;5>$P=J{10i19a@_fdX`!41G~AKTTr*EYy!Czf_GwwFRirXF#LWU>FD*j(-& zAa7bnd|W`Z{s{ZGuTS3>TnUW&b>w!J(gF^CD6F6O;G13e+Wr$Cf=beTzm3(!DIeOZ zbF#@zyK%|F5i2`WMKHihwF_j2-v49)k;^qokX@({cyWsPZZi;?$(Lz&cWH*Re`Wd(?X1k*@R_2w=PJwGF;SF)d5 z1iZPcf*$)roJh|HN@e*bGLKxZ$QRX-m!X)hhMaqi1N)b`EmHhDhcnlF`G7}bRAUsj zgQ(@jf@YbjGqOdR=9BDWj7}6_U)f{Om(Nq^WjAJ`N2jC`4$E|Qbgr`>EkcJGIT+x=Pv+}#+x)AcbRueI}8D zcb5rPOs^2>y#_q0d>=^h|9SJ+*tOxqZ-{m4wTB$X;bDf~B5(((Gs^vy9eZ1<7y zRS2mgu^B<-ej3qo-&V5MTXm4!M(H@sQqUk@5MrFPR#YLpMmcDNOgVsKy*nI zzIiBAEed}KNlz4B!(8Wad7b1izdIIMVL+oCKZ9Dn#;bRH4@krwgt`q7ZbMvlNpDLN z1v?=!T_;hk7S!-`@-jbdppk@8z*9m_&zcD20R^4Ez#fo50CJnPShT5jwP+tH8SBHjlIuwCR;?iJLT$7-zo z!2i@PL5BGu+y;tMeD;UD-DL;uo4A?!ce5V9GNE&clgSrHT< zq(pS{mT`87o1rAQBxZSRm0~L>b(RD_6qa!FeJjcNf=cp<_h^^{$I&pzB)AL%F9x=c zhM|%gKm20QUu@eQ-zdvrx2X~IYtRveAGm8N77Tcl%36JELldFoWi;^dVn)U|4wXSS zjlxU&P+|Vh8>R#Hb2cHV zSI-V8`G5dD^~>M>PW9g({^mda@GJ2Wf>}sViGyr6o0qTqZH}L}{~aA-e%#l?%TD{t zqiL2<$op4{lBi;KpMNlns)pa|J)ct&lLtD$T{J2)s;TvY z98Y98#dOkzfC~Z5ARw@G+^_1O(D#V)SIfrf;&*eiV^bhk36;NO1JBH;fr)Ef{AtN) z$pWg6yTG$0+B}k4juHS)Tg{yN9dI6z*~d4suGl&$Zo3-UaY*(ljr2j39Kj#$lG^dx z{8Tr28zNeT=$bff3LP^nj!(2><2mapcE;}1@ejWkVzDU)=d7-z@`tLTbK}TIgO%mj zK`@|JQA#9e$c=t^J{Tu0nT?#3s*Ip5iZajz&X7US6(jHFZn4}+`64)wNmt0aAW?l1 zLhFxYa5?I~Zt&azYFgm@KIa+gB#m&++?(V~v^fqX2i8w#;RwdD52f?B9)9sn?@Ke1 z<~&OQX(1#!4_a-Bah0LET+RE;1BbX9`Dk2inK)}yHrv<1~1}G5y zOIiT__#3{*Uetcs9?y&*IQsmJfrfLt{I5{jus$foI~}0f9bOvL6j(bMl)(ANdU1P@34IMwTEz@OFhv~+Mu~ThaAzUEPh#)d+l(hI(fFSwB znaI$9UCY-vdikQ@r%+1;P0yWBZs24P*q=D|@UTCYGtYeS>R0Q&U@<7|U%y`X|JeHy zxTdl!Z=d)=axr8hkUT*_0vN7L%^n_j-2k?xu5efmpRPp|GP zn+hVR1dv5mK@db`bHl9^EKn?5kqBy`0*VwWeCH&=B9TZQBy@D9{dJMNc3X!{%jWVW>frE}9&!*n`E~g1qGYB+z4dt@_uqKY5!!vXj1;e1vRpU{b+u6Ud}e zOfp3hp$Q0QonU298ML5UCfG0TmTn2qEEDv4Uh(Vk)6;HqeX^kiXBj=nx38x%YJyk; zVbs0$d}$d7V>vBP8!Bdc8o|;0lZy9AiUWfK;v}OOCwUZ;Ly=un^d8w=UN$r&Lx%$! z7AJChrK!M#tKsgEX)t!u7giO4k{liK733CoS>#8gov)>Jp$|NebE(N>ZpzBs_sDFy z%Q2l{)>tvM5vGIL@Et3xOpE>D{jZqE_$`Feb+FAqB@+!k`ZNVv;w&|qUljJ5oB^J| zpKtx+pTOXgxP{f@cid3z=5Mac;Tdhn{FHB=AZyrdd>r>LBUvUEXDh{Qp-2K1-9V=U zTT)BM`yLBiFFnfB401p6?53}%@NgU_-q(JnWGuK@kMD_BxsUa^;~gtZ_}c~Z7{0@X zVJ$@E)1t2|xyySH^zoZ#)we)CK^I!deYW5tSr&BE>lyDZZ=IkVD%m%al|fzNGP+6e zAm}l{BL|gfYQvR7PKU#XoU|agaC7#M(*XScIy2<-X+Q>C{ft-baz?RSU2e0YF_;q; za4=nG#(X#4BwK%;mARu9ZS`*QFpekHiZ`m z`k?^gGFj`DOE)>^dZsB(z6JWD*ZfgYwuU|mCn3uW_B`IV!bMXVSsI8=~fkNP^*snAC5BG99_tqRC17= z>+87Ilw2|~ou?@V9He7ZG}^?Q!3|Ki_^AYIa-e3c4hqe1`8*WZ3_fuVoiDLDXkyrhQKiiTtZq;BSF zFXZ^02T6ZTDQCmtb{Z7t)S8mF1 ze2n=K6FSIlbwfUAs$>l>iTltMTQftA>D)$sYg7@m zf38;TpSvIA2s@n1!s1jh9{1>q@D5oMzb|Z%?vQQdY@M5?SSR@8f7PRY^NVOP%{UQ? z8Y53Y1xAcaf(swvJD?Tgkkn7xPXKM6z?2Q$YwIBfVsiIPDubkB6 zEzyJw&A2!Da6C&VQ+w#ZV6jkp_*rvL4Ynk}O6Mb9AJ4CyUl`UGI271U_i>MSrHh_` z!Ls2>^zw;di;kC%X1lWpAz#C5R6$#nBsnTS#!FX8%R)QB zWbY$&eC*-8Pni<}GB`xfCE3GmQdI+!q(gqz`N5p?^kyh4KT0%6yT>-RB|&Fr>)EjV zZ`5$J=p7y>WttR_uYrD!8 zRxc&n#^PK$aGjCYvIeRCJwWNOnb@IV(4` zDuJsPe3u+&w9~7uDu_=rUm} zb4&t?PMvft^QYwkWc1@@j{_{^j?E5SOcwZVP;eTLq7L?@eLKI^5V-Aih7Br)TPvx zk=&O}343t7Y&6@QYwq9iPm7IKM=SqPKH2x$SRF7jqu4#iC)u$=zVI80g|@PK8%Qba5_|W1L1=2}iRjtc!!=fIXq{RuHEfbjuf^YVU-jHtcaU z+GvdbdZNfGSnz9+)QF=~?|tnm+5FnzsKf+EnG~a?$PUC&H~q&NS~D2A(-YB@FY1GQ zbE;$zGS485D!k>|?0yNF4R4sWYW78Cj&mJ7AgLuR*OZA($3#Igu}`!9uJ6pg^e2fC zB*GryuSfzrkT|fPTxP9X8RTKjo z*)l46H7^^?b_NY8HN{2Y>Of6;2#8aw5`9a;PjdQ&S$>E8E(w!>+>12HD=wNFP}bfg zJrs!Ye`R6lIIabK;gO3qyFy~aaZ`eKVZdkPzIBw%CdM|?@!8~9e!%R`Se}kC^1Y7x zeO-C$w71ZRo%P$V|BS>sFy95}>7zislVVaSl1xSK^(^EKc=T}xNaX@3gjUsX@&(U= zA=I+4*RPkv0MrY^2D!UAaU3izu^BSQ!3%5Pjx(OEPO}0ob^8C7dl4f(_By41NY=7* zSsXYk0Fk^=XiBD-M2c*pqT3|tBGh;(5bX8rkfRRNai6`AQUvdH?!5pVtqZUh|6uGH zW`m=#N3hS)!oIH_^fMx({|C}FBy$F-GkK(YDP|8v3Q@xx)R*{el8tU30RIpa?=cy> z3B*tg>qR*z(|>@)=V^*f9_2n=Qs9S(PBYshv0}YS)PbtQD+D*l=74Vc{r7RliLok! zhJ82vfla^AY@YUKi^|xqZ28<-eQOFs+;X3Xe$_4+b8|VLC~q&oK4+`Fu8LO(4trx%WaOVs&Dp4VuoeJg+cG#hYAjp6 zWM|pF^In>hl|}PwLLenAe?mjs9vRM18%QLT%+g@l{1i&ok31^tAOFZ5Q}dJE<9FEp z3|sEMal~ArlFh>6z<#O)K}+o4q9^k=foFALcA30#&IXRA-X|pxX#5g;SrH4MbXas> zb}Rk0(MkPH)4$CppE&ULcEj}18bOrF?}2kw=&kwrzBzQFZ!cZxkqYGMVdVt^rDJ)V>Nkr*hh+I8jN+h- zi)MgSMZ~z@nqL-HOE7tSPl|_ey~={7LN)ZQ`4F$WC&i{g*ue}x>vCZ7yiKHASU9gS zRFfKjN|_a2ahz0-G+&@^WBK=5_dz=A&3g4VP6ijR&PJvW-jc_`Lfisz^_ywxqR<9< zyKhryt!H`&9_^5~d8DcFj|AQj{X#F5H-#pOfAb#bXI2LvAO z^0t8TP)Yc#hFHH)%71&{zBHX3${3Leh7*{Mg?0^ zC4X#lPv}3(wpr*t!`R=I_Ee#!@vjs8S zWzey=OL|uh+>ti<1=U$n|E7VOrb&_`y-7zW&0I%e=hXsbKJ>6Clwb=^s2XY|u_3VO zhI}C^e%!~i7BolT*y%-$uof- zV;BrcbEQiLUE{eM79&*kZLpwxzDZgZR-@|mJOfw0oQXUkgxSh{c?mdKA#B>6pYFYH z&gEs}%sBAsWP#<|#+=|(QEkk+zz*3}x#_#+sMvp4|IXbdNtE{USKrBhw==3GDoKvZ z6&kaPoO0PwaE$AH?7DPK23Rk_#iUOa6|8vts`ouJ@ForI?(`zcD4KxnbB1KFjzO72 zwRbKJn(ksvzTg266OPFm?I?Nq*_2FZ&Of`|Zm(Dwlj+Qwy0hk-1&0m$S|C$2K$nLd z?oR{D=ontP0{X+aXVc~4^Qu*#zk#oExVOotlGTw43t=TySdQxw%uh zGcwM%l^C)iI;d1HT<%8MqEW2-}!=?wmlD#HKZ8RCDqW`*u?0s!a zMx%+zI7~6Xz&u1nZ}-Z2^XdE(igaEBeVl&sRwKV#ct5myZb5i9T^3&TrA~e$|3G+i z=vGk%xAN`kxvf#vE?84t3=%MFd5=l5s8p08s#iBh-GBGiH|oB%?cFWkY>Co>>L<2L z!_f-D(G+FVH{^q^8Qf|ReyN_DYm0gkmfwYu@5DGiC;z2L$y@j-Z3I#Ow|q~N6%M?` zPB($YW{QcYNDLK?u@#86!<2317P)VQ`Y#Z3h0OHSS-R)*VBIu7zkJGGurhhm`y1Tb z1C5~A+g-YXl48m!Qi>TSthYAk%4429-uF|l9v;+PKcV+2_XlCdJjo?j z62~!6v7rTgFKDTP?60UQB3s=^rvQ}!`i{h@2HhX$U?tF~y z+b)=A9dgF82cu5g3R>)t^UdJnMUxR+aNy*S1;K^wktY-@1p4+eyXjWN_Q+V@T=n0< z@>nP7Z7Wo;7^m>=-{j^&Tnn+%r6*)H!kfVQa z=1Mk+G__EJ9w)+Ybsc?GodAMv)h>AZhl)b(1G?TPO@RYx@;Ju{-u-~&`Nwcq2dtMC zal1h@=fUiA(#~K!($2@y(p>t2s*Z+Fr*RNX1D!02W$-QDnBq4a+$+r&JcQtWnj*vX zfZwo_QP<00X2zLsTYtAlea**t`SN*U_uY0_M>sOQ{p|lSE>{A-}F1nbG;bsd~;eon$%?_M@dnHvQ?hC7; z(wP*+Dyo3`{hAp=PNN5C6rSd!%<2|CkUbybC}$Za41}dj$}IiCt3Rq(vXW{Prn}aw zmkIDn{Xo?&BLf*O*rjZD!yWqbfd(W4uD8VPZu$fG+wM}fDhyZX2Qd3e{QbTH=HC$0 zuLP^F0OfXc*5~9#q$O_(yhOIL+j2QDBMZEyQOWEaih;I0>Ckan(lfJ%gPIF9bUo)@ z@IKiC%xwC`ao#^qr}GR|`RN>}z#nC@GDU(--~HdJ z_6EiX6xS#HACqKuKyhF{XrBpCaw!G`%`!&(?J`+AzYw&`pwOOohI1nr-Rxd+NWC*W z2Qrt~W8=`Gv*hv|Q1#I28*UtT)=3OfO1g;`&|5QglR=6#<4^ROyvwt9PA2f{z(}zW z_|?+dAZS0`4U9ajuDrc)geCs+g!4on@JsKr78NfY#11DjxK9*rM*Hz^SIAPb$$?!O zD2N)hbYxHrY?ZcC(Zyj6vh`5sgpJ-X&sZRMD7ygUO-MWrxFRX89XbOi`tEmah30%) zR0Ybryi2efH**~9@Nli$Dw+EY_No;a^#`(YMyBua?Q}BNTXxuVqXh+J z3F+e^`K2>BjS=qvL9x~nrRj|8tZnu9ZP#s`gk%`CJyX?Sg115(M=%u&@I&XG+$e(Q|1%EJojD|JT2}bLEmb-RVWq76;zY7MXY@T8i00krXOAGYW-q^|Ad7 z*FJDd`s4-?kRtaSTGIg(gbdgFigt+xS{8ZNsB%J}+T?$Ik{|z9^iO{AiI zfbjSK<0^sCNBwQt%s-Hr86@B2q3@uW6kIr>W852|FH4$YX$!OTse7Ze)vG5gUz9IE zl4(Umsb`}M1HNU#6$0%fd#|(jVp#vp8(G>}3_giVSuO02`vvrE5v?~T_rGXB9N4e0 zpv-cNb3*yYyFp8vqmGitWa+KNOP2zNAfT77`qCA@)~H=%IdwNktAEwB*9)tI(`F!W? z_shn!)j7+?Wh}k2^Z&V~=T(oXrM+|GOG)zXpiBm7)=&-&8{5C#l5{hceuZa%9?b$GhCfo@*hC_KvDAM{k* z52X_&JUrSRets$fQmjDoe4Vy8L_oXt^{Bd=Ij_BFGMsD&4yIY)WE%uCI_bp4S#%pJ z@h8b|$Oj0jXnc^Z&i3pTCxINR;Vf{cxa-4rD$fND$~7$!#i7{A8SgM01{+%KQo5uM z3PSJ-R0G+mjN@rqIV}+wtHES}rcbs#1O)OjnVYf$pt&&_rM!uK7oLA9#(6RE@6506 ze(#jOaS{6Uo}=+(*K4ygsyA6n%P0mcXb}~S`iE8FN5$oec;CYdw6qSC6d^j>AukYQ zGP&~yUGXI1A%QW3jWju;wIKs!y%fr)J&uKCGR@HOTNiqt+>k?hvC%nAF(ljXoJ-g9 zHqNv08f-h%OdKI&-TCLAeX__rVC%4Z9}B@*oxFhf-*J_Wf(S2$;0~4C_u_npJ$5^eJ)4jhoe8ik;@?u2JaXm1g8I2|yJg+27;2BDS zN<6^_$8>R${H*eP#2|2WZT#IIfBLLN%J?JM_6o~LDgLixeUr%yJFv@QAv2t%Mq1i| z&?f0=VZ3x%P`NcIQ z<@7J?kac@qRF;YcfzZEG*~s4$Q4oHZG>1;Hfv|q6Z8(os3}x9oj1ftrv`fGD{tTmo z8m0TKhAe02pgOS0(3&hUn?t4A<^u$6UFdbC?`pF({IwuRXakReuq(+Ff#AV8D&}IUE3k`Vpz?Sp z6|D^@3e|#O|2k4b-;`@kl2Xs2&;(v4C)-m~?%YoI$uKK7>{O}jr4x8Xp_*Q4s-!?! zr9g&}odLjl-k(Fjz5rmgvP^r_*vt<#qD1JtWEUxUZB`b@0gqxy*HKJ01%ji|m@`0P zD`*5_ufl<_ZfP@U?!@!1`C`kc_44g6cP%^?cza=MR3G=a>j1ey(s>sok7t+BLr&Wj z-2p3ohn)HYdimM_9pBD224`ks(>}>Lwc>VGc7|%b{o%>)8tu%tKBT@xK4NEQ9GJG$ zV`67+Qp|OVT%n@d`AH!iOdZ`C)g0U@>!m;Whu$S|s_h;-W^Wem4S5RD_QzzE{FZAT zH(iWn(FO0Md_L!$z2DD#=g1e*6dg2ZoPkNR4)w%$ra1@>U}`sEoz?(udZBBu88%mpMuTN1m}I+d7JD7 z>}=8YvMobOqb-`R{o5vzo5a{p;u*%EBeT0tHyEj^7%! zOfcw*&%Ion>UveYkKnV#)^?d;1FpDa|IdB8tiG|1>oqpLS=mJy*Pv%^&(4w|Csu!% z14neB>}S-zJC0&DQe+(!y+W`;P&@Zga1XBndUm9_*&o7Veq+bC8e`sO_k&;Sbvt3+ zlZ?&U;=oyC5H=YFh$M>HOp$nO=!rahefKHU%D)T}Cd;U0k?Epty4sr!Io>97Gd?ax zoDrLRoYjLc7uvkdHLfj4Nh3G8HL9MM!F?p_15(&B!5Y_!pg7J!rB;dT<@1nFx$;ij zKea}o`sP;ubE5w6qh3Q!-S3=z`;%`BIX(J^UEezSFWP|p;v<6As5W_$9Ir0@;;AnV zMna@3V23i#vs-G*@2&L@edOW4ihnPSIXWZu6R^2!tNi@(yUhjkr4D{N^x zNOom~RS9=-hn(=|KA|=sjVr z;LcyGUUfa%4#gFrp$s6=x5pK0GHU2D_%{V0FrPQ;Kw*BD{ok_!#k9XH^H}-~qj$Rh zU;KYSPP2>FIFc2x5N9n5tB2#mifnbBzhR_$&IOfDk|amMQ<~zV zz|HEUnb;Vyo&!lFh~Z_cTbVS)D%acmd>2ebVYwm-zo8;a?wnQ9q?w)2nh1QOkdbTe zNt5K(gW5VQzGOvZcgkuIFXdmlZeGXkun8jz1@;dmjr{7svZypgY~<-T>*x|?hdhR9 z-9=@}3T2(@;M8#<*-p6CGd9T!Ss5Sd|1In~<7qTL%|EGlpQOwn zl_uZXJc@xDv0YU3#xGnGFQabC3!xmZHR^WouzwBrDRhBspo^930(W{Y6LhLO77at> zyfb(OuVYc}`~>IK9y#hhZVc}L2gzxJ@_f7c3T&sus&ibAo{u}!kkO)r`nXbJa$mVir6?PBYfftSI+ z%5!W`Y~_$Pxy&+(IhVB6=3P)7c8#&n*K`}FN_p3PYsfw&6l&$pxgXk1$9Sdqr7034 zYE=3b^!bA3sS(RD&K4USfF-lZ99uJpWgLvbrk#F=%FCZ?TwqrHc6JQWvh!{oIQ9w@ zr%~o)55*KxBo7*PNb3PZdpxkqdx~?frzV$1=x7R*#i1x93?*D(Q$Xc!*y&hcA^p(1 z%m>0ka8nb%+NDueCcNq1&c}K$XcOL{=%Ty8|FuVijwQZK7Rym@tbq)xYs!Z!zW8M$ zVy5@~^b>My#u&hI+&~byX@ZLuiaAe_b5!&>uSWi6eVgq8;YJs&UxxQ}(1E3Cf^Hz%{;(AE!$6jb-o}9J4iQ6{kO}oj)>W z3iKAprgtdTv+e?9BlL^{l7*M7u=BD~>ocimD*n-ktKW3~<`y~g+E}7HCWyOCF_$QE zfr?(u>!f30*E=MOajx(hl5GkHGxSiHrjBK9agpRZK<-K5&_<{~JTL5~?@5vIgneZO z$U5?oXNSC>&gB+LuKI6REQ{28L}g*A9#|Wo8Rt_*2qndGym}JK>}6wdZSAjx;s>jG%b2(Y|QqKguWj zW{{6ezL>`-<}gKSkquk!{y=)we-m^LED)f$M0#Me)bKNLzBT|~L0eKLm1&Jac7FfD zUcW}^cHi{C#Q9y)GG(JQP0k=Gl6RAwOLo)cuu+`0t)lsJj5zvw!N30}+2O#1%mXF> zFQk|}3KXlNZ$Xyw&o_a=fRV`EbcuU~`!(?(B%_9S@xIxD268a`6K)K*MmY$%^+CY^ zY2>Z~^5vl5vY=7*B)Bo~m_O>^0$@FS50S@PbWiEG)7`h&_{vle??UzjQ73nxm-{c z5eHEpI4*t)Y_J=iS6qgicDW5Xosw7cHgT`QCaHVjIzfxH6Nt&#>h!>6f^Ojjf^_7y zGh)NHN8&i^W?+p$E+d{`XkuoCOG2j%tPbzdZh-pCN49(+pFNgf`@ zE%tTs5pE9sNY)+P8r9$&seT;ZOFCs8OmkG%j~_0O$vL6ZZsEG)=k*8AltR?pj_(%hT1$zyN8Wsas%c0&aM%B@~k zrh1cZJziK%_oOf0^xf~gVcrvh4JsViv{>lMiunnYOGTw1l= z@LmGxRA|uLE$pPb=|9&66bsXBSbc9{)~v^f*}`ODJX;<9_g_hO{?fR}yi;^+Ho5cK zL}k{QL}dmk<}pPcQqfE6_qEnK3Ws5B9%|d0a1QMWLCm^_@r4KEN5u8x$4LT*+O>YRV_g)3vjgXwNMFl}; z>3R1mAnb&ItS-YqvX}$;x6mU0% zg;58g6_EyMVA)__QUiME zyi(6g!j`C|bzb#edA{2uN9fhFPlUHdb%`73JjpZZ$Piu9L(cIoOSM6rK@E_n9CFIz zT=j49=zv#O{X66zDvmB(nx@#|e?h2|o%5QE<2o^9nBv6ZyGh`M)%rv2Y;_KJ)zuQ) zBnI1&?YW9`#YLkPWs5b}#f5aUd!ceTIF8pQNea3phh{bf;)BX0*!9HwrYX>>w3F_i zwDY$FbO-mzJ7rZ7P0~TPCh28T&%s*)PuOY=ts=dJ$c^?(;i998*eCloW-ZF%m%C?0^8v`$oRsSs` z?d$?C4!nC_W)k)5rI`B^`GktbSZ|&fI8(03g}Et*!lXLMkZiZxWe)b)%Tgzc@kor6rx=$; z)zX!GSkSce9al|dM8f>T-takCc!iSCVDi#u6$=gNFVMV)`p{(pC@_Nbw+hiMX}55x3c z`()v0aT%P<3BYvO>@Rqm>jyZjlVm|F0Lyn1eFxpJJ7=PA0&j(&WL`ITM<2qhoPiR$ z$+$2R_}W`P)e{(e61P~kJY(&;1EMtl?^Oo|_T{S;7DK^Wj8`XUDr#a}@&s^jA=ZA| zr-1uq^J6_QY!+2}*{}2b94lZdJCs@O|7&0LP-;mcLUV>D8(=taoDRg6M$z!LQA`p=HdE2)gW$eu zB_}Cp*eM+nfG40M^=`M^`Nx$#u8rz=U)yero_}L4yK8YPe(^cuuBYwGw{uAO z31{t1&n(Yq+V)N*F)nt%VvJ?!3PgqjuK4DRCww5ow80^(@A%6uqjrSx<+-D z^m@WIAAppHcbWShSsdvK?c&s^4h7WC&4KGM*ZXWDK`}P7HF2Nbl3}oXoYibO?iTZ` z*IRKaBaGH3^_GyLSB#f=+642}6a&GRN)#>{grcI_#ch(xIU6{*Iqmh#rEhv}emhT` z<69Mhg(V;-4-%Sp7nN|Z@VhK5cm7)U1kiohIrqWrbJ9fL`=L)=yP+WWp?90)!t9b+ zxNK=@jKv_2GY|T-ZQ|zAInc$zR`$#0V7H84R%a}O{ipw2pJcAN$i{hbV6WDKbY!*~ z*i;}gP!ZG*!UM=!sIJ?@v+Ec>f)!&0HtVg7nxjL5Z?Q+!}ssahWLhEz@^U z21-9$l=D`?TRF_R72ozcyWpy% zN4Rx%sw>^|{F%r3Wtx9$bBMxCAgRGjrr_KZd@QUb;-Uo7ih;Kl)Byqbs;Cb{ zs|KLrzn#B&M&5!V_Z;pCdd&>%3VDJ)G{>Gx&!jMIjzL?$w~XnT3xD`+^E&ERD&uIO zw)*J1*Tj34+ypq((akC>fKL`}Cf&1JJQ^j5v+~_fg_p{i@0Nb3IH;Zim`nsDwjd1S3f(;+`F2lS#-JuZ+Q*NyVcswoQT zTML=dUbOuI^D-^pUGD~NebpKKDB{v2E#zdYpXkM04hP0E=^@yYW~9B+)Y=XfD42@L zfBqS)49B!<4EcvI7W@c^fmELWj4J%q*11$ z`@+`2{sv2V)`axQ%fi|udE!$%no8>;P4Vuk;v^jpp5ky`&iCVwTnFSWzU0fkkBeTQ$fHg8`{K0-FUHr8ghiM2_k zm}H710*^BEH~{+sS)C;?Dx+Wd@8P1H3oJIAAo-Ib#V>1Yf%C(U63K#6NJa?utKuLj0Km% z_TXWMnHirBIKAqPw)8#7ktu7a6N04RpRk-T#vLf`&^_R=GSe4(kxZeV#1;5^0!P9u4Dm+uI<3Ho2DN5z1)Ob3NbDaD!hPkzP zI8+N6Q#@3nT(c;hcR^zK{T>h&Ya?4ITCh`AA08WcFCLxS+q=fR2HIhRv=+3XkjIBg z%vvblL9iSqb%7e4tVvbveO+G|1o5{t(2g1)Nq&vc1CV~|dVoF!n>ITvN2X7BqO)!~ zitXRAGeAGS{lhrMXn|HAUGHN0y##7q<^ zBL>0t7%_~<4k`DUZ@p(ufP1Ba`xa>7Yoeh8kVnW&zC5(yjKl4ic2 z8la_jdSX2@vVEZS8oU~wsEpmx(I$+5r+uUm_+Ln%Bm#epn`?eda%?Y`@!cC#jC)_^rQ%Tg%4f z{DwTfd0>_elh%P@VB?gP&+=nv4*z%B^)b#@2~;;gUk=LzD8UW4u7wLq&u*aT4Ms(5{fCJfL9RR z1a%hOAXLyHN11{OStbLllWmf_!G`stpKP07$U%Ya0?6t$I^W{%^gOiaGG~*=4QM&t zByDu=rmMgSx$b-2H(!)5uwS&rYAlbqr(^vb3yhOI#tO^RioW+T<1_-9Pyh2zWSIj) zCd~vg2^14ckquPzVL!~$;=-Xh?yQsSk)3^G3ih>QOp{fR7~?*BA7TZG>7V>8@q6YQ z7&d-~1IOAdY;Cln^bqtkDk2_w#er_}1z^D54SE#brrH|>-Fx5Ad1ERmd94z~Hg52DH(!AAfb`Zzdxt>A)N}3zCve z&^}g2W7W^&e_j^(e!8fe-n#@|Ow~T-vx&136QA8MMXx%4al}_>TvFC=zy324%Pv;$ z!26a0lLcre#iUXsnTkG5ib7FJ62l;xs^KQo;&}%`?)F7 z(XkRAqpil2AmohCmzKWj%z=LB3!g5^7BqwMy*4OG-Xm|6UK7W0TB26*Za`7|r;=;_ z+4Rzwz_qHSQ)L-g&-awf_REN_V{EP{zvIpO%=-d4Y%-?8#2!5F0fb&?C;dLOSb#nq z?ffP_Fc%|1C<$t@j{86a4BGpklZu7H=jSAG`(?;J%#OI{nWpF^kHae>&WiH|_FPd| zLuIs)u%kc0&QV?Z%_9e$jSEe~=j#4SR=qY0O{NL35-DaAMdGOFySyzb+!m)Pko$_V z#KTUAmMg+;yLI}uXeF!#k;MV*bfW!W&6!uds@#x`)3XjQ%r28_D!KQC?X+IX9N8r? zE;>o2bSgKw)=sw3P}pf|UmTX5#qbe)OKt9&I_%DORFGZddrl&V+V=GgtEux;n$K>X{xkYaUZ%>UJ zQldm9qh2~*-VK^pD}(y!Y)~y9nzzag6`;5KX49*g1oHbl>IrbCf1fJp<5to8=RzcP zF;MTu&#H#|9r4bjYfxCvTOg z>(y8&htV$dSCd7zpwmXFtOmL*x5=@~_8xArM;|v$v0KtdN=1V-c2Kr0I!vc@Qdlt^ z%BJtwVdd8+zjDVs((bTWu@F1gcSYVCuwK5#9ke8oo2rwgL{8;4W@G}b#Kg$uQ}&Y; zCZ@NG7p#68^zv9qDQB_2nBqh*LVqs6q zVR_oEz>npd?Re(a?;`VG^$m=U&Qo;A8$1Wy^v=kzmoF;hwt98byBB9b68?cCkpPRdS&;zc|4-<2P9f*P?45Jd z6c4@2e4xA`O^x>;PbkH&Q;nxVW>fY6BBxv3S{ZCvfTE2$THpK&>i;ot0)0>dVncF$ z+HDj|wF*CDJEfM6T7m?t}o0oF3*#J9mir+%vXMwU%a^MY}*xZiPGrk;`xW>-&&x-^{1EH14_SF#lueM>)rvb>K|{)ahnWuRgAu6s~hRZvSBAf zFO{QYv(V7##>QsMb_inypm8tm`7ktzHaaH#-|{_8Rygp6INik6*i13;6p2BR^loXL zWVxW(z1J_@8@m>5fC{U>hqG#>2Ee$pV)F~w8Iqu)H~t_tfk4l1|%_V7d9sLjy48s*_HhF)<&iikF7YNmGdgjhMv zqi>m#*R0IO^pK*14<-|UcVM$&AppPKS0~%9sP`K7$7UH&jZGisofABg4*NGprGCDQ zek23!E4r6=McGSk$qnaPqNW00RwByOdz}JIeOea$t~sUbm8zq&AW);DSM#tHwFV0= z@4Dw>q3$V}p{XNuQWWSn$XRMk`86x10st_V>@hF}0KnhN;--YBJ9of5>u>SmF*s=s zyo0g85zx|S6p6lAPPfkoMB1{|J3ZU^Sd*()am5V^rtz=@9OXLmU8V*)?2%(fs@W(| zx&*Zl^D112jgVNVxJ4ES(vJ(2#i1Bx-ve7x!+tbRQ7l{$pz}_W*ZZYSm65=w%Wk?y z&4-;|S~e78eD|mS;@|3e&3jM3CdR`;SL)S)%OY`yf(_2oL)JsrB6NB}IK&p! zJt;=iKnYW@MFq=IjS9QPYF3azE`~VtFLLRR-z?%n6fSqpDm$^%*n!K`A$k7J4nX_% z&%E1zVcg{2j)ATL7s|_aT0iSbR+*Lsho4$D^0z1b^;H5X-&6$Q#goIa5t-}U5+m7 zB^vyjvY_{M94P5DyTSgAz874k9*OQG*9Lpb zO|aKTF%Kxxjm5%)Zs{SKBV2>LTACt`qopqiu`Am}&N9J%u_m39OD6(35;cM|$(it& z$W^m<)7w2Z@j7Ky5orqjMHs`qqRI)0@YSWMah%1K%>Jm_q>LMz4U+tI-|#l z9`NHJqc9+;o!bqS$j$C?9+*)u{2Bo7CUC71fHRe*UZYPnoEK{8WsyzNdQPfGhhoqz zCnN_zwG49K+3?xOwd&%;OIJw;U7HoRxHq6sKxZCmbR^V`Hk{s?=3(=ojfIfvU3YX% zV9`MDvw$vzMFwLe6Vp>3{X4LKZROKsKky{)VCS4TZ2tbN>#D!JYcxfxxBS!hNbM`u zX6cGaME4xUoT11`D!RlU)joSbl)O1AotN|VYvPuuzhOI_RNocQ$hO2kS+p`BIjl?i z6pm#^#X+m<%&10wFEqGD$s_#kX20Alp>s-7qdBKoVs_(s73ETonsqC3Q zj#Mm6^(_}S(0E1On@82%G~R<(pLDKPUyzi{YUHns9AJ(@3$qFB-K>QFq}Ha1d=o6g zKUZ+0W-@-b1BU@E_~HHJnmCKTDL+p`ts&>3_qxbJ?k2ZxI*SHp+w~%+S{TEvc4_cz z@r~mQIb90G|DJ}GYRKu1OObfU>6zE2*@pwQ0os5#&V)c>%>+&sNY;GDM6Ox&)bG5x zm@J!Zn*;mdxhAG5g<>F9v4x60Dh66Ms17HB%GR(`x3Dj)N~H6y@WLq28rLrHgfT{7 zHz15LNvyCj#{Kp>^icJ-mpPTv0%LKPTPf!z^a-wAthwu+>3u+^X_IJ0+2W~ujE2Wx z<-@esGqX6ZWs5ylG{s>ZmM?$0V;5y~Sf05(J4=RM8!9E_qDLjC<0xh$Mb=T#2xEPJ z1ATG22@#tYm4Bs^*q)qho(5yF`R3H8wQr0-3 zRv^NT^KNqy3jIV zERzda$E2CVPROv#Qs)Z>>8v;R2{8-3*$vtvs}BUAB^h=)pvEFG=I#MYGZrrSFHc#*T}SXsJ}vvqE@ zON}4?Y>sM)YK^)u`xALF9k0S8|BDqp>MUbL`omBcbJ^b)apAy=i3M?ALpHihco_oo zT42*^X?++N>jzP@ASvjUd`huly%m|-A`i>RI5Q)&(OlT%HHm5Dm^dKYC?<&_o2h6l zBF$2F$kBi_OVNk;1R|+OgX)0fbZmIBd6LA&7X5|c92GQ2)|^cBGWw=JI5(|@q|PAw zO`d5!#X#UFi;6Dz+6~}_KrGqgEO{u=io1n5$~47U=lj9cF3HSQ@jcJSq)gZvRVsT- za+C)`Tch^*ubqKsa-3`DE)%R2;L&w6?D%D6^ORro>0X#Ln;o$H)f^+%;kYaJxw!va zWuCEfSdY{~DiL-bt|@`p0+hHZ4%PA7fj(cmc*^8%MnlRvw2r=GYF@KK>vZb9r9YM$ zEy$_&zIK&tX1BX_T)2}G6GNCuF%azCK}G-Nru=V@phe%t+1tFDAaC-(vs^YH-YVG@ z63?WIGrg~c$BQm;&byygKBUii?H4yXcZNO{pYwV!v6LH|Nw*GCQ)&7PzkSZ#w)f0S zcpTPpSSZTT(%S_(8fZm*@^a@!DaiK*T;_a`6LMj82guuiuru_B?%`F6Kv&+s73d^S z=yfh1*m1EN!NtVpeB^HX{bXfErsW-+{`FUVZ9zJak!+8xfB`JXQfZTv*;V3u%o(&6 z6l30O$AhduG5vdA`|z9QLY)>AE=Rgc88lFhx`$Bc0Da4xNBDPk{JTttw)W{y?FR8> zr4D2Gp^lr$e)qrrWfRBfa{T&a{|T~^oy+04e;IV!8=Z68LNN&xiG@xTN}arcU=%!0 zkRF2h_$Fw=hh*s~SYS-CZB3d59SpO(?Em(?X+Qjf`Bv6peGChG?5(p4XKtTe<2LdW z!=OoW?k9MV z)zmm{fcl4xdzQZHo0t<&Nr5$?yXir<>yU=SVv0QR4v7Zqzwpfi=R=BeA7p~rvkK=0 zAWa0iFAQRb1MHw%_hXH-%;=bC=ltT=B#E74;=r+xG84xnn_|EZ*HY2D-KrwW{Wrbc zxa6w1QFxLbAf7tmYxb^ibp;6^$k5MWHB-mGx#e z-K>gn(VQk3b6X;?wiK&}VMB(!ehVcn?+ida3;zGQzYaulnxOrj<|s5k$(_G;Q34N( zTrSCBovcyCcs+G53LSPj7FbBfxbGvSoJ+z4Uc%e)+$5+lOp>n{qod?pSq)0NgX?>})Fx>Z?uEk7^ts1fvGh16WUr@=wr7I`Ht2bA5wR&?Oa`-z zo^6)`|7squeNBia7Q)V$4MG*218>}#f0vi7f`W8ZIcT7<+mfBVTa!t_H&z4Yg*)wg z;6Hi}4$L+h5Z#->*U0wQW?NKcVp<9)24q&UsptXjJ3&R4RUGL%RiAn6>aoC46(avZlb6)s!`ei#iH>lyxM*kvKlQf z+7>)wH^|)o*B^dUZbVD+roc;N>kLwE@|bfd1~{bYR5S#{eT#)Ux;t1WN#r&wTA7sr zNG`|z0jMsL10oTz%z5E1&|J-(->i5*^8Bm4%fj$onQWQ8i`pjsgzffl^#fZr;A1TR zj_dS4u1@6&j6nHq*~~wX7zY-M%{M{t4vI;^U{7?bqSUies*_w7V?4OT{hTyOUgExb z#)a8exbdPi#naE-aa}VbO)(_f?`%I)UjVp0f7Y9s>$BQJxI3FMM|^^83c z8y2pa(HeC%d;og$;7w&q@Gg9{7hZj+IIF}vw+E#!+N-=t-xOB6w7=V|+8Ws!wK6D~ zlgDZ2Cs4!w_3j<=&d`P>nV1dnD3kSvO$McGfj9QbWZ~!}V9C)0{nVUV@35O=3pDz6 zmtEpzg1W#wzw@f)yiX;)t~xqtrUu_6%{2U*%+#@dx@_i}_284Zft8V&zRhEX#9X$^ zVUb}$h^vgQR_Qy0Hgm8luO38%23!xS6XtE0cLWqf-oM3NN&5I3V2_O67OJ~x_Q;At zhh$As4A17fr%E==s%A7b%1j1t%yrR+Q|X34lyw6I!7O!pNFR4;7Y7Lp_-P&T*zoO< zAmBr~g$7A3O*JG{3psdug;yard7h>VcHo$)!YjKtdt}`ZRjwty(lp=Lkfj=(_W{j%r#N5e-PS+9LZQ$hv)ruh}=)gYYwA$;!ENwss-3-7_RG;uVCwj)4oQ1R_ z>?UFGrcJ9!-gPI)2&ArW+}%s^9N3X=Fo8xD#Z*wFjEdGI`f6#+*$lXT;u`C*e-7wQ zqN%_pgP8y6rw6%DKbJ2Wlpk7jnlyrCc;MW}Ph6~phPRrOfQMlqH(2eRFF=K{vM>zk zXgcY`%G$s}pJAs~#RX{f5(mX+y3ll6Y2w%(4U+=F^PAZjp}+p}&<|eq1#oG$x-hJX zucHqt8~NQroh&&VvP3?+-4X)PGOQMyP_{@5!%96fC9B;(=HfXGitiPM;r0y`9tRg} z<7oW~LJz3h$+9^Eq}{Cv8Wo;Wr#N?zR#{F+J%DJbmPVmJ9jy!4;Ssw4lA9PQ!~>Yf z%$>h8JO>CF!%n-DE9RBUj(R0VK(9M|lcu;UN%U=$p>W_y=Y!5&(gSc|4!6|vnSUoA zZc%)qJWqFU;3GzUpQWHRm`kU+)y_{*0&l4Yl%KQ^Qqn@fR!+#_z}I8}eCw+}J{)Yc zTN%DtTS>74+bz(68>OaFLoo*_sI-f2=Ogg?BAXT1d1`=Y(9&QcA`xj~joeGZPJIw7 zUyv`z3DMFyA({uy9h}2~SVRxe<|Ghs$BOz*VXX|DI~JHEFO#+NJ7AsDTn!(Rm4(GJ zhXR`w8kD@O2cIfm&;xC-Cby8W2sN|ylb)NQiM?QDyr%D78~h*U^8c?%zQRIdfM#&* zn^f4|takB1HChxbm2HWrkmb@ShPzg9S3c~7|0atN>*tTr5zKgB0_aA`hwPUe8+3CMbS`^`v3R4Q}a@G{vW0M|rrj*W`if^}TmMu-my!2v=mx z1>C{8)w~+z1_l{na6L$A$Gi(gbb08-Jx;;37G)ZqjgYipxHSzPP5rJb4kT!|VttS{3SXfWf z6xg5-JU66or|G4VLF+IA6a4VYqs&A*3|Qf%lYlgIn`HI8l)z!9I>(I`T=>KNa`SAv z!^WO02mr+K1mW zZ{_=%I5lY|i(CT5#8PAf6`d2J3yo#kBo7uPGZ^W^RN7QyM~?$*05lbVG5-wDx3qnJ z#$~Gi2huep^A%%o)tOBEUW$PZXoZ;n$7a3QkT;ixxFED|Ff0pcif6%(g29GkHBMPr zdWfOBO}=2Hx&cP20<}!BgBougW=6l$(+ZtdwER? zd{+*Y2vENtHRPuY`1s9I{n>giJd5w@h&}5RCW(tl&c5Z_E#qPgjtJ%TFJ}&zQ!!qX zHOB&R;|9rvDk)7)NVVdu`l+~`&QfCzt&qZ!|Ne#8B4q!Z`@)U#4!>9rTN@Kg)5aQ1 zmIku%akp*XSzumm>9C%vg)&e>Lq+IVh5fAhIf>AmI%94h7ZpiYy^-zJCRq`rNe=H& zAfFgD=`@3`$xupii;H;&%^B!7pC`|zv*%wRS!zr&j9h2?NDFHeSXy^%i~?3BW!mRI zUlsE|tCp^v-^gtat%D3^mpENq;}7j~)|jI4pg=W(u`B{C<2DjNgIWVi}$*-!pwZFdYFD>gCf6(@xx$xA)UeWq%K zF?saXK*HpPDSET;`r2=p_ic69<;FrM);7tV+4XXLSqk~^v! z24rK>sOUynnGne_sHTpAe@zoV(-#9G7|PV_4Jr(&V5(h)oiL7C?Si}A{;;wzLjjcS z-4L^hp8#}>NHQD9#&0b{XVK5!duARraac2AAwu#oW2jF$FFXa6%xz4jM9*`<5(td2 z=F;os5UD=`xt|q+Ca)$htZ^<7WHL$eb6zz)6!>?iV9PrOeRw`=GLvyHECDswI>4-*YQL zVPl&x{&1{7G5zg((;k^O336CVVxjF121v_&`t&XdW^AU$;do&t#s|a;w@k?ac7X7x z`}P@g5A`*Pms#*y)QiENmKIV?X%u^=EJU*K2 zyS7+@=DVl-E|@p1e6g2<1Iq1SP^5fQT#!-J7dGTlDQl>`M(ZxdaNJkF; z#c;Sd)b#LQljQqIvG5Xu*V@Ghu?K}YNbFM`YuUuOuU4;KYTVE)sf`RFHyk)sGi7kf=6uFB`rD9qCLQNID)3eBBi@T;be5D&yi-gwu98?xT6H2UJ+v%A@ClC#a@!;j_ z<=f>2VvNNin*(k>qFyh@%}tde8{E|^;;i}01gKAy>QxD6w#zmB3lm;El%^;QX<RnE zjfEu6W1np9c`4T4VsG4Bt~Tf#>EX3TbxY5?U#E9RJ^0+2zz!fam3yZ0pHJV&26?>)zE{k@xU$9By*Mv;@x!KnE26=L@+rm5&dCW zQ3CH4w}p;z$)yu`ni$Yw?Epe8MpjZK3A`KfWde=iYLo}Z6K#@K4iL2q={v415eekL z8yom{Tr@XG5*V|C$}}jsL+t`IF|ajHU`n{%^ass~W`&_RFLyo`3!o+%$jjr)e9=de zKC$t-6#``N!VNcprMH8+CQb)p-43cdxLZ0XKNmPCuaIRjWnp>ZL07owIi}fSZsNwq zx;sa(#|m@P{#*K1n9{fm&e#5J6UiAXuIIRd2IQ#8(pW(;pt@LM5GrctV5Sz67qR^xS)nctY|rAsfM*GeJ=T42Ux< zh7qxG=Js~^y3B2FZz_SUY`-qhZ9dIxb|*c4Py5>OUdL6u_+ZGe-9D zI~5qPfdL=+j22Y}lM=KISUY;n4(A7qeuz31&)0`c?g_6?esFiDHs5hoa$7Z6nzSvUB!|TQM-2! zeEx#84re|hqs?K9C1621nyi$(*G&;ov?_pz&M>FHRlQW3Lm=Wl`g~KL@DnA30xTQ0!WYq)^)28Rfk0 zSv~aCIhnDyc$a59^xEckAgp#q=FC-dD<)b%-1#Mq0~osNa948nJ3^~x^7^?AhsY{! zJ*%FJcmm|BBm9^wirqkwbV`eP$E*B=coS0*S)!`($qE?sHwWN4>3raXzegAFx5#Se zY>sxdg>yWF9BrTD2f6sny&qrv{j*M~frU!girgfx3}{wiiT(nl`o4=y|2_*)37}PW zSAhZ8>m>*LEKfHGqW(4OhlgZzva7j#-XJMw^-Y)@FTsuyIo zEV0>(k=vy!Gq zihd@80oI`}jGvdq(QuC&T#g<*&f^!p*RHS^?Q@`wcSx;QSJ7vn-{!3|>=||s@2v~r zy(GuC0?6P#l`oDzrfQ4YA3DeshZx1%=#~Ec43;l8YnFyLgl&}`22F73I!7b^KQac^|w!$?DiHSsu{VQraeAqvDFR zCqkDnbrZ9i*(-N@t-gc5FM}fFm%shJ?!SKWo1gyiSC8J?NXKLJJKcNKv2_bP|K$$_ z!wKQcZyzJexGi-&*AxR+D3VHP6aVqDv_(@Fp3C1Da6TfL*Q2bA=x0g= z$Ks6QLZ5vTy$E*RRJa!+_D?_3_K!DTwF2Yq8cjVZ_2B6Idp1K|O|c+4Qb}nu!w>T> zPruFYmK(w@ijiz7gE4{>Mm=yKU6vN|be;64QF=()L0$(-DUI|&QZG0ozX^#w%Y9JG z2BmdJ$bhVN#&&Wb1UXSr>Ij)A4fKb~23{&KBVmuJ%PfxbfEa02p0;g2v}4&JR3p2}pl^ccMnt1Py}AL?!{g7BHn7qHh}(-uv2)1bcpxDS{Tc7GeJ(`M#M5+ zvmEtdZ$#{nB(b=bYtb}>=LZ}LXs7XtdckI%<+uwF=QN0(W_q4DH)iZSy0O0ck6-`$ zlFF)+_0l;2lBJXTNOx!zH4s}d>wVA?*&4EJay~@plH!h#a^#|AzY5(G}z1pWe z(KY9|IRqZNrkwTgXZOBkos?GmuKEu7fZL?x!Qt2&HbB2fvF9mrj?&&%bOvTHpGEDA z{xlR)+G(+%I&GGbM$&9!B&a_};>l#mmFTi5If6#-n=$u+5W7yWEn0U)S_-*?WXb*T zl_9DAM`DnS(gaR5{vu9J&4uZzo4y!1Og^et;Un?qlM{^^a~aHl@AX%8NpX5rQgmei zdfZ8Ise+ooGI2hX%3!_iwU}nrwU}PzYWQ(k;>I;TPMNK-{Y;*ipg0|H=7+Rm+N3i) zE53ery#E+k%ne^2On?iV3L`w>brhRQkt9lsdgNU(U}Q$1;(dxePLU&6Drcc1t`$QLQ8n~h(~RLJdaywT z`h8VpKzmd^t;6$1F`fsL12y@0Ducl|4VLz!vczquxY6~$(iw;=dQ3V(u}(7&8N-nJ z#k`vDgcUr}6}QUHPOV^cXXb7U#kwAYM2E#Y_!a3vW0}02gcBztGRnoAF&J}fnbBUi z-eZHf4wUGzO=D^J&af7h9tIo@2!>|56sg`i+_7wCyy(l`A^m}BV{^@l$6y8Y@^plclTz2)2D zzcWnNDcnA7dyHO%c6dTwwo~u+3FFH7RdK$7+pp7O*F`@&vHG-qnTmt@jzv>)qCW#k zoyu=s0DaXK&CYq}!eF`uDWy{Zjlu*`KQAvPS5g=@=zC=P1_{LEnT@g=QF$?gzQ-e= z2+4nU@Mhj2Xs0pIAGt3_J2tpH#t&y4IC%pnxOi^5E#to*e)Ct>S>~@PD_$YjJUCyp z%tpCmfMV}Zq!-f2pnQqa8dxuYbm^V6nU%JS#`MoYrZ>2sK{j=>w+e(T5CF0WaiHMm zkk{jju#|jP%vq=`S9Hg~|26xkNiK@_$C-eK0dh`y6%uoxKf2zp5ybU|yxN8N7HKQX zw+DPKqxUhAEefv6Oprho%hOSTcP(KAI{lMc6l)LPT9CD z^%QHQ$bn&(#Zr%XS_}xwU~;1C6p5liU)?R((G|f2oWWQYPpf$;K|4h|MY?P9B3_+h z>CAdwg<=DM1W)f$VuBL)^(qwEE1;pgF0fI46+}99CZGn?YlXOBJ9E)+_3a{apiMK@$_<{fkt;6N9r5$Yo z7g58*;~Y+gub#M1$2fU+|Ns5Y7p!*Whu2SUBKx^5oIDsUEjCu@J&Fa**u#|896P~~ zZ3a_dX7>09LN#@Q#iCkL!)C`FiD?0=0lY(2~0ROfylQqaR!GY0LC*8c4C`c$g2Ld^ zmzP9H#f949mtw5w>l*6ef=Ak$Xad=i3iUvJ8fo{Effm| za5JUF7%obXrHb%CU zpKzQrfKNIWV?YU~TPpiAlUOTAeDC$2CmWs{knFJmNg>5TiFYog#bVq(CJ7kXck)*I zorCyF4{aukL*@8p6$%d3#I&nERzXXDO-%XBGVzcXuAD0aj)o7&npLM^-<@z7D^kqZ z;WYN#E><6}(%E;ia;eev++ZYeASJ(DPyp!YP&UW33*YTv4$Qx!*zI@Qw*ZP*S~P{U zUVVY22caXoM^Y>q((VPjW^38?Vb!OPA$h~t-s;tm0ukVo*&?`vcwB^1J z709&ODDIWFXxf;)VMV-mI+TCC%`b^bmUPQq0jzNW=t&D?Il>Jqqh9^BVBS&tEYg!} z8V_CuJ4ku83oS&FOT#N5p@p=%J9#L>t3w^T)qXmR!va6IRR_6)3I+5~`I9Ws2bBY-YocgN zU-to0u$F>0?%F+c)Rz z;5f41<3Dcu_YChj{w+UK3+VrBxcEo=!tdwA@HnVnU&5?pHiY*@41YFjjN-fGs^*O5 zD4iS92VMVo_9*x*SdM9=i$W&GAdLa&!&r3dr&np|>-GvG9vfnDpc#T$Qa6(*?M zXp+Ng#mb*G{s#JJ=;mM~n#TVX`!~uh%@j#&8qr~Jv{lxp_%M9PtH!5URr69OgfCzs zXcv~wM_uhUvTW*^NFYAt1D{(KuUu*5W93mjuR1(S1_7}``iS?CSDmaAh%ZgdA$}(u zsfAJz46^FBz%b`Qa73@#5>z!8^1R4$4RuknHj?cJXRGESA0`y|#oms+4RlTpM|y1I zGkzC$_z91uCc6zzGlJ(D^e&CpcfVwX+1)RH<)`G}GbZbB#pXA0nqr|$>U~O!N=5lJ z68mg@)8xHdd?_R?4w%;W$6X*M=5IBDgT5J*Zo|vR!Zwm_`W+(;dl$%d0m_V^hnqni z47})y_0u=6wX{1Iml!QVOh8?vJi?B_VRa&VKKKZ66@!TbkEdNz}uV?jfV@T z!AdYE8oLo-L3mh-Q(~`hJIEUWDOm=yRarkTJ*Y$OY|Llke*tz-cd+aD7x1MYzkTmn zFRR0)e+HuuDicHbQ?opSxf%OeRCDC|cpY-lp~hUamE|r!4dj`M8>~kmhtf1aH_r+^p-t`^B7Y}}a^H=uG z|8)M=pS`7FR0VTQ2dYG8 z0P_pVig$~Lr1yQX#c(LBCUAmA->Cr~c8JWM;~Gb&Z{}t(&ifC{|4*wcGwHz(f=QSJg;#*}2A`BAO z%mZ_Q@#tQ8mJG8=&8k$O5=xypKn(O6HbJz`|8fxcAV^VVh*%V9AjyoF=x?{(T zO>Ver^#AUEziq``&fTMv$*0`PCOmj8zs3fo4=MH@MFuD>uH_+)qpoAulI%C~{i;>D zzPDv*vDhbK?kvbxXDdHe7Sd;@cSP$|sNdNQfuTNT$g9aeRnRR*x`s{^P(TKW3cxDF z0`OS<^sbm|@+01*6t-ZuL*HPXtdCzFq3ep#t5eDKSMd?0f_A@DL2JxiAh0n^Um*f~ z>(v9KYA&)#T>o0*H}C>nkXkM>0{=85)n@1#-bjK!<`}&*41YR16!o$p;IIci31ABz z{-RRK7@5Hg`VIeka6#M-V6G1?^4y*mKZB{(zChSx=S2s_z(umd0eLjG0A{KR)K`78 zLi6a&L6|__tjgynl05qE#AZRc*e0hZG#6aR$u?2P-d+*qqy`R3YbNuO;w+zf)#A{7 z5s4x!$?vASVzR?9F;D;zqHd;eCjOpo!N~N(Y=WUBxBDqvMq}i(;DQ(a)JJd{4bL_7 zy~;Pg|Lx%^j2iqpImvAb^W69r`NSqZbcteHDbhk||JK1&Kn>SUe;sTcf-SQv)hg>$ z=t{y$Wrs-xIZX7^bZGd|LxS#v_hd~%)On_*!_adq6IJ&>D329>S)WQRd9c!qC%X(U_2BeMQx-p z@(f!7d`E@!p0EL#?vt>C;kbju>kdOlWBr|s{LPGO+rh8pTNnQ@$9{_9l2i9!fH|1Z zP~Q#8j%5Z23ZlSpv7}f84i84%(f#;LW{jMGQsWWB!3i28?sNC%uXKSYW?F5}U;b*U zCfi1fKzpw31csuKS($wl3&qCelok`7{lq9XCp@v;2E|S=LvqUp(jmA$Y9lxNDgz2< zcEQRnkG>GzN#71mi#5=@AyWK-aJ{d&NhdAV2nW)p?4I7rK!Fnax(Kay>%($hbZ85> z4jd$Jc=*Y@Kii|FGcPlw{t3tVMGY)-;!(CKTCO zi+G)6w^C1X*rlY9ciR1Z36I`CMvs*_23_%V~%wfxcUZqG%vGjf=mxxQBCqz8D=LyYH0$eZzZpvQS~M9<+59rv5SAdSozW<) z3H&JHlw{AW4FxcoBg-m{rl2=%=roiWr{>~vO4uq0LcEh9_2B5hv-6h?TnrY zoRci0=QN5&A8fZB{o;T4yk;L~d9t6ygWI0UZH!AE#R5aCp3*K-KM?HY5BeUJte?8a z|LFVz{tij1AnkShJWZ#;>~l(z!*&W=G!^t(QVMdo_sH#O9rC^W^;7?UQ0S+>{)zU_ zKmE<`f320$tQVoE=$T(?nPZ)A49t&<$sTTT77qr{hc*CepxA>Hsim}^u{+g;VYem2 zpEyHhFl$LBdpiKZb6e7)ImhdzE8`x_%6w`4)GpY+<=acI2sbHC^0%|;>h7=>&DA-l z0D66~Et*Y1nc|lk4;7n&@VRT)&0!aK8(yvstsS=_kJIMn z(In;M#yAbptRMfW-QS9%@6>0XBkP{AHS;bTv}~nVXoA~V#bBY-6(u%t zBlVEpuMT2g7Siuc@ep!DZ5<4`r38HjF3f=IG;|yf%viDD_=&DLjo(?J~*S>*{Uvgy5tRnuhN;^W3@rtwq1t_KI~js+#%d<)nn0 zSL4C4T_B(zL0xr#VnIK559ky?1iwi>=ywIOI9S4vAToz^@5i^kR33r7F0C?@fq^s* zMu1Cs?dG-^jQtn;lv1S>MsY0!W@r6#%+F|nmQq}WFA}8+;I?E*xiSSb)tAT{Be~;! zR3N(kM=0qZD(IYv;5!=|6gZD8=NlZ`!}zU%#rA#QPlCjQW4R8R z)SFe|BsiWwsVy)4KgPyw8-jo#gKvZO<)>r~)5%>V54=~B1KPVu2c z5Nq_>K}M%D25&}#WTX!ToQGyk$TwCfOdyYe^`ux@oDfkAw23$AUZt*5v0H^1$zsU` zk{R5kKI^?z+Qx5D>>zi;8s$)r)ev42*`+|?IlT&LijRx-hHVD%%My`+zN3KpRL4C# zIPhb`nB&Qn%bBME&oRb|0t1hK%{q-$Ou7FP0;PN&>n?ZNOibq}wwWSLl(tTI)LdMg zBdEffI^-&9Ry9WU@v)w&4+@R3th5*!xx3}qUw1`{Gnr)uYJ^5jb9xaOIZI;)`3Hk9 zTFVgtN5JF{gwj|JH=JVF=63;zOLC%h^@0SzQE%LRDR4GI|CtbL@$%FrCUs^x1I6s< zvcnT_O4K#Uk^9>D*1rD*6g>?*kIp$p(P&gVw;0(SD&bwZ_L%PkeUj&X;spUR(z?(rbKBEEPLz^r|Ca$&x}K>gb|% z1F;!1cJTXOZ3r)s>QcR1#gzfYqU|wb$$m`~_8x)C$&Hk648+>!AUxJiZMDHi2F0$W zND8IJ93`eFk>s_I-YdtlC}Pnr`ysV2Fj1(`iJ zECK0-Wu~1M^LTR_CFH4iy%hI~Ir2+l1CDN=ZThzP$&&SH{&^&1;fn+%BU!Ait z;6Xr^?|R?e(o51$gi8Z1LD0+KlRKl2-z(@+;PD&@x;uyHr2#2HYyCKb#0eAO5*#`4 zJzOtwjH4~09Y^g6skzj7dT=Pn0k3u^jhuey@O}pZ11md!Gs5UzYe?`tRU}DVwV(_dUw)y%uD(rDYV2rjq1V;?;aoN94=se z?WXT{^w?L9Q%8SgonM|^-KU&ri<8hUyyJroY_ZRR1=T7uvGj{3riU>x3l`MQsSp>b zkA^~P0T_$Wq?dKrT!V%F*d4F{g~4%xsfb0#25a@2m*3pdXlh{fK{OCfin>fp-6@89~? znXfl}>E`@}3l}~ox=$+wJKv19LQnDEx@@w8+k)JKot_VDoSs^Wg|fjaN}I(?i+!jV zfQH`V5k-9L$*mXU^3fn+ey%G9<9roil~8L@BElo5M1?-}yka~ zK+vqIesru})vS3R@`PtJM@Z$|T}l-0Xx4N?R_~0)H8Y^o#>@Hmpd)+4>G;tv{CiA< z6+zeDyirbyJb1AM5@X1K7{#fA3c6;@ zdOV|yhpW*r%2$l(kYk5}^9Xs1-dT9z^*i%n35#J~ z3~ph3_aKO6q62#0w{e2XF~=|pV|uGxZ{juz9vj;)z7#Xlo?PE!<9iP1^3iG_;~Qqe zb0qjbb7}!|2q+-=AOLAc4}|UFZ;Eaq3z4cBR#1-uE(?)VPKU)`wJ#PzGZfHx@i1Cb zNcR&RUUq02^2=dV8H<9t=p+^m&%$v}Ib%S!Q}Yg2B->xl3E3}XR7bbjH(PovfE=_> z;>3X&BcN!YN&2oR!Bu`jd}3&4VD-$avv8raRDj$#4?^qFlN~GZ1BBwuht= z)9fzcHPA6R26;H#Ymq(lS}rh(4~FBGQQQF%Oarnb1WUIq=MCI=GJVSg#LGTs`Ae ztND1lMpI8pJve>>x#2v?GNEHAde}41^N2anhIO{x#9Q0qQw1E)H6wkl7I} zNDW_CK6CxlI=T)-=7y>Ck%78G{MI!dN_JcXPxuPZh!|-gr-mBzWZpOlg&z-+gRQY+ z%E1F&bogA<+#lLEKs_fYanLSh&O>7XO?r??b(mmD+yqWGBZkc>NQ}6CqBg&O=7SFV zZUc{vjX3Bmz>;y~#{_;&pd-`aE~FHXOivb6hgL{-kY3UV#oRdMnEU2BlwC2qrFrW8 z6CmQ_)Ub~>jktOIIn6FF$liSYclNHU$AZLxGkQgNS-O>9Mi)S!^nG4i)Hc5*y#CnD zVeP_-`MO=Nmr}d%W2Fyv@w>pSaBBQGqr<7Ej{hP~=%DyJzCCD<4hMtKDmx2g$FLqi zrN6Cyt+Mm;bQdEV`Q0=&F4sXd{-Ccxk`%Wo{&Tu1#{X}8^n~YP5IlErX#Af;yX|F5 zJ=S(O5F$nLD80IuE)Pv5ASQHLyhE{fR#H%pY@!UsBd|KVetmqy1Z{CL884jQ?)MG* zxGEP%)N{jtj>5w`6g4qNAUk*c)zWzvX633+)1ATxf-BNKMUiwf5Z(3jKnN`861mD8 z6f6NxCVj3k(hxu3`ygx!50BTxqzAbbZJf>E2(QJR82iG&;062S9T%v0u4N<+QhgRm zcqDVl0g>LOfMpsTBy(nhIv8{-*3P&XiFIvm#e&^Y;1Avd>4h|Rp0S+Gb@W>ICwSvReXZ_)$Py1u+ibBJ z`*ey;qewEPMNy(+i7_%g2s3djd~{e@bU}PUjHbbzgF4zcI6cs#-R#C=oYps!Ouzp9 z=J6B*c<}7vpctUtZ(oFoxknp9QboFCUj203+2arK9Zd6Kdapxydv>+Dmt@DG)`NS^ z#Umg%4X24Q7mxm~-);7pGs9{(mb^W64KaAI-Kep#8#^c#Fi=EkPkr&Sv}}F{a|YBx zS~REUHVTVFjN+C4{lHaV6u;g0qlWJ%y}svry4yedLoYBwWd-e)CWvm!ih^51j)6D2 zmp>5oAf!cefn>=BqKv^;eYZs)2yuhG84CtRz`?rtXy{zF!O5gh9|;HkX~^1=_T9FB zeu3Qg^4N-Pl4XOKH59v&BFiYPL2@UydfI*{(BJ(vHw?6MBlCH1aPBFtzlswaUU=iL z(^th>4aj%)mgJBM4>ll8HgGsZv2_%wp|l;!T*)5kZs587SlOpA1|`KI89^I^CGI9( zw^)}0Me$dZ7LE2MIaG~Ihe!wzV_?-t27_IN=JuRwNa}B58p2mIjgdNR1$ke*#vf(7 z65^}o=F_dvn}gZ^8xf6)RI((xA~u&G0}o07SpUNvjGu?1Ck@D$j&Oz0|aq98*EhX242j#jIw{idop7R~EOHzapervpxj(*3Wntoc2nyW-aU; zA&F3`yL3vkW}V~=l;@uI=RW#ehGiT`;qo<{ujPak>IYvt@i+TAf#)RF=Af$J$n;h= zS5oCy?uN5Y9Ke zI{B-EO6#Q5{L{(>B$Hd*#DkYTdu^t$B8r9J{Z>kgLXe7+@PQY+>R9$f0YrMO&c~)Tj?s)$(vX~n%JUI9Y8X_Zrv5sO>DUw8Kv9Y`%wwIuq zCDf?30W?<3-s%TDa1&o(;4CU!FwmSH@l$ox%&AsXbk2*qOg4G&dI!qkM~rg`#X?FZ zpVA^nZNlslx(YZ$(d@(TzgV;(4mj5=X_y6g0Zt@6-WWZD*t&R_4EwbI=`fU2ZIx{f zhS>(YeBIOS>FoGFnjl8m;54vKUd+i`om95(3k&QgmnYjgJ=ooFFoj?R0Qwm~Woymc zA>%<8I1zo!0X6o2yU&prWuUl#Wt7|9cE~YUy8i3v@vvd9!|-kzDez!3a@fXs44gpu74aRJC=3f-0Jw6rU3=m5<>9l2MF_{UMiu#xY1n z={ubiw(#cr&w78>07%z;KssfO3ZwX{Z?op~xBC8N_ZvyC*JyRA)WQ^M;Terle2zCj zii6Iu<-lbF1tZ>i^$o!h|Lr~|UNWy;xEES%uJX^pxiwStss;gGlc#PMuARODx;~BK zC6EC)1_>Vgbd4!;F?fK3JxGY&7(EzhKrw4fO0y^6UTqzX0h9WB|t)+|j0;Go4`0In3BTEHtgUC~d z(Fy!K^?rAqoPJ{_-@ZH8L0fHG)FMH4(AMaa5r)`Ih)5*!K9ent&tOnx@}ZaIH(3R; zw|Z3pzem|DKN!9|;1pdY>6w5^Atzq%BkMe3!(DIn(Q`Tdx;(e&xU@jaQK8%BXAw2f1ADxIuIH`c zSF-JXO}wQ6uqeUZ!{l@z9>@lAGyd?QOL3>^J2AQZ=IB0Po=*^U#w3XL^EUaoAl@-X z0FA)XsGE;~FxCxD0P@_Z1<$z4`|m49c%EG}W##0}LA&A(^EbSDEY& z{8*gjw}Zc1Sq3eqyOmi%Io|7L9AxeTN&W;(Ph-RSSP|^}eFtaVXMbt+PQrin(o5v( za}(ZQGHK+xsh?u|DAGe|6Gh8KH_Qv6rGf%F6(&_=hJ~n2t86=eok)i=h$c9nuP&Qk zDm(KABmjzd6^a3%2`d&Ai#mmk-fL&pDR3o)GFFR3O9eX=seI%IM?^Trw4X@(weO9>6cjMn&= zARt{6Xy&<%xXKK9A;T&#RAw;k!ZY-$IZ$Q;dTiou#`32CvwxO(g@zB(tIo_lM}tD+ z+L>VB13Z=yyYBZf5A0*YTq+nm*Y1+1M2oxSgA8(=ukZmX6j;EcrVCEb?LO@jL&x>l zGB|IM9^W$&Tk99D|G=IV#bb-09PpfULIME0loLhf9_2*Q!|;Tu6EMp>F+NTW@)LKt z{{SZ!>xJHvMfzvGt)q`Af+RtW&kjD8QJNStS@0f`1gyas6CtWAAA4{hZiS6EGDxv^DRLX5xTwR@A6gdFDXb39j_Zcx;XN8{ zhuIN1#P5_{k8T2P`}NU#!pyBeD2g%Yi;Sgs1Y%tvDi$wC(I(uVWd&T z!<>M^UXN5!zTc47!7yay!kcuMTrx3rG^`A96DaZNLwxft(z&(Dj6q#+L%DLm7tZ5X z;ZttM8likQ3+~se%zxx^|0H&~e=VsDSni)H03QX#S~CFM8O-IlldrT3EqKRk>tw*r z-^}w^kAxfedW;`55G(k~y>dPv%egtu9_+sWgYJlg!v=~?r$`#5MMYIJffs6AwhQY8 z1!{!FIZ`36=e06h`Bif#bc%D@h)u8^`&0axM5z@Sig&-*Oj4d3WNf!VhJj)?QzVnp z;;*|^R!B{zW7`vhJr0MnHRP+v|3@RD}?OU?Mk~7#UHSvxRqc|n#p`y^5LqOZzD#O@BicfE_JMA@> z5wRMN#~*GSB2FXXxmns9AJ%^5Pu9uipF2(w(#Opm@!WqINwZl~EuO?COd=0J`vF3) z*~%o)Ye(t?6Js=oMJrft6jK)R{@%18O-)Md!`S0;nF58wvle}&l3!JL9QErC?NfUR8SZJ9w57v z#i9&|&9=p3jqs3{B?ycqG#l6g;9PHu9|+W|4hCZw_##Q>VLTf6*QX%QHinC{~t$O96W(OY)`LUej{ow~t)Q@nZZ zh%5}%t$NV_rlgR*8MDG?B0JVNV#8HK|GCf6r~f!rW;G__TjJl6R1Xdcm)V$y9EydC zx-1Jf+@LRJn;%3WHyq@I(PwEZsdz0w&@#PVi9tJYXT>e>7HnnW(M3gab9Q0v> z@VSFF-V3<7>Rm5|4gGx;#`H>a{no^r*g^hLIw7JnurTmr@k6hNUIqMOzMiB6_0T(c z552JTw^6*6k9&sDYT>8uRY0>0`q&^fUch|*{qqX`MysrLM04UN)g;%0?MSVS9oa>( zU|32h?f-Nz6%cy?^N~BFU6{t#2W5uk1g_%g452xJ_X06n)Z~w=4qUfD1xyA5R|Kqz zs}Iy|VtRx1K`7jN6Lc9+mr$=l(MzZ`$uL(ixS;-T#2}0f1;h5m_H)Ez$GyqTu>AC^ zU;e2*&)hTB=jeb<4k;&V=itbjm_;I}h4QVMyV$Qgs>#2bw#bGdJ+d+EBq#@*`1gzR z-Q)y>2`plJNbL;W8vo7yT{AD3gBdHxB2jwK>A# zF=)YhzeXBM{IC?O7t|6*8aD0-vo%Go-Fu8y#|RsxS@MOye)=DI_Nuxbn?-P-pxbKp ze03GE?ZWL-D$I%Ji#~g%n`Q$<*hIgD21%_L>mJZ4-YXx!)b`_d{gkaSz~d=TbKNO! z@L2n^hr9V!qvRFpl}MIy3%_`9Xby6yBf>8k6uXupDU=o$&qcf(L0W8427#e(t&d8nue+bBCDFAFMDR81+3Es1Nn+EE{T@;4am-f zp9yc#G{MYL7*sm}jdk{8=OilZZ*%)TH&nE}+UsxM2kW4LbbE}6!33S|r20Mx1=h}) z$iU`^*XGQNA94AFZE||?o;yAN=!f2|_F81mNehSr<+4tY)Pr6hXx>qxs>6^MR>d`| ziYL1F1CZ(rwqcT)Fia1BQEwyL_y?;u%H%U5a&z)1YZ?zu`6bybn};a&K1Du*m_Bcd z`m!{MT_ie2*D2bC8PV9a-6=fBGcniXr^qeP18NtpeEqWY-43P>+?B!j4^%lmnZb?1 zdL9r4@k;r<^lrs~w8rNw&@S8#ejq??%#E@KA)Sg+{svh09-<#Au8L0u6nO7an%G^+ zciaYTZ1`u@^9t10Uu_h2 zi?hS_N{}HlS&|fudB~Iilj2W7PlN^hELs1n zy}=z({bcyz=p8wUv(NKqTzek>E{B%ppEnd8_=XjM*Jsz?B*!O{E}MnKM-+ROA|E2t zfl>Hr)VAoQ0q6Of7>KLWC8Dcy_W2blp`rwspYWU|?19WO7S(xvudgl{A`-RX z`=k}IpGMvHZJIyGzyF%64gc&v@6+*cjba@< zw`3`DSHkMaR%~fH-`0^u9vlEP*nmEbVv{Mdn$lK>U*#8oL`~9^gm_(94CV-~zxu!8 zp*aWE&x4lh&vOIoY2V~U_E`f5fq3MrH8Jb_F9+#L!pdUIjl+X<-n@D>CV2F!buV7@ z!4Uy&emW7-$EB3hKm75$1?R@W)`XAe`<%U(+cHJt~KtrdLAo=dRh z7(AH+``@+iL2~F4KT|%IqgJYQk?m2q*nz3DRE8uBx(e|%k|;7ru)v|wTqF)n5uXKo z7SI)y#_Wi1R8&Uv@pn(Z9eglsl?Dkv+-Z`=1tFJ!W%P%X_PxE#zFNv-qqz>mL;kX` zKDb7e3u{}n2-wteAmpaBS>7}Cl>8yAdQD8KPpS`8;UXzZt1LrsDqM$GUX9x6ZVKnI z5BuoD<*K8OEyrlaAj>f6-+Zm>nyN2-?KD~a+~g2}Fmpr>aTCREq+rangG`wiTRM;d z$Wk9;E{#G;pll_UT0yn$%v1bS!QF`}hv6P;aTQiR?>VOLR`uF{z7cKA=spmGqc<}1@ zvW*F8qS%uZIf4CS2!dh}bnZeoe;?Bc{^$EpyjBF+03$u%i=_;DRdZx(WG!h{eFWqx zjl$d+*uGmmx6`l0hlgqvdexxcsel@vCGw#tU7=(LKZD8PB}=X; zI|C|cLnuBGFK$+y3NY)S6w(;aH@7*$9{>k!PSfE=!t5q_T5Re36d*0C48XF@2ZCe)Jh+rb*S!u(+c9*6<(oOt8G!SeKzzo6 zQfdL_Nt)yy+Yf8V{`AZC0v69y?B+cit5r?0pt4j+X;J#*u9<%XEtYwF3!=M$*iI3c1MGG58?jea$P+p~+LI?LfE zATC%bZ}&?SeaJUXTj8Scwi^r-7dzzeTE}a!hRXN9XD?XJ#ZB_y>9EEISdZk*n}A-d zSdtKNSdb+vdf& z)5qV*$JEHGmzGS$3-G_3W-*7OWP8FTSh9VjOZJXuu_x?{@}4v}9_&6lDAY5z*;`su z^eUsUFtkxoNGD4+#k*&CpE?%Kj{Q>~;)aus?a1;O-+d5db%9EUzF$dlo*Ng)Xk$lq zQY?^{6$7Cby-Czb*U^cgoq-nZwr%td{`=xSW``sR8d3U*4%u5v!j}1-j%}T!y&SY8)Q7##DFg3h)`b^#crTTI;FL+w&Vn2ZrwmX2+$>yYtWi>l{qNDuukn9 z9eTej<%gqPjRA*4#}yn7ALQZ_zy8Ytp_4_vy>Xo~AmX16;)&RJv3;XE^zfvV$+@6b6>MgF%?tp|AZXnHpO^wb#SN66myPZn<>&nX^+xty$jTL z6n#viut#};*UFyu-UnroE&2S*QY0B` zQ-2n*|mM(Xa^hR=-`KuPdzniDBBWe@8Mr zxQZWC%110}^C@- zPTLByfm=a^`cDI~Rnt?L)ezioloz_WW6eq9F`9cYJVMNcJGkDh@c~WfusxErSkO6w zgx;X9xf<>xkR3PD7~{!#5qjQ?{61ia%6L)t7x&|YGeRu3Xh&F(?Gy`KK82L_7Ms?vN#d3??fy50vom+}{=^P?w}#iUNLLY>Q@TK-Jt6?D=^|>Gf0h_}~2pyvrTv z86#e7XCN121s}%=FVu~Viw@gUJv~!dJ_qzq7M8hU3F>SDTMt%AUJB_|Rtkr_Dg*SY zoM^N&72>zf#=P}bR}B2WJ0M~|x?_Tj{X^Wa?-(+Hr+$|3TdO(Q_Y*%kx#7X)WTj0F z(>;nEpvWCci$&K4`a>m#y*@Q}^uRhLRnQF-mV-bghovOt5$}Pwa?%^hMzH8@ASyZB zLLApD*effXnKk7kl(ba_ShUxgRoE3&EJ5KiRIA63ia}zoTv_9jD#$kIjI>d56fb-)ZPuI+ zEBb!7<8TVuIvTZu2bZSQ+e~m36btPp+bM1TE1iKTm14$BMc6t}$irec?9Cr~tva-O z%I@HOzAaW#_aQGNWX1Ew$aMMbSZpi$Oo8A-lGP?Tk|Eafnjn(vmWHD-*(}>h&b;Sp zv*j2`?>+qe75l;!E(s*hO^p!;B`vrF$Kp#Yj2uzkQWKaU;#PR;f1u4DsErB5z z_jAI}ys^v|w-Q8~!%WPs`M=Fz+L%?mgXYWMe($??=KsO%PAIz8Y7FOSO;AbSV~P0 z{lmzyauzBh@0o~g$54rVhriLji=NB;;=##g2mSSfbd|I}&hr1wnkwlDUJ|=0`l{xw z7R`+a+{pcx>+?T+bNTBnnkDk1bYpmp&nk%ert(WUyQ6xsbemT{-F248Sy+v*Hc#7h8REYG&WGQ!PY-$QG~|GkZcv&O zjm^#wvokR@foW4~Vl294t{XmQ-?x*)?Cj%QcAaDEG0H^qLcjN)B=#1?V?p7-MnLNr z^lzXfM8>PVWPd2Ev7xmY$cHLOg}J?BSHOF)&b}R3G2QSYoZQQAR8&COQ%xW&c9BdO z+lFw3i_zgk(V(whg(P(=d{$7$K_IV?UK*gg;g3&tS#H`%@I1{f*qw6?emH?=(wqMz zi+|QF3KqNV;vMlWgoy6Q?TI%iPSOv$W}vw|g%c=V@c-nauW782lg{XAS>%8R$L&6{ zaZ^uFESQ)EO8Z_MiiH^H#i3{=&@AZ6ru6YGEvG3##i9RxUBTMOPvy6K z+l7m$hhB+PP2m04@V-vP5;-)J`RcavOF^jfV`ZNL4`B$nPNAy}PYF6H-5I^$G*rG| zpbeX65s5=y20AM~1?Utlf5I4zj_XZ~oyCD&E+`tlmK&O;J-B`J?~9HhQPeBkKE0b> zxNsr%iPnepF$>#3l)f*vYVHMMVy^O=rA@zJE^?1Vt z@Sk2`#*41DWG(~NFd|&^`Ta|N*S}&lJV$n{pRqd-0Mq@P)AXY=4Zd4|gHR?<{O=K(VYmk^|gQ?XXt50dl=aBvm+b z(Dx|430!PoW)1Ip#Q-E`oNSTHfbyi_8RdxcDB@;;<}UF|nL0e5&5OEBHa#~5-p~d; zB1lt0vCveMPiZ#=_0sD@Hp*&z+DXE+3*<_i0bJl@$qlAO(<$r@L+7}kR~kFW$Mea& zQ~_SSamo#d2oCbE#0~mh2J(ip;?+D?o1VvGJWgyqevpf|-ktTy)n}bF7nwZJVBC(q z4ORx0OE<;qqxXclZ!jH>fpeqm@OHPI;ABB2O<(xV$H7($(*HyGGO}edsk8a%l~e2v zij-2?9%YZBfp;wI0&nTeE=ASc7R}kIy^tw3F~7my-ajl@n zr$?C6z znt`$hmq%;(=SaSvu6(ZEZ@{N^#>uG_Q0ah?oLsV2fb<7u{fSnYMe3@+T`aqe`|IeYhmm9HA@-IC`79tIoR^N9JsRaX0=UWvIKAW5}s0>{tU zm^gdx~|2X?^9~+a&!ND{(5f8TmYlg&Gw-rOk=P+Ba}q zo8+ys0rG)g7j0P^>vE!z*m8|eO(29*n39-vp}26~%0C)<2o}UkLh|Ohilu3^_iaAL z-9QYxOAa6Dk`s=-ra%0kS3iH)5o!4DPEzf`j>vf%NW4$6$0>4z(qeLFWoX6pR@wen zS~N(vW}&Q16PS37@|B@yp^>RxZ1g@6m>qb_w{}L02G3S1(q~}8@%%hgCTcO$ zw}E7h!g6OHtccr%>wR}i>rl;gCLV3kRJ^n@v{`daxia*OX2~nrp*LgGgK(xQS7P7% zm|gL)LuVL5PyZY%jvPa0l!PPo!v~A65-Xg_y>dPv%b%NYO1=&FHc)IjMbZF#Kz6we zB5XOzJkZfX)y|`mHG*C`A!1Wdb8K8=AWL~G;@jY%YXrMhCDBPN zUIv7cgAf``V$H<-#gYr8o|n#V@ZUV8dM2m1=|5}xh2ql z_{zdd_StDJKCkEg%ZP*Om?n8UeF=EDGx#+?L3%I@87L8~K!Q8bZm;v$b{?0mxboUx zD!={r#o4`sPUoitp@R#StrSE(F_v2%no3qPm|~jPxyR|n&_pkX&M3Bs1FVzJ4^O>t zhGb19`)nqkVu~%KNFJpSV86uW{VODXNy7#&90bXL;+iS{?l ze9lPA*;&^DfD9 zw1&Sw^cbj7l}!Qqu6F1GY~$xh`a=hxn&g`DT+oo0MbfSgWSp@c31YC>>XT}yXdx}^ zN3gKcS-6J^+I#Q{#zE8`8TSim z3%x$nIzdMv*pvx(QgJYyLy(ih?!wEyeecb}SFEOI z&AX#M;-RVo*0T(Gp{&P1R9n<;ZJs9!bH-g9-}@@)+5 zCGFw{UOtVHH65-Pk;ZYzYrwY&1X1>fn%jL(utQOW^nKs6Q!6G&0BTIH)@3iy@q@UT zc8?WsR?PW!vpuh*-sXpyO0h{4Sw(5hsyLmn!iD)Av+W27SN@0`ziYcQ&OyTNrU}@6 zP9RZuY{vgpTBn@mpH?m)nUl$08%PvUYykzLaoPk?UvQr9Hd^o3J1-BahcAR*;BA!U zhqh=8K3P*6uXHc8!7GKf4D4tr-NA>(euV-Y1#Y^$orS`R!`j+&;$iW z38ItIQ=)oat)PHkN_Gg2(dBNed>=W^P65Q$mW{mKZQGmxLfs4av%p?o&SPx`B%ntu zQXXl^b-T@-?b$(34QV}Oq>nC$ml1Sn4M zc_Hbm1FzeczIrTt92Bz_gU(YawO4jH5aXxFhN3G8D~o}eq@Wy-B-C91osA4;kD>?a z*W8*p8aaa8;W6@V_ib~6$0WtgC2IxN>1SroB?*wt@E*h2SVsoN9Ww^A%gRQdOxGuPjgX8QxuIwkCvFy7GSHG+OajN}% z^qgde9n4-P(l5?@$tbP|iKuh`bXoeB4yKDatvU4;u-EZV4PC1+23k*zEhulj;wpe%I@1j(j?50py@~uCG$Rwnuv(u7_fjA#T~rV!J3}g4zS*P)6deCWLB1i8tgTZzk%;;jfPLmutlp4KtX_6%o_=w6mjowFRq1x@; z&`Nkj7i|<`;U>~qXE596lt(}n3Ra(BpZqEf?sn6g0xIV`zHAk553>iz&~-@Ee#0MG z5pyK!}} zUTOE+7jcm34Q>u-4p{wi0sk0XGik!wfnim zsdUFEjsD)b9pyAio|_!H@cK=ukJVQFMR)d(WW_V)`a+U(g#Fq`u^ALuOKB}tOX$*K zuNPFJNNSl!@ z%gouB1_55ZRchiT^9p_Rs(tajVmCvC3n*+S6;}+s<7YSq#VGy`b$n8Mj=gI8lM5pc zu5!z^0Z$6WuA#_EN{gN8x}}0*k%7J}wPcRmjEB)j#2J)~ey4knI!1}|KbP(JhShXj zpIv{G9OqUM>ACYI0rz;seDe{-o~6i#77>^3fb%TygaR2Ls{2$+lQeo2YGSr%>SR~> z$V;kM4?u+tUWYOb2_ikIqVT#JpL<_e&#$8LqtAt{=eG-w(P@(7Q%YaXo0kB-?)6s> z3+m`0dG)ko^kG43WO+o3rXd_#WAn&i^~I^3!c(MCo*Y>tegG=pNKBc->rs@H-44Su!s0bp~~|HfbDt9DsU$`G>NKg}iVHk*RWDZEPuSo-8X6}@`O)GfslUZhCM(R5*czI&ki7L6{XJGY#*ikpgZ$i|6*nlEXHTs-##D_1;Nom%h9% za`$wk71bpLWiXb$o@7Zc-3e&`%nG2GT%&hQ;DacrcPSB-D-VS2f`lPle+=a4kP~nh zaO||oELG;{cyvWI$p>V8PR<&r#rYt4h zbk2+;zDwjoQLe;??U0TWEw&GH;oia{14?`4IWDCL9&BbD=+xy17Kw@=$7Jp?-N0KU zDxjAU)Z)e1JWBCmDV4>Dm{q^mMYcyB7xseQLrtJr3kGFmkT!BX|0=A8hP*7~fv$(n zI3eQ((_+Ehj(1i|p7nA$By8*Pu2U>wb=WM3eWZyZ>}S)fkmGHFgRxGwlA1@4Z`nTEFeq~+s{5568E94rtMUMvu z^_SV`zYkFC9g6f)+P`3TEQoDeb<(FCmg=hJmWj7Y4}|DYMK@1f&pQ}a2TD+lzV(7W zMFUjE>>xEzYWvp<;e&qNQG>Em0V^mARUsY$9l8YnhN(kd?<|Qv56mm+L4~wQ)kVKA z*5LuPw(pA@z4fX#;Kd&D`r9@7rc~FVToQj%dVbzYa~bdo3ZI}egkIw-TJfPde_G|g}0l|v;B22X{4ccv95fBCDanr!1XdwH-4IbwsV zeH6QgBIT4ebH;t)d-A1|EuYIKH*4-F8WnrUU2;{^GkNj%ivFSGpU!}w@}aL>mLf6S z4xoO^r&D3sd`Hp8?C{&+OjWxOgNU9GY`LtAM`nYQNj!RDlkA#4wg$Ec#!-k8|-DS0cm7XJy4O-xXVD zn3bEZ{DLGsW95-0HuFd}#b!}t1Et;R)1~g6k|63f_W&VbZU)mkWtD6f$Wjh@)%cXq zRj|lF;;t5U;$m~onUN9mf|H{--pd@Dk*>{Ooip!eokFO_3Fl zzy(d~4CXW0(s9bTPNCQf75^9?lY3(m^6|3CKL1+J;|%pdm% zCnPV1+z2FR07W7Q=*TUMh>LS+yVFj$mu+q+(dGp`QapSOMOBMiq`)5#QJqxStF9}vkYN?fT?@8KtwNzrz zan2Fn5Ay|D*-G!h=#5@#NddQqKF*xyHu_eCYhvPk`XjH1dqNu}Pp*~FDIxJb`08mI z#WZjDl!|SPXqy4r#yzR*0KL2QPlwI-F&3)Is_2y-bs-l#4YeJruJ8kp`1ya)@Da9! z)mQJhq3>4LfBDTvMyKTb4-+EDJqHe-tu_hM$4_P!PbNcD>=Ws96p_T>xuL|UMcSY| z>`@ioMPGnoBa{#;g$6*pr(Kd8mcqfV#H%25oe9LJ2IU7VW4a>7KExL3p!=0D6dP)W zx_*e7!y?)PT15MTvL%`Fn;u=EDV*hkoB9faCVf|2nd`9ACckab`@PnK6Wss04mIk8 zIzI9Np_^>zjKqvmp;80tE?AsbHw_yT>s;|Zt-PE&?1ZdRhR@moPIjIM8}Un(RdWqL z!$f$G0n#JKe#uM&!*-4Qq zDmKx1VUBlCz}d-LJ<4ZhDSz9+N0K^CO!>_FZru@SyeoVif4A=)*9O@&w=GevF-N|2 z_ou1vq<`=8_b@Wt%0ZR)boq@?kc*aeNAv}qBu$*dLHp+}7i5NA7I#abxl*xD*cxuT zmC$yJpx)sdVN1t%`0n*z^q7+(vkApGuA^s3&f_yE=?6}_z{lR%K+?L4``G(2uQ069 zTN88l>DB7U0@0wGR%UqBb}0C&xf%mPzG^s|y|yeig70p$F2B4;7LxQ=#=3y+{D_QF z0mVS{E{BTEc<-`!IZqRFaKYyyP0R(d50b+B_<7=FCXH9^l?q5{2rA$%i&^J(h? zD}(TNB2Y30jMz4sZvj<179`y>59mJMnE3LAZS@Tym0s1ps#WeMm8wckJEP&OWYyhe z;TvOl*~*D$IqLXtx9q5RM&|s3(r7l0#eTAe>|kd$92hLkCT8OR#q6iZJ}MSt{CdD> z<-6T9vTMO8v!KG_eItEFt3s=Q?pBvRs82!h<=as0zmIQVj6yybtkm4|`moajd8%JW z*fMaTp?u|DM3rnnf*h#zbR`sfA92BIUNxPoOmk@$)=g`N!7`_e)A(2kFUzqei?56i zV&2{2nouK%3j7M!k_u4FQKG?NU^(D+wH6>RfNXHo$16WqTeDYL;qZ zHcxG2Zbu|~+>U4nKq2F;9>elYl5@f~PJLjuZ4ACWzWMQe=Z2>(|FB1i5<9LF_~(-s z=)W(;x|5<;df^{mxRa z9KO-CyxF$yAX({5N=!^kHpQ%?NCp*K@+R&r@5nFsw9ZWlS;A|gvjgtZ2_8szvXy8_ zv+OSK0co8(fnO|(Zxu}Q_`WGNPro8&!@uCOW4|(jW2u&g_Wn+ z2G}^GR8ip82kkpmk;$IW7|z@PhU5Ug3FeiPJn+2@=N-VRjTHzDZtAJ3BBh4W0iTx| zRHHih_!d~)k`mJHiRA5OaApTNiw=PeUN8kDE66!+EP3jg+DYa@GtY0K9CxL)q>3}h ztr7RqS%GVPhMcPCo%4pAa8!yEvfa$EXo`Swo{bCG<9zS0dfYOwvl2h(GY{HXP%>)~Vx{YVxP!k!o-ZBT!c({1zqfSRsJP(7`F+#WNUrOxr1^WxwT*Q-=%+T3}5;S$F#)_mcR6u%J1D% z;QaaQC*IA~(j-#$%Iv<6nykZnCOsSfe9Se zP)r&{lBw7#U|QBdda^}v+AVVmG$ADcmw&Hx4d~3+b1fP5tG9NZM}63i7g^c5NrU&N z-206Y6+yp!?KN_3vB*6PArjH_bs95CTz@}R$@x~UiQ$d6B8f7j>Td5ua zhr0@2R?~y-U9^USMVr_k-$i3IPIZoeq7fe9Gsqu=(Yid5s!9e#QbAwJ`yy|8WX>|Q z(xY_Iuv3263fL^5Rs<4e;yroLqJ-&m{68kXTfn==>rfFYGF%;YN^;dIa2(ZDso|hZ zR3=ztAV6zXG>72ZZ}TjoYv(NkA~e?a%9tz{1EGehWdL;`9n#em3!nuDmX#)_=`D>^AulHrQU+WG(KWi7~np=Sq)0Z9m#ngQnvf_2{dwDyfwt5`rto8X) zdXrR}(TL@0e2iI=A>aA=b<{F0d>x`<8) zLSkt!1KYTm>Bh}{hF=8+6#C~BOur<)BnCbD6u7DjH>4EW?Ok_19-LS|>+J8$^9qO zUlX!!=Fz}x$sknIB{SEi;QudtbBlOQh#W4Z5H7!J6@lAt!xyiOzalfhuNjg|P~> zUD6*A=L78$@&mjE*aR-3)5{4ek?p{*URtt4}T~z8V=!I;qOSAGpR87edbdPe7D(DY_p^QtTm{i z<|+)j-0+QJ^)t$dsnR{GeLLv52;94(>_UDRUcr2i{l9K&zS~hOxLvyn%-#mgv7Jk@jCDh1~@xL_$(O|lTMKoD%Q{kV2~+7UY|7H60*??vzgB) zBJGc&Op1`-{O-S1|MT76$v!|$#emb2XfU8)0t^}?L6t3a@x!nqZA5D0Kp zc$(n{EIx|T8SVaYu8CA4b$-5~w`9-_ukD|HKlBj6J#U6znf`Tku$oo}-wMp5AJXe) zE&*Q!xDc<;QA4~RAIGYVCFB}6l~W3#J9Tg$^nK=lbBtt9w`S*r;qCYdsz8QZP4{t= z85IVSv1_H*YZXS(8e~bHpl8kFj^Me-!S<_WODJ6geKE@MY z!hqW-nXXC9GM`m1d)s8Yq$V&P*2>$VRZ6VRcnIN=wLU$+&%V!*ZELdBve zcvfKDYd3`&?|YFovOUt$Du&ua8L({ zn|g#)1a5#0fZ^$QUK8Y__PR9sHge*n_Q0A=%#Izf_1M*?^JS0B8Tx0>`sidTEZtKF zgE&1fgQ2&9wSEue13nm&-!Nk*sG={geqqr)QAI@9fZtf;OzvMXZ9?63147|F^)ARsQJ%ToqbDSaw+8zYG@yDU}K}WoU z?hA#cX?Y5#U9T^3G^`N{LLWjRD20>gs?u@}1gSds$Dq6p^Up)FwLCSw)V+UBmJ+3_ z7VhV4_FtH*K=d5}E(lfJEQ}EYeYK+3tKVy@#0+7ZwXr z{mMj{@?<9ehbg?5&~?7>N!?WQW=V(5DOqTLYz?s)HDWuGRkxwWUmyn6E$2?<*$LebzVaZFG00nXu*~7Upn`lSH)3@`AHku>E zVXcUTIlm%L;iW*O6il+d`ciK{*19)Cd>V{NnE;K*o-GIV&47 zIX2}qXNqwL)&2g(zmgR%Suvj4WS^8pF@WziRIH(n0o{=OoH|%k@c*y^iHe}!7ReLa zC0(*KmkEH$Q%k|hMm&AW{s)%v^2r-y;XfO(Azu~nEpm$8?%IJpV`#Y=vAe!ZF&8Lu z4ui@?UQbeF*XO8?id&@pa}EmXT>DAA5F4|wDHuctl9^0-tKwcHlKyRWuY`I{6-uw@ zo3>jywcyU6W4qJ)8h5e4ka_@9b$t%h52Jb#I;QE7_4HYfPv@cq%ka}*#!Gds2-yL* z!{97o9QouoF+^Ptnr&(tg=xEJs1wEUGvyG?hU!xEtZtaH66Z1ip6SNR%fomB%QnRx_HT!Uy$3PeN@FREr2Of7c6(qVSDfsn5+ zGFM@EV~mU3gk0m5uuD<#Qv48xG$`nQ!npYUD7J)+lR8=#@Z`B$fjB&5wcs+uMtn%LB zqu)BEc;#mC6C)~zKk3<*mbKe%Oe(#Dmho>968EEl_5txo;Fum;4R6!O?0!}#QoUiW7R$!n zbzH;~Xe1c1&t6M0YbcUN#bP}ZRP;hGp}q_ieUF>3t!D~Ig>Xk;6?7X-EHtbJ#sn8X zq@MfTcg?rQ&nF;XzD@A{pneE?VBlhj;H=N~*Hp<-rO;uD3bu*16|=J9w`=v}TQQa` zZ?U}S{Z1V9mr2HTM7ep_pU7h;hpj*c8q|>%AemxTQe-(5i$$5}3Ed9GbnZmtXhxfx z7tj4@Puc&P<+&ft;W-uN5@5bJVY7ZXumQ2KlQHmMVk|X2EXBJYN^<{i)XM=5UJRfK zUNVf>_!N-S8Nfl|AZ05p_FoS~)z*vvbs!*Y# zZ3m4P%KQvWC#@5nmC3DB~fsx302`;P&(yV<%uum(MJ_M zv~Jq4(`Q2bybX}p(E>?td>9n@XUf-!PDAw)3Pa-?;N?6TRdxn}w%72$OIXP?%8E6) zZf5jZff0ZoR89F4NpfHS0+q&yJ5S?VC7_XtsuPg6t-{e!9uMCnonwjqo`3-nTE2)vrvTWV;ENHc?C- z1@eQjIewS?tAy(V4$xJB$jey+J7FlCW3aArlcdBQ$|0dU^1>O8O1=6OCc5erR*U+B-`!M z!AGuEEJrY;c!r(yYaSolcI#Y`R}MqT$p5N_Z1V$IwJ$ctVaXKM!`Ul%Y;cgQ?RYft z-kzM;4>Z55?)>!(BUt)>CS680u>*?(hh7^@z_Np4pc=QFifw)UE_2*1Kd6Hb|MCXN z9Wvy9-0k>`At&5+XNWEXukR7xHc%G7K(z7>zJcZ`kz5k(jJhOWP0AzB1L;$NsEw?3 zyP>?n$5I8P7`B7O8JmT90-7g(Z4aK5y#J$p-3Xf6;6E)VTiAidfi29ZCZO3%F<^3P zFfMyYoa?uI>h?gWNU9IPo#0`pK1$~z540*hvWtE!PUmXn=tsv(bKO;Fcu<|9R&R3J zC8c86M&5C$^t$M(!dv^qkHtk`kPH->YFZ!d!vbD)aE704ERk){^8`i4zcv8Vo-g8l zG#>Ogu*tE2o=OnJ{kT)EiAiMcOP2z9oKCjKZ^OK`ZugaGyc%(VD2>w=xR#sAyDBW4 ze<}R5|2@gs$sMkZ;XMI~K}!X#G54sQKfd*T=leI_zxe)RIKKG)<@a~~xbyvif9ZN( z_kR2PXJRsWkL3FzhRrF($39)QM$gzkiRqJB?#yq^y_@Gl{|;iAjjg_B=^?r2X~su*TovH5}L^?xD}m+OE7LPXh^`R?fZ1dVXto7k_|c zakTQwQK#KCF?sZ7io4KDx!j|ke^q{$^SSS8TlNu*6IQlgpp58z?U%p)>v&jE9oPn0 zU`5U6UE_A}i|7oJE$u-8Dw#=gG&r^V*|X{*8UhrduM+7YcD&gccZUX-Yq@iqH6T*3fCXx z(TjoM5M#vlnN7=6Vf8#)9>OlKSYB>Mi|FR^zrNYLKG#ASY>~2DvV|UsI7M{8d9v8I zhu%H?+Faa~c7?V`%k`4Zr2@3Z_xPthfUpMyqf`C9^zM>vlAv5oso+3h4+!Oy12;~t z;>^4@I>|Lx*)AD$!{Z(xRw?DAi14^wqVXu?;WPLShI?D2I5Gx8w2E!qiy$?P#VGh5 zrCv!>bl{#0x}n^3sX!-*55qQ&c1dmcCMA9XL&|#4?JBf=q?24pixkJfv2n~N;?{WQ zYWGKC;O+q$p(k*%dYa7)z1>cYpkX)74qIoiFQ<9GIX%oPk^{Fu9D~p+QkMGH1fB=i zc!jjX9XNzmNK+=hjl|le$}VQ=s|KB~+oAo1F@W>vc7=podSwDQk4yqMw<+c(MY^fj z!RQQbLQoN%eYJ0Yb7IW3x7b)A z2i}N*IL?SHXeq@MQ)Dw0ixusI^wsG*yoNyj38{WlM323+I#U84NykG&h{0ZA4|H*J?_qt_sJg+@&k%9h{x6O>`2k)O)GmFmM0#s>q{sgY1b9Kh9sg#mC2tDvKl9 z=7p69+UA&V>KwM|&BAW8pJ0>KCglbN_SoWHu+yvim)UQZLbyi_?47FpK^f46e+af~ zsA3BCrAT>ZUPX9ZgpPj>w2+pNe!35Gr8~VLw}q{Z>5*lEYuqKIkH2#6iCG=|W|szD z5o|lLnR3xpD4A5#%^`Ski{KI`b4poo8Mx<|ia#XF@T=oN5gM-t_Ml0wEz(QAr-Gi+ z1LOm7?N0B3Y1z;lUPBuAP=J}w!%v#PL-s37{NrMK19rI=#T7o8yi-B=U(+r*>r*8M zJ}kZQWRMakk9JrGiItt2RD8liYu^9hu=umk@Q-ngH6hyQq##u_eHP%MmDf#+6X^Rd z?Jbb;oE5O%&^-63T`#bMMcbTN|6xvi|B86o7Wn&5iT1kn%o%dJBwp{mEFzwlDMt>F z4K5ozt^(WDkkfX^$tQSgfpQP~d-n(R%z11ltu8YdSP!UY&X3s;b~)Yfvr3-1<#&FNLD)VUcAI4 z5|d3a>nM^z#jf}6lp+gs1KlGZ0M_U89G%ZVv#gsOpOF4mHcru;V zf#S*C6Ze-TDE{YP4w*X`4hsqkE<=m75xSTDoWwv#4GTHrB5=!zgg&+xp?KjuS_ca> z6fYdgrZ=#H#iWc|MNZ}k0EdNzg^>PseQi}QeG9mta{XFEI-rrMD|}<%8R?d&Bfe)m z%XtaThj@*?cwWO@JN-b=GUwBD1FenTH+A9TpS8!}#cRF`?Ppu8$I7!^$A}}F;$!4s ztgu4;Tk(4t=2)>XB|XwXITB2w3f_UhI^JEkUiuv9ZDNU9u3vv7wgfE^ta3l07z7@4 zlt3!tZcuE36x?=?d(*J&x=OzBX(__Uk8G>saZQP>KEx>Fu$mIbEoGDc>s)VhikRoi zfgQKw*b)&80^7)%nkk3!-{4%od>5$0lNyw$6W_L<5g7$6))q2M_$p)>#WJ_kC(Qkj z=Yiq4z$O-a6C^={f`hzSAQ^zP-%AAfA_&&2>0YuYV4{okSQxj_P-VAkmSM5^YZtGY zbJIDjA7X(t52;}=Rh;UF+geD9$nHSZA3CZjB7_9yXi+l)kN5fCku><{2zx-OWI3;a zPW5Y5v;xUyvs)vtZraAE;;{ZmJDCy|;27b9j6AdtJ}iS{wEE32|FbvToG|32vLhDQ zO|HV4g7r>=?zxI3kn2x$9d=qG*z@`j{F^Tta%$mk^6Q}c1CnRj2_BxsUx%d7l!fT?c2^3ir@W^k8(Z;>*zFr~`a; z@FTa?F3r*ovSv!UeD|!Iyg_Jc>o-1ELbrIQzg`t}VYW_^#=GNH[|_Rz#&TR;cD zaQ@9%SLkZl0JJ@J5RAha-Yv>r%T;vn@hSYYCKzT3vym3b z=EoxeYZ+#v7$)i)wXKIKV`%oV`{R>j$breuHkgEFQz&K?MG~pl<78`ir&m(6Mut_9 ztK9!D(VbZVh>`O&VaKchG3h(ueY<~W1jO#2yGzNU(m8J+;lyXoFpo(*OGuDW!^@&kc zNeh++?$dbRjWp1L>F-q2Sd*vYcggVKB03?cj#0JnA3|ly(98>*D&dHG4TpJj-L$%C z@M#z<#e)yL&*Q$dkE;U(Fm>>t8`Q{58<03`0n4Z5W!L3K^sP$|I!D&BL!See&Q+Tj zuVRV;Qr~L|>Lv4Y2u`d!$39_eEMPkA2 zu^Av%ZsK>?_j`Rhx>XbZ^zMl}u)Lam@$GM&Gq)WMyD7I|4f^EkW*!Z^Du)JYXfDBK z56#z#zm_&{E!6s#aR<50QvB`Te61=_6LZB2b&%?S^{zbf!rPcvihF@h(YR;HDo~7S z561o0UpatfP>kmDw*5m&{~N|drv2UP#iaU`S!7O|EHZ~Erja7`RO~r2;0C?>&Zk6c zef9>Pog61fncd6V8@SB*vb;)pge-T~f+8)_CqhTRq;?)gcoLm+70aFR8Yt5`YhrYe zgKC$zdh~H|qlMR-p>FpMG-lBIU+>*VPE4Kda9Ekt8}3=-L_4#fk-fM;->jtV9a`KkjAY1yjooOHUUK zyVu9O_jOf{c@6FJ8yg4S=UULJIpy8V@8BO?P(kMfHqh%tJG^xKE8+rfO7vxZ!$cYp zJua-C{THCf)-PMOBsV7SiYG>IV~11z3Gmr0-#m)f>%dl{$i!({M===`NvC25+?x3* zkeJjchaz&_+XCYvkd(xz)z=DPwp~(t8wr!?wHh;EEKDA>^GU~f_ek>s+E-*fu~2S{ zxu7qEtK@h1z@wQd?}B3co>`lii5IuAnnudb5O{w0=#PK7+MM9fVS^zS$P5#x;kTDl z>7sp8uf;5xQsA~iZeZ@Ol-+_3@~y6wUhQ6|=|h4F;gC}iz1By^kB1OyiTmEb{nNX> zGN-rAKBcS{YZ*IQ2{txigb!gne16(&b%oW^Kz-M%_=Gtjyu*UU0>%5;5S;|`FlE3q z)&*%meOVI-O8}MbRniJF;wl5Hw~kifZd^?_0yAN9R1T!2?T8Qke2y6;wnBr2aae(5 z(vJ(Ls@;ta$S*IFg(UrzaX^65dqmWxfMOsxpJRyH=sA}SWIc_@zk!T}Wggj}AK$7# zzHS}xa3p)iMMLo%QW)ax$L!x5I#xtk0D{dfFB&NP^%NsUy5GO}SF*zS<#I(q319>g zvM2`3;2IQ|Etf12l#1`T>{C?x<|-QKl|)C^h*EevAwPqaa(g{$eC=&VMotqe1dM#p z&c~K*%6l5MU?NEmYcBTdU{ibW&$z0T5C7{u%H!UBc+@s z8dsx6;DRi(chQi^^s$15$&>ayVFisz?e{1B#XJ_`u+Xp&-Dm+G_a>n7ycgLc)sp~g zcEz{QiyV&Tp z%Kr9FD@h$K7UIB(TyR`R#6t2Y2HG#Os92<;%@MTG=jV5ZY7`r%=PG)|8Ggr^qpw|` zy_pQjN&+hU)RJ@&@SuQ~x;QM!dt%m|F`&lYfYFa{*|M}J@nz$YFmvFPt_2CRE)dwn z9*1~oE3p0|1Nv>Lp;_XNJP|64`owjvaSvg)jcaQ*7ZuJ72bJI!5tVCl#xIa^!8} zf|Ic3d;d!IkG9}Au+g|^vf!Men9nHkDHV&X=2s*Q^akj7LAJe3N;OAIR5i-2?nQJ~ zP&2Ro77I!t!rK(2g;ZjNuvPIfYwPl2l#OaE zSYKodO3Nr4?&)~jW|Bv541Dt1kQ1yoPtWRH z3rC!3VxV|FR4d1Ovjs~7SWmlk(=mQvArKXrdiWE?>dqI--8L2#ylw+4L_@~Gzzrm#%cGFn7cGCmpF@~KGWF0i#HXIli zy=Q7%w4I@5HLgZkee4OXk59C%c-aFhD1=a`#C9J@^Xv)#Xs2uwoL`@#+Q2^pvz#aT z2!r;6@@(Lh=z&HG)v=&>W=}XCeq^WB#W-c-n00$jI?EjK?w*}#=5bjz^{x(VO)Nxl z^XM|JAs5&uH0`Ih zIB+yB-Gsbi>0~BuGI>nJ9`Q|>eVpX_q2w1v=6d8yqmM%THann@u8#WLtCUklx};a- zy8`0_HiN=$uUnc&0k;6mQGMV_rZuKdc0Uw^QRbFU>lEUOgkyEW>V$V6A=|k4M6+;B z2tI+V_ebBm3^Hq>Y4k(8aG%=`Qh3|9GwZ<1^%QEGsznA@P&o_R5@$GXzTJvobnQo#_ z)A>C0&kw(S>1%Pns`16QbyyOM9i0P5XBtuHBvS4oMGjoAaL~kosiBx16xl|_-j*Vo zHhP_%Swv@tH~NAibGBmN^lI@Aw}&pNZ>Q=NF{(JKI6+rmf1wbR=te^?aQ5S__TQR)7deJ&Ny9a4)6oG-gmo4 z?~1Mu9E67KJ<2}rlY2BV!+~v_9KS1)ZPE83qH4qM|M_ohJTvnAXtuk;YDOG)Ij9*v z_gnMQ1vWmA14jYNOc1|;VzO`=VliC+64P(>MS|jBEzrr8g0y7{?_T6R=tkb?HSDC} z9E2P81y%dndG9m=jI2>U;#qqhx@Ps1k4M7Ef#G36!bvONJ$IK9+QGE)Ua#%0H$8fz z)qZD4?{?G3n(0Ie>n+pW>-{ExS%)=53~P}+(V-vv4?mkX)m)_V z`3gD5wSdHe6lE2CN&LWP@l+l7jB$cxZV!CorsA0MOp=(I$GYM_TMJef%eSq1cKYeG6gw+HUxs?sBYO-XWExS5L$ z#+V8Mo#n_iQ7FCLOb@zk1o7QLx^RA@vWwOM`_fMDOBm@>LMCaW7uL$)ZRg#ux!8ih zV9a1Kb!W}b#%)?GgJqPQz@(o)So(+Y5K205xW)pZBuYuBX+%R)P==ps&G0FXh>El)(>&&pj$_1Iyf;mBervJB}1L%jCu7T$orV9 zIP0-|E^xL(1H5QJf;~sMipSplk)UwulHiT>Uyr@7Kok+`erX+4ndgVWHKH#tBwLEzF#2x z#qAu$oTbR;R4nqct>@H&ce|L{>e?ee#OsgT9;jOW6U~CWfV_Ye9{oVtcAQb2gepDI zTA%hP2+hlRkTS-fZjthKz)`w|ufOGRkd|{GsDqD*&j!$i#~Ug<7rWq1c*EMT8hD$5 z(3GGq*(q;TaoB;NR>c76lVx)Xlwj$QWxF0KjE7`6&N@*MU8Tep>tmj0eYU@5^|p*9 zw#?VJF`Qe6NLJXIBuu3`I7V#!{?q=WWT^vVtH1QEi!_tM4mGV+%p zS%9t9M83q*19PsSpK6x_V1@i&J@ND^lPC34sA}Z$AbFw8>vX+oct|sQ7@wLt3z5X~t z-TZB`T_6>3Oqn}AwMA>;H3sW4E;b;`@DhC@YBCQOJFIK7(*!n8}HL|3k%!X8Pkz0!(x;@7?r4bgEs1RRwN+a?D8`m~l>MXqtpVPtpSv`daw`sQ$Yf zsgDlm2hjIbVz2IYUlqPVhF=eTjk{QY@<;|^Ox1zFzYcQ8Lpe2GXuZG^nc{`TnPq4# zj=6N&++BV}vJMuUYm6M>oVQDm`axxIiIGzXDafGWSe6hRFKwY!dtFj^$W^DK9}t!P z0r?#l*q`Zl>G^IqLU&K^m3E8sK_02Z{WJ|i4^IZskBLt}L`u*m14iYk9quS9@qkpy zT0s0E)$dF1LS?B~4dEvUGiexnC%6F3TbV$J^(k+OyN)h{fvvNFc1GQ3ptK5m*N*;e z_si}sH~MykcT11BsP-zihIi;O0xj9?lKoS1y{wxZtG+6$ZnAowts2dCH(9nWqbyvL zf71BVgugF75L}^(A#_(9rqaSsyiyp?IY1hMto?c0ZlWBQ+wedB(D%ESJ*KzSqf}Ar zhD*ljsBHeR8CAX={8m|l55!ivAZbQ){FIP$9&0?)>`x~46f#zu41`kcKFSeXo} zJLfOob~f&?V{{*=$YORifDY^h=b3Dx(kUi|BCDv_o%2pYE&vx9SZk299UYT+d(vK7 znsMt8Wa(-9UH_+dJAde7#Ds2s=w*^KTGYjXXTH{C=C@Ex2}O#iSlpT%5S0NtKd4D4 zaOv1M{f^Hm@EV6?yIt_rcFE~!cU;P5_KGvZFr2&e%}tO&%2lR_z*!B)v`-}mQ$hA~ zPprelXnv#qn%zKQ*{Z~4{kGZDh?e%(+dd!}4jj3Ka?KI(xI&77P;M^9<5rRs&K<8Z zZk=m0zgl+4JB0%TmQL9LOGqJIC&E<*UsrSb=fnvbIs5o|Vjcf656|s~1>m947g#*I5zylh6w4+bb11`YTvEls!G>N zSFIfP!<}A39(h2hS|-qe)KFJw3TKsj8)V1JTpN^lP1VRNqF1>mc;*84>eg_iyi#d6 zD^Sd8_$qQUv^7TawH^O-NgAvSZIKp2DHqZ# z<_JKHW)FQYvJc8C{*H(hGah=qYtF7|fy5GJ2_p&6)!U*!v}uvp19H1jYb#Ovt%I6 zxvHBU`=<}Oms2g$D!B6$_Ykkl@7ml%!Ew?WJ{rlUrlcjY^Lh$uFNpXv{{FVQ69AOC03Yl+*N$d55Kqnb)$b$ z8~mr`WD7h0#BtFMR7E3q;(I9uWNm7wSbbcs6?ChzW@O3}X0Cz!ai;u|cuhzmRRUFZ zD@nDl4g_rq!zzRiqMD^^r)O}Fh5^~4cc1*eYd!3P_9=VG#;D@3MC!9?o5N5P8ry0? zKMnReSgi|O^sXnv8)P;N;#l`1uxh7c(;!ae*RNF1HlpXvtEEY#j2(I$cS$8jP0+K4 zVs=ra8br`YGLz3a$jx^-5IEqrm{T6v%s;>@B^v^F2h~#P%r1^523f(82kVsJk#|b; z9lFkS8xQ}c@gU<2l~-DM60etT~OLv~=;Cc~H)g>|;fGT0V8wlk|g zOIQ8QTr!?bz{i1259rm52m+>2Ofp4QQn8yDY^J{`)_}LJ_U&-5k=?Ox*RQ3?vIY{C zp0(%oqHjg2ct(2?ob~;qWEneq;=qsq9-4$m1&n8 z2(05}PEkPt$+Aa%~9-Q*iR zK{3ZD(nQ7PiK+tckS$T`L`z&2dlu76c;~tCymZe8R$VRVgF;~2*0`qybQ2$dmeM{64a)(p}BAbWd zya5*dtMa>?zv3I_%lAdpF4!e$^vUx^vX9pAG5@|1K))8=VqwN&B!e0mpGE3%N z7OJq35i-3S1Mx2GlZM_y?B1`M{Qx-1b^$Rf=9IC%MdN)j1mH60mKMB2o;71Q6q^&X zW)#YYopwnMO+!JxB~#LTR>&WR)=kS*q(}D2FN09d+1Go7kY8RKoU6o7wTZ5DVG~ZV z3MFiurLVr%YLu~B9VexnzP=;eh$*4Vg3V;R17ivxG-7|#KrxUW+ymLb&_TBX@<2lO=N(GO|Eoz1d7NM*HhJ`-keu;w<8 zF7+SeV$a$YsJf4rZg;&tdyu%f^)e4Mv_hvM5@`EmLY+{$}0U@h#@)b9- zf)M5JyZzc^BZT6Y&pHeg2kZs_swd+5$Jk&0SM2Zp=jXp!_($SQ>%` zBsQ@SR?LqTATV87tQgZ4w^+6TU%$v*VP3Cdq0pnBli`;!eT6&+bQltly@|dCHA4FC z(OhLYgg9@>t_rVs-IrYw-{NcK%iccz&Bg^OoQ+S8hHm7 zeDt^9*?ewTPlD-`JUz3;7e8_g%aia@6(4N*_*-w7v#VMlF~xcnO^lA;C99SVh_8|k z{x)vhtV7;+XPw};N|(4`)A}jzOOhd{Q{JaTPSZ)AL*NKt?HJyL*EBJx8-Zs@%tKCz z{s^ne?t&X&Yv9-lH@03XUlH64EqSBFyc&T`h{b_p{T7NUu)04}UIPhcL$DueaQ4q- zv#l9vJ;rrx9kj>w7GrvNZp7Ua=3&|Aof8M1c?&V*;fMlm-Ta54g-~*!R|j6V@S<=} z)EC13$UO@_+Wei5utBw50=3wGUMwgRK@&Y~XXyc)9_vu>bp3mBFs7Mh^N?i&_pO@^ zDbtMG8D;W{{bb`y);9$Fha+~$TPX&-Sv3{ALV5y7<mUn!e!+YVo<#UU+!+oSwC9XDxurhA&)e6QlL zfl&*)51n3N9%b`X*r2ZC|G(Xb8EjdBb@JkW{BW~5r@sX3U8I zrz#RvmVqN+`u$Mg9;%y$y8%6M5i$%RgPt0ip6qWy*qDtG#@g(Av;8TnnQ+`#^^Mu3 zpU92g#=7*Nb7ZXpC-ACGJd0wA*-VjqpfnKn0SEhM3QXn}g4!=~YUu)&P<{MCw=Sp( zz-y4;UCi4PfXazT`_@4t^TDvw+3+m8^W=;X!3bvC!*;}o>=1nZlMnO6Mxe<4_D(BF zeP!xix0=8zk771ZB+DQH&`1vmRSn8>^3B}Zd8!=0RnU;$8wGMQhoL;a!t?z69_hB| zmg)PZW(pF#FNr54QZ_a`jhJ{l9lrkP`e)_{c|J1cxTcU;zyWflKu2)U@N0#^xr*(s zT}qX1+5oxikL4sC5E97*%9PAm*PvcxfW$>Kayll4RDCQsu+!9=3~Sl0EZRUFG~e1f ztdDA88{7Y8;X6w?7YLG^Lz0Jgikwtt^Q(Qa9%Lz}*X<-p@OT_*f4}J2SI!Ct&mOb; zk!2i=wz1l}0U4J5vogB9sEXTynOs1HX6j@Hi z8o8mdaCvRmMO#XovYDzE8wxg~+3sFec$mD{r|5^~@ll606cz%v3!8)LTvw9gWW76- zg~=*}8caxf#n70@xXUwWv4RHm-X8HPFQchw{l(5jBx|&^ ziUaRXpt*ZQUbd8C06&|dm{i*9RUFVORZs2p8i2-rb#T8hL!?61oIP+!g@Rq(oDAhn z?+j&qh??${;S>7!sv4RI(_Dr?{HqMeAnlfSVTGR&P?Jx`6lV%fpeSl*2Q)C|$^u2i* zDB$zxrO~A`+Wn7{g6Yt_=2cH81*I`*MC;ouMYjN*$z^T@%0s-O;H03d+x=Y%@ zZ=Sn<+7N`h@Rbtx{zy$sr`K5zodjQPm)C}GQr1UYgz>kL2Z77n@Y!~G6R0Nbv_)^W zP3P+A1e$&EV>o3sSB`7RH~s#n>;J(BqMuzqokR9Ga84L#rbl=q$0(+WB8RBhL2iw> zm(KUP=RM>!NZ<1=_h97im4(X*Idnb4|&?vWr<6|8g8odWT zpLs){u+Kg2A@Mj&lzmYC1ny&ixA75nC|~uz%cgoi0n@2TXGpdKg9+&OMu4e|Vju}r zNX4ejTsbpWfgKTxQ0EwsX+^R?e z^|&&T+Fl4Und33R$z%lkoOJj(7noPHy&~Hu3x#TWqklX73X${xJYy%Ti zzhFyLCa9jq&FZDw7A)Mipf!A-@aC+p@C!k$;dU}ICUfej5c%Gfn3d)YS{8bqJ|&f2 zeUY{Ev~nZ~DW{S=8oqNGeuhXQ((s+n-Yut=2#$HK4rrqjU2PEAGWq%}pRoR7Ee~bW zSJ;_yht0!%elGj%?;6*b|L*$lYvi+6W{tUFvc_DXm~#|4i)&1-;$}oCGB*O>K16J4A4CQ5tcH+HSS%vH6EA!7GW~71kz81$_{QXfpRyb$JyrD!q>~) z4v^Y;FygFcS)J=C?`GHqRtc8~@d;5QZW-LKyaHp1)A zA8YrK5_a%%;Lgz_Ch*!#F*Otf{bG*6%8EzMp#HI0uosh*ESe9hXvb*59 zp<^M}4TY7WAtzK~!W@C>mh6ttkSxOw3oCDWU~&@cwO?cj zSbW1Typ(beL* zkWG=uwQ-1>In>t2~P36`J{+iEq%K>1w( zx;%YAbc|jHc9RV|wStgP2fsi$9FeQs>)!y~AL?LCbK|Sn4+0gNxr$mk(M2n-o2H#H z=wWBGGag*o3PYofbEDDc*S9pqGZq?eYmHe?^({zw>r6}uz3Z*wua}9s zrJz9sgg|A1U(hHwhca_EGe6I!Xgq*wmt^yjaFzy&DVVFFn#z{>_pC0*!`R`KqvZsS4fN4ydx~gSDvA zHjJNEf5$9l+UhZFaU(0gW%Ab^rTo=Ook6tS=jY##supJ{_RTvz?U-MO`)Jpi^j5x@ z9+dU!m6YyyHS-P6bxV(vR8S>+_Pia;_?Q`uZ1q#KIEk!g#&MVVA8$VM2Xj_@3v}{G zHi9$*Sp2HG>AgP~%V!%o*ayGX7xTGLBFVcmRdqJp&d%V7Ibrd;H6UWdv-Uh>g$~D! zL;U-H#C`tuTacIikzonGoi|-gTKZlF-R*rPG6r>K^Oa- zm<2^yQ%`c^1lfYt@DlfCE+&-*B6kRqxLF+Cv}W!(;URGo_#9`XnwT8FUK&+mP#gIi zG$R=9!$R|G@(jNnoJ*YTULAh9vm1RiF}Qcv#FPpy!1ij_yu#obah+?sR~@g%Ggn!p z?BaKV-VENgE6^}XDQDY)k4oMegwo{woMs*tARpl#;#y;PS^rW@Ea(_dtoVshKSRhi(ujJwpM)LD$tomE-oD+66B+(gp(lh0Fr;e_jmlIY< zpz<8?ig9FDC!$@2d==i#`yD!j!N}LhF$q)`K7- zpv9dMvN8tsxAFg50?Q>ky_3W6WovB2i)Q78ua5;zFC5ecH~gCa4c{3?Bh&vg=`ym( zfsIUqiILesG219oPQ`8!Y+^e27h!#Zn9DhcxeSmrUX~I=zJu<|Ia>L0?^E<<_gc{% zphq79JxWbXKBt`Xr6gmXPLjq$7CDqT1^e;zzVqC^(1&)2zKjJ1qpUDCn0V76*jbg9 zp=mtx^pb_V3{)hc_bX2fTEdHGG&0pCP74=Lup|lpVoY^Heu2jSU{*6=`X$-*jMWS{ zZmjG7i_VQ8MticuDgOjn?!fk>$i$wkqnHeeq*JjMIMx0dXhmz6=)74Iv}QcP8ymmE3{rw}C1&(lJ!zR&s;t%)h;RRyNGTml8$_1wDfT))fKvC>bX0;kAyFude?3L^qagR9%p@yMJ;!} zB9S{Qb5=sbsSQ!?4k09jl_4?G#Xu#bQ`8ftSWBRPOYKz+|s9Te3FtCJ!%Trn(jw z*6|wbl*&Yf-~{K1mieVeHq%uyt)k3z*eQY61+f~P3?C>H6)BrxYlX=qdzdGip~lAM zi4|%leJ9b$^B;|*%1hb*M+3NSC7NcHLL)40S{gHoGomg;n!40pT08PwVpBj(z^B;+| z@8TIY7^ZN}l8;pnpt*s~%4joKS!0h_K#vuysDY(F{>(fT!6r5CxJedE0==i+xeRMd zL(mB_5ZX&)vv;lno9;R}^}?f3nwT?xu;b`&F&Z4Tb21F}Lf)jD`! znK@W2bUZISDcBKp$zK!mfTX&9a3|;wrPNt6KyRTmF;&89{|^#o{$cr4AiNc%a2}8~ z(^o>)7y4CWdIXx7TG5*6ZL%(5tMrifK;)X~U3Q{09N(PS&&05s6Nhb+`&-`sd}PkY z`#hVs1N*2J_?tUG5ob_-msc-rfCvuu9cB1^CR``l6r_plj=0NfqCbyZGu6(k;?o~6 z9ayu?*z8vvIsPjIK%hyVr|J^=Mp*j=O4-(Qe zaRODgq?l;s)rxZ0MT^E&6wUfSjm;O$dhl^PVEKz4jiAtB7aqlL_UD+ZzdLL(xCN#6 zB6_V)8|md~}z{4CM=2d0*(D+kl%cAcb?y4XGRM%U;~*JPt&dKFS8SXc|nP zw%-{m+)^`Yv>VJ*)oh|_4!kN^NNgu@D}?nSN#6RfKCQx{m*qTUokHSOY^%0^+R*8MVaD(($?!`fZ9uPC6Lr<&=Bz{99kOQLcpQlrOcR^wf8N7*!QkU~ znVlK=qRu(Yysp7v!D4|~wScUKEFp*`7AZ^pbx<&wH>)Z#Q+^D(kcOR**9Iw{mk9a; zl4pYNSrd3Z7??aB0z1PBd5+!t>c;_-ae&~(kFtWvwl2otcH$yTz?=k`!)5yCA_)JxsTg5!KB|T zaP(p`VZ%EfzeR_eEAzf09xJ4uM&zT?DJF#?tEku}x*@1^MgyqFHpt=`Rf+or&IF{I zMw>G0FfrPb_PzGYfB992dG?nLCLGvWSV&^`fVKE@r~J?NWty1Guo_WHNT279xmbdt zlQl~p`{%-D=8hNSaop2_x}_(GdSW`JtpkXuA8TJ|7)CFg^uM3*jj^#e+twWBd zsl>z#WK+yKie!L*D)eM;ThJ+wqj1ufFqi0+^BVZegI6f_lL`pFFXvnqUwP|uksUWH z=2OLPy3L=q-#IH^gZlo5f51aPauHw5U?6W_5;kQBmd*>lQ+yg4Ts&WSnwSxy)L?{u+SCuwa3ZUaO~h%IY zPY+PRp$r6&iy`fWiem6uE~h4Ns@*gxgliM#@0ZEcc3)v9GJfz3~ z6}yBK(3i#eJXLb&5kVq##`}nGovV7bdbZ*C3MhrB(j#jjC)^^f@o%Ebd@n(BdkUx8 zrGwuC*Fa<>B11Xsw1=bP_svg_tfwzPkwc60knA%>jcY#C+M(`9iFh|~y5QZ3f+D)i zuNbJd^XM++%BWnw!;+(Zm;HvFfM3mYPLgvhF=pci&J~M2s z`%Bn7_D02@j(ZpZ_=CocZDfrDZ?7v&R<6wylTVS2RBSnHt^aZ#+5q+T2lYV)$fqPt zzEDS>B@1zmij_e4!+XdDPC7MxOtvWo4!XUPXAV29<17}~%|?L%8JC$FlL4~sAy#H; zl16vSW3mw?@yll&hP0Z)mdyYnM#f@ODJF>`E2vn#m@8DF716gta}{gDc1&crWyGXd zfrt@zPuw9pL{Q(E`tRe3yg07s4_eYA@$*#H)!@|_)ETP6yJ!`b-si}5l6Fa|Uxkl7 zjmRh-5ATPBpSp-!iP(<`4> zI}e-APyz&)z5=uiCU=sV{4o0`c&wlMXZxFp00S!vGFi@_yUWUW%o!=N=@ls-i%}o9 z(KiE___oQu5N1QMPd=}MA1{p)B=Odhl;~ne-0z%s&%4fbHPOnpdMp*_fF-|9nFnc= zZGs{@F7@M>6r3Bo{UljUisNR}=0&IvsEi)S@u=xM()-H9W715vP)jB=!xVW$#n#dX zfQ5xFvt6mlHh-dA0^A!*(nEx2ndRdsBu|KF>5H2M#ZAHgu-W9VET^};J-jSh6I%z(4VsB z^$D)YX0zdbI%yMeaIdZGf6dzn*`%r+xLn^tgW)E>+IeSv{(tt~1fZ$>+8_6dHzZ#S z*$5_AP?iV=abyb>nc_^R%e0-D*Jb)%-+R;c*N(IuyH4BGzL_q#uOQ5T3TOZYB8$l4 zF1rhkD>$Giu0R0C!9h?&6!@Q$M2AG8xscF|^RFFq*Kgwee(w36@A;nd`IN}jSo>AU z)6PO7A7HZ#(dR*}mYxejS9v_~!P01q2Q2j8vGwf6cV`-nhLcFOmt>6;1$JELj2tpC z4ZA4@618_y(F-aj!~g4}7A#m0>!$tj-&WI!zbXH9Lljm3-H_b^3a`VYm9ronH14|P z6_fGrbRgEQn0y+_rwy0X@)x_DrL$oFuIs1r6n_jXpXV%gIYwWB<%X-meA~J7EYb1A z+{*lbt>9zXgFmvUYR_0~iUSWe7HmqKsKp~qgfh@Q;Tpv*=Y2jgVLEz)EAAC}?DAG) z9divUj-@en6Fbz5@yl4;Ub%YbFWxZPlzsp0_ZB(v+&CgPOsq{4#WYf+fr_q!5k)7- zQX%EK4vIU_J?-Z%_1_nGD>#jQKx4h}8L}ZVg`WwD8n~DCh8izAsYsb(xa80;HqKna ziwQi;yUhdUCfCc#ZqhAI^z9429=zIJI}2+CPJukp@hOID3_^XV)Ot&rB)=10=7maY zLr%|X$Le4qJ9L>_Aoly#0&I56wW^U1ZtHf1!$&hdLHc|DqQyKN4|P>m+sVC9)X7mAQylYO zL}rj852$Fvve9{C26zv5m-moUhI=dUVq6ZwTtXZ?8aeqO0G`dg0VMwYZf88z?fOK9 z9{G?H23BGOD+L|Wo3IDO8uaFAM?oI|Me@qsDrt3wYY$AhF+C?qzCw73lLZAxIh@)c zyaZ*aAJBWbK)<1=l@&rzuyp#Z;4^-hw$&^=!rA6l5IX48J7++OTX=mdr#c``w1(6` z9KDv5=*jTSPE8GTpCOEls z*eKre(EfYoVrowk`Z(}>8?IXQt_JI{`Q#S#U5h;S&s!gKhnpT!AJs}4Wy>d} zJGafv=5_|JnA8x}7FFz#9&*MLqxX1xCl?J0-h&bTUPkAg5OD_7?Kg7LK%hEXWt}$F zDEejW;26c1T5*eIaE#!eQGcke{oT)v$YZ%Jf_Bd*2i0S6tLc>Y@udIF&QBudbC01gH7qR~lq4 z3?5q&WDPd+S4?XN!e5n0t_$&sOw~qFE`+JUr2}137_o9i8p!e+ax#P^k2+_miZ~0j=;E*8YYtQ-k|vp`d=RSqOWp;>eVI8t5CN%E>0`&=7kn6c3a zZ1GK!w%|fYT`To!R;1DOicA&6iU4%A!uT*GjBn<4llWPyTx>Gt*#js~^!g|cShizt z3LpGnfe}LUD+-ECl(9aj($OX1m>-%&_et`^ z2bGutWMByHqzxj}dbT<^nkd>oFD~dq$S6k22pyd$8kBZ|Cy%z~raLCh!y*`+D$voH zs*Bui*LHdv7llzY3Uuf3?nG6mvLyUN_b$kG*3kvCn(4Tp<5LWKiK0vuw%!9$gOGB> zN9fioS{03o8cv_&rh6{fF&z-_bxM1}b44*>aY0R>Dw!*2<8JY&1G>JE;A;;wvbI_y zpmiu$Bl~h;4&_P^Gl-C;AjDbs6pi8xIfBDrz2YV>D+qZ?nj}RYr^OltYD^}{u?%of zrjh6*_0sBqB+yc?1Mv08`vE`*2AGl^ikn@pXlcby8^;48Q**|ZW z-~TN-IRA>cU1ku;c_f8%=i4ZSgDR&eh4Zh*DE#OZJ``W`cy#gBmliFd;;hXcvj!}z zU6iNZU>$5&85`=ieJ}o}IrXo@MwTpa5~K28dI-9yS{fEc>ZF%|lA_-YYv0XW0_&dO zSgx`b6=V61(cH@l6%!t==Vh5!QLq_;95_|T0ykeXznHU3dI^>&(tS?TO9Q*_>OXkqgY z#m1pY_ODm!$$AGKnqZw`*fcziVm4DGnTp1WO{}lY^=lQkayo^Zr^L?IC~#pcj-$po zdaL}tq&x@->FZ(ElsyjeqWb4Qnm)I>RLZn*$GS0#lk2?K& z=HdD4VbZ)Hf2P)Z0kFJA>I^I5qzGtlgi|rs_eM{HGbEdv+ zgv1V)u$knV15a2Wg>Tq|rHf)ZDe?&wjSDReFAw-#;(X$&Q6ME&qE8Kd_B-|SdtN*B zHHe=5r2IQ4=6~=SrjB<)G3=HpIWMemJ+4ahEf8*qJS@H!Q7ipanK`4(<&JBj@2W}l zQP;nj{tC!Oe>eM^TVC4vowzq|&foE?r2lFAkK}I`{c!J(*1fEKwejnzuT{Qu`Iqaz zf8jq`UTTQaD6oZg*%{BqsN)f_Q^36qxgJ#J(lGBl{b_Iq6y-uLq3d;653H4<;&m@4 zo@e@taavs&u{5y#Rs7W>g47xijo?+;^w3DqvTSii2~WhJKMwEgn-w@2*%k)Hin zj*SLSWBER7?=4Jy|L`A;MrqaRxOt@Wxf!R{nxrKUP)t8X?xDgTBttC4@z3HeAzM69TLsn@6)hg>Qok<9yTQz@GA}jQuCV*E&(G$H zn&ilagBJ`rVLlg52=ifIqv%xbmJK-_mtl5ty+U7egR$a@xtKqnO*aR}L%t3sS2Rc) zr0Lvr@Tj0hG{?P9qSH?xu~sKGA~kZz$$(kRZ(Yr6k>SjAJG~MbIi^Sl1(-x%E!{%j zbHmloB2E*dorOQSoLd=zKN)=f(csUf#>_%vyg9pG|8lV@8H%-%E&7GY(ecV%R<7vH6!@|KRtu(JuA=!1n}M;=s;Qx{0&&0mZDR$T})I zo-7N=;gkrkc%(>H!7LmWm~#c4!S+cle|o1_+09RH#13a!xib^qasC(QdyfXIKa}1j zryLj*kQ_YBgT72L7b$X{ipJ3X@M=wD+}6@Zopllvn7uAPO*9_qDo9~MYk=u3*Th(p zutd-g4tE}1L~Gp6EAbP39QkyB=9$aT}0aSD1AFC{qmH4tS2#oHqHE?8yulv70 zb<;fH_k{}YSO@|lQ?V9^lhm91y9w$@TG^}CJYhE1Sc4Ka?b+&D%P2WpT>XlRd9`)z1-N3c?T~lfN3MA+7UV#|!8(@;VSz9^M2#Oy zNIzHWz0PGk@y%}S%N|I!(5C$LIo~%g17{O*ao_|N3rg3AUN|1|neax)r!O9cXj?Xj z(xNxgN_3FIa!ab6b@9da&~jmHM29$$SmjJuc(SYxv$KD8WT7=fD=GHih|UV~;!?N1k)GaNbW;Tf1CQi**MzNe5LZAWEs2B#(~|`ttO*QBE=+7B#w%%o^e@N=8~+uB-GJI z!n+C5J?xQfWyYUw39N;+*2a#8WyGepd(U>JjIbDR8#oO_8ICJz0hzMH+=@7gSxZ61 zUv#cu`OIW^*X#Ht-V3oHY+Pp_j~}ZwUHtgsaok`vpiEr3qp!@%2nbc|>I$;Ofdi`t zO#rcjVqjjP#ngn83M>OeEG(Iv0jqqiz`q88RLIu^;Uc8B!`jXsc|N6+75etM8j4MG z1^Lv%wQfgXIe3pO(Fg4WraHkvyNnsu;N^1{bL1_1z=$1Q?oZ$N`h-V#*|zs@WWa$9 zNU8~5Vku?~MOISLSOa{5ob^)UN3GYwh4$(2XLOcTcz8w|_BhQ74-;PV`EkZ{BOYGQ zd;1--`ML2k4w#@PpJH+-k_{RK-sfRO?+7==7kOE**j$YSm|9rn*V2VV4Y~ueU7!+Z z2m>d{k1EfFs=K8rl0Mf&QH}yvXOO@pO@vxH_L8=>3S1-Gm`6wLfvpKEIm>~xwFRo|kvveO20v+ONX=J_tmVy-j+i<}l zZi%oN4rTbX$d-G4>|Mu&3d0PaX8tA8@AlAr(WC#_Y@o6MULN5_|7k1moG6{?`kK+a z+z6}sjC?eSw41z_7bxZ&MLsrYxn{nYOxjglvJ9V%9D`;^46L4B5pQ(eO)hY{Jh}|L zI8n8-+aS|=5Ga@WT~nv66x2(Tf3e zVDVCRptabH%}=lc?RQ(!Up7zTci0%Lg|z%ScU@>KlPl8F9|V@jvccD|m$&hZ&aws) z&uGFfr&&Q_;>Q*BSH@CP>c9~-3pJ&^z&42dh3a$;POa5Q>xIQ+Hv>~c#(Tj*H9GHzEH5fdfknPdDjL`JA8-nUkTj1>xmYk*9@N1}7kx(5*X6l_7TLbY zlfpE5KfM?x$U{ze5oRGGY}8kM&H_pE{&~%GtK5ERuwwy>9zGV|^Rb;^1(pekd;S_b z$q1Gx-MebCm|aZ8fhWmnCeTWtm^g~8rJ?~zLR6T)EKKxW7E*+o-m?xt z*0*P?D{sB^*1PkKn0Wojj|xfYBm!y0!_4_fius5lN2%yfmGMfA0;5Yuxj7s|Ez~_y zB2+iKv_Oh*1M^96lKfNUA*?tt9IE3(!f_<7sV2$G=BJ4UWoj%pxegph|0wfXMN%ZK zigd44R9(=JuoH?qKULyNEM^>IzA=y}mV_tChoCsGgT4#V^ghW}uM=NQ^+W|>tMiD> z9+Y+c@TAFSb;RsDt=VPwCOhof+CMH;{%R~)jE+02x0J=W!KaQ{p#Q60pN3Q-Oqb_? z!g7whgPe`%5?2I_>onQw6=Hi1OCteZ@OQEQ45OjxeM`E6Y@I|ZO$<#j#q6X=J{8?h zXNa!OZk0i?+q82eLp2oE9;#8L(Mw4$y;-r9&gQOmX^h&*t)xr&{oJ#3CVelWl5Xa= zlB3=`+*@V&u8oirbjGumk5SLTnakWXiYm`ed&tV$#+;1wyvBBnl{sna9-Fto=A}0L4NC?pE;S zAPrOz-wE3gvResKa99uWZ*Xd@8*HB4-!{KOGrQyqAA(1lf&T9u$p$}-Vp&KW)L5ZT zLuG(Ye#WR$&HL1HgRIQO#D@R6KPE9x#~m$+1u@T?v-1SY<{JLSe0j{4 zU+>#ZHRFCF3}|{VoX;NgH@h89={+M|1sFyT=I1g?H)0px=2tc;n1Gx|nl> zSK(7BTQ>z(E8%WDe>C*d;BDfAe&zc62c7Q9dgPyqyTy18Mbz+KY`Kr*%xLn{g6CEN zS=%vz@q#ihefzGZ!tEgk{aQVWSk$Vs!Xm55+lUui2A|oPoT=O)^ZZzcUEr~h@w%Jo zBbfE7rE54RAO;8ggs9|E9GD915aVfZ@Y!13(PkZC?_E|#XTrU$-M^b-bY$Nxi>e~| z4(!MtH*sW3DP})K_E6EqoGTp6cDxn*sq&y-nalOy1OELG&A$%52+#qRgjaD2ro{75 z_Hb3e5+FLsQazwi2ru7nJrL$xWP0Rty&P*}f@Lu;mqDnQ(4whRfzS`W#wg zLKG_;QU6i+-XG1^njLm@vaouL`Bf>Bj|7Jb1~f)3 z<)m=Bgr)N;NCABj_8)qllWdrF0BU~9cqc>gimK3cFI*Mt#67ah;yT6c@G+y#Y8ZWP zG_2Yh6NcbsRxqOEuf>;G$!zfv=@u^$qyT~5qHbw9T>_M_$=eTQ9S=^L#(&I(A58{gRds~!LCZUoW~tG1jcAC9EU>cHXLA`{eaqZpuh+k&$8 zpqmScE-;Bg+4oME4lN+bN!SCtO!9Z-gw@5j=EM>0;;X8^5lTkQrpdx zMuyTDAU$$}Hacn<(!bmIi=H1FQKGpi7$6M}?9KF;AVNnm%@nzc-VBhYgFsj}T>`5? z={`DsHf>;SyCK^qUKg3`+Zcs$wRTzJyz-!4GAIKPRZ+#<%gQ^h1HgoU*O&S)4=8gn z+z&*g;RAk)Iky>&;x_ZKYliEA8Rg0pNa?QR7klK-?;wTBf+;=RP8#M?ep@A1L|Y|8 ze$52GqcN&AY{+lB@&R!2)Jpe*{jHZG&%zb)jX5a7imjA$kE=AQzo6jMTx zLMnPyz+TRNYPa|p=cu<~yHK^?4Uc>xKk8i{)vjN0K;P{=2a8H_xn(XI?k?G3&p1w% zcVkow_-(q-9QR^Qukd1EPk6c*uc{_DgO5er4A$C4L;Q1Wvc&&W?9AtF+G1+_vBLB65W_~4|5K+ve(J_(sz`2!3e#$Tbr|f`lc1TWo zb;;Z2>(b9jl9Yv2>=aHVy;`0M(`od$yJ#4XVkWB#=nvcqye_kM-- ze*MC-baQS|hxIuuu!W+%DB_$4fy9crLr!-S!Noit4%_f4$%O^WHDhz>)yaik-3 zc|bc2VJwVjU6<#GaYc8FM@eXv_a#o*i|YL%o!1`Eb?!;rGDWqrPI@H#AcPHK8K5!I z_{N4*D=`%o*~IV8fxMISkUgF|LLphUPMXL?0c>@ef2({)XeV9HZ4NF8Ro6?a17ZSm z#T7sddso~}s~>G>6tN+7(oEIzfRfP5VoWzsy@_F3Ga%nSLHCB9BbH8=iwx%W9Ui0U`3O94rjpqau7W6o_xfO9*JTAVdzQWsyYC6?m9LnzFYuV6l0Kx;g0B40 z&_-Efl;J!M8C!i1d2aO`bZVGVD=Bli`<3{wV%k%a z5Wv=0xj|IRiY;?Jj2>$J&-W}O$sxi z&-&X+lIOrnJHQMxj4I|J#Q?|CZYnxm)JZ>_9}}6S!q7}*z{-$BUme{Y(XOhK;(WE4 zpH249%c2i)vmj?EH8M$_F48FQj-GIh0(rygq$tFJ;=>gI`^4+q`?<$t7&5G-S5Zxp z9$9KgwOgO0(}uI&QNwyftlMtIhaIeA|M#6MFB{SGkel%qDR*G>Trxq=X^J^Pkz-VJ z=8GG+`>9shL;ilZmf#xttY51vooNQMcZS|5Lh4Dx&`mBFl^)b zJK^R0m6PJhdM@s-nl#GM&BlNl0cN8!sH^}*{XOWL9_Gy2Pr5M(E}8>mHCnk@&ONZ97@9X%sVa46%09{ z&$B;ts|r`8Q2zmzP}_rXrw-`fvs9^28~}{V-dg&0c&B}m{@Gz>gwaTU$o2^8$dxhX z4255)0muTK;xR=!@1U|J_=2!ieoeXDbHlX5TGLli*A^fMaKkcNFFE&q|eUbetzZId9{4J?kFh>+UK*+ z8@JQ^^TYEZMm=`f1~ityH2S@`WT&ugZY+}y1P2dY2jy1;z+Q=) z8)qI<4moX-+zdu`uXeBocfxQokB#VP+vcO#v@>k@B3tf^FVw_5p8lbydGWl38ucZD z{eDos)#be069Y8r98Mo3A5?M5x#Pc1Yc(ot*yGkT^_-yM%JzkU?;7n%`n{S--c?^raa3dWEHG;Du|_se zS?S*sj_I|K;&jEY+oRGofx99I>W`A-H>VbIljQNF)w@JyxOzoUfzaTcsJkOJaqoNf zh^s*8=H}D_;R&+dw_!&26g9fDTKe3y!l^ZS1{XZpBTiIh(P{KKKJstpg6d702!GCy z6G~Af!eZ2*?3Y#p(DE=~mov`^0QVe^+%VG!KqrxEFUcCID%OD~=urASEQYw7Vu~rU z6QV$V1Acd0cX0E04IucO;gcsWlGV_)`~j#-jG2B$dI7k-k3}>@A#Ye85bkbvKBxkX zG(N^`@JxMFA1RV0@M`EsN0Du;A?l1YpSN-5s68GFrp;y`x6uA*4_gL#QSj!@sYcxR z{jO{Q*~D(5?zjj1M7|{CBP=05MB%BR&A(4SP``%3QT2_ zq&q?*srFJAy0oj%fzwaTB!|7wfkV^XZey3tv|GmrBGblfxcB$`&I0qwwC7~nZlR71 z`#)E(F#?t?K1|nwnb( z+Vsx95Or#PnNO*(G3uJ+j%1IxcTR>+j9{^MjNoHfAUGb;Ne}sb&wl<^BH-B#+4u^}2pD0~Kau;Z zx3`Im&PWD-!*vq(+&CloCeDb4Vp1ux2}xfJ#Fq!?JFw)9!Qj}4)JPpTt3%3|$iK%# z0&|hemWb?-Mg_3Lzq_|9d2pWU6+`I-%C}9@*uMWiM;yc99&vZvGDMN{P-(2PwX|dVrZ0O%g1E zI~;O2Bu!KfnHJS<*U3J}1;?;~ZDWm}_tGd3GaNm*{dr?H+`2O7fAS?p#0YN-e^25a zI1>V@nug7R({8Omt<-DD;cpj_;ZSiQQ6RAD(mw)%te03Jo)IHk0HecOK8kt<6 z!&GY`ivyeKquy~m`%33QCylz(ZP>vhHN|e!`NrFc)nh4&cHk+nrJ`ui3E$#z5Q-s_ zG_r_w%681`khaS6!wWgLq}o~QU7&o@z{LRkx_e|>VH&Kb+C_oJ zPh1=3ez`t2{L$1-I-V&D8uVG~iSqL&NQO^5QyP}4z%17M@O~~b_wHu0 z=uY~krO1^p{XR)A8if-|BAR|C$s*`?ni5j^KON6ZK4;xfH)4)9A0yenvxj%5# zyfEnt6^XM@ex&m{63|KtrYsfgrRMk$S1CL0nV5ZW3(A;kYMe#mJt*5d!IAkFpgRGzWJ@6lWGT^ zN?$V>h0apU8H${wqK^ce7O!)O2a%DFJS_el{%?HeELl5J-hu+ zf!77+)+nR}!skVtCZ_s*eeTkF@;mRQ<7h`wliXOT`swf6kkmSEs3fW39CrSwKk*h)e0!)T_kLP_! zd5}7Vvq})pD-pr9=^>coR~J;~WuK_Nl>m79AY!Q($8g-Ty%=GHnef4&e42UbnuTgG zu+!3hSX$9@ksb+8gLr5UAC;T+1Y!_d?~vwjwlXa;%;3dy>TW2|&K043-y7NCae0>d zGhuHy5b&H;ZVAMhhdoFSY|<^G8QGuFWvsfzaYyuTIQ=n7W;8Ep(|`L1k~E0`!}+j< znJkKdzELAPt?|N0- zFasu0DlMHZ(z#VkE^|5Qb<_Tf@-}WvWUlC> zB4tYKlulS_ie>7gNpduvX>_H(jz-@Y886!D3i1JLWA->?$jh;A4(}?QrGpW$!lR$g ze6dQkM9>~=_@2>BpGN(1Jbkn?Kg;_4kxml)!8`@VVH2Y)B(r;okXNio!- z;ebND2E>fsD+l}?CNA4b*XtUHkX4Xo_pgqa_mX_Fg=r;P~# z7s!T@Cd`g&HVIVK4I4|dDF%{n)2ZnAsbxMpNGq__bqZTNGDQ2S*x;jrR@p#srD_S; zO7G>~h05v}0aoj`k}RMd>fz#&VqwG?VTrQa-pfj(0*~Grum=R2J1m1|O52;a%sD{V zq-#6y+}Q$S2og&7geO8)KNdpO2W^O~l@)R@zT3vd$gjFbUK*xRpo`uli4kb&c3Hgh z9oGcjkQ2&H-vX&XEY@7dJm9FYoegfa^kIS|(KiEAL&}9MpfB9Ysg>cDp%^RIMZE+B zv9$E%ApIQ}G;DZZ;%N9FMh`Xf(dTLVY(QqX_D4{YrHYq@AHq3F!r5G(mGO6f~ z>ASMEGjqwKPh=7r5MPxaRPGWs^C5=ETRU@`I2RTocau-}89p^MRxTd*stM2L-q7Er znOqZ|YfsZNHn7+X8&5RE3M^Fm#Q$mfqK`kj#H%N_%F}&LNN>|0du6x|d_|+!O09fp zV7^9?IK6n9?F_^d4|}8o)&patyX|+39WeI%-;BBcX|y6gI7NMveC)thLVyD(_Iqy6{tM`pLOP%u~vO>-)TV)xp z8bu!|;H(UbcZZ;)q6{cZuJ~2L^__GZ_Z+{EbW1le8Qfc*u+R;^Y_GDN-~RGaO8ZK; zcS%%Z)W(@hc=e%wUrPPoI+xwv&jPLWD0@5}*=FMz0A+`=Lp!g3YFz)E#V7 zE}2<_(Yfy*`1;}3(qGa`I-U;&wI0Rx?=>negZy-ltT8~nd4@h;Z>#VAc{iu7vUySc zlf5`fKt0)zjgL(GSoGmnjE1P`)f?Gl*CcY*v(7B+wk7s zBXFi3g_7j6LCYWsuHL_rj!Yw;Od_jH=!NRyJ;smcK~*5w1M!C^81o? zF4?LI((jrsQsb_6R$OEi2k+A;Is*GZJ0j6{nRF2rt!fmVp@lAKqU0Gv!D<6@A(j*s z2y;aS5k{aPs^f!dfUHEQP7_@cUJ^n?7yd2@M?;t$a?53rfng}o7r&sK+Yxw<^a}bU zFyz3`OAo1(sk1|}Lv%F$Jp5-g$)68uA;$&Z+~~}=%vT)PI6hyzwTNA>UKRhNMP8;? zyh!oCO%L$i-QfBG8Q}dXo7*8x;e6s==-mpGTRq&rm($Cc#;97?O_H*p#e!Bkj??9_ zOV%eI;}H8aV`!DJwK&*5t>v+u`%}@J304Xly&2E zNT-Np@tL#mB#K2M@rr(ot5*KA-|m{Nqkg($dw9>h`#BlPGXB%;H(xgnW%Kv*+esC> z$|nc5CRa^d{W^*{O_38gA}@`s13L%K|%TJeQ%mH2cHJ_a)0CPmmt&aan8F2}VaZMjZ*z(uwe*vCu$A57nTj&)7ON{!8>gA9`FO95N0&? zY1FTVMPoF=iD$kDPhK!<=AW*AxzdE2BzbfZ4QYuQ6=sra=_P{I(jDB+(9v}3lMj^5 z5i3E#rb(Mz8<1T1w^>F*^WOWdrDXdga@6Dv-A6IIDN;;D*9q%GYQ;Dmj|)Pni6pr} z5TY~;GW!%Km`U5>=WQ!~W#uy;;q7|zkWC)8u@P2dl;c|Rb!YSE3ysiOoqFw8Bz6+W z4G&-a*+MZX6xm2cV`}yZS5zA5rmF+1cwJC(m@dfmOP-nze3fZGSpHHLrTx*K`3+Gg zT#xa`>+NKX4_kuZDQ!I`_-H=%9T)RZu%iZ(h=t%T#)}^Ej)bR18nR!yoRj3cp;Aqw zNR3?Ks=fpffMz~sUS`o%P%hoZWnCrx6dXM56l?6FNqbs@Haf=&yAwWnegEk{8coQm z)p7Gk=OnV$Bw{l_G5r*|2cr*&T%jJPfeTFMZzx*FL|7;Tk|3!QGLS-u9o@%W-gjnb z6t&J>^apMa#Y0Y7I(bIBssl8RG>T@$C%_VhHU0H`~% zLvunkicgh;Vc6J*ljCP@gGC?pJ~u6V**bT_k2nCCnrf8Y!uxUsxdQycAt%h@tRSTU z^$Npf>Uu@JqGE0wuLQ~{@q3#jsq$t;nG0r8KyK7>UW()+L8c0^RvLimzfiS;iJz^{ zNzle|bL2K(+4rB`M*IyhCsq^9B%pX1K7~oDXloKHD$ICr6q`_ns4gSWsUrbnoG!Lxg+~kRGZR z_^N|MVVwl4yRkeMc=nt-X(R$ZrN~nr2<@cHyjG2e#JZz`WN~Fcqn~lu`NPh`>&Z5D z(LKi_mp%_5T4k{X>2D@clAbfFp+A=BV{Uph6%)Upelu)@_&%w2=91cbl zVgy+DucOsxh1(!uTGKA^&23*GW{M8L8hxRl5U7fPSlS8y*x1~R4mBoY zsBy=(CWibgW`Q{g#B*X=EYL$B8TNH3y2-pA*}qiy_uJ-p|U;c73a3;eT1UJk3%>J-<#5&Nk% zq)-J(o!o2k)1(GQnHa(CIs2(XRfa1@s_UVs_PSWd&)~N6`$(=RE+{XehHMaJaMS2A zpO5L1s4^dW7#W+u+qR}e-%UM`FuVJ1+qP9f&VNkLE zTDdLjRjleq>k(sBb2hu^=r2B5VZM~a#y@djuhznfSdV-l_-=5S3q~J%Wcl17CkzN9 zEs;UGw#)@9q6`&JjS4J+>T)hs?D9r2OO&=15fyt7gYcAK)z|Eu1r= z-mf-Br9m9KBH$Fiaab!YoJv_42s2g6){Bls^omal??qJ7I=H48X!o{;cL5R0Ibo%2 z!z|tC*V}9hr^mQ@yfIpU`2==NHm`wV11ATL(pjk3MM^BRLg^w5uILvfbP%g+lA{17 zI51xJa9u2S4+~&eZp02}roGvEd4e(UqT06iZ)Cv9vY#_T5w8OS1C(-x1zut)W(`GF zQqkq~DnXMZO$2e-c;~t);|qhwJ@{-qJic^XHws?){#mbpKy4;)S^XPGHm${w{m?RU z%bQn-KKhMuIC+2lsuxIy1DDjUFv+XvqnNuC>87G{IO)8KfPLazQA}8!G@k?fE-JlR zwlRqU+z8MX*gMZCtOU_kek=ny3kI_?DBCmdwD{tSJGkmh6^8I1fVGJcpx1%X`8}SE zsL7N@A5^ZHTE@V}W<@zv`|D`ro4H0WA*K9szIsTBsWT09CLD-?Oq*g4;QI7!Ra^lT zgjH0(TdSgi8~`02H8P}?2VD_w2v=)C6uch_bYZRFfU2Bd0m@&CBef6>-Vj~}7Ycf0 z36L+p@DkC^0@+%+Quax(URwV7sDqiX@8H(&yBneR!>TRk$%i8mTsg46Qe@(Y$n!7I@>HbM(C=_8V@l1-fbfRWt+yP#}4!?j1QFP*}Q2dvmy2^!Yb0XCf| zkNCb@YlfZ}$3|?;b_XpFW#9Yr58nK8Eg7YTTn|E%bX=9}T;$P8n)SjcO_EkQaLc%L zM`Qv2VlAEJTRhi}K95<~vc!$q-S*p!c$n~8C!-C?zyHqrB!=BGv;$igjfr)MrA)UM zwMi%Mq8J!c3aIGPmuq3|r!guWY7uh1*K%rw2~)D^?R2U8CNd$1D1|wbJ#}pi?}xA8MtsR3RUZ{G(l`71pKGHm zPkED<71XLwUzhJyVrjLG#vrkl?j;BP9)fIOzkAo{R|p>i&?>x}TZL!uKW=}t8eJslvVhgmchT%mZ&VS;&s{s&6PW6Y)3!9akx?IY)=7|l7AR!p*pQCy2}jK8>#5-NF!8qR1iWWM zbA&5~gVZwc2Bv)X8v-Nn-YuE@H?r=zag1_JtlDOZNk%U&Iw|BtL^r>Z?vu2D$mDJQ z38n@z@=humC3|05w1mp$_JlV^ou2kk{QjbU;EqWL4}Yk(+#U^ZZ2A3`+LEk%kC>Mr zIqY1^LQxS8KV@EJUb&)HsD{Kae!XXzr~0PnSzt>%H?7cHT`_CN%pG8h^z}rEzV;99 z;^Q7om)mv60Sy8sr>Lv6%_WS_-Ddq zlcbNFa>K!BTx`Ij?qCCqNB4}!Co5n~c<1fUmIWCHmeon!g=EJ_9*qM##V1ULk1~om zK#{#r8m`Z9$f9xWXbopbQ3ovYSmmuQ2<;7T@zC)vOxY2NOV@dVg?UO8iUc`Z102P?ZI`@!=OyI#Ul=QOHeooy^Vs<4(XIT zkGrzl;XT~OsP`8A-$#BIguA?#@$fXVR=sbiI>7VxG*V;2k1c>=p^bI_Gx3eT7_H3P zHE+%*x7oR`4s3tco6M*dPhqtBOO2%aS##|POBb@J0N+xBQyo76V=EIp`C?A z(+}M{g=^i8s8GxREl8d80^xRmeJtiK~c8bZO$W|&k1pu?hvo=VhD)PYclMGTT%~CZoNGDwDTs(L6WXPDx z5ra{<$n?l#!q8>Pg2cZEG?uIE@f_gAf&7;pLB<|f!wrS~kmdJA*(c^uaacoQq4P6C zZq8oAsqkowN}Y3#U01r^cI}YD+IdJ@)XGVV z1=)1G;FdJizGMw!>3sWjJ%j^TIk&MdzrJQHqz(@39a|uESUIVde;}fTlfF@rk#y_|1u`JeWv+uvLKasvpWfXo+Md`6TaIU_m;#O1fc z*URtF-E=89)hl=vAiIENOC`!ZN}HRSv0%)erlw5S((GXzi|T*AXCX;`ZpNa$CSy?! z#bi@t8xeQoO=wcH*n4`o8h*EoOtqOTmGCQ1b-IRJ{+G;aNFY6!1LPLS-7YA6iQxK#$| z(;_m+2Rw}Z;{!~~Z4Xk8Ad&cc#oV-&{yb)icGKY`Tps}Sg4EiV!QrM(In%x`+< zMI>_jNRB(CSApW>K=7c`cDIAd3ZJ87yIZBKSlJk*^Lli?=4Fj2SOZ9fT@(I0`-)c+P6t%~eluhf(E&&?EPhlveJqnOPUNv5J(kFw;S8hURu2pN69TYu+UTmN<=(ufeh zrEBxZK6VIkVB>Pm1R=*L29(vSspt$kIiO$O&VTfY6fliZ+qm~g4ZWCiT$LZ*<&o&S zCGd*hZV_ICTSJf@UL!I;Tk2ooW1xmX!R+?w_oh`6yrY%AA}(`5)vZ+DJCa%3Nws>F>1SD2N(=*(lR7QSRtHv zEM?bK^NJFOUG}k1>as+D>DJZCG*N{xI|SDEYB;N<_U``SkNnZLRgide@i>06g2V)R zl1SIw$4I%R)8Q zbBwdKu^vPA$Ib9>`0Pa7s@IH8$BnS6&&Wp(>~yr7I2{)#24*lHLyjA#R;J@u_|!)o zqsxQp-FL|jduGuMQMitm?vW&~caI6gZ0#)`1$4G=jGzJh5g=8T-c?-;ycBRvywnBv zYNa(a?rRkLBbG{PfsNA8PO5J%cNs?mRLdG}9=%J}7+cpJ|b)YXFXb;FgxrI42Q!rw%J*nzy9CP zw2O?`y7=buFmmU)nL2-9l5M$cBC~iR8Kk09eY@%9fW5tu{qidw>4BJD*-Hjx?ZJ4Y znV%k*Me9QMMsBAoBid#+Mrr9y{@r8^DOaqH+%WAhcOA1?SxLuG|U?W zbA+a?K=N)9^`|lytVDtBc&+x#^q&$yqu9Y8>$#<}@IC zL@EEFZMDvAX}z9dkXW5&XT7>wo#V_UA}uH}9#w)yvwT3_Fc0*6mk2O!S`ylhStD8+ zGL#iHbUddP(q-c~=^@8eu_5QCmDrX$+tOngT~AnQ&34CG^@QUl-PgUP{>huh0q*Uh zU-pwab_1LPhx$GQPJ-WI`z5Mfwn-qXN&+cz6U%Dm7q2HF}E0MoKJJB zJQH4koCqw_+04Hy?g?*4rv<9z@jocar}C+VbyBR$M=j=uk%drj zvN7ToEQg;|^g{9!=(j<>&1UDt(^1DwqZo3kR>lZAg|*J*bTz3Au(e*B4WVGMTSs$( zmDQX$BWm_pz7aSto!@wftYK%I9C+bsy9vxvC}tx?5~=8B{z~2o-cfQV?7j?&g1md= zmEb$4bFTW`n4Lx!avGzKLK)>TRx`2D^wY3=*gUCKC%Q z6FAb~c%Y>*^qk?-%m>!!EY&SogFu00%-ySVk7d$4PC>n#dXHznYm5McM@5{Y&QBXR zY|>HeU^09NvB}q4Y^3UB);`*3LwbMUdx9*PM1Tu-n2rB{V%Ae+9TnX!i+9H7W&X(~ z>|ybDusdPh(`47CEv{x~zkz&&xcoW#SHFw?{a^H}!z(CeDMez&gVf=v|5}moYj2oq zY;7|EMH0n)K#}#3N&v}@1~DOh0oMk%TDo3=e;c%h#u*#d;R4^Tr~dnGev%a~9M|%G z^H}WPCK`>!vXy~H$dHr6Vi@SWhg-4r6tj*ZtEuQ+-uvg}aE=nxtHIg(zd=wy4=*dA zYI7J`(;z#jM#U~l`K=KVH|DLMNp3js+-0RnNcTR)^ire;2s~h7@~;k(BhR9D&M))o zcg2Wnb8tLSBj4R(s)O8@WzqTiQ@NspDy&e)^ahO7rHS&CIf_dELD=iz_s+r3e930# zY-b%k=+wr|m&B6-;hwSv4}4)hjYpjX?EV=Cv9W ztYJ;-v#-R7v~(7|fmf(ph{3(sh}1}&DXEJrj&bc_HCPde>gUWO0z@&Byr zZR4Ef@m==cU}Y|-O(&Os*VBlNx6VvxASom9PdRW*WS9fJDk+BmVp4yvMf#sN@AvQFuJNw@2`_q0{r{BLj~EZc;U zQguUq!F}b+Rb7OD>4zB%CVi{>fay=aw%sNggQl@uZlX1m9mcpdf^q>9f8|JYzHnSARjRpao#_$Ns+Yfi6PjE zn-u6!&!M>4H(dCJy5+t1%RyFXa?0#-KMabK`QKifGt$v9()(!HxDiye42 zn`UAR5-28)B5SE=0};ZFIZ2a2ezP9bf14z|{;@O11)io*vkt7L_l(CLJFqVPkB_dH zM>w9WFm&Kqtc9pY=d{GR<-%hT1ydF~zgJGz(4F%4u7du@E^c;60xv@~6jm;*mEE)} zr>$vkTZfm&hRyV5JMLM=%c(zSEHS4GWaDZ$FlzRgSeM7ll8dPpdDr~AvPYkgh7_@b zAZ}9p@4vYir1aKDTootD^ZskG085`tjTKt$MN}**$)TT})57s;x2iMA?ii4!qQ6!EcQN#X=1vA|LeI>TBRsPY$qK zeEdn=3_D^Kw1-_kUYn~VYR~D$A>{SEx8EU~9e4-=)ztbSS&Y z;k2vbA}j5;3~LntOgsnE9eV+Sm2IPbUDV?|mPn}sJEswxWYZR!v&=X!RY=J;& zDIe9Lkv*e=V6kc!(eaVG>jHa-R=2c=+wYn#IxkxpH7;E~iXI;q zKZ%_MDagx15lWTOM)W48%Dv=jqUV;NL*4^#*q^^W6V{Zu-$bKbUj>Kv2= zrq}yoJ=ACXC9d_r%QmFEF5cj$#zR{AtS~vCo|C1*1a{oECuBy0l*glu$=ht(9ac!0 zSibj-Y38x@FI14sLKSK=|G2E2ZsB|)?{XQGb?S>>%7XU!?DH-OhtimkI{s$oT;HAC z>+(LLQ5^T^0`)O_qUCWPj|Mi6Z?@xxW!T)9^Yu5U8I6hV#o((X)q#!4eiLJoM=?+y zl|@Cjc+^Sz-L8c|JS!+hp#Cr@SI^3I(;X>LAg2^b*^bZI1&kSo+;_}WU-RhUZ-U~w zL{(EzN%(Eo3LlLM7H$p6q4p|(84WTX8x+HaUGQNU86(9-PHcL~bu7#}j=RPLgVOMs z(T5ZhPmx$EdWj&7P74GGspPU@fmk(aO0z9(+uuU2ZLgOz*P)-d~Tzw9PTv zkax?Xsz|;A+mPcX{ai{hU{v-{(MVg+=Xx}B8FcKBBF6p5G(E|EEq@sYnEld2PC|yD zVW%WKe&)8YPGG)S#@XszA{=r`^zEegLkfF~9Lz{1Pdn=@UF=^TQSEk!mkW98;Nn95 zhx!sp583K`C%nuHv)64h99atpGeh#!U0c9~l~I}S=dbPM`5NJqKlERFNctpFZqiGI z6a#b1Tq-)9N#tTYN=Iix3V5Bg1PZM6rW^gpa(;iLUZ}Vp>ZWzHPJ(#?7=sZ&Xt0}oNlOzN|GDW-=aT_D*sS&h~Xi>1o>v5fizITxgfEDp?a%W?xR_ae93 z^#G`!s#m&W4kGF{!I)O;amyctP4~>h3=|xBz?UhIopnE;dgu-pXR6ABdbr&)@a&qY z%cf}*wX#AE3VMLMjH6dM2$%w&VAZrs!c5hW6aGQj(RTP1xbTR?x2U(80%cXHF6ur> zN%-*@a0ix*<%!otHV3OSU(AA{^J--W2QS8oG5iB{=Wd`eKUKYyY^CFP7li5qs&ak> zxlVFL-4U?FFm{kVb(|aZNy{J`!DF6Kap=&0c^QGHid|hnwv1${9M~^DXkw{$Pz+FG zYk`R{D4nxvavZRj+#@BzBp0njy_0)c&s0-M-wZrPQeir#ml*S{l{AxOA?iw1i8qSY zw#sh>H~FdYSZTnDkaKj&Nw>G1G7s52Co7B=f;pA4 zI%z(~pq|`CP*kiWygaB@y4ZWjX*q8>FFPa#67*qiFI?}7d}HXb*##`-9XlHEn73h% z{hBv8{~SwAjsveCS*Z8?>{VPV%6zfeB|~(WTdiMw92W?fcb_$oFu!L!_G?>quMaSe zGR2)mONiEi-5b!}7`6tzmtuBNq=<^n0?HaF9)}!m9k>YDA!mh)r(O<9R+dc#p0(0h z=g5(O9xhJoD(O0Yt?Zh(D^Q&pxe6!-b+Xb~$VNDX8WuKkb55`zWCrw_6 zDhH@{|8Jm0B>IC`eQWQt{19$U?UweBW3sgYX4ay6To$K?A;0x2-@f}Z<1q8uRp%aZX%gu-vF4vr%ngdPA(==K z@R8$Owa@*duf}oA_fcvoNxtXxdZ_8i@yO>bpVaeWlK+ZHDgJ5ndGTIus&ATaCEYNu zDX2Jb!GhB1>n3Xy6)vBU{1B8<%JEz(IKbUc?TJYCDO9z}_EROZ@5qmOw{kAXF?Z>T zyg->D+E3MplY!voj--{d4DPED-I%k9F^D|nhir4Jp_3@kL54G*fzZ>DkPC8*s-*gM z%C1OGDSBb)Fq8h+BZaB8B-mkfKd~4G*j+lTbB1lQvO| z-n+mxb!GeGJ>m(;#*i0aatc%=LKC#|FjT;TwlkeqJ2UM(I{$gznOpAsuXE?l?cAH` z+?hKwzED&|P|$)BKzRt_3q(cbSrDsLfuf*@3MPWJR0Kpsh2L69s3a21frN?e_?hNB z_Sp&9->kj%+G~BksvW*7=U(Gy2r@;zGOcvTsfflqE9WK%u83F4DtUMX&tsJ*1Y7+2 zBzV4w;FV%AUWK3QlYZE>;}$Nm|&UTD@Suq8>yf zDkR`}L*x!~Gw?I&_(=-~Tw&d&s^n!tvF0_;I3`Ec%iZF75Oz@0W@#6+iuWr<|16${ zRCcy;lC4Mm7F-UT^fp%=VuL-$-ADn>;?aw-9TWq3;M=KatZmnCn$-iBmY(DqLeCHY zJ_eZ*IH}|P?atdhK65Q~!Sj-k8v*L-;A+_DTsnG0M}3L5I-EF2Z1oAo?&k(SID0aQ zy$%c#3yHmXoNlNw!(G-|x>fjqGwk$vP;1~?S=F3k#j3fu$+|{ROZPy{wnlDP)!P{w z#(|jtLdG6a8()3xchCRt&y4P~e`Z@hX?EZ!4R8XEiqdpajE*AhRCH-G{hvK`cE2@jC1FGi8643$KzPTmgAJV0&Vc-rtJQMN55C0XFA>PRtL;ZmsL0Cb-80g7q&UfEeI|2yAATIs92QlpAP)Txc?P9eOwey z06deNXdXX49^P1aOO$V)V!L?(+m~d{aAa|4h#V(W*aqUZKD`i zlqXZs{d6~Tj+}+;1?2t87c9jJg36g1Rowh-@&Te(O9cI9;RWwoUQ-1EwiuYIP@m;# zCjHp-l~G1?yjk<3{p7d7$)`eyD8~V_GCuqvR|^7;hv#WjsZi`P(NuHR(0S}&SbH>{ zm9TivGIYk;c=Gt`za28)#kR2hJHW{nprR?RL&0JjCikiOp^pA@t@`FiKQKkd?7n%2 z-E}Je?z(+EqlF1bdFv!uJ=Xl=z3_+2J+^=F&M+#0&TsidLc6U@0CyGzCYNQJk2w`h@u=wZl@5TG&0Wcdz$I021PM zzk@!Y;+DWeVV*-?IDbsJDYF&2T(y%HS3IgPzIIeB3u+E`J&WS4xIo3h|2{Dygebk(sN!|*>#sb zBi}wVFL>>Y0dhlD#WT>%tD7K&ZUqp4E~D<+g*j;=BlCo7XNTcoH8PItyX{#Xmn1Vr zJ+>w)FOn?|99M-@{ZSE)9EyS2Cj)c$%V?;$rtcB`77NO9b%dlTs%Ghhxza^gy8|M$ zDBT9B0HSzaqr8jWsaWx1qIYS?kTfxDqOnlxU^OWOTlW?dYQ+x0Z=Akc`gGwY2X;s; z2shoE)fQn`q2vp0gOn!zJIB8$yeJ%!Cd-aS4RRm8tVYXt4@jVv`#?5jnkSZs6$GD> zB6dK+_+ZFEArddxE0QuP0NHV-vJ9aAIT7VM$7oghf9}4T?3hj}On$tD6tj;a1yuA- zzdj&KNSULRH8|JOTGCJKBA-QbgKx<{XXDTFY6a=>h zHb>$!yJ+kO+oeiA5*&~g&9K8@OnMyj^U{QVtK&Xoc^t+FTsdys>DPt-9`ewNJNb?hQv+>b{8lz*h#L- z)^TEpE~J4TlI|8CQ>@|je$|BIj=%S6`F|O$OzhhDaMI<#zRP+OT7^dx1EnkXuu`Qc zbby=z3WW#Cyr#%*6X5ExWQ)&P4?vh0jJ!?}y@

    n?NoPo>`%y?{#3B{_%=5W> zC8wPqqs*rF&dlKbUm0-yT$kmG;vy~zyXbb8-q7=2Cw!Jsx8@>YZan88lr)cAO$LAc zx~u|7`%ja#P}~A0o-b*%}U-cW;rp&@-mokcq5_Rxs3c0PUZSD6->iU2JVGpiey<0bPsrf?O(yG86jMu)YS>EzbCteum2%{D+subbt*kN(sdbwp3;m1zM&ChJ z<`ZP(N+a+(Be!!8-L?3tc$@D@x>lU0%2d@s5_e7LdFM-ucKWRmuI4@RYmU4}YUwr8 zkzFRkbC5=1frrW>&d57GC5Lj{xzgIG-xEtn>n@D{-s9cYJaw5%%+>!%ql%&IkWYhZoi zu=x`t&BSJGq?kC0tfiu3fUO7Uc3jUvF4KJ=;Hi)LC&0lvMA&|~t>}>*B6ccxubU?$ zJX55)77`TpagQ&!4B?qvx}Ow;HgOWeYB)8j&w@Jb$LMcyY^<{#Qxg9*`HYs4LhV__ zGq)2@qQZfLGZt(|xp1%WmIr3uLHy2`R{zhi9#$d3WUQv<#4;k}-}8IvJ>$%gaj$kd z`JA0^>bMbW5@)iK8lsp76dAxMaGswI2sYb*+do-mSSqa!u2LGfn~>~FjneENLKGND z!!+{5&^*7-{a3pVJK;?z7=rZdJn<=>8V_y@*UNGxr^$8*F=2yL2`;;qMdU-?ZLVwi z!tKGsPS-tm3-IdA`RZa)2hqa3#ic*F{k64AI^OA9vhnr)H?RCl?l+rW`#fs%+x@Q| z{#oAQmN$1Vx%_tKFFK>tD}QwD^#ea&v-tC<{#Pp&*ZpJz99{a^p{R}DEPnIi;=IK> zU(cp0UpxB7t*Dg6iPZZUoDvbR&94*~z=ZAQ9LqKa@w*dLE#gBXzTClo^~XAP$Nb=|J+eOn8E9^sV#(g%L;>*2?y9 zyF%=1tgJFiV=r9nYM(5drLh8%$uG~FX-+!luqz)6G;)|uiQ7luoD87FHaV*(>Hd1_W3_X^bB1 zb7|+s`>V|bR~^=V95PuXe<7jT8;bc-NNL+5(Mi=sp~Q#ZO-DeL$^ zOBCD|SSHNzjgP42#&art+GIMZMxHmP#jnnN2V>V=Y?gy>HtE=B;IUU$Q_Zwe=N0s1 zs5BgR^UM;J#-XqkzI~Ey!P|nn+|T*7^IN~&3@%MPFehiu#!c2#PV(3EELfZk3~so# z^EZn6A^txFbJN5iWj`p5^t@ny$0hSj2R5^n1B1ju4n(S?)hnBe0*tv!t^qtSOkbzG zBFW}* zWyB1P46W^7*W3`E7Z}&s?-kAXGl?6kYRrKl0fgG4a;mmdObUM3=x*Rt$%aKv!t)8w zXSw!@_p6R6&WbBJMIjyZ33Ao%=Deb>uUfq0JALzS*pSuqDdWQy1W)PCX16U*LbArU zx#p|jFNrO&psja)$)TtwP7mGAN4~CtP~hqU#;yDFHoa0O-9%3K6pNd9C-h=_W%NbP z86oC6w34B3Eqm?LH)4KX_QQ3r+IQ+N79w#mG+m12@8l1N2o z(YXP)p5M<^r_+s)t6M@hGwM~$-Vjhws&n2cj(M?IVb^tzIhfd)3-hP#cV4^dZsgaD z$cQT8caTbU$Z%kfM7%CFA*rZ2@`|k8Wo^VAriYiLy0M@` zRlto4`&gd9+ZT0R_PIL}gQZG}pdDWCk|l;6<|a`46_?4)`2~x5=oP?9dDE+FzIMT| zA6{8GS1aA)dIi`k@d~D>Bv7}bbWx>&vCjJyH|M9iYZnYT8Lp-JABT*+4prvtoAcMZ zX%|euwz0$1s0oPGA$#n--CO!;fDu!F`K$Ig+2z3bbSF$ua)e^QA1S1wcPk1cD+GDs zE_%13T7-gYgRZ!7xEq4^=b7HnOp-(P!sbt(YbM#PK)#e>a?nRdWAOF>=VMM=P&?!) zWUEk-sEaP+BEx)cKnL8fqZ6E4;7Je%V*k}ze!`Yuu>3CU?^yO}$B5xS*RsC6;8_=8 z#E5-C=w|<}pk7%*KuI{{G-8wlwOiN9w#l@ztxPpfqteK)`W{xPvF)I&81m6I@|w_i z*k(+YB`rD3Jt$OHDIR*%xp&cLzokFLP6p|$MU^Vd=)}FD9zN<&*#t(mgwKe1VGouk zpI`@{%Xg(Y=5#!4mProm=U8ABS|`6MtC$6Hg;R8h?FgDM^`54x5Za9>wHP zWG5AkDe1NJeh}*3<#yD+gWdtOuv=eT0em?B*NdvTH~^LOfKVMTdgO)Nv32||;YRLY zgjUujtPgw?bbzBq%Emh`_NrN*(2$G_jsDdZ1Mp-Y?1sc)b5sBJpyAFtMyy==Y0(4H zI98O_fp>{}On|MWm^O-Bp`r&PlDO)U@EZXd`CeWkp`{P)h5egUR*~Cy;~%=LqKEmo_{;PJRqB{QSKHGJMH79d8!Tfal1M9h1?gQEToZ> zECW^Jd;Xw;RYu1HJs~FJt&Ln8S?p5mqLJf!4?DHVfX6aLxlNAjw+5BPUM^^aCiyl9 zOI_fj3(7A?1^zQB20WE*RP^yzuZsVE9ZJj!p};bqyE^imXkSF3>=AFL_?%!L zPe*qJ4LKe6KPRZhRPwKl!19;~MzRSwG7P6HF+o{$P!o$7xFB==VM z>t>-LKyL|On?hP$)UCpHen(I|CnI!oSOJCq&|+lM>NZG>z)m2=Gh_qP8InsEkq)3Z zRA{yNR0B7C8YI5DPms=8kGz55!ABGWA;1r*Xqc?_ zMs<=7aN94!Yrd#KxRYDW%Y{+JA9WkKxoMWoQ&|bdAH^mEG8;vh;hzN3NAo5-&=n@+A z!!bUXPG@@}s}=htg00Z;g}JbBzy=xEmes`e+_ZjS-|viA`s>Enmq@1rV`;TX4Z#4# z+@(kl)~X>*YL|PS$AIJu#rhqLZdNZB0)eUx!9&uu?mgmE{u;rKS-|{p&g&BojqiqF zwc+Y2?=``Qk&j4^SRLomPr3n=A4qNmp_SSM0$MrrpiGUVtNEhVz&fDyR%3Gv5XQq| z6mO`jL`~xBo+xs!ksp)qrL$<%sULKOwivBWDC(L3EXfa6~hG9Xd6kt<=V+y{cH&7pyYk%c z&P|3|(}8`ebQ7bONHOsgSx-e{#6s(H!Uxu7S{V+xx+hF0u~AH!m7TECc(Ll!7LEIk zE$J~YiTjd-lr0p?wFc(;qb?rkQlFInla!HhAaADfe}tuW`+WP=C?-ZHnq z!g-OT>*XSvBTK{H?I(R8xu}tAUV68{uU~jsS!1Gp$oAW0-`*Nr7(EyEoloM444+@^ z@iRIuf7n;Ok?eF}$EDI_J}9CXkWk!1Meld%P@Rz@XHbV~sbL|lk!J*}kMoOsieS)> zxHd~0y-K}~E3^>QlWReas7%o+grtTAtwLn!C|9-$m*#N>05KOqX|a;C=7lvc83z%_K zZxh~->D|l>pwhVpfz-V-mpj*l?&qj;xhu$^>qaiFe(-JbMY!=iD6aHDT}(Clo~S5; zS-TT=4s1DKkB4s5(Gx>q-33+y=D6XFci#2??o1<+T3%{;k0h}R+&J*ORA};p$f6iv zf6##U$|<2zMHk4~;PS*8~4HNb0{dFu834lQxRELXj3KdYNGN zE8FMnSL_N%eP#4Yxc*P2YP84IK zY~}GPO)*ejn@vTZ58tZ~@nP6bjVj@7?pk^B zT-Y3)1+*esAk9mb75l*SDHwJ#WCAvEaPBOJu;Pa=6nR!Du>|vQL>zCUw-v+t^e=jh z7L&%r7=7B}lI2;J8i3{jy)qrRZ^^QHuX?WI2Cp7vcYqwqdE!p6}A)4JLK1dqHq*7!n z6#qh6j)>q0;XC9v49mcGhAr`!jb3caCCiX_)P9#ds1SOb?W z0w>s>6>zEb?h(_>wWHa1B90pzAQn`mad)ysf^^}q$%k`|MvfUS?cis$NG>gc%bo6d z#&o@}I+h0h*tqKVM(eR;PmMSEl-=waFm~hG%eH<$`-_P zKJiiK`e0>E8FXeS#QIpCI`u1m_J;49m+(3)JCTK=+&eQ4gx{f8 zkSzKtFo2}Ub5u86Q~k5(?Y^o0D@Zo?9+@f=W0*}?Oj;A1ekS%9yGiTY)amz@7|qf{ zZsyNP*)tYwxoqO-oTZr46gf#nYj0%YTQ_T%210M0CfY{m;cH|H4v zwYV)Wj^wkO=Nxz=;iL)Fj#3O%`W%FfgfNZUaY3DOr9dmqbk)c|B5PcRWGCr*8Yh&P z=^4u2MXRS*z$DWWSU9&)qLY+Rb>W$!?csG&e4;T@Lw)|eJwIvtbgxJLn@t4lE5od z>|ij*E|>nu@1k(nsaKjPs8H2-p#2$k+U1tdZSc^_P@4t&t=0Ey%V?i%v12C*Y9nVz1AWJqw!-A6I_beT>CJX0mgl1}LQKz*<`Zk4b&d%`i9s@uZ=MgY z@GlBQqF(H zIYco<6i{?VV-7Zw$1h!K0MU8o8c?~niU&c1E0&6-Zg7`H;Nmx5q?IANFSdmp%48V^ zw2LH%Wg5AGd;XADUPvKl>C(>>I>;b592fe343=qoa6DX}kjwkTz}gr%dKf+NCKI^8 z4mgkg@uH_W!8n`fpac8G7U;-v$y5xbF75nG*G+CYV3l^cwFFl4`a_X$;FLCR@ivakH2)5U08{-VC{-ypQe3S*ADbwYpw zVU#8MfMSvzcQlqF}qr3qG1J%jgQb-GhxSCF);S?*7R@#=%>(j=HB{ZJa zoz3ot{dsP~U3(_2rd@HXK=LXzZ*d==?IZK9Qr9d6D3?i6kXBJGKnTj*Z09L+tBWad#E>Ue*TTap(*iSuXGCS;{7~q4>g&3|h#=D#J zfEe3V@jtH1&Y@5&=p1R}+4N><49S3`$D3r^!Zbw|GYDyr*TktaugPkZ@e#4gtVP%O zS&K47jlo&2k9gSXv-x(yjdj>~Y~G$cvJX-$V`GdZG4)Pz@^W)YKZlK;KmzwDZ(=jW zBv51n75#{Jm|N`C>bgtu`SYNR#P6T~NOTb>t*-nNxPPrgg?)y8oLx;NYbIVJk zvXgNbb0n<9hQUyb*@Jy9vckr+d#6@*np1r|>2)}+*F?%q@bU$rN6EZA&#w86k@)(j zJnEKsH}$9Uk^L{_aqvnleGddn8aZ9_8zde4#>kNtcy>N=AgG5vE7{_=Oi&~oxq}Pe zdM1NO7ZurfdCLk%Pfk0duhy}zr#$F&XUOK3J z#eb;OT1MxI9|cX-?|n49tOD!k2d3zl9a#O6SWZ9eqWWz(XD1cDH&iP@Ck^ z`+Z>dBVT||^@d(1uxmzE1p^mlE{&UQz+!_WPK~h8NS57k8|f9_8aGkU!Nf==q$=VE zh;xZ^fhe4ONP5IMKCo8`l4!<{kxuUl{#zN1efQ5#o?-ZGS*nhpTpBs(?t6_id;&RE z4eRLH-cHBnXkj<34$Is6`{gw?<~uhIYlAH8#ptX4J3=7Vj#OkDm!!_TKvoDcL_*d5)6J&svncUsydp~&B8l4_T1 zDj`|ZNtKC7I!G}GD6*G|Zs*s-PJA5i@I0L)1>~HeMp)Vym^8aN@)EOrW-h0EwniSu zjqyJ+_oQ6o+ZfnRPXFd+R4&!`>yD_Cinbt#b9vsC!ZT3fS07l$*{;ya`XMcL2N2^v zVHqa-%}jK3%ni(BKZp7AtVYLim!MuNmN!kNV#I+r{47+A)P8dfuS{5_#O2SL;5ETe zMArt2bF0}WU~VY z-@v$ya=>;{Oa?{LsOS~XcQWye;lGWMWl-XABl3JCFz_-}Ouk<{Q!CQQcRXL|gKP>| zWTBCt@X^Zfiv5S5?YJ9_n!Sge9pfeSX!!mD^OV^q!*Gtf!Xp+EV_}};>*(GHtP<*# zm3-qtRO#1hU%U9-Q$M@&>%OQOP!K+*IKZut9`geq{kCwutbBHBAh_5g1D0oxp&bc3`J7%VuMY2;#Ik@9z79Icbl> z!ovbV53*~czNdb%ymD@a=(^{CYe{$_lz1hEKA_KuV!&I;hoVo&Q`7kNho=8F^(}R} zs0%ce)L3LSsK|zpKAo^t~C+dXX~D zAxO8Qk)wyXIj{z5-V6N`2wv|X*Z2)#_Av{VM#b8(2FqZn3%gurXD(K1vlZqv+73Ip zSRlrZ=UgD4EqU(-h{5ayec*0LX0yLq(HLWA6(YuX&fb?;A%ZG%=?U39B3c7b|;)NnGnhq*ewy3W0p?)E(ARVIw*G|U27$6yd{00Fhgp*ijR zK1m&a*eNY6Cm4^G366+S8{{6jIZxdptK=1v^c#stTrZx}mqtGp-5O z@>+#eutOEkx#5}{@X)m|JO@foH!#JDVnwSuSO0D;Mq+mI>=;CxKdzI-Lf}84macp&e7AuWCAs)`4PbeE|G~UlLFp z{=v)FzFHeOIl%^2$z<6Y&RL?~<(4eF8hCzA0;FIMNZMp)y;p(8*;;qFbs=P`jG&bu ze>}~Y{zmL}*5mfwo0G|UbleS0OLcu^ix2wr%6jN?i}jTo7s1(W=kMcYa8T1YhnuPo zkxUgV^mYd74avX%=YOgH=U2b~^?(0PyqaQGQe@d@epi@Z z$y(iwE1Xl`dE*M%G!`+J0|&xD6J}H(YzM`F=e(VYMuJGB-~&hbmPdkf3o2u_NRZy7 zOt{e(5_vJr4%D0>K_CuR8^ca55`ADE7b)YdIBYNytmlsrT-Z5yV%bUhY{6ScoJaDG zUat5%S;ekm+<`anc9_5?iDD8d5>G`R@2v4U!`7SxzylwHzJ z&_1eno;r}QHYY4bcrs1*e6#iE&w4&e9kt-yv{MwY*R)`O07h~@5VcE|D@GS+UcYDdE_&6N@~CI&fyVm44@ z9d2yl%&Y+(@m`r4H_|J4A9_u(9oBzKR(7h3=8RCQD9C{0|3VbF`l&xOy=#5TtRnYFz?vkQB60yZKi!O!Y z^W{_-T_-6Z`P`T#%~6lUO}5X?HX_g}*gtkupFFZ2$*fv__u8D07mTjyo4fz@-^uo| z)-Mjcem!bp?+PdeT#pWkuVR^Y!69gi_v~gSCkqNzBh8|*Wm}l`&Y#~h z*B^A)E#5*CXndh4m<@VO!909+H~a4GJTpGf_y4Ja4a8%a9nuF`+`=>i)pjXjKY^-#J7wo-wOZ zW&)l46tjmS`BZd_a`l2^9#?t$L5u)jn0$VdK#qF(M6*E;Tbca-=!!`}k*ad(H*C~x^(^68yf2G=WSY;jSQ;89`k6s3tn$R5CX+^77{|N7yxPIgo8-s{sC z8S~cpNZfD6g5yEAvOuh6?jsulEd^n1y2$bA%$rRCI>COPE4RT(5-WDn=hSM&^C*!4E(E`L$nQWmY=9 zSBfj0ZkLms>z;%3$P*g*CFKG7c1Y>$O7L?}$}@P6+idqOQsAA-%El02c(oy!5-aqK zo@2&7wi&9}p{H|Y^xXe2qUUD##uv#=c2kf8$79!+Xg}Sjn0|`%f#0ri$4 zBB#b0EUfY@ou!qna@V-MKR`AxkO`d|kSs%wc^9qaR4H}*RM%YK`-^qaKajxF${x@g zf-#~lhqQYL`z`~eCyo39+3HeG>PQjfyOq&c zNeyMC{#V8CBiA*)N3tw2v7^ zOY)t%SpFA=!&-y%#RF4xz{(&^YyQ*U|HXWB*m=!P4sb!` z3v~3-V<3ZKsI$-?=DKWhd#JDIS&Eu-Rf@l-Mr^bSa_~BzazeD2l*=NHWvOl-CU95{K!0+*;E7v<8TZm>Fg zA#+8`X?4+SoaOasWFoT|>fPIAL$)uk-)^W=Kyo3`3EC@{;U2s0pv4w_;KJR(>jwS z(?g26Pmz8q+Mu6|d9%r~Ca4mFM9CgrEGO0fVpN)ECfUSOH_#b8)Z~fhJPNuSl&#tZ z{tn(i;<*;d1@9u6dbgrLqLMU6<7;JzJLImbliuOgaAG}r zxd(mjg;nv@xdBL;&<=bZ8aaNT%`;GtE}I^7jq`}}DB#9Y_?KLp-9%1sEA$^ZpQt-I z1ssQ6HT8AYtTfkpXEU`qF#Iei!-H%{=;pBZ-+vz`usqd(>n+~qKp;tRZI)RFXU4U zI1yRML8+AudR+8e7NOU$C<$u{!4%gVh-#n;+ptr<=%|0P?3hX;Z-;zFjMg?YvG5~A zuFy2k4v@&nr7tbo!?mAVI*D&tZ~n*rjwXMV)q2Hoe{v55%5R#NFI%V*M)|#D8HPFT z$nrgsAhu-R-4q#=QQt*{H3($K?pjI>4i=fvD{smx{11`UZCeSti)XwLku?z`D7Ui|2O5Id=XX zdB*=trfkfCL1IA;E}Omsf%!B=9CxqKu+#?D_2B1M>Qc`KnQ7Phz+CzRe|2wYNjPMZ%s4rtMN;AiD#TjZGC?d^ZI`e+ zt8aR&pQGX0aXPID zj4n{jClslpqDufft+aN*9oIH;c4jLb%h@t}kNAe`Cej$W30w}H*X4N!=uY?ZSfMH( zly&g)R5c20(Hxn_IYGv2KOnr}n(1~zgyuTL$lf1n-{{4f_o_ReD_@ z_q>PwH1fvCLDy?D4d?q6=ObC7m<0?in@y|kutTo+wZ9Sb{BJfct^)(hLT30Lajv45 zRC;J3JppD_2;pIhL4W8fPTK7K+~&wz3oZk(wMLcWd_anB@M+S^9dykD2z=^#KD^QW z*7FBBjX`N1=fl(eS%Sd&p_wdPtbZqVEz^5G|E>81WnmsbwpM*Mz!z@Kbv{55g^MK>772@uEt>;-U3`p|D_3Uy%K4`?;oR2y#sggli zS>Oeae8CNFn&;>48(pe-Q2UO>i&z|wyjyq)o7&A37;(l1zb5m9jc{tI8&>#zZduRD zKby;kJxPP^xG*Ocw9fip(#^W=*|*?m2rS-`Wm?Y_5G&S65BmHQB7@c<#nOn{>$+v6 zymIs619Mb3Y@U+^1YqW*e!5E%m4uhN;0|ArJ{!*drs0VJd-f18Ml$!bH?Aaq^3`%p z_X<0N2@z}Mc^t#hPLCoV+}F+XU-$e@F*>J@uWy%}6_;^yAP`&RS?7$TO&a;S=S!JJ z_hea#YU={K%}3UP%VZK7b@BvSZFHe-;Qjab#wGINrmY{6bq*Xm*=4fmO{JKv6iK3@ zYm^@gG1M3DcZl9j#Xd{5+^ikpX;1-~v9Ni>2Lk z1_w7w&=J(jnxy;VgUbY%Ki?6Q!NG@-+!Z~8e1Xk}jP1~j`VpEO+Ioms4UOaadf$2F z_H)mAO8za6A!#;jPoGl`%CMvikyt^m46r#CpN=8x(E9X#Y~6w#T95tZznvyS<>k0= zCl;u@aOTOTTWK}gHY3a1q42Zf+!t0&@irffymcT#>t<`1jNjkP&>LCiOgs*|g0Vmm zl&vbGk;8n2Aa(|5?$ByvOZg`l3hM@W{PD2MT~?Ojxz>Hx8_n6>92OE5_}X?!62f&} zSObVO_N`t`ytQE)eVZfkES~eZf0?i^v@8&nBBu@~)|&vmlVcYojMXVCbWo{(5&Xei zHS(DXgdH&Pp1)9%!~_uF3>_)&!c<8(&7v`?dQ6|k+$CHac7dpO3wAJBsuo!}r^_v0 z)B@_FUG^qiXNjCqKJr-OZM$ohkrPOcc`i1tX&1czNjoVWi$}$AQz}V|2};gT%qa?# zB%{xV?+wxW5CD&SK{HezKDQfv#>mXo7XoIr`^vWaXatOu7gi;5<;+NS3Yg z%UYBwL5k9}*_-?)nz1`^JdJ|ZxQ7$Nm}NZu__LpW6mG)n&d zDN;#AL&3!XJu~%HS=ziI*+J=mcptA=bYtN?l0b5}#p343eY}%&t@zU1+Jys5d|)Bx zkV`GSc}5lQ1UaUN@jgvzM8zV*Ga%=!Z}ouS6K8|Vloy0Gs^y$Ds&dlp`GMrJ($H6_ zO9g#ad=ART)~M1a=M>o!H50>&;VH|g87rW2X8-PH^U4bgB@39mwv)R$xL1||4C1N& zD@ZnXrGEl1)&CCsXpi$(`mYdl1jR$n+kPLr&jDY2L#&74izn@Sz-kyA*YYo#vEsP7 zW~~LW*Hcmy`Yj4gU?B#WEuCN*s#R9;%U!?)ueE3Cro3vmT|rEk3)fzsbzOugGnsZvR@|2vg67d zT&PRS7oj~wuXPW%E2L4LLzfc>)0EG?8#I-h%q&r3N1T{;{auwH!hFhkrk2ze=0`&Y z_4S};c{zbXi+qt@y_80+3AI!eVhRw~&*;%cjbg?4@&j6KCEdSOUM< zIaaBamd@^S&zrZ$;>GdFWTI@ zWe+Ix*~oDI1ww6f@Gnf_Y4a4vCk=`NuW>A-J!Wz*cn^@PVI^Qa(t&Xl5gS%qAHjjVLr{UaYLTH2}t`=tYF>zB9U6VL?i!f(TeAxc-%`} z6WT?eBw6$Z(LiXiz4hZvM%hZt7>$ICoQ0b2% zU`eN#?G#C&qIZjrDf&5v8VnQ%Qx^n(>WvpVqzTUI?arTi?{ow9BEyE_0ZtR=xT<0* zj0`)3jLar>1&6-#IeodglUc$z@LXeI3DYe}<_^j^#N1 zi2PV;$mR`boLZJ1Ho&m;;E&he)~Jb^KgT|NJUG z%0L~Z`$8LoGF(6+imu}ixGv`ugbvbuUN=cjNM1-hXS<><5TDq;XdUV2^hPuZb6xU9 zg`V9ss){Y=Z1+ZD(ScFDYUJ4F0jZ!nq#deVlikvpI9yD@^W>T7@UuO$yv%vG`tOC_ zMj+`H1zjcE960Ty!~`_EDF$kFvZ-i|?_S|cGBEYc!5AUmbfGo|eb*)7<|-7~XDHSC0FvGZi|F8j+}a6U5T?hZRim-2Se zx8~OIO9L~>Lve=kiU_aYRb2t6B;PMqd6=6iI`7^fN%uKTfSUwb<_m7gS^_)C=D_W~ znc|yX8FNcr9?k2WtCfzv79ZHhEwGWmueAVraze3q#U{vN1t6+1WV`2lBY@s3jjSXE z4h$ejlN}W!IZiP!D;=Vu^Pohd+Zk1GkPFU`E2|!GEeY>-M%A9fq9U)o!g|P`8g}Y& zPn8@4_WO3d%4L%10vUF~tn79=70Py+<$2;$ykVz2&HyRnVo;@>A5S_%vgr;QiCb*R zXkl~s*aSS|?GBGID?PWV=o`IUBYd3vomP+-2Zj#_432_NGR177$VMu9-^^@C85tnz zVvihp1EZCF#P10Mk)EC~!-c7xnpoh%Mk8)rH>}LcbJX{oSA`mJ@q@#=Gf2@`izEjQ z)YqF}il9+b3&Qpal;<*B2D%d@pu~OUbh1Phw6l>VILe^jT85XxOO;O4uLJWx|cVMokVRH!z&;@Wm?Xqy8xLGU@CC7Ra(V zLydjIIs{p2Wf@uT<<0VY)rg{)&EJ2Alsm9>X*EI2If?;y=OZfm_G?$gd%?RY)o179 z;Q+%0liV z223z8lPudHiS=0FHso|yYPeDhRQo$zQ&bIf4R0!EEmqhVcQkE#lN~nRzjD}ZuF=|b ze)m=($@|i5F4mgZrV@&Q-0UJMdXUaqR0(O4NVA;ll1@J$s5e&>+9JspsMAMPCoa_j zPeRh{r5fK9FJu6L;zbZ&amOZl=5x%Hs)SONN)Aeu9rDun=DQS71yn4hm0gt72M!Bu z6x(5iv@Zj(W_r)jF@nJ(0_?+oeoOIqn9AL6a$-f znN&2E`k?ku%!?^XHAYG`s>(UTPD#Eu#2X?zq&p<7fvKPqSqA^Af>E$E+H&8xZ2CGyVIw&#yEmMPRcj?Z9a*7AO=Bc((H|aeG`rFu?=X zPl$hr-aNnBwMEh@O!8Hq6nD|}?#l$L+>&QZl>uT6a#m2Ccw?WH)4uLzv?IT1CGV2N zvDVTKoVc>j#OP&G3{b6XqoPrUy@G#7mMS?7>I?{3t?9?sZ7h%e7%SarC;yo3@=K%D(0>2TH%ZHM0$Pou0-(1j zrjsH%D!P%A;dvlzP%Ie& zsFP(n3Onx+1COk7-?eRjki?=mpK( zvcP0m3~OZvfbtrJ5~1B$=d;ekiY^EW>-hEj0ippK2qV|3-kch zq#krmH!u|v1L=0+({r}|S-%sDKM>=M&FSi-$>@nWFu*M7i5a|N)sPg%e3_t2XwY}p z10Y$}E4wGvOY7(e|h|tSu*L(+Ol2jGF28P)t4rrR33T zzpa-hZjQVja>uJqTF1XC>v2Eozg3h@$9Pn8(`KDgX;fRhS1-r}*|C*#@duZ*P&dDt zdk6G$8aWmGyV7cIlBj55jdG1(cZi)_QFdSSShJJuzTDA0eM#mj|I7d156l^CGfeQZ ziDEWVB#w&Ca4vJ}hj8fa5cTTdI%!*Aq5mrPTF~sW)5wqhw#|;M1cljC_CI9>3dapp ze9f!+Q}dcL3)R98zHt@y>FQhug6_^w?!`O+WL}qH<+C=c43_YHHD8B{I-X z*dGp~-hmY|9M`o?>0x|C#_1)Kzom`DvvX}77xSdR#Oi4%W*bF5prVn35H^5fcpZ=q zlPrVuI%$q?Nw~T!5Lu@ldSJ$;Mui;jBaFvW0(cfU|6)!Y1xLFbo{tV`k{SV${=$F! zfh2!v08(TEkZg(pHtcjNx^7|lER9=}LS0D@2)pPm;jnn4?|m;QC&JYZ=7irCG{G+Q z4hc|{E!ZMbH#o0=!{eNUFuZ8z>2o4D83BylKmr{%ZBTiAwG2s%B-sVGsx)$xg;2+f zkSrZjqAm)N1|cRu3x%wZ4&F=S4J7dSN*j{2bDuQLF&dKopS!OnJElWn&gg|qA;my2 zy?~0|E`JE2fLO9zpx*e?=BU(}7f3s>!eWBz(5vmhSh6D$7K#{AZ-+=i8;EXcY=!u&KK224Z3clhMn;DitOenKA0FzUy{8;3;P7Q9Av}9EKGHrOP(Mj`0U~-F&pNnvj*7a&rHSnU*{#Q2{D=x zp>xzuvj0otkDN8Jh7}Z3N|B>f^i|n0MLUF3b#!mUF=vg+u%WDOr>hh=vBbG_(dWar zL_lJ{AU9yy0ok3tS1cG{)Og zNGItRX3ZI1kO_S0xdBPO$unZ9L$_}|hm{%2qNG6@t+a-7G1-0}>WbL5Rj zUpJbVd#``v*QAP_iE-T3KDlOMVj3vs97R5+qC4EzlXiZW_Z872UZqdr+yq`55O?=O zLe?I?i<}&8KDSJ?Idr*moiJW`mefgZxZtX&0emtbUguWOw|FN!a22#U6luXu06AC| z-6b1x`iR#D*M^+d%GQB=`w?=2T=7okZVqjZJOL#sEi$dFMAYpWABg=X@D7FJ=Q!?P zK3gO2;k5*A4xIopCjmEW%@4F2A6jGGRlKqLZvjS7{pGLP<7Afu=TM$7fyohyfy}-_ zDjN6gj44c7S?R1a|C0*f8@=v{dHi9_u)!p8OjE)cFrFmkQ zhM*L-5FSE*Tmyt14_z_c1(rt-UAF`mD_}aqLtIgnLXFx427atB-S%J@>s{Divuu0D zNOpLxesj>>$+$cYY>zCsJjKEWq4uSAew}+fr!l0P2KhRu8bSt!JY}b6>FgZek3_3^ zI_WtDTAp@3vTk1t(Mgb;(9oiG;c2EZ^GU@1DxJNboK`(i)fpdR7(Ed_NuNQ)%Sy zCXzaSMaa9uPB3%9o7`Af>6O8QEhyY{>Y`hLyKIFZL!`!j6Ufn*H1bAwJYMB@4YaW0 zIh8&&l54@O-mTu1bnXkfkPVFb9MtNz$ZBDUubZ!x-BAuZL9`%LD{Cejf-{2Qo+MtR zu6HlswDWUSXI z>ZZH=+WAFZr%5)gE&#zZjc;S1y51w5&UO7@;bwpJhujajJAAR;dP!Nvyp}7GW z!AR4s;S?+4TpEP^;L}#n>p;q^j=#(8q@qG~nyw-hs(4P;qFZyX(Yxr4FZ6j8gx;7} zOfGV|JvH)E(oa0&MH`@(HPicFLJY2TKT8sLhW6LQz0xYTRqmJ~mI76fA8_MbdZA(= zkF(J$9=!5xkUc#h+2qy{+5eIbin38TwBEg)j&p$`N+`j^_teOd0}Xq?>!{&pxV~)e z1u%B!1jj_N)A4Qw=Xy}B@m>YL(6dwjIW+P@X&={N&D3wd@wZ=o-w4ClKODbH&N%SO z{+7wgzLjDwQKX5A&YZ24WzC5dR7X7m=jf_iCa%tDrCj)%#8mYzsk6RRu5CN-gt=7_LdqSx>WB`upr$^eZbuqMvo0=|f%^nuc6) z&IxCYJY~*s&=eE1j*N>nkQiw@m9MNI@!XMj{^A z6xnCUR&I<$(~(YpB0>qYgPyUn>zp#-$*@#OIWKQci{BBb!O0}6BF=do5uYV11fRYH zQDlfa9|CiSs$8u>ConXsRf2~Us|=%BW|)bt)na$pA){LxVkY8J)7q^F^xD}5e$wMf>xspBHr zg0g@NN?kIVGz!^R^;A;%qGZ`BPOD3fbDt}&P1^!Bs(sv6pbGWgA>7=&_@cAZkfLkZM!viapOSh3hR~^(S78Cx8LOS4R-HddI z%?&FKxf@Ek_sy*5vkEFt28O1fwB~FS&Yp7k%i`bd`i2o$C;s*FXQbMJ?N7T2R2nG; z=vmJKgN(SJbOSdtD&6Bg&JoyL>Vl+C%n#n>mMqH%#`5_avMOGS_NH=H*)|nf?%FUc@=)G5#=2;X{>4Ww8kAe{}yWIk$)2JfqbG+hY_5_mf!Lwy6TjQ*iObZFctRmHTU6|Ai4&zc{x2b#jTF zy>j53(g70?eo8SN6xgun8wYi>^zTgJ8jQ(h0 zDR|JBY6@GLP(0T~KlIoW(j`-Of?_$QD|b8ZROLY6uNSBsu6xFEiU8CJ&Mm_GQR*_q z4kkmC4E#vLPT4?pj{Ve!AjZ{Lk9Wwc5v!kQ&c=9ciWd^FpMA!sH>WEw# z>icjn^%G&Ht43~Et70c(qjsvGIgW*U0+x8;cu3Tdc}B1;Zp(`!`3`KuPMW~yD8+!V z`9UhWPkM{D-UY(~g$fWSuO^tU)ge9Sg%La0kWQ90${z*Y4MNh*&)xM^4{O~o1mYf) zUP$9}_sfttTK+t2frAyyq1F4jmpQ1hf;^oUy!St^Mrt(_0kfIRhCSdg=J>L_$jVYp z`~Kob&nb=IDeOA1g5*A9v}!dbfI3Vu2PtxZiteUch1tNvl);G?_0X{EsgL%?2PXM$ zVD?HIJhD~k^v(GP=>GXVb5@W+?g!yIH^?^jiRUHDD)_a)s?;Bf0+Snon7aFg{1<`O zRZr0NDb(!k<6*_#yJK4kmZLS&+?Hq>>C+akmeC|i+H^!^M9!(`zi%dmUz$0t&ID6c z6jM%-<5V=3f@|bUyWIQd_s`$I5FjN#9^Ou`5TtoR6{5>hayRZA`A<2 znmJ3WgR6rlRg!2+cs(&cO#){Dd)>~P7kfTgV&lNu2^K0-H+h_QM_HZ$5a#KdQ^`4| zi1AJXHNkrCPDre{GA|B@->Sk&A*rDo=B+f3-HL-gCE=wD+aL`&W_oF0Ea$E?&13cS zZNAx%E05%HU33Y3(8s1W$1rZTfyN{-VBt~AmZvo5`QJWov^;*lFMXGMFdYPNM#pA$ zQ4G+Q?4Y6zZXOVIC;4LjTZ=?57!1scN{zgo>;Uf+_`|>-)fWku%JYE&33Bs9J>u1_ zokU$H#L!Kle}erYzLWj-Eqr;CzHIs1AEWZXaln%cHLsLcCd3-03Yf=nT?$ub z8CoBWDhb;G$^!jiZREOV>YSnps*XJUtFd|ro|wFi7oPqW6KY|32*xlYjyug?57juE z(+fCkS&Rk3fOHyn-f)jjW8koP@WC;*iOQ%twVAyGeKr36#TjdVaIyK4y8rjd52{r}{JJmx(PA92{ z8kB579JfVg$fYnm)D7I!CE+Nllkoh~T+u#mvaC|E!xt}3YGFHi7}&vK6f!Iwj_!OC z$LuUi_8m8#`RayEHH+iUqLvnE9b7C&>VXP_qEM_SSm#pVQWgkRy~AN0Gny`e;#Sn82Kqwuh_kmcjDtdi9xmEQC0?p`sME`KN1^JBK^l}jbx((qhyZ> zO42E2J4I5c=qkm2!G3{$-zS$ovT*q2!`zJlYCZ8L>He9n zalF#NE!_7q1S@#k!&eF(@&5SuD(v#oTex{#jXZ|AF|RqYiPJ+T2iNh7=uM;`^tStX zuUKWnEYu#|=KHvVoRGYb~oB*ac-P zDc+?n4?vMU&+C9_kj@nS)nLugy8Z9_$4A+>E#Ac*mliz~YhLg2Z;nK%1ii(3XPg;q zqA;`M*s!~98I!DFMwM07<(p@Len~tV3se;C^ks5TSVZrFT;g3}#}_n5?iKC|D|Oiw zwmjl9em8%dA-DXcPyCMcFq-&#Oyo)T|NJ<9iP8Q%K^BS;jf5Xi%g|YDD|P9!WQOJd$Rr zs(II7u~o|3KC_w^<8_t4-MPVIr&tTy+l?MQb5=uf?B1EZ+;;vsAn4jh*N2@TmlhR6 zJ!V_bavpFkiOz|JoX)8-m?5V;zjgD5oD%(j=q7A%qCm7BQ0y$u_({s=Ls^LQK zzZe0vV7YD{X?5TU?STnIx+vyTigZxXW&A-&vMfc3a$umOo-8wPP9*tW5o5>&^E~_K zr>J^?ye|b(rm(phsD@=aw@(Qiba8Izp8&?&Ah?qo5a*!-O;bZd+;C2(54?~|A5*My zhXA#{BjATHtjyc_Wy14-ky4kBoV#RtQfth#F(|TMhduDq4A4Jz1?eOh3T=YT zCGktzmt2dAUsCyY2mH7F+c%@uEItp=#7r$f=Br(=#Vx)TRq8tQ&HGE*qIA&a{+n06 zvH$HR_;1xa=U+SZvyyKnzJ3?3UyfR}`2L#@qVGmMiu&M3XWzaXmG{=l*Ux=t_?t!F zY5KSAKUwyx);BX3-}qi8w8~r@v-pD_U0HH}$<3%<_*|KbG10#2>lr{8p~tO^!IQeY;>{-%e5tq-<4D(Y2z6 zS=Hj>!hI1vo=2RU^{f!P1%(Fo2mZ(0O1f z^&Crx7#V?aU+oE-ao_kUH!Z_vtUb5iWq)+7u3m70#749U8|Xg%VM(Uz(K(ogn8G@McYsjiNjgzfr9?v1a!wt5 z3cOm%>jxdXO3t#m)!?ULrqDL;H6Z=C-%3v~#EJolgJ1%=EUm zb8*K7+&~31fU*bz$|8!gyMQa;pfKW!LELZ<6d6?bpOXZKL|)Ac3BPE6olni1cX@z3 z-<;<>=X}riTqU)1wYy#(C(`+*I-^DgNI!Wbd%za~ShI(~r%dEvciAF1uk&?Sag<5- zet2mto$VZW3&7CX?v6N*q(S4vCe6M#uNtD6&`wEKCy6J8oODI6a5t}aZ1(8|X5yL4 zZ0y-H58Ls?(y#n@l4YkdhfTgiPiIvljt1RsW!xybeQbl74-Bo#`TRm)E66D*w*UBq|)tn}c z?yZu@y+5~mBEkGqv~C8ByA|^%?fZ_^3^=aAeS6a;H%mD{hqV_B0)Hq3c?SFp?Okc& z^bWNaX%+@uM$DjJ6>su`rZtdvH@0LLH$%b(xI>$5`$E<@IxFB#+!E~fearQPO&Hp7 z*{+0P^~n(=tF$?Si{ch8wDV{-H3VsQtDHQh4bW(&)>;USX6&&%6S?9?pSziM)W5k% z7LxT2yrYKhSHql)G)kePNHP^=TjrGX%@wOyM|;&BckZToU_W zT?|6?So<<3OjBa*tPdHD@ml-X8II2~nifk<7&-pVw=aUxEWYC-RmDaLu;%+7<9tAThyVxK^wFuVgwBV zT1;1DK&hNgeNvS?LLR`1PnPX1qeBSWpThqB4Tlo{eJuWm10#gN|Io{ef_AA4U`U~J zC6~mHLDeMMCC2VX;@Qcu5){uqXy-F;sy?ZvCth7_UKD=6W82Ea3=9YPH78Qe7eIM`*a>M$$&19c?rJ$C+ zB*xsxVR9+_=InaUZ0;uRCO@pSMs^))frBXUv@Fg&-uVUPQ-JL!?~&b*K9cmh?R9VG z9|=STc}H+OM=xIu%H_zx1BrRlC=F2EkSeN^VP@ru$08_tDWB4!!Ey&=Be&65?4nbp zD(`!4axJ#5ii^d&BMRj^{Eov{`aZof9Iy0#1y#%HK()O^ zvjy^7@vkojuJ=y2b?!$)x2vP6wVbuwIA}Rg1nJx@k_><3P$sJNk=tR$J7mb3yb^Ga z7eDP`r1ARhh*m|kpj$cM)J*!IonQgCca)Wu<1k6q%*iwNFi+A*IbGtY;m!qTdt{a! z>KyfV1!=Y6P=DHH7aQi3&fldff?B71a!Ux?;C{CQpp^!Vb=~_Uc(h1!(p%fa+b4KXI!Bp)E_cnTkJJNum%ON}IUzW3H6;hZtMJ#`s^|v!4 zodgsy` z;}(Bm3k2(@a%^`p17U=%*aY8Cj;)$MWJVs;C6Ij%oL4?;W-@9i#bJt6Q&AnzaxKaK zoUD!h{}i0pUm;jfwp+PSDbT_NlMb-L?ZC<#dj?}Ny#(Kl%a`bz2iBuVPuk@vfTZPLrL{o;bE zA=H931fzu>TO=`C7ltdN1Qa_YEm#x%?;$26$^KirjqGNZeRklD&nYw9RZ@yFij*4N zmlnMhgDb+ghk2*-*l^HfbD$So1rrF zOS4;k7XO57AItLNz^qAzw7o&iFlz-bvL-)50a9GHY-^( zJA!WjtG`cjLRdIIlZGN%?1~)E?E?|#Dp?AU{68aouBd{QMmF$@gcmsX=SB;(y8?10 zZS?11@xIG7g|lF^{j(r&rz(dQE@K2HYrHXdHP4nw?F)i0lVk^W zg-gvAlwFhpHclCkodgdS`A#{4Y_T?a?n7BSJ=E^x2``QZk$j|UqjFjY-3W!%1)Nqz z7TrzC)amL?ZVm9XRxdC2#^xzwo@b_682kJ)jXbvV?Ci~guuF@+VnWNiH_s%K0}hOq z7BjRQrxdjmISiUKUh(s4=Oy`K_Bh%13VlP2@p|+;^SE0iSjV^5Z|H1!Q2ac+uL&Jq z*>z1Kb?kR_Z=U!~=GU@+zU_5vdt2zg7qS4RyQCS^eJFNr(Im)g=_6s?1g{(k#8GQO zQRaWH$#;XaL;>i|uXTwZ1CM6xV0zY!3_CMwM6m)BCA|IMyOwc(Hc2A~Uc4Bx1RcS+ z;Vue;a?;R}nMKM<5UulZ{E^Z@Nn2CXrmT zudpr8qvKZYJgldieMdHWZ03c|sfvP#U$LAe8s7AY< zU+-%&8M%W$DIuw^Obn&M%<}D_6u|e-p`xxqZF-BQ2&h}B-leWP=@Y`Uo;y@hW)}cj z1F+$CaHL~!adKq4S;(39CT?%mQ3SqM?Wd&^Up?^6Mw)^;5eXAWeG+tahWBKn)*!i#4ou1^{ zB(iAYaP0I@iY|(DQc>p>i#)nWw%;xF5l)2=sgP@Zt9>ej+0$Er>~lhu{gt?&Yr(b9 zekgzDjwy$E)jlX^d4Fyeh+7^Bj0;-CTgTJOV+45R4AcywlFnxY`&8iw=_Fa+Jsn5c zNvF}8OgM=tx?(a;vzKFPCK zduEbtcY+O!ZTC}OJQ6dERX<8CqfGYbiQALYWWWjPM2Fj>I7+daA~95yaf@^`AWmfL zOJsjB<`MreD?Z~9pR@N#=FfN}8iB(yy5mdl=U7tb7}#hv0oj~X9>&uTs`20bvqlnx zwD;$Z<6D6VyVnVi^mo1@nArdIuWz1eG7{g}_1pg>TOBxbd%(l2~zJr?^ z7VBZhRHIcmey&e55(>wDjvbDFvEL_ZERAFx*Lx=njbxkOsQb?4w^|~HUa^MlAzA;% zM;m{(;_JuXhM@j0cl~6`J1vm|PAk4~|Mli?VVEEKcSY=RTPp{@JZP;v#lM#RAs#ni{A-#~@I{Ar z;N*-Qa?iu#=%H8m-CyqtHrbps-}Fr+|0SdALLJ?(h(HCUI7pHGRMg)-z7~u+8+U`X zm`g-S7o$-^bttG0n%ly`^Mg)*{}=_KQk@Gq8R`mpuTQ=Z3=)v_t||ADjX{~}b0p3; zo73kCDN5{z-oW3-?dKXVWC^lyjTFd)+`#S z02hVvzT=c7viNhX`jjo68IQNjpYlt!CEEA?;?FD_xH&A34&u(ku==!bwn5V;EL(h> zr#%>OTHGT$JHIC2xH>lEusBwc2*uXz;m4+Hb7|Bo?(-;u%gs<#s69ry#ogiAoMeyn zJZ*=|dABim8qY_ZRTzGbtMUBNM}L(ahW{#V{)wfEsl!5rL0$!%P@7w3P$!jN#>kFQ(=5*VyfBFbB#R?iDJa4a4te`V%;N zDGL-j8l~7Y6cUP{=rG_ECu#(>xzvy$ie0?#J=bw9a%BajaUoV1!TC^__0GSTypGnl zZ)TF>k(^ZrhQJxKW#=%ZsHR8-6}61FjJI31j;94}7Y@$)u7EqPw|N_bVyHF)(Q>Wt zQfTkcDc4>E9_<4$78DhQVSgEj&!9SsvAXfL=LxdRvsH4=>xdVq(q(!)k$2L|JX7cv zg?5iihq?_Kn#75;>)k4R6XeI7^Fp-;==)*#b}SAVJFq-wZ5sK~sLc|y-Kkr-e2LNo zqo!Y%yiXEe88X8@Gk|4N3Rt~%P*HKBZaO0*Jv2>(c4duwl3+>18jrg`+*%T0q$cD- zS3smF92VCE#BuKWUzJA-*6@l|8R}9<6}3w4hgFQa5Y3t?921JH`2nN6XWg>Lf0=B; z%8mCf{)4P!hZP5=P-dB7C6Q8q3$~GpO5zkMvwSk>2d;@~ZJUenzd@%IdRxG{DIlkO z$#1(i*2vn;$9-m|SYqRu=k0ak|NK3rz|Vw@4|X43OLnlsh668Dur3YT9~DswP|D3i zspYm{>?H6I3J$dw#r=~}rso7{mlgZKi4xC;hJzcN98nMVmMhe{;ytM$dO55$#+$Aw zjrU}WTP5|Tl38Qn>{!;Ik!F}BYDQjtjAut0e$?mq%?Dj9S04tmT^860COVVWVuN{u zZUOfAY6k_$uDJ$YBNkO;(Ou33%3>dgujruMB1f==49L)lKyzL>FAi#NZ2i%G`L~bx zL;i9on;tPQ{v*Ux9M?wI{@0#1%hs{aGEyA(_a_VuVsiu_oxOqAE@_n*)cLXDW3sA5 zwp{@Q@$^c0vV8nj7r-4G6x$T0FBps&#oVr0QGYfqJuxfe=95kb&NZ(#W9~hn6#W!= zNJaHW#B&>^&`>O7C1m+Bp>Qlg-r#&j*hxdehTN6b1n5)-!F8n4RR$(|Lif)IG;HPG zt6loSdc(9=gAe#)12N1JrU0!XT7a7&+!?J8%l4~O9sqV8=%}X99qQx=oeFO+Qr0b~ z2*T?37*gw>&B>r)1En24`o$o;ccpuV7zb)K^fJ>y%DV}+Tv+9X)I7tmr&QOJ$QWF# zSmD_WJF?`6R(X=gc6AFJVc*cy5bX_VHm6aFe-!>n1~CUVSbN@@Wb13pSR27&PDow! z(?^zlBN*Bdr^sWDxf6W$FZxEDId;i}hVHO9$Fgo5J>RGU@KQjr=0Jm7k0X z#>35rEigO680=uI`TZxE-li2z9lNrUYD?Wn_w;gCsSy!t? z1J;hAlAdXDM}?ZcP;#Uw%H<N^hGE3Q zd>r(Bbt)XeS}fTGFN{Ef3+Aw3e8+AB^$qprN0v2e4oj+HD4xrrE9V&!8@HhW4Jy0; z->IIn1`v$1X2%Px05P#mdNIHf5U&W8iV;3pQ(B#{fgKPhCKVmpd&y>SVfN027OAj99qn@V>RSr=vkC9%TR@M+WGtf$g(dMU=4#- z+bS#N{612_-KIpj_IeR^{(ItiFd!Z3&FV#Xu2)?o zLoq4)n2l#8-O@41%{z4&alWBcw6`9ZVK58TnEeojS~!p z&!KgY82-fg9N*3q`df0J*g(R8*_8~esEhJdvPxwixhk*btaWR2L)HoA0kt~7-5hhA$)$}1C!EjQcr?4}g3x@1#Pd8Cl@ z&}#!}RP78cmgSN-QJ*U|{4EPzEx+e_&-MDe0jHaDie<&JdQao~7LATu6wn~eav5++ z30O6yC;;!-&C4a%=cV|c544Y+VS$(-uhjC@b}$(>(ucp2hw;+@fy1*YVz+_ey@Me{bQ!vzh^?#hgpZCe1^t6cqfE9I&&fO! zSW@J_%prRn7%ZQff#nFL01=l;DyjxLIJ9W)Prg5SuV1#HjejWeQ_+AdE@+2atTXg? zpbM4Rf+UaA1PsQErGDiu)jpNe4trMnwrHyOg~~2ETX4oL161ml^J;-@y&JNO7X!LU zGxUbIBJQ2jJ$-<;)_1_EQy9l7cWIZ`0@E&EcE-IR<$u?cOm-`k1Mj+^tK%>lW(lPzqR1X9>L7HkZ4e<2K5oF znyL{si}mtVoN{k%I^F5t=X#Yi1ZkmJz2{?hsAP)^Dsn#>0KJP6-UgGc3F8Ic)lwp>N({&l}P< zei2=)tRhXCK2jh{5832e1>Aq%H+$HQZ`j$D z14}C1{$R2ze_tE(8oA-X>(g?x?N%S9ctDYEYzP~xNb-k5XeFe7jGbY>kY&(lF&=qu z^M(XuzDM`;>+%|262Fb#9I(}0o1sRdP!W`^PV&ehokE?;c=@h0YbFfROZp=X@(jka zjm{v=d{T(1r$JS|@O|u{c2Zc)IqsP?zk&wZO?bMW)_Ci(z%u_!;RCXwFf2tVS2U># zIZLI>0*?fo_2_c$muvCF2Bd1 zBQfZNdV62UazxOC)~zB4{+gF?qcxIiw9AX!vt)EStYu>8Yt$o;Rb+E&!E?rtG1OXP z6&jYkp~$p-b4t&e5bG_{v(MV|3@eK?A+1T|X97kFHd0gn_yI|HWgKAecZWH^8I%GLlrp4PpCHG*c9y>$ zqJ~gJ(oL^`&M7s4s28I>ujrBEx`5@caU#fiqaHzq8np)&3vvV>+Y?mQBWQ#TlpU?j z#;Dm}SVm_Uf*`vDmpB)NY14Mn$7S6F7v&0Y1iHoXzCH4?xeyGIw!x`01qTJj$p)&*f;G_)Ih zjVwAf1e{c?vqzoy;_3bFU@!XV^aa&W_?{rY0mX?oLQz4p(@q767AP7GaxI3i%Uf2+ zoiOiDZC5R~N3Te-n_)AxP){=GtMVNI3u7ph_f7?Y?{;yOUyO@3Lr_PzO6~x0ii4c0 zMS>O;^hff6P6T@S9cLqvGsU+Kp56zviHd-|BD_+m9E4hf*bwb*?nPl63~FP^5{m$CL`8)ma;h4Z)VX6`%%@( z<`F#sM&~-7vC$LMJwe-B?M}CTHw?s_~Sb&0y16KqcHVhOPk zykpa;v0}>Nc{=&M|B#wCP{-f-_9e1m5`@`@?>x3s3h0%zm5NFcC98^rt&;BWCe=AF zEsA9HyEe!`@c@X|sUmbT`d#(%`(gK82c1y&C)q0zwkN4}B3TdrLq_@O4j;lU$Lzkq z4%tDZfi8fi0-XHj$Ad6p_FifhMG2B2Y)>32* z<{@Jh12RJa$}aK2fViM59?La$?4tFYDYOzU&l$l!m%jO7(2vJL33gzXI0GfP%caW& z^?h_IkY43?3Y%27-LdWjy3SJgpB+O&BTb=|zyQZ$q^Iq7?yxBL&u^Hl#ba*TyQIQ_ zvx^tZzW=8w#R-bkfgB?At3E+L;#}9{)0w`}f}5U2(5FZ*{{#+aKO}hhx~7}ASa3;v z6QmHUc|Bp9{0c<8da{V@Z#dxlp7cD3cGTyyDYz=Sd&Y`nC@>WHj`_S1m z^4KtPdC=Xdcy4sqvK}JL9_b6Nj0#80OU(#5L#pfFX10Wy03~#e+(Gt?#29v9dve+g zWL1;`sO1NY{T&T;YZzJ?+QICCC<=DI!2~dtuo+5IP;v}ddwyLKz5Ef~>mNH~(8>7F zBYH5P+`F0+7t{`WB6Qc+dldWJo)R-7)&n*|XJ7=ZTUf(28me{(ql^wp!{^yfup0g& zMs>^#E5ZBkt{qE%M+e?KGxT>{<@0KY6v=?6R8_znJ-;B&2JBEf!E4Ox;J4=d7;MZMr$q&`WbD`fXoyJ`mz733*7WsFnGVE;L0aRk0s=%ed zW$ZSG!61wkD-6Sq`SgKn{c9~15gj(klxMa+rBDix^hl(lPIAkHr@`0xl#6Q42gA=V z=vM9~4U+31B=x`*ubp3T%d6jY(bR5b4Ij_jZJN#IV-$04_Po8`FfTtNc{S;QS8HB& zMcy|4mT#B-zT724odNlr_2F9rDgrto+fz?#F(1?c{BOLZT@kWvnyGuj<{5PXt&%66 z8NSz)>%)qtHwfd%K8a4%?`kLJ)F{wm>GIkIFs#rr@%Imof5+YA1OKpkTQk}C%7kE` zN9-^r%nnKcF_~>t6ehP*CU?@wzTNc3pnk~c^#e09m%Dk!WA{$FUDhuLVy7YR-Ah^^ z5RIiO$N*26Y=0{>N=(@WN5cnoE*5=cSupTyB-eqRS%$KJM(Jhw!GJb?J6%O*xSxWe zlQMpEgcduk>E-Fzkzn4|koVtzUk|KIY%I_(-wCQ8HKZ!A9T0NBAG1MPyjTYT-!Csg zYMNi-S?`GiTU+dlhNfba@US#HL+^)%9rG)*$$6KnC&4E3GSs@94duasUE7mp=H+Qq z^JTnsyiVzWY>f*_c(+Q{`sRCS`+zr|8(Bf8K&cbDvD(Bbo4k6&EBwz0H_eO*TX`No@V2_H) zUHsyfs@G3l$6GJo={Epf7j5tvKIg9;!^iuaQOACX`I|eEYw56=#y`y{%yBiDnbNO+ z>kL^v2_zMUFaODuVhaVc0Oj;3!GeJ=3}%^>_EEga@SQ`?5@q-M$SF6Qe6OsUBG%}z zYSgjc#0njbyZUc)ORId_gpP?lzq><@IB??Nni*`*QVM7%TTeyRsZx1&r__rsdfuF~ zOMFJyAl0d|r?>hciK0b=$`H4CN8PWuq)uK-Et-N=)m3z&ssn1WJJfr)&~BJh?SoB+ zV?$E>@xN2x@#<8)VfT3bTph^p;?p|L-hefr$(pV1o~u*qREMT+_bm(D7JkN{D_=AH zTJX^L#!*w~9Ijq|=&cIYBv96XA97jPzOwudumU{g_4gUH#Dpv1J>h?mb+63cDc=l* zsgzuCO2Go~B*yX71Bd@lx6q-0L+nq2Jnk=F&i`R!YKpK0e+dB87p?z;OeRVs% z7Bs-ux$hA7{_IlZcGy?7XwZaR=BG^Wkj4G>3Y>f5*CFflx33q*t_h%-t+5L_z8utd z=~0elt2H+JB*$HR2}8%YRtZw2+W2wgMrZ~^^s$5%5{p?gJLq*htWAaP=Wr^QTJ3fO znwkUC8(IvE3m0N^w&csXwAWGb#P2<_tYu__5eME&GgLWN@oVYzZUxXRQoG0fvL9}D zVdXiXUI&4i@$=42^JzUS>;-_4$IY3z>7AFI+K-%rb-X11Pxgu`0#<=Ly20g~*NUnC z*fjHRnV~J3Zgsah4jPHtdv*Bo1Gav$4XebL58C;NorC&)>B9A65nOd(?}I^bHSvd+ z#s9b_|DOlY>Z&5NU6MBcVvxOr5T85M)&c0bL)hiw_kH$Hon_jy-1zS8LXypH&*H$4 zs5RTOlv0Wk3Y^ray+Z7h)-Fk$j@{9!Bs+-CH`V2iteUrqtaRQi+B7W&@;|rK7eF~9 zb6V##-IR?%$CM~1_1LqNen@uFSuSm%)v)u47Hk%6_pVT+sVfW})><@;&QL<_xoPHF zku6P=b=Wa;PnqA?5Jiidmz=cZ^E)i}fPuS@ZO4J}27(&EIH?NEo(}ccoEw~4P8QuB znlKr1J#j(3q(ZnqtSP9&Pz-B0Y0K#yX+}mp>#4<(i4xBwkIO+ib(w$VG@K24f7Z`@Jl1~r&pc@7 z6YM^GhYilp|1^SDm{8&RsP{bC?7-pIy=HLDq7=Z{PottxD+65cB0m->Gmo$-Hrc*72h^!r>VAevIVV$rW zkl8_)AELbSRTEP3rgQF*Dt69^14ltFnc=E|Qkq=Y4g`sN zZ3UFLa!26@s1d0+6=U?Yjmt_n^J0I?3M3=iA! zG&`)8-1}g|f0)*#xr^FolZy`Q!9dr5VGegErMN|r4l1fLu+Q}nqZ{3VlGcfRKEF5Z=-Ie0%b5dwfBOJ=_;nLIZS~(@wttbpD zfewq01VWE2)DS}!)7{{6UVEktI^mcl;pudPbA>QAT&KE5KXi+pYt#zwB{%2jVYx}6 zPrDB~p$aBe8X++Qi$vOJqh7eA-CK+2as>DuYi4^uedPei6O9-EwhnI$%ZklTKMP!T zc>7biV%Ag>-oCTzxBp4Dvh#x+I7oBA44t`@0%$s!&~bQDHs@gY75X!;TDm_1^fo-p zgZd;D!d-$w<2WJ8oM%+X>3~6ey2y zVkX&b|2S4~JqruloZ<3*A+cmXIqZtXK+7r$OARR!riifaE1jV9AQmVNIxU;A5tuC- zgP?Z+s@md7yUnlVY_Oeg%jfO&o|Q3~7`c9avt>}`6-gU1#Fmf~g#=IB=Hzm4p@Py^ z5N#fD#jw{TnZ~seLw5cxo*jNQ@6QMMrX}X}=FNx5Dt1eZ11FPrnk_L&lwvbQ;$f=| z@qIKO@9X6QGQGTOZWmMs>3pAf-ZJ_?Su|Ff;a&gq2If@NALoumg2RF1FbonLt&*j# zH=zt917a&^FxHW!K*K*f{{&PhqY#J9H}TJYBZr@{(rTc?I{Y!aAFueAH+FMPR>R5P zX)#&mz#E|*W_BQfQf#2eS}H0zq6fA?>1tdd3PrtPYuxXLm3tR?9aKS$#&~P3Ot5;U z?LN2WKY*Q&@vl*Tp8n5eFiJcVxaF$Ff^2aojjigB1I;x#aoVhg z(fHbn2_K#Q*|o6%!-03Q3@b@uV23Q8lkc&}`P%m`NABQeD%Q^5!)fHVXrKXoSR?=c z0mFdUvi$p9|6JVpcG}3DH%bc|GdCEV#b{TyxG4xCWwPyxY;l%oHy1nRt@W**J}zLf z-gbN$ATwVq?DOt>MG@zIU~)F(tNp)4>K)kG=rnUSE>j8+0Q(HO3CV7V_xj`u8>Pmc zLYSd5Xx&uMc-pLudLNW^tFO6@oZA{aAOrU0r(6w9m*D$dnb}MKE5rPTBb--RKzI((K#gPAYUGnS*_zs0l1tlX9fM>#)!JlBM<-+VeR zj}*Kzo8WpgJXTYR3W^*=@q+~U`mk(245?$Vf7y)0K$JYr_G^UQk5;c-?5~$&pskHx z9lo9iIMk^TlDFMaaYqZfgj1JF-;bW#Oh1(3{|!B8reAa^g)n|QU>oj<6CtG^G{c~w z1xD^5Sq;yOIrd^d#6625HfIDM{OY6EOc*Mi`PUU>7ds3&u!nQZ3_}Mg1++fdi?R9? z`Vsww=L4AJ8nTnF32dX=_*pa2=V{UG5|q8&^j3RhivRuTse#J^2mLbWBa#N_TDE*H zokN%sWzyWXBs#V>d1@~W2_3li@^-!;5C;M7ZO$po`8de{Kux%d91bD-1T zgr3?rHk6X|kt7HmIG$2thLK`Q0mb?GR8$cu0=fw*1h&&@uEwMwc3?p1WNlesCw-k8 z9nmAjpi8f!Ck$7d^BJ*OV8Qk-^WgpIG^;jG@M$6Ef6D0*P3smk6!htOyg4@F_O6!i8>4`6dRW7QaU|PbV1~!BfBfJLUz1hI9sEfNNp)b)861;ge1$!f z0#xI2z=A+RurKVo>q^K5A9KzN1*QV}tH>)r=G!rV0kr)vGBUrU1Z)gCrRtaWgvI-! z`bvo>y42t~uazQ{P|3;mfn3(`VDNrz3l1oCWtHUXR{rV7W61|Oup7gW4?>E8o?a1> zEF5>d0E_|NmX|Xvea8V8~AY%b##%a>9xktK<*+8r&dVPaZgY>J5y24e*i>R-2Cy5?ho$rHti zl|;vGqwK(&pfWR1cphP_E`~S2%)D5X#o*aQ5s$RY(VyRaKtzAr#1*%{y@5L|q#ique1Ngg+6FPn4&6b};QEvf>*(rUM6dikUQr)85aiq|QOM>&EVf$?Y^cb8=O zB#=dzQ#Q3llg&9wdSv$kl02@@g%xX39A~YsUS1uZ$-POIPioN=DnC_Lb1wVs_UNUr z$rnM^a7RE7N#)?vN6DU$*pMQhPnBcBnH}Ku6m?_488f_gO)Y+FjtNR1mTRg>?n?%x z<7QAQqZFkSDFH5}>hxEl1r@>!LB4S7^iDY*=Z1}Zwn=kV1Mu+Q>xb93PLCF}XrSB} zucP{l9XPlG~HW%8A2Wh)0y7k0KAKDC8LBg@VF(KJOtn18V%%Pbr)QQm&q#`XqT) z`_y{2L6hUs`G(7D-0_BFRpG35Zat*g@TO{?{7~a9gHA|7=}>3D(y?*!MHiiFPx$_M zI**$&Tuk)9SBdvUmDoY24D}-#2alJ2K620}rC96PsP=l1Yd0h8L0 z+&iGo3&oDzppO7c&li%4ARHw@ZY0UH@%!oY1=Z>d`jQ%(D`)#*XK!t(__iFy)3ZSA zwt|a#-X>Cktta%UF&M)E0)cr#M-bO`-0(){WuO22=V}m+g{W9Wl&&`H$a6$sry>ex zEfys3>cCmjHcFTAn&1GF*f}B;56`44L6Z=CD-^Uz5mgHFh1UehLcq7`mg=qxNP=L4 zPdn`8ZHN7_n677^{molHo%pg>3fz_UyVhf4ckJ|z^n^xfoahDzdXqaBa>k<_jM>L( z{Y;xZX7@MjtX%kq-mR9Yq-U864(wDj)UUSj54~|&{P%0}vyl&E4bmrmEgC$L8+lD$ z0cVmVeI7dBrl3s`9cr67GB281>k(sI=w39Qt*$d8=C`GvobWa|I_lV!m1LU(@7E8S z`NVmY0!S@cR8$ImQdJa|g#9eTIzW~OWsb2bM08}Q zGAs@KSdD&2(FN=;9xcmGd4!>nI>*{uWhPjcNjLb2gh}~ z2t$*bV)wq#PGOfUTF@J|Pm=7u)-#UNrOXg$qXnM__PO6PoVekdO4rhvf*YK}vP0Zx z7j35Ct~AE(wQZxulo+=%BF3qd=+6i3`mw6;fP7n;@75J|U>~>w^RxN4K;Fddw+e@~39Ofm2 zV2e%T)5YRG@lxKk;98*Mwo59wSA_A>C&4FeE8S@&&PG_7o;_ttz{!rC`rw}{FP))I z7PLd42C~!3X0Czi=^R0oY^h@WO^=lfIa|)Qbz>d9<(oa0K8Fr$O_=%|PI~0JKD=GF zIeht~tMWbw;%-xR(~l!My`lx(N}can&nmjlqezBvl8Y2ciuLkdIvZNP^gJP_6KrU z1`I>TALKJ&rK!&M!t`8k?NM?tV42rpa#7ehj_o*yX2|M^9(rJ$UfCIucfWd#YbmPj zu&`i|$!>GmE8fH{bsKt3kYAa09aP;mxi-yf(O~X4A*2E5vCuCptVNUTdxBHymMYE+ z%?>T|Uj*WjrSpsZN?!lCIP#LXLzc54&QFjhPDzE1Lz}$zx}RV0So}Bb z!FM|STQt|eV~OLOBjs;2M7C(~s8{${d__4lLVo0E>Qk+Tl%>^R_ntKnSwV_&={!lv zO?&6f>jOU{o7nB09XRX;TjgQvQYNJU6{%Dzs+|V;baku(5<%EOBra%~XOpK+4R3b^ za4s*jCLku_su1|xopz-0juBR@;bQFruigD_xTVU= zv)LX8h66(%R3jfE&!b4aSa8f4tCx|@pr4!SUCmhqsX^^xK{r1RYTO|#+bKlCLI-W= zBLn?q;D7&oz6rB^S>(U{O|{9E9FO{BBPnEOOB|Tx-(Y4-YA8h|MarnC2ImGT8i`KG z{5+8_cB`3wafoSvC1|N3IWE93KnYwV9_*EM(TfGCBD5`nF6SK4`mhH2AKFi9{bO9T zt@4iGMyPGg0)8mooUYc(|6c2#U~J*6(iZxkG@P#|r&R6y&(v|84!VEx=;1d;JM*GN z%v!VZ%zZz*`1PGa6NpwOU->kxr_MqON_DEP3=#&m|l*VOV&1b^8SvG%WJl-)cG9%=UC;Z~k)t`=~6QToqEexFy z8>J}6a)BUqtyQvhIx5kv56=;u!3Mh*g$KehR*P#5@Yu4|X97T5tlWDMFN+4nuq~3? z!RX5%_p=Hb4e4cPJvVVN8*nf{t5cl;v8Q66Y7VSOIf4`#L|$xyOY4yKqGiSWx-F2{ z|Fy5Z_p-Cw+Je`S4DdGea#S*fS`kD5E*=S!@jsi)pY@-$^)qe#n6|i+olW8VpWQ!x z*>M4#)=^s}-nWxRA2f^Za*pEx6$$jPk2#NfDcUQ6i;3rEXufaRqnk}2-+?zdnP%R~ z7D|yokqsbx35!ON48*&93uj(eKp`XY+nS(>o$QP|>H}&QMWg_0+KyXci*LX6Kr${D z7LVKQ?O$~#m%i-k85p%U?lOSph(+r=!N2HtjpOulS8xWM8l+V`ovK5iQ*Xg;Q*IoX-OBQhUw+d&z_i1inexBuNu~ocV-K6zgc3?o zM3FsIRF6DX(eH+qq1$yQEGN)mu`Do2l0t)BXq6<$v6tnbQx60@E(*0hGVD|~=(K~I zAiv}n8=_4Xz&|JumnDh!#T5w#xXwKbx*2*o*0UP-!ZyUysDbkoLFNax!V^2-{C;Ll zndQ>+QiWg{7ML?{Zl{*KJ@^&`zh)Z39Xj7c&uhx{T%B(czsxThdPmt13s%e>E4WxO zo~>?ShKuJ1J&!4qO?l$>)@~Pqv>hW-vd30ypDb6^<(F?M*_8d%2h=AGRBmw&E?gvK09s{~8Iw0*9Ppo!lq**A3+ z?2YU+_slbAEh?Btu+Noqd;XgJ6_Yda?#(mFGsg6?$h|eFUqKple6Xf^9 z(!gKocijeMg4%XEe`dK$yr@T91Zf^f|B_h63SP0#0F(;0DDp8%#*A_PM>DlnJ;Tm}3HfzO zZluZdeCNoI_mDCNj;)+GGbSe~1$5~+N<}pTUGbA^@-}`9rwFpc9dsvr!?7KmfgoU{ z)q)-i3al&=-1a;pj1in5@HQRZ;^AFsH3!pudu9$g4e75e3(SDf3}}nNP$)UJlylKV z+b8LFOOP82l`-scQgvarA(WFf6Dj|CIf}K%xQuFL8zY>ou_kbjRR}V>i4~kCG`;@K zx{pn8I{%|(bIE;nE13g3#2d^+hn7xMESg9LsHpX#vZ+a&96|ByhkUFEYLFE9v}n*x zZs2zcb-oRfqoFObo`@Dr6|6sBzCY&i5aM}P+0BAa`9Rubo& z*BvNWY~Xh(i)DRcG;Zs7Iozcj9T$($iT=WKC;dQu2}r(|!f(z_Q}(c{+!e#jJ9JJCb;kJl?J>qLF#>~Y?*?%82o zBZf-mR!OR;jgQOIli&xzNX7+O%x%gRRf7Dnq(^p@JeFut2o2RWN=5D7$7RQ5=p2_u zMn`B5(D%czq29LeOvphX&t-?3L0inYre)1+aPHw=47wL)yWFI;$QkzSeeR*Hki`l) z6aKSs#tcjC<(Ddxz@W#xlkRr|Ac*65vA$Vq1Jel^+sJY%R3f-W<;#qIj@Y1O^go6D zlgw!OecunhKh5OvEd9`H4M}-rJf3nhcY8Ob$fZa&6?I%z0i6*K`c*-l>owVFvW|yc zICK4ag+&1+0SkAA*6u!qVEpI9sEP+1m+IP5CKPy~_}x=VZzdelKkrioJ%fSJb&Anvr= zBSwHcej}AGhC+o5+$eQ`L|}s*YkiUZ1(u* zR9&b*2?~{dhk1$QyrM@Q@0$hteQgI_!H-p-A;M-M@7K~;l!P$c6tXuUHDsl{P>DjS zG1LWU41Co62HD3w#MMKsZL0HULTL87D+Kjk6XaN&5<_LEj|Rj+O$m^<`n>sg&&!@8-vG*#(bP5hX9hK~s=yT?b>Ul`Zv|I#P`Pr&B#hcL z$eM+zuD613NMk_Ny^>csEt6XVtg9@SF0v$|-t#kIAqYI*3N8#d3dLo)Ts*P^c31#} z(SI3D$nQm%Vg|?vYZ=wruw%QYX)SxVe!^LjG>Jen8cj@Mp%->7a`s!5uu9wAYeYa1kh6vdb&Qd6O$0%2{;2>!59Pll5-Q-$D7CWbS zH}k4J>pVL|c9@Hy&tzTjr-nYrK-19aMJGJ1jl>&1XOO5>|`sv85Y}>>u8S84^HLW*iXP=)yK4Z6ca^MscY?X#Bn>Q)Nb&6c0 zqDsPh!b+(gc`3EVxhb>?sDI|~TQrZqc|Nj9^M{3nZ{Lip{q^n0!nZ%l`jzghgFnJI zhO=;{{HII4QSjA2bZCx$_x{@_-?<1`RnYV<13w89?QJfQF%n>6y+NAg(xL%=q5BC= zduSsa7qK$P_=AnV1pfFl^!0i9!WPYCc{hFdyGAYWC9^ul*9Nw|}o;$1$bL>z5_GioFBZK7EV&7)oUU5v|VsLVI zx}j22Ej=W%H14>t5*DN25R45wtYG1|*84AiEOZx`)};@NCjXVJVYe2Qt(cZX( z=ZG5VQ?KW_Tpk^vc2u)7>}zVx;fp7M@ebzM87b=Ju-=L9rj&5Yn#flq!opB3Swx?9 zFAr)1V5iVM@)~cb(~c8mhSUTu7C`ed(8(+WvAXf}wpiI8Mm*a7LTSlA1}ytiy&}+H z=vtFUNd~D0Rg14AFkgbIk?B<*l8bcHV(>G0v7Oe2^(BM(e^rPGs%|=nPutSsM zvuWBTpwR9IvTcWEi|~?ivEr_>LvuRdl5z>pa3sNh*ZLZ-*cc=XyEpOR|O1iNtGH4OhDvIw1eBgFP5z)^#Y8BWkHYUqdsxGMPWNa z;yBl##KE#{T`5q5U;A-2w`^b83)>JK=H|P_#O0Lh#0-ZVtl5|S@p_epX#cne> zqGWmZ^wn-9o;r-6G)nPuim227?t@bf0 zb&o3IG74-ku1bs-Mu8w!re?y0@9#U~W)dMH8vd^EO!dt< zc{6&2-9*o6(PXGQg7JD4jCOW@GqAq1Ifr>0+}im&>21@XLQZvaPA#x~l6_HUq}t~K zr(WFvqup{Xk)2X?L#M2cV7w2*$m@7GLKpD$7G4w%$jYWBK)aNMRv>y*fMm2z)?Roq zc$IB5dC7v3p%%@uNi8!!56ukd=ic@2o%11fiT}&;Np}JcK}mCqM$f4Uydu6Q*UQ(q zH&DU&BwV(GUlSLoQza_SFDMRg7B+*|a!uJT zi-8(`e21-ntDsUpgWf07sa8xH8lhQO9NsEFzrZH8%9el|0_G5C#`5~%uQZ>&xEREN zBQH#auWpsRHO}1&)_AvQGU>kHoo-dKYxIB<{#)*HgjXfA$+TMi(X*I;tH-p(jm!q; zvtQ(>gG|O{WkUBJlE==tI53~(gqd+ErxXV$vJd$zRfdu>EG^r^y#U$ptK7lB4p}|) zenmoyHVt&CGN8gW0kmkKxd1F@2~eNbNk4FTMB{CE-0ymQ9=4vi%FP5ZU{H@SK9A$! z7+BI);+Yzv?dIqC*aU(skTqmDUifMYWIgyu_XC*;KPglH_yI|9;1CRmI1dZKWKaqq z^rcWyx7~}NKPOHi)F5K`wn0`QEcDltA<|O?7q4WfD`*gWPLLOHS`}q3*`U`9Q9Y5p z8%>O1=4~;-jAJm{lo6N|6OVp#TCgQ=&SBF>3`{uiewCGz+xS=pk4??`fC85yLIDpD z`3^Wf9b~?JibQ7uIzq-T&ulFX_^B~qoKG8HUc`@pG?oaVB4Xt{Xt9Ya3RVGS(TUSbWZM$Y|uC8)DHZ*vbkukP-MObJl=A*<^WqQX)KWSbj^CfflUxM zApG+h%Yww0N*juyND=ohhaeH%3bJA?f!I=F&>H&YOUbWR%`^6vc zmv>kRfG-~yrz2(njI^uyb6V#eOHv#guB`({+!!cyi$QebtU+|68v?QMzwRTCTvz#R zm~``7i@x2tV8f)CN!grqI*U`M+CmNdYVgg^e^C-yHn)hgg11eHlB0HZW2}VL2n&jR z@%787Yt>RynB@37-@ZgPunV&}a5`|0Sx{v=rGW0gTdAl$E*;>{R?f?%a;XG)yS$cc z^@v5C<kAf2nqSbc)pZP65R&IfkW zDA3*<&_fDD{dB7MJbl>n3q|6TXh8);^Ye%v9Nn5gdl*IQaP-UqVt!)S{M+h~8Armo zgWsHM!pv_zotH-nM#78(yBYOnn5m`|6%;vWWF2WS<<=@e4Sp^5bH}$^NWW_g>5+qP zah}`lS`m~Ys)UxDhW?!uK?Q=_0${HQ@XmhM6xjHJL}ia0A2z+kEpnqYU5$yK^=<{i z3RMAD`+$BhN3VijpKbJGS4;zr1zR@3kU65hd^ZL_F=L3_ns|7D2}4akS(-sgUl}L1 z*$iFBD8*5V9Kr@iNG`c0?Iq_Vb@XALQE`1`_@GlG5Lw$L7l>Y7A*_bz+R#<)U1`?L z6u;bQNe+9X2R!H$Cn{97gYJ6)CyPG6K&Oi1Xm^C1bH}{& z7EWn^PE`?9?tO+%q2sx+;j!VUygPcF*%f7Ds9O9&^iI)C6PEt=cYPVzIg&kcT)zxD z)(p=j@23=dDN;y9q58{C`fq3X)zgb0K^!ee5oJSX>3&!38o7 zs7D=iLZL8B1+~$5)gT#G#jmBec{^nMeGEQG2z1CwAc&^W&<(fVuWBWwYNo~^)T6$ z*4LUpB=HVRWhykYA?cJNjUqZKYMb%_P_RFK_=78v^`3ggLG_YZ6`b3_b>Zbx_K8|F zSekfa?jlb9ydrVA%VTkGSoa((eXizQ5q_qy6FQzw$*5uSbOiexu!Big_jeRyQKod@ zJ=Q}rP(4+qG-i_x%#}6n$H*p8PvBu*g%CGSJHm_9m|M&dXfFUW4iqcrCr%#>#y)&0 zqOBni@4<~!Tu^1;e({!o0;NtJt3ciMcwgffIu#Np3;ovtl`1!++$BeZm1S$(aUh)v zf7WqXH_>iz&QxPr$qBLyy89guM_)PL1KT6vEobOHSFLfpDmq<_7jR35m$XG;@$)Od zQ(wo!ftVbpQ6uxq9oJZoMuaxEceRVJ$t2|t{-lJYIC>_CU47&> zFnN{J}XQOC19=C_?vya5TEkH2hEROLcJM(mKR#3GikKq+bNir}TJ_99> z3dn5h)UDFipaNwty=CSFsQx%9O!m!1tT`k$;8? z6~H}1LRszaKmP9qI)*v~(mDB&N8Y@m+~t=pZifXPyMJR}UHq>N8b5RjQC1S`FqU~< z5*yEF&1~a#6V%nnq941XtH88ovz;hb=&^1!XZtIRct0!jIBr1fn}>cs-?CvjQ(ti4 zAMW`@s1o->(Q-<9Pt2_ib;6Y=xE)r+f@2UlFwQChh-)yUEu4uzFiF**J+5 znAw^glmb|6+o-7Wpwq9#i8{z~%}&?h$4#0F;X`sAvVpbqqQE}?9KZY-J6&r2Fu~L|Td{IBfKeK^6h|4u<`)k!syOjvAC^Sg{7G zW)F6=95l#DM-g%KgnvdmWod(ML$IzV!0}^v~vUq9HjQ| zg`{?lD2Aju$B+tNtVz5q?wYGpl?5&j$(o59|CB1#8J(^hoHk(SprCv2oR#tf!&tjzr(Y}f zL5ALL8HQd4Wc!`7)mTVS6ozZKww-PbN>^_LuMM*9k0pB5Ht!}C^sJ_#XLX;X-z|&2 zIR`b%dgL)K+8jZJ(m1Xb8(c%CJ-AWYq*}+ToL3}t*shB2Up?@JB^`@__0;IqO?E&( zNkc&f7Jgw%)fT%4Sx5NVTJzN(Ln?QKpJ?4FR$qO`{pMDGxUWrNvPrIwde4*1>}-9zh3!y)Rcm)H8Qaj*5qA`>iOtDI*vRCrSXxgpIA zsrEq=X5{)BcX_tj(=p)8I>3*-$72}MCZ{7kwWlrp88)#w2Sy5m_md(*=66L9m_ef? zl$~duX_~A-1kQJ*3J11w*9kjMV+U0!O zIa|D#^mC&HAUKx_u1p0TBN!iCSObG+%z{1V-!wM!O8Bvo^H)o*FB_BLxSKn|zyiJ@ zP8DC3e<6d4Rppr=oo|DrF1$sPAiqKO`Yl{VHV{1ZK$_>Smo3z(lAwJ!6hpjqS)9k+ z09@z*X>=mecRgY-Mj7G_I6$6-P&zk z_wCKPcijQ;geL?A71RK75(E#NENVD;z-mRXs3@MmM6s4u0WC#^@4AwpBofU735)Jq ze}m^7%=Le+`?|0Dy8eG7WN}b;%Czd7<+Gc);%Ayg|5N%E9XIqS_=-UvVc%^GQQtCl z$xl{POtQqO-`W1=TVyl0IMt0^2{1biTBazZSWpVdrDA$Wttv^pIP4yygXM?rsFD!m zx!uMbp)14JdzVU@G|2X!qF6;T#pl>l8mv8N!MU~ogVzjYhv+QR4itsD3D;dmB<47{ z9(5gfWDp*lx;f$Eg74UdCT{CR?1UjXYOJ=!{v$0eUnYdt%;})te!E$@U6@N(%xafq zL^R6&p76txsBCp*`00Qv5xFC(>37fo!!x;q4QK3EhsAL^H>vD1yZtorfnqR(t9E4 zd8H^#xM+3_=qniH-O`(0nc~Y6R?I#I#L|aUdasL;G~r@Cp3M*%tmjU=}$5b^&S!+NeW5Pt|$Qd%mjXk2p6?8#M(t+ zoiwhX>y#x?tpVwDdf4d9emHZ5xpH`U0A<*)kLEo;4%+WwiCGH2eCvG@J5;T|8|Otp zT(GkcDjCI_8j|^D$Av45mi2C<_uXk;$o!mCOwNr+5Kj?&i)s)pf zDT+*Qq+>bAuL)?Q4T?2^W+;@QI?S6?_Q#k0+tHbDn{Asn}_(6~I6 z>AiO6AbXsmsHPjB z*c7sXyju__Vbehy2z&N=!zwe4bS1S&MX(Ez%{=r>d~%WE6;@9>tPl0Dkm7 zS0Awt;JNQi2z=UGhHlaqU|dKGMvcS-kYoRUG47bJTc1ULP3P-w6qTWz3g1q;%H zOG3&+aH%IjoWVaAcAtOKuaZISGM)OCPnSG_xd#1dvBD$bdkPnLJjdYiq#?oiV6hL6 zp+aHbtUtbKOKr(zjnR!Aj2$9OB%H&QBrVpvGGCy*E@IsKhLRycN^DauR$<57+>nE1^ zo||(|vU~o8xgEjHn&tB@NfMaT z&JPPb+?-5q#*f~&Wx#wiR&nEbksY4PDpdm08r~v>yzZwVwW`$-I5*Lj&_%xb@K|5i zsE3U<9&B8NjeVY=Ad(xA7c&<+QjQ}{q0)(SC*#}PB&vunj7N>LH=YBou^8hZ< z!E>Gjid{*Ocq*nzX@I`-l(4G`OJ&)6E*RgR{4a9kJAU$0oPUM=?|6s^z! zIM@P+y7_A>NcJ;Rvrub=>r#q^{-+&O%mKQ}FN41;=oY;LmXH>#k)0>i(scR^vz5Lg zTjkj#yC}Kow^>;)Y}RazK0~L22GKd_l{epTp0twKuzMhz*C)Ru(a-%DI6v|kyyGO@ zHTBjsXTKtjp#L$$&cjnqegWS9@ZKt0wnHuvt^2N+5<5JIX~DN3d@%&v4>~YuYan#t z(ubltVVxjHxHxFDbVGEF^!=sd%nHxPNEu{`P{8OISrgbM81kay$~z=u%3<$nx>-`%%VI3OD zEc2nNDFviInxVK-i^aRh{rego_NW~8bahaIb{VG0M{Lq3J%H1oIp$Yy zUY9@Zd{Jgwqidom3Fl;>5M8ihT$cK%`|6O*GL;dX44%>8# z-HV^95T4jc&Rp?pl=b>tj;IF5Qy?^PFrzM>E-=-9yKpPJWy&g^Im}ocp>2f~Mm$+b z?-x|k>Eb>Zz>rNun?o;AH7T3I*Ux_N5F{LHqs^qN16L!dvwrd_9$*kxF0mOcMG*@V z^&|3&l8tOhRE`=o*M{NHhL1GsaOLrPj|J6`C)5A;ez$Gm)H8x7J0(?+(KTr+3X=mN z*!d@Rn>hxFA+6@pgBQKddrz^@__e0_ct?S_ zKsYk0bh}@RV_&ti-z~d;&Vv)20!jDHzPvU&@xN>}72S3UWJm7%5R(Bz%vmYW5Jv8i zYIT8nejNE6xDaw8GQE32)VnLHigW;9RgN0Lic8Pig(t*I#6^(>AV-uO7*7@oSMdtO z<^H-r?L%IKj0hxV2dPEVW6H=nXjMz~ch>C58NjxI{R420WP=m1$NWJ0UkSD(?_35U zH(m{}L;v0){ZfwWh~{~f7U?fA^3YGdAZimJNk$d_gnyGVm#!jOWWm0;gIQkqqN2gs(dFfb1)d(T%cn`Xg8d$?$9hS%6`L zuU6mrV|PXZl*c}JgePuetabHwKEGi8m_qwn%kMS%@}~UliFQUFoB}s;Gtf zzI&2%`o2FDU7CAm%l(bgj0o*D8lh4kPUK@7zDq$e5_d8}96bK^@1EZLQ}ih&jZd5$<4D!6z3bm{zxS90esg=T{|pIx9baY?d_kJX_a z(`!}n>>laoQApl$IN(Ikl5ugui<0a9u|mE2o?<0a1Rb=oVI^wpD#(IL(sX(&ok=eu ztqhPfJ^m}_cy=WI#TAjxz@J~R;hoL4bqjV%0kY@h`DFs3OZR=P=(PL&3r7T#{= z3^Z(H`GD}0a))a0n{@BpuQijPxrO}sN(D2I!RUz z)#vEO^C2Ky7?eobM6nwwfD8zzVK}Q(E>Y#s#lagHT#9Uwf-FRf^mEZl#pZ}4@urFT z$XXDlcMeV+02Nm#JMc)STw))}*Kfa<>SaO4U$m$HNaCL|bb!FsAarb`*c6Jag<2P$ zfk}mgOCmp(>V*izG~vMxAK?D>HYm)j;G>2MFd`>tIK}9WJ;XjVcD^*L^k1uX`-)%p zj83v&t*wyUP#`0Cdt`%Wd*nYuPuM}YIA&ufUbU!s%xnKzjFc_X`~EqB+8e%RmdQe@ ziTT9))T|!w61kap_4LFNg%%}SSNb41_!>or0vZU*{kwRXQ$Um`dedx|jzA6~#X3NG zu;)Z**gWdL!d~wfV;Nc0S$qCW?t9qxHBZ;F^Cqh?XEnvHpvW?dU@-QlUY&_bRa*St zXbv%E12~8W8*90Kw|vVyqpf<-qs1!jI}rmJ@xhVC7K%-y$Od4n1sUHhepp6pU^+xS zQsaxO+4Ga2gcdx)RbG>1Yaos$d7%m6_j#$x(ZYj+D0tupTwI;<0ABW8z{+ot?+7gZ zV&0hz2gnL;c{Df9wr;aRP8!8-phz}^yeUu&PQsDJQ7uL6NHvU8Jfm>EU za^{YGpD|R}u;TRvvVW~rn?@Pd3*<(e6!nli`idO2;A=xxGDfJqts`2{KcI0a0@((2 zZO9#xuhA)Y33mw#nNjd7!^M;{@iWx3HYzs%%2un6%i5kB=gaI!?j84SknduUl{8JJ zMddW(4=k0AnLbs1m2|>rz+sRnI3Z)ueXid8li|jE#xm6W{=+**$Ral$YCxKKQ1)vb z#U@f@H5GFboDuG|=u%}U45{dOmu8B|_M3HqwV!P)8WZ6g2Mk(`(+Tur! z#{7L9DSpZ_>kU@ErIuoWRewLuTK20@1v{Q#;bIZ5BNDP*=6MRj!obw3I;MjLCxj$A z13aS?Jqy+; zqVyM(o89!>$W(+FrAbi@G?pMHfQtbKghk#R(15?0+T!Q(!D_gG7@4`Be7Ai-3^C&w z6XJh-jO{{=+k(UnZI0Q2T#MWiav0WeaU>a?M!+F#;TR}}xY5OjIDum9Uz1}KXIX|H z@7qUG$R0NiV^3KPHFXqwkRmlyOmpx@KnR(lSnPdWpj{G`K0QT|$Sl&NiaSKw{ZrZm zAF3}&Zp_*`9omaS8)SGG1qRdUBu#Q)Nz^VB4#Xde#D-{{aw{zUn%hk;c%?+PLGBG| z1#Wt#D5{`d1_B>*|DaKjFVCS1=azfp0}>dFZHDcJ97EEO1BycjhYL7PNE*{P?+@uh z3zGg&GXBqG6}QB!8>jROtZ=uPVpH*tjmewT?cX41klhbLVz}LO)%f)j7tdZK+$l^C zN|Sc-g^tmUtf_kfn|%7@Nzrk_%~K(HvQSt8l(g8Hru~p~O4?{hscdD9lfgfEwus98 z3xq|HDdSxU4X$Xg@+ZSJwFCFrcP)SJ_)g$t3r^IlmsgQ3;|SC(4oY7XQ7pWWc~r~- zoRu~RY8cddOX9ap*QwL#MYGSV)<&K1zbMUTl7e>mU7UP;a)Y2#ZYB{KD2L4^ZGz9| zUxKN}f(6y|XLLTZdBT2`ODr)W6Ikrwz&FbI68k0?B6xIP=PbVI@n5zKv~C;K+GCrA z4x1ak(0zq7W1vw7<(W~fM|_AK$H4mV`qA3r1lF-qUf93;S$s&-8|814)Mq9X*<^*Bs=*> z=sJn3v10`RE+apT3HJS%9VQY<*ebyQ3_^p6(+ zCl~(Lq%=mZhRILWOf7bW>XgXsb~Utgs*CtGhBhV_Y`Cxj<&i(`ha1)g+6ojtx*X`n z&cTkF;t9xm+*V&y=S}X6&hptFc>)OVlX*Qbsk-8~o|j5$m{YPQP1&~_pw;t~rq5%) zHQS`0C<>pC3+z%PFbUzt7fxyPUirLkX;)M)of83c`%-+uWz!St z$u(c6I2&$2c^H5^ZtD@N595IS_;%5*|NQ3DUT0wDyuzivIy#YQF!B7EGx<(2DyQ$; zK?HDmrZXrutFG;>JUZ9ZqfJm}rqUDz&@3uBoxcarEU7(3p;Z z#2H5C`V9LZ`QY1vb+%>mTmn=#hKikn`uo#wPtV5|8qB=nVliwQX|r$-Oo4Dg2+1Fl z#i;(JpL{%QRD6vi$gv$vxT0V<*Y6zn@8P!Xz0U|f>@@y%%Jv4whE-zCJn%CD9}<>M ze;T4wu41dEAaC3aUu?-j&bU!Q23n0PI)-EE;oPIRaQA4~z#QhgH;PFgX(N3eBP zbC?Q>EvHBc6|;u-p>PfFlB7wK%Ui?y3x=t2Bpo#Dn&Cj1=s4)J*N4RWc0d{a4$)1K zPPsjN?N?TL<_DaCMLkqySrK%5YLQni5Z_;x?3ULH>)^^V5t8Efkb`8mypBF6tQR^F zSzNJW>t(F>hl7%#?@3zumAkeJdv?~!u(2A&rqjcKpErI6@I}yjq`>nG@<)}tQ&SW9 zp9MEEYdwo+VtJr8ojxn7R$Y7H^u%4yRr0XX!-?PK5dW4De8EFJ&3T9H5BftzhHfhr zc}56Kxv4p_aMWfWUXk)2e`mMI-TI)~i5)pWmX|D;vI*P-&CZ(HF%zP-nVeo6nJu z1x=bIujc&p)1Ov-_1xT38t39@1h}z&C!L&`;Dj6Z_293TEx5wB;3hQnwIiS}&FNp} z#*uBd75vvz>^h1hQZY#{bj``3zL?hv#nW1BB8T44S)T$4BQ}X^v2?(>0Dgk2;)3rd zJj$u3vsM1TwykG;MnYOU)sL1=(CyOgQZu_W2q540NBX@{1q52rQBXM5;to9Qvu)`s zY#X-Rb~(>Z3sy!1CP7w4RQt4mcEL4u)f7`2MiM+&6$C5L6E;pUb9LXb03$G247l zQ`GSN0ew(dlJP;+&}%O)PKripS;j$l z47~gJ+r_{zKb{*rer(wC)qkxDkz=GNvR-{p5+|$;(ewK~%#4aCe5CD^q(qoST?)KL z(cf@G(x`yQ^dWWQch4Pb$!;uO7Ip}v72Gz1N48i65^E@S6-AZ<#j0>k;Fr=SCDa=l z6w76oM$!v4l5gG-8;|7q!@ZZAxA@U}ak-N$i1^O-H{T+g-MAxfuN5K+DHayva;caU zARy8yv3tLLVIP zXHqN(mTjV9)`hoCw^*Z^PKxf1>h{jySIjDz8#`P3iO2}*E=f_igNvge$_s?a*lDee z9sl0I|6F4T$FoU8n81AKv0v!$PjWMb+dA%`Gk=lW4l-Qkwr-q4w=*OiktFaFfEW+p zivdTrxcsb-_9`^tjoqOxkx^7R4*Qk4``tM}%Gqb1yZJC3z9my1%M3Dq4 zW>r9obOi&&LhUoPU6T8f7HOfsw!t$CHVTBUURt)bqC>E7sb{nOi_H6{#_jSE3X~+rrkD&MSxixMDM|>`9^}y_WZ+<4=oLTIttlZW;2Q50Prbd>vqRS8 z8!yZYz2l4Z@7ldD?0o^b7hU6HIKai;2syO9M|do?yNR=^^Rj9%%x9rO=>^ z7ak+pQ?qtN#t9R?^{<7M`%|B5{k~U+LLk=ao1yhUBFHi_6mLuL(s;iNP{0qGDmcy`c8f7I)J zNdh+=jqadP`W!htbLgUpzzU0uu+{v$&{FynzdZWt%v`m$S(&0J_unUXq{q(*{>7f( zdHnGw_#bZJcfd=0$u}(sQ!SbHALOtbBg|lhu+tR#F-01vm^OitzAfpBGAQ<|bRfkT zJA1k8GzcLV2vMf#f@p7aqi-U=LzY8>wwg)V#@z9VDSf;WtDfV8$HTh3I%snm7gKMs z!aLxEHbZL-CUy;qyY!0Z^m9-M6z?%J^(6;pjQ^1A_QR*9C>muDJF8nEC!9A4YO77W z!ba{Dc%uibCm_k@G&f)!{_?Liwv^*`xVrO!F4runRVqK|kM(Dbq)pnaIixCt;9~3K z4v^2nBd+oNHvdZVE;kzt`3RRBJ@VA)LF>Isfkfrx zkJ|A?fonBcN4myW`RYizcHH;1uqT6&Ii;AXrVDu1E#*4i31PkDgs{)U++v;{azbdN zJEo78o8*ifr(xsq4|?s(lWT0dcHDMh*-jr0W}QtWddMk$({n%2=$T`fJrLcLK%}$M zXY~XOdhH9R9i2(_ZViP<`0@xUyQhf!QxemozGyLv}}( zO8Pvq={8ag)2oA{2MUuJIUvJ%W_RF8oRC3niM{$;+lm7_WDYp5$b;DQjwo3S*>aSZ zF+fK?%sFM{p>41Nlt&}q@+c})aWDan;SPLx&bfq>3EVf}oAUS9Z`h{(-F8f|lgu|) zze5xWXhl+bq@*tpw}6nzsM3y@apfppor6NhpTY^RW3uM|tkG7n@u^}ID6^W!5=&>_haU2dDue6VzkiF4m3d@XlLz3z&S?|Vqwo)rp3HT;(Jj^6(`Iu5Qak>INcpr&&`;`V5X)IeYDDox!k*-}a z91$ou8k)=H_CYaJd>Zqgp8siEn(~aSCfcd6!Kx>-IvbFN#|0ZF8(>JnEUuA`ovrtZ zoqZDOZnD++YJ*Z+46Nml&2Nz|6W(yqy5S(x4i`~_9^sPBHw)w0K1&RD;k)C$MeaT` z@~=P=KBy33CB?>5WC<0MK-wc~_y_z`#=rfx3&)kmTIE{Q%I+CY0%?CR=ol*@DNA~ zoGz{utqBCyP8YKvm~H!HfD93jxC|B#AN;pd^3NPxiowymMY=~CE7amEhhOH~W{N^Q zqh0B9NR7;>SfD;yuOatu_2IwVm0O&Kb@!dwzE!jCzAY1p+ss2I*re#3 z3i2?tLD5R`LP5`4aN6am?qQq5b#Jce$=uB0wq^Fe{8ohCma5f`@+w3l<8^+(S7;12 zDAxJ*`f5u==b1Lp%uE-%>SeTNM2Asf?*T44&IuLnTS}Vwa@&8~@?Je7Sr9w)T^ksj z)9h2js?mNGN{Yj52B$QuCv@?y^5RJBY%_QGofnLg8)ZxQ10`){RYxT4Oi`HgZ4m#M zH*S07s6n?tv{G>s{Hs&n5QpJEIFzxo?aVh}skFb$)|bQ z&ssq>ofe!o9$92h1 z_^&3p5&a%D{F5~Hs3!3HWZF;sy5!ei?56MgFY(92m7-LY>H0i6DQLNDi%eTGt$bD- ztCxeO4zD<}*SD4VZ%2`5$EmpNFel`ViBCzKXsa=7N6NQu&LSpLogBE6x8(W7L3yEi zzMf6wrv=}lGrf(1X60_bHbK0qaLOP4Z?mS{Ya0_QJQ;rDxng;@%O)ZZUc}+r;=u!4 zw86<1?pyX3yXcjdpR>59px@~ikd1Eap31FKvfC&Y_}I2mG1ury!D<%su+Rq9Nl;Mf zG}+|^EYc@|l@xc2LwAU}q6z}@AgDp^oJ)aw)cqcfvOIe8gw@kxg&m@6vL#{S8@(BF3$*slI` zWZo&+8F{Y)xtsAs(WH3j(<)|a`A&@{ozTnvz~4L9P@Jsez8>ZWTTXoP4GU}zzj^*s za?p)EOq&&+n<%!CBK5$$CN1*qoW7nn@H<73F)bgmMC-kqW;SciFr9QtM1w#d-pMb7 z#N93LPP%t?kED<6V^Wp*j6Qt#++1E4ocU0=QQAp215aDF#0V0^O*3;KpHu;B%tt_D zC7IL&bj=3szlf%pZGwZ}MA@B1Vg282g5OEeJwOE!E36d3dAJ`r-geB@?wEIRF4xy# zgx=@@+Ml?=@pa&a_x7tRzGFMvdbIq>eX|Tu>NGgkUPrNs6j=?c&HQV0jvC}01G*I2 zM%j5{P~@na$!Z8}xf+_|_y6eSjw0pJJ6(IkK2rD_-#zoRi$i8Gg-qs@Ho-Q!&Fj*% z^}KTbIAN@?mzMy;=61=xkc99%HRiYlmV5B4e#6abA-Fm9B?`lM+oJ%y)N=7pBXO?Bc;%o$B$nt_{*RZN$Qn0}Zi}p(LpsHRs9P!(gZu_HVB!usfvE{F%v>ee8rZCXg=VHr za9MEEvvlfBL8kg3zfazwNC>|);jG{2L?XkmxeF}(=p(reHa{0?rp&M)wi>@8B1qiau|pk%FkFIdlW#NkG%DZtQWp(!O#ak3Hlj1=Eh#; zx)mx~D7Kj*r>Gb#Wx<7j8s>^$_lyz(LrrDKRu-27iXzP`i3PrC(qklZ3d;ZOePNTA z7Figw)doc+ohZzKl?D`lRb=UmWcAqzP0B<*C2aC626_UMGOA}xL%7Ij1VP>V8ixb2Y4oJ;Ux zjh-R*o&+FH9%{@hs>wfj+MAYd`Iq`#rY}j-)LW*MOz9S%7v=IQydf6@$K%L;0iM}9 zxhZ1JC{}B5M8}hi>a_Em&_R9iNAKsA7T=>e_WFL3`OJ)8RaTyAH^uItfCeGvBT=>* zv(DS-OmG*F<(r6ExY^f|pMDrR;>iGsY)d2yjH$kiNtf{kZyuPl{|?R=G^D-8km4)8J~D z(H{hLgwtE6*Fyb7>o-6D>g~C?^HSSba3r|um_V^BDH2b`AkAN; z0P}8@qCWXQVd=+g)hC0*$<}x9arQwnEi-70t@xQX~r0N5<%x!K9jO|ulVGG5kQDg%ZgY9B{ z9!Pdj5>-B{UIHL#6qE#-#LW<)Gkv`E)Sl!a*VltD3Cou>{fiw;7`sKgfB@Np`D zca~=vNj!o*;>Z#=PO$+e#-LFng<{uIWDOOQtww#gc(1qfXq0Y3nf4l3BK{|-#X0Ub zoZ$@|e+D;P{3;|Y!nT*pZT*IwCbup{)65!Q12m>qLLL-(r_VD>La|~1iEanB7cmLp z72=zI$$&(N$H`z10^7C^a`HYWoKW9Bvp@W4A73mQ)p@`q zZKS(Jt63=dM_O5P!((m8PEtA*ifl1mo-RH&rNQOUFj;gs_;4~5hI||+e2o3!uVZ{| zC4KB@;j9~f!OIly?Nb*;w$Uqrx!2Y2;L!iT6pE0y-g#I1Kd=ApjX(ZIvXo*MQ6$#t zEgbSj7a!#I#k+0yDYY2<<%M;v#pB90D<{e&XdsF*wClB8L4G1{PP z5R@@T=sPoddFerY9(QK!rgzJefW$3n;(17ySJMrGtlE?{yoSm*Dka7mJP=PS-mAA-LdmRae5J`vV8@e;TF6t zPwCu6irg4)$E@(Cr&y2!+C#-)t42vwo8bQR+tV{5%z_Fy3zDJs4@$aV8(P1|5xi;Ois_UVX^K~#G=VX7swIe_`6JYzXqO$49O8o{OcPTLf|yVNXHs0isdqNP zrD%JQuw#*-54!?CLw^#tqnz+FruWTK{zw*e+_)Co&Z5qMDsGxgi^T&F7%Q`(E@iJ= zn-R4x+9}X*++dD_$nj@7=0byM*Fx^U#AE@7BpyI_y zX9H0X)|ngtSs~z+ONttxB5Vd~T&%x@!#&=gdLJW8H9#$9P^8V$Plk;YMd{RD-wtqW z&^FryB)Tb(_R&Xw0lGmt<=bz^K^N@}1wPU-csh=%k&a;x?so`uGfus`J;;Kp-|sxQ zhGcO=l^d@h*I1#dgkqs7w+MJ|1;tE#$okMcdSSp(A6;NmcpKSGV`V|RET2gOT;Z6u zlixoV=Cu$8Hq4x)O^5dC@cV|fqjAKY}$SjtNoP-)K zAL-Dw!P|PaJ?)iD>GUqsVflP^sMBy@k&Hmz1NL7>%B52m;0aa zzvAcUTkg=8_n3=@VeiL0&3UJ|1u(JUfj$!~KBeV_Gk+k-ZX5s?TX~6WiiLhL9jN^9 zFGv>7UKYG5@VuyGO0(v+ywBr`qR*pjT5R9}rb%;hsvh^m#1k^Ejdev6bL6BB$HK!LR#4>}GGn*WZRSvvc-m{);Cri>M4)!NU3s$dvZZE@I2bQg0A0RhERT3(u15 z3vxOXhBhI0fDHW{XCHk=0Mb2eve=fRhf77I8&7lWPz_bnJLvR~K6xH^x$HR?{EK`O z!jCXTk}jY$I3v9Z_-~Na(bqKl!$A*l(Pm2Oq2iYY z;?`HL=aN!4j$2P!g%}4Z7DmZRD#i$T63nWagyp)uuh6K%-XiUchK^*EpTw5rPT=E7 zlWJE-w~DR;4H*1@LDR2|F87BHAVIe`q)$XwpmQump;LE&E7V>EVho+qe6ZhRnSY)- z588_=AvLp5*h#A8OG3MNr2*Yjb&v>lFds6c35N#^dt;^T@gYCbp@W>9;28g+_f|jc z+z3bvkE^0B(v?11+(=;Zyw>{MlOQvEfv}Lm|BSSG4|=(fj8dmGA9nSKIc$_TWb5Hy z#0e#1Q!8qI`n1A>33qY)$RMVxBZ9WA&gQ9INaO_AH zs_&baR5k^6(t59plKcER8fR;$YFjHf4>5}QXfCf^c3DzB%lw2=d5`3te;2P8n!;S- zRUC$l?N~Bse^{7!!cE*Dvo-&@_@|v)@Dm6z{(-Zy3?^+tJ?Wskzq|3P#V@~ic&<*l zo&Iuqv*r%zljG`OOL%cufjEu9JC>Q&pX(=P@)DRHsY@jeM+TKA*n0%0xj|+2k)PhT z?F@6jMrE!)YAr8fkB2Q`kvi~>WC>)5 zE+8FXcpQveX~B2sCB9iOXYP*DPi~Oc1Yj#NW*q;eMyYn+@BN1@C(onQfNq?ox5Lx3 zC!~{34^HHF(3K%&;*{~GYJiO)l_3TtkVr~^zXBO=JLrR?+}{YZA!NjLiQ#Qtc`lCT zL4acO3|H*hcQn1ppT6`hOE6n9;oj?{&W(dvqg60_hGLs2(g^he(vFGAGqLJqGw|kg z%5&*OvyUj7H8o5wFHO2IDl1}JWUXK!vsiYQ80lusVbVkJjvD4Ja7r$g)sh-0&pbvB ziPw#9)*K^UynVjM*arD72zqbu5BuiO`;;{SyXlS6oZy;(GfaXCRv9H_;y#a);S2eP zBb#C6aT&kF@AivE!EQ$Hg{?>Uw3_H7etPgKHisii)UZ(W08X4XIDySRiiT)2qw?Ys zLT$^u-8PW1Q-5xRT#Bjw+^mdR;37)5CtAS~L_F~!F1UabBF3s3YhJf4o3vAJ*fu>8 zT(?01!iT78e2i4hEQ-t+|27^^alPcM^;hM{w{L#ytnYL69;a{LeTxgPCyX8IWf_A0 zqCNdb67R-CP_ESww2@+=Gh{93g8Uz>{!g6uvA9`tT6!)lL$bxUL9j+#ODZBVNeVAk z_<_qM(yZX-IA%v<-R9w+#tE}j*vYTX`PV8yL77(*Gc9_*IOSqhzsGJMP8#J2v&mi# zLg=AwM``QOk<)*+B{q0;QschL#9u>-TR1s}Ew4#0ya2slx z<43s$$wBkuz&A>UH-C(M4tj$259kn{qt4qUa%6q*72N)lazw&09?jCh6~xjiEhJH^TZ9?K5ha8h#DTcrHse zl7$oWQ&2~wm+toO@vfNg1xbg>NAoii!jmQ(jqc(dq4oTd=+8wb=^k(A%!@go#nB0h zQ&43eEkn&&8eZG`*3+)gk*V&Vp;PY*s-{n<8)O~yUHT@-F{kFaHwCbeTrH@h zlcLPmmiyQF=g`@qOZ>Ix*e*r6|GMzD={QSjk!FZ%fHc`Gfr#y6xG;d9$ZaHZ0j%u9 zj|1R1!D>v{Pk;Pvsl_pgZi(I@iQMXF-FPemp2$J#iy0IPGwC!a&JKhD3p!%45iu{c ze|iF>>a@$p_Yf>iNl`5JIWa0>AXnm-IumW2|trTKz2ca@5l3F-FD21AiB% zmtFeHZ>_NH2j%hrH;z2=t-Qr1irq+o&0=;lEi7sum}GqPUTuO((JA&Cw1%7El`l!) zn*@BYfOfRp2Nx`vMxapwRNeo5WzoM@^95>1u9%fcuG71r+_9QI#(wUf8F+}d8)S2l z^i8MCk`#b0Wv=@AixmNR!X*9$uNr>VXe9ADgJhJDB75zfRcgyD_GrxC*OB69W)=%< z7J~wgT8afdg#AEo0ZQt~`v>`vy}>DpX7DI&^fKQ(I+346>6KbM-@1{+g zY&V_V{Ag|{d4ETFqwR|EGcx(HvzCm^=oa$)TVY!Rw+8Bz zEz&iCpj_#(-8>xto$K)gB+9ryQZgrxO$hO8e)w+&!Q~ru#W!c~ncO_N+wz-`AOE2K zX)j_#-o<2bKD1C^nzdq9DKrkR^KA*&#(P%GGAImF^OY`o-R*#b3uZ?%_sR*gW6qDc zxO*h4|86|xwX^!411o&#^d@0?V3lZx2Cc*8c_c&e&G?8HWENtFuSL8GiB3bS759<-kU zTH5G1(#tCljxJg_xP_70pVR&J!IGc)ji1{t=eccjm7N8;d}#`=Pu{F4B412wm(-F{ z*uMVaO?rVLY!#3Kc8Hoajk3l3?es}{!D!)u0}{-`11Eb9|6=<{7@{Mq|xdb z$#>Z47rOzAGC0f9*3RmZ^m)|Is+d(0b;I|@j8WH>1sgjIqT%CVSa<5b=Tq>sQ%f|$ zvWZa&Nr}`D{c{VS!^{QdDv)tvH*?q%)GN|l)8cRE8+X{&=q~I$!xe`)IgByEKWvtb zM7h9?$roF#@?Z~?3$o{|=jp>cf>-f->70lPDqhts?TUiTnEdWn4DwVJj6kAXVFk7J zE4M@y)B-%Xpn~e~-=#i74Du#Tv*r~1zdZUJyFvPSq<%`K`XE2g3kEQ9PNi4vo!P8u zqd|hux0uOQ7t5<@?Ath^%;1;H_54oyW4g%KAi(=J2W|;DuWEv=8m5Cuii%fdin$JU zoG|!cRx^wZjzE3+^wMRv(d#oZ4B3fm4>A2S>ecronB9ZHXrOn(w8W1KI-pCz)wmqL zbF0fa0mJG}*Y2eK`|p;7SkmsH>HkO~xo*6Yd(g^(lv6B7OYWj#?nY<%q^i38*G@rA zz#3+w^o08TMPYxdq_;}W1z{^zA)Pj1s~TIl^3`b*49Xn!UHUWsbC6VPhc$cV62|3PgCh@n7f zQ69a_>&kRo9yV)!P`u4f4^Tb&+cevCoUADJEmbh5m%kg|~F zvU*tF>63rp34J>zO|XR74WPM*YRh|HzzP(IuTm6!&!1;XL`_NqEK;Y0T?PJG4ryU$ z0`UnrHJx@8tlLiiw!zN=t)@51-zKSU?34CbfhC_}ftDbPiaAMV@Ug-*MR7S~(KHx< zGN-f&QWQl~N~XY$soYGymAylnS?tkk1ttxwElP@78hFFMVpg6oPM9w4019H)jZq`C zzN5zEd>tn@GUI5@+mnWY>PD{I_0&YAT!(17Z#X~M{oc1P$8kG6$`T*>&HRn6b z{k8obJLD&Qjq(!unw*m}=JIccYe!%fv#aL#d5&Jr$t><01I)Pmo2K7e@bYfsFGIZ~pe*$7j?yNG4T{f#8yUQ|Nrrc~hZ%q>Jb}SN zRD62kr9R!>&^`e}NWVvpx|!7}5un;OL4yD_3Mxg_^oFlAghM$=Dw`p0 z6C72idRD`y*H2u}Pxs24((iGNHs6Q+<#jYRgTlDf#9+BJRxF+L-lL8a!rz1u4=9*U zEP>bp3uU#HfsM?6j~ITA#0$f?<{5#X=12PP+tPDARf|n_a>Lj%Xx2u+Od%F6nOcI{ z0AW=#T^H&M?Ytbu@X)ur?5KSV-!uM)Y@Wp{&3NzQHllxK=3FgSgYj{SJxY-}DyB4` zOHuEcF4h`Jqaa%X1#%7Y5>bjGLEK5>A)WF#6y0xRDnhjAI5Q%aE3|l|CIsp%XBUUn zd)6wn>w+(Zr_ zRW%BrTH?5}TG|P!9~lwcgK~$FHVltvbqhyK{<62X>Rk(-;+H4PA|2z%N~=-sKE?J@ z-5Q!qH$pw$a132lMCeBG(GTFx2l#aVuhC@^j>(hi+9)?KeHDG#0%t<|0#c+@=m~= zfOZ)^9)>h%Hw0!{n57>i3E}C1N9Y6~BP{~8+DgG{R$E5kM~}3chAK(off0F#PI&~>dGr7@AagW_oEVO~D};P#@gS{&4pb+7)2RB;>d+;}nUg4KX` zf?_|S$cI$S3Q{@s_J5z7oAl~ds^hitxn*xIp11OcO|Lw9EJe{qVkhYWi$e>@y7745 zn!v^2N-s(_vUn6kARwRJH@9-$3b}#V@cfaGW=)D>2c1Lr&Mu7lJSsVGzu+v}BfaHa zH6;xO(uen50kMc3bFV1=a$Alp&cnhWcSKnYWDe~N2UJ&mUX$9=uyb)aZj2~91nrGq zyCg|cEE%_cV$#IpVN1s)gdZoL02Yjsdtf}=IqRbMoGMEaE36cq;csL1sw8jzPBoWV$;*tHZ{1N02QYML>=kiuS8ECnozEQ!LZ z;rkFt6;h+RObW(y92d^E{NtZj74+LiTY7#jxNu|7VTa5=R=68#K(pyfk|s_4{8kth zu+X9<`cp**Xnqs~o)9Ilv7Yz*7xAyo)cf>#WQ8X&ncnHa2PHdr#|5Q)V`L+JkcW31 zYSye~8~8VQH{p1R>e}>P-a%gB-0jiLnr%>Am>|9Ztj(7tE1B4^^g2}4i9z0rS>yB zLzRwhR;}Cmf+hH^2>tUivi+F}em}Gti1tzJ2NWrzVhr>(bvKl(itF1&FkdHdoBysPj1o}BlVL{X;7L7atMO^ ze0_K)zc32#zofbf3$W&A@1DCWWa*64gq!1doar6M+T;AigL9AjC{E`tzJEN^HjHsw zr(!2CxdthkDpDX`Jh6k$rLmezdw`b-4E|Nm?VLO6nwgGUnv=XE^sW#8gE-#F4IQCh|9-T>GUjYZQk^C1-MI9o)M^yTrC4wy8C1+!(&OE%xkezf z3{6ZaiZ0R|-77c~((8>C#zy)`bdzSw8(F{D`EO@l%b$;<$q~9Td?~+*pTL}iq_*?M zh9OPG5hxEv23Oata{>z0b#URtkx-tx@hsU6<>^h&MwxlZd>Ow|RI5r31iE!7(;5AR zONZykL*mbs2bcWTyU7+6ZX0yiv2#j9y7>BwYZRSy9AwVos4I$tb8pL$H{ZpMVF<%G z=qrY})3t}V`HHhD{}*f}m!FZ$Nx9V!w2fjxhIcC!6DRGKe&SsQwdV_p#ruP?(;iqX zd4*A_Q~Rf5k#HxyFGN45S%dr8{!;y!6pt-t)&N7$VRf_Am2h(qSc3+iL7Ohb zR%`m41Jf<>M`mF5da}cf;}2M|8N};TNwH8-_yHADAU@%r#7EW1L#ozr?WZ9Jc$Xx3 z%y!AX5TvfuM}GklLMEO;Y)aW5T1C!?j)iJBDU8yaqDKEisyrI$p|snDTiJS2$zXJn zqSy}XsMq8bv#{bE1VbWB+=OKzi^IC3J`Hh3pgST+3_^#^=7_<<$=8fo-rB6Stq11f zYuuPds@Mu{*%S){n2w4`r`IUXL;ZMd2r9nly|6<7`NpvNG-*OIQTpR)rOs}j&(UC^PLTw$K%delzyAC_?i1Y_v=ZY2Pd4larnB~3ML5@yOJXD zR18*;?^ks|q4_^6Xf;{n$zJC1hI7@~zx`z9f|2+QH(uSb<2OEm^_=z@iN38bFMPE@ zus7N;16N_%1lxQXWG6zBCZq-5f|Zv38MPt%B?kFzSp}6nCsW+(od*+XSB54I!o8i<6W z+7i?b)o8u5>m*;wuD^IjbOU%{N})GD)4P){nYLp>1r^6$hR(JH>!{P;`sf#3->Z1V zICs}8tLI(#Nz3=n{`A6k7vQpPrzANrCG4W2fYuUO5dC{FMU%I>I7ChiN_&07sM!P*=(`XZPVd11pOW^S+Yum(7t zHb%Aq36-`oWGh?h3B2qB=k}>rvn5eDnKCHiNhRIueFgOq@j+NPnIpah66Hma&?yVk zr*@{0=~5V!xsmx$*|OyM#X%7D=BV?7%6->M<7c;u){Vd5rBhyiv5qz{cj*h^x1vr$ zQ8RpQ8+~nh0#hKYCwFP^i|O=7%Jkq4dZiCGk()jdKP`{O&(dZ^VChV4NF^v%RKZ(- z&(S^`#lvGJ#=ZWFS1sU+|NZ{k;)cqc5nq#1*xqk*W3^aIVNibEjuJ(L9&QxU2YrElgkC&bTOL{wc}r34oi}O6goU0^U>TGYtu2H)^Hf-@h92=cm0@bH zZ@)(#ojUcJ2}~&pv%oZttw}zWfe0|@bqHTQ;Rc8{Ix@+3rNgs!e^z@>^o8~v-kH~D zgxIcAbD6Na@l4Flsy6z!EMCPdY%%GeclwlIJp*KRbV|$yj%d7lq8%Q+z+uEZdaFy1 ze7QI{%~nC$j!<(AztJ~Vn9I9P9&V*5ZiF=m(&-LStZ*rGh8zy)nw}kWe%d90QGj6w zUfrzeq%VKx0?dPAAcMWT!ya`XQWvJm#(UfBR=!QJZIEB(fucgB_9u!X_o+ra;&ty&!`5*?8wT zH)~5gY0n`9J?SAXIIsWv&ojqcqTlvkU-%nY?#5HEJgZ1Im14o$t)pVH)#d)V5vJr$ z8iTo>hie3@*(N1$jT8vuy&Q=yo%IDfWbKFJoU=amiO=EWFUG!f=fXJK`dmAOt##rr z$YNO?y=>eLwf?zy)s5#$nOar+xMq!cKShyA?nxF0t&ZsPsDI%QbA_pvcfkryuH?h7 zT+`$SeL>m;MZOty1EfUoDVwL3hvMCfWlg^Op_#TkG+_ehT1;JVY>w03UXDh3Xf8Wj zPyF^BynvID+&4P;YsJ>0uUkCJ7ytH+H%KkF>TWll_guE}LnkTr1VuigVsg}lQR(80 zh!jOWM0QC`V%1?lNq)B;Jnxn3ychI?k{T_|77Af}VXF>p@ z#h#=}Iw`7w){ELbwJ5=r7ohjji_9D;{T}&3JXQd$;P`)#&k4{T4oKjqC>9h36b39X z9Y_%83sgIY;m3?03V?D`PtB;A>&JY_hs1YpexSiAM3~C@G^@SkBx@#T=iv<*O(@Zva{%Z4z%M?+#Ew7!+;udcFHS zc6nU`@}lL?HK0A5;srUKJlI;o8TRoazk|t6VpRwI8LFUo@SFWEgWxbQQVwy zXrgmL=LnV_=L4+HIT2X)v7}q)seIbo1W?Tr6ob=()9EdKZ35)dGtHrqzYPhsGCY$5 z2M&%_mn(N9IMLjXd6a!543YC0JA2O`nr%hO-8Qwd%gTdnrC68^Z>D09Fb%7>wOBS~ zq6cadU_K?CE`r6VJv<}5BQ#B#9ja}WKph`&10iYah42R8sc~A2ZO0(oL1J+bcAn>~ z!`wh}_*cmreJn%IFE5Y%Or&3_yb`KQk#-N0Ch!;noFzHhpyjrra7upGxvI z`HD-DQu#suHG#V)g0eli^CHj}Sz^ux`J&)a=r=JLe^`ie7Hl5#oL67j9wV|~WcjAc zzap!j8H|9)=b-A~EfkwZkqr>uLNJmPbf0h1Brp+<8<;veW6ELg&7PZm^~q_J#*PFZW>|H~1CzD}nuZ$Wlm)5rjEFqo z8OvvS2#zfUfwdvHty7{*Uz4)HYZL+!&PF$nGA=ebbT@Izx_9tv?gv|3NZtIk6(oD8 z8IBtdCbd?MqLgBHQ)CAfa}?GkVfB$+KC8kfO?n8}z<2UmW_HMMC@~dN(Raut2{5z( zQCR%`?oNY1QgFI_0~ozP9&2QomlcpzGgs${b;q&z{*WGpp4u(03dr zl&t18xZ`YW*@x1PUiv)$X_u={U}^#|a?NKJY@$B%ThQmRU^D(s4?ZN_Dos_Q(oK%~ zD(@7H&_Dno2_ry;_uN_(bvdgn8TNv3yy5KLm`Z=yEO z8?7MJ;lAbpv^-)(R~_Jl4oa|X?A9MzB9dcs!(SyQpR$PLrd34JMzNPDa-NDoa?;B1 zW`86mO;O;`Qc0f-U(7cs`sAgeHo^62cNwH!xaoP9*)Dvy36>{yKDb=jq``%oQ<`*f zl_-APDbtN5A|L{1j>vJO%7R{mcbhebRPR1K21n9^uW7LC7b)mhv$0_({5JvTFRp`b zibxOM6wwL`H?Xc6fkZ%wq}%(5`~S7zsBKotGiTBkNZk2;;4u>C#>FZbRz_Y+v1=%@%AA$O_Z=5<^VMaM zcVTfx=W{0HpI~T5%skp*3S!B=PZ%PC9nU%)DUt?GUd^s?U%x+X9v+GX{gYi2_QJFd*TJ{!(CEEw>_E`Ra3S;B2s;#WlO zRoFu0X-l#p_vD_SgM5&<`DUsqrMBCzUV;qRFy)E=YAO^^<*0M~I%K_y9DsKoef`DV zlSad-IOatLAZz*Ua^!OLA`*UCjBT{xwh*xsKO_elrP*rCwranayu=@cVMoJxJ!(sb zAo1vJESV-6bpqr<%xVFT_8qU zo-NT(C|fNxPBz!fo)m#b3>0b}BeB9YGCf>&fQJ-d>BuokqmEW5)Tt#f6|*w<+G;+? zR?F|8Vpc~;t;#4pNyE}{8~yg%sP=d<+5pVj2Y9urT+gb>3qjmVn-{u4e1;@2nSQO| z2gxo!DE*vxDf%!l13@2cVbC4Xg1pe=z|1Lm3{G|KyqK!)f~Lv0-^OWJ4)lvvhA%kk zzfjn#SYW3{$|0QDcpm5PJ_=iII6L3AljS!~J5=L0*n^Tl!k3WU(Y2BW!A6oW8$?z;P{kgo z<2q#`zt}IGzBy5QC}5AYBDx0Fk<#hAbP+L-T{F`ryl1YaML|*gm=Sasa)Ji;NM+cV z;pF|seAM;EZDIjZu}8*nvTUgIzZ)|q0EgwE{ObmaO{Pc^6_YHEosAuX`I3IHrtWHD}?lA~vb-cOuIenH9tD#Q0iekL<03WWG@6Ytj@QL=cSvgmo8 zI-Sl}Kpp)pdZBQMe*vgeZ^a0rlC^n zZky2EB>YRBt*)lqX6x)IW#-YVphK@Ds)wyoeE?-{Ciz$_cq*E3fM?>z?NFSR>H?d> z3nA&-GhCfQF6h8(hl54IU#rKzKb~vZAo|B79e(Lbe4LzlPiAtOa_>Lk5rpE zrA>g9PWQZ%1MTF)hV#8TdXM9{aP*1JxsDqyUQzt*2;VXog{HoCge>J&RN%(*<7}%j zbv?zdqevnZlP^8!p9lQhK*g@iRbTZ^3(lrD2c`w*@Zv}g?+Y@@L5#b?=(k_}=*-hz zT98L?l-`(*EneC?FBm7I+_FJ&fOlo?DBBobYzNKk92dahDV)5;nBT4c&2`%bucvA{ zW2cSF%pBS4lN*r;ins}k(dQFr%!rR@CfWjV0)HUU(eE*kQHk|+cEhk8n>mOZn>g6c z!<=~7K5mX?T>q88GNQ~ov*7?);l^1t;71;m_(-GJ4HQYHV(vgE1_%u@8Ny6;4YQ5j z74#`@&orYnJ#dHU2rS_JwIg^Vdl-aFN6(RP1cDPy{rE@O2f`2&C`$k7f0o%U?YV6j zYG=*yAi?hPWy0-1Fk2F7gw@+3qQz>IE?#NCZpew|2Yu|jK@~4dA{PU0sdUPlpc+*P zT%x_+C+K)83;K*31tkHv=#($fwn%%S+%RCe=Z8s!%u@E zj(YDO?=));nhx!s_p!K+-a7BSI0}wc&&A(5WdgGpC}k4i+@i2<_}Q$%$F&L0!*)KC z11pGF68-o0{{GHz{eiO#_0VKCTik7Ax`UIU+;=W;{Q2j97HAoh3j2RlPBNaESx%+Z z2)2u2izre+#hm=!MHs_&LZ{$a=oGvy|6ul`N9ut20BJH-lEMfiU+R-vU(?0Q1Zr!% z=C-`v^Ml#L1@~jI=M3aO_5+4_@-7{gXv_8Hwj-4t61p=00c;kn1PCD0{(=N?s;WWI zAj;yUu?@0|!d0OQ>gYJng%e;h3@1wEJ(9wx<)6sNjG`5idwY$H75xkiiST=OIx^qrc8jrzPGvD)uih}dl>FMOUSKb9*aHt1wr`yMZ+#m( zOBm_o+O9JjoD&hV1XM4Fo!~?Z0{`f#^bkQkfa0HpE4^yHK1z1qMAuNiHtiXP^|ous zIB7R1e4wrGd=u4Fwm z2vx54Db)LZ=(RjBZeEkrbmm4Fek-F7)v{oMwsb7teQ=Be>sT!iM9`Qzp~Eqes;Fb-P|3hi61#T%Oq$ZkV_*X zF=eaBe@AGQOBS_WfPQpY_%N>zGHt91X;KtO_sYO$lfzX!J?ai`nFEjz9> zHaYS1>q7Z~AlNZQ!rbNOc%mb@gs2WK-S;{N&kelIf$e|&mME0xCK5pSP`ttupF$ob77B2jxv zufDFV6I8jFwjrO#H-j|LAo;3HqWgfc&`36c0+Fd^T89P;m0?X%S9k(BsR~7o^NA*C zp6qIpMmEN5o<;3)<)FuvtWse0|@(CIiz_W-*)GNcvH=PM7wLTBM zOi2`bW%q$;8FRm|_6gjhLiaRT3lu{bg)j_GKRW$0AIKMLlbxNp%g1p*v2|tiSwF}j z&d^h=utLgI-^|ylZH!U3fWAwI8p;04)oH40!Hwj`1wG0u(u##W!UUc{+Xr&Zh5Q~E zzB=4WVfezU%BCX)$5Cm60IyAxr^)N*XYmtwnE?hZ9;p#F1Z-6y5|@0Fpp*jncS&a_e3=doX7A9FWm&#{Nom`&Q{0{SrEHPl0{GLy;4fA)smq z(5FJU@1po9`2`R)OrCLh+A}s`;_2m+w$+gBK1XCWCLfhe|2-8yS}>TqJ~w1aj9g*b zLeQ!BVf1T2`y5Cyk2Q7QTlcwj4*3h^iDX*9;|XP5sVBRGDp*?JjCk;HY%)3z>2245D=tq;pJ+| zeNg5%-=|Q7d51d{r(U%aCyA{{8DT-j+_VRpn2|C{M0V<*_PO1%&S~ImS{6h80*ddp z$sqSPetMZtGI5#$8qifZS8sPw4Lf=MDa!K-y4C?Q^! zZiVSgJb#BcF9fM$AhzOtOxPP?r}1X2?qms_;jw6}Htli#-`?%KY+VuitkcS2$_|FQ z*Gxg>!b3BTxFaQIp-4ACRW8gRV}K0^8MX^0c}mBRFC)8$MfZ=q&Eeb3mg66$=JQ{4 zsa+#5;bP?Uj*AX+C($(6Khe#GKC-FzZD_~dmoWGKXwGQXnS}q(nj`mFQ^Da8gMl{} z*8CypwM|vIQ1|)FM?jc~vzW;>52lWN@nCYu-KWx8WQt9cl*4W*h?I?}mCYt-h>mO} zB1@pieqU==env1DPjz$RUCA`OIvFcI#@1EJW@N;W-gAGPFJnTRj#|TTZ zj-cZSDi+m-4lMx3k$J>a=T+JgzCJ!eg2JI$T(EIFdQ@wn{@LVQOb8~(M*(J>5 z_jo3{$NCzyYw0HbdGUIxOQBcaQLd%C6!#q)Yo7>I><5fzTl%2;ds|*c1@6LcU6%gE zlxAwsrmJpv8okQM0$vTdQiS#H$d7Q8Yy@u7dsH)N(3bk0pgyS7Tnh$0mQx|+iY7%P zc~yB~FW2UOnOqC zO02DF7emzyZr%?3HX)Y?r%K9J>;=Xn_HSWDjy&LHzhs9EoZlv4CmTGV>Y z84{Rop`ho>@UMewz~L>CCIHpS<~BOef&K>SEPyVS$wQn z9m_J>7iQM5d(QqCh}@%ZpTRT3>{rM7k5bD=OLpb39m=u*(*}Z0B&fASWV9$j^~eeT zf-x}>oOTGmq}I$?IS14|CR5*oWiWP({<7z3=CSNr&GO)2Gb*xFIUA`RFBmF5wLnER zLBr&F9}(FC{>`9k7bp*69%5z)l3U)LUF8zBsB+;6TCdIr@;C#XG988=$SeSP#4g}p zC<{+f9((u1duE-9LAQqR3^HA$D+5}JTCNsh{L=zQD#W&BxfF(d&idUpIAqZ?{4PqA8p!Hbmi^e%FV;i z@pry^iCWL$s_|VGL(MjVhQg~YMC4xnm$2YBtZ9T(pHnj*NQaz;oOXIN3y=6DK$6la z4_#E)2XWI=1CEl-#4IsBgTP{kK)R zXCyJBz$j~zt(>zWu$TZIoL5%J9SoLHyVWKHje4hJ4}QLk5Lmk%oQ=gXxEMxZF8PP5 z8I5FT*pP3Os2GZNcL-vA&%)dlPyG>3U07T~<7tC-Ao!tgr}FUyQGxXW+#Yn~=wnmo zk8VHUsN(>@%6t%Cd3dk1F5`PiGEo=`{2pg2nS#AK9H4SE`~9-*Tj_06|Jn#4o7%GH zMM){|ORdwBIBP!`5{^njJCzq^HK}y@Wa{jLE{6p6CTX^Q zJHgP;Mr9*qB7)F$k!$G#u%0 z@3TQK4#(0NN_(-K+XxBFH|Ik(bF&;n9eJ9wKpeb>D))gByc_ZYsOe4=7=f$$miuba zEm%CB6cE`88%t+&^sct&QD)dUZ8`M^Ynfy=!wH8YrVMJ&xTbfGmnT^rQot+ptroY) z2VAPf?Ll|KqeZ0=7=}+#R!ZyV;?Zt#MMP5gM*ih_JuW5gNMy9d|XgOQG?C4bGj4R(WR_s#8UY4BCF`k>Z-N zY8J%ssf`lTcczTXa`7)qc#FcwADILNDNxa}Lc?5`nTS9xt8gg*a zR?lAYVn9o1E?E}gKv0bABlZ9pdAH-Xw?DdY-8%Np5aQ0Hm(RuAhZ|Hc+4%bQ(96<7 z&`-MfdixSxbYuc=tLG_?c1S3^=9vp+`Td}B-Q;?G0VvZ&5#yGWv-RK2{^^=6v#nn+ z&g+;xgHhH&xcjbsl5s57*}Op0_*%`MsTCYvpaB-&5rtXF1f4`s2}I<{z-}^b0jj5T zk=-Qfr~ILsyrwh>s7#U(c!DCE@%rSB$=P6q*D2p}E6%iL!FowN5e6~3BR=P3M(ww3n0*y&AwFxiw>z>m}bLa9%r=MWQ_N81I3!A7Uz|i+3zpPIM~o zlX$)9c$RCQ^1S%aEUfj|;&B6XdrYk%XL^IoZjiJYYhyaArVH$a#Q9P8ei>mlDx2~Z znN-Czs>R~nJwnij391H*DuKU$hAxrUM7C2+Ab3*lvvldh;8?*4`T`$BP2qS9f&VSV z%{Xn&+}1ZB>8rvk-lY$!U|Qd(P;3K1iM8~Q6KZMIz>92ztIh4%>q1s= zC03oAwTv>Gm)%a8C#lsO_F15mYlP3TiJ+4SDv5|}la(ruL+&C3-EjGCw_hTKL`NB9 zm3KZl=!V;q3J)8RF>Igg)$6vj!wMNwvR6d?Hq4BS|NLx0303;iq}rdhz)>whgR=4g ztSK$`Oamp{g4sy|456eb>-=*-!=l|hm|uR<{B3J7ZqBX<0SHD|mnZVtgNu^+SRuHRUjruP9#u!x z3hDw-Xb$2gkSmQL*n^rH@N5&Q6(KF!4p0<46SP859Te~NwQS|00^aIr+0*vF8U4oo zH)E#ZeGVd$wg(7oF(YiRVP$p*-PA`5U-U`zy_vm&BUFh?vZh)6P>{!871kGiz}F~j z7FLZb#?JPMTU>7|EG%xCgdJ9}nDXeS_d>+x(I__k%Fn2^9QIR+EG)@Zf=(l-O+;j} zDxU-{anKiy6BWybC`@V{axxvo+KG6VDna_9vT)3d9hx+c)>eEx88;>Y8~7FM@bN{@ zXE(;C^5ig45(AazA3y$Iy*e)hMQ<<#1(upql!t-&1P^oomuK6*#J@g?&MiBzOcGL_ z3zn<7Z=Kj`Qd5!uPB3rJD+$Zmx}uuYb%I?V5u z$GF50%ZO^SOgxET8SZ|Y5i@yU8UCid_cv{_mP2O~mE|xl7{t@pxHV|dYz!-eq8&l` zWE3^%Qyv7Utc^%t1QmW6B&aE?3MCa@$EaSZ4mrEJNr(bA%M)PYQw`@X&Mx*g!1U;G zQj5J2Wiw))j7SU{c1G0dbH6EHY_=pV@2|+EDmmrsnz znIRd|JIEHO?aU(&QOF(u9UPSFwgu)fC3G6C>-Rehaor|GnRv))J1>D(=Qe<+JPz|e zll4gx7nVEYP1&HB(=A47rKFivp^p?v5go!@O@S}ohaux!SVS6f>Qv&H91TX5^F3>M z@e}ZZ?S_ugK{Ia5uya@T9REj`nPw~*7YARa(x*|C7F||E(AxI;i(m3k6HbFHY%jS0n*UDB+L#9hsThC96*dspYc`NYKR}a2&jXVR)=7=hn&q+f>9k9S{4r-1-mcS^{nV?%U4yreJIKUW9!~{KcXDqCTU3MO|9dof> zBxEyJ;&9l7VM*~Wd0KKl@O+?dv)4+mOYk4W>G*ii&uJGpHwvscF3?x zr`yy>O6Cu4|G;cbGKs-=s8gd+;&8bCzp$_wmk7FzpbSJL^09T1XCpW;Sg!$LWnp{}^sq)T{V6h=&T&o9UC?KiJ zg|(xtp1`CdfoYVi8yXoLpp#YW1v|v+1-OcR zRii^uj(V!c*$69*cp4{F_=6x92Z;0WOO80(Qd`0fMo&%31h>Y767z8N@r}kpYaWzm z{TU7ejDZ~efPX1pS15v%g>L0}S(;zbn|ZJv;D9k@tY>c??`AH-J*pd z3ms%=`H@qTuv?tzTDJx2Q4YbE8`mj44YyT#AWedl0jY-eVJm%IK|5{3t7^OjZ2XYwrb_jZbV! z?=GrjG>3=7lcW4!B62&q+Y^fd4ccf?y1Gu<1il4kyJSi(OY=nC5w-4} z%0Xo@f1h-j=y<3>n-G)Xs;U{G*$SU&B@%1za&w4 z9QI}EEjqY@pvwqq7ZJJ9r9yU!loW8D+Bn)c#9?buZULq|f`)8{Osrvv7D1XzR9Kn# z0$nC97}4nLEj#t0+PRPSn|8%@3%vUJt-nf z(+TS`7Xx~!Tfw!wbX5+dx}H>LK(W@NS%WYOxgozq-HW)cH6=(taw=7>|HgIKBUGly zD2xZT*8<*3_X~XFI@s!oceHvH@Vk{s5jW(6{D$yESYO7Z>V!p?g%j?x33beK-tf^L zHk)J29W%<{&Rw^0U2M%iThEYQk_a#Z;UMO^Ah}C1kwmaWYJUxnZPq>k2o8Fn9RRf{&7zAO6x==k9y#`lp^vFxtwgjcI}pR|vUO~yv#*?%mX z5Nbx)$%&%G`u^;!+rJ1fqh@z^X*8A34mBM1bQ&yBvzMU3^(iGHHv}D@Uk=L%L&_cE zVZl1FE_ora^8^-3`+3I&x&qBUSEz7QBTb7bCw!MUT9i+ADH0*|YC}*riPSFGhOS!< zIeW0EP$f&c92sO;#0j2dCj3 zWM6QKvO`!E1lor#-7p_JP3lfjxZ5PV?y_#tshL=|Vg!{c)AFNU(@W}IHq*K*p7>p5 z;pfHKZrw1IuUMEMfq-<`^g6dyZreq}zXX4AxuV_+L?L+DZkX7z=#G>8*vI$Nj z2BFE?WaY$9kuSJ3|3xpfM+1Xs%h8w>%GASa&`8{GzY)d8_uE!f*w~!iE@Ov^_s+kg zvR=q$GlSyrFvYOMjr9h3{F@Y#VXb$2Ak7E=WYWQdy=@DC?1l;!4gJNN_ufPx@#1AY zGz4ZySA=ZwGeGW(ctr>%LHt#c9&Ccdr1R7xlfhN>-PFezFf3dRvfZ=T7m=>3H53ec+Vj3@?)kU1!0aRGvvB<^|TbI7TM>V z5~6pj322rV^ES?n7hs8lUcF2($nPd&rrFnu*kT@R^fXw!{9|9rYJ4Fi4}X?rP1o^k zq7;Wm5e8C@1OBI^hdhC3szu$Y#!R<6?pbc8%4uW-Yfz`rG-F)pTK9sIAol2tV_qjv|14Nn|ncEr_4 z_>B#mF|J_7mFNF?GtqiL;;i*y@H!Hm+r&|#le2e(_J<$hff&$Vg`yw2z?SYGqo*8l znz{Q&S&^8!wD4Sxk9j8jkUkMh>8DW#Ey5Bz2paefwiA&x!bb4$5=9-t4k79lBOw!# zpd*j|ef51chQ}{b`ExTv&Ma(FAWOPl-78HAX#*m_tRG#aZn-tf^U1yP7O%#6O$vwU zbJ)@^Bj2}Yw;j?g?D!72>-y>3Dy!$20rUFRf_SQsT_BgcYhct-3*7G~=)DA0PDGY2 z*+FLb*N~5#G6mhrCV5$SrQ|$R%$}uAsuLu~e9wzP7gN{}&>~ zgXa_7+22qx9FE)NS-@d4L8lPZdLr@*e~j1_f?x>7!?3D8gG5rzLP^RnkE!l{@Z`#1 zvg&O6_2rl?u^Nb`Y-&$TaDu3 zn9}NFK7N)L*o-2U?N0RaZ&ZKlr^3hM&ACm7sW=Xs45;WIv7nPi(3=P9=p2DY$FNY}e0i-EGABmOZV+5U~>t>CYK4Wx}O4mlLiG})@SFCL0l zFTQhBXC6&XM9dOUy&N7*5-k>vqo>kC1oeQ3EDounG5)%D(arfqWVd*bucr;#BGEQ_ z@1j8^uIF_KD7=_yto#mU{IH;R8tf2NMSmIGkciB zL|mWuNU`0mI_N6@l*XWK2)Ia92SI)6@Lr*$mRBsO_l|MSekDn=BJgYBpt4Kb6<8g# z#rcdoezR#Ngp{bJW5`8i(Bk4RXc6(S@~wvdX+%ERD^HW{4XO&j{Lx2>Q<^KDSUj@) z^+DG>eqT8LijUGfwrNT|Pf0)XY35gwrOqZkykEZ@3&@1W@517(IXpNUhGXFC=;Sq$ ztsYIXe4j#59fa}QWP@(0v%54{sEEh+Njuavko|<)JIFk{uprwolG$#N9#3#>Xw(ju zuo_pVR$ui9QJAe%#w-8xD=KBQO&j#!1) zra4FbbsK`JN!*AQ^+KYpuFD0o{rAdICGShB4@lpKH&^pE`Z_FVh+WJcJC8AhpLPHV zLlex{N%`X{p>-_o*)<^!yEzOIwRGi89cv_pONvF*~pa4=G~!cIP3~tvM?x31bvdAjuVmS^wek?+?#}Z z`KKi-=?qB?wb50tz7l#R6l$S6)kejBRYk;rc!Sp!;NjTID{!9O<5~IV#IFVb!CT32P`AzYWkEYC%-Cemp^9F!IM-y5@N|?4YKXx)4xGo>S?rdUw5wUQmmg zy+yIH-^PKJC)UKiBbDV3e*Zt?y{DS7vSM{$1CZyivPB#=C0i}9vW}qR2`UyVV;^N;u~R_9ceNmOZj-{O*iJ4B zIxf3QCIp>z-sPSs*yUc~e%ITe-N_%I?m$*PB+gA%^v@=MOp1}P#NDizVY4M|1b|PVats3sCEude(qZg7~KT@IYD(1 zk?G3b*5_T*b04_k7v+;* zhF*S8uLiYXyaF8EKE-LT0gp1p(qz^3H|oEe_x7jXYg+R0>Ni%*TpglAfnmHY1u%M% z8kB!2u2SpT#rq@h2PgCJ6c#zHdb8opr9~l-Gv=}^q*-1hy2PstZ5GCiV5wqnZ7fWW zy?36N9A<2dHZ%IAIQNv78E-#6J*AaOW9RyEc<|b1fzCpL&L^lGB67WZE5!Mai7P;nRSjpHxMBbaZo@!Q~1_I3P-~_tVa~0JgjH8x` zqC}11Ur7goZ;8`9&f8Dj`N?a^yur$lFgIiOGpq~=QL^h-ArtX}IXo|7@PcEW6rq zeQVV+E6Xu8Zq}kW>r8*n0)wFh0VD888itW~Bo0H0sWP!q5l7+CzVOP0ol4!~U5v@& zX-w8#A=kU%&3J6-4PIsl7A>?(Gp0MlTCfDns6MJyU)yDu6=0?;$e$zsAF~^BbK$x< z)Xf)c6?(Nr{>;||-A_<`sFBr8ZcziNlX5_|&AUmC+4;cJ-3+74*`P!Lu1sf;m7qm} zoIb_;YVkEP-o07g;+`){6qwe4asfTS?BWz9>e4I^-0i2kB{<^Vsye6AtLwdLRUJTj zfD1`FOrW{JJ2a~u*j&J*d7cs*#Cr8dtn4eQFnv1%=i_ocWJXcth<>jJ;m^*lA;+66>0xG;4t^nj)gcq8tU zYj{Rwp{Vf*;kZ5UI4VfTe{dMj%y{(M^Tk%{@b3#1%V7xee)9392qRe;aV`|P^$oN} zb^Z=9L1Wkpn-DQZ8;&^5jEJA_m>y-F6T>F_%i*{mL$-{OY?5VBxHxonW|>k~474zV zu16JlP$*p{E(?!yM^>0@-{a(Jzgtq>Zof~}SE){(kv#2{;dwID(GHSL2pZi!I086N z+Pw4HqaTsxQLF!lsuNT+hbPaO7L(`o1ig-+;)zIH0!$1`3~Lv+L!9Iqc`GnI1nZdp z3YNmQZhrFgCvAh-P>d3On>zHzfBS};85IB7PJKWnaTpXkEkKb)&`=4TPDJKvQk3hR zcQ5$bt6p$mxcF9wOqD}Uk8==iyXY#Vx0L86REraVnt9L_)ngnN_J7>jwqfL1Tk!aP zM}4xxNN#WXcV2Xipo9n(mG+W7&V_Tf`ZvjXoa^WJLI@V9I%H4x_39&`Uj(lf8Ps>f z23;q^tc(jIjsnJrc3EMBh`aLERcnmAP;M!MZ83sUa4zTsqqNmM1!UhIamMuAQl9SK z5@f$knXbb%N0|JC7lIS=5Q9bU2P{Km!kB@fgy`pPNs@ zy0Lo<5ko{xccA2u`MKND@6UQ4jyDp_HewVXb&B!l-;~JAC=uTg|B^~%w?xe02`=cg zj#wkiBxo=sX+&hUfAk!^y4*LFzE9x%hXw!Kh&X`Y$B@ za=1rf(Hks!+TuP(@31l*Q%`)TzyEWy=?M66#tiEE3l^zgVG*hCC+I$c>LDU`dm70a z%@_W;5|E4;7BUQLB(aEJ*Z=B6WmN>=C!ahEeEX%mBjAkUY*#mO`im^aMg9kU%czEc ze!sPIbV(lNKFHW|RT(AHedd`IF%%jNDG9nmObgh+3pWrI|i{xzF9%m<83x zH$O5&swnl;8DW%5QQinGqi%#AagJL62C5irm65E6p)p0-1>dN{-VY&v0`f;m;RYJ%-kUw@rCTFj!0OhwU4CG&R?B1K>vA}) zi9tsXX`bRBFZWN^lx0i4f-39`sKCAzxkGWrU9VmlFyy<5$oQvQOFllbq(gWsGK1`) zV(%Q= zBr@9NfsNUS%{@J4OZL5h)qqBEiHotPAV6!Z%Xy2b@(F^568Q#@?t__d zn=C`pCM%lx4fC;4U_90#RxqEsBBAp86RC0LaF~Xn#u=$zYJCpPs!XKC% zWR-W7cbhCK(C7^_@j+J*C2H0z~$pU^XO{XC?*?Ru*Q8&=^7dZ3&)Xev?T*28N+7 za>xl4qbJRjjO<(65H#|R$=P1;Gs6eg6J<7m77mX(3~^#pqOD$yVL#0O&nLHOv-!Fx z(V-c&yiCa-#^ZsEzpre=#jwp7|Cfy8&a>0YfA4A@ZYuxoJ7=gh93E~8EQXtOg5FHP za~U@f_n!ip}G-iGR#&Zk1kz`Zx z`6*M(=+I>E{XO;22{Oq>uEnh-=s1E}jbr|P`a`{c<&wdg4%isRw=O6q=x{9Ee6I7HVLnJmm4sH8RfwgDB)!zzyfg^$XL>v_> z%KMaPgZry zw|KO`XT{u-x;4XL5@cZ3xi-X3zK9V zSfPaY#{9qgck3lQHscC+*AqY!YNTh9N6;{Z%>-WdIeVlf?lJU7(W)$oA>thM>&Hi0 zv|r5oY~DVfbkM(S)ut#-d#zqsl4jvU$alkqrW{uXYrc#VR>R=3uY+RQn*X`>*Q^8T z43X&df_*-y6soJ6-7Y>mbC*xM)WC<3Q9Psw-r;q5U3XnMXF!%D=!8Z1Nwe+jpR47+ z-%)$4URUlerT;b~?bNG}BOG&oznUr8o& zhoXHaD?g*w zau_}!`8=XzXDdOc5!5CkGHcdu9;#CwC2JIo`wJh!a%>;pa8aK{A=D zBrYkB_)Lx=VNK9AX;E2r_&D-UzlpTYWqe6|6^4{dtnWa|d-Q!vJv)Sna}yz+Ujqe( zz2u-Bj%QgBDiL4M30=nx_eS^{sK>XFlhfG(k4eT)JEK$sr4DNoJZ2s0ktVM$M)|sxb5{83u1GP1wo;zqR^{@k;`8}Bv_7Ea(ycr%%YpgQI3UO# zPS15>7*1M3#g;v0I7yCAbFsEIoLvH9uq+)yBl(rIFC6Mlf&bA6&`9#QIsfPkC^*J} zrGvy`zh?QMYm2(ddmO!L%M*U)8rD#FrUl0wV1~jd4(*i1)VEUo&DLga#$S@CJPtdw z^%j<;f}nwQZ5I(46;=~)j;ip&DKb(p7V|gF$(#D7-E`a+1R9Y80ObPlACNA4Kek1%{?rL1fCBR@K74Ex`mVv*=sg50&Y)uzYnU z=+=xzGMkUuxmZh@;kOks>XG{mgqk7Bk`R;>+(@pThIizF*~t`T(hme1`7zTR2(j@& z>#+%9z=a(Sr*%t%T+9Q}C%Uu0p<+gJdpSHS0=2ji(U#2wokCFSiO2`9_mZ(tY`$~J zHDwJ{o1dU!ryI1Tyf(U#jDM#Vj@YX}+{R8|0{|PFw9948w&m=?AH;i@q0#p0xsRyi zmxfhmw*_c&2s(?P^h6}A=g0@BLdkwu@PT>GhM(!~7SROt07UlDqwOs6gE3 zDz(gci~kN)H`yM9hNklgW3*j*Wt2wP0O+yZFn^BHn*9#F7W_+hsTn4Uzr58#t(!)b zSUlehg5E+a%{N{WVuc_sop!P@`Nj#n?>XhQS_=t1T zvPlYqj^okV@A*BaCFXZ?lsN@=ZFJ6$mX}#0gtI;?19Yzk@1k_qJe%cRP;P-fYCV(sc8!uVufZK z^7^?=hpD)iW~kY2VG7d-dJ{n<6OouaV1#_ESY^8M1a&qjS7U^AW+YhBEuVWHgf}p& zpxno?q!pIk&k{J(6%fUJTqL-WZ#{x!L- zXJ%{AquEWu)PUW}afM6A*%@2~0JAOm-=DnxV{2$|c4T3IL5fR`Bu$9b6nkHZ<3Y%y zO0ZXUT%g+`Y+Z=VHfbKG9PV3M0EQh5EN*q^4l@i!@rI`MHUzdpP+8ekAG8_^aEY|n=MOpc|XgdS5GTm zk{U3`ubo{cb!ZIo$s01RSMSN)j@w{m;ki5CEt*sJExCC%bNrp}UZU1dgY~kJVaja; z4K=!3h{$|0S*3TY3BWhfbE_wEp{Ds|Kn4#c7^f)L(wON!iR$ftTEw8{BKXrX8_ zyuxvTWD=&9tU$uuSwTdrwqceTBr86e8Bb-rG*e!nLK+d0-$~G59SeZ7oTyTtpiBLV zqzC+aJkmg;Pltzc$z;g1?^ITk8Iom!%PL4O+3HyqTIDAG;snxl(Tbn;MLV2$#L z$N*3`3L)6)+62ixO&~>&-a!sXB$;vshkb>EV3Dr#fzP>6f+XX4AaH^X;C8n&C{oc5 zdKLCGLH2p#PddPEkDisO;qHW?FZyhiwJ7?tnUCBxl@xxEILEry|5g03d014Lw>a+iNE+4SaRX&UbineIB|dz!aVwmh)Mo*64a-5udYR2aMfq2&h$?F{C~5lY>+ihm+Y3S)ihn zpdm@A2qgkR`*#jzpM9!M^N1HDtJcy-r4535L9Nd^)%67iEuKpV8sxVJt@SWy?}V2r z`uLsB)YK4cY?UGH1LVQ%&^h{JH%OnN`~b1$O9l>n50O%qZbGBJ8!JrPh2( z475e)s3hDqmf5RC2wmteW8gamBR*#;$e#-W5u3a(_onj7q zFbq1yrr;|k(BhKvZF&COWYrlls`+%PAGjB~v%aQ+u%OR*x~?)k8F~+=DlezI&)*Qut{-6+7k$&`W7{&r2$D!XtO+Cyi;V#9YC4- zj$y@~Ni%e8`+H=cIC7hnd7Jw4-~IA!!t9MabbEM;dgR2a0q)vI2pEnGtRxWh8iI-; zBK5NB;X3vH`E4`|k%c11>lTmJ-L}AIwEG-=Z0W0KM6Z|u@ov#ie@<NV{TSr%Qk3bcRN{AMq(>DuY1W!mSYyZUX3TWI$RHuUB8bo zV&m^po9<*~+NOH@e|z^ts_zd%{< zGZvGzdV;PcsA?kefMzZINRjD%&<``KdLZ}VV!)<3SH;(rOV_(SkS6m}l%IN3fyA`? z5r3UdH$bhWAt3A5E6pHFTn4Dya-=|i;I2cGltRh2;r!VWK$fW-$P~j(iYN>3RF;J5 z9+)fH3nl0x4_6cBlY8ako)(TDMkDOh^EPd=M9*Hd``=z0uohS1Y_cqaTv9e~Yk(ZKD(?#S z0jhnjZntNyXSaM%x!vs|&!iNJr5r#Wn-Q4pmaIxvwfeXEcPg*)bOWxU1BqR;_k$dV&G^h)3&`*>(#R0@uvZ}-kwS@ zcY+A>R{xvx%N2>#B+jAO0%zjZ$+i}A+y8lr8np0 z^Aq_+GR$1Ntk^5KN}rY(vynW}!TNrNrpnsf?yD6aO;`nK_N2 zJsv$m@Z7f1*A!*$I~9(bo{azf8-5dQc(~)cgjJVtH%y^2{{DSyBEOf!Rb`<4OZVU7 ze{1nZmukPKTPCt%ly+()AILuC7y6p6Xq4#H+h){G4*$=1kUl=_vb!tw-xkF=nTMQwE}5W{2r2<`lT@i*29rY1aquH>w=BHJc@MP#jGyC-*XR%0=8hiy zc848hXVMb>cxak+r}mD85sk?1)M-+kmq<`Df*PT zq8;Lsfd*Rt$|-)0rdw{%URLg%cN%J!k>9Lw!3nx0bdaof2!puI9rM)5dv?nK-NUM5 zxEnl3xzykNZ)QW{mvyzDGEAfTE#B7~1Z^a!4kEJ31y@~7%+J6a0dMZGkT`0JBi%Q~ zqwoOa;9T$d@u**Xw!66K<`MJyLW* z4E~yDjjTqtGGKqGE^AhYuuwEO9j?Zv;G$kA5lET7bg6a7-7Zjj7JTD4Vu@E)P~)y! zBX0OC#>VWqCln+o&DJYT3Bk!Hc*Nr&h_`%}toLjL1?zSvdIKv^a5n%k=g0F;Sr_}T z@pZWicP6QITr0%pCp=-~HKNL$jna0js8-JCzJ}gy$v~af!;NiW5l~RgCxdDHw?UCNxEKLvS5ceWx7LV zUgJf9qm2ivfpJPtN9eq_%yXEM*}OYc%{1zgg?G|K&?gD%I1!l+k@#gU*Ige7cX)h8 zb}Cvy7&?KvInSWovM@KKFC33;l9Vb-+^d4l`wWtO!41Ibo5?e1N8W{1$xZS`GE;p` z+$7I;F=#V++sUVQt(@BlZ+$#JgUl5fwDJ5M;%L=TUM5*8X%b_ab2j9LHIgv_Ig2yZ zb<%MS3ijza%YL-pv&o0Cx`M8H z7ZEgML*^2Z9m0aYUmn={_GM|7Yp0^yYq_XXk>#4F$>PTZ?B?rLO|oinE8VS3iZE!) z6$8>bf5@nTBPZ1flDlCS7VFg&fglqXn64U^m9q(86NI)+-!Yzxn4$gt>ea1pn$fay z!#})F9eQbMf!Zx_bef<)C8%Q{LT*9_8m~!lM{pzX3q_jjmczyhGpS!b9Q%&;g3pRT-DJlX7C`h?S{SH(IG8LfRhnh3p z4S4}?dEh2qg;y8SrJDFwjZi6sOQcXh6sL{T>eaZWs>8DRd~%o1Qfy*ro!fxSdcpE} zeLX>}S&y(|>_6|9Se`rhUwT*T97#4)aP9^dfF)+6WqP7Fh#Xz{AhCUds-4{?yA`HG z#smyA-4bjF!orZGnN^{VGdh-!ANN>6!2jw2iXkZZ*PmRYL&iXGBa&-?Qp;^V{yQo*vOK z^Vl<5CVpe-+JCTy2pj9cVZVj}P&Kl~aH!)F>R(?TcyGxgg+Yrm!`(A4^7`jjE-aQ+ z0VPNyq}b?{^@3(EV`yXe$-uv9K&|@kZiDse@e4(1&9Ky+8qg#x_d(^iogU4?vop(- zIF4ZD`ZI}4BVI4ZY}#y09P{;ztuh-E;mLPpFFHd8x`{>pJ3^~mvZ(a}Q$-H=kb0;t z%^=%?iu{MQ4?t`Zm?H;Wi)CfugRbbTg3Bs&v`MfT5!Q|*V_(|m1T!K=8D6Gt+W3Ea z-#1TpF8piReX51sbce&kORt4<+DXt?2?#w#uBVbUrzKIYrJglpQbecnq4X+!GVrz+ z3@VVmSvWgy_8nLZxGb!iVbEUjyey3pU6!sEWPm^v-f8sA0D5zyVx8ocRIkP~;ra!s z+M)>HNASuNq(q$YPKmf8O;^=X2i+k)VDl8R(iy5aJ+ z1KV!TRbja#mKvRwd#%WBHNQdGf z9l}a-{h~Xr+3vM)xF@VfrdOv3I#fV`h|}dYZZS|Qcug7OqI(puJFruUGN(H{K;95r z>;gHtLAUClWxlNnz51{S>JWLTKk~pGBGjIEbmg2l3hvF-RFg-^>Y&tsbaj(_5Ewd+ zdw&+5H&>S+X;oBtW1>Wg676BMC`#1HOIGPtsdhl+Vi^S%Gg(#OJ|KI1Yg14P|1`BD zRQI$ZKSg=O*(fU;rtmOX&c7Cvo`bw$l(DmKYi^y1v(|~>>Q?zVaELpVM-^MtMS);NfVU?vFf;I^aJK^3@~EP~DA6icJc4oUbQMZ2q*19L_AaKZ-)YoP_9xPDcy$S3O+ zxtb=jUAi>J?wp4c#r~7`U~D+Ey5dUra^i1GhHA|>aemZDo*)ki(jbCdCCTPD39~dH zAOx}G1j$9-%DG*_v+hm8{ShbV3!r|JrZUL7gw;XuUU{&ta4^6KUeOhpXEgzn6M8^; z@O|qf=b+Vo>Jg-k93@-S9Td>9mhu>p23_Sr8Y_LMh=oeqS@{J`z3hcofIK6eT7iRttMb8-_Srpg7W$*lT+%42w4VAF-}gVzW%eVMm9dW(qzzmjdTWo>(`&U485&_cms>$EF2h~RK!nIR!7C8SB%Asi&{dgZ`Y z4MdD@&O52Dk)dq4Ufn-GYZk_2I)qm|Gr+@X3GE?k$hhelV-lin-6Ox7Cfl}vHiA*W>CZqJ5*BI%G5t`C{6 zK{MYbTjRFhFGYEX*Cd#X8rg9_V23?gQ^V|h1&XUN(tm1=n{Q16^paQ=2GXEfubV#T zb{qwX-OaMonhbK&oM@0#UFp^D@HNh7Flj5gpSjV&2RdJM`QbvdZ|nNi(G+SgJKvVO z5hSYB!nbWCXjmnv#h_w`uv&Z#%G;7vHRSM8(tvA&Y;aLBzff|L+Bi3Jx~?nq?7Ukp zOY`RzXwp1-Uk7Go87x`#DR%jz6C5`?DdG&&NG!!M=yOelJOxyahC(xZdt^hQxK@xL zx$2!D!GMjfgEV=*vF`oKPGxNbtViLO*FY}lIL9hmku-h>c6^Y#yLVdKPs~XA?q%mb z>LQ0H&Vv@%_=2Es64Z6f3?HD7m^3QT=)KGR2z6YxLv+`>hFlJnFh}_H{M3LGvnq)T zf%)_9lE59o&jpepgZ75JQ;8(o8J;IYQ)i^kIvEH>l*29H!V}5@JaX!v-{JPi>0!V{ z-XkYaCgGQbpAn}97=;ZIbZ%OBSE*ZWu=uIGs@&`UF!+l2T5uz2YO7QENTJvCk{hO% zlKC*(YLHzCf$=&Nv(l}9(-r9fNu{Khd@jS5;{o)1=rO^I)j&&p^rpDNnkDlk8P6EV zKKmA(dKK5Pd+1KONnQpk=1oxBk9&`tHo9(>nYQkFm-(2s+QqfQTU9Q#vet#Q02@3$ zPOHPzX%@#Pge;4e?7OphX!D2GmHsbOo*x4(0p?Jw7Ss&0DHp2pz}V~zEEHu?S3-2z z{?T(fA!TN#;?%3iSK!EQx9BXk7_hk6fji9B<=3GeKek2(8%x6B2n+*m`d{4zG-{ zHLX$c(6ZAr_yXG2&^^muhK^BJ(I)uE_ z;?-wsL>@aF7L7PyhuMgH?^Imv472y^`-`d%sLj);3JWK5J3;3X)HWir)~CSt3t-`X zAZ^j!eW&j`4d2}KAF*$=f3H=WtXfN#c$8`I3UnQ-$&29dVcB6VIUt7X;(;uA28e}e_Q$NwecxKnjk7+|g1%GGEhlMbs9 z)K~ZNJC&$4(5c){-VMk4gS8Rqi(-hB5L~W1B(PzzZX8dK=@u5f?O5w&^#fTMvne@F z%Ky=rhnM4#f8Idt=5TPO$zt5ABj`f}b&!a}w8jX?pL;{Ajz+oS17@Q-B4~&4a}YGP`YwlmP6H3 zUBP&^LwJbSAE8$_@?zyRy6AbRYK*z#!w3bwnfKs5EeS!awd zx`6>&aNya~Rm+XC{TkgLs-b2!S(T$PQMu~M6rJFreyXnUN_HSOgJmzV#)jqX4%=ge zjVbxY;`oU;B^+J{U~o#R{kPFD=>^JnFhaw%aY#R}S2qjII*1cB8mxMj4T@MbJi&_- zzm7WNV;*=O(kEgm{b(F89Pa&t7K70af`*j;?L_3TVEij9aP@==A;Q$F1p9xD;si34XA-o|Cf-5U@8*AZ1i+k3*yH{G<(!TGE}svOOfb2@iAu9 zj$#7~rtxOfh9`KC*|_+ooIh*5&dgaamSN?%P&DMk^xvdf1?(kTAuz!nWqms{1V)*S zPtA>=eTp*Mk=;(2C#lu!)|ojRWQC~C2+rV51f5J!NuU(2-XSgv?*%5rk~y79^jWt8 z`(cZ^1$@%2o*Uf^4qb|QwmWUbk%NOc&-}1I#s24MsTmoHzr58#t>f_2ro^JJGYEPM zL8TIrn1;X2E0=ua)a13&>ke5%EfZw%yCAf>&8u9xoir-iy=tgl>WXwA7`MyE$B+2T zvm5IGi#~rmdW_YiUCy(z9Yo{1PW9HcK%5;?7^-_dIKJevw0Oxi<>@6~DQaL+x?@R| z_pv2|{2Zx~YSH!zVqngj6oIUY8RWgNKB`pN8P*}hSBKr^4uP`9Y)V>#_lIWYW`en z&G^sRVP=1saAlXXSWj7qk);vYT$+gHVp0uML>Zt89Kj~2- z4BUO?{(V=~0+rblEgpKmg35epLLdK+y)S`lDn0j}Bc71FF=Qi+GD?MX&cRq<8ulmBrE5=5nUHkr@{oXgqJMIh}Xx|J?oUw~Fm0emypP2hqAgIpYk9O{2&X zP@|>_zKZ>xJH7Yv7K+zT(9`-T{pfJ?+l_)#<8OD*XzV&3|Dj?6pS|%l=I_h@H9?8( zH9sW$(xQ}5Mm#7r`aFm!-U4B3{^FRqT>J*!8g zUB$y_A4)e`^k=fbc`oAR0(XKQq|)ITO}z7@I=W&OtlxT|_4#zPNsiRZn3&3z_sK`V z=;IXbY|IO{o#A3Md?o+9ef^5ZS{DZuDH}uYLvK5HA1$Par|p(y#bz;O0T+b8Nl?M- zjYmS5TU8ZL~F27e9^+czHib6AW7Fto8|K-&x;!Hs*cV=h6_Qjq*h>v zES46F-OG#25j}QhVnhIW=FN`L^RwoJ|N6caDbm2-zC;>4II4Hu1{Ien7DVvQ;Zmn4 z)MVx~$%5IUN@!sTG{rTkt}Bp&dB6)NiYLcgMe`#Hpz2Mh*5WMDEC)K2U=#+f7MW;F zNdwVeRIv;$P=S12ig-^%x}XA5BUKYB;*pbod6-#44^x>K*fU1F_uhNx2WO4bqL@IV zB3Ez)MAObf=nvf{oQ0Aon8L8t=*o1H94A1~5ioZ&Khv^}3{H0F;<)=PJUNEbP!5k_ z!Q#*DMHru6H+gWmwgaifa_OEhjL{t?%R*X31_a3JgZtkl>l<24+6Qbx<(IW&%)Hx=NzMK()&`reZt#8 z4!It@$f8Njp+?PI@n-Q&D6CJ5?&qfo>SU*3nc53+u?|JX=t$UE|3Xf$`{}phwj+*T z_Yhxd&kdc9XJ65XtP|QBXI6be7Jq3no1hdtXn~$hv8yPO2~su?=q;9_FJv%Jf;Hrp z=!BnHsE}edL>LK;#F(*J}$f!y8I}q2VI89JAy%x^h zpV<}k&})nDvJkxPp;wZ4W$3+_6#{*jQFE5GGkrd3fq0MQPW{Xx-gEA zL1kPw(V`?`(`Ztob9>v*dOxl`=lE?8F(G-b`?&p=r8^m`%~_QZah9wcOSam4jSDFj z`ayH4#PY~OdaloT-d84<{7|&lMEMb!aSGdEOnG;@~B`xq=O^x_P2|>Loip(ofy#-UFO02(|p& z)rS+U6U@Bd2Q4LfZcAeiwjfnD)6r&%EulyeZxtlV=aBo{Pp(2 zQ(Wd84=$T>AXD`yb-aO}BQ}A?9me}d+uE4N(HTg%e!!pq#K!YuUz%w(BB67aY$Q9m z84=I*zsYGEBXW>p_fupam6+#qOWrCo3C|h<5^uhN>Y)LpSLHDNj z71IO*-4TUPK+^RcGj@kpy|#u2Mx6R#N>3RRCl5RyFpjVIw+(Z$ z?dum_sKk(inueN}E7WTz>80~0AMw@V46`!!f&^Sw^jQ}QswMj(Mp`22_?-@QhST^D z^#C^=bL_pO|Ec>9k7T9*v}t-(zyQ`|Mbtf(tk1A|D(Vm9y>YjQC2Y;Hx0%nUh5%A+0#`uTd$btslL zGg!l*HGyN(HVc_7VG~vA?nk7H3!q70i?BmILJx%#T%0r$*2_i&B*$f@xXRAWvo$?6u+8(TiFLG&x8EV;C6P4s~MYJ2b>Y%GhMbf;T$qY#s{I!~HW3=1GT_QI` zU}X-eQ)97$W$UCmYRIB-eMEbO*(X3r83TQFsuug=(DH3i z7Za$`LI=9rCo`m<8St_+!C~3LLoYntNATh^{%8Dqpk89Y>!2!EV7UW}yzmO03Uea^ zUPe`)92=$@MDuw4j3IKVAN~pr-VNT5c+ng>>8?X1H(U4U>ZjmnJ>WC&lYdlMai%%^ z&K{CCR4&qk(F6>-gVNDu6uXHc8>z$<%pCD`#b#x?pfpa43RZnSmiEYEX z-E=~Nja*0!rv2C#%M;s~wKEz->nB>A!2z!vX*)^v%LOuYEMU~BYQ~|T-^42lEec)h z-|zb|FOAnHSS0RL=zLutOmq5uKW)}+kGS~dbNbeL?w9ULW7Hv@6*j;7wD%yH>%q~z zJR3Z(pxAVZq*93$bXBy0u9>dIRjDBoD{hfQxyA*;9-KkrEL=H1c!W1_gR6h_y@gH| z^$!lE>r_A@h-Nw`Bq?Za)UhCQq85k7^>)GW`*3y$tmeV-gIs)u+YorHQ2ajr%YU_> zXgtF9|LQ20s-_rFKU3Oxhc)#RD6G&a7LZ3rUSd z&ZtHx(JYZSPD&ORiA2&8+8f@BH78!j0~zr}p;bZ%0N)XJ zPTS5_39%8>W^=WQX=K;xQPkAq2k|nZ(&Pt1l_hw}mT?s55WxU1HfwY|3(o|!?dfirnWQKR*#&SJ)nGswy9aZGIkrZRT~2@ zKyN>!k61I`7}ii=asXT0bgIl5O9N}w`~A4iGEa>fCk#27P=-AI)JJhT@ISwt_gaM2 zl_~EkOC|;nUNb}8+@Q?h4vK|-r!p$>W8cadTSHgB@!pA97iAsFLViy?y0p-sN*4X7 zVOAQ?sM*Y~rS%~jd{chfWZo|jZ}8pqGu&>{v;p5YhHkMM0yRnSzFttEuKR8p@6V4O z=t_IzQY>&9?Pp)iX=b8Mjd}g2_SzB-L>c;h`hE712H!Rwc!0P845CmB-6yP>SVE0{ z;FH7qfA$M=@mWsq-*b(3LrngL@m7=aosGZzHCa0r>edE-b4w_;h$4kl;%!Blw34cr zvQ3sgqe*l9yM`aFeIxhZduE-|bkj(~k3~uauxiDE*j3V-5<_@NY&kDqmYHDG91Gtc zVUQgcl?BwQ>t$zLU3BzhY&pQi4F>wj?_s~i2`-+Sp>O-zzK>q?g1R$;&w`NW5_!g2 zMP1*%B-=5?B+n#Qe5xg9VESp|!9v)#L7g3j%6OaT!jL5KDTO}xFyOiyN-lGGdqG3( zmb^i&O4;o_wDWadi+#Zu7+vkSpxBpPAinZDq^*fOXL~MP4*I#8>>?z?Yn{ADghoKwUMO8c0h&y>cGu2 zT16FKv3MZ>Pjf49JGeY zrPy^8$p#q{SP>_Slf-4b1Kwsatt{r4Un?I8=^&Y&PZ_`2Jt@>-{P62vC8vk-NIcj$+_EtaS17iPBA21aCc0`um9Qzg zIeM#i7WgAjo`4>0PHanbkFqV`#fp zPMN$qNSi)+18C&vfzq&8yj_?xbHJ;ew=becc{?VHLE#a6;zO@&-kM1RG4lg2%8=r+ zK{+>Un{1mjm2FaKE8_7mP_HKFR4~nr8}Ql_(M4b3C(YCz0&e*)WG&HsKH3Z5{^0Mq zW>Tl(77Y&VlcVadG|D(9m@ZzCV5MR=7PGoyR&Z4(R0yXyo9y$J+$W zX{+TALR&>gNe_P+zj(3%Izn#w^zid0o(yi}kD@i=285vp%^gP_17Qf~g_8f(=YQ;- z7YDOGCUv)nQ5#j;sIat3Ad~t9$zI^C#Z$Sl76$DBuRMO20=ReHd@q?Ag%RO45|6%+ z?s)gM&*U@`o*Sb3Y*)DUMVEgyP@jG6`agd$>%lCYY6JfcsizCx?Eluh|49Cd<>!?* zjsIlSKtBhomsV(Ws@>sba8*n6|3~j15`X?*?HSd*zz~1;rMs2nK7q4O; z^&KUgMy6Br8nT$1be974irfs9;TbtJcD6>z)qC8G+|2${Dtj&$2P7zYWVyf+dcr{H z+%VK+cJn(P`rSD5=7+x1y+=8{dCy%({C3e6-x#aq$P39|Nj8loM{K^J)fBs%B0DW2 z-hDp3Q~P3CVoY*Wa{(_OQ}h;XZwzLcMa1?4OIxF;K?#Ab9qMNJ>7WWmyLgu?;Lv<)??Q<4t(o_^gw8b;>3gpQwBm?n1zDX@&-jW zZT!{+$ZG zBGUzZK04JMC17nsY-$2tUI_A>Sxh#sUv@#bSy>S`5V|?^f^xGGAJE`xN?i^(d+yZf zO8jLXTkiWGR<9*2uLT$ZzBEanUK{7DlVVL2X$5+JML&bS)~d+|!>>)-!b=wS`#{lt zT68zEv}7@7B^iQ#M!RC%1!1G8BDlcJrO~A7fk3bI)PkT>(b=M6X*-h-L0xVCEXx7y zTGf4-`9IculX6u;C0=np^79zn(SGW0{!#GeoVV6a%Jnv1jK9x*6ya)+q>-(%4iWIY zV4-Hek2X(zb?OfOR>^h6;(z~QmNp%V%JDZwhb}V&`n0x6ilvyFI{(ICqv=YhJr|45 zyI>GjwSW5^??tENMpB*S!e+^Ws3N~|-qopFl?M}Q>5};K5#0oddur)E`DVdg#nq{9 z=p!2RE^Ris4!F(DZrtHChf!)G?h3V9qTg+-&_7K)ToiT(Wj z@?#RCq+0T1t4V{^xN~_RAQE6XdmBg?vY5TWH4`n@q68Kull0PRUd=>2yDKaU)_X;v z+h-gE&8liit{dVDTruXrF1Q?Ri_Y!&?nt`QJQy_&y3%?`Py8W8PyAf*8uI(h*L&jI zl|?ELw-Eec@R=;;B)yVf#Mi0jvsb1-VI2=j?qVB6wRE?+NwRlZE-c7$LIyveFs>+W zPuLu_QS+&IwY-~7`)6FVVVU_~-m7`zQBk;2vodI#-?`WB%bLI11yVH^fL9Nz$2;E~ zJPA27&@RH11E>u;?94AS=o+oAV5a>l>xH6Hb+GcnNF-2aZi=h0#)8eQ-o;Y%CURrD zMwy9&!{`qh8~1NIp0&c#4}4`Lt1QnAJvmxsAz%9m8g-!`#N3Y2))8<^eNl!4`9{p5 zXhnV|-(w^jSmn&p5ygJla-?!L7bal5J85P+P?c0f!$BW#Y6 zU)B+~#J7YRacI&C3l5kXWFDTi=^DIk>i@9#?@MmP0uhX3mr$1jl4kbvE9nLh5|}S2 z7GsSj(l<9MkZJ<=l7ey=?RwTUA&Ys!>&$gJ;WShuK?5f-RX5g0_NL$NB$5Q$1P zsF0NI-7Kk^fT9$|QtW8FGQErG7y^_uD8Q;TS>bGNNS%glNJEe$^)OuLIX%baUXpm(1Lr)wmwpRw7 zi)vOP~{91E5Tx$V^hlcB|aq(FGLc=YoI0TQsfQ8$UHS2Jv!h@?Xw!Q_!`u>U#a6J2xgoMCyYy???5CLSEPh)Aq$$0 zem*jXRflJ@jfxT;E^OBN&kH&tDS`RsCMY1GNY(~lK#_TwbQB~^niNM>jd0ebbIMc0 zI0B?R^(&WA==15KTNv!F*Qr+XD#Nuqz4^$rQ{ZwY@OD1?dw+%{@ZOs1NTHVF4S6(%|2 zs7$#z3aeH@H-gy_-c5f3Pfqun2kWUK*h1zK5D};nR?sVhI@BiR1=*HxJw#p7{qp#j z9v{9zaxuW{K{sr`IR@C<+uy$@vw}(ax34ylWgZ;KE42ZRo?_QhWDS+r_Lb}J?)uIS zeo|ER8y~*+*LT;mP5jfb?^X%(>1E7D&`Q&(PQG>QhgEN6hO{XPLn>!f39&$B_`~ev zI1CpJPCkCvmpMkm5HUH*cf!0szGp?qxgX4*Ms9g9LY9vi9A%p~hMhBpJfsrO5Cb?G zbA$(>!8Kd{sraT(&BO-LR^>y`Si9nrB%UkYC~jv8)!5>C#V3u2dmCaKU~jGJnEBj> z*rkC+jpYevC3|^>@n6K{LV2E1Q^>!eMB%=6CXJWM!d;Togq_}c`m7||zaiRk%^p#Q z(sDlCZ?zaJqbmFkgc~&lkZQ_e@To1)3#J+9W2C~T%IBfiLDCA1vB$`RDMhd7R2|CW zfgR@0?&lwkEB0OL~1aD}&S`e%`!)qH$^@47Q&1vm5I!R|Dutqg+C28zw4$T})9C!{>!9;txf zZVSm3oR$MwUqXYtE7T&ZRuj=O^S10##FZ(RWm^Ke=~~jEKKZpyvS8Y2H>Vd61fGhY z`K8-mnQBGJ@2fO>Ny%7p$maF#q1Z}_YzKZsSr_zz0DA?LqiInZq*?x{x3+g`oeX6v zJK~CfqE_3^v_l{ymF=MMiWcZ##bq+)me&QCEqObG&&iFwJ@u?Nu3RjQ{`uC`MB%cIMPZ1HZeG0q`o$VqU}K?iUg!Qlnp`q`^5x^U;m zP}H&whP;I_@)m~aRG29|s%i_+VvDWC@4?czexF>anZoSUj7@O|!qY$)7OynW4T`pa zd+rR{jco8oFuKpVmD_BwebuiGf3!|FQu%)*kQ*Kxu1m8C*CmZ%A5f&9N<^ZoB(cT6 zu9n`FQAU!etKGmQHT@9JWO7%lOZ4c9~s+P1tc?#B9V#eAMG#T(R&^q-6$u6kf zM7KJh80Z_`*y`UG(+^eP8SxE3a|d$v3ascXP}eG#`Wu3`y_Uu^*M~PM^3_Hawj5;p zn&hbwa2Ulbt0B#m?VA>TG5CZK^vQVxUPn;Tmg+ea2DQiKCuolre`$bWel zl#e$mknYY>Hrmc$*(l~k4|tp8%i#-Pf}kJ>Tq^cO6hQH5ZY+5CBZJ@|6XfDeZV>dA zM*rnSr-(SD)&VX`;$ed7Xs?V-Z%^oBKA*l{XQp%j9FHY^UXl3m3sU5OQ%RSx0RCTt^t8LXRuRTOtB!Qluso- zY9+bmw*=czkfgIT4y0{V7%n(K8oW2ikWqYl;PJqZ&D|`mqLhfOQ#R5WsX`al4u!|#m}9%2w1HS($&#es+7$(DC+6E0ae2L0jN*34g{>O z2@d$J4QZP2&?}e5|89Z*>ciTZeg+Glps)bklY4Ye%G72r##r~BfUG3BomhvF6W(n6Tgks+Zd)CKn8d4CaPj0EV21_l1lP;3)L8lhOvrzd^~Z?3pf zW(Xg=?P^AcV}o=9A2B*~yg8WB;j=fN=mK62c*OtkzP9E&@4fGQyWH@;AdkN(+&7&3 z?nr<353}rL&>g6aV~Wx2TAn3+KF~!9`NP~;?GoSp>YcJ?Z!BzF?t5PbaSbu3Y1YyU zzz0GeGN>ontwu%Hbpcw72v)7AQL#NR+aIN)2fRM^Jwd|)7ZaB*4C1@^eLML!pMM!A zgXFn^_}2ga+F$HvEsw1(cQBn|F0ojs?R~{Gp;(HNzHXjUhmYRpBXA^Fa6@n5U;nh+zR%5LLE@lutu~@)6879z zOU}L4&%Z10_F4SK=nIAKRg;sU) z3mmOyNI`6#v|X74LZK_XQzF`x2S{$ft-u?^B-*dsCL4`$AI7M=uo1~I5;$2AD*2^f zO|s7bJzbLP!M4HyadtnPX~>; zAM^U9iH8}XWO0!UY6Y^PPEunY-IIx53_ZW2^hYyCl;xsV0Hdi?T1M-i=|0X zb%N#5*>V@I#|MwX@Bm?BT?RkSW!sMDpm8sErrF2uJ$5DIAR^zS`QVhMTT!IKhOT9T zJHXmDD|&^$X|Y3zt5*p`p_Ek zP@+33~br^kUb_A1W_~*3#)Sj=ru7S)6c1Gshof2hG>Fiq3>Iyj(zkLhlGa zLGMaHN%Sk4g=5!f()?3|(3uUNZPe6^D-l$OFJ_0AOX7q;8~ejyXBc?>OUL(q?rmKH z)xTQ%5t%ns2fhcpx9e?Yrc8>>phy~(i1D*|q1RpmGSygcZIi`Fa|sOIl4T)HDmP6w z4qw0XgYWPWZaLxD{49HWed;S#SX4&-c_G=zEeqtqfcex0FuN%hs+_h_iRi4L2E@7m z6VzU|iqc4DoVm2AK)o(tV|a;t*DKeiEzp=1h_PK2v(2-Sop+bod|3;#UY$9kIvRw< zKqG>WGAWg^YYcMp;-d^ukb(4?$OD4sT;dEnk1OYmh%10L+aJ;BbNL{S*E+T;Lpe2nt-~-_3y%Dfg*(W;yDB7wl ziC+Yj-kX#s{0bqer=Qg2+n}5mS{!_kh9lmq<%Suz@ z)d&v^_nJKxEZI$`j{Y$3o2@i%_rF!*)&TEFkm4?Dg6KJx^_3={N51Y`-#S5-d}%C7 zk&Q*kq1e?Fj6&jeCG-&i^$L_7AQ{W?h<;{Ecq+Rs92)tXRHnF87J|Ex74iWu7dfhq zzNCYWb4JIYd)>Jyc(1JP@5?_~srW+D5tkO7?R|pogTTis>BWF0yeeUxc$4I|Vm+un z07&73Q!SGG!;G_@kj)+MH%#`|x9^|>EeFFmDk*W|+)FZN3yH-Y9q-k!M6 zo}I-3tw|viuB`_a$pX4nwIy7eF6g3*m;(9>azfHF^ST1*axAAR;%^Gl1%+@~b94>J z501jOa=5S8O>gz_V>vA@JU0q)@$<2zU$xFTrIUC!Nv#L3kT2TIA`KLKlmaT5#1>vn z#7>!|OI4@ZHwkG@p&)Zx=n1->-W^`XKM)Q%WpN2iBbX#a&ZizeYQ0)+G||Tcp;sDi z(y4ZgU(D0i%1sks0%}yWFi5;=Ng8hE*V6jXeXr?M3t~Z+y+K(a!CU%(q#1-HAqx%-b%?A!(?tdCI!q?i4DxiRR#V0_=Kjir|s z$eMTtx;!kM09lcqz6Z6FnIT4Eadc+S9-W}?aIoYY8b)?i!j=A70w9O--tMyzjoH(GXMW(NOHTbfp z{XU)EEdN5XgejKZ0Cl0wde0=`P zIu!x{uWVZJ^H!r{R z?UB?qdGMTb$p&xtK?F7kk}VIul5+V+6tu)K*EpvdKJU@Z_x z6@}_U(s)<+r|1LWMWGlt!hglm;*e@O)wkcb6{w?-=;_fv?pDp+2(*IIY2-b+@cnNU zo&1T_H2q}v+OLyy+=BWZJZbgWVCpl9?Vw0I7O>Ty5Mu=ji4>z3#K)Spn_9@{Dshte#JPn@S@!43u)jWBHc_owi9_%@7urWAU6uX=v%c#VBHBta-=L?Xm5TkH~{Hvr` zdKeNEdOAn!mXUV+&Yj1A<7c?zjAIN8Q4KYw<<8N6x3B(x+LCzi);j3oR@ws6>mn04%Jr?BI-{sN~EWPj5r>=%AmbMPBhLvrAk z=RYTPUz%letBu)dqS!`?9HSBoW3eUGNY0RySeTZU`XxmbLq&p~J{+0BwEEtHlBaY5 z)-Ti$ZK+?W-vj0Wb42;sRCIEa#0H-MuidjwN{e`A4m@d%1byNr6?B1o!f)p_D3ioB zFB>9vd#`#`4+2*MUK?e*!$C?X_(7=7w>H8MStZb^8WfjgX*}fUzAJ|Uh$QiPzv^gg z6yYol;1sOv%+PRmxuB7LKYG7?5uL|2wK^!&{y~EUzROm}x zIMzbVVW2+&++<4Eh0KW^brmA#VdUiayXZ2W|DQLCt-$zw#kfC{rQCqw!9EK#2Mt;> zt)gZ(BOV1IbJ`qkB?yx9BjhyrFu+A<+ zJ;V*iHoZURzwKRG2fkp3GTpm2B3E$BrwyiMG}S9(w}p?^3mp8nJM}UL-!MwsoJ_Xo z2De`d`uI=wWlkRJdpOXhOBU;;m}Y4U==Z%G3l*`_WO2Wbqcn+I@!=F&{}&3EIHC2W zZ!P%a>m#9T@?5wR2i$;-3O#*9RT1ARDo}$+gAoF5Sot)&m2Q@!Se2Pw7bEz{u8RlZ z+%-S<7tK0{r>7SwZUs3!d{WZ+KaZp+!Gm2;2So`*s&-|b&+n08<$>T2mOV!t*r>%@ z5`Ql0E}ILz%sC;Jn@~V@PuLu_QFGI$o34x9!8;({8(0)#)b#Vy1OsvUP?PdZc%C%R z=Y#h?D&+UXL!6JFEZ8GkDlqRFHII0)&wX<+FZTP|bN|9+W*#o4o<^AyM>qg6ZpzQw z8LRiYDkI`7S;@_N^jwok+L&W`VO&eDIk<|}i)k}~!HSva=zSJ%GQGwa~5mj57SESF*DFcQ{FM*tM= zAI9O1s~=s-uwUjpod)#aNyx!s<_@zgq!lzX_e3m+&4PS!TMXm^nTukr396TnKhE5f ziXw@H{0_By$Nuqfa0kKTTiv?h*rUlT`29Hh-gb|5X&rQ{_mOOJU0z3$#2e#s1*zn9WoRS8v8B9fz+xs0V2qBd=-szJVi ze`pefU_Y#;bu-JO8iRD8>3h=$^}7eWPy*2qi9Dy4>gsx4pS*;!{6kZeEH=>l!mFUQ z-^`R;EUi#trfAoA%pazZZfJAP;CBaH6JQDU$NbB^Y&A&-v;lgpNTJM={_Tr$x} zD0xSD2Z~FyEyNI+#ULl3PL-OFGZQQ5v0Z3XO9{A*_J{_M)6JYL%$ObdyMFniPd_Hm z(uNidtgS~qY;(aorfNzAt-gnWqTtrFdirX3HdMNg?EB4O7`TCBbUY3G^|^uLx}1OT zx7O+B&C>lr9P#YSempJMM&&Nt1#(X$#mWo*$SUxQ*3T39rj@Cu;HZR_`pvq+Bx-XkiAzPn#Jo zazXh_r$Xsh?MCsJL|`3KoJ9D?#Q*$l;(!0?onOEEN7(|3olB9V|Ah>K?e}Z*A2VmZ z=xL*7`4ZpF%5;JDacBjZ6Hxwc)GX1Xm+GuVdUXA0{p4nmI$nS8+YwfaRNhsVObpx{ zVh;}B?zeG>cTg-ubjzs39Z+*yGOLaG2$xu`qVwNAH!I_?M5gW^GG5pV8S>VKH}X3CrTuXW6-hjTjBN1K#->Ba=3=CZ7W?xr`x^Xecf zyw<;mpH9xYv}(awoDJ#b+8=Zw2qzzHebv6z4y!)s9u=rsee zpd8y+(UIP{G46T|@*E*$meKdN8(`n?dVMJXBhx9ud?gDdx|ZEVCAiiKQrF_n0p zv@?yWy}WhdI@L;_t+MM1EP_2uO5?Q0)eXR~0!_|=3qtbEJwOXWa>S5*M#iobachVc z1!7Ie@_46b#;f{-|=ig3~1Qk~(qn8T(1dQ@tqS zU|Q*=JE7SF+dCG22b(&KnhQ+#Yse`1?q<5cyGq!goI4$)>!cY>BRLMi|80^x(z(2} znRWbfbNN3>>R7VT<~zNPVzVi-3cJboh1W#%iZ99@D(hq|Giyizy>p`W zzE5>HFz>a}YrT6V*hKLt*5%T-?eI??#xOX1mMc$i8V1ju%Kx(OXI1uTXD&{P2fMcp z64DJbJ_}l;Sf{w6c{h#CryqJHvw1+Dy`Ehv#GcetAb)9+&W+kAE)U2WH#!&}M(WX( zP=3a_^&EH=^ZuQIe7MyF7L zleOJ_?%w^9pyH(ASJ5-9sPMUcFpKPX!9>MQ*+96SV(Tcf7a@#Pvz1a4$mF616!z!g zYGp6TTsA;AAcT%1214 z8GYw9E0UW2ZC(MX+yq?Cs?gmkpWMe`Pzr(#MOv6|-JhtoZ9T-l-;kgUQ>>0)%+IO%& zU6Jj<1y2sT+ATggcKZM)ALa@n>rA&#>i644--_2}N-JmVo%FfSM%kXoD^q~+56hdt zQ7MvDO+e*Ki10QlF#MYtk{>xg@J<-IHg2#)&f1#Cm>zP&^+0zF?xWxS>8>cNqgnf{JZ>|u%=gcKO*hUM{6{~}&VXb0HW%VeQw{>)_W@_^%kH4`!f4T_dXqvo!% zh`GY-oN3gY@H+uj!Zj0(nlv_9JXf5?ONzR}ln3ah^pP_mCxbhu-Q*cHedL&|h*uua z5{boy2Hs_6?v$LsE8syv$5VJWeU7aQe02YD`i$`Nkba*NbhqSG$S|?!B%mHodfYdL z2ew<&bNXNOvhW`M{6H+S1Hq{-`iKNMMNmUt+hr!*zANwcnHQQZa;1m!ym!rUw4e7l zm!0S40sU!HPi~;q*t~uGrBfuEo3ZiWb=XcDV^d7AK$V)0G(VPxlOkR{7@gI;%J4eD zC*!w?3xIR1I($8wBt98XAJ!zpu9JIEaDG%(0^T*&D6b}4gDYp?a&2dPQ}i&6gfrvJ z5n6V3#FSYXQ>anOmkndUhc|*8@8V=jUYh({WtiM*OC&cX|4q_|N=A9Gd$P&KV&qcn zI*MddiS5&1bqACht-?HM$@qcL&NrK8A*Y5>)0I#!wH(cyzF^umsZj&zDv>^HAk>Z4 z;mH^d12&_QXLSsl!~b;LJJ<@52khZRL^qb~v3cv8C>9E(KE}Ezof@b3E6j@M>tR^c zU<%V>1PpV7EzHNfMg^~axLKYgeh>3_xl(9AZin6twATi@iFZe7{$4k!4D{)s zEib$N`#hX6urt`^-6x~!nk~m;Fhn54b7%4BxyD?D6)=bX@!J>4a&G>K2V0vhHn?9; zvB00VmP+hV-i|p4;=5b;=XqI7zfUe$6-#_@lk%<%Q{Q!btnfFe`+@prkKbxCAlnDo zk5RBV!^663@cgW&{Cdh(U;FN9kM&C&^h3`dyI|}wk}hb5A-^HKB|qc8VSJTvgRG1` zHRB8vbv4MZ2qD-fDhIxfqpDleHqsBLp(YjTq!k66g(BT@_^(cdSMQZK$m?XSP*|GA z)6+$v2D%pno70&a-cV%6TLVUMt@jDPeasiHE#ntY9yxG^hl`=Cpv|#o4Y^b2IN3nz zU%vj;{~5_r$b+MC4wgchA#EVt+y#Q1w-a{#P>ae`nDX8PYw7|uHn80SfUO9_K1u_P zmkxNHC4D~F>xt`e1D!0MH?sgz5+jF@qwyRev>ZKl#K%+qc}|x--HQXB6iY4BC$Nab z=;uY*1rWL1q0nNkri7QxGeOHvQ*?857kwcbnKj(`HV6ON!+5iUZ+F`^r!TGNe&;^g zAol%()p2187%}-`EJ?LVEe%lYeTv+r5-p`@Sxlp%?Uz_Kgl*EONr)wIdKwc<_>J52 zhKXM31GPgaU}&P7qoELj?tw5xJChc@nN$f^_+5gpVh@wbVrb)$r;J@c+CGxX9wvDc zu|oI?zlUiB#X=)H;Dw!I28eRZY;T@FVE!!SF7m$FFI!7lQZHQ=zgC| z!1r&Wf&9NnJ@3sgM|R0`UFUxF?8djO2)y(4Z~UC>^WenzWgBPs1jQbw$Pp?LX8gfL40pZc#c1&o*>of|A}}h-#~ALMI+Q>K=wgc6o6koGsDE?gi|vt z%&gj_!sOU&5ehQ`*OgRTE2^h=Ps)$Xfi>r~XXFkRCCmkbYYWo7U zy_k*1F2Eef(;#VcvmBGtNZecvYQgVT#O)2%UJ!1IYiHJa--{XW!V2L6^+s71lO1*w z=EhahMg=CZmk95?vUx`FWS#21UxT8HULrgb4*KQ7Wz=U=mxb(w^7StIy+(ymmBk#H zpj|GRqd796Sh_v%zMr;Ic8#%Aw(C?IWbM2(o+ z(-7ijRUBRIG}atLXsD=E&774DBjFDB;NXt~?(i#9I!J-IAnwu6CQY^E4!NQ^1%c%o z%%{P{p?J1livQKAj*|sG&SS#qK^o%1Vc=V?|BnTi-fC zmU!?Qy2u70ITX8^f;SG~B`gF7C6Wq>McoDng4$J8afbvz29NykIu%mjHOsM%+~s3F z^yqL$!qCrg^-;%27|K6gbM+tN?fuiI7sDQGQyh4qC=1XQkOqR8CXu<)PGatjyf$qM zZ(h(sz{@B6W9s{|76xQF!N`=%`eJk#pSh!Ebo$C2qvqIye|}|})qecqVqQeFVkkQk;&400&(w zPmA9B8g`gvkP~#SAYYxDfQf?YXzNq(&eFI6uO{Fg(}7;1PKBXUq-=)Hd+&^xwE+X+ z!_8y4Wf*J+XIO*c_)9+2)Znpft9NIt*kVtQ_(IKD4oDCi6?c`zQcN`!&}Rq+@U{l` zKs^Lf6z36F*at4CaT>&29_oT?9fNr7cmF!Up0Vsp(u?GPcnxWCph8BF#)EobTt4N- zHY!k)e0v~>ith{Wg;@t%b%8~vQGv%s1w>9GXOKyA0WhBX2=9015c_Q63uQMPWEkIn zuRwh%P#>KcViKM8FQCiAmM}YgjLHSkC@P^n4H5rrNF{31qSJVnLUrm^=sZGM0&`jE zigAb8ENCc76MXS%nxFv6w6vXZO{&hg)6ph*TL3;0#e<7v+Evn2wi@z{XrQzoyl08{ zXNTp}9WZH@#_duYRgf9Li2p57gZVnk$s3%n>F@OCJ_h3;)KM4MsW`*yrdRRy1}-O!svZ1xx|Xb*cw5#q z{hZYF@_aT`vp29ZVf|<+-oYQu@wE>7HF3Ss@wajaf7Wv|X`*1mukB+nT>KUfHX#n8 zGnT@zEM^6j_U6(z8fIy?0~-}eu7RipmT*GJrumF<7PdtgV0CO{-EKU%1Kl#+a7FDf zuK(AqzoERX>*xAcYd<3M#*+0m>*q|0&7epcm6#T7qHim3>8m{<*#Ofp%FI9`3KXpk zcR4$ZU)Xw!(}rXCA2F;uIT;Sm{d&DKO%rbq3y-Z?ae&um`MwFWXU}f*s}`lQ+XJug z=RjrV4N(^j2QB|Q>bujo_Q|e0?SRvG9m>%h780Gii5p(;MxWdxuv&}AHQzl*7I2H$ zdM@Ch`FU`}b|uBGph!BEm?GXG#Ymx1RrH9dP)~2=VSLZ>&z+!odI&kd=;@o?cx2xn zqVK=x`Vkw&IWy56C{T9+@nr=K|Del#p3hYB%9@yuar!_fFXm=TUNcPn@klsT zJ=m5wc>~ge$7IFgVUL%S zU)-lJb9$hkzSXTqIGLC+?enjlQ(4^=&Ea?UkUS4AP^qzTK*}f<2p2X|iJAOdK{9D$ z9xC^R-=G(07A4&0ZHVt<4@~UkZ2^Y6>ruBvsR;+j^58qL`30(!1| z<-`XuM$IDTmi$K465xHx;`dHlIk8Rh?l3LDGe*~G%(%k>JnQ%367ir;@@sxfte7eH z$~{UJdN5`RY%sHmVlyd{K_#||wuCo9RlOc5nj1yQVw0>kRJ%vr4ZN8qrk*rGCffDm z?|29}A;a-wTz-a=HF;^tH_A-*g|i;JCI-5MK?Nuo6q`noB~)Tf#I|6pgP5JKM%wlZ z(Kn&>&<(|m&mSu1F!B87xcUlCm>Bby{|Wn*uhm-o;sSY>WN=#$dv55MY_>sA9>s#j zP7akgM}5HOsH#?uS%hvuqv}Xtig>H+p;v!!Ejd7n6IS}vl5Rms{6nwZfwiP6x+Gp7 zW*QCt4X(Ka%Piz-_zdy9^gkAuUUY`R`_d($$z+e{lFXM8TTX63PCAWuJ?t>qFKzNW zs7x00PPK45RnwO9CeZ`P_1j-#)Es{Uxf9m;G?4^^nx|%tqn zyXbtnpO2-@+G;wDhw-$WnLWTEgDs07AyFm7>vXDZBnQf%H_vQli)3(0t-TPfW3RoO z{cRq-oi_Txj}2ld^`_#hjJn9%c)TntBMm z-3qLerLuaunx0G6gkF-l-jvu3!tg<1^B9-!I0nT~wJ`5eJa=!9ZDJ z!#VVTV*4p_2Q5w(QwS1QM^xDVRSOhg2T4xs)?jSSs8#RxTj*ygp-yEB;XH`%&CKM3 zf_N5FO`DVp;6msmFbmF$w(-utj-6wxCMOGO)i^g{;k9;kcp1p=)j)e$57QH`r|;3w zS_t*nYhxFH5J?jc$u+R~>~>6pa<%`Jn1@PjZftn~UfwFg!ng+I$pFYg@z$sdXzj=J zHga{!p3rvYn0JwEz$=;b^ScvJ)lv(T3KCS!(T6S+tP8jHvn8AiFqg&L6fJ}T-UbEm z$M`La{&Fx5jY;{oD&F3s_Sh(mgKm5ipa=U@U;ulR1L6IC(2H2cOBSDl+8ZvcI<7DU z&RZo{e46vm;AEA?{P?`|HG4)nk41$8syq}00o5Zcxwllg<%03)P}fYXh=*FXpd3b9 z!vA!77SqWyL@pFIc&7{8JYiV|%Xxr2aYvjV=HXw-2_P?(*Zu3^pIN7^X>&|d$OUeJ zJP*!B-?Q76xC}@bdoTI zGWRZlx-FB%E^rzDlI)bms9CHr1)e5zsBUr) z)Pfg|EdceIe4muD+43?_^8%MhZjv6CV%J$l!f~lyS}k>}i5kh^+l`k?cR86R>cWx5 z|LJFi%J=uKJ404*GfkdLMMKJLfVF{Qb1AY87iuONm*oAvI`z76T$H1tuwL4oa37*N zW}3n4F<7Ei6^)FW8~Bwo%)Pt!eU1gW6;ObJVPndr2zvTq|M6>4%tR|n=KVfsDbahd zL8-DC_stXwE>98Uts!%?GU#gHBEMdyT5>RH2$C}SJ1)~F_!CWuW}C>hK+$I^U2c#=z=5M?8*WDzm2so=Ji;g*g<)2 zqXHWc9x8VNN6V&J?L<2OGKY@AH}2tML>Z?hj10LPe4G&W#N0f!C3N}QdhTE8?ClX zeD7V2PzTTB1l%$InDmDq+t+P7=;rG4X^h6;#DEu;$m?l+6vW+q@B5|tyTQsj_~RZQ zcJ}Q*zG0L$IK6$(4Oyu*eZR2xOur-!hyy(;@fOq-1&>xJzF*Rjnl7M7rD*fyjn=MzF;K|Cu|nJbrcKrFx43FKsl2J zaswnsy6JuVE7Kn^owT+vwq5mEK#%g=Yv;#nkEps84^>^@R5mGkVnCL}r!WMgA1H>> zKvxhwjn6R0K|uaj0;#`7MLm5GT7~*S7^Q>8WTgROy%@UHOLHN3iy}sS)h0&qJk5-U!4~~^I*g$3<#ljk950$vWJDIn_J6HA~ zG)Y`7*+=#YKBhPGkHEU6TGAHK1O(`L>N>&3gl(bqbROM78d=oEv%TxlHF)<1{+bB9 zcUj1J-u%e(5#3}juS%FqlEmjD_CoaK2Du*A6nDcFWP%gONpHaSw*lT=xgQ1 zZj|i~N6m)Dz~b>b)pFkk0dn)ewRsb3<#=@wlp$Cy&0?0#0BsOh`$E?&UQ(-^$95(bQV4O} z)|hmnh5d5aPlrlhpkzfVWmFxyDKwLUZHnuSN0!!_=~?5nC`W_;H_2Slx0omt#WDnA>bqIKWBiWs@Nrn)1#>Hn6l6M8u^2wX?6yaC z#B}nSR3|<1O!<^G*X4n@#0x&1cjo~N!VtarUb4@?r zKAWuJmX7q?XcgIEgU}+11#7l}N;Il&$qU3QC)SWnPylGu9DePBFhhW;$Q|KZ0uHmu zzU?ZLG9}`!vWqDt?ciTmhc9NUgthdeeeC;x41-6n-Gx6#ge({S9`=t4TEaVH9~FL~ zcrhJ>u$$#s!n_G3)b>E+TgVJ4mTrmEfr|RosqP$Q`?qV^cXI59zS)gO9b1*R+jcFO zWVIl@Zz~p%_1r9o=LW7wt&IgKr&#EsETIxXsC4F`sSQA$Tdv$nYYY7LNYe#KM~9S_ zr~L|0_0xbgz)i@Edo`{u@^cbS zPE0zk#ny+@{DZu#peecg>kmj0H_&)+QW6?#2em7tQ*0_l7E_5E_^=wfCb$xbqD#oJ zcY655Eh8EWrywz+PZ-|KAH;oYxxW<>&99vKJ;~$-2@kd-WfJ*dI&tp?6yY*^_ds}i;dJJiKetduJFf9$;rTvO?_HtrU8NL~!N5y;*XP$YsN zjtGPb+Njg}^z^hnr#(IWdi&ZAZO?qY%{eEvo#_Qp6crH^P(cGI7ePb?L{YhV0q>xw zDBeIK&NwK7B7+M5wMcMCB$^EgCwl&!Uu9=s-ay_bYpr*^Ydw!!Ht4x(QM9qt^N1n| z=zs7(JNqW9U^1fJ7~S++`&5S1>kI$S2$PT2#=cGNOd`uo8b5j|_7O$8fZ|e;GA+gH zJwt*NyFf}o*|{Ee^0oP}d>wGYMQSpGONp9T4=0}O`aS}WeAGH3&pK- zI_PN9%~UB4`f6_}NHM5i zjpCPSUKOA{uWc1}LH4B#RyI1HHNm|kin_%_QI8b&<2 zo(W);(j|!HWziYa_Q7wY?I9_yUDD*>ba-FzF*AWp3aV+ls^Wg&mZ|+tn`S2Pa;N0_bh;gx`CgMI&a=#;i(c!KEjpzt z2lxIn>&uXSm1v+<5ynBU`9WG&JB|G z#>Bj3CLU!6#croaHWk?eZT?;IW8Qmsh90172!$GhwD`h6jENMP+CC;#h}_Ov!@3$< zgxZ*=PznyTJc9-}zOXc_EF%`Wg$RF@}76 zcUVS1GP4~TpmnNSblSXJzcyCe1bLBT&f{7c!>RwnE^}&OP)_!3g4>i|oSJAfBg<9< z9wh@#oCa`Q0}1^@!?P)CDK?HGE2&8R;I7f2@*gBs)-J z7}Q}gVWDpL!Q@Y+(IgidRGTzKycFM!vW>En;>Nk#-3rG`&-wESvkn$7sT}!yD4QN} zf(7-jKl|qiGNbblKM?<(B)l;(C~z(hOT}kUEO?VtDiZ1HmiZ0chx9Rz^2fO;Ng;(<+}GxA^?@96*4Tn~av;>&@H1VG7o z7({kcEGSXsQ<1ml9^h}D-T|uL72@H?jT&r>FY?>#dwcHY>DUCX1NMY;+MvbuNC8E= zii=aK2tHE+B?Vc&&wPi*vCCp*RFE;jk9`odPqKK2`C5ewWE?jOY)M0bhk?&i)v-z_3}Z;ZRCHL)dS6kAG> zVqB|al63;CbsMBJ+~Hpsl0Q>>*z=q#?(wbkdKfm~WPl6uuM|o2QwIxgAGlV|-5jh% zwy(XUUa^-f10usdH_VR=K}x5>=URdO5NN+_DBET|cFZh^rI$8>)X&yeea&0|(P1N4 z3zCLebSGO7oXkhwgJv0Iw6G#^74Q8T|G%wel9@E0F1p(LnJ3WH!LqhYRWlX&ckJRi zOebY@b8b3@JzjANW+!G!HGQ5&>+<8X6E2Wc2QH!AZ(=XWq~>_9SlH+TzgNBgiPIGSdowY=}@Y=Y5E=cLqJOzlob_=kIL@b zt$%fNU>Rl|?12;*lM`4bOskvx_ADb<7PjQYk$i5h$brM!IumFepjaq|EkV)GYFQ>f zIe6(jWOTyzpbe5%VGn&nx{rTOTraN>BePSvq+hW?(nIP0U%GHGruTC|5wee3#MIOt;zh3*bdgeIK>ey}%#QOGv8J-k~=ELp+MpW0btpYInZ z#wG}Oao91R;bv@B{l}#Lx5$W>r~J$xlZw}@+4ZUkhR#szNs82AbZhWIho@}y*$Rxf zcc4Epo1ZWPYG7|EcZgE`GU=#!ZA=#3C~XjGQ~eGJ@t{?h>Q}KKYMvG!!Tzwp&^Vng zp9)xlKC*b;V1o#DWF#}Vtd8g5om#_tA*0I?=rMWbd4v>d3N#Qr#q+SS%rMUTdD3Td zIPN3a1Z2bP%S#S90nakX&TROXUz(TDI&3g(p(wUlhB8($f>e(iK~LXZoUjPVyX|2^ zUiFzDFxBTWpzA-v+@&nHK7zd7K zLH}V`&OVu9*HL6G6^R9NAQpo4*?Cgr$ry0TQddKZ`5LURA8*WB30V04j~iA8--sJv zljogZWv+`o123 zf|**(%yEn$M=_1oW5m==jN+|0J;DtmkH7WQGs0*)Tp!gXlYI_sJ1&^mj#`R6LXm1J zay3tT8f2%^{nxmG-p46bsz)Z|)Yq_|E$ChVK?JH1H2`_-4c|MkNKSLn?({AW>k?Li z><8YP;J$8VC8-cx_HJg{6rU-&`D=af_TYT!#6q%@n%{vxf$sg#=GMVfPL62wCcnXKM{{0I006)9ub%!A#7i36rQw9rAy2Flp44%Lk7h8+# zRMj#}pJG318#C%i!>XwqEl%tW6;3#r@Zq%&&S;D{(FadSB?lcCCznlda*|?U3-Blv z`2d=YqeK_v$YFzQH+iarh`zu~*S(~Lq+W1DkTNTmt`|3Hx&UDn{HM|iVZz`T zwKU%O%yWP)QS?fWdpr(rbvqU^dc=(ZO_n3t&#zd8=#g&XgeK~p4;Q~n8`0GBFFq$p zG`HZ^fh*!b=Ve%MyN+VlQY4Ow#1LmNv}xoe1DUB|W&J01m)LXymH2dt>kj^J3jBkMN+KPjNC76F2>S!0&aha@$18 zTz9(f<6$PGRd`>%kC!RADQy+j3-|H2J71y;!jMGmO28iaen|PWwLW|7uJmM>V(Sp{ zxo@y{Kr#e8P6(lneR)x(IpvlGt|}BCt#|&+^N65dF-YEkVh7E#_K;)FgOwg=qid#@ zyWnYz6q{vF1xLfQNPO2x@8+$ISQ~Lvaa5tbDJ`H;U$9w*o#;?(h6z;LW=KadHzpG` z{Ke4cYKxzo=EiZo^q+nyd+!5d{E_jvb`rVw#>5|SCj3AB6bmAmy&ym)EC%vWWLaus zaCL+Ql^Ma+yf!9Jol9raF#&+%Lw-9%*Ll5j;sdm|=G4+rf%#Ho*ugbYVpu($2TdHA z>Mo@gt1nxDpQ!^1Fe@N2gVmRYej#M`hlLjomE=<-L6OZ$!CBAZux>?+R~L<0oTX$d zsE%)irqzld{GGII43ZD&RJZv_(+o<$15P^iZT=Ebx8l(p{EaGtt_EzF(IURh98j)YN#;I`0IsBB=P3-er2p8aAGnOFJwJQ2a;V$u^gd;nEV~ zz~z<}#AYf5IsEds9tgVxpEB!U@pVbOIjoN#4WwUKv0FyRIyY&m=onElbJJgM|Ll#? z7;?6=wlJbrjB)2{KU4q8XrO-Iyxo)Ba^U6m5)+ftO|hL6>7XL>CCSVxMH_q#+Z74U z_(dHwI=lj2mbxP7ruZ^Ds0wn0?FNg4tsH0#Y66;6dw_YpSN4f3Fk$F`{-qDT&IeU^ z+d;VPGEbYO&KE6{FO%!sw51UhLHkw9ytcS!03UHzc)R)zjDM2s1eE~P&cl52s<+Bn zY(By(K;RVED=QAWMy`9;_}?c`y4eleplIFtT(>9-73bXXHZA_s#>Ba&d6kFM1XRf? zgq?$8y|L8i{|#TnbH>qd2W{))Vffl%f%`=@Sc+ey4y z=pW-%uh7$n`E5Mp_a8pH!#+?u9$JYZJN^bE53TF`__TQ?ox`rnE$Hr|f=C9hiY}(> zg&ks0=w_QVxY|l)aCMd`X<^O~oiYQ64^dmBNwbT;MVZWOWXq@79|u;>ySbGZd9rMH zmXqzEjw@&UTd2`v{o?%Gon(&#d#tl2NxEu^h1TbTROC+gHbohAnqNj0dSrsGdkeEd zuyyK6sC|eE%;4$N7iH@M&bh5||5F~dgZfiOSQC_CAM`Ju-=xXl4TvuWB?yLJJLgst zP|a%<7O_w0?K~{XP7pN9qC~q$Ic!cCROdH4pL0LYFNzo|%0`JHb4T!V47=gj<@uD9* ze9xS&LFc=?Z)iZ;Evk3vm8}TH)5srFFHUl=_);dnLj?;p z!!P(M?`AE&UUW}mJo5?&F82BAIo60T*C#z!$c9N|ugS#ZQY={ZOezxDNsT&%m@Z4^ zm#ehLop*?ihT&npq?ySVbC8{_$==tj|Mk3y5j;wP#ejX=8GN z_W{S8jm(18^Jg6_z;0MQqHS)p42uWagEJUoEW05w;4(?!mVp>c82)c){;=suF3tu$Y%VuSs(>phbL3o;zik{Hb)A zV25;*IziCFT$vx^dBx7;%vM9_Mek!9tXLSO)^CMiEb6)r9IRSU*Tqf~bTc~DP5(z> z+jtv7vF<5e05wl%+z<1S(lT1q=c>h!)lh-3my~$Jq9ZAnH;mZetEZVH}pebn}y$1BK-8Cs+9a{b_zfZ?fp2w!c4kTK)IPBz7|`xz;G-T37@3Fc)i4!icTP?%Dt zDiovYPx(|Rrqro1?{953-Iia(!Lh7Gh0Wji6$5fY#l)bLxMwqsHs!;Q+76NJ4s277 znb?&56bsu9MO0))aK1>Vy78}>&=|juUn0IHJ`kQou97w;{nz(_oB=FEjG)1N4V&bH z^|RH`bC5Z0Rd|i0QC<^J#8%L^*|LZR;YoI_n||t9qL&>ZAIp!BbV)fZnsutvs#{9i z#=O=-=S90Q!kJgVfy-r1(3$YFAE*7hxu%fA8WamEK?dsq4&_2bsRCIP{%1sjtlCTF zGLyEtmlGl;e*Yhy@nd1FkZ|A}2#KVOyhO zD=|hTMNS!~$qEOKO0rF&k_{A_M3F=)GRdb+Wo&xISQW|F+ts-zRF1_8I#G7A$w98PaQFy|5}^K)gawtgM-SYfd%q6X6}C zlp2>V)KTHVhVyP|D9d;l!L(2dmmK?%dHtiqdK?QCja~9oXt=<|Ya3HO74^uVKDUF$ zM4A>~9rEpzb%vrYgSLg_&P-;2s~?Cva_L+2V>fL&+XibOB#lT8ylrEf@%bE$3N6pg z#o)o`gW2$i6Iv!#R|S3lb(d>c`sEwawIEQIO{aLF;7^BmV7z4w4fAIWBy2d-Ru9Dq z5|rzat>62DF*f;Q!!Kr%PhYdxWR*!1{8Nhkj3PZ$WHuBWHOR3TdW$kafSY;kSk}51 z8YOCJdk2rPRw}NhM_Gf$VS&E`Pe2!A#3-_v^4D zu<5M~Hrh`=HBa5b6!|4U4sO3|lj8Qw;xG(g-ai^pK%*oK+O@|DE$m2wp^9^l=XHpY zM?t-C5b;B|wr@Wms|8?j5uM!q1+n9`CSP#43Q&!VArKJ(~l&hf>GLBzp zH~BkD_P`pmWIumo>9^u2i$}Trd9U+gqglH0y_it)a1vQ(@=aehkp=#Xek!sW+RMx6 zcp#r>2`UcD1613O1 zmR>RG5#1|_opeDHAGBP2m1j8DtgNCt37&7%Y=;f-gi| z(A)-+^@0VrfRVC?E)I(lRjI4gC+Q9*F}z8$DXdv}j9{Bay|7cM<6oKoD7-{^d+vVc zDs?<>xh9EOHt(!+56KW*l;JhKaMfE00cjo?F6H6Xvs#5WAzO2kj`7T$+OOE`wN9Mw zb|s)mb5q*rp5}pfyoTTSk)l|Kl_ZIr_wRn@Y~Nf8{T;Z-uE1nXucO#Bifp7JpMF6< zYt<~nzm1v-@ndou01_0wudzOUhYtt4mgn@~aL`PwYoXW4S#6|1e<^O*bP4qO4ui%`g!Fp zIbc-r_=x(NSAw9Ae(0$eJ_FW{>$5WGPv-Q@?wa1DX(drpF7dkPJ0a1c>(Y-sw*>C? z-%j_@8A?3Tsh04nd8k=m4CnCaI-ny14e5~8%y#-tXr{V~?v*uZHhUZy93RU0c7Vgf(OgmfCN{umA7F=oqsyL&oy?&My~;+WQrfXh}!`{mNx)D=S1Lc}q2ssqwn*c60hb5}{e$RKny;A9A< zKc|@L;U)RBK!L7KRi$1!PpjkY^~dT<>?67-$MCpa@q}peVNs6_A}ibz1X`r&%y3Tf z*+h;IV`4GMr(Sp<9GGK3pZo~HjfbIewG~07E*J+l$lnZ~=(vRndT|i*mJxlvzbk)_ zY~)sW?!e9er6xXcJH^7r>sBffzY)mbm@jE#dPs#BX-5sa?U~bBh5P)wpp~VAPLiMu zc7w247Bgd6;6b1?+C+*$VW(MEIa#NC(eAwPg&+Niv6!&Yzw4`?{L9+Z<>s60T*4>E z{SpuhTk-~2pm%B&#sp+T8UYds8G=|+0Hllgl8o-aMW`3=TA zuy)wIwS`dm&^tGv2Y($&1#YPdaUcHz+sr&Bwe%9veb4>U2Z4CDOA0;fL`-feTM~zg|_^35N7LhYg8GG0t8WcJXES4EE}c^WEU32 z1Pu6QsWFbpp}RtN1Vi^5{QO7TcgkHpYWohT$qGLaCQpu~FxJ&WLD=0;&<=D5>JoW( z*xk_1kj{`D{spp3vQALqjO3hnAD$@;(F^hJF(HmSj;s%d$6{JI;b>xv_PDRPxWS9n zJPsU_SP(pDmRoh7Q>9`&_Y2WspP)Pu^x~1nb-V=VMqA-c-|!qBSpgrP9pIUpn- zq?pzr*P6pv_8VqD2`JNLn?@yXP3B zm6Mt4;p_c$s=cI280!LQEmXjS^kn!lc?TWqG8(9k7JardMwYO-WqnhkIfV$9CAR}d zLKetH(8nR&A@oTgpZI_i)*&7UdAX0V^zn`Dyp^CB+nbEy>D<=JiLOQ?Q}&g2&ym#* z9Ps3sn2j`w-AKXb3SyH$FlMGzyJ1`(AdOY{0iwh@NM_NiRdOu1=3s*|_1yTHFjpnnUwy@1HmS;4KO;1{p;id!k7$x0Q?o z7fBp7aU=UE78t7bU`T=_hX{at(e4Nxlr^-ILh%amQeJTwBx6H%1y*=HgkpyB@Xe}N z-wiVg=^USIx`nw*?oEz|yO#uYLPh85;A>FO5Y4uHal?!Td7DcMbA%Tqf;;A|46-e# zxFTl=K_mRMDVTCk)l=qlsxNvQ2gZj5($rRAC$knBD0Ar^a)}oY+ks6Qy$qTy`B$g! z1J>O-#S#zO&hRUyg$n>)F_cX&TXs1C^8a!-uLFHUT#kiuR;V+lk(S<2qh1*hJFi8Y zF&PUCh7!FQU@%&Rn5=DybBS?Ku`=6dlw?G-X~v@4eRg9!Clc$T%R!)BJlUdf+0u| z4<@V62J|ua{2F+Nr$FCV05as9@k;S(49S!@Z0Szm{QtW5N270xT^S!iI=J~Z2ac3i zn-q>cqu4%*JfHt_*A%t^2+>?ARI(<7U=Ob z)`+D9@AY3l`I<|S-$sDcV>k`%<97weV8#B7e}^D-G@Qs*=lP!ZZs6DWqd3#WplH&^ zuW?%@2hEF^fZ~t_AvQCtVOy9}ss^YH#CO7-O?na7CEr!gpbTDGV(FxO?dX7^j0 zz+2YPho*Gff$Y^fgp3d@o1U^2FrJTN86hL=OH%(-ankQ~M-58*VE(ro8tQsTu~Lh9 zom}XZX<`SQ3_ZfjHN9czeel)+rzlZ}c!&dbz^TaZ6l_Q*g2e2A6A~$50S!DoImzdN ztIgwa8==PLX&_FhnXvPnpG3WHv^B-vXB4E}fo)BUNl5dUVjoiE0TsDjw0^QSn%sxJ z%fZ7TP@o`stI)tgfl)0M!9%0y9^O6qUl&JHkNHqszK6GKA?}11OSKL1CP2>sRIj(f z|DkeTT~8kXCLMzc6?TJG1mU?p*Hv?$mr2w^p@dF_^L;Mp4jYt7eB9E2q!Whb_=~2~m&Fx9xiV~DJtqfw76aQ;GE?Ha3)GY$ zn=QEkTd0)*dRQrF<2{q6IczY)4SB+!gsCPQk$2}e*ZxLUIxzBbO}yMDiiMKJ^`OWt z!V>WP(kHIT{KT*VSqkW8Cr>^XkqZUmTYc8}fNI?2@y$SUvP#cAjr|WeVPj(Z;$uHD z-??(wB2f!_QMZ&2f!iWIFk`ZIqwMY+G)Dg;!Wh=I*x4TeFT4XMNK6Ph^yHR#y{E%M z!a{8)7QX1|24GgIhNMX{)Z(Q}GJKD=81U!&H%{k`&*ykW20obj?qno+7A%1WRvNjNq#HnY7Fm&)6R~ zR-wdn5rnTC3nf%)=akKXLxILW-z_3}Z%mG_)+8(`qu5f46jPB=p>gwC<{(jMf?#I| z{>)>vxFTYT$XJ& z{>v6U8y5Dv>Riv4=6Vf^U2Cr3k zC^SiZGh_{WGo+FtWzR& z^JDpPvUL)vGWo6+QS2^?6ad#RAKSMNsv6U1#L`5RO(jyQ!j+b;ngF&zBDzeh~ zvHL*iYH}3foqFLq_X{9#oX0yx){&1r@oFElO^R(WHNpzvHlUQK7uJyNG~Ra(@+1S9p_$5uOofX?6A_gF{qZ_r`YC($2PD-Hv29Nb6B>BoUYE>!vB^L zM2il0|B#$;;H=6W6IfiK*h>^?q#`kBX_l>a0|S`HAQ40dSi2m^$f3pt(9m9Cmv~eF z`O|ryR-ct}%oJm4*pMbRHqK%F&Mjr$?EKk-PZ;jZlR`A(sx@qq<0EH)e*)}msytPW zLDKEH9njy6LD6|1!#Hs9=*%1%*YD7O;B^a><>CD|dgV^R)8#IY`LQCa@;ml{iwSl{ zcrna&TR7xoIw$rz^;DXt?6^cj4!nrYUTV=KM=^s8>xMHC_vA}Wm zO>KhR=AJ7F;7@+w#1An#DY5f{ZDjW(a>m3LR8ed>MGgSZ5mY82Een)%pA+|zYNi^Z z-T|k4@9WBPmr8M=82MX}eCnAemO5Erh|RQ>1k3BLkvqIRCY!!b%7ZrfAA?E?dwQsm z=D=!PjP$S_&;8`$&8_CLE^i1fEQr9|WG@J@tg{mpjSZInKP(Jb zyQ2Khle3JzWnoKR9LeYATO4?qQ)gl>4p1ya!X;GXhmh1qMW_|QIixr&Yi27fcd-vj zi>vQkx?Pdrym;~Ao62Gy-jP7syw?Y{&dG;W&=VRie56HINkdIB7O=$5!xAo}716da zQS-_oN~ZTgoYe{(N2qj{NVIv+*R@aW09<1tyB-+b$t!d$q6i%oIhES?3K zk5-|92YVkBkDvAYB(w@vL%A-?1c%i+bw~M|KvWvP1k&KH%3y4ymC~GJP_T0UJrH_9$D0u;Yab zcU-jN+J9``;?*;1HE_&*$g77Q(;b$91N*ee=Dj0th^Mj8Rf7A>P46~>E&~b$J%rYR zOa7Tt8t2*@8zW4hb#NKsK|7yfcTbR*cLBU1ps>(svhGXB)><0g>QbyJqc)MPnoLQf zCi8o%-pN>)|L+gKo%q#1<^TJ}cenlWlW$+pr2C@-!H|0RFGRM^Go zNmGivs_1ICy4L3^uf+E(d0>MqpxJC&hZ=JW^Wtl>VaNP(M%KQ%X-0pu>;tbjlJ3B} zhL9K@=391AEYSGoQIRNpixh>?ewN<-Ade9%+vHx0uo5C9;Sg)Px_qPgq+}sOSXpA(C_Bh80 z8WWXA8V05qjmfum{PI7@=1H)VGyFSSK(WBfk^==G>eC>9hTMN$;ZLQ}q9z_xCFV;G z3AgyicwU!kqe80!@}_t5Gx^Q1=Q`k&Hm^?AJ10H>nc(u_b=zQi0yQG>s-5PE>#W${ zTpQ+?)2ujdPHM^*+n$akVdA)d8L^NsN%Sc3zC%O58M0*}d2TneesYbVn_uGnaJo)) zA>t$pjcT`)*bk50?%0eLHCc;|W!OKke)1j=-R+{W?(csc=zY)s^ovrXaZ&u^OO0gh zBxpAtZd}qSb~8mbfz%P~2SH*-u55%(|tlnTw^v z`QKRfqx3IbU(~O-At19 z#fl@qBs1{NvoE(V$flkwT=`Curif=4b+2^&%bxc8RVLHoCXD+vsaDitoXT zuGQp>ZN3-sosuAx5kQ)gKRZM+9e6PWg|x#KLwhK;kRm&fMZQW_rLLu~2CNIKA<%v3 zk^uz*8N3@zA3t8uDm(!>l%;|D;NLv=A|BFuA0wX(p2I6c=RoNwpk7$PyE~^Y9LKQH z#G6c$O;BJmqHSJqvW65B`lDaGiE7KO!l%DzU(ik?7c;JnP^b6?_9hqvm{-aZfM5Z$ zwgXPPMY};b`KaLI*{Q7!dy+955pocM+Rqc~TT=-oRTTsw5jdKAVGgX;EL^@O~ZgICt$< zRzEdMs)w51c*$O0;EP~-3D8@)HwQfhv_<{KMB*B4A7FeiZCM$ztw@@S%Nf$BIUo-u(7z_TF&b%zf z_!ohf%pCKNAFRtxg1T&Nd&m}lT>c>?xIHYAx&0WGg)$$+4zD>`s0qvCU4OzFA>#U^ z=L*@tEi7{2W!7F3K;=?w7DY0tNYtU!BQ04jeMsF!R=~D&O#tK<3@@+DKjT#sfNO$S z@(|cPb*d(|5UfrM)8|^_pD;6FCSJ2AzD9+sp?Jp*cp4h=(ShrK{of_-hf=uNfRr9` z&kuD{`&_YE78^l|c-MI`Ggc9M;$Y}=|NJKhWPJY6I2<*8xo6hv-fy3!#(9O!Wpp9W z<~*8%#B@l;gDpdP=(E&A6Du}AZ}Hb+?17Bp^>8cRW^WX5nD-$&VEAe0x##9-$bU{8 zFy9*BvIXhDWepZK61KVJ&fEtnXkh*&m)QrdmD5s%+HG!oc(;^S=4&(6AJ0DSds%Gr zeq775&JjqK2eR>%@16bE`{sq%4!dh)p|rYD6HQ&vY!5!@+a=AN**WDT2nVnALAi;Y z{EiTNn+Wp>v3jP>AGGspP9}p|Azbhm^Li&PCc}Yqb`~n3$|B+=NcR%YD+R8wBIYKv zEZ+*L3|J}OBHtC*=h~#{nVlQaL6^8abk!?U2~xrI%GS!q0bf=_%ESlR8BZ|$Ka@MQ zi;OFnr~J$xlL`l3=3_>natz_Zz8Ee ztnP4jgrPyMM>;^_u zH|V+WZCkK-^*h(4?TQ?JcUY5VyIVZ(A}RPu{Q}e|*~PD-(|xwNrTrVe$`);NYyGFI zaHUam3+zYXf?G`If=3GMcgClB#80Is7RB@aJT#n5{gBmAc!klx%eI6AH*@mKw|?au zYy^(tFWN0+mjeUmvJ8c0aTO#N=$ZP;5ZU|cN-7oyv&4{31TqEz1 zL2i#ZDJy-AZb ze;>b9cr~CgB!T_tPaj^8r_3x3-^R-kWrNs4-@I+U(ANd)lzQv^trA0UcL7`KQOP!br?P>r5o~>{fz4DO z0ryqUzdL8S?FlWpOAo(*{lOEq{{oagJn-1O@X}#Zw-zcwTS%R7qb%F6I1EJ9rWOU} z_(l6PY0kKoMbtqV2y6j)#dzjVKLg3PTzZ`#P1WURFAHZnKcldBrUTgN6*qg=E$O*$ zUibfom=z1{3$4PMsk?o;0&yD=H6W3gF!xJwEWQSwFpxdcsR|Y6+>@r&(&Y;pU@d)J z`l;~QoIZXn{kO$`{d2(rO!Ib4M>(@vI>$bn2crPj3pT|b_^K+~(Q97!|DyDn<9bm` z)%~k9S#)#I2Rfg6*?-1NGQ3(oDVa(0L7IdvK`ak1B?^kY^T}@a3m!VvadOPNMvyjd z8H-n%G+lJtS8qtO=>)qmvWyuM_5cKoY1!bsb4kDFE#uPI_jl#*k&WDPwT>IQlTwp~ z^mdBPrpQ(*au6x1B%fRIljOq+IvW^rchF0mb*erlRhZ(HDJgb8=3g&fA9Mg3*iHlg z%4Ux&dT;n4Au>s$U}rA9hP9{v8WV`@1RxxC%ljuzPBcPg*^0oUWWZ^J4n_wC1k^(g z%dD)W*f@%;1S&+RP{N*{R;Wkk#fXT6WP;qTPL)N)EnJ+iNT)(wU_G=gp>l&kBCj$aO|(38 z!08dcPBGwgmh^-nfzyCfx?j3qpX+Y_gT4mQhXJQu{0v@9z)}C9*Ud76p0GjeY`_U6 zI}N2zD7}ZN;61!%7Athyfp7VqA3l$YWTOG06##l60vW{#&QP-qppooHrAN}YMMhA4 zP(1n1ByJMPH<`Q56uSxIm&oMc5^3XH!=_BZg2h>jhIUb!G}v^}4+1c_$pS&**W!8h zG}*&H63(+Q{9!vDbD9Om4G#~!{qZz&2{Q{yVfms))u3LgtTR*_<(0;)X7-Iop~nl8 zXcZ)0c*_1KI6=a3cjbRI`|#Rt8ZE|0c?Uiwb#F`zd)vhDHBsyZik!m|d_Dabd`2yJ zsCIR`F#fF=KVarie&(4r?}2ns(WDi28~@ht^!M}1d-B|w_0G`e15)*R`kownbd#AC zf-U}r2SC9H`~SO{L!oWVD)+UZAihpq0p9I0>2=kj9xO6AluSP{9UT^meP*e1LF_)4 z#?ER`H@_HA8L%I+e~-ys=B&D2jttQ{)n#R$yVYLIs&B|Be3j)#towE?zWd2r0wc6$ zZu)K=S^maA3z|8GMZ+l+yPhHmpqm4&Kq%S}E!sRCDp(Tvxl@vS(&;54RB+INWW;#+ zQNsjr0ETrlYNNZBefR%ve%&i>VXD>V=w4FlS*qB|LgD@n(b2%9S*tzAuLo?{Y*;_j z!yX!kLrzX|!nF`_n|V6_4N2BoNdB+;;#z*b_iA2>{JuD4W+oqn6`M4@WZ$eyA!kHw z%!)~svIaVK(m8sWT&KDioH|ZTKi1>vxsw3&)Ab%f>b#FHmJ zT47E!=&V;7C<^(Z_b3L#i_dDT}SLRpQA^6~on&;NSK$5t{sgAq2BSj{0Y8J&pRhACg+F>+2?jM7}b0$ zf)2cmU4=+r*5?|}!zwtP>Y)D~ewAVwkQ$DIMH(%rY>pzc0cSqGddOU8%wdg-1?jLh zrhMvwx3#4+QKAmX>MlfF0>6{MtEcw|CW}_e2Rl$#gpQZn;apl`HYfIb$;pIFQ2PIQ zs<+XEJY!F;BsvF9=0jEWu*GX3#qOZUb}AB+6WDNmmCf_52r8eNOK;H{3`+dcH5O7 z=^Msr3Pu{hPUkp%XB{_7UvTj|8^@A2bKrQyLf))QRVZE;gd2A|L>PA%^z*P{>5e=j z*y=>6tzcz6zHRl6){S#N;{W)V7gRfJ1Y)7m8^yw-<|U~+{1d^Mlt!EZ{o50Yts3b6 z>vYx&FMFU1!i{6#!%7CStg?t)x^gb8k##Cm4{LyURWEC&4}{n+nQHy?oA{K^UsyMr zobTdf-X>(s;N3NE8MM%u*RLqzbqNmgw!3{ywgcx_->fQnk1~%}CF@sUvsH<5ttV6$ z3pn>X`t$-+9Y`?1iI6eECaoO%tbEd$MFxYzwR_kdPOX+Uas>=n!8+3Dd?gzzIH=e08Pb)uA`gM zE;{?`hBo61(1(QFw1pl^oNqHVHfaZa`2<-z@6TN!#nr`PigN z27xEs%Dqb3m;(zICoO_l`00g;IdyS??V zd);HFvXFm;zd>?_l)ChiCQX9in0Fm-<-9II84%~h@Q!&OQ&f@0paBqjL6lU<^uUte zscd!I=DEeINfRaN3A--sRDPz|A&m+=Ns1H?`F8`4fL7sgj}(uBupQwBE`+;*^}-BQ zhd9OS2tS+N09k#V>h@fekZERGg~*3+01Qd* z&dzrm-Z4+7IPA*ELY`+Wzg~V31n3^h4he5a8#ULyazlDSvqy1Vx?{@n&{}#0h>t8E z$IL9p+${Gg9`zI4uu;Ate~Q?cn_Zc9^Vehzx9uGVF3N|*?XXPl7K%-!$ObC%qC2)* z;5e zfBTBUh>!Fszx@MAcHrDSQ0WYt@ob6({nK_7Z_b0QQA zd?We36@!G~U8HiFc86#y+s5DqJak969+!Q%exRejjoVZ}{{nIOugG8jF7o&P{j>jg@4uwWDRwDEqHIdEG@3u=(Cyf13yroU zbn*FCQa+MKuH)u@NVAE3IZd%AC{jyB)<-;)WlAyvu1h{xM>0ey{5^v&{`2$m$;>*k ziG2upLc9xwm%D(X(7 zC&+hH>=O%=_YLB+UJnFoL87Bklk**nX;IVi&f6K%y<~}c9f_9pJ4K0N;NNWeqW52` z$Yl~Afb1?u04zs5@zQ7}+v<533o5LBlTqAg85JXxzEO&E_fMENm~zS5I&hS0q4hLL zlBX(}-si5Zkn3en-LXOMmZtETg z_EONuGb{?OrC8Y5s-_}4{QBr_#`*gwe1)q0H zR(t08q_{TDZDS1YIW|LO6_Q@*fJVGwOsG3^=FVhiBa&*rQ1y4Rj2n_1*tWp(X&91{ zC^nHI@l+(Xw6`&_g8hNe;8j3(hhbi?$@4*2E1e-3+az}(Om{BNA}5%Oa)~v z&l>*=uVr^HTZcQ)6s`Q_hgG2H2bjwZQ{9Ysmp~nJ0%`?*JvvFyFC(Vv{J6NJZw!j(9$x zZ%9w^`xVO}j)?J0^D6c%0n*BP{zY=ktK4;b9f&n^Z8TLjm_h4@;AF(8FINR#G4Bnu z&{T&CNvNlUJq;i~E7Bs}8P+B33>hpc%Z2)mW?7AYxj>8c@Ns-}A>%M7942Cz>#+U< z9qpZ*Cc<&O$$Q1W{_Fp)x~1HiNj{)x3jbi?5*8#}3>qp(7zaF?;LU1qn>=Oz7d>~E z-7?RhSx8h3{Bu1J-=KQQQAw(-gGq#4_`8Aaiap9!;lr?vvVFWbK>_e^?w)^EoE-eX zRj0ZeSjevu7AYTk=hDZ>?r{5l=vIHk!>q@67~Vl%_fe#Tiu^|fT`xaA_jYKz8a1^$pq8=mXHDOYrwSJII~k64Ehu*}92G9W z`D>sHskkJdv2vX_i4ezFdlZ0tlG#*zz{U7+k?&|Iozn z4&1;|+IsqTe>8$4c4d47=@@B)*>Qt(vU=k1Me{R??W4$J)S$x@Urazd&^KnQ>*+q% zEy_VzFiiASdqYMZs&X;m*C6j!qy=Nb?-YNDC;_B=fJCc2s2)hODgz3UMR~qH+J|Cs_H}kLKeW(=(htd&kXz zXXV|kdEHG#JJ6Nq@xUFd49}OW@c?IYK;25`(l!1E)Z-0UE8%LxC6IL}<^?i&YQ(%U zgm~GzaKFY0F%!QwG3Ly>M#L0*JpCD|ifyDw0~Prou#CzI>2p0E zcn4y}XhB=(W3tz`Rd_wTm$cIHK?lRn`Ddz=e3k}2l+_FCh5Puqz<;h&rMMml$Gi7W z+oG(ZZ)>6gI~LxM=JD15Raj~GF;SZW8k6YDnzNpH{1kqQZ(?|=tanZubI?D{C*QXz z{7)sbqG$biQ$UsQ%KS5G=#mW0RDV2Mr#huNp~kUSyS0TjYA$HFc0}3^hpz-GC+pXD z{wJTA^N_x%_Tso6mTWV@Pcp@>qsUsMp~XeWc_mWOYWoSkCRLAq{yV;$=e zGpcmls6n&f(_(+6(GxZPy!3st$$?`R*bo?2osdVdITYB7$Q`1zd0NPj2-mQO0y^kJ z0icLmJ~he5pt+%qV=JeDMvu>6(9*`l^O7Pef-qAH75Vb@?un$EuZIr`7gO=^ZNK4|1MU7uHUlMb%Ar73vv#_#{nyiJgK1B)W%MB9WXG4p88*?aN zc`zjV?SvHz6UG5T7Dlko>2F?K`MX6%1M-xg`D0S?#)Pt0P0Y<1iakk@Iw}%1EI(22 zQ-7lVLoqNg1lbX?^GVN*7Fjn#fG4_+SPkKho3#^ni0^XSx2r&40jZU-IXRX z+kN|;@UEM5wg{iQ6>>%L`Ppbuq4=!v{$m8s7i&^n4Nt@dq`0R0;qNd?9C84Z1)KD$ z)1w2NM_qE$?~Q=k|1&o^x$D3lX{AZ({|UwRQsfa8nN1f3=F*=EKM@SpS{t&nMS&~j zpqTN7UV04^|0DUhn$Z03na4wKh>x=nJ!Q>oos%Rfizt)mVe>ng*$3SvTG&O3XjEZ2 zd%gtffSe7&WUJ#eq@A{#LmacE8l53})EBCK?{#KY;YRdfKyUbZNn2k{c>*5`|^D_8Sy7cyD!l=*lz zN_iLXkI(H_bTHaH(ROfmaX<}IB8E*;=v8_;H9I`T3;X;HkLQHsxkD9TpKCj8s_!D@ zf*e|FxTjTEAEZ;~i3)hvmD!@LAk*0bjEd=WSwubmvbb2WMTXb3U67WCr~d!I)3r2b^Oo12=dSBY+!LbD=Gqw`C93n&7t92&jdriZiJ8uuQilcMyl-Zn_?{7jzufbyKGN{U zfbhOa{NF;c*C`+qh)f5mwFkc}U)cENr(XtUtbk~@HeOA@4F*M~`{otM5Ty0O1L4@* zrSmBiACbEaVBN<&mWj9S9Y%K-x0EKCu%u36v@Dszlok@t&9XRf2oL-E z!;-m~gkT}QqCz|TxZ&JrY992dqskJ@fI=3z1Il*|^&M+F9 zk^~U;AW&KgwrOw6@dHw#T(!Htr0gLN35MeK7P%Lnpc>tz&@kc9*T5R zk*7oj5KyGg$(N+lyPQj=A9#DeYm;I*=pNh^Y!1?<1R<-0;pB1GLUDuoDs)ez)3KyD zY`5oK!DXKID$@*w#;rnFM@8gDd^}6r%oHiFGFO?hh~ltRRg4DWouzY9CS&n55X}`S zabWy;l}QfK=FZ$A(aZE9Pa;~x7beG2`I1A>ThSK$AmE8}GkaVet$7lx#d~wUwH6W# zc=CS8#_7+33m05d;BDGxt_|{~ffuzMcD}FSo>~d8M+&OF?HV5$ZN#YTlNI)5thfANda6%=(LNP*?1?71lL)j- z4T~7|QEUlC_E3>k!fM{LS(iw?qTgw)XT73HlRL9YhMS;GnltqE@baK80rEU!GVozo zMNmAi+PgIbpNNOlNIt(DQYwA(io?#+dHkKyNAfbsWnKq;T2&<*D(JE8M2zYq9|xac z`3pP3Mug*j2aEsu+4r1`$SL^jKRzN+4!l37Gl6sh#jc^qDk`#gkWAr1M9Ms!Dwjq< z$aed#mMnc;t--3LXYF~O6ILhuHtAT-TSiFu{;vEzvT>wf$AQ;Mr6w@hPO-p)vK1RU zdI+wQU=MF9=zN&&By){5xZW%7(;%%Ou5ttq1tOVGc2eVxc%SgNj55Bs1%OP~xpU6q=;Q z_0ZZye_i|+ovM!B$ZD^W+?lPywTnJl{FfYhnb!&sd%Lvg(s#Em(y8!Oh5uGInaK%0 zO>FD-p3kKJ06rb=N0tO6EodbqM_dOKhkUuVJg6RhbeD~j93 z4zia+(BjosK#~*4CQLk<*g4$@vYsEymy@jy46-T{kQGrZEQSgUTLrrm2?B#p(qh5$ zprp?&Wr?&bBAdodfLx$>)ZTT6)(F|4_i6>-)i4OmElMQJFAk~p*5V)?bme3n6oBYy zXeY+oZMtD#Q_#Eu2wX060?ov;Z~Zp-+eXm*vHg#?$*I>Y+WEu;NLMNLGDR+8(c1oL zhyg$d0--~UhHE#klSZZt+@a2KE)?8RmQNe(`~rq1ys8f=pVllpKtBxAskTz3ZrX09 zOMbwwTMnrcy#JiImmDEcp~#Lf*dH4WUFx`Fu~Bw+4&IbRZgx(6IAw;x-@0eie3dLTcNFx>bVZRfWu`0wyvA9NI2nRk1C z)bF&zzd&}z`M}%I!hBf0!#`mrl&jXuZoA|J|Fz#~kFs8oL$@n{F|)`ofglz3-};@3 z#f3g+oYN&*><>!yI79y44}2(l{ekqy(AvC>|2X^~{Z7}!*PwW@Rfy3vL`yM?8W%*( z({A=}V?J73@Qq>udPKms9nc*X>(V+|rz)TNY)-!u`YbKBS|3t(5!_+5jBD;te9;Mw z;fy%~pTEwV)|q8aKIpK86&6?#vD0h73I8*AP@G91OmS&q@!S?oq4?p{CibwW4*tV4 zS?XffJIdW+JkEvMR-1RVKmT+M2VIixuyL6Sh{Z< zIC8PjebdG~p|Q5J4HjqVKHJ>3ksjV19@rR2!!*ik0+M`k={8lDYny7j8}L{o#G{3eVSy>B2(Q=VzWr z1cw)_e36RQG{}z@Rd(C=g0LJ+`VF^Z$MxI{$Si^g^dHXfJwQ z8-kK&=L2gkBzE`kn%Pzw*}1fJstkdF)Al+v8VqslLzU-c8J6DRowt;+yka`tuY%vk zRPyQtPwY@6VIy>m1qdTPYZ)CQWeRzVrkyh<27E)pS__P$?doKHE(l=u(6^Wsf@tzc zk*Mxh9DVzgU4xMkr^;cMM$Q-F?wY6aUkp(kx6}*D!G{+>WKt~DQ>9UnMa;E9IUt~5 zlbsc0OZ0Tgv{R}k&0gPK{P*@K>-<5vBfKoUpVUC`;1-=zQ%O|0xM;B}@#|HE`1Fe_Ob@ZL4-Dl&BM5nuzr>-hStr(yhbKdbE%N5Ghy8_us^4wKKXS~fEy1TEBz zc3zNqC9j2oS!9J>J2PE^v<=4O#5n9CfBuI(KeFbNe}3~g9&q~P9oJ@`I@ne&FnXhz zo4#8|mOJpqa+ZmINuk*F6iJ{W3&jPpQ)K_NCQX+JW0NcM6PccXb5Lb>W&S;7hnv;k za@)Yi8pIk5f^FW<+EF+mmU?T?g}5$!txvUg zg}5NBF8ua5>s?yC`(x7ti8V<2)&3h76-I-RKION6Aj#aqR|m#Lv59%frdSZxP6rMe zCd0Yd`-J+8v^cC)m`M@^y(E`D3srLQK@Y{kCn4jj8y(6Vz9T*MgP0(~w z;)%(NAsfNu3?N8^VVRM|P!R{6>2%sow`uR~ZGZRnw*R)Ki?`c!Zf4r)Z3j^l96?l2 zK@Ffn1aSe}P=kOXj4LQABd(Z;%it)uFsSf4u?#ST1Ay&VQp@K*M<>HMkdneSO~B$53` z9J%er4obRB4|U>b<}pRODa~#Fa_FaT6t{_+p*MBGn6kM0(iGUBbSm#l>*;yHcy6oc zFxjc@pl{I4aTld1+xb9oU%FD(LHB<5Jg`eOCw%_Zi&B*Eyd=$0?+l(Bkw0+-@0z?7 zsD^-kl-EJyrzA5Kfh)XHCCv$srB`3+pchX#O^kF-ND?XJANINdIQv2{H zuFxy;$PPD#AgC|&YhgM_F~IszMQO50yPr`}LEVvV^2wL#m z2GVXDAeB%Il)DsCnyp}q=1^CZ%?aoyHwbV0>iG3^o%D|1y5MT&AgQ7k5)At+9b2fj zoGFnth|kc){QJPWUWclQ{Ms2OB|83R(&4UVtaCS9xUdd9%oE(;@~gk~Jg}z};gVx= zUf96Wz#yNy1h(_3~0Tp3z-}8fM`vMY-k5> z6dD=rK3;a1PW4&nlJU@X2PzUqWl?ax@5fIU{d{4{dKDB(7CT@Yhw#Q|FKhfCvJgXN8DU*^FISsA0J6|5az}Pz-nbR(4=e# zc3ZNMCgmH2o>o)!q4xB9BxNA?-Hj0jN)i3m!Icz~MUkbHrkU9c&2dmwgKgXE)x{#5 z(=264^dUuRoFN9)W=6Kwlar|$rupG=JG-g6DxchI9~oi^sQ{z2xpBT&f?bNY$7U&; zKy!I*5O&r#GcGg#EdNf2ziO8IU3HlKs~*6KaNkER)bP9D-&jq?=*K^vL`>WwL+%?S zCv$D~b@wQygCd_(ns#|NxWmASj_J%?x|&xk+2N-P(q;x(-W(MOqK0_5G>9iy$3n`D^x)Xqsb5}>^YTR{{hL;lX?-l$KmJr}m(Wj$R$yZmdc13Ar2J@16oE*Vr}op0r(D zPahEA*p1k<=DJp9=Rs8QGRc7i7fMV7blb?sG8IeR;8Ve)q56ms@dn z=v&{uK$g0(SH0C{Sz1dmxfEGVY4Uh`#x7KDiNB}ldUYE=+3ztZ%=ajAL*|Gs0qIhu zU`O<=h{Zv(Y6Ru{#qb9RucdQCR;!MQTwh}0u+_BNg1gFE_KnNQYErAx?o71LbG;zG zG)q=2ds1)IBBq;vGgQZD^T|>{^2=Kyv`1AY**4X28T7@CtRxyY81+rE>whNy^x)gX zYCXz4@{W=PZfre(rKq1%yMkgeDYAspbbF`AZw~76HvqxMY1wTkq5z$DJ$;!#^MKmo zfsLrI#bvHtKf_;s0w6ZL;6)2|87~XS*kKuy`bn!+472kMl7!B*Z*{&XX5TN6uR>?fL)4T5U?* zSE~L&<~?U&6-bHqvmBWevxFk)KqaGoNFz61iwqmxaYF%tHF*hzf1E$N&zKfRC;u74 zIQ=yH^5ZuqTJ2|r#ZQipIz@8EkX<(4bTP$1b?tgeGiUr7pEh-K!dg08o#Imv-75V2 zjSFv{odIk-Z(M%8^6hyub*g)icShCg74W1UBglUM`RO+w5WVL*>9Nogq3fhZSJovs z`u#t7%xqpb{cC?RAp7?}K%B*WD-$lx|M9}4A~JZ5sy`2O4u#RPl*j!VgqDkt!l{zpCT)W9g*uS! zx+6^*XYhI~fc)T_H@<%BJ3D5aN+_AI-2d?_2O&e4FL@-aB25XOy|MgXcfEaT=KERy zes{+ENpIYRtA=U>hKD5^U=94{2|BpFctNoA_M1_^uqSfm;=#D@Q$lvwJQ7bRUR#Bz zL01-03#lWE@*AYO{f?>ibR~b`xLo>EaUWFt~F+rr9T#sB+L z#jlUtEG~PsI2e!DNeW@}j5e<**rk))e-5}#c>jA{`I?)FYpv6D+P7W5AeMxsA$zW1 zk-Am5HlRjYHa%I|MUI7@0ZxY+sGeE{WiuD0CU|vT+@nk%Up~H7xS#$^xjlFZ&>x~u z<^^7osEZgp;dMl!Xoche{{pX2-7c>hZlTP8fZ&R-0awZaM;_k6b3gt`x;?cN7tFZv z{+We{s>^F9_~cp26wyveMPQLidzjfCSU@ACc$Ts`{A#%QpqWVwt^iF)F{FV|l4mmz zbQVi?haiuY3oX;|kTN_p40iX4+U)oLxvI`Lcvno@7JOjF2ib32fNr=#K33mt@J{EQ zCUr9|&A?J&e0^X1NV;|A2Y*ZF{iW}+ri2Y35jk|1d$4)vJ2xzx^bO*IpVuzj`um$! zCuK&3u#Hr6i`lxdKYP)}rD>#?V-z_|X-buyv5N$y%53s(kzc+fMQKzl=RX$w`OD*3 z%9HeRei4|MWajECmRjMg=mudqT@u^snJ+00o-4{TuUh>F07d4Qy&-cT((s80tB@1F zR&B0AMpXXwgUs*tp{+S#x&L;bv(q0*Prlh60xKS14}rOP23IpREKCi!WU(CNh`Y?L zbBFB};oP<<7pi^x1p=OEyrH}(lGNya!|iD13(s6OOaa`EpJ$q0Vi--aDK$qZ>QHENMnd zMjE-x7n1C-Cf*70I_NP+8v|udNE2@s-%-|)xnwPk3{@_WF&=1AhlHnr26FPTs{h*l zb$iWPw>2azDzsIy0@4YR0h!@T!?#WP%&Ru66C4sOx4jUT>~{l7r;n36g4!@#O#ln7 zlFG548SYbVKZ0xCb2A}zdp8OFtzNHbdelX->V@%oD{TCkjTEziA_bJ@w13Nl-r#q; zNKqU|_mnn{=h$Qu0_}NnhYmlQ45?B6W>y4Tbs!c-`x=Kp%H( z+4-xkf!3%czxcT#(u$VUtd6Z@lN$$e58KSYfnuO7X(y%GPZtKQpMD{xUa>R2Te&m- z`ef|Et&&}c!`B$8Z}3V4E~*0B5LPj53o!^TE1P)+=qSN+xgi}uQey}}@sU&0jzL{4 zsIFVbv5dMsxQ_nvGo0nx4jU160C5&MA~$Q}Gp+Df_FZf*NYsv{=WT&t4Ok0QkHoGK ztc}v1o!Qo-JZClqcUA z;2O&3KI@*FYvXTy)oN}wP2$}lRooUQH!h>QU}KybDCQ_d4pABtRPaNG(Om`he3gLo z?fJ;tz74`2=&7z0FAg&D8o=IciEm-F2cQ|dW73@nBly0(9u}22Jl)N&gA%K&{97T~ z21U6TuRZN=Qb76L?Xi~PntkHU;#`OuVnwFG}^iRKcg>I=blV?+Mbche}({&-f=qJp?~3AXMLnB0{dbzP;L2ZA#Yu<0^No}RuJb&Z#$MB=>lR(Hl z@~R2F#w%kU(u+MiCiQw?+sQHilX0o?R_~ReH(!A&PPp4FvczPnO~Wj!8ye~?*n7s` zLw^+D&fw;M`__M1V=I4KoboEU#x1tu#*WH-8_I<)in<4ocG~-{4&gQc3M}1+*z@ z%akU`I=%@w5(?pffv<&tFW=x<2!H2-%-Y>4z&rvh<5t%NPv+81pnH?0+^gCa2t8~l z7`h~OS;z@5%UwJCj{9BlHG4f>q+%NS&Xx46g)hx27b8}&kNO-BPU9ySJkeh=DS&01 zz9&Vwm0bGfOJ*kty{0T>sqePHR!=Q9Or|k;(MS~mWD)r3hsmaxBvFn$Z=7~rbRoUQ zw^i6io1o<%%FrV#`A9N^sf@iUwKUWMU*Y^ym_FFKalfOx7h^&I@srA$E z$y=C|u#%V0PTx*C1ovNSPIw4EAN$RPs@^kX>1%tX7bHczrJyyN&+n$!Pp?*2`?=6V z8lIgRmMI$a=4{{c+MiqP)T3LQW#obzr<@+yM0Iab%uR}11Iy!+#%v)ELersN0O}O* zC2HLlOUf06>I$Hhg3S_&Z{c&>IccvCM*X!F@CL7(A6XuY^)j3MJHbnCr)ST;Pb$SI zON*Daf~~+u-qH<=X#&hMph#_&5{b+@6|16Qbi4)k@EK!U;5V=gdtaO>S?66Id&L*^ z;2|-A<9$L(ef2;`(?%zW3{Y`yxfDk*U!5c>k{*J3;sKj)hXsYY@}YE&Ab4}KAoD_I(O5km9W%8V(TM?RZkvc_pw zk1ZUV1tQPL*OTa(LZwheP&U`?)iQH2R5jpe=Lyi}C|ZT>;?CGAnKAynC$iDuwa$6` zT`*0y-}eyTahoRS$swNaX@32C%^&{zXTP2Go^(FN%%w=8>!QI9{EM@Zxy&&Ha~jeq7qcT+lNj?& z`^2}q0+_I*vXQF3Rk-!__czaMirFhRcrJbY{rVaB612VI*6|8O9W*4F{I^TDy<(Wy zCN%PzVxX`>ctorZ|6HsOZHv*VKc1;Ao%w(mCT@3;^qXVQ7<4gl3KZ-IVTZ;*oBUrD zmEKiaot_m-!agO-xoyteI7qn7#ziWm7~p@*qco^#ox?-g=nMfaryrh_{H+q?0%1|T zCP&ZZXvFe~`CgsM0tg#rDS@=r>=J`?0-S?DBIYh!Xo-dgD|;|F?|oI<%D_Kbf%2!u zUxbs-xBG^94ojQ<_82O7hVa#V?5!4(Dn(gnI#cSK#Xm!L z(05}qgF2Nap@o{fT2KIN^||z}iFtxArZo8&0ygk0pIE3a_e~4JDzQAsqT{cW1iaSx zI~DQ%HtVl<`5R-lC{F`T@3}Cvfk%nS{Ht{lovJ&&lFsKFK$xMD-zwY(xBmT#vc`Pp z-z|5ppPt4TBD>8T=$)j9e~@WzM^PTZ}5c*LfBT_rnmB1=>}lS z%nm~lyBt45zzK+Xe~|tb3Nu`bSxQdHKc_207doHQc6dhX%Qez@=E^eCs{1d!A+(}q z=9v}y$wF?ZapTYCFN5(h)mp#i&kQgN_Rszq<61rp+xo~L#?f8u$u zwh8hXSa?(7w;ld24o>yi9=a<+i+oG%bgTES2!rRExWedOk4n`8asndMTlnY29dTO( zPS28Z#le?m!TyzV(ZI=J9v%Lz#9S|{NvZhS+b77!1CgV-v9|*X8T|~$DvDW2!OUn- zOlMnQr#YllNnen*5|ju_CZ-tUlm}CwRCvlJrb%v`a%+kUonO0&<%k%&yIuE|9Wnnh zG5U%FZ>R;mRRWNNpl0d5gCgaWW}eq|VYxU_bkF~&Vl9+r zV8!0KdnYC6!#nxg0#m5u*kaM+=o56FAVaWBmM6`L zIj5|myJSr&tm7%9k%(STn*u5$_mKrtmis$bj@xmh*)K=~ymun>qW^2=n18WiN87dX-8_g71rW#-Em5fzV55 zY__^Ouq+}cu#h)9cY9TOhg#k(XSkqAHK64sF>*jR z$qIUyj7 z1&kwmLQDDD{c0VvY+ReL3KEU!yjJhc^mfH;JmYd}VVx|_NErSkJ2V125(e68|8?co zaQl?D+oFRdNu3>gXWYFw%il;E(wu-Q$7nL{j6>?@?YvUD9TJRn^y!%MerOiv3-*Cz z+XlM*mD}`HS0<=&#yA7ziEVIbzK47fC!kQGcWXbhCyVCd#=5b`!a^;LRp^zWIZUJf zHCnqW`f}v;=tkb{2?qkShtzc;D?QJX8(!Kw-ndPkRTw?X=_!B#Cw53+8gL}%9{l!8 zX{+tw@pO6J^9Oi7 z41Fz$H39iEs&ir58M{iL$e7k@P4w+4R+)9inhK870(9cHS8vi{Ad2^$S`xe4P zzgFU5AA5P)D!FmUmnFi27RjibC)eU!RLOG94RLAAvqP%RfA$@;Mb7`=XN<^hi|$`I zQup;fTlC9wpOFLHY>^worPXG?-9#~s6gh_UMQA0A%3WTW5+vy~*EtxKhJbC8z~~@J zQKq^D7=ZCmx6f^1ug6UvYymF;4uLz+u&J#HP3N`9Zuq0bJqiW(dYE9_l_{y1cK=mV zz&Wy?j|KxNi7dC6KOi|QF}{Fa$*Ty>i^k+Ah6A&d9pphA)Q%SM5=HHhxTuU=9bL^y zy2?SoJ#nIk_2L;I>wo&wu=RzZR#=7Sezb%Xa)Xr{`$h+BV6}r{$|z=DW2XR^XI;@!P+V#1{rQ zIve0*P)r&{7E&6_1tf}QZ;4M7T~?Muw{9+dn;H4!fTLUUVb1RmZuk86xzp_Z7A~7@ zH(sJxJeqT)GBQhfmERr$)l)!dUmm+zal&V0_SAM0VNa4y&eY$ zp0T`3msnbvbSiXnFd+#!N<3{;b}Nw@DT|*HZ@HvLv3dO2=}l4Byi+`%y^0MNGVEX) z{LHhU#=c=1Aa5|L{?i}6KiUegc?%+HNw0_7w*Er;xqlV!Vv1QrkyJ`!X26>FD$>@s zhg^=wx6huwXMK%BIPVZ5hA{4MZ{dW9(Yc!U57^5Evgqt#q0ky%EJI5aoswgdO&2hn zniN<0JAqTf5Lpu2K`)1rqr>#|$yMsZkiuz@KMvj`-7f8>Z$?#vP}U{s){qXb4*GMr z3L3TkaZiD!-}$LOho)xm=MK*fcP5^?pVQR1aee*bv(#{08aLiEv$!;tP>504rrtLZ zD_y<-uIIvN9hB_l(4No$7PxW!M8%(G3bMGF2T?bSkl_ z`ep>Oi)$_9F&%M5yhG~3Xe~AxU~CN|z@=~=qieaeAwXLR5tw`*qY5uT1}bf%Plmt* z6~dbY=*2iGQ8hf2uvyIEV}{)@22A>wi}pR)EX~YVQ{P6H!%EboDv~~qTCYZ3wGL@k zG}c)|*|eZrnIo)*zIa1GuH>5UnIP@Hmog_JI~kC!V$AD;x&R*HTXIYemcMe4%--RL zT1c3J(bESMHKC8A?nQxTdIcKiyA|eUd2C3liYtY*3O1+$BOVP~pHkS$U?!-NNno7PoY3>~Ja5!|s^(P-x}tVXT;uC}Cyd=HaI!+| z!#Mp6C(KeheR!!oIknr?V6o6muUBJ{ca!RbSi6DN1vSW!=5tpBGAv?Nsgs`%yDTk| zE%phnj0jFR0mQm0IUNlH49E!}qgCJh{GPpDDT^#ANM2& zNf*Ji|I$b(96%p6d(v`mbR?5Rm&r|U++1{f?^b7x`@ebk?R!7_v*zc&`5;kggDHrp zGu|hC%6J!ehnXaAaId0_Q$AHUjs+?1cUHfR4Ndc>WIB_GafsMq@wA)=DGBqA(!LAw zxeDpA^g0?mF;Z*EEt;U%DTCKQKNh45Ad86v(_0ifU(wTvBJgGKOtY%g*BE#ZDu7Y^ z%hegEp<}^$^FGi>?5@kK#^U0CE;;qXOexPequa}(`elj#Whv1Qt_elZ^(+z&~6 zxiS(%dxxSk<*r|UaL$H+oq?v>>BlBb*mIg#V{2s|c}K~DF(4`3KZLb{VlpYRgwmk( z#fNT(VlAV^XzAZAm5W8{ya4oP7ef4XWWG$xx4`Kub@~^_J|O$!_*{9gt3Fv*&$lAO zBg7+#B)>2Bv`RGbGr ziaTD~vXCRH?C`wk0)CrXTMs4oh!L!n+D8(nY>7vz%iI13W3~hsdD;ch-QI;X4xkUW z)LcCT2j(zY9JE??0G6IdVX07IhnLUceXOjLp5Y~WYR^KBs?V^d`rKEp^S(?jM{04m zWaQO^VbTm+{!hj%oT|M&7S%9JW|fRQ$$pSU-YIHSnPfM_>4UccI1D8gw{7nRKoR=4I#r#dgWeU9CAvE$2P8Kh z`(2-WYT6n41B}Bh3*PDpj*7C5q(dqd`p7nls?@bUmZ3J%58|A@$T|xjgU$6&&avZT zpp|q^O1&V^YM34~hf|5pjnjmCZH&-nirGk!4U{H3J`B6Hh_MP9uyx+biecQ^K~S9uIz2itBzU+4b-GhpzrCFM7% zFQOQBnXjduH$!m8>zGd~W)D%K)+Br6)i$CYIzICb+m@jih+*8rZd>N6=Dlmb_OO_E z++SiYrBHnuWbZINTMf;AP;Ry{8XJ`Hz!l#H;ZB$W%n}w!wMTfDqncFM_;r<^7>~6n zY2&egX-RBRu(lIc`T=#QhVk#T-sMC0{b)!Wa8bnMCIbxfBV9|>&R{W z6_!4ti;<~vbGopcJd*BJSJB!;QLbdKx<%~t^}`kePrq{%txw@MDuo4!1M~hkm?liQ>E{j{u9ezc;JlLWdl@G=9K-ojvSMG|d5iM<-MG_{MZ|6K>pA03 z6Wj;oN-(Fa)dgky6^m|<-QicDZophoE)CVYaB$7U+y2Mpmw{wH7c`RR6YEwS>tC{z z=(YiGX1r^RmJ?7vBve1^Cf`lWC56@YdZ zb*SgrZQz*)%Uk}_ij)7`v-&M^){R}6E*nhTpqMs_7%9z(Y29Kh@P{rD;YI1~DM%}Y ztd~`ac}&*0+1SfyqOZktdugx5VEh(4i7M&)A=%J(u-Qwe!uzfSZ5Gci_1!4S<6*DD zez<4W-AUCj(!%hbpjjA}ncbwygZ2hkdBN^L-f^P{e`=e7OMZ!WWl$Ed%Uf>NsTL({ zoSy0fjoG-o;_#?WfWOvx&#j)#yh3_*3S~iZf%+kBIhmzg;CUsgSq0qmPuDpN7g&QV zs+>;Q1CSk91MTuc?wr5Fvx4o9pWLq}b6=Ru6FAlV%+)fA$)HFYr8(_iuEdOkxoHIz zt+l8}VDa=^W`*o0$@<=546)qrs>6O^-aq&2bCZvcIz@8a*aYpendf4P(NknSrRgGR z4DO#E`6V&$7g6V>8>#of41iK-5_3p#K$;%k|9DC4#<=DLOGV5INqT&lqC>v_)gpe6 zM-r1MaS<*Yo_wAK`!!tLU^gzUWAS(}HP{t(IjTH1J#wGyR!ALv zD`dVGQanHKeh6*fNI!_&3vE%DBd}C*m=qb-%1jr5=?N5ujH)LixaP%w$W|3HR-@we z@cvn{Vhq`C^X%#=2K1})DGes?T4bfZvn^(Mqi8MDO(%!x!>i(&c{c^iW05?)h}T8- z|FGhl`~8X~t-=Cz8-2%fU367g+p8;OH=~@Ni0SvO4EZa@2f5!k&VI@1!FAuA#=55P zd0(?4<-uED|2e4{LoV4M<7P5x;&? ziwxc%%W1Kw5t`cdv~`cuBFj>m?_C7@7odZK&nV$*FDtu9O(?dH6_XAj#~$Wkfr@p7 zU_0SB^j;89^=$B45^n`n@Z66#l3i}>)10D{Oe?izz7Rd5lRm^#ik}C3LOqIQ zkr`w$zeMl=e4lmjvzWhk2ENWx9(nzuv{SJ)6vx>WQLoCEc22_UPkPzOd@)IQ$m#ty6k73l$=VyNz11%N}Cdv zf3E=)y-y`v0+z}4O9^Kk3BNyD*{5-NTKes66CPl8{Vp~aP6YR>hL`rZitQ11kOe@5E%Rn`$ zFj^0xlQhQ6oZATENyj{wMqd_RRcfn*bx@#(sgrbaEv5ieA9Wz21ho*>s(M)_La$KJJSLkz@tkUj#}UifqzGdZk) z#yWH4!zhonh7}7S*kd6|gVyee=8$zOeM+dGa6YmD$bii{f5jrZ{ECyF1h$oz)A9a~ z52q(#)Zbga$+M5_xNVx8C02u+&z2&rzPimCQWaLpb1vg<|L>UnwS%Z&AHYRtxS=9s z4)w^sm&t8W!P2)>=6l|Ad*G5+cLg2~>sHnTm}HfaJ3O^*KzO55rFd#_y9oUM(kePf zgfs(2B|i8%6<$3I-^HQ%a}Xs@+W?0~ANobyP;&TJJ+ffyV)e%@2NsicZoF7k+bmY) z6tk5gn~+Vu3T7UuH?x$tfl)|nd1F+b@Ip;k(2%fLmh--IW$9s+L0Wxys{9zq5R}OF zh3%47ifh7jst(^uZ}XK{Ue*k?WCs)m?=zux3&zl=#d_zEQR8A9aYN02%2feVt*FUW z{;{>pK*Iqub4Cbj&yIxzKJVdt`9vXMb@fZx+u7BL}+ydccFvc z6a2XpBXy{YbY7f4@itu}*va2E%qlAz(Jlz|T_Rush_q zA0#I;r&UH`aH?ohi*t!-972nOXmw~TXI{h!t)p+o{qXibSBuW2$cmT8V*VL6 zD=jiip`)SeRGNkpg&E2$I*1~$XG0l%==c2K&Yxu8wJtil{$KD<$YD2jRj%8thAk8W zwTGubgbTJg=b)?#gU&573kTzkS1pb9+9VOSn^=kjwcAOj61lT*51t#8&P!twBh~~g zRu+gXue)LwE48M8E;27ZV=V55u_Ce80}CQ{_~E_@N$88z*x--b;>*zD@<>(}W)PIe z;tmT)%8|c3g&>P3l1c9Hngb8@xNlq4{>?yXSI^BlzJmp0ln!wg0t$`;pb& zoc;ImhotF+DVXW7u}(&cxlEA@lxB+}L$C>2+cE@&WA9If!YyH&V#QZJ1)9BfdY80i zY+iJh5)BO2F6-$Auo^we3nFACYLsKK74psETBcL=aVoA?8{$eq|)U(tRtQN&8R zz;~?-^-u9L_k}soNbU&ChHjA9?|c><{ZaygCRt&2|0nYKWGy$CxvzIms%&6ZMlnEg zRYGZ6g}0*8qN{*Kf6drZ-y2@tUTb`J%1WXuBX0_@kp4cICvS)ZndIQR!V^A={O*%& z;?=Sq4?NrL)y(9EAm2X9@yxmfzIIx4k4Kvt)-mr{PPbpq0?o5P!1a9grb&h|R?sAv z-qn&hZtM%?+Q485#iUc@V@gvgJ{Puu-lpi1wnjY;PM4%b-&3@^JZv}6H|@|Y4|Jo8 zk8$!~DCXCrmfF)2ZL<05b17yuMY1W)a_cIN zuyET<5z9nd`tjDQQGEicxHiWirjMs*iB+&9`aBP2m3jVzL+J_sk-?(rpKe3zW~mR9HJaH z8m?OVSnSQiPyfEd-e_=HOWb(d!eU}B#C4H8HS)+*3l<5M2bY2%U$uW%>=OTT6GjdX z*6DB*7A&*pvcty~|N9~P;Jw?fB`l0}eWkyB@pm&{8PbzsIQ|?N;kkBnrbpUs;Ix>G zI_KB%vCO)di0_F1Kr#kuiFIS6u-RtG$fFq0aLl1JMbN-s#JfFZ1ytD{eyv0sGdXDLgeWC?Z}m=%W3!%NZ<8tK_Fle2o#VFC=VI^ddURB5mH9wx<;8{tj2m+JvB z6f}&)MuZ&=1GKP?E>nM$_>$F#1pnSJi>!3xsMih~gRy~PfB<+crMc&omXOZxjG9Bu zjVPb-fK-YX2SFc;V$pb=3R@KmAuM_{!Vr^7&yA=+e^lJAESgjpJ*y({bHDfT?hVm< zf}J1feBu!d5h{IFofZ$bGxEAj!P!|C2>ekSlgiodd zy?ZWnD=y+|L2wB&&G+dp*^z*4Q+hn|X=Cu>phE$eGwzU=i}SrYv7Szp0xbzk;wmH4 zn3bWWzDR)h$<%zld0_m6ByyF1Q3_2}P~}1pH`wxbmhud*1IjELgb(~X1dXa~zKi{~Nmu&a^Lh-GxN8E8Q+k7& z6Y!cWep7U!s7>ADadx^%VR?nea_RX~44zBL1Ajc$L5>P*!y1sJRgmMglkc<@aG2Js z&x$?y=`1V8&3f&dIOPZS4YY1+y;wSXGri`DQo@XiJn5IOEM03A(gWd_l^p;a47lV; zn-l6JWf2+Ej0%HiIlX@R7Jh1cqUf=}X`mQz5{H450mpOJL2gL->8~~_W?Bu+ln+j} z5`!BXnid;FbChDh^3(xG_{@t^Q_S3mRQL(n>8C)ord7CD*$zw4+z42t#+O51T$yj) z#9a|s;I9wCMX5QVhExk`!ZP`dAd0(sQh6Yh4nsLq(pY?rwJnuGJtm5M^&#n!Txf&@ zair+WWald;8NzPCm`kxKW1182J(5>h-q(aJ4<91zhJmLi)@JY*LqixhJe|yY{RdyS z;_2Wo&wWM?xG|nuZSd4YF^v>CMrl6m@mLUD7(F*+S44~aC@7Uf12?p51DZk*7usY& zu?U&XvOM)Ng8-bN<9w_r!AiSckH7VJREDl)@&tYTNT_(7@~_~vAe&jf*X~I=@U~ET zIZOHZSJEX|bDr|GTOm3XGC=%Kk4JJ?Vmuzq=cmSDHlhhgh0MMXEP7a^uJ_~=K6X8< z`h(7Kh;f3|s5^i4|8{^CR_e6WePoRr!)lKWtTs^$C_(C>b`xlcmnr9fytO{uQYnQl zIbdavdf!wml|uNe;LQ|&wXoz;^ z8B^p0ygcI*hIX18UN--~z?A2`F`!-Et!$TD*?&Kkbt`e%YLQ)?>Y9lkY+4)y#b86Y z# zF1AP~p#~Uw9&3TNqt^o)UQxXUd+2p4>^4HLR*QZ{8+}BKeugOk4`HmOjmF~G!^~Oe zBv}$$Nn;}(C*Ol*BA!0X1v35J4V%|b4?FB5?2wsnSSAX$LgvH28TXP6+=?6A*vcHV z0n;vufem^YrNL-or4Se!npEv{6E?|Ox?&q)i8>vL1>NZqaA;7D3Ai+)?~^w98sB);@%qgI90m8l)8`$xN z6*|ew-uV&P#|=7eoa8)j1D)d(bA%$FP?|j7O+Ir)hj~ST8|a-8+vqPKLTDsC6VK3j z(N~q(gI$dp5?0b1ac$vL3Mg$YQT1=!AbUe z?01}hX^4#~&PS9DyztGCOxDLMfAy4OoYJi`%_{qj`rP z)=Ts#%AC%%;K+#3iK1@yl(GH(U+u)<5T5^7!($rb(`KgxNalod2$dVPpt(59N_%fgWbT6 zkpXN8wdd~pNBph0F-?!UNLINqZXg5Mubh7)#cZHJKpAMsaF-1GWjZ%Pi~FE8A@!j; z)kjLD>y1LK~vHyb7bM1A&3easas*4#qSqT(qekK3WEw z^P!*PgqBf@fAdAXeLJ+M7Gmh^sWM?S>> zMpmJxkr$TETNN~isx;HD7(6W@jRT_1;_}c2d2>RdsM{+`X;d5xtnfBMDc;At)AEtP zj6)c)cRd}O=3!mM2_w{zH}6r;J7?*FXfs(8)^?Xll4N@{58l)sRaKcQzefTI{in-e zRP?`T1YUDO#puc{lU9D!YB@Gd;@u%tV}O^W{{z}UF-IwKh|=tzv6;>d?NBx+pmO4d zm@aX~*j6Dvl=X^bV;h901CxL$sfzB7-vi0DCZFbnRlpL6%Q~J#b*XfT0WyL@6Jvn#AM#t_*cu9&uL+#W^uD=T@qlTOeZ zLPpT&a6_#EVD3|8NnDy(-FUadTEx`tof(wI)P$lw)`@_MKt1H_%wE{~AaYNUD zhR`Jt+w$n7zo*rhe5gJB9!YUyV^U~iOjc4%7Dbl!#n7tMwLa~19`is}1&qnbFQ50s zEHHMhtsbBDifd+-WlFeAHp>t$In2qFjIQ|Q+9P3BXq2^=g7gQstgsuK5>TY+w<_(V zm~9j(r8G^dGzKfFG3jrjACTqYI&~k_wiYA$t+1_5kHk}0X{$vZH>|(K+F6XnY4?Wh z4Lbm3xv;~cPs>`pkNN9V+ZDHI6kpmmwbxvS>qI1S!P$T&#pa;%AlJ1vNd3GM^_%Cj zNmZn(m{tTa_ua4V22xWk7QvpNyGVy+I?^iE}*A ztPFkL2O7eT*Z;8pz$0IL}-RHWIrj`TZ21RnNP}YNC@wrJ_O66 z&U8dIN4-rtM_ou?pS(?xLC%1}U?Dw6-7XxtFTL-pbMVXWyL`lca{BV!H(r+iBU56p zHOKVB%Rd-J9(i#3m$~siCdX!_Swu0Z6q!$H?!0q&M&JJ<4`(YrpM~lXgR$KQ#ZURe zvOueQ@t8TBN_5<|weAaaveyO3R1jqt5DmV902t8)R2K+hL<~42qF*WaWy@bYai|&|$spf813$ z!Dj)+X0tlr8@riy(%Vlpwh0vXFj<@h7*KleL>zlUPZKr<4*10E!d-YjlCpY)ieb-MACF~!5ucMd) z6se{(XJ%ZKeso3o5A>~$@{41NWA3RpsE_-dk31WeOB)&RxJcrJv(s}y_q0APRZz%p zirg48C%R_Rr{d;>blz>cB>IrT$W#eam8+B+)Sr;#F?h{!zxiWwLyoFS_{aS=@J>X2 zDsD>HD9VvRIUcWGp{GBOIh9ZwRwwCH>=;&q&JZ4me$U1;UJd|sM-SbQf7&9pA}Mv% zrC*b@fzmZ@j28%!^h?*Qp%{p&t)MhZJ-6_RW9sS8LUAK}Ku{fcLtV?O6we2Wwd=3o zvHH-H}+duWEc)94T3G+tGu-+QHx;| zeRxf1u4JEhvlwKORjw`@v6~+b3yR&Hu6?_v@&}urcc$TTF%V!O`OL=95*bFxaF>D< z(Pbe=RM?ic$R|;BKi0xmvxPhe#VAjSY-4BvKWP1O zvYDG_>&9jUxNG`(DwPxiCDFT}*-wNO&$E%859r8cnG#rd035R~MqU|TBdMTM#u}j+ zu-hv?V7;VOh=Sc(%PE6r7ulmKpu4?L!U0rpeM&;Lx}a>o%P(Q{c9&PPV)v_8!w-hG zhnjmyD{0Hf-O3!{@QsxV;ATeSa4|Ri?f?1RL@R*iz3ab-{cFfuVw}% zMr2CngGB7ING;YXBOg^W?>6InU(!im^>Dx0r|#$WWJ3`Z@qAD7>)&hs@ZUfC?X35t z^C@O7MH2rv`bTa4Dg4iWto^AKF~)bk^BvOS#zpjB*g*Lv#ayF^iPB)5v!(b5ayuV; z*Mb4DNKKjpZB@cLI&lgb5fhX)n_|oq1UChE6IzM+P-kBg&?4I!4kT=b08ActLHRSN zQ=%O~;lgg88)9=>sLKoSj!Pgi{n2REJ19ZAO=`nc5n)mF} zHsFCboF0SGvxFtznqfuTcWQsQmF#t6w4Jp<+YyTSgdzu#AQa1~F|lQUrXB6osC!ZE z^xsQj*Go;r;F%q^E8eIqfdHf~s6qBd9V@yl9;Yxv&vYAir_#wF|G6gMd|~cu9m=EA@zYr+l7V2i};aq8nLC@mCm%%^eXq z*<;0l_nhD~YSy1ZKD1Y6c&=D5fDE#qSEQ#H;B(6duZS+78->{2g&s|X>JB(La|9QI znpCD3;EZgB4OfdSOVA2__OeLSm&x|4<`s(!0kgLN*&p(G4PZeW`1T#WWLW5N^h1Vl z9XopFfA`Zha;xDv^sR4SAWO%Ptu|l&T8aUFwAGYmZp22>9o21xj^E?4$+I-(G~MIT z71apZPdy$hWp|h^2)EYDuTS1Jao5D=gac#?|1tzv?~X)y$Z(;x52iWDf zZ#ebbJ0Jb`kF6ll-4s10rv?fKyYVKg(+1HdiUCcAi3UKzEz8rON}=#iqZ}QOR$&US9Ml(3%?fyCi(%K*4Jt6z>LwN5jF+VI+QqH3QHjIN zMoFoC~BS5ZPXc>sF>vS;}j$?xjVy z3g_|8d)BL}Wd-zp{%QZMFJ+AbQf29ypmou^)th3pW#CEO@!B^rWzf{<&>o6sdpZtX zG020${!n~}w^D3xjJQM(-8ihwVpGxmJs{2JSB13+wMD!x0%X<9_j*AL2MZZkma+hd zV{+-mLD*b|nFiBHI9g9nup?l3`YzWWX{Nun;Z>{8QW5@_1!N<)kg@xEW59LZe{-~l zVs=qvJEhs@i>0J_W0y@#mz4RY%5T6s)~c@n2l|3BsLN0piS)E51s#DkVd!i$ zClpKY$aaz)mg?V>V0<+v{8U1Z$E}xgf;zma=%$3T(^oL35(?DKOp0g97!=2Io=tKD zJr4OBmW!MQtAEVPEDf=ur|$Ko6=Xd(^tkczQ)7dkZ4^^VkzcW4gT({O_;l^RSl+pdoNH@hipvYZH zvw3{Sqz3V_$Sl4=)uYh%PRZxv%g?1rf@?&3jjW#pd2ToWSxxK_!1G;X-_+I7ADAFI z(`xP!z%e$8F7W!s(SA8XtMHy@XDsx^c7njJAz%?7@5qb(;1bAoWGQct#Xt=%W)*P< zKuH1bJEF=NkA8Ay&VoXxuFIokyqHax5(at#sn!(l?ajk8|H0BlDBp;e4*bH@w$s9{j;-m3IBiTTX7eu~ka7u}Tjqri&u? zC`}iXP2kh7r8j?iY$3pu)s!>^T5Xj4IJ%>S_JYJ=+8m(=Lj|IJP+N_vo|U27B{}1d5j_Y4 z%)a8g0_Yo8K<8av?6SDo6%tdx0f-Hy^Y;0sjqm#f!N+p^vEbk$4f|5AsjcxVqql@w z(HEZk(GpVl!njgUgxoJnv4diu%WkVhg0nn!qjXPbxuQ^A9#ainy0GLm@v@XU-u6%f z1G;&1A+1F!yX3ImU=#+}Kx5t&8%8*`$P1e8)UG?H z!pw<=c?tE?$M!zByH*Pj7O#3$P?_&TQcjmnUN^Zm?E2*9gid~@x-h7VT$^Hu>=5h; zZB(_XTgiu8L-A0DphHj&Q9kSeETlUG4H#KgJn%nFE{0uw<&t!(ZwH+>7MKqw&!2Ld zIBy$fnKh>Y!HFQip|~x>EeX_4dn(nZbDwS;DrTWq#Y};@Nw+%iv0p3wv3IYBC16|) zMH0wWmK=tr#!lqM0v!&iJQh3Moq7 zaTUp)R7IK+@+WS7xhvN4&YAID_Q~Rchdx}4L{TU^9tPU;KN4N}yM4^XZGBXhV9?f3 zlzWF@UJ`AcnL9{SSxi`$D=vJER)KNFB0NSmLz_C(EuTnvs`ZakGuSHocu-wRM z>4vmgJ+V^Gcrh?;*>QIIdzyvcu==&5JAd^VspS@Sb7RwV#m3`3MKLERa-7nX`F43V zC)`tXLL{}7zNow$HAm1YTq>F)S}E&_+ACbdn1qFN=G5kdI@2T7u6QykJZ;XRe~N@xnU zjH^>E_u2+tRuK<_cKQi<(gvvG!Ow1hazUBG-p>NFVv(HRj=)c)20C3U2bmkM6r20M`tq7`T*KN{tKD z4Rj*pmGHO0+;-}?E^<{_Bf#T2Rgn~izN%cW&WpYfmoYZOXHK+sgZBvv=8vKrU&k$d{S{3kf{Wqx`2h`nNe4xcNh^x>aLsMKmS6d!%Hi7a;Gip)(mW+0bh zR#PM!8U`R3ws7*H$$dYY5-P3K1cu~o-SyYx%%wmcxf*g+Wu53Njsx&3k&A4FR zhMtK@619x|Ux0-ppcpA1vOn55fr84_$EA(3fQU54WXUBe1{dx8X>mn1vKs zKxsc$1IH*L^! zo??Kf>*fP)J>Q-O5jdU|h2W>Bw35->{M#KDnma-U$|M9c@V zmLYrzt7oFo*28t$yDxu(fnWXoO7s`@ZTD{LmasI^_t%5R`VweW#d4FfkkyltUp@j| z(4UYD<=siuJj@8D#jS$A=LW^x5CBhNM6op4?}F5*td0ZL#D~yi)+IJ7yFB+z)T!31 zDy7{=BFtVw}}FUHi@XzmwGGEYzyEu|e4s z1K3|iX^y;pQCcx=TW}ry+02eNKlu9vX`5m$X%9gSiqoWyzWdeY*U!v!Zd=#j6U7lP zgAeA!v+Q_jOPKvz`#wUqwJa>1f!$sgrHg|WzS0R5lN-mM0Isk{@g2Tx32kJvxG3yo zU^jnhB$R1~S4=Z1SE;tqhR|zpc0=%9agOXLh<>GrP6i&0_*|MRsSWFfVy7kJ8|0?I zt5Y|4l2%e0Hq9)%t_7m{C|7kn~ssW(d8pMYk>=7cX4 z8>QR4?|PjWn}n%fE&ExX~ z+D&f_pCX?7JUE63?&A*kj#c&N-?vvua@!S&MG>h(o+HD$YBNL2Qs9PKPj?EntwL-l z%?Zh+*Z5vgXF!2jQE;nwBj16I$#B0^^Zky^xXttJl+S(EAe z1Vy~|sE#R?O!94dE3XMkPVkQ_npHU>$z~bq_vQ%#ob`eoNzso3_t|$Fx~)lJ=`2LX z8x)*BNFd;*+zWG!`QLJL1{g_^M$VF$=gV;rIBR+x8;2bP1MGcAm+bn_diyd0w*>}E zVSsT;Z}1sj2dpcJR32TgI;1{Gs$WU=d#E@rDdg7+4?&$`5-ArWB~FK+V%i7+#c_c2 z*#yTXc(_+_0tj{Xjdxxh-xnNt^~}2@lUs1ajq6OxY*wQ66a)QTI!bdWt{s|aviPT= zo&*)f4W1VQQ>^MTj|J%h?S;U@ObW>Hr7>H5Fe}r^Ukr`TEpMV@>uPLq2i3^9Guqer zd+k|L+}4h;Frw5;v2_m>98i6#ETl3{yDS8oqA(L!0BxS_&W@iB{wQ;q>%m8I?paO- zWK}(VTEWQ{RHb{?mDgVn@?^zt8xAeG87;+LRqOD0$LU^Mx!qJ~OE4XU*SF zr~Xk%8v`E)Uz0Y{cZ7!-%fXK}&`5(|d6P(`F&H)9G`<`vyH~tiAF5O3s}H!czhZc> za2^coM`DM?0B$Um@+dZ z`BqO~ErIHN}#RK>nBHY%C2CI0Gy@&czNV;86dX@%uO*s|WKh)oad>Wn;hr=pV&eM=?;Z zx`xsm8xJw4paX)#L>JVp?D4n|+~JE#-`$YUuY-o2+z32tgxE!?@{rm%r8(hKvEFn4 zh*WkAaWb6y%?~gdx5;qZ;?&0O)c+P)QSxs2xWAA^+&0Q?>;V_sfPFQ^WMj#L=DfH% za0AF~>H*mEBDCd_P0$otF%8)%4!?HI7uYG<#krEMK%h+kc1b6?vIBtV$sz72e{!p9 zUwps*y8Y_&f>Z~v)G`%71!JSUCUg-f?ZT#7UBd66Z$_+&UK3Xcdc7vP3iLsg`UxP+ z3*s*Y>Q%s?lchZEvkE8%)0odf_k$9!wlLb#00O%75m};Cd9TNo@r^>(B!kUMY& zSf;FpoQb53tSZ0l|OC`R?~g4=@*UNHGm*3 z*oY$!n-w*!Gnq`1nPfJTY+k$h{j%AWY}QRO$?m^nGT9`^D1st_3JPii@)AT40Tq>( zFYpZxDvGEe26RM1pqqa?$$SD`Rkv)AHN6sW)gAl*G|an$hnz;9FI61%ER&p-o)gtddg#rb9iI0+ zo_cPY8PCgFI4)~tODI`_Y&_8V7u{Q}fhH(*y?=KP$>Rnk2cDv8&7f3HF|Zylp&~F# zrHdD3;8(&t`lU-ZOlfgh>E1fiti@%wiw^Ve^JX;rEnVYwOoMeG&9XIa zx>WxMG?IB?bqMIQ*lMgT$79@MZ`q-ZjyZYguWi}oHFL4a%FJE%NgJscE#T|G(^9jU zEjdFmCn-`3x@UBbZvt~Qphg80kgr$2UNOBfw1>vb{o`cyV#wd8pGw<=8Lkwyps zzh-ARGP3<|r?4^fK0UyX7d+(e42_p`%g2JlabwJKLK^LO47g***r~6txM5i_#Km)V z;8}~MzG#m}2Yrg!vS0^|`v|bal9z;b(&y=PLEM4?C@(rZ1DR{V>0wHpBj>itGbAZw zA3wtt%8caIf=cmDQ7=D@?wz_?u|tj2e^(Zs^#V3ZWuG!z@<5pa`I?Dgd49)Tb9~2w zpwS1{#<&{&OU7`vw`Tn>#jb^!b2Zim!=D9 z1nE+wrQPO%t&TZwfJoKtIR>&?*$E(yv>hH;hH9)ZVd>6hr|-e=!U z{=+K12@p>HPEll~1Cu3WnhkLh#jK~uI?$O5E1!@1%$RZ;C5rSi5T7G8F-|8o-CVY2 zG+sW^!>)uiz1gk}r}uo)%xAYgv}EgY*mV#Kjo49;?Qk!B?Z32+6Wd>`-@(M^hNR0@L~@mC+vyuG#Ti2){=>b!OIa~m^o}d z!Y}KP9RZ_wh?Fxws>IvmQE1l2R*`KE>`@#v^C)&v4D12sQV}TN(i_&U#I&Z%;zU?u ztn>Zc^Q2}e$_Cz@yPLPPTdw!b@kjOK)x4#-bndJo#YtJZ^U`LhB81&<)qpZpxb!MH zr0F450r8;OZii@^m3SF;H`so^Z1_1}=Y*F@#qo1fyi9oMSQvbrY#A+%=D@Q`xfzCb zQVi6JWm6Gp^r3)M874BN@?blLid>LtSbHW-(WOIneo5GUs6Gc*wON)+@AgT8EJ*xs zn{Q@FH~rKd=y_1-81+q&hsRFFX#5D`3=XSqU`LS0KegPQWM3`WKZLyMEC!beN`IyLCokS7=?`vK-uJrhnc~@`MPaT#?~~Nh{b3Ja8)vKY zA^&Rs2U5GD;`yL)$H((~?fHotKK}L0|5yJF)6Dbuop1hv9G*;Wm>Kx<6myOur;Rbk zB1M7#axrL7_{-;a!iQXb2E-zCMf_?*6-5^0d^F1*E1G48y>%!{1k^I5+8^fq4mwYD z*c*8B?D`h}W6~>J4N}G>u&duq zZ*kGp8YuR%G^a?mrisz3c7+~RH^^hC<)Wl1TwGD^W5?!ZT65m|(dIOEj_W-B%c8sC zmf_bILn;m&b!Q35=7gk>RN)ti1(@8A;GUik0l ze_%2;63^dEBK6!-m>hUJ@1B{>yg@N76uAOHd}YS8YZBwnB;|UN!aVfUtFB2JBu!fM zc~YjIq)&m?Mh7JSKXC1#vqSRu*CZ7|Sm_)^-|c>3DtxP7HT}@L#HU7` z!S4_@z&B=0A5v`b#lI?!*{Z(E+cfi#f7NuWU~E13tg(>RwXp`i3zPr)v5yI2gUqQ| zqUQ!N$2F?tkQu}ZDF#TNc2E(R>(nf}0@Q9mQHq35D|js~LkeWK%@^Hw&f}#EQe_(@ zdd*St*gqRQYF&9~oA3<8a`0fUVmo6jQUcmi)DN`R-jYoKV+{g0w()tWc6rkTn5X=V zpOH##fN@}(bkz(nXDQ|sMe3-CV*z*h2c01+iB3?SpI;zYIr*$Q#{Fcia1FX$K=VO8%|5AFdq6buPBSi_Rht0nOAXJ9-)YSm*NDrrEF;BSGM*e5P%LrfY)zMZ2r`%I!t;1(nq+q0OTS@k4W_*@^CM4_0s7gwNf$`!WU}Av&CRD6U>?q*A`;)&?^EHDO?~`s)x51C z7@zM|>{e%k+ABudPEJMQ@!NBnWv9KjxU>n4skVs}bM3$QtCotofE*gVV4NiN34f!ACtTbWo(ox*^QFJ@E1w$t<)x?c?J z>vhs&0xnB_Ye4d{Hr9jtWj)#UdiZDAa?8yomi?Q$_pXC*{e6#8MZfg2xFl>ni2(uW zCM_-%@Z7cVOmYuIr4P@z8Eibed*N!5!|&s@(OZ0b1Tnl~#qNc8m;H>0^^bYf>D#u? zc0_NSY>4Adqv zxUw)CR@hBUq1ezGX<^!Vhk*h!NokBYqSjqJZ|O-*t-oFsDXR5vls3y7gt~5*O5PcA z9d4^vl!%9%?s)VnDt#MZg;mH;6}ANF866VJ)XTf+LBXJ)V9_o9cDlsW+(34OQcubWA3b6amY@B)368CE~1 zm_Ca1QW2FPLjwu?iv7+2Oe_Ye1c&XJUp7=*BWf|1FW>LD&m{mTjx~m0bd|ueicmX6 zV*PZjkny)GG(liMcAF3tXet=Io@2b;Pu4QH+&5Lk-IF6tQJLXnEToU0fHXl}9#9<} zR_j$QK@ULe1dsG8?0mR8LoMA^(hkYyS7#T~1>~N*mE`d%7$~vBzc|;wo#`V{ZUsEp zsV(5)wbZF+Nu)~*)vFwE!{cr42vlh8utRIVm%&Gd@t z5Z#-eEot#QM?cWO1ReURa?{tEw3{T2o|URgp*v_iE~JNMAc@G3qRg+BUNw2SXtjTB zSj=SmP2XRQH2{@TDgEs4%hg6TP7E*OtOq5u;CS1mR)2&?#s1MsACSERJ6&1g_Ybk} zn5M7aUP%2uIpV<1Q>$4R;S-9fr^p#9;@-ERrc8r1M89tUvdlQae?LECl>0tP%KV~4 z#o@#4jAt?LJdVlJq)s*7ww@G(wJIY;$*!f+L-Nus zu-|YmP1;QUt$7WLZ!B8&7vt7Ok`i0DDsREWG2DuqVX_h@qb?LUp5%lmZfk{q8Z3VH zsweaytA~-z7xPR%(%<+bvkc-;B>lAXZ7PnPC@e{x8R7|IzMRq+yJLu;0` zD@!N5@Xcy_!wCZ8Z~Q^`m0+!DUg`*$mP!tCTUI%4xPn|Zo5fC1OdUmzQ4zbq4UVF2 zD*yIK0p75n#WN_B2A5&^N|ZmpgLlM`*mL= zj=}u2cIBVPJho$f_SP<+#@zo{Z~SZOmLE+wnV1#7_lhHFUl|iqVP;~ACVSH?lp8Nc7wRC9-;~|N3<($4{g0%*CH87F^<0YpQ|EDxDW8 zsuXuC6T|Z*IUzlCq3iY8?TTdB`K}b(2Te9GBVz@{m)^H!_fLQHs?!dw6m0Z0#^P=& zfzb{%1=d3FZRyfJD1X`p6G|?fykKZ1h^iU)TY30n zv-}(*G;A>T!?Oj;?=Kx<){C6H+DX6ugHw=Y#weEw$AR}XS#n8{m#j0mL5{okb<#Xg zq(CM`td7D()EL3QIvOv#U^R+f){%{`ulUQ`r@m>LbB_P})lbRM$)wH9ST|Bk14YhK z5%}nml-oR7TyQOk(RID*vAj@>RAFn}ZYm)4yiwJz!V7ttC72}FLz(OCOMeeLP_iRlY-Qd2KNhJ9$uvOHqG;AVa+Fh>8X;tdfBIvqc(K^1a zSGktKouVje+20@#ZfudIIXM-`eh6@Zd#;|h%H6t=wR${Q-q|snwrb?Kz10qzM&ac> zly1p5&&Aw0uzy=*_I6}YjGiJ}sE9@t=zJR9iNWAU!LXAwNIGOdO$A}QqYB)H(v^{3 zWsm%Tf2U#NDU+&Dp+T|^u2x}(rO_Cn&eEIhPH)JY^R1_*YQNL7(jJk|og5a=WRuyb z$5Ko*MLwn?V&2L6VEs2u4u}1Ra+HT}&FGGDo83=wGi3kq*7sa3?Zt})Ew7$cs25g> zBSrfm4+6=0_bIo^btg3`&TEwWfUFsHag4mqwLoHrM9`?0H8DoP+C{L%JZE>iM8qLF2VSzjnsfAQ}&Dz5oz@urY`@;l^=8 z;n$+y6Irg39X5Z3W!+mXtOW{@=#UN>QiMQWFPgdmx15k1x{BD15`+y;)&rKU2m74o z1lUPgksn3+n(Rrzvmfsz>8}{sOr@Fe+f6aMD3VV_7!x{oO+5kyBIS?`SGedCNt$mJ zWKCwOn^d)=IOLfVF1nMHMxlyEShDCauQQ9p%8v>%&Al&=asj z*0}XBGP81)2R6qNJB(6S#HgKyA@liv{+UEh2Km&HxwZ)uvz8*OsR%4>MsLIr_J|Ut zNxJC@Z(GkC*8J10ddk*xWaBpMPkEHu636vQzOL_!wp5dTagpP=vm;?qn2v&dmvz1s z;U)B=;3y_XolCcpDkynqoK`Yr$mtZX4(6d&vWI6pQ{py3(6$QAW){N6ux06v9nHdj z{^K3X%Eng;AT0HZcTAZC%PHLMPj`LryoJ@(P!hphoReIciHM3fg%c z3cYW=Y~|!Omzy5(yj|pUcz1C0WMH-mKQZHi)?V=8ju$iV*xaWch7UVlnB=a9mZEA7 z>#nlMr0wwQ7Vnv3{Mo3j6h9`nwEy*A7qsY3Q#5d58klX*bhHz z({Juju;{=}hwM-oWg?_fQ*Qrhj>!f2eTBA$6gY4yEvSf$;9);RF;HK(ci2Nkk%t0a zmo#09d;AZe6#9W6Q;G%AUGj~-)ru8N8h!V5lzh;ks#*aro38PxnvM;J6v!W=E9R}9 z*G5;kWb*e^+%=Iw4sh(_OZWh@L}y>a=?+bUD~8w zM(W~!9FOLV10#b)b7pUNue5M#gL`avE6JVJEc;x6xzz>yxbQ}8DsMSeNT;chr!|+Z zSDO6_q{r!Q+W6FXx9oCl)LW>D}?DzuPP&j;zs!Z zY!l%%+u>xyz}f;bBf7KM#{+H!VV3px4(sT$5CdElX9}tp-f}%G+$lY%-Xv*ex;<}t zB!)F=lc=<>*MD$o(FGvf927Jfc*`z8wzXcB>{aL4DY*dLK!?@E%JpBn<(k8<7usd$ z4P!OV@weG{TXw|BaiRXWar?LZO@43fo9oL-)>kHv_OO|kTuL#Zv|dO>tekxMozfpg zPi_-_2HRP=yj6m3B`&7(cxQf|KrSmUya}bqQ$WZHwy7||sz#P5XcMN8?Xnu#h51LN zjoR};&7dmW1&gn&#n<2bL&Dn^v`9RK`d7Aw+c-h?d}gvl$vBTY)K_*{YN0u-k%7(2 z5zD14iUD>PJua7$lm|73)tLMUs~#*|H&&}h`FD6@rTVVWrBR|Dp^>&BQq(O*Rw(dV zf>88mIC0k2BW@a)@elksN0{YyPS$0T=lrX&zcE>tH+LQLB6l2kRUKo-m@q&wk16ud z$TXZUN|ga?JQ5BgeN`oF6(S9=u8;JH*9Z0ibr@#M4zH{AsxyJzunI4yi-GN6^*qSb zY>+1@VQsC+gPLaC!oWOGW1__~Cp0oh;|l7qX`l@oLr@(Six4rBf1Af}F1Jxl*dR9B z8H%;h&zv&dig`CZnxI=D+qd5DTD0d|oByTr+d1#Wz8Upa*5a0*-d#NS z&d2|JZ_(4YcK_SMH|?-FZ8ltEM~PWCHb3BmlGheZ{n4ZuCY1F3Odd_Pe`Q?gYBNmj zp%_T$E}$Z=&8}5kon`!KWMinM+eqZJ+p`~&)nrS^py%y57v^8WDbY92qg&pj1&-gq zvf0(7hAx=a4pmF*eDPcx=@FDlx6`Z1AXuSW_9!$NJH$pBAe$p>($T)!*IxDQC=|3u znik9yZKLs1z9^A&i}QfQFH2LyOX9~Y(5oP`YTAl135<{FsdF9hF@4DxZsKNB{Ju5! zeLs^;`LO8dI+FR7u_-lXHl>7Oc2i^*RM5L$6J2sIZ#;18c=v*;2(v+Bee%_@`KE?gVft{B6I$x1!G0@A8@NTwIOezI+CJ|A- z+r48KTnGBa>$7{qa>H;DW)vjIm=5ACLCXah{76w9@3;cPg|@~4+n{mA2U|xrKA(5# zN4=JzVU`H2G23yE|7ploj1(OUxaPNXK)N<$z;$VbcY>gn-t7St{b#*0_$@)%f~Aq7 zPEFPHogpP*$o8iz;GdWWgqT6Q`B!<3?un8uE_U%GIX`Bzch7WS<426$aWeIeTc5YvPcnTKOc8S;sU!EhA;aUTm3M< ztP?w2<76}(H*c4U_geR=Zw&15NQ75CTAM8B3S2iY4;V`!SULkGiuWyu)($!K3b&|+ zoK|{9Yt#JW<|GPQT$;4mzI*1~4NM8JyNh8pek~?DJ3!W~OHTNuJ`IiAIi7qe$9(}n zmw2S7x`|>oQY4XzKyrqh5RB9UDe*o34mu0ApVB3bswPh$zl6DF$YltyDxFzf8LRYsruXGAKR8)IyTTv4EtI z4$uL*CjAt&#>(8UEPN`hrOOw0`!{McNz8k91B)O-!0zwpSm_<@{&2aurjDC>Hgx#y z{g$%~7uVEroo>RiS-n!wNmmMy(&fI;C{c>UGHLWxl1txm(dn)RY?|3Bz6AuQ@xX?_yXr?B*uGcx301=j5GE`lauw_k<=N zWbvgJ<_0CJy0l574|*AQOWp$(dVvOcx424NDE{j%7u{hsemV<+J{`0!ch+ZO z+%KC@%gp70+T=-D*$Jn2({ZQd_GzE})Djn5oDIkIwh0@&6^(B+do&n+f}kc8AztD= zQuHe86u>xZCw5tOOj!+Y_Acyl_7^xW z{oRXE2gsU%AN58%pJcadm$IGydg(6Wv$b`;t!vp&2AtEDuWq@TF#w%qcSc)yWm6v5ZrRN|M>@5GF*CDSC zx`4v}+FaUPbeJbp$wyhIjlOyA{j&Qc&(GMb(W^#~0y)sw6e)Dtgx%tLXWdbf60qCl zC@E3&%kYwo!DS6>Mq9hQcy1Vow)xF&NCu);B#%db#esc}?PexsBgG_AB%X>`O)5gK zYVMF8kMxjR+9hf9WpR>n!1bYX3s4c-7!NNUJ*xrp(w=PF;QeJAzY}DdUt*Jbc9UJ) z)~^mc^PDn+MFqtipvXR~;JKTlqg-IL%Peu>Q#bjdZl2cV8E@LZV759 z-7v%Dgd7gAjjLldeqOXN>}PBMrT34ndHq$FO2Be1l)s3ctJidkW5X{3$x@t4g5apA z5D1u70KpQH$QHU3x_mCgEu%ZG7v@9Whk=4R$8d*zt;smX-UcJg-Ua()Gm48!ZCb|v zTCmB|WcX$!lR^iskT`8-J1Qv#D46zB5lA==9OSj|E0_Klif9tODngG1Gz#{~454&P zg1hFkSCUJg)aYWBxx5>Y!BQ!{1?YzS1#%BK1Wbmxzr<_ zxw5cZuG`^uo;K#wAtCksENbI<5}V$9gxk0 zx^S&Qd;C|(YeRL<)k<~wqQPMM8WPJ_nFB5?Z?)?&CtSRCZo!p*RGAEm_SDZ0k&Mw; zL><_Ftuiw<#S~LWk)04|r#m$X?wc3?;iKK^B;`dCEBQFEm!HnGzhR^|0qr+_2_56S z%D=+<60Zm4Zy|BHghqK&W4r4Cw`P~n+ac8%``WyCw>o<6MPhr5alnyX?TIBQ*>QAY zb;g@hOql8VRm)#U>|~N_HVi2g1IA?o6=84$VG9ij&Nuo#a9!3<&bZ-3unbLfgRm0# z_BMSj#y)w=mj5EdG1tdEYA1F$$^N)U@dq4t;ZV8O>-E2{f|^H)qn=2e)TG_`9%{5> zf!-s~knB)wb3a8_3GRDT@)F!*{Xmah)5~8q8Lwya_b8*t13GzD%;YBR>e=Tc`({^b zjKzJ%^XEa3F=+S-Y|Mg+qyxR=?_wYdQ}raodD7 zf;d649}3o{OS}EI`Cd}j2W*``frD%q9jiy$=41Kg9deohDYst^{gc{cGG?XyEs^9n z@QV1TnGx7aF(niv+eRD=uL{^Nt#Ij(9Q4zxdPqADm`Qeu9x5)%I)p1{o|NgzG@ata zIh$riLF&=r8Bps5<-hunc;1jxvkV*PRe0^06IS7!Ri9VynZMt;WOg1O&)Ph4Z#8hf z%uh94`*J97yUGbTuldh8v*~vx;Kann&nMm7<{HQKy~r9fW}QKb8KB5xDxwT@hH6cb zks4mS4j(6Fs7!F{FQ;qL~42jD>NYdsDbqxyc!4I&}H&FAX&Fv>9FKF-cAYUer>|v z?fL6wlG_e!xK^1V>T`V|1P%-EUv=zdW8oX9b_Xx{8FK9tqS<$JvA#IlLk)0u@RZ!CdtyfKp47@H(b{!NX z%X&%ooF?rPQbP9#Iu%dIpl6fzpt>5El931(d#Z356r}q1(K-B6uIK5*@P5)Y7u(;4 z`))`FJrAmD=T*35XpVv~Qk}3z5aZmW-Q)3zN10zOou@%n=ESLu^2Di7TQnDs3xV{# zNgJzd_qs2}N`{g}Is78Y$cfpOw$BRl_~lJTm`6^wkJ_9%@8}{E_P%@Uhr7ukZnn>X z{p>4dQ|xJq0oL5(D1kW`T+BmX+NibLDsBj>@Jf*1Kw?J6)vmEN%|z%p4~1`b&VOExQJ`#bwjl(-dGZLSlu8; zZunK2rZ=!LRI$Pqm?<|STa)Kkw4j+8b2Qri;2VJ`4j&cwc5WXJ=duG&r7WDw$Yt0d zM<}KGKY&RR2;Q=RJ1$ANTDvC@$PIDSZ+euG4Nz(ZRPGwMeo%c@+O8Z_UYM_Y7UsJd{LVap_3MPl_V+STY%MVRkE8ZZwQhI@zaf)U3#;i`uqNlgJ<0Iix8S6A8h247yNt&u69(5&;Wm{xtSs(d z|H4~MndZ1bx52aE*1l2^RH(QdS%n!-F>5FiLq!-9{ys6Vd7QuOmMOSy3FJ49N6h$sWQPOQZ039|_%tmx!h|K*)(nGnjQn!tg8}WESduSl1k0zT3bzXHxaQJO`G|FqD?K;N*CBu= zJ^`HrE;x+p_?rg2^_!nq7NIz7nhi^BbGD`g>Jls5a|nv{PVhW2atN#$NC5m zF^Z%4TF>jha<$AIcUVZUWP?9)!FcQ4z#TBekLZkuctw3P18?#bAs3XynNtW5`nM^SoD3U-$^ejBchsc+zUWn1M>&jE~+S%#6YS}u^ zYC1}EZlWuPt$+S6e`dC9!R`4wY~1UYzQ6jVCDS>JzAf7A<38~M{J(+Hq%#-=O6%pd zkaZR>0e*3xYaTH!&z$nU-}dfAy`$E{!(_Q_iGeTaixWH?Hym*0?wl)@B%&`?zB=%Z zOq!XIOQe{1imbt~Xq3oUA`g5Ytz@+#(=AP6n~jq%0mW#StOM|9ciQ(HJAi-t@D~p) zNt#|MOt26)VcrQQnHZIybd9_u$PNKkQ80TL!O9oGwp==_UYcds{O)<`kXr%Ra=o@F zVRd6?b7vh$p8KZHKmGp&lDM}&)XGc_hxmc`w@x62NPR|K`wt3p!!qsU`&3ZkYc@w6%gui+N(R#gJY{W zuFuYp3n!&pEdi z=svQf6lLM!EaGp}+++@_cfn@o0lJxa=sE}%x_({}lzm3KW=i)4HfeW}gX-nfHjmx^ zw@$HT>VWbd&=b@MGNpEW=Em1&tWVzNFP}w+>`#7_;H2Z)X4i{XD*x4FQsyr2m`z%^ z%{mV3$UZSMHr*6+ha#O=9*w-SY4o80U5RI#3S1P-az%gTzH_>ue12kB8f?6L;&B9G zCy8P8^a(+f+p;8dS<*qS8-191d9`4}lz5)O;W4Jf_A2(yPlDvzCopkgA@w@nYtlXb zn|X%5dljiX%;PNLV_t2UdVipUwpej{GSvA#CXjg~Dx zJ%4u|i`{luX_m@>@nFCWy+_ywYviF=>8{XzAaHC{>2NdjGjU&7lqgMt`CB(V8sr0R zH^A3Sli;JU`@&@G&-eMWu*WA(X3241-jg3q{;|SjmeQvE`a_a5nLsVa2rn|5Vqg;` z4ZKLtVr73=yW)~ML-PBNXO+8VxTZ`0d*oT8q+gOUC7@n7+yI4@2BpcarFNdd#*P!p z9`N|0Pfj>--03SRr#j41A&f;@>xM@qy~`yF+{@*nI=WiXCR@ww@iJrVda zMoxW0z@^yZthA4}(N~A1YhChorQUbXEUXI3o%Jj@9ysaH=LjpGkE-36asM!^GU#|{ zlLGSC(!(}{FFgw1gd%8Fy77rI_+771U!9Sndq%0s1H%GD%<{uOtiy+i>_u+AU6+ zqWhR^r*okct%_s=uhb)HC#^%KiMxU>UhN|L+Rbqw$9$-5e8i5N={qjGVM#sqB1#;0 zf+b~UMk$wKU=uV02uBuQm;U*t@;F&e-2_U80e-xoKX}lu9E!whB{@(qcAKo4{Oqm! zU*GrcsYNl9(FLn;Sv5KD`;FhY`U~T+U7<1#q}bK*Vn@mkFKCvIC)Lq`F~pMUh!`pb znbQgS0e+f zet2E_abU^2*QNRVRN4ONhZiA}%fas=lS>qHg;_N%Rn{M@*Q}a+FSyhr%e6~dLSLF+ zz#rgWV34~dpI;>?r@NK*Lxt7jKe{(!Hwds>jL`y(UR!@neKohA^F`pqHmcfHsJaP$ zZ!V2)t^M%8HlSAGgza{_ooB~MbE?z9#1 ziGRl@S@)9L?OkT=nEfRmWioKwB&q+DM1PxSvPd7E=&K_ikG7HRxPL#afJVlmH&9Fh zMb=UgMT!+6nS9+&x|r6*ODe@TX)N1mSJp|72__oF)j# zMReqYAp4--Ro)%fy^8G&Of9EX8+~#5YH=y^F;MTKRkR|kW}D%%8aBgjux-BfrM=wx zVEy~An@on^Zz`6M%?=#N-D_qZc2EqY?r*0e_Ij=Y&ZXVqCwQlrVt*a7oM6~BR-5UD zgcgx=8|3Hdbiqd8-cAh5^Sd5_@-wi-NkcVyWzaD%jL+J2XOr{b4}-w)`=w(z<_7FG zW0Yxc(v#jvnU=h{4!dGzA@nT|-KkEKKx%*2s1A&-3&Oza^hDy<|z9 zI(|V`$O>R0PEu~CYhZm?N8c7VY4KvO;sF$Pw+VX{LyBCw6wby&R#9b8t)h^(+D|V^ zhrCzZn_J7f>V=>0$#Z9Q$PxtUJiV%0+H8|buCWfv@ZgO4d$gy_@f`c0jAG>-H{5;u z_-^mN*+~WoDyaO7&280tK*pdheLk)6W8=V=_1*qPrXpuJYEBT2)Hi-)E z70mjOBxSKji;E$1u0f8@D9#hZnaxKM6M8kq-sXnD4vI zqfMBt>7)|``R=K*;V-t%DY6UhM|~MaoV4N{;Dq~0{K?;U z+k4$TW>i^$jPW#KTHfAoS*Pf*UNB1mB^ni^QNneyks?F)oHIvPqCg$R-mqei9&!lS zY3)7IrqSR&;Nuxa>+j`cMqU$sT2W%e_P{GmQhsZ&`|5O-t=^<&a{pC$eIrl;VULmXSEJ4lN_^n3&JM-o8A{ za(ToB6Ao-pSXOyy^hwZwM;G|^JFtEZ&-Fhh+b%P(6UW1%tWucke_ou)$Bno(AuHt1 zoF2I7GDGrs8|QTZD?~1@Zw|-?kQ-3Rn58)YobTPCn5|i-ZVhe|#_(FC1!3{LI)!n& z@31-x>{5wM7$dAh&`29&#b+CU?0*aDHoxkDT&zXuhHTp+=XJ_%=kx;(0 z8x$RkPAz)6s0MLm;Y3J3j}am)18cj}+#vGa>?N(1v3@S|kpr)WS#}0?$>Ib{mgMn@ z_-(>N!YFcGdO`jfXf^0ncLP5W^v%hbek@>9_ykqT8-|@Dl57W+m-N(n;lO1#Qy}i& zTF4TT=)i%vBC|kT2E{;%%oZvFs~u6o5jo#7VBSCh@DwIbig6B%g~gLDD7?URa_oX3 z#VS!1Nq6p-*>Pt1;+bLdp%(qs* zx$7qv-fPl6Rv+O%_~ziEPOW*%T6&#liTDWr9Eb|x!lj2Ch0^$(!1TZY{yl#b&_t2n zI;aRb?sG|2OJC)c`sGL?ZROsM6)fzl8QZ{!9W0}`vXsw}AL}g(RUOuqWvQUWv;)2B zQ+49hDA6%R7j05VfFy+)-u-E(2y*%8RolXILU!@HLnkV1G6sCu1^_SVg%ds|OO{WI=Fsp@S#9uGgfu)k993We*tSxyM#}K+fkg&vjlK=S9!WlpgfkFMTLWQ`hq^ zi%S&SWMyO{7@Z^Z8L!59*n6jKjj@V* z42WVkHlts833f)m%n4CcY3Z9ET9QR^Ss6KSq>hC~3PW&7{0i05<320oSdZLIubjC7 zwg)Rzr9L{0+2~bSp#IdWK=q6M~(k7BwyU{wgb@QRhQD#h8bf8VbBjJ_KTvXSsA*8+h~K^T$` zqcRFsO|%J-JZF7am;5IE7|PW#g!=dQn7O^c4$FZ1tCI+&aS{)u@N-jLW0 zl|*1~%7GuVT{+-dN#BMY$a4V|u3JR9jq@vmP}LQe`ADF7E?~D9gM&Ds>hgG)2a`Js zw;3v2cg@VD4@~QKEuVkK6}0ZIk~o(H!MT6|Wr-pZ68!L_*_$;^n=|_SHO8|eO%YmD4r9T!R+9{%9Y_qrAxv@?go$T6_8 z&S`kG&2F{p0Vi-wnpN65$x`~`l?vjp$bp>UFNvIcE&QUmOIj%ed*fLg-lV-6x|#n# z`P7b;(4&v6)plg`+wFFi9V#*N*EGHAYi_J3#TAk+-MNdL@VhM59SbNh#Cv;{hvp(1 z(8`%B0(A{jk4mxG1TDI)#=@}WfZuilz=)37u`o(Z=C$7*=>EvkY;aj2IdFapi$O7L zMt0D>z-Ecn$H+jLFDjsn^|oeS!-$7!ua8!P$XtnwODaz{yjN*q}iG+qPPVsytXVrDclZWgoaj)LfXXH2`FDk`Wx7K1 zq$eS(2Agt4CO34An3Onwkdp%5ZY*H^W#y#XurM78QXL zPP@{uj^60oeY0aI8{? zL>blgBljIYjJUzW+Sb4SX5ste0gD4~U$cOv#4}yB-y46H%s}qATDoL*yCUB&nco|B zT$Ijqiu+xUNqT0$J|yfkMv6*i;2z{-QcG`e-V#zW<1j=~R><{kcJyh+2Nr9Go_o8^ zEo*lNnk-A#`*-({JZ_f7fx%L1W?9N9W-kRbLlLzQ=-(sh3riI2_BhAC0d<$vbUy!K z*r(D`={eDTi0$o^KK1OTOQlQdVQO1aCCK656l3OXf}n=iDMU`7Hn#>)l<1Mixpcz@ z_R3kE!uuq~IX1kPe-U;JR=H&Rp48Y5Oyk4O3zHUikg(iwdFtlxS+0n_WGMk;pv^J^ z9TysRTubPVJ9o&^of|^A&fqM6kd>#oJWngTv`I7e`G2(asKW;QKae}`rc5Sf&~5N6 z8FF&izm;q?GZOI>vxb5_y@)z#AGuGXQwz!oSW7ms2pxu#HAo$9JCVP+A=URkd>&gm zBn}&mXK^vA0tWrJ@fznDf7e0=ZN2cZ*Lu>RIwfwWt4R*MT+=uYw=<)rpi+5@s9AYld{P0LY#X6iX=a_eN&9i&CY~|fppX8TwDL-TO$NCVT7q&y zwn8Q003VZc?O|xRVrImg^!%gcfZO0W{{L#HD@~Z$lo)WCB!6WvQ)-5p9Et%_^>iwt zOp)Ss{`K=TPBmT9v!rfXy7azBB}8b~iylExq)nLYw^=si)J7twrTKRVQ$akeUHNpu zHt!*)KIvJTRGNOtz>OU!U(%@!F1_i0FUXQa#bF&-7U3%td_gHYy=q9Y-#JSIt8dhJ zxhMZ~CEYvYkUNm{G|NyX2ARQ@CGk%&C1FPexk!1O;a24l51MrLunnfs8T-iFcaq`c zyH0xVY-+o|$)MD}xxSoays1ZH5+7ZK3#t$HD);!GR_*mX=Xuiu7SVY;BtBayffaR=DqfOH zcZ4EsX)&*vDX?h{?f9Vde8Mq&vVrj!?t&dkzg_kdZ_7BE!6}bfnn@cxrkIBmc|b*^hqQ>-h5*R~G6nDTjB`m3(Q7u& z%c7fuFbNoF$AKPCQ9eJH#*KB{+5ob>2HefpAbqf*4W^LYCzUiF0N0-AzOX@#v^e8|SE{)FiXj0)eZPh&ZI?OBXk>lZ&g}NKz7v>{dwISsH0PemK-UO<*;Cuud z{2vLSK_5^N;azXv5dF33K?%dF^wv={Of% zD@e{&O|O9p_jI1asv2;5^Bwoa?e9#yWLf#)urn1)NlLrVpx=3Vm7qym5p+kGgCpcO8M=Oy#;ll|s3E|$l_ zW3j6-;sj$VAipH_`zAZ3T;u;PImc}RbKrPgx7h@Cono$1Ed$EjrTp^Eh$DFEL{3Zh0c2_Pf{p*1SqS{#>?f8Q5m$BqeTe z$GLaV8Ki_%cq1LWUWMJjbxEG4M4T?UJ?B{{WI=@CAF@mKXb!IOupKL--8k_$3Pib{ zdT|!SGjHB@x%=+$OgQ!F&q?i924Nj$2)jfvjTC91B3elvuV!J*!c@bg1pf=A%_EWUGh9ITiKd=AVH1i9aeY43_QTc*)&0v?e@&6F{l|R4U?3x^irPG z=zSZ58iFtYgy(ew-feEVnq19*YY)t+nQl;&$LQLG2()bIK@WJzqN9Vl1Gh6nPWTg# zbctRSq21uPJ)on=fa~gcT*A>4jI7aNG|E|Za`t{#6SB&`@$PxD#(~p%^UScag<>{S z@LVIX?hIUL&_pr_pry>v`=)qq^gS=`CorgqVOYtgE0^9?rn{ufu-{~6_H? zJK!|4z4rJ+XCKReAeT)_$Mx{oSe28M`JxVZ$CH%lA(*~{CZQfU3io7ooVb@VC>U$4e2O%7Q8QM7J zTfL5xgb7zIGYpe8gKX#V9JPB+Q_7@$Q$N`EzRBU(|5Lx8kyG4USqGj{ZkxF$%@or_ zkqcDBDY7fHS+>T_z_rm&(j;lHIbc+(OrvukZy&`Z^s39sLNT%k4Tl2_C5>2Hj|(Cl zUd?drRc6zzP$i!)K}Ot_q>D7l4+r4q>hLb2JEZ9$D?P7CtLZdAmX8tnZr8_2!Y@pOTXf?9O(X;pGa&T%t%L716FNr*{VC1YRZG!8NiRdIw!EzAZWf zj0t+xdGDjrq6H_F`(#fc!@p>OF|}byq^Kh3wr9%g=}Z?{f|UV1^j~f&4=;M8Fm^Gd zIID?!D+kI1j5qao6p^RDyyLM;3WR5>L(&V{Chd}#fV|n|)I(W@Bx_;2!qib4FbP3d z$IQAq_a=6wS7B$?poMcb%}b$RMFVqSjDEz)bWwg=-o7}^1g`Jy{P+JPTPKqPW@A`D zF|bC-0i~EJEiU+u0qgZ`dJjm{UGzGl(cO1m>1mLy%cYNaUYVua?K2dx(z9NU>((#c zbbN_>1BL?d!~sPt1efil=xmdhc=QK5uCts*#c}6@<8%I{!gBHCu+9%lHU-i!8?wV! z@H&J^O4Ogf3b{181m-F)fjOu;r^%G|`LCU+%Y_wEyAtomQ?qQ9=&sVb$z;r6 z84rNR{H0dh$O*uc_U`>w%?Bp4l1>f0N6t_o>$|8nNDy{ebjSMkjlGHKXi>%W@wHI zK9_ZP*F#Opfa{Rc5kAzWNSd^L;&kWZpoe!_+$Y`>f;n{`^CH7G^YeYr!8YY%;MhnE zZo5xT4llZ2T({+GCL+&lfeShMd;<9u?@?A-sP{jr!8Y zR4%KWSQ`$NwHTVPwDRG9+w46VA1*eTin+@^X(JVr2~0gBl331A%t?yWQV~^C^d8xo zUL{1mWalt_a-X^f77K7H3`yg0(L8a}|!YJDsUXiiJN+cAF*}6xxLH|TX*=o`DcqQwMw{H83zswvnZQnko5u(1oZ|~8sj<* z!tK6|kRXJKqq<%Ng1wk`+!>e;OL)b+)${Q8t_i3N9xKwV!|T{vwb8c%EjbKXr~-?o zHV8rNTe!=|U`e7xhS1pRDQPrPwG7L}U@VKH6&U+p&-JN}{hPSm$;o~=?pyYYrISAV zhG{-}{LVLjK@LwQH_Yav^AvNABB!Z{hcoVb=pjvS)wCWuQyS01bxxDEUD>XRRUQUP zjE6I7cy)9Go#|E>+Nm|Ru5ihe*6<4Wn`ZWjp8(TAn-HYRf{)48yJge&osTRkq|?;t z9uGYuMX0no+(U&6e!A~bM`M4b3OzqDtciI-4$a@|k_j}vSqqc-dN;h*sM@X_0~EP} zYLo?z6{olX3R`mmmE*35e)Svw>yu2N(q!)YBN=q!^e=PV*9Ek3M*3=NDP}cARzXM^ z_NviD)+|JEKCV6zcc6@Z$ID?o`hE63yYav4d@Z?Q95ztL!uIlCpF%{w?~PK$4*zVr z%%d@AHHqgnX%DK;O7nOr%qsuZphv-LnU(H?!FVN)_ZfYLWG!6F#Pi}M=>Q7kBgbUW zHX$ZkWz*@rED|rlw#P&J!;l`vF01CYs1MDP&4_j%K}Wc zWKVZ-6v=hq*vc_8>#&bvpnS5JinujJR~ghbcPCv8=`7U&#{xR(;%T}{bt==YjAN>% z>u_7LS%w=E$ixj5bVzKE62fti9kWK6D$AM$q&HZSi!Lu7cGLgctJKxY%jp=>O=FP^ z{x^oHaI`VAEpiY-qkRkabDWU#+WPnRR)5EYoZEA2J|ibaE24AUS(J2|q3JTkT%^b+ zRKzZqG+6dyR5*{f>1(kN4c-%Y60*q?`MJ}QX4Jz6LP;El&shaMalRLjxaYxgnO=0%H=JuYzf{+TAp_#{kDiLbKUw*X__Oti+m-$7CvKHnAR7^MCgS*_(f+|4|UGmm-X)a8UkmhUp zmU>A3ZB<@bxJ?}od3?t2yXnyw$^|EuLuXgqSbiCYV|Lt(Ql3XS{~9;J&_@M{p5V4Djb2s{e*#9Oba)-< zE|B-NGq^#HWWiVuRqS5`WJ{n@m@=bNgL=6u1M+zNBug{kmP-$~og&91Idl~`@8IrZ zrSdK1VL_2wsvE>#!|_`y`APhk1wc&gu(hbZL&bmkR}=Q;F7KF4S|*bxX1+!@#oVF5 zkBu;@pzR9XLAQu&d=5ZeS}wiWUw2nf;gYIFhLBuHl|;@5y(-$Zi#&6x5>!pcC#l12 z=oBVZ4k}2k;# zn_nOD%qcVEKDopja5GBh08E~`Lym}xx%uI2BwqARZFR=Xq;Q&Y9M>ZJ`nxSM9~0gN znNzVuKbagd!&@Q6z^t-^in!qUnLI<%uE=oBw)Q`0L`;I?DvX7X_Le7w~K~RG1ZzMyCa)1?qa0YndcS3tY?8_|}4Gk0I6k zuA6aZ6iId})dQk#$RcZV+aCy%QKSf?Oy*|GR;V%cGagXe_)jtGjvdr<%YHy!HyIYc z-&8Cio4Gkrj+-J&_L|wp9TbyIk?m9jZldc@twq-#+!|aGhPlu9@D|?d8b5vCa8Q^baWlb zoD51PBjYP26tkNmyC@T7iZO%d2$UAwbMK7sEjI?StPx~F;;~wd>jP}sRevbW}9VL?DxQ0A& z*-u69m{9ZE3#s2HM@9=pJFppPHG}vk6jM);GgO3uIKmUsn?oSuFW;|H+^*~n-R)x7 zz{sYPy|U=50a2p-Y0sQ4F!7Rh9;7f!iukx%f>l&+*eRyM`zYDyTLp_PV0LR)R?RCB z>rV4pgOik}RjE@`c`F05>Bs(hje)JFSYed-G1gpRh8w0uLtpL6lQM&)5_T%kvcYVJ z5@YbS940#3y#ga&2e!}FzY^@Pz)vD0_Kins+i@K?!XmZZL+@C8U3w@WZc0+fW!U&A zQTLFFpk&uH-zxESWeO|v)6j1uUo1M+5^<0lDllj3t( zkE_knF039St6^bvkG5$TxaFg|@xa1?Z3zo3Fz}M4 zq5Dh>s>h3#eg0z`&krvj0c$bw@}6wlLRd5>>Qyh9#Ly0M5N!0t+`Fw)*UmP&2opZY z)+RgHb_lKMjot4YC4Tc-TGyY2Q%r`U>sKv*A+cW>L&2iU)2>VoDN{pcXIda)2ZOKe z$|rPzpv?^fg1L}#hvz2{QPvs^qho0e25i&jL4o?=bQ3mK{N5{$q&aZLU4_}SP(*<; z0m(!8gNmST>E3W)aPV2f8*(~LPLm3k4oSYxr@}TNYUng+F;)Dis89^F*goC<#r0Ik;DA@%Ymt?^3X z%vk@r@Zl>1{BF5@#^CuseaxfrybHU0;x-x%TS5Dw{?e!4GQs4v-d}!7jybSlxoL*q z3lwvnBIl@x&*Zz@+ZB!4eRAkSs#d?XP?ZoXUnAe_Hsmxo`z-k+V7;bO+fUX4fq6PmpKO?d1%C%M z<@0;|*D@viZs_4CNtK-tK~x@yrm-(QFlcR>i8U^JJcblm1pAH@Rn6o2f{YpPW){eb z1CFl+H^8s>xTJJE&Wr;`=UJSYouci`X>kD$Gppi*ddMM7hp>nw!W4#Q^Cj^?Wq!5x z6n}bQJUJohg}d!{V$}led;ejwImf?y&zH12u+51w%foy~F%KwmkBX=luku_@T?c}@ zjsJK9_FXXfzn`=+b;9MMbX6Zgi-3MD7X2PxbO0>EO6gr$llCgPqd|(JY*Cs=AyC?{ z;U!OR(w@`Q%d6?TveL^|XLZt=aN*fcX;bzhMXUKide z-Y-prdov*~4+{-y=_GzLgLebf5@Zn9(vRlslMSDPM6Nk?>TF1aDfXySXSy|MxA;E) z{(AZIyQ}HW7vGu9@SkmP2PesliNG=CH9hGoSLo6RU+D0NlKu1j}GZ@KOb zKMnfTyW#Ckf?c)Sa|#w_YvP#bxx3Yne+}2O=~HB#FWUTS*^nYB1g|7PPSQo$I{qDH zqqc;;0D_TMc*kLOiuWuI=+s;oOTo6~zuWxSR--dkjjgG^5NZGTqeE!P>n?J9M zd4qI~mc8J>!P{sv3)W9Dk0{auwDx4Rwo;r%chKkPoX|?uW1@=`b&F9H*O;}_0>!yI zMc2UniRV3Vy-(}#;GX9*alZ^CkMN)~7*oUBg!$^BnMMYL+*wauH}d1b^p$w%ianB) z)xugD#FA7yMSDE-A)Rgz8NMff3W>#+d7TEi45;HYCW$90k?cpWN}O8igLKH4MV<_F z2Ranpz*GxV^VsWN?1@}F1FpJ8RX05>w+d3eZNgl-!5P#%WCbBu_qQa~ZAqnAml;w< z3PXzct*@e}(d(HPr@ZR?*2Y0aQeaGSj}jqm6IYjb{I&>su*OW<H}Yn3l@u&8vj&s#YL)Y!AudG3W6+to5Ghj8+d7C6_!MfD&ZiGG|8LpqHsueLVF}9)^`r`SH{$UJg7`-DK zpJhkGC^lr${&)4UmRxuaix3ueJCt)q`h}%SZAvr6aa^px1IHe1)0`6?Ui-$k|MU~f z#HSY%@f|pr_EP?MRHrbiv_&qMD1XCjB&`4nxE*G$w{u9T z1MdgHmh*`C#ZiiZ`o~I8FW8H#gJAT=YCpTq9`U^ z^4>5Vx+WNQZBq3rngVn+poN^RL2=@aF>fwf2bt&2D(6pQ1QB9n3XeaVzRlDNqv-wYL7DJBK~{)oh|9rG((9+R!^iIAjrj(<_TabEP? zCT-k;*5FN&HsNE4L}AXyW?6smgoT`}gvS@Jq}}jf(eYOV9@)>H^jJnszaq6fEb&y_ zN^6#3;%ySYLWNWr<&q9>tUqoHN?x!dWcNhU)B&hC0NetC_B}lDmfQb%)l+UaX)y<- zS8*<&QI5Kg=U}$G!YrS9E9%m4pk_25gv=7f`u;dSIg9=uE#RX1zX-)#$~-4+&hN}vn01SgjI&v@alz2THU9C7Hm>-!rLh*AmupGsP$(e8h79AaeA2vL{)VK9@HyL{1 zoxr?XgYqn$D{6%D=8V`*AhF4m?Ta7rMb1%JI69rqx{entM(>*bZMCg);kG8jjw5_j z-VGJjCJpF92@uYjSolU94={fYLBd({%wrsWkhoj4=citlXxDq(`9F~*V?Zx{U<`CE z#b!}tC6$PAWIc^UhRKrsZ^8d!`aCc;fXB`Bnj@UPaA*cByFMNdaM+jfngRDsU47-j zH&)0j(AXqi-Aq#5ILiVZ!vo^>I*Ntd>{==j>%!kVL6(LTK^`92%xg?U)_op-*%ouf z3p=mg_s*2{d2AQv3NXGgLI~!A-2Sk|uwm)I0Q`{q$%XQ%p%#Gj_s8%Fg}QGJfY@pC zDHIlkTnI~t4)apK3eS}UR9Y1GL^bq6$-}s(4Dw^A%GSu9GL@oU;bGPMNFy*TA~QBN zg{}ck^~R`HkT899DigZwk-J#u~72%k&1$pQOr9pifxbRJUMy*UnEF@?n zw?QXLD`*{9SZgmx+kk^WFKCX!UKh|oI7Z6+Yv_8=kKLW0koCjc0xPa-m&M&Ey9(nbR>)u28x`f60xWm-LtFA zmXP%LIu%HlgHCLp#~snCsB}?jOm{4Bo=-)dOv-;&Q{Npl*eYO`aJMcrm9GiLw~Bu zr;%V1al1bhxH%>iL;c(k7FuW7vw^m4u=l)-8C*Q~*BR4Z_gz&Mb1Fv1?ulPG`HB~p zF}^a!1*F*yJy(Y&*`bk~dl5fNvp{K(4o7C*ee)B zSayWnnAng7G}Pk{!!KVq@z%u0Gj@zrZM{Q)vM~eBh0frkE!z1ocA8}|`o)LSw~;b8 z_JTmse8BFimSU?Ya)3%V%Psh5*G(HEr+;Qrt{j0ot|*-#!`>6NW%0E(u=;kO`{ zY`(5M_EW{e8J}wJLc{>?-9puVsa|y^;J7w3_-xby;J%nI*aEFLJK0BcF}*YnrKvV} z76q5Y_6qBQ>io;4YeSt8>=_CwY(Vnj$H6C>*HK<(hWg9zG++0&U})c0-Z@K_yD^3e ztuT~Lv1=)qg~TG+wGi~9=Seu^uwUSPTYNlp&YpV|37TQ50gN zH!E~rYk^448PfhA(`#s>wFm>$V91J7MVcp?|LwwlcUP;99ePf~f@J zYDl)aYZ|aT)GKa!ZHz0>jNH6CGT8$b3nxsYV?*$>76H+#;$C-IT{8^{PzIAuc@;0jP?DKT6}IN&4RJo~XnH40c7~=C5Ld$Z zYhQG^R3A|ZK&zIPN>OVyS)vcZj-4R{zIaa5^UHZc>G_Q=e%DxWWb5lrBw*(H+)Y0E ze3#4!q4XAl(ef%_ZCvU5K-Dq6OIQq@;jMy3-&Ws}Sl5_}7pIBmjE`828$aXGF?s*5 zZSld!c(`%+%nmID=wJIo>{fDjS&3vX~?Zz8wif z@kSFaKre#Q=|?0#)5O{~1-~Kqu+uJ#)!pTUOQzC<%ckdP>pZzSPottdrj6!C0tf?V z&|y#vylDj9C+%ze;f@1qjU)P=D3Y!uwqGzhSRds)I|z)lXmF>HY* zHGo5`?1?uY!~=aEje;!Zyl+=vP55F>Rn!W&#B#cYHY(C1`aGJM9};oZ&iEePKy6)(7^M3x&|+=0DNC!##xam@8)$8|Q7km2>#4+LQgi>I%0QFZsaKUl+HDy*PIF6{kC`IqQrFW5!WJkq z;_z};?N5jMpiPU_(4OT>@8V@!+;?^Q`?r2@(>AVeCwz{Q=~!3zS6)~hwfn-QgKEp_l5pau6=V~?I!O1mgQVJO zku0NFC|4?`643{NULhKUCNpBI=&XsC*iKmcKrapyR+9clq$woz`_xPIkaaQS3^Jq)~~nH2JzW=-4ms!e!izhx$PLrn|ZreytfX;P58CxpFE~m&6DzQn?$}CXopkP0T zM)iTYBa6^}7&(B{{!2&nl^;@t-!5Ek%a3G-Sg4d<=Xoyv0)1GqS#d{H=lgiZ@)=J( z>OgD09m)l3f(l1i_rZY)=r_IYI|1(NiPpWhwPc*dRD}GlayD7(#?^{@tW3pLiUkaB zMpJ>y{+vDWr{$(MH{~e_7^qkk$|3vnS-Kgj1{TaXByW$tC}^jV1(rMLbmVrPf#H5E+q@1ks1pwHGa@_Gvo-`Qs>9t320KF zJIdV&_-KMo+e0vEctvgy_IVUZ%IV4|Q+fy`wqKby;5)a&2bcwKe82b=*E0)^H&qtM z?x22oVy<#BMFViwgak2Zft!a6!2&L^9kXF*E z$`w6=G}FPrQ|ew2&dQPDE;ln+Pv=9IYC1`x=1I~;+hTy-FJM0u8a$v+sWVe7?j_|Nx)W22@2QPN+wGShG-8v7(Vp`T@gkL19 z-nd|JdhkF9=F@KBBz8O zw%PJZzX%RDj;-1u!Nzh2ST?W=177OFQlX2yLcJ>pRB|u@j0dWsGL>Di&?XN5xnfCq zaeny0!?dcrIEV|5*oTMg=&q75%OVqz^Y;u=;Ku&zA**F$55EN1t!m`5?qS}+pe*c#~v^77clzPP5q>0TdvlPl7Lrfe#?1>JtgE#1lh7sVWG z-RPKbNCYG)zEC{X8nbOOD!|Th+LmSrAnEr%27h%DVC;ir5Z=^fgYuwZzMXJ%GbSkdpxO!Z7v`wCUF5?@h6wbfuKYVJOp<`@k*EbE<>I8A@i&VD^xEI#gx;qce`@${D)pWr0oRs#?^+QUB4Mt~kbpOo`S;(@O$_V{H#y5f zITJZ1G&z6@M=r6CoCRYKq}sBT^63e3V{>GO+jQwW*W{(MZZPLeLg$sF;AWPsVW1Ab=qm%4(|%85cdkt(s?wrs-2db3t~+5 zp4a6ibfs7S2&u9g{%7ePa)N9EUX8sF9;^soq-kYvv$aIj44;w6hr^q?VvVYLX@*}r zJwNpQNRT&8VGA@y8Ua}qRiN2T@GqLPUQs12^DmNAMI9uyqC1hzm!Ao-)K$D0%(kQTZz;#T?nVAxz!ux6 zK18laGlQR%uZ-yOWHAR0nWUGP4p*JxXOiN(&)tc$n4~p&RW8}XFVN@4{&1t!3Rp+6 zA5f%*N?aFSFK+mDvHVf=6{#ViMT~V0>-|oXc4arys4Q2lmK_SHh2(XIBE@r)aEs4< z|BBGPAUTvJJS3_V)drsl$O7(`!jO9D7GVjE?vYO2J?*gRj3EInls=+a*fDKc#M9tg z$}^DQ)~Ruf-Eo!P*b+WWtU3fcd&}KRj}H%!$G`WZ<8NX#=LlWuH_!7}$VKVl@Jqsj zlXMW%y+u;k#S;t(C&+5qdhsW*Sdq7zlu93eE``NJ_bqvk-|88e6A@m(3kBELMAZ@0 zt16FeCWXN9zg^fWI6|LbFA4FwVM4-s;+zQ(>yd}|$-Z5>{JUjBk;Mi@to!a!vdE3y zqC6|BvYKL7Q6!y8%)^SyeDI30fx{t5g5Jt4B@%Dw)L4?YYq&pqp3j~K2zZV>yvyx_ zU=V+K)VrcNzk1#M72I1boV+fa!!lc>#oraw$vE`~@8FC$1PKSna^_{cATfH|-fy!L zEVklX#lQSD*)WFexB5wzP%Ll=7f^}&=#Bm@(d98a{f$tAUwNE`= zCGK!;TF4$W;5cv8zGN(Cp5lkv%Ixi7|7@`(x;v65F+LS%Rg{150E2gDN>=IgG9J(;+8KGtv(U#zAJIn9uBcKu*=E zDiLxZu?nt>`81qNCfDT8PWE{qFO{CY5_N>WDtP{__?_HeSNZ1CwQ=~Jz+DA7^Nct& zweNd2s^LFnXEZv#xXR|9QqQed}gfxN_{ZXAvUiFRI<5Dp+L>mi*ZT7SP=4uW8Fu|{5J+6P<1r`A*(1hogyhz;zcN~ z%ct**`}4i98?vV!rnl>ZE29d8i>8eX7M@R;!>D+E%ZR)y5BK}Jt&huRd2!=fO*Lbj)MWY%;fo|c>2)JDdT@I(mW zWJFL>xRvZ;EsKixqwY&&%@_iOj03E(o?;>Dxrs`gZO(jPnB+deyar}Vw)mt8izZ_k z8rH|Qh?63Xfd#=eP+8O{z+YjD_=>c0<_+myA%@0X`mBfWh@b5VT=}edZ5P}(qo)0b z`Coe72 zoZnhuORDf<&!QXq8Fq*kOnhWbiq)FkVH|F0kzKh$yfF-GP4)!nw6M9&Vh&EaBF$pV zSD~ud0`^&1f@>WR9@u$i<(|E9A$aU#=R47B;%#%pZX1oYQ{A5rbKjuE$kj%F$nFI< z&*%x{(C5euzEAfA_J(rZg3MraLYoxGhZ5c^ ze0xJ5240bK=E6I!E+|!mEk;hVM8kr~GrSCZ;~ebRhshwRz|kE`&d&C9_)QOoM9NCPjtzJ}Fn-5#14WLNh(q zIqverdTv+nklM_Yqc9n0P0a6Y26FO8>4$gYW?HOA;|~|)lYMTS>A7HK4~|hRtVADB ziJ7L#?Hsy=aPNB?R7moLTiB)JxfXF*Tnovk%OSj;2MW1}21u;1hz+9(1FMgHas6q5 z+)I<5=~P&{vPcARUkqbW4V~G7TLaFXCnJBSV7? zua#)j^trVkPPJg^-O7YoQZj}dx0z=xlF(j;^PRv^5-4

    slo4H!HBkZ`gJJyhJN{vE3TOpN}6qzh3gXt<0F7h^? zyEJp1_PzO}IHXYC2P(GRu^Ypn06QipWQddU!XtC~EWPjzLp#80mfUv}J;ndMTw5mG zHzZnXhc~zTt$R#?2KGw5KA-+CFgy37%>i8ZUpk_%fA9R&wdodnlEeI@i0mFiKD4qY zhbi_T1)SxHS(*m1kw#4vtSZCmBraFf9QD-WsJd2BM=*1bJrU;ShO}IqJmeW zAAl+ZuG~BMWm5TR)V&i5n4QM_@9w2lwo@Xz7A_c z>%asIZq*$+I`lc?Wmw#|)Je3j@=LbFQG5)G`0LwLRc1iZ%T4duhJ-VcogmL!;AxE6Co=>%YkhnL(D^s`xwut z)0i}-0eCPQeA7inD6r#@qi_>g$0})MVD-eUG00p9rDc#6=VU{6ZXVXqT>fFH?}(<)Hfh9Q%bQQ%evJ} z@L@=}5c@b_p5$ssi@5!3+v9Jk&(c}Ke0m$3L+AS=yVSxl=-&6fy=V+}1?>*W(%{i7 zW;X8Hk|MLA#i%@BpSqMT2%hZ{+vE(uI0^zj*L8RQW|f1gWoJNvDhWtt4GFuUSaG3b zVR)`*HMA<%i`%8<_ebfSf|Z}&6_^@TqN!wih#{d(wIQ+;I**O2jPV_cOi~hai=;;F z^67FJv%#<*G@qx(<62&hjr;x#LsA8CwxM4;G1F_zu8?)?L9&%j|Kc_Idc`qlWZ&TV z;B%X_heR&nIkI zWUHt|Og(in2LnMZ326+=>1Do^QOkU}BuP7vFIPonjIWGYHlitPMbzQ&L=Sz+T8#X{`y2>2vn=afC9UfkiY6K)c=z&dz=&h^|! zB?ot_Qlkn5mu8-kb*OqoyELtA3e?+g65a|jBy^Kj7A19BNYaGe>!ZJ)s< zlU`u7I1&z5VTDw=s>h_!-sh3RntLXY$TW*ND?JXq<$WHgZEH#+pVmH>-|<}z3RJ@i z#904xo&C(MM<4npUdxpGrguMX-1%KwsyaSl88=?0?2y<2Ef7>MR8Pzfes69olO$Qo zR(j=6Iv4+!xmj_SBRDgu^99l{vB5_VWQSKnHUoPmX1n`5{yKME!ZE;ns-Q=)N(r>j zJL7v4s1lIHqypv?~GU#g|_m}v>gI& zRTMIT-xNYj8ZN3=)`++spk2tP7lv1Y{&WEineJZiR6%)cHvyq6NOyAUD8sBZlaGe! zfq`RHRExOl?F$g!!Ey8P>W{R0l*1zhcL7qKtyDw9i1{x2rt)dnJB5F+VCatxzlb0o zyK&-snHA&IV~YKhBHh^HhzeKOE(YbAMY1-S`xKJ`c|~{%L|Y(LjrO5Qu~>44!Id#X zxLw#CdzoMo{1Lq;GF9-@;{sg_0XWoI+Zno6a5%hJ&b39S3J@ue6{kUn1PAE#UMtua zRs+rAC9piF3Lb(=CEkdyh5i?rhvcMsz2A9|FG1QWB&cc;XT)+qGUeX`1f4D5f>+UX zzS|WXQeT~54FNguxAuQgS4c$%HYF;GC*Pe$jOcg*U+@ohBq5s}OFAq1h@*2f& z2I17GpPc>igl}5lcY9jxC*&AEKirLtTbq?}yGXI;DM;Iy=+TN~E4*}Cvw6zZT%hlb z-7dT$#R{&q0T0L5&McJzH#i6%ctT{koz4XtWrVcFx=C$XK(Ya#IA_l69h#M)8F8jq zA=LV;7a(a_GpUc_a5#)qcaXx64!Rb&&KlKc0??y;8lE(s%Z2K~8_}Rv%%Buck79G+ ziO**TK#53$5}DGtouMb;3p!Pq!Ds1Sd5Y%Mj-)G%-ZSer*qc{a#D6(g@vomPqE0h$ zU6!yCvJfe8*)x!yQy+0Ktct!*swe8ySEBC4zBk03*DD_O;HP;(WAGtda`Zvt-AY^a z0=KO*wxdVz(Oc`;_1+I*6~n^f99gcWQTZg`_SAaSJrT6lDDmk1NsnM{TQGjHMyJMW z4GDO((N{0HBHbUC59;H)pkcrN+#^A;#PH?DuP*sQDs^a94`~j=dz^?DUT_%f7jJdU zdC0Ku9t~1yOuch3qx7>Sdv7jgZxLIRb>iD-lQt+WSg6nh`!8||ib}GWGJpKEY`V#J zYEocU^AiGGctA~(L+@qK9+_g1ibBZAulbx4p=AMslMv=SXjw)|@R+3hwZ+c~|IN5@ z~4zfq(}#qIG=2$Q$;&skhQ^FcaD?IRZmY9l}q&zPYCY2O`9$QP~GX& zCs@pcXEFOHr1&9J@V!yors9gjN}xXZxQNT}+ZCjTO&nJqtB(MFJPmm7c%AtJ=%|>6 z=2l9oM0uJEV4XHYKKv}5D{55hwClnu$ITlDe*pGPR7F90U4qP6RZs=XVLIP@zSZQm z;|cCS$cA8ec__~SIWcbDxHBQ}f^Ic}F-g)*a%3A69C~7Ux|_Tq4tCPH4S%%Fjk)bg zXD4abpxg}`dz_^^zDbaJ9MoO{bA++R&RA5tNEOrufwicjQw8_Lu;GDH$<7%N9mw%N6@%w= z>McG^E@iTJ7(7;vj^%^XA@h7d``|hD>m4dvpt!C5umg%ufhnd!{7KB+P>ipn1*6ZK zGih&N4ru381eL({`st)hVKdQ3n9nsapF)WFLcmUSPS{hp*Tp7g=%9E$fv$LG9~6Um z#lnX__|hbchtvIIWlya^I{TC zJ{Mme@R3)is#3%m1n1(h%my=59BSYI$q-#EX;8L^jpB=b$n5;rPVuto{Fjg6&B6;m zW_t&$p0Fby?EjDlSrObfH2?SS8U(+$fa(38`Y6dAH{Ogdwc3n7qSzjaJfIR0)+Y9e zg_BoBVSd2OYt=$$OD!^iRjU(QH}@d6)_vT5mbRNNMu%PPSISIuZk*?J@$qL z5m!IVm+`lyj@Y%~u6tl$9s3&ISB1)#NGDprvf$nED@cwTo2p8y>EBMVAUIfP7EZne znRE~f18yaqdS}ED|NEgJoU|w4Q$_)9->O?*yySDAt*mD~R6>@=_OKra0Z4`N%0LsM zs)ShiVOssrjm~oQE`Go~@&wX=J6-#>F8C{6wmhJA2sE=aMdUuc@=Mp`C7=wt56XG# zl}|zp3Cp1029MOxy(*{648Qze&+n?}lDtzT7pY`6ct^I4oW=hzm~ff`$9~bmu#=R*bPYAPL5IIWmu*x9YyTK$2tS zu4GVbDn*u4iP*rKBmvsT%D{5q!}&zIoT-SoImHzo$17mx7_?t8hzm|1_;J`2TS5vx zQ9C!rgdJWDSCfig20UH=SNS8621EPP$4GVdj6=bKjOBzB|DDN;~6`=%w*z#0HE5!JuVdLiYun& z(_4JkD=_v^8S_*%g6ksR=^pCz@x71ViWu^J?)013+|bFS3b3j-3nYnk>J>j&`!!G& zOA~I7KNWd(%2mM)rhwk3JsY3v`6;_7*d)n=hjr@eh!RL%e$3>_Gqq6u0%hzCO6+r3 zNtOsQNo`Q)jJg{!m}- zQz+g7goIe*SY_l6Er^ps~lRH z)gee>QY1koWHUKQE7J~Sa7M^gLTP!EVySYY($q=tfe?#^>&2;{!gIqP7qRN_9&avL z(#k9nEl}bx4NA;%7w`l;nA#g5&k8WU*7dL zI8W`st&(RFk-r7tJ@Ca%!liNZBQ8oYS%Rg>FR_QN!r5@~Xdh?OexIvxkns;GxH5xp zD7&ZLpJYftWmuF7t_!*d1#@kx&5BwDND7l?0z%~SQol>NDsfjVFkWRnUZy|vwt8Yn|`L1FA;`@J*&YqF}afkbAe?|t$N4f9Td*^SD z-?b$wc&*Ho9Xf)NS##gJ6PY4Q``#AerJ3i)XEXOF9h7m`<$Hyx;q7$$SGLbC`IpVq zC%<$WP+tWozCeD}c+NU+AIJ}X{*xF7#WxoG^yQf*g@?N|~%N z;qI)&G#s%p>L(n1Z?wN0ggwnuu%RmEJrp!vSe1T(6xYlh8F(FeRiu$-|$23 zH$6Z1kG24LahY-7WfH3K1_jOI84ITzk(vL$F`-KOklal8P0CkSf3f>p`QP09h2NY@ z$YN@z6onMPehUitv?HvB>3vMjO{2wZxkEUz6U~QEyZS0WF?i@ z0NbY?g1Wax@UJTa_rJwqysi-r>LXg2Tgs#I!z0QI#tSr7)9Ts}A(=n?t8Jm?Yh^p^ zlw%@e#OjIrr!1Bf#~2ctm}8QI@-5^7Fd1E!KMv@IdWB8$+fzG;KDJ(LNXXU9&J?Em z)8Mlq;+F4KvG-st2X4r#a4WMF| z=5>FPxsj}dLedDR`2>p2I!~jb0IJ@M^zcYS9G$75fWgtRhjX2MFbv{9kKSZ_@Snuu zrtb8}Jw+C~v74H2wQQ`R*i4FKP>I;)3Y|re4Z^au2EjR~NwO>-n6Rmyem-gmKnK#r0ppd zASX|QxI73c&7i9UwMTn_tEz>DJPoLBU>N|1?C?G0l+)<#2KmKVAL0)M>-$~xiI;gB zrBg1E+Ey~TEj;YhCLIV{IKDy)tf2i36Hit7!_!fAVniJ^`@^!s>WV)Q<7lC^nZO8_^lfVsb@os&mSXzJD&Him5+m$QDF)_*V&1WxM2U zVr*lppchMWJ$n>}gdT!t5l0K<^C!64O>zt+w(Ezhcdy{E@hMlM)?Rw|&vPv(x%7V) zP9t4zjFOdB{bCD7v-3uir|2i>=AHl}(c9z00B7{i*7G&oogw zMA)&?4fi&gvVHMQ3f+WzUdWz+?axWmunh~lD_a?zV0T!fI(O_|#i78piaXIr4b&wy z(ph7n`w;@IJ>L7`vn9~fA;V6!MNy^D^MmzsODJqhgH{m8UEfj`LL-@;-VvdPZda_F z!f=dv>1(IA2uosdq&uSgs5~uHGQjy7MT62vCr!hemL$m{?;N@kd?l<$#%@deH*8no zzrhG>pQX^tX8!LDX&k%q_ixSnYzf1)GWX*iL!-0+(U$?lJg&A^gReUdSh2xxbn(&M zK7XF3wAiKN-}vTLvT_XBVYNtXrdaSKHc*Mz#n+^bni^G}cVoi2gzVr&)0RUobRPU@ z4oOO-hJ@7|9y|hMBKyL+IthI^}c;F5^-H~4)VCML(5pKJAZXNw-htL8N zk5G?fvXEZ^zZ)lYwpd|3i(*$&B#la}h~MLN0t(^QD=NM9@m#%XPb7Z45%46W!Sf|` zXRi6RIu38we80U>}7s7231(ExP~p1W5yC-7D{430`AD zhT!R>zg&}_OE4rfkTZ~vSR*?m#7nc`BH*{2&;tB(c+oLZ=HEf@5Y*5QqH9FAr=Ef$ z?Nc$QAs3CCqZ7m!(?OpB+TIO*CuqDrNwOMvoKKRIWSg*FI?pSaoqdEhs#Z@Nnxi&k zJlR;MmwX(Of%?vzrk{)$SD%lg;R7N{&(9C;X8Hr@7bg>c}H;} zJQnkm?eqgLygCP(8}s4Ynwb0ksldW}DdJMZZm-?Y3W*eGn8gB0XSk*v=#`Nxx(Yn< zZj$U%9-BI&-e-wj3Oh#-_>#qD_^0e(*ggW4-`TNajshNYkiSRz%O80@5++kI{iaQ@=1SUL#dVvzUTu~Zl z-q@b8FxhhM^mwFZ?u>>w9d5o*F1uM!4MaZ&!pelLf-7FYxeMLRNs`5qTS{n#ZYHJP zD0ago%aUcCs+&SoWS*Pkk@Gh?fUf9TZHRNtUs&9*r>FYXGX97}40qBv` ziq3+%PPsHy^i;`x0nyM;>Jt?Hm_6vn_N1%ky?Qe|cd zyRB(}c)wX)00`qtW14_6zRzR1Piss*z0tQA+vCy1RA}!L^x<5#UJlnH*)b@?K5XA2 z4dT9ynm=yw2e$RQe0K3}yjt3+?9JC=BCCZ?mK4g7qL)Jf!6UH?C!>|&aCea5Qx$Ah zB`EX1;?<Q{5O9r5hWH7YFsC0iX_ykFIW1|8I@Bxm} zN{UUR$OeQ;12+;rI9WRh4n`tGvWAwNF`quvh zT0o&qSz1jtx-lpYSb<_2#THRSPbGFT=%w^{W2^fIpx$SsjiLADS(;h_Qgk#+8={Ie zsIZ2tac6-N_^uM$LT(Dr(ZH$7WW*MRVB2-Jra1~VcwJ>19L9@1PuE4b;f0q`HRZEf zU)|8`#%qe5hUOwzIxtIs>FBJn4a%DE!b#|aAZ}pi17!^Qh|FMDVPmkV<%O5Q25`|a z`*@c6ni~E zR{14GR!pn(LeiL2*|mvhkuz%Im8h-s&afL0QU(=mm-e#qLd!r4V}9>4)F=b~?JIE> zw5-vqa><@Cq|xdz*HP>T6se&Sm%MXL{`VV=!>IddEkP<;+tA=*bZvDQY>aG+X^`@@?`Z;n$|L3eG7F2_-bn zkUlJVa;og2Vz26u5Y%K;=!c^85R(3!QrAHJfqt0fxWk-*7w6>pkzvh_eH%YWq{4j{ zex34OfNgskp9RT{-Ca9fc2&Ugo5gJMX;o(if2t~y?T4*2w`&sc%I&ASVzG+Y)ozOw z?)c!_>P8oT__pyUU-Yn8n37L_{TGt-#;iv=D?^!1u_+W;hQ;K#BIQ7FcPG=}cSzo* zO7|R5DD=;$BarC7b|hclI##sQHYxpv*a|x-Yz)x@AR?h<&8*rwvcBrG$#qvm%nJUW zh*ptpr3R&ET2B+2Z>kO7}!IOtEwmB`yyG91go8} zqbE;W>X!~-i50%PWEkS?7=K2YVEO8S&hY!#Td8A#~_><^SsEH$p8o=ExUT?jw1F6-c@7 zPK+G1GKjk<7FuIVsl=n|T-d0Cwr}iaNa61ANduW0tlVy9mO~XwV*(a!--m>7SsYTr z6$Z?gY=8UOEJMOcD3`+PvY6}s#UUqtR3Pk(ZA@rXW-(b%`K?pG_g|O4e=cE&2F~6z zInVIf`-Y(%w{Q6Vw(MO`TZNd{iiX3EoXuS&+6;{J++d(oDN|eNT?#zasP*~(<>mW9 zWFHp)KHfOhw!zG8JzG01W2WL;)EDd~RnqcU6c96eIhF?E{--c9ua29Bk@(njSrV-X6zWvQBkO*8zHijLc3;jz(7wChoOz$}kmcg9l{;75TD=)B6 z3;*L-*JmrZnhK7oF~+PIbU^?@)w z=&~)JId@Ct-KSlN;Z~nuu4U=#mdvp3@ z`B}PIFkkR6tQlG!@mNLZX2l-GM!_f2?%4T}f7>vpe%1+65WffLpz^id0;8h&d-uPo zQ+N1h1~2k%Vo>q$5$TME_@203+9^Cw8kH;jo-#?2vr%`(r3Ei=7U}UYF*OLz1^{*# zZ|!61ss2!>tzw-WVYrWH)hla6dK#Z5XWmR8VK78tp@pT-1DX8}c&_=}NTx}6KJQ^X zL4LQ|f8K-qhTXTWt3(yB+cw2xCzDbnsT6$(j9B>p5Q*|{SFWCj?}4-=G>0yM+VvXk z(`kl;`}B@4T$9hEZZM@F!CdCQS#g5gp6WU)Z=N@vIC=g%Cw4GwX`z@prJH|IlQ=_uH^on%R>haqqRC;BEv>v_N=NTb{&1}OD4FlZO*ue#J)l!-6~(4gB!x=M9>0#gnE-(hg-%-n+LMNa z`lt;GozHsh$U+ry<{(^M0?C&R$`7s)QzJLqa;3X%)ODv7DxUG8&+`L*PtBPCAkUHB zXVQd=VWG&C<+5qAv-AnyZYEEAk2yk@(y4;oA)V2f*UR;6^gTa*v1kNLoP#iVz9LxO zy=;n|agHBMPP~=5&z39w4XF!&cE16pW&_2pLx(BRT$R@AZ3IbfY%4bc?;CPeWR0C; zZu(vAQ%_%)UWDepk=v8;29q<8;=mZe?d^R(vaBtko0_|cVGQZEdak!9)<}^SDzP1E zL5;v#l@nYog*4h&oj+>X&%QIw5M~GiHYQ|h;BukVy;jh~;At#cs-cZOC*z)aWQjV! z8O#Yum*~_-Lg$C}d6=&}8+SSOJ}?09i`S`ZgeT*gqvwe#y;f^>%0P@Kc$c&!BrVi5 z=qKavFa~v*e-3b*qy=CQm~qekCEZ zYv?K|);}3wr&Z~NEfs^WIlSlSg`fVg4-7hEyngWRyK64^)0h7gY=K%y-w*eY+&5-7 zRAu#x*+H?}DA0X?VBHc8$~EteEDf#;`bc#_kxBFri`y{1jvyeDB;BzKLQ*t`1-Y6^ zuj`r%iUrE+8dN0P1l0e?Vg|j)1k(bQUKka|cY{NBP~13v&>6^g{M#JAvMc3Rw`W=) z)A+*$`DC9Pd%hQ}KyZv=>nQR8f&dxQfv6Oi&2I`f`WvC-z(i58NR!KU3U>u{O~M{^ zDm-i+zB=k4~9j~&Cj$`8amAvfYs4sru zS7o+srf%ym+i4udsW#Gk0(!mAK!tBbMA8fnadSnA42Wj_lOfIU^JmKM9Uo-y8+mwF z^FziTa>4|*PY^2!r6j_7)TPO-s>1CuJ$dN01PrwsVs)&2w zy$-5fjj);&OS)9&p~eYRH=x0Igb?g7aG$LjyfHYW8~6eD|IGgJ5!Qka??>I2$Qn1s z$8IZp=qVODK{p{jvX~-?kwz-EO0Sym#h`MET#%@7SqS=b7kujc%}23j63<*>I~f#y z$)^|l>gjYBYbNtY%E&;57g9$3Xo7h0>t4B1EXkfRckbNOsM?v}Wa=e)*vMxJ^>nVr zNMk_7z%HTiwN}u|>;oqi1r5=}JS0WXYLpC;Km{>{;K(k4PcAYu!vKo~HAAu*ue^yL zSPCbdSTdXfN%vh)iJk6qr0X)*O+e&G(Wz>W0fN!(AOeJOJLJP~5nk*}91j@T8NoG& zwWHpWds)_-_qg+aB1`ziPuv(7KoL41lDn2-vnaBXN-PfO^0tuD_Ao`0xg+7gE(^Tb zJ}V~Zp<}}s*r~b}Jd*I>0S&Y$uED_|qwfFOnT58svu?Y$wNod1X=V*LsCOc>HPzwo z%W+?&Qx|AzgScMrYw{jaC9D$f@Zyp^je%PQse-lfNh)q6>;lgzH!mcEOp&t>@mf?y zPgocB)Rv~eZ2@A3cpzU}99E*hcn4=r>tu}`2>`q>agHG3g_}m;053$0I$CvP^Xpzz zfI*lh#d066%pcQyPdza9iILBdToWua>j)}FbWQ00yZ9t~N54OMZuMk~kKrNH?j`yB zd<-|PSOey-0Y1iViiMhl?Ns6xc}Xm`Xk;<#1^ED>0{CB^{HceLw5dK8HbUU&3pF8Qt;q zU7Fz#(Yx_>#SRfY=I(CzFAYhy7#GIY{SMQ8W;SK+KRok?M;ifbUdV{D@ zgR!PQ58O12DEWtDzf{Zf9Pm$e^;>?wRJY}{QHHke3%7V9JKM{WiT;i8Mh;tfBYP

    c5DA{uZz2-b9b-Sja}Jm?Kp zr43A7o##EM@EP=8&0R5j*%&gw4!#lg+ZG33j!Tc%c5%#5`p-}M8_1&3HfiqrDnM#; zBw*H2@)U|BQE?aEss=tF)BK*EGYHBbkhsVIDMeT0tzPim+dKC!7{=+BZCk_6U!+a2 za}ByEyePcoe_HtY6d*F-oFa#nS)7dln|+5ub#pW|oCdmxo-u9G&8aICuYO({z>W^K+?pc_W>y-Q| zMJ`csySdf0UVQ;JyXDlW8D0MEGU%cVT* zVEwcLh!*JuIdEl@fBLj`S*QA5csbQBs|!5N>t(j7chE_Zb%GA>a;ht&57g}~$IKVh zlTP&@T@9&`eqpC-IrMH`_FKuh&Al?CX_{7%?VZByR9^@`HKUwLCv|~|Jj?6&{EPQ< z4!tZEZ;nF;&%Pa2XTG5JTo9O1_HoIiKao}JQ0Bhx4*~VF5uVOwN}i6dMBE)ICRCwM z2?Ib##oNwQOzrj@^!_v)#en@Q{97e^l^4QRGL76@{x6uH8H(0Cm{j?#W{U3nxZ$V{E&xCM{7YUv&%^7ovlJ05e@1AU-w(aDS-$|wkpN5 zRcK@Z$dDX_UfLnqe$OhXBd+k;$6X!vAS4H?#PT?c!h4|e0X1rRVtc?oL69d;TrstE z@=Yn;XJCyDVlflagk5(DjqX?1J6wE_FN{w%XuI-G3agQH)+p*)pPD~ zXd<37p0h5onh5vZl3uv;-d*eZKO1FsIpCaOwQUxZR^FV6f6*+lCl=Y=326ut^d<#vPrIkwgG$MT5lyl?(soWgjn^}2omcYY zlF97~(}<;h>mn+6{lFELD}Mw@Ivh!>I4(TEttQ*WN4bUEQrUpGfvy)f$sUN$3iawW zqQvkgzePu1BLz_v_Q;j7n@j6=*3sh^Pd=-;bl-4Tq40aZ`p1&m!_F6zoVLx{{wHKQreX z$RBlhX)bVkys_57Rnz@sE*zfrCj+?XCo5C;+SM5kw_A7rv+;r5*H#i6TmiepyL{Va zALr0mz8TN7%gSQ6idKy&kMKwb^RspQ7 z^`aiSNR=K?I%EH|$2{lZ&tasEzMeV_ci)%nFjyBAzarSNQEq%Tsv!*eROr^ptr3eS zjy%FHm`-&qSrUb3@%kBcJ&me2x0(AgZBR=@8)@V3g38W;$-PjMek$Ohcb@9BvH*Oy z2A~hv6qNq{gV@9TCaKOVpW9CkP3eW&&g!oj=B;@b&vgoShMo1=9o|46nlct-IT=fq z2pa>kxJ7Bpe{Xi5mn~18L%Q5JO1#pdO!zS+AEd|w(2kRDnhVMl5SrDr@r#rdUfQ6f zCtGJ^%)B*7USzu58QCnYGCw@*aWvF4 z4A9Sq^m#*OQM?hDC$Pn;!mCR<$Xxf+GzK!AQAN5gIbK<^%*OR`wo|j!~sT?-f1{%{I z<}@Ylv@Qv|d;q3cDvm3F<^<JIM~dVP83( zXie;CgFh2_@Gb}6@$ZZ2^WGG74eIb0zu|PFV9(cJ&o^(+IL^L@)i>|HiOP?^^Q~zU z%>GB>((t2@pmE#2V`4+qD`0-(b9|pWOibDVfWNG z%hW+N>~S3E5pZPF)mG-1D&2D>vrnqAe8iYHp-?l#O` z6E@y9%qRCdS6_AoHz<6Nln`S%Cul*=((D8(8|~aZ;jNM~xm`(4XRZ}?PxLs?G`qDb zZq9=v)^fzp*8aM2u3(=j5yk!yaBpgo|(d3-vf0MCB>jIx;@Wjtq{ zvW=g{GhdIS&3;Y)`$5M^0=qT>_x0g{A9!TF$689hnj)*9dRS2?%mE*!UzjR7%Qy8t z?x#1|Q%~Wn>3A}ZeQv&B=$}60IUbeO#_-XO717 z#e9NZXj*s&k#0(&6gAOXC8#yITXyQrc1G`iW)8HH zU}@Du;4HcqGwgx;p#wnQQ8h{Hr&Z*OfY-iN0(lM7qjjDR+9V{;jGH=}{qQiq?~cs11JAI7$8YlQ?|<3N23m<*zED1E)E*Aao37bOFDAWAZA_aF z>Y0q64}#Oh1{V%Yufvb9!^PV{-@G^3?3=W{dG2G9%FZ`&<6y}i3*RK0l0!gMi@ZSR z=-qQ$<$0VU?jG5V$y!A($pT{5vjHUxgi2(WAyiVs%K^f%ejqM>!Xi{p*9D$)eOU?{ zP>pmi97cgDG71;q zP8W&_{$Dl36%ocB zj!NNm8Vah%%@6tV%eT`=eEGmQeR})HB}MbhnE2k&9~Y5B>@eZRM&g16CQeZDPbpGI z#i4Ltrf5i(8+F=57lxf4@q!-E`@EkOol6(X8@C{l^_Xl}dvoL4Yd_>~gVq z_B7q8$|msvr|DYwx5f~yqf>a;(RxW(7~LQ~${Y5;f3SzBpI|KJA*WC{=)J`c3vGJ2 zTc>HT+NO=hfZSQS$e~!Rop2ht+_3-b00g-UbuHF?&o&x{dtwceN4}L12dJ2;Rh0O* z2{p)Jp`$yawW@yT2F@4h1CxNNysCRKirI~T5&uJe5$%@RGN+n}&ruh@S68DC;%HVA(qES-46-Hf{n-!-pH55DouUy;LZTs9BX zaU&-CG$n7O$O%|$jOjUSgg-7X4%rEEGu@L9gSFF`y5kSe-l@C@%@W2NiA5 z6;?iH*yAM0r3Z)xYlUzGyg8d5goWoGeb3mP0V7~IwKXD*$$}b^{eeBub3N=~3jMeF zG;@JRSw|ZnV4D?%tjeWV2YulWDsSeaQm;fW zy?S-f;n2&RLT|YH&g+YPVeABXyx`tfQn|HLp>8)AGLn8DvIz2w309*)?YZad$6@ea zs)whWVb%Y0=@PQ-m4Q{Y1+4Z^^4%0EpyHOOw}E2Z4KO%*b;6|E%mY#?yd)e3>fI_y zGhNF)L@yO2i&h3@iHAIHhh~ZE$Zl>P&>0%|k9k_frpaqS8b`0LCdb4ys3;9}wX(I><{P>lYp zE6?W6_HVOR7*Ky-Z;=YJ$Rk2%pd^E(znq4dBTnb9T_(P<6sQIjvPR$ts1Yp|XQU zCOh0sr7`QLWiy>JTqi7K53mVGq6Hu+M<&Q;tt(kR_C!KIiHWcgw%~z&ftywpPeSq;7+#*8hgI z)DM~LniZf1DsEA4;5-Oe3rBdiRP01OTYDyi&1Bm%p0n;@WjCmQ**#rt-Hyt}Zn$w4 z(nj;Dsguy$@t(#Yd4|ok!C7-M+Bo*$XSC;BeXelQioB1^o1ed|UiK!r;l`2PB^Irx z1C;zeMS4KTQ@v|mEve*JNwWMhBt2x+q{I9(>dUb=XY|TWa?%8Nrj1`n=LU6yp!R*) zrGUr&H)rbAClv);bSf5c4#u8VVs=ok-YFRV+Qo16&b$84(|INGHi&ll=J`L41nyR*mJ=_)v*^p-qZhxv zGZwg5C%5r++;b4zd2$s)fUdDl_L!d+__Z#vMC_KY`QPZcJHw2a`PcGRk$g8sOoIht z4p4GvMl8iX`8-ZN=n$-)yka&`T+-R!elQ<3D0E>feSl#skCVj%et6nO6DUJ{y^y&v_}x zIRQLY*m4e*Sb^uYx}llxMw-F1H7I*MDRg7-oUi~-B_)SKj{Q_zHr?$PA8z_L~?OqA0Fj$Z$i~!eE6oIA=wt+%#go!36FA?2nU%AD2wSK zdxQ1rX8$#Bmc(A1aUE>bZf>G6-ad2BGs6C+1P#j~>E&UsO7jbcA8Y>nWW^Q>?&;cbJ*2iB|7 zE3&Dwv4#!3v6@-KEcV9+v|a{7ai~Inng-sNV!xe&4>ReKz*AgMoaU!Z>+*&}yq1_|a6&IS~{et~;Z)p{E-nme$s96d=_62z_lNF_-d$||| zen{i5seou;705OYdn}iJP8KYemiyosc6YnD0xHizk==wvf3(jp+YPTp1 z(Mj@34;Yzv49SkNdPd$69{8?#7|$K5OboY@(h&N-z#twhpGWO%89 z{W~l`%IlB=KcPtB9sl|jv5u1miaJhNeQcu8KERJ^J#3EYj0@EhlPY;;q=cpasDUhD zhZHwXFaa<|l#8yXxyfiLwmKshL!=oJ?V(}I#Fqtc^mkzh4z zY=B|)s7nu%R@D$|V7P5^$;MRolPltW`Z}XmA5?xp3S!bH7l-K8t0r$JsX^(}t6da= zvpFsHPqob}Y(#jkF#`m+D?t*xbh=hC=zS=(5)?+a z_#soWfzB5lmF<#u!Oq9DkxmfcRyu`r8W}E2gvDGe>8v6K`o6f6Ugm|;u&%Kq_M>Rj z)dWwDGm8F$_xqn&SN^)KDJip9HJ;Y;n$nA{l1oBxoKu6)kH)H~PFael=J51OLQ?}k zCY=}D6)-A=G`DE5E$7N{YgA^o%!1hsBFrBwvsM-5VZb2ScMS zT=9BhKo^8#77C2y%ohLJI<1ZW$hVDtm zG~Isz)F8crp|O6(4XJnu`j{~8Ne*gvQf?%X){m*-w~d6;r$ zNQO*Mr-jfX-747_(nCy9*3>zx{1>eBIV8Bi9na}zG1sp0IYvbWs}*O$k?@qZ6sP1=mUfDCRq1!@hxWX-FBuD z`V|}fjmF+ZzQ~hlVTX{BmjB5JuKCFdA+P;=$HI4`%pnrle`~go-EJJYKWTxGDoTEc zBIQ)v0c7m~nqw@W#zN}^!PcpJ!qY$*{7W7Mt!j|FPf`#C^1|ZoS;*&~4BY!U0u0t8 zFaLI*1wFByDm-zKv;#*zCK4;AUKAbxMl#GbR(RD#_cHNb%czIk1=szY0WQZ8G!ji# zpPc|Ib_kkz;=p<99ugZ|!}tO>^GgDCw2=%$N7s2`*s_@8jBk_Of4H6BJ-b1ja5uZ} z-fd&_h1aj%vzGULHjd)Pt&BD#e?P=sNVBwnlg=?r#opE(&<4F(SupPb(RpnbpCkqe zN*e0bi^8)++W8Maz&x3Aiqww-!tF$~mAPniuJUV%jdF{FJ@0b3EEO+Cz zM`yA2q*HS6qt;Pz4`c?~M8$x;WNRju2A0HN_j6UKrZX~`(*p99xpdWd24mUbVB}1> z_Oopqj8aznTJF35zVtOSHdNd9{fRt&WvJ&iSs*EilCPl1QYx-A5X6LJMp9fL-jHd= ze@(>sw*x({g^7*Pxdo5k8w+El8^_LUFjhAErtl0rgG(1bwW~k9d@m80_jipB*rc8}dMPJFVgnB2W`^e0CE+q##^DHW^QrSZWVv79|p`d-hdm*P%m`NL=TNE1HM^*~)6o{yh zN@0>5nv6{~g0D-OEYRox&@G;K z)xny=7U^8GDG6S*awpk8niz~5?|9Ez*u4fyev~4IsW|BS;%14O18zV~*?myOgzYa> zmF=6+D7Y`Jr?2`|LU^`^mms(-zUf~_){^xBk9is1`ydRLu1evZo?0v3J@**1Px45t zRp3#NUzhB*@@NcZkDK@hVs9&N0Dnsh=dQ3f5{aUaods*!;=^ywJgmG!2K)ZRwo?%2F5V!b+18ZTJITQ_On-2T|Qz^&Xzp((y~ z(PNrUN4|1hc^V|m6l1V-5sQ-kxbe$9FdisKx*nvZpKybKN!tGp)#ij<`^VT36>QY~91bYY- zY>82ZsRIOs9WrU8?S9D7lcS+K#ls%Am4`#?h)Hu92S{U#RE6OKk{W3KSTTDCh)eC0 z)K0yp%7af%HCc}0%o=xiY!80yy{B){e_fH7kt4h({2fVQhaC5H-$|hba{!Vj5<3GLS)pjoO$w% z&mTVwGDD50g)m~4jZHVC!%)PUWzras9y)v$t@g22`)}LIzeC* zCIV+xtH8!T%~f#$)YhJ%(WW%XN`%mggm?VyGf^HV7YZ3J3NH#>fFTDV^Ndk(CR}hV z{;>FgHDR*bx-B*+kem3G!h_0Ut_D|x&6AU6JSi=3!92d``?q(VUo?)hue0(}CLG*0 z>5po&Au&Wv$sh;aI8b}u!ZaMG_&3BQbTi0XX3|~4MYEa}`?xn}mdbF;d?;k0>d9#B@*;47chbva@ktz8 zuWph)4(_3=U|&@@54%2e^gtv&yG5DIb9_Ov7fsKY3Hwes3^G<|qE58`it~TWXu3IP z^&8~oM6%R^X6Xw`-cONUDsGS8cB(>QkfiXsqyq%ocu~DYgXD}x7REHDk=w4Q;q?o# zG^LuS3rms%l?mGXGzL#62+qTN8XrPm^Jq*L_?39!)~q5<8y}Own$$U!5lzB%5nXgz z6rP5b3nqhqM~aRFHtTIw-JFS!AZ26`M{|QS5IN|RO9RIw1oh$1!`x3Y54ScqB>}< z(2Tw1o38#FNpfRLw$s9rZK32D6xo0(SNTRS=FgvrmO)+#J(LBXl0I+5DjZHPkKPt^ z-EXH7e`ZbF>^VM_xQ^H+hYb#k2U*#s2@Mw~ssGcAkKZ?c5JqmZbD7WS#s^o{^clcdAzELkoE z)?W@56kqqt3N@3>}1;jx%!#qgy4Xh3TOh4EC5* z0ZdMR;A4Lft^whI1i|5{nxe2B+#Rr{XnJDvI2n^q5sk6;gwB4eMx_fYmgTBOv>H0i zB-sTn`}SnUyPSo(GedcW0Z;bvZ$(}qo7~v@Ew}J~cT#eoY0shJ;=^ls_jt!twDg9{!9BIvV)y3Ou22t1L(qz%uVg5#hl)EPL>E}A*s5yh zTo$)Mz+m0`cS2Uk+BqnxrdQwM)%kWqkxeg{pmpyLd3^TvkVgu&_yk z_O&%%&!vVu27lS}ts#$!cTfGWY+fB%L-o$PKkxAm{*+5)hBy;;V+5L|333+7V4QDf z?mX{phEDmr@0}(qN2{-KZZzF?#<*DT#A2?#&F>Rv>S+Br|1Mp!mgWg)j)hHD0?v>q-Ok=i^wY(>#Dp;XnioyfG z5-jcGQmWwCT#V&gw(nV1bK$9kketG-2ePV^x*XrK5pJ6Sc=(do6c-x3Qa0VscK#^3V_c-c8Yp2=p5s__DtuK4&@1I;*jEI+=SXxZV!jgvk=5jDc(6jE|1f!RUD zp?8Gkyy??BRrkHH6|D@@I$jc_v@7v za4&`qd)x_W<}^z)MU_w%?joUqF#zT%cI*Kr+k3jpTK>hF_4cLW_bIc0#ZxBSJV~v8 z0=WkyVJ+%)?_Gtx%Ng}<|3v1p_`GnRx=}Rbu{>m7NOn{j2rph17egK1A;kcy$OD&h zlXng8ia6iXr6y!7fM7exZ0@(cF2%0=Xs)l>w)|7gmUCpC8)pcMEnu;ok^__37Ao#M z8SvJthh&SwH*oID@`H~99w$v{Gu2tnZkplB{8`q6?oD$>mu5K-N(T1l(o_Apyx2?3w;AAJlw;i zF~~=^2I^FsVTY6NX)qhaF(17f<}AHEXHc0>%cNDt4qDR#x!Z6eq*KP~M>P zgNvDXK|9A4v;Km&-A+)vU=-(HW(A4~5x;1E$C`NK*pG%jJZLdspq}pP? zvxk!Jrbq!5ml&M=)<#ab@)|Hw=Bkd0kcwYN*8uI){m3R+6Rp8u$O%Qpb`&TcqaEI-d4j2zP$ zpD!<(f8}LgV37Cgx?kU{TeCFy;}T3WXm8+-X1tAt-B!aHQ^#fu+g4+gUG9W`J0CFr z8)i%LyE7T@lOwN;CAnr{Nj{_GO%ypv#bt!tqqq2_ahJ_aCO4uNhiXcqJAsL%h*J}I z$qR}~pm~+4i(Va+4J%P^=wg4(dC)2Q^5hQc_&lBeu_>E97fxwYY@4IOPFPF_Y4B#P z0xR~9$zVmpi&{mwqEOh*R7`zNb8^Y^SW?jZC_bZcx_x56QZqBRvxR zz4ATbt&&q>Cr9p_2_3s}VaJX8|9(>NvIl2Z1F^#2FkrvzJEnlWUcE1PQMgXE(f`it zi|1BGpM6uWF6HjwRK8W?4@@3!4T#r+wOQ}2bA6kbqw{XZ)Q#yDR?8N3kYAXnG%s6W znSWnHa@{zDUu&@tl~HoY6c*X3)d84_$xdY(eccbc&2@B(SC&A^@dn~Y?P04}GPD{mY3d#z* zEXseQAZmZOR&jgAXERp5q2ts7xjRyD=S^=7Hb^kKb42wS*v0|xf|!D+O_Q~Xhnzy! zw;Vc(X=h>A);P8?J*D+Gsvn!J%<*{qB*4n3HY`n6y#yD-^j%#WhR86w=q> z|8Ec!Q7qtsti|;7(EQ*d^H#&LHt5Wp#?XgfeH?i|Qd4hKYfd2dWv4-*CRQy?oz0V1NQ}Q#GntB*`+brDrkdNcotJD1sE9rsY8*0!*DG#=Ct+jx!K*oEgV4Hl<8&{!> z$!9?pzLlw(Ht3V(;u>trnY9lZ=GWfxQI{UEg2sf9pMO{LkHy?wP2UP{0`fSl<2IJq z9aQa#eu^xOQI%=Xr&4$*qzc8cIM`%_P3|tYKCA#WVmWasnsD%6uBrKh*?KTJjDXx_ z7m9IX$2ZA>hH#jYKcdJNR9qW>@x+C%pQa7;l8G(qk8=L?@wZmZ&wOX++c)Pe`$p$4 zZqEDYjJg^ss50r#{LtH3Jn_10dBkaYmyhxO4C?on-Y)p&`tP*3)`)afldK{lE2=}ibmBFVK0PyNyU%G_7nlk=$WB!sc;tzYD$}b` zY%GtXqfz#)p0xSY@V0U@gOWMbkk`l;WpE7QE8<)go^x9!!vFZ{pBQT2~Fcn$hojjA>lizDl$sVQk!bP8`9C&y2#=#*`fj7wYYQ{%$QIz4^L^&ffy68`dz z**fi-&bddb-PpIeY~kB9QSy@%1e)S7jD$p`<-nq%!Q}{@oCWcMk{F-~BQ=3FoFc80_>0z0fQ$O6{CWUU+X7hc&`jfn>xI#{MHaV6pRw zzBM!StaW4Tv+*1^-l*GXmTi_|Lt|deYfV zh2`Ds0Au;AE01ggCjRyBYOO^n*sMiv3>zB~7KcKYL}?X^{Vzb{y&l5cnv^%T&)Q3F zf|P~JePZi5wihqf&$;|qbm`<+woz{EBij%_Dv;cpG&sxjZR4jM@Nnd&uyW`!A>|b?c-4?ZhIve zB~mw`y8b(@|5y@y2H4|ZD8^JB3^F9QeA1Y|-yooVRU*8|dCK*ILRcNMF*Uj+HaALR z;^=Qvd=7clL$feWyiL^tY^lkZ3BY?h={77FoPzNnEHB(0Kt-s50Yj5EV} z#+@>sFCK)IWqIxV$;Xvb&HH3k(()>@<&}xz9kke0@1o@S6vS1aRAZ%AH^~bg3f&W~ zSLghA_-obQD0)w;$Wsi3V%t6HL=Qw>i@Ykv#3SBu^&QjIawx$-&Av@RIX}+j-Xx2? zue&Vy#2Q`J=Htacap;ca(%Jt;%y`-3k#&+ReaW)bIToncK*>`n;J%L2(fg#kNITIW zMRJ~4lOA@;_Xgd@)g*e~1Re&6n!?ZVGA%GO?B?I0sdVTe+ty{)DuK{C#mr_0?8ebp z8||`fAmV7C8~q_>hHUHTuH5p)h}k&e#5{?1!o{H}AIn4k^@C+U_{Rc8LWp51GJs(< z)FuJAnwU)dztZ7muMHW?{5tgPmZ7m87#wMyl_lxgDYE=V7 zb11Y$Q6b+F3$BG;t_!Q-oeJ1Wu10BgU@qTCtZ#63gV2-7vU~C^jL67?CnGucCo2;& zVTmd!OKx6cydU;oARAtpeJ$t!jc85LQSxkxY^CCs&%YwxH^U%7(zt<`Rg+2rGXw@% zGABczS9dXMVv{5A9PV2m`Ja{K(CY)v%sEF_3L9c`Rk@y-z)1E)axRdpMq8VE z=1h86(%y#61&t~r;#q8{e8}DkNzi@;dsiEYR6gf!69Ve!` z$G%xmQ48LOLaWj%ilMkDCn~}7GLtD^CeXC|C4>}nTIG3AcRT2(nGKS4Q0tw=d6NqgR!BLCE`Xd zSugrLCPao!eJ$%BE5?f5&FiO~;|#$3HbOz6iCiHY$Q1_XrgC#ZtRBluI{6z!yTq$_ zWwE_6S)rFjeK7`^PPvs^7Q4$=7rlO36CY*nYI$Qrh3%;uGhS>DJ*F!&T0htkWVSg4 z!#^z}S?p|%8&?g0i5X#YiYWOmisVyqhN(wG5hZ8*mxUIKvIOl=h>ev28I!lEdgx0L zT6qWeGQ{FrI46W<{<^T!!Y1hnpca* zVZ+A55xe}TUzj1&{=@HmkF-vNisg~t^({(%gCYjtEr9lj&-^-OK;ba-QK1|;uxmxM zEBa#kXO^h6iVkii4O!N!!AV>LzmYTn^F6d%g?B)NGgE}QSD-FGCcDVNG4Rq#QV<2y zxEYUsaO~2=z$H{?R?HZrTAXD=e>dLH}q*ZM3yW~|% zk|Xfd?~&f(^>UX_8us`?mM>c6mCVVC(hM49aMwVQ@*zcOAcWJlPHW@0GY{$Z@P4RO zv$142W13#HB)sVQXfwsi_i*2muCFA_nQCjU;pD*fBZ+;zZ*0prc zR%pAidt{@&u8n{D$4E7l$6dwC49<{zR2I96cR^k(S}16ewULj2K8SZ#(oXt?>8}@q zdPof?&vy@}AlPLw!Y`-V(Utbg5nS`hHgFDX`Ky*^UcAE6f7C#hxN-IgLVF|p<@J<& z9Ys>8xRfwMbdp?y1yLs7Y}g|utX6zixRTLM=@|D4Y!)DJ&1~Bk)iwx5u^zAK{hDT5 zYvH{jStqRpjGl`0-j}Wq{zB5@eG$qSG{;m0+(!S4uBYf2WwPu7&?wKj`c&!9KYaMI zXR5KW2c!Fg-r!|kgL>Gi&@^U?FNlbw24NMYrf_o4bnKU}n!Pznv&!ql)PeC(Pqb^w zpYcCjJ>gJz#p;!xe@yTL>o!9+1Gw)Vk6gB}8Bd!M$+1#<#UR{GW>OPp|>U}Xi@l6(TRQn z*-oQ`n)TC~d=aKXAw0nx<|RspBP~u`@+yp8#9?_wY9jZ?unbap~;$IG~1?XJhCV~cQ3ujYM58W$k^gpaj z56g~jWd^-32UpXDvhvA#^=)p2ca|LcffjPMak4mlF-yFQVlM}mOiv-_AXk~kN%3st zHwx3DE<2C16~SN^sO)h|;T~s#>*(%>yREfIY>0UL{W|lP`!c<{hIh$dA6qOMq)UDJ zyyL@b{q^c%rkC_d9?#G!R&ugD?@M?2wx~0y@^9Upmpy;`2gm1~QD;ov!)c)F#hvPY zG6bVkMC_loKd@9bAg<=`^4&aVA7^twQFMlXa>T938qUhljohQ5#j+BmUcFtM%;}q6 z!_jiJ+}m$;arz`ksB}~QklQZ9QFr)MO=G(_ICDClJ8LZG%Jw@xO7P2lW8!aY|G~Zg zGH-|8zurhl??kf3B7L`Lf;@f#c}&GMLRKY*UPg+1;sp<45A$ednQP(Fo($w zx+eO5^m0xfr!#tsGR1d!@DNi^uaXx*{ZZ}Ax@f(66)!bt`L}bwT`0sl$0)}zo8&QMYfi^lrusr({ z6EW8(&HN_WRdF{6z_c?hK(t>4d)i&xuINv&3cQWJtbc69vaxN znDaoirUg{`7Ki50Sy7qvAa})V{P(VC&7co_@HwnRZ>JO8Y*d`U?#FH*6+;!`Y84I! zQD)3lp`T=GT!6q`8nYNUrON!v>4k!PQHp1)u!~+nee%gCtEgd*0jMduEbf`P$*X`e z{loJ0a0uzoU3qO?WCwM2NPqan=ZCDx4WHecxo^zD7KI@)R%#Vhvy*uxLga5A^oDuI z_$`75KI&+#3dKo`W)zNDPCmOSG2ZPUUL5#M>>ja=mv!Gi5I)gtY}AI2HDsY18=FiE zW3z^mCsSl46?a0hC_Iq7qgi75sTyN$Qgm?r){h%_W~K4)@)M-LLO?eI4)XEs5d?T|-?)FA1MOy&#{ z!%R>egj?O#!(YKK{gHLimD~1D*(k7D7TP0T<74_aRFy*qT{5V$U5Y4?)p9q8lIIS2 zSf0y(NKuw(gLC6LqgxictkW|N%IIS`^DH~-w1BSuiZypCn>@Mu#wUmkR@EF;#@tNN z&6%s=eXdg`i&it4A@y|SyifXnoc;B)@05OB|Mm%xm&^=ty@T^)+F3&B$zv}6dGVv~ zr~}PQ*JJtd}tY8^v0a+`2pq64q#7+M#Q#1HSIa|4ToGhrfJj!WM;u)<1wke{WV$ADY zBh3+d^=@u8JrGk5m6}H-c*l8lwPZ-vpxhf=Pj8v}pFmx-XS=iC+;(SYwzFks*SUYQnReQ5b`VAJ0)iS; z&;TkR7X<`Vl)EV21q({WD-zKvEf+5o6~5<5LQ5jCIY8K>JM~vMIXUMfocDj8_j#Yo z|F8LexCu%Bw7(#elyZ0;YBs~hVT!4xNEKq^tVfd&>N>xX^A8DcCA|J#>>J1_qjk=e zb<=ubmm-gMlr#yup>hg~K2edi9_WKIq2a7TdIzY0?kp(geHLCC29%m}dO;7|SSE@b zWXJtJ7VCu@eNHkLfs3OU3SG9#kd6w9{}jD0?2A928Wwxu;;9c}xr5aNLp`qFxn!{k zMQQx7c|c4(Flx~BS7&8h_L#aK}%FN6LV z)P(DJsL|Tvw2^<9|Hvf|GD7Q=TE0f_H9#6j{(^pfo^nT|vAuDd`)AU`i1es&87F&z zZ?xrNi95sLh2I_8t~9}EQ%c~+By}vlAnq3Tkuo#TI6o?&7FP^#5-Jv9gOGoEB64yT{qn<*Qu`lto_^l z%3+6V?y*s&;W&wvDzf;fV6#qsM;Rx0Fn`!#pT|u;v|g8nALe&MKj)y!S`Y^wbnSxF zAzn%Y-^C|Q2ql{}9jrhxZN_hIzi-*k^nzrLth8k{ym3{U#ZL-P^^Et( zrjdHJTHMO449x+3`r+5FN=w|g%btF}E&3F^*GF5P_b}pUkU1w5>!bI1B#>(QhB9AL?=|JJ5Vqe@);DEBC$M{}IU76Mu;hEn zH%z`y{D!xGNNOC(Wi!9?EXACn$Z;wLXZS8?OT}EyA*W5R3_Fz4Rl!|~CPlB{!OMCt zB-Ms$Rjsl@z1HKpG}TY1!o#)7)2e3ZbsKiTeH}C|T^a;?=k5_76;uiE*iYr1N=*Nu zj=-=(oS>S%9CX7#Kp1~e$D19k!@IA#ouIRw1;D4)RQ(X<^=6M9Z< z{yrlIvO*Mhmn~8AoBSq1`@rF-q7@mP)bKLJ-GH6GhhJ;`MO*X}0fvW)7o~b;Nn%A^ z>Z$gGY=-v;?xyJXtA&a`{K{l$K3tdZGP%Vry2D+>lNDxbvH^;@Pmw+>?mcGQR<1G(8 z3vuI(J|OppC00l=u|~OH-m_p6qv`Z6=4a6D3|#Tk5WNiFPYo>(FsS?>Ie9vqi>$v` z?%x6R6?wcgsI*8Hlmb_027O)KUf>Xi;Joyn`QAn=jc`P z_j;U0S{d}P2y0`&D(hwXFqA$9SxowMv*V&hYDIIw7? z&k3t)ZiF<_n!|j(EKW2O7z=8~LCU?NURQer_(EJNL9I(%C=)vrK^T40_S3d8Gf%O4;Wh22@u12dwjliul6M-EEl1+w4^8jg4VGKEk z1_P4kcSN>bcE=5wpu0l`WyxeYHW%VD&3^D<0hO^V!$c8b^?Y_vxhj4nd)}e4l)9m; zT8zv}9l}!jmOLSFRRk(4f!3eA78o?s{0kSydn8A;$_pYI>Gk9uy#iPquw0@_kO^V7 zt!~?b?1vK=sEI(u>Nror8?2y09sRdgRs@(9P;)c>nnH5M%4c$T`3yBUqsqhgQA{aC zim~t_^P5+taR4HqAX;=Ta!X*7uz_EuC?q-ja-b#JBw6Q?>XODw;8pl^sR#KPen8iy zND(yBAU!B3gCp|c{3{W+=K<+bWRE*UaDmA6NZ`lfMiBlg4WDFaO$1ElhG-HHH0l!8 z!L;Kg<&)>$#>~tNMNhQ~c*5JFjp4=ukx?SBJ7_!e7&zQ8_PC24y|380ORWcS6TJ{) zkdq9AQTHA;CjtxeD^2bz>#!IjB}wTn{JiadS5xc`K88(Dg zLjIg@`Il$><&UOGNh14~2-3sh+KD8yI9=>?<_SdxsTi#4+y#nn$7ENeXJ+;= zLk|7?y}X^$A!RMO2}Kj-Vl98CONTJt5%-;=S1iD_a!J^6VJ$GLVd>{C`YcIPuW+oT ztA&GLpD?93q`;%&1+Y>M*bO|Ny&kzY;#m&ztiG zNLRE`m#{;KEp^a^Eqnkv3g}1O;U9A9_1d`bXxPfxZPCW>r&Hm1NXLiY16U-;uCOvi z5!7l_(A(TA1B-o-f3k$eqi|1?yu_)=<)+aupA=Aifv(BfKvc1_S;@bje%6wWoK19( z!yc0rQuV*ySKN%g6xhJuMQ>qN`j@$H2<{MSsqU|}|7)xK0{G#7c}Ok=t`F~V-vfty z_>hboRpUNvZ${$rW6BQJY1AdZws-p_OF3N5#%d0kdCI#fCZ7UkvzU1BbK9bk9ynIi z2TQFkbt|(zs3Nk=FPA>0z)JJB=r!a4KCMz8)n>9t`u}Dmj%_*-AY`pVBNw} z=VNs0oV&2jZQ)%O9|YoER13Ho_%NbBWYDEByhm|Meni#;<(He7BZ^d)GAY&~T?(vI zwM8cj21vIOm2u;o_D2r#H%iv<&dVD~kD|)C%)LrbL{bF%6`+v=eFet?8X#-9Qc&vt zFk%QOzq-dOuVovcCNlBc1|Z9$SOJQ=^ZPer=N+@80^sbNWTmM2XsKD`tsZJ zH99pitezyP;l-e(+-=ZoI!)3&HK?0C9I!iVO7f87gRUjeCjT`n=uTU@E9&i6Ov|hG zzq_47%Gec6a5zhO&dhQ)P)r>~s;QU)(RK#QzjNoN@r=ZUwRBEsvH%6W+EkattK?0< ztC$$7L3*<~i1=Wx(vY}HlH73L7f~mu0uG7Tpd&m@9CI63YL4)_Ly+E8tL#^9m{CF( zd3EwMDg4~o-E^&NlaZKTwZXG&L6vi@V69+6^Y-{aXQTa`02cKmWTngmBXO_zwuTdwgys{RbKg$d{O2MJA=V&Tv#E)>`eQI;>-#)td#G870(@RrjMOax7gvk_O13%)I!iW+Z zBW)wWwGr+~>^GdM+h=w+Y`K?aCrr@`R;x?yn%@7)JGRwQjhM565i9z_Sm)ob%ohPe z>UD_nL5d~>iLT;A+uSo}K>ciGV44ixi}TleVbgUlkg0-;k;Vf)thv1|Hxy|yIIKV+TbG-S@tzS=Y)TxZ`7e#G~cu1zq2(pEYUGOo+MXf{T z&D@hR%jx9>aZs(~qp+O2XF;=Dr>b4nrAT(d(r$Y%T*nv^8}m_Po@w8mzkB~Dp_Y)~ z?DS$~knafYfIMT(!X2(9JcxTj^Vh{iQ)UH5505PtVf67+wQb!Zj1m5te)aDv?^*^p zI133YA&d1udXVR~m)9pq;Kc%4)Mi<>f3-Nzw;j}j2V}?Sz4onud}?5AL59hnc}nW50rfz*(2|{Ld4iCZs6-tl372IGklUZH9^(im9YX1r*~izD2N~7gi=8+=4=)6$E>j;S~l! zalX;cUjz9!%_V|4HRF2&q)(nEs|+;t1DzyUb5K5}AfQmzF3Xbwk<$bJ#FvI0kcO(q zsZv!2wY8c&wVp2VFAd-4v7}vwIlPH^C=)`<)A%vlGlIzR{I-skF=nA@zZx3+#Y947 z9JWGMLS@bJd`XEj{?n`sUF&Rw&?z5S34h|!8kr%{L>#N>g2%=m+z+Oc9&HIM6g6 zaIik-FE1aWE%Uf9$Xdoq))#rcao&Mu!Un}H`Y^;o`lQ-W%-dlzVlJ)s>VQm59WUSK z2rpTHLW%Z@mY9#$I3QxzX=&u!?fgGy=2=37vyQBlK~LjlNRkDy(B5Nw__3{G&?Qr1 zh>)Rcf&{14!N`e<&dJm<&T)0nS4AJ9$jS+sM}Mfv~~47EiY ze`}R-P<)Ae277r;;;P`%@U3nGQv5k3%j5AtWaQbk=WkXJp|-mH+UojXnmtrgqu>*{gX|<7SK=cv~?vRf*e?@8+^IpSElo zl@h_P-}!mWuYdNR@BVS5>Ws5)2rPX^C1AeLOFgTu^F9U!1Y;w}us!H<6$(9W@(%@^ zCB4wGht&UPNV4F%`(7Y7Zj#+}Zi~*K(`3g%$tI6?M*Ucd>Jh-B0_RrF){5*j|JV!* zvqkOb7b62XxxcNOmYZ|F^{QpY&Po!jLwHwl8^p+0EXby@%T`k#ki=`3HHr5|n%Jx} z=u@f&X@{_K!Nw_C9gd^YW^(VWwX0db<6{IHxND1l|L>InmSnY@wH;PibhFjFXe~y=mNc-83!Y0xi!(o6}>5cj5&8t$Rb!&^>%d4gH`D;8kdo-)gii#i=*c~$DfIoX( z@aGgkgS}y~8yYMxtUP=>Z6aO?hm&?zyplX1vC%3|$o2-of-H-yheD({h{bD~g{9u5 z;Vlrgj}x8u)L;TQPIU0)ZaTqPb7w(MSWg%nkKK9Ln`b}Zus0)SzKP}dPTK3AY~!tb z-fN67E88vuPAqJ+MEBz6!uY5hr#%7?dcy9^+}X?J_HUE>a!A;Y^_OSbcgwoBF~(AS zn$zmA?+HyF$ZHoi9VV;Uc_17%A>g=T(Wp|cLjR1VIL_s9W)>8X^PVF@Ms;%C&=f7Np2hGhYom6|R?L$qhLYa+99c89ne zh%%eRpUJ!b`A-FJ?2p<96#eNwyQNt!or=do=ynIKqf3E(B+vIG(2VtwP9Dmx*v~A( zpTzPOPmNi|@5LsHz+G?d$|28h&oP;i<$v&8OEO*@<#M}3OeO;SFQ0-@Mq!dR`;Pwe1z!_l9{fbUf^27p816m`1XP_zO+0%k zridbiRLl;3vud3{A661o6m`h&v9w!J6x9}eP1>n!^Dc@yp-qyJBT*fvNyAQRo48g3^0BNM*!(>{o00BD&Pr4z2@{vTg3WLH8n;E+t)R zxbmdTs*{66h1N-=jR za*K+&xv~XJE#Lq3D z^ZhO^YGsV91y6tGhTuM6E^LeL1ySfMegiOG9E)gWZqJXiHX}|YguVpR=&iAV#mb6N z>T`GAQJVmFBIdt0kiBDNm^f^!n#@30PcbzVsT`?y+8w3=5hvuoX_x60@8L|>VX#qG zK*O%bDL$}~6hdMfCOSw+^#~JZ>Qq>u-^LgYEmB}Y66L!!+j-gUxVT>`GFm)Pz@y*G z7p2JtWWCNyuv#0-WzbNeb%ObJ0u!Zufi*U0Y}soaSeJ?qoczBg5Gk{jBJvrBtyZEL zdHOKLfEoK7gy-dHye_$3xSY2`ddRQRzh+LaP#>l-mf{zHVj*aY$xecl$8EAK>X>Sm z>!E;}IT|e5$>TwaNti5I>4pMg8Nh$=X-KKscr83-2mgu3ArO&m7V2SIGp^)v>5V*0 zVd6XI#5?$#{rjNXs6*H+za{SwB4esf)#2LA*Ca{$=(8SIWM@6j1$V%VcU`Gh;B1Fb z#tRxb;&$*WA~jelTo^vc-|BgkBuP-bvoL%W=$vhHi+9`RwpnKU9&lLi7l(DpIjePO z9c^O-9;e^?>G$h?Z$eXH$BK7#vVK0Y zK0#$hN_aJ$<%QD3JNR9S3%uIkLn_TRckH0V_hX|rsC}EDU0ZNx>(X^eC-!u1tf*=PbWd6~|<_r$zxYspui7 zs(lfe!Zqap^in1Z639|fz9c6UdY#s|Hb!li=~P>MheMj>*QB}h&5&XOG%|Nwjb<=U zJ^GN+aI6E_GjpjEHpvV-bb&6L5q$d;+Zmk53q&=fB{($-D~DQOy^sgZQ2e9p5iI^N@q=N9#!-#hbqx4-pKwR&48@`mkDnaHQGKi4``;^uy1i^61o zGG_n!cSOr>{^PKN58`U00$n#BXl(ZZ@*KC znpv8fz{}#y&_S1+(1g%K0qCFanR#*1etxN^eIaE#f}S!K>|w#;4BocCHdwaIvB?>6 z7$;VGFtJQILo(#hvtY=fUV>Y-g50_Jl3W@&-2os;yl(MEpPj&~jj^a)8kvUd3l8IW zjSC_yhI+h*{^>9N`z=X@UJ!^_p&&xiA#B6W_iGei3aRs6vUO&|%M)3_kN4Kw28r=r zef;Nz%=qMIzcS4(AFfMyncQ+DE6kQ90~B+gB7IcM2))LV-C=d!7$reX1+DXG5-aK! zpM%~-P^b&*0vWHWMVgKLsztlQ1_%aMmLT=U9YJ<*2DHa$m4^aK_^YEe&2oG$@exMu zuhr3pLaoNY;<-!kawW>^Y7E}&TezRUPH@!+7t$DjUeYaZ5EO=f0?ka=W3M%cmtlMc zJKuDw?tpTI=9WB6;jQu~f+E3fvWu7NT*j{_dqZz7 zY>VD)JmjGIoNOv)Di}Z8EYoOTW+H=b`?-HTGr+i;EK|!n`<9R`?EGF1!(+dh-@z)`ARUrZOJtOon`~xZ4vp?6EQmH?|Kw>mz@Dy?o9aCalD7c>9N>W~@{P zha-WP%@A>xVop)yI2F^cOk@UKyC76jGiMi6CS$bDM10m1oB}0r+n^XRc0o}ThHEr9 z!66Y1Hd!IV)kOx#eZd>NC{7;N(pa*5ZtiC4t|pYwPi7fdy`*H zC&7Ct2C5kfjP*Pcm`h&<4|>>Pv+Pa?&dW6r7#at_CWM<&5PJ3|YuG%asP21L6nlbA zh+3uXD<-?ySsxB#?W7sjswn0lMfRhbgfY_G6_Nz*6H3_R&^1JJlizAct6`+DN-*q@ z8G2a^(nj|cO$sbiYm!60RiViV#U(nP+Npd5N+m;bh${An+yV(=of@E-4?0LthGL*Y zLM)-9_^gPY%|mG)WZC@jRu4SAc&Ns*ABU@s7*r{aj>CMRKC{v>_dC6dkI`L1Qog$8neht9S0%Z?GKiXJ2d*-(}^KQ-!;eBb`mw zYRXsfa6_9Z15rT2Z3|*zf^0i4Z z2)2$XJ0RJ5)rlWhtGGJ-H4`)zy!TNDsd{0c(QXEf(-d=pB8^lGh7NF=coK3y1k+U2 z^m3?t)+(2~+!lgQgKuL3vH}Fn(RVRcedJ5fpQ#pNZ5ozNV_MX7%GQ}V>Kalh9OR#) zchdQituu3J+yh9fpf{>Z!#B9$WQB!ju`@KOuf4aFTF1|aY-6}sk$`HgeUcm{qK_}4xEh$5NQ1!WjqooW;I1tg7I*}1oHqyZVW7} zc9%FC`zG5m!egIl-@UA+o@q&Wi)Q`6gop2)rrsnUIg(zp+kS&$pi1;ItUkiANx#8) z+YHR|mVsV3w$))px=vNIaBHZhlUESW%Z^*a0oY0I0<`0v=AgYA`T}xu05Dlhbpqw zo{0u_gUQ(^`3Ij$FI%#(an}8{!mEZ0BCK;Z2ryy|4hE}@%VGA@!!rl{tKYO0u3vqn z$vuJio zKh>_U*dZgmv%mePrdg@;?YG_{?Hn$$`rK@~yhAa!D59rg%3&I6q%~TYgF1o!_Yr;6 zI|*}QbX736^x;ee1#_{YnpvOvH}h6_wt@(B2K4o0y&fMJ8#V0E?4g4tQtsSpezOPE zHaENV^Vd_GzNR_<{R^*Y)>9e(T(`K%2^FM=9WKLMtW|#E-mlo`)9a$a%os8)74rM& zL1mgSHmGvJKuA*HDz|Kys+$z0knh^59*FpahvenQq)EWPv{6zK+2pi!X5zeI2mCHj z)3t+KCn%dR@}0-9a;(rckxBBI=ROTwR%n}+^7aqnX%pHWe9z}3iDS3M;&5sq(+r#I zDP|o-)>1M1mG3?xMf?KU9`V|!Oy?qgjeECplVtl;(GiXsF&iN?>Wrz`+sUfenJFJgf4>5L2U{!n@v(IDD zHCcdh&^q5L@CdcaWYX*ld=SdB!eU-xM0%7Cdnuq6KORyrxDPos@fnSU(t)kQ*2p`8 z4Nz>A5tI>>;G9dNkB1FJ+uTP!1zl1McTP~#$ma2RrZZ>t&2c}etfn6B8tCow-r!r- zM{zcWVx{7#N7m|*0I2L&Ug9Al(|vO30}F>=sdCO(*hdD!<6n9pt`qE#9tmucPnmgP zi`#Jnk>!c3a7*oR*ZiyH9K^;m;jYgGwNRte0AI{b1p#}&?FJG*4920`jjfe6;yvQ? zq}X{%lZQC4wxft83?_JZBd_iFbkiKPd}YuPGVH(#Sll&;WSbc-)=|t_imakyx||Cs z?3To?5e;VVrW^~#A+rS$6CB)eoi6!F_lIsKMEq0TwhLs#3lr`rF+;=-ih=mYHYz4h zbWI95AzBAD;X0qHuy~Iv(p~g`OOp^|_-)bnPKU7D6B`nZN1RsG%8;g^i z8rR6PIq5HFfAX>=haa2ygu|;HD^!H*-R>(|<)6{1bJs$HYg=?2bN$t3VN&qmyv@7} zaJZ3Q1bgkzsR!Iv31UG`r8m4(0o3IZ(7YEe6@BJmU$l%9DF_}m1JC-{&ek9ND8yt> zvV3-?l0A;VAvAi*uBMnn6ghyceCeV)vsSot2+I@=9zza?0^&Whp+)436dN|$qC4oM zU}X9^BKu5yT6G_Ud>@5;?4IGJQ+>?f&;Y3!zn0GW=2hthvRj%W*(cjh29;-|NaV6^ zZieK}qKza2+K@~5D_kD&_Dt5}JApet#-lrV1Hx(|=k5dfAFEmxS-P&AUD;T1Mj!Fo zNSr8BqDdw@0=h$z$?{NaWjEHi<=ab0$=b7OJ7BF{!9M$00e0FiK8)j2Cd2Z?^~o7B z#Lji)@RDY$nSD&6m^Bnh0CG?u=F>urL9s{bjQ&W?nb+R4@04O?lx=~Jl~e7y`?6bP zktHBFYbdM?G4v%v4|~1@iH5Lv>3?K^ZCS3`btJ7`oRx)`e$^%R&r*|xQ2h04Eo9wT zfp-osVRxI^feeZP)NH}#i+m9_T^Lz>44pjK#e=HR)qF#9Ppz<#&g9hsu_D&RVlVDg z0Ld1^@RY}6b*v}v4eO}+&Bbq@n`=VI2OoA;l3g#%V)v*Sb`DSstZVjCF}t8`aQB>A z(dnReWsW+7E_H8Hhu@%C|j56$yBH0#%vm zA-X;X3)%7|`67GET(AL(g)OivT-X8S^PjBqnrWJS+F!o#2eQ!-O!Vl`$WDrZ!V4W0 zgVVjHL)by%B%jD+t6`d#4=S-O1sT`sC8go@UeL3WLb}DhE;Vy>DwOL5m1n5WNSj=P z{PYbR|7bjYt;yWOYBad(2L5^9AJ)Ax0+YYHokPmlIVs%z%g8x1v^P*p9Yw0Cn08r% zpC+5$84-*xRE4*PQX2iPUQiYjktM^V}mi4O_Y1@5kse{8}<}7IT>kQCzKKz|I=V$ zsm6Rc{^ujKw`Hv*c@Jl&B`d5#)X_;$fcN9VI)Mi-6M zDw;IjK97DraM45gi)O8}UfD>1{EgnZs7{S(8z@!l5E@_B`(758@Y;C!l8?pEAxImP zRqzwE5d$(K7z}*6+uRQUvpvvAqr0?&4^kOW>I9UhOMz~vTb@NUdanUe>Q@1vxo)x|o$QJ`UeYxLmX@zU2Vy&Q&Ug=ojQsJ85*b@D# zwAXk2-LH%s^eY`(78>@I{}}dc1br&w@#je^^vC6dwno~^rJk_i8NY3L=9#RfVd~Dm zRbR6V+He*tRsu9h5aLZ*X7(`n}6oTr$NDAGj5Xq86Uu5wZ< zYyzt56AG+HIT6uKpCz5DvmSfII=6fL%)m=LO_NhUKhH1E?~GFq%sKZ%ZYUZs)F*fp zvPub^KfNx!E{D~f&>ZwQuyo!-=-b#2C3H2sTVbo@11>PtA<9Nti*UWr1V3jhr@|htwQkMw25B|$Xk1)WOAje>d74%7npxZFWn@V{(8p_xyYnR*eKg3* zjVf@$*306m1$B#7UN7pRZ?oVTd!%HtKfd(9L%q!fI-=iVy zv&9$N_N#**%-^EM6_TOTBOx$O5G#5Rj(zt!)n#$jBBM%^{qSKLwDH4))u^*R*g8x~ zhj)K>iD|AX^**B@T^ydP;?1aBA5shm1oYx+e4S69y9VP-$hCsCI9erA#??zMK=aS( zpuzx+aWz#Pyw;;xUPD&8Ac0aSzaK;Zvh#v>fmnra&V6N zke>!cpSl!hJ#H>Ms@C}c+gJvy1#I2Z-fj3-I(UH zMLi1m68(zn&~#hQD}mOPn+so%HCIGi(Y1;6qH%bdv(k%Z(08hk-cjbxee9A;-vE7_ zc2LPdi)OER%c`ZY)lyl#fSul0we8TymNH+QU0qp`^vV`tKxg;NQV9ACgqP77l57Su zSP+m&cG~NXZMN7I)#C((yC&*NO`Pbmw9p_;^9Y>Z9wD%%0v8n0N~gG1`{*Esh)(gy z)j$d}mo9eiqcPgEA~+ka^N?U){6ukND=chsgGV1ULAY3-z{W^$P5VImfmC9e)Ut%B z*GbY?brsyLCGmW-4Y2>GtSUR8YI6N4=9HI~9F&&+7&b<`3>GP7}SX z5d;bkMzn;j5}XD0g`;8ZQ`5ga9?(n>4>p}{9W=MsH>|W2r{S!Xu_iv5cO6y~cIuk*PKoKBy6 z023n6y#obfBaMtx&|=Lm^S%eviFsjrd3OVb7HBZaXXJ&_hwb#$sZg}JOMOpf5FSM# z#xx#E0A>SYVj^QyQ#nYV0MV5JSv_+!AA@8@pzlae0&(VM1;?6<9H z-y??|$#t`d>?4YS)yrurrY10(PW9a7s#PM1KCDOhxwAW53qv)#cuBnFe&=Pw4j5?d zfO>{bWry&zXR7Cl1)s`u=f*iz_*AV0XafP;zK*xbgE1pNL;1Ta?7Zvrrv?^;|eZk7J=1N$K45`eWUias&Ams z6mZUe_uUM2DnCo|?%!rQ(-ZLCD;t(tuB6!X*>V@{#LBv>Qr$;N70_=j->Amc&njjc zXjdBrVE+aS6D<05(}OItPgW8-&7dQW;>z(b5n*r0pi93J1;fF?MDg+6VLSK`?kc3Q zV(BOuQe5QM`wlzc>MGNzIn2&?eEc&CyBUJ*(J#o1^4qn^U$Kuj?a<6>%kaYnp^I7j&7tIz$@3F@!Mk% zw@Q~ZDy~VF2j%hl6?GAhB^MXzRE>(|ayy8E+5&ttA9?H-ll~MW{P@ofmNta5v0N+0 zO}DVjyCC3%EH+95O|>d*&`qVCC$_PN!Im)|`%L@pWoO*t_vo*RO*Ujz`n6w@M0PfW z!$mO#W;SFS#iUVW6GWq3GU!|vj8@bK+~OgV_HubMT}YjX*cW*VM9lu*Rq2o-CE_B# z&*edc&UK2gWeb|jjf8zsG;;OkOFs{^T*q?OVpv(38fppJWy!NJP}1wH$pJ#lolv?z zMMFDsE8(||qde-2so7*_GuD3pweu6HM&Yo{uu_e(I}8ey^4-!PfY+kZ#QCGJbOqh( zGU(PLoN5%=h#@;OK|PJjCK3M|HYNmHcu=M6(JMH#@Dj3hav_$9fyA%(^ zxh`dX*nY8tbO_6VJ+v0Mpsu-B(}xyj`0WuV1^=a*j`P6HCV7c_rLZFM67M7LJA%{p z7zvwiwz+M#c=hznn&=)@4=i_Mi5qs#{hg&h$|fSlVM}7gdqpNwEY&weY-(ocK++8kW2BBROog zhApQUNDLIAI((PnkY5WB`SdH-GRIW4K&Dyl*GNBlHI?6~Xo~#PS&!XrY5sUnYxH&H zZD=KfwU4k1QZaq>DQJ-&IS`L14ZzykO5s`tU!4Obf=RYargDote9IDPfV!WKG1q z;@^^F4tM!Lf#s-*%S?&^_alvpiG6K@8wNWdPOgPQlvT>}Gdoo~+;hU3<%PU#^&>hy zurzGY^^qGgDCksBs^gqMP4ZJe=0MpDk0k_~+1pb@#xqwh7H3b`PG^`~0g%3w>TPPLtSv=~aRfOT!ff^=hxCi;jQ zeNKrr7@jl;9$*Hq1AIq(#_jb)k84my571;AX4BZR8WirjCpZ3Y?{dqkSI$l|R_a*0 z!p?_jaJky@oul8;WWV0^_iMj4^beoE9`}YFvS-hJv~5P~voAC0@3KSdkz?t3&pW>@ zHdF!WjhDu#m7FJ?iW%hTakBvsPoFdedsqQt+R2#L*3U4_DYyRP@?Xd*N0Mtcr);4Z zFdG{{Knx_)u@Mq)d%g%GN*%x`hly@zo&_4EDgg?AP32(NFa%%D7gKQSM?Qkp5O8+} zDI~fLzcgXv=v$xqkWLP}q;Y1A=l3Y4mm;51F*%`Y0ui`a=MG&FUY*ePeZw7LJm}Ic z8+3ts_WHm)sAp~hy5r=zdx6mQ_WUD~G@0fRuRt{5eZ_m7&jlY%56Sb}Dyvm=$PNVE zSAYcfQF4p7L7AuA=h3ao7d6s{oX(RPvLp0@&v7S+0zIN@=A?t-OD(kfU1a)|x98{5 zC)7|RSH#bR>&Cyd-0xHWR=n~NUO98YC&NF(-+1&TQpf9p#_}Sslvk5q)!dTjLFfqm z>2?`DUMA5YRE6K+dXwJ@^(N~gM8bh8?$zPXw#yVcI_rA6Ph%dQ}d`Rjkyj zX7YBblY&uVq}xB&Jxj99Ez=o;BTzBmJeBs;Fj}^PXvEk})$b1-emv}AvPKo(_|`eH zhQrn<&uj`!rHAjStHe8{lgy;F$Yti-e zXxTd6Us#E!FJYHGtc=F=-d`P8T9Oi2>HQz%mrISdHdb^9>cWd6Q$2?q*7)2F9CD}> zW=J#q^}@F3bv`#2_CdExx!Z@DBd>S+FAMr_P0DNC{ut1F0?JH-E_g(T@ZzGglGXDv zeQ-)RD_I)##3RFZ_w2UlW2#I+wxm~??vupZA;odx9X_qWFUoW`pSyg;zSF3-2!4~apy&Rp)<7F{=|YEEY06$yx6 z@p{}LR^+=8Xi3v}*XT^?dQup^-!F@A_xWeMn4TD7zxr(JnD)r_j{lQsCi>ouvxM|J zk`y!dbJ=tzb~<@N#nh8J0W?MOTIKbS3Tc&B0q18PFU>!TZj~2Av_#(~yLcDGRl)I0 zf?tD2reu@kX7s)AEa`2J%%GKykG<}PCpZo{taMCW*bROi)^Lk{UG{JBzqKljr2=oN;e4u5e}rf(~Vq^ZGiLdD)Hw-7yOy7!fP47Pw%Mjlji^7$QO7qIFPveu>0oR z`IV(wGn*n7?*3)Oinen(IrDNK-6u%k?RhE1FV$tG#|BB4xGg$s4(@LfKJhD8bcr{T zI`3lk_YyS59uv zTP{E9qQP$UEV5pJiAvBw_>%TxDp0}}AO=s(?&z2{?eaf(pSP4>wW3?P(xb|`m%b`( ziN5r2SEX1a`3duh=f*h={2cXOMU#A;2Pl;W={%MT@*(E2*`qlE^&?iz%arE(3`l$E zdVzg!^tf)Z(Id9&jkfMWj8W*qU3+|*{PF9SrBF}1#T+&mR_eA6%*lqTlM4|1%N8AS zDh#_t>I3pb59S{hp@NZS0E7Yz#1pvD4XqmZ&%Qu5qH$bswkwcL-^=R!@wThL^Xn#$ zcyWcWi_~#AHgnZ1Hq%ToXDD(Km4mmrU!d!SgRWPkee>fTb*_W1ae~eTD;$BC#N%H0 zKzO5MmAox_>CCMG2LdaoN?>@a0#e&P-?r${M}w}ztU^JTuqdjXxvsnyjb{{g1qYHlMG;e_kBtQRX z;cimmNY0ww-8zb?rpO^wywoZU1fN=dmFk2qgka0u*LfWE&YfFLBi&mL|2A1tA<=uG zG|NU{Ft`zdN-8L}0Q7G-8oBg!=o>*oKfHuidBPVZgBo7P@|g~y&Zl5eHjU&%IdtCw zd}qI`!7G=}oxNlvKQoR$5KkHB&9{7F$9g@^Z)7};)$`2V$8kqO^R$;u7^;~2{z_8t z!Wf$4W{<)lia9`$a$|KE5_{Eo_X0cOkV1o52d#27silwbV?~JJy8-w!3xZbbe0l}D zLi?4PnuS|Kw*u>z2A5J;1kgb8q?#M<1tQH_x1IF%z_##DWtkuwhl{Yi{^fpIq)C48 z-Fu5C#XQ;LA(#;QO(&Q9{k|$2mr=u#d$J_I-g%$SUKMdoSnQnTG9c`NRKqSBX#+CA zQ_w2$y^~?7bFT2(UQGh|#C7~5;md$IXJ8AW=p1>`(dd_i#{WUjy|0lidR5dqns2w8?_2Ml~rD=)F4cdhkT9Tx~u}G z^Y@IUbKphRc|Y;!3fmTtJZpl!$=ElM^{7m`Z(`+pSzW=75s=|-Kx6&m1f^x@|LK4X zcRem^0sag+O?F&#Pn?J{n*%O)UHbSRML_Qy3V1wrLp4N0HG8{kdDL3LRN0tuV0mot zP4Zsr7A9*>UZsc0n*8MKv{sVF&Y|V*6iUE<9o2Y}Pccxdx|512Tzp^gr%L$I7JVtO zfsZIU;B}9`i@xHwE*RSQrJseLCOh0?LvPP-fZC0#5qU2^BK`c6q%=Y=%u*jx@8MSo z_Q`a1tX>@lj7A&9Ne0iCmn(no8;65TV7c|(J9|lI8q=tt+Adq;x};r(#s2u_F{;EU}qZ{7H}tUtTY)jagWKeSBy zkqJAxJEA9~bu1AN4$njVW{Ta3?=mFA z4!3wvkSx38cPl7~*A109HR3(u+_}we1L9sf7fwAYs1o2OBj5I&S(_PbmFuAUL$GvM zt30i0hGSrA#CG#Sm>{*v>fp=b9(t{ZR;iuw%%{g555*Tx$5(AUC z%@`OtVTyiR$G}*I(ecyeU$^A&=WL7H8nX(qCloVCk%v@Foo^nj=!ye2KzjmoQ(uC| zI%Q^=H_oO#;!TWZTSQK%J`4oE={lZa=Yma4mGhN|<#Hh8m-mNcL;JNB==(R!0C`TI z<#Npf@dNSF=%vxd>+~xt7v#gDygnd_*B}M?@s|o^RnA2u(QOr60QrrM(dkh+^pWsR zWmga~{oj-(Mx)`;;G4Ly(I-QaLEnNYAHN<3|1{S?2&+?h#qaif%_?QOQm5JlcB(#5 zr+N}{KV%U6uVIH13j9;>=qhLYgV02QjyVvJfZcjAYQv`e=e%`t%bxA0*I^uXr>*pH z*UE+zAo$CRUC^n@4LHZ z2)JYO^;Ay`c(-i2%TsY%Ek48gT^?(4;H(_ck{=#;XvuBKCN9Ncf7%KYCJG-KD({RP zyAXfGb)$C1)Pvh5XxoWbzyi%?FF*UuS*K|xGojwG|9$er;RQiE0bRsVA)7>sSxu3Z zR16{@HMBgyxa5Bi{+oI;NElk<7DUL|eXzb|;+IX$YyJ;EfOcE&PN#D4%4O%S>nd|-S4ISHXGHT5+ zvN?H>81qW3a6$=7|9SX#CY=1f`R8-V-Layz9A2%iHjAJrb@|yyB={SBuF}Z$y^M@Rk2_6YE=plzf**W@CXnE^< z^$0Hun%Yw5oTF2eLk^pm4HDQ(Z<~jEw>hnr?W8XSJ|xS7bUwRWlLM=qH_cn`*A}hw zX;ReEnV~nM->qU!x(x_6``0;lxhC*mX1C;^-s8=k`)@PaOh8F!^| zWT&B*qVx4ji?N&HEPW`TGBgD)*cH^K+Vs*XRd+-)-Q(UCeTA3DtBus&``jzDwLe;W~M9g# zsjf{>-<2t7F$BjlLGIG%a5_W`niLt5%+O(n%uxJY3g=aMWx8El1hpZ?6OMWt z55{obu)|LEJ`dcggV0Te1h+8Zo-YEk>b!bJ23^Ce@;b%49)jm?lvG4ws9L9j0N#8& z#!%S`h8O4TVF%;C8}2*lThKt9Cyf*3s=H~<#`j*@=ifKW1VX>xRJ}_!zc3(#9POxe z)-H<4roh*ZNfz`=ix&4AvQ~P!gU$tmrRfSu;vp`(XbozE_A3hcyI-mf)~N=iiJ@gq zTIFuH-EOh)4>?F;MY$8@Ef{n5GVPq0C41JDC-&Z3k(XCmc0F-6xnQNk2_2*(b58hz zpl+l8pjU@0*n@-qw;*dcAlvSB^3`+?4I-jJTJO~)M>So1r0wlZ!LW>P%qKRUww!H! z_8KF?IPJl^@BZ0R#E7#9u_8+p@Az1<#&0*#Lpy(mFzf4A!6msHa87)V9w67G81R4M zw}Y&Xx;c;rlk9uW9PSpgDdi}|9HvMu z`Zc99)&yRbR)YTkwf@2~#efuw6oOiT`sT0)_R(Aj*y>iqA5xa7+JG$l5lQtt0JZFA zqyr&M!h;cKJ&r|e^H~~-UQt{0$Q3?uDfc@iUg_AW98zMr)?VHf9+qq6@wPhYBk-y^ z)mlNBD#3Ar25Nl9#}WH4EOj`%!argLn*9`0Mv)RK1{)i8 zOK-X){A0`F6$^?3EMKEvyOv4!!TmY>e15h^=YmdUonhb)1zeif7JXi}L!IhUCcUr7 z6zGLV2_8A9?1K0Jww)9QOb`|pgExsNvAFWYURp=V7*iCr!t=p(%kC7;PEl6a2XGmJ zs>54ln<{CuEPM7hJ1-x)c?WrQvx} z!-8aDH~pDsPY5Z`yz<1}WQCMz^FOZk``=YGd^F&W@*JH5F`7(XjvBZZ>E(VW)E{}T zcWsM4NwVf-^01DzazTd>o#mr0IX>8;T}^-D(i)j7OOa&qI)qnvUH0t0oe;F_n45^- z=k!mOT#Rf|*c=WrT48QH=Ta5iCB{JwljLK z%)n~!dOK8FXcAu?xf-?&?XYdC0bs$-2r?8lVA*gckiOLi;_go6Bj^fmmM1vdd1=iO z4rUJRq>jJE#jMXmk!$mQZy91?V@5a}61EbM8RYAjdkbPc)0LXy&|Es3##PQ4&&N^? zf=jE+a_NU^%EXBDC}e7c3a}}nR39I9EC$*RFTEW;+Wo()4gzXq4O)H21#zxW5sY=a z9|`*0tDG~P24qt=w;0FLW?0x^=%01Te>qbqwIpg~V>meMe}K~3sJUk+#ef2rj*4mZ z=v2hcZk2yVcPmpO+M@UKcDtqWJA`-MxGMedhVme!IFsFS=w(5fk`J?(6v;kWTl8NZ zk}J|4_cNr?&Q6KVFfp5R3@@HF$`gKX?O$BKYuP7jgI9cJIWBGzFsVOm(nUS*t$c_K{nMa2vTH1~;`&zSON-bsjRviJ=-iI7xs3 zI*sw%b{U-6B)jR{EXRW;(~B85#J<4VsEtX4-3RaP*lpRe$O?6wwVDJL!=pTuFB%+k z={pN@={jg($x~_KU&^K9U%DdArLRCsnIC|`R_uLF?sF6 zro&`4J7dCO&lQv>N8uukVm47^Bg%#ixU?*MBG}LGRK`(ByhhMJIWXsd_om<$`EIvN zdQ&ipFC|T#SH*T-`pPk#v;#U$1pe$BCU87_^PBIG`mq9M9EQL(GYFic7~o<&Ma6Uo z?}-+j&rQrj)MHYuIZ);Y5B^aBV_o6X-fpf@=@N(I*M6KkyRkr z6ObfG;@yYBS51qmQ}sX9f^E~%c^pGM;Zr{TyVH_4Oo)iz@b(W$4Lkpp!zr7~W{5aT zF{dbU9IM@qgqQm@(rcVMl>@VCXC3xvk`MAz!mH_G=Pn?o>w$9V!yX4_1BnN`HqtZp zMQn;ddvheP5M=(E zX*jH-K&R9gxzQ@|nn@Cz#W02g)%D{VMOJeVOMk_5__)5;|0$5|p2&)b!~Tht6_K&! zdC3xE9|-n;7@FgeVFTMOs+_TGyIuA$L~Cc=lrP7NH8{RJ)vjAk@5j5oZ8A3rzpK1Y zPII{O@s61RxlAz^DRKd(l!YgG>7mA-EzzxzSg55}lO#~w%jZ`)Hwllr><0*BN{%Ud z+fjTgOx#E76$gDT1f)4?HK0@Z;d`;vb{daP z;ynO{GTY1xYs8Gtq>MhxCZt(+X2!}*+~3lc|Gz80j2?}q#J$WrUj*r8l$UNG9UY%wM;7x@9pWXEQ@qI8g_cd&F(bcG*t#b&|(3Gy)zcebAtn zCf8JXbwkKr=aU#(9bCA0(51ldQ+e*(etse_cNg(b0Ucq!Bo~OkGDEu|i)5QY7{`#< zKK!H2z?a+O_(+>hVP!+6IlL3fvvgoM3l=Lb$u8(JO=O@mFisTjf&VYTB+j7A>7XjF zwfxl!yXb>qxHW||4Q#<-)5uQ3LF~X$vpzL*iOGQ}^**B@T`$afDc-E>`60#Jqew3m zGZ?jzk8UjXjul57=bdCq{57~kr`i&*H>h29#1sG5=)DGrR(S*Z&ElvVN+b$|H8%=o zVIQhiiJ9sfz`=SSJQ|#?^mHj{mkqjTXK0LE_ppIWqFX?1h!m*R^yT2*aM+0%%8?W| z@ZZBb!D~GZ0@Lu4B|uGF2r)q;(OMGEP`Qb0uGqMYVq6_GwnswR5+jFLN`sC#9)K>^7H;VOdyNurbcJ*REZjM-EtK z1fHv?ua&s}#h`X3kAFy&B}oshrH$RIgY@+|sdJuy5hY($3)pH}JoVvIBQ0<`zZt7_mbd+J5DR8JC?kmtc}vB~KP4Ne1bii|W)E zLIVm6zu2H0rxHa+NHPJ}IG0W)I|87hf1$DX8m0i~IaA+*;(e^e9(F($Zfsr6hPfhE zgw3d$IDfrY?%XD)etxz(&##B>aD{rS&^%zlDCd>XnNDkA7Rp6hT-a~?=r7;FBbpSw zf)3$*=l<|~p!P<-4gA(8)JTs%>`W8q++n^NzMf{<-Fl zCbUUp{}MrZ*v(NKPK76#c}KC+nI{w(q+&LH`>J$3N#HdJd)%>ax=H>7h<_fCtI~ey z;JjM8#C;dtEZ!2B7ky4VC~Avt7I%xWE#Na%t)g2z3?UoiQ5`@Av=>Ty?|3xQ_-+T( z`*zSx@hrF3WUUFUqFt!3HKprAUyTDBTD3(!0gR*Z96pRmk7; z5@@Nzwe*TwKpzIFdfc0!z4jzIA~{R+veOWSMowx}U+E7SgsfVk&$1x=D|V{u13snq zGI@T7{2s&S>`_$A+(jelJ)Uc@xEuiNQ{YDgXZ<0(IMf6|#h*3XND(^-ayX%J+RQxG zP)sF7Dj*)|br>jmH2wT0;avqVc;AwDLOq0D2rR82T>$V=Ku>Koy-%{yCxOBYMhA_) zyk7W;&>)ArXXa8T3>_HN5U_!zQ;7nCcSEYd)LkN-(8>LXE_Tz)+GX`1aC^_I)O#|5 z2zElym}ZR4PTjuGvn(BLD;}yIRuUL~<6T^Y_h<=HI#oG$!fep5ghpk0&}A}il%v3tw1_auQ?g_({Ko4!GBSA(!R4F?>C1G4EAq(j&z zNZ=i#&$(S8X|njhBLPj~9gq^sbgFjk73`39@}5YGLy@KK(~!i_NrsQD;CRYucxngB z(f^9Wt((64zn>TS9TRMBFR1&B9CsvLW+U-2#hjzcN7mBuqU#|`1&K_*_i|5MUsXBZRMtb}Ze!$Hk1A)BvH{^QASXO7 z?1Q4rw&+~CK-3jd!fR(@gVKFAcs2<3N8&eV;(sh=Goq6>Tu+VP(>uOmh(^C{eYnO- zH*$BTikgx7yc0{K1nqEGhF_&$t~y_mD$7^5%ML;WkPXa65A6iDZA0q>K46@uvO?=L z^`fD7+)M`NpX#<nSU~Y3by0 zHvMVEZAHBS(6YxQ6e=qNBc`It^+JG=+;*z%hs^_R`xs5ik?f2|$Xh?SGsk2+mjA(X zEy>_;Ai2uS928MZAw}}2m~HM?A>bA(S{~KFzs@TRZy*oI72c(}>w!c(*>}6_U5J{6 z7l0DFJ`5XA9(t}PdFnj%X^54j&Rq-KhBx`GbdlnoOE-=0?cg7^50S3Th^qUW}*Z(0Jx`a(n)snf-iiXrAAlkWPD~GA9E_ z_GnwM4y3m~o%hq{o#3!>;d-K1q(VvN6T!LQrGkpcGQXn5SEOBv9DaXzTXa$6S;<4s ze7{w0#$z_in&fy)^J^o=*msJW04gjD!sNcRj*79?UM;lx(eqxW74Nv0&Sa`pJ;2sn zOYfPPEGXh%RW>;tQGBe#a`BJE`;{f`liOkc6=TK@9A9yQZ$!H zYb&ke_bK=Am%0JZX<(&Fr!L~3bZLp+>%RWgJaUdc>4MTYr&Z_Vw*}|C+oBJK?sZRd zzO5MaKSn3y;-Wcv`>jFT^2RpXZJA2F^YS>c2#xV!!xo@ur7yWiGml)HYDqYv`?CmmMg|eQ*0(iq^j%d<<43 z?vV+qQuqLB9I7JXAXkJ!_fTna#_3*omdioknmPADUTWAOTaC(ATBT7*{<1h>!7aI7 z2DK<4Jg?Z~JM3@|M9J}BO#{JoFqkH|TrE`O^%-jPMry}*symgJgPZtRxd%DEdVIq)~+&{*xxSqZHB=@G4#D`qzM<8V~p zih#fa$V&r{V?dp()H~i0S59rwn`hn($&yyi9)KwSVUKpt3p01jUMAItrAU?sort{0 z9|&)h6ngYSClu(R{OUtEFHh8>Dc!G|q~YBzB@yLv=z9 z)B|gHxv|WYGqx*d`Ul;lQh|&Qghx-=d*HlnyiqI1#*~1#}2z; zhg#Wv#ihVaFX5S*>*AYA&8is#P+hmrBZ;@x*;qS=?;Fd4HF{YxSjB4LJr`ta)GG5y zH{cG{0Hb&NvCa2Z2mV0h<%XIEiKwFBSW{&aQVUVoUDoYi| zosc3HNyxAI-3m(})j%_y>Z!pKyXlcr6PbG7kHy*QYVJ0$#Os#*f-?EkPh6jzf%avx z)y#EGqL?)lNuXk|f4*IoJWGRg*`+{$IkjYSBhQ*G5Aw*dQ~8^ngTmP&h1`EErd&;c z`1xh>E=h4DKv+3yEt5qtItmo$W71W(+$)9myiNywxz!R~Eq+LDLI-{$eO59Y)HyHT z&-lAhqEnqEd4Bda^TrXk1tU+H`Y8YlD~xd02e^NAexGG;(+e_ZS?PMh8KYU=6s!ZT z>g!NNjQ;hoL$mxGjm`aC^eM3hTLGp5+a{Q@gZLC++w>8Cn*OagOSUu426?S8n=Ox8 zDjJY>(knyL{88vSgMJ7e>;S1Jb%Hvcaqm`DE+0jvAO20Rk|hYUu?12WVoGTGAQtvVos9XVePUt|DU}#fot+i*T(yb?@RJv$VMPZfeJ)u z5G#wJ0#4L!bEebT&YU^Z^L=NgGoA83XWH3%=H&Fu%$X@#QBm0xP(edb7Fm@=6qU6I zxB~@6aX|#pT3i5;qQd{alV}MgnimKscD|Wk_079HiFvNv&;8uXbvbE_?FUV~Vh`+W zhJH773UEthf!lag91#XQ&RSKapg@3LuYsh<+E#;|sTi{^TG?utUkxutHF%BJ6kvri zn{}^<8EJ$D1yC*Fmr{elIrX`#c}r4 z%m95aBv#lEQU#mM2Vo<2HRO=78e~Q2L{b#DjaJjF>}<81(djw>?HDUv$>;P2%}A3QTMJ=WT&!#!V0iOzn4oIqx%iPKjVlF8L(AGcExLwM29S~W{6_+Dopwu4z63aw)ssr1 zW%er9IInSrOP1loFj+H-5{vL*{9J5&+gX2QmOuEd5mU?a54)4kzcJQotqGsQFeSf7 zkpYZ612@zQt0a#YK9CFK`oelzQ$}wG0$@#t>vmU6c4pBS@IS6dc0nfKVkvSJ>Dj5! zK5b!ENv<(@(2!@~vcg-j2Dn_lF|gX>p&M!vKjx;YQR1yRUiZOr4a=^P^m*_$$kt|f z3&?Y}$(x{h1KaPj=p4SLOmsu0AASA<#|lm+e6C~UAn;*Tg9g-ybiaKToWD&ca5Osu zF(;5Hi{oxpHPbpSoWNYgQPtLYnkrsB-6Uy|^at2%V)MB{akQNge0g8~`Y$Bn8v{NF zYLAI>$5HZVimao;u}B9ib_`4m2BsP|Mu&wRc=D~5Sg$9KV4bVi#nHbSUx35PpxN$b z9U9igKw={$kEO^4^q$(q#Y8Vdtf!i&v+ow;@-!nZx9KEW0t)nPOrNs;tGC~MxAEy) zbQ=!ync1SFc5nu4zzp>>=-M7s<`WH4SEX^HLeha*3b;p*+OWd^aBv^l&CxO4G&V1_ zC^T5eg1fhN@dGPzkXQ-{vo5cd&|!X0_9wSIw`aXDJi#jqwfHU&5WxXL;5*!_kRH%` zCw;D{QZTbi1`7;WX|^XG);tC#+Lsqy`(X2IqqCvz{!l|!*zoQN(g0)Dig-#MLy`4V zc%(doZs8jKL}L0jHU2rXXfVu$MVOd)+N^q(6(*?X{(Ju1zpu)=1nR$Z%adI?WW%y- z?|S<4g|R>^2Pf;7LRN`-xb(mFzj|zGi#qxzcV55sE~Y7?Q67;KTwI98~yX z(J$7C_k&=KrpTwqEea2A3D!fXW}ifJl9VWhxDoC%;+IYi23EB!`oYbcmj@qzc62=T zhRv2Lm?6_ZcM_$kW0>OLY6ht(&(V8DT~f_upk)2(0!9`L7m%j)n)4?93}9k%h_L=% z#XKS=a6)FCzq|hC>qay3NH5f8QPN>zG9GJ#BA@!OQxAkyQKlkd5!N5zo+Lq)8JM8j z?unWMU{2Z;DWF$UGw%ZA1u7Sh+P$})Sz}9740(saS11jQbY8 zFP>%L4owzR<`)1yZEErOD$8$3%*0Hw=pzV%EKY{#sLPjbwOq{=*MUVV zxCeYj?JyUGog+<>$~jtOcZPm|n?9?pxJ&+OW+#cLuLiKv(djw$kuFBV)Be(#4@tr| zCe&JFVy4n5Idn^Esqj8gVjwPPC`AD&rCjAo{(+akp#iPF`7Me#?h%3l7sUd7J}~U7T1C30Hh9A@^kns%n18AnRA3BTQ7$QI%x74R4YUnoN8? zC5LjTEU0hbY;lWp>rqwFpSecx^MZ#ts3d*Q{UOvuMfjatk`#Q;{S&{G;A6}^_bg6A z$koN#Wd_=g4$&4!Od9S;aft^)%Y#;{FrNwntO)~FOApm_wjt> zSVbD8)gFg@Dm_o}vmC0~coLTX#GZJkg{XMq7}mJR4i$m(e<~Qy!SMBxIvbAH7n@j? z3`!0xbg5K$q3Z>Qqtf$Cqy0_@y+dfeO1wm4F zKDe@?0zsFod~O^$VD*HKh1i+ML$wBE9*=1rJ6rapjE)bguyI^%x2c7g3AaE)wXh`! zq(^wop+M#@$`JN(dsIWg$J|OphCiQ8{g;QRN$bM;jt>xDj8XGnfW-2N>TsuiEx18IxsB6)Ru%Q4H z=;$N^V$g|^)_Gi%p`1~R`o?Q(-x~Sxf35qM6|YAwYlT3_y0?!lE&O)%($$V}q<{Vn z@wp`@RVRtoD+9O*k^c7FlHH^uB-I=HpGWNQmDcN_+g&)u3nVP8?lQBjtfoZrTXCALjiq`F}{6t*P?wKR+ecpXd7#7*c?AXk|lqeN82C& z=DlN)aU=BZXN&Vm;WHNRKV{;=R8#T_iX5cEtHf)#Il@F)Jp}vvIoYC{k{H;4}* zRec@#LVDZ1)2WGr-VlcWSm9-&xw96ACVDS6=ged1{y(SvS!u-3{M3)*NhZ6Piw)bY zS`&1YQgZN^_fX;0bX&+a`{N;x{%%nZ(*tg;>MmJz2->Vf*@LiNcW6M;--Ty@arcc& zMoldp_t2txx`P_H0ThdFNv0y`wvfd-NW=jiVa|%e($BNfk_`2f%BKx zwJ(utv&m|cYStl2ewQMB@Oy?I<#1T>oVAppg0ws&k#kIbPTntzcgcdz9?fn0%gjlr zvesOX4TvLTn$_xD{@s9P2#9JGnC**~6wiTQbqW2sE2vAj8lJCK75O14M)zDzjPUxx z;yESW1L9`7c+N2zN?=oc>XiM`llG1NnjsLX=%nL0Suey0bx`E0Rn$vYNkCQ)6sAU` z=Lr_I!Vrd0vCb8=$>>^cY*?4<2#JClPm|N+bH~peEBqVnuSpIBzK8!Htpfi^?`e7^ zP0lgQZP#K^wEPaKngd0%He1VT-V99OxY%xw$7tJsB)w;Z-qOmjKysO#&9dEp85uRP zS$8P;O^V#0!m)L>n{K9uh`#F{vrf1`);JA`-_H^10uyBs)Fx-p5pI%TzUzVbB)JXD z6PN_u&UD$AJD>}HO4TR<)H>HQ`f|New^MF-<1w&a*ne4f&Si0Ui(y<&%|r_pNWfwj|VW4&t#|?^*u6C z&V#b`4uip3AeY_1>7*}+y2S=Tcx1sxNf&H|8N;hGeddmhKW6sc$98aCn3P2iV9)~d)S4l+``Q1hEf z>l0Nsfq1W7ypvNFI%=1=(2(Dp3CObm`d2^rX@j%*fsX#y`Q76gAig0XL^IqFoeo!+ z2z~{WK;)Ayl0dG4s%xU`%0J*1y%qecD(QZZz{sB0?v1V#Y8LOJ@y;1Q$)rGHy_&tp zsOeG+JwKje$Y!f<%#j8G)5q2yjM!C~y$GMC?aZSA*jEKBuyC|l{I zC$W6|kGA>&yVD4pH(lij-2} zYi4h!E_`q8Y@`xg6&T}oRCQ9cO48ztXVa;Y<*UC_y0pkOm2M493Tu?^0UM4 znXthJ(V*1m%>ZN@k9_f{7@|Mmv92X6_*wMUc|9r}Gwh@bjAI%>!CaC_cLT9Vw{&yJ zdEX(IN-~YTOqN2&oJkme{(l8J&t?2^Ny_-1=&z@=Y>frt=vYX9gAlD}z z^l=RH-!mhh24As7R$}~$0k5ZO1)lc2c`GqNnt>{C{jhL+&nuhghz(l~Gaa%miWY@l z5(imYE(i*Ow2CgMyBQM&f7&-ZUDqu{&~&|>#p7HmpX=)I$T!uT{v%my!@g;ziEp}v zk|$DRBNdKNU+Uorr7For$N?yz=iNVPs?HZ6qEQ%0Ho%cGk^_NtBcqC4<~ zWO8t#N+=r+gPk?8A;&0r14Rx)e-O8t^H6k=dh~ajx|ZA{Wv~8c(bBFTTwdCyKC^t^ z8xjA|xOBuW@#h;~+49}^Nd(c71VYUaQDJNmC zEX2$NehQmY>@j8D>|FSnbBm)9O=YjY*-X~63%A;^iODv>SqdfJLc#om7Yoiar`?NX z8suGV795q1+8OwMHu+rgf9RIsrNLBv4nH$+=5{x%al;-0lX}AQdi4K{4}RF}rp8R5 zBb7c)`dpDrA&t)9^ve#aV*VEziQwzEWRfQ}OH+Kvym6Rd3p4ANf9>medM}X;FF$5_ zi41$+2zJ-hs){~5FQZBg&i?~Et0URXSHmSD&15@PYYu)onf zc7FeE|1X~P9B(nGxnhl(<~)=+AqnW=;QXMID&TpWU%^@LR6aLHn8TmOi22jMe|9){ z`gqfIhj|=KWJ5mv(JX2_HrRwgY}mS(sTOE(t5Y3RYE?zhd)OuGns@8v*cV3ZK7C=B z*X=%HSFYINR_&HQql&PskThK|v0xnLZN=N}kG3~U0xpph8xDs+0)GsNQ!XXnMUf0D z9E-)UM+10zx@7lhEc?{c|88)I<6f2xktWF0Ka{nG-f^t-G*CuDq8F2PJG{z5?{gkF zBU^^1N1v_JbUIjNVPQ4@W|kHUrjzw8>@3NJlXr`sb#loFzarOtoGVO+CpMGC_$NAV zhyJk7UHhdm%+a}4nfxcrip98i!Wh=L#0nRb(&v*ujSsZj?8e1R=-lXg7?So`^9Oj% z&clwUrn~3HlDV=B6PAoLz4x%f#H=45U48o5XN1EFj(S>Oy@4$~`|WqTWI+1?%H{%5 zP8^8?SG5VMy|i9M^EGP%wJK~X>LfFetF;Uz6D=jvi;D51u|vt?S3mjL?~MM({&$=t zMGUwrm5a9p_DdTeBGxa9mhX0H;-Q=}3WR7Cbsj~2CL>(sv_cbVD%2g}js4K} z(kLw!AiYmAu;esZ2 z4St|b8pR&)rB;>40oJ-9ajDxuDSWyZrcsJra^3VbrgCuwT`WjfNH%qX>zNy?3{Q;-w zY(~e;@;K^};WT2m&Ar!St784M9A)R1x9lsh*;kEgN^t-=5{HyV* zMKx|pY$n|q9(Kk{!$>F^6{pdhlUy-Q|MXo>!6Qr+jGP`6@k{Q9j;IP*N z*)_XlrP3YpQ9GnA>XP-jmh!-1i}61(XH)0~=U$0cTrv+}@#tkIy_?)doTRHN{3|#; zu%f04@nSOXrX*`VjEpN9$^=$$j**)_O+X>==vl3z+T()5h%}3?@K5m4zf*!&^}H6h za<^*kDIRu5j@m`~76;)CWvUHsE6Gk!%-HIy!H=H9$Ki8;$N))%C$yZ?Y62&)ik6(3 z$Wp$a7|9v0U2L<4&WsxI4ZjrdliwVrNr)+rh*lS{3E>ZIH6gf+U7tonr zS`zD3tSnV-ch)(4*!piBZ>{)6&(i&OOfs<1!I zlDVGfd$#o4q;X>D8%**(b~sUF&HZ3}U6;+`#7w1DsT+nl!4b*h^wGTm!_WoUBR%DI z9+GOQK-O43chqjDvR)YPUT%>AnB@0v`WRbum_BX2SFF(S+|^izy75G>&s0;JIXc&1 z!I~R!bl&5 zR8z+V@dQ)|V3&d$ek7QN;N-f+owU9VsD(jxD%?Tu4*MLW)HSH3KIDiSE)+no^3m_G zx`Wnmvq3Y`V7~x;eUfBxD)`rUH$K=8Fzu3I46~cpBdwXM--LC^j?UZfm_A?AE(Q%S zS4ieyswYu)jN}K6Nc9w0bG6V*lPFv7bXaxNbsim_Bi;|aK|q7QO_?Y|k2->n_h8#qp?#aW zPdW<18TCsuIlG0x>AAQ`f0Pk&YUxKG8_o>`Z~k^B=K`QH$$hVL`|=C=JEqCIW(};T z8kOlmS;{K!qA=sS=Kaa~T(X~Ctk8xH&uJ6G(?H1&Q>2CpzbE~icZTWYb&@`A71<)c zBQFuPs5dj+(mrm92wy+swyL+d6$navi|LJIt^K~x9jadU0znk7n}-@L<^J744P+%* zy{JuHOm9=gI}M0?7q+Q4k~r=TdW$@Qe@C7r?ok2gK=wX@KS=N@b`oC`ZB_lxM}N-b ztenEGZTg{Pg0)HGs_JmYc#hUfgMa9_SYD7S9FcBS>gFcGuJ5>F9n%G= z5r}$kQ^s?8=|AJuyrsjO2QLp1ybA|d7g9Y}vn~YlBGn9@%n_!}jdlSyt9$Mc!PKrM zpR>xZncK`whTQI+#mO!kmua$98IVe;Q0|a#T$Up1n|I3BW)mcD27hsDd}FB1?(WUB zdbW$#JLL$`pN4tCUMyrJ!QB5eEkP`P*cO@?79&p2t>)dJiGt5XALSi?*0sVg-IV0i zsNZcNJ@xy8WCZzud4-%iekJzlDt&`6ZicS7Hc84IFewjFA*fHf;;Lz&?*W;wR*}Rl z22+$R+6L4|RsoZRcrjUL#=n>zgqV)P3NOzwzy7x`pY>eUCypol_JyqzKl-~xT_Lzf zu7U`AJzc#Z#~ZKJ3$=>NhSuas>%K7mZBMk% zB(@r9*gG2Z?>{zrB|^8~K1Uj76VQ|yvz&KO@(UC>M}=1~9lS`n;m;Oz^(zBEDqY$N zEJD@XGu$h#N$wYgK-DiOa2fF|cgPp*g*YaJ;PtOB3Ojh)l>2=j!oCw1^(4oZ#cgVA z*i8&<I zaD{(A@XlZZmLgi7p}HixA}doh3pSweUtpQ)-gJ^KKV>FZ!{aIAP1j{s@UYz#UsYak zVyV#reedv3_K-vD_C>b)F9V}F){q{jYJwNW;#cs82$nu*6{sgzLZ86Ghd#*xA8fgL;G6-);92wv_3^ozKmiX+!&{hK z{tkH$4SlWLJ&YzkI7fJ%WX*qcOO3L?Wrckzy+^N4c*w_WHTly}%KBGf))1P%@U+h` z(e)SZ+_2La;dCs-lS}&8!O4b0FmWa+rj@hgpa%BGQFDgKRap&p)DEQ$b%A&*HbF#(q=b#yMagRBL3fS*Y@+^r1mAF_PlCOChI4oZ zB~+2o2j2!NYxpD5+dR!lQ6h9x^exMzwO)--X^5@WnE=|gBuAJ-p&^i4W=Em2l0!Y*w5L7=A)!{kCSBed=Oa>KIQf?Cm6qv40K z*<91dUO`pQIpUz?65NIxI`+U-YZkQY_vHn_ z1L92OjIw;;KTc$L%t;**qMpF zcR-lnkR^^m(b0JS0&Wv8i(bh;q1fVMeLKPqGmrfgR&zhuo6Tb;zao9e#pvO*zjWq9 zlEBWxvE7{_NcN0%nA0hF8b!2Jc&hN2TZ!xD1&{u2frtP`1+Ff>8{DeMgzEHrbXIVO zXOnlN97!kdN@45BtMsf8^a`wwBFl}=V*u$7CJPKppIaUI{TRv!lZQ?ZPm&S4XDTv9 zKumN@7%GO6ucydb=(RJJyl8GX8(I~tKEo4!;}%T&gwI*~qUG?wE@u>k@4NDR; zX@E<-jo<2$M!KY_ux9P3s%udnUHi%<(f!`)e`0U6 z67?@v{f(@$;qKcVCKe%qlE+gdh6=~{Mt4vgcUbZ{5KN#{N^5A5Upusqr25S`Ff4<` z%mafJES`H1?-R;3!eai`A2pEG?0gIx_D<7HK$JwuH&G;x3fFoS2vz~n4Hn^Iz(dn0 zC<+~sBK@e=>mjfn&1^>6QW&8t@pT(uy)&u)mz~5$Ui;ku!y2ywKItSM8X=;)R}c9_8(va2n8Z{^DEWPg452vFFhn-jl1rjtN366@^X%m5 z0#mUPUviRk0$WQ3eMP2&vi@P&28T?aE*WwYZ*a(VZI>OO2ZBMu%Rea$79e3YlOt>w zH_yc?`W#_|UoCeaIM$((U{rM7^Q(ZdZ;0pu8=xlyiDmGv8fClefpag7=WtXE*Y?sa z%u4(2xh+sR)vJ%T;`l9!s~oK2&k^F7EljeDA#9sU-v>fLBz@?lD&aINEZarKz&m_wg`KWpZ+m`U+gre`EkCyT$AHFO!asFLz)=2x!9_WHVJ8 zw;`W%4ERTyp-H4o4HTBV+awd3Lt=zQuIH9K5an^&W*{s(#+a}H(wN7repq+rW`x3M zo7BhNIYiPX>e04g092YRQiYTpR;^qrdWN`3~wcoAQ-*HcLT$1CR?2}G!^TIleZ3?{S4kY0-fvwFN&NCZim^CL> zz=?SiQ*%1uf4*(ROx>@}e?e-WG0b$DV5WtVH&Nsy6@HcsIYMO(;NascvQDy*l!$Hw zR5M3aTGfbM0rXved{=5XT_PP4WpZ>57X;U#SvGRE;VgR0x7|-kdU!(tF}|>}xowc& zXI9Tfb+BZgziyC6$<7Me)OUauce~flKy0VP_ZSLz=-muU*rSKr)Cr+j&{H93mG7cQ zxD6qH))X$oep@#1n-*A^#oM?GEZ}DzR(~3P`OqvQtX8e@I}CKw=5L;0>u$rIG&s&< zB6qQrd;>+IAdCyGh&{AskW>hCz?Fth^UQ9ljPqs@R#?pOblrM$W$w;kBO-phFIPv3 z+3l@tc-aDrGX@c#Qt~Q_ltbj^%?n_DpheoPWWgm_Hq$8R7I%0K(`6K%X;D`#Tl4On z*P~y#4uyvGuqf5jN9pVZWz;#kM=S7{iKU{E~Uq-F;y3SucyG8xM7q65qKk>cNj9^O{NcYz;mw*x(L@EwcEKeae#}V4mjS+R+32EhYShkSbvR=nz#X zM+0)*3PVbuDxrj*!EfTPbgHtx*282cp2Zwa>Q+{(2eo%W$?GzsvDqB&dzK{4CM71n z(o9OelY&5ec(P9_^xG5%ReR(JQ@r|UEH%?A*6_ClRxZ}6F3avPs1G1T?>{5EZfRoJ@8;@W(i*FWSO~Sj~s+222a0-`ntp5h%ZFP5wvn={IIo zx@>~?&nS5#MNUxR!}P{)r7uYnKKeURcH5zWflf#4lh&868_LledFB3nTr5aW@oE%& zxj@HHa~!6d6sQex$Y%wRd!S!C95y7n|E&SBPTmP#Z9m+7gug}HM;CtgvVQGD(vQgH zmzG`>^+*lZky{3L?SIk4FYEPb(@P~{LHpiin!B&DT(+k2c&gFNTQ`_R%c?qk?Vr+njH^Rsg?w z*8Fnr8h$0w*$;t?RfOL$f=sgg(h^rBNn0~J!cR+bL)OmjbJofPQIfqK$JP%}IG}PEN@vzBS zbHN`sP&y$H-mT;#N8^kn99Y!XJ+=zgU1E*Kgq7uZZuxNDsU=3!)AG|*8KlgH^Al%G zjL8v7UPqBnVdJ!5i#VM=NvcAU9GmF^mm2|VxOY`6=Nw}0dLAJ$!XEXm`=UpInU#9AA-Tvu>@K8mkzf z@2wETdaZHZDe7X7s2+O0KsTitWO5orhq+dVg6Wi64h+-hto@3W{g@@bx%5ZgMtAo2 zyK7^~&Tq_4y4u9~EvDpqD3VWwpOdd}X;A36F~S{luP$y-9D3mn5DWCu7yMDBF57#U zbJHhL)*;&bqRtmrDI^|B4Al8n(H){J`UV#@P%g`MsB$>Kls7K{^au?qg;`eMkhUCS zW0$6{9Zx&tK79VoqtCkPPK%<^AN2-fn21G9l2p14S`qK6E~;7>l=E$u-EqnQ{-;b( zs4V9g7+A4U5!F*O;LQ7dynK(uIq&y{^r)I7=_=HN+buo$Qi9ir9UgBF$qamSjH+=7 zUU2hT3MDVQ>4-3F%EqDkeEMYAl2G6n<%DXLXw~{TB_0Fw>X|;D60mSX;@rS=Rjss# zKFQf`i8lE(VdiTfdh&^)i1U6C01qhU+WLxL0Y;U5Yh6dqN2usJbINV!}`Eh$t_qU8%U2d z$-S1-K$mknd3XmVwj%g_-?}I)3)!Jca!+>51cDM^bmFv1tmCJ%z>B$+Y#LGW)lhE_ z-`g@?K8_7uY&W$FQqN<{axy46bXcWQ;gQ@N=VUP|^k|A*Am;?)JGU3zUUbH#maA9q zF=+8~VNJnz`c)L-2=a9ou5T3UVLJV^8(7EY(NfF0~MlL1wc3%d=#J-a*Ujx4{7gHs$AtxQ6c1{+CXThLzL~klQ$nAm^y7&U-IN zp>DL#RIT9Oay!f^q>quqz*e5@)xf*$ej=#BIsxbD0Mpc9Gx5_u`q$pK$CJoBQ+%&b zVm78SG?9{Tq)04sq}3?1LBpw6QaL}HUmT1^Xp8$!=XUY77joUFXc;E{DxYxH;@N+~ z7}mIC-jw7y{@>;w7){9^dj4>g9RJ3clFv;LbDol)rO0V2yj>ReBASslbsQI?gZ15@jnWXu!RUjPVk5P_+uwxczGdjRCqu*Q$Y)gedpX~}B-#!AaDUdzkS^%tRnfq7&AaWY1*+Caeu=c!t(v(JkfGOU zE2jInt&+XMI2;tvxL4Q+Vz|c@$u1Z98|JsEk4ho>3uAsM?w4LBy)S`kplAS0nxS0u?!tr8QRQbL zmODk$K7~&%kNI`aMSk}PemS>5+U$(H#EG(G z24%)EXM%c%bGw)tU*xJal(uRWm^Q=w9VAplsNcGqlyfyrk`oG@{a$j)8=tLH#sIVI zCw_hKP$Ohx4!oqvX7D9)g+ryef_;5k&o>kwuB zzdEDvU;gHu@c;hhZ~yR{a1kxHBNR2V?DRrnG*m@)x)WrL4Lea8Cad#iN}fQGcq;s~ z#}!b8Y8Qj*6IPgEF4JHnpwk##t6YAjRin{gtzL#BR%`3{hpqV9ys?O>T!4<7nohJ!li;hZ#SI{wL-&=x;dkEO_CPa-QzlS(+3nOup6V)4fEx7 z9x*_yevcFjO6(J5CrG#WhBJ_2<2X6|+x92zQ@y*PYx$0Y<*u1U2w;V=C;Da{4spICZ^36 zu9{J~%usdH2HC>O#SrL8l-1Mt`|ohw?h0}m444UMh29o8gAlP08^$01^v?g2Mq`JK z+tm)y|DVwT`Ca=iH*&>>9gr0!yP|$dew!k_RQM6nC+Z1GmS}3ZS#&Mv6RlI%2uJOP zWrr5+Dd@j4HW^v!&@iX=j6RA&zyIwGj+Iw$W3^0_P; zW0n2VWS2|+QIb9psJ@^i$FNfc=d^Pw)SNX)iv=j}(&=>#>c;Ta&zy&4M;2?+9Q&cd zF;RvA(Nv$~0wkTgBHg3Ztdn*z?c%-=ymwtlB1bn@lMQ$7aUPi46Qm2?1Q(1NuCVmM z`}9eQ^_nt|am@~OPh5GX<>DhGvO?W+*MFGc;bcVJyBElNB;JNmx7)<+O{3&milk8C zyXYcU+}X6Lw~#@8CvD(uiuE7hM9WW$H_gfPYmvkVha3~U3gFDX1t-V`Zi(XYL_YfB ztnl}wFk*$zF@8u^7@@lNUpoHxRrG^}tetd<0)jK`;&}IRC`d&O9t^C!pUiv-m8@}G zv=&{mi{3fHTcq57tFlg>Bg8CPFa3VcbUV4G^EW9}&A{p>S>yRDJEJYh`|{U+Arb8I ztTtQ@1cqiz)*_CQM^j`S6^^uB_@4&jav7=$dap-;^;H&3(V%~25hNyf&f2GRy1;*O zG(w{6^*5Wz`q?Dg$+zK;S*DIm~pk|YHWK%xL6>W13UmE47Vue$z@bO&Qf1Ley#<(8+{>Weh`Gnni zWWx?Dh=+}FU^h|nIEqA5;nt z-o?BTdG~{=UoSQqj8xgrbI4vBj-GvHVge3R@*0X%QsH;zJeb!fL+0TUMPW!4**dR^ z0O?g2_R?&Rx1p})lI<|T_u zoUtajQIOy~$ZT||q6Zv5Vd^0|cN^lj+u>RZQ{;*)eKG!3be*cwCB}EEX2wFJF@YcY zlv7g#n0d@hkYc9Z`rFBCenuQ!`~Hn0l5N8{syD$=86_{Jpm{Dl$vy9d-i0NiP9Ane z4Le>Eog&E~=#u7no?K>73VVZm&Fl<%?d)Q~VGzjf2wD&Pjza-mGM#WR_^xQPU(}p< zk}BLEHe$EqrTPUUc6WSl`i$6Z{?-Yh5iesOXNMipjTZ+0 z#fY7xzrOJ*IXzK@oekGF4Vd8RDkZ-{kq!g1N#yK2P9KzeCflDCUsN^IccuF{E4?x~ z+mx$k>tz+NJh)Ar4e-MH^A`Vl{moa^NQ?5Mhe0zH2JHfx^-Ow1wt;(99RXaD4boOg zZqQZ70#FP(3Dxwap&)jpiV>a#pdR+^g6AiG zo0+Xp*>^%wC2pqMp3>Mbt>Al#EgN@vEY`kuY#w}`m*$FJFaqs+xxe}~*=EB=3aZM- z_`7+O9HNeyRCp(EAh=t4IpBcLFsDI^0zvWqQQRnQ8hIef4XNh-d6?6qu7bkjPAEOj z4asH_e|*LD_7AQu$PGgEj;#3`_}$_n<|32K-|I2ewd-r&z4649p=Qo__0cOSy}C?Q$R~bnOds7zc85i}o#Xtyf$O7#OxE95@(4*rHJe{V zPJv;_CAN!uq;dAvn)WCC^ewVCPa42V*Y2IC!_>wFr#tY4WO9&QIL3x82()RAS)Ptj za>y4Proz*KSfEXPK(KY*TJIZzA<+QP&+51-3p)ktNV{K~I!2f->IR9#tLpcT*bg#E zo}VvlQR5ALvMffYr~#d`EIMXEA6){)5M8pIpa=5@7lALLUMWARY!$YsZ-`C{3czDZ zor_O3C`;jq3{|SIK+q!0n%^#74=g#5nF!9AyrXRqn#Oq~Y+grh`MF=Pda&e`|h-{um% zAd+myPHsR{9#agx$3Qbu?XlIjkXa|T?lLq5kUVXkr*xeiNOW!QhK(1kv)OEt85z2g zHwx&js%D>Z&%OUt@oK_%+Ll{)=FLry1&A>Bs8!FiLd0|0u!~Ed_532zMYI4*Ue*h% zBpMX>ZVep_E)LdLIbpU46`r+mUtilonJQc@1!)T&p zQL;qYK~;WG2Cztg7xU;2U0_l$)`Um!HP}as8Skz|1%d`qljNXEgHqiQemSBw&Umtt zF8tEWtxJJu?t_wc^QX*fK)o_s%($K_ED=l|AqqNu~ngkHX z1C!-z%h1~v>RdS$hEE&bE16;V+zPWE>xK*?U9t_%t&%ek)?2|n29ohQzpIP$o)Uhw z>|14Vm*v1&B~m+pL#jM$tRXad1&LFJW~};in_^KNIX|yfbVU~B zZ4FPyauhtl3QWVSj~{~7Iz@4Z6Fn@f+JzkkI0{gF*v7PmVpC6XFc!gU)&w>S(p3k1 z^lrpXr$UenJh#Le-}v}{DZFMs2@?!rg;T61!FJbrZ+_ORe_vHnsdO%XKzbUI*2v#+ zb#b=qb)e6{NXqn;*;OTlgFc++)X{ePe3ylV7q^XD;lNRjk3H zo>7w1{^br)l1`_|K!wFDP99Y(-eeNDN*Cq7;cPTLKd#<>hHPTDt+rvD6q*>Sos?Wh zk?lZAJ{u#t8&s|7Ag)ewO;0O#-pq~rpN<8{yU9xME zwcP!FR)z@7-Xu{niFs%D$d&JJe*Re}*gJUYD^I8(Tg;#D3*c6 ztjl3F(69o_bH~rsjd&Wtvg$+k4J4J_I%C7_isdHI+D*wpj5`~ol?kv1de?H3)yL$T z7Dc$hm1TAe1MRlbS>3)0UUC5$Gq=$j@Fj{`&``2Ep`tG&Yvc_CS2n8vSWg#Fr6x(+{)O0{*Mx11aB-8aOju+#}p-VA)7lRYN=CK_G)&C)=ya z4XGz}q{L~!5swmI2W_zV@t+_qt>vCG72 zBvbOu6iJ}Mk;n8?AU-|pi+^fR5pX5HW-g>)vgkVdZt)2})I4cXSe$e*^Zl_n%@P6> zi=n4On*+CrKZ5_2y5Z6mvFI9e)a3wFHNcvOJlvUqHNIG{ zkRTZdo+8R71t9b0WXu(IfEjbpqI(eI!ERQr{HYeMy&LhctK%5o4 zA-I~`2PjGA1jepMd(VWsI6bhxI2O2!G7RC+ropLOyKBSwW`wP6K~-X=1T z>X>o|)Y|A3#4snowb;OoR`&-dampQv7i1}8-KH?v7R}k1^_<<|@~A#ZVt5mwP7+$c#tr#A?6UJ_;fkS@pp$r)@UngU$e1Bwm6 z*^fT?Cyqz{eCS={-Y9bWN>6ghhLc1OO?ErCDESu@xlV;QNziWXbizKLUV2@qHf-~Y zDZJ`ANA0p19ViQeJR5Wfu2$>CtrKNy z9Pkz-JvH10Hcz=%s<{Ko#lw!VvR+vx9W5X6E%J+!r9hSUIv~dF4;XgTtXB8AcG8zQ zIw5AK3@Ixd1wRi&X@D=Lk#`_u>%4meE5(hPM>^M%1%se6h_S?7bd>X;Y`w!^@C5i| zgCTYXlHJ|o(ZscC~ILQRVQzovP)VUyb8t{mZB_BJ6#%M(4=^XzM_emLH_K4KX~eDxIE*V$8|GVEU%p z+(_o_%ip%}W{sDpW#iY_aAePn%5J|5W%YXL6wm<0;9NIwDkZH#gQhO9M^a3$Auwx?tFq3mpgb9UyX$l~B)DHjOBpG(n1xE0X@iiHAk~@yL%Lzxs2w0VG zCTU|jmko1r_>Hu2ox@wYpirYnis7&l=`J7ra*UKhAoFJ&BWr$7kXD7C11af|$Yzi7 zu7l00?0%6|e13f6{nz;uZPE*)q&;J$j8CuKN~P)5D&eu`)V3<}A%LoHuW*yrpjQ#+P#1Y`2n`;>;{? zV|0PnpxY3OGGiFXHa1D}7<_@b#%idKxyBe2jdO!;`ou6--Pcm@zu(Af1~YP51X%>@ zVPZ;%psbe`H;;%3g4oZ!-T(Vvi;X^y;Ev#TB+iC6WCbSvjgFE-D}FK+o+?D*1+*dE z&tVDyWG)iG)dA4R5O=QcW(iEhX+OmhePUtEL+f)Bk=%{>Ov z5EZ;(D^t3__iw?(PZ&da`qWMB4t9p)$H_wH@%w3;g^8KHb1{9yFPjtbLL;wNkjZJ_ zodI6KgTh{)YPXiahrn9eD0%4JN8frCNuSF?ORUlfJ;9u@2gn42Sm6{afKYaCzd6U< z=zMf93AjX3Y}olIGjTq0DLIsEW>DcNUVU@~KhjyNSVv-*?e1wDkXZs~Xcaw@Lhw)1 zJb_2dDF*Ub8FaoobP+AecSum(CJ(tM6jxk9uU?!=@053W?6e|u*rUb7x>x}xh9TLZ zCHBg{S3m1aVk`L9II)~dq89amrI&y(uuYxhd0;_|5W^H_1o7ND@+H~m9nXCL%JSD2 z_R<5v75;_NLo=a0GC7b~4I7UKW(N|Led$e=(WD#=|Jx=~WWy$<(Zr}9S!NFuX0lOZHknzAx90W*FaNj*edsw zFf9H~X7EfmgiTMlAO#eVl&m97AaR2Ja-;nzw>AYWIF};JaJy9jIAe*KKYB~ zx8MHN_}+S(b#u*h%dg;P(*4q+K&Zd*KMLuf3c(fnrX*{AqkSBBGMZhW32;Jj$FXMK z1wle+g?~|CfmgIVo^t_2pmU)0(OOe(oRrDW{N!iVk|$YNm09!$ts7?>mnL=hhZ?ei zT_)Ow!}>szIwtWLPsw8_vYraZ+B_^X(395PaL(Z;K*P1w(K6A@S%#pAp0w_{@BG#= zf1VK&ga0a7O?KF@yHaH`%SDtNM8xwz^$3hZad4sPJ_mh@&&2(rEuj(8Mrnm0)2rNJ zU|uyWCFLRapqsS;M6!}W;ki+;(&;|QQbq8w%CFR6pI-y8xBz)JjpuhT>4BSk*7#V# zs8|MrCwL(GOVbYtR)&PiuUzm)xe+vu_Xp3B&Fl<`4VSd+HGx|hbW82m`fXQt)uhNM;av!(!Ji$D9L1VjxYs54Ny9IS(Yk9SGNMh8(ZFK`%w(_ zH!Gm&D;!J(S6_`X{YUpWgMS%c{8Vc46+f;Ty77Fua|%f3_JuqU4JaIS{6M__tD6ulZs8Khg5y!6ep3J;03w4A#%cLvi*Kf{l`z`J2w2H zpZXS)x`EYi*LEvFK1zA{hksuR5q;dX%oV0eS|+^esFRz%YE$opc9M0VTx7U9V%NvL zC^8(ksWAwQ_m_yu{gWIYa&bkA3W?zNF4Vqp4TO*qWq5VnoK$+>f^ISPS4BFXgG>^> z8)%dWboLd@X+f3rOSf8X5AT{}^0sN}KZ++Vc9Z%+;6pb*l5ZVfsA98wA~QuPM?mmuoW5qJhvh@k(K8C-aM0g^0h z6(u3t;hABpi4WnoJf+#k0j%^tRwj{BUy}alStky{0tD27K&K;x4~?%qbOkS3eo$4T z>=pC}!)alWq8^lBN1*`|npZhkRF_I;(wZ3I_K?iLc5#=iL3Eg#M@`-qJY_Pc1QxIo zPZ@sFmsx>j*3VmmUYkk@rVVeR&6Hr4hF|f^TAQ+$S7wr>XlN=#&5O!`;T!7?xDqR*jYL}sE2NJ!*+&+3QQwWqt zI&P+F)NZ}gN!8~8MM~_KDOzl^C78OBf4DXk9wHkCpcx*bbXBQai{dILB21rXz<>@k zogg!H;zEq-;%ZqLicZ%Zf@~@B! z{ftiHWOLTPxI?~aPA)%bJ|5i*so|~WR&ryY8#I>L0YoF6fo+bXy=XI;rO~&rG$t{yAtXllUJ|EdLA!iK z;6?sAaUH))+9iLl3u+|d{Trcb;z~e01Z+Tj7}k|+{r*U2U%=GUDPG0SS7im9W~k0m{w5A=;UD!(gn9OkgnVnsA*GV(Oc*BsB}z~&)~wX zDr^>k@MRD5Wa|QBm|pNjcfqDBhrgCeV2WH1`{qui>f!5?%nCOXtQTWXeSP%FpJ9cY z=aST0&Wvv%`G(X%m}xjcLbGcvNT6Xp)0n=Wzji&tF}!}2~(4Y%u}D#9eA^Le1Cw=t~+Kb{tF%8|E?t-HS^92 z5}_`ta`8;AGGBeg7R>QiH_zY^E4wjkso$H<<0~C(77=D@7qDCb+YW)Bu0qfyOYn-7 zZJ(=E{4c=50!+YiF@s<*z7;!6Y&p9ld3?`^nbr)<+ph9HN6ta~X_a^E^D)9<`+VUr zM=RIt3DJ^rhjP{ZAP5pk3%oR?ZfK`>VgRhvtv6_o0~}g)jx3vvvFrC&$JU%9R;rPi z^10ySrjs>E3g)}=t5?7Vhw6niNS&c=+kqXr3lbb(#^ zuHr0cOxK`Wc^>%w?*`QP7Pw#^SfTV3U@uFIMNF&YZBo<$f%<;Vdf#5Cpf!x#3B)W0 zF-#1E0@W@HiuS1>>nn^|d}J{gSs}}I!<#SFRto1Bk#+6gFZ`86v8$S~-G3PY@rp6! zcUvhrEVi4daE#RA9!Hbw{Q&grAx_&aUco)!qvv<7@JGq}n-bh5S=C^U#Ti?OkA2pf z$G2BEFL~BmOOm{|h?{}dVxOYiKU3AF9&(I^FzN+|PTn<9F_bH0fF4__;!AO4SmcXG zRVA*2vSzw{@&4)E_f9-XY^U8|!zLbUI?uAQY*gq^zUMIBfSCE^A?dM!8yf{TnNi<3a0Z*e)3qXroeVh@Xoy#)<5ftSVnR2CJE{SCwLk9fg=1W z1T(fgJcf#eXnlPDOdYU8Yx|aO-+k8mpRt%JS$u@Qp6Qho1wN3)x{cUX@NP?E7-(H| z*}#u=!&e!6Owa8KS~XKd2^Iomj59nl5Mc!ns``0~Or@5xogV2dS_hpqVH-Oe{^mKZlI zEk9kALCR*6GbWz?5lUW1kxyYO?01Z*^Ux|Tx~>PVq)vza;5v^2XfnS4e|Emxyu5qq z;eR=|wEvY>b-in=*9K0c56DLG_dxT7;kiWFI*{KjcZdtkQexZj1wr`|ts-BP1~v2@ zF58s*eYb@c%7%iwq!IiIK{MSVO<$74Ss8c@h*0m5p^z*G=oJK-x2dcXY$svNFZ=0- zQKJ8qt{1yPO2V=`gza*-3d&DIsBArQ zc2ebvQnz8JIBqpi$2HP5b2SL6^Q24m=kmGG1pOvFg8AZNy4-)Qb?R{~g3wd!+7yst z_EGZ?@_J=e-T0KwHzXBoCIM8rIEGs+(DyxM(YNgrWtbm?IPuJz91GE6c2=!*;rP9_ z565!{nxW|1>N_aw58KDt3rytM9w%SA;E*disKm%S>S_-=;;T*`R{j*wsml0}!_qY~ z^F>?u-J1Lst@VA59oPIXaRL@c+uaAu`{R{6zcsEh%kvMrlh4@&-EB8fMb^$58+0G0 zxrF-tTj*1m?M3I*-Et zL;GZg3h9+0Uf(1+Xs=Z@@jyKdIx?|UWYi9cEPFr(AW=5#l*7;A@0Ol?Ni*z-h1r)? zu*|i9K z9y$CP-)bh!x!E}vxKb)TQ@lQ*&XQERM_LxzrtX50!g7aPx>vA@+YEZmnRK!0ZUDHG zpl*b`2N^Vu(gjrM2bCv9mkG)pX3{IUbslHLc+aD&$QPH*M1x-6L8VsJF21@rmHx~X zg%Bt0c33)VCQBTP7n7Y~R}y-8zq(ilpl2>JEz{-F8KcYgcJKYk!uP02r@NW|CBJK6Ck z``j-E1&3ZX+L-o#xSmN$YZ6l=s+N`uhl0<_lKGn#>|1bIc8BSP+(0tK78@n?a6Oei!z442vwMQJ z|NrcL30zZWy1qv|A^9+5Bbbu{6^S5AD~n-7Y_v1&Y@MysnLBrx|LvWbGK;sfPsX`3 zbEn`6E(i)LpaE1s5Kur=RMv{P0~!@YR1gHUD1stIh5!2|p^`{6CnVgo_tszGtltUe zd-8to`n>+@m}G9Wpi{YG$_)=~Si46NJwUKPUVpzvf#;8jb3MApQ4nvKF?$I;YXLEG zA3LFU=bd*vc?K8-ro4NMtZ-l$0aMGUB$SPmd;>)is3@!+Y-h9zHKtqT03S<{M~jQD zkv%3%Bf+74^YupFyfwfux!0b%tbj3zlk@H`-3-9^X$$#)B(gKB95;{-r5>YUlu5}o z6iK6^nkAU$IqcLySJ8#g46=vh!m3j*t%7vEdQUa>l*Y{}(2+~YGG4mhuoL==?V6CU z$Aguk#-3o(jopyj*ksE6!$vW2T$RgQ>dX8gsBfWdxHE|7) ztW@#_TpRdlp%udOa;%QWc4*bSOs?^Dc2!}lgo;j`u_s;@7Dj(;>ZNhme0?*17fj&o z;-DWc+VH3Jba23cI5=1NA@nNx4mNI(b z8aZxX5?+<42bdON0xw0rEC@a;ScLRaa!uSbCmy=p92cz&&7t>r4?A^94?+{HPUX^| zVn2uVyRcfzCV%qymB(R*wM^)|U_03}g`6>31ga=`IYka4F>z>FPfP#W-`EA#m zE~_AYrAwNtt2OTDB>CW7d%?$_tHdx7-qi;5%4?_Dh|O#~U;O8c8N&fv2n1F$<+uw% z%zqD0HMPMVR(_eWs;dCOEiR>y#DNX<)ZNS-;7EBWzC7QqE7F)CvJkCf-fPc;tk62? ztFHvgzH5*xzw7+nO>&Bzj_kN_C%28n;}uGNi6R%UM7mubAB;2#$;|E7Q{^T8`~2JG zdqC7x@l!dP*UN-0;vuJ1bBpKI1SRm=$a=C@yj@%`sD_q|E9UH-x&JlnGh7mm1{<~t zZD%%yYkUf2`QbPcWCTqsfxPl0PA;d2Iib89kmIu6iaqdJ zVCSUw_{T>kifRJdXuM&qq$KO(mI8R&U+_m?PM?H@jiJpU$9LN**qGpz?E1m10lc&Z~m~K-h>X?pw$(N8U zV2*<5<|2%MHb{|=EZQZO$`D0IR)hO^tsF9TKn4F)QV|S-M0IL<@>DezCABkaBhn&w zC_Aw=MH6)4iWi|xzdAI-*(Qi&1vLxBhz*c7iHX~GO<@3!>g4-}Nu~ok{h(pVD5qZ$ zC5QZvT~yR+w^|5Z)%ch5)ai7LAWwv`<6)=WvIHKs3C6aq@xdK#8W-SJxZyGAc0IHT zT9KeT5Nc(TndP%K1n&h+2BcKLA=isGZ?^Gd^!ju2q33XS`w~RoW?~pU>)*=U1gFi6>(p5^{Opzuks?@(k zR5o{uPx>3rwwhFR^wT-nZzQ~t?VtWet>l;j&*%AS6uY@lB2MQ5p?n=Mo1dM2L|p5& z+%w^gCeI`PH6BNrAY)Ju$r6^QmVlWUyV~+??A= zQj|-m#nDtYcM!+V~}&=`;Jpg+Nx2>Gss~DGwPjm+Nj>=kuxO4SKs9vKU#m zoUwzWg$ve;FS7{98Q)!hdb$B%%C&KoWUB+qo5~JN3Pw2_ zoCgGjLU0XM$>M=d3X7XiEMY(ssVVXVy3^R`;wA%v30a@daUny*=O(vXn&y>0y~{pg z$`%N)Xi8aL#7-dn+AH#}fg@|x-`Wq8o$LgX11rZ+j5A6bIzY*xokcMSB+|zuLjnF; zd0AK+|BP!Es6dI~xgdPhyFyXb#bmv+!Oe245Y7`NhChae)vHvSp}j@4AYByeqP{|M z0<+u@n@a@s+#{e&?eo-dO8vEpJV9c31&ue`Fw|uWppB9&Ea7DS^6>5aOj8aSHX$ts z4ttuR!s*nNEr3UGqut)C{hF}NyFvmJl4T7{?upJvw2Gps_c50532{p?mD2Ud&E8?8ch zl>7)qs;Q`tU(e?DMW*>Q`*-6r$79`oM@muI~{b4 ze~EKi*rYn)(c{sRf8(|6LlKxrTLnhYN~h#oD3UVL;5eCSlx7LmaM8u5V_M9Ho+&$^yDTt|gm=;%;4m!W zG|S5&C&TuBzcJo9TaS-^7RMaO#s|%#VT@w1$zNOZs}Fq*2>I>qBkRca&zV4i{Nqt( zR0$>DLy-cgtP&LiF-?}=0BL}(C((jBazXg%8&5p%E6b)G;nw)C=46Fk7x$APsEtWc z?sO}bJpk6eeJ(p80E*|9avHdoguQeOb%QBd@KA;#On=o_l3{#PXd}5Yz7H|R-R6<> zxaV6Rm@))>L2N5CoWHtKyUW~OSM1VM8ITX@RkjDh7;|8H$P%oW-ZK&V|B|@j7p7S? zV#53ICqIq)@cmzX@JBH%cOsMn+tg+~tEI z9%(8x%TZ@Dl^%9Ln#EW$(LPOfZHA$df!6)gB2AD1F5 zTJHWd1iE8}-h_^e8617(WNc`yq>8Q;WzI+r$|li4yWqv@4be63@xfa{6M>{CZpr|p zE9G!vLsv~n<rY{j3Q1mhCSl)?< zZXjA1bVZ6*H3opuSCvg##Cd^vB0Q~2IJgWP3H8D*m>KOG=iX=tweO5rIU6WBZ20b< zABx>K4*ynDw>j74PoqW#wm+L<793k7e$xu5~KEK4{*nAzg z`w{_kFd87I-E368!eYKGMg@y8ZE-0pR7_g+rELd%4ALTh_(%Il#uO-tA8j7&q2%D) z%%h^Px~R>qk(-V5>fyJ8cF@Pe>y-__`co<|4yxv4E1^`(qrtVBwT~0{SMNM+`%FrYqeNXTm$~*2lDW45? zC7y4C)R8VAV(uOn#j5FxA8X%gkc>5$4Ue$`iQ~GYI@@|GO^ZgJcd0qBkHkzN33^6u z$d(7FGn~7~eWoFJzeKlL<6lbFLgerC!g~wS=_dKasVFRgY(uENcqsEg&U&x!j%g0J znN;b9fUA;=Wci$I8hs@nd>jFhGoL4c=K{(^l>ryYGF3L+@PB1HFcIe8_UFnCa=5Ca zV~#-``RTd21*FJ!FMNIz7({Pe4jR zDM4C~n{#UeF3@{`OdRdXLDs4PCmPQ!Zys~ zm6q4q{fRE?r$dJVJ{?*MwkHa@ah-I_h|PJ4?p zUyWvLB6iMf{_?M98pKEckEJWfwy|qSUpx!L9ac4~5;)6?l>*)RdEB#u8S2$YL$!~XlWA|6jf2B$F@aN}Mb~k#gg@jG# ze~iJ*Lt+dwhnjgp9P|T&Z24Ky&j(54=S;TTGml>911nz3a>i~zYPk!__Q3?vG& zv1k-`cpHMT&TGHu;Ohrp*C^5y2V9$ED}vRv0r_+mT{f?s!TvkwL93Y#cXdFk;+%PH zQ2SLj4|Xwm5eLFy)LKPDFru-4cG)~kq*=$i=A#>@5z-@%$hJ5?4uSji;_W`&4Z+Y@ z3jgJTvnJCO-GB#NlSC_m^{GzkqyF)tWJZ7Ci@<9w?o8#7Rr_P$ZRC)7+y(m9|G&2l z7<=^9ul@4yj{@zgxIhR!^E#sCV8}o&V}J@zS9C@UxA-FXAdQ$CTin3GnvF_~1UXN}sRT3M}sj7b)?;E)T0@{A*elMDkkr!a{dt&Cv z{Q={GF_W1}iNuuWyAU<{RG%&A@`)wroOkll-YiSA0DhCB>W zW1Bwo_@bNVzKi;(GkS!JIQF)qba1l{=oTP#oR<_2_nwm(#7_FFAO4mkJFv5+#7G`y zQF8F~rb9A#Ko;Ez=`z?n3;G{nD+z4mn7~6GIq>qA2=^;7ul9nM8jrL=9H^ZspT3JO z0!qrO;xrd41efKn-MX3AFFck%h5ePRMBJog@#96SvvfVcPo~v`4(r=9 zQ|8#t^o0z%=Ft7n!{{zp_zlpw2umNER0)9tq?*&t?B-U^co>rDdT?fw>iO%?*NdhP z-iIwEF~U>i3pOrYIzI5-OUlkdoP9}e2D~Wet zs|bpZN9nH`N}fiM%~aH?zy|unG>{$0v3-*pBS4<*L~agv@H8L|pu@1f@%{GDmf!|1 zRJGAI4hM{S`=d6k0?4TQ?X}4YAd_0&pOI~vp6#&0#Y`&pHKv%;E_=e?=o>Al5N@Ux zCsT`a|6jRFBDENHvZ$ON-VJ#wu9B@8lMl%r^S~M#_V^HEjL8Zc)S(C0i(L)EqwMW( zo+YatSa{?b36C^NzL|nqh^k+BP5jXf<{u9s=eCy@%USK#EzR?dgMP);5!u{ZWIOk0 z*lt-#*s8Bz_>nD_k1<~UXNSHR=-|KWUuK$J>#&l;OeSqQ-N)n!z}1!~SP8AO;yFj1 zJ0#lw%cT7bs_e=MCi8nXwwS7_Y|8UW*cJdwU@kQKJ*IfTtqE8$1zQ4Va!>P9 z{Z1~#Ho-gQrwXt|ur~NY@KNu5vJccKk?@Wb{|B?vMOlJM;otPP6;69ZQvE97wm~p; z8X)d9`bL`xC)U#EdCSJ=$TO^rZ!5Sc_Iu;WhH~H#kC|*Jbe@&FVB0ecJ`BVmEbIV<@ zCrA$GB;-)4@3^ma!|I_P=NuZ@__0%v8tKuCfQYG=99F7JWNmy6S?7}&(Hwm9wZ8D9 z-tjy0E6ojEp3RZFXLWIP;3{dvCnbX=)*VL6$5t_Oz~{ z1GGD9QkBu2P%+Rg9EODNJfND4oz@YSPIvP2xxEmDyZ}LG;JUT2SH}xqza=NxL&uU) z$A8Vw3JOn~mWG(Az(9XKR$b_57PqOkb0KS}U6CAgbE4kiv%l8TxB2XjiPd)E2?1S$P+;R%6%I1R-z-1$)e|AxOlJk#8M)#^y1qc5}!TbZy)7t z_|V{5yESj!EON_%U27{RkDl#MDS1Cd`mhYSjc9y|1DhpBotq^wyn}w~c8?`6%Y)oi zbMa?E*x*!kD|0hs55aanNce}j0&5~S1Yh~aFtFkm2QKI7xJDHSju-(NL|49nTreH7 zO>cmlK?Gqh#WT=_{`eapS1{`hauA@51WQPe5MPa*VelXhw+M?1x}}4RZe!QN%Wp5L z;AVtEH%+|svb+y+8FB*4UGNN&=T|H|ttgy!UHmX)QIF)#oE*B3oBOTn3-Bp{!_xYW-lY>n8Y#r8dXNIBnw4xu@A?VB?*e1X8O1b-9agRqfy=q>J zppAb+I^cRq0z`&DqJZU()f_E%&>63;;sc#Jt@WzrVKlM_Vv!a6+dz}rDI0I6j?cc> zCL30+O{148lM!vR%SsKSG@HI1?PQQv`JevkZzS4*rIp5LZAzfzYbmmdirOVei_E9A zpq$)<=Dc~LKI#8}mjx%!5=hu)lG%;I3KEn0gZ>s{N-oJpxp80x*bEgVro3Sx2WG2a z=E^Z~EiYd5!1V^C!)4f+DfRi8v;;*uDf0ZSc0Cdkd*AIB22pYJ+jo6Py90}gB}R(P z14@3MB6q1MR3M;Z^UMR{UEbHHJ(R`<6u9fWVr_JfbG>M}Xb@_On&gXh|J{DG1?m&? zM3|hAPPdV}3tJZqJMEaCMK32?{nh0zX>VT_=hJK4_xm0JryeHxm4qXw&#+UC=%QGw z=m=XY$1LuWpc9~K+;%-bFM5tfSxKHfB8QIF}D#hmE`l3q&Y$>VYtM#W7MO%Y*%hoY%{4a=3^UA$H{@Bz>hiB3KfCG`+-6kGH7Oqg1Pri+=>`CZ<%h#|%jr58#(|)NALbXk* zHWK~ndFLMIeBSaY8@;ZDM!#C%7SDm0ttk54?5~_$Fc_IH*+maQa}|wZ`IJpm>356Y zTt_v%b?7TArj)xZpR!^~l21D)IW!-{|F-Gf^LJ0XWCwq+g|B?am2t6gd0yQ1D&_LH*x*TE@P}LDny27v^8nuaLX(L~!+H^^C zIk?d*x$KtDv!etVA9kN%&DLA2usdmi-|OK#gJ20vdG{Du!7jt!alub^7}?YtDftG9 zBv4V9mEFc)MKVR*vmeT~x-SnH_RXTpu&oUyZVpv-_jo6pz8VF3tTX4m?y#%lp$FM|KO|3JttUJ01_<3klit3`hWvUT^db2lqH65_0TFn(l7y3~)I!n3LWB)w;z5Sd_s&~Ql1vQd&%!!%ngKdj03z0LHer z5@=LtcaZDUIH}TYfYya$zxf2uXanw0yrIk$j>8-~Rk{v0fNCQj7)$E%MK&IYx!pCe1om4{n4^4QpV2Kb>Q5li$=EYaY}xaA|F#x(1L+m zqRZ^O?~+g9ziRKNo@fs>14*13IpW%&+#Q!8B(BVt?0pY8-a5mMI;%15x|dYW7;?hP zmd{EF7Wx%-Qn~;vHr={ zg48ir%xi_bmgK7L@TjN%a>7OV^CB)JVzCP~;>PmEhCjf_YCx9%sbr zEx~R4^FPS`x+eHE$>Ed|9aZ4hj=xj;v+9|RGu4Q&c391x&MTa@oT#rWZ_RF_8-%6v zFu2#oRPigOCo?O*U%lY^oIKGDkJEEDKsGrg ziIO|KEkGajm~Qf`h*0C|dGPh@S!L{s-|UX5QE+1euP?X1zf*3&l97+xzn@r#i(~)1^;bjFQ`#*SRh5}LowoSj|SvD^RWRi}<5)0?^L`goFu)8C0 zMev5;c7+;~?wTc#I0tF!>X+4bu$cqnYsZZgc5eBtJhn2kq-T(R%@tg|b4G3sWm|KrM zK5Pbo7Y>wl@vD164fv4#mwGGN%?=+9Y*U{$!bc?~KTMG_Dyn_Xuy2~rby2M+ZXNck zoB?5Epct&-)e9a?v?ETk`?dwDJXz2qiS!^;lK4a46}s_Wu` z5LmsY#`)g%y+DAOHRuQr_Xe4x=5L@66T~4g32lRL5l3BqBaEAiie!A8Vc#zVLif}>%HPmJ}|Z@vV4l? zcF?C4H$32kMmbXP3Ofywn=V@;vO|ZR+L;0ch_cFk#V}8grlJ+@k5>RFYv0apWCq=S=3W(dTz&fNy zhrNWbp-dJadmNa%FQJ*9QVMDSj<3sQeSO1)h&|Ts~{x$yHbS1BwJ_wW~>Lp~58~@sp zsf*%8_*i@bia;YqMK(x$(K=?6bE7al!m%o-y&9n?jC%2%Uh2IaZw9;d^wJwX#EPY?^vX^h@tFC2~H z6sI)=QP-xYwn6lcVURcq+(scyxB2#?-fIg-#M|O?ZKT|R)$V2^0G+1fCn!=!MK!u7 z@Vcc2H7A|k0yyg9ubPYVi}D23WMIeJ<)1If>jf#hKEw74P733(jbC<}wh0(7rm!oG)lF;X@ezuv) z!Jt-tcwEGmX$d^^31}3!2aW<<_PXKUSy)_>L-9WFB0AT5o)vEA?o=iOZkSp`ALCTc z7IDDWG|1qjHBN!32)=?=2lHq$t8P$u$^k_u4e@h>X+}l zTJ%o_J?Q(VsjrfAQwS7Nk0Oz}LCN9U-a67xbNYp-Va?jJ{F7cdhaOHG>TjtIySShXU3~7{Z z3C-m|k5H&QL3iKsStlfUE{pJ)S3r|R7fv>vCmN;V4nQXH5mFb>uGsEYC>-I2P~)>l zuF4Y~&@o&kDaX`JN4;P6lR4&>*}8+HK8NWhtI2d+lUsP?Vc~QG%9Ly4D#=!M#+?Iu zpbi;*Sqdn59t8|rQTN5A;sW=h-ept~vxK))V7jkWrBHjmzvj)31vGlj^*x zNwvkfnzx2GNIwYg58vv4()|H9Q+b4!W4G3DjO0 z7j??&J&XMkz)DZ@QQ!Mgya;_v>SSiEdq((>Q|t83*W)=E;0oC2dww2EpH0IW8x#BHQic>3TkfuYE_9jkr zP@2zG-^2)wqF`3Pu+giRRD0Xp8Ez%Wo?Ad!A;^M@7yT^wy8;7jeqS>6FJ#RZ1~#zp zj8faSQ1TR9HlxnY>lQu~Yv+#qZc-&P*>nulU!l>}PHX3G^}jWHuMDkro5R4|jIf1= zxp8cLe0#;p@9+)qSa@;M$7GcQy9FUKI%>jGDfuRfBvDbQLMMXW46TJbB4#Pt0*>xmiqs%pKf+b&g)4Xn?LVNdPp9Bg;PF1o!%vQAj7K7Ji&y) zVDz`u5(zN+`ia_R1&K+npHACp8bD{mr02kfjhWDTo_7m#E#d)PM;FY!HM<+M4(#An z!7I}3?{tHX!!srnP?i8`M6H-85aDw!1pFY-)3B8MxPH=kk~&rZ%vVp`1G@kvoA4ts0mW0<=-~Ye>mIalOp`_J;^^Rd9VBDwVcPVm*iYoRP zaBWh3!fbk@fxmGEn6psa-5G?ZG5~A_s-G-1k}<{ zyzaXAK9dvJM0Ior{~`yyhv|O(q?gWs!tenI1#FxVEyx#ubrX=x zT=YNdo(tvTnAU!gvo``;k!qCDfc5~rVkW$}LmnPK>k6nW8{zEg@`CRol z<)0)RVD5U<(f6k=i|ho6c1n@J(uB|a(Y93!>ge<9^zVyrg zY`~o6j^HUd&(0KbTz4Z7oR8LwJ1BV@MXq7-j6ULxZJIKs>Zr3sXaqfQeF~U4!rSJ9 zm2~K;(I`+)))`3h#~gI!v@I2^g_z2!%LkEE&^?kQ zlH*t5lQH$?)Cz80z<{JeSOE_1v5+_#1pi~G9Q`52*#3<=O#M3p&SK-@=aFs)mTjwz zSY@A5a!B5J1gRF%(~1md^>vR5I-XM!e%1}@f1w}pA>SurjbfksMoANxR#lRY>FRjs zq=EG2@toR#)?gs7&-KOgUDE#9nIIGgT-*4yl5J43S1dvv^kJt5bL!~++0npct{CzE4 zKvFn+!V27OdsPFZ?}~L~3f;oK@GcsS!TqkWvo?Em)2)yWn+$)#pxbU)gFA3R6}TYp z2KXyETir7t4+tZ#Ul4lP>&JcTDcDZ}++hVita=XHn`(SM`k`s3REN!~Fw^@KXEk?bv+w?Cy)z%otOk8~t3T>0 z4ctpY)MVn}eC5zv!g&kK$uEW^DQ{MH6WCrhOVFc9Qujt1+P7? z#84XL4&_0=4oOE?mJ;bnZ26X7@Q1=FmUHmp!EAVz)f_l(F`(CqUuqQwRatfN{lg@4 zET56%etuGEWOx-(@gY}HF8?XG zqCt1YrcH2(?CkVZe!Qq%fqfSAnaJ(Bx(uu{R;$azcbJR9PT2#{S!{}amH;;|VBm}m z@%@5U-k(lC$2SdxKVQS;z$%HEaCZk?GqXjQ%ov)yYdQK#8kBcC-5_~FUcuksgSkk)Y$2YrE{R^lc%Mt~XcD;D!NF6xjqxDJG96nlkRIptwT$xh|j;0x{p z;Z3T}pxdsAUg%TQN)xB;iiqbt^~|B`g~h)2pw)nWoK2#N;4>sHph-1y-!fhmzgK)| z+C>g#W$c!nCf!~o;$KHo>Y{~O>8}qifZv@e{nM85t^+CaRdJ{6v+Evx`DFsmw%~iA zNHae9lJASh5x0e=T!d^4ZwH2$8Ro*2uU->x5)Fj+F&8;;OwG)XxtZL0&#Mx2Z>Iz_ z@LL%u<>Pngt~yJflk|xP#aksWo8OnbW{XZ+g%OKJ9^XYNd+t7$aA8Ir*5?*COLS>; zXrq?;FALH!@2!nUi!2HMn2R0_^_FRCUz>rL;W%1qS#r_z90LN7)|H7Fr znKGgdcl=&WV%V)wj{BF9bfYyYk&?$#WHl9a#(CHWyt)tO6wHEx%zd7~xufiaVw#Bp z^fQE62<>M#OwiW4*8}~&h+*~4pc6`s^1%F}`I*W?$|hCPyo%r*;D-S!b>^-#CoqAR zqC5tzZ9Dit%mrZxyfT4o^=Kt{{b7DTT)SVf*`B@t3n5{wBYd`JKYW;%cC(951e%sV zJ#QO1u$VAY5{35k@>$D*(nVQf=w+W1xPvrH_DiA#>I~;D-Db1oxbG!>3LvI@O^gJw z3Ly5w#b+a%2gu_H&4J$=R+x-ObyLXQDP)aN2ylp!17E}dh5(UN=AiHs?zwqbeVF8P z402r#2doI~ri&wrBKqlDv!N(z2NWzOGf0f4=ceAx-Oh#T`q}^t;^I#Dlu#%m*-7;-wLxDbrL<2eT-9zxt%zrB-=4(gb$-`CxY z<=gnEHvTagu>HZfRgmGAIRgophMlyWY;NU@`@A+935hR3ZcdWVQKkBnaE&`Q02;Yv zet(7_VM*Yj$G9Fz`j(aZd;Sh((LgMT0rYD4<~9ZaR)tV(GJW%a-n#&m@v= z!@O7;OG|%nyQS8{AgZ)K*|&(Kuv3N|*qYvNqzva$a>za1PDQmre~^bEC&4|VQD~Hh ze5;@*JobKX5th#%mL7h!8|vgLKtwhN7eksD)>?Jb+os1!wbBOXIvVdSpZ_?-Zm@zm z92c^T!C=$9AC&!_9#4aB2gZt-2H%go%7CY$)cx4Pq_?{vE46IiUf+WY7ylhyY1!Pq zaC@TQ`Ke=pi|22);}PC}yp!!?kRkcQKiWq!zA!SR!bpbfq2$nhCXb3Ld*{!OzjjTW z!D$tiiRwKYz2!zw6dim~h8z6PWq`&=H8 zqvR&H0{Dy5!;gAjgnYP8Wqfc0|6oLnph0#e04gCl_ri1hwW{t2)WrLlw9pUgMFoLN zBabOC>8@V1oBrUu${zDS&c#Fz1Ut+>d@fc#d$-He<`aX6DR#;@O;)lKF^;RM0}biu zxY8y{o)(-lixG(&SSgU9NtDr-+ij)Y8xDa%ui<}qXMNeG8LxwHp+H|Q@KzSE0 zsuUIq+Z4AI?F>?vL33Q3J{#MjtiF*ZOS;Lf@yi|O9=o}LVshg0$EaWYI_fundjD4+ z{87Auk}su5v`zC|toe(g-}>nvOqU!-{jBB|5q)qonL)0+k8Ut};=OC)XY^^Bm)3t+ z^Xkmr!v?Zu2eHEmR*;<((z<8!R}JFiH|JBoPd;X+j5x3z+-f8h&r$LQikzmRkUS4M zLoqwOZqI3O-Az*Y>3$kzjeja1NkLY)uJZoaa}c^~UY?&F);niN))|xyP02P;4;MUG z03~=Xy&k(sIjISqHOS>B?zDutM5`EZmah^J}p61M(+xW7UaU2d>W6XC=NwDoSn$+ zkZpQHqgd{_^3@pU8at9mjC1~3&#R?3u#!R4mhVnVG3_AauxK&U6Q~o)8b`X$s1Lq= z@O6L;IQ`X0K5OMI!3|uj7CX#G`f-f{`$(ll>XI1rX_nD~WkJZ`mrbkng_Mo_{@J>i z*JJulP=gH8fUz914^(GVtgB>sz8y9RSdJ4v=46PaFfn@_EBs7O4*$0_Q@_8%;>V09 zA4`kwxI@NwE3=!sW2RQo4pimU98B2NwJ?kqH3c-vw9;fIUeq9K4Q_xUSu8sN2G`J< z01Q-W0dOxbfi?TeUwUGkSb6FA)?UC4EB|r(TgUwk`c7c_M~NicflFfkpeAUbPOl}uy7`iw2)Y4713|o*cCHQ;EWP7)ox*NSXe>|CX z4y-7f$+U}GcnwOoZ+K|Kl9~PQU&r!*iH{TMrp_`1jXlfuJo>{=TU7xgp4{%DB_!u_ zR*Y0@Bw+Sa^1T!(8fmtxMmjX4VnBb-u+t5XLAUFnRkS{%3To_!WX+P|dFoou@Em>n zWsR~!0&ZRO>768Jp-UOoAYq*}HiNE$e4mQoK5piW1Wp5u`N*~tGnV0FlpuNWrp+K@ zUcyY8U$iOGfE?eYt9OwD?96TlPRP4pgsEec96(m1XBPQP$-#~r`p%fpnLC+j1hpnE zV!(BsCj&`F=ZfH4zdlX<%GjkxM zH3a$ZDHtOZ%mzi9`#ouwG|wj+_Ukyeg-6IvPOdnNB|jL%y52!(cr> zp4UHhy>mY&gHuCx(s-;jBnEO)uZ#E1LoR_#k2{Q>hYR=a%qfvI_<|gsbA^*de-6`5|1tUIJD>MP zt_{Ks!2_fhsM^-blax2)?ckH?0b6R7syIm9b z{5OenU{fBLAxBl*rciQlP;JmN5jR223Npv2>y=HU$m2L^q?d*2ynp@l2ESfuiSWMN z+8aDSM=Y`N`~Y@3!U`LcZ#H`E_>ut|&0oKjP4-S9XN{)d2qmwkNChgUh0wkoiSH5u z%jRVO-Ceo}isE~KuP#0qZIu$bJUDTx2I#;SErRuc*GYFkpOtndnOm-i4<2;yfxuCl zTO$`L0t$cr=g(U?o2+KbaTA&n{(ii2mH|diBISOPHHCa^1fx<)UQChQKq5c=h`86ITXx3( ztbgjXZOXf^#({GUe`*wG-Ou{(jy$8xoKfVVm9B|g=Cw(bqAY(C-6?-b7Y)gl29XRaieHiI2nv+{ixqik7TnXWl#R zUq&Elc+^6cPRX}WBn2xG(tT4Um*>X_micaUMM>DCYLIn+>8{gxLVD-4@hj$6dzL^G zJ0uc)Ol*qhF~NiftpK{&v21x_*Khuz`n+f4wDEUASEen%y-ymrWp$SJMKt90mMU?_CU;z!#fhGhMx`(iW#$+l-mBukFE5AM7e~c_;x2!m@*ErK? zS=m6z6DYElio(zXh9Gt*QMhC(^`VB1YXQYbs?UL`wbCq~_uymm__`6WZx{^EGWzrFrkrtgyexs z_lUlWU==SXaEV|$c;cXxzm?qZIP#{hDJ$~#9(l8spjwKQ;CP+-7-Y!z&n{OS^^WJY zGxvkmaP30}VF|jPx4})0*a2VPS{wggrbPwMvw%3Roo=oI0_~3y@XjKw9MXKNclj5F zVRZ_G6?ci-na9vHqf^!>EeX$tqLmX$EK=!o0YXIJr{E)-Vit{s_pMAWWJ09VTcLmg z`{d$3TPrU%fd(SQ#_%vNc=ldQ{xr;>jR>6=Y$tozX(JA7f6p!2<+os{5&A>HE6xjl2 z2?KHeZAw*Trm0e~hj#8$-dXyZI9F9gUj$wjh}JD!{Eu@YjH#vjYctV26aa?ij`Ax#j#VGRNIM`lgdGcd5r?kAqhy9vVz8##E=a2 zon!`6U&^6r$^S0kfA6gDzfH*;9M+X$hMr;2t($HME}k1Vb(QKO%+10Io%AczYG<-iUT)i6B~lY%#W1U##-9 zj#^DV+k8%XjrRa6={G5J(qe@vAD+WP#tioz+CrE%P&pUfZ+l!il{oKU2qB9a9y+8f z2pf_m0wBt#-<)f&-ZiEZXT|bpI)q*JSpj77_o^~aecn5ppzV7)s1MeI4Z*eI79I7`r^$Pd5yT9XRTZwPlSFm&1+OK7 zAz%)w4Zt&9(qsNdJdwBcDKFEtk4a{hMIQ025w}4<%OP2pbTf&MY*LlV3`gnsU~O<6 zohn%Y)<6v8#w9a2Vj+zqYB_k-c4*XoP22&D^EgH~lNhmA)_c9cG_LN-eCfz4_yWPd)9oPpMZ$YF98G->FH(I}0RZk2AMm)a(BZ4Y3*95ye% z;@AKFz0N-w;PbP(A1omE*}=zwUB~N;vWR0Q%ZDiPn2Nf}fzDmfCAadG;{t3xfU^(Y zKEF#6%*wka##}8(Z-D(9VQW=Mex=G%W$dijS;#0H$J~Y{l39XEVFz7BWBcZQSzL;SFja#Kjjp%Lj$DhflB8byy+y)+v9 zcu+nNrrryw$;h-C=*O4fg2RN_5GLT4K#2QGju zRFsa11RaQK%nQmeio{V-diNN-u4+yNl#x{_ zOTw(L0D1WvGkb+)uhQ)8c02I}5gT8*{8Pm=10tGVzxaES)%1iVcFAA zQF|cYAy;t0>#A>$B#GomPLe@EJ*nkvg4&h>I9Isf2D4OfL~tK0-cQ`SrCa=~Ipw^) zGiyUzZQFHrEU;OI-{%FvSYz4p7%TizujTmm{@MVOHw$XL$ZdA3ivyc9u|^J;K}!CJ zA`hr2J(DM_6H0X(|CA6zRa>VIIqmb@y@Sgqq$b09*|Vg6+WGAy%C*YkfR#+2x?VgDGX;||Dnw3J46>fulbZQ`CxwC4N>EX z=*WG6CE@YG8c7cyO8fN9i~1NW21M`70oNoEn)i_M1CK$EjJ|NUn+C1|YoYo)cx`*+ z@pnFaLph?_MEzp}DP{*g2hOPgw$D)#t%j0UQsgj@XOWbE4BlbL+lmn^54b0|NNZKA zzI%DWc~z6Dh_2zCh*&eV#5q_?M!t9#?x1QxA|LIRXUv* zu{|z?+okyF5VaS?-zc4YDwx!oSEC+06C<(fB3b4o>PztHi&BUU4@T0k7v;eO&j7yin$ za~B(DodXLOGkw0=_!-VO=T>@P_2U`;SogiMA@Ipp(@z4o`}P5a2dq*|i}02#)en=2 zw{si#t6u5gYgJ1Gy#%+c;Q(fXEQ`)@tOQKta&mErL&@PF1 zB+G|ZgsGm*UdS+zMPp;aN3v@^`oTsGV`FhS4se*jh4uWu@CNhJ@T1Di$p2lc^N?@5 zq7m$@K}KDzOVimm?~J%~`Z0z2Fj#2leyrwTuLe*_;l!R<>btMB!{S;ax+vCbsYtvx z?9}YLn_Eh%_-Z_{L)<~M!Og)-1WgONmGgM*Jb zFzVwNnb$ot$8mX%PRZVz^sfe)B@k@*9=XU)b#~l!n>;jOe(OT4x_3Wb=U-`1mCAJfQCi2k2U2{_Kv9PDY!YT3+IPnl3ISf35<4#6)o1>>z1A&_tIc@RFcE%s}{&*}L5~NbwzG=xCoThpv~{ zcx@T=r;YP1j~dCgo9369F#^$!tC~JOzp;xph^7AT`1J(Yt;-Es{?}G(Sk}I|!gv|oaGHjc@-EJo;GZ)Ny)c}fb@A}!V$QE{> za9p@US>9+jPd+8jrARgv6&E7J*Iv!(EQ8C7RuB6i#wC2s zfXo1w^jAOpElHk2fO>wEJ1vWnZ=*;$5ZL)1;Z@Hp*X61u`9PlvaE+>arP&J;zjpo0 zH-CbT(RfbkH1+a;Vc%Y94y5f4I|1Eq=t}opp~+KY?abmjju&j7H)|M`)mr1Y68T4e z`MUUj7tKIt*iq*VP#XamS%=(j&4z+IddK`~ySC4ji1VW;+{h%@L(5N(Fa}2r&n^ift=bNtJFAP|4i7};R`T#q0{@O%{R>SxWIR| zOMJ+`4|*vf_cOdVeb`9fAP4w&LEYaH3IQbjc|45Q7s)UaDn@l`dK*8H=;AqKNQQ~I z*}4qyz20c2sPDU^*i`h`+K}bL|cZ>i}Vwq)sDPnFqvJcKf2Enn=DuwOx2| z)5m1h7se9WX{4d2Qu0j{Nur`K932DBp#85McfKD!KpvAvUh7Gt;`02hQ0m#^QSW&y zVyjO~sNDhdY}VL8>Dhhu-rBw_<$I<~YHR|Q4(!s^7=a^!lCP!6Dk=(f(P~b-s8QOE zb(rE>UaZTa|A}No3n5`|1UxtXsc*gG^Lgh3KrJ{)R5lOI7VH>ML?07A?pzQyL9ZX*9lI&8%iA7vP*zjmxc-q<76S;wn~%_FhMTA%ue!8rSPH%1|b9ezhi z5<80$R`{Jfb=r|!Q}6ue%DOe{L6w2L^Czv+6pVJ}9Bu~U8G}yPE8@-V1;kMTk z;9%^MX46#=m8=b@k_^eJBpv)B4~$J7q#poN{MqS=Uh9G%aP1}(t%r~vPNUFh(@pab zQkBj8t*Hs)uy(E)qac2S<%j4I|GeWW5-=HgYKF$-hl2qWmdTfiVSxaYy z9S|%Xb^OR05vkyTqq$sw9hO-226@!^0Kp(*m++*~oV zAUNpP1txCVdpA5D%!%Z;R0^!V9Cl(irt& zWiMIITh40MA?uc6(MLrvev`I(9FtkA z|5);c7_H1$v2E42=!*NQ5}^S}acS3oLDsTEk^?*Fq1b*@czY`)Po>BvD(V5<7Frkn zP`p_J>{f4G14gN2<^dfOd3k;*$)Z#J?n4bogY>B{bZUgF?53oc%ob~WnA~r#-FesN z&YN$*hxSLyvPjt!a?$9;9;f6-DQKY;l|pSytIAJf$1^S;uz}g z*Y&3kE=;BVvii%ve;`Q#B99GLU!`-5DvAK6hx_~{gD=bH;E(IDo{m#v~>O4U@ zUFUamZk=DcC<7Zl0#~P@01~-6%5>ZwcSFts{YRd_+NI>BU-j|B_@$4q>H=0Uo}~Df zf)!B)$@3`qzhtD%forX!jmlzsDfvB$d_qN`1~NDm)1%^;IHo(KLvohR;9;4pKA#@J zGwif4qMDcc8v0gt3AV`_p`~nh2>M#qhN@f0ZI{O(sH2<}Zs(qcx)V&oKM=Gp0tc;` znc>{VZ4kA<1Z%m=^{~eTpokYYO7WtlftQ5)Jn>T?r!;18qt{2D zdW}CDLFZ|fl!vth4?AI1OQRG&Pd60BCNtZnZ-GeW7|!vC*pCP04Rlq?3wDr%x#`*FM9!n1k&xpqCf0eW{WH zV}hmH`J64&*1iUPj1zcxMLV<0_X+bU)CB;yVxH(6?}$vJc*1Cul?2pqOq41J!(Q(i zPALSnb*0&*?k!yWUzhZ=VQP!ZNqJi2P8zo|LgS#hawgZnXN-uvCxoCy;AsW2@x+p7 zmsqM%+RESu1a%!gKo@7DTLX<=OJsyB6edqqcX5!-vCR!x_X=H;r7P1SkTRSq6Sb1m1mNmWO$7R33r%JY1$yixA2eZ3C2J+{44?LD3s{nEb} zFR*qtEE;?q7qOZR$2IrX-tj3jrF?xp2IjykSeB6l+C<5dD3VA;m4~gGr;Zl1xRiRU zvD6#X>M%75kRndAt`Taw=-~fF{Nd|mHqYwqu>X9|lwHK^#bbp;9F|v*m?23)Wv12( zeAIm*NZxr+7AIZdnyVNJXlIgq?8oA2MhICh_{qU)JSIe8@okiD+vgrcXB>PkA zSPm0oJKRp__-xp(y|qGYnic;1!s5VSFq0{M?@f$Gq+(Yzt$lsn^w!i)+jZkN;xFMBh65d zrcqQW<3Phswa+D)sVg5tjTmeICG2h;}9F;##(t{EIn|~ws0{Ik}-4(YWune z@0t=uKQCPzSZy`KZe8u25>UihB~c&pU+#5S3FN*yij7y7Nylf?S~FJ`BgLA*$8nE& zq_qD0+6`xeqpt2tRsSH%*g5JPcr}8m?@@8`Bubu0k$9wrc}!z2mtGC2^;)4S4$?|L zR-${%#)9MJGiE7ZUOtj-ugEPw^ogmBYsPHEZZujbdQFG2RmiYy5hnQn;b<=BiDwp_ z|RAI8l>ommnozo8(?t`HUf_gjeuq zkJmNNa+i8(J}=#`N!83C&9p|jNwikprmLVpCHL^FIndBB_q77|GHN5Yz( z{#>`<&$pxU-}>|4&O`3ihN(5YOepK`i)`cffCbdYMRI`4@|bC-VDC;4hJ-LF50qJ5 zc*sW6K6;XD1A);a+Gd-TD5QK}y|&5JEH-0lHcBxcvYYOpA42(gKS>9A9HdcDHDaW>#~mAP9Jo)5OV!48b~4DyL4Q+gBycs|b#D&vow>R?+#MSai}Ytl^#H z9pU{Set+`e`%!=S$*+I+VU(DbI}wT+L#?9R|Md@Z7mnn?EI!vp${jfQs@X^%IZer- zsc0P)wUJu0@bLou&p+pWzxpe~-!K1m=6CLWx$nCx7iv|=R@zTWJPHEqxv7%lf+5*G z@p;w7x6_#3nWN8TICnwPRXTmnqe<02yLV*BXLkWvrpFJoHy&3EM^)W zKMR?}8pYLbtoT6%eN1L0@~oAIW9TJj4p}EFS;<4{!Qhgt@tDt!D^Q6U^EqAA$E;H2 zLM~Necpr4?gw&DBxh243i2tm_f^{%Qj}-^EJdm>m1_wCj@$C{b?T{}=Mq$#IiL(u`2DmXfcc$V&9US3n+qo}gKh9CXwE z;eInS@lrgPxzFB*+2P@pO<5_Xy&T!3mN>8iYNjP67Jm%Qdde(Dst@R56Po^9R=mBZ zNH_jV@iW1S#mFE1t(*V)m+u=S1hd-jo1~tdg5ki3=r)ok*C_cFid>?iu%a31p|GC0 zlZNJdT1AIsZKzhEce29b)5?hvd?JnxYUi{D#4(#h9rPe~4^NHm*AB@Ma1-u#jh(gG zOTF2zi|Dvon}bivbsKAamIk#mkLd)m3|6saQjJ?lINpYJL+by}-n+mxm7e+Io`G{r zUJSVrOwNFcL=eQ0TUb#Wota){rk!ptyS;69ds*4G?rhuXZg%=_r!6p`s0gCq1vP+j zk&7szC@MEc#L-c~0Y&i!5?z(@)LGxh0(Uo9BJr=koow zDh$2V$frBtp07%u92aW1C#IkePaAUC?r{dlXHa(!TRM`0@o2rQi;i=@A$Z|($rgl- zoT%)#J&YyGuR#^_dd+#Ho~dP46R`L z9kqdF^N`H0-@R{6qYQID2Ff}(7jqEU?gfpIpm3x?D(Eyty8j`;9-l6dJt?1u?X~To zyOk+M6Vwc!)B&xMTc$+|(VXPz8BoyP+0KE&2}kxYBQ}D{I*u#^Zzpv0E&WZz@Cs>P zVJ0bcVr(^-VC!Rwt)@r?71bsvSy)VG1Uv#o;yk8Vy@FanWqK9xE;5N!B2_Qw5*6@j zJlhqxzalu(Yk03#`R8ONn)--r1^ME$P*b~;R~k?jpj95zL$Qtq@f-##=;{SpC00VU zC~EkFGuO|qaBBfDmB~}5k9m>pC=}TU7OSUmLJ^hE|Nbw`(}`aakJ(CA6uS=F6&2HT zl4?Jk!GO4IS}o*`dZ(58prSEK7ngZwOuD1*J61em9*fUKR?zvf<9?1#3| z0-1n-Sl=vhDb&4XiF>6xJhK;+yIb1VsICa!KYhOs(kWqMNSea%KGnTle$;36yzB)D z><-V3q{4mmymom5ctH9>A4#K~`JhLG94pplY)~<~+B$MRpL(kOKaI%wb8^CKq?23l z$B8ku)P#zrpJGA!vl|kRP@B0NbOZi+O;Nt^p1e)a{FQZ6;%D}Wa_RW+ZJ~+YX`#rj z)iwJLeSL0mL>t*Y_Y=>9^2gpOQyP3~13wQx!r#Vg1BbfrwI2CKRlO{E-W?gRekk$U zW_2U8W@5d}aNT{+T7Eg5%p~&;2egsssnNnpQL}oJFKi3<@&@_qXC5UWG3GPKKNfk6 zP7yb&H%~so2iIKj;T0%H+dOfzCmwDMiNT&0@YmZ&H@_M>SNbJ2{8Os7kn^H=SKQO! zbCE;`9q~ToUo}0EeaI|z9pSs%5tLT2%4PRcWM;Yk{$JdU&@2DiJ7>sBCtjU_^4^I2 z)h3GFK*4j3Lf*W3`Js800V~*kk`%m~DRZl%Yxz~v^Jld9<+)Y3;dZq=aq@@(dHM(( zvc++{9r7J4-PF1fvaoiij_e9(P#Dp%J|*xxS@)$eOfa{Nz)~*7La%ft5D-z-f?^;< zLvf@uAnS+@()t#Qwn+~9A@ASgkiH0z0}kC6s`XtO-YDD2%Vat|@%&}3m)wfO5=Q>y zEj|9R?&Aon5jJ70pZxsm-M(>b_up3!J%*K0^8ytwZp$Enk>k62?r_TB|9a?+aLIaQ zbBTx1fc)|jSxi!#czW4sViK|_7I>{ULE@WUN@A%&dHccvP;gy0<8x=;zp}3A4 zM*jJ2$s+UUisAkVo;-ME_9IG@<0m9c3@{aBhu4vVZfsUN|+9+l7PWalO#xb zaZcABzlzZfggrj_`S;*75YAi2D0<_eo4Jlwk%>XqQY@4~Rfh-NOS)}^f9O!)hbAI8`Q2^wH_q4|eUlM@irv9hc zTm^me(s7|lkT*Cb845vqAI(j#9{Ey!eJ~^sP`dacX;-u=`k|nv zN6|&2GYfVRNlmdFyg+zo;t@EZ%R$l|qd@(LRgHer9w2|_9`AkS;cMotY0s9+IC137 zO1s)l9xh}ve7AXEM!Jr^`tGtfj=z-zN$m|}-=Yt;F8XbUI!8EswEeA5zj^W-hNH(8 zeNgiHZyoB_js*}XcBV;>J7B2L{A-E@BVqfM?q`9 z{t+lgcgqeyKYW{@{~I3#?S|rql*mjt6Y6DD+0dk0?_THcP&~j8jv5kn9DYl;vJMGL zfwbQjfAvQ{G_Hd${G{|DX|$9Za$fHWV)i4$W$hH(N|DP{RHJg!{9_=eg~bY(8ouCB z#MkrLV*S&QONQ?;$QUmr(0K^)_G5JaWaubB3R}?h^GV_lxggy;&M^3{YLeC~S@Bo` zy#>g!ylU22$W0%YBbj`eHxeBu`R30liWuNQKgxhR=DTl)WccO?kI@}L+ek-HhTo9O zC#1wHH?&=O{k486PJdW9fwLI)FX*E+`fnN|tPIYjm(P2_oHlCcnNLuocy9@E!~3j5 z@0VwHWSB3dx%fCvJe^qy&g|rMO*uv6eJi~j z6wX)k&w=^UDnI-nno4GtyE)oyjoNSx53<8vqXYj#CpDTP(OuDR$!b?pZ1U^Qq*$op zPdD~}om5ss*M|n#(GB{(I%yGwzb*XDV`Vg!2ed0vBEVnd#rcg(=yP;WyDGmxsJZYUmKjQx-mzC)m5mn2GOHzAZ|x7lI@TR zG7zpi-fRqGi6g{^FLU_oMq%;O~O1n7u01?ZUd|YfMT4C0!NDgPUrZHIHyT(f~NY~J3&VGqx0Lhc9VQ3&Is0- z_(kOuTSk#mDhgK;DEp^Xwn%sTcX(?IeI^(;L;21IPiU5w^^+DUR!c4^4qGm3k>-Q* zY0!(&U|)zKB?08OUbkSKR2WbkhNoyg_CDjj%MDxC()e4{_A@5jvGam8i=Cjw4LgCp zkFUELv7^?#uOTr`jGYVw=6Z@)GxO~t-MG586Bz@#4{W*d^Uk-+@_+Z zf8pP==vb`JlV_4LSJG%Qe;=jTk10|OZN~HplEmDB=FbY~+Kge2tBNAh zppyMh*A(ZWl0Du3&s)5TA`bIwm|Gzz^B3=0%Zq1oc}v4LNc%&uHoS^|+Iy+1R*ACG zi7ft`<8>zZyl8Qc0uOb{_WBK9(yYG4Kk8XcI#h!$?Q{pUg~hWSshNpWS0|rsHaR2UQMwT6gfae?IjsLO@c~s%=}D#!^8q0 zgMW5UBXszr!J>GRZ>Qw4SnHcQw^Md7;3DV;X?@$nhd|}y9Mr(wgX9e!uO|5K9!WL* z7lxtAU3-)3)AM^Z3%W|EEJ3-yB%o_u^Ao-1%N>n)6=8;G5lLB-Kg`1U+B%0d5 zBDz$BEKtzz+^%R4lp@7EmWSfawaQ(>UC`a3`IKo5N%qyMFbRN}p&ABuwkyyKArXV& zW*K^3M6GhA2e-M*MnG9|x47=J9Zhhi&&fr3-bO(E^z?)#lJ3Nw%zhJy6i_Ur`Ld}f zXabDrCf#!mv#E2JhNnx9tFlxVXM=A73u|3il6c6kjpRbkuS9+tT(VXY!lh{$BGxKl`s={P%xK zmQw6T6p4Pmu^8P)ak{D{bxE#y;>SvU<@mQR!7sE?yt2!&K^}OqO85*}^sB4N?>YjX)Z#TVs%G!DP%u%u>s6l{) zT!Suc0>gRvOhxdO;F~k-IT&O68Jmp_r^ovIleztjotBgKcgOyA_GoW_3GrUV+zL4m-gd z^(m(^-BU@UI__T|ez)nTc2hPp88W+I#^jFndBX`a)W076r++pVXLZ^nlNBLWEM01c z7P#v?48HV6^ny-rzNE^liyr^gwXLYI&sf`j7fz^{u&pHK;aFruoOmg2MMeZ^;TmKE z9+knc5XXMXe5M~-#xOPAAupe&Rc>I)1MKN;y=drc1(JSxdC|@6czgwW?cZtRB-8gj z-xDO3Ta?C$*GV8HF(OK{hGLT`l1N2mFGy7!lcgyh0d-jRg4P=CjC4)oEyDX2B>6lrJ=^cY`us0ra8na^E944YcZA_Lh2o@q@N)?8mEO z^6B@rKC^u9hd?NR5YFPzV6|{_XCBj0PLeqQ(BxkabfN^;)3EEF)_E^yl_NWV;iGJ&u|nWIx63 zrO0k7>Z71;Iw?4lX$-ET3&XlVo=&SQ@GB7~@(W43yqrp5dm=Ja19UlcH>eksqqD?E zeHzuxU)}!JhBr2RcLPipy!I+pYZR?NkoAgxj_YELh*L&%K-ge=WG*7l7VP23=K|Mv;Mo^70#7PaOlk^*jl z5+|;3I%YD79iZ5K6xl;XHON{OTUEKCS`|iO%jUH+T`-N|HuRWc6)?7&0(-_oS_kZt zkG=Ee_3%~!u9}sXM4Ih%bdYWu5NC8p>ZCxs0qLD1@@~3(UXmD(M+X%^HA50Ks|e79RGFh&FVFtO|qrXO4F=P=6$XhlsBvE#Jv%pOO}S)VZw3@WhaU76!xwnmE5EdM8lvEE(Q=7fc0cc5kIKf_mgB!Vf-hkaW8pp0Vu@j>DPU ze%4M~1$A%&8Jcc{O5FQX5=n*=$7-OLXat6LQY=Jy^QowQ_l96>T!sqjrg_CMspuoR z?TW{K7n$WU%@GNf_ZuX}p%hySLAzZ680oz}SF_#YBD0Fk3`L<$L+`U*lgu#l*}+1X z0Ll(fHq1;WcOCm1PEdX&t!913-&doEk<}X=;AR%9^RX}9-r%rh7;OTOEjP0BIg4Dx zXXQpCk}>tY-;uPJEKN{qVpwu07W%g{sHnczbkjVdv2|znyh+qAYnw{Id6d z6!(;EX)HOMAxXwyl(=E?J zx&b8Ys4Z z9ui=~uvU34Aa`Q3I(_nu*#*9Bf~~3!@y(Fk0f(ku8waC4%E2(V19pGUX)v5Op!oy$ z9d%=IP@K4I%8G+hFE174v+I0-JEPR6)(6_&kyR22DIxs#^gjpKm1i(GDr`S<7yIw1 z`P;0xH;txa#|-8!sdnOc+GP`)R!^}fDRP{OLTUqJFszNf<%ui@PzYB|cgi+TIZA+E z^!0S<39{U?Pl07|+jtsm%Gl%^Jtd2GYKrEnq=-Jrv@4>8eGyt^g8MeAokE=>gc@a^MHeXS?&jB4Bzyi-xrcRH@t1T;+UkCufcM=XknMAQP~LP4jhp+ zDv&(!skm%#0}>|#_{!nqU-`8;k=RRB%WsACtdC$i<^mft`P8pK1%`hfY7^pJ^8zaT zk{DzYBYnAM)k z{RS0hK!mCpitbjauZ8U78TR2>-Sjp|QN&*Hc8{*v9KmEh#Et`(`6qFEZyj8f|HOau z#Z4hjJS|yi3c=DSO{=UO>TlD6t}B*U`fu2dB=I{O>UoU`72F1z)7jQh@eX^e)I3?} zwCf`)Sxn^B%%k(B#>sH&@H7KAM%i4=$!Q5QTj#^eYJLV}DhJ#LW0*~Tl4I52h9WJfaHjtLfySP&V(V;vS2LUI#+{(k>_bC1_)p<=~{Z6*1PKJABI z4kii>b^n}w3r@d2%O~3i3dCSLGnirt87 z&L}Lc&7)(5Ft?;LdGu<20_g(vI-Q_HzG`wVT}2Ly&q)fzYl0yh)8S@^DqCAd)JDjR zbX{$^?U>K84w=7rY5seP5gxh)Ay>#ICteVhn?RwEVz*Nyhl<(}wo|^EzuiNt#9DzQ zF`ys}iy%uF%m8*kWo*83ldzREE4w4n(Lj!`6tcLBP9QPDt~tmWR~rEVyNI=plW-pe zGDf1rVM#&j$gnWOyz+V6AEuf!rhG}tY_0ICbhyQXOWH+$7$ejwOX-T>RNq_SyO_8s zxir>QBKcq+lL}0MfBN9BTa6SuC|p?e>MCF!EaD9)+Gw4$oW3H#tTWcJ=;S@nV3a_O zaeecQO&BY}p1|#yJCAM0dsE(NGoOH*)+@0x*sa~X=Y{XRYsAItANTy29Csz1CXf08#h#;pvp)*?jF8b43U2X#CRG%j4N?f)?p!2zVM-qO%^kFY|;L zDyHeyPCEUn7TPKdDsY!brh6$5FM%#Fy4kye54v!L!pjWqMx`*60$ApWa|WzvXW?Ef zvbSyIZ62>DR(s4FHjTIS@v`JDIBn8-;Wz)*Y0lSVh3E)5^s1riBU*ULZLg08Yt1f+ zx;&s#wMNxV?*nj_|m!C=~|yW>42g=JYV&hXNS5^ls@~^#fhOFe)8sx+Uarq{Y+C;G%C>Wn8G<>)KS{s}L@fL`) zG%^KZ!@~O<1AS?RnHY*%RCen>ec@no#?1@wXrB}QDQV5@8Af#U{ZzJ;Z2i)pquK-= zyD1j8Ls77V#w6}^!;L9(WI z&=2H}OubopFdUw~rOO?*kJHn4-ub-y;Nq(M zaX4Vhv}8pRh*BAtI}#|&K;(wDEAjtLGq9W{TG%7F?57iqongu@aI<8gUI3>T9gg|A z`@fl!{y44Cu|nh10YeViKsOC2n)zZVS--iB7aep23=F&}^ENR0Xo^DHNUX3TtOCH} z_;QBPj`@qmY^=vQnV1Q0PuKi(lF`I;{`%5i$Z~Eb#);ETAa^!mxwVmE0r_jGsB);Y zEsaQ1Y<%mA1Up%~=^T39lxFo6#m>ovkt@`t>J8ov_F@CGX)rk<$)-En=Oo3oDew6h zvGJHav7Bg~cyWEmWVDMZwvZy*si^Ie{gQ*8ODEl0uyxuh7L>DO)$AHr5?_@xhGe=S z7S@Yvm^ym5*W#^A53k6(7i#|Vm=b1jvpUvy&+MLvRw2e(y7Z)BiR>@(=5cj07AsJ|$G^lz?4z^HZaYeY-AB z{#1#q+7}j7>r0b>e6m3NKpH>41Bl{)Vxu%7GgM#0-$p->#mREfRV70%ct&QZCJk~x zTBU&iRr9GRI=nh?8T7>WOAEx<`%%POKT~s&q$w)(4Ip(i7C-4g$gfbM>6GO|J?gBpy0$U^9D!q$q0 z2=iL*XS<4=cnM^szFSAfOzi@;rvdjvfzk8Z=)=!ps(-iFt_QT@N zI=V?xEIK$bpXmrecI6)4eWJmZYs0;Ic*muN_5;1fI9|~M5gKf?ZY6%~h`qx=6Zn{%)Wpz-w}W&9ySP4V2CCJN$=j)@m%% z*yyfRKHi;aUf%ocMCHU@v6Yfvou1UaYOV&|qXID!D}%3tWVxCXWJlO3)g#GD56w!C zS`|!U0Tq15hh)Gcj0yy{nvyY`XMM_gd$s2u%wu$1ItH9L0%Ikjmj-mU8T1bJR?rb} zbx(`?eLhvj6X*jF?2ue{-{F}Q+{?QTsqeqtC)eGJd50jEJuWlaXs|N|NZ5So5B~A? zk445AZ26|Ezap!gcnJkzml097Efkwhk@ZwmZQxa}4#8>J89F60)pv!uLjWltl@8)= zCzaLGZaN_}->WnH#vCn_%WWFB=o{w|w|UIQcA|A_VPPdiZLbM*`-u@BLgCu)kaI6t z*yWxHW;!Sq*5R#GRFb$sR>bJl9pqiIUBX?$G{{GT064H%YL$9+4b&vU1MP|~y2R@+ zRHetyO5`;JYi`T(W}#CI!8~b-7;Dh0NWBaQbCCz6U4hH1g8>x_D%{q1Jd%_~VDUPB zSQUhKu7L+#F1ZEbtw6*J)$QmYX_fU8SMv)Y_k=7WnUHWwi9i7p{A>-LoFh!+wFw$z zTII>elT0psitJHlfZ=+cGtL=(qw_zkU1b%b*3sub_}yyraw{%#l@t5YR*JKb|38rz z@1N?cqp>+5&M$kx$EwC)q#pyC2CTLMk9a}$g7(+iUxN+`(7=(L5pCl&0+$2!*g~mL zgRD{61xc&{_g2+y*hYyGXpsiLc3*M^fCDsuO+u- ztJr$c+F;bH$yF_nJToocJ6->IsI8P4x>H^eR={fwIs8^K50JG>h>^!M#RgslH1i-p z@#1X~T<$??3l?@7EVk$f$z<@rpj`9x_z}rA$u6G$TSIFvb4fBVX%&>VE$NPgKLNh= zDmIl@8SJ#nD(hCt!W=olOL68yb03O}<#OWXnH7%<1Jh}Wd{uWiba^2PxB8+mQolhi zbeVUas#t!;bH~qeV0qonTyh&17kf18Gg}PHn9uPUT)E%85Y%ZaC4k&}gpZv{u_+Wu zqM~$yT>8)Z#FxcerPmZ~D@|4PP=Yubj*VwPg9% z-ZK{%c3N0ikqkCWN=W`+DY!#}k1{CVKQW1)#B|GW>d94s3Q7i@KDj_#UM;Sm`n7(&pAS zPDa@Hw{sZ*tB=b82RMy@^G?j-zkKnUc{PO7+72s~51FBtM354|AbuWwV-EE6#tJd1 zFurpMV$4RcIslH(za=NIPWau`ho6~~2|b$&aN=dM6@sC3MhA@ojrvtAR5Ao~lXX*y zc&ELSygv`Xo|kXprZ`{oD1mzG;d>v-Sr$t`X8k1U;E- zp4oUl5Jr8}I)YBU^4@_T8kfr#ep337G&=Eex!c4kZKv2)id?3mj_|I}z0BnK4!Z1- zAMrjd%yU~6a$9yPXwYRRACL8f6w`HN(B&}M9eRE4y6Kgo9?-?ZbBiL5(Yv9CO($Q@ zG^=yz7*)FeT@VQFqxXXv-hRpOwTIFjMkHSjF_szFPt=1~!HrB|9V)!K>M$ydylvPJm;;)r+7prsI!d@`UCDyiV)V zSRvcRngxThxMnpdrL-&3z4xhLRkYUg%J`%iM~r~Y<2>S;@z~^KKPJ3V_}hc#t!-91 z$a)kN;>NH$bT_?Y;T6eWuPJseyeI!C2sRGDMw_5r-sN2{FA)#W57=I5zIPt84^qTC zzMubt*f%;Cp8S608z;aGZB}phNEh};)X_1Nrex8^z-0UX9&H|)7g%h24;QDQabC+@ z*!)GJxojtwc&rn<$W}x?x3UeNNR0zxHO>nB4a+vJ`1`7pN$||~ z?4NR#O`W?md`DP~52T*MFDET~~G{CW5$*?IK1ww`| zb?ECK-uLJe82#8q=hlBrRygr;35a<|#PHK8c0EN>si=d>V$pqB9`vNe3KM2(p(r6q zoHtJ=Jv#*yM#$CBXkoW(t>=B&_>8oExVR0sLr2);=A3523EzCfH^DrTZzW(J>${K1 z3n-^Hl0%@6sZ}QOibYA{ED+<{LE1?h9mgd5cF`NzXkoqdx?2T*oW_FNIG|f%^0*xK zhSNAWuSas|FCQhD^G`YL+Qtf_RLtaf*E0e4HB4>YMByuNYw_v>;vf zTFG2ka4$k9U+6C-heFLm zQ};^t%r5sRr{YyDEJS&OjsdSMy1SR>CJ3=@zfI8P9m8fqrROfsrpaeS^<#I1Y=w~} z1N6LoV+EU=^!VE{BaDuJ^IMn68du;c9qBf1rC3PZZKk3UT(3${VysQDjaM&t>~~rp zX&-diuN)F#%)J^~WpGS@axnPRYe@T~Ro*8Vwswq;ryRWZ=jk60CdLa|3F@-Y?F3SIJ_%zYTLTmgXf&f;D3 zI_gs@`6%cn1d)McO#cqq>eu?mFe?>0!A(K4IxQ4mlY$$A>!3ikn(k4w26oDTkQul? zDuWZf`{o>)n&e#xXC{5EE97wCdh!4F!I_1v3!x}ie?R;jlu&O~>h12YQ1RhL z@ci<#xjRV7mj*niO~6w_v6U1#NJUjkcg@9qp?1ZY;5dG11p3EkL>n31G&Cs7!aKuJ z9OUHkOh`)K z>u(<6b6N~piP<4zssS#kzCcK2C_u;VSUZfe4G-Mp6kAbYa?v=v;e?7;HVJvpl*~CT zDy$U2A(O6NsP8^073rXTFjm;iY=Ls!A(sYOJxKSR_O29F%j0LAm7bMi=xy<3za;Tx zzr~uW>GwkJ@|^rJzaGMmV1|szCA#B;0}?a-=zEhOF=h_+Od# zY46NX*n2yqG9aBTn7Lyz5>`SAFrCd;AvpqWl}40#-*vB=4r9GccuS#4nNktfAX{Pc z>eL=%VQprxxZlR%`jQx+zSzzmd|(X5T==K>aMJ5a)|e0;#J$4Cyg~-4s61M$!XF*x zc@GE1@o}|~@|w;&jn^Z;7lIi%bRdvk?xeDa_rUuIKZh=j*iK^^!ZrH7M{jr?beC>b zVK6mGyxR-)tRB+wo{!;fc&|1RJ@tzZyfzaPOkM5SFh2}H&-lCg^ z#O1N(ttZd+!#VHwCsusxQ$cs+d2?$cx=D@C!GN1nD%>7>-Q_nz{4UkEYDRO!;-a~Y zvdezmUQOy|^$C*fdnvdUB#?E|)%-hCGW@j4PaG?NJQnL}qoEwtY1Xal>9_xO)|{Y$ zOXSaaeL-S{&H?$L(LN$sr@?bV+2+fBx4r*w+5QAv>AL&h`HFEK`{|7{o5%qt&TclF z_-w~1wvHl4kgv4^a!h~8_gm`TNq~N}SY9qRqzG>DK|EQx-@RF07@+3}FAvLsQ>*EE z`3bhfU+;*Nd%3yS) z!Gl9aSV$g7+LD|es{r?tA;68X*b|H1n8lqYA8FGY>-NM#BjkjF5BoQJ z3}>h=k;Np1TZYPsZ44-mj!3L$QLL6Cn?MW&fR>}G2kPAd;o}g^TJI{PjJm+~0C5m- zg3M1=ltutCugBp4pa`j$hCxJh?AsOgNHRTl1V@Xg=ibx~2X|NJ|C{%)!8tvliKKJG zh!ba^_M2d&fMP*GCmU#mz(nX{g^kcFiwXp>!hY!rwFVR8DP+Z5P0C~gUwtQ?Md$EV z&qO*jtx`{oCcvenrbUW{@>?QvLmdR0(Sqfvm9YadEZo91_0ex$tWn^^kg%drfD|B^ zjHZSu5Mq=dH&jy{m_MVCP966STAODc3!~%EUDjuwwxRzzYA&nv?3%}kL19H|Nk^}q z*do&0@+>9baiFhybRwn_(i9ke$`J3BtcKNnbWkk>@-Xyd;F;GX=?`>??#P$RcJR6& z4}ns~jY>x;?2O%3fYopSbgTnL6V_8JH%>VjQ{Vd?NpoT!x71`3$)Q+S-DE&1j69A- z?;&l@fIFZgZ`J_sPIxbJX0|JCdE&l@OtDCFG$2WwrdY1rrW{!joHC$Aks<^_7UppB0(+KXO?jJ8_uBO1I=1X_iE@vC>B+?wa6MEmFZCQJQ*(B;0Ps?YaRJ!_mb$hUNssKzyGROOg1>N5h*h< zBHJl8hk__c)Pp}habi1i z!~|Gn6bsYUE-DHYDLw?wR6*sNmlhsZ?F@?#9P&f1viQKI;kk5o_|CB9lML0Cn4fPG z;03YNHAM`CtR*MN@=2FO7n!w_F1f|KHn`qdlK!oZHv*%wvK+fcO|J| z8ex7I<@4n9M;cG&HU^{lgwb8YX$+j#bT7YDxZgZe$i=yJ;`OwZ93rYhV5fCj=w9$- zp&T!8K%r>@CLn{H=;~L8Tu!T&y0=I-_;*9v`y#VVa?o$c1)KUlh0~9`+q=-A@aG2j zVMAhg9UYWyyDGJ!w#J<5|JnG8^CF#CA@pDT8NH5|&pR3x$LxvR5^%@sj{F$iKrc=T zZWHa5)X|2+TIIm}%3#d(ICSuz48%(Skejc z)410)ZGs$CA`jEb*LhbYI(Qv&X;o}hb%w2CJHtRwc-AB7DT3NX=siKoFNfaZbr32q zwaTW~(-b?&aUfWLtnn5vLzH&cEYK?QUKM&f1kO!`gmbDqkG>>n=4rB3?TR~-*Fc(D zC+Lt5x!~l4f>`)f4CR`+G>U;5J_&d>E2ia#?PmI=n%amR@MTeqV8yEypfH+CW8q>Y z?*#bWmh_Ko0gve{8g*Bifkj4XA9B-_a3>^h2FOOe%7 zR60`_P$XTh*aQJ|gJ=}C>BG2eQg$iMLtWFfD8oRFU?E zM+-I0Ok?Pg?_6A%tvaaE`fl`YlpPLCd2PRs-Wk5IpjRCI&8rJhlnv_?j!2LZbI1iV z)6wB5ae^9KB}^xX2(|&Co>n!$Gq7e3xm;LK!gSDCdNPd76Lpb8E_nMteel;9;Q{}$ zd8Z^LUZ?c8J1jm24S0J4F3i&4<`Mo@-)t48s5IJlJHJ}^rdD||aQ)0ACP8>-A{JYh zLL26C+1X$u?;LX3$gWh>c{h3%1!y{%E}7HTlm71aTcgd*(wAg;WyQMU@~SkVT2RU2358R?SP{Gn5|)CpZ~M2XtaO--oF<8!DyBgISL{9!iim@RVIdJh+-d6 zG!9?w2U4om6O(_-o>y>~2?zfL?cTB-a!sM-+*(&@#MGogDs zM!1^aMHf*hB3U3TqQ=)OaY$(5$d2QGKFMp>{mN*Vj=p=_m$Y;9mz+2N6l-F}dMWlU zMQ&43g~CIkUd3mg+xQKF{ZlKH6(HtV6OhlOD7O2dCBdWJ^i^q$+X?qfK62Ke-N5s^ zY1~u7Hn3-aSr`cQfyWku0L|)7{yBaRFNvw~xv<~>OnisJOPLC{Taw<0N0DDZ6fw?k zjYo$%-X9rolV5%Mxp92_m1!9PmRDc>z`$=b{5-Zj_B8a{gtmyGQzI&*O&f6mOJr$ z2nC!Ys(v?8EcB?ah52yO=YeNMH|b&W*U*?B3A7qIZkY4J~qL|UW(mKk)2djjw;8iOV&51#A`>`ClhysVMGu+ zRhG1pHU3)YCXV&jO*=ceO@ItF9rO}B-prjPV5oqn_y`Z#crTr zWTMbM+*NL`0AN|-tyfhby-lK0?!=_-bRh3r>L^NH3n#L!;-pC^T5;SuYA8| z%Q>>vm6Vvc0NW@Qn$WgTQJZI!iu--G(>44C&uHP|#ad;I{}(hi7Iux;H^AEq;;`G9 zLq0_ernw)IHDo|mD;~?Ee(}>7!(hDl=3_X6+h90tYV_bI=T?rT=hlh)vaR&o?)K8r z$AM4!1iNW|yP_)$s?Oat*gsVuhBT+V7&>l;T#&j+t4#OLh8QuBXEk_Mk-K!4Oh@bJ zdy?Tir^A3h2F&XX&lq6++)KZB@AsjU(WE@~czg;{G@O10P7IdKCYCXYVpmcm0T* z%pORJl^L3*C=KfNHxyg7NHZKfyFD{fFCOV(*W#H=9lC?l@BEe9ce8%sYQ%+F_r8Y2 zaEsPDaTEnafkwprQz$lxA}gt=>%1l~6M3^b!q&~#lm^txj*;W44wb{E2nermCztR2z8Z>$}^%as2u8%HwUIo zYLHb1rzr+J8kKlf2aSu2`;iCyGbbUTPqqrFTMmR-DweeE0ogVT!#`&3ueN=QXV2pF zLw4Q?d*O{A9`rM!>32JiB$I8HqOndKRReLi5q6}MVnKdv2Ra;s@(gidWT&i7jB)yG zRUW+@lD*eq~BAPAOHPj9AGLZYBH zFimlG$_mm)I^+eU3-XbA$uq1!Pw}YqH87v`@CwMF9KTJ~T+~RYgkW*EEM02&evQi2 z{FI1F_-?>Vk+0gP=T}4Z6+ERtyv280=mDtI!Ic)m)?P^CZf2iz1%59sA^r(-J-(P@ zt|#|w_mmU+K2}7XVd)H!UnD{94`~g-ZB*94SENkFR~&SF#rop{q1_FUWplwDG?wpU z@4ePR^ZMWZRz-}H+HRN3lVlk;x7m58YXU42Bc{9c6q`zsl;QO?CK?TO-LXQBrx@vQW#*WBTSS2RfXj;A&MhjsEr)uPc)69>)Trh1IGgxXN(d{)zam z0}8kLzzd4;KLSj#$nf4JS|DduC;gEmax-5} zJh2s+n6J$gyAgjEQ3cSll%_B?71jn80kLpOWV@m_q8IqshFowDb|7Y}P)%a=QHJL| z^kD|N`w&_AIgWm}#Symdv|6ZR zaNeKmuCI8~<^-W!<~-;9%ZL@~P!zK4g7rL74Oaysq@l~XHZVRAU7t;g?ipGo{*r*% zPi{Ck}k!wQEJUAKZJ{E4Nm< z^+)tYT=Uu(a8abAGa>ZzkVctHbYE^uzmQy-k_+z1HpwoYj&7q5_+vF_7rl(XB>*{c zp6hT}du$x`k9y=ee8y-_j}Lt{9YG1mQd+j0G(!oI6o91Fc4jbk7W^4TiZjTZIZ#s~pKC zJ8Gclr^3-dz_x2R0cApT!M`V&Q{;ePz=+_;YKmP&krh-F_9~)y#2SxBl8wN@cG%-% zzN3GQ7XM^hp1j2i9lVp$gI zJVUpUfx}_E7B!|A`%*ylmA1y9@|T?%0M+9Hf$L<)d;9GU7-BTnCIlJ$fvq@SGh zsPZy&9hCy%YXkVIJKz57Ydw;6^D*1s5t785Ak99*eMy*$+Bl;_l@ied{jIpSn!d20 zX&#E?;jIn(P^l+J6)n8!UNH@3E}tggpvanj2T1dgHXrX%NS8p=`MNvqxkwU03$~K( zCecBKP_C?JK+$MGz8h5?b5sU_K5VrcQf7wYCv2RR8JffEP#L~evoepy@6^c@F-}WE z!p&Gdj1KcPCk*AXP;%ldmKEO9ETDhDAzrg+@T*V&EzG4ggYrD4O>mlM@K_#wNR{e4 zoRKuOIu3egi!mFy&Uet)c;>e4e!pMz_f-dq4V0^ydV0@{)sV5a7QA`pLD_r9iveNp zPqgiR-=8+&pgF~n6>gw5K_;`Cxe(Y*?~oj1FxuXwNQum3vQpwmk`Z zoyEyoJ8veYVZ-Jw^T;ljX~c=+ZB_!*P^49~@Sc39d^I1)K*9>d=t5v&P99wYB41c3 zQy_+uis7UlUcRJ?K=3$SeS$3z_sF_LJrP(1bI5-Q9xJDJxK#>PfkwUl44i$g^J0Yk zWEG3*a(GqxafIUImq zR;8fV9et*}SqC8M4YZY>OD>YUS;#0^MRF%D>2R~O*nF`G8aX35XxqZC>~2L;$=_E4 z?(uK8LQ!C|8U>a!d`^&x;5DR4_66jxKM7hse>c;ms1WB(t&-LYERCauM`(v#9MNHi zPqgh8y&t*@XB(H;i(2y&Nr5YYTH_H>mje{Lk0N`hs2071JJi#)NMra1{Pnzv7ZwaJ zMjB+NNFF40G&kLw=EVxPiDQL(1BSCrT4rxRq8kzzCc2^CaF1lO0B(ef14B_9ESZd$mw}XDzJ~47bjF)lY4DuJD(@w{)G;b|@*;%XBfX>rVx=?a3 zpxx)OA4L7r6c|sze4$o}hjfAlL4DwO4K<7mx6!uKP}*=CZYC}AN2%Lh_Uz#b_G4y% zu9-dj+NiD+-6z+fEUyms7Wm<@IkE0>O4JRXJJ{!h* zET?e73Dt10Vz;?Ew$t_{Skb#~SKtf*TGr9?3j+-K!e+86RC7s$v9khU)$~ijLoPk? z)80qsWbtz>x8(7UCL3MYmek7CG&XLfK?4=%1%V))Ws%-$dxEoPn=nc#X_P;BCZyvsVsTz~4 z#F{}io3E3$NLP9cx#U30TrNyGiEb5qT-&39fW6URX(UGMvmGoWE*ghT>tLyKUB56bqal8>y&mp{pXXrMt#w7qiZzo*wQkOq;Zxi4k@HMEc1I zk~KXg;vqDOpY{A??)bpMb~Fsz2uDbxKMODGXs}f2dtiR#JLY-fXHAFmCMbxNY;Xph z5mZN`K)_D<{mAV!Rzd>hHvHQu$&{=LJ>}UnuS4GM$a**Pq2UG?^Q#^A-a26Z$0Pmm z%TEu3`Z?%5BrzInTka?2^JV8Lj-5_9j{MYM5R z>HD7V2@-3Gyupcce?XfzVtKQMVv{M7NJXJ^*p|rLP|Y&$T^`NK7Ei4*&!-KXkMWsN zMh>-&KpA<-xNKVoN>ly1P;*8PDjFuWy9`+V|^ej z{9>JV{sxYqa`VX9v;Kz@XJD*E)7C0tgw^y?f=d?c7}KD$S|G-<0?amIj=x0;B5vgn zbwgecEKbU!zX-EGfX$INNK#NGfp?RiA%eqO*w!d_`lL}B3jJ>(d3)TcrD2h;u7ET5)0?OiVzJ^_O$mGmA--b3{@s%2%aNzUz+3!96nup5K3NrvVyy8=FBi z@}hCr-Zt@m#qW(qC1H8eJksUFMrEam>eXY49iYg4jI$O9fo%oDVFr1JSa6pz`IVw; z5JA-!J@-cFE}0>Tm_P=2$3{iGGAb)zv%N7+Q=|HjQC!4r(mHmj zgcTA^+=-Ty8sH_j@bS0LUMygMj!l0GhD z_|4z)#HuAIdPN>kgS5@K1ZlX!#(ykRrE`$B`shZ$#M$KdY;8N;`UqI?Hl25&di3WfcKRFPQuq3r za+2-D{?1_&9Pgo65WXm;qIwjo*lg8+N4pZ~FFK%&y)t<9#1>HoT`^7X?lc0fF{z9@ z!Kif6nf!9-94zo{4TfgGHoBg7UIYrfFA}f#|{6p@7miAcpe6~Z9uj0 zEKYcy@cl_wzi!U@%%uUud4CYr$g7bj=(zl%p3v}VQ>tEJ0(po)ywQI(S*h43)AYL^ z3abp>Jmn}51fc*$X+-hl)qG9gYdY*%1(FVAoVy*OgM#!lh2dJgUB&WDw8RZERBpt( z=;U3pRcvM`F4EEzCzaLu8uromD%ufy!x%JwwGnNmr@eObn<8_9%x9y0PQ2o_LW9`> z9R-FgbgFM22>t>;?*UlX0Am$Wgx#2WXDqJUH`;F3sl)-!|7r9b|MBT7vIB)TvO zP#STBUkM&d15g``<8X5sY`ev6ZGeS!17X40o>24^?>8AEG=BF<-!bx$D}h4Q5ev_? z6uX)tt8fGoWoe3~p@#zx39d-mXJpXZptUl7=B4p8%f<&_>%+ISBkcW*2{Y!u8-W44KiHZpfS-P?>5GPHAw#dDDdnQ zog4^Q?AM!U5n;9>*xYHeOXZ}ULi_hvP6yL}-_7o@6k!4Cz|2lri~IJP*fSsll3)9=0Oj%QZ5;TdP3 ziXRxC_*hVzrbwRGtVRWS$IL=wLW}`M_JN4mT{t0zI{TZ(tL8NBP8+nhLd=d@4%p&+ zg6se`T`aEF_X}#SdPUDK6&;fw^-1DqiZz%0u0`S|)F>cvAO-~o+%v_cA`Nn_<8l|- zc0pb|;+ALhlugJ=+UZkGUsfS+!X1+D*XB_xMJ|LkQ7x#DW0vUDly!Vq86mMUnlHRh z4TEh&8FZaLzT@Zu*#T1)4V0ywX0UngTTYmo@c7%Zt}vq&5=~sVjqGwIK%qaviK(Vo z`1K#4qEKzcz>=jOvM#1ohM})gP>@Lrx~^CWrBtT_A1FTa)iljRs^m0?LSzJG1Zj{E z3rV&>n}|*Y4d};;rg_*`5t628ki`m{J+q-I1^P7d=t3b<#iHa^mwzFh>Wf!$92N`D z#$h96;j_5x;x}{$BdTumM@iR9D>ced{biYWmRlSWjn=^=+&Op<}IFvZ4#bCHo2RG4ZmdchP!nGn~eDPCg3*jWKxx=~98P zX(8t$-*JQ z<3O#l9lD7IL6qpEvK!Ju4FYs*^O-J?Ype-K0Wa7gkQyzFhT(*-BSznDo7Q2}{-2+2 zHlKBz*05NeNAxV|Ydx>{Vnv#M71JO?LN_!yPzDFECFpB3p}#$(i{25Y?+Yl6z{GFu z!cOE8L}tKjRhc)g&ovL^8)v|&0xVa+B%`K}R6?HFAcd?sE{_o+<8t}DO#UXt*~yw% zVVO5nPkCVG7%5M+$})NCbYz3Vdf;@~kjv)*gW*UVIBJ-3h22QIW6SnvA^Gn&dT%Z; z0;%zbaXF;i(!}J%9_%?2xO_sfM=A30FfqG9imDQ#F6+FT#pf#7i#IWQGsHV{_^N*6raJqvuIQc*`pC)22G2hqlg zuuH+&s>%~r+g@NQMJ}k_K}SG9q*GQR&sAw|c@}^!fyjFP-}obk`c!tZRGhdWG!SecQcn6 ztWDnsI@730jtxiMq-r{@)T;$`&@b5`0DUyL7BixF20GZEkkt@IhEL@cN;b}bd5!Gj zos(>uQt8pG?v&*-7uX&J_Tu8*FZ-oL9uBxo$1p1uhXe878*}P_nz@eN?bWVGVs3?W z&}WQ)&8Nzbf+~YkeOJuAEj#Mn!y9S-#)v&LD30;%2sr+jS+~u+@7ie>?Av25@MrVqcjVurhD0h+y{^G@-ifWjh9-BOLf=7x~lQU7`WyEuE%;qRZ z-7#Ng#9ghsycWDz%Ksmi`HGG9DNC@fl_WWFeFT(4j7XVkDRvV@HbBFX;(Ev~;D=hO z#sG~bdXlEhyNZ9>TT>W%iD0S|6~;mPIkE)8d^$Ro|M%_JastSN3FrPFdtU<{X#WGA}`%!SuUr!3%9M=>DJT0OiPR}Xj$c&LvPRA*?Hq)%5T2oeN&vGw7m ze$`}v(hj00nSnzQYF&7#pj5Ej`)NcDlxq#SL30Y=P8Yu1do>R-?t1w{lIXQA_9ArB zs&j%{q`3Voe(8i|{z3OTf*U3b%5fSbJLa1a)@2+YzojYJH^Yn|!OCX z#@=l}18zLYT;f9Q77uflr@RXyCyWYP12%EA*t+t>wuguO%n)KmOIXfdQb`dzw79Un zX|zDgUW(CDq>_rSiAqypwL>OT%iH3U^tu{qb3jlAJN(g{48*2!u>e#nybR$xlu8G4 z1eHKwkZ=<$vN{zXLEmQO94y!xm=4wm`K``Ev&vm*aoAxl3i>(F9hx|BMp~SSfo0^` zac}>6wKb6)oAtO00|%skNBBMI6tjUM>#^k*S2Dn!dnjZbw?b&FBF^RHDzV0^Zobnj z$H<31;xh|ej5y2rA8TqiKCx5G7hAcEb=Et4GD8MsrmkrRzB+dQW9@pDc3;hL2eNV< zUH66hNbHmTHS-*E^v9P!CXFt<=;^e0`CBQbg@W4A_#L1tRqJu)EhJ^pgLGL3IS>rg zEkU~j&q>Q-HqA){zx4*QgK3K^1DVX5aaX2do#>I!j_@2s38#nPIT}SSP-PZ#j?wMm zMbN^ya@rD3mQNXHn`G6rwzv~9-9Y*NNQT$L%3VE27oruxyyDpw*GYGfu81V&n%H36 z4|=SsX5OcMjl2eVO%&@_#xA^%oHoY!neCgv3hz{}hww*ln_=~YoA-U9b75Frwt&@X ziaAMv+E)A{`G)BGo;ktEoa^FC%o3j}UyRzV;GI$Rakp|(V>it?7Sa}nq%fv_dt`+(@KE$%c)0nas8l*}pTw)u5|Y5*Q-y5@E9-UTS=t5Gz|AUP>S#)(6r z7|W{y#K@LT>xt+nTFW;fMN?J^Y6vRBcYC0{U+p9WQaKnoDHqO z@6beYl}AMTV9s6Ud7H1^#u#Q9YbB6DarvB$;eB#-FEBdvN{i_%$z=F5Hz%{J%KKP`_U_t`C%TzGN4!6Lt~Y$~&4DjB5WAI_^@U|I)Z*L|-vOVSWn zqQL)aJq+}cfDaT|K3Mj!Av_C6W1(*%tX0t*j$)e{MP*2-s6?CuoMXr~kJLjbRMAZ* z3QEM?G?F@`%MJ7%P&mPgQsi)GQy`xj)|z6iO@+@AcMKt>fCgt1ojs#A@=)kHZ>+kMVk&Rw^EZ+OE54CL-Mv3&h z?u-7(k40yr&DiM2r7#V^8QzG~>JgPa7u(xyHwU=0B} zBDF4L8-w~<*mnm?_Oa#B1*+t5Jp>I=3}v-ETfP``>nO^l?iHsi)8(D~Jk>#!Ms+Z# zo=y(mN%Ua1hTJYg>q8!pB>_cbF>RQ&Gz|NR*ZH-}K`D6AIGrcQL*DvT7#r7P?6nQ~ zUWwZpYhQ_tv*o(0Rba&$iLYlodl;?gmadj!#3GMu;33A47^`l+X|h8@dxy6wbQ^tL ziIsypWvAa5KO)8$6zg5%v6t4@aK`uPys5q4F&mmyzpi;ePPy>p1X zn4H2O^WESUPQF(KcpwJa5WKidu#Fjjx%e=*RDiAqZdC7GV33@en;@v=_4*>IRls&e;6Xd#6G_jF+M<)%{?n^~AVFh6|VU;0jlLOqzH`OR(Ou9o} z0x>F_=nV8{L*cWA#Gd`*fs;Mgo#-)LvEBb3VKzW{0Rp z#DHSk>OA^g_y5c>1ISIJ+D8i6t@d5Gnh+XiMmW)XCC^6Y(-$ z8>g4Xef!^fB5sT8p4X+i$-NrX2i5S&jxG~(9HJ(MAnUWYJ*;G=nQ;?mct=f^xUlDw zV}XQJibL93!kd;(laSkmGkT=(shYooiseoyKs*f{`3TRj;Me)t}& z2EcU}oV!ooyYjNLF=B@$R!11?BF%B?CJI zs_T?`Kyy=uq!jp|aUFzhE06|Ps`0FT&*}x*==Ho)b0O^8{;VJrcVV0k%Y&Eu!Rd<= z1zi!z98^{BrXA2L#U4-gt!*Z1oNVK1dT?#^L{z+7cy+VQ!lzB6m{f`+Q}H{w{p6|N z(z!bqbdh^Zn_nI`&uh8g`k7b6PyHsz>^z?i>;=s8GbZQL@vr_t%a|80o{t}1BAeJP zdR;iIw#Nb%#S~LO0l9HJW;aX3YSiNzCIjd~>b|{xX#v#`JSi1rl3ED88G|FpZ#5_@ z73HgR{9}p+2@phJGTWK+mc2-^VNjd~o7Mjme8akS+-1EJ8`a?bo-Ib|PW9gT_oMDd z54o8X*|z%L@#`nr=myUEfc;S0ixBx>Rx)R?o--i6;*m;{e2%)~61f0Uj_bK=<{Qyd zN=#G|I_~;_WgMWB@ZIa+y5I>y!4e@31cvQdX0)~NPxk+?QfM|YtFy2Flq9)ue5B06 zXl$jJEQ)NR;?p>3P_L5Xdn&S2)GP(#kRZS$kdZi4{y=tGTpWgEX;4-t_ylDACTV(M zj}HfeV2n$#!^bzjU#@=LjE}cEO4gB5c2kh+uJxhjb7b800LAR1$ZjgW!T*X!yJX+Y zH5`c4qXOlS+d&l)h2UCRM-Rvk8tI0N#Cux<@|9T41nX>6eJd3;5-e0mV(>`yvY^x8 z7B9}5by~by+Nd<;h0sfeoRFxVQ{gZyvWJ?b^%>>c7yz<=*YE#RuYcLO(Qvw{o34S_ zi7Af4|APm|q~#290I_H2jS{sEiVk*qAsRO$Es{+4xj$}5sH z)>v4U6pBfr$Qmjh18=3^qN+;;NT!JFRZgN&{ETt53qYT74QK2wzE#?2%{yd+um>40 zFf5NGg;08dO<$*GE%*D#^W1`VNsb5@o<5p^!HD}&Hy!Nvjqwk+>$zGO+c7TYz+LQK zOqV4BzqM}gKR#c5j#%E*E~RThkBhic7R}}5li?Fg=i#Y5b;=y z)%du7WHoFX@L^(I8wlw{VBHypP>0;GqCR16skmQK1|EumJ`_@>*dg8lrAX+=Ixs+D zHy|u*-dJB80~l8JZc542!+lX^`26U@`8&w2S7wQG#sX{&6a)Ow2dQ{$tVd=|Q=t;n zS)-jnwG{ZM0|{9Z(T83RA$|ph{f#nAh^Rs!#M40sTPZL|V@KO|W5;zFY%7Fdep*1O zhZ}$qQ>ks!XCYdg5 zf?%CHqTr~6VnEj~--HVFPFVw}vu=wUa@)y;!nK;v70N_GPEfN_-Gnv$s&#(JqCAdz zeLzDXbgp=}k!8}I&88zq{-32}&nq+id}x8{BNTIpB6U>!TCXFJInEdBImw)25AF=yCA4=@& zL%ZI{4C{uuX|?n;QJ;i*{w;yI?nq`(9cl4Kjlf&dPCj@yAb9aKCOwCm($!5%W)Xbh+AI4Ip0`Kms3ho z)7yFGSt%^@yT`~17j|f&c3^}a zRPYIL(n~-30I~tb=vQ#gR(8O6z59Dl{>BW9_OIVABD-H1cjc@Fj2bDXfg(C8K8OAQ zcrY;sR6yT?JP@pGPe78dRWbavIBcaia<`NOCCtr+bRnLAt7$wXTXFOCPa%2OEZZ0i zy*`jTGSH@NYo=$08u*7oz_a1xc%YV->3TWzdbq|>cY`Dc=CxO~)8_}x`_@~R)4o_O z>AGIKt-ioqPKW&V>yJQo1Zk$QcU=?E8eY!ZI;%3IHDa?*J*|(enYoSXTX^S3$mTpr z7Kn}QG5M^{u{G9LwQYNf0~d;|Cm>kK`xsG+H>PSV?50hQ1w^7ah9xQHvzeDQDlk51e8rNn<)TO zd}i3kA-Cu=sv);){`a}vQsmGXkROl@x^0`)>uW4X0B&!=R^LfPIt!2+?a5>S9Pjak z;`DD@lXkmoA*2l&?gz3)Zax3j+;h?aPi+1?9nj}hOjiK;;APcgqNO*G1F8aO^laug zax*zO^opot(N6A3AndMIHL9v)>%#}#dI&~FwzxM75Ag@xS|f&pjehufDJY{i^7hI4 z$@=j7+$-WG-hDm?RF0X2*kZzt;D|$D!^-ec@~rp%o@2Hzzdru(7+K2BzPNB4CC|dX zY^0bJiX>6-EmDo5P_T}hE>C7ML+-t%QFKLa4e1TuO;3J{sl6C-2ryV&kFNVViN5shH@;;y6t|-q zJ|V~1844G6M7k^t#YKuaPmvF)cpdNtU=-zw@IXvjz&+1mNC&HtTL)P$pu-ER+HQ$+ zJ=EpgUb+(4{88no*5eR4ChG$I;lrS3R7?X|PYTcUV)VEt%x#fw@u`}R$z_e=oKLf4 zx!+Y|4@Zj>S$R-E054P~xXS5|NDf^Slg`bZrS9bSkdmNvOeLp?9^Q3IF(9w?FdbmG zBV^YrV#hpX{~>n1L4vDav95%BMT#SB6!c+azEgMv3RY61n;|$?P8HLcfnBNrue%W# zu`j2}DdbJL=TS~gnlWVy4?72U!iTN_2NC67a z=;-6~YmJ2|*lUV;yb>`;&#cCN`Q&fy3X}*W<1OT zWjA1K46L&@x4$36_c7a$_BSrPL(*J0!vmB^BQiV%6az~c4HcgyIZ1Lv+K`TjCAA7uzTRGOoX!r zeE7PyEyXBtKq{@b`F-n!&x?V27q$^LR%zHXV5)9}WJ^f02zl0=4vFC*82Qs4n2bDr zvOe2}$!%F_!z{DqP$jLdBU{-md|WuI1FU`{?92{|0dKXKia!x~R&s#84+78+Ip?Ie z=-q)meEhCa^zyo7_X8U^2Sb+6=#44i-1odR7ezu)GjC7K@xX3~)^*8n?@4k5$i9n! z#n@Revj<65w#M?Slc4nPL#!7uugDtL#`4D0&hs!hX^z^YxL^u<91Ph()F`=t zv6XhofTuS4FxMbSm{BBJ#;f!NO;)jx)cg3~iHK-_4xwAt(PWwJq-U>w5XFg+*i^=cxaNFZ4wZ(1UJEs3KCK zN)+UIn$%gb`k)n5h&FM`#GRZYoO5w)ao5GyIKV~Hc%PYbJm8A6LEIiD$A*e#W(wGll4*WFIDtRq?yo# z0cArpvC8*Ibg$=SqL-icITGE;&;M%I!kpl(05E-2q3<|{*ZQ&A{oQU{*nLIE9JuQ@ zvY44>=l0uWKmIw{GL0Ot@CwT*1`_o}RD3#MNy z6;Um^WntC-WCp~Kcl~65w7g`A;7=?-(m^p-Dbh~G zCs4OQ4i}cPkw#h6OWa;h)VuryGCL-(`6S7ktxk>B(P{Im6&1n*GY|WzJ3tYxUUkI# z08p%{bCoE_Wc-*gU9ICirxV3o@*aO=U&C!XWyr^d7ggtgUge>{8xf@f?9?a;dMHd2 zWg2tJm^}inum0ESUpwrl1EnbSX_Do)2e>qVo3t28P;N!#dDcTaYJ%Xr^041n_RSvL z4Z9pp_-PZywsH59%K!I~b-}gEx+*qGsM{sQGnYv(b9M*ih;kqdj|y^~^c8U%7h`hC zqLn4fOb+%3vf zCNPSw2MZXIMZF*a-YeAv zBq>2JtPoVfK8bwdi40D?;a{|Cdj#)plKW9k7}zVA6c_?U;#nJ5|yC7-pA z6^f?JkZZU7#B7s()xO=2+<3`6ZqSk+QR2`?F+CLNrs9#8_$V0!qCp7NHG=3|X9RfB z@VOH{BjkqMYTw)eDul`0z6j`l4p^xifZ6V*{16YSF5S=4_dJK(vLyAvc$B(Fa?Y=O zRw2EaF_2a6K*-(4@8y0NGvwAm?sysL8MNLz;a)m0ew`6oxcKw?|K*1KPv+ah$M=Wt zZguUmKtk3h8vrVpow6dJrW$g?UuxNml}h8y8+d1D)$@wOcKMzb54jC^HY$-j&paJE z6N%%U91*3qk@)miMLVtArC*Wsrwy&5BcW$!-t}EQtyom-f0)jgy+)ENJ|kN0w`N*f z+);9Kej~XRb;hqnIOw*XTcGR~PRKNUmWg=2pFMP*pXUC^icC znq^ucvcEt*PC{(m{8S%IMNAS3EJo5UNLZXPIbXC@-JvggilLL=4O&r$LM{t03(=!B zGErj41qq+``XL9UM)5Ra7bkrZ%dqSk@R9St<9`1d!GDt0x#E%tqotAW&~ zLxn{ic#dfg5;mi*hN1jvlPg zMnNs_IJwVlmDWX_exogJ-`qUW4c|ex3~nxS)W0=u=c4Q1Iu|$S_K6g~ZKw8s<6K-$ z@J+6czN+etaj4aDBy{XRgx!Y>EHQV|G-O1?znL1Z?(me`Z2}yQ{({^Z{WA^ zbLiSx$Vr+30bn&!eJy8->61(!H;FNS{YaJu5o8_RF2}^gauCNnPL70{4uSj<)&&=k zPCn+kaA!ZMUeEvx-HRX^C|b15T}Nw$>eceq^5Fw8wqukZ!YEK7T?B`nr3bh}ZWyTE z;gP07@@~A6HX4r@ay##Re-{3#wcI`^`b66W5l0|XRaf}*2KLDY<^4Xr(kkCuq#zKv z<)L5)R?@icH~C@EM8heh)ukex6yl4H#M4-&iS0hyOY2d5qba?Y3-4PCFS=}~(1w&^ z7k7`}P>|{G*0^(VNX)LIFK~M4Rh;U8F4dig-Lii1ZgE>&6NH&f-+H1n0cp?{g?|q@ ztxv^d1SWtKyRFE04rGNC%D=SrN5N)0)%5m;YEtmZOmRTgJR(rIn__lRqyktqfmyjN zE|DpDJqx&pE^rL;CUVcyAju3|^fah|UWA@fB7+~>0CV+pGryF(jHBTu$EJ#sq0;=C z{H(u5wJ1SQH-C!~*v*r}FAAIJMFa8%&M~@$J`}phffL*WKsJhn84sLBKW9triW=*1 zjg2@53N@64R74up8HUx!9QcSv{{=_cYS*(IW=rz$9|BI2 z#Azhg;uYLLG3zO^4#i?Lz(|Tgkj29?9wzmUiL!;VA1b>*0k0DKD^Kigl3lWoL$Gwbk=GiI&HU(k zR0w-!oPo;G_io5{Lhk7>aP3xcmIw+w*L?3J+2efv274iGJwZ$iu%CUlZAh=l`C+1m z86|&IpZ_gcH5y7>*YrU%%}A7NrkHe!Yyf%(&Nfb=ZwhZGBp?&1C4zSi&;yBPPrZCj zW~i|eW~c1*8}ClqY{(WGpFOCKIl?wHrYryXzlj+dRc^Vb$VwMpM;BVaU=zipQ6v@k zPXmV{bRk(xtw)kFTX`8uV+vF)?xg{p{72APQv81ezh)5GM!{%n+}XeVws5f-8Cw6- zGScP3$XIR>VR=L`_bGA@w3;J#DVt?WgG&XN|5?iG_0-GT$RpAbu3mJTSHd~ul_NS5 z+C-nF)%aCA9lsXSu-t&Umo}25+{wKnL@v6OGYwD$3_*-^IbKtvSfrhfa#jbFct)RW zw?bV^TW)INEE;mdgVO>`*McJBOywbQ#_V#AF8qMk8Q~I-E^PVJg(r)igxz;9hiSb>ZQC_Bf2v6 z8}X$)>q0CxxoQ`7cWjhxl>krS!+GsYJ)Oww|tJCiJYt;Z^k*4P{cC3_lZC&tuJCw-Er_j|U&qrosL zkE7scvdDA^jRK|X7cYjsV;=TfVc5BDK6X1`=m59(lLk(j^nNs|Y<2QYXLZqdC(ZbT zCY*?k=dc+EO4!YW)hCbc{iPX6Z|^uVgM8$|NLpnv6Anp6!)QH^A0tWkv=>4}@&^&Ishq7|<5KozasK;mlj zv9T^4a|XAVHKf*KxoWlG5kEno=WOxW%z>E_^QCI2T<4qaV^Ubc$OI%i0M>Xj(=AO} z3>#;x5vcKYJ?=T8I$?$IFjvoMhE8H^z3Jp?g~w#Z;tteC*9zfJbOvvkG%a8aIRXE) zm8!NkiamD+soS89P@~8UJwd-Bli`_2f40gIY|F<)Ns1hD}n&wt!LrKJmqxfK0z)XeQW0P(s*99dk^!#HTU@5=r>lr|_`g zxAV&;0|THa733;mAoX(WsH*cFkpB^N1n^c`VYa+C5Y#_KLvEe?w15KDUOzkt`?c}B z@k}ePgu!s#g>-CfLh^Jom~Q>-%Ad(C*b#xGOp1Yg+(r{)?24%DfJA{l zFyE^}2v9DjTii=j#n3JR&3Wna11fB+7!O#-G4a4DEFnI=n~dWux0%xRKr=QTeqXwR zY;)mRuHIs_t0-nC1&T}Y%XzJevqJSA(g(y7-SUgjQ3ZAlx1FW?2`-?JV_XaROFOyy z1Ztehl0`X!0nYk>Pa)S0O;_^MGjeDxOd;L$m2h>UpqTELUjfuOpbE!wOjyF;xx4ei zC(G+j^cy?ZNb%sU-&rTZZRDFVHH9rFM#1oWuU(uQzU`7rZ>k&opZc{+)_RpMyhV!W z74Uy&L<&*wgs+A0wN!N62NlPS?rh3y&U;TMI3N>!?l#Y7H6X4#_53{gE;rO{UaGox zC6Z!xQ;!QT>W^3$qkR;!n!ebKNUAophr(i7{?wIJS;Fc*8-v2@iy-;&nD5-T~NcGgYqIU4VE z#;=*5`bML;n(T$+j>X`FWE~s@UHl^#q5G;R9oB`RkO7EtB=SPx%ymEebi7;Tii}=J-toUEk7H+g!jbz3Oy9` z{!@r>qs7=x8~ADS|2XId4H}gpIl=F*V{k)d$cm_s=Vkfe(N1y&Ck`!^n6dNlC0}pg zD9kn}?&J?@NS+JZAZVEwu^Qh+F%=XkLq^Oh-?q3DF?FPy&XAOVLlg2*OSB3J&M74exHBVy9i>Zf3nL^IRtRlv)5;Qw+DE(0k_U)=N#**A~tDS7xv3- z)EhOs=Zo7VdqS4YRbv#RU4p3%H4?R?%d>+2wAEJ|i5Qr?DoeJYv1h;?ag6_8s;5mG zo~sP+sL2u+o~v>!Y)mS}BvWK96@N^e%)vCNdWp|*(knI+xnWnaU2n4^9^$iW*g4A2 zj-RX#woSt*^+Z$u@YL59#t|emNJ@>ly`8 zc(qINJP!mS>9-+Rov*x0mvG9tZGQh}FrMtr{)NO04&goFuSm*hj*1Jfcfh|L;qc~C z%vOqI0R+B|B|90jH*oJt8+duJ-f8A1b9#KQglj>>dku(nKZrgjZR8HRJ#U1W6a!lQHiK^)Dg;1~}m zh7saSaM%XLXcJ8ymAckC)9@m3g$sM1Hu4ErOL~?z(3`%nIZ)lfsTJyC5=5I~)N7_^ z23GkR=p&(J;>Ee5L0J)$t!h-2LBK4zj8`T|S&%X7OsKlX`zF-wlmXjSIs|^(p)$IB z;SF$G@6lU*^T-CCIuB%+i{5H=nDF65&@_xHi=PgHm~Au-OzXXCy`r(P6nN+XDaj7` zHq~}7kTCEZ^3qOUErrqojmp$~wOX1yxpn94p0=X3M+C^mZG_u#i$Zw|@=ug;2#RKUkyjWUjWz7g%$J54ChNOVRU-g3XWR;*(xJTS9 zeHxS+v@iB44<(n%>8pZuVVgt^ocr9HulM+-1?-C*aGpFdju<E*x|N?c5R0%SMVxp-2)oc9#k&p=AMi`PXqvd^;i7 zbs)HqK2MzLlstpS$Pacxc;qRQ^4T_oKb>}Kw{=JND-tcU(U^Y6+sH7B6e7caCX+t> z`FPs}mgk30+NUl5{*ydw%|e%D5wW2{INV>36C8Mu8+fPRf=F4Z04wG&PHxz>$?m zJ{0M=tqPPB>LoQQbxuIO*ZrtMLB1D$g<=NTgNQQ0RVX|c7{*boXBR{)XTtG)V+k(X z_GC2g>1ynaU&Nbdo=4$-kde-pZ01R@sE6sL7?A(?7)lp_URa|r@#z<+FfpjUNcuf< zfKT5*FG}Xr@{UHXoQeJ6+v%F%y4i~w7ry)F^;~qrDxi!I<@IucQ+R*8AzuQzA*MQY zAR(@&Q)5g`E(NNSBqO?#ldD`8+YzzqOAllx$&Sdg^xa?psrj^1pksxViQxBdf_8Hq z_c)Xd7eNP3-fZAa0&Z+071J(GJA(|`-SoD%)^pWr^#_V9AEe_4RwuY55T1Z2ARct- zO-K~Le<(kXT<(3*)eBa7yKDuFZ6My;{=+NQ?hc!Y?d5cY7XJ~-ukyv{6D@uZ)_quZ zwM&#JDf;1swGACCPjSEL`Yb+K)6%v`!x&pder&$ z#x%~l=~=D1?v((sUMQU8UUUvIIV-bZ!;@h>zIL2odm0(VpP?T9{a2PvH3MeZ%DG3# zkPFM%LBeOms(C%dtfR)!?!gE_8`L6xz60q3K3Joc7G#k zt{D-x{_ai{DRE&F0$rIS5K&Dr5HQqI@rS`gpzDeyY(Q8d0PUWZ@Y-2G1Q7yNMcm6^ z7cReEHUEn6F$YxYx>R+u%YZEyQbT1*U3eW?=Wz>;ND$zTE~$aPjDttxIt!OqMamQ& z>i!s8+Z;9R*n=G#Mr8EO4gwS_>`a+D#pA=5T{9JP#}_#|kzSq$)%*}Y_Qk6H(}1T* zQ@D6OcJ{*J`5BY*i4`oU>Nhqgz3g?hxWe7*cQxpqXWi@`em7JtomVcIF(5atR52kn z;B}Z=I}3v{s7qJqYuasUuFnqWg|+GGAQYP`74^w$qL4%@QM%R{g|+r#W^}tTyw&kI zVuhJ0zY~6Ew$MBct17<{i;OlnFKlCjq8Cnf z;)Xfj^9Xs%jEksuKkOv6>~P`2MJw$VxHv^ICn(Z{RJ6z|Ug+CMQrz2xjRb`)pt&CC zL)0jp0d><}g{o${jk61!TI1%5NbC&U;{G@y!DAJLp;n}KgmCS7AB_Y(XKWHT>-@Aq z?;u`32g;;Cbh%p!L`V{&g64VeI&+mLuRVK?3OlRfDmnmn|6C;Awb{8rM&j z*!$s<@Lx?tK*NRY%ux$GJrmGq@z(K^n3YO|f%-5vkBf9y@4fq{0T6(C1hxDyCPm=x~1j>tt$9Bzs`1zWm^swU2jx7$>GuQ%D2Qmh*NN#q|j_s3|dMsW= zu{8=WK2tVZSm1)qWhZ>>ii9qnb)ACCCZ%mu9+*h34u#b6)bA}3l!`zQ8#I?DcYKWl zVc#H6c+6N|fR#5p<#*qkE3_`CwNay5!dbJRckcE#9}y&%#4weHtYVf5AYXse^GWP- zKaC=lrw67By*v|Apl4JYK_Uwn1AKAcP4Sahq+iv}&2(3vpJ9+4nei|NDqYVnXbZVH z-$($19A=ITpxQS;c2DK32g23~tOm$+y~1Dp`s+2;Dsjyg6Ze4! zN=4}Mb<>^vI$ja?hOk$f1)krbkcXs)PUD`Mj$Ytn+Ayo1912O8TPikfshw3j3)yn9 zencBxpepAeO`%2s6({o$H#ogx{TgZ`H2*z1j){U~M_2_1gUF0FThOq7`zr#XAGZ8r zsdW~JO)A-iom?9^BNK^AaabPr43Kf6VAp$pg~`jlM6&&Gblj1#-SKy=rAb}ZN3kJ8 zic#14m^BOfxOqUKf~^=wK+M7r*%k++EWWr5!9AsdTTzoY^ehdT;|BCIuEffaQPJyu zTJaC&Ad7r$@HfaA7Y@#LTg-;nCMJ&Rq+Cr^aLl{Gl?Slf1W#et#ML z=Vf#al(g-VZ47@0|LQ00@?A20)Gg9PU-{<$W`FY*`Au7#X-@~~rK{7a(?&Z*IpZmjI3Z6#&hA~rUDB*W>CyVilksq3l^~?2oA=y#jWHO&^qDzfK4J@ zc=dwAzzJ3#I(atAIF9ziXzXEk^xm(1a&fiw;+c)B;<}zZv9Z)f7aLVE>UcenCQ1}Q z9i(Gf3GE#sdmv%&wH$fenxA?8#hQZW6@i2eavoDjb;4X!C%G!_Bu1TC6_#ovy?Y0_ zpUf))~GGr!Cm+zBdZ{=FA zBPw+b9~2A14UuKgJTT;DfJ&O0DA2$aXGyT82CBLr$bbS7D)Dyv!`FSU8({6&%s&v* zADsm>3`1@kzOYr5#$6wdH6DTv(>_JA-xZQeimnkS;xAMRaDGF{l! z*=I4elu!(`*5*_3Si-W7$?`cH@|e>`w9%L~%urSVsk}*&cz}C1xYok}{LLCg`J7@& zEwe!h?eJZ)6!&tf6>1wgfe^LQiNwGY>X&p1378ogtelpqid5|vMP{tz@iRI|@-(2# z9+}M5P)s&OHdFDfidIFb2HN#4U$R0I(wu$)6P79PnP_?(9^oB!$vy< zHvgGQp4F}48L_B8dwz!2xsC15f6CiEpS7Vtyp)$IZxODRXUxV=t#PQH zrxn(4x>Rj(Ey8BmA@NdP3*8oz6?~np0BZiLzD;za68fAcp;FSWaer~O;q;+rWznc# ztN8z7EokntD6yfmt`pYNgP?bn%+W^oiqnxCd*SEC9 zyh<6?BYs7kP+$idQ91iey*SWV9|5yAjK?Yd*~Ml9`c_GxHa}Ch0C5>(E+^bs@!Uqo8j%UN7J8SrDdCU>QQP zs3z(%6xe?xOz_z2*CEKj)v`g+sJKLzR z7M3)XyivN8rLScpJn_5RMTU=v~2}S#8pd!L?T>bRZY|I~y`D z45hsodf`MTZ#Wb;?Z2!wT-nSkE?h)nL-BRv^j-9M{+juTf&=vZ=$f#cpnhSwf4M&@ zyzYUL5~Qa;CO!h~KK(*0jqQ_HM7Btgk;eq1e&9_pMC!N$UZ(S~jZ8033rO(jaW@^& z7j05=!;CWKfb!^0ZXvgm+Z5c%$8&S&+F4hG6+qpxi*wXn8;yy8P8wT{vLwd^x_~b8 zl{Jc6Qaw=cV@Xyg+`Wy#Y(R@t8;xx0_!}W1DrN!jzIZ-}&cgp11zzLb=NPvae3sVB zY2Thf`l2%7o7SDDuL$^HUK`=mq)|*NMUtub?AT)fmK^$`aJ6@#SZ$&LGx7D074RY@pQ>(a^vZU-pYIEA8mq^}`e1f`gUiO{ z>+{R!Si{6+*U2{UJ1jpAO>Z8#&4rURtrn*5D8(G6 zNIez5N4PzBJAFmIMz(}gDrk#)IInualdzleUV6RX$MQ{*v(O3F7Ki(Vy?So$v()jhhXAuaKeJas~tMTG^G7>A- z5!^B4h7`I68rxc+b2$Q!s3R$1c}ggX8vgAXfM_0n6qDBQ(KV(LS? z;D+tYklRX+WjwqTu#Ly0@=&bmxifAj@+?h|GryT3-Bwxgch)6NHY$mF=^LPG@TVK{ zi%>~lq`EyX#XC!~8gwRgA={Yj0Q8-k`8~cGt|558d%Y@|(-+Y!(}k9Bj{NBI!qq_+ zBMp*z)k$Z@?kDz;o%fR4o(x8#Wpi0iIqKgOo2Hws)UBUg`7>GV!Xe^f3wI}zV!)^0 zNX6%fazq*hE|t}dV!Z;(+~8fuMI8ogiXk9!-FJzgRq;IP;XGKFuFMj&&i-|tU7rJ%vE}y3Hy+HSpb<+>=S4^8cM6krgvk{w| z53F!OQF#|He%)*{ruO~hV{*iWZN?1?n{kd}&QjzI6~BC1r@Nls8QbZ#Vj32w48(4J z>r!x9u--pia8ceC2kq9JCVDNwZx83y%xsI>;iFaEr#FyJ_a2aB$1~gF^0^n~rQ88- z4gW*mn{lgya~WOmN1m;5M`W9sb8&-imBPN5`+_Ti*0|gAOj|m=mb+_wv*)(P<%rgL znYQWumiy@x4V>1vnjq6Qwzy&&lEC$|v5ykp8J;d2EUJX~d8lp81Jzt^6c8J4@X`io??7$ku};f=|R6 z#U4`6SMTKJa;oKnG818Y-fZ=9P|)j^KKEJe9l%!tQB#VPCE3;qa5n357fx*2NSB}Y zFB9ZI_4yG_ieQhd2O0JoNxD2W8VI8lSA?c4R56_ym>fRfnJdyk&!SFvE$CKcIdw2( z7x!}-HvHfJ_op9QcPGB6f9Jx1KpQ=XC&2efq;AO1!0K+jyjxz(ZHq(ghm6=RRhHz! zWVfB!Z&d6xQ;s>#_B0jui>(}MwlEtsTS&N8A}^#%MIXq(Zr6nsLig1HncCEoKH!-n zx*l*n0L=WA0Em_~%5+>vK+)(|D~i0vR9;p!gu*;AS=tkf*8&NruOfhRI zvWklDh*%%6Tedy8H>QO16qp*F?8MpnH`(zdZJp}W?X2b#*Y)_7^yQAvSCQs*IkJ~w z%aHmL;lt!1o* z?Q}w8$fS7YvT+1A9P!}oYH@eSjaXjf_vO<_ta;)HTDrE3?04ZP+Ifo^?HI)XN#-Fc zzFDXXG{|;vcZ<9Dd&n}TO@@>Pt=`9Ed7k~zWAAN+uCsG-=}_!a9HN(Hsxr8R^lpfc zE#a(%5?K5juhABV+=fs*%TMt>8aaF~jG{I8!`gb(do9P~;W_|9O}WYE^9JR71ra25Z&GfN2Vj zwM$T#utf?X)^xeLUeY3M@lN4kHC}3T4xPb+6?>@&)mLHhj$H%ol3bA*B`J|2xm~hL z4oocyW7$G={x*q5ac6F} z(%g>#_do&({H4@g^jYCqpz?qM)f~DcC>4q#hTM?ZMO`M?25j&vc{<^tQ21?2I9RqY zVoBvp5DtTp&Cg*Z+6E!F>z@O@X9k?+j^HUd=fbwD&jMTqis_`tHRSz5%W_(^KXy=7 z3>7DBjGlAFBT+Em34l5Mh5>eC%SpDfBYah8=X7g@ULAUUj zBApA7TZETJkuhs1s#|)-FO%8HT_L?T?_DhK&!Jb5Qc)`ue-=UE3tag!8Su(g7J^i8 z4y>TClOacxt5m0i^+V&rA%0m*Cx7QGJqL~~1{&C>5f36Vyh?=H=uZ9>P#9R8uf+SN z$<}Z>6(7O%QD=2D3+NJ<8V1scpSCeITF^+bDEx`FFKwe!=oqaFPnX|~xDtL#tfgBN zt&wZK{>eft)*tQhjQEY5@X^kP?}R&#eMwdi-*r9CcYhw7_Of%N8ekg28o52hNC{@3 zs}~f~m0@+hxzp8spqGrY%`4*6w;4z@PM;(M+X>qdQ_0wHbJSi|u$^-HN0;4Zn&+l( zm;LzXWQz;WO$RLIrgDk_O5GypF`C;6q^BjE+VG4SU7{vB^$p~IyUtHmmWk_pGd$bC zZ{Dx?M6o#%|g^?m=P-&15wDa?WB@nkc4`A`Mi0Mo9KXh*@21a>oqbd2;YAb@Gqf;)Wv6_%-tzcu7q6j5O$s)-NpLVvph0xd)+$ z^CowT;+i;#LGG1wxsKjAy(aR!B#9}eZ$Jsl;G9jbXUyoHp^e@>u|(iR2h4K#EWcR~ z0-Nvb2FztkG<}krZUmZ<^zi%A6=d5qXz~~t39h1;ofI(B#iLGPx*Yw(Qt&sQM(pC; zl@`IuC55QL{WEGp;o_q}g@q;!fEPR@7mBG5%-lb7&~0y6nrgtiho2Li6Wl{g`lvY~ zEM9T^%kCi1u)%{}$Fc&A>-sW!#_LDcaY2`@-Lw(uD^TI0c9n;E>D*iqRy6hS9gjmh zKw%G5?YNA?4`Bz@gN47jW}Q%SSwz^#bm2sXxx(7;8=^G})LERr3LfkU{W^TG?YSU3 zKi!-r%u{oO+Hm2iTm&!k~zhqom?y!FBBw0`&h0Ls=S1#+KKn> zlAZCxcx{OQ?~c42_jnb8D;_JAX#vYahk~{SyNl2y|nelc>*vzcPjDY5~Xx6hi(n)^wI9Mtm^ ztEK0|kum{fOs~lghTMn%ofk63Tjbz>pgm7}#0%))yW75;?Wa9Th}?BoHa_otu*%B} zjUQbhi%9BAmXw60`iLs7JcC@h|;hn~b$<)pGAX-nJ$}5%XN@X!!2vQJL zoU5E=(iYm-3!Ot3dk%=pLPkx~*haxP_OJx>Gi;CJR;*634d~XN=BwwJG2$jt?IVS+ zOzkZ&e2hTP9*U`=$WAIAX}>T+c#`DJ#>jticqw#pLy~x<5_c5QA9ZUHD^?Om0QqjL^}!$}@g8GJ>pz%zW^85%mmO)*r_6 zON??(fxaqDYR1ipZ+`14+2F!%Z>0qiwowdFb!;)Yz1aDq1xa}9f!?_w6=M9&k_zFF z8=~YU7y6`21Nw|RjSXW5RR(#Jy}B-=JT<#Fjd$4XzQH&Tv~5U6n;C!i!}q^s?VGr4 zAjO8$WPolk6m``S{NMEX(wjTDU6ArI)hBd2lhonGabg3F7teOy7Ix6^Ru8pV>w&l| zG;An*-11FR;i6bOKMf0+N+2-^MI$-%9lv~TO;}|}H~sDhv+&R24tWXGHerL3HnPEg z1gWHRhJy`6EanI2{5sFB$_gS=etG-czL!0XoJThBme0uyRX58%iNal51kz4%^SIp<8Fj*_9ozGShFBOpy#y3 z;np+4M(#(xeX_ICUQeei=O+#s=TslCf(%uVCzM+^GruBD?>5?;O=Y+S*`_b3k>VIv zFh|`D5@a1a=)UQ1COhiJ0!G%x!KvT>oBPk(t?S!f5n$L*46j!m@h+q@W~@*)%l7)! z%?Bpcn$Q(WjL@gc2jy2GScsaJ_X8)<+hu_c<4ojKNQ}7nnFB+))~s$}>Zv(tub z48N)>gtX2&?vl6hxJ}DLL^~0H>RDhwcBcZzRHS-&qFS| z0=6+hB3C`+WJGnsV!GJB*R#zpZxYFro;}J}WP+Ts?9W$ftld@{Eaa6jnwjf8vG;Po ztH-?_x^6(Z#}rt=P<9EY2vYC0vnJ77VfUVm7&|9_WM|f0)<^lzM?`Cljh8A2VgtyS z&{H4vPv#-Hj*)Vy)}t&0iYebMmuQXYzIUG;QYvtI+84mc4#>mPf|F5E*ncgv@GiDeOcq5p0ZU4#HWCy2Cl$vO`Ch<7qT?qqccoZ1)-Hjt z26FZxN6KD5gXF-ZHCc=U3(g0^m{=HP$vdUv`l@#sGd?^YKfFXXjaJ0!!i2+nEHG0{ zF_4-TI4H`<{2u+Zs*g4Q!!y#hD(M|4CWVW2ZnGyx1bHHCBb z&g}y)`&3;v9{lZuWx+^>Oz>w*Q34P$yRNeDw zV;m#?{A4G7vTZd+GX(SFuUvfD6TAhH52PEukRs?ONsNxW)%V8C^D}P7<$}(2t2Zva zdgwH6uE!E(q2L(Z9^Sxt2r9qDg2R4oaYsV;E;#OfM%o7CC7(y@zzQqW{g~H1qRsxv zKkh5bCEAzFKWVY>PYzQIFlN+J@k;;~^|B1^hA-%4|C_1O%l5|Pi3;A-%j#uMd1LN0 zQL?1VcZh4lF+OWL7*A`9yTd!`y+gc@GeGA8M_74uB0-v#CegLm^F)aO1D@Nx@VJSh zZX8x<*moKG=L9X3ZDfT}*+$2HTfa3onARj^Lm4S>CChMt@qK8dfD=@Rnu^( z4btBLoktd%`IJYKI_Wl#!k2Cfifjevh)hbvK)+roX9w zpDcT2Ow=|D6O~3WsT4`3;xBRXMD5Hb(e}t?oF$w?(C5=Eye-J!p5$e|Hrm@h(PuyA z`St>4%!@c-=Rdq7FR&(#b=ef54Mtn6mdC0zbz5kQtWZ!Pyb`Wquohn{aanlcg-ltr3Su9 zPP3a*T$ljoV+%}OrI>b#v{CWcelhs1ZJa`1b*>V_GmT>Su!*xoa8Ok$!k^v3zGzf9 zMLPVf;CryDE*0s*vBdHeS=<8Ll`F`fYG+~M32JQrQakIe^q>ktzWK`I?%9fbpf=D( zt8tC>#Ipe;z}I>mfjpE+GY>OYr6Sb5)Bs6(qI9j-;vz^#+*Y=53RFFy%37n??XT|g zP^WO3;VQ*IanKD(dQ{oNn~eqPal+9Go#Xs!i9m~QzyD^!cLipEy`!D}d$P`j%}}WY zWVTRDCOY8pJu}Wo%Xyt%C7h)hX&eAydCDiJww$)+$+1C40}@Ll0kSlWeuJ`Zd6jRsJT;~*?li%zU8+>x z7Es{9&j$We9+Jy9a(kdY;O6|#@zZ7Xwz_WE>Y>E9nwaOCD!1HIWTgwwH-#4S%_fRT zqev=}vT13g6D$B*PzQzj>Qzvjzty*kR-3lop6A>N#rnNuWe}`ScH(zdkeyn)|5rbI z*$b^Z`6vm7T(JgzmrO5j@*N791e^8Z7}^7g7teD37FLj$a(wF`W4(4^lPYxK#j%Z* z&f_m63bp|8fZ8ytAF@3=pgAa+v%%|$xRWl1iY8-;tzKR`D@U|aa7R8#(6R?i7EY@( zd(uZ>^xId5hQ48*OSC>uen6Vo#ZFu}lWee1gxlsI?14@BBGR<@tkIQG~#@Z-4f1W&O&Kh?!Y?+7*1diV#kX+&|621Yxvc51$(^)s<{I!*fddC6ko%&$=yu5x z-htrBH5oWsb`J9-46NlX*qM!0F=tm==WAZ{OkFqrM{MM9mhx8mouBc}O1}$qk?N-B zZ%p6Y;vUVoC0_k{uDD;E6F3lC4}1n~ajWGY%XiBf{gOEi{)29KcptYW%BhI56FkrD z!-!9o!|m8_w!w4AJLNrVt-n_!c49*r(A0~H8ZU>yCspT0Z;Wv&CTxwQJy@}It~0kc zEdNVMn0W>B$M^I#WIMZXi3>;5j#{jD_EQXW5mr(0J4i9edmVV~5z%w%RZqE_<|M)@ zM;G2rSA$Dy>|)0v50qqTV-^p@ra@03lBHBcK8(H}RU%en^+{PsMWpd#0ew-pi=$Cp zdUJ6xjV?+aw_94}ix+j2+=@NW8DUxM_~;lwCH_)&_eW;vEZ>?qn_OcD9T$$2Jhgz% zClmu^OSh?b1C1ak6*L5X?3t}-dL82>MNkZM=1Yyd2LB`6EgvXgi%Z`-%|vbdcgOkl>IGz#gpp*99cS zlnJ&mNY(wyagAh8Bh?l| zRYoz}DN;zq;|kQs>-!5uOvXb`yQfqdp@CY_#n^j1tUn-M4s2O<$LFyq1l#8w5KL;e3lf%D;{sBER6PQF;rF-fK`^HQUqs+XfAvyu!T0G}wiaAA* z6IA@FZ(IZV#9SaAx-V(wAN526#dBXwdb{{5Mcmt9m#ci?sm2b-*G$i$m$>(eH%`Zz zp6q~L`keF=I&Hx^pu;j9-%H~aG>Y{B%Q%%GyJSstD({Tyob-Y$E#N5F`<>hmgj?rc z4ax$Vk>RVFj@|`aAh}ISM_oX`#T(ppA8fH!1~$Oar6uiKEHrxU@+}pd3oN8dMMZQ9L{|R_PO*I$8sA(u5nTPU_A9N{6|Zbi;leJijrv%W zc&iLcQYJ#b1T2J6(E8xQ4@D^MfbBc2lhG7BV#MqP2KYxK&T{_8A8(AhGZE5i7d99F zpS^d1iz21eF0d{KtjU69=S>;7q#?F<#p|r`j`LpC>u;7!NFL*cGZH;XOe)0qbW} z1?^K9^K;Km@?s9=p zrewR#S|*cXAhnwY8U)dxcMmn2cNEK|S$?>xsixaB1yEA)Ik`1sz-=g6+YejqnwzR) z?&AY0Zg?7mmHp7WWM6;tgPtJk`n9xk*CJBzm0>U*wL!>UiUErKE-Gfo`9S5G4tkSU ze^9FjXr4*S)V)f~1%h(0{;}e=*D7EWqH=x(2%Kfo`Ef`Rr~e>n@<{eD=zVs70YCXhlqKb&PbizJ=%v&dhB_rIZO%g2@C95*MyEV{oBh(TB z8SMI%J|!Q767hsYrhyo{K(ZpikYG4XHu{>vlE{D?9$ro%F?sW}C8~I#tLQ0XM~h2l zp)Y9;H#9wc`n_}{%fLq4>=)insWhY%jryoSA|ESd^>3;O>@pdD!5OW z!K7;Xr*90$?@+lhbDV@d9Y@I1S%hs+oqC)TLZ}g$tkpEFniVh(>bog940+IwBJ0nI6jIZlzI(8GdRIOXKLtBU`+ zu1ODEMDEiS@*H8kI8gvv_8{9>Be)%QQjZd@ag&OLU39&;R&+e7Pm|8X9*{)nlLbuJ%x1)wHdid*sW# z3cb!s3p{FIhIZ>$bIC8elY;FJGU?2aLGN{N z1KDWV^kdO&2(%?f76xFED5U%bdy*pMetcB#F&lFcky7ClrN!YJWjzUY-Y{37o7haSto74b|D$93IL}`Kt zepwI$j*~WoUJEyoo2rGPi?U`zpRkVZp`VC44A&Gn;rKl^ggPFT4Brl2+WVV3XN%lT4ly0azTLNz~elX3=P|h$A0L{@%>~U`w@KFi4*gV zUs7AAl+0KE-|tB(H=mZr#-q$;0?MNp2>54GF^j01{38!kk(`GI;EI9I_et5{lVIk*!orSJ0`s_-~_OyH^&yG5j~1e^~X_Z=n82a$G&Q zeQ6%-9aY|yEy3?9-d*+9Za5bI7d;)9eQ*)aQNx|lbF1jSRhRF*qj&!>D z85;zi!Nb{Rhm&EU-ubToPwW>lTv*jSW;n4dd9Kk*C5fO{VLYO^DM*Yi@hS{FdF#|t~r{azWz`CR!dSm@UwD~#bZlSX=6!tQVe7Vi>R2Ypbdj! z^dD=}r(6zN|5}4;xrZ@(nn~QC!p`Mb1Zq~M_*?kU4XSO5gSuKBUUZ6J+i@(v7;wW* z>5sj(M_A7K-*pVlxDkpP!AY6W`);(-o(so9BD?KX1oC#4T)nX)4Atg3=~fU>tPx!h zUH7)g#-kSNL0z1$$+JN?;C7DA65)A?@{!KT(BkzsgiSyVxe~BZ>*R6b@|Hd`?IUkF z^L^!H?D=b}|7pYi-`X1+F0&AiL$)lY2}8CGDwJ5;rZ`Ju$#jFDi9ReeD;G&`%?%M8Tavkcy00_Ue46J2{T7 zokvFcv!l+$yyag0?$ICH*BkRT`pQxW{RK2B#2&ODyXn6lQi5@lmQFgYZ2&3iMnSqD zZc_K$vpy@L(tmivuuz&SIY)05ES-cD%4P$ee>XD4|4v{7oOjVX)2~T&L$Y*IhBlYJ zphGXl=}>T%33N6TIGI4I=iGs2`w}4D!iJ^Xr+m&C1=M+9OC;1K9b5;eYBmTfoXnhd zV5A%cg#)*8>M5K+G4cJ}yYpXmTVfJ7`!vnj5O@=5ZmMMAZcs*99c4LE4bw@zqy_>g z%O=fVKB>{rWM~)ES#~nCi+!8sfSL-Vq-6U=l6wucKx}9A} z<|u5QO(1MrPDKqj8ZC8Y_1oLWB7DwYe~hpQpT~xz3y^`2H6R+}CTDmXO*DQy340i3X|#>K z7o0eCIS*(EK<9z2zDD{j`>+d3ykeiESzS+Wr8g;edKg!#+Xa||L)$;#h6$Dw=g;(C z`k%?+iGJyQoc$#?i_cs6%{9^wHwvv`7q>vSMI`Pk6X?&e3DTre%o>WUreZ2XQ#IAV zj>~{jpCV}|jXL*_=w3;^_@bu7`FcJtz@5YJe!+d5{dMxG4+dVb;v(R;74yjk9*5(~ zZAN-4#X!UBEmRDu{9*lQhIS8G8=B#t=7Yjws7sKcFOuF6R*F06WYP&!Lz}D%aym;? zy%Ojc3cTB7VTOOY_(=GXa3{_U&kvEqgFk{d^8DSLcZ8E|o3QrxlfF9^g>oJTQ&<$r z&q)pi=gusiIrQ0RsFZz9Zh)IxM`uniGnB}R0`U7fs1M4WdTS=O+Hfh5hK7DLfEv2i zm7jg0_dJQ!nv}X_ohHk^GP4Yr8qvT0?%$06@r!?*|3}4oibkR?oIDEFfKlJL7^0%eZnwhAXjPdvE`P{ddOF8lA^$8kW}RCNWmB zWA3LBdfgbY+gH7NrY^zLi1p~uk|{%v`oq0sIl+#$dw=s@(JS`F zJ}(tjjimsnEb>m+xjE>B7E|bqw#rPNRUv0lP!;cLfnye_b(=W4c5Lbzt*zpd`i4Lh+Qc2yC&{tHgqJ%yKp}Kw2t`Zr2|5 zJ?e{1OCLuY59+r^p!~%#(yumRsXNrGLRYwjQZtDEA=e37Ei|a&qp>MhW#}qT=Or{p z0jd%F-C>8EWNV7P7Wl6H(wAk;XpxHVj!c2-sx$5#>VC}?>0?E^rcaV3XX{Ajq`z{6 z+c^9{N4VCNhuGiJ5wiFEow$Ej^1*NHqw_D-oWqj+OAR-JZ0yPzCbDl@i7cBwNU5Z1>S$~ycSQ*j$L=>3Ssnv!eTszeLVaBz)Ga@v8$v7mN&`QJ6Vrm* zLT>_9&bTQ&S)L_G6gbaXSOMxV#)sL3kvE--cy^5M2)Y_!pXuYmDc~_>mL)Hk72YaK z2yYkELWHHy^ICX?2;3NP>ZoLK#yCoj!DWm;GK$O7SG+UHKJL!j^e;<701EFw&J)ux z7&PnF#6usI0k_*;sArJjY~s_4pRr>@jke&$_j2YL?3Q7q*)Qh!(O=s~M4q2ycpMsI ziF=sD%Y1fumkgGHUjA+?d^Q=ng{%FW4V|)T`o`?c|4knaf{q<6L6wbt->}BY=Is}E zkeW$Qb2~i9)Ic$3C~}I5dGtn0=vi?~D`#wzt~5M) zqr#))wbLP{koB)O8g7e@dEwS4Ax(y6D5N!sv7;Cnp_7$Qmg_vfGxD_fZqOPKeM)nUb2djd5q}C*QKFCdU&}KSU zpTOJ{K#pBD;D(756MalnA?lO#N%H7NFk6^irYG+Nyo>@3FTR8mUiiBN5`NIE&{%EI z`sARCWF3#Qe;^Y&ECQcTF|dfwqGGbi8WFJ8PCuj+Fjly5k@Q^EQ{!HVcc!32!1gZM|TD{1grn_lXvXv1YasQ zhNT)XHT?Fhoq`@givNx&I|b`sJFbKI(z}u_)9r+QfP=|n8oxmSo%Fz5x=^m^A8%Z=51>46OZYo!!lW~NB3eT+t0@vs#XyMN&(d8Kiz`|*sUe5J ze|CDr%#pEh3>1#s$+_pSgCgPG@0=J*euu|VH5Sh6Y3)(LK4R%Em!l{OD4=i6G;Z|I z)1%I8k-Uu*LMNVXD9JG%hSrD`9#~)a#OKz`3@rv{s?|%F)7nBsf`5VDh*zx7i<)JV(&k^9 z_ot%2i+gAJq@zK5XPA}gfd}VRiXW=GHNBDXlgp;YN>4|$Y3>Wm>U@29NTMKJ>LPrM z2^xbIt`Jp1Gu~6ATbQn_5uDVgP2Z$j zPBP{FYV4GHgy1)*<%i<<_^tDyGj`Ot20GkO88)1w$CVSE*zv&8S`!swdokKCOwBC%-7W2$f-yYv5(BplF5LZGyf?nm$;Jw1)$VMpM zD2_ZuR{L+5a|9-=qd`jqtNroYCz_@?E(OTofQ%9)LxbtEPtO@^vR-!m>W{8#_G>Q6 z3ItWtEUMzdudq3@Op_9-d?YDK7C?)m|T5gW~zyKJC`-I&viL*2aR z^|xAW2de|TMk2f9Rg zgLlL)auf&Qj<-dtJi689s1wtB!)~cF1L1V=Au{Q!P&VF0pM}K^TzU>pUJcFNxDWPd zMsCoFBBrZX?Irs?u3=zbpIe>Zp$8=a^l@z<&*>l9$Pjs|z3n?sBhp1Y9`? z?h&qZ?V+pR3q4`q&FA@m9FHB?OdG41OfiWRSw+RP$~rxbm~QTO!))`=@`$@t8uR0K zY~;uMGn{Y0Zq|Mz`PG8i*2yPR^G`)&4>u>4$7uP)1}#S^26)+3RE!zcEU{2`Uf~A~ zH!}2ymchd`{$U|>!a$G zH$09iuZ06klRhHaN_VSw>RLlL>0CPtIW+de-|uJJE(HSngfmh=KRY!~YF~uRTcog* z6@!$Cs!wCQsLb$3IV9+cafHgZS_Q$87bTsgADK4xx=LC<5 zS(A5ozqkmQ$DRmF5%Q3uYD}K8dn&?Wz^zPn8Q2Bphy=#Mm>h8HfJVsMr=59yOkM-0 zF&zbHINg7=C$ocQ@}jDryzDd0+R!dZyicF>bLC@obYvZba_f zc6A%@X|YJ%rSKU(lw6@S{90FTzdO0*(R3>~dj3JRlx!J^B+KJ@rrHJqr4$3K*9vBa0&KU^=+o$fMJ5fZWnPH_ z{I*YWI;6sHj5KI6{r-t*}GeqJuOd< zO*RsuHbxtJmEAx>05z4V#=XiTK*miBt(m(*wP1mZykIAw$oUY0JAUJq*MSHtR%Gt; za>!0@SmANT7z!eX1xczYrh*~|sF-Rob}H@AFP(Kkx5*2eA(utgc(rS)qgKwT)SVH= zM=$Xm2&nM8sY+++K^Hwmn6HluIttYug@V|~4Z>vKiio}1ez!JK5L_eJJ^h|IYuYhM zzgwZ;7{N5xLLfweoR@oswt)UZSs8kI@~1vdF-uN>kc~;g9cN(bx7zBMXe)#Q7O&by z_Hly{kIl__8weez80aNZOT{epFOebp1p86~jepRPRLvfYzn1&pL2T%9&HtKzZZMWg z9tkjR@XhgBP8LcbmP@Bi?|}IXc`B(MJJg-@`|rPhG^iL>;}>O>p((-?|D99&gW`P7 z(b!X=Qn%fn;8 z`K*oQsi7EX;&PCR+2p?P)pGR~rgnO(Y%oPMsBsSVCXa5{RQRQeQbFwtQ$ohuK%hhR zMFD+8ZPuE!pK6UQ$`yJOM*E&wAY_C?Cn_!4lSR^21M=)`Ll5_l7$8cD1E>+>6RABm) z^Q0pJSj!py%U-P(rA&iP=9|P#b28;%lCk&1I6h8p3`NHtg>%m0^ltKZ=B_K->1MB! z!G(3e9kfeu%zcorpDw7B^+3U4 zBDlE7@az+2FcluHq4y(Q%2D8EK%ROPmJR1Xd&Il^eqzZkd+Hl+Gt4YBH>@``(L)h8 zGi}U@TpKlJS8>G0^@4+rjq(d_;)1U^S+xnJn}5{xvaf$Xl{C#M_c=DT!VlR8DgGb5 z|L0F79r}2WLhp7>=2ZN$U9*QIF!=RBP|q!rq7d;BlCD}N-4+3rJV9jztZNr@_^avUAdlBB7j4!qrzqw)MUDbzO_RW+ zYOo6_(h7E{kyc==)NNsqH`}5tR1DH+OG8d7Pmz@d;~w3B8(%(9ql z0cW)`6ieAb7Cl$ru0gxB{|uTa4`1>giU*zoc#=7ZK<0K>+;a!T z0DZBLipdgOVir#6h)5J*k$8p2{+S17E_PoSo)w+}e&vAoYR!^>gz)+ELG%_{S1$=z zHtCV1B4|0ZsQ#SXB=znVaeM4*-V3ks?U092u8EU!delyu`yCl2_7`-dYzuEo3;*rs zKhBX^5wl|BwSOTAlgKt3#B8RRbc(E}Vy?fPL^c0^o8SE6yJ)tNSJbA#2;4#4E>O-| z<$p?WN^sA&#wR(ld+aF>M}}bV9M1S~_|9WHkrO`nJH@{;?_h_$iXd+zEi8(G$W=mZ z{Uwl9HZ}{I=AcW0iYuHKD`SHMTwy6aPi^#_RUlJDe8Zyfue}MXBes z_PY0jS!P%@Ll@N92~1WvF0oTJr?u6}4DB-MZQ!I}7gUoN&q0TFw>vt$SOnm7pVfYB z9R-#B>}Y=Ej10j9dhp z(Sqd}c$^Ir%l)!62~DqA!Qv*@?UQ|g`PodImPD^qWK zVIz2e%g5UEDX5YSL|eo4u->r5Ac(JJhM2@DZ{D4kt1y!yc|q`Iuf{Or-hd+zp);Np zUz8z>t{`}cZ@18h@naK>rV7%e&TTF~cHBHe5NtUOL|^keT3j zGpt*fBP^0bLF+h)WOD{I8V1*3*ygV6fL;*(gO0Hjeef6kgr(>sEqIaV3e9dsqTrtQ zS!sbsjiAx6Td`Z+3;m@qwqHpfn)f*b*9tt!bV%d}ktubbrdZf)$d4$DI22elZJlJd z!j&d+R45vnVr<6UxjX+aKkl=4t$xZ|hr9otoSZ~%+PtEdDCPo18mXALDcSP8@cKD_ zSSQ@>)vnG9UlIL8^zTDQQ#GH^>x6|6I*DU$%xE@T5OmS0;q~G&x;QeOtnzPHuM=j0 zJl975;g?$;|3ZDU;e>R9vWqVI4oY~Tbl9KYKR54+qC=f8?2c?Utn=ODh3T30?{q@_ zX)6ePB?``y>u)t1@UutuSaHpJ=%g3Zh}-%W=Lj; z>($g9j=+orQb&+D+H*J|c)~AV`Oa#4MYxwr`wIk+hj}$?DJF>`2~ZtmNEL1f#Y7>< z>qEt1wK9+Hi|&Orj;o3b*~iOaTgQG2mwdyHl-&Kd{?49{&m}I$V?Tz4y}#P0!lPEv z5mrr~Q9KqDD^pZk={&iau87(edQJ9O*hQvQ=1MQ+xdC<C-O)f4RcaF5UiGTmg-G7*9b#Ip}3pq*#+}Q2S2)-4MW44=Z zrkvFj6Hk#9R7{H!eGjPY+9vJuya%Pn(B;szNEvy!9m3DZm%8+*&MTX;CrYyL5wTh- zOVr?%1fj|IOCnZ4=o=oUMN|HL?4)C&yT~CuJnKbL2vC-(EBv;L+M>6M>SkEVfp>Wq zN#Dn^y!>gM(=xP}W5&|Fc;SHvV{ z=NCRlL>TSJnruuECx|(Zq(j4C_x_G%m+P2udJFj*&U$v1ZpeL4BP?_VKnxP|$7 z?4m#?m|-sM7K#Dl|0W8EHXpxrRnhLfi|(S+1Fw6>i`v!o;wJ&fgljfn|DZDHcUd^S zB_M^|kW_e7PQM|klwELaJ$dlB#h%}9ee>Ev5Q_PVsZbz@nSX0gRL&RaQ( zc0sv1Wm@Xw>iskumI_$_GbGo6EPUbY#O_=gKuOI1$W|WwZ){;^l7jo3w%n`K| zQ$>+ORLpU0mS`8`{`13)=fH!4atm0r0GUAFNk1#cz?yWZya0(}{W-dqK>aJKpcE+c z^o0S3h3$eJ(JjnYs1WR;(}L>752n|EG&M4*fnl)OJD*M>#(1wBA(nP@ATPUvP9|}$ z<%f@vpyCLKM&lphHDkdOClF0sJUM>N%f3AB0dZrHFjiLU(*w~HuaM^m>&3g(s|aRz znhgblo|*OP`=auY9!);fY4i!Rh1h#!5y^%|+^)7jqsGZtK;Xc&oNz+*Mji68Z~w!^ zGvTrOy4}V!Wm3#0ilk977wBzK7nxoq<^tOx19(!jp6Pdc^u{{h;%WVE-C>=c$EH&emD*bJoBAnAHVjB)rSrEZN+@Dfm>9B z$IvLZnO3$^3|NsZRE$Lx9hbevSSbC&#;1e8z~CQdgfp~vW!TB(u_WgQmr1)OmU@5L zH$Vyl+!{Z$EErjIsfIc?s66lCaszi=T)-6Sr&d>Sn`IQ9iv(eP~X)vP@klNMOMI zg!Ays#OP#@D!DbIU9iq~udr8w=Vxf!G?(UfhSiKh+I+0T@%#Xdc*OB>u%5}yedKN3 zLvz^|-=1$p)x4_caMCu34A`LRKE>Rj$ZcRXi5j&@5MsB`+6UasbZm5nHY2(;Bvpfq zjuz!cAR?xUOyCF%GCZ(?7hl_N*HK?#O5jpikmjR1JQ2>i#P%!B8OcEuMn-MhvL&m`JqgbS7rp-`TjFJbK94L}j2X(xT<;f`7caFCFx7bJ6#tCyH0@3!MEi5@< zj=%n7r{MD`?^`i<=_hfs$O9e+Zq`g39^79tky$vA^dpDAU%gzgPGa`Vo>r*On)XHS#|ukjoBb<9w`MjPvX}>QJU5?Y3p-@%l|{O{bhF_m zyex}JhkBE?Eo@IztruRAuebbGv!PL7xjL7?l&Wmjp4O&H%HcPzDT?WZ(vuOJlu016 zS}3Tb69wn!%2`Le(v=S)A4Klaubs15e@4H1I{sHR3QS_$!thfrd5=HuhwaHVM{ z-ckJL%dQ(#_GZfux6+NjjDPb_Tc}gaC1HifLv^<%S)L8_(w;YSz72;P4Y+apl`>rq zsiU_mDnjdi`N_9$&a+;Tq0N>YnBD}4*zF?5j1%BuYX_aQ#|7ZgTZfq$~%g~-8+as`9V!zuBO|58+=VJw|=m({`UtOXa zaBCFc$(Bxt8QRZuW*WPa7}2&IQCE0g&<(ityl(bO)m#U6^1rS_?g}S-V;l6HH`CzU zM;!qr?q==A=xj+hXxgP}Fmj1gJ)DF~#)_!V)dOxw>iJAp8`L0ZqOoOtyjM0cgG^Dq z>XV3-P|cg8*eNtV3Ah_k;dc!h+k>Ti^=?FUPztFH0^xb+8#_{9+X)!7wKb#Ldk(qm z7(5X>AuZgBH}wa`&146UXGv&lJuC%YMKOmcvY(1U>IZT;urLi*dU?{-{znvT(J1vE zH|XGDwMx0q`B1ZRoo6#+>;(cS2%bSbWB~{X^(q%bN6s!)y7Hj51Of_W>U{74?}=xCSlH4L;`s>}VN*ku|X{^v+fLc6?l7 z);ta_ur%+>1TNPD^?9hwE>WNJ$_vky^kKfMR@W)(j;snw9IUW}{27vwMzRu|2gH#J zxP!@&yN>;whUpXj$-YXPw<&d&y6f_&+_`DKmj7=wUwT z;b#LZNwKzRx1w=&v!MQoZ!@*9+;!}SIoT2_Gid(L z?a6>#YzdG3A{I_#8v)NG9F-4I7ipLG?y2a6RMW;>eY3XQr)heU2qW87vp0oXnkHdX zvR%`!28SfDor#@wC#*Fz5rW`(^i3esTE4+!7#1m#*Q<*o6XnoU38fb?_*peO5xU)D zPG!KYQdk>=$IOEb3J#3ABzcV3vjO6WJEMf*#viNysI~ewhEu;dM6xGAOYh+>#4d^{ zp+JxbSh_vvZq}eEBRrh|88eJ?2+z0mSdb=bRN(7fAf7}&(vV47bp!(B8P%3 zGKBl(Wh9{52hu ze(}+6KkIa(Pgqe!HuHEXeb8ne+)gpD)+vAve9CO$I(Z*JHACB>encufj%!abmi>~5 zURd%hmVB%oh{yzfc^6&EWYV?MFUpR&_p6%?n2N6yo+7LLF}QQaJ<|okycP&L1mqDt z3F|=@A(<19Da`{(in|prbziIgh%6Z?)W>5=*A^QLrBF;VMG~o)bdn31;9E1>XJKZ2 z4S7VmL`mdez;UwEfg~&fiu(iFq6d4E z;7X?(QVKggI`j}f`5#bFbQCUQ;tr1jS`*hC`fqy~DlVYlu@zyFgH6#_tId#jOwkt# zt|<2@4$UePV7RSPw%Y#$X^p-U-S2jS>=ibtvh_zqt_!T#4wl3C7@B=-FL&7<^}Dpo z_C|xZQ8pGckRsndr(Lr`vr#ihuhje+h z$n43a8@0xC0ahf$Mx(r8kvvJXi1YwY2ldEIV&w2(R}qYOLG9bevW0=zYotrOeJXN` zFdsO`62#WiwYvRupU^2Qb4LL*WF{Q@96QRm0qEM=@0Y)B1(vExzvhb67&{9;FkOsvMf~ElwosAaU!Vt)XrJUb#8j84mO{|3U4Thm z4A@xIE>TmcKD0H|h#AKIpjclB_=WGD+O4?^8J;fE6IkNEIcm(MPR~Zy`RI9e@3Cxg zL(c*A+B^1h(n}>G#4=9}+KXI$xlg0^9Hfi)fpr-NVW|;EBf!kw}W6w^feN_?a$wN{4z2WDw98v-)f1TSQxUfT|W$uP6FT2q&(>sHKM1;aJW^spltrzT|Qu~mC;qcG#vKm~C#&|_N z*pV=j(RlC!!5_bGol~8aW@4MxKc56<`N`&^lQ7jT*2YynlL_ z2JDdP>*0ALslIM|_x+C?RDA^lz!`~8C$dp)3>lpld``X8yw+aco40Wl7AfyRQtX^X zfo8vU&0k`p%Ocw~yQlYg8JnS(MB4NYAWmZ3H6dz{C6cPyENFMf9jB~=zx?rzcGSQ8 zu4Dg}oBg@|?@6yjTkTK4;#J$oK5q7hze#Qa=4Ke-_Bh2HrARFm18i*ws$e7r=|g9Q z?b>!tdf+v>Qd|+0DlB>36j3Uu*BAJ0@Ok3Xjlv&{`G^oNjNW;Y@w1e>;T#heT(4vGs zpx&dKDn|=E4#s9wC@k%7dC#6zz*}>}LU+K<&@h(m8nb$d%6QD(tUTzx{%iKdZ=-rV z+%9r(kh$H|31@S%E))MdC*p>^6f0WlIAxw5Zu|RZ#%z&|zmB>n_EOfUpFB_l% z8pLt4q{pUSfdYd$s$!~*6a-@+KS`7vc}$Y;KmN0kg9~47 zL+y;Sy+17Vn{Hjy^!$TrDcQo~IDNIvGNqJaV0I~{Vp3y_-EljPna z8%hf2(H*p9{enl0g@RsknN)^izPMcjm7+y*OY#`c8*n?Q-yUJf1s)^)P(A8wq>SEb zxOT!}w;Cfv;wHwfJT}X|(G_ofRhHIOr-gg`@L#O+YBrR~E(?x_gZzP_jJ~PJQuHd; zkafP*BoTxb?s=P`_t6T?XPVvWBR+|OnyLM6`T84@ZsDO>&c?^^S9(lX8Ga)tZ*#&5 zWxOu$wV!oZ%=MsvK3Hk+3Cq=67-Np-F2Qvu-Muu|m5+JEr{n0eAMs)r9^~}w`8#1} z{c^@{?TbaZxFS3b<+GHQ;$k?D-l&c9xvDv+gPBL)H|tZsW&;wO@MxZVllLi-rav0g z_-chHjfs__$yq4H@4is3msbjxitq{-z6p+E-fkwc?z#^2!G+)W^<=Aya`V@%e<3Tl z#a(&4@_}~G!{V-K6a%y9S}LX+lerOX_Wz?gvK)}c-U zsVWzO!EQJ?V#Drg*X=IbscW-mu6;!e2@B8dkfK0$NpL|l^!cn#-K;K;%9cD9W0`s$ zeL!6s;qf$7XJc9RQ%pHU%BUEVxKooHRU+RZxGO6Z6bmZ? zHo0%|>UXP&N*48o6_Lz9RC?U4-Wpf~h*;uvRFoT#DA*&|1=s+29Ejy^_Rk4V3)(*g z{9=&rsub?mub)-nQ5|# zTK?!hmemfAO$EzpC(AEYxH8J5?RUE@yDS66wpsTj&>&B4w1~LY2&&+a5p~-z$BzsC z)gF5_jjHeGYv+K z5LZFNdf(KlMe4B#vGEuoEJAG9yvWiH5(WN>NiQdZ<^(N>>1PYDJJHDc%Pty@ zQNY7Re~ImFa}V87+Y?Dx_!kv zwmC&y`MV%YZN2gX`;^UBWVynU;Av1j2JZcS5dQ>GG!#TDlNIXg0*%-+1QZ>CKVSkG zcq1|mi{uu5>>;g1VOE1V!XoJbNCslfm#cgb2O)3ccG&E==mm)Q?@9kW$GVP@JUpIC z_I_p7F-m^z9%LXjGfG7*?$dlc#Vy#Yz0+abrMX3MWAOe9@-0;2GpAzPRt`Q6CV z1SP;8(JQ=qG@YtXyc-RPp_{_51MehN(?+t~@rv88bc-=C(keqeV!U!2y&@`Ic?y~r z9sBnEd6;V46SdR3d=7q>;a?h@4FW3VUtjn2v0-e?NV0*I&CZxGwq#buX8RxsZxh8V zVGm@sT8spCk-n))_N@G$Adol~_88M>jq+knGvvgd{gaZ`SdGj2Jr=Q8BBXj+tbq(pval7@*$@*J75gITJ$IA%P|)V6?zv}uP#t4zlg5jOz;ebY zCKpFJ$>rIWB3uL)ZA?hsZFb~$<&l@D&k3DW1;Y>4+dPPc?^((Mb{PvP~uXYBWE zH0)8Vd81vupSmK&KJwTGX8V-G!g|nIsf03|kNu88`+rUX~QmqTF^>n=Y@7ZvV0#mkT$FaRElSwY2E8dtO^YyCje3`3YH|LYbEf6AQP0{+hG-x=5={p(U>%MyH{?$nsX z>wJ$3iY1WB@k*P1W=5r~iD@(({(AQhm%W+#e;&QlWH>{r)p5b&GY5^rO&N|w7j9qK zEb@QNYBAj8`n@EN+bqOij4!jc#dx;*>)+}Jq^ zisvqL@d5Gs*X9IT9gw=W*6b&_Jg(rXwQ(tSQ_L=kl;B!wv-i5-cv!?f3R|IhpsW|) zjaUVX+-m(%FKoF|3p$R?h8`%%%F~dK+PGR^R2a!v#!g|xTwO!Yz4MoZR;2TdeCy@%9S9Ax(K#;qT zifPnl=ridyJyN)k=#>?op)HF{1^Uh-I#$}DubO7uquUpPD?nUP;#x6Je~-qDy;-?; z&Ou$pT*%nx!O{VCb>~B4eN!j9ViX*I)L&xOSs&Qwzu@-1622x^e_AqUw%Cf8H!rL| zOjb@JTWt`NPBH5#l0wBCrz=C(Ng8L@(dz=z$OTchq&}$MEq88Sc)NNXbfm5bZC4+g zI<6EYZr~XGU~|2a6FB&r?nwIVn`iA)!h8iXVM+3~FiCS#HE~c-l}nedq*TDc>tuDCiOF3+Wd22zn*;u2#0;hT!2Qjw>d)ozD%y zAIyCFAMJfsF761AV`eNK?&*+~vrz}fh;6G+k?d(mea)jQgg9B|!>18j4(1YUK)6C< zTqG@2oCHz5<*x5L$iX=o4ky^=h7itY|7z|pqwP6I&&wh47z!*rq#?y4;}*XI>c@&j z(o1v4IV@%!3Wsb7=V2TB1?;vY>9wlB0IT=$`yEGClN=teb!%+gqB4pB)36;o_Vufe zi26VV3svW@d++u(ZcyG1>ytEUtAi^1DrGf-y{g0dMEOlYjR0jCv2krycs|$y<9fkD zY1OocVR$9#!JUobDNkdK~BJ_xvh$7HU8C> zyR>Vr1#}AwBKn{kb2YtV&Q%5eh84$71Rx78xgNHlkP`r@A12p647LLB&mXoQB3pS3 zz+*OMYahkzp-3qeQ#196Pm6Mur!kLSO7iI>V$6q)K1~U9R5_@J)=ofUYZRnv@Cf$c zE}$`W+{_qH>JuPT@TsI}PI6==UF)~e!?=5T)0|@JYUuSTwek!tCJ);+kLV(O-`AWt z5jq<>!_3c^HaS6O!Vly8>+H)to_B)yyCftmMIC>>t~ogGAvB?V9Np==TivTj1SH&$ z?4tAi+Xdaij<7QCW`jxmB;c&i9{Cku@E@C+D5??M4r?}~hTjfrXA(n~dDOeEs+3F2 zZX^>jCWw6L#q9X`-iK0_LyIR3xM9w2hv!Dm45+Swe!-!GtC)_kn*wZxkpY##&UHDS z{kM<#1%7rbr*5$iz(~#qZzFlkyq}-3kK}zt*5E7w$&QH4x>~;sZN;>w^`_;@98crX zpkg{s6emJt7s>C5JO8%}RPFqw&j~}aSn_&V8NEw>Euxn+8_q>NQ09_&FUzqEZCltv z(Mf%VzDJhjp5>7?{jARg@E-RoZfoMbaJxX61YY7P^3bIc6Js(YPBk62H{)bT_&dcX zZ@cj;`(Pe#qrNQRIehgq>67Y>QfPM3t?7pH4b%rc5C|oK8sIR&9Ef7oC{c|WE$nNY zPsizFrLLO%o*`wF=zVqr7j1Du@5Fb+$4|c4kL7VNkHz}IdK}oY*?tSkeG^a_kvarD6;@EZ6BRGw?@B>&Ig7J9Gko#9}1caXp=ZkA1_4sxrzui zO8g9SYFz^$cKo~-^Nn3EyFtM;IjHc_8-$&*d?7Gc6omqmJHdkER>f{u_+WFv9$9I4 zycZ;s^yigXqMf=b;(B{B%6PwPKrsBS?0^`-uck6;-|?MlUFP(JNu&;miTIbFv&EzOYX;i#$t2{j&z# z@I-7FYwT8^^(hjbegmq5+BKzuD*{{>xnP{e{-WYLJN$_{XPeWw@YjML{LR6Ef3X7Q z*t>W9Njr~&aEolBzz-s2kvd*Tdjz4C^~(>KdG z@58WpY>b;jZxgroA>><UJY)J9%0tY8mxKr1nECa0;lV^i4AI__p`gD+o9;%Z+M~xjP z{~h+XW5-TfoZkcck^tVWqF6}&*qWtX(<1=e(4|NXN5{8nc82!8hpT*B`^Woi-%$WO zyJCE{*a0wtlQYo}`(5>47i~9I+9^9nw+oD&G!&8PFF>V0fdE3!((#i?cm@PVLF59b z>)9VM!q(=Lf>2_0WlG($PLpMmMnDgb=af8~Ib}V?q);RolHB4_$ogdKYkZE7#Lztf zb@Z}HRo(+M9$7vK1=3wTZ`eL;hsWO)Zny29Vzb9?M85y+9PgKX7FjYiLytm{>B{BO zB6*P%9a;Qme8+z9fjb5Oo3R;}9d-bWUw?)}#$fDBUS^jSRTTKR7s>sTfP6eD}P;zXE;!KROL+OVT|L~7vQM=}G zkt&PYHHsdfr(s;>c~sx7sqjk&Q7@wy6bi)7>H*%HNsM|bsA-K9 zrUEPuV4%(kLyk)g#uWMfIT_lF=+Y1qeAghrD=but^>37gIMu!k8;ilsu>i&XUe4U# zOaJRZd*77{Q1}~Q1YXVXNXb*L)>8J4)vIw3g1(`t6+zs2C>(}8po+#Ji#TiuxgcZM zWiH?R=gaf^?OBp6WW*Nb-d9(`vhZ_KCfniHF2Lk%C0+GOf+x@o>jKLnP2v>jwN)CJ z1hwQF!`njdgq1<{VkM2uxGYBpD;6{KU39GPMvW0Q39c#%4SVIAz46^`hMSK@4vp<0 zae3wr_yBA_!}&P->_37d$=_FRQen!0prHkE=dMMhfXA`@qc&z@FU7zFvkTQlyCYH3 zy29h6`mzA?$SumAz=YSZ!#Yy?Q#H#xFnW(6wk%OwXd?(EbcR(56Xn;vah_>`ih>G{ zxS%|kiwfu-k`s=qPPMu_bfx%#q&4(@iw zB%F8_v?yC;SYU$cq-gwdJlDUmLNnl&s@W;T-V7II$yB!F0LXS!%2xYh`Fyf#i|5Az zE6(PR!vz^jWZCwi*yk6~Jca{{P6D=8&k*d@m#-xC>hS=QX z>^F+9y88$FB@b^?dn}8XEeuqO$+t<{1+_u3Aw>$5|1h>ev;?)8>H?8EGaf_y^dWN@ z>!%lu!v?#h7|Ef&Q>;_@S*?dYVMP_$JW?Wu$CJuI8ym5mVxaL;0Tpu)rjOn3sTwF3 zLs|)zwhz^V4Y=)?0xqgQ_RMHerV$Y5IipA;Ro=z{w>@6nK*B2bxrc@F==XN$E&8I5 zy<7nE(ofGKti-E!MZ+v9~x@Px-vHWnlS7aKyWy^@B|K1nBZR5-=#3M}`j z)itQrhOU3@xX#$1IvvtUm(J@Jqgh%DEF;S$Ii4VH4&@wJbO<520^00}JZ@LuXsoqj zgMe}YjwVdl`)^IZl34A_{%?Nw99hN9JK-^23T-UUMvB=$!TUT|rV3W)pnj*Iz+;Uk zQstnkZF@w`)N`_vYU5VV%YqGoT_irbQ!{QVVBGDDOTh5lO~29a+pe^tLv~;GJCelX zs7Z;=pJRjoSaFb8^R=0XGbLY@5yi z-;e)v|HN2YXY&|BEUmM170caMDboWp>Ex(Pc`h_-Ss7p=Tl`ZsO+LAE3x!(@SpciJ zN%{05NO^8&mQEU<3277vaRvZhc#ofbJ2pf+ZK#UYaYl`%3^Hb!O_#enC#4UEjU z@*%O;H0J_+kHJ~$V{Q5rvv|uZ*~0p7mCwu2-ecOC?#MV+8Ag@{%UY;BLEus&IH;DnVF!m+QNpjP~hr ze_q_K;N~RoRu$!Mf0?w$KF!Bl=awa{IYKir1$XjMD}uyi$90TjF1_&Nl(St(Z?XaGP&<97rp-nQyjI3}k?|40T0B`t$|C zKE)~N2El#^x)uu{TdajX?3gx&plSx_RN~Qk_eA*j9^IkdKL@!5#}!9KElQWlV>k+| z=jRqY;yRQ?HOqb}vO*;|?Qf2gr96(R<=PDXI*M6Kkt8YxDbLxGZq48*cSr7@vtMrZ zL*fz0P^y=Pk54-H*${CU>Sq^?!xkqSGU3mweiUHefReXU3n(WHTk3A6m~@J)r(&{z z7nwHwbAsX4+oEG$d!zd_{cf1w>v!vv?G*OAl})=W>v!u{uMN#~y)$GTV-EPRSc7rE z2L1#md{BRSC9NaWYB*M;cI_nFxy>;=P7|N9K~n|A9H2lLGNwUwecn~YN7pqU-6SYk zFeuQ0>KZ6EfM&O@l;?fRG{yw~#Ubc}-+pEJYe2IvR_^rZfJRaeA_uAV6&}d4 z?oxFN6Br~x4wCULd1MqLtrXvLMr8(Pw2eFiJ$H1hpp4wS!EK%uKkpv>KRd}G9#1@% zY!LA=#hjqXF`RfR#R&|?K^973@deFfjOl^*VJBUALb@)x|FwiEJ9Wkl?`8RA`dw2z zh%~E@c-@4IPLgQ5N0;WYvRmCo9xJP$EoEu&Itfl=mGl-W?N@hydQOIej{=qrpmlje z*ezZ;8!HPh2dC*F!93Ntbk;_HV_Xn063QY|{4?~%OP3;be{` z{IqD&;g>zkk6LX==Pze&&oZ`=+n)9AXcr(OjNH}~{a#RYGhPkArej@EjoML+HTO{% zjsDUoeueub+~6@Idg)AiQrq)OQBV@` zSo5h2^$YGzSrIV4c;P@s8W)^A`;qseA04o7Hpkob4@(!kb99zyp>(0NQJbsD3I~bD zc#kYd7$-+|lD*!Wrj*V$tD$@t{SFiWJQ>wFC?SWdrr4090#rpZ?hxW`k-QNmqI|lF zxI}H3gVnJ0Ds1+ijh(Z$*saD0?2`$5=lwL(UNn@qh+z>3#Z&?=RQlbr$r_QBS<)y# znh7QpQZWB3}eXRrhe5S)wjs{siN9Bh)IA~+Uii5wx{WNY^`IoFq z-ycgxkjJZMwv3<`i2SL#BaPTjtVQ{`+De(W=*?D(>*#Wy?GeVRpbZQr3*)`a^mZnX zZqxLVOu8E?15c3*XdDE1YX(U`xI7m$4>IY!a!|8%?wmD7gblYz++p-3SNz+x_jZm& z0-wj;EsF&Hak4jh4K&(0F}niB+#=6r#7PbFmkp3S?xg?ws^X67o~!#9J{uLT;CObW zYq###O$nN84X51vb?aZq3Lb}33T(nDX%qv~3@z%zg3oFa=l&ozx?Qt-dY=~vnoqkH4l$A&!do-Z4Jn}OLt&n9 z?yNBU@ju(<1`zAi@a%GzZE*s~1UJd2|Ni3W2#ylNZS$1TglW1r)^jiEVCF-Iv< zOU0yus|l)FnmU3qNnmQQg4l8(L)!_q#UySLLzfus(a3}P`yr>r=jgpc<4skeph#Lr zHc<aUlTDEfDh4IXQ#DA& zLqSM{f@KX|?vpt=1^Qp6PBzx+^8fj)@T@SyKUaTirm<7irzw&j2*mZZi%rw$VKQ0( zICkCqknf(KV}*(2;qg?mm&eXZlg+@_QOpsF)S$CcEV$-f2`w{vmDM0Ga8b~%y$H+L zN*b#^jnzs#!6YuB)7-1+J^E!*JZur6H=5|~uu^8B^a^yT#4dvv(}&WqB6*FV(W_wa zN?bwX&7kK4w0p;E2izJ|eVz>}oZj%lG2&}e*kaj9=p=rvv1EB$U} zb^OeJH>9O(qWUD8y^Z)i?m{)kTLa(TC1_zzYpVfTRlr6tR@248I3`zL8M^GpHS>yK zmP26;%wnUuj)lyBTgDNsoTB?Adfd9k=l||&~QC85{;bFv{45^xxvyZC_ z=mUXS@>EC5-d$^^Mlxk18Kvisc_x1EKgVKtc$`vXsbeqo#{N--(p}zLl=Z4Qx?8^clNfdOEO}ad3c}T&tekspCHYXq9hz=+JI{WR~E=2wCU#x~_sIw0jRSX`Z zgQXA9Lg`(f?E(4z^{OWUmQG_=qT02$L+ZsV-0uOWrGUN;%cTUULByV8Tl58j%YwMz zFMN)O`l3^WCNW-rI|T1wh7O!J6%9Ta+H~bnx#j$9{SnbBQX{DFz^$$B7wF>1B3+Tr za@_{ylYnLe-l0wN1$ayO!XhEwy+dpQ$@UXclh|_SZejK`>~gk={cZS5*#EuvdDU%>up7zBIOUi<0D}&#_{;lN*j@Tb+${2d$A0x1n}$+LCNc{r zl71?tR+mYy)|5pa2)r~G7UPf>N?=kosA_HLSBPSgdHOuiL|ra`*~Q~h$SGM_WRiOe zbbZIZQ^2p&ZDDYcg)TC(_KM^y4YhJOE<`vKz`uQ(PN+n)+z%xwyHpwe3E?=&#rrg< zsz@&U1uI`!;)^||u*Gz(?tnf?P)~23wnSwi|K;jU;v{lGkQ3fXFDDD7C)}T1mF!y) zfeZcuIxBpU6dOqODovi9G%DEP%vz+s1^4Tu7eO2WUzIzm9B}6wRB7Zrz5vS}ot&KV z3HAGOD(q!*xmZ6Q2b5UUe=&@drYj8r?p`xfBD?KX3)qEdb%kFweV;bgyXPp7^oLS( z#`G!Y$$coN{Wzp-DxR^PN%6n*X8c?6Z(sWPVIdSxs&@(s_2+cBjh)rnG_leX0T&@N z|Cje~e%nbA1Lp(H-WH8+|ATi|ueVQeae)qxCr*}x*lBH-!Ypn!%*T0g`|Mlbo39AC zN&4MZ&c3Z^*7Qamnc1f)7S4Clz&kK39I@fR9i4LuT{->lCt7E>CCfsN0?C?_UE;A> z0p*}!39;1_6Hk#9RLr1K6q0nX@)z~`t7eZM42BJwBZwGw`nYZW_)>A_0xKfQrk+ug zj!7iW#;y3AVjfWBK8EW|;zb~+o+UzC-NFFA)1!)kf8|)I-D5w@! z1>F~w&+b-K>dI$thqxOC{#!I=HOAwNX@Y9K?TL51Hgb;1-_T5s`RIkVCCZbZ$Xp91$-@_L@nc`_daTL$TmSTVpdWed_MpX+M#P#ZH zihHmW%Ji=Sf&MRiVndz;6a-g@J`F3?b*d_a#!bG*$V2rR1%6o#sjxl?wsEcy-Bc9> zA0yWk#e+>zunPd(M83kSAh_QxQ{Eo1BFb2;##7_HGUfMZ6Vy&6%In>Wq+?H{ItCRM zYx8A?#|RvDTjTeu(*LqwWV5VJTa?&Lra^^{%2h?CE+3S#k#7schh6i!h0CSJUEWyF zx<~&AhOJ-SsTs^EWr?udHIse>ZpB?0q_I1JkBJfrX@-p<6&|ZSPLT!Jy~d)rjkj~w zeBkua7~?1mUCurlBjqM}>%6S?{cWzjbO>*Y7|LyI%~LUuSm|ZKp`cFsP>}I5!EOV#Y1P6|i9RqhFn9}Vwl>n_Q7kwB@#kv_ zCrG4jLD3g$m66e?{WRN!t$VzGDU7#P0V#BS%N_lXx8RyKZB5z(u7EPZ7QwhM1hvFGz z6qS1StG`gjMOTI{6>Wex!2i$Qm%ufZ-TQmR80H-Nx?G~=)yVP#V+ARYP=rSfs(F;5 z=hLc2)e6r8LHH@hJ-3BJ+M*0w;AiP3^pHCD3VpSOPKGVmPU6EfbkAEnncib*0IeIk zRf>D*P^1GBFH;SDhQM%6&z z6*3he$0J=?$-OFCP1st3;SN_aMab{}@S*yDe)ZqK{?q&7#T2uEBC)1MKz}yO4`TWF z(!L$xla1!WMWieuY3w{WPFz!Q$i%qpp%}<=-bKZnXOL*ERgM)Du=Yt)HZuF6E>n$b z=M?&``$nH4B?wxqWcp-qO`|}u$8E~IpjHKNP)1jYPKPwA z(v;dD!_s8e^u-{1QOSuFqzSfq6a{wD$O>FbZsLm6>bGRa8$$tt9Y5Y`SJvmDdldqI z^Qt{IbB+_67ArM=peZXa0hk!_mx1;OOC~VHX`nZ53*Q{rD%Z1YVOlC~>OUR{>mktnh?(w=i7pHY z6lQbElaaLJfpvSb`gNDDnTz8*JLNbrTvANn@whltt3z*I7RN&A!(Z>q{&zPVe{xk`6@6cp$GJgH`s74j z8`&fp+Yn*}a#;N6_QdI@XVYjVu;)Ux`6Dv|eF~t-qq+-sI9`{!H zP509?c7~VvflrODo_BPbhAP)lTA+SM1%1kvc_BxdurWoOS{QJslHWSz%j*qK?Z$0v`u+n;(PB6Yc)K4yje6m!=B%Y zGr#6-?w)npptTi;EBe$9aiKDxECegM+vr1dt51>AD!s8CR z#Ll#cc0LRlP9>%jn-(jTnA)(5!j)m_(=!TXO_GGy`dyGDtr$XqSLJHtknVsGG+;^_ zxh87F5XA;d!?p!1ke8xbw$FS=?6hXX%I^2njQJjO^#6R+sQP20>eP%newv_U{76>X z8oqLyN14S4*`p~di$PWYFt_bxpGtNBq0C8GgF;kh5idP#XLRgL@Q)#DNPplGfrgq} zGLjs6>=@+jG2%A%!3r3o{(bbV*Z#R&C99~#?!EG@P!gNn^($(k`F*TnaW5btCz+O)=Cn!=&#T2{81?}-wmxi>-&JvIoAjxE_csCcA z%aG-#Pxjd7#wV9s-zA04;+}gmM*{6J$p9!`m3w^8a&oxqCNCVfFQ5?Ug)prE z+0)`7?Jd)Pp?{9QMu8u(REn!0C^l)~FABmmxQ8@8QaWKt*urso36&O5H^5Z>!@lvn z%t=lDtly>~z8dD%#)k9fy$H|Vd;91Au|CL%sl4{w`6R=MLp{LoJ1EpsL@@;vn8}zF zx=+?4(M?;)To~U@_j*92&#O+NuXFp%Yvt4|5&Gzj#N(M9E$@_YD;M3h8@(=vXp~Rd zB43Xh^&qz?^?3-WL99{S^4!YBS~=Du{_J->aK+wO80dGnKUtZ{(f|1)!5ikQ9;fxb zSy{)dkMFg~;2J7mxr9B>gVs0Xv1A(v7hCO_q0hFAl+kHjL<` z(Zw@bAzs+;f=?vNG3=A;ebKG@cXR(z#p#bgZlAp9JJH+3I`Y@Ke<}1?5_Z@XLyEZ^ z)PWrK^4Q)88Zav(i=Yk3lV_N3vz!(|RyIzpoP$9a?}_ug8mh)Zp8UX*{uub%F*#SM zJ{;Z%1=AV8^jkbb&y#tG+fLsK-V{~`wQ3p#>KgALtpK$YS_dUxj>>wn16+oS42#dR zGB~4@tsRmd8KrCk_U6m`rRf6a(v_3sg)Czxkas8o6uY1b;2#XNkJF zdt|yOeIkAtzsSFmcW*+zZvqr}#Ss);+UeCf>8`wq*&?~igGP5j-u9k*qQ22UllpOW zummcDY=~yk9~uw&4ai9bdK@{Eb49&i*jj+^8Uc&Ot=|QO!A7m?@F(J`;GZ zHvnMKT4sy)(wimV57-*EG78sRSV-9*edM_#9E%yBYvt?!$pNFq`lm%K8ZQ>yOj>46 zJM?Upr4vKR3P;l4ugVWXo}58Z19ZBuUL&8=9oZ!5h)~By;IarQ84d)2Zy2PBH}dgj zq0dR5tKR)Cc-#@u0(^A z+v!C?P(h1V3nt;O1Z?}U5of|`rl8KYzMR$@c-7OOSt+WD6^nEeP;3mdCU%6sgjKTN zd&FC8F5=>}v!NBi7Cot1k9&PclcZ=o3~>i(lNHY}jRSP#Cx=Slc_9?Vzy8u1?r+XAb$${jW6DmiMP#XGnz%`BIKCp>5cJ&HWVqwP3MXds&2T^a zdC#-M$sN_t76^?vS-R=U@5oAa3nVA@4`rL6YBR-bq{uodCJngiwx89pA-@; zJA63ts;A5ZAG;}L2Su`|m?O$vt`}fYi@w3#;PIJH&d<=xH^mC71F8cW#QBlQK(3hK ze#>`moA@-1ZSsY4&oiq)%itamE9FEN1HpNU=zy?6G4~kfYREq6^3Y`YU+is-SVP9l z2t7HqGk(r}&-bt9oV`xFD6zuA`^BtI-db*ftVdkxb|GkWBp#(GG~QZep7(;8C)`$t zEtJ-U*n32uHxR5w`*|JN`5vq3hWf?7d7OLM7p%xQ*CfF*!%c4Lbf|9MC^*8?a*zTZ z9LAAtj@R)r9YQw9Fn!XoXRMGhTB11>Zk}u9v{qxQ$<&cdF{>%Eii$zi0dKx;lkruh>v+mywl>sN|Ae#Un@k|7tU;# z0&61p(Z=_Jt85)-o}iMP6JjYy3t1{pbx-iDg9?xnZ(fd`J0dd+8^bZMSTG$*$DFYP z$gI3oe=;{5Y!a`WIFrVTnbC6gxZMWLnsmPoQpdk2UNIBc!$ProvCC^GcSN8wXy9!| z(V%BW;_&@bUGpg8&gHSk;}amEzz$eWo6`Y`vx8&W@f5S1B5|;DA;{ahA9QN(sM41P2}-T< zB)@)_gs6BugUX14!T9Cdgo%NH8o{5pI+xxw_l!Op)qDB;pGTD&T$js>q(;eDeVRdo zXQBuwx+H3h2xCq`bd?wzrG<8(6n z?7Bl_nG;7?fZt$H7TiXPSx1p1Dke#k2?_jAaa=L!3-_GJUT(5{lR__BR!!f2t&Yxu zkUT_Rz|}EgmVqr0hB3CR9<>exOZ*0sH#~2?>^{(ouzKi|L3(~z!}L~p@$5^{iBbCl zFHGGK*dcEhZi777J)GR(=cTb21yN%bC>1m()u9Dw_9DsCct!Q`PX`i~V zMpdj}VQ{j1XZTKG#+2?z$V3yK65r?6fSE$B^!L|MAFZMO`2J#QJ~j8#_t#MG+X;=f zV(iJv*`BPu^(se_X1gMEAlKk$v>r?Pc-3omgU}XYXU;K^{onl6@ahuy` z3~lU%l{p#ZvU=SQ&1J@%)@`maF%VCL!!RiV^tcairMLfw((0hG_&dL zXVs&98nc4L=%9wSpT6u2ZAXEe9@0{GL0W1jy<%*>Z#wq|x&EuAZ>IdbmfV=usLJFl z7<2cx)o*3a-Zp#wTMNfj(=~LbG}HSWXuD@|cgZRwyP|t%;<>|sWD{DRGe6h|8_)K_ z3N6&5)X)FZT+z~saAZl);xL`0!xd9rb_&7woGh=GW{dX+X3z$uc1+|$p|KsFaa01Y z2U36QrCFi_q(QNVAaARK)2xOwAkTh{_PfGrK%6&`>YcvN3tsj$$SqGy`%`C&4^BEK ztfsLp>&7%ld9MIT_UqEa|Hs5;Ks_tOj@EwI?QR~KwGz^-5qFa^AyzY~kr!f{^2RiE zrr?$*%4FHKj?$*LmxaNwsXu!)V)yksZT#}l(!lzcJqavJPzId8Wl#m$8L@l%MlYx| z!LpNlsDEjdtB-nJ_f8(`uoZ>{Jh6bRgBVRL{8Dg<6=X*pKJ~L6FQXHr>GiW8kt8Sf zse?AoAfNg+iUDaZ4HbiAtf(E>sM-)z$FJmefEL>x5sFPVhtA`5d8PQ*$Uaj%0L_9r zemb|qJCk!5N}#kHJa#x@Sixk_8rAaP+V3hWm{7;te&qSGd)IYzwiqkx^!|=(+yx-C z4l8FYb?J9$WgbF0so~h+ka#kPEV1$AIY%F`!p7(vMcn03qX|j%+qRbMw4{*YysK$) z%*3peQOtgd?8UMbR8($~9GvjpJi#5Ga@nFOUEBiSdy_PZr@Z4jn1HZJzl)(cvQ{L^ zx6x@_q%1!GWeNGgt>nEjUaX*92&Cm94Zu-><>*<04%Z$@w>$%uzIVZikN?>Pc`QLS z$mCc*vgYIXz2sJ?4yyBa zhwq))2&LUX_y|(ko1@ahQedU4{>=4Z=mB!p?>u7{+hAox*bTm!nXuBQJuX`}8J02t z>NtBsCK>0U?q5h3lPylXlr1-zbMh!=7e#WYm{v|Tog&7_Os1k_MzXw)Uj$rn`Kma< zfuLA{x?Hx9F(|iP^}ZLOlc`(i3`r`ml{86u-E}gJVyn0?7%nzR8h!3XuJT(5+>EbV{Nl!`O=O=FgQn30G)F0>nj)1{%vJAR zk1IeiSQY{^3DPuR7^lw#y~4{lkDz)I)5`QE-SyIT$S!N)SI^8L>TN*qpq&sWP@nNk z(<`ZHL5W&Vt#uW)NLnBrS;j#vbM*$0SO?{E;LpL&SmC2S6OQB%YNYK#4Rm$6L^nYv zSskpCtn%71eW#Fp`kLLqG#FSGjtAkyrbBjM$}Y?PiFuxh74_&{T-1#|?0ao`viwv) z>BLgEdRQ82{ObNsJ^IPOu=?_yH&=XarDRo*(Zsy0JRu$BsI8RslXa$!9sla~< zvQ8M1LT%AAAv#Huq*ab9qT$KrJb?`!*xR0%bUus|R*)W@{Ae*>U_{PG`D6b~R#@6u zJ1_E~xPP$2?lX$nfWK@^a#(@5VfuED;u!`0yQ1eVe(RFBO_t8>ifmNfgm7~oEM($2 zXGk@@e0Ds?kqh6!Bgq;vgCBO_v31Bq#s2y`=3f1mDqr3TKaqi_95hCs)_tohI1dR1 zta3QG>KM_jF*{@8KqQ!VZ5`) zDQhtJySB+P41QpQj;)Wsh3zqQZ2E-#660hdyes?zNw74TII%Bnr^#fZrI^hW*+|7K z;$?XziPR^9dgXDzYKPKW{VoRnb>v&uDD+z9dm|e{4vo+EJ_y`)b*@{*pwWl_b0nN> zf{leeivE%VA!Hpk`&M4N?P`Qb^;gP2B@3N+GYpJngMxuc6q87icq#^!?9sKFLZ<{n zq$dO1s<(U#r=)=&kzPDIYxK9n|Noo5f%}1KW`zW>=Naj z*rqg@fao~IfN*g&B+AoWu-!QaDzs(vfII%c#3K22y2WFuw1wZwS>z9FtQbu+gncxM zKG|N^WchJLxdcUOA*%*yhD(ArN_2FCA_1H!NTs(O*hvh&noKe1g(|_hg8H`e1t}su z;Z>8QV$$l!)sSVc_rsu^ODm+_FO?2y!NUf9*6h%Wud|2gFPCmvK(2t1 zd;6abkxh_z+I+;0SV3ZRz`y+;lXVWsX`@d<*<$WG}!ZZr7S zb0EMABvj!VWp(sbX@~c5VSLDCuXElF(tY&3$P=X3)A6{m2_*yO;^|L^qlFbpsI%HJ zXKCa1ru+MTwPgM)v%LW^xCxJ<8XQZy5_Bp~bv32m6`=6J+BGxWAW z=$XeHexdPJ)8D5WK{2~Ia|Ov7LuyRE&3zP8M3DlFX0*v#nM+YUWFBXeUkSfW)+=v~ zXpMLfkwr?~&Pc1>cZ0(3z!hB6r9-g=>XVha^@KKuHma)W`?4O=#Z8?EN@vljvv+=H z-Q@YaL(CzcUU{c99pY%I?)xU$add3^)}Owy&p)!?V6b~}ownTK)?>-fUiPGt&s-~b zD|o&1y-5|E4(}RBR=DP_;aa}h0Ft6{LEC2C6OReDNELR>U*Z@`1&VsP?`B{2TCT?t7&3>8~q}F-k}eR z`FqXw{P{!3>RoqU7nWPtrPL3M==!k!H&e(Lugn^G+33M*(ZsAFQUw zI1)lN(H}_0n@PqLNXiw)1LXIRa{g6syxlJCm23Qt2I(X_$xUIuOA)s~sCVgK@=uJ? zA7s!OQ_|^UXeM_3KPo{O)e6gIn)Y8q!KH3ELQO1?2g*!Mxh2 z>LCY71+RjGS5xTE6e!%6!F@m+ysUXTQnn5F(^>NAla4)NHIq_H#lBR8al-n|=hJtP zTua$5PMkV<(q!#cK`|hnw4aJu60yqjK=87u3wg=%7XC8-GLe?IgGr;SU+VzA9sIYR zxhTHkeI*KJCK0l`Wr@lHO9H##nMTznzYbXuch`)|;+qkVeU?OA@WQ?tRe9dooGp-r z`;cs&v}Hn-sB_WP}U;0$?>;XQuxO$4P=N#Dqa@KLR>Zk91ZL|?X3zq~} zfuO3>x*P~3Lk)8Et)iF}6j@5e7-VyjgV%B)K^x*r+4RT~hz<;pO`|{1cm#i1AAL)i zFv~7ay*kr~2>%7kv&mj|%U~zAH)l-nQ$sOT6gf!6Y~&<;uMzSuGdU}nMwNDIs;E`= zh+oc6p^qs(b3Nr*6PQ9@5pSmt(+}UMlGI7FxHa_Z$O}P@s$-D#a^u^V;aQC$SALj2 zExiLG$vVl?KGL~${8G1+NoRTIg9{W*44&6XtLb}^h0^&R*bC%=&%DeZmi4g-c+asg z&$~4ooSuo>9_(v`(qpD}DbcWllJo8rNr?%Rc2W%JLT;yG8abzDbWcm2SjxlV#Ujbv zo9?G)VC_edXtVd+eF4dGv``uaW^#frV5#&>$W`xpDN+;}GO;$iwsu^pI~I_Xg`{#% z29-{Dkyl_S17sh-4CRxb|IG0puiW^S5ipC^yz>)sfE{3**!EmB0nAB?0Y-)+RLt_P zT^9dukMNrNy~wTNTso7pF`$H>!MXn3bwA!h&3o$<-6bwnRO@3SjjBgb#8}BIdrUE(w~g)#jC=dSkBh$jVD@7749RjCnwb;I?KAdA?fS{)*=^Ac zsy!r&K;L8o760Z%@JlzUu82G7oan3aPwvZ5x}%nKPfMc<#O$q+eGv6LL!`g=eE3+n zyUF9>k8ETMYnEsD=19I&sGB^E2RFUDWCtd00`FahuXe=qMF zyO#XH+q&<(>;iCEqBLdcgk{s$dengrtzCXja!d#+#Bi)pWVvnkxB%;!+o9-1GfX<{ z-|8X3hy`+u@B5HGhV?`2kej_XH05RIpF}T<*8OIvA#?jc&|WbXmOq|P5ytus`^NJ! z#p;S#%chUqD6km2tJ^nsrs-y=__=Yb|9rn9ut4bwLV z-k4S(OCV=QZkf!|mW`M(V+9YYwd~hB&9nZWEpT^UlSiyn0+fFjEeCQ=eu|-5#L0o_ zpFyeW#gL4EewQ|&1H^2X4U^iG*<%eAXd1;x1pnB=#;``UhuwI7AJ$>>hRYje=GjqB z8wRwJ@1&EVk87iTLEIS0?Pmc)sDs^E7#j-I>}hz0iYDV`*zCvijRu>$*3q=|Vn(vLfq51=PCN%$F*@2X z1Vz0x9vmf@{?{rmoq(0gN2A*5dVaody);8l&NOfrxz?LOKjajB*WKZJvZ2C+&5#eJ zt#z1e{qKo)Z1kmO&{={8sJJjt1tSe8kbWvwdpr^&Q-#C7TeH90=C9i9IY*yb|Eew7 zcc=BGdbmqM-Zn;U{&0HZ_sJnArqpOQnPfhvm^z9ar()vBL-%$_foxT%dp(wTqP5Vv z=X0ShMFGiay&kQK21SosnNUYBm~u|o0B2dEBamgNz7*9Cszddj?JzraLB?ZQ;1=e3 zc!T14_zvzGff{MB>tt&J9{Z&EsZrPRn&6tC4JzdIrH?1ZrE+7(b#c`ikHryZ#^uBF z+dUvTI6X`kzLGf>ndVnAZvS|2*X{Prp=^F@Y=*npR5R35TVE4aa3B4*$=~?cyuic? zIVduDHUQ%uZZmF7!~bKw7ciqF@D_TWg$z&waRu%{@`Z+f^$jq+{;1dLfO5JL0{Az_ z_03eL_*F&baJpnDM4REO&W~&hjpyX}tQ&_ykkBKpDAY(3YS_Hs{DDR7X^^Vl2(-0n zf|{XKn2p4k**Z>x+p!xnr!`SO-P3haW;9VLmWrNMDSv% z`!W>0v)S4VMvKkI!+^WZpD~n%M{mtZHFwroNo;KkU+8{=4?c_cQeans=Ae@p(%qL$ zp9}v(*7NH?bDbV`FbG8v+Jkl`(D|7#!yc4;er}xE+DP{phVr^Q<4CV7| zedgpxKWq#!0%d7(=Wep&l>thv2~bKYW*)x=SC=MLJbMncYQnxwFRCA69iURk9U-oaOx58dgjs8C!T~1$@ z=0l93J_Nirs2YH#^a7M870F>y9a03V<>Ag!L$jAQz{^&kSZTL^x8}&po|b)#J5K;~ z$FLI5pdY!PnxLNd50P26DU4`k7#w<;-JIAP@}0@%yp?Rq!<=_gCDvFjJzEtU7|e9k zTN30M#Bw2sO{4#AxNknop|#ymuWo7a;%ZPd_8g=h+$Ex0E; zBiSiC<=HIXN!Q8l+NDm5^*Gov{H!~*!vofb-+~8u^o@g`>{J+6rK;Lrm5@|R1%pmJ zQyef^d*)IMBp+ns+S9=9v>`Arxao%n-pmfnkzDmod_6BXQ=BEZ#?5tAH!6B0i5_Ks z$w1POEH6yStD$Fz){7n?Io`UZ+CeJ(5 zzYP+X3T0cEBmU=vjj9r5l3yo`ZFl4r=A!Tvea&^A?_Ku;6Gv{3!g9RrJE+g;&HB7z zVSPeP{MnQro6j{?jC2a!%gyC1n~pzsgv0cXcKL{_8Z+yWRta59>fLT_%6pOO7H%g= z^1BA^2yjw5dQ4c3#M4*qIp3l^Zm}8(=e5Y6><_tV4iTr_)mXvq-@Et-+!PVc5%UFy zUDbF}!CAr065+lV#a+;?o)S>W+v1n+8|UV5Ke_>7L&L5;(E0p6*kSkoJ@TZc8RwXf zN>vpk$BE|{(Bc>rb}6BlVu}<}G4Y}~Tm0s9$x9-z2K9t%T?i=N7Wx2F77tI#=#_r` zE*FRvSOUxGcoAlqXcX%?#eq4Iy<9aWT7Ra1v=Wre#aK+XV9sZXN8-;ES^jgNBV4(( z%8}@+P*%W!TlyPuZo{;VLxay#3uyi5`S5trU3T#C@+thAxrMRvB3F&uq-gfuIBno( zgQ`sUfLv3Z2DV$gp5?z-g+HrB>99k^#FI?_VSdLi`dbb6hcXhj^*p>*tVY6li}_`{ zj{jshVnNRPml11XLHt8+8U*Tg8j?!rO+o8|z7REpl|xqTuIMr#{%Z#6J=F5a2%PVy zo!TKw6rBzMxzw2&MT^g1~o& zTMY-azu`Qa^iAw_$Ll*hTlsx1IsU8PXjC=JwENnJZ3W{f=^Ca2s>Lpt12 z0(8S;))_pY_D0X(zK3&MlHzuHI7unayVwWj`N6BBCrK%2Wp;Sw0}*m|V7{;vayQfH z;u*>E-GOKLn0a0jgy~pWf_xb&kKr*=AQX*X2Hv~QS+(|LnYT1>LkEhb-t2cRw0m*5 zxd_CwWgbrKu(cxfU?|@J&RTvEciGq!{Zgw-zQ*>1Y5RS}whg`gKF`*ktS`Kd{pwb9 zxN+_gx<+p$yT_0dCa+yN#gtNH9~FbCJL)C8&EDx@eX?T{3L>!JN{tcPl@M1sg2eegnfi+VfcW`=iRwyHNswf?-cuJ|#!U~y);i!ehjjNI5=SnJHvqSG zpRCtI7Y^Cm$AtSJ+7Fb$up1uggm3(MY#S8|^Z5%e+4}ukZfAW|EO^J9_r07t@3_lX zj1X%2`i*o_Fov8mflwvIR8XW0RatcOeQ{T0mu#W5oNfc*(!X`d%ds*v;FB(S1IIw+ zXn3I?kbI$*lR~G{$!-~vdTBe@ag8#bT!-qv@~8w}EVafHiJ^~innN{81ReIBz}pdC zFFio;xDDojcF4?6Qr?zg>;Hohp_~^+Q?h>1{s$wB_WsI4N^XrIOHGK&A5u&YMeb8E z*@7($x_p!67(G!h3d#@9L>qre9IZ;O`9~q6$u{9Q%pq99L*votC$pl@W8^wdVgEo}2T} zm0#B+0s5hO>3l)2$39T$of{j`JuP2}%4!+(`hXG4t-{HX#T0FME_Kk=RV)7dJ@bMP zHnD9d-kMvfYq`Qt6QqUTleACUD2Y>c(r12o879>m->w6e+eTHnq)+yMRMT2T4{3LB zk&w!fBSk>yveglf%$tsPzioxGgvYp1JF^ zE66K<^@Z7lpXXh{LAl7S;#S8phYcMnj^+7b2bC}XaQ`nH z;~wdQBi%J*0lP?n^X~pgs)@&74aFo-WF-}Y+G;w`8wCy#Nt5K7c%=#md@g{a=)efl zW*XmjTQT$^^I-4$vnK!Y59ZZkugEH7vk5AmNPg@NoG+-Dr0yX=wgJ26A|P6I@Kj@Y z%xpn~ub$tK}eP>(7P*{^7o zuabZ6yLWnCaP}MdzHO>D*vTdcYGj!~pZnf(Zv$zv0@pNBC25cz47xTwJLxx&+@+` zeaQVdoeSm9suJ>5_Yv` z+AlqfkowInh_a#Q;Watx^^M>Ds}UoIe|zBzQppY@PCQe! zm|&!VV(KY~x5s?RtmPJQTV?f%r$@>1D_-qB9d3XLMLx&eFri5*1Sib@4m=Pmc+7ih znR;jTh4CeDUpJ$`|6}#JH!nqB61T|=#_6EZ94&vVFkVKvwYu{gFJYLP$jzG<%MvCU*W)>6zGiX>1mTG#84s{mqx{SjT< z(=*mW79*0#Y5X4h+;pEWh#g@n3l`s~?O%r-TCy_z&g<+xkU#lb=5Bna^}tzi%wv!k zqsbb@HTgl_MYo;d>h9Nd6F{JzSD{!lwq1zK$$3Lqb_^LNEIj9q4nsZ)kTO;VNmts9Urbg;>D6nS!A*_Krxp%o4^d+&LpEM2VkTa$rZ8G|6 z&(OPvTOwl1Sh0i6z(iq7&%xFIj;pA7*Y@pcOQJ~aSWAYTFOTx58Mg9ruq&)3z<$i1knp|A7y#f70`^=L{U@FJkcj z4W66)QYM{nzr{NaG`Urrn!s|p4z8n*Ey*vFvpedRBqef-axJF_G6zZ&4}_&|oxXdx z4mK1+wi6Csp<1^S7N!>}Bjerg4<|3piPx)E^3rbc)`Z;hMDgrS=|;&>rN*y^T#sm{ ziy_pLCAuDw0n*6rkVRY0Jvza`5PJHavjx-BCmnmV)V1grFZ+r#&s#?~Le(@7d0>5e zY;dpVF75*|!ilLkPBxCRV~5m2@2pJ5sPVNfd~@y{D-<`2#;ov3mMpBZxmt!um8U*KcOwTLuIr6f*zzTh; zA}>s;6l1-`qM$U|K7m0ZH|Qr$No8tjb*ua#! zuxryBRr{kVxecBNNtv)Fs$6y!_*YiU#QVE=d&PJ^gFBz3JvmPaI3MJYdD9+{VG9Ra z+Nb~CT`<>Zd(QuGQ3Sa+hO9Q3D;ADs=8YzOR7@XKn-<7g_;vip{`j583d)4#L1#%k z=XOAuaNXqNL?c!!`(gN#UUr9YXaIs z4d(~Nb^PP-Dxad${jRA>L-tPG>)U7S0sF-s4wHI*hwJ^&I@yCL!&B&?(Yo&BY^PHv zW^nFJs^A!|Cy@%@u9;P$&PfI@&UU&=bPF~I_o$zxu#{QNA3%uEBLGQ@$`uaXD) zaKukmTR1BD_q9uZYcwP~+#)8ED`Q~p8vF)!P)s{TZc;G^r`H8#1ZKg0q==gn`FT_- zcVBpiY&$6Uc6b-aG~82kVw6U4R(6MOR8<9Gy8>B{_^9%PxP@OJtMo~ux6@cRs8Pg= z+ND=HaX~elR@relTkVPEV4ZY{e!lCIp{zp}spV*yM?MXzWNN|Nh41EkwJ`d$YCApf zoJMg#k}6v3XJ~Q$7rDR2`^OclJv*m0sM3@V>66?mAl8{MB~RW$x9}S{TG(u;%SvdgbZh7?X?NAgba2g>$IAi0Y%aunkV{N$&gJ`#)}s7PAYXWjiM5Y zO1HdL;alp~Az9^D;d{|7(|R|F#k zhrTlYM@F+$8y)fvIrWk$l608Z&=!ihOpyynm6a^7mm0h-uq%hOyfiw&b6(h*v6z_G z9f>OF?`@}XN1h)J@+HbVZ)A@FZ_mtT;c-PhM^6{Cl_~}G`ZmxHSjqgYjJ_&YcXN{b zHhb4VeG^WDxzU@w^)3P}4c8bg;GOrbYzt(FfRYzg`jnlvq-yjj?E#x5%w~d(2unj$@&|y ze$ub9#~DGUn-zMQY+|>FcHZp)0k-Hti|A~M*-nu(DrOUKYsL!lByGw(d0y}Vpj8E? zOp>8&R$P|X$rkzCMJ5k|({?$(DyW5D9JNK+6Z+VH$(X{ZZb9t$Jb8V{C2GW=wm4H|Ih#-@*l4(?R z$qiSwdtd}`m0vrJv}>m(REqJs;}BvPwISn#&7cpVt?%#8^ncrERc=I7+#yFSODT=DvVe z;GG-Q!bd0Jc8?zNU|bR?wXXMBP3{W2X2!X$ml#@ab4Rl09&#zFPqs&Nl5B*cxq&+^ zK=!zU3xtngjT$)lkq@HEX)L{a%-iUW76~^C`=d?;w(xhj9p|r?JO+8#;;3p7gkKmo zs2ehZhB9lhK$7(t>+PJI&E^IDPP?hMQlh_0hGAFz@&v>!re&e!O7v93xr)>D6 zf4@dnCPa_%bt%%U;0u`;mHmwc!3AHx414zT!54yBWsAJla@&0tKr!?{+jM2RE1TDh z)i-C$s66lWWA$My8nYS|=QZwUPdBfcV1$cu<ze@Z$vPi^6?TX*7mTA=n8?>%gt)q7B2eV{C@<^|9kXA_f1!7%OPE(Q3R} z2q9KffOH7-hYNdrC0L-@=BB&et136&B024ZV`bB`h-X-`Rzv0n6mBNLI`tZNvp4Dj zpcBxL(6Mm_XvyR593g?ad~sBhq|6(sQsn=UMfQw=mE+*asET3^Qse*?V;B^CqYR|Z8b#jp?nnsdOfw{7sde<@ z&=$zf($^j}a10`fU6Cb0?fSglEWuIOk3cJp0{tea3SxLrJ62zzmL)wU;NbN_;0%fM&iuTw3W^HIaEdcRlj z(?9Az8PD}+`gzl_FV6E`D(m;qApSjI^fdF0)w9LCPE0XkWwZ5YW~VpKdh4QqIK0p& z$*;w2Bov9CnPY4k_h+7R`~~XMNVUst4po} zrSz4|J>hLA21eku%Wui`PM-#dL1CtHlVphiE*%Vv=U{qkhpSGujNF^l9(oW059k?M z&WR1a6ZvH*lCI^1K^ikS5)F!RTe-Kmy&hM>)MeZxzYQ?+#Cx`QXcToI>$Jn`5Vswq zwRiiebrX6BE(4M6uU>lVwNlR%SUT!J3Jj>YdhMRAvH`C+whye9li%cZ1qVmSB@tA`kLPAr|GX|B*pIa_@JU zFVHtx&D$w72nw#|sTc7Y6pusihe9)4!n8!E&=-ih&8^=B|JxUilTxmo{pNNAf&Hu_ z^u)?Qj`~EfGJ`VOp~oJNPmn$r0y)2f@)zPMW;sRTs2F6AFLl#F0UtPffGhDIvX`-` zde5%kr>&i_``(?_{geG+Ug68`bnf*)%A1bpF8;FVk33gG4PLT*vuAc-3d9>wo)lQv zTIG-24}x>~y?a8uk2HcwBB&KqqS{!tB0uE4B1t7CZjE=)f@}EZ!&YFjYsU`8>pKfS zH78VZ+8M`+VEo#N?ZVI9d*yBXBZBpkPCDaTdQBxhQVy^3IYSKh8&!M!pSE-mC`^ZH zC=fC2!7*p7AVO6IFP`~q)0avvrASqbz{I?8k1z! zGe)wcdTA}ux|Vrtr;*5UkIzx%d8P{kArtQ_aMK$v>ZP~Ms!(D83|al&yDx4Bb>!6^ zkHndr!(>M|rj}y-q%E{w+9OGwuu%Fr_r2r6+vo+P*Buf}=?+N(QD=u;Ql>!;J=O?k z1EberkIfT+rg&dKy|ln};|#pH8TA%BquP}9()-Fw%KPFP*kxn;&$(sekuIXpt(737AZ5Q5vXZNUa_DAhtD*(6jWI@- z9DEHH$msELumKy2=`w7R9AmNudbMu4ChYmkT)xpv%|5&C5LxELW(pFD2d&UIQp`Gv zBte;;JcoN^I<9uMdgOAhLhu30^*Ei-X5#&i`350=ha`;c{K(OR1p4skx z$h_jyX_q-xYB*yf>ZRzn1 z`Al8#JmHtpL{Xp16-irgIkZlfi4^W!raCTaooJmhWG3da7 z_%L*x5Z5X+C;qZRF@Hk&NQ``ZMce}6y@*%=s7IGUS@2cwdlBfR!^;?X9uZq^_~qHX zdJg|CpL_r-i|)Jumh#`;%l_v|BQ}Jz4*?BV8I55ijieZk9*}$?mP&3@R`61}C%md6 zD`v%Wknu~SCuc%WUJ8BdHT3%7_9u^XiIXd%SjeA|_y zb6kxxPW4yHKP3w-B|11S-U)1Y2kmx}C?=62@l*`v65sMX2`LVG)uv5~ykL!DciW%J-mM7=C6&gcyxQGgj{-M79x*K0CSsSZc*eq71I{J zIq=qNhketOXgn5%p=0j4xZVfSqe8HT`-q^E&VTE&I0LG2%0h~!>LhvIae^${v&>efRtKqWiQ^I*^y{AN-f^f1)&cv$>0>~hiO zk+#dH+1`nDqhn#`OZ|7#*JuB8?TGHYd0AXdUvujeW3|6V(F5mS$U4SdCzpU4a2vhQ zzfGuBZuDBoVB*4N?#nu?h zZ`$ZmF;>s$B*1GsA>Vfkr1Ped)ec+6S{WbPabcwc2VLG@eNghU`|WXeGLJ&>M~&jB za%C9G)y{3G*RE!5hPU$G`u^VX}R=4R*jmqx(#^tfrPxm#%o$ii}H+RQ}p# z63=dp>%>MQ$7F$`p_ok+Sr0sM`ceRl2}1tAAt>1IvfQIeiEg4M$#M^U)!r_zdVZ5+ z(c}$5j&;I589ugfLQp+<*ntPS+ad3p``Mn24m+`P-ilK&ON5#JT3+#t>gd>+c$Pr} zGdY31GDd&VIj9qt`;VA8Ve4si;E*=C!R;FaOE}2w1%{? zBnM$Ji`QcX$*{js#|nVAbbP-{Jf~gQD%ZV(`Ag~IRdV&=@Fn97JVXgRyjc>F20lNm zsD#f{_PZo;yQggc35ay4qdV#I;t6eNP%_=_4C&0w@|d4mw?t&Qdm_hZi9R^eT|*W) zvF9Y!#I~%Vm;{Qf1n~}!y}`#6sSqyik*HhvJ(4a+Bflyu8c{JN~^s1mGu;wwe(P)A9CY z@yA!?2cz%H8dWEzUXt7u*70{ods^qDV(~3G z)}L*sSL-F&)0A!e_?egacjzSnb22&A1a%DCWLqS2E{4?ERq@h(NU-m|h?T*i+%xx2 zm|?_?`~8|^QuN9goYN*KtEL#RKowNX3KADquUJjmM5)UDh?0N<6ZeQN!}6vZ$QN6L zd&KKRD`uV(w(uW-16Rx2!Ceo<{41Cp%q9Q%yw->Y>EZ|oM*Fo!><&Dp*q}rWk{bFn z^BBxf11HbBIdmScT%zTb^Y;fX8duKGnp_pMR+6P$5PbbRm&H8FVRo9 zP7EDLfEzS)nFM}Gq!?L!MQfQK~+qbicYKg zT=v5~oos19jo^@=oSYP7xK(iaTsnCbJgqo({9{nL!e{zik|&-NW%)1lTuxE~@;JlM z8M75(rqhefXG6xKb%f1(Z^d|X1~jM5_P4@mX5d+97cPldwu$Hu>e#ZOR&YjcXi`*5(IxlDpzO?uK#aIST}Tz9VgM ziwjyHSRptJpH~%hPg2K!K-P&$KvSqIaxJ%*Y6)wVY3V;-o|^AdOs(Y>OfBax4=^EME5S>P6Czj8--hMU%H zURb>p*(dck%7Lm65#0Cf!suSP`ku_7$Fb9Eo-4W-ugMSc)YUXn&~8)W?~EG76|?H4 zIh-t@9L1{7Cq$P`l1y>G3v4VQytYl8PA4f}ks?X{@3z*Nr-;5H$rVh}~)ySYW5HB6xrt4W_@S-a=8?hI4Was-A3ppPUqgl#x(Hcdi ziWoI6B-O*XVWQ84l|~4rcP|(gN{5bSFs2`KIF5oM6exq-PF5gW`HbM@>umL$k)** z*1GM2d>s(hory_0DWcobN1j-)pkLsYx+O-IK|H<@^x-1cMD@EM!9zM^3+@ya`e1z% zbW<;l4bBoL%dbtZCF-?qr-Zm1PkKG+b>Oici)sr{B;#>&T$4P-zYf{iu}M4KD=&$_ z`LPZ-k<=-42XG?ont2gi+gJ^*ZiWDT2ZO)qzJU1x4A#R(ydrDg_YSQ5&d-h53rKD5 zCXMW(8cv)X2EZKTe!NaGI*PPVF?w-BJ>?bdjUcmFR^Y2fBrOb!_vwz@piJO((_3g@ z-9zBza+;VXfNf`#P6D(xo=D*)tSm>xh{O$B#q*xdrlapveq){c+nDG8BpQGgOZsxtk4HHUk2vK0#|r9MA$KQ zSqu5^t9&+vo$x$819ArHr40(b(sb#BUiW4WZYeF>*5Me%r_<)Jj-IuMaE0oW|CBL8 z&HZ8bd9scj)SNht2x20Gpq4=~+bEJcP@{)&Re%Lzrd5uE-5&^oA6LEOdF{ePpeEU- zOmJ-y_Pbz?>LYi^QN@JS47YelV9gR(8)S>$Fx!j6LkinZv%USh)SM=B9XmD~awj%Q zRut|EeR@4WkAz-3?kv#-XcRiSg@(Yr5~eyG9Xm61W`Qe8$IK_OGu3NCT0$}BOkEO$ zWYbxqvXCm-4low)Wq@j{!Lg)Kpl4^B5?UP$M3L1LdXoUiaIV9AisV?E1Zi|B7saVd zf;LB?CJp+Zdb!o~UC1PXj(R=HxcD4AbHJyBE_FkXR15#Qv@>eGB~RotgF8HI4bl2M z^IJT7;lq;e_L)~qu)&rSTOljFNX8{tEpV>g@>~wAzv}GS>n7I(sv$$<4VYMJ$r`V6 zT74!Q3(QBx888%ZJUf*3zq0dtkFQTPPEoU)Ggpu-c2krS8y*O$4Vt3%QA`m<3aA*I zYtRo^>V^g*KT@w#)xb%XqxN$nPu?vzazA<{~>qhVt zPxCzDOUZ8YI z9&35I%W~`-fF8TJ68`<#FVyC>I8N&f5u)SX_v+><_b8AcmNoz-aD*rz?3v2qYmiY=t{TSm+B*UWu4NDVv7;>4clYbKVZ zkzybh@f76ClNH=spk;7cni9Cuqid%5d4uY9^!&FPzSjND%GvvV_u=RNRyuq2yH{t& zf9YA*Z9mu?h}G(O^2cn7kXZro1+&$QZ`%&h z=uiJ7G51lkSq?ifw5)ivw|f-NDDeMu?w;9~#BDMxJ8o2M_lW15A+g_Th>qtpsE!)N z#qnCV5ac#yH^TKl4>WhE{TdEVY#(*@Eo)Bx=yUTJxs~8DPVBAn(h2b)%LF+RHPW%+ zC1hz!33lk0XW`GX`TMl+tRpY6ngE=4Tfy(0yx&{{oQ?T#;@z|rDf8~vbra%6yCGkq zRsInCEs3G(?XagtS8lureRi$#Rc;V_h!uDpJJfOcn1};ZPr?@#-oO= zrw;~gr|(Uw;24e}v_!5zDRD|*C%F7OygOugJyluByDr7#ERCW-R_U`9Bvv&{qpF9j zWWMydAKnQ$@meTI&WC^PZ%J*om6ibW^qSqF^JYu`PB$ZJK31Rk6Nz);_-MKb4%Sml zGDTKXF_<)$B+4|hPvZ(iqgdnF6rzrEEuH~VU6V2$9V<5*7FL68_N*gM*-eZWMqSD^ z7m{<@U9%NwxP5dJqvLDCAl_)uV{4LBO*}eDA4$wq?g`)O(eHw@PLWr}40Pl_q%%2M zUIuVY$8$D7Xt7ao96c)WA`ow(u>f=jX(bTRELN`YsgsonmvBx%lqsI0z5+fw1Gyf4 zPA6#srwyvVp&;wiC$VXg^j;(ss!Q2MpICwFsl^z|5i1O`16AFqT{p~2i=7r!R*HqM zdavO1$tr<_V9CrP$rjJmo{cKhU1{N;@^6(@_~25cM)p{Im+4{Rfe~c|uacYMTQI3n z)eXM8o!kpklVI-#~q%*3+3m7coIUQd4kx)M) zJ0b4y>foy{aq77#q65M>%KnR$r$aX^>^$wp?nl9(B=~qM zNiFG~)=ux3m_M<79F}bY>$C{#VAcXlYdak;x-V-Ho|aYxCGt*!h;m+3URbHy=TW<& zG>Ybs`=Jf0r%&R&`=O^*_Q%Sw5b|_#8bO3uNE4(sg&vK6*?CYpy?4813GPLzG3JP( zw+VtY4|QiiljPzsi@n(oj?JiG-}SJbux>Xjpkma|8#cciY1|zBpeS2Q@|}2dRBy65 zIz%z$6e-1u-~?VV$jfDkP%lM|iBQSY^qC5n3)(5&D8Z=U02x!OyinE-dKw8lbkxp+ zB6Kxo#_9BhQz?Erc{OPb-!T~p(xI|gqg)4B?J2;IwA-ys*(5nLZnv+xMA=C$i#sCX zId|Q1hBrCcjH0J@#O9;nz=#!!MqgN!`kOIE6sdF{smVM`CB#m=asdjrLFo;N6cbO8 z~mNc=Eb8Tv)Du>=9(?ums+`yclb&CGnw;oH=J=#L| zN9RO7eQKXi^^^Xfiox^VggxTCp{hg;4Svty@i`ZUhRpEae)Y(3lBk?G9&IIw>O3>g zD_7ns%kp0!_#*JKxD}jr+OU&8jVdf9z?{sQz)GKkK^2@1*IxGpg2z)2`(h2u36j7Y zcn)^nka(r>)+*anT22lpL$r&#$A9@uEfAR&c$f3ApzkxO{j8du(K9rtSa|Bw#XgeZ zlo&Yi0>?^;!EGt7b3ZAgugXsb^~$qE7 zgCqc%;0x|`@8s|Bg}h8VC1zR}kWs@#ilr{?a;m(;!H|tW&Jh@p z2!c35FjUmWnRYszPM6y|z1?nm+dEU~GI!d}^xkyZxwkDcqN0GHpn@7efyk-?f++i- zIO-@^R1{Yv;4+K|h=>dS_f4XbNH7Ny{?WGctDHS>z9;YZe(&-;oK%j^r-QDGu${#i zW6;I}kue@;vs2~r{jTQGT!)R)S%};kwBw-XQgNOt4uG@}_GF%?nw*z@@>0Cwl(1H| zv_-lUe}WzpbwIRqMAk{&<2>k64UI{5L&11Pex%1?v!OP75Gx;e%H7Y>H%~;H#epXs z3)(C?kO1oditK3Mzm8g@kzerNEQiJ<%m-*+T5dx>!)#ne-?bh^qu+0@qqBA@*ZUgR zMh}=Hv6RMv*UyzEtDPc($tS2CMC1@zNF5hN2`Y#R@h)W>W4_;_YE_+7^>Klzh%8&! z?-eE3A8|JLAY=v_NhF56Huny!c$y>{zm|~7m$ddGyS5$zBhAOiZTo&&_sp<>`cIN`q7SvJHak}6xgMSQ#8OT~tw|Mn}JeMYN)NJO&5jYZ!* zyfE3X%qw1zp%`{T0bJ3pP&x?YSvu7 zYDsj7k@&Gzww!s_a21M>)=IF8Z5LNZ#xqsW0=XTwH1cym?2yXrCDllyc31p> z9(FqFj2eX%P*$abOcT&F41^a6tMoJvF|$uYQ16!Ogdk>5^YXOOu>Oc;^D+vIrj#G8 zq0D=)9d<=!p_>{BLCy=i-Sn!|Q7Kcl|-&?))#?ANMXcD{LS$sZGcx#pK;OHPnU5$$9ta8+Sta0lHT zrjh50x7gA8ScQ4Z9;|%;4^Qp%egs3#4l)bev0u0%@vrWJ1V;&_?ny zjVqy8X#;Sn0kv>5a8>EmFp(cPIzZ=>POhR#!m*MDlS_CVRD~DlchbqD@>3F&5!1*U zcvU<#_PipkU!1#+ERnVIlfq$=Gf)dZ2)!Meq1Z!Ro0~$fbcvP1;!9c|)IlHfUM|%) z_rvUnnXAg66^cF@d!vA?W(81XC2(TAE{HKT0n9d4F6dERrg9Y9-HyuBJUZwC?#9{g zei6{DXTH(@1z6dGECKC&?VKu}!;-H3D97&;p3zXvN&VMD)Cvc7kb#_Pgz4EpFzX2_ zk%&A;b&l-7ZU}JcOB6hr~=5c=ES(uq|e!}=}V>h)jm$fpb z5g3Of|GJ(kc3@yMnE<1nU}^}e5*YKyF4?)4lDtYp$daR!&jaK=>}VwV(`hN!%Q=+&ht1E@=Q>#kF)l+`B=vEAXK2X~;c6P=zxv z50w^=P;32wU0-w}b~%bg-x;&VGeL~c5*K(SI&=Glmqj}r_^pwRP!yj1vo|y$TPNC;OH4ovhC?R zzbkwH>qhf(@E7Mkry5=u2f5wE7_|@#6b~N*p#|WuOP8n4Y9`ML?#eETp|={;dC_se z_O?6$ii_d7OX9MH>13MZQ>ZHL4BbWHRjn#Ka|QgQq4}p^Vyp2?q4D@0*>Y*I;y#$H zLDxDS9*LUY1#%9T>1?uF|LpCess+cy9drY?heXZaRu!l;z_SuWcr~s^ixlTWSg-h5 z!0IU@#O457(Cu54@-y=~8Vdy|%Tyahdwf3_{Er8`zkB|{um?RL;QSd5-QcZp=_c`b zkN0tDEjhqR24jDOIm=D*D3F|hs@zQhtlt}pzh^?Pn8oL?y_DSy;IJ$7|M~KYn-N9> zv{5U|pvoNBziBZsDF+D#{j+}N`-@Bo^RvBYXa8+q&BLX1#~P3hf56*{m@f~xWni}5Og z0p6K%BJw~`x+sb1lidz&=bv17n#?9!q$zYOb4`&A#TCHQN!Rgo{3hu#-y2>vKCP;J zaT+I%z71tuE2r&_NP)K74Ks^`tERQ8(xhq9?0LtjCTTi(np5WvWD~vzeQz$Vf)LTU zmmYYna838K)dXyFM6eon%NMgUMa1tDdUkjl@uOQDdXd`f!1$>!!B0NHK;9}FsmGC& zQ$LwCOS^@@T_#oU3+Rv^ia_F`yGw>#FNnJX)Ll|YQ-<{N+Q8UUi;9GO;p!65hsNCs zstYP1I_M1Nt9I&#u>o2v5BUVnutLk!o4@j^G^fC_z#Y{I{fb&nBe}~nU%boxq(?Wr z-U9BixmCPDV9%`}3&NWuCB9WIL#_qkSTVIX_>de$3fE6R@A5Ge<+k%n#a;YtvUowP zZOBkZdr1==mC>GQi({-t#c|8C2e$m~f6cqN*~D@j_b&s*gOO9&dV)zLsCZ0i9sBx4 z@gFa~c2V3-Cq?XV%i|7!WYqQ_7c#b*cdW*jwcxh;fi}8OdueT{d9S5~mc|si5tx2Z zZwK-;uYLZloOjNA>%yCLOR8K(?(KaQ6N4IgEl7(XmG@-4X=`7u$(qAPFL12KydajN zw0PwQ&pQoTy9@qjP?~H8hZGmrU1VM0prp*JbFN0&$?cxpG!q-lj)gH>pJRpDsRtCl z4Kf!bw4m{akq2xmtn#Vysi*ZpiQWJVMci=T10+A{Tf7s(B0;j>2JiS8J@b3!?*v^C zWd6SfEHC3;DO&q>*o=mW6&dGB(dyre%?$0hoTfLnHT9cksb7r03`%E4`W}yS+WH01 z4R^f+(;;yJF9FKB0px)9c~~~4`6ub3{wpAgAKGXoyYmL=P2m@wU7;WI4!d3B-JMk( zQsl8K;$xnMN>WaSlVS3ewbovk+%yS0tgN@=di2@DZ#K3gFmmnUivCoHO(nIBO;R|lktC_1pi?I5?o zD)lU@F>38Mz;w18eb4F-wc^4Ok8On12%W><{MH3(-E-E<1oX`#;AXo z#87*d@-mgnQKQZg%K0Hr71n*t(^0JHDA%ZApVF28feUZV_>#Tvc@gO16%S-WPMyzPrB}U5_q5=&?gt#6(r`P z$Obg#_u6y6ZVP80ZL}X7lY-7r8yq+WUut42atH?crZb4htH8u^gsKm027Uz{d6ww| zdX5yjkG>;=Tr-|*C!+-@Y1~b&a?$s#<%yF0?#K$mQl_sW@ISjkfz8mHPZ;(bddv1= zw4S2xc~vhH8-XJG+gDqtwGIrFLKC2*6U-KZN+lu(U7JG;D-}H}=Dq-pQYWdQKsQb2 zb8=x32s5S!>&vPWL^_Es%%DqbS8N!@cYFx3CpL_~7pqC)dz5_7^WMCOq)s<@2i<$g zgAhhPsw{$4EY@mwht>zRsG{N6NmZ3gFIn`}i{jhLdip-Ckr#?j^Ezbr`FBE7c%Qtq z&97Ui4eX;Th_Z;j@D|lQkh?kwMYt7^6xK=71oc5EfZoop(JR4T6qnCky{OnTP4Mxe z2?K3>FfunePw!6@Xsiqk(fH?YzGqJ7zzT;BTf+r4-Xne8$K(&#rjbIUB4{26dnN%q zK9koXXTRl5&91)zGk$_o8Flz}aj)<%SMg=5w zD3Jk7{EFwCTv!Vw24Q)E37Zh}cl*qJD?#zh=h*s!&iq$u{;`%kyFJglsPhCZa`-3N zPhBrM4Z4nCH)4XK7U`)vvdNgBHK1`aSs<|z8c*gMTVue&^O6_+=(XP&?Zw`oyGf}V z&soSMcIrs0aF1XH3FPbt0A}h?x3R9CCgBpqaex*GYU|8L~LG!9532AaI%jERbYKV&1)?Q zq~O@^c*JKN2gx5+xU_INhwW3e?DB+t``14(Z&b6;54MrB3|cSXYs*8rc9`+E6ICDvvXMFMaoR&eYiRXy z*K0uBlcV^Ai(KwqVQH`ehSpJhVeR}i^j2jyiFuf%Ip4VbFN5KouCa@Dyx#P!b8jtw zMf>XdpWppq4W~-J6h|DyqYnfb`csp_o5+p8!!aQ%s>7DYWwbS~ z!+MfOE=I)v)@Z_BEj-{&U1#U=IB>#Xm5Fg1BA7b_)lWoj3%?x>+J3xGJql*$N0{!l zs8XHx`1g1wMHtR@selqq)~QX3)BMo3%W zJ@`976|i6;Ld=XF89W_obDTQfHY>WB@`se{78)50#vY0u`rOo7o?8=khg58!i?~>gUZFBVtVvm0G-m&fG zWp$3qbwXLF_d}AY4T?R1>OHa^vfXWO@LInH9`bwuCte?IoFuP!7_tg*Tneon%~Gsx zus5Qu0q;}aS3F=Fj9PpOEAUR;`KOjFbNW^byr(!6`dDmw05&l`lTaCb&$7?3Y#Z8-9eJ)vzRiQ{y>2)?7Vqqcp0eVtdf?7s~RtV)_+jqQD5W4@##0iHc+A zG&^vdFwX>nTL>lbOJsvT(z)Amigb;g+`MRyZO?usCWlXdIOXAh&9+&f=MH&jYQ_Oz z0MWH?JXmjzj~B#dSU^*g^Aj@fI2T~AYlawbBanFyL}=Vj2q7=UZF1hutAqq4o}MII z;;3jE{HE(W6DjO<;5o%YVebk>eUO$+oz~CFF+8tNO(`y;1kbX3W_>ZyrNl}C(5Bq z0fROJ7JKN7G)hlz+sYXIGVGnW=bghD`_9_=$$r&9Y*(>xU|#0T&-uOcfK)-!$1M&{ z_UnLb2Bs!$Hki+T`Rs=K+0V1p6|Cl+sha$-v%yBcK_kQrk56U=u%a}e|wutkW$I9T`WiCo1BAyfZvS!TlX zp6n^}LJ6X^E*L=7lKFz#phEX_(GGC~=pr@n@HT^>Y+^Ct`DR^H6A7n+kqes zs%y}z!LnSn8VhubWGCEj&};pcOH%{4`n+54kJ*6o$9Ll2pK0`Bmw(_LN2O1rs!SZi zodi=rP`N}Ts=?smPfKP1Q?WXeT4yBU>ysV#+TwgKR9!34`KS*^bd$*8H|&%j*e2Qc zvR;NW&K($^B>920l63t~^>xoa8K!UTVDrcI?S0Azw0-85-M80a>z=+v{%x=My2oLC zR|{*L%^=~pgX|R^B8Q-Pq!ko_>t^lZ^vMb!TX$93L0^Bl%6*8wEW9GcG|gx70TAfJ zD{>b%@b>%P6t}7jSDY8#gR0B~Q4f{te>rqtzz}_Q@qO`Ss@#9u42|5DQv*WM4?;DLeme+z&&sH{( zXBR`<9IO-gk%QD{Zg7DuU;R_BVIpJk)Wd9efYn$yZdh|T@2K>7U;X3@wlUf{*fmrR zs=-FdPDA$){3}kO(e+h-A1Xi=6de*=$M}k8m?;OG)@-*Q&k2CK*TFLxhM4+BwAQZ4?AsKdNb_4>yRrr zUr@$xU-9X20p+n1Wh+=b{Tx<6nOgK;f4Tm=o05|rJ#DWY@1hvx%(VJW?vbH(%QV*Ao>;JAzzAy1E~ZI7@+h%D6iw!3lt z^S!#Qr>XVq3X&Xm8l`rdEaSEl40JMXB_dno$m@dZFRTO~cEZrhkZTLXYVU!hh>nC_ z{0p4Bl6HQ!GS?rQTe8U{@2W7wn0?)<31G_JRkruem(@;xlE-n8krpPf}Pz1N<5 zmaWRMFQsYC1*;r(J#0aO>VWKo4|=2};h^;KXbJ&#^s10O)cO!yg4rQXGsbYSLcti% zvGpm-C>SM;Gxe>YcUGHIP&zC!ERaeT1B=hy&|Tn-4vFIxYUEN)_A7zz)DHT?&2mig zy!)=*r-diO$O;&?58*xeMa;HX0b|NXWk0>(X`D`4-Y$EWN*!%FabVY?++;e*B^aOp z+YaK1Af1Wj0bO1o;gal!V2Q)jAFhP^xL77I?1U_G&C+DQZdmhXh|~xq!-?LAE+~-x z)VVOsj$BYI@$+~=^Jv3t*lo6J89(RVWR^;dSP|Y5{*Fp?;PTfZ6I^5v%vOR*GqfNY z7q?xqL}xYjt7A`egZL6GWWedrvn3!QSc*%QD2=aQ2IH0Xc09%n2b&H2=nt|ChtcMl z+$9~u=E8_#iuc~z??k@$^WQA}Z!yU@QG^4h^DM|8Vv)2)-a)5%Tny1DmkF9c1ps{# zJv9L?V^A;(>AWlbdnA}4t_nNMgUnxFXbOG7KPe&>Dvt65KsT4h$>3EiObYMhmW%L2 zo(Qkbp$iCPPskHh2BC=5AcfRE_sA=r{qqgl;`m|1PKRaZI4J9Eq*gX06mcQW3Z3JI z+9T7$;^*j~8vUqcs2!da@yA(4%QNs(=?ZF_16!V26U$RfFwkLIKty8t=E}eH(AVd; zNVk1|ct+No>+=(-vNr2f!Xx)z8Wg|O0A?;*+|xsP5eCn zYT==XLlK>>El|}~5Pn>sbvdcB!|;zWP%Mx!##3y6iXAfk^3`7rzH2lmCBDaGRM#{r z#)N3%4#7Yl=Pe?#m$wSYL=$=0xH)d&@okLGCyA@^OXF9$9Fi^b*af{;15gN!K`wP0 z-4&X^K^Z}fJcT|UG(_X4>Mx){W$@8+)5!{QEzC;5^Je%Rx}}&x9;Eg|A79~Yn0?Tr z#Y(qSN{#2Tl=}mZiEBKE=wT;RbjDwt9tKzDV ztFArJqS4^>fQPhA54;XQV+J1E9Nam{{l_dIYd3g~_Fz_!C360A^B>KV%@#SB=eo2jl>KG@(D(QdPjz2`c{ocaYP(9&{0y5`16;IE^x)1Kdq$l9e73ksmY3ZKf&xJs69kvp}0fV zFS$tX58Mc`lwQxn${gjH@W$|B?^+;PJ3?j7JQ;ogeCR!ao8}CJH@PBDZmSB_9~#4B zW^3eY{hFi$;g`7?qE;1BUnT(C_eN2%XXeaH3S5sbHGDvstXI6$R)}X|U!HbOMmsW^ zu^7iOcJ?KBOGbg3c3=Tr9*B>;ko|h+(B3dUT|yR8Hi9@A&tGt zrGVy`aoI?3HoF51cg$r}C%a>%w0nJsZP4mer7<&7gjeddy94!Z!_W`c!ef=ME% z1Qhwk%+d|WENKNG13zbwQrFCBlbjc;3um7Vw3Vf>_zF!=SU=FF_t|@ol|h*@uVdR+ z%&S@~6pCOV9jR&t0^(=mmuSfClA;zCp-5DlCot4$PcCbX^+Ny#+}`hf#Pit=fx|9y z7Ww?-ug`mlZ8S6Fdzmu)+oC$DYE_j8cZjw+x2iVyZ;~AG$`B8V2dE}tEron*J)XLN zR#hzhIbAAi@Jis+*?SSh3c)5`$OyREVUrbti63~@JHKl5O73$re@azPqt2O_@M8pX zgrE))k*lWNWn!3^pke>zyn%VK^m_lj;$AZL6_kzZ58tl58n%u)L0+5JL1%%q!Xfz) zrW084(!qViUcCY5dVV=XUUjl2sn(?n&fJ1Cy8>!>E2njZ>Aa77)p1r$d+3zxo*}&& zc2as>ndkR$@J4VI8aeHLczqUJk3xk5&Z*Ap{d2~fT^|3qnGE6>JL4NvcH`!-mdEvQ zSA==AmQCBLJDy17L!5ZWF5Gc=_JoV!57OjD2q<;eEb|u!vS0ev_J+z2YY4MBagL zY*@r+K2C?-{8ECbNpB-J5DBlm7+EF9U9=wXuh?VGRq3#cDhuqBcO}T}yj{GD3o4D6 zO+6tzEVHvU8U4F$osAj&UV9#8HK|UWSDz(N8J!`W;-xfdKf4f(12Zh2F`2uL5X>P0 z!Zwi=3ll{7f@06RAan!;ecXItGHw#aF&n_!Dk8DDv7LY2`-oQocbofV_fX1R#R@EhO{j4p{b@E0k+4Z(GjyD9s9Tb~B7qoG<7TA&R zaj(d7IR~jN&Rg6!csFn}U3L6aDpQo{+A#rQ5bLlvhQBuUsd2;ecv*7gOGY$(JO3BI zp|&`1xOkrlJ_-mXm!NWp$ju&IP~>ogycc>{c3AdVa5vOXEEBY<4w2O%9dsjk@5_VX zrE?p+>X2iex7M%9C5P$fX3p$_xpm0(H0K@^W%O}}#T~F^w;<0%0Wj_ne#VWgj1Mt; z`yc1{8{zW%oegWL?G6l=Iup2*5Dd@&6cUjguU=eo5$F|Srs-t2puk~=+XL>3X~>XU zAGAF9Hn&0?qikbZRdMc10acJ}sFBq2ddUnz8jW6d6&I$=`69N^oG*01?fIJb{K_HWSn) z5TBMdih4kp=&Ce@-udSCw@SZ~|C6(?0!pTr_~vl0LC3*uSsz{cy$7$X{o2R>S+>OH zP#ArT*aKwr=h*rvJAh=a;duRHNqc@FEfN+;vKjSdki4r5sti(hljlQLDJGG}@X?T4 z3!z8XPu9-`JtEHcj8;SB_1h^_lLK3gFHEe)C4xCmP-ltA9j~;6RSH{G$GkiF`4NZ6 zT3K=69r~)c(kmeZOMh>Mt)Xg_wPcPmlMCw8!u*I__hOFWI*+`kbhB>XOzh0rMLA%88VL9Xug*B$jTF^Y2fNb*93goe~J7?-{ zpZdNLVn>#Qe3v@GZY|}wkprs71d#0nbCH0Ud8B~_Ge?0G?x?Y(10wf$<`eFN(A%L< zMcB$Dh@yh?#8oakxfz^WbSAeO2k?M13$YL8OtNc8dpLo4sOnF0{{- zc_D*%F0Ao1eoYd+erR>b?Rn{Bj-Hz2SuunEFOW3gdQ4!>dWZXy=RJsn8m>`-ni=|; z3L9CFPP1HMjing#-LOAt-6*qv?nLikWq77sUU$;zA1eW|&97o%3f(3te`z%bqTwmf zzi<=m0X9O^`{|?Hy3-gv0LC1ZY;f7u15bSfpZzOpwS2JgYfW`&{lbs&mogw*Bn0w|If8-k@F^nF(31N(DEvav9(WVL z|4JTKs5@X|2tu`>F&^GQ*DTngP{%`V9x|T5V6|Cs+)IDxfMlCwQ$R0O8?;gpCCCY^ z4vF)6Y>lvC_n5O9^=#@SCl_kTQ{KB~s8235sAs7+D~>6(zZh)fXmlHCzHQtO*1s1L%rp$-~cXdQ{Ol3r<`L|6_ANF}uT z2!)CSYuqNqT8x=%1J718U-z8 zBV3R?f`N8~EF!YqrQHQw|Etg{hIH@g;M4=vJ91`2SzJHAiEG?MnZr(Tyb6lvro$Jq zJ5(n*!%3u;E53llU>)co*37)+@@$8HJnzQ(F|jz$jOSyYc@L{GaokB}g?qTloH5S= z3EfWac_BI-l|d^MrM`W%`jQxzq8*%i=p@@iU7HJpAmS9o5hj+-R-WeMgyn`MaPX#c zkS9jz%mYFBV!PVwiDBiLLq6^cSYgF+r|tCjx}z2w4NJ=pmuFKIqp>SG?(a{XHnD=A z5)A0PeuCYQ1>pvty9ClT>P~Jnby1wb+vTp2qx;+fbiK$A0-=!*9WdWKlo=91!%hc+ z3dpPw?ZOKSRtZpl4W0BVAH4>(mW-923Mme5cT*ph6@=r1HSz*Q5l~rR#O0hzBlJ&z z!XvYXN*Co&=h4c3iOYcNuA1y*Di&@Q?3z4*MjRnFFc{~E zj_WqdFc^hpfVlgUm0jlki^JA`TX3ul3v#pZhDh^iPUsU+=3fB)mw1esZ8?Zbv1W#>^3lL-7#0o4^ zh10%!z?}KvX^Kt=Dwn2)t=7e;ep zRJ*qF%;PIBh^?>?jVW4kh3$PNdYQq22AHEA;=}mf;ju81^9+Y3B=aY_!%H zbEpg_tpU==NyCaq+T@bkKRhk{jnP;H{nv~cRHp-n-jTUr>VKhwe4xk%k7tM*HwYk6_J{#kVpeQkn9$Te<_ zi-w)jf&4C=1Oi33Pr#d~0RQ$hr3=UtofkGjvHu!iwo@Mn!cVw1S4YOU4}^obJx2=_ z`WPzC^;ZuBB=PcII<8dfWM%w2vc_=rC)_KZIf^Xh8aK6;T7a;?^x{It2EbuJ=mqcGq*+o|K6Et>M)=A8tlmoPd|}r0pP9H+H3S0-{R$#-gIkkc?-+9WpyI$u7u^?Cx|ez` z=PiSFBRyoBAxWi^AP#H!jgZW3l2r@4g;-?_77w$#g;WP+JYcvsgM(+Ylx1Gc(tPp$ zz*NZo>!9##*eM!XWex%Z&%|}>*v_0IXVI~@CpLl3d1s5-&xoLc;UAV!8SD_`z+vcW z69nxd7%0)oBO-Shyr=`NR~q+-LP8u_Fy?ASnK)G@TvK~YA3?(@z-<7wz8c(Pc2>R zh1V@zuGr08MOX^fO=t+%0Knq(F`?gJ^qX0&=%@TDB=+)rO3rK&w zr;+aq7@#kRKj$BkAxZ5mpd~*6mPjMV>vP3unXn@ppO-6cQEm9<)g`u(AhQYL%%M0s zh*%xw4PJSI(|%8iTkTl2I8msu8h;iSe>p6m_#ex0 zX@eH5^kC?T+okMu-3=M6y)$=ymCDIj`ev=k)839R1pPIP%ab3hr1f#%t2wXwqL3Ew~^y9NGKT7C2&aM6rU6 z`6M%9f1)rkzyIqi{`=27BW&iR{_7!Xg#(BGfIxag$>#=wSx-=j$Rtzcg2gt5##;3q zdZllNXLJB$CTCU)k<-QIqiOyYP4-y~waI<<-eKk0PW|!CKSzAc2#h0#zUg@62>SRCNywBzwrmNtQfA`!9K@KRaa??QAL%m^U zkr3Oa6M3mY*AzzuT@XdPDby?iMj+p8=$K(Q_AJJ3#Fuhho2-VL)z~?1Nx}P@<|lg? zLDly1=?|#nX;iTZsImwKdeSsRWF9O_k>2~Zbc267xht?HXs>vSNBmrH4l&M|qfl3f zZvbpo3T}YbEI5u(wp$%i8>CUz*qb4qT}`(sI9QE3>x)^z#BtXRGwx@9*Sv+2O`hCw z|1uCj7&(PKZn)gwohicHkCv=ipx&=63|r&YB*a$pHE!$-7;DHpeS@_&@aa44d*mCj z34HUexaX>>+(HXqJ3p6G930KtF!QQ(RZwh1E{&YPwfqdu3E%i2og_Xeo2-W;`0oHG zj#sBF78*TI*3bZ3JVyp2hp0mZ{syNeiFuaaqGYv`W28{(Stb3r1 zuCk1akhuRU`=t>VziZp!MO|UHN_OCxWtmBBL?6Kb?`Ago^@z8QK~{;niia5*^63uRug`4aO?7b3c)&F=9`1C^tw663#3GBqyB zx|IEr{ekV^?WdEC0d0~F%scQ8`W6RkwM%29n*%d0uffw03O)O6)!Dt9- zGe%pgLR#bK1uCQ!ClX!O}r29Ms55 zo%cZ%)pl-|?6kg>+lJ1wZH!{st+Ev%>JnKyzfV>exJTR< zt~MOQq&ZekB#TZ#6)Gw=vqYP15jEE2#SU#>kNEM2=0vY~CPT4>U{cX}j;xBtXaMtcE9t;tJ%KSk@wnjo7IHi zxNDQwzWb+kSEIqH_}ZH%s5K7k&44b`2otiIU^Wo|kI1wDEvW3^bb>)>BV(&q`-=VwqQ&)Lem*#t445KJvWRS}V^iPc0l zcRMGJh$Fg$$3?Lbdwi>AWQkWfKlJaQS5A9Kec^RW*d*;z9+|aj8rZ&B*(BbB)OmQ% z%4wYonuLwvH-qpjp4m6!(V2(T_Q2f{$=tPm7nEzg?tdxWAGh&Yw_I0En>Ya38pN@t zk`d=7Y%47r#Dni-uX^4Exlepc<5wgES&LVY&9?)1es{}qlqu7Po$z=Hy@^{Drmy(E zN@mT<7c4W-`{H#)9x;@gYpVM3r7?}u7(pJh;tvTT}~`1<~1=Iyr*OBP|F zp%zJDYZgFa3q=SFJBF#}$-fh~0am>EN0+cc#-n>C<0m_0gd|;|-5*Uvbz4tU>mArl zfr^$9ZpwCo$snL5H!_;iaxX#=$9p!yb_1WNJK z!iP?!#{HeL1dr2`;B2!&i7h9MXCaLhN{Cy3_^)B}j5Hg6$8p_sONsSFNFgjEuF%m0 zwvRM^|HKbVUs?P5@gMDe``R~7edFRgyWTwUTJ1aS{0-c`@a}*i66?sjK@4S?;GF%5 z)5ZxZ;#r`W==savdp*>A&BF#J4jg2$utEae2+!DsP-p=(i+Q3WR3G=c>&b=Cbem4D z49JzINnlgEu#Y_W8;oP8`ZK~)C9 z0HH`45s8vDZO~Q?RW0e{QF)mc5 zvXbZ}uj|#)k149X^TM(Povx@ydy?r~Fc}781d^|ucjNUaKh^1z^O0Ao0HYCnN%DYCKjme^o2~Q-h~w?`(DQCwMm{w3GYg!*YRVXMI3- zXtEzfKiVW+vP9=ez>GBrF|`x8XCoMl@tBX7ZkY4D%ZXzJk$$amLYNv<3PRA6ob-S4 zdaWDfC)ZEb7CU$2Ve{<*xiQ3{I`Z>MDwAD^!+{qkH6|gB-2?;6-+Ur6HK<@Vcn?a9 zl`l2fX;D=R?@*np3$JVcCFS4aURm*K#T#qhdhptX|Lp(nNmV2HV1CY*6Th4jnD*rZ zl0)*eFXw@h01|0n*;%f??G1}{2z%O+m>sb>@P7So*MBH9VkUO;rC(9;>@efN!F8Z= z9}#TbN-${zwGqXQu-_Ghxf-FcK0WXN=NumPMZvr|(lLB{} zzj~7gvd4$X?RML_##A2oz7ZcsmV|tlI`P8b19~(@;G>;jE)vu^BJ%9wvx^J4g}~|8 z&PUpGtarr)@eSS$o;r<_1I5ybJgjpC5qnUYKyB}bGQEm4UZIgclKTT)v)YAW@dgS zzB?*Vm$d!vXXf?yPj?qO@G{3jl|68YV)PwHNncQ34nB@S!FbpyhS%i0OR!4-kTMiP z0?#D$*Z?0GctucpKS*UMt3%Ss67YpfJo|k*jP;VxaR^UIlcq_pcofXG>AK0b@ECQZ z$92%M7xeQj?|oq&b%W}S5nt=a;^Kzd9)06FzFbsP?WH<6HhqCs{&iNqXDgp(yUQ%U z=h3Ra9F|YN^hd-D^ZwHpB%ETQGqu#SiN69UIiNHNs@U{42ZL^TBJ2(8Chee=HWuit zgg4G5V?D;^=Tt;yC%{(hJM5I~7thp5PA*ihRCEIqySh%&$Bk#Oa3n*aC9}!$msWFHI8DwdUt{Pq zHKYc!l=}kCF1EcLWC4S*-3l86ft{7}>i+Ea&%5?weqd1;va730r~)#6c2YPn^mowM z0JvIZ+&JKEh`n^s_aO3y-Nx8qRqlBfLiYQ}{_rMAr_0ue90gEW+a*G86R_AADEK&5 zV43pyd&?J@gT;a#X0|fh6UpZdI=U%zePFx1&;_aVq2C)ft9bXt(LiGli4lmCV_4!c zT^x`Ir5ab=8+d7aqp0pd`vpNKIwp1)4*et9jfum$Dc!jr|KIZ-!QJkroueLfZH6Ar zaCAF{ouJ?dTF~8?SbMhsJtw6Fsr*uhtIdmECjIft% z=NBwVy>hu5Af|>f_6nHHdR%&3 z8p&wCZ9cJ!Du}GGRC$5x4)QG7;Pq&)Rh2Kvmo$>Af{t(w`aU|kgT!ke?b*Z5Sc#cQ z^;z{;X0wJRiaioVzcDU$o>xt)?%-|AeXkFCypK!&5HroNw_+MPL)o62{c6db3qO~w z;3ac!%-_XX&e;v!5(h*bbX`Oa_Zp?;)Xdb#V}cF@;_a=f6b1e~=v(EJLT~e{SlBav zBQ(jU6WEBHz^U^+u3X`I%3CMHwN$HWfO;s)Qr5Y5&?kf?(=~D=x%fhoF%wm&K-69w z1xz?+gLZ?4^v2n(s_k$!RJMG@a5f=iH`G@qa889GZ%(!+J_WBZoQL0OTHoR--Z^1b z*b3A3Qkwv8Y*<=-cH}(BY6>HYuYB!>zY%!9-`TL1+RiRv;J``9I+Hbd3Bl|ls6wdT zS9U_9SgT4WtK$vvQQM`NbCOJxd@5Kj9S%+8ZE?=?Gn}ghEm#9rN`H8D$ac3@6<*cq z)gPV?mH+7^3JW8l*lKFCB2(GGv%Pt<5ynQs!zSDAt82ew9>`^5?;JP{X(1ln;@YISaL4L7|3y-V&#)eed{lC=a?Ik=PJK!!CV~k?&KFPi1Fy< z7y7F6#ZA&Y?r`Y4$l(T<;-<^<1phP_05Qv8$^QO1Z@$rxymETuC)8>OHY7Vt3`rWn zY$T{;BC>|cm6yr3FSHNWhVE=j4F3r zQcrF*E=$ZzTJG2~X6l@mykTCX@33JxP%j;kPfjP8Ed-THMAq_)JzG_Kgr9M%g4+2z zxiQl&h;MrJExHr2V%m`F9{)YUweb>$od zWR6Ja4!fpk=Ab8wmS~mY6Ru8TP0|&`0EO$QRmUFi;8u=&O4i3cpr9-aqOjPhKAUY``bNr$NC#CKDen}Jj|1_^d zmMT9fO>;gsw+dE6JKZ+BMF~EiU+JTfui;eA-yc|Lw|4ajqk`2~TYs@-v$O1nr*@kQ zb2#iIWI=?Z3&^OkYFeE_ALn++(R#!yHZQ7yS_h-uP}gzLaOfM8@swU$w^>@ixh*>; zt|gC&n?f{lJYgpvcfz=^Q>B|_T#PauO=;ww+F`y7enEn)78c0*s;6d9L%^c93Q*k6 z2B_(v+Z4H@bI2;Lo$zAeo6Q177M@_A!_o-1&LHD5xVU>aFyBq14wy{vYLgs?#-Q@RUJ@IGpdmSNhHa5+1>CF! zw-t}F=_Nn(`r%8@yXMMym!hy}0Zt$p)7APU9j;L-}2>^%}sP}gN`1YO_ocW z0SES2s+bRrwWB`LcEPs!V{E?9%A`&CsPyk?uNmi@B^CTGs?LE!uh1DdVzt{$Fvke$ z2oX8(vd-tacZUp1n0AtVRHg_g#F%MYVEM99)WBUat&Wo*Iw^f9YvATbXyl!)y-@VK z)=x{u@vx%0Rdpk{iN8Ohf!j`&xE~C}rlcKY60gKJfrHISFp+Y4$n(N`p|Ny|qQtjP zc0074Ulu$-XMxhn`sur&tRL&N(?N~7Nm>idhaFH-JRUPNPE46W>>0b`KpZPf5g%sP zmYA2+vYCw>INp+Og1{t#Ng${-L?reCMhg&UWe7&niVmJCn)ZaG6SYVd{|zOfaD#n9T&W3DkSvz-lCvZNqw_1Wu1<9=FP+ ziLd1h%&X_8lUkR0s*ZPCR?dy~tK-}Vog}bW2au5_WO4w~*w_@r7y;sXZ{RGoaT>M9 z3y^ToGEe^d%ZDwam9C&)E zFadl%!9cP(n}}=>f4ZOwT1v}l?VN7u5HM%t36lMQ(+wJxlsSZMiSD%vOVoHxn`Dh! zo@k|Dn-468P>B)`4sh_MHc4tgf4JR{KOxZT@oyjGpN{^Pn|T|u!$xK;G#2mXv@u8l zsm5wWV`gSMw_8{eZcu5Oq^ZL=KrBYuK0|8ZsqD<#?|(yWs4fRy_{5k9 z+uk9VeuBD%@{xUX<$PTBBH_qm=~^|YriTJ$Jv1%0a6sD+*i$req)NHLTOZOUX_D%% zK0#*iTI3pK9u(_82)!Mep}O2AJ(ROnCXppMS6Ft%l-1qXnmo;*uav}c_4qZD3nF5jAx`sL&5v>CWPGvOFxAg21bqom0SHWDl4^gM_TR*)`k`tC2Rv)+z z2piP`nR3uipIE=|=Ef{q_03P_YgSKCOv6lu9r#ajz#L z_2t=6u?uy#r$Tz1??BfNCX*6)2^=(N8hJXbz3t}cj5hnQ!^UV&u+K4e*!bDaKh~LN z5iEq?p+9R5ez%H)uj&_B$QK$dNDcfaN%fO~WcTtNwn2X9*_L0Kiz{1FC58kq*ip3H z?eaVk)(WK1NGJ_dTBvj1Esd3;PjbVnf8G$SD4?ycxHJRpNI8zx=B-F^$ zJXQ&iL@k5Y$z3KuX8=8!Y*_XuLWos8#Zuj5`_U4OlBh-C`W^+uBI0Us zcNk0|<^zVHSp=iQ>g|EivoS>g9y=tjaK5tVs%=6Dwi1owxr`Npr`&hmylW!F{f?Wk zx1>zBKM?wk$YqLF)nVXkhysneR#iI{_0rkJ=>f;R4n?$6wd4@@P{b|rk}#EXI`D-1 z9;(JOiWkdkvVRfHctJHDps=}TVZo2DJ@3VL?fewr6C0FV@&p1(I!#i;>!6$XgRXgg z)xwV#9g;&PL0-knoz>=lk$*gBCwH^rq0=Twc0dPJpcqJBw-?C#t+-VoU#Gv9Afom2tU0$mldcjjK- zRO-={e=ME*FR)Qh$*ArRVr?=i-6F04TgzF#l{j?c`d71h*-E!&UY~` znLkpW`fq9#yL^M=I$~6g$pU*5!K4t>I;bvm?W3`v0Yf&p%B!1ICBH&qNTq;87G30+ zvZtweY8cr7>8YFTc;GvVxSr>|_h^S(JulB6)@f7$FkIstl*JSU4$%eNR#mQhPFRP0 zJ2#Ko;Z_;Y4U4k2kmBIU8G1HjZZsHn9O);IW@kKJOWy1}-)KDE`TK%GYWFnixXHJ! zj$pvf?264k~RinBy2K6B9q`&XYGj?k#ph<**T$h5vc!BlEbN!}=^10*2cI>wO?W zggRrf?gR81mw|8~?uX2keb>RGLdROTj(W2l53+*mlw%7H9Wj?Hv!L^}UEIOxhL($Z zsx}C-bOIe|i?2QFS7FVIXZ=tc+`#I^JFYvo?4?UbpZ9w2(@-gwAX@H{skkIY^AJnd z&(g@Rgbs&dJ4<~~h62yE`9}v}ZC5%O8&NHckC+TJTLa#wCNdj<^I4Bz1zzHoQ{$k( z=#P9*GW}0f+_X`a$&QP03TS&q)PQdxm{g4NMIu)=E=}~pb!`%4R#)#=)(2uz5RwIk zIvJ=Eq>6*FabPk4V>R#TNYp%Qv~6ZDL5b}^DaJIYM;;B*V^mEy@ zdCeXT;>iJu_3(Hyhu8rPPd$ zYm@ZM$4&+8aI94JQrYC*nLV;@GGBpP-O|4BX7_Am69~~^;ut?qN2W|q2&oC$#9R-p z@x*6TM1WMC@Cup2t&%5&_t8IcPx3JClUO9#N0#UG;&)u`{u76+}zfGs}F`e#_iz$pI9l!saQ$>x!?7o;=MLfmbW z4)zyE3HC?ycveHSqJr!a6>vAs*2ven-4|nZ{Bi0Ql&E1C zeh*d4PobA;zUOEgjofSUD~> zG2~_s8FK_#eYYLA)@QbxOX0jnA(tvu9aZ4Ki>AXSxUVD_Xh|z0A`QJD;C#?PgiWHh z5N4mYda89~(VXZ29mqM9&cYD08f!EkRdrS^E)2^P9iiY~7iiT?q4&s^OR5XP z4SE7c2EjHGOfo?w5s}eQFZL016bGlyXjfbiWQv-kjbxp?fs5P*JKQh@IJPCQ=C@~Y z4VwXA@z@Dp%4z@{*RcFxaoPt`BQg$u^II3Fb?gdL9QQAyK+PG!SdWXx zw@k{$>bRXBDvZWM_Gk50YUc}MJU%lq9yJ70Nl+C;WINvw=Y7O<+D#`#w5oc^bxD-Fbw&nL&aIgr6|RL^ePlt^$g5m4MW+LY9$)itRujL0yPd+yIAl((mFV~fxILal z9?O8hd+(y2b+$4$dilwV&w5kq3pLFKT4!VumBeKmnTKCdfjka}TMe5NVF|5+Vsv@V%` zSsSP#2VV0WHL(}f1Ow#Bdx^*l#d%@8qD6i}2uj`1%2wu;J{{IQZrJ;16nKXw)Z4N? z8meoApKzOGB~T!_b5R>}PS3BM!71Q0@v&$!fs-fbaV`l5ZsqWu)CrIUU+Y&Y?2X8F ziweF^C(ld@Zz9hwwh_u$0q7Za!f@Iq_!%QwJ6fdfYdhZit*;S4511pdl*WMp1oO)X zZ?%YEK=NS+5t%Efa>)vd<0Z~X5Mi195FHB>POT(^v&OADd)&n~_%zVXfOfL7J{S1yV{ZYcIqJMh|x zykoYX#pZfe7RYhKSf9T6{zmh<7&fsN$F;4N$}n**)yPkJG)W%N(QZft(Je)-#LKi! zc1V^r>tNXDp$DCH5>#WhE3!wTjV;PY-fPco&);``&7AL_4LlsUV#b1O0ao2rx#;+3 zJhI7xC2P4@)uNH(QRHY!CvVI^>K@?oILpmaw$ZubVW(BEH+)O;)!sKV->Lb^%~#Zi zF}OUywx4#1DVzxmT_w@phJd8-wm)_5eMA2=R-0S(E@#ofEHxEd#O*z zFT_xcJLFnRKIBzHm>;i0&s1GaCUB6OvQnAsw-Qc0_FwXOMYs=1XNcEMn_QXn0(IjUBh`&w{pR=NVv_}(BoRoPvjjW zb&_aaF_3!qKtA9=5YRS3R^U-2{?*T|0ru5wS{teiIT%n?qmFG|0)sD{YY z85ca)aB3wv&Rxnw5&gnm;aWdPZ_0IaBXv@>bf+xO&mc)r5eDSVf)q}b%dNnSz-+;G zi0|R;8aY1T7*#BK^jMAja&T_g4swn2It%RqhJl|%YME7}n9B$;#uILPnoWo5)hhfh~rG zrIHS06ccA)ivdsT6W{1%MVWXo>(~n;J$Kof=22rb9cI2>U;(Cbnv?p#o+Ix1QjP^BEUf5XvuVPYdRSbWT{zWvv)n8)UxrlxdYvtc2S zh+G@k2BMe#=z#XJCYic}uJU;h3d^EI-f4~+bsseHYVbmVv)?yIF&T5wxKLnm>}+?! z@+36`D-m`v>%3M-&5YZd_Pg{dwI}RUF5Nv^DwA5xXNWfd3I)*Y#Y-iL1|}o zJ9$&MhK~0r5&-2^A9uYMQaWanQ6Ltx-5ql*b{mOtz5Wp<+jcM9^huUq|7hiB4jb|6 zh`inXk5$f(_9f1xil-whL^>aIIIxGkMcVM{hx-5J3pOvZ8=6fq1bZyu{nY*TI#8JW z;V@~8s0_U2cZ7;|;CjgnlNiZ5f>}#YaRwDtEqUMdn0F>uoe-iU?@(J5NuqyhiaY@g zJ1aJvvixhW7(p@r!@svvRihDaII!c#qW8NP)SNO2NG`Te82u$&hnl&>3RQo-sgRv z=lTD6-Lf=#k?=A>MbJAEh~xNgWDIhkNs3Ap@^WPGmW4xFRRC0{qWQo*)o9_Rd1t&0 zd#YO8d)-Plh@*t1qBhS0SmMMp18$hMg8!nvbMmyO!P-lw2$1I*XtB&O-`TYMWsN`GBH=T^=$p7>g zeiBXfVLD^v699|1!>!<+I%xo_)r_A4B@?%V+*i95krXr8I3_6YU z&r>H$|BR0jw(8ZI&e0YI$+nld{WcwTi94&S%voTZM;2VzaD*&(;CTcDx`&m1rBLh! zimam|>KR6G{N=B0hsibv2FEcIa2%jmh`;QiBG-QT&io`rDZP_U1A57dd3zxj z?K(v1*ZJNLTkYB@-y*G;cP$h+0wEN$d`_e0n&)QEH06Nu60b^rU3Q)vQtQqw{Dei#ZOEeCB=EO9*8l$FSOXCQ4W|6{9_;fLi5{s=)qogTuyx699%iXYWJH257*! z^1sbveh!;oVj;eVJL}pyS+63V!F@9{2$%y4NVD87HG9ABNK;)4 zy)i34>^NOaw-TIdPAmI}PFYOfp4~P35!nGW{ddCF`7W$hE-VyPNz470xHV2mnN+Gu zf&7VqPKp;Vc^#BNhPB5^ML?4VN`+s~H3CfyRZrT+bFp{j98RY0wc@}5S4Hq#m9 z^Jg{z-8WXc?6z_leMGHO=K!xwS9qs-gLw7iST8)gMwbRY^v?Gx^gm4NQpHPF9px2;T@T4tUJ87|ZxvK{ zts!;FB;HwZx$$XCP(lx;9oAO$LAPGTes9b=%182pe4zWfJf(&l<=X`Ob}$>plAN(w z4^ER5RkHHSB_A3g6t^rsf^<7DgjT#Vd?)4!#ST#9Ay#u31lX}~L%+r;6jjU9#f73S z5+zIw%k_hpAVvgnA+y!}snZJgDs?vR2I){#@^dBH9w=WghWKJ1odcXaI^R_xklVCO zkqTwY?Sh*Eq`qD1kwYi&waEQk>3#&NlS_l@WKn^65}mq?)Ol!Q0}4Zt#DBo;Do9$@ zf6|l>p(RCxr27jZWRE#f0#iN zpbb!5N*{1*lrIU23f${A^rbO!c`X!p{Oy6i&(CDsuW-WOgiF7Y%{0&5|3a;=EM(|^ zD$Nej<}$}=EDpL3Vny}H`vgnz^PxA%_$Z4!Gb7eR<=Hdiaby`PKaj8c+i#3!>BG98 z&LE#SuvuDeVwN6L?57me2mzrsW=(=dw`)4TIHDMGoEBoXY`$MRozAaf7A7fj)kqK1 z9*X-YLv0MirWlia>eRp-;U(~~AISFW4Nd|Uy2^kfymsmC=}_TnxK!`5*j=aWR_&E* zS3==X1uWnnl0s2o7`A)%(K$4x1j19-J+})hRrp0xq;0cx${S=rfe&bR&x>Qq^py5{ zV1q}0Da}94AM-5YnQWjIN}kp)#VRAiH)^FHdtiw^bW2u+4E?{pBvy&Io?s^%;rI3J zKl=sQl9{Aj?gIn!^%)p15)fAfeOM5Mx71 zqD=LTSx`&@s|a`s<|x<$&?#~65JmTTd8=S?W)r8$+^Se_nJiC4=6%}Y+_GgFX{BGQ zsrgWBgiD5CeG7?qV7TO)z(q%~sTA2rMQ&H^42>2Zd`&wfvv?iU(saNUv5R*~)DG+` z;3?(MNbG|eL%ZZ>Ub71z%m;5Y05R{w9_O3@GGXz%XJ!9fG(ZN1ZoU1flQl`x#g!_& z^d$H}u-+gUCJfkm%z>)d9Jm9)1*}QZB1=%^6wps`Nu!?#3P989gT-VE9nbHiVHqd~ zaU=te&=Z(B@xG0GsnOufvPsh4u?g&EF01+Xe=g+I$xa7GlL5D0!SXq=FzVg35i9L< z0h2i|i>`(B+F?PjV89JRQpoqAQxEXa6-f_)xl529@(=`ufkMdem`+(I$e|DV54h>XYu?cFhlWqI%z3FY7l-lP7UnQv{Wd=L&CNs#=4D)NYQ!FH6WKofK zJ!=JvsH>6&&7rSfm7LT3{ko!2b6Aij*&Y6bpU6j2??%mCQcC0bQC>>;ZvLW4jhfTS z2XvkHIoTFE82mXSJyuMa%~sF(ByP8sG9M6g!99l! zgIbW#!|LcfVLo*kggOgFmD8F2!P?2@4fZ#ms1OApcRf%@|#MKwu4;Y<9y(5XsTFiUd3e3E>0-0NKQ}`Me(ZLI*?iUExRGqYFh=`I~XRimj!{X(|%y9rt;m&;*1Sz;WF! z+`@K2w#5;sbV7O$tVv3vQ=})L_IhDr*duyjsc)w{W?E&@u>pDFGYeq*W^o|!Ou>F2 z{tVLL*Ag#I@Hp$Y%p(h2o7?{I>*G`+12AQ*V)YxW-Lhpel$Fuf?_Ze{Cu~&a(3PqZzBWacB8w%PRdKG?{_d>*t|PS^ z#d{s88=GI_^s9B;HAfBqldb0Pa9E$%0&dGeE2~kn+INlnhUwL+ZS)~ktT!IrbL*k| zNE#jMbw!e-K;e5k#Rf(K7gyYl)Q3H;al$QiZTjDSG|f2M^!-S-glusptd2I2D^~$F2 zLgn)Y23rj})rbnh2c5QfWzk)dDB%^!Uh;&WVtc6}u47}tm9~-F` z^vuh+pLgW#tCA>CHEz_bA+3-kfzs)B<<)c@T^UfS!p&Yxgvp$@o~LuitA*jaRgIdh z#LybsHU|~T@v8l%&M`uJI2vtwY#G{_xgY$_JUQJ`4mgB)amzz5r&rC}M1!S-E zy@tVHjPyao1{+`yxoFy&@ks(7%`3O*yS31nNxWhIPR(fT9}OyO;8L`OIGp+Kw=nGwFaHyG&>}zBExa5XeGOzcUE01yXcuVWh*SH zi>BQb*u)}@=9ge;ZXLRA0o(_i%@-e@7?ZkDy=`IgspTE+eY1Ev9e_I9DK_o5J_tsZdfA#UZPrse;wFXU2$g1va*63FUy~<}AZOnp}+?6EHfo)8+ z$v_{ZSYUW8f~u&mUIx37Gxff7IcXKF_sbBM!&FhhoRgK)_j%hxF9kNSxx6~kMGwr4 zpL<2PTv|*mCq*+m+!{58zbVWnk6J-C@1Xhu)2N9Hs9}yxt(9So)dl8^YL7mMco8W?h$w9eM5EBHo$%}D-cHCZxJijCtg0!XF_=+;y0T8bo4k;$qR({)NzSBe%c z^M+Uxrs-i$aF!a!zWLw*IG$6O#!C!$-C7~ZJaOSVd$@T=)POfGYblm%dm1Cqk9 zR>8uQNoU|@JW~|DSyjp_m&eXn*Z`Zqm861S!#~PL2j+%+t7Lol_VA85slI985$}q) zD#1r`J@$aNY`9wv5?dFTg3u~nFYho^J(UHdO+!H>O!}*JsivD0Sj}`xQNi}VaS8Ya zYT1o!G}Q~PO{<_Cc$@Gl-p!)(c#9%>dFAh2CTY_;rZ#G_)Ym=tLKf2SHW1;3d)wTr z<&ehA`^RNi{--bwTN?GBleV{f!-&BTJ3s6or<}=66UX2(#a^Vyc`6c$)Syrn^QSXm z%jEFnI;cVuluOH@CPb$!3A*du9SoUTn>~*bY!mAh=dDHMUo3o0qqht1%g;-7N?;tq zt&p`rt-(ksf|-`sTdAs91UXtdbt{eSaa9aSEL+_R7^p-Ey);k%P@(8Db6kw+c_mXo z_~@$S?rf07QttCw6%IruWdSHpk9v?pT5Hyb$Xa7HhWn8=JAQ%Tv1~O)m|Glo|Kne; zd@b!&BfR|nr(_}7=uGyRz-t@DW>K(B0wrtuy^xdPt%66qb!yb}#DP9JZM*8I7_<E6fz!ju_{mR1#f@E+hc~xh`o;e;!6URnW+y| z9!98qyJGVNvetnY>;)zm-b%4RdB2&8+~!^!QRH{_joSGiZTi*~Nv9nB-MqN!w0i5duuJnEtG%(7!RW5~%TONufG5q$McIJ5{klv0nz3vR#V3*>t;k^dK8) zVtpQLxZV5XC(}24(Tj9DU?W*y@036PAvr86oVnRI&+l?Th4NN#j_ctWI^Irsy~n=r zJF>Wd#lro()Nnf+6ypetV#v9T-6&ggvn#(&puegz+Lak;|2L6jJMipt)WlNkqu4@< z?4}~m2O;Uh8c`{T7GzUUybYP$xsp5~hLrH1s(^fYpBJ{u6E*<-jQfGWR&ctag-NbC z(mGi_g*7laGxVhR`u#pEvoLg1;<{IU*OoyB7}QY!XCxzm*G37Ekssp(oC%%0`d;%~ z8;9NPvycRIZdy0JHfZt8PDu5R3D_w@A*O7fZl65yNw@=xx+J}*N{rhL0j}siu$zl* zG9aVDhe-!^I^l+o*sV?Z&PKzcY5%R3EOKDOl4fF95-B#GA}grKm9FU_NBJfE9QsaZ z3jDWO(#T7iI6l*h=^$E#3DZZ$<%}C9e!b@2JLbH?T=E?pIH+WSaTuM^W>G5?;~$lm zxL;6d%jE|o4IURo+RG#l_LQ(F%BGI!i$>ilAbmF7yr?(Z-EYa!{BA7k8wXC!v#`EN z3ef?-RDtIfDVAJUN!Pd^Hzxc>^J_$y^xLRuV4?!E^nChJ3m}_~q@DEp0kyQb$P@|GY56ZhCdwefh$9n+T%j><5@>&H)dFy;zAxC`w z+#9nx1^vOGbnP7LwTQVtEzTJ?^-s#LNCp(0f*$@l;5fR*?i8IB_X}#=p73|@HcGE5 zkBfUmYXXrscQLa;YGo({Bt{+XWB7XG-;5Wp&A<*~r@pP@hBCXBW&q z(8^PnudaMxzH$47idI_KWv-JYDV9Rc=W1VYYjJ(EYf_ycc6xFE(0(3P_mDc7ad%mf z1Xa52Qm6(mV4B#@P|VH){el5);jD8Xsd}e zNut;_6j@D0;uz~kc%5&X5(85hXRyx}nH6uI0|u<<&2~*JV_>A9)U0&_TNK7QX+vVb zMY7(3S3X51xX7kh07^P+sd(owOZitNI;C!IG+Qcq?A4^$FU=2W7OnGb)Et`{&%ZJA ze%Mmy-f4qQ7wAE!oxDM(DqfuPVL_B|@!V?r@FqSRP!2%(Y(6j?JnO?AN1T8{)d+Wc znHvx;%OMALUM!fKb-)XI1h#WOmED+?p?*5M4f273YDis9pOfW#+1zV=!4O-Gq8IdI z>)T#$T&?+{Z>=GfZj$1zbRB^@x+ch^yB&;qxhK`>E}in?2t6%bod@2pbz zR*(2It6}TG9kD-7c4I=;2XDUqb>jr{#*yCdlat&QJr2B$(q%FoU82|v6ltI$>tvVa z?SR#Ck1RzBY%z*b)fTo6kLfsqh2aC>vz{P563E#tmk!E%<*6aCe;X~-Q~nwnBiq?t zvWiVo=zNn6nKKv!*zAEM`Nfi3ii7G*H6HH>)#v_Vl2j+%E5OzV{q?=!KCu2?uVNL8 z2{%cKCh2Kqx=SX#TA{^{*)Ciy%VjQFlZ!D5C_V$_=#NH(-|&kElm9Z|i_UC@8~3^D zbe}E0QLb?ma%kO}x5alO(0Y%LnHl~*tsLj!kBrMPHXuj!oE@G+oVsFUm%oC0aTCBu}>6~GLF#DWn|9t8nAx1Yv z{%7rGvWuIW;=s$B(zKeGv>gkN5M(p-^(i-mDu1X<9pNE7$HS|n)~R6#xTS#hJL2ucK^pyF_V zht~?jk`!mf?Scl`F1ij%2#O^8!6~lrN|UtGQIPI}>{V3(o2Ef8-Wl$8^L54w7n@*X zgihxCUqs4s|hu9YA2EKuZy zZlP;tUY&W?tAeQ$EOAbomdIc0y(l7$Zh%6ejyYABtKIDn1(Eaq`cd@)JWBPwJHJ=Z zF{e{rC0*^6s!a012hMn`l-Nj}lpBJ~Cmu`ZW5?Ch`44m4jQ&jVTmM>5Rygp!KFA>r z&a9BI(Q?N@n;EsG(r$lkC+A8lV;Ls}-41j2x-R2Lo zq2N5;X3yK}l+z0Nl_MX@{_9O6I%3!U+jmLXOXKY}n_%fI#h#+b2`X}(+g?eIDhX`K zPSqx;n%?ZO#5v1%&zyET)_IAuPKi9tjT&TXZUYvnv*KiRhFiJ30*W1X@#4jK!lIcE z1-0^fpfyk{?Qm@ao`(UR4tN((k6^KXlOmT%^*Sajrq=V;yxt}+aowak5Kv4Vo4Ugz z*DFI@?6!@@-gL@*-iGPN>1sN4?&uK98Gn{cjOFX2g6Ng{mA^A5T5#C-vISCwI>DX8+{Xv zrl#w^oBvLhjl{s=z!g6^CI%&$Vgc4`smT7Y4!31dZ(ibt7x?jyYUoGxkY&95wzxo78Q1sBJ!KJ)F3V3Zy zx2T@JF|!czQ?5zkgHDFtrg!l$rGFb8Bg6dc-Mlu%{jk!Yb^*Gu+vp;{WVV1g=3*}l zal*?u8k2c4a?Csw^!cn=2Mz*S2>4*Vu|X0T!-3I4kb^~Sq;@GvpkY!hM0uix$6O52 zL3>Uj!++|dK*8|G?RIS0E16ZF`Ym&h)q)d<`pfvTA@lfE?H1qS2*^AR&^9RBfC8gU zabRu+Xf&sXjMuBicg68_+wJw5cWn7}kCc__uoVC6J{3$b-#D?bcG zow7?7Kerr=aBuK{TVdG4aEP}FG9l=C8xT?pReR^5WntJ3*V8~~gl&PW$8{0&*rPBk zKA;DRWH!vKr?d4;*QJ6U3I2p{%sL9Y5p9YnVej05*-4Pox}T}#XN6V)S2mI~rHMEA zgD6F*a?ok3-+Ax6sXIlZvh*AgHIgBJ;Wcx#SaZ$^QI6~Wm2}>`%Np0mZjbsdkqypd zugQ`#hhkyFB7=(D9G)F=fwcQ0`O@I@%S@rD2ox+K1s*AQP&Rj;7cxa$B>iqnLK?l0 zPy!`jusNEeif}s{qOl;wnh_su8%{`Z+#z^;tSFmN9Z?f~Kb8KZfE6F?l-lzcqn zd$H!6rWU9}ZwEIAtq<9&0uk<9rjnEgAoav{)y~jXm{-OL4WEsa)v);NiSaq$1PjOA zi+b~oSB`cWL%D45>QNDGo`>hKFnq7)p4guVj-HGD|r&EEe$G=XR8ZAAf{J=ETg!1 z(-TwX(Ct&=!JjS+d#c90 zB~a`tiYx~jTScmJPaxz6@QS8F-7s#d+u8rMFgR8XvV~{ud6}Cf`;q@wwwg0zTOgaR zlco9R3C{uro`KTpf9JJZwHIQ~_iM3lZZFEF5=zMJ#N* z<4#{Lv)3=4?QlyF9;X+U1jVxpfkT$vFcZ!SY2(?d$uWF5Uqj&B{&2AtAJ3c%{gK^K z@;-gPR%-MuME69$CJ8T1?%7TgPb8gUVS-GdB6l0&xUlWQ+XEa)N4=i#i{$;1T0xuQ zSZEA$KzT=b*uTu@tT4wFQp9*?h3#~Au>ApI)dXpVj`6W4uRMMAhSCTTP0jmjZ^Uo1`f2)1Yi%ZU?uA z;&rU$!hPS4IZvH3+k%c6j zbYKQq6#~>k?R29O+3NMAGMJ8|?F}oAi1u%v(xBA&?i8ho(?bS9TOu(m*RPAtoTiPM zm`~No3d0OF+jc9M8}mE1!~4Z9IXSM5yF>8zqd$J<9iw>(h@JTza*UgKabORn&BVN% zr`V4vQVSY&px%MZ)LHa;@11nE?>cfWsDeojs1@`=RMPUjB*n9*){zP(i_T&CX5n)> zWtF&hZlmS`(*u-$>qs?d^d?e;Q;~UXBWV|0eghfvu>_+(xFz_2Sg$tdUog}C0Q`f*TY18}F6l5$l}y3L+Kt|2 zGqRwluk6)-B=D<#a&AFA-A}Mxu5vISPuL91OxpC2Wr|k8M(^T?jv48Ej7gRT)w<*f zAIysN0ud2NjeqKd^b`gL{iwjpGh>)$S8XY8v*$9|O`=`lUZqBIQsBu@mGFVt;xlUf zRji%`BRymli``!3?NXrFjrLP{7JY}*i^}Pixq0F?1@?f?fr>ixVAzxxKO{W|WahWH zTH04mEILbl#fZE4#ez;!F;b4T1Dma@CT6RSVnHFi22>7xPSOK(F;yGT8+J!#dN!o_ z+{;_-+e2TQw8P_qYJlD=$q;9S=#*K)OY_b}gX6d2LZ-_2x|s*)NL#yVuVC3iWht3=vY|c9C)Q>ffdduWY$MY zMGeYs+OROwVohS3(!QzrBMm7B6pi$_-L7f=vEfRnaYhok%->3Ozcem3OijazRm&;1 zgdzulPGbs^9it)1nS5PQ50u9s+)Qs)o%N~?U&gQH9q`EWjb%ILM3X)c1VUB9&fv>r zz`b3H>-I*{son;ufZO~kW=4|%UW}|Cw8axZ^m$o$vAQj^A-HWeBoE3Hg?O}Ga!8e< zE+ZV>=8;F61BA>S6^rL1e-gK=?+2*Q%{M~m>s9~0n;do~mrUOMDT+Nokz=SzXNctL z`CuEAsECaTZyDd)ya2qtv!JH}7h#WHn?|342yYTNqOk#4v|*7CB2BpTN7oU3T}*<} zLF-;clLA-?z=?h8WFR3nd#hSN=6m% zvd1`let@`*0k-BeAvmr*_gVSf@4s$@(~oY{r;>v&jYVoS`Bj{x*y9vAN<|)0>C|cT zHPQCaMIlI&-VCh#f4i_3Nzg#WGj~1FNVN*?LF01zdN5G?!`-KTKx4=r zaA});9hO~qW%HD_*-!oM1)dhv)BAZ)`wVBT0z(VrZ73jJvR(3$&}~qff@LW;W);cL zdesMw-b~r~H^KeKV0?m|ar-A@aeEwhZ;V)QdydmcH;EFwNaM1VDp8^#>?4h=PFXo+ z^dmgFA+Q=3qw9XuPkp@O53kHI8ksc3f98?B+>DIlnoj~7iNoBUDvAZAkTNRrT*S`M zMrQroXmTf1+YhS0`=ndlwPyoL+|c~s-rq%tjwXWib2oi9;8O+Cm{u?!`+i2-4)VnD zq${*AG=rzDVY`FVARQTr-eca@Uy2fTi~6O#VQq?1x>uqtS66xTki#GdF(?OSRnUs= z1}%V5!_R6QeSRXf>dI(=l=y?+2;7Z0`e`#+ND{f>$bq+dcA4NPgJL0lE0v0jXK#DP z2lab2YC1`svPAt9veYlDQA9kQ-95EdeuA9$uA%S3)g;An8r$5ZcaLf-hU=J&3L{)U zWEAcDesg4k%!rYbUwiioS>wQ)&AUxTeGA0`k?AHXGE1H1iQBxbg2Vou>U&dqJhMP5 zI!}DprIb#ec#LQ_PbuM}w+#+d0y~E&m@QLnAtAa$~!kLOP_z3#H(RzNa5TZ`da8Ifm{MbQQ3gRPTq&hC}(4KENExg;s}M4XF&RpAOnsbK5mWVR2X-i)A0i$t#;lrq zUeM@y_~VALKl=Rr?VlSlcI@l7eM#F(Gg(EO#K`Vb>^+LyrXv5COU1wU$1DH2Z~kJc zBD8?bnYofzOyy5s$#0X#;H~mE7gLu4?@j!8n*KEHjgi`Ir~a7ejwr=Vqee{*-2!F% zF@E`e@l4O8Zg7K2UX6VWKIl^=HzQk=(aM-N% z3p-j^Vw|n^IHjK^OS#Qf4!m`fW#T_=pxAX3Nd!F-k5z2$v?f1XS?hPm+nD2Ys`ov0 zo;W>hZMeM>1uf#yDu5d;zr{}Ma6|n=|C?WQ3d2LH3NMHxx!+OdOIu!FxM|X=nT?vh z*;V{v`m}e2S0#UZEgFZRw!?s0?vj%&qhxP=N%uupnaLn)_)XH-=`n#?tlXeb+dVT>3%bBhui&OKZsO8@7AfMzJjvxk5$O zdKHA16T^QolM@#M8@(^wO5odLtWvu-mPKI?b(Z%D)MCZdNJmO#s}^7gLIo(uUM}Oq*kzS z&}m_eES{~W+i0kkX;Un`BH2sQ+4bIynw@mNTMN5%^1^yLIRI02k`#vTby!eJFKnkD z&#i~A0(aAfmMNcygtagmd66;ptDpZyhAW(4Mon7tCDB+4wH-L5V4+Yu&F_$^4)P(A zCbqh^x<(80gqwZ3LDLB7!0TlB45r=~(r(7ja&g3xrJ=X^^*hJ^ElFrJLccAX_$RW` zfsIg}i4odFvB?;xh-?+qN7MrC)gDjW($8gf@oLCwUtEWsRHMOI>YP9Sql@$LW}NeQ z4LQdFxBtGc5jxzA$B%!s;+FXyiNmsTTZjxr37;s=PSN7}JQ`$UVy7=B4YD)d8IW** zJPz8$OoOZv)QBFzc2+X*rs2g8ZmRdmp1=2MgP$U2j10{wHu&u4b-@WE6IL%7e0QqR zggjwumJyvZ0fnex?!r!rh051$&+@D?)j4$89PMQSN&Iv?gCt`HdNHi`?+MRkE=W5R zpx4k1d0DGM^2C_uk2%7Dhf0L zrt7_1dq}PW&pD6^H4Mqc6bm(hg;XRGcVhUakU(wA9nlKK;ee>Q+OcB~EQK0gwfj+5< z=?)%4c7h4cYys+F_{Kw%{zrSl3Dgs+yas)|jM(|n*$L-J3b)-i2Tp7_V1l$fiUrQt zOe*p?Si#Nw7G8;KMR+1VTfIlV)%~1or*DyG3iv!&N>}1~Dey9oeBOHVs-(a_A)p&1 zN$`oGRx8;}swMVxe8vW?(E{hi&;8sj{E7L}$HJPSRe(QLEN!d{>J6)vqnB z$%-J=0@Y=`@+ZLsk7lw%>P5)r-9mC^+?G`^t%AgmI;d)k z3(58DcS{v!(U_^87?vXK1!N}(PY2p+mKgzJ9ARuW$Jl6=8R>g*A(eUMz3;F412yF3 zx<5HX2Aw$l4jee_v&jTo@f5p)B5_n?foIWFsBIYXX3FM_*V5T^tgL{K=>y|*_HREt zd(oUP!eN)j7KkjW$s?ju*1WNA{yjRGJ?xJ)C=Ef!=_JKH*s4utx4NHJo`4pom6i16 z094v6aV-tny*^8kI_QV0FpFWkb;X49z0##1N$6YTQ)M0fGF&YsJ9@`ZzBMrH=5E|)OdtM7jdC%KCD41bliPutW0!3CK zt8T)ph6N$mW5dohviqRqpmx09r?s!sHj6{+2Y}Nr(s7;AV}Db$1{>Xs4ByQ4WTyi= zrJ#&CY;UBLVh>T|02P@I{^s^~uL51n&iU6B|9Bwp5q8rl{%Q0c=BRL|=p>yPnhXN+ zhdnaHZSn@qo^KwQf8(ul8Z2r3*mFR5S%_7Z^&a)A1b)<8jnK{C4Mk^$2g^Lm$Z>kN zXM<)bb5)YC;A1go`j?SWWi_&gCKxOJj{Hg7u5v;)b?pzYP96(4k>ma}hy`wiooVk*KEaP?`lW_b++oI+WpkQU_uZZ-P(-|9 zD(b3gs{*pr8>UC|+d;&&NgCyys_0iVi*!odrqgZ!O0`Ecwkz{#m!{F@z3ucw%|Sz73MzraPfX*oy0QlhpMzAZu{du^^qQ_t?$6I z7HrfFbCOahb^}G$fi^u5MIr-Qe^?RyXsWirwZIjkMM&d`+qgEYpcxS5Ne zcK)r*oZ7`<5n+LI<_PS(C$pfjc~V|V=ZS&ZkSSKB0SRIb__3>bDT3Y1v8hlBkGT@t zg=M5cnMc-61pYDkqDkGd5+2Ac3AeC?svgq8izc~#hDY#jo*4J>dU?od12nZ*0bdw4 zbP2QdadN5(!x<|vtQ01tf>8@wSOKRO&Mz-$Fe+eigHvo|r-!Tp#*t*~I_&rO`Y2 z>*~1INS6cqxl2q+PWvhL0Y!SK$dln+lKx2rbR{2&!R||I1$eIWZ6%rN3J=Jg#?wVi zhwBBp5;6*U$jXV8g2w@c3?^n}^KOuZ4Vt}Q@0x%9yRi$}-`x6c=2y2;Xp=p_oXqiX!oHnJm@s&l;%I6NMS#lR!0G*7g(aP z-5vrRC(}fgT#88kqSqANq#MIC>1c9Q(x6%QUNday;Q;S*y(LZO-E(ru`->rd*26rhru{@>6-9c)%Sb41hIL z+Yb44WuRPOs6|381N~Hns-~qcmdj}4#Rg$>!&*B}$YnFme2kudZauOr zn3CsrKkB)b{;reUUsq(R|oEWdmr?AelBNy(IbUMoH7 zWQ}2Ep5thGxPfETW6k+Nr}aPmp2TQKU&h>-d-~l&`9T$cTH8~%8ljP9HP+>hu zfj)18vY!lzcC`fQ)Vp}iurq^Kj_?dpUR|^{RnacJ5*{Pdexgd8)DE;~SpHfSkWbgj z*ZObryazjd*&z^h(5tKD&}Y2Y^R*kLw}SDrn_({zGu+G7g`w72HI85$(MWN=Xa#J) zlX5f5Jhhuk!kq(GiC9Pk-v->w!1o@po|nt4mlcSYYob7mViEJ$10^YR%EPkAcd*&)R=K2=R;yWrJr^tnm3Qu~>*F`~taaUWfiujjv=`X!?=YWkNO z*`(;DF-G+!#^@-;R#2oAn7!3ILJxtKQNjF0kea1U^H1P+%d~}Ii-Z*l3=klpF2)@? zgP#OH2!^;MR?F4qKVB?BbRrxq>;B( z6+K<2JQs0+{^+9wYTtZpf0Pu+ACf{*KJ~$9Qyrc+Hnsu9^FCp8_bj_bu@UjJ=NaMB z@ZFd!Qq0ZjIPe^G!30PrDE1gdj!=@duvh-nXSC0 zJ42BR5%Vn3kJT4Or_ssN;z)DQ2~rh+<ltRhxZU5ddZJzjca5sP(Yb~y*ktiCW9?z7xLSAvNyfI)EHG|R_VQM-+(18s!*K!s(1&sO zYZ?3*Ym5FvVZ_sh#DI%ry#sF~7MWltn_@xOB%O+^CKnY~{1U@gN-}tRJn;xKL?4m9 zuxu)4@}1B;Dv!#BfAgp%)L-)HUuILkY|z~H2ek=aU)VX#ak{{+L5DGcg2NzlZE6`5 zBUnT#DRA|{e=d2P5=i`90*PPRXki*;g?@niB27}r4r`JoDV7Rvy?_Xw>u|iFJtq{s zpf6kA{>#tX!+nhgB>(Am_mOmGSk4Z&@w+J&VnlhsOelEb_tbAMq-pDvYsh)>)Nf5- zE^}A0QCcfQHo!Bgy`)F<)UTfu(s4{-*nVD$|2?`aK&QN`IIG6ijhfxOHpQiRk0m=u z8!2Fpxr}AiYVkF2de`PYiSgWG=VQ3R?HB8?6{>Fx{PNxxJr0o_nj3mn+z`?(sDc`% zXtLbz2C!(h!mcYd{ z|8r^3=7$aN&b$`Sv_hr?s2O;j0iD_H!YdNpB%N0xyV~m^SrZ=ZjuU8#e;>V*zBlo% zqLKa7>pFw?mS_q*lR(z7Dj+@xStQQ9b{F(PQk9wW;+YtJfucn{)aNcENXUn3^@Dj(W_!NiP-3{WC+%`ka%avMi}m}=qNUoA{(j59j>Q*@ZSav zW=B?dtpSO$v(o$0eo3teNtWY5t|M9+HTBFJS0$TguJ*Obd;9E*T0G$XW-WGQ(;M7= z$`0$Ed`r5*C(;Ozhe3aqlU8oKzz!Tvh%!k;@1aHPA5&ewA2MBV^Ti)wH9_!|CzqMsx)?Nd-D zs~f0kPO8tmlSRh{K(MJ!)}p|`Ix@KGeA8W$3D#Xg_GwrQmky-Pyj#wO450DPoaatp+qvA%pjah%k=Uh{g_!WlzF1i19%uJ9?PoMq#9;|eE*PMEUOL*LYL1%$M1&gg9 zgG<>gS_U+KgHFd>u!6r~jFHG^Zs{YAXDM=6UpU=l5?w>Ft0}UQ zio_&PNDPOFB=9}Q$aL;@(S=3T&l$pMSh!9*$b{nI_2H|C@6tq5mn-RSV|4>;`vF6 zJCc0h{ljkBLzhlA9Ic$#;c7VAMlaE<_Qp>eHU~TxBhS5QC4e~15zqZVD{pcF5cQk# zxQ(pQ!Et-kcZqCxX&jurCN?OCVlyd{fvLvrAda&>WUorw<+)p>o4Y5_z@ezmPDj_K zPIhc6+NgHA7m|^&5u_tyNa}%WZ&)V;P!sGavotnbSqYL+G+XrPf}3(9Qqm^>>Nh0G zfulo(CS#vPv0EsTMnyJ@4u(g;4nDe^b%MgF7%q$>+UuU}f{UPunLmFu?{B@ZBaMK_ zROd<_(OQE#Efy@{TxAo>Vd>?MW#ldOX^ZQeEXsu6tlsAPV4;VUN%Dq=k@ z@oHsH{Lm}Lgqt{z>%pa>9YFfIozC;yFW&5_zXOwGepruN8|*W0x4ilZrg1h&F@sO1#G6{AqQQTg zB_HhbTIX9Y$`i*2)yWRft%5ZArraLNk2b^1&|ur>FDJ}Q_}=t=o6X}2FUbnqLKGs0 z)~Qjz;?}(NuIW>tz)qY)mj>x;b)$rrA;6j|$)N{8x5chAoXKEwgp$d#_B=W8yJ&79 z$LR*moqsW~0<^$38=g1(C!UW3RycyhIMdh@9)`YM7DiW?rw2<2|5rcxY2>fo|HZ=J zOK8@KP!0?Q3rWCdp-^QJbzQM%{sYh&*OB#p`C%P%O2P-_<@7etYOEFLK)emFwF>ZT zIq8>FFvSs=voSDtvu7)bnv5+5=61}6v(gj%uAc<&SWjag^CmN;V?UOyR_ zdx05hrBhD zQ9qV6i&`Q1d%!(CM4zvG1+2_pK8m95%!f!pnqTHSSLf$S+NG^f12^E-4Wdi-C_FF@ z*cd@EhGBpHqn1H2Lf+kkA18;0{&Q)Y|K%f8nX89_$W}px_hnIRKo*1r68NQ%(vIaU zw`I$zahQcZ8$?bx`s^wDAG|ZY=%-`J$aLV45#;p_Bj`+`*fkVcO-1GjfhC|U;D7|T zX39i6MYu8d)Twe>vk1BL#uF4Uyw>Bw{#Y3P3Hv)eerUeXaoC`dg{8}KP3F8P;VE)* zlA+LT!0jrO_4fz21h>=e{#vZFZjx5Qnr8gQ-1_0x`wzoM#IDEOrjuRw{^uv=>ADUJ z4hwn2bu#?jqbItTpdJz?W*3Gf!9K~|+1fKMOEi$P-2k*QNca5M4X$B>Bt;Ebj(4Dy zODzPSiX@#N>3CSFS5!I(mAAHE2Xlmsncq3=)W!hf2ASe5v-|$XxHf)cN0m3Z=}h8G z5=jOq_8~>?W1J<+^9--tKQ`d09|)SUjb0z~8X*4L7nc4;&-{-ek!|-^dgg1(rC6eq z1%rs8m(Ebk4Z((3x^Y{EStc+}rK&9~?%m^!`wI+DuXz`qu9*4+HqLVBoEeAIS@d5^ zUi|>8qEn>h{w4Evgnocho%nA#baGC1L0;zhAgt!~54Nz^<)uL@y$*uY`oT2`9uCl% z^C}=RR2W?dA0`yI)12^g2-;-7Za@X1H76SsGMEYJ(Qf?PnV`J&;wN z9@3#%>Wt@~O53OGrS*~t4iyocV!Tjgh)SP)t@r=1Wv&rWev4OZCkGrDQ0GkmRZX!~6sg2=+jMbVP{|aO zd#_aW^1A6#6|jmRD-j+5NjbZ8a*1mU+bG{D@1A6!6VX0ZrHeD%Fj!tH>QmGQ-3!#N zfvDsz-YHQ#-Ad!F?ZQ2tPo4IFq^bT^J5W7hd3m}RYt+Fj!(AclEj{GC_wH9m?{e84 zN?b8yY|jl!i-rARZyBNV@T+hCgj9}H&F#Po^=l^Zs;Ahq6gfpjBIm22mV2c~NqDmG z++@hQ0P?;m})uJTSR`({hNxI?nWeRP-6e8;69$%+E?rmT>eL( ztu^m>n5{KEznxnBMK5Y83&;dxhtbGZ_iKU+WCPO^o**9ez&?AG+>fjS1^0f(`UEFXyt1OR z!Zpyi66$*I)*h0}Edk4c*GR`rCe&hzg&DRG60j79RG72bt9YV_cJB?tOg+Pzsh!SM z_srHQ7yBOw)W!!ThM`e856Q(kWv#qWbjJIPcaQ8SU!Mw!E)L3A>XbF4N(@G*8wA9V z7*adL3~9Sr*E*zGuq-289xKF#o)7p_zpD{Df6<=*J&7A>745iFD=>Nv4|i{**d&Uq zp&~0)(d1*CO^lIlZ8~#GnKG$CnM3cFLDg)V;sU9a+cAS?==~1;@!m$MEA%Qb_RoJZbw`6LgYB>#f=22|99i_Yca5gx#_WnaJ{jtx{2bRt%?)xXV0ZXgQ90cSA?XaCLHRy;2_MB(IzuZ<`XeyPrCt0@SWaZOZGOdkK0tJEmg{NS^~%3QEdN>@gSM z@3gz4JZh+q2vFNy{KqdgNX!W%PMN&iQi?r9kpm!!Be)fa=_n6L9Y|4wvNb&j8BnRd zJL%Q%UKhEp_vwBmJ9ZH9vAn}cRj1=ME+xm^>*aBuVd72-lAc7?#Eeu^CdAdHPbThWi*zFoEydTz0XZW-V9(%O>MEB*mHy~(@ zfygY+rCvFpw$cJyz&a)Fuq^dR;3J?hhz(N9b7Q@rX)e=GDws6-lyaSKHeE(+gmQqL zje%}8kYCi$GSK>^LIpJ#LUimqY5+2bip|*8i**W}XAD{1IL+uR&SIT#+o6Hj85T zk3=!DGu}CBof7N0wFcv;r%j`8fFqSP7njwScs*{LT#7_m+{sBKMbj#$tc2{Fr%qY! zIgkK|Bo)zAuVDEctwHby({l=#OY_E*o@`^(e9maT2pyL0TSmi*J2wUNXmJ zx#hrnITncR>tvTn&W!YsQzEE+Al<gfsdMqK-J=nshm|Ej*ycK6_Fnp((I-}9DTL&X18?1|GRdTWO0kbA@+lQrCrb}$ z7Oe`YlNo4;j_@#P^IXKvP@H+feU1>2zf&atWS{9axu zlPAtyu;}Z1-lz)LE_v#NwVEqkv8bk$J{vFqseIVJPKF8Yz4FwMOY_b}l*-zrkbmr{ zZDh_x;0=smg5ns+Y>;>{Uc~JAq*<5d!Diyn7d;uiFQSV{cU=Tu89T?Xux?H!-2{H4 zwn0`Ikfb=_vID+Z`;s;8cP@BI(&>G|zWdEz8WH%{)p4(p zE^ezb2R2npOz_%Iu@5NHLq%?2u1c0d1_%sx9JwmNaC&bTmT7^YrDUBiti>RQo!`M* zB&;DBK8ZZ!2-W#Mb?N~^&{ZL2q)U=6SvIksbSNtMjE6^x^yV&`c08Dy6S0hA0N{Z=LKY$#8E2 zS4`gl&am~a^{xk|o(}A$>*n_ga6ws0Z-IJ}M4mR0r}HiFJOjU!Xa8(-ucx7WIZ2V_ zSvjS{8@@%a++l%Qvn}KUb0Sd-44ir~izL?)ewtW&H~@rkW)w#hhE*_yb1KEQb6Ksk z=~mAr)^%sIR@__^hutQ6^Y-^1co+@Rw<|VZAZxi9BnMvG6`1T~Y^7M>A>B+xq7NSL z8xN~HZJIbsSOzmAuFfEw)k1QB1Wxaxw*z-@dPt8XDI^J)XrXDY(2$#KU-(-M_Ro3R zl+A!WQb*tU-N$Dp8d1{q-_3s~%N*Eo$}t)DWQv6x=e3}t>X$V4dhl_&L$k1VS{l7n z7{fnKFB4vz+Ndez8P3xE_VVmIz$P!3+i06SYS$A^rj6Q{^1Ca1BQ|Cvf3un_aV9{n zIV|kBo?_QhBmv6wkt$rZPlenrP*>y{M`fvz@uNquV`?`f;f~|G#}{b*`?u@s{#@sr z-oN8|gx@R3`VaGTJ}%302cB9iFj_ZBkGbSiI9~v-^_+YPnai zr^^*WqqX>L;lw|Yl@2_Uvo6uX!rQTFwCtav@df(CaH(Bti#kI4{gUf=Gs2S8CPz_3?yP@()NP#GNMZ4SKR zw-Yq%PzUM&q?KpVkblOj6zt+x3OeRAXz;&JRTmgLlIn4Vi@C#T#rqjm6K-&6zPon5 z`OdGy!o|W~ZdpLCCuD3sqH{g2%z2>Y{94;hSC|da*E#J_<5$ zkHSy9e$Df$pu_+YU zfXqcDKAnOLpA8Vbc>sZDR7SvrQ9~-|_V6yr6>%nA1{L|$kTN<>VB(G@TSDeJeZBeq zMM=;QK9YKNlN~P&K7fX3*iK&w#loU!KNT4Z5#~Msm9frCH2q;!OoDKUbCmE5(Rrq;ZG|uH)@azQHPYd8Fu~smJ zkw#jlEEcyDqJ9#BD0;=byZ6UNY&hzK(_RNZf#J0@k3kz69rnlSlJP@{mFO6n zKFr3@ey5-Kw1oUzxsgqZ9N2TIHyQ7v6k9=&QYd`^mn4loPkKDFL+T-0M0-tw zUJ7i-AxR?YP3O$OY=4|&YDMWXL&iT6GGKNiPCxo2w#bCNtWh$hO3FkN-xd zQq`?jywE8(g(Ql#*jZQTfSVRsky=6Z8c)Wwy~`08LjbYz@92TUZR)9=_%%_e5k?}H z`CG~Em&VAPF#%mU#gfB6-|A@tC-pbQaJ;PC9iIk9P&fb%_Yu*PO;8qKKCHI zv1HyK@9|4T1H2NR+2x9?IqyM#S^zQIY9)@!O?_`uCT;iPO zZv%s6AHal$R9GH?#zJsXVS&^vqKUrvMl z!rpAxfRlB3#%Dc3&pPn>4PK%$`+h;jdLlE&?$i~NNFJGEetCSssahf zJR$0=-xpP?uzVnHT-0(nVPt6Du^kjF_QVMz6U1Mu`cjy22Kx5??dhb@fgPDTlR4!G z#g7NN zKKT|_U%}W(Bh?G;kyK8*D9RI-@;Yff1s+sBf~Etq+2P__i-M9^A%fJ-yW~YHoQWM(Z8ruDn6VpBsJH8HZ|^ipT+BhI-K+LgT>=e$=QBU|ZVc&1d<19D?A zA-!R!Yj~P0h0oj~HMG|$SGdN=OtWW(?gaix*$4)V{lqoELEt+(SX$dZ*^TVzRc0t)%Jae90RV0IaX&TgmA-_cq-A8ti zHW1n9RhkQ#{foTP=w@rQGhk3@0nxW0dT=i~MI=@ziJZ(#z>tt=N8>SES5HoSZbVKUt z7_m7j*qP5nBc6`}Yr`Hn^rze$^W;wpIhKpmDQ;VVi}yaA&X$u4@*EI(8+5uLUqc$^ z3_7K=`+04U`Er5u2=H%|aFZMUJv6h6ZKKot+wHyGKl(R3>~+}TklPR2VP~Z+@42m4 z8ZC~d=KaGY!dEpVT2&U4Or&iQ^R7MK!uP_ap2$Jvdu_iEdGfYu{N zP)nDIF*fRoLN1tYD5YCJb)m>C2A4*chJffO#yIcC%?4?b7SN*229cygz{LQuQ!Vh` zL#`Opz#G_I!W?#8fJ=F%oX5-%Uqu&UhZkntw|VK@%l8APSuyjM_f4f_JHJ`Veg86Y z#0L8NDRwVK3aQxAkW?i~9%7kBYEbFixY>YX4Byj5m&5d5DK3YdqHhLa+dR74SUG`b z)_|jrf{KS`p*r`6{dnlw(;JPn?OTPi| zrEt}kAmi3IVgszuxz8R?AjY9sr`P`fabqYo+d%14iakb=S}OJo(8Yo< zr8I8l#b^Y`--6CE-7`v`-xu1=w99(f+^O}G%4Jt*+`m({Z!Gx3AU25RB;GT3YLBvh z(&?D9%)ZbTA6T`bTOlfI4=>X8usMPyK7AgR>vID#rKf1TW1T8j-lY3nkts5=mSd)X zC8AT1_NkwAeA+TkXMs3;qzy3@58mY@;7W3TRc7xPKO==Y9R7W3U!Sb^~m;Miz z6e&!ZxB^i2CxqHXMV+b)UUxxP-od$FtN&;zS~hAvQf3C3NV%*k#90hDG(^JzibEgZ zKVeum{jkp!bW;K=;Wpg zEctFax_?`#n&!y@g)kGol9A_Bl8>0@j(>_S%tr$8QAOHr>*$nV2{nIY~DTojK$r zcLeW@#L}DFGjkOB6R-@1>sX@*=m(&52-y$L9=(m%6w!CTZH;hVI^6Xma?FjZC$8DV zEaxcp3`H8K*v~_|CRh7genK>1ewP-Q1N2dCprw#T-J08=yYa@Fw|2f-{C45D*S?{@ z_9#SZ{5!%e=2#=79(k*h7c1Qz0 z;iJ(r9Dj*=~5*9-1%Z>1((I*tRg=e0TISxFB;4(NM z8Q!+Tc;jUVss8W%q)lQ4*Ss_950mAdkggbdbp$%%cQ7pK#Mr@sHHP@TNa>He^B%+4^4@7~wT7=Z!_&@eFrNQb0 zZewkPi7A`dLq7Gf935dC^T+Ri5pg|c2j7t&eubS0yX|d=+Xgut#tmEBuK2Y^?xFi8 z#rx+Uz>HqDM1H?MB6ndm% zWgx#BxNQJOK-hiI0?>MjO`*tYDi)Jxy~4c;Or2Tu0-JOv6rZu>J}r{-&##WIc9m+6 zgX8HSj&Qzl@Q$uI&BthZ`Dyn#E;Go5oGH!mYgg%!$1;D?i5HM;>cVfGd~<}hlze}y!(AWu_VIkqp1H!znSdvB%j($?uRHA$~{Uz zQ+#48hz4~r^V?uO9acuv`@OLE2b-Cj)zwjYBvEOAB^1n?U=lwa^v9C~*v-66TWRhm zvQ*)lv;5dAkSp)-S~4?5yV}fGnKR{tZwu&Rp-?o$u>&FN{NauXf6gb4F|gk~%rSY6 zdr*d0PI2Ea@z3UOdBeW%jn5LX`~GEwqf4$?nLKql=!9!DIP&yFWQqWZb9vZnGRgJ8tqRiU? zU?9??9D6S~&h4@dR1QdDVmn<$NZSjiC+r^Y^p|49ER^-z;p~^9H>POTt5!QSw?xuL zDxaBH30kHG#Y*)Q3$xyFsQC{zY7aku71o&3g3c;?2vjghcY5zp?@}j2-tK7hRqt|1 zx>u9Vq~0*$YxuWFYEp0UZjs>L3OXZVrN8C;_3{g{vN_lvg$i2f*co8H4^pPBmz$V8 zVVh*NJZEzA6jXleVlYNq8`K`&#bB2&>H3c^ zNzGGMW_iU1IgJzx{39o+*dFf%g1jhLEK=Vgd#7L{Z7u!zD}Vcb;=JWAZGW@v-GHECkyjB5#)^$$Pvz=qk~tnml3B>{Rpi0#M+!+%qdOClYc# z(u2XJk!#0RiTXTB#?^>AiAivToR(Fn%f!V~I{fhZCLQYWwn*?d$fUo7Pr*JX?0e#i zpiw~&LWn1VxToiwgkPR}%ZjV_a}Hc5N1qv6b=3xOO%&TmkyEJcT+ilA!OTu3WC+jE z9cC+%BdB8Q**=dX_*n~D&TYDN6FZn3L6LuzdGBHP`DtkKc<7U1SDMdg(rwdz3P`!H zUKQBqQRKfW5D=wV6^PgMd8EnlwMmB-%EVmN)y_Kpjg`Vaj}tEpY;6cG^6&H5s4)OC zi!~Vm83BDB`ICw!wFmKJCK)x*@x)zg(s!NhxE4RJ-4_4SL zOGt_#?VeL{Pg$cn&%KK2KUlEoz5o;YW zYStSV^crr_jj9UkX^pBg>Vv`g7_YG=0ZM{y4D6&)+&Ei_F71)XCUvv2*EqaU#o#s0!H}ofXj2HFp8myAq#^)uSt(>626UM~qjrblU@-POZ8OJv7{>AS zaGvz-7%h*xUGn;zsl#bRapU9>M zH4nA0>tt)g^eAqOJX}dY_gF4LS!4ZWrU#g4ausXCGGX3>dY)Wq#>8z}OrDM;aCia5 z9GDOK@wky5=CNrftP28d^j8J)Wu=5_Dw;XMt8x>`ICxP_fO*1Qjxrd|2H@(8ZlE|jK%tym?l7ImspmGvs4#u!*QEhVNrtZ?p*DL{=ht=H%C z2xPJP=pRcZXJOgoqO?pDCmwuL7dtDCpk$Cec zBQ70#CNN1*2W)4>%Bm23_Y18PO`22Qz_10i&;nT5Q+5fn19GL+ibOUpG~G)d7g`UE zJD*4O5WHwr;65>42;vPhi?#h8t@P)9i=c6C%cN|~agdfc_= zjlE-^;mldQ49}Png(SotDbGkG#DS|8vWDzY>{0h9D%A1Q_R2rF!W2S#brxij>r^L$ zQBS2wmoMHayEN&T)~KzM009I^3o4hDiPpb(R#p+RG-3o%#WOClGaAm|#nEdz;{I^l zzIl=lQrvio;^+ZwWoikQ2KIT>36U~3SK6Ry&}3@wLa_;o&W&{7^T^2XF*%Q1Jc2vI z4^MsF!J@#_pj&jdpoJV~SBX;~QEPyXu!YnWW&!mzWZ3H^CKk;2taaXH;w8FD{}KJ7 z4SFDa9<2E*H|W}t+~j4;-8U(9#3yIR(=L#mqfH&R8|1A!>3c71^xmUbL-L}4|4>*E zT|tg`fov0XI4l(5?aZHhsi|DtY7BZvNoA zAnhpj`QiB2V_&}fwCf(11|6D|CA?06Q4f}tDgz2=++$)YqjB$O*|mZ3aTpf^7mU&? zKV0-(+H=j<8r6Kaa`PFomS2$U#+g}Icpel3Wl(H71?joi(rNMjma^6kFJOOBZP4kj z&BTRB%g$r!avD$RW{aX$`{Mq0Px}@{o%YS9&$^JaZfGz#j468?Gg3yz`fKX%9e&{j zD@sbH{dpr_edMz0#>RwY<> z)qx0n2OUFngKraA;$0*=s{T|1T9&%S5g9(8YLeug&WxTLD<_D8~lkL}eIF?Qh2FFjh zAZz#0rv#dyIC|6r#DRz6$b%m^cQjx5J$Sb@65&(n$mgt6j5eeA&*Z*`+y1U(qs`ox zNU_T)vJ~b%5W_$o=nWxxBK@9kj)pcD~Ew|Goe$cXRAG@GM|j8`zCY<2ix+Em)|)CL{I5{xBZjZhR7qrYGE=vm#jVVLOa`fMI*($i8cw z_pA8dTY>S1`d>^VU-AQn8&5FHZEC{rQS9F+(oMzIL?wl!jRp1}$bY0q=qrBOLf`XF z2?S7_etq+s>;I!gg3Lh$f-}G%w~=X8Vn21U66>;2eyfdMJY`d8uDp%T)MolE6Zd-{ zpJ$Ew)Wn^E*shJOzD}Yst@GlQCY2zpDA{ z^r*Gb{T>6qIY@w&)A)y)YkRi{-L@orWybCw*+)Q+HubwP$T-3^q`x%RHesK~3LpIq zUnlE8hWYJx$m|UBGn~1c+Y8Q~y)({UoraHVi##zjd*HvUQ`#X=yIM#aW^ zt_D@J8o?6JWZx}9=s6OZ1P6o5A}!*v#=ts&L>gV-)eG8ayM*{YVM+^qZq5enUrxm| zjaC&yy{x|S3^GtN$(4XUar7K}|P6ARXs?QhK3m%sY1G!s4u$LoJ)GfXew>Hg zLo36vH7w})O)t%|pP(LfS=>0n=9q>G{Hi7!HQ7NXHXAAn?nGD6m~-x9%Aozd2E?$N znQMxA|5iz_utxm}a4%nwS=`S8PVF=AxFkxzzliy z$nE^rvG2d_ZJm!wUi;=Lvf`PUk8*70qm2~%5e1_Vdz>wP`OLhAmktImp0G6HD_M6` zf!_&;h!?#6gT`9@eRQK4N}4oU&Lqo`N{a zVpsWhs8L_KmR>;3@0hz4RPUQ~ACc`oxYMX9oIAhtrFz|{0fX;MGpy0(b#bBhq^(M; z;gH>u{f;ENalivovV#(|=@bh&lT<2p3oPvzwIG!VBa4g)W_|@y?f01}|5JJx3U>H|r85%EBR#4pc zxqpK6xp8^HCYxbSqSzG_Sw_VoPl`FUinJA|^YHGw?~aC9uzT!|4YA!tqwtoOl^FBK z(Cij_AA@6Z#$4nXk|anGHfV}fz!j1w!1i3!CC!zhuJ=vV^%`VvL9Z}XiC4GjFy+;x`1u zPyYR{?L*$DDtpEe*kW^elkRSGhELP9O@XKZa8Yqj(xm$u_82v|Z?Py%+$sDlfd9zz z9un&XJQ3sU$MC+68!>xd`(dDcyWdlV4i3SBR>>{zBD$9@mF=2+AM&Xu6qgj8e*3+? zm{2Be(v`_B%CFDZqmF0aY4@p@$FrjZ5uW$}W5Ro=ow|S! z<2k9sW=`5cu^?2H2}?2%&^9aMCg?2$0~cgFBaIpqgw;1QosxWz88I-sK%Kh_76Du7 zg+Z+ot4Kx2skuhYsjyq2`od{T!m_mSP!NK;W+UYJ@<+xaGmTRt|D=n!kug*Vd@<#} z1y6g&MH0CdneMeZ{M;OUE3hV+7_2NACHr7A8V*6i=8DmI!w(WC|I7y2m%F)b6AniK z809WoaOfVROBB`9E-?p#%^arM1-f0c&j4#_ z61cR9YPE0Ur5>%JM^fe)QutQL{1Pau|k(^@2tk@0p-#@u>$A zFnzPq%wmePrigMB#>l1*M1sUXBHO7t4lT~Bqg&t;4+Jfgx@`d)PxyUUyVagLluvP! z8>jR+R2bjs^Y}#G=W)fi&*QW&bh48Riaw9Kl0J{G;9uI5K95pKfnOUmvyYxlbI6iC zoXNlxk>u4<~@bgHCBK4N&{Kt zN%Czx&@_sLl zFrp-!51)k-5B=fABe=!=(1-9k-Pc54`NV6ty*vC=*%ywcWcXL4V)`p?Z9<)~&H7&HIlN!kXaT-}B9%a}#0vOALGHO^c ze^Br84}b6t>s<2V>!&u715cTjP?OEXcZ_0dDN;?v?p5UbZkd=Q7o43 zc$?JsBzd6hw^NN1<_41Qdn5o;eCcx3Awm@y5ci0xlkAJ`k*)PVLW%`n42KNeamYC4 ziP)ayJj}RlO!~Y1o9$ZXx4hE#9ph`uJZsf^r{FT;cArC&jzAgQ(cs0RUSX2Ra^B6k z56>(UFZC??&Yu_mWW%egeslmXZP0Cu&Y~N2Mh)I%_+J}7-vwKn^4NF$!EblPjoUr* z>>1rMY^=Y5V%Jb)6%?G2H6pO>Y_1fNPZ)|8$qv(|ajnW`C3bEO{~6i+YL9>G_yc|X zeO!H%mvM35>IJ{wo^Skt)rEQcQ|fi{nH#SZ-Le7f6^i|wA{T&2W$rbS1?iBM@MHqn ze*=>ReLJX^smFY&5%{a~BeB0Ujs8Sqn6U!ZnfpC@XK(gfDNeU0cfqkmX|c_IH^c8M zD(FVGmb6drfLmI0>G6K?qGS>1ueO3KywATu0~Y4Olq z;cgL3#muqE)p9+qKw*zikFv!_k0MtFrVF&FZcjEvW+<>Xn)^HDIQ$=LKb$?t`K{pv z|1sy6{BoMWin8B-+omNu**EzIjQ}B0ej-Mvg_d24S5`moQ;$ zuJ45~?h)mDq{BSmAX0|;Df}k z=oc4S4gGiL-^9h4<*(#Q@r=rV`~C&6oN{Q^E_FUw50o^|mFFmshG6;l#S@%W&v0;j zoc*Tab^HehuHYV@A-rVw&8o}(>rU)EtBIL8|FafS=Ef6QvyD0Vlwyxjq!zeXpgsLU zSQYqPSY8+JY5RVOC-yk!MQsc;k-tsUjBoTmLf>Z=`k+cH>Ob#NS1Ma1183ifj`uV%2y%$UVa>BrfSemkJd1_S zny?WEx>GRc33TeM-)*h-AIR;Te6EqCx-p)BOm$GDe=fy>2HiF)7MGEe1nsi>;a`Uz z7PcyJB^-@hj-Usm=h2lhFeoR6YfZTzK;K3~b4htv$IKPXN^z11(>i$R&LUuX)Ku|CE zvZXWiNpj1Z%E`M0+q4IQ(30s55r+h2nife<)K%|<*=J;J>asa^!;isIrcn#rJLwS# z%qO69Wniw)$d7#YU0mr*3A{0Lw*b2yKY>2SuM`VK*GLP!ctSRfd>wF6HL+ZVB@YHB z-XAYSUA%e~QhlHv(I?Od^|2Q?>-fX5u7_d5(8)>r$)NqxD`|?#3an#qed{7wE>W6J^Xs_a@9#di- zvrRpcX2<&EogMZk58Uw1;AeeKW~~r^*9sE3|9?J5>f9J4H*G-jImKR}K%mS_hH7At z4+BGPtmrh?H#NXoftl|INxYDxu0Ob)j+@=@Q6|1DyUGABS%Uy~BbO?ggKlzPluc(( zJD|Pi3;bcwN#CQapqrK08QBsD1XDm94*Di763h&=L7Z+F+wXy)`$84w1&Sgo1N0dY zsK1v+pOuw|HTf@ArOA7g`eYHhNI7uZ{74M;JEM(Cj24~T=oosk@-* zW#zcbf~Edvw3tU;FWBr?p*>Ec4povI&jnHsWt_Chw^8M+PyG0?upS)_R2=#MCr{!9 z73$K$?PFfFLZ$hQ>se%v8$;!k4OFTrwvr;{RIDK)ed2B*u7M{o8yN$W>zk?skEL9= zF%S($O5nC>)rykHOh_qoMU}*$NM1@HiuP_G=Yq`Qy$yn;zL&LlhhkxmH_G^$n7XiJ z(ZEUiPXw34p^y~rb~~z7jVJc0jmrJNaP~x#@`Pa=4vYqwkD)f7@X3!q%nL@(O`7(b zS@tPIJ|&>;o2dY4#lg|qgJfcF(yuGUVXtTjFi9;H9U`wo8uUI z^xBc!GuyZF=l1%(d{E)O3o(R44H!L}Q*=6T9~!l&2d%%R$`K^VTWMT;Fni9}=)EA| zm=CZA2{2{Zr-tQ>-O}CCql$XfF10>^Nnr3~i1jCQ25kr~3~j&E5ySPb#Khh0vOxdRt+!(xbK59*iQ(}cgVI<@bw zTJ{3D=EhF#5*tGA9*X^%BArw$vghtl91L#IoHDaprUv~rFA8}GAtVak3(IQ#9+(Ku z)aFGQ7!)qD><6YxSW2(++BIvHHw5Drf7hr%&4~j+1|}8wb2s~8w^P5zUR+(6-2+og zwQ%ti!&v<`t%(7NFQAogj5$MpEbW_&rx=(Mf}@xA-f0i(O@K5VpHXP^;4ani{wcwH$ z^v3L1`TW)yR_MJ~rmG}*{Bp}~T>b^CUW1k`N+}jvf%j3d-K5%oA-P3gQe0xvf?NHo zNtX8d48&u1RB1?-SDCn1^%3w;?443eX9~NRZQiZwPl0K3QK&&!7raV%2?XJKWJP3~ z_eaW}@C;!|%qdx&s!4Z=x#`y~>rn3m{o4}xSr8H}^mCQl;~KKfLHKgOggJCdVRzY1 zdkQo@92hKc;Pw*Wj_#z7 zzIAD2r@(=6a>gDFTsxAl1==stgRK^3TH1%JN!BxCVXAE`%sz?*R`J~^JGmSfnmYV) zg;-Wvr@B7{eJ7-8zB=n8|4h1L!n)bV$t_s403I9p$MTEPmDBe4t@2L`!s8Wq90Z8L z&yr;TyG@0p&~F?4r5u~?lEQ9B?HuPat6~{;XMi)vlnmN*EsV0MH`m*DV!CaYI7ipz z7vgoZ%$0j}UKt95uoY_iOd8YI+gQ z=>bRj3bR~*&Ud2(>eC9u1!BE1WSwfYtVp`w2N$JWY(iY;pSae9@GL+3>q^5~`z#J0 zZ^n)De;i$#152VHF>S}Jy1M6Y+8bn%oV|2xrUJ%=1^=gboz6=G>~h%(H;qR`P{&$awfjyzP? z+UUnW%z>l)EV^+M74eKZh{hD+sFXfLb z0{y-$dZDNU*c7wrZqgtyG1pYLqdo)vm_y?eeXi=dfj#SkD-Z{@Na`d9qK$#+Ugtw9 zHGk?-Ln~q6F>Sh6UQ|MKgDz*@s(F7(ry6w)zGtB>q9`~89Oaakj5Com;=KXJUIsCM*u#HVa+S>3|*5^? z?}3eGNYjP>a)Txl6#Og&s`_?Sg6j6n5o|W&X?6!$;DOBnL+RRAT~@(<#%izDuMRm& z);%-!YLCs7oJFzFeV9(gwn*OTrMJzg)UFD$vyA$ ze@(E$CkxsmVXdZKRUQTt=}7)^9QStUS#n%ElDnS~CwR;YhgLnAEZB`L0vA~@#KTxO zhxIF`WI$EZN@Gs|U1)Ew5vg~DQ@ycFirwp-HqCqAk6sqp` z=q8zdxF3ZfF~Bm;E)4CAZqt@%cFIt9+Hzi;co~%}&4lVK>~6th*>tNKSJ90cyc#J) zlfor}Zev2FD?>Kt! znw2>E7A$@4w^6z@A`1fWmEtlHmOr43WZX>LR}_^b&!#hJJrwV4A%|wQo4fPU#u~;h zqeeLEH(VG!3INya`PX0H+iUOSy6pl1hudll+$mcuxFE(RC^PAAovbjXPrXu{EHVb( z_y59sgqX=WMu5*ybKGJ?c5_>dA*!*S`={S7dBt9h=TU}6H?|cV(n0B7*T_O?4%BF! zQ0()^cgM7e5f%!9KJ^Jj;<(+xqvOR`2l?rP%g-)9x76O*J;8Z z-2!B%!mK;0A!O6lYEx)8(PNhha)K8KnpteVT^rT{68BY;FPOXY7fN%b?_$sXJz&vs zsl&qGOgtuJd2~FT-owiu8S|~zcjwqUs!x?a!r{DT(q+PmxpA|r;NR+~Jf%Jp=x4S< z8`3he4}^6dADlY<>^U#w0T#}CjKl$MuuQ&rBtFJ!PrScAY9PgKY)=|(>`5)fLh)oJ z6}y$rhgxh)J$8fc%3GIU)@amy$*hOk>Soz$=Cth4B#@|>yv`f9nshmW!iaZDgAPr? zUF#bMgm9T;df1uq|#&=(KXJjS+g92y3kxKT4i$V0t;@>HmOUZ0QO;>Lb0$MOU&0&mkI zaMH)>7tc=i!uqGta!>}3t3v=9e8DJf@B6mtmr|>7d9P^fpUFyoi^XmnALrVbqD>UL z0sZdSOuGHeFT5LLnw4pCJ@Qkn6lc??C+r^I@6qF3AJ)WfRay{(N!`(cibL=iY(7Q_ z9`CLG;9u2M;%nbn0#3ahPH&C{jknRz_a*%Z%I~dVKQ3ttQ)uX)N!fTY!0qdR|ahI+60=oXJCrD z8MMT^mWJsJ=<2h8s2&BkYU!_Dxun3(${OgZ96r`~qi8UqxVMLmwbwjvUbByg_{=+Q zY-Bj%r-jl^;HQG+aEynJni*osCtvk8F?V5ku>qvlK-BxR@9l}mLbevvjV+>-c=#OK z1rb$;7YNN|Put_wB59RCKEFX#71FG%2`E;UL{~!BnVjuCXl=sHILvqWS3e zuuw$rg+5#T#pqPawvj@Z-i)RH+Sq zc2X>4Z?|K)Tc#F>8kDUQ&{fHl-(t!kyuk|G&t*5lPfD}Jm4XD(Sy?A(W|8%yNw?H9 z-78_jHL}dJD9Un0f!|i;7GZ19x(U@zbF^{}iRHlX0rP|paPlxWVAj8S`(N$50Ni%1 zkfS{zL$NQk4$4-tftpO8;d@GUXj&%J0rYxrgu1uNsg+YlcpdoBLE(pkN3S2LJ={1L zB6famYQ&x&{C!o`b}?o0D}D`;xk3|@$i{_s%q<5GxE%;pzj)~mSvoV`e~Y$Slr63z z+aWnq0ThWhL`Qr}=A{Mpd6?#AN?q!sJPHDSuzE1#z{aEZa^;?L6MhwL&pGyJ9@mX$ z9S&x(Lcc!s0%=9`F(x6VNtY1QLl#K4Ogg5#{lW+j^bmvN;Mfmw<;d>hHYu9_x%86# zlA+s9EgXx0o%EO9=LJYb*gXU2FzD6MMomvtt-M%RDEO450h!p!=(9{o%x|_(zZoq| zIEaLYU;mK>z=OMUBVmZih05PE$Ny!kry9TZpZ>%>@wyhXbT5{M zVD4+X&nfzf-!5Ug;6AfAz*5B6%XG}dHNIRaUfrcmfs|RVcZDc>8V0q!%o6Wr9Q8MK)8h zmzm<=HR3vTwl*oO2Q!YN-Vj*xvZC!t5TJUA^V zG2{+$4yL@IW1tB6_3!?m|L>pv=gQ!};Od+lXW($^z zj+5&%uq45#!Qz0RkOwh(-0(yKuS z?Ce}9T5wDT;E7{2BBt+&pZeqzxN$PXob}v_FC`y*$BLalwEf{K`NWN}^Mws|&Qt7J zikzlmTcP`_h?p^hGo4whI2|O{49qblB->0(n*7H2BLT2+&F}V%W9svQed?`jy+q$b z@FYyJeHpwjxJ`Z1AI_;$7s=qIUIL;On5S!{yQ0zp56Lp6hahOmgpA7-zq^7C@)7t- z%ZQm6W^eFOwZ04VRkGhbyC z5rNl}jM%{O$DoWCyvF=w$vggPE4{X@)akh z9|4c*4p}DsEV4yYHc7urctM7&Eot;oMUDE33K$MiO&~rblioZD+L6YW`Q}AldBP|^ z=0Q7nXc#zg@DrRnjTcUy^GP`|)xM^|Z38QgDg|@D*$rQ_Rw`2Q7J;h79lLBy#}oG!;?W!O!^j8U zHTl?MgwK++`$m3*15x?n<>ZtXByS=3M+c*HW{MWewp%23$u+@(2#~>9>YoiW5%Omk zH6~^SP@$#PFvej7Z~r&8>Cilk7n z7i5(IWn!GcjT)dFiM%}%gcopq7G!}SS)*?ec9N7pS1rX3{TS``@_njymyg(6Uf)yq zt^D{-!_H0i<=Pw-&^zg@z)bowgWcryVaPv$MXFs8T(p8JLmEBunF*W0bQ~BqmyP%v zUchi)`+M;=b?9l&^dt8{VbG?@H6cJk8rdr>3iu2d?e?lW=%wQiP3rf+=obCe(R5NB zhppM?7;(4`d;mY#{*eCuCSsja3O&+4B}@4gRJt$j2?(SO3WC>DYzjqIQ?ch|=gsOc z*hZ8`4$ZbNW_w1U!BP6QHvsu}c8%K_81H`(mcSQE zn*>l3E7l_$JE*ft4WU;V!`Klp4_u?b2OFah&_4JzZ+iE5+v|BeRW(W+svSs^Zk(PH zXcF8KRR$WTR|fQns{NDXX*7`0!aF1h{g`xu=)zudoMiZzfT?Lz;J^U4R(FJ15%f)R(v6d-?KTd^C5k;yk+W1RCP^;_T!ht^lOca&kS=QjC_n4B z&B+9gr1+4fzQ*bL3ScG&^nMQn%gK<=pt#Un#Y$hi58kCk zgGT0>-(Fw1)AfjYLAn5oYEbmb#NZU9$G4vank#K)Vh>sC#2x}oXF$Y-%XsjS?{D_Q zJLSqxf%1n%(4k3VU~p*>m13V9R5^pD!G4I-+fP6uH=2f6Kpr#U19FbFBFy`4_c^lO zjS;rj24UG03%Q;xR4lroZQ=E*w4iqSe-}vgHBkq({ZqDSzotLcY#NtMav}?%TPNeC z<1b^#3^ZQj^~f`tO6?5Hrgum)Ub-~*$;17Gr~bIZ!$ZpB?mWC>{E*TYowj8-u8bQ8 z%^a>wg?hi@%iz@_43^8okUr2X4T41q=EcI}BtH^^pHZ2-%tOr+NyCE&kE`#Wx|{e- zs}(64ckid9){U)5tBn;oL$M7MsfRiNSSV=Houku1gRWZ94$0xG!gA;(McH`ZSrOhP z6?Cq!Sg=l*!c=N96emF}X7`*{$+elqAT{6Xz1iQ+yMGhSA_?I^TlhSG%i_$ zO=g!AtCdle&tOk!>b8{y9Q37a;n!e+FO9xSCrgn+GSkPX z!DZu2xJ=JUz2&2fnodQM94o8POUkA{3hE?7twoL^)*ge?M`6ed zYtN;A_0N6w%h7I&HI4;lL&W;2M`2hS!d5Y9LA7)oHUGdo&`k_!f#CZxsS5*1kr}mC3izo0iEss8E2lhCJ5eIHO%u{$_gt~nAv-h6%iZ~02ayh-*SC5^~ zyM22|rz%wmUCv-NBt-!O-*q%dmxUk2tPM*F(HASLCw`>75uOD4^)4#G9eUuL!^s1) zF!W7bu$LE3o=Z7Bu_f5*5e+DDx~(k%dZe!tev{^B9>k3kggZ> z2j6)SD_*kcbh&Gzu3-Y_0fNjME&$BHEqDQE%;dS>IXc}6oOzdXR+8MOtY_q?4VVs4 zY%xXlP_a9KWaVq`CY^CyiBF59Q?v}KSf|XYJ%S}YBu|(i<@-aqz|0(L~x7C=`?v!aFcF3z2KGOb2~x*_d-~m5XqU>MeP=# zTjjF3%%y+=R1E-PxUbxR`lH8{ePgWPS-do~26*zJUuW=_dlkj5q(}m$2@@$8kNRTJ)i*Z5}JEAg+4I&7@N{6zG)3JyH)vw20%=YL-WgjglK z)HfRn{414>G4JNW>SQ@>5SD4)-R_f3XHPo@Gz;~<<#gjqz05N4P&p5tkJypZp9l}L zYijeuk=qXY9MR8$S$hJ)M^~!dcM^p3?qH+vfM)PWSM#c%O*Z}A8Z=8n=_dN2- zs{tQOvramHdEZn@w!85n!4Vt8?59{*dn}}4kt#A(fFzVjf)2kpX}PRMSU?{Pj-Rl8 zs>!c_?t$cgmUfE}&%n%T?v!%b(ddD5awa!Vxu$~Ufm!WLs-S}&*x#l`4JkaDp)k3Y zy5E&BaUK+pUdauUA>3mEAj6k$7k0i?w4d4u6860}O{CH1g;bPDi>|Q7A|4GHgvw7MbXVSw_tn z0i^9v8(>uo{Lidnj2c|g!h?xy)#S6XHuaf-1%ym(6RHKkeS5+$hj$4RSp!oQydwhp zB%BTmU5UM+&DEp0a%4}}i>G)0eMy(ygF2aQ;Na-DDO$;C-;ZbEl6IXJx<8ogMS^bg ztcDCU^{OO!61fKTH%HamCZpojwa7&FTI5pc4Q5y~*5ij}xL|qw-JL(WJL_nKeM-e` z189z14^}IH>IeKdB-<%#R$>>xD7myY58kmcwz*_f-td}?#{8sf($T*!BIR;ANr0tO zCP7t*(Pzm_q-rl>^p^PP>+m9IVa470W>|YRG3K*0cukM8Rhd7oNc(O^L}GM}pgr8c zAdhZQ6xR7=3l0P0GKy-tHX3&fWJAvi#iyK^Bc7XoNR0 zJyGXn$7b|-w1;kG@ik7mB=Fb_*N)~BhJeptKjBs`J_YcMa+PeA7NL=e9! zfh;$Rimg}e^M@*?s$dLr%yezIflVL{n)#WacDE7O+;TxxF%2rY)1d?f1EG&3`ead; z!oVO!>s8@ra=Zrf9>r<_8@sMkaNv0boZV%f)V=u)P zQb5LI?h(Nxf3vbGxJ;xUQ2saptdb~y*hK?~!}rhrdWQb17n9`J zPjn{mD`M^|N(tq4l*+-6uX)tNmMMR%WxuTRwA==n>yL6F!UJI#^p*I zMqk#>QK&rr%kTQ5+)#NdU9fSS6&B4eoOzF=c*4@k;3vI}VxbhoNX25IayH!-Suhvc zx;i9@4!!B(htK`whxn4Z?Y0sCmNcO#q_4t|Kf@DwI0+bs1 zPDy@1zXv8#Pz4A%Hby+w8H6uhm>qQ8NN@7P?4Lp_GiF*X$S*#dl}`#h$w?avQc1Dp z6gf!6-hKI!Vj*bzRfD!)lddMJlXT2XnUxy!mv-iu)-PD3yUA%~tGeF1 zea0SOEx)HK4O$)4q}!z~i#!v23JOI|!piRFik@&hZ&yss`1To_CLhu6pS0PpLDxeP z+2ZM6244g5^#k*c&u!GXU`%-p)zF4zh`qxfNM1wbz6vY*bH8`Wo>!1pXPWz#z;Vzc zVvoFmx%}pKI`yCK{%GmEkN;uit7Y?YUp8vGNp8R<x&XV5qKqHL9lh)D z&(pr4@v_&MdNi-%zTk#}`@#NH7R6>zB%O-g;lFO$?HM(KWNEzT!|f%Wwe-E&?Gy4u z&}k}LqWkh2mznF+@0e9oH%iw}T`1b)x5z6^o*ps6(BKD$!Paf~L;kosaGSiEa;f*z z&NFz2WYfDu7i9V?e%BP~;$mgl#Fd~nRvE2dBP|!jN!w-F5E7RQpMM_WB&q z_WJbtEYN*2afA)QkfY`hCWgFnWOwqz#B(41zkjnYy?=CSao@0vtH2&vF|hkeze8~V z*n9pj00P7OWQ>d~7P#)tjgcu{wHZ%)a`pD>mlV6^T7s-g!d=W?$|j;X;azC=OBdt` zQU#~H%fdRu>%5ETPSK{|ZS*G1jZkE8s*B0@T_;@g4U;+@I5z4f>x3rtsps1#oF=Qp zn?ut@r9Q)Dj{yC9}^(xt#+Es}8QcZ=hv?iQ~UE|lJmnx7>2!n;h|7;^>~D$=Gb zlw#t~tkh%FbU`0-uTLFJRn5xd@E4n1hbUXLIftW%S8*e1=ZAm&e?lvIJc2wHlK5w)pc*2bK}mZ9#jc^q zDk>J`5|O^i!gH%PFk6HPK(*34yF%L^TsGPcjpK-UIGUrewdMOcW9$j_-L~F@gXzDG zE|(;-_$|MwLUNpXpwZb&3IwR=u44s(p#tqi(Ak zujO$Nqo>I$gS!}gn!HZEFSOrd-J~>m0@RXJNH54)^bH)T|*Y8`Q4A5IZZLY^y+ zbEBp1Hw*shX)lmRsfA``{L~F0Kpfho?)B}EjIx!mjFls{!j+@y=)Jf{e+XMKW=Y+R zqd&1a7e6_;>2-3>jnf`IHtxa~6x&7-6BU~h*bZxU$)Ys+3Zpj$ZUC(>a4Xv)HwPaJ zF$U@jG;C7~4biEs`SLa-uH}JE%2k4-q|* z&>FR$frk6)=yDwcxet_e{p4`Yu)CIA}0kM7GjM%>)9emR>IW1+dHK^thhzEU~^u*-gH9 zp(lJ$0?!e!Jv25WzV!f$yukLH$BMZhJnbG!g{080NoN8D(ktP$^oEc+33in3P#h<# z0(VB9B%^GL90$gOF&mW)UcjJcy!4wJ_8!YKvcSjT%UIe$yJwsVt_?D3jF7nZ0wfW+!IelL<)=cj~?pZ}614wX7_V_$ZM&78HF zVpA!y9ym_Gi@6zo0mNLeI!%w({|dcvc6N}7&8CsHgk#aKah_bO~+&%)aiOkG*hX&x4~P9z|92C!K&+EbOdlrMFCKku)niqcI*H zb*tehh(-`EL*80eyJwnxq{;^tZd`80k%?dCc~rVXd0rbAS|~pR5$+< z$%am=QHF-2z!>cb<$P4|x1M&6&WBNjzFC=|0%d{*&3*r~{@7eK%ERnH5y)u984qm% zzgc8lXGFQZJRFDOTE(25!n~**(b&4y;M*lcs-OPg&!Rz=M3^LMku<8hK}^IXXjH-a zxoT%%F?3{)%Ah#@jT_xbo!3aXuice)@46FUO=R3>k0%i0QfCB+98yjeGCkLJE;VE@*rf&E)Z3%NfUZm4s3csM_d)D|}$hDgIuiQf=MsDk9f~Ib#v&FQB8*P!p2AeVTk+k!^V>O z6=Ae>!s&iH@Hkoc%#@9#+e}4kD0US^RsyAjq)v@0U07L|qtN35Yc)BbO&m8W{*%ps zIS3(}YewUZc*Vh3xfK*!BIV7Ajm zy4&>HvAZLxXWaE~(%lf=CTmCnv(!JIq-&~NU6af)P7V(FV5frPPOja}&uSQ2?_ID5 z2*)^NYFnX74;iDXCMJur>FYBLWAmsyszrir{ZPTDz06<-H71%i`{hyZ4ytEx?Vo6K zHz&qnRFcxX#=(6TSbh-mt2OUgElBzu&m_ETpTZnBC{Cyrfg7cJp9=M|0=TWCFk`+Z|Pcv#;m4^s6o3}_7NDC$mP(POd z^|Ogt9aTs7dKX0*nB~5!!V-bbSzjlMlipNiO3y=wpAkNx&a2LAr>F#|eJezZq$@(8 ziU?&Tw`p^wi&aLS^oS%;lHj0Mj`S3LhwPeY66ou^P^l|>T9^8>kYo|oP#^Km3N(zZ zS7C}bSK329^?^wdCEq_592Oy4a7OULsRkxVj=u%ag=)43e zMr!-yKWDCd+G`k%vkOC!fks~#luUA^DS@M(cr)8fN5RI=N~(b-C|u++N=>KehiWYUL)SLQ;G464m$0w=&}-`f+f&b_3# zr?{gA+YqueB2JnW{ADl_UG;n5d<*PWBROb?imydfyHJ>X-(O!JtNBZeJ@$s6}82?Uh~I+Jq9LUQSZBL&cJ~bvQAi}yAOJb zC`q%)q9$ai^bo;ys5(`)xJ(=;?T*^yn>M*g zhefJo;zgeKzga@v{n7H5H+^rhXPv5+E@ArA-K0{}q-zK2iCQSV?^Nya!aLT|yZu~^ z4=0c?BaFc*A(nhvm?F6eoq!VibcaotX*sdFI1z8%9r?^6OT6I`U_v!l}x}(|JHtz8}=! z7OUc);5G~4cwsJMlU3n6sPEubQ3TIu9qJSE}OHR zMs1Yq)9Y06o(*B;vL&8rps3!Y>!d&a*5?q3;fl7a=RC*jVuqgT9_$@tGhMRzsn`YrITQ=p zIvG@K1-S!>+_=y*`o7|CY-U7z(2;;75Z8%kuTF2!p~Mbq@E!C@2|NjPFOZc9Fa#yC zHOjAKTSzNJQ+TdRct#(Ef_0A*hZtQ%VNmNt1A`Q8Q1FyZ7su2^jCh*PX><&K z@^Db*a|oHM>wPwDmB%@9g2g&$5;Kdf8!_i^?6x7vRj z{ItunRR!a!KW0w!xTcQ^o~WC(aYDIIRdBsZ->EQw`+>WW-gCr<+A>7^2i(;|JM{gy zEyfV}*)e~OmA_-}kGgG>D2MNvMyF^SG_Bx$;;TVerh$YMunRORFUTOpmPVh2WJjyA z62g&DD>!&I&~a?|W;sb}l#Qxh21hc}uYE=@ zvuk!UdpELLv`&6o*fHUxs!6vo@Z;CZV!MPF#OQe}6?M}4Ub&=bkyO8ZYwn-g<}Lci zhPf9NEyQxd#poUagj+9{iN6B&f^0$Yl&#~FMH_-QYI>QpAY6dNb6o%s7ouVS9q{TT z6b4`2r?WnSe zdj$n_sA=Q2Yl+v&9(-Jfm9pMR=LGPO>|?LLEQN1>-U3 z5x}OaM#uK9!+*zbblkR>B`%rzi+#%>pTNkC18I(CLr$J zZ5$4HjvHrdImmO`VOGNk3QD$t6iR<7tV-UjY!ZxCVs_8~IEaZsSB%c)x?`(Chr@*7 z#zusL31eT(38;KKG2xy9q9>zvsi<656#`o4%tC1=Db+3$k1jeK1cuo?868MiKILGJ z;`S74Nc8(%)vHK`8@G~G+5|a86bmKS`BZE#y=_jV78T5m8q_ey+782v6+Rdr-vA*& zZ1O15EtBf={L=%|17U8c^U6@%@WnJOCg9eFWkQpozQC_V0`WDH_?6Skq$Z|pVw&95 z5jGzj4DwZn+H@&QcsWaBzI9yQ|wWS)KIZdm*~01_eS_0 z#Yf7$^6lEphyrb@b{RAUFZJByWhA%8Hi`Er?nGPm)T_$F+UR#oOslfO*Rtzlk|f_a zA#EbcL|XQ=O3F01nAXT#X_Id4*ej}RI#t^xT;`b;loo{DI9=hz%xyW8HO^?493glH z)F_c_&PO6H9JHYZ(&^u3y2UT;f9b}&P5+SmPLu8qv=8B)Wj|nw4)|+eyuM*wz zjuUmzN603xI@v96{FxWkq^qUR2{FTPOSH-B%B&_G5*y`27DA-;H`VnS1HVC`pbF?d zFgpQ^4w3`WS;9@hOFWN&mJO;|p7g&9%nwbvs>z$Y{uZ7Xjr)w6rJkLN69mu3za~jg z8L-R~bf{z*!ak3<(2apd$VT}A`UFXaPu}E<2zL^LLz-5X?S$W}a(y~kF~xArZl8gogdlnktBE9V&CmvZ~eud3r=KJ&e_Ih)V8Vfhh>u{*b zWly{3y>r~TIfrCBWk$^w?*o%JOzww8lB97(<6IrAKlw}j^jV8P{!aCa-&viQ{Xg|l zksHs96O&-$#N4G=P#XFg2&GiZe6K)1EY?(DeI7P1=~1FCDGU`m4B~9)HtP4draBPu z4t5S+J$$~n1CA(XI83}z`_&L~>D(IjeUY!IYt*De27n-X8rEln6Ssitk zZc~@U7&VCJeve#vhu5yzcgb;9-{70Tv}u>j)MsjMsxA=A@7O!G-=jd?3bY}hMRbAe zkX{r#{5{;(g_&|>%9t1{`nvqPJDtI{&T zH7gCFR~abiSwHn?us+$V6R7@smHi&+qGcp6su!eH(XMTbIO(tNAtf(H?Nts5u$8bY=>Pkt|M~O(`i)`<#V(>q+~5bt@yCwrZhnr1+b)zZf9cCl?5n%n)(_^W z*V;cZX*~XG(lw~|PALsJDaUOCQvtJ2pU0QX`fn^F7a{mAp_2qj<8KBv>AoT-{ci@< z3-FB3Ba+6$QpaSY2KP4Uu0^gPCI*ihHEYOO8Q$5o(08GAa4E)lL7MvTPg(ET%TsYE zJ{=0_mRf9xeqGdV!KTUSUJIqY-VnRsB6O!>E0A{?wTZ0$qT(aAz%N(o+UL#jkenF? zj$65OH?LuE-?{y@thW~qN7&AN|1!cMY}cS^r;#@Ve3V`vbW&UBOrW_(8O^rv9wn+~ zjtZ~`(8vei1J{hk>%T4fuj*H<{%rna!7Wn3&!2VUWZgv@XR3~3Kc&bqDmGnILNAcE zNGyx)N6A8xMjr`i_gUjrIIB-xgp8rUs3-t3qKK5h_ZPsT``ljPy%*C(xYE8kFeBoC z7SpV$g2tGgk(qQNn;rqVo&!PVqz^VioQ1HySiadWK?Olt)#M8@Y=ppMbC<9QE;4Fw zNA-D}o6{HrS4`>!@z#~%b~?u|k1yXiA4m;$(fDuk1LCIJcXm1G*lwpg)#;&ME1Ld4 z_TB}qsr1Yr_lPGXFNWN>Bu9cM5d?AMVyJ+Fb7?!hZ1=ab?Y7(9cKdHTwA*Q?-D#WJ z&UCk+qNso{f)~(m5r`myh@vQ06>+?RqB4Vc!5}&gGYZNusPKQDBr1uC=0L(8-F7~e zlXKql2G09E&-1*`<@+$5ZhFN7K`Ps4!0tD*Iogg#Y4cuZUSOquxce&Z_}MoWPd5pf z?`-?|ugE5A1^L|l{mDKvQBq1VQ29~>(It4f^@=odhkuG3fq%6stasAv4>~DbF~3Qw z!DtdlheCCsPgd+Prd^e(sNfxT)v4Qgu-jGKQ~*aJ=K3_KQ-MLoNf}*sVEfq;!3iL@ zjqQK`-i|+8=6`Z_A7mq=vp{Xo2m1mu6$zdxOoAtR=~8(UYI7JD&A#<_!U8{IO55vi zEgEp(tMd<9VuDSI4Tpt`4FnlEo^Q+p@8}kf3f@(p%y}O#oTv`GW^2b)1o@VIb+#o5 z(324&?xu~|qWURRCwdLKcMG>jQ+b~Xk@2NQm?Ldf>3ti>Tg^Ua{Zo1IA$w%oqTZS? zabh0?9_tY^{5bK0Z6P8J>fQFd_oPXPZ3$f#Qzm}+vpx2%ykAi$+Tzg^3EKy^VV6(j z*%Q>0+qj6VaNxvsx18EG4y-5>>ev1`0!tE1Hp+yvcHx#gOvsFbHd9}PnrLiDxY2yca2zzQF1NI0HAryXlC zp43#krXTa57X&5W_I0g$5i=YyFSSFq$u(j6+fBixfNx`(+b!1$pUw#`L}P>DF~f08 zH`>PVYm;mCDNGKq_36PMk#!s{QUl_RQOs~f6a$HDxerS|wtB?J9FZfrPyK@2sRrfO zK&MCP6$w;)Os3)v(iC?Y@?UeIupdu?GaPc37B60$^!oKh&QpbBP=6Dm3T=bJie)2u*7aqhAzL@7w$&ab5<@~PrqEZ@1rTvbax0}CW)B&BV-u$Hj4$(805T>q z!15+m0Ga&8e=dl1Gf9&7H0S<6k~l0$fVOs&B-ubQkU#VRxXpvGV8Re*!-I$H>zX3q zJTFi;LKkPe<~&jH!~(H{#uK+Y?F2h$q-3uMwM+uxtn9Fnosz9eRJJX6cVJdUR0H|F z@e`&G710U6UyI$1^mN;T%|SQj?UJXF^AqGxXNM1iT5%dCMo-KNAJlg7+aJhGqCb_2 z@;ZK7?3%B&|8OH!{AS`S&ZWnGw02_v)R=3mAAPAE2Ym~F@Bu#O=89N5lm}AKJ$>G9}mAs=o!{Ry_lQrUg zZ{TCHp2HiN-DdWMj$(kbDUXUv6tsG!3Azn+;x*vXtqIob47y5VETA(~%NQ6ix@h#B zZKGF4X;pX;LIiE9HWk*!lt{W|PNl>cA5t6+kVg};LdxXO^qE0HCZx>C`CB?E;xJO6 zW@^-yw~AsaDd?+f;!DA*I$U_9dHXC(Qz^dagE69$UI+P=k%$`LZc26o?xxqZsz%w_ zXah~a1!($p=v7TI^c`ypT^ED*W-2s$WQY8)SEdf?`>>fdR%tno5Nv$Fc~aUuTKRDv zumaAcZ~R&M;HM_w-2bdhMlQ13@NhR6Meduy>@$kFNs;SR9C})-lvq=Vt=)I~X)uXM z2aPdz%FBS=0<)T-YaBHD#*)yD!7W0q>Xz^3nLB~1z|fd36=+uKLhH$=@*DG@A{jg} zkiMAVf7@LLGH%4B)w7h>p%?--icH0jM=#wO+9kwUK9%1Bo|IdpIG|o|N;%@vWZ{CZ~3m*vZ`?`hrjgTt|Z@?8&%)8}#HWZ^>qyKAH4MIa2WD**E)|&owG(4XKtQ}{p zKsoYH3;u1M32^V$#x{~tc7WsVCIAxWN6WPuim9eZ6%~iA5_9N#1UA&>sBx`EbWuf%HUO+_(p5RC9f zUZJ|3fs}d>hucK?jBZXHA3HJD`jtpZV8dw*5LVXWDW{wSpQn7IeV4lR9@&;Xec7a? zaM(d}&?Xl|u^zsefcl>hwT4WXUiICceqF3_i zjlpa9l@Z-C9BmXhhu!x;vh^`?W8N?r=a;2@^i}A9Shnb@;{CgFBVSGeI9Wi zRMp+E1n(54On%x+&pQyTRkenKvO}W^UM0bn!UFkK-_8siN#fi}i6o_W6F`^Bz1vMipeDxEX@o55JW4xo1GL zYXdFe{-6rJCXaj|NSqDb%?`VE%dpVPaR#&XhiW)GrWJq6e2%)r%9ftIZB=5$|1Hy| zMqCflqD8(mutmN$utT;x=;JxYoXK3MM#u9ydads&pxH%^X5@Gucxgwt4!U0<#@cwB zz4xoorPw0B4aGfuK6M_j@yGm#OvQ1m5Cl_w#O12bo%y=y%RI*^RUTntOklH=B%?n+ znt6N&Z{&XaN0xyFHY$q4ma&ca$l-`~xv2=q$dPkcPzwxUj#_9I%V8fTSWM`kp~^p> z>3k7V8S2D_REAt$Zn$y5D|j6Y?pRc;?`0#&k_{<(!S3u5 zF0Ko`6WJ}#l5O@GQher_t>~Mzhrdg_k$~SAskP8ovYRvroflPQ2bPDR8THeSg&)fh zc3`>pRi2wAedG(GTWpX`Vg<+N5r`-B1-E);MAv~g4Ywx~C}~U!YbS=r0)uhgvvTj* zuUT3gPZkG0e|81d?qJi+41c|k@D zE;+$rwE`afQIVuhiusr#=c%}L{G{Mw)lyM~WQcyZBp@kRFTVp$_TP^_pQ$)Quj3bk zi!713Hn%5x*aJ*4O#&_WF3?ed$MU_5yiV=PQ?>Nc9}K{EzLi_;Wyf zUwYZ+;pmh%7dI`ktLWV>ER9k;jyZfb9&KaEO7=F@z3KZ&Gfntf_O5R-$$4S?|Fvdl zETC_lFI+CkBrM7_4CkR7PG= z7>wi1UKKn{y@B%>34m;Iv8F-R?bo0_E_S#Ya}>UwjTm+#c3xc+X-UjwgV72d2FNMh zhTU|r8VAca6KAgt1s4}~4Lc-to}p;%k7D&mtxe#hbF5xj?&^POLjBqokuwMF7o;fu zP_V=!N0J-RIfLhA)6~T)3>TZ9^A11Kzf=T{k;jYDss%W)sBW*dc7EM;SsJDF%uf)>3iU;AM!v z8)AQ_qmf&*i^NZFk3A_p!_;|n2`@}fn0{?;&)8dLjbZs%p=J#8kM)+;p4Y&nIDdxxSzI#gB4qKlEUS7$5%|Y zYR5Nj6p>xlIN3NX7EYUqg+mkr%@ymYxD))#!ZOTB-fHe2n3mJNmTl+)t{;ZvZsX!W0cOfYKu zY1P|gBZs%td(0qHOfgWPxP^+Vb!+v^kqpa|U5g-YoI^wQ5gH@u9$(0{Dg$c+<}hgv z1m#G&70sjv(wCr3f|0mSuV@wKV@FPm(;2a_5N|oO41^|LfV7#UnOqG>ob7nTkR87NMTkD?SKbQA7G^wtt@?H=|g(N!IFl zS*oLTkl#@Pg`#^ryJhA4209nCK}xujpEm19JnD6H&cij%%1X3zH-qqBwtergC2i9S zqH`+Dr}75KM_}q+_bld}l=gXQ{ZqmZ z1eMIV=u^zwz(;jd>@_HJzEl|_*qw=U!`wGqzE@%CY_q}nY?LR!?2hEl8U9E~P$^Df znq;4gH3$4gT)Jue7oD|ThNCuUb&?=DDz5`loF;jSH;V7fsaLF}N+J6Dhc5Z$j%gFMB>OFeqS9)#Tm+jKEjb2winJbbxXV)o%} zPXEaYvyIzT3|Ldc|_&Hv~(;SUk= z7tLl;&TePJU0aJlgOyP`n|g|=p-43qS8eD>iCM6#yzt+6QB&~6fI2!=&=T4S<%#tR zhTMq{9QBd%o+*Gm843# z=~DsNnQ(aiSteD|BCix<*T6f_&f76y92?$7DbXhn9RkLiJ&QkbF(Ih*%U``u;yEm1 zv}Tw{qnH$mtO9BtLqB%3o;pI0kpl|N=84UdeY6mEE%Kw$6L-K)sQ7jK=qJxR!6bIM z+%fxfga%Q8RazgC56K<(JMP!`?Ia}}vuO|V3&G_c9|M&Xr^Q?N-Sh?0ElUIQrH;;Z zb#g~9yMXfK8*ez{OxUmkO8bc=lqH+b3$h=!K}T6Z4=c7wv;B{eCdt~VE2pND&*)nu z&FgT)f#CMoE2NL!CG8Ow((xhL{@Vi%MmR=yJafg_!|9n*I%GiU?Lz%8P0Ho)cRuqc z`WMDLSz^{+`wqq2rpRYh+-bi=WtI@bIcSQ2cd(6jZ9%$XBUFE_oH|IL)=RuLG?k}w z)2d2lmqjcQ6-%Ll-FR11WFM56HbEoJTqylQhk&Lkx?7gvk3N+{(i8@!LsNPKzcvD= zfw_&>&<--*h{@XtyrZvy#~;(X^w1*eh-zS-=HT=MQC(;SZ^Q*_3sY08goT5hS?$R+M3azUf8(N^&z7Suu@bh+t;#ecAH7+}QJ2mBa|KdR*U47Nnq zJX|=3yxdQnge5edI+r6aE|nJi<#{g#8lHWDS1C5qA7?6Vdg&D{@;WibmSOR)KryA*RWqvGrvS~o~%)3hilMZt0@XcvUQ9|*ZS2))Hgz7W^ikwD5W`8)WlrY-Z`?3D{XaUG4sR>nUC ztMtvhlt^?m*f-(5*#73+M48ftRbT~-$Lr?wBd>*fwoWqvBjBCdx5$PU z1{hUlz}QMLP`Qv##T5fLCa{6Y&MHg&n&|_iA-E}6&(kCd3Z<={y@q1xE|6F_NOOqj z4F}ukW~d{1YuQU%qu&BRb&4~9uFP4c4=boxlMPM+&DX!b_#Z9%8M3i~IqVtUY6g`o zidj#Q45RL9_3RQNF*j%}^=`5j3R(3EOr}HIN3S?Wu7nOdn#k#C>&FfpPn*fH*H#Vc zzdoJ>e-0}c8wvh;#m-l9{a{n9R~%Pmd1HM%Y>tp;{FHjR`%N#@Iup<_Yt%Bb!-kVG z<3{~H4-elfU--|KiItp9@VAjFiRoCkWHo_pUiZlEpbc{d-A_QKBqmw3xTcYNBu{+^ z!qaGBCeO=Ot(UBd-5F9cJ5N0_KTUw+wW{rI8ER*#41CfoSKnN5JX|U#{A9I3qBb>s zHrK;skNm8YyhYO4*&`gT2q-sGGIFb_)MdOV6nRVjQNF-`*^U`}wKy%$OBFYIEGM_T~$1kgXWx`1Cucu8T z*Ek$tTxPZ{9-x?hirk{&GWfaSBSAp*cbs=641H~FDrB_94gjT8DK!+6!PlI1JuvGk zTtwy!t!nMmPPgsemFm*SAwH7-Rrnm3g{&=lASNxMukq{X90~4dHScLo1rI3>%<2^D z6_v_LNbT18KTrZ)a~n{X*3szD?Gmn+;2>z-JpXL;KxAq7(vT~Pc#)=*UN2cc`$!O; z?4lo?E`(jPf$WqnQs0=jlXsDKHhRVMLGMbr23@?kDK~z{h|BvL{AMFX?rPDg;4cl< z(HNm*E~Fg&2~fpKI8APRtuxq?w1bTW%wZ$P22EIfa0Y)3)Q@!2SlWuo-{7PPcTQES z`Acs3$vy;G-sSwyRo}f4`n+dUZFgHYbA_m$^g(pHeQRze7A$6P&PP^KgPM0%dg6KKE?uE+e`$$m6*)?4 zfK-xP4ERh5Ne;bquAn-4smCCxj)uJtay=EQivtD%?IKkf|JsZJ9}}|cVTF%Lzq)ie z#IpYR1<@)ts->|7XNd@yLRvi^5+yjgMTEWJvKK=3&gZCcFMs3a_bnavPu47Q*rjH} zEnmMNRZuBLPHTwZ93t203VD~u-VlQ@>45$UJ7z#D-iQBl=62b*Q1&jrnT~N*QQ;`; z*p?ti#lJlCyoa0PMHSF<_mZ?N_KQVVAYuD#Y=t~IHdmSo8Bd3l*TLg@QkDqi?`^T` zd3Pc|mF|+_=gT3t0gkVA0Qsm&0_2n-WWMy;fS4{Za(3X1|Y+_QdUwRNRna`Lrz_SSpFBE1Hc#=Y;u8x1SDz%k9iW?_2B#k%Lk4jOn~r z{5ac^){Cvup>HT*@|vI?3vK7K#Kk4?#sd}_|8`!qn@+U}>ob!L}k6JE|D!$w=(96Be&&>89isSAa!x}Fu1%DMfK8>d0&j@Ri1Mko3* zkev%6yqMA0?o|N3d*`Et4Mav69HSP~Iftxt%4CUfrMG1mn6nVE5!gMYY!Pl36hS|g z)R;?GM@Z> z4tq##Gn>R{DMcX=jy@nF>4|?`Ft4>X2>WS4Vfjp4Yi~BWA0|4o*gC zV>W@4KFkh{2(fgAvC&W*-tlfTvq(2nOcq7fQ*jvJ$uzWSLOR(h<*Eg}bdsCqD0o!+ z71&D-z3X7X^n|7IR)}s+P%_CF@Zlsxm`^MD*;n*)OtHQ9-q+WVtuIXc_plj0_EHQ` z!tA8t&cg=yzppCZy9@4hOp~qTmC;51t3A%sm~UF)lP|+Ya_zA#p3O8~8H%ao74z_V zGo9zYZx$4Q2=_&mhSxGU9$fUY?g%_{XI3NcG^EFBRi*rPr30;-F+gXOHDh^bzWn3) zbgAz37=IIVN=JTBMRM6ehr^xI}c1*0c(t zLOl{26m~~w@CuG<6?O@se|XFl1(b84d$vcO=IXr6S6I>UA;kfZurbA%jel4oG6|P= zE2sXMB(oDP9L`HBF%vGEC}t!6vT^-jvuIV~JE%Bd*NpxNvRTFe4;!oL@Y8z(tEXFf)NDAebi9j1Pq)Wrf?Q63`YNmr*VAna zHV569w@aQz&hs3otL^whS-f#}%;vZu>^7c&SmO4FxL>{#_uD`IF@@cnvpJA6AH+~v@-&VSTF5w^I zZ;LAC9VVC#`PPo9W&B~6Rlt8-#@`Us7t}!Ch34G#KFz#gmyekQZzP*coYLpHYToiG zS{3#({LuT0C`?ZQ;uT(j`mXwZOt0^%ShRun$hL}xT`tS-1J5$*v3Sv@pf4BY(@R5I zU(KgmJR1bpBed!a^^T|=LHNfk4Z(YNNgqH0%;2mmVR>&L#ZbJcR^98{9(z~*sqhyp zUKvYs*QuXvzcN#1DHd6#)jn6@WE;t|4RnhzmDeZD49CpL%E&hG!|d_I?8A5nF{2}9 zlkZ`2PTU}e2y-_OX%2bxs?LQ@Kp<$0782SB?S+pfaY&>?%^})6VY17Jp z`n4Oj68ImM^y?vubO5@@;?<3_M#AtcvKHW&iO_3fdh8B)uuq@F_G7max8?4(;+2-2 zapTE= zkgaDv z5}Rs6_s-lqV|{3w=XS4LU`@KCTn|M-sk{y3FN?EZ{p|#RbQIiJ;C0kp&Oc;@SL%;` z^B-7tW`3?VMYj4clRDt?>4lOV*MZMQt3tk=I{G~Qg=eBN4S0DBh7fxHq5nq?eO%QgL@pjp z{QRDXUEmtg(MzCZB6KA(e#}&~2$zB9ScB806AUBZX3&wP@b-f035{v%go6P0r)kIE zw%i?YR?pbj?-;XG(EjaLXv)1?g^9B_2LCSz4tqmE4q7@E)7TarRt(h1v-Y$V@l6J* zOQ6dVlE~rR(H1k?Ba>o2pvW334r9l;Q#I{SqHO4ajQgMqzzWw2+<>@)`MdFf@L`9G zNt5dWeNDLNUKnTZMlq0ASxChVfT2+bR;UK)L@}Wjw!H1i+mZ02Q?UxV z)0sZx(Dxuud&TsA{%v;+Zm+sQ%d8D;lA=>Z3p?G+c_ktz$rwlI3Kqh_?xF0^vgz0B ze{G2tHi-({-J1ae|LFbBBWyv#b&xJdeK6<_Y};CuxxePYG(0F*X8P))og|>fk1f05!fIetXVWh%lPo!FhuBD)e8}JT&^NYf z0Veb7=ImG3`k~m!k@U~%bSsf`K(9+@q%2u~;H+2N`ZNwd%TAiKyx)J8GU=WN9uH2E zVHb82a5mZwe!kJtB$Z-TQzVIs>yTB#P6XQtf&UaYGK-xg5}v-Y_Q2=qGdSn!uY>k= z&oTjG;76XzNxrot1r95b2D4?~Nih((EHx7Nod%Ku*wv~HUVOAx3)Q_4v_#E<25pCI zvyY~Z-$m=7VS63FP@N2<KKRX709yVaQN5*4O`pS z(dQiv5Zi*`$fVC_S%QX*O~T=nI2&+U!t{h3mi5sGUb!Ug5^DY1LQ4gSQ{qLf;%<*J zx`~%2T0RB38^+X3Un(dSVXUYlbZ1C=EIL?w*G%(wtlf_E94L9 z+?IG``Riuf=H&#gm{J~nF<>jLeIwzuq$zkUhaMmsRfY7y>G}Q-6iYl3r)+n7z|WOj zl6Jp|xl0+6YA0Ez4iG4gUw)X;wm`AsKA}EId@~@zgcq^vqAg?xhw*aK3@;57Q%jM3 zRNUbi6+HB-8c6i9vj!@d!N_Jwt1qU1U&CNKZZV+MPu;ff{VbQKc-G*Frk~ zN#AS2LP$~5s@D1DN$|>EX@_gCXH|%S;TmbEO0#7#$%653s2ungLAdc+N#;g2lzb@Y(?RR!!RwdzaK5=oU(3#Z>Uc)o$3Uaz?C z^>(Ekn;yW~yKV`hArJge$FC>%y|n%tl}T<+%P$-Eia6z!X8BTc*N%MYoj>pUtx4MK z{fUPsxydfepSyn(l*^+H@1g=^{; zKo4R6A$LQakXBXUvug%YIOR(8P{7y%Zwn5<0MFwXk|u>&J|FWkjSU?}Tyn!#DjxPi z+#=P3$SRf^IK!%x*f^-gRZ}8?vRUOt=@L;pbQLP3*ZDRtI2P0*Um9r4(MY1OETu}B z>;|3B)tE73JP!>@aBz#fDkM`;AADz)rXS4s;(%_5OAUGLhF;TH>$GlWj^wNo&3eve zzBD(5x`mq1w=%PD2ig9@pzoNOysf1eXkWdDio+mNrlLlLHAx!u_#hn^;G)&DLpDfG zc=SM^pK@5f7C61CV(=A@;amt7nPiz6?T>n9fuVEyDM|rkhbBIuMMzxQBP) zvbTDsN-z+K6JrS65t4b_LfEQO9KU2v{$cH1JBv{A$Oi%p>}h&ESHiP~HW`)T3%ql_ zl?n}#e3r{kzBGRGfyH0M?vb*Xc$`DoSsI*mi)DN}zUz5!$=d3Pp{BmTogueeOJ?in z8}l-!I$`7dY{1xq)o0J*_z|qII;r7@>txeS%H})Ue*P=6X$tIBMmtYSDW;f$_Hc1s z!ks?7igc*2+3r^Dk-|V7ptylP1=?S$#?}W~RRZsvY%ASI^gMKSLC=mhAr6|-kmtjb-}Nyo2;aYk0F!d;hG*{}cKF3XS-8)qMfEn^!2C}YZ; z#%R)3OS=8gNH!$X;V!Wn2n-WMgQR{=A&nHto&HWkwizhj6Wj~fsBmqMZ3<2b*2~ugXjO?`55juI z%LF^+cfeL%FWDH}BD_BT6lwA7rLjiw2;B;WY`tQqU}y}Gc|uS+12#49zy5EQ6t|p> zQrn=j#T=+oUbtEjj^0oq61;Db310m+l3Jo@?s@DW;BS9<_tW}Jy=qaAMs<^cXgGXKB-O}Wkt}sltt`V0m!t(XY(iU;LcOSh6 zM*j7xq7DWQ#SD5oMJt#snLW$?giaW@8{h0#u=;_O+@RihZ*^vjX-gBZboDl}*IE_> zhb@hBX42sZ#WYdm02NoxZ=hF9xlJ!2pMNE3N_*@{x{UOMtn*ErT|DdKfFAW}x=uVu z+GCds`uRC@wrCsu`B%W}#jBZ*SGB4l=9IEM7OSz-!#Bw)=Pm;tgWW+N&$;d16LKu* zM966v58Mgrs%WH6-Jtx!cLi`Od>oJj@yQ0kQqdK1)n{B793#A}AnwUiR&itJN!##R z?f+)X^GB;l=DWsMa zLY!4bHfny_!i#;1_&3Of;6Ack@S!M`cPKnhvR_qA?Uv?(hiADU%?q#VsKbj|-t7B# zt?CHfEIlPTEUKn*MR}6LGcGC)kVZi-{e^rm@{|M_2Zt0r!a;Yf>I1KvVP8Z+iLsrp zph3$K6W7<4x6^^X3Fa*#WYjo5LVWo3+X{%Gy^Yn z`{nzW#~3>dX;sa0jRpAFXwJy_j$#gl+K9_d0j5nrZ&v70XYB88>;#ZIb?k!9k%V+{tFS-v7SZ~biP!LV zK;__h9yS8GEWNGxkn};K-DzGwslyVp+N6m0jLorYvQbom~v`FW7iZz{~w#Pbk>dg@ zL)w)q70bdwJV>^9K)cG&Rs#>l&WKhTyfP>!azd~x#C1>TGQ{r#J*C+xsR`T%gs0Gk zvw&P7#Zc9nLw`BHObkKLv5Skx3q5A|Sd|iwPCovEF&?pk&*Z6Jj~E$`>z~6q%ZBS8 zOWAZm#eA)5Svb(Cpd!L<-iBros91))ZV5F#bhpgd64Mb-v1292f+jnr9ruu*5B}0; zJbB<84h`AJ13&vJnlJG&+ob!0Zh3#?1JbBh>MX$w=$w} zMu%*P=V}iSLpd~58yM>dK2cWzQ61*5?2AB4`%YwAsN<0|7SKErG@}p34HNSTK7IM8 z(-kJ2v_3ufBeIU2Ai~}gA-FMm!= z9ZyFQ4g26uHhFntYtwGkHZU!pl7U5N1gM2js=41YYx#!6czmeC-ySAV`_8Oq($x-D3 zyz1yxf4nT?T>r-tIrxaxCWyMCy6yVB2L$d$uJK6xq1qJvF33y1f|Hrgw8(`j?6qi+deUyOt;nTp-=Wu9w8^}H%&n+nKRRZ#15j5w#& zkR`ZHpIKsjtS)&Q-){T#^60xKEMtjJa!_)36JsO3Sg5{7=aV7Du>7PnHS}s&GoT>b ze+$2M=54`s&r@pj=%dVQrj6%x(B-U`J|@@#JDdl~LxL?H7y&pW#Ue|r@7X5Jf)SV_ z37ew8&*_VzWDS|Un+-Pwy5vSte2rru3fcpR-s=@s3wC~&PVp= zrr^6_nk^6+*atkS`-3X@P4X7`Y5H~uQh<*MeU3!glejaUVUMv%lfL}k;v<%&V{CL3 zhp}X%4HE{Gu$r(sB46Dp-WS#7(aP6!$n=oa&=3qYuDrvpe^1>7DE&*|PrmaTla#p; zy=FGKF@>x!t0n%DVg@MEPsJhAPM6TgmV;hzL%am3cTT8=Or1`bVFT{9iUiS$@Lsxz z)~x3xh_F#1mTI=~()n8dEZGn?n;TMO3G=7sPepp!R@ZB8nk14S$^s7BUU~>#zCA&G zGzRQUqYuGux84UYV0WM^;%nf8Z}Za0>O${DqzTRk6}dLhgR&OaUV7EET1Im(Y$cSP zuJ%|Ng((7>W=S7p?B5jJ4mlwNK~kWu3q3|!g*niSGhTF54UxvW(40VB;2rl77d?a? zvGjO<&~~EXY$dU6rC}wXiM#e?=b{_4q$cdi|Mq$tS@XhZxb0>dE{9@(xNIX9L}L}S zp?u&7*%^Xmx*9CEHxx(jd#S?bq*SB#SrUll_mxCbB7#oB9kP89V^T&e6vdeJ7K`r~ z%h134>4RpO2_fR!;@^@q4yQJhnXP#)#cZZX78SSerF}1HRoD2XvTZ?{bh2~Nkoy`w z5Fh}>u5M1IqLA*ECi9O7u_I@Wq+8KU+JI@n*$a+~F{VWy%PO+Ghn;kzuI=};6e67U z^V$##Xsno$=P;WWdgKAQuI6S)b~tu`)KmE;5 z;{N>8cmC@)aZ;LbArxgr|4sHQT(9;r36bwMZaz=ea#)B|mMs;^PdC-}**-Y|gCPQNzIUb6%Y9cb)%uTmgNZ zzae^;Pp)C)S@`Z$<|ut4a@8WfDDCydxeh)WG3?T}Fc%c0ab%yNpcelYH2t-z z7sFPk+BTKVb2G`R_cZ7JK$18ttBTBI)dq^mq{s)D5)N$f!F#23p~=b&e`9L6v1!>& zFVtjc@J@vGdL$^=2&hLlO~3&=0kuk3vpc{9jo+0YT0^$5+bwa}$=qlLq)Li`63gvW zToF?fm@8NjxRjR~TEv_qYeQ3mvv_zk=zcl&O01+mv{cW%d%(93*z~+ z(c#8eERx{bCtW*LQv?Z_b$n#5nNV+^aT)Ezgu~bTSrhV4P5TGNBuU)w4SY=2KWBxs zyUpZXWDOr6h_@s+_o>#>h*!~MT%DDQcoq%4!DUJ&VFBjpOqNOD5DWk5=Z zb!nN32KoMgHeRNpQd|=dFVZvxm(q74@3`*_x(XVnQoKZzD}gvTjsn_B@MdQ!I%GYN z&YK$3;^BBTW5Af5a%CdbjN$hx`6|E6lGuZ@TPPb;9S;d!G8K9Jl~D(Rp>OctsX9hW zHdaRKq=B5=tanYSCijcxDdaPDMk{xN7bMv%N;XU}_b4(*#bxkwp;I&Nb2JzY+Yij9 zC)M|*CsYkGC=o&NQJ`MUix0sIh4jiO?;75U&l@93y!%^-$>6FWkU9EJZmxtC?r`qRODPp-93Dy|>hT^v=LE0{)=YiZ<0A zn0f=aiVEpG=#f$>zUb2-10UnGq}gpCuFm=Hm^R|lp}0c&e6Y$+uPA`|Kbxip+~Zxj zl0G1@y2h`gKU0pl40#NBte2EX^r{Sh2u*a+eNo-A`%vG@+3mOASFbIKGy(9td$#40 z$|wkE|H~qmGFmMzVQ8k7>^q0Rgl|%%cq22>((M{@q7?4>~cE# ztY^-QVY4oSe`m4@Max!%G?5V(&gLV6 zyLfbJ>KckkrpQWkAwz%Hn_ivb41erfkG#uHvQM6l6nmih^ckFUl@(Mc|ME{sC6?Y8 z&e}RQ{42RdN_*@VOt38_3TGejSAcO*I6gL@a#74L56lp?UcTDxe z_|)bo=s(cLL;fNI;0DMr6i*s+j1Zpviaq~u`CFg-TRP=HS>;$PISNTnqbVAcWmH;r z8Me{PHwCQER!_`f>J?vGP#K9i+0Y#gQ%#Tv-hq$NoKpS?C3pOs z1|$20p>?ChzQt!hYR&Ok?^>o%Jy%wVjkKzdnSMig(w2YJs{XpT;}=>Lro#+G_PA9n z>KCqBu*(%%lf#I85mid;Ih^4?F)b&?gah`=%K8r5O7Wf9zxugljT>ioJ2tA;ir>I` z*rb>|$sJjjuveMDg+Ux)1dawh+llMF8jBOxT$*-d% zgE@Rl=g4y@t7Q^VNbkVhIYrzQpDGHGH>7B2~EU@9d8 zq@B^KQo=EMn8M({^#rskl?YLy8NKVvB3P-K{{{O-dvf@NwE=+J5%-Ewk>`zf-QibLOvF~3BETEY+>3Ww~t@E$se z!Ymk!P?U&nx#rME<=G+~KM!I(Iyy~IO>a^`YDt0mkQ7_BSF5ndX~d<(3)5gw%A^x) zO1&?_NIiq0%4D)P(m|dZ`#@z))!07g2%y;p)rtFW4t&o9pr2R#Vu-YISgL$(29_%n zbCDt)R2&k17#lGb(m*Jn!lp^f{ddvT?hPR~mDSNYh(s%JW2K`zfnEp#&ed?;c%@jH zDp(?l7xmI_7pgb=;O$p^aMQlrzdE2pHsoI7rZFEcDvD!!qT` zMix&j|>Rd+b-TMiRu#Gbpw6f>WL;LCbpyrMqIE_NG8A5 zkh=_8XDuWW_0szSaY&u;7M(dYiK@2FAjPvf0(5Sd6n*q3K=|<9RYgC)S(@#?idse8@;O2u64VNk zsbuOx@Bq0V)9bs=eU~&|R3~nb>3Epxi??Hb%rUwiIEPjSj0BX1*FtEjc-FPBMjrBV zVq1#aQ*T56wgb?q>lhiB*Gp%|U=g*B?vt*UJiKR@{E+gt0Q-p6O#dv5orYu&EZfra z8CO_AZkf06!t>sWI73n$of^6i3KiFSogolU-0LyoQU|rK*sIS-Hk3F!Nezj&{R-Sv zW6dHqMRGP%3@pT0mz~ubj$tr6t(kn{$N$PRNuS?+IBcJnyFWt11m=)@?9y)q=O6&IB@l(Qau2%QzL-UiFH5pzKnZ-0=2T z?5>V?)P-L4K>~<6+SpQ(owf4J1#%p0p7~X*24r1&Ao4K5=4YMcEt1X-HXN2oay+ykgG0c)TNvO!)F&`mc5Z<9`l zKXPmsVYS?28tm@?-)hcioP*YDId0|7DO=_;d-2K&32!R!L~)Xu%oOuW<0| z|Lo^(@8UmuCdZDjZQ}fDx^LcGljwNmN^vqN;joc=#7r3MqZmkE*+s=A&&yVoMZo?v zczO6BDRR}vH1evW3JrvaTGbl=#u>K-b>j14-JAo$VV4T|DKhMn=iUwm4W8fPQ5Ml2 z+vb_6z>XGmL0MqE;FuI<$g?o&;oJ?7Q{_N7*$Wi=3=GGC#Wql^gbb;#_piQZS+U63 zP=Jk6N~}u7-Eytl8pUS6esIrqgK5zpHspRnmG7|RN_$txZd}+qp@YV;!o{SY%xJn| zS(?UKT-Yc!%N6BG4$o*1?)2H@eO2BbdqB1Zas#kdzDtM!E$ZEusl8LJg<*lBC3d5KU=FBbYBtZK*3?Yo_52SitgbN&7+JmFStpZMeiS#J#9I#UYiuOmZWBQ-lWw|@P9w|0^&n7&+|iiT z#EWty2l@32Hic`D+$mRrdtBqc_!;=Zw}~oVt&YaDssc#YzV%aQA&3Q5tP33{f#s`r zUwyE^Bt&u)|4~AAvlAj5#tMXnMhTH7iUD6rJyJ0D7!oY2qp!_bFUeQ`wT`~3fIPRn znO#DoODjX;s7Wy{=ea z)cDHcCDa=Jss+#i=o+L*lt>z)DtI^^XKMBBBmKeS!PvNQWF1|`v*WMErkGPh5Q!Ee?JEdz$2S52zq zh$>gGCVJ&mbg5}Q8iG*_pbpvt425l8>5?rTnoI>IaHC2p5jl}sjP3GU$^uqv%3?4p zyLR#w*~haiU2dG+2HBuu!7@}x*2+`wS7!l7|il_wcB6DW2taB$YaA&$p)AF5`kWRA^5gJHz$#74r-4bn3p%RpT9FitLjx&Q^~G+ z`JTC$uAd9V&bhqWnT@=q?&Z-~Dyvr%hWAINiPpKNfCYkuve4jh!O6K7g8TUuyo}fK z>6;-REl3Tm^lFcNIM=|uzFFt#Wu7Je+oHx}r66+bN1de@ddBB*!gAZ^F@KUjw0Nig z=658MoyE&x995cWu0o0d>68PS%daW8oPX9cpUDIU+s!_g1J;BNDSCt#$Q=j(-izo_ zBgaR3>;>|tVrm=pCnTS%ifR!a;twghBO1nygr_gqIB@dxiN|^Mos8ihTeh3A(bWYT z7aDyN>E63$;L*n5)=*>DvWJ=d<9(f2{WWZ!DXWRcIhdWq=WKaL#C-3$mc+>~$Y#d| z|1(x3c0;3xREQP#3AaPWI5<=;@Tw)vcND_{|KN2D+vvQ)t^kZNOXGlmBYR?row zO00uKI%ML2z_Q>NWTx+iKoLfxjdYe5NIJwKJKaWH_F>UrYwc-{xOqBYqv{ou*aGKV+Dwvr6gr-Q1Ywa7L85>f~4MIb;F^dPK)XUI5O z<6q~$csJ=+^m*`h8p^gGuuREbdmq|1vlmaGo;TQPerR;%=xppV zL5q8%z$wq4?X!MDCc_D5Ja*eJskInCXKk1L7s8S)x6z#SuGrWZkGy(G+6~z!79Avn z2Lwl=TKFG{&w1VPd+SbQt>5K4N)GymgCIKPFC}$B=Xcz^#FP zytrov@I$`~@Ucu*U}FPwIQncOwP8IkL6j)KJ~SGv2I#<^HO2}7L$Lr_tGi|md5ySq ziW_I3n!4)KDy*D=-4_egiL*BbkGSC05Q|BiSLE#ga6Lw<1CM-_9qvCgX{vSq{q@($ zISxkv?wG02YZP;pBA2PS9{NB~eQ*)vIvF`M?#Co5Az2?uFz(Av60EH)RNoKYfpxeT z1UT!agOKzsX0s0#+77~wr&XBd+5w~w#(BDgmx4D2BbEPh|Ej1Rkic!|tF}H0cT(}9 zWg*bnTADKzN|AO4t&dtCgJV8Ha)_Li z!VdVfxJuazh1w%7f5Vv5my3qn<6|_q@5YW-*QL2`y>#b4T~j~{taURx#;KWXEWoJK zW5&p+Ih=9G4kNqubC-R?B#b8a|Kd~9#9?6svS(Bp!&!z5JMXBv2TYx_%R;e}ObBWR3q$UK2l;DUG}a?T~YsD`fr5oxEN0 z1A3FOQJvo~_vX3TC;_ZJ+zFK2s!F2 zWTFMA-HUIwG|vBETYA3n=RLy`lR(N7uDe1~Ic)rtni)S@ipi$PhDY`JnF?5b@Cj=g zgw6D6I#&Rhm}=acl}29oEaoM9Az5N+|F;n@*FaBuR zvx2ibC>#AL$|42~xmE`A%1GfbAqcQsO?wbwdCx?AW`~IR(>+TqIW9Si2pi;^*t@eQ zVsj9r>uw1@5Y*~vCVrJQ(N*aHfX5EFp|v+rO9ib-Ba@=ti1^hn1`=ien69M&T{ z&GbkC#Q^b(*4UZ^Ti6(ygk<=md$u(E3+RKE9p2}&-TNRPyHGD)+{T0Ut=^U1My@d= zXz!(+AAhtHCZlx8qeHW0b@T^zF!^ff`65fsC|2r-vyn8i!wgjU6a#zxO-AG7u&d_2 zdnM7-1e@$sv``!AL9qry0$8pF^g@+BM;2nk2n~8>CnQg79FTIVI4W<8zGZ20Jh@fj z?k3+>U-1fF2ZM{o49KvRt^EoA}2+zLjsuw*7*L4;ul6ES{0UE6KGw8bX4hor+b_tsxUD zHH?U|3))t}=iHO!mAx}mrVUH%v7gkCJa(RV?xG#ann%Y7D=4OnBHPdsHM&MK=aoQY zPJ=9`XpNw`sX@BaTck%`9c`>igDitpQ*-E%pjtmnA1G9J)7Z?a*K?3yq`5LOHB>9x zH4Cz{n|Xt>#gG5z5Qy1@84E%9tdEXB)0D-3Y=7R#OtIZii3HP=s-w zlob1yiO)u#r-xn2-#ARxeC^El&(pomV4<~lZJz$#Ibt=;=U}xq+^x>@X{SE4Oy%dS z6V(O{|0&f?MHcTyOoc}SjdiEDBP->aYUxc-7uer7RniJQd#`&Qkv2hz&qN6q*0?o? zm(xD6!tG?8OZjJ(K6V?fY-3^77QQBjj*mfqlHmZ`NlnxvfAq}kefE!Tn4kkz&z`%6 z-RGg!2l3K)6-S}-JwjzI3LLj{e!cz(JLe0bGy zU}uNtfrWpNiB0y&%IwR(Bq=YSVn?8AJhdC#HL{3}+KozOqjH^l z+kytVPFx%M0qKBNnbmZ9R2Av*{8Em%(5frs8=xEb2Lin!TeWw3du)OFjF(>VRz5i% zwI%$9U{7?u8mfhEDR$3rlA0X@Tpmgq)6b_b`&h50b${=imr0y_w{i1%vX-4V;V@i) zO@CB!^%jZ&lWQ~dZs2VUZV^5RXc3;KS3q%CF%Q$9*Yhx`@fOLIo{{81704CE5unab zVOC79jLcMQjabI34R8`PMlIbKQ8H=<=NyJN6opy3N}i12aTp6Wdcm~u)J8b477w#}1-6K{7eG|7&(A1o^*)$C*kcWrY5 zZC*ynjt?p3FhvehaqD@Pqz%w5dB~%gE&%caE!DT^=rpUolp%anj-6Ht^Y=) zrVlLP5($=n6{-t;-pYS%`KwinE(_B{IlQ+t%Xpn`$C#DCmz=4%0!`)ntj+~aUCd2E zNA0k@04mUQb1LOUYWx}`uNuQ zAWIi?3m-VXbl*{+85=Zg-=%7slJu`80i?YtdO*%{I8CkJOr&&EOczBiQE}&kk_FY$ zr6M46 zmTl&GvX`)7Dv44B)NJLIRvLH>A=e|^wAN${n z|EGxG53%%zH@{~*(H{GuXoIYeUiOVfXt>xOTOGYC;QhsguUwS&$nl2u*v?ljNuj+S zx%NHl^2kxhVIzPn##)7sZKY-<6i)q2_MN4cdv4CUYHaMHb@b=r)r#tfLfVi~e;nwW zu%*;dd8rIsE=bmYjbBIa4$`Vt#qJ17o}fOfW59>i^>p+}R`{57Wm@2;Uopv{j&Iy3 zBD-D~LkJ2|M;StgD5jAD78(Ov|1$R*Ld^!*&9KUd9(tz_2BfNss%=5pZW<&ZgCDJ)TGcVK zG64IK=ST)+TG?K(r!@QEV=uiu!n(k*2MA*TorO$!>PLHi1*>D;`wzEE&wF|M(tsth zl~8P)IHf!~L6oh!D?ck(8n_$^{klk>;9Bg)S!MhJ_kRAa8Fzrae3B%xkvq%W-$BRwP;S}`P#!GYa_Ml&z@bdneq?gF`=L|2) z%?i{9DdsLk`mh5?rsA*A8D6?A;JzXa0HXPytBP`{SjW=Q%{~LX41W#&lPhTT%A;@5brC{>Z(GPU(nCyr_QK`*mP^49j6egRIrF!ly1Yi}!HU z3DVCWmTUH_%EU_oyZtKXenHCldmtSzXGV=5m?}qSRnDjpE}O0?_3I+TF5SWn%&<$M zaw9bKz3q;e9dXH5Cy`51uywkGMsHDZKy&bEzmc!p78K3wotfd^2}DCh5HsDu)0~cO z5uT^N0J8}Hf^{(&Gp{Iq;Q;8_#N>D`*oL0?&8dGIF_oPB%awngMDDu~i1dvzGm|N1 zB}JAS%WVu9jmS0YWCBr}FW|GEu+3YYdcpP+ey$|{S0+c!cQHw#(l3AYK8a_ydE&6G z1C4n{p*oFXQYf;Diu-usSqAfPGoTeOZj&^K?g_SV49I*ovOxkk6~opipL%ZTzgm`` zvT+7;7!o#$RgvIeH^D-dC8B*1hRyRK(k1NkT5P zP~orzw{M~;-<8i1niR?_=hq)3t2i7L*lMOCvM6RfMKZv@_Fdoy%$a&&ZY5t+6L3qW zSJe77$P?%652{_Lg|6gdroYs!%;X#Y%~W)e`02M~H|Il@n|D95XUiaD=QW_X zDD5K6UoHD((YLRAo|{|imCCy<7=)^QJrMAG@=~o=m$-?i<6Q|&h3<9b(5Woje}JU& z8iV#q@BjPY&oFhNBXnoTVAz$gV%}lAGNUml9kMkqOH1j5$SvVbk!PcIZmpty@_nEQ zZ_LBdc#iEh+8$gzDU`-=h9wvae!jTHvdEpY8p%f4I+6=#&g=8J4D-g7w9n+F-+VuYm^L;$U2;#56&&8!6q;>p)>BLdMbfFb{SbvQ zbXUNuqEW zhmiq2i$-}xb0}sLMK)4##)J{f2Ql{7K$qASKD0KHqg$*lv_u2}nUE4`qVl2?(@Bz? zZnGL!l@&saNO3$Itjx#`A$^}jercIEz{XDDaOBuVM!_1lLxPj)`_e2=O+93{)%y%7 zG+RBoWf{?Rp?4zJOb~h=E*uLKJ-lT?e%}1^qE9V%zBV?PK(Q21NH;(?lt%DD;Y+Gn zAuxtijh!CRFWtz)IQuHOo~J>AG~8Zl41FIHVvqYdQ}wfWzz@aw(YsM{;)+Rzd=jC$(m5)mX}Wcce<`2P~gRe55tCSKGl zE|H`H%QqIxVB^qKNZwfI(Sprlk=zL7jcFQ~qc>*U){*0%`3oXM6#ukZ z=W9Yt_rkDCB%8x#YPA`Zw^0m)Y6_{iPatLW{i}+Dh9(%j9v_e*egZE`c2vEQYFo5? zJ~qa<9I=Ie#s{y(i?Cf!8?Q@PK%ZpV7VP0K6BMcQ_*kqwD07-L883_-4`AbrtYAz@ zf17>Aa=ZInrMlYK0uQ{@?Wb4NFW3~`#zXo7blW3i+fArQM|K^p%4xu`y09$JV)cG! zowSXXjeV{EpCvLl>m0Lzg1njQ!4sqD6s+u+x zu3VCrCfTcyUgvv|-!0Rs z`{>3QWf5=+m_j}9l#%DtSx6ZlOaM7fLJBK^GU>?jkhi9q1j_aQ?EG)Catf3sjNUMA zq!_@>S}G0`@=Bmk0UH2!dF%~2N3V^-V+>w3`R;oOk4v4+J#F^TSS^9sozDHhZV5Qs zB;v;_;%C2O!p5uHn|#S9FHGigl38`d5XF2!kvmjeq5$bA9+y%WwUg#-bd|gv+85Wj zL0V!mbgjL{PY|7kN>5FdJfFD^9gnX=stM*Oj<|F{d0;kBUl{y{ebUTutV14@)z8#J zO#=?dRwanqAg>km9-7A~Z)%Dag`o3##Sc%PA}yZF-Y_IU7C{{;-c?N}h)N{DX>9ls ztoW&4pg|Ri0Vgde&TTU;Nppghh``Z6LkH)A83T0fv_e|Publ?JbWa#oo9Y!EvXj0k z!3R86dcIvs>7hQFv-K4UuLYOBYXYA{`OYNLGKG9*CYUZ$OeaM?27^L;frssHz;?i> zn~px`TdB}=$cEgzg+21Mf%*Puaulm~M`+rBN}`chB1saRfyFSEX&Tw{(G`>IRSF?~ zO*`ZquZr!39qj+l-n+m>b)Na-9`Ot_FAj5KV9p7s$Osw4$i<~7sQ+iW{Yn{2yyLGgy5;01I5hT@laU;$k5ShITR|pMd`jv z0#}DuLosQ(dy8(7q$uhfDezh7eri&M=uq$;KRiwrt^giNWpEqv>4nRgUBdHHusm!` z$eySS*`k2;!U_*OGOC(22KA*m*L+Wbz?{Q9Cr9D=DbO6&C0=-@+?0`t3JYwcx1{fq zgcoL61nK=jsq1u#1=d^&6}>#_zULZ0Gxf4hl^B&L!ycwt+Df{@qgu6u)mKBSsE5$r zU7sGl94K@Lf|3Gy$2%N8PtNV|0QKZJjy}_JtKvJhsRNhYd~xP9kPQiH?xJ|XC8j}{ zBPa@j#j#SemBsyew=zYMMmNH8mPN-&N+Q>WrYK-R#I^G(0|#B=b~;mAtGO-h5yD0+ zK~PVx^?E2z^;qEGBIV&>VCbdMkvQOGRL1-=d?9O_f8lVPdWcHuW#OUYw6Z(zp~MGopD(s!m)CsgTi0#w4L7NRPOW@ zvsMYqsUCtlU)$)#5;NJgjxM0H16qArL%)KXy4BiDME|2rQj+n#W1 zIo%lqyLwPA$eFszqdv6X%`DuGkN1GLj9WYG)}3XlLNWi{=0AoTvctidHR?w6i4(HC zY~+|pKl=2aLM<-u^5pJ9vXh_7>%vt4$E;l53W^04?lS1E3vffZLC`LxV&|Rw)|ofn zGf@9k;PaKLlU@_BM34??%(KF7Pwa@3E7&1V6rI#UfMv!x&}?7L)+_7iMpdD9Ex8?; zJLjObNoj=4`7U9b7_Zin?T`{Z_}z)WKB8bn0H-Gz$y(3{~ zy6z%PIM|uBnnFpT#8k)pD6o7M2BXqs4Pt$+#N5M-+hXwfUhTb?2QKU#(joUvJp79K zU5j1W`;M2A+;HIt%yO%k@dJvzPmw!RbWddYEVItvHeoVzAuwKcArOk2jSMb9=cP#j zMS{cbSRS%Nf`R45K{vxrlcnJ~5KF!dB2L@9^eI8NOs8W)(%tVX^tcvbBJ}a~MG}Md zLSTcR9+$ZSC`0O`O?rd)OKLSe!tDy?$fMK7{!-^rCvxIF?^}C^& zg7~r)!`3-;yCR294>wcNPf524@d=bUG^oE)REFx0M(vq(P0+2}7qrUzQ)#uL-aVOF z z2x}6c*PFOkfl5$&-3{tB0blt+5&@G8PT?Wi8V?@!X%HH0M{)e;n+F5pY!xf{AjF0J z9%%75DAbohu@G%YrJ{EUk6?_Lxum+HS})k^Wl$H+ECEU8^PwdwtWJSyMNPb{)VD%( zM03zD8w>_+-$r9K8x}?!4~Zv6%qYD+^S72Y?E0)VFO%yoyoN2aTEo7g*t-A&bIfv;RZd8t_ykuDu?^z(0#2M>=QO(>j;Y%WJ4xCo34Uv zx-+fSu&mU;n)ZZkaomZ2{rLMI+Qye%Hkil}_#V*zXjWWhYJ=BLXb)dEs&u2rvtiF9 zKfZYs4|o}mF^8A82Wl(}*VMHCNhH}Wyl@?~a#!|HEEMq-q8uuS3C zXw1}u@(Pa|qH}^=Z3=9bOv?{utyX!XP)w^&C$tgAr2_MpHz16Ix83rzXt)>2N_6H^ zpjvo+I<7MgG#B;&XTTUd`={fO#Sb{i!QXyIZvoD+sL3MI{lY{86RiBBMPu21iaeyE zw@lhI38Qq?(4*#{@MA#%MDlRgzQq^4-xx`gvUm0t-&)cH9kTI`l%R&Wm*##fsE#V} zEf9X8N)bSvw|En@Hfqsb3^ZsfLibMI>mNV+LSTz-iLyc5CQg^+3d^Im!alz`>M$u* zZ6qnL6}@$P?$Xzmez#}tVUqP)?7a7r-+uqp4;#PR%XElabQb~>yrJK=cuUCj>4zbr z*e))gl?ChFNxy@_Q}k}XquL%}#$-f4^!^eT`(g;n{Fj&bRZfIuGwSLz=-Ca?^W^X^ zRCR)W`j*#iXnE2o*d%D5gnt3PRxax(y}8HZ4bg!0+VrC%Wa&5p{PID2n)MXBjv@(E zboE;o<%>g3yooJd_G?ba4}g+(H+{gb&+V8rjXo%M;0tcgD`~$GuxB*K-NVZWxbDuj zXywFj*lJk0tfvF~nn8$NL$RwUvYd*>t(A$yxXr6ba0Oo7vvWonq-I#z169WX(EgEl zL3Qj8qt9*BT1>^!=)bNb#Y1VqyKpqG-U=nP6kA1+161@?VUwxH(P`4{n;m{is!t8S z9C}s0TDKOWT#SJt7%YTcWW%j({FhB zh~`jmaZm?roGYNk8j7NI(zobEq<~I?2y$Lvr8Hf1#1AWS^fyFt!CUD|;Ufa1J-~Xl z12ps<_QUAOJGp_C#e}B?T0r!lyAG})+g=!_2U;EtGD0O3TS$?eR5UmxP{DUl9zQ#q zZqZdt-Vt??=~iY4zl_-KmB3~R%b?@%C3!df z3wK29mX}Uj4*o}_@Sz-%9pZ(vw`k4_N6Jomev7(f{HEaf4&5L<`-|o?)6X8wad&WAjv?we z$G*X?F1PJs&e8t+f#AL;kf3P!F%; z?!a6LQmarCwO&!J!1&i8#ij5|P;F4{dw47$$2t^GsE3WI)zWup4;CZi9 zcp%sTls|Jn;Qp|9(FF+nSJNqeyV)AhKP#2C!Jalna42|#%zQ3Iu+XpDx6kd7FaCe5 z-;!6dAlSc4_^Gs8Ze;3}y`ZZ3NYpFDu1F*DMTW8ge5}mSUWobZP-Bne94{+F9hq~a zOF7{9e($XdWUUKF_`vBHDKRkcmVC;`2%v_8a+F_$cg7m_8SF-9VIZxMks( zK;Bs57{}44|NeIC_ie*$d@_tKj1G=yA$Dkt^U%l4I4`a7PeXYw%y17#bfwX0A)|!} z(_HcahNCC{Y2$t4@Z0|XX)znWZ}|08@&!M$;lios6;?@rhZOr2Meb73t%}-7z@`(N z_gGLV#AbW9#b=}qK&wMST?bt+jVBmL!5a)vi+NwMpjB~_&IPJ4=BQ4+Zcu;nv&^@e zUjyd4MI#P_(W`w5HGOWV-U;={Kr+vTzM#nr-i>4}4BzL_CxKJ3N!K2MgJKdilU~F0 zzEM9HXokIB38FT50EgItqk2ARp^Ty7QVj6^bd8}@PGM(hGK~Ye(#lx{}Wj?R7}rxJ$eG|KL@!x zn<;h^{;|=wMBC(fq{5?X2L6OX8TlTQN_-og7KqFkyuKTvd6&cX1&x{=;$wr*?|qux zaO5>^N2g}jJTKdJu`U~T;b;wuRF|UQ^HS6L)5>%K2RB`|N|+Z|#AJup_?skg4o9@8 zTA6lL4tN>kSlrkc zV&^~RyIbSr-WGH6Y#Y_?{Z zOS@s=jJbcxn7m*yChx{JTliwhvI0G2`W<@d)<1tE^lyu!+fLsjDE7KUX@L3x=qr9-Q5Ld!VlyZtVZbF` zfMvBPE`f|}(7J{lO%>TGkJBZ{mU$;m_;{j|TVL31b%yX&910Da!Q|wZB0GB(T=vKV zMa;5^D+Eb0EI%vt?_sWoHwes4@-8!ZkoduNW^EpAVBQT{^d<3SZB;A@OmO-;IqWak z<_mH3M_y*#bt6}oez_~xw(7sfYPD~t*fxq>L}B`Rx@(3(jX_Q*3|u7n#J!V-=+;1_ z0k7|iK$*Hyh@W@4BS!k&7D>9rJN;2=Zr%9vfS+1Tg~+rE!QK~Ggn=E7@WyOYFk*}E zK5YVlGDZxFG$?z7%cG!V!_1G|<9ASgL$ofmTn#5O;rdNwy|P-=5w_CnkT%ETIJp#| zUojIKy`rXG8=WA^kAR1<@(c+oD3QjUU>-KkhIp}ub&3O^ym02afv{V;f7xLx!N(`| z>%!3&4kf{(QN^HqcHc9@vj+-Gw$pcEQ{Bm!Yx@_^D-38O*pK#H@KVoS#yA5DXOFAz zLKYrBeWAHVcKcU&9M$fcy;kE%oQm<+sj~AK+J{S{pmG0Yn*{~S8ZbR9SdjgSJ%M%lt#A^^WAniSpOSu z3dhS~bvi7fWsWTm;st>i2c08ha*kkKV2kjgc(vCRaGYxdS7&}c;eKQ_dv|*Glonl^ z7+X$W5mc+P1Ayc`?SlL~h?*=>)`wn_W0iJ%=v7Eh8Pr%+Zr9XXecyI|7%5!r{58C9 zJ2vL$ul{XSZ12|;39EwJVV%v=Zt=z7KNh*|Qk{4OOIpmzq@X?kjvy9BwtH2LhU^~9 z+b|Hqd$x!3z|!3dHhsqu$P130_A_#rUs}qA*DRxzi+7r0Kch%J6+MRU$l5I;|q2f^F2b=$dp`7YUoPkXAMq_?s1zGqeXqIUdl=g`5qprm>u4 zvEVYV235VxV^vJPBgSQ{ zO&ll55t_-JUx<&Bc14D;VzNP9F*%24Bf4P#Ygn!uJj4rM3-9#5YwHYgcoF8!?M*P! zs0h;F*Wh<3xG-FgLLq0_9%x?P>zOV?We~Gp-tV?s)dfyM&eUY)3r(*NI@a~>#WR(wvM%3q)37OZz4B3@L54}%|imqc(qRX6ux@UT4)XG4htDbGC33reoLSX9;gW#>wSZG&~NkwCX3mIGeZp|b|usEnvoF+>T*Y~}APS6*a z=+mRROiIX}S*9>X=G1kedKAMiA^OegUMQE`42pm$f@%=GE)%T{eX1z?GaQVT2?}X-Dtnn)S+jhRWWOedJ|i$rNguDrri?ZIM&_FG zu)C9^(AG1?4;$9cI`dPEd+6ekV8ewC2}hfo9hw5@;C?q~^~-wHx6YGon$iKZDzjpi zLH$9uZYsXuq?ScydFrKqBddqvV7ah$*kQGl zY@yf`ifo{wFEd-U+r6-5VGg}R0$D7z9w{}A>Qhp)YacIbP?~!g#(6y3YaaBCjpUFo z3PuYD+#b`T-vxGhSn%{ot_JQK@?W1;Z z>Wg{be%`fSN~DXh>n(P!!-@|q4f$V~Vf!*Vqc*XFei5&QZOrTAzWZ;sbs>D#57%88 zxhh8Ll{p^!HL(-U0fDDkn&VOAiwR;QlP}C559KwY3}K084~V#Q(XG-^1D1UTY0R#Pcm6Hu*4k(~F?mnjZTstjEPCH0-m8AUIhIj2V3CdMZ|yrasd zml72IU9gqz6)smE42Py&4)a6!V&iL+a2z=ZG2DixGVT}i!Y#gs>d*QuWEVf*!-b8? zaVy`Wiee8?q@0Rwel6vnu71b-|1G+MC5 zmA=`CXCS?tGP}@g4J4!*L)S2AL3Q+BKUn#Ci>_JFP4Aq!UvudP9~_z2qT4w$L-YP$ zKR5)>>=O2V`~2U3@aJ3NV{9GW8nIu~qHB%F7@2$HaBK~N+(=;da9!taUkid7f3;@; z+2q3YMP*i)$fa0_c5I`f2bvl0RxMGcd!L$fa|=0PKu68g-3(3O}uV=hj=$B z0*W_CN0CROc#*z~G@CwRYk?%??MUY>LL;30kxcMYpXIhJL+s7RJX-hGyMJ5dBtT@E zl+0iSz)r6m|4wC%riU@Bc(j97FQ_nV4QRLUD4ho$n;GPbC7Z{c_VMjc1UGHb;j%So z90DA`HkUWf-0oFAt2FSxVVSrf?@(n6dq8~l^SKx0eJU(iJx%K9<8Q~0T6mA|e2-T2 zgO}}~zIbic_iaP&E{hJ1AbOfCV^YSeJ;F5seX0`KPI-+W2}GVQ_@09n)|cd0XKvQ~ zrE~OZ-{D2(1r4XY_)iYZ4;tGl3*LX;VN#hh z2_grBLh!iX>6dJ2d%Jo($Q?Z~ZC-CBxbe>Z?gz%_H3zO6Y1oq4|J6(jIMSGZ&Lg||?Rj1I zUq*m8Jt*FOh+=CgQbk4QfLKNc6t^vd5E?{NV*TPK91<7>wZTTF9F*{5JmRPZWdpEh z4bb#F*>n!Cnz#YjJFhl42f{4ak4Aq$93#0$u1~}3H3Dp+hB81{IhQ>BIA{Zz!!5uo z$e0W<7YXj7?t!H0NSs@~ufw2??XwRDusw{Qt51K2OtIkT?k|-~$=2tLBvNgKmST#9 z&T08nbiAyau9|!&tit2GG;Rh8Dj9%DXfPSW1W~g ze2Sn8jOU7%Dp>4xJy0jFKkVMF*zB`;HBw$~F@u`6`y?3-cdL(^rJ4>)(o zm`BPFE^r25*N^0V9O@tb{7uAA@d=QrA|8K;{3uzWUuY?^>`k(43l&IdNeRn5&hwu+;zd zlpbN0wuLQ`776vcNe=@hWK69nJG?e{#YOxAfR%4q+F0+M z5!k+MhosoI-|Z+UNML^9KwyjD0c~ohT`agDs1j66(&MP8bApceWq2EqQHd^jdr-d{ z^0cePb#y!Yk^?h9A}^WgG5w=fd85uSJ-;mkv53iYz)IX zaS!IeE4-}J*p?|fPS}R?_*f?wc6KpY%qFz}?XMOXtHy=$q?0-qpD{Ld9 z2nBlp@U?d{Dsyl8CENz&{U82xgRO3c%PuB6t=!-=irq|+O;j|@e2w5nSc*@As93dL zxR%5TVtuxFLj^Rf`Srpx(rYuy0uw}d{)Fa~CPh&*I=MrBXmGes%Ly-TGz{hJW+(q* zjV(g>EGI7P>~i1ab~cYJu(j(M^V6Lt*b;9K4izh5V?VB$YNxqE~x=7$Lvc1AL+FtLte6DYEp zitb_h1CyBxvaO0MOdI*AeAZtIHD~ASWQ{ZW14q#XG0mfWsDKs6!L$CX;r?Yo!uN;; z6~%7p$3b_*Wn;!9%W9rCP;3%K5~=82&|R#_@2Y2!te1`#)WPuzkVm=ic^N9RP+GNM zls2ma-?iOyJ<7Aq&&$}ze17bC7f^49lOAXUsgVX{quQXo%wYc@EbugIGwV-y@9_XnQ+kiN ztyrwcBnJuVd-bSVf}IZ|9`@wxt?Ve+7H;6oNc^qFwgB|G0trW%=N>=QNr#%9xZt+& zfD?7ew0mTiugxBKsIzxMvA|JU&McYk<&Zp_FJ2KD->J;K`H7G0SpXDYVlS)u!kIYlq@xTZW2aTBgri0(#S z)y@P^d(MUmKe()a_}gzuH`{UwIFj+0 zf5-NKsJD-j&yDIvwLLRv$Ct_P7JDGUZ$!u3%?lE)yV~!3G^@h4O@qq@Q%kIjVHU-1 zrAQhTokn9$4y@IzRU)IRoXYXPtyn>dRKg%Z+&G;ujUg{b>V;+2TQy`c|hAu8kAX* zRYJY_$6t1Od?t${`h}8(l64+Nb%U}pwB9{+_GlSK-caJ4e8*mhY+Xo?u{bW;ZDoHV zkKBe>7hTr~l2j|hvYKL7P-Gdj>jskLYBoK5!L^7+^|gr6nr55HdG@T^JTOWhxedrk zf3hTaAZFHCw20)maLf!ehzG^X%P6*#B1PD22Aeuz>lTB00kYOge2>h*$VlwV#X;y5 zAp^QndPG$T9SzJ1Y5kgtNf<-99a$RKD@=@B9E8-MT|NzByx-`%C@7iPO@L+;WMm6e zryMJW^lX$2xyTHC(t!x#M#&H+V(eRcdOmu|f}FtLRm>+Fha%X!uq6SO;~+F`r&tKd zZpBFZY8JU-$n?Ol2{6wN1TPIgFKrVql3)+R1q-hDf;=PJ0)!%idj4_sm(p!s^O5FN z9Fzb}RhER~{dEp!aSzP&h~YCZkb{46!zXUl2f;E6R5HZtFOzr|HZu8EurN?8tT-Fd z$Yl90osk-lBF6SRXe|y2(!(*Pwj{6#c&|$W``yg{O%F%9;3%y?{K4ShS?3sz8w^9( zk};pGRZpE}0m!_|xvNOtIC8{l9`;cz$U~G;(R;ps0T?-E4f0L3 zcOac&fufi~MIQ=AUUP-apl*!lf>>XRu7vK;9-$wYJ28kdq34Q0ZSJ+QV)ikTB1vFJ z#U2d@>B9gn??-MP`NO|Yut21JPS{0~>cR_9xz(iapjg=TW>L`#ea;XxP1wA&MHeI4 ztT`aY76NHNNY4bhEIcx(uYo}Hs_ChaDU1&-@VOC&Qd>smeq^2OE9Ld+kX?pwvh$!Y zocSMQ>v+$ry$|rSbaiK@{N9$(>au}04)SV)xWpfo3L4e=^U@msRt5T|{cgwASA3U` zFM=vjdXys4Q;6Z^w z==Hh;zAGpk9TIeg8MJkDJ#3D*dFeCg%b{`Mjp`h_OlhiqNmZ{7uLohAxNsx$<(vx9 zfAgcv&MgQPmD2wD<2g`36?FoUd!)#1niVmksaz>NLVq?4c8GS!}|>@f!) zEr(w3RZq9DU`Tdr&Q3MilPql|6i=;~>AdyDb}r1H?MBMbk2~!wH&W8R@A2OwQ5$#R z5Iu+5IELWN_-JLX2{N_j{&JT|i+3iJ)5TGvj2XKjGs+ysKS#d(N8A2cF6(-5bi_hR zYJWspV5)k(XMf<@*UbuRol$kNYPw#WG;`^U9ezboyJh#HYQ;vT#?(O$?={Y-6@3OR z7Lu7vIxWblv{w74-)b?jd~lk<7{hHbhKQO_JER}g*sd)vNTQEp)j?UtG}%4~&L_CX z`owvlvW=0!hL)pjKL7%DV#s-5;TYL32-S|cy z5OzX>W7?c~`ppM_eSIV)G%lRP;V7ZOn%zF;g2Y6WZVq@LSR_f8{jY$+J~QFW*&llB zh7-$Dg_fPHTaeo#63Z`Y>cTGQHmiMaGR3Z?$QsPx;P#}IMFj^`nt(RRRD~Mk04nJc zDac?sOlW@SFWly=eYh}t95#huj2Gd?$3XUMR_v8K;y^`CG5A0T31BWPttx4$N@$sUFBsjpoUUbbxPypvFG|62ba5ue~7^QTJ4)_Dd$) zQzGq&M4W!CHL7Dmb^;@|&iw(=Z}-ZDLSp<48(v-{YFAX!#WU|l9u^)H=E*RKZw}I6c;F432BT4AST5~0Y{Cb+js``xe# z8QGCe*)Z|)(T0l7;<>NQk;9G(yjv#d1}G%q6I00kq3Euz>zt7@R(qvm6bl5qLsaxb z<@#w@oY+O*C;K(20j2UXA=U{eMs3n8A|=uqL7(VIREzFLNWC~-usZCp;`$Pq!G z+ooVVvmW}VTwwa#@cK!=N@=Hb)6`tyZH4(v?6~uOJDIDVTfCQ!tMI@l4C>f%1$1@* zo=FQju0EmZ4C|h;#XC_pd}qlanBl_G2++j67Jg^%xNEj~9zN#Bg`+PVS)p1@mZY7j zBiO(}pDQu93e=Y~eJbQ593vT%QOg|i$OdhLGS~aQ;)dv;JYlLqjk=9jah*gG(>-O4 zpi4Y_q&(Fej2uQ!9e&ssgV*I#wi;P38hpiGEKWN;X1q);xiBrM z&uSs+qS#J~d`?Ao(U+p~=Prcm_`hCZis#-@HR;+TT6FuPPH8HI54|&>&i!W4QqKc( z@6FgLzv|hd!@%to=017vf6Vv)SBq}&jWN`~jfaA<7(R{8k6bft8`~P1M;`haXDkMR zY7}=|1g%GL_-(>=RgZADzd?PMWclK+Xw;R=TmIUbdCT8y)%ATl@y*k3H0fG&+XX$$ z(iw%CZNfs0amJ3QFF^J;Io$k2i!K{PWlMYud}2YtxKI=8ci~AQ`LH0zy*ds1iY*v< zL5>OuV*g%af!x%z|4Ag-*0k^)S7_82d00`kEJA2=AlPe&OabU%G0$6mdAy`p|3!DAQp zh&h7C_al$I)F5uBQU}!(I~H z2x2J1!%WL0-^;e$ieC_`!O`3JLR5nLf#3wu9gj3RW_C#=+6-)cu_Y=yT)$j-I@O6a6Rh-Tfb_u4u+tL?U(%!A z!CI$M^!XxXV2LsPbP`H%*d<$Djr6i0<=5xQe3IzGffZ;FIcWQlL9xJ>O{JoN9<4nE z(S3cKwZ7|<7W$my0D|cArRWb#Dmvq4&3GjVs}eiJe-e;Atu*8pbYV#lTUF2rKGdx zC0hr?Wn=anE(H=3u)%OA#5Zqw8q}CU1~c+lggV_lF{}+LGDeebZyyo{Ee}pcz`*d_ zNEj+!|IV3!m9~{TE{hJ1TAZUQTqw}Z)R#!l32hkTch=p zA1bo_es;%Jj@)HY!67}4eHOR)rubY8UG{T>`r+(YKhx&p^B*kvM^sS46K3xMyw8WT zyM0f2pQInkH~ZcaI6aPGFGvO%5A&gu(ZUNPV`eum4Ph)E>4wDMGi3b>vv&bCzd`ch&&LEgmt~XeiRD5v&sC`kH^9Vf58ngZd0~ z%*vCj^3XTU+M{ZN1aDf1ewpM3<8(`6_j+S%N(N1jQx5r=lJDL8UiP;wreti-zkWdu z@iQeZyrsEfWlEYT_7p`tht(cTNzTYj4-mLER>;aY< zwq64^QqEL!&Tf(9>>Ps{)s*_(8kDgeqQbn=AQZ1*VjWf;(vgq59o3y*;F86?MeM&hnJCd2y zfE>k2Re{e!_k*GeQKj&x)}Z~^Zx>k?8Xvrxy*A^Zs72QivX#ZVt|>FL*QbN@)r=D$ zWb>hZ@w{s@V3i05RWumwl=P_T+Ke>%v^Z&2pX$S2;lpn@AV}_fQ%r;eY;8K%G>k|no2k!&Z-yH=R8=n)P(gvOB>B|Jk_77-Wl^T@! zrb^02h`$-s@q#w+ZLtU=3p&<2z?eNbV~#LyPmVhBGrY{(7~QYgy8)JUD8K*bdr0~? zQfajg6;dqlRr9FmFWTf3~eLW*LQ<#m#|3ng~qH1ds2Kxid5wLK0UyX zuV;^dp6n%giDr)`UXbJQfV7dAi3T-3y*K2Dx>9<@Is1+mVrE=-;hEi(cW;&jGM;yi zB$Kjnq{#|0brcIU#~Ld7B%K{msYapIx@k+nORNaJH{pC_!mE954blW@!mHbZ``k+C zUhysbsk=wtdnFNyCe})Bi4KIsN&4J!r?2$eGO0raN5bPFiA+LJpW89%Rz=>crF5TL zCS6K@qQcScy;2}8pOpbukB~x-`SbhS*0AaF#>i8zrUi}U<#z;wA(#1!@-B$gOA`wUiffbvgpLye^+kXc!Q6N z>$?9ks45!_^T%yEuK1c&M~|zk$m)O`t^S&FU${YC=93?&M~#Rqx=p-GXs&ndi7bgU zG7AMc0+R}9E3;eE%AA;B*_NltcB{H*tTNr>7bn@{R{~1d{ciJb%vtY|?l|H7;k*Ex z1`#fd0P|Ll{=dA?KmTnBe3?jY&;^;~$=bxw?IjX(lA~e-t-g3W^Gr6+jK;^mdi?2< ztpvWyZjCrZ?Q1n@0RshsxbH>n9^4GKLDh1e1bdZYm)eu`O(jOm%&j&?1p~ve1EVqN zycWW-UEA}RS1kTW`P4rxBRlv7dtKMlBA-~TH~T3TN>F!G(dm-QlQLvw!b8GXwp`Zb zTQTXqGC_vymb^D)y)a$=Ur%0d)HOluaUJvyF{;i5uZJ8Z+!5P=$f())*JNE zz8=|8KpYO%p0?&VYSDQ~{@HBX1`2$<6&FSiN4o{AXYYgf{VGLH(eSDK}lLrS5SWhZlkDHE1QISLz3 z+LRHZ<;ejZbmp(ezW5(o)+8UaxUg@_!LdvaPiBgP)`dc8Gj!WfR82lB?DhnT11M~G zC5U!t@)f2i-z)md%vS9ZFMXx5T~S5$X`S!J3>za3!UESYhD*PnW!ue`4@O+rqHy%S zO&8<~R(aQw8c~UEbI8J=#)!sHj4c)ll7$AUNR&2hGozTnddUaZ$NGy;PMgObFf~ucJ|bsTY!? zN9fAXW~g01;hv+#dz=jn%L~jOA6##~W6$4jy7gOIK}wh1`fw;^;&$hjM?RHLb<2}s zB`NYP^>3%~^GdHn+Ezuarb3iFeu49OEsv*$=WIWI&e;e2_B$qlDt}6?As=8S0zs9QY4WqIu5Eh)&-$Y+OAqbVuLd#?~rG^7pZm$QBMg) zDV=Q`c;e*it%{R@!V4#q@$XJZY{~i@ocLBnhUdO1rg$ax4_+pMoFRyOmQ?mrT~#Ghgz&^A3!l0ErQcA_kaL>J;5E)})Qys^}3G zhZxieqD`kQMv7s+4S zlR-yLfHFTM|0GRuBy=G4m)u=Qb`B*K=)#-aV^#sR3X0uFkuoZ}!_>hE`(vZ&tAiMP zk8pWZq4#abVkfxkD>YkL{XsdN1CjMYZxmV2SLioE)V@pHNiR6A?qv)C#ZyZndq4|3 zJN!6VkOeW`%FyJf1VKH$0RNK*w9Iz#7pkrac{1d6$A_Nrg-78(!-ktfFf|BXwui%o z8#kst3;fO2=RKN_I@oKVFAox7j>r#*a_AU|{$pV~gz_PY2dw=if%`Sh0Vam+d1)8D z$}gF55}C<*f`$hN>oJ`1i5m_>6p)R{yfMFdti{AESr&2#NMpPPaNX6GK#Su+B#AW? zyNV*qv4Y){1w8By0>*$TeeBCiM!(JsCf+XmI0GQh7<0+%-^jKlyYY#WxG*9(ID5H* zbWwv6tJ^mP1M^OOLRcNt@75zkiS{(vHQ#1n-e+mI_%*zHViet>o5o)ti(rsv0NpJm=bH)P+mbd@OF`RLR z8wo?X9v?O4{@@LZ>#=i+;1;QVVTc45tXz+JiiN7Bqf~VH4=&12i83P+W}P4#eLjuq zl*R>T$Zjjjsa|@Ybkj=}9_hYiLZj-ie>qj+w_cbbTRywd`;4@XHmcW!E_CmfuNyyq zzWLsZfkw4?toxp+l6qwwZLXNc9r#+ZiLID~1b--9vW&At)F$qH>!SRi9DnnUsEdKu zBJjT3e#5Y0`~hgQ3fc9MKSU=qi#tOskh=DR8^t8og;jRytUyvuv3n>|LPf`pTh6Ru z?u6YDqq2S7yo>VuxmTD1&@sNMyDV4%RUY|~3q^gZ99V`rrCqACUiqTfabNiv)l0{< z=&lLQk;>5Q@B~pB9Y|#`3Zx@q%kgBdGFxIj2ITM_Y1Ax+xe>>pQ}(3EIS{ z{RZ)AvV!RED^>=yd96Whb@~*&RnQ|m426~bZdf8;tHEJ_+hU3_X5qRe*zu#lWA>*snPuKQQ&`G>~Lm z&El!dsj+_dg0kq9UYRguR~YDHkg3He7c!O|`8Au~Ge^DB{9oFqKf&vjcHNz5#q|33 zo_9$F-1kCdNdid;x&=+4@?==czB2;b{*=paFt{Ox^1`VG*c4YmRB5#OFxVGYvGG5i z4m!tNu}}iM1ysSKAW#vtZ_?+J-aAO{k*m66 zuWkui698Qdu#~V(d`ETAuTRy>7WEFDBfxkW3Nsn|FV&|KmHq`_&)*L%x(^7gHq0=H-0+k#ihJUwA>#LI3dE$86Jw zFNjCWkyuO$I0?x_Q%2edZGd_`?+GtL?#L9mIVd+WY4jN>%C=ysZJw+KJlncy8(F9q z1G25YG880MzEMh7OunhinVO_JFK-d{yEO_>TV;bRF|rP*mH6;BVSePmRfoJQ!#0wS z-;5^@9Pvs7ZFt-G9m^UN96Rl2~S32BS8B(hY#Y}_WcnZ{@1#$ind z_MJWyj3M9)@-E*Kno}BlwoQ!Dzj|fPq&~N%faZWNgRj2Q=XT0}b#SE`^<>(`6_cUH zANUaKMJ18tvp%M;Ppk6nr!PoWP2~w7XF%mi@ZcGK0ODscf4OS^XCtBFx~^l-RhiuZ zJ8H9VQ=>W$gw;P)m#Bb%NACpVQyIFDc!x&>u%%??-=Lo--n(Z-t*1V z3Jm>QpM92>^%y(vW%u1-mSrg;Aagw_7>X6;x=uU!)M}}!q}crw*-J&|`gST$gjacd z1zH^|rtO%X7Su*(Og=E9Ky}KyMR%8Uh!+ct%yzH6>O|2}-DdT1xbMKk9FMggi%Dwu zVRB%`**TlEA9V&Too7&||1?`3FW4?R%slv{H~E><~AE9*=O0 znc)wrp%*bq(e?}fB8dX z0$q1GA4h-o#>*B{<0jLVk*pUcMftIn5!y|$#T3~^Mb|>U>A$Zq<&bV_(dC-*ibg09 zND2C|S7`@Qa;-BG26@1Xl; z{8#L_m2VGxSUN75IW?dxFRCC&798t+{-zQiBTmc5q!%{*xOyHxZDvXbpk2Z?un^;-PAM3GT&j9j1I zqU)x2LYUyAD@+OC*ZiwCgFYV*3_1c?)&}WmpX|rppFcV0X4@g+1u;b&{Ui#aZu-to zW=efinP%S@Nu~6vXF*h{<`m?)dCaH%-|*A!v=a21lVxKE)4so=*t-{hjYMOP`BQF3cf%g?{YEYK%q<2D_&8>ngUws$!&Fb|+YD5nxu|h1MeNvvR zR)fQLLrqJ++ahv>SwwP#*GMi1m?1GBf#^2~7D}q=rR1hEL8Qml1a0EX$aHzByeG2X zEl;*D=$ar!(5~7SfP~~WQ=kFb6$lqf;ym>KYA^PlW@nsnEoz>R|2HkzyZ5a(e?@9t zIIMihYV|rru_r0mzelGC+PwCLtY#ZQymjZy8oye3JxIMTmbXWg)3w2+LL<|nJ0QQU z_=+v3%0fDWHfeWyw~1@Xjj$t8@v;N*4ESHayiK}Wj!)Jr9|c}h9wT=lSd5xfpO6IL zcchR_#Hd1b_%s@2<*(}Iw}}&G9e)+8>f(Jq_pF9!bWBKUfVsjcbwdBJ#LPwsCcLc} zuPr3C>=!>j`;rAzfxoMmPc}YhK_MXj4XRezPO(`O*$Vj!8f&#n=B;||*4+BJSnqaR z{iU>@~Ev^=ekEDKnjm}=YK0VQhM%={jS(zROU5r z_?WEV=OwvrV2faWZpz z#tHZ3UUk4mcNQjiBj8D|)hRIGHXK6)xySVXR=nhID<;?`M(sm%geyYh0Ut+HNC|CK zT!t)Ei*CjEG`h%ltHPi@G&wG~mo}p@!jsrEsXpvc#4b?5ck&7h(C_l#{Xo zRiUO$jBCyo-+FPLuuRybTqf#d?r0N4rBk}VYNWq;QQpmL6BcUXL$lp)dmhyqXE@0M z44S!N8iPS2IOmWTe5nss+{^xj#W#^i*8Pw)yRdI^$I3V9pje>tU8bV*B&If2ZLm98 z2(8^3p)mih2P(y^W;3;&VeN`VS`LlfKC#{(CBRD*X`sK1kN3O9it;2KP|ACXP8UHpd!J6aQ>x#ks`thc ze;^F1_PXO}cr!k3=^v0opAM}NtPR3HO5Z-|KEYu>3%KEBz6bB-gzPYp_M~q!%rjg# z<%P7dKmO5ge*6s!(ptZLJ)4xiFk2+(^f<_8K1i`O6m)%xE}wOEDpYNST@K5cTIpZu z4^WtSMiA@QE7V{6Cp}&*o`8L|Y6J%%5CCZV~Nt_BgE=1XJ zzG3f&e#8Q!Vz>0;WSI*aEQrhv0>%c4O`=F56^$0BUR5cr7WII|f!fI&dZXe-819xA z?D50C1J2qY4gCf4n(Lv5aLQR;24>99XZr=&HYqV!O+W(0uBONeDjL@dY|L6uy66to zR)yYK>T=KeTI~AfpEZaRuJiilT{n0YKXKy{+k#s@DQ(yNmvMC!gdC%3ONYxqI}nJo z!`HsL1@sqclHfFh_PF-wubx(crK8a+>0AEBI0lY3Qa|DQvxRo92M{A(X&;D$R$OK%VfL zchjt7<}BMZ%b-53$fEJiEFs>}s5~renuX_37T_-ZP+kGG3g)Yve(cDmYUF?!+3)9i zdf$HrRY)uj$$KT^|3p@~utNgD+(8b>W{TZ}*(+!VXkR0{qmmhCq{7230`aCnDm+*hisKKIoeSrkmDiuBR`uu z@Uu}@Da9)7CLPH7ymB&l#k9OYyj~|m(g(h7xq+iH8N6o#?6z~hd+Sko&4lZwY2AL7 zbHd$XLF&Fy{a3Pt-_F>DDOOvp>`fBICQ>AxiY{YvruIT7MI&=Wwa}+fqc@tOGD!h> z5R(}dNDl&%-RK_l%xD~TK6wAWZSCWe>;o5`e2!)#*dzd_9$RyQmmx6~VKn|MKn%FKu;f8Hghpkpu7PVe!{nOZE%)b0i{_HNSRr$=&% z0l3x4OrP=$5NDIemjI%!Deq3-l1tY^j$YXdp@qiaX#=K=D`H_~D=f`C`(n*O6 zuRIM_E6>LiTTPJ)D*A>u27xzCxueR`c1j;Hi@YENh6xoUQFns!$EKIEB$pYKSum)v zM;2D?CdtZYRd}@1WkT#T@v*Q)-ATuJpyqcaZ5|@gr_{f~<4|OSI8zH;3sTHJpx2G> zgyw439e!wfs%VeM(iY5HBa9bx`Ic~JOh@_;;eOt}ub%sPXIso{clH4so}9kqDKW)8*L_xe*6~7HV5`*j=%A z_I>%lwOm07-Soq3VNYa>u0J9za6A1Oyf2SA83JOjnW+LAW)t=85t&mk6%q)}KfMqfbXj#1AGGQ^N)fPNFkG!v`d`_i}?~h>bb*jY%sGd02q-q5kwANZdG*Z8cvTDK?oR zYpLkX;5GJ&QQr9z_fDXy=`+Y$=xzm~x)@4B22s^a2Y8jX{~qqo{APfA97kW_H3P2e z1TFpM)}Ptt16S#bdV?2I% zZIw1vX)PAxX!Kv#kzyBKdg`rASuMp@QRDy>-Q?co-oq4!#Eu&{S_*QR9gyNQD|u}E zF^Xvz)QQylDL>mAas-N|O@~mVIxdll8&@JdPL?RU>AOIZTsCe6sF|VM=A9oj=}@5q zq%-BmNm+Q)tZsVQxXWahsxcI=KX|)IH=-p6=h1EuQT+b%42(SUGA|HO@6=8DJKIu` z7laYTQSnkNTj|xPJ|*p>i+$T^J*F1S?UIMz-}9Ms!f}JW5aM{;X{W#Yt%8N&7R#dg zvwjQN<-(Q)8YB&JYO5#~bf(Iw=#A9s*XqUlqEdYx1`j?qGBr`CC z2_8npAvob@R%$PP@LQP$K^fxpmr49kq6#jYtIoGVm4RYYDS$>ain{Fo;^?vg8#+7Nqc9OX`56{D#w80Su;2c&^&8Mhg|0dO>p0Yd|Pt%3$pHT zaLo72Iw8l_v1!1;)mKRiq?m6Q9e7yJrCngLeq>ZW-RrpE^RC5a?0v^eNp3u6Heq*aW}zgD#=cD@5j9X1 z)Jtd1snNy;_Xq0R#J$2DfnO=Uq)}|D8q5sHMTKJ7G&xavnwo$-B#$R0!EMdgp6^bQ3cJ$I0Yx?!^M&g<@@guUDQ7D(rlVf_AC&mEtP0E;Nfa1Z?ti zSp_L>gY@I-ADytxaz7alcHJNW*=e<6rBUo=ifp2yQ`MJb8(64dG^lZl)$fK)XuEt-w5^*C$y8X+@-&yR_Yda53AfLOi zR~%<$i+U;c9z||b(GAKJ!C9#CPf>g#i487^fGhz9>}xe?Py&l#{48xfoeM>1kHU(e zTXpa3B%vwR-$B=E8kB_~+k3%xrC@IeS}(8>tJ$21_Z0@;s`2`6Wr_kb4Co5!%Vq78 zk{Kh@smuVX#exMp!ZG9BuIg0wf<#@;R4A|oo0B{~j*6r1D|#XuJ=Td~U8GG3ckrwj zG(X3I+LK7|!QEtuNZ+n7X!9YBfyDDV*t=x{CA}&*TeuENd()sDg6Ypm7hs_any4<4 z8;JkH=g<@l5gs0+-8BFH^G@!=#%`6cZAx%2i8N*KF%pQC*O+#~6WIP7&$UNvcHiu6 zgqTjA?D}+AjhyZ#4bvJqoz>25jfMcpSpSP#YivcwT{ck6AtPQ+R|moJWRka3)+>>v zcSB^-6uQh@Cf%@JB{L@bl0&CTFq5fY`K-Fjj zRjUuyztt-daB$i&kQy#-o?O@-pK-_v#|t-Oe&MEh_jyk+G^&l@))>@F*n$Xsm%F*~ zdONHYqt-rh(8Spdi$Twf#vwmgB>eZpPlXnT$1MnSW@BC0;n`;8{w7oGT8gZpqOo;S z67=uVFPVUDLN0m%%Ia5-e=q>Se*jug{bD zB=Lo*PugXLq6~_KU`{Hk-J$vmc0o^Ouu*l2?ig!O=S*!5&}U3;6PqtzCOI@FI3Id9 zDrH~74Y7=SH zZ6h&~&6>Y14*4IX_nn<{A+Sl87E~`PpLKT5L-`lt9guveqnpG_Jy%6_lk~}t{2b%D z0HB5koHSF_hP;_*eu!&VWto5tjPjo`Rd?YT%fg<*Xp6 z^R{3dzj?$4g-#(k;5AyTo1ewz2)OqEXXOJ`ln@NhYPzbpIBjGKgI5)$ZjgS z5roCQghohdbj<7+N!3eV&F%FpkpgXIaY(w%$Yj#DJoimp5)RRinM=a;r@e1aLg|_& zwk<5y?}FTdi=)uSq#cx`psdm}UA8(Zbv8tJJHm44j<9%vxm#pAh`u|jy4KN_1Plw; z6!SXI%dr_#cK1JGgDmK&du?qw$#h}#)LNmZlwzUDdjS;^<^}^~<5N0zT*n zYfvV7t@eubTNi}*aNLsej46T3ywLLO;T(1cH(LI5{5J>3S`5mPWg&-1znjbI!;&pl z24xM!uA<0tDtfyY7L}a-$;$6E>dr%ni36x}!}{j!nEhdmJJN@^XMczgE!F>vhttPf z5OM9_&;Nxi9|v;kgBeVlC>CP2>!|2msszz}k|#r-?}(~&N+*3NIG@H|4cMQ)5Tqn8 z29!D09zWx|uwx3o-Vr=wEJt4EH3hC~hbMl}`Kt*QY_!h_yGT;U0h4VoHg-@fv|q>q zA5oAR&`ocTC{>|rSK-k`pOnG1c=pGywKvhafs4?dGw)=1ud2`p7+zbchaIC;F=)(x`J7^TgAM(QQm>(RU6S<4J=r8{TzBuxXyCV)lj~r%Wf?$<@{>BZ8t2J;L%9?|$`%|9ISKgbRpV zD0B3AUZ@?j)1#Mt-iu9;e2d$)oe?G$S~4VjpbZ3&-HvKkhbIL?#lJU=3)))hj(5L%b)0dl`eaPX#Ai75B+;FP>KJ z$$55`-Fxyk@jJs0L6eNfFWByzT-LmBZ0J6xTcR>)Y)sOqOZvfiXer*L!^Xb1BU2Rh zVk7C94V^oI=CnMjObDe*f&@*rdowI}NaS4M{jo519A2$aEOk&B(opj?LbMDuq(jc} zLJM_n?oTh+cHiSee{kLS0paM?heGZ(qJ%Tb^?F`m@?_OfopjyY^vHGabHql6W6np2E~oj-9!oOI{%G-v z|E2CHbIC1!$!HheTCK4ni7y<>_EF>klEhmTXT7ds+oLoSQ@qz}K`+^op za^-j@q_|a8FRk|aXpbNAHjfb;eKDk~%op=yccveMfz1D(%wR^#pl(o>L}L9e4j&I? zcwe~J%gtBX=?af@fr)f($(-lO8r9frsne7V!@`$jrpUL{{}WQ=i+>Eh4lc_Tzu5E1 zEl&#u%fI>EQ)K0HRvMLSwT(=r*o_pt@6ne@t=FO;3{4_3Os>x2sb_%kpGH?q(xVD} zjsGJ*gL;oDX|j`}J&&yte>?Q}QKueouX3?J8mqSzhUT-Pxo(V;a7jjM8r=6z-r}34 zZg4M~)voH6_6Jsko`$0C{K#HmuAqn6?$sYyN0(_TCiRbwCO*tPZ{#=?FJBLyF?&&* zt>!Nue7La3Q)y*a3Mm$<;PR;GcAz$-$@D-R6Q)218#vq=a5AtYGMOoxgn8@Ro*AS; zZ0T^51GP;RB7K8+D{Rz|b#>#VxR+sv)g4tSO%P#V5p<|4CiQ?iY`&w#=y^kiYj5uu zy3AwDC|~Z#n{HWf-mB2nko=(*92YhmN3DQyfMQ|awTFt%k|Yc6D|(sOaRW!FH)9qs z%a1IBZP(t(S>feWO~iWPd+9&O_;vf-ja2uWOJ7_3Ry*jNeWgrN%b5$dr!f1R|zmi_Qt zF{L=D8r$gxafv@rZi9fPnF-tG-I}DC>67(^GYdeE&m_9J#PhoN|6;>WsWdtJ#vcW? zf~-5#+`k>_t@QHAp!DWCh9EM1l*!O=VET<>B5{w`qc z`L7(A;wIV-9tMfiXBq0kNbQm$=NCF#4(|&BBM!!ILDc%d`49b)!xJP0QAXy%%sTpr zra;)|RK! z0+P3hC$ZC~M7uGlo2>f($$JyHrm{4B*el+U91PhA>iiJv}2mGc`TEd_C!!e^pl(qPWYV-~t*zWfhP`QA&e=ATCu37AT4v2%=?Cg`kM2 z@V##mEE0+4l7vaw^3bC04vjFML(4y5q@&X%T)J{Ul zdMNw{fs0(I3&gjsgA8)6cf1HVG$f$=dLII?MUrzN-AX)t+Y?pq4Rtl0@?5%sq%#*l zN2kyq#~yPTT2!{ULv~}aCZFu44+f)11|}~JsRispD(7_|ZyIykv(6X9oaDz`wMtGv zv=T#;tt*_*aY3DB3{B<*|4ALR-Q3o2S)#stRtUM}QzHigFpiQB0eS2*nMP-L!c7a0uw}}ug2x0?V#{Z7xtC3Q_xI)u9bDky#3o~>$-?zS67*w-+|53O(i>A|7z5tcNgi_KJirqkw zwU`4egf7+nihe_>Q-h!)C?{arXBcM8s!5)%nVs-C+?Kw=Zcrch)971!>gF9#q5!HAr)V)(5u7=E9NA4UdOIJ41VbIJZ>M=annk<#CJdmEVmViiC_X z&=A6vP=AZpp)pE;nuD>goD=HYvSNloel5f6j_Me>x9EUuD~tEzz4lwctV7QPpY+*> zc7xXm-M-ev#)T$}`R5|C--#C*=vtYu&@@sk#8Qt?aaW>lEj%DAmt^^8$l7C3O{t5{ z@Y<=alzak$dY#NLL^mGdkr>)a+zMRJ)5VebZo5E@9AoVl`2+5K^kC?S`_Q7pa}DYw zm&CinE21l+@4%j>Bjm=yLB+5v%lo77_E;=r!_T_r-!pHJpEKX~N=iE;;|tu_h!IXp z6ZMA&-tU^TCs-gP$n>l5&I;%x>pV`Yvjyj5t3}y@R;HAX>f}Zzv10rp6+wn9k7K2-lxo#zq#RF z%HJZnyXXcp>QWzFDjs#g#-OZ#WWNVKb-tI~PIx5?vgj-RP;umWE&Tm8Zb#XpKJBr1 z+ivw)?*VyU@C{ItEq=Ksj#6wbMRZhLHm{h6Evwny$9(UCsBC`F<&Zm( zNPi2watlh_`j~oIGSeP=%r{Ap&Rh+=AJrvX8QZJABCU?Dj@qd1u6@7V00El9C?+NW^g{gXcsF-D|pNe#R} zHal^61xRcsgjez?7TDOcskp1mAypHE+|t9!R4XI2%1ePc>T*16}-!C$YN_Oa2zivo7OFmb#NK?CL9ZqiKAPF%liHg%FMIFFG$MALN58t zn^&dpT?aRHfX)uA6Xftwz_*CMf~VzAyMeJ5FB9i-I*xyKZ}aEoA$X_V`dY{hbV8YP zt{PbGnXPa81gyaK(K#^pSju+XUZ{ai**3LCr2TlXqc(G;MFiENEjO&BL zZ33eg8no^E$H~fB1pN95&faE<-9(WT?71yeHu0|dmAzc{vL;PZ9i0@iOt5bONQHJY z9|xsLrqwcAS*YOQ9nQTRU}b z2T*fEpjF*goNl<|-G+9MRf0Ua-d&7j`A9r2!dt&!70BiqE7ZKT*_imZp&7qD&{ z2$0r^Kw>Zfb|wSlgMevXL-f$G5)>R@W}i#B8M3jHqeTmh3(3&WD_y{!3}Y^7e%;D02$7}--(WN-r7@~F=!%Mjs9&4T z7$mfzHMq)0%iFJN)sLV-VajpZEqlA7R)S`HYhwAQeX#Om6mFpT^GKjFYCHy(ugE4< z+@hsUOkC1x0dQsE?FMlDiSQsQU^{0o|k6=R53Px{V{P_Ny>C6QI^aMuEUQ#xOZwogM4uu_=R8fNVgxGK@J0PR zy`zq#7g%6&2R8iyBMe!@@6aiiWo3M7JL8@%CM>Cxr_Ev$%CEf-$J5007qnvl$8$VGC z>B?RFd>V#DImP{C1Kj7dWNh4Q*uQM~+8u=vdRcRS^E;C6#L%lWu}ZlVyOSbWR9t;< zouJ=M7g8KqviR1bTHy)F4)1&FCIRS?@Yh2or!lZ9>LfiBmEm<5T%_Hvo{`rJuSt<` zz@9WXmLB;j8y!mn+UYzeoKTN`{I|dPTO&@ER114a-7Ip|WI9?X_B2IKQE~mRUzNUl zohbz}2ncHlR`b@9jUXa<5b5%w^7yw%m+(8mjssSnY*%e}y*HI`+LWix()uvzdG7PcZR(tMlYaHY;J3{wRh)Jw zW`QJSng2CE?B2+tvw}6ZeR^cKJuge!!7mzfY2|f6cRH>zeL%aO>2=g+HI*c2^}Nj_ zh(_EEt~I6yjzXi}36cDR?u%ZDk#WBYd3b%c#qw__hF4fUHqeorJd`ANlea_OK z1=t+2avZ+7V|wBpobWyKP4NMNc@XeRl~1w|v0Npn^hV|agBa_TL#Pbcfa$~?;6*vgFp5EIXU&h?9XnQEWI5R+fI=-P}g+J zgnxPo&HJPhw)X9yfwhVVgyLg#T_l?gZFKNPD0ry55<9~d=lP^j~y`aHPf>%q}K%I^U(=GEY8BO_G4 zZ8Q&_lD=rv1w6r+J0K;n6OvvHAQq6Pu25|BZJ3IG_N-~LA8elWDcs+C z{2^8AX*5bdtlM#(Y;xia7ijiQh&1k^ScpgOpyE&j*r#$c6r;^hhmY1o?5}O{v_iow3neD`(MGKtS zJ@jqQN(>KSun^MXiYU_ZT5Ywo4l!@$-0qk zyXvd=7T1{*AUo~W$r9Ny)X{awkov8g=?e!PZ}c&+db>dYEk)!LcLN>MwJ><=TG+Y2 zDHRt5}UV<{fOG-ziLlHISIX zur}Yq2@=$f;$LguGWsQzzNZzW_l4OkC7NU`?o;etirk^%_LFkn!++3h@`T3s*EBW! z2TMBnElV_oklh+_$9+rLYsvgEmmaVv6#-Dv38FC??5mF#9eL^g8}TukVfX6jOeox{ zq3agxQB^BT#nAd1hs22Q6+ zRs>ggZ+SUEfT@O6q9c&r->W$NidI?9JLWp(QoLkPxOQaG>)2g{ zeh;cRqP|SB+bOck19#m=7Hf{V_QF}Mom8DHs)@-&_&l_Ol?52h0X(|iZAdJL9 zEOwkgG*PhL4Ijzf?wD9O1F_bEutWZuoAN>)YBL$80HYuy?(w2DzdP=ZT|=x{EGB%h z6Az#F9oh6gr^Ui~cNd*oRX;b^l6g87?8MEE78GML0ziJ)yOhrJix;)Y`p!m_C1$S&an_L%QQI#GSx2in#Ycie|VANyA>=z`{@Yf^BMd~)17LQqNs&+nM0 zU3f?85TM(D?X(N}&Me^GnHcvdr`S>ow4UN_K^4oG zBqL}UuSV4+-sgQV_>pvocc0nCU+K{vn=dmQ<*6I^#ZbrJ9-A;bjX5VC@u&$o>{$!T z{>I=W-r4A-O*9S!d1>EN0S@PYiLl2sIHK^|@-$2m1##ZTwEAls5B|`I8Hvw-&mhg* zoMq?z%gARY(7#5pmnm|QibGMv6jBvb!d?UaxRaE6EpzR6-A&_}A%cZ8C;Tsncfn(T z1wNfjyhy7&!qY)>Lpsyq1%zhc;q)p}T=5F@avstv-EWc(rjyjl`sAxv%xQsetoJa` zS$!zJ?rj+PbMZRnO4O+040+^P=Go@et9VFsZdic^5ob8lAlT@UB)Jwgs>qk*2VL{e zWcKn$mPs$V9Wz~TN8)b6Z^UkBSwnPA(7-AM8fFm(FPy_iq99Ov% zc7VS9##L#T5c{D{LFrX{EVgIwi8`>bM|O;F_jG~DcR%q3XM9Y&*@4HdH`kttHku&! zd&kqsfypG$oYzx>bmhb~wuxdJC{jnot>bAvf{0F*YJ&$zZhi>tg*l{IzJYzPWN?XQ z#2rgcTOjvz)&FLA9}T(A65qqKCh_5Cod0;%4O^^>6H5bF zurA}h{+b?W*u!`pE;!w;1MZ*6^MTO!h`3a|U$tL#*{uvRQb;X!RUjNtlZ;mG*ydM3MG&yl$M30H{a-L$LVWx$OJN36$rTIX`n*y$D zKJNi6FzvB>6|Epolm&`I$CX{e3icu0BYX^uR0m|Kuuo}^Z5H+ceQ&>a1OK#iy&%mm zan{<{%`v^|m9z5Hw-)VFH2BnruR|~a_e?e7HrQ$9@%z0oh?z@w(RI9GuS&)E_(^tsCyu*tLtVvTXb`mTGWGn4vl>OuAIh#* zeD^CGqJoVU=jL~BSCB$(Q5+{;nwm^-QBAQ`6se@*AQmIX#1r(=&EM{dIY@mWC@xcv z0&_je+_0zun0uic4rqHfLIkg&< zp|i`50J9o9&oM8Cx9ta?bf4s7=1EJZEx>{P;0abDhhjnNCXX{nautPqq?16ms4%lXo%yxX;`BZ1!e)Bu|NSxN483*6_ z!(DTxffr=mvA|2v7F`>lRcien`+VY_Ac_}hPLq5ZyI<^R5hkCRr_Ndlp{MS1^og@W z|NOBzk1rQcIPs!mf%O-e!VdxpG><%Sf33m(Uu+Q0f{0yZWIBW5Nx5M;?jy9O+^u|? z6mYDB$HY~|(a6w`@6o^J-`tHpN%dF1d6ul>=94&a*uT(Z70RU8trW~c++qJ=vf01E zr#0**De~(Qp5dQUeG*!sGDwot@KXdAgRgsU=Eu*SJ|n^r9n*pg{E*jc{`~{<<^e9o z!->&hq18a^*9v{zBm5PDZl;U0LMurvjsIV(KkoO&<2-(A$PnEp#%sHv?)PTIu-BNs zH{Qq76JZ_2`=D%imjl?UMlVkq*|6oY@X9BuN7o? z<$;wx!0!=i;mgg#GDh*OMO8rLkJOomWjoM=>3c zyP_Yc_1or?&YY)*LAgwGCJ-b(jhC9YTBYp!#jW=u~0hZlFHZmWeAb82m=;l zE^BAwh26>(FGF7wdE-BW#LXl*L3>ax}&x%b0+ZT+v3`F&+4ZlY_+=1jB zee`PTdAz-LL)aHsv-;>+pyY%w=XDX^9@A!-mvM4gznnM{XQ3Ju10;GLU>^*sXbfBK z4yfnN3I>n($l~o`SgNhbj6tjyiN;)NKpY1PQ>%OoWH<)@#P;pDkW~cD7g*g%c63LNl4b&=6@iz1KMeR^^(Kq;QdIEq_ z#aU0CTN3GKQuuphTIEuQ8XbeJn40Oavz?h_IBpPG_t7uc_!uGb^D{HrNXD0}3hST= zY>Fs0p8^KuI9!2Ngdwl!E-1Kym8vEro#}V&h4@XsD^`0Z`>YJR9SmJgn841bdua?6 z>zgi+W({8yNruTs9)0v%4hk1f4L~N|*d)_$`zgx+iU=5d&z7oTC-#0UR1ISj4(6M% zf361*QLFCLVu<$3R7baIB6x9}Ws)ig?i#=o2rX zM-_Qce4p!Au~@5&7Zv$sDs^u0BK#aI-07h&i~E3GvNtq|X9!nYfml0$&jg9qzIYxy z<7+Q{!kx;IUpM zy3If?6Aej=XNmy#WXK$VH4!n}2G}{Q0oBwh7J+j7${a)}_CiuPEC>FBto2a-h;pLD@MVtI{8k^mifoK!7zM2F9q`17sQABA4x|QMUpyRd~kyR>4&v`tL7r-zTc$<#elXZNU=bI z1Q_^JqcpqqgW*f8lF*-m_JhFqm|O@3>bbLVQ*~L~B^(eY@id^K{aUB^2BY=c5SAf3 zA<7QxW~yOFRx8L4Ibt-9^|rJh8yBM=d_;YXeEh=fckY0qLQxsmBlWwY)=a*r%q@X@9-hNLM|P1r0)4$GwoBno3sYPW zRhfH7fP;X{LOSGjkW7U$Ld0oOJax**^VuvY$6RsuUkeh}7TI3CirX`c(mNUw} zz?;2}CPiMf{|3sIqAe%hRa+?9N@q?0wZ5h|9NXfwN+2=9z+0}m&>v5rPVPCcTpH(h z`ZTFHqhr#!wExNNKX*_4>x#KxT03oig6QI_8YHe+9*`dPFmw%=q$(f$Ip&fPup1Oy za)FW><$>DOqY*8_oqk6+_^(r)Ce9!+7>=oaoYnVzhvz!>8{xG!RzPkhoL(&;gB%8ZJu1NzOo$Rp9k$0g4I76=K_D%>` z=h28YJTk*=he6f=$;rnHCki~$!-_<}V(wYYUk#k$x#}I>EppsOb@B7(e)L)=zha42 z`M{?^0K2R{D51E)*ORK@7@8#Lugy5c39%9EczOq%Rek!ISYCEK6xdBP2$!0f(tG>PIrV%?y6Y*;9I_M zil*v`r(=0!+uqULe&7AeF2~#~{2&Taf#j{XFszf#+>R*Y)d>#BmIp6?Z6`eJMK4cp znGf8lG<{K4vUN!G+aoVOh3*9Q$q!)m1=CS|Y=xoVV>1DV5zMn1C z{y6I!1@wA;VgSCb|H+mgL<=}6P>&(Gl zddSIbC-!mHnD{skDR!75_oz5*X3D1zsXDx%+Snjk*x|L$wMUo>)l@sYHT9D7(jo~q z)MD2{KD~1uI1zXf_PJ2*(DKGUS5!jK&C{&?C-8^Aj=^1HtwWnvzw2sImCxF!bbj_c zO(7^JmU-TtmmsQ#F~(ftBU%IK7?XMY(75-k=y;1M5P>Q@0 zx_QujS3nPjBCY94_x?F&Z zpqb5+7GXAYyVnci<^1ec$ouA-bf+~=7Pj4k%zaWD)f$!@-5$Gx3<9y|+Sq!gQdOlo z0NKrZia{oi-z;to9TQKRY2udGoIcaUjS-)!{Uf{3JiG9vB966?L&*H`Rq24BIC5v? zV*&Ds>{?I`GEJw*^x@;l+GE`!_T=(uJFvWnB{hyMH&@x<(yrjd8zT#v97E))^e@*T zQj;ypQP=U1it(}FSa=898V*{Xbaix+XMz8DMt>LxwHRj%;T)Gm-h>@0Y7ayxD zM%;n4ehH|E#U7V73eYB92=9?)GdomwxuSU-bL~D?Zbj zOEN|-O3hU^xWLDWGnPOgF`?=vg<>~QWGxlf0iH*ipC(CC8@MVAr@uoMFZw?)HP$Uk z*36-uuPMF$8&C6Cu+usr7Aop)Lr@J9(+MD5p&5dNG{)Vq4gn;()BLc-|2m^N5WQ1f z;#me2w>`pMMP=l^1=B$TF=b0&=V>+{&hip7$u@e%_uPJc%zPXDv@hbsNh1qe>qJ4K z0CJ7$4eY1_s~W&kuopQ-vA}gwOU3Qw zAM>pUhSFImmczFDqd+^=9vd&hxNJIeFLa-xfW8;n3!2y6&=&-~*|J)?LE6o%@i^nHiQi14 znOU(axRH#xaQe%fcXb6t?1=^_nPS&dWQ~DdvqQFNj;2|$DZoJQF@0#BfH`Y`J>l|c zJN)~V1-Jjrh=>xmh?mJVZe=Gv&{tr~0XvOqd+yeg|Qyw2B9j+G!^2MUvY9#?=L8gq&5 zyj1}>4oHO>wqxzQwrDKRyBi7HbgjCR6o-Le$}TBx-7$*@3^uUjeIEYE{izyHYasVz zJ#y%QDVleD?VX47jF3}rSaXEzcwr!S$OLjF6bnI^d@2qLuC>a6C~%uwwF)dy-Q+is@u637PAx)3aRtK(rKQZAJ0X_cvw%}`dSFW^oQ<$5@x zE!IAGF`a9BymGnKc-R&b<-E%=v0$Rc)wC}xxuE zsEC{uH_Izkn?g`t5(9UbTPz|6_{CwX!gOw{!gAnsZ!BKBpLB*U6AaKzz<{($a7cX) z{Oo37+Tvbyk$;W2a6yOv@u&g8YF_sIMj4)k`(Mp*?-AmNE8Ta>@U_lumq&gO9-+S6 zX#%w5vQdRW94yQC6j;t1|8ml=%cEI#13Y=WZkM#1mlRU2-YhGc+=sLed`1~3P7PULl*y88o?j}~i|0TsFp3iN(DA`Vu$mx2NEeNgJA(`Ad=IJXRanGZ zL-#OEWW?iIm=3Z?b@VV4zU)`@J2En58zd&1O#7U(>?a@mMS|2EEL<{-PHdknz$iXq zP*JA3DINC0PNeo&EO0`#6fCO3s4+GxX_cE}(&pcTqNYC;QkSKBs6Q?9>=AYev*?Sg z)~{PtOE>ZkO>PzJF=4jB$Z-`iin4HpExnRXY;!F1N}?tOIypxm4XGywtrR1x5ES2` zL%SxVWd1&11Dj~Bx}AN>i?s(BCK+Q2_?ToA`=9z**PA~wHzrP-o3>ycx`g|_t|wXy zJP)OkeNbFkAEgUvQR+gP#cLP#3P*&4{HsFTRMkMRxa_qG(g%VC1A>89Ps-|4wXg1e zbu-AzZnd+Lo5@^FVfrU{874#7^@ZhCX0kX_jF-AU+_;kxyDr&9%4U(%CR>nNiq%o% zgYmYYTskpO-x7qJU|qr<`koZ!WuQ8?mDfO!Z|;L|4Q@H09n`f$49{o-*~-sR>%ukI zor4YV*y^nBzbW(_meuj{VV7c{RyFQW@S!de^}AxM1)6`(g%?vcZB}fXsR@%d*`W<{ z$7!aV*J`g>?mzT%BZS%(UYtQLaNCSKvGa1@4RkD?{MqLtToffx?+vM57@gkkuO|o)URsa&@ zHwbq6ou!A#DUvgP)hxqjV@jh7M7<3zRF@F1$)#7%YLCs}Rd|sjyC3} z|6;4Y<)?3R_!-OJ@+5AB^FDyJ!bTVKqN}HC@|^eghpyL&NvbD>SRW~Cp>6^41*BRO)YgJa$yY*GPsQtKGoz7hK-|MASHY++qrjJpGBVnTB!X{5)#xfEnvuJ9M zcm7sj^gM#Leg8N(VnDDx@mbtVv70E8f(&zm40fD%3CosTl@|CP7H?DHx@=%8J#Tk8 zou~dX(-YP|iKh2C`W2_|-+4C)>wH8%WsSgae>ilJY?(#&n*bx9VgV^RR9yB~PyK59 zPv6_~vx~1U|Nd>kI=4bz9^LKx*#CCKCEktL0(!#RNs<9b_}y3R^TjvQne4CPZKJO_ z$hp%`w)IH)!f-!1av)q@5TqRb(?xf4tywNnekb-yEGW~~s`Fe6!j1$Ul-7h`wP&qj z1*^eW54tM);4Qw_3*r@NeklU9o3Mbn6;qA`xdEqBgt`MkZWwyU=&>J~D+4;M6JkN@ z4+9W+VX&Ru6&5e*6R%eL1g#9dmi-ZPcG03fSCT_-wdvS$UDg*WcqrrJzQeYEZOMkGqJ@#T?Ii2iRJ|~SoOsW(* zNa(lu9P=&Y@AFJrSRXj*QmX3l(S@81NDOR^-NM^1-S{_~`PnSKZi_AC9V2}{%h>D% z_va+>vI6!**?X>RnVu&r5gyoit~gDP^ZF_o-QVpnUxA#~x>#6K(){}99R0Q?dp0Z~ z`u4L!s?*Zzio`%ow(Cuz7a{zD{$GP>h$^2VQAxPH(P{>=se{*SAUnV01l}1h|66yf zd6UTt;-y$&LN*TakuMg z6PmdZiiOgFyHp&e5>cwEPd*rl*?M$dHKfg^lA94)#wyWdm>uwAuFd1}N+Rn$?oozvhF?9>DCI{sAis>7WI$Gy(&3l;$ z2g2V%3wqm0P>PAam&R~4_8Q=Y8Sr~#`Qe*sgYNHPJ`s))dpO_&hu-W|`{% zs39Y-%ofR&s9}Pg1j~cZNOU1O@fpb&c-RHsYG1O$8}Ec#2fyx!OJ3a(?XfA)nQ%zG zJb3)~=>er>Ed9ftg?-Xu_fm2E+%bRbSGdS7^Lo7SJi#u6l@Z+$_aoM`Wzfgl!94Ob zAGf{M=6!onrj#)2g#7rv>fLHdS?c8)4Y`tE#uiwigD52Tc5_ zLW+gV_AV;!r24c}H~%rH{WZ!y7h}>|H$UI4Lsl)l%^*41F@6KT)iZnki2H!P`qT8? zrq>HR@CrwIRa07Br@^aw#pTzpKcD*0i9Ku!)Q5RtyXg`BS(t5DJN3m3EpptbJKT4E z+Lf_(>Yu*Tv1cqhRzLd3m*3$T4ao0K3>_!Szhnh-IVNUe6UC-bWCIWniF#<r+|t}5g3Hmpz<93@j~7(LBZ<5$m(cptvxVzEl)EpRbecd+9N%r!j?#^%Y_DO z=7MYu_}y65t5tS+91L~@(5!^g_-ZowPX|CMtv^3sWdux6*88a>&x!Y94JHt%qS#7` zl!1bOpUNTb~W3q*yZ0R2Bm6#apV>CfW32Jc7cK(Aoj(B zWvepDCZ4L%{La-po8Yt=Cky$3$~Ue`^Le$PKzw#?vv^}9Du#9myGZ=p0y>%5{pu-l zYtbHhG{O-E!NS~e1BZn{9B`T&I4&%Uh}>RtHq!H^xk#m6n$$^du(IiV?l=d`j}Rz$IS{hT$Bs0Enb#D z{i$?`;X+43i5qszmKX@>irUwfkaMP3c^0q)t zxm&rLzBCiXU$O73MLyzD69S2cTy+nP+`g*>wKP^P*NBt-S{;xLw-I1ptXLDkVUNSn zkN&_eF+%1$jsH+a4!gQw&Qlg>MCx>(=lVdbvaM=QbgQ^ZZ+X!_qg}|M$x-xP~(22D*sm z5G}ABGHJlxn-sDal*m_u#OnaC8{*H@S0Lx@nc+16jm=mPUav}b%a>`6@~frCe9uEp zJBbIZiLw!QeO9}l)Bp!rA5co6Hmqh_V6zZH!NM~>%|iVW*2y4sOgHnPGQ(?){3#DcJ+UWa(jDq{KhX4@0Z9bfO69;k58d ztK7bz&1X1tz#SB+z~IM=4n^U6j-MNAC&SDzbN+I<-oR-xoHzV6=e_fp=4=d3Ydb73 zFSN*^vr&y=tV!f%I5J> z2E=KhVZpN1$=af;6g%17YOT6Qc7$jyfrs1T)c^(HhT`x3*^im*`N!Pz=<^N{%xi?X_@+)A#j{XRY-nujrGdx0+&i7wJYOeB<0k`e z(SCAkVKLoLdKG)Z57Ce9o;p1V)!CVXif>j1$@rCE?Z(YX@BLjfNo|t z@P~gEkR5m2{)*!rG821Yy$v(D&EYTpVJq)Is4-BST&Q#w(-uZ&U!YHB2h7a%M|H7rZ5jy zJkchXOldkeJ=f0rSpKD9@vlaWzWwFz&iW>K=t7)!x3$g0gsh|3)f7pj;@bbYXGt2r z$p6k)r^hr*hL`8Sbh2BfsX%8OLX-cr53rO;KTNfSBf@!CQDT83VrWrybS~tCH$YS+ zL68$uE*YWgrOiT!ho~BST6k@A4T$-5l6?!%W!lTHr5`BvNC){KQcG_Px$5WO^=#@u zu@4|Djgqu@&-Vh1Hm2$Ijn!l~Hyh)`K&dyeG5aVMBFLqn@)xx(>}K%I;1#}oK;2sc z8faMoKpw!I1Fh z&*a+{^r|snhupKZGP{Fl>aa1jE|#W$s^(@`}4lYZ@-=6cc*9CMPBUprH`?$p=hMr54) z*0-;ajbE~UlQI*u?4;N%ifpIiP?E72x&@Bv#U6(v%e|XOmk`uK!Z*V99Jl4Qv_U>L zpEtA!3#Rr&4f1=1b~g!6PckRqJh2N;k7MsMmVxu}A9Zh;yC+XmoH=pg$bu8n0s+}> z}_nzZPekQ3ytJ5+pQ!~JwtG3Y13YEYO=nd4tr28H_f zOFa3;CB`MdC4nSPw!}E`?iX0?CdB8{DRv`8k|96>%HyrdFQ}`AK$fqV*9D$unji8d zI@UOtY7usad3`EO$S+A=OKZd>*Tb7NtF??g{=5+B?%4@d{ZgON!*gYc814Y287 zVt8fc_dMCH4n1f2d!D4#%XuHRQ;V4Fg+^oXT4&*UQZ$*daN=$3aT8~D0 zBLjjk|=;wASZ+Ic;~tt^ghN26mU(w}&!&pxmp_$?#- zli71Nm|GO5P1#s5b(>~ic7xt7DboH|`E)a#K#+KrzB5g^U6Wa|4ge;%Ov?+)0GPzo z&8WV#b^i~I7DeLo-!n+_WERDVH!q)=Sd?oNdzm5^u@Mn;exWW6_alFP`~N}C+3Oqu z;!EJD^s5LuPH$z6=dl{CkvA4G&YOv4Y>mItJh>;G>1N`oi)@(++F5JpBRrk*66h$t zhi5uunM}{Tbmm?BfX2WlpEc%kONDyAe!}~EJa~2lsiyp-SP5f*W zYkcGhy^lqbb*~Ky*Dlt40GhYUVn3XxRd&eCS8?lL_iSrpr*}=@Y#DY1HUD+>%T5K< z$rSLCcnyNjp%5`0mZ2Q?Zp9;gH<(s=c3zhU3SDILilL0DJp6ku~T_rKafo2(=Fd)86=a~rkTknHoala z9dCh-9gD}tE61Q6c#rB7$yDk){x$`4fuKos^ih5{JWcS32H#-p{(W2ci>ELR7Dlq~ zX-?DNytY4S<|k$5`imB14v~^=HCZNt_|&!q2HB_vY0H92-*`98`yWe?iTH>ZsS4uV zR!@VNlzuw7qr<)^ur%5Wf{w$#*tg%DE6QozT?@QWJLwXF40Ly(fw(pH`nRq^y-8c_ zJ@@w5Pnj*UQm_4<3G5v~Wi0H3$frDLAd+>HQI{NPn&(z^rNi?*+yQKKJS-v4PUkrR zm>RqN&xzkLdcdzA9sDUdIg8vhaVjrT?0Jf`LblZ>h1X1DxKo411lZDpkfNehc~-26 z7a_g+75_}7&JCmFTBUB@<4_Z5K}8U{6HpX#N3hPLDMsgpN817q0tNd8 zaSL52%~vA@r4`%VDbJxzKIHQ!n$sLQ?-KR!2bb3P7{T@PGc(#q2Dgd?=l#o|8w@a>k$SJNdpA~s%_=dC;Tr!pEQtp=-uB!>4-aQes-%1 zc}?_vvfsCkS0vKGL-V%bKPy1|d^nk0VJFO+jA@S-uLa(&`{VdtT=ugl((lA|2uUWQ z0P!=~#}pZ%;;?he;9sRfgBKdYQN<~sfx0Kfcfd7I3b`%V6oMQbXrb|ZD#v<<@8{rR=G)ZUZGVV^4DYws-rQN zc^W7@Qo$MA<-M0^uuKNbsrRukfU;i`qbP(9Q^gCq|u32sp}?)XrtJ(6ge|aQ-|&~ z^dz?OuuRU-`;2Rmp_BoZsuJny*d!i$-0MA&p6;P%4OBvH05a8GQcbHoMQ~qH5`ItG zOK)KDNOMSKpmn(8eq_Gp9FX~9HWqdO-3-Vud5pQ>#~@n_Qa=pH$K?1nMjpET*Fj}zrNc9SDPtz6Gc*> zDbdK~r9seaQ|Vq>_7a#KrU{(hPI6?iRz1CT#~On(JzH{9#Bc*f*XYN5bK1kF>+761 zX-_PW8@4NlWk{5|E%0;kcF$3W7-+DmW3S@$D`PG-G!TXXOB^~gKxpX+#sn1M^yM_Y z;f6L62jmS*rilZ3%ngJ?t7t5v=g8t_Pji*?r;*~st)R$WDy{=IH0M=m z?&CjOW3kg2Neio@awPe@6kZER@1gV+($KZXHc1BP63-k-Z|H-CY3`2&t+7YnSn;DH zZ(RPni%Z({?BMZCEoqHS|NCO-nYOo0vI#d6kz|5K+h175P2)cv{gwIB!=-)EiK8?Y z7AkZ_FkOAlvr>G;JJsuRP!kw)>1K4Gh=fc-_Qu3>X3u)aJZB&~zx~bO1@01~ubU&> z+)0vNnDAY(iK)?2Y$ipvQgJ9zyN6yAg>Ee@NGqUp3JSV^0in1SIXW^U9*|B2=~_LD zPd8KJw;{@UM|JA6V?9778W4wbPtEw>z`McbT$WF-I8JOIEb?1+$jX9H>oqAP(Wli1 z2zZgUr9)OIoo3Sj+lX|EkpKMc?=}DP%m4n>AOB0bl46%rB;I7A@x|29HMKh}uP49% zs_5#Mor>;Epw9iQco!d6i594!$fA+17#P3m1i3V-5?9kZyz2y={7>B5qR)ArSeyaO zn;XKA?AYN5u@({L5SiLMTYr~jL`+iNG9w{2aJ!3faVdWApYM`*Ck~`&O_qxkirqkw zwb0`p(iK|cdq${1k!UEpDVFsprei&LYEGaEvSiO>%*$w9c33WTA zQ)Uajx)XaN7UYWqBG~{!s`P|y_UIlFh_W$;acSlre0vfjB%cV3A16%;Z$dglhJqH z$MI2J-zoFlAD6I*6Nh9iWC}-^QlPCovRG59IP2LgT*0#eoKpo4-e(Q4Z5e;6_a1I4 z`Gfh=W5H%`0__xJAL(Wu(w`HZ00IV^LeA=oJ#KsAcMPWCq~a13En?8y@oT^SPqyZJ zH8l)V{ie9x$H`zg?_>4G_kYQn$5EZueYFr7E%yfc+BBfY*4qIr99rkmsLqhJZ$S93!we^uNcl}4W=_S7#AlN@ewR3}c^95HdN$|)8I1NTsIyP;wm zsZKxtyQ|Xn*qd^oHlMYHTIoJ0UGKSiR#HfdYngYl-wGf>#a6MTkRe{0-(mk7U=?@M z88Igp?Sts+s##l~+;LO>5y^LmndE{nQ`b9KQ6yV);Mu^ z1ZrO=R32@kSTK8=sJII6VW`zVBg~Q9Wrn0lAs4){Mg(g>E|6jOEV|G?JM4_lbQ%;+ zvpwm&IOZfLdqTbYo%@^2dpo#zrq1hg6APUnDN8?jbI%X%EWt*@l%;2u?D@gXB^TLc zULZsT`zcth0=n52`7S>AuES&ln@cAK?o|x(_mj++&0&{A91Jl|sG49l938a(VG30t zG}@5Y&TlzN*3KdYCXZ?c#crWU8Wm?qZxo4Id6&c`{J!vf&@6yjtRpX70-rStih}!8 z`xPj4>)@|)!sEs9Tb*#HV-Gn!tIq4Keyj6gy|)n@t-q>zi)?e^1qOIBCM+<86bs@~ zyQsKqUYBq!khz`oD3{!GKP=7`WwVb2Ad0GP5Y+LI7^9xl2_UNDS1PVzd!eJKHT;?s zrBa)gncBqCU_xm~$(i}l-VjLWq(&GeTN*$B>j=3|!Wx;V`2iAh?QcLcWEY$`` z_oGPa0e+jjGHMK*(tOp`h2@eQNj0@^;qqx^apHs|$67#>(Nl-gea$uLEJ(j~2@L`W z)wE918s1OW(j)vN!G-W1*rjy8U4 zwk_>{&bxS8YX2L0rDxtHa+KdkC(J&@>waZ#G|J!hfkt~j-Q@uVuIbEf(DmqdeMs`c z>H`}k0}IA4ZC73fc8!w?EPYM(b6nF77tEV3M}}99g8je$YTI6OekLw!j1z;!0-IBZ zY;#xzy4LjN5G2>x?7mN`FP)!O8>pX8D?#C?xwkkj6}|VGd4bsr0tySIXP80xW!YQ*i>9H+31|6*i#fa zPQ@)P_wIL1jeJPg`Cfv3O+Trjo8=>X(ECy~LC*RT59`x3czT522y*L3=I`>@2fe%`@Y?|C<~0Upu+2aWHs-QNw3EHbJES@-O_CrX1C$iV z)0y$1F7dX!TnipB*4a19^QRODuo_U4Ej-qqdkQYkKT+kI|J83A0afYs=$E948=#zc zf74?EsPh!tN|6>SuI%-z(yL4ff2(WGPZkJNvWbX$~_8Qe%_R_*~ zE|rwif~+S9d_JPaulP~$Z&V>hWUWabC?h3K9AyDpGa<^Nqu389a)64vBFd3;D+ZY) z-c{z1sz{O-hUIP7y=%m6Y`fAPJ${b4faoDhwWDDGs_*atpNiBCq8F>m++?ijys4fg0i=L=KlE; z#2FpWAJQ&Ya6-q-b%MQD%_|O`?!RzicgI2%LY;4K_)YmSS3~t2aMP8FQP8DXeq*sg z5Nx_&Vwh77V6h7btR2KMShk+MT43G^%4Of=#2G>hO`^SYjc@~-5paiI2FeRXl14>* z4Dc#pm+=996I8PG%o`@inS6tf)MbqdY%Xq=x3i@{#$+h$Y!2z64+q7^^vL#zD!k9q zEBsHZHG>MBV1#tO^cMbSH`sAN)Dx%USqIi4=6Tn>xxp*fJkam7%Z-K5{brA58Ydd- zOYTAJZ8e3BT#sVP+|GKQRgEfAqY^~h7uatupZA?|3o}pWX4% zy|M)Cab%DSzwL<}I{!BXu~u1&=flwMS1h_ItBo{nZhm-RPd2HX1$)bhdx)bHTT2lg z5TnTlL(`dDph8O!U3~TQ9KC7QDqEq>@POQ?dtbpqWI+MM55iJCHOv7Zmjq!hk4w<% zpyO*!kaBNBhyn`ifn}|RPV~7Vg;t0bIYvJ&L2SJSq8eGj7+KdUH-u?AnIch=pw~Jo zNXy@jRbT(Jtn)M8v;3*ge}~)S>$F*-pMPDp>?=kTb$tC+9@*!_TVx7$V1@pr;)+*33)uNm9dwRh zKn>3ThrVp12VOB4I^u4)Mytf1Ipp?BI=9lrWxP9rtOdYxD_G~&C9LpnlAHxvK%1Qs zN8$VV2!8ezobXKzwTE0-V1&`o&*dx0PA7&D7{v)Ns-W1t6e*_S>LEz>m+Q;sAnQE`R5Y+x2G z2|oq()|e++=YgTNT=fobWP3--=KZ8rxSMpyGUQnSeZU2XTcFS6dle}H4Hn{p@l7^I6k<8wuZ$NEdK+3llXLqpu! z#g*Z`G~QhjZpbaP151x_yd8l>OhqtM znJP7X@-^}tAsWbBeZ8dZ=W^6y}Q8V9&o)v$Xvz}ye zLyr^31$8Fq*-Np&*HcKvA=f1K*8+afYr=@~2G+nMiOqJ01A!$TV5CWL#R~8XVnaW# z=5c5_&wwz^}XJ+8m1mI3d6D^^B?#$ z2ui#v)mmkODDmZ!dRA@w&*!tARSwqSS;N@K%!2&~zQE4K_ z9kkHqejc6{z23(W5B&3I&nlQae;B)8VHqY-=e%DxCmyxH<(aGg*c)@e_aIAriZlpT z&B~&0LTyQ#eD$pM*fjny8IYe;Uy&yJjJhQIJP0ULwa0FiJ@$_m9g4cQs9sg3a_lo} z)t7I!J{w=!{R^Kspye-rk^rpp7Mc96>fNnjMn^>P|1>+uUMJo~ePrTV9HH356se}- z5@(gFwyPd92_7qFJr2DWn(vk!)*kzi-{#Y;$Z;R`yyKYyjU>71b`}e#aC3AccIi#f z>OkE!o!d=VkI)_I5@yjo;q9?Vb6x0vMAjr(Bgfi@z5L7Kn^1ezAzSBBB&i7*;b-~V z1@ksQdvd_T*@+OH>M=KH^WJ$y_)8;#zJ1kokX&+N1dW&==u?WlMUfj++@EXc>rA#l zQ^G$3|Fp_3p)Mqavz|d=GPbB_Om6u>BO-Z3v@sA z3oeV(!Cxy0M}DykdFGUAdTpOIEAV6chfHCFe}1?>@NJ`&xcH;Q2y%B8*=Q1jT``kg zHj|7}ak}qaS+df-b#7sBXXrAB#*Ric$~u_`3MBWbR;SJ#pvR!#wmlY$BQFMD3WTVp zAln-R^F-@|uxDtG^pGlZ&MFbQ3f+oQ_a0$tJ7(VqB!Ozkv*OA+y?Xk&j4T8j3 zEk5{Hycn@q=k>0~Vt4SRDL>dlXw`AYCsRl5QnE!s_UK{)eG=<%+Z~pgmR>vXQKY z5J0nhkZ3(N`Cx}kdu%0rlrG^ni&sX#>MX$St4iPF!XDvsC;xeqFzNetK+L4$+Iq4! zxFOb@&-n$ZLbkyEjCpdDP{W-V#)z;E75CJa!TJ~^)>`HDfi2K(mlV_NfxQrq+>fe1 zmN^C&Ci6EnkeCeNkQa>)m!324Vc-H1C(dG8=x*4{uca|Q7$0*t@-B-1c1i~!6Q0gw zOYr2hcw`f%k^?X%+~xSgH@8yh=ENdgfZ@dJjRk%bj1>;XoPMQ;egLk;PFe@`UVC^M z!r{;X_X3Y&uJ6A6PBJff8YxmZKm(l($790y31XLbPyOEL@UC8y6ixcLIlNA6GuD{| z7#>qBM4RtZaY+&^c-H#0h8Q(E=YpsvgnW-fEkinkPHUgIo9QAe1?N@AU%R+O`|YYF z!|p91TJ-=pyEQybCetI^JSW=~hru_BBptF5aeE}nP~*QWIte5R^vvetmla7cE_6nc zD8WUmM|i=j1W3zqSu@;(^<0J0Bta4UZp{$nek(pANst5%aTX0DjgNvKX3WC2^5P@# z7sP0D}x%R06t`*F(ob$6UrE3kIpzX=|aA$wH$v{t>J?DOq`L}U$J&ijzp8QKNSZnqb}?GJ`MchRhM|5;&{|DwWH!FrgP2- zH{({$bO?vMvWy#by!XehMvOFlrS>mmg%hW>cbecKjbc+Nk_2@)0)1bCp~`p6rB#W= zyXbRb0q=C04V*ixA>fdSwmhh12uxxVX6nlBdzo{HSfJ`CfL@U^{)7A+{@n;@z%XcZ z@ANAY)d%XkDQne@o_qP(yfW1p`;;$P{u`gXWA!9V<1>iA9H&We-sDk=o9x@>3`I{@ z{yK3m#R6;5Dc)UKk|f!$&bJ1CasEI1*v%StP8>p`M1Y5213vc$fOfn~6le(UCMZQ<>7-Y0`t;O$Hlv_LVaK2*{n!}3pTa{IsG z2Cxz?w$1$%?%teQcEvpNKjIPH864EkI6A^ zQCufple$fu+jA8AF-4jU{oU=c)gCAiTR>Mu9TVh8Hv4P&NinGs2%H9_M0bj+smg`j ziX#8wmurP9IT>|Uf^+D6)&n)x5uKk zLI0B9f@r0B4J>eb!UyPn@8k60=qfiY{|}?z`2X4a5`d=CYk%+H4atWg8-d(ApeO+* zh$CC5h>gyhcG}l{d9SbSYdc+r_SNY&ZBzU2iVG@^vM9KKvM7rnpnw}Ht0;_)0uCyQ z8xR(C7!h0;RQR8h1SOGZE)ZU{Zx*=gmvFz|J?DI9`JJwXzc<{Tu^^2uq3+82LbuM> zDVj9rc_k8LPvU~zvBK~vTv~I014I)Uh1|Syg^PREj)7&N0H=j1@{NP~?-D3F=u~W% zB{G}9X`c|*DBnpZgyHcrdMG5FpX6pJP}o6Q$Qf}x5T}&U*Hv}=gfK4i&S5HymyQD| z+&_%N9e?TwFaM@d1f@{}UnZyRc*Xp&iBP#hvCR}|q9QiSZU%yeg3g^~0qFS6bnliA zFi;5H2`s;V%?riYPF|=!Q5-kc1!ehWgA;DZp+R0wF8K|(Ak}IfS@Y_O=|c|Lbd}1m z)gb=_Mo0xZwIPQVQst7O#A|a&H}IvQy6Y#WT(H6)kKzXoIW#kPtW`KbXDp~xNBIr; zWYX9ycUnI|cXW<45sGaaH2om-M{1yrG=DAPSNBwCD0v_O9YEe0er)GMHoK`Q-# z?f_sVczr0|=+sSrKoa?z*;1#R8CT}5bL-=;gLBL2tCMo|~-bb$rA zCp2{m1A80k9flmzLO0;f7Vll6E4)MMhpyuU#4{#m+*oz+(3@VXRbiQ?p)GrY*SN%q_0z~5OxBM*o-6BoUi)^S z(d0?@%GgK>UYb?chb9J68O0u^pixi6JGII*!7k|*Z>-44cdJx=%C@*4lU;+7$V6r} zdtT5>Z}?IrWcp_Fy5yCrc&OjmJ9E2N)Kt9gj7yc@N+@FA?o|P#6Uf(wcP@2H1HVry zeV$h@yehoGOXM{|o^P7_4PF8NtY%^m#vMRowrM!w>35y}k#0r+HNSrTJ(6I@Hgk~) zBr+%#+ODKP+J$HnFICbwFPE=lb}zcR@S?PbSuN_67l@lQr-T(M9fYqRNjJ-S>EeY2 zbgJr8;E_%BUM;FnWZ0gfe5@1CX%@$DF()-d?bzR3>}iCFdR=rm+4j=N84xQY*!%J+ z7Lum2sfb%p;`g^UrbJ&RvWcye6uPev3_7$=Pf(XS?FelZ_I+k$$X@A5l0K`OE^=Qg zx&*y3prb9lG`#b(jhb`s6unXK^X@le7oFAYce+m|Mv6Q&&tpZ`6XEjIkP~v?t%x^6 zUvzFFEagOkX>eL1oX<2WGQMW z+svcag>};h!q*C7r>7`e#BFkIT2LyD`l^J^gKO}5J#$`PXvzFb1nUXP=#N!hq=x=$ ziqd#~NQET@!^i*f`1p<>q*OHcY3dzwMW>W`KpcTNga?(kl(_EVl*7pZW>2KZShqMJ z(lX4nE8aOR88(828(T=sOQW2CDQ|>wN}xb&9j&x-LZqU~53S;+9- zr78$+mK}OsFMM=NG?NTnOVi1E0kXShs(XN5vojD|llHoI(-YVVA5r&`RQjrPn+n_5G$Kupk!sgSe~O%@Iw0za5uA2CJ0G8T`Mj+`w)1j5 zdEjGsx+t}u3}B-z%W`{^FtYvDQoMTYzjwaqx!fO$Tj`j9`p=a*CH~v(g$Z6d-bwZ% zzy5Q7#@#!M4t#!+^n?*<91v?GY>qxGC&W^h>b|w+H%9B@;M*>8a+90&VaLiP+N6B# z0mb%HxR5I`mV@p07;uTx=Tf!PgNjFBlBCkBAgczuF(pINO>%f5 zvxSWmH30n$R=Q|0KOtEJoZxjW2h(G9zYRIR&Wj!yRC6O%1=?3 zLu4Ho&eyo+O4hipb%mOSMDRCP3Q}D1Uelft@8DCg%rTZ21SH$gW%i~&5XfM&zASh`2jrX<2cy3*T#s7$;&UB#+}Cq@ z@$L7$y?V8Gy1BXgLL~uPFltwM^oxO!gN_wfsk&($ISHx2YxxCY9k~nrGw^4oWcZp% z(_~g+`SAjI+#t1b0Lxg;On5g-XjE8>&Tl$O)^Jl;cAQYP(?s$mQS2s)Bv27Z{1QnS zIDfFgMkalIZX48BFX#0K>X@UTuhzU8?Q_mnPTJ9Cj2mo58^Jcm{BQoG*vANt-|nuC zB|Baka6ogX5x$gSiiI^^J{56Tl@G?sD#2dJtZU-m7VM;JRi#c%nk>ibbF2J110($# z(Lc*ThOkE^@8+F)pGyr*_$Lq?&kuog`QCC75CD26{n@KUY~lL zp({9LUs^Yt+)(pa=9q6-_hB5aYFPEFI++nQ;@jfikhqryH3cT9NuyZc-%O$+bj-D& zlOB~G`+~3fAO%d5=D70#@aLW<&Fscm8=Y^cZvb=D9cGLAsxXaB_1ztwI5|`wV}S^V znI7|{on8ks%}ZYGwurz&dF!n&>SKb%5EHz@Tvt34RtRGP_B-w5HM;6tyTg#l^MF6N z@TDcxmEb}qTeU)y1@56qutml)>yzO7;Z{~h4g4VUl|@D|60+o@R#G}zY{-u756Iyi zp<+%^>5@eMR;MlMCtHck{;+h(ak5r`#IPHk8|Pu#K9Ii#)eDR1UO`^y zNAz*gKIt;AKGL9U(j@m&fP0!UvM`S~l-PDK{fur+B;fTaUH zH_7gW+6QknX>jwN_@VTg`|u@&Z^ZD%h{v&E%v|?8YkzFQ#j>Tl=F{I3^Mw$X(gFKj z;t>l=BttbDw#+UB-jUs{|0 z{Enk}DfLSN4+|CHt#n`DHFrb9B`m{7&(}Thhk-lD(05I{-y?O}e#Lfn(!gPTS^lu( z;5yA?y#DnaU-{llqcVELo{T0bcAOapZBIrho&t*9MUkCU1SYItx1>ZS$}f}7;_s6h zcq>=?EfZ)zRt<(=(`D$=hs>6#bgAcRQN2owIWOh&x*%5{Z%-5S6J$1oiV(dB8Kd%e z<|J?z5W`p)@%x#dFt(GHHOnZWt0`T-SdnJ#isd3s>^JsBEO?ZgWtga4Oc&F6PT)sx z6~^;ZAhXsM5P8l7T8E(L3}S=RtC~N*&wSOx1rv5`fmm20Vb&aSmFd%&usP#a#|MFE za;$dJ=BV%RPbn*8hv++@(57 za#nteE_Ess$9ko!s|DF~8{`?>ehaz#JLWx-K;8j3OfD|S0P3GuuT9>MBs&C~p<<`# zyJs~ukS9>;RPJ{Wl3rsZkd03gow8;3$iZ2eBDj{l4`;C|1pE4&q{{`Vbf@rs*m<`d zGmgwyJFSv8=#b)7FDrFAG9ypV^M45%0p6Z=Rk~&@p&_fG^R&w0aKsE3R_t09Khp0r z1xh2RG$-FaLeja38auXD%1z)_NU`A8-$g|<2#aTCLjC-Ja~WNv)Ls@J3d^L26R@?- z!5IM=0V&EI?v)D?r)jY^f2}Lt^3W9$Op5e5)){K3kVe|pVS1zBgu%Vby;j}SWBC^m^Ao3Q&HmjCrJHwC9( zub|hEW4u~b)NAJ#Ko?M2TO8UeO$^F{f7f{}_g~|>#&wc(5O?{(VcN!Vsb%>w%BuL4 z{@$BAzGTElvD?G9NsS%jqum4_=P9<4A`Mi;`R^~rmavz^m!QSoQqd>!Om!dAP9tS{ zEs#B92arh73KuQr1Vb|-I(tSNb5C?n1ifK}<$m4JxgH#BsF$*!Q_%{S{jdWQN4GME?#3 zyt`;GS-pnhj;!YjmFHYw9o)DfOZWEL@Bi3{tdOPcbI9eFMq52F!C^PW-k?aQp~eXF z%?A|+9kn{<0j>AxCrC=>H_PrT^c;4OPYJC3Dau3Y{9rxHS@!I#pj^rIxrR)2^ynG7 z92e84T}uO6`AFIUU_?%aPIa81PYP^9EJiI;T!0`3-8zua&a3h>w9<-}54ad|ErBP7 zz8rv;Ch%i{NTi28poV_nURR{0PAEH#d6fnh8+==a9Co@~(v!aq`bAFHyBtqXe>h4M zV)D^x0la3Q-7k)QYMuU)_;M_)>Myc=5Fkio+7-#+E1ehPTI>h}7z-A*K{ZJ>f19ep zuU2{tEc8gA6`CZiXt(hKzxFjp&8yI!&bqW?6WKzo8HNvU3N{2_absTS9v%vUb#9>2 z`V`4Q25O+PfK{%=bZNkS-;&_WNpk;=K}wjt&$ijMti!ytC4U}|Bgc-RW5JPAKI?Aa zGFV&by*QV}RRRNR3(AZolO-;OXW44J4Bs$Gzu)B_-8RE$#e8+w&wfd^ae{q+sNBqvoupNrNk~`(&GExo@sdnYdd#=#b-)ttz2T z@k^+E(vyM`7d%%b+!aW)WN%3sCQRVK$O$!(!fU} zp#lPixv$kK`eqKeO!8D)InQ8CK6x)EW6Sb#brfelSDkWI(5fJ$2wU&+5V(^QnLENWu1MiMPS{xwBMa*ii-*StHcl{_ z@{OFq-3Q(5FNI>|zutvL9%p`dVbMtvJ*^L7FolYbRUdq}XL_AHYHFrBKCD8~ zCXRtldS$|H@g48YBqn^^;^n!J8#CIUJM#E0xPCVF@8*1y;650!9N0v$2^5K^B2bJQ z`nAUiq5qq?Jks$!ByM+e5+L^LRvMiD%P8|ApQpnz z_WS*bg;Jh&`lf5H2qL$KA<}zV{ZQI0dmlox*g!seb~DoiG2jiZ_vOt@u4vMhHC&*q zH$=E0a?Dp-hH^?>{UP(hbqf{W$G_D0g{-$q!m6Q$rc!WE43-Gg)a(`O9Fdd$rmH?< zbq7_dOqHZcU|tgErc(b``}>Ete;@IaU%egi=b!xgcfX2|(yRlas8N;{_WJ_<=-X>H zjK>18WABIs3k0)?az%|wEiR9$VUeWIHOl!LZl; zp8F|fn>#cbL#Ns)WsD@5frghMo7kOnT-Xuyk_k)_u?jdAonud(umA(@hq(dg$K78q zH>Ym1TUla(j4jpo^z?0ESwa1zjGv`WrHkFs1BmA&{p>8@|@KvK!dE;o`_s9uukzG4BRJ%-M&_#+pPmx9>;d(%0mrBT9hTx8{hQ{P* zltPKjT}2-g0}bb}DZD|x^fBvLZ19eOplz;?_}WLzG8b(--R_$T6t$oDrYNsSOXl|n zwgesvE%L`iX-t{MS~J|g`6~8dP4N2Ui+u~D7UAAH_L!vC8>0kW5LvEwjPi>N$Gf3u zgsCC=Vxe3~i~zf8a_D==4T?|6r7;|^4xA;ae^S2ak|)`#m&UY^<7Bg}4#<;s^4f)2 zbd<*t-&WyX=_5fMX_pl;E$aW1JREtuHc;Oklo}5a?Yxr=Q&8eavk7ETUtt@+4UsR_a{pwBMs+&ZOXE zECRv0S9*L`aneoHt*?Hez}1M5Z&z+RPd3;wLJCa~vV&sND6$Q!1kp^bQZ>j=g2kI3 zywRl|%885IBg6CEiq$E~1WAq(SPy{ryoueV%9FLy&9WRRc09E`P{x81&=uo@fE9Oh zLdleOsiltQNhcO^JFsLoGQ2nxnTz5B3Iw^LWn_TA*Lz^HsxwFY@m4~$e=^OcPUUvz#HeQ=;zh6y4=4oH#HsB8~j$D*q&QlzKK$)3F-pjkHH9OW@-@kc9h zJHn>4L)^aNI{iz3ebN1Eks)it2V5|TsMOP-Kqd_Yrc-JSBuwv^oIOsW>=nOX@9A?g zS{r}Sp8Eq?JrxR1MrH$Srr1P^tfwN%zuE-(7I{$7cobNDJ2b1NuGFMTPKvVW<+Hni zc6GDM=E+SJv}Oi|-^j^%_J&X4Gz0dVb#tWm|J6=2LZj<{F8`fGzcfygY!ifUp;%Dp z8>oo8(!NkE`CbJP>NvqEWwW@CSu+QcVwKzHYvWzYyb_rY9494EJq%Ya5T67ab@*#d zN%U*zc+oY<4wph+oBKteF2Nku#m%x0d@hI=PgaM&$B)v=Vu#6Cf{gf1xXDS5z2ZOX zvn$P`x)#Er*O+^xOxy|nLM=4mr0>w{fw#L}nC6URQx%F8?4+gP=>K9aFPYgRvgt0j z=Wn-*l7If}%5L+Xk@mWPf^gkP+h;q)LRrjKD&p*XY!V9ezc-xE&WB}4pQ4F(cD^=I zQ0F(~kQS6H>ZbMS@Bdf|Te#G&78K1Na=^A~o#Rbpe-6gDpm*3BGX8$f$m2WB3B6NF zHvH{>-Y^;)5BcdoAZ2#!^}1vNlG7A>k|MR}_1ZfV{a#qa&;cZTDN5Yeo(8KK$>c%W zp|2|gdR8BSGf?m#F7tvk0bJc?)95101#XbTY9yJ_)7BtcO>t(uEZ8fGkiUJmn*!u0qWbK z-;Fn4*x0Sqv9RL8$Xd^gYG5hqmuXK;J4qgpOnS+2$95Tp%CI^U;WmlNk|Q*@&hVJ8 z=LC&c-YO{BJKZR2+7|>~CCPRyYf4OH%`S?CT!Bm~0`m=bkT^#}N@1pYn|mjC(*X`Z z09vegByJP;g$_73Dm(Zs>H!xluR@X1sTjVaL5>c#JSvZ}oyVOcTFj-zn9=aO8@ZvS zwC_j%@HV35-Vfv}$@Z6K8C+q4ts;sAqVhZ{;(eE6>SkG$s#E}MdGEXAN_H#a1fYT% zdF`}8U18pKAC!&h6&hxt7!G4&Z-X8)Hc?4Yra?pRk^fFG+Qtd75s-TxXtm1}!yJN(|a7ZJf8$1)FVcShS?@jp1fmO&aVh zItbbc8*SCl8DAQ6b1lyE^Ws6q=O-)f&okKxGyzd=3NIX15%3 z2A~f)IYFO*QTBR+LRu!uovr7F-Oc|{jBBcPx*q(q#i2Up=BqlU$-vfGEzMNd)5tW9 zY@g%s?{UV_2=!vkCTCz9Z?G(Sei%6I(JMwAtytxMj0`#0t(yq)Mn<|`VkvelMWP|> z1QhiLWEa%9;90!iqg+tu0Z!Vd{+VFI;_0~<>+C;0(!?IIJo}^M(@%-ZN6&o7heOOlh zw>D-wuTC`R5H+=4aa(Xacm;2(<6beIZiCYwK@=3vH)#xCrd4=GTrXcAlss&A&3Sp5ue#L9jK1C7l3J<>!UYS8>J*)O>tXQ)~$~^xX%lPDkHA+@h^ShrJ zv6k-?@*26uZRKRg=FgxBGJ7btn<6)WKTE#cKOf+Mw7*c}rHGoo&Hswer8(*999|Z^ z&3})oc72?27bb;Q*7PiKLeN z6We=i3~F`RCCZ{ZG|NO+q}M(0b7BGt)wvSnguG1O76K7JjGOF*>D}jK_+3(z+x(%u z2gp&}>u@gMH)$Ray#5{;fYlscnJKzVw}>}N_Cj55550%isov>6ifH}c1?Vt@p9bo` zzqz2#*9f_qH`bSs3~rFK-`ZLN7SsrihP@OExt9f0#4?ybgL4PPpB8u9m_BKoV97DR zJmBszoZIiw005CPm)_lNRwvdeXCt_T>sHU_NZ`}6y*hRmgAs9zEc~M<#?PNrL!E1 zcvWCk>n~X*(7`3SB%jT4O!kVT@V2|5b^iCK8@>aM)v4J%|I{4AcOCvt>&YtL^(2E% zn>FOqD=>Vk;eGr@Pf$H3>^%nIgl{}XV|MMIy3DH=EtLHYgeUN$<<&t0bjE^8H9Gc@ z^|^zo5|lb&r-I_p&fr`Tx@G#9BO$F&8yNN4W&N4XU}8> z=ro9FcLpDKEa!E*d4&m@S*`^51m)deXP~|r z{u%db;xrC1%t@Y=$4~uv8=Tfg_G>Wwy)^rmR;F<^%Pxz%U9@fPyXbgwSE_Gidyn)$ zj0kgJ23%~N`C>LBT!3MAt4()b5@7uHb8~*Rz^L&W|7%4iDY0YE5fnX-SS)=&vBxQL zl!~|(c2(LQT&P&?zt=r(POorxcoFkR-`=uGlf(m`&tL8;+Jk`!orznt6r4LJ!n)~= zvNNh{sv>{nyn`lYbaHsVGg4^gKzo;V&r|-Z$e_bvRn&AHuX1LpFLFK9(w~4{yC&Vt zD^US`Dv6wd=a-R4(J{V`xBb=1nNM^KSLAU4`4}`1*RS*VdQKpxwhew}nfZFl!a}9y z&7QAmW8bR!eC8X+|2va<^wsJ&Cy?vn@%Ml3hE)&#xuZ<{1)Ltd{i>ao{)V8h8>P$d zI)B$eKD1-$0xjl7NS8|#dx0Y7sE8u>ZutOnOZ+}_70jb%StM`{>YEa#c&(c8!Mt<8 zb*y7B4X0P0O()5;mq;#n@KU@svv)$<6-f|A|3u%t5c<{ORXA9&qE#3l+6Sew)er(3 zV6^xn(!W}8lAQ94m#pwl@!H~1syE0|yzqD)A1-S5>?g?GoFv23!-H27!|;Xfr=|CJ z30~Hv23m#tF-V$c-eSdT!U^|N2Ih4CSYw1&d(ey|a@daH1!x)puagv8OOaz#L@NCu z@cyhP@h&%J#q+ag+s$DZJ zl|HOGEiG_64zbzD5ep?NrptKAq@c_TJbvhg4XYqwcF$z&a)ZPdPkk8ruhpBZnN4HY3YLo+=RFd=07 z?KV5z0y}t~>>cqMvF|3QZ` z@mV%6G%vJc9yoPmu}(?6y^1sR$}!oiTp#|Jkihjr#&Er58T4Mpv~cr|W_GKIEc7VD z6-<{j&8eM+x`QguA%_@2=VbS$FieVd^FM|O9Oo^T=YO;$!Sry^Bllrf?@#tGCR<*b z;N}4niIPLHPJEO{_=?2G1r7_C89qD?Y;a?cnDU#i%(`vf>&$Ld z(S8#sJ#N8_A$AO$Cp#_1>ZzroRQgjzu0&tL04ix!P=Xf}E;P&HULA5cqV9!y-h;Dx z6b+DTIpmN^mqFI$!|+~Preci`D&qmev+w1kpa zbe+`Nu||T5h7l{ACW<{vkuy|8g>01oxv|kHi~7juvfuA9sDK5sQrp68S7Um2SwOwx z?yo&s2pt^5Yn9pqs$R&xI3q?D?d;jn(1#N#zK`+^w>QfYynsj`!Rv&23lu{n2eky0 zIvr70$Uxsg>iJq%90^&xqh33|;FG}1P$tnV+ZeP@r9G~S0m+okj}AMmD&!A2?4N&H zieJsT5E9NcC$J2wp7H#)5@6iKj(63U9mgYf>^J^OEQlR!mv+7Vr|1kLmNk>{TF=o>Dwxj%pWf@A__dek8}2@-0-`bG1;Vg$?e zkjh)+13LywrwLdtP%JE?K0>;q7Lq+1%0*;pkl*=$ZV+O&=n3_CFKscs#20#Tfb@E=|S4;7+XtClFj=l zwud6!RK!Q$ytZ(qbBb4~pc>M=i&en)8*&N!GD&QyQ&-3XNcv9o9bi_ziqpK4{^*KL5F`1IxM9I z9SV#?cWUtXQrit%}Z zM$F&i&4~io;{hjV@?(X0g7e{6nB(e?k%$oUcr{x%Rs$dmSQ<3+SM_nD>xoygAs8cpH9e(t>&J-?!le^7dZy#WLUyTf&^cNK~LO`1%>MW1)J zL-6y8=$*?xJ<^eS#QJABwktUyV{9KWrn@*HgUbJI|IQcP$qY?(^emz6bYw__9IE6% z=pb3Bp&x?Y2G%jy;AO}GS!o;O)nvVRspdmZZ4*#nq$s<1dC>38hS=c*m8V3*m{4HV zO`M=IMM5{fX-?9`B^AexYk+f2Oq#6}y9GVs5tv<$Mv9KXHB*mc?rXSI`*h)UdMABZ zd?+lN-t4^aesJ$1*<{D- z@O>r%A)8`>x<4KBB7kHG7RcqZ*11&Bkk#pb$uC9O!dzSs6CN+Y?x6|%L-SJj51}-! zm)r~-V3NRIsgc|dyTrpQ;%$}KYkbb!=l-#_88K56zsX6_Q2y};4=*wzC1lA*kkrU+ z?QF-2rr88nrzrLWMQVVx271!%3P=*2p;rJ6ptvDmhf6W|{2tIy`}6=fCn81pLgE1j!RrY)AL)AGc6Iu!Y@QBc=_yLQ6>rlPYwpmv zGD2#CM5aBs8{!nXP>`jA$|qQRAM?Y=SO&R7iXmyEz->(C&^XZd7>3W^;Ev9dbuHgD zZw~l0yQm%O9}ArfSFi=a!wkc#owPe$F7YzJ37unGV|MiCZ|&TVez)z8T9!O@Rc=3Z zG)kG8&s6+_tgvH!wB1B*Bv5QTMPiUZ)Id!GZ!q$RqGt^w(%K9)S|%J=Bqm}4tc3u) z;uG<4R*zvB0;42*zVg#qgAOlxyJ}3eDp1tYz4AD6fr*+LNu^O&zOiy@FJ1eEtJ1%; zF^8ds;3U~MiT)A};6N2%_2iECLQdeI?(ePoQ?OBHh#eR1AbadsW}G&W85I;;N|8fU z#Ca0KOACSwnK{~0r}SB^!d=qCDjkFM+1d0xvIO~RvgsqfgNiFY$X;JZZVN!5Br+w^ zo3OTNW(J&Nc(L*o@+6)ikEWb-)0YFTGKW=ozkc8x-f>bfO$VQmqKprHKqv5Hc;lui z;0U4TTN1-pj1xngAoR+Q2L90PZ-h|S*KQV(oY9InAd&}<6%mx@ChTrp`qEvSdqVpH{rw~HnUO2ZJCsNdhKD+={70_L~7t7FLy zJJwf~CJ-p5SZF?%PettDCo*Y4hQ>Z%=b~&V4n@&Y>U5G#4&MT%E;c9K>^lVM1f`y- z672J(y(Yz#GG2y^5B)@o46ZuRUG=akK0SS-;1aKL;@3$Ng@cJUfGmrYQB2?|8^u3A zWuDe)w}l555;Z&Kbu-B7qI+YhD2JB`zB_=7-{lv z*P1)uxI_(KJhxo43@e##(^sY2RAu5=FBC&suz`VB)D{yz@?k%fFo+l;->n`Mjpoj-PNmK}Osi^)sako4fZt6qzji9T@bp3Gj995a<{JijA` z=cHez+)Pw|)|{h(i@31kN;V7Z4OhZ+P&<1}vJ)aQZHkq^-2wTyi+bhhB--^J5a)G^ zhoo4A)GJN!s^oP+cigVv0q4n(B-RQK6D=}g=54k4k3W3l-=)S?jpCngHj-Fw%V0ZJ zDUeD#qW&$FVz*Lc3l)LR{bpIzYbRujvB7bfaF3IrFk9QfzbZ`=V7}~G7Ap?=!uz26 zWQuxB0X)tVX0Hf2z5g5kVb0iKx9Y_L2SbzqlW8}G?&0;ylVq@hMIQ#HwWWgBwvS&N zuuEC$wAcH-LVIs+`YcpD&9VXKvx>8dQqNLMZA0yZYhjy-iZRXp2w}k)ZdTsFP2*hf zdhq2J9V@NE4UQZ6`L0)Z84Fs4o!_}CtrVp5FZrzxJp;avCe4$xcRcng`lZQUl>(ij zFZ7gKoZ!hhn*jypZCqHfa^A;s87HhzN#DxU{=lethTFy6Hxb#7yRzZiiv8|WlsW3O zS=+rTRq;^2)hyfMr^TKDNwPjBRbogEL+`kuBSDICy<^e5WujhT3=bBR9sH%DT**V% zGP;o0%xXL7LB)Z$mi%2?MJi#ePZzfE8=%w~8wzM!_-Dizq{)7B$v>C0Kqkm09+E9I zDyvCFSezhok#-BO-4k!Df!hifmO7c-o~dqP5BaWR@y<@V4;WJm7Y;bb@^c|U-|%(P zo@z=kt|=WM2FAZ>I`N#Q)P8M&gRE3xo=^EgWvy7qn#4p&=z~Ewk##|dPzc!|&x0Bc zq?0$0Rbl)Q()uj zrUgYaMegl%S1<&UGoe9fBGVv85x6?M1v0z)W#zN9UAzO}ub&Ln!xc_?TWQ3!+iLSi z?jcKO83D55z3FR7svQT8A+KqK(_}Zr0tRxZhz~^h{9=vaXQQS}e4lh^-hKGr7rx%M zXyeyTE`0Z_CXqSkxOQ%_Cf8@TOR*+JiOJDO!cZe=(x5@oOD8L1oKsafOgGr^m|JZN zOpFNw=0>v(fWS$lP|;sswKl|vn4f+$FP{|JF=FaW5K~F9P}6>xiulW*!wTMIrvv0ec+0%j+aemsmo0`5IE1~ zgkZ=VQ^tMHUuMJ&oG|ptr{DCxXYSawTdT!_t2RoIEBad*G#v(7m+aYVowDid*=a#o z$R0hdpR@|^I`u&Wcr{fVy2=$ZS5lOdW4(+CAtPqeR-2pU0S2zhY@u16H)mF1;(3GE;5;WxyyD&5qVhIso_jx#uO!=Fu#DOY6KzyP zu@EWDqau(*7n9h|h)*cMbI5L&L5`z=Z|<@(+l2GTLth$8cr zdv>d8Ec61r>585-Ah2A=rqUm|BNyiXh7_^Nmou^RYTam-XMZ$b&UaS)=^gVfR$MX< z?08KKZ6Zc2km4wI9Yxj{=nb%gS)XE`qTGFeS)-|v{2zcuV5@}Rljv#GEb4X9)bk^-`={Qbu5%8qgQ1^zy(`nAxwn288OWhwz;BqqnkOI&VG{z{g@cN z9F%ipNV{SU0ne25j2LYIY&uKE1UveMu?vaYaBHPYucX>xcu>Lj;qp1P2oE<--0P901YKLq0gd1iuaA!M`m{+S99kPt3!9c5Ap= zh)13jS9(C7as5vNXT?&{Eswp@EPAOLL&K0cAziArSr^1nq>5EG?2~C7Z3NpK<0S5= zi*Mh$6kwDp(TTl#NIo~IV!r`Ra?(Vqlv3;=iX6Z$z-161eB=>7yu&!hZ4emAWntD!J@g?tZWz37;@ z*$o6dwJ~#Ig5X!5B3=Z)F1mD0K<@3t3703BgspKy@IT)>y}-Qk%R+%wgS>doeX^T> z0G6QZSnZRYYrzJ(t4Srj!oQr`Lml_CvU&hcudw}EEE#trZYYehC3VKHeoGQZbFWuwFawV0#NPk$_2%9`s$Xmk?hS$$4_tX}+U0kr4e*(z2 zwdY*#c(&6ObeFtXT|DiA42w!9*zPcc6*o{%tTZtDDki&@LH+IDr|f&tOM8~dbW9C| zg3pmNP8Y-}9?N7`rG5N(Ai9Y3f8W5EH0%|S4)08A;+C;F$ReTD&~^-5|ogZPfKyvR?d9nf!gMMmm!BL)j8m+D1+32Ny3TM@OZ3I zCI=oahkpExFB+G(uk`)=V{*)n^B3Dpma=Cl_6$YpsEEI{F;xq@2(Vw#+Dtm%DVv`p z`#{vkPoIVS_1ZpW9h<|ep;NrJcU2SCi(9YaahTf6(Eo&lXkH ze4C1w@j%6zDIe=KoS;Je*XeJ(@uGL>!34BS^`{{P;_TVw1TzIN&l$RJO*)bN?z{1%>Jdu@12U@zS(_{e=#P`dikoIwR%Z4+TLHYh&H87lBkHY^9zETWl&X)P1m=3 zVgJc<(~BAH0RMe;!Ytq{=jmJYtYd-le*S4G6b8;%A)gFgIz~X*Lgbi?`L_4>I?P)k z*sW{CLfeBq3$IFdFKlCqp`%KZrg+xI=~?`xyh8WwvLx9FbqT$e9s0NSg?m&hLz*>Jr!T(mvPadZNfXpbFw^y&`_ubnD?`qavzp_i`pbA~HFBudI?Y~l}hHod_Zm8>G|S>%U5xC5$YEbX zFZV@q!rl;w=%lEtyf2G)D-O)TK2UHV0a(~#Bo&IN+2!H_G2}~Nsv~w|DGt>!g}mdA zKx@B_?SyDiANT~oyNj+opmS5@2d}19QxhU?tOm~$ZS>qN>)`S1tA4Td?f*7kJ9CL7 z+p$owK&4|Kkc$%x`9RF@F!WH*cB&@Xbd@T_3o6WSh}&fi@?7<9MK}GorND~xiu?K0#>DGe^2@NT#hhL~pqfb-e@`mbut)sO}G{m(Y0 z4s6LrS-Wht-z|BL$AJGveoVkc|7w1bV)OV|VmN{s*Mb3kavswLQ|A@#U7_lbt_O;EYoN3X``oC`#hAMQpQN|AcmOp zs61*r+wO^}PWwG_Mft&t7q<%EcR8lU*2l+Z9(OMbSnE+7dK5}~Is@_eVf8w2^ z0>4VtAn%py1sD!!*j=(@34VbXULoj4Xkv#_4k`HIIR1eZvY$_whu3+?^Aa(MPL$dOee>Z)0VJv?Ql$1`$ z^2X{3vkn=0bz$|G6LF!}^v~{|ZnR$77X)4<$#!hLl$gx*E{cWP^h_$E0ooXM((Bkb z!4bc0VXbtoq=P@efA|@+65_)k9F83{mLLsR9Iu|PnNuIs=o&9+AuRz~EbzI>zvA-= zun>324?*RyjbOoH{w>%H6EWLQkHO94`1JDNg>M_tBJue36jE=;X!*nhE!QaaGDR*@ z5otk9q+iw-3X5I2q0^wsK5Vj4t#0ii^6txPrRe+?4QqSJ79A2!` zP7U0K;lvWm({EJ93ATHsDEFw=vduCK7J&&|A?p<8N(Nkd+%V`k5T3wK3reN4Xt)s* zsLK5|v#^PuglV?fGq?EN11?c)GpGOf%i?uGm&H)jUKRl4w!qPYdTp%}_;0{`4YrKW z@k7i4@~k`jG`5UKxIygt`+l+)or5+?fEm~OJ@%=9x=mY8k;+Q2IJ2lvVFEDP#GOq8i>igAVVk@JaK$ zSlO%#Ld#dXOw>Y>T)?FhDe4v@PhgCobL>6qY>pv|D0%h?ZWt2&a@FqHMlto5zv_;V zom0ti6KQaeV)s#`h>F-UD|=1}l|QS5`k0?RXTOt<#a=r}vi8~C@_2S5zkzp_PM&_l zIg%VucW7#P#}&7kG9a2rpY@4v1Js4$9o_Oek0#AZaFWGHa6d(v&+iUyrK{DKr|$_E zCo8b|bD78)G9P<9`@$!qRqG-@=I8EL$7Ubjn%vQrH_ z`RAd9OIna+tuh`=SiIMI5L+C=CMPU|_~Gxe-Zjt0EI^>^!zx zo*u*oM=pOa=%P8P*GqyB3lv}%AOhLrUdF5AXQ}JK!J6XL7|;s*ho!GJY2qa-{15uq zDjOX8-6m1Ea07=8P+(!mm_NS0-JFH=X|-a_Mld^DWUQ^sMcity00@l&{VV8Kv_?^Ng%`v|U*`T6@iG=t9-WS8$=)6}N2YC(0XXd+}S1lLW z3MkxA@mOp;*|0FH<2`Iy-E9AP>&*dp#=wD=<} zsFHWv87!a_H3k!wib`2rs%;Fd@YS9q>9ceUa0Q7shGMh{JGesEp+qKDz5)h?)mF3Y z;(~mBFEG1e^a(#e|Izr3aif~7ay{&Lj=X@Orah~^++<$t%0-XaF(@sR#q0ELF9Fv~b3i7oW1^X4aL;UTz3QdKA7@<8s5IMH_j09nV`;!$*7y}H(tD&sA-?AE8ckbxd&G|Yk5M69l>6pL$>F-;WF}$>( zWwLH@a!`}z9xuV`w*M(%A-HI>=!&o=P09SD;Ex%yP1VLzf9;F+#780dv0tk)d75yu zxwFNB{{oXwS{S50!7`sJKkf3_7sO=&1&TASD1mKdIX;C5!vX7YVPOCp9k%=?j)DvO z&Hrd<`kmaojfcIae?eORNcYL(#vGVui{T?+O`iO5iakn^3M%4#h@Nejc2~Nb*QdBE&2}2{xd^==aj~$} zwMAU-xMA95r_H<)6|P8Y=uSnPV`m^X4(+Bd2Nd$w%-c!FxvrVF3jS}>qw0NMkd zWDCLBl_p8>TFYj54+X{unlxSV95^=oY1wptAS_|#4U9MM#H#svvNYwe^{m-={y9$b zWxqP-bnVRlQW}AzIr;VxlK#?E4V0U_4}}y9MUK0uh$sjMVluZOX>z{@#w4*nJ9L3D z9KzKuG9v4kWF@R>kZx>9wMqm5XYAi_d?vQa#5zZ5A`F`t0&CV&=EIt7FLyZkVy(w}DieV5XR2_fQZJh`7Wn6%0CT2;7=+P+OtEl$C_b(eV_|ak2EbEwO9^81 zn7d`|R+?CHv?>mH9pIgq*(I-$42fHv(j|iqsp0ECdwgb-<_6sqmO(Egw?WP{Y7)NI z0!8^}H9h2lxQ5Pq-kOL zff)vaG?@5%(Is!fs~4Z1_%UMY>47Hri0h8x*Ufvu*sVaZ(D&sY2npyq=UT*D2E5s& zk{$edVGJ))bXr;`s^ry)vYiGUhL7zG-V5|-QPc6L;he3Fl`+n|je<2#;KNOtc>C?B zm}!(IkJyvZB;}=%CPz%9Ndd*~qR37vVwtEgY}0F{PO;3Puq!_CVNlA*&+;nuyas$f z(M*+hlLo`st-=I#mKO#HOX!Dwx|xrBpbdfGUTB8~@82C>>V&_uWfNZ^ZJfz5!|So# z^#l?Vxz+f_vFlDo!2Csf?hj=3RN(Cz=_K7uv56E}PemO4!c}P&f1k95?gGw}F6SOa zq<=NE6&#skX6v`qx9cfa3vZ4 zT;+1YO^XQE)!+b#iSFmLLmV#lP;N$!Qw<_DYfKp zvHX(+Nl8+aail<;36;rpq+Y&^TC$Xi1wxY{2Lq887EzRXu9>Hg%i)lnfeqlp$niN% zDnTn?DuFhK*AF}z9 zmUe5WSm35i45|&wQm4|z?kCAHNe#Vf>OIFQ;p(ZFCb3G;&EQ_T8hc@O$yGh;EpjUt#Q6E5eU#PRj zZWKlA0~+}XL zjD60=_DDI*^w>UNjCXO8FjLl~2gZNLC}F;Jih7HD#7)B3ai+>`6A9Bsv0$}bhU_N4 z_Zh6&8UCHf;8qs@I@#*h2Z5gyuL3CIsqn>ACM;L2pzDQ^qQhae^dsmSJAC#G3L+r) zuA$REH|S9IR;g1qb>VBP7A=W;BLfOSu%%6tCPNJ^Ro=Mu{XvHnZ>;*aD+^1V?)-bI z?}|5eD@ud$BbqcH3tLEnmyU;D;RJLZco5j6!8R7>#hr>)@jjmfuQvHkms-U&pS|Ay zh~+&`ZF0b*)dI(=3!j5k@#Qr?pp7WI_bsoJWZ6^#fw&Rf8rD;6EJfBLH&2Q(Q?!W z?&hcQddPYA)8bY-0g7OuQpIsVl_to7_rIG?cKtA{Qcx${A=N$ib)O)3S%>qHa$tf! z_&LM4e!l+e?nbH6_>=vM$(EN!Hvv8O2;rAQu@LXtK}GBY3WdM7fhPbN6b_Q;ux-x*?#;bpYpn&+vB=REgko#u{} z$-&LrZw*D6uausyEVN^hVPQS={^zbr|Ix<$rUv9!(nxbGLq^d*h@OWD;vCw!19bi5;T^# zcx2BmbU!V{h$gJmws?1f;KDobRFAM5+HE0yUW%6%lf{vvtVDVfyhIK13y`0aC)*EH z()r*-Z4kE8t$}f*FZAwQ^p9$>>>n@Bl^pU#zosFeVmDufq z+(MxU`n#)C=-|fme(h4Xv>@a(ET0E#2S^Hzt8`#OY?d96r6`d?a|d*nMBjLo>WpXh z%rlP5WP=W8JRi;NCFQfW_#F&4c+2y6R=KDxJ=xKYPZ$bD8+;tcEQ_Jh>MvKF{(aAX z7)8_13V%L88eTAxk6R`JsD)y$P^6iPD5gt+xe|Ff(5~6;NUU4^iC~DR7i>!VPYWH&`F}xej8ETY8tDS}%kV`XDjpO3Ew(WsTo`t7LAxT_^&Yc=S4bzxPLfB015nn6_vgZP z$^5-On@PK{WYJ16%(I<#ND~&d@b1IfKhE`d@J^T1@EZCrOa2ne+u|1sx24b8C9Rj& z=&klM0XJVYFkZEgO4%pAiTqM!tk()&x^p`)Vj+o9(Y$lAH1|6mgATodEl}gIp5V_< zX58^e^=%hE8ZAM6%owsl%2`_4%PAHz zbW5m+!>SF_9stE^so9hE~m6>sUi)^X|=U(p;0a3iE)hum~pk-rQbq zLtRuqDOPQkwJXlD-E%5o>Mn?J`#!0m-^uWK7!r-mK~ zFXEkLJ0_&ro=qm|>F3V*2)Bu{+vJ}9tTl(tOQK#9-L~K4Mc7HP(B@z}w5R1)^Kyd? zKd(Xx+r#kg>6m$*uC5kfrcagMaR_@<3#$AKVhAH@htz$1tcbyqK)mjm1`cN^ncXlR z#G4H@p}URM^MmIah0S}Vno5%Q(s%(*m_X$S#X^^a{ZvE-y-l@-_fVQRce8IV!7vOq zQqtaZMdLX|Ipl!+^L4!Ys=nE+!d`h2P&4Y7DyLi_{-2^muYeX(R>f`5N-4pslBZ*q zYCiNFau{Cqqa_X&xgdSkwV-FL*SXE&c;GX-=HmAGmahJ3-|9tg8sU@g&AUx1?ASVj zMc4>ubUnp_Ry{#Q)GFGb`VELszH$kwrvB2#ECaVR5}NOocFfzVI_OtI<@%Jko7O1`tA3tTmQAns?wbY=zLeh&qhs?rjNN#Iy76TrEjc#xGCaX+i>k;!ckTw?qkJ74siJEY#bR(aVc!&R8hP#JBd_NU7b`PmxlYjd;R4c! z>xWu#{kt`SMhT(>5CcK0XTsgk&zuG>z=yi{AT4}~1#TGBWBBHPDc4Mnv8x-gZ{7uo%?i@vSG_|Oyv zSiYHZ$V`T?0mdS;XS53+1ZzK)MTNJ^ss*Xgz3UTs_Ut4#o$^|+At^by1XAU5CD1n8 zF$*G=cvr3B0&{l$d0O8V3_m?B2*YIqu$DRx_h->@_z8ygV!4v|&=R_dK=7e6FouUM z>K0??+)MCp6ntR5(?=4}V#`F)ax6Sqe1QP#IZ_wK3E(n0{$1DuSDQ*-BHi*AkY{`p zpWu{Daeq9z#E86${}B~J?r>9DcAOBh-UO#BUSXHMLI$abR$-FcZ8}Gt0rRP2_dqph zljhd+n@k@+R-B^jn3pMPQkKtZ)Kt)2!G-(+u`c|qX3!yy>`+w`gj}cM0f~n7{5JPY zQG=>LJm|0|cnze*XFz>wOtgR93n6@tc%A;21xk>IJI|$*WdPJtC|X z*)3=t-u;Wyi{687q@Si=Rz535xsOx`wdH<>DCilw)G3FT2Go$yybzjn8!9nCI^n@q z;V!y_#ZxH0qNfAbpXjK^9w#`RsV-1l4{G2U5;n9Q)AV88{fg}@rhX*O#XNy=YR@sK z#ZeaUqa7Nby5zLJpYm^i3OQ(Ap<$sAqKD-8U?q5QX!h*-fH;9RNtPLw;<7RX@-B2t zg4Z~oC&#~%gO>^bydD77Bc*gJ3_J;mOEg2*he!P;83n;$N{^y=Rk0Mz`_YXUmj{T}Tys1$3E-y=OBeo~?KqKzpFxJ9rX2i^xoWSQzRu+iFO zZSG_CJ9EJJXmZ4gV`IZQH;gy@W7`|%OZ=ykWbOA=0AkCLp6bU7`S)FpsjF0_0!$E! zdToneh8jtW(E-#f>t-(4fI{|}vt>P4o;l2fpR^1Xm;b!@O}SCVocQuru8{RFjTsH} z#3Oc7PI>?DJVQb(!dj zkKz34se=xyrVd|qMe<~ScpI7CHab5!UL3ktYS_b3uvOiYy%I@}d#UH@si{!&m&n`( z+R_3#RfV9-@+#1m9k%Mf5sq2$|JnNzxTvx#Z(q~zQFXDXje@EtAV>iXV#{I?5xeY8 z(wWINnItolNoMIuVx<$4?#@;*Usl|4Lr`!5EkIcWQC3lu-36Cc0Z~v~Sw!2W8^MJ} zh3}lILaCC{dQinzI^X0s^4_cWF7&(qI``aj{^x(_&CPE9C4xAV7G_fbm6bl`X8Ym44KxuC>wf+}G!}3yw=9M~SJsnG-+Dgo< ze%G=&s4~?_-;?x3q6!L1tt3qza!QEY6_h7UrHcYTR-(}4A=OH#y~0n@$$D6uPu3E#04PLoZ zpIgIaKEdNd+f&?Gdu0CmEuC9NTmD;rEgKK>)PXT)fqA+txWZ?#aKLrQX~=1@Fdc}1 zLrz$b(;ikw8`N76b}cMkA8^CaaU4?}-1X*p?}8~5e4MJnEqQ;qLquQqx$R1Z%C1gtayrrv{vX|m;I%kIS} z&g^Sh9oIKdMrZc1+vC$@&9$?d&6Z1q$4W9?tY)0EJG*ANR$31j&uqOeixS$nQd;_`thJ`D|si zCC)w0sFN~$VaTb?rOibf&+n5MMBn0IQ(YcdHLV}i*RirQ#Rc{-eTu7sJ7Aw|Zi%b` zzp2HsbY|PH#yq-;0hTeDS#u%+4 zkA5UQ4UD!d8VLoE+u&t3W9nJE3d14-uImzggsCMl8BYEuPu*)S$ zF)=1)3$C1dC^p$OCz=Vi1!g-P~+zlY0)^LU<|W4)<-O( zV6+wG{clW<4Kvz}bia&^WX~(Z6#3Z1qEu2WbdcLmX|NK%MujAzZhHP<=w*7>zlNEg z35_`(2b}RbN$WgX)F|=M9JJNF$8%6I=#<6F<=vF@dki{V3SJ*M$lvNb=rrh5Enhah z%;N}u(5W=!KHP-_rlX{TUIq;qjzD!JK82d-N6Gy8Hozq(xQ$p#pMN-3kUf2_{9pXP zZf;t*EU6A0C9z=sTn@XbbFMq4KPXe+J_3Hkj|HZ}>$=T#)gi zv26Ld^VHvGna9r@cG0mA5i=C|>)$T(YG@PRB_hY;2MgP0DOK zOH~D(&+d~!`F6i*ejD^%X<&{p$DpojOYoto+T1sGd#sc#7VeX3FHBuEqf^=n&)XU# zTyXGQt6)nIz*w9Xq)Gn}xa!>ht`4RR?(zIiP_{1*#;DtM)k%^*ZE5u9l2++WTDwcQ zon67-=2`BZ6%NWS^nhEXs2#eYtx{J)`C=zsM>b1uL$J0Ih;104vpuYgXc#}bYz!;o zaG!{HJzq0AFdv>v{vJ8xz~w%mA3b7|bCzQ3DUjsTocr?U^Zsz+oh1I}Uy65b@YpUb zfVSBCqD!R~KR^yPh;=-dr;#*dTd+NP%; za)c8nD@4&1(I=vesM?e*Pa~z=++qjzYJlQ40)s~>_7Fv?hZ~7OVHlRLwo0*FUN0Js zf-s%(r7}nZpqxWpV25WReJ`wD4!7Xw8?Oxc}mqa_mjw9Hf-X5OY^fU8r zn6HR?e9Z(%Pnvof<`-S!;R?oFVH2<)L9?M-&txkK+{`q{5H|>@KE+{4g-`yR1O-;v zLod)u9+t+V{#1=`)wFI{<;rIjGg|D}nZU~i;v-h5X)A-a>#1-6DhMwuQ8ZVVf`fAcoSn_E^MhXKI5k1@}Tr zk1OM9xAF9gI?B;_r+<_oZE}g*Qufx|ZI8T-OIhdKu&X4+fgL1})*P`o?V{M76eKt_ zHBe=ERnRT!kzWhxk;jA__Npd3-Qt+@c!PZ{g-_T8G`VTj0lLfFX%> z(Z-2;q0(n%lulVU-^*|c!0yE$`u!Oy#y3n-p9j;%lA1d7*`0~?~9j3y-Si(mYi z#5k}C(V3W#c#2(1kyVriQ%2a_DKAosqRnM=hJO#OwT~%y?rd2HiRTVsmk&8XViGOU z{J`A^iSO5JYbF~;%QrgkHmJ}9COaq=jLA0K2368^|D(M`YM6TG9Rb>8rbl{vRy@B_ z86(uDPiqq#CJ?QWuOh{f+7nUT^g6c&Hz*a#P}MNFQ?e_btV7O7ukz`!T_9r|)G~5D zTL1Xhfku?ny|ccIWN_O?J8*4AtqFcgC>8*CO;DOnDgDy_jw!y}whVB=^CMS-`3hyVET?$+p+-}Ssa!OJhb?%hr z%TK?lQ{G`ZY}@H+Jb-!%E#r|55mtOVylp(?>kbSX3+3zCbRwv)bkdl_)|QDoeKJ%B zRo!$MY>zwbYg9&=HmhMV%Cq*o^rO5jUkfo>lbio^yNKjCup3fmVjs#V7Fd(Tl;$CG z6G9(AiX^$>WAuUWt^O(DP3j}TSm250E#w@WT?bjeB;JtXFn?)ylkAA}bJ5c9;z+Dp z?qCjtr-ZktFA?nabyJqi)Wsp3pf>;3 zOZ(48?8JUlevh1bo!mB=qRSL}ks{5MW|45eavkrgq)ENxzphG9>!hD8@vN2XAk8pw z8>jVzCmRWV2VAB#GZaDxyv@jr2)+53d0jv^h+(biMKE z;X82T_2h=tAyb=9<0v!x%q1(Qn%guwEc^A&me&uPOKUjnzS)9|Mkf>%8%m8A3t<^* zm0k$|Az!?T_3Ul1IO&TFv5#Pbpx75rvHIAS#%6TTj1jw|4`#!woUl9TsBUKFhei{$ z|2;Pux&4xDnO2%8>^`K}eu~_qG>PJ}IfHs>1Ka_Dyn%k9f-O{%VjnLi8nJU()FuD{ z19E*zA2bgHWF2@@uknnnZc+NS#J7_!j!aTy3saO>S_0i|lN5$Tf&qLwr7jwK3D&yx zOQF~?Ns;bjyc4_IV%Q(k0OiUOUw9as?q;Yuz;lX;eh`?WY7?Z=ozN5)OALmbQ06XP znCXVOr~6QVkRn|F);5qygwz9enlEw2j+&s$tI7#5Jdi8KECt?_DozXE!Fv$(v;U`k zDxa~mPcJ-&Yq9l%$6qlI>{*DPRZZ4;9trL7!_ z(>LF7w9yoHVy6uLy{{l^;#-xfIEhquu~F>ybQPvxFKT?o7dWJ8M6+=Prt=xM^-|T z`)!;+JW2KsKl!V<-ZGa=w&Q*a07WCSPDvEIo+9fE?QyVcvEJ)R2;Bp`K0Li2xb_Y0 zCULIVeux=uE;*oRw8!ms?ZSn7Z_PH^kW|Hg=aRkLVtfu9BR*?_iz5^ZLBA?WgAVWx z9y%z7ZW7vrNW+0ndDo}dS9{G9`z94fb|@e+jC7g?Sss`s?Eb!%P4zq#5yv0C|8vo; z@EDg^Dj^bM$l3HRVHHUrXWe&`1&f8q!2x?_;GOpIN__A7e<6QJmpHG98i(Kj2OOCh zBCAfvjI%{UZ+#>*;^;TUlmAH8yfQe-HNnwVirs>RSQ-fW&8ZB zCxcE~o$cX_5)^K>_(Vfo2rGM^+kuwltJF)Zm63v7oJO4Co$6pMr# z1MbjQCE2`oMT`1Q$VO=r4+F@@NiVI7+z35qH!(F{z5GM;v8mN$J>(heKj&7;ftlM5 z+fA;;kvLfr>e8hJ8{RfzWY)s7ZKUEQ3*EJv;N}#?LR0HHP|b!gZYI6m&rm&UY@C+E zG`Lp^*0Kqa^>nIuL8dBKd>|0~P=l=L(V#eSr{IDI#@aA6yv*w&4_32GRfZ}{b)M`c zCBCs#Z&a5GWq|HR-gRBWPxQkWV5X`b7OZ9_mTG~>TZ3}d?4-#H(C^7%`pI zi0zHR%^9O*F-P5t&Mjie(g&6?mZ{!x-@LJi!)6OCG!fDH)k{l!A;$a8(tr*|yDT^p z(16bD`oK(g=%6>jX=@H3uryzGxo#N*qogS(Ise01nK>|CDs3POU|BJ>MR|c;=U3yK z1;z%c`nx0tgW^KZO8X=Ol5PLL-xN)Azf(53s$(Gr23wS`w(nH2O~#<0__vv*mA~*9x2bxsjaVA z#)$j-bn#?aXCgCMdsVob+7{eOOTbGmV-^QS*|l#`vWlcOe^hy4`0 zmm)=&9?653l>-e2cLg05qlq4HHN>#7HCSI{Izep=h>5$VlU+ebSHsN#h?m0)FJwMu zu7|bJxN*R5Z#P}Z!z6R|bW}Hnje>J4JZWPea_sNiVCF~uYx>s4cV_uJJDQ8=uoS4@ zU;XsGuNa}T+n>2ZsvS6xa>WGG^%Q%GA}1(KGg<8yFWkth@p^icq&Q1&RFg zUM=c6s4ME_uZ-T}JrGjk1)ZQJd*pb%#;Z%UgLh;0`Y&hkFwpkcD;ZjZlmt|~b7kHY zNtfJk^(0v_H73OHbi25hKOg|AXE=_W=-0s<;S~aFaHG$%sTD5!qnCLta_*Ik$r9NT z>mzYL@^~z$=Y)0YS6eRr(wqqHu=`{S)b6!WDbd+Ee$kU6mUxP3qKNeMFL8nN(r1iCxO1SU^=Or9tTzq=@eo zZ3&8>6BBZT8FK2RtLZqg;mW5#ughWPr!n)hnY{r+PDnB@^+G=0Qkl-~Z2+uC)kyQ9 zKejE5%tpJm{?_ElMx5OI`Q^Wom9LX*lNY>&V!@tlFy!<*6lI`qp+%7ksKpRu(fY&T zKJU9=Q)-7)o7tcI^pxXEefprCk2rl*9hdZz^!uI<&9_4@)t1@9rU$J=8q=qUV-A7G zlBB?9n1^`XvTQg{V~~`4*1ZxsBpRaa7=OPGG7Bw0jT@7n!3k^}Aw6s?IQ@>X+-w=r zAN_dMsTT)U92e!pLSW^t@}}qY=>|zv2)wmQSE=<kF$Fd@5NKxg)#|+HzTTofq3Tk#(S`Cl zdYM%bR+OL{!pIpmw*3_@ldiI7)BN-+1O4mkJ)C;od&uF|8uc9Pf8%WaO-FRoe>mnvE5;fy> z%3DCXefrdAbPkj3@zASDhR^Dhm7!}O6kJB%llPMzQZKmA^hH++bbfmS=EsSPJh2nf zebVMOzeNqwtE7d^VYWJNbzkY+<5@?i_&xGE%u5T2=Wh(yEWIVqe^STK1$`zfjJ-~b z80LgN>RXflEsimIJs%yve~c`7WxSqr6R&3j#l}-)Eu<6XT?LdK;q`dNF`o){1m(-q zd9Cj}czYdhqL(VGnV2@UPSy;zHF@fQxlFtTHTFbaHgH3R%hXY4R$rZt3B(>K+eJ#E z7T1kdX^upT+{UhmZ9%uE#J-`;P$fW=12c;UU?XGrUu{sippu48>kYDIKD1pYAP;A| zX|eHF)|xkI)jgcPvW}bc&zR?vVeaB^;r2W3e;Kjh5akLHp2VmO=K^5(_X_XNg8=I!Q9hsY`i&O(E_=?J;gWQyHHkwi+9 z#|K9SxDXBQeY~}7wG?#+u>LP;axC@8brp2qydhbu)+r}&!nsWMM0&v7%vtYp?SH-O z!pS#y2ZL_T>IE5p?EQUNls>JS-c34)PI)b$#J@$|CTI>yc1agM_PRCm+?;0pXPF>~ z+o8BMqc8Gcz)62ppR;R^8g5`Q_fc(T3Z5B=6IiI5#}X39Q`qdl(G?4Y&6qWAgDp#< zUw>qs7zErEE5f(HRtidpz>VD{Y-K<5#N#H}@I7C6-r)nM8w7WO2Lt>r{ikr}VSoGt z&(5iS4)B`kdXT0|QfwjV;w_{C>f#`xMI|1AyeZyR!ydA%i{OiaDetxLg~IqPsU~BtU=| zSbN09kc%KH7RU5>V$Q&zq@hzT4@WN>M91ihyjuSurv_QOXPg*W%Fx^flbNX@^`uxe zDAyK8rn~kj3O#H6p?}RKQW=^GV;_U216lAJ3V4^80k<<=6)r1fr=bDdlRrhfN}Vw` zNiqDlya37LgyV?iitFL?@jB>z=qKh9QCv3K4$O_PAShKW9hBqJrSE)1L+0U1)Kb}U z?{lFs!lfaYj(~DkLnQ%Z$T2703EhdIyG=a*FMW#Q$YN3qzId(}Rn%TE=g)uhmOmXo zRxFP_o=@JD-29a}Rvb3HU}5PMRdK%xdWrC`{g^W2o&E95@bjMKv~W=# z|5$fyl5ycu?b!E6^4N*n01jIR0fG!8{pGb3yNV(!D9zQcF3g*^aNmrL)7S9N@s|YJ z$v8Z3ir_OV(DQ;3?DO$|e)VyohY=60Z#MsiB)$$sMk8Nj2E{^}MMr54y@OemoghxI z$D;u{gnc3ZI3U^Y&Xhes2PmJqW^xUaN%xWtrZQ+KtV_LKfA<|&ts6xH_WM>2zZEOq z@2Cs(@bmWi`HFnGdNa-@{qKtV$4@VtlSyaMCH@2M=;g*lYp+Q<6v(nm_S&bKNJ=_l zlLNl(1?5r4eRFlA++5%G*(i$xd!rV_c>CF-Q&&&DHKT(mj>J|Vr@c>mZ;Q&ZhXMr-cPBU8-9D%Obr-Tf1jp)OOIXt_;l;R!>f)AJTtbxWlcG zY4>ac{pnjXt0y;#mc6kwU@`eTLYt$=q?ZurRL8HL47vs&{D(b*iX*zlsgq*eq>Q@c zyy)V%Cd~4rj1tCh+^^*yj(+yv!A1lXbr&us*$$jeJYoXveH2?lkwQv?JJBnO9Rcxk zj>@%oVvtqHqw=WR{0$(#p-1y-w9`9@hysNU!=JK%+7u=Y-~O8%U7@BM17Bj|lPC zQfw7P%AxNjb33e_cR^GydmK;^S}#g>{ajfs&!A&MvVH3Xs84xDnC(&nJK^)ei>6$o zituZ(7BS}YE{n7+(08Fy)DBEqZHY7T zNG^zS#WhSHuUiB`(>B36@qGy@9oIYK$ZOpW1|668DB|CO`%kD&OXlMsNs0@k-xYtO zp`$_^Gk?e_ecEMFwO$PZA1oAqEY$7|Y7Ibwc5!6+^lUm=S~GVQXmYFhx{(a8d`@Ynpv1y)LoS&n_G=TxCQ>8;s!l*VxteZ-|4;VU9)K2l z%lOOqxlonV4Lw6LRM=2!LZdKTAUl3jWBmx+Ap6MG?*}h?AGS+ONdV}MiM1z5qYtV_ zRro-I&o%)<0wcQ6Z87AOpb;9@FGzE~3C-3P3y%j}5P^0>_H@WG%}Vi01~+IgsTM3) z20m0KtwoYUH~rU22ulshAv1s}gtZ<8LD)pb7Kv>x*s~&tUui+%t`9{2Yhaa4i3WIOszCpUhCUKZHjUgqlmIzS3O!SzTctP0qJ3&I2RLEV4?_*trQl6(81|f6hxOXx zSH|&!jPB0#)W0W?EC+UX4x6|;r4$RYF$I(cn+BkVlk9Od^y;)mFgkIJwnipKjU2{4Z_I&9M0Lh$o)P%k-2j)))gpB}ZSo76|=ZU5oX zpXz?J<=xKrcD{d3ouqga-R+J0J*?4g3|+)GJeBMUAtAW>S>STl7WFZZjzA=>cfQY* zNFMX=MsAiKC6zq97oSV?JF2n`l5B{eVQkq1Nt_TwB~)*dM;J|y$a&rlQZSmjyaUsT zPn(#dYKpC($N@^z>C;bggn6*BzZ#H9XY=lni*q3=F*jyb?Hu4&`7fW9NneZD0hRt) zs$$RUiWng*7;}?C(%d#q@8dybzxz4aHAywJ!)fx&@!UzTci9rW3ABPcBL;XYz)md% z9e_3by68NT==bS#oibT?*~KQfvIDGaQWR_%R?9q7Qp}6!9M%X~sO~#MRznNyQV1V* zKx57|yfZT7oOh5LqSfJpPFI*LRoc`+r^Y#pT<%2;IyKYBq^kissmeM!pLY`ilr8Ei zCK&`Fu#47Gcy=lOB;6yA3;jei-zKOSX&G#Vkdb5B;@C1mM#;KRNhhNJ@nuI%8dFPZ z_@7F4lTLYwyhVMS+~t+X_rcM+=|v%*$tp!H>IY==oKzs`!L{h@>08vt`8Qm9JyDEg z!1ab}jBttYN>n3#B)BsIdu^vdIP%z(7CWd(*a$VY0E)>N+)z{H$!lYc%a!ZH`xnWk zSB7#7!p0-gAlVcPwV3IYCPRg3v0Ndj(HBJNra}a~-NPU)oup`$*7ySV{2zsaGn!&9hgw0dxUu?K_*!UlK*gFFyRW_V@37@Z;b8QnH+4mrx{T z*rj>FPm`f+deXOg?N>!cq^wN2_C8t54JnQr94EU>kg|x?b|C~yCzRpYMHD9OP?NO{+EIfvt{%pQBJv)VCvpdg@W9utZmR$H} z^I)KbC>oXr4a)aJ1$2i3`{v!4j*$$V64mLF6gB=i(7(G?3ej2LiFl!g7s2(l+WPfb z8jq7XciivTFMVaD=9S#fGP)diTV$cS8wsg@!yav&@`}(L-nNh{0rj#vdY$4*z@^|C z*gF?`ZkN`}vTd`Mu>=y6>l~46@~k~iEF*dPAzkxp#-(NHijX5@h#P&}abZuknJg{q zD0U4+R>E!?>|3r-i+r2mZm|=vbUZt|)?nq?r|f^_c7$4KF81uOTVM++)P_#=NY&U9 zd;+)|AYrqGg-IXofpl2}E+~2!6E^eK%=SqC(*E;o_zjjV!zdfw*IIlOio{Zyl|T@`s?lbUlk7^xCvR{2vb`JHj6U(!z-RO)?R)ij z_wU{^_bnXOSXl5Q>S(M`yg2uBQKM^_G6RC3sMx&PEtblj-YZKd1FmV}nCJ=VhBZ7{ z)=c+PYcc$k{eN<@7L%%eSF*@lr_O@F8xrpNBloJ%WN-C&1l^2K6&Xr^|?WOY>8FGK#In>OLZ)&N!BeI8e7{cF0X&A6VK&$jOAWV_KrGrDC#ffb!%MA ziUYf*7Ao3MmIM0+YWvA4x1`Bf6qQX^hK{!oYD`mQUEq8yLyzkjcgcT!Xii45!1-yA zB`G#Bi-ZmC@nC|rJ4hzVW>6F`;IK zzO;_}ott^^y`|=){bysQj_YI*3q<~W-p8Wi$WkxV`_X6HlGP{Kf*_=A>r^*pPk1Li zE-)SmGkYIfhVfS-zxuJcJHln1abV0?a9)mqzHbcFAFY)6s|#j%wEpwllkWYpeNQ24xS8Pr-P+yge_uYq%2*q@j8j)ma=tR z(@646B4s*?1)Y}7l%^7tL9h$`VqvEHwQwlhz&*Xruh(-IT?p8z6s?1}N`p*WDZ(=_Pxj$5yN6wyIVH_9+7NTjTUdUR<0Q(L%oe~Sjw09%#0kjnHU4ep5X8k7Uj31jIr|aB8+v z>=ulKXp+KO4s?-_}<~ex&IB>0wy58>69k`CCI%%@>+i z#9_6t1mR}AReB!UmtY}JocPj=qS<#8M}P~m>5XGSRa4f&!juK-rgnsmxkcm#CG)54 zcmBtrvmcswzH`{%s)b&5nW`H+%ogcs9St%RSkc~4)bb5=>2@WIjxsSEW_pyT?RSZj z(ReLs(=UGXTcZP_$W#c)=iIh04!p}IR8`VDs0vWuPf`H;bSm_F zhy8A~ppHgW_C&w6Y=?7+s9xT~Vzw9}gr`-HByDuA7^^@*O-`7{Yw;R#!cx$K$~__+ zrYo#Lc3jZq)(yFxRB;wY2%)ABy21`X|E2TZY5abYGZ!F#^=7Tq3#S{P z>820fRR>qhP4q+cD8v8kxu9{dYtO>Ner9dbz1?Rj%|$ex-A6mHJ*qQ7Lm9=EQluE| z5w2-#*|lt?piM90|P?_Ym z;$zXlpfhw`aH8KG8)mVdu?yqF&@;o??1JD?(Oc#L8He?gEkue6Jx@zeTqm18Fm;(1 zh*oR}&7Rumv&^f+`R1$|&@`yVio{VG%mgkL_D0>uuY9-cEJ(^_)0p)^-*JM_ z@MJ=*hQ*U7Cgy+>EGB(eG?Xyi2#Y`cxucxye8~g}j+#JaKgGhbQbcJw=q}Y6Y2i2j z@TY>HCiTj|iqPe=P}JhC*B;T5kn6G%avS7)yHv}exn4f6n1=Sx&MN{de43eF8oV08 zV}3E+%g>8^Ag?A}s*Z1L^)rl<#x9?gN}u&Sr(Qnmye%z|)o>cc4SDvG?Vx5EPI+g0 z+RaP)9CpF6P;1!Zj0HAn;Yg=G?|qpE6&E^HlUIco)cVr>A}vPe?m-u;L@+fIWnefX z!FH4V+>tCJVYKz;E5&d6nX4&tiRn6SYJ*tNUT&m80|hj^@*;Od?GM=>vN!53@0uiA zstdm-Kg`Qg#W6MXGVc!WV!BJ*0)1tY1DC&jm{;wY>Vuyz=burnVGBTCx5xV~?=bY7 zs^Bl>qox-=d7s`(HoDptft+#nteF|U@?6|-{gB?c^`M_|Ny;1gK`BXdU{9seWcewe z*xeMQK{XA~Z$% z8pTR5vcEYHn)*JZR|sz_^e%B5jp>_5uKK3F?V>u6iL7EOloc)wkb_E6q|%l2WtY9s z7!QIoIiT$Z>8N*rdX5J>{z;o7JQdPS7tI@R?Gxa3T|3~`?%4>jVq~;GbdPs#1rZ%> zp-Un3gxv|5@3kz*sp$XC4xp-6_#7q}6olevToSwJc+mAm)>ij3{Zz(+%V8hJra#T_ zIVkz~^EZr0a}ulekxXt#b6{uqkO|WEQY=t(^C`{Z$rtAi`j>d#mENJRN}ANC-?;|- zuXF0Vu5rP!>NLrq6aKr?tvs+FRN%7cMXFeJVN^9#YagO(yzuAz3*Pu+A8)xDe>Q|} zbj6=5qx0!%TW-8@s%(cFvxy()X}#B-sA)#rEd7o58j|Y3t}rM;jtJW2Q!K1$Ih3YP zUzxJg?UJ5)tN&KVK@AxuvkRGK=1B1M&|Q*)L7?p8+ZAzBcGMXi;ZFK7Z$oHX7-o&K z=|`@6gO-M9_xN=vK2ep*?BWR91_R@q@6jITw5GlGw~(9B5vVY*txlz{t15ZW`}s3j_VgHG)f8+3mq{mrAJpT023(j#zcYH) zOq+geY!6}7X3?oS6t$!}c*i(*Xk4Mi!BZI*;w(dJwl4Nfb9V1565QKw0=Lfz_W;ek zn`s16#D(A%HL7ISL0OXh4AwYdG%lz-@8Peh=l3}oZBE`7zxXqWdC6>!&IA(i6uXup zt0)bUITv6A4`X@S&%!Q;u2$@m*>8R=zGLeq-s0nS`@pjO_#i}m*PPSK#eO*O&f5Y5 z7#W+nurb!+P8M5@R?XJ#3d)mWIT`|KVqs$)V)W~b-Ow@Gz(HFRg69}bh3mazNn{_l zeX;{PF6T^)NFBvOk$(-P0h*lhn6gHi1!0l4uy~@y;{I+~Ke-=xKQIgUmB7eFj1+lx z(pA&$g<<;+jD36-)+yB&s#UmL7M+t-LRSZrT1XRv7C`p&3YQM1HaJO<%=EzhNs98o zs%bkU9kAId1v&mj%xXn)U@DD8)p@*F2)c|NeH@U)!EJeR>u4ac>e|$=f9t=@MI#+H z_-jE1vQ?S`9ED3v#awV?LE;P0F!qVE>aC4_lQnZZ`jhs(YI%JbB_Ta&;|%4$zHYQF z>dxP2$s%sH#evtHR1@2hK(TQYSxsqDqVGqhiBYOt+pV}DtqxuhUMk;i2eJDXPO~)# zvS&G(r9bfd*S_~V^JobdKg5CKvKAPY8FSaa)l7Qdx*GoUs7YNZdO&WdcS2l~&N zWW8@a@SQMYV;wx6EDAQGU0Z_1@by@a6I0!cCgUfU$$XOFz~R??6C;pLv9N7Pp)?P_ zjDfT@9-jfF0E1^eo|i+5fp4C!W`z`GC{gcU*t}m)<}9?aNM%%we+V9K2~nv*HPVU0uJy2;D05-_I&pFvT7DS$ck_^#!|qNjguj;Yaie-As#{hbTppY& zMopE8!@&sKV+|xm+&*E4+$`CzSFWph-{@-`{q|>mq=Vb`#DQJW#U|y z-#fQnUN5Val(_dt3{Gtm9(rW3#GDbzW2-d_{W4E8Vq<>;d&#NH@FW^YHf#%IoO$zV(8Q z({xI_XQOmW!0Di7;Ok=dC&OPo{1<=zY?R*`kUn*A?RR93{@OEf%2~rM?%CP9NudL~ zxMxgU+!~4n_U*ypI#lf2rT4Cov;UW~;tWYAgv2h1E{Pzw*8n=TcnH^H2`YGFL7`E> z8~X|kf8GYA=RMAsB_JRNdYLV3UU-gDn+6Rin}Cx$D9;u5IHR4z+;gr3Zy2A&vm+|Z zCz{haC)4$sYEJR@UiMfUs2D)5I78JYNGI$0xuAM5zM19px<}k>1tLb=J#mMe5HacN z{@+YD7xZyh|HgtuPete|(lp!fZwpk8Ax8~cA6L>Hbkpo@LAPd>$m2;fGl4j<0vsb& z4hIDXL6Ns9pIIspkPw$?%qi+OIlhj5h zq%(YKGCBA8^dCJQ&Z%*3+K=jTuHrQmkFM354lbt zHyaPij>P46n)g4@eE9qKelh=dPg;?3{k68f|3}~X<#uy&(n}?`Vu6B$DcfzH`EnG0 z@1&8+-=VlTtsc9C=~Qv#CWy(ccFPq%@`__J>6KBB+;>N$P0pkARD(9??{R--um4Kf z;qW|)Gh(fsqVXcz+TX+VKFi2H_-E&zyzG@@=&?RgO~E>iBt@ZTiSKyQ`dnY|@cmW- zVEFb4`)L^fqos=H{l_mZPBE@W0UuP%C!4v&VjMRtO-fCc)twXzD*W3iO*Wk?%nFAZ z)zjW-LM?Ji4n(c<`@#$Qy?7;po>O`7@^EA=ZFH;0U9n!ws}t5-gk6RzR~YMp*X(9j zxYMsq7ft)&?Be;;pL}41$~(J{c#}_FGIC?ANpNm}VjocCE>>+6h3J%j(*9Zdm*U6{ zD7u1XAxN#p)nv%2PZ1~nYr1aKXX*2=bjpOty1-mvg%6fir$+zGO50_MtO+=?Y`eY^SA5`qmz`>m zq5p%H2Rnr0cB&SpPm3YEPEf(blhtlVREJ3#XiY-ESfyPSoJnu<+Yx?1mFcNHOiH~f z1-;LPXKiL6#t9;u{T=25ojZB{pFNFss_Dn23&<93OPT|x7xtMfX*m>|MUfqprcKb} znGn^YF7o^=Fpu9xA9PBy?XC)~GX5E59+!lKj%f!(Uv zK)f#W#3;TUT@(v5VC5`Zc-nrKEiWshxWunTQNe#P#{`$4oddhX77%1`cCpk8LruWb z_G=oY;@p1o^K@jZ5cKr+2|N0Sw9>E6G_EPXsZiIDJZ@`>17qTZiB&15*fI*XPMW=> zPl07zsCdw)XmU>wAL4Z|I;EkP42avJt^yVn<}m#P<*}2X6;+HdmgGea@Ge12g(~t!i7Xf)6J@);0+H&gV=1-8@P3 zikK7&DWwKkuGp}NI_j*=752$nf}w4DlA?_)4cGgytg77(Z$p_7nJu^ALehs=iFt{mdUiJM6$oiE=vBiBW5LKKI1MQkNpIsIn0 zC|}-+7`IbZ6s#u?%4%FIlQMw zPs#6fbPg{!AXz%(lEbTmUI>O{Z%%-HU#sCieM~QRJa5g)XE`~Ej=M*^LT)WJuQ#<& zsM#uwpMo=w()QZZq(s@zkMT}%(GDx|O#~XAP7ViLJbh>ajyU1MaTE4YB|DqHVf0{& zJs!PB>RuTSw$o&lX{OjFiZoD~-fwo!8~pnIuP*uW70HHBEt1&`G&n;FuFf0Z>!@CW4cG7JSn#IrXdf(g% zelCP-F(HNOV9E~A=_u2GyL3v79_L8n_*(q6hGsNrksG3tCPP-;%CtI%FUZsNXnaRj zeZ79R%=c&J`{kD{hEf`}Te>f*omtQTf&5OXUhhEv&&1k5c;D8}vQ<-K!<)F7!T&7q zp8S!~+WflVC)3HN+^mh`+EKE~B#ZQjV!xoseM*Df>d!f+i`xXqPry*T!Qi9_J0a(i zsluN+B@!D-e3KNJymTmxDi=W4`fBN*9K``^gOF%DEZ+kGTmKr!YY#t&vQT}z65lxH zHvg#m&cH={?WaM5v$XZ{GXdJ|e%;VV97kPDu1d0*>xwF7M*uGE%iQA_?HOK)e>Yt( zFNdx)mA=~owbvB`ZuQPsFCWL;l;tpe@_S)P3LGj!RTH$`PkYoijt?qjaEQZh_XC|F zAQJcSibC)wDis!mz;AWfJkk!A50;qkw;k3bS=d1%YYPK%?eqgTZGrD#@G`GVRZajZ zB4BfnlW+9~>y(&xoH!F@t!*+cIySUE&mW4LZCdb$z?;5C+w_<{xsvF(*(S&J{Yklr zC$xuRcTr>~r8yD|T<>Yw%o(pMP+-y-v0`d3-8=J&V?CT*XBv+gP0a$=q^;*C?*dlc8*SFymXR-!Qi( zuXgkfxd&cGv2wuV~3kIk?DXSFE(l2 zAt~T#(QVAtx0d-_bU!jis9p7yqrhg*q_qRQ$NbIG+hNc-^Zm|wM}l#;o$`&l^Rz2m zmh!O=8fK8L1SCXd0nNq0mP+(H{H?67?flN>?{4N*M3ngMAlJ#wS&!Uts8)76v|QaP zI5#I%ti2*BmiN*(*s?ibv64M>ek-PytByn_3obDw0e5G?pV28z^otSZdpwprheB#4 z%tl#{Y;(yH9Ir5){eec9)xER6jAU?wnFA*dYE58PLa~Ju*@N88E&fT0s@drTy<-f* zXi+?@JUEpu57zEucDrxn=~Q{V1gL0){9->Tn}e!AWzfY+yZ){A@Xx}K3U}6XZ8)xQ zz^1|hGeFz)o=YF-2R#P^mge9&*KG=$=lmY7AI1OtH~r(G(mAmITWSKECzM~5%Uv2C z6B6?-DznFho+K+lu(VBp-OW(fp)WEnG|{gE{Ns9oPPxeC7@fc`0gtC?c0TEpSIc`m zEA3nRHXf*OoqjI&&zmy!-@j(GFdv>v{vJ8xz$MeL5{*bUo~77&ikzY}H)a=jK>ZV` z4Sq;c#I?aUWw&PPlm#Ar{w?Z6W}9a(UG9#@w`TVF56a7DU1B~BejL!EK1RA_>Fy`p zLD^Q;FMUY%@mkau=bogOgrr78PpdfT;jk8Qfr$QIP8)ElCt4i;vKOM% zJD1J5&DSRSA;o2zr*?0^-N=i)T7TOvAy!R)2^6orf91RD-!z(v1})%Aj-L5o0QqadJOuuisAHsIUG+c0_1=?^(nFTIQULn?h( zlFi$ygvK(FR|TB{TSfh>MUU0y!$t_=gdVE;!OzqFceP;rVn&X#%(YxH{Idx(4I%d^ zuHo-?ADmVZnirWdHz9IxS_9-8Q|ZT&jjkmij)y&84c8xdSGWw{{lL4CM*jHlST$ZX zUOBenLjNJEBv%`wiIZ zG%1e;WkaSLJImP0epojd9A@9T0c~`bWh*n~`Xf*C(rPYoJqM2ASSY$SbnRS#>>sFs zOQk;sdIZ*zHNDvkO*w%ggo_oX0B-R56m4{eVwbSUQ>WbE{y>%x*#e}7#ll=67MR)3 z2jT#u(bp>*!Nz=;zFC9B+=w`AzR`lgh-1zZ%rM{e$A5LoM(4%SPC8dy26c_{v1+N=#hfEQ$qKr5Pf0U8*`7MMd;o zACTjJKd>u8+YK`QeRFTmJs7miyAm|%Zim#%OMKh(MY7t5yaJGV$fi%ox2x?5wQ(5` zJ3?#Gwd6O55hz7YX{X2v2j2Za9sP(v%O;9Vq(}m#x$RyI^@V!98Z4Y{V)l|-GqP!< zZfC0M>0)BPbz#(*;WF2w4q&H?oJ`1T{}HQrdzuj$!+jB75kJI27sW0bBle4g+FbD! z)dBi$dErv1vd2vIdGAiScG(+CLl)Z9v~a=K=-$ZFtG0xiPwE%Hf7xwGC*3JPz5#MD zLDACSh$c^Q;esnsw&vL&vJWASR{2!Cl@eEz&H_Dyp0mGNJfo6vasDHi&O_fnb$ z8Oml<5uNI)q!Cn+4RqeNux{7|)y`WBD$Oa1UjJkP7^PFnG-regb_3V832?Q^qRZwO zlv|KJwbePxxm!`;a-09iy;}iKBq;WW^pkbGc178|N{A14i?qoekSX#jnhvmQ`k!0# z3WBtT@k_miX7%y3)q)1uF?utLlFm&(OZGTV^eKl*`sS=;k60mOxDS&mf-B}4w&`bS z&Nm5+Q0g_{7SG?oOH!BZG#vi5NLSa`#n9¦o4y>nhtBfDO%TRq zEAKwvjiya2qQDs$EB+x*PXR65r^SvovI zlEUyeA<4pFfyHUboa9$C>qDs#hN3&74@tZO!*Gwu#w?9uL8LXA(j1eZc4%!7E`+Ea zs>M^C5>GQ!rCy*Jo+HHCgA4-F-_l*e&-``?FGbp8oikga#3e_t zlGq55biu~!B#zr6>A;a~>f6pC!dPL4%^n;9&W+AHrdXpYU+q9j_Y^bD9O zee4)*&!y>oU_}z5CP<5BEO4>|dY&0Ff8H*}Jhjbb3-7>$D+`J69?yC~2ECYI_opUx z>^H7T&Z&!oN`qD^)+rjiKXL6PYbMtMTc%c!?ULfxF>k=NnL#bPRKHANaZtSsJ9YKa z!=q&AN4fuX9HmXrMtb?x^ciu6Yx&eZ8+zkoL64~eZgTbv3dVZj(0AC}@vuA{*t}R^ zd1C6RJ*@odUew<&DKijh&CD_#7%q?f(#(j6vy zIyR0Q3X9|G{I+ySGYMU8>pMdU-m3z8iRVhnXsk8%~OkS zqmNE`*T052MmNy4!4nM*YjB}o73>cMli$Dz7u32N{@?UAI*E)mrU3f6w5}q zsk(*s>ZyCeDS)iOI~{>f|{H`YAKTyp7f?;>}uUx`j=#+M#e@ zAbOkUD&~~37j*6}1=lFs1UFoY#j9rSiqf8dkL&0(W}RQ5_ua_KP>k(dm83C=yneE2 z#=aRlc=duy!A%kCyf*SmqmD7j!WMOZWVb9uh)26;$4^NXrUn|GOcpMQjun17(*}UE z0lG|}W;hu?#{R|k>;D~Uv^Xo1dJD+zS7yt6(!>^3Q0xJU?8CS+Y9-zWm@J~&73b#N z1FMq|SqWh{2m@772Gr4hGfww?pu`Ei9hE#u_i6vx_V%R3bk& zXHNtY`gSr6G7J=^O&)T>{^u#*SqT3(egN4AKjUv|{j-~b#e5@vrf>P~F|wQ+ejIp5 z4GMW9*0PNhyMZF{AZA4WrCo6VV6!9y4$9!5P0;9?E(U4Nh%VL1iLAuNuRj-HjNe>X zKf(veh2J!1OE@eM&jSBpkuX)VGO8-@zT&)SKXA4oUCHhee)?96xx@{-wo11UZG$@?C1TZ-G<`Aj+NhLhy{5C_4%`r7a{EYRTrduU@wYx+I&=ia8*L6L zt1)B2kPKfQH%QC|;Dnh;|IJo?TV}+}iEn=E3R&;Km?<#9%yx7De~@FIQIumx*t&BU$6XAjvfW$ySQpg08bBDSVxHCtXYzD)vRK z6AwDACx)ZV(#*j8=vFopszXjG8<|=h5I)^z*R=#~zvd#q=cw7AY6nNdf zrowC{xj@70QM;aUgGT=RH5Xp?%?+wWEf%JND2p~#46>45XOMr=8jdIaBfhhF6q`elEJ}k>@=h?@8{Dt@Rk%EW z@KlHYfV)n4)o-6FQCtRX=hj4DfU+G_F~xJ#X}S*y{z?c%?TM)L+c3EsK238@2(x3E z$9QMO#*^`6IBv7zu6hI3qonO7Qn#P1mm;Cw!$&`)(Ax~z}o(G%|;<(|- zFIpnfJ&XuxeY5#DB=MCAycU@NJ%eJ?DWb#PdS}0O6{2vLL|tB4yjvt$dR$mEd*x(Q za^53f?b4xm;CY8%0|m4DeXmcSe~0Pvy5Ty%NnQ7^d%kysvZ z|5uTZUv^IgJ655T_b?^-r!urE(Hg8z0?_p_8f(`tve?ht;)@xU;Zmu`8vrq`SNa|SV$mmpftPP zn}fEfI^;E8_XI-`sR6_1_n{v~`Md$%qc3&LIyfO6)P~;$riZ!%;lL}-O%u>wq}XPPG(qPFIkbHVULKyLSQ)j_{l4N0FKwQ*TE4~u z*OL_C?U3c+*eD#EL*T7-^i4?&Tulkbhbl$wz^UEivBOwOGG^8Q&|JzQdPys) z*7eGgrC8Q+T+kaeJ)ty)k=V`)e>C}ZaT6cXCW43r5dyJZ8gBRiqesS@7Y z?|YIQRGmDUP#0vUk!9U}>huu8?t=f8?CH6FsdR-;ZIEF!hpmGB z?SKAAGUiEj%0rMi25$A6ojz+k291rZU=z*TC(R4* z<(JNBflgE{${vU%*M}xB^`ZL|u`u3o=&gspp^}#YRrXoo)*A`)@2uU8hWRhF@|>H| zaM;~%))DrRIa%hV+PzyK(8RGvd53w1e^FqyMP1{yo@@(9lK@=RK&X*-srQmbT1#zma;AK+<;2aPI`yrU;wUqiR6Y|GU7G?xRs#ga+_s2$T(Nx}GUPi2%|OWCFu9Ok#GFud$;;-nD!`H}Zq1m$oHDof&E<`a z?_SIAdzAgLdJBbg4oH9KVDR_&!& z$Ta0sn!mOyVyJua7IjMGI^G%H{GD{ZsK;~Aso%RHWYDQPAX_o$lG zyv^1q6FTNg-0a=y4L@BUWW>$%)W0W?EN;6l2aX3GHbGe_#THYffYOwDp%Uw5QKDZV zbJSUT399BR>DYIXqzBz!y>_}m=_9(>SXBhf8}#3C&r|B(AlMGPn!UbzL|Uvg!c!FA z%cd(sD|yRRSrJ*`SbBh;+0Y!>5@xnEM?WZ*Ju)v?bJ)5P3nggnp1XusgOjO0+@f0| z&%59CNu@7{+C3jK6)snU*BdV9@*2Ex6?$?Xl+OB8_&~#8SKQ3)@whm*Vczv`x2PL^ zili45pLjo#;I(e>eYXZZfKQf@i*qr>vEggiziks-*cK56vovOFGvYyRhR5@xE^qUK zH4BwqkOER*a_6$BO#q<{d4*eZ5PV5>stgrQL86WOGK~078ugIF`ycgrdtK!8{v9`3 z_S+x4yW#uB6-n&%!6Z`uI{D1RlwYIR%M`hY0kj*uMMC7CB&$!d)zGnH$J9?|HHvod zRtUHIK%a**Uf6-?0tBwKwc=*UidmWTidj&{j-qqjbQT!)LdA7e`m}4|+K-v*Vf|s( z!*KW1p%@A}8}*s2hCxApEjD9>G?54`eYL8{Q!j6bku8jur9nKW?yI28i}9TnwuQxZ z)Wwn5Zd_aMjDOeh*%c_YqNrBkb{8MkDIo&SO~Kvor1@FJb#{{vdO#W28O3b%_hJ6IfGcXc2VLAUq9m- zl{fT*Qj+GtyX{I7lTbjhKw!?LG;yKzq6CjNK_5vX75p8N4yIjn0qE-;pxV3a^?qq1 z-4wq3^eAco!@XBJLIueuE*F$xxeBNad%7WvHk_Lsb7ubI3 zf&`Be-&SdbA4oWN`s|f=Fva9&YjcH{zI+Rh~ldhirq}XoN3Ah*F4h2aZGB^TDOz_ZFIf#hD$e{5s@w~ zg#Rt)cQECG8n086PkK!po;Y9$*B*OA=*jq6H{M^(3uGO?H#>PG+*!vBA41E$kqgRt zid{#MHI(MMBHQ;Cn-Kw3ys*JpP{W)KY7E*VZ*-ry^<#Kut%T6bFsBGDS8)WiIe}V}u8z2Av*7cY6=;;`qJvK3Mn`!)~?+ zcC&*{dGp#8iy-x!#k12Mc9c1^5(Q6pv7?M(k4u&f$=>#e0XHK)esY=2Ckd}Z+J59K zPp4QNMN%luC$8t99{xaJj8GqD%$~Yb))msjHiLY1C2znLs|8yXHT<4grI2*Zk3Ox+ z6=SE|3HUB@e(m}Ru*Zuw9ny~cM7gwIz=AtTx2 zjCAWx8k#?|Rl#wgi$hAqJs!0_69DU{VB!Msr%&1c=(~CQGtJqJTwD$ZUUw{zG#5)_ zd>4~U9(MfL{ZH5QTdiEl9Ytm3> zMcsvqN%kuf;5cHkLhhqj=o4K?X>ynylGyO2;amJ40J~%-z0LW-ltHHlib4OP@Ij|8 z^2y8`9>|Ef#40NN@#kr>F#tbrdFzn`C~=V`p0RK~zs<7%P8YM6M0X||e$Gz~te4+# zT{F4D#g=x58*-j{Eu$V;K<8L4&;0I3R*stBV?V|2rAQH_ z>36-&-vHg@AY+U2*ERkXkigA_sxX6V1&-B`;5_I(hho>Q(nhEy$>SZLoutSYuN6b4 zSFCN7;#RrA_|7Gt@;F>(swQ*nZeMue+*$5_Yfht5Oi%4s{04ea zEv+X&`h-6Y1hms9NjyK7?j`$tAA&);?Yb(!R#Z^dVPRA^#L_{Vz{T>sGm1H(wib83 zXRbKJWp(7h4vqyaB5ZR;QJA3y*&^3nzStI+bakvJ+6)piX5l50LiBJJzorHIo)H94VaKy6|>J7MxI zTimf478ajipA(#5F(E1cSoTZEo>X73Y!x`X^WKNO>Sd3K8Pl)J{Rf|s&yw-k0W%Zy8e9QESAG;h1rP-gYyR+3T}vv;1B2%6f-hbG3rtMlM%OAAJdVUVN5J(xM(SUL zlG)gX-E55Pu*=QXEkCW@4mU=7}=8Jz>zfz#>BvN3wap#6%L>TzsQzAT=WvaE)6Rq>7`326+b1WWd@2 z)HO12s^fqgm~fCx8*2`Lp>&n(Oz>I|`{+>B02St-6Gr6P1-GP$Kn+?s2iMRAyjIVB z5y!kWH{G$?09Vf%{nS3vF%rW& z2~d_T%P47}iLLUirQbC|Dj;UY&&Xj1hLp|(QfDakbBZ*A#ws}?%i&!a>M%;^cD({6d(lAoIv$6TcQ6|rooTuT~34g%?MQfBBq2Y~y_aQAs$w`&ith?n8M z-n~^_D@|h_z?E^#+9_K>F%Lugcpn}~Rp!#`KuRKwNgx$2t?Eut4v2Ryb3g3g!)uWp z9{Ma5F58)oH+y&+UDH%8Tm>aw9I!ayisM@U(kl_gBa>DY$HcR_L4wH11jp^W?J9r#$^xS9J@`9R}^q)ByO-{l6-+e-&CV|S+ z@WnM*jZ*RqN7$*1^lBS8>AXgd`Hbbo`3t!`= z3(KZAhOhDkWlvQX{ZW}Hjb905%TQK}^^)l7>Of}|GKXT#jhTaCcxS#4srcJ{pU{QH z@jGagC@P$1IOk{Oz@LQu~1dJjfy<|YK5dTESBvItB_#*z;-5%(YxISv5h1}PdKidR|2JF zHY~mKJb*}tNdTq`t+EWNvh9yDt1vav$czUrR^99IPV~GV7*Vq4=f1xn#~gU>xnVNy zZ4}!|kroh5bxDIr{i7M1r7dhXo%e%g=mu+5k#M#N9Zb-+G%?v4NwTxE=Ue&6jTMTCk5d;MY`+P%3AsPdE5MJ zLqt_3AFL)u4^B1@6Q z2M#)6@C_M9k&;v|2WA&JPE#+!gghA_ZrFHoo833K%}?vsmr(y{v^=w-^)tvtZu64^ z2mc?K*r_gx{emJLRAd8bCNEvp?U6K@Yf`}U+a0lIOBjrJ_&5%5GEx(M^Yb~)a-)$-ef7WpNRqf2 zDF+Vif?~t4qK9mXg^JBoD)Kl;1p`YhKOuZA|B^Tx#ns3Wd2cxCjBAzUl4j6{X@@u^ zp1Z0zOKJkDT(Abbn(m+5XZLvJm{Bzj#8`9de^WbFo0q^itfOP0+@)EWP1lfH&ij?+ z(==rw2*2#`FAB@E`qqh#u=}3&aG<06^V1IgLCO7=Nf*EcdKXV>?)UJWmia- z?6iNY8nqk;UaR-Eauvj7Jgo%HbNmP{ztOsDIe~`i`tJQN&80g=)=fF^v|>SLNH7r5f$*babr*6>&lmS;W1_U9~v#phT5eZ#y~%3)_1 z3)N3ZTnyRG9g=e4+AtCwjXXwPqYk!(F$AQ zmuq6pxdpkD(K)a+vA{Zb$*Tlh(49Q&Rz;CcY-o=0y6<@i*!26t56wg|@Dh3{vwOx1 zD&!s-NN#9)ae(@|;unMFQMhjiJ}kuSx|of;E8fsps7ed1l@7=s@Gj5jQN%Ir(9_&a zXOQC&wbI%bTFA`_8JosrAUZKr5{R4w~{=PY;;Zb#D4gzirxNOe7b4;U5&1l zQ;K+1d|g;PSe{-8t|htT4tL5MH&@36t#`A9mH3}{% zK9}m`he4HWQJTl9h((6$3jHhm^U0!G>2|k${x$Yk3>ijeyjU?GV_WTW!^*!OJuiOQ zV=GU(MNqk;%k42Y%ZRtzR^PDY=U9gc%lF&s3@1!XjF&A)H7~h!*wv5)a@)S}b&9sI z4v4H2`=3-jglV8vT_2zqR6)uAGJeXG4AqN;hjqvpnltPLf>EE#2^rLfYGmT0CEaLd_tn`vR)i z41X@?F?jMV*>{867&vTsEIs*JgL$9YH)IC0(ABmoqFD9eY2OlYa$qen*-26qN_I*5 zX}!GCXE(1OGSIc8j5!Uaqg8#D>3(wjlDLd62*)D6^&kna7c?oc9cjN`G|YpTm+lMc zrt!hBQkzrm`?2Q8FTlcqBuNk4a}8E;nC0I1o+fMImHEJi13SHnKu9Md}n) zQ;~;G(*(0&19aUYpSMoY=ThShQwzpwFy#P!vi&euRkm>Bf`4vVZVygW{PZ4nt8gxVb@+35(uXP`W02K?SfMQ_o0SvK%NkB3Tt?w)ufKW z%FcEl&6S9jITi8@7M&66-+fNsr1{0%4Tt#??6c2tVqGA3`>n_B^oAe(3#oBnclx4< z;W$OH$0_m|6^RtM?SeLGziUjupl>!^7LL_)g<+c0bQ-@|8PCUx%G)7^hNa`mlqqFl z$cu@s{LpIvWqH65a)Dryvsw_#_7S}0v45LOmxB1_GTETh4(M`1{}zbX6b+=^4THv- zo$`~f=Zmqy3H)hnDZ{adWC) z9D|%4nnWLTN>jxH^8Eo_5a#Pq0F^HKGB|VHmsSTiD^H8gkv>;Eg~ZKFi>%0}Mb--e z8@!~Hxy0<0@18N1Y3BLkFg93>cbc){b6GSLf9lqBGsa2hkDv7(AxkHbbQ3$ifnwt+ z5=%uU`sk%uK$x$om9F4_E$R)A3d)z%s#3)-M0L^hkF@rSHoeRKdzQ`jXeCV!8=U#s zh4x>+Vnl`S@2eJ(&D_?|4(y~wM+M|6rY3GTOEHbeVXDijoYA8F0EB%tG0RI^_VegS*WR07KwVo%b;}{TiQ^E&xSU`N|cOd zAVxW6KGtL3w~Ug}=ETcwhxq=+sV6Y?(*%;^z;TsBCX-18#X`GTF%^kLn6W-JlkZ8- zh)&Zhy!2C}g^jYEb02voI-gbPrj>a_3v=kx^q@~)xSn3Dp9-Ch8@=>&hEJ(S`LxwD zmJ-csb&q5hX_V)xa_QW_JZ6gmAGV9=AT6+SZi-;+0oRcO`HcY1gsh0-Jafu-hYi?T zAaPG~+v;2nUWcL8aKB#-s2HTM(CqrdMj28oWT?7jXS_N*@o2O9@Ow*s6#dqnU$&@Q z)rs$4mc)=YxV9ELP&&C%6pjRi#UtZHi^b=T_XH<1Llx`@DR|9jW}L*T-6VUoV4DN8 z!yGg*J-aBjlp;H+$SQU{(C(IaTy)Va^I8##lMoZ(6MY1RAbmDK=Ry|+!{nY z_aSz;z%n8(zqh*HJZ$f<^NxiOJ!W_YH|vT z_Xr-_FUSnHp~w2{ak2&q!!5wGpJuzWmZ30;KTF*R>|9{p>gurYu+Y}pBtyw4*64w_KIlD9HccWR+&--K zHJ4&xK+>tmc>dvUU6$N+#qbpJ*%b41ihAaVzmAS!;{!_}A#y$-QxYZY4c80m_~|@g z!=BlpTJL@nNa1eGT*B|M^GnNNK!(4txNXJPI`&IA4T$50sP3+xQ1wS6QvTHR>p=3Q z122hJn{=-|q}clu>Baae>cB_KV%ccf2Cwp2P~s-dp*6_704W6MjLr@02# z)A0U$a@&D%vChOSMNec0DDr@cY?2KZ{LNOKAr-)Kh%K(rSHN_+l`vSKbe6s+!~bKo z4tggyd zWqVZ2M(*e%;td0B$ZpD(t!g!RVJ2fC&d;}B#(-$q?h~bltVuRW+AUZAo&ic z%f>smi8Ogk8vm|qv@p>JQGyQokjf5DRvjt>D`Oh}Iu9?cKOl+R5aYn19f$x9TRmh^ETlHIC`Ozbq`?NK8jm<1{6D0x zqGpw{HGaToMIO#w^V%7*N|qmd%!7|!Ji*Q zQP&`roFN4s5fvWzvmFe@u;Fr4Fj)4Be*RUde1g%Ms4{o|l{|7_L4qwN_92#HS5ssK z6?qU!TEJ_dmM>hpAm(lRj_|O-Hrsse*qdGdqINBh{V37C39>i7+~;P*!>=!rMI?dS z?Bc)%Wv2;x(kT{XbGA^C-SqK=NJOtwY?j^(*(A=9?3uRZ&3@ije%&+@S5-sznpW8$ zMmEJ=z76zQvgOT&7b2tcTttnwrhWmiIvM`YYhL#CGz!&V7^X`WCB$&bpc7Pwf+$56 zlqD!z|PUYitP7T#ATy#`^A5I)}~&#;DJPjm}%<;yU&!v>P=#$Mdu1 zZRBa)j?hgsb_v|(56GeW@saOFzSbpniY8mN)bF$?lNT5K(7jP+P+lsRAM!;0ylM#j zU}HHxtx2|O?tr{Uz9n?fX_>6pMW<|%;k^}}*XS*CuaZV*yvq71(P|)$W_q|E8;-SS zZqN9ndH**tYaQdc3qm2${L3h#gk=c>90u!|-x#-(){Hy1C)(y}%LeGphdrUbMq~5G zorl(u%t@rq(nM=ekt6+#}thRKIH>sH>W!$mAhE$5llSuM}ePcm;O1TmNEN`*m0| z5LUcr*8V!Hud_T5qj-^y>jb7x+5a7h5jFBp-)<)B9601#XaeU{irq?)WGb?kv_fL( z7TxX!*{4g)ZPyy_uVhJz^YcK5Br_D~;Z;aRqCqJwBGp^~QIT|S)D5`hyn4nCf6W@B&%0f7S={b()2)TJQ3dbVVDn5HXcGu=g3ScoAG@;5yXjx5aGQk=d;Qe< zfL8TY+2OF{x!v@q*A&O-8opMUPD;di0af17bGP|-gl+S04#VYf9`o@vk58J_2j_oS zJAc5bcK(Oa)PPep?zO0M>D}^+Fy}(zTv)|cx>Ztr;mNAR*IU)a9%UZcyxr0c_0fnP zd6&PI)baXUQNLs-Z|SW4qL%r2s*Qn=yYjD-E@jeOcKa85aJ6T4K-uUJ4>v7rB?+Q`=K2kYHr+7f8u^YWw;9)y? zX^f76nqlMyL7g`}y-$Vp7nq}59$X~SO)sAg_ZjBK{l2##cl3bPP49Qt;H_oh`N9Iq z8ce+~91Ta86_~P&qfz{`iNAkqjs8_5w!XjPH@_oWxvkV4*TExuO^{eXv3V59p(3${ zJXIVuGu?YYeuS@r7J8EWG zcr?_cMG2e4r^L0YM|^8=V#Sb+ykxxK8u>)q9JxJXy)f7amsLsK#iVewUIPaP#W535 zR8cITWH)%yOrp=1{#n!0_*ea#)g9km`}UUaeW?Fgi~71cle#cJlggx0sAl&bpp@F| zcYrPvRnF;!N`=GD4Rj`T`Co@-x4QRMzjgSnI4T2XI^2%=mSH=adh~C1=Hs?jiNE7( z$r2j@XS^sGZdr`mHi3y{6Xg86xDZz(ek#BH-f6PNfu}OC7sHa!TPSuj1>bmNQs|fB z=Fsc2A4--kFc8w^F(@Ugm+L*Ng3}~dye|psJS!k17 zl@3whRf_dfv2wGOG4Pr}7gH|Kv6js5cg5EIRB_9kT0aODmdRV@oTUfkS0`Vc3{R;K zD4KV%JVg+I**o@+#v@6t6p$eu|AoYG;h_7RG$r^o>+GG7c8 zZ1e-VCLneCrHFpFOA(k+PK4%F_>RDBfKzv-*D<$f6k+Y>Y1t~5MNTslt13?V=&6^wB`^eJK(slasB(g&gP+JhwXr|5Ci}FWl0~=D!=;0 zWl6JomEw@+h6R@;E$V2fk7)tjl~#2bUG9-MyF=AKcP+boRqz7#-3THDoTZ2&!*KmR}<(*Oc=?tR*T5vUQ zEykPoi>aQpY~zFqS6-M8V$ zN^(u{!Or)ReslC&u(!@9-uZ(QKVSK?@^44~YTR{DHX9>z>(604M!S^T7&$C|QSAN$ zcb$yjDY*Z;Pe>FupU8pBf}!eW*d!cJv9T0cO+{kcE@sV}vFpojrkvL|TZ{d_O)khzFgt{3O zo>_E*U%t5Ctw)C8D>LHZ~uIL}|u=wt%6f1s=;seMrW4Km>PeI|* zB!%&cxZrMqt(in^0Rwwj5N^o`R!ZX#)KzPNT)yC5OFCwwR&iT!JxOaX+5R!#0BRW4UW*A?x8yDo6LO_~G)Z^`KN5dR@Jg zf7&^Q8hmR7bxr~6eeYzIVQ=l^>*{-oJFgEo;juz!f`VEMY+BgjRqeLT9fjm#Cgls0 z*)Dqdq`2T_MG|k9Z&GO4wA+f^Ge9UKY@=77>jBaXj`UekhrB1`D$H*Q;$=Z`!B~G> zF7K1WMl}7jZrfS1;TwY{sQDVU zpw6V&G>U8kw(B?3#U+aJ8CvBd$uXeS!;E%Ps9tatDg~>iHUMAaF8A~E9!ds((B;>x z&LnjUF3-Oiv{T%o9`oQ^GY~`9j1eVPKYy$je*DirA?AW_4jbvUphSm)*FEx6p1UH7 zJ(2?s@h;gXE7#4fqcyAL zItF`>aBPwk3E_=&cf_DmzW5NqJ1gPNeX3)^W5PivlxXY%0$%Ve6gASCfMOvy8_4rr z%}eKX$?%tqJ8H&lu!h^_=WUN~JUI=P{#lSbpmJ?V?O%DSs0F(9Ul-m+|3EI z6XtG8m>Ob)M!HYdMpDGhM{(f59msVJ15!1`?xV;aDsqQppFi3>4BKtiIUeF;2&WE*guZG0)3&W7JDo@fyZ>3Jrn!0I^JeLS@chCuM z?w@>57Z=MCHHhDRpXxOJtZ*`6Fye;|Qp|~XC z&yV+#6c|Fd=he+`QntDr6O~KwbS*jTx*IFj=tmN~$DjnPdp&bnoo62eLxt zqmUAapkqT1^1kJ{_emjo-tQX6q=o7f=;_5qY?%u+m`$(_en8Sy-5v%q4-HbqWvilu zyQUS1x+O`W6`pt{iok-;Nyh~@$u6mYaTr==C66Rnm9Y2K9xz-np78t(ay%lJJ|s`{ z(KNZ&kR0bWX|($IR8V}4W%HOmg5o9}^!cFUp6MVXH|P`x&2jL_U9tNXlLE+Ohc@s$ zJK*~bp)7RVqN(h+|3VwV*ZUKnV`PZ~yT1^m9kwD`PqFJLvX+Wm3Ow1lOszc2wSc!a zqE%f2v7t0*Q?bJw`8*3@o%MO%Q*3|Uvh^8l4qbP6(;BG}6r$UrKah9_21StxDAFhv z#4b{($SzsG!f4BGx}z(L)lgWq-r}4tu48IE`XTIx-Ic?sGG7E@tb>T%y{Zuy;{=h3 z@BDm)?qyf(O5#;1F$kjRQN*(SZc(#zN<38Ruc`LWU~xvpUDP@+^^1ohm+tAegMba7 zYATW!=@cCyam@Z%2chBmuq$M#p?5)(r$T0aY%(#Vw`~$Rc|1~M0*)Dup6)(=>iBJ?|)jMo@(o!sNFK(tHvw>N)Np@9{p~ANAoo-v) zP+L?__k>>sML=|PAkBd?ODTT0>3fjHco9};OaZy^W5tRm%vjd(A1}LQdBE&|il`;R zn1E*4c4pVKH1DtcmV=SlB~EdLW9s^!x2T(C4;SwJSzk!E>s68)h;6l7rDY!V%;2n> zuVhG@)kOx$BgA%=#ftIea0w^NLjBXfub*Jf z-s`Z5XbVie#fBo4xZtE&snfR86+v;#a&?iY*zFjp7xxSMT-)6s1{9ne&@S2~*&&H1 z9rDUK+k7%ZQ+@mGzS15uOgIerST5xTlkfgXC^xTaaafpGs8c!Y9|v`GRlyzdMtQYp zpMMDt1WcGNAUVYU4mu&{N0rNl;Ccz-BP|pb<%ga1(w*{?uiMykj2y4AqGaTgYm)*bW>KYikm~a9C!(|#w1$wh+_LGa*v8k3vEL+fXP}9=zUI7G`q)%_0Ttq z;LxB~8z+t1t%G$F9=$O`2aUQ_qD$fdc^8Q8 z3=zj9DUc!%6Olb2DybuQ5%6n5pLQSFAAnUGNEuteYXza4?s;j7^`x8srDDLTEn>;! z!=#taWtNae5Yxd2;-?yPs)kO4Ay3-yyBp|6`Cj^HsAfxOx5p*P8t|dw7|od}x9E$a z1JiqXS4o<)CM^_L?!6DHK9^pUOuTv~G!(er4eyfGh3Vw0Ju9glq(rf5ZnOIKFAlw3 z`gV&tDRjRGOElXQr$tABX}X8k9M-BnD&H@9GC*`IXz zX>hGltEzX`DpyZVrFW59!O-$JNzw1tE8qErU)-!0?tH;6f_=8Q4Ti(!zb^#WykjopZb2`-Bh2suS3|jN z84WJ(aZxPW=zN?v;9cdNG#Oo0t+GwpA&+-{0k|-ZDrb1u2Mr^h`VE1{?^IWIeb=~v z3W%BUGjjMFldaO3tOL(b?B^6|q$1byibQQv1Jbt$F&KpLoq<_Uy%W#J4AsD_d@*XS zV4n?U(D9%1^QIQDRbmb={=rz~USgnqgLd~*Gkafo`t#YX52%8-qPA!;L zFs;hv;nZSbF=WD!mcK)uAw~bW4H)0{Id6JnnXCyu!=Mw!FSkNNBF9&Y^ZQ{odvVzv z@e(}aK+ccP9x!id;o<>1?wSyogNLUl*$_tH{P}Gb~)N0VcvWztoDS z=yQH|#9VsPg2Lk_D)xufKdo4pK%HEmS@-S>eYKwcCC~pNpI-YSx4t3%iT8zl+vG-D zk^1U?|B)ndvlR|pO;cuKE3zpzgCeO^$(d{SOCy~P@U+^A^-9>?ba3uCiYoRVKnLXxS7QO_?pQ3`AWXt)`Z!JxmJm~Zp zDe=&p0+G*C1OpZ*_qj!gmxG~hcNzWAdwL|tF~PEEs5o3VjTA`BRfpqfNd0;LXl0#xL7r{ zC~TF?P{TRIyO>KiO0Ti;)H(*3b3yBvuU#*5{mIKN zRDg97Dbk_lm~M|XvOyv+06He(Z-ZI zrJW!vxY-y7o`F(nyVC)XR2)P{W;aZRx-=q^vE*fv!+3-yNz=I!&Z^6n6P0VUA~ z+1zkEc1oD*QpdEaaV3kjEqZA@-*7CI-r!d-UsJN+X3#kp;eBD*9yDWyuwf%aR(%$9@8rU7X%5%VY3JD!mfRk0Mt2?v$tVV%dl8Cy3#hhr{amg}lvD z{1mO~M4vpSHlmfeq__}J$8^A_+8$Oe-$|#+( zSK3VsPcy|r@4!h4+E*@$VnadDI9rwHtB2$u=%6SJGtzNfGJoZ#r!#qZz6W`yMNq!m z!>fi}Xr(miI(@4c6(rogQ9i~`_cGOaleLyv*RbqG@vT#X? zlZIybbg~{6t%FX9;vOLFFJY>K>*ufI^}8lb)?j~Y0{J`c0nHfb($;wy!t+`s_Cg;B zh-INUI!RF-e0o|AG@$}z8iSXWKs#$rsPS|HYOcY;I7%4r+$Kt!Jm!(JdIY%5^iQvT zGS=KSIQamM`wkYZ*PM9Y2&CVX{8v9|a$ukLYZHWDrPxapX`>Ne(H3PP0}PoK?XsD7)3*0gTV|DDuSy$?C<8&e%Y?MTqqkxq%O*aMc~bY4=9w z)1pc^+{im8D&*%t+9{Vd6kqi4w#>z1+bWk<<(9c8pnY?R5R|Yx$u8e>(3-Sgumu`9 zx6EyY8;}J0t_wDZ_ei=uG@Z;Ld0J>Eyn)w3TA}w=tF$6lWq0sDeGSaEp8Ya#0{;a6 z{TJUiryF*jj)}gEiioWc6qK3~8-ACUDG?@u;J1ON@>B3&wcRsloGo zrEzLgAN!B}B;A3hwi=VEt%PEMp>YQgGQR%6Gmmt=UgdHJ*sWUBt!lmG9=YPJ7lV+x z2+y^ukBM45VO_0GQk?PXR^0Mf0aGC6IFqN}BfFRqx(irskYyMf^spcMK!82g*ET`H z7)Sb9w{fyG)EwoW&E|P>hn=b{WWG_z;=1AhA6XiquRhi99Pmt_kVQwBR@EfaDj!Nu zlG{Ni1aPu2EK3E15qwP_IpGHTpfHH-CjD+Z<%MCeZx5u5iAB%MJ+aSvixXBRTs|H5 z9rIqDEMqRJ)NKaMH!%b~l;R%pcE?uNu7pI(+6 z{K>`n$V$@}4zw|F%075``hJhu$~=99UA8PE=i#<#znN;Bg;cStYRI-p1T<-eg~JLd z7Mf>ssYpGIkq__?(GNVd5P4MhfsPEu@{wT*)DNTsa%d}*7xMIB{chF4wE=n4fq3Jl z8&GIK$rCPJa_AZYl)XYcYM;Rv^AJ3Jg?gbha~8u7AhMNmDAre$!w zRDT(0MGlAkI*MIOkyThZew)q$?q76qv0=4YX&>M|4|P@m`+1MB%XtrZUNobdb;PqKpqsyv+0EPNn(C6~{rR*TGYjUmst>DLrTOA3bd>Nc+330|;xIuq z9<6e-v{A4s0{mUqLMW<$iHgR%^@7CNpLCL9|Gtnx-zed-pc7Ee)v88?_2siZ_rRmq zJVsnt9o#Fr7co{WT7jmg25~G~W5n8<-su~EX|ylD+_&{R;gEp(5(Q^gr7#cNJ3UGy@%)>fcV=E!EJEqx0eKn|>$5cI zsOn2mvZ@!{ng&=UVE?S9j%w^J*$E*zP0b|CEq(g4_DtU?2d@C@NW1- z*>HDMklsjr9v587Tw@A9PQIRQ3lK^m{0H{9O8L(Hg}vW}Q!o;;QO1vP!lH zqJjNxyXM6K1!yMeWk5HXu7={I66UZolwP!w*wC-T@|cd0QV%TVXaxC7IMGfU4&Yio z4rld+wgt@~YZfKM8LNTr4Y=i;@q&d~ zI3VNkpuLcvoRC4~<^AUrC!;+oxc|FPNR$KHBdy8g7f-RV6j@D0VkQF1OEnojJLh6` zDbmh-H0mT`-x075688NJqdlDyBqsik%lH2LvPb;#m^5G(!Yn2BjzbC`3-sEg_xxa{ zFwntb;qjnTn>5Y2lreBuLzv;N{5Bs7XkE&}Q4qHW7GNM?X&)e2hn3-`f5i4^Lp|ic zHXOgI=pzYkC88^mZhF_-mnDC{rq~a$!*xD6;IU*2Yk#u#?K}VR_~ur1eSqnayYfDQ z+{slgc@X>R62&oFolmADR`f%Fl>!mk^Ho~!RjE&tR7ZtzvgFM#|0YCdXBBnae~IgPyf90joC)S zldAale6s5sQyK~cdc$Jj^%Q%6A~jSbEQACXf%bm8TO4zSEd7raZ()jaC$xvB`eeGT zeDBhH+|L$fx*h$&(s$#S4)xy_efqf=FK+vFN$CQ9_a zH~*umg@1VZ$Od}Nf{#wmha>pV2D)kfM^_g9VT@Sig1ToJ8q4R#hR1xLH-9tJh@%gx z)ODob8-t^xCOFzpvA{G?K}A*z;`v4XAd~>{zGkUjij!TrZ!#M%Ug6#(I4smbFlhH2 z(aahsX}l^s>#u{(n_~ZM?vQeqCQmn9n@Ki_+XXsu(KA_jNv!w!VpcYt z1H4LlIo>x$u#6l^Tf=AMZy4K=s{hs1YM#q-*cr+~l5MZ*w6u?;G3avS`?L!xA`0Dg z@;cxRZSycqV{t46SVf5&frN49=r=BB+SFe;$fJPHV3 zBQquYeDj!%Q)6Boa6*VgJI8o7&_^I?P|jO9w>aXq=YZ2OqV*YYO7q_8oahq+`(Mg$ zd8PVY^zSAW@^WaTit$_vMP2)>Un_AP$ngQmek`9JH-0n!LnB;%a_tl$J>1~pz-$8v zCW(lp6WOSV)N`LFHc` zK}Ft9aBN^{4dTw&jByoI%=ZUbiGEw8r~la4H}L_jP31+<50WF_s}4?rNv8()}D@5bv_fJUCv%g=;Ua#?qSW zFI;}_ujXZg4(mi)C=t97+60X(ebC)y&@t50`17g?$se1QyJbs+*JtlmY?#~ydd-z{ zw8}V;^(`V_h`Q|NFhBj~$2jg!A7G~~%bw3DKJkPdUl#rRWlu}gOfG;@^lJYUAVJY7 zP%8!qGki0ot!k8Iy2(H6o+9lK3^?Ve@V{~5{CNXTJExq3YSX=Nq!$)pnD4ZUaWIbY z7_njq%h(Ag_*hO@p?18{veCRlcw`~613Nqx`hZab9tkY+K`aBxEV@Kx;myisd&*B2!XTWO$Opri7oTb|9m&5B&X;zXFCXdOc*9)up zx5!yWPdAV|f<~S}Zxbt0ZITeR5$zgOC_D#$s9I&M8d-6wt<>8a#=ltc~(Pr$%WytTcxR3I<-g zzzp~%Oaa{_9jJnV0QZng-vyE{#CUC~-vwC5+L8wU@i)YIbVhq2?cf%08Xd=-;bVXK z(-w&lHu6v3ZYJwS3j{mv=O?iC9ft8#irq?)WGd1y8MR4)ucS$av&9g_M%>&+x7@%g zejZeQ#xb}U&Ve}zPPR!;dZpQ%5A-pHhy5HIoQFNYPMenDe1GoM#oA*J`o|GEKB6>d74&bbHLWnTXehIu$teo4E5OXW!Yv9KQork^Z1?}!2#RMiAEjPYq3Bb8qebng@!FCBjrOypYXMf(vhEx@(UDvm^P_E_tB0syX0~ zWHC8*lp#0$gE@cSrnlQ|&+_*@%8Tl_xsiGD-Dg4{yR(1OY$H3lIV%oqTTYl@yoO@; zQ>2oLjG1&e?2zYC)uuO=Ps#xf{H@Ad-V&0_>tj=a4xoEhH$Ow2raT}yKktTTN61dz z4bcf@9J}a%e;QM-Dj~&wt?Ct%dSB0D;9vudXIs_1OpftX9+Nfi=$xZ-cJksuiQ$o_ ze&&GZieOtr`8dA_!#xp8+ctPTCLR9K^p~BR=Btp4AiRR1e2os2hx=Xif(|+I5PTJK z#;f=hTql-W<{Pcf_v`<;nC$0fbsTt!bl$}39H-dNC{S9AyopRk{M^7y zw;}OLEL^~t9!#}Y6?ej`gRv4qJGV3l7U0nCl1~q@*bFVhu?7Ol3E*6n%SBP$G@YU} zC{F^5^hBS`g?K6^K!Zn?2A+dvsw72|Y>T3MI_i9&&i0@a);%l{?hVtNXL}&beApQ~ zYF5uqn#|oO*$RA?tda3JXF&d`$h}c$gwqFQlmANAa)XoOPR~%LFg#kbm12`|#*NIS z^M%V~S~tz3kURx=WqNw3kFH73D8C}P6{NwSaK3PHws0-a&{b!X*#C0|W^C|y&O>c* zy7RY_{$?(y?XXK}3liCPBQl{C`xr^$r3lV~Do;1v8dwmX18L=yDH*D6MVZ*%31;(w z;56FiH`{Ta(@aDysL1)1xw@&tV#9)7X&JrQvw8Lhb)Lzd<&r-e_K-s_hmKYmD!_Gg z^XzT@H)fX0;~_K1-SCfPxH#d%VEM=LYdGJ-4IlsVSEe)AIJK-w>MkaQ+@=->ULPGZ znOdqS7HWHUQ;{nJa{cc}E2(yuJ0YyX-w(f9RIza1+YkQj&ip0cYgJ$Vo0S7M&OlY(hWR04k(Fv=u z2jqL4v4=ZKxMCLY{U`chU(_aXhN|J;xzC220QEd8WBEL10B-urZ>!7;U0$kefQ90h zuiw24lrimsM!_wQ{c|c@4)L%=_YkjET4dLf%25Zv+QA?7W;-6V?ACnxhx+#wMmw`9 zA>bU@$j#0;unSXeVrOzFHj^T0RAigtGvPYkWl6KT9ta{@)VFCA&&8t$Bun`lyfeWG zZdI2u6)sEpC&L=(B;I=87HFGH6n9G2d9HQlZW1^RthEptX&N~F8qOEr9$fz2IYywk z-aV2;b~`X9NQ=qPH&E;$iquh&Epv*(kS$tsDxhCcIcG@HqHBs)c`D$(B1iJMkD&?Y zz8otKurjhZ3RZ}~FqBtrjUy=yL4MU}wFRTU=hLIao#yIX{xXtoR z)HQjC7hE~8$7>;>{y0wiDCBO)X*4qzL|I{nr^WG+=?qG?qd~8yS&3&}M1Xc^ zFuB17Y{(cOFqT^~9tAgSrjN#ESeGxl$IAoD`#ny9Z*SB<+92> zdr0Uo<}j)mYzMH zC?_jq_(vv zwt(|vgpUWl=J#@f^TZ5&U!*zvl*6LN0-sd7An)6czSaHiNAZyUg&q(Ua_1iK*D5O^ zrqRk4NQXoh5~y_x065`sVuQ8RDj$WEFq#9?F9H?ofq{a`E8UH4AM@^0j+PjIUS z@Ax3E2L!#fNT^+ym+u<60is5=W2m^!TSaHn{9_ z!qJ4(X}g@vsmDi#S{&G4u|P-OE=UM(WidE*cpCP&V&rx}-lfpw2Hy9~0ait91HgK~ z?R1UDGT(z!&caG)W8i^+L8o>>i|lI}1Pg&nNUJQ3Xq4vANPEyIyFObBB=rMwO)`@k zxYZfk<;#HfJTCZ=SL#%)G8%LShX`h72ar0|9N)J5yKAZB^$nTV?IaRf;vL zMKKAkKBwwBZq;{uOwYR|Yh}odi-**{Up6!h)~o zw?ds27G7Z?l&w%TX7pGx6E=bpC-h8cPgI1NufMn~&K)>RY+)IBo{gT89MC90rp$ZKUR(_wjLsNF+(u*i?8k62Cli1BU;p2u>Bh;Z_ZPC|B*TFxBWPnB zHW`&t>`salP>~J5+WYA>Maz(UAXZNui8vB*efBM{1GBST+eN#;kLe(df^UnCr4!-9M|)J5F2>dOqL)B$m3NpYgIJin^xw7}>-t z^}ep3nxxRndwAv3fDEP?$jPt~MK50_(_rEtNs;2WU$E5APMF{{{4d0U7=Xmdj!f9> z{mD^t=fpzsYzB+p_gSB75eIoW5V(i~Nvl<|MmG%(M5*7_>Cm^{COt;tn3G=G@iv=V z{Zfo!7_5HwcrN5L432A!e_xUG(eI7U$-=^VFY*Pqk{HKLi;$HjjH~?=dygV_P?@xv zeyo;L>jy=2P3}4+a@r&*F8QJwT(>L*$i$Nrcjb>@zfuX8?2zp9j|;{VS{1Cj@mQO* zeEL1P#&9tbA}0D22X5zW2X&QZ&nnLy{x|tLWsjh4dX*AV7x?@pAh$|k>bx$9P}T<{ zdq}zp3MnA;qgCQ>!;5pFHV?8LRV<$TTC#-S2?~X{7V8SBlMXuF}LM5~A6|FhE^d=5L4S%6cV z_!?6!S|uywbwPDyY^bK+ZPRNayk;;g4BBW2Ks+oRBA@}SyAUMA9{$Wmm@N)&+Bphj4m6dw+)nNbknJ?S;* zl%$9gAC*_r_$aM1lUF$hJ2>I~$~kf3B$r(2IR*fWP($43(91!3_?m}43~P9>^c1u2 z`vZ_aDE%<`4`ovx{s(E`=7T$MzCmw- zhqDyhOpzuk5)_R=?4puJ>40^-{MR5?&QwipRd0gaRvdFkeko#^?4ENcI8-@4#s1y0 z^69PWo4kF_H)eEz@*Hv?4LI$VR`D}QA@5H39aqq8acX4>Ji!5TE&?gB*2yh%2AoR$ z;z^eK8QALe8qAB09$>o7PPOpG$rlobo&2%U#b-$$a%J zM@sYOBg<^=8T&3LtS!l2^gZ(;pO>mQ!$RSYp@ptZdTZ7uR-r5E*u{DB~uS_y|;s97M2KzitWA+RqdLL2TG zkZDII&pQ6Yxksn2o{??iVCTA!-?}0WtZy5@slv_-smc4i1 zocO55#PXL=Y!O9vpakSiNjHBxv>)DyxG4c`EzsI3qq|`(nkxR>V>{jb=E~Rccn^01$JebViK+&=Z`vX(o2d%Mom1k1S1JN;N^I?D!q>=oXhyZ^HX4s%kKr^l zj_Vq1J9_jD-s8oSpVJZ&!_Db&U?&(_?}p`2lPGpQMb^Kh9<8#;tD9V4Hoaye ze`b6D8iu7Oo1++mF&yFspgFCnu1-b(72N;bCnSp7bmYLR8m$T3;wd(kBCDy$PWMB6 z|6%V<;F?O${c(?Y4#|rl8^PoVh!SCfIIIA2I zJKfvcdwVmbzg^t|=rD|1uThrZevy_rr8PUO{Va+&TKf4?rlhP))I{%6;Jju$qO520bE z`o|UK)9f};_F?F$g&FaNy(qM8gnpACy6$gAmpWm28QC-xx?I?-z@#-OShAgBA$*%l z#iWu(VII^{VbKY8Sz^VLHjn8c(9J{}*`Raj2Vs!0R_;`-2||Y#x0MZR16|{@#3Rcla|>S8k)!c-@o?@dd~XK7oR& zt?IP_CH_ysVm-8-iX$^I=Tk1tW$=)x#No=^F0nSzZrGEPL4*YoDD`O#Ku<6`4|Es1(0Fwmxd>ogZM^_q)*w2rO8cAjth zeH|j*;<~QFm&BLX`I#pp!`!gzB%Pc0>$?BXKrUtQG*v*cKy#E$#dL-h(fJXEh-Lgd zrkt)e5;N*~hM*%eiBSu@`>!fa2SE(84H(n#x4kIilxP3eqPe^N$Cgq>u3r zsr62zZycd@u@l1FPxI>CO(ETqVfP%+yeoS(pEhMUxj zZ_m)y2=2M((ii=-_uR3$A5~AV9Y!cBZ>)oE- zHtTu>E|$fGlL9u%D%v78sTxIV14?Nfb2F@pMy9t%;Jy|{#zWCt5{ZL;x&}qQ5G9zC zBdg&^ry^gBs5$N1;e(4Nta!qLx#b|DlH}e&*GxF(jH(L9lvRuf9YbH7{8*AZ*=%^8 zuxFMNy$gqaLBDO#qPd7-p`~vdDi>$_+@^Q?tqnk-?TwQPph4v9yiK16!2xI+(%zNT zME81j3RcY07J2RsteBh{yo*^Oov$OeB1|1oTh*1Sm0rgoX45X3AM16(-6?F04l)}< zkK+jTa`M8*&Yk*}zjdL)XT-?ZC~N2v)Ouqd5mejClcS9MX^MUF&1@kxT1m!1gKIB{ z23;~Ln?Ei4(@(6m3|zK;#fCaTD*2S%CEnzJUi@q;MX|;`KDY(qB)Gp`gThNip8cW< zemuD*UQG(2kk8JiTN%nBrt-sMWXs zkTVjhU>Ej?ZBPa8n|%%#Yx9Lrsf*@jlTW_*9<;ln_li>$EoOCGwy3ru`0vT(% zH1s&rGvS1~UYerV;)UBK9w{D8KDY(q*fpWpW|AjF9~mz>&RfFXkYF)puV<#Xn{4sg z?u8^+^_pD~SE2j+IPacMGc<>k@QWkb=q8`W5e9)d;^?WFv14OIIO^Y-xGBo){jSfI zXOev`?EOLy(ID^l1jRycV>z@kf@VP{&j3s1WgNbz$0jgY*tG)wYWeBg3{yPK1ULJ;TKZr`c-E{h|-15i^i>yJemw%f>;R zbntksr`WX=Nv2}9cDwY z$fKMFz;#0xqSU*;k1!)c~vuhbIt;?ZXc{&NQVZZciUt)W!vao z^htVGP;;1pJ~(@O;6@e1`srkQPNxeev4WB#Na@+H3dfRc!=nN{>DK}XcEt%@0l+^}uOY%_4m z2!YNC+tj1)R%G+d5So_y-IHXg3qxqD1%%d7>>7$BQ89ZacPf&YYPvYGQ{F`Cy|)5k z3rrzpzzmJGFvAg}9?2MR0LDmuhV!+YfH5{f_Kn{%W?=m8RQE};*oA?SWdV#e6cjFz z1d!?qeJnA^Q|B~D&yhrrGX$CFH$~-yk2FZn`O(7#cA{c9w;s_o+)#1Me^U3i&8YZ& z+wWV+2cMZ-(nl5;x=OK^DRL3g(TW5Xg&ueCG8q)wfr_arI$NDCOw*j2Wh7NB2uHmy zqcWr*fz_UxRRYvMl~aw!(llq-t)$hj)epH`7U0)K!VED~c4GTdc@T00mC}cS{A!U9 z`mV2%(oo=`ht|zw{Csf*bB1*CT0K7U)Z&HZAYrLj>L-0P^BAPQAA82l$qrah2Z^pu z`A!v{pjWPdzbJQFjq+v1qIQCDU`BG_=ZmqL{*OiD|6+#M_immiq|=3sPO=3V+TyWn z>{#-IiYbb`KfW3m7Ezk`@tk%j;JmKL3D4%`^3oOc(g(6typw_x1dqUKSq{WOq6=&l zsbH=KS1^Ur+y3zi+*7Y9r%OdGKD+1+y4W`@;IX(~iF@iatq{K}r@QG3niHfMXkpfj zuYfuUJaP$+teTS_h^0b!U(?n2OY1!MNN~><$(r#;g36|pib_ST;C7uLJrEge7wMQ1 z4}+qLU`u}v6cbgFay6EdK9=Cs8R9d-g+W+(aYK^FtofmKPxSTZn$S|FT)jMSPhc;v z1CnEP5#{t}gT{(F`Lpjlwr-4b*?6;!RyU0EHAq`$VOjyW&3391ChP4&46Tcnbdu+=~vtrTtjrGOdjqhEw9dB6-zUS{MxOCKBfCh->A+IY= z9C~FM8y9}wS9?QJG@+A+`N>RBB`g))q_+hX1uS&O3l1pb{7@DbDsXVF3D^1JalQI* zV2KBoPC!zkO@44TF2DNSYGiG&Si|4cE0^(~%z)+u`IWilbYcKrsO`McKD!Y=FQ+Nl3Ui-Rqa zEsEu`>!I4x(Ct(1c!3DU2$PC0d8R`iwi5tDUhB-AwgE6iC&bv+ujh@kF7$L+P?T9% znWv0c9ehJXvez2-<%&$nUdegi_~1GX@MF!pMKUFaqH=gkXD(!PzPLwD~8zr7C@@;fswB`9PGXTRGLpw3^VmmOLAAZR-oG?S>euu4_VVO>GD)r+cJXQ(6?u#$6I$QXOFo^i2_VV&yUIVF_NB z#CLt#MOEU5-j^b_^EOBwh9rl;XArnvf(pA2bA!*l&TDU5n;aW&c7duW0;~Ell?@^% z$JOW5k0r<;xf0Yfu;!#+6+5TdNCkG!{fKhUWKCBD;J{1-T7UbPeWj^TJL5 z@C_o{Xv6SJ3~W9U2MvwQ`w#a3r=f9O2U8KJ%9?04If1_^d6#V9mWXiO|7T!~2G@3M zqgW{R*i6N2oKQ9eDd5Xy8EZR^@q0nL=c3=viI6o&QQ*-|P&gYwv9{E}|-= z7iPDBn;Z+RD@ef`7g9^#hXeqspfrOTckVPS-FifS0t^6pl60)wDNP_JngJOHMVa(b_N>CASA&Y2@d!yqTdk;vdQTT*P`54X=ZW2boXIXP)_ zVeBliNMAmr*e;6Pr(%qGY`R_63qEh|G^E|smPEAEH|T?t@6#=>*U}~YyRu5vV`y8^ z7Di(iWfIZuf~8WvxLL4*RM9IY#`(dqtc}uPAQGPV!vCx5{Xax z-AdqZzAO7|q%6n0{foZ$e{V)!{BqrF(mqtY(1lCWS6ZO)3B~qOy2r=L7j)1wWfM?J0~bh+4uwtkSD< z)t98Wf4}thjALLQO{MMm!hGQo23Uq6Dc+#yiY$fl=M=@jZ>A`!=yuUb!5ZNOS-s|h zEKb(y*egzRZba7MyJ22Ne)EW-9jKkeF&)v4I;Wt0Z2(b<& zKCg=4!U<3t;l*Onezmz^r%Y5N0_8D~4J(N_&Hul|#(+O(8x#iv;^$e{QP?O~fL3Z; z(botz2Q)}CDD4q-2Pq3aASq%}{O%eg}3#)hTO5$g5^*=dBB?m9+qfLDvo48&Gib z1nm@&7IRq5z+ZDA+~9B8eUEPTGA`G1>9qlg0cR&vOidRi@$tkXGqX_`E{Sg{ZtqkS zMckD&DGGU~X3a-YSya<(kS6(OD*P9VpgZMK6G@5Ki*FU+^5+u~BP zE`6n zcviWzPmbk0kRO$&1%bmX>Ty~@$B2(iWsZ@%h9I+)b1*Az+zg8gIQ4_2*EJu1%ig-tT6BKJyLCL8v%g<6WS_ zn;=DUV!$yzs(I9cgu_-f{uuaNTh%Q<(bURo}NwhRp4bcvJ;baeEdta z9Z@S4hXWr&m>_pr+JtiibnJ2`#mR>#S=ZN%&fo8D%ZpX1ivV|Mi zT-eR2umD>z#qOlYb`(8p6qU@qDN0eC2IjSGbjEjgec?fLo#sUJ^8cy&Y1`N0=577A zV_zwF)A04!FZ6zU)tf8dyr6-0K;|q{$Xov!67cJl9X>b0OGU_kQ7Xzertpt^ZT}Ys zf-uiLH{|3Ok6HU*%HEd;eqGr6wGsGjkS4h|h_p*yiwn_K zi$Gxz8>&n-#m+YcmLp>~P%JNV`ZwD^d31^TsfXEQ{7rlDPbA)jO-7D|$=E=#DHK^v z#WY82cg;J;w#iDx4}iS!`h?3q^_nWC9|{6b%G)CJ>J#&H%xFTvZWy2&>GZ*W_5-$I zFw{EwitmOK*3jUxj&|V`k_{lWFeu4>l-I&6@x3O=7w6Kzixmzr*bFhYGvLnnQLR>Y zRUm!u7hk&asxwYtzymW@X%jZ9c27K>Gz$o|<*ihewM4WHk#cbuQU(z$y$8H}+) z-y5q$=DB2f`i)++@pTZ!Sg_9*DR1hO@xqspJve4G?M2hzAckXp`NemM_><=^>`XFaf3ktjQzlH< z9fDD$2NEd#mYPasNB{k_`+>gr|G&eJa(X>o_w71(zNuPbwj1()eW8vdaol56Ow4Dw|w5+tOlTW-d}JP)1i%G2OrgwUD~0sMjnRx0_ig-0E4v?~YtL zZoPQPxDWZ2!I$KfUd5WT-&;Cv5x)vd(S7F3cklU~2UYE~pu^D*f%|!H;0_)Nm!A3V zv-9DoqzicOZ%FnAUX0kmTgo>;yFoX-IOsq~A-~zHo7us`!jyIzi^$Qy(-X{?yQ?k^@__BicP1;1}dh5-tE(_NRF!4WO;T#C(0I} z`Z%mw;-?GkRVqN#(XIa@_?V*>3@-Jc@;Y>=vvPFwodpggh_XXq0}_E7yvVAZA057N|3$PBCZL z{;=hWbVZ5+TgJ6m5wIkVe-P^@5fTFKXo|91PH=8Ny)jH8_1Q`ddXK7R0%8(%UTkcVIX%1=n8 z3(qh&EKEWz#h$0g2ULtMv|{dEV1|MkwL=lWF%^#WVtJC&e60bxq4!LMR;sWwr~)(si-AQ7E7j_N1MMNXJsp^+ z>VcGZa{v~>t@6gxo{%K|zzu6;Se)OYxE6>P914P8-L2ZKN}M)C;h9Z9+J|-vdohA9 zHvf$04sb)e>hk(~ulhpUBsL1{42wfD#V3`Se5@$xEhlhLKM(!ejNh53sNdK8bQ<~Cg{P<$7LCu3DfUx}bYsQZa#<3;L8_m0 zE)e;RI($~j=BFq!-FGu48Y9$Ux+^P|_eik=zBmF?h4V{Qz<-!Q*LWxK4RnpHURfI2 zAxo21yYE)rgr=|XjZ%XE{oqXgR=O&voZbXA)F_C6E}Je?2jRXIn%yBweAfh3yX)1U zR0=3&-YG~Lgl|4WLd-p|KiAjOYlUX?lt{Uz%m1^0js=Co%f|chmXwyD<&34 z?(a*PBc>3$^N8}@8A^!frW8ll2o?4)4 z2HJ(LnY(9SfHDVQjhJMT_exQiYUYyo|G;dNHqi@-pK(S+3Bopk7BB6SpCY$rZUnYZ z_=$A(`EhB#+tuLpujP|= zk#PmPt zUpPwh5Vtu9omp;k$#Bo+GzYnE_V_>lx#x3J&DN>=ZP`+?*@dlBxyAZzH^uItNFi|4 z0RPZI`XhN2%H66iGiOCl_>Y(ZVXqnd6a=zb z0dWk7Yn0O~NS5a(%5r%Hlfx^Uf{egU>Wt)$=Ao8n!^N!a>o}pA+V|Fk3)VdV&(nLj zF6fDkz5q;_=Rq4I_K;~e2kZdqNa$hQE5}a!10lvQNF14B%&q^Q{S*G`Pw)IO<|n^? zC+4p|`OWWt9V4MxH$qWl>^I)GUX0mT^HjTUk#q+oyr#Y9j`U(MK_gG$GT|MCHg)1U zzY@b&4PVVSA&m2 zv1SInc+PoWT*=1rmj{mqp14nsql2e0iOsPe$jlHq{nj_Gk<~7|j)9)_L6!cSDK>*5 z8>yJO*B{bbX1}|aqz80HUYWauEtr@Skr#16Ruq5*ok%N)F^)~%7od`=En>fPPxvNp zq$YI!$PdeojDjWMWJjn+ZsTwN)(nyZ?|8|`9d35Sg;zAoEdqCsC>B^b9+*0vk?VBd zY*fc705xr}6-J>QEo$K)b4$6PithC)4^9iZ4~Zw8xI8#-YK!8e8oLOw{n zSqc^qcx20iOW`&y8@b!_iy5T~Gy2vi2VW-ZxS`K=oqDpz!l30)EbuyJQZYzAw*GUM zMNmbYqS)mFi|Q*Zw$|4(7iJr^nHCB2B#`IM7vJ;+7DJyFyg_zz}1pL zm-su&ybaT+J#&_vb*bz?SiHm zI^|ExPe*mo?`@>s!&kLl6UWExPotvN6&NFaf3wStayV4Z~f4wq? zS1vdycOV!>JpRv+;7~7z+oX1}V){?4X&Fmb6*;I&Myl@;2J}lAu^XsZ=8@Yy*Q$cYq2-ycXrq);rvV7F5rZcF+ZtT z^3iw9lTZBbjyxb|U$KhuI~I1Pkz%h<0@~O3*8njDy;rhV zyoFap=LFQkuO5oJ&)*Am%Jv$sVPnq1B60X)C-iJ%?tgxN`Qz8k_&PZ4uggfmIATLZ z7#RV}rMZl0jf+)J)$XY7*d(ff_05y82VvS?o}Sg>c{;J_EQ8<)R6i`2VbD#l)ItBp z36&1|cRRsFFhG)cqu&XjG(9*Q6VJCK+FYrQIp?Pbb~$^W_)8vu;|Js=KWOJ++XFI0 zwYTdoU*1_lSAN}$p5h6Q-yv16m~{fi=Rsc3C5o-1NDURU3zXijDKq_(-gvfErzsUZ zB(0ib-`Mu06L0N{F8|(!&#(FJ?dS_&HJ;4Oc_ZnK995?Orl6!ZplZ*aBJK^*@Y*tSu2z5&M5Y`+G-&ud)bZbnws=gR*{;#@c` z1WizbP_&j}lPRJD1_uU1bv3 z%y3gAE=5rz%L&)3n-oc}t$0nZY~^A25c_eBGEKc+xOU4{f9SKFw2-Z8MBqN(s^A`J zPI#>-lR3)6;?RD#8$O$Sa>9|U;ymvdXq9c9n@uNRurCr_!+tkZp>u@ca~kNlVeGlH zC+S*prA=_NoG@Y;j#sP}mOOf5cmM7RZtI!{dGqdS>%^@aB+~|wjCSn07 zB96!cUEC~joo3O`*M9ecX0sH}uLo7;9Xvdz8x)@KYH0r^OFc6~khl2+T}m~;Xt1NU3&Pc%B zDh0_BaEa~$v3Q%hYzk7>lm#aSoRjPfcPKDkFn*jd^MZRh@{n!J9DARVTW26#HdJUM zr(l3`ZX9(JebH9+KHoZJ3a`ZYKXgxuBddHtwryIv2kI2&gS-iL;U!1imFcC8l6r3I z9qz-z85B<)7w*5sGIh7A&%cXZ4FP9Pa`rDOAC>!tN)guB(o zLcCLw13_fy1D=$^TksTL9zY`qL^Y7MbS7=ub0f?NM9*FAydCu!AJlHSxZ0p1u;Wnw_D~}vg|VxvCFmq(0YnpOOa$M=EP)! zpdPf^v94zC(>4+8-M%M)i5PoeQ^t?xoNFJB%xk2P0E6STuTOk`mo+bz4MHg- ze#v91K;#bJ%?bu7b{~IzLDHLgS)Xya_5au7B;M z3Ne{W-=rUUX|F<_CyluzS`wT?cg#E-_=(fhUx&gAC(I6Y11Ig{gxRrg^UG-KC9=yx z!^WZ)nM9tF8QqoDDRK5F@rRD4CT4|Wqwp^uGy7rj@_RY*3~sR4O?WNi&3?zNUpBr^ zmb=Ka)@HVquxWs^uKvsQmr+y41zhz7{ zl(tpN%K2Tqf(VqOzYJ;N9Epx;^j)vHCq;jyPP3WX^MhMoy87+PZ{L3V6DXp?E2`*| z^b_B7MJm}NX#+h^-JD0jE9F4<+A-i6fRf`t#0fl9-j7JhTr+s;{yi?69CTsuT(SVq zDT;;CgQHXowxof*;I~2T5yr@F&p>~rMidvWy%=hcn)Rmh#m5Mqfors$$R+&45LH7P zgx@i$)*waB2@rl&fUsl#=hg%r36^IqqtXH%Ma)ZC$uPHg^RqPf_i+u zFb9al``pgGZW#3vTqsNgQBv}&-~L|v=XZYdlRv#DSxT{sDH1#Ar8jO4bF_wk87G@T zefjm@J+UrVa@p0fjnbsF3HfukOIkgf>GXhQ<8bTh^elB@Wc;{Y^1Xpov}s?<%u0TK zVEi}-x*#tcBs=l(3@U<>#{afR5)ufwdHhT}_chSSB;EA@_Xr zy#3P+(nX}3zAi~%mjzdcUYT1Z+2k`?co-N#yKyma!6)UOLu zO0Bk_f{7c;E*wkxsF=K|4S^L5b}26t7DUu*wtV#($hh^35_!#nv&0}>%lmX@7F`5u z=yE~3U`yZ`CSOif6QGkOricx5;NAv(c|!gZqns;CU6hlWEo9@3)04f!=_2 z8L~p)(SyL(+$`O%J_)~=s;<*?)A(!9Yt5Sej$1Rcc*~=*cmsb&hH~xf4B8NZe@mO@ zfg10bKetp+ueljk1%@?~b=eY@^sSFyn_z}v!|Rv+K-P|f{)xdBVJpREQAAJ0oFr}X zdy_86-Y*5k_zZEM93DDbE$>*@g6Jo%a8F+$`=zIQ?Ehc=pFJd&-Lss zr!zzi(Dk3ZD zxgqHFObxy!F7eo<$Ye`EIOUpAP4rq+4_W1%tv(1fPuIjX;#^f0&vbpIq*`!4FsyFrR%ME!2XqN~gzVTO2-5X9B1rL7)yp(Xqh50q6|B>T{7Ib;e@ zZqUejoyx9M=%qWS^}8XT9n$eN10{VAIix-{3sR%&#ZLm)@h^Jkz-{+SkNIQ!|I+Yk zIZ^!4( za&UlvUe6jq@#Nyj1U451Zb2#svXR&skBw^h#lZmKh=L*CcGFJBjQ+q~I89b#zVgE# z-<~JD(3@x7ADNkDhqkY!G%pL52Lgv>2%=Mdl*$t zYm4b~-X)>U&}`7}hP(Gb-EzJVx8~zaRIbJXFjEi82$1MF)C~p50Ss~ftSq?r+gF{2 zA(uWdtN(LA9u;_Sc7wD9B%ILw*g35o{EP#wemZ@+U>K_l~|UAvXgiKg(v4D z4~)6Zq?yJV3DlOGUorWKZ!^6Th}CLj#gV9^Y_IcS*9bTo40ipf9Vc-zL1R1rF*nn? zd4P*Q*S1>3th%G0maJv+g+V#qM zO$9$kjo(Y@EnZMz3`~jdzBd9p@{&;;A+la__u(H7qL$6JF5u&W6&Ln>ZIt_+lw=E6 z@z(OtF*+i+EvYuV?+G(kFK#m10n$#gJ*&_tcP!p3wBA=7`*XPIwe4Rhd*lah;G|DZ6B2d%xbl;cOh_(lWNb9$?eu6< z8(U<%UpM&Q^r;rafv0*+c%OeL0vA>gxycuPLV7*HIlf4i%O1~Zhy>n&d+w$aiv@;= z1Qt8BuLU}5N*Xq@2CRvbkX5x>Imx={#zt)mHhf^r9a*{_5`92jXdxe!k5a}%$v^gz zj%sHzrw8h!hd0!1IlZc`yZStou{0vUJl7QV|NB0YITS^&>;C<~S2%d}ypv+LQzYNW zK_UQsYM{Y{8uD6^HZ5o?1mQqzxIAhXQx5bbXfOKR46sygkYBjt~K?el+rz<#yg3h<6bu9 zfuxL7ivU@rs?F>HNrU?;?;6>)z`RKZl|`ZqI^R1@dRK-^Y3$a%CfUG#2xKu>Ik;di zX#x536vYKWw{ho!dL_=k3otae%V$%-mT)W}nV`{ARXkE69*oIJUwmRD<4yb}s6kEKIez|NG(B<{ z;m|m!J&%UV>t@gW9CZ$-=kL1yqR)k|J+Q9Kb6Kllqh#>0cRfi6&jh)xBz|*vi@eu! zvG7`8GLV4fscTgapi;1M>Sa-E$OocAsH8a-*bN%S52Lc7f%1gAqutl%;Kc>MPmKs{ zJa;1p@8E=ADy=wsinVeHm-wygz5v9ArpXe1n)DQOztxKROGT4+-+{b9-w?MMdrF9oo-*OAF{pUY<-=xzjLxT zW4==P$2VB>!pZYd_Z709+rr6(bBiDyFsQsCmtw&W%QAKNZzdl^)oXTm-SaB(NSxLk zc~<#=ZULQudJP`I{LvOKEND0lDkbH?C3Eqd1h(3HgIF1~$c*KPi3%%_*D5AH{vxCljvl<#gP6=BS)$-oS?O`1=>oqBgl<;an zlj0I71M2K`o_i#y=Yev1i}+c*p2*B8AItCgEuFa&q93N?PJ7rRc+SR0@2rQ->HT%x z*YEmIL;n(DhRdyQ-Puj@xEU4~c4VNVV31)sNU{4UQVgzzVrFiXhniQ(0AkJZs zwj0SrFwlXm2b~I(E#5P^hd`Gmd_#R=P*+m!nT&yzRo=K?uf#L9Ksvirkia4x2{P3b zd+(6z)du-Zg(KgGk-h>JCdh&BgD3OrG^rUoqWhxXlBA(>1FnmG0;QUR5Ryr;n<$b- z#ps1uo(0TiX$O51EaM{Trs8Py1Nj;H_6(Dt==piqB}LIU73X1Z$=n_CxS%bgLy|Zg zAUJTmZQjbZElkSp|FU(d!t=QR7fv+WC|@Y@+#6UcN)6WL`NV}6sItgv{=V7gBuid{ zT^+JC>3&`zC|PCDd4Vg)X)qskp%+7q(&rau7e%y1Bu#W`j&wLUd1@qVj`GM4*v82p z{t=>j^%*l8r1xeTm%_y&>;-`kVHMM?{0K7Rqnj9lk+L5GBOg{Scd!kCA;Pp{e%YqD z{(JKb6Te(Ho3wvsypxp{?b=T$wwEFgshBe?)?^<0|rOEndVp1C=C3Or^11F2Kc=#vXB3q{# zp#MU{gD7KE$8P#+Y@TNVHJ5cy(G9z zzFF$BecFGl%lzUu&G>usXy=c}X>Ry)Vc+MLh3|8PVu5L|j*4lJEe_dl_Ik>}zp2+? zkZm_}jVV^`gErd~#U`J`&@NuQ_sy`ph}}$^1|vN6nr!t&5$<2cFZN0HDh)_cqzB*? zYe|0K4S_*$1{!x^T^Wx1^vXRxYeI{eA`z0PZy*aNw9~u959F)(y)ac4M)vY_bFM=h ze;tD&7r3npZ59mtawZ?U+d25x_M5_|bKnRz>^Fw*zrD~#Uc(epVu(;HIka2UO-d!!dVqK5rvMEg) zwO6MEZC>RtIq0A;1+Xx!F|^GKO7D;~D|ebse2Zl9o{Y{K+&oI0pke>Wj`&PY&=|XP z*R_X~c?x>s_2eAz{%Z zdO46dPHK?$!otaPDw0H?Fq;-PPC`;U2ij(F$2J;Ihz&Uc^?CSxD`>g3KqYak@Cjds zS#69M<3&1NXV^!tr?HsTHh?ulm7wEHdzf~QdxHYMQ~{||si0S?t&>&D9|%5~c@JnG zu<RysQCwFV|)+ztck<|6J%#9sp%fd-9eX(^n$L+Umwj^VhEc&T=&RVi*(G+r>+p@`p*W^zu zX1Gr%_G60Nrebgvyhn1+XHV3A2((<6+%syg^zw>CYXde2i$R38*1y>+f$g4=YHY2= zU`*~L;E4k#_!QBJv9m5CXp`!kdv$2N2KTHLZ;;~NcTF0|@0yGhhGGV;TJ4R;``p^p zUD6Vd)ZkLkj9bBeqN>*%4?aa!gSaxDJ&=6Yd7mLz)6)z@Hwmoi7jx9NXY8MlJG)+! z$L#h2HVf$)k`rDD`Q&`w**At-zYhsg#&JocaSLO-!qq~%gO>cVk98~U}_jhmua1ZAFwp>iH&epWiUI?Kig+5;dPmpXO# zHwstmu`b%;0tpv(h;7uK9hY8ZKrR86LZ}#q9oaRYF{oE#^$pesn~tX_+TH}kTKPS* z%<5g)HuxUTKMloy1J}iSILX>@NupZY8Pj)XVdi(AzCB%Hc1Yy^`a&H^5ro{jTenIWFBZ2pf(b5dD7Q1b+EqbdO+lrj#UknVR z9f*2<5`jGf*E{QTS_VH^X!bMyP(1#xWED3*!-d__d<#EgBgLj-cp)Zsrm+Bv2h9$w zwlfmzys?=HYdd#An+X;*9goab<4L39Ul>Q0Gd!Gy2dn#5C#L`BtKKWy4z=9%noFu$ zAn0x7EdukPSGJ0h`K6+hQK&Sl3+EqW z9R8KlOhfHqCw5OXTam!ul)OteaI+$=8}}p7Q8H+?vW;T1DYBW0Iq$oTNelXicF}iN zvGtl&6Z#alWG#w4@*YXAB8|lZSw7w1GT2Wq8ifsRR^;2|Uw+HFFZwg$uY%CiAS<$lVv{J60J)?Y`NE7j5M?i$ zg&Gn0;!==)=!`@HAt#!Kp$Fbx3=Ms?^Y&iLo8d9WJh!N~?Efoy;>HbDu3PQ|3B^IE zNTApi6j?^aRDr_+^aljyiFphDzIpVaU>r<4@Zh{bf5Dyie{N5pbs*hkYaMNb#(TXC zQeXy?#03^XHBkw_I}$39ActJYBzpINn|fRLAzc!h3SPtsf+7+ZLvKV@(aE8gMVMUA z6sAtuMrVc;M0JzfGn{59>;S-!OV)v%FBroS9pd)(PxxBrFUOnB$E{yBzE75Ob4^`% zftzb#K2j+bUV}BD8UdY(NL8sVP!&a>mi_$81jsLUN5{@7@kk5$XY8C7#k=QuHv-=s zt@Z6NkTfEM*n6?({PqLZz9<)m!-bbEHj1oo1VU*Vusp;HyCcs{&*wEt3L{HAT3%nr zYxOMR=|U5|_fL14K^baTxIkm5>z%c4?gy)n`VV*=3CqjKCT?cKg=43OEzHJtiiJQ~ zE)|m#u2*OBF&qXoO~}1uV(N$14B%2|H)Ut#oiJT#H?szRpv;3xk!8X=3Mgyk<-z)P z>C9|8PyNUfvvkg*%;SQWfr-fK(PSg)-8K=F-Ka zO+BjC;ptG>G3-yz9nB5fzQ~OFul8EAt-Gw7YJ&zIE8|q-Y7{`{{d-<_6Anb zC*&Up2X+E~8ujFhXTIJ&@4*)ee%2NJ$4=-V>44nEd2x$8b;=Uqx+!fUC%(Y;zpPfX zv+=9*USMt`;P2bYg?l5P`*`yO;(R$Vq_eK1>%}PTvWsmZvK^rfQLQ2%RzfL? zeZD5~OfB}8qcAi6Y>Fbz^R{4n#IPbK_M&Dei)jBo?&z{@6C?fp^25nyJUwC0EGK$y zt6&!nkQ}iLI*%gz^rs$Pfo`11Q%+z_Y)IO(%5or^Y*ZCHD}Z9h_`hThap0g2x|Qw8 zWr+U2F>CXEerFvuu@RiWf;|&GeJTBT&Q1^Qfb#Kqh-%mKM%gE{|M3rcHlwh~>G8Yn zsvu<2j8A3edFAw5-?&Cre`Xw!ofh-TW{L%7t&LR7C2&OkbyHDnBxrj)r`>NMvl42| z@|Yvh3+~QKWbTAqo`pgRJLJU?Cz%uSp2#Y(=@h3U-+ruHS{!G?{V!>i|7OiZ>$0IW z8?3V{p!Oq|K?SdRP0P&w!0jGu)rX^MRoi(Rq&d7tqI%6`Q9G0=H}UVukp;hAb4+rK zpQY}Otd_OTa`w;Rmmo+OknjWyO{l*tjs|TffV5Ci6>`_SPni^pG-RhV|(IZ&I zJa*2Cb%>GSJm5pz!x_6cS(CAGUyOM8sxNR15YfOUm@U9*o}xI2G%tb!At#kaK0Yz9 z6C=45QX4`4Hi$vz3jslO$`)mdqAg4Zm`R;_OV$PVz&eGRQ0#H=O|$xF;vH zjM*ELIbC6PWY;GLUnc9gIkK+nqLV!qj%*Ia!on?+idic@I2+ln9?G>y2t8nK5N@%4 z>57$#{W5Khpm=g-_;Lj<;XY8-gldndJIFHV(5aTLomd&{OeX!2!OBUPg4@cx`#<+% zGf;xEnz~8-IG4>vk#38jzeTa2FxO1Q6sQWNScwRt@q_+z8l<-o6!+YbkVdaWR6tw> zPcHGm;zg`4D2_a#(ured8U&Xo6h?OnR?N~XQJ%#BEfzN?7<>0H0DuKZ+EQLGy>)Jd z8q1WDBdh6N&rLp1!-Rxvn|x3QCR<%iV~G;%%z<5{JS1bov**uGa&HiI%JYR7((jPv z(wJSS2L2|!a*g*xS-x1uukkkhZUP(snqI17=AWAvKag=68F;J>lq0^I0LjfJ#qUt| ziOf)2o_^yOB;ggy8bRoOP-u1&#imhYJrz@@tWzEef~g8wjG6`6^ugIjdBvjQ$Ob96 z!Xd4SyB^veX(#Y?;g?~hbcXT%4+As9_!aJ#*v8BmP02TYV8+MKcKy7U)Nq@#TzI|o zi3M_+DE1mf8i2)Ih$7qC^~^%yjX;AxDswLiGWhqaVzF6AuUXEMU{)9Av z8pKgwJv0Ou=xhWlsym<*bndjr-XNa|e+jCKGz*?tIlQ+4?3dZbb@H-6IXo}0*{kMd z`t-vB^49SQ8cIJc@-vwk=f5-}>B8oGD-U_dYk|vcxK9^fb$+9SOW3K!rb%Wb z?#<$z55{=OiOJ{u{?X@lS=tL_1j$jAQyWDMQs)KQMvRpgjQ2{6J?HD z-HaGLR;GR=h#7qCX9PXbotLLrGgrH8AkYR&busi6nsuA2MQcKv>1V}KrJn>a7B@&##5JF>hJ=kL-$qxejtY-L@Btzfc|y}bIBt5d{~_4* zxA3&F!gdiB)wa>v3P^UJ^>P}c^bj!ONc%eE9h|fCQ+a%=bqk2g&L8_M#`^A2WSi!vLL^||s)vTvxUt?RDa zL5gy)sW?HgM=4TH#hjbaG4onrdDLUUI>=5pg)QV?i|QeVReB(w$H;z;`!0~IY*Oq8 zFf|K4`RaAaKe1l>7_Uv9Lql;4tjD&qyX5E9dbM7upR_Bu7glF^4Bi%xbVa7B|GD6w zIlSc_dL{nG{yDd2toS;p*?ljE*9klvrK$&EBk*=c0$0zVgznpnX{i+8=AU9~vp5y>G$}wI+)M@$0sy2BGNc!xJ_z)Jr zx8=IfPI*OCyP|A@E;t9u;6^Y-*gt~9fr$NgAO4w~AVS3+ikK2+_EfS0wyq;ZE?mm+ zfrVF6O0htfa)634W#!VCB&M3~lOJ$5MC6N&-3O3`tDOxhf_cp1kPc9|s9?%K0}IJO zv8thkY4p%SO3EMg*PH1Q4-6dZm4{W^NsFS8xBj(w5A8OuT1Yh9ib!CAV*{eZeQu`z zIs#$C0ni{@GxX;0ar5*_P5>Iq>&@BwzpMHmg~ig?YFV|QII=}y0!^#vyl<-}b9x>R zw~%w8#A9F7>Nj9*W&WNK9i*W*h6kyk@8!t79sBOAvStr;S!cxtALwqVTgs(t zMW^Mt^akmjFo=#U<8M;sgm)@NVPq{w!`^YWyvq6CIGK(yKMsxfk#z_2E5*xXqfdFC zZ;PVd_kN@{G2pQ~s=k_bKw~>@Bu(3`Xbc4{ovY;Fgj^}Ei623}dv{O~D9BcaLX)CS zTsEZX|^EVIY;C^GA5 zQxs*v2P7LOBuv*U``vV6EH_9DNMP@RB2tNeN7QI6L5K^?pgqqN#JH6SEf`4nCom837uGJ%N@r9^UJeOYYaN#J0~2E`D5Mp#k6^` z-2J_PGi1?cCO4RAv9w-Iv56E}1-U_{DJ(^S6;fSO^~$cuSm83^#_&-ye}jkEJ_rrI zVsy6eyq>guju{kzi&qwq11=mvxo9!WCn*-{cq^%x8+4`w2U)M&9dwo1Gr3ctg&+vQ zNiZ{f2d`0t-?oM4(rW`gR$-W>Tpq_lpCZtzh+7m{Q>sC9xZiEXthGR7hXoQv6BfF+ z!)mD;D4gm-8$yBlsYYgy=7Ao%-V3MFK6x&4OQ!cK2g-Exf3fswB&f0v9M5bC$E{%_ ziIaU9J5`eZwOM8;WhnkTpX`0b(uN;e!0R~0R#K#lidiJ>lIq0EWvNqh0xBS%Q^>3L zJ?FQMZxG>oJZudcR>li$W9x%| zuRCiUlXF?O#sM$LFT+%P(^R_I{}vDe#lwDEr}M1Hk#>v)kPXH1nHvC$C5I# zeL}H%ejA9jIA5M+*nk+>c)#>&+khCtDIZfi@r$0;WVg>pUtBnWY=Z#T1e|1;w6xXo zViCqwjD4&rGmNRmY~d;%CI?1n6ka+Q&raY7muD9`dCN9jLf(w_;hX27X{p~mNtU{> zOS9Et&RIvXYbcUL#hjL3g?{GSGa!j`6&e7rgB7E1HL_K6+JOogls85X0WgqUK;YzY zx46A+5DZ~6#{QuD^zzS}fw5%GcmAD}xiByqEr4;JVj+Eaii&A@JzltCMzPo+$e`1q zDFpfeu^YJ)I24lj=v^4)Xw!8m`EA`52kb7%_tE{UMI$*+$TlIlrp; zj@kai|L({Ga@K|I&m9Y7G*avpid>>%ws=*!KO%LSJ0a=et(p!0k@tPW=O6x8^Z#7X z)N2+qdjq$4oh6S*pW7C%I&ahGCFT*d(55KryfLh&V|3yZlP`<5c2c@ z&0J&fo~Eny%1$|+VK4SZj3rB(^YU+YV#+p_hM3=|Cp&Mv>ubi>+vmqzAZeeO03UQt z4l4f7r&v%Z+e*bW3ot6IS04^6r_YdDf4w?ZsNF{IksO?z4Ky6sQ@M9?k?148$C4{^ z4bomu)Zi|bV|RR4WNB!%d?&L6vD*{Pe?F+sUqR zL+r z=GAL<@Or=6_`L=H@4ny+TSYfS9EKX~hKMYcy>IZ)&tZfQ|3&Zd&_8JBUT*KT%O;?r zW-NU7Rp+m(5_iyN(M~AVzou7inOZR!_%oo|-=qV9N1#$uag}-Kwc76K#7-30K8-7G zZQseUN4VkE>j!bCU-gAJHtE;WJs@)_Jtf45zm7=^08NMekPv7$hVVvap*W5Y7hrI_ z^@vXA1dOr&@Vz$9(>(Va{Lg%LgN4ta%Cuf(HWGqDKInI4DbKypEhalW&>+-cuzBj(Mpg$>quYL|`w z+sHgONGts@H`u8#&JS_>Zc!Fbz(v>JEK#7df|J9=;X*w#TMSd*`tpjJHx-W2HJW1Kfki(@pSV% zGOsCb70Dk*PFk4eLlg^)O?#jegjcWW49k?H@LR~bDOc!rQI0Uv1Gl!(>!!S0H)(fx z8e68W*DRa#5&fZ}#G_BX8}hOSdK>gw6iPG0R}1%f+z_<@(^m=G$%9Ke<)_Gplac5P zzN7-NdM1+)tP9;8;iv(!^TmD9BE|BUg9d`ri|e|v&-bVO>^HA^?+`X~>?K7&fq?&B zX7@u_esM^qxPn>fVJf35iYNjyI{a6!JSp#vL^Ve|_Y=uUd7I`k>G7=BAaQQ)G`ymm z#7CB1R(c#E;5Mc;6WdC@}p%d2hD_o`|7xr&$u*lTN z04Ox}yyw!(y{Z@_k93r^<7Axd0T8DfI_Tuv?t+AGnJ1(JKMH)CoN?g@|80wzvVmgj zDRO~|Stx9Q;{7XgwM)PO>GeG4r#%lu2XTQNg88_C;#7L&fk{WEwNA&ZVUcGR@1#<@ zRbn9LgHitw7s^lgy}ad$bOkiM;|E@cyBieOCG86BE^+FV8d-{B-OP6S+T;o`_KKW& zi2_W6r2Q#5tS08+^mtk z9D?@X#d5B6$hViGij%<^^G@ow{wf%7cP_0zN>*@-@4K+UfwbeG_c?END@+ZDnKZ(`brJ*{e0vzsR65SkKs6k#X?#_f?G^SwRe7(ad-Ryh?%UM1&968|2_3&gWmu{)V+-eutL z1zvFe!1b{q2LDdr1jp9bYY+>lU|JN5XWokFlV1z0h)M?gwQOAR_|&QWVe#W2KMlXo z<54QwBWf1>pS_`ScB~KnmD#TeesAJL((1xqbKKa$o=P{xc2eXX6_dy0(Dz6wox@Ay zU6&Ly^&0&5r2KSL9`i7|iP1xjyG~}>*Tnlry#@~yMgt8QoRp&YyJ<(!(;e^EYhpvn zrPn1-0+SVnh{YkBd`#Oree}u?#hZNg|9|9t3tW@;xqn~r4aqNt+yIkTq975<#mdDn z5sU5Gvu^9@?q@q^|86_$WIMInS-Z2$op$T+g5r&ff)~($auWmrQ4|#f6;Y`Q78ONQ zkbu^r2#ORH{?C&Hn?#~{A>l-~eg<#eTk!pT-{<%Ip3C>k{Sr#c=Ut}L$rZ`G_Q-|8 zW=So4K{#mcWaq5a&bJ4akPjIzk)9X1-&0uE8dy)mw9guF!kzf3c73!%5+1yF*cmn$=3P~I^A^N?RPN%u0Si1BTa%vb6{lkxH}U>O0XcgLAjd2 zzVo%CYLWXIs&+9wU*CSjTb4~v&G%P)Ji&mU>;KjKce4D2S+_$Op3%B}BgH_Xb}bcE zsz?t}FA}1M-OS$sUiY9=wSPX1%Al#jwLv?7(Ij2S+cy@RLPNvJ4mCqp+5dM*)WX}0 z0UIt4dM}do&zWu6ZGwOtih+)nEmTx`NHaLMU=@aB*KoZx=+q=doeI=eLt=_<9U@Q} zKqnfD1Qv}-pf&?*vBc$uz@9xW?C@YbY2vp1aADbs{F*EKmxcc9jb2?Ob_W8hjaBYn0()0nR6z>cU?z=R-o3|UTxhc73MRvh> zoUOo$xiJ~(XXCGZ<~A!+LCtmA^*Uuh)+5(Pr{KXqUlwJP3EJW)W+g>pF`sa7mU^FW zvb@TxBlNo4QD=KjF_y=~8jx7N+I~kYgXF(cC;h-Y+V^Cmn*%R_EKrB!iEk>9!)_IG zNpwlH#r>v-x&%m_Q75rY7Z(0_#hXz2+b_Qkxym#;8~)1^=g>zW?J($+%w5H=3(|Pw zQH>IFpCGYV7Ze>);?*j+K6B7%NJn~TsG zJ!0K^YoIhz1EafcglMx2CGV7NlRh-qme}QSvq|T3=JBpHi2y#Ln0|`f2U#~2uBUtC z9if~34tYXnE@-=j9p)y6J)k>6+h;urz84Jr1>6#lL5D@Q8y4C0$O;3owQ}tQbq%on zejd7Fmb%+5apH{`$0n&8mC15s0qLS!T}#61oVC&(?rH|VvW81Rn(#a}N1c&FR#M72{3Q z=mw?wgscKK(DzOg!wT*eaGrFikR$dKshX1o3YhqB!j?NsyEZDDMZdK>{zeF1%lPA* zptL>){Qa@;NIc1KV6UmlWF9M_7~mTz0Ivz6FDbrvV4T$&?~RJ&8T>gpAM@4S$(sgbe5E)ZUnLG0=dT>)*S%ljII(wH&wFk<^$Vsf=PuDN=$~D~;X) z*@|`CUDMlSN4=8eu|j~2pqZa7juvhY)hIVjDdIOq;@r1G)bG9wWc$9F+wXFJCMuki z^Lv6pImL6YXJcdyXAo@ICtE>EC64paAyO9<>)$53=eAgQ*6ob+Pc{RU4e>OZ3GN9{ zS;o`!SMLAZTpN^4p`Zi1RTd;pJKdJMsnH?A2~4BB$n?n3ZHf~QIxQw!={V4{9(2lr zvM((3LOZqD17|0YqK4^9|M_cC>IB|#=Yr5dr!vV+X!QnRFNh#ja508tQw2#!#nr=G zznEs8qODDULTMsugeVLRU{3; zSrtLOGaiAnlPo{vR}Sm0m9Eu5_!xT5Cf#B-MwA{~qXU!fYRicFkUF~CoZ9}mLXQO= zer%=63 zd!Oy}I`G2Cg43JA(Lz&Q3#kvnTF`zjw&fwk^!$2|^(yzo^3m%a$LJbK4SgA!tdQz+ z=)_Li-pLB~Fd3SKtQhN2oMIU!qMsg}z%@9%PQFfyNQ?tJy%{D>Z!*QKp-4PbVR=^j zp9m=D^#z{>Ymwm*9imZo07qlFD$U3KY{2HjVl@WLFSPS#%NQ8Wu{}F&ftNY+j0L9E zZr^L-#hgBnM(O4~;vEJFzH06g(AJA_uO+)A9iywyX8%yEe*Si>K^g9^X7~IZwv;M_wkmOA0(3hSd1zDj2To3CHw9?pp0#Z)1AV<7Mdc-rqePKvHojIpUrH%$gX^f-lqM;B; z9oZOLZ)%<#JS+E%XgpbsnB%&D**{$nXHJgCChG6NxuiT3Pi7OvY{Wk{YTeXN+;jb6ByEM2R^k{Ia3?CU}M}Q}`C?COBjOY@- zLe)>rsiWDz!+`_-7D%Pn231S?f?I-t>UZKUuK_1)>@LwM)(tu#Ng60kf&m2mx(ZPO z)Yo8%!4jVY;Sta?#ykOv?5J@;w8=Tv-(K)xi?;^2|UaZb%$%EX;6ZtQEp_K`6HbepFJFS#;BlT-B*n0tna+K@}Y+TDh7i+t4 zpXv-yDlUTV$8<<1Ym}?sxH5Oa=3i=*XBA(}ys5Z5qaS(<>}4lLH~L^TMl`fmT*S_K z`uM`XZ~f<@Fwz1gBQ~6LlY>*$IO|l=eYT>tvSN6?KE*2Y@%8Crw)K}9mqtH>`3Ok+ zM8LXjdWs%l!Ua&3@W3sbeOuo~Aot(tp>^Q(&Q25jKkf-H7M~{t?q$?2j|bfIehEVJ z&kdl7fw~xvrXTg%!|me7`k(PFpNe!gIU}z3wie z8LZsR^|6`hvp7~GM2W@w+3LXmyiph{HlSpSU>*4W>`>ykW|HKaphQD4sTA2jMeSD= zgf>Z6x?+jjX<(NbbQ&rsz(#;l1zW7%GKH*8@jP%4-i{H;OX~yP-4G8$m73^9ZDAe!sDCIyA{C&IfT(m z!^?w8rZ0xMWxroB_YB>}Q3DMxpe7UKjV?p%3<)w}_DPsD7Di-55wpjGXvYm!_!w6= z<3D4}D@z>KidZOlK_-nsr$Zp*8qco<8WD}_(wRuka*f{US2^Nz@`#7Xdc!f|uNcvV z?0~Ux@Vcwquwq%C=zoE%8?Hgnanscxl3^T*&!!ltS4*d&jw#c5n5n}u1*~1n~6&?pED#9@+TAXv?Wu&Vl;dk=S)mqwYy?ITMfZcF;Pap29^*SJ^@p0V{%EQ9BR zAMLF8F9T4Q=pI8LN=;)&duf)LJ!0xPWvqU+yo4mhtxdlBtH)mer9N|@Yq5k9Eh{nhk zQs<0YN!|%hG;xh$D4!VHQSC6Z7M6!{xY9$1Eq!a<|I=~iEOZv=+pvP^s8<~CVMGH5 zxVRMu19IqQ$tmYFagsL@4FHRmEWz2nFSh9*u)>DvrFQ+q3LB0al=_diste6y_YUjT zSct|ahnxmF_%?X~zf+#!3Z(Fpwo1$S{Vo;s-LTI=#B*Il(p0ELg5pW+!nx|PQ24h6 zYkykK|J#C`|G;bP3TIaMH8WoJ1TZW7QeS<)>c2))(&o4?0I^WgR_B}{ImpTIfgI)$ z$q^p5tDk;vj}J^~)5?|VBENq3CTW^jeH7SPuvyIB9U6Klj08;f1kupFxl?^VG-pe9w3rq{oG%v)RpRUh5O={>2G9Y`rzx2nj-tb|L{IU3v(n{RW-T?H+JCJ9|fHD-{19v_G>vj81FibSK_?w1b{a(zWx8l4uHHYS-i z!_dAE;BeefXUSD>w=l!h^5Z?*(@6=tsl{>sWx%Lx9K=0LF_jc4r=nt=%Vy`y?%}Qt zY7H*(Qio(Kz$m^Hfucr4w6d#2ePAV&$Qs4 zCEGyyU5>$>w1P*x#>i@aJS&bfbk91NhwN7vy_*Je5w1URS z9@)L%O6htajLY@ulGVBObFH&(R^ic#Ifun5roglC>HW--rRIre3%T6O;(Bqb0E>un zINkD2`K7R$u;?k-sv9#ZUDi#`knHzCwYRt+{FJNOAKnf{qu+0vwvnVMHhcw@_;n3Ger*U_CmVcsEx3VVp=G2 znTqNGrghYIMaBiK47-xl+khRNNRi=Bnh*7iF!A* zE7i~^R2mhszu`S{0&agJPQ1r&neRPG;VdY-qw%s85Ug#Ix5@DsCY%nDN`F)uRyS~( zMf<|DfaYnx^ZIY-MP{3%>0+IRwf=8e;s(6`-4c|eZ6vEh)C+~#^cmo3PlG>Yg*%ku zK<^YX+&<-v8~NM9!U9=5)$;pv^YCw&=e;r;SNuT9e{f?1b&&ndcw6ZVWm18hGe>~oMj4WnncpNv(1WISdhG#9sBv52E6?H(>O`?UF zP%M=Nt7n4@A`-lHhmSdqticez<|DzuXs-GE`vaekCR^yhsIZVNY*2O@CKQ9R#vM6Q z2w}%-GLS^p>E^0U&f0JtU~*^_q0SR866WfN&XH6I)J1~hI9J2!V7h@AOhr(VH%>Kp zqK$E7Eov-I5RVUx1|v^iyLJ9)Uvtf~7bM4QNg+*kR&aM(o!cXyJ|HDVDe@dTU3}h8 zD~sb)c{N7f|b#a$dzdp&D@NBMq-y4Y@;Sgbg^|r?cpR2t0lv z{BBq}WRb3lABhX3i-dS|$7HQy{dAc1e6fe1Q#9bzEGniu!B2`7mh*PV&bT*5BJFk! z-J$zyuJG=McR~g#U3@gOFN_`ZY=^AJ5VN%LBz&H-=P!8~St`RUbt3B1wPdFQ^JdhU z%sYoD1{l>3fWFz=pap&vyvQtW7k!${L&U6}vK@4N@5y@P`P>rv2x$Fo^Iga*=da>v zLb5^hzMC!t%Iy5P_hh+pP%RHX^-8NCOVKv7k950K`c}dj`98g(G@lgd&e_Lc{fxYn z3x&58u#TT{PKtdfC3N@b&5ylc@sts_>JTeehx*Tamp_fBn89&ZkHE=e?CLx&W~h`s z;-P$F?AxF<Wipi|wZ0zqew#!M42ClDmrJIk0Ux zW-`$GCsPIv5ppouv=Wvh24 z=;w6?S4mO5Ouqs~7LXDd*cN*4RQ;Yr}s$OWce-!rlf~_&`R(3J0*z=LeI9GUnjf6i;GwublojSn&fR0 zm{=Gklih5gWOA_`Ka{%JhULflk^kpOV>hP4hcR_k{nP! zDUj51^5<^!#|E-ukFweIf;I?t;F>j7*yRk-akn<0xmg%dII&HBjO6=+5}0I%s4)_y zop$>aa%<>iYi&ksz zXRGsAq4lK&`L7heV*thLhkO2moN!=JTsHy5MT$93kp?QNmTUnI+Hy}lIiXHz_imV0 zh7#d@BnL8uXZ%+2kGzf&yEy>~yax9R;PWPi4?3YB&~+#}I6He?1nOO`;$yi3(3Xbe zi5EtG=80OC*{~n)bc9p{rEp?}tN2>by*y5eJ+NgUhkgjksu`|#c!N&sISYlEs;fR) z&qQtxh*D{w31!f!-fgX7z)Am~^~Ta#a6M`FMzqIj?6Cva`fre{|6EC?@;MKJ`$)39 z*aLb5lDylLt@KTI;QffE7Q}wj`r)^FJf3!<^%#2EU2K2M=aYAy;E{7!|JDLC+)fV+ z<6_y+;=G|GN?PQV_Y#erqvTOla3NBJPEbMhDsz4b#rWF~As zyu;m*mD!+n=AAu0(=aE!_xY>>Qv96JcAPeur>ZEXf+G8=s7*l5P&OO+c6BVrhVm0E zK53Q|geGz874?d0|Dq7J*0Yc7^i3cr|E?|z?cr93#rfCCia6@@kY*8-rj#nSxa^x2 z$58{duQ(?FX?*ejZF1a=4!yb&`X?}U)1U-SU!>+9m^%s#vd7gBzO4B@9Ad27XBk)T z7FFMs8ZacfBl>TWFrI**t&vZePBFj%m_kL(!+OQlA(&(7llDl|+acwKT%Yh1(|mIw zn!1B73B!`5^)vUV)U~1(@G0t?zi=7F^R)6`$5y__@UX4B`^e9-{M3i5wslzYjf!cf zzceq>WD|sQ;MzM2wWdj26z@L5+v;871tepQ9IXT@>-t?WfqNPx5|%r@_)vjs%fTapItrRhpCu|k8h}z|ATpxm7kX^gxM+Fy{BcJ)gkMIO`aB-em z!u-Jim)8rbCzCH67%s6UWN7^qbDtu2sVLNV8IbLDPL@|nZ%JZ>Wen~Cv(oRHz{BqC zOoG9go9@xTc&W|_SQ~`(NjlXitg_Qz+Rp6}Y+#`MyPV%kx>^H{xD+&Ab>)5dhn0jKl*aiZqgE%K zD%S#O=LuPdLLI~Fc1w{~`u=vb;Iqad9?3bZz&Gx9%fH`cPN~Yq_i@}PH?g45**POa zvY3NukS=J0(m(>F&aDPIX$ogQ;@u3c;Pow)YO{XQ!@qCss`2@v43iSxKCLxVtiL6f=N3H{@rf#0X z_IUufTP8>zUiL{{W_wbf5I#)Jp%3S9bd?34|BumRl8PnIcEEfy70O zSmfG9C?6jChP@+s1=yin{g>`8R|A;dYbNtZq634e&;(3dCmy*>42;$JlFG( zq}ZcXaB)s?cthlFS&ewLXSxIhi0Wazofq2ew$Hbg-ocFzy31{*SB(jw$%rsAl0j#A z4J(XLV*e0#^S*vI9*hIGs#stFzp%E7NAaprz_^7I!mqZH$xMMFqeg-(jPOsw_HROJbmxM^H_K-v-c z-*35r?=n3O1O>LzTLCKDg(m}x=f-n8U@Z;JC(<@f1r&KT%U8&%IjfoM5FL>4V5)GX zyb@~3oB5g0P0~tV^XQ=QTlPpdDT|#)9}Q%Klp(}e^Lw;NvOEd><()Fh+hASt2Y*&d z(#I3X>lwY4MHB-xf_b!&m4 zR6$oe_L*N0VZnZnJXu~5v`_WW9V)Sv?a*j=9SW}P@*dk_6h`~Rhr8bLC=8Qfa@<8x z(l_t^pYmJ8w2<@k8uvW+&fv{Z6O~4ng!d6x1?_jP<5h4sP1MT3`n1X8y-!Z8 znwBa&L$493+dLNOlBnq+<;@$I~<3smi%D0!NR!Q zJC;oLIPkLPtcmHXp%@S(uA-tElnFd7eV?eec%G6RoS^YgcR(rJ{Ok~ns6thOM%g62 zEUBIh6y3<@fch7Ovp!SS5%r1*X>><$y_?qc3!u8guq~QNbsOi9e@?(E_q9G9s%?;y zkM}-A>L9wDrAiCTp3pi$qde-O73eZqmAqPBE8Ql)9$3pu9%07H78dbY^VCRS#STf! z_H6GrFZE@!IC5N*3i3C`Q|{x!;3g?jvmEt{1I!>pB65W`NptOFZ=O1W)nx&n&}n|xLyF+7Es}8Rnixu8&TU0jKB6)gr*xZRCo_QOrt;#A1dLQq$rp zP9Hg`1$X)w{tgp=*V24?r-)fC}+?ow6j#W8y9G}z{4C@vb;+G+37stGkC#l zFrd#y(2(tW;MFjW#J`=%$$46#0^h!p@CC89J_N zELYHHFQM1~vO&?pCh4ikhAbvzf&$8BBQ0i|JepE3q*gI$VqJdoBGUuCAYINnaKLhC z{(OB33CYT@gXeoaG*>cz{#}{Y6GbMe{8h>N?H+gGH}H#WVdUC z>Kcg2@OOkp&uo_;;?19r=_XY8J3FSd_%})6fk-MS2o(A!aMhmDh)4|7r2*EdpKG_V%Qn_ynVtwmfvFQSI|$n zhdmCl!Ug5)alKn?z{M89x)u`Wz__r`TB4`7L}E;AFGH?pT~HPXF@Y)yS>iF2qJi@M zMM7L(k5w1-2n>fk@bJVvSpj6+{F?XQH_t3TNf7VAzKexya~l0nu`!?}=(1O-I0txA z)5J(F)#z0K(YTFFw6IgOm|*8K6;KqIxnnvicjGN2f-ktRF z(BvXk?IBqXY(x&5OfI`A29y{Jsi^(oOJ^#@D4$yL+lV+$kGUw@du|`=SX?IyVv&ausV^P?sYjO)zXH`NKOPRRK zqldeNn;i0J^u%IDA-#H!>Kx+yYd?R|P%O+JRH>CbSK2n0!@m7=E83R|OaS2b|LR%|QcB{laSC2i)V= zgp7fLOC0SCJFRL*KQ5ykf#I)bH!cp_fb_kq-`!_k0dAqtI!l%1@h~Drfub&G&GnWX zv>-P4c9Z0gX8vxkYU$MwY&S=lvWF3g!Y$$fr%R$H<^(<9)baBHr!>C~ZW*X3;ce7O zq?Q9@gUX@%*1KWze0oT-yb*+0k2|Z804jx3$W7XzCBIxM3J@B{Wd~NMbjcq^dCtA&hvg4f&KRf z1Zm;{S*AowfSnsE$l)8)dHErWgm9OR;1{FN+HpOJ|M#Dl?wR-IJ8T!W1xn{km9BIj zYld}IFz7VjbdPx%<%I2I5{EZPu-9r^Y4HyBIl#(Rjtfr>{DZ&2_u1K5yoltmi~2cm zRb91-r?iJ+ph>bA6PEohopdXSo&pQ4(~3Glfh0B(HcP^ie458fU}# zN5k#25$wX^*R+_ktU9bcvA~;($$~Nr@5(x%7`r7{7Xz+$!@BS|v8@Ks+UITcKwEo5 z8=c3>puChG`_o=?M)@Zf+zxCuEbzu-ZNepz6M!PY(Sb+6uSyK7ptXW}0XStj0n6Nr zxR-!g2xB!%XO{EzzZY_l8G~DzA>v z>uyJ#$5Lc^1XvgvF1!A=jD=yA+T;GcsO-RagPRhm{YXs~I&g?4&BXCYq?kC0tfZov zq|KrvZ~O^e`CG^uC|zoTZqw11*0LTaD>1?PzN5W>9VTkBmL!|kcss1ku)q)$Exf~= z_se!h*|%cZX^lG@Bn{HaZcXZA8Y3&XdG05|QD^VI?sz$0zyGSFN&e9N6zO;GP;FprG2d8;A>-NP z+1sN95M0j+L)0hrCtrEqsdu%Ip1wF|7dRjK8aPZz>&*-1le$0!67_>^G3M3|4=Zd~ zds7>n#|j%S{rcM{Cn|^LoZ4RDHt0XBm<22324$T~ z`>b3}Hr=YswmtJRD}+7O?$`iUMz;w4!|!jq8!+?Zs?F!gS_k%1icL_EK`~HjyP1l* z9Ny`6!2f{1Zgk@~*tf3+X)Jk);FNnSez@s=$h8l|0MzyFpd8_r1#D?Zzphf$g-I!+Tzk9 zKdI~mQq9|8Kn?$m4NRqPwY1SQWg=c+;2H1R9|6Tz3 z(t%;K!ld%4QpI<})cTzo~wN%7WM$~>>jP%K&&UL&sZf^-&sY?2@3Wca9Gkjz$p z)VKHf8Q>R?_K!r8{hYD1Lc@}g*HcO{P#0W;J=JyezE_~lKH``R2%~f#L)G}JAIN=? z0tEIZ%l8Iu@=JqWYjm2!{4#i|rAV&+K9<0=x zf7A4s`t`W;}NrX1m=MAkBD{Lv)c>bXy5 z<4MqHpht-zqdVDlvnpl(hvxE+Y}(iyx7~(V5Q}UT6mpYDyUV@cI$0&XUc5}W9f*i_ zlS;usX@y@69W&{nBz5vKVPhopKhj&KUJdC5j>c92w$`Te?nW%}+3y$QU+q`UuT~wD z4t=Vmp>k_#V`Lw}$@GihoqihCexF;u^&Gqo&$j`#MhPM_$7<-~=mGWDo|y+H8_bP% zPVg0y`ofqS&~Y=G!M0NjwDV=*45sf;!K8Jb5IO+pbS%QrIZeVcuR*6&?`}HRy-#|; ze`&}rucaY%vMNrK^pK}o1Isv250xE<#5Yoj>~XUJFoqq%QKMzpyN-V6tUvtkH_ds& z*_a#$&SYAkKQ}PQU

    6e64#^ka`;@j)M(N^XE4xTLmT4KU3TQvd*&EJ49JD6hBnZ z2L$nUrh>+S95F;jkAkP(A@heZ-Tz$7Q~D-2pm^LfWd$saI^FumVEBg{1FIl0+=X`D z_ue}@H%&6kF+P7NpGP)0@KU_gWY*Y5F+eW0m5OSgRm60Y)BaV{zuszytPtHN?U83A zljT*@^L+HjYan;MbrR~FL)9j*1Hr=Et4)r3>@{2HhR>?OAMzW_e%o{WEw6NjNnKMv zd^7xSqj5(Z*r&DNjt`Q=`h#ZYZt5c)o}~Bf1Y;lcjg7$Y&=y5HY66xz1RECJI!C`(n@y&72SZWSazS( zK*P+zsq5^o#bXXeYr$=PoxMN*u{VE%i@{)&{mXaHkd@;JNIMuU6H_T>0|kG6RBXh3 zqH(PUQyU#vOV$LPCQCpFb{l7QQmMk!h*qFdNamI+d%`|brcA_w zZEV>{B)AW>2S5r0tFk*nk2;_DW;_%!REVz8M?)KBO;9Ga%smHK-;*cU2@|X#V78pG z7baMlxtB71C-s|qAPx%{3qA$V^Yb7a$~z}rLmrT{X-jpN(JYz;ld+O8@;p&vQ!kP;ydPDP8HRPJ^cIJaGcA093Tp z%OLVzM(0aUDoecZ__@g?ayuCnGdNg@CS%KDb_e@>ywZCu=ASF6*ZYdra}PH)yh*-O zwJu`d<;KVcPET;Z3(8;N{_~<7{+EJ0@12~+$kpD91C!;gf?nt^yy$wV+0^F z29$;E(F3bNHu#<28J0Om-?`;Y+McuGlSL-=PMmKH;(G_Q_d4jK-)(p+ z`&$}i22~CfQVUXmPkFP)F$tC@VV9WBiOtm68q{dSuCu}5I35;G^>Wa`T~P0q%c%kE zRL~DN$X$0#k|nGmcr6yDEXZ)p3BbrcatR&xxgb(6IO(1lTFyTVEFWmj@EHEza5PWc zDoE!o6rzOFj)~pV@Xtbp;U;FQ^bYhQ;Fm6d{~ePpoS4Dg3%)ges&_-KvSfLj7{7D3 z*A7vSEER0xf(GStSO@i+A8gt%*8sbBs{gHs9AKAAcHC$Zxo8q3_>5wXQ{*TWg+TfQ z=d)798J8FZ%xFp8+x>dOK7;ny`*U^V4XFB< z?^p*3qYK7pbHbsLT8Y8J$o}z01BriOEDUI47=EOCn#uHscn3%Qv-ahyuoUi&kYj|e7!8haxpIj`&23WFhe_w-9=Ns0rPB%i{ZuaV8%%DZc}2XG*-mwh%r&>szVpyP4e!CCNrd#0ypKP zGTk}We>rIAAd_w^-%g9}kq-aYcEdhEU=_0&H+a=6*_==JNtO-=wj35XcFV+Vbe8W{ zI-8VFLQ2C^uV|so2@h4$d;EAfSTdZ2%8gdZX#wQdcrED$nISbI_RY5W6cTZ z2u>3p6rG(0uCUfq->I~EVm7^+!Q{lC6H*c(vmWXfu=IPZhTGS+XvA9pQ-6mUK zshO@(-VLh=T7a{koo23~kr_EmhVEdW-#uNYem}!tdp;_UtRngBY>xxOW<$R=-)5?HQYRXx+9Z-&OL{;L^EV?06$9q?K?Fg+U z8t>)qz4GjTE7|)0%wrS3$>_yoTg+pbtQZd()t0P2V5DDuQBph(`_| zi^j(Cm$5i9_lH0GmbpvAW|`!`@Uh_RpeV^r-r686090@H+YyRxatVQE{jsO~VBzf8 ze<@^wnm>NO?;qwBl@6PW72=dm(JUN|7C+qq@{v5A-?mDLZ=|PY$WI z7<%$b`|gRoHPL&P!4FB3|1yv4X6J`E@T&Q96F;PyVu0M^AQkn1vmF{tPfmX1)*(Zd zgL(CC`J8$4P=$Ejymb8878Id7XtSwtrj$ zCbw2W+O!mJP~V2)#0K{^_iU&H$X4x?J(R9Q=rTY$)=)2)-f<%a%^o>T17 zv?G!RH!S2lL*Md+Xv`{pIS-S>r3#?%c2=L5h|!b&34>0%Nss)LdjhFfZc&wosb60lL&$PyZ9B}x<&wI=?T~*tzlXca>+ZDOiW49Z zSRIxsT*5TU@S2?)_Urzd3SOybFkn52{t!5XFF?@ z8t>I0nSuZ5lL*NnC~$F*H|SI*xhcoG29&ezBUrOgZ@X4_TjSyX#BRQn_^9ajfd=$k z|K81=B=?0ePcYxyK*gC&B-~z}AGzZ-Q`NvqQ-VLj8PYxD}F0-ySHwL$%Q(*kY zVa1NozW@04&&Qinw>(!#YzqV}@%((xhY{T#U2dBr8zW0(Teyp-mn$)y-RafJ(0&OMOs-60^{_=(#wx&Rcyh6n$cJYc;AZQ(Fa?8bG4Lay~MFADY%I!qa!7P@53$ z(ZN@@a=U?uGnQ}7dWOaDj07T&oeZ-L7GL|!Q!PVexWKUgyRZJiJZHcrFzmo9aSIuS z^pIxJ>X1h03@}LkbvftFK$S`$d0x%bIhO`(oo&xK=7|Ah2c#!1x7&{JWp59A)>4^MfL91Y=2l^`5$`mIa zI+CrrF{7V*%eN!)Zws(F!z?B!<|Xtc(6E zEX!Mc)LAP{qZ`JsM;``|9cZi^aT}aAJ+%1d^UgJfVYL+BUQi}U577b@d=spB>qvfN)%GWHP97F*UUK1=J4zgYhV5mJ#uz~Z)ze&^o3Sl zWZ7a2=eNw*aqZtmBU$3Wev1Xk5`7#O*2fc&`)bhXrlM`;87Q6vtCHuPO{Xe%g5qA@ z42|-N_a2qHl}?i7(5ok+vM0=8bUD<{Bse#T)V*P#W`bcoyh{#!V}_00Vi@hwAj|}B zRwBpR)0+OYFqda=XaY9=@EBR@z)1n1*ER~at)rN=6iLuGEvC^+y+M<@-1D$PU8+c% zfWcYyLZ7`;bLi#r)G_M8*FSVtLv8Ii+2A}@#^j|oLFV7x48S zOprMbj8WM&`if)83M9q`zy?5g+AUatWZXw@7B2X`0VFqO$G=Q&I50?-ne^a)MKQe; z>A@`Y+6m2k%;^h(z4#9kK!@Q>ABWJno}zfuul3aJ^f%$4V&l z&?S#5=xW&+&lCZ6+L!a|-7beGIAfj&V>;Dydju;$X$05e12T2J8``yn!cLi1aD68H z0rhS;m7R1=o}j+W`AVARi^3HTfF(ekz>5{G)_wDKArc}b@a{=^!a6`3FwFRMdk%I&sF@_LN!nygy|Y1k!UxMyM&|R;^KRQ^pVf9@yE#_$#Ie--XM(~4}_^f z+h{ziPSE9{fpA_M(211z)wwMje_(bg{}`y1oOLVmiydz#$9cpC?6EO>ni(4TG0O(* zOaqjh8H`o$kGyNiqTymU4!o2Gfi@%eXAQ-~Q)CqtbyZ#@EE1x{$fvOM8=T)Gl4NfE zG&_y_Q^v+hP(0-h_Bmu36o1P2#q0lEMZgAnM66uc>35AzjX2=bB~2H1(&e6G4j3!( z^6c6s8^b0%u6p%oNX;C0C2fJ!4D~>;5CIzxJ3wH#M~VHYwSLkYLvDIT`Tq1oK%BO=^WgZC(aX&6(say z!2%dh44yrA99wpv2c#aXsX(F6I~W)z^W3u1SiO zv?ZY(G@jKVE1QjTiq?0)DNWqWualuN^q^A%(9>bXD~MkC z7(2ILkmd5nu0qcD4Xd9IJ3eeDpS>`pbzhjQi!M>j1&W-bqSENkl*kct9C83Rp<6OW zQO++^BtzIw-4VJApj6B`&nXGJ>3(qz?8<)a3y&BAAwh9P(1zDA>H>AW6@V=0uDw9| zT(Ctzt-o1Sc#`)$Stoc%$#P6FAahkM(<)f%y#tDUdnEZ`!0w>o?DlH_9N!kLlw&7@ zI$GE%Lj0rS4MjM3^P_%o@S)hFnXfL9^-1x)d6uDMXWbtAMWf#^HD*tII#JTM)4c!M zVf$f%j?_saHWXJU?42F`JZpa4Xt_dR@}zsr?7kS4(oiBE%=)cF&L7aPU$Df zG6yy!Ak=C!U96{=B#I;gU&pIeyaM++!8(w5uc3Rxi@CUcKv1ER**t?8ewz>f({5-D zWDVBgX}7WYF;(`#KCU*lfWVQo z5)?Kc^k?0W6)46zm%j1gXh;bixDmtxDPeE;7IBp#*ZZnqE4@f+`l&%D42`WKDP)m1 zo~IQ&;zbML|MDD8J%nvBcDhT_FbfQmxSQTJ{mS&+lJmTNmrg}@@FQNC%NFsY=~o4N zIJ>5w=T#D$q0&mQ4dbvi5jkMfK^~hS|pD-mzq|XSj+J2VSt8HL)`_6mx_k zRaDd&I-Q4IhW8{8BU5FmQiag6d`WZ(LhQN>A#OG(wbF}o3b?D>kP&w6Xl{YDlsi zi55m492}8|SRsDN@71 z(*6oT0`IJ1CwC1wWtng0^;IGlqoODBI`b^vwJ}xf^WFkE=GH zCu_%(Vw2Z5gJQsM*i1!Xy=yH~%(?X%l!}!qO2iF7o1aGS4a5d>U=YO@5#Ki*DR#4{ zRggm;b;f)}F=yoOoR$CTk-wT&e!&P&v;1lf6ZLjni~i>R#lL*hfRedof(}yk!UWf@ zn1H>WVop({mWs;dWYIY3oc8aRV?O}8DD8Y5$=7NB2d`d}t)7TU<$jkzrzW35?ru^k zz-SB}!@M*^Tm&65Wb{UwRdtgT^Lf`~dh)O=RhGvRPBXtkknednOe@0+kUgx5cUq8! zRAT6iBV|=PccBpPh|Kax_TCB2`mDX5k-r+o#^~AGY~PTm|NX^F(+rsEeb;R%+4{m@ zs?r2gJ1M4+BKcHQgDzOB&kq8XD&~+OOT1RTP>3u}IOs{<4IK1UZYr>rp;^+%g_JPT zy#q5mMEtWu+dw`~UB$_vt2it9ZSoDX^vrf6fW^~?%XXl7`uDIsn$xAtHLtsM*yXkb zI$eDgOR^kkbvuG}aM0J8jCC~}JbY`%ne8}@_Iy_J=D7D)(*I)~p0f}I`(NZ3P+tpf zRNnP$Q(`4EMn8~ut=p|mvO_c?6FZ83t+fN7v%jPG9=4aV8vut5-0t&v|4X4^lKH4) z!e7ZMc9V<)FM#q)CYenXvl0K;sI}j`G#9c!KnWL%sz_jp`x_F zOo3WYSf-&TAJH@94?5Mm;bNdwP!G|ro(MfX>~?y2cn3J~g|bsni8|<1s_1u3;9((Y zinK?r#um<2K~F?a7@m&qUoR9rVY$aL;Z3;>DYvY9@Xz-5zdpk@1Hfv=aonYHUY6(N z$p-vr=LBCNsqAJP$NiTP&>u2NEN-WmZ4}9(q7uU%(CheG=_26{@`WgU0_ry(C&$TM zm%A<+Wfy%}QaI}~AZ`Kj7FmgoQf`s3AC9!TE_Yit^9TcIYq1?JLTG9YV=?02~$8T6}>qzktIb#2$=D~f9#4WSz4DhTEG)480Z zq}}~8ly&BCdS@IGw9*ya%^rtH6>q12NCiWcsbB5#j$K(0RO6$g`NoLJ-9@2XL*oR!D}2}KBz2sIJq{AHrr zIiP5jL$?5nf14cdic@Y4o#)*uh!*Y#W;nE3xc|D_J=JBEMtRZoj%z7@i>liH3)1eY zuA%XSa@8QcZgQVXIloD@N7XsG6^vMg=#YQ4|Cf>$$us??QCb}%Sh3xFI_4}-wG%I0 zo)}=j*vFr=9U$8rxX|UO37+;+%x((Ap`%dY0o^KK(WnSYmM`ON_DGp{B`8Oj0|o^$ zjynAYNESkxu+#0dVv!KLmC_;QM-of0 zHA*A~TFrEc5?%2#WWv%S`dknYTU>hN`&Fo;XvfKEbWpK==-Hmh3M%8?{>k@_O)yxQ z>;KjKce0$FkL zpn=)nj?gh1QkFoTxC|1ktuz59;4yp_JZ>`a&ect75=J)M*|ue$dg+rJ4xDr5lo{jhb$v5&nta*i11xGLJo7=lsJ*4 zNGpA-yox#f?p=a|lX7U(h|8z4RM_f#nw(Tt@rpUoe7yDx4``b9x)y>2&M1wA`)nZu zX(xC;hUrK**unel;Ke_p3@Cc!`sfrHa2n28a@d3k@E;l@D~@7TQY4m&YLn;s+<65W zl|geH3+u*iqB0q`XM@D#^fCKj86M+ecf2JG46ubal6bsL)>w>3K&nyh8RxBAIwyHU-oqU)xlQuKRUB+uyBJ_k8)79;hIe{w4v(*I z$POr9%5S7UTdc@|mqR;DW|MS^fhPVG{c;HX4=jMf9y)!y3AQNSbWaX!2R4>Lr~Q7# zTqJ?P%o^s*?1`4g<1-T2JigeD8Nc%@3$#!7375V^-+}2WxDWOOa z6@_*EseJs2-L5sTw9?l|Rr?==T7QjF|EUEiqpovs>6H_3$PegsAqIy_=oG~s)mhnD z87{8!_=|*b;v8BVn$DC5pQeL|(>5ye}d%_wcx4Fh~&XJYhygD}y zm~acZm2{eShHL9@@KcsB!}aVh<2VhG*Py0m$J}f3uDSPQXQ7@3ujv&%l$60$e;vA+ zZDD3Rl-ahN#yfb;i5GEBN}62}RL;KyT1IJLKKfmtw9Vv4mD+x_R z7%Dc7&e+fRj|d3%`*R-O*|NC_PAV!lXLWi(%Bn1IKw+{#qiiHQL_2*qOa<>9*XXqHlb#ZuDAm;ICj>U%O=5wE#^PgHE4(*ZW`Si&4MkIL`bB;PUMz#P^ zc@145>Xo0Nk^I88w)UC-c(J7X`~`}Px2GlyLeTZoXY{mHhF zEEfWG@MFv~gioq8IBqh5SjYonq8OVeKha0i zFzzt>gPn}Z=dG{Qna4(75YNLxCJ*Eay{<|x`>lU@k&c8|eT>GU z>qMYPtRsVn(GBdhnY!lB2h95mpIq5G zZm7mmGvvCdx5MVOc@)#pfhgj!k9%5NNp34ni!aUrc}W=mvmj4cLt}yEc}^Y(A+!>$ zZhyQtrd<-a>!v>Z##gT{fu@dD59xL{!9DV&UejmSKE5Z@hAs|F zhL4|Klb1lQt;i!+)jhMDz8`)w@b0vj@df-x+;s63Nz)IOf|g|-_wF03LmDHI;Hywp zCGDnn%q{08kQ+04-kz#bsSZrEGO3vU*ZKI*caA3Sxpvb6{YCS%8{jWuzY``DuGuI8q1yXoAPS^?UFHYRdnNTHE@RJnz29)d`H$+;dF&z zZT#-(m(G$Dc57n?j<)VKSsUk33=gIVY+w=1(r|{$biG1UE8M^-^IOg6$l8!wSWnniMfZtf`Re&&vY>rDBvt`q=!JiLmHjtZ z0b^X)-1vu{hIvJ`dig=JncckNz$@haCi6-G#eg4@Lq(-?@;UL|OE^8;PFf4qs&{zZ zq{^!?@!yjdj~czmZJ7toRqfF^KQ&I!pruHi@AQRn!vH5L83YJ5CPFF6GaIo^P+0urzsh0tyO-9#7x@1SF!bHh(n}iI!O(%@b-gCm>N>?}Dbh+s z;Y@{m`Dzrz?sqAM7~K}??zgMvZhEcLtsh!Xy9uzO$M_HWWI-et84V7oP{G^y#D4iw zh~=m=LJs*=!1-IIHV54ao4;jhs{lC+VE=V6Z5fg_;QPtp?-Z|eMFNGBARCgUD&oKi zsO1f3wF>Z@d$Mvq{zkDZJns%)UC8}RgdelHH%O@>j&p}U|2*^z9iDp~s<8UV(C*=v ziX*jn#%}lMx1Thd_uhW%m7fh+yq}jmAPp}J)VfWe)A7IAof)P>9pBTY zkSdP4N1h8t0*fwj3-g=cmEe_3Fm zM_;@#vQ2(ty2d+w`cdZ((x?yq3{B6zD}?K~XXruSW3n3h&mTW!EQg+*ZGN*whHbv} z*>~A~K{ngXV1Q1%{jZnF8V7D}2EAS*UwbRX0Gq}pDhkP@Is_GhWNwD)uBS$MUUnPO z)iC<-ZXe-x(dE3}FpYATN4rZ32Ulo0^j()-P>VMPgU@!@^>fU#hi$ZHu!#v13Hvj$|@&?3#+E3HOn(=>zV`>)=ureUyuD%`e6Omz-S)O|JcVzW&#L5orWo$Q@ zc{Wo_3PsjaQ74sguqZC%B3S`Qfbg$OU&n2eJrpdQnC^UvB=9=v7WqED&?Rya#l399e((Ndy`1v5cnS zvgGT2v+H~2+{vXT1G$Z2U@fr~-N|CQg>%WjAry#;C*CJ5@;e}8`sof?`PA*<*{UsE zOf6oVv)K<%!OGee4jx&;!(&)ci~FDPP>uf~thYxsBCNm7`U_Z%h~pajcVAof>1dJ_ z4x9(IkgUiFc<9n3UH|eiKd8#1n}gDX{SXY?4V1bEd=F0DqPjF)-NroN^m8>#D~*CV z8QikjYOLZyjuv}D9l$qov|Eqdk(}x7ef~ppi^3+R;=nMmU>Gs`i3K?Z9panr=O!0T zf{^bnud>;R6BmIr@djUHDAXvg(f3^kov>;cXE8Mr|D?F3xHa>$L+n|}GjgctCZ&;} zz^ZGOuXg>w)!^Q|*G%S-#PPtEWgN@cLNTxsO@)Nev??ZPR*~QC>0fU(L{^CIllI87 zkZ8H?F)z@EYbrhW_6(kwycn_(OwNF^1Q?JRSqvlKL}zZ>X{Wb)=k|X6+HblHy&b=< zw@qzl+72!VE+8nNq6QEFK|nEXs9_rwM_hmbjpE88E`y_>$e_aaJV{UziRM7U7wxUT z%E?(?ob&$kywCeA|37ntB8RBh{Xr<%Q99*Pqr28ky&b5)W`8S%gfMEfUo2Gwg#( zr}S_QS;nn`O|Sz(Hsx+o2^mNYqh2FQ@G5{-gbHpRjgO70Xxw%**}lDb@=o@UV}+&( z4?pLkob#k@;k|EKF}1?u2JZ$>8>x|OpN^r~ z)AT(lT9+0Hw$#_tiIh=H#Hc$0(U*CS?LV-6-tA!%Y)1?`>+yR2;X=#Sd#CkKtTfZN z%W*HFQ6!QwPAZh!sSgJwixyHHvO-QCja=F2tzb#88cDIMJ=*2=k=}UboQw`4&m4Bl zH?V@pOa9k0eqq^*Vx{E+DKLhFg-mox7}4D5mmB`Tqhxfe^cenYS-k%-{1*0)vU>l{ zn`c$|>{xE8(ZMEI;=KQ51Zoh56(yunOfp55W1SPKUG=-8wEW-SG?5SuV@PfX&(Ep)=f&!veIk1LKU^{QPmM9k8t&!qz{OQWtsymYZt<34_- zfh9T|YXydl&T6frZf^Tls%1@q)0z(}^#eCwZt%Sz%A?!m3k3J3?vOv^4Y*`Xz2=qS zThAYG$?{JlNBCKR*IvCxuxNF>5W*544&(Y71`YPsgL-)HKYV>n)?@J>r>-DbPHa7N zX4YdH#lU(~NX2UDGS4bmT*wi+N>5b9Hc788ZqOv_OP=kC(G{6w&ZECLU7UnU)lTgE$9h7px6m{QuWrL zGP;|48+iLJh}A99^5_D3lS-`xF5y#Cwd2ASJgQ8o|G0%3;Wy808rCxUS*3VA#|kB@8v(Bu2Z>*&KZ zf7|ko$q~^`dGH>oW!K~7#E^jQ!C{-8(-hN0kp?QZL5d0tO^QaLL5UsXdRPkCCOad{ zQ)0vw_sbQ~hw+85a+XGkMfM#KrXO_K9nvdDb<0vZ-*5i>`H;)=HpWr10z09|76&n8 z^%RD2(U`@*auk%tQ5;C!8P*`p@mwB^6_4=nf|vWBTLSKQQB zIn^9gR=(?6L0{!0+QO7-8?v7CX~uTu+4z(%i};hJv2ofhkQMWhJ*mRIMY2hi%T*(- zSBGqz0`5<|U7IHPiKiTYg4O!-()*3q>Vi$iW=B`qLQ>$wyPiX4=3@`VfD&#Q6?-8H z+IfMBN25rNxybLA7EAQ1Ly~%F7rh1>gWXCzR`3>qg6TliA-K{?cZO||^h6$pZjO2> zN?YCtsTJu$cSk43bbuoJwHZ5QRdk=^lK7ZnQOMD6UxcP%`|hM|LXK^=z^-s&g&fL9 z7q!yH+)bVmNlrTW<mj<^3MF#-sPr zIG47A!}txgD*~P}7%Ol%uM2qR{@RC@GGb2af>@CZ8w}bvt&Y)ouMNCP+C1AlvlQua zb?TIzJ_Yn`jv70+Hb&)#JAMcJ7%eP8`}jczAF%@MgpiB;4bQuKdX5xK&kk3sA(00x zJ)sSJtAeod`@>ncg=&n%|2HcXp zpUk->TTJf~#J_?+wJQ93FOd4|G0#3|82kaQK5lzNJHI#qUey7EF66Iqt5;JYE*vByNs4?^@442T@Cd z&gT|*|LvT-aqHV>oxywhd$Z$Jc2nTAJ6ZQ{&7A(PRfUL!MxSsOOB>(NDB=W_FKa!I z%pT8;%wr7NWL+NLK5j6nF6L+(|+@NlFlwWIog2lNvTUWshN$%rnL zV`s8JMCNHl^hjF5m+s4sag z5ZvK6del!UrdlM+-3LiAmCaiqD3s>`u?XC-8Vc0#Lmkm5N+OCWY~-t=myjY+3eOQ8 z-J?f=88qw;304c?OK~;-_1oW>FthhPcPY8?!q}FjW_-r?DW;Dicd1xogJ%k_14(JL zk!WZhhTRRz4)2hpM%F_uG_nE|Dr)@C8+!2!B0x=?#7X7RaWQGi!H6PI_kpUZgQS24 z3NE7?RwQWicvMh5=(07a6mHx1ni>a3&aTpkrBaaME|{J!-^uB7tB0J=ZF-~U3SOGu z)}TR`OP&>U#m@@pRzC8!WYKXknUYThpiYJL`Iq4v0PGi%6M$=@@zaB;D;M98_qi4N z-|^RgajfUp%PNHXgUYy=N6zv`Ax@}_G0IxOVrnO>ngQ1yA1?m4=Q&G)3a1Ut?l!Yn zkGOb{!ZA*`f}1{d2gt3vhcL^`c3&qAW11<{z6#%5>Z{42LH0P*NqufHT^0>eqq*u&Xujmrq0=E96?{?4m%SZTGuR83L42r3SGdVxd zxpsUPE`5lqA9v|?PMY7L;4So0h29T31-W`vp7Oq7m{vZXL%o?QTBGcpvCb}qw_^jF z6`S%1GNV8E_IqDxuq-;TlG|?uUf#8mhXIQN+N6UZ#kgeO^jzn%B%x%x!oo#&n3qB%Mv#*p^D5wg;2z-6V^jgWqBGVi)PKO7=^ z&Cyr@UdNo}s#BvbMOC_2Nx*qH#On&{mkhep%d|p`vYXx!2(+SE^k7hPbRDC<4F|=q z6u@;<9BSqF2Bz|grkzx_N8~FXh!d$jl7rrBB{#wi4_9uIAEZ~%*TBi>l)N)ez+WAwRVdHsr_Itg-lB2kf_QM z4af}j+?^3UGJWu%OBrX`tbGxco`!pnE5lJn(PnJfb=`S3G8>P6_mJ4MWs-gTZ%t&S z6UPHf%&bTj#cZI+I*bGx0x&nc8lg|EH`6{BzZu|wRP%1a6xpD=-Lc*lb6m5 z*UA=H3X%PBH15T-U#Hm}qnmTHAsq67l|6Z>Xm03g+Juna@B7t}h3q!f&Rg_HfO2VA z-^VJ7SxJ!>`ICHs7W=d?E{c&C~06pvs$bt`czdaU*#Cf;<)*O=kK^kSh+eLA^|BfKN zDvc9Q7KP;hRIlo$w?o&)0;)}(45D&95NK-zg`b7Wt20wPa=7b)>-mQXYK9p*;m-wE zJFs46jM%b+4H<&&N4mK!HVOQYFxbC6=HDN+)2q_kc&@y3G0n4u%y%VNgQQ;7` z!NF(22O1i~6qdprm>-iXAj(@t!UtoTK@Y0OvXRprsiB3zSnuBZbJU5ug|BE3o%n(bN7?-DktmJ7P+ za=IqCkxq?jhW^D8ZU-nx>0B>}bCd@i+B`QVWZ8s~(cMJ%{aD~_GCInnrIlnOyS2-S zw@UlXOjHTQ0HaI+tX-a)U_Xu8V(JFz(YMmyRNtGk>?>Gnll~QmNXPTyc~~rAAPR#b ztioU&i2dBBvqkEQ0a!cynI~%D8fqZ0Nb57ML*g%E!poy2YIK-*+EYzL%>S+lFNwcD zaF-lshZiTdEjP^Ya-L$&Qsguho8em_if3+l9v417(yJD6b_FC%G@dz4DnyABn@r-~ zH|K2rUf;iI=U)7A!uJ|}-mE&!>82mVbWQD$<@;&9GW{O%GTf@Yv~)B7j#mrRRA}i{ zq&)hb%sAE_@UV?X>;1CmWKj|m7jjjs1xr;2zyE|_kzkvkmA@hA%Q@(+X9EM(sj1g# zebhy9Y7}O7ZQa<}2VakX_UwbFfoj;bf0M8JgQWzJ)9%u(i27`I?NpwE`G&A=qcBlW zXej19HT66-B=eNNcvg|n^;ZKaDuxN$#6IwTc*EimumA6sW|*h^z`IQofa^IO>8 z%gV)hDe`JX(05E`=ntKLxI&ISXR?OZ&5X@CiUA3SQ&jAO*SqBDT>O%^%Mb8Ng$MX~ zf-9;O+*09#*Y8Nqh77pi&*f8%KUckB{HdioRe6FHT)cJw65;r>PB`FFEXfjmE^bn- z^vj1$dy#0rvL@)Z{GxoNU;K29qFGu?iX{(3^)mc@dR4vbKy;TpHA=5C{wtD0<2_Ex z24u!jj?f2rZDGqf%_`IHwq2EV^pJbDSsCdLTft`?a#>!#@tS0USMBRn|0Ijqt#VH6 z6+tEXu#H<9#iUXsnTpM#k14c~t$ZLRJ0d>Bs}z z{<&#(u_dvk75>OQ+zvj5T(yzu@|NUG4ZjPie>9qOZ zcP{tLwPf0ULAK&nc)76-v&jD}vtUXA-AfKhKJ!Tq#duA=e_miKe>HDG%z(=YH$9XN zB8yGlvouicg)H_sx7Tle`Z&6V-idyum8N*`N4rV369@e&%+`+W6a(4qA}Tgbgq;QH z^1W_)sPl*Vex3Ia86h?uY=yjJX+&=%xVx~gMv=mH`H9y$=yf3Ht=t!Q8DcN$bz~hG zbgA}v=(WoI0q9ZTo%VNx*@NJbg&ThG^cfw+RaVCCrRA^wdQYecGE39Dx04blhD@Cq zWONj>k0N`h*h`*qF(@hag%HJ?TLW4lN}NRx1gv1xTWCCeTv5z@AZ`H%R-?G%t@Y8l zVxi1JD4S7Z)FoYxbYsYPr6mu<>VCIWvTW9%3swu(1VNzLu;s$ai2`~b{W(<4*ab{> z!q0FljU_7l2x+YFGvPO3Kb%IH@bl39;R!O}!fpVkU2O>z*9>!RlPP97MG}z}Q=`CU zDAe%AKXw7bqqlGK!L#9G^H6rZasChg{^!?BfGD5-*95Ya-EPN;V|t&O0b)PJ?4`&q zDt33&VV}Jm*d#|~`FHcT(AVWhIN833Bqzbz^tmNY>gOMaQi^sry(&H2cph002V}=5 zqffthT860FCzC;5#6o4}l*glW^H-6?NqRSpLeJUge^OLB-F{GeHhRqL%CkP$4MDD3 zuTkGOp+~+v@Evl}iI<};Gdx_Rm~#|4gZXS!9>aBKGb9`hD$CeApx!p6fxjc9!)whO z+7M{jULD#Rwop(gp4TAlq>~sV+U<8c2~j-kZN&c{l3jUq-g>`ox=P*!S?w-ig|yEd zQU!;}fXtwhnHAXYjt2*cM%hKAWiebU)C>QNUm_> zVvK*EA$@LmcF<*ws7TbVXnLlOqMb4GEKCi*{#1OiTW_{|q<+sb8soGvVJqN0d)2VJM-XLs9*_7RMl&N?Si$0@v zbouGwm%|qD8if@>n9OdMCv$q}3jz7uJkL(m$+;KB8ozo;!o(W+zCcVj<$U|lyH~!m zaqhA=zxc}GA1?Vtvnm&&OC?i2;qLIf1r-*BoO(%yDEr&la|gbDH}=I}h*+VAN_rieh4_! zvqV)KJasvEsa%aUnERFZ0F-7f6fBh;tFT~tf@@7s9m%5C3-G2iKaFC0(2|*J8T%QKC&4u)2tPTT-OjLsaC(m54I7&v z(SBgW!ytGI)k(5`U`^C0km{lln%f6mTKTA$-U%I%MS_z^>4BUZsJgCkcuoJPC&O;9 zedK}IO@`A3DgNzSlizyYwH1-irP`;AJ`h;xy8>EH^1+eq0x%~rpTPbY)g%h&TyA;v zxUeUh(83b7BV)_~XIa5^f;?)*Vpr2@^P&3GpGcw;uQvH+tIb-9NvFsvDz-^+OqecD z7R7}W85m*;1N0uvQq+3KG(h^~)W{0DP@F#5!P@VS=aMBfU>p1RSw|kRgN7>Pjd$El z(D=o9@;*s*VrXnLgGLU;fC$Zch{SrJ9&dAWZzRU<) zA8V_716O?@xI@-29|!Kc(YO9P5+67=JB$Rl{I8r>Oc-$yDff^(C&mc03=G?D?V=c{ zr`txwmPBB%1Xb#bL{(8`9B_V`6kT*O5A!+?!^${nq#V=IT3W9_A1W)vkP@okZuHEf zt0W-vTt`YGP5@s=XT-*!OsN|Cs>|jq5OmQ)1KUvun!Rlq9wmF?*luuQhnv~y?@qN; z8gbfKiWLbIV-OeEoXMgXlo?ZFY7zA#u#icOtSzV{parMeky{zF31kZnvxgn{fgLzJ zzB;(slFsf0DPq}eW`Z7(-(j##ugd1M^5ZA&mmL%!eP=ai$;6(>_Lu|HDneUA6DD4g z+@5@p)1=z=PTad&-fC7QPSmT=2Av=&JiXTu2lhb%FxUcXxSeqzt|qK(&p2Zls%HZg z=M4i9D}nub+3LWBlRCUxB+JNA+91cdaMC?li{uP+$Bvt!G0myXuy7;>Mi|?5O5~p{ z%iyfkk99;OOp6QIqe>LCi4K!IZZ4+{(rX9BCq;=+qr5+871=oTp;x|A>y;>I3C)-~ zPOsSJH~Z+DHo|Lgz$sRi+<6oA|EGMV9HDJ;5tz@qg`>m(U7L6km1B{C`RP1_1 z1OL3&#Q=>m$9D}!>vh;?cT|Z-9p|vy9teuu54j(r^{SYrS8eiG2f5!8@fDC$)T=tY z`$#tD>dgB@OLz11s?X(@AR>E--@)I*X_uY!N(jbfWi{-ZbfK#tD};^uePJaYrBL5r z5R}R+f^P+mSR!s=#>jQa9^;R|{>+21K%X7P)Be}(WzRc3P>Td3!G>~x%HX}?iqI99 zJXS!xy;H518L_Oe@gu zKkYq27CG?>lw-C6t)`e1iX>qfQ=Vsr>$aGFpksL`Z;oz}0F^t2*wXy$81b-e>TE>C zc!gRY{iE5T;`5!Kp0U)eaavQ6Wrm7Wib}RGt`rJ7MFl z_0dd?-vKL3yyX5#%Q=C`Qv6Xn`LAS!6I%-4JQx;Z-9Rzxa1#*QA+IGHrX;-zK1GqJ z6Kdvfx!!U`ZL~p`il9wU*_bsMIsuCWwVXbO@k>52zc!nUN8aWWLpbCFD|k$}{i_LY zTaw(qAR!Pdl(@*~TsaH85U6{CB=C9R^)RSoy)3*U$4qdMV0=vt3lup7E!GBIdHAip zLX+88y8hCCktB9i_D(E!wbjhXY^0b>imahxF@;kS(MJp_i`zs=qAm35$>&{5rf7lv zbZwv-H3P0p#|s##zc*)maKZu~Lw3g6l-l+DuaBjCFWQ6;x4TEu$(|PmAI)a?sHK>L z6se|SjmkmToU`Wj;}P**7sT!Iq^R{ViLUDNzy*ip^*vlH%Ev7)h}#?M?SY7=NQ7rv zVH;dP=PN6Tp@L1LG_uF{Ms8Lj)tnkj<&BA8?Ffpdk4$mTG8}$j8?~--S|* zojyI#jj-FRV0u?rKDYu+iWD9S)ZGkh-~c-tZ?QXajSjjr3Uj#yK!t?zs5n2^6omPFAJMK(lYd&2rb7^h39jXW3J4F$1_LqMJgM2nDS=?gj` z@O)@Jw_l#h8*s_E3i|x zNfH;S3(zTb(D8b0My>sF(f0lpj288l&kfXkcc+iZ8a-s{mJ*EJ~6?cmhAx1R;eAc4*@Sw(0N%VpW=z}ty`)QC!r7=22PS zNtj#tUE8|P?13%Yyq?E**#bXy*ou^Xwq$|{Tgu!$e!VTlHsS<*|i$ z;>ZyoF^xGVDtM=5jii|He)IER#Qx^J58nTim}XoE<-}o{r`AoPP{dl!VRA%KElHQ7 zHnpM4y$kq~dqAiNw|e83Lt+a72hO_L(AF25-!FN4`Sae@gMMqFGM(Gtk4d6K(5aC~ z;kgyAD_nKF^vU>Vya`lTBEf979QnaI5=P>uJ00*tOR4fu7tzeD-M^+(cC9p^=5TIzbMsY!WQLJ-4Lf-|UunS|>Kw$G~SWN+&fya0+ zt0{2a(zj23|23&4<1w3jk@NnS5i9)2*})fpXJ^p+k|YRw_)hhlO*FDuFACAm zH6H0z>-{hQgZy8sLffLY3vYTM`;H?W4+9=qp?#S9F-9;y^YQD0uhm#e5;<)pf)&+6 zV}}OlP%()Dh#!l9QU~RCHHur%?uDJB3k7%F9?I25)>F`bWiou!7+TgS8bWp6*i#4z zYpl4&t{=y^Bcq-FVNh^L_aG~4H=()cH{bu)+SlkpyTN>DBR6;)^tvi-Qhi1zxHYS8 z(~HP$dV`{uUI?}1J&}6VDzXKTcUIUg-wbQ1MzLh({s54<;$=?B6dm<(q%9N6lmM}!ol(H)ol(Ot;y&=|;p*PLM^H7Tz_o@{^Y+cDpyLFKfYhzt}t7H$wo868pmXqRe2LXS~xOWinx$n;SrNXrP zQPsV7faFXhm1eu2GKwjs$W|&g9UAZAKx|MuXE9wq=dNtPW#!bfoc5W60e`9HeBsp@ zvWCGAsZlrSGmv7wBTO>8!k7$AV-PbGPy^Nu^R`a$Fb9BF`% z9sz~Fv*^+(8b$moM?ueOz2YV?H6@GID?TMTnh32!SsFf)E)7PVxeMv=z7HY0_ck?U1huG1#(BY3rrES>C_c{ZPi zQO&yf5Dxjo&e)}24?Sl|7|F(_IIl}atk6rME7IroFrYlTmag(22rBYFH8tMrgj;#E zUWM28aEl{XhqlMG%WuhbFUP+;;Brge!C%YhTyM#jMqP)Ek!XcTx1` zUPWkaa34%#(>Nw~OfzEhJlkysZ5xP{O?fHggShx-Ck36@a#%?Up5q*n>n^VtR}wjGv~_KLdwB$h%#BA>cFzl=O7oYnX#PNGgHo zPhcsz4t)vx6{W&VNioQLCowqA4UdOjd47vyx}fP~(4|LQ99a+f^eiM)5E*KD*EJq~G7~lx+wp6}E-d((w~7@%INUnTUIp9{KwnGY(8W07|JX5t~Cdhj#F@ z!t17Hg&%-0O9%Zv$ejl3T$cf1{`(J*XMc8v@B0q2GGm{8i)pjdN9%se2qi3(z2N>ko)vh^ zFPT$H=liKE_}g83Wak3ZD_yrM@qo>GVtvwC0m{~e#rm(a`HFQwZTr>qNzZ#&r5z~H z@5}^g!X^be8f~6;X5O0A1QaX_9U> zOPZtGgvKo5%{eQnPv1`e&y91JOdJom*p885UWkKmF?`rJ3*LMy(B!MuzPYNLYkT{+TH&xOq@8+hO2mP-{cC zD%9JjgYLFRJDj^UQ{64ilpvD^wvs}uq-{#O0*NG%NO0afWOl$!k^LYtDj*EQgvH_T zxg66mD}Q9dilbk<@LLmX-YhvZm0TZbi{r$hw#1i)cboK6%sqM5QLr6oJ0=af^t-n!+GP*;8L-Q#5MGv_AR5nXP`~SwRQgs3 zw}kHx=oY7k7tqz5LU9W3t}7_!fuwskoi0~5NRea=+upJH{id{zN;o)EzVfk@_vJ~BJ_Ru8`=8tXTLlA)F^bk-bmcTWgU2yarTjVbW^CZ~<_T1gspN&DsL zaugm@$MaHnT~c*gbVcZZEHx^LS@yDS_Ml5#NP`riYovb~GbU}q&M1Iu)3{^3i5+lP z+&l6WOUK!1@nOXk?x2^-*8@Th0E-p$48_H0Bku(kDqHzkbVr0bO;o7ZtZbKOdi8t2 z%7a`S3kArC0J|snpF_bi3XD8jMaGJQr#;UOBM(yBU$g8#d#<|QpW2fKEXfCy-K0TU z&#wwz=~oq;5DxsO9dsUjKJ<2Eiv&pL8laQqu55kyR=NzS*qapX21Mz+zaTo^pi2Rb ze5#HC(pV5=g3{A~FzRd85tQ=z)ixKC%ToO12Op8R7ba;9LUzNL^in7$i6YCulZLn* zvgxf?TzU0NAWKM%?6Ko6ip3P!0u~m-+3nVnYv;dV8PavyMaD{ywn3U5z9^)duBUNN zGqYV`TuN|Pvzb1?Um#E~11n+JT|1N_W%#7nGT)EZ6FeCHumg{+8;6m6=b!zZ@wN=< zK3B_&6?rHlKrq&V+tDvxPZr^J6gm0rlkj6Tfwn@0)u48`**a8aM=$fVEZMhG>)s$; z9HQ~72|BK55%$QJ0kt_&FRGKGP)Ivno&X9NWuE6jD5W5pd#QpO>^6>%n|M3HVW&d zS)vM|&b5lJ5LU_Jr|VUyI`)9SOPnLREndi5!8^#kMW3BrMYnsF*^jGsgh{s47OQFK&3M3_FmL-c~ zHc(_86?_%3%n$X&o`2EL3qK?bZamt+s*+ql`heY4hxl4CA# z7D~p)ix~w%tjwu>Pg_Tb>&as|jA`w0yWe}3ta0LEg|l6v@F1b2cYAv~)>l0 z%yl>lR17UC_Jx7PXx33N!WN9W{qf(HS`tjYARet1p3IatQlbw87I80!{o|qd0RLR@ z>d-opAvyrmJDbHFWOK+ZApUwd9*u=V6UPooLr*#W=f+O${eP`ke4Y{n7=TCxC51qi ztcSZa>i7KD3Nj{Cr}v?0ACRqVnCfnR-i|WHyV66K{&Gs4AP34BI(1!6;k^wxgO;%+*6g z*r9M4X@+fu&`8fZ^4$Cr7w3hVOvlpn?(L+6U1rybK~ZOBXmk{_k0N`h*xuJVyrD2s zeqzo&8Pt|Eanj{m+|O~Eqjy6Y9OzzRyH9SAo;&EGhm(6GM|s8cWnlR;T)?V=6|URf z23Hm&SBp7Nn8=6feHx=CHU8&?El`68W8}J*aoRj_q_a@TWDl^}iXRK>^Yo8)0G3Gd z!AH;g(vvCamw)E7$~_-?!3(CRay5SHFD3CY#UKaPB&`X)6S>m$rqp<0(8Z{7xDirM zB@yIl(E`Vc#jpcpO#d!6E(dkH>b+0pCg7|| z4L(CMUYO-+ry0!hDF!lz*;H&6jRFyym4zOCk|r;mS0%~xYMHuQwuvb)ltil2{PO5N z@|mPZtd;dipw1g>x^byFG!>nkHBt7%&i^QlOw;~34m+`hQW$J~29o|^h)Z-65seL4g zF<#Y1f>I{5`ZOszxtE3ep-0GYLl%7u^n@{Wv>D2>_bZ{L2-FdG2^WGEL`{$m;=Y3} zc=aC93d>|kQ@~Q8(W;l$(#8a(Mo~rA^W%dD1JtO6s8^H$Wk<5;j08{b6nFBU>Rye- z5DkHD^gpelY(mMYzxsa11hYSM{^1HaHj!L6o1b$ObA}?Ps94Qw4Za#hp5VG0etmII za#6lNNTXQC09De^tL>2`LuYUyr!(Rz_?!3$hQs>ZE{ZdOF}RaoL9dv+(HoD{O9RTF ziUjF$vY^aqU&Ln8$?xQM#1u?3wEJij552C

    hzggtEFAyrM!keQT{86mK@b4)QZG--F357f z?yUb-Vr2)OIBEN=<>W?Ky1dEbipl`)ur|-bK3Z`-^s9F8Ghy$HpN`PmXI-7SCm@9f z`9;n{UJrje5Lg~0hQ*Ha_?+>nTxgv;r z?%&>c?=vzV<*y^P!8vXkMLp3f^XY|??#Ws}7WGqJGQ{`x2bDn&?X_3ifiemh-8rxt z2XBxLx)>vU^EzN0ETn3M9$gIm*|Sgzv0)3IY;$Ami?mlRP5y{yLe2Db-#bE!H7UEolevBwi; z`;0#_2#+6j;F0yEWdzS+Lbc-Rhv6o}A^V$pBiY8za5(XDbKJ~uR8kD+PnA=#H#{(F z6c3Vg7ehD0))qFooc!=CI-gc+h37bNf-F%tIiNI>&EYy?AaQ7k2&nWQr2zXC9 z7&8)IJ^R|uh$n+x}UZAn|oCIallfU!bon$2nQ=;7umb9^s_7fi=8?rzY)>@ig98jm!J zQ_u&3F})p}3!F~n9b_C7A0{XQagJ2_R&%y;HbEi_?>ZrF4ArZ4cy57cVW$kw99L9< zPuokzZb}}3!&X>5#eNLC^k~>IAAJ97>+VLHR;3^8*_utX>{cZwc4r&RR;5oU20BP| zkY^HQhofWDBR?=R9sHQeS;&F_#SrzNf>=3AD=bvj2C8$&D&7#Cn0ggGwWi zsJ}c~qd;O746^AN?0*JoDSQx}5F?cAXfrG)WNfoZ=cOz<+TK|eV;p_ybl zagoekvsI;tVxX>nGZov-OXTbn*UPGXc0sf(KcX<=@fm1TCDlGXF}Hwm=RjbkZ?3o< znnjG)lR39Mcgp(2#}u`qetD)Hb}v?E%$Di5SQ3VPv^wmVFR(Hn)W`GUe`=X#WMe*@ z*kQ4fEv+LpK@Fhuk}k)(O=Lt>=hNlU2l&@!Xd#CUUdK}TQh6)vj*(Xm?*JmiZg5tB z>mgC_nV%Ztx3xf92|kKxfSXdxeya}!7t;?sFfG1Nut0Up8y+bV;T{?e3ORkC6a#FR zpel1lSP`oCS}ax9!6{=M;G=&^G7pDtQec3{`1=n9=h0P?XZSf|1LH7UvOjz(U^l)O zc)^la*9v{@ZST`iP~GlPNBX&oIcdHp#L1y)eo0I_rRj@09r$; zf}1?*KwleUsi3M5e%t#tZ-2l8@zsd4v-PUS<6Mo%7TMj4u-E&3>dPC&MRs3);;c^P zoqw*#7nzo@9Dc@Sk~~s(y%X=Wi_Mlb4aKad$XY6Pt6&pz*X@9?)Vt9ev0BY5=IK>c z{>w-;5YrqF>5^A-kYXmAr%}{Sx^K@N^V7h@9?ef1&3@Ocqj_#L* zAbnDlF}0YxhpShWkR6?!Xr^tvK3PXT&lfRk;~rgoGLma;I?3Jebp5|6KnovvzWa?#~K58$1eQ$sR!L z*O}4msI3ENgp>mH;TyjDrkOzXt*yWOfNXGLP(c&Qu!fv=zLMwO&oH75E?*YAZisV&cTTAxEH^ zu0Lj@=Q7T5hjllN87}rSF=ivNgA3=Azg@LtE_`x>?!^0RD{O|y^^Wm6jlz((#CS|x zOpBzN?h~p}Zp27-W@NA%g{?4qTnUkI1WSNiRpRY32ZHTe2^uidjgoupA1 zKW$bPOvh^t(n?Z2yHx0K9I*zFX@MFeKA!pbpMQM)j;{$Oi~l%v1<7Iu6DRh8b!ITx zMls+N7gDi#%EACFdBCLPGFdxk%_~QOl0|Vb<x`=1+TalvQqa+FUbPZ zZVVLTRXp0xZI>Tcz!Ij=a}okO_}BT40nca=Gt9Gj=IoE=hJ<!{LiCd`F6&^PqJ>+pX2 zH!d6K9n$4-f~yhrKti>O=%PNC<1@M_kde74z6u*K=~w=8L3oHZ;2UdHfE&m)rc9@$mtzHoKi%w%v{jS5%eVlbX~!8T3u6Hht*1S<!#Ke%pHd+#SvRGfP`hnRbEiT5ZXKnM#`46vmLW38G6$r{IvDUKf68VWeGp37=$9L=)SlZ8=c-v#8{p^S$o45R(J={-}8s*vl-SnpBt7qpn{_^%aOTUuDoEPT!?h5MX zo{^VFV<@FVa&zh~CRua|$}M|mlt)7ijj}7KMSel7nWk4&`)ucAx#sw`KwU%pE4za1 zZ@cV-G5dkbG}4}b&x^K<;XCabWF=k?Y}+XGoJTJq3n=8y#oFy|&r9LOWR2*Uuqvtr zXxR0fZRFwD6*+C40JigO>$qdRkCjcKe*MKwxg|QD%ejKZnKT=7K+KE$R?!W-X_Qsw9E5p{q)cLTK$f58Yb-rNtq&1MQEyfg|V>#kR!z-m6^cv9WVgo~#pggX=?EMZ~hqu3n_!v?etixFb7RA?~w zkfrd$3lfvHA_;L!*iA2_+T{PZE2|;ZK;ly-U-iZv$vRT))hTLH4SeIvf4}>0%_^O1 ziTJ)(zgv^){brR$(JVbJb10l}21g62VYBa$!>q6T*Y9ksmHfC(u7xH4OIY+j#udMc zju+sH-%9F854V+Ui&>)D5waqtDJ**u#IgoN`r6|M>^lXcecxu^W#b3CzIS$CVW+K?R8iDliD?BKQvSc!&J zGOeMQG>W8BvD!#vpsIjB8U^xepi2@r6?+i7mv0wr7t~9x zxMD36cAl;Di*rq+E<+5y8;Il5ThvtGgXQ$O2g-*hFim#tyPCe^}#Tsj}&`o8wQ-tjKS|tK>`c=>c!di8f<;}wvA@LyR4h} z5jGsoTXeeRq1$Vg-U=I7Iq{lg#gD<jkGv2a!!iS5uuf-2c!M>Z*r zfInLS4(rMYPdo0^!xJ<6EDL6zt6-azvRI=aO~st~d*=>`%@|c^*1e@SQ>588_kR>= zhTvu0+T9L6vTrN}vrcS4tQ5?~K^UcU&Q*eCChBc-+T`2d!)Ecx;3820eZ>_MgI77{ zNrN;y94$;t55?A%a+>Zm!dXA!A1y)<`5ZJB&She zb}0Up#Q}D6SBJi{Px;9T8c!L=-j}HVNSp6tg2qFpZYj~QgNF0^Sp+!ahPm3M6tk5g zTd3Guuno$mQ;7PbyTH%0grq>lWJ!d2x%(j{_HH2I`n+ywrUbg;4c$NK@;(W+gc)Nl zSni@e3YD}SvWsE^d;3D@)-IT~cx*g8PfjZq%)aiUo_4cQ?Ek^~QZ!OX zFmXZpWXo-k(}Kjx#%2$9CzM-%5m5omKNgJck+{owD+1EY5et5@L1k*g(VT zAP%^|4jQ?lKZic=e3Rv(4sR4s?3Vs*rCgmFStLkwRrk3ygkqEgvVWJQ?TY14#dI?G zOF58uo#fmL>LXdAK1nifxu9Kt|mL2xazIm%nzxdm`VyZ#<3uh#Y+y|6;MmB;GGU?m#tH1{OTphBiJU? zt59sLm4Aw?53i!vaXP%KIe2e7XXswzeNECGfx7~R&gnsq(^#%hE9$0~k%Ay>Ue*}K zJ;YfeJRj6e&T;I3D?5N_1Z<5RSFAup{q)BN_gSibJXhOjE7~6Du(pQb_9rXw7O9|d zO=?nTBX3PI?ttt705(%)J6PC^I`(_5gXMR#mwQ^a|G?g1SS0I_j_y`LFF|pGWKNgx zD#X51BlV2>25&Wxj+aNbLy1tCEFZ)K3zZ!aoBg3tPqc#3Mq*V6Yz{EOl}qUqMWUoB zJ=`QDXl$43ys_dHiHDJR5nDR#IbF8T#R%U|%Tr^Yf=44*pNkQkNaqbv{9(T@-?CJn z)w<@i^?ksZG93TwDP}DNU;-YMn}Hp)HZl&r2U(&G5;f9dB5`(yWKVQlND;Ij$Aw&w z+>~CMVNV=C0?tSWY(YNK7>+o%wMG+dsV@J5prgzT9k~<(ovItDSR`CG#%~LzCx;d) z27=RFx5=Ovy-`>&t#bM&Ud?nf<`dD_#d~d?x^*hjuNMi@{OoJ?Mw(5wD0yW4MjFhH z=Wgx)?;Dol4Nkj5wj#&S@wJQMWuz%IQ_?pDDJ7Fb_fIMgPvNbNS_KYmp|UZkm=`}$ z7kYF?4xP(eGO?Jag?5iY>49lken;q*(1eM%y|j}PCpM{eM`u5itU%AlI)6O-ZuMIvbVrQ7GID}pvL zSt3mEkNxG^_!ijCfsMnC`8HN_;Jj(sAG{-2%r$|-CD3IdiJwSv&A^dPF{>!D66j&U z8qaT#B74YYXeGbJH%MQ{3A!KxRuS4stH%u!IC*SvG3Lo;F#s!EIIsKqGBu+~V#0;+ zw(xf(Wuzt!C+;*YH3M-r#ejH6CKcPn!Cuo9G5Z3OqSnXgRci#7$s)leG48`!`77o7 zqCS^hn{kl0$#2ao3wc#EicgITuG!#uXjNeW%TeEI9nW3g-1SzJ2_kC>WZ7i569?X! z%;0dCVm_fr6;gzvl2b_p25a|CLxa{Cp-~>-_j4N|xifDbP%mLhua!o(6!}Co3dD`N zTv00Q0d3quaREryr_0a7=4Q|ZrJ_(m0TYTP5%_S9TaP@C&g3L91+;E<2?*Wmp`l}; zV2Lt6q5J2>;$@+u9@=4e1O1RmRd4*e!}Kdpmiq&@tvF8=(Ep9x6CgFo6$ zvYi-66=p!%PBA4EDKa*PWpeu5>iNl>me2>{Pe5B4a)#ZIF~n1-2T?#H9}l49(b=xB ze#woHdRYot9KKt&2`~~TfDI7VzGx#l9ph!$2Quv7V9VX?Ak%%R@#~gB{}*KcY$Zj| z32i++&_LG8-|h9Kq#{(^6Nx0W_c^7)gJfaIL2_l5p*|{IejzGlQiDelQzQcRr((%r z?n36WRIPK}Md~4lXsq+U8HPNz+d0@U1`4Y?zok(e_UU(nNMw0*y=+ln@tn9A99oTi zX();}mNn2mkh5a1Y{8b5dsMq_?6ce$y&xd6vcHSx8L0tJ2J75VwYPxY7Bc99i(j`? zM~~y~j-4Mrt_>iwrhRT1p|cX&I>|j8vC?m|8?*%j1EZlT9yiMBba@rO zmTr(@DtDpa#>{%Un8H74uR*a;L=w?SAiBvgU;;ncHRN=oC;4q!@Fk*bcfQ@QyHrhwAZhG4}!sm96d#TvSjk=4n-d!a zkQ5tMBvDN_%5u2kd-0>2%jdw@kiz7g-&6C^CyKm7&CCuDeFDI&kIm zK^F|HRD@Ot+vxg9s{>KmL!;P79}ORy;bDOkD_fPPoE`(3tb0-yBEKWC)ZKkSjEfbO zUfkd0)3Bl4?Nvu|d@)ov=yDF&&re6}S85c8WR2dL{Dp$fuw)K0p+l8T9&C6uifu91 zrJZD9Ah4+juHB}4D$By+uhfM?9=^pl|Ll3Q6jS(WWFa02*Rb!bYbXss& zNnf_|PeY$5(C=sz#o;=piq@;J7d_3lmR|DK`~}qS#tjoVv21YRP#{QJ1h?eq!nZ4wRNmI_Q-ye}ARlX+bWOCG~MufZQoI z1#aUO`G9bxe;OC)N0q757KNM;_PZ?*WILo)&^SA6z%qi9VKK$_`^XL~0n=8`v}D?` zLQ&HqNqHHPZ-(MQc>7qxtVaH&%E=l})Gya4u$*lCzOXDOzf_29U7BK4f+XlOtzoob*IOqb8YGW0aR4oSNFx?9FnjUx9)Nk7aE zKQBxQH|DeRlod3NtWkj7gCX!3=`1lSt;T-~{<})2#1o9)09TZ2*|f(wYL;Jy5e$-@ zPno_XtiB8r6f>u8`QIkY1?F7tCHj#z*v?y^Lwe0{c$H#0DAG#BZZ>Fm#Kpj*bB{{Y zt3yvui+^QZuzGFaRf4^%t@I6#T&@8;NKUEqPT|$k+db8}+}c2FIfxTHl%MoA8d$8r zMT`{&YL?qRBP`=A6r7z6Cl?A(YouY$W^o5GYFlMaIYF#=>COPZk<7An z_${Ba?dl8@)c&YbRg+>THg8AGV0D0E$|-n0bH?hG!s$r%f5Hz%fP#^YL zEUi*@fyt?$w?`iZX9wd%1`db25eIoorCZz^X()!#a>_W%W&xWn7bvN4CA?3kOllF9 zL||igo2SmhR{9RM__2eBWD~xuTcZBTp9Nd)*4S7bC-$7IY&u)HDNwr!vz`StF*?^? zQpQ18)$1%*SEY(f zu*l(OTqem*42xnjSZF9_Jw?`1u|=YE`6bUD`4_@o;Ood1sV{kM^lVb3@JhIe&;_;I z3tZnUdaK}$?35%^dPtNh85gXuu>?aahmCVQrrX#7WMi58r(ZPzS)aP;}L zu{u}7K{}-a)%a{)O)#V>4R^Q5&(1~#9hARnm+ypaRIWSju~eKVCkkjN!;PY4zLwxWWO?n*GI5xzR0605W+aCLwm#z#YG}i0fEr?Zm&^HIlGB4 zyPG9i%${}R2kU5g{oOkp7n8Xu{_=y5NF2M;Oee;S#>_aSP)rg>%O9dzwRLkk~^pJ1J6z zoE#XRtDw)j?(#uVA@p~#nXO+anom``VMUXA5%JEIv>EhS8ll{rA6Ivvh7> zklnBqSGa+%^}(GH#&j{3d*kKASJbr3Mv-u05T$9mENIsu4n|R z=st;+G>pAo+{okDHq9eH>&Ua#-Hj2p#1lHczV`B12$h{Ua&LuD`3&iETQKFgqQk34 zo{Zhi;U}St2Hn`9btOl6kheGhC4HO2whB^s>2oiNF#)+*iFXu0w6|(@T*$Qx{P%Q|L2hL_hc2YFYeD2!~nd?>h>?w22! zsdL4`p>iM=Hxfgxo?OS2N2pV|zj74R9$BYR9$1IkNbBKit+jQ(F~RRz^vYMrwdc%= zC740-ONxOaoF3q|gEj1=SA4KuRVZlXR|5@1Q$(RMIchUq%q`^HmL~|ZXdTo?eCD$X zs0qSq3DnNIJ~ckLpFYT4=X;bccQ2<(#CN3yuj*B;e56FvC<^C3oRcH2;5S9U zIq0xL8gh-IcSepc-h=;K)bP@)RtP#IS@fcq$M+xR9)v>IL&1dr-BLQ=FLla2(jT+Y zGhdm*&E{=ToQB%irE{M{JoYgEm~W}(=d`8DR+RfN;tSdm!An3jt2Cn0SI;qQ*PCFM z-owRck4DkXsR{0P@13y@Y6p+deKM%7!mtq1iOe7Kg7?&gX$&UnspF3IB0CfH-Tmj= zEIS?8Z1kNtYG|b|qLQNn{8Y}u+!vPiA-|vc3*mD4ULc?;`P%qw>mN-r3lu$iW;_mf zKW!^^H!WSiI8WXusZQ)UZ8LMWb0`J^#_OrrR@YWnHD)uBvoBec;d^Bk6x4&N$-}TR zNPkz(>LWL$MuLzllhqnUjX$0okNrQ3Ss(7GIDFz+*^ZY^4`|dcnJ}_AA>{fqBnb$M=7S3A_uA1JCbHE{69upABf9Dy3l@ju^DbV#d_6xKV4|9&t=yJMSetK z#DTyCO1ujgy&Riro)uU>#>iX%Zz9;zq5oHM_5&QR&3UUdaJ-;z0}$cO4B zZ=IW?KKf&$;BT++0mNP|c@EvUfKsn}>jG#1k8T!in1$ir-q9S7YPMK?37t2F^6Hkt zPsUOQ?8L5#l{#SbS~ZFmrjVP(xdq>iVs9jg7*hDuMpXMOlrDE49au+&Bvv(f<4G9t z!v_nx{Y{3Z_b1XNWRnxSCsk$^a0kVJzfw%a0xJmJ6mdX^ow;?qG`BmFOP-zlHqWLA zZDbC&O4%P%N3>C2fFKUEt@v(@N{;z_26!)$ey+~9z*Vm*;&w}OxY{`tbiPst8{KWP zJ+Btc?dLiGANH|yql4hno?>U|v}d+e{n-SXU)KI`4!J#%tTc;oEO?0-pvZkHb~`K} zZ4s&RWuzNo@EM{z+%r?!fng8>U}fATZtcu{2n?K>^7yJoQ5n2VP{ltpC6CT$a+&1N zHLsr$BD?B`&}@C#`(!W#890p|m*^E9dPP2SH!LYQIdm!S?-#|1q*W5fgLd23Zqt{9 zeWWqy3;IS_6~r;FO7kb3^eU4rg{%E;1wlF7BXmP(jeD;9VS>M;&uxiYIm9zA377J! zf_o!%-dHG=#yOx&^BcOpRK8xmjvQ6q=ILBF`flgm<{^Pq6O@e`{{!2ZtiW~5VV}dE z_#gEgMmk1_ygBcp=J#KpPWK)1ng2L-1<9I7tjNxncxYeU5WR<6;_>*XNu>+j>$}%y zefSksv&vu@44>y()UeW5i@9-^t5S`+qAb$ z`#(LW=eCmG=Jd2ldq>-(4@B`Tf`Sjw0hNb56cI&HK~NB)QGuYM_yQTh2L?eAQQ@vN zGbl4KI-8l{jyeDKldzw^J=wqSUcdEQzqP(!I%|RfG1}QPu96f79_pZYZqQ15F2w+I zNd^_QfwK?-#q|nwZ?rTnEVO_jjgA2FR^Jo=XqPM{I1S=3cnlS3H43y*>)mST1_mpH zH%)0(B7sfPcsmSk27p-!;}OI3WjC^d@t7a&qANgp-Wrp;H zWuO>xo81qv0>#)ZYg9e&8%)UDV_m->Cmh&>v>BO@ixhL7BK63_*(FCw^B6(47y`B( z&Kk1DC)Ym}EQC7OKTm=MZKs{LPJ*(wEM>g{EBlk=D`0tPr+XKy5CPFZi+s(jHBhO0 zG;pB`S1MPifK&jREOANWq0RYR{l#AIQ;_4QqXp2{#{fZjJtu~2qcgaRNLLtU7q&nf z{Z2?)-s^fQ@CLVwzhdSQqNN{*k;1%9n$0=tpDlhm!PPA(@Jc1fm=>#KoM>EO|97scdKpd=KPE7$TRcC~kNu zP+rzy!=JYvUc}EP})fTd>c`-8?syDJQJM zrZAi_8Y`ahzm|JH^{Nazq->O|bW{JcjbzeGNRB99hSW(VE|_$#^I%Y94Ed4Nm_DVa zrO~-2==iEJ?1ZInSmAOYVlh)DD+SKZT7C&PN7N|K5r8{{)eSi!oF?I;=PtdXgeAEM*cjIlDaKd28m5CYme-lFo)$gAA7Up8e(IZLv>^uqP*Y+6_WF95aE6~50*MY+Q@d&1nK8`p32OSY8kc^7j!{bOIOjK?(vH!c;M3p zRT+PeCt9#XpGFBD)+oxST;l9xl9|=aWuXP+=?(xoWulZ^e7nmL(Su6=^mEkcIIIoo^O{q~dsBmkk3r>Q!!#YH(4A}zuoH{(K z={~18PM^Hb>5Nx}7tqVVeU+SQsE6s+UAKwplf$AGJ#ZW+%Nhl?X}vYF6E2?h??^XT z4OW!oqs$JF&ES7>BT+Yc6c zl7z5H>ULls#tePPk;pu%5OOv$=$xP5j@(3@czfedSIxaRr}?K<^d8Us)03^NlVtb) zS#g=MZQ82qKlJ%>mZ|=d!%m0IC@LXSLx(exmt{j771w_B9KEZ!sUln?9%UPrxMK`nR496XkZ%01;gEjJx4m_}! zk$+6)B!um z5%f6QM~vutRwjsfi;uP39=G?>5#sHM2I!?E^D z!2esNW5(9(zVB}U z(6rQlCXh^akrD@XUMh^>Rzxv{6xmHhwa~4~Mqvvu&eW@txErP~L5?A>EN}d~SD3(A z5_%Sfv-H{ZA&UZ|1)cnP8bzyeNhp+(a7%)gIWP4qnS^C`P-)M9$}0(4OO`p;!WGTr zi<8pb{!2p7+19zSCCn_skA)ZA{mY4;3Jk#ctZ@8a$f|D*aB_@*vzcNxAN=4 z@bbB7oQ?@+qb;99zW*^S4Q9#g?2 zd&o4Y^{4=?k$7iy_v_k8g~C)3&}nEen-+NDZ7_~l{Hj@XwZ*Mke<7;@XY7%#-9@Ie zsK$*FD<nbWu3?HH_Qg&tF{#E!b>=F#)c*QaAllOG*6ENhO9nODL? z4e~Nt-Rs)o-08Y$O8?7q;I02AhqPc(_Yb+t{s+u=|1cGdW8d36ZL`YH5+rOg?IsPt5-dZF-JlM($5Wt(a==JXD*}mIeL(-rod9+|&pVPeV=>@QIQxv$9 zw2&tGjhXsmMS;nj0y>V9<E^`tynFpz&R1uVto+C?ww6wZH+@ffd=%&k9McNGGmEGMbqb_0v45K^kyJH?0 zv)iK=%``;PQsw`iL-vg$UmJy>DkOKO;=)q7rgT(k_4&??erH&2b&{8 zSzR>nQ2KHyAU42o5f;keeAKW`0)eY!$PIY{yvL>VLr7#^5}M+Z%+C?&laSk?TK6Pb z?t1_(uk|-DaG_sf`ZJiGTKJf)MadA~ez8$=fHDJhCLSA0scUpk%YKX4^v_XERf zIcI;w%{M^D&#h2Z;)2Tw(Sjwe>E1WrsNj`*pN^H&rJS(QZTYb z?QlaUxr!cO%WP1cfDGF{7)Oy>vx+|Q>yZ*?j^wjv+2131#E9-OZ;6H(QK?w+Uvo_d zE{8>jnL#Sir$JWYQscZ`93!|fL5-OU8bz68Aqd89@!1}H5T;rGzaNZPVg-#cvflC~ za|{Dl=z_1CNr?mJw=^0#?WZW_1VyT-s3()+p)x&HlopJsOe@^i1QmwmiVyiWDffxD zLwa^va3=WhP_5nt)1@rPd1?xZ=hT4z-S30~40u5n{lswAl(f@uUYgrA?lRi zrXa1j%(vg^3a6IaC%qtVr(4Knp}Hxkj_#6`d#mfb+d)<0F?jpPzmr98_Ea@%d_2Y z?VF8Ek>ng{kU*h)$hFDEuF-S4f7KePQ4~oY@=)zMNghuUJhn;lB2SVMmulU1VI;^c z!9N9XtX_rx#yZ$;s8?MtxiJ&(L`mASs#euNYxu+y-^C~1`|xJuE@6?RiXLd|Bta)? z;K~Ipkz1)hW>9#|V&1@2dGNGd@SK;r^rHjGp-6riNdMO|?G#g;$X9A2W=0J%Gq|13 zq*v=0UpG5{72Xq8@kXXk6@u0GUOo{=; znlvgZQB*ujPhe=s%0fxnTK9M!YF**ls5-%8uXl+Hs?-*SVHMm7p1Ri!`7%p`8oU-u z+wHd0g`H&;P+r#CR-FI4(DiqwN=6PlwKAhkRIkt|H9j?x9v%BBB;Og>pO6viU`Wa+ zu7I=bX)F4T{jKIln_;HhW3#@PbMJo}49M80Kc7Oh!-?=Y@M6>=qp;rtis_`tH&j$o zu$Imc-IHeq*9l=%$Q2*sg1)0`bi_Kp_8Dq8M>o@PLABC_ysU}skO&U7nz-(f7JSV! zI-)SF1+-)fK#(ydxB+x~^c2e|rnZmN0Jle>une5tT)`ENdegL+=?ft|sFz_oI;F-L zxy$CIK~!#&a^Ae|@Odzh@e0Ft3Dv3eR|=%-&6swMw8*>opl~onoljP~YbA{mwSGmX zUZKzT$PwYhumxsv^LiwZc#frV{Z4vOw=Q5-Ju-Q@SW*7Ut}F?GTFy#Gx&wR9X_vTU5u zMGdM!cVq}seX_kl1@yg(k@cWwa8^}C-wv$dS9xl~(*#?U3nR9R%9QQWI>e@dUL* zHOhaKhz2fbQ0*qn^k=zV7DO4JBz*-)-DB>f$|BL zg!`dyBSVv;WGkcSaged`24k6bl+4Nmd%vhag8e3 zV+5jfiUCGV4HdOS^>70IgX%q{G?ug>!&Hs%p`=#PDcr!60TtcD314~Xwd6MYSJ4S` zaMf!OY4@`ma0YsDqyQSY-(G*)&;2IURAOsrXwQMcU`BT91S~OE`!~(l{aOuQGcz3& zqtV5}tumNF&ntJCS13#m*+@>vU@itnA;_ob1rGOXbPTyhXLB-OAj;=%fhyj4nyK?} zyNPpyo9){uNfX%^EF*=+&>_jTATe)S=C)_{T9bX@z|$}@*%v*oHT*_NuPf?P;G$Qe zPuk?o&bj^=odNM#l#=N7TRjzHIHml2**%!plydJ%(k3H~872?*JAFwi18VsDx!BCb z*&WPNd5l}1la|){CHm~2s0Gq^)X2&8SJz9A254sP^@dGk+k_Gh8Xwm0bV`xtj3T!n zO`%L6z2aPdAei^xNz_~1bN!F|$B9aVu6ts$b#xj}T^W!^w}r$`-4viv7@Er`2V7AC z1FvZdK@IH{Gu8V@86SQ%SI@6j15>dW!Nu9S)m_^mb0}7=IWpnbC$**thOd;-GLsEC zFLqoF|0u6IJdfK+m-F^ZP$;!7Xd!RAxP_xpoKS9>vSeJY_#(5=1EuT!d0k!+`9OA( zbioXN@U6|led48pB3UXO>#lFGMz$LosNr4^E*V#^s^Om`apQJJt{Im+r;VId4P0-X zVY4GztTJro?O62Vg?CJuTWgJ+i(?c61561OwaNebl<0|NbdMwjQlmB!E&qOKs;EJA z#Wx|mMP3fdAIOv2L69K^$xSOrkynfSiny6%E0aAMRHdTE!2Qrom0>A@o*A94*Th+L zov?8Hz?m$%XGS@vP1@&FGP#vsFU{mu^HcrsO6xiW-<+K~#;Q)y{jP6^KU((m?b@Fk zCS{u2f~Vx{aFa3zcA>hAAg`sEW{O;;qB7|iUI%Fm+7nnuSB9OPmE}^ey7H5=s?&mp z1VY!cI{GBjIR$HOZm5=yTl|A%9A(@<1ufo_P!sm(UeTd3=7eUIakH$VlD8%FCVW_ z6o#ev-2R~CdyD5JQ;omEW9ZMO0TEG(&!ur|fAjd08y{uAwV3W1_vn+RpFa2`54OJ^ zcmAXPUuL|k{qfol@DB4Y((HeU1t)CK2^Kts{Uzo*VVJcC$NjQz_2ljjHdrRvU)5Vk zKD$*_2lmrX8QH8O6ay+=#Z=VY$hH5{Jtvha{{HPbXWmbrv-*R|pQgV3$??d`ZzsKX z{=3_#_74yK?BWmGA}{`2eK%73R_@0af8IT(Z0>=%t>0TQcN^9GZqwY&KPdT4`kZ?o z=KQ$ot-Fy4-#hlsksqyn=knXff2)0K;EHY3!nf8?$#ZZ1a`hZrd>S_R5QbVF_P8H% z$KvrvJ~ovlWix6!FoMj8m}+UPzEHP@;KG?2TyE!;^8Hfv8lMVYrQ}*LO!n%8$lX@T zZH2jDhom+LN;l$2TL{*yHp9(GfdOk|m?(=&XtqFyX}8z}EvAjyVE^OqExF- zVuuS@nT;`m-}X(MZZI2jnzC1soN=Vu=zBXzG5aa97uOC;g;ycX1=RAC#Zazx)}uzE z<6KRjR_NUexfaM!ggF*K&8qAaob7U;_3$(IBx7J#D!sW&lf z+X7Ku-Gby+?(%R<;I!E|U`ZdDeuppKw@rt!>LbT>?e3nNAOEfaJcS-lenqMrxW-0n zgzEDYQ%{jvDhgw~m`IG-o~yj_MXN<6E^(seAew~xs~IeN0mrk{e;ZQ_tDR-^CCJ6X zbt<@DvS`#CZeUvE=LoC?&>hY=iHYN6 z2E%9q$jzeTM5|nIA-qu%$Lo`hlJjdhatFqXVS#4zgRGD{=B+JR%pV`oPmeSg8t2m>N~7p=@C2ON0jbIu46)f7`nk)u@9mGGO>OM`j{Vt%y1bfjFOI&J$4!PF3*FizaHIIdYb>j=*b3Z_COs|oyaRIkfwJM@% zMibP1gEA8k%oKWN(6JyYx=9(&(?*oIeC@r~twD7SCic(I-VCV$8k!nu6=eSH6z}$5 zKN*wIt%9H>P#S6&EIVQi=vYCC5-;<8A~QfKb>i>;NRqxaLlVel464Y?pqOnGNu{E+ zFfT^_Eu8f(cD>05^5i9|M#;g~)rAn)J2zqXYse(GGbC9GY;+r@HVJRYk=`O}+U1}c zF4ELlg_@W3tTm{5S=-i}-}%cG-!~nEhUVfqu5WH`VA54(HcBoFA)JD!z;m zGGm=C%>R#13}Z>5_h}hv9YRlgF%zcX718$*+_6dEmgzy^W8Yq-Zr56Qu(|aQq zQPq+28+CtrJi53Isxrv~S-$@8F5RC#r$YU1v0?Wp{M{2dzf$)H_*ViR$m;Zm&guT# z59{%0RY3~v7a;f;QXs1WS&M-O>Yp_KG+bC6`TT+UC(S>sJKX1F*%M-@>+;WsK0@F2 z{8zBzb(juz+(p{oyqAzFHXu*-&+pcgH4f|$?=UwKH%jh@l?L@Y zncK#*8bt@47MxBujZ*!vMI)23TWijo*z`C%+f=<`%m12WK!|en@*`x610w|V1qXSq zJ1GXD_E}Vve%*2}=Q@<1WN_Ea>LKmI8#6Tu6b3xTy+28>%8ctn7l63!OHw?mOcpzB zv9y-fkp&=)8s3n{sfWV!dWD^$_#?oJF?#Hz7QzlMZ*7R6UtWUixE7I^DZzdG=6de3 zfc>24!0S+2q75l=N#h}}eS*hBf^l#_9u$k_gx?wu%XXElfgDTkvYFmH=vd`(MSHgrfUi$5_ z**9GA>JpbT(mY&4zkX-ssm!)Ln99Ovh^PF8M}Iv45dgiDJ2XH)Q;PY(r%L(%b+M4zHVDfPe1U}Wq+ zC#(_T8JioZ=T^ZwQ}w(ZTW>8EACETWw=qK++$CSlAbs6xre0C6(61${F+&Mc^9m!s z7D3h#E(#(!S~UyG%%zgAK?vV^{Gu?=#g>H6u@oX=qW@WP%9CrhsD;fo%Wu>o(6~cqv9$Aoitg+42Hbh#mWJ!j(w@2Jd-KdqE7z za^QFltauNaM(w8!Fuja_q6Hhv(v=UfoI;iDn_-_GhWr4zBuqw@=PlT0%RD58nQ>8X{_e;xe{Aqd9G zO9;#Mz2?6sPSw1^uLP^NcfE*XN z?*%JhBgLZ$JV&}~7R)uw-g48G?r(|rVrE_ZkUy#jmeTbhhlP6orIa7fTj$&&JMO*H z{UmraJ!Cc07P=@fP0;RDIK9{QO9;dE#t0BHY&*6-7_s!U=sN}*)cF0InPjg6&y4Df ztV#vNKpjCT6?H7vgAWKxx5x)}P&rdi z35xm)`U|YW&9nZr3xJ-rKJpD#02;GttLGkXgB^OxoLo*c?0g}|4PO%=p&As=*+ntA z6xjjA+}uvrJ^qD4-5N$+(7iyEe`u5g&CrqY8qY^!6oLktHX#T|9fFmR91)75@A1D* zKR<=xK+My@iYF`=eSwEE;(oLZyhi+Zvo{9KKd`{G>U!wh+ks(Wrnb7%wJhK!Ipb9p zuryeGKdfZZHQz3-8rQjFryiN4zRg<~w8t-=s0%oa46@^o%4~cjATZl=Haub8#ss|k zquV|Pdy?1xQ4vWS2d={44{2v%?KFyY$y+6j;9xW-dms(yc0D8|dfa3c9V0m{E6Hgo$BIm~GdD zjD8gCn8(C07GcbneD|pegW*t}{MBKy)q&^GM~sX_0mba1NG_D#3%7H(x*nXOPpMO1 zBBen|@?5$AQt4{=2g47Ctdye!!MgEBdA9>og6rtJew!y82S%ZiNj$fB_XU4JJ?SF}b|1_xT zrCKPezQpNs+UKu70)p|XOAuU1rL%p{kV_o98lAy?%@$09@3rUlf%yG_rtF8nBspk! zkEfW`6j=c=b@)WPNe8W?{(4rTIBIY3`CG8;T|U2lw6<7%n2zh-P5!Ua$)=;sD>dOY zGxDU0T3*iu3cGYVMS%I8+qrpRtvVu?Roq>|MhQkk`khcjCykTGITCOHL^4`u9D!ub z3UNHIRCtzA=LkBSljO^o-)WhteizlV}9iCVe7{@uKIPD)7wBO?-xhg+OcF;w$&Uo$i&!8`H zVy7jN6Osni1Lca>T4j}-B_Nh?P}h2I#8y!m9V^%+Tk+ZnWB z%sy!n=Lm2GdHMJrs5^l)^sf}j9tV}Mx@@+~%)K&J$F1mhR?lI3v3UbD+;pkyx1USi zG@#^{H|tW!!EcRGYB0jo35uzrNCg$u6L!NjfvKXGj?3oOhUf9x`OC&F;e9DM%17pJ zbl_0PCXd(3&xB<33rvq>2}sr)4QNpHx?UBpW{?vtnbSdc(kX$>B#(E;KS8vEK0Fag z53Y&PRN*%2=;ZNs(5NEcpt>sG!zq(>g4pF#-=e^7d8x2E+zNDB0X++B3R}QEzdr5H zvM)^Iybc?qF%y8sDwsw|yZnNlq8jA{K_o(#fSFD(K!XIJ12tr7+(`*;3W^bElqpl1 z-E^9M8OjouRMGm$yM+3Rhdl0vsaEv%^Y?4CWg`$6wP}kTcJ?RXPeEf%9qDh$)Z2`U zeSt&=-qICtH}k9Krp-AoY?Z|Z^!qLyr=eEAmG)uNTX=NYI9y@OgcOQozkBo+&uhXG z=En3gB?wO{azpaM?z=wnisN*8#=9@~edJYP7dl#-Aj}q!GHGM(-~R0#Vi>dbIHjE; zOC8v=0r~7fp3Me|Nu)>u)M0~N(b5>x)z!QPo`C!Y^$vO;xjDT>-o@R@wL`GZge#ZXo>z5f=d!_xv9B?-En6O8hm-eD&7N#pE8ws=F;f?S z>Q{A8O$jUAWpp1;UuCONY?R!dQ5%#QYxBg=+_ zfD2^31CK0Fi8g3t$)p$nN*WclR#3__VqL~m%TlPXE|7!dx+2GWB{0b3h%rg73#71n zCFPO@d;Gpq;AIJrV7+4I0()Y|a-bNS58a;4;qkmdRzMl^J>`#nXqwCIum;3Tb~e^$ z-{6+g*C8|i9zToDpgTQbt#F^qK9@3Xr{^jzWZ@u%s>Ww^a5G&5?q;%wJ&`fc7b8N* zz|D62%MKyEJ%9Zx$6!kS_+@uBS;Q`|;<#%cWUGUa=G*Mx)TPyY06T z>KF2*>0&b*ggCI>FtcX9mK=c4zRqSGch*XdkjpNk?!M~!(;`UNujkFS9Q21D)|ytZ z4qa?^VAEixsx?*gWYSLYa^U^X^Ui~WxP7t?C@RzqA7ykxcsiXY$@Ry~-h=f0Fdc(T zj;O==9^EcQ#xXm)A{K$c$T$q_nC(ZKoG|Z&{NS+vD$}soP?$KdeJ~R$JL<2MSCI;^ z5xRu?I$@3g_~ziym*jqgLEg;HU@+=AR>&~A)4n_AkQQN_+%<-oqVC7c?Mu$181QigI+>%MENq5?*>;BgzvO8Qi#P)-4vi+KM%do z3&PT%U9vj&7nTa!qC;50VZcV{|Jvg5X4^0ihhf%%$IhLmI$%ng%f{_+;N+AdBS7t- z7_c+jsHkqz08FQ!mbf(Y&v@0y%INObwUbULPm8ybwY(Z>Cn*#{b|j~nkBpV&;R(EE zw_0u{y@`Vr={fGn5-oj z0P%xNur4*+9$45f4Z?J{RG&CbLRbS+HamTGgQ_wh3z7)F2}|^umqkD3UKSn<2a9y~ zwPUl&W?!ZoR2>t}25skF5!nGdRzc>OZ8B?8MtoxPv_JpEG{M?TX0k4Yr#0l@>-|m` zkvt!A2!fWm6EqW#d7^>ow&ascaepB_<-NeS7@7U!_U|sE{`0JA z*|=hH#k6I?aU5jGC<)5(T^7(H>vijJ-s#@%*Guo9k9*J0rn)svL)=%{mfLnbi*3>vqCqOANclvyy|?DD6m*6Y=^k%F>VWIM=&N>=ZHYh8;;+d z(QcjU9s91Z92o4|w&o62z!nu!gn^}a2#HgS&!(sFguHz0lOZ`b7)An6{<>#clNle`B?7axXi5H<7v zaa~>lt)ik=F`X*vlMMrG%fa{3(aDMnhJG~Wtak zwoptmMK(}TNP6)|mLo#;cL=()g3bddLxTfyNErt?N+44NC93{;5=>Fb5oC);$2A`b z7KZkm?T?tp!Y~$QOs;nI{K*Dt=NG>|0PCf%BL6*ug|8U_IPR`TxSrL z#*`&w%;|Tkm8L?jD;Au^kY3j=`3>g_idLDPVo;6A*LWXt31T|j5|tLHLQp{~*6WJc zJ_`aP!PgL@VtaUv`Q~?bvV#qXk^NP@h2)PTr;J|p5sCq>qhcy*>*PLJA>B(Sa$4jM zWL@&j&N)5{U8?B>P*uii2b7S?pjV7frSa%K-zqu{^ciaS?GXOLd-0w$MT`LS7=XvE z#3k9MJA4-?2B9MJ^Ltb2^PGXJhOyZrcm+0j^~QZMf?qN87FM6O<2nyhnKu(m$E2aF zMGkCP%#2f5zfmhqS7y<@Zck@Cz!ifeZtWC3SsV^x$Tq0Yr~L#qr0u>wphM8ez&sRu zwmxDW%&vdAe#TT~$YC?w%;*N87H1R5nr0wa0b&DL3kSt&T4nLhx>B9v{?UTGuwE|G zNtOo1^D=_By7q#URi`-7XWjS*;!L_T^se;g8@4fN!}Q+Bfbra{4b!Ae4w(nWa1PAV zZJit6HVj9R#r#%MKHN;ofsM&kqp`D=Vop=!Bo(C>Fw0b)Bei@j2uXE^p`cNYAjbt1 zK;n5laDtoiL|gud|kc{1M6=!42*f%*aF)l7Z)EP18!^d7GUd&tL~hYjO4* zaWCCsg+KE+d-B2FG^qhoCw}i_bT2AkF|ak1ijd6VL;oXr&!3N*!3=}SFXQ!q<|Q~qgdTy9uUJAp0Tpd zOAAeD;v9CAGD8WcTk*Oiyd=EsuBX?qd*2V+Ar=5)*!%2#kR2dC{rHLc4TIJAtVC5# z@*LP|oG`K)hbg9*0`cD{%z{j%*Z5=tkv617-sji3>jm{-XO%mSEc?ij$w;_@g&Z2i z8lOjAW%LoE{z_QRDFET2IL;a$7;7#;NahGBcW;sJ2BuqU0vlk99ejRtx^;$W4jr4^1_yR=%w*T$lxV-CQ|H_ONp_-7F}+Zr=QpmD z?PNw>d>I+62=S28SJK%;Z++bvGI~Tl|mA+#I|<0=SG5!%hWf(br|_ z^x*THEV?u(gD#WkgK?uY46*5Oy@9Z}&#bzTmEjoEUV3Fym|@)c$^P6lQs}_rPOZ_n zbBtojC{hA+O&?r^g`Eaat2^se;(}=i3CtCFF?F1m?Ntg>k?qP-el_PfDBfKYcX9W6 zx6^5y0_W4@8~QecSz3UnCwANk$u^&d;Y-FfsPegG^ljcsmv*|C?u}@YKZ0OwHk2Nw z_$&*)#5u;Tk!loYyi$Uf1!sF&4>k5E8Zy`ncfcwH;mxd2MCGM%uT3zZNIQGRRg%JP zA;59NjSz_*ysVTAnrBK4OOl1j=XVVnmCTr zFI&T13~Lr^e0B-@Wr;rL?5}+{(vG)vN1L|C&ZvCxw0^89*#R4$)`5+R8ES+|P8#S( zpB8Tr#dK&Y4zzW^8WRKHCy^jw#bR{*s*h z)_ASJV>>7rs*z%VRProJXLil#b;TSBoX=F#wJ@+O4vrS|x+eN$b1Eg@2=BQz^Usq4 z?!fs)ftvzq_yc=&P~NIhJoRnmH>fU=7I~VeT82~{)i43yE7Nc{akP`Lcx`YC4JvHE z8fH-gt>Be~VfYcXgWhA-wXNYo zO(VBef8>yICy*v!;$)(ax<=9v?HFdQ4Dv*iLQJ05S@Bb7TnapeDcJ7 zX9&nxX=NvM4VS_q<7&nh;a@vp$6}0_yRbg~V8;{#cDjElT|%}Ew=V6#E)E1{1|eh* z#pF{YkBUn5!>Gzt@qYhKI@M2C1)D7Ck*HU>qy?Ay?}NIjY{7Y-Be-~+u517;8NDFy zk?~3Li*CAAXw*k+=BLuy5L`WIr}H54qizNUjuaQ`gE51>6_w1{5RmF~I5YzUkWogtoh}i!c#p#B?a=>x%ii3uPvTXGO3fsgz14DOQ9g`7;EI{Kb#gDS=uudk!ay;8-= zUi+AO_mfP$>iTbPMxIqQsPIU+cO4`m_i`6SR4WoEKK1Mlt5+R)=d7yFX^qc4Ug9kM zk;0#~!5w{0=iSy#IK#ihsaN&?58UtbTA8*9w`Keg3`7}x@U;k~yow zIa0lQ=M=oIL8YD4Lr_!#ugie!igWHuXV$Bb3Ze?Utax5JNZ9M|sQ@JaSf7~q$jc^9 zl06_~3k&!2GW7`yJ)q0NWuaxX8UsEU1PZ$gD}k$A)(7wNx5*n~^Iv1>+ho(GEv{wd z0Xgn$_4vQ0B~3Ee73J#XN641p=28wkfE_Zj9y=)pvLv#os2J)xT8hbeVMj?h559ly zS!o?mdn5(th2@6laLYVXC)MyPxSwyM|GtPc!D=H6J z>y_&F0-+$Quw*H?Uoe(f^?Y+Tn+JE~iIn$DjmXgD1qa59nK|E1aCpABF3%Nba?@NY zc-wr|gIWj{td6?Dvvg(2T;JL4xOq@~)ffEke=nM;=+Gw-?GrHk)}Vsw=3=OBzT}(@ zCC_`s#k2AwN_bdgUE{i6a-8T(QTk*F;hFUESx7#$d{zxF3ua&TgqBs7#cmK9Gv>7i zAslDJcvJPAvVV}p?EF~A9Z7+GY_K1jNHGZ%iKC)wWQ9T~TXWS3LpMs&L{JT2D3N_6 zX|?~P+Nd*Gz{2Q0d++?h_Xnja11z+`6O+jy2TruRVARnkDW;ktl~7g`h7*-6x-RHm zV3Es1QsR;y(a3xg@=ZuLr&f9@P#@eyLvUBR(Z9e`AEsXl`kF@rn&dfxbY+qpb2Xui zq+U_4=z*1}ILN3&R;)N)Hm8bS<=&v|3$0h|m$uV6qSJFA4t$+{=!Q=zanZQ-x;4tn zd6*48^58!Ue7wLXdBLd>0?GW?Gfcd3%*w>uhhKHUm_0vF|7fdgC!Oq_dqgqa z6zRee+SN>w9J{z9^x*3q6V6D1;2Q#N6$Hy5Q70W^N43(EWb5R+(pEY{d4s!9u!-Q} zuuK4^Uemi|K(Zl@Ga#Rf!26iK?cw|vhLjRJGTk|84<3Zy(+U>2N_SpvIEK zW)K6>C}G;8E0J!LpiQqIu<5pV!!K)t9Wq|ln(fZBLWbk!vm99@e)|UoqY@B1{b%I( za4}@Z{r#lH$O4|Dn6D{PLq+w{>9fm~1G&bnAzS?8U+Z_mflUj-S+&w$w`;*5X`z$2 z(X(8kx@(+BUy!Gl?n{!F@vC6?L;Z_HQSmIpM}R4`n7oXl2iV|qA{#W3#!bU~P=6X# zmvA7&&q$+43uZaZWA_zU;S|BGY`zimU%@t^VC=69zZx= zG+R-uKqo#&@W}Yi>?nIS0ny#8qmzYEWE^dAjavyitydQwv89-1Xu<# zygxy=JOkE6@eHPyt#gY-`lv{ufr5q^7GL(<$HqY}upPZ0g;vDtwrjsMq z$c@PXM-y%@1M6TonupuFUJcEr5PQ9MtVS zk~ybU=gC9*wCZrclHkgK)r?L3ufZ?bWScb>OzyYWu6ZmB6IP;J_%5Hi3^>o~bAKkw z#*s{;FMK1#BvE8774`X5G0uS-RGHjT&i?5axkpKDP(By4GN1BFTu_W6$F|<@OI{JH z9(YMxwmi+Q2OPFoLb-U}JEqWZSkqw!v$x$*=~104*v6D{TjVEXd6G3=g%IJ{>9NT_ zN7N{ff32N1w(+(^!qB!@VV0H4G4xq%jZRM0YxgI`+-ItP5bUqQEs(dcXQA%%;tTL_*zyS;kYyF z)N$g8rfZ@OYdp-Xa&F`l`y*}Mc5d-3B%f)fF`j!_nCQcDmNH^647$-$!!hWVQQ2gL z*Ri3qCZ6#%xEp`WuUJF2J8+!1+{h0qq!{4N+et-L1}vRc@74WA@vJ=)E^+pY8dMvo zC9ZgEpZ~^bJLs*hH>NeH4!*u$RO)}rztF!30)}m#`~0)K zNe=>bR;TBD?Hfl$+Hjluu9l)^kd+&D)7DU8g&Jz<@i$sb3+o)#n3yS*Yk}3x8>$44 zPRRWJTz&Un??o<1pcX*i|8UC>mOuu1pMQho&YW{|`kc;i5BWZf#I1bt(Eqw;47K2| zH&mZDsG8)_g7n!XyeE;T-ulZw)tPgiy#GyPgX+HMxO1PhgqIh&p#S5f$Q0@?e^(di z&RGX1TVZDyKv@MRcG$Uk_3@9!87z)U`sbz~XQ)kesk5w<9)y`Vw6DJq7q#1@&3x<3$jf`oGML!fjJ*7J zo%^p3J{oT@9c};n^54nwagc^R_?z2AF)$9TL;74ji84kfTjZD;a6+;tFyEt=|Cl_K z?{GUZd2wLXa1t>u{}m&=g4IhM;WI{Z8!PMWxW+eUTX@W?9#_qx4=X!K?SxKp%xk+_ zS3qBYy3Qj5=vgt$_tbYiFGqw*Cda(g>%N-^2@ju)N}z-la0iUSNRY_#9#+5|(>d?V zCsPf!!%3vvPcnvE=5gHlJ2_@#==MaB+CgzC*)N(}6o^+q9SWrSwaXXt@aX4@0{{6`+^WK1C!U<=q$*(e z@qX%CObToqKOPa51~6s(mnB+We7SinwfDtqoea3id-S_6NVEe_kTpgaiKm#=6j?El zzLzy|v2?pz0q43WCg|DQhZy#KThjf*-e~88=DUBGC7Q9J%ynRqi=WpSe0fi?ULkaSw(byHAGwUBOMi%hRCc};j{ z1}0~9$n0~B#UCwDY6yW&#>$fb@>`p%iM4C zG|DIcdOLFdI_hR5o&a8jIgfs>jZF1}0ErNfplT@2gkd(v>kHj3ZsFK<;9`RmW9z{# z9f<`z?97J)g95n92APko6a$`Z3Kdna&;oUxE?=#lldeR~iN#ynN`}>-OHhK)L^dg}&dKEVLJ7jxOsQKl|KWtM z;J!5zpF-PJqI&@cBDTAM2m)ub&mGshetk~2LoabEc;()GPVsNz{!+i9X?Oin{i^5# z5wLtjYMgN^(I@`RA)rhFX0!bPYsG4+L~V8Z$s=n_lmkzd%$TSw z`V4@rhJPublQfh3i24~@+{ww zXVHiJ7fTbub41l{)ozR7LP&tdnruva2Dz?X9v5e0VN0B-&N&9k6R*$cpJA7I(}-{} z#ID#L8s;sH=~8yQn8(8~K{{%e;;oCOy0Q+7k*h}T z?lTS9JKSVa6#iMP=aYytqB8O){eC?V7@k!n$#Il<;Ce-{T#yDzeweBm$2ms|y;38UnRs0*Bpw`6 z#sNPelnV8_VU2zb9q-&IRAbSOzJ!cbceCk&VXkF`mSOI<*CF$0dE*BgZv-2rS+c*X zw~%}Xo@SjgvNlI3<}d{sTTxpkmH6X74XW+jeZIJrNJ@asvXi%rb5@n(yByR&%foRj z0){(z84a2Qf~M)82Y6c8JrzcR(=+T=CzG~R_@FXF}W zmhhhXcGA~@2MxLNY=bFQXtBKsF$*^HXc=z2`kJ2B7HL4x`;|Z4O%AgQ<~Xo_e9;JR zUs24L6z~B=E#f4B-g&AhEm*z7ElFNEp>hJwxbfdQQc$ciiw!2Yr*bNj6>-??;88(p(q~8heMOV93d1Nw*qa%quG@9s>q0IJe zm2H!x1$S|QT?ESvwV;b}E~p0X8UY^d0KQ=$Y8GQ=Pt?wDDu30}VbY*x`Gz=6d?4_N zxEq+4al1ja*?$8EO;RddAwH{$VG`W7bMnG6M~&l*b&s3vBio)}XKwPc_PzJtD+?C{ zqC_ymju9Mu9scMOAzDKkMrBH6+|w2!$oQsF*!%M01W&7zIFkIQJ0MnB}Q%7V}~6|O8+Bv@>l~(7B3B~B>hefTVxH07`*hjhGJGxWH}Xe zY%c0sr^#bUnrMTlf_vu;`}h8EJ!b(ThP%z~$Jim_bYb-FC<7uM2K`k=njP3~L>nzj zc2dkY6uCo1E%7-7d9tg+QzF!LlVZKr1;z80_;iJ-4*^-qAw>qHY?pIRle$S)d^na&|l2?}Q90xQ-2MEu0*8Bwf){Bo=u! zcrBK$a>x5Lk`BH`(M-3<)os#jC=V&6aS0qoHmD`RACm*(TzFvyG`Iu=rWz#&UsvO0 z1)N3(coIde5Gh>3TMWgH>QvxHEDS?=O)R0Uab6eH646XIk!%yM$O6zkAC>I8Wdvka zpc~T{$XoEf0d#YZb^U^z7)RQSzV?e0bDkpgR8-;grXXa+g>{${%1zFvC#>fgsFjgS z>4Z|FED2g3o)mrrWb)1s$g_qe-rU)JP7mSl75L||cPivuKc;c zeSzctegcvfgHmobQ_Ln@e2=;{cMS+fLP%y4B-vo7ME#Ah8<g(7UFic%`$@YyFZ1zucE#!96$0B#pss1bHToq@5-c7Dx0|z6p z{h?jlMS;8MOrH*-wR;6Cc7SEQZsM(2*mKrgX5L5)w=OXIB6-Jj&5=#!uLEaLnOTj@ z3){`nt6V39Uk^`$y0v!t+=N1TWk8K{JG}w|G^1-mSm4MSSg^jgqb~ao-!pg|e>j`` zV{(k$ije~&qshn@`I=&CC~}I5+9NvbhglD(Sbj~6gA1llUUsRMZg#G7#!1lGnHoiT zcmiZTl=D{c*83KSdg(MV+MkV*r@nwAa&p?)nJ45~Afka2rZ{fbj0dva{#*Rxcz<8; zd52^@6!8OkZvSse4Va4erQ{ab!fAurcT=sbO zgr!EX1SJbpnV2ZHyX=7FJdeBGGzk1Hu{UNy#z^ztpsJPTad6_BFFQpxi1u)DYMD2975fs=ND_FFdRI7te|D>aHovUYibM{YPOL$U;eWnE+o z$XNCymRFrkJnC;SMAK6LnLsifIG9smWLk6oP4#SLl{e&LZyf6^#0|gyibLxe1Hu+i(AyD~UHy{F+4(k47B~lc8G!T4Gh@z@WCsS$0VCk# zPz*4Sq*GBCRnw(PAs6zOz88cm-A*G7eXm<9{ngLwrAhL29~lxFioAA)=z?}w4%-T| zYWBrOU%a@bi9mY4MU2v4&D%h>8WA02zu3-Q}wQk3kmhx7Ps`ANg8hc-3@D1TG!K@LEFWB}>jp5n+^UwK#|5F}@CD^I5cX)`vD)Iyza?p18xj zDO%7j#C1q4ox8U}usStsNAZ)X=t4JPKq58t~&)_!YD43NGL@<_H(Oe#e-g9N(s zc5#fL+i&$$NNhq~sa9pn>-)SlibA?1C6^q|76Gc$BPk&p_61vTkR3KGXwXKNjEgT3hZ(HOPxj}gk-~AL z*64FQMlod+DWRfp@b7W04JvU#>y4SH;0Sa<_G{6s)8ZrJHJ+OS%D^}LCPV{Lh9{YQ zzS%##SAq_T-cljOtnW}(9A6&#il|fWvF4yS+r#ZI~^?u(9 z-X?+`Wd=w%TlW)YHoz8`zn>%{ugf*D01gWLBn+J0)iJ9?+!ODDAqADkO zME)pPz(M`w%Ucz;3xhOMq$AXOZ~TjI(@co zrC6iDBUo2l8E}edCDz&&cs;CpWLR5bdeGAMZvHy+Ju=KN@3<*O>o5O4-ZZVwOg`Gc zSftj{4+%&VN|Ic%Byk*=rPc{Cn6X@1F*u@a{?*y&3@c>V=$V!@!wMOWJMd7yRIV~j zxGwLF*y6dBTc$itKYp`@58NK%4XQFppX{nQk#j8a zYmp61nPGJ2k~@k!b3rmur^4YPMYLd)2;?p@ZvZ`U*ttzmw$mNEwzGd<(Ai*`6WR$!mpIOMvc5@;00UU z(E~S((qC3qgWCTNJ?&K&gx3P|U_7sfs1rrSP?S+V1!L_{!%iQbc4IoQ^=$FP*!n1a z|JchGLqq3Ij1n&Jcd?m-jLDnyw>Z<(YBmix@GQnmBKCf68{Pf7cG6+r4vAJ0&$}Wn z1uph_&K>y{P6d?mXDWBkUd(HhL!~`k2q^|zc!jWRv)41(=dNd*s5MXvONG_`6<&6P z8ars2y07+w+S-{*e>mMRrGaK>b{(fs8g#q6iZUMdQ5LzD?&HK44!#S?tf zW_}y6#o>A{`Vw0_@zfdmt`x(oYWx@Gk(?x3C!@K!?Vjk<19KW3?O+nOM3Lj3Nv~$~ z7iH0@q8>?y$6jDxK)wuhrLC%0t%HLpJaFGO5MgC$#yWqlxZ-C3Pu1MD#U$NBoGvYsSd0Y*DG2St+IIMPji&e;a5DHLmFSJc&%2ngU<0@$vMiu3iX6NB*&*%)cD$ZS6lY| z%espdHeS}44bQW}1{Lvte!bn*U{Z?TdAE+N97jNia8TAy3dL-s;6FC%y664SBKHC> z`1G{Et<=If8i2J|$GABHTtSN#biz9QPJ4Ye+WqwA;0^>r|SRntN;36fBsML z5{g+wk?7$Jw(gh>-}4KbV$}(bn`3(Nq$L0UviBu$O=Vf$KJh)08$&h%c`2||A_!v1 zVrW4Nt7@i~>Ak0?*XjBCt8STZ#-5(;>X}US^iqwe<%TQ@DyRWu6GRYjDO5H$P)h|1 z6-87a5iAQuPzq7uJ12=sBEh_nFtK{dU&+f`ZZP+sbI(2JfBr{pLdU6?4>poA7hZ6n zaD4=9k5Eb=$R5Tzo~w<^OvIrw-i9 zRH^eYQ{=QH&x4GS7slNH&V9CVp%a$ze14A>D?B#%;NQzR2N&fWBW z-z5G`UKQN|v^nI*rC=|r5^t?v*_{2p$M|KSx^Mmmt1%MzK!|}5ic_r!$8en4vf_*x ze>NsJGxFNurtkjH1eCu&Lw$vO;=(|=X9koGN_m4KtyGMjer6L^1RoAE*@QyK1Zq_W zcz4BnXY{ds?BV&xgW4clyx$k=F(9UrL4T|o^!z+>N3_xDF|18``T@{A_xtK)$xsla z!MGig;XddSt_8c2A*tjQi(xz7EzhOV3~8Y$S|6K5Ht>_Z_s+mwc!y8rYk71sIjq_X zaXUR!3Kd9!nZx$_BruI$-88O@NDIZ1p?>g$wgO#y44`s1FlI;_=Zj~cdf|W@%2fIz zfvpwdW|q=Kpr zULDjRui-x;SA#Z0pf=Q9`V86e;~jJ@KaqFLE7zwvI>k@NpsYKHzJrjT8N6D5Bul(23 zl<9KK7TN94K{{tqEz}%Vg;epdS4MQX3lU9r37UjEfs_G zrJK_A(U54oCOtl_hDLdI?Dng{V6#?*jmNm3q?8|%a7 zH1Xyu%2>mW`e)JiJCVLwwf__P*S490aZj%u z18_scuHUb7w=8OR*_@%3GIjJl*FaL)C}vPO9fGxKq+NI}=po1;q_VJw4Zg{0KJ~Z| z)v2h6(v zOpQN5h}id>F9hxnMdC^0Pt(aHl0+a;PD7h#9g7;rS{2$N=khDLz`@)ZfmcpMgmtr$ zf1zjde3OIvi%%BpCZ#Uypq?{xP-`hA$Xp+yVopVN)BDt2j9$?y-4lpr>3u(+ zwmdjibeVmA_dU2b=h?mEHZ|kI%L@Q{eqy{(vW}NgyjSYN$)F#NXBl!~0pR&By_ znb@mDx7i6okuA7+?qT531b_zKVt1vKe%s`sKIUcpj8wU>hkDh_Lp@6=PgA6xidi|O zo^0{};pwmsr)X85dK{A?^#q%9XNq)+G%^s;8_`jHZiH7N7o`t_N?+-YHMd+d!~dtOh(!nd*0EuXFM*~bDUJb^u8^l8oj zLE8u;tqNV@>Y%+dG-p%=pjw-cDzvF#&@9!UR%@iHEbVdci@=)bCE2{nnYVnO_%;fc zXbL3_UVDxGM3fuoU?t zx`@I*V>?gdgVsoxaXuasW6Fmf#g^G0E*8jzmnbW_qV9ke<$rznR$J&YXzbDR_I2st z!pG7)UOKekC}0d1+ZV_GQyrbGLxWo71dImx!_pL3&;7l zc-na{t$V%mfABXCe_>jgR&I}*Nv@A3+~vaS&l9ut=PsrEj3T$Fm_ByBvPp&M;mk1O zLW9K_cVJZl^7Q~&58ZjPW*Qd~5Zaxd%>!LrC=*(;)3YjYNdgG&z-^gfy^-+(Lv_1B zHy5t2^-2usA>B~nj8AFw(+5ZvX^TL61zB>#fl5^scx+0`&9| zYnc<-!;Tvap3Rce53IvrlB38jyL5B(-H0>flB}4Q%s)X&css-QD)O09=k&}SIq%${^V}Yc9K#vM z)}iwsRP$%FY0c{Yd%x2p&V|E#nP%(L21=Pok$5VmSDr2#@c7I_gJ$G@gkI4+9ZQGW z)MM{Lobo~2R+#UMw;pp$>i`(Vc1-%zqqsq60>xXGHXkKxxoxps*!kUQ2BCCHxtSs< zR1C_hplfsV^&Nhi&d^4Am2l8gAC6V;$WZvmU8_12xSKea#$rB$9BqmD!%lovBi?^Y zW&(%!p7`HMk_%f5=nXcaA~lmzZl_2(6;lsgR@#M4DpYRjV>Jnq)j>MNBM>RRFY6Zf zfV9D#P;3-cBij_VFVNsQq8Bm&=7i&)=EZtHraTFJ}A9-eg z+CnM8;BUan)l_!n?2RFs+F1oELuH1Zj$Mc`I7~?!`i#Yj@+Xj)%@Ih>N08|q&VB82 zM5xJnWclT6BE{U678iC{KQ@C#6{Q3n?2o`Zp;NwbUHaz^_87F<1*7xED&KU!bLza{ zE>QK)mNrKxgzOi5PU`4YQ_=(Wywybt79CS(2Ctri3Qy~my>vZ1zKrT%|A4Zlr3&M% zW%PY|^^`71TW$|b;Mak=I#iAY7$0?ro#F_f=f3P1M~*<*-%TkEw(Q#CvJ8DI9a;?b z01X4KD#LEO!y#nBw1TqEHUH$uSdiw)pHr?V6b*WyWcq~p9NcHz5%c--F}STgU01ez z*|MgG%i80@;W_YKMmRHxlro+oYoJhU)-LaJY8|8L3ctCaD?BcE!VqIVcedci{E3Nq z^-lBeyDhoauT&(&3jcbCI-j{DULUq1EWx(`SlBDVn&f)AB5Ett03Mg^kmaefc$@q$ zPFo)a-P+klG~Nfz1bbUl;GQ%nZ8hs$(VBch0v#xMxuf zy_WR&Gzeow{Y)Ice_kJpf7j9GS9g;#2JiTs)H*tfN!d{J#u+;5J4py*?@eoZ0mWV57%SD-#0Sj4{v5-Yvn*6tSM(8e>X zrwl1R@hM=FBjZTr^a1a3M~bA3GTrv!%3xnc8O?F0DklG=+%ksp1wo3Hh|3;8${#w| zbFc?$j;@Ss5cbYU_rAgIh)xpq@yeKW)3*3Fs;(@8%?#7ybI!LuvOe$QLv(!lE zVo_Xp+havhy;1(OZQ}Adpo6hTfK4N=JJjhcf5yd-)8^O(y!$L8yAk$ zTJeHWvJd$Cz;D9c(gUC!V2{mgUR{9ZDV{WsW%=}F26ad~A*{AlRH8Wh#sqAt?FGqr zLGZHYx4n}cV%Zz&#SOFzLjlB$M+8n%C}lE55~!F`X6=GrQay9nO;5K&6L5^h6)Vnr z;s1{O^1tl!HC(rK@@*Q-fj3!ShWe(L-2Lmy@|-3|0vqLNP;Dejv#8nE*2}VsdQ-jZ+vV!;=4|mb_`o^8D=9`K{z2w=9nf+mdE8TXKq0 z)=}gba`e&pxJY_^a6!?k%AiHi`%ipx{f-9}K~v{?)%)$jU7|9842qBHSmSX5P!v&j zU#r>^7_Y1ppypTqyfifm^S66`6#PC`dLZvsgK8^CoHRa{EA1uk53;pPRS0rk61^@n z@2?W9;D4&#FFz2NXrJrC2~1`ViR~|*0))%Yf2eE|+`-0ro$0F4%d`{ALV9Jj18bXl-isT({ed??qda>Dyag$I7; zbJY!WXDCSRsP?Hf*ObQ=!LAufe^S{@U{*tGh6$n-*Z~l3Gi~SHoB%R;@!l^rTgFwm z?08&wqiiJ#b6%XyLnj4yK3dgE-zz*2dwUR?$`<%sW^t2){TPPbR?c1)wJ)%4W;(PL zE00KJuPvydSNe9anl^HqJr0qYHgdvw1UNguWwGkmiJq4{mRcJ%IL+kA=of`vCtF>Z z?N(vto9v>Lz;?`~Vk%U{;#^wCRzP>Tig2Au4+OUc`5>=pdU>>FcFe7LB{`dAxA5ustJQjt6X$ zCLKxnB-OIGkjv`i!YM>6Wrqb{HL}~9qwjklwQUJ@RoyF154cCiVSh2YO1xwugE5?t zF}}&>^mI<2CWvoZh~}hyubzD9!V6K3*+R65Qf{P35*34n;1P40B>MFNu`PQ>o1jsZ z0lVYUw?C1Th;p4QWB#r6zd2kkK?^+ z=yEz`!H4tDiMIM(3BDSH1-{MEH>6j1eWCdJ7;jhP7Lu@Vz~jl>OT@`bhwUb0go|J} zbgBiJ#La{}&MW=xznj*UkAChcC%0XAZCNvUWHRRwr5vEh11cstWDWlkMAFvqWBEgh z!YwOlS1@ z89L>Wph1Y$C4^)OFh!g;>y+ot@Y)$jvCm`@fD_rr_Da|AOL*A?3d2;%5XJkC&O1Wa2c00;z9pMiuV@Ir?A{z*3q9@dJ`4nBK>HRHzAd7+P17j{ zA~F`{Lw9*pA}jM*!@uW&qhpWZ4z@Y^R$wgu&|*A(4r%+*XNg9KynY60Zs?o4=tn!= zHKFhEonjfe#w}Uk!ana~Ge77KrMyj%E)Z1U9iHDy7xV5(Ysd=!4*6cF0%;falQ_@& zvtmV6;z|LY#iP71Js{~io5C9)gS-U(N78g=S#-KzZuo=9N3doAFU_z3L{&HYD+viPl1_@ZQvW@U3X4ud{~&3%tr0yt`a+NWI6e zJn)A2c4#egPLwYga$C)mM;r=VIjx`6ky_@&j0AoygXgQ5Bcy@`W|sJzZ&6sKvKC&~ z1SMFDO8?Kp^BE$|4nkrMZkl2OT(tf@4Ozwwa4x({&oBdAGNnwQ$T}(pvr)#b&e-b^ zM3lWISj^?w=WiP(p4~oS2T$^Em4uiuu_m>5583U)OCYFQj^Hy?Qc8&Y9-w0G3mc#z zbN-x88vDxS@oqt0aRt*Vz}}as5M#+xYgNdG!pfs&l@_`<+?tm<^@dLcT_kIqWfX8J z;nl!$StM&?2S|B@j!h%Sq#!0X^F~-X^q%U208Q%DLQ#SA6#w}$G`>l)4G|m`sPX+e z&z|PAh)w!=_tz`FXTnb=HTVs3b_)5-Y?`i7$`*<=Q!#(R<-AH<9|@`jp(}mk`Bmc0 zvzw!L0x`drkFU`P*D^zH$&xeEM!xp0Qp)1LUIdyG8S%fIEDEydA?+#r50 zy8YU{;7!w4zP?YA5fDpN(MjYIvvF#BDDYkk@6{@ zvHxr-UhR_GHq#ED$Cpp}v$-1UfYUs>t}ju$GP7ztb!{%ZXj!Ri>*KA7JSNqFrs4hh zor)y)VK>yWe>xxj6vlqnQRreao4`QT{@ z!GXD_)M>M}`&ILE{IAZG0Io_L?2xW3(yFw6$C+eFwr6v6m%M;U2*J+G>jZGc?+MgD zYW?;EHbtXLRKT18-BMH;YYROXuzNv=5028pnvc=M;Momc>)i+50984ra{7g6V_oDm zwq3AhAxdY~K$dhLDHh{$X+HqDpy?^9IQ-;*AM;!PujfW$$(V3iN{tl`M9;KyAoG9C zKWR=GI}myq;zarCdmwAr9F12xg=NsiYS5#OjuTz;KS6HIO9U4U34KKox*20rq%d9BQL)ly0Q1yFn&z+@ z?yrnpy`ilD?6X<{YOwdBH+eQIhuvzK_1>`ymo9CQ9i;C^XjLd=H|&N1z#MfJIpBk` z%foJHoKhqeQ7ur0fShz36L~t=_NbFi3*4$oeq;0OI3V+!{8rDBVYlRnDjMGl6(jI@ zJF_3-g`X|9$R0$_iS4NLH?3yBDLIx%vR!yM@_;W`@@Ov`A;k8s=h#T-XP)!ZAKQt4}ewuut&> zqRGn^7DenMoA^rx7Y@0do~09>AbrwPuVttbLi%~@y_=)6X6_YVS@g&E54_bK0E?jb z5APpZj7R09@U1e^9pHFa;oLd^T4O@Q-iu34yxIC4lZp9l=WlP4kH0Vu^QUG8=_;iJ z5Bee%)4}Za>JenlJT5yY*(Jr*NC$nifRN4{@_#H%WA4vCqso94YIRWBYdPU9$}EZT z-l~uuX!E#%tak4eHhbjKcV1gQznj*QM7CFG95a;~a;x{P4C&=3@yqF6{%UuizI$Nu z;-2v4T=RQ3!o9|)(x?-pK%F`B=_rX*rNDS4=_SnsM$EYb;AP7q_UxV!Z$0LryZeG) z|8>P~GWma89$@r~3PncW$T=2e%+K96gY%LhoN{?nQ9_s{t=I~C7x=n#i|mBECRg38 z)PhjhCE)8CkAYIz#ndtz0)Y5}EkZKr%i<$}(2>=%dMc1D(R>FL7!s6Jl1aKHmf z6rhyldgOUeO`IrU?irA8SehMn0%$l7=3&MP9{~b+qr6a(Pajn84#&-~g8{-72G4wj z7Y;_7O#9DqvUHOob9bMx6w&0eIdWZ7Nvue2p7MU^dyH@BSW?QgD31hI@-@}sVzHsP zHlNn2v7Qd?3o7MhO1fzW<7V`kwFgb3kL1|1>C^igEKB=cHsotXl6fcH&EAHp`~oSe zxI@VsPIUp)pTvo(=me&oZgjkL7;_dlP4<{0Irbtq)A9Wl?T=@h)*Cm8`T)rtt;vxK zM^BELtwH-JB{Ui-p<=3;1Cb3uW$HnXf$$sN8FVrI;+dxCY-<1Ce*QNZUrzbf+II@y z*-SM@Ux=<&ZuVFy*&yfJ~SE>RG$+5_2Ki!RCY$mR?wT zMxK}`Ed74*qPspO+%&zreZJIB za>y-ndWA=c;w}gdb-evKh?^Z!^zpQ+M86#Mk)Y=29JP+EeyeTq=OiyUUyw)NoSRRt z8A11T>{P*M<3o|df1PIo#`O4o_a97H`DN)p50XX~_G#~$&3+rDyhf206a?8vQg}Lp zBAF(Ytq#iLU2xBmvq|T#;&o$ZkXLRDp_^X;PwuBqD<9#r4Gk5WT9{zN1ZHH6v{g*W+gZ8TEkMI(A$c{dblwu*LEopjY-5F_rlq0*sW}ws2L)S(K9Q}vd|D^hT+#AA}EQFL(Zokdwe46lTH*oo=>WcKzaVa z1RYriO3Z`az2l+Wx$wT(3gr&rf|BsCe>x!FMVCSp&e)U!Yoppo0TVy5Qh7FF#;~$X z5L^szli|6)F5Kj@h}{=wlRYl%vYa(@S!yU{B}EQWF%7~VA1$w*-mB;ah3@9)j@R$c z+DZ5LK-~Z@BcNUQAfka^$?t`Nw&v)SUTgTq`~Fa0Ga)8CMo~IY@Hj;EYVR)2^oS)D(1N2l6uwrgx8+_Y>ICE$?$jY{Pgx)2Y!n5 z+f*R?WZW|zI>lk(6X_)=cg>J|=y71-pn=MJmS}x4 zq5iE#Q5HD(*BDdbwiYS2{cWApWO5XL`I{z^I0a~oBa1RJDCIVaqyhCFl6U3rpsMLk zs2fb1nLoEGa1hGc6a6swQ7n#A9%I@eOP4|CisF>TGgB8%%ud6O@1fn_EiMmon+W2n zU(x2tCI9-@r5vBhG`MTkhGdeCEfne49-&4D>Js@(^~?lD(+)ajc(G5BEYj4_4Z;Rt zyJw5+zQ+?#&vP>5?ZCQe?csu4?)MgzA1e8-$&9T2&EW^+%ooOU0VRhKHQ23`@(M*R zQ8D@SIf&&pN3VwVSt#?MRmJnKC^Mva>H~B2vThptBw>B+6KQT(PI#6i(XV=DmAFYh z1XaN4U^TJ+TE`agR!xZ&Jqj9TkIdakXL-~J2iatZ)jXh2sk1!`X0`aQp3=``GHc{X z{50ib>6R!xlslnpNCWS(xNCkAUpu!-Toi%i@w0*F1IrY7Vfa*{UsH5FU8YF%y9e)Q zdsfPRd~tDcM%-&7)E;XjC;qPqU-Osg=aNK<4VAPVK6rdT0s=Z~SWF>k20R)=Fj^cZIupFf z9iA(X*btmg>t!9@n#?fFHDt|f2PYff2Ga?JqJ!b7Q#Iw#%(jBQF3sd`_{oFC#>;o) z$!`sMB=R))gqh8oa0HD2ljY$^aB*6VCodfO$2H$K;Z5TE>q(@+h4FUB3~x6mWh+Il z7^!vZm6^~K^3%|=z^cHtK)ph-qRQ#*p>=d$(7m8GD2CIj(v-Cj1Ii|7U(Whn1R0v< zXiP$3Gsy(Lq2F+D0@EcQ1j~~eW*p&$bbl0%dL#&C=`V(Wa;NlAU=qKZ#?g)kl`FQx zUhH^KDqAB%L8kNK26+a(9aQm8E3fd5@pTNo%N=CTs^LPa_ znxeH|sfDZhaNGSHYjbHUiY!-}m%g_4hBlt(gd)mA;jZDCRxh^zw>YwL3h-1$y0NL0 zasx#YshC`Ku1}$)h20aDPv_Fx+&`TLenG5gp8!+J6aF=}{o}Xsb92yboIZZn4bPVi zUMU(+-&)uGmk}#{Yjx~RUhKkMfh{3dlznWWWWR562>8kyz1oF+5h-e85q4%+Dq9w? zF$7#_><*S1*2l}5nX4}FH&$)iEV z3)+g}?m8cQ;v(BiS?&3_Uwp=PV7&OOGtXQcO!>BD|7(}^XRUO!Hu5!%b(N^~p%+#P zkWYz|jzuI`OoFkzM|8RPD6o9<=>&;0*NkWLW&5N*Up$^Y?8F03-qPgD^Y6ZADPI0c z@$dGUdDeN95?t*}Dh8V^VZ^OZaotaI;q3}lx-3zE@^+dw(6Fxx(W($WkKfYVR~YD3 z0?jdk2TdwXJ-HvTiaG1NCQLB-_JHR(jGVTTj3H=#q-X!TmU-qQ4WMY@kfT%5n)werQrUoT4s}fJRccZT9r{@ z)QQ$;d@ypJ5eoRrx8AUZ&#Y%1MXKx}w;&w2GS04Rz+gmH&x&T_^*o9K78syyx zfIiq>pS@Wpd3w*U)_(2gH)_7I=AB<%h}NolXJmrDy`i%@kSRWUv(O*UXjP5A`}`da zpYeg_xfOCWe9n7r^{@=dasiDCyRlXRxR{f=$=mMpB(#))mMS{-G|8Gd99TKs;dB1d zxv~)yFCE9xH~h*y?D$_7DD)5nsb0{aC>N&%HOlj#`o2(flh>x6cud&nr8s)ZUp`*e zg%2ltnFc@+uWfn!WR19ivR0>Y1Gu?qf2SRn|l%hg-D!bLM zMVSFDKCH@-VK?QN)cLvt4W z*aQ>pZP62Q!G&ScX9g2JrEI6jb?9Nj4tgH;$_v{m>-Igs+?PQIQK&}KBuI=q;%@e` ztVMQZk*1BLvez>N4p4v?}LK->VYuK$qdYi@W z+VIB7o;@eVfFdWHO-lIwh4h$fs-=`w6gh;&*v4obu4q^ltwA*@ zv@ZsB)qdY3rh&$+&eA0Q?r@~Cw8++Z8XVvKUQ3@UPaWmy!K6W&2y36RX67?B^bo&B zv2u=PoA+Z$f&^b43M_@~b(Oe-UW&J)$SX9g9mle0E8^_2Gvf*TpR3OZ4>n<_>)W@> zNP!ERmO3+3R8Y$O6e-6fx1N5)OM&)prxZJ)HJB8PBe4rVR;7D?5}BvQvio7TUfw;Z z;$H#mx%c0H|FIWx^|t$zN8qNYHKc*3xk?H}+x@b`jCxaW&!_7Bat+EgBl+dRtRz0B z414H80bCmg5DbRFX=t#|>^yslll__W-(O8QVOd4{VrbWebA?uFWlQ*br8gJs7HpF5 zl=SiHr0KFt(jBrJ(j%lX1Pb&e7bdI(-#9TG(DH01C*r_5T0(#DD>o05RjK&$*Uyu6 zE^Ji_%&f{*O1XuCPdsL)C+NBZzcL5xO7((VbpzcDdn6>mb<#=vePr(p)boViPqOh^ z1Ka0{0}^l{v3=li-f0~PqZG(a-c*;f{Wm6b{I>BIbI4~djE=QtZAu?g%7+x`r(%%* zuT}N<^eF0>cE5a}Y-6bi=##a9t|g`pE9mrv*h^%;tTQx|DPnTe%Y!#F`SfznVYhZ+ zqr3>*js{@@)2^z4ofPQ4;7xHrw=DLK73mmFycd$mcZte;$^w$00o(;;z3`r=PFW~9 z%Ih{bKf`WSA!)NZ)i~Cnz}>LQ#Z5gYi52OETcR{Od^4n`L*pXX$GgevV|(0Ddsa_Z z3L0ePGc+~AI@od*F{^y@)u~?FWm;G)HIGOpA6qY9CIcRtB3>_zZ(R*N4gQ%U{Ot2( z?0AY3{3c0Wzo@Yk#d~qta$z!*6`?#dM=LyT2{rj1d9%>J$@j>R>{Im68t_=&Ph}Gp zCQCY@@dSuKOpL{{11{sVL$*C`9WGycOV>Z1o+mEs>saY|Qs$G)Kjhyo+%f&3Z?9|ieC+#5>*wF47`5a32CAE(i(ZXusIrw{c#(AFx=Sw81I6o zByh)qJ%M;v4ZVU%^usy?Y}?c?ehf_w@U9axPC&Ilw)Bwy8n1r#1Zn=;)Svv<-XCrJ zs(sM318Q9DjBVJmxa-I!-wejI)cxjUe?3|Lim5C@EO|t5c_XDvqR4tGCN->fMgimx zZ@;!fwt~?M+l5=yr^P+uOh5bP&*Ed_^wP$;=QO4xZnkb_c6)W+_>l<}SAJUlh%|8v z+_q72XNmS%yLz+)(D z+o~$}jq|(ae~g#qp$FS^JgAI0!mp7n1KZP1KZGi`OnwHvOmtMORi_0t2=Pv>Dw+DP ztiQ`r56>TDv(!7htPM9Ns%-=ySA%2I{bK=jrMBVZ%lGFl?C4n8pBpn1Si+df8e@Jb zeulmY$Zy!HIupEh!I<+k)@NqVH7@gS&-nH}W8DB9*nK>5uE_ws)m9Ks3b`2|7mgf5 zAIT95)S&%%hPH4t4`x?f1ImdrZ0#no*)GBZz}yP`Xg0r^gd0 zem>b;U}cl7fSaC=ZvT;Bdj6t0Ch+{>kGjKTrwec2Pnf~sBT5OCBxO`g5wD!ip<_k) z^fCU(V-yl<^1{viZlWXIp}kP}e@V7R6f4>+ImRni9FNF@{ZkPymzN2SYz~zF7r>tW z7=I1BX(klivRfsIl6oNV?eugougZ~Nuz%`beA2p&871;Vh3uRC(^ow%cp<7d+#qs> zp};s1q{fj7+```T)NJz1pR>6jGTm~~;lkZ?;VqMuCF;UguEP>Dq$p!>x3bf-K2WR5RHmsasP&;~ubqNjOoOmn zT<+7U-aNY=^z(Gk1bMRr8Xrq<`Bq7a=irq}!T!j@(%oL0A+FpkyDYAQkaM@VWMLWm zgpb!9>Lzmp+L$iOsED!ir1!gvH;GLC%6sKge^26F*k36$^H;V}$}~i7OdqR*;#J(; z3`8J_uS2{OO7QPZMG;qH3*q#D96HwryI?zXQpfttxlZ_4V>$8~C(A+rfVa))xcXvnQd zRu;9zcgXD$>6M-g9&(G1(EIKZp#bPUh4I;pfCNywUje+rr|(q@4H0jL?w6h8W1g&# zw}R0=7a8V?yfJN!11>&AFE``~woO?$(}cVg?@fy*8KVVOB+9i~-h$014 zjGkV}Z}dvy8?)mE=4b|eI7m+q$OH;LBT$|Ua)&ctYE3}KgXFj2czgI2*&->I*OwH^Y$XkcB5RpAQ#4x zm9D8J-WQ^73QB#>N;M_k=Yj^=PdzR~b%tk1hTTxxc+9bz+mSapP#*h#7qVMd!C?S|*>Y zIB@vej{CZe{!6}i*2a(X#Rr{y#`@QClu(E3e$)Q))s$Z^G_5-R%h&B9AC0!^xbS}I zqS>lbPbrU6Egv5Z295`ii^`q*mmLGxqH+)g^u)#b&AT64M3&I zkm%+Q`xpAxGextTqMQFV@hAP?x)7ZpfsRekDSe;x8043Z@$=MWpkcR8CXc z52<#0)%rF?@1UxFw&k7Ei$D1XbYL(KWq1d}_~ZJr*ob?(+?DZsHLT-)d)Hfimg>@6 zmLnHVBUlm1Mm@|%IS6=qf|jRNwM?{Mkjf(Kd^ME|RN+Q>8;fZJq*xhf=jVZjGcHir zIB|Ai$HuY8bK6JijEhrD#wJ?-o`x*@!X%n9%*;?SrA(m6Ix6M>uP7p!*DBWB{ZfJ? z+w*`7dbR6iDPNfg6qwJHjhHZhXaZh+;}hZDS3S(z26;YX1zIYbOK+NXiRc6Rczb8u zp9MRYHgek&y-uggg^!#p8)w_JdW?M1xyRO5kI}rQs{N&>CqLa1uL?d68YwQToKLo! zS%5@J8BdWlRLp+G4(0x+vr?!qdg31+&=Arvfei5xlW7YgM%+GOhnx^G`O$}SDl9iK zE}Imxvb)+C(jKZ~u`o@?{wa200=(=c$gb3fPx!K63o6h^Oc)$YZ^Q`|lYXT8QopCk zPW+;kEFsBX7{8*#%&KHjN-agUqP)C5e7(W0qz2z)H3h1nAZ!uZ;Zwk5LTw_H1cXBM zK_^`ibrb6k!a8ZJSDS!Rej7y`Uf8xc9boj&H-}qxHgQ?YvBzxIcTh@jFSk)KophD3 z8d~mD(TQv~MAUobTEF#S@q&B^)hF;1fWw6GFsyG}Gj)L6WRE~8exYbWcn7u`67z|7 z`mB$kFgBXQA^OqrAGnzeNzuc9{UeDTEgIsw(Hv0v7#V|4qLk|?vX+XeSJldp*_Hs^ z9CPVj;qDm|>Wx01W?NwL{O*Z6vcBHr?rZ(b?;_-yWNfsDCTvHI%Sn2tEG4w?24xAT^~HZFlM z7q%}}s$(->OPaG2=#PmW1?qI~dmdWVpRXt}%~C`E@e0u5w$0L@B==D-G(QvM?E^JU_*+1bSf|msK+j{N`xnAl1`rphwJBzsi}Ff|iiVnHd4q zK9w`qD=~ULA=A!%(kEtbW8GMc;)YP)7k}qFmIlPiHxEs4qp&;zh{`H0biy2V!&nn4 z7O$LrN@nO2wJWeV9E;W`WC9TXTqoh>-x;T4XvX7SgdhVZ=t&fUI)5R@Dius|p3HIjXsH1-n(}{OONC&9bid(b0uViej(PLHjq82QXZ9CJd&9ykEQakFm2NSuH% z>0#TvY#);|+w|`Kr6kRTomnUz9uZ0_pp>A!lTF2J^1nE3eb_elLQsCn1i{ElhSe2TCxVBlq~uL=Ok7yl6#i zm~(q?;e-|H$14M;T8gv3DAnY`PKOm;{tll(&l_PiP)ZCW0i)Kjj;)s-cPb=CpE4VP zGWx?#JZl{&3WnbDb&~-p_kR3yQa6R@&Az9XC}k5x8mX8+47s7wF!tcalHJt;Am!mF z52CPtS|~uyL|p{6>%K%*$ul`ecm~>~h_&>*=_Ed-G=|WL=nhWfHmkg<%zZ zZuSMaMJaDmqyt4RuQJa?E;ZSnrA&vnFlqYnZyIuLKqrJ=J`tO#Xs zF^*EAxDYkSLf#p^H{J3e|;A3^S%6<32Xkpu3SR4ykc3#{bq>TNhv|CX9pE? zZgCgMqmzi{gD!$9+?qdkk@IvWKb0-`b|MeYR`HrthUR)-gK%FTh;YCUw8Kp*WVp10 zH5~NR3om=`b`Ckp5r7zdO2>EBx()qHd+VnX6I8NSE~ka%|Y#2s*g zgzM(*zxwgJ&98cA*2<8BiiD8c!B6CmeG`~o$m?{=o1-6*3`v_ZE%2y1KMM3oc|CNy zXNKytH=oG6JV3-}U7qaKbLNx7y#j2#R|;tp*DbE0ArR~{;E|)wpg&e6d$$YM{^Y{XcfNCL@ix@~8Quc| zEPNekA~?}r+5@fUNP9U%?7idm2YbA~Yw|%>|K{)ma>j+DcDK!(qgG0Jg(8=zm|dc+ z3qj*=yH8%&kfJ;ye@;Ghl`0e!it@rx?MN8L=_42K*)+yPMY?jy=<=#6^rQeD~~58o~>(mlx^SQs#ua1~b-3!5;G8 zJ7Z|}Ri+%^quKM$-WgCo2pc0EY-S7;sr|m^e4&*VI%XMk9+U-i!-lC@)hb>iw~5I+ zKEyo74%efXA&HYQnl!ic=rv1a1D8z~SkWq2GDzPIzZ%p|_s{E{^@yDEER`fM#o;~l z9!M`8rW5_}70BF*6QBcD3GXcIj~Zk*1yC%?))9OrQ@$Qr0r!)3Pp4U!@uKBrvu^Ko zPH3UN({<%d%X}o4G_?y?w_C|xg21+KhkQtp&ooDG_3NeYyjJx3QF@0O&t8%aNV}jY z85%#pE2DOMtZ2}qUU41HAyW+R_*k_}ov;wQ+xJ2a3y+)vvhe~8&5*qy8tb|0S@(W5 zWE#2k+f%+y9=Q?F)*7*iUPmccQ)DF-gUZ6jK4T7O^)%M+G{&`;e~e>1?8Hl)zNfAm z>HViAbqn4xxic#_e)mVD+J)VjRx_BMqm&<0x;=9Q%K|f84+L{S%8T1$WZ-hP`v`eZJ2JfF^=qd|V<=E(DOS@1dVsLKQQ2WpmiXNI8zjW00g znE*<2%S2f`J+!KU9F^r&j!PuNEj+U&SV<0rQ=_*oCG#* z$QtGDTk=%>c;Ms09*q_FB!p~;Oo=|J$^}FWM6|uxsqO=q`ByvsseqzK2=HZ1vkKO7tt1bgV0?rF=IE_pq@MVU+Ql{)Oxvgx0~;Ul$aJV)Kc$^5(S*U#sdzuO=*txIoR+I*C( zbzx6ur`bxAPANB2Bn4{Up}3`lm*(CFk(i~}cpP1hH0W!O7Yd=mbO6FAiC*Hn}3Q=rZA$W761N(qw0`BV(b3_ymKz8Ioa8Jn(P z08EnyE4!f+GuDWr|c(|3ROp1ZdRFD-9edXJ<`fo>Wj@sdL+K?Frh#heICfb9bI zBZR1Fyr5B@K{t~szGgYuK_~D#X*jEgFz8|7_P~|C*QA={X_bbK-G+mevn!oSg&P-4 z#)pJ4pXLM;*G)dQSHJrYtO+KYlY=gkO<$P3&|WjN=TXXRieyqTyTf-$4+WMekV~j( z6W?SphLpe?Zdur9 zrG9WF*o}3fmBl|s1jd*HssKb+-L8M z4j<&rfjD}H;<6Z3Nc-5sD(p_q1(ajpV$4f+fqBdk9dpq-n7_6;_Jk$v!DTazADLmH zkWxb1p&Tlv5lFXsStc;kGx@7YQN$n*!f=(-GbQo7e7Z#piReaop+pPh+5;%GpUIpF z#=>c!TFN&DXjONn?Nes*E25Gk9ytvpb^&9g9dis0UKqtXFh-&LP5RE%_anxmGv~T- z0b)gGZfN!?pCQGA$ZXGA2(#sqb%J)`8ev~V9#mSN5Z?pYk8Oc}*e$?Sw~gS5KAv8B z)tQu-r&DMjqMknOyq_(F*~=~a$-AsUf%>%(DbFpGGL<45s2CkvL0@9}Ne_*Z3`p6} zRTnEZ29!r!4F8m93PneG&h=0&r-##ITR!Z>*PP79Gjgdc z$hfF4kp_h(ha^C4;Se9U4k&lgO}B<~Bv{OkY!5$xL3N>uTiyE zVWhF#jJ`N6pV{JjCAd-5E^HUq2j0%cueAklPg6c_nTOM)-YqXg2MjkweLwzpqMn!sNZhs)P6yh7%RT{$r*{$hW%IY>p0rH>?V1Ibu@z4vd=C2n!ttqJum7)~f8Vsq zNPK@ii8PEBf^}in^^V!fbAwU>1@j7o-Nc^((WuZ5;{H>U`C8SI6v;BdM+?_QEXfGi z9@rcm8@yx(-62on-=nvwkcnN$izA1H83CQq`sv4m@NumQ@2^*si;;Ty=^LNTYmQz! z6-1XM#tcNl!ZYLy)V4!0%Djs3&gc?oC6pp*gBsrD1V_x290@XxeZ!|i4*#ig(egQCKwl4)h|v^E+M$>&gDxkAvEr5S$`1I0GO=(i6+|yFcF-z52!(D2(eGWNenUrry^3sU ztF$Nrc2Fdjz7v|l%TxEz2h>Y*)KF&tWlwrx;ru=x_LW$&TpZ6^vR4|~ocHu?$ zCY7=OK)WZjIy>d9mo>_(=cEO~!>R1>d?-c;e-yqa5XpV*f(|CbuUh=i|3fHyDLk#C zY_w3qp}0jYmOCt$^@*(PnV|jy3fxeXW{}QVR0}G?)j@6I4!H&|Y;%7UTEOJdUGj2u zoupMfF>9F}d#0C~569fV4Ibb8fZ3`rt$i7<|J!d!>K7)UQ*O2@=2A+K)Xbn_niy<& zwoH^y-=Q~3fQ0+4Yq0Uk2_N)08Qd<6ou9;1)+#x^RF8 zxKkq{da0Ch14R<47$or=2}C{TlBoN0KLK{XR#hb&lw}02_jD3()C>ydFKh>_`E$;G z#R;&J+>`kQewTt&jD`)PCMB3&~`rO+I%yEA8 z%o;^m;Awt}$9>s<6{xZ#Iz^hgLB5J#?z4{E4sDJeAV-2WOBz6C1|QWc%0L!+zk|zV z>_NwKoR}Rs6eQNojWVNW!>hjJAW;`YI8bTuu)3FAS#({xJ+J`k6k%s1E}pP0kZHzj z$ACkp_@#qc$G|8X<;j6Q>+&o+OuSO!y;i!pR)t_GOOp!lvP-l>iFhfLbb>H!<#cCn zW?q;fTaog@gHAqS9Vz8Mdh>U`F|9JcZT!U?@)@`NvFjGckhPOXy0ecdH?+s}_aN5&7HTW0! zEo`H_2>h1>W&j#gq~(kQuX^v;R_YLpvWIT|NtLjW*= z`NhltIUF?M_Q%#bNGF?{7l*fQfJR%Yyuau3iH&t_sI-AGZtSlHUmA?CZ<}_4U~8*E z5Z*}itDSL=UnRa1fX!t3*{xxkDsex%M}Yk7HPf6tLAEXjW^SV4#fzVyBhPS~4VMkB zB(rTREJez=Y@%JaVvksnG&9z#^+i-fr5Y%xE0oO&O&in{WPmt~=KSNSu&_{2|x z$*UgsRCbLz=e5GwQ0uA0&TBxm*yeuT9R>4aIrmw!6@d1G{@K^S*0nWIkj)UoHMw?pB52Hi=6hj2b>P`pE30Ypp?w4<^9F_rCt!lbV7 zBhYFOoE>8#vPpGR4OQXDX~sTnXij0M6}2byuKEyI&rN=hqcL|8-eHUfV?MmP<7a{>}XVR!;ba4 zW0jH9ECmx?*6p<-fry$CXf9HrYeb<4d@qdFG{)nA) z77x$ju~yaLbHd#yRM{>}^epxp^g0y;OcBrHvOGE^swyP^^}En7r$6GHc#mhiau3t= z^XBg*P&r?B*ca8-jmy}k{rOcprepL+I8Bo4Qs^e1Sar)rL8{a+J&n0*+(JRt8Zy?H zr!E(3(4WM-vqLTP=%4z?v)sPEBcF8cF(+iW?qa{?{|weyMqsQ2Ix>}sg3WV{e>X*6 zhz9NmNKYmT?oM<-9wydCR5%v`qm0f86|QS9Iv%~#AUCZxr@rycYh(jA=hlV8!h6gX z%^j37gCg62Oe}0r+>rJJYk=VEm+gtx?|@8$i(ge>FLF7By*?kg8~a_Im;p}kj_D75 zv(zaQP`F?t)}OC_Pme~yrz21O`226$xnVfi|AdL+_Ky(8}NWq_EO|7s`jQq zOEmO8Q~Y!)+}Pk&7?l_j3~_YSX~2>dXqI-Am&L<|Xj=uQW&%|-irIA1Nx(ffZm>TI zHz*FBB&9y!G{Jrtx6D=0N(N|3jB{GEht3n#(7|&miSpxj^P#33@ejh1gwKJahc80HmWCUE>$SxD*flCgpzBV(< zM7i{bkOsrw=6!7HR=wbT^wsP}(o;k^<0t-JT*7vf{Wn%Gw$xC3F_PoD#+6vnRm@dC zqCax4Q71rJX|I2pGTSp>-5IV`Zx6oazdd-FXdfuJ#V_oh$o^2=0OLsGGWJ;30TU2& zN@~fha@no56|PjPcs=w|G^RB4v_8O)*v_AGkZu7*aWo{i=GD=LTt{d433m|JY8T)i z$L+?fefZqq@XT0P8_r?pt;1oo^wOkT+kY_Gv~I0f6?~iwySZ$Rg8+MO1SvC-QpQtc z4Ha`Ju$1>Xxiwd_g;yEAJu(mCYtEnkQRd4QM2zy7(@$|jgipfxum9GBh}CNnqDc1? zvd+w5ctR-$De^hCXho(r$lf){3)M~XhYDDGiao;G3<832Sc>TNl5g=J|pWq zFVQ>ccE7IRGpaL;rd^1`<4GeW;h9)0j@H7~@^{a??TNQQh-ilC?%qfY*|{u#fZH^? z?6-XAdme35O!$k|zo#L~T-cL_{n7}FluRiTD6)=0ZJ4`%-sPGF@v250?ZstFTa|I+$rvgQl31qao<5eQ17l;EIkq+*ca zWZ3i|IipF1YIzw8H5!f4=Na_KqigufMc9IRqLw*77u##q2~k=&qfl%V=o{mTqXH)Qv8fo1^u+_`=$eV>DKVpi-!){3ILtVhe=iAMescSCbViz_^R$4-B5+u!E zV)!eS1NC(z4*@$o(Qh58W%8N+ zdGY)*9-gU>><8u~Cb*#?Bb`|5S+^5B*5cXDKb zR~dt&)zftjDMfV3*V3+!+x#I;eue9P2bZtj{(W~7hU&gl^8s1m!rn3ne2myPr%=jd ziX>1mxJ^RkhfYP3d!zR;>4m7`@OIkR$YdhzgaIJ7;?*2Y&VR)TuajF7zsa|hjd9rx zwiO|md}xQ#51dO4vNYpFXM$s2(_9y1NU(x7M_m9VW#E}K$&Uo3Km+S8X{KNxv@Ntr z-WQ7D7pxI*Dt2teiJ1dodf`m8aKg!?b&bkDT6Q;aS)5quWYWr(vA5ml%{) z)cQ4rd=y--s;AF;B`}qO?OsoKp9K!hcPc1GOe|-7jCj_Whn(<1)$jfLP)l)6m&Jz_ zxy@7Z6EouZ`=k|AoBV|662DTAMzSOoR3AGCf>Ou+8N43;IdwOURgfTE=c}6_jBrMX zBlfm7K#Lnfev){IvJBT*iHf1hdZnPp2b@+#Hw%1inO@lAQ^n}mCAWRcqaV&%6SV}m z+|Y`%IU1Whw+maPPZXJwHBlIVz+0Ldvv@KxtvMvwZI#y ziCpKG73HumxgB#fn!$f|&W4ZQrdznpk;~={YXs-JY|L9@nOFf+2iojNTWeAFN>js( zTr#Au8T)~vqfi>uD%La#%V(^dodz`kAam9R%BUbQHhtKwM&8 z)ZetT{!PiTM3Oy))S7+V<&+Z0*t@BiJ%Ud4=16RtR|)N{4oM!*x#NSCfQA||?A3{Z zbJ)mWg_#CBA44IT3L~LLYOCh9uU^(FZRQoIhJuh{+bZ2OLxY3119R8G*-fF%=#OSIR83lFUDW^VkWa%3a4Uowj0FE2dXdo%|Y06#G3MKjU zUNHyw7f$wWQskt`p|83qo}O-zA>UhrGQdqL9cVo1*y8Xta?Qo?VK)aa zb1$714!C&fSdP4b6D}s7TRG!)z|-*m+Z!rKjtiTXBW9*$FQqJ{NHG<&YRZA&c>d}s zxU>75pnmjLzqFtRVX^lz2n3|EIg2U+vm{%?N*11C&(kZQYej-27dBbjeK0zVPvI5g zlfXR&8Yh%~lahrvs@-P8+LUsD&6w81+W7XK=L8$dP5OV9`IunS^zQzpB+Z3k1C*)} zS)&3<2~9_{shEe53Qd>oBZ-n>{{e{k7cH)$`=G?_B9LF&g~R?u1o?QU)Sc=ByyoZ} z^+0GsP&a*Z!M4DMi@SlGuyJZB-969Y02wiX%zO| zGBaf5P)aBq*HSSZ>=p_mM;jJe~LiYWYpu_D-Gv&i@@0PT?bbi&-t zv*Y+hlr!;Uz=6G>ax{o6$6NKw`0u>xCIuy7u;Bq#u4xXbcZZ)KLyBRyCZO|f3~3KF zw0=XsHy=u7Gw3Qhk?m%oj13EFVi&IW&O>dfPu2V7&husLo&9HPk>#{G=rt#kGU>{? ze|Y7yO06OHA~ZFMDo91&nx{$T>lmapZDA?`^XbPP8kF(f8I}Y23W)mPi`d{QpuaiY zJa7!OIU5(x;l&Lh|M({lFNw)P$r5gABMC0-pcI)cBU(zil_FcH7%a+f2WDi4&&r@& z;3{O$mD797K(u&%P}ez6V4^D6FvnAEBQ7}yR1if zR)z`9k|-TE_%pxkrV|-MR2RPnq?loS!7F$BnQlVOhIU#0OC2xiP{R%s%lLgXy0r;{ zU2v){xcgNdqO1ot-b0?4f{K^7z!F!f=mmkp7WoAT#`MV{H5Re#mH5C~0*Z+kluin* zUjQ}PTujyYF!jjzaC^Z4YTPhYcWliHC1y~og6Y5~`ET@U(@T|wYloK$<9KksUxdG$7m=kWcBug$^ zAQukmvrsZo5)o7|A5}COp;c|C@-q;dubsCgoQr92!2H<{A_sgJhatqlGsMrky*Q#y zg0Os=vfOu>X&a#jM{m6b?&ru#MS@d|Hg-WAwhBR~ww@vs0}@!8VlzlGKwz#@d2T@xuhqLwkp(1P=QSCbm{l>^7$@QDyW9VC?|;`` z7iC;<*8@A=4v3U>DuJnq4D$=(9lWG*mUhtH zA+@|z{{7%a`U=_k%C123yEbjSpkHw}4CTQ(#MvIlh(U2bI01AUKJvaZ<9=|BSL)0w zVf%oFWWFBHq0)8+uR0JU2>A(uI{KdcXtblDWgl)#{>9UOjETze9~=r^^+($}WG=XI zVOzpdwTzNpD0-k*p|62!OYO`di>^+Z(yWAe=loYS4(IQy@H z?_1oDUzGoHh_txyUiT9#hvEvwT%<@l6>AjM&io_FVgG6ESIT~H^h*h^A9ujvUwB{2Lvy=r8H7NPN(aaqo&dAe<}|Lij`#BG=S=imI-LtCLHmkq_T zNH1l9M%y;JC#Xo0^^zef!J~j~4uz0xw)-{G21T-k(xZ;2L_cj3xFF+c<2d>pCuB@H znxXkUZLvNB-wilMmb@^VpKL2zwT@!eQY3+j#R5f?xTxX@B;&fZ_pjY!%T@8>3Rug&o=hR#RU_G2p3er(%2NtWp}FBCCS8ho2AJ zMyzK?ZgZUI9zQSI;0q3m;y_HBww@G3 z!H4{O{>RQz(6$d2Pg_bHj0ekb?3N|y@X}qgMy$w><;%&|X{63-4k{>S2SrM#*w1*y zqBdj?mEooIRTB&z>%F`ea0w`Kg|iOQ@{Gt+4p8x42dRjRV~Tn=8m{!cKhiAbPp8;^g~5GFP)s_>6QVx5#)ga z(t+suP|*SHcsrH)fPttsUZ!_z7%*xF30~I&JO$^3_3cF0)7B(sSi&hAPMytu-it(m zTjGvVJ%h6Ts2Z`UsZ^3l2@tsgyP`h?w**fOd7bpbHZ|?C4dIn)T@%pnvA=Vh>H-vB zBuVNN%Vwi|AXe<0h0u$+SE9qw;1D2~VB%O)!_CZm=eqa*nFx!`h0V-vD>L(m#fCiy zF%qY%QEz3CW^eARSR_G~e1;4*(tYv_S(9lLSa_wxr^zHUK>!RP)P zq|t?gw?-=@oS~RjinLI%7S$$A61f&JL~mPAuSo#}6p507CLIp}TazA&g>iCCUdYSS zWcs2+0W#IL(Q%@>pknndNml5v+cvNnd72*aT`7pAq$$mZP1Ng3|1xm~568Uk-3)Zt zCh2+c0S%nOn$s@6!C%*^!qscH##96tA^=>g}6|F|QgYYQ`MPkq6mPQ*=vr$aZCN+5I!i zDzr3YEq@8HU0b#LzSg$zk`z-#Yxy^oO%cncS9uPL6ZqHUB~cgV_fF!Dnj;b%IoVGg ziyaAL*bi#K(LTz{vNHYr0(q09aN9|{aC8>}iKA3gawrC9PMfLNmBEWeC&cETt$-uM z(*5~8)B6=W)hL{Jm`U&?nI_l#lgqm`uxX*04v9X>nGZGJtvzo^`Bq-`&Xx1Y>II~ z*d?-gti6y6Z;q?1EJrcLY^O*bO&RLQSS-(>2w!9a>6X`q$xb_ZQ0y`gLt z?g0zd&)?;LXV#GCkmn~+sR8D{BJHITDtqP@_acuHYH3Ixjn)tZ7Sf;y;5JaNboj>< zNe+1DhwFhG)6VEs$jt$fga*Njq0BwiE%kb!+ycL&-}vT5vW}Yrs<&=>1J1_>Z0mABguOILO8lf%KQwH>ODgmIB~qHvunmhM;Y_QN}Ldto2% zB%}3c5gNDkdNhJ_Ui_+o|6ei-NW{0rzaz=hK(}%9)6J%sOp0Vsv3F$6!VX~v!46`2 z)fM<6uiLqxgJ91xXFpccLmAJrbc_AtNwZ>Ukg3)TgL-CN_elZv2%NY8eB_?V_^r?V zMnTfhMxlg&iEwe;l=Dtf!b{iX?)ljyH1Pu;;0PE}_s%Fl#VPg52X1Q)K5n zKXIS44{$PMQ~z-5)Kc5Pt;=qzS%S2vmDTIsA#7FEh)*;9QIEFO*ew^7LZgQG*fH0n zh>+~wM`P#Q$uJ3aBE)Lr>;wYV6M*A{kSWXf$#ZR26qiK^%NlVsx_`ld5`CdT&o=F4 z;Wp1YAge6pHA=98-4K0#&Sc=>@kDSy#^akN;T0!jP`<5yc*VoAqP(m7=+9)O3$G}k zRXu7&*+emE6j?{bUiMuq$`42RN38hJ;afu#@(0W^k?FIB!zLr2^LTDKK*K2L$wMqI(WJ9p`v#;CnB z&-mWwJ#?!MT`5}ZzD{&T{LrnBtcrefulpoe7f!f&wB%UNu`}MW2&oj)G-k{ruW_?~`)h;_z;|I2^mEOv1#Eo;3SGF#0a%zh(!)7|FmXC#Rngi7e~N z9}O@4jU;f3g}Ja<0T%73RcH&vq~jkOn-;!H+B$E$S5?fez^aAs>{xVB+O5dv_eZs9 zuS1CcAt)0k@jfDr^zb*6CdG9{NZ1c1i#c&Vj!d86^}C40whk$m5QhuHgT;MqmoE)D z3GGy%_%!^cbZOwHehq?R=`p$%6or?N+jO;La_tNUWK0q!%pX&ASyd5ZF&v(Eo6^W0 z7j{TbTbY7JifN#rtzj(k+AwQ_97oiJ#)hYbQrNqu@(bwfa4g|K3M4K$x-@@Ge5Sxs5?#^oe-99I)}o3*qo$+<4-)1c?&A-)P(B>$&n@bFI*oMltIsvX+V+ z0QS;9u0qS_{ec5B)A&cohUwRY*&(@_8ety2Vfy5n5e_TNxTk$wH?m_z^|xb<&%4y$ z-ml_HvQ+S)a+B<&rk-Aqx$PUeJ0;i^IiGtS4y#7Q}iW2ft%S#TZU375!!ZdA5x+xOgcpoX}-yv#~wAUx~{T zYTIL&9vbrC;)T7shusX~CdCp_wnS$Tr$i4&G?1#miL5dlrgl7V;xPQUFYWu;+Aj(% z&dQ=Q8xN7yW6_gbm@*A)q)~gB42s!EkyLPCBsJn9(f08Dft|`G>0yBmnezE2mDpQy zklGP|C(})LswQdraelZ$VBB-c@kVxw^5YXzR=$Y-M_PEje_zxdo%Sv=1O6onIW zF3N6;w=F1_lzLCH`(ZyWIL_-tFK5RE|KHYx{lsE4ezJedSI9XR-r)>d8Hno?(@l|1 zDi(J(Sv2lh(v;ZP0Cp-FJX|=SEFBsR2P_Rq79hV2>O^60(kD#|M`A2AK?tgy*-L8Z zA$|#@_&%Q7NbCBQNjx*R3|l9dw@#J3avDz_4hC@sNc~+GX9u5xLLt4XhEC$)nPIm) zO~2Aen?D@OHdZM&D|Psp)gFVMRWU~4Ss!!Lg)I8U=Wth~E7V+38Y8yN(yJ~-pLo>( zS}gTYz|3aDxEdS|8z6fx9TS7>z#7AanKCDS-Wxx#fbF{E=#ayPy z1u7Q9Xa9Xw`R-jsn-*$#e0vqA=$R59L77uyiyFs7RG2PlO_%w5)r;+2;9!q4U z5X`$0YE*2DzM$?SdqbL3z2rUf2juec!%xSUKd@E%k1nl#(O)xt?<)SQOZ)!9)7qz* z#Q#y0jfpSwK~5;6^5XcZ!4{NN^pr0l1ul%T!&WHULovH3QclITs*po#DgjIbQR`1g z7H#y>tq(mcedvZ}MOPr%0d2udFVVz0uwZF9ho&lv&e!B= zy5%X+u)oTp?+34%eg~GrB_TUKfH7U@Jt%|X)E+udV}uo|j=lnV4A=nXp#tmackqvk zw?R!wkp!F2VaWi-5r^G!!_B3%)l4;$r|G2=posoymS>E4bvUG07rHS%{3(aHK`Qae z`(3tGiY&!|$eS;X*rvf+9n)6VAkLC}hFw?mnj#76v>gt<652r@ad$ppdF0>Ec@jo$ zbnep;_>I$U;Q@rcYk@~tvCNHT~0AeDH1ms+^DBMh!qbnGS7R)IteIGeZH5& zGrhAt+O)U6wsui2judlMBKY=8{~6U z>js*Wqg|9o0X@tJ;$jCLN#U&aK#=VcouqHco5iT|h3UleJnY|6J8$w#2M5%m8!^G} zi2b*7Lv45eZ!>Izd@c(PmW1MdC{|AluU_z3SR;+S=5*)|85Dyu+&WLTF|$9QH+hgS zfrlF|f{VBKMOgMVV)uo)Wan6!0T-sRKv(Tid!0Ipsiw$2Dt0Sf3LVATw523hS`xKQ zdY_jI1SdSYgj}XkR^#e7S|NghQV8paA!=CM6rJSL6y47sR<~(yD({9B@)nD#XLc%g z%ufkz(=LN+Zpb$V&zmo+UctfvC{K+rWPRzJ6(2U@MCDUAwH+u0<$5q6DN0CIvo@|4L<7;r%W z`x=(t+k(zzk+H|hkQ7o3u=sOPJLU4sLelKs3n7g+x5spA-W*m} z&v@X~j}$4$D(nVT>}rpi2!l2&XgIJ#2+_ERHtlx0o8$^^gk%LJfgtG+ztV{+JzTId z+A3IHJ02vK5jkOn`mY&pH2=Y}5UpI56hnGkIKZ~XN)zD$#SBs89u`-Q9L zVH9oH4Ug4?egrzF28B_E0tq#t1H7XGXls}udtbEOE5l1KGxgv-uFCKkeO)21k#2^z z%a}VFB#F#sW#6m<8gmim_qU+GSFf$V-9h`-&B3+xqtdu@b}}BQ%;kN7lsJF z*UGXLQA`0vwm~96TSu=Zy}r%D6R)Eqbm%Ll^i>>%`Kt$kC;ghDk81Rq7WbXL2Q;Yx zd73KPz&q)8Q3kmxp!PwgYyZN#z#OlXI&CvHQD`~KwoPRo9D3gCn{gLymN7MFsJkH% z&_QGWOQ%gI>vY&T&#d=5>jXCwviTQXAKKQExU9RwQXNu97x6QKkm$cO1b^oM;lGYv zCORo9@TlkE&o*tLrh-?%pTt4{JK^zY*>DydY{%sU59<8~)#B$p`HZnSvr3b0&=bj* z;JpA{#w3xQ$RN{3*B@0CqgUb%| zd?au`WLeO6`EYQtREMt7t*{HBjxs@9k@E;bY%gpfFwXa!kTd1_gBc$`?~&aHfv2L+ ziHrEv9?ilVVfDNsey8%(*YA-VVQ5k=`&LUTcs=^a8PRu zAHQO&=H;@h4vRilh2#L}2ckkk*Q;m2zuYO>DH@pDHw(?d4t_STTXE8FfVYF6CraS& znPZft2N#KMD-#(+yi z6f-&ioK7(S{`FLBCx|o|{6F-?-tIk+>THETUa_b{e3-X6JUiSdONnZR7h3*TFZ|Uee4Hu zi#;(wazZLD3icO^TI8kDBp;BEY>}gyK3ba9q7oUr-KN37Nt*Iv^npl&xE$`!lIT^* z?kzqh+FhEmPgsf!sv_?U8M;7lOC7{ziX@w2j9wGx-*AEno0~G$@d+Y{-QeW^i_yC!$1izPoXOt+J%cIxH-bmu9WM+a9{ONF#k$YA!PX>OSho_Q}_Kn<?eZJ#8!EIh#mM!_^o4@|#dABK5-^N<$4ifkBcCSxVgVHnN5i~uqQmn=l44O0L~VLtGUqFnR@rCe#$%>zmRGKC?ZREauOk0-js;zdt`sJa zqUUUX-DHKKeG~&!nO#&Ywp`t&*%wqKGEG=6-$>&w8XJP_;dcnzG!G+(-7pLTk{>hn ztATqSvBl?o#jsm^$Zii{HmE-YbENC!9SrVNt_}FqZ@;>F0Zja``Xd^}bjuVyKGhz} z$Z9ZD_5tA8v*B=vJo_W?6~-ISxBC3asCV9&9Id6V{v|V_*LIK zkrhG7WVv@uC{!W}`uROH7V1A#^eE5qOZ-dxQBe0Bzt2;z>ZO;?*Ck2JhqE;qKDlJ0 zBw3IMEnV6}D?+YCo&=F*q-Dj4P&{PVtxniTHZJ>_fd-_X?wD|~y&(xAN|=7eRM z@`!(uz|>>9ShSVF59*;WgtjZ8vZhFaZx52XV9Zw-=~j6!xfOII;$|4KueN&`W!d40 z3}7C8E;e{O+&pnz&sfSAP7nR{W82*o7qGdod18t0pn43pj6q)+f1V5;0-9a{Jp=*X zAumi3k^FM*#og@wEn${@~=t(??EBBzb? zWK8j+&tUgg#>!BvA%S19Er91@Gh8?#!&1ke8qld~m!*Ytfzxq)dhv8!u2(tls_(f4 zx;W8!ahkGLvJ3XyC!~Fz7bd0sDoz{+yJ2=G9?iCMU;Mq%Z-s4>#1{k|EG-rDpzsL0 zumUTWmkYvWHx<3pjkGaHw@216V~bDX9N137ZaEFIrYLp@OgZQmuRBWyJzL}*K&R6O zZ;%%P8!uPd2{b!Uh)GkH(Rp;If2UFxAD9d3RViYFDD5TGed!RQf)zw~K)`nh0yaid zmB2=*dBXyhMCL(QIY{*8V^IJg67wa7DBP$OiFBCr?x5Rc<-VsN_wAi5K!IQEi-S~) zUMAC5mnF%P6ukUN1bR^UG>XgURo7qLIA>5c?AGA{RRq|f5kJ3%@3Pec+HYnDDlJi) zjVZxr$p&sQK^OLUK}Tp*$z(poz#cuDij5QX04uWw3a#s)|0bj|4N=%upw<_#gq?(Q zzm1|U7sTgDr*h>NhTZbRKOjwt#Xb!rL;lbW|8wf5owJ=9wfS+TrHWmzs8v{ilJ)X$ zeoxX~7@!!efRaZsTPc!7#l}NF)NJ<}FR%y@?|Y|3Xz*_Ug>OuQr>|^i@hjRD%Mq(_P~k+;JJpqO*f{`VGQHd-wd!A-IA$8bA|6jq^;6=Fm55 z+$^Xm`Sh*#NSq6oMe40^kW4X&6j_aZuQGiNQF$8R$K=zQz6H=yev+akHeX!3CfVkm z$#~0Y+cahV*SlV~^-NrLL(SrAoT9JE3waoQ&C?VH8ifx8*iLn^V2IzQMOBX~P0Gv3 zf`63JS?Y(1N>w|v?zO+(^4Y+z*CauUVT|EA+5lM1vt2+L^`#&E(rO?Fzu^QF>U(kj z*#Ep&2AK207&O-%ROdm#aTegDSXL~<EX9=gT%q0upN#QNTw!d)cE|?vcSAvdU!gy{=&E=E3K;ZhbZPAMeb0s%LKS-1ui#8 zSOWQ>+M|${EkHQk3Tok{2Y(_q(uGn8^?L#l5X3d$f zLQ#S@?vU|=AG)1#KM?uQtz-V)`Pf=^*(+xj7^2pRhCsI(%l36xj~y3sAQCoWAw{B7 z?n`8!ktALT@3XK~${rf`{^`L-X&kp$RN)&Rm<&OJ-HNwec9Xq_x?E*zpIrzYj+8TcHyt}~fUXDH;b8yxLB!Qa}!1?^ej?s(a_^|C-v^b9Gh|9Sx zTE9A6!bDpZt&IhWY_fYSN|g)e2U@LmN=GQ>5Jl?Dv^*2@E&_OTi$&*OOA~$Y+F50q zvPC{9tD}+puwPxy*Q;8C(}SU``nc*^NFUunu(1KE0PpteMk(h^-(t~PSlrHn9PqH4 znY@r1FyxsXZfbkb0flcsSisP7N>l=jW)LqCy+5l1q@0J{CSqvp0@r9?ncM4eV*LcTJ#OoJ zo%xRaagUJ&blorMSI!s=?iM4{`1!hjkod8Tj0?y4AzC{s&Ywy#DHKVfVzK(QfUZ>Q z(7Wh_*xFffj_1Bv*!I==n#Rk`mb zp5c{Lk|iq%tM~1ouWulBKTsCIQq8Dc- zC*1JH@CBHG0^kl{%meuWgJD;~3c#y;r%r~Y9NTG3?(g<1;ctAU!{ePUM; zJN-rD_ZO1eE}R%#YZZxIJe7G!k3`TK38*`c{x$ zmFKlH{1pAb59mZWz=}E$*#V*)AE>h3Q7N^I&JW0=cR|zgP8DVr^(rtryjwx1#JM2M zI{<}jYXj>2&qLe{6HOR(ONsi7E|`7a6Dj-20=(w~)fO3Eg`Te=D#yE1xr7Yydw5+? zD?TV&=QZqRqEM0R;>{kq?hstj7K3gGh@sBj4sxydp;t(OKMwJM>Y_%M0x9hw&t@9c ziLm;h14dXCb5&XIzd0PM28_aAnkw?sj zyk#RuzURioziN~mgICRpkLP|Q9+=yv?N@XvQ~7=LI*(P+2ZJ}wa<(0NWNNq?v`05O z`Nike9WVdyDjUJxe0g+L%pnM>^(%I&hvcVYGK9Js^;QOBMWqpD?Zk6Tzo``)UZPjM zjsJ5pShzyuv2WyDjQr=+FMKV;vWi{%_KgZs=)$X5qg6bmieh$C#6ZPnXpg~4@R9fX z^V(q9D4;iZHDedo+?TIMrUoFF7S*$hL~nP~D}z^gn}LD%bU?=EfUpL6ai&5eU;Jsl zBk`5DF``(*$6Hea@IH*qItnD?N5f-WSiN!(NI0!NQwIO=&pO+@5SP`*bt8|2g-B(V zkiQ`QjHl~VUJS$zN%^#+K*41W?E(;&0UdO0<;!_r|D45~1pc=AO|t2QAv1uj85PXk zPBBobw3Uj*egau^R*3F~-zjkxw028?-me#=JHe@i%4(qJE)KjLpjXui3T6+x-SE?^ zE=b$>`vdPwQC9&qk`lZfXII9WMZ2J4WlToB8EZfX9jkuzm1f&8uFGz+Spu`!%4R50 zJ}uX)Hq1^C=tXD6J@gL#_Ly=A$zVu-l1+pCQ|+KRv~w6v%L*m9kn)R(@cCVMQ^UgN z-yG0K*N98Wd6KSKA*iF#Jm^&zyld0y13JhhkhI$u)FIqvDu-SXj#5H{#GuHOl|`j1 zoeeVkQ8mhN;7dEvG2YAGxt@N;wyyOBaX(n9U{}dM2rZKIx>rZ_h~q>>5PQjfsYiT5 z+8?!LmacZ}Q&4Wj2&*L{aZjEitp8Do+0a9*w-kxvJ|-# zlTCs3h&9*v^z(COq87o$z_|HuqY_s)JqVTM8FEk;tWoz7y#}-*!y#X+)S1Nw(v&C} z=j6z_-SD;9d^mgm!3n-o0=|DkPFqki@ZEr8WC^#pnCs#ngbPN;#nw^GT8bo4v1!V3 zUnn3(9nmBSGN4h{ZH4HD)3>W5_T}+!J0bM=hDmz0-aqL-W>`R>NnBM+GP!}mg*{Vf z05U3AT1qjH&Mlx~Hv}$`HH(woKO?EiCiOw-br-0Kw8T)F!dOtAI|Q zm8QHc?FQlFG$pDsCwMP=Wj$1Z4SL@6Y!AIM>todT8F;mT-bdewf+$O~tjogy!aajB zq?d5;<|YiQ5qR14+kOnb7Pk8@m^Z6OF)(DVk z0xAF1FWsh((Y4StaoL-At~H`fYly;5XKmi6<*gBWB$;XslW6~A9d+fnXMWTe&Unem zhoP4I?AslT#r$}FHgJw?bYV}o(#oI|Pz+>na&S-F13Fzpo~WmeUN5ShCo)Fa5z?wE zq0m`LfLU$P)&VU@6( z)IVp1JD04`(?T&PC~^#%XVIvwUeD`SYz{B-SV^@&;_q%)vtsl7CRMeEc^3tpZ_}W>#!3%P&Ii|t7~^^l zm>J`4=N-!V&F^ZZ7HgyU=NDVaS{GiIO06tQ7R79#NII}cpe{X2VsTiJmp|;*4LP0K z;1YlIWp#CQZSW;2)(0Z@x;o-0xUbG#iOloGVa`Xs@=jifEPof5MDLXMwhXjA@9mqB zBhUydMn*s->6zQDIRz@A)ic_(b+Q$L%D|B`ACbvGh2^`<4IIuh0JbBuFCt?E_Nm0* zUbz`%F(2RCQ=Cl<(+KQ7M};pAQA{00s;StVnT>QE-7HKG9spsPHPF1Ec4jaA&<&5} zO4kQ%4X6>z{F@Z&zs4j-Q{ z|K}ekBFyi?AYu{bKSmC!5h}RD?dR|EzcUMCwAfq-!GpWq{-~;$3|{rZ$#ZZx;%3}9 z;CkaHRlet+cjokneuYWQ0t5VzWH$F^(j9RJ9X|&a+t~AXIZd@a13Ks$b`vp%D?4@F zHP5?ta#YR-4|(ZT-HMZbhyC}+s-o+8rX_yJGftEmaNE<#NqyU=!{PDU-s;r%oF2dH z8r?b2_tTbrz>>DVNA^kwB2y%v%JQN1yWT(TrTv;x5JQ_htNUm|?E}T5%O~vxCs4R< z&h=ZbzInm6E2_)xSXersj`Ynv{_lF#YHdYGN>rY&4uxTn{EGQnbRLG?hUTa8AqfA0 zYJem#J9+u2JbRX4wde6))&0NqZ)|*J3FHTJscO|BehyzB&@5|K?Bt!$+?N*2?v8*C zEk8_HoQG}(@p)dHC`!|qho@b>{R(?RI~hd|B7W{rl-cwfIc_mzY< zM>5K_^TWxNu`<)VS zH88Jr^n9DW@=W3tn6$<5N)P};@h>!3o0Ue{J%z4F1eak0chK`u;+Cp6GTD8N z5{E;*Fk{fL8@4Lb86k$9E%>|cA4`J_AeeJ24E55D(o526j~y|Wqz@IKfUPbiI?On1 z4!;t5QF=kF!z0*Z2(=X~rVq1_EvmAB0uW%;9an7Sw>4f@w& z|Kl5K3ub0#{WFE+yRfNgumbaLiUCgMPNdcyRAc*hNCA{a><1fh!+Q;cm`t@I*F8@I z|29ntpr)&547+{m*C4pZUkOy+d;FcC!?R6u!yB8jB6_!boi{n#*kEJW&D@`Ln`U3o zxdje2IF3QIdC3_C$O*sXgy^Yzru@)o+lq>dQ|-bf1T2lKhWM4zJ~~cR9r`h>XBkj7 z&`6({zF~UuYqx^hv>!4xUMHsCm*y|}YwDtW{tii-_B^EHuYj;Z9QDo;Dxd$zA(`+@ zjXA(XU^V7|4m$X5_~p%Fi-lRW`SPzwq6=G?Vk--iNii7|*@zqVUG8Zn%5J9;SWz02 z0&aFV+L{{i8QzE9Ekax_izMhxr@b^uG3;^7xP#{O>BDd`BvbC5S>^S-=ae#nZkuQc zy+X)8pC<*g^TP{3D)bN^b|)D=`yh_(&=utAkDo)~Ei59>8q1OAxdEfKtTT3|#c;&` zae4yDng&#;(e$Vt6jMTxLMqlAz(@^P9^R~g{VH-tT2*`Gx=YGizJt&vt6+9N-|X`0 zRVhqAKZ#fGuiH*v0uiN;UPae7iPzxQBd(p9=v4p{iTXDV3Ts$DzHxkG$Jkn0kLiea zoF>L~V;xn$%97b;Aze1Hz*54W%3lkjZE>Pb9z;|M=mtfpC@uuIXb?SVhY0bc8WQ_U z_$SwpOyLGtf#2RA+N$!fh~Hqa9Ydj5LaYx+_k)#wiD2rhiAt!tV9?Kt!90v}l67p4h|3gAraQ=O3S2e^xB1_G za@ODCraoYfts;7EwehZN56J;5Z@Y|QV0*kBcf)DQA`zs4uxiyv_eVjNTF}Lhn~(bE zosf2H^=_5-$di0>$Rjc?uBk0TxOjf6ygOo-G#|bkr4hO+=?!JMB46VS8FWs#AH&Ta zImGt88t?JHocPIf+p4h_Byz@5W7f_j$Fwt5)CPX3_Xd7Hc({FZPtYA%j-*pj0&%p< z(hgytV2gSKzW`bv7fJHNC&$Yj2SUa>>74Gn^|e1WEwY%D$ag;MAl2MViVMe9+pSE> z2NZLZB8^mRop8(ab>cpH`=U$Il;}2Xhj3Z&^2kEoifM5n_mykNjj)x|^x^dj-pdMj zH&2uWU0(2bo3@63MsSvw6nb5p#7myPbbgz*RsNyR$Ptj{Cz)Er*G})#KIW^s0&Bc# zeTRWtvp=v+yV$)KYMom`TcP!GAuk<7UUI!k`Rm6iS2qqkjTjkwe{;go$i;tY_`>W3 z7Cg27-}pRI<-#G@Ggb(BpJI+s3BnY zgSwE1R~?pZl^rAdH3!v4{LhM)1s_)D!CG}q-l@con1Eq&E;>f$n)`1~ei*g#lWqSp z5!>X#?hcD>8u8;%G9U8d^|Nk*hzB%L26=@-ULy@rn%U4!Razn|@ZUNUbK=?I5Cy&h zSTv{5;V!+`@eg}JB`^|SPwn@1r=5et7>O%4rruTOv8l5N&X#vjrc0PY}z$nEE?V%UN2idZRIq(tllIc z>=D%Pg*|Y4`LmN}Y-{je5Py5S6~ZzoW+O#XQ7xv2&Qd3Ofp#~&6Ogo#cNeI2rQQc- zpHUCdb(*u{wD7yYw6bej?Rmt{zJ?n(o;{wUuJ}>%eQ(>CA4_QMvhQM1tE!8i!1TWS zz9257NZK8?HsDjg6n??W4T5UHc&ze7{uvw|=|nz;^Szw>3D->~ee|+=_;t&|bf2H| zV^ZV7PBW}5qhupaP|Pt3xZyzatD$##bZgo`$)#P^7oFjA0`$Z&$I%OX29VVS-V;P| z)&|6bh3El-0}`KF<^5q;Q=P=y>4C|R0fLG@NgyH(y*DAAiF7K1Vo+w5F~s3b!5#UK zSFd{=tWPOXZOpJ6_C1T2>%G7g!cQ-tkrzD9q$&rbTdfXU-f)IF+`tLxQ%YrXKMb@0 z>i0Vu){@+5q~2=Y3={)BbxT356u8c=aO+a~NI!@#AvbtV4aSV2>>;ByRmaX)M%jPt9TS4Vfqq|&* zL6S40c5biY@vByCjrbnv(*8P=(*NMV7hAu&_{+aO4V`K5x}w0HYWzJ!7X|i1KpEpZ zz0T}RGci~=3ItD%!w!qF$_M8C+5fg}rPB+tPhzQoS_RSOIO?kM-Mb1Si1RKLtn=vc zD&&`hSMb)qbWm`F>{lY!)6AG)13d7W;fW zcbB4T-Wf5dNQ3%#o3?_t+hgQ!X{2rcl1fR}Jk%u5@VTtqw&0;+880_{h30TDHc`;4 zE(Q*dB`Q2IgmUqto_6h%2>tudPJH=!XKb}VG0bIYZ-{P)Xv>5*CBasRNUcJ*N<;sD02kyh70@KlWNC)Dbkm@l7OYes$>HnS1@QlyQS+iRUC> zhy!p&eaD>#n30j$ffFrBuYcJBl5Z9N;w`epg?H%ttN>9$F(5~k4?Df-HzkXC5BzdG z2cznFI|My*dSQ*(#Xm-_geLK}EvQm21yIC?=mYS|D!N;Oj{eOKjj{ZjxU%K6Xk;g4c!G+-hT}wvUk1Z6Fj$6Fg z9g1Xu5!Jc-`KKZCq*uku+Z51@;uBN-{!vXjlwAYw-OOS=;yyWrDC2>LqoFX?c%0xd zr8Mq4Z_l#8W8lZ~DspqL#LDWPJIi%U%6BL#F8B+;`ZOGCEARC)FI z6_XxuuVRQ_21$cn0)fE!M#!7TiI#?J3ojOxO84^;1VePSN4av{Yh9igXiC-O2MqC7 zK{Rp0%w}g0OveMU$M$FB)mR{z*s-MV{l7Lh3usC{ed|3EHw{{-j(*z76q87i)g!IW zAPzduq&(Ly!+_}Id_k@MjrKiD>w6~S9jC9W>*fu9@YfYrU&Iy^Z=J>WrZBv=Zpwf&)v0vhNEpVl*7^+s4~3ST$lu@>>bc|3x!0A0&B!g(M{1^o{On_ zq}6o%-Fu{wZX$RPmNWVuNX{68@ZD+ch*J9gK=Vz8h4(;r1oh7MyaBxpv}-(4Ntfr^ zfE3AI>G>}u@i4Gs)vbQKWqi>1lgq+ zlpgUv;`@nTRYb|Z;eB26Hu=?v4+bAJU1J`l4m37U?C|en4O~qBa?;^xC&G{LhGzfP z0>alz4^Jo8xq;AynO`fdKsZD(_b76QibVk}ECsvZaa&pBoeJ7mc;ukuj5H7Ik0H7{ z0$b$h3N@#IR)PA`SHxXP9sbBRReG!jJg$0akZk5d+SCHqDEJCIO=JE zf*@RoL!&AZ`cqg#3=Gw4yc|uR@0(!tk_B0GNfdsXUUfLQSPH!QG$kfuO#;^irq*!S zWDbrWRNwao;X%{i)2}=bnP!rB!s|@A-ZUkC*bTuEZ~U!S;oo-wNI_}JC8Bswmj&OV zwt*kpmhic39)qQ}aJSz{zmo91Aqb$=y!+5z2VX4_t>$4`*1Ref2IhG7E9(RWvkg%- zp)D%i2|$7qvbIkwN#nLUPmJW)1MG|97<>FF)pG6C=e>J+3h01l2=U8}bz{1Q@P3o+$r#ZgTnb&^n*4UE&&Y6NemJ-~ zbc4)vj6ToHmF^FW7xd9r;o3U-up~p~?AWnAI2qH$a57pr;e_I!_(Gwr3<;O)rwe;G z(13DO_-6yftfxpa6|0|L5n@)YDWH=iI%x1k*OGSGAwgLLxCKFF;>it8aDkDNkz+Od z7Rj7m+Y}`iV7P9ig0M(Xw5odOp*cy+(s_d(DZJBupYbt`X%hWx2Ns#L?TOXX=Tbp# zbEA{r?#bKoo9A8E%G_=iTZZnMHRKI#FgleNq&=`JIOA6;)%E$-`T~9PzBjm{a-$5l z6_YaeTr9!J!s6t~HVw~dL7DRYqWpOi>7?Vrc7vspPK&(Ry$8e&hCDNU6Xvdd#X|jU zm!YOG=%*YGE`hTb#3nJ7NkI}9>oJ)~nlj~=Z;{wAOE4uTVA}>#_M8zMj$1huH59X- zB71@00I@l&qRi)iOh&(M)!qo%9KJjpIeQpfs|czLtfQ|%ansR%yS=atn0v^M>h(3> z-Ak7v+nHau0QptscIsDv7^+w413CzfUrrwkJ_UIz2f=7OCT9dNWB+!HL1H&KKPztd z@$=pj9qV=r=+po#sVr94^U%A{nco!9*)zAe>#q9ViM$hq<&yQh-Ls$(wcM2KhI&g< zvhYJmA#~Jm5CX=UuZf{ztPve^j2#tY#6+eZxGG8VwwRjl)n}d|>tC3-El7=z+S%k% z4D=Stq+&4=3KeklhrSOLDbc`|z{V^|k`_6pjdm-xGKMHq;;2Xx7g?@6s=48}Qnbvk zP>N|MC&I}z`QyfvW1z!2$~SoF-}_e!O8(IOhc0rQ8%kWb0^zz9O3qWvS&DoFJ#{?? zB5(Pf3(R-#)-*x2!Obsx@Gm((NL+Ml;kqv${{b#ig}fu=3TV5fDX%CSp;Y@G(W@>C zOI|rKX9#qF>lE8Ovw|)QjlvFLlLG4$3VG0t61ubW&^5dv&zs82!h^x6K9Ll9KDduG z3+og`fz1kodbfIspph=4^8>DH;vq|PT$L=q8qIXj)42m--;cy~if-XG`4RH*oO`4~ zSv|vURE`^7W3OV*Jj#kEn-hK|F=gMjz-#Xh1Ak18y)f|l*a}|l6w^kL(^PDQuTfYD zDHcQ2`8lQ1F3--$+tN1e83@a@YA;Hg<<;=&v=-^-26`pW5OtP0$H!NjWcB_J-LQ~r zZNP4iE$+vG;TI=rmT!^`hCOs!6;wy(N)HO|N)75}`HE>x(0yt+Fk67l6Lv{^<{E`H z{0z{qFsS>;%4uy{e2pQ&jgk&w?W4~=Yx-=Yp*9F|Ikb45WOUhJi2d!aW2--mu%Juq zzA%^UoJLMqO-micfOhOYq~UZc_6Kg2=~Zhy;^mD%eEK9RH6Sj;I3t;y;iW+5iTLRQ zQASxZ^yR`nT&-`keJ6=Juu!jRQXFw_q|NX5gUi|~AAKGq<{yorSK(NA7K=|-I!3e8 z8PA^miQ0dLlQZJFtHgii{cDql1w!xYKKe6!dtA0A5QxL0{hmz}lSYwsRBXNf13%Or z)UETnFYTaBT(fxI1rN*18{SAc)2r-WOzZ~%%at54%QA}N4|Bro)IVL!`qK;xG&G5; zYDwl;dvzC10KnFKl$%^iF~BY^pkkMZGJ~4b=qOjuD4;PebBJ%0EfE#N?z@LJZ_d(` z_%4ap<%y{R6r$N9zd|~dU3@GMhkj2oJP)F69?n7)M+`mbu{a(+rV-gO^uzp9zqe%) zyDV3Mg&X|m(~Lf#HMD~?%O3>x(FdVYPOmCg9uwaN`qgQsNoYQcG%Zkk3_BQH67|q+ zd(dJ5{yy!~>wZ)HFtQ}7+M|g+<(&xK?X~-<6$ykfS4iUaG0`1bxmOx`bR0B8?(WXGYoa3#|voK{R zwx2bO!_OzYwe|-*i=FxX`vXm6DYuQd>%yJnSb<_a#UxWCk&2BI?FI9g1S&VHd=}POvlA6|>6)>A$T&_IE2ar;>up&|SW#f(5Fd3F24%G~_eyc~ zF|8Q04Km0eBaN|^e4xs9&*qizbbDg(7k(el1UQ{fV(-(nSsv_@ksHTT9l3?mdg8iP zxAgYAm9_##F6-B_C@Yz}0HHIR7f>+|d(-Or6?aXIXgigiAi0FCE(>Vv2$Vhx(v4n2 zUXL!Q(;SAb$mUCGWqo2@q3BcDS#hVb5E7Fpqv&9Oj~mWU;M=kGgcF>nUi`@)PJhd? z;JkjQ|A*u#w*|*_zYpY^)q-=5VnDmN6%q@e$AwMY3wfp*ChQ7}F!`9J%Jn+8pr2p0u-x5_31xqztY;j60oJi6 za)OgNqWa^bQma$*0Mr6Wyggo!utAaL6PXh0KSN`|jibn+^Kl7a49B!WX9 zMWD%W4s6DlTCPYMV+f}lS{L@40~6tZyKZ)auyDYe#rI?_s(Z3*DJBs6K_Y)JY!@Ua zd+6oblagJNy>GT37*3gg)y(aQt)!TT!-YB3f@KD1@XSN;x4?Nu&9e;-97d~KqnTFlpT(mzs>X- zuPe$S&jSlTfn0eTQzriZp)^IZ=1<wpm^SCfeVuk z4qNSZ_fX6(ij-5ar$COXO?yFdQc~)Xs;TzKqkE@kcvc2>g0|33&n#eg$BFLIS-^V2 z_thTwZh*AVNU+1Z@=ezl^41Z&=Q4wL+?Z4E*Q(8&w&R;yzI0maSVO`Y3#{AI|G}sV zmGah`3x8)p&fYh?w#@;t<^o}o+S zEp|Uh>VjM3X-cG(l`XtR^6BO9f8;{xQUkKxQv=coXi6>YQR?tkOkLv6WGgQZWZ{S1 z@GgVpt|AHQoI&|WTx7!hD`6;2pGC)qKsTH;#*r5U9j`68IBOymvo5@iW~rD>lJr4F;TSM@TO+nDSRB+E zk)~WWd)tD`%n>-^*!An#z_T0P&mPL5SLna<+KJy=)~r9Y{CqaK`NCuZR$DE-_bKL6 ziVOh5jvNr;LWqmt5|Tq-Be<|Z-&^bwjiGlaaKeso!2gE5cM1Rfd1=ZW^RLTM{G(H8 z()L*&S|q|lM>To0J^%|@uFJbY6XsfEg8*cLI-xWH>*BW2y$qhXB{!YeuZAU&SE*k8 zxoQUQ`2hY1JgpBHQmz!$^R7kevgq41-emyYS<^jtqKv=}%AfkK%UpDM^ML9?s2o8^DL9%G}XC%*c^{ejBJLvhpiwC|@%Q&+xM7i?WN-EYKvc6IS@fiAtgjpkIf=h8eQ=MH?me{dO+Us|I*` zL$ZQ0)nz{C7Bs>}a{!9m(u13Yee|-gUXt!ucvX24WbQXgb}rcBUN1m3;e7r8uYdkB z($1K#@1M{4J2T$zfX!FpnQys%A=tkI@6Mln`FT$()X7@p$WOp116JE6yB`S7jP9Yk z_)CL8Q>cZ9c?ZsO$@LkJx}FmZM&0M^m+UYYgAP8W|M%ppl*riP82j> z9t00Hk?E!lbaiNtbJN$HP-SC6oQbS&=6o76$Fedlx>A@xid;B?++;QL`zU4)MRox} zHxe7YRYR8GNJN9*QO(BBBlYFrtP5f+F+jd9AOgD`#zl@?V6N2%K2<9qrdVATfr+S* zlTeK$(&cH&dF>1il*nX{t3Ndc7kA#VB4;2X`HAA>g^MJ9EA! z**&0xF8tQ4(v=foY`QQkSQwk7(v=|by)Qa5EaPWcP<>8Gq-ihNJmS8_ZzkCsC}w+kMJ zB^C4P!6)vcFZDMg8A*!OcLQ-SoPZ&~m3y;ws&$S<;_JV6n zYsCGa-Ozh&-|Ry&i_H<=7XOYUKWAI1GAqlHO);4i$-q!w0UtF$ZiKWmg|L*x2c8Q~ z;uUycrmH9*eRi2Rkr|k~N{QtVz{>SM03~zw&!*jv_sI*#1B53>ar_bXrRvKY|0Di+ zZ`5Q2r8`rC3-#l7`Rfj<2R+vXbVDRKo0lDKs;|tF)QZc*<~}f!f&{yRaHGEdj-U52 zMr4P=7+a&MvhZWyuqD;FEa#hrCi5wzw9Du!iBZu9IwLpv=^C^ey^8yiJjqtN-oI11 zb%tKm;w;Iff;fz@fwyw_|L{-z=mY5wyyiF>xe8nGtFX8 zw8lT`$YK``u|S*IQTyH$ib2=ePp{%pbQWqvpO z(v&qOc_}CY54#dJ3FeN2jdR)xF&4zwVe|PP?f!9;#cX_UPjNOejJ45qVTeGAaFi)I zL@^*0R87SuNl?id$t!x*3PByH6k>oCD*vH671qU_A!XAcTpK5<{ah=sQm{-QSx_XY z2{nkz`B`*6t-GqYsyL-SrOppOCGMe(a}Ch=zHnB1pl)Dp-zc4vH^IPtm7j zZzA`6Fe-@$$sjoN&@DZieg)9l~|5Uy`EQDQL_4-iSQUt0bv>bNn@4-XT1s(QB3k zFOTevysWMWG4+rs zpnKghCN`dj_~f5&?-34<-`*h|avi6~@48w2%Zs;8|E^`7QmzU5IyuR0opRlO8R@aI zOqVF;JVnk@u`P0JYTPc{8PXShOP=7Jr|Ab>jaJnuaA&Y3u?)pB^Fa$qS0d=}S|UQB zdGKdCm7mc@VYi~x=MZEG3h4WuJNf5%P=Q-88=K4GwK^o$)~F4V4jKz`w)hUKhuz}m zC%b3S+2MH_BaOScQ!`Q}U9;bL=bglWr6Gly8XCvftM+=fDQ*bXhnf}c;q-Q|Q&2C1 ze;USlu+{Q2x=C{QqKs{WCUlhjJ2gA+<{jJkjLWWOEP*3XStT1@*Iginhw`z^w z?D#?UB4iS;BS4uH8-<+;D6O1bNzU?meGfzOsE&qjv{(J)te+ep^5}c-STKw|8YhTQ zS--7a^}MHyjsmH2edwA%XAIy3IjU93FVR0x7L^9mUNsH`p#>d#x z>F}eR@G<4%|IYEZZKlM9ZRx_h>`E)|rGR2!Y080=(Tu~v%jTD94obJvJNU``Oxd#e zUGw5ZxSy_qI^R9Aa!t$heD`XvBOof(BW}~;6}iyX0DHVO3pdPkuKUgzSC6cZW2>@l zf!U+`I8il{3^?}20g^L~)LK2_a*6>XUrfb53QHQxS$T! zFEyZxzZj+1Dxg09rSIBt6*7mg9@t)QPw zF^Lpe4N)5)y_^aK0)r6JOc(<=5WLIZ!RnZHLCZd5+clI!FRJ{<*MB9fP8}m2zTZba&lh`jVwI#=KL5>RtTA(Y*sD#rNib+SmD%RZW zpptYV)n8QMiO$qmFy*FNl}Y9dG#1HnQ)sVHB2>0|rKi zye>nC1S){e Q#AwJB*hMz^E|A8(L_5;>&V6Zh&3#QbNA1z@f|Hf<>~hO`FZ@ z5FTAr?Q#4|L;NIYS=0@hQTM*Pm9F+U@`J6gDXY?4j6N9b+&!(;+}IBeG+0)*I`s`F za8O~IO@`+^D36teCb4#`CA`Kv1jNofa50dDf@+4tZtqSaEbJx7nO8BVBEe>GPQaLw ze)jVrFI)D!-zxsaTV#t1lP33BIi4jH16pSJRP1%n(=qL`IFSx>H0F-di$z6}PE%1= zo>v}}bm1kBeJ6xWC8-ml&$J5^{mo4P%>I?LFvQJ{Se&2Hzig)2QE$1yW$ABE{c<9e zs4g5LVW~vL0-@y3T?wrSf-bZXy~;r-ayCM~7wg-dg%#GhwHHLJ132kKwIS`sJc|{{ zQvS4vRJyPg`OwOW9Htnk)~JP^WXftuYeX-7e}2#O;^-#2CbS;dYFh#f>hr;^5q02u z%u{k16HUjqy9hk-UP16^x7Zy9e6|XV8})wc?TSj2nKOPFsX8a+uqxId)sT;f9*EY+uWJoOcys47Z4mk0S%xmvMK}@lpr`N?t+7g;))=s z!;FF=g9`t1NSHAhf_X#2jkceDrg^vL3BKo>^PJ}_-@`z~*;Y!sVagUiq{sM>e~YK< zQ(c&fyPzA~GsC$fx-=*>HqpW6(^!)T`E`&Og3YS)1F&H*lnLcP!~qk(@zm51a=cI) zP2x_0*Ju-u^IFOFs3QSpz%>72=R%U76|7Lhdn%#Jj>ahTEJH=e1pv|oIVEB39 zd@Kq*q-v2Y<6R?=N~SVM+M&mgk+(LaQJPHJq07+X*j0jbNRe&hW1;~*TSp%N3d(92 zi5+G2icZ;P3DSmTL(Hd$lTQ^)t&AuaHo$VV0s1dCO3{m75n!6f=zWCricK?a5+4StZ-wPb5>xs9}Es5nCt*nmccz zv?;tEpBtL2kaY_d|M1O$FRggxPlFz}B65B1sTT)s3~3HBdgswOQ;Ws9J~wCJa+OW* zoO^`yh!2K4kLJski0wd$V@raqFSPSIE0j<@9bcC}>z;jVLe!#ZmzQ!E6o(FQKMFH? zub+NcS)jVhE9I`Ap24XSw(&ovZ%;&xs#!=Z!C|Lp$Nv~RaQyz^!0XnF3Y$CwH%^JP zu>_gfoWGR(jS)8Ie=b2!qVA{H?raD&n;AblHR%k=^d!42bayeuKo811pxJ|3ypLX5 z8B@#OC&7%;o3pB9hvWs^3*k4(S-*Qh@t$}$*QZy0DsZ`{UU53|JY<-*%CYM1vuAgR zTqppRzfvo8{^Wac?lHSxsoUClD`tHErDr|GAxfAp7Ez>l^Is*~TnE)u%*8?*vYDNGAl5u^T`OH&d0%#r~=- z;Ktt&xPx@620cC$=Zn_C(KhaJ;T`yFpj){3bS5|lAM;$M_{IyD5%D_vlO0^9MQ%7^ zovXygIC0}}p^eNY%t$B>trcdAa-tS_Cs0k&JZQeztk?nM7*5TyvrDjcqsXq=ocGXB z-W%^&GalF=7HHPIdq(p}TIQcRuQn1}-+ZXdn}(ELz4F7d#s1L9Lu6uS#b=U%q|YGS z09_AI#ivE#aT|ZZ+Siu<6Pw?^-NLZ4VH`STAuD!2X}7gxj{JM^pRLPF zpA%hUqZoC+vP4oM!MGq)Z=jlKj7;VYGp(nPZs862e%?CYM`El&9Hw^Yq}vq+S-)hV z%L_alMmh`BI`Zh;*95)k6fW3df z2l3LZzzpzW(H*aeZZgFR8ji@=!7+2heY3>oO|$IpUv4IA*<~KMu^Xk#0vg#Avymbh zlr~#*bJkk#@1#-8BUDS-UhY&Yb9;Yo3a)4s$DN;*mv0?{{0xJR_(**yoBA|NSF{UL@ z4}>ZUc5zpGpHr?4DV}chzBtps-RL(M-Y6a5E`T)5{mLfkJs?sCdUZ%k!66m{vcoI~ zU^&1VCT|+_Nh}u`;w+Ajjg-rT5!j-#hevUQO7(sXL&_Trc-F zZ?$*UtC!*)kiD{Mf9$lNS8R(bm!frSsWIpS8CTN->I|?X}c6ln4h3-|AMQ76`f_t*F zkY~Nx`4kEx?|Sr4Hpm!#2G?9=g^U-Hy(+V;w>fT`RbXS|a(D8F;w6ykhRmYvuU?Kr zqNViE8*>{WN$mu`lfTusW_%R0Yz6gPK!bW?&Tqb99UF98XxIo^9wTMZt&-Kgx~t++ zdeNLFU=6JjSI;wLNN3Y6OtH7F@8ymlq}QtqtBnC>Y|K8`8Hi;+1Ace-(B!2S3Qr%; zb{YK=8?cPN%CLo}eb2MgBY(Id_Mc)lN;>AxzCyCt?ReaHPh4$b!jw=96r|=;+T5va z`~%{4#oAZ9<`lpB_Pd$>XCa04k>57Rdep<`x~F2}qV)&*=D`P}5`SdHAmv;jYCFSmnEQNy}qKJKRX z2f0WYSPIP%h*-|xnlp9~xxC=??0M#8$LIdxG_uQ$_rhl^7Lo>vIY5zmO54Ui2k9n@ zsq6A3)O}ftCWXrQ;r1V{e6{&2U2#jOBcbPoRlyYjns#+Y zuZEZMrnN!O_CXmcm%D0wpi5fB>5{g~pPWXoKo)h}Bwvo%0;|-)@TCz?-Y< z(&&cFXjdIa87m}FGq2}wuwIhHg!j8&{8Ib7pZ{jT|A=YEgHUdqWMCuZx=ETQzb3Do zyK*uR{2=XMqZgjsoK*pgkcQ|^ddQk^ves$;nyhx?WeJEeo(KnLQOpJk-ncd`s)O$3E{b>n`2jj4D6Xgbg?s3u zQ9A6v(F->1zPWluJM)m!;i~P+=Y+x<37?z6H5Y6nVU&IwFC0;Obk8uOT*Wljm~JF#E&npfvTG{EGMvM48bL%HdoMxFR;F>%ud` zoVMBHV%XLYJR zJA#sYc>%o`$T6^jxSejEyGe;92FXBWJq|;bjiA^Agw5=ZKf5!_*AQd|h$?aAKC)>v z1I3NKZF?(i>n|9HAV>&Dw28wPEo z*A{WBB!2o1rEc-OhN(N07iNInvPD@2HVZ@$JEISIE$pNd8RYIs1AfV(`Kjtf>QuF^ ziC;_ik~KV6CO(c2CSyu^TsLjQB%2Iwv+kGsOznToShcS%2tX2HjB-acCp};Xz9xZ z+}jbK9W`reg%8M$;q=D(nZ==|`wb$!V)Kjj!9Z^@Owx;N?*REoiV4WW@s!k%E>1TzN$MvUPpRh& z5D3&}Ov8GtT5x0`LnGeU9eizuUV&$>lgfBR0)lawR9K0`eOL7R*SgO>>*;O9f;}(e z<^)^xl~Sb?Rz>Fz)3jk;!8PdWv;~MpbxQ=5VacR>_>n@l$P>RXpyi zCW#G#;|{t!x^LJizi+M%X|Q!jc#Uk2=&&8~FztQxiodPp_R4h_EWE&=GZ#xHyXEJ> zp^w3!JbI&IAhw&{M7acXY`|r$V@A$hfiUXZRjmumCT-ll*eG)CxpA)zSs1`~Ddr=J z+=ikr&90fymQ0f`x+}+{E!(Bk($2!%p32r0FxR$#*Fe1cw#K4zgxeg23oK@ZHL9^~5Db_DD;jyTMy?RlFjq ziEmUr1eQ2tX3rx@P^Sx(y8KGnJ;<2JhUU60njWYl!SSi&hEK65(H|S`;%L)s%?f<3 zS#i$)9AsLdMc4+JeeDVyv0_g5aC_d-KTmscxX%PpSX#)o!$I|J|X z$_1Ca;RKDVElS*zE%G}!Wym92R1#Pblc(CEED44ngEw@x3C)bf>2(+VJ4v~4fDDjp z^7N^lD*Ni~|4%S93TCV?*0u9j<)1$_q;{l1zjJ_#v1YX=?N;x zjW+UhmxRXoKzq+S$4Q&)#GQ#<|8V>kbZXn}BJf zCj%Cyz7OVV()?~JUFxm?EYmc9=e#5L$5>_V7Bi-!w-EO=)TK&S4*>#U8W`H6OK-b8S zY^6d~NFO5SB|XwiNfD=?93#0?Q$<*xhEABOX#;+`&e=`Uo%2qL`z0Sh%`5JHQn|+Y zWzn_Vim+`m9fnJ=v4=s8w|cp`Q}rq&@=uJ(iXHN(7T=bm>o{MvhCx+^)khb@*H3Sf z4tbOaHZyqs$PfM0m?T^bbxT7YnCjOfPK!#D|80QVB*kqUPQ8zJDzHZim+%|&;qS&Q zKx5GG-`K`)k~S+6nd<@f=HO-y^kv};?7LRV!Kmsv-W|S)(TpWppRT-|^Ted=#_p0I zn(_DF6~7uFP3-XJ#`!cKTdZ~0DCROnS}ASq%;SnA&Uw;9|MfcbXgEf8h;R9y2Zz=D?}O^mto(h*)c)ZcHVFU-%5Jse#aln*VAQbX<#Q>Tc7 zlFgC^XjRlJujjoFSm_CX^u1X<(^hz56yqV+pl;*0@jGYZ6OH^k0UfYuFw9&8pv%x? zML=>7jaV<`)N(fktdyQ4$-I6^kqUcg8OS|#n{P!_f~Os;=>HAI#?pnRdp|M!&nMQb zL7z{daAPEm$qjVt)s)vv%s+VaPi`RFc(A=P{yAXs3YA$49o7ErJyOX|3%T(s*l3}J z_EXG0iqufr4hVVSdNa&9hc%rL&*3 zN{;Y5eRVhoI)0W2io{@t<9-iKF(fO4$J2cq!Gnm^cDf|e72QT-jYCy%2aSZAc4!lu za&)PhvH7h}zwy;OuY6>kspYnNb{kn)=))_e>cWer;byB?P(oMG6IW#LyEx`qM&ZPDCJowmUN1Fk0QyG))<*1Um{o{xjRe>()yz-Vk8B* z=ktJ^m+y?JBjfqyJ0N1aU^HyYIxzV-hD`5_dfz%P>+?2=8?P%i`uE)vV$B+ETl$2l zBH%l10>J+I-@r!dt#b_26bUm zV3s0{>nfnwj|dxEoN-8ftf96QOZL)P9BY1Vw_R9lFm4|g;;ywQG&{6hc$e2hOzE6i zipq!`@{1EyI`Zl3w;KbWj+mGO+ZY&SaiPQ>FGX5M7}+c?ZmeBw)E1ymqgQ!;(<}5Led>;&eo`zt#Vw=@RC-A0a+O9gW5aGx zeNL~q7FoQsQ+-z>PTi76*DF)2ijg%HW8;@eI%|Ljhip~-pGHrKp zR=sFsFwdoM#;O^Obj3?)lS`=rdQ(6x7dSwmbFa%00&;c`B(6mW%s#DCcw1lw$(x4B zZ<96bK;p)RDwMH3QDwc6VlwcLg(~YRar4}*zSZ-pLTlpQ+Wg98ai^?++pBKT+yYjR zK}dYT4k8Wop&utr*aF2G9xexl=?j*|e7FBuFRM4X&Krewf{VhuXz*=9ViW`qKkzFO zJRk?dOTHTR%gxN;(mUnb*gKCck?gfm*tl6y7GKUQraVEUHDbNU(OsaYd%TUQy<6 zP$h}5PU+(*x8;XO>by09PE|7Yf#Gx8lu<8?L-wB=%1)moeD&0u)^3sKWczEwiP8XD zl&jGg?S(jh5vLhC#OQVgW>3B(+!M1*+T`jDu9du4pvY>ZYfp3kH)Mm%yv}65^t(Tj zwC84>F<7iK`4j`Gm)Stk0hL1~&|qu-`Mid4Ayj_$)kAMT>+w1)GeaR_7pLy>1s)OuFFHF%xlM8!O*9lB~AgySS~HYGXHIszV(n-AB$^tmZ@J z(QB%S6+Q~mP5?}V%q_mdHULJ6zr670ms&lneWo^?jhIHW+Bbn*!9;fY1rvF~9!ij@@H*DLU9!IrOCS9Y=4 zpSZD0-9~|0?$iumo1IaVc^@TMV*X3BPgQOZaXMyV5=ATbz$$O?@7!C`F z+vPgF>ISzY^wAe;xg$4p7`Qz^vZkyM40#+ON60;&dp?D9lb4CSf5^kAXrp_ig>2Dg@8HCNXpEr7hlM%h5{7A}4Pe5*md+BaUj^rddv?u`z6-S~u>bmVXZ zA5ZzTtsB4n{)2he+Zi_M#(fnjY;m7-uclB;5=B;H4r}=ggL++LCPZu?(+Cqzn-vL^ z-7PoVn2e@AoP!4&v;3;}J0X8CuSvUq?kgp? zpRqM*rA4XNLyGCA$bD#@3^}NUG;~|9pJr4*A8o7_-3K=KX4Qi)t>qdZ1`JFUu!n4eg@(?C#{Ls;63LG2Mwfqx)kQ0wvs7Bd# zjzNu^s-(!OX~SQGOV|x4r_rnW2~cG`g4L>*i_gKI&?Fu3ZDlq~?(%RM-0F+1ixPk$ zqgQq+&~#mw>E8~2aGHFDAO8JD*s2YAV6{&tJYrCP_8{4Z+jzl;`f~3g3s-+P}gW5=y7@uj;R0bezO%W)FM(HVK zySz(U?jP@poMh*S!N%`(9G7j2VU!)N`|d#Y{z!M$x_PFJ9+kKKu)5O32ngN6`bll% zfDckw?+YKCb{4{MmB6Y4rK)m_odG8*vC6;O-abc4tC#v8T3AU(hgov zF{u$9*T`JVuvG|8uB`%=oh&^-2ie3A@$ zu)@3>?~pcHAR~!lR#PN_(pGR*&D|3Xg?c7t!Xoki!QKeekUbiYt+sN;^I3nZng5R! zDemD7^7IN*>lfW-=DMtt#!FSf3Em5_9JFal7jBsbd=?2l2tY52K{#v#-%}fd3hAq% zy)G-uL3!GP6PpI9GXP;@j<7Ep3V&)=9O-}k>X*q`Pjb)V7k7hVu2bYHrPYTtN~(ez zy}Id#bT1ItoFWH!Et)%C2cib2RRy%i*1QD013i0ZU!IoWIp~qYWQ1=C9l6@c>5KVT znBZwpZ>6ySL#PnS&JmbUmh3^(8c(%)10*%nGunzhcn*Ls-yshQh;iw&# zglvk=o*OnqW4_EvkGro~+xE`1EAMhm5hNi1jZ=rrKzBlo z#qEf!C{#-u{S$$DbjYJw(GiUdVBmep1Ip1xFC-700PB^d(9J5#iPE((6)Vm5!Vy7} zILy3KHN3j6n&f$s{T3@#6~#aXd^x4XoYgL10?tFb`N-AaBFdgz9hWBEXIUY!GN!>i(Do+S24GbZ1zjx{gW-5h z!w&IQAf?07hO;w`LF>YdLARz1#u|e#ue}Yj7K_A3c`3fNA)jTi>lH44!;ZaQ7q*=; zi;mqd7D}Iq_y1^r{d?^n{`~XbEchSsa*A0>kwvcSj%6ppUhx0Ty+E@n`Pr#SXGkVH zRpQ2e``s3*q?lp~DUwHNcg3#Zt(^isqVR!XKRsq;8`LLXLF2kra$)|8ux8*-%MZZ& zNumx(M>LY9f&JwX`jzLqVdDyQKqiT^rvd;d7jP

    t!(!}z)#PkQY8z4-A(bvojsQVW=yIpF@iQgu zb9M#}pJFCJVpMHpQse{afO9>y;|9W~h;Xh#WT*VTe7Nf0t&=ZptEp zyk1o7wJ!9&biZ<|Z%x<|!EVJ82J(f3i#eM)N9js==Nv5T+{V2mG@WhX7J7doD-kC} zLXSl5wSY%Z_=z*Lf%ys4uaTzFd7(?aTQtCy#-zw|yfdb?C{93Y^$Wgjz`JMq?whmH zcsN6gW`&?6dJ7c%EsA&~uT`!B5|g!TGiseFzb8aF$|1YoI4k+NueC_2qPG8+dGVUF zxML={`a zugH`bWLxR3Ij0DAX2KEe!H+{XNE-PU!|Q=2Eh#odl*C!exf)tEZI$K$NhW#)Rus2r zj**V&OjR4fnE`srV&z${jwxH<3ig5Op=-c|T`$_|TdK}f)$(!X77fm6nz2-}HE6-0 zN4dJf>jUAS#}cMTyq23ChF{eN0fTHvbIDAg< z)~=q}Zq4b>CQ#tU#?n@cpG_vktfxpSr9Grffd}N%fKZ}OmXtvIPaa-faMo!B{uY~Y6ep*zG@FZyfvK5R~ zUL?KT=IG>CI{D4fcm7%IJ`5vv`b^?ibJ03&&TWxlBcBhOy6Rk9B?)(iu(pPh>h%UY_*ZE zYNT_bfbAT)TQ&t?iR=>TYG7W$9#t1P#`!}Yt9=iu3b;4Hy1F6X&s*pFNQ{j+&+cXU=^u|q^K@DxC#bwd;!EO9@;KHrvBt@R} zYM}Q`%?g1=BMGtBgax$m#b#-CSR4PocP$iapCwIxWzo7VQHe46D(tYn;Gk-+sw_I6 zUKz6hIn*&`S2zP%+aXt-bke#Mb+HKljD4hozAx^ivEiQXXiWDoFMzI7mh9!W+2Q;9 zuf|!I(y~FH`|b>gjf&unevSOKA_HA6Hby3RwPz`^t{WVA3+aS%v^7t(mdt< ztoyUx=Qc_q^tr}wK%5x0k0km-d&oU8z<-K8WV~v9`#?FQs2ibikgivj1-)KNl;zNr z0rl9B2eweVCf}?=bGl2~4*6yk!uT1OPgX}?gW^hbQlX)38;C3yL0g+PN3YvlIN2?N z!jJWPtvkiIt$k#pD-3p&I>l{Q7~_gUlYkHgX-r$0;of?{8inheF$l>9iUrMzx8CZN zwS;Fti2SX$Y|GZ>%Vl-LVT7>y7WNObL&)x+nZFp%LgmJ0l8uF`gU(YS+n6qmYXs7Y z!k3mwfmAAi7}Nz(Sk#ut=(^}!I-i?7eQXuMUNp8xH~=QwlR5Fuk9yY3nr~jHnt!}3 zpH#C8V!5#$bIxK}J3=wQhWI|EEmvLjzpXqIIP%%7Sx=pK{f4IaCmVjY{Iz3o*!=@+ zUXl$Xd3m`aynbNG@?8X zKL=%1u=e%P>p8jPk$7j!m6?YG>(r)lmqvcKa2(pp0W7gtrW^;?%OC8yVa?nAc`hn9 zMvM)1_Oz%Sq5*j+_nLeK?3DKL3b-BoYI*W*T)Gecom=L&RB$Nxw`)`zVG}ecUjs@ZiSyexGsA}G)crQ|yKO(a{9hO19{=d>ue!c+ zAr8N+IqpwKRnTc@_8~#LteSfG#eVgDX)<)FGkwu-aT=WD(vJ+duASsw8`1 z)Gk$!8FQ+{mHWshPqNnnb7d3*HfABE-K&J4qhOGCIHupHIP@@G$$21eV$SfI{T=~Q z>X=4ZzxM{owE3$@8AE)-XB7|!DlONcxCZ-6MuOddJiEF$1=_(?-PonvL1-J zmk9R#=bBdsJqA6D-mU_xC1MEA*LQALlA_UQY24TVZm>{) z)fBUnA_huZC%GbCI`^Dtz-OtTja~uuo85E)Hw{*=3*q}AW21!bmn@4g@_J~@3b`#l z%-;np@^!NK8C%5d5DRUUH_&&XWo!DBe#jGP6V?H>(H5~@a9-F=9|=7L9lvvZ_Dpjo z(j35(O|9t6L{r;%diz)Bz7}rAPGxt+5>m)+Bksn2nu8Xo+C?!{6se%J@4STpA?!Qf z6xpn3l@Iuoa(n1)ze3NkUQrOWKOAC7b;?fFyTcEr$=~PR1`DkbYVe>M0y4@V3WmPJ zUi9|@Kh}Oy7v4dirn{s?oc+po*UEJmtSS){I4Sb(R7{!B<`k{9C@S=G*%P1NcDpe? zY-ChqhjhprNDlbyhFPC6cF-!>6aaDaLtZIrq)r^mzweY+YcUBc`CE*1?TKxjvgG5R z4q5jqVH2cuW6ap-Y=W`xC4x*T#F4SBz+PzW3KhH1gwdd`ng$mOxxLVPF)y-kUK_vP zCsj4%G30?=!TV-a&&#HFiE@K<$WWgWSq_^z3^w(JB~D+saI5cW;30<&<0k5QbSXQD zn1mImY%fty-f;@rzW3IzuUe;Mv4M^oyJ>8wo7M7NQeBEalnhIU^^HmGVD|d$2&5HR zBFGjw(?ofim{=j>X_Gne{;j~5FIXeP#w*Q%veuik8t8uYfmyY%reBk1OnU%%Df^~^ zqeXZca-t^kg|ez`b>a(q`c$?x_%r2D`8Xu%PcNBG*l(Bo_ixBXb|$PFZ;bX>n6RZ3 z1MDCLl=hsm+Gh=~E^@D;P_l38VOgy(A5z~q7X#U~v_aVeG@F-U z>x6~8Et=Gj9!QcvCuE(gehjP>$&(Aor^io6up>{Xt6uJ~_NTB}V%*p|v61PJuga#; z?T;BN>!&Z_Vb+^L{gE__;5q?#OOx{5aNiyS&#X< zvG8BULwDxJMu?5ZSf_;<;SeJZ$rl`^TO)O*(^};Y5hPI!TP2u>RZAyvOx8&e=Ynr3 zuU2UgZIg9^8>PclI>aP82XNwI9b)>sUc9FjiOdU6E>y-}z(k{_=L2nd9q{$aWtRtPWk$Wx%mm{z;sS*dU z!eU5c-P#|&X$DHpg9cCX5j&u`u}!kZf^vS4VxZi!pVBU4%A@xylfWQF?szCj6RwsY zod(&sz2LAP@U4?zQD%H#w~ww_k;q_CTOqy9uhn1I5Bc8%zL&(sGwLMo2QCj8⪚Y zpZC2peN#hD!+!6QdU;5>u#LVRa9micyd*RR4S8S#11t{M1C;?o9(}VOOOaE4fXK0~+U}vD)#@63n@B^N;$>X*Oy*7-P zr4h@%dlj@!i>8-sRIK4`5f_t<{=M8@{s+PN;&iSNLPl8DdyKPEzB_y|XSKW@=mQG4 znUaGd*YdTZCU98wuuvn8Jj!lGt33Sf=BZ{ywd~C)Ye=>yF_Ov|%o%^4`&z>JtgO(nMpsvN`st=DrhagO zl=;68vEP$Z9>LH5@$4*6$!oojkuAP0nszWa4h9!Wu5-Gft^{XX%4wH3$_j!1H!HS9 zqgVCNS3>)^)iL!DGkzS{L-+gKoaL|~;siK-2CPwzJ_885qvr%pl~LROXdPSneALm6 zU3xYmNX-ETbrHuvH_&>5soq&3RRQT!Ptzrmny}+!x1t7;0Zq>sp?GqEWyC>Xoq`H5$QL*X zM~P9xts+__S@Oz=ArFl7VWX@|Vu%fc!53Qz+nHT2tpc0YiSK!o_b`rEKgw**yJY*- zj}lLCUng1Q`h_2lhcm>Dm5&Y1ker|z{t|MPz891ibwhlF+ZE9kQp;Hz@_1%tKpJpJOLen07vW=b#<0BXUomvNf>(JN{4 zQpuTDo2D&c?$O&-=jbcp=af)ss05la?2Wc5WJ}Z|ENRD3O?KFMiOWJXj&Djv43nq5 zE(hd+qdBhb$=tvo*TE$W(yNOm10sTl0<2nr!5k6inn<|JCyXu z>UsNtZ0Oe9GcO~7{lsQXA@Jv(Eliq~sMV&Mzrq%GVh`CshGXcy_cE8#YMPd4M^o`j)j+NqYUlYXhcP~Uv zGAj<%mR)a?#~#3`{G`V{kz!U+WCf+&6IMv)!?LnVUK!P4f8S=i->7{LJ{=GC&ExE+ zSmD8a7m~RVJ|QoRxZPGn93(>?ZtL(Nn=F93mSWaWWF-)S4=2(V3v?${kHo8eL6z;B z>)L(N=wNy}R6$pbKGdaGSwZ!M17{L^tQpEa?+0;T4F_pVPpY#G6q81gbwFPjoW!dP zOD4Lj0gGY{YFzlRUp?3p#ezFtxZQ5ybW8+G{D`dWg3G6?+{h@`f7*tN_y?s2zh+hz zhyMHGhva}K0cxEm!nMs5(?pSzl(r^(c~qyYdR~hr-gCKTzqp3KSGgSep8d6!#(qOv zxXYt@XmCnYe+G>&5Zm0%gAAiU&kJb*+r2mN?X7!3horXd0Bp6x+LUJ zrVrCKQ~W9$(kd}p)kGs;F0k={3USv0Ih#TNO`hpI5>@gu>*o}Oe&fyjL^+S`eLa0WpljthEE4+2X z*>2x$lF`)GFzat^Xnr8)k$9XOtqJ-bac> zcf1x0@_dFou94fmrWP0mSu(FgP#Ms{Hv~XsStAWZBOIhfMUq7r27Ob?>4Fd)G~89C za@PVU;xKojdA=nCDh3oC;<`}mwc*gB#p+kV-ZUE?;mEJ!r!Q4S{9^@ux@X>@Z_G7_ z?#Q>oN`%b4J+cCNXG9Wbhxn#!r5vevd$~1p91^o5B4FXLby&=f{atUS-21k*AKz`W z<7{~5uR~_sU#`pF#XRr*q@LIPpSvRtLyPSM&w5FtWCDeZBCuZzI3_ShN3mKPsl9!# zJ+|&bZKHj2Ata-&`{E{5y-J5go$=H61r2&U6fTqYbJxz*rStZ@+#%_loy19txFLUM zf`nne5JvQ{qpy|SI95V%-(|dLo#(HtTW@{dL+8E*maTSNc|c3q8C^&lNT&)qsx4e8 z&ywE-PWS;io-ND_LxOly1^j?Kli#aORkh0(QKgc5L55)6J0AozN2bRPKe{ka1@2%l zDts0T`iUtQyB28FuxmjZ0oF3cK_lZcys^jl2-ctW`Pp|ka%3Ch(;KgbTUWE%C_+!f@#x$KUJR-4`m)*IRyK`um()Es z{5Z%hj{iJjD3coh*Tv6#udd2!LoEPevIU}cdJ145&rzx_K#@aT# zeK+C+Uw4DMSBV51y8TLs2!-AANt4$`V$-S<5hq?*3m#e=iRb5(1rW(V4=^%=ja0BA zYX?j~bUWvr6t586h7LImbTX%r?vTAV`I2`%lssX8Brzr{Hi24*eqKDhz#xm^9j`Qb zu4-t`fZrYda33w*N?=SH8Q``xtE_$(lTID({5j7IzZn_dKMeI{>?UyEgsVIY#hOkr z?@=U~(l$wvMJy>20|vM&Uo%WNzF%2CA5uo0$4_amrn5`Enrs$(r*z75@!`t@erE>8 z&3S95lbh^-;l3J|tgz5xpHNI6MS3aiwz%tZU6VAP*Q`k9p(^O0)4np`)5{+&u3zfg zJsUem>fX5yTPr-eKYVk@Bc6w;$NQDYkGau5ai*?X)DeWru;g34{2>pFFtvRLk3Rpu zWL~crwb3%E-ZyRXkjHt_?}Id|9dr|v|943CPrD>+l~e~0SM$RTPkl)|FnQ-Z*w#sU zxtIXfDlw7LWQ6An>XeC0oe~0E(A=*WcmWD$T%Ub?c8j7#v4{Q;cx>K>=d(pQQ3+H6 zWir57U~f#T*@-uYvHP3K-}U&KvG>bMWC2NeZm_q_LMi1^448jel=eOOV!?2at@WJ6 zuz)TnrE$if0iR;gUR5#=KzG{@E9i7*=3&#d+kW65JP>stDv5MOJBj(yrp|C2BQ~dW z%8i;YeR<`0QfJ&a@5)B%%njJ3p^jQiVX!LKr(a#q!QB@$KoWJ$S?<$7UnB+5MbWqC zT;w!M-t%5p$&?%;gYw0kJk@?>epIf{fYYg} z9LJ57HgO7CKAYrE-}?Xf-#6=)9BSa3a*Ps-Sq{uL8TAb5nT)stC{YiqtP&mJb9S;8lwc1)t-s@Wff#AdZW|ApvL&sxIlYSua$to`vUr*q0QaU)1kOM zDe%e{wP@=7`h|z0t3_>5Cxk7UwSln8icgBSa;l-_#BRk#zir%cwP&h^$YwDrxt z*&Tn%r+ne}UNOUD&RZX}5u+P>2f=E3LhvUj<|suPD6NSr)wHbz<406I4=QufI>JN- zY%kI(KO3#v#=Rs&PNLmjP0~_Mqt_h>TeScK){ycngpG9+cU`;U7z_-yd? zsOwR6^f~{t{)IGn60tL}=^OSdyJ7Q;(Ws(Oz3(|?r|eYVkO%hpLqb*^ZqQ2|QCyB9 zWHi*+zQ!IYtklj6b21MFTi3_AZM4cpEnFSY4rVBuz~DX@*b#`i?e&~m?n&ioHp1K1ZKG0U92fE^#f3L4bSDGy&AkJH7CS0o!wm)pEUe#XeYm?r+9N4xAy z;K%T}bjG3JL620Qt&(0f9_rhE|G#5I*F;1Z?7;6x#v zF|AVtjVeh!C*8LZ2vlL^gPu2sygDQu(de44;IQ3bIRq)&3)NV6?8B1zsg&2vYMHOM&WZ)e|GNB9H9-jkZpp>np zA4toE*Wvm^{8AnIt+IMSjy}vv`@3(x+q<;##&|kZyYIh@*yvD=9k4Jc4F!%^w}(W~ zY4RGyKRZ-E@k{7Cv}KL8_uq!wJYO{sn@rBjkH&11)l1s=@t&7sYUZ3G?=JV$hcrrd zkXxWP>Yz~!2H%!)Ob?Y!y(y@sQkZmZ#*2DIyl2_e!rrz^R-q1if40l9^CzAj49911&)mzc3P!60x9VuSoSXtt@_bcM=n7t~ zaDRBS0!xWMCPt?bQ9JK9;M`4@YXw7BpFkhHKuE)>3qQ%jf*WX$ARJmjX1^>{ObSJe%EZ% zzR~%{4RYL#y>_=OjO~jQbDkn+DQylgc-*2pRekgoakJ*5uVM>L5vGWlRBmS-dpt4&AJ|5PpX~2X5S(!V{1=ToI5>U-CYqsp94_eZoG@XZJIq zEd<_a(LjeSUb+7czvAd^5vS=JKG^U4h9+Z1{@g*2GH>j_HRzE(bqVQ^4|-gWP7hoz z9rV~JFW^=M40RcE2}^Ki2URyI;TKU*XV5 zcCX)UGaj4X>N#e;RdQSBu8qylC1HK=RysaBpKjtCf;WWq5ZxBv3eNR_3-c!$t#*6i z9Kzf<2C;(c3!Y13rU=a1;!VTkx5*keUW|$@)YwLf$-upYwtsfJJi%+T_~cC8 z=vAUDP@N?Z5XPwSHTnJ^U7l(Wv_NyYh(D76J7Iq|&ecb(^v9&1T$+1FY1SVzvj3h! z3Z5JNalk@CPA=-bn-8f z4shP4&uGz{ru#{&yhD=AE94fbN}^jd*W|mDX%U|Uw((0LnfCy&6x>(sjJe@spwB7G z{4PcIJCF8o7%A*v_QYEC`9r57?`}-L$eJdGO^LqyIvJtO+moBIEQ;AcffTeY-?##s zXq5Ue-B1pv5$bBwqEhEBpMyM8dPQc~UAmI9)i-70T8gkmhf^Ui#$<;l{r$N09@lO6 zGB!5fSg@m4H1SK6XQvqC=E!cEe2E}EYKL+{ml}N0_DgyUvAEzCtL5f}nQ{Mn)tcDD zZE<0Ps-v1I7h-@9tGk9fSuA5(fMEf;y-F_$C(r;H7N*^>aDnK@nvxwX-d*#Ze+@EQ zA*#fc`^cu}#tPYMv6z%m4D68$DQ#lN(ugba20GKfG4imiUH*Xd05eCe;<&s;)Aw>m z5SF)A1hnysMUTNyY35{b&iLP()yrKOlcUP?Z_#`rg~VFu_S+Gi%iYCw73%Cp*XNB5 z)0MFxGLC!vW&Dou@LIZY9MDEHGh{NwCd|Ohf-OLyZ$KMAU$6#JkAOzGNqUVGaWbWu z(pus53Cf#cvm%+%F%E&V8G36dXU1H!GRceozX9P& zaJoUWKf0b%$4`n(;^1=!!}G))q>+2x6-JSBAZ7^G0y z0uvXe4p!Ku^3P@lmHGk=T%h z<6^flp_4AN(=2~HkmDO{)+~>iV=IZ?jaR4|3tdr8F_7N7mC{}fsFTzv&kOa6M%i{w z3)#*fanMGIAz(fA_9ki-kaV5~=Ne=PL>%CiMx};qSMCkX0Og@q;L#`2o3onzaCj}h zH0nP0Fzd!R_DRlhzK20d6!;KuguMhkc)QOs(J zBv4wQ$iOrM)eH2BjPPOs zW?2t;v_R9DHPSSBqwFqkkzluYJT|PgnCwBqrz+#A2iCJYe$Gx0tXX=dxk;& zMW%o0{CpbkvdEuwF1VhGKDz_XtiVORH`nK<|5!!wH~$s0EAPt=K-v3&z&v^hso`Uu zCGz~VXtr?MNdN4+aJ3Y&6w;z@i@Rfx9kD~UGOU!i%4hefDsdE4pHASSgD?K#)*qgA zHjR8$2ec|`muCsGV(+Q%`G9@0(3z`^y|3Or8QVLX^PXd+I3{J}zLjRJIBd{RpwDZG zK(89^ri!fpX>w!aAT%i)-~x+45ho|AE__9l$v2kk(<{#hLK`H2+6Zax$^YD){_f}6 zw}0{bH-4uT(~Jk9s8PBCxNpT>LSlK6brXr_B;~+HFO157e#ofLS7pPO53p~eGnqz+1l7{_=48XQ?66do>1m)j3vHkmH1g{?cwvEyRvi0u z&{#%^^YqS>Gu6)PY!#Vp-CTb9HInq)_-{)sEMq;zWKm=T1w|jFoHl+oofE#!@AmBd z;alk{S-<$4u+TdtMz1)lJbLLgdY56A=;1fuJ!|O-N zkABO%sN9@W|1ml8+$<`c7K_SxiaAS>4=62Wqa-jFB9291crb%=SDqxoF0F?H)2E>2 zaaphi5@nk?Ss^|2rNA}3HhwuI#TJ0gVhA{|+N@eVWdpE09RQx&Mt--jC*Ud&e?iEI zUjQ`8e;0szI$`a%T8VnxUM&+ij*2fia zF)weBGzl+@jFF}4y)hm9d^TAR4l1ItESNFwS>k7`Gr_yD_h~b>`hE~{j4W|u=Uk2j zTGmm_T8gZpv`FlNdo+W(SYTpp%ZAQY9c;ui)+=Ij&pt7;nbA4V+Fox)*^Rx>@$I~z zNoG)}w(NSFJoa$g3Pzwpe=>7Akz!U+WCf)yq|y2?M0e7<9i)ywL1w45w4RO!>k$)k z&I%8crfxl66JW+e^DlNTBpGh3Bz9Tgp@?FD6L1UWnXCZ64fgY{fDUHE>=n(NCLqG@ zmi7~Dy^!Kx6}&{yC)6Dvy|R|TyJ}rQR3cMIf8YxZ83N7|Lo{j?7v^9;9V=sKe7NN* zRx{bfqx5+n&qdUfA#l8;& zBCSxu51oF}BpT##|0L{TO#gS1FpiLw+b#dHSYF4gT9 z6Q=8Oc>M%VPEJ`g1-r%+LW(Ys{pG7l#Cw(4#M9I^xQ&mUK04`S9@2ohgxJnGxOW8E zV~^SKUq_RzD^H&jz}TR-y*_(r_ML#k^Z`;h!%Y5ogj57zfB6io3?qE z6*O$l;`|$u$EW|rx)g{FG~C#^Vxw>f2{5~9Y+R>nl@z|TOsd1RuT|11ONQQe6Nw3s zrv0#RzUarA@y$--FQ=GomX7(euaGP^wr{H~jHwcef$EKXN?R$a3CmOM3~W}M#x99X zQnZSiB8@?8*2FXVj*t$wFmnd?9(qeNbZo{SO=?D zxO@q76eDAZ*D1Hce%o0nF{>frUEv={GCMWo#$*WP7HTMmVm47ElhRfM99AN?V2dU` zDuHuO*vl>Oxiz4{fs{30B?8jIv(^qfOZK?PE?JmM{Melu-bQ+ zv>2`7b?{89u z?JvR(>yj`vL>ROm!%{9WTP2kqIVZ>pH{MF;TZn!= z#iUXsh0>y^F)Gka_s)if%E0D_6bBeDfTqfLI>C+;HnO%;Oip=>oeKG8-o4}DBgp8k zSV9UtM~OYVu~mA|0xi2J1~f+nrOhRe#D(-muZ@2Cnf1XX^ctRC(awR`lHjgSn!I!N zHtxo0gR+#EU0meu>ZYqWDLnnmMS?~?X0Qw?jX`O0-0`;2D*_8u7_DlT7fm}xI%N8p zu;1ZSIN7{mgC-jb(#U0}plD96<%h)wc|#sc zzSR5On{jWA+c;qhi4nE&w9oNE#vUGE?xYtMEe&~`7glf%5M-;xu5c!L^g?xY zn2+wR-|hLst$ftT$AhIlz*ZZ&SGh>Ah51B}W-$KfTLmu_KGH!w+ope!}4g%?E%I1mtPsPlr53mE$=?!<< zRc0U^)&A{0Qu*9?Z5u6Mx1VD6QIO%HEe?riZmDkZM?PCL=b%pvW^>mE_VAh%8@Br)V&G zb_VWHHY*^=T@#M&Rlzxf-n|lKd9=|R?H?pcHGmcxu1IoyR2xJ$KHbDh_N*RWrv9?* z8)j`|@PG7k(%{C+R)>ZDI7czCjx|x*0r++I&>7+P0`-a>-mNJI#76G{x`&s+U_fBd zV}QOFIOvfRa4)cka|pUiHS!k;_Co35F7D%*=YX`Rm*g;m9vf80go7TdGzJY+hJs>IV{Q+fK4pjBCvn|jOC#RVHNYn_ z$qPA63e)OTC2pPz>lScqtO%?RED9;}O9)TqK&WcwrNC-Yol`kCoQfS5*m&Bb?64C) zYiR1T-d-qWiZOcCTyVo+DoZl2S3KZb8`(E2LzT>Fq;&)0#Heg~k)Y6LVXFi)QO3+d zdD`2x`@SRauzT`xJje=gdKJcXKk+@X?xiGjVh9re{3=l?^o9 z*rQ^j9yVWN(ONYdncq9JkH7-UcvLKP+81Ftm)HiO4Sw?yzLwi^lt4u+S~ zn2Xy_5(GtxWWl8n-GEPm;DBRAlXl33C00JwC7&MH@g0slGUGqKnf$Et;I{Eo)U|YT zbd3vEV(OO0F?ZJ3tBXr0t+qyre1C4_l`JJl2 zbP+>f&leYS9tU=+aE?WSzL-IeWubexgC6zt4uUZR_Na9Lfg?*7n@`sGe0%%8%hnlo zZo4$u$eFw4b7o$vqy_jwbdN(Xa(2o#Gp)#+kOw)HC4vtAdH=F#@YmKxc23xShqbs_ z3(m0ttQ5|q=poHbYl^VvWN&1HQ0#Qjc4d#WkUpx?-JY`@7-R}(#FI@E-+vXPYbS&p z!@)R4X|L2SdDg9zZ2GvOjlWHH@@3sIQWlLq67Z08K%zmrylUEjpAL!Tkc|ilDbbZO zAyFJVC2qMMj6;(44rQe@Ug&8*U8gZyDIHNSWs<#~9MI( zv3Fn>Y1Rz*+692#%Katr#&PUh9TJ+Q2**0JS4N+Ej+M~dxAsCiC$H6d336Ne#Kxk6$^;mak*Kmp z)hZm$S%NG|P%G^1$j`}{-35Em%5K*2Kg|lQlg@}D4q8)byDcbe&}P58O|*#8D?XT! zr>c!C=Oy}M)*#pF zdQlHAfivh)#LW>8dXxlK#N+|m4saiXI}f=Nk~o*igP3I4T_-?}?PY>aUd+E8;BM!c zvg!1Svgx_L^_<(You{)L!kPu}EKbG?u}{ap*K*&3F{8)lVc!L^o*jDJxa4_<1$qi8 z1~LzFvE&(7BOv%OGliY{u%eIAc_ zczClDH#lLJN-9Y@Y)k#+vJhb*^U9((tLxR7Oa_$6qY~z|8SPQG=#x>0=M`!)r=nVr zF;cHO%X|{n4yp>}>Rhjz(gbGLqSlBH$ z2FCm4@Q?fH>ecr?JlEEZwk>VSWu-@Q2>@=)caSeiL?i^jlDARmEnU61&lq<;Qs z*$_F;gN9aVMnr#fAupLJh|Hd~M9|IKFW3?Y3n*cwPrrP3@G9@5$piCG_!jUF(Mu+G zL(l#I$sl!t%gP~WG`g&;SH%N6Ocq@d)eX37*Y0tz^f83?%rSpUorn2qLuempVs!~G zE4RgF@VmTP`33ZDO)WGp)vH%9Iht-*e5Ja7AiFUqo7e8$s2Tm8$xIsjRW=BoFz`l$ za6;SGuniCW=o4GU7nhyNSlC?>pquoscNCx^k}4>TM6R(^!Ft{zQ4;8z;o3NvxuS%u zVw2$Hbm*jRgm7}6W`Lg)pa(6YOxZBdrY@bE4&7)u0hc0lJ&~K$d#CH-MZIDyow={r zAoFD>&LxYe+WAIiQ1(D1wGMM_xji^$5K{tA)@tgd0g}e{= zsAhjq(@3xK#tWPLFL(@l_X!`k?E&iL6rRBYg&j?8)u%xXryi<;$3ig&tW9XPdE(7u zV3!-#q7@!hn|lWce~shH7O7Q zE>mw(XG^oCSLfZ4=gsd{)KJTqi=r-JpHHH^B%+RKq*KDL%sDGQ9JE5QE_5yQLmvp) zqw4jo040e;d7@w4{4VISZiLFH8hW*|7j@^!AP;#zb}#sCpI=tc@*vFHwz`-5^wJ{_ z9SbUBjM8I=oDQeQzR|t|oZv+L^4sm}pZ6&Z4GYF0bxDFG0V+$F#3Y7DJwJYKfo5!D zVBc>!o~tU|m2PK_gThb6k3y zqzXF4+3xMy1W}zwt*}$trd{#PPU$J_(43Tzbpfav{W*jV^P&nu3cV|NBTwjXAFt(j z>JV~((3q3WQ@>!G7qCO<{r`FUwHV8+Q%E*-1k4aBU*S@|nKsK$WV@%$!ThE!jo?!=l$XgzArp|Fo+xU@gwp7P4oA&c zymx8eI`28zf+Vf+T^(7(EiB~1LBuR8+@(=W3Po0$bz;$t-VmZYs7{fLFatTwy08zU zeF$Q|-&tpN-(7pt785QTv0)iyTruIY>VX?pTy;UvH!Hj2j*EbEmIE7JqJLC6OR-H?qmpH2#C2tgU3{eo)Q06)?5x)3X>bx5Cb z&v!5i2>Z|TkiE-vOSex#FiIGnn_t1hXMw6=J;=45;b9~NW{AD&dT39*=dj`vy9Q{q zVSd{76Ml3qr??G}%VxU=61|oLTV|N%*VfdKJQwy=_FByu+bO1kB4t$E=73(YUbR$k zSd`;w6k=qzOK23fg38oE|FhyA1+o!-tVjptD)d5jD064DX^r13`COdg-5nPPEJ_`vd%V#!0C?yxVtj%W(p@UHl2LbDVa6|{|a~yqw8#*qZ;aA)8 z|G2D+!oukD_U$RlV{1GY&x3~8%VY^GR7^*VzVS-nE>!m`9w^^_cDS&&$H+4IIePfn zpT`Xt3-|I4Jnv>>Bq2~3v_e}MmK2;Eu`xE0ft0cY+Uvnf>Q>fG?{n7&wmchMGt=ooByrFb~(hQ_R58sf>>5{ z1B#7+EMV_|{~75d0fMhcQX3j?q6I=a=2XElQcq_|FxiBDjwx7&Ei3UNtkcq~YD14G zFy>bkRU2xQ<{9&>a#(Yx?+ zsMcyZR7Nqt^s$wSgMLik4%rDX8M+q!Sy15Yl`lkj)O9O0LI(PS;C`a3 zkr=1r#RKzC#b5!A!>q$&Lp5QnJoZG#9N>nPfiIPw3AGrPzr5dFOSW>0=DYB;bkNEa z?V=dyt*Hc+K4mgfBteW^jK1ZYIB#`eI)8Owr5Y>m7Kx5}1BV^j7vQ!N;|aZLv#d*| zO9(bjZ{(-(&3EWkU9w`CZk2Zm55qi0aI8>tX@z%@q|5t6P#x&yIBp#_VW3%iJdQyb zJY`!oojT6~oh;>>MP$ch@{!eVet=>iJk~(PwF>)`8DZ=A2~!}o>p86Kfk>F`5sZZ0 zla6S77!|t%&!{s!awJFqbx$yI{dQ^Xt9!vS=<`VrEK`)L@mcL!{P)h->$C3(+Psf4 z&Gg@HEzo@t*`v8S@2)idhrPiQ8w>{z!06$a&|q+S(_Oa`;D6o{P|tfmc0c{`f==lo zDEb{DSk`ndZ0o#3yhhXhK*J^F1S~EWr2`=FG%)P`?jeH=ms+GWi`ja zbyX;R{nKl6{<*+WECPYP1W~`R)E|~p=SjD+d%+eOMAIMz8c`mZ>fRxSUhI|f54DeOCvf+u9{;m6t#w>2u#kx>LDFZR##DT?Op!3^b1qd!ah) z!i5X5IoxqF_T$FFXuoLmHrGLCxt`OUGwIW1IbXBoe1E*W(uHf8SeV^SEZvvM4X>1N z3<|H%UWx=Gg4yXVVS3=s*lRu)LiMVN`O7Q=F!o_?a2fj<&N#{rm%yI{Q7X&2M|9;Qa`Fe(i6A>O5cb^JrV>-z9(2psFuh%yCP@rKS^FCQ`)=z)8hjo=>;ACkDh3Hwu|oF*f9Mrwz=j(> zM%#^1+fD!oJ3huJhN9wsS3PGU@|`XmW?_-% zJY5y5OJvBix)9r8^EEBND5O^n!*Z!fqr-Y$EZs_xoRwV`Uy(1PjC2Q-dUY$8D>p(; z344ncHpu3T#k3%3g3Vk-4;|62RnZ|pzEFfIg zK`uO7v2ccDDF4SIyUa0iE#0(Otp<)f&|)n=jhX}ASAV$b`+pq#O51DouW$d_A3yr$u7AZgwkja5GgCgbYt2v2CI9&2 zU;p@*B~;0Ri+NPy>q*pu*OyY6!L`ES*aFKp4#ib}<;-j2oRheK!^*PQy)!<1aI-Q0 z**%zOOOU|D#<*_51Yx00xC6Da34)6t`+UnY+h-UeYI~wTUw-w-ZDc zrPoc{2Z>o@4Ug9f>uLPds}}KqCDC)7(0Tg6P83{E|Az5hQ4*0fL|H7)J+FJmftb%_ z8zc$nIFF(ZOQD#R6j?^a9nmC5oc60yzSH8KD@EF;XFI8k{P6Y+@yNB$@{^POm^A$7 zGXb{RA&+MOUDy+0(GxMyX@ZaVI{epFem#BnOUp!6Q3DD%x^UD!Wi;l#ZHm1^zULeX&SW}m&JNzmul>03rO&ufb;;AhH` zd^0^4#~xDYRabo4BB9Svxl`RGELl*%ABe)^`Si|NMp6UQRPf|1Y#+zU`dzf&hkP zHacQ5MhVPG_ibUBye^>A_irc8em55sK@^HKy=;-0Vz(jt##)G`X| zO|W%2-z(n>xVkWQvWD6Y6>deMjUKQI1&~11wb8?Ig=?0wbN@VcQr!PZ| z;~fx!z~i780(N8NS30GCF%fPf346Qt6j+t?SBE_p0oOKabsFd^3ZwAYlmY%F$s)gY zEpF*mP5xz(#iA|Dws|dbEMr1ag+gAV+9<3G$fHX=3VHcZQ(sgYx2hMJx?v5a~#O55cl+*r(lv3$f)2C6l-k5xGeF{ zQ}Ua1KaC9%3Dg}$MR01!DrS>?7H$FD1 zY`Y3ww(gpxPq256QFd;AqlIrDOOQ;=WDt@y6C3D|HzA9@tO3Tdo;gG0D)4QreWO0S zfnaw!R@B~(`q)o*TGmIGl2URzDktE9TS3Ggbp>BHEHiy9=HZ8kn_9JoT~#GC{wodz*aV!X9$96$rLcOg=@n zP;rB=7^j=99Hx5W|1P8@gk;j#x2>~}lw3^zo)|6j(*wq6RCK%EPxjaZzm41S#on~^hvL8nYs z=aCtl2cLN4q&SyvYAH0(8)o!?XVog#tFUyzG|iybRTX90#E?#_929_|WY=8MfYzvJ zW}>54fA(O^J|Ebzc>D8@zd86diyisxsmvdeeJ(uZT(+_!A5jc6`WyqlgO|$fIvZeKcbQX%T|lxTGKg1jMCxpdal#iJyCY7G6`S#`{tQz4T%C zJBmCFmgmOLZB$gM>%~{-`)*BwowB-s<=))Qnd5NsL^N>*5hq(jz4ph3-`Ub7bFoD( zT=LbqPAq}@VOWZTO0n(Q>r9O&MkAMiVn7OS#hX<>dGJFNJU_$3s!r@|E#O}# z6+gPZAoahh7Z^NprR!&uM6HZ~qT(N{22B9HDk)^i+%17CW%z{H7szo>$=A!Jb5qd=yS zF{FeoHpKCWd!l1C*2yugSFwjMYL43m-Cxte!AdX5ya~c%8tOchhI_6?-kf`3N{>wz7!7jWYVu=PfYS#JX zNN|nQs6lOS-6E<}y2-u3yc7mR*|&vyT3>vWiYQG3VPp@V40!?@X!@?Q?_w`D_qL6$1eX+NrK zO;0f!De?gocZh6TP!*X2ObZu0bnh<`p+LB48~;_|bHU^K?9xbGCFr76MK<~OL~d4> z`#*!p%3e6xSP44-F|kY5Ca*vBTZ{j(>*ro_a>IrFkL8m_5BV1qGenU)zy~jHAf3`} z3$8%53d|Jgfks&u1W<=PO@&-pPfREqf$&#NPH}ked$%z zzMBIuFrvd-p=2w3U92vZmy>yVspz}p zH@4=vEx1b3~vM>u;ZdUpxBO5OmKilC-*a375=0O?36)t@2BeS@bHVOj;(6 z4=a+^^N&qMbD|r3B|F0WrRYFsC_7}gqo9KP*wj=(-OKr~l0P=pCX8P=RT2RZqj!+Gbj>sF}KMdY^L3%T4}H^pPeuMLExcv74?f zB^SJH>&PF6iVJ%=EVbnq#Ji*OX<+B>R<07HLx--TL%4PhmOY5D^HC1Hj2#j6Ue#)m zW#W3bYRaF;Di@x(imVL9CW_gBRxK`@mkm|BN204^u<+!XyjzJgLsisC@rICg?NX0K z0n|zHu=Buad>)UFmOW4y?X^Ds1m_%Khr(C{GJoFW|1U}lJl3a$oF(ffk9C>rHlu8} zf=mI$HNc6xBMS|8Fd=9EOa9?6BUZMr^u*z34o4j(zmDw!{6 zOz*%-+bF{|^Fcu!BzBx>FD5YOIDE9N%ntjUGlnNJ=|A5)7-q`|^Y{wKg%`st^fk+s z`}t?ZH|Crb*Fs#SKvTwV*M6wF=Y27ysyTJv&%F_}+J; z3v`8}{7ezWz&z@>VCzDvyn+{qPw*r#RQn;2-&p@8=6`<0CFR&9KBm zt_h5hNx{elRxccu=~7>IyqECUc#d_zo-rQlK4%|eHxXmaCka3ORvmN+Xfpr1;s^IGcVxdwzT8y7^;|BB$Tpbv^x`cI-KBD@pXYli|95 zKVeZjG}H5BDEF#Use21dITDPmnGYj_6|QkRWbuAFY%r|$#WJY$z-N#$v=h(PlZErA zizl!#d$ga3v-Lx{#6w)z-C*%hlLScutek{W$$ojJtWAYkJIEtqcG+B|>(u3ot-;{} z8QXiEx!=3~RI#lJz={`I%EY!?!E; z`y#u=uva^;Btox2k@aWw>vH*FO)sP~f6YO3%8f08Z)Cl({B29I})!jFZ^L46r#p>qCtUfs((C8m5IhSDHR z(l6HsCPyHFLIIQ(93xPcl&8s#aPUjP8T+;~(!>2xz;(c8z%M^|^-Gq;ZO8w2y?|_Y z;Vj-sD{Hr(Vjyl>PsP=R*LV$ykASq@qpdbACi%~hQt3Ls9?2R~5(S#^p4)+YB}4h> zs#a+B8IB!-qYompg1S5s0@I>eUqii!HtmIP+z7kvaZ1~+MJ4&qe{@mWqcERI{p!Mh z;GX&X1cBl>U_3;KWn*IK;ARw4ckKCOfdxk2Iq>6hQae_*&xQTkb5_v$kYWx~>-M*`RDv$sRVV?)R;sVU`iT-qu9$opJK@9Z8;gW#Qf zdTWGf4M}Qo8#;}h}lApo8!z5u|pnt-}~oMAISg~_IRMv`0E`- zyLQRX`WJMGcSx($C>`D+>=hp+2P7v%29G9Qo3`+4#-ATtaOo#Czc{5WQ=IhP#?R*A zHG0*j({B6TjI0YdAJXO|hC~NIX<}xHXmMy7@(uw zL&agrrfxorI-EyVm>zB8AdcSgH0Xx|P@D26_+@{dne*v^r*c^k#L& zj237Ty-Z5L(Lk=QCQybs5Rw5DDb28Csi$j$8v^!*B*ykjZ_j`hpqTpbYx0zc)q##_ zDRjV(Kd~2<<;Q1#kH?6Mxvu~GUPr~xp7-u}WCJ_spCTv{WlYhP$a*yGs*EWU^iWtv z{mF;z!s?SB;gqwSusZ1_?^o8^b_8)*Qn|3h!_q&r$Hd4DMcXwJ<8-WwG}8GrR{Yn^ z|6)$w3|+rZgHH<2$q8kvk+%y;kL;S&z0TapXq)*D+jc^huDxyiMwnu{goD7Dh3&$6 zRRZ-HXtrbkqt*~~^CfxXWCX1?KfkzOSlK=I8NIL^e_8mAo%wq{dN!wb)OF{RFWr=6 z+jcv1F&?h_Uq)DZqtisy(Lk*clpSzDc0yvk*QVY7!xgVrzGhmM^ZKFjxGtc=rzWO~ zUPI2!Hwy2Wm@V%~&ya=?5SyNQo8K>N*KSfDkhRGB>C{&j{{6l3R~Np&n%61KhD);r z%~9+8vefO`ZT!#3j-X>CKI{ysk?s#22F;Qal6XiXzc?COW7qlZ=O0q1#E$$) z-HMSv$M`CKNdbQ~KU;7SeDL?on`y9#OaH2Gc0xU8i*R@JB2kN?UVJ{JpEU83yw5=} z^*F)zl>1Bc$W}RR3;wRZ1g({S$Knycwr}9y$q{Z<3obl$U9)oR&r!@7inKvxZm2nw zSLd-Eyr5jKow6F}gvM1n5`93q!r`E^a5zKRB}@<%O0fuTkNUW@YSw3D8-J%LK5V}L z8|ihId=`naC5eLfb${LLqXTLe!QXU$hpleqeFX&3p+O!ief5EZQC)PN=B}p>S;Wl* z;>fi6z%4_WN3T|vMGi(4N%A%Aj1HS}FDo%7iHihmc04r$c@CFRPqFm>ci?(tjmO&R z{LY)7*p|e&Y_$VRkxNH(rYu{PKn(m&X`8m`)r&AWpVFp;G{Ty%QP@cK`_@Pbe9OhR z=qBG|AT55&vt7GG;Y{V&hrjgrhW+8~PTSx%F~2>yYo4u=*5kQu7fx5QsB0NP|FL$i zUWG5X$+BD~H@q-ciHk4Lc8uN$K^xr_pGHL@^C>8CY4=k&_8`gp;sM( zsU$scW5h_50SKoIp^dVeqBFFSHU?vrrvqK=@xh6Ga&i>D*x@8T-Tb?MvrJz48=`yU z)C)6t4OmTHMvCd8NGBC{lFk;OE>ZNLfG^lOBkX zVMAzoV37ouXz77^O}?fDl3qu^Gs3%J)vr7b!DyW0@QPD0>tZhuEV@|bQ76dP^oj@g z+x>6P=mITLoC%kC)wk8RjdwxW8Ip$7%WxF0>X5aB zHwp&8b$N=b^5h@duCwDriCxd0$VHpKH)-GREht&GJSB$oazlv=yV5JI8v5^1%rHeh zqvEy)<5+2c?1|id--F9@OQ*@5wOf!J#{?2gai=@c{rS8a-g4*v7wLQt}xs7zg3c z9I{4$b99>uIRpCKYvC_M54@du;D+jB@E<9`|4978Jm|f0*`|Ko6!xEXn ze@lQ~)eZb}=IV%o2;gJDz98qnzc%lfy|1&)ZBE~1cRR*NSGcZ)U;V|GqHN2-9(S)@ z*!Hm0uItQ)-XhPU_kv}>EI0~77D>$BPxh2&hiZGl{Lue!I*iR< zzz*iAXQd}?a~F@}!*%gZSTYpY<#yW7!23jbHaty$+9k#G^;tS3a>Y}7Lr{q1nMwt4 z!^SfL27C=WHojExNtSJI0+-OV3tJVI77JvvtX6Dca5c063jcbQH?HXvTnf0)YI(DRVk#-Jjfy)JGZ=N;9fY(`#biY0OD+@a zQZvPA453+b&<(m5ff=`L?pKt$%&FZW8-SZ`+teE|-O6=S24Wx%gXeZk=_F@kbl5)I zF6fn4P6bICxJHK!)((uzI68OWFr2U(jxju>mr85te|TAN^VS9OHc8{QJaXYFXqy$v zawrBw2{ux3W-{I)Nte(_-;pA9j1Hv;@*&>x&Ux{9F-Thbukt`XR$UA9-saQEOphrR z<8&X~KHKupNqM;8FzB#6=#-@VF=|yV6OIe#LRs9DO959s8x={R_=&P5NulVhnA>MD z8mw2{6ATEQ4v10X!3i6q?sN9$|K9oTXZ(?QC)h24ByqDJE{qMJ3>wv6ok1~cD6*Q0 zOBD^%eZgk^#+2~&UpgypruT(q`($dXBr9e;OHAO{+Kmfm1HtC`pI@5s@3!oOE}LYm zwi@Y&OoTP6GMEi8n1w9e=zN|a2V}$EeUK(bu1N^pt1e2{Md{3B9Svk%)DFopf~1() zQEQ{kmEX<;gx!PxaH<%!VKZQk+~wqgOp17w{M6P3VVS%4Lq$j4{A{|AcaS7$I+*>y z+qGQTOLqwqU&;41#@2ac2ORV)pbIqDXT#D`T^ZIAapIY@$c<|#9Qm8GJd4v%xUMlC z4Eoi-f6X$@j3}7DATG#dE93J}?|fOvPF@Xl%J(1y&#yz33y?uU_6IA|yGf&B)l`t~ zoYl?C3a*7x#Nny?_*>~u)c5E1MCvtJplLJ!rJu&&cI_a!ufR5!q{+F0TMRy-n=F|O zIc&HZtFKUG69X?Pv6-IDwE6sfx1^bgK_lA#Kf6azU2J&-+vZpj9sJ4#MY z-zn2w5I6as(4fRHXq<%UdgaTZZ31#K#iA`tgP?$q%D{~wJ+BndD4ey#^V&GZp8eN< z?94=9^4P;1c;G+6)%!dxFsb?KH&2q4E*x7av>N!06!QTEziV7aEEO639Irx2(!qIF!Q-49%h=?OSbWE^GlrLba6z?nDd4+fWr+jiv~aZhHU`f zW!FY5A^Bnv&=jV5C3_k8pUY|_dez<7hFP~l+M=GF`4|~UJ3%pW`7`_E1d5k_d`x|Y zv6vChF9y$%^^?gCtGCraF~Gr+L&fFtjZlVYq*27u%;A>l+u^564qP0DZ|7Bhd~8)z z1{69kL3wsOWv;wHl|)160Dt+Er`9e{=tr`1K*oGKo!O+OA!|g@&VK@n#;D4 zx2RB)IdwQT9Xdk0yjuB1uQ|!}nfUnIj+|rmIZdaNt`6p<&H6_gDVR+5TMbGz z#Z*zG95b|a)3N%rNQC;B7*f!C zbKKU+cRgU(@v zctK{KdA?^t=2={pEajU;WXE&Hi}#V0#W+AQdnwWY(OxKx#J0mBg6^-mLD4ju43T6e zJ+M)9Mm-pDH=uTnC6k20 zCixm{NJNU-li?R-a4~*aq3V9f2{NHVOh&!htoA#5}jSJo@e(gQ|*SmgvN?RShjLDHS`4$H549k%; z!jq1^er~~g89%+f;Jr-h_cdR?_U%(zN4PuoLv6%<*zq|&bXnb{n4>DS%^`8|b6q$H z%#v~Hkd5eO89}~AWt1)EbxRlX+Mhvhv(+yTJ9dJ?_BN;Quh8B}pJuTr-zomZugE4B zo@jPk*@zN~f%Q!R1*)!@yn1yebBJz*w&QwnlJ~tidR1F^NBA&b&p;g4<0wIY1Uuo+ zdas?DG&zCk6dm9#ncPTUlb7?mp*Xvkw?&5G9!D5}F$c{~0I-^ZF(2xHgPbNHDsSzp zf3me9&s!eaEFcd)YC(j79^mgF*JnYv-J?h2w6SL$CObjFdcU(ya010kwLPr`ww=i? z3ksGl;nzE*pUN{l zQw2!)hU;FezC8*_yJvtD&LAr^b=qfX}|qnD9x zCMCdIf;3Oh00`D&XfG_R?{fYgCs<4xnwBi}vG}KL|5g1C+2F$dDJ)q>l}Z*;OaVo5 zAIWR>OEV%&!LjoMM0~#%AKR_05})Vg5TndMpY|(%8Kh`R=wdaHx_}lSwwglfSfp!} z7m1dNZa6**`Dm=|#mb|J#(L|~lN`In2`evsF=?OZdG|xoUfn|92s=9uW)F~nsEbSw zM9QL#rkKU@*cxh+dRK^Ewa0x=WSL^iGczij@!?n)us)3wJ}7BU@)cV#Mwj(dS%eoG zg(tmRp_H>x(IvY>A5(3V>r5CIFqI<5{p z!3`qEOTxAPT-0x7Ax6C_Gx!$0NYpE@iPBp0G32?#6O(~hSq^l4wLyKdcoCLaE%N}GiE@t)+3jc8 z3c%4`InW|K`BCil>cEL9YFbob&^4b2Q1^;yU_83YBRk@n95qVNW~4=#xgqfk#`Mg5j}C63 z=Li#Pw|I=bCl9{;XW3ygR^ITlZR?nq>vuxMfEKMHg1&IKj7oKk`qc^l6ltthwWOMWl4)f zu}BP?E~$x04@_XPBvsHjG#Jw7t{aF#a#x(A^GJf|Fig42Jl;w4ybjz+I+Rjul_NJ& z4OBxNb}x03Y#sbq;~o;rOFORL+m4&K4T;Nk6zQbL*W32NyR1cF>1^MlL1inQQQ1tg zAuxlKUO*qW%L@?-yy{BmQdMA;e+g9#dclRDJ8W4*$0ZoThpvZ=s7<3vK3OD(f7Ar{0-lp0? zS`;`19+Abv(;t#uo8pO>yu|K@+n`cgK$nS&Bq`zPl0uVeEpT&iA$rwHucWD|g3?G`BfW$aN%kN+F0I2@O`SjoVmDFf+@1ET zQkr^aQSru^Q{9Bjh|R0-$lZBUV`857d?Y&BpURQ1^xm91b95AlXl@2${+AEzU1;$)&izZ$TymS6zv05}=o%|Ap~OkdqDkaF z6}N2ihL9HF2F*eLfv|2U$2}BXO1eqC7%2#^%xN|$BcJyBM0rTP-*<1wHF+lUflRMj zBr2dQgKkP&!%;gqTXi3ba zTR*drK1#a1&1btoIsnz1Ga?@SO($Vp)vhg%xHD(zWW%&RMI${VYYksE8Pil)Hf;V& zvQ;jdSdvU$_``oL(&bxIDq?Dd2_(-%VVC#KPLtke4hRu9kTR!7gsm(F?AUG0Y;Sb( zzV5dd|KV#fPd_{H(kYVZ!u3_VtUQz=ih*on9u@cZE6OS!hV1@Y52HIo8iUXCDrINE zfEa~6ibB;+Nx!@p%JEv{r+iD)ANijpL!PbPuVzyJHf{HWHdQH1_V_-k9h@-Dl z>~_#UMUVkqik-8%71>a>SORQ#lsMV2O)k4UxS#8*On|wjBz*Kd#;RbK8{@m&LRT`^GHmta|b4!260` zWv6m4X%gr)Yu#IfNR!>?1OBa454=!DVFCTQ?|y+}5%8p8)k1T~{N4F{?q7@_}HY$zN&FZU|3F@UW=>wPM zk!oKg)lL=Q0Z5?>TNEjhexM6+EJ`>cfyG3a4m%VH+~yH|N zFF-fCnXdB~k~IWygq;IX^(4e`h8-tVP2%USJZ($1>$0d~VFA7?Ss9dLS`O!Y#k?GD z*RJts5w0K`V2C#bVh!+Tq(F1kvt7Gt>aLJ%=1}w}>H_z6Z5scM;?p^2#E!K$wnkzC zNU?pCL-#o$g)*+ssJ9iTaap9WsOBWitW=v!OsjmQ*Ky4v|07<#^v2jFo;u_X!zkVr zpM88I@GC(1Zb;V@P&N8m*4K(e&cz1~>NqjtP&|zFFisem^ycB$e>~f=W_q_y+eAuS zIP7x7Y86yVF~EOTO~uv5RC}Bd89bVJ@3tsvWB%}HKSSA~xFv`e;gMctT6815f~0_t zTg0yeYlA_VKC)T8U2$Jo=Yj1_jdYTzBx-pylntt@)dyr-fcdsEct=oe%n4BwuT74h z4no#A;q!!vXDl#qKgJ0@)ahSTw?FUoyq|cMiPwhb(?bMQWQHND2$iIoL%t~ZeM4ZN zbNS%G9P`rE%Z;+#z{m>i9;hu7FX|QdI6VtB0hDkbdJcDRLdi>sRROuLSd2_f=%1F7 z;uj_`bJ)rx?V%V54ex-|Q%D*Zr*883tQJMLVn10Fl^bziQOU~;&Y~BK(34GL(nT!+ zHPqtJ3Q4)9o7W7bceezX&e5wrBWZIE(alm!72%y50(ygMsGVUOUhayxD9VxG)#pR{ z30~tMa|B~Q0j!N>l{ld3$(LrNwB$uw(6ip4$R^d#8I@t16-EwH3}{Fwk;OIJ)34U z2e&8|2kR2#A9?9jo1+V$P<&62Zy-&*adHNz4ZyPK%>$v`xsgJ!ayV6 z3L5JvCY>T_RGhg##T**Mnafn0ca)Sy9@pHJcF`TOm0r&f7u=BHuwl33LBIUgPls)( zyInRA!@?=vB}|%JrR`H-0dFyV4U|~1l<&l%^{HcZrmwPPg5cL}>oGDeRqb?|$?-tfG4w4n;? zHJ*WPaWAGTXjJKi)Io-_8Ok^f({4uMtYPp6N};r<78(@Z>gnzN1IlBnMp-|p)o={! za~_ik_!gYMi^~n1U^1!fgFogfN0votA!o@tZqtnm=d7W)VU#CQKry)#$)@6}qlT6IWbDX3Ks-V7dZTP0ZO@SXUpiSisqS9n5uLgXTW$rc5v{ps#yX~4@23g7~g0oS# z(#dnt_+iBbB;2o^f03>ZvH<0G+xD*}xh@>yZL%8rDvBwmNGUj>vD;$nm{o$2Pc!@W z-d8)N@uIr$>%MrrR(wF#0_}~k(CHEGoMn_%dm#DYw%8*|sG@~saAxrPJwnqHJceU? zgz=&!p19{|Sd1SBHfvz+L;b`%IGGp9YwymV`&&?Rk2$)W=wBG~Qfq~=GKv8z)~!@r zf-K&DkWL`ALX?_r_|64Woh~FML76X2(Bf_sUW87B9Cb0TU7Hb6%*zz6^2TGuysh-5 zh^FXb@NXIghR}X;i*5!THb>*iGPo0m*Ar1P?vqP5PLilBfYBcPc`eC#&cbvJRd=*u`xB)yW_^d*f+8p zp6dF$@7N{=q(t!R|9UI#*FXQ&+y75WGj4=(;kX-1_#b0#SE1Wgw}T8Q8$$Ex^Q2qZ zH@8oAm0wR^3d|uOes)p}!B)! zZ0j4iRBgI2u2`xekjwWfzYBCc2jso4jKmWfs;p(s+4? ztbAq)Z=nMK^29+KH+r7<)QO+`+2NR1o_C!9?CrvNWRO(w&VZAVsXfYUnZ8@$>@gkV z*SBw&$GFkShuDq87!j72mS=CBHF?BUF}|xKi@5o=E*!+qvT|b5C?E2%h)%%%r| zE%5J!GT;=RnPlky06_M|=$wp}u^wOb;q4$>QbR72i3?+bg#vNds|RL>6EJz`Zb#LJ z7tm+LE&L%F&L-xjG;@mi9zhA|rE_%|vu}e;@x|y~d2I+DDxh^(Ro)?67f~rep7|{L zBcJ<aFrXMD!;LdHBc77E67tnSVK{^cZ#IZ0d^b^yp<$FMps z43N!MMq@R_tfI(rG#bdcf(%kRR073C%m0PNu*+nOYv3k$6Z@o-HgquER%qn$b&U%n z0-|oCBCuO12IiGbP*0%fQB?R;_~=y~vaK=wK00(whP^IDKs^&g@^6Q(2-RWkuR~_y z@M#Ru2e!!YKyRG0kuz@NFhLMJ?RIvM{Qc$MN^HH5=L#e&zJ{5n1x<%(epu~YIujQn z&IZ6317RN~#<92mppC-ZSeqk2nds;5`D1 zY_{@0YAB|fB2^R!Y?knH`86?Hz3SDy^lIP!`Gve>rj!_X>jFLjnvHuwCw%v)>*lv> zPcaofXUKX;)zr;PlrIq&746zq)g@_j@Mol5yH~JXeA;gl2&R3aJ}cg$y(lYN&>Rfq z)~{V!u!?cyOyy8COcVm6?&oH9{_Aj|hi%m(OSvGbMuEa@WHQs|KEjg@G3o;$Wf*u{ODlz*tXk&M{6S;JTse@BjG+8-y0K^6HuO`^bv1(ta+y)Y)of zYBDKiJw?)~I8=x?)iZX{hos2Njd2kaRn#FlajJU<)JEscuY=~Y@s-by1616Q40Fji zzGFPMup{}H@*kUQd3s)upc)IS&pu$zKn|87(IuZOx;eNiYPo!n)XqIlo}tOJ45FQQ z8PAYU=$^gHejEI<#bgBirtWR>feU*g;CPIRglwf4sD<7_#WgA}`do}2B5kS`kbvqY zmjc%D+q8|sX1aKOnYhAdbznb4N#aEp2(kmBoZ?;I0seNMtDf<5*UvRy?^HzCIrw9F zoKCeH?8aiO^-j)bpB37&1wL0X5~y4p1)mIxSwoT4R9w=`#IWX&^gvyfBwq^TU;gi; zgb$KFm|#9s?Fc&UJT;OrUtA7Y8S{Q;9c0H!yY%GOpLbGhOrGRvu(MCMLeb4bm$@=H zm8?`=pY7}l=*I`lZU{Vnt5c7#LtqSZF)3u)YjbYcY zxjJ4n?75XLWiAFK2)gIqcU$L|Pvf~Zd0lvqWTjX4-2T~_Q`@y|^1UHBn(MQXj{@(@ z7Mzyhv#@u!O^%6rR6)Q!hZ=ahM`%owZhg>^L&j-@TsQu7Y|Xx=V2e}I{MwotlJ~;Y z1njk%gSJx)2%D85pL&L}2-KZ*7kn1WQbOtjAGjgGy6%e4?YV_g9WLZMWV>I{p)@oe z%aZ6&Wh+BD%sWCa>sukr-scWQvoDIwT6Ju5g>fTC5Xu9*dt?R%$R zpVjRowHF7V!cjo5dJely%y`iMU0dnr7i2}uA{344Q-^?kU9V~omzntA_JD{kvdQgL zXY(Go6-hFb7O6gQf+$OZ-Q*548&;!Yvojv>*gV21yWh%|O}6d4aM^BAmR^c3Rk^y- zPp?`dSSCWv23@XXY49-r0MTs;L{dOQ=*`HRk=@G9Z{1wb2hhOlAGl#;Z9tg}85;wA zZoaObzgu$``}iW+bzLvQ z!cm*fFAvI#C=(x0v_SSIkq0pqI!(DX=#;!gxJlJ3zDXbZGS;XWoHV?)_?&?v_P0opJh*QU@g$HODtyuBgW zf|4jBecLll0EC`c&5vUoaQ^J%b= zPV=jCY+QrmSUmRPjy~YX@nIi}vEE*nRaRY)tk1Be%W>Ixh=l;>vK)Ec>ccyO`o!@f z;ExS26Mrg)-YoB|@C%`(-UUFAvqyE7r^ANz4CNiqMWP}}k*IDukR9U=&O}YTVbAnH zYy(3rVM{I(mB6tt=O{TjU3V$s=9g~DTcI=`dUi6D9kPtDs}t=ub|kn*1AU@PA8z1o z{`0(Zw%I|K#Slv-Fo&!W;QIH|Ily87f*s-J^md0VY38;tQ#KG$f-7LBciD_sDbC;+ zxzWj=oWLuh!al`x_(tbJz7zXaUUGQxY z-gj#Vx!`@@?TR`(5C~D|gFuV2Tihfl^==Us@(vPE*oDUPJ-{$p7k>PetYD+CUAsE) z>b!n>o!<`czvw<+&>Ors?610J_(Tr=dxCcDaVdH}*-Wo`-SkS6g47{Zwjf_}OSv&H zg_kQe$O?nbGufWw^pH7&$=aOQu{S={aRL+dxA60?KE3_eh3gPl+K-$3H-uF1R>*Yu zK3mMn;DxIF(&O$|=jk$eWrF0XS@f&jDcx9jXuJrcY zKC$kI;ykY`(kMhzR?Xa)>+n^!<8U-Sys+KHj-$)J`HIYz{FqBH#)ZRrEL6#;s(Y1x z3Y-+(8gfFiN1Y9&`=`GGZR!`m--iuz*w#3Gh^+G7FKF`5j##d25@gX?l8=0JsK;BS z$f8SR5anILJaFsrFkfFIy`dc7!%_iQxa-A(N=R=8IW7R~IAD#kRcvVCaQyZE`L*X8 zmieo62JaSWQ!nM2oO4yd)}VYK|m7;*@WWBpc{$HwKP8q9C#`vNm`T zs2QPn_@5VUU+_pf`~uLtpl*1>tL<9+90Zeg$!FyJq5oL;56)`hPJz@YbUoVSfG4=L z?TxRhY(>Ez59PS9-C>dbu8P7=A>;_ubO^$p{@h<)ao{f>Tyt?SI?OE=m&1NS_YjC)v)J9Df+2B@-i{74QD7(90)Vu;E;fuf$H9PQ&Pxj zg0+<69m5HY&!bn7eQSKf^Ah{`j1lN@UDq>@o*nvQ%RF{;LHPH{$;qV8Y69q@m`;kE zr{b~&i$usv6IU2{^2Zar_2sF7YmiX_-GTp9^5x@=;y9`eMuJqtJ}bVIU%sGdjz zUGA?#_Y_l$+h?T4te9=4GN@7MP~y2m_L&0iMthK;M%)w@M)3eQVDJZ|0p~NuE_1UG7x^87*P)c~NPZP9*;G@|g=yZ^= z!w=7}LXZ>4Tz6*hJ-c+f?acOq_}VOUo{>%xm>cQQ6~>7AEhW4SXl20-hfS>bLexX4 zMOkErtWf$+x@17QSFM+ohdMnAIuW?DV$}gBe&@F&^?Md$^!uLQUnL(tX9X1>TUpNY z6mym$r>VGN>23|!BV>m9D+ZYdK;kS@Hmrn#iVsy?Lfp&>J}cW2SQGQW4U?c~#D=|g zYapnMC|a!C08C@uyfYw!1>xzeDd<-@KUf|xV+{y2P6WbJZU})jb=#=8&tYC1~l^27OBkTweOaw$v zHiu6>`!t-f+sy2~ErYhXQkM-#v*bl_9)lHmm@bE29h1+$I!ia?(Gbuf%a%a;0ht=M zDUQDaS-O)@l6BI_N@*h_F%Be*-0IX%b|j49qP_HswZ?U}oe9s?@~hg)_&n@Z7*=Kq zbG<5{j1RL@P+EbL&|b+IX}abLgR0|Ml769XxpJfO@{Ch6^qP07V~$I+qt-^(%ez%~ zy^ADR%;iWb7uzu(2Yg0b2;0XwY@Z!ISJv+Tv+X?PvWA3ZPRb1a`*wf5YCws~AE2}s z+)EG48TJIK?rtTP?m=xRPCZ8YBqUL5LpB4oWfpx|gA{XDd@_Uow%xx_YCuV>T~PS> znco^9F2p)7Bz)IZ(Z{{(!n^q9mHijIjj_P<-43!fjfzADD+thvRr@B&^&G zAH7KQ1+7~gmMz(?0j{nqKwPS!se$8VxIs+fPOvJCw$ARGRz>Ru6nEyV^J-@j!uoxxqOiB3(=Uhrz%3`B zONiPx9m>1F*}Hq14qc_Rs4r+#xyJij+`HYAeB(hiX^_;5&1EYYQ^p~U^sM0bWUxJc z_OIXg>A%_*p0ZSI-VaKfS{ZzvXIhJ6NI%~zAEaKuxd63!-&;|29+kl#z^WUR@vvYV zv)>K#p88%s{wHxhGS1(Y^Y?JQfzzw)x^bXy{^a7d85Wx}_%rzuvc-k-@r_niq>^HQ zGPMNj)&~_`!kzrZ9^DYhL@nuRVViGC_^qfas3H7NT*4cUP3Lc=mkKtkcS0Y-{@MM& z2Z8fjJ-&UYeny%>F1Ctp z2*?GMf%<2(kTMS=Cu}%uu5fR7t8JUxThPEVq*xBbb{qx*#w?q}TM*ed23U z-74P=Gax?#je=eDQr`hZxz9n*2X5Eo_&9vjNgAB&%Vc-NwyIwv4y2686+)qyR-dhtoRM+3{ZRo=*jQw6pzi#A3saz{DZcJD?_GIMW^ zZl`-sm=QudM>Jce!@~`sT~NempwogxyuQBFVVg3pW1Ryxa&e-#%0c%H>1Whvh*B0FA~b=5~!liLA`fxgcMDsIR#U9)paH+0c# z2x$%1p(LnTYxYw)uCg*9`ctdIFo3D!vnpz(mq}!{Tn+j680^)7<=74o0L`MmfQk;1tNa8qAY3sILdhcn*(hd zRMTPGY0f2`<+`f?!qPAwKetuhB`l_E7)-Hd(d`UYYM`zXZtGS3(u@d{8ouChI@~DB z)8L}8Ri2^T=bP(QsJahi7ntud(EEJPh30tfp0UaMeh3!To8yugm7EAp#u!LPLT8M} zkNeR5sav{!XE8y)Z+$D2e8SBHxiEBASou(QDduyE4C0~;g-*=*`4)Lq)Mnl->3{;V zVc08Px8NEnpi@aDe1b;>H3Yid!5PXenqKH4G8>Sp$R>YGjv=}DW#v_%q{c4f2W~69 z4rmM>do-znx|e~CJU*;QIw-FPHy*=;wLyI{3|^G@nohtylG?c+FkO7eqNxC$p@Z`M zAXi+aUK?HqWeauF@#!YX>}&FFWqM3~xL&n7rX;GISLbnL_IA1|3Zqj78nP^xl|v+p zK20FT1b3elx5$CV1!73@BFS>OuFInf$|7CXi;@L>V?bbmU+9J(90DcJu>z1T+m=M~ ztme6O6tk8hsZ`tw?-W4_w%}Y0(yQvj8+>#hi&JqRQ|IyImkV2;Emputr(R0s7S^{peUBX$iGTUcD%(ti%dVqXatryIUNMpbw|bob-(~HTJJMUitF!v) z8vmQV5Fw1tk!;W$^zWUW50b`R!ZYe&&sBn98as>IwFdtq^me+2N?>|4Ig&QZen)sE zbY1KqOFU0$izJ;eRi#A@MB&P8<;*QiyEZv+aaf(llDR9D-HIA&16dZ51q=7Rq%|DB zeVqQV-QXFoJIr2%_Cx@(gJ+DHh5AmNaGkeh0{T(Y<}+li3rB^3DrXdbTQ0?9Q)Dw0 zcaB*!HQ_6rQoZVsVi9mc9rRC)+3s^O(xC2E=F!9alk_bBT8*U7r%Z7Y@}Ro{cKH|a z6M1_6IEVg;zrXd#aiYZfAx_xihLZQck@AkO1tmW_@zN=h>B1=4WrdO=ih)3L9u?Om ztnuG2HItj148JHt+G;)co*BxF*%!oEX#K$LKvX}uF}IwDh1Zx8PmvVU>4A7euS#%l zjoc}7>LfKzuviR?6HsD#dGU*3R*_Gy$DKE%w+ycQ$&#Uz!wY*+Z4*E+; zyd6+bxL{U3G`J@~4!v21G30hA`%4K{%g?=57~aG>03{vy0nkt(=%TeQIf7z(eGr1N znN=8G>Qd@rC$ls@cp=K{1F_jVT9jWkT9?1(@4ikBGuwkZaCya56I;|w$xczEk&5|r zzzJ(hDODHzVN(1Qn;nPpM_8Fwud@i4J;C z3=Rt$G#qaiG!ZuVtIDCCe?YR+8MoV=upfWu-r#Sgy&FHYn0raqfU$NOyEW7Zt>vj* z+epz111%`j8RiUWDcNp{?4V*QAs~a*IEX9Fw8dgoBDueMt2C z!b=6+!*2D?7@)5g*8u%!0!x3WUHX8&Lt`0ePb3hRl}BbmimKlYGMb2>^90{O<%Zfw zNZIce@80iLOiI*kzEoAj^0 zzrOmnMtijL12-|bIhias$qYWAWPKF5N5$L{=Y@CDIrO4H*cD(of)4nNs%PV_7Wxn# zu(tBn`sM{Uf`eEuPSwB42&`haOI3JX0_JVQ9XcNpT^`NI3$OKACCdXg6n$?D9&QjD z>N$BXchTJ4(KL8K~XW za}!t#cp6skOeiXARQYGcCtf#Nq*4iuC&nwcCbMHaD_`F4a_56LWU8CpSp|`P)l0?4_WrqK0{?chFzFVV`IB5#J zpoSEN=lgen%1;}20d+o#0=@)xx@ejEm9SSZo8nRf!oTb&2ERiDd(40a?5>lL=~gAX4_{TtU-i*KKa?BGD5_p6xROFzhXp0xyQo~NFB2vo&$$@btd>ZL&=&b z(m=(m`x1~E^K{&JP6vOT=%66Wx0{XJq&GY@^d^t?Xtn%aWSa}i!@|-nwLG3vA@5LTvM?T)8GaR- z%4-z+gz?TO#@R*3f}5@d7r%{LBwVJ%(dxOSfl0vziQisQE033L3}R;9tij>aPV|_L zJR5jsXE~wYYcUt^enE;OENGB-KvP{Gfm&`gYkLqbSU9caQN*-m;5@gX|X`PHZOyn0X@7>^ukhmX0V&Y^Oo@joZu+@HLqEjDxC)# zZioRlE0K@=4AHO-LTVW!!8!D$kUNt8=!9ryUw$UD#0U*f+Xp5e#pqhg3)Z%xuQ-|a z3^?r8#zKFeQ<1VJ!j#vc$4ha03~}skv`}wfrhVCX&oeqt3tcP z{p?%fA$|Zui@ibSuq)~zq=A5RtwCJrg_Uub7uwG@_K6`ATeBQR?h1HKu!Tb1EQ7$- zG6IzQ522>&ENd0?Tw)dpQ&JfrSPii{L??=OUorc52rvLZZYmenE`8?8@R>s_fP+qF zNuMhg=Iwy)+)Q4(zWM{;uReBFP>JfDY{E)S7Ie$|>14iWE^XS3Ki6 z1CZRjDqJ)5u)K%8kkdOe7iJ4h6lN>BXu}mZT(<{cmBMoAj_B(%YkfY>refcY{Z`34 z%f5d5^~RV@uOI$_YU4Nl0No5Q2xcg7AR8O2u^T7*8Y{~SUfKTtcien&d#=1=3rpA~ z?gh^!ZmCdPRU^r9{rF4}Fn9(1!R+WCQl*B?x#2Xn+&&WG^)`lu*~&8xl$ecV+qafc z68Ova-XccZQs$I-oGfJ)xN+btQLc&ASx3oID3XlTSM~fdU@2}9;WC3gZ-Y*YURTR+ z^Lrz2M?7@hI@bQSF-{o+%#869#&+9R-#jaL-a99e)$_RbUFahziEaT->gp*PpA^4V z;py;}&{eWi-WQ{rmF1E+ukOeuucPFyYqhT$_-YL2ksl%#x=rr;+zoCKo$xpq*{noX zwp8h5(h;UV3eMBbSsP#tVWr4f_emIIp_3j|0y(1pQjaj53RhB+PhIyzxIyGF8e zf^7^;ueBH&8xqETm+NoyLjJu9#E;37ID=n#^v%uGkuM+qo}u+xBgz7<;Zxk*fh7VU z5GfJd5@U6Esr0e}t%`lU7USeui;&?aW$g2A{hfn8r<&c2F3QhZ$vl$mz`eA<<2kG} zC5w`&DY5~b4t;zDCE~kCwXbR+>1XSkjSdDE(zQO1yf%5CP|U{;TD`r{|CI+UF*%A3 zdcIv}e=M^;gaGU3kolF2)|N?|f6DpD*NBnDA5V=Z84ipQtqDd-C|Mx|BJwdv4UJS~ z=#r}CJL#o7(8xRplH{8ez@n!}AlD-BR409C?$)n_9ohBn zy_g$6J~#JfNWEk`sqnE2a>j!gdm@JsVqWSwxg*rdqDthNL~f0IR$0hG&Su3`Hc}_l zNaCiW@{>-K#Ob8-pt8Cg3L%e@gMzEd3_%sY)U%D7&E7TVc4Ug*3E@3SGH1YPOZWgJ zZSN2egNphLD86^h(243@YvBy~1o{3~=tB`pryU}db58p#1z|>Dwp)3m;Zuq| z;>cL%)E3h=G<_rfWd0r3-@$0w95=}2{f&|Vb0%R6guM%SsQa~)cN4kdkB4;9`{#ZV zOC_@LyNz6#NNA{oUFle7n)yU9#0K*Z7~>s`u;IAbnBTwp_B;P(H4I@5;%b5okOl>f zi_$$o2B8AnVBewbNZK$m)R-WA4R2zpljeou0gT(@=JX6Rv$_>&PbMjXLD2bm!w@v}qc6h;$rB<3G$ zNEx#bzT>W&&@4TCH@%ONRa0ay71Js1m!cHmGH|4?NOOcbgOckPAP0=m5|qk`cey6) z^T;9Cm3cC_U)rsgaK7PsYK{v3Iwnu|NN3?VM@ctZwKTk4UM|7P2ZZGiE3l#Ou>&%a zYu)PUD()S688u#8#Rzhy*67*q#)-rYD;guswd0rsXB`+sl_ns1tSZ+8VK|gLKmwmW zQGuWp0&p30i}(}Nmu>;_-F|j&nsIV5nHA$gPoJqkTOoR{%xO^dpKi=ateOVxV3aYMzc&9`hcr zWneNtjG)0};Bj92<~uWPe9yQP{k-%S{iK1}QslrXp4%o?t_#c#TKJ0LI#r7lg#_{T1HyJW4uK!lCQoHw;x~ofAv*3!k`7y|dtvnw z`1mRsgGYzyhR}x4Rc_edpDobNhTm#0$nb8MdKa$T6b0S<&zNsU7<=~G#Qf|tey^W= zw2o~~^z5*Xumw`*!%%XU3L4rNIn3lBFI|D((eC0*IObML!W=Uw2pNH${7~1E|YXwy|~fmBoItjg>JK5tEAOnv(jeR3U#8_vXiLLmNYhZo?E<*q@ddo~ZqRV1Tsk?OwV0FW*Q`X<^%EqU^Q^UE zEVE!UB#h<5pK%W(vqSxt)6O?SjJUb>jhkg8pPAWlU}w0_1ZBG@8EmA=sTfF3Lau9* zr%J=?3Ovcy&wvI|tCnZ6-s)e)b8) zLt!hwjf>FHx}6C!^dPj+yP;#HpS@x>{te**xt7Cy8{?hZuG2cWZBCnylK z%TxSpnl-a6a2OzCluJ7!Z%j5X?RHof*g_e1>gyMUd%0Jn-E35EgMtOFX8O`;xUp%Y zo7}r07_;7^@RdII`%&@H1%Aujd!fstC$cIu+s#H+;Tg|(jPw4C5p8pd+0vAKzP`}B zZNyk?OENhOACd%Qboq~cbco?0$_D8#C%ZJ>(0m3M)K z{NHCqC`*7fSIzP+PoTlWv5F!J1T|5eblu!U8>?B25jBq&m0>Sz0~g~cj8HS_qm~x;0 zNVv4CZIMS{w7LQ{>c~Fa<3=|J+3GX9nxhAk(Hs8Q15A?r!fl{oe zh-YJ)qoD=T@P7V|Imh@l!8!CM(8*AtSKR?M87-o2$%gO^F7Q!Ue>BujKGKqEV^9vW zZVXiox@eUQRE`wqf$I|GUay0)5vB+?dA8BmIi0ImF(m`;#UuFL zN|QYl_L1XmDSnN7{Hmr0-lb!CduQd+*?Q4owPH6^IG`r6TAmiv=vU0gw&W7G4Q|j{ z4X^h3pXKAIRkY#LPAcjlriWF*N5Egf&&G2gPwa}nUy2{5tIJpySc_S$E_fCGuATI5 zfDm*<0&TcvdMAz9N`tns)&-?1^4QnO5f4miHi)&d=dc%ggvKJr4WaBfCd>V@5r%Vj zaXUyYvlY>S(_t4)R>THMc7h^D@!M2AMeW%anFz^pOpL(_W>}Q)ONGw&fCOm=VtKK= zlk#@C%CLf=+1nAl(oK~?8(eg}#voCf7v9KMVaLoxb_rAStyf9QFwZ!tocl;n^+O=v)9YW&P}>EXHwxF&aV!YunJJ%nW72eClkN z5mQ-yIcrI=1N-O4Oi)rq$@WlWCyKhS_R9=cWwMZc52J*rxn2)llY?Js@a^OSWHx!C zT(XOef*Asyx|-SA81{9aH_TgQm{j^WaH7mYW6eEDxAY3I1;je9<_-9#u+{P_tZhKU zzjX2$)^?v|lLwrZPF^;-NsNC(k1wchoaR6DLS+oRW9ekn2-_`OHu+#MXwvy@_c>#` zbePFp+5lfBpZCf41Vbi=yY@0kaA1=Iy;Q@JrD{qBf==tH7!+|Uk4*FLP}~gJ z5S{_^c$Y2%H;@%6u%%zsEJqQY z1XiUBsI(_XR?%DVMJ`n691%5e(FITOLm@c59vp7qEBw{b%@ViE7tWxPwCvApd8J&~&0^3~3l*@MB1ecjQfZ20-{!D{$OwtanC@%?k= z>=IcfYrz^ymPCe*`f{Cz=HApi9!3>G8^XgNPX#Zr(*Ko+EZL3m>HHDz?%JnIT;vHtY(^Q=+Ff6nc@ zYcga;khnhRJxA6#uz$PVWZE^93sYZo z?2K6&u6+%|mwAv>5T%A(m3}e+HDYb_P4^rHhzQy=q4o5fjD79_4Nnhevm?yxT-o1x zkC`jATaY+!e6#GmnqQuqJM^3Bu75kE!l?`t0tGg*IDCs+6{L1?f1sAfad4N4 z?IB4dk&R6rS+2YN^MzVqcs}ag_#gIs?$$r#VOC@4C8Nj2>|#bH#Bp8NQ+YiKz7Z?S zH(dH9S;fqSbzI{}wwgeG6D3Qh$T}(}LvVd&qICD`3GZH*3+%wNIuy$w_Hk=g9sSo& zR{lKY{Y~#*`LAYWMnIn^#kY?B-%obEV^4Swfo@DNAG+C&Ka*5^-M<%sUi%vtpup?X zE7F5{nah0nA)jN4VjyU$_N?Jn`87axd)4$qWFff=m81`X7fn0lHxU3CQ|*Kuf$@0g zSLdzXZ%$3*u;Ez?oJo*9h7f8&Kox(XbF2r*En_p}1G>Sh0b*eXeRhZ*uuvdpLgv}< zV42y2$CF|A7mSv@mtI@G{#)jzEa*n?YJf(j0JHuZe)h#jx;LCndWRFT57;GJN)WFPVH}W+K!pZS(7M&Ey|$|Q^n z4x@}cp3|0X&POMT{_*eCyNgA2I{Kc_Q1aX@#ljb?KdA7*#bzPo5c0xtZERqj0y2OH z^tM;F1{I2M72Ex)VY$mu)Ux&~%J~|4z366G2UIzz_2tr#K>+2jV|*ORmM9w+s?1ll z@xaLJ-V1^AtHwjR=D_C10_j?*XRC|85CPNYz$CMPr)qG2;z$iTp~Y!|HW!rFR<*P0 z^x;TU_d=JZ&21+rf2F!$`_$u;6_Nv9*>2dSRl>%*+vwx+8erbq!7iVAV(LTTpcC?> zZ}eC-V|=Men}Wy;RA#tf^!95P0%G1XIz5xRfAKjvIGJ2A@fJ=|vPO!W#KP)0PJ=t} z#fXQB)eKb(dxO(O8BoGj^h&x%JG3G!h{pebPeyQBlt-`fFtjBkyH!CgQkC9~9CCJZ z6a_qVE3rS)z;3oXq6+Beb>1lGRuzhk1o$0q=AD2)03yF8&gm!@m;5SmTG>j?4 zU@qf$)O|>4YgoNA6WOWK^`b`sZQNYCT#*r2$|_{9dj(|!t6yF<0UhB+7%{~yGxUx9 zYRf*(2)L!_;-3z@Z-hs1+>HMshhCVyR=Ww9PE)eaC?FAy+2vZ~u66CFYsiyVsnR0h zc40rQlkD=_!A8;b8dA<`gjK6qc?M{cbQ180I9bsb1lrI-U^L#~rwtuC)*#P|$dA|? z7^`Sj_DA;#`>_2aGAE*0xob{EzyeBD^gc8VOY+#=aN`wyH}Ywc{|(iOcY%qAb(-|@ z9||vd4qc+1o4~B1oGA=VO(L^*OhB2rZ1=Ax8-bxg##8}TJyh#v*{Y#zdJ!{YqfPo*$TTA+zP|P6 zzakqQ7!$iqFi}X!@+p!>#q5A@uv)$+C~oS4kPUthg-z~o}X`t+s&GcmXsr zw8jRS;de0t&`Xo*xAmBp)-XwEIq=5ZLWK}EszH-&q+u(Ek-^;&Dy=tGdNx1<`7+)e zXH}+4Z1^<@6{D_u?bIX=ZUSK!ujj@AqFC4oayukUcX7if>6x#`jjU&mKK7H}$v;>- z9s)K8PJUV-U^^C&Iirq!hV(&i(R$HI1tt|X2DNb)P{njcP$p*?X;va_NiMr3w2#~G ziea2sUZ%Vus!18k+abl*15UBL3g|h(SB>t6gYZ>G;3*}JYRk|c%& zS=P6GF30H9Oi%mvVX|a0fjaDA{^VLpwuT}}RE*lMUW7tmTHoaX9by#x`o|~nZzi(R zf)kT_>aS$MWC9A1!+?=Q$yQNh z1(FA;Z+qfey>`PFZKulmymzxnWw9|@OGvPBeG@c65ve_^i= zf5y0sAmO;n_V2#))+O^2aEGnFvQX6hk1NoR&>dD1lIq*zTEuV2(J|HQQ|@KCFBM1Q!T87ReY^?5osr?9bza;2Tlc5GP@N> zf*ZuDc=vJ-q$rz6C|M07w12im$}>lMQ1bfgDZ^L7^^^=&p4C*$xscC= zT9+=`Cv)X431~emY1=>F;UhA@ z!0_Ab{@OAIMyMi~v^1cIHYZ$hSX@|Of@$L>kZXVdoj%`UV3t?Q3#kcT>tJrJM1|c^ zVEzHjP?7fg4}W96sxaBsK7Z@YN@)#*$kp=2uGhJZprxd`?ZjhZJhvkN0w!eXYmH=I4GN+xDM9x6Z zYdKvKT`#!{jU#;?cj!|3VQ@LCU0NH3{33?yfJ;s~?=$}5XzVO$bjP==T~70_@;j9K z;B)_WMS51>QEU1crIHGtI=VHeGUUnUKOmVL{AUdR{JzLkDSo#3=Q=FQ^aa-1(C3}^ zr)16r!2(_t|2!*hdUv$xJA+;*ZkR#5mU~iAPB+N$iV7%Zxx(K8N-22H|M|NyfBEn? zzyEcNkd`?SiW*)Co#cx7K&CNn-ycMj&J$1uquvl*!l(8aOFp*bYC)${yVJmY$2YnmtfO(g<$p)yEOD| z`B=xKp&z>L3)0$LGxAxFnbA}~Yakn+W8^=-RL@KN(GQK5X6@hKd5fHO;Bv_x6Wei> zl0oyzCCF@ZwZ1K)wBUAWL&P;!+~hy!QwiT#M429549EU_;eRUUE}}a5hq=bLy5VnY z-W>5J%@0OAobkpYDxKs+tdlj;8{GFu@Z%O!Td6-UncSp21&nJMbYbKTsN~G0lOUV9 z!LOg*<+h8ejQRrFX3r{fA`bc7kfJvQapRy`QMr#LW5Wn&BiM>@AF=>YMnI$7|NG@{ z6C=>doHCD-rOcKz#|>{=yQ!_odGHl!_h;e zi|*vt2|#-ruC9>u3AYNXLgT&G3GT30^J7895|5uCmtN%h?%L|V9P=~Uo5I2GSCUKz zwl`HK_GTL;E22mN6_d}(p%4^^P*b1kFUc0)4m|Y!Lsmr37`jD6x&b#GiHJbGM+Jv%}+5y20xRXJ|1-= z2X<>Ms2k;i`DhSh3>kz8xpBd{bO)`^b0Sf|O?S<-#n1%uNgiFr@01=ORiWvkfAJtL zgLbD&ygSg40yErZxCkxRC$@(k6MPM`H0CdDj<24SZ+~}^ars%iG~^%|bQ&Q9=)kzx zWU?eBQnGl8ET>|O>kG05AxAK%L&|MGsaDK}HS=!8FgCx$vR(O~rNV9I3bQXr4il>C zhV>$3Q8H*w*g(Y?G#XQ-n8TbuAH)^Og6IMPR!vmDe8^puL-gvS^hP-d7}o)BSnX>G ztb@V$w_TH$Fz42w$J|61-`gB9!w0;y`rv<=SH&>Fhy&Xb3pF&zH*=e)P*FC6PF8MU zx6#|YF1Q}?SE(aFx6|2f-Gjp)`{UP4aPj!233zdM&o^Am8-SSL!hthI7W$ab|2W1@ z2wmW^j6&w17SZlk|8Jog2DOIwe)5*P(ZTp`?WQwi4YNqA0|!Y;O_q~vN|s5HO;pTZ zuSien3kfjVQ5bo3_ExWD(zD)Qgf-6TP(qc#-|j(hu8o`Mwbr#Xuq~{HhDOpkdm~mt zl;o@}A|1mX{!=fR7!K2g<*A`;dLUV_T4x@0wGiCw5!G`OSdgFn#zkR*V0T0bFGJ7_ z;>Wk?c;E*pVb{=?rPpWN2MO|KtO9 zLbj*rZK#I!Ku?$220VJ03L11mZBBzOr(SZpNj@|NMrBgOJ&~}1Hhgp&eV(A)N>bDw z>40Q!aJzhyuT^yl?_$3Q0_oHCF=rcC%D;Sx@Nq0TOGt76#7PggJ+|=^SaDzgT3TaB( zr7fa<*Idudq6YEjL8pU|P<^?3Wym3tr>KYm<~1mrwd&h5-hcC$reXrRK$%Si@FWtK zKgRs>H!;8c(+9tr_s7Sr`{R8)@M-XQOb#Ejq-`}(qD!?0ns z81Do>>wZQUroML4Gs--b^E6w#1G_91(>wdwI#G+LPuj|d4$7etIm8TzxdHR;gxI+! z^TY%#Pwuk+E6Zs4LEg(n=0&#-YeX!R(e`+k`?QFVpiPyp$e`1sGXxo+$8goJJQDcS zMW?5pp1SR2RSS@>SA}MVA66VzXkAW8wBDTnk8+@fK^FxDd@;qNmM?ejHTW)x?%Uj3 zMBv&~2kS-X1(&@OSj%{-)9!_HH+jZVn>@=Sx4CO*Y;IJ6_^+Z9YMc69(}D|l=^@Ec zjKecbz{A8HHNTcA+$@7Qt7u=1M1+EE&ZBFm2I&UO))7R%Vv5>(Zd4^g<`N9>-6-xB5 zH1r+kZt3E1AZiDP%uUBV>w#&*Rp5SM>X0!$EYF*5kz@IDn4V#ToR_votkS$~M9%$p z-u;m5n+%1n!&51Zl+iVIflTAL1_yMU{siA#4e9|e%Tz*ERfmI{EUVxW)L53G~1n@ zGh_EmSEIq%_4Zd9$;uZd#0Qkh!&1;2DA{@nW+CQ?q%N>nk`Ykh4;e)a=B*0vBfI8g z&>Gj)(7Tbf9E~W&58`=nMy`o^91ZNJb6%@eVDEo{;FLdD5r2F~-CF`TeJ!t-b{^n+7z5-NoB%NM+SZmqUZatzw zK5VzV^4V$Rp7=~q88ikUT~s5#UfcF7n?7si(=8X&H7OB{CRbmu4AJ`fw1uOnam+6A(j+4AMUre&Ln|FfBr zflSRtD&{d!GzNCsfUO8Xp9A%npzNu2Neiy$8oZfY8qenoH7NO+Dn%#Ao-o&=a_m^Q z5vVN;)B3C1DSRVHmT$Q9OR|a?NF3Nn-f9ApO_VI1BI~G_TVm7~QOk=WR9WPB2oet& z^k1RTWZDXrYM)dawLyAcG9RTraUV4yC7WZyhlM${<>A3!9skbr&dZGoGpNU?VQ&h$ zCWa^q_#4>vfiB2|`WiSf)=so7C()X*m@RsT4ok)6Ym38Xoh+;t3q=cja)6;-!@ekN zQa<|HWk_M4QZ_3uaR>an*~RRm{Clvyz|BLY%R*i;zrgP#e;I4%L~txy3ldYW$eL*$ z-!+UtLS6k*^q_gtnn?i0fgKPF>1;#bb_3+CRW6RSmps=pIx za+u+0eAr-ecgcUMy2jHN;=o>qg}#s>k$F^+(xzKTa+Y{O2Iw3yBlNn-w zizhctz$^dG-#;+Ng~Ki@7G`>v@TRn$Urv__G}HI9>%~VzJ+G?eS#*BDLD6RWR+Nu0 zLsrl92>06a5Tli35~m{aa)fc0!grpVO}063O~MJ2eMb!?gO=poRLokpdOAg~FCIIs zRG1=Zkd$*mV=qZ@A$%pag+4=B<{#WcIuKt)rF=p=nhjIRtCWfj^d zD3+SOFK7#0>~Ze3L8m5pC;cd_l+{L;N9HIn&kV<)G@==#>-t0kPI0paoZ`I(oRT94 zoE|_I64og14MvdVL0$-3r)&LJLrST7c7b4DP>bl?YdD%(p1?}rtdgw?xWP|gVQCfi zzF}RTj{dZpi%F*{?kZWOSH_eSzc#4dN}j5XfFiF8tOmB>SA|M2=)zBxKAhX{ihDMs zpu%#UL8oib;Ls;(^Mn5UKGA1>^XKdItqD*p7U8fws@H#*`Q2X_0r=aNEneixWD;vK zbKR8e4n;btnEvP#c7qt}*;XhGAuClQjXT#a@u!I}d8#;01F)rZiu>SXKp}*ypSTlXa%ZsNDI+d`28WVqOwcpZd|C*hmxFU1d>0)MG z95x8N{3j`%v=Mc^-|;(27BVw?4!j40F3Mpg600d$B1Ph%3yNRNZ-Ne6tR3DOR2+f) zs#YcfdYx)lFz!rsA>K-sz4mB(6Befi)D?EeFS_c2>u`Yp>tmL=*bFewoZvB_>6zo%>SV=F z|GMNmM)M+F8Tb`)(t*uOr-^yFNXei_;Vc!?B~BOJLu3+JcHqh^3)G2C ztU`7dUF7m8;FK5ate`(;t5<21)}=VIN0b+&ma8RKJe!qNN6#Fn;*#ux#0EX!$vexps|6X|@oP=;iHkim#Rk(k9_#g-8;2v!xn~?U>}-KeX1Ts}b5D?g z#W{g>#}kQi`+{vSs{7EiJbuX0jBeATxZ5Q zomSt+Xwji+;=Xv>93>VeBOS!QbLd5Zx`-6Na%fV*I!A2cNR?)hhoCF7fOq26JwZ#t z>&1y)^^yUn6R#qfDTZFGS;MXV_lf{sC{<45JSQ7!YoK)<-rHJ2u9z!dTtlQSTf2zB02* zP050=03+glS8^bcWHSr#Ij}QaYl6>mN(QAoMaUzbN1qEm!$P{Qz1&SfxRT-ftIFYL zdg64&qp73O6K`kbDvUtDJK}Sio!G-f_l$sMAx+^`tx4K73Tx z4XWtd+w&1Z9a=`4E zPyV>x|6tvzRU)Ik;os%|mLxgw3Ri4mK{6>B@Ia@7%w+IpSwUzAIr#FES53+){(W*4 zRFLZE>#~A?gD=+!@?gQiHuRGu|H-m3uBSK3&10GLQHHe1X=Y1O(V}-lp7-+5U2g4? z3lTk{p`T4q8j1p0s1%twBOk=k3nO#s9U+%}vjh!bP!J|-T`T^tJoJM<#C-Vc4@Q@Z z{0aHs594XCbl?@pLVIN!S4ZCyLUK}}!sQ2OFQ5t?`msPE5{_4eCX-zD8h5>>A*5eg3T{*Kg%*mNm=c z^$T5HaGxk?dLtjgNnOHJ5C}_^)`Es^i)aDw>g+t&?y7dmZ_AN^17gx+F=(SY$~L&* zKL5Sn4;mN7|GfO2m&ijWSQv(9l;bJca*8aWVs5yen$s@TdN1IqcK^V(YrOTgdDQ1< z{ZKZ&lF`E8xL?_`+c-yEj84s`s?&cY%N*FL$uq&}dPr^iJ_)i!_`43H)!OFgz+|&A-PsMW*G!m|@n7qyC+_Gs9%3vS{(>@D% zcy0o4#TZh0h*^e$t~_yK&-46dpQ*^FwH~L!fd*S?y~<+MizwnQmx-^g%KP_V*as) zlrf9YIqtGa>P?`ukCIhWWG@v{1te~I=eEzK+Be&+MYKaS=+qI`AYLxTl4PjoRaE=V zNA~FfCzQf22jV^89LRPnWnJ;a3St8TsE*bJ&TkFsB3MuiBz!2-HRyyQaLDe5Uh67Z zlf01X2)o8=lcTuc3aDKkA3m8u&U^)X=KFEs^47fQ;y@#a>fTtri{wlmfsVj|7n^-1 zP^q9~kUT4gb+&~X;vk*bR4M+M^48@ zyk=z|NmmpG=;)nLpZ<_lLuUk@3vO06`lzFi$t%P0{=3lhf0?z(_7qsQ0>`+JFv882 z?a|$J?g8ct6BF;nfm`h?EMo?Ws>Q5U7ZvV_GXlG4Y%N$ZTZOyga}eE4@zcPG@fwN| zroqx7rm(Rzj;)V3fA4!gH>Z+e0uBdWm@F{jT$XNzHi%>L6u;9`ABpuH54ebEUDizL z;9}xXm7_59n6%N@nsS-+yJo^JYS0M@`t1!FrZ^e4+SnQzCgZvM;=NdNoH%TOjD?xU z?m6H#V=ds7ddBiHFTFF}LyUhKbf4lTES%$C!Jl#veB^;fkgWn0Sw+Ebi zZJn&cz8v-Ntj9d_Ci7(bHWm*gtQ|Ps~Fo4r@Ow1Vzq+yk`b<5^22) z1h|8NWUyD0m+B^zIaPZ5O9_JgUfFKQvu1w~SUv5-)(dz!t|_(V zW=6(ilG7g6AAV!p;Ji_Ea4Nasz#)`nCY_u8ldq&J4#-OO>X3Byp-jxf)IO z-3KTcR8>_`G1r9Yej1;*!pdM>^$v$GIFrH8}x~0y^EUc&-uDfSP)}SmSbtBuaAx2U&d{ zU6MSyG9+RARaW)_-4pnX?N$0~U;mbQEuMuUwgmRP^y%}Opr)W%d3x&SQlz8W2c{sN zUCf6QiO`dXhZaIk82PD+zHv#uyUJl3SlqPpkmzPI76L*GKKZ?v6ycVS1@7a1pOnvM6zL97TGCFvCeTd4}z zT9{2W$@sYd&@KYZ#d=Dk!LbkvZw3 z!ve?xjbk<>zCTsfzkW#%jArE{uXNX4b1DK3eMi;$%L~E|3F5^@vp)LsXBr%|W<=Xu z!wea!N!$NnUhnC!$gohsiA{VrUj3J|s$Q}tu->_yrRtWP@vCrI?4r^LO4tVBo$6o> zPs2-&%zzF+)pF@_U~D=f)p82l(-k{J>q#f%#yg?y|eS+NJ z?+Z>~WkaJuD!YGH0q|uZElC1viyJa_J#>wCzYE*|YeL(j>&4Ym3OTBDmd3j)synj8 zD|IT~h_~TE`XZ28CvoE3mvgi({~~HPr~+#RrgLnLs)EEuyR?*b$hiT8@Y>;b8Fac# z&f$kjRaN{`vKk02fF1+?I;^To#131*$q0QfeSYtQnbV9L{I75Q`LD=E2aYc7HrcWl zQZkT(%A;bs{kH-w8|c=B7YJ&BU=0b?=3`QG69k-)ylYGNo&Z(4D38oXkgI2A8^Dm9IDFJ%2&G@69IQN}*)Q z6iJ|BHu}$l{||41nO6`1N9sU{HR(TYjo`zYKVcibX96M{e(b12ZC?JyuID{QT}NZ| z_$qrLK^8_NN&_uT==4otBk2v~!e0Yvgp4;Y%vH;gtFJ(yp|8&bvAMJSs?dEwo%D*? z$mplPsv``Gw=fcjY>#^bG6FLobPdcBL-aZOf*Qp2&QxH&mJ(+1jXr@b=KQiMq$x*Nd&TFQZc&Po-yr6Upnr-k_WU)}olOyXP_FXRN6RGl9 zedI_~Ctbx(AlINCOh>~JS5*l+6YL{u1EU3eL6)#(?}}HOD?%}`Ne*1CVL>Y@mlw~@ zga#9Yr|OC)ibD6e=7sl5u_dDl8dxTdy97;xqgtip6My{Sbfa(c>0jITk}c0!Y2YD~ zC21!m+fI=(DyAdkL{MYUnV?4ZEZ^%h6aC_*S4g%5HY?ku39KyFE_wxUp48F0h~0tc z*I;qS7B{U6UYFt<=XC+jefr7H*Dr_dhI_7rf!+JYMF{1fLw%a86Vya43R(Jt&9>UU zGX;o+h|I9Vqpn6)UtrnyasAY@#m5MpN3x^KiTZ`fd+jv=RWT(4a=$H9OrfY=vQ~Og z+C|^>s_^WkFA6_yf4?4bD}PA$xFjud8IaV_WvmU66`p;o@pa62(nT_b3~GKI`e3WMXTEVaDv`)8mf!$fBZON1HD~4!euDw5*6uPmTA^ z)&FfszMbJ~T$+_YO&y3K!U+uR;B;6&)#IH!X7^jR8Y9(yy>kDS+PprUiFe|_8FUMk z>qWu__ab4Jcz56dVeiZ&_Eyoxpq;_HsGQd>lU`B{DcX(B%Yt)=hDHLeW+m>_@mw1> z9sJ}?o@=M}xgKP7C_b0$38;aH%?*z{I+In%$_S_?Hiw|~ATu)|W^b86X!NUhEt=dT zFfKq@+_jfUf&<583Qd+2H6`0Xk@ZweDQktO)+a4kb<-VL@UdFBK_7al^t#H98($_4+2t=F>aUgC5J2k2H?0k1?_|k6T}d!#iFfZk8aTWn0K0A1L40s z$$qP6Lu>YVa+%z8!vKAWM+5hhPc(Yc*6pMn2o~_@M0IaJe=r!qCz%xgcfkP9xVU^= zKIJbY-hmgF0+YpMBPB~iqZXr$SVm%bM>r~sy2ZQagti0OO!gb~bG5E{ifW=dPP*AG zQf&lW(ZwzDnV5*k3jvRpTL1I<4s%(l7sO??AVQVMz9~d4%T(#P*Ny-~%z*z|-}Ry` zbS@Abt@Z6A`((E~u_Vz@hKP}tUUC5%zqg7GLxu8M-!rUO7d#JQKrD^V(CdSo9gmOK zOFG?fB&|yuzk$YAHt}`F0ys0z^{KbE@M~&o|Li3+S{TVc-)bU>BZb}^IPzL#Vs$bo z*+z<_Q8A6q3q>b{gJG%!0lqvy4hs(gIS{@^&Yx6>z^UbleqVSUrZ+@m5sbZCRby#NzpE%i&!PAea(>Toxg}|hU95gXa_yN zL5wtXedM^CUA|r;Jql(|bpm?SjK8F1`#r+G!f`nfKKb9Uhvtio!$#*VNS(Axla>8$ zRiUX<4D^F2McGJaavnh&eg<9Q{I@QsOjsDG^G=nPK-#8c?iDF$6ar=M5O*PR9yW+^ zm~^*vx4SNQ+#UdDie>YqL;rrHFUdHMF~joFpMG@GyyBb5Cdh$(TMPB!rJk*T34|3e z4Y{3n(HA~GjE?ilRc!KIEIt7N`%b#i8|qRT#0P}!umI%?ucCA%RGg45*P1|my~;A) z!|3wJN-rc5S-@N3r)m~e&pztNg+ zuK7I^@t)Z?*J1nF*1JyYd)_0mhavG_MeC$>1S`%|`8)$_3|85eh2Y=YeTpY0B5wqq zHJE+;9iIrW4*gz?7K@+DzxltDjmAQ$`&dO{9XKMJVPeISDOmzVRzevzC|IXT(RDuL ztP`QuEHWvN-@r4UG^;SdeAsc{!N`tH%5J~Z=VZi0;TOO9D_Ov-2FZav5VZ+{k|^0K zimaewbaZh<0Z)Z05^Y>4u5EyB4z=ys>PDYOs{mwwqL8P4(RFucMLq9{n{LUrAhlmp z=wTqHyy3b%KrLT7xzzL2oMn@blLurA{V=+7s@kuO9G5rPqk!zGXR1G7-Go10M<#6Q z8CeWU()FuI^Cm+Jy?hn?Hm=&QZ2FB?kIGa#qOW#|?mV{5^EgyldVKa`TZvPu#CrG~l#ihK`#(P3t@0 zbZ6>CFRe>6eU{exHquu-n<5rP9|ravY-QN{=0#zj>m^SdvzCJz98?I>!xbi>Dm_4_BiHv+M1zG?JCfJQ&CQml5bZ9JZ6+ zK*`oq0H&B4ZXvtbt45^b9wi3_P+lK#fSse*LSsAWm52sewQs$+j&5vCC|A;@HvNA3%{ z28C^XVF|$yV_7jnwR_sBIUNzZ0Wh|Rm|F3$85G|?27oogj{6$RD`S=Xte?4Zqr>KS zENC$nNAv>Yf4)$KA_1w=3MA{+y5s{NAw=6iK`Mn5)5zeBJ)*s&ioPR*@HJn1%kVHrQ` zL#};&iV-W<{$K0g$?}n~;=rLdjR{uLC>hwuHB`(M&$A?h?v^eNzZ|qUyp+BBRTX9e z+d!2BOR(od3%Tf!vrUc1DD%Y-8>5V7yVI7ju_|VLdAJc9lK)k0A|=eWI}VJR<0h!7 zrex4Xx{Hd*3||gG|AoA=z$Bu|S2TzXA`^Z5c4<9-P3WE=Efgvxu-fP~q1Cf7BfMU8 zGYnTK!;Mv;Ys0q(WzY}cOsh+A1W4IH_!l*7szRH+7NZWq1G;DCE*Yd1pvb=)Xh83} z6xtv##SmL2tJE{zj2mR)_sOL3(5pBwf-KOhw20zfIU<7f3Dl5eZQ}gBhp1jfhM94 z6zXymweGm9coc9nqJul&bi}(mdcY~yT^o?kIz*69K*#Nn9Pp|K7Onwc=Ca$9XZD<* z{=(V%`3{X|*|d+~Hac!_irW(y<854r6swk3lTDM!9uoszM9HA7MnlCI`lwQ+=nkso zDBO;+wcXOavv99qIHE$v-{^N6^qD+wdnNyf|)T;+y~YQH9)y7v<3p_L3|IE z9@zv%`&#cUz;pZ1bulZ^uLYC=fc!R~XQocl?{?BlCt1QlRoAnm&lP#zmbxc#45Bz) zp2ZQET|`!Qps1DhL{H0$vLAE}ZA*7@|_W;}%e6^gOJc0mHY6 z(a__%nAr?CY$g7ygR1YF?|7aj&vD!x8?mqp+7na?Bsph1vZtxKBX`ZoQJ`)J@Lzum zNUy`n6`eHp9OG%V99NJ*r`}g}-pQPH)+HgP6fqcA5(0sL3<6y7?3p!0+hl{K!wh!J z06xN^VJGBU2Kbv_eeX}^Ym5caAMg?BKj%|hsXwFo*OJL~^ikP{=su5fiR~Yx*>g32 zYwdG3zt7(9EkEbyD%5scx*)>cxDEbp?WQwi%}D7$#|;${5St&iAZAlCfb=FR20BOS z2f)OoVIlWcntz94mk($+a4;sA%)xYV0xBEkEu_ti zhP@H}l4VF|2YgU#Ubb(6)~F=9O@0D40;$p__Z^}G-i_V|gk7GU62tKti0x+x8bfcs zyk%6XnjGg4`_A0arT>)b6elY`TmX{e!N6k7GgO4hXE(Unadwm zz4`H_pr?+j-(C4e(zl1M+W%(d+_WE8&K+u^Qzw{dFS|dE9Y*Z z?|!S|_4v0mR3f$M7hQ8d-XJdv*+M73`gqigK;*qYKO+mQ!=jK(!GXC;-q1yiW)+ws z%F*Db}4H3(l<@d3;xSIsas%c4F9?%`qT@dQ>@woA3%`8MGku=ZgyoNjA}9rtyN%-T!8c(m|O|6WuWwfwfgjDdx)!6Mw{k0byZr+TwPewMvGfJs??VBtzE#NU*>fReNx4TELsg`dWr9 z9=4fyYeorVeX&zyv=;ok{NIu!W@WSv9BqMGm0{K*lac`;P&yS8?_4HbC94+giqeL5 zy*40eR-WM3OIp1uUAC}tA`h@qd=sJzBlGA3?5%p{$V^4L$Anmv(e206Ylt;dJl<<| z{(SH=whFA2lwlA_rDUrqk_cJ~K+dJgBC9#4r>sAjssS|>Sp?PD1i=lfa;^&3`%t$oX(tSbJ6${F{xSpHgQAcE+5Ji>6*7Poe3YRFX}3B>CQcz2xR zr*yNiy0bhoRoWmfWc7)#>OML00o@XMbw+n2l!S(F2)`g4fFcEGU5&^HjCV=m;Apst zLESJANE8S%=!d~&)8io*l{g#6d*lXW?4o|xDsI21MTDd-hQDLq>Yke_t(uz4%LrQK zre$XY?)1GnyOy)xtIefBT+2dosbWc^`!!%yegbeN=zH2e+5E_3VEd8x-!&(3vcQ#; z_(r4Y9U44dM%ehOpy2M6b7EcK7twXx0 zd3YL^wm(wp9GN?MF@ghowHAsIO4v0tFknC#L7M-4P{O^*IpluyXQf|T|K?}k-tm=` zuV058L)yG8@N?h0wexO$|N7i!CFVEUXuNWbGYV3c3eJ!s_gz%l)V1s)_rg~e*~LVT z1sSF|fp@kC#xpXCjdyRq_iLjAe0@&hE95#e2iSp2Tb525?f`#5$$BZ$O~vGbql1-= zpNW?7J`s%dONX^&5;i{l?S-0f2Fn{VniK>yG%{FwlTod=us(WFv9(bsm zy&z<9z^d@=k~<)?EW-M(Go)UEr!WD~8hSTU)g(X0&rwu`E_KIKYI(C)YbbmKGI!`i z8$DLd*c6o)u1cP&jp&P9F?$OTW+g`rI$@?^3-ET_Cpg#%4-K6Qlpa9VQ3K2``-0Jp zDUU3UsPao--3o-hwuq(EpvopeFt5)w5hNqpxL2erSV_}&&6%f`-wV@v=a6kP9af$B z{hx*pnzzea=s@oyd5SD{IlYXvPgu`Ii34Ph>T{{$uY09gxl@?VI`>-n%Zq_0Dm$`S zS;Fomoycnd?W1?3ak5sg0`M;r?8*W+uCKlwzVbsuGID<$cWvmn@u6-!3{VcdQMbSV zh43wrByn`K%E0CU^}J{vu#+CUomH8j5|cW!0Cr3hMvlrHI?wOvR|MX0zd!!d< z(%qTvn)D={G~$K}2nsHsEFvHXD667^Kone}0zpM_#bPl^f`CX+@PAHKVW^VQdQio$ zv}f{}=G9v+`0lsPJ@=gN`JNPRtC;QL9*X0~u3~#A7UIHLRHTMkZNGuv>sk_8Dvhyk zqRV`+6s`9g>-kqD)nq@vRGQ(nlC6fV`gzhI?^uvHyV7y*!fR6eeAD#-PuKbbatc!G zxy;J(TFjTXua7qH63QovZ=?d zcAnG1H@)T^)rZCw#AY#J!X}`X*d>n^q$=adieL=9p&~|x6IAhN&!{!b0~#-3tgk<` zAJ~DQ3g@d~uLPz-=2^R_M2l97ZZlRYzlOOs6D~00K*9_)0Y4Iozs({LD}$n zV}b^i&K~+=1W2SqDDG zqn_-4=ZtaB4lelv8{Sr%Xg#7MOvMW@KX8*#_kjH56k)*d-J635!zZTZ(|ZN~8|JBZ zQHR-6Ye?#N1b-WzcMxhClTR$9Sg3K{t!vhR`;-3ppf-3?fpp2J@I=)f(K)vS_eXXQ zU63Q$$KaBaMz^p?n``$NBzv!l5-ystG8%{`>T)~4zcBwov3fzB^iq?yjS$`OBcqE}xKvkcm`uNd1z*03KB{$0!ZW7CoY{ZI5LUTtdiKYu zRhgdR_3F=lu3vV}&Al{@T(n_%mHP%u%q@!TpvX;(*!P7(otOkAt1$wm;SC2B@qn~~ zKU&$-K#z=~ZZaR;Q{cxVuf4yFLZw?2enV%qDsYQrr|RM2K9Ji`w+Z85e}wX6SR1cK zE-lhtzL0AlD-TlxPaza@=sEONY41EWD&e96ZZ6oOTpt`iRk=r0>>34@CyUMo;=)0o z<0U!f$Zm*l%A4i2{_9=vXqa3Nf$>3pke4T^3QVK-i24=hWsnCLFA8lA)*hP@s~tQE zS58=)R-y{~d@?aw8+Lw7#AuI@28ZnqC46;UAUZU~u34@*^kG$^msb1sGHBnT>PAUO zr`7zLapnw^2*_`L_XqW#-g);IfBaABDvDiBktoxsd2$G=oQc}2q*JkiS;Ij6s(%K*GH@fg?B7c;J$2S?(+upqT?#UYn544) z5;j2_VwJfbJ18d`LR~rbJDzb)@JT@h8;+ltNML5E@@0GIk@>d;>m0l3cxXZM@lB?3 z@jXR4CT-ui_nobOSMb{X9}dhr43Y{1^icRk2K&`qm$nIW=(BVa2)GQ+?6YQn?g`>Y zZ&$|7xMhz`<0p7}%E#Z;`!>`2eszZ&dttIZHw`?Ya}?V|k+W1}dDv~im%=ao&D=MHtob0PnoRt)SJl(L{A-`>; z!fDr{ZKP3I=~&>>7n&{tvLt}6#;sG8$?sUO-S6|)viW_GP#<T~cl^@e}H zD(zJ0PV0uvfG$bVUX_9U5{^c0a)>j%pgnno8zTQ@eQ2z4mA%cbR3^&*2MDtNdtiO4 z05a`*K`G2};==WAH%OcCtSHB^hPR)OTW}=Ra-249FLUhOTqk_&v(`K|ZDPjRQc=g3 zCoW+1X2p;&LLtk$nmpw6aOk!(l~sdJeu zI!}Vc1K7R{G08&Mi1XWlz@*kz4j_98^8kIJ=+#;cIb4xq{FS#BS~wwP`hO@|E1!3y ze5#BWT>n+?cgh#1LZkMK+07AW-?%2l#xSUUoBYEL{y7kQ>XK`iC2J#Yh?jK92Y5(` zJH=k&Z~E~azmfULLaV_9rg>6Gu@+45@8Xi~p#uMLd57`V*~FgnkY53)G47(H$&k}E zD5n%g2Osj$auPJ`10IUK zq$-a?bx;j2UeGCA#v6815Aj!HBp4Mfqa0%>jMTNTF%-6SXq9ya#qkRGn?x~i#j;T6 z2xsHPXY6N-f}0bZr*~|e{Y&H8Mw@kGO;jsl$XCaHgzVy`z_y@vSuM$h;Ox$brASn< z?7Tl~vN+)4SwmU$IwxFATmDn_ZQiIgIWcgYEVl=jaBOmS6UD|;B$kRi{x-(Da$Smf zpZevfQoYhW@N0@iH&grK@oky;@-6ye%^8f-%eURY)!{$Ba#yZL#mTq7ca?0k;WY-9 zlrda{ofHc!$?a4mM#)fSFqVxMXcTpx1uhyzKCs-7lTstdkiO-u1TkaJ#8j>$1WXkY!3b=cMmu89)17D6Zz#Qqq8>5NidOfuJS99=o?7k1~+Z0 z%iu@n&>WPE-aC~)O`}Bi>dC(6-O1SLwaK^4@6sGK${b^LxrWK--%;F}m#RFiEb*M; zrOq4-=*+}aB7ozDcKWBs=WZkV<)qjy{WMu^!^=sQ!E&;NViPHn0Ih(qiQpj>M8cp; z_Hku8C~%d#VGJam*Fa;7t0`utn2(I9SUJ8%pK0qbPWW{CnZMlke%>4YM2S`eoK{xL zdg(2^cEv%j@~{hnD{k$=VL_BAKXk-ycrFxu3iEu*q(e@P%8evH6f!K*>je`tJoZ5e@;wMu5lB4+R8_i_H3zH5mFj%0{C>GQLw^5PH10XgQQ64tHw-j~xhprJh(Brh76bxLQ?BtM@8uVDuyg%X;>QpZLc8uu8{v#ZBgZ zm|AyVqSs1a%sQX|llqQdg1(j1vX?Z{LOR<^oe8_= z)X4Y0>{!p!9$a$n3eX-v4JutoZ*xdh9`iil2Lm*K+-{7cx?0x7vsvi^PAl2;y0096 zHD8arUw*crfE3v$6ni`6&qF}v>toG0Kz8#PfT<(^tTDsL4Dg-a_dz}*bnNkg} zg1#oL;#d1G?ULv6kXM_fYJjSbD$vHx4~-G2u~AvVj4W-^Dk@x!ipuE@IM@|&r63D* zmXDC7C9+!o-OkWWQxz3@Lf9LoWl-V?yfCSyY|`u1SUY8O{7jSclpDBV=ttl3Y4_7( zXinN+6G*l_IchLf#(cMYJMOjXKdgTH)Z*2_jl!C+1i`1`d@u@_ z+?~{HJo!Q_H$P8~J()9{=ErseyFbj@|Nq(Pfs_B$Z$Bha+}t7?W@V=u;6I*XV=1zh ziabRMLQqbpV*%{m3qsDQ9!is^Mj6ZqZe}6d=$Wbbxau#<<;DcJ7i4E;f)2MlJlDU8 zJ}>T3q$w|l9P`hi*9W&mwE94|K0!6?w~qPukVHv6Y|qy_SlGOoo=~&kF=oLrJ(6Xw z#D@Oc9l_{Qv+YO6$toLmSAfqpCKa@qVmDDF9?DV`sE1L+t8)M)u5S9{A^yWx^CSmo z)cP9G%H~hmG~?uOnQF^U=F>;xw8l+q>ZXVPw(Q0=X2;wSrjFd;bp{-M<&a;hsw3!L z5M~+Mg8KODRri8YRfl})NK@bnYj^SfdLp@i3C+jXPg?iL!?mT%IQ05yME~Vq20eXuK#*LoDLJ! zVq4gBc1W7>E90X7B4M zH4hVC-#=AHJ~O_U={y=fyF!+bgz=%8vxXGjxWthsb?3bi~oh57?a0?BwRY0r-?F4jZk`^Wxq zA31EpQMF43lm98jo}fq_6`9Vv#vJnNaZe9!kPrK;`g!l-|5^Q}uADfQy)Ld~F&mgZ zyAC+lXUTRM>K7!=PUIhxB+F13qXrU4Wo{=`cvo`-HfLxP>IZ>pAm{wh!JuKERIld1 zHf?cX9f=E1kencKl3u|!|EPczufrkhn04f$IL-Gfx1$pcxLDHi7z0T6*?8*B`^?Gm zP$eH;m}eYp;j&S(VRzX?;bR)e1Lld&((3(wy`a3*Er$YSy#!n{Z=XxQ#~POb&>3u( z)z8*=Th+vTJ)hjM^!2mWJepmnay6DwvDw-$6Ot)=`B%i)eLhv$Mr)v(Lkmc6l`D1v zp9$O5X|IGA2lUU6a&L*)>D4COUP!S~^*6FlpRpe~i5_ z7|F69xLFeSUsfM7rfR<+mc#_HTk9RqS|L|s167bd*x{PEuqqHc((O{!0D zIpN5%PvAt?z>2VT@!=2k0@MIR>01qhvV&!$TBu$gu!UFd7RRgg)85_!YvT}mZ7%TMb!>1zT znHP5?2?fS%e=f$whBrv11`x`oSg2u6M?vKHZw&Fvq`lBswO0@=x+=ZpHVhO4ba(H8 zKJtiN8;P2=&pClf^WE&8?9>u51?9RFqb>NQp4w&P@r;-=$B%EG!UIlz$@KbwpNQt_SD)8zvcn{IZ|xL^`7KQ!JScheT! z?wA9fXRjKgcUTTxoWMg(`SRnj2DHc%yD`}}fwbp)59v*PV8 zJ+8+=@3@Vm`8F=DqOC;yehw5veD&s&`SHw?sp^=bf=_)sTJsgFt>wN+1~|Z z9#yT^Mt-0qK)k_gH=XWZ1(+Z9IqP;jEZsjZr1GUfr&IK;d8_U5uUSWK2Mze1Wa7f| z0o|j&xru(@6l4E%3|khKqwqBA=TkBK=7#llDm~CnMt;1B6x*uCr~V5KE^?j*gQuzww(`hiCkk{;JoWwfX$EN4!W z$4a+kQ2Af!R_z-Xi1+u(ulY>=JHQ#;W-Tj|=Vfw3_mER}wsF%kn~e#ZXhYVZC=r*t zopFCCRbxNoPiE!Nhx`U*5Of)BtE|pd?VZ^uED23vJ`og#OwA5#LIAOz)q6ZLCxA@* z{npnbXX;m>)>kgPM-uHxvB4^|i()}3ER~8}X`kgBFRJlO30`HN;z{G6uRRII-b8r)*O*Z@CqKEi|FHDah&}3@$tWJ{DTS227G15AI)-SIn9RcE=GqP zv#TBSFsXmF`eU-fhCSY$2KY^+*aV8iQIUuJR|U7rT4fu&YP{7Y;#OIduDSA6pD)~| z#$=3vmHBub^UxF=a>DDhRkg4FoiQx9M0#yFA#DOgTX_|JP{}OG6Jb9gb*t=Gh>UPcm0_KCr>UPfH?>ap#{~cq7lnH84jvokag$xH_FThq=xQ=tIocWw)W&k?E7D{B`M_c5bgL9}hU}L;vfCp)e%|7YUQfaAT9l@*69K>}iu=3Ic4RZX{uE_T4_lBr1D5TbTE(<*}y9d~v zDEHevw};;Ce8}&}>`AO>69&`xhSccnxHvKEfoWMk|#zFPOyCKKngSX7Zms`o!eV?SCB z2xCSvZC9?|Y`!~BkCyi;BWg&#JvnJG0f#9TW~3YnEkF#iIU-6_5R!Jq*;U5dwG=Ywxci{QR`dbNFJ*i~taeUzv~+!E0p zRxbgiCsGY}Mv2Zyazc9rXC$|!7MmUm%*fc;vDjAC^o%@LMQWGHVqarrmKS6RGNI#w zDP|=7FB4sWJwCQ;MiFKWljm|Q`S@3>!@I^4d z`HC%xXkPH%{jgf$e`}bUFx(+c)O?zcCIh2r{)Y*|?q8@Mbxj6MHXO<|fuycEY&<;j`tvO3Q=!Fl%*=n)MR7?j04sY0pK?BE(tby9FAjS--# zz)l*KSRdK7Lj#ivka(UDpsbE2z}-XuRsBUlhOq#l&2F(xND69sy|sQv{NtcJBvsie zL%>XEmL~EB$2?=!B#*g&st!3hSkw1@wfGyxZm`Wl!i0a`9#Z0}6#^3?4JB3C<7JSTKUnXP^#)Bj$_)tsUJ$7w|-rMQ= z&6SVHn&)iKonzqBY@=8J|0XID#TN861RECaxQ{ujvSs?UYG+WbnkLD8VN3w3J^iW<*z5tWYEIzCmI$uDO1h9?34?z(ha z#0f#3Pp&sKie|G_8T_PRY=V;NT2wE2wf)sq$~W%4Ig6NdR-?#pY7?%s-{hcmo_#cmbas_} z8kC$=Lv>pp|F~ki@|rZ3J+`oVVH^*))aYC{I%Lg)I(_K91{8AWj&z)^3dE}FB&I=e z3RvJnP)_^E?>wFE+V8egR;^e$$LtlyG$1GXT^MJX8FR|??=$zxyKlMw$hh|+7ZBNS zn8su~XiO-Lu3fFNI)^IpO&U9i;0E0SQ*G|6*WPC~E;x-k`I{ZC&dD>zh0XdXCT6-q z+j+W87~{Wd(U+lL1yw7~i@T-Z)`F~J;UsIuxW8drooQ~X;l{`x^H}}X$mzkrC9;Lv z*2sqCJjx8VMmZD<>7)!QGKo2?EcMi~>}F`iH%=<5V1;S-O;v6RJSJ_GT@70yAMlNR z8Hjfn(R*aKi@xmNEk*`vNoba;g^i)CX;sX|)tGhT@ga8@oBc%Piq!kYsvS@EdA4CE z#Dq3Sf#VRJN_79LV?;rM$fXHHqWt3e0w@a(Gy3zeN&f zZ7uqV!@=yCozRV@&5Sd%_;XrsQU-qPb&5oDGbuLgf}|Uml#LX-fgx>CdwD7hI57Pai>OMt( zXbbP$LiM?Yg;2(t7GTZQW86726BHJU+>iT5jC1l^pRyN?)m&bX&58-_mmJ6<@1cw6 zI9`X-5Hw7@61axB18Lq4MiY+lPz@+8Ag2xsSiTH-=!C5PG^*6IS*ZcV!&Z` zQynyLvC*cUF*uwPEgBDs~hE0n}_C%jtR~-*T;O0of$FqR$ma)V#1Hj z62lXr)IFUF2BgxY{j8l@5`RCa7oC0MkF!8TT>TI(f)FBaW7y55Y&F$z!P8dI<16b*2)9#v@xAi;V4|5ORBFA5t9dL)i4!DJ4&rzg_ibVf(n<@@E z-J~in`F9GN=vDL8So3jYc9*0eQ6mI=Lq?C?b{+`hcz#T!Elk9Pn;yd~XDn zgtiH>Qx1NvrZ4j{W@9}oK9t0Kr5)gr9kvozA9vFz6Zv1w$6r-GCxcuM(Koiw46YS? zC5xYbMLc46GpGfqlgIpZE)#Cb_lSxilxOzVYr@d_dfDQ1WG)i#eKP;M#_?JcflhpP zt3`W4u(J^!_c76efqBt_K7J3$^(p4n3RXF7bRG)d;hHU39gNCe zmUEMM#pb@P3EgMf`j3;`x7~QB??>xXo_F!wM)^Kbn-CNPm(5HMZk6SUIDCUAH1{SR zG6$+A9<}OePEeixRnsE1F%QLNbAToYGVOF6i66 zN+Q^$($Z-$xRS)&3u;hgIv0Y6p}s+~gEK*FX1`3mRR-b;Kr-p1)o}voj7C?3B8S4)a;CoX1Sl7j0^Rh;he5nddz%(&wu=uY`0;|95ld8 zKE*;8s%$C}O$9_%K$8k%v8R>QP}vSWW14xZf@}TPFewZ!J>YLkbLcNapd=mydZ0o( z8+4LK6bBqiLaPG%Jj+M}Kn>$@Rx%uuhl}MPGIk_RaG5s!@P(tEdbkX;r`C|v7iK4Y z*Z^FG6bsDk-AGTl&Y+$XmXxWn5ljw5?+_=;Azd^aRK)A1A?^bIwQK7H5qWSU0#O_0 z!zdmivevIskTgSgr!^2Zd3af8X^Q=C&$r}yz?^*hdsoRu8@4a|3?Q(RVxb-Qb}F*{ zLi%&R#~nz#c?U{48+MH!w|I@%k+zWc{k*Z z6TR*Vs>Hg^nhpGm;#^)j1j_Cc&`5T?PY%4%MKS;VImHRBFwQ((#BSji^e?w~%#x=vJ|0-x1bgGWbR%0`~cDe~P3Dl=Z#_U#^uBFvZ zRV;f#5bgX0uxJ~Fz)k^u%0{=mS)e26fpPpkCKigvhMduzNn-Z#HOv+9TG(1+GP>Ot zgcEw@$+9)!BZ@BhZN?&ntY+b7bXR2VPYWRH-#e-bjKu}Z4PN}?w%VJ0GWo||`Ba4! z<ns^xhxOKJVSZvp~zFO^9tVaYJ*1$p!s_O*5`IcGIIh zF;P?ndHB~|cet(;+;(dc=YXnH8W7#q^YbJ%JPkEk+a_0L~1Q5AWrt^U6F6 zA$eE=OsE=OAABoNpry^Zk&i;T*Ik#1ZbO^-9j_4?stcs?#s>Eu8_<)AVd+*`mTxb8288oIzR6_r^FvqoW6ue2R^B&UD5Ho9MK^Ndx*{3r{vr1}Zk}{p6XeF;k%aKF(I1C5f3&I{}385`4*2X&*@ou{2UL1H%2ketZ-WCZMT5R ztANdZ-jz~{U3L8F{R11bG}y*>GY8<1?+fn7g+>|SStVjoas5P^gmw|&fMVYT0qL{+!nDyKM5 zY-)0@^`D&ab)A^Czd6#|D1JKnN5uxGgySFi$ZofD69fr!=)UJ7)>e$5-I#Sb^7JcZn+Ywkd)&rnvXqt1y!EN>H9(n+(V%HoYJGcSRcGqxn z$iR{nP;4GWa;V5uucLwko*M0`=6)V7DK{6gR&J*oqHB8Q&N+@!{`?7=6AlQfW@iJ2CiF)gk z^oXB^OmiXs*xnf&JZ*F2=5*Nff)0A{(g4E8-f@-88fW z_so?(3W*lwsMP0OPAu4^YIclw?D)S9B04;MS^rV9actLSc~U0&m9@iGqEX%@?`AH7 z`w48^Jkio@@H%$^W5J4F^>M$DW`F<3ecE~_IW2k9yLSBhx#zu%sF|0Ut;^1JjssP5F^oYZ7-8JNxG=c^D47OUcm1Tu{LbXhKa%Q z03WP}5bem^U@}acUbfdeAQA2FsmZbz#sLAE?wDGV1d5HL$T}+W3#bTam0{ij%nV8g z>QWolhRZm!ZXSZhdC6asuUuOsgzSqAn3%=Al$N9Hx6D z@eljdf(dC>V$;JUxZ}M0zL(Ua8tZMqeZiACvsTEh0Au@M_}Lm*;ABFkd3eqE zld)fFv;Kz(PqW&;a?w|^UWlUBg|sNtx!&2qEy|SeFXTzg=GmR9DK-w~F0zglK2 zC}^_=!-Pa2_Or}VJ)ryerJkE+s2{p$l&1q)*ft?bIb}hxqFS&V<(yE|1SQ`(_-T@S zS-(e#XOyFQoBv)u^ekjnzPu^0)Ke?^w0U-#_p(UwB4(MGVzh9sF~2 zn(qMVrMu$J9fvxIx${zTmo3 zhQwIx^OR3(ypIsPb41Y%8m5~C*iO2Nzgdt)Uk|<@tA=i?J+2KxUB9SCp$4ky_lUk? zPX~ z%`*xk!+N#Zxq)tB)i+&-oUaF0)47&wLvHpAnu#E@$FtOpN4C*wV?n1U1^8{4IIzQD zG2KA1>nXB^io}o`MoZMXzTO@D|2GAPX7%y5)Le2I$F$}6*?&xX{Y6B?Q9-4kRTd>u zH#lTUb%H#TUl_Q&a{ayLSb_-*JLyMEBjU@q{N8ue`>-FV&;OCc*n?up*l6WeicO-( zMk=y5tQ{g)NEyo$0mBLp+Ud@qMtOGdet8;QMc;wu)>HY~&3W68UCl|qL-PkP{kF%^ z1GHW9`==!Tm0o&i0pK#@EuvO%2et5K9hq$=0WNQ08( zZQgeUY2HYatfEmA^VFioMJ*Ax1?!+1F^!G}M!-i00tU%vuJLm+Uqas$6fybrbOrsH zYpU|x;sXIClFXN^@r~ew{KsDbqXY8+ZT6iZ`>JLWJ!FdQ(od7s&sk69ECXO{q1Z%< zBv6r9zw?z0Rr8UYsnC!*eKuX z9y=3)P(#i`&M5Jo#2f_S?rypu1hp8DfQ^4iEDJ))&>R429V?`_8Pqcbt~E$cwYaSS z2Ot}6wOM+#pmvF!aR%VYy^sx~40_RxNg;2h*i96P2k9p|S6LFc2Wiqc(XFx_AX<&Y zHpmu&(wD^|wq_a&4u;9h5iN6@X~XoVKfSy7c~2SEee;?$n!7A= z+VV1XHl5Oy38u?z3|MAr%^t-x2F7vAsmK#@zD&PhiSLSkN8)W*7Y?LJ#~6oniiOBY zG8Gvo>4oK|ibk1Zb(^q97Eem0ox+2@y5d>1?O0^fEk7!VliUE|WBhL_;9@pJ#=5L# zPxQoZH%n!TSea%WP#_4 zA}a0lc85}NyR0M>gbmit(0C$Y>5<(Hk~0TcE=X=kRU+DApwub{8pxpXDRxf1?UEYv3_1B0)DVR=AsvwJ&^Z2qZn$(M;*sg?Fv2igu*q^s;l>{i-e zAUz-mu*$v~>H#q{id)VS=+!hxV%aEBr6W{^$_^_Ny{^;g(#^B6W2u#9;_*A2)Ra6P z#qtN-ro?7xD&P3l>@SUbH#`|YvE6J2G11SVTRt$aRR$g`IHJgq8<=-gik-M0(9pcD zD)0#DmG1)@70RTy%GNojI70iF4n8){jThhss#VtRd}PKWJB+?mC<~!LL^qn{jn1S1 zZbhSG8sH~X*ZuQ--@rn>a=ivv!|wB5O*MfhXVLMRh3G-d)| z0wYiE5z_z|C$KV2{`Z}&f7E*_64}3mk)9VO6A)*R35c4`4pU@^ifoi$ht3|lG~#i( z5%$t?Jlr&6jJ`#QO%E|h(;!|U-y_0(IJ7Br)Bc2LJmvs$=+%xlgK}X5-A*?`puLYN zcLO@E3izIQ62tls_w|5Q&hgOu5GzNtqw}GfguXm8Pl8E@ToEd|l+cyh-ZVQsx)*`0 zhK5;7PQtHS5!~`hj#IYxBfINBIPH>am<-Y;+~bEygRKinU}3*O%6ytY&IJ+~_!xF$ z#A4_MS%+gB?~Hqn3hBZn;xFVkNPCD*@L9vN*<`>7Km7ouhv2aD@EI~<2W=F_`lN9b zyN)6;RAd^R!&75@A5P$1_^@>8L7u)t}hdlX;5USKnrxn|Fg}T>8Bp5PxX= ziMV1TTUw?qJqU$`tBgTU3Y5ojLND>w=| z>iy782!nKW4lNNFv@3SK>r^JqnRQ)~<_#6@Y%%Pij)HAEq%7v=JDrnv3L#Y~xU{I! zL&tEyQQE^w!uI)rJ&~*iW^Wojpx8)`@>6Ce2ynC!Fe6(eCEkS~p1UV3KQ!LIO{n`@ zvS4hyGlx#(q5Vkly0qx7Q;By!A4e?-YZtaev@&?*s%ZV3%l;cfkAl^Tai|1Vagl0( zhj;bOJot4HlzHEaubm%jAB}#OIDIdupIrI-Ya%`T-Yc2$XR>}g+Jp@Ub3o8(%!YnD z#csntHuB1A*Q8O9m-_e~GyxbO*QAAuZ!+f=-&6F`C3FlABRg#b6@^L^hrXWE(RXO2 zpPV6MrN6^?SksVc986yKx5e#7&V#P=L!0pA%&34gPuO>@{0GbLsf;J#iO=St<%!3v zexfWTxy3kTV6$5*6S<7F3vLV2fosyD%!T4IZS~Uam$K+1q|O1Drq0#=Q#;@0O|^Cf zvHJ9m{RO5C!#LZk=|XpJ#asHdq-5s9cL-=%+AJ{I4c3hd6bo6(M%*wZD&BwpeVrNt zgk-?4(UFZ{i*sC>D>@2+U3H%5)3A0r-l0_tRUqvid6GT|MV=zv^wOp6Q0tgRS9(-> zln^yWCPA-<&g26}2D?3U(U{#-_k~t0Oy(7{%Y9Kwb@hA<=M*ho?hCzu6T;Ssh8c|# zWi4ZZlV%)KG$@j0q$=TW>U4zvx#MtA)lI{}a5%sWo-vV zKfVrMoRBr`_umL|FpdD)EV4{Q``Uz~MP{4bkMbY+od?a(YGAdmbjzl%x~2ydDk^5C zDzPs}Hn^%8t{JLCuVU9PeA6Vk0bj{)&8zdw_0jOo$(!iju!mA?-FZUT8)gkVWf9Db zH6s=}=7brF+zfGkLvM`=ym@y?wGCUNs|MDnfnrZnz4Ta~+_Ru5qSK4>J*0p$*{UF)laYl7LWSburPJ;d% zf4g!goy)_BO!0yVrui)bsb?6diJ*)dr1%n1K!6@nT|c;8OmbhC_+h;P04pfAoFXMu zWIn%_zR1Kx98@GRxM#|u&pBj(R!lq;Ai#>I3kwVs7jM?Z7!$pah@1wr;dToW1ucxUiGXU{XH*#rbBq(7HoJtG$iz0vtHgt$ zN9V!ZXQ@!E?uh>Zuq&~1QK11?euGXG3p+uj>>zNr)!16EFC62Shd@so!oxF`A{{GB z)v1VguvQ>*^0PngFPGH9M>rJGDR*zL6c&+ zZ=R?Jcy?dO(u38Qpx>#=0R{GMahLqoTuxk)8-n6^y{_v(GbV`vCSbcS^7kOj z2U=Cos$PcT6?jXz+gV6JJhE#6#Np>wSvIf$x5{n@;mD}xj_$gy@Bu)GQN}bHt00<{ z* zis+cHN7yevTTnoXY#3o@4G>mCu~0*Ih>9#z^w7xwj_nc25iz1=Ajsb_qhp4)lm&NB zM+G%d>4K+@$v*R}bW9B|4k&@;CXGHO>Gf7W5LZJi@B*aKZvb-y%2smVZrxxzT>71m z*oe=z38O?;J+H#G7*Q6jjustGOC4*Ms|kYkQJsHku?(slMI3#0`Y4!SBmCP|jT;u7uK z?r{@ZF7+{>2hhdC0(H~~IU-nE2oWCaLy|+^nx`&NXp~1uZy2s)S*kqAL0?#@21&(2 zd|t!sq7MtB!jqU>738ky_7M0DNC#>J(WXzug@R-mCgySdgAO{*>A~jU)LdwpSbHXR z+6>fjg4Q&jKm24!u7}pix4(ClY4+g&_l>*loBZ^t3Wp3YFd} zFfizfKvkaZpihw?dSqAW*eI`-trnm(8ul|mpsD-kuNK{g?Yb4#ritOp*q}@tN2Z=E zDt+U3@9PnB=^s|kC;j$hqXCO*#dLPrbTUFk)aOO&aqu^h8Yu-^or$WQxc!iL9 zXmH9_;WhmCG&w}eM(nB;E9caaI9{b7-oq zyxN4@m1jhlPH%}g=6{7WV1m#DOx_HcG9;5b&ke?9uNVI8 zU-fAD*F)RCO)l9mS_Tc!@;SwJQba>VBF6%^>Kdlhd&p&J9}~;QNs!!?s>C&HmkPH^ z=?;UCf!Bt8a#c-!70_#^BosO8u!o1|b2 zuX8@8N`)Rhy5^y}mMm*us$@~eMvWyK8(nS(EyW;p54pf=6XN}c9^YRTi1#9QUi()z z%Evb3z^Jq-X@tpx&3OHrz~^yr$q8)J{KrJBt^}~YSD7H7}Djsc0oRW%ggFS{symYa|*>d^g-VmZ|tsd z(09lQvLT6H?Vz=UPbCYm6URfRE7BIex+?Gl>4%=RJ6sVYc$_SFWQSa$tqYHW@MIzR zT&(VvmxOLzxGVyT!f=F4=OW%s*BtuNBF?_n)M08I;5_9NH%!H+52YG+w6)n}tBD@9 zDQ<;y0*iRjkfhfG()l`e93Cj5!l|h3$IG|`7J4MsS?KV62dSvn8aY} zJ;iOzqgDJPWVMSm);x>gkHpiPMp7f6aMFdPAd8B@BX0v z(>w3};*b9+T}83WDH8P*dp^&@B#Hc<+}T^*8~&2(cAI zsbWxC>G&|b+Z~TuA~5){Tic`|N^~e}yF-)9qmVS$=Dp|EtU*tJ29Z5~#N>PWZyAn+XPK})h%-VY^P{h^f}agyu7EfF}%u+Ley z8eW??F=%i2Ik!PzOzQq6_?KQ0R|@J#p`yvwSU4B*{{>uyKJ&PUUWG!ahYYHr||lF9<@ROd$-E65K+nY4zTPXC;|F zE8W(Fk0?H)`{@C3nN(dIR?A$T^%?(>-8O$Lc8&>N5qvDPm_gxF-AF0QJ$`E=N`Mlz zV`h>QQvArpjR%$QtzOc=p> zQ4NHnb_efr+Rx8rH+y2u2N*(^V-kISwrAEUf}SDB&_F*{^$AksI@(1761KS8w@Pot zXD+Hw@eK_xS$1VDD?VQGUA-$cz3*3d$T1te6GhHak-5Bd0jeONK0*?h zpe|li@CWMUg5wZeM4bc_V$pU9$x`ibMU#bM%pc@aA3=(w(qqVZui^}Fju1%d77$#} zFvt|a)CdqCb>S&(J|$67u@LkN=R<@FTX=#gXmn{4c7q&pt1Ky?gHbOJ=+(xlkhq4N z0N{zB)I$q<3c~7V+#4ZwWQI>?Iiwik=INiFjVjab*EqiS^sF?c;CaUtbgfc-&_#}d z;>Bu^w>m4iwCIc~o4;y4CQ*_YOwv>a=7(0$>)G9Y56I|Mo%mRkXhlFS4`ntUkgoY9 zbgy7ZW#C3~J>&r?nwL56kYB25pI4#yzI(fHSkNoT55+vvAoO*+?|z;vgR+uN6_%3} zyJo90_&3Np=QFBw!Fk!}FYO`iLNmZLW?H8ZS7Uzbvky7piaJ!0Q|76+SHtY7H6)c= z%GY*tLgcW4bt|M;2zl(LB4hj$SdAh*I90XUzEPgf(<#WrL0MxJR5G@E+z_LVOd*}^ z)!=XfvXYo#Nn%jzAx@Gp`-&K^q$;<0pAXbs=Z2bR&!7H|$-~R+U&ZOhiUpr9c-~dL z^F-xtnBWBmWyrZd^mF$#kl0_(MJqyZ5(5K zu2k$y1Vl06Q0Z7OyJYsxg}ER$1=$Sb4D4_Ph7o8>-I$~9W-fx?j8xik#ZGo=z+*mU z9=j>l0g2J5rZJOS?3tv{FEnYh{_FQ7X*?ws8{RXO7%V4Q6uXlmX&7V76KP%D8&Ck$ zDqG>a+_^hwIOtvwMil5q!6*&}ZZMt*%~_=-aDB7SRv;OzbKmMVVs};oqCe!SWW;W(N_Ujz zi-J!ZyuNZvpIzrvPj7Z#>4tJrX#seymu}%*p0&?!#7@JkbXz;OS$@H-1o~31bnBDi z?`oDe&pOEKrPI9ogxOpxMJ5NiG3a~dp#jpL`CW|!Ie!zr!L+RzXQMr>e?!C%p7+)h(a@)+Z&4w1J>S3|@m>bUae;}F z43Z9+iy_j-j|#|x!k(2cZ{W8?+!Zv)qe*gj-=dP)EfH5$@dC_kq$;yRDi*Ghw>h^& znrqOm;&YV;w{-#2HGRKYd6B9c*AU(}& zJn^eIo!|r{>aVYzJ*m>$o0E|rZz9Dn%tF>+V4Z3y7BpcFQ;{9OtHnlaWzvnWeKewI zi71vF_Pw)s!#AsGP(Ot@V3cSe6mw4-zSZ($5KWXOdbLC}gH&NJz45hFWv%3bv`2ht z(M?y~-9?JCkk3lwB`(||X%zN2G${7*8k}kb@q#iCN`(%QMGjG-G)X48K_+e0tOYqv z_!{SCMV;Mmj0Ng!mgxgUU1MrC5-2u~BI~F~ZBN@~9(2A7ISBN`(W>}K&)GPy>lvra zo8@N=Vuj0`49oO?{m~C%9Q1~z{?+P_$qH_U#fH-xN=CilV2p zk0|mGJ9IX>=vqbW@ynN~_s(n-qCyJx=EN@3P>>E7z_~{Ob>zLAR9xnjggzkI^nM>C z6Q7YbL8w1rW~ryHHS2ZPsBlbhXVDc4D;=}pfAhRjl=vi8gJJCN5%gI@EMdbIyc_St zd!Sa`NsSM5fRGEiI(Smo-?fhc=Bjg*mrivRe;Jf1!XMU6$57b$4gU?g0U^5zEno9N z1V+OkOCRHlcxRMj422zMYyFVm-wS1~Hp`hX4Y=|4S_7AMylX7A`hqYlObD}P2X7fO$ldB^;3Kzw_!Kl~}xI9SXD7&e@h zG!aJr-2M9OtI{$u`ZI~a&|V|-mDos*yKQz~;k6ZriTlI*>8)-}Q1{ir?*nRl8BpYR zJD2%b8#G*SVz9yvg+dYo?KYA%wIUsBhZwhP@H^`Cr zn@J9ZHTj((@q#w8H$2v1Ltslpf@DR2?#=<>W!{M5j1;2BlF9|C$}I0R?`Ei_X=F3m zWr9-gmWV;p6jWy_$tIVRS__qbyUDD|N zlI;%LUG_^_B6{hvH?B#WBYIpLgpDA-dM*O1tc#$^8vn+<3kHK)BJgjIcDS$TQ=D}x z3%lWxJQH>6?z*jX>yoe2%1>h|V#E&ru4Dd3$vp^2>i)hP%zlt;eK9;A>W7J<-Clb4 zJO9x?m6P-rW)BbQbz}DM`zRKqJn|rQ!K?Mpp-V`os@yG!U#ZOD?ecD?vuGgVC|CNo z0RaP5G~Y)lpsGMrrfP7A<0Z{d_wln-xjyA?cz>$$j^81lcDOT%S?yA`D9X_a`3wu# zIkb3|G;!LoO8>HCyPF;|KR-L|97!H8fo{W|&jABqVT)g=G)_-n^NS0l6Iph9NdR2NyxZMw`A{M`Y>=#&MahbPOmgQTq zu+*oXKKX+SZzX=$I_n<1{K-PX;#klRXnN`M{-W1zGBH9MhLMRFVIDMK>J~qeZu+}@ z-?<4bnh-D7U2n?&^8We7aXhO*WBjQ#A28!TZ>_Tl18$!hcfGb*@5e+R>vX|dQl(rI zelnoqm6U*5P$MxR)+~Tog7)Y7ap_+3-H=g2DW`|JpIuA6aYK zwd0yL7UP6wsCheoUurk%pnUb)4@nd^2gQboT&V`@P&~!PQe-U^iP5$S&p6&4=)MNc z7|Y#agUwT)wDhc*1t3fRQZvTujQw@x3_T>eeskp`vc`r(#^76w@j|yzESR}XR3rw9 zu}CFXs#WO5U~ir%pBhz^z|gTNITAD8&_r%AYqlR-ii!M&i9N%~Tui_DF!;yD>|dK* zSWGZ|+i8S@Ha2^M+y;fo!^%<+##W<%HWGe;&{@StNs}ag3`FnEDn6RnyC3@$H+<-T zGWL0E9dp9Rv=bej9tDi>qx{N{WLcqNfQPl=yF)4kx$G)v?1{mQ zjSda+-62qGE?qt=RXIp@vdt00-@CSW#4hQZ*TH>BRYr-HiE=!!rASp^weVboRUDUz z0PL|nvZCYCDp6pGzS!Ji-5E5)#>hF%?by#zLZ1jw=6{CmN#E@z9{NWc&{PS)>I_BU8xYQfAGbrSU43sQd zx1?YVqg7}8$yqZW15Z9{%`>JkFrEc@_xAPL=N%Wf1)Bo*gTPc7*%TN>brBTVN7mWF zS;g+aIJ4%<9%pQ`aWT%b);z@x7w)U>ZWQVFFz=Pj_%m5=54q5>j^B2Q-G*LvWSbEC z$APjKPovm2dk^oSTeYH1c1(s{El`>ugY4&}1aEn{-+NdvWs_m{Ea=U$HQ)T$u{q6x z?QUXrX{sI=w;1C>bF<;?4b&@++1_lT*m#P>Li|;?;W=eCRb#Jp0%gdnM z^HeW9=4W-s0&gmg6>rSAWGqk41r;_-@i!r6{=tY{yOuS5z^O%n;^QC`h4SRdZe{G! z4@c}ifn^D6m(v0uEdyeB;$zU@64hz*pcO})1Y1a>kYSnMfq;UUpYXM0>}DQTOhc;) z2}(upm#Gs%b#jey@azW1Q>0CJ&#^BwhRT?YC~T+WW|qv3%cDgGh2YLm8YRHHpAq)ho7okUujbc|jYM3HkokOY;*$d-swYgf-Q9swhKcB$w z&!UgbpRQk_R;&&5v5(O}k$I}}g1A$$(*F_2n_drFCb}Ei$8QZR4uGP~ z{}V?M{q~RV{CDIp{_CB{KmX$0Km1pulxFP+MUAtkuwAdk`PUiY#!l^X6)rL1f$vh? znuo@`9b}+znd#$e8)YyWX2_)nR+5a_*Mrf*q(O7S|MS32dL-Y>FCY5bQtj1MZ6kZ7IY$s0VpP!d}1c8YB9I!o?CYYkL$#c`kt4NB;>W1V8N z=p%4ykQDJ+**x*G=pLHPh+5@p?CyeH_3-aVRrU~@CmZ&W zO|TQNsZO$7eOh_b6{l&4KM)!X>Id2=9@e}fC%Q$6MY7;YpCW6lUK6DCT+DHKj9u9O zw=Pf1A2)np?C_ejW#g>NHoWIH;T{(`7CEZ*ioHjn&u%yUVZOBrp`SLl=3(V&BUtC! zjhX-4VjP8e(qh=IKSoT%T2B7p8d&v~h}s2MM0j_`r>a64`U5Dg&zLgrH}{q= znt1#Nm^d)KHR|Y7?U=9M2L1B01qGzY9<;H?Zi8wl7Pvo$K;XnDL$aU$-c_c;`G1m0 zv{#0tQTgAxlg$z5K-u`LTZ7QxDuy$v==+LNY4d`XhU;oYj>3s8lk|K-0e+llJ+*F!5EyfnyoX;j+ zxOt~dPs=zK*>=B~zOM?S7(IrzB=}z>n{AjeTWSFAY>LgKNIDgX;W3D!f$+d!P@I2k z#LXb|L^>5aWl2npAj26ysndWJnI;*OReE&Owca;_pj`zSBx=xZf_8u(Ob(+@N6e%U z{PdwGb(Ir>r{yk5IAP4Tx7jG13I2S)hep}L7K$4jy6Fyn6(His0&C8^$>}i%5+)zE z;ss8Sn7%aO>xNmJnMmy%^) zf-rwG?~ZVfw9#S36xU^$gAD!JWAQu>a zHp;64H#p9qFF`2#ZCx)0Buh4c{i9L=LSgX z^w4dS>Zr^CcjmZOhW9)PC$rC-ihWZ;>5(+-JbZ?X*l_Y^y8$-hD0Ur1Vsy!$R#~nT z$ZR=I7u*V@RwxNK%_$}GaZLzW4%^5jpU{LZY7OG3KA*tz?>3olWT!FpM z2Ig&3)zg<0t@53+4p85zh5$>ckoJ1zwJ0CPd%j1q0f&4SFXn5qJ*uvg;S$CwN>(4#QC| zspiE2_YYkY?CN=(t^!>`9I=P46?Dn51C|AWG$v`2lsn7KpMw_CgE8V zZ7!5_hT&PGS?>A||Gi_;kM-bS*7?3g&b~0<=r#byHHy7Vk&CFNo)E1o%#Pk)`9 zuX4Zz=dmaAz&W*N-uLvD=hpn1FUg7V}{fiWPo+6)7kvge$l;P7b4`^-I zUTtFzlIycfaLo%$2cu4w<@3+GtX|Nu;5-OG#yH?f4b!opo6&e*^HL)pI*GZjC?&4NW0`*Z3r6^najAh2GnHm44N#bbyK`n6h7fU|UIO{FZ zFRzd#B*BI=MSBgbP6ow7HZesfzLmtBfLQy|M%eDA3lhCjA;8}%dl;@pVXm$2ZNkIz zvw+(6Fnpim(&ck50xME`yY=Q8qq|I)vO)wyl zn_c9pm1Jt>q2>XSv{61*R|%wUR(9~S={!jeIKWdT&omKj(5FqS#jbI~$IA(c$K-k& za`Nr(T_qcB*oN#gupv7s78*}&$4;!!RUz*GXYWnmno7_8agTUH@?ywFFv)-l1TcUz zEQX3WI7_$N+urFi?Z5xt{(ik}rHi*S?X*qpHXU4WXH!r?4WKN7fGlpPtZs~s3JfYE zu1LfgMFkWYM&S26NmLR!V-6(g8QYnUaPsDyH{rbB^SsaceI#JwM4-qhU;*h z+hp~cDuFJSUPj^t@q#=$nLw+KWu#DY0LZ0voe`@&Y*i)3x;kSPA`ajY9{D%00?EYi z#PsX`TBeYJ?vk&R@0oK-S}4+C+BMRwwDJ}_Dg$wDtdR|hMy|4>kK4unw>;V`JOY-i z-a+=UbHZ>~T?LzHgo)fhsg6*jo{BxsGfcfCX&0U&`Mk@@H07Xog`%0pmb-O=TslK` z0*WXSC7mFP@}OwcAna7MDL;jVh+9I|PhG@2u0280wXHFIJcA5p>Qr0}?R(KMy($#> z_wkipT^@2=c5l{R879lL(}!Q`Azd_f3&M<0z51w6dC0gFA2v{WWYpODdpvOa>ATZO zm>EjhemUz&35WNg4=lh@O{or0WFHlaWsqsgyC>x-Th7$)%`!v!0g2;&urg|;!VmP|IQdJ^vDu)7&~5e(dB^` zu%1~_@6fy*a#& zcQLdP$WK>$;AXf|emLCZ5N@aOxJR%8?DRELhLr0jCkPXSpUUE+3q^xrtdVJd42~|e z*y8ag^QS7k%&2Pl>7E57ox@@Y;?g7btO81vN0Dt*Y=7i2$qvzjFAce*&K`0(F}up` z!0g1(rONIX*MPMR#51yPpgGtaeUvQa<&uLyQd%3;$~zu3;92E{ZFi3aou(g3SuZJz zTRpDTXYn~~Ph}^8-r5>US@-E+*m z<@Nw9RQwEh8IW;oSXm;*OOUn!v$K$!5-)$>fhCLMf|dh~FwMY*zqy$O%U^Y8|3Fr7 zSg?Q`86j9UQmQnHtfgYF`t-~8!v=H_3ZStL6}KZ?=L)(iwAlA1Wav7%5P;2Mvy}zo zIiHmrnV4Vq$CMY$&|YChS>PLttq=xL7ykhW7Qs2|u$#pi?i zDGvc*f;_jwE;ytIl^9yQ`s6|ddUa@~#K`lL3?(Of`ANa0Zf&xjs0~q$bmg!C)+ma{ zc(C)i*ZykJl|NZKW;pBJWr#YPi`g4pw@KS+WT;#$+TvTvH`ccw1X3ACDjrYYTox<+ z^u-PwVI@K)$d3N?&GDp8ao91#kUC}R`&C6E~o#E#X}U;dJ3(HA{zXrOChFUY6YhnIt|t$0d>R#)ZLsvVk@1GJ@^ z)xAa*^Z}11c{`2i+=uw!&~*+w7`s(}v{_g!kpo_`6CnTi!PgtDdoVmFwhTini4S;J z!_t&l?iKtC;tN2Ve<5hVqe`&CO?O-t7p+5+1tI}^ysz_%WRBD{ELNU9mMkuJ*o$KK zjYocO)=-tXN)fp;i6l%MX(J3%s)rQ$l#0#tKlMVL_d8wQXQ8Y0?@g~-V-Azu0o`=U zlsfO$m~*@yc$gc7H>6deH{9%&H3Q5#Sd(pfQad6sAn+pqGTiODeouW^jpCSP>R0XrLF-sonzz}sbXvQ$lXs-nzB zG&xKTjWY2kA4I?KMJH5flc9-=-V|f#xlNYr+9q~x#MyLbZCm6~R>8qTtc2LaUwynK zf09{N#2DVukwudTB#w;mXr)l9M2f5iM}%LAc$>70?&a_Dx#y2L^T)O4{G9q=nQxs< zx8A|Gd9>+S-FohNqTYy9-~Eai6A!-nwV#suFIW=y6$=12Q!1dS`hbePBwZJZy&Et? zE`{hJ9Zq+Fq!8?+epp(mE&;_>w?8mV89!qk6oz183sOB6iU!<=U67*v(MSsfvXSrc z5yNsjy~}lV5Smaeyo9Jt(P>H}6J?#C3kVf(UL*5lm0JNX3%qsZf$KsS2O6g1w#`Df3?d^gudeh~s=?4Qd=?~Np(t$LQ z^|L>5&!?}uB?ni|NfI36Ef!&sQ5U@c{$sf}jQPg0UhOe|61xjo3AqXX@#{B;b*2GB zh&~tM?d`&W=!(cxQs>hf-5c4WZiF;L6Q2^K&Nqvj>3&FwZWrciks{XU>cZ%}4pZ)j zT}(X&9U01Hy(ZS1J-b_N^VzJ{guCLo(tAdzbqW(hrckdEtBLmq9?)X16x~&lrd$`= z9%gz_3nduqA9le}tKVIA8=eVkaadF|Pr$OJeiI9#ofOQ65&QFDUWNObF~Jj97fcH*JifnJS`YIK;QxO^l&_`PBfkLG#<4hm*)Qb_NZ1O*LSi94RX*C>0c*6;rWUf;7=b$}4XyJUvhM z;VTP^=Dl-i=Gw^LKQ(-e#__6o1JS0pFjRex{s@S|j?wWzkD4hd3U7`1ShG>}PW}(q zf4y^F-D`WlpZWUlTVwwI=XXmz?ILq*!xW$)x)ftzT~W+KN{bslOsyTw-XK{IKAbW*I=?1M#IO z)WH)&Q-O}9LenVE0K2K09{l?7w=afvM`mCXBIppKSHJ)I{K7YS=FK1ak1OB2Ez9yn zMIQMS?1CB7BH0%?j~9O22Z#Toa5im5PTzO^PLRbM&X~=zz`|Nel|+#>ShBJIMKG~) zwdYm6N>n@UyF&qn8j9cTxPZ3ZX;&|*o){$~{;)KHr zD-`*R5GVDNs)iy5sn~VOMIto1474sU90}JUJLLq)_Qjmon%Rkh3h^-X)HHeS+krkX zGX#A0>SO`7dLIk}|2~@bx8$Ht z)EZ-aS8Bejb7 z;2JT8hOy7ouuHA%1}`tXS!`;w3XP?1@alLg<;g&gaEvbWx#R@{%^*`3Ljp_9?9JiE z_Y);?A%)V*N@(B#ITd%IP(4|&mwyOo16z1Hyr_r==M{k#^r>vINDthm1rfSh;W4O@ zPxr46%BPP(4dVb1m)_uEDov_i7kz`5A6zA9ruXu1KwUgYb5SQy7S@nj1%6v(ST(gr zT`F&)+l4)JXV?v1BlIq5Q(~j~;jlyjHiqSFQt*!&E+s~o0VsBzw~dst10Z*0GC5@- z=V~cc6$SkNvB;O(EIzJC@>|No_PJ@w4ZO?J_pT|2T}s_LwQ;iwMMH{Q>3V*ZpeGO+ zIdpBxL_hcev~(@JT2l?ZU?4h@u1`p zc~iTv5%z-2z)nTw9L!=mCaI(ED@kD5(xTuBh?NpLQGCz|YT=Y*JVG&UX-Io-Er~(eq>m z#~fvSE$a9GTKou@msJVzAdCqWzS=l1W6D{`?Z#J0ubPfgN&RazeN8#wUNwCi|E{o$ z&XK-9`+(+@bPu?A4G{?{N8)9yP%(-c$DqVG>y#V+(2S>ymBZlyhBDT4)kEo0_m73T ziv(bvs&0} zHXJs4GA;C$$MsG5be$lF>g9LIbz4=2>D%dQu!8%DBSizl`q+jEhKn6I{MtXlPw`Y30?vBzR#c+9|s4iD5a^GS*5kT)a-2_n(}}m+auL)g`1t_)HoqF0&(^P z*n|n;KPzgprZj-&DI>yn$&@OABCDv_y|Tm8*LvOzgMR1vbSIS36+=HZJKrXo{y3ZN zrA;&0_at_YE8(gbum0XQKYGh7M_w&HGL?M7PL6Qca#>*!F&v;&4=8evip2&VnD$og z1&ME1dej+)p*PHI8+Jhs{WN6_xhzHE29P8zye9dzuwyi^s4F1v0*m1z8|l5!i}j*- zk7An#oRTPk#)mKio`S(W$e5dQey$GZPv_xu9m+l6dxOaXv8xQWfYw8wnX6iz>C55WvioG%A(^8a1pOV({n=>!FtmaE?p~ek_*9 z=La8x{=w!ZHW@0t2Oj21_lk6#P%z8cXrcE*LzJ~VVq+q4coSm~s|NZ6(L<-rAq576 zcli%^I<*Ju6A@^OSbcJ~^N!|!Z|2&GW_h%9dB_n6baGa9k}Vc{XA-4aLy?u(U$s$j zClZ3WO~S2y@T)?6!I|_6j~C09HFhFfn%*mr- z_e7mnZIEsD-{5u~+Gi)fZ2cPR2M1^6y`21Vo@V|LS%Z3os?aC-dOOGei$6=F*A=HQV4w@#|m?!xKF!?vcR0%f9ZbI&15f56y zw?bVR8W@G5t!u5+uUzN;L#ay)?c5U>31dudOQhF zPFWICJ-5!gQ5YxNP4Xl0O5{l1tteAm38@etqwh_70BQbD<~?UftR8dK=A zny1&E($s4p<&DPAx((j~+QXCEUR}`au7Y18Hl1}R?6y}if5@eY{zN+Da@*?@X=}{M zkS&lFcLT!F+5E+#-DK6ACPgOQslXZ}oU4gW;;#|q(Ou#@>T|qAD4?zt9tkR)dsewtX3_)EX!rDxM)@c1Ml?qVMpBu9q0(kT@LW7a`YvvfZ&q3iBN-h*s~vUvmUQ2&0I z3~L_0KB6FH3WQ!!n;O6OB=X`H4(mH8{4K_GzKr^Jc7$O#h8|acLMT zsYBT`b^p{MDC_7UHNHm)0>^34U|XU1^ddTHWY4cJ6$hKQmcRb3;~?42;jQJUg(%oZ zsdiJO9AknRviC))kUB7=*ys9*`*SSs6C7i4*nldZSxo+J*PXL$@k^ zF1?z!F0d@H!d^|gpI-j)g6ruS?K4vQYSf!zv!q$M>B=uj0y{r0haGcJ3^pPQXA7mu zpvVR)7RkG|c(n@?$&HBkpiUYbFegZH#3}7VDf;T5hs&d!TkOJJC~}T-ZY=n)4+1`a zmPN1p{F=gykIZSm`7KG~@J<9Q2_sfKmr?=sPA1l~e@1r;GW`emiO_xU7^If;!ERGc zUw}~0L7yUCV(8{6_rubZMKA3M&I5P-QL>xACm0>~J3T65?7GK{a^2Qx!l)3CV z#A@52RG;qI_^(y1u^iG&3Pq=B-ChkO4|yg-vkl11b#hN-=1X0-zLeU)faPk2Pv_TCs~vVl2bCMSmX=>)^<8)ZXJ-a!~TICTH1vM`o46zC#ta6 znn?HP49J!UMtC?{j=+-G2m~vA)uy!^*GbHxBro@!0_$uwuBLSohWxV+LI1=uVC4D3 zI#8rKal&Fhy&|?O|I=4FeuR1XM=98taAapvf^}{Z8^;KT83EkdI%2N{z4YnU_Qjxg^8C6!@14|I*># zX84yy9VOXRX5fH(t`}a_8ncssSz123FLJd1KHg{KmU4q^TX?EuTU57Ty>Bz^_#Ss- zK!ZtG*k#R`Ct+4veh|~(5pCX!0+y`aLH2RjC34nccWIzhz=K~;#n$-@dK>!{=h4X| zBe0uBavMxa0wWe3WJ8LSs6x*S+4At^;kp8C#uVepo5Ih`F{&zS+|JB7N1vHf5|N`t zCN7;`i~Kk(>K3)$FWvhX()0zT`avnsVJUj{J7%TuLCb*}XJc)TL1u`^)mGDGyj8Qu znD8`a*nI9zWBDdCIGEpIhRrCt?!=tB)>x4lO7B!m{xey_POaf^fEcCx+JZ-X5y3dUf zOw1^-(>ve%!~JE}d67@=NE|kJ7_uLigk*u8lPfYPkAcnHF2oYRtZ*zUagGKua(QfA z`;nJ8<(OHJtf(!13Kh?ZCd+`AM9CH50gWke(*ZpxPfg8+%;{@k!(so%%Ofzlrpn&6 zF7ji83Jz~23?+dV#i@SVJlbS?WxDdfWU^YhSMGc`!0oq1f;?I!Xa1fPTFAFWFXz`XST?Ef2>`MdyTOAFr7$3vJdQA7A5)1Lmz_G@Y~b#!RJk z#GkXUV2I%(7qAXBhl%r=SpXMN|%hS8#S9Cm|1`q>D}BcD=1wthAhi;a|9W3Y)X@+V>6 zggDW|;7s~F5Jlwi3rM>#o|N-4c$whz7*e#;XxAEK$%6CzN9Q>fM(ok^*jja{kYPs8 z6u+-lSyx7IR+2E3Y#=pjCjFroIy`ihP~5J|1X81B@ismhB`w}9-i}>$tEFQD0;@|M z`jQz0qiit~ioWzp={&Ot`PPx|my&}V_K=;o5Ema%s`n{!l#14uaDUW8&-;o}O#!b0Xk!+OE=fQ3ZjE6&rwS^fOy|UVXSgrr zB|%#3&gkRHrGmuJ-2unwcCr)l5wA!SLl;BMRsAcMrKYoS{NUKL64}7%u{2?EFjma5 zf)mA)@9O&3$}KX`70}lYC?NNEC-WM?FOn6GWJgXFMW(g24MAhY!Pu{xAA5DOH6}Q_ z_b~ANJ|Yg+8QTB?wVsLO_1~-f7#6JL#6)rOQGvA!9CVAw%`lRo`wj&{Ify+^C6&`pPl;6_H?YhF$A z0%(1e5?M{J2wLf>SMQZM?fJCjevjO&opsNd@AoLmAa|n^Et9`8>6>OP_2XG}cgXvr z8M+)67o8TG>qnI8EJZ$q!mC+pywjl4?(nN={3>;_@~*3?fC?L^r76?GYk&!0wZ{=- z8zoZ`>l#lyQ!hL2pD+c@p-f4yaKQpo+cY$T@LgfTEk;qCwz5T7cRF6y0`U=3E|*wVzj&l@VLD!!DSGgsM;`-K*4n zDvOVH7%A@faby!LSf8{{Eq!eI+$ZJyk4Cp zTOJ5p&}qtE=)83>C_@G$0+1Lo?6O0&Rh1u%BbhWXAmwTeG@iCeeL#yC$fv7Ap{XRa zdaO`DgBf$Mf z_-B1~h-&F-F&fzD21W|lVV6e5y1*Txa<68(K&m^$&!>C&jjok*paFMOr&iZa;}CWL zG!Ds65gii1=4hc6hY^(qka@=y(A@{VZQa~N-hzFREur|eo)3zX2`z)I4}Hz}Tl&t_ zH6)XrwBoQc7#d2Ba0c(BR7Dgiz#>3Qd1{S0KQ~2nLtYeF4X)mF{|0)4;E8hIU6{V6sWxCEht@P^w$R8-$kK9m<5LP0@Jr)N^_nb*X4UbQha3L@(_so<@0TIeT~+pz zb)FZ3SCB=b=VTZA(!!13wsr}!$$;Uo&duOJY{fdm+Ms4}3$KHRjX2Vbr9-wERpS!^ zR#)4$Pgz~$gqQgnqN*ikiIgo|f0ZOog1#psRRTSw+C-6!RBU=sqp&cbTKs?{^6F%k zuUcc4@OB408iV(J>YU{UvhB`}v~rzrTj3*^cY#ryK-^Uj2g)xsS(gPd)YEK-s2ubl zFPe8>(HgVpg(5+x@Q5Ts)*5q5x+DN8p&R9c0q5v?NIL1J7n5SQ%TlL2iR=;LoC3b* z1R=YW$}85?#SEOk>y(Q`2B`nm6)1D*dd(Uj*}NUrMgI*nN<7Y|SIUd|T@kt_PeX*W zdQ^-79ZnG-tkA(-L*mEZ`R$Lb9etdQLo>+J$kiT!o^rE5N|D-gl#H8#%{2MrR2& zy>x!=kbjoC9!%iHu7>HUe&t@5r8h`R%p%b}_iQL9!!okCS1!P{x8-GuB);S1-56l( zls%CZjH!O@Uli7K2MnwNhxi8BP0x$scGyn99>rwme@X&?P3g^!aWZ_q9Ab6nxvSg% z;e5nF>m)ihRttAkE5eX+S4=ucF1uewAml=OvF09kq3j&8J|a+<471M_&|J9urAa9!m(X=CA^m29@^o>cniq`u#Sy>@WIF&y?r4rb7QPihkEb0f zcV#YNXorfu1Ca+XL*9j%z%4u^bk8AGp7()lBcHw#p{w`5C_d_HkZqolFjLn7S`xAd z>eb6UE{d;;ugM{K_ye^mXKygC1?-3&fl!=1?3A%OLTekGSUvF92Cc6;0R*^g^~kW z7{EIA%s>=P)pW1_K~EUS(V{0VPK18jVV4F7Yql6?8FuM+ZvyVyk6+TQamx$rP!1~3 z%mL20bpK9mesHSafIG0s?GHpb)hk>zTd!UvE8`6)6T*!jQpVfQ)@ZaRNS-Bc*no(Y z>NX*1{x6P3nj!VQy*si<1&1NkWC5u|lnO*m6%~7hzCoIz+J#L~t0T{f`=p;}KM@x2 zszNtwmb}sx*cy{?bmH)~RiBTZ2a!u#4}LAGx`Z=)#ndh#vrVm zgJY?@VnHT7XdKn6VSXrZSthy+`Rc`j92!FtnReiGaS)2w0A|#4S)qvX7XRWW)_OK) z4Rr?18RN`LqYryuF*3Pye7(`gpHD@DG*a5@u!%1RpKGBX zZJ<=C6iK0Cu|rl3XuJ^5J_rqnhBasDi{cft^y<5AMZ7IhdErj>eX_xk!$332#G=hR zS6CM`a8_qXd_PMU}-7JE7{7G@5#gnT-Gu*qvmQAXqyTU58*JI}I z^ZZi*RV>*-{V64K@01mgwz1y5ByvOa0EDSlNA^Zunz_c^s``ci^JCSDhq}7A#@#}NFO$M=v6*y z+R+Mq7opoOrlRN52LrJIW0}`3AIP7p4&9+Ci6|5S-yAUaOVUF6Wy3CjGGpR;7xc~G6e57TrB{irKb*@{JF`P`n5>F9O!mL1%L;FlU3aSuD)8@@wR@yQrYTp6iezwA z3(r*OfXu4-bg9pf0^P~OE|3ThzTbQPDBeu&!ri1+2vufHa-_8BP*xk8Utx}*Eo#TuzX>m7E}^{4@N4<$o&_X*v?44H z+gN)o6q*7`l}C|nRO}7usi~>6m(9B8UaHw9JrGd98{jXAKHzymg@sUg-~;?vu}zu` zk#-!z(*GX%^y@~u>FYlNs7pZFRvk)02j@pEQ&+sU9uYH?TDVv^5c zQ3X7!BMJ}pQmS1PDWhUD=|aA#?z<>*0W#VZMPjP%0qB~Dkq0xKDn9)E>e2VUw*8IcUj?^`6jKm)NDpY*h3i~b z1a&BoylADm)$g!(o!@c4E6|a+H|#1z-gh}ps*D#cb^^y!)38DdbwFu2V~rLzDH$9# zzZo*Xwy1j}vGxta=@&>E)F5=x$f>ATZ+6}A^7UXH*1w-nl?7s}`gR&?f8#xG5a?VV zA5^8%V;tr5+7T8Vtj9kd^V#_FkMows_V?Xc1KkzZ{N`tBsD~X=NyAt%4 z7_lprP%0qU+m5WvJEO4zQr8!lFhjRnahGlfpK4|Z<}`G|@#>&q7u1^d-hc=QeH#sT z)OVZYy-M^P*N96n-@#abfJ7CJao2Da0<6T#gx?&IzsolR#wE~YF^T6eV76HRCXG_9 zrAQJLi}l-ebideG=i8zDjII#31B*BM?waH+Y6q&Vqdm2@<6^X{9XG}b7ZWdSdG|$Y zLGqkzI}BoHAUc(=Lq^5|$fQYA-VAAkq%B~$8SqGkMvY+q(#C;aMKVo^^NZa=ynd4{Y~T_RpC zx~aro73nHS)53I{9imRiziIKj^CgFoN0^sy9AL0J07g#8ENk+s{v*@6E|RluVTMY_ zU9jIE*EgsrrovaumvY3jKJu21dTiVJk9w&?2btG@G=up2le>N^H;azr|M|^JWG%Z0 zBZqT4OD%-TR!Rks+Kds#LQ&deqc}-~4!}+F1XWh}ViD%lfg{$GvUg#!fxaBI$gM=2 z2hQUz`l?UAtdm}7{~g5=3ARDX89R#Q_j>ir*~```Q_qP)!jNE92#w*>l=#!=&Bu_y zYLC0ZF0gnL!n4AiRkQbK8CU@0(Irm#y+1?U^)s{dh~Kii zV@B1``sy6-_>@mxeBTv2!0Yab6Cmp!{U6|#m7O3f19V4%(AR+;TFBY&5?6<|dM}lO zhT5UT&IA|5br9uQLblS0g0jF0vF@rAS^*pT0%y6VfU|Ay9FQ%a5%kO@doBX}Jo;3f z>8$*~9%<2x#)uk0jqiSKt`|@dVMxgQW^kQguPjl51QL4n!N3~ded;@ET>&rIH9xpB zY}3@Li29%kMZc^90!F$U?r|baLC*@hz|)P8SS&C!iiENE&=YvF+OH-kgEyYDcKx#v zMH~(wF$4;>dsNQZ70^C!Vf?F?q@9X9es5%J%yuXqIZN(;vvnTSsz5`}S~?S2pmcnr z(fB%7w9RAfH(TJiCFUCVic98QQ{IJO66VU`b$#N8(&euv^8Wm2!VxfevM^&jvMbKU z_#H;OtaZ!Yb=F1_XSH7jgXqJr0v|YaO_>vb`-LU9jCYyT7*8hmImgMxD%#gRBG$CH&8HSL?(oHUfsj3Nb?()=f`z z)PjXw#%og<`|d$6#$qT)+se!JGWFVm4vwfcI~+ofR-LU3d32qlf1mpD?j_b9+vfxv z4F26h;IYtQp<5n!-rx{BuEkC51&2(Q+oJFrJf}gzxU?*A;bUMwj&pCC_28eE z{bGPLO(KwDKO(|%l~O^GSsN9*-nWGQOgQ9H6?H?Jqa1SSn%Su>_8oGmju`Sk;hPzl z$m@}&`t6E|hdwlQky$}S;gF8c*Q;vjjk3W2$fpM`%_=v&$ACv`OtZKV3KWZHELLR& z!Ay!9v$BE;ct=THWV|*DvYGViM^^)HgQ|qSA2sBX6Im^86lVq=hWX3AvfcOl=hBIi z;u%Xrw#IDpxD=2Q_2eqdPU)fx{M!O-S!NcCGs-S7#vzMk82tr1`47EGiQ6AyfB9DI z+kgDYujl_kx{OjSp-7zbay>`?D<-`9zA)RGo0F^G6)N3Eg8wnwWp3uu+F3-JT8an1@}rC)1I zhjRW^n4~6XhtK?Gdbdr`WVoe_!*+o(u) zqk<{{)aq{Ws)i)1e};S5Mg)x=3S+!d^UJhb)6F}THeqEA*}`Fu%>fH3QcS6!mvlaQ zYz)Sv1YM6b4Y)LS&}G0bo+!|(SGbw76OhXi%Bj^&Thz^BY!O(;I}3~=ZL$t!lG_O~>~d=Asi_7bjv;+{9`sevt1-jdL;<3! z<7G*%M3l>x%>pX&9=bp5YM5SK<9jg_i0&_nS1BqYajFLC5x`uNyqd-~30dJ+!N-hK z4JjI3JG{z_msN)z;#V7ww$oVg)Gsptb9$re{jf%b4rkPBi}<(Y7fG4dRvBJ2?6TZ# zso(&xl-H1L{50hX(kQX@1e|k! zy{7yd=5cl?dLs+?>14Y{gQnQCF5;kcJ^%7d%;#*4*`?^0f+t6|NK~%aNV@3v36AXY zS{K^m)A3@3W{*a%juZ7NFU-t#Un9s3Pqx3QZZFUtAkIGtDfZzP01$zb2cX z8v(W7LO>N!DyYBBqhb$e%7D3rD7vR82`pqVa(3QDYs>7x6{h8c@t*JKhCV_p)x&>nCZ*B4;8nEuVl|6|?M zma|qALj&6+??yRLF}~bAIf;K6O7OD;Wk&vo)|gshg2zsNq0a%uX3utUy|hDF3$@s( zP%%^@E~Qr}kI|{5ENYFbLzevU9UsTqKc2%82dp$7>g(%&ar}$kyR?CB3%wL|DJt1D z*|kstT}D6*In)Q-i-cgJqFs0- zvMn^1ZjCwMy2-DXkMH2f`{dJ?digl_3D1MVZILi4yAs|S(@xidOUX1AFKF|+tLzDD zjY;s!0&?WDW9K9p2b`Xzh#21=E7aJ*X{N`w=2&-IXQSM3ID3zwH#;Wk!CnM4O1hh# z9m=J0J@7h0&1JI~3zUi?+rdD~)fNOGXP&Ozt=-MBQpvo`V?onc5yIxWzkjFetJd|8 zoZXlhY8`uo#XeBZiuU;)@BF|Hl@5Qj3$bmSuF*A<-X!Yf<4;|gB;8lHBfLX-DPWOk zkto|!?^hlSSy@f;jk23ziGof$&{d6g_s6pORzhU7X~#XrN|sDWDO#-^Pd*lhW0VZ} zSh;iwovC>UYEC{6B2n5##e*oQJ<(p~pAz2!#4| zE?jsd$dr?Xaq%HVN>rg_Z3Lz)t{KTut&;&tCowuXBEwd$v-3INbC>I;5H17t#cqDQ7GFt1k~XPc*}B>GF^x z;4R~97>#VPu&R?N)f$Sdq++j6h7<==#TwRMx5-v}IHY>=+)O}b!u0T>$7jqI!6C(Va_iWe`E7EuKfUF*8%;qRNrItUlv~YhEtT3 z=K>y->zwW%Si9+@MDUxR{w(%4#;golO( zXtRBhC$q*^cN8)pkUu)jEnk)hZ$BWAOOEz&o2&&=z#SN^G(XH^LC<4ZV(@sTvZw(a{Q5=VY4Ud zoE4F{Md?neYh~E=ShvmNniu4C%37U}sW(RO*b^tC&hZqm=3p1w9R*t8pKl&-+wWKRaJj1rfm$yaJzPvsNoR6r22dKmcv0Z_WI0=02Y) zw>DLlDqoYX)vK>e$(EqZIU~OVLP@W#6`a=$!F#93w*}8(~S{Gj|CC)!KnDKd9eBDAAa|$8AuQL**_-L z99CK`SU}?xr8+^8225}0qBrs?HCd2Xiv10sph#E*+;!-dyA5$caM5C)6bu;V(zP;O zkJoDXf(2NvghOe{&Hl#%A4KW4@zKWz2@N!~q-~%}HQ?AaL?oz+W*DaTDoWjScs8DJ zf^>Ldhx8iXZ5~4|t=gekLoOR;G-z&zA-jlPebndrY~6Cv4wg(+j)Tb4;ISDWa3b?Y zHA?mE1kro*em~O;s8_BQtRaOQ22_Iup!QQL@B;6mV&7dP!V);h4bO)fNvt2pkORr0 ziQ*8M8W05sAlK>$KVAji;ZMX^WqUHDi~a{Tw^%l-Dx!uo!RD0FbN#kiIZ0S_CLFr3Q@JQIWGOSPV6@O-as%Q>=c5)jLcv{bfH zsw@f|9(`30D z@7ZJlkpxP$iXzLQh=8sZ4=KQE(QB6o8r=U$AbBQ6{z+wL%orJ-os(=`0?XOR4nt9^ znf@K4IzZYUvk5}R$YF_GhM7Pt=VUBA8v~C_Qae2SET&+_z$gM>;)1l-4X$Q6aqLUA z{~${_Y|3u6kO`@jDup76RP4IY_OK46kx4rXjGwsZ?C4#tIvmBuM*oI|5j?D5G2x99 zKe=L^8t|N?Ju)N`80gz!Z2>xr68Ff#83IuoCnLx7jN2ksSV#lAo%+WA3BN^~1x4R? z{Z5d@92OMdJ{b}DT}!EwD6)o%?NWXu1FZol5t7^nXVwJmlj_wSyo+xCwuT zc+BrBKend4eLCLD;k-o#Vp}8(OH+cg9QkZBp%d$>+4vJU#;`~b-|bZ%oQ3W;y>`GI zNHxNZg{sioWp-_pLAF@5+5>0_S9=_RQlF~n`E^d@lo00MkJsLA2h1RP z??%G}>wSf@YAC}_(E)5uyFBi>?*a-=x z_hC0O(1t*V??yHOXEb?XI|jydCNpsU@ZUj(FS_-*fvWhyXFoH{`{cE?U#orP_Pm?( zwo-N9t^I1jD@j!UmmN-0YqcV_W5DV{2ft*-z$n6F!XJP6<*$2~mC~PGB=bqiB(l>& z_sgbKdWvj9{@~;n@cB9f99V6#-I}E#cU?0jS>fn8xZ+h!-;1mcg&I3}Vw3zXNh=~d zlo)PzhFrIODMmb#4qgVFp4C!tSH~+!y6ItUxUey0x%-z91~bPPg@=y3(}3z0kAVgd z2`t)VarWmIcm@t_#`ZJjvG?0o^M82ST5zxd1&4FX8023XzfrT8bkkUQzC8A_6iH?4 zNu2=4OX&~fg#qWhmP0jgx;jsL-93li>{m7K0r^*!e|!G-{dMD^sNimL0b!u1Xn`(- zSXs3}vM8uOupppBeCow^;pw1HmA#R=8n0oX*VzEchATt|wADggLojBc;LF~~>$7|L zcuj&zSMSj;tMO_AhC^s1=73gWHj9)IyX9zy_7*9Z0&hgSn01k&Pk;S4631Z&8R$eK z9AwFqDuE)asMrqRH5zcQnY}q2v)9v@plyJs5l=7v0te>(v!>Rj85EF9(=N|Bas!xZM5yQ-8KrMmg*MVo)#F`8Lys$vR(C zLockoG*#>V6BgPQF+u$kV)5+6MC^}`&imI=G}4STe5SVQ=E&`Ky7}8RuVuYn<(BvJ z_E%4;uXzCtR*|gAtx-_{ttn4JEB6Gye60?NCa_^feNds_Qn_gg#~{x(aE+!~JRTeI zV@jeV2~D(uRs*jDfE$U}UC>;|1p=kk+h=o0*v zgk0w@3Dm3ctsY>xMz+sGeBA-fDQTK=ew=8jJWXk;q#JgbpR2tcm=LbR3Wa#@OVV^z zfz*hM9BrDilitbC;_dS3Rbo>?J7L2PV0erQ(}8^e!A{cb-v0KlzUbK=g`&!UtI^#+ zlxgaD+$3xg7K+k?FbZd<#kPHEY=wpGGur1&Rtu7zY9$O=@D-9gx;C^R z=#=X+@N8g9+iLm(@7}DH%1uhs>!kt=*g6p+wt~Z+^1v4OumT6wyxsQ|>zZ!Pf`g%c zJA+sG)jBAL%#gh=YK>Vkdsw{AyHQxc%XYt`u2md^#yhn@<$48(7m;`xm}rDu;#x%= zG{#;Ua6TxJm+xSbGj>R^*hp+y@G+mm4k>32X8qNgJ%o+9%VCEZ1J?+0wjSblLGk}* z^uA~e9_>7;fYNFf~^1P^S7hMf{8fPGiOhiwo~}efPh<==;ni zx-PPVUJh}?&FaPE0rbG=C+J77<1G*GhOusXG0B8VJ*OE5$jZWik`&QsY*j~}l4 z=JN5h2;{J{jG;wfp(y!vwu=0`^UVuv3wy z#Dr#SB(znA7d6QX;0M;o;jD(jfQ3W+-dB8PTj`k#$_(Z);ePhB&=l4giq#NQ_VNKQs&h}hm)(5MoTl{Zk5!qg?pu&Qr)J=O)55# z+=y5kp;y1J-6YDT*H68mEc2@nS4nkCB<;d{`mmzPqh-c|d*V-dSWMFxl^oLHUq1yh zU0Xnwq$!g_E+}>7UKapWTa8#lSMTB&gHYE;ZoP15UJ@UZUeU{PTvi@}6IX{KaYUi0 zU0d$;NI6OeMh`n39PY$;h*yGqmUJ+oxVWwgQ3@p z$6b0U#29O|pYkrvIYl6^c8}i&qH2#L(c48U+!o}}J6$WpPvn_>amXa5Y#x6MWM&)| z{Akz0@o32$-sc#!&v@vI3Y9w@{<;1o5i8}ohtcugrEZ0ozcRCsl% z;J~{W3K|Hl>rf6T>!8ozN8%c1I&feiSYY5;%VyX2nK3Ymp)^qu{mNyHS*iqQzL!Gs zI2@WjY#~ARP^t=wlv1(9q=R7doI9^ymhPN)O}Qjw_|?nOA~@<&rbNLeBf8_gOPa{L z>xIqSu)A{?T{`!MG)p?X7(J`Md_jF3-F2Ws?S? zw|JFINfva6?1rAnExZK3V|0072^^>L>NUN59ItY_4}HP(>W{=t0Mk^z4jvu>0D~}u z)vG{A;(7PO#-X=74F>xG&Au5~KMkC|>|ZN;3jKcvwLK(P(yhE0dOS+6?RPJP9`oR1 z!S1G;q8otIQ?FhuiW7mV)1|x#Rm$h*8r$w(B;A@GvYqacS8KA}dzIUycJ&I5K#d*X zGGaFS9bRztyK4;>hIIJapmU^JoJ#6^dZXXT5>3Z#u!r4`)w(Gg}grY5Tv4ZP_ zp7WoItpnN5iMqm&Ue%yR&S}(o^}r_GF1-BZ%OL3t!Zm^ppTU5Baw%#XG^ZR4aFCQ6 zZ3)>)zR_lI%vjFr$$it!B14<7vW9Hou*f)IAu@_76*NuBr(!X%_<`EwZR_B5fwl>* zdt`VvmUSTO8b+h@>5pI1#S4%}y3jL^PEY|e-(G%PG&b2e7@QLfVNCS6WJRuva)_G2 z4gsTEJ@!>bO^9MAOeP0pEwgS5$k0zOSydv&GG5)0pduO@O5F4;gb4AlOJ{_s2bL2b zKloEw+<(VBY7Dcoy8qnuaqb(O@sm)qAX%B#Q%Z`zU_M%CT`;1`u!>TFqP>@jErd8= z8dMr0_k3`>5QEjZPzSKib1%P8(g|(*pmrT7oYRdW3tMN@8_8VO;x~XIr^0vo=@w9Ln zeZ0?*retGnq1Gok|Cs+vzWvO^Si_hz!k0#Vy3RW3!gVsk;L9Dv4MsE$EC)mhTx zg&k_KeR?IG11w)aU7esxRHbXWfsP1!=@@7%T5FSG^SDe&4XD}j2rB$n(W@l)eO z)nZHo)vHT2+dMEbP^+jG_svR?Ky(s|ne=K@h??ymNwQk6aR9|?Ci}kR{cgpkSTj%_ zg#JZA+SvhxyS_#eXVEOXhf>|4$So?iU6=(8-ZBEap`@)P;-I)SCX3fD%%QJAlFLEi zH9*8Qua<~fNP6i|>!J^e`$-EZ z)<+}SR5+s^h=0lz4VtTA?Ls_ji)Kh+kR=Pc={??IEXz>P2oTBYs6zg<(Fo_AldB;@bcl>fL74669a2ZfzN)3=0LZsX@y(hZ0fiJ2P+&QqY*OY~llh8;AJFjQRAUCE^ za#z?zH+j~OJXgr#Obo>~>=(#p*F4u;Z6>|N%ORTola*lQb+B0dk#jiWxHtDFKV4uJ zOcg$-6r_{Ge!qANQS*RO-J?i1B#y%_bl<(pJ3*BdULZm}7qyowaI9Ak`F}bqC9+kO z>UULngx8>{ax3#X1-dR)q&$@BwRp~E|4Q&2Ci%7S3Pd?1)h`zakn5t5Q*n>mEzbjr zF8Y!vTXF(~%{GrcZuxX7ua52)chRP@OT5x3L+ItaQv^lNG13n8FRMMyfn33J8pTY73i}S|M3H6Jg682D36B54HD)pXB2Pug#W$#;sfhc3C^a9 zGbBIsD(;7+DYyDn&Vhc$z=|2PEL?~D4j35GtI^yx)~(ffI*(q4H8{q-?6yt73LF#P ztGGE=WEL>*R80OeS;Nkk!(o*c@+d~cJ~vaUblf;&F=&tB%|el`W_E>m$R%aUE}xp& zI_%QgqQ;+C_HSZdb(+q?8ca^2R2;0~J&ZKN<9mB|WRVIE?>|iz3x0@F)l#GiGil1a zF1p1*-|iivtq@~1X3=1K=3V{}3Ce*wqtakqPgF7lg!I}j@a3(butfHh>;8xZCV$$H zpdo7M)FEJih@%!RqI76gSJOE1-okfvABQ*bHis7g1whqwV2;2*?tsUB*ynU>#fxG- zm=5jT_iKTRc`W-8dt5x)WLW&RO|Dp+#Y)Ib_|b~qI_rY#r>$cStF;Ve)o~$rzV?XH z>I3M2`GNWvcyBJx>`;_W{!sRbF_HkgQ)FF4zUGW-&=C!&ELO$*f<0(Po5lWP%%I8q zRsGH=vzXbCugD^MCXp74r}-$QIz*9LD)zA3mdT}V<&xzh-33xCC=@07K|Dr;g@C#6 z%%v%h`sDcQ)uz+(MN4L)_V|(aM$xcK++mvd25i^#&PsSu84$9tW5H zu*;6{PI^apKRM~)5CLNwV4fB#BS#zo9(I7av^DRZwXyuP9?N03J%dTyCNt4Bn4&$p z9(e||uG*n7&`7YLI|7~9kP=+isYrI+5|!s9j^{Bz#t{g4YEEXz9G?>Q+bL!(w(Z4p z?~v5z#+OoVp^WEHDkyZ*n-ZwNshcV(PsP1;o+Y%N7NyX`PYj9jpluxg-|K1Mev zFv(x9Zh%B1ER3#lJ0jaE+wO{l@K|EtkPT(D6=A#Jqs`%nF=mM|ifuKqNFCs7-64QY z)R@C=l&uz0B$ZO7P$ZFxZ4_6977^3W)|j(XZz+cq{b7j!pCTu7z`*L;3L-}U<@4qY zy*O*Dbt%bn0t<$UnDcWF0E@;+A@FZ}eaD-%-#9;S__f+MDx(;RY$y07V{PqIS~)WdxYbX8=;Om5_m#$itjIU0(QJIaV9I zFYJNAyI^}&MjZf1r}3+RXt*Te7xqEaXeb-s7^`h;!vFowOLJvr{E558Z?+mzc*M$~C#YmRmjK3b9^(_#PgvcYF1(_@Ubv}y(dPv>w)arKf zuR`q-nnvw(wcH?k=vf>AQOZi`P2~WcGq+CL24%aMbk+26+fHM{kS!u)e7ItTp@}bs zOenT?_&@EE;c%7$LlD52-*h|7AZwE~c|(`AWAqV4qarUn4{FsgA>0WAfe*nW0UD#(J$rDFR@t*}SbMYofR>}DR{*{qN;QTIQG7k<%`Cs$5=D1GPmt-zDl;)%?< zeUYit%BQx-Q>Ud1Y9Y@$lTMvh0`-M`k&crSR%^xv94I%eE_c{VR^Xt%w~e1;-Cdls zo?M20;n*PD=#Yne$UbqZ#L-@Prlqq95lk05aGVt)CSEsuy~)~n%UMJ)xMNSMYh~^H z9$;gqE1=j5Qe!b`zn#A6xf)s@IZZ^E*3>3g7&QwAVqlc1*}-C!^uyfo6fbkwr_WHl zd?e@y^pJ*X(3eWZEioAilq|@kS3tuMW0OwIpx)`RP7V%_ z-2^yVph)?0K!311G^9MEMP)5p1;kbO$Cju z5+(Phcc`&mxkZj%U}NMtM~ls`ZhCH=K7NLmF`;M#pg!ktT)_I%?cYxdFk3Re-FY~P zY-5*$z+p*KXJPkLP^waj6jQPKuVY`~)|h1;w}RIDR)?}2&K2F0oe<~o8z8J* zB(ncjXD8H*3n8OjJbls8|3BP}8pU69Tgc8y?HDUX zg1b-WFHimVJN{+__5E1BjBMqw2Lya5{A>pccMj*|Y!GOzoJd;WF22D*uE zRO|`qRNVQJUftrEBC!j8EpcTZG+WMTk5}x_{G$rh8S8{O&RRi)&efDiPf;nP0#%=j$a3P_JWNWCeNOF zTxYO?$%L;e1IU-nvSnVSu#?nrShifYkS$G=>J&vzP_eZ#WI5O^dFXwDbkSSzV&wjNSTko~T6lnXvyI(EFYaw=MycYfa)#4j+ z)OQ9oi$L>jRFr~Fvxs+AhVQ@@4L>(P`oRJ98C5pvCnrdb_R;6!C+-rTSB*t7jTuuG zTjw)gi~(J&Fh$kMFE9F{`*q_)2HEmCABHVlxbVEHSJ|QkXAOZD3#)~n@(Lx`&NtO> z`K&cBr9>7*4)F1vu`D>7wHPxv*nG$s&SeFT30Z!>E27K-XV7Et6!Z_{Y{eVdY$0$G zDb;FI%&?ffr#+^~Ynl)s~mo0tm}1ob!qmASSM8`sS-&bnUh( zbeGqrDG!6%h3~r-Mck!J!A4G$V4lIlpbEN>-@)sJHjL>%Oyb>FBtbr1W5h>bvKpwUL5Vh`q1qRW1`>;4R^`dM}kH3osP{2oi|FduQ~8Y3M?! z6BIzFUWXVt3=7B-9SpPBf0!LQ4kptcVlz5cZo2YIlE7hf?65$`7D|;tkqxHupETvZ zkXqn&0+tI59o9m=iGkiN`G|Bt(r!V(dfz71U9U5$&p>`SKvImhiY)Qr5J@rW_hRGC zpWyPkF4a0@#1L!A2t?LbbauxF< zM|ZSNzJJ*wOdtUL^={=*l&svYUFv4fJS_H@`~uk7%2q>fyG|2auw zn1RItt@eYW4KuMr>xyWX%paLI7LjP(x5+tn5g-od%HFrwhB_%#2Su(@v0334#R=iC z0U>!8wtdLe76c$C@*z-T&xt#QMmuPLUrDb~m1_DR+K6rED_&2|UEl_RR-(N&|!w^Z4eWK4*4Qx1G)w5qu&K1r?N+yaW;DC8F{i0pEawiXtiq zf-?93AIKp5zb^?&0?}MZ*l7QqDmV9@dxH1;?sv}l9=~55eU|a5-u|&Hr-(oPH48;cQ6j#AuGQ-c7X;!l!W05jWJU*+Xxg)oRzu zKkT1HpHb{^SH}p#yc)$1{2tgB`gTD{56-*i=y=)m*s#d~WP%Yj^Qmt{h5e_U&J_Jw zec`{z`k5qC?;YMj$$(#G3%DyF#i&J8i zx??sOSnV6uI&V+5VWQk?Vmpka)phA8a6#k^g6D${O^Oy#J)O-$#_Z``=K#D+MfZ!4 zXo+j*)`l%GgoV{c1dR+dmW09FUS^lUi(nV}X>_QN7Xft^rq_ls7-&nl%U%Q!W@gOx zWbDso>LwCZ=SOO?`W4gROxBzAVklW8MK+D{r=(-xW+^}h_d*DlX*UO^8iW^(sAaAh z$H{LmbAN2OxOh!w5F=}(_2ilmML)BE;v9Ai;hZ5y5LuNZ$>W`q>3w!Nf^(qbGM(5l zbr#X=Ml6Gm348r&e7)FFXD+^9m3WbCWwr&jUXKsS*E1w(lnlrM6RF@%`rtyOGX^gT zH-bH4gIRhw=V;55TfLKHhe5a(DGj#y6hqSWhz%Z3kF8;ZlII)O2}jnPCxg}${8*+# ziT#(4E|cxA4N8jjP?AN-K+Qar3Pw595MHj|dLGCJH_(M-v_KHN4@pYI%OLxXdy5D$ zqOO=z3+CmTFkAJxQ#;)v>K1Ct2rXz}CW#asX3U8V{l50wWt}jL)i9ZAtk|9yiR`T7 z6gr{SF0x2Ia5_D~%(mQVd${F7pPoTDup5vbcPB)GB6jYSo3=>b&IBQ0V(O_F{Y{&I% zd__7zW%(Pm@BY40tea#`e)C&b$rfgljP+)G6DYqJlSRFolEEft7ZtouctANw9~CwP zl-Z9xZc?3BVGVMaID{9$IV!y9F-SLp{~O)sMZ%_~(||2Tkb1Tf8HWI~pYZ?IoevbF z+uKt=J4#Y#k_x@2n@`Drqi8P`e3Tm{jgwY;;AGtgHBcJ5&1#QFbQDl)j@V$`Kf0Gt z4+&{~4?1^9_s?$Rs^g@OB`9e;V$%sfwI}ftGIVADHJ16G{^TC^v!_0KVD^X2=>4_c znS1EJA~(M3Fm zcn(@X7^}M~+zJeX5I)AFgKgB^pdlH%#@3qm>St==`llV_eJljyMMOtpUh%Grw+AIyZQX5Ilx!?_rGTwTAT4STUNYSF+!g| zDD7C1FR7!qddJEee_Z=+>hjy~RRxtTk6aP4e69Uwvd^3cq1q8!ost#1RdQnI)zNid zT|+|UO)BIdy}WEl*s}c5dqW>wTy7zPQ5?Rh!@?XO7~h9+3v|6f$PXBGt~@**M@n8B zYja-D+SF09V-%^Pg3*T$?pUB(Qls|`5BmN3{5MIDxPdmeB1)R*aY{C_7~K4*F9+!! zZslN0aIuy_mj;{yO-$qiEu_o58^yi+)4VPovZc3(?(#Q*0z8TqfzKCv#-?~u+c=wd zg5Yiq8%!y>G!Jn;JhN^ee( zrZ*G4t~xcV(!II~u3xTx|6tG`Ln*VYK}{H3lOQv85-_1pb2KpynfK>^!^2Cr9_jv0 zPXWoc;&O{Ry%k0&B|AuwLMnK}@@v9WlIWDiMuK5Xj;mO@%V)&qyzH*9G$RFuZo(bgx5ByRg`3`qP zW;8;KaWWP&<`c#kW5U=l?~V~eQRBh0Hsl*TR%Ku-HcagAr2H~17E_}!kHYJV6=uQ7ce#a>2i zE&$)`X4x~rcZ?g|v&D!}w-~BXW94B!Wp=UhojxevT`AfQ%o?|#5?OLY53mvRFv5kfgx4)QHU*uRglZKbwUsY^~=rl2zlm zO9Tt$M&Q`*@!IM0X`)BC^HrgG^tT%Q<+EsvfcE+8mds`b>Uii*;tV47?Wo_bFkJk; zCXq)YYxodet6j0HI>okAl%xc>bSO~J0mF>bHMu>rGNyv!*$vZlPVSyURUv2w9l?*BFv%#j71~pNxqk9;9T26vP&T z?|yJi_?H{f_MnFnP_6{Y0}z>uB?puVyp4egynJ^5{B~}h$1aahQQ@Koq7Yu4-$UUk z`qtYRG{}HNsB6L@VI@5<8?|97SZQwE!hzX`ypAhM7Ine|k-$4)UODbbfMg6_^bgI6 z9^)XHV5viKdXi5Y(xSX3^9^w7kAYbNCEHGsSSmP&r%BMhLgJ)FuKm&q&o5yO^MIRT ziv$|$oU2@(u%}MlvXa(*gmr5)=j_~@TwL(h6Kn-*&o19qA*@_5jD+mk zmIDq$py(SJ3ugRB4;}sbEFDm~f7$ve*2&%Y zjBT>(DWDVu3ODXC_Fn&XZh_N*Wyjc-8{6@3BBRkBf4jxbFq>iKDHpzBxTCS!Tsb3q zlu%KK>mxdpWUvt<3A3+D&x*=Dhv!cx8g#>H8YnC__C`Bq28#Tjp8xbsox%9_-k<-9 z?6P8maahk_LQTG@kc2JC=(*$c`RnjMxw({FSCJ;0hPLMv~u%rXDP5UKk z)E8i_&e2;Sx*Oy7CEYDP3~Y`U9X30l5*)5G>!5r5T;qv{lviki-YxmNu!{`R$830?(`J z=q%Sp*K>ftW>usBTU`=0x?TaI(k%`{!g$#^pzF^9rm#v_F;$ANEy{|l<823TZnFye zy0Q2Y|16Dnc8lwPLHs`Sh3IeKG>F5U6J4f0tD42?85lwMIC*oYyR*Bs=iv`@cI@Z* zzZfKqRy^(9)tmO(DA`qtv{1p|qNxWLO$n>%tyt0^$#xtNHE<96M|eG4+~k<;+pI#q zm5sCZJA}%b-&y0HBPehx=QePMfhp{vq|EMsPZPV`1Cl-+!@RcI4%?*ghHV~*hHZMh zhHc_Khi&Selifo(B_d>ODG}XzJJ+WblptD#yC69??ItPCY)Z)yTk3E~7XPS(@umC+51si=7QIGVQw zBUH@T*S6=M4GZdx)Tkf!jguk?OOj%j?U##zPPS8oQDPHbAoKnMP4$bJ`&(b$AG1eq zwQHJhIVCt4mMU58&fG}xQMq?Dsp6cMmB0%5xTKs9`3-X(``=!y#z1lxy;-Kdv$)k8 z>sxU`_>z^!$I{k8`)M#FrWy{t*}xpDjF$kI8G3)Z;NvB4=%y28TFL*Ar#6f_uwr(( zoqChcW=gh^BI__J*)M7KR3kSB9{eBJJsiN7qvs{VY@y2swsOB~n09BiM#D&oU9Wgh z8p?|YT?X{tgz~azj5PPsG3=~2G0I#?uO&HxgU(5E%Q$5w9GaI8xW1v7uoDZMWcIcP z-t~EC=%!W}7P8OdpvSNc7LDQgVVnD`GT@5Lr<2{A zRhad*Q;w~U3%AaSWbF_=wVdjuY0%KOBV#|kWP~Su**G*N%1J25{kJ1Qx+TZAYkqiu z9JS&d>}9=0#A!-)f+Drxh6V|r6Tp&wMGBdT;eJnn@wLdQ!m}0%gThtS!rkms3os1O zD?Z9rYcK2&oKSvAclp)v!9Ee4S7}AuAMIQL&G&0C77wgX)LsiA2UZ4 zBRIX`71p`Yurkd^N!NX0$C4`Nb#JywFE4u_z#E3qg8mb|n;b9rw>>rwH*Z%m=>B;NyJioM2zV#xqsRBV^h;N&|}AYJygUQ2W2y z0Nw2%*r-|b)VW5~Ab!fh%Snnqty)brs-`fdSa!ge57g)#j0VhlUFC1Sqe@$*GfYkY z7?wdwU$H#Qi+aZD1SPAb$Y)eAu7@+}z9qOG&f^!k0>d>%sG5MydNcU*2Z*|XeTmAC6o-3t@5eh(|n{C+GU#qyQ(8W(VYCCJUSVu=|bhr z{+-;Xock_mP*+e)Zuy|h%mH?XbkP33eVHAe$C`tI#kYW*=(e&%*{r$<3c!)xp`289 z9-S#~klg3oR)(>rU^yokQ!`>}g5D>8iaYclazpZ|@mnk__PZD%q{2Y>phGn0(OaD^ z==hJeZQ*O`pUdg>oHnRs!g#%Thxo;Vzz6^@?#irJjT@DSPp*IMJ)Ke6H=lK%R9f-M z8N!KU@^c#~83<6Hq=E~a*7^YP2B@qiDKM8NlvnM7fgf!(Tefcj>kE(JB_IP&M>s|} z=CajYiwm62_+-1mT8gFCX5aJ)<9yWU*TDm*vMm+%NJDtN{Brg>$90bT*(sp+fXTn= z9qdj~%-k~HVM&aCncaZj#-(bUgw{E3_B=~gHi}VFvD)J_Kht**NH&>)j}d%~FON;R z!U$f*;WNQz>c<>**vHzY-8qd8hf|pj=$lv?Ve#}K9dRpE1swl^(<5rC5xcQ z1}b(l&eSrPX)sJdgqgmIQ8!2az2#pFRlZ*nL>SQo59jQcM057} z0Op3xH0_ zKGA8nV~=R(+fmLyJ!ZSwaT?}~$qhTRA)HKWM#~Rs{cnTk7%DR}u_aaiWFen2vs3ayUrX|W-@3Q2cK~#UH}6VFBwP6L>u10dH>txu0v4W ziY=ui-HH)Zt%sl@N(OQvc~o#4>2b*66-cw_n@gKjcf?6RC72Ec@`LtC3UqZq4I8=R zlq2Y)v9b+(&rsL)h`u3BhZ?tTta{{z!N&J2*&jH_Z&vkrG<#f>H_WN=tRQ*j&jU3V zY)nLh$;}wShWgR3-DemkzPu(26(cDz=qSU84ifS7(p{2#I4Xxa>1kNo8jhi95Hh^W z@}D29i4QYOd|?6#D|Vh4DT>@Zr^+?dwj^MSV-q{#?VXBYn{zHnK6q4Sw`oBkt4Wni zCwoCk%v+lloL6m@wR3j|l-V6!7!GA02S`4!gYUB41s+-im&Jf&8eGQBM9)rG3OGOb z=)YFl=%$|BC%^iHgfNQEn1Zlz4yWZS87YX7t3Yxu&6pp5ve&eiU2ngPV zG&o$1RFKb{wQV*?cS@h#g26z%%|h;a6XEh~7MbW^>fjjP!}2qc`|NeIO6ga>agJ=V z;tdT@D~#C^@1$fqD0r^H+Hf#%^B@W8UXNT+p&xv}%%gN=hZLnAk`x-*8?5G^CQ~tq z@g6@o7C6Ib-!fyi?GJwacRF-dCQ>3v)p$3Dic=H9oj02b~JRvj=J?DfU`#w=F8nG?@mI4aD*@xrdmdXX78$_w%q z7>0FTDKEZ};7yAN3xp~?*TI`#&X1B$(^Obj@G;DJe7&4kwQPdy9zm?#ApB@s9xMk2ofxUC% zJ#Ngu!U|v5A9#wc;Dpb-ZJ%bhb|xmVrV5V&YYJ8=ujOPy5n?2(Pjkm$V4_nccZ2^1 zw?ySxWx>)7u++WkR^yqiOc0gQUEC1fTKl7KM$D>$oVH|-D$x|YCMLkfgh|C%lgT5@ zIBX_x;8Wk|%idxbAv6+nt0qlg!}rjw-w#`!q*%?`H5aG-X4REN^~!$sz6H6WBn48) z_BlOrn!&*rHr0Jc5hFOkQ zyP7qUw}@qcD-q8aV1_&-|DA!;X4J+4#}ufJZ9iSdjG#JW=^ws&#Sjrj9($MS|rk|Z@xbU!A8)0Ncw8^1Hc|^I-w?(wc zL5+2R$OMI?*1&t63Q6&G^Z_R%HceOdIBsI<*qAJ&Ou((dbT-*- zgARiZO^PNE-@W6GO(5ojpvw_#^LxOL=3Hcr*kD9ZlM`pLX^J_r^nu0n&Pt={3hA+TinX*l9U%jd+1}lJn-udOX7e$`G%x>$&jNuOVrQq1;^~? zPF;&5S^bh2&^4>^JVqeh8Yt_U6=xvN%~E)oDqPIFw0;h!``7*y%h#Ef4Lh&@C)xbk zn3lbIrX`V*B~WDhXg~UxRJY9zn;o8OZ-W>&+Bv(rFPYRrNpgOcs0NHP_?<-Kddzuu63g3j4i??TBj z_WppYf+FED_C5Njdkm+Zmt|MQN|=4reVbpAq*yR)b5faNTP!HE>#&Vd6Z!6)1+MsJQ~je%w;nE>LsRj?2W#&l~O5EBGsW~je&vdB?4&wRfs@gmu3 z#j!mopcpfAq){@+eov%=f$0m#K_56bD(ZQ4bb(-#Ly^M+`w(8XG6|T63xNQzSvBOG zE2@(;NZNhze1fP(kS6SyhTMS3p~O;qk>@aBgc2(IDMpVn{C^^`X>wfT}84up-IDXrnLk z->6f3a(H>OSKDr!6E-&*3bxAZHq06zWiCnb%|4iJu$q_U{=mLbtX|`_ofS%MvRCtx zloh1V4-?kPx#t5kgGn5*~+4VE=_49O? z3j1ho1WA5vc06TzXxmT8fZZk=DgmVfq>skTlo0BM^wWouFb;D0T^07djo{1WAe(KN zI0>|=`p9)*1FUScXTxXSke;LWIYhE9sS2G6ojM%`L^o9@>9sR0V%5xnC1xlwt@VsO zfA}}U1uhfMi51%?BZ>+c$^$GEbVP_jJgIZRg&Ub^B(} zU0}ojKhS8OuSIa7H%gNPiC!O;_^|ntUq12InU(EXk`!`~nOU*ot#gx}S*fLDpHZX| z#8SbVmPHTRtJ}Fr@`8mKbgVpTMMKc4KZVU*=cE3Noyx`*anes#{V8fWCL8=E-&URJ zbA}k4{sd09b8%<&msNk-Kl=j7pfx?*aDqEGIh4~k+0Y_gdDy*;WV3(`5NEJ(&kU$x z$7Gxa8lsft-c5=c))6I^!JCM|NdjpKlbC6+TftjW=d1(Lzt+ETo+McDPU(;yG;$~z z2#ZTR;1(iUEUMlajdBUkck;$2G2vkYW=B5XMgzj`M)F4R&0Or^z2UxC4=m% z1S%L=r%^BUSzNH%Ba(I1@0#B>m(_t-k8{CWjkE;=e#hPh1y)fRXPRopCOPQLMsO0X z8Cjhff9~8cV3_3enphnpDN!k4FHkZpmksKZUMC?X{DEJ&cbbF7?%<-Mm7nLE1{>+) z!WvU#CMPjO<8NUMrtw`^iB)n_eV;eXof` zG216u@n(9T-eiyLdF5Tmt zMYjr@+3L@nFE6|dlwBrnVJ1I5cqcRLXcrt)`$p65ez^OiIxihUy1&y?K(c3&I=x3* zO34mVq>u_mp@|S)rpHZH0xN|BeoZ=D)+3CPKXL99wTR+9b9hZGP*Rt!=HcovLRRfi zsA}h8n|{Bij@=;NgvXKIMf2}*qdA4H8y98KQF1(c fRV_lOUSa4v0IY(nM58dcm z&$Q=hQY+&x>_poRt1ayLRdwiZext+At$?Tnoyhz3ZVU-*JPrvr93 zoQlb2S&ZLW`G&dbaOVS1&xH00SI%h1O5N`J_0o;3POfGJG}?CW+!32baf>LLm*D_D z9sEV`m-_QHA?m-wh6~cM-V76~k)R_(xmF(T+_kt-{0pf6mOjW`@lDe z4toVQDQCz!W;2@=FNiYq%^l0g*A_BtbJ+_WJ(s=uIO(tz zM?DvFn$DNoF$Oq#d1q#9(I4)WX;|rQq*! z?uyYHX}vD>i+{J~?%x*4+fm>vzR3pPSSvtR-s>PK)E+Cfi;SAS$H~0af;Lk+M&YelCndA zXX|Kf?KP^mG}!iu54%HgQ9o3N;8(Ug9P(-b;l&KOdIzf%^v9CvFG00oWN|x<1tnPv zhwKw(pCxK!VhCeJ$#d9e-cjQ(0$+#~a82Mw)}{xlFM#-Fi|7ksn-D@&S@dD|Y=fjg zlK{tjZ8DiV^iY@khvtojUK%D&JSzsA5&zAfSDgDu8RvD=U<~yhoH;)*2ZuSo-FRPvm;cN2RZrJ!5fyt3fTtF-q z_AWnyYN)fY=I{qk>X$Ad{!?cAZjdO@b`4pBis3;#G5lzc(I_dbw-WUvdS%W&uh zoB@u^1rd|_`O6>uA`B;&@gaa$Zwi5lUxCY5ob`Jgf%#8aF;88^3R%(tH6+@Mls5JV zFv3kwts#yD0|;nrh)LsThDOFsN$Q<@WrmfiMoK)-+BSH{Nke62p!!dyS=?wpdnimF`E>0kj6d=ToVeB-Nb$cI^zk2bTcPzEA_GY8 zE0*=WNzbwSl#&fnE z$4S+3nv9k^?rTA@V`t!T#XV9E`W=P9@mA>f*uRo)=O)vgzUoew0dn1`&K-M+)AWG} zDnFR9o$CLzBQZmT3K8(KlBt^aFpcFTS+U(fT zpb@Mp@j|2ksc6#|-jLztYL7;7CPC(c5<&XHVj(n&@%xhQmO=xlT))fTB;DejAc9?H z7mZH9PhypYet09E#Q;`>EJ?A&`2SO!PnABucw+gC889c2JwJsx67$_b~>-8cG{_T$kXQ~ zIOE48#ba{Y9;;{D$RU`bJL#>nsy#+?=&GfC^Cyqpi7;hGJWtg9q|Y(K^9{SAzZwb% zS#ABB5lJA-y+gr`!iY^5=ni5}s5ChBfilCOOf+Dl z&vw#3jStEzwFM)#HEI(Ha*x0|^)YPEH_Oz6bo#O?WhgI;E+kz53pFOUp^9ZPy`HFR zJt75|rr#K}jjdiQ&*OLbR=cSGw?W(~L8L@;G|e%HlC|)#5?H>_E04&UI$B1D)6A$D z?RLsPUlY{4`-|gO6}pX+>eSDUl2j|szN*mM(dAPzs3q7-1s6CC*n>ycxsx1YV^lEL zZ!bt??&T%8cLW_{m;07`7JB6RHA5O<>^H9o|9nHbUsBG!PVaE;@!IE>DIW$_p&Y?p zpbOdQSjQQ#-vHa~SoW8eV1O_Wn5T^0aUC!MGqvgL>3=j7H?!I$uMxSj7{69crO^O6 z;{dz!)PHYm*BB^^s%zNb!^m zu)dWFZV_$sDTX`*%tlxb=?xjmr@|N@o7qP?$c2Cdc1enD_C=Nx!<=CJO+%7y4K>02 zmO6De`gWLM>Xg;m3nPk``A)5NYRu$SBS&S62usnXFM2U%WK0|6F*i)pAtQ4!WBYG6 zUX<$0#r7D_OJtiBFFT6#3_vC&OQT2%6^xwQka%QwQ(DbIwwHKWwor`}3|Vv0X z=N51~6i^?f03IkT!f2;4qMk(q%YUN9O@{&X4>MrLG2(~SeaGb~}WTC5l;N<>Oa zaI|1CPLrYo#{HgczEhDaAY`vc9waSyU;IAdd)(v`f)?f98Tl9u*iu!*M?3Q?r^Qvgc#Yg zUiFW*#j=iM5I}=KAF9D%m=AJE;y{ggt9QTjzMVSXF5hn2a^V<{wnfHve5Z_z$Bgg) z{evxr_0Lv|4I_owoix&~v_Q^EvKLG+2$J;#mlv6I1=s;yDMrP-hl@jzmU7;pgF1v4 z0{Kv9ZIj)4>C}0xAe{Qd84_V2r-(kINU==;{g@&9HOqI?cqVIMA-e*wXTDuCW9&@0 zJ{fazJPamvh7oq&aK7zQV5sxXWQSwL8M#JO;gPiIAgAh&h%f1r5eGeaH1zt$$l$4@h{5I=sn(rlq*&MFznOa}Uhv zgc7c3PBPu&v0?5>mu6L`?>bHjE0=X+F{WqUUXtn6p-6NwU*K`#p<<2`oqd_OJF}fF zDcrciaD!uHAA!4t!*BI0QE!qK%~$8r{p=#wK05X5*MuwMq#KrB6Q+_aoFdniSu{GW z?vaZubXHmZMV#!THQO^Z(h)OrZnZocA^Ufz469YGc8y}B{uCp<9sl+)D3hduvetNSzwD-?U{SQk=ZIS$ z*#%kwO{{zLak^hp#sN-m$1>koV4mraMzTWP8~8D^Q&^uSI$aR$TYBs5S^{OToJ{sD za>VwCZ?oz&t8H<%GJ;d?u*)fvzO1YZ!mevjZ5)w$dC#XWAO^6yg)JYfT$mm zT-xQE92l)xr+4(mkCL%7j7HCTSCzjl+Hzqkjv6bD2pDnH#CtYy*ZV}W_Bo~s*6>QH zZ0A(LpnWNo<+@n~75MZuDRO+reX|=l;k(3b)1_a*9OuT6Y2NeQfB9A6Y@OX{dF$dw zB-V;!iv@ZHF`bfuFkTWByvLz5pvZ08@~grwNhZ4=l9!Ne`QCT02H|vu$B+0!!rLoW zg;0^K{~2x7q`Dn64Qz(Y2Yz%mGVeee_QuG-QPH#hex8dCCd$nlDo7$Thp!dekt2Gb z+DFNNA~lN&Zs+a;c;NnOHT6(3FuNa?HVN*bkO$Q(J>+#RF@e3b9=^E;u-(GU` z&4^i-q{kJtq@Ufzy(GoV@9>#e{d7%eAw3Bg(#fOCvO6uWxD|bdg&s`&G*%oSFj5wR zv@1!9T)#m!dTdezY3w-Z)5Q-LTY8Ue_+Fa~iV1GC@F|Cl30tQE3hP~FL59oNsL>Wm z7DbT=Dmau^PakxSmG^ixKn8gv3*9pqhW`Jccm}obz&$IM>t>cEt3GchWz1$4E4CCZ zdZ0K%$xc#0^cvjAEp%SVzc z{uSWF#C$x6M?|tTg=>>b&YSYgGrG)q$z z?*_3NA=4sCf(;Omp0@I_%o`XbFuWG% zin3aNu@Sc0y`sH1Ct=D|p6?^1GKOpi4VCdWzcSy$zAz(MJYqC<$QU7G?2Q&aW`v9x zi9cQbJC1IBvoQX9$H`h|>l-VckJ9zlH`^%LR*FPZ!ABs`8I-=PLeWK zP#VxE#i$)HZBg!PjMPDfSz}i+Y8T}Au zh?U36hkzgB{|_|)iJ8Utplj>*0(BO{ZOx{=`h9cEeFis$&DU$dz zT?q<^5J;*NRYFb^kWhM@^T=g4J0ss2=1^^xJzbp7D)&a#|4+*48`3*sJT+odD(sPB zR6X3Yj)W}1#wc5MNO4zz=B!44?I$P1s#l23GA_Fq+(v_ecP32Op{KNhk zkjj1dE&P;AA>mYkXu50=W6YR7>{DuK95erXVr`8b@9*)y>!dg-t^qAE1vh12wn`I$hwZ2`FVVhGN z{JWh0NDB4sHhbxOx|(~1Rc4pVJ`ZX+*!w-Y8p7vD#**u|!QVW1VvLsYcEs?^lu$By zvvIUc!0zC zv)*D=L!(cA%ui$Sq#BYxXftvntN#n(*eL26KDh zCIb?~X~q)JF%FW6W}Ts;ydOB|fb$=%Eu?-OPN;-bUw$Ft#cd1b#$0?3#szmI_sL}WB+p?P-UK#%8M<) zq!9JuF3oZ^<;M-TOr-eA7i?p1m#tG)kUb8^JuZ7*1fhg%=YGEqNt0tg`=;tcBz}WH ziXW^Bm-VwT5R&aSV25;mnBb8`*EsfyKX*Dqwt==~B&)=uiuF(#%}I3ZGe4!tq#pMK zZ?IXO@t97Sy}(wR_}#sC_ky43EXz-h?D`tH%*?V_@m_jB&$8U6WL*^LfOKMyTyXmz zehbtKcPkrNpD8y;p@6MKbj&5jp`n+6>{j8~3}vDd*4$$V;ehLiO){OsMxDe`sMJ?qA_MkVvL6Z|%Q{#@ zbop+cX?i?5g|v%XdyzJpM);cPKB86hH~w1Xcy)SAKd! z8qHY^NtQAGG5$9sr+}WLl^Ji`!%>FA(*rD+`^@odpXY;OZ znhM|g-G3w%R=jg+)th+gDcNa?oS=g91=kjR^u?mj7ael_7?0zm+7HMD*R-@*b&=Ia z$N4mHE1}M{S+!Gs-nrDiXiX;Z&=F)eX3WS>Pd*y8V6JSz-U9Y2nwMMjPe@t-8Kq5(bDldQ6=(^xI7ckr$jaj-H9O zRR2BS&i30nM6LL&|DVZ8D@Ihe9-=N&vWpaH0wsF)UU2@`(r>|z zQME@I@I0f84oH7M{3uUE(>`kCkj|owO_UfTIdHi$q9e2>S zPy%Jpi8;MqC^el$Z*ypXEGS%Utn=xm>pbpz4%>9hKjW1?r!)Xqp)q%=k^|-Yexqy= zBR0o4WgJalbQkx00K`ZaqV}SR;i>5Y*I1~%08`H{F#^|&!~eL~nXLntjk`@K31_yd zw&ElRptl>t7ZXRxwooJr2`^hA243Iy4_B&HjOv6F=1@$=y`*Ab&BhWeKi-nsvw15J2h z=sa@Gib3;K4>Wfv*%uVKMFq#pTJ2Vj*sNTcOz(He5v*TuaOujG`(5sPYL521M7*s% zm95 zw`jC+k2C&SD>IZCPFQOW0!s5wk-eg5b~md-Qw_e~B}rMg24zly93`IPWP3Q!23iXe+!AdrH^zbX*wb zsNYMZAc1wClf|y2Tlu#nXXg$HwQY8Yo&xDa2P=oYLzK(jzTkys(mV#w&;XkD`9;Rj zxWW6!A1-AXmhl^@e?Q`pq*FFbhRRdaEsO!ik5=h1c7;P5dTvun(sIkWyJ+uHaupWv0Z-go)GfBn;M8J6+Cuv)(5y(>YZj*n*59tU)X zG%i2*chBqP4T|NrteA#kc_#i?{r7FM<8R`9uL(?=Q*YJ2>e*WD-29+CI@u#1yn0!{ z?~o+Vb%|$n!s-IKpF1Vdz_JJ#fUeVI*bqR>jGKv&Gea|GsLfn=|E<5R@@y5o1 zNy)&ic$I6sXFhA=v}FuGA7c|C^!$ctIx28^8ubI+BKPO{zZfKq%se#KJ7to)dduH7 zN_LeZEmUx)OO0ol>6*@rC%;?@HKKxMIOD?+);$%_MROKl^e5J-e|jlW7u5Q=EYM-hMxZPq8O$(d#opo1^e|RT$zV~op9;pTVid&g zlt9`3X~i0cUK;CPfa4Y81R5a&Bzrv=utSvLUe3pqbWAL|Y2PG%M9TRgUgdm{u@uBf z^I0u2ZDppW+&hb|wY?M+#Wt^l*{ryEDH=>~Af4Mr84AZTNuRJ{+!&FRMN*lQqDHQ| zz&)C5l9uz;n4^yQGbrVER@mTuQ4SF+mwIQYxQvs=_+`7Dotr%$u~Y7?WdgvOeyGrc z^}ctfNE@Rv|4PYBh%p@sCbpbE+`cYAXKCyo9*-j>Gs$_qSG10jL85OJ6}*~v*Lj;? zk)*?Rz%SA{AFAVv1tC1t>usFdN6I}CWUze_L&TZBEX#6D;pGbk>D}OksFIy@d?@Y` zuXk!zowY9w=#XxmyE>rEu3m7)3z#?UA_c{6#lm992F5{qDm=jMkX|JD?BjGJP{?Mu z*ScO5L6RqY)6|B=OhYh{`8V-lPuz^zn$&9Z1V58~y}>Y|X|;7PMq-p0k4RGVIOf2P z2^c(+0t>uuINkT`b1(2ZYrAv0Ns3Gd?8Lyb)LBM=eIx&2Q00d@3lmfk=tJ6GF{ki| zo;kcv$?j032kG!STtmFB`<~->kRk39(F0Mkpji~-_qkJ~cahu1fJ*Npu$5Z7WPmiQ zzI^8{_kpMbgtBYtERP6I1Gkr-$v)0)QkAWUeLwc^{!kRO>QBGkMQ!~??E6C>{Glai z)n9*|LH$qJid88qE`B5Yef1x=eq+_4f8X)V>p|yLrT)92yd1aoPo*0d6+l5Agvlg> z+zk2@Nu7tO*yVJj_ibe&Sk1?heZrgI-%MdeIG*NruH+}XLtQ8A?CLbVKK8GMJ1@xwM8&LpGqMzz89in*YR1tM``iBa49W1V zb|y2z(%%U?=tl9N!*!oZ*eDf3ausBf3$Q{1i|~*M4Yy6{$~bAc_eq~val75r7FVVY z%OucXYS&Y0XB-U^L_KEwwoy@UV)Uu_JagUqL+Co~q2BlqG zw9BdXom@Pgo?&32lLf9yL8gU z0u2?GuJA-#gOa33@foo>;Fm7Hp_2a+x-=8f2BJ<005!Xq9~9<0P%FGxYn z0JDQmx_+q{-7=G~EH7M|m<-vMG-QP68POkql;N#gxon8*J3#hXanjByy;VvXB|A)! z5-NCrT;pEnUUhmT{3v;T*_=x9@%S^^WcGSNnRlaNm<-5@gxaPloO(eAoh$0)S2>46 zqD7n(aTm$D=aj=PvuogL?`V)zx$d&vEY*dO4dM;Ggt<3k?eD9@XU5D5{^BNlkT#u;zV!;VMSwD z5V3`5OClai?ut)}_c^8rpBF?71`Y@Z76UuyvwpQ@5rp9)5)AV}=ucPyz9JmsV-ci9#}CgDYoOxunn!#0`hYEH+JWOuxdj?W}Tt!uVh z$CAfOYDt-0iBC7F`{BCfBR^SdpCm8i9D}xn9_x7>z@?Jxj*+dNIcx2i;po}G896W` z&WwD+~T>>LIZ*8}#A-dJB_UUcY>-ITDR-<4TEZMk_m+mOTJ zmEvYI!c@`9-|W-MhdOQCZg)t_T-Nb=fVdnt%s{#nN3qA$HTs+6F_)07s?)z)Y=v?J+maP>`L;%EXFuC5*_oV_p^H@-1x2=9>}Og6BU8L@fv zR;NpfPrp!07kY``A{v%n<(XwsI|W4O!DH-)35S^BNoH%489%yr*>G=cwQE@;yWL1x z95C)>(dz>@iav)}@@n@C!NFz8P|w?>00Id~F8i2sHwXjPc((HE=x(|QO0PD(z1nUU z#G;#3u&8yq%DQAirY^&Q7}I=8SQ*ogxepm(h59##;W$IT?)bg275mwY_~ys zD4=G8-=R|!&55!*1*_#wZkL3)p_&rX#+srRA2SA`aio1XbLg01&YqFfI1KeQDKNXJ z%msI==lE&tCLpED1hV;Q7PWu}VDh78{II5WgYic_!RFL@_mR#c(tzy?j(ez2D&uW8NlVO4;cUXc*u-aVLQm#x z!syvsugl@Pj_-;L6Qag1PpsGA0YS}J1Se6lofO#tB-1`^-W}4-{yPIT-udld z%ns$eFhO>UL%ByE(Yfv$BzHmKXIc!3iD=eaN0=A46Zd8u&Fg;s?d^s+l}tP;Rty~@ z8JHnUk_F}bDEW@L_3tzSUPf#N9S$o)$xU`YcjMAf-fGopml2yvQG)7}EDChNrU^Qx zFP)yFXX3#&^VwAne`jZ_%Q2|?O6C8OFlIRh)@vNeZaw1?OUYs=5{a^ax(E&N*bziK@$X8=yQSSzZUr(olyI@Vi|TZkAn?w>z954#$ysk%yDCU;+bII zMcu91ecg~7#%i}WMo4QqA#ix4DxfQHhanyun4Q7;c!h3B!<11|1%?yP%YogE3?soB=^a$J|Q8m&9Vh_ zcE;@Lqbb>Difn{EQQ&dKAp4H{HlJb-t*B(~bPB;eA8=Cu^t|=-9W7ry{9Ta_65we1 z9f^KzAd#mBi4;nfNRb38xYo8)RP0gg(Q0?VPNQr+VBf&SO>&aMMUOH&tbU4QX|vxj z#{k7l9!-nu%@l}?TcAt>QilJ#@mGeaj!d@0R=n9Uk^~vT>vK{!*ro|lnO1!RRC1EN zdc}hp(if20Ns`AqpYurc%A(;arpz97yScQAb&8yG`LiWgDlpcQ!iWYmmIaFKNw(yd zIx{lk=|9aUomRa3STkenT=9^S^-(}46^x1|f4(8T1iP4KRXoX;UUh5dZkXA>tje|5 zHIlW_Eua12(V&epkJCE?pEw`mWYF#WVqy5q3X&w~+PNuC@t)X1t6jQXux?3`@_BFY6*Zp7 z?21jsoQ_()c4m|6i+9t%mi4oU_rsQdTKwJpUu*s8c~!G2BCu_734HlF4&L(|o|8w? zI2&Y+6VKquY(=y1`r9c|9qzWrcwQpgm`z_+9E672jxj5mOiBj9&J-%ROU&`CZBo-aQb(Y=Bgj6QqKq*qx)hBq@r4X<75?hL#aHrZL?7RkyC?hPees zLinhVn8T~5)0GFEn-nKVzSCvd>e=^52qgT(21Ga`L*U?;`PZVL)^yJ3p<4dedG-yA zW_athyFNQl{L4(;jH2rNNKIBx#M@)Nzdr$|+n5DL3?+-C$R-Q`^g_W`Be%h))%(65 z%8CBpWI>p?*?9)5={qnEU?=5|KA5A!gxhb+J|sJstum}Qa(qw^Eqf>#l=SSTf=BBB zW98~ABuiv?({wV1-XpdZVjocc~y+UY}wGcUEFmb<;TgE=>dDPRA z;bcZuaQsfjVyODT{CS`H>7erclD#RU(294=je3Ls86~TvNEsE36@W(}w_a0%u|a&n zyMtBjsdc2KD;sTd6uLF(CRJJlx#Km}|j6E|2 z)EGA|6GR8!kTm|#-!Lto$?DaL4UUl{{!uyUb~v^%5^{2th+%mNb3CPkB? zUvj`-jcM1&2!srQP#)C8ijxj4kCXO^v*-k%8dPIdv!%8_Q%BYm;PZl8>wmI%>8mcx zHsF`OEP>Z1z0WEpYdI*rc0S-DJ@MJ=PS=Dx<$HXqx%JMGETrbq5DvmU_kfZ66sZxU z+7I|uu{H*tvcz6z>M%fy@`CZ45gD{j9C#_oE9E8VhBn#LhvRRRw3}v${MSxlg4fLRWnp4%XopZ~4t35+3 z&hu#M10FlKYmP1aq#3=V);ouPw!G+fi*-0jm8Wkbc~-o}1zGbk+u?Fb28WkzP+YR^AR-620WQTn1TS%%3EP>qfvYShfk-L6p$TiMm$1cfk z(LNA7?4=P&D6&}SmjacANs8UdH3Ze&dO^SZ3M+hJf1r6VVPx2*0u>`&n&mVzsC@Dd zbw6;`*_x-aQyWOqYvX!3s%M$wzE^Dva0H+wbfuyUVoqCCyV62*~I zf!8zJV}XV6PP;(Owzlq#~M=WASCycynMPD8}Dt;WI3+ zIh+|fD-yQOs|L8pfpKK)!XS#0MNnh|6|5n-%9EUPEcET7)%i|Ez*BBfJ8k5jZqlPP za<3&%Gcw=S`!dC}ja>YR4iP^&vg>Q)@nH2o`sj-! zl?(~DcqI|Yd-REt4@2@3l0Sx$Y&UFij=nx`3;U)jnI5)@V{edVy?KM*thx`MHp#OB z3Kb);xJidko!rE^sq{Y3jXug9wh41f5FKWneG3E~;M%EoqU7D|Jh(55J#6#XHPWY0 zbp4UD|L;M?^JX`;m38jmdL6n1AYVU&8ko`Y;)b`06U9) z$hAvVEGc%YqqYBn7rb&Ny%}Ss81o()f47Xk8Grg+ceLSp_cigh7+D=h%kudBex(5i z?7EbgZmaHb?xhbw&~eZ{(Fy0MI4PdqC4yvKq+-u^YPCaNxb+m&2T7_ukXbNRZYftT zjNuePWE=%kypmk;`=3oEp~-r^2E<50Q?*B<7=wwCtSEVuUElmHIzk)=5;2EBBdCMb zI^C-n(}6L?k!E=Ng8I4Yy*^#^UY`uQQeF9Dezf#&i%9>_yp!UOI)IIrBOcKX~@ zrTw&M#d(nT=<#acV(Ictdu=1FvfVcOB3J`!ueM2(1t|DXBt)r?W>xF*YeEofC3W=C zAa(ePR>qnPFC1%=arQFnqq+XVuVQrZCxId^U$*2QL~{!1GH)hTiEB4jQCATNqQ(G|#*6URe4N|hl6nQ`eV?IhIkf6tVHgJnv)u+fl zNO(HvSk6A^2=3NlNt2?%8*NCDYln27FOV9aAv@>x((Q|*NhYnmps7ei`FQQlCqb;n z4pApQonFnmBgQ^QY&z*i$98V7xZkfuR7f89srS$s^tQQceHtWM1uO_yOM1MboWp$b z*&)1SK{l(F6v93~8%PC#5H;VaL6W}WdQeE9TFc>~=>(}g5S}A8CncEDm88IW{4XGt zW}|ASqQ!?lidf@VlNkPmcb2PO^#;3ZtZF1+S_=-o68AHw9Vr~gA|k-`&0=aFj@StKjo zGe6avm+n%sFDP=03P!>vOtQ##IxECVRZS+`iKS%>?xq|=vG0e{dPbo#5q zJu~RM+5PMx$9=x#-Vx4foK-XFwIoMS?SWsY)eG)~&7a}?X)3O!W>ieNTMw#75U*&?MZY2E4gL>>6-14kbP&xRlqShH z`(R>krV=`i<9<;kA#93{tnJlI<9wWX3;z@ym}5*|(WQrmT1~XN*KV z@GPuZQk5f1)R&j#`C>^NR1!3bv8(`S$>L_MTxo81@(V_iA&OqmlX)-wwfJ0{p~e{# zTv+jL%7`M`b#D5yR6(ulM|(-Cpi%Mh=x5wt#YyX(8@SpFTIXPquxY7QKRC{(QBg^2 zwO{Li5hT`r4eui7sYUeArX04!DapJ*(e*Z?FxTFYr;Z6;tauA%gw?JSDj+rU6BZ$b ziEHgabd_B=7gOPr1)b75sE4eW*Xj*3Rkt{sl|>&__7Uv`waF`yNe&f5qw;*6GRf~V z^|m`-|8uD!L$}o~NsO>@kF3D7--OxMrE66;?L$Q%6yjK*sX}@#5Po3UIr}p2dv>5i z#89f))F(Nv9rGa{*M&t+Fj|dJYw!Hm#znfl(DJtI2$EyPd!gfcd!fUW3@RIosNgeR zX|gWf+L^_qUkYmB^LwR_{VVBgW#1A^W@uKSa7oFcPEjPQ Jl3(t}q-qXbm>`VN7 zr`;lK2T4*w;Mau$g-+i{?|jJ@9u4A8j)DkQlDx_L7-zSew0dW7s$8$I*4l4z0HJyMBI}^nBWW+c&h|^v2Z`h@-h1Hh10G3=M{lRj zs}b$Cocz^P!DTiy8FchNt_}T$wuZ@y#fsr##4;fhU8~&y*(~d$@4+-SEGcn2=@iY` z@}oWU@CP05HNV^Z^Jdi^2yG8bjuE_8N24R=-rRa15o+f?gy_-!cRPdd99}JQt)x4o z$@G_Wr=$VYfU#Yf6<4nqr5CFQ z-Xo|xiRLsw-b9HIQ7(RnA5aj#Dh$R-}mwVrU1f2oELXm5~^zd6* zbcz5XpDDK8P{Q9yS3>9!6+M&bCgB5@Lf4a`kU&F^|D*=Ryf7IB&7^)6y_<|rVE^Fb z`iH+Q%H}>kB%>=LSx^npgFiK9~X)lR7eUqd)$_`S*tDdQ2uHE4DU9 zlKWQAPVqSya29eGDwcwoNP-yRap8V;3M-P;Wl?Q_iAK&8SZFN6L^oUT+~S8n_^qL8 z(JM987||vI@5!5CVhA&$<8@^ME73m_y!g$EFWH^k4dS!F6ZHSsdlR^(vNUbHu6RRo zFk~Z`dkat`fI(EY&>}Wky-ZIp(=*f4%k(nS)7{cNUH$LruAWId)7@V+DlRA@D5!vh zpe%xjvWTLxy5KHYsFa8cg18ixf+B?q|MyLTl1MZ+Bz!67D}P1qdII-6_dVyl%k%67 zMPRM+psJK_pm@)k5Q=xJaWFDW;HLxof23*4W65^ap1*v-xbSKF+dKKBVyq+wCpHfd zL>!fTcZ6c=DN;klBAr-;?>4`6tPY4R*M(_N$1juVl0h1tv@SrCL;CpZ*bW+N?JImC zO+15_;Casb#pQ-}SkuGgMU9&fDQ}o^DZ|7o#ht?ZuTnP?kvr%|b8j(qbV}5VEgH#Mfvg;U- zwjGAAHA);$z6w{%F#KXd%N%pU-d86DUXaDZ!pZD~q zL&9s)mYDrtZh=q`n)`<|5>~B=?ot$lpQ2L)tucer1Tg)rF*`wUFJFDjJ)15HI74o^ zJ03MwqQrbmuqz_QYK#*~C}w$JnzZ+3`sXh zh&n+k6a&&;|ErLo&?T#k0;UG4GAfO%W_Nj3QP*ZC3a(=0S)EUJkO9&fGjb%wqsFg+ z=@Ras(`Qw?S9omn`=uGYQukX9+F9`>` z^vb1HzbKT_ckUu1x+0c*)IqAbq05OAGuuqib%tV3Qlt@!Q4AqzO@?fc$y7E-H0>bI zm&WudGJyzD)1cZT+9QH`6}mPki3e3EHL7iFb|BWE?Q(Aiei`*Dh>(xUK zyVM44VRO|70&j>xJ+hy~%aK)a*rkPs6E8fOB*+eImT!uz5rWK8SOI-N4PTQI*(|?7 zYC^JoPZ4azdL$XTcEa#vHIfV$p2*4iv$?SBtDYDq4mMbjkG&a+9SEs3hAPM1inDXH$|6x0?f6mwTY>YhrQ=C>A)aa;VtnPytyCoB*xhgT)|~=7MVuQ2tg=&z7v=RZoZ5HWcS09h%|v zp!;z}o(L(<$|HdoSbb2n3#b6NS_lq3W-Epa!w!cp;&|zl^n+{1Bhu%@X2ya@Ul&~( zSnW3GaUEjaC~@)>>edi9Pz%|~Z1Twh`iIAEeUYmtB-Cx_Eso&e*u}PAPVN0=khx#? zb2ZXf@bwxp>~hsNy&gpriGW82OR86ets1BBkxdWU+P=4`@p0T_d2pFCec5-+=O1rL zmXw8=iGd6gW~c%xj_(rVJ2d%6LCJuvee5GIGF>+=M@E>NDGV!L(1_v}a3S zS$>7b@Xk2zD)VRL=P@2eF#WupEG8*V99`UPg7O@Sg^)oe6^q)8y0C0II~a1NHVBX- zMlTg`ns*~KjX4C#+L}V%0OJLP>~6H8}dpO@~MHsvtCL6t43 z5wG_%u#_!XQtpQTmMqy8)(m7pYv!d*NdaM~9(u_V`vT5Zbd6ySZ2^(Rd(LmCr0>VV>9m1p3lg4b zP&tOPr=`?m5F2hI-irJsBsC87&yCm*btO$yW#wyniG3LR2#_myHo1IQF{nmTu zgZD~ zP47O)Yk*1sQmLpG^g@~88ge-(`Q3q^-C&1(cSg0woD=uDU*aVNHPXqRJ@SbfcbpYX zGvA$aN@QHpd{92+ze(a)87@xjTP-qK=4_?dbcAdydfqXCQd0x!0E6xYv%2YIK?i*x zFpcRBKjzv)7gNY)i2vKT*0$k|c2*ErHYsUq=KkY&codv?HDiHC0V$DjQfiQ=D3pTG z0BYB2m6-dlFEcXic0oyyzB*_?kRZA(*LrBS$nJ&ahcyIs(@CIM-p7=PSG#eK^;!ih z&OjPHP8qkY`XY`OWu<@bUsXnHGAsM9DJ0*C9VACgEJr29g1qq_sQiHO<1d(7nLc$Fr@#O4}86!oS|P+6_xnk1J2Hx8<1brd76!D`lG`kNELEtFUx2 zrosPXaSvSrN43hm;nxE2TBt0WRc&W7Fnd735s0vCQ^s)lQJW6_^M_wH0%w;$?;)w< z7AAJ$Y06a-m^D%C8H${wVmpM3Pw@&=$3%tyM$WH0flT z%sVz#fzX>wfE@uihLsvQ z^rqk9)$^GPTi*SV5mF-2#(yA}xIxN^=RA*1Ak|H=ofNr2#ny;#LY3D!U_I4`Bbwy} zG+Lo#L7%h^2>uPR#*(R*7ohT*PLk+WEm+}wRf;txT6LyhFWsbA9$2T23(oR<&ez0; zC3$x%s^*_16;Kch+#N|%%cp|cY!dJq;;r?(PN>j7DnO+Ij9-_jZqk|2m|oQ%2^w#8 z!ZNRB-UYgidElDHH0p1`Z`&z7sKR;{q<{ECuWLK(f;E{q6GI-6*Nm7GgR~LEz54ko z#tcWUf5-hd=H(@C$r{T77w?Lh8^g}i!=W%pl4Gj#z_|>vkMs?4P;#DL1|(CseFbDC z6VLIUA)EYAo zneuMKyBVT*L8jk6-d!d|a%fs*)GleKqKZlu4brQFV9UKH@EF15UOc;*w}M^GT$L8_ z`}OR2_P~sdsCd-~j6PxqY;c<<4Y>Tk%{&c;i=}bmG0j4POl{D9X?G~*I_Q%g3^c1q zFZG&n!VcXsW)s$8Xv~`(c+e;N;{o$D7z_C$;Pis@10>0~;(tV{8QH+G9EmUr=w#Pn z7f6q04n^QOtrB%u?eYG->Cd_vQ>ZbHV!^NIrq*5k};X= zH?i?W6bp6fxl}CDq8PkqSEXy2ZR#W*B&FqtZTIK^lH+|cy}HbW`KMUaZ{2X=d1AC6WiGliWYt)K0w?yvS`B#j{SXd?OCiR~)WLz^rMl;2prN}At&}9VHE^1oX1HRyW!H;BB)L!X^ zFXo1C;y1~G43}OpxoYmgxyxNwP6nZBRWUFyT$f@6qCvrI*VM!OHl`K)!dm5d-*rKI z6%W8kawohAuEP{kR9Hm1+%_g-YL~FrORGFVo_V*0lm!EWgJ>`B$`^qgNP)++bM8R# zFRDA%dTo-}Vv#su%4D46urmS_4xg4W_1&#&UeP6X+O&qFCivM$u|P;xij3TQ!IIR` zO~N?Q!6>b=I^^t}?7%KzYs`ATTl0EB%X6>iUo=Plr6+KC@Lx2i{--r7cq0+S=^cu;WOsY(JWGBT!MPD8jTgq!=ak+fct3uih@iVj@Jus%j8hNLx zBs?yn6nOEw=rhWjUXbpLmEfgbRr9gL3(3@zXN}Z=LQUu~R|i!OkFQhoJ!QiqfAcNe zIt5aCh}(b0{_3}}|M6cx{q^Gil&+-MEy9{DN{uqf~C|U zX`-Nj2ck0|THnXNAio#fBs|VLrZ`1oLXA$OW1k5&@pIHiAl6(MblaV2}kTf)0z+L#XEVD#OHmhcDg7X109*PmKq>K6X- zz``My@`Zn_gAdqSMqP-(ujo|tMHceYA@`zGl0nj8fJi4BzF45z<$27vK!wdNrS{2w zX&1bXIa-eIpi4=F&zmV!o!3K2EKsQSNgvB|)sXdxDtd(-a3zg3Lw>jtDTj+i|;=EmmB<^hf4A46l~jJ25sauq~89*%3NiFs{`uJ`k1a z_Yjh(2Bqtu*8XY$Fv|BST7$4w?f-$pGC1b$|I$VCP~%$?fVL1Z#rZu#rp3G38+8IY zAho5Cch4L7dE2F}e9h6IV&Lsa6Fis(`V}bb*)A<~XyKC4C~&ka#sR}|7GS9AOfZnLhQXv zg=$h>=5Asx1DO-ge=K;r zYl8}W>x z43*xs0uaKg3o7*Q0KLw-pnJiYVH3CfdHD`o)$uPcpRg^vJHPf&EO2`lQL($i+ue36k_9je)JB&DqbODFj6n~4+hB$%8v-VTJdmJDhN3|{ zKkQN`D43PTta95a+Bl`%Z40kTl@h6E1x%fv>ZeuTm#6zX& zoe_{iqS!7S2)zNd7x;6#N5QNX&r@tIua~~TtP%`_ehd@X1kZl3=%?7$m~6?AqE3)H zqme%L{rZK=$ud7so%cMiI_g^>s}`iIlLdzsX-XI6^V=2nc=7E6Av28G0a2VFB6e3RiF3dhsN1hp8&W>W6C4J1rwAh(pN8ZDL^+Cm-2{_e7Xvp z9_RQ+1R7mfCLBb)PxRZIl{UF*m5s9>wh!k!qzGh<~?2g&Z_BhB5D(NXdI%=+sn}vpCr;*=_NVi{Z0NcnerQkh^fH2pVdZ z_q*Y3d1|D&Tbw#Cmu?GLJnWLdWHO*VEZpU*Ria`<4Zj1T1|tlcT8L@K^V;=yavs6P zYk6Kp*w{z7LW~=np5FQS_sxq=oEA*@?giU1pa~#rv8=hQ&J|yRZ5w;8gV~U=-(EA={EZL7p@?8(51{tb)2XM!2X%{u9^MfGs;4>R(Z@lAHaV! zNavLvbc}S6xM`T)KJts_l85fSbhcNv*HhxO(I?rz&HCb(i~wA*@gIIbYA2I+lXcBG ziakw{69&dHwD~#ieel4NBrwOQ*GoNDPp_UHFK-Gt8njai#gYZ|PEZ_b_HU6l1hpu& zek~!(1Q<+O9;}m$q=qK(>V5m?c6d~~wTqernv0~@D?4z>Wwwtwt<0E;-;(J0Y2;_2 zJ)!lod>Fd4%JpAe{%`mN*ixGe*GITz3$>YbmSgN8g2PeH7KW4Ma9&6J&SgiI{LF}` zXZK5F|b2-ezUK(2-^&}Mlt z@2Yf2(M=oDrM1eG$ije~5RPb*WlQ>Cp^pwyT^O*F3-Q$g39kcD2i5D?`|>JKWYUMS zaRYMkAZ&4%-wCzI%gA+*SRaHaITSCa>GOSdE2_hhXr^eO?w=Xv$W&#SQNd1=h)=@1}pmln;!MS>2Z{rb};Ij$4*-dOjFje5Jj z!_NHBd`ZV8(&@b3Fj8p(xEF;Im+2~xv(j4f$p5*ZFA}+BYv{7TnkcO@ULGe(6f{N` zLqXkjX|)?}X_b$qx905&PM^^dGx%>U?_G$24)^6e3m^5w-?%v*sJyZ9aa{23^6NW4 z@5$@~q?bkt*y4pZnPose-68BCy@CwCR#~^OYJQUB{v<|=~hkyEw$viXRUe0 z4Hxs`#N(NT#GyWB9Sdw1Btd?be8wICclI$=3BAjs!$V_~b0SUh9#HK?x=769(iF0r zd@s^_++es%_rEQF|7@$`IW*^*kZA+!nFfCg5k6*EuDG)1C8cRbf2Y*88p9nAz74=a)WVsbLF;b^pf z?Z=v$&%5vVy4Q8D{4lK5eoAB7cAM-hLBHO_+cPWUQ;zP`SKH@g&?cRkzcx!eq-Qm_WKSxsUoeDgwX-p}rWB>cS z0O}Hz#qQ;i6;gD!;_|mRa**E|b6T9njGWa;lXw@&4F)|Vo7^iP)u;o;)3h+1)UaHx zqX%U6L!Pa6W5j-#eZGne8;@89RP)7*nn_0R#IFiI3X?X=7mtyJ=EQ-AEE8C*r`SY_ zBv7&UJW=8t{L_`*ElQ+^o`CUWtZuUgD`VZ~=wlCD!%qM2D%$Q4V%awOQ9cG$R1EKw zOkK)5=iSL-G()SzG>ID38n%(%@2TwA=ITNxG zK-MEv`MmrkkNI{9b%KjBooExkkY6Dka=9wnHa!9O3!F9<_2aVb2gg&K;KYtx3&jcP zAoz72M zuD2$jbjoC?s-{@b?Ab@fVuDxyoMR*_>OzPn7o=)8%{o1Osi!8JZuW#gH^>cC_&)*P zTdKM(1jKBQk>03hpu&>j-R7gYF2&-@r}Xl`xadL83jYUFGi1Xq{d4*hC;Xd3^25MI zr&V`AioRBj*&GG*9$!r7vBO-*TKt$zcHXdSjt1`g_v7S02r}ZQtfzDtDRAPPVhdDz zy@J)e9--l9Ys@9cg~#m*X?4gZ_veB*(Mh(6-9Tr64;H^B1o0su$phfyu|E^cGoO;FBm_kR1Rps%z|sg;A}Jf0NT3mnwfYTO>Ck z=j1oPb(L(G3<64{W5C-eHk%?_A+064Rgn~O({ne3+Y0C=c`I82Oj(eZi4oXmPWp-@mG}Pp4kE@Hd0OM>i;*TA^2w<$oHvRR(*EJysYZ~r&AIde zNu5l}OhA%LvEb#^Qn8iu&q@nn0gA40&3W;5KIU{IgKB~tMP{_hmxAS&7wEH(vZC@K z(GG=+>|&x{*N?Wt5#jIl@ghND{YMj1w=oi%rn3rPxf0Y@uRP zB>Sa@>HE^Nu8$&n>Bn=g3Cn@MYN=@0B|j{i&JXLAriGy{gVtkZMD=us<9h^&+;Ki~ zr(-{V^QC`3`ph`FK0Pb@G5O5p^W}HhYyyJ>id{*O{ecZ5c`COy1_@Fwbo)mL47EZhpK52rILlg_tqm@+b?&y=Bnn?~X=a?J2TXimGnqJUoK=you2*P$G&>k0azMc_iD!zc#y5p@XES z^X}Cl58P7K$K0!;wyB>BGTaS!CR9VR>V6+Ft028@?SSFv4e6`ad2g zE4jhPiN~5;6ZmYR*o_oP9#KP1eixD$4fI8ti@wW5NsxqaU3gHX>2T|?U$0xyOPqjV zMSJ$Tq;{bG;&dZ0)aw#zNftL?II&$hXabB~6bri91yt-^MZda~$qegcZt*jLBe5UW zXI=DR{^ps-fpLB}zn_=t*WuC3uM_r4@hnD^hFrQ8H$tl2E{f|EI@c23(V$M|HsoXU zJ1kbo2`?rl!?ZQL?;63WSWOUj0x>`tqIAN49Kvy|NVP{Hd&kq zD^l9UsebSG&wVb3r44;0XhY;$kE3eM29IY_g9^j@SYL#7xCfysw}!^ESO-u;C-EM- zjoSmn0VZRa56j2K2_MT-+^_w9%dhxG0J#LZEF&wp4MooD5F*=6pp{0k8z{1#iXCvT zQtknV&wlq*sBhZju`H%W_<&wbC3?4q^hQpsS>YhVG)@F?y**^GCEDGHi>j}F<2+gG z#NPTslb+6`*ew*iiLvO3FQ9jLq%k_>T3(azqFi%T`oIggG*y^5-Ks1FH6*-f!kiu) z`rENHidW!Z*+Puv{a87_uH8IW@^eL0W`TtFxU`FDB>IfV8sR#2V@MB;GgMsN8uL~L z>D&eN>Y9+Vs`IJ_Rkff4it;DGX4p2cywb~dKVTU)4}{A@0*wZw@x2XIB-e=zNP~$1 zsi0WUBQK$1>wTYlr?VMyO(By9lnqeBhwAld{5>kY@I8hNkk}4G2rbHsK6~EPv*>F5 z(!DdLUWBsaPUf25KJ~CmibsuK8iSl_TII0I$*@#Co3EXLV;e-qo^sd|Czc^HhGltY z$-k&{pZ9U6K~+V6Bq;>-5`&&}t{Q7VL9!MzsZn4W>7}q9KlvAQ&2o?ZN7#v2r#9CO4Tt88t zoM1tz{_vC7FBr#~w!giTPbwyp^Cl+!2*m1y|W`|x<2hY?0UZzqdMiW5&+cAMZkhhl+yH6Q*1(FtZPX>!KmJiSO~59n$M3w3%p02S($9No z7&<}G-`dG*2m*Q&-U{!l(oUv6YJ!o9ptBCG697-Mui%8%cfKZQ(FPj_l-u9FS4Ik* zctB}18BnSy7FbWpsn}akn{iFr3vzZVCf9p!ARDJ-d9IxN$p6;7Wo+u4q>$&Kjr8K} z{3c{`PqIE}Yj?LUG~fm`^^%!T_qHV*BSB zLNuCIcAE@EgyKZqieZ;jzaIJ)f2l}wJ-l3uL+53F1Mt$P2h~9e@2BCglw$hl9+B3= zF?%6s%pM*Otj286P6xkne9hlXGa8w{+xd%MldVo{WT3EORJgC0V!?Hv4|yZ1Ub-n{ zI5dge4$p%GT#!hYcS6x<3AqdD+wU7H&r>3sXrS`C8kOvdL|rGQu4B)7cJP5(M}fd7 zn3Y1-%+uV0}Nri|I}I1h5NEAO}s?hZ6}uyM&HaN>ZUg)k5!{ z)GuwL(}ULYyJN79y)_17gjs?c{KV)3fvqukQ(e^Z;HA`7<>>pildYptxZcgrB}OkyUBjJWxre9C{5L~dC~PK+qXe;s8Kw^D36{<5*xF^&Lm`4GkG zt6-@NvP$wqL65kTF8_Yxm--fNaxZpAf%@~}!3lE}at4ll@nNyG?_@{+#at@fX~TXN zWWiSmP-m!%P5_o-6ngC^HDZJ0ub}{|P1eaAmmU?^*%nxQrf`B=cfYPB`<}CGC&n5K zKdI^bp?L}{mj#XUj;6#ya%^8@)%-ri_7V*)Z47JX zHl|*%-tUN@lj#8-{z~r$G^Qdn0~2+N5}4lB`^AS9Ksk@*kphpGsczC(>UxCWUXA!E z-K9t-@gQ7p4^(o4+v|p&g_e9iOHO^b|!|t3wcK+z5b|Gou@7g z*e5$O{m`rx-rGHn`Idy+0~40IZ(LZh)TkYfFJ;)N<{tf5*T+umBeCG+FQ=D0Pp@_ZMUW&O>ZKJ)8axJM8q||0_uQ?p6UhnJ!E2Poc+I9A0P>FI!7KEq zS=arIcBcO)@|9#8H#_6RsH!uuGi4MD*$Kr|ELK<)fTBwp)8u;2`=-~Uh~6l@j6yuy zsz3&gBGD>O&AKpM7_tJkDm#H`cRRm0vOKbr(WFFP0WX~{EP+XhOoArxp3O520d@Nf zNTa)ZTn1+}q((PthwV?FeiUV%Y3;OAUkh2){p2w?-x^q*G--_I_G^n zs*QO<8tDz>l-T@?BL|lk-C|3pzqr|f8-X~7%vHnRk_oT{J#kE*Ip*30 z!&E-ZO%4Ah1Tf}(ZViCU@0p1G;=*TvKE|P@YeDEWlIg@TL0}#jHA&q`v9N;9qhgN+ z9p^og-BGlAAM?#6HzCcSCS=f~OtFQO1HnL1z-O`*(d{8k{2cfGh(caDL=E>!FU%;F z6i4oiY6U{aV)}OYT94b|j?A>OrvFC|4rbd9g90}La>~oC)ZD$o#eg`m_sxPc=Cm?d zuvNA6Z>~uVq@t(T%lt>O%&>BpbEGjfG?u1khTRC6K%>G9N)AKLD-Bo4|MM&6w0&H_ z;l%D03nYJ-qKO_l4Nh2#L{;IM482PDty#%}oEaU$Z2GjaLkP9WO9j{b^vZ?TLK5US zgKd*lvej;yHB+*|8V#$er{e+%TlrMHmO1Rb1UVm3MjXel9=nONa6^un2^wqn6%5#R zpkVf<)s%4%dhYT3EE#ff+SG604jerQB~k2Jimaw$v8O=q63BfuZL$RUk%@HAj(X}= zbn&RGCvKC|I5R0ZI_v-V839o|{DVrObK+S^jR_$3P%Mb=6j8AyKvUfTG}Wy!8%Vce zms>B?@;ncH99rOZL%z)8Bf;R*!~7mPGrEypF}aRz5-xpr$mN!}Lb3GSm6O5!uec^1 z^j)Ps7l68OOL+~dl`%V}J7kox0y1O0`zFU60VGb4p=$oLeOH7LGC%)l-Y!x)RzTT_ zJtOB#U{p`BfUAR4Y@Db^j6Qa~7yh&uXC25)dIa*VTReOHbsd&0$lqk#Bi}4 z{3iCml?ljFdCKL?LM-eZ0F{Pb>b;46giGsOP5nD@Fee|Vv*4RDRT;WjWpVtd+%0z zW=89+OE-;72d7BRj80}#*j^CLtqWQemL%B`sJSON=8579*?}6&pF=@~Oi)l5a>1CO zR=LBYMy2Tvzi8(`(3?>+miZX-m|ekB=d$K~-;wDodn$owop|mTW0uF5vHY?_>Az|a8v&D$dHq*p9XB)M z#BLQp(UrIvc&4c~n@lBZHB=OdxthV{k4yIZ6TMXvlCBYEL0*k2@gcAomnBdLUw9$r7}?FQifWW-mF22RRk_bWWp}t%eez2+;N$CoL^@n6eYSAW9VkWh zcdvze?KE7yibWey|6;^?wMF|J`OSmI+5fw0mYYbq`cVWV&<^so?4t-&v)ti<914Zf z>ghNN;rX84Zc`Aa17=EXo<7lI2ll+OI3#fJA-2ySaL5RL!LQsxJs$Y@hK%o7?tX4b^ z>z_B8Vk;W7(Auw_-L?C%$(0Z|^N^{nD~;HhkmM zd#y2LfqMh+c=e)>zJ>|?X-sv2b#2fl_tuz{$i7Gj za0^edI2AtWnGn@8_gYXr@3wr%B{%$-Z*@o&wK}L%u~$(Aj^;kH%kz=!%#>5!`TXPH zRz@I|%qy4f1(4;qf2z!N&*z_Vx5IGB%_P5=A-`c`{B}A2FF*Qvq!C2_v~Q=5lsj=W zq1gl$hbb2F{i~_i9Q7vmeaa_3%XxLeGOsMJ%M0%M91Tj5#0Nj|siRxx+z-qGzrfJU zlOWq|kbJ>o9YQ>t?Upm6kKa#gqYL>06S-{Au)T$vjNZ7axd7hjrmsSSbpJ8(x74kuBl)R4uS0sq2nD)Hlt+x z2I93Xa6-wX%lGRxnOE04ZJ@zIz5HR|>P11i&5~;jC^^izEx!aIhZU0_LdYqX?gVMI zBEJSzu`EYj=+G6lNw;u?hsn(j+~EWdO1kyK*cnD6690iuBFT1QBT{W*M0QgwsGb&5 zv8jHY$~IZIBH2}6l&Fm^3r2tcSs%@3igbU-aiDtx%YwB^&BtWGy;8A_#jM=Js)p%_ zqT@=z5U_Yep1&yg8MIOC>Myek!BF8_>n0GQ)#8{Yzm^VA_eNKp(q?O$B`o0k{ z-#J754f$v?d1%rxH!1cyMcOgZZ9h3qU*XlLGT26X_zO9K%|2cHYr>oCF7}TfvjejH zKs<#!kw}wWHNPU@z``XvfEeYfv|G^w1S-4vb@W!hEY)Ftxw!h>mzP2QKD8us)P*;g zI8id$CmCsHt19^`BaIC58|f}#KKQJ#(~gIW+^q~Qnhf%W#ytF zmr6K(NVrcD7ug%N?+yMMJ9VaoK7Q@t*LFCk&U9WUThG_imzWnZaS1*+uOow4D0%8* zy686MfomFr6uf;<4XagV(}e+jzJnfXgP`iV1?pfie`_cx&PS`<6LyUB&#juTG30)x zF^7UyLSEJ~Q3FOU9i(S(5756By*8fgm=zcKU87CdGNi(-qxf7%1QxnsAgJQ2z zqz$X95ieS0y=t4^?&vnQS&s4qknai9YMK^hSNJ*>>rb!{A0rZNveVNmBD6{z8*|hU zz8#@^zAe8Wnm9Gp?<0QM{LHB-g8Es#u+-{S*Um8Lm83BR^fFPwtlAmv5dYQJZ{vW8 zQYI%zg)5M@&N7_rqVwooQ|lpxFiy0^O|zdyR)Jv`9BNmI?m?nFlG1m9d-hG^sy#6F zTJve?AQn_Kwd2KBLp0iPlvs{YJp@^1>-RfE|NE8gxrCxQ4 zdR{;9J#Td_@@tLBC6Cph6V1knx`g#nDKIfF2A;$semw6u-Ap$GUGq8Ycx@Y7Ve*Dv z9`|X@2R$ozMwtBWME`NJoEuD>Hy?liDn`L%BgH0DWSv3xSzj{O9a`^OJEKMn=7uqF z4!!AJANSKXZ@^kqnBV8<{r~CsL|{%cWPzV&FI0v=-XwG3y`A(>ga&C@kdY;gLG!6W zGg~zuv$3%NQGv>g&W7#KP1U^YL(9o=s;5X9sJrQnmxQ$_i#7nUE`|wCyST^W&9e*J!Ief4^6UTjz2?7u`rDuV=WnDdDRwzU;w)kOmE$&l z@f&XBubBMUJWK0yl}%wG7i-1ji$0y?T)@i7`Q*wsdO~|6ZpCy>KN@sN{Hb!|lv=@F zUMWm}D-=5doA@7#<3#0=NmI+Gb_g^5Dij0KbD}a>K3x*`%lDGDRcJw7SK0Co8?+%k9?5PTp>lG1L|EsiioDBS&j^f_dIdw zgrrsRymmLO(qW_LOAy(Ho|pGKYD>fWiLZGcj?;z=EF=gWjXbFC4hQwtW+BdpwghTA zgeWj^gIOio3Bd!LGL?tdLqhljbhM$PxOTUry4Sp2Zqug8b+_yO&bYpL?{MD_$w_YO z8|R&Ok=rKgo68h?i6Si+qfC`l%}1i%Bj9~Iq{LVy%)<@L5FH?*r=PTRGEbpCA>IGB z{1L4|&Ip}EQ|bMg3=6Z7omG?I^F-A{uO@v=c_fMlYx+TQWLX2yGtWKLEmrcnC=qz{)PquWTMH@=AGaY$2kk**$kh62z zy$8hAZV!BlW+rEQp|aS?60jOdUc(PA{)4@{z{UqY`r#Y4 z?c4>N-eBi_0bL@JE`7~7F*){&E1!@f+!jSn>{aP7nZ~qGEaU^9rD9PyLKlWv2}T*Q^B>Ap(!@JhX zw}4`Hk?&fb26y`uD}e(bPn5{lV97zF*E5+`sgqojp7(tqubO`}2-WL)5qscV9~Pm#h&Hpgyn$E)wlt4BDzL0^0C!Zh=cv(rvxECi9e!dF3nxG%Io z-3WDBz3vy}nD2dGSsDfTmMj8?(*C#-vhg3PqBrSoE>9s)r)lWEXwP z8K9x65g$~Yl^%4hB`p)-pEc=JtD$0Ym*aOVLnZ55-#+{IMp*p5`}eoVX>Nm(^A5M< zt_fhSQ0&JPxk$xsnlV6b#he2P&uZ|x?c){B&JEw?*%H&gDDL}>|FCY+iXUHyX^Bap zmT3Q`ec>NZfAtE~_HW|%MRqD4%Xi4CsN4lT-q)C3$SN=3m(ba&LSBLIBiS1AkS?Jc z>6AHrAlHTrpV@S#2(=4VOn!MD89vvA6@#ri8dT<0EQlBM()Ye}O_~&PLtZVYqHg&# zDM}TesG8*c6o!A+3%>*#b6nmB; zr>NLV;%-HY=LRTv#1(fvuS2+-e_Fi#tCb5Arw&2R8fs!yLke;ybA^{P1F|iUjuT@> zY0Q17>~3TBf=hSUC5>6-*2$o+H&@*Xajy}+Dl?;j#WS+ch85TZp~O09jJnFQc^aeoc+x+!#R~JnI4(}@I@J29(_r?B50Lo4-H~CmEM*A)VuJ;-(LUH<%R1P6)rd;D5i4NgB~^f zZm>Buw5E|aATxM^hg~ovkjCVGQ>#23T<)t+n!rhL&Wt8G-dQAC8gqKO;o?HKLbqb- z22&(T5_HqWRE^v&?qtVdZWMrQ*!<4shd)*rEmQWi|L=DsZLIl_6Jw>^1S@$I3lpSl zDi+;Yn@?)k%jC5TW={;@*bYoDWo0W46>!0T9ue2TmuW!fRXDwr3j8W*FIUV2UFEHYw za(`N!2=ls6dXz#L)JnngfRQ6z z^f_-Nuc~%?aYV6!TovU=Zn*=Oo)}5(@tV(M`K~*{4g#%1&WuL-#kIz}R|@JO6TDG6 z~{Dd@IQ15yNpNHvA4tHI62>Jf9y7( z!S*lXaHnN7jA3mk_LEB;?nZ=Eef1mX$yz5yNTCTrGAVWo1-%e^Jg79PLb@`hLfSGf zA@BinEPU0pdU^n;@j96@ra5G90G>+-+~hNH0}@AcaOmIhUCIp||KpoA@J~i`NW6dh z4r!W9?wfSNb&74L$Ym<_JgtM-eGlDFvgyq;mx17TLe!^`Ilx+21*KP-xCpEn%Z6a` zO|L?r>nu>?dX`C*^hr@LB1xi2m0TnpbOR}bIp0vUATfF79UCpT{op-r${f9r4ziWh8z;CM;lPykcfGSO}zk_1)D6$R6 z$6+zNTagTyN2P-Scch`kjJyi}K~KolL%O)9^hVFN5dAtCiK;Oa6(?F3l^G2ZDM$j0 z>A&wg%sh=3!>Z=Czzm7jM^l_V(=wJ3*O&;e7(0^CpH z4ZFn4FZw=Y%7FaSVc;?EU#?&>zu!^&ao^1eG}l&f+IXb}wG`yCGK_BLf(Dsn*EVsz zAV*yteqWAkGlTBy`9Lz#O|N5Dh2_)Wu^aSA7T~$1q9RGE9}aOaIId&!A-^2$N&$UD zk*98DabUJPLb&3<;xIRU>M9dRaWX3J{PNSBrcmSfljE1WiIluGajMfMqg4&X9;C>A zD)xfvHd&lUUtkv-I2yN5P2X(&O5#@rel+A__{13bx}XM+P3}v14}5M*o4s3O(4B`e z3Pmu(>Ihpy9{3an=g}FmXWmJ$jPIq(cnyM8lkYFg|JrJZMI8}r^Q)r=qqDrymEHEy zYOx>OUINhsH#Y9LcIO|RZv>G07f(d z(4l(+=jTtNFfgUp8C~wt6AC#u5<|{^54{B0IkO>srbN73(Id};VgaC{Mvc-MA+`hK z6s=DphDmuR)1`=)!#!ADP$yd@XaT~cLH8oj^H3}xSaRgOd!9w2L@0Y&V&fp@2yDDW z32xa24o;?MQdnLUFVKjj#`iW-0>WWg12ZupQwgPOIc#c50pIC1-6y`#D5g)8I4=s_qVh!YJ$R{5|CR#@k% zYeE`SwLyDC#vCg2&RhX*nI>0cRUCFfb!}t=&>*c$t{Nx5c5~EN1vS%U&KvgZ1G#_u z#gmXQBX|`5qsbz>xq-)tb05x_fTxyXK@GPG#c$7v9z)D&zj9cd#5=6ksAHiy)JyCrJ|N?#Ph+NlHRYj@Bu&XBm^i=IaWXBL!*b*}>o!iPSO z5njsok>W5;)Fa+0y2?uwfKiaI7UcmfwoCW$*Rf5m2EW@WveJFORD)^2x$0{HyJ4<_ zH*9q+=Wq3_ov~Vk%j9fHy8m$K7NLDvSnZ1hv+f+nF1KN8_P2UYPcmAb_*KD2$*>Ej z2Ap`QoMmEp)>CXEMG~mk9dt(EUn+cIbv)&{AW2Z1N&q+8}ldxfGDyi`4|Rq)Sm9Qpm>|f?oM{=_iUi zbL|8+YvA*Sxr^CpyP)Tt-+pzit9cUdTT(J$A>|k8LY_k^o(3t2kmUKSX9obtz^}ae z|E2sIaCIlMnp&JkYT@2D+K0GZU}W zq7xj@9~pph=~a^l$Vma7&xJNtP0p2GnT>m0`g2|693QB1fI7#pHROU@tC`i4@wcdh z0_;_j>*yiax6ZoRhPs#dvti3Bpjt-Vv6=He7!UEc6MN+>5P$c9Yi_`;$~^}P>^p>) zcs)!?RBOyx=_T2%&{}r0`wB>2-$MFDNxasWBV-4i8wqk!LA`)QD2ffqrn|!P)n$s3 zsN;^KcpjgOx&ueNg)TR|iP0XFx#-T2RZcu$L6-KY0c$hGrcxw@irvrCK^|o~q-~>$ zQL5iDNFG?6+syFY=7jdx}Z*I+gS6H?gV^tJQYM%D~c$Pj+ z>`}3>^ubb>Y~SqGAp4zk+E!%*Q|;27WXQAj0AplPquXf$80+dVvND zrt?7KDm$PF`OGjKu0yyYD330WY;uLW(AS2gBv4lGvns4cwT*2TcL+5ajUnF&4NV{~Df+dTBmz?TFZR9*7Z)Cp14rP{5ZJfQXMVzA+~GGl5!gHcGxn~1{9 zy*>sDQ!4zsfG6#J$oq`k7^VLWum}q$_U~7E$IF}PQH3g2!_9^XjSaWFfTHg&SFJRs z+k91+!ijB-1yaq#=y;Id!?fHS_dcf5yAKF~G;t!NZqa+>bYYu$NMn{6z2376YTb%N zx+xRlf3Xb+FZ#=28uU=QTZX_`^OycVBpmy^*UY22M=_C8QQ7`!49LnOnhk_@F?GT+ zuTG{9lKElUt9-<72x<!h(64KZQztlYfTUrA;$w4( z5~m$jEQpSvmsS^6vAAK^o2?2@;9Hyr&+N3eG^roi!62j;Oa`?m6ZyuLHdP z94FMjGymDY&i)(YD(XKjWPFDlerr}yH%wMhA5m-*Mb1#MSW<>%L6$UXz0@O+=5H3-z%eEWd-uu6K5`9w7oF$fWer}ThK>!HCr z45tU(p#?1EJR_>InSU)Jd%2;?iCw=RnV{+@#Wql+mWmySER@}M-{zUF zMp7wMD8Qu5Aw?N4kKW|7V)8*nlkl-3%X8)ANB*~r9V?EaVx0P|Rx*UE@J zsz&-cy-PY6oyczI?@{$bnd)(}PjWS&IBb{n7A&*yn(@B!R(6(EFff^@SlRG6FS9($ zIrp0{T=6#AjGvr)=K{%a;y}`V6Dv?eu>};#rDE5AKL_TnnmwvE*{yj?mMlqhyF^O) z)ox39x7;p41+!MYQmIMe-F3g|^(f-16kNgRP(^x5B5lfmqJA^&*zDSG-;8jZ+B$uR_Os_uiGeUESpd`v)$9g{FMKLp zCb$h1v!!0`wq|PD2}mP=8spP8XmJD5^?zS`*>7Z~GW>%|qMJ-=O!{jN#qOd=5fyt# zm@P&2(T9Q>@#dMA7Yy=C^xHdu4Y|{+6mn1M=psH|)vP=x((*5`UBY6LrP@CuXGQ_< zP+;v0ys|@hIjGRTEnvN0jd<6aB{FS!2KML~TfX;A_i*YN=bZ^m^ShSgYlO^m_EZAV zIHY?TzDJw1v-Qk;!ne$flCs&R_AHTAOULt zB$RNiTh9N?`!)}2kihAJZk#oWPTqEpHd(o~G7@jXi0S#a0?F_EA^ptLAR}PPdP*U` zjoWI-iG!0zP2jbUVk;<8ibGo((+m{NnnHDl5c8uB35#W$zphp8RB5*Gsz9xxc2Qx# zwmI9_mN3k)sS{KxpcDmD$2A!9z8aP6TFsV6b}}oxx3SyiG(yAz@6({0&%j}ZB`S8C zOWOlyqX#fk6x+V%{>LZb=5t~$)dNn991An(I(23=2CFr7P-SFr^F0L-wkKqpUyry& zerL8}ORIcH=hDx;J9+JH8dQ`3$u=-xNc{lv#p)L6Ha}Y_q~6$D;|9wczrnT(Tz*mc z#cvqRO!?GjKP8Ra%#8DnxTMR(%v_>am{K=Wu{ziCn52-Kp_tO5bM2g?RW2?8vgeDQ zn`br);{|C@I=xf2QwFKM(~kyq2$zaVy^3KpYl%7X^(JFlF*c!9Cc;!ME_iQ17txZ0 zs5sGb5$Y%DT-(Lxf!D0wd(d0wn(7)axB{iuaFt>i*&b<#pyK_gyYncb5z>=&kSMAP zI|nRz6(F3uRrWZPlZhHPvPRpWH*9mkSo+2_iRL+;T+EOYyYnn$gq8&+6a8p}8tA>G zcE@fC54QX5?soB)e02Q1Dlb_zFSSZv1bCr++8b@)GwHfR4&J>=xxTVrz) zCXJpsKc(0Kiae%b@BHd^#52!E)WGpt%EN7}GkEAe9FfESB=k1a-R}js>;!_DTOg~8 zN1u}Z$d;eB{$mpLrzY1szdFi$=818J;fTIbr~Z7NUe7O@nG*GZ;nKfo44?m}+Rd=> zVIytWFkT61$zT72R(VEQ%-`^eyAWct*;IJ1;O z0WiZ=%P6#AjkMuU+9unh+U|jWkQUn9JhRo#3D{*y(i6TWJ6i@>82J`~M=j$wQS3&FBt!I4&oF;p4D)3qJ*@Y=I`7b|TeCHLNu98O zPJpQQ#9+nPe_SBLrOU^AZOckY)VD`QsUHUXgq(6>j}7FEkLv0+ifyIH1u8Z_yfD0s zr&Zn$)l|+uD}^4<;N_4Ff@A?!SZbR1=nTdL@#mqBLvPDZkYt$sqOQQO%OIV*pkA$M z@NX9%bM2DpT${t+N0koDNyDpwT=k;w%`mN6lTSDC2V_V^l?zPb5T3Zf43K;L)5?n= zDG8Ll>Z6`#lsX`c*v58E*C3bLGxt(24ayum0~%k=`uT=$+$TFS9f=^`Fh$~m7ds9c z*wkwwW!c1xF}zIDoo#jD8PWB-6aB}@^2xwWI=T}#Qfx9s)={x-vcu|+WRTo{)+dLr zLE?T~ytK)1syU%=D*X9YKm6A|D*K=0_V9_bX4Gvj{njRP^+yYWihCCA40r}_Yepav zz!}_IxbDDAr*ou9UcX4wHGO~_fcG5Su8$(S6#AsMk?7DvTO&Bu^_|r?8RHGL?i*~o z>Mg;^Z~j@Un_;vb@gMjkl58g)o2pIB+HQ)4q=-T)wnO;zYuBWIy~&*OZW12w{s?lR zaD{U(xJlS6KMOd2Qlxmb`qsSA2YXQ@z? zTCWHOEu%+?QiUR!vpt*Tns;V3k;u6NJ=PfWd~-dk|w6U$$Egi99J9# z3Db5t0fhQR%hTijBUa-l@|9%U=ZtNp&cu?GQS5FCfE2s?g*y|}?7{~VN z)SySfTW{x;oVAwbC6Fv`nId7FdzOK8wf4{R#uJEiVn3XPKxDHgYCfVrq}mPh6#61D zxn{RyouoW631%^QG!TQoJOrFEpwpM9Zlbf{f&x0-zgt}}t6ho(QRsXbcIlNPFLW(g z7X?Y7x-if%5qB#ZRJyPPMx)~u>(fRr`s{gk*rh{w8}?TO4F}{wMo%A*1RH)4NTjm2+xCvmF~}koA?g( zAB_hPR`;$A7g{z#<=tB|DdWiX+~fIKpa*kWUl9R;^HDzEB#K>2k=0ae7oDo!Db|WmHya>lE9oB{iJm=={I@2w<#880UcB< zYJU`oZt>%yF%E+%k9wNyd+)BE36h6CnpA!P)J~UrRrq4;j&)T~=!FEzl5OeqcXYWhknmW_H zJ<@YXUJTg?B&VP(0SaP8FjUluU8Xa=v))m^pTbIdde#?~$Z=!Zt+0)Gw3tC+psko|02;zn>j>0H#X$Otg0MgqXm~&B%C^nz1u2c<=+-;y{o#yC^&xL_Bn70} zK9bZj>635JXT5KboZv&=-GXXR{mEdCX+E7+D_`uL#Fz%P(dJ=;PwFy!j%j*HqXNa+ zvZ5UC8IJ~FLtLz(n03f;$zSt+|Gc{=+hpm%m@hofZj@)R7nwqVL3Lp{FG^UV(;%HMSSymCzf&-vzhjl-|=GAPuXHRr2schel(Q2ZvEc1HPhr#n*T=+*k5 zi@`wYEjnh$J7TIAIe^uF`vr3M*b@6#ej2l9Wd`@6sSS=^uB{p}xq6DOlt zH$qXvgf?6^)t@l+Yp*}=im^u6i<>pZt7n*K&RXOtf=iPB51}|aEr-J`!?Rw<%W{mr z{>#&gpLe!Z?4h_}l0(BrX}!8v4x)?RC&5Fg9IO2D<4Lg-77nj9dHfp;f5ZsKm|)A^=ixqv@18ub)QYU z6uCBPcUZsMF-^Mra#cB3_@DQzI1CSuA;SB2814aX+cKOpUG<0GrHd`-d9QTBpUH}s zfcA0aN4;yf6Qs09})!k{8 z{$pk5dC$UdADI)b@tOtKjb8uN>y;z7I_He>0AtN%_`@=MmZaK#evMGcACZ=zV}SzbrQosym<_mx*< z7iQO($j{961gG8t4h-R0c`3bwxgqQIsgfLxEtFPC#tJI-;p6f0GFCqB6rTnEW=k2y z$7r~&IV2neV~wWHi!@brII?epH_}S-qPBP;?`45(e$T5$|FwZ@C(Z?$C`GGkgY-^J z^@MyRf|BPm2Bt#-t*6)ozg~W88Y+|<<)zW>L)C^H^}Xf|m?791a_FdE=h1r}$nzK4 z**&=D2WArd<&3F(bVd%{5}7Sn8r!JY=>EktEDqJHOPI?HI)l5(nn>goEseS6i(66W zQxuQ*1Ag>TIOpI6@kah@nT{D}S$H&C%KuCry1gKOGXh$|Ll&Nu6uXQfi>bK2S@E;V zg8RI6^-Qna@q1c3(e`Xu+xa2~4?8J;_a|x#9<;}Qc7SBN@Yc85YKDs`7NpL$QE~e< z=32Jy;M#x{AXJBJ>PYN{bw*}}WMr+xBXfa!n(8>&9&J8LQ{f&NIe2PBSapXc5u+gm zDrdML&ql$JHG=EV!3ejVv*&^F5yQ}O7sbw2HjiPp?Gf?ux1djwZPrSql7r0?lQIPPaZA|0^ z@n1PYJ4U(&!e$$zkOmhdHou@(27M5brb>+1<6$bt{amtKwJ|D1x){iy>%zwbHoSuO z7Hi`KlnnFA)t%pZDa^9!OwRm!3dwWfRp*e^sb0>UEJ>$$mENL((P7Q0S1Gy-%=QlNyMD)&aL9 zpoXi8T;Z1wn!ni0j=ae=;q6eE?MRf2#D;LKK#tr6bK8(%Vy)v&RlLDYumI=Ae_Z+> zvebov0}>BI9MW`(g<|bBRNPKwvT%DemhoeFEKN|U)}>6W1g4fo#cg^4xi07yEt>%{ zRq&4^0Wi$haPSg;Yz#i>+&}qYLlyah2HVG6MI7dM(ry z6#K3QKH6sOIZ$yu1a@RKX@@@K4bpwl&DtA^0%kE|P|a-+EsdQEZ6Km^&Q*Hl539!p z$C<#|nenraqAf;5;yz~!*~KsT>cR=)lU63DhGHROypM`2V6s5>5ct~a!WtEsK}Jvc7H?CeM_S8dC^8g>C$98?QafC#blu8CMM^9n z$E$h*+OGk(MnyhIkOK)jDmKCx$StSDRMERU+BLcb;!}dru3@8r&|@$Sdo(J96%L3h zMniDy!nTHk;5e5Y)0~v0sjko12$VZ#M2CST`|QMaKfSsl_~U>&(O9Qo2Wfnuh}W6^Q-iS{z9)tK_1Wz7?{)kUDGcz&DumE z`mMdwx;!gA*1d8x3@h4p5~o+R@tAs>Z$9723hEc`e|MfFlmN!0|CKvXF zIjU0{MGeZUvd#!oJ*}x2vNph=x}iAYKDxQLV;XFS*b5ii=bU}M@T)&ev+d;K0}2;j zv^d(*vB*Hz37h4LnLSgp124-!YcpTEQ;-I9*89~NK4-}~)(D#*qa5ta7~#ZTbUZc^ z&c=oL{at_AG)Z7tdE7$X7LbMfw$CoS_kjlEArZSYid{{SRa9ILq=Jg(TvJufxvK>6 z%n8jP)X@bHX#?`v4}UyHaCkJicH`pF6Jv8w^R|DoZCa2c)7Gs@0;c%Q8mvODQY10Q zlvVV)R~85d8LD-+ypiD?ViX57T?8~V86<=lgK((bpY7u>9UeLT1Lj`9hl!HA?wb6? z>2JN}Gw8IwdhR`v`ofrw5-Zb@L$Scy9 zCjJpapSkcP2jnbKy!U2QBU1upB@19nYhdcaksH9kZ1t)NyXuQ|cDP;EtCNW#0NEhx zLz`p^#Fin6hS~P#rzw36+^5X}_=!l=i_+g2s&^8+8vW^Ewe5+J)5$cleOwW|+iUNuPn$*ZGiNw$U z3yML^@dCv-Mx;0}8cl5%j(0%o@Q`>%A;m(;QVtb|wIuD5bH3SuWs=q61o6_?Pc)y1 z`rYvOoNpV+_B5Y2D3_6QzUJdH-&+xb$GxOof>)i0I6Li)^)~&{*&q-c|7V@g_vYFL z$6VHeaD=;zzzc7bx5y#W7Ga_XsE063oA`N|MY6-cMcBw3EN{mhJDb< z7rCv+oIRWKY$4&YD+tH*V?<*mB+RnCn$(3JI; zGym-U-O(@_xv&Sy!D!S)EGT^zuqJ**nLlS&RJ7WKBiOxtdMBdNqczQewKHD$K-bg-x29e03&G15s%J ziYNeOK>lPDPACg-vV$@5X?NHR5BnlsW@X&B{^dK>8jJIBEbjd^q~s;iV6~ptQ7rV_ z9-!j#=~O0_IVx*Z6fxVxmzf&n!pVi=YRL`>h*0z^!K<6rs3=z$!kIdY<=0cNr8VB!Z#l))O9CGnYvdP zKXG(<-;vMNuwPfJi^EQWofmn{mFs$dcTa9Bv$ZrY$i{}lE>(x`61RzR=uf4eDmy3k zgSIxLZm>Qv6ZSj0ByuAfl~_QM4XBC_k#<_A(=~&V3R3Vno4j!pSho|kBOBIQbz+t+ z`LoL|M;xrwSet??5vuZGfA1z{wYx5U`mUf}&x_&fK*7fty^@|M zoQ!NX?SG1&jnkLND{NyVE{hV5s7EtU`qOh6H=}h$;Q)L$>IG%OpWqz1=d#6tLyqk(7X`^B# zn-v9Y5taTYCt^Q~L3IsWYrR*Wm(!OFo4K?L7Px%aJm>6Fey}(o`!xR@%Z_Gz_b)#s zhldNNx^Q{=HLLB{8H$B2nNz6i*{$mJ?3r2!GFHW^W17pdHIcej6}GrMkn7cHD%2dn z8UVo39>0TvE#RZvifB+4O813A!FrCSlN5wh&nT4)xK%;V9(b&3D8PmO@KiyQIuDpj z8k7gZ7e!@8K*g=mzn2tBkxUPMYCs-c>{}O%rD=&EDs1Ac#rCXRXs1CczU;s@kzRd{ z?uh0ASi7-i>#mIG+NqU-|Cuy(>3ka_%0Zt%SS`fvX< z-m)eoE{-@18e-gohe4X_hJHXYcxd^^DvDh}k)_Z99Jn{%dmEN;8)Q21o4}1U=k=52j0&7F7XMF-UK<$FD9-XQYdn zozUZO4I~iGONyBs&4t;(aMh^j^V+A`9<4h>P-)t%UV#kGHQ?7eDP+O}EW@rOc+t-J z=QWHMSjPRdZ|(Chl?H&&do7yY0s16~QS(00>;ZZ5G^bFnK4T{A!^AUQ;)L70FfsnK z!@7U`kHP)#!yg9|qYKmHeK3Bg_i&eDyD0KG71yXZ2JNTFP7&;{@g+N=H@f%GS7f@kP2JZ((>YokETjMojUn z72;TnRa%#?%p!YUn6BR@s~yl0iakh?8Y*tdOYveoIXnIWP`Uy@UNR~xBMSr#q7Dg) zoalqv$fB3F(f6k{fv7alH`0X=o5>KJlic^a=##)KS6%hJA}gP=LRd6s6_Xym3dSHw zY{krj^w#Jy<%gm!@T^SRZn7!}iMI}iw}KDUFDi9rRis{>2|Fuu z;D4dt2g-7~-YEekYmPOtr(e8>+PF6^?jTS4PI#h#_e$5h-l z@rGHTlb)tRK9%>oRoyYLw1obxTZQ>Wq#;-ym8i(qz`yhnk`ampllJ27`Wb><;yszt&?#j2qDesJM7Jd>s>2~CF82QJ`2n+We$%*+xl7MDbH zPwVp>aBHW{&lXBi^BCOZTn$Q;Vub|Oa^9C$MHW$Zm^WV7vVSRh#>i*Gtnv0boh|hM z2jzFFCKpm}i?k&XXT28#PcC-p_AsV>g}KQJ3d)k*TgFWGjCaM*vp)?7jo`NO&sAvr zN3Z{Jip*knlz)G#iLByhcU*W$+HPfcGAVW=Mba@vT^e)CyHE@pUz8#?WshSjX6n_A z!8;_^lpSz;xz|aJUL8NZQGra2P73bX;&3F^!}cA{IPcAqw%E4$aAa@se4n}uT za#4~`FI8+8@XEJ4?AN#Jvvk-`v;RfBjE3tj=?AFe|6*Gu`+}5GbJWyg0Pljp#Nlc# zY)O?CK&doTsX!k9E|bX5SsPrSuJqUywkC4G%?NR}o)~0?wZn0+A3(#6vEnS5 ztLL13@~iNsfAY1s98K@;nMcz3xg4%*7fHF5%TYkFur1y~#i1!M0y5H|5DpdVp{^+n z%HIsCbJ6=Xn0QQ^plcJQO1ot#F{!{Oo=;zr!R;_Xf0$l2Y(M+cnjJdQBI-R8BCkwy$N^eZr6utU0Q2QQ(&zaAs`}m9Jtk3lEdwDskE?hhKjektFrOA3hTo4XY zEo|Py_$qE|bp^8zfk?q|a$nx%IR;>U3?I{xo!GW{%K4{1{@wfkVJrFRvc*drBA#8M z?E#S9&!3#DX@<^>4tie691uZQc7m6YPbUkZsK8v2(5+hd%9_ZrnS47Z{IMN#p5I`F zB3{O9+?k}8tNbnQsb=NUYO;x+d+Ne2$9^mKbUVcYy7H;G^YT7bzcMBE`V3vBG*^a% z=DH*)7J}r{x%5Kd0Pa$BDX=;OilCq$1ABqYk@9)he0xYSa}H`xb_%i>ptbNU1W6R9 z6w!_VT|<#(KDNSzV@?iFzq0AeMHPOQJPchE*(O4S+=5N(cfVB2xQ(W;OsMWzvx3J}%{;=$M~ z>NX#gJs-TyBV%&P?}I0<>hJwD`WM}!iH*7NR*WMymJzyh&b?S;bd9K<=uF3avJ zulgp4dwsCtEK67yu2_MnU=QS< z^fWA2oCyBl_Mll_7Dx5xm&ja_@|^V$@3eA9vndwRiW{i7i@__{G*u4x551n4m(}5g z98H3-L!rC+)?JVM$tMZQTjAD6uSNj^GnI)mF<_p?5hB5f7_j_aoH31*Izz{=EEoy@ z^~*1l>o3e^Cef-YzlUPGDbk5m`IR2k;duho<;)eY5P&8pYCW?@86BD!(*e?Z>D5aW zZC))Z5tSql&8JnZ16E;1v6!}F=AgWJmptS4oq4xp_a)GY!~US z=TkeDy%)&0u7?y&dH4**qjnL&-eK%3|yvv6`dNSU7{nW4}$pgg8LAgZD>rO+Ax z^0mN6hK_U+Q%#!G=Jx%yrVjzn(I)7^rVEOe1M)kcj53(mAr+`d^~Q>X7YEas`y=fTfN{cuZ|3p+JN zt0n9l#X_BZBNcZ;VNikYUieMYr^1TY{b71dpC^ho>eW{zm29)Nh)$h}SLc?@E}EV1 zaY1^HDSTx|bmGK^({8=I$a8K@_<6R{Bg5w;nR`jKOj#v48asCzy%fsn@txOIb;2yq zPo}jiTS+c`0R+473`-$z36FShV?LK`V|u3cyWzOw>;*x;+nNb=!tUU0Out*dur}zV zY+zbusQq9Z4YEdHQS1gIH?qY0f7)WJ%=M)Gap73q9xG%$QtRsT0)EQS93UwN+tUiY zUlmfL{~o_?fez!n-5_Lr+kCG;mJb(D#!Cl_ix8b8DjtoBQ%s+XdWBiX_UY{VccrdS@ zTX&EwbKyqjtyXJF2F0$UNGhlXgHLOg_A?b5H9@Stn&5&_8iN2CaN8EP1E^xhpp@~^ zyz&IcqvxD`zz-ZJzW2p)IKtFxCS3B^2T7qm?@o~}H7K8M_+Lhy zp_rUAf6_@?GA@^0OgMT9HmZn-7dP1~A@~Z}4|FBI5a75VaX(R$^ zgm9wK)Zw^CxtU!;AaN3^n@n}zme^-S zAu-ONkQ)}?BMDrwuP&Rc7FK$oH{0*lp;)RI^lzPGC>}v^grIi%euo>>&5C^$lPoy- z_O@UAifrVUHge&J?LI366;Ui~ne(W)UeNHVpf^bG#$FFdnAz!`6xk<861J%Bk$QSh ztU-As;Eb>!phD^6uXTgTdBC(*~_3?2dMw@ z>GWXmE6B|$7G|v`kO<0yvfJx|4?=aAms=E77hW1;9+w5$L4~q55i(VOVOnbE6$=)5 z>eY#!H$_fxen$%ni;Lo5P(bTG=Uz8smJS8FF?y|ZZ)$dBwoD8b9;1cKwiXY z|BRr+biBAgV^Gg+(xwXXkg@=7XKMzg58mGFovy(lXSB`Qxp$SB($pDSWW{v;q&t4) zA-iV7b9a^9D(DxHF7iyBks>UL*~M&;6_PIq4$07zhwxp^9At}z+cnu<<7KaweDg&4 zR12!!tJKz#B7UfHVW0Sz6{-$UYz0O3P;nRhQ3Japv~`x50{L^I)5CtQQGu2EduBAa zV@ZNu-8!o@=5vxy=V_`5a$OV&KAcg^w6MtXg8VwBPLh5l27K}b8`U}#Qb2YdEQUcR z8^=11Bf|;s;RO?}!Qc7)^k3O-oLqLn;@Bh&azqa4L&ie=tf<<6W%6R6@6+Kns*P@u zFMQ>ge86ohENs~7(IU^El&Ekv3(6>Fo@>$^a-G|W876Eu{+&Muil$i1%8^H@gw#Q~I(jO#2x5qNl>Dvk1$wpYpZVQiw z9rCLY9S(0$E>)Qqtf8Y1O+BXJ9_7GI)hM85y9U9zLr`XViQw1Qm80n-a^W?q#0oq) z6q`*EJr&mgI;xF|gMz%MUb0jHO=J;;;=)%FSzWP5vB#i5q1kR3MNMi0Xj3AO2GD7s zE|~)%G2A4M945oP&2iV3?*CJ)Z6^`uPHPw=<-+^$Oe>hAP;3%KmQ!){^s#88GBK)L z4xv4-d4v4c^PLQUVW!MZObqj+b1!|T`}>P+GtDmRpK#=ki{@;mOC`D{brpTuKP`3% z7=W>#V3;O*=GeKjIsDssc_QOhl)t;#wyN#9DuCqZd`Hu-TP@rhut{@LgX>2oT@bRE zISCv~(wzeEIv|T>?hL6?Br#(IAbTKV?K%uU#g4wp4Vh_w5NLE3NA-AYppbOEFpg@n zRb$A4@$3La9w60u-<0)CX7B>iC%6;S6lqkhi!FwBl&b_0WVzh#dZ@U6m;sc z$lsC;wQG*HUytAS!~5>jZOH^(*1_SR2TYP?d-ld8dmo9;6WoNFmRhED8U|wws-v>Q zW5G7{_~V2lyJy@~?&dVko$iA7PBY(a2K$PHLOE{AW4!q|NhQ{Y{~w5l%5#t}H$ zSuT&=co}LCh8{Z!I=t{g{pgKd|Mi;1?`e7GW*#YfVf-FYU>5%BcYo0R*Ux@6_m8qe z6kAJ?Dk|=SS2KebO+V@f|ABu|PoHN_NFIVO0{ktSY-wd=e!!yXP1?`qe)i4Pb1%#( z`9a$|U2_(^+5O#1Uq7S8D!Pi;&6;MdL3PRxyb7;c=pb804k+>jUq}iBz4CbRDaFy) zezyeiQMyK%;ITF^$1^kZPE1u~pDKeLH5Nw=x}g9Zay(irdj0pES8Uk_UDoVy&>~iY zdfDHusSbkQ+pJwH=kE%|S=EkUoPSz7?R>Y_QJhOO*uiaP`^-5#Vvi^q@}sjv>L4@nSF^ot$`GbWzeEX%Ov;?NhajGBh7V+H+(BD?3kRlU_~_wcx4#jnx$-=Q*p;tg}K*8N~uW#||p4TeT9Xf%?1*Y81C@ zr;Xs^V&zIEUFmNsRcV3}1{6!hi;!CAptFQX!cZ908VQvxZFGa+yd+uJ>)D_*(syOY zHODpkLb0618I(c}1c#l|Zr?C&)AX&-58_5cN9Mw*Mh-eMtd+`#0=Yzmu3OdT)hKFI zcLID;gQ`h;beRlE7EBE>$3Q5k3VPdg$hVt6d7lQfL3+uNuzb2rl&isEBeWpToU&1) z=$SuZWCKnY{gIJcOiZ?DZ7WG~;ZEoxD|4c!*bNj}PsM>u2Tc0as2T{hVlk9KRq1~v zYMFM5o{#Ghmi~Sa zx;Mfk$DJh2Vv3kdX|)8|!S+JvA8&WkZC<(}&?QS#-Hx~&aX5Te)cmauwl5=r7>B#U z4@1uweJS$7@3j`t7^5a-ko_)fRnA+%;5fxX&D>!sZt+W-eV4q1R&C7Wq~tiMbe16*TfS$$|J74mQs~0rzcQCWvDG zvR+}k@#fgyVpSD+;5QEp4EXnd=lI{=swSakb+5P{4zuZOoDrj z=CJq3KV0s=-jjDmFcD9_#0gjUO@zx5@OEeE-?UXbsdCISVIW#rN?;KeN zNuS-($22uGaGge)laodCK4ll(7&!oiCtD!U(kIZvGIoJY@WB*ilJsiSCCR?nHm}3b zO;8t{tw{~4l2ixoCW(-GDgkcRMZ$*2<7BDhH1X_$&v0uG$2CXcD0|kuLb=6*k=&O2 zjwJJQdtJE5Wrvl&lSQ#m*_lDb)rBEtE(T*j2GP7-LZ)VDlf>&)k`CBVXGu!|Or{V_ z{B(mlPg6PueD))Rh@+6e`JTeWf4ym2tLd`&0FJuMWa0MclbQ#z7Wr~7Q#)+0XJaI; zDU}|$lQvT=%)=U{7WonPvGJAcuW@-~132f%p*MIrOXCxN?wSAP@<~8r*Y}c#iH)v? zPKhvA2VQvlDl{yu{gEKRA7M&H5vgF?itoiRWcD#3TI5 zO=bq=iW?_(kseT8DDd47(KGeHBz(x>u&|lV5u?QBJ&r!*g%awQY5zIbmb~Q!36XG6 zz2wseG+pG9tUwJbQW1U4_s;a?iuGO$ce7_)r&!R|R$XLOQh$iVQ1(x&$?y%xO2wbaVlIP6qH*IQu&}I@$| zsZj4%H+mx*iw;j01?_-*Pj^hk%$}*`Qhcged3`bp%#{1C^nn*PG2IZGYf^VZx8q_D z{8@*D_>-;#m1|6bLFh`C`x?}xK)i6yFJYoV_0aDSxJQeDuq+Q~ygS`1J%IN^h;CIT z-OQREN>lArZjU}q_mZoz9dreqApT6zgfSbJ?H=Rzwb+reppLPi@An_}93>yPF#3QU zVMr`^L)Lv2zSSfVB#yU2swaso4iHIt zJ@7WdexymAEW9i0hTTvyy93Jfw@<7RY@gUG?~ZAP8rwA0#!0(8(k3ANSD`dl(*Py? zj<2M)+r&JM%MO3!V3;+1@2{_EEVwxq_x>7E!VfntTxbGBw?j5Gbrf4o0qJ(!QMz5! zJFQKW6P+!sp`qAL{cy^WfEB`QX|1r)R13K+Hcf@>CMN}zffvJT)$8RCLr&9+g-{(S z*rBWp+!lKpRxIddM(k+TG4=E=-v-60uvQg{9Bzx<3w=O^g7PUwdF7lUUGPBdAS;(S4(;o@|$I^4$VPCSSguISrn0v$oNv zhg8t#n1=AfA+^kP&nE3<=}Ok1ObjUu=yz*U?$WF#dG6Oe&uI6DC1`=_HfZg{PeE99 zf#$|^q*2Bbcc1}D6(dGQfr@c>yo%KeC_e0?MpMTc- z6s$Ha-@Y*?SA(5|c=2*TI7id~l*`D317bJ4` z)BWG0X$$;%z8`d)Ea0~+x^Rpp%gWZRrr1>!S%Gax)idVLpI_`-7BGK){Pb#qy@`X- z+7TI@kr)KiGd!%-h=0aN9=s=;JI~g~cG*-lhu2&q$PZf*-bZipKP4S{Y))L?^}H-z zDqcJBvNYEer>~;#``wLhP=6eNp^#mF{yv8=6#aL&zQKd;{;MzLwxL1icO(N5-J-dMV2so zJnHEJ`C-8V@hK?SzA@cM61{iMax_MsGEL@N_JZh{FLLkz<$ouXk4Esyg=2mkf>+o% z-=)|WS{JrLutSm>w31!naVug&6lDB%n$#YW-qz(%Ek6dX?eyR@RimN<>KybzZ6sBb zOLv6r1_viig+9-2QX_H}C851=v9d4LCk_UR`E`HU9qR38u_FKX)VMPw!-YNKy;gRi zkYb@ABZrF10)`f30Xs@pzFa|9hIf)$;dFf+pB&szT`|?>-r1iuzPLBpJ+SvRrzlPagf5qGWEJ^qj0vx;iqG{_N#>jo!(^ zwINxur4xEdR@4J;ya{~9Z})l5Tj{Yq0GbXP75IwJ6z%j?dZQYycc@Zgt9|=G(({^f zUVOy0X=UyiGe?<`vIi5^D~R1^qXG;sn=dv9sMg2^WvQej zv?MHXR&J2)m?qtQxvCt5iS_|2M=r#Fp$ZeYboE|+UMD9$kb&q7Z44c7!>mG9lfon`%IM6XGjIfhx5b4#(2A&=GaRAhH;SWH9m@KFU#hYcyUfEf)&I5nc zfLp1gJLcxpl-Lx(Wu-xlzkg}K$G%Y87+b2*4TPPg&A(N3;0qF$QT-p%_G1eM^|!@w>2Y>ux&~`karl5+v9caS{UC}5P}{HRpqEgX z(29q&#KQ1_5HrCPk_I&@P*n5~kZ%H0NDL&rB>JG!KsAHx;8l?*C3?aqC8mMShZ;Yl zs9ibW7C&LYt;=`$Gc~!KXzQM36rXAf`PsT3PxbuiWDCyz`ZvP?vh^i$#A-VCQtWPu zlu&WIWuE}MMwO(CPW)~wqzl?aHxvfdNm+`j>D5~G8McOQRoz!2Lm`wn&isOOs|rI_ zLB4W}Y=6kQSDLjSNsx;fb4fTZL;j%{l>!bkizGG5g_HH_3$xESB6sxo!*S#cJ$~!x zk9f_H>&8U9Zk^d;TaNx@!H(o&ST_GFzB_3Tk!vuQcL%gPK6uXroxl~-6$gFl%8eJ&v^MuAY z3(+njoeTI_w-qdUeqK(ZGt zPucD|1wWx5eCNC^UpJqefeWLBgULIGUZrY}&ZkRByJjBzvsAS~1yt@3{l&9+8KKx) zxnF~l^`I;x>W)bT60vxyNPfq^G(h*)kA0gW*UmBxorfZ6I*|GE1h1W#do`;B$3fjb zlWrq=c*UrfF9*Pe15u+xQ>)KwzqteITpzIw+hs4@=O~R13uI z(~=l-!#RNDs_!P>3$riG-Y(bkVRVkl}(yd0fv)-_z+oD*E1Di z*hlwy^aD=_|BrDF+zwz2e(Og##Up#H^CvzlzGqpZvZ%gql9MltBmAkABYc@+LF=X& z_tj=v6O@Vo@#amT|j}S zHmJK|QQaCW&Dn_$z|I`>JM9T~TI9*%Trb_Wup&7MoPf)-O}v@KVF%P@0RX|2m_liq zs#{PNa3>g>vYMHLLL6P`fp1vJ*2Fe54T^46nn%7oFRCuw#EsP|GjVd}(lt;D^mIRo z2fzk_=D0tff;}>zIXIeHJ{L|0aMbc$w}>1yYd;X~di6pe#-+Ci8bphyFZ8RQahz?6 zZq{z4Px^GS_f)r;dt`;!2(z0j-XS?kl9=~9LKjbOBbY+yjoBUGoQ;+P8m68y<)gXf zZt(b)kHsFfyn60ElKPza%+P{4Wc!^%vC!qM2SF_e?wX0K5<;6m!K6yEDoEF&LMAIr zk#2)@HZnEm)0v^0q45rSzYVGa21jApg|kp#nW>T6j?ql@v!CU5aI02N>a$HHz99Y# zN0QOJnW>(EVPGWlJu~why+^s3HB!311wQ8U${~SOUT{t@fhAxfpoz?LYq6++UBT?$?D; z!BO{%3Mm*NnLocp{<&ni>YDPQAM~;PFCr}9P|S^;`l`j%sF?ic#bn!Xu7(TypC4Me z8v7|0dJA?_as6a(;CsEE3Cuakf8ueI_LSd-sO#FTUtRU~<~Ityy=PA4cQ3w`_ti66 zY;9|XyA2}q@O+PE?G;rCgI9LVMiI`fUP-e$>3%o!6|Az{BIsA;33`~!&|aON&jeDu<~RvlM7o7j_R6jWplV*oy0V_n~yqaT|M+r=LfWT zRYg|&Rz=?TgO0|aBnCF9Mp;dWqY&`)=@|h&p8i4y-Q~u|m5E-9Y(uav+Y-wWR5j8M zW_2nyX?4dXNzCcki%v@WJ!>xPnEYqG!YTK;Cx4jIs_{SZx)`XoEH~QYKRZCOUATe? z2w{foVv8vj#AUWoacLk6y#Po*+Mv|&ytc$t@Yp7*6|R6@l45BeMAmMy#nN+O+0x^5 zwXlOOezW^$#@Cy*$jO={?e)wPAc?R!d-%mqZohH)&HKQG>I@TN2WFnVmN)@5yf8yu z4Ey=lLM)j1{mw(H$QFLlSr>L*YON4fO0l~rvYm>Hk4W&i?)gBJMHhK~9D98VYAt>w zsS?~)H7M%o9duq$E7X*)^(=#K$tET~;^3SebY_?_`XdS62SI7yJFq@I9@8GJhhqPu z%n_yIQRWbgIiV2r=O=bEUoPeiZ1nbJG3jOmbu z*DRKXjd8k8QV>!oZh~@xOOdx@=A9?Kp68-3d9+8jl6UKTI0n%z$$hx_9BCvQ96v>V(SB}0#*Sle6H95-+p^1h93g#ee3AQ zal11&rm92%7T4t=dweO;yRf?pP1Hl&-5nGQo8zriTn6)D$OBn}WQ!of6G(lgW(Tfh zo7BzPwIO*CKo@4Yx5eu=&>CLWBn#ukDHHpZ=KGk%$%`|u)3QE?}r8WIZ% zFqKzBn*^_bY}FG&CAgLhxS0d>xpbQ;P>;7(%`A%9#pw1ykLm;O?WR&X%;=nr*5yD& zl-NS)9=~egDZg~}{#c_jWn!m1mFSRb zs4RF>U?Q|VW2Fm9s33?N6@^kn1Doj>XG=sUp=g{d>EBAc~WWvM|0Ook>YJX5+8 z>V@KC)(f(jBG`x`ap_r~6mmE`NtzXS#PgJq$NMz0oA4U3intk(4f!ro@28Rkx~J)h9rOx&AS0UbF(9 zJ;1hlV2n<_ec=c0KeMA18zAQ^hGv6wd79a$ydt)@LFSV8;$QORsjKnx-odbttPzv z)nwsE61}>ZImhUl)hS+8L0Ix@F3?4Gm%g`wKj!sXCWT=ye0|#0+mq?Ew(C z5p>9%VIj%OGLBn#?T-Sf1#a1*wXG!Sh4GX@vUrH6tf$xw6j@KjfwB&r3A=YA-9xVV zVl#D%d|z}OsR=J;j<{cy<9idsasZd;1b1NpvXCtAlT&FP8-4gH!2=$h+^sxvoF*L%KJWW*=TozpCUk82HmS#z$xWiUSpk#!9o&i}q z@#T=^$ND8tM<+jg9o{)^|FpqXDJW+x+2O+Y`p61jRTR6QB6~3uc+Izq#@NaYStF#R ztHa9z_J=K;cvM*>S;=NaK~w-UO9oY?|4H|*SYSq!?gR(?8odD)9_&cM5$VAL z)2hNU*o(}%SE{{gh|{AqT$#bGJwC|*GFUm}nZ9J2o`r`R%z6jO1i z?~oT|q%qE`yP+sxs(}i{s9a`J6>3s<(6z6x66h9&=bBP6NY|e1-Xg(NA*A>=vqt~z z(aFNW@g}|aTu`Sn%2AXkpDs~#nL;Mda1(iaF7{85wLi#l#^WbHeJH;Pa#>!(Phb1b zYFk(L$r2jZjbIZF_t)V2ad3m=rnrK-sl4P@BWjB>D2Kb&1Pv%>Ks_lF!~pNe8$_Q+ zR|cj;E)_OwVb23nfmf!i6{NqMCCd^Ult%)Vgx`_ZkT#FLh*Q#s!tJsn0mp?$0&=2T z9K{3dV6?-R;OQu~_bxZuRky@%D=enx*f+m*nXDeJi^+u(1yJ`p#7=Ie*h~s^k>fI@ zxw3X;vU{V?Az3(wk!D3f*aZHsi;pM;sv!J7uyUO8YjR3pEPXue`N7I zzJG%H8u{3TqqMiJyo_rUdzB)WkPNCDazUj;S0nFNJ|qS;DC^Nc2GvHliQ2rX15bly zh|b`Q`#KXiOl?ObVIZ&Ag zH4}{rEY53H704b=8*oEjMKf0+S5ANrZicZy4!ATt~h{=i|D#@Yac= zf4`Q=1%>o%Xf)`XRWi9}YFBWx_K={HKB8%cH7XkpafVmjnMH)nv9(R@Hf(1b<*o|S=*zGJLx?Pz+d4-=*{^CaoF8o81=*xwk z-LlnR$&IR@a%LC3GW6KI(cVprp~&T6-d?AZZ0UlaK2L+PQJqdcP@#-5YG&X@JQG@6 zPDDV~VA4rJw&su!xgn0yB}(Kd(@#QrvKJpclRY;0eYw&|2|kUA8;bN`9qO)SFt=4b zWC2?p9zT6SRF&i!tYVpT0Pov$s z=yBD{DZ4GlEN#OtPoudSoHFHWT8m3$jGB}|_Pel4bl%D(I!>`iDF|A}U6@@%FZ8=C zD}ZGPi9jG7IJrq(uF#?0Ub(m<`m>iUtf~#lowE$8j%Xb^M=c=V*8w_fNDhs}&?nSM z(!0}ksCG>3iiL-p)Td3Z4&EvZsgj_f*~RpEHb4fvBf8QfJ-i&Y=bdz$mkzn1^C8Al z7*agathQtD@pKPpBsk&c`Y2L<{@vFsXli-qW*#YfVT$ukTVd}I#R45b6;_V!0MdD= z9L=W>fpY1o@IJ_4uZ)6*6vhC4`Wn!7#0swYt*SdQRnU`(gsRzr%e0mLy6nI<5gy!t zNZ^fVAaSMVeLyw?6>C*wJDoN`S34WlEj2jNT?%CU)@!Q4htFabgc|)Xf#}~INwTm@ zxP;+yrtO$*Cxgz8-+*Jtothuax4`Ja4;M}+clg1`g+1uiRs{8l z#uZtgpfV&ySgbrSGf4_YO^gbmTWM5n#3SIf0(4CMTE_p1DN)h7IM$wMw3&Qn2c)THK z@V9D^r1gqfl7)IGA%{qAc?%yrKVL8k;~!)i!M%+{I6PjX}WS$Ssg3eU?@ zoXkGYd*rlt9u#KWRd%bIrVU7Pye~y!Q%|}&m&R*rC+5-{-5VrDBroD3@Sao&@Xjia zTKYD9gY1T@TQnULN9A@q(?)SvAG%J+C)pperWdH?!Au1DE3Z4EAU4*C{Bgd(|_$({vEE&*)~ReY%r;Op=1{h z95ASJMR4aoSilj!aLSL!5f}Es4OY{AmSR7qNP{UX88fJ;tyeD)9)}#pbwNQ0E-AR2 zR0nPa4(WE~X1T5|3{UW)UcFdYC4p_5drws|!|v)15no|`(pPM=gc;WQla)U z6F^{?J}_3+gg-#|O!LUc@&X8T^L*%Aw!Jb>*7CS;E}f%)#-x*hjqRz@ikX$5{uV#0 zGzJMRM&*i(Xjbe5%81@QnuoZ7k|bc=<(B10a$E8{lI+6E5wJWC@jkLB7P@#dsJKQ& ze!z7>M<`^;jQ(vjPCcGtxj6Ph!%jF@uNb45XFH$ehQlx!_VI5= zy}8uQ0+FKozxo@A=jYD4utm{Z0VfoqzYOO#sV)-OrJdf zdg7rmxRBA5@(o+noF|tO7mmzuXyaUn_{_I_R`OQ{|7_A$N$!#BCLZ!bw54CbXW$v%-+o0dteTQZLPeg$Uh~0;f4A_JG3Lq}ZG}2_W27rRyvH{}E=f zAd@ryobZC{ zHIAJW?ztGo!n^MGG_<3!OlAR&^6zgokyS42oou&4bSA|DZFxEsw}52Qdu8QeU67q# zD_BhuprrhcXO*CU*(-~WFfjV48&D;2Npe$p_qF7&oKf_Q5ew4@pzBzSa=eWnbY%re z9o`m%{GfKzIkLuu5mIc0kS!DoJDN>k^CUfz1Z{#=uQpeTx5#hM^N^e*O|_eRE`wI< ztG)@4emotSAjI_3ynK3_*r+UpLK&y~B_lwI#iVgeuY)f0Ldm$2#8{oD1tk@4f8#V+ z?!qW3utLcOid|2^EX17y-9F5wtO`nGwt5|Lucv$ec@>i$z6$a@*ty-{d3{D|V7exO zn*aXq9K6hNO&Wi6aJ|W~*Lk6XGW@W#{maEAouk>~XYNih)qXZAPN=io%^m-yXbPTQ zpK^T)+%@_aO7rPj1}PlJVAsqaQD$@FSo}QsHtL}&;`#4IfGjQO^WhpY$SXB(9`drT z^=g35dV@s_7I-|Gm{kfao5`QNm4R{>ARKY7G}lO{-Px1R%aDxw&A7LEZ1pf)HU!V1 zaFH5xO;MvcFTvSIb&DM8;ydeJIU0rxs;NPB^h3Wzl8aEPf@zsN&CRJ>NWZezrxa+V z+h`oKPKr;S22$!gx@&qbsS@BQ9O3viX>_P@($))X2QSo6;cvZBW}7a4GMM7ReJUJD zVZ_1Zkk9;}Z@<$$O?5doQE}0y7IJx+bfrJugw>@%f!K09GM+TxrYn2_!|@u2U*-l( z-zWbje#zpwYK`ydNWu#fe1%5ZA<_I4icO-(aw-nFUs~kFz9x;~df!cO0G%p@V<~4~ znO{56;&3!P>xDn8Ts?zWFj3-`b%HE@VK9+vg^6_(n@W)sDh?5VX&v2Wkfz60%IX=q z8$x4rrW9M7K(1QU=fG}}XMO%Wi%XvUI){zO-1F|=YzbhVTuNLxA+yD5{?jOSHAPkp zCS=ST6jYk73%&#;+Ro#FdisReX%`$tuJH6}ENiQaDIg}{a&s4M$|Zj4+BL`Bsz#vsNq%yiiW ziD8~|_NDU)6(?DXtKs=Y&jqq>IQkaXO(22b^iU@=pJJh@KAVb5(3A@>C73#6(R3^% zgCwqYp8$_Ap<5Y=Da22et&-Fk?V3~4lGtw5E?`kT>bENl$D9Up;@mxpU3joI4})h; zg#`B!Gk?fi3G zD(y!MpSO;F>}LH><eV6YE|Al_KUihDxVWDH1QC<~k zP@x{@C(2X@Nt`@&2JD~oLF<5jKJ}FjxyfxNhABT7Uz#)f!{#3ve*mk zai$<((eyj=b72*-1EL-DR(hwPDy&a@8ak4jwRq=~i74@!N#Bh|S>VKwt0Jszu7mtu z4c!Wo6PKY{9!EH5p&1!!h9byx>=2|3eF-nrP+8jFTpvx=&V`*84h7g;O+m<+pmRZY zrq|Pp!y7<{X+622FsOD&+9k^-_d{spZ-09Kcg@=4Q8SAGE!(zzfRy+CM~-fkI(PbUh8j=d#ZA~uiB zvy641Lpx$PJYi$-g5z+(&%gxT{oTL#TENrv?w)xh-G#veSQ!GI0*cL}$QCLtF{CKu z-o)Ku3pJ0A9v8Kysd8p#3f9gj48algQoVZf%e4VWv(iT5CqcL*z@Wk_4ur-NlQ}A^ zJn=TiUEqh6P4U`Wwpe*Vc0U}*TK$z%(i)SLU$XlO;KOb8-@@W7<%Vr>N3ou`FC=Nmg=aQ(zs5PPGAo76Bm}e(~{6n9YLA_s- z#1wky7Kr!KrJyq0?}q<6+$xO@Fpt+R+X)}-`7f%kzUgmSh%_sgR+CLH5m0d%V(+(8 zY#~MR!C4Tj^;{bJnV&(0q_WLg5bt?)3Dd7ERs!Xb>B?1ECjCIxCMr`_dgKbO$eOf< zw=crh<&1U#xu(h&+=a2bLl%hB*`=`#|AZ}{#Hc^$(SD*F&-1hSE=$_Jpl|PS+wPnf zWHI9C+qpjDhHq9>r|)|C#qc&!m#`)BMntcCf7rz-&Ds^h)F8+Kg%-#*%5y^-Ksjrp zdlvk+*b_>YMC+IYahqhf484!y@`j&T0w72h38q)MjVNp z$8&{8qr{Q*%N%(22e##{85Xo;s{XZ*?B<6S7cPASK9nJD*kOvTqewLs2Nl-PI9wTt zOkRil8ibX?E#Aw3d#gcoSbjCCOL0el$H?-fS2wfk*}e$_Zt>G=>6`RGc-Lf;Y|b%F zeP}b2?!HW0r_yDCk{-SdnFpKH%?v)RSDz)jJaqBX8$hzJ(qm6V4^SLlW{L#sUf~i2 zwsTHLiKC&`=!s)He}!R(#r)ELy)DBD2a(9(m{vta^kz-<qzX|_O*6MMmop4+FVe&PuYrHjU-~4r;#n2?a zH*p2Ye2G+AeM5IrEFh+UifeG+B16Xj$kfwR4eqBUE2b`crI*w)^?q%l5_Pj@E{#Qp zU9o2;-tyD60vThQl*Oy{+D3F_(8-hjyOyP+<5)XT-YlHZj>QQ(<+KxL6N1nG_8P? zW4w5;?B)~$quU==EN}vBpN%WKfc@;(IDABB=jZ!u(~2DVxL(i3$QF6|q-6|P5p1l& z-Gpu->^D);r3yICuBm{QxC)43?}M!J9=}gK&&>Q%chT|sZn>`2knWmwD8^zneo($G zi7`{3N_r0iK)B|Y;!e-uCvc^PUg6y(%$eOC(?{n+(JfG@3yydqe|5iVr?MKfKMc-dNEs;v9@~i# z&+=0rpw0k;(%1-|C8OVoqUpS&qB2^hvLd9J%*F@(Ni(<%$vMGG;ZfF^VRQI zKs3Jlm!FcuFN}+G%?fg7C>GQnPEm2F&|vN`+XLy-?a{yh3AJXYB6kZBjK;z2O3`AgEXh3PD)}aRJr>nK+q`S0z|YFML?ub;Xfw{N+Qu* zNa$$&Pj$=P1NVD#zVn^sePbg~65B{Y3c7qW=%InyOc-`K8+I_tP)#Fz(bvhPJLz>i zt!g>=N5-o`Aj_{?k?)574{{HpqFSOj&8H;%Xh<%-NSMRBBH@Hf$Dz<1leqU%=ncCd zxXk0`G!w2|#OnPd+lh_SCuT_5OR?bB?80u;UQ+Cp9uOb#YzyMJ>Vb$9#a-z=32vn; z3l-Nv?_eh=CE(zvfn4-yppo1s-RglMTHv+=o0OvH3%eecO`i~SMLm{WS6rHBe?lj_ zAz|1k+GCTGJ)*u5TldA!S57-mh?u!5$q4R@&_$QH?c!;ftCC%F5*f58R~4WY>wP)0 zi;fGa3htS@kU^Z)${vzFvQ^T?^hU-CYu&3tFwE^h9oeyh%N{@)=^2N*-buDB{C&O= z+Ib_{#7Z&n&iN&893w~I-ve_`suv0`Fo_{YLiBPh3D(PN-H}*7IqJaNdm*}L6p5*y zQ!A*RTRmwvl3CCApT0IWEyFstNiz z_!A>JNcp6v^+2MHK1_)lhBrAmAfp?Ur+0Be&C|Oazi~ytR(;KcjrSW)Vt7S?8GsF;-=RqEu(S~^u$C&R^3zehj0{ssi+ilI-o&mBX0#zdM@587ra3Z1gq z&5%{=jyyQ$pjhwQDu)1GDsu@`VQ_QM1sPt@$1a)y;c`rr<0>uPe`Bzr!ww^Z=<=ex z$Dmu$j8?Cnh#E-KCNi3B9+#N}R;x0s3Nih52)uMI4cnYhWo~#zvu86L&wS^?70b-_ zD>YqNNyD!TD4LN!Et!87M5ycC}rP;`|j&Q{mLQ`!{x z6Bp@E!U~x|7Yy;VDR9EN3<05B8mFu}*`N!GfZ~xBL4y+2svgJ}d1#i)Hp?>T9N1kQ zfWJyr$aV?*ma9DLWDn#WVJE@w?&c-1_#5O&a%mvEcPoS68omX#;&L=o)adV-Fd<;! z-+ue5bE57Q_W4yYag&Tc<0nDYUeq04p4VrxWN3%WpbH|g)(&XJLz9GUk_#TC0R{XP zk79?m#F?zh=!V_#Cds-v8mU<9_*=5+e3OZqzVQdgNjx_b<;06AppqF@VY8lM*HR>j zia~F@B8MP?dpf`*1tXMT*f=F6KT4y{7b{mq{a-}$^Dn1 zo$b^IoFK^oQKO$;Om_|g%u6>vI`QofEt_ziHdAh;t#(Ivgo@-I1d(a3TAp_8mn7k}nX9&& z%~hKzHkBeD198OMO7BGkyPlvy`0K_7;Zw5Hvi!gt^$oz$1<~ReMT%3Ra+tZwiQ^Dq zK1!VNVSc;A_6}Zcy<}NN{(St-iCtMM#p3(CGbZN?wJMl}&>$qaHi^b-sLjU8mT@<` zVVriX&pFv;dv1Kb=9hcI3sBZ_l#i8q`GJd>m86QlX-Xyk@!Srd^nhAvA$^KeiFMJ} zBst-E;h)K%^h$u|vgsrRp2wid=3vZSI}T#D;OGS|tHV&me7gJMk4^K7P`LK{^gFwo^y; zdDn)ft2MiL*P*2NfxJetI^Z+!avC+}Ti{=9^XJG2O}@U9(#xDbdvABs%RDU&q^Vf~aZPU6E*|<(dO~3?WCIJZM0`DhkmSCl6t@MW5=BON%As2~HI-@K%=n87v$*C8>iMmqVQAqTw2rbv3Fh~ zcr04g0^jv8mXL@`iB z6pJcJhFDV-Stn?Li*xDyGML8d>27(ND&K8nfxpq6KNuTM@_Vz)X_E@D&|W~pp-Gf8AnD0G!u%H%OJ%)E3Y#sG$wK6@+|ST{Sz z{&M9zmL!c%n|!ju9%-CTjJ-(4$}Ajj=w-sCP^T-hcWU%dRfFqnY=$xlaA~%xa#Tk{ z+UWs#E-ceadBw2KD)PdHCVMT3%!X_X*mySGm_5-xJJzu=g2z6gA*gTLnV#wy_+3UjJ zT>SEnEwx~|Z~{2-G-O33)>yX}J6F>a*#o&ottwTPDobEtZ3J$HUV4leqI#$!J8mek z5gbpAiMtxx&RlHvM9CW2!^I9nUOrED$iA4mihYCYh*sRCLdBCb43Iw+HWpxJ1$$nmfVCQ*idy9COu6alRgZBm8>2 ze5Vb(?FPYXd%@K)MdbLc3-kbv>0B861UQkAjZcA-#H!LRqNQ9qVIE@zYLr#>yvnU*gcbjhLP37$&gKb-$q zqG^?<2lNq5o+6t*%WLKh$m6^*K2qxovHsMc`;cY+OrA~O_AZ-lJW&rP>}Y&!wkl7T zEHB=8u?@5RtZt1PQJ9ePadOB-vfhc4J23RaQrS5ayOknZDB2fc6-nycTeDUxP#G&N z;L^OfX^$m3=_7)`z~d*T6iiDM#Rpv5B3eb*Ce}qGUFP3f72Q)#I!u^r`1|KH=tKU*@CP_(o16K_ zT^>+tDG9(O4C2HAbt}pSD@ZY2=D!t`MY7d*gRp5mUwDJJ$1epmMzA6T6EF?R^FF(Q z!)4F~`{~;u5ZFOuz7CJ{K{0L_(bEqkJ&}059sN`@M2rewLpU@3G<%-Cc8>h(_g2#; zU!?bk0jJ0!Zp$4fMvT_Xb6G>Nt0}Sy5_9q`kZ*yC>Fd)kgFx0j(Mn*eN2(qN`>rF5 zn2n$q;ZDaM`rmJNuwwy*6MG?6K+*Fx6ujCPeABCqE|fnc)wEt-LAU)KFkvJAg<}CQ z($iO;`N4mAn=Hjo&Q53~soX4u6NBP_nWf05SnyS|sF+;89_b2Fhk z2#1{G+t>Vzw3W($#d+qH6B*fcnyUGS!$iKjFfrhK~Z|XTLFdtVI}cx&wSy!0GJeD ze;2P=RLERlVk0vBn~g;2$TP4|xKOBL3Zd?Olk1}}2h9~WK;!AGF!BbM0k{2H>(CgX zig4mb0^ZjrnOu-){qHqoAvYJqc}*k9FmpkYDK?QJE2$Vgv=OH(4b3&E+B?Km>mZc8 zXxwZ=(2Jhnm~)4=F{?jc#7bL9FViA8D7hkC?|(;e2DF1(+}h+%W)_NZA8D3ig~qe! zhAvgr2~c*XOE`2@4(~YK5wwbbRFS~$m)_^^;vEmJ^zI^Oy|;RxQ@br_A-@GQifThv z@jKnt`?rhgeagIs?#9mmNeK9yjDW2kHNnd~6Gz#?ZX2RTwnjD{9TntCOF)~VCn6u(w2qSUkV58|>!3?jNC8wUE)FV=Y*Tbe)&=zV7W?6q>B@BV z;lR6L&7n{tpFv*)frkdqWbqPN9R$*@liME9QB2nfiv3a)TX`#@)0B`8&ZQr_pZ6(N zjLw}O2Z(SJ^e_R6+qCtq?7!X^i@J;xd&E}MWy-vchIELk#Vvv|uW~`&)H`A2be!jt zumOHWKyO49e|uD}TaSF7cb|0k^pjCpb5jK^f+Nr~T?y0$Rgqop#$)NqBuErwg2Kr5 zs3d;7Ko?yayoJSQmxg8p9dIp}-_Lveg~c-t%QtzpMc@^8Wx3N+r{vFj_FP?bGRVqg zdK`u|e0+3Y2?dqq%C?yeC#-GF0-veYt_G9*HK|i6gW-%#2qC#cra= zMo^z8r`WW)Kv{d%dtHE`l@$6ojeUGagSSJqih-b~TzY24irMYZRa`FrG;DPCh1&}s zB@VcK$SAJ1ZjJt-y_US{l?KSjkKBO3vX9+Z*r|1lU{6*rQzvTzHVPDP7*B_o6AVT% zoTI(PIt)fwOHD|9KWE{qo=xr-qPKf{#=z7>@y>Z!^ZFFa6-(cQ5?H@d@iB6a)V>A{(!72kKpGclm=Bk2Krz2(Joc;uB`7QC-@agSI-0(ID~IfH;?)so9S$pAtEAWp zij-0@D4vNTo2XET1fWpyC_N$U3rh^C7xj^BD4{PSaUpv>6Q*>~=>bdMEamN*l1=A} z^MQUAnUl1xH%TtY*)0<03oo(V3Jl{VzughEh;#&HN%V$ur)sdkw^eWN4hFG9{&{Ircm;dFw*}^>1_sHC&k8Pz4)|~O`tt$*hUhLDE2IZp%Av5j z#0|@j_ zm`qTSi}nmz>cl1}+sp)gOtI@Il8maFHIj96U>yrYe0#|rzuPm?R7+HaJk9>#E*iyu z$9X#9f{d|_HkbQ3A!FiCn<~y&S`{l_-bMVjh8-4Bv*XHZz$GfKol<+Iz zfyKc$W%$r^Wu|+#;*7F)?B@R2%i=Ob!|b}1&1cUY>lUXWa^6^$dvwEd^({_2BU!0f z(W=m;$(^1Q1S+;aKhKn?mR`C>en3q+mHe7)Nd`T1=C69T_96f~a>te0!@UTDa8fJMU~ttY9mJR~vlaptErDyH@}sjnL-P zBRwzLsNBfzrgw@CdJJhU+^07M91$!cnILDUi|!*AM5mP&&6HH2%y zLr+oMlpGhV^r%tCM`VL%o`K4Uu)RxO5}xa(i*6CbP3ot6yjLqyc&I!WKk2}XSQ}3JWw`y{pSGuyyzaPpTFj#YC_A*d^SURarwq$axt9J6TNe$6; zvRt}Z(g7_>;O{eap`}cIKn`8wv!0(xuBkYhw=n{1Bs*s1*mKx<;eq?7zx92~!b~f* zj*xJHbc?3Yt`Bh&NIX-`367e9SBq?r_U+RN`^8hU1wn84VgN( z6*^|CNnE>as`JS#@(hzy!XAxUs%fFz9yj* zE8@AkC*Sq@*zLWxx01d%Y_lP{RP>PC0AZnO+HfvE&~#BXNW3R~@oo?Em*5a0;DD-M}a}-RUgA!NV zKKB3z9pHr9iIu-SwSqS^4F!SRW}Fkl1Nb3_h0RhZb`3>VQ!!;B$0VhptGv^dpFscE zWrZeTM!9?mdq%03?opm~nBlTAezxPn$^#s9m>Vt%xA=W)fe9DIel>E^&doh>;*eY1 z#NjAxlhF$5nN>Z6-X*t4V1I{rKdeK$VuBR$e>$+rVE)xJK#jW`(SX1uV$-9 z?(|$5t5@=cnq2WNUW%eaW-QG@aa;|S*&g#t-7z=P{OeqlRKpLsIl zu~p@rA7GgX;>pj7aa9np z9UN#Ro*zU<9qGFEy9iIyMD?SJ%@@ep*Jh$BG;?>hQY`cmZKh%tu@Aym@sMm1GlZYO z7WN~eXkZRq9tCajExhAkMv(dicCZk*m@WbRZJ4?2ryXf!IJu!@*mPp$dGHhpkTo1)WIKMN~Gus#Hs66O}o7F9rUpu zU7O$UuIc7&4DJZ3li{IrU%574gS%PkHU&1JY4D~$*U4fdPK&gvBc4Y*H97QVuXg(0 zyiU2M{!5w-f<03;S@dc}apbL8_QBH_a5Kz)m~Gm{YH-8Nug=X~YuOKErR`=reJSJ< zUK<41mZ=h%0hcyriQy+|!Y1?En9*|CjP|eQFaV?dh*90gX#kuz0Oxrlu{hl1pom=O zZza1Yks7m^y_{l8C~^p@K76Y}R)C7d;u$v}AzM#lsoehE-yZkRygQ(~JX?JnBw$e`ql?}y{FsRq#=EbcvwU(l5KylLP2qKOR(aou&hzY( z?vb<$5=BlELuHb7WvBFYOhxbBOJGGzJ$YuoW6x9cn_&2iKqw+k9 zn6xP+{(U6fo~?}52h0XIJZHG9Kh1t;tixfXzK)BPdu5g$iPJ*FimQZ0uSG%26)P3{ zWtvrNnLmmsXi^kFcL2T1$QY-0RN>l)!^R02BR;_~2RR{Q;!haW=J!n=$$@u0WaRcM z#@V&pEQ0@tV*4oa0P;Yhoj|Zt9#Shi3(MrP+5H~vph4L^wZbn|c!IBGG;4wAHk&>i zTq7wCukZs#LPMEkT*zRc25KgG+XFw7-lISBuH=^k`F9dOMRi8muh3jp=RhzWC3`i< z1BC^X$eD0hy#u(_H>x$M((<5^c{{=fqm<)<~{OhKe@vw=w>9TGt{5`eYXg``!CJ^2H08Y_GzgU2;vS zpp2{yg|?G`jowZh#kY>gAC>>0=08m6+u;^5ja=gvuybBdpA49}Ih_>SL6J|Xn7+u2 z;0^vz{9>5iwnRSR?~A&_+pX@W8=|j&`Q}%WzR?&BQ9IQcQHk4CfgU&paXSST0_o~j zaQs*Cc8dBUchbug+i2ACt`(IFACeq!{-H@gfOn(?^hKsCb9lE%i{M(+RY^OvSoe{B z-a$C8i^iuM4aTP!@7G1|mDkDk$onKarx<^?0hhkWO5tLDsn0su07Q&d@v?=-$j~!E zltP`vzZ+D?oN#?8%ko6^k`Wy>PUi8sDbzj(Ml!Iy*u+W3JGQ_QpE@6vh8G$c~15 zgylkYvH}Je?xtf76L^RCM?+9)vyU7d6_44tU}Pi^yyTd5gEf+KC4T>$=<`+SSrXM7 zbzgW@k)ntTLS4>nba&V>UNYMvIQ!+RlHHzhleY3YMC$|NC+VWEBGZGai@rt|M?c{| zB#ms3^tPf@)Ht%)c*)4vdU{OAvh_noah>wh-v(K(-ksL^SXrsJDQbKkO9ou_^NM{m zMnT>YXMj<>9!|H|GRB-9Xs3%gS&xb9o8lrVlMD33<4Fw}a3MfDJFK=Tkz!X;WEmB6 zCF)2-rKC*>FFolE&88^{GZ*|lUc4>ueH@d%Pe+a0dxu&=;`YawU;ZZM{r`UVzuxQU^m%`$jSr`^`+dH8j&}3zNYI0=6n4zK3(iPQaHX&{{H|Af^mb{Bs6F}& z{Qn&MzeY5EaJLEM&%D%e!^-SCzjCoO7tV@gYjzAtlFMd)_~B3o>$ujgJR2!nhWOdl z;8D`V*9$oZaa%-jr1H)C)EmD;A(BL`L2|2~;ybB=U)~4)<+z@n= zK=soOUI(92mZH>Gj z&1HILRd{!S)(5ipY~^8J6VkP}2(N^IW*o^MuyHV zRekCxJ+#sEIGY;{$H0XfUMSW3H6OT`7BdB({nwvJ>}wOx1y+_}YqBJYO`ymMD&|30 zB6A)ht=$T(BzelMS*bw}{(1bo)M0~ei@_Oo`M7N^+;>(aHbEj&u>Lwp=>s}jxK(F&Kn~_DanG{KZ?3{%v-ft6 z?z=aYKYgMy!9@PQn$2X_Byz^={gzSe5ej%7V~oY{`NC@|TnyZi-^Y|5p2-$Zd2G<;no zcY`3c!X&Vu486w{{YaFlvr{eoXkWegx7K{m9^=5tXLMeB9rd&D5KH+Or*&qnNX3-% zbEmh<56)?lZi(K?D`cA37P^a+`H#~aG8|2|p>_Bju~dbT@!MqdOAYLT>P4=N3mMOv!;M2%;ySiuBSt z*}8zN>2clSz z+@RUCH%6p!g3yGiKL0GR4A?s@gsjkhp(4y-&n;2|7yr6|v)-txI_P35Cl|&?7*3S) z&wATwuCrl;IHSXcXWHWuZs<_0u5huOb*vb7{2m{j+6+CNWwSNtTVScp+4=Y8kH2-d z@!i|wb#i(Jr+4qXv`FvnnO$Hh9B)N&y-C_E!oJN#!j0@H&rII(C_SwY8h=FCNW;fz zH0^WoYl4ZNE66#t%2_Bt{93ecesTiI4o>`{u8Q87hceeQ9SI%aHPAILX* z#!tHUg+y_;93{wbCfQ5M1u2Rx>Z=kwb|?5=WV`&{7q0WP;zTi4AVSq*_`x}wLBd-X zoyQy_cvTyCHCpjX#&~|;j9j%4LBdC9YWHQ3*|R>E6o~%VuVb`AcD=(X+_| zF}rAtIOha#s&@P2eU>FguSr7CNk8n5=Mr$=jHIy7~;$}70T*=sJsm#sAYB;a2MRLsMstpvILXkC8 z%=u8g)L4n4S>=he2#F9)xi_zbf0@bj$5ZE|-Qy}H!3jtlfH2lecR&2ze|_ELm`v>c z#T`<`EkN(YO&YCc&}gLCdWxK-Vp656c}w|MCA#QElh^v67saXexs}s;dK*(9)v7j4 zIYsA9Z3>K=v{G?6IDXQAOWdTgkewpq)@=!HL8w`DG(;EuFe*R%G;H5+U*&ph))w{= zF+L$-X0;q>HI*f9H^a4(3Z^D7ky#|&DO%0bMdKAU;YWkDOaap~>xN>*>}^2GS;Xvf ztEL|_Ib20vY?_wL+?CPpa0FVxdr`8Vm~hoN^U?%z(TN@2hh|8*MX@(1(n`hb;*|p` z*85j7@g69^gEVg_#*+@@Z8xB_V5_8!$u>0SZ1yr zR=5E=F}4zNJHqg@KVA0KwT8FNPq$`(;8My=LvP70KHPRM5c?r zE6Y)1lhAHe!V3|_$tsN&Io9rb0iw7~Z$C}R%6!$SE%(Yh=@yRzGmKP&DQ>yaMCip? z59EVolheH%?&nyHh0VCIxX;nMoNzHA;^wY@{k6#k-I}v{8oA}fHfX6?tLtYJ+e?vd zDh3nAmzk1zSPuz&kbNNcp2W|fcetIHT>|lc6oWrOx+I#Vs&q(a;;e#wr6^Ta#9U#T zLu+|u5ZApU=!RB4kk(W$nUNl_d@^*{1DB-6$R71K=vsv%Hxx%fbQq3)R7z)v2ju&C z-Ez>XpK?;YAX{Av`G!*ZCfOSZWFmq?QO$09{WSPVEp)A`UfSt;`;AUl9jn)Z>1!G-7YKwO*H&){1XRV@b7vB z8(l}}8=ElTed+2{6ZU$4B8w+mUK_uv+-$ZiqF7kf6;LrKbyp}qORy4Vt49;7SpBY6Kwp5(=FQsFWjtenoHAF9B-)oGFT>GLfF%7D|u*8r$*S3g3d+QsG0%lgcmgh{vm=0|CqPFTEOu#wzAD8ct zN+(`iwU|Mnfnw_@pk|C&LH0|#_=1g|S``?KYhdO4nW_F28mx`1{5 z=qF#A2i^?$JT;PTUZ#JAXR9X`lXIH6><1Jp_!)MX3vxKwBPu@Tul1j=R0U3MsABT0 z$nW_hoPfY(a7Mb83y?;--%-aJA8*xLD$+ksj_Jg?Dl5A4b+V6{tCG^-ej$`>cDf!1 zmnMgYD!)KDb^?y|3oi?Coovtot^=}nEb45MUYBI5HDw_=;O(|SyJIO-up>`ihiFjJ z>vcu|BErDoqiTZ;gJu`686>4w%&sKR8w;IsNlYzNg5opF=@qk66rVv$%XOj!$v&t| zOHtIxwt3*EteB0Bv=~FmpdayUEePcTpJ66t44bwA&bl4?;*x+fmPHlM1Iu}9->tE( zW<%LhIK^frT#ztoyoabwmfkT>EPxSJ26&P zn#Gx)P;5U%9#Sz#bht~N#P6WjPHuLChUx=zo84NV#~KC7w+EL8L8pwMn$`iBQWR&D zT34(Sj|<5cpH`+$DTLbcVi7RN$CEBusx)0WAV1)G5Q4l=o-T+F)u7r@CAmlR9+hNu zK$jN;luK!2`_n111Nye{`g}1m)gw9mGHBtsOn`izw*pGuQPbiQ+YQ?2B_JM|OCJh1 zlq(+sR>pMY!sye!+vv}vrJ;se_C9HMB&PjSX53H=x*(%H%BYlvmVj>hj_^_*(kgZ@EKe{m;T8O_2G~{_R>)?8NIaSbhzwF#UvL zA*WD6#UO2K94A=2!c(jrI?D4L#4Uk0}&OQI(yW*Z_t4a(H$Q|xN@Ch0@DCL;)G zT1$Cl1Sw#RC$PaUUHMRsdzZnV!8_X&kHB_p@dY=kfk{!c2#l9PqU_B|ev`Cap5!_< z(Ax#xBYy$L1bR;Jp5VRt$JsMYD0=&P-YSys#3(v$hN43h3)E+t6Lh?|}N^ z1U7a?+l-`=TH?SeP-nYcJF?wuxF&k9-g2(L9W7^>a9_rT@ zzp8UJfv5Tl<$onhoESV?%)qmbVv{M7NX0-&yCNUblGp@{t`G9hpzwW-&tk=b1-7n% z#`Myk-OK?N&u$urEl#kQ_=6AJfACEcEQ+T*ewS1`F)Z|Eu(&|64HT&Z!4|{n8fwtB ze24`r<3g$l%B~q!>Bw}}PA@=G@?5%s>7biKOWcrFyaUn_i-d(y11_zK70}A?#Cx@3 z`J4fl2l4}R@z?=Ka-%6KQe0C3TX<{OU|0v;>6WG}6c4!6GpqPJ17Lp3fTr|?!df|~ zCWtgGG?Mdn&}g@6W%=R~w^L-fBA$O%J?MgABU^0}2V{*rCviP$BPzKeOZ@)&4=lB7 zt%$xFQ%L9q)Cr0W{2a!HTFr{d8Sph93_cfy-L$wc`FPqQDDl(N&E&m0!L3;fg*o(l z21a>3l-~4_93XSAz}yptNZ^`Updc_@lchFFRO3iI0G0>LNNo(M6FiL4U`-!BzLc(# z*UI5Qt^90Y|CCR7M?+cwB0I^aB3PDO^D3o_z}6H;egYAf>L9!pu3QVMU52i~2Ih{u z(+4}ow63Kd_)PTIVKsxi2C)%os#^IrRkxy!(Mz3nWy5KJop%}i4`nM0$C7t<-oK1k z$-84HFm*}|l=tX>fPA4aYu>F{R{}q5_3DYh3Pml`BDk)|5NlQ4fZlZFrm59IAKGe- zY%~rdUN$OAHsX`)dvLb+SG!(yE>F}TK*?(y;C6Ytd_2Z;gi(0;5RY)vI2__+%O?Kc z$h9lxnA{-u2gg&$ekXQ=8qM6GYKnzqdIc4OIfW)E5Kjn=^Ql&ay~~}VI-ofM8Eecd z6h~@-iPNx)3vKk0Dv9yLK5wn6C$hvhUAb4X7hIQuuciriLvIeIL5)w+iVZj7P1$Na z$RL3Jk9Wpo;PT9+ck{a3k&g$x@Q+x{8eKERZ3?7Z!?17*1cFi&WzsGhB@cV17Eo0o;I9dg4`oOs zxjhV1W5{0uu5y#KPkJt(fGVJJ{SY;Y491+%fi|$U8P3?lq&Wmtqd<}6E!I)BN2B|N zrDWJ^(hz7xV5~`cMBPP}^09L(9nz)fGHX(qa`0;Q&)lQzRgJ9zm!@O#*$ot|nn0-f;G8XUG*uxu zbsU^?V;rh6xxCt8=Gy@;)BKUvzm6j$8k~255I6IaGh?BFb>6=W=o^QpC7#m2Vo+d@ zU$nlvIw{WA6H{Bgd$ZR@asF&hLL7p2QZ(N4sjd)R_ z7`G0|m&-Jzp{u;{fiES`*4+m$`c_1Kk4gSsH)j}_MkykbS zvGlgXXv<9|#1^YFY#tr9=~x@YX+EX?u>N9KA3kveh$&0*%{s&iAkq% z&Uka+GqL~3zvzjaP%-h_Tfh31z#&xZs$5OBjwHP8#FI^h8B&TVb~goyp%~;BD4l-} ztObe;_s@)*lBd>1chR+9x+=-#9T9ZWyCI~~BIs1SrxzTdS3{33-qJ@B*i%Fw4ry%B z@~A6bE167nB_B!>-R_brk`n}PwY$9;t$i6j+u-(#P%)Zw){*p!ymMQ_rq$~b@IV|^qsTE91>fec4Q+w7kD=tZ-?LLN=mM$kxv(ViiH|Ot z0rsjviLOnP6ia?JH)apWlcO4xxCE3la0r1)f~NTxqBSugcW57;T% z~zCBpt^w^ z6nA=IZ78_7fz|Yi=u`X~4!{iuY*;!P!}rGuh3D?KZv6J#yjg15X=#P`d66&{dSA2D zr%1i;f&~y9eRGLyyz8&=FSGF_Klc(GeZcxl9>EE5-Wlf2dp};tnw${#N4=NG$J|m! z&ij{kcay9EeP0km|WxzHzF71~bF!On4h}Gz6PD?EFO?>2-~6 zBu09YeUEV(3Fn>J-bH8Sd#zcm?--_`wmI=SP-hGT_UY2;&y{fr{L zRLsd)d!VIt)6`Euf;*cA&iV}TL*8zF!jzM<27%inUD+hPF*}29lHHZ%PXELcE-_v( zAjf<{*7W6yG$>WP9Wm%~U6C|x#WY|5nwhTdqLI@B$$mGmdBDrr0hQ)?fepXd9ek8N zPmDzPgAvFeq{&mKPPs_>+;LU3*Ka2jHS7*9js!+WV1>|B!+oU&mZ4;yl5mut!;Ilt z&&$)XZ5PDsV1<2amZ5AjON~@RPz+;`!_5U!wp-+0WHGP~L2pT^OfSW|j3VMrTOP>` ze*foCP_Sh%#%WEGmGBA*Xs!#`4wRi3GBL;|phR*pRI{uL!0v=RkRse3tU3A4hksqo z)7%ZJgYL79A**@C;d(%oA<%*Y?kEdptDzdUf;QiDkj3JLlyB91PivWIc3PxZ$$)2) zHGCuj29_A;E`zjpwbx*9>|CgoZUY0<@BRcj&sGXs1SqWyoh&_(rJ-0Yr_nNa=PFM; z-$G{+G)4|W%b3Ap_v#fR`nZ9mcIoebDKgD!%hRs?k|a!mVAybnVl&02QsiSQ23v@< zD#$ON2D4qpl=xz8LfTx6eBzJ{x->AkL2c|lZ>*tDne3p~Zuu4CJ`9#maLhiZVQ}6; zgy?_z!zI26AJaGf;5dnQVy_2+_`^Km^%T37B1u%tsjpm>6wGf`Gy?ygE;?)8D*iSa z$W=+FSFNB`zK|Iwqz>WCcEk?dF+M*zA(s06=|9hQHv!||SHD?LRyr{-^2~sdMzI?x z7=@S`pG;n?Pd9Y^K(Q6ZZNaaFnwgazHR@*3Eg-r>6}fV0Idt)i8%AtL$M`Iqhp@*9 z9TPp*Xg{!&eS?OYVc*Uaid{pI)y85sJ!I}RSjGm*LQo!9Hd`~)4mFN%-td=lc>P0f zW?X)9d;P>|{k@tEr!503uSprLl`zcXxx3v=6|nb2D-~yXx4oMsI&Sg-_Cc6!(6-NK zae2x=J$SI+%QTZT{A}L>veAiW5}>ghX2kL+7DSG>QZWX$3wjSd;B_t%r;tRTpDU$T zx`Ei}R9$4e+cmeA$aH1N)MX@d`hb@%vWwKp&PQHTJcgF9p`+&_9XZL^Hu~nv*{6G4 zhLhv9oD3T^n@Fa8z9cA$f{Vd9f$1?zjj=DcLbnAO^g1;jhLV(Z>K^x2`U0sPKmI4K zK;nppQD6SI?KN{P9n#liYO&(J#_>^v0&2P&l|TPgD!0f9Yk8j+dnqGbDKt) z%kbFgT27P6MD<^FH!T+_&$pdBZ)^q}kKxY9(-qZF8!=QGm4Qe@Z$z>f32^n$g??Pp zAsgoepiNjA-?gj(ZPu}pAYAp{SUM}5w>pMc>8$u@K%Q&hP+#s-Oa@*4)JZ>Ldf$SY z#|EaKm(J$WH9qV4j`U6~zW*FBGhP#j;WykH*Y{J)I#(_}hVz!ikTNq{^0X@U3@_o$ zeeylb3h6<~4UmU!5j=G7@m(!G7_~*6uDlYc7i7~d^bXzzDV9n@cTQj;lgO`je{TzY zl+AK~PZzy`=mi<{qPhLSQPq za!Ythh-|}#%xw$ZpsYt8Z7iP77ovZPd82$`D)2y#lMlnqsEpr?ROS{i@I_Pn<0rT3 z)5swwj(_OP%*08Gt)@sN6>~)h27+p0v%)jQ#_e2jo4iiYCLcQ50FtbY(a5}hf!P>R zE7&8i^hUk;tBO3&B4+XQgYKE`T1fw=24%lp#AqcqX6vFG7*v$b7hej|@-yg#!UyUk zeh%9M^h#RhGQs;g=)Wu!;>G!5lq4TZvUfD_FoMLUe=TM>+jp(|DI+u!Q43q2gikYp z$wjQ*PqLjDOrMy6X)nbVQDhetgVoqr6Oy7R_S!e+q;IpNe{yMPGtisYLQw6!m9yGC z5}7{fd!QpkibW}tmr)Bo71harxVOi#M9jcInH;3ke zpHe_2MJ0w{03SI*Kj#zDN7|xqe7+_htaU)p^eLJ}P@vyMkO`^FYiGD!kzkWKu+j9` z^elT`)Gz(P`cXUY?|2eR`tv{?=9@03l#+YrSsT<*L30Pj{RdspnL+B}#;I}M zkSxX@`Qi$uQ*a+toE-$TQDMsrX;#27#={-)d7S)0~+Bj!t_(nEQGU#%Otc@CU$pe;9)RJ$f zH-_Myb+A%iA$MTl)+TtF84b%F$H48{;IZ9rm;e*_o09j)25x|H;wZ^JGr(-4SV$Fa z84^>dll78PkSo+L9E-eY!&iI9VHXHK5-e zD22tXAhxVIME67*n*JV0Iz-sZTPJI$ah<$kHq<4gfcoGe_4>$?aPWl@L^pWdf)(KQ zcG5sdjH!Y$ekWY4L19#kxM*(kQ0mZVo|XzXsY*lffk|Z4#>s8~8-}&vo1?{@G`D@S zrM&5Dl5yT-2BSoZT}hE;;MRb6P6niyHKl^xZr$?Bt8m@HIeP*N=YRO;E0T734zD{>7rowRuji$Ck0s^b?FiCEH@|%qWLwUY ztao(tbAKx4pT9W>~X<ndm!KJd3XMke~kOap>GekoO_gWz;5q*x&Aump|bU?9=4&C<3sImIVZGE{9T%Gg{6Xz6``EN!G*lf zq&X_hLCI}Jk$9y?H?Ll#mv7}gbUf*m<$JO1<&RV6!?0hD+sk)aQ&I3~L!MU zr6UOgIk2aDw5wL(c?;wDs_$eu>_zQT*BAr=_EelbfpL9c)v{`vrEW>8$zvvtUqqz zfTa!(m$}D@!){hIC|kW6!s|nq0Cw)viA)}oO&=9phNAo=aslEvy>ID#k3(3#U7n)I zV>)J~&8Vi2)0ZJzSnr!Qqh~5Me1JA!R5Cwp#;&>B=-ydJ-i&u&<8hQUIG`@hPI%%5 z78|b4oz}hiRaZ7f6LFAE6QX@dS9Z&H2d74#VlgD5DGlDjV!09?Xj7z03wRd^?xWEe zhq-HXqi(~njqwueR%9fv_J`SPzhy~-;ykOZ-b@W&htStCw4ik zNbsPR%W35X?>)SFVEj!KchMWX&x4DbD(H<|E(0l4=7yqNP!-ZUt7|IiBjDN9vL&+e zs6*g*B&c-JNv;*l(co?LMz#!&T$El>J7gZ>2BVQ3*|EUI+P&6c6!BF}rKKsc@{L8A zdc*e>%X)@N2{mZqu`DcK=#V#R?Jwd|6K3@pjyPfcRgW;qJMTjC&jUV>6sCpp$H^fV z$$D<(1kU@Hkv(R<>k5?T!m>7CIGB^HNLO}*^-GtzL3hV;S*1$L zoQY~vVh>mbtq;Ezjxq?f&^30|u?3I-)G?SAN5K5)Gyf&_`_xnuP5y!wf7|C>#y<-gAWfYo%wY6g8H#+{*@lh*G!~Q7)A6>; z#-?>qTfS`q&8>)vd*sAOVK^stRNBp;agk!rQ{m218)H(?{=&0qhPq>N;DoEQPcW(dfp*ew*vzz&X5s3d@CNd@ys(NEK1 zWu2lZ4P6(|Bs~i=O9hkr=AcUsC>mp)c|d-^b*o1q)9Quh1>4l^JnLAF(dcc6<&*5Y zZ{3EJP4UpXns8G6h4R0WB_rX)iCq_f#jtqXI*LuENFws^8)cac8e2$P(5A>2_PeJk zw+D}75Y9A;Hp9ZcqyEyTS%<}MbA6u+$#O~dI5A+X1aRA-sbXC~E@bo$Li~QUq67F; zN@>)g%opd<-MmlT$GP&+X3&6t@zRmD;{{U-|Ha)>L&u6#&8{hRf zX7BxPu3$8{qCk<7asDPc>l?NHHGHK4DgT)EV|}n`Dw&?~w`7tt3CQ$@TjzZg3)A}U zp~l@puV%Me!L3=9lI4M#_V8m+^I@E=TigsH^cnPNCH7!u!Pl`xur8ogRtkC8`~1sH z$-Hdv0$PEKB12p#&rz4S#q;&@K^MHbm@cM4sotJ5Js04NP^5m%v| z=O!uf@0rEvD`8j6n|%SE-{JC=;DmXqWzXSX|HcHZcXm|ykWaZSk)7C&S!O1R+fT6% zDbhp5X!#H1#o@U6=~tu*GG%a+&U*dJSH-W?pTv{?e1H^3c|ND7IgL8v= z+%@UY(P})r&%21Z!juYn=4w7>77C9B(EA-5?DyNmi{l87yer!qU57&}t3pN&9SfN%LmB2~f2702=YRw_ zy$sUj=>brknFZ?H_yMJ~(fVLU*B&bl2hEIJkTv49Z)V)ew#-jCt+}z1aVqiMDXNid zqx(J1dJnk7^Owj5T*~|tJnDQ7LBgN~x|-X8hUI7oOa;u{pmb$C|E#)%|NH8Hy!hQ+ zU+egX#xMT?Y92&&bl%jazz1Q)>K6KpvRryoQs-@V>tin%j5wBKILOWFeCqjG#;e|_ znoA#Z)f9^s2{Xio&_XVa!39l~|H*OT9~?ebHiKgLP2;f52^154RkJoE#55HZbrdcl zxv$MsRAn|5?Wb6%ktn2MHcd&GsX0iW6K6^EFn^%H5VGUq1caInKzzFniuNk`6@FFH zYpN#cMYT?;!6p^F0w{v~pclXxa5h#5X|Q>^l-KWmgBLqDU)%)2XS*`v+Ke+RE|2A5 zdqHsHmoHppO~`S7)O(41?8Hm?y=GX*rC11KW>PUNAn?)%H43OAq>Dc5T?_lYa###W zYXz9V&VVIRqwF3m*)yho;++`U5Lykc$#v2o8-(JB3|i}N_qoPwc*ZdHX3w$r4ksH! zh5q1+Pb_7rtmrx60>BTT$(!Le z055r%9Zz!_0OyTtY+hM++A?+hn&h6WB(?Q4LOg@cm;sHh4F;+?Y$7l?GdQ0VhMz~S zf4txXe~|~$x93SP3l1xwe)lVZ2`UGW_&)_EH@G`$W_kMSJl&}I>Xct;OmoetD1Scb za^krr$t+s8Xd*jEk;hcbj_;v{5cCUy8-Y$MWI!alc_@duVs<-yC>*+-_jw}^T)wz1 ze5FS_6k%q2RextSa1IT+oEBjvXM5O925)bdCxJhb&s)mW%6Bt{)$~oKm%(SIC|bxC zX}@B?r9-xJN_s%iblmD!;7a?7WCM#wh8~9x=#piUe#M{*KE6eOS7t*hCUr{Rlw$cF zQL(&5u;Y8#;O{ifxj;9_px`Q?7G$sR>ev~T{4V!m#&{3N3joJ+iCaVH`nmW^8Lz2z z&HAT?`A!?P<77oAHYA6yl9@nw@>}1&Le@Aj2zQ$~JXgCd)Nh+A|=fjNt2(XCl$ zl!@X)bQAQsB?neVZUYCXTu|iI0&&_b*uNNg7Tn-V^Lilk+wuP!+}`~|%+G%PZp{Du z>^Hyrb&P~&T?j>ukg}!f+6(^MlK6+qjON6lJuCc0Xmt!76c1*_d1K3j21!+M!yXa1 zr&EAU8#d2hHa*RL7wO(^R7^CPjwMTHRDo0}H{0X1ltyH;nW0%tv8yPu9LhH#0;|b% z?}q6F6FbNd=;7ZTx_ilT^=xI4+QwAVEhP5MTQeV$<8-}eBBM!v zYms!Rce?UCxeMj_r(xE=DTx)d;qj%PY2Y z6^Hg6`Ho+cZ`fZlez53O58E1mJq|_YNa(T^ts~v*$Yb2lx}<*I z>ED_rl;73;d^)-Fiq-P2Fk2)&rr6IY(hGS)AZN!)frIpZaT>H5VME$|H%)TnWdhn{ zWDP%?E`y+LL+JU?eex|VoZ7`}7GY-_F0Cr2ZV$fW4hQiu>B^+($n=hs9a?_ujC=`J zN^B(CgYh^f6*GM^ebbdFDT5K+EO@{p-oZIY%6(G4B^n-c&a+LqJ-8BtC-5R{3T;p( z@pI|riWSi%zEFu4T{I&t_yU9iZ}8Unw<$_|V}*t*g_{CMgUV zVlNe5Q^u6|rWpNdTOm1QG{@s7LftwfW0zemuxv-M(#O%}RY-Ty<#e1do!7`d;H{cc zDmosei>}~jf}B*XAfLB{N#~8H8Q}D~9r=~Fc((P|J<>wMX+6RYZHkMfHjUHfrmd*W zV1+n-RSol%A;Z|}g&{exA0hWd_xU@8&2Af*gYd<#@yUb$4TR^gq`|1fSt=^02cTpL z>HViSDEG@W1%@o8{V~o7$=0og1rEo)%L&O7R)2iyy=kVoY)JW;%aYlNJ1RdhTO2*r ziH@7p2b>`Bldgw82uoM?&dT(!l5UaqL{!trpsI_`r5l(wh|^Rt#{psw!fNGs{>Ym- z>Z%Y_czqD21^S09^+sh^1T@)()zJemh4m;hf@vyOttJRLpt5MZ!cVg50mzBr9=?o3?_)3X3CK1lejNx9gxC8SEe7d(L5~ zN4VQ@hd3FOiPzSBef3O}L3#Un-YSyMt){_=4a#vd7#*V6{S;6?#N;ujm^Eatq|>cV zSq;uspA_QpwDjc)NN{9e2=^j3d%FRj!e%8+#x9mJu0acwD~=;r#cA-s9B%6o}RHE^!$ITn))LXc*Nen znLuitILd$53=Y>QwwWTAs2Cjtgj&M}fy*v~zQ8mSln+|>=DIhbh)#7h1gld(G9FeV zy-@gYz(X&scRS$sh_@|TlP}Iz>(~~kaL8l2d3*d!)DPYAWZ{BrdMSur49GR*iW9D= zj<#2l7}CZR(wZY7@v02^5xq^-=MF>>8fBTzb<~Xx5i^(p4eb zCGbHhisZ<=z&P(K5=~LiX`tFW>-|_-1A);OO<(o}%~SN)w($atak5VnzEnMJ&DTs2 zd%rRDhvX9{hS+s8h@GR@T8f;ZVz5m2X|YBeAM1PnAA4T{*Ho6~TUWf7Hdggk{d$?6US_(dr{_&CMbC^gue)pdrK_i_ryCa(7Z4RxKmnCS5J5yy z)C6$Das>(%MO2UoN?{==Qk25`z9f_+a5WbaUaX$-SCV^k?+M=X&v(A_E&u;eL*;FA z8|B?Oh_m&Ayw@`Vc9LG;$4m|=Rh2{SE-Ja639ggs^L1&Esl@$u9^@pi2XCHvaKT#P zx~bSPV3rkL>4Rfp9<*Ow1nMa(yxIXE*clKP+)NrYPrUFo=0~?EI>KTloj^un3~l4P zYEB@ko`a#0Z}CG_*qVDYbyifJiTU$JvX9@Y=e|`U1Ryuec{)b16%;v2#dIsvfRFd2 zu3Q5$)TzV}nguNgcu*SIFQ`$LLwz`u4)x9{klmTxBCe#@gCgkez)#++_c1B*=xYKf z8*fz`=>z^n(yZ_zVZsz_GRd2DSog$d50D@Ydf=a0WxqEj&kqN|SQh&j58<2VS*ejT zIpU1#AUEQpc%4}OL1@Ddtswek$*=lJEk6*sar*KT8xS>9>QdAfwPmW^ zbf+{oJV$3{J3=kUD&}&i2}GVSmWb6i*vs5Ir&NdN0Tn+8Wfue*rFGIX(2kl1^l_M| z!GVqjRZIu!B*ChvZQ9)MD*psFD_n0G7m+HknTP99XqprVIdeMwjZjMh(FN zOVZ3Rtw+!ao?MOi04%XXqhj+^hNekc6yX2@U5Byfua_?Z+?zL^uD9o*aNB(w2j>LR zHDU*{#mh-#aEepZJy82=nt0J9M<@D46XWdn`K@2!`frrH?7{|Sn;Ae0TAJ#H1^$JMY^k$wyMZy0PQJkSghLQiMgjqeWKxXzmaP>i2@&cK~wRRz{96=BF8ZbVDJnwF9^ql+7 zNAtX_eoOh+zF9}sOeXm@3w}DqZlT~G7LzF3Kq~1^BF~e3pafq{=RkEolF+T3yh^gp zuQ#Yh8b2ADJs`X2!Y6F^ujKi_?e2EnA@>7+sZvt9z_m?sD>_tv`AR)m_l5Cl_t=0u zlVU+=I2F0m`WKY>RjTy)5==eoyOoE7DhP6P=F;6zU2deIi>XI_bZ)lpw(rBaaWhka zN^t`4&J*ZXBjY-N4~_g*YRYe-tO(hhr^+H_{1D>CA>(=*gq)&SNJ>>wF=wJW>5Z>H z5Ps_0=hYBhEXV@3opgf72$+1c21qKTjSKKA`^9=bpWfzi`kODsOgoP;Xk4Be5yO1ek&*f%yL9psIdGwTAz3AV8T ziF?s{_RbI_#`g1FQTKnytU%f>-qcJICj-6i@bEPz4pTqfPVx#18Z!@v5-D&;mMZa~NR%pCg{jpfNYHuCtc;i`fHN(g|Re6HaO3sSf~w7qhgM~gX~0`1ZRbt0v{@J1$Rk{^Z@YvZc#i9#FgkV zsRmhdlrLWe!3~oN)wTbskv{nKuLRkPF#)_r7NCk*f@DdjrVjLw2y6XP(v?M7!V0gC(@w8l12TM?oz(lWs)c9g7{lno})%k%rbR`s0tWsVMb!r`sQ*pxw7%#Y%A8yie zYwN{UQ}ouA&Bw`VH#SAPY)nxa#crlZG8I#!d>~v(as)T!_0hY&RRdIvm!cnfB?dq6 z>y4^}?M}7q!K+uLS@Qo`d7e`YK3XVf?>}a|^>ZsYp4=}|kQ;6cjwd$YxKFVi6lsUd zp!|py=_Weq0a%?Zd)O+b6c`ozy}EsS1PLs%EoK4%eSxf0Wtg^Ba5A`%&YqPIY%+UA zZ93cu$4dH00?5i*q#`l%Yw(#1*l|i|J6?5_2wS}0*ERZ4D zH7BLDQ?1chmNpJbYkt<~7oTZO_iNLy>qY#!`(1zBDj}{C{9vADSUYrhVwziTufJPCC^;zH3iH z!@=ItB2N9fg%0&BSqFXZwX{KWOuSup0xBaLqR(mbbd{>7u#c^qpAcO!{WM*vN(`;` zeF9G86ZJi*tGx(E9{s2W+Qsw0{pd&e;<;}my?*hvD}P@<5Ew#pb#?TLxf3?=j!yWHMRV~y ze?B}vYnD`9qPA~R;^UXPuQ?C0XPc|=p!hu5n)|Uq@Ig*lD%%coLsP3PAVzZK#{|c*aq!a69hI;`xVRRk&WS$ zf1goc(J2Y-$mW4s{r%H!Y0|Lb%sgqFWIbuGzs zV+a9ahHYkxDHe()@~M~}b-(v9vWq?~Uk)6YCxv-}#Lz5J>I~$P%7WIkHZh8xR56|4 zpV}ly`h|ByD?&p*&t>=#gA%Al9z`wsz3nr;au#SZ-~a|X(E$6-x!)C z?Du-$WfHfkj{0CNLWl5_&l=`32oB`XcOm|>E?~K`k~z<01~o*Vkav55KDsi?a~YEm zPz~Mnjm~U)walRtU}LsDJP_f~rTk!Xe^ynzy^INm{ztp-PMKj|oX=C}No`fdO0vMJ zW0|~BS{rJZ))0L=teU=-I74%gzDe6|>q8gq% z9`wM#e5ur|#D(ddyK|ru5y}x_!O_9D8WlGJ@4W#ESxwTzQTi78IwWl{)zcP+-P1M7 zn{sS0x~*OBjR_-UvdYjD&tI?BW5XF1G2E1Q%>`D#TIDL`1p>>U21l@xOT7R|!`^lU z*oOoE+3R5+>cC50V1H%W@5(0q){3e4Rf+RS=VY?RhHUI9#r9L=F%{GNlfar3yCq5U4@6uH9`xAdT}SJmjk$-kl0u=rMT|5-mXSa$OPU*iDZvu; z8C@ESgXjsr7} zNlG9V)7=sD0E=^pFSIk7FT*tBV^S(@S9Z_PW0xTUEk(UA=!-$d?7lrefBV~i^RuUV zeW`kuplEZLNB%jXYmMhl`BKUI=KqlCZmqzwotvIs(L~kIisy#l- zB-evCL>fe`s!qQv+7ekmeTra9fl1s?r|N2y)pVyC`CV^=Lsjd!ZQ2@g6=?Pj1Xj}r zUKh?8VNYFbr#8{_wfU=gw3JI4YH4o4KYna!kFz|eUTX;XKGT_ zD0i@C0_#VjRyF<>c=G)k@jq9CAwa<~^UjN}dg9eo@vUN~~ zR}XbE7+Bah|I&PYTJ%}gFzq4CE56?ru}zD?AM}hfL$Gfpbq0of`n@hfx(r7-qP-x` zhl-mVfHH3EZ~1QXMY6O$QW}X7?E^l!G`7!xb$b|TQUR0NluWu=cVCIjeW(QG06)<1 zNi)BmAViJ$+!vkThp0bZy;c0OOR-l2G1MvlcD|9MBEs9 zXI|R;#W^~|v|UUsv<9z~bg03QoRenj4hI4Dy5{D{_2^dT%`$46q?TQ~lh-uoq2)#M z&+mPHXfcMqz0iKg<+j~VTze?kYJ+}(t^?>2D?mUFLde)puq`S#+>$wOG3V5sJpa#| z6@J)w-bl`Uj~6y3NqyeDH6Gd!H{Nk^(1v6&g$$Cyq6pjtau3wkk;tK%#`7)W3g0Z@ z128E~p=pYCWvhB?c$JIN>h`nF4^L)a#(oHw9DjEChf^c1-pdaT?#?2m{6c$fjF4Iz zU*|Z*R#L=3#bAP~Av&MgrMRaygtn>*0?R3b*B+mo@M^jr+V?L_*{?b#uGF1`{@Ufr zSjk%9rs+k}vZznly}B$(IrUIcPTib+m|5zX@3W6FgdUoH2of4gM5kddzl*u#cTcrx z`faIEfNfPPB^5yGiC1ize#Hq$8v}|Qaf`=|EXkjO54`N9fJo?kQP8SM@@$l%LeFBe z$`kIf@r4s=v-mt^9uv)H*l}OWZ~6JnN5?wsONTjXZhN6~^^dpJCl+C#EQ?7D-4l4k z^P%DfZJ5403W8xSmMOP?Fkgu0@4mMAwf|nbD8#zzwEs=VK9WC~RNI*Aa*Bo2OeqT7 zH@$UJUbN`8x)iXH7F4{*ASBjF6B!OMb0etF39@Jc!> zJV}Vhn`IbGSt)!vry;s6%=S)5&*p2%lIhtEmlSfd{m!nGuYF@05=n3lwm4a&0t^}@`rKDfb7IDf`zn$uQ=d(p6 zPVh1u!We&)J2dRLFZ$WC|0;OhYFfUt`5<9#j}Z zojN5>n2FD)#)%WbkA%a0^L4tVAN^^D)ynk#RJnrem<*JZ!^@8LQEV{<#ECH};m}Pc zJ0WjYZy+n7%B~Qk9TGz^b88S~FFYk_Rh?yeB5%kK09y+}W& z-0|8|VH%KJF7-S(|D52cv@&XyE5xXs0~%+-#WEK6*LReJjQe^n-X+Ekd*Vd5jnr~b zA>w`*X<=|-!PtH|AXT4BClO#KLo%}mUKP_@!pw)os)Msz=r!Jp^Cm(id92uVK3Z)@ zS}^6%2?8Wwojh-k)sge zJ@7bhDD9ZM+kVI6wswPK4|5bMp_YYoy^YOFd*lyaJv4o_7$jx!Zw&#$pO(G zpY5{KboIn>l{q@$L!Qt$=h5NYdD#r=(wl$yk$u4nADiL6UKZgfk2y>BL|o84kz>VE zni7g*GIUrWWvPWqn}QmASS6F9u90pGMcOlcmJnsp8x_kzo(&ol8WsD3l87ED6q}(U zA(Jlk)i0GGttYffpq|oYpNG>dcbvRR`_DHh7| z_E0f9be&M?U9SPffl3XpKjn560AAXPgvwhZ(k3Zwa zv0aYj=V^H_F-kEGi1Jw#^Y%n!f@*A^0$EveUD`%uHz}4whaoe;n2{wF9GPQvJl-=z zSCPWWz{Wm2=ysH1%PHvTiis1ZDL#aVcLJLnfb7@VI`j4tMSkN6c2V?^vt;}952qB) zM1kJ8;FDUtX<8rX3!T(rDJssxhk;k610Gn%dRrA2Y>xBdTRAWxalsd0z%(UD`G$6j zoz>^%_;YvI#T#{g`JUwds*VzVb6@Z7a-Z=(?F+R>E;sHQJ|rB)+m-Zc(&bwtM#=j+ z+7No-jTULLPcLJLs)%?HGT^b4WYa~^L{uAb%DDpoxxS)(Xq1&%U%tkiGDMA9B@He($Ed5c2UyAs-JpDe?rG z`3<3`Bzt_$0x>`p(-Af{kJbh7xBs5-MZWoZZurN~?3Mo7YLE{7v$v9TxN+(KDjN!c z&nUK!B9EvTXe^^^H3nge7?sF#AhXy>=jc`hlme3U*b{&+EY@C z)1jrUf~gUjlVwd0tR|rFOHu3Z2)-Z$0^hiKbN*y2`l3zm>B*9p zEYqE7vv#ki*hGq~0WMDk)_&_7*c}iNhD|p@2E}N{W>v=YmABsHphX(bM~vxSZj1E$ zk86+H@3Kd(aqhcDlMEY7BvEVvMOLHa=|11R4BCu-?=H^_4KnOaJRV@O9D&4`=5(Al zaYI7-o9|tBP*}DXta&du#|_bUXD?wsC54fxQ%ZnQtY2`Jbdg$3K9f45lI$eO6BzSw zgo8oW+m5LF3LYUC_i!XU+!2^gwc)k{g9APqKp z(Dzu&T%wVPXuq^n)d^Y?7aD7rCe1*BWM)*Vh25wLLc)3UDbMFKe)2^AurGlV zUOwlGzzg}W{3Pi|Z`xL@n2U`S|s@A977Zcv`;mgC%(w{Mq$xKnV`(|kHgxR*F zsOvSY8tkraS3ZP2jq7|K2Z6<9T3mxF7}cj~n!kNC*lq0xM|>A^)leUWwNp^zg8P|V zL7NWBrndQJ(pRD7xy^-4Q6p!D*Hn+Z-DPKZ*^o&M-~08i{?+QaOnUm`8N}qqp35?u z;L9V5?WV{lKr^N)AfVg`f8PUMMPO%CVbQ^$M~84@Xb#<|?uWX6Gy$fF1h!Gp<+;(b z21sRVRHag=U1$Zhv@Ds$;~4O`=vSl43^Lb3;5vn4>3iQWO%3B!H!BBNh0W_B zvZ6y_GY}K&Z$oRs>Zq(~10Jc`tZA7+7DAF?Hj5YTxD1UG&kq5a7w%qpZ~B3szwG_? z2#mENgAivqMPSgFC;p7;I+H_hn!4Xy+==xJu3K$}9{uM(`C;Sp+g$#F7d9sSOfIGD zHT*_~U)(rB$)(gsm%^q9)~-gyYDT})2aBRGt%h=sNX?+f=2?ri*y>s5GoV@@wOeun z{h16M5`I9gOwp%sT@Nik^o%qiu7VLadcvuH|E|=!daX*o^)DoWUlEo2Mij_yn_c=g zicO=)W-8{obboNJ;ItT3bo#xF(Ae819`LxU{VbwEaf+-J#07Oty+XH%GlTlV?y4rr z^zcTGtDs^x?mKnO8|<}-M_Lj$MgfOjafaXJsrBX%!)P`OHP6_UeTIu?E+q8G* z>{lfbXK*FhJy7mLeh}au$Jy_28;`~L%2xZPwU-JU936JcLN=RY{Frv!$J}6A)rEpY zSxWeR-!w7Ea7Fe;T9|f|e3pbaMC0Ou+QWCeiUh{cwSJovcjjSnUWR6)tW0DO?P4~` zjs(_tEelD0t24;y28ADNMy@$SXC@9dU;dAS86xYN^GE&V-;sDX-ZACctRh<|HiaS^ zFgb|IuxK%mqFLV$0hKD<5uFKYrBc+VU`L~0r92VtI!|N7Z1IG_i2Gf2g4=qGlAo9~ zyRkL*i^%y7Tyw9leYJ|&@2@XY?O-vc=5#~g$P7Ay5=X{y_BGrn`PXj+X8x_!Sj7M4 z=p%C8jg1A!-4A2@Z=%?16uCmhJS+F>q%j89s959OHT9wr`r>D8oBrUdWqz5oVH%RD z)@$lRkI7JFyg-8IFw2ulYLz#?1sXo8I*5`agx`=4sPbrxmEgS@ngP`zPf(_ZW4xkA z{#1^A-N^yP3+^d5iL5-CzzmOC1DP_c6t`@LEIs)sLz}?fhk6xDKFD!kI)bI4#l%EP@cbX;K zD_Sa9HFeMIA_rN4#4ycS)uqpy#W6zac|$toy(>T58Xh{dOHS$DOZJRb;^fBV zEN5*DlYwG^-Qyq?v+29Bz%-W@WOyf2b_mkdg@SJSB(qAHCPwDl3{6JB4f#d##3xPM ztS*p%w2%CRuoOhc(!^LUca}|GP#`-Rydu0}9%QET=z1WaXkhiV;ug9U!Z2B~ig{bZ z^_I%I6!lu458l7@cTQOL#)y&^*r4$Ph1>A#ZTxkmy~oSpSf10QO-W-meKU4JAuvZJ z!vw*jO$)-}*=0e4+LaP4DRja^boigk>uEbY>i93@e%hm$dH4NlF8SfFBkf6+p?6@I z8?l>WA&8Mf#SFEZwnab^FJGNQ?*Ns)9A7}k1A2+%Q&kQq#U=>>E|{yu^a(1>SIpC| z6%;{kugni>7?H_1i>YF;v)v^i;)NEDZ{+x~F>(Z6XmQ^){lWaYo?luQqCKLyGs!J} z;ad0o9*_YW(05U6Cq?d2F?ZC*gr`WcAYXn)sE-TIp|NJ+h;B~=2KQsKDN(|L(ol*cY71$)5t=kU@ z>zElwH(ZZ_zyq(|B-hv;Sjg^(^jN`Vjt;7EIqUPPhWf1nS-$Rx&z=Z$(k-8GO#8$K zIHP+6+ovA};+!&{O^S5&_UTv^erf(45mtdOl_bsVQ;cH7o$<}#_(JhKZTszMPW5@m zzP$Sj(gw#-_>CwqXPL0Y8YdTQsLKqd?ZdV^au5@Cce(y>%}4%O_T}AfTMWZd%H04p zku^#vCXn`++cdBV-X%-oi>AYoNq*57u6Uo{lKJ!XU)?IRu1wp-o0>`DXj>#VP81c| ztUnnPn@*7}R16Zv{N=W~2%5|qqSsEjPhSMu*IKB3sq)|Ixzz`AM3_Iyq@T*KkbAm& z;)%hKBcQP}`>w#qh%vcA^U0aQ`F>V_{Pf~0mq?l$2l5WtfT4h5K`7vE&qW*TSi?(rN0EQh- z;BdsWj{gKVFh<+1tqs5bbNdbR7sTn{*!Lqu45Fv%r6AOpNB4WByo&Xiw|()c|=e_ucVxbs}PX2X`j8QQ7a0yVQz#6KWrZX3PM4Z0f1tTXft$8-)sg8Vi0Zr>J4+( zd^uG8Us8SoVg(16qEq<^T(>LsU>v168Sc+ zNb_7o$ZkJZW0-UMAzX6IZ8S#lQmM25R`oOc&=^NdsZmi%pcxdsJacI2l&48yl-{1u zejCdZbMVo-zi8`j08eAXXUlHHR**>re5=Lp5Z_VNHoDS#dR zGj_QTEc%QX-&*-j{nRQ*-Ya0i z>M0plBNOz_?n5rc`>JX`wrOj_SH0|AUh7`!LwCr|RH5dnI_|0GrXrY|NLc><(cM)n^CTvAOan$qIVE3hz<)4 zy}6zf^|_ivdJJb7X_PzB$5I=8Dws}htoB4k-DWjPk!E0l#%J#HHa|Wn@Kl-6pt$aIx0N)qIT3lTyT;fYHI|g$-dS|d;QPL7L(M;(9<+L(wP!;cs%R(|tg?`g8kjf)Q< zXFsg{XCuWXQ6zzi+2>nLAA9Sjya+T8{`g2$P45MU#`35tp&??zEXUB7-ch&?T|ZGj zdBK&s#3ui8Jmg32yKr)mAN4Ot7BnhuMQqn0-CCL7Z58&WAcaM>teejNcHCDnoW5PK zDhkLyC#V*{F*0)-gEv8NN>15u=pC!oIQ7fxACnXO3WeMlA1yXErk-MJDRKc8Ny#5vy?Bv45aRuf zidYFWQDuf;#;-+;x9yp7T#y50&y~VuK$?$r)L0XNAJi|X1D0%kzjwbkuE7UXT_B-} zF&HeYZqSr~nakCklojyBqoYR323lh};D^WfgVU#f#R{uEGnfaY@+E7PxM2gH8j3wn zfdE!aPWaM*O@eaj8eI^0MVll+iIm5|)zBYtGWdb|7^xEWs135z;JDz1=t}ywDmG*n zENAy$f9RJB{O9!%`9MU`5S_29)8e&dZ*B99mE^sC09dXwwN;9S=o;x>T{%@3m^6RD zqk3MMwiS5WI)gg>T2*y>Y#^k^=UmrQ3vJjUX z2Arh4I;Qz{Y%a!!3dcTUELU-(VwAkkq^OGe6)!sv{zXBnW=OdhH`n@1S&lqWxIDy# zv(GUz4uXQ?Zr7dW2Z|%t-;@ZfE0RYLD7D2+hHdlkT}TSWZlK6ID&~(6yxnPL=KVn5 z1_D{vrPm?mrau~-rs(k5GCvlg855T*aYzPd*oyAD=WRDKzsK*k;o3_d|J|qd8zi?~ zRyg=x8x`1@x0Hn@{`Bx>=$Sz8bTv5xY&A&Y*CK``1+_G?fJ~zr3Zy5-hB#<9hGEwg zVPZSf!XZE4wl{v~9s9-P3z8J%kYKszb!mPcnVl=BU}7=>4h_eghsgss}<82#=dO;BlRa zo8~+D%RaQA%74J4WkG|=Almq9lc+di)vJbSf3*|O%WI|1Z)<&MFXc6|CER_J znZQIcd~3dmVmDGGiHg}R0p>ESR*MJyRQ)m86{x5}<;e3P$7HLcswRZNb=a4hkChm9 zo6BEvgQfq-p>No?QE>DhU?Zow9&k%gH_4P4rRxH=MdgOO{2Ms(={WY7N8ab^Gu)4P zl(3}xe%Z?RJkHcucQesv|M@7{K3cNIjWc}}HpZfaVv8uq4#%VgWdZ>}ix{%Mk!>_c z7!(VZ$vS*)&fKpsDc-LX?*dU-R9$NUW(oYinPhqD>y^iXudCl}Rddj{|zY48Z|%u|?L zTH-V=T#RexpMNtqE=E~GsUI!qIudEMIzKqLJByT#W_8?mr6o{;GOX_HIK@J^%>ZsK z2%r~8^5wPSQvVin3;AmGYOo$yd7?ib(n;s&YJf0p^}HluB59967Ho@5Xd7trl=$xV z2VGEFZ`AH!QQ}~wq(co9y_UtTN1ec?sF6M?L$jT15F+<4>I81`uAzX#nuRS=)E|1rN^S%~3OshMA*u%Ufk+d6KWGI= zCjT{KMAJUM|7B;-8q!26^ZQg`3~hsj#w_->RI+}mA!0RZp7fFP-fhsn2HkS5DkzqJD-X(F=>6h!SM3jZ@ip&hk6J*I# zgsr3@x;nf;{c-R(ory6_oP$Ui!^q>ig&QgUlj2X>2WQ=O$HNhbZ4o~b<auTO(K&g+Te{kfyiovVK$&ZncE`XO|K`9dPP7FkR2eaQtVvIJ_=cu@}T7!#>5lrfo2IWI?E1*omK%if~Zt_!x!aH zZu<^>LM;3W?L^-q#v3q`ZTZj~x_mw=sH_v-1bw+o`kdxhj)U@fyCCPk#|_H&56s+T z-;B!9^LY+&t)M;WSas{`JdA)3+Y|-hx3aJ(~nAx+STlYEeU*| z_xQfrj^EAwyhkwt?&}O_{=Kn{Sgq4OkF4`#r5iUG=Gbh6H&bjfMb=X>7#>+F*-e+A zLz^fop_^gHlL$>2S>gEBg{c!E0XYmgd>}jQHkZHT1=&eI`8Zx-PvGaa4h@H7=t^kK zKq8zb(ZSg!=$qc+cUze$Z>LLr(5`8_BqFx_aZ9Q2Uxm0oo!)|v3m*`u* z)j`-@{HFnr|K1Aw7Kp$$kwgaf+ex!bid?#a#8Zo#NryM)gUpm}iWc!+?aJ`naQ${D zTWF&bSR9~Y3UYlU>%eGb$u7+pqrn;0Pdo!~{f(FJK+r>l>KmrQrkfsMi`N^{K`A371rRRMl>sd`q^dmWLiZQK-`*2Gs zSSHAV0gXb9YFSXT+63$zcVG`_?v+D1$y7F9{{B(Te*apr$r+sgu0h+_5H|d-eE$HQ zeIN43MazA+aX`yIJV=vAu@K_hPQ`3df60C$7)uFKiucM9tR$tGwS%?;s_llc&MzI)~D6*z!O+T-Bti0umU*E>ieP$AhJXw|hVr2Woz}P2JV2jqJyt;-d zr0~xeZT{%bXB=}e!tjs&31d6UJr|>_1n%oso?WxRe}NS^fy>tHCWqYE*STzi=+hK? zk|M{bn9AS_KQk@*R8bFjG=UH#(vM|mQfCZ?8KyNVx}$*i@^BE;L(w(jN>$vf49!Z( z?g)K5lNDa8tnlp&Yt`&y@y~HVr&)c{LYj~qkn*ZNLsuGABrT9F^G1#7L67A5Ml10d z+)51WsfY#@%Fe}(D-YpJjE?~QIPwtuFdlU&DbT*{-EDW>9R2u>iZfF-1nDcJMN;f@ zMjmu@I9h;RXF?4w4)xL#C1&S^fT(SWL-cXIjy>_P6${Qa$0p05M ziLzn=cAwVy)M;x&mrreoj+MkpklU$~w%nVc$)C{}=+p%hCf{+~jAAW@pK>ygc;SXh zyz|px@?NR9Y}V0To8=I)#m!jHx2=EAuP^1iT(k~~orvxXFgH$+=r14lTVTP{ow zz`J54PiE!&>>DRjhhumdZd)9FJuU=so1FjpNmcmEuIN&)c^dX8tPiN%Azp_pf5)_a zUK;{(Bn2XC;X=Noipis^81!Z)$c*rZ!nkJG;hVXkFxrl-cvPBfg*xKWRP}@|zX6<#x&lng26;#*LRMRL8-qQ5szmwPNmIU@X;6_G+=r=?c?L zH>iiwPvemZ+$m2QpqNgXmfnbXvn^iw~OJMB?-2seht%}fmj zdCCGAhNDZRix*D>KjPScGjgJxeP;fb&P2WJb#iIopY}pAY-8vXc~j_v$V6tRw;lz( zbHk4XR}iCkuV`W>**ZBq=i1p*$~hjlkr>T=ZGSMi#=h_$l5fLeTb~o17^SV6dUI1J zytp$D8ujqZ?XbZxWQD~JJ|~-D&VQo$4L;BR`6D^^U4GBsZ5`F#Z~yB@_9kPbE8@nn zGmdiGW7=NQMVmLmOQvLn=g|d{WT=iW$5kToFf;=EPr0q->&59ly!pXK$s%i<{yQiC zZZA3N#&P;#ePVUQ&db@RI6&c*FFLM{Zwdsl%n3~86Vtg7KkzeFIYn~Ze>!! zYs4E!VPr#edO(XPm9(k~1O`!!I5#3c;%Lwk;FK(%zaO?g`u2Bk$lFx8f*!aOt5Fh} z%cOOYzVGdtMc~51L9g_Q{Fd*~MSp((@LOY~zvW{AM{Lp^JTVsA$jbs!q5qosseSQ^ z+ol6K3RQZca_P_5d9rerQG1n5Wge&-1C}T=G?kw3-t%gqYp0x$KTy|6Yv5Ge$tE{8Jo{}7 zPcFqmCOM0WX%ppmTa>lSKt22N)SUu|+JmXV0AHD37nC>Ml*bcPW*K^?R#_m+n^mtl zsO+S-Ik#qJ;&=BUzPjyWsBZ z$FN0OIXU6A^C7!DuK;0b{N(o>Ckgj8w!GzMF7DuV~5M(KW)5mIi@flgc6 zbMDTm5vPaON0-v^OhWjj=nPE_e68D+tCT0ezxa$i8}Vq6QHrUyLd8~1tbY!0#5F{p zrVADMOgyv6>z>zTa$1{7FOlTJdE~l2t*sGvjU{C>{@ECsmFN9@_LLoc#p|o?zTedL z-!|;A=UH%Dw}yjL0h58&3T=zpPpdjP01Eb-q@CWbbXP`$ioO`tY-SZpYVVnN0|lZr`X+kMVL4il}~Zut>ysxC>m6F3Kr zf_|@qvyt;MCpy;}fXVaa^ zKJZE)$s@ZdA5e9w^O@$r^8|YKnevFEb9Y5l`B&3dV09~`chi?WuPy9^vXwrNb8Zjm zQ(zTbouCi4)-}p%x>B*3DWY>A&A6ULY5O8UnmE_cfDi({ySC}y?wmrAIO=}%#9ld@!D;mb&X=LP^2D`KOm@@0)0P_?a9z&3A5>? zBz8fz+(ecy_(a|k){d?8y>H~vgWA{y(9Kq(xFbTI=s^!;=}1wp6s>;Ez-H*MEeIjYdeP;XNvBH=_z!wug$8~kmThF~nn^y> z3a;}?VX1E?-4r^YLgqT7^fs-p6z}srrY)!Y1&*?jxe9Q0_#1xc5Ei)sm#+}K;b{d| z^_!J{Cd(%iSPh0P56Kj}o+62$QE8^|D)U2W(I)BPs8%4O#WH#PGw!ddL;s9n7diT* zhutzEFS(!eC|_;&{mvFL`5)L*ce}0Q%R$qZT6T{s$o6Y>NWRe3Wsr{yPwE&#iuU>^RxAMLfX z5zY(DlkW6xnPOk9%fW`7qs#F<L)%ML$T#}|EwH1M!qF0 ziPE!+Qq=i^eIZb2hkeK=Bai5?Ee%1J+F3)QEFTXzmLB9;mT0vaUF1mO=y3;OSOUH^{XPX!M@pz$^ zYW`t(b%@n-Y5$v!eI)-SbBn8O+^lknEu&y(9&^a6LDlE=v6sGMZU?j>Wzxl@K=x4- zk}`lpTnovkY~3-@F;Oc6vYdEf$7_291qAzg^XQd6M#zO5rh$FT(G^9kQnmoK-9=fP ztdow1tATR0%&(0$gce4gRxO$0B&3ZMPor8t^GSPTjO~~kPgj2O?cdwiMexZjxN)T& zN5#Z(=qzefYzl9ZLg;VvOxP0P#-|=?f%F&;%+P@Jk!T{S>#S4B6EkB2z!>h@^`!N~ zNmi?}d}YWof)k@Dux&!pDEQH=K1~}#Q-H+IWNwTGi?L30mf(iS1zRDG z*>nUpHp>aeILMaJX@j0D7T^MS-#@7V=9Oz`p>*3K`G@T^PA4${@mm zQ9LTumT6nWSV41a#_cdHv)V2HNYNutpPD(XXx5d5+ou&p6#eyzEVd_~NVF$l{(``S zgPQp>X1AnXog(OXZHEFXsB|ZNu;%UnNfz|0mq3&^-ZP(RpKH*xsdgwV@0I}Z(`BDr zdczM_&g&yaLHpd(bovy$rXjjAs)5-WelxURojkpl45*fYGIW)&nH-$I)BC++-~00s zEpVHuQ5~y~T#y5$4SWXPc|c`~N+lJ(85)$Q`Ok*veZGBO%b4>d9SB)>3SvV}fKKqo zo&(eI*L<&d!9MkXM|`|X<^t+lJ3T-sB#<2vzuaJoywtNGJkwij*VDgpUOc>y*u+HJ;GV^H(-*_huCDmior()NcC@Ha_OcjkiY}y=~pTn?x81 zHbfYNNSp}CBk*tZH|E7lc588CT`|vmurX+3&;}t=*LBRjO8UGq>3GSN8F};x71UJp zEIc4j4oF~=RmrM;_0hTUB1^I)R)W^YnV=gd%v=UKpUY?8`wy#`xjTQ|OmcTJS!u)F z|0%`xQltlJC?UOP$tax?qFVb(pL__gmoTS12R*Q&-@;IqB&{$>_lNA(?-L7I&W63GDW>As?TekA7mU4Dou*AkY07X$l@nEb<%97aO_85 zHh9?v_f1&^X!@VN>?tcut)`10bO#+RkOsM`YlVW&HZxPmQEC6Hi%mQbXKiir)`%rrhW z@eAtc&NqNoMqV2=;8Dd~2psU(KsHUyrt>0qs7=!vqI*<%-tFFXw8^^xRA^0#W8Pi9 z`OFf}GZA%)hUkZp$pQC>(fQSc&H@=<5WVR7w{!lGZyyrkh@}(@N?+}U`rnIw4biuJ zxA|TXX3p3oJUe4R)e!x$xJ;Bft6grQj|-gX9lYSz=l4`!FqjkH=JnNe-)~-7`-)u^((G39j}#Y>y@X0 z7-ZOSwBfs-y_Ndb@0Y#3;-B_?tMv!4!wxzSh__#gPEptUWYU{wE(PwX6F@(IBdTA$ zUwWQcZm*Bt^@B^%$Y+(I!J(b<1>gs5$Gj$IAeoWr-xV45POE;Uk=cU1F36xvNBWF2VqB)M<2QdibB28!||$P!Ig^G$Sr~c zswZ+hu5$*fyzS%-9gP=UL*qF6r*~q|k6w26-cDt@1j(OUVamIsHUk-Aw=b}V)OhX$ ziGj=zXK#33FaZ<*6=HFW(^8NGF*U3jOS=rSC8`#{YSm+XMvN$YA<$TD~z@lV%pBI*o>dXAP zm1&BD0`ozwj}bP_ma9<6qF0&>wR>1&3_dJeZQ-InujyCk&`?&>@6{t{ozd^3Kc;P# z-d1)iGXfIAk%zugF`z2*t5lTFFYrlI7#8k{(4$VNB}?^w9(^>ZlYSq&WiprpTKr`o z=WbGAiEO_*O`HRr7GoW0 zCA~fhNurTE3#-~QH260VF@oL>h74Q0YX$LCi`YV^hLQF4ku`L`*EUcHgNR^{8Y?=g z=^G3Rt>?>~rlRb8Tya8$*BhT2H@;$D#O1c#3mgSoCB7|U_}U{WbT)*8v4)-d3(o85 z@VnTNIp#N_bFSvL9iwD4Cq@5@`oHaqh20hr9EHKB<-NoxHIrWDOQ5ge_2tR}Q6W|9 zlfWJgwklfV7v0?RlE5C*8fk20t9`YQ!oHS91$Kfz7CaU_fV=L<6JT@HCtV}z6IDzH zI_bPwu@cjQ7R{zW=V|&lf*oGVi*um!oONe3?EG|H_%?s*TJ)4XyNYB?CJ-hYX2gpq zb~ioqNY$Yt{ zC*q>v=BsXL znr^>5mtLYP2U5Lk^H#b;c*^HabhmU9knjz7bW3lC<4>bsNmL~n@TmFx=L5e=;$kp3 zI`tzL9oHg&pJAK6#rx;M)&*x@XUS5M=f+VOP-Ppo5+0;j;7lu_Vz8*j+O>pQ0!7Q@<0!JyVVg^jXtSdFhj8?NeK-7nnj}kNTuyhXM;6+QJf< zm68s19i7T9#_CPn2UW~V6qZW$ZK8D1`}m&|P&IbsjKtEjGvh}UFXX(kIOgVnePY^e z-CT~0c}GZk_=@O!fQ-H}IDf{m;6$d=T&1N)DN7WvwA2$%;C_fNIL1Hij*Y+z9FvOv zBV4!GYF0}9FQ`bH8=I9lo2tRb6#I}O52%=AK^}cS44FIh7?&$mr71GOv=qo=nM0XCyxUBGVbvm4POJv%z6I5BO5Pe&LqmoKhW8bgzOmW1Q$9(4kXCa+Rl z4PBVec)0E>&6~JvUB7o5Z4@W4#i|Q$fQS$NIzzwo4&XTfg!4t-Gt=YRSGl@MQ$ofJpiM3%5Q6bX+~bypX~LS0_Jmp+~;rJwlhiD*^-$w=Q;52&h8 z!6{-1|LWs}8Q`Zck?6A`H%2*my>up+9KeXf zm)waDzwwTBl{@vz>mQR7ZoJC1*f`Ji6kAJ?3sg*}vR|E|egGtEP|I$~aW5rz1Z&x? zGqRvOK99!OpC#B2MeWe0Ip~2w|4hhk+THfwmV5M1d^%~!G9`f)DHKi+9CK-qXIMd{-T-#!HoFgX(_jx1`9B8A)P zj1mNVCF;-OB74a%xAm+!;N$SLv9E?IVt)-_WZ(pyk7UsjBqgzV_>f#dIPy3&(>i{ zW@h9LT?V8gv*dT?T>-*@!%#Zj22AYxNLd}9UkwI-|;%5OVv3aGq$rkW(3&| z$g$=tIW-)DK zImfs2efku7V<<)mvE#Qz@JU4ZTk8bS28?prZS*;5C$tC~wL4gh3Sx${Ev#0FQ9yH{ zR-+p30+7c_Ov-9YK$H^*k4uC%btW z13>IXaLwWWHJm+6TLIGh{eZJ%>12>o8lFGgK(Xs6vX+X;*4>?>M`75MIcvPJ?;pZg z1u~3ZOw1!3_5d7xn!|3Hke9q{zxxK)zx>%pf3T0N@bPxtxLSrI4zrzX5aPmQW)(>e zFhyX_AP4jYlRV8V)aVaY%-b5iRns$JeZh``{PP84LLmOiD~Fn9zZW*NrUYbfB1NOM z8@Ta~=bVjwGg9mkiX1{+4bXrBUcGYJRnVn8r@0{;@c4|}kQqeBWW@_^$WSlJ7!uEP^r++50gwHG>!zY^)G30RmUwTsFy0e*Vd?X;m$MBjJhNU{F|8fK z$9bZi@^0w{XQ_elhRNr6;dJuUh~-m#YxBT z(4F&CUoZ;yLmzGP;kIn!OMOo45m^!PNB!mBk$8S?tNSjkB-aM$TPQY#A{(ffR*1`8 zl~&Dwp6f@d3-U_3PP$Woe8AX9RQ$j66W(sX58*F>r4!%&+RfkQe#;7t5AqM+C8tMQ zecadsxnl!>28snihB_qQ!m^WA^?xssK&1#{!Po!(q7f6z@&#TJ6 zNxsBK|45VulBa3RJyo*qRz!xbG_s0mRih@49y^rsW?i0Isy2zS?LR}45zt~C4AR$I zXI%382pAagWP$83w6;MmF*B%}?olny(QWfB1+6I5+{mMs2oDEABsfFY0v&EBM_Z6=WN~)yR!qVeov1 zA!ZN77EmOQiYfQnF8h6rG?kp8pUzHEpAz0s=Y(&U0pVZt(_dEnwD6t%Kqk`wbTV0* zV{-Gw=!&h6%wAX@o%Z&_MZYh7>+v^AzVdJpDzdH>xQO1WW7=%9z~W&rw`k_jLeU zifWo?SdB^VPn9dkj?w%VH#QlSHYQ^q#THYfkc!zUFid|Ca$bB;a2CkNJ87)G-7G5> zJW=0+>hSIvHzWqlAy25u5($h66 zqqLQ25;@{(jDnWuo#Kt~QSHa`K4EO{aU;BZ?f?DvMOM&!=j7k*B}YdCjT<|?*K9y@ zj$)zZ{uC8+XVFbL`Xql|5zygRK0i;=BE~XGeU|V8ES^NiHAP+G+a$73FF-hDMF7?? zd;>hrge+&g}_alMddyK< z4AItWjF67jV-9-L^mcW6cylOfj%H{q$~u`Kv!ciJwo#kLV()=PMx$OcbTJlk4&9sq zEeyE*S}HlJ>n50_bZp=Af?0A6T3+y>4!q5cmeH2i<7>75%f4}vPlmva*EeXE8|D({ zQY`GMvZ$DQ>Q84Udl>~ap-IAg337!RLmx((RG-l23C15GuIt|rokusQvKW&hb;ixv zSF~s7yOLIQCv8$?`0}+X4ounbOU4eX;iEYJ^`ngn#@-ip+kzMlCw&REPGF)95&G@A zBfxlAAUhudOBU23Sjf_sN~-)@)$Phtp1=d`;%Oe5ApX-mbf;^7^0FeYZ2z0sRrZ-A zj^xZy|3aokd~@n?U19j4kQ^H0rTKxlvfqhDN1$E#2}r~4cnzoG;oNHr*NA$daC-eb zOcdOZe=IH)lrq`m;oK}pINg0|_rZ@7wGH+I@E=icjoCwd*B+oa7I%yvH z&y($vKk~>q^T%tR+}Bv=_N=@pwED%fQonbatZ?Jbl5CrWWfR42q)3t_I%jrA_e8Am z)>{N}<7O@oF?%ukBRVXMnkcgZ;Npo`=fKf{QFtMCQh|v1p*`P++b%mC3?7*>le!v= z6gs4O)HgvtG*eaqUanI-Q3oFSvr8O?#j{%`Coy5{Frd-3Fck%HWqOrF?$ zE$>ro2SwVc7~D=YDw@fLAael-mLy_o1)qNr1T;`(tu|D@g;@^u^TyC)+6U^ZQp-sY zh&5kSuff)36f&uU4vkvqFz%#x=n5D0sP#C8WnzI(48?75cZwTQ%ivYBx_vJKr++(? zp|nTjctdS|6>~&;L7o+kZDJXk-k^kVy+MTJQ*}T@mq%BVJb%4OzE*HrToHUozG=b0 z%q-DI{(~Oaz+4(xtip}Dw4+xT`1%c3$C@ts#XgGoMVzFwLLu8_Y{%HuerKndx zUPzDE_NZIT)wz#VkAa*`kCw7U{7BWK?*AGvGU%}3c8G?vH1f7Drl27~f)wZ4?wLX@MD-|Y{2}-=X#ppRDgy%DT1Q}`)nG?X*nyb4^OlC5Bi(XU;FnFF&0k{IM*Z&jW!wxdD+KT+8;ei{jn8m86A?Rt;YsyCW?iY?3+{!PW{E;esxP& z66qr!VC`|O4^|&Tvm4Tlq^KJ-&`-TGypz7?b!k3Qj5I1t3-VsqBapDjAMe(uzP4$q z0n(dD(h4CyiN1tsWUnoRy}V`kYYQ8gOY>2o$)v~%*QbQsQR~g4U`+*%vYy~Luy2ga z3O9!KO~a$UX)rpCyT~3vnQ8~CFBUXOal|~jZyKJ!t_G~*$N~zKQMZ;~2%Js&!1Se< zxmKM0;-h(cNQoQc?4k|MswfutAC6ElYeiY%8M-(&hhDF)W#g%S?|Ka$q4|oZ3S?=} zg5LuTZy!ofMea~!B7@jN8s7{scVNTTn*FjZOt}_s)0?LCksKPAMRWl1UcFI(C(*CZ zfm3)@GO*c{`&{Q-g(0rbbM!nfTQ%t$zgU%N9~9yfT6W_s7)NZ&L_a1bV31xd zDOBiD^st2n2Y=?E2LcP~h?zXWQjl@Z!wxCG2fmoVI2pWKKIqY6W_D~>9)_S*EX1Vq z7^G|)^w=Zl3B%3eULTP138)V(1U3q6dNC=G)`-IbIXau4U*`h2eLnVhUVQKTzkKa| zs}Z{P!??NR;TNV{XM+s|*YZj1l1XF$1i;1DLaqm=2Ot?atZBd|Q6s({+)Ypd{A6&A z64aW-Cxb0yb`HV^YRJdQ&2U6i@qHf90 zlRT!20KK1C7|Jp_*1Df1CuKbX$Zcrw({bF|(1z&EvUVjpTV1{=>51>)YtTScSw6Yq zwFFSD!_Lth9e!GiIO)syJD4I?&&I3oNH6*gsEmTd(4r|lf?IMddAuFgps5UQ13f&` zw3ETN*(-Rk5o_a!%uj?g!LL*ZkIREV;fk9{#Z1X)Y{!si_-6^f|b2=RC zrvZ=00+G?iarU|2obdaqeLbPuW;!@(2YcTzO~cNv`iP^_0+~^qB1R*{o7m)Le`O1|toI zvqA7Y-JVg(cc@<(10v>GeUnV}|1Ka0+<2eHF;%g-lWRs4*Le3S%aEF%|}%(k9BfUT^Vkd%b7s z;)>vMU{l9B#7`C)gvj8yRB{m5ix+=J@GT=Sw*e9Hn1SXh+npKHnQ>-!rvLuZ z&bHGj-WPHxsDNB@$suxyqJ+Z>Djq;lsCaUSN2v;^6czrT4+%DjMDswxrq0$^<;in= zgZaMm`F_5i{<;N44k%@t10&m3fkuzb- zXcs+5eo~v2mOBd`g4nA9H<~poM-0^a8Ne=z`;3sKm+>9rA6Gp7+D)!(y;|^HA9~ z4_f6=gEmEBh3~B8(=@hCu;l0DXNui4*)1!HIAzXfu6q2m&ux8&#If^X?Km@8V)Ql0 zq?mLHqWtO#`jR+}*P>_^CIuYw8zhjq#;87~#JA483_{Bo=Z6Z61pm){M;u~jYT;Xu z9GV@3Ul^MksfnX~@`!%g2#bFH{j;Xb<7|Sqb_^3>2QcW3+Cec1xG#gP4r+_FDONdM z=W8;2SGuYq56O_V&sVwxuq_*ai%PE%*ZpT#dr?H2}(!8+YvMRpSc+hO-&vLM^8Flja zng`2u$HWgzi5Km5=`q6yIqGZ3$~?4=xtuiMhd^E9F8KrSO7D@;Jz0yQiG$5MScpZF}~!c4lB#;VcJjPDWHdURdkB<9BI$S?3+U<`U=z@?T$Rll? zx!2{~;)mi7H62TOecDv%bB1L}p7_q7C+K^gv+dVo~@m|i|$&`w_r$)CL=|0T$wgwBe^-hj^6O|{vWic&Zuk-2!nZIWkL*|v-M%j z(%=8;>rMfIg^}^{*5C_{u(iPwHVukY;MYHkzA*9Pvx$B5mQVD=#-Hd@=H1otb~tGR ze~5M0n~XP4Pdh^rpBt00-^gU-Q4H9SOkgeplEsfZ<mKhghqS$%H zb8C4IKKiX5EXUsX!jp8_akO=XQFGV;#oVFD7nB;i?3W9mQCg}=cV6Y3O=nQij+!lj z+g`4RoLwsYfP38aGoc21z_RI^3yL@+4y7R}l4c=t(2p`JR(e-@BPDSH(e%p;VF$f$ z7L?p#|9zISn9~aVuh-`xw+OcRgR*%Ry?;)-WR-I{^l`!R1j5vbKByI?=@s@vw`^6Q z*0D+261X)C*dTi0-g5U__VZ;IZWD_Mr;n1Ev6rM^a zN;E_AM1Gzq0aOUF^Bbt^|Au##yMIFZ#LMiq8H|;Unl|faYyV(cs9?8oIx`g%7X=$6 z71Tk+Pylo!dR%lH;zsf6Xx-V>oJgMTS}I)#bP+kP?0uzz+Q&Ig=f0ApsGypHyC9Ji z56vC7Wi?PGmr3Vv?veXY;Dr1*Evj8~p2tScutUB_z6YMSp2IU~Lu6{7o?&1tz03+f z)amT)+Btf&wd$RjktEfQ&6d_^)!IWbAncb5Ipv^Zf(q&!eN50LOLyu7=eC4%5tK)B zy-Pzz*<1tbAcR-zQ3%~1sPTwBr=JO%px+EDE;F19p`jY96Lj}v5L+pzTLCeaf`HY1 z>=3iD>cRVeU;4&roT$L{ygN?7&noYbMDRYYqR~gX?%ON>@RImT>GgT2V*90X7v}*1 zN*VqzuRip?ODhz(99xp$7j5GzhDo4`6-wCN9Qps=>iD`_mO=RiVT7Ou`qq(t2-;Aw zeIrG&mo&;obI$9Cc6OWloe=Oo=zDWZeAdfh^MzGS$f0%xRCKV>ND~Edy0yG2;UT{q zD0Z<$vC||lV7Lmgo#JQS&&smA@X|Y#b*4+fWS6BgM+B>i1 z{be#CUpr1pnhE*VdOULP0n*oH*iMG@c*73(M^{Z}n0Pv=NSlFGFsOWux-e6bqnt3P zCXbt^j!@n3dK9JiZ67&5Iz@&Z#tHM-ar_^IEe(pIq9|q~Mb=U3drPWc?GDf3WO~)n z7u09pn92~Y!3X~AX*ale3J=ZOlKp4?aC@E}5btPJwWPp~H#^6TCclzmDkxG;spDR| z0ztYK)mryf`X1C=qqx@9g|+V5z)SqYA?2YF@&W!KC|{_61YuNQoT!f0U5)w1ab<%r zN7*4vliuT8C%T_dTN#|PFr7fWBn-FFb@X}=QMtswPMYaY1Xcsqv!P<>hCKV&T4*v4 zmGKOP=%bzhPd$7JM*g*&q>U%7W5@OcSnCYzNfE_B;Zz=_UO{rWSiBGAx%@-UtGtm) zW3Qlpafxhz8^!s;2TOI4M1SRst4=pvDi<8}>*pe0)Ml6d#YZ6D92?Tg!E0E`od!)T z7;&*0bPT838Za^(&sxXkVdUojho74=)Y+{|Yld?!*Sk;{>)OgWPa3(O3AJpkVUx?jjl3k!K6=QhUWQld=^lQkTT5`G^t5k|a;twA=L*zc)=D{+a&Dp=;%?ECJo73{E z@?|99xyeE8GqOgx6a#w48I<~#xO&dr`87bPv4;zh7txs+$OD(DL?!hspiN+Gix;eSj*Et7Y(jc4y`|%L+_DkgIRk`wgHm?uJnhj zkOVjX+k3m-GEEYc8qGo$#ekSeDy1G}=9YdTMNNxQYVD5s#iKmQzzfg6M?;*oPrAiL zr&A6}<-1)I@Oz8#R0y2u6!)z0wKVfDoa6^=X4J`_X5~fLZ=2}LgZ(W;Z*)o=(oT}K z(@B;QPPS1@97SR&^&ajCMWSmKo$OmrSNar$tPicD8-@2JO^QZ-QD}y^+FFjIg_FYq z85WLco!eL;!+tlbfx?|Cl^z-0L9-IcK|7Asd}ajM28yYt$Wcn&FVE#*>bVi5jz5=O zpNBi02Q*3;gH;=Hj^RBBj^Uy2Q4L9g5eL*CJvHM2>4Ub|M%n0XCGwJh-L9p|6b0(h zjX31{7D$R5dz1qUfK~!u=%_!CvpYqpiw0vhp-DRAT!kcdm420e9dw@P)DrW$6;Fp9 z3z$E-Y*>8e>A%BxS6RV)TGl(w{cr2h^xy5f+)3wj0)*HG(U5+MxlNH?N?j1%DPKX| zm1V;YB!(0UHO1lAh>lG5B-tH&T$$hn0yVMD)h~YP3~JS&5{%WBH9neLv37R3`#ncg z5{88=Vu6k;P{+mch!@3+e^K)hW)VNC0=3~Xj{z^uXh)jYkPCX5pxhK8NEM;JFw{=P z%dv$Kb)gDExbaQ@vXaTt<+ER08(j8Em#l*7 zC+VUF&M}V)s!NvVeazz+Ps_^&JGp(5?01$xoTbGYyM0n$2v!iM{;GZZuxZP}a}w); zZK1)s97i#+6p5nL?UGbck)TMB7Ib??72PLIo;@V*3vE$UyG?>yccL?9DTpRI{zRYq z#y@4%Ooltves_w*EO%^$A6<6uE1_&{M)c^L34mxfT>U&nf{us-3 zM;tP^nhQdFz>1&ySO0Qr%)nQpS?|dF42+Xtw%@O5zKr*qzb`%Hqaj`Lc&=8V8R8<5 z!x!@}1cGumszaiKv(2%OK*pnbp-sR%F8o*`eB3c?aWc&N=-RJ4YqM^7IWB|t%A?}d zFDLkE(07dmF<3BWkLKIRW|^A}HZyH*3|pLMXUnp`^uD3cuP~~Hw<<~cbCbfZHd-i) zDF#Hb@+tMF^yY8emhV)RiOa-S7p~ylo0ko2(JOfU+})&vV|?|h>b6@W|LAD-=%w#r z4z)Gx2f_URobxzYAJD@1OOU`7rmuby9D~ zAvuszH*lBQCjCb548-V_#R!)@PE*eE83WE!Uy0dsr#{rTn-XYoR z4SA^`d?rP42U6*+oPFLcs@m_a``+1q>ia2HL7Woja?Viq=Jk19Rkgg;_M<53gPjsA z-s_^;mmEKKlYV;?3vQ~12aD6+R^E0dom>uTQ(X`q2+g89_`0VH zp=vmbzAHN=#vo0KVn|-g>3~`Z)K?k;`gu$6XEA1;w%)M4W&w8CS)TVkZWu8oeYV@C z1TzHBsAON_mI^W8Zn~LH3mSGP2`vu2L$C6V_KcgCVB?5$EIh{$n@Nxm>!CBjR{Od1SE1$Zm}eM$SEQWgkn9aKxwgNK^N#S zRxm&Zw3pQIu~MoV*q|ZCd(;n$rdr_Y3PHL1)rA?(BMvcW9j|lW0n=zUUX%L0wjuES$G?$?XN(Ah%a4%T_`oy=Q`Z zGJO2N9|hG|gV0!?GFq{Z_uUHCtRu_8=_q%{BD>Sv6o}Ak^2G;z2l&k{m=y%RsnHPq ztafI^0VgC+6#M24w?^Tx3?@bEh;MvS(dmXPR~>Y&7*uVXAh_&Y%;|JvS*J6VQ#bx3 zjenNi)Y)ypRcqR}+Jf|$lKqb+ne4G+19s8~?9~)gNs$UleU4e~TBn>2x0tD?6$Z4TK*cYkvyeNo&a%jWKZt^Zc1GV!oO!{V*H zMw!+zWnm(J*dd!-9cZyLGe%hIF@AFu|a`KU;~3vPBE~y6jACzItzA8hrGJL z?H+RK_0e1*@$wEn#$lp8zjVh2gj&u;Wejh~DNWQ!LmOWK&}GHTk8-Poy~3Q)`i*^3 zTyln-;Np2Qh7ja;+9j~*L%|)S>pTpZ%`VvB^-52>TYZsvauO$s zktav8?ip4XdEuAMw>uZ>jZ4eFuF4`6c5GZ)jf~4D6myJ%l09{VpxA8%xdW6(Evkn< z!1VP;o$|u)yRy5oetDW=g`iNlJunvZB;({-$K`@F#Tg*~y%?Ivi3q?uzIbVMSXFR8 zw}Hg)Vt5@SVph7iC9q9()UOT#GO50soFSY@-=>Q|d9s7l&}HKD%H7=Uz%RXa`t^Cy zj(A6Z*t+SHFhCQT_mR_J_H%n99?Vr;*&l@+oATqcf*r##6DF4Hs- zo34G`J>CN9S$`&B6`t}*o4dpU8Rmwt$$3`Dc;WcHdmgXq7o~^X^fyV>bOPx^1IO|d z#Q0i4Urv>Uvm~C#rTElK(Z?O@lw1=|OTx)C$qrBRtJn7o-6H zWT7sQ0fpJ;NtZnOn-9M?;*b`yB@_d}$Pbh3nGy8RWp}U!sfMt~SVsxo;oJoA=5^FM zsxP!h*(A*q;pX_9^VJ#fn6l3eJ=9$BI*RSAvzWK>ZM>ytSh~XPz(rY7l7a(T?$<)J&;~3@{I*fy<+)s<*>s>=Tq=^w@ml*m`A$vumkor zt>ASnI02eq>B{T#n(4i=m7IFtz20l5>&93Gf^o&6IMSrNUUNu&Huf#CcIHN(`1#-b zB0O(0LUH!{bs!+}WU#tE7K$sBHM>N+LpxK2_@?|CFxOuT#)6nAP64;g{U~4pQ&9!1 zSG$Q%f#tPXqY#k;bI+fzk`s0uT)ANc zfb$ds)J12Z_Yov2vXpI5P*DS93m|j>IwrHP1%C-k#!+R85={cc;95BG@*z5NaV->W zws2PaWzmJMeO^t{s=$w+h9XY@H7*}77oUWOz=1L2_y0KS-*$ zm~_HQnq}~xIp@tfD~u(pkePeTz|K6B#%a)-cA=v5@N_-ARNe1r$nxnxn`daqV<{$z zA{!}nH;wWDqhvujG`hAekj-nHBwvP^f5g*ITlUdEeGFS%W%be9?@GSpqtZ`Im9*_P z24_Ya8>_)lEWMwL-|{YnCe9O!u~1uS6Y~f-AT=>O9f#{_ZVr6YHAv%7?*Ry8Q} z&7qUzeO{U(-y3tWDD1@SW6l~Z1B>(A32axV!9d_+*~%18FidL&`OxU{IFG&l#lcYy z*fWbhpRYmD!(_J}I);av{5sB{M5hK;K$$fxPaxAWxSO#O6FB|EFJR^bE4WXqzb3zC zTGPO$_Qih51c{jnhyux=Fxpf1H;N&*sg?_Fxko_?Alm&lP`1|48wkpt=5e)-jr=m% z5x7?CF$^k|D?l}P!ZajbooomsiYY2yg=_;{!E#lMh9c2O8R-cIEi zp=k%jB%lo(MH4D-m%C$6Hzcx~qz_%vUc_|<(?{rNYpy`93A0COFaDi_$wl`58nGT8 zvXB3dS0;R@0;)p|4V5Iv2<%Wg2qaNihw`DnW_w0KFGYi{b z+p^OomZ#f4*(zsf@=a%NKq$&&2%EfQXF%pQ{-z{AkC1EMzga?Z*%=V~_2UT;(Hr2Y zf?~=kQi@!6I=-Ou!pvYzB(%)6OR)Y$ql?p^QHg+rb-Wxkj&(IsU}SVnc34xDqNrSO z9t3c^=^8pynd4t0!y8MvjnX{euQ&zV-X1?}};0flzkrxtb}ptL62&r!j}b;C}|{jc#Jx?WdyR_N ziJYb-EkO8yitUZkI{LJGE_W;VtXrX5zx~xq23J-|BLtDWdd_ZfKa9C`j+J0BHi_1Q z$JmJ0`Dxw|egC@;tNrvwr=+(SG_ve?SvqQDM)pw*sFM~`>U_U}x%hvJ>SjQr>^`{) z8=NyLaAoO-VZAfkRE@7)5)aV>?x~^%!7(mMd8NVnG~k}%*%Ek+r1AP?MZS2PB`6Qb zacooJQ(1z_Svig%zaz+X>y=s;Ce}lVbyqR-C>bYUN0oj5r_=a) zCTEusni45y8%5$Mbv@9nMTI3g-3r0>>6;6RI4H@a6CxdU&^qn|)+#(}T}(Xnd(I9p zPaVpt7tI6aS1GAOGxgx;UOfL2NwQpH zI~t;?bll_GL2qMVuag!O>9S667`PFEZ?(#ICxZ{H6YL9vAReA8haI%0QQAuniEBM< z^O`XmSL6}ka^6suLm3rri4Q6<$or|w zSzqirL%$U^PkUYBr)nPz6TuQ3w!p!-W3mFr3rhnI|9LWn3U(YLMXN%q|!eEY|Ip{KkSe+?ws4^WO`KpOQuOfHlchw#)g^PC-xxr z&<7xQJmP?*-MQj6MZA0}u`o817Fn0Eo2T$+^cO3>4AZ+F-YYlell|<%D0b|zo;7k* z>L~^`M29GK9nm^xI_)MooNUo5PQ<)K&vfrZ&q3jGNt>z<%7wSijF?sH(j~8;Tf}`N zkGqpz1F;jluT8a^ToRXgZ-QR&mBDwOTy#L9Ph-=Q}EXWjDCV>?SXR=;(dc?Zj>xpmt-iW$<{l`l)CP7-;-sL|gtn$ZQLjKH zB^&v#I6IS_{mIPm)1GDJrcAq;urT+fF|+eu+NEUsb7OWOSZffy-9<6bF};&guTk|t z(q}KZ!N23AYaz=QYy-p62p#f0Acu|D*N3J-XMKd=nz)BwEZZ)rqnDlwxEQ=!d{>$- z!jnsFxO7bpSWkn-TN4`^Kwjgz-$~@JBEar{c^F}1xKAkO+UBD z7h9e;g(ihT5dv0c=aF7v;|$m@`|c00p@A@&LQBA5;BS~bZM&CA9hLLE_4eXVdyYht z40iUyjy>91BYROwF(7}mn^Ip4MfvGXOhmwK(hZ6=x7`Z49sHXCN4OcvD*hSwt4`~~ zQWVFOz)%-B!0n~eI0aD5bVa_;_oygQqPyNqpX8(9T8^Xv>U6V|JwVIZY=a_EmZ&iR zkkO&l;q&?mU*9=;)U0}EW+X|q;}ERY2nTy8rhp>3lo|?qVW+D(3m$2^Bt>~CRMP{h zCx?77SDZXYlg7b9;L((EL_of{gT6v?ImvV2iCzUhQyQ{407?@iz@A5IIvkPS3B^Nu z==ZH`RIG#n0EBg6z{-Vv;d|1D@BDpLFkvlj4wS_H4Vr6vzMVkM5?$%RU`U-P(e+Z` zS&}EZ&Bs`a4XB>XO*MNknH$Nv=Rch5|Nhs#0J2Yd91KdgY#p#FW6PILDqzimHi;%X zD2uM3>wNM=wVW4oSLa=2lW@#}3H}BZs7l zVh&PdKcz-$23`BckW;5zbJGQ#4;Yutn~wX! z*n~~&6*~j7qab&d=|0HJ1|mzDBtZ5X)XqbZu~BZV5}zLTBP81M2yet;Wk4U@#4iQ5 z>~yl3r*$j|I4;rX7$WzQF3xDScDJCG_h|mz`6zIR>w&IV(dw&$^(JRxlVI*Cc9Ubb zWXbC{{YSF9-t4@2dfFM1I32#ph7rv?ih)dQCWfg7zJ+?;r`@p&W>CCc^2i&r_}#J; z#r1h*vKZ3JxeD^Ck9aMr=x-kXEq2mec7yBC3zOt|m0LwODN21CZ5i5{99Yav_2kd9 z0*n0yB+58v%3bte`AKbZE7@YlYZTDi8pP!@CkyH~eZ$R~Rn<~jO#l4i?6V{@tr|ZPW$WU0F_)F>ad99$5GwiTNl|l-+!w#Dy zrS!1FP4Q~iYhtZq6!Yn#Mq#Wg>Tg{Wr-@I!a%NH5j4J0m@AZ((*1f;h5wEoR9^tL_ zJL+qA1?_Mu-7Cc38h_KLBYVQQ9M?5sb@Hh|UfX!+Pt){BQ)cY@3wh)~l8mNk6UA(x z$XZH0^i6EILJ8*U^ZwVEo}Qh|ed<5gb{AQF!R@y!DD2yBeM67tnR!58St7-DQM8gLLok14%sAyPS51yZ40MCb`}@biQD0 zD4^*$5Ug22hPZ>|Bxv9~-8hVc%Biv!NR@b^_dch=>Tg}u`C(rC|G zFHMx_%%Z~JwSt>+Yqf_xb#9C=T8JRyho<0_qbljgu6hg8KDYH968AN0>nJg@I++xc zPLUMgev-EYR|@mEebRpJ&LxGi4DMQ|K53gOUY;$E;&hQLdMK=l#;ezYOTx#lrzi$p ztlj=I&Lmn0h9~@rafY($Df8xJ$w1Y|FY96AAX4rlS$4b-9X5i*UWx&X>;V?RC4VUc z?^HXd5~`K7Osj8~EP>NKJ6^7xlMj2RQ{v`8?VKFPqkdV+m9Uzm0oz6ToNmb#c|W%# zAVq(l^w=?OR=;5A zz}%;p0gBu~?L7$j(74TRm(5y}mJPOFbbTQ#gL7)qf-2=V zvhoS@)@dB%y|DIA&B>;cS7x-L_R4P47la+4oQMP|;9}xO74MiM!Txd$5~$?Rm<;;h zQxP&q;@M41a%grCL{)TDyR02b7G`@=)4}3>*xt%&I_!6gb!DUCk!dTv-4=bDX`7E& zkVWT!CAtRznucH3t8a#4CxAhn$6MeAQI^XQkisX&y56URAE zme2U?4a{+26leX5#he?jCiq2LIq^9uL>n6wwxE6L{j3m8 zt-C$^JyV66=cI-cSeFfU^63-O{hK(fQ!iE{Br*4gc8^zQ#%GxP{vOoi$E4_DRJ> z@%4Fas!Yt0h-*BT3$L(33Fy&K{r`Xb>u+)YV|s3&K0)N5QHChYF`Y6qshg ziCM1f=3iT|GN9n)-7fEyhsFuAg5Hapd%-==J6)9PUL$+N1D-(mdnKoaf7GuSXg2ET z(R&k!HW2O|_Yu}O8F&6{0G@H*hwaJd1Uyk6`)nN67oJ}D{`BvX2M*);&vpxbU?wv} z`bLUbN0HT(x=Ff1fLk#gdonU8kNsyISXu%Xme93iNUPjq9=bm-dLeK!MZor(+b3pd z3=f5B1H0(8L5V(<{1qgfz7I|_XzYMQRxJn5rqVkUS=<#Q6Ru1da##Y#%5~%BF*d8H zK>GJZp4-9!Y0o8?#NI7#=B@L)tGuW@=T}Tedlt`3Q4EAc2>PT0+^Y}>yhpT7>7oi) zOlx_OJZ&IwOi4Q3A_Qlo{UKvdM+HWSdZqabJ7vlK4U4S~C$_LL=Teymz{IZn`=3v| zJ(*0C9rw4HA!^7M>|{FUw<%f_ed73_n^3W^N4D=3t53KE(_s-NEEvbyx2*n|;?o!O zPV1_*0Y}M*gZX9Y>*bLo8+jzr6cb62^^_VTzsF{DOH%3VAY^mzrvDcbk1&mvAi~R9ac}jFI%fzGdT};1sEhjQcgdu)S^G4WpvT<0!s@&wkTJN&yz%&b@^>>$74z&iqzhEb2C=p% zirGk!b;y+qc?j%#9BX;}Kht3__tVTz{&*|LuV1;C`nKNC{CE2I`#r;UjU$;uJ;Kg1sKsNT%0S)&E2qGI*v2&Qa#Mk&UMNuWZCoc$DADZNuO|O9FsQOT>yu3O}qKYHn zU4!!~#( z6q7-L(w=%1(k~7?0CGqbzvVF4AE#%+o(y3U32SzYHa(=eL&FgiJ;Q| z>MOdo>O4WeEGfu_2%4NV_-bOUeZ&eV)10>*{f;S6%*+Rt(PGyUb|~uF*E{%$J_nX8 z7u*IajV5Upz1j~{eV7zwW%zOsv&)uOyJbNy$U_&b$&C%Q;auQnIPv+g_6!)0$8J9C zHX8n=M`yZeyBnJnH9&2;U>2Q~5|1lZ-dJS^E4eUs9VRt^e1V%TjDw}$SLUw`F3 zNm10(kiimXI^jK9#~r@={VJ#nzexYbo8frIhPaOv-+j1SB~U!w`I*>5N$ZUrc=Uhx=o+9-D_4R zj{5D#o=YbapR?Z*FX)mq^kg2#=c?!2o!>7z?T$*WSTHdal`7-SwgvbZ?=hR7oAJQ! zbBW$t72Xm4H;I`}3XML>G>U=GHW4*iq2>qpiFGBEZIF`fl2{!Gb-t$4ZOe^nl5?DPrLp_(<6(S z(dSg-H{x*AuK@JJJKY8rV7FCJj@2}CYI1PvTsCvIDGSVVVqwgX+4M;h zB^W@_6gl=NbLc41vN8Iga@kHMUvQBF0*nxWk)9HF@O8qAJ%G8`kX&zkLKEv;1s?KR zM^p+x+STjxb_b8jP*-u&g4Q`&HIZOd)JzU2Mo;^27oGc|-t?^dQ{`Q9(vAadH;s(R z1&TRGk=8L|(j-Mq)jhHp9?W7Q2WJ`$1V+_jczAl7Xs5CQ;y=;yUb&`w_8@5jA_T-x zivruJyJ=*jLc3Drk7Vze$}Txtlym~UE$tF)haGX~2XS)j(?%90?5l2a*~ILTotm2h zO6Dobdvm*Iqs?iOc7V!Jm;9kiFO9Ul*`WLy3)H2$9~T$qm0r$is`4wZ7i=uiq$A`4JyU22(QX>#WB2I zAy}yfvNfa#nop|unm%cY=cb?|P_y2_KOws@uNRWg8eHuzhOUz4(21@ZJowZF?}Ent zco<|EhAvUG;3h9Cuqsd&$n5v3bb-NkP|0qiw>FTm!>cni5ev30=%D+-6fQ#=>DzP% z^x>t_cW78>GwDjc27Bw;DT_`b2~O{8;qdzkYc{1$URDpIf8mc z8V|vAo}f@kzx+#e@{*N51OYyfbS(22rE@6eY*>grO`3jgna6+^LL*-O$s3x6pKShi z`?qv2OP8cAfyVwmaf(8BV)e^g0zU&4d%X9ecngqvE_ZE`R?$(MUiv~H+R4%otk8;= z!#nY@8*22u+Pzo!pySs0-FP`R0fK|&xQ@bYYcrk3Ym#D+9e+*Kzkd8H^n*iIpno6u zov#2VAX?kfkaYM*js9lWq`4g3Z*RANVrBPUICgsPKbkhQu<^$1*rhkqm(wn}z2Lyy zR8f`~D>1O!DN7LTiK}RA2r>+UPp&8l_>2rXB}#@yHM~+3c;~9HEYCxtR$!?HR%@Ua zjSsjgKN=DRshHG&EFgYFU6w}fXOI9#>Jor3ja^imn6#i{zEPrE!4I9Vr0^PC!}BTJ zBOpl_FTXwWv~;f$$H&G~)~?)?Exu+JS#ziV{lMM1o_ecRF!HbEB#qs2YR6fODkJBv zh+;r&J&#gX_{NcRQH)@>xW>=$xJA{=+wT#{OA5LQqP7^?*1axxS$F??1uq}-$XuK* zdM`70w@W*dCcY+JuGCq-)s}f+<|bkM^VWgP_+wAz?0*c%4$Rg=M!9KyHA%MPg%H?e z4N~d5DF%8Iav+udgDcRUwPnUSzb;t>@3ecd>@;Zx1+O+$0=X-#BoORvuyBbc(W(zmHoZd+2_e zd)a$ySOQdWq5{&@g}rjT_kuVw?6k5`ipTiO0eT=LYZgu^I;-h(E(fL3Clt*t&Au_b zTd!Uc=U7cxc+%4}kzJDo^TckN>^9N9mUv$=ZD|k_-XDMUYxN)B`u);(#WdqUC_CQQ znQ2^TlA?!;${RYj8C~N49X>&e0Bix^tbL1l0K5L~;|PDfP5I!%u1d0NIyq)!Q}$EL zUW$}Z>MQagXLJN1VURDl4~5{j=hV?i##$BlP+r4(=s)am1ojUY-Fpb1zowKQAC#-xf5Tk7@l&*h1nL=ZZO;YEfPa7!09La7gG$69X8U_X>h-Z6Vl9T4Wze>G zIIG?Gs>5G5n&#W>wmip7A;8G@FM&(jrospeat73q;-CR9h?98V6PEgxL#_BDsOPR% zGyrw&g}_o-h4_g7lt93;NjElNR-4(U48+QwQA@TQY~t!y90wnV6=d~uvr&fXVuL+W zJjHCLNHp-G1gCnR5MGh*giUTYKW^cOcRZ|*4V)a`Pw5fwDS?i`>{*N=gJ-7VF#kI| zpBZ{w)cr%v-^eO+{AN(G0*&`d#`|Xl0`5{u{SIjv! z!`g~(&zM7tVPVVa&T)Rn!$0|m{*fLQY1Gg^lT++qVaIW`&y8Slkz(2@(nhI!ebPne z{WF8xnHti^t5~#)E)Okr|0JNmsfK7B$Bx@nrH~ZeMJM>x@Zal_^>Qw~a?As7*ey;_ zytm!wynlgH8srp4uQ!J_aB5%!znS18CU@-PCO{nTy);3BVufeAXh4E@Y6J1k4XT8f za1g_5TE_(CF~N0Nh0`&Q_llJGTb)#-hCI`Zu>hFy5;nR+ciXk!NG^2Sr^aXL7m3i-sKDqH_k3ExEY>li)&|h@?f1gWDh6nyzjF{8U7^S z%TW)I4>V`rCF`CWfMgp1WIM&gQ)Dxx-ny{w8!JhyeC%hN>aGm9M1oJv%o1cc9}PLP zD93S%K%tu)Hqdz1bXxOz(Vy>!n9`ToZOFt7Az2KlbdAzIpm-OqCMe()0iBwoqFtF# z0$RQBj^i>cJa3n5be@`dXEWjY<;9xwYIZDEOUn!Lus*?LEQkHZkBJ$}kww=?06D2) zH)d)bkNV%3c~X%pULnvEzmIK?Sr?9&osVZ;UXkp&rkM+ zA@E^j!?Suz)FdEg2sbz$7bz1y8U3PbxIzyaRl{4ABz-!81|fssLovlb0wrIk-dW-k z1KsL5bT?Y=jgISHu72s5%jmaOr#n$YH}X5e_J?WE35|-v=c#ovbugtYU?# zPR?8tJ5w8M_0M_|L=HCnte=-oN@~{a4=0llwqrwMCLxS60^pgfp_Z%0~wRlbZQuv>iWID8N6^akcImtw%5&7jm}9y>j+ zNl}s{MX^S>ozo@1E#Jz+v%?N43%4`6qg1al58Y82klWx9$Rf#FIb|OAK;q~No7NOq zh7}W&H}=wosQKZiKdk$@?}V{LGR3n{SP-;7Ad#OhKIq%QPk5~nlq(K1huu-NB8xug zo1)Y%elUN^%mmvBx8WQc4b94rrQSZf{l^}9Q}T#uSWi;yI3!uZ<1cm8{uf7SQP>*4Q3{AkPTHd@9_KO_r3@vm0# zDf;luf8yg5#M|$VaNY;%*Zx6oGupp%J)4x-vA+R4{|5fX5sCpSmnurFV?YB>Lnq{0 zu?UFHp#N*g1({g8`8_htfUJ8qrud*!T-HrPK}pUilSY>uw>}-tONyu6?Yk)m z?@CeR2bG3I0%0LO7yqJ-eJ6{j{)zSD@hsWofSp}yKc@V{WC|4Rc!O)EK=Hirybuc% z@6afD5A5-?HHUn21=&Ga$_Mm1KoL@qX>tXn(0hR#hLdzXES=+tz{%3FCv#s}z<+s} z0_4fw@UI^XuD&?pfUIA-yz)&KjCr8mKsDAgl^Zc0)e7G`QEs)?{po*u!A# zN!$K3Z&kb_-$|ZKdzKw1cg?nEr6_@=Z9Z1U>lPl|&Wt$VpDBXLSCefiV7_`{YL3+Z zLw`y1d)^X9ikyfAP13z2j++&f?leT7_8Wrg zoNiDaozjoh(vK7u_Nn#D2e4mwv7&nT+nP9j`Tqi~wFnF~t30qtWUb|{8Mu0~Af=m`2o9FRC|(CH>`OWfIC5U$UZT<$*C|Y$6E&1Fnoo7qNvgSRktlSIlsyzh%p(9FyS@-k-<4M=E`)=d`IR#R)HU$&QGl z!g^FgVRd4BHpfxxh*xtQfg@3r=iR{TlATi>mllh3>q5e^BZf@Sor6IMjUOgc90H`gT z3v`)xCc!gRbcXM}XLo>`g^dLlCWCkjhM({)Kfg0KGh7dqH;(>G5vgPc6+8BFfWqCt zzMi0%Pbi=UQ{S7L7Nny%j+a-2wnA=alzteOI;|trc|$EzDv}QiG&NvlQxv5@Lx}z9 zx#AoU_=A&pN4&h&D;K%Ma3p+rbHV6Kbd_2Z=pl^`8X?H{T@_I2hw^~gqr~lXZbL2+ zVVcGn8R9m@Af!G<8_QU)s*`~e8!O5dSHF~atuYP5unF_q?{)zCG7RDLxRJQreaJb6 zm%)vfcX1B+p&H0eD4fM`OcP(zC@Y1&!ZZ=^XP*JRkrYK_SeuOXWV4+FyniA9naKOF zeWq|;#HXf9lij+<<`$cv3&Kr78%T~X252K)(l`YiTw-t=l%lv6yotFMoCm!i*f*Ll z=yYoV8-vl>V;)hwc==AwYGoX;DF`Mq`D0dr?S!(qpOt$x?TvG1e;cA-ntrk`KaG@* zSHfh+7Ng0?VjQLzX!p{Buku7gI;ILb~La-C)ZA(nN3){|u6o&han8JqcbT z^E;W{3=W`aQS^d%bfd6UIbrlo23cbWd)iMnXfnLlJhHquPOG!95v+rM-Mxk`^v|M` zXJWC)S#hmLVenDdZ*{{Of8D(wc4ZNQ9lpo?;yjzei@%k~Ipg|;TilX|;!nAmB!+iU z+&6Et^JQ^1UEm(WyCRPeoSOMiJmJsU@ZV>0Y?Wy+8Pggs%}>KPp+EcGKP9{wU-Wga z)ZgKHeO@XZ?U_oS=adIPy`{~@Pn?ml^rKaiWNKH*gf3x!BvKH!z+cU69f*QP3S zIVx%iybHUdeZG|os)U_T{8yojbjJXmEnp0gn>>=Of!im?V29+(b-&;8_f>zL!R-~6 zhG61ZQ|*@~N>#KAZG?~~=f?1iCHOJCf2xj;)okDQN4@iMeNptw=PYmFp@ zzNDCY6zNBGsHppSj@ zt~}B08L6Tw;jSPggKJ|(98mnHLOv|Rl)|tqHnc^dK{=m1Q7JHkq39Y?+;-|D*Rnu? zZMoni>E~`@(9yvSHwa^;&p|4NE_wCBYji|drdNsr1pqWzO3)($tG3!t$LW3C6}+r0 zWwP6t$`DkBCiREjmfxA``fdub(q1-V0CpNFTGwyE&TxtBm(jC0v7B5co+>+@n=%RJ7gGN1)8&hw(U z-1j`^OZQS?tZOenc>!Ms^IGf&%b*kayT&5ZrJ!5iwle{CAfvs*i2COnrzcPMv` zCahPePC^|OXojgiovA5;6$LkE8eMD_lidei-oRKrOFi~oHT^je`+{FU z@?u}TljEQIaV*ImFG6F-DS{(L5t?#}DWyn}u4#3rCrY*^fL8P&-)h(?U3OazwU&{N z>qx3-Shg|%MC9OSrnsAER`XC&0R#X2++0By=w3zuFVAjnFHpdw(z_i8iKa%9D~NOi zwODQqy)<*l+HYIa8cqaEW=*0MQDq)XZwTAIbk+ms*B8iA5^KkuM!?BrkVi_Vm=ua6 zQEE`DoQ;lYEH}k7k*_)OD$+Dz)DDO=$9o0*}S)9 z!g#Ra+>bYuc|4309h&yHe_1!vlsM3CS06J}faUI32{oDyY?oa0{=!|;F26k^O>{{- z3=zep{jyfiGRIX@$Ro-U8sp8PP0q1{#*rIy_n7*!b_)$NPWERYP+RL>N}m>K1ACO8 z`<2NyNQy$QLy1o`_jCD8Nf+rLClshFxsRL0?NK(vjnR{K%Eq*BeP?PYp38_a}15)GS%;+S0tg(aC(PRI0 z+*An2Zmp0RX`dc`3An3qo)?5A@*zBxqJ^oSI4ngK4rD7Y5Hv#UVRf z?7q+%Cexdm)LDQ06N$HDQv(|w1Lro2VxXBM74=uoNZVA)UCTowd5yx}`F#u8RGEwS zaMy<&;jZGvadJ7Op(Av?YzH*PkJi!&4}@mXS4fV3JD3QYtV0G9Z81)a>;b$m7{eBa z>~L~o?nZyp@Cut+DLdX*n+e8X)V+M+O&6%G!&Cqg*A}>IA_S?TK2oXF#6bZ>2h=BI z0Ob>QN?#E6P6^fegy3Wg{FsbcSGk@m)EkoZNtb_1HrcTu0e(0GLy}A}i4@sJsdJUZ z;X_XS{8Ty#Vj`s>Re`O+2bBe&wlSK6sSx2N8pSu3<8MQIvVz=HKIVOA6_1&qE6vh4pk3jk=LhfpN-S-#a{; zP0<8q#mKC!S=H` z5kL4ZJvv_9eRL+dF{6(8I>grQOtdc3{dJmw-)!E5Lwy22+MQNlNzU3ZZEx# zoaJkF%(yZ09+RTL3l)ldaa3Tt>uGXSqUn<+N(KoQ58y4?J}HWRaY0B_;1#cIWi^4S z0KO*C^-^HA&p~Jb7$eWEgbScuEL9Mw; z+ATSteCVP%Px1sE+&&WLiOH&VNiGzHrqao7=Q-NN`W8!;fHbn9wt9j1HV??R+@pzU z#lrJpbcvbzhM};{ye|2Je>?WpgV!rUoA~!bv$;DZ-L%aXSMvtOGPsz3-d0z3pR8(< z>MhdofBxP@vUxfr=M0yHofHGa_B$x`a@U?&{WH5|+d`9FdX#OdcInE1(@QRi|B6hJ z`Qk2l1t?+fn0buE2X(<0zEzmx7!|ml7hyYJ=JAJO*{A!&*Y5H2wmLBXWgI6NW4}&~ z?5yNVQ_JzS;(ak=AF%cb8N@ET?F!lzG)M-VI_VtoY5xD&hsR$C^9Y``=KHVxliqUp zt$yWa!n9{*wRu(?L~M zCj}uL{Ihe*WW%x?PNOgi`kXhzxwCUyR8@)t+_Q6$l_H1JD!gWk>d$d8$qMFZ9>yKT z`X@K7>dyVTSF@}UV;@{MJxUGJIloPT9YUjZ_P6Mhu!3BZqQo2|e*g?nIeuQ}}v4YODe{~D^+?1otZc7l%u#W9=#_AJo`1^ap zzbZf3@wEOrg~;LC;Sj!q5AfNZ-M-!sx$nOxG=m^emnfG?2HO(qu_~q z!iVC6;ZdBwj16TSv^<_K>%r#nkXAWh1siI=MD@q7d*upJ%{}hTf~^j0QSt=0XH*4l za>eE`E6@A0CdqPKJZmKDUT1}iY2RL!srJ&ZL=#ckwO{v~)EKQ>kCk?7J$iz#Nl|{F zMzWTdsjQI;A@k6o~O-Pbr# zE|f4}@WxsT)nhzZ4zXvgR*#3Y%E8IE6DU*9_Bq-4n5pWHpgQH1u;{d@w$54Od?2`n zzhmZR;XcVOK~&&smvh`^<)F)y+8;L1u*#fUIEY~P36*IC-)`L#GpPk^@i@s@J)=UQ z=@@138g!~y6dSTnwBIRCgqcH4yCl}R3W8q9vy8bu%C^5b;&f%4=j!R%+#SAarYHLRHImUfCi;9F$wY~I#XDTOm?99>yyIkh<&cSD%GPm| z&wsS5`!zjER&V*?zmjS@-Xvczn)lNbbAlqDQ0nH;wVv8JJ^V-FtFQFYooIrZEY57kxe*M|^`@(T&R6$=74?+U~UT>G0RL(ec(!f4Otz?&}(J5&kp$*Va#Bn`|t?y$pd(nT=egN(yB zoNmgsj-75_ja9_$l4IF-jq<2VhI5LdR&vs6{B~By+C+dg&NDaX#r|N49z6@*|FnZ> z*`ddd3#r6)+Z!R*#lfDNaNDM?-llk z08@*O)uqX$pIc4fWeW${#P(t?_#I`O_rMVUv6g0=m!I-bnfv2h^t)KeeseBfa!Phz*OGW3WT z*q1npiKR#sw$0aohI2tkDU`i6@c|u3f0`Dw9>Th5BGxnY)Q{`g7cHLXXAfo7J6J7J z(~gEzPB-l(dQLXiW;&17@U~0(=$@Y;1%ACE4G2?@gKM1N*T_#-jx5OLUMDT8qJLZa zy`5C=k4Ao&`h#rMwO{ququY09L%vg_n%E`yV>Bxjw5$wD!~&+ zCF8L>`j&9m0VBa$M-0^=`xmS|@$yc$fdzTcb7J*C+>@XF*crCy7<k>ub;MejWA3BVH5_ zO6KF`B|b60X|~R91y6(Jl$~@^*oD9xA4F6J7wkp1qz4Wz1VWHcr*hQK#h^bj01i1_ zf#6Fw-8chF@Z*woazE)sBODW_nw%E-YeBi<$*vKcNIlc%NJaVV*j= zS&-+A*Xl&su!%;}%3SX$vQ_NiWEmVLdzaTQelpELZ+Hs6{QZX{g5CbvjvW@Dhceh5 z#Zb&9ifo|N2Yrh{n5I#-(g)PEU}IFxN${H@5ExIEC6F+FVrpJ3`5ni1GKoYx4yl+) zBw{S1m#(1CL4vu8j${gjntFOWvw>83rBLU?wM$06nZ9HyF=7cc9xdd}e`p3O zgj$z$m-JUt6uR5~U_&&pn*7Ho7(1{X8d!NARI9V!;B@`MqUwG}LzdgIM+@C+21`XO z#Y9nLBc=XAa!Y(+X0{Si%Q_C^n*m8d|JNI605MjWcwyj!vZ@7oO#J%Og}X_y9b@9O z5hiLW1_S{sNdn@vpZ;tf#4)(OJ$ zsR2=(!{EB!U83sYSl31B26rTAUE38bLd9wH6&-GYIxhZ%=&e$*hxfRoQ=Qql}0B2Vh7Xhh-R4C`a%3Utl=i5hb0@`g*xI!7dZrC2<$eE;Pev`Ca zvQ}_Y4#5-*v*2%$A`W}k1ncP-}`y#kZO znTauC|I!@b*x}6e%?`1t?2O&Ub(4*77ELjc6j@KH5BaVIC89%qDT+1DX^uG4)>w1Q zeZJ<;TYw1j58LnxJ477W^W#gVd*~%V~*OYuK#kIDK6~wcZF?@VF)FjVm4DGno<|K#t{4;<7~O!UGmfJ z7bWE~Eo_ow$a2@IHX0TK*7}%y>WHiWOZmBdFMPg!Ta)gYxs?=7Cs5gJ5W%XVn1dAA zPpLcjSRh#DQ3G|#Ni&a!?vUj856fyfUGfZ(7HD3P5~EzUlW9{O_0!#(=yQjz6b?gE z&2i}b#p}T2=hUX^;Ez3@qS)XS8`>`a|Ji#JxTexHUECv{kbIb(KynIHAc7z&0-;4L ztgc?C*XeC~rn~>Q?{u5Kcbr~s^-QXIs+U3pcUcrrQCU=oETSxmD1@cJwNy}46juas zDXI_@DOCKwZxX7KKs1LWOmw>Hr#U$rc)sVn-}}AG^K>dun5rPgIs}GDyU;k z3qXbDMraSL^v(pr!Q-5Pd6nWqv37PVERga0u9^CUagKW#2iL5)AO>$%7nZI_^xYGo zDc}qe9T)2l)k*&wQ+0Piy7gorg^p@!6KU%V^xGaFI z@TMS@QWt2%w@V+(m!i^l%4`&piE_H(zjo=qx@Zy*Yg8ezg3nc3rGJfqG28g}>? z=uPpltSWF>d|0UyFvQ3<2lbJ4F+GYbWgDc09W=eMd;GRQg54b*b2ckTOg^;c+}jI{ zPK)5tv2;@D!m-%%X5Q3MiUEPTdf>1P>X(#FYnL`CihVav%Y!HBVl8)uVPniWQP3ool@NFP1fJR*Jzg{7;83 zL2dOR$tuO#s0{yMw`YOLz8#_l$#UL?aOk`0RMdgsdOa^qyo=tZyfrcWSmPN7yB@HO z7qc1%*DVdK^rZ4$_1OnAJea-d;6DT|w|;IiRJs0r_A$XRE38i1e(i7hfyUYA4|^Mv z$xauZed^6-pK^);N`yU>`e{thv|RrRQ32%m2BCNO7@ZIZG@^db0tY0gMbP^BmDK3A z*nIlk0%~Ah5%unI;b!S|NjYc)4M-}vg#i_!n~IA-4~P4*gX^WxAnJG3<0u{P?)X5= zcr4tQbBoD{J7SX+YN)4G?u#w!RYqrVT-OjnRlaFd=<^zuUY}2UzX}Pw5cEWPL0yh*rbN7Lb|X+5x2SV zq@TZLf&LV{8y2?c(`E=b5N!3!7Uwv`II+oxncANhZacs@OhecSA?n`e$pI3zeGFYrl+-ZPk^!kzCpkK1lIVsKw@ z59_ni7-{H9ML%%=dVtZ&6hHsz0g}Vc%DAu}1f{zs*|B{Tvxg#ulzN1@6mng3qHl#H ziPP&-As>_^dOZ{jkOWx=e-l$0T}N;q`a{onnS#Oj&jM@08pOkHiM|_1jmLe(Cgv|X zx@TUus+QMFKX6E?uH~4H8(^lU#qy>O_`ccqvnO9NLZ9#YMZ2brnslGfN&-);k%TXb)qiq{6-(C^Fr>NWkoI{GA}S8+eO z)UC0(U%jc{pF3M{oIr95xh|H@5S;^`RJJVsPEpck_S!;Ll$(;^y(! zzwxi6;kB8uZkWOUEXBZL>my3t&g}6To>?khF5E&)x6>8p=nQTFbO$6dx95BzK3@*P z@*N(z;=;h?!W?*Mwg&70db}<^ZgPSO;|%nhV~m74?gfIItI0644B6 zuqw1De3zJM%y;wdC8PL*v5%;coU)OZ}@>-gQGw%8kz;_#k18G)-Hu3O=gMFvSD zH(!+j)mQr#eG1;_O5aPO)zOt)2R_iz|4e%x@#wobZqI8nXC3;=r=gZ&(XUB7(2C~u z3As*RL)EJ6RQ9?fUtm#iu{4)#;2|f=-FYZjb|g4SjJ(ZQPoA&3K)RrdKo^Ev9k3Ta zvy^zH(~xfu>T%+69L~+}wBuUF;~7T7qFTSYmTY(7#1+(>nyjYwP)s33w3K=^I1j}j z5YnONYDaa@4$?;#C{m+W#12Lc%g_1L^R{`foPtMgDsp)Hf>%vB7J2C<&^Fshk< zC>oC$NXDE60ecbPEC{f(d*46umlLl#Ug|*Ga8vX)Wj`dRw9>AKeC}Eqtk`I+_;ToZ z-{r!!94zFx+^TKs3kPx zMks2mwcQnU(2!-L+G{c$Ss{BFp}*Cjt~f@B_2L1zfxfN=t37r6UI%28%niBqZ!8Uu z^&2_($g!vAch56UJ@3@S)|2AbW_1lr$|e(4HN{j>CT1C-d$M@2{NF|_SWFWIWIg*-?CRA(U+S;!H#rDp!L`(NfKkn0)%cto(G~!+(QHZ z;Vu}$%^Hi72!weku$%z_Izx*qSxaXz-frSBN_BI@HyXC1A}uwYgjQa zaY}N)XJEau{iZ%AC z(X@1>6zH&q1l{tgh3GWx3r|62*G!HsIFmtk*PL*q6v?MGX#sf8Ealeln_-%G;aNeY zO4Grw68DI(J|3t?@D`UDq=?nu#K3BkM<<3H2QH+1=)THQUKJgRwjEg$2P(a_Ge^z~ zzWq$*E3>4{J76~B?4}stTi!vbi=eq1gReQ_LvwKg(Cp>r0PAgeRQfz5(boKRn{r4H z7Xwj|dF8+s1$(gmp@VK$)blpIxmvNC{z#eW?ZifKd!O+O9+O>%Sj``<%Ub%od4E1M z&FD3CEt-FwY@I^DIW_TCizsF{Me@K~)sq|dNz%hn;~tRE4R0(qq z(Vz`7e^>Bm7|2Co} za#`0La@B=>>?dY0?xC1_6zRr{evYt58^kB(uU|E~w;_ zUnHg6V{{8`IIBlg9`$L|DbhoysSZWs{(JMVLjtzIB`)P0aj*4%G`}tO*8HRTD{xP` zqMlR7PvSI%Y?DIK1zpE4^GV=9dzB1*XCwy)b^?AU)D7a<6`~F(bB5wV#nu4)8O^eO zP+P*S1M&)GGxw4R9U{EGgI@t!VW`T4vIo{)B)eaPm-*K(J8tE@Wo0(0ym4d<52jerO(h{J?7MhHgOY~8-ZOjE(YwE zTeBdxJx|tW*$xcWx3=qooAkV|35~#br+n(4$T}CE+91zmGSu5BW(x)w)S2F$O6W6< zN)q3SI7@GeDvLTvAf2HqgY19B%v%wSkl-kj=j)G5*s$9^;D#y1u2E!hF}s;aAf7Yl zRo7`7aogo7?v0XGa6g_vUqh#&bLNoO1CI&AgK>y#h6Rfub;PNxU@`gk{wLl1j8@|r zb8y zBVZEVnYNB(yD(sC%mA~OVu1Fb0E$8(9JeYwiPJ9a=RPK=@W}$e(Di{N{*zm-IM!D7fme&%-&GaR6eRQfR{p zF_YDr)U}rM6R%WLuobS18$Qc~<-jGqj++>i6jrLLCC%=)1ts93^b0G5ZLw>lrJR$z zMqUlC#v`4>y4-ivdurnj_Moxx#&(`&9W*|}|MDk$mNc##ZX#BeG$`|kn#0(H*%Y-E zVv+gb=NHBcwR90(3PH&U_GN5`jGY)@dHI(2YU%%3wKgp>ga*b%5*dt;?Tu+?u*|bX zR!1j5ASE?w!UWh1V`V!uoHEl`1x)CogMSV(x*8X$QvS^8hW-U!ghB~C&&X}b}(J>nLh5dRk=RA zS&pF4H*0?(AGw#*IJh^94G^q5u=*ob2X#CLtWPUrRPRmx`IO%^TCRde`ywtpp;%Ec zZwSs74{}OGdSXt@3j@jZ710%=?mgSs=Xl1!LFaLE zp*k~@^MFnUO|J*MhcoLr&_Ep76oLwXLju#inbLB3Wgu`MDJ$f=gZ9iC5S>yMsMuqX zHL6}RBUw4j+9a_;71a^E`{$OW3$ICX&`JfyQ5yS|v@}*#R$6mW8Y z;C?4nqr_E_VJ~L)a1Yi@)$-2xHoI%3sp1-CPSnQe^r`GIW9wkri7i`y4D0hgQC$9t z%xHs7eEU1s$Oad-K|sJ~lHb`uF(4SZjZ)+2mUv%Yv^_c>Dr;(ZniP@{uB8tJot_4D zTUSLo(JdM!BXJR=gUJ5TkU?pU$3D+Nsq@C$I1u9(-I8A>Y2?7g}1pkRqr_+0;5L8Mof&l&e=QIVIu7Nl3SK*9yZg=tG62Jf=o`K zP+#cTF5Su1vtD6iN?c6E%)bZMjOhHE_PY&Jjq}N!|9kbXWc657L@qq|TC+K33&j9K z+eS?8RIB<3Rvq2=&hpbc;R7iS#N+NUteLgqB_lwuXBuqO5tKgNw%aGvP8gNswl9v;%3p?kP$Gt=L z#?<&j^F0)#ouyNwOXU}!siRbGI5q@jv|2jOS$cnVfk+P}hhpdRn^_0ZlO|D9ta0M$ z3ZI@y4o@MM%xw8diaAD+CQ9AOL4DO`c{vbP!iq6GF0^Zgermr00gEAe*P?n=D%3us zI0TYibkW#%V_0N%(I{|)e3zM=%%H}_aWNtT>cdFdT}SNMDV~W>iuy}81@g4`HMgL+vm@m)rR!}8sjZ@-7c*I|M}~tH;rie z#oe=8$sreB7Ppzf`~<}urAQ;C-ss-QMe#T-eM_Dkw#>^QActZ6Tkz3@nrM8J4$rRm|K5*Bqm3 zcvr|CUOrv*mZtPAEK;cP=v)k&kIA*@M(%-y=Vce=NIgDIjI&a<91UO__OODs$;qrx zPW|}s2LiqkOmAJ>^dVW}!eD}?CX*OTCdF)`NE)Ti;h?_TcCT*LQThqNxPOo6xujX9 z^|%vR9C7fiYm49e+cxD6ddGrs54OqkwP};B3AE{^4mgP&IF?L59Oi8VNBi81?~pVX z1_zLTnt)>$#pF^Xi&7hkW{2em-P>Z1_$BdBFXe)$PJV`aBM>DoG1Z*weSvHvrQ|bg+n;fUC4$oTuvveJOXx4r)MtU>7Pm@;geV+*uXPuDd@Id=iS@fmj zf(})out#xUam2ktb$L;da0i28x7nec*j%RkjFSi)4pm=j1v}OUl@4`$-Hh6)dwsIz zCNIiT-kr5yzFmwRzPhk=zF4z)c~K9IbhY~+PjE+-!o!un;qI-RQ_Ah)^`W|Xwe#}1 zCk2HEtA2~N&fj{q#tW(jo;Ajy3W_2JhbO-n; z^jGSA!>6;_7X4L)!bL?Q~!V9(sX147X#XzI}RZ6{u z^!RqltEjuoBN=`^lvYvKA})t^%B>%1jcxtafyGC@HMpec%bTh7U+#LVEfzVeQ9r7W z-Xp0AyTRY)zA6O$tv(ug*(&FBNi(>sLb~`v(5co1RIisqO6gA{YW#CmjXbSKTkKtO zIb@JzL827sl0dn+4_=98KAzDXu|4=oNGog=MV}Fu(;ME@&c5n{aOMl0a6M$LI)O>v&Fx>5V z1yYf%u!_VAEL`wm`3$TCZ<70-n0+h_dQ5FruJYExQnwM`Hu3fx#2tbk(CILQBUBAu};BuWI2y(E3@L zG%-$!<)UjoS3qbin_fkhViBN0q62EUu&%3CpB7CE0EVMZMdO!Kc#k~#qw?v#$W_$N z@K&!nxxM9GEaApt#>pVoWcfv7;uurvzx+P*ptdJGesLHCYd7ZC^e2d zRM!BVNqN+vpe|`1l>Qb9yC%9G97n@8P>eZlPTe_g!OlU;nsAq$ajY;Zrp(TNb2#u> zU|VdjJ1)@k-^>QV?6%lONvGoUw6f6i@|qbFH`~}e*oGJ1siSLgb5;`$wNT*wN2SqT z#Gd?RHOY111$wQS1t_JM5{eWdxkn4X+z*)y;-DMklyY;B2EQP`$Kn4K7X$PLhoW(+ zIQR|t5*Kp}c;=A^IwxF%9MQW}?b2*9l#Mscc^H+d*=!B!LiLi z*fv?1-`T!ztS~d_#)oZ_1xCzNeeGLk$yyi2Oo16@wo=Sy3PwSV_5Xv?MyQ}euOkP7 zH$#FgA(g^H(EKxOVa}u#`WYwAsIWwbeM8RTYF6l&ykl?bf4%DQ-~pmDwgIG1HE2g- zdr`ZzQ-O?yeLgx*!`_L*hK;_(i0NQ4x^|w*4jliqVqa9S(R9qt{veg)vvW^fH>UeXLMW*^NzTsDY#~F6=Wf|R&SS{N=FxSK$EpHl_#we_ zw>IS<1lEqY;~5_*OT3qH4ypD;Y!BEOt|^a73QHwG7377T3fTErxv&|C5FI&3v9v)W zc(d|pC$>y3kBB{W>Q%3<$&@C0$w_%DE~YC2Npx36rtq%7H_h|l9=a4Nu7JHgOWEhM zR@N`S<+F6DBj@%}!)OmmoSB+|-;{(UYa(2$H3#SvIGQwom+DI`e z6j@KHKMG7_YB{M;GI+#qwO6`gC$C#I?6%i?N5BNGl#DR95!|C*3TN;5+27`TY}u2> zCdK2rRuPsvrl|P5gAJo>ZOVoj`koT3IbFIG);v-$Zl5h45)^tYU788WM9dmqgIa_v z)zYO%c3KaD6RA;HURo3kr;hyRSRmO55)1ES%&!i`P1eA-?@TtDj)aw=O|TXiqr%jM zlQrATppZ;4>nO6CQlqc5%0truWt;6%B>SJp_?pQG*)r}Xr%%|X^_lOfl)HM45fN|Q zELcYhU3gIgmFOnZ!y$^Pq(}v&#vZdcA#w&jjM8_SAsoPYKH%G`yv<#vFExsbS?`Uq z44{9rF|1?0k^LB1=O55pg_>;m>#zz|FG@`tQvxOKknkN8pt{yBRUL#>V1FwIlYs@2 zt>drE{bIar0m_)>;DtNCxc_*!tOI2XJ28pZ@z!@Ejo|s!xdnSjDLYTZg$)fzR+~Vr zo?>(qISgIX3T$1#p0(5{9i2KC)j$9nILTl;{nMCz9-7V5I{3A-^XLlqULbDYIrF4% zqHi{wbyd>FMJ5nzf4eYkA21DJjsR&dp)n$87YGn{LzRM#?hi@y?SYnwL(nGqkuohH zpROhLPfyN4oc-?lp(du#vT1;g-{QjFtd$mnYliD~Z67W#n+j!uXPmZvP>W=J`>vv{9Ns8FbUM_5Tj+t8Zp;R8e#XUrW~rkI(GnoR`&j+?!tpl1}TGeH?I| z1W9vFi!-4ww^3E@)9jwYOAbij#DzBK8=%(W7OCgO32`r$-#Y4@u>U16wJNYZF369* zNb-L|KmC#sJ?-DTn@=iU8}yttLr){c)Kf$U=%K6pTKH+)L6Q{KBD%cjg6LL6CNu1o z%uV9lia5zxt{8T^6>&(k^g3YX?_oB3Bl8oBG;*Fd$l;oP{;GM{|<_8cvUjo1q(hr?(DRilx?YL9n~u*5QK zVvkwhfW+=YX_RuwAUvxJ=5w@U6UzN_ z$p#)qYT7}wN>jUFTR4yn3AF-!8^H3}mqZ4B0{xO5)F_3=BuzZcanclmx)bfveX}%` z(#9E^rT6F8g%r%p0zo$&goAWpfn;6e{ez!4S6J3$ITOJmWSqubz$`!RFG&R@l#iYPjobx#ZcQba;&elZb-(6?oG}Vg!WQK z5BHhpv(seQjnxfY*tl#n^Kg?WW-Uclji{1-$nAn00vXanAj1VOhtD`~Z0BXU z^#7}vZUn@4i+=rYWZM*S$m~fLQw$VF=Tqu6P$iT&B~JJ#I!hTZERD|aZ{p;Hub6^b zu35?(KE+X3Cxnt6C`H*JYT|60*@k?!Gy1uBB-f{jvjs}+kI{q96P0`UbAGY;wexXW zy`0rs=DJ?%e@uGw&EFg0a`2a)GIIa5Ss|}B>%Mm{hQ7!xe?hrSL2lvrmUA4HzqiOLF_vx1) zGz!(h`SkV~$ZoTnv)lb5T^zMec!b*|ZRTwO6^x6ZRckbk_P#_fe#ORf)c9VlY}`A~ zh`x7fV(UrqSe%M3JV%`{v#iw=14PaTFxhD+4?OQtC{E(-gF0=^d5?6(dZwGkLmFKe zmW3IF*o%bMK&1jmwPb@t(+}tDQr-9JosR7(8XXr)#Wf$hcYv7dbGM=+dmCD2_#l6sI3RATUzOv=|8rV@MM#ga^c0e6~WJ}xoN<(05tTp zro=lx_`FwAjAkWI2TiajgtXdgK!L1O2}T)Hqnc?j8#CHsaiQE6 zaL#X7o*Lb**fz7usXp>JM*oWm!->Oh-E54pj-pyVu2)zpn!F~a1VlegswB2kOeRG( zQR*adAC21Jm_SV-$HAS^(Q%POifap23DW{nc(uSJ*&@3_hCE9>N*z&7W7Al(Mw3l< zwg03kKji*NYV=e@4@AEwDK1=s1+CF0Bc4MsP?48Ose5EYiaffD+>hJ`HB@!xQ&AcGK%t$T z%dJxx&cgHic&j)$!dz9oGl4Wd0F6vTcH_kA4Xgv`A98(vI>~5HR68sGM4q|1Yz+q4 zYGzN?Q_LEQtfbUf!@JY7QmR?2(1|8GeEaz`fBD#de(Hp7SReZ_{EtcBXpMQxQZehM z*<=4)n5x-*r$nEQOAF|sR|&JBCUIDP0J!^e{jt1pn-t0$Z!fqFj9jbv26F#a-#REk zTRungge;%aBj}H+a8LBD@u;5GBdCr<@x*RWh=eD2mFLz=)WA2vvM`QVMbqz?JC z833#X&vloPpS}H+Dz4GWxCObzlSCKxT6UVjJ)L4UP$U`E>Ob_oJUh|1Xqx6D&dvF| zU?$N)cV#!d%A*7Rp8)s76g9R3*7PdpZ?gkz(9<8UvUF8kHl1j#HWZSHNt|AvLg40Z z=68^|=`G3x2$b%oAAvM%g?vzwuev*L8E4zH8jt{UN_n$`T$@%( zuO;{AubARQ0NDq*{klV;CyMAoq+5K2nC%u=r0&H8}RUgtDSd< z6&@x{iITlyevGfk5{j24q^O}U%tV~hmAgH7_H>g^E z)k9MYrQL<%Ay0iR6s8Cd`0k0& z&A@5tfrrxu_k}}Z)5yQ@4$e7f-GHNn%J#M?AEB)Q{1Jvw~Hn&cw(z%aJDaD2^Ch@ zn7rj#>rKnjlF>^U7mk-$DIn?KSJ6nuk|bUqUPm`5`exppvzNOs^f*LNvQ)LC#$!-g z>z@Z@j>q{&1kSV}rsKmFGmdICHvJvO+P03FBHxVFj4_@jcrtLAY?=an;HIlbEyX}{ zUM}WB`bf136}t`xK9xXaXkT=t>^&{0USKsTl25*eO;K12h;#|bzFWgvWx0ZWMIG-` z*;NsY6;uT}aU@tj{E;_T|2fuUbId;L4}Yv&tji|ilVl72U2H^3E&pQ|YgN+D|%l``iTyYax- z%+xuI;ldpGtYAF3<4fqsXZyvy*qMBnb#2cFZQ5q(+On}J zuIrOo^Uzy4IhZ=PG@AFcXbGqRdc$eMI=|;B`x42d)o)G>zp8kabIfQ|xUg}sG9f;gTDtH79Zz!gxxa*OUNPqW$~bYn z;;1}v7u^<%RoEr+vtEw`H=!8mI7t#8qLU&2x83KuWTon&2;^_=Bg1*|VA>3p&wWtS zjThhBo}<>m^8a0#nq?`lJQ^%6ysEJxqkKbA9CdpE78N0L0Z^gF#q63{9)(Jmnhekv zISY)mh2o>Yw;vY+6$JS8HaDKc#ngBhb}r-GP++mF<}%Z-IN*B*D7l}zb?}qOorv-X zY&buIl?QpM?LJ5Zrk$Pu%W?asBl`ek3MR8{d$6Ow);fR!mJ9qXnfci`DJ~2xE1doD zL0vQA0WHgf=v<#9dm`)~_wfvn4Okh^ZSB5hiL6-ss>`9qQ@0eC7C(~b(<{i|hUIOs zSU=zAb0fT%U|9(=t(~3GDA^lx!l`}aW@BN)csur`H~Z<G0lOd>(oFq+92HhS81<=gZu>9~|AFQfI0PUvnZ=~r? z;%M-Mc4?P=r?-{JOfkxug18%{JbY>{OKGbbhsi{{LqZJ8x@*?KiA8DWfi zl}Dv346Xxm-5`~ij*F5n^_(8~TNidpnd6R*MUOLmQ?nPC^@y8Y?Yu9nY{{h4|L<|b zOk-+i;1{wLWXEem)KzC@OUfu_F9kspbtSBM%IOVch{jZ9JtvoxinW}9c@6w8wl;~e z#~jc6;FjWVkL1fk;j=n==0JZ_EH^j40xFRUSEZdJbc=3umud9#jL z8WlDZkP9!It(c${v6scI3*NcIlrnV$vflU&6;bo)Wmqx|94DQ0>w@jUcjr~eQ^-YU zk`kDV^hD=ss*bwl!*?C!je;2acwb+iRB8RKz)UFOO z)?%ud8X;r^I*j?iq6bl~VyaSF!$ThPiN&sMpf$eq69QMm4`Bz|hkoAQf7Qt+-ajW! z1Xa>PvA*}pz;2=y@8#mwEybD%avw&9m<29IE}4)|cDRW6eNni?xRQ~7@Re4Q?7~HA zd(74**%Sk;sMqsvth(+@K~e6 z>}uzIsXU$Z?|h?ydF$e)56K#K2F8WczEE;+f{sjz*+h{vO1(0?W)aG&=7X`=7oF&v z52gW&ol#f$)Qt40g~DgRy)|*jfU#u<4(EZv^3L5Cj#md5!BPDDrw2&R6wqNbeX9E? z2E52ZO1+JLeqp>Y4&OMmtXNQtnJ}FE3&rV*K|u=$nPW%_3e2-q__a`Yz&AnGLhFl+ zHwN|0hu1NnY<50fwXl)2Dz$V9&k>avmj5Pp&-!yqY`I~1Tf4ur@(Wz|t^3}wf4gsE zAZ?#Vw@P0WZD2tJ@-kd`9-iuu*_hRM*oX;M_jJU`tT5ra^YlM`^H0CFM~BCC(N3)B z@F4L-5~n;WCp^pl4s_t`m-j*+t3hcJ`*FY@#crHECy5El+sW zDIO1e7eL~vIjq_fktwU0G4U2sZ_Qpl!j3R{?C-(OOs;TmWGn@bUz7Ea72W2B;17i# z3Ncmn+zq#K!uzAD7IrEM!GvH*OI^rHy|cDGIzN0X)GHad&o3{U0GPcTQe)W%_7Guv z#rMODCRtX)x@^Ugwfa_2&g!A{{6|>#j6`Y^FruTMv@K&i`YPw|VP!TZ{p_!@!WqW3 zkZS$vTC&}RBm0NV7HNAZ1|+w%Bl4ozbatp_05oxy2j%bvNfmuCG)oyLERGuDV)nmK z{4nYvw03F=#VEUUo7+a#hgXNJXX?Ugu^$obNSktyoB~>Wr{KvJyo_ZPO!qkhJ?!uj zsBC%vr$)S7{#p4G((1x^=`+JiC&hp~!!=5Mh<+HA>ZuFp@;#ygiT)%b;G#IE>!^s^e4fu%?#Q_cY|MBGea{- z)-%gwCkcAOWkDY)F;0+8?+du^TQ++K2mt1U*U6u|;oTZ$q8u)QvyRqE%T+xzvcTYX z$UPVo!gg}27G^1TMwhFeyJ5d9{)8u-BB6DJIs#^};K1v;tYzX_I9?umSza742NzLW5F0NJ1urtss`mth@ zG@nikzsKAIiky!X1CbTb>2pI;&&z>^(t?ogIa^{gJlbL_d=5rvp*b*#(-w>8u2gk9 zmHv@!VBc&88PAZ~d=0BPh8q4LAkLCOc64v4>&72Rj@eAIfnt&=vJM&NY8Pla6-~T8 zx*Zn75Gn&+x%8<{a%*2WI5rxI7v94e2dx7oJB&}sjgCk5^xyu0q`R_S+&qFPx2Wdm5Kf@_}isrbJ^ zw!q0iu@MtyR%7IgGq5mnPwSW%qfc;>>gUT#ql`HDQDspMDR*I5t_X*_hcP3 z?rj#8E!12Uou)B{dwCHa$BVKc#rc573VlP0v?&V0yB<9X4CX@>Rd!I9^yZ9Q0Gxd= zvj;^ZXtMe_0*P^avO>}1$=g2uqa{O%%Q`Do*igE_9|e3=tNKVTcXjmk;4Ec+cteN= zYp^Gp4#jp+PMBGuYKj3y$pe)7LvEM! zzE_(P(}B<3Zuvm6kjBv6Dq((jm-Lx$UVwp>tx9rV(JVU(@t5n0X8wh_JtFJ?&`MWC zc7e9wb2lu)02!ySHYmZ!QuWbK#TSXD(6)9V2}1M%I5yjm;oKv<8s0|XMOb2; zAxqln!_(uZ9h!TVE}@qkrI%E~BzV@ZTsrKg1Ig7Bia7DGTP2WB#)Yqt4ZEG=Y3Vzm z`JVc-+hPki>wK?@a{cwcO*Y*FQJg~VN>04+nq&!7YWfd(zBG{IaBP_)&FVK(9F4WD z{c+R7FB`G-l$-ktqMJglm<`HliaAMa-YIrDWj~+*-0t>6K@28k*3Mr;r zIOoA-*$H&A8qD$GMOYZyE{&J1^}Hi1;FN_7#q=mHNmF=xBJiS@KK3#G$rCaXWBlBg z?P52jxGafatow~^uX^7A6kLX$o~AjW%2M6p^oX!3W<7(9a5_#BC*K|9fCMGfl4*^|QV!9L&<}7!(!t-$#R4Zhc7)r= zZH>4T(Iv&L@t7}m!!k~ku=^q7IGYtpsO6Q*@+`@YU#Y;J6-wh1@+;tcz*?)7Qxse* zeHeLYF0QISRF#XKyVZqs&rRW#a@rXUPJ)MmU}p3z1qGOP>E%TraO&*5z&N1;p2#@x zz~U0?(0TupIUffZjmgn3Z>S=>*!jsWoJeXgGd&d)1GcDyQXBM3A$hIQ@oPz10Cu+> zbU#Txfpm4QKTKfj!|w^O;2kS$MrI`2KOJkI#t*1b-YpBby!_NZ&UkRilGVp$V>VWJ zdXOLzXuGkcp^S5lgDNx{EMv&0xBF;kV@KLkFCC4v-L0~QkVLO%^HZaApd6Gg20@cah)-=mi&Yhy}a`BxUSEf`2jp-?8p zJ&*2@=^@t%>f8t2aUR=3p1Wm>^;j$pKOwIJ3xt@%IvNdbJftf`o$lx^1L_=VYaEHO zmm3h<&o~>FmP3rFDRav?MOM17yPIc*gH05ZMv+uX-2n06bZ!l_O!c~_D|&%cupH72 z*b=3s2ffSX=@Z--!(yy$A1DW$$@m|7<1JFl_uI^yXRMldwq&IXdwy$cf8TbsrOsv9v50DX_Yi4zrPGfvr(KibE$Sv)^jcq@I8JUFElpcI3|mhwhSN>}DSqM$2t8tJ6j?=P7cQ zQXe6Mqyv})v*~-0w|rV7ut&NrHZ^QDPwP?7S?6_3d6Fp$t?@@Cn3Yp9r8WLFGgeK> zgQ}Y}?k?{d|Dx!F9K%*yEZ*k<+~)}3Y7hjP(;!(f#D&Z{oKvqqYvgWiv3PAm$Y;{T z7+qKgAGzy3QnpEt(ktG2w`Fk|=?ThGp66eHnom5D4G`^w)>v?}zRvDlh6~%A17^V3O)=p5?4ZbP3Ek+&0$*U^bWly2T6-YeIEA;2@Tl#k{$@z%v0 z3UVrTSU_czZEz4O#=K2dUtgVNbG6H^sIAZ^#tRE1Jv53jXz39qpo<*GE>p&Qc%ydO zHp-(ebNUW;4(`hQe~A4@qxo3%hw4Y<)D&{x?7h26F_$TFky78{9P~>TXM5yv+ryfq zZLzxGOeS$w9=vONJWlaD6@x2rPM!+ymMk8Sq$`es2FEt{ZC>~r-jQvVJ`1#+f5Wz69QB@#e}|0#U^R@c zYsBw|OrNSS;%ifC$R(2T+DuOsX0J&;#q6Y@R9M}wSjQYwW+@+pp5mu*_KWMIp*}dG z8(Nt9ct=Px|4{T(uSfEI!3CU}uq-7OE8O)xLN<_$X<5od-%Mw#fsY^WHUf8?r?p78 zq^gWSIiddBMpE|L0HxUsC=C=-OOa|yo$KGJ=p_Y#N#YG&S>AW$xSp<<*$ebz+4Sz9 zl7P0@^#RL6Q^g6Pd%0aeo^YHTkndn_!#Zh82#7`mG)wDX9hC=^c-cUrcv{j7a5^I^ z=Jm+V0Ik_xZXMmr*Lt+hKOB5oQqSpB>{gxey*uxyBN9~O2hoe{<~WgKa+$}P?*Fp1 zHls6=u4^rc6`PW!Od(gGEfE*F?*RE6aQJVUbqr=#>`nS(ma@S!Auw5SOqs*`h;wuP z+F!r-{(m+ue(%qT)bqDjQlCw(`09JhsMFuGeNr@eq_%9@4+mR*3j1@c&!OCpe)n^G zSnXUGLRMJq5@d10bv%sJ=;&f@A6+?pi((Q!Nu!uliX>5LgW|6~^>dTl_dLIF095a^vp-U1dKZT~ zb@@BSdup?R7~@)J?y_zm#z-zs9)9fkeamEk%dVfTg!eHOePLR;qKiMoU8}0$6#}`_ zr=qorq_Az_sbWo`5WnnKb$fyc6*4n@=G&>n(Eh!-nM{_lgw7A>kJ7}2Vn2SmvkZmJ zXtC-{I0rMV&@$F>MNVqPqn0Lu)=fDB3s8T z5p_pw4@SmJ%`KlKadN5|!K zpQM;$6lpS)!68q3ry>OcV1?qk=u!?eSFaD>5WYV8rs%c;h@)?kGvoxsVeuGN{yz}Z z`ojXLQ*j80F`v67K|6mErzzxSn7&{WJ((e|l|uazDp7SOv`1g4na{lo^>Ya^DegJp z*}wv>!RpPLh(a-TPW90Rl2Ry}8=uu=Zvt(acFa#=i48oRl?j~myLmq^4l|-@b$Z`E zvS$i`PBjxRqK0A)QKS;2VieolFVF9iVv+2yTfN^i-wr64UF*4Y2eh%F8bj5WG`m2f za!7DMaCp{salIeDwnyAGsZsH<J42 z^FR{idxF~@_XK(dJg{`!6ElErG1C*R7#7LI3(r746c*`)l0tEd7qBiu5om^>mgKu@ zio&{jW*o-Xd^uUP=J&BgS^;C%NaPEI%`63&PdL7iHI1XwN5*EvXltr31WYJ6=E zeJ~Y^)Ne1y5Mc8-wt(+a>5JRfd9}zALCX+-{Ve6Sw`)9((VLl5GkW>Zc-IE0up-If zU~H7xLEnVrJs3ylyh7%NZam+ZD~e>xLSQ!Kp04{=5i4cGyHt1QY5EmiG|Dj{Eusd~ zP|Ms`dkw&3sP$+dbs>NEktR+0gO*IThbYm7kzs`>5!vv%_$~YzuD*u_JIgb|%fuRe zTnA^?z4Fs@$LHu+Ok6g8Woa~QysQ1E{o-TkHp`Xw=%k|yPfJ$91c|(~093ls(YK)3 zDvwSKfmtQPvze~tt?~eO8r|m>*-Zi!cR;i{K}Lcl2w)Phe}HYdgLUUiwqx1TS3Q|j zAiOpID1C%32#FImK{LRe$TQHz215_jW)n2xz*=F_VU3V&;yrcIAmQ0y0n_vW% z1(P-X=d74j^fFMQJs99Hz;kwvNlhM1#QjzXMuPmPN)+N3YGzd%cfcVpc>riQwm4V z_%`$RNLIz-QN#5Z&AlT_Bew-Q(M1lP<)|6SHfzUXooRvrdp23&hf<{f`WKc;0#*q7 zn?)PL+NB9G>v)=LDhxE}XlxTsS7c4U7n!BR!&{(Gl{r7P?7VkbGi9Zh{t_7!q=y&tc0<&wHLRGIIHg-uzM$M+Cs{FN z*v;^*K$Rg_F{Q=>>u(KvihZBPte7%Dj*3%x<$g(=J{n|!l}oz8ZpHJOrLAEo%eKT( zMri~^4ukW^4IT8!3eHs5Pd}X|H3oY{4@AEwDX&eix5Ugp%Apw0S<0l;EwVb*A)uiG zVSxj_L!K!RH-$ERr9KTc_t4xdB_=MqXw4;1*g;L5bj1m2QLx6*@EJYY;{?>`8##E- zk{PuXmNngMVzMqAc(GFLjU*k};+4Yt3Jub(mPBCbCPdm>m7PlbQz!;;M`cF9G36TXa`M`6oIo*5mktJp>Fp{LPJGW&V(T>liWNbwI(}Ywu3!i%G2?_+J@)&7 zu+ny)ZW@(0pSxjRI*(oru9%kYp5G_Slma1@E_eq6ZW5IJJoP#Roz%@TO^XZ}>$TGp zWKLS7*@YLQlVv*2j@s@N{9&$T1jc1O7Av8lTRwTTPBaYtAa(Ktrd?6bgKEnOW_mG< zhfSmGsG(zVA6EXziZf86>`=An+lr1aco)`E3v%9AAxMp` z4!INQ6o11ejJ#x7#bW4fK9vXj4TTDnrEIZ9_w-O25eKTs5Ur9m0VkRp>0 zC3stWZ@#ga$)oEz7_r$mdliNMYw2Eh9gU|0XR2&S@!YMLJESj+p2 zEYR5P1`ZL_vS8JV;i5d+a2b}Dr7O+?qt#7P2}L%*rZ^^{B0E^V#CVPW@F-+I4fx() zxJID41-ZqOL>IO~&>3%%D@do94HQWRL0REC?tz7;CFqZSzVOB64h~N3f(d5zU@_98 zM#GsqSQ(&6&#t}kRZFRxT(if&kz!IPvfdzImZgMPLy|b(y*jd-yVLW(qbyqg>}-Ab z&V0~g-J9LRcUk)XHzRM|n{Kog-!1y}zmaV&Y%QQT*u+{CQw+2;=Tqt%J}JE6Stop| zMF&Gu+?!=NA=g75M|bcaMrKO$qO+9szG4#z5>< za!*jeN%zo7w@G)tTo`05EZDNoRvcTK32gZp<2i>Fj;T|8wZ_uLeN93sRvh9uVeK23 ztWH-n1ZRtDL46wPU~4>ZW;(y{B6Q4enx&;t1$N>*VwO-E4=ToW&E~(H{F%SemL$A0 zZ5_#WVOvsTW=r-`Off|YFbdxdHPUD4K?0K>R!X0qb{hD14o0!~ofu$?WqH#T=)|5lXG&*7Ne{Qt^hsTAx93nxyasK?yiZxpGR+{2}R1uU_Ca zh0+FonGfp6_tKTB6!$n-niK}))ARqK~6 zTmy`!N#g3rG~n`H3)%JE^wZf##YcTkK=?KlzY{n{Ky5jc~g2z58XPUggzUl(eRJixLHv>@Eca7dZTVDBw*6SRSWg$msRdn3oB+0 zf!-#D)*DGDY{RnC!Lf7457D@c{HTjq4Uy~S+%D2z{YT3V4=b$^Sn-a_P?#Z5+4%_B zDd~jrluiZm(qiwZGYtNy=l;bGHpuwmD(7#pLWb)WupA8R__>eKs|`Bb}j@`ww(GPDUgf?3_bb>9L|PBV3X zk||~#MOIVl`(CJU4|4Z4Gte$-N$2ZDYUn^?k3po_#vEU{N*@ z>38sNdw}*?x}r5KBL*!!d62tSqHoJ|-<72}+khb3|X(IX>Ob7d=$7 zV>XTam?p!yJwU8p{7Jw4AZPx!jFzL^`{^&q(bpzF)MaKlE>cV@MOr8|nu}|JpZNC6 z>-{vRfJLEr@eudvmqvO7utgZUhU0^}W^~ZlBy6aETSeWG9SS!{+GyGp>5ALjZOScF zmh$dbAS}~G<6X1q`)_2!Y^85w#sa#yx8h=+yY+J`{O-;x1@+){ykWOBu?-5nEH3^I z!)`b`mU4hXpi1MX7Vb(-jGUnf#ooC}E!sT?YI8N<@XiR!)^)h6nPA zq)-^&`?C%@4$!{z@-dXx&#zjznQ=CpY=X{MZ@iO1WF0z}ej<2exh7^4?RQ~s*2>cU z0@)g#5T4}QMVAAC+yKcIZ<7N5M!y0nEYi4GWILhBXI7YTGEPiy|M(;MU;T`8$}{HV zYLdmyTXA8U3fR|o+ zpI)E}i3oe<*Gn;f0BOcxl%?q;UC=7OKkzJl6p{mn&A3CgrtlcOUH$Km zIwVHG$Upc>D@kSt7#E(4_Lu=Cn_{+6WDBJ}I;S}t8aV_xVht7(r7J#}uM0cxoy)~t z_($IWw~veT?FVC;!{dcHVrNf5I4?fZD3t^i zn7}el14t+K%OB0h{WTtgf@Xd`x7VjF_G3PF%~sColBS1e3YR-mylgy>_+ok)V=NuC z%ewO#v+Y{~#t0U{lYz@*6T2D5g`L;^W-!uH3@9b$8UiZu!a|{zP78nuCOl^l?4noV=w|PVTwFj0( zO)lH-W2H>#(BgCAeAP`(GE+yM1zujX9g3B7BG6)k1m#WMMqyj*U#?3W$$>ETqcq*i za=1+|b?UZtlQGt0@@PTHHOo1N%^KK+-4ZLaQO2~pz8G!Q(I@0rfI__&6v*!So&__4 zv(3Xu6oQBg-5Rir^VI8^?+GYNseNO;XSxE<(m^50!zf)?uHev2CmIu$(E4H`(r-Br zUEzQF&j2suob#jl?H9>L7Y?XoGhJP>r(p;nAsG9w!m88gp z4F}Xcn&fB>Qw&Tb`yod|KAUqq^htycRwqxu^|&EP3#j28m-opGN9v$X0FT58S4X$S zR!QzFM)qrIU6_u$(W`T24sfJnRL!s=8aZC-)4^}_*go}j!G%HcR}X0Z_u788io zucHTcoO@Uq6l!bkhAT27dQN=%JJ-kt7e>!MGxY4B80cW%MyanWV4WLX6V?K3QaicH z0STOg%1@QmLACz-#rws#7a%)WA-DdG9Hv5YM!1*PMfc3l4!S%M9K_=T$v8)z#bxY3 za(8*-k1a#?R$|>P{4P2}dN#NRIKWHA+2W@&o&XzLAAJ$LfKTS{m-o`$@*Pr5t*AuQ zF4arLWYZT&pHF`HS&;@`?2TTJWjb!0@CgmHlxIa-fFRZh#(sR`GV0}J@x*OAmDRYo z?qvR(=Zha(X8&AvS!^YrSL^?1emoT0H$}BTI~auGuZq$Fc6vUH%oan<(pp8W|J5+Z z%_c902P??F{B{mIgB4^a{U#}QyJeN5%Ob-{-69er={fHAig(TI0A{wr3GCo988=%H zVRHI}ZL&hd#NVh&W}qf}EDY-6hq{wbXf9Da4uVfm0J6jOI{8bY+Q z`?&^;0l5eEIW7XH@-%Y3wSiC~q-b2p^bX_28V!-7e;5w7dBv-89A zpkVR|eL}Ux9W54`GM(tTTL;}gAF32&iN4n)or?NsxVvFTemI7c(-yAt-A&`K@FBMo z3SNdh&p-_o_RD1{`?dT6r6TS&QIzaquGk8bg9lST<;v<(wuTJtfmF-CVZaQWII za?ph{!57S4{$mu=M3DwcT_ZXpzCfM_j(lP*+B3;TkWcC3t>RQjx@Bj?&v<7SCNlND ztqTliB{I7~MW~KG7M;i}O z+w;H`Z|8ISff_J7a6h>wc^%2c*e%w`Hk2I|`|u!f{3;A|v`Z~64lC@*!JtD`Gmu7Q z{QRc}NRA6*MrVeZeG~(2eub2}OP}h~l#7l5^$E&XCW-s#a`IHNp2-Q<=^43zM)a#6pPGW4;aqF5XBEpo<%tx1oWHMvePS159cQez^gOOAY!Y2t*yV{f3kV1epv za4}D_pY$t+JU7ic232@h!?MA+yE??f<>~x^GVPFLR^0VfiC|e+ z2#OS?@YYI#l9;18B;g>Y=qib^d#*|G?o1UvT}NMx(97^i2cgGCNN}(mjzJ11vqE(qiaTnV;D6aRkxi_R$RyfM zo8CBRk~vL@O^Dz$qGN96gouf~g~6fr^q9JEf`7l60eZrCkm27m^{hc((y)GR0cy$J zYGjo>Kee2B1hER1&|O*xbx@ZU!YRL%)7qiz1Q`8pkiw8u=FS^eLi1)A$tIlzJDZ^O zJX7NsBt0|A4}VouGu7mO^b5nTk@QIfcIHQ=x^_@X=n~JTVm2~yqSWv0TJ&HMl<6pT z^DvSF)hwhxO8{!CA;qv;lROEC&}!(dBvW;SH$3l*CdD5&>32oc1i%*UO`p8~)d){J zuSUdGKW0wfVAsup{vj}PttDDskr_nWQ@dT0!UkNP?8t0wlmk$M2lO&pz{Fj~Iyq{c??pkz{Y1IyXUBYUno><5Nz;D55Ezb67b=hH+@ct zE1{TOZ`jOfk~g|%Y94(Vvn-IRZ4c-Ia&xB}m$Addi5P*Q*`10tbF${QH-7!Ho0Cq6 zGYqxNMUa%Q4c-qt7<#$_LeF|$GU@O$2qs`t1#ThZ>B686WqR~+kQh9!&VsbNGmVK| zv8Pjk!U-o6cMP`umnEO9%bF7whS+>)u4$5|1Z1jEk2>GyQS@?O=$lDpR!;`qaqfUj}=x67XNtXz2)}x zbXf}L&wkYVRmPMd+q^#bBH84^yT5zQvS9_3GM6IRP+;xfDQp*>Sd=WNU35*lJ}Q;j z?_V!S2+omw5>f46&pS?c3X#6;8gr7qJ$KNvQ;2j08EU-R6466D*-UUm@TeuP^U#YU zb3!o1|6|U-f5QZn-*x`(7CH6GG=_d;2CK`I@*+jfQ!!oAa%kx*_un*ChdQ`5p@$c( z<3kDUJ>jK=Df92jFTeh7tI8-SnF_^^7`GhMpt&4&JH@*Zw*QN%OyvUHi~2BPH568U zKr-Pf6jdU5B$8Mc(m>jD1Gw=^`KQDyLU)NfBar1-ml?3tw>%Q1=nAk;y*x4_AUFH~ zG&VzsR2#fSSsn?AU;idK=KbK*=~=7icS4rGPVrILix=f%2jeJQj^*$LSYG^p{{7FE zWEn0S-LlYMrLB3B1Q%5%{u)54ZI-93Cs6k<-8c--LlUAv&h`YV+dFUF# z8^QIv}I@Y2hArK#bU$x(>3;>mdJ;&3+#_RhO)&vQb@gas$ge$~t5p?-3O zEGFwGky10$odu!XLY1dtxrowd#MZ5m^R>G@Tf8WBZkoJ)z-UlQ} za>A#_=eWATH&eCEd%d99yIGwqX!*u9>3G2R3C6hK_{l@#aKMh^9n715oNO{5T|d9_ zce2WbS3;n(Gpa6NE2RWCxRHv<=7FFAe@8?vKOwkNbKG~+)GWFIB-`G1e<G?clU*uu<@6_b{rzxt<{mZZq*byYKE8nloBL& z_F(rH)E*iH&oMK5D0oHq9q%H*3kI)PSW?1Io|!wZ8;T6z&dFP3cfyPWTF}4@dcrkg zR3OIM=pCY7L98f=7c0UsaGjH>DmTzcI(YYo6P&C^v`@~CK*15O>}G}j$$zE)-UOZO zhaiG|_mx>UuQelh8={n-QDlIMsSys!>SaKabUnNd@U=d&kuK2W@Q;JUBuJ|SC-J() zHyC8u?&BjRBt%^_teI*+I&P;lZ()bwfm&eLj`c{OP?<7;S|6FsL!!4D7!H@&%jB6} z2~?f%7+ojJr%^?0yVr4b0(Hpyfm}E8c6}t&)&h@Qv4??D4)>u-XO+$_q_=6|$1Mj^ zr*^0j?H1R;R+RBq?l26FqS?jpqj!k5!D>Al1Tk^Mouq^0)3~oyL$B5*@$d)g_PVDs z7`|67Xbg8*j<;|8&i8LOnSVS`>|L~!I=<+mx9=@# z)fNYoNn)8^pX=czk!!sR{cekje3z3pdV{prr(5CtaF`oLoZ^?P;)@eTs6#6T z-sGDcm|KurJXvln_Ho@MAAwDRQL%C=rQAS~byQ4h_;BDcS}!c1kCS>O(y3gPY?BQK zwv&Od8)42zhxydlj1}|8oc@j-E24zEZuaQgy6}!Oi@q%e9I4DfO%hLsY_doFk_0uO zYkg}%v26@xDr-Uqy?cbZ?OvC8NXKui&>n{*p3UHSVq**kP6yA~haY^)QnJ)#-3yBd z=`k83&L$uR9%Ryl?+MQ>C>F(vQYY&Udl>&@sxku}`+nrP(L)av`Fa^_$@E5bLMc?P zroerWNu8WPITHe7oO=@zoCg7R!(lBDsfx~f?`4-PHIku>6yQU}+bn3#z?8mmU6BcP zqnRELRl%o|$`Kq-A9V5&H*oywr*Fi4*A#(l`ui6@PtIG5KwLQca^EZhxkV{&Qsg=n zgZsVMzyW)P%RwKZMwqB9maGg<31}B~%kyd7b!iEm>)j14J?+BU5IBJpv+-n*|LAjg z9ZwZYa``2ZTh&K`OXyqkyXmd!^Wj5oogO;efyS=WGM_6RI=s3Hmf!!l$rMYH1>FXg zjX^KmQ`WV>0vAY&Lk?($-LQZYm$(_APiNeh?1Z0)zape_=yG}qa;6V@)r4-7K}Q)d za-_)bN2X3*V%@%R1fRKYa?m?#IA#amr{DcD^RgGiw8>Bs3TM1cc2Hf$*I}3fX$qjK z?_-h-tW7Auj&vw%yv}ISl$l8pkgqinqKHxgNn9>$E(ufSFAqMV zP7ly4>IH+sy@~;8oCfK<4o0n*7sorHF{+{N^13+2?lioi{$-^RlPxH%`9d(Z?K`hr}nvB9QL6!Q79m2?aO8Y@IO- z+B(M5&YWJ051rd=TRiNHyIhGq4>D?&lh%2VtKX7dcrph;!fqDV3>ER{*zL3sh{sHzjel^=0gsC)ExhnV|~a zjQMx4_8_~x`P+l&Z@=u@fw8kTym3vck7yU-w*98S4n>!AU`hf1HW`@G$8XhM1gbye zM@MSXcG6CFFauMP$E8zjOn9(1fet>$jt47CU+PCGUDqrXo?JGk&!X9c`!#K{%4w^4 zz;wPV0#&(V{EgX&uU@pUD(LUz1S%L%g7g-HuGs7(u&kgR{MCj~BCe&1lWIN3)j z%PF#(ia8}tW$I*l;hkBHLS#6(KO>G;M-M=D>+t-6$UJgya+7e-a|65{C8hkOP%wRs zIZWyW7sT&@1nKEn$Ax#q^}JRs-icS^reuewGRCWCq&LJ4#uW$PST0B`-rHcnINZVw zIGP38&n)ZHxD@xh@EVwIYZ!7zEj%yB@` z3r2G6l=0|bKBGSpr+MkH4!(ow7T=nW+d4?wgc2Kd!JTs) zf;0~Bu`oK%duu?ol)f>gXcxJwS@FN0B-s0V+F=`V|GXgbjhf8QU3<>s6zS#)e+)>L{-VU`~_&lrd*r3>ywZ({h1#(EYF)&p-+u zU!eBPuaLeM*)0d1>|`)TP4aYa>>tD=eLI~LTA{hfL$Xm+qTB9u00?ZagyMq^Cm?m0 z^#Kq%5@}(doGtP;`Jr#uXIvl~UDywmnfakyN|{ZO45$Q{ek3@XSHgpJ4O19$MV#Zm z(z`~uc+=E#^2N<`+-t2`EUhn9JdDD|(eAl}z(3XpRf)~w6Haf_8S}h5a+0l##coQh zW##|Azxg+xGohsIOScQi-dCo&q}dEp4V1E;A~jS@)mJbJlOE7ZZy*V<%3dzQeV)vK zb~3Eamh?ia7ifeC)58a*oLf-n0a1|Rk?(e|TJe>In zD>&kG>WlMw&c5sfJ13wY@UJ(SIO>6-RhtdM7<+j8JdV-X?loT39*x2SQ6>D%GfN^% z0`=pUu;)78oF>hVo4Fa8)_rvYmK2DpZp}uem1A^l z?w2Km(S{GeB2l!5age3MrrT`NeSR(h|RJU|-h1d0Q| z#s(B~zhe748&Lwcak4!V$~VXDvn+|^VtZV8!OgieMjyb(=>7WM!PH}@|Qo(*UvJ+sq5=^_K>1U zCLJ$2S(eTleNUINdJ1^+luR)8BE~lOE8(gC+{$ zG9HdTGGZI^kQ@n4R6@E%Hl)DjiqFC}3h)jnzTuV84Ri76QShq3gX%2xkXyZ98I4j= zSwXhiDF>r##2T@`T7#06bDYrSy33UB&QS)w?76^PuQJ|EpGymuG`iY!@`BlICRy|C6FV4|oIw`#ZZibN&64xsbvj%eXwboL2-Ck)%36T{P^ZS{X+ znm=w(G53XI`hhk+^ik1_wU%rMTy-D{8#xz#NcO z?O^ixHFSldhh7ppc?niYpx}71q(#*%1`mbtI|}eRnD5kVs_X#@z{YGWN50AE{OJOT z$)@B8H{BpfuZ&G8F|#R|lro(nTd0`7+++?d!a#d)_Ca+ch|g5c-a2ECPY*o+l%jRA zI8gz;0wl}^WV=Hy1f0_B8mFCxF#yDoIA9$s;XB_fwsbly*82Ec*BJOnL2q=0V3|CL zH{^!L>m`S#L0PN@I{iXBp`k0=16N4UNUpA4*s9$?F@D%}sN5a}Pp2=YR_JjW8WT^}9H|CgGvW=87 znIef)%)SL@q*#b;Oa|+^nTycVp2Rz-X_7nDmsvyGi{WZqfmnOinddmcb>gI>Kl*}Y zQ;5s!}zxCtmOt=B#w|`t0p)~IaexpK8JSDK)%1VXa!!*Ivh553lw};`Pw01UD5!ib1AT z+s}&^?e#e<)ys-_@$MNujZk2CpPw~-Ysdr7A-AmQy@D(J>(V~{ef~bBemaU4envO& zdqCE>LD0|Z74$B|C$EH_fJ*w8++^#)Omph;*< zpO-4SVc<_f2UtMR}ZTO1Bvn+jx7kIt&7Tt^w>+WOX3SlBwAqwe$lIPuiW9<-cL^g-)7?!dLl4@-0pgQ_A7 zOW!@bt2``*9lyF)!>rgk)#&jYHI256!^!_m__J6dvgF@+MeGF&ThGo#e?GXV+W!#j z@voS-RMesP)AA{s_?5E_4NNG7)0WA{XUoTOzHOOe%V);pJvUQvA?}}lvP_|JF%>Qxk+LLQ4}o@V zzT^hcwff$Rj29J2ms}wFpkC?yNIWsl{>G;CiD`K?9^iTMGk&l*>AiPNE0s(Cv3wr6 z@4^X@4Q3Lx%O)!0CXyj4W+M|PY7Z-q*vyL)l`4{9@u8=Y6tr1fEW+9^G$B}RfaDez zA`U{un4ZR3{0f44$H$peelxwvvrV>>zfput!q~!v_a73L$na{WYAY%CGg6FRCPoEW zAN86z;kzsD%xR^o8mcweKu0=T@qsmT;sz`Jiu4dSvwwVce; z#P2*@8a>Yh!=HROe;wF1H8O)8n)dj6nW_hhl5wk;=RD)johQd#GF2&&%L1F}#K0jana1Pu6Y+`us7zH_P!8SC zYX^$l7NBa`zF?F8`H3I%ivtHeQzSX0mu~gjqPPHD<)r$ zv^O;2f}^K#hYd$|B8(kJ-^%*qA1s@@T-F7#v}+le;c#v8v3mWKJKj$?NHMLO56q)k z^k?+_$orn-m%iZ&1Uu)~=EvEAV1)&2V(G8nD)BQJna9eLt4O9bBjdtU7>CS^%q~g^ zLCkh4rVYv=Yv?3iX*9BWom+5$?%;u}Z5r>f@9nwCf(uh}rvVd7A)U_T_+OoUOmR7+ zGolh&@QZvdGW~ON{0l>>1-fdFPvlo%y}WjEoxnMPD_0O%wIen|!}c55L3FEqS;bTn zVDt;au95Ufq{?g#c2G+A7UzTTlClpt${)!N2bV|I^G*wI&t4zdCp;OIHKV}27Ag_4 zRCU6UGrRd6Ou1r{B-aad{R%_+dDY5I{GHyP`sRAwb7D+s=LFdPiE_M?eFCg}i(NN7 z99|;NvD|ESS>MC5ecUZinsSuv3rU&+=0P&jWdjIDUJ@~Bf! zal`A%O?$p#nTc~*Sg_>QvIW&|Tm|AlUmi6>cf4+OvEk+{PCnAt4b zrn%sdeMeg{WYyWQopb=YxM66~{RcmO*-Z=TuIGATBECop!r8b;-Zdr8{WE&o^i|{D z$6+%p95B^fhTsH?3BP}9?m^4WVwW{7EWN_@5ygRJl6JaIn9WO*tq>rmKw-q7U*&AQ zto!wA(o9u;cwR(~q~5RAzeb!UvUg2u<80UrkS8|6#)0j;jT1nqtxaq1Th_C{>pWk~(?KpH!%SlUTLRpU6}!nV5oC!l}>J15-c$2lQm z!ulV5?ayJRrA7wfC z1#~hoRIJ8{;;HKCgPz5bgFtWZ2>N>N-{8@+_MBszXTN?veY)>oxV%qQ{o~#;*u0 zSh|yoq$dPGXPQN4&B|0AQtkEA(^#O7fBqjTOt5R;T;}g%@-{6$-nWEoeaSMfm1eH7 zh*H8w$glIxRMn`r0j)JA@5_AJg@?Qg z!s}>VmGlmfX?4l#L-s-?f%AZ3tDRu+cTO`K>|Rj*bNz>wA`7pG1t~H^Q97mELc#pR zV4YEHG^)-O(&wkfM(52qt|%AhX|N@+kj|oW!`G8;@s-efc5mp}KAjwKWBb5%-Z^j0 zmtM29Brdz6VX+!5zEysOqRxm6b+M>!G6)f#h(bX@q|U!V3Pb9FXRk)xCoH5t5wBrd zeA8sG_&E@i63{#ypDT|Xx#|B|jovnSIYP^ZJTd5|%tIF4DT? ztK?gikYg(Z+Rx>|`E)iguMb2XlWd+DFECco0vFVN&sOMp)g9pt&%*^9MlU*iI&I2% za?dkQbCNVq*P*;FK6xQ>9gwfIL$^^}^q{Bl+#NyOR9(H_W8ZeV5JWYr1@`SmbF?=s z?#lXUdrVH+(#Bs}+M8DdLoC**O@_Cf@-H?=~d~GsYikv$ZB!|Dq~9*tY@xCR}<&iChi#+GgR0Eh8H~RGv_Cdq^2lW z@z2kDh5ppmfK@R3L<0cj7})Rj}>)lAlVZy`e+VNe{PZ8 z3G3n&(lw#w5uHw}i8Y^^F$rZ1m7!+R#wTlqJcRE}yw<7Oa{pr>*Ac>IKcy(Fs9baH(J8^nHqLkTEpe z`RUw&xxhFZeSfY&jk#Eak|B;aEn|X%ZQk^fVP8=fdaMYZl+nhHTaR> zNaPYdT?au&Jufl5O@M%JGVUoayy!sK(r?u&BCZf-uug}ZppnE9MKN?A>TABr(j0m2F>Mbj2q z9$5+0LWT4;?XcTD&szUXRhz6oXp`Tt+YRP81SOfUjRm_?Q8%Drr%g%6?AO2Ut(jzkkXHYD9f^BoAe3bWq4kt9i6U#L7(_-YlO0~Q;55+x>5jVpV~yA9 z`3{(3+dp*Eoi<{E^SJgtpnkh~sy%9_E*#3TsGY`6*7ZtL!#n9D$(}{4--059BHl5& z1PueCsz7&xy&6Otc zx&$@qz0wkp#V+*NEZHG#7v{mf&z|r{3|>u&UKLs$bv~jvB2(4CzaYNCl#}yQ3zUs~ zhvQr&FphdR?~SX{PAH@81A4CQUP-+3r0grLT72ynKa!R$diS46yg!fJ(xSZymA4?N z@s|gRb6ST(=E+psiJd3oIQoDcJKy_3`Q+D4IGOkEhwY@=g{@1Q8759s$`cgGX2x{! zN_eyH5ool(52HZd?f3GHq%GI zXQ9J(v1Q?9{8dakgZJq{Bl-4RNW3M77+E)xLN^A*f%YcquU-pSf`lA-nk@QJaGnH4 zY=xGgTg+hr2F)m>e0y%{WX{VYJF#PFe%Rl01Q$DYtSqc2%$)b9j(;shrwY=?bkd2+ zRA$htNRu`N#a1EQ-4Tw7+{*;Z`0zC9b3^ols!wg!AkXOtn3*;)oPZ zxBI2fd)=NA1s9HeSxOXA0`w7y$~G`9D8pc=yt&uxRgbwMc1k0MP*Srnyz_}wKY706b8pZKMn<#jF!!e)L)1%T8rHvSWjU#=~xhG`*lEn&Ww3My4tW z==mDmo%P*knJC+V!7_l8PICjsw+b^(xS67ml23o}R}#xD3UT2WHPb8#Nv4#E6j==f zfWl++WpY^EO&?dpP1&tJ61+n?&d@Q=ye$B75=NdoY|N#CLzX(DTx!`|7!v!;An`=p z8cR-!MJ+1b6VV?mE-~~I4on{e!bjZ>=|Od8L?%ed8F!ZPI=UXb2ey-q1q;qw)7SzP z%ZS75{c|UL>=zH6c;gE*P1eIrqNybL+^mNShpX?KL8XFH?x9F26?2w5bI4o?{0(D^I zKv=sFpRXh9BkM!=N;hf#noS>9{qs-nFLJ0UvKu#}Ks0ugJo}_<(~QYeOt@*AbK&erS{9d_%4i1zR81e0gF z1W^`_`5*Gj_n_dBfU$t!t>Zb;d}FLcpAKGh!U8GQdfQyYHQ zr=L_w61=ZSJ3(e)fNZBv(0HxdV==y5m*D_-%SzvB4`;|N*a3r4cEa(U?3^N6YTnYk z-@NSNgAHFt5}kDcAkw2bDNd7B5Qn-2o4=SgAj0Oc9C;%*L=4<}H+r(kjCB3{%HPQ< zZr;X)1KL6}qqCJ#Lh@%L)FnxGi*}2^vp`wnup1s{L*-*J6ra_E9)@d^e~ zKkbW35y4B!oGZS?fjN>Q-#W!(fsMDdXouX4ddE#P>K$Xhm+?^?#4*sZ4Gm9igbhay z!3H;1^GoW%278L`UD%wkbnZaUISs`pwe#N#&q8oOed8XV4&LG5c+oX!&pcg@SDj#+ zW~~6~Ad#7=CKPvqhuxawxDyN|1c%j4@&lUH^HV36Q256&XtJ1kM}W&>NQb-)z4hVB zf2~$P^f|W#y9^sXor+|49ae}um>cV%i=9`>&*s&H)@jP20&?TzD;`<2KH?;3hdTul z7V|eIgjkH>jJNla^#aQ!FqdSk3mX)c1#<^fOW*!>rs{0)AY-hIZgkI;7HOIm8P8#P z(GG}Y@ek@ZmGLg{ih%%QcQ|HmN9w6}tJ8s-wt)U%ek#)=L>U$*plD2>dB$-$9yXG5 zbN~FZ6F#Dw_-hS)6MROpK)0Ts8L(T8YR0m7{&2#IfW?`e*5*LeU=pWT&#)<=j~a@?@=m#@9g;5Mf^XNF=u^qB!jcE zrF*3<^Vcdr1AgMm!M&tq{_VMY6(y0YqKbHWVW*-6O4uHfCK|ZcdHr*6Hc+s@jHuEnU#qrnK4a>2@6yMK{Y~pN5GRI@oJ(a^dK{BL_TO}VFZ6PJLf;8< zj?)=E+hfGk=pmeP{7y|=m}N-IC1P>m96k%!s`GCGm2|_3$%OHwj?Lw2Rp_wKu9#u4_>>i*Y2~o{$xx zeUZhI9(Sxy04-DhPT`?>@6)CHqn^4J`QhLm_f)1`*rIBb^|)t&U@|n5A^SSq*E5ugf_`hJsHT^TcJrFX%w&tzznQo0hO)3 zjO%@)amDV*{;O0p!Jh677X~6r_ePn|UO!#AxKeX|YB>Yq?pH$h1SOMIAZKefm7_mC zXC2!L7iax^ws9xF_tAe@$_lY7Z>sA2;O%Tfye z{2}Ojt)8(ys+mUCX85s}M~595IbM*+$IJ=pzSIkHphWs*qYj^) z=`o*Shln++7r*E~OSLB0=)()<~KhMk7R zqNb@`w5~6*YC)#TNIPmcaY+ef`YNLX#rq)C#n$yKi4MJOD)1j5CEQBU4%nPF4!2}U zACsue(+qlbf|4%g_p<18;1mQz7mM13mr1L4H_-AUbpkBA^wWzaeNZ4VhH%-T*c$~y z%R{ap?E0(vYm^C1kG&qBAwzDQ;ON3@AmA+?RmGn~Dc4XW0Xx_WV5OVN+?my=*bD1x z=es7WbMj(9SbfatC+|*L_IJy6AeZe%W@+%z!}lCXZc>?jzSw%9yTRKYxM~45A!n*k zFuQH!LOv8UA>mhsPui4gp@tSQ4BA1L5QYE`)j%$ayXhM|d=yJtjE~!$^*5Wv7bC&) zOa7eqW+U78hMOJ8*!lCOpPLNH#K-?NgXmuwx3qNPXhZUVQub2h6DkIy7xXB}{2j8B zqW485yjI_3asW+}JeAoI_z3BBK7L~zUsplOc)MPQgEAi|-&SMgCCGg22!vZVGI53y z@6Eh8QL!Xh0H=^1`;Irzt=u4~OulBPau+zUSg$HSxRxY%C|@6W1w!oV>9_&4b0+*S zR7mWkd%W{P)+&L5nSL0Qk5o~21REhESOnEIO~Q5>YqsDdQe8Fy{b?b@ZOemk_1U6I z_W*9FN_8)N-a8*oLsWK3mB}=D>6QyVgfshmtLWXlpE?w!&jB#kOS5G+zxS^tPKWA2 zu%W3gk)e%?7r}04k2|W7?Tm1&VY0=1*bEC>26Ny&f2rGX)KU(WMRgQcEDsd>AUmF~ zsiO75&60Xvqxn1n+|91GFmI+>$4RBae zw)sBWWNk9NVr#TPRiV(YkiHk$EniY8YnYBSE4n;bvXn=b0=3F`g|{3~Yc%oC;LABD zIH7i8M*NiBmfpl=!NDS4iqS8weoec|7|-hNhT+N=nN}Q{QF-x{ae&2(2Xo*Z++gv8 z0uSZCR*ioJEPP_AlE`LpjT*_ivcu!y9oB8*8C|_^-og(fR?lyRh4faSVQGiO(zsX} z4tRPJNjeZVtV7i&ZTKI{YFw5w)@?t?{PJxmgg>|70*$-rZL;Lo-n~KYc;#yfWd$0{ zNo)()6=B#V!VmoR+$FFY)#w_91D+KGFQcdmK3PbgnZ2Y{)u&i5a{SkNET8m~k9^bt z>qj=m3!J{WuAA(vxv=>UQ%BOc3&XCFbQjM4SDB^Kc2G))OY*6hy?%WR{*UZ_MS&H8 z2f|ymDbVn~b4t0U9~P-SG#0R&Rz0Ab>Ge?^Op)|GS*EIo6a^lUZIcbTedxErt3{5! z%6YcKP8g37q%VGq6O5_vCoKNOTPCzDsuFgRdT!Yj7f#b&Gs9LBr39&$lT=KKq-sG6 zq~2Rps6;sErH?rBWn;ro3$ISrByUkcwnW|+GC;8Mw}!qJbp|#G65MM;ASD6>*_gds z4;jEmOs&t-kTS_`AjQ7scb{LN#3#YrL!|#^svU$l+&I=iva8MM&+)NS=+Z^%e zVbo!vZt1Meh7&dP;qYO%qv}eow$@Grjxwb$JbW?iM%}ULCnfXjX-jk6B_3gEOS>;? z6mFW@z~AGOEx5)MXl~EV)GYVSrz@ienX5k8AnLv{yn11)_Kx7FcP~8%1RiCIjiI%{ z$jF{0$&udn&7RevXcE>7j{0?xV|4bc9Z|QKW|BYLjfEq(vM1`ZYXt z1y1GBLo-^npZY$C=%wQ)_cB9n{d6I3$gN4dQZS_GW*m(K$G`yWhJ&VmWEgJV>qj5` zYN%x8RJ_SV%Xr)2vJ5T_gjZ=vLZ7A4ywO zyXb-doskST$G=^8JY+}2-Pdl<)!heKhqZ#eD(Jp<*H6D3xh^2?wOSu=Lyf|2akWQ} zPmdy2)I|3}4nHB5J_}Bz@EhGLWmpZIsk$Pr7U*_}wkbQ9 zkHND&P!xt_c<)mT&1e+rU>_l5j|6uVTGa>UG}G0B=Vp-|hR7E>Gix`D;P2Vjl}lG2 znPb9ORmh)Lk{vD_Pk&&BnuC;bKSe62nB4GgI$wigAP?laN6@Eqk#|^KL;q!+sfqdq zb1Q5&L5+|Hz+w3~>_M1rX$Y*-VUCj|#+laB$O*I7yL^hF66dZwH~ayVZZ~??3kLbe z6}nR(`~74*M_4PT;NjWn<$Q*dg__X+)=&TJZ$i(q-%njjvbb3&7mkHMC~;I7L@A|& zUX>y$W>e%|up>3%k=GXOwa;h2g=Ci(c)j%2;6`D*Z0ppEer?KZX_4Rd5@cNX5;1fPQK6SDR3Hq#kiWOeH^jY6V;RSV- zbgyE-NzfQIO=CsLs3DwlSo2BnLX%x-`Ojtfq-qkmVD^RlfKnc%$oo`Go3h(?6F=?s z8_cJQUi#SU*MKm;kKd|2B~A(*2)hw>jNUaR&V4)mZu0A^zx3In)2cO~jGL+2C+pPY zdKHFLdl(0JA2N#t%0u%?C5e(&Z9f@OTnTOD!$aputE^p!m+sH&0RaZ%?fLXsO(Qsy z0rFT{Ex=-#u{(vaPJ%T@Z{^&WCy0|PaNX}^`PIvJEpbJJ)1;eqXJvCnYagHzmmCjfz|jQA$Wd@1tT)s|Fd& z&^^%`!QGKIS$5>cs7w|1G-6TuKHovlvvdPVoKnVv#SjXO6w+6vy>o}%T6n3<$QU<$ z2EDF_qp)}$)J7q>CRT&s=ook-$)Mk zR_0G!)wHk6D}U!@m!TQ(%zSM-UCGaUZGF^9k^@Tdc?&b=4McX#?_lcvyZs&XrMDGT zwj>{$}WnJ0p<5$SQ=r)_*&&| z-p9hipk8M#f@}wnnF}$!aST|jjlvF~FRc7(r==C)vV3;o4b40=7;U1I8!3`Z#oSZ1(_ z)&+p7Q6Y_Q%9s-xNZRpQR8DEd=Yk1qhQx5r2_RuNBY#`<{^u>V6JHTC!lJZT0{cGM z;RpRv`0D~v8NFgDzebz|dyxHekuFA87;-3hMR*=GDK7V3C1|JhUc2b3gDJYU*W~bLt%C!_(Ma69QO5&X-Dc{)qW{dU;WZ}+f$2l4_ z{nj>olu!SC<9NjWWwf%4aouR|Q;*;LG29fbD*i{8K}xyp__(kKI&B8*T1p9ps8vS! zsybOCus`)NnVMc2or$ruw?);V%J5&y*FE-K?pqQ8M4F9?jX~gZvV2ZMPf-%E*wvCu zO?4=iSfNno-7p<)7_a3olQ$VQ9HF`FGPw&rEsOrp=N#+-0C&`^tD1KW|DF7|-e5bv zLpFR7Se?;}FO2=06E<%TWyz=diufKDrd8~C%;Z(5TU41EWAnXpYd6{F#wJX#Kkl?E zRYhI#mb#nIMz5}G6(PSmdJX(UU9;L_`J}432j?zzzoku+os$nKHVIZtLLNC3LoD~Z zAE{U1CRI1Ez9d1H^etv%DC`YG*l|BH7Wfp(6laMYCg8>MWfLG+tig*%ANw8b-tzkY zj@fC+yx^)iKf=PLusdWse>*H-fYv3I$(f!f8Q(SR$h6op`6Ks^=PUcz=m(mSvZxW)^_Lqf14pfKb#DdgekkYZD;^6qE0syc-U zlUf|`HQ)?yGps-7jh$ZuZt(uqo6CP^8KJN`eGKom(?@t)VWCj#pXh~2DPWOEWq=SV z;wVrU15eUvRkr(Lt?bSUB0lxlCRrkBNP?}?ma#Oo~_#WfRC6*!gI>uz^bTinhxejB$9l^&bmFfV_vn# zAk!e|nAhZ9uBe{w)FAaKVz?vs>9fu};)Gl(@7Ejtk7d*AOSRfz>CnZ360Cwvm>oN9 z$Sqd1*>lM40zDuea@(rDIb(%(v*+nq$Av>~npvr-3&Nt`Q#& z#?2c;Z4qqg1Vf6wnLY4_uh(;Wt8cNx-CjK(iAx&YZSX zvKto8LB#}iSXf&P)vQnc9?ul*{`Q0Z<7BA|N4q&@(e6e{nM{#HD&`~4a|;UTN1m}F z-4@xMuufp(&*tU(BzPn^MW;KXT%-;0YenDFJluL*eB*`7KW6w}ltN!nE( zsj59yQ6o8JN5xs9+(j z+#yPjz6^xdYrKwVZZU@iNxVXj4PK9=#XxWDRKs9x3SAWz^mjfNYPmzq1r{zGvaoD) z=hJnvB}b-$E@Jp{UxRERE^AlG&xaz-1yGe2(mO=0JQV$H6fWcKj2IWJ*ou$QVTU!G zIr`i;e^d9!vUc8Om%S`?@<83jAep%i8+?)Ae5(>VfXjjTrb&n$pD2z7bT^L-x-1z! zJ0qIBkq9o{{h=b&@Sq{I=nPBZ_4=%lFTwvDj)}3%|7a|X96Ngc*`rZw8P>=9;)X2c zFI^TbEP|HUNRjImFRGM5(JHi#iw{JtR&?-ky&OL0FPtkILGi+I9DM^fP>5&M{mgQQ zhKq}F;bjZUe$E~5ED01ut8|@;WcQ{i4N@Jt7o?v&FsE0V8jdv&pNQ8mgWf$tLxT=h zzV1-(iZJkNV>H_iMycH4)NfC4U?Z?bjLI{wPX==ELB!X8K0R<`ho-b)9m(ZZLGQv) zsW(fYlvB#x6xj)7cR(tN`3_v?)&&>MKw;z-Rldd`Ox8{pOOSbB*lobGXvTd5y#-do zKqY*nk*XlwLw90I(Tui0WE)6jZjd70VNw@d0PF*fU;EgLnkO)FFlr**E`EEm2{m0m zzw&po>Xktav^Z<@5S|3pWJ2$uZkP$&f_4oODP;n?H_ak-3 zf(x9&WMX~tIAX*4AWk^T{Nli{WtW!AMkFllV>$d)Q5kC7iP3!$@yIhJV7%!sO!H#P z96MnOUpOSEaAacDBiUDVCTnpr%3nZwUAPJ$*-R%aexh=iB9DN{nx7q>MHkSxQHxBU zsJeyak=<|(loP^|Nh!&P#;Gb0MZGIWz0{MW8>nj%NSp{NTH&58>Kb+GU_U|Mr zi#@WUD+uOb@i!`@M~XF5C*yBcBkZAZTe*WN;iL8uCSnRh9(jI1l7Tt207fh{T*v(G zHRhmtePknTl+44?kgL0dw^DQm0xAf>Z2{pEM>B2*9JdK4?)$;f97`>!S0pB8k;FPK z?I%3|Nqt0vw444X=+(8!ibasEM|vgXQhKOf?mLc4EN(b@p$&4_4KZ=QoAR&KeqPGI zBF@ub=$buQcirb$5C*sJ&CuaW$r#dN>Lra=N7sn!d7Hdim9_r4UP+RRvO>RhFUN{wEtB?B1?XQ{K*c96!Gg6Ye+q>R8Y^mHGjxWuQ(8JL7nE21N_AezK;dg?%6## znWq=BU)u{Dqiaa1;5dCuI^@PS!d_|RtX)&KNpd9Tr?#rrN!mSjP1zK{!HC%bE;FRr z_QrMurs)4WR%rs4=V$$w$mU5vCOvv7Qb;L*lqQFYLD8TQnRQ&W4$^rG>orJymrs|@ zzaqutda;MDc7BqeTnq&y@YjK9<>GB1d@=0SCfn(auN*!y+y4>9JMH_3x&AcU@8=Y> zy6!H^U%vFeDN9N}E`f~e{>#{y2``hB091Fzs!5Dyp>DDe`0i`?xGr-ll&yiwc6eKR z+Ii;+694q>uS^!?_JVaY$!!;2Wvw(@WqnF1`zg{##jNyA@Xn#HNn5l9{B`_J5b!|i zmX7sV)jd*xDEwZzkeNj7Ql~H+czL&0} zoh78_6P5G5)XZ>RUYo-SE@-*Bm%e{%M<**)%yTcIU)Ao9Z*gY!O`KIFEVoS^#@ zH+@>QXJzS7-iyEQT~K&9W?=NmR2_)uQFO^m<>j!B+p4@kqiDp4{k0uUws{=8?=k_D zld+`EugUz9rwMyiU-)t}S;H+fa$%cPWafm^DdiRl<}3y|knec!^F0BnVck)`Bth&{ zU3NISUR=-L;2px_B|ZGC zZ@TJkK%z1`9GYnRnCx&!ra;?rrChgl#&Fmj@69uhgV?5SyBBJ1w!>CKF28EQup6$n zDguwuTh*7vdU{C-)%cAPCp3_7$ADFfF?#wOa?Ht|O#I`P?LTre*^`n_fALomYprOD)|1sJHwDNsRkQcTl_Voct5|3^~TT52C^Yq>w%fEdVvbWI?Q` zI`p>hCeM*D)5)l^5$2;@jgx;N)9~NYX3dH+`H{_qifppanjdjt6ttN6k)xFIeF}6j z4NV|hWRC;0=(WoAq7+ew;;7$gVW*~5o3B}`#3Srj2i`8vtI{eYDu#WilE^{+E$!&L zfjLl9e!aJgQ0J>r=KWzPG5_K4~XOfvsl3Z9tnV%hbw#nXJc!NCp__w3_V(|YAQF^wqo1^L z%Z#|NZ-SKFs7k3WN~xzvI~B8h(mG!~T}A1Iwf^y5D<<{MYt^pvy(&E{UB~}Gbni)a z@2{BJYt`yWui&b*Ffu7Ro)q%y$Sv(v>2g6G8FK4WRI2ecYeu6SUpLE|qVP4veg6V{ zU0g?6wB?a$k_O){Icoo2m-Q*|Lb(|Kz9wAliLdFZJAU~3(6h|&cmlJ@aQ@&z(VOCE&bR|3t65I$dTZBUKf=2bSe@hV{we*f+lv*uwL1v{IJm;X-pSJ6N@w^B4Uu) zB{sq#*Q<}F7HzQp(waX%W*1@~t0j2>ggz}2G+Y8zQ^mLY_fs~Wps6)Zyoi*Yx(=VDF zhhFCvkL~)mnN>oQFM9pL=J&~JYYT4|wh`OSJWd*=+)R-a5Xti^W7Y*+{8s&!S{5O7 zOEXX>0CsxiIRx8Bo`7f^QNsZ+hRDOAsCdp`?0~60GgWQrTApPbbm1j4ivu~KxD2ui z*-)X20zkF%AwEDYpyh&tP_ct7PKi^xd~`>GOXw6$K8?&9x8`HR+dXwY?SuwLv$^4l zl2Jjb**K1V!;X?8lKON@CAwFn&V@yzu8+Z1BRy@_G+XVFLyY341D=fvJ*^Kqttyfx z1{7*^i2+Mzl}BRpG?eA>_0WNYt+qNtyY*N?Q5)vg4j7pbWy22{!!7^z)%O*a8-}k4 zSXj0a5eZNS2qp)+7mXW>j&G8-VJd8bgbm|3`j**?CYM@n7`kj|%CeK$CPQ`=B#r}) z?%h$xLk8x2D(m$e$LQk8tk^WouD+8wVfKFy{&SWoF75jIojs(8TaLtq({ablBGxKO zxsM{{5V7)WK=c&(_WS90az)a}zbLL3+J1Y&4;NktvxFbW zij3JWU`0-X{lW6cJKn`2Ck36)okE+y^4t-ebMbS3`Nt2wWU>|$`+oW{IpV_Q*f-5A z!#PUXOp&u7nhpIvNxUAPy^03kHQ*uYd3VB=1|1AsK8fwhib?4K)~@C98--^@g&y_1 zT+qeEaj^KHRr~yVt3C9>PKEKg^WOKTr-tKCFVNhci(ZK<+H69~b6kndmpB57-K<#I z>7=$t&HTV}&zZ|g$AwJ|%hq|aU}r?01{qw|c$HAS@-*2U!NyPnwG?(yKZ@@4X_Kv* z3f-*+0q#8Ls;=R8I**$F$@GpDP)`PM(t%=M<>1Snls^HTAMq3xx}}CMg9QDm$#>~I zS}$8GDB$0I-tMeZK2w`#+9^MseNb+uV?l3;UT$(pC%*F4t7L-YG<-+2*nJRpU ztG8hP(Aloo3=(rk!Zw`4&T|6E1mEE+8|-Q2abf?$(#mr;3^)EO)H+D8Lby{Hvex^) z=VRc$D5tAKb$7$`GW0#gqFVneUDzVNUVPL}i0-eZcyt2Z|#B73AKWBLU|E+UYK^9$gV-%ngW7 zuS-j4!~J~Yy{lBMa7QY$O_QiJl7MXwPmgYf)*xrt7tSa!w=BkY&!5mSUFBp^sGLi= z$L(QWap8?X7UmVq$Y;@6v!Gf?Pw!Bd$Z%ig{yd;SEfzKLkQjA*Gq{{^G$urh?V*Z` z7pJ~qvL=uCIo~HWE^JLMn^}`Hl=37+j#DwnRBGJat%uy{R(yf8#{WbR{1zS3!sS`ZM^2?@Au0XLy8 zF+HI4%U9p(T-5N@z8{vnz3iR11)BudrRf2WztsEX&PDNWo`3!JqNe!+kpd8-z+ zLE43>U#W++R#Iqz`+d($4e~6sY8wbLkX?sP^(=K#=p*k&`3})$fcZ`3PNicUofGiJ z5n0x6Kls7ax=<5>R;BjtCOci&B!LAQwMMF@lm{qMNyY5(L0Xtq%qk|8sne_o2d2vP zq+1*xbW+(nBb8a4%g1wz4RlY1bUl=)w5V$5Uymq~CV1$8{SERfjG=QB%Y)R(HZ ziG}|ZC4{in1d82|b>*%cn^tZzJ)|hm;O2UL(Cv(8H#0}PaF}uXZe_>IwYsH0d)b-& zu}Ioj8eK?(&eSL3e%SeeO`sE+lz?1=v^XFHRSi=D*1eH7_3kwaNMENjRd}-d+CiEqQ8BqF)=xL)^oHE_ z`65%#QFpzpMtl<{YX~N)SW*+3#H%HPo^`>Ua}ZB!Kn;4xnM`^1aqC()GLAXR$y7}U zTJI4$$z-au`rqqF+$&?MVAo+3u+~$`B#Nw|Vh+MO{JOYWX%H+yEVZgmgRX#M)<&GB z&^GhIX-J3MeCwkQ^*2nI_+97kZjn>mFyX>=R3DjP;xeU#_Qdm4%#q+yMZIi~&tPUH?Z@hi%rZfGz*$g8X^px^7eo}vdr{|A+E=l`2(`a4#!d>|Kw=>#>5Y0MZ3V{5LVD|rPR*cawqzW+0gRKu z4Fa}Kp#Ar;`>4d`UrztmV(Z8UHt$xD6GxGN$4sNbAb)hmyM6q~2cOQTt$2C*#P}R_ z1vI`dFnOeqt=~LOR=hGE35cObMJbyoBR4eG!J z03h73#`6C0Fdt9mnG1}bJTwjm$=^IR{a>Ga-=ZKn&a zpaO!53MvEzB8bYOAVSy#bktFBP?^CsvbYQ`;KHE7?|G8okVrHK5*qElemW#)c?0MD z=6Rp@S-xNTf}l-aGZTgJDxoe@y@z{=Qvv7E9%0I8o#zE#WZeLjSGtBA2a)k17X(hGF8dqTjWLna=>>e_CUF4{B+Z(9S{q|!v_M=6xAUW;oDR_G!DJS_Y3bE zvN{Ov;Gjb;`J5(_8d*)JPTlK+d->F9x^OHW1f;XUpM~{u;M+-)R{!?HoBvqNNGE9x z4{7I6IdRL>q*tI`rarJpSS0M=9HM)a8eU#losXfi=1a#ofjfJJ2RzvdyQ>)~PqY%kl55J2~$j==sPp zVd%8EWh*Jb`oMdnTGS2MW7F3ip)es|b<3HvZT>Cqa{PPi^FNA{JDDG+dhNeVd!osY z{>_hcvfDR1v6H*htls4=#q?6-CKZb%Dd>k7Y)Fet4}wA~WxKdN;p%q9oycl%3eprP z_BY^x^(lE_d7$8jgkn`eE5c5wnq@6AqiTJ#43-{xK(aG%;lhP2vKG%~`KO9ChRd*E zvo!Lumj=pN7B1}L?)ACjr{m)Q>C7Ik){qM<;S^H%4@+E@15vLYwAETX`@n_4>y4YU z#nAnvrEiLiSHj0EL?Tpdzr{~D1mZZ@;-IB_Njsyi55$U@1FBrF8?rP7-U8P>Nx-c!)inT)(m6ki#^vs+svDd zey$y@-ovTwVSXmy-216)1=;qT@xj!Z!J~v?Aih^f#r}C0UjrQlf4;2v^IbWT!c64zTqK^^{ zC{v`90{@lJL9mUo)_)HHzt` zNCy>*1(sFnEi*?=;0<2e=Pk&s5%`^&V>D04QAzm065|a`O`O zY<6+6t`z-tRJXttY>5zM?w>5&vKz2$G)q=5vjC=bV2$EZP0Cq(>~EIM2y9YRPV7=y zX;a9Ux<@_;Vd>qoc1j>*V?gC)`PvY5v$Tuf5|Zq{C;kLVE;I%tK*9rB4I3No-HSexGmtH}7!9}_0UFqKBovy3|UNf{8 zH(r9yNnQff!=bwj^>(3k(tM!nT{Q*#6>xF%0=x&W*YSt11^q^{C)%E1VTT5*R>63} zj^JQrSEwKV_kZ5CQLNm!LrLKshS1nqQYP)8E8t;&3N-s*`i(EJuD@pXsX6LXjBc*6 z>CCW=SW31**;{ppLu38pn-DwBfW=U(Cd7H|^xFTL^N2FJJC8gb9S0p2U~w6lPfVtm z)f7pfVvz}Z1HXR??khmR-tNAl#RInX_St~2cBf;HzaW6PebATt8xtVj+|@XZTyrG} zX5o~98{j_jGoMEA)a-X88bk-FaRoM=IJc!6+jzygEgs28jS9_36`rGG6A z?P;@aMQmiSDdetvbC?>_FzwRbs7~l`#mN7?S3755^j|M{d`T;3hX*Et@|g`jeexo{ z8aL3!u97^}Ly3-$N-@|unWk6*L6iZHA(yz|44?|=G5sZKuM3h=2o7T48Okq`#VB)&5NelpC=N>+C#0_COH=o3bNSeb`SPX- zc*PE}+e9&imbH9N3k)mOc?0gokR4h_brrvcx0#!NhWWD{7d8xL z$4gk@V$xrheeM4&dm!0tpq~Y6rE<{ow{3}=v*$%A7hG)bJJ0i(qs2e8_KT!(dv?jy~ z)2f!%gkdwIdfk+IUN*EU_eEEStOC9IG{ufNygPQjQ-ND(6Nq^Q6l`* z{~hT5<*!Z1xjJXf%jBvPBPZT$pZ^8L+@nYz^wL9$=s0YZ_qwGiu6wQqVqf)uB13=# z8c>P=MKy*Rb1dY%6*1&;mRus&p&S@$Dd|o*UI@Ic_g?#q=>rxkEsbRhhTfDcIzG&> z`7h@t2kjMgxTTZ6NF$d>0gwgnk)tJ6@&~R>H|>H2|q7UW(#gtQIud#{?OJ9up zCL?(#24^mroMd3%yXA(ZT5SwghhotePzVi|Pj!d3$l4*ynHJooP;Zi655+oCW7Y!% z4KOC&BU>qhmtY9(R$FQ-tVUn%Mm=g5c=?~;|9U*&apLW^74YCPWDIEUlvDxz$#5q?1d$INOF}&rx4@qH)Kc2u5bfO z71mA^%S$20nP)FS3}*#TU*+++ma4KZNUX++-fRgNlf#f8y#Zo6t7i?$wNOEm!UO+j zAE%gZBFF)=Q+S@!9oiJWjI+h9$!*Z3*}chaA^KE~mWmuZO6+1utPb++R8N^DHHg!O zX|l~$oOKkFLXjjY_HJZ9gmm*rsrZ4C+DqmP?e?USytLT8QOE+-=A2i zr>mL0B@DOAYOQz=a4B97ZkA)Ary4YNq%E?Km9-G5(L}C`nmD_|63DjVaNO5f2l8NR zg6BMwars{3kBUf@EBVaKxE!OHk0{b)R5wDePA_OxKJs4feL%9vNc?3icj(~M2jz1x zEtAaI=bx!O$UWhG5a?L4piUOs4Y4Ew)fh2`zX*30NYDY5`=$}G%q^Lt7c`Sn;CZj6 zjh}#ZxBKL`=tYYb9phYrn~)A%OP_>h*RdsXSkBO}DPi|>4D)4uhDKQuocF6b=k0_| zmfR6mC>pTV8e52#`5#w7^Lz>^^X>vU;4JZ=yg*!Oh)4~&U`z`)9fmzdEl)ekpj@(s zmp*H^=%NGzZJ%Bpm>u9U1~5GP7FKX^-ks-f|G@9ER4sN|hu4ZiG1AVXGZ5z@qSFyva_aM2u$h0th_AkpL$Q@U* z)+|Q9Y%;TWG8v>|kCVL+&_Y5=RBUY`$X_sMU@SN}r&%!Qa{QI^(%Yh}fMx-%L7jq{ znKhhd-k{6p0!#~@4sVu~cwGuz%qyhV$d<{P!W$tQgif*1t-vwcV)9g{$v$!rLSDJt zW?7c9GH|nR&Ri{bi|>YMpNq3qLlH(QvMl-b zS&#zW_hPqR$Tlri90jf-Br7DDg^XcFRwC7SenD z)0r((8+hlun*~@YUP`|h;ds9zI4j%1&M6wcdQ7&2mF;j|H2peB+!A0y)1f!lR*;<0x-y;D zy(9Hz(BDfj5dGUt#iEKvE`1Ly8ip1?%K!+~`nc!#*g%e{O3aU9tzVz~ZYU-&o8_gD zL0<_ZId|!gm3Jc7`!*2#@BZjbs$x!)o0k4;uGT<-4^qkYjt1Hf<70xd)tjti-0M35 zAB~5#&xrwLg|%-ZXR&ZOFNLRWcEhHetzJ+L-z=-e_n1V_RANe9OW%TM)pZ_J!Vh?0 zFQ*!J`}T&5#e9teJr*O`_igpmH7A~TSGt>qsr60wPE*_hAwqSptRtd^bSdh2wRFE! z{m|3c^zYya9yu*6Cwt_r4t;IiRE!oac-`Z8%hye-&Tmg;yiE>1XRFR7vsLF4ifN|E z$5bpbg>)!x(}~QA2m^y-BBQ=Dqfen175Kw@JOO%O=Q;WQSmm`M?3id!4&^%Cp~=7n zjA1C`PpctFadr&ka4=()rikYwVF-q?^8KM|2K#&(gMcB;K%>^DNao;8mHrPQ`0&|W zLk|uHq{?L&{;CSfRMrGx{-d8?4I{D$JhDXEuroErpSFY3I?~pB>&>{=OsINa7gI+H z*%ebbua`!SnBlF8Vk#&oNsql?V6DkiEq7faERL$BTSAsi!$Ph8$Tf-;u9?a;R0Wj+ z|Ihbc8I84sTjnGLZQypu4|4IU1$qH)DxrfgO6!GO0qEYh$kxw6mDC#E1NxHU5N%}E zL7D&-*m>$~+8tO={WRN8UcdiOf3nobaN5w%4l_VxP|QY(tfyihM)yoxMT&#A2A}s| z!TU&<3=P}^(JOd(JjXwoQyx8_D5mQtvg?W6blW~%% z^T^)kY{fch2ABql0j~RMDpvPWE_5+uDvLSo%zJ>eW~!gA@R9!fKI` z3Ja_IKp%`82#10!2q$EF6?2UG zg6cHht?Bv7!64XCLyH`4o5yDs*v%13xUd_199LV1OJ~@|P8ZW!RQSbze@Nok`MFLU zd;x`o5rOO!ibK1DwaT9MGxfJw{v?1yL=p^h5M}evIQ*98p9FiZ?(+(j@$%^ ztm(ggkED&}o;dLa2dEZDOnffIY@!#yc=A&=TG~r~~`_mFh z)(c}gbY=+JLotOE$;Wze&q$iuL-)HQFCUVg?e!b*To!zf zgRUx`AMm^-NSfNG81UF5KmIEEyxBAL0*ukLOA~~jI6$)rr)^wVGJ?z@hwVrm zR%7>?VSYj5K*z^oRP5MCR(hCoIEKrL7Sct?zB^n(hP6CLNzLpYIsqAZ9T?;t=iEDF z-VK;!(-1J}zx3~~_#cyP`EX6btK`aPYmoDn{*V=B#T@+z#_dI5Z%Ss><88ky*cr}oHE zdmRqz`Fg&3;H$_4S0AY3sdu^+a*zxgiIk9v0BMW$q&pM}*MhgfM$JG*48K&Ws8qNg zx<`9t0}A+a3OSg~>cQmZIN@k@#T)Ml>&x1v&wrr)$4W;w3;NwLkG=fmbs<1t(-_n& zDCagyvw<@e-qHv>vVQi8h}M{GRL>8}=6x{mbNw&(&OiEY-dp`|or*yZ+J0X1*x+E* zR6L%}@xHKbOh${wym#+gGcB99*({Duyy{sIY^e`KrnEXx@2RG*2XBTDv1VqI`$Pt2 z9?!6C5PE$1guUqcmHa~+EAdt;HY*If5NL182ljvTEe(5Ac)GWktzKzU)F>6ur?^D& zK`sj`N-%w!qtXkG2c8T(?A@VUEX<>K%t-{o&(mZhUoSYwg&=$%5Qu*yRO7WRoKrC9 zfJdUIkujvn=Zv(Dm#3;BK=6RN7->+Rg3-55fqG0|Z`-E9>NnS>6Jwb>R^MdjjeRZr zR?&AXo9di)eX`Q+m+4UpYfzl9m&O*EA(vE-4rQ-99)oI`Rzu@W7BI{wi$9ef6RDT; z5WKmnW*TWe!HYpklXhvvoQgR++@R75+i@^!WKfREf|l&v(CV+91x%$PwHjK!Yo_L^ zkSMueynqP{S;-Czc(g+Q{y|zNL1LsG^gg1Klz9)h z#|iT#Tcinmb#>r2W=z)W%csif6)R}48hLydvxCNg){MmQWDK0xeX)`;Fp|+2sU9Hr z9MS4*(a?~Sp36FUIkm(hLMNR6sX^KSd z@&MGI{5v?cwx9wXrU?Xu;iHCL{l?1@Qw&f3x39L6H7`tlV3!#xvM2_K_0y?XBb`j9 z^7FvK2xH{8LxDU(tw3@P4z^L_4U+(R8hvvtgqaXKVk<&My44ebmagT0rN*P7?8G_9 z)n@VEM-($ak^59^E!`0j@4b9lrtg4a571I_|60pyI_$oY#`AO`t z!e+e_VqHj!63_01=DJMZ2Ht=om3Kl_K-NuZi$PAB3=Tw%AtW=P=!h6}+2Foc(iPFI zTFg13>WH`xNA^T#`VP8aOWvB<1JFviU9#P8K(XSbtbpX`4EIdmMot4{vy#!#4!smQ z=z^o`_$9Mf!uWf*%Q)xw*Lh8zi{0=uaKw8MV9r!F`e>!b-)7L|D8aw)%i5f=;BO=r z?f%*X?!clOJX!(c-wm4(nz!n zSo)?H4Cj!5yG#XY3m}VyS!Y~h7LD6JIO^Gt*FLavI-S=+NPYFeHcN&HHm;8od&5?U zGO&3Glc2E2-bb)a37H!ErFxG7ae)xAja}lMG-kv*r3Rw%4#gI?EuN5b>QG>Qe8}Z= zIL51nTxx^z{R}LRNV#TMBr(ZUD__Q_k$$v5_$aiPbCBB!^yc`^D8YJ^tek>fGtF+O zQBpH|Odgk`z7^({iv4%w=RYmcj7QbodA}&oNIBAz)=I3{P2=rMw)EL56gFINU0NZmbaMM ziGvgaqRBcc_6*s`fsR@C4sbm*;;XORCz;CQ9(P0AVvdOlIh*Fh$$RKF)nJ5u-&s|Tl-HXNN`1_!Ks#mTF?uc07j!JREXI4OLrH9T}L0VpZ z0pdBu^ohWW{(JrG^38CBwHYtg@qX)gFVqFxRhleI%+a4!k!*G$V z!i@Zn>aAqYXr;nVjEs-XkWoW1RTQZJs%d!%_l|Uz=kU+gn00QO!@3|qoTgakmJrk- zPbV3qj;8^$f#H;4BCg@<>%8-Q+hT^ov-~xZf#`mT@#-ys9d8=Bw~$@yujJOAz8PHnO)N26x)NbhKN#fbxMDP}3CC6k#UiaexZ z`?$Td#;wG_WxUl3MV*`J1L2vL*y8mG9S% z%r!s^DNHBJd3&bj(lxMo8v+F{qx^3i`Ou zVKc(~xBqFgw+VzlJu&GN$zbOYIWauo^g5^G~PxH@p37_$PJ zFvy@6M&e-<_~|Frfu&;PoL$SRj6Xu-sH1sx2 z!90_JGbDw#3v3$_5+O??$j6cO06D+p3y&2lnqeQOG`!Ad3v~e+D-6EM zR-&5sU*6iW;}UTqv@ zJ%+*gPwMZN|ID=NeD{J|A34iz)p25^K+V>ORp)bxxk`~P3~Hm`piyWLd3=Y9b<~~V z@jUDU$L??>2Zj|0R~~owaK38Q?1!&+f^2GXh=#M= z^HHdYRn{;N=Sn#G&G(^CT}vMhyY{O3oJcRgk5i{k(*aG_ZLo)Zf>jn;(c^=g1^UMK zb{J`fpPxTCobZwfZvMZ~Eg+kmIE7Va2A~}jlS@ISB(^`Y*z2zR0wl1Cyt07)H(88R zjdcCF!bEPoYlk91STbwS1yWof-i$Q)dI2`CbPBrVx7;)w4E{OL(u|Kl!fdwe{Koov zHOda2S}gSX>GKYiJ<&%LS6|ChEe$?Hc1d(0g^?9e+f;+{+NeDGh+^fOwwOb7t+Z2c z((`2CMTi1kR-B|yfcZHmc_d%V%ca|hV`*zxB-NfY0(M4+*>LQj5^?nNes_~qsrcHr zPmQ8 zBW*Te@P_T0V|l-$-WR``koh8RlrXP}fw{Jrn_mN21MiNAEM>C@nT&4D0CIB9dhSF^ zZ?i&xqo#Y5ajjdE*e~9aEHqh;QlH~;(#_6tIC0Qrxmf`DKE>Rj$Zb%9;c9sKkln5h z!L-vFDB8=Bbdnn~?JU$a1DyiwtIUDGZGjl0b6DAxA?=q!QSgj@(yBzVBvi>T27Ym% zI?i=Z)HlN63^!M*l9;0;dnVKqV@qn4w}UdSz+?YMb5rQMqu+>AD=0F&xE|+wY)oveC|qCnN8V4iur;f_prz| zi$11AsV-=orXSF8!hJs2q@-?@)d!_1YB<~ddqV4jP|y&{x-kTsFD&GomY$Y2hYW>g zDnW&^S-viyftccET9`A`Q!SSsG-RT+kPSEoJVA!;Y8>+@x%V1=P?a`bNB-JsOEMY$GQy^V=4L-_CBtDQ4ary^BNmiQ?197g*`ablg(SGi@>3xt0ES zwq$PM+_sovuj3NDMoTwFCI{g$Eg!aQG-{jG(%Gtekym6Lpsk?gmqs2ZnJR=iej@T^ z8OOL5S`s+s4V|Oy(+HEp>ZSEHa}=CY-~HMv{}esEuV1=)C)qz*y488V4&=0%pM98O z8YvKz0@I7BQY>ljC&=g40Q6}svcr-N4P9CwMPL@nKYV=>#@>^(9D`8yCfV&Vu=ReILwmh?AiS*?lrU#F$w9aU0IkokeSziv z4ZIZ*MS(hgxwKSzg52hPF4!sQlV@;hgC2z14@%=lkJ+xxF37<_E_mcEdERN2%A$_Y zSA|Fg+AJ!iQ+Rq3H(kdhMSTkTm}fZYetM7kAp6av4UeA9WZN*D-7aCZ4pB*)ezWDP zCd0DJk8_*UvNJ4B>=~aoGc3&%0~OClsMrE=^H)Ju%zd{ASrM>^2lMDPvkSyn|5eO6 z110U6Fs$fnmo|Y=CSI9Kqb5+kGFuF>gs5-w)Or5}3s%L{D=L)( z^cC3(9v;b5?jw!T1q*N$nh?t~mI)ur1hTz{oyAfAdr6zf1do-QFaC-oIWatTn!#f$ z#bi)qBNb~*Ss4k=EV23SnSn3R!g1Y<17LNVZ;3>7(eTiKjHuJ%|7 z#iTuSwQsivkfkGePoKPmbDmQzxZ|z{?GIo*!5h#S)(fzV9>p}TD?r!|Yj9>EA+IK^ zi`zz$!s7S|6bnR-30N#v8S_hRH$UuvRTKQv1D5^6Y-$Ug*wwMp25m@BB#X1eor*)` zAh!c#cH4o-1Uo?|#v+ZJN}EA6@}>!RebfJ8!uuwBqq%>`mE2(G0XuOdCebV{H%KuL zDAI52j!)#DGvzPqXVghgf$7Qe%@tNa9_*U?sW~VFkChrLWqOaJohu;1>ThslOBoDQEy{!KntU&BIGWHRmWDyDns z=AgB2pls`<&;tPjik);ZI8k^861Rd|((4i1B^{hhRTmfX4(FqwL*CGk(<8bbQN%&+ z!e+Nh|Cw>$xX-n-pT#7|gB;$b`7{zCc4I zK!La_KntvRNz9PTz1Om5rU9wFc2)_qSa{LDb4G`vGO%VQAXk+Xg~bE-Nx6_THn@C8 z$rW-{zJqhp(2v$YmH~e_w41@D$vmgc4J$s|@{^yKAQ+H+=^klwh5Y@7K zT*UeL%}_XzNxIzYJ=-Y-C>Y=;&_z&u*!_dUkJe^VH z?`bBUw%oCEc3=}r`2QRGtKY=__WyqN-wXa%vVvlkQY3C7{Qal>6_Yk?{pW+0UC(U1 zFehGat@KD|fGpo#Nna#-_3CP99W?F<(Aj68PYOuEOl&lQ#V5_5tRW@)NI1%-PR&kJ548xVXT|A zpxzI3Q%#s?fA!4!Wc>@{7nPWyC5K|bbYxPo7>qVFh_-m@1wFD{`m)!Z2-x$>R>}-M zigD#U_@jz%?F50oTQqHDU)(>l07h=6QbNPRAawZYV~puPE@Ytb5+0 z%$O(INe@OKY5!^i?RUkT-E<*UAVk8_LTU-KVa^@ccN7S5&puH&u^Au3hQP6~csx2Q zd`ym-o)P*3)8e9`2EId%yOPh%tnzt^X{SgV6}wKcScv-?+|8(~f_mtxAay>Z3NM4U zX`6C$w2@LBV*O|4A+vM=#8wyJaV&6IfYsd~aY!Hc%;oC2TK7`XVYzx!Kri`hZY`8e zWzqY9C*U-+K{j#@2O11OoG>r2OdJQC)_scjALVk_aMddn>3;3f1L5i%0!^8~AqJ|^*j#Gr#*f?p0y`mu$o8b-3@H- zJO6){e92C`ELq`3#)Nnm7d83r&KmH*6vIRr8FN_J09ErFCg>eIsP4~Rb>Fgv%S!Rn zBVMmcvey z$ZC?Dw@&V7udM0$j%mUAF#o_+a+uw2$ayU#a2t$_G`3O9DT_tW0Ggg-z;JL9I$Ge$S(ku_-mOi@Ov`Nh`z{x~-t& zqSF)?{EI|bql|apzh18zuNH_RB7qbB!<1=@Y`@19j`4kU2+|&8LoByd*&?gs+z~W$ z52$v9v$BVF|Asu}`^4tU6ZXSbl3p`G>wR5J9Vv9;jQbHYuvAe@1x3oJ*erU5FJ`R@ z#AV)Rzp?czkRdOmv7rxrtQEdFssqyz5f6ZpB8gF-B^`=V&pmQAy4WLKCUuHiyQoiI zAcTHGOn}|xSJUgJ^oZ&>&2Cyl`Uy(Me3!|d?poRU@ErmOTR;3M*Re90lUCRCT>Y8} zKKH-=jsGO|>@1V>+FWwc3~VPU<^)BKL2WYc^Uy3hD>zND**D%z$D9#;B7NX~E#i@P z9+dyI#T-!`m#h_?^ueT5Mu4#=wT{y>RZD-$0fu+w=@6`1#yitFbsXeSK=Qb@m_mB9 zZ!aV;tLbj}3Ga&%?03`)aHOh;T)MzF5BTQ~aez$<8X1vUsuq&@Ww49D-v{qtWsQvU z?+^$+VV9nQH7lpKY}*uVDfekbC-aJ|G<@BZb9|GQ1Gbym=j&udMY;tw^wl^#V?5!xj=g`MUYxrk#q3FOU8jdAVDKGsUNNH`ZI&j?dk3RAX0*V>S zE8kO?7N3o&L7$QhPVBqvHCs;dC}ulFG*oQ1N2UBjq#hlV_rofH{p4BnL&;u`JMR6` zc-Z9i()o&9I*wXO70e&t>O~s9mTvHFi;3r4kerzJiGxGLHkq|2Oupg3=+Lv%QC0}1 z_WAc`3QRzme`e!BvdW2pvcn8085FaTBI_wogG9wT6n5>T*NZFW==j)(Il%1^X{IDl z_zwx)QiWLUG(rDk3yul;wVopOFj}@CZQ?m4o6q&hJ;_LzYqn_|E(+Dye( zsd`B~uUb?eq{iYoY&JpPbFa_M*OP${urg2wrCLl*c#rR zHbsxm*@mIf-4!_ccpQ7bh~2opay%^QPK*>QEa@j;?iWKBH}qgr?`@{GOzq>In7_ln zO0@CiCA_wn#vr|@k#lE8E#2u+Go@2d&E4t+MJ-S|s^RHH`y>ZceZULVEV?X86{ZK( zc{s?W87E#G2ARq4Dhnhgmqh-zueOpkPVAEGGFw`*C}s;q(v3@tv3Os7WGaSVtLa|B zqCB9Ut>z-d1GGaS6Bm~2;(=~jE9?yzzxMbuzq(_b^Jk7b-V3a+Gov}Ji*H~1fu%#j zrjqvgb5)SsI!A@&FltN)VTm}hfvop+kZpL%eA)=1r;Om3Q+__)-Ik29PkO0NoD8(W z8@q>lPJ{{@%VaGyD6&-3%`{ZpApd={;6T(85+_XcFZD@cs)5{mqR?O?2EOz)95dxl zjB4GC{NdZ*I%e6LZ>0miQiV-}&>Vy6NHsy&CsWJQ&QfdX3RR+a8^|T8wbO5gYe0E& zjqJK36$nQBeXO6j10A{%zZSb_gY(+mj^9oFV94YG|LD@&lgPu-G~S)qctCgl2sVdA zidjjK6;$kU9>{*l4J5IDTJ_dMnTL@8f!EK*uEh-#^wzqyxO?dq8Bb!@iGgAzv5Q@P z#xfHOksajL2O3#?fge^}8Mu5J9tEynS*J9alfYM_KhhW3C)d)c-1G7dCH6H!8{mxU zAn5k!fYOoeB&bC18eYEGp* zk4CNBWN|(e^EQVZa_gVa`bky@U^|WDxZb+KajDfL{IhA#oA6%MUGlLjxo&3C&ru9; z>Yf3UP7Qq}o~oFWEIy*jfk0Cy5X7as;p5KW0DHF>Bx&52&pJfkRwOYsL7SLvXfTa; z#rjA1=7O@xcF@#DjwdV{XyWN0ZhjditCfTF9O19IOeRsP0&&L@|C#7$Rbf+}T)qL5oC)%t9k3Bgx+DG(Nc25LUE7#e?K zyahIem|aAT@6^a@`hr9YdSTmCV{lTQF{!qo=CLJv+KXp=4%^eZH>9tLO{mEhY`8>{ zoj5jIXoiDKirGw&O^_~@)bVc0a>Isywt~WY8#Ea#lf?-Wc{$;AywX=%pn3{}wtFS5 z7BX#YpP0?pv!8OChaT63>{h^cZ|tAXH37sWz-0+p?n#6(l%)xbYV;oXX=-8W8>_D8}a2m6Y%u`;0GD3n4^rkJA7-P zEzfES&7vmWvUiLxBA^1rD{EW-7`y^ zD}k-a$5YnLK|(QgDeP|$dRYia6nm>t3o|DY#2m= zV{A&)faWr*3fu_=l`Uv8UBe1glmGG^nHSGwu4bkG;4oR?#O5l;%v^1tm~|9Mp<;K1 zXRF$!t3A{xDQ6_LQ=`LznmEQz&qP;|gQme*@!It5b6pcP?g zBjWh#8m~=^dO)$_r3nQP%(TrI8aFD~4bXifEY~s-^n&b#v}VR7onl~ZT}Q>96V>^& zOS8o4htWm+>z<9%3`DZTERqte2RxvrP ze`2L#35f`h0@a`hfVyhPP^z1;*YDIEHCD}>2>L>S{B8$87i6Mr0?YZHu&ofg_2k6z zk8a199F5Mf=^3Qbi5-nk&1}I@iaAV?Mk@B<{KsT}SVf?x+odBKc(4T!IkwLUo1s}p zJL{HPP0)TxYGfNQVx#783hxXDD^lG4bV6CY=@6&7Qg+_B%q3Kmq#JM^2sG#Q!Sw(syJ zmz~(iEH>j=@1vMo6zQR2QQQ@k&TrF8;8mZ%q?C;$*s!;mb(3^)PZJ276u{0F+rVzf zkc}ItMp1kO2rk6sQR*zZfLkC2VON1Vk*^8c#a|V3@{Ke_HND_I+2^xB-KtDt(iHg| zkVjfjD(d6j$C}ZTZyZ-@!Zrg54^pe2e50IOGz$`mSnh;F9|%u_e$|EeYxM#3opx3# zNg(Is6+X9tU2kD7^ar4ZUq56eyMdtx2U9O0xm>(feJ^rpum%Rj8eq&$6}u&ng#}`# zrB-K!x=GW1_wAQ{V?y1VyBepFYtI=Ea)KEW2Poz~Mebk)W;cCOR3NU6D&`c<14=`D zfyD7`%GIEzjzaOdLTCrUuuKBMPWWKM>bQaVGb-$Xz!|{w`^(m?%U_})|?el;~HFvveyF_gWM8b=H(yqu< zw7}m`!JLJ1AEX#URX6o>2*|&j`k>++6Zqb#-12qunG>75elx%78pU)|q=SmZm|hnb z^9?=pqDAK=gYp*H@}SaiJydSCDo=`%IX$9Bp?5=XfIjx3MY*c2UMLh&=8eY|E!rnv z2CU0T48F@$E?Q*Ntj!8O5j3DsZ=jZaA4|r%!<#^$7Nr8)8T=6B>%Am90mGn64<(|( zU5Z7E7Nsdre!XnYb>0c6p2raYO{z1#a0*Al;L-z-3WT>Uf>NIybej@89F_>>c?&vubH|4z4gc{Lp(Qir0XkL>OV`I13 zVy2-ODCf+eVo}%#rDS&nmIV$dG6Wwhj|UzlE1{$MoG5!{4~@sKzE%u=Fb0-y^MILf zV#+(m0tCCy`X~chUxIRf>$&lFCWs`gOrAq}*cEy@Z_^f8ZRT=6qL=}S+y{{@<(BTLf1)FfWi1EH_Y5E((yOUEQI<_ zt&iSgpZpfBZV@!gFG_9&cM2NC*)vlkAAp;_9dw_PIct2oBFa?@QTMn$5Ia?t2PB1M z1!pSJ!-nMjB~eRAGbvT|(9N<#v~k#?;N_}?wbITRYHW46BHK0fAomt&7U0=Jj+VX$ zd&egC^C~?JMf~8G!(PCxV$Me-1)RBD)dS$004u31;-2L+3-lff^>kmT`n<|mV1mEC zmUddNds1J)g83%=MJ)P6PjpV~TLbU?h+2-1DF#Sa4^gptftHI~Gc24)yLiCXCuuN2`Gn7&%x`kCo4CAdIc^m@q4nX7}J#=BPew8=M5I}&s+@(}&UMVsHe zG;b-{!0iSmh<0FZy+V4ykbUB_)*}yon&127oy&N5@9M8K8}1!T+sc$>6X!>< zP0u=H-FA&q!9npG3!5zUEnbl9nH9}t+(Dh1vmv@DFjLv9h!0TjcW;xg4atx~$;Pli z+#zvMpmq*Y{7#fpJq9#A3kY^OKl3jOfB$`xz1jco{y!y0*{yd@>?~b1vrg?4(?*d~ zP`Y7=faeRVxyACb;QQnd9q+oFI`notwVcvWiSy&Bc<-&8&Av-tT2B4Jcwjko`4{oj z`|uurE~mOa&_MRH(TGze-p2_ z!FY`UT~C-I_6ME~z>@15KIfaz^(Xb||0M~dIowWMt(<3up-mK%Mv=9~=n*;|mw`6^ zM3BzCK%6o40#Lv235K||@qar!szVCHj>sp5V{uFXc=r5_`wHue)xRW6|MfRa&}jO_ z+0RJ>JHz9|pwXLwrj=q^C~_PWQJKxrhiH@}*3t*Lr>7OYbn4AC1!}cr1bjFcfl@&C zqZ0W#D0&45VMb$6zTYiJg_)T&MGEPFB1Sxof?xNefOdYsBZY?y^{3v%gh?m8FfO76 z!j(p~q5Dw{5E#q|_{(4fBwF;KrV5fk;lOGCiS!d`pQKNsiCh;|Z06LTr~d`jkUCsv7h57(*=cMy&0Tm%n1e3i9kHYwHzkkYVeX_PB0M@sR^66Ecp( z{&gKGaV5=WkF=g*YA8|#@f6`IX$5tTzjXGOuUet&4LjeCc-`{q6r2%h`GYPy!s~fj zep?J|UfgrSbnbD}jb|?@;^9k-Vg34`wipd3IXV?+9rS_@d9gyn9DgNSd`CLyvc|hu zk*=(T%@E%8g?EEJlo{DPeTy&GQ{ImM`Lp^GI}faLuOgqt+a~#y5{-b9?CW=mEtodY>HeTgbT>nZ%%0cUuf*{tenW z5x3oY=})0utD4?T=lPz5NR1=y%(#F(e*9V;!8))zesT71Z>u3&hy^1>Q8Wi)q3J=u&9NsN*f3o*@`;Pavg{$P20N=A$SR z#`lbl4_1-nsutiCnaTwTw$f>gnA~x}>?s!1;?#&wwC&~pZEz{J6oqrz>a&l`;8962 z`zf-IibY8-*eL-9uJF{!;$FH%hWT-HUQ0OVIV)!0d_CL!fFDMUlNe1HRywZtRcFuC z3oxCI0)s1NpA_M;n(MUxt zkB+&1^;vUad7eFHEJtw6d+V6#c-!R@U(-@mIP~K(qG7jGIkCf2ZDxy#CD5+}SDd2&XBWbyP~uf5RI=MeXYJ(w((8hGmnI8AMQr*P|xlXN_< zGrTPZ%LB8?#V{TJ7^t#cQlu&FL{@`3+dVQU&rvO&+5tphm2O9PA3=mOMOX%2bGE7$ z8rswNu>Y3#$jZ4#cpJFoa{EBU9+Q( z)bo~r7KUDM%Dc;bW%QuSE&gH7K8cRMT(w-48o7IFeAw03CJ@oKA$fQ)aX5e;>WqdLTL>?D6h2Ipjg;IZ>M8rc4*I%d<&1;6C7f zMb_+Qya`+P&WR3y@J*UxV20jfFYH+wIX$6j@^c}TtzioUV;6N!cEI7VsGN?_Sb z2)iGZrZ_K62pe>13Tz{Iqq0x^q<5J-cS7`wcjJni%dbLM&;3_CmQTNIY1g!Tp)~{rVJN#Eh^~xSXH%uN~ zLAqr+elfXX$L?S|W9LZ_jPVF7>^SfGMIwY;xd}U2(|`RQNqf$+2&HB~&!rgXH_f7A zYXv2s>DA8Ub8m~9pduNgFNKjSqw=BQsbk&;E%UE=^#ouMdpY+`wza^(fKJaOlD|9#D@FuQuIg9n?Gg-^iP3!pXAtrzMiC7iNc~GegiGiUIo3d@2@I z=mwAVhyZ|%`)sEZHviO6-FA5 zZ1%lNf2=HoE{II!7QaIQrGd?MG>Ri9b39lwEwLkqv)k)G{Oe!dwcIT^ZJNN!HqJcOU>(KCDVXhsdF}Ve|>GZI(yB*O5o9TZ~a$sL9zsiM(scW^RQByfUzudy+P= zn<1s%fV(E_@9@}d<05k81GY9cy9e*IB#c|bXE2s(q)uxktmun&L=5@&(gWYWAo zMd`e|atx#m{Je8sDw7b@1l_USs-nP4P@mb$+aO&Uyq|MNnh;b0`?)L{eegAW8;KmP zLE%})ZL=MtZ?~Q`_IS>-f-v<-VfQ}E0$!&@ij`vB4ntQE%I0MHHt;&295PFM#p{eO zR>F0lm|XhKj&D5gd>VXc$9vz^d@o zinVTesuMyjzmC%%`pCa%ZlV9$H`Yz5=hg6T0h8whErzTBVlW|&1;+5`tN=pQtC~LY zG3{O+F-KRDOm@2$=Z*b9jp)cg&u)r=`h*?ON~t*Rspn#_&(z*%Y~@cN>TLHuV71Ab zg4MX{TO{AV=A|T$x~V#Tr$9?%j(Nz%z&HZJV0fm?J5!1U;!lt84NnW(pX}i^$}+|F zLv6VJX!CdPT88zVcCoP%!cPb~AO2DB8sA=8^GZIqfUDy-3p%~?xhRn_5Zy273AiK8 z78g#GCBdi8Zn|xlCp%rj&YJLjy358>*zd&AK&yrQn2JE%k4{>>Gq6yKYx+8j+4Qu2|}_W zq${tb6Zw7IcEy0Zk-=?(+!dT#HsEkfLDJ|5`^E2O$t_tM*?_}|oe(P=k8#3k(HdXi z^Ay#}w}+u*K`oTg8r4cC5D>=cuoV`gPoCqBzOz8T^?8qz*7Hy+vv8HIVHNndDM<5Pn}0&^aT~(`Y{S!&O5r8t7#+T$iItQ(RIU02--< z37}lN)hlg^+W2B&pFCB#T(Wd}lBXuJ7P#;5f8z(Ht=43e?aI>*S%Z(w$IH&A{yf}- z9Fg0+?WD+wYhjO@VWE~{AQg9jij4NL^VWKO|}lxiv?f8u|H4?TTZ{6Cr~xjna!yrgBrXT2V9eT!^~Nw{r## zlBrxe>r*I=DR$4I_qesupVGRST_HDK-Vg0PcG%!L2o&Qknr3BuCjD*x|6HJkgC!o1 zj*~$br%fM_EoKl(rkK?fNuXjYRoIM^#?>j0_$5s>Fd>ZpJ8$!s@bQH<0pju16Ze80 zAb!%6{$DO8KooxQ-yf1VC$=e>W`Ibcm?VmWD6uF`gg+5V+9H71IfEFmWf<8dl@I*=vv7~mx_>s8I(|v-LNz=M^zyFdsq~k zcDH{I{A}E`yZ(8X(n39o4&Yo?9|%ZO)cK?b6?0Ze6QHs8x@RfW_Y_hSJL9&P3X?NA zp)=0PR80Q5D&@B!rqyF*T5l2A#cnU-#I^HB&DI(n#T=l3)gZQDcG2uA!H@!bXpm~~ zwC9}}`Kn#a5cF!r^HEl^+3kpY7yn8`1CS2l`;DOeIKVvwq9^(On4f8sZj-jfY$u;8 zikLG~jU%9Qqm@Lk_=g0z*>csx) z9y1$~O)=oWZl+@U6iaybUekicNt0(bc_`T$eBOVxN4DyUtTOQIE1>ilteepmlh0`) zT278%0XLq5M;rnQi#Z$qA0o=n-@C$M?x7?P4M(Yu1q{xKcnDbFZVOcb?fP`M*v@1|DCf^^`5mw67 zS7iB|yfEZD>rgzPlfpKK)p2&vS3pj!oZAmcw34vKpkAVGonG#r6my%0HKl#SV8U0fWdGW4GgvB6W)lW8tB)IDMN8h|YPY>!;cxHQ8wLly2 z;42BPD|pC-jzf&y8DR}3c5Bg?F~tf@&bx1o4c6yc#&fNNO@{*GqILX3UUAU}=sS_1 zXO=HHVh?Ue)>B{;Bv_9+{u@|9!g=k>qc6U#`+><>(NF{5A;+C~%k;UK`*EIPK>VSN zimmf`{m*;toA@kEkya#QKhb7Oi&rG7#iLHCp88h>OsR z8o}GX?6pK#5|BdFXNg`=C8(ymnD&rTc@N#{t@Adx@dcjOdHn{O$@Y-FegmG!+i&cH zzNKgotIaalDrng^kl`X<&sHy_5JXlmy+;dIT_0FY=Sv=iVk7@Et)xxJXJ-Iyy82n) z4R*-y9z4_by!ZK(c=ZzvM=!V@jD5v7!*_V3Fc|?k^l1prlt$&bwef)^Lw<-puQ(gD zp1aR8L0A+3@jYn#(aqc$_2&=%cv7@s)-&h(SPU;fp8mCo4ou{&8{1>q67w^nU4-KafOr zJ89>Q5di()$dJ<(ib+SeF}78So_j8R186z0%vz1wp8%#7&uRhi<|B#sdTxtn4Z*Gs z8`coJOs!2|GGAKk^106%_goMG%=x^fVs+Q>Ic6V$@JKSEcHOJz$o&8Y;81d}~CzMXgFl zf@Tar83~(b&G(qklYb68YPqslG2mE!Rv-k*-(s(GeD$q}6ktzs;L2_FPqOJzTD{wG zC)gQqr!61L?ba=^WWQlk#^Sttd?0%oX*wSB;_UL&xi|S-mVYYB64%dOutcbfx-VbQ z7L($Z=-D1p5?m~AmNjs1(Wn8D%sJ)O7E>SC$A#1*1bz;K+KXeq1UA8aq^RvJ}G5){NEg+j*NtxLL-a#?YJGBk7MY&!m^R`mfE-jaW{(7bgMdPtR`MT$7 zByn5q(Wpx1U%2tA)Pl8M64vF5z|Hdb}2*wv&qN7`y`uz+q_k^`}L}T122E_lKPV3f^@y_ zZr*)vq9-uH&B;{mlkDd0{RgItwGL=?`G5SvlCaNdx42fA1arbQOf7l;o=6w+8KaT> z-uS*X=H^@5-@X5i=9lX7x3|7M`1YxoGzC__3}2T=Cy?`EUC3I}9)caE)xbF17NZL} zNE@%XKdmihEy;?^CY^$2kv32xv703vr{Vi@sWH8>mnF20lTj#?sL5A)gU8c+?8KJG zO7pR?pR7Zn34_HGWmSqf?aUJ4fpFN`7%MV5C$#&R6^`r$5*x<3{GG%N&wJ}nO;8D^ zIwVc;g}jfO8rcjfY(sn39*+zUUsgq}=NJ{+vBIOx_Y)um{=`@B ztK*;YJ>%QpUFX{^KjE$JP@M4{Z zoN8DWP;Byi)Sw(ENXN7)?G!t8z#KjO;oEqY)f_o*F!RlYy1z^^fk(By{14=j6Zeg6 zHbZq1#jK)8JStck!mY-Ty*=!LjSw)u?AbsyzkZ_Lu>C5UHoBwAS1D-O*37n>kIH5ypN$d+?$)33i&{s?q|%$~y-gP4rza+z zA{p!=^-dga16jEdk@^CP(NZLbiakUl4Ry0%!^|s0$KNos!|e)eck)&DNVepvws(<3RaL^=r?zNXJ zW3yI*n*F3w`5DuvsDxzMu9sm`aLTVhoHSJp*$)wxqhi^Tx=g-)W*bMnA+iYcATw1p zpssKNXzjH$RAUI#|9}tJ`uB8{eW&wo8(;NF?y<3wzZ=9ojn({piX^683i6<+QjLW` zTch$|dx?%`cW5$j#OM2?D^LXPERx|Fww=a1`$@%|F76T%7maij7o#em%S55tF7s1t6B`jLaoc=pgZC1GVY;@M zbY-5Z%N>g)xAz#;EAOGp=PFtSxTgp9%V?xGz zAKg1lmO61&ezw^Tcpb&0P$UT>Y6U`LIIWN>q;^mJY%aRm+F824LmP>WaD5U_4nD#T zuB(a+2%Bs5oebihUGQLP3Q=>?Z1i508=t-hg7W@8{pwzj^SjZExxY zm?%!=odCh(d`=tZ+O(2sUwFNrLB-E+p7(nlzhv4YulKL~EP4L#57Aqu=Bd(wJmaF5 zo*2i`3)+=q5V82?SILtzsa)8E0NV8`V(d2eZD z^EXYnnB4cv&q*V@sI2pD+{tA#6rG}&lN32Y#p*+Q=q~P7FX;6`g3O>YV?{b01yWCkbb=+?8rUcA0nXV@x-$SK zjmG8-?8xW>GN*^H7l^UjuvD%Wbke(kETm@ws#L~^1=}Zej8|G;shIEjYJE-Cr*P=U zWklmjs?EN(MHI7(f=-!OWHz`Lko0nQXp0PksX)}014_wQB@!p>Qxu4+g7Dcd4JZ(o z%2OlzA|b4r5Tu_{&+VEG>=G%QW_lO5Cw#~yb(${x-fMafM=cs-oCV8o-WZ>?_1mn^ z!gCezKfdRP-)s|PvK0Sc0omtDJ~4w#BgNEHq=t%J&Rfny)4nUb&If2Ru-zAuZQPS0 zD7`k;4wcfnkX9wq^B^_Cc5Wv`z0}AQJmhj%so^b~79WPzyo7U}b4#F`q2sUO=w=ML z;I?MSWw_%PG!EbcQh0dBA-c#5tGJiQRs+CT*zqUl%}zM7#FgRwlb>e$5mq>I-UNTD z_b)%OlnrCk{o%w*7f9HSaGx_N2I^flVd+AeqH21YaHVHi@Mm+|AZDKwwK-ayrubB` z8&u)-g1Cq?Kr)!A%!g9DB|@!gT=2CSP-DQk&0~+}!o1$F080<%1p$Z^-zHa84U}^` z=oZ<4$3;kpehf&+R3Z{!Ekzx8U@nvWKlZ)^psDQK-z(fpaxi2gklX<`BFN&1U>H#w zb*7!^GHu_qeeZR;&3mo1ul9AirZfMUc8dGTrr-i&02LyL%BF}K1VvC86&z3$HxR^S za1<07RQP{i5|l)uxscF|^J+(K&b{XZbAI{GcfRv|zn_sVK}W9#rw3_4@G5DP!=W+= zlnmL@*@B+Pw6JfzX4rI)T@y|Z-x4xA~E8iXPvrUt%K;#BY%AV&y~wuxd3vo*cxzYeO{w2pfWw9-Zg5> z?@ZYGFmBn)PD6Rvhfeu!3Cs#-WhvpqZ>m=U|kEg9~@caa!2Ns*$=56SRBM-3=(1;~Mhn*)sw zxO9^giK@21JAqds8>G=f&GoPvCRv&&J```+bzt@4szk*e2ar}qYO z8uG>WDR;mr=hZ2>M00dP8|>y2cv`<6d4sfCls_GI?M;Go)%D3eat#iF?&D>jp07WA zrCo*$FD)}JjMCs=m|Fh9duiX&-0-NIe#PhRRQL^kJ*)tG9sv8@x9-YS&TV`DWfSy@ zgWq3Dw!JWB?Wmbq+efjGy|;^s+W+b$&>UGfkS$zSQ7$x8ro!wo?R{CIMP;bNSgWiRZGt7p9#hkG@ni^~H|=m}2!X9CIO zh92kLHIQ;M^z5cs$V$xZD~LD*9C}C|5Is6Yu~JbaGdN9Hu+&6139z0a+4}@onmZxK z1-&kq9EOR+#?9nb&r`5cq9%B@umyvF5HZq$%{va;$UOHvE+Ii1Cd!V$vEkipe`LXl zBhxHpzOATfWAuNE>ZaGC8ErD$v_r}rYBWR(b0{piHkMo4>x*W~zkt(cYRiCjdd9ku zAHbP#-fx~4)Be6?sL*N4X0609*RdzXH$!UJzMn~od-4uNz542`3UURQ;Jaxc*`ynp z%4t<>ZFmK*oL8@2$XrtF3p^X%BK>HiPt)A0S?Q7z0o)|>sf|9h;kO+USLTG}eh!7% zrJeBGe;7_!rqZt~f9`Esh8kWz^Cyz%#AlvxG~O?0vxQE7`rB+?1uB6X^Uy3K^J3KB;YL=%-`W}N6P3Yxd`!{%)^q*AwUvq;=>*4>o zYH3-le1VOpJQzsE1aSx>-*$uDh99!PYieYPeqE4{kUqXifQwU$pjCc#S_iKVNF*Vq zywfwq6Mkd?vk=rHppH;1vyg0pMZP4|{sPK_pBb~Su-$8jIL(anhFH&?eR-nDWLcJ{ zT>cG-abnA|&CIfFq}XJNtfQjR!rDZ=k~W#1mk+)ZRv}?YTLsxh=Ru`(ry>P3dv(F+ zouZ1OgDuHp8^9U3k9#`yf*WvmEdSZhUpL|7-MZu-kV7xb2B^^tO`lS1HARk7Q7Pg| zI6G@*4hHE#`nz!4-3h=KjntP$T~~~k)W}Lz1#n=Vp+c3T6JEO{diMpSKvWgjcZiO) zI2c&mENN!+?s0x!_+mI^`@|e6ob^Idg?N|bA-_=4D!_W0`Qwhk-UtOYwh1?}y5Jq& z^W89baV5oO)3J?W zK{X(Qippo!`W}V&S)sffVr%z_rjg+O2=h#0nO!r}7+9__jiFE8WUXg4E@!%cW)@4J zR(Tb8>*%w<#(>A2Tt7b>|5-!-m~DI`j*oHssyl7|aOmGwKeCjRctOsutSDpDP2UlY zwRV^y5-nURye7M;FhqpwnY!sY5#2mI@}|9Uc5)cEp=UEG0y^Aj5fsj1Y0@ABA}= zAdNA3P?j)i!h-A68jeO{7>tAEOxgws%U+Iu9g_VwQxLS_F7@cy(Gky>|32ulm_-a8B4Fcm(MV zCnVV3BseMV7NoJ6^ks4Cgd6hnW9E;!;{ocfkUN+mz5~0TSZ1?CCk4SYHqFDX0J9A` zw(VW)38?7*SmOWHaz(HfuK%A@aR+4JwT2tGg^(GdPl~#)vIT$I6$@78Z%`1iGL}7? z?~Oe1LrXF#E(?zH1`@5&Okrx3jyCPemkMtxh9Sjnpc%6T5d-yebX)kw{YxTD>&*{~ zwxyAR7iPVwFeU!^ z>pyx0*~zPuRg%rToQTSChwQ-LVGD|ETX@W#8ZyYfUUIK)D0wPQ?6p|QQ_21j5;8QA z=rQ*oGT0+;Qydm2OHa|5keSS2dpm#a=vMItCHgdYI$h4&N`J;*?N%*& zGc!M{DK?HGtHAt#TcTC9)0o->9BcZq$bHxro@?jCB(qQ1$_pDoVWqd@nyY?GP5iQ# zSZ{|c6sNgqMy+b@uZlpghqZa z-2Lju+IdxYYI8`X|&?LA+*U3t0%!;X#-G=l79lZ)XlFn~klHBt+8j{6}C(Yv?kaHpD zgI4+K-Hp1M*uGSFfV9yoMx`41)biR!w=un5b#e|x4 zzb$LY4ku1U`NRwdB^0}lB73N)v!u)8Q&@RQAaS5>PNFzPSS@dcqIsa8b6q?B#H2eO z8Ny~}nQM>BQdcBo$dRrZRWNRuYg*Vz-^&smeU5i<0@|M(={CA&>P|T4OHyFJQo2Nh zK5m&b0a5~z19Uv&6Lvy}ogmQ6;Y~Jd${& ze4Qf|ZLNP8wtVo`dpqt9rw`tF_W-~C)dxkEOx;!}uA2mzyz8XQs~Prymjo>k53g6B zC(Vk2S??B2IvND35bEvxO!y)vDdHnBI9@8q_o)P>4tzLIzJK~jU#y4Ng>lJ3O7IA8{h zJc@abU5b6D7EN;YV1ow#!3exf6OuAJLRqDX~G$T=!L4ydNR*CIEZ>fr&0pr~# zO@iY}eFVG!#a{-=^9siS@g>p$!S4mkY4g>eK#c%~aDYz}1sj2t)mKvx0lY1~9lSdp z*41@5W5zm~26e|8w74x$qb^G)ylZl5_PpaMBR5^iax>zS2Nc^yk$dPnl!V5H8b77` zq--V>1{$ge&qHxWen31>1Ehte0%z!LpK8HX-Wg`2S1eSyK#)wa+uMt+#NP>uzi6=-b zQz_rUXi5c@bg~IRrSvrhb(Cr$_}amXVU5X0xHZ}aLDpI&9t3StrH|J!E%J4d%>*5o zYrL>))&)z1CB;IPLKzjcfOPZldw^B5Ce7D8-V*SW5poB{-X~65%|2C&^rlw@=olfL z)Zk(8-1)HjfK33h+S_qABzs?de<)eq&TBS_m8@>#+PoXi7`Tub=c2(2oH z)wIwZ25RpO1j_hPnXLtg!83rc=YBww0G~6|s3Kc<;ep7xsHqWU|(Bb4A+5FbeeDK4C-{$|?v@X5du+^7b;dYqq#QU=O zX3kv)#one!8x>U}!vHeYg=DC*{BKX)5q@~|67@DiaWib$(N#9Z5RH4YdS->x)QyM_&ig_#Pjpn2lW1+C1L22Rm)36Csk|0&dr1>H(hY8vV(`WKMh|2#1vShtyM1a zxGuoQ=gsT(M1hu@At_-{{tt^RMy*;Qz_p81%UkFiYU2}mA^W3_6GWeSxDSd4>3V^u z33b1`K<1JJC-#PRnt4O%6bs8#3Kg}2JPgMF>eV|$CxrFt733DE$f7HhpxEk<$JsTa zYQc5VB*lAi;K8io?GO#evNfn#@0%J2{^aE6P;nuP>%Qy>6HS8c;|s>OjlNCiGM$jh z-37|QTRm62)GN9=ZL9Q#LLUn0SrmA-(sWR+=$xD!@ZjgQv#yb2sx(NRK_bIzvStNR z8DgI^qC)=-Y)+V-y)0SrQWae!%~MtJ%lWmkOyMahzK5OAH7Ec*^=0qhU>{U*1JH(P z-}}2ItrQnW*@?X(DrWEB?Y|6Q8ysNH)qxrMd_@2au9FyKF415sDM20b>en;CLK3@`6y)B z*dK6Ud>(y}U#q^!ukbql-RvJfdVTqvoQUO-bs!8`Fe{GP5QLkahe6Bz_3r88yCT+j zwG-S@ckq^vItaCmNQarPSTv&|d=Z1AVTNL@dc*8b{$tf^@!xr`PCa-G*=SxKgM#D5 zpfP0KybN%dK?#nI?sYSPsq&T5kI14ICN{Il40?$an?RA+KJuT9%wpl4QCjbLcwTjg z*CK6@t_{l{c3E?qK{RMM46b9xy0$Bp^=n)lRwv%qSP>M(is&Roi73tw0&Q3Z4d<(> zb_jjN`PB~pQGm_3aCF&{1$=%x@)yfllezGC(0PumbKeX1!TrJP`uTicfdZ;hL*SCEVA`9gERHgF8!7E7_lfZ8VWh{K9^{Wlt6LLhk*cC~A z2ft`oe4GQmw0-WMT2cla&t8q(78d>Z9dcI_Qq(Pf(vbNt%)*jthN}dMjity+D$1BS zo;FHzASjd9O2@GIfdz2n<5;i^IJveVXuz%xZ5;E`j~0K~!>wAsYm(cda>;Jru5nQO zevVxnlpFYoGTA#zrS;2)BG+CQTsx8!*tXNVL!KMBWNM0E2EB|z-r8K=d8S&lCbX|z zbB)PicKgOMwO$G0X!m7QjjSXTvxe7%!mcP=yjQh7+#U$B4H8colxH{C0W6%1%7~7v zpZ>edWK@oQ=go^`wG%HtP}kaT`PoFVU^hOdq8b%lq=a{E>XNWZ8cuTL2O+QSwqVud zwd0Ga13{foG_^>kg@e9M;Pda2HhDr`CvY6Eb?~BNn99L{#11HU&JdhHLe-uN6Tf5v ziU05QbIE!q2FY$SkZh$`;G^6GQV$Bf_%r$K(R)IgK`%TyB+cipZ;#7rvLWcMZv`o! zSCei!HK1D1CephjryQKpPToJ}px``HKO1SCk+0GoX2a^l3=S$rb7FN+JKW#~m9m^h z+LG7w=_srdZ)U8pmllreP+;i*60;nbX4uYDk%NIXppMi6$#uEG!%QK@1ub>}fz3l$ zN6Xtjimg908Iih4=SGmT1L-X~Zz(3ZZ)RAoQ|wiWT%n@YtExn|NsC}(V4(yJN1Wdx zW~*m~+X8;)WV~0Sw03u=yi9tV9D%guOOgV{_$-cdeR84v5rpUxnP!D?+&Z<9=(2C@ zT6N2uIyjT=jL^HsGM@=r=G1ysfrwtI^pfO=e7^TVQW~_;=aQsDf{7?v*koab!;!$o zDVHScURmO*cVFy^@AstTS*DE()}HO&4C3xHkY^ohN}qd0mf4n0TbE!Ze=>2}7P=vH zw|o=3K$tCTgT()|uxHKEGd`%H-Lhr+hxQJG>ur7d2MEiKjO}Q9XS~V$Ec(;vRV0<$ z?%9bAlitiQ?W94h0ol1) z$Y(4ZcQ9zf1Y^AmzPbkqa_ls1%}o_}txylF-eb%P3>kqx})`WRix2l_$g4 z{mOG$@#0VuR+cBV=aTJCyhExon|(dS?xjc(6_xJMPFKr16s19(5%p@VU$qonlU%Pl zZ)6#35?t`D28xkGpy9B8jLyA@PVvK*QfZ4|pVuXb_wJJC@Js1dI2~*x)uC_?GXD2@ z?HX4PCyNzi_jF{ptP0M7{MTwhirtp7At7QQvtlP~a9W+HRh!oT!_wAR`I25xEbwgn zn(=RA=4uh|(WEnweYhTKR*fy^g7WB70s8Ujs&c*#auTnTb3rZAZeADBlJ`@^Y5cv@ z9QZ9A+LYMo^B-UsZd2m4(nF_Dy!(san{3UT?T1H`8%}I%mYJpVcT+6vqV7^rct*T| z)$_YOOGI;V*Rou3MKM>ayef%NAL2DLJ7+YohD|mI#Je>!M#^^dg(0UugFZk|Mk|+K zX>1FG{I;kHL3;>A_6|(@ILuI{xt%nV+w=w4uWO|Tra^jTTG)L>9Pjq{9=VRm76XS8 zolW$i*(5cf*9F7KRjhvO`haWj8ppzMdR%az!&~RBTdVDceS<@r|CkbH*vC z%GBU5s(1e^uxBb(k77biXF$wE4W`ndv^0uL8!KK6nR-L}Rt#jR?9m2EIh;U z@~rL-yJsCEZ+t#0(=zMQX?NOIGAxZXsu;-Alm=x@!1>oy%L_!^_LT^+Wp-=;fZiNLn zprbzzKScxaf{t#TQtQ>h(;OU&4-AXNuoWI=D}z08U}Y>$@EFk@*tA<=T0_<)1f3;o zT}h$YEM!tFBo(DmQ9zv`EP?g7MAW80H3d`;i|1b+pU3-@&Qn~PwA~Bu>)kQDnE|Rm z^O<8RY-wiZGY`Gu9GxRQsQDO%FEFS1aNfw)rLV^=pJW2c?8dBBB>RPlz#K8d`(BEL z-CiLTb(@@FP-gdFP)TUL?;_bX!z-;a*K3DJ7rfLxfd{<3z`lUzPDzRcAa=;0fr~*I z!q2!iq#gkf7D_>|1&4d3A zm{}>)3l4AmoJU{fx8-}7z*GE`9;>nWIpDCDcb)c-_3*aD&I z2zoLoN(|edD?T{&O*9tGLnOfCJqPap`%esaboB4-xa-mw?L$j+Sov6MzJ5va;T2eR zR>H+cR}}T?d}d!{IWIZj4!wYs^E%=04Bj5e9?4ZlMpeu{3(g;_S#QU^ctIAP>l1J1 z{lsKDgu*rdM$T}v9ZtOey=P`SS}C@fB8^m3LExIu9=Yb&6W>w;WoYeLUYNSp=*ufB$N!WX-JPeR6CI!}`z zUKEl_8{QVnUv~#*i?7JRqTvIb3{nvxJKD#7O%b{74bo^UTtl3{Fas@4Lv0)2dq<*X zW&PNMH0@2{BU0zY!MqML?6gp96GbjjQN|z_GC%G1)dizIIMSol$htk3dTEt$ibo*j zdC$Eg!a#`jS>VB-RbJcOk3$MxoFYz95K$n#!@v2`vWYOKjo`%07lMmpxV42OGchc3 z-K~gR5m`l6Kd`w=15lbk}3R1}UK1=$-NtFGBz?o?jzt zlV#A`B>MtydgTTKY1F(0!dS5mxTMa}Ma17iL=-S_({q`v^p)>loOK-7n=mV;67W++){3L4_qNQr;nNndTdJK@ zug+xm2UL(uwp&pUn9Ib5E_E%CM+=v^4sp9RI2769uULVHjm?};L|tl^o*W9Dv=h74 zR_LVD!s>w`6Jo_l3f!54OOm0gXSK>*^4s(kpDRA85XCY0%t#))NLE7+%f5V2XtEIz zLwW2+N5jqvOv}*hGwTkK72LL6P8_n?YPM)4Q|vm5BvMiFur49rVWX%LiWJ2V%?FoB;RX&aBM=hqr2#hRB&0XAP6a-+58k{tH+UFV zAoCF%^XjAd?m>Pg+mX}zpSww#6T7e_X4W8=Vz*NOf+(O(1}@0kQ{&yrNxWM-UE^IY zF>t|eQQ_9MM0CZ#T?rG8q%9hw9Evff+v}1akS1)Fr@`L2g}%zGqYFSTRI74`_K8g# zP&POpc1HpycVy%bx^|xVY86y(3)~TIP&>g=rY(kyCd_vk#?_^7%55_veNPYH&*k5_ zCmU@tBP}zhTp}slToC6yOeMu;MrRwvf=`k`MWHA)oN9cCv_W^J9TDedZVWu>dn9r> zznT{tn&#dre?Tg}9XsdYf5rbg=GE%(vu+D#Z=#dQ6`xx5$ARsjLh@b>G$3!#P{@Ct zImNc2 z)6>)V*t%PKl0S?_g4=WtdiFSX1S(gK2N62piJm@W>$;5@lQ>A(pXBR2}Po#RYv%pX;LG=DzjarRA6_2Jo(lS z4JuE@vG)}xsEka0`=?thPja|KNt`%@X64*xr#zmIa%>o1#ls%d#@gu9BS~qgyk(GGYp*rPCOACPE3vrAb>6K!&uzikbJ^t|mS=ia4#Rc_=R{P^ zTn5s+J0VqUvuD-JHD2{xB=(dt9zgt57AYTBY{-n63OPA2)NJhQ~b} zd%#&Rt6;Sj>u2ZtA|Bqsy~e2YiFCPbiYo+aJlx{XX*YXlktb0<#A;53loDc zGqZdJ6bqD=+o`A|MLKyX(Yxn~Kk-c$LrxNDmd8SE`6^JD${hvN{Qf6|3H*l=t+G*3 z&lE-$iR;x!&H;j~u6v+1v5{N~tyNb=&HwJBpJ=~Pr(Wh-N$=!Uj636iRg|MqwD3_L zzqC7!g5J6O{h=rVIj@~0Ruq9+!#4&V9BU|1#YJ_X^pIZbA>XA2_CItc{hLMV*^Y^CUsd|sg zP5v5#Ks#2v;Ge$Pi4JKQyW2M=3@T+R0%JkM6u+yeyn~4v4t8dTFdg>%^q(4w({wm* zNb+arZkAalkT`AN1h}&M0r)Y;{^?LuYPslN<7>)N#XXjtt-@3DKVXAxU>&flLmN4P zb%f-5>Hn0NY=``xuho(`Cth?Q?W`XrQYrRhifo{wP+d@y2YDfBe4vlyLjVlQskACh zD%4A5sH*91k4uuGz;?O>7y+Sb3#%?ZOtJsr-BYt_8$zC1R-PKi-VK}(GV;*o@rS-) zLdgBs-}rBG(1{Up*$g44DE1^pKEY5;44VZt5Ien(hhViR&~o*WT`ix{BflF_?Um-% zB~2SO5Ba&Y%Htu6*mRE$sP+RE1Fv|#=5tXpq!J`4?#WL`;{3kwEd@|jhk|nw9g;06 z=eNn0g3pplXVMzXe$5Za4~Qr6AYq4RIaoysdfbyvsycYvg!w*;CoTf5?k3N+aL|rV zmUc-wt*V0r)l(?5seuFXxPhu+lxU76rwW&29p?=oTBC;9_g%10OsW{jUVQQ(++P}$ z&YLg%+;=remX^{bqIz`~DL1sX^OrGO{2%$>_Ng||$y5ufZzz?zndWo0}^?0z=pB+r50N6etj}f zeI+UCm3cl8W#;So4}DvJ(6-m573lu371OJ+Wp_xYLUUP?L+gScdK@6fBG-&AoZ92k z?b$9X8Hbh%efX2UXxU2n8l8A2^vAa_H5?AY!k>L^hj;~|HunrbBJCap5s7|9Gp_ku2nMC1Pn3y%%}PTu3)ZB!$PJ8-Vd);-bOLp8)^lg@ zd7q^9`$L<1>*Egf4yQG1#PYZ6&RKE?a#^#SIAUjoX)qka-N(vQqEOM9RhN8;@ zxn_6rz;OY#+GdCY>1S#o^}2``%Utz1NXqyL>{?!#3K+ivz}f^32QX|yg2j5^H~>5z z*}5?ihyJ0}k}dlMF&tJ%zj3dXq`1ah1J)7)MGjKDHPH)%sp2+znL7~m?Dj==oHkjk zceF6uU4uoVNRML7rd%;ivq0D`fQ>w`AaL3pJ)Vu{T?`6E!vLzXQIG!ovdNwl2fx3R zY;)q>_E9t2v5#W+P-GVsg$6UMJ?6~|KeUI{u?HS3;9(b94C_{9047_idsXHQ%NmZS<;cj_72~peE5_BT zZ~U}i_R%+%zgnk8K4hRr1RBy7!TRZUB0AuZOBdW3k?g%mwIQflu$fma_yF@Wv3cp3 zl+njPYsCIg;{d0pOiaH9y94a6J)W^MOrY^!xN;lWo8sc7A- zd-6;=o@oAgMe+A*1dpRN$K=Q4(bRn5DTw3tx-1YT%h1)pTie6C$hN>@&HRu;#<(1IULS)o#0{>Wp=i8}tJ*&Z}Nhj%%pM?m@ z7xHXzJb#d-?REl3Kb#D7$sIe^ZOs4+(}*Vaoi{AmFr4=2(F(sshr;OA_P&z9@ACM} zLsJsCiA_@M@u-KYF+ICFEI0Bf(?si89UVVx<)@tC!F{F&)q@*6V%{FP$+9q+%MQtj zJsK-z%m)KO>RH$dXHA+cAr_gPo4K8*NoLRlfkLZTH%>?2iD;7zNb6=+Oha;h9Z+fP@*mQKbD9*6=-@Q=P#)%HWa=XC9&|GqnaWp6 zKO&2q*vLRuXupiAM2bzINGuh#+qYSPc`vKn^?XR;$C5Xq_1iWxgFfm#e0VU;nXRxG z$`lW#FE?0(x9&f#G{Hh$_0E2h?!>S-U$5KN5w>yG$=<*W8Qo3y`{c775o?$qH43rvlS0! zordAX;`2SLhmywR#799^(wHvNpMs1}ieKu4L%w(Tn^{T zk!tFu?+8cGK=f3&(FYt;)JHf=TIq%SyV4a<4s=&qEzr+eJi9BRWL`<@I$G-+89Z2$r$31=Fb;sm74lrGu(J{anj%d3655?atSuvb8PrwS#1#UY( z7p+tj$>ueN<_9d2RfZ4Sy!M|Ho94KGhhe(5?raQ@IWqF%v03LW3AvmW5muPOF8h>9 zGga3opPN}EeJHsxB|oqWM7h!+Il7ElE{o@X>en>9q6RiW!({x22^EI7;RFrpQvx)5EU(k z>kaJEfO_@eplD%EL^sda?dqf^K{05tpP~~YwaR?pkk#@rwJ3&C|1(vtv6YTReuW5BdC=;{X#g=v`~9CqK#*`+R!02eSnMK8 z_O_?bYMnTE4vC8WQb8Uwx*Y>vot?9;C^A&tGqQY&CbdZGrk|k`d3&cFd1-E=;sk-S z78SmX6|WlA6>(NEw_be!`pl=UDDdeP`C5-^K?geFMz@Kj1_|+=u!@A-XFX`EHGf${8c}u2dK=nkEXAP9_HIrfweQ2|y zM_$I{&&-*r54|gGo_c?3DKCMC*PqE7MSGOBp=(Fy@bvC|{WnXR83@O^I{;NSf#&ft zW9`EJ2%5X;)21(bG_3*h4i8LA8nuhWK>1aTEQdk1+0TG$wlrvISc7!Yj8(%{er1>; zE|YElDYZT#r=@RX)}Ma$NNw^%TEfO8lYLH{pnBHKL8_uy&{;W5MeXF3(zX0jQs}eF zZHa#+oV%^@zb1PGDXV8hOZ@BAM}_O9mn1r-QSpVmRlS@!Jx1bJ7ZCmwcUg ziMmIzUi!%JbUWQXGfifJ}OMWDZY9XTUKo`Y)}jYfKK)B&Z@CyfvdPxEl+=niZH|k1Ls0g{?@xW~9aj zai3}$xlpvqn8=hF1re*Dvd-y2`-{AW+D74J3Kn;cf@f|Rj+^oRrFnkxS5`T%-DJi4%;Rk^NP8VFl zOLNNy@|GN?*9B^z1KUKgjGnyOCq_CK{$H!EQ`f6Y=@lfFx$Tn;{!X=^P+smE%an~9yj6ORaTw?} z7hnw3-%)MN3-6txOs46P=c5y($Ayz^b7Ip3e8~M`II$GFk|N8H^}Qq%_zE;fK}Adh z+0eS+{}WN~hpz9Ae6#lXp*ts@b}ON~1HhfsA#V%pknfTo@ab0k-H?-nCB~0rH$0LB z$=+JOn6TDxttw|$H}ByqO>`Lp!6beTuUnxx9F#*J6rYhKsuESFy)x+oP$7CJ@CqC( zVrihJU9f#z(kQzo#&NM1&o=#dx8M8@@zT0+a4)ZDWK5Ri4@WzXkcHfw94AhWgd?JU zmShdZ##1DQio(?T9T9h>nStP{U_me{4&ru9s~py~1*gRB2(-Z6&8M`jVc@08h%Yy;Wq3OZ0jU*o*_GQux=Cv2rG|_zu=x_S;*(Kkg!s; z7w>n$y?Jucm=%mh@80cs8+?gW`m>k*1*ov?i*JWX<~Rg5OawI@`kN&s5|{mn6Z<1p z2$V8ltth3df!DK1i7E?C?pK(39TCOTwaP`KmxiUxNK$~ZdJwX(EgQCsd)vU&&)?t( ziorg<`Y)SV%U#OT=gCg&L0H|39GDjGR^z=Uv_Xp6J^upxfm8DQX&>yw4XvMb8&de{ zqAE-IQKwx)Hk-MRNff)9B5|n0q8HbO-=>=cb-vvm%hb0?&-fxn7ntQ+MPHvh+)#Kv z225Z3p-n00-qtZNfNw#4@XC>qZ)QM)=P1tOG%6T^rXdXDxDTqK%XBIKRrM4q&QAIckUf}0ZFH^ z-C{4Ejd*saXR-W|NHZkIv3ww+!v z>2r9h6VCq*GX<0IJ-cVD2|GP{1=&R3nSOn8lb~MRJ+7VJ8kxy8b;4#?^;=X1cgYP| z++ekI?1ekT1gm_PwBux{6T>RQ3|8wXHjyF;RMZ`ieL5dhrd-P__FO(g%gc#q5?qDc zqz|qr@_`hz8cv3W)#h+P2DiB#+SA;SQJVgXDWgotxb{C6{y~;IF*0;!$k;%!uw<^G zq8g+a$uNd6456-ls_o&s<%#2)=yNkQ`vvobM#|RVg9jIEfVCLfgmdW42^%B+UiTw^ zOLh_~v>y2&S+tF=gtM|}VTq_#c8=HXVZ7EV!72wG;%N>B!d5}BLVXan4Ygj&74iHH zs2RlB+prDl1GG(`wA({nXMf0#Fs(5^ zEZUYv3b;kqo%rw;Jl=kh^+Oa31onF1tM}E1-iwIyE9aMuOLMy?Lz#@N^u36qU)xI0 zAJZbhwmhH4Df9X3#CJUw2WypEW?mH6jyd(UyAdsNoiv?yhH0m_%-oHDK=)%$<2?Db{et(sK_2gO4-nf9`geONR795R|7 zy?(s=%f4kcsinvBkxKw|bC4Jqd0a7d3q7c$-4aP80^`N2d`|?P3OpnK9AfsZ@{_)f0>@CX;iyfKb!-e2 zkNe4p=xK-%k07m-hzPkKbtEIpQ+)7T%^qV=A0X$cV_}D?@)^ zT5=RC1HMg8x{@|C%YKPs&r{?q6_w(b@AH5Z(Cz#JI&bpk$%likPp((*0~t@^V)TA zoT#Z06pX_LEFL^PI`<_FW-fLH#7wjaf#&l**jLP-p@LYr+m@!JC(`!9E-$AjJYf{C+CRkgXxV;&aU{Sr#3FVcbqd25cyr1eg?9E^iL$Bxz8WP!kH0 zBZ+<}sdr%fy(t$$??hl8Ncwo#du&lfhg_LbNoV_)h?Y?;1`6lskmJgv3DrQ0hH+la zSx7xaf`YAvht{#l>M~}iQ^;*zY_lJRr?^jrSr6;;GJvn?yx-41-+t)&2NQ_?So6zZ z@|hEFnpT(<^gN{4FDTMUMI}o+6scm3L4--%Cad+@LU;2{(Rb(qVi4OY5S?d0iqx&d zvoqrJa1HV>p<4DaY4i*Ynw=(kZ@}V-8-jXWIuy;ywDBK9#@vDNF`mdlp6q>s06%22 z&w_v?MYW)a*TE|s2j?D>ZbMFO(x_gSMsi?OBPoGO8{~d^Kp8Ju8UjpA4@~%`)Tg}7idol9*c8{eKP4IIOtBOblxA3v^ezC|QGgG{a zVu9UfC(1%$DQuGfX-XFdmC^;_Ep#Io;RUjl5co}{H~MIjy-`}uSc0Z0Mvst26`%s( zg2@dwtc|w~F`$$5ZYUONrE9=)_qt%UExPkVHMr(hE{P)Kx9|Q@^VfHN|KET4on#5c zE~H2_2Q&NRQL}-E_#$a?OW8{%-1z@qKbNd`CA-Z&@U0XJyWmY!6iQgv`!15<7Zl1q z$x2o46bz``45^myf(>z#;E23LR7#hVS}zn*nh#&z`;oQk@;7(S`uxqyZ_aa)%adJfSwFRzCm_OkQ}gp z+@bYilx~k@qJ<|UCxvAUwr21)`jjy@r(71@oN^3=G3(X2##F!$?#c6fHc8|BcEboI zfh*W0voC>noJy|_PT&{Hw+1{xv&}er;+{D2>M?Y!<7|LkBQ@)-Z|?bOl`Vp&3+J9! zdGmz^a)blZG$?bLq{#Cr9NwCEm@%6%^mulM=i-%jE3Q0$p>pEWDJu&ViZU2-t}w;3 ziC*ccxyfG>+C&?-=1;!c!}2NHJo%0~%e3;Zv~EED=i2W^d72igUtS<{NrEfvH~Xg_ zq*E+Zk)}{ldE$8CpMX79mF=I#q)GmauZ?UJNeYbpJ-L-hSBlZXK9a2UD3myU4s8F} z&CIox56gbd)<5=v?6%V;alXH8<2NlSES>gX(+aodcA!bQ2+O{)q`g^z%9L=Njhu3O z+;qX1ij3#mTF=s;9IA>$hZOO;K)_Vb?~HIbeg_F(huaWN@ExfPyDYRMuom|hs56IX{W zo_I{^a0BufOtvH0un<1(=jaRTNPf#R*yqb$ZM2x*YT#UpQ?x0v#T}4BcT5$7!nMr` z{n!i;&o+|k3?B?^--HY=3^;PG?+944#FDs)ORAj{AM;qDbgGsw3_$9HWw3e0O$~S` z^}cr^@Z52i{5D+>f#-uUVQFE#F13u2PN9;*+dWoCf= z*2p>hxBvEKXYE1J^ENop-7Vk58kU;`VEJ4z&7NQK0YB-d8*RhH)BPOX!U+>2Qa-xt z7i{uXKlreDKiSId*w2ZVoC-59b`QnwqDVd!RqeJ}f}*Akpj4R}qR9zFtA2eMGkkDsh{r-Y~>))dPrT)7s=k&O2 zpmKiDHFxEl@;A@E-ZHytma#oG;6D5+$bXAjQ(68eB)i!Bkn_x?$jzQNREL=?WlzwV zXHpK?9X!@y_4J)*0F9HCqGo1%qsmgu%xMF1kigUrYLDf~vb~a}7^&|fn)U1sQStO5 zakHXhO1DS1S0$tAil`KqL36X>D({X**#u}pKI+5L%b);|C9!k9(o+*N1eiQEmOUFd z!DPgt5^wPclSNW(E_$Cla&g)iGaNAVhe-^@uAsrYSbchAeh1g52r-qCFv6Wf^_ie~D|Xcy1Bz0l{-h9W%F? zX%nRd8Ovi&j7|rR%5I*PUF`J;ie#GvD$FwV|rB~e>AiWwSiIB-s5z8!OutY01 zv^KPoHjMYscP{p7vqMcJXH;2$%*x9lAqqQNN0sdT@4aT}m~e^MIWe}ZxG`Hji>DX* zpPlnw@$8F|R(U3`Bcfhy$e4P)c2+E7yzUe|loZc;?}J$8{l154)i)qL=_{HeocjN! zZnrE?dqH4grEo0?33muzwiAVBYfvV|0#|Ap6@>?cDPdU?GTmy`O(c2@Y;hrF2Dh@8 zB{}pZ#lFBqDC%tzJe<)sy2s^;&nkY8yvOClq>Iu!BGmnHuqVR_4rc3$?Nd9LC#^3v zbJ=|lEMxp!Y={#_>a3)&V9JWI{=QLE6AIDZB5C8DxenxC00m+FdJcC;F?j$vRJFsRE&$|Cn-uq z`vMn_F2Rf^ED_sA=h2%zZ+aC;HMq4-4TM5nJgd%P*N!u#`9tNCr=y}QgEuyhEze+V z$#eT``EbdfXw&|u^9O!aWWfs){0D*QenGg^6dOm8RX`jhIt&NJdUrG}I=Yk0L!!-A z$oscjwEnptdfx-Kf}-#F;rjWLfZ&;yx|v)oh!eXi&rsQ1>e~BO^DkRxEpu%WoFRF< z9=VSBLf)oGQq+1abu9^9<-5$a$7QMO<}kdfSJyIqPh~+8%tAPoY^Up0JyT0WsN%cS z^{Crfk+H|B8C_GBxwfj8y7qjlZuZshlz!(ryvrh4CB4+OCH$z+_*6^y(lN_i2ft|9 zil)Z~h1XyI~M zwddd`c`&oU0T_cBe$ai^+)r5dpQg3v+)oM~klGh!?P)h#dzvY>ks=qVs05zYPwThA zcbnuu&>jAJeg8A)7vjBn^KvNsgVEj$)dL#E>-$@ z1Mxad7E>!b2%30(!l&=y2zU^wNprhKHn6~I)eX{jIL@spI#}ec@=z8pTa2Q3SLC^Y z#&<+zJcEEMlQve}b68AdBU$RcSM?AST;b^3_^UpFG6$G( zvpj=)`Y|l$24DZLuNLuclb~99Kk_r)F3ASS(nZL$(a8a+z?5~&=eAFaAoZ;SueQB* zZceu%87c~m+ws`Y-R}2EoZlU&Ufk`z#AB0mPl$eOtvZ3~nDy`*r@rxMcKfTlepUT% zxP3ny)J0-L?*|^@RnoQL^*{lDPi^|XJwkg9LeXP14EAF2TcajPO~y)eNAxa<=Vq*& z*eBXyW~|aEb|XcSk%z6FZeTZz&gW%>%@6pT?gsH@q%S%Hh*`_q?v?GoP>7Tf%?ixh z(<(0unuaIsdoU0g?6h;~SN+#bBP`1exi~sbTn=ER5D^n}k%25}a=jAv(@h}5`_LoJ z_c+-wq4)nP=1tf+u(;}fhoXw-#IUfUiihzQ^hgXfss}_nL`H2A2+bIeu#w3pmHq_s z?+WBy(qYBeInW)bV}4_*XoSt%#y9pFzz^>GngH3bArYQI=FEAER!qx z-NXN$OZK~xb7nr=Clq^>A{A8BJ$VOjfpo3McF9$bt)9L9mal8on7MjYz46WTZ)|(( z-0c0oZUM%0l-cSs^Fh@D>)|k~KC!()sEO>!EF0nK2iYrL$jL&#o9$EttV6?Cm z_($+PDo7?To0qQYojRY1^}XYf=W{A_K7R9vIWCCW_-Isp1x}cU`~&be42^L{s!Rpr7iGOJ^}fqtXdoO;7=37rj!sfs4@{Qn-FLHk_g)v=e_{Js#THez(D2x5Dq5Hm zh9ve|K@C(#!}+L0(@1o|NT|8W{R(~GvyEOmx@bl|#1QenXP+F(|OtZ91Fr%5!;|L=_Ke5M~9^Nfw~~CRZy#5`dap@ ze@v(5|5w+y>(pE6n<3@A`OL<#_39IWhl6y|DzceZJiShK+o4X^kl^!Vw(Wjp9X<`R zkH!pzT-SLQcWdOjXurBWt^jL8y&CzbnUdQo;$#uIYOWE+9|*5fo=U(2fKeSEeGT@ z+m@Xj_}aSt7-$23BdUNxjI_wd1iQz18kmtSR4kL zCP%edHzBEGUXQUcsMF<96S_s!NU|7%XhSJq7u-EclQXL=AdyLAE=QtDVu8mA4^4`q zeF8*hAQK>4Y|wmw^E{9SuaULMx)nR+@vxUEh-g-9f#&zJg*y6k(MrYnpk(h1NE85l zt1Scvs3QARH(?tM9h1Fx%5U;9a$n=tB>1p|E`Zvi_tGNPL?(^m8pY=_L;Y89?$_Mr z()tVyVC0?mdny^3FSe26vPPDqIOMz86D!FzNUqUEa8G$4ID>|QcKVPnN_Xgjag&OP zw)kjI(2nq7u;_!s2)Fqj%+oIjMs}PM?0e4?mpS*N#Z$>$CyvXkHjB$F8p+NdNqSJP z34(9*eNyPtsb~|`s~7USB5rsz3A#OSx0274%C)L}!0un~dp81wnzn+L(_Wz5E(vUh zozG{G!h$k2XfYRiH5+Ojkj$6S+C(y`12Fs6%LSIKLipGCK-@gp|wTa3}A#bPrvSh`}JuZ7B%Y|2` zE%$mclx--|b^790D+N6>>=5sS>Y*fArs?w7Gp3A*@y>@h;_!RPHv7qgv64gTvu527 z$Bs7Hrk0sgE|C-`woS!mfx2xJ3+$O0R8(vD1yP%4(kKm<8kEspq8^v^OmX1auw8)~ zOs!lYZ)WhBL_d_r!m1TqUz-G*;lDN!-qm0RMLKVJq*-`&NFS_4=fN$p{hct20qyjX z(-J!}D*9@uCCkGLl6hr?6CzcdA<^9Q$`)eXo8}AoP9TjjX7=K~Ak96Ne~nIjIq_w! z^6E4stQrPy4Q752XtEPto;{3PQ`|3kCTrpn;Ie=$9*E_^i4(1$ptfHvQxe6lrbrwW z6)RrKFA*IKLjN3#+1eu?N}$@Uk4STPCjIe8Z~N>&?lLSF)@?~()%t~@CR{8}YR@Ix zop_b0GQ)tLV)s&{h>Gfbxn=Ar|Fygg%JU+fTlb94$?X%aDv}{dw-&SgE;I=x47P4>`Nt2`&}fMe>~DW_!HBKI(xfj8uqH~OM^ z@`oUT9_L#_m-CXnb0h5qDciBqf311OwVjY;9V-KD|EWX&_4g~5mHSScq-CW_e~n_k z5Vfx2`E7KDszb5K?|vj4VS_*?iVY_zU^U@)2IK~6Ray`o)~a^gsMdeU?#TG zXTDxFt2(reE}rgSIGD{ICv2GYbnFXG*q~hhel*uoo!V)Qh!xpvte?gNnS&E9h~nKE zM27M+CBh9yS!q&OL#jWDt$|42zqCz>e<#QvCLY}Y7oK71{&6(fQ}Q0p0NN8cxwna z;{>7+hecziSQ0zGAUmK`GruQ+Vq+ zAG^&T#qmyr)T?Vm z?TWSj>!oY`t3|PnrW`Y$5-!j%@9EfAoS;F)$bWj)QnAKqR~aiBH0v3CXyUXbQ}JhR zU;*6{ep0Lttx>LL_A`Yb=$ON+omQ{LJ+iUASX4t7O7bGBe9tQI&Su`#X?tcHi7=Z4 z#Ug!Zum4uMSaguh4Lq(avp1}`0H@D@4E)o+K;i}*Vfk-<es0z^ zSpvTqGFux$^Fy+H&MGR%#}g_8%K4l8@T4{^$RVDCJ!Zzfb7CBg+srsEV|x7DBdL~g ze5bWFR$}%i#fyD6`IU&S&^00*(-u%0(&N$=P)QP)>~TBnz@W?xJ~jiE+pt4^nj2u< zK6c}O+1H>Vrij-pFQ!iU!`~h!eJx+ts$0Jm{aW>ptKY7gRd29;nD>(CQ9&9)_55^I zIbY|FIotX24B-wD7KLn*UYV20yCB~fxNyc1d&g;Q2F}0>Prue7W5sL$Zou)P-;jT` zlC+Hydk`|guv|)5W2laP=#d}L3R!**SrN@9$Kr*p7_#W*@R#Ll_y3#N zb{^VGLa(~xBf|g}TLIF~n!u&Q@iC(Gn=wCs*94F~?|90{O(zbZEH}%Xd_b{X6uC!5 zAtBgCpDH-S$&Eb9~UocqcM(gckF$6;e>84O@}-M#b8tEcArypCS4;#cDc=}Evhr& z7ogS@hrI0HG8PL3@Yn!#je0sX~SDzEkdDB zL~X$1jHlFC`;esA?o~+N@i;JEv-f39o==m&5YOA~i>j~)oMEs`U}FYhoE&eN;2m=2 zepqgqlwjrT|L*ITB=zb-d6#6xj9rp5qE`6YHin+`#mr}H>2y2AEDMT7*6xY zSurXV&iwnHbnpi~h)>wG5;*pQ=H_EM?QVL{Ti02az5vV@zx|Lzb8~Z?xB7)>&D@-L zijASj3MwiotQj)(aV-KC1;cf-!oe6G7>TwZd*GgqZ5{JTeqw+L5|wjS7n3dAAmPM? z@CVHxQAn{MIkAI^x(#LRwTdJVBB@tziCF4;%B@783pmUy5iNDC2tUkU=9)_1rZ)#x z(zyZc9(zKLDC^aWMlTKPP+Sr`^i5JMbw4RO2}K+S1to!SnjqNcmhZFCyLraJ@KTlC zp=Aqf%zfd3U)dW()4g7sXPMr_CHugMEs4$yHXA4w;*D#ls1_RYf72Nhr_RRI~aoMEqacQ5q1vHM%2ep%YdG1T~YUEN#_C5jm_WH=VTIKE0 zX<>G^0hXUW_jxzk1j}*k{hFJTMBQKW8f#j3JRWqOBkKmr(0AhH1(e161y*zv3kfah zR8+oC9h=7AKNUDOLaGIap&s$7$4b9SI(>ROUHgkmlCCL_{7=z0{y%%~0T5Ms?SXs6 zdztxgn8x7TyPyIiNO7e&MAV?BY+lOCOL=+wPyR1u^CZQ6DVt<=^Rg*mFNmNNMIAuO zND-ByqI4CkD<~?84Fs_)u7ZWd1^(wS!^(_|&V?B=n*SEL<-6bD{eE-K`A+$rpcAU2 zVThm2-@hi?qr~3RCscQw$|dJG+X85$=m88_8!$%mnMcnz6cS;G0Vxv4`RQagKNH1{ zbES+F-X|MT@tj29yxV|uA=brdDn+i(c^*F zD@dJVpF_1TQt#F|BGnJ3%5_eA5CFNCdGIjaxTnzhS1OnLYgN#17~8G&IUw0j3uIa9 z{F`No-lJDsqX5_Nc8lSv5mH8Q^6u;}95;2K*sV%3!?BE2`}K+zQ4Fc^ZKvDlX4ul` zra%iTNQvC$wm4o|GH0XfHLoW5YIX}d`lYn&E%RS+%jO;JDD%=}n5vF(6+g@&3WJ0> z{>^V)B^zEbU9{Usj%=qmNTJa*=9| zgl-Y*d|J7CL01B`Dt$S~n83v{fEe2twui^KaqP{x+R*^Sj%`;nfn^=Nd0IPU>ko!z zG_j=B1Cp;B6jj1Z8fy!Y3#CXB%{?ZKT-=-0(tl}y)Xo;IoHMu$esnZ}?0dybk8?(* zM=ixc#%&c9fn3B1Eb_|VoW03Co7n(i=PcC*(kAK^Hil!Tp~!&JSr46i!9Cw5s_?87 z>8}kss*KQ4NO7 zXbon@lmS=QV>)tfu5V1BR=(RG*bno74o|v~;8E7I2^+Vese+fBUl9(v*Jw-iaH9`Je zq`2Y@;Oh75piBJ*_@tozO#m&|u+duC)FwH!nAI1&ZIa&&8c3XKaA~1C6$gCU>DMIf z>Eok6@cXhsX6>BDJRp^?jm)}gB(v%$?i59iQxVuPslW~4f+U96%OTC@Bd;=Hfm9PK z+!@kMFnp448c4)^;3^GQ*dd%BoD<#ZOb}I)P6Q?FFHb}NYkB^k3ou6d0<9zewM#;xd zDDELe9-uWgq(fvwVB*{+`T5C(tmX`zDr^y5^E|_J&BYQats0Ltnd)tBxW=d|%O#ph zHMYanec%GwhkK3HM7XAs8yjl)zgvVLzkKD@&V0@`X47v zRwOG5d~eVLWQ+X{`n+fitlR?ei#ewnWHlkbppFZnwN)#8EywKIxN`ablVtEG>J>89SXZDHqVfHf-QyyW38HGBOo$RJ+QyfP9= zK0n1~$M)4pBdC;9+(C-$ry|a|KT`Arq&c@s-isDyyDt_0;n@|a;2fyiL^tSJI#G}< zNOUfe>4%|js#n%7x!|@*p6SyVp5?en4jDrK&o!A2ETn5J@ZE^pMA5<>E)|}GGeRdi z=9G1SH5ve!v&LV74U29+eEh}kU;}`Z|EJkTcGc#mzpqnz)7nl{nw`ZjIlmgdYso#oV zY?ja>$^;EoB)cw;aqoo=z1{LY;nA=QUMHcG?@H#H?9RfowhG6M9$YU#$S7|8YIEr( zQ=%h2#*H0s#mqG6&QL?OFV_F!7O6+D)Thh|8+qyI$4~*R6D%Q_aBbAWgvXNbf(DOy zM?19Pea*Cgt&E>0c`Mr^cnrxLxL3xG709eMmRJ~EGm9`WICuZvRU||z3dbBHE)~XHw|KfJn zdPUZNU+}Il6T)SgZlhyfwPc+vTKM5CP)jBs#Tb4@$c?CP?HO-?$+DFJM@YYe-ByK? zZALIzPjS%{Sq&*17&yfy1FyR2I}Y1z z3i=@GSDYORD^OTg+8v{F?6QRpR?Uz4DqRuHUD0?7N zY=FYZ3+z@|jmOqG#{%jcbHi@V)~Ys9kLJJsm&<;+P*R1tWzR-vRlTwv7WdbieX-@1 zK3s$?uww$T=xXJJ&AusK*&eGVG=^XJk883e)D8A8kD#NQ{&K;*ym{Cy8y`0LZ5|44 z&Pbtma@cbHD8WxT;)lmQtj=kPPutY{biCk>TbJ?}=>*D!Vr43m$8^dPf}>sAl$mt9 zNUJiM&~76Jzgv20<7m3I+3^M`)kw^3pt$uEiKZejiDF}5q5DqPfnH0WIY){iRD{tV ztI8)!WQwIS%#yjRIcDP1Zya?rNTu2@RQ{bTvtu)OyU{YoQ(PQHVj$Uq{>Zf~uthMK zB{+~c*rjYHS;6UJzUQ%Yp-1gZ*Qr@Nbx75eP51hz%gMn8fi ziw4Nt0b{4^pU{@TLJ08$!03m2etq{+VvrRD4ymWeN`4lM9V?JbBgv9Paq$$1qax68 za@TdO3jpDgXPp<6;6YMBhPn!z`kCr>$b`5zMvPU4#j_9@1BUPD*W2#Wd7uy(qnJMzV;r46<-mjluA8t#N zg3r1?x9hQd8F`Q7rBtZN?}zL*Wv#N?owFGps&eQijhRbLa&)!lQTVSdSKjQBSON*N z58LntFG!4Y*%Bpq)#;y;$xc}eqyO{0Pkwy)%g4U*hsN;3q}Y3>%y9M0cdmW=@|WQ% zRzK~Xv){8`1Zk0=KT_y?Y&s4emF2?i+KEn>$8>P;hCq>HM|tT!p*KgjL`@s|<@3 zW(ktz59n>cgSUeoJWdtjR+YqH(bi@cj9z3%NH zT6Nj{^-x5&&1Q~!o=6%3D#PLPAh606r`7|%GL^d%+sRf%f-M7N? zQIJ+u?z^4CB)1H8D~n|!Ht$_-w069(Vf3VJPYzs&4KghOGE-@@X2C#aX(WXa(fw0e zVH>)J+;^)J)rmTx6=xd+Wm{m=slRsB`>+7xr8+14GYh_v$;@WII?=wl=jAP=6KW8X zxYN=ewq0ZcNzdzLKmRqf*|*WWNFS!s#(t%A$HjMl@v0Lj9uMmPTFleZ9Px5@EWg*+ z!>$T$1*FC>b;?e+a+f8fN3eQwA+Z5MhyPOZ#qIF3SpOj>NPe7=7_sVj`%us7$}>1$(n9Ai&oZR5Lq;v^#<+P`E>ncgVowLFX$TC zV#iqmB}R5=F2(JjNG2AWw$lUr8_0#IISw}FuK73EE?EoM7n*!H{P2+y2|M~7x&%cY zkCYuyxm5w9{&Is&_KBjlxuuZ}!U48CeUBq+(#8y~l*Q|BR`Cto7@W>(KY0m}aevHU zfBr`UTBP#-H;ml3W3lD2}OY>}G)Ehl6<@0ty84u2YjVe+hbd0#G zp8rT0BdrmvaM?h50d<%C8bV^Z9zmgRLr5Qn1Vqta!s_O3(&?NgxXxak5%0KTHaxFh z$6fNyP!V?ekdaR z@A#Q3!Xw4$U_lT4MA&o9Xp`f-cALG%3xZU3-)%<$*VE`s1GczG&SfsI6~Mt>bX~Hs*mWM|#i~+a>7eROv?7 z)@hm@kR`E>OAXd!s1Nwm0Et8WM9m?=ImcDdAFhoq4CUh_w2~Zrv6Osv*J3$&HrgTO zU#pNP3!4GzJaqJf&<=$rQyeY$CnZo?TGVRsT3T!1msZubnx@LVQr=!O32~VFq!%L6 z78(*~3niBsJQz#=u3?QV6$jQ@QcLGCuan;YzH4AJUM^Yhl_n^MvXN}>`^xKKDc;&ix0RKm z0=FINn^4C*IHpaWrA`xUQQ;HkxEyJgt-9}63Z-Yt!Su*daPS*%DgI}-iAqD_{q2Kd}ZZ# zf1mU9-*2XVzj@K`6FnQkGZxmtUS?^)lkZ)FJUPHM4-qL^_L&iWM4%z9NC^+MtCe^Bv!pl{CuxB2Hzs>UX zGW|+gaz64^4N$chIx3 z;*O>Ts08CkraA=%Bn~&`apZQ9k&(S+_i1K*SJepU#PW!~Pu$7PR*N$x#Gi9(R zRk6aVkPT$l+**2xIA4+n&fmuH{Ll+;l&dxdo|iA-o`^3p51bpr?jLq-OOOg2 zD@%N8>0N?Ow>sf1MFKcqbijgt9s=11rKPi;l5FwqnO)OHAqt0GPpiqtVP_xt#rI`j z`YknR>$>!QMEV@egKQWTXU8IEtC6COp}4gaSw%&pNXsSA6i~T%uY1bGc&`V}W7Vb0 zc#$jtgz*z&a~4%};gqQ@%SVvdvGdK0<%<;L1AArku2lltJTi#_>4J_f6Cay)TzQ$X zwlH|%N?HPpL9O$`iL89{@rQ3-Gfif(+ss-s=_#)GMle^~BUMOs7KYwJsPbDC!-1*u{BNxu8K67m7L8xVLFg;h#GC zz9UkzKoC1F^ke6}?(x!AC@bnzVuEI-b~`XV7^Y# zPT)z-qmBp1D?W2>r=fI%S2f#+a5AWPc>lB`q=5QQ~t#K ziCUFjKaGWy7-l_Q4v1kM8iPyx0P($>1;wUOz7(Txa~#FRP-HFa1c2tZXKv|~PCA;T zNiT7c;ymEt{AaxVm%sY4)cw!al-H}bUl)RB%k?f(?gP6O5@y&8bWAf8gzfOJVs!Eo zq{8o-?4Fm_snJVQ9+)ESmYs$`cOcNlhd3R7V^mchn z`jjWg5$Qfw=f8t)WG+m@=NiLvflwc_rI2z9^N4_o3$n$p`{I*pNfondf>sqXH(sFg zUpZw5o$6WRjPt#h^j6J}&i;7wg7hC(E@=OH$JbKc+V|zB3s(P|rsb_$^Z&3SygPK8 z?~S0R@WBOcwGep5^z>fgduyEvA;Q!{V*|=RXjRzsqgRH}#}&?f!YZ#waoHPaIpbgB z(!6vRoj?8>HEu~WkkP+|L{7m`N5gJGAF$UX*@-#0PG-u!@w{i>eq}4!W29WuDGpK+ zwowt7wx;#^IH*vSsRMTbFf{8 zs(_>PBV|MQdbWdoq&x?$Ji5Vke_UA(qpP4g@ie3%oOZt``B=6;AVYmuSmSy#97+43 z!xUZTcf>o*`4iYiT^Gd!?+@s6I7)9}@h9F{PNKwkj3cg#E;H9fZ8Py4f0HxxRX04o zF1qS=OuWIrOIb|T`3+v_opF^6K7C!(9?o{lvhGI%Ps--RKCE(_+d_G@}%u*eD1EW#O3toPoDWs+Jfy1S1u^} zRSZ1+3^<>!6yF2eAdc^E@%(uIiA}3aTWs5{4Q{5LHl{vdu68z)%q$Dms!pjM(RqTT z>0P0X;pbf+PCY5fV?H8PVGW9-bQMs0Bbk3=_(e&VpgrJL0JWkUelU}Hc+ zJ-^^QuJDsa&E9SEOqr-&lL(a=-m2A#^MN%UF~S@%&jnr#!R^^VqeS0T6S_l-J=(*M_+2FUWQ(ZbC)YqT?UbzwA3DAR zxRvwiEU#^bC#;DYi_l=H{W5wGd_Ds|aU$w#{P0zeYvP6q+Hi*4hb>licov=ISWBl< z@0ZMP43DDnmDQvnBsXmTSjMsWVdL|qv-tr(Y)t=Q^oh|VL)dY4jG1JJos$aHJLoHt z17SLPlaFQx-ND{vo1mKVl&XTP^U{s6+Tj5sn?q;l2kiQ(@|>qZ;VgS^ax_WdwE8)g~DXTzcV3N9Q^-P69 z(j5Qhx2}>66NYoq*lp%I*=;0Gwo@ELW4EHY`;^}A(FbkEG3wCFu7o=3a>)gzo76(v z?^gDfBA327Ti;7QOPxX&x@lFZ&Ub>WX~3FUrIA6w%w$%)Y~FS`MTHzRWxeFHCAQ=6 zf*Fp?61vbAcS29dNmcTsRCb$h^xOuhn14WnVj`W2Zs*OUYI2G@UONR`yxpXNN%Sv) zdyU~Kz|&hQ$_*QQ%4U!n8IYc-B`adbXsUe^)BH`7X86Fxj?G##DK-N-=ve`DgX;fJ z`W}ji3EdwponcTi3!1Z|Ne(}A){f=JQ6u$pfa3O1q=<^xuj-c^4ZALTJhOA!eYY&N zRfE1Y4S|23>*zz|4|*M%}z@3GJq$q z|Lwgbl^=NQ*s_IS(2!c$-4wTzA~{q2di~j!s*MaVwT7u%CE~C`ue9j)7r!g02FC-y>zdq$|{VNcns@ ztpl0Qr?c9@ilmY%xdAeg2a?~CSUdLD<{N<^mEwR|Ihl$m_C6`8r5Cl)iwZ-Rh*Mop zfZ4iuGnXw}TudrOIa1(OLHdmzm(A1K)tU4*w@P_~szJ5N$=0f8fj?xcf4YdC3TQB93dJXwes6C;Gz>4-39Uh4d!VC_m%=f2%$B$#p6H_3CvIZn{R zV}Q#Ye;fz{rmqfK&x~2&c_nZxg$FMXjA%JW_K5^U*l>?O64gaqhR% z<)RAkx}H=ODqB^W+rjlr+~l&5nvj)Kpks7b5OU0wsyj)&0_os1m2@3gt?S(jCA%Q) z9E+)A!Sg@Rr94GiDBlM?f-z;m`pDyjFY~fxWFUIMQ=$J?C^B`Z*sUlr<05(Z7N+MG zK-y&#lNWk(w)I!<1uyW~t(PzHv$11l%wyK3%!Z-K#1nf1Zh7E(5WG#RP>Qxtcc zBDGWm66W`~fZ`TKyDmCD>7wMrMeVSa7?9$BIYUF!CP*IqSk^2<<$Nk2F7&8iE4OC$ z!2L+^DoAe4k?wFQ_w5pNKm&$yx*b@&GkiK~=e2D1E4b$przxiVKXxl%%y9i+ z=_}eRnp%2~bcy`@5Ay&02sq3%%Q3I5UZJCN={@vb_tWk%Qlw<)4cap%2`eK)E>9GW zWH>M6ju%f|G0(J&gpW(#jyvcKP84}+_o zuM!){*L2_+zUw{W590YCFBwAp;X4%~)Am+&t7ObHtiq&|R;Jodqd^zi9k(uJUEr0# zeh2)wz~>}nnY4&jP1nk^f~)=XSNk2%u+qBpxZp_r)A~EhoNwErNe-+bU%U)sA@M|l zWy5*lW!$u{gikk3#J5|#m`SZigQN|3I4WjiXZUgfW+IK5*ciGfmOx_Y_^~?X1&Q%V zAFWB7Yfv!WOV{R-{rtT4c5H^6H&RDODeeeGs;P)-NWI3;@E*`Njo}Aw5=nY(pj2!GB^||qcQ=iS=%ni>x6Qmf<8HvCuw#NmPuTp+u#PY_cr}|O-RN^r zbx_qJN*C_tGHLAodz#cKQs{I}=YL302<5RGM+bzLF4icJ^3u6SGoZQWx4o}=TeLL# z2HP0^k?g8EjxA6o3c6;df;Vmn_+;_=X8%ZW;$*!3z&TYms;@hr7tMFU`OH0^3{O;TNY{OiAfp6Ss*O3OmfZemrxtI(pJaWSOF#J}g`o+z|eh$rV@o z_Bk9>_c1QO3bW&10WQV)I4-C)iB_vvxxNU|M!-u4^Wr#Tb{>N10h&^IE}daazY zYtn8>JS4T@f|t|Fyz`)OUXSy&z_O{g6<1}uCY>k1o{V{K3G5YFB=iu|pbd-VJm_mV zWILuZOO}7w@3`Tjt$@HwU5<~7U$%de0W3{#UU-kh+c8)QjKGpXagg4or6RIeO}%TZ z5dCp|%KeV|#$yA0T2p}hRFlO%aMEB`aI6u_11C(j?6|Psq0_CxtzLnAs{FM9kF^^O zOn7_@@1eY4GHy|D$IV^?CG?`QbKfLA4#TCh*sbwJz@a}R5poU1t)$2@DgwQ6`2S+~ z$EGkazdkPqi1}x2dB;3J4!kQo^Qxy*x3WoJ=%kwzO9O5KBOzuj9vJYd?R331qe#}n zg1tCvt79wsIP|)A(X?Ijx6@B%jR`wmX2X=&wOC*f)QQQ|MS>jx=LlCd7#`;WvxcY@d2S`I!v6i0J6sGogtQGzt-SNK%2JGAzdgI)LHa8{<6kOKp!GihCk%{P2dOg86BtoY+;{? zMHN6HyQrDy{1ezNK_^`&Nf%(JAfQcB_2M zlqaKuR;ya;azvdaxJ@*h1Z5#z${g`cwpuvmg3MA>*b)K5&CLrHY-F?9EIO7XF`p`PqzBcp z;Ms?*xM(wMjho-OGXI1BgF*TLQJGWoJI!#Rl`V7X6~@3otT9ZQe$+RCE&1xQFMV)! zzGl~?Ogcw;BdFgY-Yc0Z_~G99(F^uE?O=5N4bTI4k!0|c9uR<_#dt`q*RRo!v!clgW| zI1fkr#rMv6QeIEz+=kK zshzTfV4eTX*;#7vNJ%$O>krawowRNSPN`Kb_o<$f$KY>OF1bNhgFQZsOtTu|UaH7m zJirvNybw43gEvw?{MYIn90hvo;y_rlMN%wVKA}r_*6$46B405fN!kkCDE9?i_e=!z zUUgf`6wTP_T0&iM*ZGdz$}wFZ^XDT%l=HIz09diR0e^aE_OYeG`+!npP(_kKP1I_=j03y2WWUXH0rpX;-V?Cnu79rxDtx&r6QW-?LjCh4i7l&&Ih8bxTi>9F?S_Z8ZE$n4UoIiB)=*va{>ynCV7$f zX6`A;gv{PnwwAQGw6bfN7MFepXrl-QPh99z2+ePvc1v*pqL$5WZ6J(b1LwX6vT=sl z`-}{kmV66Lhv;9Gy)s$>nKB`FS?y(RJ! zEq08WA|ug~MRDmANu?stIz2&Fdk^F>Y<0|X+)O}#;MLt>$C!+1nkFD4Okitg?}dz= z+cS5D^gCb>0=*@tRH=@(NzQq;;(Q@wBtf1{WA&p&zsX%_N&scI-d;15PN-8_*>k~d zQDJ?DF4Xq%x8zE{P)XRKzVsE4$vS%A*JnnHDmesQ@-mP#;%Tc;4hn z<}}HVE8E>|EtH3Q4Hmum;U2W{HC}Jtes_e;-#_cvpW<>U zvV)39Vxq*DF}^b-N7@>)b2>8Uqs^fw%*~Lsk#ln%8?weV=jQBHw!l_)i8u!o=1TEh zw%-9OI4~hMMu_e@Td7ftaQ{r?40mLcZ@|}^rDdkY%+K#`?Kpa4roef-x&xSb+LTqo z9#Jcs<6SQ4kYJ8=i|Xp^M9(!dwW?*_2Ys8}a;EPVV4eI$h#24M$9-9Z;5#WS8vZk_dH zR+@90D246?zNA+6<}9qM$%W+j2i#wKW$*O}uvjWxP$tZvSIm0A^*P)UbO;`BWllP- zo_oOkp)owxwI*b%@5Tim6wY7IthX{Z!AKWkn6KxM*Esq^^9wP|w!wbCoRQ`K*k-z) zvD;m*nIORxU=YM^$a~l_u9m*GAZAQxub+pI1yFtdp)t6^52|nbc+56+rt(RLv|r1M zm~pcr#RdkNPN^y(hzFr2jOn4v9UDk#ZYi>vYTw87`jIyF8Nmu$fFvXNHiC~B(IXSv z&+q!zDl~0V<~hIrM49OP`9VYYx$p$`w0oT>Z%X3iy+sgd&rZ5h29P zL%x;YcsSW$oV3jgx<fkN{sU#et6{lZwEcN0c8YzMJut=J~)a>Ux%7 z>vpJ})%kWX&5FWMhyg-}v3`ehbMiyll_hhID=#xS)AK_N0q-&=%;aDB!-w;2W$lg{ zEoN4Gw3k1OT%T{cMYda`#LT7_l?tXT+@jOyOhN4QfpU!m_K{OpC{kp{P5MNJWC;BO zP3j;hAM%3tQ5?kG_fQeH=eD?AaeqoS`2cIu-Sq-MV~{)yhw4cM~cs`IoD+O6!DJ0-%9$F z`IGJjZSh(&yIivQjcnHfrc)O0RVFN#K=~eBVC7xxC|>#J@7{1=)mP8&i)^=CeCZ3X zemKp5oX7vISV6YiaTpLvPlrIdfZ`x8I**FD&n|-haD&qrUMTt4<PBvMzuhEE))H@PSKAmh4@xhA>|oH`dIrLK5& z^R#YjbFNzt%~l07Ke87p_s`3-Of&86*3dPRJHLl%;`9vF{SIjG>S?*wdv&@kKIpyG ztBWlRjdtxOrJ^OIT(n17r@-MY{W3y~SQik(%*IRNjr;Q#o4Q{0mSanWN1gM4 zxMjW9UMCG6#4zij?;Iq!)+^GS5hB~e(*=;F25D(F+wGdH1b=jxOkkr^e&`U;EBq{E z@WR>!WO_VG*9WG%Ks}2b_M}>MoIp3FoQ58!w`bNV&yz0YV|BGFB)yN zN321GMq{6Aqrc6*x%`*J*fO!vitrQa_Uo$pTSLPKru3y|7(q8qyyAp{Jx1ClZwtXQ zO}fh-=7!s)d05wi{g`Zqg5g$?4>E>(-d5LmeOc`n8Q*F6)ge>v<9s6rgdtoi^7p%`5CGQ zkE{)5epr4-^vg%4y$tPkcXQrI+lFj><_`9vyqjQOhZVtHko-3I?+8I=Ehr5_>iM4|0~s$Ujqj`e(gjnpNV|nn ziV>9JC@zK~YpIAf8aPpAVp(1ug+>(qpGS>wSx5HL4NqBwiBWyR$R6g0iU0cR#7KJ3 znReA)s~n{+9XGAM$Bl}s_@JAPstOq=-t~Sr1s)9;rH(C z)(!jNH@^B;Cj%({qB;LZvWj1z$9@ek0zBM9oVJ@NE`cH&^d<6G&xfT(T2;0T%0wH$ zsJSWF=9VKZ^~a+;`B+LB%MfnRBQFM}RqvL6`18)`252}))%!@M9ed~x87+1(#X%DF zF6bckCA2xa1lK@>l}HoiZOQ_lGxYKaYrS-H9)LwrOrKKSmA8^Mdc}k~<<0O#gEtl} zS{~fW9#)q@9I8UyrEFIhD{sl_!xBBW_}`UR`*pZHCd-&y>jH%r5}r3KjILM{3cO?; zCI96uVZT8l|DgGMdUo5?y8&B*ZvhR` zt&pmD<&yPYnhwQTa0YfjmTV7@N5)IDRhjf5!D@D&V69W${4L^Wr!woNHV%u=mt);X zhzy<#JC`*t&N%#+_f3n*`6N->Z@~bhJ`4p(I>mv_y^V@U@o7@z2yV~1Dl3_DQ*f2R zzU#TN&C@dIrobcecW*Czw=ukI>bqL%T^-#bI;q<2`|j}{yxZsS?uM`8pE+NDH3yt`4B^KYbmmdib(Wt5yd(-Ni+vST11$F_RleoELnUTt?vt`HNW`7l#`9W zuzO#CX|mBPB{<9^Af-7My0Ol$(N5P?`hoCfc>a&HU*0<(>n^ayVjnc{$`({A z7H-;}$G`Oyu@R)3HdFY(TteVj0u+8X!+cPf3_FYUFD@f{E_p)aS;Hqf!-w} z-8bIz&Xx0~?Ymwnd(})!uT&;O{TRBub-MMsuO&x>TYN7-lj9@uI)!afG29Ybh?Lz9p&~FVq($^ZnL(eLu_GioELnbn zKp$Ls6|-_er1%Kk%C?Dmg{vksgad(*BF(GM;Ry82-6!Y+7T*hWn%BATGmuMJF#jVi zo5_S+lSpyd8(90NAC*EMmLZxt=#>-j6P9z<+RIpo79)Z3k`t?-$UItxv#Eck`qiD$ zn6CDlkx9&$u7g|^18Hnk9=-0!l?9X)*fhN>v}8^?XsWvbb;^W^`n;jBkwYUulU302 zvPu5Zru=DuLBibn_T2)KZO0O()=0vXP~2XM6jBkX&~zeNkS{57LY77-ngI6ibdo51 z8j_`MSHy)ihBv6{0#lhHN9blBkOK{y+Gd`FW+$b7=~EiRQ|L(XS+}Qt`Oq@AMRb|z zc82i`>a9=RTUCwWciE>TmC3Vet>FmJU|zJmY>9c?=On-}Qpd0GS(}Saf&p@DN$wVmq zkbE?G^ECa)g0N<{#_+@BR7igiK5(MPvd;rLO56#l zuDQOIKAPpi{7G#gD*~aI9ET(`c$h`EE4TXoBVi$QdGkF_(gk^;2c~oh`W>2OIbsw} z_-I$C=Df!Z1OMXyB}Thi=Z}pi2fCzVTD88GFimhC>?lo+G~R2gW6~E_1a1t(EeLizq5DH0!`umMG1(ZHrN+syBi-O{QRLVqKpGA` zw`5_+FRTSV^MY(-cX)ws+-EiAm8P|V*+!ya3&m}s;4cf=r~`HbzBznr2SPO1CND-9 z1L~$XXb*EwfvL~>PB>#@M2zFp2vq+zE&BeBeh{xkuwR?<{(C!j{c9y9w0!IOK=Aqe z}Wm~7kEBE=vNtXratCvoLSUIQo1VEqSal+<7=n;T@j3BQG z+#6$BQOYNE#eOrl$UY-cm`!nzQkzaiM9#ue$1Y_aC@Va|Fg2(o=J8By|5gYQupB9N z$7qsg%xitK-{Fq)xjC`G(WI$!Ob61YL9SvOO3Be_r_l>Qb2A*>@&0?Jv~I7IU(XDA z+Y+u$v`_wCx~y9C2mNsZn=0&>e4WNim;E})4wowCC|wJQE!Ws;a!%1Jyc<--RH{$V zhuqVpk?bnwrt==IQf*~^zaa}TI!u5t88VkO56vq}pWkbKXzH!8+aeyb{+A5(UZ7;l zaT*|2>VXD=>$q0cfR)^*Y=d;jHYHY@_CV^qYh~7Z3So#84XndU$PSMv5+17pbW?#TOh!0AIwValp5o#t5`(r3 zw7_<1S1%Di3>|2{yhI#J4!jAjRa7Qpj3kRdWNWc74ANo(ztjH%11wI=5Bv@}GhrCp z)_$`;$$g{6Z=twr6uC@A9RBt-S^oSR(4psJWvUyhj~u}=*z(rXNMTwnIziu_ne3y3 zDr0n%M2f4JZU_+_COwKs@u!O0Gf#_kPF3Jbt`l68G=}FZ^Fp)eT4{^q2$L7u9kfe` z1gHOqVW23;sVpR4(mAuwAv&}~mJ<>aSUv5cq?W!-Dm>GnuSc5tu<$E+@cHcyZ)YqcPfX*ra5nYk|;JPde@FPWD|Ig@{LWWE7p z|9<2LyGf}X?~E=P0qP{h!B(k;iuhE%d~Ry6CKIAC^$=TF$8|&e=6oQu9E6Z{3f;v5 z8z4#?wKYHH|(5&mt>ygNmmoCVQ23QWF7tLb*d+&b4?3rm0G4GXy zSCKqB7BR<-L`*5gL0DifHIR7~y)$@TRJC+@060$al zpOO^sC~-5p9sIp*qIRW5f2n%nPEj+fGr9<+j&Xuyh0X&hdTU9Gi{XiN+$tv|lFMUK z=(Z4i{D-puav z?W9+KOTTB!oBv3;W^S?dQ}kZ6DsazYz`SH0Yu`TL@$I9g-aI~`0Xt4_O2gk@7%byTO~r~f;eEEhYL6lStq4lP6)j#ZG^P#sX_ zw1hjstn#iVRl+FoC+=G4RJev*WSau|9Ps=J>2+T^qib#v)YP9+-El_u4rEP1I;Sl$ zWD!=(SuIvYhw)^ZgI99IgSKqp#6Oeh2_(nJWZX(|n-LKand){?6tkAy;+092d@0ql z%(KiXovK*a>e1`1^{SIU39A-mg=B}UB^xK&@>LrB1^B*KqlayHVqT9Aq@CZNg(^8W5ig|!5jIg|i;(+=+mx=(_syvB#IIRktv`CKqnB5ZG;Hp(UR3@;y z$gwxT8dWBJaXYux9kxJF7JY)83th%sae~%X&}m}utc8vVEcQM_*U}vU+qv3MTcDmV zz=xF~HUPSLfXn~%pWdbx$*UG$JSo985NKnx(Ya9eiVoNhY_YsATN#U%|7DZd;J*3g zA4a~7+fcpih$)@1-9o|)&vJp=0Le954ISChC$9x?Acy%s0TmX3!d5wuF8~uSP>i2= zjvfCcgKk;A@jGvm3VuOn`_-D{ijmeiO>rkFatwmbq6f}Z?w!JVRaqdm%Z319V4p*$ zFq>J;-khE2TuWzp?VvNI^>m*@p1f07LRXUHNv%-UxPe>`yX{`(4vsdLI!Pq~CR~VK z7Sf04Jb8!YJSmbr;K~Afp|$D`my$U}-uK>Wn12&hG{DORtD>U zyklf5YyRE7^p2&Vc4nQqXxoMh1J&E!ijjnV!R^fXX@`6Y!`N|&o{_vezuDpdtEz> zY?o^kcbOs=F<~x&-3!jN)L?8Vg*hlExGi(5ed`rlgX5$f^a|eVvC={ zB!5!pw34tg6^3)NnOYi?Suh*-@vJy$O-Q4%K~e1oDY!#Jt>GdUY6P=Hi$( z8b7H`9xH64y9rh}<4mnUUlj>`W>$)C&RDI`=fdPd?p{0HMn?$`_@Fq$!rwLv8J8`^ zqlIG1l3^AhvhOd8a!qNa_}DUbj4(5VQ*Cs%YTcxQ0LYX=7k`udG!2;t|CEGGyN&jo znKbQ!sc+>qL4}!Uk6w9UN`HpBQKh3xp(o>*+qU?PwM0-@^v$;NpkKJ}|4-AN4R&jx znCZe0!`ye%lJ#C)f+7!$*YpW5kTxGp0m$(JpUX@at2xJA5;nR&2-VRyf*Ms&!-2_3 zH-mN)Ta#5=pI8)fHsh1aiS3SXkwO zR_7M4GZQnx!a&MFt?IbPw(w*~vPTQE%qhw5raz85rMdysQ1~344UURu;$ak$vlsZp z0=V~z&U)a6*~(hTDu^Bcw>5D5Yz|(CrktC7e_-09?v+YzGt-IAAQiF6;$O~6wW`aK zj1bU59TGf<5=V*40~=HioNTR)hAazo$Rof0?e8^z`sr`p`Qy8?6%@CWB9TL+oAKeZ zCvAIV9xkU8i+*P6owZv-#jL*-s}7;OG#27Ns2-tq-B7J=WpVJ*RkB!KymTh(4loaZ zVIqU$19R8*zv{dhF;HHCj0SzmxUgd1Ead=}$l zP)?Yc$J&>9;bYtekE&b?^OIIiXd#vKstL;!=>(4>r<|d&cy85%#&GO)a7+vw0?Nbg zeaa4Y%)YFASHaqE+uVG-R>6KFtq0@2HQvi$r{?v)y_clgv7K6OWT) zDKyfloKigoyA+?hrMS-Om<|hlpXR_}<~m<24+@r{Kg~=%s>}Sad~J8sPfd9w`MA*R z_sc-c&|zX7ah<5fd7l7sWSRSr;f}~w}Bj_T;jZ>vZmG|B5E6>m=wt>yCYs(W$ z!_Hxgi~r{9{rHClfs**Qx4uly+p$1-Xe3Z>Qrr!ST&E)H6+JFci;Pvrx#I0ywQr{M z0bB0d19oc+vyKDKjAo+quVT&zw#m=ayPY&wWk)6tlosd84*Kc`4h*YToKmGaYHl$) z8cX0$L6J_sLp`+M)~c~JD*g|J&B-w5UU%@&s@v$^pn)Q2J?mt@gZ`J$G^k>Qoh-Q^*HAUPJPjExEg+UENVU@=_a?}co;>^6FNhsj}?!f z2OlpnH2ze{U`3ymw*0U*_Qe| zp1B!{z1o%AA$s$8uxS0T5S=KE1H{CcYWxYKo7hfLs)jNSQ3|(89IuA@^#)ags?6!C z$5n8^#DzYkilaFSvJ9dA{5sy6}EQ_)I6Y^Te@q&c?M&2uako=j{ zpmWkhiOnSDwaHq^Gty96irYewO;m(Fdq7i6x>!s>X&vYvqbc{@&JEDUDKXsG@32yQ z7ZSP0u&IoNqR+@&8JRHS_mrY`@deX1s;`vy-b~NeL4q6%8YejEOuJd=W*FoWIS(I#yV~b%L*P~U%D3*Cv0W3>Vw{RYk-Xu zUA;|;dXXj-JiQ2^&M?fSyeQcNorjRm_CSc00A)E)j3mViPDTq6{qyF5GK`CVoO^iJ z_um-Y>HMZ_5!qzN;Mr>go*fjINs;YT#1e6*+gfNsG(vVWIQ(uKd6K` z%)cfYlrtBS4v{r>43!;5P)Vk^B#OjC+JO{dp`-WD?*GC*!FqpiY?Mvy5}c#!70VQj z%6^Bfj&~+wDz?qA@cbgYaf{TiEne(4ehfXzhR}|tQyc`8x1l|{d0IPU zGV36l`j{gqo$|m07KzU@)g2Nn(*i*wiU9|Hze5WBnO~+0c<0hxb#$`a_Nn?K79r2! z&AJ!(iIA46s~`T@AVPk8aO+#-(rdHF={CaD9g1tGNGlbAmTUrx4a@*f=~S7jJG54c z?ccF*Jz3rg=+eZwlzSzxF@e~Ltecbz9)`9l;~-|FLBh8UE}zMofKmmzzFc?P=2jSb z({Zm8nnX7pHCLd6Vv0`+okgcIInY5m4yaDIx#fsK74C&dYFro`v`y_MInuMPTCa7` z{R}8{wSe%y_Xs}wg5FW5(N9ptB)Zgr18|@v0Fp%A1cKG}fgUu4p$RPd1eb_$EB1W& zYQ)fdfhjbiTPDbUaBt=h{>=ck@m;_8m>l5;Hv9D+5eOv=5n$&i?hHjvQxS!p_oghJ zRpXp4+e7yXdjz#KUMYd(xTDYu7?bDl=N7LNI+|PV-mY$iU3ZV*I&|{v5yZgtLeIp> zMY1Agx2%TTo3c8r#-mq=XQe(BoO8@5b-OxOhJ65Z^hsr>EZ&Q+-92WwGQrKuPOE=@VvT946rUKS{kout znZzm`T?6W{9il2j4vLp8qjm4^>gzfB1%t3@`nTIzq==ue zv13qxY#AbK4pUqeMarp&8s8WxT1x4HHl?OR(ZlW&6iIU6QYz}1s%d4wdRzou9y;g` z%jq8H1ghU*iTIdf2b~Sk%5udbX54sseI+h&;_>~MkCNfLqCY;@y6Yye>LqJZKgf5K75P&mvYEQ z>7Wnz>~&g2;VVKb_JEHLt|Wlx6zlC8RJWMp%5oZ`b0u?5L0<;23Q~g;f>G%|5ax>) zw>xPfXKi9ifKDYxeBTkL)j`fdBQQQ%?JI8em(KI$nj6l~h6T1Q;`GOeU;ZZIw}1NS zuNVFC^W{WN=wA1AxOjwBbLi;mEdS3*xBr>2ZTzF)vi3 z_3B|yO@-^#zMr{v07%Orz`c$`79iZLVfplEM8k`hYm^v%@#Lep%MZg}oxJly)5J}C zg|#`xyTQ&<_k7T03aO;t!8etT~TIdBI~m zmE2ixT9C%a!nNa_lbO1@9l#s4nS9C}W$F~Qbepn((F#7E*BG8crvyX?JPN{8t5ink z|C#dx;gvwGV4t8;lp4=K0?KgcSvTaT_R_7%VVVLf;F+y zm9CAdJ#N_QLtl@(VwSCBgt@h|7%=80u+jBBlNN0?Wq>l%;&ka(uF0aP8|(pK8E*_9 zY`;|KyVo}_w9nzN;Iw#$OEoD7&I@g2&MERjQ|Mx8kDwIngTsPc=-6@}nA^HZv#`$r z1I<@td7;?0DT>+RwZjF^(9H$iqJSmWdW&Q6Yd4Y9W=^fP225DVMSidM``J`C(BHm)hVBD+KT>wF-J~6A z02XD?gSQqngr};yNuNWiYKa(sE)naG&V{F_a%C}r=-G{eYg*L_=BnE&Hea$&u$D>m z>~py4ww7slvzWvN=S%t=Hc76!;n}C^e90yfHK{SYmTrfNgM7)dxzWOBze{6yHnT~P zD;}Kp46T0-XNe1K4 z;OEuSzp(PR!pxV2|7CvPinZKriP`|Z;}L(~NDAy2e04@%ylRRA^5s%0B2HR5rA3rH z=_7(Qd0B#nz)ypA2O|KlvEYOgW~%c-3n34$KCqlwIich~lB9KkPlM2H7b88zEO)FC zmI+U(PD|QDwCYyab7@tNl=%|mwZuMI$l<()sd20i+yS}fOI)y}o>q05BwM4<(l#I( zy@cflqVr2nP2vn7a(?>w5=pXS5EUDNNJnuQ6iK5ZcK;7D{HHR91ReCd7Z=={zx|u{ z<~MwO-GVhgS+?L^AaZc+66pLNhn6{2dE|#6W64$TxUhRZE1a%5bKcpNHe|kx7K#Jsny8|F*iVM2fN|NHc&gB|g0u6tm z_lz(bG))b?N17%rq3Y(GU|N+m9u47b-%MMW@z%NUF@lxla2m8;#{x{_#=YUb{)faM zISL$7PXS4T-3Gb{r2P(&97z-xPmwsRcl_DaW@8ZEb!)5t+ydFW%GbS_fd4nhT=s-*iBrBg5#xk7PmW>a860E7ZON@p#fzRY{H znB#?`L9t=U&rxEImkOeC{Vx|yG~npgude)!thQqu>5Oo+nc^Tix{-=F4o;ai`X=Oy zrwbC;Wx+A-k7q(sR4a?wDpw@=A^KL|qg`+w>%tEkJSHCHP+r&=w<_iP_kM4{#;VmZ zb4Z6BV`Ht68?Tq*dMNUQiU1l9rB>DLj2XY0Zh05fnsmwZ$>e+7t~hBrK{LfjA41|n zw{v}9|NqC{m%ufZWqJFA7m|x18^PoiAV>gNtYR^=hz(X%chyw&GQH0B)pUQ=U35>| z-8J1klc||nx)H@)1O*q+fXX7PvMNeg6cx2pus~5%6oa@FmVyg~3g0)gcd3F9-+tvXd5Y*1(Q7Uyh9rliN%{$94_AfvfTs9u z2nJ%Cte^&Q3ee)I@G)Ff;fLcyiOp*04AD`h#2+8)1%ew@hNxen@!B;tO`^Y%L+hEr z(YEQA>-e}z4TiE`@^VD-CZ((m-TDm^Wu&)ztaI52dmq{u*f|mQf{so7yaw(f-g)6Q zS*sK@*OWEMX2@YH)8$p^fqD@7+I^fg zdTo_~WdVdtY=h~kZ6IYHLOkO`SIjR6fFwwzW*j5d4djCk0Vho^6T$J1a>y5@#H((M53TXq>F-=oB9n))L_VpAKgRqh3>t}$!BE+)V%&VtfjPJXkr zkb_Xk;a&rAM2NcPI&TwSf{1XlTaL1Xqbdl)B2L`U+)fqvJ=6t#FT3MY_yjfq!8ML5 zYi>53FWL#S-TWsXlq!##wNtQDpzF8ZGs7c?t_=hRa=Ii^h1WB~ux%FvtM$K_*0+y~ zl`DAGE~yAO8wADO%RDn9fGh$$t8)ZcT!o38+9_%%$h-C$d{(>UATV@2sS9NZxe-Ro zHiyIT{YhwB^*5_18xv&I3O)H%FK;O)-*1^)oS;V7%QtjBk0Qt@llUz2%K=Eq^6}yG3 z;U&!27^q4V)lSLvE}oI?t5;gnNsb`BZ!Ud_fFuc|jlgDMGrvi8QG6>{wQFX%qMsM% zX`A5e@Yq<238Tqw_++>f8~ikniBUpwlM~<7{F`Ff4?X(JYO;mh3fhj7xS<$r1gc6X z2I@x&sMrc!sZayxi#O65UL(;24#+$CeeZPWh~*MPkW3NPUXyfpu1b%QT*%ryE^dM{ z{o_*)M4+~8uUNx-G&4`Joi6qKf}1N@A+yE&H7lFbGt>{ro`)Q(YPy6&p8m$>vGT=XQ!6X6rVKyB5NxH zk754`=Dfdnm@<3VEt?d$DmUU#_fQ;s@e8zl#vgrcY zOmu4fjCfut3Y1-xC-M?G`iM9LpRM=ZvFs1DP~>RBUp)L?Y|OF!Zk5#i{Jk}%sZDlk zD$FD+X`Bn_Dh8!p_37{rqjowM(t4ur6dz2szYihx1A&F~BG}_O;nBoTBv->RSEYh1 z;@k*sf_NW>(&KDcj5f}x!qRUG>vpVQJn239Xa5*B9Ay0;)n>AT-GaxC11ZOhrlDGj zfh?#hR0pjFa-$qNiKvj}C|%y`dzAEXhk&=JYR*wo1O-N2AY-yO@?-##%7;gLbhpYbx=R#sW(fzB={J*EeAXgcmfXn@3>gQ z$c__lrml(poHe?he&7thtNb?x1bx`P-Gf|yX%Mu)s0^u(Vs25Shl)kL`-f2p9;=wl zFx9as%lN7C0hfi*Y6X(l&Bug=RQXw^Po6J2HYJzFonKWySp|w~$EM&(27PpHf=4~0 z!Zv874#?JF=@2GU_0X4uSL7=E2l~bj=>t{kN)IGKRNlOjp1JTZf!NpdZG@u-%A+ig5xpS;{N~<=gYl72c`~K=F#4Ao?~STZ(nOSZ!Iu zzvBWl#M{DFMHYnBbMTXH@UlZxc3XV(+Q0wjJh=gN>yrb|kag@(XTS3|XmyQrpK~Y% zMBOr}*d0+Co-P*;bMUY}5_1_=yB7OEDnY&g84|C~ii^4~)yYe=Ga6WrpLZ?}&jKw9 zY+$S8%wdBtF7q_pi4A^!|BusmecJ$(_pY5xCwtfd#g5nAZAL&jLNP#N-hh?ZWuW19 zSB6#Co87vV6@G2eeX^YaUkOe_O3M1kPLd`$Ev`|Na2G^u3)@Jx`r!Uk4c$Mrk9*U< znWL3sTWr@!?n5kD?b8x6AkXlw_G*hh%A`jfqL(;6AVowAso*u-rjVtM7zt{UtP?$X z@+7@xJc_#G0@LU--H5AWM_IYs&ypzvm>#)4Iu7)3_FI?+TjL{^;fWNpf+9;H>07W8 z3afFz7rUMZ?mKJvCS!P`&mZD3bZUi1*tyUDwR+)s(@I4)%SbzRpUu=Y>ZR@V*-Kq? zrfBis|L*|k`S_XW(2X8^AO5D`%kLUY#QnG4exK~KUCL`TP6N03fU&<~CQ)WZ*WOUk_`YWx`SEYLc&dpRSQf4onh0nG{pXT(^ z*m^0aETUhGeU*!=lmpx?K5fw32~lZeAIvl@(aGXlqPr1016rdSJ$44*vnOHs)E159 z^ZKW7ZwjxE0cjS%WW>xc))A0hjL3L{3Kt@?3)NdqI2vwzdnYLJEb~x}c6t%h?xr37B0IqncbXSM!6fJv#Yy zk{tR2B-zw>?eje^#O&H;X}%y|aA>;LN*f3^z#4mCjq=26-`ly$RPn}cBZ6k6cz$2? z_J+4g|G6?|E%oR-*T1vw`@c_c%%6M8J3Fk$e=#_qMPg0JBadEsS!{=RVekTA%S?{k z&((6W=p1sGB#QJcuS!q4K@yF3u@K{ms9(I1S>iuHKk{g#n`y+x7);D4ZyYaK84YGW zh5coAmZw$hTw$6x!^ZO1v4dkK%?1t4fJ>%G%RJ!h3y2f+2ycc$nm)=nVSc_UUNGRY zl3B@IhW|3d^euD*MFTEzf<;obtMwo>%HWO%RikugwM*S|x~j%Q?_$R|0r~V1>!>`6 z$)U(LD)t`6S5jOOgg$>mAT2Z?(yk_89Vk_YK#!ElVTt(JzpsvXT|9`ZKj+x47L&r!)&KRNXW>CWmKh6rFVCoa17|K_`EJ&YD}VEDNs( zHt=&C&0Orv7S?3Ez|gtLCWC5?Q}a`k+s8GlO=|(zOha}ICNm`s=Y;2kxzjGoGkN<0 zkn;t15LR$bNz*_AsfTWtLu6l(COJwr5?$;cNgF045qM14u?9Gv)5pKMu3a#mNQ@mL z#7rcni$?pPTFEUI>V?{DIF&r>iN?8_-^7mt7qn5<1wzw!6;mDV5nId?qmPUw<2@pfGI}RAxN3g(7$tRz`oBy5 zYvJ=5#q4 za{8ds6v!ArRJ<^zM~K(7@-M=Vk(ImxPn%(iC4@$xNq=Yq^ys=Y520W;w+hn)%$LLg zF_W%|*%j7DoH`0}`KX0q6-%Hw`^Ohmn|6yy| zPssk4W_S6L(IV#yiUE0}V^nM^$aunjazwshAF1aZ4{nm(@qHwZcTI3S{T=O_yS}rL zy7@0FUtjEal&*ypPM0i^b6QfTtl(_*x$NE&zQhslZ+)xkN4TeeJI5$hzKW^i@0!^K z-pr%W1A#fpYp>-gd!g>&vvy|xG}xZzoOe!eT1Fv< z@`$^U^;V|ksCvo*vAe($eq@@r>o?2TU}Oy`Gtwk|ax^@U>va6pRZN;RP1?eGSXsX9 zrIEn?FHu(f4KCH!Po>Q%=`y(ke#uy<0X#P$k?2s7*H8#qGIC&yTw@Yia(Jy zW5A_D4B2#jvb&*3h%qSJ?Su@fDvYruM>GM;3kKfmIsK9hdX4{(LqR~HM1^=Ni@>$g zkOQ(HZVar0ytl8w14@;vnt_kIi_QbiZe;YpM3@XwAHUr9riZHAdH(#BZo6GlL}d}Z z(liM$%iNSAHTaOj9$@lNJ#^~|2cj8zLGL>Rk*ZqROa5Bq{Dmk>+y%-t7{|3~V9pYIV}6FPvXL3l z-A<0?UrkAK*&xM^LoZNBGQxs9rqjJ3C=qF8jdYzX-A8}CkK71tj4YmdA0!>xqSJl8 zmX^y-3y&x_`qs%#2|-{R{>3!zCC*J;ltRj-bA#f7H%;B`RHsa|mB#(}0E18(G0Ijv zob{^(Y0icz=VR5W{~$}*O*wY#-olRY2*7Nhm{f|a(fcfDTGWco(|2-JxD>|y<7_1s zYp4}yc_wz^VFe8<4*m1Gn)XWJVqXI^?!6~nOt#vwgIZ@a*rgP+g8~J`*bI7Ma2DMm z-W-7PFBziU)9Yk4U;}Q$^=kg`^%D6J<*Kkc8A=^(58DfAluw@9N4A4)NeHSCL4~7i zD6~v?DRhA?DyLb&!0g=eyg_AupOp(sU0C&Dr>P5TxBEY4oYE|%&h@}G@DZ=mGF@(x z>qW?0%NL|hQEeVO5hVMU4^Obb#hCgZ>qD$?F*&#By>wHx#i!Fx?AZN)*vN?4O9Qn)KwIze$U~ot-Oro~&lDeY*#)k~At+(Y6l^5*9uWQ@#WHMD z7%{+oEj03XIfN2ycUC1aL`YEBoa4NE?$d z-}Fcx-brDman|L49b9n?#*?lD6YwUBcOGOUYcY;nCsuW{9K`nZQl zFKiMaT~t{F(nD$GEuooEK6soi;bQeM(no1$<&CAFgXJh$juIo|!SYksHxGZfx#mR= zxo+itB}@PU{;F@Ou$*Gk)P`4W8tE|mmMt1&vrpUZhWWW;w28bEC zVGli$+1_@qU*EGu;281cxApHc+&?>T-27!w^qU576!~&)lR9?65j%EgFBn0oiDH1b z

    Z{iZ8qG6zcy?m0z5G1?CNOIrYwLYFHB=9aM~}oef?ZT})?+bHek&tNj;1en8cn za=#rhAN}ouxLc-oSaZ2e(hhzXP<0&OmVkIj7r!lfoo9vH*|`tJ%`q2adidSc-V%nnd_@3~g_@eu8Kil6OLsX6}07qbv)!{I(EdBYwch{}&7$6l>#qTEd zj-Rh~^RfV}48;Wug35R8@spCe?0LcgZZ@g>aX98zD2!3YA)D}T4z zRH4;wfl_E>71b1zPLT~y>U}Ue_i&z8uJJbmhp~FVew@2 z*aPIrv$i~91(3=0vjYENJi9@5?C_Y`4XP5|avyLx1PoxP1=R+O3YlS-gy~^gp~q8CWesH~fQZ2Cd6D*&>mJ<4S3IIVLoh%goV8$LJX?-oxs1;~r$ zj$}ag^kD&h(Q64vd1G2V7KdSyYy;UdEx|*zNBPjL0kj5@Lj^M=CuGpaViU&qZMW=c zR_@58tKSn>nKEtLEkMjrfX55sJb^J3nVcTb$HhaTKwA+9%#j~W2o6S$ilvws`OHKd zv%^HD=Id!Mdcbxc)X!+;b+SX`6V-}0F2<LoA*&X z_2Iaz5Uza^{eIGSHzw;h9uJE>K6pF)J_{H8x zj=a(qUCLP`y*f)PZ;1I&tr(D}OvgFn>Z}_6Dl4gkSvVpVeob}FtTJHcM`V<}fBXGX z-q`nE{&78#p+|5bl9S73_xf55o2n9cV=9GDMDVTbXqe6T+4$7ed;+Yc_}_fpR)WF z4Fmf-%qFKzZgIi+;z0+EGA?2R_lVE>8I9l_m&)!~JD6%_H0%`z<_(@k^bQkXn1X^b z{+3K~?AX<9Fq#`HDF$dlcT%yq-a!Ww?Ll?m$|(?AN4jA=J^c?_&Kdb_XUK+qKWYemoS2hBb+mJ&BRFi+&AqE1s z`PQOv8v{t5Geep|Bg@O?K{Cp$IC+V~Qr-*$a2!O+T_oF)>^B;p3W_PE$POyDI&d*m zMI=JL2yWf$N$T;Mjv8QzTa9qFkG=_vQyVE3mO0M`TN}3K|Qw^ijnU`6mvJZ zZ3gY;9B5@#OJ+K*mLg@e*~YXlYo4=1iqCJCz5RTj){gTC%=oleOo&`L&C)_n6Mq+o zz9&c2fGDhc^{9^f&c=JB((JTQ-%H&cJpJRV+rJ6;!q z!#g6iKbvA;hRQ%LCL|r*J$pOpPGzKMl z3g{9#jdP2R7nDRU_9+%FpLs~6alY@p*&mN~N8NsP>nru%dA{v_AXX`9;tzn}TMeX} zX`Jyo9*yVwKISnPZ89*uHm0|)f5`l|c$7tl{eCeYov2Ob4KJfN{U4$7|K=4YRP6h2F6|^U=&eIK1Fh=*m$RM_bV=UeCs*)T^~T6S|h!{88!r! zO9a~>VdY^^4siS78I+#$vO%==D~}q^!ypZBrN=UPgGafnUwlhg;`)_vlM|}@*a}9L z zjlP>e=ClVmb_yfn1W>}><$60*)iY(k4VB`b9ywMxdirsjU1{%+4KYQAnOB;`D~uVG zoflmtP24lW!my?A)x0PMMqs5jyoQgfQCocpNB`?tywcG-u+dd^UroC$>G5T&ehV~bLwmkiR2WAf zR$Cl%S$+$mbD*ja^kr!J)Q5q8{>w$ba^HpXR&(P7g}}bkBt1kwjOy`x7+5YIh&o3$ zPw5js95;i`^M-EBD0yDTv}v}fPKy~ilx$@iBzh%sfNV3UkK08rrLvR_vI@VO z(rm#}DmNgHTNBdjUh6(VT1dJ-ob78J=i9RQaaI$L{hH^UNq^d9N|a)T0vecRAZEEp zSM{K};ieB{POKNAA03MzVevp(uOTZ)*l%<@_S`6 z16qXb$L@r=1M6XWxF_@2{`r5r8T-vrhSNa>$=a~~tB%freV+dfGiXt!{_yxshr z@kJjR0wBA8=_(~RUYY>NvdJSiZyr!gKSl0RvCYy{d9PR_QY+5R&7xQPsFuk)#AoN$ z1*-bw)w34M_H%neo1lU?!~eKb6VTG$^lyU1Ueygyv~7}QEA#wP=3v?2E=ikmaMm_g zwE}M=!cygnf-9ZNBw&#G*Gx4t~G^UH9Wq#5m;8Pq6I+PLr&j?iZ@=hHAX- zhAx=8ldF{v%5pi|Twx(r8<-B#9%w0tHCNjuXSu(zj-@hZ!Nz`lHus(CnFjRDyOg(* z$Zl;@Z7cuAc1Z91N80Se@n=(z8mGi0^ky-pH0t!4Ih z*J5rW(NL|3<>dSTeMcbOqx19QvtkU!fxRO!B($y`2|QdKTh9?$Z_fxo_5Em zQy%A`=&r97jhuDCV@2b({$Jjd8o+bpd*8o6*1R;}*=YoxtrP=gtec?7h}Z0f@(L(* zxz4k0?iDC1s&u|j4$<}R{@3QW+M+)P%hJie{oOk;SL7!__~&y_1j4dLE26>7_>B=H zW_q*Ux_XKMZS%cU>}K~GelO{zf%*#0?qVp{!#kcYi-eM$Z~RnTO`(mWb*dR-bnf)27P;hfuPmnlL79-Z!HzHys|-3 z$7_r3Wp`! z*t~pJbND!i$Nc)_g-Sx$zu>6<6a$8?{@eM_$TB|v+7vXfqq zUhAcs!Og&Gjw?}R;E#3)>%BLDnzmZ8+GmaHc!9_UFRVbzrsHw8We3{1g?*h012&?M z{<4~Ev14r18eyY^Vu~qJK*g2``y#HrmK+KCU(=#|d7f?j_W zMtwK>)(VS-U%|feS44$}Tc&sNdlXs)mX7sCb}O-ElT#me6;sdA%JKQ$$n{X9y3@H^ z8Rxvz+CgB8Q1dK%^Z1^1VH5kUfK{dxUTiFh9q$*I;dsN+G>unjaGqFo+5OI}1D+bM z5B3v5vbg z3n2e$iDQkhl_`;xP1z>i>$lXgfwv80IFJ>th`T}3CDR1zrPH^=6IW(7Np0`qjs*)~ zA4VTQD_%A4r0oCEXDZX!$$_Tes1uQ!y=ZKhKsy-5R4{DJ-n< zS#0_+4~w~>D=SU6jO?~{(99-K^&1`H9C`o}lL{rf7TqpC zL}zkV3cA45By!ROXQJ0mDWTOqodiGmq4+X6>WuwA^0@X|<15Rf(gY>kKHnvd!`Hz; z@fx@pK6&mv(xTaEf~AfJ$;H2mCvk!!ygc{*h%@ph&s>rpCu;XhrT&vzqVcKK^44gz z?ztz1%~GU2c1mn}@EDBdkv3X5=yJp_4hsw(@t@14{5M(YNFZ;1M48qmirI*3q1a6E zjnH}+tlbciOXA&jL56}{dTA7JWpQzf3|FvC+^3=<^XvC{uNvS{74+XrNWn|9U^;As^1T$Zn<5pE(C?o?7fMpO8pTrNz zX)uKVY)qfQJpgokYJ{pTx=-Fo>+^6`L*HtiUK_YuoEerpwcOiqv5tR;ZgD#SiVlsk zGGQ*=D^3lA8YSgDdS}2cuC?4|wqF7ZEYZ`59>bMi|F)yslsC?9r#3Sjab0u@6gqa% zaZzOvT6uw@EF@l_Rt`X+$AsCq$8X-kaemx>q7GOYyGdC)CvE$q!KS=kbik8bcO*-V zl1m0C<~~L4VDzGfj+=#*ZUA4)`<3&_Mn?jvBqRKtJc^|jYyoUq zmcXXvLtZJj&}E%xwIWe;k-vD>3i-x>J_w9mc5ex}JT9Gw1@r9oZ@HXE1tK8M#x5%XUnMIu)W;Jd8T! z*#(_XvUq;M`k?HZ>Nw-V# z1uA_33LK=$aqYZm>Z+Mo2|I!K27$~}ijd!Z_$SqW{_>CS|Htpeiz#LyMdBvhYcl;> zS&I{+wYQK;&8xj_xE98*;*dr61*7RbV8 z_^kFxMcG!h5^DocB(W@{0H9+_olJIE83jV7gKVYqhJV|yHcf0WlMxUn$PnRs%~B#^ zx~g?foqwflD`W;_1gqj^HUHC@H@jl$-#!Lw84&T&%9A)vbS=}qgpc}XUp|M)*1QKpd?yN%D82@$2r6K1T1 zWy=A{R!Iw|Po_URFDwpM6@@49`bh@l(Wz2Id5Sx;Zp?yGpc}lkA%n8hJhc+o=J*i9 z%>#?m9H!b z)>)n+UsTUo1x-}V{3NG#VK-C_Re=s-8fOhz%1Pq&@|JS+><$HjWDqY&n4!WT&?R!i zwTH&v_BQZS4LOuSU>V#zQ4c7Os9GQiM-zU8r$SQ)Md=C+4{MzgIW52k@sfmyIt$ic zws7a^p!XJ0iKtmI{1asbp*&2N+^A?}M90H*;%g3S|Gj>y(FS=0VJt1niz$$o<` z6~FA~m=dMht!*+xii$ZA=*j4Wyw!>d*)3XMZH*D(Rm@&xzsnj-r793?Wwi2U-*$Pv zz~+l;zz$pJ>SobrVa`hTSV4EPgzw8SP4G378JiD#0ibSsUI;%naH<0KflBEu*w~wh zmu>uREPVCG4^6-syH{_w;r$YU$ivkTlKa(p@&QSHY4)pk7zOaQPz-RDq*JlzL6m!= zeo~&IK)g5L1UUdM%IBO*q3x0tuKA+epd{W#4Cb7BVe_ z`Un>t51-Yeb?yz6CJRk-8tisvF_Wc$GfTeUmJ_bM&dMN+qE;Z&HZIuW!!FAppn~gS z%r;1sm$;XMJBr?AO~?kxjnE8ma8)_5_n1r9gw$|Xx?!@z3V~jMF*#BNyP}dS;QH#j z#O+J#n8~r>!2}-0r&a+3JJa_w-uwK&H&~cIb^qxK`TV7^FxQPN%sGlVLjj|7EY=R( zetRPUrv?7a4NCEu-!JKl$Wq>LgMv#)Vo3-r3&*o25IliKvLvS=hdg;U-AT4ea3>*~ z-Z2$-xYdgFKTi17N9w<8i}~pIJG;2$!j;UB!{>^$sT$Z=U(OkFc<7c*V}dzU%>8u8 z0b8rEwW{N%n0r$$#E3eZ31YrX(8bfomGYjB)<$IYj ze*?S@y}qW3WU+&n{W|nOL_Tr{s+?jV>%16pJjCm`u)O7JoSXO~k6WT|1gD2xiC+EP zlHaVFx99!V=tF?0xM`P|Mt-?Cn|nv{hnmN7SA}m?R*-8m)rxZOK1lvS z%cNQYft!562^xr9^u;K!d4>UD4Ef;nxaU>o4!HC{%x2ZpxG2?j;F(#)WQO75 z%!(MX7CQ8{!z#egbz**oD|`6cuS_%a*z7UbvAHpm&sWDUb#IFvaM|phqb!x3AjRA- z`I*j31O?$%1cHS_XCYoJ?9F=3GESfRorhr(dhf)fR+47Ni4BmpJ0h_mpJJdkKZ}a( zlcQwY2`IBp@wqHj?F(p@9wjL3Ttk15O|SDTpcgoOpu!yAKG_KmOlMGG`C|!Ww+XjnM>qj?g;*Q__@BYHA#E=EajnDm3=niZzJ4QA*Ol6hZn zaj!|Y^@}WnxJ2EnEDB%a`cSO9c-^ISPKL-fuY1H_`1p|J5Ila))(5PHzyxH1QnjrIl{7vFa2rK>qLu7eKx_Fs6fakYNHh9&feEzVOvt^HMy9CY=U z7$zU#ZQ+L`iQNjwjxhs7AS0%#Op4h|ku)kcOPR+x1Z5<+)qWmwz*RK?TY(xLcjQam zZ~Ax9T~Lz(#PCUQOkyT6 zu#3)88ggPV^p`MWV<1pC+;H0l%U_^{!QkGhkS+-l_)Y*WbR)tV98Zsh?RVOPbO&ew&BoaBg!8dc*Vv0#jiADEysR_@WH zdC9wP{May~WKsj)BgfgzD0XZvzBIzm1&V2>NE;Q4WG+ZosSk-HG8+SKzk+$gs*{k~ zf`R_KVvL_w(Z3YYA_>_BMjSJ61H8;#D#x#}1ZcUDv>h5^v3tp1vNjl#fO6_@H z%GZDUD6fI1=9UFi1|XGkVMJlXmD%ykGI@gMgc&2Y1OB{$vetc8_J{iCxSBhrD`YnA ziX8{p%q;Xv-EX_Z^H2?~lMiYE+&=DVpB8DVe9*VcAJ5f_`(y%10$|u#A<0@Wv7n#7 zeEi;Z)BJzCH63O$`fmg`@vpdD5Nr;<;D>=UwPHZN*dd`u9^S^e>%+6sFMNOb!m95O>+%gpnbK`SY#6)$(LqEe6S`v-y*LtbN1^B=#E*+q?GXzcnqHwcFK{nIhN} zQPrGEj{(<2L z$s!<2vY7!KABb(2;>`}k@AuqYX__);CJSweN0acjPF-r(yo&~Psny(VQJwrXq#w3< zQ5G?WE|qQJ_IQ*;wnguMb&1C*;h@9*S96@2=?wowP6O!(ydSCNc1pCIt3YTu;93(> z;oK|6UaUP%;~k#IrpL-ghSl)cZ_F$29p9gu`X_8wM0UJCZN|;25uOyHPg?H%*8wgh z+Ui&gD?Lzsd}1m|Mgq}tbdP+>_Q#7|;vbr3Qrhj*V3;Q*LATBX->pFvCHc@0F72C=0cdij; z!yK0!@qkYEUFW;qwVrcp+6LEVXg7v>dSeTV$vzt zkKu3jIw8`+#$zFO{fq;F2P8JrA7KrE$8*W>!MMkI)DBoVCDgyXlj-x#nsWLXRN6}v z6%j4b0*@kQXrEB74O2CRoDW$Yb=*^x=iBawcTrIM5W$?LdN=I^#E{2`s8La1&5Jd^ zi^%()VbA81ygk=QqaANPUomn;+bE`$A}2v{QFeH?2FRT@O)YX$PrE+rKm>%wLNA44 zm3)&E)W{>TX0~!C2Z^~a$>V@w3D>kBIFHX2xuNh`O$e~1N#ecZfnotRZ@RqdIhDTp z66MMuRk}|vFnhxTophyd4}BGOe&d3Bz^~owR^hG6Quhof!?(P`evDulwmoC{X%VJanV(5_H~!^Mrsn5`s#P`P*OvMxfuKycY|vqa zyq7-Z+3hjtQ1AW71M?+uGf;(`{kZiQ_jn$?T8(RTg?!3evFloI=T5gIToxg z5DYndvQKncJiuL{Oqm`pXyX8l;WBx;T&+lpy5qkLm^VM&Om)TFjoC(N6iAhd)G)a; zCQYf%NS42%nLFh0S+f+)5H6*yxWPF_%VKUOvT)>|-^{#aX>CsX#oGqQ=HRc-eMuVZ z*sGGL%{7&bC-bYCTZ;4}^U@T%SmpY=fNRG0B zdymvBuXVU!R$!7OmoQAM1qDD?lh{mN=;%TjbL?>Vop%xb1L?e%kl;o(t1gk z1Xtm`?$%lM`C%V#n+;J5c4{{m-6n zR^dP+P?8OBE5yAzs-L##q!alQ>Z;!?{MQ$~W_lYJ>QQHJ3R7vri)Q4|MGRyy zOyoo{3@VGDI^1{yf3kw=q+8$nyT8shn34?nFY-wRyPSPHPF(oH$RZw~n0*wfg~SC{ zB--zbI6F6kUPC1~8lNwAMAG@T=!a1~o@$?#kgMYUsakn`BqkRR%GQSDdUv|w^+H}- zbh^*t=u^Tj2>-2{KIohL>Yzh$WQM4K1Dey4Vqqba8-4j&BV9eMEqb>inYTE)fpdM_ zY#tV&?m1?0)Z61ksA^7thBn~n-cNjvl6ZDFvg7bgrV)i9iZwA8P9x3vBBDq#7Z0*9fUX}QU z9NX**(D0TzPmm8_*?S#nVxKuoR=de=+PB;CpR_3doiBQ2#}VbKFeHA{No^e_2Le|* z=ef5_PI;EeRah^WDsPrza-3c}9<i_75r1(r_#8j2-+Y>^btJ$wJ%VUy8^yrvy_t$F;Vhqx8HT;2cDkyL ze~8AJ8u$xbR9N1NcBNf{d$d@6Q_s=L*G25)AhFI7 z*|9#RYu*=X`_K$`7=jidtrNSs;fA_l6Ep=HBh3={8b2{HXXfF1{M$iF7lYlX`tG|Y z$qIIB6+2$@<{6p4bc)$P!T7{h(TR|wguDccqnFE@_z%Sw#hrXS?iQYQ?~7RCh9bnL z!Vh{+9E2>v&4^iNVn73*G9*xM7a7p8g}?3+NpvKIMz2*(G3gZ90EuT1Bh~XYVi*-w zNvr*;{g8})zy+0g%49vj-FRj8pez}x)&IYX%4Y3nzd_Vu>ceu=EPcD(kuj5Z-%Ecj z>kt=+>v>yrxdV{A7`@5qj8N-dLidvnc_)8sEbD29s-Fr`edIP! z;()wb+%%e?DTXzeY_+H{LDQpuk=%IE6O$UnsH@l+s#f4SR+ZwTg@a|%)Zl96kb~N1 zRoErr6&U)HUY)K(4ye$O2F%u*-SuKlW#RSSpv;2)Yc5?NC}qwG^;%N7^hVzT!39ny ze`i44EbQXA^O`Woxo(&AMt1TO0^_34QNbJd8L(Q5*b6IXjaQFfzeKfz`?;_z0v%zN zskZEaGQurkd1!)#XP&aMI+H)#ko1d4gWIw`SC&aC?Km;1#mJvJNHIXjP)EgL7InJ> z!+EMY$pXP%WF8|+>=;PdMnGCmF)0*DreZfKS`@kTP68x9 zg}hGJ?(kf1IK<#ii{hjZYf@$mG+yJh4VH}I zSWtrVk0pJNrwuEVOzMcf{WsHO#;4=8Od)&mCSXG-#S~ZkLzAq$pP{t0kM+nH2ktGm^bVCZ&jCApDYxiAR0% zE}bKbg~!dXXMBRjBCreM(m7Bt7B3iZQ7h299dOk+Cvpb3OE@ZwyX*r;ry2NF+9d-n zUGhUj~Ev%dJ{SN2xlq%ctRAi=9xlDMD6E8>-WV^-- z2Bl}9NLAIM&?GJbTs23nz^iT1XPE_4@4Ieut$^g5 z!-8D_o19L%UFRKFwnd*I`X?5;d=C5``qsxuu0+FAE1D!%oDT$Eh}1&iYjR{gZ-wia zVK-Jdnao#aPsa+fK0%@8Mc--6cHb&dA<3KSE5T`cn`>EwE_ZWSDNNNX-Q!=w?bZo4 z?^PypY~_y_kHMHm+_(#yoUxjcCi`TsEjCSUVzbj|$G)$bq^IPm<=#4u6;z(Xy~UpY z1vz&Zc}wtP6M#IY)3=8re(|EqOf8x+=zv?tH{7uPP_4k+no}WKML|GuIPUiCc{O#) zm`(k2-ruOVEg9@#+dpatHoC~pSfqXscFQ!g+6*;4swn_@8h22p>OvoHwob=7^)<-3 z*Kg#Og@eZAFk|PCL#iB!)S9J*oEqU}Z~;&jr(3pFVmn*TW=MvI#^|SA42j+5%*HfN zJt{Z&vFnoq&yaO5jUQWXvr8enIy~41-3;W0)RsW>MUgVx)OT8Nv$~G)M)Dzsj=hTMnE@oD}1kPk%Iv!gf z6B6f0)JAVOiDFh!WH}YvF8NxPFS_N_O{;bV^wRKZrI{0J#|^);MeowKuWi)cS($VD zjV6Spz89i2K;lU3U)PdS$I-$mc5C}clM#ORQA{mGs;SszK!(&7eUH9BD+7wrwsE^; zrNKueZP5oJvO(f-6@wwPfxj<#bK$#pW3;oXXB{M2)Rtdec)j8Khu_Y8y(8wMoF8|@ zw0hJ5$MGhrH>UhMZ886yNfn8&xL^`yytGsvH*Epuh!QoDhbxYOtzlL{`bguYMFeo{d5~UfG!0D10L0jjZVFXT(E@L(4hutkP5I*9i9pBzQs& zCEh)zsFQ#N4Z9to3F(&MjT)i8cBv(#9V(f$-Uq!^r^0PPi(nlnvA~FhP3J#YC=WGQ zoh?3D>qxO3XDxkhWMOJ326lRO1Kry@9b(|i>H}7!b2G2ZF6CgX{z7D$e>S&O`iOT` zS{RY;b3uHbi~NuagD*`hG(uj^geQvUgHLIUDB`U z%OEsW53F$~X>1EWCF3?My1}syB#bPMg@gX=X6CKSX!FsBbp?*$1^@}2Vz!Z;b_}57 zMgXd#7!5`CP_YGqt*|wmDo^3IK)6)x(*$d>g8{iP>fJzJjJt-QZR^mv;*cX9ys2Z;=7D%Cpl&$np}XlCKJX>OM`X~K zAV1R@AhH33tdbVR3PO`rr|wKQT{YS5)}onp&?SP^$tgaVKU+gz2I^7N?6@i2Axq+^ zE_q|pif#(33tSSG$JrZD6R@0teIdkG4PEJx;*-St$Qp<+^VJ?7P|S2{h0CmfGO6TU za*%7VH4gp`@g%{HZ4IoCM^sd&Qp_5PtfpcUMaYhhS^fQztzK%yhBcQwvh&c475m{-yyhXon*7XT);;!D|HKN|wbVyqasB*Q26f|YZ92a>;8 zdp}62p$kSo2|U{nCFaN368xBd+IDwtB(~p6F2YDMDA_=yT1S~B9p_Cf#CcB4G9XLhs8mS26oZl=NOn0F~} zCCO(O)VJS|BWX19QT9*_OitxgEJ|8tDfjrNI_F9-v!Df5)P)giL$pyB#kq8;f42wZ zOKQBX(i-Rc;JhS?dV!W03&hrjNEaZpltj^qZlI&BeE2UZ8310VwP zqm zNH!I_-Ss5MRo<2L!Kx-V{3zYY$5qV2s7A%0!=uoP!lMjqc{*hHr};nP4RAHi2T7X$ z%Bkg%-Lk4VWf4eOT@JE8@l(FE884G%i3O%(JmQ@7xLW!ED^o)Wl_&GQS!|FPrBy>j zBqR8vk3PbaN{=HvI2;PxtSH2TiE~w0nh;A7Fz$B2lV+<;m6O?IE7@^!q#2#%q93+F zx!4lrY2hAa2k6CDfwp`f8hnl; z-~0Xrvc?f8?MFtVwo(ibv2LPbH6f2?YGryElDOdQZ+D2XYzmS!USI7~;(3nrN~`(Z zd|(0yIH8OaWYMd+83FZ@eSs73J!Cr;`a!Y&yEpBT)mYfC?N$6FZ__u6%&kV)%4{HA z{Hi(SewzX|zkNa6E$i`bi{2o~_SawA@uvRT1fbi1DwZHO#iGs5OYVM_^izWg5qbP^ z5^1vIWy?(?+jfy+&Qs(p727(eI9wO^!bI4eP)J-6kRgISmv+f@VVY}l!~#LS=rS

    ?byUe3XDs*I6}6JL+Ee|`^&i#8@2!H1ZbW7Zb=F~)<(k31A+SPYC=wV<_94wLK^2z1oRN{ zP`rxC3{wq1lcl`&+Vl1XS%T*AY-NRmX9A9uWt_bI^tn_g1H=w}tM0F4ksU*9s}aOf zC?=UAiBxRAOCR@3*)313%0NER`8y*MK-Qop;5_7&Oq>i#<1g|3lNTncdn9GqNc80uZA;FYIwGq=x@%mmt62M4AcxU|I+Y+l>4( zdSwNQ$;bZV(|vOc{)Nlk##FM)j{S>PBmd$M#Q=jrJr#RFutl^is#;M7s-ZVt*(%BN zOO=i*g1BJN6_7uAJl(W{btXbUGpCII%Wu3h@>$) zrq2M&GjFj%67~0Qb=JPk7~N!#o=tKLk}kxmR9q|k>P>N&?@g^KyFW3~e086QhvGM@g& z`mA}FEGWD2dx>G*5#APlNRrs43)pYSkQ5uuQkfJ3)#hnb>=9s%%~n33OJq4ppjL8e zi#{UDVK#EQyf!)ILq_acr>&AZk|MVw((JH>Ah*%ZR9GleJ=W1NJZA}bOy{0)exyZ@ zFRU0^WSa2&lI-Z2Nsqq3X>n_dzUAIXtGNYYOo|?Gxlh#GT<0XW1z~5r<6U!=KSMq6dye<2~j^OiaX^GJ`G2nD&Q%C#jAE z^od906=zcn#HKQ+*h0zYvgHJ`ZMM^i%n9Lz$R*0NPHoT z(IxrH<10w9-szlgUG42rkT!OEV%y-`54`jK@Xf+WSd$l5osH)&Jhh7Za z_BaQj%FqfF*g$@KLyGi(>qXexsR=<{umKkvd%fepiA{m>!>>Nd4KmC-AAh2)Cfi4= zMYrR`)`LcqkGS+}p%_%)9UeL=g^0uyM z?RWt4)O5fU6qZ)***W`d8497rN<&^ zGyFNK>&i-LKRE`}*&}>CtHSFkys^FT1dQh4HHup{$@5Paa{k=_xj!}kDu{e($B!y-JY`~>MtXlNCf%K}?H;Fgo&>PvU+zX=K$kGi-6w6=HFu$=c_&tFu zJNUiv@q>uh4DkE2Cc2&!vJ3dw@icnG$Uat63{(hIQn8tWB5~H-xZwR-SLwS>aJZb~@{99QxpD#UklJ7uDC2gs4<`PFMvg_r5934Z8rzO8eP zGWdmG6`39dB^kPmq*OWD3bmqw)bWu8aGh%#Gs4o3`PZ}HLH2Lff=;c_gx%NJZWs4I zzxt!X@l>qZv3)VaR*GU%J#@YZ|Am&_K_e&7HH1ba;=5<;cot&WG=E=vfGhlDZe*Dv482k|JkACn2 z((XvWxf~HJzCkfpDbfWMeZa|<8=#eKRIcM5ipk-oNm|13T86SR@}4kGuv^>`-b>?^ zPcO^sV(!Wo27h*0{^?y=Cx2sL6JOu#D1GgC z!||Fy>m4`Y7d|;hx6WzuvrG{O1{8a25@rL{_`vUaRzv%hoi*o6F0UI1zgdXb|Jd43+ z9zCO2j!9u>T^P!Ms7J1kj+4O`%)o3iLRliktf0tJDmL9y>x;@85JnL+ihHF0E1s<- z1@jh$T3h44pW1jA=|(@)r4$wdkBg`x180AqvTSEP26Sc7!a$%Lf4sjY>Z|= zYI424Ynq5BF+D{ zj#knpfOvW+#{!Y3d$Il%b`bf`ccapy40DX%!W9K%w;g*er;O&4Mv6H=k$qHb4V?%c z1Nj5moNS1?u>wx_oQ68;^PT}mtuxO~B4=K()QkH@Xt_o)S159sid{S9 zyfey~9aErr+QH4_;BFz#X-Q7)x@~Y3%I9>#rl*FdBf&hT7?5KkMn6&Q6P*@k)7xG3 zyPPYTLBE{=pn#_KSrvBNc}h|$}t?waYU3=#goRcgH8+G{=Z&d5z}34xU!=-R@fw~l`ZehgpE7Zn39ToMl- zv2Lb};Y(+oK^WIJc+`i@55g#}tNljJ{{79+xEI~0z9@!Rv?HeQ6bLSGs^%n$nq;?m zN8}pbV#fkHJ*ti_o_Y{eo;OUr5{g&Rn2c|(85I#vUmOc|pYFx_S6E?}%DK7yUDE=7 zHo3@lyj^Lg+`nCt8L>7}74P2#sbPKcWD(?VHA^u;5aW8f({x2Q4Z%V-QM~bt;~0=M zo{zKm*`Z%&e_vskaH5a?vYKppX`JL*qj|H0VqisI0MS0525z~mU))F^5?4m{MO0BO z;mve~xJmlRua|Cm1GSMl`Cq->AubW!m2DB77hZb}bm+OsJWP~Ak6C|x`IHP;-Sv_a z9_2DM*B0hz448S^EwLC0mN)!?_os=D!pezKgM9kJs% z1}Jw&>`kRn%zBEXP_glIy1be|@vMj5s$5F0&u9_D8YxaN;F2mYl_y6ZBs;j6@;i<& z4C}$O5K6;Ff_496eW2mb|Bt;lfotl_`p3QEy(A9}*$5<8pdtYTv9cH{;9xuLOlO~Y zXLk1T}|l=Rf@nBzJj&_xa{L=Q-zlzDIarG(|2>3=WQBC@gM5j(U3N4TzB(3`v$E zt^6L{;(-+PGd|dyCYiYt)E!X6bg7D@XUJ*~?HbRE({YbhQicPD69;U!Spg%PVj&g1o{C>- z>colFBUp%iF{B2Xo$Yn4)+93vf~qxjq7Gnw?Q+}ZnJYab*1IheZyhH&Dn1CYSO?Bo z9m^Bk2+_+v|Kn)(xXv5kAsl;La|#L7DPe0{5Ya*JnA@c)^S%>g1nwSGd*FFI>)^Fy z$n^3=&5%1C`|{NE+PEl-_4v`AtvRIJiT5k@R+i!r#a2@g0glfLeDvxeaRXgVA11A; z^K{-Q4eIt+w^!`mEJsx(Amm>7kI z%y;aI%%kfRi$W{t9jlS(3srY)Z|di?Oj#SnMnnJ-5T8g6Sc zOaz!ZANkhXwid=|$qO7d1N{z2we-AVoRTcf7axN}bpcciVTBJaLVEhJvNg1ajDE`P zo*RDHuzT21U&sp^6X$&Wr|Uc{7Umb1$UL%&-|p9mJ(p4|dz4GDpih-W#h+&vctOB~ zM#(xz2>Y(rWJPoe+mvm}YY=5!HtD?EiRpQ-;D>b_#tZiyqDbp z_h;GKj~6=%uc!9~U67mUQXN51Qo>XO>(r~FV22A<2+Lhhc{n8H$ocK(g$>S8#{R?) zrp6w&vp+<#H0!G-eR3@MMU5|1%7Ery*qBm3y&$3iC_qJ0%(O$!9Mu#0+&kzCkg`AJ z;plH4ucte7X&?IsYBeH1Tkf<~hlvaRl0Su607Ku*#-^^a|xNbs)fMJ z9n+?DPjEA0EDy!C1L9@;uv{Km`CHqTB2MeG7Ft0ji(=PNFbnZm_AwYU(;A`vBcHB< zEf9FD0ezkjO~Y~yGBpV;gS@#;AD4cT`V3iz_<=-u@ z6J-SMVQT#f1UE@<__grIU}Oq4cw6hNhqLhLB}wkoI?-b0zA(jqF%Ts|A3>BhyG>IZ z)k&vA2=8<9#RyA0y98T2YF*#ew!FEC&hbux{Z%LFfeubB;pLI($J}lf4cWc&GS(>1ibkZx}(`!i& z^k%yxc@D)p{uVh!W}7v9TCCTJcSjt#!E-Pn3#jX=zdcYk$K{A~H8l*@#^#GlB&pEB zt4`1uSsA>N9Z%<%r^#}xofJEQoQ7s7V}j5-8=rI)rtL`g;u7hn_}-%xpQ#4L0*PT3DypxQf=ywNV8NBvNhAV$4buoCI2nmHws1L}uResmPL_E1ed zj!9{rYty*lgBLg^6#ikx^fxW*)!YhE8>t!2opxd_%N45)SUts_rpO5@{xfBfw9g}n z?GJwtZcaLO(75^D5RuH3NVdv>LK$NY?e)ys8;d(Ltbi;EtdumWPq10Bxa02=?2P`{ zTc^e=usJ!F)MJB-I#*+45_^~+95FChDY{2r_ueYbiCF*DgxEUQhkkt_ohsz<=0sG6 z)d)|8=1nf1+a0u=`r8Ps7hjxN@8#`%XJudPdc%#gYvJFFwKX^|f`k0wns^+*drc=0_*(u8m(hCgIoyreMsSqi;Bba8pp7~Q3@dd|I132gaH#qA5oIySB zqEp>pHvyqJ?t!v&@-!$xuei|Ja(y5t_jis zThCOls;yHl3ThFx%aWP8umK1ornn_Dz4AROY-D^vUKds^G=S``UT&CT1Rl>WQl}iz z;x>wJ#V}u#;SY`ETT=e3-a1?MuG4OgIGDOEBGh1Rj$0MgrRoT;ktc_B&1q2RHRt48 zqcMszq6r(x!!+Mx|5T6sGe&YJKLeY8$W>9j}6mTd?eF6qI-B@V?y7eS6;DYc5{%WjdY>f;Fc->2tpP&1PQTs zB_FD~p?63UJAYAVj?~;KV$indYnnk8eO$`9cus`b;Ew3g^1F%OL^!RNGBrB9*S3}L z(_3IC4y$mq7hVEN{vQDkDpiOsH3nC@qROLk{vNu@_92)b`o`jFZUkSNP!RZAi|e}g zZ4U*xJzT8Ki5=7xR{5gODYlm)52*Nyl8chNL+!+LEA)WeQz@yleGq$^7MJxV)g>O#OfXhc@hc^z^0s_Bq*;Vfa;)~=i z$);Dt;+-akwFLzI+ZFpvl6SlON>JLnNNWC6Gp)v)y+VlA`t4w@zYZdK+XQQ1yydWI zH1T#eG5OwIf)3gw?{bk8iBa@yfw&U5J8dWw-sh1jG#@%c>J(dG1f=V%Wfq9jRhdC~ zaH;k=OdGu}YP|Qn2VTsHJM(fmzfv;rYW>vDZ{Rdc9)u+RR$!a4M&ad^#b ze})OXOkCOX`90fMEgypK^G9>I|I(+j`}9JvDpzEUaRcAG473OAc%H*97G9O73Nw69 zNT9xTe6(83c(5BvW z43Lcfsp)I^z}@^P@DqvluB0cbEiyrD68E|s-79Zb-~zv2Sa+@ zy9A#qA;RL>J>@fbr{lW#@tc3MFK+$QF?NNYDOq}G$*ngmXxSFP+$YuiB6Cjc4PUmx zRvpEjqR4S7KAS$NE@YBeQ0{RpV^HG_diz1^Kd6w-k(wy(_ef#jfJ?q6-+O~+hhkGq zTR65vM@{bmmqI2-*dp3WSFq-r@G9*V8V~5ysBZ_=kon%%pwrO>c}3v9S4_>j%x72l zmJ{td-(J$8z*bzXD)V98%i3Z{k|Ujl-R*4l>`yTkNVz{amQMD( zFpz4pg49upJxq}rDn3DsoQAReiNHi@SM#Xabb&#Qm;pLfK3x&BDyotm`2Oy>D9CEQ zf**FQC)A-xvi>Mu?;fprSY}(S;>M6sX;w1JA((-p7QD>@|E(%LDzcYD(OAnWJd4SBHAo#WvU@M zoD~68kzUI*Ju}y9vV)*#;hw9Jen{FCIY7U}%dy+ufD5Xg)+_cwPe$A&W8+XPlfhVX zrQ}%jBcEd05dBc8-At;cJ_yf=SRB#ku0qV`qKK!Go~-so48ha$U#wsd9}o*2npMy`45%d^i6m@7jXn>1gru zx7AMZhyGx6MnrR5CQ~X%6)a@u;h_y)i+!OUMcyK+hq{(4vfgnkf#N441YiBJ1}x`z z+(<5(ME%V+ZpYCKFI!q=GQ3#5tli>qky!zK@v!IgIGT#Vf1$l5WAOU%`px~IhYOE6 zt;tyZLDS~R7FQ+kH->p+-6UwKJ9LfTLb3T2*+j+fiP0VsWP|swcR!|F;8hvCiA{jw zvYZH9r?H=Nv0&{q^EMJ%Li-vb>lN5o=z_}p8rEnB|?v5K&MV!9Mvbd z6OMAp<0T+90^~XnuU`z!54obr-=m&)HTF`0k>rI!#c9y{?c~&Jx4b)o_D)F->xOdF z<~WpUZ&I97USkSiUxUVNk6@X2_sE!RUp!^IUX1`Er^(WDq+>>Z_}W)>YE%k>jz&wS9hbweT34eS zn>jkpof;8ZhOUctVUqtvywEbSq;2hct`_Iy=&RN5lf{$35gBUO(yLbY-+>$r!^`Omla>2nU?pcDL_zSH`wIRWR@wu;J5>xgh@ zBo9*c+MvNuRS4?PX>*vIh=kBgrWN`rnoq-_Ljp|wAi26c>I?*PF37P>nxipoFMx+y z49jZ=LDxJiF91)t`@ygO+m@64f^ZTz7~Fe*gbf^01vh5vgBv4rrB$Q^+?EPinIKoX znVe(BEzoC~u5^l!|M~6jwg2bs-~9Yf|0`Qcv5P2@Fy5$n3X*G1`}{~^SyxtMUHv6V z8je8cygM=IQ&N`TZ;aazo9*pFLubM!*eVYhaZSzIKG|czRZzI`ZZ4dFqyNIUxLP$&bHdaYAm+s<}%(;zts2;xt~H)splH z#h#_0Yj35W zpgRDsPY`d7#ujARbR}J{-WXh`tntI;D-}H3=D0gSOWlvs=Ln9DS0VF$iKq35)^`*L zcF8b^ca*LNMFx~CJ3RLRuVCJnV{4eDYNUsFpwef;kH0eCf~|7D(@N6j#Dv==R@ER6 zDYly;_o?`NO(l4$$ze5ffK0pFr`cz*0_h&5(j+NVYBZ`_fiig%uBZeuZnhNm&se;= zUz5aQsjBuc0aorU_okS`2d2MC5;lN1wdDLdd| z5Vo15!r-G(-u5hT0eyq&XREDJ~x}jic^h=|IED;ZzGyJ8zoTLej^xI#5 z-cyW7g)<2}i!cd*9mOBJ^s5$nt1(modR@}FlWG?7dvHZNdJhM1QkiUq;KJSzSZC>g4xk1)rAaW}qJ ziqwgLDeGqMbj=kk@xeSJ_Lo>U8`?s{yCd>Cg-(4${;^*Y+vT^|H;KIh8sNE_O>zg~ zW&~gv;?+0^DKIQAuu$J(vL4xL#=jt2E)IqHQe~>pNH@^EM4JUtMhnHc9^g8mkE>&L z1w(MKQ;`YX5YR(H-G-i-6)~-cS(=FHT0>2uS=22I5z-56~PK36TuBBT< zXvXe}&(V8BkEy$&w27W|it@Oeh#oiaiY};*(yJmiK!QA7rELK>`hX}KdJgvsAA06{ z=Yz;5IKBDeWafsTiB9yaR}@3{*bd2N!NHJYnsd;boVQOjmWZ;(m+d=#c)L4w>n2-= z__^ZbayZE=e0wCPWmkjqquWTvlrq}(ePdh`tZvzKA-F%;0p*jEXY2wQ_+xZsSRZV= zJL&Zyb)p(!qA(+HhqRB*fwUC{0Sf3Zjx8sbrAEJYo*ncpNqvF z6@T_W?~#NT#vVZ{ts&k~D#a#IWF>UGn7n+No<55u3I{Zuq%8P4o#yXASBG<2>;ent z-HtnPFYw=g5MhCY^8ac#lF}Cj5>O301QJygdw?PpSX_k4@V($RpEMO5q7wtY3<_zq zsxAkboEj8>@AGI>XM+jR=9xU4Y&uJ9gaXs?= z52(sLp0MSI+A-bHJQK#eP&;8l^@hJpwb+ftfACEt+5A=>C(c+Itju32#X|i+Ar+51 zY3Ml;a3L^D)fw>aI}524ra!Pv*%5R+;IOnG!ZWR^Ji1TN0I^|^LR3PNwkcPo*uCha z1ZfnDg^+PjW(V#JI3A!=SA|@aW(PRJm*I)8;n$dFB4uPJ*9se=Y&SoAA}~&j7>;dG zg4i(GF!}tPJZMVO9A^| (`x-tbmcmjB+^8o!*WweC8S;-4CFAZUR`r|y`!6Sh7_ zLLdk@+2D2`eggRhP?P0eSr<{X%SLT}}6nVfb%M(DMbR zMh2B(?h0N$Rpe!{Hy2(x_Xm>3&)zt3Fs97P-sDj%RATC=c+`4Hg3egzl4=)1;|CCm zlb;0>G~jYoR!q0~RK(l_k5;E%D8|mF+Vd>*J*6MHADw&4dx@i~Uq=p;Vdi!W4{iVU z_f|}@z$DK22Q69f!oURdI)-eBR#9v+MLwkBt7az)%R$}Hf&xrUW7jK3tSnzLPxfKr zO9pYk>AUf<2PY1?D$7ET0Arq)uj9NPI@xH2i4=-Wq{s>?zJN9eu(?^6z$9-u;4+^6 z?T{I?2N6RqAGeoL+8=%LkEJS>8Fzxk?yx(&Ps^~WruL)R1}Iw81M@%+y3rj$+OwXO z^hdLI`+P(?W!lQH?4TQht$~kTy`$La(d~}6nHHGcK4+k1-hj&^P_=unh3NdtrZ0V+-I!5DcnA0G zfz%i7$LL=1!Q;kXHw)b@MyBFx-)bNq@>_VE*t;#XGBQuv5#LobP2Cx_ermeP5Y|WM zO3zErNLQ1Oy|1ax&*=t#H^Xy zYk;1P$C~4gD}fgZq;);~m3R+yX;b&PUy&uT1~(innfX*?eEk&k5WS|-!`{n$WtRUv z?>Z66dFs?gs~$y_5yx{Yk8Lzq&+HAvI1}i{vRJYwp)n zg=~kQy%yzZE`ZJzC|qFyws{F^0bXcDpw{5F#0Lpi^XC_UNFkhpRqKIb{vP?^DSCQ? zSJI5pM)`*t37)p-OO~pkBaiMiFW+}UV%4c~4~ymc#U(P2ta@SAwNk5fEtg^;3z-E; zM`*;?7JCmy(pBB^I;E*y5{BonR6$Sg8HX(CNKJ*K(@orq z&rg0C_OlnUyG@s>Br=~)m{lxa9Jra;?UTyXK}*l%@o(yBuYDNZcYP#xbAuxF@zQ^t zVDUaQoA>;M^t-?kJv0uQM6oL=vJB#&P*rGA%D{}S!x!V>|Dj+0y&L{MM<3;W#LpEa zp}V#*%$D!uw3Gx622;Ifs<1u1VRVQJi+NPU)`hfG4~SyB#B4Q++&wkk*1p;{{3K_3EPAwAtaD@O|DksYe?s20&7 z!9`}tbZpCtX~i+OnPUZ+(SyW#1YVGt`29alUTCX4_FTEI95QJS1>4z7dERuWKkT3@ zfO&5Y;x_}|rI;FXELhLLQ}67C%2Rhb_5e3jwy%3vVVhRu$j4>Vk3u@5*GB8q+uhnU z+AVGynx2w9581JgbNJD@AM!BXiSzCgk6b-;+_oT*PaMRF6U`jui$*9`hvf$= zf{c*SYlRvQ!<0VHdm_^zJzXci;fA@xjnM^=7Q~1Nrlng|z}U_XGB0730_^zdfW^lG zHG42WtYnuU9;B-vY7-jzfd(wDCrV?EnsX|C9V5e&3eIwl^%c-{pOI%Vk*W`Y`QX77Yw3PTmsjEG&qU;(<_ zR|zZWMX|@!Wz0b#hSt;kyS_Fqj*?vf8C(!7zn?X+V|9)nKt59ZjV->I7*3t$s}Xg5$A10F`&@6n(JY?Fill(kkhy z<-$+1agSjVf z#|mILj;H>J$D0<|%&ic$ks2o^r(CfDLOsQvrpO5@z7SIX*>tgdxobsC_mn(MzhFH? z(vE%mifkcBd6j{bSC;_OFquJRb30T^CY6RBB}*rnrC6HdkPT1~-KxroSfRQRH!oj& z!~L@$oD<#cwOO=MOn?9TTo8wwUFf5R=E9qqBh_orJ2JT1@ zpC;==dQ=7eXCu49M}VKPM9CAlJoSPd{*_e>j?2WhXX89JR`+uJvALNniM< z!8$eiInX$MB-77-Vr=iQ2SB#>jK}-;J~{8L7Na55{LK`b zLjg5Aej7O_S*$uEttKVFhg=kzBETNudI9oFQonZ9B&k})tR^kww!9St5tfj3(HXus zB%h9hHn%+}dCF>xt-i{zpu6{#Uv9T`a-sfi$m4y|zX$cQP!RuGSc?bt+t8l&Jnh*c zXc1`ZT)PCACBcCBV)vUOBT$@c=B%7O0&_gqUPe z7Bf=jl5>hIn*=S{hXQ0R#imhY6%~&O4-80|Wg4N4c2_uRJSH>O!v~-UuO=FV`o;wk zmKo%Qj4_&Zdxm*Ikm2#^^~~uOWW4)c>jAQb-vZ;jm?yPX0N+coyC_md#V-_NliLn@ zK58bRW^M~z2knFgjbAosn|>O5Nd^+BZ6--woqE~iPQ|COMN-^d!vg0Fq9zt0|H~#UnhDnFE3M1y zt3o~@8T{;$6Yr&WTCHRS6bnmR4i#S-dosFOR3d3q)Ctm6)uLm{bk!+!F?jgRaeCP3 zeF9wTYSAVZ`wl}Na1fcqvp}|frQ8>fJ9w$%1r#eU#<8gRxB1PXwr$UxHjl{B$ZUxa zBJzEzOc9nQV|CvJ`G+1_EC;+R`B0TlcZ^&28Pi{9{;RX5r;2i=j z_WY=Vz-)C$VSqMGP~w#%#6ykh0tX|Bk5#c+KOBrCZmaUMtr7dbVevsGcK!FKr=Q9o^@lQ@^vtBVBb!p!aL>*zMCG+3nLo zzk|Dy#R{EA3yq0ltYgQLu5{J<>7NF-(`PjAY>r5Zu9KV)pAe_N)XwOq9}*a*oS%~q z3Qeejk>K`@*_$1VNMhq>;K8*sNWU$TW(ReGquXYRS>xw6GwWS3s{M}r7Hkw4vVoLc zZ>??LGv0AyXqc60LfqAIPh0BK3$pFuAWB`5ZG-=G>UB&-u(m1$wP%qkY5&AO2MemX>;bqSE$|$I6y) zOz>3Fi0xe$Y>CKd87nZ4md$mfd%3GfN!FnyWI5OgF3Hi;R@S zeUb2QzV4k2l14^Rn{r>I1DB*>zKZsInJsJ7Fn2rd5U*9rc|+i={;P7^%JvszmEtJS zH&qH%h9NVqQgYqF+i6bxeL1W-Z*}O2!qR^e{bTKV>J@2%958N0{@21bvBoLMVatU{ zUgK-z%v1X1m;giP(Agh;@H1<5;cAQD5_jV510dh zXN8$!iiJAE0xI5|0myj?SWfAxRRSHzQ$AMibv3xvd+wC9x~>PsbbPM|h3qzEsw+rn zq^tDPy9AYVv$_yk@D@pn#Ble%NF)m0oaqR#@Ju8R!o&La@bf*&qoGpYPFtS7)24wq znD=^myI+b>PhXNX3sCC%|ChVA0R=yo`M)~eJ!BhewC00D%khGInRsg?GR3=8rZNKTae7Elr^wedvn@2T+mP9&UE_ngDC1R{ zK7tt^%ZcS2i=Qo-cD3M}wmc=L#RmuTs4h&Wc__Z=gFS9Pr#n>JA|Hf55amOt${+)2 zE^|e*QQ4t@YJ%N9eIBVoU?G5iVWdBfTsfP!z&+=a=QuF}pgi#_IEEaP!~B3U^I%f3 zE$4vGqT|GFD+e0^z16GUM}y3{J~Jm37zqbNJ@f{3xg<|!BwDmRm4ZHZ!xUgbHnT{1 zYE^CUs+zjl!;!1asNB^t0T0(8`74(GO>EiS{Gt4%zmUY?c=}EZh9avy%m#{GkM2_Z zai}9kRksUrbYhPy+kC*3w5rOO%goMzK2LK57zo!F-$y$8Q%F&lgTE$~0diQ&wK6Nk%gTA7Ut6x&RZCMrHN2s97CXqc#LEgpMAw@riU zR}7P(7CD-e^K*dvfIT^qnHs+g;U<3+!@zYU#lI%Nq+E^%io}hexr8aZd`)o_*qMu< zse$d7?TK2cMQ@GJUXVuFLR%Sixgd?t_S(&OgWXn&gHG(z&fF(}gK2?bX{i#+JmJTJj0a}p? zgyhduVWL}3gid3A*9PzLPq+^h@;=pp5TmkB)j;3DvXhP8ZQ=EbrU3j-z;gTd0_C%z zJz+xPfB*cR?SjX#QY<9bRn2jQOx4u0uN;fcmm2BRkgBN(!r~}=Jx(sgldqKB5A?|$ zQXht^Z_Z{NVkAr znJi3Z%>2bW3anVplrI+-Ng)lS-laih1>ASFs<10VTX-^qTd3Trm*8%!II4(o5TZr_ zB@42~1UXL)(%pEYF4zK?>DeEwA_c>3y`9&SB8RPDv72JcDYBi4?|~w*oEe3J!jN`v z!_)+ErnoV3xMN4@%%F#J4bm^JEf8eN8zYe{7;-VhJc3SLDyRkrXCQEl zVKRDn4BDTkW_VNwxGl|Js(NUV1vsyrTYHGCx8R_f`;5FW;;D;6$g#vGinh-{*hg>5v5mcg%Z{F58islt3{H!`e;G)kNG z#CVP7#nTX+FyZZ4#lN<7gPk@Y$Kmppcr9@4fJWLykX=OKtTyEtA8lGptLpBn$jq0A66)|Lt$=HfTDxUhGg>{`(59&8LRy!>mwWq+_NTR9JpP9k?+m}M)`D&P2+Vn3%y zFBQLmH9#2$I>QHpS4CCQ4KcVU(y3QP)zO&Y?hoGqZMYR(;e||&FkPj^JawbmvObx` zPO+dhhr#p5=msFvpn%{xf}ERl6%Mw<=duj9TgRh2=%e66VY>k2?^*guV6jQNVjpM@ zRx1;|V4EJ=D>t~EBXvr=u{^3R{DE5{d)e<`@Lf|rVt34GsOH>9ACJ&h%}y4UL)rq` zE}{@8h9dF40@rd_`H~sXBb?r-Mm0`Y2WM}Hz&~uj1zCkoThqEq7}01e!Tf^QBn~BM zb78%yg7HX5F`Xub^~)7HuSV+qv^dq{$jUMq5?*WD7@D3h8!93G=ijz%CGuQtwys$D z%}?5kv`03op^_|c;}q>?C_PIMXZx2(j?wkB_F!s%(dc7gr+2w>qv;5oY*J)Q$2-g?_0tVPJc4)fj3A98e0Twf8$QE1 z)!%$Sah@%0#%XtB9DE*hk@YWwAep*6vc^w)Tet>VgF`RQE+zKg9B-GI2QI#B4H?t( z{BY6qmiDS`5&qN5mlJQPvaQU|DvC{}$cI=~-AUgL&5A(X1E({taIz=O@9IWN5#UTNK+ykyec8o4$cx^f)vTZ0i+W zsw6fYdftBMaYU07kr{NKUH}PW{q+5sRAIg1j^YzY4x7Yt%%_lZSU@i&e=|(Eqc}mf zMQV*xdWjBVW`A8EzN5gq;QD&VD;wqcz~0^LT{GwA%%il{d^?`UHYXS4>27*?G6dc_ z6d!w={*>jO`5=RK4LF&3qzj@qmn6AUEgKv&?IBr;zDFyR9b5{#A2o~-v1|TD=Qh}N zx9!w(&)+ci-+6=mzf)Fde_#RI^4}eJKu+=7+BYU!lLxLpKmZ_}y%H~0PX#@80Vmi^5}sKjlKD+}5cxNOq$NllPF$KHE~Uf$&2 zAvvYa^j{a2JgGS@Wzyl9ozQ;|M||v3z%HHq*d>{{$m~#_(HMYBcu?4o!?Xz@@dN&~uf3vKS>y(|Kxx!WLa zkID+z%b`wW)(&&<+A{Lz%kf~w=rAuc$);Xnxi(sq1gHKJSeMa7M4vZIUeLlT|9nVQ`18MX#!S#zRsn*dunZ6|;H1 zJf_;_PRG9GWz;5g)b9V>c5nTH1cy0x;#hB#tGPK7TT?(|iaOWLvMQ!DREq-N<&kY; z)jQ43N#=o~an&^2_?x#MFqSptcOUf}Ba56kjF)S*O01#SREi`~@yO-9?TPvC`Q}P& zQ`iNYPE^h28HcrVHn}gef1bI=aKss2)?#AV_Xi%?%9S~7B#uLX?8=*0WVrR-=#6CF z1aXt%u@ahtlhx#b2AhE&7F-TY2{RulpVLC)(NAX`n9>}FQQiXAZuhIQ9wrTXl&y7K z<6ryXGW`DA5m3P~dBKI6vfrhbwiuC~ANrpl3!T`AKojU8ZHre^YzjpZsraqIC6Ptq zeXnS-eNDRR0P6i>-Rv;Y7Dqkw!Ex;c#o+bh^_v$cCj2fXV!yiu6cu0lRs;F)h1sou z!r>4IWl`)p3g#?+zq&+Xq6Nz_DrjzDf+U7rq*{ax66-ya*xNq0UUrQ1RFvdX@klLzI!Ai5pdv_LNSnzPSBq-3&oH^#w4_!KA`C&%Y>DZgCPSh zSZrhj8N(w{Sz-)r;5{vL_pCx-CpZYsBZu)IqK9B_3}5ns@r0jA*Fg(R;=iF1Bgk=fzbRmjT z{`M0|Hvv&w&okT$1dI&ZqGN2KR5y+(BdYK5A+Led2S1 zx@1^!X(tZ8bXLY3C|~-C_FcdYlwqh56DNhf8#|iKwDWrSgkv+E>wZ_=BJUTXU3#d@ za0gxAuukCx2-sZ}op7&K=DFjU^oag(lvCu1CHsId23NhkVZvYkv9#fnV^>Lwg$G5S z1|NWM%8|KUswI<__?-0VWzR^L0^_7bvcGp1FD(da zjNCILo6Zi(bYv!xCsu40s4)Nq#}JWWznN!SCE~Q9eU7@1Ob{AN@UPJ5G+3UK0CkEW zD`+Xa)~PQthB;%kdT=~Mo5MWj+U8EjzU8;Z{NUVwP4cifD8INw=8;uS?4XodIVial z3uTL0RQ%=WUXsIPGI_F8p^3&;msn{caA&DcT`H6*=T{2c$ksH1Or(69k-$&&bsMxfJLpP5xf1ta+oRXTEDZzoT*DOn#FL13nEAke z@9CWM?-rbi#cO^*&N=b=1d7Z89vM(|cw zOv-V+F}py}5x#O#HH{3oj_}S|50o7s4qWDq{TT6ex~g>UbyX#(H8;m)dffz7qI6YB z$Q4~49stV6O;o;Ec(_Wea9nK>+|6J85nAWRfjiq6kjAYL4p;ocmFIKs=W z%J72>!ZH}{NJzTsl<4M+4EOauS7j}tZEmMT`9a3W&aWiG9Vf_P;dNzy(9M~LpS|7s zdt!$JHm1Y;R=KEEx84r4AnkXhhf~Pr;bMW#OB5zGR_H6I*zFYAW)@A{2xM9`A{K%0 zb%I=tAri{qBeW=EW}=ms3Qj`8z?{E^_I+Kdjow(ymc+tE7vu{Sz^3a^eIhRaYlb=@ z7eFryWR5(lUK-0qaZHi)ENJjN$Bm?4vF~TOT2OQJ)#~@jVkfpwo2(F$MzO0Xl1#

    8N%Sd=aPVw1_DEO*iLvi=?9uDr3hZ$;LgLuhYX43aJ8|+4-u+R7 zoK7*TD3U_O?Ur^M)*g-gP4XH}X0V~4e8Sp%Oh(9>fi}GJCbv0!hn2H2dDZ{<-zUtS z4X5=~<`xhLV^)eHPy7MCC2y&1@zW1Llfbs=gS+3Ccy3N+u3 znpNs32E=I&0HY}fN#YZOyS+D)6CNAA_f9=UfEt^pQN;hSCl;CnrDvf$X%~c9mrm&c z%_4pCAP9jMbCyogedmjJu77{|+euTJpo{+tVKFEdHLF@lJ!uHVe*McqdG2QvrQDUw zam7VdH@$~rR*2^A*&vmW#JP?%V zTP*|@22g{jl`dxV5z%s{Mk&PZH>v1vqF2`@P!o}o&X!%FKmz^Z{NHF$eT=3q@ z?WBjKU9!xX2Zc3I{xRTr&L<`Gao93whAg66Nzu%SOUAM8+i*NE44=RX9@Gr?pRby0 zoO;|J9?u|q*{yM%*m1dNGEp6)7?3}zqvAFL6VJhEji3j4UWm2SO`JvvDr}C_6zS;F z*k1nfD2U&m7xt4&p*mniXp?;@M*jwjws(4C>5)z^!G~;7)$}g^UfDfK0`<`+sa*AK zZ>&w)tkjprHE}SuTR>lj4?^Y6LLlu%)nFTpne}GL!qoJF!{fziR!d9o*NsT}*`4#* zWWN(TG0i5}I7uSu~Lo1xRF-3Em<2g75$=y;CzJC--2nOpge&j^^# zO}a?3oES{|Ou$q^F+~*FOvRn2p`(z0*7vb{MNos}R=@-9*62cI9yc@iFt;LF$1j&Y z;5Mo9p@Q*($2so)D1;b_&0`N@Kk0E#e0^(lt#^)}0|Rijfo0rSc;Sug<}d;GVS>LA zEKitI%ZP^E9OlGv+Jh$0+D&`RkAoJy&VFvGPe9<^c}9~ty==;Kfiq4O)_ZuHKOA;A>( z#g+$l_&o6`ke5>E9n}Y2<@UJO5WGjDFg)SKsV94&>;*p2n5SO^yxk`Qt6ugMc{&%X zSTynmvA*1X!1GJ74k%GuWF=6oWP=m?qVI0(^VW~_*!Mf?D68pe^6yT*>1$q2YoQKx zr?-}0ExZja3+H^)S@N3TeGo+LlI&GJrDelLE<=~Zgbj!F!!YvRIO>ZxeJ;CUa9R?e zzp4N8-M5Uh(;zqRXQZ0l?BvAm%r%qQ=`6*ZqR4S7?&FV!zmZ8=U&pm_a&&f#`U_wm z%$j;L6w4vidqA5U58Rfn4*ly#O)xiAxh?U%7@8OyA2TG~=c+CT64oO6w)7mi&-)u( z&=k5>e**Fis`3!bYn6xWiiQ@VHrei=eIe=^@8zl-y2AgHkMI%aIFEov93Kcd3b}d< z-T&6*`5FbzO1jJzM?`)P`P%XO<;y-ELbvQ)0U5-WZLd zhF8eFExgR>p>qYf%pkPQ*V6sstiYY$=_DNzI}h*J12LXKANxu6J<4hprS`rz3!4J1Pr)u^xi^udh2{D=pm5Mt@KQ0fS-$(>yakN2~3Gu z<&nav<5bRd{CJFKFu-Wr_nRN%Y3v5WY3+3CPyXUN#b`)Wx<9MQLUt~V6EC=POw7(I zibzlL6W5NU+MQVz##rPEH|4Oq9Fio*_$L8B7$I zU}7D`q*G)S6}O6X30uUwgZjmrl?RlS5$!Y8#foe}A}^5#ZZ2qF1G%f?P2JPcwGJ6Z z-}LEi4&U0;bf?t3KHF)l=`2)SuXL;8Wd&}gkr_^-*f66(iu4C4HUa!QbdEfSE)aY+ z=ZQ~YI0(xeCPm0R=m9Mu=xNqWzaHvX9IzUB)5VEHN1xSnH)+SewP%`>Q@$eW8VgjG z@q%NK4@Jmv*CtyvznboltctF3>yE9avjrU>lm*lq(|2;0zm@|Xi1%XGee2OTHqPJn zPV)SFv4d~posLBn>yn74$Z{yECeg{6$IXR0Eht=;aZwy*>9ijDiO+!`U^a(9bgtm2 zf17kAZ!NdZF?kZpf4%iDs^w=m;tZ=7)p=PW7WB=Fc-cerJKQ>;#S|O}l#t&Hyz!?M zdwdQ{F*k$%Irx<~2iGeB*h%TRW&dLCfxIF!iv@qBhyI(oaK5@>{}q6fwt zXaYi6-#^~)&JI6D-wnwQA96Vti${*pRiXQ)@AJj)x-!ez4J@!-Q7FC&dh*@Z@DiHv&rU@fh7$hYC2oj7}Q)?`^*OEFax*-yn`Lure+(*KGQx;*NG z4k))p6wu3})!MLapofPUn%wXb!D?P6QCA0TV({|ZaNHe`b@QP|q|mhlNZ6|Bm0q`H zddjvU`mj{XPa&O9C<*70FkuTPg`-#RL=isxD2)PF)3&`n*1ZYlcEqZEn_iF<>B9HS z84lRYKF<4c5DR<=h=~#047(sVACjiDh(}0Mum`OpVq*|y10jo-Ls$8s((16w;qb%Z zFRl==X*8U6Rnx#M^5GaQ&7V&69S0E$RyNd$Elr+@rCCEUX%tBYdL4+P>YzS0$D<;$ z2x_sLnT*#y`luqZSky&-Y@01$4H4^&2kS@Mbc1Y5?{8l-qN4M+*FPc4oH%C(g|ee^ zc9|3dVZJp~Tt@iPDO)^irhlev4NaPIpS~@tr3a+>P)(c4 zFg@5be(Y{$^}xOT(zP)mO z69~g}(HB(?0~;%3jp2K+!;AG-hpe9UZm1C}%QAX)kZtU+;>5EH6djGSI#m?2p91{E zVM%pAsSsY{=;=~WT@T~+IvTl&<6}0@NC?#V<&o9A;h?1g^)9+ z!+fNV-d8ajqlaBu$z9KD(2!OuzBIKz4Ev07xnDWxg?kHCr|nbNZ!3PrfRtSf zSW)s1mzk%0E#yiw!gDFUh3tqwfST5JuZ0E=^mAHd-y&40iOXMb+>B< zG`!}8^oq7cs8d1{eVTkU%ERP3c)=P4QZ@~{pe!%mh)>i&=lr@i)Q!;d(ItVEIb4uK z@Ab$7&L_w@(>wfuhYqrju-rzyJy0D5bv#^aV~}dtvEQHK9jD#Ud>GzcX7OY_{qZ8q4T7^w^|HhyV2Te==w8c|{i07I=Q@{I9!Z zK)333H+8yP7m>^$5had!iD07C7@pcgt6}lL4|7`6)o71y6ZLNU?Dwe)C!7qWF zFzTQc(bz&$8(cgC8AySn#n9t}U4>~gyXdr;>d(<((UX!oeulMr)JH{ z|1xifl~Mj251s57Y0-0OsWl30oJyZi(xj2E&Z=*IKjA&M!RGt-SV#-6-eZ3;2S(WeE&F%3o5nP59cAE^k-b33s z$LO2FI{)^V3a*2Vja7IzwKB#%$AbTm?2Ql2o4uTNezDMnRttsO&8mW5{O89*B-_o&o~UHbk3n4(|BH;7LAOzO$uBM z*F;7nk-fRr4N*4AA;XX?@HrwWK4b$u+v$}&x!^S!-u+>Pj} z{N{JhlND1)vB@B2Q_Ok_{@A$l!pz_*zrIL_f9@v9j7HHP+3vn9rdPTPzTSLKZ0Q_! z+2N1MDlhu(UhpSaf12$eSpQ_(u4Fd|PU`?4;yl`AUOCAooZ`f(Bnt{bW!y&pEh>#7 z&%G2Ba@%BE1t+7ns4f5%SPyq%iE$BtR>Rviw6eb2|C5Vr{x~ux{lZg1?y~b|oVa#n znMsKEF~#&#79f$fuY${W|b)s@uF10yAXW=|XwFCtw|H z`ERU8-wUTKyIwG9)wQ4S&7*nGk}W%NYSKc4PY3io6$H?=YU(b3eX%!ECFnOcE8Ao> zQ7OC%A-F0%(H&taoHL4C$Pge+q2n74t+qqv1I^m9u)D$XH2Y(!@?@CNLgmRfuO-{r zStut)&uJ6)teRpDP-Gt!SLN2uy(!A&p)T1TAEdNUALdo@FlD2D#7z=>=6-Q*5uM_p zUg?Huf^JtG|H@QIEj3Eg>s&-{6{u^0K696-gOA=dN{C=vTNj-KLeF)a z!(h^`&fCSg?v~_cdsW`Jp=Ay=vwe0zdiLr^-<`M0oKd&LF`4M##4S-R0gnM* z%)Qf=bB0}Z`)?PvyDjJFTQv?SQ#e=&0QBgyab2TP*#4D0`^%1TWE|h?7*DbPAnHr^uY1>;&q1$`#6^(azzKc8Y9lr3pqu=F4sMV+P0lhH4dojjfI_mjb|SB$N$`L^IhW{boAHP zJ|~CS%|TAQdeWL$!X}Doq{vw+u9SB?`iZE0)`0tQ&Vc(a5z?*XD*8#+>?*f-LA7wn zoF>&pu$2orh4TAAkDC*@NolxN$4`&S6|_k&li`47RS%s3k-seQWe^C+me=tbz2XJM z-le>HKHh(AR-Uq1bxQ>G5#e~O%FS>Ea`v1IseCg##Bd#oqE(0X%|1A-gWn7OEX)6m z7glBy(F=E3{#ug){`tf2zwF|f$Te6%r^(7=A3`3fMY1U5GcYO(soSzoA4==QI`PNq zjCVHC%L1_9xj}LcGTxc4tVfXTH_KwdZ8ECtKZordtY%Fr@9IhA2S&qF;eS?2I@lQ= z=bd&*qRI075yd>9$bF>B&Es}S)JvhbcWJnKvvQ9QdaUVkHKrc0tf}Aq2oSuZgQm{+ zESjm+Yhw+&l>Gy)y)opAOmONGq(|S6qArwI&ehV`hJ;Hkts<9~L!)okM|N^9!_o_! z>jLG`pcIb`Xj3Wh&iCBOx$WH|zB&(y8wa573#&CUWXH)B=Hrk4OTB3RhWXvDrEkF6 z4sY7-e%m{T&I!5hwi^OVSLfwGMe>ozl;B~PVooj(Lkv*S5uWEhKwtT}8my_w0^S-m1-&Ms*jMnJn_ujifR! zfB*0gaUz;=Ar$rd182VY&7rQCTiQ%)tWXgI%F#{BNM8z4(a8KzoX<4dM z%x2MvsQvVV$m66P^uz0eT7~Vh9*IVt9KKPWA3o@{En-NjUIr}Pxo(*=@Z_3rV0%G* z(4bc}t&>9iC+d3UD+^sUiu-hz1VTI5h&$+osk%)JPDgn1x?3rw#?R{&l|n)CLf%a| z(8eE7RJrwvoe6L-c~EY_7-hs3Z#5>RnfbJ&3Fh#qc@Ka!xyOU0=ofm~X`2{LAYO zk>yUDW#3|Akg_Od9YxZS54C7kL3kx~BccK7L=TF(d^XQJ=dn=GAl2$g-n-~>?}dVE z6DbdE-PAhbSlD^~e9q%oO|4Yc{hJ5PljJW|!ofmDze9=112w+aTZ2l)6_G1Ei$zC* zOQDD9q;!)?FUp3#^hA=tvqKCZ=Q*Qn&r_BWv-t06|Mz8Au}P0Q7OIevo`G$wtr`JdjNW^_TmzxCI@BOBPcz)l=4hSu9rMWUq?0~1mq6=!G&&HpKs zMa=0?Vie}6|4}G5KCS2MHk>^SK5470-~9|fUD*YmX|oEb5M`;dBjN_h-xg7y8160M zc6vDoajd}gm8}`Q;TWJz{nOXKGu;T6#eeo)NphUnV+9VOQ2;2Tm{JPL4CAgt+d!3D zBOggV)OG$H5uK8wuFaf_bN5I$F(m>6#bpuwMMMb%+I9pb2zCdlQ#`JSb2-(apZV@# zs{FQij!aEBF{x2s$N@O89Rr95`{RI>ftkF2bvyMnBVt;mC(vdX+6Q!_4Uu8nRv2!MyZ;Sy+C-;qcCkQ&h+PS7vRdgelKWOu zton59_1F|nlWIMwl|JDuROGU{)9K=a`)Gd$Q_a$ zW%KYpXBcc6`R6>>)>Ev^4VC)uWjkMXU6H;wbHhP0n16-Q7ky%W2p2&>ndz5FAbkXl z{kS+nvAoSn3c{3dQJp;*=$- z{>U9(-OQ3Hb>apPyt)~1XI3TMBH1L~_Qt_Em*(COuL*5dY5Cb9T2ertXH#Hcm%aF>rL=0mAu`IlbIurdgFY zy@THvTFn7HHt})AvKg0%#_feLvm0*4fYS30y~sMTvQ5;C)_j>Q$|X*mk+Ps%vNlqq z#AMMi?~OnwVjvdZ#N3wdhfMu{2hBJO9KC4h?RtLNd!b>pan|blk^B@%m_l+*%>8PL zNu|h2D(at zK&vl#J@7-VXJ~8+|5BV7Q_bxFM#e0k)|vMtkRT`kE}9J4IdVpEcIuJH@~CC6#k)bp zJT|x3>@Hh}?B~rK7S6p0x-U4-3fYs^{x~)9WzRr2@+Av$xq;bmm$Gux~Nre*!d!ZJ5 z)Y9h=#ne)yii#WHE{jg#-3!zxG>Qk2RelS316(y`n|07+gqcJnnYqh5>J=YzTv`#? z0Cg)L(>2r#9JH-!@siulkmEd<=@qqny3^Mzjv9CwtucNQL*|w=%d@3N_%X!_MU#iaQkifgilqOg-bl)(kTWKO zRYNfcDN;$rH9~AOL$*v-74WG zJQUm-*vP7B6v1elvt3vKRk%e+@Sm;7R;==^idY_{?srH2!^1=`41j6F8s0e{^fJ}C zk!zxD7^d+SaF0DK=10drOz?ZLX%2GUrNg_i$Fp8`FQ*S+ibV@Z*4rUXa|QRNLb}mw zE5u*)n@?m3&;kub%+;cox9g#qt^;Sw{Q>|w{z&lA?mGyO-x1P z8Bz$kOz8VmNDIL4(Sc!F8jehw=>6EGCc-f$$QDc+?q!AS$zDFZJjQ5(+#mN{CF`7c zBHL|Z8;d9g@3!p@Bp3dd9$x=9|U9wYlQn@!~B`?*tEOx{ET)!qC?7vNyH%jhF zPQ%gBISwale1*Vh1CO` z*N8&#_-F?yg<@7vB#DY^aLpI>1t!0N1>-%@>uj3{$3IXnhR^t)ZOfgXYv(9lcFhY^ z0>*HE79Xo!DkFLbEP@Xz;fI0h0ZJHJ#EEZY20PY5j2ea)kG82BH0oA|ZnApUC;wqr z(PeY(&u1%ioi}@@%Oqfpc?@4^rnF2)8B-C^*_|z~WGJ zm_6JwSPnF$mw(eZi#+<)yT2s$PV8^oGMN|7Q_MMvoTlQ|aTW?twgRcz)lCZIv@4{s zU>he3HELCC@YXNLHiP6^hAi=oHKE|Y*3CQo3}QP(TY zhL+1D%32h%0g$QP8RwoEP%Q0Q2GAHqypt3q4-cACMLTV**OF4@8;hAeJ}a4;=`G?; z=!WfyUh1d{N=sk5;kY%Bc>WU{bCw+>YLAxw^kt9U8p84Fdy;nULV>fvXfJb=u5uGNgszVc-1(bRvA2G+`27d}_It*?E3v}%vfvYgccYAG$&vlMgzR==19QP-{EtvfJw<9j#1H(^c&am2od|S4X*_+q zGtNIqW~q@I=q%M)kTLFpI=Eq%bsl(f6}X~bL=3vO1EDh-7d`c*M!pA56e??ox*2j! z;Qc~Y0lK3zXLLaACf;UX-^5DJRa}ic3$SGCV#;P@jAetK9%BoLudZihbtYvczcJn1 z>Nu@qYr&3G&~>~9{*LIw+$PnrZ|{k{2*q%{-$8c6`yhLT45#sei;^J-{O=NK`OPXk zibtx0mNWgNS$|JT==IPeQv1(MBg0}NT1FmY_n)lLLQS4k^gZ)lMmA|KCl1GeuIQ)~ z_IirRpvY?Eu}^zlqsY+<6szllQaC!PMo}|8*%R!}V{R7gW{FQYIJ$2#=v70?U1DnBO>@Z@7Imb$5(U1Ip{cWSuGP(CRpOYi6jEj5Q#C%+& z80b$sN5z$U91QGsdmzs9(DDy_`=;n|*fQv+KQHWYzap%e4%A_?Rnc0|!a$awBD$4a z5jH|SUd{AoRgv;bX%AiL-X_h1pdGr!&~gU4Z-%Gc65WR)yuR5OEpC&Qa!2k(DsOyR zvnm~&>!R=76fF+f5VSA2k~+h!q_zj$fNH&5*IIh})HL2{AzL7h7gtXaX62TZcpACv z`uT_o^YUD$bz&^k4c7$s3D7-FmoM_aECs7lpsc3%D<`gOZsg0fI_CfMh6&nY<*Q8E z|JTQ=m%Z+wR$L36sAbYTP!B+6(`07pv}1HrL^s_zr+;cQbT4l4G@L!iZvfuRWzow5 za6ePIJr;_%QaP!P)n~KCAUF~fqMV#((_*B-pA=uXe_($lOIsWR778! zRn4yvFP)MD8YbO($Kw`#pDy7Jhhf{;lYkCBTgPG4*ZkZ~*4QxjC!YKC@t?GejWLpi z)RnzupG6q)B6N+-Cp*}=xlX(kI%|TiT8aUhkNs2}+L3Dn6V>h91QH*kM#*CIL$Ut} z18v~Xw8Y!OIk-D$Ux;B}ZMX;30c{1M&44op;K=^?v8KQCoxZDIw?&4_i6LY`hHLrjRr5|U zSGg_X^1w8r&KI?D3TT|0u$a7no}ldG$gr`%#mFra^7Gp9bysH@P0Re&;+3T2l`$>H zO-##virGt%UD%dz4H!mq!wsb=s1JFFs{`h__>eCI@gbLK^;HJ_7p$mT%BvRE2SdyS zyVt7xP^Gw5x)^G-<3lcms^jOBaZd}|xxKO*(E~D#@<2clT{IJDz4wX_`xVi4*n&E|At&-%GeY&tBEirw~JGe1ilFTg=p-}T!6rbhA6M<1nqGd`wL zbXSI5Dowt1K_spg#|)55{W#^H3&b>#Q=? zaM=YEV?Ef$W0qZ;f>3eR%dR?6?%fakaRnZ&1PS$%1ckC8mwL~B@sLY_+W^#;SGm>F zo0-J{2ZE6Lsabx(-w|W~>=2Cu4bMKr0Y@yOVJt`2{p)}I>C0Z9-SwWHmSaEAa-;%N zOP$YZ(x_+#h{nY7|z?jkW>I9B|g|zZitSXqe5IcM&;l=FmdN5H_bKx#8ev235Vj z@OA7H(ZX;RbJmORhIQB+BYW|;XmW-1Lv3=K!*^J@FOxs})oZ^p2L>DW#fewZ7GQKM z8vWg~?zyg=q{L$z@8cE*EC-egskuhO|44f%O?sN3drXod&dBK4NJ^`e* zNf2n1`cQYgAc-6cY!aoz@^d|>GN@X35xKZ<>2O9-${mv35MX5)(4f~UFEZ&Mqgq7A z3+%!$Vd*Pj`xTo`$d;aJ`*Bt_+<8rK-T(TY%A5q(0(-2MUlyzW?5*b5N{?=DECvB0 z;@FS={>NkVf!GTk#hgNV$mP?IKGyFq3aNr)LoP>RKl&679FAQSa&SH#IE=+0a3pR1 zqL2r0q?mJ*9E$xY0gqkN?-$s^ORx<^U)4fMN}c}` zpR@DT+al`xQ4V$3#gKTZi0Gh)UG9ZJs%8#S3hH3OX=WU1;zq-YJ#a_g?YM1r=5Fxo z?H$3!#d3Mq&ID5A#APZ+OqRcUDFy;3JE=HaW+w6yd1-UpgXg7!a~=u@HBMx zoF(aUWYTDpA>~^~SfivxwgReNyXfnoSTc!sKv_O4bu~(zZyN8qFlo-gfNpUcRBBiG z_1Za*%N{vT4V&4PU17uuIg>A*-d^ysBd5Z@gPf-?hm8DeQsLIlDVIZ*SX>Ueq6nj$)snH-e2y*0hv-Eb|mJ^l4_)DGW^Jp?RyJ$ zmI3qbA?WhJdTceU)PZ>;jn_pNbNYdP>D1d>p+Fz5!t9Kpt9`9!19S(WN`oVi!3qIp ze(YDf4nP7c8#g)VjkT*e#tG+7C;E<)MN`J8VsYXeVV;S>SVJ*s6iEhR7Fhn22-H|_ zfCN4@QNu1RV$Aeq-CT;y?s5#yoA z<3?4}YTdMUN&Hv~l&{Q(H5+ySj%A>XWh-Q%zfby;(WHFX_^X-Zb9My@&Kr{^%O{Wa zP6jFFON#VSaiHn|a+uY^A?ZF>ECeVPw8<7rdZmkjE*`}(8zsd8)O=_Wmx`aju>wH~ zN4+Z2a5>f}G(!P49&8cor0c$ph1VXU>XnGn6V1sq4a@!4*gKN z?Tu0h*n9yD9;xy#eTJk^quZoFr7CEtZxN#cNs}UtS1efRRV=s!b7|4czS)^de9Dlt zWnPY8IWc^0o&WZTx*4hRRG2*Lcv?PIKLDf52|*+0BK*c}@R%flx-$a!&PZnjw)5-j zCw}7)uv!D?7hLn!l6?AQ7lqj7)fLzuSw@fi%#f`nnal%V=(!$H1__SK9B6y@d?3E< ztwXLB_fF{`zlG6)M&4TA-CkGc4T_Gu-v!I?Z28rYYr*TMo{rucT^E$c)$tqQ`iyXV zO5XG#X|^1nyOP-ubd9s_je~(T({1;Z9E1y-%@`{qM*Xex(^Ol^@0@svWug4;@a!J) zrO#pB2Ju<2DLaB51YYB`&(z2-b51JHia>fVx*5Vdw?wV7y+H?p4CmuRHuAB)>5}hB zNAq%zA67Oa=eZ9E>R3n2U8QFmMw}Q97Tl*q?(9sJtHkOKj>Tbz;c89zkEoKi*}Yx7UbmM3cu$UKkGkr_xvq^7DDru6?)HeI z{*G3Nu%X7+W{f=l0m4syzTh`Tm->$_TYSiECr%+OG;xx9DW;nuT~OYw%ogm4+Af53 z)iH2nz(*~C0-YR>GQT}ha4dt1E3Y!YW3vovut$F7ey6^k69`>u**<&pI!cfHvVAV{ z)hY69A9dy2m5lmJX_*iNAPHpr4_`CI8q z2AyIJh}cy5<#_Cm*yGbdtNTeZqvacLh^Il-9xhCCT>-5K?c5ED5)UlzYLwIvJbv4| z3)X)bvYns?iY*HT%8lL_Z9t9d5nEn}Cnuit zfW3EAnk$E5Kzt;ViqrD*!PBX7+bpUIuIFFl7`SB?AU$HfD2212B4|$#&<{bx04Km| z`Z1^9ePqm^{|4K3c3gO}?IW$aJ2>*6Kb&l|ON*C=905Tnb`W#g6boe7Mk6GZVpdXQ z85M`2M6_J$ipaZOC9vB5Z-^)2&{%^AtA^0d7j0khuOFM!jIps9P8|HPKuU(aDA+>Z zD?2!60C;2mitQ%qWnM)yC&DtA44*Ziz-hzerU}?N{M~Q8{y%FXsTJRr9SkUzHLEhG zpQLX@_YsWD<#4X+jLr{}3CL%Nf~^__ioRmwSv&V4 zgn4u5Byun8I&{k>k?fe`)-;gWV}RHL#2%me;%CJC=lxBu8vzpV?^O%PdM921?J*hh zEfiBgkxf+Ge)>V=D$iH5RIZIut)a62e{ZM z_Q~_eoW3;tG-!`64NsS^356@^I?e|UfK7fQGaIgiy5n zgB&dxl3wsWMxP-?oW;K7A?ulDRWYZ*y^e={utTcpv{1L zO;m-@(W@RCaPs20#D*&_T^C5qYk8lI_B$~eEY$bnK(7fsBC2xJNtY8H{Sbtlku0j2 z(JMaplghc+Yu8V@!Y;lw?9!yr@@sBK0qP^DQwSI%vN61|MqN~uyRL)OW`a)TxXYGLhoeMN8=U23%^ zq75(U(_7hq=ILFI-+FU>c&)jM;wjT0cr`P+;cJ zj^4HY1FgK0om>0M-M;(3Wpr!*croioAVFCR(UKASvuwE z-APll{MzXDS=eJyCLNNN`n1nl>s!mR)yYm69F0dy+g75_aM_L7KPdgsh^nMzDRW7e z6R)dRm}C~8P|N^D9#L_)IPRiL1m}Q$;Sd*j05B}yD?{#XtO4)k-jb^?y5@PbN40aa zVuoGT`&Q0HKIL2TTk>T<-D416gd$O7^2qjut1-DV;NCmqpch_)gv`S(DIT}vc~JR~ z9F9_tZ9uvUzEOtkhAck-x^6)7XjNo!z+*D();e=lWHW=M$A)7ikdnlOK9nHKkOo4d z$dDB%3%v1)Zda{h(_G-MYUghBE~0Zh`o#uu9R1-&FI_}3Q{r(+tW|ub$oEob@Vckz zVCRaD)7Iayvuc}uwKCWoe-?(hQG!E?#r;^veI{hxEd0V(cor&Gc=keDKv_%+1BxKG zM0@7qNsyq5QXdJ%R)P%K719q8HoRik#n3r0O3E4MDQqM-uBZboi8k3OMU`K+U?nI^ z>qGhYFu(&hH#`?6sgaZ0$!9|KIsO%4nRJ&zztaPz1Zk!QJew3v0Kn>?O-wxKA9YIZ z$}(OXc4?Hf%9csCal6269^h^a%5(3J#1Gi4ECNQn6H`myC=#!b>cl8%o$FfdIpB`| z^aL6XO9o}Ei?^H9@*S~!8*AuUp~--%u^wahV{_homm@a1)5810Ka#X5 z1Y-1~YK?L!21-k^sJKJ#-V|+*y)CN*R^?{ZrQmh5mV`Veo!&a13OdDSqt~H#7lj=9 z)=g1KK(+9E=t5p$P<7~pys+c=ZYQLofZdzwv>r$M>k;XHHbNx#(XlDy9=mnC6Pu8g zChM~yig`?tekv|8xPZo0SsMg~+GO3Zl+w2sNh%=yfa|j};O?qx$kmWjkm-CV!a&g; zA9TsD3+w#V2H*TNr!@@IpIC`G>@rfVX^24K>Mw`B;H{Us*77$_@Af?>sd59$C2xSt z1_q$iC!(qp+uYJ2we^|rAgmD1imuL6ACaE-doZV1H0-h&DngMFHZk}T=?`lS(-(Bo zJ)~y3Ix%=hP}x+}&2AC*MDOhTT9hnV7?eit`Lv2ZVXD0`EgOM}6BHn2cy_GEk$s zRtIRE>)U|0cv-1l<`8qpJ}cZ#cHxWYH;nLb@lft11!K_$I4``(Arr{# zrkHYylp$oQX%vDy0JITzy;gXti>CF9HQo?#Qq%{nFF18yyiMB{g6>!F2VLjAcjAV(1k?|6#yAJ|-o!jBEumyNP*WN4^vmU-rl? z6Y|sYm;F+$Ki!V)KObncD97GfT}d{xbBmq0YOCG^;JYXW7`?VraT(#A(H+WTz~)#s z_w1YLQA_6Jiw*>21?JJGmfm=Kaxp&Ahk3x7YYU%W-`*Z@? z?UCfQ9%#QH@;v{hD4x0vO-Y-?CzWmXXIf){ev9S>PuTmF==o$*l`R1oCx(rMfXw-6 zsXpsw8U8iU@um87Lj<9gUJe9&-QFeKDz^sy2>ISb*+|y-us0AaJn!t%v{mzrW~J$0 z78j7pu^bgAM!{th<8gvwj#A*f#1;E&^Lijo3T_E~y7iPLVy!Q7K_E-SYFN$!Aqh7( z_=bl@v69yyZIdpGxdCVIyZ6xP?p5wx(|8n>zMl5OW z3Z9UQ?4~FuF6intL79$Xv=q5X#Z|dsjgGzy5Hp)veokN;(AA(s&VxvuG>r#k>?Iyp zG>WP>Rc?kf>n{IZsOrIx@DpANmXS95JcvYTVvQn&b4h$jjFA>3pV6n;cL#MtkS2Hf zz7Q;gZ{S`Lo*`T4)gh3|t)}r&8b!KClN{`6B< zMK;ixL773f;J+^*6}kQjMia>-m9``IE>Cth4xP{3|7LPw2^ z)YW`UxmOEQ(bs<0MMtDv7@&1==64S=bl znN_?d6}J1=2P2QM;pk#XD)3xCkhaP2XbnNRFGsmeHY3S8V6eVw#qRk(`>kcDivzO5TSN66p@!JZDdxW0 zu!|ww5A{OZBXoQ%u%1DKH%7K`qC%dBVHb>p4{+}Vrp-(vi^KH~_(Hfsw#ZYRJ{Oou z<>fQ6WUOz_5nfl=fcWm59UK%%2AT`J{EBQx5SD~tSsH3+R|_AAv%~vHwNPykA*&Ap znM24VV0Od!32oj9pz)p$D8?s%PNuk+S?fRSvQ@B&Dbm*%JrFlQ(V2RbD7Ve%8a+GN zW!nHm%XTX6w_PIhwmu8pb8Dwx7H?MG1e?DQ7^${J-I;wLsAhVz3gwG8E7Q6CM9Z({ z?WFO{kV`F{C>Ysl;Nw;~$H4^;wtV}nUsx75SpN3M2oO21;hzDiav!6K`q{Zj7fBX7 z@7sC*GP2KPN-LolsOs5__0L*9csu+j0q1=7f(xusBu>HYI?kU9UA1BVv66YrSw5?c zSu*7sK`YeDJtn#vR^+BpCXwS(WJxXo59p*)qjc1CHlA#TiCtjb-pS7P{5Dbaq8#hOBh<_SyFW5A*8$HHy=Utf{Tu zC2yP!O7g0o$SN*&CXUsJ*>o{G6F2RBx!Al&$7u`vEL0Drc;GsyQ<4;%0Oi7&Gls*C z5?o>z(aCaTqEcta^m28;-ujq6>k6yn1i^J-`b4BS*+FE)7}zxWR^7}BB9nfWnf+sP zh`b`}CkueenK9_mCF})Xxtgxz6bPDC$j5do`VQ2-Z)55`%3{*IxcxkGI~VE1E|sWJ8HXgN6uyQDTWP1oH%}CVM2%KltXfBOxK~wmk^fE6;)HS;Us#OZ< zg;1e}D@ok1;dRsZC%TZ|`ft?yZ+7fHEBo!dzW)8+N33|U-i8y~1`B!{B_0^y049JG z=#j#nI#k`jiWrSUctoCyHESXK+=n>eq-6-JbxR+a(@Hq4M{0pWBAH2Ms`=IYibyR# zHymwPH?8$lpLgG@#84YX@vjTFBV`2CSLr8@gTYI}p`9{1I|dXF1khFRm_Fja*N*j3 zW@BhAE=HWvXMUPJ&tAKhaWO`KW>WFlvg#<~(l$r-^AfU~T^z)TZOa9dh42xIfy$X0 zDsHj+9sY50MXc_W><`y(U|cqZm%{0hbaFKE?X+`E9CFIGbyPwLCaQ(0k^e) zid!D_K)lVX$~_ZWwM+b0aB#trLto}B^Xik`7c7MENt5dFKka!d=e^c%T~r;MmLaGQ z-U9stk7=ybtQOvx)do{o8^@7>6&o>O)3`tHX3Lhy_uUHRFO6s!ytiFUZm_dDPK={L z6CB;6n7b6|q~gv)(_1bFyWWt*Mx%H{Dug(Z^>VQfu2wt*N(Z~)cs@5XxK(zab9G)V zohpAs(g;e>7brWzN(5_X_KWvIbwgL+IzgVtGD#W_7zU=Hibmh;o*C_`rBkqxu2FJ3 z>=-SuWEXcTSX-90eisIM`W|II`ry1NNEP6@U!7lu-XF86><7z;{yFL7tezP zw0uzr{*HOvT4l2eJTl;4>7I51677v&ag&>kzl~7*$`*}(!|WF_?!Hqj|In6c&WQnM zVVX;LEh)Ul&G2s%WT!Dhy_-BD39o59D-_MD!I&=J0>Qy)MQ%eb$mmxf*b{<^yFK(V zsQv`eS1w-BE8P)FzI{=ZH~ko$5ssCkJ<>uN6J{`Jk_|z4=PquoxJ!5@q)%!O z=n2H6E5277Tz3f}HH>Xgost!98pSRU->RS0AgSP#c+`L}z_1G{3n0m- zqki%R!rHCgDyi6dyz8`U|niegeIvVw{` zpsWwldED>@@YV#UaJq!B7KfC;gazfE!p1tVKD~aTwpf96a$8#FOmk1!X;&4r+_j&!I!dY4J|4tEi>(FA+(Qp`eSfPdLn9-X4p3#Eb znN#-#IWg9{%86Iu9VS-g3dLNa$OS46$(OWz@N1z)3E8V~Z49&!TRcm+gI-IeU`X!2Xc=je(^^u3S=K%*>B76*LhTL(qq4N`Qsl41@`!&S}% zi#xN2&hrM`e&@Eo_WfW_qw)FGb+Ul08q4@NaUExwiSfy!7!5_TskpPsLD6OtN=GGy^{8YOtLh#uq}iFCZr!yXBG^I*Rrx9}v(M&wtY zO0Swzp0P<>J283yETbIaG>S>4$Z{&q;QyimJSsNj086cIVrm!6$Hp44nBOxI`>}tx z^T%Hs&BivjId70#?4o;4?9B|BK%$#sx+wA)71sei48j20_y;`egD}R2Ijc6=3E-&N z3uV^}Inb7(1zOPtVXh>HE~N32EuQ-WGh`3==e?Q~=Y{A9WdqmounU?cU;-HMOyhNu z-2w4}>Y&}wZ+B6&U0A694D34$R10rU zxP7sDoXrQq!TXlyp)uk@lW%HHZZh{@UXi)Uf@kybkPGtEV6;CgFo)Jc{@?^rqhjeL z|J&hRpsUaaX6d#J`cY1U>OyUB$D8#(O#DV?tixR!s}W^p8XOF$Npp5pE>jvOtCMkm zUqi~-O;%2v835k%QT|Ik#Q^vIK`O3~Y>(I=J|e!xdAil4`r_>eZ#BI$`2D5xF;j3D zrX9oC-Lg9WW)+Ycdu7Y-^QwgvK@E~Kiatn(*7 z)48K__AUFq(c-*ysQ0Jjq!T9vI!$cIRf+)yi6$xz6V@djT0YJy)uBanb?BLp0wutJ zzdUNdvrVC;d&4qh&{hrpEO1Cek3ClJ)$oc1x#7T%7Fpz8OK+!hf(is!;fFO`7k-7t<(c&MNAz3fF3+>w{!VOfEI2j@38) zq(Ryeh!fm%VNtElkR5f^i5Jbh0!6-Rbcc|%vyOwlF{m4IMl>&*^&fCpTIG2;2<~Z= z0G+Hdg>*)M(izkMwaPX^k;Sk}1yoq%c&N7ql>*<;{x{W!e9pV?aBC4ihy-?|V&%4} z70}faA9Gx~{JRyvt%bj(MuFtkDZpqA`iZ!CjkL)M=f}rr6wSOWm1a^h>$_nT-QVpt_4N_Qdho1zkzgW z`VB1TT^zHL2ZD;ofUBpY+3CH@|Ay$K6bdfR(|LYpMOf063$juUn&a4k<=OcSk2nH2 zKe#)fri}C06VE4S$&icF783%6{^)pK3dO9TNRlC&t!I~RaLtFTQ%=Z4ta_du3@e8C z*_$Td2&QL+m?wO-e zAfHr^C|!Oa;0q9DJ_squEY3r>-bmwF-chg1^qSWmiVXXB72b_qPfe2g;2Q4%sF*-y z%tAV!TN;T^!18<9=v7G95PUKa_Ca9MO=5J7`omHndwI$0MAXH(TIjB6X0pWi`}9LULx!x5gLNB; zJk&*8C@AqLP&R;PhM$dMc$qU^LZ_!bCwloTBMj%a7Ox~FPV8SFH?d3mDF%8vcTsT* zj)DK16O`;(puFI{jkFO=7GzS3=b!r5UGG=9{ul zfWDX|b1uQPDIARWT=GwV5MjLFxW^!%?uN1_x<0s(c@lscEl))F^iwnJXYZ`AVPnrt zV;{Afbz^12sK2Dn44rHQ&*G&aN64@Xs|Pr-bpx@hQ7NrdidjjKWmFtmkW=Z!v}ec6E!-`j6I&CUrMl+>th4kG@F!lJk;bd?%kjvXTR&?N z*+i%CAcUmFA7S__W7TN^2;hxrv8(~l*mlN#N5_IBJKN#6Yw_lpMyP%K*S3RXixWc) z*w06qlYJDkn@A521~4p{2^+DX^b&}FxDVE|Cs%u z^Zd`V_AD#tOfJqG4l!3>aM~g+3wjDiqz%$qdWGPLtc%{rNfg|FO*6HKz9H%WQ;`E5 zz3M}-0p;Z&Mh}LpCdy6#c)@^-10OGVqJ2)Y!^hyKzX{?RO^ZvQO9DxBV$+guVp=jN zW;I1pu~M{4A5q2%Eu`$Nh`df3CC5dJJ@L%MpyBD*S&fUQw@klshg_;cyM4RDhFms-y3fIBjr5R9qEC|V5zdfHb8w?H!E=Ln$fe=6 z?esDkFjZr(LxyZ2wZ_Z1SvMaxn>dED3MAFR574I`upYP>Dz}!a^z1rl(b^QdW(;^? z?g8$u@_PaTF;I4_uUYPuB{LXC&`FmQ9gVV;2AHd}I2fJ8($+HR**76Qf**#nDKI%% zjXwMNH~75XL(-wxUN??&{!we5&6`OJyYmt=yz0g8D2-2O>2FZfqpluNgz_)1u zu267uW}YJTL1^j>Sy%L%F(p>Uy$-?Z0r$k<;egxT_`vE=R8rg-JK&zm1&B0CFs^gk z`@tMkbHxkvDtU;yrGmRq8z0pxLTX7JzXOQ&tNhOTp7qrz^2kj&oI0Rr=iYa3lYy`s z_q4EXMxFnysp(P;B+YKh_bS!<1H0%Xw^sywEvJiCzUz5gn2B0HGH} zKhVA%mfjBQBOeUJq;ru*r$M?Vutkj4DJMi-7_N&*W}&s`Ppr9wYS}}*Obc6ZpnUe=#ed=9!?x3vLNcQV}2?(hpzHNfE-1& z)^67&bJCSL9#^Io(0@+_$6QY)hv+%_-FrV+jC>tXQRao*U`3pK5i)wfA%l!$iQt0| zKEUNY{J{Uc^5~MkMXfWh8hb_Zsumh@P_1K^zp)S_ThYvgX)Pn9U8Y~S*>AwDdi`10 z9IKvZuZx~s{%x1J1gq1o*)3?pE(EE|Be4%b_3)^F3Qr@~l1@?o)WV43h!vzXJYUpi zkJbeiU%0c5HLLlBN*w&bA8euAa^m^L0_7I!{`5+lAya8+3f3r2N*4+$BFp)0ilY&| zvg33kxHlQHdy>2Uu!6!$_Wg7pq&^3@8b+VCz}Vqkak_jf-6d4tf(2K>+v(rW4NngL zV2_W1Hx32Kw$Km59*3zrBbpSQ5%_`XRlZdb%cHPFe*DuA3)ES)iZZ*11^!Jhjr)iH zeOzK(U!DBkdpF2xCthFeFp2tXq8K1m+Car+%Ry9ER!Og9P(%zU)fBpjD!<`?X4N8+ zBxsv68~|!_(2tuT>++4Cwn=r&H{V&5W2Cb+#0J zIdK=Cg`%%zz|z;ODi#dHZ1g@wc8j*sLvxx{WuO?ppvkAnr;vLbHdNPt`Y^WcyGy=X z@#EX`@&6Z9IrJ4-gX9=p4BQ0`l1Jo<%wZrKAGDs@4Lc#}na8v?MD0==Ey}6r={(Zo z#1GasMUQ9u)Lvlx%o68$G)Qyg#r{2X^LLwK zk*idrDCPFh*JiAn(Ev5Z$EDT6d{1op%qQiX#gZo9Iv$2}9&(RE>ND~6K~a5>;o4>u z=E&+eJEiA{;m8(vN(UdGx9*Ki5Rx?i-H_vSEkOu@kp&tuTS?6N(w2$Rq0i zvG*l#O{M4F9`PKK7eh7z$r*5g2!c4W7%FP9bGzIwxA(Twd)s?&=ic6FrHk#?cG{*h zU*~I!xZ#GNpaL4AvdAKc;DSH^MR9bLWl-ZT2%@7Xf+B+o-}5BFA(2Q9BsAJve;rQF zdCwa-@BcjSv;LpxOnUQ_RS^6}$071u6hkU%!W#oF=*>o+(tei)(Q^3j2A{5l*}S_7 zJaeAZ%kg|%7-q0Tni;5AK$#gAmgAj0Jr}r^HOkHYXXg*N#7s$3tW_M8)sSpDW{NIS zz2yt1!|UbC0x+BZG3gN9k{#u#tH23}49+WNUU7pWLq?7BodXYN0>QMgZWwtv2=82_ zDuR?UZJ4@S{s;;g(IJmn{OFZT^eUNtnR$XaZ^mDySag|rI5$R!Nozac#k(XuH)Zb| zU@IDE5}gyR6jk^1K2ezPw`uLa@Y4t;(WUh^}d==z;Gx`BeCP9K4*}%WG8> z1|uI&0sjeamq+$2Xkjnd=Y0m!9J}bd?w1*?L$3n$d!r|gsB%x_RZn-=&S%_mM!)ng zY*}_a!A(%zn)z;-mkB7B-aP*vN#Z7`oLIHmX9lYbiUmiI2KtmoUSzs`kVRV`)iLEx zaEZJZn6w|xsAe|%uj4m|F6DR2`xVRP-U`1H+zz3Uqi*WD8J{^uzrUnr(T2Kw% z@2H=R+snWIbrVb`cK`G)Ip)Oj3Q4l6p3|WNuxEBG#e>) z14Y(R(fhpH=sd|`$R$-H01Va>+sFiqC<_^Gp|gZFa2RLFTG>Bgv0IN6_s# zCGuzd>yifwtkOBH%yC@bX0vs)5l=&R8MFHx^`SX!-xn_RO^Gf9RE80Vk1g6uE$rV!Xn&O*18&WESl_G~rmQgIc z0R)a(@WFXG2@-13s)|mfNsvUH`cXOA;l#TO5V{S?UoWCqXvdpJMYjl!1eU4xglD@I z1r&m7u3vu0RgJVv2YfKc4=dv!`vK!mU9wc^9sZ^eLwc1?&>@N!-|*Ikt%l^8bh=5Y zz9+*(cdfXC{!q_ic7&Z#G+N9Il^(|9!PU0UPfnPbFgf%m)T>@a+~RR4_<{nmK_AQF znOXLU%Lgei1ixbENZrX?wn=_CwIz;5$rr4Vl zxnb<*fJwl);#4MGjJj+)jTCKA{B%QpL*5wJ7>H~qNEFyOzYk*GKoW?RydNu#wvX{- z5@{AeU&b89_z@?-ge_f2fxHnCkCTC*ph1Q=7#Vx8*kZs1Nso7XAa`7=BFFy*lv}ha z9vJQ$cT}aQI6ctK8VEGRN;vd){A%9d+OV+tTt(P{e)3r->)F&>?DglBG z$b3;M+4aoui0=FTy(_LJtkr&@>MvyZYlF3&W>`z2*hGrNQ_&3&3P3kI8qfnSn2rY_ z0rb9Kk`|Nk7F#yMV))f_*x8%vx`(g&o`lP6Ea61bk|e+JUR%@RjJ(7STg8`7%xzDi zudlu=G|7qgiYI?S)^U>)PVC3eHIozDC^iLu+331xOp$0IUHq-m&F(;A49s0CJxWCt zAuH$RFdIpm;2Hm}phd9R{oweq2969S^F8CG2(aTtHuc0;zi5KTJGUFt$l+Iv_(X38 zf|C?mOOYC5>Q$HQa6r9ljIc4}&Kt|8+?S}Elwb?~-)$N~S-Y z(ar0CjvhFwjm{Ng;Zl>bNtsXS0(Y`#5@Fp6Bx{Lo^3Ku~Zg_hps4;imz-lQq_Lk14 zFjS2_r4#rM+}C&_Hwmgcqo+FN3K??%+3;zUFtRRxJU@&`wDiWk=(Km@3{NXQH@y;9 z7`Sd^1zsydG9ZybM<1SAi|jXsGe~k=riy2hCu7YLjvb#ZYZM3><(hN5%{oH5x7Gi~ zviF$NdgJoVM2m)E(^>Y!gKq zsOWQ^mA>oQd!e{}U^w}qylq-5)XLV=)pRdd%w>Ul2v)SUDsT@%XB;w9?20T_f$@Vy zOZx&6fOG3!=)KUGDTkrx?UB4&Q6I7;a=-<>dPsEI1SIQNe0ht>rVq;Sx{1RA+1GQ$ zvBF|0)@NdQRGW0bB~3X1wRu>-8P9Zj7fY<##=`=KI}UBws;mte_BR;*Bb}I#|5N!-H-=H4@6_<{#Q>_GgbnP_tZn|BT7^zWi z3dwT4Bt7GG40vtFXTO4l;fjvo*V6uHFGl|G;j2zEQ3wsk3!sJ@t7+9cd|L#a0thst zBMcRkEZ1>cWW(}sSZVt&=Sybfv>#3Qcz%U0e{+%e>Jd)_fSUaV1-`D4oy>_l)y z^aR?fX~pCF(2ak$O^^O1kH}G%xIKEOWn^8H-zpZEc9Jih-*S|!;kJ`Fu}fvQ8MIR= zb_+$4z?u$tOt%P(rD)GYH$>eL+gx+~kP$6kx(e7w%AmBVK~U`I%-W&2vvB1qPQ20Vh^7tjN7Pp6HQ>EeD++pSqu&kJKD7!oPvwvkeoDUH{=`|Jl3$ zF(=3*F!ps6#gc3%PA)iRCISvpEM(yqQPFy)3s_%}Sv(hNkgB9PihB2h(C)2qT8*S| z!AZq+$-&u=$>NNK&1A7YN(c2rSVdhxXF;nb3dSlq^G zk(}4%z;uOpjD@?yi4_$q+#Tn^1o{(_T0N3iNvl-1Bw9&{q>x`Cxj9ahi$j29E4YW; zGA@VF|1C=7n`FwYl<%G(D<_c*GnulPVmDDFfr?J2JKVGBn{IKW!@Y;Uecmlb7t$cp zhNVxAqwt?S1-1=;yRA=jywjFn6LI1G>t6)F>a~Fxsslbqw>;p2Nj^xh*#H?trII^} zcrj)QCHZATK2j00W*+7i;O=)xs&#D$H?}nec0wCqIEEp%BiK?Q8N-*_a4RPSPpsV3 z@J(-%1ZjHrz!H+;#QN+(Gtrhqu~`(^MMXc9Zh+eILiySGeF`9Wu7SEP^pYZDA=Z;( z##xfV!o4Qxa|bsX`qMI0ErKM!0T(oMVucmq8W`X>z>FR%cGQH1!44;^O!!TH`TPYY ztoScmn@0}4VhIN4%ThXF0|A& zd{dg+fQumm&bQA42DH1ue@2PGF~ZAIJ`MAE80^?^$~wHx{Gn>rPfbwz^`+fD>6e4wrbVzaqyW5nz+O-xYK4jt%y2NJ(AK+R`fL)i+(Uyz=ohr!D+|(NG|CMy zVH{QAX52b~Kw+lZ)Epdga`E@5aLWY-2E4E&X0nMZ(ebDo28bh_0#sT@ouuX7JC+G_ zPMfk|B_*zh^vg3;Tcy}Q=OMk3Y!ANbuM?eR^-;Z{Edo4K7Fol)7h2+-;XX=nV#Awl zF@!4=EN*w$N9#}+E@7cFsCWN2OhPGW#q95sJ>I*(yk%f?#e{On5XrFtF$2CPQy=7*lbX)L%%O^7)MLbqK zV5_9pB%QPxyh?O^h*=NY`#7MfK%eVe)g(UY@&@dNQf z>odY*;&hqQT63CtY=b$8oQXdV^!>oiBy#?wKKFYPH;I5zGeou6O0mfl*+@mF2Bg#5 zy^;5;orcnzRs~Y|!^VzgFH}$HWNp%?o{ltPET0Rf6&Oc94!0F>+R*gQJ$I;Aoi+3L z^yLQA`<`q)`z*8wxwkRv4TFs87|7t)0{b=!6+N3Pyy8|64vHx@ZO2+MN6Qk9=>2>v zuzU^d{FxhiA6EV6rYMvAXbYQ~O3IwreRIJ~u$`hr~Jh=5egb2yCH>b=h5DktW zY*Lrxx1PNs>6_d^U-Zc5Cr_@0;_l?hLxAdd@Qr{?v&)|Yl^ghe+Bf}e%WNAfX)5Tx zFy>HW`xrciRdKYx|MS0V$dOHte#rRoI^^`|o!4&teZagQxR|!L{Lg;zA&D8zP3Xk2 zZH?KMmq4-WDYAx&-t4>c*;idJL7Q^)SV1@rN<2?U|W!1{Q9@*Km6$by!-oK zN>)gDsE`k4US(x^kB;{EtS<6&qdOuyT!bhX*b;oob3{&(; zE9W!;1wDiD`Te}ML^)Ak`TT;*S^%(BJCfR3!Bx<;c?0jyEY|hs#=@gvS z%WH)o0oZV-9|!0C(*=9o^iiEsJ^VK%Xs=r{IXCY72sRgb*m)gCoX)zC7|y%;o>?1jr8jC1Lr!YpORKxkZ0EWcTG3N+Wferg?Wi+J!WDt#XrE~eAXD)n^<{&m=fq?ASy z?tI>B(ooN79XboNe~-ox2G)@hXut&xD{UCI8^vaT5I3E@ zZ(hBtMu~kTbA@pbK}YL76LOewM*HYbquSOiw>k6ABT6z`2FQQp;c2<8y(Wr}l}%75 zxB)Kf>}Z%}VrTA9u8 zGM6l4`eY0Ks@H$vA10gNb?fI>{!CU6XMsDf2Z3apfhdJy!4%j;Mep?Mj@aqf>%L`X zEcC@`qfxrFDmuc|n55MvSC@wGRBiD4%xC<>3`YdmvUqkGV_nN$-S*>p<-}>|$x4xSG4XA{o z(!bX8qwQbnqATHXe|SSt2Nfcx$m+?^_#7ydt9Tzr-V0diUJoRwom1*$_IchfdpM(9 ziA6U@yfzBXelwYrO|ekBvjfZj?(hqtD-Eg_ zfOiQ0m^^%$-W6_od2s}?#tgVXGz$ho_*{(j5lF>)K~&9*Llc5g0^T5m?1F47lLShi z{Flgtl+|h1e@51GLy8l}5%SECvYld6DYAu%z7Kti^ij7-#=?!$Yv~n}(jqI~K+4^f zllFRM(MNfD<>N?j90sNU@m;^%@uJYUFr4vWWj*bE+U27S;;)$Sv8YtgPO6<)U%6(s z;tdpgh9ak^=o)bM0aZgpt0GQ#z^52^$&8*72nI+oQBoVR3D}~rxGzW5H8)qR_gM|& zv*_f>XybsL0?A_rXGx|io8Ir%;1SQnxGgDz5P8Xrvu|anvLKL+e22g))k&X}BZt%~ z;T^?*%izCky3b?1R~Iyb#M!n=?}QrdAkN_8j1pUvm|@0nMG>cMbHerq|M@#h0uiST zHdvtq+3a@9d!0|cAvr>=3(_XTIuhXW4NC#@4}bA7}kkLL=Gk!Q5e z2yWnbW7$bxPm{p;UiJ3#WD~ay$%)PCA~W%{i(;Wo(sn8u)3J=ru+)8$XQE{uhZeL6 znv^;Sr8J7v<*F`{L0^*`2mK`z+_^Y$YuEQhl{P0Vb zYKzl~5G#tvo`1d$#+W{;T+%r83DGJ$L^{w#w|i>cI%q8e))|zrh^PPPp=e3a{K=3>y)xDHr4^X6-ieAAl55oG& zELwfX6Ac=ybHifqgS?mpm<`$rg}pc7xE0!+T=P?3_r}C{BQ}clx_@&o3s=v%R3C;V}r0810;vSljHFSc!bNpF6;Z{1QWxX)>iI~ME?SY%lj z`AW5)uu{aCsk$XQ9Hvn&&R?XX@5&b&xcb0Br=t&wOXs6a+)1~1j3XFE2EfpF-5yJ8 z+Z@*4{$K5%R$-~Oyi(Ax(iwIycwRFNrAWQfN?)DGkP3&%qi6{|p<{$PrXi$a8kX8i zGE|qO>2!LqI&n%xI7$(FCBIo)9Bwcf3K_@R)Yf4Vy>vsiWsmXKBwfx*M{@M0VI5S4 z>Tm!?2r)ep|Ba(G8HA5*a2Xsuem`eNW)xbs)Z+pQCr-hImf}OyjChJ&OOZG#8gvWr zpP*S(GQAs|a3=1a@q6UPZ?XvxDBnVOgux`)v+C*(`L9za4$VnjM!I{#eA^$2~Gmbnp%75Vo9 zV&)oW!h-4+0nVgR;`~`O=2@=e#|XPZlIE2NIWV21UQi#a zM<3O!=q2$X58XFI@yq_mz1}ZuZ=->j1>)@T(NdgQhuAP~7s_|tRWD2Z#c88LRy2}! zd7Y0TWgqkRckxQ+t9#}Yt4hP4@J7pYX!%TS{%kWSY@X3R137_WVsdNom8m9K^zFQ# z{DN$I#kQ-%X4_Rh#X`TUEGjxLupk^OdVnS|S%DeY2IHntw2PMv)MLG&%|LC3e1$2q zAXkCj?54&JB`4*Vnbk6AgCA9IyKst-qac5Q~xOQnZ>U^8?U>kzuW zeD5D4O>o(gE!UC*PP}zN3%wzl%yEi6N|7om`Xkrg&|mJNep%u6%M9w5ErL|(vt7y@0I@OclxMbCQ-j!N#SjO*h1m_$%bvJcilGu&o0XX?PBS3O0c4QN?%g~^kV4G60Sfp-;(ZLy#p z30ri5S43EQXp9mc4!3=UU-pzozQ!3JBTs3U5uETb@yN1m4=od&tYiosg_PAoXd+Vu zS^N5^HDn`+5gsB>gZdS>=C+3x1r&?&n4QuVK^H%cIUN2#RPjdhf=7~5^zAv0UlPuX zVSM%M^1xp<38%$y-gw&w^R87{sHOk|RB6U6&`&@Ffw0*GERBxP zPhN%~+jF5_K8yXY-fFv95Mt6%RwsAuC3~D$M>%Dtqf}5V=(7irFJcSz!FDP}SZzpa zYzWaYCNtt&cRiU-F@&)^)*V-o`It6c^KM^-&*r<5%9h?9$;p0c|eq_lR=CpvY z!Vgx&Tw)W17Z)0;kQ$LE_(HC)dn0?J+K5)j^W!#Cr^w zy+iEa{S*uF;zBCAdipb;GpeVO64yJPW=twe26-iOfbHOx3|y*$IvS zu`-++ZYr02sP#AD<~M~k8^|sv#!a;uZi*=uDyjES(XEOm&rBe?xaECVv~T{v+#`Wo zrHTBfKD}&8&_~cfWIw4AU=h3VdIsbhbj`>So>ppIlNGBydw@#=@9FeDL2Gzhr<@Gi z4iRRp+YXQY4(JeUC)8M3`_Z4~gc@px%U8Z`$@ZUVwm4fTHi;sMRP-_?8I}Mm_%Nc9 zP1neG@!F;#RS9@Pjde8RD)qT6z~~<*_d7W)fb&*Oski>mJC<2(Tnu36HSMi6E!g9> z_?kou;pRL^3fro9OuD@FQD;a3xNQ5u_K9cKGCNh{32`n6A5X<*dF+0TCHwJ`oo0dt@PHV|n zNpnW3w(Njb_L6jS*#5x3#k%8Cj=1*YxZXp*%KF+56NQW*itPWPU(3nUtPDx5bmk5tXgwNl`2%a!^{lPP@dRPvl{m3NonN$n5D2C&(93owu zj?@-F1PWb2py0DK@_cw3U1T4}r{(ev^VyaaH?vlUnZpsMI0}rP1w5T*vR_r}S67nl zPHevdAKMVW#~zA>28h{Ibgr-yEYWgE#;7^ zEmfK%UV^NUSm&23L@H%aKg)k_m*A$ z^*@^Ma{MQk?~)pBrG8GFC);9%t|p3YpvW028WW68LDfww`-B0L`(CI5I( z6*3&-%v^RnJAv>y8b>&=8LV~B>fUDt-Dn^6oFL1bSPbnj1Ii|fO`ynnD*7bQTVUEk z5u`0VQYG+Xg;)Kd&s0P=Z~S#y8`s(9^|9ALPQr%z>4Fc$mUUcCYu#9>*}DGiCcwp$ zFE*1>s?KMRXKH}4OJg;^aauMn*5@kSG_8ker2BYRLTc%90iO9}T7_GzPuGii%l2A^ z!FghrdJ~d2hEU!ncr?9xUT<`8tF3IQXWB0`{ zPRMOZoR;F{-8p^=KND8+2YzsXXeSY!8CLdEEHwAarJ`e~etG_)KHe(V0a2gg6xreX zNUjamN2La2fRT_FnJZaAu0o%&yK~E@^(m5klX&~&4|)2iCttiT$OzvdUT!GMY!O`b zucDhgTRrirLk!R{L&ia`G$+V7uY&hKM;jMgS~X5<{8}-$9?RQkHEut#0oZ`|tf1I; zQ@~LkSYV>}hyD*e{Fs2|aFV0ZW7a|P%``11%uaAScz&X#!fHj+L$7v*>h>JiYOaVb zGY5PyUtFWy=~p^GSJa*hT_^r<2hzr~vztiqA?}sg}_ciS@%io)_j-+!l zWSmzRk_xlUvXEl)DUw4)@AK{@`lw8L^OSy<67LN@6>h20yr~bS?E-GRdeJJ=Q5>_u zCu!cwx#H=GVU2B&(M_^WP<@W{w4QPCZ4|2JMnK=7rz@{LfV?^CL&# ze!|1RoA3O6G_`|{>!W0uuBJIB1kE+6ByA> z36BTHLzu49yHlRyozF+-dpy$~ZlpX)6XwzVb1{2Bqr@xu(z9>vr0c{faL@k0Ly&m; zz`fGfI754QI{gexx5Yj5L!WZN<|#X)G~Rb5i}7x|2n`z`VJ;DD_%zaT+j5=L<~Y$y zTYAefK=*1YB&l=M1=1 zN!zBiDw;|5tZqd-Qx>>~v?{72SIg=>u=ymi`(0dkW{O66P_a!G=f(wZj)0X_~Bx=hvmhb0TwYvRba!q0V5rBNk^X;fwY`lxjJ6Ysl$88bF|UWOfOhkKP^7ilKT znGJr~bPvh#U+Jz<;x#U~;ILHYLq`KB^BElTX@)eU=vAjOS<7TYai=j#i7ngm8Fh@X zLxla4GF6%09kQM|C|2@>SDEC=aa0_Iw0(FsKkSjbLxCN#FM|w8fIjh0NR4L=-#E(1 z^oGSv~o2 zF*4vXObWIWC%6IcBS^M7H~j|W@MnRRf;tT{&8m^d1UlKukbxA9 za2>6k*5p$$t%F9>=fkDKTrp}Z>Wqb%^i#;6djKQKR8V3JqgsFo^r4{9(=-xVQtk=Q zP+`l-2B7`xRmdV$E87ly^| zFX})xzcCg{i=98d@N(t-weiPqo1msx8}Ln1=fo)r9cJLUMzNPEa)F9|LV6-jPhKs+ zQgV&58S<5YTra3%ArO4gc>av*t-1LTePDiHT=?k}BNtAUV4GK#Dpi0c5c1CAl`ikg z(1oQ4Ji_fj8}_j>)hpHOS_n2FO;+rWZ1%+K8s*R#^-+zUjSvmNyN{EL3kyj#6E9xQ z7KGQS>ScYNwTwRMx@48Gk_`Ti8YL=EFXH2cg@F@)1I!HfYrJrdla!^tR=V{)z6oKo zQoehFtbA>R8MM+Kq7QAR*i95kprS8HyP)MtuJ~|3@?66tV7Gqe-xdyCk`p6zG}2NR$oH_v zDk}eO-$`#*`!(7W{8V{6FEBNiSPOu(KGj#a%@t%IcxbY3RteKo{$Oy_6 zK5)m8IE^uUF&e|oc0i43_u`DIWgpl4lO-vJ(|V4ruys@kYC+N1$V=uOf+bepk2no_ zj`|bA89yq799&Q2$o#V6zP587<`{{I(sCmYoT1_-gtmTxO36BDb z&7;U}Dmu-ti{7EC;b*(ulC^;u*vC5|PgZE54iT9CydTeP3h#te%|x%q5YOK2(HwFt zs1aE7+kl|BS%K{f^CAy{f?Y4G3v5RzBm!*?h6wc=6nmW_SEy*T#P4_>oVo^!`?3us0q>$q2ciuQY z9n6a-bb;4WSB>)FoSO>tw3g3Gr}r{C-ZlSBux9o$5LU*{vrp&^^Wnn{?>ynM1$*c9 zyL8CXgy{js;PXvIg+z_Dr$!SfO{mW0{kdCV$n@+Gbtq$m=R(x&!PsWKRe`))Nq%aK zRaZ%MqMd#zLxJu#T<5pIh-6G44f9zd9C4N@mp_$GaT9Z0x=)hit7E!(o2Q(RV?Xcfk_U=aP+`5?zfSBRp)@>-xL8=$W^l}T_4aRl zXxUPkOR&;;U0YCYKU9jQP%LO@o6u&$go*)|bC4vVqn99!x}m=|imgY%thid2RzD0nxJHhpWgZ`b~8f7oZ9_6ZYTar;>gxgn)YBV>D zEc(h<=1ehRq-|m7HInAU7%4TwNFK%Rrbs3gjR{hyb7a$)&X6k{aKZAT13p@QDUgL3 zlQ4m2$0ZLUKN#z2kYOA!R)`v`dI#Ca#}E%zYFf3cN3{@UH;oD{LxhIeWOn?(2`v-; zF!55k(u9_%Q$H#vJDeCTm1bxuqFBhg%%h@nfX68d6evjR#R|)%!YXl+UW6AVszm^8ZVx}516bM6WjseWA!ar4!DpqJ@AzM8@P-WGoQo9 zj(~xwIALbul=|QMTPE^4t(V_QYHl}lXUP>N2-U{COf+vyCH3v}N$F~T@FZepHzp?} z&%upK12}S^GCR@hnl6eU?( zTN|MA`F>+@WNl7+oIkR^c#>t3E*JNj^F|wpmGs@`y>PSXr|xH_?BM;4q=zktu*V8> zo#W}EqXCvY3a?Z)zZGtU7Rh7>M1 zwQ&3#?zTDv2ZaIGbsY70>&nOX7G6!3nD!O$vz7DIyF7}R8{U_s7Z)N)cBXgA;zx^i@K#4wxEF}V!&z*LI2eS& z^Pi($W7}`Ej)viK%vXqiPJP3Kkf0T_zfX?8HkwDP8J^Bj>?ahdr@+IOZz!I|^21D( zo{bkjoKek`$ah0#cEz-Gx@f9KStKt6LL#7KpVcDR;oGVx3tTb9$RxOq-OFUrS3>G} ztAs_gF;};rU*mH;yjyXCZlH0lR>eTzCy^Q@`e};;4+K`YrLe04j(8m9Ym~dgdqsV` zV-c17Iv@1h9t)~b6?_gqBKx=j z%Zc@rdNW`hp;+*|lvB|qqE=VzPIpO~DpOauf!9rdj5LoV7``f|j|5+F+dut)Pn-0H zx51eHSd^^LgGy|LRSXcqyIr!77sH#G1`V zW&)#(Vh>SdKNStdLLqHZq}M?ID^eirl<7bKq>Jn2#Ugb%qXVxRrW7>EjsztuvcPY< zY<9mqF%pwb(Em2za+KG~WI$)bI}k29KD}ZZdf{?~4dA4`FIo#zZuf3dHj2h59fqzd zC-m5`y9}LsEXUl?!%WO$Eai%ogy)@pg$s~x#po8t5K0qsQhpi8wUOBZSAye)J(j-` z?&YumWBev=w{uzm=iS^du3r0hMwQ$izC93|cEGl@V& z!^6v2;eGaN|M1mWrv2tmAGVg0-A=sU95>r<4pHoWitMAJvF~wd_{~{RqV01@dS;5c zz6f4;u)ze!Qw`9j%)s>m128(~dMAe0#&a{jtb4fmN-mR5bi9TZ^ zp$ydlA8@w<&6Ck%n@(SV(9I=js{(!d8YO1MCh?90>HuHQLdOSRHsERK_J_v~!_d%) z5Bx{{EwM>O*&*26OyZr`PXZMCLoyCD6q`notyDDfWtB`vM;LbYXpmtaT4VKR59ybq z=L2it(HP%JXMrg`J}b_}^J`lAp$p@21oCUV_XQ4%m zxjY50cfet&gEmFeNBuUD?L;{ zH*#>J!5lN7vw-@*HYmJ6=X{Y{k=tK@_s`&|cc3O>21br3A1C`}!i0`Le2x08Np{4o zj-O9DCJmEV>%8zLYbOpBbk8WZk0Ot$=seeKI%S5jkpi;(A)$ABs5VFk`LYLmHcjq{ z&?t`tcF=$6{NJ`z~f?x-WsAftd%T(!flSI8|UV-RvK#QO| zA`T|Z6`~_2j>5T+?btY5jPNpZkzMLlKmRxs5dr&{&f9pHJ*!EnZUsWQwMWW`T8qs&}&ao8v8vh-K$7JfF(gg@2#)sOMyEb8C6>8rg_@~Z_g=_Cy?{w zl%_Wd6xeU6Uow+*dE)Q&iQB7Qua1lum`AGxqq#WZpv+)6<961r(uMTQNl{@yNd1uR zj@UH68}>0A?@*u0;7W`F4t6$(2Y(aupVr|poG$Z=@TIXblL$Hat$(^oHgel?oLCv! zYbIQFQfxX!wo%b-VC$6!tqi*)&5@|};A=4)GyF2=o%LxZSH^% zn%7xjjz`t#Amw;er7Zu}pBWQKe)Cb!39^hENSrt}w8IP}niUv-djl^?k0{VPyVtXux1KfbD8>?jBML0Yz?cyh`v4F2 zB4eg(ytrMC^&Snfd?r(s1UgECpwrdZQ07zjv-8n z_`TmmC&*18OP~6y-;iV{23fHg$TBGw8fv9e(dEjL>5#jEuEaLTSIAV|DhaN;R60SMe2e=sAQ0^I)``xtRq{5t(jh-icGLa*4SsR#UCAc#Y2`rZ zY2_Xd8-xhAhRPvs-#YSW>rnFlX7Oa62{tYPF0o{V6N{H!W)MlH*o_p~Kt-RO zUjWS271NGS-xrL*AnXc;DVq(l0cFd0WiUKn1e-DN{EqQCu#SP@%%bmg-(G0RV9O;? z;KV)`D}1;0qSSy^d7XIAtlfe8!z)7aBlJ;q;-;|E%0%$Ao$%Q_C0Wrny~Y2kWIwRo zR)ll`18$83+U701#HOf9W@iUh`()9bQ|fpnhB56?#dLl|E!_tr_KT{0cGFs6cSL4b zu79;pCX6ePKl2$JtB=}BcJb;3rSnUGWj6!*2Q|_UY|(LsfsW1i8!B+P9S`%ZBcEDD zO`4)GNt!K*K^Ms8Nu+>|=IF_C{WwT^>;5;K1)Fl9dJ@@5oPSEF(C*;m-Sl zo{a6-KGg)2Z|D8w7i60gZ%K#E5S&l3ITXpFqVIY@-nQx{bj?HzYnd(L6XK_e#*jO2 zH1o9LWx*%p%Ya;4AC<$L4{Y>Y;kB99CV0kgRbXLiJX66>@Z7X*!w3p9MAhC&3ojSrSrnwTFf#=_(%>Vk&=4B<|#EiZE#*>nx9s@Mcm}s>UlFN z&9CXrRKG+fR=6v|a5ie=;@Iz`e>wh}17ELv>ui*9)DHJHL4E|@Sx0Ykt%E(qI9C7t z4PP4gdJ47o?e;~3W4C*4^E8gVxUf@E;{7Oc=(lB3?8Q(q0+ys{Zb~KYWPL^ zU2=I3zwwydbT8l)2i60@!{VKE-4`o^a#V{e+|vBup3UEC{K<-c#@?`1LvH$P(PPr- z{ro1a_d_52_zK1>KYO_U%a-_a+DZZ|@$qJ0fhvL?t=Pv5I^ex^O|p@6M`(qY*c{$N zsa8_qn->``uIK5a*7~IcX@wPTJ!G?cm*^m{{U46J%&eGI4c?h9*-AHk)TiJ?*7{`$ zTLkUUNihQqx5afbjdbw<%vwNa`6aN&LA2#X=0)C|w=SZZe=`cPvE4g~cT&|s;~93F;8yvSODd8pfF z9E0}2mlT`Dt0S{i6~65-Jz9^S|6)mgACX&gcS>JDtPP{7Q_}1D3nNn`I7?%kTnb(@sfz9h?V4LlubQM~?s>exGT(6|??4w6hK)})S_X;S+sb{|Cwu{l*8T?OrX6ZltnD|`lAFkcHYYAT^1vygX1 zR4<5!F4D_E`e0rtdZ&#Q*8?s(C>+Kf0^PiGBo``%FGH#54&N(~KT#rS7o7CMw&Zq? zXmP>FpiwaVPy3?9I!1;ODHA`LRy^4Jo=9Z|G0(LMh_a`J z)%oCAjdEW=yHAV|($TR_CZ8HFnZp@M_C<`<*&gbGe{Sh9s_&OWIUd-?T>bd%y!7bG@q24 z^d`<~eYpMV_OC7LC%H%tC)Q!DR9ogS=cM|mNA6dJtA$y-TwXnt!R1Fl9$$D)_yfrv z&mwV`N0q!c^tN21+@We6r#!wDRGsRz&JD99mehpRgl(GKF)qa2W-DW>I5WFx96oRpXRAvmtg_^@bXrhYA=>Qz z#x+S>xISthQztH_GiDXjD<;JVb~IN-qSD>#8-;=4coO8(zh zswO%WJIO3_NgHv-Y7LyY;EbevW$J(!6_^xFcA{SBO#MrP>pold@ zNVO3wgc_nwS4{9;g2Yi3)7J4G%aL=cB6*f-$#`%`aKguP$>B(380RJv ze)>ngNp{Ri`R)m_(uw16KsVQ(u6DLs$+#Q zQ-E|On?{oTaarQQ#p1TqgLgUpCpR4=TC?YhrO(`H)lw_Y{Bu*5FR1asi2OB4Q&iqR zH7&X(IUDuK^bL?Lvd-&Jkd{3WndRE&rH?uH{Z) zP-*vt-}{b9bj1Cp{1G|LP3>@EpZOg#L34#-FH+<@6`mVv8{wgiOuZN>5rlId94Rf7NS@1lyj}XjS0nSmx(R} zhqfxz#SpQ{RITH+L%s-Rk)+dDwP+ZH#cWvA)}+j&D~VANQlD1#KydSxYpQHFea5vu zMBN+O916vfz<#IB3(S`St+;T&r7EcX&2)$+Co5LDArm6kT$FJ{lxH`UjPrN}`u)i;M?AqRRF6`i5lPxg>jqF0vAKcT!Q zLsbK)o)m?_DKj)myrL6zfbG3hcri%5-FL4Wo>Hgz?epI5Rc5%Z2ky6B_OEMxzACn3 zWmu*4`8@VM_(S-=(!Vlc<@WpyGs$gkSaD(@vdWCo@H2|-p-8u};TWb)aWG0ORk`*Wd7_Pgx$d^ioYFX(pSymW?N3(i^c_r)NECK5JCM3zg0&03~> zZdu?Sg2XJHbc=^NgWwWk7y35(>4H9wPFEvErsH?EQ3@z~%cB?O)6~vRu=nHf!aP4U%A;Ii9eq!>T9F2Q zn;pLJqg;1~r#D&GdxtB$xG<@1zh&odF1oN2d;6^P3GWCf5XA)7@LT7$tA-voMV$q{ z%eL??`Y^9fTm>r8fTTonb9NWmGV?^>Qs$XYyGkE*JG6`b=!>UG6~TK;BpT(#g=bV* zyp?mCqVUK#bG@KlbvDX=5F9PmhE<8|m|^#auPnPgmN29fqsK}ZQp>am_Cj6hLH|p zl7D?R^=)#Lo09Cr2350}LG=m6)>Gs(6|HBELNzf`S*lAt@vvTAxIhiXrOfri0!xGnRjIUTpw2Rb2cL(LkNX zidh-J=yO4_SY5~iE(54q&yTR*t}^`d9`&jZKb_sqa?)!ig#1jzTDo$a79&=iGe|g8 z!#gWGFYllmJsY80z_!3Ra!t|*6h!){X2rq4Rl+>T@sDLwy;=nO=5O`sg?RWD=9Z)) z|!? zY6XNDK?i}jAx0d})A`i+RT9IBaD&E+OX#RGoS-pr>TmzE?p5Dv5`>x1q2m-=F>PzW zEu!AwS3M)sV;i{PlKhrK`ab^K7x1w%SA2`vD?|tIGeNa@91Li#7#UO}9EFU}=jDWv z30pQ#UNhaK-sfJT zES~Hb`NK`n41&nw=jUs0wiAx)_0|0q5~7o&OW`vGPBM~0voPKcSJjlyjjp` z8P*^ogR~nOAF3-w1)>A<&q$1Izqvd{4u9E;L%FWOi!(Up45u}4UcIsK|NZA*d`tr8 zow^BUNvabIoI_>;CzoQO1TBM##`>WYqRyhb=H_@`P@ItWMD)Slf}!SWrkl6J_Yl2G zSR!h5O@Y4A*TW!3H+$B8s6l%e(I(#nO|P=)ZC>kVqsHMdRC7hKr37&dhQWB_5+661Pc3~cTqbbhNRbtf%r!;-jNIwhrRPL@+}fjz#|5GSpa;#JzLvQz-Q`i`IM58Y z=v;s_-2IL^x096rVJ!4+&MP4y!)~aW^SpfoX0aIa_c49pu^dvRj69N9)TSBl{t>q0 z!chnqe%^~89{#?3bU0iq*&nR%AumMQ{r~t<+3;`F~SgkD~>oO z-@_J&I_M=x7YghOqFc}sIGesNTB}emn{`2SBoI#=hQ>0LWC>Y(yfeN%CSztQ8V zM6bkly_{6h;h(PgS=%vSxxH9?HlKFPpSW!@AMGjs?F18IRJ#s-K%TjfG_!?SPqAw# zvI_0VT%c1_w}8G{=#l2Y{MDB}f{`ElOK*4BMQ)GXY3cvszyGh~?`vnly0B~@AZ}Hx zkAxK1GN|%up^uY(xf)w4j<>^PWOzT99u9-@a1%fK*>PWglbrZXVa*1z%ZYcEYBK>= zOtH{pbq^I?O2D*^s%38|@`U=RLS8S4Wsj)3R5fHw%Q}KJkOK~$17eti;g)ht=QxQuD*T%> zYb|L3t+4NDne~3lg0p8;$s1(YGrd8!oL|j65M?)#szc5Eds1&Mn)gJW+(?~@J!%$$ol1{z?*`EE?Aqqn{E*dPJ>>3?Ejn% z=TP~%OFD=5Vi$%+$3T8DUWYk`sH5}Z<>|xLRVxvq{7C!7V=h4su325)=X*<^tzYM z*C>1D?C@RB_C&-$vTcmxSmIt>RBlju@qS00iSZ9R9%|Z+cKBs%CIwEc-F|Ggzvw8o zj3S4q=pLx->X&yYR{3d&M3SG8C?m_a!}ql0yyQ4ZQ|=3X z=3mX)N&Ikwgf07reL%uZx&&^MOt(yW<04(0IBVZZHugU6HW~wz>2!Lq+NgkZx;~ty zfpmj&EV`wG z?u`sDyF4&nj*9aCLuMaf|H-Cfg_XX29!r*F!F7xcR2otq+Gz|J6w7<1@4x^4$TDk` z;ITz6y==@&U}L;wT^w~SoXPy$1gnY9{(S~%du=SKWoA|Nk0`c_BKN6iEu^?z5v>nX zZ;sf@CV*UGIgmUA3W9Cusi%s5H%=wYZIuC1?58$oRJh$ zqdYPro34>pc$^PZ7mE(648lxoLTxPR;~8*^wt z-xr{ZRJVu5Gp&jof2gDO!tbc#)$rBYurwvS7+8sG%&^(Ys|YVs9SLlZL2eFu(shQ2 zc(S5FhWu9X%x1S+-s^no{or>^RwQFR+}{Y1FGH*0T1raOyO!Z;D{6woBZrX2v^|ySOR3qmN;(nv_$kr&6 zNNNk4no7!?Se?FLwyT|@*b@{v1~e~pS{NqV>cQGZ=gfdh--0AwJW$DC27JN>h6fr`K^lX z&^~u#J6Am^5UH<&)Jj$)kOb1o)Eb;WD+1QT+YhrFzkKAKYZw%Ag87Iq{O|*Y`Pvar zafA7vcj`VI3#qsBz9+;AsrNPiR^^lMVs9*Mt&wzjZ-D%&fweGK_`p4G zJneRFrAZ^Mzx@XMsONwB15A>p_N|SjB;&PFevX*Qr4ot-!jA%DdIZKO*LoaN#ZgPj zRF(74tK34j2|g9H%g-~Y3Rg2r?z$#W#^Pltnx4NTaY{vaJOjK`z!g*xZlD2IuL#OC zs6$EO()lrSk#}g*)fZNrvkY)si6d{|t@3#QWXC=94TU~xgI^Y% zrbuMgGMjyyqIQu-@?H?p4PVM;o{AqzmkPB?WJ3Pvi#b7foS~|v_jz9q-^%NN!t_4g z`k>o$9?tEYt5G&ik9Po7scnyY_^o}&+0i`Fm)^rk1#sR>);A?aJQ=LP1hSEp+n@URWI;8D3?k6Wu^Uu3-$Q`S;uw5yD*v2y`W7|k09{3LR~ z1xdS(1ZIWz0WHRW3%Wqihbg*z8rX+WdiJ)=~K-nSJqQ-ajkh^7VlD8hNHX|j)gY1?{?ywpRa z)0hiQZTp>Y6sH$`;(KjdlnYIgqUi_AGfAlvJ5|n`z55?g>~V@5rJ~Q1Qc2CMUN>aO zEn@e0HVAr2hvN3^R%XSd5_yeJhp0AS<)jL?6_e_Gu1fJ(A9ayk>Xqi#=dKmULuZ>B zpby(3#x4?lyjnVg&hamiUxscGIv$S0^L3&+A9Vj`(|ClVuX`IurPFxNgE5tmaKqA2 zT-k7JJ9J(c$q7r;SO4R;r@v^z(mS^s)5zgTL~piaCn>g;A~l%x-ApRyX@I9TM%W>0 zq+6MCkp}w3B=I)*p|7eEII%wi^|fC9fFIAOjczWZ4XNJY+agGectUr{TH$&;b5>FS z#K>)QVR(Va7&6SBRpF*(P6IdXfXgK*a#z;MJAhU;o*8iI2yK(*Fj_t`fob8~s5c6S zt{E)r3yWdn2~KO~ye8+wpe+eX6PTh-{ivMmm_#bgz*I!B1r*7nqAy8z3uB;O7_-~J zxtU30HVT#yce*Aku1fml<$`S9cA(`~R|VzFz)BF~WenkI#lXYz)OC4SpZkCdCb;1} zcm2D0*yUr?*s!)#_QH&{`A2)66K0%O?JfM^hjD=>%*;yvYa+?w=C^ZRix6by&;qst z6kAM@y;SsWkJ9-i{(pY|)R(S8i=r%EcZ5D_w?{m4j%0tiX;D0EVaQjP?zhXM^h?c# z&zZtq9%Wy;2FFcNH=tiv!J-?A&P9*p=uW`T0*9}G@t&zHbn_&4l&qXqjg z_Z8KxdN{)boSt`NE6L7Dq{{4x@1s~qS;?oOyA|tMVAAnur`NMbNL7&95LyveicXPZ z;#2HJussG`nv@CrHlUfr^x8__M-Y`UM2Rm+v5lS@BeFL_YavcsuIeHkq5(;jD#_3O zelxp0?O~qSSWdkr(&kqet4b_G)m+@YPAqM#MBEdZbE2*DuY2gD)&|`Le*Z^uNc9VU z%2WXhVr|$hQsUh>Zu!Fo0CGY2SP;wg-W7LUHd$7vyx8l=d6h3>rJM#0?IoGQbvz@1 z2%6yu{PUny@A4>OZg^i@Xn!I0(ns&C>u+7cg{%JM-zp$ilRQ@SD#%Iv92yc8H*DGf)M%#N3>Lu1{XpJe>lq=c;49_vf4 zy*AqGGc#T4KE>Xp$Zab6nxu1Xl{`*Z&CjIwOLBNA%2Mi1@Fj5J92fSxq>CHEultr# z8-c~W-{rAl9lsrpN~s*)L+K82h1*5?I9*C@=I!y^%u9JAJz%NuWO(+}3P=`ilkf7^ zM?Ic`g_)_+a?ws!uf$Bc4g3?z(~xbak7^MVg5KLL(efKSv*`Rqn`h-O%7fmJ^|HD^ zeN;OARPva_@+(4~@wNP=)Iq5}>XM>Ga9t1|b}0B@;9l78;VG;bpch{U5OTgNBEkR->V2x=EmGdyz)lTOHVya~2yy4eu z*av36leR#@e&1&MOmvrbx~1mAMX7!DbQ-M32VM`%rfracS5S#mRJUj~Mnca`dE}O? zc=W$XQ2?O+u9Ls?Gi^`#13x%Gv`!o;(3x#Gd;dRsUjo-udhYERJRx~8WFwfI0ThWK z$S8}Uq879>({-k|x9z>X?d^AO7in*&(`DM5+Rn5czzr8R1sBi&$|8s;AS%l0D7efh zIH)M1f*>G+gG7-*h3|Qipd=E_Aqih}zRs_5a<;&E|L1w1_gVgbirGVvJSt+hUbcqAO=1~)A*YGCylC0>>E%4=`v`vM{5X)3EXOFn zT4BTVpM?Wr_5)^nBC7$hU*r6mFylADdZY*)!*`N>c5GaLB6SpJY8Ax*=hR^;;&V_V z%Hgzw4g_iw?c~<+7Gir$JH!^pW1Y_m+f>$vn%*y0$I>93xu_aQb&o(^ zT3Z5v6}-qX1OaFc_6AkZSbEzmY~@48wjfswTxYs!t|Q7=MwQGhS7>Bs-3O)?%YZ2M z;o?Dg3cm_m;+)yqUB$9izS^s3b`}iCXkDeeNE@J68(2%2bC*8C!+_d zInL7X%jGW8HTA3w)iPGtZO4-4%nv6Uv#+x8nC*BGWrC|Uhh8~rg}+Ky{2S>ML)`$| zrJK4j@u61aiU*ys{P2K#CGV^|I8=ks4mTuM?VDBU*2w|x);z&x=)}ZRY_#~Y+@y_V zkSyQlR<$|SSmn`f>#$7dHl|2U3Nl>V6mcS*0fM*Kv52D;0l=!!Dxdw}O-D@wA7|Zc6)-3njuYK+btfsh0r436_m^N7k{65ZbT*2Na#7eX3N7 z*+RjXMRZCJ2VGutR#d>*=cX!>cg^jZyAO(bdgTKWEEPH^*cenV{bJrm*PAw#*D@GK zwpcQ_*oGf|=-iMf*P~-|qTeO5i5)uZI3Ky+03F#B1AM;eR79+EIt}$>bMN?n!jBYe z^KIqp_C!u=2*a&=&cz|c?t7-yhTIdM4_oe!`_e>_({SH8Ix*Bnlh4Q)vjxf%6EkuL z+x)!ol{5MC^*~9J|5qL<8H+N)j^XgB0UVA~40N?sQW4$$wW1925DLC}AXtxE+ub4s zM?$ahH-xl+_|cGTt1^LiLbgp&L3f1V@p7j<;tc0D*+YJt=Ob@aux|{*v#*OYxL5o( zKrzi;Zj{prpq{TJSS30HojKTFgZr?IaGPVER{*0`V46tPFYFgag++xyczCyO zrt+{4FwitWp8M{+RbjxifX;3j9YY{?jAXCbbR~fmmZdm&=5Sd8l4su7!h@zk@>j*C zImV?`cI)w)sQJQp@EVW%zC#W>xupvV-7r%y=u!dkNnGO`_3fwQffTWdU#o3=?Q`4b z)FfHsp^Agj>HatcQhgc=~eF02w22 z{K^L#cKnIfTbBOsdz~iHb{q~&H?SfbC?=jF>#2yXKe#4t3@hU7TeQ`$j(>+v^J%Ah z0xt?}Y9F3{@VLi3P>g-8E%&kl1$F(~fBb{7lCTNE-o3H}2&pD{_0S(y&{zSBgmo3* zD+0@b=VN$qju9i9mU({l{7=mEY0_6L#*KdI&A-n7o!)>f*>l{3+a(UpZ4MYM-7Ci{jm4AcFDP3$b84>w_r4ZE2qrgSP2GYRMt!?*(qDdMT>9 zRdDyYBKui7Cp$Pdcn7CESk+AGrI-=LxGqNN>ZK`D&k&?byF@MtwJgF&wDflj@B)!A zt-S#6l}(qw3Qm#e5ie4(CiDW?D}#oN3f~Gk*Q(cXJ$I$g_8I%QJ<|TU!wzSatDG^bU<)^9Z1cw2XpVgwtM4@pg)y9)Nq?OB zm7B(0Dt7A;o9H`fP*l-1o^ia}Qk*eSxw=Jq!2KpSN^nb-q|6h=PCYd3Fo?6^k&1w1 zInoaF&n;b$r0nrcmT&PmG)=AOlU8b{-VQJ@!wxXt(ltO|j@wp86l>rZ&9+ZIS{FO) zfb-TD?-Ur@6%()TUFRFn1&1ruPO++BY7{)!0}D?0R{3F<+yfdHH23Bva)Ah}FBEL7 zb^j_;cVj;artW0j{j5g8eq-pGIm$O))lWO#zbju#woE1m46OSuih;uZ9XjhCFW)>X zSsoVdY@Myq!SZ3;(TJ(GlD(iGzVO*K5}bW)IaGtU*f zo${PiNZ2_bX>i)e$7NWyMZ6EJ$AAqoSl`@&3&JM<+|&8p?jn+F$4efdpcobQ zFQpimC5x$u@_<7#4te%Nb8aH<^WgFs+q|~>Uh!)PHnqPYY~8=DUy}bH>wmF(N!+(L zzVTo&_E6VC-74OvMSgkFcHcOthIw*Fyqh(BEjDO)$|@M|uwe9!FJ(7Ndh!1-@P-~c z;idczQfy%_!p;Mt# zK9x^3C}J5D!LE|Dadcmy3Aq~DBQ5hQ11JBHJlZ9eN%gp+WLrIBRZJOU%;2*zYM%X> z6{aRtpZjZ@aV^BtZI<@ypb--l4^e_XxAhCRyC%!~IZ0j_Gjce2-W9?%uKn^j(H&qW zC}PrpvbB6V{C`FyoO@8DR($HY!6jSSHFy2O6&}{KK4Xm6_`x&A&8)b~GTnA}ERNg9L+(pXXSQz`SQX)|Iy1XZjhB6$G_%A$EjK)H>HU9bKYm}{ z{pN9vrydiJnA2;BnjI$Wc%CUUz(fJXfQaBOD&inFH5BT&gNrz|^aHZXrE*3M{QJ2A zLy2wjeCTcI0S%^p(n}u&4|j#ME}$_ims6wF89=(ca;M~={E#(pk1D9MvjqjQc5uwI zg$7=S(TB%lY}JjwS6zF{xWeuw=?O4Vdx!ZVo%YNK!*DsX0c0*ta>=CUO<;T5>O3tw8a>V_HSI3TVkLfJXZDOula>$Y5LuKwDJkb}m5G;dGN5 zO*DMW&X)%qhCY{LBr5FV>Da-2OlI*sY&njMVD({}(BLYkcFNUBBA1 zP2MT3akP56@Oi*87v8qaM2z)2?!5lRp}dDb9Z9CFb&+c6)~(5OSAURdOhgdeIFk}O~8Uge`!T;c5ZMx8;`CsWq&TcPMH zO{8o2M~aYYUitL3^9yBlQeEqL8ja*Kd$}JAS2=f*JW-yOWB-eA4S((YOmRE#={F0L zyxJv5CHjD~TcGL@YC=$zOs%*CGOD?T+-T;5=<#G?vyIVUVr3&I&HBlwpBamCy(FHF z2?;NiF4i}IVx%ZHSsVZxD8B{1i4esqnU|3MI2OA*eSZtZJm#d z6KQmu=!P3wvQMB{GS(|csnRv(Ch&AUb$9_(PGK<@Hu5$C87N=u0p9W z?2#Lf^_w8{tL4=*xPcQdTj4H=t|j6a|HNS;_}Ao?j8LGjxp{RA&Op0foA| zR&bH53nlE-ik0qHf$A*>s%TZ6(p*lx6ywpWgW438MB|yxsR+0o_ySh%c@KuoHD)nq zW6|u_Z8Sy3zKhP7(JUMYSv$YZF^cohRjoMUyVBzdXY;I%!0j`7pz@Z9>I zwITzMHYiVofI$!43;}~W=W3`O)g9{#eJHkInPOXmW|r+5^CWLy=bgs78*IR2$I(O+ zDjhq6kkLO$dN-s>dE6CrDU+bnyK`QKa}~5`UI_ZgU1h;AkNJ$Bxswero_QlHAN;_5 z>R-R2pXr*u*_BQ9vs=R1v6VSz;MgCd7!W+GpdyOgdg*wP$6C6A-v_NUSccO-cPk`NK_{{&QqchJWt-=y<+3Ym$a(YKb^9-e6T#==9zWgwj>uHAz(bFd_C>f6)xXi>(lsU+K98nhfSK+7JAf%b|VDk2(msbXCkT-#i?`exG6m$N^+sIq@QB z0_NMw$XJLMn+2-}%3Hdoj<~vWF1|;SUK+d<8Q>+8VlpUTo{p&T-0;=Ak{#3g=;7~P z0iU$cZHHuwq+vd`QKWiR$ugZs&M0^4ncJZp`Bdy%XSZ4qeofHGa0NbgE5>Ec~&lc+cCHG0B;IoB|VFS)xAWxPi zN`nHu!*g&?vi#w~9uE`-J#1xWx(#dlUvQfV_#ZZVA{19qfDJrzP|P^ujuEN6|4P^ zlVJz&=|&snc#2t1ku^|&r_|P#d?fft(Cwo01I%S(kc3gM|75<}~%Iy@BLXpjwpX=sz^ROmK+W|5kg5;H?gT}ItI$^@RxW4k8j8r<`VPietmpSR;7G9&n?< zXKLGW}WVd^~>dSq$^IPi2JVi z>4QJ7C{GKk$kv%v6A)P+J-94zH2N}|Z@)QZz5i=^uXS-t?s}4E$6o6R1F!Wk#Q^p1 zekuZMsbic}SYH1REvv={8oADgJvvTD&XimyVw$n%cnj9OWy*zeg6mDLs+9~N+UY*x(M zunaM*%*rI+Y2V_TNk+K*jcek3kXwcl3VO)l&XggC4P=me4ASLy=xv-fx=_{=S}VFs zr!y1yX0jax%bu21s9pA@K@?*R0JFk z$x_yVI*nSPyAs)vTY<4W2P|8Q6PT$iWhn^KV*9fSl?Lv~H1cfa1!zO72&i7H;m1&M z(3nvr0lRk87u{>Pf$W-CGqc(g8;4R{as{=V0f|cO)ew+HcZ$=Uiegl_d79_Pjx&5F@;7m$GZ^`_L2o!Tbzu&)OWc_l zpV#;E+o$v^*jCz_3HAS05c0fF8qIathv&dJyy2HgyZ<5lL}@_YCzZvE}rb ztV7S~1&vuy#(ZG$hBrFGm(>8d0ba3bVTW>n=(xd{$V8rpOk^v6aN$bVP0stZWp62x zQxMl{k}mh(1LtVO$bwfDa72+ouXN4hJ`Asbg6%p!o^7F5{TNJ7ZCHx>r*3Ga9e;WI z-JM@9rYU4`z$VmE2F)gB7IMXLJev*tGsej*Og!UyTOIo7$LmfR_w&3^{)~wppIdGk zStEr0TcGJUpIR;GmgiI1pgDpxmq}chtLO?Y>X&p(R9*=N&@2o*ndNLXVQv9cGsxmDM}bQM$b`Z9N z54xOnSLH|&MVSPG^sb4JueROG$@1|(83^6hQ;*GBJ4`?R7s?rVXL)mou_dwFe31#O zaa)??b-;bl8T8=qb1x9|9opo>z864d5d=qop#`PpQGnc<*%nG)X z{xkaDOBU!|Rp)yrlF7lzt-|r9}_3=Qr&xedO#zpJy@-L0QF}pc@+UElD2txF5Y7ZYSk*| zweoWXZSw|*TJhkko46kZ03Y5nqaQ%#WW+-_E z6)!wo&SdPYi)z3^*#=-E<4^Y$b9YuJjxB?x68B2nZC)YUvnG zv6k$|789pcpllM5<3GR(%9Aer^u!ip@e#Y7nLt^3RI${4ih(Yl0_-zRmM1SplmIbP zwBWuBM*p~9zE=MgDcdnZlPt&jYgHSBlX8_0IE|z@1eZNP?cN4pLRKHOB1_y8SRA6^ zR?hFCmyMK1lqs8pY-!t{HF#Em&0{n8tZ!m>j2&#MV?9oPx$@Y_+&bSo&~DmKUlTWk z-Tvt{5V1!^*^d=Hp*HE|B|ZZnyC`~6ePSO{5?sq<0+`X zz@Vg443v|nAmQ@oLJ*CHBoRoB=S#3C3T#J%qCGg4X_6%S-;}G$lzk*W{8LdbCktZH znBuYJ@@!-{SPc4O69F3|m)qoS6%osiLqmdWE((a3&kW;WCIm} zRBmaYG~OeJW-h*}Pr7Cb_|Kr|s-+3M>fK3HEo6g_EgX-rN0B8o#=eOS_OgP;q?1wa zd%x(=Ky_W@M%8A9)4bZ{wFFBl)0rAtPtAv!rHMlZ%h;L?Fks@Atf2AAU-}OH^kr0$ zY2cs_ij1`BLCBMn?szXWSzZeexugVN6c*=o!fTL}se*ZCjr^ee7`N6<&AlzH<6aVe zGJTV?je*3{u8f*?OpLmfcAEyz&DhK+E$htC2M$mZDH-F%t@S?{|TAe)r>+Umm@7k;HM@=brGjiCM7%Vr=5fvO)~?qd#=l z8w)|OX&|!W3MdnL7A<7&)UCk&etyBAe9(E7@Crw@`KxX6m9v_iRMocDF*7y>=7Y%i zO51)`l*FGl?kh9Vkfk&4(E_z{}=~WfJ#iAz3cFqRJj=(rhoBYPq9J);rBPg1! zD}RDgXo!E(YvielT5T6#K}e+^sN(WlPH_@lKXk z(0$T|kdHv@v|b8S^DY$usL+A6lgUB%WGi`jf@Jv(c`{fFWQlJkxL7^vdstb)uY^SI zW=@pg7`YqLF548+L!-JW9=|C+3OWjTP(OVkc*p_&OO~JUxXIr$f8|s>ZiUe`XRORI zB!=f$K6>oN%5HTOnXiBLJu|1KdZ~_uRX5EfbIW1c$(fzx6+>avVt@c6y)q5*ZlDVD z(6xs?2fMC7wQhdMAXj@boT}345Uh}@T`IVEk~Zv-C)&zs<)_g}K+RDR06SYqHx%xo zmhPbQ+W8iP$TkqNJb3z&K*;xJ(|>PVAZfQ#lL>LnR;LV68#G!ZKzl{8o>^fc!uC(w zA+x~x^kp{R!wRggZ01UTVZ0EsTTGZ(+GH+D;>LU34(t)`UA-^cwIF^L>9XU*-)e(4r7tL^pCTG60%a}! zj_tGi#U1?P91VBhuQrR2?QxL4>)y_P#47|9-o~&N`9WXMz#!MeRrF?2p7#O&ZHgSu zkVCQCN+6xxyD-fsQcy*&B_mg?nR1&(%GrUC63@>C7wPNbK5|`LOXFuezacC+D4W|w z_HqXZlsxfo!mqKK7ZZl5gciaZ|VtEu@fg`xLtt4|}w zjNMCdqQtqwxMIa_=PDCi3OBhG^bL86)tvhhOaU~Avu5sb?F>#Ar;AU1wK43Z=gF^PGh`bx5wq*Ktj7d^gcU}p>i2)> zGewV)&i`rt2U%ms7|Aif$X1F0JGYUFzy{Z3d9>h?f2!ivoMLQFN6E}yNfN;trnRoI zAZ$6I%OQ5ym;g}dzk(GuCjFsk{qEoCv9V;&aSw9Kj7F4Y_qu2k*iHVd!2Y2>(iK0HUY%Owx`pYue}Ma|Ef86&`8 zPBGB3e#ZNZr_+#XRe&I?T7g>f`n%%gDAw8}De_4mz>knde+onkNT}YR07ND&s0cvT zx!Zxw0Z`A|3Z=V!;$%ow0(COV+2e_oQ-IzT)@nMXS2Sg)!inOn+ue$Tt36vlKJ)UT zJx~vfj51j0TkZMCJ#yaWdFhT7F7PEe^b074IY##Sm6JI@!leh_LvF@TNrfGQ?}~w8 zIzut1DRP2}_}IVLvz9*Ua-Ss0yIxZ(c2Jl8UA0kkz-imeVTV(p1J3<(2Lf9H2b^zm z&(qkvSjj1$uIiq*kF)mG6&~G7bk}!&x9G=3Z>kkrsq5jV!&4+N5U)-kr<5H^ydR!3 z>@XmyoDt)C29#0Dd5DmmL0vp-^j95wW1C~PV~;$ElLqy@<1>Qntl2od<8c^4cGhfH z^4;bay#fz;43TmjMe4OY&_le+8LA;(8&e#TCurqY2yZ~A`$Tv@BNGY~jgsH|{*S8v z{qNuX{LjA?ucDac6p1u}fpLuSI4@xZjaT|Mp6Z&dw>9s7&{jrvv7=P5U$O{NV_=yM zQA`O%il_*Xg(k5KS_qVKZkV>xJ=cF{`;I@>QXCv9g{t72#@%l5krb(o8vjii>`FP4MKP0GdE)nxwUQ3 zm|rrEo&TQr?)Uxl&?)LEj3znkpku!&dve?WItM9cKSc_WfvTNGwz>v|TB*Y9DL6MB zbS$%1c4m%Rv25A0OmY->@9L#kqXnlb_Rs7Al)-^|X@#_1QVE$h@X$w&q(OCuhL2t1 z*lK&4Xt=RVo)eti2*6ZZAy92?f1jT*$-7mlzW0F#`v(p(_>)SKC5;jtoffWy6wP_#=zGTcmrRr z3&&tdH7|+k4&Oeb*W)bI?sUi!Y_Bb5U;$W+wZV0^{=m+5guC>18cWHrDeAUk+hIZx zr`ogGDNU3mR$+$#!UkGKw(-v|u;OyW+#xm>LgwDjTIZGCKGaWDJ#%F6@#c*wt7LErcI*quEtOpS|vzqMfVoUdVh zm=$cO`JXKP*&p;)=#TZkm`y&jV=J`QAPM=9V!ohAKNS(joPxXzc8=6bF>Hp~W|04D zVFyeFA~pL8mjmuY zvcynmr3L{tpm)v_t#^v!L2;p^J1~}6=~o;K-B&vpjN@X3dm5eNtkagjmWW{oq{kd~ zSm(Gw*+aK+keG2m(heb89ldj=GCe3+Ug*~Bw9WAWutCF(MpCKePOIi6K@jlL)Z*}X zE~;O^A65{oYvM)rcvT#yNBDB2AtxO@f9@+%J^W66=dG(`1H1UH9edCF45rK-6a&TO z+o*^rU|22$hUF@H70ISCk*)u*SC+sVfd>@fK;Ai1*{E3O^3Lj@B(GMNdal(in6YUY zuSGMykF`J7g3d2T50XEe>_0)4PbL`#Pj@55Bv2$4=v73e3pPUhJBNPg42pFMBrC@4 z65I48Kl3j&^Km}&9=15hYJ#xe1j>-7>pg`Y6=A3UyNqP8GadH(ml0^o8^u>wNHM@S zwVR4SHoCp=A9DXTI|28{6m5QSN}7*`yPF=G*FoRp=1Yo0y5{2gDo=F4J#ESt66~2O z4%_8|-BWeYpMc%N9Zq&ujzi#|*9aFz%Cmp_sW%^v+nrTbMeCJt>eu+JM4vj+ZoH?hHS(>Cu<3xC?{sy8D)sNQ~&Y_wws7oI2SZ-flsk@8{5NdvpHkfFXy>MZ4hNSY?55$RWq~xS%m?m9p`x zyXN2JpL0FvTQ<-37%>YGPa6UoV8HN}rV%klxg7P$obI={dXzZ$I7E|`lgBvCj;EBJ z2FOXKm<<$($3m4){LA>!WYAf|Upv18L~UZoeONK6;=TIZK-S9k_!vDZEU+^AGMjIk z#>!95-k$cV9v9xfD_=^s*s=9EV1R*L6a(!dJ5YV`V+E)ck=yf97Y=f7`@{%7^@5rt zthUv4?>7h3LEnECv~Q!1m8wa;p4Uesb>%LX3g3dMmlxGZ&-%V&JE*XL#nXnwhCPp_ zVKIhnc*SYPM;{oQke4JUV8Xa;n^_9JYUPwqT?R<9Tq7%j`nFF00*`X1D^9f>+u>zw zdtf$-#=hKUd;Nd+;g%OYwVb6acWRLr%0V0+#ME?}Krv3sGJ##ajy@c`>==|ePULu- zJ=SP1Y&Nn_-l0y#!#`$3cwfj}*P6S~p)Dobz#K{I3p9z*gvfbr}$I-uP@RAr{PB{vPdaA;MlQIxo7~KlN1BJ zJs(jKcO~hHtAe$XNWoEYC#a-e_A8|0y$Ye&DV>)AJ%7a^t^B)^I{pc|oR{vuJ9O#N zBHtS{ZZ!avZVefLMj*8!o4d&a_uZ7AlQf21lb?`ncRNa|wO6PWw*rfNGsN5d%bm)m z-2g_L6A-Fz<@fP6dDM{e?x@za$pW%eD`IIZgpC_t-}=P8^hICzbh|bSlf1wr#ECEv zk|wGU#(T9vAQpr0ZSoW1;}ffl%4V#`J#gc=kk!KHl_@`aV}miZHXFCrj;jJph?hnB zr+U@#ABvm2qkWR)_(>J(+8{l!05@Z(EM_g6&H0Z4^C|02N!n3ogRilSg>K(C0Q8(oK7T)OttAeP9nVW99GX(+ zQs)d&uWaNe&Nhu<4bl>C-JUogjXJ)dYC#+*J0{CfpZlV)Lv|zZ5=h6Ui%!Up!2KLl zE$L>PQqd(K*~GKjWrw|xT*}Jeu?}z$KF>h{%fp$+7oMA{cW9dCTzrotv2$qb*whpm zI5e3Q12r>hXfPG52bJbK;JR#wK2lJp!q7ej`_Y8tc;?V|LVBdC^dKEaMv1Ifn+4cD z1Da8hI>VZpc2AF$B`|sVZr0lq{6E)1=IddSCjYNIQZiQl&yKZK^_yGEM8+bi3z*y}&&JnYcMsim*G-IJk%jGT8>bhff*oLos<2$$?xn^pzc)ww9x{Dj0LQ;Z_e;=sMJ?WhyHI(m?D2YP(UdC|Zyvg6Zpw z@}U?deKoR9>ER%iMh^4_&~{WTOA{5-Cxa`?ydD zHmqS88CG*KV&tBTTyC?Uzdk;z)HqvYw{$5c@;~Ri3xwxnx_@ywd>^t{@w%OP|ePdQ`2z4akgP+Pq7C+P8{R2t1W2Y6;J)`nJSrh{^M;aA@ zyxIp}YnR-eH|V@+c8TbyUt`#%;C4xJPzV2pTSIVT*eT^1@#i6#^m%D1y=itLuR+=m z+nL&JTVsY5PDY!Cr-#<6d)VP5`($C%EIm#fM9PCCi=De_$Meib2B<2b7znQHr6RC- zuoAf1u@6JFnR89N-Zh<5B`KSCh0_JZTm2#J{3KcI)E;_SwBRFm)pc<`q!-u8lb!a; zfPf2?*`ft`f@Qka_%?Y95KQ7Or1gOeqas%^x zMV9Dze4VW~T#h?f{^Vr(kMgf~Ft_}Jjpx6Y#;^^B~v$?Ao--}m+5TWx$}#MteWWdh1*7*H?) z-{%46I)0vaA!LyuVaVU%l|x?w@(L9u3GYrv27N3ofHF4c{o>u?23Yqb@US&x&{?gx zGp7LfqB?-sw1+m+4zRgTSXcAZNlhsVl5+woGU(O`x; zMKQG$Ii^z}LQa%kph3CD-3tO~t^6i=yR6=^ZC*8#IyI4(!rwgW1T=zFX@7RZ+PY1C z$!*ZJF)UHE!ha{Xl|)X<;Z~9TkOYSxc_``pRn= z%3Mw?(;|-)oD1lFRjnwM*E`}~3v8%W0A_MpG7C#%o@5$edEZIf^G7|HV%EehAU*6H zAv-Q~SZBZy^@w5yDRQ5RsB;{U9F*5dqo9T%(mziO4X&thRY4yEW;(T^QBgT#*ddMX zm&Glt19jmnWh04&q+_F^ItZMcd`_MyXLbsw)vFV-2$BBv(i?$WXI9b80cEq1Q620X zP)4b)kct32dC=t)us&hKEa;82^6{IyU)>?eq7OnDaeHtqqrIhCagaM8xh<{cG%5<+ z@X)YByw?}*RrFP50#Dm&i+PN)d6>{lqjSCS+f~?%`>`+|+$pv4i1#s4%{eBHW!gcV z77q_Q^paffUeX9P$FZSyOA>W2Ei1!MkH7KI`rEOYGePTL3KR{?Ne8JS+5UGJG*y5E zP_^z6Vu9W%S(M;&dZSYg7_P=J;7=5P8rnw>ayJB=qlX}KaoxLpdR)MObG$r`Q>Cne zqP;pPh@J|!xwiAmA$fIPdehkg12%^Bu%b;HgCy_b#|5-W_V)(R|MHW;58&xs_E9o1VbNsFtBr&fh37uF-o* zsvk^L?;Qi5DH{r_{!u=?D14CH7`D~#rgMgKpL89kM|zqZnhpiQP@Pw(yvsc_Jqgr5 z(&$6e^F&qr0rE&(&O7cm=-SJ_HLpO}7>2G=1s7c9q_SKzx^S(Dt@4CFb`(#c#H<2;9inpeD)Z!gcaKsOBveb?{ZF6@W~r zkd+j_gQ?10E>}1h%ERm@*45Z#)jo+e0LCx{W{)b16Io44lX^csas5Rv_&pYMNF48^ z`&ZXyQY`$KJbJA$?7l~*c)weWd_cT>PA9*LpfB6&)(%;UHS*8o`(?-6<2coxLk`)> zo6Z`pE$PNAL<@`YXLl+qv{1PRR{YgiOu}x%f+pl44uYg~W7s|CH9l!>Y2GOkjclK< z+RIjj-p?Ej^FZ;;TiE0vJ5XqTywc&z#gU9n1t68EW#>lOI-GQn*|$wzL9dagLIcT0 z*K1xXm8w&k%V|(xL2nl^y%JWo&{JLjr%_AGSfNBt`1Z<_Tm&A*-1&Sdp|+ONC6`N81#1$s=p zS3KpfWIekmul=Ua2y}~#s-oCNF z`H#Go`>@b2;xnb7xhT3a2-3llne z4IOV!K2-x699Yt4o2cotN6Jo_HDI9lVl^-so5huy}#Ot=v_bU)C;U7VTWyw9s&|b0R9`3s7Zbg!_0l9LnoQumUqRoH=ZPxa2IatRkRr(eX*pFP zq~un+u9-40kqZ^JczHaaHr_Ce7lrt*|N1q(L%MR~_kKhwU$7F%W&`i}EX90Gk&{%! zDQ!gk5U0|qYz|h87fk&^)*P_ScO<&r7*^rCmf1yPq;^$#)Uo8i1Lm8{zf-Fq8fgF;2pAcEJWFHoa~?hqH-xF zn<6`@i0q(GNV(HT+zc-Ar^I?K_s5_Z+90SSagOI=X?N0+NAeQStzdHSzf5Htn5AU- zN?r#ji&cdda(ZZ6`m`2dWn3sQ?;_J!nJ@FEU)N)0&rHr8QuWdVPOciDsE%Sjp~xvJ zqI9taYWk68Cli>xDgxHKUxu|kpu?|YP8xEw-|Vzx!GI@rA(cDfRXaf$Ad2(I zU4@aG9B5ENTCrG$Juoc-n|X#wcb0h=dwf>KFYlKcOU9VcT8W+l`F6Lv66_>u2+Lg3 z0CUFqu)EH+^p$DpVq3nckAK@y8)iSyQP)hwrs=nLtb#(wC=jPVNBsJC5x@W6|NhO= zKZ{pU%yNoEPE1$pGoHtV&xogSQHcq%wleQt3DRX9fY$;W^r8h9-_Tj~VTU}ghYMp! z6qKE7c?^Kuu~%LYoDXI1x-&||7vD&p^3KTVHUp8_Xfg4sJ>yAPbchvN?AHm12%mI% zo_=Eb#i#T4kU~4ISUYPl)l^eV1x1cf5v^nc^!IcJR&$>zg_u#|oTSM}2Q&^X`gcFbM3eG#~B+htc?d&I+itLC9YT(5{yRzV^5 z1}M?T?}-$YziJJ{TMF(mF>ev9br-V&HgNC6Yig`eh z0V)ES1c7)^{?HZpTl{x0C1hj3S-(N3JyD@xUMqj=tD8K^fr5F%kM4Z;{F~)YTbJGo z-%71~qapkpofx{sx58IvS8>IL`%k?-?9c!`)EV3!=>S35_$+0X8{YE#f`Z^?ae;@n zU9ypLegUv$US4#5K@+UyK-aaMe`zXqSGUrdK)^gK-0GzV+*2e2)76R|-x67j6R;(Y z{uS>)re$FI5Viz?YCVAS`S$lJX3e1a~^ah*=@&=cYebH zV=7F0ou0v9jn0BSCbwMYavUmVyImnakm8fWE$7~F+sjSoTz7*$F>Y}P@>!)96%Kx2uvwxj+udE@@W7b=QuqRTL22J~*Y47#lJJMNnW1Y{*nOFvQ61srn8 zl`qv3dp81g2iE;((Nz*4(oGD(rj9x~hN_;~ric;bLPq|a5Ms`F-hial3!7yjX)*s3 zXnf6~YaCNy(y)eKq#e|8;A2dGXk@3gkYd`_6c7H4xa%FEpPrp`o}}2ZGkwTl9?GK_ zNIqnOGtEC3ektt4H?D~vievuuba+khKuAkq6fYZkyo%h0y?4=jxh8FT>ia>`J{nmG)1uLApAotV`)e+E=nk>(owSpOf z@HWIpWJkPpTa`2J1rBo6j0)-KibL=q#JojmbdK{%zbyJ{=pAQ7RW{uR0l<2GvFH8J zKGF)D9PPn6#uXGrilGui>ZN#|h8s;Zav+8QY8CDGriW|{g^)?EvIBmLRc{RhwgmP_ z%em!FDgQFET?+Jjs@HTCmftdQqxJdGJv`@UiiaXs1z?GbEKktN&!poa6xbrXDNhf|nT?sshptG%v#eekIS+Xa(}R+ysN#4joRMC7 z)!}KUN$0!_0`f^nTs0_9;o}XrIc7VmKI|mvK`ER%X>rJDqTvoZL{Xh2SDrlt_*XP zy02F&>Ldk1tZ#2nB=EFKJXoNOzjY4ML4G9vuHh`!#*oM6eSGwcesGHd)20EjL z+I$p^x@ZD6HJji6_ZMBASeH?#pfhJeMggmCHo4T%hm|VqgTo<52YVvR(9Iufv>OvJ z$oj8fXED}&YxO(E7Q@7=hD3cjG4Jw#{E+;RVon{*yYcc$|9G#P^Cx;WWA@{pxF@gu zCRUH%eghhRxbby2q2AS4lY0HvB-V~yjok*W#&(KHp~z+`0?EJjyWDkd7S=fS&+VHD zjNU*zUF1{blRy$k!PHMkvoJpdxg}E_hXaNN-RWxj70b5<3Xs;}R2cAbaZ({&XL6D~`bUPO?F*6ZsEKSf% z#7GwP{Yq~=e&4jqo!ppAq70Uz8j9(qNDmc}>eWZDp1hkr>mD!Z-G#I6|+UMYd)6rwgg~@KwLnD z@TlJ|Q1&f$!)qG?+Q5LsiDF+b{FkoqyED_FZ@R>-A)qm=mL3#b5nL4y$upG4cxuH- z(g6g{S3`G_9Ow$ZJF}Xz()Cf`(=+0DhHJzMSn=mmhJoFyZ?~1CGrg|{y{bo@_wUM= zk}Z?T0fVVw7sX^zWQVQ`1&O+k1L-7^bgLCHVdp_^^E8R~DutL?+4KXF_<+0IUV0y} zfSpy=N$-Y43gS39PUTL!f>&@*mhGBd z0Y}|(&I58*kmFR%!H&!afl*T%!yW{#kz=U?az3mF$!cs!*ywZLHOuXY=Xy|DLhW+2 zq6w}w!41+Q!C7uL!?GnfMw9n^M2_zab~uhs_{BHBT;xNj#-f+SL9((^QTy~{`EK7j zBn;7!iB$8DUbR!+DC@RLZLt~S$R12)+|&xEY!?45=|!*StoD55o+7OW){-1LJqR4y zEG0%&aKVr3qj&CxRDb|^GGvi7;Da5V4gyTL2{ObBBjG=6%h*5j==3aQD}C4pF0Tr= zg-v0PfHC|gD~!1DEm)b8Ne^rO@ZKwWM`guo|Kre>NPt^r6xmEX#jK~u8Y<$1BAwUD zZ&17s@oj2X6gu*QFo@n^Fhyzxa7@R{9%w_Dehd1c!Fe7STXleu-icj#!6 z#l;@KYd^*gzm=qj)8&4YS0dWwQVD*_er^$GzYITDb1Hcg1ADW8ZfH~{0Ope>y~+9R zY~%8~mn5}gqF4{vhfqQ|XZAW5t@KQmwm2bAa1$VdJ+&~);5pd!-3hFzoxv$mRSAh@ z(bo(Rhvj)3-z3bV!9gg2yB^nk<+e2xQLhEN;V zr|0AI5br{oF^!Q4PADBit&RsutAVKFAT;BTfBvt*?hNxDCcArD<7`$_lKsX(fA(Lm zzG|#MVz*67CNxFrrLCY*j>*DgIo6sWRfsAj>@=8&^64n{2MnqmbOk?_IZ5t1uLigG zCU?7=Dw%tlxgg0D_ehhJ#UTg1k+Qjxs4hs_Ns^Xv^PsnO$4B8BJ{BvejzPzMjpN$+ zsytDJv|W-P^bm}Q>SIO9RE>1}nFaFzJ}#4>yJ(zfX9e&{dxB~gyXc+WUo?}YB+-sD zNqY^v@C=H9dd^fTLRa>Lq2fKkjS6Uaiy;r3v*`o^y>EB4vkj`7_P|&VI{-l%ER}F| z+(>M16~~E@agMzCm$AagE03J|qGsta;viBUBw6e_a_u*dNUk&Mm^Wd5vKi5SH8JTM`_@X?04PzY6wr%{}6s@vZCOx=w)_}%Z=hwToDSD)9r08?oenJ^zc+45(F)p+DE_QQF z$ZzibcBbA|^#4@4itKo4Y(Dk1158H2j{`vi7Z`6E651OWLc4d?OlgT-Qr+tiKswq-IMU?T&gjMuDmowti zbk|(9VmD`pq?Z0-<|gkQlJ^qjWx_uVImFA?djB+dnl3DWI!lD;SKhbp&>D;zU`>HKA30tApA=iM$eY$a_FcTD4oS14w6@5$OfJFd+!4xkS&t#zzZZbNLR`WH}X%1Tw`|7 zQNK!p9ZsuV)942@o>Am`1J>L%!HqJ=Q8hAWrXCML(WGilp5Otjzs!dJgkXNuEASM! zP3&NW`B#2lS+>krW$h)23!BhnEAv|wgbJ^ySCKPqg;W(S(D}a;Q*$?Z)Xas@=xZio zTf+IxQ}qgV8@n?hf7C(8GDnng990Dk4e>zofgjKxR-vLAO5T(^)$@0FVbLXvx8AbH zZ|;CUdF?IxK69^TXEWsQc!nFVj@Z}?`wdzX6aCkBii!F3_Lib9Jcyy@)&aHbkl^acW;wj6%9Gm3AZY* z%G3RDyG^DE$>nY(Lk{K6iT)WP-QIgLjeN-A%xhOAcOZIuPpS!NlWp?&Oq9#1AtSW( zYQ-jx4J6H{oe#z4qCUD>zS4cCORo2AAG5vN$7X{a2A`mj<<{fxGk3$I;Op~^JI|h` zEwkgP#zdPO3W2Fw=}KTA26D{``T@xlUk}!8wUg_?>_^btCpy-7EJlB<%WbyLv<13w z^W@1FT|TFc(<8+H*LXI#qG~6`*0N|&{Dr7mgR~BcEY%8>7vG`G;p`UIyEnMzaz5g= zacb$yeuc2OjpNuBB`i>3YzAzG2)warRE)s_L!ECC{n41m*KU0l6U@9=oRlUig9em3 z5K_{WF>RbuNZ;Y923*(O1RsKzDdu+$QgOtsV9^I_1oc z^#9!IqS?zUZCSyrwwH;Z#HkSDPZhM&gD%ZtoWyMF7d;tx)>wLSsZF=GrvG)faqGxS61p?d0n;SW zE~P+u5_Fa^K@D8s6NQ4&LunQwMqe;N+kgGEaFHdgBX(>r_87P)X%w@KB3r45-M$6E zO_E%(R#_lfUg);pvmU7j(KAK^rbfs&`XQedn#KWP=@NpZ6Jz`3{PKK7?&lMD1&rnKMCy&L}_KPA5-EQWp6_(?*iA zLU;q#I(Y)D(Qb6V6M~J1gU&6W=vNOK0`0aaRxuieaSgleI-4K=cC6fvSNs=-A22S{ zW;3zaZ}A0`+m0@!d|cFBKjY!TOy~L;bwMj9KiOV2`69u6jbRr&ZqbdQ*rV1DDx{f< zZaP;i8k{;2Q_r{%^?2NEx6!%l!&YNbC^i_eRqVYiWPVna`7 zC+XKFHv6O-U0Z3SUQQ3X;IZ#j{Gv}i>$$21=~2Ip0r*?06+^N^jwQYWpu<%@eb@m* zczb1dVx#*CY58;$xE+@%GWKyS`aHHdD-G%|F=n&`YM)Uh^P4Fqi6V(qM7=Z_hNA~c zR||yuJsOIHS)^&c_MtJP5kX+f3wAVSv*pUzl+Tc*!c@~+e+`^<)jYP8==iP zvUGA_GDv@nMn(?RC;nPhs$qxzIqlQ2DrrPOSjR>RN_NTeT`q{W9J*LGAkixW zUxlt@9YHNh+KnH4EPyngH;@f>49;Z*B>T1bzg*or*kKB?L(Ht47? zQYNT!1?fR)^d6^!+%|a}CziS6TscF}-LE5%La%5UF57`fC{w9rXlR$+mlcO#xfUoB zpqykeT?}p%c4{Lh(XviyF6Ro!T40ZD2i;CDV{bgIj+n7cD{JVOHk$V?d1bulDs7uU z>Z(ol#cOrKQfe#75#@64O2*pP5Y`9{NsVDSp2hBkqCq+he70P_PPg(I$h4VbOQ9so zfWdma&Az}iFviLrQTKCx{7t>#c}4T9&&ctYX41N0V3N*L3^Zb$r6M*{pMLw;kI(-4 z&YMHuZ~Dn*s_E^+;i{^|o2lW~xBfyCzWi&Q;l=+l^mhEum+kvr_Z#cqzV)>Y-{|^w z%#zO+9|*7h+RY_h-`@TDhBqGlcv3q0Pp7uNx$zsSg0JmgQunX-!c{p`>yiy` zHou{MGj{Q&@2r0-`fC~gvEj|KZ&xjD{MNF369Dys$E;Nu_gkK-(Z|3Sfmq|WjDf{QeKXpuK6nxXEY@jsJ)!hUu!^JzXgZG7YX z7 z@%Vmm2UGw;Gy?qNW4vll0OM=cgqa1PdU~d@BFw)3_ty*fdgv^^xcMWp){fJdy9~gP zLNS{ul0-#hhSd2si!ORaNe4({*fC|Z2MSd5&}Gv!+{UoebS1BB+WINgHZK1)7!-EH zZE&egKbW>5WBI-R_WOToUi3xMKClPx-;hUB$Vq?Me*<@LUMv6M!XA&~q7+FFjV%O& z&Nsku**3EWrlO-zVb>#14E=<(!_ngK4t|UvbJ3>R88$=@JFq-u4r~S(c3}DKl-*x5 zQz&4KYnFJ(VK-d}k`ihKrUIh{=^}0A2}Z^?&pH9Q!QFJ%YpVT@$?{&ugwP}MA=%Z? z71A=lG7yY-B#(BnUbt(F=V#f7jB!IN?P4_|_8SN*S$zKkNm0Wp?bZ zKrz)QefLv~Rm0IwDStTGe}XI@D>r1v*~1KjiDM(hBv2$4iNZM?^XgRbl;a$%YG;W<4p`@_;$4Ox~t$xUb{ zw&I)m_(`xFgR$S$$9J>dhUqaF>y@?J^115feWk`)R%~j#?Rb52!~pkuC2}bb~wPuF5VjPUzUJ-(n&~6BmF^M)e{rQq#y| znPj<6-vs9#ZSi*E)cu|v!O(G5EWE-Ql0^!JWN9MRDcKdy0oe|w$me54o~Re%e@R|< zoikvWjH{hKX$H}7v4k%2|FZWka7|s={Nr z@Lwy5l|+u_goKICzkVv`wRiCBZ`NLW?X|w&@Kiy&Q?22M!C-9_tp;=?C!B2H7~>S} znkkmKX?90p5-ECN_5vHNOwBHe*-4QKDjqW71ikb**+E6UsxGLU-xyo%**5MFUCc9| z0k_nY+c7s(4+V{}cNOQQX>m0X_sD5pW9+G*WZo9VQBo39&Z~}2gAKv7ph8}mqQ$3* z?h%_?gG$kq98IRI%$Y{WxlkJkxUAcj(*A$4tEk#>t3@3^!*K^n=5L2xAGa~6QL$cqUb>tq zk4cWW7})@1LFbu@l9!qc0cceS&0G^GQzM6R*PXRHKxj8ws(=FOL>n*nGwA@g)E3N z6}LPN`C;rv*AMbodiAx)_K0H9cE60M9T2dAa>+I&dPbi|=0{a~o`}l_Z2?r$+ZWm_ zZv~mGB&H`4JJhR}C`Xtt;`MNrD0FJxXZSH$!$GqmdYyQ!~0L)S-O)+D!$e_*B>z^R zYKNE$?!)DT#TKi*apbpo+=dg7sQR4W{G07YgWCd$Wm5r5^RnoyaJ{BS*+U~+UIG0; zb!F;k7j5Dg*lAuK57#IiaYDtIA1(OB|Jr6WxGgGJaugD#?GbJWMa79iDayBw@ayjA zI6WH>j`lpVcH96F9P)oBZN*>R77)M&GbCvxm12@9vJCbfq0$yv86N%LCQFhhOa2zH zcs3?rZW!6{I@FpQCcan7jIk}RcUw%b)csegOnxhBycQsJs=9$b$3rQZ2ayAOq)t6E zb5sFh7cwl4gF{2?U?XnGsPBp1WSjr=WZcD#vlm#hoK`dWusGEmn|!(c4O@CK z7GAE~1o&t?GpI$gvuI;<359=ZBG6zDc(qHj#+$F${yiF|+}@Y__Ns1d6_yAM(5JxndVt~FhlZwyeRf1mT$Khr0H^x-A^Uj1d z&^@3IQW;z7Q$D3j(Io5Qo9>K;CA2#j zGSn2h4n#wGo)~>~@!J+7a`2~TzaaH5%ob3mmC0$Pm}ZI`qv9`3U+tM7**rP(^~Xmb z8$>^tmj8On>-nL0r%{c6576eHMZs7-y4O@uRtO@$4Rp)Idg%fGL{-Y`P-$wCJz7cD zhTT+^N~O^JeWZKoosvY; zGpIRn5z@YK*~Jl!j+m*T7_w@w`dtNT^TuXXXRbExp0KFHUs)tQvy*lxgL$cdb_q=S-n(wZboF-ZHDW18-_K+&- zoG3@qAG3Bst2WImMbs3RBgqOX4ER`-%sZjII-!DRlzrj7Q}l)RZP+PlQ|_FCRjkR9 zZUwBu{f;^ylVul9?6(hWPjW^VHxK7`e<}Orvo&Sh*pFe+l&KM)U@ptMKsfrMxJkS! zq6b!U3n81XH6mli0ZH#9tfVz)>wLPDXL*T}`(`}!F!Mg)(I;NHlQV-1vTor+k1wOE z1G|OQo-HBxljH1S*Vs7)M|OP@w};qqw1+PFQfcvdR;PxYCK+QP+id9iB%fkHt}dI3 z?+!1Z(xD!shA!YWDw1L|y`WSp3KEVWv)g>VdLm|UED*H5XZT(+b$Iq>+I+NY_EGh1 z;))kObU3-xeB-}UX^45dz)jaRJe<%ntUsNdVfoh;Gb$Lv6bf@}r_5Tmn$6o+$ zV5R)Kequ}K^@2oMSctyniS|u80ZE8?z^T%uJSJ@tx6lV<-CCpY+RRfEJB9hC>s6BN zaV_*&-rQDgQ$*s}Y^aB-qUMR36!X0oj&&*A>@tEZ9_wQiIR9wVPgEyt%gfw$4a8RW zMy^8qqR_8IfUH_rxYnht16G4Zx+WrBi7BgvewdzErY!UCjok@*6DR4do=D?$AnY~| zl`~;Aq&mY-fb6qXK6hL-Nsa^pW@I>j__0S$riD89vp@aW-{PtoXGWeU>t2}fT$Po_ zT}&~+0hUL_pLh#*0$N_a>9r!L%(sufR9miHqiR-nDd!#x*eCcTv~u!XgSI9jNj3Mh zk5QOQZ;=?Zr+xbPM&UW>*3d#8?n_t1ZV26}NaoFTjg=WWwp^3a!u;QUoo!+B3o;S0 zlsw-Uzb!5oB!_N*`Vvj_GGHF7o>?0G5X!nSpqEegdo@RF^E)O*E`D6knF;*$ z>H{&mHAve(=+VkEUrP)d49=rFAv0gEUaIKg%@@tn9+?PQHf_pniZaDUUbkYSEGqZHj!&Jv&uZxZsP$!Qpz@E@1Z4fB2;>;V}!ZaCQ_B;`?t6SmlpWhy5D~8NXc4LG9&FL4znx{l%C`^v<5sfgIl~G#DnqbxgknCW$U!Q8MbHA? zEk%#A+OsYe3OYgD1<#TrG6W0cU6Ypkf?Lr|x6s)F2;Tv*?L~Q+?*e&K1nfMgdUvT8 zj_Z*2%W4A$ygNw2KsICR&mdiY9Tnd-m`1E1KMO=%ChJsa?=_wBjzTg?_L$GiojYCTuSTahZrZCh0U6M zu<>56RDQwBI-H}-HP2WbkgL%vo4uQXO5CZ)asr^f0-7`J=zqVx)A#QyU3dQY3{~X> z^WUu|DYJ5Mm3N}QL7NbMnPf@M!BHW9#mgC9d+A=msn}!}lJyTANe2)zw7JWUZ~ViL zGeRHvC|Q3^CHdS8iyKc;^;U+Zl48m!vXzR@=Aqy+#_jSTNOs--xE#H$8oG$T$r~vX zHU##F=k}Yzb6-Z!H45)5YyBD(^GT~58C+`Uor-Qn3ONPx%?2&DuH!WTJwgejXWpKg z6WA*s7=fbfFvIKsaK17nwg@9?G9{&)6FTXF*y4grj8Ui33Y4=5K-*eKJh>n2sJ zmw6Y7{uXuPcD_!BUyGv+*{A+tJX`mTUP;^T|C(cl-jKCeI>oG{ z$O(tAm4#?1`=`Q-+oc5)7nX)#pJ;Fp>2^!Aeu-=a81a;jjQ(YKn zvl%^3h(X-gw0G4hJ7Uh<*+2DPEmlJ$TJ>FW%8jkYEi0?hO)*^*=|HaRHd&SgHHmbk z(q!KRQI=%T<6JCCE1v@SdffjjktK#52t}WvZ&F>rT4JI$MF#C_a*Uzfh5Y|Qi0KY^ zEd~z4Y(bI4OuB=cg>^HNd4nE*UBMslI>YnZ! z)wXd)m96tJ(sLGAvZHF4iF3?`?t!=dzSxzqaM6+)nJUOGe@$!hJfV$f2qTj)Q?7#!E;}X64^Jz)41kr-cB-Oy4K5c;w+Pw*R95MvXk2i z+I_>+5!*FZ6_q^?7unFo=#BEch0=4D4AU!@+>Uan2RZ41lAbmhZfJRS6CmKVbM zI9+LexOO^hH60AV<@b8^eQ6$D6<5uGi~*x7kHIVsb0;uIg>Gap%LY4t)ECGg7^9an z$cb4UeMm3@f65Ugahbgw1zLh?@;7WbI@~symxVl}Ibu+}gDhuoySZCs;I9n>ze9bU zt^+g_&<03hII1>iL8xZjR==&$IUrSoT%@C6KpYx%M~9L_J5CM^wZ^CR2eyeXEEyx~ zKoGltzD?FebwsqvyA>S~IN#$Hz4|aY5qFli`W2)E-tL#{b23g3s_VOAJH4CvAM?7D zm*x1JYlp?50Wp6MIlNx(k8&Fjw-pjOwmV5xrY6q?g&yCIj~ zETq&dpHF~>(7Prm8@MaYNf)S?hWSxo|NPQ4wbRSRD42$tBRS-ZqKC!>H)`t@i_Y=d zBbuBW=(mhGH_WiOB`!b@opOG<#Ny`W2s1iJ@(bhULh{0pIIx~#)=^{)Qc0Z>BVA=a z-8-#Ff|SR)6LA?csy&+&TY2R%=}IF|CFm+ucV#sZ-4ORN+L>`Z!ohI_AS3)pyP9ul zI{wf$(UOZ%apR>jOA6)*=7Hc8ua~X~ET=PNhlx?~0CXJI5Tmdvt|GX7?m|dx*{i4# ze-fB7y9Npvd+8;!&r3H4?-AFCdz2{SyW3}lOBc_MVC5^5V%MlU_RwqJsy=B;vCjo7 zZae|85b~R+pHyO=d@5DTW7o|vWwMwrKunIeFRV*xx|{{;WwY!eGOz1u!zu#$K(@14 z-U%ANC}{)@R1 zBdyU7l}Q47z0?tX!@v5fb2zZ`3;%*Fq6&F`+H4z=;bN%Vcc-4P1aQ(|szN!KhFJzu zsp?YB7xeLUxRgp)7Ka+?E@e8uTD_Rk2bGD9^rvK-U!@B5N$tg}o$wj%F~^`jTu;uu z_Vb1%AKC5^umsnyzHweUpSq|lpLGjtQ6eu%P(`hxP`4>ZaxkD%xJTNf%AI*!g4=;T z^me~_qO+rupXSKBb?#dnzPA^|jL5!u@2%G@i>Du7X<0{hx$y?K!D=~lh+;r(yq=2R z!q4WdorHR31#dLc{k~nwGxR~=-^iWW$ZORu|AzkU?Agd%mCZ{PbiqzCDwpZi#&`PO zNt#pimR@~aUST2|1C2$&L7!wvOI(vMJ1S{>p_u-EnR!2O!}yErp1J`{%AoSsd{08d8e!l&ZiUun$!EK_(VZBy@lT{ zE|T<FHTnkBA8e=dp;Obw{Q=^ zNGd1qr^qcTel<1l*1cJW2?p3Peew<h$pbzOJbWt0tCzlNDoa2~ za7;!(k|&;Gfw$Y(u}BqjAj1!EoPzz2rYP zw%gLIz93N>7PeTk=n-yxLzQ|qKweAe`p^@q&y_{NSh<$XL#6LUkYCp+ObA~WR2^90*-J1r zs1`gSZ0Tb3*aehfZJzyu?#Po_M>>A{l4a&vuqgZh8T4@56dwYbM?+TED<~$3B1@?F zZTy80dO6Ijo1y3HdWEF{>El`R`W<;k_O`KWjP32riO+GG=EfL)v*LTtd%UGnm`pB5 zV>SaWgwWBNC(?oZH{DM#*1X*>&##kYPrTq)K5Kw~|FvAH5!mhOcnRV4f_oFQCb}3s zcHv}LTi|fW%`P-?!%5DU@6HlgEXVK4$N!!rjU`1^U(odwvlb_k_*@O{ng9(GRQSl3 z$a6KwbTH_FO6^%8+v9*J5Mp7d(_X`Gb#<`ot2e=U6uxTd+5=9b;JzDfhj(6jdAtQ4 zm;bx{&t%DSM*315mX%6 z2K6MHM$;MP24vTQcWCeIK#sd~BVTO6#*%dxeomIVaY(<|3I`h~CX*tosrdQh=Zj8? zkA<3<5#}>h;$u+YaeL~fs3cVm{|hDNKcy>+f;xqZd7rqfkZkDt#$#?+fH}st0~ejM zV_}%GMe6RUx68&^@X_$v-oKCqZj2Au;ThtTrcq2PMUrvOGf#4Tid&K9{3ax7MR^kk z1EM1t8GB(dlFx9YEhkvK^g&Hhh1LQKW7Nx;WH+}kz8f!UPg{ZLFvT=ddfY;f;I#n*+7I70=Vq`z*jdRj&g<0kEob(G}qUxCcA3mH%^@JF- z9m)g|p8usM#K9G^z1SPx;kDhX$jxi^lGir3I}}m5UdTO_uc1xHFAf6z$?p;e1ho$UJ>DfBn3)WY$IH zN4FHW6n)AHMZdJ3ES=FAdH(A)bmG`!Qgau#r!E|8&>9s@UfGh{G1#etSb1_JosFI+ zKf$pp2^!%i#Dbpf9@$69A~&9*@~rT%nqtx@l8PJ-xf)d3geunhIhg0B2YQbpNt?1r zgj<35$7$ne&s*j)*7o%0tY^5*Qge4Vd}PaA=(c68ENp~Z_;(fOrMTr(9nq~w2sci| z{id^;voZa$ULYInRx~MYK=OYp&%FCvM&}15h)&4Oq1NPxoWOSRW%-&oBT$U#y$8Gx zlNCWXVITQ8x#9fONiMK?+H~S9xIF5I6Kts2bHDT8?`vNfv-wSwm&jDP#kX0Qu3V~J z2S#XAjgD!w?1a#xksY<4gK2xe{rBbZ_C)Rr+wF5&RtFyK7VtbuAa%G?czepRaHFC; zrjB=6+ATZ`1kp%cxR*`~Oye(|aTxgDF+rilwr@Xz@*&?~5i9s~R`aRXfMhN72C+E#4}_4^g@kmCc*J`zrW0DJci-{CeV zeP4d`uWwkK+{9Ji{Sm2sVU|ekRxa^ziaA1&PpSBPh~Q!RWL6BOysS`yOtiWV3=QPt zkmQJclP*RAd#a>7v`Jn7l6>jPbwO#=yl*GIlNFN8yB)K@Uk_{_mjcSf52V<3pAa&| zzjkZZ2`%;lbV8nanp`kccrQB)Kk3F#d%pHy_KDfpevRxwF!THxdbdTIf zAN5>JHOt#%y|H@Dpa;4_3xacJ>eZI5GECw{*Sb?(B!?lGrcahM6;4qN6bA;*9zvI^zRl| z`1N{~@iT=7By)kg28g5{H_@w)N&6#_o2~_nkt;yp6qE6^v2^erw<$~T+XeNu`7F;B zPaR8+OFMulnRk+$p3^={w{!OSS);||ZZm53pv0!hsGPE+8~^IkZn;MdleLz(v|x{F}oAeML^vGM8EKVy}C#==z;0y@Oy}RymdQe^%K|oE}sG- zOGOg5O)q}w^Y3@ql9#z{B#?zwZE@I5SzpX`&tAIgomNPd{K&K!umW}xDq~B13ZgRt zn}om13B(Pk>WEJQKMDAM^Me|B)e!>#v)1Vu{PR=pPMPmQ(hqy=(UZQ)S65tSo7nBq za2wvA2nOWOp0(WA`DK~4+GGp7t_a~9-U!@5slFG%ytu-3dFCkEgOX=<;(`a*QNsTf zZ~yb20&&#`$uMztG4B#!0%oORBZEQTQHOy&5J0owf>C~Im+TN2#!$TEaXHP)Hgk$g zUcVcg5SBbEkQD%~Tr&@cu1aE@h{^p(G+z~454wyenA6}C6wvp*^t>Iw1V2i+vKLKX zO*Er~tJhB7_@=D};|qch7AZ+&hQc~J5Y!VKQ9n@Wg9eoPpoP=Wb*+Yg?M5pP}A=lg5EIq%!c-#7uGT72f?mv0OE$pYSrAiesADp57ywZZS2EIaVHrU_)U zMhq?c!SdLS*mE>eNU_gJ_3s2MvCZn{Vqe@i1&<})8;chD$Vu_ZXrPu>XHC+pcZVkW zS12|^a@_iHMc~rpLtuBrP~b3RoKVA|-;q9=9MdSVEeLVjiw*3`sN;cw68t8nHbJ>)o=a$HKAn& zntf{ZezNq1iNkHSnwv5yW;I39sQ53p?iT9 z^J;jLFqu~j%2?$fH$R`0iQ9ye+lMud7S?ZlExe(PYZlUw&?7&nAWp#QR zp@O)UbVjcaMRwg>vOwPNg)MrZ!1$!+Az17FY_|GbR4(N`R9px)PDBHNXK9)|pt8lpqZLXW2Fg%|QHR1s>2P%^uX>{H zjQ1tjA4Dpz^60t%7ubkL5E~hUANAt=pPV2(rgUEUuWXsPpQ}6r7TWM+(DbaK4>Px8 zx`j!q#0mGM{iKGUq&lFu5A|fIy;+OS_|MBg35SJY~nZo$zGG+`1F5v;|pwosz*H zRg!Esc8F`NyqB#MQ%V69M0`Kk7c=Q|`IIFys;Dk6ckzsNMJnS5CPLHv}W0(at zHQCVcrO*F`!SKj|FU~m{`a&v$~kFEb}x?bb{!rPdQY& z>WmZn307RE@S0!|-btH^-7weEoMl<$34E0WG^#t|#U7WW)j1HodD6acLXDjd|MdGS z1Lj)bGOKn*6zOnd3o~d1fEyHZjUrd6cw8T~$+m~2z-Xh}qSx#W2Cb02!Uh@%fIjx_ ziNtL}tZm#E+AKH8gFs4mC>G}@^ZO!iO)ZVqRmOpKBjniD3U+BqqkH5np(fzmj4>)| z_~p^v^oO{9N(e8K-iz*`zlhF=*&)&0elCHKAa7AN^sT#=tCRYJ{tN&Q8{)`d34jIQI>Btx)zTb%%V7vVK{o zH~xb2<;%Q3d^E#3!$duas}WqWpp27=8nbWePtNcxnETDAw;RdFZj8AcE6lB;m=uaE zr{ar4%c9$4MIzl{QVIvlymcUM9=46QR8~Sg8ZA4&f+K#+Zd8o;lel%j2^BB>@c$L9 zpJYMB?H|h*l8wXdO}cUPX0H`ewo?pBBToW0>>H;|rbF}MNc zSx>VA>4($G@}w5fDE|7jRZ;{i$XrQq}(US6rPiI!Zc#OG%8#|Mt68CgtT?p zAZm|^nxkUzY{%Ii{Mfb`jq|I2{qrWa>P7Att@*=A-^0-J`CC|{M1;yjue+YG_Hj8YCKO6)rhM-bNCtlhcw+9S6UFe zvGQnk8Txa`*gE@=9V@rg-v9jfHK|}wtxE~}O0B#lP_X>BFe#?twFfKC#(_Qk&5`@w zu@$I#(t5b@q>^vtqpqQtbc(D*akH!8$&!UpmT3fotlg1W5~H$#ph80)++G*uO5Ls`&$)#LvzNCtm^PsRvF$&r@(RjJ+sHWH#EC1i9@pm!-&Op?0li zvC{l^D=cRX+RKp9zecqrHW@_aSE#N-vQ3{t&%duq=3Vj{EnC7FA5IzlXLrI0AJk&` zU;f3%Vq1Ptx8W37#cf;AjosO8R+eZJ#bi^U5fVS2Im|n!+&RSvvDG4ewWprHS99HO zuVzUw3Xtb&Y9h|e)N3%w>$+bfZN95l<9*j7i!-Lo=H}U;7~vDY{%`+Vwrsd=YeiW2 zZLj?4=+#Ha z5t7WqiA86mv6K)FaQ3njz( z-O+l@N4nJ6177n+FR;cHN+X4cz3pFmjlTX(%LEjbIOU(nK{s}8yR4>^lN9qAMVhGi z@)(m{z%70tDd|$;mR>n*>cKWJCKsRp9UL~xkEkm77ozrp)kzSw$p*Z1DP%#E=^%^Flzl3~1vCckFGdam(cFiR{>!3`(R;&I{Seg3 zAb?lG1FjY5KBgujpGJBV6pG#}_*evNESOF=;DxZum1mk84i7vHHr%XE_2*Nk z$t=JU-w^+rqzt!j?8dgG%nEGT6a(ShAW(A;20!|lu0(F0o%{z=P7%;IQ#}xr@|(nU>Z0HrNrB2ps>Aluy^2z4qily~ zdU)$}bGw_99_=BfOzYuaG2L7Nd%q4**+chcz}FM89NXmbcvYzD_s2o2LNXW+dkjr$g;5n@(zc@KGspp8VY{>_$9#3mOx!p)&N^EsBX~7yyU2S`qK0~(R`*@ zc~I7>UH#V9|2hmrpw4`8!~D)eE^&OVLt47_kp0X0TxAzrm3;F@V=d^=8h@uF^M;#z z+&F0oiWY}JIhA6PDYA@;&x45E9(6^45z=UT#95MlSsvshIVs}C30d|5m{XTdd6*r* z+jCcU*ewlAR)F2v|_nYN4 z;4LA&^?U+M{lV8%?{5z|qJ}vrv`fAC4YRCsneR59DM_|Xw(C{w6T1U`(3*(s@!MzV zx;SJRb?JiF<|o*aU8j=;(aV;oQHMIa|w zjk)yp|J;1uOUU#=YgH|w8^YUU`zAtoqkv8kAb~!z$QDT|f$Yl^jX|HKhOSfg04dX1 zar*c{58PXXl>F%@eLIDzF)+v8J@Nr% znPMXk6Tl7HLmItiBYmA-tHEnnC)5Cz=E~F?-)fa-yj%9Ytgp5BtR7u;HIr$B9};<8|Ou_$SW2B`FT05OPF?lMvp(TK$%>aHL|Ob)t-qe*g+s^ zL8G~@=7uAe!RCpM>k`NAr!1!9f%k)BK#s%>T5el`0<2I&V=T!Ovy38(srW;&Jv34x z_bGFM=757^`oy~%(S>INVnp3L+<4`pe-MnMkll^_6P7~u8%(cYQAjzxc$`tOMuru^ z52S~t_9!~kSCyRvq+`ep2Gu{1RpYN`nCnc}w5&;Wj6s_maY(j3{Alf)Is zGKxb>WSAHU+s&w+VXl6!Cnx>4x{|c#@o+hUg2%00cf>wryc7DZ9|c?Hm)~ruUqLpx zv1z4-6?TLStLmW)J(gg2R$DBF{txhvfb$n%(!B^lN)Awz5Xfdzbu%!f2~YL&bcvW z?ptBz8pQy+;AJZQc-ZP#q%6U`pCVDaXIZokvmDK8K<3?|ZW?xSvgnhf*U#Y7L+8Sp z*u=cv0c=R8A%8y)azB3jad`yklky};&lU}MVTlovd|`^_e9=k&a^;{0 z=0_p~jj}^gY4RR0WPH@{?Nc@>Zbh0q>=dqw=#0GK)upVMG~ff(Ntxc*Vo%tBR}cJ7 zb-V)#Q{_>s`iR&Z`h&uzD&@t<9@tB=wslSc+!JQYp$k^FjvL?(?fl@Ev;}atzaMm% z%pWfH@5W&`DE=5S$E~E86%czS0#6F<7>eJc# zWe3XVG5iHnETEX(QJ6%Eo-*RtL6*F2AbT@TM4h4ald>dG&lq}} z#2jp_a)vUaWJxdGEQfqK`ZLL5V39g2J}btep;nN;+bJI%Q{w;;g5dW5_^?Ot0NA{ z2k0(EHqfqMbx?)0byV~-?Lv?3gyVYfjJcTA`4_$gBOW0h^GV_hgArhT8-kH^idjjK z6;ynM-?7jY{C4fhx0Bwv`^I@`ogfDmG)=-H{sNwUGz4UW1PXjbc(Ml1#YBV6oHwT1=qun#)NS> zoL1&d81U9rNnyvCE|}aTH>gV}*93I|Np@q!c6xbq{!BDF_Q3B}eDQ}!%hZw+l$$}y z+&JoX#L6htPz*>LeL}_W1{QPlJwbe9DzH=I!nI3TMW12H6r~I3v^{aS~)NUo9z;U{Kctj(kYy#+V+mU&DT*YJof- z2o;>L!Wl6n9znzEr(-Y}+KSy*9cJ6`rQc5Yrz^I;iQATYuy`BI^0kw?HNbRH15&x4 zjDzs(X>lE|OW7Wg8e9PaS56r0p)dFu#~g6+jGml(n-ea^Pz|-O+3prT84Gb^4}@h0 zu~pqew}2=bR&cb*dSms78w1GsbZ9y>Mq!O`K-L*_Ib4UNkwrmW>LSUNsbEPOXw>95 zOm0tYk{jtFPnS(JBc5rT3!2BhvBT$Irf<72-7*JdDgUF0?BKS%abpkxnZ}Ui%>jzp zN0C}89w=$)Thdmw9`*=Lr9Ic=m@hl%(J4#}%cE0Cn=)lwn|RQp8KU<_nfak4A6x-7 zL9sjXX>8*!^h)N{1#E`={jI#s-Zi0!$0E_7N3vu;zYDAv(rjRW7Gib%UZ~DlBVM6e z7ln_E%%E{FP*ziu9Vd_B;=s||-;W=svf${)$H$x?nQrXMd}4)+B8mYy?p!Lqn#tf- zQ3G@-$>4V>3PQUSPhM$-Y$(i#YM|GKH3=K$oDHi3s=+g1+vt{%>cAXLy`aEm!_I-1 z$=+mS!_VIDNbj&4m|^@DD#Unv=CgPI+}O0R?EYPlVk(O{H9@y0Y+;mMjgrfYM_sDL z?lsRy$DMO~cAymh!%u(kyela5!Hz$afC_5F`@*1D5Z6NJ&*pgTnxt2s@NELg)*=zy zDv!aeT4W;CCHSvXE>rH5>#ju_WmpF0I`3qp?g{6?wzmO0D2A~fFAb>P6!=@F8sp5! z^JLu%2)DT3{OnC7yES7j1eZsnybQM2bd zh^iKeTEUl@FT(9(Oq4315oUuP$cQ^q=i_PPIkK6_6)8~W-?r<+l}n%ZH8Lg>9Z)nW z?!?0O-ZB4dkq#0A;_~T!*>Yw>Z2r^f*Pb?L4&CE0(hl|SXd`Y`Bq{W;&-0FpJWVcL z7_6THT7>GeVZCHWaMQSh|9NEg@_$d9o&0YjTWD|}1-o&yH*P zS6bGQU2eRPF<8wPhbRVk+UkKoa!QYCkNB2it^ebrbmeDs2EPy#k1M1bLR+UBv`E#K z!S9Rd2^pXllXDV-b}%|CI8k$K%HpsSsvOB*^5LdI`>AM+te4*Z7HV^3`0wyJGxNSQ zb#_k(l(2(d2NsGYP`cEw8MJx)CV7?Qy5DWkX*O5BkAR8aZZO%gupRCZM-jyhOuExw zG}=-exNU%zh0XvAife?R&;q0dm*ri`d`%xeL3Hh9y}A_0SH2LY#nnXIBluK?e+hp> z&;}oa_6r3#DnfiJchX`}9^EbZn1|1$@qr2fNHOq9bLhXCSsv?LL*+!|I0}+s+i^qA z&EyS7-?EsYU!TbQKH2ZaW~jr;41Gp1O%yq57R)W=HP8*S`h-hCs^NIpfD-8qF8a+U zm-sj6`H;I-6R}ga*i%;%x{+xSZ{j2K0fuyeeg&6LY15BO+93B|*DEXyfNLk>%A!+a z+6nF;7D_Rhg^18{H~qd5T2iKrUR9NAnO7C^&ZODM+v0GNX+d^uR#Ox=EehgB~3qyp7qmTjH`M z=cSxDyhix8Iip~NpJ#7#cH{Dp|E=>emg(sSJBzbP`E!;`3q*25Hi!08%wCFAQ}OF& zG^(%q3;^>%2~|RE1D3g_ad?E%3)gz1_oHx;b_EWhBJz6CLOj@d(p!796O4} zt(m=CWI@sI%E$knB)PGZQ)Gph^%S!fsXrh#H~5ch-qNdA1a0%r3ayhFwAnmVgXy7D zzMC-_^n~F0*4V4Ew7@@pm_292gUt4lt*>*3f^~aWov}ko%}}7iWv=*aEA!k z2Up=@NWYvQGG<0m?$2$bcw8J>H}+sy@}-JH%c8O513cI6$YdVQCHX-AUJGlfsbywS{PkyH-Xkt#Pv$SEs?d`d9~DYBo6PggGUyrNnbR5v9xrZBitQYk5>(}L!`yiKZWRA&p! z8o98lO$+KLHT+vsA4Cp%4A8kV>p;;3rX0+wZIeOP?lSKKfU|0WE{|SF^63+@KjSlPAyObEz%LR2iD2{&+=>YZ|!_p8~*>%Vr zd)%<}{ht@SW-H$DT;*G`D2-IA4wEIpx+eMdaMOg-${XO{j=>P#m(h?{Str;8MPjI8 z*CJjU=D_H49~Zk5TZZ7mkx${?h8;0)W*of}YOyqvv;LY&@?V&Ig?cNyR7o-A6xm9} z7toCmU%eKhS7Wqo(4z~gK``=yKXnIqJ&|~VKXsLm(y#Yk%V4HL7M&HYSMN|lF6RTpNhOs7=CNpjku;A8y{ z?iYfVJ4$Y(0Aqd@y(er3S z^vIpLoT-i&pda^E7Fr!q$jAQlYUmF>|AFetRQz2UC#+TH(I1_%hk(;jj6K4dV?Un_ zN=_J~{(I$(Nw$h)FGw(sMOdvi?6#-?5^{BYihkK%8In7pC;K;;oa{r;vq#+7PyBll zdEXWjZaX)zjB>g7EU!E!UAYrBjddugP(as2bknY%dY%{z`$l;*!t~a^8Yeq3W&+nzTLS`fi{J>gAu{k6^GPuwH61~81-@{**%t=w(=7XQw$I=9-!iz<=KLM0xJql6-0CuYuR-98CYm2@c|DC+_sX0F!&!JgkkMQUt6K~8F9CFqsw zgO*42DUPcSX_iBIUrykKsP{8NZkRe=&6_8>Cf}_|5!m+3ebDV3v>{()hYmT9G^asx z-`&-XKNVfHr5bhHDTswu)JP`^ntf6P5E7|W1R;(5FNOV_bOSQ^=lxSFg$U!UTrxxx<$QSI$N? z%g1oO%njAoe^lHQVlhPxb5>T7+!rRG4-^SQ(&8#8W*bGyP;fQR@7(moqOCqBd><;h zlx44^X>=F;n&oM*Xx6Jg8uZxookA&+S?jj&+r^I_N9n1iahs$?qI<~O;bq#x#*Br# zGIXGI*bS(x(l=wA=}S8oWSp_Y+Tipb8vZD>z-IQT)%(fP=PYLuicyDvBa>oQQzVUw z$H~ubgvoFVYk-kf)n-Oju= z^Dy)GwMn=U+^hBKWZreZa!{@5gMA^0`R)L(8RwyT=?+o>A-xQkrA);kNGyg=Wl5^} zKz=ti#w#A?vFVQQ$MV4sIbjF za9dA?CC{K!*bJmReKY!Eu(lo5mXUeKw9oUoAB=0jMR2vdf%(m~#TZs&!qtRWHFn_% zy9pU)ukfXkU0?SWSmvC`Yroe>7CvWlPOjCQlR+`7D3Su|xsVZ;C`c64lUwp;;BP{i zjv9!(R>mCzZzIW-6gkTz(`hR!8c@5Y5#D8$w5I?IUc6*`u?av0sTPBjr|K0v)vc!$|kAco@$UZ%DcuV*@f2hhQU*VjyuX3nLtub5c8febjuA`7%`r ze-XMbtW(&dJ{j>97cpLfWNI0#YHK3aN2N;gqkzq;2#Nq*%kjnuCqoQ}>ClO|ud;6*c0EbUpb}bc%*eqViabp}XZRzx!&=2NQ zA0C5Fw(tkO(f<7p|8PTim}#KfqIZWHv~AHjBT3zLE#PcnWdHfZx9?qf$pWDTi^30( zL66}kJGYexCL64PvVvlgD6)i#Upo7|^sfBM)T>{wob|z>Ioo;r1kW@I_P;gzpR{M( z?eEQbPqG^t)msnswnc<2wR)&-QAhWSbvGxL!^*ND%p}uR8-}M?1<~Z)Nw>+mlq)@8z=f#QxHU`Odj-tU(%-_8ej} ztWh@Tfs{JtREm2%19|9in_LT75cSaGQ_@T?fKqsD+bqx1Y@b{v#&$^7Gy+(3RVec8U}oh!~Pf7>Dk9wn3|Es{W?muG=2*>r|E1w|6GNTm)39b}fj z;>i9m4{*>d$G}XV5_KLjWs6Pi2fSO==JNJ4GgEzm)XSNCOwL8lvk_tM6P%DUCa3XF z)x_etZ1>1MN)~Z*e%&~vmuKbEtfrVWilkEU`}sM1)EvT+!K`pywP)@(&(DHvEU!M` zL4F`e7Nn0KfjzNzlpa6F1xJrtx&EkvT72z!C)jzEz;r6Q-4jW!Qv_wvw_`e?6x6^J ziB3kRkH@S7v*j7JLFY2yIIXyTsLh|i$#A?B_x($O(=3MLC!b9(CEMIMAamTx6x2~n z4MlcS@q1-htb>v+UkJVs+?V4{V5j=ykWa$z5FKtK=W4F{m?t4jB@ixWiX@$p6BgH^{$a*RsYuZuSt`|s6 z&9p12G56?)VF~_v^@GS;k*FecE;eNxDjQ-X?xNoj;LjKpB{Rn$H8Qx^)A}9Vth_fY zAW2;H-5-(K7X~ElRv0|EOq#38hSPIZ7MAsIR$uVWuNVoEU|0@oH!HU(f zAd8vI!{qV1iafelx;wN}m`7*$8|e(MZW^j!X(N3g>^3nl=Idavpr#-)h2#=kJgrmq zPPzegk)_euQ3ld32F}~dyb{>xx}@AE=#FfbFA`l<)8~w_`s6)tIl216AZp1=LwBg|QA|HYZc*_FkQ4#X_GC#GMd2)$}q*HrbcrWsr84 z9FgvMA=*TJyCd{E6j=9$*cXb^>40}l#6D6dXoQ`zc_FQ! zF`LW-!-usAoiy^YI+n08N`SJqRd$^D|7qfT&-)g-IjN?M&JRiuosidrE@iM#|1Ogx zxJ){Q{coR_Zd0s^XrOmUQ{oE44ccR&EumZZpGR&FO$$2c$lPSddwu+f!z2H=3zr;m z^0uiN)4oTZ_cWY>$pzpU?+b-WB=L4-vk=f2c2FC4|;ScOkqy@EzVU^!3uZLPt zbeI+fRnQLu!FxfM>4xVcF2uTk2PzZ$VVR=E$4rmtXrT%xAPw(YJmc!{KzY_xPC$C; zdq28T@VxU0q07-R@uHsrM2^y^-n&=cXUh9FvyEphi(;TysNTQM39-!|ciXZPmW=R3)j4S~Y(-uN?j9Yk zyHQ4F6cF&l*x3b%CtA7w0w+j}Ij>fH7-*S)7W~dXiDbF)`~yi;L()LEP$0%Y3L$I* zGm2iF9koQ6FfAvrfQNNS71C~4;p)|GGTd&$GPOc!0S!r_wPd-kxdA?kjewL{Y0mAy z@-(BYJlaJxF?J#58B>WXHYKjfO)7+!+xE|I<_{3g1mA_ZE2?X^F+I&zGSrN3r9_BDd zjljqNuDtr)@UYa|ITb)Uk`=rnvH$FV+3h$45xS@P)Edh*z#FjKT+k z{>TTxdbRnv(V3zgK*+O)*9lLsBP224)eC->38?kID&Hq-Uzn)ePOIsvkYe&FvWbfS z%%@X$%qx#Z5vyLm`>JGNvJj<;n$`J1m|xQdwM|*B*wqEK6~qCNcH9+Kn6&PuJ-=x?K#GKv&a@mR@*p_0Ae zoS0?5JB5|1Rug555ePT31-d=L8(vACDFSm61wMuv0H$3f*eQpNr|OA!NCCY?&<>Rh zxUuOJC8{T0`iM~ALOihBl3~(N$NcbD;{~>uc|mGsSrB$efu$x;QUU=J9U6-+U}xB- zJV=%(zfhKvF4zXhkE#nm1Eaf6r}3NR1@yj2r-1xH0xHIUs40*ZFiQ^B_)!R}9*ZHqr^y$7I0k4pb$r z2r_7se6P`Il3<KF5bQ{%p=Z)T|+cX7}zzlKA>`^QPCs+(24kV zfJo@^Wc9?6!*Cv^Av?lP%Leq5Eui`6k6nAnW;b?hK*neYKt7?E9TeG4#c$MH;-lPr z-_&er1<+&$|SGH7f9^ix?CO$2QVmgm!E1y>VW;ddkizb8Ex;_!R-0q&RcUT|Ff=a!2rG z;!NXXAI_ebjfTJAMBM#s{h@!h)tYzP9U>O>_>{@{uh&glH32;9@Oe}Y^UwpYT%DH0 z%!ex_v+zo<<(WS{I>FoX;P!lI;LL%3a%@I0^iOu?#4FrP%y)FW^-&g&MezsS2C~JC zJ(i(CwO4L+9{MI>U;L071hW5YyF>Lx~%LQukGg11t!7z#Lve^lpqG0L&1dw~S(n zDPXRO@ApEb&86fJt;0RVs^EPSOQ`p8ePdATAvd;HSQ?N%{(W39p@KY8szPcDvVCj} z>Y?8^mrr&>s(6R#Q6zWnQ#zW*!ajh`0u1{nICh5$WlIX3)-OZUtV7 z+)6$c+?S+{)2kb3+(29<$(A;YwkvD>Fa^_mZ(nqM^h4#YXoGfBRFbNKpF9JfGWSsF zX?|vBaO-sQ)%&Wg6SAf0;M068Xq4qeZ=RXMA9xv`KT2S)o7X!1xERkzGQl|uOT$_s z$6#b?#R*H)R{4rO&-;SAUfQKRFRlZsitGLz;=8iM8CmoZHEdz4>!nA0PVekBAHC`J!Xv?gIuN2r|8jMQ5j=#@01Z#(E&Q ztEzV{>4-C-GXkVM(Uco>V!D$;ZMl27EUw(RaGQncw^^RA)Lnz*h)UHZ{suoCG6WP% zP7BH&uj}=z1wwf|fSFFOUL-AzcK-DB#DKvCN3b+pKhC~3XP?h^WfreSd_(+elH$f* z4eUq`@oKUu2H1!)G1O}e=~C{4LJ*^{TLI)|#t^LDHW&M3csJ1oki=<|VTqcq2E5_> zKBJ_c=E$IrMj1CJ12QJz%|%il3q*cUx8W37<;D;JDTyHv*+enf6mV9>BWYgKxFo1# z1Eu<_KDgBg>6poq6LDqHxWj!gAX!iok)*;3pf98Ed!>Q9tE&rG8@7$N)QSA=!+v`m zLkapL!*=Jii=0sMlIAD#mc492$+wGt`afj78>3{G6-r7d24<3cDt`UAIzf&knYYe2 zRj`YfBdHVAsrJ&>{SDeq;TArqP6Xqcxk=a~#wYqBfx$?a?RQgE?Kwc-RJG8lv3m8M z=w-@<)3%O;yzc2E=r~-=F6`5Lcl<6VxKMYa7X9pb&*L4SI|8eyFXcP=cc%*94ip^Gnb`aZhvvM^@Vn!4v*`L$E{IqWb4X|A9@G=txz$n?FSyvyWBs;JzV+4N4h z(#HpqRdO}5F&3nPT{Lk#fwyNP*X(0F{`-k8TylXEa;doUq3=KMk^BxNavyH+Lv7Gx z;J;YTl=_wf&p{1+oZOCSqtLVX%2@U zACM=AcJQ`A@z4U^?$EPS62cqlhp-dSF1|9gh+LY!*ypOCFnT*LRj^I5CvLrOz6PWh zVHW`JbqYa@kyLwP`xYoU+BfMzurV6xJM`*@%5x%v77OwBO==g{OP4c{LIuxNDk@+D zs?7JyOkfR%J%`QocE8&MNZ06np?Lqerb*r=8}Q1Hu8M1tUGg+&FZvaWcJfPN_Jq_1 zHv3?fuBfROg7p&JAelKwGVF=owS|HaU zaoTkuaFSH2a9<;T9A`362cgDs62r&20By{eKh*kHL|W#zoS@tcQpRn5b7OxBh#80M zKh#jnZi;+D#owSe_#t5x`be#jgW_H~?;9UvfM~{jDKb8bC36XYxM5}3r&aOmLTzPKxa?`WF*P2Gz)Cq?9dXsiB+Z0%xb5N|@6p+~wt4?{ zA;PB3Xq*ld>eoE)+;&^RW%=l$GJ(3KFlckgeJK_#U6y8ho|AThd|cHx5@+9>)hXT~ z#Y~xEDB{~I*rF(p9*8gcw}WrqoaKx!{8NU?u8|*h>Uc`0PPoPHw+vTm(3WyNoVH9o zx4$2Bn9O(MZA#d?9x|D%q?i>HNuuI$@o{rnkqG>YbY+^PDh}5t>!k~mjMI0R!98?r>VG$k01r!-n_`h!w zlth9#kZ_}OJHNuoS-xPtC-43~?^~xXn0m-|+hJnH8?Ju)lK5Y35Wd^!bW?nd$qR`M zSQbz~-{j^6osqN!Z6ryYJ9G)JkJl7_fMnBI$`W3(;#0YGY{=ms{j*X z9*{Kx^j%fUs7m?xU2KAb$z&;LS5Pi<|d}L~VKw>!$$dv&7ET^F-&T;R$Aai(e2o8M8f{`51 zq!2k33O=hPoivW&*x+31x5l#U*cg%c1d?MpumrS6|8@Fpo&j5a+ulAxRN zM%da+F`Fn7M@1A1Q=qF!J5vRd3xZ=9B=gn=w<{2V%3N~|Lh!>0lye7pSTU(hzZ16Io z(6AOjOxF3V>-!I0^pZPlweJ<45nlA#JiRTjjz%K*opg=IfI}Nc>r~0v;F+RGhG7r` z4pHs{4m)NJIP4A^aJc9-;4m1tM8k~~EV(8u4E@Ti}KD)759zceGstdzA~XCznaU2}DlCWo)oyarM}@R>xuQFc`coY=((Ynt zrH`mSRwPY(;N8kkp*vtMj*b&(xi`$7<)+Q*qRZ(z@7v;~xdNQmrdA*g!@b#7j?7w) z<`LD&@RL?WG%GYuy1nUlPA_^{UL!az9Wt0BgUs)_BFrvBUKz(lW+-*^b!wMY^HRVQ zDG(x?l-Y6st8T;Cn#)+G?8fvLvjWKE?|tj?PfY>BCaGh`_AKyz53^_06q8Jmt-!4r zdY;)L+3tF8<`G3IFMiqsg7!?3_kl%cIgR1jbiJ@snI^*Pm&MC@r^6>mi?WwDV?qb( zhqkXxUh|@pz4Vg%vI;Lu1J{%Fyngx+w^LT=wo`&>on7LG+&0-qBv$mu{hG90epGoz zQW|=ew*lfS_g_2UW7D=G{cf<|`XTMYdN)};z)F%({Y`&e_@eisB|=JHt=R6E>-j)d zO{dOUCfFd)n7u}Dhu*^hYSbrJQ(?aB(D|nDYz`9Ht$pQcV7;uyW4mLNUx_T++h%fO z>~OJ>e`5d@JGlHNtG)?~XU}?8&BAGFE4nJs1d4wa- zo5vQl^0VmWyn0fstfnELTLYu_Hh8KPcyFS7Nv5(>nH+{KHBh3|4lLmP&iVz8LK<7G zFkC`{Xnc_ll~d{5N@<6rid6eP3`Eli1AECJ+gbwD45l*IyRro8c;PoByPp4>m9Uy5 z;~oCPcMOp^`8w~flaqFwWbQJuO)gOk)Wn^mBHDrBw@X+pI4>xtZgTf18^f!=llaT) zi?&g>UqAF8r^7M&a&+d}xf@+?Koss?NM%5}ODn${$f(LVpTelw{;(ZxE2V>halB&p zQXs{|Nd9_WiQ=-jbb$sq;m<3t1ut2G>87q>>s;`yQMB?OdtDV)EGTuV_pakf!yy+Ga#(_ z)w6#h33iOIA|r%lP)r&{)Kml#@Zwm$8?I$D)&hI9s^6s(T&Ec3J*?_}99YQ7RMx{7 z997J;BL55}zStKGvL;c!&%29W<78bDUXv{v7fejnvgS)xFqzc;%6jhSYgmQ8*c;i* z^d33D3b}+Tg>67#-i8Uv3#2mO8Yhh#9onTkrGKJ01aI3dE1QdoOs%kT2xk}s#)T}9 zDZ`es>LGVa`1KdPD-P#VcgV^`cii@ep7M#~A;bMv$%$!q0&Y>`44!zPRu0=5DHE8#*_26z@K4Lay( z#6LHYB0JVW^+r1A0L8#CiV7-1L#vgsqHYN47DK;J9LCI08!5@uwBpqZV9M|nueIrS` zt^BlE?eeNw_XE?M_s!0v>z#43t6sL(if+iYPV2+^U3Aj`@VQ)pl?A>PcZc1_|D3y7 z?T-CspnfX~-ZRetn_qk~Kc5u7G_?;WjnH31F_jeAPemk6yFRxmJjb<&gMGI*eFo;K z6}_(=5yUx`I%NY_<~mZ#i{kZgJC(VjPeRkUt32btwNB?PSpowMz#}hM)F%JnzATT| z2~!M`Bu(MTp0~f8_NMmrba5weqD6%@zIxApS@;pU0_GZPJX%4>_2Xy1vf@|a ziI~EJ=OF#LmpjToi!`N#exZ6F&Co<+LnGGJUG%!;u4-^^aIf{<6Qugotw+|R*fu+J zW_2& z+ex5y{2vq13TyHewG;p*kJs*iVKV8B-`*6NGHSjgX-qTxo2i^uD5dBRJImQ3 z((ON?PQPxZAmq63Doy`VQ6u@;=+P$w-iuDYGu6Q!h*l+ zTZyl(ebEz{$l=i;=>ZEo-RCZE2}Vy6Jc}h&vztN^W)w}o;oigjgxkSQ0`8f1&Rt%m zQ0s(dJM`W&Dc%X5hiy#0HMXtS0l|U|9Q!S-WC``h1w8siAC^@5U0rciqC8Si?1BG> zN?g~Daa#*x*aFKiF@}Z5`3NgaOkO8z$~J9CVxwQ|xYxv7`x3tn29orWJK)vdbw*RO zmozbGsqXU0lVER28_>C^A#oH-t_9_9w<@Wr zWy%I5_Uri*GflDm&iTPDPC)HEV!&-$4gECa3_o?h`hb;Nzz1GXQbHFn5Bt(`e|AF z$TCSW0Un9TKWxYJDS^`ssOVT2c!?z2acZaBXrFgcOg2R_sfYtUyJtS~h0*oFQPY|r zt`*B^gsiZtk%JS$v6`t}ju$!uuW>F1L+wGmw2nT%uuH0n<;73e1`o*MLvhNz-q5g- zvufe~KVv^UJz=tOR(i?GBAxW=yYGM3aVUrLh3bEjmF#4Q9UB`vjj)wKG4T|Mr6O8{ z`$9{Ey&-jU+u{q0?l01LIQoJE)+Aly-ZG0pH12m|x4@sa9q}>kPklDNZpWK%rVDky zGC%l)@)6Xroe-~cQ6+ibrI!hEMRC0RU}!&Iy2dx!6vNbc+*%bX{WmeG08Y&&O_qEWiXx0CtO7;cix#nEcPXtsi~gU~C*)6!2|w%8pIb zJ4WV1E5%%*$OS55Kn7eT{5#~9x7xEY;6wi-ifXzz1TE6Mpa+mHxe|~Zq^kAJqfx~m zgH!WD6{?gJIxpl_V6hUr;5SVzq$~K_6!R<_qm{~72(o8YjaAN=MRWGI^@Ij(&Y zRg#OIDGpfTiiI~`Su+P_CbrA_UF$fYrI0lay53Xhhuq>2B->Hpz-JUsF-)eiQGq?s ztDV~A4bpy>%79&qVI)xpeP4`ZPq~7AXX`*{R4|N9R9FQltfau?lvKA~Q~nG#0arWr zjm*%191-_I-w-BXRLGx=l>it>(fP-SrKL+pQyx~?qG5}$L(#AYZG5=6?yL3VN#WUX zr@NUH9!`$Gxv-nJnOo(%8QLqZ$j>j_D#`SQibm*Wt>HDeMhdRT&yp%%wc_%$Wr9e- zhvXE%$2!tv)a_&kk|EhM>XKHucs%B($Nde4?zMlvQABdsMb+#$TT^FbgOpQDDMgB@ z2;|R?^L;?Jx}W#2mzDbK$yCs=Rbis{#N2(ag1#zK>NFN23fotA0G~dleLLwm-cYnw z)hszdK$%6(L&CUTu(LY=OKondGXWivM0tgx#KStk8XI1nNLG7FtC#+P>3yBn5onHq61IGg)f!7^SGD%FmGy^(^TlpO_q|WFGQ7ewfD(Msv zUT+GI6x&(ynNhAnI(BNPg~4T7sYFE@Fcl?iGIdn=d*?3bfYn zle{bVm2?rOOE>aY1AcA>q;lHiagJ5M7uXcOmXyrB>YO9i&Wv-!BBpJ=_55s?BR)2= zKc4QCB^y7IxOlpPZH`z84(id5-&t!)JNayxiXDgC%#cfVDyw0Z0(9KCpl93;RJS=^ zP2nlwc&E+bqTg)xps@`|Ojk7znviS0Qv^eXN%7S8$$C3BOn@kV zSkdctirI$9jkv+P0rPdM=BuiN1wy^H7<51zMVIGQ*1D?0iu@DhgFdBF>lhWLAMZ%7 zSwmr@MQm`+JQPMKQkk6agKudm13VtPJU&4N9PBsyyWI#Lu@tk8B5SCKW{D0Q70Pb1 z8d~)Z_)cV&=IIVv0*I$~PuvkZK=ka?ta35{;_s?6eIR#=xIf57_H3$nwioD~<#0d)6X-lB-)Yh<9XmE3uzUOR!M)_C1 zeu}J{GJ;64lV-#YUAToH2RxAHZd zjM>!_k~6dEI8vW;ROVX%pcSS`3Ahp7rs zVFU!B7+i-EBp9;boGMy7UpIIGukC{({Lei$w)uQZWy&M*l3=662pd@x0|iH^FfqX$ zL4ka+s^w550=CI=!4?L7wb} zrZUsbR=zH4A@bB@(I44J2g!D)`;MnW%Z`PJnGUUc{CZ#<@1*mn6z_Ujk-Ud{3d+*( zXu={LPxsPNSUi1VVvZ^W-}ehKn4{k<+m%X+Uz&`6gOU6=NHH)5N{fCC=1{NCp9t>YJ&#(l!f*kL_tVHmUWcTWzT>EWm+`7AOHpJhQ=b2Nd}y2c{_S6U&441`sJTBN zhuNXXj;mbTjZk!kVm_frJ=TpR3)QK-J~Y(+iO73Qr>w z1?Zqe-bvpYPKV27s7>0=KfPeR6Am@i#om+}W$pA@s7q3v65@b_8cw&gW?Glx93V1R z(CeBj&^iw~AO|11Pingd!OqifrEL=SWFAX?c54$KXfpGUF zc?sFh8E|Nve^ET(kRsaVbItAE>;Z?((`)GMz`5Qntpv2LBzcmf9yrG14vF>M?pWxS z0%N`5m0LZ3C8^+F65oYl-ble|X{FCvrjJ~Itwi>im*iX}8T!}ZU-_7af8VbrSY$ne zVsj_H{j*EUXYoSjLDA5d-2VlG>S%uRdKM|M5$Fmj2KcCVQ4!0jHu*o$TUseUDJl_X zaqES9fexXH*UsE=YYgmtt$Q|f4Fz_}c8T}5JO`BJ5HIy zFifFaiARet!E=yUeZwu-7>h7r!8+Ey<@=(rRMXs&-8S@?$re@fBmMisk~nwf-F5Eg z;?U(3(RGsSnJFm{X45&6>#rRWSMn>ot3d$Zxrfk+eA_)6+6S8yoytQ&HpYWRTR);+ zvG!&gug6N^P;c(6cw1&rI4LuJ`v;O}$97AxkvPbt7)T$bKw1%ab~JRZ0DF0Q8HQf{LQrX$!-OTnK~zpUuR&0RNvuqaK=v03}qA4sq{OmHqS~CJ@!JtG)OOiEvvTi zc!OauVvCMpkJ|K*6*?w+zf$ufp+Sm7Ctv8UHnj0*P`S|ly z!4MyUzBG~TUyI_=mjjWSB`J)oF-6uC=9bV>W=2WGcGJ1}(a z6mpuOtRR-BRv^`yD&0ln*9u&-katOUDdb?m!TE)%KId9}D@TKCyP}nj6m+`{I;`;T zrVohDiqTrd5|%`HHJxu5SPO)28X70k7P_4f_reG#)vlmc8cPw9I7pt;9Pmios8D@G zYJFoxmqIhV+I7`1xzOv5M)TFcJ>-J#QRO*dJWo}|EfH=7CXAiR6nYI&MNU;^z+aH2 zyfAmuRE;0j^8&ScNIl)n&-X~2y7U_7S}=ApADsrly=%c5+HSLY=213M1nx&Gk462^ zfV}hnQT&iJ+A;FFjpR=|#ayAtMJnQAXdlGybGg-Yl;99)d}wW1~9ZpdMB*STueN@+c~Q6;i$?<~)nP_?35e*Lvm z0q2CPrqs*!aH?rMz8b!yh zj8_hLBp+~y4ano-&$F@vKp55~tP?GP31s_~$9RW$)w1H}kn6G$W+S@aV?Q~npkoCy zs=4)Nrx^pxoFDd{CtK_oW_yibrlFV&ilk8yfI&DA1Gl6xjE+4}LvNei zsl>B-sRa_K5O^lh6K{~}8yHmuvQw^EHr0gVM9QeUJ)VoKkTUsHUCsG74M>^X^Q)Vr z){c?VW`vZ}6a%TIlT^ey7qtQlDz#3#dg6WRDs-n*v$`(UnpqfjEf|t?-4R zPXzc(3;gal0;91kNzwrIoQ4**Zs|VnbZ%ZqqvDjX4~U<7q%qT=cpN7;Al&pys%u&t zZ|EkVdIFV+>~mTte0yN>hT^>&P?-m7;*i~RZlZi(&IL}fkc~w)PF%5-KNdrc6|N?I zxMLvFw9%1G+Q5!|b~9~}+s0R#ohxcA zj3P&5E?Rp*6fn!6`Zmpn|W^ zB}Si@&nMQpmEBkZ_2I+%jr(paiH#B$=TqElb55VP2oju zb-HV@qPujL9N!hVF`?ArabO{59EDo!0cA67WdyY9kbk~zT2sYF;@Gh=GE;38E9#~< zQc5v_PG<51YGoIFm3v9t zN2+FNnIdkMGL73s-xY6$2u_PoyRZ$M?^I3--4UEKyV+}^Y!AzJeAriif_9vh1xvl# zD!60HlgvhF*s&IZez#%e+C>zzn<9A_*f~Yl`o;je>3Ze;z-k!#kQdU=)o^-W+v{|T z93h>uZOVFSg0}E`I>YA+<<$_8Il!;-L^OB#PN9!+D%*$6Y52K`_3esLa%E*7mXn)=@u;%=ebnILKN^lkp<_kY+CVX$W( z1pG}#T1QGm*|CukX_SKKrkL9lxk*KA;VctG@!}xskRPlnrnLds?*Y-4v{{{e?Lx!3 z;*fT^>KKXkQf+t4a@H@c&BCxSP$COcA#a%~%2cXSMe)!G+3M23U77+h z-y!mTTv|8s>ayDoH^bbS8qPibDOZg0sTD9dWGPbqCwTS|pz@a=1nsA)^4%jy4Jvi- zktKLm^E7_V5@fK-fhyoDH=1u;g66Pl{sEFrukl^sT^#sC!A%frh`JMlX`1x<&iFp6#zx(5FBg8c0KqzX2OcXW$ zR&M_cgHZbBu3!F!Y=3E-?0rV&O&-NS#a0$h)~laYPL=ya%gd=Y!5XF~Tg;Wzw~ zgRX`*g`W-w(td2M-Y2Z#<-6AN>-l%wZUokFj*%+?u^d~0ZQO$9soJsX(LN#X-*^VN z_-%Xp2wC;gzy;Dx!>r-W6tjsUaa6>zX**myp;Pvp{{x7dEAlst#Xx#|tYwRLsq~!4 zPu22|6qLD*KE9EbHLui15ajl2Hr9OQE9*@ATkST$l4c|XHc?C*MPjgN4d+93c&i!} z{mxxHfv_=LGvlHtk=q`2Iy{b7Ot0aji`zNaR+i-5 zOQRio*=?C%_4L)#lljTO*ODWM72zF7Sci0FYQ?FUtu9UBdxF%;ULZTjchPtpB1zM7 zlxjt6K%ysfr6{wqfAd^)vO*aZQ!amlZ$R1WXSWxCgJa<>EbOS4J zOs;(+snoPy+zeHmzG+!kNr|LinA<-huH|8wk+sIzTK*AO{OMc1q*d-@_vza$0s1BN z&ntorGDGOND4pzK=i=C}B}PsdNtkMi(Nbg|710Fan&alIoudoh?N`PCvyQ4;nk2zE zUb#ZG66Pyr(&Y;>ln=by6}O2RRM8poC~VyvRdo28Icw&q6%V{y`Ou6IqB`KxC#iA+ zYH)ffmh9>0-66+Allw|(7Of^5JfW<2>2iS@$b!1)HREeju=txW1}!z7uf*!v4vgtg z=V!?~>^R78W^{AAe4B5yR0ZWtb7DE*wRKC60UK026z%3h;R>7Tw()$rL(j6<{h>!D zHUEbt=dI472LQOVi1t$Ur#v5ueDvTeq`73s)_-@um@eT&Tk@wke{y`$y*(7$X zYSczzCXQlaD6*D{$c028`mlpO+LF zVvrZ!X;*ql6FYff$3;uMM)Kks#dJ`lm5L}4m&j^3Hv$voSC}JweAOpD8gecK`^q2& z2I>2z@DlOW`K|nWA(a93+-&bg=|iHw7t?tK)0T18hwbw|OisGDGpW3r`~+^Y=g_K0 zm1;-)ql!27z%)YXxL%AV)@FWqdasKCYQJb9OGx}kRV{W5C@6Xx zX6L3+3}~ceDx%13K&Ef3u9u$lY30{&vgvegJs-PhFN^EB=Y*-8JWir~wNr;Ahl3TD z=NDQMSRL)NwCviLkYL%;66HATVd*ZP6OLEBZHp3%G4ZT=uo)h3Ve^<6!6%+{ z(~UdB)u3X&TeIUV+4O>`m_j2JlTI;EM7x8EXm=~5`^bsu>!mCG6Fd+2riviDH@9DW zz_;JIDf|R|*kh%?mWM4rhq#r(Uio2CB1;$Ng+zzi7B#vZf+3IOH^2X*>M!s7?&p7g zSGUoEFswxsKJwdj*Z-jeF8Mx~# zX@pMr{qENpRXa4ms}&V4MGM~3{SWE_#Es&EKDhQy&+-Kq$*QTiWEIgUJ7uaAuO!KV zsrgW-kQiDbhLz)ho+XY$8JlS z%;a)3^qJ{TW z_a~pPA$qAnEy9=}pj4@(_sZ_jmw`GrYR*x~paZI@YMz$YBZ~>TLQ;cNo%|wsET;xW z?B!1@R^m`-bfdAla?s%dNtEN`J@N{OxoBuqJi3u-#})VHMG8I?s}-0UXomuYTHh-H zgAPc&aX*liMtb(E!sg3hJB{@0-?iz_^T-zsM$AvIpGqeC*ada%IKI+kWbz!Pm^z9a z!XbJeE3nZk)v+PCpUzlVqXageOyv%@<$}(@SPo7Bh90fRc@0wVt9fVwmofvsQSPfe z3xE($XACEJcF|RSnRGQ&n_#ilnV>U4szYR(&mpqb51PQ_SICFJ81Jf|*J&Tp+p1HW`Z(lGC+p=kEwqr1v3Fzu0 zH7Rt8e~P~f4P@lc-ON*Io$`Z=#iymW>EntvHzeE55f27KiDgz05Gca@hp}w?b3gy# zH;fe{!`HCI@u|+ik4&@9&xUR7SM3rrx#i^oq*VsVkwz{DAqH@?tIc2O>CWB!kMu{_&A z$0BeHQvxF#+5{)&!7)PYmI{ohPV+MW=I{S#uOQiW44A`4fccPOfRnOFPs6So5s%ac zP>P=%c0_T)QA30EiY7%%fSTDi^QS*rh;<$R1If9wuSRx zdN;2@0VHsulcJ5JnC?^#I2@n8k>GC(-f06mm0G9u;ni@(HX_DkTgL($lO=5OaunFC z3eGg;S+UzdoEhenzrz2HD04lNBpaL>!>{~2^=BJ?{K?zdUwicH)8P}c?N9g2asWJi zW+IM80l)|X;rzD8nJ;=}+{Q0;0>7W0R(8MviFHu(M9xD$7dgFJ`K9hB-P_zI5FnO= z19itlz(W5XtiUlTi+`Z%!r2YlVCEqf#Q7%(V3q{UT5trtsiTeHXoHG{Y zyKZ3`gCqS9ynadC5q#P0k~m(zOn`N3P2odV_j{$zs`g3aRx1nWEYDAsE5YO)TA|8! zvvV%@gERc5@FR2;uTz#NUsAf@p8Ei=##8Hr1afK4eIYfRC5_=Ni?bF#UUV90GKLn$ zx6}CTmZfpG`yBjrs@n!`6PU`~^a{U1Q9l&LY$Ui$U*L6_TCr7f(gMB44tB<7%Q#*= z7r*9Vm*cItWm>MlCbVM57LA!g2Al~G1U)peqjU+o-7W`J(|tgPeZjYzw36Lnje$F0 zFhHNAlDorgi)c4DiZnBIFv@jznC<1)81eJm{U7fQQGZ^jdeKdn-bL8Zo+HSn@5>qk zw)w_#c7+ytZ}+kNu0HeWTei*5JZAF~=EcNF-tzrTkKTFF2R8J`b2xQ$qP&Dt4eSGj zZhJlQVQ`CDi36%I8lM6z2>s4EoIZl&Z%{4iiQQy4#0pDmku~48@UaXjBdQLYqJ=>^V>voY3N3vt=I7)xU$c{NeF|`yqKtJh#2*GTU|7d2!UUSdXfPV%Jj1L$zun*rE7p*jp4f=LosV9 zvYLvx3bx5V+vLTI?#qC7bn5@k)>}Zc=Kf+g^YB{uby=hZ28;{10Q-o#+ zzmTI&Ph{NYhACsU;*ishsj;5X(J> z;>brNlTPrw9sCH`1+BOF&)F@u*m%xzR=fP$2mgKFG)But-Po~UF%!M*m|5#vE$IaU z@@_iGI|afi7kqn39Z-&^hoo_{f|kL!pA>ODKShLrwiB;b%*+|Zqu#oe_2xQu&InF&MebG$TkFLV{)eHOu#EyRY)nz0^Rg2Ng49kmLB z?g~HlP6bJ0r@QRf8>%wWU4;|_Y3N;4gnGuO^AhDv3UHNob3cYwkp^k52x}Oi4ysv_ zBUWY7`xRQJEcmGoEAlS}z7|~pWw&(4%nW6RL_H%{a24XWs%)1FoERoG2;D0v%?#OO z?QS&dVa5_9Sboo}@WyYy=@O^6t#YQ4IJFI$(<-mQ3#n z9(1S-XyEHw4%fZj?~FzNg`BlOwb2<|BEA?J=h)zyLg#oDhqSrjwOm1hr}b)`bQ?Nm z;EbSS#)J?v53vHyq+4%@KQnc+?Y5f5jN^}$1|)@G5NYrqUVc_3gpAkO@%_Ep*AGeH&<}S@fcGze*(Rop$)SX z6+`z=)NgjE;C{5mb+$pt^!`-3itMywB~xvrWS$I~%TQt{MPJl}`J5`!98eI9tZ&f8 zUh3ZOatdZ=4-sl0a&EYnFSt(kgw*i5=wkN*VgJ-Y2W;4Y_8!RzQnOI4h?bs#HZbc8 z8cRVlf<763%Ia8R1)53DU&#Dm%8=L7@h25oX2+mOF#=6I#l%u%9TkDa)kq9i9Fi!% z7jh4BgG-l|x+i!poj8^Q!!BBk39Ik==nJyL#N=z&zV`X5R?>z>%U8MRe>TE^C#YO{ z>yo&Ky8-Id?#m9jpI>-CG*!Hhds3tcL17dr=-?lL`E-L5Fzd%+ObwSR6A=49)wXp; zut9PRk)l6adTYmdo)bpW zoeA`%m&JKB*0&8|G}IV@Vho!RUX3nVo`2Q6@X7m2&Tsx?fKcj#x+&zA z9V?&pMx!nVDCQwW`lyIXh@7?s)p(bs}*}mH9t~N^-7|=nqNnE(AD(qkmIB$u-{o5 zkQai3>vXg*Xc%c|n06^1bm)CGH3;~+B``Xqk&_kF5gfxHAa^M~QLKS=w@-_i+wW2t zaANvNb zH591?f-x8aw8QNxcbNb!Bvqx)PUfI%l4QT4K!~+>ILZXg>J)mF=OOWat6lAa{>(nfc!yQ);SlhL)mrO*h|1WybDY3Tep=Y3QOp7! zza?YZif5Ox`WaSRVZRwTfJ`aqf&vlBqN|81yA_>^#E{StR zs787@37(nCOg9YeXn^P)csCN|buA0 z+Sil9+Gl6Z)<1zMP4w81MF=B7alBk7Fv|2{q(nL12)S$BJ3%T#FJxL_)mYJ>Ln~1D zV!d$>w?o$A1b+yY_Z;xqG_?zK<>rt*95{tF$r`@{KDX#f{yk^a`qw`31QNXwzExxT z5{x8g#&z!5?+~jm!K9uFzpuS%z}n=VU)?0N>4`+)vsg2KCOi7UcTDNP+0^&gahSmj zQFtdE!|Yeaf~liYonBA?{V-j0OMt4+xgh|lXeKsv$p#kYH_`f6{<%r?tMSz2*s-`V zQ(U$KiJsP+rj8hU7nU8>w_kgdLSSxCZ9pMRvo=ploxV!ewX|NkgvA6?j{2c zcvn++ylAC=?`t`Vdgo$N>su=sdS(QDH15q}gBi0rYutauvkNrMt_U^7jNLk8W>A21 zR2#j)yl4XU_(QVeid4r)|6FmosAJ}F|6Y1I0X7*v4#`0i2;xor8}qKy8Gb4xCd$+e z9$4oBQ~A)!(K>CK+TzqItc9peFR?lhtbzi7$Fuu5zBBC18ZWmS|GCg$)-?WbWhN=N zV-xzUk)iSt#Q;aqK^#b)EYK^J13n${I`F<#_)G7d4>}-zT4}^g3Y{#_F@m?b>AT}v zgk3Heq=ulNFhPVt9o<-43{_*z-@X|-|NKMe3_tYZHwCoI4?+bHCh(gj-Jv(!Tf|L3 z7L2r;#UbMv>}M@3SzFq$!qVjLy|e0#IR-2_h?Hd{lbwld$EM6dBjoL+m?DboM(3@@ zXQyWz@AR~K$$&!=XW4=eok1@N%eZ}kR{~SK?~oj?QvY6Py}ad}P1j0xLOcj#)U_e) z^4i7y&W+-zIoo`JsT1x&;#TbJgaXu5QM)`5()m`FS=K|@_#~9+)3*|emzj3|o8bq- zWH+`T!Pr$SxNQ+auDe-++@~XXo~CbP)z=n(npS<+=P$Hb3_JFQp~QSxQYw>TAZn6AMdW<- zk~o`g3%=&M)jiU`n-?j-0ZC2axt^z)B2KleF}#;l3fqG3NuuS2Zms+>`FdU!R~ww` zSt+zVRh~s?7|9s5v4}9Q`yPuv7&PT|U?V#0*fur8_<*L8uFUwbV}}IkeXVX!TDF;% zV8ybPtb5njo4$K(Jj%q55n+bvJBb5BvO+3_SZsrpqn)yo?pAg&EZZV`g~Wxdn6J$< zIIzF?WPUy=Waq%zv6?w)q-JU;rjjE2sR*FXl$S5Sf-K~n+T-(X7WFO^06KrH7z|GF z!V!(ho>??zak98OB{-DiDs<}Kg+k+ZH&gE>%1h<9zwmSoOw%-lzq=E!99P6bu4fx_ zUsftdx;yP>A3_-LPg~ZayW+xmR8{iP#7iQfL06Uv9bFtSD=r**;EFN2PbB> z*ea*PH4A3GVvsdnztb)uTiMAPJH|<=k*vw4m`sZ7q#`izg@shh1z?tNw)sMkABP2F z1yqM@OX%*fW=6FqNJD2r(auE<-ieiSdl@Yc>L6~obkWISD`A+d?a(qFsDN0p3M@ui zByHdu5n=;WWdC}jk!)ZG6+2c-`9@Glp_uIy*@hAQL^(>69X_{15BcvG%!&x&KpUI*G*0z>gHs z_18AYQ{HS`^z^7P9A6K9>GU@)e)&957-JP|QdqUmazSBen)2E_sL{|pg$dXrwEp;z zSpVBbU&pb0(#ERY+Mlt0J@X#s2t|zc`^EeH(!YMybU5QnVx^dwF^Rp{`g!ynv`PcK z%dw)recvkQI{-W*`SOLSuQ9+E$W95ec$wG^|D zB4t!WHU9|bu&jpD!;PMbGaUK?yTOg@CcWeew2N(!cggPqyFSc$k@d{3aJ>&TJsP@0 zf+ewdcRy5jM@_jUNu70Hw!?MRlzJsTkwUNV$KM)m54XevxLkCr^uQg~+6XpZ)Mq~Q zkxq>JBiAqRlIWp3g7v>QHEqqk($$v)AxD->&VI`vVXn`sxlKN@;{v5lBhhz`V$M+H6Dnd8%p8dW%4O(yEriLc zhsikzhjmMnB)Y+iIZ(K^BQy)R$F6c$cn>;Y+^k%=-aVFFgPzJ*;1|0U2!;EI^By^x zFKq>JJzt7heJioD~{wHE$6Y&`+-W*2CX_%YdyvcK-1y z`Iw!xW5+R=PmPdxfnv^4)9b$Ziv@ie{C?hS#n1>=nySE6oi`h9`u|vNb{hiyH)Y2VzUdh zxGDNwaH8K)r*5rnX+N0w|M zASWD*2wMvPdGp=&v!*31c3Y2YrYHr+JL+m8u@4d2W}$_e6#2AG953sW&)LFVescdr z9k3HQzJ1^N0nZ?E{I%FMfa^T0Xk|Oa zY{O7j#Dhh@S1W3{H^gd1n)1TjH0OPw7fQJ2eKBxY%e~`<9Gdt#UEI#DdutpU!D_=T zlZ83mW2R%vau6&pd8pa zKnxP2$RYJOS^d(e5=a;hlP_B+CV?XHR7Am|Hu(qF2qv7>iY{rtyzMI)^alABD(V~Y zUtRMhJTsB1gjG~-qELbF6Xfr(?(=12pmPBp4Mqxe?u{H?fTInSZ}HQ$!E5JR?%7&p zpDYv^IwQj>Z)o8nStq|2=sP4bx(b&Q%}XOPU}D-Zkx@o5B@`*7BKDH&b6e@2K%D|< z<&s#{#-?>Wl8U+;uN;y9>?588D6$7%=(U z{BF9ekhC*7%Denp-zvAJ@ZB)u=q#sSeALc=eW4OUwO_H9&9QN z?6-!3nDK(E=O^-ZGWws5;g!M%%o)T+2 z7Aa;*tXuhMv)bi_@)OF0`K#t>oz5@RA3>ZAVr-ev7j{f|+8z7Awo3W}@fh#RU^P*FW#qqXrIZeMX;eN zotq3v)DG}2clqQ=VCoBoBwB%`09lUFz0?im?hI^@B3ps=Mf)T+4uV;^4i~JxPpo#v ze!J=xrGqFlT<|-)OTJ;+U~IR^elsn@_hmiu{9rUdGnGxuasM8t3EEVXPt<}v zHhE%VUb2&9gLxZ1c-5d|%KhG7O?I&hw%V^KBp(^6%Ka4cAw~955hy2iQuh{ZnX{Z1 zAM%hs?!U?tr(>$J;bhu(dS5MdZuXadJe# zv?zkqwRad8v6)8$LsD2ZPvdu!e@eK`ce`UQw4}x`R>}lgfVI)IyhDvyZG1Xn^vjO@5Bk%3n(=713d-VuZm1Qb%uJkRaGa z4htH?bH1}`arv96Z=4Rt7PZFtP2o7xH3jAq#d3~ML$kb}KE_G4NwpeVWQ-&!tZ!_C z8(=oN`X4U7_g{1q1-+^a=p65$G0%%5IR4K>87^GR@BRq=iIYOV_1DiBgNfq&u=hOK zGG(OO?02NcUL!+9LorZbc2o4mEjxy9xfPGiqi6n`-j${tjdttam?4iu z<6*DQc|Vn&TL4j!p~U`&i5DRj0>Fw}YuQrf0WgwWSpD0B2|)(&A^V$Z2ieWe!?EK) z{Ba|}QbjS)PG3$%#7^rZwI1ajjp6U+e7W(<=l*^3H@eny+Vm`O-Ic z{y6QMjp1iW6%e~_5|uhdIhBRpf33v{6U%ks1FsH>!No zIfv=f;W)UgUfR#ChD6eC_<0IAi>-l93-LChjbiqsWsn;6uKQCy{hqf0L3N8amXi#2 z2(n`@=YSD{N+_m~A_drilL&l{(1YIMwn7r?TM5(O(UO4}+r4>Ew1x_+S+eg{Rk82_ z2hXL5RLG-~DDUT9;OvyBQfRd#R}?dC&;bUtz|{F%5f;ko#+j)ifioEYvj$>x+h%Rp zjEV&o7Yn}=^znZfU?UK0`ZhT`h1@s#ygMnTog!DLh#q+ikb0kAn9pr+jpO0?oYMJF zyOk)fJPYnd{Q|Nx~qc7;k@Kke?6-Q_4s=t88YUY)h9sHQkPNk}2W*^BF zMfxA&waL_;I*$58Ff*=@B$uJf*!AqPI@z>3_EtLX~CyR0T^Mn=(-v#eqMuZ?aP4@dlO~<>ezX_~A?vbT~)jxYy zQpJAF%WoC_CHjvB)U1h)T|m0*xH4g#(fHrT6w^p~96 zc&=tf7bsO#JQqo;Q$z!@B<~7-C7n7mSF{!f$6*ox#R^tcpXB0Hj%9+OTQZbuL!pV- zr2)ojpt>De^?*+U3<28sdds4CF05TG8IWZt^Lc8|-jMBzewZShp&JbYZ>pKmM~d9k zo~vghN&10AO#jxfE~j0gx+cxxptB#(Er*$TIUJlzTKgB<$14yf}nhmBqbIF3F31N_`4p25c-x6{84 zH4;_~Iw0u^DlzcaVV+cR$V1W~#Xfe-Na}FaAiX6;Hrqi5y%OB(vlr+nkxB(7IO9}t zB*oYZDJhIHuGaaV7;)a<`I-Bm!-3f?LhMijy%JQ(%?cVX_Qr^;A%vO#Z3-2ZugxRt z^bb9MVLC9%%%r8$GVRR%FPtGuo*WF7*0(Ia`NpNipMLq+qS7xly?yrW+dyF16dudl z>$NJ}@Jy4ABr3cnz+Rf)nC;peMmKmn<3Y=RfRYvF=t3TX>v@ z=C@*m@UH!u&j&oa9!m|<=jb=yx=1$KvH1+n-!N}*C&d8M%62LO39*}?2|&lhgQ2M| znAePriHUOL>ODeMuycg$*uqIOk~^CyCXOO8RKyKe9D1ob zAc_>I8u(|Xt3I06z+dO0YLH@RZ(^tDvBA*zkzx4?OSFDk)0y06hFLf0?7soohmIM_ z>fkEheZi_q9$JglKAlP{S`-%j@fLrc7OiHzn^=7(?Kid{c;DG^?oia^O3r$cJB1uE z`j+jZm@@lxh-thXRMXJKrACPKU9cc>xn6pgzn3wGm%Q-{# zFM4ko_4}sqcKPaWR?->Hk9h+QW&inrTqnu%2OMs~Gi5#%^FEW_p$8nUeD49ta=hw1 z;GkXHv`n{Tewe# zI5dVFbWicT27HbuNyUQBz+zDz(4IyLZaNOkYIV8D$>9_@4?3JBv7AfLBn8pn36hs% z0J@c+Jo+-f{!3%;c(~W?*cLIvy*?CS*FZynD#5cOIEF!|=Y}iv93AAQaZe~8i5sMN zDnohwH84?HAy1wt-yAkDCoAY67amc=l=yyUovVaTnaz@q{>hJibqshh7KFN2qrb;k zuJ-@BZ`urlI{N0WU;c({pF)6YeVDVEM=?1R$pY?m$d`2Ux))|S7RuV^Wh}fZZx6%C zOuIN~(4)SGt4}o+hm<;HbCxq_7sO7h@NNpnqeCQ$6mbst+@V*6M~6P*q=(qds2T6) zJeJSUl4o8L1=Jc5DH;zcnEk3&VupDX9hx1p)QUUPPeP)7gXdk|PSP*eI%O>E2}u>C zc-sugBcsZ4e2@IRt0g!&`@=WS#>eci@%^>m|FfF`8;w6NT}rmuF*d+EA7&rrPz=y+ zq=S7#=W*&3rdfqyf;#glS+Uno~w+~n2)H7`#3`8pfA=Aj& z1_;7y_xPt=b3JVva~+Euc_LA4h!=KP+4#oysy<)UlzH54dcFI3Xazu|>?JFCw*oQg zb5Qw-MAZsonB#b7IjW~KnI}XkjfE=2h8TLf^h5I1hA$dq$xpAJN+$c*h2!lwNYZ2^ zd5%&{9Yqd7!2!S61Ji)Hg0dNFIVz+!)>pmhQ-SwFmSP!PqFn!S@Rba|8eYG6hXhN> z)rxpt3p7Q?`Q?Qy7bJ4kGjz3eUEnc8Yui=sGC_jpDIq@DC5@eiY#lA0z>2$Vb|=&| zpd*cow9Bu!B~D#KZJ)Ms`q;oXF46QvMvZ##6d^2*_9m?UpYaGAJ9c=?ly2XhdQ+Uq zy(Df7Z-?;kCGqL-4WW0Ti5jUbGr9L=F%WvbB<_V*Y4j(evF$dC4KUow8XxT6VoXm2 zeRBV{DF@U`5|K4SWu+Ub>voJF({Z9a77F1Yf(AzMK4pGzi<3@x=(||7PRA8_++tcA z@LrPlT_}~+KZD-X`Guz!#E}i$*nn70zv~@-mn~WVSVLw+=`ki$jQY@Q=KB|ZW6)2& z@4oU1x%$%Rri@lm?}rV* zRr9yQe5p)jJ&k`j{zH~~|H?YR8xf7;Iw@Eqe#_MWyzkcRI7>E-RBvg=8Vi~qhdEy9 z6azilJE(}1AECC2mLGD*0)7qXuG8WFYJ*vyi9t8SH+|B$tEb$hGr3Xjk9ljRG=+DC zVuQ>{$z5Ka?FW936$_(E5G!0X?{cmDUhB6_t6A)}a@$OON`|u3y-{&W2!<-gZ{tI- z&{nNj$rJ>uCdy(N+9Zn+F!aC#{Wgz)krKyuRquUl+U#Mss4&yMvE$z^f+AT)waF_N z-IwiCp5;Ih1M@v_@9)*5@vo6Ks@>^IN-NK?4JOsnPVv7E;6meMZe{1Y?u?KpuM{Od#F|M%WMV?;HxwTf%=9=Re(M1d@y(K;P{t%O!>3h0-s zmHWN2=|cHYCEkRRRfS!wE4@xAtweV@!j6s=RYT%&ge9$VYH!v*vfnfKJm3bE!-VgX}P>((l|DTp>hhlkC||XVbev%b-O$ zQqVX<%PZsV@o8q_7Bn;ERCMSXU#(NU@Sfz*%yR0UB;WP%+5u0$v}VRiX_{NTw2qGSPxjQ#(l~X7 z*ZSts`gbqXt&H#42Y-=T+0Vsk=GC-DIT&oW4z_s_cIn&avR-u7L?jZzsqsMBtsh!8 z9wTZTUv5i8%{Z1H^P}c>e*}kW@}SSoxIY;LQ|g1dDdd(NXS~-N384Xsc}S5yY;;B1 zC~RX@p)~5C?~C<~tjJM@rah8QA-zMR0_d3&<&6rpvX7hz=#U)bc1T);t$a)&!6bMG zn=^O#_qNQ>==p1n@uUP_3EX|Vc zP#mkEp^qz49diV0T{l2WXDw$-=%52~+2Fqu1D5!YdI81H2mqXP?rKV2fB}HfiQRih zz8wSbm{H_UOEFNRT1G{by0`Lo(d7)FPN}QkMNS68px@qCJ7!|K@}Rp8duLoPPfMX| zI1m+f+AT{3k~2N#3f)?qV5m|Jr;=YBjE`%b?#p&AP{oH-()sgu__Qi^g4fv~Ep;!I z?PRPX@+=WxLw1qnzsD@Mo;=7-fX(~!FaGkT0Y8&_esz=7vP)ywu?5~{gtpTZbBZD- zkp(W6Q#C8sYlEk%M0UtsPxgVclT=mYk~kxv+V{j_9oAa;vE*8CCJp3zyjuS%-w&av zu!Gi1K76v3ZkLyfZ^#>5&oAufF2%;(q02}wa@Zr5S03CaQ6WNGi1x>qe|>N98mBx8 z|BL0v@>N?neZXyhPFNs2<*I)J6&~WdTFf_V8x)R)*C#+5^RA3=nTIqrlV*I~aaZ$V}U@4OIA~$AnNHY*-r1{w7U^0#gHs^dRqg&O;< zo%(>DR_JUHE9GDL`YE!Gow~B$e;I-HoMCCvWQy5J0hl7tn8_tAPR){!995?m6a+`k z%Ujd~N$RfG)(B4e_QQB{4Vbn4&Z!et#lyO0aQyJKY=w_sNe`H9Ku7OSrK`wJcIdEU zV+J^hhM}W~Vu0-`kBT@&rvjb*1;~-0)qrG0s!Zh$w|*Ctefm*5#})lR9oz5R0xZJ! z#l<0mKKa4A+K6gptt(E&$XHmT9CX0yRQ&^L1xzycvw8&ET9VN+C|G`%*IQn)7rmiD z?_r}~t=D&riVg|n&5>YOH|_oq(bEZAeN)YT%q(}?>^j!Iz)BBJE}k_fTy793TjG7s zk{ch>PNGNyoL)7BO;xZW?HNU>(j7@89Sc1pcezG=VhOK8FJfAK~|H*)1gip)A zPx(4|=wSZl5#l&@ta-K>L1-Pttf5F06>)X(vY#FQ?$sY{{OZIBm0??+-QL?P{W{)m(>Ar8Zn)#dqTm7=Z~+8S*<4V=CWyP>phj^;B03H;3W^LW ze$SHxC6S{!S-2T{ufJYS&ho;X_n+r|-e>v$GnHAg-bs4(-T12LO#!gxHvzq_UX7P8 zhh>YB;jvcPr{bkh09ngh7MRP^%RUU%Rq<;Dd5}sdBnQr0PG==L>#sR+(b47jrdi6GbXIwRDQ6CfQPWzgGReilG3vvRL z%znsmE@z-Z=~6Cb(j|pDGY3axjC^eGDT7lQ3!FmjwmQ8g!}^b3tMCsun!BO zPbRvCPLW`*7GD08KK@1<-5YZizJ(Y?fGk4@4xqaLEjb&eT?#)VyX)WQZxWh|l6M7l z(D0>$f$t|koNtZ@{x;d(@JwZAL?1tSu$Jnue0LoF&OQB-75rB5BYW4!YYTW-+?@L} z41B}Cr1EtbGQtoel7>`I(cyUvD4h}?Q9Sm+G!xK%9A@E>Vi!L+9Lb!3hgXDNY{ zbx^1!onlfbvJ~@1CVFLbgLJ`bYa?|?*1SV;;&l&Oh@oG*W3xQ;vz|Q5&2G&9h9S#7 zL*%s!3`^=NDJqM%E1@{PRJCeCwW^CQ=I==0w}E0?bmdXm80qeP?psZcYVq^DD;F(Xhvwex)K=5 z&G*BBY%O`x4%g)m_jdCOAt{C-zXFgEO7lfQlnUQW<*h&*@Ht9hj?5i*5Df;@i;f-n z3x>Xj6NpB<(^vHCA6P+j{zv5xNQ)PPsLKXK?G$sBBA2Md{8+2Mks>KgSgh=oRf)0o zpc8s2%2l28#puUA*ovuJH4ZN)``(2XM<}SmBEMYT_R-z^wc~G2>Vi@OKwlNz9Bl~3 z9Z8ECH~-x+t+Gqf6H@c8C54y5E05nTH_4y?qD_e52vjD3T?_clrZ@x?RO*OU^KhIE z<1UIJ05QInx17n1)eZb*xP?M4lyhSI14?-ZeV>m*&%*3QFl1NV<_0s3BGkve-R>oE zf?1mLawK`utE3=Aca&_4Lw3eiSzQDKl5m4HOv~I8``|E!o;)!;2i3n{^w2&ZDJ6nm z{^!pUfBDn*X8vAEGd_g!;#OUjtYfzlOVo{YN8E+TmbjeIJW-wSHfi#&g&50r#nyy9 z(2|!ES|>Cq>iL< ziavfLU5Q+M^mT2GtnXi!y;b-2ig!Nydd|CT!uEGBOVb3{dDSGZ5flcN@hYU(&AkOD zBFZ4#*8QzN?*4Z7cW-^^)LR$6JLfA03C7^+gj*A^PVYp-sfadV?zn^gsPcBuESY$e zG)PX-=?TY)36{aGNlqisJvd_rp|c3=RQvApv7)K?v-kc&lDMJCi=#7o8#JX+%yNn> zp%Sqa_;}Q+aXL(f-$}QTx{qLCEw{LQ%*i)$XuLaBh-4&o$sINIZxS^U#NP zC#FuY+L90c(!QT;@#5|EK^yC^hhlb8qzq#;nBl{IlH006)zJyXv0Laxnv@`;7I)bu z#e8U#+$t(mRr{@-mMb+x-yZ{s_Q-kRZ&7|Mva{?|Mmf!(JaqT+@63$;vW&qeZ**cTS&k|fbi z>G`OZDFZ)Ii!h&_W6o_?K?7ZdVvQ_a&@!CcXKqk>I@O*&dS_JpKgJAff~Ln@CfOrN zrOnrOE5*QT&Z80+PR`SN^JdK*We@P!>9Tme{>{Cw6Snn;6`CIa&ulo(GLb@}* zC&@#VX?pM954s$KQ;usWCX*s7sKkM?g?1(GSarwLS)*|=>7-FsRTvU*OOwC}3j-$2rQhh-!|YfX!V7sNs;-o8U*^Qc3-RLk2uo2^2VECo zDJ3?n!>QtQ?L7Be-U;O=+#yHB2gQtT^G>z{Hh7&dp2wKh7Ut<-ziv` z;pG@SoV_b{@*iTSTi2MDAI#4qm0rBlIcu|?9Hto9`W&DVH&6dW)j($EjPIM4_J+-e zM*7j@{5R6x$d6e0MuY5-dgU8^^dYt7#-cDJTBrerh(gdh=;QB%Jy4R!5PcJvI6=@& zw3~lr+_u0~W3v2>+6%nvQ?8P=njJG{|2?1I$2V#}WU2z%golDwjVTX08`Le`9*1{2 z8)tPM9Rn8V1wWmH2PdOLEgbpt|F#eDd+nmc60|?h%o*LMY@ggF#MY5@zWKhL%H4dV z?$E1uDohCmKlE9QT0`{Wgze&d{!Nm>?r)U<|z)@4q#<#D(CC zhtcFDmbmgZK1g32WG_+6WeM|QznE2N_o&Fi3jiIlXMrlDNVLok*k3ZEt}Bb&8*6#- z96Ao77mwx6JGSSvRes+48W)LfDRo`)EZKRc+bmO&r&*%SR_CcUf_$j1Au?0BP+JD| z`_3*W+~2q3vpozx$)G=<(*pEL%Qx1>e%0#Cl#hP+Qv%FY&lV9~z=+VfSigdJk_0oU_O}`>NL02g;1!shI(nl0W z6nUVJ0`+#~@!g8-SZwaVe^}Qta1#osn4fBv_XR9Z(B0sr3A&}3%0B=30hyy_&z?Om zs*Bc#t__0nm#XL-K+?TJ>kNL#WF z4}#VZhwNxOb?j2Vz13k+VXM(pPl1X^qzZ19R{>F1dVFCR((qddx=e9~;JKpn(_iq4 zhWP88KE>^IJN;`oKgj;Nhxpcd?{a?lM#OAaWJkT&95Sa@Vx)z+6nG@UN+SSRVl=s-@t|C+USx~(|*GON2P)!pQeay9UIn+7 z1F>u;J4r%OyER>J>r61^!?NZ&gAn`K^u*_^ zFODBG&We!v?~hqVa$XvQ7;F%-gJPhssSpV`+m*@z6fhq^>Otz)hRlk#%EW&llOMn*U!&b^|j=09G!tgv7UjDGl+S{ zV0ZvB#qS4Xe9?-S1uMV#15)k9vEhq07&u8W$0%~xf&t_Us)X!#=yq`)`>0KRc7ABV~SzrsEd9m z2iDEsFs@B#xl^|_wqQc8-<;TH`Ko{>8ujO9^~(A{6KvogH)G@$j64IHAy05Z5;q_I zU7u?|wC^YJ+6XaAhe*d)%;lBkz&n@op`Z5L zSx%@JF(dWgZ`!xaz9f-emIm8(qnA$vPg9Isbx8KT(EkQ8cd_9C`Z5KOql6H`L0|Ms zOLBDbSf^x-%&=xHo9nqWs9>ZE*Z%j(0tkq%p#@7FOkF4`zhXqLD2k|i!Qw%Vr=A&!@Q1cL2Vn7~fTY1c-GC}i2zuqjzyGGuzCzP$7bcc!OzaH8 zUUm!Fs=-em+9az0GlM(fTHdMvBQwmoXD-li!;CvT1Se>WXses`%LuD+nUM33bdvAI z#--lIxa^`BaAHcRM5vOBK}vcgKQJi}6#1c*Q<0z^ywg$zZlJ3pD?+ZQvZv+Kl~W&u zZ6e+Lg79{bsqRp0l))`+!lY^Cl3D?teC%^ETF+p`X9@oT59%6liF1oFu}&cu;0!h! zLmxYgix=L=4LCK+zZ-OS{ArLXHEPS0+xQlF z^%m`>FJA;H<U>GNGFNg)ah1&RP2m;u_DstF1U=ww6_tc`MfHR0ylTqf*ZY0CqPB~3!acCN zau1x`1EImA{N|%IJ@$H9UhC+xC}d%clwOUTKD&bSYT$P2h{N9a7RWuXS7Av*9#aI_ z7YUb8P|LK8%A@gWrqcQm^2yk7Kj=&Msc3|w#F<;J3ZAS9U7b>l4~ zKK4ly4!?$LG+uUd6oTa}eepq&zh{+>Qog!}gey zv-8S~#mufatCuYHow13Oc(Ip!!p2Lkp%}c~{Mtt7;#>}n@jXSi(x_cOt`m&XYDt@L;E9yj)KI;;gbxJJZ@0jb zwn@5MSuo+W7W&~rud4QieJb4<*3O%GNU?c*UcgTN%;!MEc0$-cc?Lr4zt7PtoFGJX zP=DBFuejy4rm5BjLgf?#oDJKkM6B4fh$N>fJC$Yv-gUln#+nmN$a#2CS_nJzm7_YL zw4pBIC|L{zg#($!4*HH@yXc}HFfWdt0Y=`=htuSzZ?ijP8rL z{O0T}@NwIe4H0YJoPFp`qjm?s@5>jzJ^Rmhgh!c1x;1WJgi+fXw;{qQn7Iv0&sn{k zK^7-0QF#w@zcIJtQ>`2a~qAq>HF2ag0s;^D1c6p-1aqUYlLMg%k?bTVNq4Yab82@%}Exq z_GIw0*9GdyI^T7HM(xT`TNB!7ynj9~)qUgz?gpDPz#8hA?;YCrpY}x&TsCB09PnW& zp)f&jQ`)Pu@-#X0>euE%x21_Lhm?2qlskUzj0X<&vY$EUFe;vT(z7RoUwu~~8Zav5 zqy9veaWg7jTs&N4V^r2r%v!X7iO2^%R|ISTO}?9?J>kn=(_N0fHo8u7m~;UPYXQA6 zc&}oCpaXJOH3`i#oLXX`kFOXHReZR$$11-zdhb{>8~J)D>ITJ%u^=zy`Kn*;7_orGnR(~(%*}D_ zD*LjTXCuX4ynL{f((Hq}2iz^0l#Rd!Z&F;Ac8{+O0X?G2QY?wU)7atG1jO{s@;niK zE~b#OULSx$S772DRAjT4Vs=xc z0;uQtm!wApo1{mX^Zb)RE!qsK<(;E5PHWS{N>sfGAMy6i_)t>F%Lj|IjK3hISHena(foSnAMW-@7VUx6-hmDk=R}iPyTv6RqYzE=FeuW#M<~*#PtW1Lr zxR{aW-+SWg%zAqboR`FVWl_x8r>Tpum>wPez$|u=q)0UEo~y$gtIwZ64_v(`|MoY3 zV>KwhZT{H=^0603b(h#^^*yAR&nVJEC0Zz!F-DK!f2%4M2w}l41!`p^JAeT`57>A0 zYV@mZifQ9nou^dRR*$|GF?@~ zMh000a-h)5WWKYHpM{J9xq@C&9$y!1?s#aHchJijRC&UE*G{sW$y9bL>w@9FpS#z{ z46-*a9Mg-RkU>Y8$13q@ z$UETW>}hUXkX~RnI6~{ zYZJxfQe-`qh&o%J$eRPv|IlsL>AO! z&*2GY&;?9A1%L;R^y*L6zh^~B zNn7nv&$mO5lDS@Nk6_suCVp= zz5#GV27@A%VwO>4F_qY)C>OWNK+Jjr!Ok`e*Lc?BVEIcOK?KXA9zD$s5f9(`>FRJR zA{yUXQAswBBy~0~zJg-FP25f;nrM>@*>^1L7`RF!_YP`KA#uu8g8pHV2orcH08k!p zrT=e%MBSM&C9mr4#(zi*el^ki;O?qu3}md4-HLe_m?G!|+6*_(p(C^Nbf@vWBlk`g zr`hq|ox>})-wq*G*p&I?9tRB}?=>WOHh5l5F&Pv|rxG!W0)1`>Y4~B;(3BJ4ylYVx zNivnlPUd!4J`YUXAp5+L+aIR#j@MK z>Z-m`MW>OoF3@&6H!BXK!nxs{eu@(+sN?@?x;xowFMjsnlx?Kci|xfp8+%bpF;x`V zMrEW`Jtpuo3J~k&mZ_zq4H@N zQvx+Asi75OcNhUKU~<@Ca2eDM_i_R(ReE7UG-Jhu|AU_MWHqC+Z*2hgVliiSL1hQkUL(yu%z1Lbmoj;`YG~& zO2jH6WHVi?%vNS9kvY4YzjzdsR$q~JE3$Yd+1l}$%1gYRq()f9uM+M6+i(Ypo64pv z9)-N8Iuq!O;Z;xp)hW1(uyLHy|c?UtyxJ zNDJs%UV+RMhmPA}vRDo&NXs9R!fT}Q@58?n|3F3vyn=?oM0GNH?dKYPp^xBh$1XwwIL|s z*23GUSq}WYpT=i|niL(1lR?XwyYVK)>MxuVXT~;4YRvZmPWhjBg1X_UrE%0;J-di` zb{V_$o6)j(D^$L-duuK!9|`I*gB^tf6ayR@2Bb~K_$#WubN~}v6{yUfl8<;E7!dQCNaTe<#ob$M~+^$#^A zPtHA^VW*BbIj7zm+U=UVXR_Riq9b4X`X#c$ix)j;uOH+GZ=@Kg>0U=At`mObKUaiW z51=t?2)+$%^Ywy0Si?T@KRwmV!C?^))nR=?t88h2t}dbqD0v^nBzszi&3W`X1xkk> z;)LE2UGqC=dpQcPjqVaY4$rgmH*!MAE3)6${ldOEjmwtGi;J{b+SW|wY&5WX>ovE1p-eYE zJ>Cd~+500G#x4lDB!vrW$9F)<6Kbn{A}^S*R+A>s6-wud3N-jX4;Sk?g(~aP=sad# z{ABD`N}KJYx-SV%Sc20g8tdF}PgW!en;!U$*+Fy0XIADwNd|{B}s*!_o%<6wi= z2iiGIK4a@wiev+MD7`tc9d_%+po`H)Z6WU7b=*~+^<+2FS_EZ36ZOAY?95~;OnHtX` zW2n((S4F?DPXu}GzLq7mgLMxW^@qizh4(nZ!aSHGDUL4zQO@C>YsV%U9H60Q!-MD8 z&B72ZV6W_c_-?9w!yrpvo=KG>YKlycKgCOqTs7`vW#2UD##Hx7oBfeZlsZF^drrD&*kA9o|Gq!sqKNM=7rOw zc<%~+=|A?xeBO&=A5MD+Nj@+jEkOo|kmgFvq@}ljSlQC(cTsBO52r2R1X)kvbBJ*{ zLH3p7AMdNQ4|Y7;obJV;FP4x7Hm94z=-d2qAg}b90u_iX{1vzgsXOTp{ueCZ^{@uE ze)7tDk}p{|Ei)>G9i(=sO^X);XVAzfYs0lcz(S?;g6u?~&4wpvw-u z6>^ZI@W7C*mv5x!LFSSRCOH`(+vSq`ry&dt+bMQTt(j7(ecg(w$_am3M7F*(l>>)u z(6*0a_E2OemADx+K)Pf0Dmnv<+Eft!?T*PH^&lRyOl(LvDLyFy)iaSHp-FsmEI3?n zHI=^<>I!y>mx)*SmDBkeyuQLOM>UV%V;w06L?gD)CdK0L&GDV$&GBVZ=lfMf^oh72 zhi&F8=iQ8}=l?PeSL^!sx4vOT&rfci$|ie1XBNq5gOMW?(@2qeDltXU1yv8rf(q#K z)Ai~_g7py@q1WWg8PJQcG!vemetK$gd>`N9=UCd@_Ceu^MZE|^_ucWSz|WK;$qO=! zgZ%P!KGaVDOC|Qe6-?Nmc|6&2t7U*32{1R|#=9smOJ?DLpCi)k3cBsQUcH+kGlY5K zfRjPDvYVon_kZ}vzVGyN<@d1kqL%Vn8DNI9bdt^52O{AABRp|l8%|+I#NpAkYwg>_ zo<)TB?(v}7dT`~*N{UIN$Z{$%i#e(3<0p%Xet{?prekednHc z=%}4|-WZ;^#*T{C1!HnF)L3v{j^ALJC3DtRm#@cqj;< z^WzHRbR99x@_hPWM6V1B^}3ZlVE0A8E&1kKLDM(0VnTL)C0qzRBR`I-`Y28$%!{0-we=?oyT zT%xUmI_zG^e`hL-!Kvw_`^@}tb44?UBN6T-q+G*`15bZ__P%`ofz8o7Up+;Za@!nv zZ~Z0#o$*06k=YcphJsm0L|?_o(;3vqqB26;zkl&Z^S-)~dN4!(^?NfmY4UvyQ1n|s z9}YX`el7;~>EMbE_L1CsCGlUTe8=89;bKC(I8ejlx3tPGiZep7iFsA*&Il0i1xkcg zS-bL(v_atFs=|3{97YA_kzIS>?V1}mD67qQ6!7Q-1e>_6Oy0Y&th2EhDHOAmBFRYk zqF3YQ#;69_j$U#+;D3Mw8|KUrK)5!w&OQC*vh!cFC(q$xCcJnNVc`I>)Ua-i&5s2x zAo(%z9Y0XeH29e&E%e9A>G@&BVaPgdqG!);mM29ONi&sc)Ixt@x%u{JU_|Y%z6?$2u zZJWe9rOAQ3b?C8VM;p|~xsRklb3A9H9!Fm|IAiB>1*8X}8n{9mhViL4X~EK2-3G~mAAiseB) z7*`SRBCuH>*DY>4Wgl&QNtPv+IP5ubr!_hPic$MCxEprYBEtHXhq~YhB8Ga-)91eM zv2vvSZkJ`7xQxGD+zxux-AZ77&Qx|lz+j+36q2~+eL;O}ee6ByA?FhqT>o<&d95B< zsbBDia#~uvcTXtl`1)mgd*QW3k1Q4-ht7#IQ~6=PXRai36u0dD3ntp}%Z@1viQt|se;O8$DUvjlW)gJWRZCPl^jywb))tHZ>uOrvNd6A!u2UF@~+7x{36M&plZR% zAbidc?5QQj5e(UyJG)QrgN@Bd>=-J)n9!oN;>8p@HjC^V3NPMk0m(TVyd0$%5SDA8 z5_2aSpiwU;EK8Om*-YP1=`~$+qw1jy(lQ2dJJitQNj$(3@{huCRyYzl8nKql;T zEq6Tj!Fv{ostFF(3;VEr5!*-w?;5WN5=6MP_A4%hne#j;l0s2K$cKTD3zF&#eubhO z+7#Ew>y@>I-iUJAp1F)BJ8(Vw(FPuc*>UDj$Wp ztioBJ#9$j{YG}Pc&x5FEj-(pK!g@1vF=MOj{Pb;nr(gUKZ^F%UF~o0h?iBl6e6I3i zzw21N+rA}`g>Nr2wj>}G2q99HwfuJFZrMXAkP}g>Bqb4#Ca;os%CfV~j>EWM8^A;7 zIpM;4-GK>zx|ZyUf{zz3Tr3JcSh3%t&eIe^sja#C3~R7Y(Y-`>CfX#^2lNGO3ec;u zQwYUf+CjM|Nd#X8zoTSRfTzLlFxc%^8V`YdVA${fP!Mgu4Px1CAk8mU3+iqK7QuT1 zsm9^Rr4Rg}Pk%b}DxcmyY)72zhWCEEew9}G&3CPi$gfXleVZKc;&P;`HXhQ46a)21 z$I*q|JiU_2rYfm6zq?Q}pA)e>AW3wGVLOkR1f#Z|wDXpbD)r)N_s3*`L=!$?xss{G z41-?1gI@za?;~l4HZwL?x;G)8zi{#?uqQVsl|*ERmZ|PA^CsVjTduu2?vi8+obO3! z7M`Nhp@n&1T<&(pA-LOH4ep;ge@!^Sor*s7oe3dUR6SyjE++byru1aL4FP>5>ncYtbp9d~)6?P4Q+jy85jQkcf*!E@c1PdB9tJRf!+-G&yu}ysj-^ zPt>;8+W63l+zU+7O#!X46Jy*~XY|}JvopAT?tpH($O*S2zVnYRMS>L`s|yslWY?e{MeJENshS!NI3|j!3(y(Z+1Se$d~gZ^PLHu)^eL7s*VL&J8ABTXIz36RP609}5AB2PX2>aw*p1t?hU!Hhn^qJS&gofx9sIvaIYs$(Q_Y{9@eDkvOFV~bG+*4rf zcn*D^tdE)_D)2LEF~XYyycym6R6yL>h;C9U>7#R{9*DoV1egKfcz`_ao$c?gwAa$` z+L|~Pk&hDo1ztnsRp16K;B6j%L~}w~3;J@8CwD{K<%;ql)S2{<>r*iET`pb|UI+XA z9sDJVb{^K!9nt7D_*A!YIg_usImxI!CSEN`BlBZh z3YIZi>&@VmNJ^DD}iK)%I=-%lbdFL$jHgnM2Qy#qV7gy2`4?wNYCeaxZ4#$m`mfEwW5yIYeG|@K5_1wR>Yv(fEw(@A?HV#o4KN!LTl0<~9{x z8+lo+Slnz+Im)G4--~xdECi+I%(=Oed{HcJhVyA1s&H7gL&M|KJ_R2KbU%G)7>-W1 z{65Xk>fL^zJM(*z?8N~&a8CyXy=SfNZk0v#lMloUc90d*i8Rgih<47N-EJpcx3Sua!28U0u}%Y@i4C1MwiEL z))a|yM|s>O9yFT{;b7Rl_D}yXb{q^*(mmq6-){SV_Ss@CONtlAM_7{HCn5S#JmZ?O z3_>CK^g{A?LvS+LEHx3M*1+tDx;Ui`ItNmC#}kf&rgj(oa9Zb>a=yW@UvMa43ox`_ zpR!0})Gp$!lw5i1Lyr|Ea01r=$Oiv>iU+qI{j`Y?+GcRdWw-FQ?Ps9w1ZRco?pgc6n49E_j}_zWmZs#?}&d* z(zwOry%;XrZQzniF;EwtMJ3h>^60v#>;AW8r-O=vuaIQY$A9<*T@`;*U>?1ZX^Dq^ z>9vt1ezOW_%<&#mXN~@JSis~Ucm`VvJmvrc=WhMs?n&PYvqI!IJL*%(CNFkDYHcRH zoMM1|ZyS}^5P9T3E=!N`SIIvW&pfAU0IvfZD}YB@(InhSR*icYv{jZS$Q7&zDUV;} zU#?x{zccLNgo20yD4jY>r;lk9mZ|y`pr8zQ=+#R}gCL(y0mYUoq1*iyCvjkp3%3J< z-G~g~oKUBJed#TG^%XA8i5Ex6SVUxIq6AFew0#n2T@2K#50mtSir9YTCDF`&pP5_W*m2a z;AU4Y<}P~gH7inlB%0kM&x?@)RcV9bAv-As%8hnl12i(U-~xn=eckbuQ!T-<0vh=o zb#q01{x$xW!hl1GHnSvEMQ@=|4KPKLN8cd2s%R)sgDN+yKhmq0$LfkCwY(-!49uaE zMJuN+2y$->%>~>%vnIK7QR0M~S3apaw8EbDkA*J}^6PV`dkUi#4SKzxHbhTSL(3y} zMI0qPlT3=$2}d;R19KteS1P%uSSIMBy9CAYYe_Ow5K$YK&MW{cj-@~4{G)Gl2$9sc zZ`uRilCNz4;gQfbw>N6KZFb!JwdOQC-W!Sd@lSvJf9=EeURzSY5{RD*tpIn#SH%NA z%ffet;b-jDz^c|7x=DDJ6!6Z9&qo@yb4L~M4vSJEJ-ZlL(B>3)-3-*BhT{g_J^Q}+ zZ}vo8UW*YH2C-d18KKwtb3{dw+d@=DdF<23D*z@FqOGF7X*S%@{#=(k{=S!aS$ z$jzE|<;t)<(q>^(tb09|A!gqxi(__Uh8WC^r-FYi`~E){O?4$G9H1B55T7l| z<%#|7i)Y9YK)iSq_unz*!}AmD+1$9qPrNuV%fj<^DXf(_Jf%10YD|+nPh(`x1SXR? zqBBtOST9JAZ=(DBQ+S6-niBck+LdLzOT1D=zUHxy5!%C!2XxRtQ7~SwUK!pQ)9tBd zbC_mF@bc`oXV4kyfS(gl*unP9M4!6xeP~GWWr>(I35je!9mPR9t?|?%Zh(nlf%{ zzt`3;D$d=mx8K&V>>f-&(2Vsv=C*!I@*Xh(JUYO1M?=C@4xDXz^#Oi6&)`?8sGb5s zHj2doxuZO8&z;7|p!xK~At#J@Z=8E6nSXVfHPRL~Z|PRD=cP#vp0SC`9io^9iqrwO z0dSQg+_&*pigcyGNPUo;lbMA3BeP`rB!|Y1uB2&ZRa;CkCrz7$q~>XYwTvNnmW8$} zQ;Cg&kAfO_^T=5=uXPbkzSupAf?<|^Q{BfA{gX45sGU$FgLYA>2z0~>CLjfFVO$vy zM>f&Pl#`n*r@s+{)~DwmI`@6a_|a_L7p+iQu=1NfAl2NUk}qhZJ*3DzCw}4tl`4U|+Oa za!8%YPYZlN+LbpawnQ5Qri8=b)bxrw#i(+7E2>4?EM6x+Pg_RFgZ7V03B9C}PJsrH z9+JgN4gD+#y32q9XxrHtWX#>uE|CKo)9DCZXH$c>+M|G|o=|7!g1404AX zqP*BWP8~5gX)tF5(@&8HsDXQ2-KV@wY6NI3Ht{Q`uHjWifWiWhd+|)c=(Y8PCymiv zkrvR`R4LH^-YGW#Crjm26o=~b&y_s(K^@(C!HR%8QOAYVygSgX-uQOM41?d*Z|T*K z1StY^tnqI1$Z5PLpt;@Sn;DxCde5&gY_6{<*rNV|11(9&x|Kg41`Iym=&@W~9(#UEq!oi@ouzY0!B7QNUTnKy)f<$a z+D$PP6e)$S=8+%1dFk6{zqD{YUQ3%?`~IZ3@6M&O6yx? z=B?P}=xd|vBs$!uVG7y3-!F$Iz%fWTG?ZI!;sgor-Hd(fOr*-5=k{6YZSVd22@5GN zZnr?SvRb@dyld(SDbhdv4}jOPu%26S>EK}O@RiK1eRq=8`*`1=ttG`TjR$-?T8&=4S2Kv{=en}a!VrBT;#33{a=9Ud<(OS$@i5s>I|v{4+x~R- zYjO522$%eV7n>RuPpV0=V_G$DZ6sFbpeDQ7=vY`vb#3&1O&8F7U6!Wskn+ppdB}l@ z;SP_1Cq4Uf#J*n~nP9apAN-|#KiM+W-o|@fD00xohU}pjsM;)}5-mmj<#8QQF6_O??RpCET}$aR&&y(SR#%Lk;`Flx6y}Txf;Pn`c)a zAWOWsLUW4^OtL6uHAONkkzmV&>w5jMXSpKmoT_Zf(dp)u2V*YSytq6edz$-Z`UO+N z6$mdF&TW_3fiTpb>&r!7T5nG!=(V*{EaZS~!jc)6r8SUnzO8B!ZV5OPktC|o^h3d@ z{`Ff?P0%H>!QYrrAuZ(X7eXoAL`uBax}31FE;STmpg=spys#{1@Y`QK1$HVe zynPy^zTOsh5jHosK@>>04`LuGyaGSHdPTrPDX`f>Ag(UpfB*`7nxaw4BTIHzv~zkh z7Wx=L@?~j2zOq^cL6i;S+JsHMsiE_sUav3=qbE2(!GxQWlKh;mVR$lK4uFEqWIX@; zli%Rn0VkvKO8JNXc<2g=sTX@PEF`7~p9QZSk?HQsQ=pt>J8&ZRylx7<_;1-W@4kfs zF;`R=fU!6@z%$mu!yFb$+JSO;+-bT~c?%pI2nz3()d#N|y?mlOx6@Cj+%52WI-J{% za^Ae!VPAITwL2u1LMrr5cLiaI6U^@glBql%)d?=_^(p;|^n_}%Eb#PO`M@B!oLS~) z%ThM#vFCk)>e1{MM@#JQIB~2s@>*5UEOB z0EWlhSB)hU+eWvKh60rcT{*ClYc9vk`Ir zf#2zJM6Pr!d2yi}OUKe3x_eq9oi5qHfAFJq?>v}skTiYi*4zE>UW<7UgKo?f>8J6F zMjCk^jky@SBRF|vx?mqLfUH&DmF$^z(J$?d20C5P&TEd&yPNcEwew|sOg`4TI;c|s;lq$xi8Cwi<}I+_kIbJwNnDs z)>w@8=uh{P4ct-!-uquhs%;`Sr4$3ybz7m{g?jwOMbs5l6;$`HmK69GiFSsZl)PWW zKQQA%|KA~dSvlPa)qa(*937v&UEIV!C(e>BPB_a;0Un1dSOYHK8FHI62kxIyEPEes zZlw3Z;qTo29dH%;Pppm$FdypC*Ot=n*i&9|@m9RJzJY}d6Y>E8S7K`*6HqAa^9NTL z`vR~{$ql7UL%reWoO2Wu&l$oamsU>yXws`zLlX8Y!%VVfsL~cMHW#~W%*7Up$)m_d zDsl7lRs2e-Nqkqf;*EBN?WY#)l6TL1d)JHs(V#;0Dt^17Ua&EAleC?84MHb)yvGw_ zK~Lwy3fGl=5qizcdciV3qz0O)S62y;?y7@S z`yQ0_j{YPDwm_2H*aai!h)lv$bb7)oC4mGB59R^^ieW;*%Gi2nFzk|N$;>1ZSlUvlM%I@i$yO;?7_~Y1h?UuG z=03p6gtKwbQuZcRPs>P$!#xL1KG2o-F zrxGz&Q#iggtU^%_L0GI+f^cmc@IVzzI3?c1Pnz5gib8czmaPw1IQdE(L|?0-QzSRY z2L2<@`Xev}PGjNO>2vEDc9-PQk&WL`SiRSrvA_Hc$@JoqgmN21HIaqm zKa+L&>rgMdfX)R$F&(B1EkRc70fK^RY(Jha*9VwamYPVz^o-DAYT(~LqkUk<#88E{ z#{1v@lxM}sZw~hyBJ*CFa0y6s52~PCNik^@SxzPPkbXsGtV!4gq54u~+SKm&8YtSY zA@xC?M9a`$dk5{x&<}g??2qPMxEEwa#ZOMYa++klG^p5PgNh=G0X3q{P|45lS8S%2 zF>^$@>Uq3Ix=npH^oIPfa896wa-)p5%g>mAk+CGvO>w_+7nDlo3eLz{rWm5rB#Q#; zgK9%hLHf{x7;+FSHh##{51vEIUvJ!AW$%!@B(5uqtI`3P;a1s6a7SCzSCt!~%w&Hg z>dHKhNSdZscf>Tyv8JV9LXoIQlpkwpy|?rldJrDg33d`Co`Z*VH$el>wVyFVflmgyua&yvcQVJp??Z8BsLm!ehkGYL) zV%wm0y9>mc+g~sAWTmZ>P_fz+&mrdKqQBR=vKRK^m@La)xJ`JJIl!N<>`BOuh3<~Y z>!z*OVu>2^>KL{6u~Ub9u;&b9VHZ(FOE5H>h$yZ$4c> z%JYVG`P%y*{M%>Wv-&AVXGDLKoEk}P+br;H6myv(7pTOm%7Ty0=bJ4`i8d)4Aq93Nu9KdP%$F(P1OOve$aXRIf8s5(q2$?GtT@}Oek-XM4>#i> zHhN}GDkr+Dq?T8&xi|)57lNFyCi!*#Zm1QS-7Lo%XzVPnCLK`Gh&%?jfJ08N-s5jn z6sng69>dSj!ZSOcUazX+qw-+~jfMh6*iqr(IV*$3OW) zktyk`!^%>%!keMolWHcR!!`+O@he8PR^k_96PG;a=i*_jlxf+DF@;?{^_8Fq*x z14)qx+c)>V-Yg$rQhtufXaBogxZ?1ucHwt8^C(bWW`z9IaAPo_=i&@@DDneNu5)&=AY0+Hp= zodEj7ysu>0_jz$aju%@P7JeSAOWLTxBJg&C#g4u4^^>+uXo%3I@NTQhr75wgq0r(? z+R5;hJ#qzyI|1;#u~PPb@~SHe-(GB1SQNfZ!3|LVe3!m6Y7Q?y^bY7x_3?K~PX!x- z5hx3_S=y74`z9Sy^u%wKBA3QiQLezPc7km>xIzRi1>0EeJRf`c)j9T^7+%|Q%+exq zOl_21Q(@PQUcE!IT#^F~zDo9x)uw0KlUwhZRNKHeyFVx+1aP2frlC4a;^O^MXyjjr0!1 ziPsH&_Y}P(OP(dK5|#jGhNs=tXTRtUqv+X3J$gnm{nESk{5f80Mpzhe3TTXmE)8g$ z)}**2=vEjbhh=4WW?mea<7b{3juYH0#q5SWrPLZRQ~cxY7Lw}4#cJDZLU}n9vyLKb zsl+C6IbABLq*n2_1+LYo#JA_p@cko|JdO{ zi~|UByUwqoSB~lw*U%ftmDsDwBGB{cRwAPk zJ`NrO{fo>s%U&T2a<@^oS+jEY(L>{|BMLYW&*r6`#0g^0~PjRIj>!CH(U z;x)1^`XkzyFqdqGm}`r6)xTZ}LQPovtwUs?7q4NPZPu_=6tj{dX;k7G*{zroQHiKLK0h`; zHb=4r_C8=dPml_AuOMm6aJ#S^fZ-Qp^k2Yr5#N@HC<`_+65s-KdFN6q^q!P?pPO;m4qQzhJ^O zm9ACU1?}b-#{DGbVl+^G)xr`w^E@+WbQx2|%#XcF){dVEi**VA0^HTd*E87qf&4qV zEdi$>Jz58`84E4HD@KU}u(&Ucc3|Q1ksVlVS^xQ-D@4a$9Gzn!I=+{1_>J1Yc4eP` z72Obj8p_}x?jOEUx?j8}supUqI_Xd8Wugag*A_Zunn?&zd>Vy_@wo$|&njHOhXZrw zJcJw=+=W}_Zd)b%=L+MV3~Cah>s$pQ!zN+%xIEec}mz z-AWXVtl~lNw}po_IH*>=JhXaTUw{txex1s5vX!Ii1RDr2sZT4IP{pqZ$doL?wY;m!4E`c$K)~w;KfJD2H^kRP;4}678qf&G z``Tz65C2=z1bK9pq(xrH!*$bZ<(1htO+!?NQeRE~`FpNZPkC_-3`_MChM%!a$|7KR zl@#)Hs11WIPaAzJ5DCYTxc!$=D~(Nr* zTXmA5tZUq(fJY}tzmM1EH3+0F2el-mP|Q+_BvXl3q`Ri3krJSK({%|`c%Uo(^#43R zvri_1+Y~=}%;N`~R-0GCOIA18mpX9qk-gYvuoOn1y})?wTu~ns)SVbJ-_zM6mKk>d zKP*pr_VlK{zTd@*dzT+wg3Vn5B=xHUQ-zSkOAXtXbR7$E##FXl&A{m z&tz56=y{wCZIM^29)p94C6{>LqLKQEwIO;h7XZY`x==q=N%HIq z?7TKL%~CRV&wq|6BXq9lKoAC4swd>pi?xfPU?1u?qOL1p=^4f~heNX|fWU=m=KLrp zT)gtPAD7nqTHTPwFVy^<%=hB0Hc)*HazipGCY>TF7CAEHiOi9}uB1HVnhHw!6%`>6 ztt}F@411@w%+hX~4nIOUynCsF9iZuNSxtB*ZUx z#g5GJ3x@E>Wp1`5d}QoA`zkpu8yzpMPhhF3Ln)~&*-^4`)D`I+R4iKQjn4tT zgbSo+QrGxKx4^-&B{?bJW5=PaBu`wLz)DSVCm_g55Chl zYNT}y(wg4akvZJfATJKT=Gd%3=@gSfk)>23#>^Iha&lcHq?Oy1^QIkAUlc#!Ik_S{ zeUsde9K!_nab11lgCiq*XIL@ujfU@)kp0{+;l;M&ybUIfQOseA9HbJlgSjvaZ3Jdr z%A$^|Q=%;vqbB0C@Zgjhx>xoIL{eL$A42xW1lgL0(t$g&czB0i{mHAXaxB0*E$oF% z6Ea0%-{RxA98fx1tbEb^bEXa4?KBV_gVT^k#-WqW!fKrV!at0$!pSr}<}%6l;@wZB z4M4V13^X$5K{E;eR!nwi`HU*MdfaVD6SipoUPYgfmhmhf`+bVx5(?-JoH?z{q31{S zkooc>8odToeWZ`VHj&iOUGB@v*w2Tn^KU<%yFPPrT&UOXUyrhnp>T0rym(W@5^=Gp zS(a<`>fK|P@N`)7oh0fMABoM7BiWbJ_Q1LMa~ws^jqmJJUwD1-B>S2tuifIXR59Vc zc5B4;xFS)ZRF@u56^iQamYS5|4XeQ5BX6#nytweQlNH z&^b}MdgzXS=zn624v8rr`X`MUW;5Y9f}F&G^N$IB{LW^7tG`kC&R0*7rC#iB6x#S3 z*%Y&ef^TMGo+d@ouB_u<vpZ_hLeFC)E zd$)e_&lN4}fr4sCbbmzRe)n#BUN-D3+3wq5?mNj`4aXA(H}1v*(8zwSD{z z{H*Xh^jwn1FOTrF(cx%no^D^C%&2D$e)+A)-`MjaaM_%Aab$#rMIlL491m@O>fV^E zG3D`hlx@QE)3d@;Lvsa2b*XBNtVwRv;+ZZwO=ATUTS4k-?)1Vv+{TB@Y;RGspkNzgD_Fd;V< z^K_t(4h)c%e|V|McayXy{D7bw$h!7Lq(m=-bX|)wldm)KGnEfOhq9g52^5leD@NbC z;|qD~Rq1@Fenrv-^lIxOKB1B4TCd)#c^J4YPCxplLTAt<#TUkv@v5Si1@+OJr`Kxo z`B}0huNtQ3(c2iyc)CpGBEM4AML!+ZZ{ziR;LmO!)1lFI#AJ^##MMQ1(Cum?kk6MZ zdVy`~vNT0v8BDLv^)2H;xow`t$b1eftlxk8uf6|ah2P{kritVtx6rK@?}r}Pc*>m= z^ASZl&{HmyJ~Fe5)J3cubrFc2W+AsoUBtkI&N@Zs3J_v>&}-O1>p>i*iarxNXH1_` zha2+EP`(5TAnR2p_%-1-pt!!xOeNGQPT?(Q@-!Km4tmzEpuT`K!N7>f#tvmM2`HUe z7jc8vEpHPhi54fEAU{mb!XZj(XFn3QEMRy+AnY+Dj(~ImKCzkvi zMn}y7u4Z|4WRJ*_=`D{ptHv!El|z4|*=6PvZKP49(^k+bjqS{r9wt;D`FHjxr;tCbhK(dZONs5y}ivrUGcZ18J zUFQ(ER7Pz>VDt3$$!;;6F4#xb<;NYCGFG(sw0+a- zv*Np6?AEZfwq}HO(xs{*(Ld&hy7`^-@`UVZk9{l-NN#Kajm2|GQ4NB<;EL#~qMPWX zs3K`KL4RV&tMkX`ngj92`oLVjPvmtGdo^ysz%%pZ22eaRraMlsFPevDOA`I87Uu)q znctJ-ktE;dtGb3_GAXhG<&XRLr9Axq+dzWK>?*Du|0%G|op`g8o;9C$(eG2~4dR&> zJLF5?_O^!{!Xqcy-}Vswp5FUST%!I}k-Zv-*CrTQ^gXI0+k`j3^e>3a1GO?>3d@7t zPB--aluQ5!Z`Lde#5&sY_?3ZJ+&@g%<%U~aEj$9#X9nel+wH#~^)KT>V*+wyNC}j} zru85>C+m`D$#j!SV^P&9#JmniS>Y(@cuw zq2vLYU*p`=18*0Cbae$ z#q?058&uUbSS;2l?}JT30gY0J>&M+5a}thQ)VF<$UhDJEm6#jA@zNo(mK=h4s)o3M z4rMivN~cqvEXt!ZLfh4w>0%jtsL>t+trHo8h~g1;%>_m~(AO?iB7?5)5iA^0YnHF#r3ObYK-U;%B4Gg}?^Oy^Kj z?8>>9h9R}<7bDfbvSRP8Z4G0{M_!x&Pqz6|^ij-xirl3V(GF#W=g8fY%bR4!I# z!*>Wr)1}AnmTaD|HKL8aro#KGqW8Z3*r!cs@y{1TmV?8HDhBXX%LC>r{O?$=>Y&qr z=lHNMd?}Jf&DtA|EAX|qQu*k@&lSzleh#C}L-}`qo$#&0p;qt}KmNfkk~@-A*?8C` z6a&eoA}Z0M#|#GOt=;?zahjl0o&zTDgW1h*-kZ_KUkv6bQ+eVmIy7M`qj8==^uvEw z5Ok)hyQCSR@EUO)f@Ve12RtgbfS|cS_NmwdevP~AmCNgm-2m*%&rmbtwaf?qo&RqO zTu~zT;xN!I8{}`H7!di~h@1{xG#Z}O6L(L<{JTMd!brI2&*zl_4|FZ>{qC>LdF#%M z-|5whM$RQkqN~aiM6b$)BGEL`r^GvxM|!}~8UUDUh#wfm{Xc_#&%5^(t96N7F!B53 z;85vRF9t-r4Is`?%!d^4`6kYlC57}raRmx5wrH!r(FEW)t?gH|kv*hCV@Q~pN0)(X zR!^$<>s3b-`M|zL(>KY_s``CW{jLZXLa)KnS5uXX6E;gtq*Rd-*`!ET zW+(LfTm~vmyxQyE8`C|0F`Q|p7f$Z?!KXWg2T3xyrfiU%@L%R{)apeW{5$CqaSy00 z&5wev2WF>)8`fNd4<{#%6UNvLX2nOpTVM|?uigE#fCwcDn&sv3RneX5Gq7Mj_Q8N~ zUjVWJ9RM{1lwnBrd+dV_4j!ZP-^bV0$V|b&By|B)hVg6;C>iKLdI{W;$3E?CAMS{m#tSnfuMO(wXUzGq2o(1pl&AldBT_meJFk~a>L%4>mi-RhULs`b8oa|V=J@3Me_(8CZ|z`aDw zuRe`{#HM$%u@$D5?C>dz_%wnIj3y97tU?Z!K{VE?8?6@p&Ch7D0y6%YNOGqTXoeou zVqZxykTNZyqSm=IO)C(62-P-5kqj&j$F|s3!&VP-M+Q}j49R|9^m7frH5WZ zsy5D#R@6Yvg^vC-1QMwtO}p&8^axOA8aTS`>Koa2epr2vAA2*q`QfZr^sDM?pMA#& zoRzRN!`&Jkse%X;5t zDEDhu?uWV>OhlfXcFl9W?A zuJyiKC2isOGl_>|?3y?775qy$@+x5leU{fOZ<4iz$9jJdm>AZjeDF?iRII6p7DbD8 z1-7V~QFbw%Df5FKMqO7dvtjJxk2{OHkE^b4Yrt$sHJ zcSu|K+85Gbsk-@%%7{NUzl_?C{rY9>Z<)@=T) z!jIWA6Y;Wj#?rU`y(o+_Bp6DDd<^cMY8ImW+;#=zJPv>cT?hE<1N?p(PwVIyQN9Qy z(%oA9HwRVK5wxQBSB@uBG!KJqzLCDmn>FRFY&ZXwYHt zbSI8~e%?wE@>OI?2IgeV$Y!}Ae~1OEkYt6kWbu}Ve~Q)7yi&i*V&l5AD(&JgNi2ug zodT0}C!JzaDUw1(wK2tF?6iXE)Ip(f>{R67g~Ls=aMmx+>w^?r?9^sC7Pz)Swfw}a zKsL-ctC_R)W>#y@q&ZXn;9|b^STNel1T9cr2h|UU1N-O}`SO?J`4jU=n|%`2eDG$E zOvp=iA3SF_$dlf1iJWJ&7#Yg{%qKfIY%w6rJ1XSUL@|3QQit6OwKVo;cED;eBwpM{ z1B-Wyd!a1lTgVfDi$ETPZmtRPCC9`!pz&rU1yTTlY^dt_7(#sAvTj8&o$EaSQaFRN zV6Pw4ByEO%8Kf1co0qYmkV@s@=$A-ykRham>TSu&^qH%Cwz}c?x;k+;8R7(zcg>zV&m0vx1^0&Xy$5qJLPA+5!sQSel4CVdO78dQ!f?0T<~UP#QP=R zx%y(?qT7+N-@YD^L#6z%J0j)hOCt;3d_OO8-<#SWd>UE5=<+-9RNL2GUWNWd`Sn;&|I<-1EPrEGAdQ8hoi$VGRYEyi3Y6{Jpc@AD`WNym1+qReSu< zd;PjM>0;NY$>A+hj7k0RAjRCI$aN~JoF5Nm zrri*k#X??)zv6D&u!pYn#Q5DM5cu2bromd8%wQ~fxl0#9==Us0*J@R3)tz(nMZ?&Z zhvK#Evi`8WvreOk6H)|dXZI*NXpA7Y%aEm7U%`QHED{AC};Jh`tOJUUzxLshe!r)mYTy9^ASb%Q&C83qk~4+Y<<6ZYWN}M;Jkc^ zj@~i@w%FLpo~On}65Imo0!HS<3E(t#?MBay6Cm}Z7Dz|Wxn$6Bl1>jzN!URi_Cp+m zl(z;|)iH~X8p--J`T&%EV9p2$9W!~F6y6p zq-Bf|4De$a%`p%Z%mwPl20_jjdt|&(e*q&pjx=23LJI5oxBh{g%r_IAj7TZ`{MUaa z(HxGfXiermfns7QvYLv*%9eIn#jI2r@>Cq0hgQUqb7}F@wPunnKJ35?tQMn5ML&LJ zo4KFzB`Jfk;ML#)fXotubFnm2A8<_w*a&11!0jOY%vVR(dfBo%*gm7y0O^ax!}dSJ z`pH=%+g|vU(OOKt|KnMto5R-P1CxOMEsE)<$fs1)Uptk1B6{`FkE8xWO7u1IC8@H5 zqAn%YLmc)0T!~T}T}q_d12m*gy#SeJ$S@8oIu*l;V}6>ZMFT)sjn&_nf7S)$sGDgd zJ<4~{li@&|9$0RWha&kIO_lVT5=ER4Vg~XYlylPPe+($nAR4bhuD*6)t2DPg4K$xW)a)Fps#PSv}4H#XiaY4xiOcE{+ z&SWZpDRzQ<8#eeb7>tQP2V2w8_m1S(?-`e@%+K~uA=k$u3gPeudacPaHcToI&o`wns>AKil~a(z(2t2Qe?SSPaoi)2-gA&z~y;u;d#7$q{!tKXi+tQ zbvsIS*h?Cfy#;%c<6|ESXBjMG)!elEedmwnjo43BL8pc0Z2YQ27*F|;`sO!)rZUMJ zd1Oi@F|tjh+-pdAV(Ja&7}2H)%d*4Nvcrpguwcz7mhp1Wx1z+~Xk5zrN@IwQU7(M{ zIm#vzQ?!#}poyTAit1Bh1LdG+gRENyJpcZVRGyxVvtb?S39NR{HWfWX?h!VhHCn%*xaIqEP0w?72<;u-7V-bpTuIM3leS zgoX*I8@G&xF&vZ0cVGN^$aoY~IlSqzprBd_9VqFjCzq|RbnjB3G#v`kO>A)3U<|B; zkmD?eg^}ZnwCl~+70#Ls3u{Q#^Y>?;7j+1?^H)YjFQ}TILm#5+1WSBdL)yYGN{gp1 zaZV2YV8Q2}jZg@>Dr}XHeLV&XA1qt<6kQ36L)rBLE2}|~-}a1sO|`&r)jX70M?tmv(AYhI3dZ;Oo)xJ(kPgDnRW%kn1iuG$mBo&?Z^T1 zS~|{}5DV3DmYv;MFZfy8Fs0C&IVUE2iY~D^%BA zYi4|;UOA;&h?=}ZN~Ff$^J`OBlAtrQx&qeurWFe zMYM_hA^uuHlJ|hJUa^#E;KfO5#hvs?k`{~wG^fSb${H;?AZ-#%m_1?-1Dj@*^+T3T zknh^nXWWhE=KBrlr^tE^FF~ay24pM6Ky)M>8(LA|yl$T6AIBwH)$JDwWSINyQpWM3 z=VpRl=72CsUG3khN>X2+b6u_rz%Ex!Chw9G*%|EUl#Dek>=E(6HjFiz{mxlN#8?5g zPiDR8^MVm2@707ikV1Bg5{I)p&;>B61Ff23swh%PMJ?fdq~0nmf%5!Lk4!-q589z! z8w5H!6G-O^MA%zj6jH@>fnG@yzYEgE+Z81OdrNGyQEZxP88u_ukWPWbT05(ILbdgJGx3?&n3r zPBlThX72JDcG^eZ{pzq&m9#Nn*y-ZTQvR^hdgqn?U5X6fW?>GU;LvJ*!wlIt@AjHr zi{r2|GSupTNP#&s5ND$;Ste#Br zR6P2mW6!Wc#pIX&qapS=|KQST%Y^S0Il-8%N^xC$7>v=91yx=tuES2yQ3StnwdoMvQZ#z5 z5cDavs#D?~MYpg83{io!OW8N4f3BTzXf4W2SC+9}*%x&|+co>2_I@HHNbHkC$zr<< zDh1kQ{~9(im>26N*}fBK&1mdQ$#;LZ^Y+uez$6Q9OwUsd`|Dd%vFx@{u|%{)l;k~O zd)$ZfVjU_bD)~3+eX(lkxY?GUYc3DVS+~Q2ZtdW=+CZMME&TouZ+*M+)mtxa`qsH` zG&}dv2AyB6ca0!#VWVK@f-5hyg=g|QggY1H1?JGrVuOfn2l>nU|Hxj5Y(h0btHO%K zTJJ_d_QFkeE@^zbDD;d=d`Vn}F`$ zsO@kL#ek+&1{H-3?dRUQ6;bugYZ3Y`vd~r)_A>X;@9OA{vTJj&t!%^znd2b*?^iPe z>+MmSI)g?oIy?Iu=96was?0~S@9Y1~2n;pX17KyFIIDC(ET{(*h_*1v%DsM#5yvHY zyd3B>jN>6!8K`+eE1ZdA-Uuj}hz7U;mng?qQq?k*X8kfNt!(M;{wDuCj3c{T3)U?uEm{u+Dle8kr?l@q1n$uFQQ`Ev<%;9 z(G|}@&+Opy5Z%8ZyfV9sx1TQHZIQLgwX>6e6SZSOo4p!!9{<7H0*A+sXrHT=;V?!p zpQ2{}r2tp``L?5M&2gZ=ME`zCz!aM1&r*24=YEhWc-(#3tF*pS2$dHU_qBlpEblHfl}V; zdHJF(?#a)L$A7T$bAI$TYY_41Q;t7#{=LtC_O!1iogTH~6VTsxz^|X26R(_G8TCT$Het|qIJ z`%6d>yVZold*MSSt4R&T?54<0Dr#s!68~gK2AxZvTQu@Lt?CdlG;`^rwC}HnM!$|Z zT2%@U`?~h16CkScz7F(-E-6t2xZh(9?1Yk)Sdkq)w?$a)g_|VI3gd3MjnUuQLD$N= zWlO?8p0OJm#rkM_hzr?-wnyM+eq#J^{69B7S!doj%GvOng;vm70mejtohhh?PJZU> zw^E)v;HO_&u z&WmaU(JrG0wX;F8ClSkXc^ z7ZmKb%RUvaQSP4J@1fc3e9T=(pXTL;Y@e7WbX)N8a26eR#10?p-+%2trW;qL*9+eM zHQ78?V=;$!GqBu@+RYSF3}~z6Qc;T+4=8HA+rsN*7)3+XneFp4B^RU}?$rSm0mWim zP&WC+Ehq&}jKt6k=r3*yucCWk^IYR{UYaj~G%$Uae-0uqo1MGE9OUoX0wL4I#W66l z3?X&C@?&#z@>B)cEJ#3idBphQ7AIL*=6e?F*c3wx2Kc2O6i)7|n?~c2yloXXog0cy|okI-?9t5ygNgiH?dg z2n?k`Nt1!|NQ3UGzD2#8E)KbEXe^5sA-NOww~shsm_CC(E?G&26)5=F2dY6S&&2?! z*6=#q_WL>nS~eib)X=~yM_}p4pZ>?5@zl(6cw=OtX4W7^rBxv|VnlgTC|qfng^UNN z7+vk3FR2ZR)swbsRj5lp;)H>H=owOB&yb#C4rvb{0pOJ5rlY%M*X4;Mh8hq?Q=0uW zlF?zg0^vLjz~U6Tp7h~2r9T%M zK@)i@2}W+V}m75%GMa*gmWpQVu(1(1o(3MdkF~z`rl~LG;fyL1{IfN?B}ns#fpx8uW#- z>B@j2#?ZN12?`IH!G%=4a6fFaviPe&4G~8!q_(>00#G)jHfXJ9jA+QUkjj(pl*RFu z>O;V3%Klk z>mcR#(;{UqrR+bM@^BWU9CYQ?c(70#CkWRHS1 zD_oWe1{CPYI2)$zNKggRc zZWeELE)kW87GprK?%7zkILR5h(X)!mp{l4I(gY~Jtp}d!3dJ_x8kanE8B|&pD~`E$ z@sc23yp`PMofGd4%K^&KjluirhofPY=0y;#A9A$|5N!q9a=?tA^f6bXB!cc$j@uaKVbqw+$*QIXE20bp`~-Y;zsB+S>kRZ*pqeWF(A9mrjD zeo^E1>VI$|JY9W6z0Ca#KM^{g!50eE@jnBnrcm(Nc+&-rfQA7!)$B>)!XQVCs4%tf@1PwgH+081vk^xuj^5wJiFC2ZI?JSIb z){)2TW`VQC@#p_@!~cBOXlF91p;yU~DdZEAuk(3|X{SgV6=jH$LvN=N`i9rLr^-;2 z;0`%NvSu2hsWGCGfEdwd3J8+ia86dg&Stx2tOffP3XOT~(0UC#J&qVk~RJ6TyI#pUfoK43qaAd8^}LO;B<{9-Yl8*xH5ObuqO zFtLeSXiOTE^7c88ICaox7=yr(VUT>UVsSs|;@y6sdOC2i2W?oODf9gpLP{80xP^(H3V`ux`0N{Q_z6Wqrhy|IF@h6LgO)W z|Ab5I?B%WBTuS#bu6w`V-WX4|vI`}0czf1hvV50QObJDbsHm)&72+(?1>!*&^dY7X zYPT*aPmwx*sIl-$R>m^Xf&Fy+ix;3b_o6T^cvs-CqAmPbP_=KoXQAr{fx{#_WCK4j ztXbHjZiWoVR*<#Hpv!s9iV`b=eNULWM=sk5V~>nG-V^hBB2JnUH9T5(!{H^%f*35W zH@HTf6W>s323`BfQf7kMl+0$ts);vyW+Gl%HZDI}aiP*YVa3^+U<=8qf|tqN0Y1s&Gc(-9DE-m;0W}JT*v&i`4EHgR@BK^ktqpdiQh?~RDem0?HmVQwh~luIExSjVfGk*vHe zPLiAy*0^Aixz`@zFOY&uHK1to*PswUi)V{x zlYdF@h|`eUkXr}F>M|r5(1UV8nisCA2hU5ZsudTJMtx}?#_3S>H%62Rx1@&^5e>R# zn`Vu`mU_hRBn^SSJW&@9WqU{bj)&Rel(7Pi$-=_s1$&TL2F@6UW>Q|^+W+!3f@%Mv zbyXyLET4wMhGwq`z$z#PNWF`xsB@7Qq<`yF{smW}E99u>fiV0T$J4pw&FZ9k<<0W@ zK5gL_lwHbH?|exauU9YFTgLk!@HV+1jiEY~HL!}U6W#O7b1&m{km%`W7H0V3I0rzb z)6VOd{h-7iE{}V{GL(8;zj{k$giZLN|65J6*oEde>?GHj0IHN?KnSUTipo*9%T_7V zKqDICyoMeFi1g;Do9Sp#ExpWt#0OrqG0mXuiw&2MIlvp?7F|FW^wnWk%UVIfG!4d$ zb##Xig2|!S(lV|veE@qH{gt+(Ibbhm%+!pJ4BN!-TnNCQ?>tc2&s2K8D_mqzY)O< zX1-{xi^g!JUA8z)wRps7@e)zDa4~Mt7hjJ^|K`nKT#KlWO#I1W&Eng1S)k#RS&1P`0#2o0R;I{!Pa$m85W=Syz*JcX+7V=W?_&7A)}>s7fq0}W>be-=1pTI48Y z)2h;7Hyp~WU3tOI%gg7CPp#^*a<5<$uX_e=qp|=dZQ-|{@1D^i#G=0qt`*Wdq~8_q!4mT9 zg*o&R=YGc(6k7wA$2r1|0uC#y< z7n7i-;@m6m|7D9K_wVC_m7_p4Y0@`V&i(vpUug{G4Y)l%GVPvJGwdHd?f~|KV8{1_W9*0BrGEKY~x6EMZ9#*%J*ojO?elb^8S&1+1p16Z) zNwHf=-Lu!uR;q)5}B|en&+|q(?;!ylZnVisAp#boY8XpO4qDg1jxf znSWZm*Y7$V$E^EmXKcBHubn=7 z)EGg6Havce^HI;6u5~jXF^*Qj?r~0!ovxV!dgS$ppFX)HhQq#!g_4-^S1w4G3+e>R zrfe7Gh~m9J_gp^Zkx4QCFdXsPnu&MB-+j`UmRGERe|}qM&cw-Pb>eUU&w_9{eq&RG zT}u3RYvSEn+)#pgU~1XC!snVpG`KH9juB0jYJJe=7w@^pGEE9Ray*`_0nZqQ;Yo8p z?(2CW_ote>v}_>4UDgW_`5e7i#Z$~$imaldx|D-+hg^}BNOR2{#RTjgo3(CItU-iz zBiZ+^ijeX{=Dq!#4W3x&*6)|=M6I5Qp#~vm9UUXuIlnnr(=R_J#=4Zm&;bxZ$`_rS zcFhxIptU64TayiBnmthV5<@QY2ZXEVA?0SH0LMucc6toX)e8%RrU76YUDBUPU<#yr z{Mx5h%!1(b11ae`a?1rq4(Op11jx$@EeBaMF$vH?hLj!jSx8$ZgvI&86WeCo74}P$ zgEvnFf)e#zx+$~-Drn;8uX+-dv^A(1!;aYD)ZJS!u>w5ryaJ^#fR4AlU z?2@K^C(Z%g9glqMR!_Uhdh*CHc0a++bX>Xpk9S;*0hqsNPW}&B$u0oHVV^eF#HZaz zG07BJhpF@oy25J!q+4S|T6LLk0)g(dYIQ%U^c+?shPA4?fk0?N79Zo(vjURc;bM8f zxU?nD4;u+e&b=|^6>`^!vw3Z@$z-zEP|Qk-ETf`|e(`B!cI4K#dcQuAGiv&oK4E@M zZ=S$ctPI7Zp`Xuu-h4HAw2O?x=DtmKcP)(~^ynVqk&t1~R4#w0IfMc?*n7(i6gMG^7p(ycS_^(CY1{)Y1-KaRM%5;AB-2}mabc0(n zT`Rmnr!MM_(A4rVBZC)rE?6swp9X^Swlse|VQSd`<}o(o34ewCWp;r1>c;Q?_$?z~ zmZitcBo{agnENJx`IKUSA@2$m)gep{-Z1Tid?&0H*g2B!k*Yi)ABHZHo1>1QLfZ9& zd>_HHc+?vH*G}aLsBQh@QTTBpydKt`22~?JGq3=7D~DXOBzD z&uF2Ceu7ssHd(+iX4G+~EMw)Tp9RL6ySUMwBlqq*215BmiP1nGiDhS9E`Ib@DBni~p^A##La1MZl)-9w5uVG+DlE6Tobs znDrD%prW#Ya;I*d26smp^bmjbJp5S{qQOFxtip_`mg|lYwH!;R|mcNH%>_WNgyZKg&rpz@Ojkk`^Dc8F|{^Z+4y*}}qwS%FIy6o%qGeKcM@Ctcz` zq-+f@3@zu~4qeJy?m7Tli=Oa4MUDR@g7;nZ=<_|{nJ|Box;1=HSbE?RYOk;ro??X|?s{{R8V?`W!2)PWnHeoD+P-xa72d^GYtMVCScDH-SZtnXH_R6a$RcwWv>V8g$~b zK|?-Yk{~DwK}r-uIm!XL7PNb>&soY8hiG8GORo^1P*Is5o>+|DEe1+H00L~(!$R3{ ztxDg@gHBqi?4T$KDpfFiJRr)!jc1_G_DPmi_^wG zjHk4o!xqUxX+5$J#)$4Kq4NewooZb4k~I1bopH^uE#GKsQDMv2<9SB(gR~lRojlIE zE*6yUHjzuu#)yiZgLy6xZDF=g)lL6fp==WYm%BggFd5*(VUvF$U9Q0Xu5MwEq63mO z{j~lG(Ee3Gah&_=dHRRf2bD$?%2xbP)4*$&Z3=If6+JiN^pWB$h)HJ#W4sPELG0nW zTMLLMEiY!*$Abm^v+UfQKhC><+S|B36^{IUC=3-Lqj8EPUjob5S%TVBhGGWIA*N5SKG_SkDhJpQge)7(n=MAhmyQAc3(NjjKpVkw`+MGpNX%1~Aj~v@#5#(Jr^s3= zDj&AHpijrE_PHUuJnt?&7<$(cRI=7gigi*`CcWlQUz1 zg1Z)w+%W;gRf_4MNEg&PxqL9MOL2t1gPfNPN=#oRE7y@t%%G=M_27D2_$gjLT`kR$ z#DXKz7Jk#UpI$52z&|P}r~3s}^Y=#N@&`iO!uLZgrXx5zq@6h`DFQ7!)bGW8@rG$T z+|Mk$C*AD#7tQApeX#S!%BF5%aY$is4YNanvNubo-43mBNe(ttJZV+ued=lK0BG>b z;&%&M;IC%Q99E=x<9l1nZ1y|LOL?x^cMN8w2hgw%Y-3v(dz^jnKFohx23zuHkqgb4 zvN^kVvcP@a$*YnM(Orrm{yGqoZ40jn>VfE6k$WbzAvf@9T*_g4q*YZ(8w2olhj4L? zOR>0(EY6{U&VN<`F9)hLH~QuK8PvirN-=`5~4j%6eQ_8o@E= zS>a+*YyN9L3NbERV&{mhq+|*?VzL(0Q%ns-c2iMx{$)H2sDG{);3tMQL-&N9i+;OX zsvD}$cFR5p%#>tG){9OlZbJ}NgENG-{Vt6GF^s>jiM&IGo$f{4fxKiKFE==rK^mNE z?p;dE`arHSTGR)$1lUA=iRAIB{j2@|f5_v1q%`A1DDH+qzrDF{;nN<^|sWjb%oh65VROll4jt zsKeX}VS^z$dtrlmg{My4>7k=hUZ7r)RR638uSwmOmE zAJL10#7R!eTI7YiI)CIF#2h9Dhm)T%>VP!xYWbkOM;pq>Ec_h9PjUpDtOwqMJ0Hax zJ2>tA>+}oenM(^Pzk1kq?BwqWJG1bJI+wqEZn2^S2xA{Thf1hz;f-X6&!*7qg%9pF zoc+YFm#lrKEgZz3)TvXC&2JV)i#CP!N>hOxf0KHHs}43R{et^oQ=rt3dGKso_~m(J z`pwW7^YG{=VCjq_XUx*ywB|gk8RPDzxc$4=B<7{yob{P4lzr!@b3Bq=mMqAeQ91vJ zRA15C;I+xWl($-D57){XG(5b=YP>#tWI~Qu;dS!DdC9+i+UW!hq245KblkP7!jN9l zA&euvAw6_A=&>flQiaW#b(b}+L zmvaLzE~swa=HPB&=bUEwpezma5U~EJote0KU^595^v%@I9ji&;uE%l1?dQiOMs#Ee zH(Vld>=KdO{g;tK6Od~u2FPeOVl(wE#SvJ7(t$<7Bvfd z<;z?*-p!_VXL->)Rmd;Zj^D7h#glbiv zh4n&8smXtFwRl(H7k4ieJiH%z4#eo(817t*Y}!Rr%hikbgNkacu$i>VZz^M%)nU!D zdO9u$FKJbH)A-Df{Za7+RK}QV>rb&;J!5xuHBd$ybl>g%5eVtC8i2czY_i$J2E|d# z8j7rh-sw4)<``5yMx2^L3nRvd0_*3+?9L~{$LxU#dO2s_;(taO0WtUe;~k`i!@He! zlPNw-F;LvQA6bA7t8O^gL&pg4praP1Ub3TEQ7*pknFDn(*!E!%5YFQ@3-7pI_if=} zS_`+x-M%Quo9kSvz@-W$;IRXxR?pPmA}scou!j{Z`!M<4KYjg-yAe;{Z%98y z*0XbRIBZ}*Nq$su?^cR|Vytv3s$y0-T@$2P>Uss-5!6zx<>N7cCR<&u!1jSeqQClp zlxp;E8*svA25rX_o4>;T-oH(W{N*$wO1c-$xj@p`p@hSGr79DY6j01IisVpH2k3T4 zzpvy0?J3arXG4xX-7mwtRvho$7JfeL%4}?1O7|O_wO3fLZgSVrxAjiY9={^e2Wh6; zV4Ltc)HHJ(WNb%^-XO7uft63Z-t_KlKOmQ%;c*B#6RU zPY@V7BM12ULSEF+#M%wLgnfBvva(@e9Ma|*b?+bIasM0L~k^yX>%2s$lD_R{8wnSdRBy8<;G3wAAlEb~5c3B;ua_*pXz4YL`( zL#{mvtoA{dNRte-CIgT{I48yi+EQ_!a=T`T1c8}iJ4Kz=L0lOPU?wndftu|=ytj;+t^0h7rf{Pv{yh46IjjUr$yp*tHWt-|8?4TwNRaRl?;ZkBRs626; zJ?B4;ecrgpR0aHL1u0;+$Z*&@fdZ*fJDxoh106{_sHi+1Rxm}2(EI36p>7D2Heo&K zIdO+jM=>srxg7ZLrrP9xW;88uL>xVu(E2%b;K$98BH5_Et_Wb zsPWMTUMX*%GnP*cxvrjPuR;fw=y`}E%STT{hBY@l|AWS7&9g?FU2QDnfiTP5O>0_t z*h!tN+z-xct@v7CTCirL?Ajb?mZml6m&AyMTu+fwKBkK%!fe9II!F+hwgxu*6mZZ>I>dG)tXhxG^LJCR&M2Y zL!;5V_^6IYgdJw)y(i$An-MlXEQhzk7Rbr4 zW)gR?;D~ex5g7GQ)^BHQJZ5?xUa|^64{x8aBl43!{%@GkaZS-FGD+o^#&K;maa{LN z%!d@Ir=oU{Sf&H|Oo6vMx<8zXfCpb5|i;=DtycoO_t{ET1YRj3Zz=~J$Il)R6;`=hH`OlTqn?65?@xXv`;;^l}KY{M{Gsc=;sc9<<9^37_syF z?TztdD?99PI6DZmm!l#o+^0QL}PY+yS-8a4MRW{4u6QBea>=JwF# z=%XvXiDL13T*uw}`gt206uB=D(jDBneRH z-s^G7yHt5eT>)W0sDg>(trQ)H@*7~Yh8RfnRTfDIYy>Cu08FAX;-KRf#$4>ZTW!0>?`BrvC!&D>w+Mhr5 z9!Yx2Xk^MvY*RMH07a%2tGsn|vErEfW*OFcr#yE6`YhYRi^Y3U&bnW#wK0Bpdgri$Zou+noh*%Re@!-X7*)_5FvU*28xub(IUt6QE)j;F()Zen>@tdJUCXNSQQXUT@J7}8P zgYHRV8N)NJs#9WAoWTE^N8bZ zoMpW+8b7wiFgTApWhKH`VQbP4Cx4~UJfq1b(#&B$+(Q1-kl5bhjiElQ%PaFu0AqCB zce$&kTHQ}jxj+YL?JeG0pri`dD2>DYBAi)U;AABt#&uh)>n}2;M&qOS=S!_5p2Nlm z2%JY5pA3qDHj_cS;1c6p{ zl~?Qla=Wo6!G<{>YZ$wqV`W09pKo;YFsEtdY*B;-O4c|@ow!(BEA9k=k~Cm}N6+Sf zv`H|I_`j{QVkIbS9p2{WSb<{l4fXr~ZN9(eEGR7OrK=;3OAb%n>60KRpWUU%_o;Ci zo^{^iHaX1$>QnyCz*_895QVGD0U*pm&tZ}JO+>tJy(gpr;LO_|)9`LdiL61Oc^a(3b z^0?dBA!z%wce)ft1Z7PB&F@JvhY$Qn7^r4HIR)$R230JR{@4iRD4< zaD^34CV#SJ-&cK&IN85wT@}fm0%BC7r=x;mplPs}iuy<$8;o(RJZYD5Gt?<*y<>yR z6(`AZ*o(d&DGGbv3EeL_s2GGVE8f*IwGmjqP-;d;-*C;FAnhgEK>6K$?WH-AcCWlksZ6;9qh+?3QqLGSfUxa>*R<$*>#x=$}OY(`NHTZ=5 z7;tyAg=4eH9bU0`rGFlOt#h6i${$^uQ{&PWegLTC?}2*39_U(Y3)@ETRP+jJ{VqyV zB-{CWmG?kQsfMYWnKUz6)X2x=)CqZya2QBqFlmXS)wuM^yOg)dVQ9!Lke-m&d5i<7 zY@ugjC7<}hc67e?Qr_tqMlh*kSJjd9DIixp`U_n|G5HkHAzxXVs2e`h#XCKAdTi#W zJQpKajIE+rkkIKdICsS9toW?BJQUps%*|+3SsvGv8a(WStZ^0Hp~_yEL+4BSNe%== z9rV4o#|Lfu$7%UVcHeQ%vI&0sAE{;LxQJuE}V!WEm4 zZec8Yo|gRm&;4IBnv{qtVGn8Gu=8@kWYSwG2E-8#VUP4qe)IycU?8jBs>-43$PMRe zRRxsHK&4%>@_-c6MOc%f$rEJ;=SyxlXYvYpwc-=ZRYekTN#VWSdR}md{DM=DgfXDH z1&s{x-u)ghMw@ELt&VgnR+Db}mNZV%>5%~(^LnXBw5j;KCNUJtit{BgqD$nsKh*oR z&gq}UmR#N@lss|48T$q{C}L+1e{yHvT64lx3%sUz?q$5J>6y|BX`xK(**Y&;bW*$w z%CfcmZEnSiZYU5M$1=|5Uue9S#RS+ey6um!ngH&aZO`@ear34X&W3y~w4q?roB97I z7Njh&YCGa=f07keC%@9v{_+1B7ofNI{~&_gWVb=$Zd`!Gn^bu&nam7R6SjqSKr!4A>E(HS(AQWi z?1M8{)>2ECdVCBYXdHN>I~0AM8~8_NpMse{F7Fzbe6KY;Y;vpd)~W4iQCO^LV|xlF zXKl9&=1;P+6Wm?XrhJn2s`;A6W|Q^wHDc=9dk6S=(ne_)lw|4X8gZAG9RX_&6pOE| zLbk=H9e1I5-@doZ6A7Gkb1Y;yaH-I@`DZ%!60}NC3anjF7T6>^?$tp<;0P5=Q2lT9 zGa2+%btNoS$x1Z&+a<9Ql&uB%I9QN0wLwUrm?6Q{DPPj2D)rc*?&bB-b}gH*3J4}< z>@hFwhznLGZt|~AfBTxqxD35l{>-1qS`P2UA z#)`_Iji6%IXA0oY>{1p4$IF+>vA}!c2qRd-!;TPm+)%9GG3my@f9x|Szq7y^xB8iz z5bi*4WH2;I&^5Pa23A!M@O#60Al`9_e?(m-!}hX^!a-S{=RU6*mp$c;AT zqgP)$Pu7jq#KhrE6!cS$vNKyK2Kv1>Q&DNbC+4>XZGfmB?p-jzcbR{JUh4w1Awezj z%;4ByjNx5#PYkVvZ$cc{qVmve;Le)>H^nwY+lEp$PgzEEaOCeao<@-T^ys7$B$eGB zoWm|G5S)#IPd>%yD3VP@A)8{G>W+b_2BnNwlaFXkQ)rxj27TXi#OV^yLL63gdTi%s zLO{0j8NBxJCu#7#RlPKjmPVdJy7>%-yY_P%N-k-%@{ffDG^_OJ7V?pZr z6tCWYMerGZKU97L@fl7(Zh=tivqA8or(HeSCr<kh#k6sWLv^xBfIEx?a z+!h|aplW_x&^`AYTIbaaeQTeJuL<_Kub6U>^iAvX>YI8q^r+{`DZ@@HrX*6`k@1nY zBRgMCq~ar2PAM0+5j(C-EFdzR`guQ z1T~46&X85JZ~DlDys|v|V>Q73?Gx9{;|Z&A*rm0QSZW0p2IQZD0$&}J;NW4V#Ew?M z>PhA*fc&m)lZlOP>Z-Jhza+67j*k?WIG*VglS+{kLwux5iAj?z*WI(UD(oq0RpHKL z#0hJ$3}<73GU&<+4l24D0|!>k_~`Xvj4>Q>ik->uKIgyng$MDon(ziv$Zi|MVc!IF zkw$G}swoCkNh+zRn=fCGMuTV4WE9ZL2a(uiqG6|I;a%Qjvd?}o>@pmIec|I^UJWTz}qpq+XJt_=#C`H6xO z54_WEmWVajMgwn*V|xK_!GHhqb8}XB&PMGm@VsM$zFk)8fn`ic(1S%x*v77B@jzCO zZ1t5FR=Q|%>Gbgw$g=v#Sy z_J3J3(?ml+>$&G&qnT{YCqL##P`O5 z!bHJQsB3|?<;>vA{5H}-56rxE?=OL_6Y&Vsyw8=Ui?Ry#BJsORPI#aER*ph=VD z-6&Y^+ZK*mXb&DYoGlAjY#%?DJ$_8lG{!5t7|{39q{^p#f4d5Z_qE>J$u?T+eIL5b zhZTuoM@iDmW_hVc=Y$2s9!!VTGyUL!3Hr@yL8P3|?mWO4?TYK&+h<4$JG;W+z}OBG zyP~6*Y>H%I4kvBFiTTAL=(=K1tOL{^kYf}}rnRbrq|_tHyE`P7Ne;XWi;xzIUQf)= z@QR%Z;)-r9d{pYSkEO5#2di=QsHw0tq+QO_|7dziX+%m&qW@{KVXRt24&$N11P{3s z17WmG>_66My<39{1?8dGnWe!NEZo#!WxS65G(@Yy+W2xYwoPS7^i}kD*fb58`B88a zIrSZgh;hIH0AiP!A9E1PW+v#nijDtXt}O0`5j1)Z?Dat%b8s7tBC3sw8n;t^0I9lp z$jqT`t$(~i4j z**c6-Ei~z8ag+aUUgpkb^ULAQjfDdECVv!40A4x2LP*lIg?|uO9+?%{6nbKQx!<{< zCPk(I7)V_QAk=!w@0k0Q*{EH$19b5w0A*G{>QT?bfeXz8kd~{f{LHyd*not?5fTf` zD0hJ`zrrIPb~=|`yFxivO2ZM>xdRYCmUQVq_D`K`vSx!Wggx1 z=4ty#)qJFi>XpK(R0a+UD(Q78Ps`dBjiH)M=UO+ddRT!?&F!+?&}gJp=_SzcZ>|bk z<%6~H23?}iqXhX-c-`P9BS%-cN4v6H!OrKg4O3&l%$idV0At1Pw68V${EyM6mB`)< zBR4o)lNM(}NDw`lxlfTHDhk{6E4)^?_W^OhQHXqetg7@Z3elV;`Jx`+1iTiAmkm7; zIUx1a;@&482<;7NRrQ9P4C*ox;6Q0S%5HWkbEuvOgp^*`7ad$|(7}dKBpE{)hG@{& zteJr+fsYiJ6v&s<2Bq>6Lo?{;1vgzg6{#`tC}9QY^E9Q8m?;1PcM41;OY80fS&@~Zp{ zTEU$W-C*%R@+DchA$W)1`e)9$l=*H1b)wJUG#EN;>*%KM^Lq(~ec(inI?w00R||iy zUz}dlD25F2_qr{WCqJWE=3gG77bpg8ze~!GRV}~}jDHD30{GWff)EY*X3=wxyQlcC z6eW`LUN`-_lsNJLXl30O#0geXSR0-%8gfOhnDWp}C=WI~4UN?(|JTYxIrTUm=KA_* z{hQB$Dy;+_XHDGC!oD&oz=*%U{I#o^Y-1NSB;i((**SVl$RgkSWCUVuv! zmTC6PNe{~nZkN@BEcMYChM2q_6>)k(-k2vA09!1rBEbV0HR6X8a*w1w7eXhsDXJg&k!p|%`qW)xV^^759 zYxv!tEcz!49H+kRd=VqW2v-!=|6> z=vPmk`_Xq#+_B}kA0vo3`L|ny=gh;(Y&KFHUUGmPc~oI$JjJY~$SNwz5Fj*c!`o$L zzCE7*8UQ^RKNAoh90Qi`N1yjJ*DYg%2oC2mEvTHKyOBXJ_eRClByZ$WZkJ&m`(JXy5G^*4aMzJ@rtV7Kx-^0?Ool)eWa*mp3L`{bBKl8~BcBtX7$9miZHBA%)l!A3s z)K17t{jF1ZRNtGrGjNDLD%Z*~gWJM$)P5@n5^EuG- z+To&`T}35{VrJL4WJvPqaRSS3IC=zM;{(>aE-CZfj5vDh99c{fIh=6VZi0*~ih&yS zGz0sRL2iDJdaqxJs6+&$bVL`hOaMy`gwjxAHG^3x!Z>awNNwJwSBi2cWC^lq*2g!Y zRz1zmk^TAB%A_9}F>)v(=sV;XhcR-)1S1_31G&tzRMav71~Y+c3)4TXD$Q-*Vh~%d z24w|EPk^{y>f}Do?cEm=0IFFg}0WUFTt27wwRy> zeFlw*BJ6~I7fbV6$h`5;=<;C0`#DRhszHpsyJbyRdqZaACEze3iyVb#XF z@3zQqlN)qRM6dp@+!Pdx&begGgnhUE%_a!)T^<{Uvk7F6m=Rk~SjHPWkV&%}H=EZ3 zTWBkd4s3zOpOZm4Iz|MoKl^8Q(9O=6jmrqk2z=~EW% z>-QUR+U9nQPH=7)S9<1)AThsDc5O}%Jw#_OY*0rBo}GrY z;41Oa6T=b$qaStXv1OC^_1JICGVeOzY`d(5K7z~qKHtR9yEM}1FY~SU$p}12Qd~5R z1Wz@U16zbffYhPRYdfj)fzPRoi@25~%KaT(A z)faC^W>8h%ycQWxHGdmU=S3!eW9<(UUK)9^{l~o#gNy1TOWqn>6#GWg4>P`YJ>q&q z>$hSf&whL0mCVSiU;F4~7$xu9gNw3$oEN$FhvnZ{z3B3y*thMLZWtHROmSy;WjnY` zp5&EoIrlpwc6R+=H#xaBRx**p9?L3|iod%QbB7|gs3?@D?hxX(r-wFdUo;;CrttRD zB_5gq#Vu%_Jw%E^YFw7`&bi<|3%RqA4hT0k=-L!PPfNXQi6{@si?4Ye6W0agOV+w$ z8Ot)yWu7dCjLu*ofZ_-Od*b1e=ZB|`CAynce^R?482737nxOy3QbT4{e^wwZY z%vyfKEUhXrG|xwaIzt_F+BCh0M5Sk9Xfu5r`Z7nHQm1yQ3T4 zm*zsi8gJ^CCI=@gamci3#WT8iN#2~Tgt2VR_8-mEJ?*;n-Si%HtZ~Pk{K6VHL+xcZ z-2uU5ELXw*qD5yxa|&f=4X?4F1q(du<~ee8!!NqccE`j@*#CuPOpLL@O*-eRe*NDo zZ<)edB0@Wa7HpaSN9s+p9F3H*=hzDTjQya4&whRJm3(tD88&%K4ksHdP|s|mvt0W@ z2li8_@@^5{a9;gv;d4kQgPDXz0iHSZ|JeHyxTex`eUETL@?pqEFgb#PL>Lf97DGk6 zsO@d%ww-P>y|?%F_Ws*xO_%BPw$t8p+L_x9qKGIUD5#(YP!>T%0mVTLiwn4nvWSWz zE=WYjGQ*!i!Rw%6bJT^EtF$omp?4KyM%*oEncgnS<~L!IIRo(lFDHG2+Gcz+#4Wv zgQdBJu9MP^J1&34Zul8}<7h=h_gltw(%nVtUm|y(vvpFO$vWv%is`3FA4;`$(YLuV zBsx-k4hpfm;Qx2$H^@3=J0$5$9ephV^zT}fLksfh)5<2g#|_K42ivi;_KR z!;X^bKpmY1$xw}AgF7@)!D*P@SmYmP}^PDpa*6pGU4bSl;8{kAYIOj-Cb zE}pT?-;aDeDifrZfE~1PAMfi9K+vzxPDW{7Z0YzPUo0!=@W7Xr5uV zsa&>$L4(rlQNuwZqIUj6&#k2JrIsipOIt+^N$)Humtps0JB_BIo&T=-y+8KQ+kBsR zX1iUZo4p%@hlR;JJHdhd@S7Wpi43ykYc1Q2G2$`QUtf3jo=DZO1N+Jrs)nKKn?|Zd z%!Cg+Ve54vP{4wZJV}1>sFAZ4LZdF3l+DMk%CzTwk*xz!UnstW23qyLxjIR!MBON@ z4lJ0nn=YYug{JzdAA1}piC|!A^+EVU{?M`55Ih{xHPK?sXO57yqi$|4>|H+2%5Y2x zPrUx!H;t}rna?>H>2%<_y;zea(?g26Pmvx<_3kaX7Ak)8V1dydl1$VIkvBsw!w#tB zs}SkDu^Je}!@b}m7U9JRGNfH}uTHDl}rR zb3dWP`Z_4|gM^_F034G{%%QWyVX5dpwQEYks5rO_J`w$UA7VU@D-PJc{h3R7c6Z zkdqNxeQ$VH(pLjFgf}W`$pLYzFgLVAb}%S`*CB&Yu-ATZwXzRZIXU!0f>-3bmIX9O zTcS!iUGjX8y|9yKqqkt=wo+)eo@oQee*!BjMg6eE|H|t|tcbjRGleubFjhV`!O9Ja zX`{#$N_Fy;>te7})nJY8h|B0k#qw;oYX08>h89E%6S&>OmSP@y~ydG937Lz2Kd5f~k`&%ZWoAiO1N?aQmAT4i+j z2Z|#=7@Xr59aI)j9?&5#;wEwOL4~VLcvsfI?-It#cX3ku_IX0PH!u&M_N=2%Lfi)9 zHy`__&fPF~Rk*EeoO-*q56&AH}!y zaX2K`DugNOCa6eHx>sy9DxNuxwmpFrDyGVRGcE0(jE3h&cbm47Lk?_sT1*VjX^Mej zt>cucChQ!2bYUsyAop}c7k!dG;*Wa58pXbdqr9Es4h~ofZaW{? zx%kn7!x1|E6QK0NE~s33(@d@F3c)8{nf^MyMo|==3_RxroLbUPhGhG>r5wEH-TC`v zPyFh=DuDK{TB7s5B&-+Wbf-~V2|Q|rC9yctR)fiE>oNL7i(mW91FXPAt&zX=M~M+k z!XDv&lVk@5Q?Ut{vM2`nxig?g1_Y?L%6sRe$t#v1yBetcB}0cxGvpmMF4gZqmOebA zYT>p}=$eTUWQy#+7#mlf3Ni4^A)e?7tRON)oA}=Q&PIqt>E2e86%GuMOcRKtP)rg< z)?o$+5F_MetjfoK8{pJOiyCdp$1bHVO_izB)_-?mFAt8g=V2C=+7 z5tux6Ej!%2cF4)oJilZiG1DP0@Wbl=y71VSz%g*`R1Q07T{lkaQr+Wi z493=8b#x@M#&*mZ^iZRIT&o1M<3{Ex+gg$@gdjVBf8nicdhlY!C+~`l$dJANm1eS@ z9Woralyi>>GBPP<8%5G76)^hp2i>-NW^)r={>G0J><(**LSd%zhyme7ucY7uesy%5 z^HFgFEPV#}XWZ5}PYz7jAw)k5*%S$uZgxpW{GR{JQl`EcT#dyCgKk=Hh~U@owB8?w zv<0iRuE-^xPe10J2v_6gO?jc9onN-3fRp!snWZ=0`myRyZ~W$WZ>hvI<3uQGtoZ)@ zTHP9NqZLsm##fQ;4(zrbHn9tPCbpOMO$UM~v*0OB0=?XO_S&A$!=3aaf6`9ac-P`ULi~ za+IcC`tCO#^Ne_zoBq9%WVHk11ys34c^I20W+O$CDb+E^^4x*mvpZqnQCuY*a%|1- zljkT8`4~Mq?Fs6LvoAt978G+&U;A9#(=)l0ZU;>OfiaAb^ z8cKB=x&t8j)3h{U>Gf}4S<=pLTGYTV@!TQl<8~@rqVT_EOI8SyT@KFENiHn9M1K&3 zDH(*%B?!>=O83v{qVXD?EQxayGShe-9s^2~fEr0l)QP}cC2nCdd8_Zp_iS@%0*M7| zQ5bZa5cF6e>1osQj6D+p7%L=E+kfVgZC)kiusI$Jbx&wZuupM~D|VydAM8oZ=S*mL z$2SDlfnt33p2$_KKr!`e-Rt(77jru-C@d6i-+lF(UXdx^Cq1x;cYn^?ZOg8Fv+FCn zUO)M7XTF-aH1h}DU)ld7C^+yx;CJiwBHmH{@7wt)GmC|IuGY0su-*%|G>UAuLvwQJ z2I&(&T>l^i-k@6@4Uz@?c$aQyY`Y_c`b^re9*P#+3&N`=3Owsj^b8v^*5Qdj2(D#? zBI>&xAE!-(q0@oa%oZ3rKUAy?It{4%>n*5MsDO;&$5K=^?xKshn_oKW-orzwu))af zp8G=MA&c0|>5-=^8>DT)4ctyy2Ltk9uxB9TvJm&KgR&S1GIFm%NlZ<|L?QLm$lD5I zPu+NeJBHtVrQqL;uFY#_sjrdq>>@1=Occ^%;t1WMm>U#nqg1#SLe42%QRUOvQ-Z4L zY4UbHv;u^qNQxROdLMC$g*hO8{n%~LO??s?;GhyNv}*Q)kn}mjA?h3ABD#Zl#OVvK zSTxAhKwTW(0z09ODbGu^t~aBAddI~0uo!HY%F7D#XgHztdjn4 zws~vz)2vnwY=JB^VB@gmDt8Cv1*OUNbMAW|3%nNu{D}PYzy^Mw{5r(ywk=8}E9Y*J z)Ps+GB5;p5KcH6dk=p@3Koci}gJllo9!M2obA$8QSWFNTpWVmS8E6N`V4VnDmP#GoRQM<>l# zE zC^ot3fV?>d_E!d1&BaJsRp2(oIYA3lxYaDSrIPKq2WV^)ZT6VQ|B=-IIc{?9(aO4h zr4c!&Re#+`N*y?w(_jMWqZCs`kt38!>zdEq9?+uL8kr9!rG-&%jXbG{5onaXbg6$b zDvx3#TO&*@duB!lWx3#)hE64|D>7D<2Hc`sAiUEg+{rcG0e7R55hF0%;SsHGRxn6W z;`PWzg%J9H~TN?xMbQNcb|5!#B&->%KnYa0y z+zv%tS)96XB?tK6IR(m$>F8K)nplv``X|7C6s*7ZM9yS23XbbAejNErl{rr|o6c$n zURGJ4yY2_NIc&5wFr3{FuLQ|QXu*fXC6f3~&}Ohkl-2NNedmc?(CvC!B{a@G@mp{F zf+RYV-6ke}JH=#BWHY6T^S?-vf(N;EbPoNo|Lv$s&K>C<*K~-H4LP-VX(7*}bv@48 z;Qpr^Kjg=lq$h0kgZ<1dPwatNd-V%qH(r&@pZbr*YM18P3kxD3!W@#-aPpL~UK4gO ztc8r}dDZ~i^oq&()L;5v|1ryGF}}6?XTKub9N1euWMaySC5bDKhz(&jB<8`WSP$JLt(Fx6S-92}IXf}~65ZQe z?Plk!0l;TQZHx_j!GWD+n>j0UuagljMW6oaJreD}Z~CJLIT|weHI$uiS#ht;_sQua1zmen58HHH82~?UX22#`KC{-V5&I8vEkk~x|AwJCKK*vU9zz(L-Bbldz zXbG5=5sI-!Dd>&BLS@(q#g$T~@AuKVqUv9nUStXCr%)~lN$gU6yQDRsVP3xIoEQ>F zBH&0Y3LoUA$rC`m3n#0>msT-}%*No3$YzgizUs~plnIOz+>?XIMF-H@X}#k;`sM3f z`@>oJH`YUKj2VgbuCZY?)?Uj{d-Er|s4p2|)%q`Y^T>X72}Q?U(~~9>z#XR;$V*pJ zsw_baua&8w`Xkl^cF6X4Tnb1QJR;d{{R^T6dE7$ZzVJ=l3JQ-M5sn;B@H(N-NI!ty z#WIiDfX=91$RB2Lj*~}W$-E0*1>BaXn{>tWUK+O>_?u>@6K!ypf3~QHE~cBKN<16s zO|xylP}rYuBX60gen4D`<>dG#P&{e?@1$Mt0crpal(T zgAinotGa+QKq|S3OppB*HeWO(){OBNU2m^rmgkqTa)HI&`<9uv3q4(4;J`6e3r#^- z?cXU&c2Og3LQn8smww@uIVG~aE)NN)Bi-;$T{P%+I&!V7gQI@rvj0o%bT7HDoP>pv zHK=*o`LGvOKEJPJ)Cl9Jeqvryz$PB=z+Q=k0)$e}qVO_KS4i&yL#g)}a(6z;-Ryhm zW=QG6{_u3=*`+(^B1tn8ROW*7n=q}+XE!~x;2#)1zl{svO&@VxU%vD6}zR^H%9*F<+qO4aIUsx0)*eOzf0S!>BuM{lIUg+`4YA!uNvMk^?4a&!-sZ zGRme@fLcJ|=v0q?r zFTBAok7)AC7HJd@LbTp?#ryabjqJ?6vg_BsOEizXu|bLhdn^{>bPdy>P+wjxOPD@F zwSU9&y0}bVR%>4jjA5>Jp0>;7wh+zq2&;M}AYPUgo$AXq9 z>`Ure07@99l7(%w!e^gXXpqX?T(~V4B%!Xr9!QP{XR%)ufFBk4u3rmW7Wl= z8x(>aR_+PQrQ;)@d$>?AY40H0A+XN_fBLShhQM*{@2SN_8v=~jn49^23dwWeoDMXg zj`E?)DW;4f`+$)NSR%K5ZOu;&Wzv1&9rD^&di5O2nZPVs1XYVIEp)g;#l7YpC)gWAo?|m9$QKju|>4GaWVV+ zG&IM#haI9{Z~d{`Y$I@1y*)F5WI8Z#v?jpWOEE5x=M8?R zaDa5mG>W}+W5huSoY&FqBzjh(V!v!ae1eUj(at5?IN;h1BiiL`eL|@=jd0LmJFJ!_=43R^|W&- z2G$UnlX%?%)i%?E#v;Ua~FVCMPqp zCF8uoJe48}je0((U6Ms&Rs3oZn{s zWvoErxB<8Hh2`bul(ZI@U(3C8lB`f9NJ8%tRoo|a)iY{ntir^o^CxSQ6($^a0rc7> zugfnRy^@N#e~Kf!9oUaOWis;}rWg<_J3y(9eDAtApVJHE1rKC>^1X|$iRy*DZpXRZ zp8I9Vfv}91W%~|tPezo{Yg{yn)eA0?9=CyzCw?XJ*qKATlMyXZTYa}ka{WIL=g_%y zZrDJ`&5)L;!{53thBU-Z#duxABp>d|XIDmD+{ zJ8T5bLQsAK?~dmYza!B6q4lnrqwb`S26PFNVAZV7g2i*HZ-)X~ar7hvANcGBf+yYl zK{pWf0rI9oeG_Y{uU+THi6O;SxOn26)4XBFVUc2?lPH#lrH^WjZ@O2S z{2UObtoL1oI2ox-)c8VXU7{od%N$aCca^2rMmJV`W>!W*WT#}xULQfa8w%l=roBroVFf7E6g@KE;9 z+dYfvJl|-+1!*tu9QUr*1?dxBHPoKeD2{V0rk~(l3A!3s<5LT<-AuYzhPP;m8g-L) z{+e%G7e`aK*_p?=yXb3v$GJ@&O#$z&mv=#DWVE0u z_}+{>WvQ$sO8>JIxl56)Y&Bw@E2X>Tq znK;TB6tkHksg&wCxxMf>=@z1j$FLJ>I2eS!6GM^ZAS)EfI3`))gPnP@moa+!zU<7? z@q!)y`x_%TmhCw)lYGR^JUMR6fUGewPlFWmkRtafm4R&^B1qWmp${dM39oXZK_UpN zCyyxA%?cgR=eA0)bZgiN<4iZ<|Av0{gWOMqIl!sgDv1^VnODTtNYoS8$<`9gxL%7$ zb~zH5$P|jsknGuAbOqhx2Kx0qZXmZAl?JNQ2L)-fQ^Nbg+ap)bQy-euMdR+o2oTn6 zhvNVEMX_GZkTG9FT_%RBt_ss;_qZkVVgw((+##-YgWrZqM>{zm3e+bgp9pb7qeu@_ zpJBA#ct3SsDE|0DQ9g}i9Z$S+(Zy#6)h}AC+4fjB_rf>7ZUkRo?7V*?wa%o&!~k5R znDZ2Apj5ze9kLHuO%9Mgy-+IF&~?$Rl0mn2*8(uekoYdshkEppL7faWf1sxr`f?x? zR3hu5D?#}ZYG!c+wXTg29n9k4Js@es1*LajNWRUzU|>=)$egyjH*y=LNupbx$k~)f ztIv@25d6=9Gj%j*t)|S>Lg|y?qeiiIAtWMuJ+19T`#X%t)0Se4J?xB!!v>XO3ctsg z_rS6#yK>-SUke?x*e0^xx0l}Jw?)#ZxI{91tBKtgaAN?-s!F`EuCw>PyIyZ4Kkv)5 z=mls0DTy=WbcJ&Q%BwQ`kMOs9K9(5#>1rb-t`5@RIrL-6qzyh>Q)cgh9=Sg&Q$|hu z_R2fYdrJ|vrnU(|kN$ALIq{J|Eo0Z{-pF%)`fqCx@$^;p-@^(KQ~#sXJNZA1&dH{~ zeC;dbq606RK!ko&X8ktB+@i=0tYbiaLv)l3%?FQ{HqI&YQFn%I4=`v!*PzrujqZIwv0s#usSIuqegso(WfdD zAWdQ=z2B!#ZlI3?H5-gCo(NA0EbvU2rghBWM0AI#-ivSro)&9)w^+1)H)mL1U6URpNGJh}OF@0tTIv@B#|-&GINIg4wQv4O*W zTR0T~t2k}09iY8vX*4Z7OdIUB5+^qJCU&>5jFXps_{w~9OgL;%*~0km4k!*Q6cmV| zC^OrwPo6((m2{135jFY4ZC{?sT$M10~3n&k03O?-DOMe83%OWUeh^8LM z4hOW8UfzE3y^tDChJVo#tti9){q3HYg@td<<$MdGtWY0nQ3uq2G{Gr*idzGS5yWZl}-4V!CIRNOAi* zKXc9#uYD26z3|f@U&lWUwVfl|mpNyJxP3#YWsdk)%`WuI^pB=0=@R-V+_gFChG(H) zH^J|D?=fiCBd5`!C65CyM8WSA4}^5ci-7Cwme64Xc9sp+=C^u&XdZESuACSPag}3U zO>S*sh&vtwC$*2O-m1W;-~Y{FpV>3Zkl3>Mdp|eV3}-W6Id170v7j(+5EH4RGpBXP zG5kL0cGUez2nc?lM3zReC#=m?pKwU!q(W8!7e`&v)xm0m7D=D%j`UEd8olI^szu3Ic8ul!o_g@c+s}K5u77FOGToAV zPR^VieR*Z~ST`fB5SeGstZR4$-YLadP`9 zrjH`slnS+aTP5q<^=ZpqX@*1x^n}KY%Pco!_h|5CCbShw+RS#egKA9(2!x(hBv~(3`>8N~Af=3{s<16+{$; zsnYx5m_UUjYyMmuoIrMDDGDZLS zr4JUGbJIEO6lH;JPDi738Bp|d8|Nei?s*vOXo z?QHh>+=Dw#=HxIISXEFfOuuy4I1SejC7$}!%6|GI&q0rFJ2fv`{C8UaJT2bBE(ch> zsE+%dUH*CcC1SK7rA}FAN!)W5&I7YBDw?~QVp1uRg2~9!^5eWX!S1jD(h{|a+ZVo4 z-sN8+$^kyx=%B~)$(?Pi@`3Gc0>*j|EW`Nfr>(IHBQm1S{P+mTc3^)OWXeXNpoC(I zDG<8RlXDNat(PZp2HjJAo9N4)nch0Sb~dW>8ulX=5FshdF3Eb|LeXYX6W!sn1@;)O zXpq(cVsfGRejj&8u7$vgwN-axJ#EYTtT%HbFd5@-%eNo*5zF`(!&RkffARepQAW(@ zLT6==!w!s@%O;pPLop{QkaSd)a4vE3=_`v*1QyYi0d2xMU}FdVT!@FQot6m|ocZ)8 zVH@V6o^G#H7gi=q3LFvHEEJV{^+EwB@-IdUx`bFQ-buF!2P7NjY7`qi&O?2{kn95& zTnTA;y`H&BBx-L*;zp=Hz;D3iPFZ-G9GmPWVxeq;rZE6!+%-99ntE)YtCBXNssAwa}#6XO*Uv>w@UP7YFRFo?q?NAqC_Suv{kN?>BAr>$Nj(+xuB9c4fOKPZ!8Dy zCzQn9k4%12My&Cw*Of5M}fU?5#zE>^xIQME`Gc>w)d(>DH3N{DqIKGAbC6?hV9hlPeyiY>afdz9kio*dV+-lh>M&FgvD%r`cSTs3< zK6*H;v}dC)nUrnI02ss1ow~~PzkX_79c7`cXqWp{a3`x4sxNw+T9~6mpAhDQG&vG@ zqbHClQcvb6So!=7FX5y-dBc~mdj5`Eo_fXE|4R#uF6&RvFWf^)#tNr6?(Zk{CT?sE z#Z*${Fr@+!DjpVQqqY>bq%CiEy~Vk}d=#=A9Vb{62WNWZFG`aye=NG|6(<~UTfRKg z;(t@w*ya!05G?o|3+YLmqr=}YcP*~5czwLGIUin|N z-fPJ6vx*{cys+$h{I3Wrz?;uK&;C5_SHv>l=ast8{hkp+S=8V+$vFpJ_WaERFV`rh zl_D*a>LwJ%K{0)R);mp(?s5(13@M9LKXA=(&-6o5wH->RRLO-zAs)?;W=J2AT)HeW znRirlNxULbU$uWtJS0P%XTz#FBkBxO!^xySRNy5C0=gx4=;J)~Wubn#&IQN4-S%59$D!^krdXa31%e&&Sfc^Vc|61%4pT z3e_m$oRMC$n_lCL8Rt0X1P?stu(FDr2rux%`=DNI4%rr308X>voFbnhpONF^H$CQq zW@Gf2-+tVmS)qq|UzPd)%zGQzRN6W4WOkOf590Vs+SR1q&RpJ@j7Q{&^3im!R=AN1PMX z?9t)TB~Jkcyb}3&dgVftPQ+kv61OiTT~QFEQ8Y-ei!*(9y0_6cg|;&oV?TYiBmdc~ zGWP9kwwIk5`S**zOz<{(Im*QNDzcqj>4F2Bk;5iF(H@EcyOK|-u;URaH6B60AIld| zFBS^1a=;1N?$M!$3(chCV7}Vm{#Xng9j)?imozBsP2!-tqfzvGT<7ofQ}5)W8`B_d zb8WOwCVt%L`V40F=}7tQiZ@>KH{zvk*@g;|OT9VAeMCgN%a#^pfCk^w5YyR1K5SU+S-_MpoFh&4{ za=ue=GzV*Wd31FI`lWV6 zkkw%Md;%G@Wk0~M0?U+-gMa_+|2Bf7wu)x`Cf2&kJ1vq&rZiZ>VbPkzbeJ3Tc5>9gR|rL3(X)_sp)4 z7{Lmk)9&gRL0W|RIN2af575q7F(XE>Yi_T9uO}$1#d@uo*(WQ9{KhXfl&TD?zktz^#L5U0VVh;hWBhOKG zmBRhc`vhi?;=?YnVJB=&L+_?p(IotPkiJ-Hlk8`FEZmtDDyAgxyoq@)fW!K`7V7@@ zEndT^;+>>RA+DCQIA?Lk%N;L+Q2<;FkmCf&$?P~ajWG&VA_HfrF|M}XE_TRhp0#T2 zJI2}S%J*Xzko(Wsl6iy447O@2vtlY4f`A;5ygZKVn)!r>)GoLvMv(&&=aiinc6__3EbgaWZ|*5UqDh6ig3X4Rb4`4%pojc#U4&o(Te-1h`eW z)ygN(EpRWS4s>yJ-X{XLNK$x@y?W^-MeQUhi}0Q(^g+?@C{lgJI>ScKEDpc09~q;Nt^-Zp_pT_rIj z|AomJoG|gn4^j*Sc}ghNrrCFZ_%IXrq#LCe$I~b}I9)&*}g zb2rK*quu~Sm{<_g&QJB7+-Z$PL)$Y(#@2^f zO>9$r%9q89j20!Ezv(7Pat2A((Mw|u#cZX>7D_cJFNQk*ZNBZ{`&>Jft&;oU>UO#> zBu$RXAdRAF+QndGfbR=|HL@%o*as&^x@J6YoUNbvSodQ!4vw45JL<}%zjU}3u^u3_1q3PgkKr@@*l57t->^C_WXJKtVn@YyQ zr0w^S<*UE!ZnPgiZ6nJ`iUWsnKyG?e%rBc_AcMG-QpL-WTsbQg3Brn}uMP%xB1f4& zYe3Q^zbxr6_!d`~J~_5G;u0Ispg+(4GtPKy+hi=HSlPCzwLf}w*^i7ExiIg_6mprJ z*XqC~1VnB|A?hy0+@{Daj3R3kSnHtgd$>u~$~rhZ-S2rElmN7nM7{b@O*K8l%XT{i zeJM#C?8!?5-?c3m_tuiVTx{7Z;nac-Db|}8LYx^&V$yI!UPlf}j(b0X3i55!6XzLD zfWw`1A_LwpMujkfmpo(G2|IX`$Q@{LLYJvgI>KwGJ`BZrwU8}UW33H{^Wg8*dspxt zlzSP9`gXZD!A--kD|ZYdH%?@JW}R%b@svS}s2UWz8L?LJ)o(PBbq;Kg3QW+mm14F~ z@LT~s#x|K|PA*-s2zxIpBGR~8h5EiwqfBImWH|&a5$>|T4Kh4p*5m!u$k=BGD@!zm z`~6rSa}F4X4e(hYr^)cogpS78z_PGHk)df`-6tC$Ya*NIc=ySI!^mJ-4U3TrCh2ch zu$U_UMq*g7(P+e{_3k5k9N1`_F)=BV^MnU_Y^xpzSX>YmpnuVJTt{}f)~ta@b=UBX@MX3I#FRZqhD=Wo^RY<3c>=kK_S z=b!w>b>W*v!&2sRPDVOk7|%D>B%$z-Vn7t9hf=MG`p`QM<|l*Oxz7EBa-BQ0q3sE) z5b3;&maGtbEXCzhK5+A<&)76m>y|CLwCvQf<|X$$&{>8iZE$dqUvqarg#vllYIyr1 zwk+!PyvH-rC2kC^0Ht+(8}V+z4hDHZt{O0z&hvWtHIo&VJUujWNkA-*Tae*m(DRt96v zUhAEs1V0(hz~f0G@TI*{eT%~!HivMaYRhr+E)+IpzmEG(5eq#lC=)px5FI(>bZAyK zIYi$NA9g}z&psd%y3LJ_L~i*Q0R$~802TMcdr1)o7xm4+XLNC9nZ27X;?q3!Ng$_5lYb&yC+`f= z_?FJC@*D(FSrowPlfv3~K`GL4jWVl01R9ff<-tGMToD|-bpcLMb|GKusa0F8fxpavvPkAB`L5eO? zs&69mSYF3DDC?AU$k+R7dHvxnQP-zm_QXB=BJN2BLheEcF?>1 zpMN-Mp6g^YlR5C3%0gxonTHISAxs?=h}HLIta-_>V%j%DPl}!%7%zP%SGnD@6H-Vz z8c##cX`33Z(IYmpJzhkOTyL+xfAHF$wh4{qDSqpXUy#JH3Pl|l39zIZmC@WzF&Pxu zOsUYL(eeAZ4g8B_Ybc}}(K2kCGbBriILPe^8H8q;LAMS$aup%bsNKikFn;V&V6SmE zy{~1vB2GyD^KXp!ShnZHO!AQfPhV?HrmsPYfpqtM0~cDep1mO}bYFnMJ|Lrmx=^sj z6%x<`l8exhGIFF4h-2P)=N%+a0~?{t6m|J)6!n3Fu-FF48@Ce9+J)*1i%NqGJZg;) z8U!`$n65_r&aX4d)gx zOIOkdBNKS*XKECM0>mc%8FtziP!^8zD|e;!WVIWF1PTQMlJh`(l@eY_Kl-Ol+`W++ zMN#-5_tx}Wx{7qWT#Fcz6^a}-%x4*BW3+8fd5xU-cXP5=hfRN4Aa+F(5km?T)I4Wt z6lKsx1wBwz3z5&bQ1lptQD9iH-T~620cq2UsLAPvVQ-4ST`{#9cKO5#Ayd7Xb5}x* z=1KNv^>(t?fz8udlc}nTVjzH1L8+?wrJgNOvCcYirSQV+EM~QHJO80yc&;G z%VOtN)4H%^-hRJyrUU$A+st<3LeLlcyf#4;D+E!koe9q7^`aIEEw2eWWHFLhWwfAD z*bc;V>46QfVy+Zk4eTW~@W3A~se~rSeAgxj*xKe1Tk$;Dj)E1pw8<&#M!{kE#;4BS z=xQE5cGxM%LO^)=dT#g3dg*eFqEog5a*nkgx4}@H;G7J|h4v{t)EZnGG%P(b3tO77 zLGu)eml`gDrY7k8{ZKgQ@sY5LhQ+<)wyT|5^4KDIEK_0==$hSx9g@Ghy;AtR(~pPidb2Rb|C=L0z=A_}$w>!x zQ=lYx)QY-=VlGgm2?e?^@0d$>B1IK8G2xLC2#nUlHiP`EyI3|R@?Zb-dlTVvbztCF;B(zY zr-VO(i7R7z24`z%y;nB3OFHD#p=gJ4xpqG8(J1O<7%e`lfGAz7u`nlzlQI1q8Fs?b zth0)0z1$S$4#wy*dDduHbDTYro|=xV`5M;e?3M(-`n(6@FlHO;@gxMa#-kUMd*P-b zs;l1RUI>!bPeN{UD`{wSq~xW3oKiYvBI@snu}s{lA`Bo z!m+@-7t7l#OXjIjaiW^mXA##%6mhe>ae|oyaF~pm)krb9%KjfLBjx|}Y z{_{J3g$~+O-ws8*^dd;K@9_TA3rHw#Kqk?!qu<~9_3C(DVyn%?Q$7Uj7|$14c?~;| zq+OJ}${0c7_Ne~~**sQ}f&)9GP=Plpq>@iD&~25CjkEfO>ohqwNgJA@)ff{&x_R}H zK%@eSjzpm7PeXD;48G(jYashGNarlBQFhv=^xHU!6GD$2@nac1b9VnzskxKN#vVCv zwYP=*ay!46&hy>N{ZM$z^ZtU5X4U{bu#W$Pht+U>vP}9BRDsunglMb$JhxTeCf+mw zX*MfgZvAQFzz&<+kNezjy~g_=^Qef!CUh)BKvt3##SZ%DoEo3>K&VHNycsV&9*WV~ zI{E}>V0weBhu6XBgEsv9ut6U?EmNNw0QLj^)Gh3?!?Lj$V+}L4|IzYLxMQ!~zQue*C{-OItlsVe)mFwaGVkjhP;1_ZaaxX~Z1PQQ0!D_|^ zez&KFn-rM>)&jeTx;-2C`(*>-zUhaQ$-H(FJuYQiaGwOg?1kOF~`s) zxW$3v9$J%$axcXI-ERS=3aI91fi!12sc_4p>qX5`YrgkR;lEccYx$Q8OD{y( z1!`7d$#klG>XSWC)wZj%*w=`dqTwHulPq?Kao}icr3qs8Q4EC63Mtj!WZg4|WM@eB zY(upf#8{2irkbBmcgm7o8mAu@tC2Mo=y@vniAtriC$T+U+1#K9kz46@? zvx_3R1_fZ0?`=_3%~t2=Wwo&C5i1!{+!d06kKIcBO9QaD5o$;vT7MGSHZWp8?1bfw z27L9&b#@HXPZ%w>gvh9S`40Vg)-R1{nfm14=MWt`v^cPVS!vQ|_<&-1De?)WLZ`)0 z(h6;b>Qkh|v)oI&Sfd@p-&pd2IkUQ;Ka?N z(MX*FF5MUbHUV}?`{(a+*C>#6H%EEMA1h$-8*acl@t(&OF5bUTlqP>5>jJ{vZkM

    (;xWa{Eg=F4p^2n`tiPjqrmxX6YP6^@`tRb!NMfx8`r_brcK1F>sE>_mv2!-~- zrGp?{tnXFB3f!AP30~Nch?k7ak`u<=h>5b*-$v{uE_m1&V8q?r%=c4B-dM>(2X=Ul zo7kdqiU9(#eUxhL!e~LFJl45+R>7Pcw+7k6h}F({B!=|UvCjGai4ht8F{FoH?R+~b z&bgDjnUlo1BRv$_y%g1tme{Ls!mfAZEg?ywA3>LOj{EvY#0-`B8&O!dY(oXf8OwY*uq$=c#C+|i7>INhBOybY zTr=yO-)H>q`E#*x!=Uuo!8sb30!)`R5gG;Z4S_!ed-`NKemZ_X$qPLiILM9TbkPL@ zn3Te5d>Vr@{h_}ki4za(oL3cj@_%^a^}F}Ko%NCtHcletL6YmhbyUYpAaa0WN-44z zD52@Zh)f`+%kxziaeMU};PgavD)Cbib9WX>5#V7BxZ`cXpPhuej*E{5;3?>RpCkH1 z`E);U64m%r(pRBV2=!5b)&c%!BdpkNf*Wywo}R46v&ZrTH&(uTSXlgP1NATLTplxylv>%lO;5xK7heMrrqF zL&yp*pIv4D4a<0Wc_FEtOk`4#9DKRFV+I@=Yp~%OS3Y*3B)%>s=S_})X zo$e(-w}R|1*mPLnhXRg+@)eO43do@?xWj1>CXp`bASn2?Ft9AEnX?a4Zs?63hU>~QETSVt7b7298C*n#8*ZdVeYNT-O1OcWTdYd9gtRukcqF?ysg0f?- zCzj@5wbrLHJ?WxGIV6LhhCK)PD?#QMbMS{i^Q4PLYG5bGes=8m>Gw^j0olz*%?SBxSiiSmKC z1<9e%IA9cMBM?hrSf*8yC7N7#SceR~HL=Z*TX~nM^6uZ8Gjv!W!oYmByin=yp<;CdX=vrbP|>0?#ff9P`x9Zizyho-aY-{d_u0 zv?ZW0z)l0e;z?UD-c~+3mZz~Y zE$iIX?fh*4mxY+oO6KheQ={sleUUOwlptMzRh2j`h={WcBDjeW_jdSe{bKKYIHc z=GC10O&)a~#lW-Ag5_Um9``QUML&@B$uD#2$ncDuxzRx}|9oAn3oG$&4~h5am-oVB z?5WaRHmacTT>=&L0 zMAspo&Yz2)hMgLvMG%%<9W@Z1qpS?RDm(;%`y|eK;B?#M-ohjWx; zu+Lu0aLN1V=f5-Wg0|3Xj5>uK@=E$7oy5uGs+$$MutWyS9gi$HM(^d;LbMt~mFkOO zxNxw;%Q)7vVe`Zd@8no_vHd>FPkgKw)M0yGwLhBqMxfDhly;TGkbGxy!sKN9$Q|prH z-wU)kY4ZEwI@wNc#iB}X_Uv7uIc^EAyF#JZI!(SNvMTTc@j6a)P_-;!y6s@a9yum9 z-TZ^iAoPMDhnw;3-+sl2oIP_mJ*3)!9hGY)kZzzDVCX+hsSG}8UT89KVXuMIEs~6? zZ}{oFtLc;oFojE^BjW^{7YsY0-2+K`dY_Lv&ZCRIDZDLLXN96b?_&=LiXah?A(z|E z-^+#MgA3I50=wWw|E>`A9Zpv0W4NeMy2-tPuHvb6-UB3yQz*&`#q>lKuii_I{hg0J znx#hr4?z(z$o_`1dk<`kb#v^G{>Jy6j z8%6Gdl@#8h>%F!Dq)^QjGf#*89*djk0iv$qB#D5ggjA9xxO*ZfBLnw*Kj)CW!Q zSwJ!1WACI?TE#g5{z;Rs_bL;f6MPJcB>5mUyiL&spPJ|xC>+1SteD;{OX6H4^?_#< zCpaxpd%1XK+^k+&tGGly{)cQn|M@-gyJ)InI;0T_DL{W<(4OHqAq`ma=*q1g2FIS(A=%Q0ae{)mgM59q4dq19&Pal_Q z-7f{G`((E~`~4GN!oLr_Y*!qyL(Vg2j)?;1v+FG*hbdm1I1wgp2gZ*DCT<-Rw}Cz* zh(X+QU*~>8d03eUt_u3u8$DWsJLLQ441#N12!D`y^e_fo+TO&~SM(8A;5THQk zZ<#mjR5ve89v##rteba|Uo7kkFM!@1bqWv?b;wW6tD86E6fL+9|IH%3-MUdrT z-VJ1-!A3`8tU-NEyea%1FE@Dq97x>izz3{YbYglpA3iinTP1az0f{{@My#Dij($Ro2*eHGM$$pODO=S{P-=v@Cb&UJKo z&_iM$$g#!OXsj8x?P8Dq*6{OR%Z#Da%vo>!mZUjwC>2OqMuk#yDP{*nGAR|d@tsy` zl%@X3kU8ApTd@cSky>B^iuA~R58>F`vHqn=8^@Zij3Q zxh5LDohIKv(wX}p_=ZPXqAqcI=}Q5FZV9|WS_gLj_QJ{~dECBm&|DA2Gq(EP_P8cU z;WhF<5Mv%BEwq8ZW%`E^yCe2_T$rv=T=T1@?|YZ|Bn8(5WOH@=Mz}}z>>=4!U;Lf~ zW?R5j&gPen_*cyy-zz2CV-mnW z<{R;>xsDb1r~GO4@QLSr{e&C@yW9^4RLutJ0rz?@B#^T_ZqGI!<37V0NR0dD6S$BS zB&Pmn;n@arwIYXIQd`g;GE%5wuc@9EHPsjEL-nztESEb#GnhXsA6R*Yosc>eb3J%y z5-?*8LY{u>CouF@Udj$3zUAL+G;eNavxsusa1yc5qko0zGjLJLjw~7g7djzw)l8sp zi4pYj_xNC+h@u13zPEd86bF2Yg_t%3lFLE&WL~bafi4iA=BZI9t&)pmEUCVDrdNmKX&YHiQXCi-&@MCz7TFX7 z(cG<|mIwv1slK?KLkGeUv(#=`i$=|p{qc}>CNxMV-5d0QA@{ruIGb@5|w8sOei2%jgYr~fmmp|_d zoDxy`Z0uC3q@!nMdt;=&K5$U39`RGRDsY!t>$+7@?p5zq=amYr8vi%x{P_2(H{SZO z>Q8U{=67$Y#5ChXC~B;9cfc=#iYLPB=D0CM(2p4H#%!jTREnffs%}!?x0jm*l(dKa zc1Z5c(0Z)mrOCH~gl`sbwjbm6aVG;kpTEH@Y{%ypO~TgsFS~x*&p5sOcJJ}^WGB1n z#er!cYfO+)MlnE?um|P9TP0=Wez>~UMJL%a``pXJPP+q&!-kx)$p#*>vX^@$hGvCA zU)jds_7Gg_cF|gIpb)*y?Qts$-w>Q30rv)jwLx>`>#t0Tak*jan8}QVjUOA5f}tue?xjW)}?#)dP}2@La1Ff;hZ>*KZaZv=X0t8D}}94$_Z_{*CUGME#bv8)uj%6e5$%F8PRH6Ryqz^{%z^VQscZHanF%7i<-wVk#@86QQFL@vXH5*a}0#z%6 zD&fCP&}6U1v+%MTo;Mb6X=o8Q=8nR`P&OL3~n{1~rTSCnodZwSh(C|EUP#ksp zPtW@zyG@wti?kqyE+q)U6>zYL9sAA-TD-F0stIv=Ni!g=QW8Z1V_>{*r z=1+}~?RQu3dllldON?5dt_FubN8%v_q9G} zp-|ww=$3r7Gp2i5q7Dc*&Ab=1i%ttI;^hiX(nn?uk<;|0d6fZ!lTejnElNz?0s9gd zBZ^o}C)DneJ;%Rpv>k!5^Zt$0Ir5O?f!?n7O z)+nn;1$4t|6zw6&$9|6If2)zuJ^e+U`Q__NAmI(${ldrM8=fOqbbLLDGd`Tz5KopYmKErE`V_D9Km5R) zFZTsmCRt$8y#*P92V$(-#aYcr+gaIk3xf)5Lh3r09Pr3!nibm2J1D7z)_`Fr4A!NKuafwW=~-KbXk85>kHVuzz$2` zeXRfEzY2}>Rs7Z)zaWVYJYVfLLGN~o$)Lz)N_Cd#Bn{H5oHp@+fGu-xl4HYv8hm;P@_J2i)MF7OXQ5|5}K$Vn9{B z&E=z+m~|O;(nOX9{a=*TY>Zeve8^)+Tdi;tY`>EzHDR) zQy&TJ!fs}gAvq;z@!8-Heb4!{))m8oS-dmC^nhc)zioeO!`jXmd|5MK+WuqifOFmt zjWdftu<^U(k~4W=V!%5o27<0PDb*Q*9vt*=VICfnHSO>bd^|(TXcD;-_GY6L^YqnM*~*!kVk9~L`K%To*`P-lj0LRbvxY|(kMkna^zFT z!+32q^!uSNm_BD1Lbm8JqIY>(c9qjANuHr+<3}BnBp$ku9iTjM-z|IgjZi3<)F_d& zzlf6(-awCh6;umAVT+zLvx^>k1w76nde)+Y)$y?g_yw^>|9tDUE@vasqI7So$%^Mp zz6hM~QC@ip#UxQ=9i=MdcF2pDByp-dhv}B6*smpV&XKxthSJ6we(U#H1z_uMXR|Xc z1NeK1zb)e#F)=s&dnd_i2li`oOvZl`#cZTVGNroy>L1hpATu{dA?mVAl< z^kh@2ERcyo5~2&fmp}|}c|S)di4H0hYhB;Y6LdxVz8=;{XBCYQ^iS|WsaiKCYZ}Uvso%R?jq2VAoKxJNcMykQRv8* z194hbXc2|1FU7uxVDH37n%$6CjSF_So$%SS-ur)p=E90>aN)ou3>Neu*FvVaB`R7_ zpxiG`67A*UyMJ1C+zTuXztnS!WFX|q;%??L=dADvZ_-d=H9U+4V#4EY3dd@QuP>a{@);zEP&l~Boz^|(#+$NsR?FALA7V+42SPZ}&>lvp9d%2Bn&Y3F(C zK7QHgmQ>9BQykgtz$=?mCXU8oiur&d2Pjpn*Zn#1qK*Z_exGk)F-kj$p4Bt0k*+53 z{!c_5uZp-hEMwfnrQrl_km6{*+cT{Sg(!+@uFX2 ziOTA)T?1y$y&_v-h7}$xxf7%J+X4)TQ&zJ^?NU+0bR&9NU%vDTE;wp@^Q(>LXDJY85X-Anfi_juI8>SchWzr0DZ zGN_lAO&-Hyx|+tn?}!_{bo@#3QpN))3mCBJZu!ISUNn!oS_ou*^pk@>Ok9>Gum8!- zkQESk&!kZ<@c*;-C2&!l=lUM;3^O0j&cMtGxG_QoF|s%m#EvFyx};6o+orv3Z+rW1 zuiUn#%k6EaxlP&>aRV0w6;x0bWf26GMG$2d1vLs7)KT2n#Kd3}2ohBIzwgXon2|Aa zV1_%|8-Ip5>)`pG^M3F5F3)2t!_4MiW!1xEb3+@Q%kJ%T*lGgJskeUPX`Eet+A!rj zNp#@Z1zeF)QTH5*0RfwI1AClK^VUx4t!j{Et^i5<2jqTe56mwIxLxuGB+09UyB-*l z)540JYgCt@vTYA}D~)<o~Sz5qR5w5!D_aRe`C%~!79!suWtH% zX!~gJY~&Za?1i)?E^~C63^fL5N3Mhp5uhNf3f0*wqCauKJRW14eqaX}_w2pDFz;Vx z<@`FV-2_sf(Y7dsVs=tw2iPKc59waerT$~Nf1{{K^4}xd7;?>|R{&)XcDi@a%^|x~ zA1)}8$C8Wof{@9$j|)#Gcd*Zf;_vi~f3Bd~?N_c#fE~e*Bdh`9l`dJnJSHqooFXn% zRq`^q?V@J4IPpEfRY`8J%_kxYqsST_7Vcx~-L+Ep@4aYrPn^W+5|S~E95Z?I4p9s& zIuB6cpLt&NO#gDN=S45=L&YYqGbF(~iYoAk;cWChDT)vReO9DTt-MOUAChMc>KfO4 zTD#FV3k3~ahn?aTy^Hn7+xVz+SH+FvHu>+T&o4SX|N4A5uNw4yQhdg)6&csRne_v= z;1C#hkL!0B>mpY6-El*fFaL1sJ@eJD1?JC_&PZWj4zk1N=u}?S^kn5hmo9a-BvLpi zz^=~YGlrcKTpQ_hb%&ygZj|S^L`_agWIPZVYe;Rq?>T{p>W$YfijAfvji1m;VxF61 zajuDlOs1G5ifqT~q}G4~OEQG-qTXL-N)t>FCWJ4N%10D0%#jxw^>ftSW7Oose0#X!+X0To`t zEfVb@NYEAGSIa95$pC|P$gP)a*grtD72E^vna~ch__a{nbus9u8cD$_XxzFDz3nzc$+*B^ac=zLLALe`Rz`0sRrLq8IoSrAb&ms= z5Llq*0D1#O9dNEAdDpHv+AB+Y`Fg6Svj8@%;%!i`5ez$ZD1h%vp-op;_-|xV=p<-5 zz&ZmP;%BV!96zSUx|{9x1zkJx^0S`VF;v#?QANHOuec>ib<5j2wdHV5f6 zCG$I_nRIhdnJCqx&+`PgmW%sO1hp{DL3Xh)n`ut9Zt}U54!|1qlE)KX( zk{sA^DK&9i_EF4U3h33tb)55y9{U}T3|@p)i79^CE_x%g!KIsyA{BJI;*M-1qi5Q| zRuDWL%TA!DEk@ia?F`CLSI{`USZeb4{E3b4lcSD5EAui{%>Q8FSBz-Mo6EUJsvHW7G1|3m=n;4FyX4wR5-Z-|sZ*X|Uz}|VNR6^C7C*8D zJUd9;NV@rvc`2O3=7}s6w-tL1$>T!fr)%%2QdNVBqv{yow}5uHHsD>)5gIn_GOJh^ z8dk`#a1VRzW`_*Tdk1DsHyWE;zrOl+679go2C~bejLi;;0Ss@Upl>5td4ADFx?6rt zTFKerdqa{-pW#=jI+TNg67en|St)nFu2axbQ#L{d2`#PRWymo%GSrqoL z`r0|N-hn3}Ad?(52_;d?b_$+rcr~wDnX2p-r7!N2_6B1uZK3ByRjRNF+|c)x!>S8( z23;6BsHnEh3vbmUW;xzg-P0B)v2szU_9OCE^Kgs9ImAB2zM;QA%|IIC!Z!{-MTeCNi90xWh^(N-z2*s38qzKqYq39tU1TI^o zJAB_pNyHLvL|C#CHOjIjC`I-*Qfc5yIY!(=Mje*`2I0vsoZjM4x z8mMGO3bSTvQRoBJr*gdT?1(T>Pz-w8HfWkSc%H!1lSmlH9#-I)dg^Z9Vsl;|3E}`x>5iVkamEU@ofaO<0el-tS5)68Y=My6S*)9hAXm!PV9p4*NW&T{jZ z)j)E5mp%wJyIFLs_$&lpY=exa58AkJ^YmNV-gn5UXnulV$m#n0 zJh$}4kzOZAp5%~Z5URl-)(3Y?hdYkk0do=f9V^2MI5j-H`$NxKZeQ>QVV>J*g&!uc%Ar-RW#QK6`dt`KeV+UkWoSVQu|&e~Q*ju6=uI)mU4kY62^;K5-h zJh{SO=ebIT_bHm)3rbg=G}4^L>-8D&j#!9h0~2T;Y)#zAB`^HW3G+T!Hi=9JMv#TB zSzLaXl43tlHUm;yx(hVwu^2)puLMoAVW$j9Dd`d-4@Y5e7rk2p3wCs|P}Z|Win>|D zPL0kDo&`R!1Zxu0p}qpx&JK}Pj82Y(YslF#?1YlCJKZ-uv^(17nB^(zPyzF?dDXAOI?fj0fa|Nc;0k(3o&~Yt^&DVI z^aWbQL07CszUq&%d6-$$q5^QfM-@nO#zFo1|G~jZQTlTqo%SLEI3o|-r4mn+w=(x+#Svw#<`_ec>YA5(FWZ~IO_?y_>fZelq zSkrMfeAP@3BPg1GQM{b&V7EYa;5tnR5|3ISXHyJ_4(*}B+xR`aC?Em)dppo@T$qnx z8TeR54??u&jGs<(#;;qkQx!>82=Q}m-~c%;?t-YuO27B;idJ$QQ~qqgH$qD0{JZo!*+2o>0 z*kEVb`u{7$+q@Ur0yE#WkSJk`|3isEEC)lSF`Pr*&HgR^ps)vb5+KYLyao`PL(TTvtW3z1O+6=-<`hrQ2l|NxrmCpd z41DVkAs3YBlLcKLwcMK_lK_|}hT19wJ#m%&cP!rjtAFz``lWR*Z!RV24qTs5W8!HR zQ4Fkc_fz3z>Kl`7}D_S=A-u+UP5C{gL+~v4$gYPF_ zLT;b^@^fyOZRrc{57vj^TV-@V*#tHxCS8icLmnkN^!M zN>hl5v&5ia6d^PS96`EIuR?{{y!DC_nKsEQVn!PG?2N7eymP+dlsb(Y9a`gyx=J}> z^z38Y3*9l`kKbsJJ{rHFvNM*(fg-~ttgtlo@4paUH5a#I)ak~m z%=iWCU18E{hDy0Q8Ve~4p`N#rU?ED9qGxWl$9b)U@ZZoR+jbp6YkAm41R)NLrn#tI_Qu$||^T;CVOI7cYRny3BlV`b= zV&W(ggH-wVB!k?C;=AOwWXS2jqC!oSuoPzady+}wpC0`@ta_gCevQ6nGCo^=`JSnC zu}5OyyDu1F;r+YvWn}v_QfvZ?y%Yl-#=EKTCRrtc0&qX1Z*#49@GMYHt+>80Y%S@DYdxS0XGg#%JMy_eQP#j-WF4{%sP#c>_H zeMdej;v1o|^y0Q-WSs*qgMfr%RKhQjVzyCaD>SS^cLz$R4hpc2z|b&)b&B9RT;}!& zjyiAq5{7OkC17g=8%xyM2cl!#l@)5Ie*NlywVJPM95%LZVWpDq0p#l-+>2^)SlbL$ zYkEas1juq5Qm2mbu%Fn)0u@hOYtJ3*P%*`8#*5|&K{l&h$IWFC3wg~GvtxMiid=3F zl*C@+REUz5l_AK!p3OPH&Gx?IGblLjdrMLW`OO`3?KBNOIs7cZq@Np~yn}uA6s_Hn zAv4Y>r@r;=Yh?3tvrIW?GXLzR7|8zYq{2HmS6y19SVOfdG&2zN+}8veRxQI$*nqLc z|F{IhYY{?V;W0#Y`-9O6O~Qo479=BHlzpJ_`CVT=dsc4V?9Rr9IIcrQEVRbs0=7k( z%_Pi*BD*^Qol=bJukwx+YB5ZI*Eu(MmG{QLL-MPt%pW=Ae#+?Ha$WDS=QYhwIqFU)-9 z^~lA`jc_TPeO6AorjgYq&7JotrjH``V7bdV4~y7`ZrwCmX($bgS779F*{AZPz-=#r zh_!a*yi1_7auqzrWF>MaLG($_R&s3SZZCZ}6BoYNTVC!`0%l%Mt+v6fVetoxla=>L ztN*eMZYA6l|D(=LGOQa})&Q*4u&;97<2~m@d$@bt9x9;g%D(`#viG>QFeheTlxnYv z+W7Gb=uoGz=@a+?bKoP=ANO)GDcq|tthqDP*&ug#O=3Jhi>~IB`QjocQ{T9kt?pR} zxgF1Yyh7D8*w8sfpv7^G{fb0>=(Apz-^=Y%qX!eOSUW8_u*>y^Q^K76qNr)jnn&E8 zg+0J$`G~uE+JhghnHCMgyAja$e2X*+9&yiW?Cy)!D%UemZglSjE(PQvoHaml z#D}4u6dI24V;--99{$pPn5G5<9}D_|aS47RNGIRJE%tEQS*-W!k z1vIG5N|3?R^N|7)1iV=gas1W-ef6LHP@Lx3H`= zLtVzt4Zfsr@7I>7dZ6pOf^T?f>7C9y-7tQtkBeo^TqU!=VIOBd_c)#`%S%`dgyUvZ zeUVq;Ip*>X zPgbQkuFDPjF{7uN$70^cgAa1502K|sg9;SJ!rGB23Kwe0UOPM?@*>g30lD_g5cFW6 zXvI*^w$%#>DeWGJV+9=(Kjbm`?DD~~wfof@Kfh+Kz{@5;=fEKt3wpf|IQxV*HM{0^ z$~P(E71^A6vS0K_vWKdpvpD?=bebxD9gQtbiD5DRHFMiRIs_NuYXqBJ8wg%8a=sl+ z()eLyduMhu9wXoVeIjFYRRs_FFOzK!?5ctU)hJgri(=9#l17Ci13!AHz;^6+O1h8E zr!j#H`NGUVtj^D(BZU}>8Fp#|amK89O|lqqau|khqGdbgMhcOS(2gU|qD|Vwd>A%H zkCm-ITQ(tMMREdnY;Jwl%MCgdt2qhWG_g)o9Gd6WC)H`%JzB{IZaUp9%H)1XYRF!C zh0iDUg@frZvNH^(*V=Q76*#Dw{~qgYUd{GQxh57$@33`agIly)zv7W6FkHKxX8HuE z3X%f-7{|m&EU6r%%xBX01#27oS+*3QE3)oE@>DpbS3x!d4*QFNKE^(>{X z36rgm`<+bWjRNXukd67-GV-RUiu24{;vKfM(n6Cvs@}zhc0iW9SqYi*bP(~V46%DG zWvmBn-T04nrG58V9{;gsTFL&9`R2=LHZd0m4%%5*bg$;E=A9;KbD)L5z&_les0=9- zX|vV$=AZFf1y*Cj^i_cMc&>O3ftuq)oI!__E@VDiy^ z*jn`U+L}4WIcn7(XKWxT&)6IVN>ihPFb60G#No54@V(%IVeQX@<{X zNmoEAb%J+Ml_oZb_LeRx@;>mDYmzQ`w~tP9%J)Q2{m-hG?50yaHX+BLY`b)iQ{XSd>#G)7!6uMlnbxdl27t6w}NEt}ayYqi;& zI;e!Y`9 zG3O}KK!xj>&-sTWCuU!Rlz5x63iJ}v1fM9er8vR+JhT&CVg_Jc6t7t2z1Fh`D*n6F zI?XY4vKr4i&N;5evr?TaTy>fl@ePj@dOxHIF8Xe8sqsx#9&)(@i!kKd&-U(}TRmH+ z>6H)3d*ul{;AW6_@;~`fj2J1^Pm^Bmir_tg==kc~&dVB|d%2C=OCq*GA+{(qnWfll zutegozqax}%$-~dzGN1ZWo`&tH3LHK@}#ge0_aGr0J%7}PlDwyIM(UbJ_!bxxEyLq3~@ zkK@KDEy*k7dlb#?go^G*-g`)oENKSt>2P~zS9mRtA-957`^GEo1f)7+0&;ni?2Izi z*|2*O+7+z>)Dv@tIWBF!y{z^}jpn(b=L8}avPMm^-Rf>d8{H#2;?qTUN;Yt-RlB`X z#CrqNCsoE-Ka|$O!sg@t1vf0XdUN+g`p6vEa#-ji!-CF2XdlF`l+7;3XY7Mns8L=h z8X@v(gN~I4oB@Gd<059GVl6t1X2a}i`)!`>f1yrpT-T(`{Oun|JiB#`1EZtRWL=X% zF=!F+ggZ zi$pHPo`bHdd6@H#Yv-`a+mD6E#8d~0@G=5m3)M5VLnsekyeQM z6~{UCDxKyyIFKz0eD!snc-~&%d5YnbyY$ViMiao@_97OAtnOF0(aef{5Uq*n1 zYE&X%8^vs;NE{V@=+%2m&b`*WWaMYOVo3h*OOHT|7g6~X11y@wmG zh6NOo z=B*0b0PKVhCCOk%uS<;QL{A?eaqjyDpG?wr;$?7HjQVzu_~6`7p?b^gn~Jpz45@ z{TOk(hxU#Plh#c#ymsW~!%oGE&PwVyHX*{Z4<@V{_NSi2YCfHk*v}Q2OM^To)0PDx zk!B8xWDY74`7xXu-kq~BkDV`XAOiyIAVDgKZo0;`cX7GPkn9rh*@CfI)G;Xyu;W71 zQ?X#1gD?K0JLh-C3GC&(<1@%d>?SY=c5R|eC`bk=<~~LGsPG{@5M`2QjLI~4ZKpSLtet|ldi%8Rj;h1l1T?>Im2!eC)}ad?@oEybT^*8kNQiw z%jnzXxi!nv#HG|dzdmULl>R-wVV>LB`FNm@-l1McYCRv{^2saUrkKt=F8xH5%2^}t zfvQ4>Wg(dp{VI2g5q#=BCI3SnIXP_Jj3k-(LmMe(9Yxk+$$JW&;+I7u{n-CXzEp1q zzRFA12=utI;5EP!DE*bFxNAshb{tmc%d-~4d#NtMyZn3wX8IFjkW zxrJI2AETIJAkB1;3Qu3Wktz2dq*riGsdIx(x0*Fwn)6WWHnO*#F7v_j?$R|79l7GW zIk-qZAju7m4#l>l+~A9#%zH?ZAiyIZE+~>$(KcBzj~-4oz#Kh!(bGtl$@uJ`7%(zhF!zI@i+DxmuOdJ@MT_(or-9WUDFYAx5O+mU z_-7K|UUow;@keXz9as&4<4)K=Sov|T2 zb5g|E!Hc!I@gQEepU(8s7IE4+6@1h$L&Yhm{q}*d>#0x#`U5Dz>y;OfD44=l`*a5( zdn$6j#06)2$C5p&PCkBxwr1{CaCY%qQ6hRmB@*lqZemre*svDQPoBpL;Zu3P_}3py zL{rFdUlC$KQ)r7%cR(-DwXEi$%^Y}R+bbU~nMA|*`G8pipwG{qyieN(mTi0S8uyRgm7ARVullHblLRT)l5_G*NtRWu?f zdWU8ti%y0V)Wx8qP)if7NOdlpjqFpvT-Hi%+qy2;8rWE2VDzWh5>(5wzkko)2%G5m zo`WRMf&JXmCO{~s7@*}Yp~CZ*+#=Zul>JUt#*%HE6~at<4LKl><)P+Xj<7~m&b#M= zZHT$E_KUFW8<$MGW|i}F&;Z*a#r{_$N=@=C4l#7Y9+-uniaob_p-f(fqBvxOTa&Cs znkcoM*K12e*b)}(W}=Q$bw7F5@sr@)3ET}mJ~*744tZC*0Y{Rm)^i96g85yOl9C=5 z2k-<&-s(#)X3okrkG!z)fE^ej76M2}zfc$&$xHG|p*Jwu>IzjXZ$N;mdgo@viOYO8 zG9Pf$iG2-*>F}^JAEuYub*t_NYp0rrbDtAzSPK5qgWPivVgcqlWrHZj|4~3bblRiB zb1!$9PMPnKtXy{UrAt!X>^kQM(Dr=Xw>Mx=aFl;;b{oIVHCmQIqvre?UL9};btvAA z;bhUB;3@BSPvsp^_j0d_$`{)bSge5aMGGYxLTT{dyx%mJ3duJy*n21@l_I;S@YFyZ zxhmQi2*voo@I@Qe7B#bmov@vwQNGfrFgRY(7u+W}#?^uj&8#+Vp{QP-tlTWB@UrEj z9P3{{?#DaU&1|@j)ndbOeZ9MB@Aa6sIXY}2(L&?mZQd6DNZ~FIZ8C&O4$j-l-3#@$ zARde{k!GdV-Uu}X6$`wYUTEhnc6b%%#>>qG(JUy5Rr(&5>~!BaKUsO;wU+NZSXw1N z>a9hmKMM$rQv232HWij?a z@YB#|ExMm}4|^Q@`O;ZaXBk7Z>W$HrWY=?3YkSnhlggtQQ1Z>9!jrsu=pAGS$%asF zKIhiLYEG)MTa>=IPud%-)8sQv{JjuxD-#WXNO%RmQhEal_*Xz#|9Q{e;57>iRgJPv z@T-o598?vmdhE9Tu@*0rY6<~>RK&_Zi;Z|m<0rI|7_5{ToqevAMp69k#TEaoP{Leq3p^jG|%io|NO*|;rF{266DA?dv zBxw!-URelNU7w%lmglyElyG;bAMu(s_g+2r`rcRma_p7A#DD#hrKgvi*Ie~X6X&^Y zWAfZCs*2_xRPQ5Iq*-%$>7HLk{48}ToVoR1YJak6>BQs;$29^L-&eB(6Br;?4vXW) zrNX`&^5+>wOVhbH;5tc~MoLY9w2xx;QqV0Lo&g!TNNPxq!Guo8t<)>~xHZ0qyqlG= zyglj)KIG&4RtQVN1{H(ctNx9$K|uwbB3{QSpQY2B;dCf6B@vLb`$%xp#V($$ek>=5 zmI>hD>6bnyXxVoC%YU9o2;XrXcT03p81U($Yg8%XPDLHjvn^vUBYaR}HG&qqhHi2( zSkxnl55%-|vogw#0Dla#@ge>3<#zk9{CnP8o^_VFeX9LIIUpF56<{pVs)c<5UQOd>add*lF+DqmXiDYWBt&sf2^C99{m zGa2A!O)i+M!;#)QvKwZ`x^YJ_X*Z9}vx%2D@PuL^#ZaRf>4Dd#1_D_O%owolIT=z8 zLd|OgJA9k`?=H-cXcIN38LTC^BpT%Q1L;_faGgBK-Oiw4ffCy}(qqBGY~NSUa?7R~ zozzup{f?7i2ksVy)c+`6l}!}0fg;gVIPw)>>9ZCq6XO*{bCZ>J9!;}{Xxy}BjK{vx zzI)ltC~v3!@pgz25h;p)&mo7Nn*hlNCipo{F;F{ONrj_N5-F^ir6*d6S5$@ObM6T$ z=OCF>Bb^O3`Iu>Fqf-O1B({=7xo9!PfGkyoL0xnKu%7mUgyNv!j32nG{B*3_Br38GhCK%0%}@=}ng=8EVrYAK8S^Y7OelOp4W;6T#)FAAM6~ zp59?I|2Xc#o><6I?Q%8GFN$l?a?>S&c6WYwq!g2n0q;=LTp5NeDco$_=Vl zBB8))w?_eZUae>4oN|{;ZVvZE&_HlMXRG~E-V=q(L?H3BYnQxn@bCXzO&fi|75>24 zM3?*D0X_`WuF9n*uj_c!SXc*%Q5Q_o$8Aafvc;UNiH!wuU=P*;wNnT9F@u7OP^EuJ zeisCRq@A-cN)oZK(jx}Ke7R6_Fzl2tCqh`uK~lv==WZx}9TW_5lNnf0$B6IJ2RXfp zvV~BZ(4n~PSp&l_mRCJ{LNhjte-`Tp%i`V>IhB=5>9{j_;lCexEx?G6G_UjovY(x= z>$v|i^1cboD<}rC$%m=%wZfZFMRQtk*}GZO@bamzoY$nQ59!;gE9f|G*35Ma3`eo# zF^wepd?sBTR<16a*QF^_UJ;e5Yt?w&ZMbd~&v0eo>?&|<@kkfHK(TBYURAZYLz%v~ zRK0_AEVRk1!umZJF(FU+IT8GiT>E1xvX?djsQ-Ijr%B{knq>~0Glt^KQA_I06tjsU z8>sM8?w!)R&d4yBLhG321xp465yB5WMPxdO9_qVtIdcgexw1^$!rg6`zA4=}XT7GPkSAWcbv=$T=UjH%gX^WkfK{1v@ zSxdu`>Ul<>yz_2H8QJT=KsjLol*1GQ@|Oiv_(u?kf3IC}1mf^I4Hg#F(>FvYb(^!e zf*&D7F7_C4SCZa{Zk zydqM#-ZhuPaMnJ_QB@0w4|YmBgRo&HpL3DZP3KMut;)DY!iG=W(ydsHgySwUPoKW` zSM$`M!eCs$364P-6B%O3mL zL3ZZikFS|KC~N{N4(!ibaA!`*v5o{)Ku|fR(RtO3^2OlM7&^V~1Qe=csWpNa(2+or zNDRf{Up)_djQ=LrCRaPdj18{O$wpmT(e0w!nG zGaeL}440iYEE|`ZH9m>YdPS0f;RM61o4h{r#8^fO#8K_Hbg?{0R*k>q%k6f=^7xOj zsF@nXo%^lX#zoB|=5#bkcHl)#naQGNKgB?FcrO*cdfJt5UtSXT>Urn8&dAnw$#c!L z6uJ%(uPSxKG1hUCWSOL}OS3uAnOd@829ucg2u!^xh^q>mQwF{oC z0*v{5VYY1#)UW~!^_Tzt-FcZ2FsHut?Q3MS0|Vxu31D_p3`jrkL^_HJ`e5iDMKt`0 zR}}wp^;cm=siQIdU*Ug@Tl~utuOM0F^5u9!45u470Qaf3Gsn1Fmf8gwqegYyAQ^QJ zdwgUCk}1i@ujVZtXMS0VV9#Z7duV{w; zzx%3V+?$#$Upc)*r>O-EpuMh!bE4gboOaICV#o=v!B5yTXi#jA;r0fkskeF! z3bMVk=zjejlauAn=@uDWQig_bRBa<7#j+gq1B)>$w9v}`c!x?~TxQ@)*B z9&|^VI_Ea;q1&>od5PS#IW_}`$$;7bEhhJ~)ker$AA3!t7R7zH6W|QUBUED?6YH?U84Omp0z-rebnp9;0w# z6)4a#7B^6dj|z5m8N@rn=V1a3Cepsi+Z&ev&B^dLU4kzYmk@N$FG1pNzBK@K>{YxISg zLD@kdouJlfP71e6d+1f)F^C42E-Lc=>)VmkHA$B|lZ(6s2{R74U0(c9a`c=!+h-0&DVd+Qr-M^G#{Fsw$_h}${ApVp$16|$(X?7!DM>*i)v zU{c^^Zq3|bCyc$W3Pc8#uHzqw$apV8dhdgFEh$#!6K zQ)gmsN-3t8B860VuOb`B_u>_|Jhu20ib{i^&?H`wp+_-ub-|JBNSnDkHb2k!TZr#|IQMK#@+GjybQ!Bh;~`;;s(!jkYO)WEo*bxz#DRk@-1~OcexgHL|z!&DC(T`nP&(1nE4+2 zgcY1NmqWlMzaebZ3@j?sqP}Ck0_&I$0AXUhVp%;YT@)8u6u3;=3aVVW+$dokz0-N8 z8{YZ2Z;b!QuMifkaI5G7rp3R&4|D$AP=%&rkeUa(EpBSc{SVJMJo|#E zN*?DQ&ASDZSrNh;qG;Yl(KR&+^fdY&_fA&c2pxLaq@!PCibNWc70c9_RoV1)g|{*-Vj5RCtXFd74mQY{;pZvq_dM%$kP^ zQ`iPDX(sXU=vj@6$5%|whmzbgZE7PfPKCd>g%mtDxM(!NMGeJNQlyLuPxMKlcX_n& z3p~1LAmP~Rg{qy`BniRVTe56ULU4IdvF91Tbs@l(ovyCrbh{$0!k|Elf-PAySIJ<1 z8~w2g`ST(9)S)O)wbMgRcLJ{Z4>_USVU=%Em^MaS<+~P0z(CPMa-6qZPxpc9@W={1 z{z$deAR5DtShH)qkh}K%Kio6tPI$8J(s5^2OAHBJE>LFKP3vO^O|mP%jF1}G1ztE7 z=VsB(oVXd~K>(C=HGXK4rGRMJ8iHco!%q8zyBVa0Tglr^M|mS5xV>0jYygV!j{8%u zv+Q-%z0t1}8g0%W3#b1N*}%@`IIa&*fbw|MT(py7cHl1y1$q&}$m!+&iGuEcHPgBS zdKRt(Cgf(f8cD*Oj)m!qZv>^#D}WO7V<3tj{{ro)!|~)nw!nGv4))pe_Wp`(<`tP1 z3LkHO>$;?$Zdd3uEBs4Um1I9Gno~Kg3Y_O5XhH86t?*C#oSa$ z)$DWhMNmT7OUH6=hOF>E8(6}`FYQe z^jGbbexxY(?~vzsmr{LF{6^er<2QWex+I&kf5`{pb9A-_-=qQ15lZDW%IfH}IomjR zeZriNL7MDl$c^dN+Ng;G|H)V3se7z^PuKzfdxxaAI7WQ^;beb3SvgkMyaO+<(@Ze8 zg<@hUvXKf$o-pJS)UNU04_!osMBC$W!?Q(-6075PQmq^Uv(v0cirIyB-n5LAR}){F z7Gy+)$a%>ga?pXT&RG)#R8dSh1x|1HmaplR-XU5nONaDj(cGfBHM44FK{0qI-Qtf; zCRy|;`ECvDsNg2JHqxg_9`w^!_@5@Iz1$TL71j#TnFGMYu?phP)w8d<0E?WVkbLB( z7_^4#v*>EbR{>oebW>j?)*28?qia1}o@y?hJZR|)t=w-JE@RB7Q~uV-`|p1>Lg?~O z3LlVWc8;(Euc&%V0M|h=trWRNg*Wm)mK^p=<=yb^<02cOUe>z~q#m;7web)8LA#2c zPENZoaL3#(P^OCxMRvJv`kfWR9C6|N`vkJO@&?aD>7W9I9|m13$wgJBb763?CrW0n zb<-lgfT#?zB@xsL;mwef&I9C*?8^LnIQ zB2m*wrpfqhr$4xL>LNU-iolk`us&TG*o|2_O zo&alxkE-&a>@O-%TcO&`MEb--gY0p6qa5ZSNDGh@afN7;SNwFnhCGFIbJ^;Wu#;cD z4>HYYNPxHw)#a&71^zL7wk70QLHfxT-E@x)5zY$IQ+^mZ=S_1a=I5kE#R6|LddOFy z6AnEe?H=waP*$8|Aw|PrTZJDR;=%I+Q5|>*HPcr5t)7Ae+Us@BIoqv()4|ydS*sJXFG_7UFWCrC zBOtRHFkf_{Wk9X}%;(SMVLvuAkORZYLXa@}wd>Hbu#$H(Bu-Jm-Q|(R>0eMquOoLo z8u>L48cg&#u%uj^Doj&XLwLLWWzhMCGivxpRmtk}OEAHZFe7e8TG%ycStxSt6J*gR zAYFxl@T1-X!6zU*-r_tEoFKq^rFiWZ)h+4vc@Xq5zlYus)*KY;{HeNnMvA9Slg(M> zcU$qHXCC(>vUXabv=J6%1A-NtE{);y70&8;QPZr&rj3y|f1lG?CjS;`UvRU(e!aVojt)hJ z@6|7E@n4RPGsJ+Pq%bUfZk+!Eu0H#%j|FM(I%8OhI2`T->f}G z56LS@4(S%9!8a_gaUGBzgT?$8#Um#jM33S1bTnBuU&r&@J~Xe$aoBb|3pF+vJ;q$y zuoIH=p~urn=lh^&HTqE28i1!iLFcDPvfr*Jprx-{zBxAYrChF~E>g^<|=+|3+)brKeZ(NuBqg^qQR_grjUv^OEmqx#7 zI9?^qpfkB2lERnY2LZ|2SFYPv63c3sJbuML|6mh5Shxu*m{8wqU-P!g2$Oj!|A-@* z>|)0b?ETc5fU1~c3Mq1s3LhXly?TS&$P z6M@<1fWuD~@G{QbZ~pF$TD}oDOD}FaM%Fp7eF44MQS~&56tj&YTOnHxK*0K6qzHKE zdsv_YBHx29U22H%C9Bibuu4BAzrw)+y)W_`#`zGRyo~j5KY0tg>|te9rtbW1*sW*X zDM^J$4Qi}4K}NDwkp-0SmqZo*UGyGqk}%%t(gj-$hSj*R)fZZMA9lEKxtsQ~c?!v4 zomC6DEv)A)_H4#fbh3JXaGg9FQhlh^n++t>vAj=RdV?z`oS3%q>FZCk8XC5r1fKe9 zvV+FT<3CCNnsGk){rSZ2kz>z|BLX?=QM1nn6ayvYXVAO7JEK8%$U6brOE6LgA|iq& zw^C}gx`jd4J!fxxSh?~8)JD>TO(?L7zq-8wM`{=RxNS zkN%&|mjOhq@Ij@y9RGoN+lmFP%oZtzxeWjRRImehA@quOki7wVN{nM-aolm~Cy*Y- zCPQuctnyj;GxKL^52+5RD)#&}*p2bl*y61nw^rs& z0kzS5;mjuAbSYl(f1BQ9^g`!wKvzAvy?^0EW+UcXOq`jQ1jN#ZLzikP0 zEHQ5BHa6d%R`y5bbfdp=>(^KRPNE&yU&%7@S9VYg%!*s6@B$$Fz&uZ~5&|eK(yJnn zqc4@*R%Fe4D9D#*s;g%|^3>J_)&?%0EZtX@qdjTsBz6|<1`9E_Q2d2cCLB=X2B=JC9Xf!{>WaRR(KdLIn) z#|Fo2`4xigx+!#>PmcI7-NsL$x1uIs7Co`HYs-UY8@yP)|9FoK&&w|Qxe+hlzV6&h zt~l^)G-!f=k16Iuirl2ayRZmdm_>ggTCZqOZiF6^LQy_fo8y+KHy8UTubh4KOEuv4?LS*h0! z1*GBn`G^k+Ny3+9s8LQb#|1jP;}y-ye2*fCwG4nlb?ibc3XCDQf--?i0FUZCSLyq| zwWmd`q}UU?WU`>C!SHo>ud$sZ+W^Qq8)WgUF;K>8f}3*o%jJzeM!40zyt$O5Pa`!Z zqgq5U`4riY$`cRh19Vep|4W^-3~%n`#X3zh#~`wpLOv0|ks{4ldWH0lU0``OxSU^f zk=EiS&S~3TjD+Ih<+#vVo}o?>Li-(l>b2iyyY``?k@fe7Las6LQjlmtb8c zw(b(eT%gDYumqzk_#2r4==E<_A`fG{BHsh#n7U=rvWSKKq(U?#KMKA#u8R;rP{nCz z=)T}~GUOw{`9)>kIrD&=kwgN%j;gSaTZ?^DvAj|sDQ#B(eKoxz40d*slObz`yJuI^*co0P1QDSlLD-~e z74qzZt*1lK3TRDXhb{Hxj9|GDTPZVt`v(&5z}SM~%~1iK42s!JkrXQYHm?U9(HvnP zsQMi9JUkOpWVby}lhv@=E1z}BGi#A{uj>^KaHR+0@)}f6^$ZiR5{H^Uaqrwx<=a*b zxHS)djB_T0lQHgRt6i*cGFA4P{7=6&;$-ThAI&A5WBI@i92{I}Lht`6#Q?3uCscR} z1ZdvFR+3mAw31+`d4;e>)hfEF=#eFRZ488fA+DP${C7G(2tbw_ZOz=B&KOA87@8D@ zhwjL-=GCa0UG=>v75-RBiih-<^#ylA2g)bGg+cnh7_8j688Yl-{8Eq+XeCE{kSGEf zPQZE%x|T0a0n-P+O0nmlYb+0s6$X#+RKU(kUIUbE?C?Yqg{|C5630#U(r)E$a9bC$ zn}gA+R@l`||DyuxlZWIxIGwZbcZ~y96_6Hq28sROUGPcjSBUcfO`{v>`PSXlD0-350F*}rHi_Xo;=A0mD zf`g)!^ZG;2*}}aZVEMw-}>-k@sYqUB2Kb5T^yV(U39XJwHWn!TU zDCPh~a;flKZV!Dlv=g+wE2VYfEzXU6jQK#F20u}-f!V3*)pMD2aH18B@^k)|fD8~# zNNQk(f0EZ8?vNrv_z*~!@XUOlwtxY0GvLzzJkJh(6>B&>VOVg7?Ge>-`@65dU<8f# z@5+~v?G8Lq6`R0dFU4d~WH$t)6dx;6ywR!ICoJ$dNAFf^Z*q4kuS+&ERdgdC*ZVP? zHrM^28gfstpDvQudM0v;!PVa6b(!gfrPe(j%L#?$I812pS$-Fr(=FSQv0_zit#@_L zdS?tabU)x;AO@DB2izQCyaEdsEBJTm|Lcqy!44~2OsVZu`kJ#9KPOdz7D$(|#^Mpo z71|5_oze_-CyhNcn683_QM|$sGs@<`Iw?myK+d?r3hBLG{<@Gj(j;q=6@pk9(Ei18 z8tJ0BK);V^DWKW+O%mFA+L@5q89o1u;KD{fKUu2tG3UHt0~iO+dRkz^IjZU>cj@h% zQkA}I;WD>JP)RT1b>Ijhik8gdtxA#))hNBAbC{1V}al&#yl`*Q78yEZgFI zFOh_2ETRLHeWRj*nG^%e5~+rAuWAm)`8pK9B!Z4u+dM~?a8r~hx7bE& zYxw&@SHHN*dvLn;fUsVaqTDfe*a<7*p#@J`=TkR77MvS&TcWJXG_eID3)qj$DNp=l z`<>RDzqRG<{pL$#Htvc8XBR9is4@Q9B8?CN|J5KDjiR=hbHk$-Li!Ni7o>_K!iJqN zWE&|=Mc3m_Kque_IghS+BM;Ju%#;1B<^-W)&8Uy-oQj#Z+QUYoDJ$3SBZtRwP8>MU z23hY>&PhGR0ON2C72ZF8*Nj_|F8)Q%JwgBcvwl!}M-NELXF)$V=aFZIR|M$KoC=M1 z$&xn8u1jJ#v7x8c1EA`CcE(w9U2=CJgkU^Up?-BQCw0zF=gJp0x~^CB03>RHwn3 z<&pdElc3th!?5ldn9{1C(}A?2jucMcGQDeJjCx?Gc_g z$B$)uF@_6Hee_M9(7Yw{Iq_U9G;kWapP&c}i&P(Q_KULUXhqFja8U}OObfHs5yA^I z4!$r65Hb!OOC*^OxZSqbef_TooQ(#g?n_ngkyY%ng$}$5g7T_SPRv${iK9pi)Qvk= z@i0gWZI-@Sbe(UDdycr+6C{(>_Rao`IfT~2V$6%}xO3@^S-IwQzido~1H;0C0#uV* z8^2rD8+_ZXLovwhhhjW!4;cs^5PU4qcFAL%o7`(9+T1{ByCoN96e=D{jJ5CUfZ-RE zw^ik{urn?N7T49BC}F0;SnUJ?K2Uw^Ql&obwtq>!Gjx=@Jg(=%1(II-XF_JkS_qBy zRjvIz5roZt;cI^iHCmQ!S@Kj;?7*%`vx%)ZK{3ZDQbmQ|qq}8gzMXRIKybIL8ITdJ zNK$O#UH8(in$gR}R?JOat2m#^yA|<@kDx;8y>@8shQJ|)_s@P4K8~Drg>)#ce_z|H z*vPa<&nQ!!2SL!V8s;KxgkOP21sA8JTKPr>meFy*_>AGkLIO079|Wz_3`p0FGxBIp zXc!?hE`+dhU#7h8|K^aJ5lG*!-gS{|dB#@L(C|5GBHBYSsTA2og%7`cU9y5|R~%dN zDNxV%atk&-H^+$vHbXT| zHnVBX%aaN@+qSq{AP;TM3D`s?S)mPjn2txBddo&oZQOF z`EmW;mM*gLIa|8iZu>1i=P?!-jyu^0i2S~39z98*h|8bCfAzcY-~ajNzghMt$r_4T zNs$OU6jiYs9ZQngTYT;=)J7~E{?ahaEq%fL!B<5`LNjPXBM7k5K-Ddl#%f0zRVU<3 zSo(c8XU=m)j5v$Y+d{uU#?BDx|v?%V3$q38*HC;|L*D=tZqk8v{0-*{jH z7@7zoth&}RY-%(=tDi{L$AN?V7P3A_2ZHS)8NxlTr+IrMr+Kwd@Sm+d6tu@xr)luK zCdJy=eZt*LwlJ3y=bA6d5q~CK9hM-y8L~~dQ`N1gwmkvG=usRei1fB-^nGmo**_#U z);yHru#=62;0Ufmj&U!{sF!Qw6^XpnVFP3CWs~Tw*>T>;~HWYWrd09Kxu4LtL75Re%emE@+W|>o_81g+o)aGzI}8f`UTmV8jysv(KK%XPKkKpt zz4Fzb>2##$J>If%m-m4{c(rE_sh<5Y=$Oa7uufCp*TBERn$~?Hs9}ZKiF^^uYuRD; z-P}K{e%7IJ)U#c28kl!KgGeh%gzB3}dlfO9Rt3^gchZxT4}LsQ&jmE5w$*&_BXc5W zhxJ%3P&4;K*gcNwT@v-3BkEpm8;O{y#Yp@au13H0msiO$68GF>hM{eAR6stBVv;Ffxe7n4ekh4otOsXEyJh-S zmw1J~J$=3FarI``hZ5{tNA_GH&Ddo<6^WeLlsleQC#Q zMvN4?Km0kVn?^cK9_~eof!taX75>hU6V{VP2!Ywb2{}uVj$+6ueO?cFD8VB8R+7zu zq-lq~wipj055=(4yF*Sm0xi<@u3F^AY?8I{6TDL;+7$1@Bs%n-q)P^ko+zlABR;AI zCJJDY_#328;BIN$Zkd6-K%44Y!71}y11hXqBrxcvk1yHf@kvlN=RrU;Oe49_8(KL% zYxYs^hY~BR09d}{pD!*Z@X4A%Wc89eZfJ9L6!Q~-5m}|LzIKkRci>fXwh82uC}uka z@Brov>?6S}e-lvRHY+Q*KmdNjBL$dpla&YLCG)Wd(4Y@&cMCEC9k#F?xz64{*}?X` z#^wd)yiaTvat>S-Yk?(7r|jY9fr4K|Sh43#ZUtQ^LIS89VH3Y_wobX7M1K;|(i}8v!ZoRw^)=(%vZ2yjCwy1h+1MLi}12Ml0F)xp`P=&hb)mvZ9c=ORuFvNvH zt!!_dCdq5_tGEvYz6sP1`r<0%lb@}BIpP1My?J}dEfDbT44Q;DcYIIR&W{#u&+46Z z+}O!~{dj4GhjFTD{zdU}vV)zY>%hjO#AIU1rWjC!+(U)8dt`G`c^KHg&pi?X(L({$ z9X;UI&4;Qb@PXT)IyOSs$=~mOc`>?ES6vS2iQ-2lD{S$>Ww#xnKR(Ehm=*g$fBV^O z|M{#JwVhIL3~ZOT%g?FKskEuiw+L37o%FsW(l)ta1sZZv#z4w1o!+F#2uu?ULJ{1c zqTK^e1^to=8rf3Nb@_0CHd}qg_Y{<|6$ZCMFFF*|WT-Pd1{I0?6RH>>wM4a9t>I(V zyi@KO>Jsl--X&2SZ*cq#>Jx*;2t+LZ94~lSS;HykfB3^pvC-(H@e^7}%yVOOa!rg* zGQ|La^mZyd#vfZqw~9*_-I6_;pZ$^+TL)XD=%8qEJ=g+j7Ab!E6<}#-t>l*MNT3}J z%;QlW7n~knZNCriyigc1%?J=p=O49X1v`K^?$QoOPe+Ge;wUDDBI~K}tzH|MHZBs{ zLRCtR5Hz>2r!ZML87MFvM=LR5deJ0oeS1l2hB+oIJj_lI41z*-w>l1#BO~|XfV9oa z{tJOI9)fibbBqh^yo;5+cHFn=(!MX{n)6$*SqeL@Uv7z}BR()yk`laIjb#YQO4P}T z5JEe;^L|hnuBWTylWZ@n!#DaHb_TTB?Y|lEZU2A~55p@xf$V2DfjBTQ-Zz;*DkugB zFAh`TRs8tC2ZAEWRxlKcJ8N5DAR%o!qQh52+n5OEcIf?KQJ<+UeJr^)85owhdc%KS8S)U-A72)`Kd zfZ)pfd|Nf3S`Q%0POIq=GbD_2vuW-(J~Jmsd6H<>aV;Y#kB{~<9xLi41BL%SVI)%y zLzx49W~)C0WmgCbgDd>8fy=->jZ&Xo0eC`$uwY(iP)V4ci7AzXyONcdgszgFqkH8W zWcDeEvjr-~Iq*-t%<{A;7rppfD?2ds>Ir(LJpA7$MB6T^gJr0lg9>B*yR3fBV_t2) zotB^T7)6GT`x34?p^=*Nv^i{vfdx*s{BLQCU)SDvUEB6;?W*sTyXam6<{K3-ukGRH z2FEL|c_)F{D4Y#NDWJ{a`p^x6X$isQL50DUAQ-a6N4rg&BW_kkx!9?E)@lTsP7-EY zwns861W$R%wR;xF=&tYoc04dOu{GVf5dk)qI3>I>2akgQ8qUXyfA zjQQy4kJ-XeD=mwKtL?XY$9wk*{fxHa)*H7ANH#lL;lLT2Iulz_N-@x-Q%HqlApyoq zk^d%Mk>u6CparBZG5K4lYLsbF5&97_+u7yU^^2-lG@<{$zEujt^8)+~xAk`F8LBT>eymxVfXNROw zI`Xqw6Tei?R}o1;i7a=Mmrmc>(Uko?$Wn;C*1Jtd-O`u$Ilm^AV6=yiDf2aLdJ^}YmA6BSa|w5mfhRK zl3%^>&s9)>^stGXwj~i^?<*7C*Dj3ZRWFSQ>zp+}YDr&kpCE;9Rd=}{ODHsWK&{eM zpuk7Hg=S^4x|@FJmKW5m=&~!5AA5T<3Pn!`!i0{-NK5|xzl`=KDmrEn>7GW`n~1$V zqL@L7+^53veM|B>=iMPsWnwrWUI8<|PSeA?yttZ|`{E-I;fhymagOrske?8gaFLe? zgIR|2>p(QJUXE-c$;vCfn}f@IFyMKaS>b*Y79PpUi=r+$-ZNiO&U}8)WaSp;ZT`6b z3fT_K;#qVoFFx=QZ?7xvtA~4~12ZgEMd>s(>>5 zP21JQveTYPz=b8EkI{s#|Y)XU%A@pj=a94R`6C&;?#AHLdgpO&Y8 zG!gQ*->r`}C&aM8JCQ;(iugbsa3~_>-j0ZPK@T)+>jsviM`8AQ$1FZJ>z=-&u^5V=VdfM zTi-bMAxUKy5^`dtvd2WJWK#^plQrPc4I6M@NmjU4fm*2}wY;^SXTTj)BN>!!<2Lcn zNYBXz{VWz^Ud)zPjO;{!$R|vo4j?Q8U>0C zC?@CXWN2~f3rTW;xS)#LMQ;kcLsxPgh!ShC@x=ag3@-3nQVyIw@S;1<&-!ZJYUw_5 z*avu&qKmonwW?H3sc^UVIqxInC@FV$;5@@4Wii!`n`iu8Y}5etLCq5CO`|oMusROLuI`ERg8M@Tbf_$=H8vBuKs5J+%&5DUCnpezOmx% z_lL)cUSFBFSARcCW8B31Bao$Sj&j2C zF7p6FFv!c+#62%O==#amFmdD%rx=o{3!_VdHb@L57Q;tcRCsL?r!1g>zUf&JdF}OF z`Va@N#eCTYdXIN~a8+m{sTbiLss6d(YO4xe?Xw5I33@rtd?>yfxp`VXC)pLx=K?3( zfP6)?);Q{ANRHDev*{e}R>?+kM~HW~aF7o@kG{jvs>)>eE7Se2&{vGVzM7bR*BGnwuC6w^nM zUMlw3%u0o!N27X^s*$OilFYm2i@X!yzfdj@Ug4=xqzl%C>K$LWiy;8uoQ6x)cx!n? zrvi66fb0gZgO;-%$-LW=n~{bK@fvkKv}wYN@|CAV38Ed4a-T;R3Tk+H(~}|fUfnHS zA?=E&o1$)&SB36%!|}7l$7ULPNuKpQ&*>+7JrDErUwSa;qJ{dR^CdmmP+ zYDtk3i@@V1YO;c2$|s>eatR-cnSB}YiMUdQrb;}+6lfKvZ z55m^*7I9X|uY*0FA=<#?09QX+I5m7AHzjNQkz})$mqRy6d*yp&rHWEv2Op1R1P-__ zo?S+2`1QdjNV;#Qyoie}$!xY7v=Tm6o7T3%4?Fm5F5J8BTSlEUvG+G$kouQqp?B3p zgPoz6(-b*H#a2(PWbR7Jsm-$%3vNKl<`KWAkF==XU-$j|A1(b>^}M)WFPZy3=9X(! zi@$+g2s8>LW5%NVjnjt02Drt-L6;q*lT-=2Ve7WrQO;HG0U#i&66Q~94Zv&htVXd* zc`m4&TF=d7R?M!Et@dhB70x^P({Zbz?Bt1 zyMCoiH?O6BI!@ulK`RTb_qt@A47LxtrkZ~@N+-kB-VJifrzncuw6S`Y7B04O>Mh)D zzr&UV(DgGR9==9lqni_UnWVoo!UnA45s5Q}6a#GUIaKU%BNBxV6hy5mJJk>Mv=8tET-&s@KgPfqp-}c z7%*|ur+ikPK%MS&$Om0I7}Id1pkW0KTg|%_<3zM|Of(O3I_*-(Lf{gLSRrh=5J~t~ zdai+h6~6V3oZR8(Y}F(WU+ln7%X(y#T-^z=KYH+e^Smmjg@lDnDU26%|5D@M5?aL5 zc@)z%bT7B!)q!8%{EybqJt2<9hE=#gt>8dR;8(H21@()FN8aWwVqPdi)I!@BTq$c+ z=Y{u313kZyx5E?H7R!Sd3y%2B*C_hk>p4jA)FMwCLkww*6KXYJjHbjGzlQZ~mI1T+ zN=Wq_qe0s8vn4sC+=)H1=S+--V-$0QB8RBh5=rHhYEI>p4t@r$i%|E6wTJDT-3pEJ zmV`j75RKw6NqB8zu)1zq8<5prn6raBDPAPvY#&knA7H>VZCmL;T1ITjo}Yldxyv1fe?`*W{(DbAZ)vTPPPPK+80 zc5fRvX!rUgbDHQSQjLc?3Bsw}_Rr@XWoE2`#3=XK`%vt@O09X-85>AA@w(SS@f;S7 zsCPtY6}vfYLOp?cBgvYI5J5HLV7u2USm;%ay@6mlK7ZLOWfG%)5#AAgK$2e?{Q^Cj zN4P^WDP{{rHbH>^kU~L@MR+?bj_aniO0J0$RcZ_^W8gR+QXgtLeVimtl4wA1QLs!< zO)_2kCAL-^rjuaUvmPQNO_I?@vhA^gwBnK%Jwkj`+((eiDMM5(#6lrGCxW55o`VH} z^|_q3oyF&mW_dv~if3e^;#|o)=A`>h>lv}o4WKHhhinh;l|v%a_V5gPPY~pgv5rsc?r^32>!j_Rs*g zFRE4YWdwHXtWefC(zdh`97e5RCnStAEIV)n|Id)ucmgj@ye6>_cqx(TB=Le%un=jI z-jnIz6EfR6~89iZ<<*(4Kfl(`x8kh20QY|`KFnRhj1vo>*JXjtfhij++G zV)a8$jS~Bw?uY;j6jFKJPeH+An*25&ppr+U53D4j%Tc0+9Xe3jj0!TFANu=y*Y$N52GScDtiy)O;$QKf~sCIA))f zBAM{dRR_Y%rGaCgX7W(kBe9JdO1{v?cC z-Z|(UgL!5*2~PV4e)#_uh?=r2vm8WX?AdD7tXWi);EV)*##T4#(_C zfBwg5=Jlu+I1`OTGdke(j1%blFQ&;;qVfZ>resagC^WMj951)#i4IE($2?2Jx3H2G z&g(h8y12dZMbGXysH_XB4leQopBC128H!*Q;1NUPAV+vEpFr>B0BratkCu;zsoHrF zPb{>)M#=Wc*God64MGtpkxu!US?6dd_p6BP(05PR!?^@0>bu;!l+i95=}mzfYhd2xBfLo{0HOeWZsT?U-CJ-0E!bQ zt1maHU>%^Cdlb1##SX|jBiDr@LtDE4HtsrC-0zIM7l}C{+r)cA3IYs83P+<`R5z*4 z-}e8s{kJ#Z-!rQGX(`Y^`!03;x1ay^&jZv__uG;IdAe5$?~~V#25D8hxSjI5k=wW! zpE<-F@-dtnbZI0S{{i_fj~ZU;)MLDh)9{W#mwLZE$q|ymJIpie_eIw8@Cw7;#c5sq zExZy5?qvjG>8nUzXlGU!qr+~zsx(gs~kE8Vc}B_mJ={Y~+DWC4Wq5K#}EAzrOY ze|3v@m#jVX$*etoOt#nF%NvnFhMp>%tt&PrM1jByV~p4|=P&N}#|q zaDT|&StrF93bY><#x(yURfm11pPe6Q;=4y4{?vR4#m05-#HO-^bb5W0*219Th-%Kx z5cNS|xVR>-<3XxXGcLZM_`(sw6HJqNKzkMC?1$xL^%!~Y_I`>#Q~@=e1&Ga`lh+ea9Ryx zK?>;TR!N2^N1X5)B*bHaaR=Qd?DiOPsS7F)JPLEH*X8A)hnv`UC9uJqh1O|tVS#Zr zN7*Yc6BR{ek+r;=uPlEJT1|G!Q$@YvZQN%5-N^YRa&>0-{P~c4sNBhIv!lOcfYVhl=1$k_>1Cs7{k7 z@sc>m#|I__8WbQ<^gBtMLtFreu+nX-KR9QDhFq>O*CP^w;|1-c&O?ojV^(`UxffHA zQb=7;v&}42R%jTxbg{HO+SHEW7(2Og{rdl9%)^6hmM~6is#=J0qFliYq*e*!AnghK zRB_e2LQ(36Z0Sf8Xumb@=yS;eMxzg4r=veoHLQC++l&*d6$@-L`I1M{^PcOxy6IK& zgF%(h2w{-dB)elrG7yUsWZl$TyxndmEK87%hQiJleer@G(JDHXTyyXA-c5fM)+AFW z1n1E0B%gE4E7{RFunq&LGVC_pqYi8t1EbVKOelz2{9C!v0{J8*_&izv(!>EvO#(N$ z6a#DGOe(eyxB~uuRSwNn=qA3#Kht#yXE$BN-Qs;Wssy5WMUoWx2A`tvJ)9bPU%+9W zhFcN@InZH)E<2%9?!lnysrL1RSP!eGtdUXSV5@_F{!{74R;DlR@jgc{4qVL9O+kL| zB6=scMD)3Pwo(Tqz>V~m5w|1n0@H6N*oO9flq3IimXCGht#;jJ^|m^1(0%hqx8lu9 zXr3k(ablapLRn2+&}!f`#L~MV7ekKHkPCJo-UMmyU181AWrCY>jnbi(Cp#qD6}%>c zTSoG{|4ja_zfs2&4gG8{$#mjyf0c>W*hMirC{lpo{=TSUVI{9hUt6VB#f7B%-wi}o zvWuM3fRm&)8p8QBGS}h$KH#oHvji=clAt^~UwJx6qsW`y7X=*jz;{e%U>j~$)gX;Dm~5xsn5*on}M#CR>?_Ujbd%+Wq)<@^zvC9bf;XO6nAExMsa3V zaRf4xcgsOoDN>@E^gJWBK%XtZ_7ntd0Jv{22!CK+`^^SEPOPXbaET)!qP}UIo__{Y zjPj=6^wDxR`Lz1lVy4+liq)uKGpG%&{ZDf2cA!xYHOyOEPO_a?57n9Ip;C(3O_5?M zcGEvqek13H36l(;ODCntF}W;9kni0sy(mQDv%S$f#LJlue#M)Y#ecmjZzMNnXcWoh zJgj(flmoMItAn5E)-$cl54VrLvhQ`BWK&>y%spajpjZurr(}TH0UJ#I$RXJ2qW89h zb^K!`pJAs8w01F*NS|z`RSrf%JIdRxm&AiWmxf-K7DX~7#b(eC=mG9B4&GsF^o*P_ zRd}yo|T@%v<#4(K7GYn>S+K&!+POw5kDanYahIFq`;y zX-qS%;Ae<2`L~0_y*lVpH?u!#&;>gzVY;bJCS+`n-0b}|+qK2hXZH>}ZQWl;A^*rk zBXX833#lhVPFy7nOhhB9gx65aYKp8dFg2G%rTW(h&IR>FXZT-pcW@hPR?@*1;A1xi zE2vIP`hG*A%!r7LSAYLUlJ?SgWJ^r&lS45;ftEqV?&Q|chBgwLBrOsB1YEQH2D(YO zOIc3cXIAr83y#2gy4mv~!EXaN@LpVQC!rAj+acZoB13%_ymE)h zCaJoJdu%4=3Z%&g=R zRBRVr7j(&c6{k8liPH^0QRfQ`705t(gmoBoLHWq#EN>tkG?KGA$k#s>fUuu#_IFJF z+vL?!BS4OS_j{MfT6O?&-Vhc69_JC(?N*8b`n}Cm?AkZAdUp9_?zNfSK5af(BIwn} z%i=bZ>ogW;V(7RKY?N*~lT*Yg@LL7SX#G_C+a)k}V*!aH(h&y!O)}qnQNzal>cl$8 z!V(AzYvKivmX_jC<&Q~e4iXdQFR(R;Fu%tU``_Fim}!m(r%iOVFvpl9ucMKDsR8sr z?VOb}9c^@(ioQGT5!ChnwTkYZ z(WR_(E1O$ICorc3d2Y3KUz?iFffc|QwjWJzwDTS-U{H@iA;fu&`OLq5??(zFLR5|KRgf%pYgy<0%Sg2e(05Y|bd@MX zS6LIC8NS?2eVvm!t12{=TTL3Kc1}ZYYF#k?i|IjHk_h$SC==KCq=)PK(&(o26Vy^C zgH8wG4XC9Im*8o8h|MuTf24LA-nKdTslm3I@gxH{vH!zDSQ95YB z$^OZX@)+!Jh?U-P-Zw4%|4Cjo*E>%q%RBGyx1?QaBuU@_?cit6$r7FNx&L}aWIw^) zRXNI{s0`70$UW_ro(dS0X;sOto5Qb$HPAUUboGW};57OCr0?8~N%&#OcjvdLwvmKM zV3_+Ir%UOjlg_AYk9zh1A66(PpEifMWq>DUr{^$6TgClB-v#msJ6pwx(Ntz)tK?A( zcw4fl*g|?R^8Lbjx|ro}e;L#D?ZF>?6;u1}`#m(~HP%W}cv@8_tfrHx?zgu5=F1q3 z0wrdV1n=tLABJl%@zM6kc{1hW!qAgpY`%4T_p%?D^G!IdW6c7CL>bxqY8lxDc1NWc z*;(^wjdB^(UQ~tdfjWxBaCN)9Ku{lL&*?Rr7k0C4cC#IKElZ41$|@!n87pwno>*U9$37P(~qmxN-Ls@sGh){K#gLl zf3YLVBSuc8^}rZ;m;JX{0b}A@f876f^I)I_^4z7WTA#%s3BHYx#1$vVQr7xlwL}AL zI9@_u65sRf<(?&*x)@25A9IDidZnOA^rYTjTf$&ktG*1YdDb?&INP6rD>jII#xOQN#R{<#gcAmX z{%Ev9{;T;vrjsw6*a}%`(!b&%#oVVz9~FyLT^hx0@7uBlPKSRUebpPQzy{obb1!c? zW+j!=n`W1hV&Ps%R~V)uW-Awme8NG?ddQ`N-z+TxMw=#n5~o8^15wRHMvWK63o@Xg zM4X_LqsFCC6)bV~c<%{J2HX4$bCiF_qmwBR9pZMwqDcq2uT`O#(12qd4BA1u>G;W~ zL~#Ngf5R;GR-X*9akC@}%guYa*SNj%+pb#ZxuT;h-I^pNerl}jz}uJkpV$Tsp9AgBO&E=Q* zfPw$)wtCjEfA_1SX=I-hOQ!b#XIZTmStmEn@l~ZduJ@i$GYaLQ*l;zR2yq1U_ z$d?|JVP)4%?pO4BQ9QXrXHUD!EsWSiR?JQy`=>4rxf#;RF zv_xoCee`8<>LssGIBy$;ArG+cHj1o zY5d09Mj7+LnN2?;huFoCoHy5tw3|qz&nTveA}5AR!&(&GKzl#n-mU~Doh#l~yi1~L zIel&mFsiV||GrNRz1<^C-U%Huo9F>|;Ax^y3pWOr3Oo4f-EO#sY(6MQ}HLjT<*xzb32VVp3F#BA2d) z+I>Tfz!h;GU7_p&3VqBVJTJsXJz4GtrfC$pkY{C&bg7XuHV&YS9LDC`oPYZIA73?! zpYIp``gdfr6Pq{(Oaw|1#Xzg>Tq-t?J`kaP!b1(~@_c0ueVWb$E);bJUBJZ>!E#tE zL5Cfj`(8N`mxCZ6w`p~gY;hpmftE2g9qa(1R3Km{1gx5raR6Y1RIuu+SY5=*j+waZ z)qh!JzG!mVIJSjlP(1S)@5B12GqdvC{u{SjR9e*;RRfJ|EgAmXJPfA>U9OA!z4nJ3 zRUD;HdnP#|dv0tHVKv)hx|kJ2sPg6wY34L2y^xV7{rPI;*ib}VA;6CV-@0t#tBYqdOF?nrWsE#fpEuJ^! zZDK7~eUrB?6ry?p#f~A1Bgw)h#rd z9~XX-5J~Q`^S(NAYhByu0r{oqCjM>jQdz%vtLH`~u8o0T|DDU?4t@ve;ho|43bm?5UiBOt zZ*kzA$#^7_cUiuQvo3IVU~+H=f7zsCWREOPa4>rLq(U0a2EdD_A;bdk%TGeQ=kP~FJvfbSk032u5y#37L@Z7!`oDz4|XO5(sM(%FbA**f{gPP$q* zrY)}O{Mr`Dc=&RiSKwOW%PkP?^=S&#C^4G~b#65{xzYveLiKK}>(T>I7_~2`DySi_ zj3mOkt4)|D*O%@)pqiTyX(JC7QXMe)w~o|8NA4=&-jF`| z{MLY4&Y(*p?=Y_}7~6E?!7Sxr-o@Y^Qp?E*ED0J48@EyQ%*SdtIi^XsV%+gw!b-(Z zhZdaIc^Fm9f3%VLB*lr%o1G>~Ba31*6iJ8ta2d9YYLzS%>ty=eIkZBsJ^<_&An$*5 ziBw;;kBt1-`VjM@p!wn1h$1_(V2%SLC?CcPg6G`N3L_KS4$*by95GHSMl7(tpgCF* ziHULQLtLGto}0k*aw{USU36V=f!}@DS|>Os>m=&FH^AQXSR^>i)A7~Wz|Oxyn#5T) z{n&I{xKb?6%b28z12hRcq+C1m-+wW0*X*=NvCx`%J6$H;IK5Lg=u#QFIp~OAK6lWi zY1R=zKDW~Cn&%~m1?141ygJ<-p<|42VFwP27_<2aR^Xtfe!8&UoU7deB|0uEF{@qO zMe9IXw+XxL)71ERPsExB|GY8mah}zKcV0jF#qXV{a5ows<=_0yX|ie(gkeS|D5X=( z1`1{&7Socju5^IgBrT_r$uOU@+NYMc$|G5{3`!<0aO%MZILApEzeCOH39|+rk43cI zb^4Q^%?cgP8~*)6;LhKhqr+*Hi3QlAM>$QNqr4S}%wifv!|TVs4jwr4zIDo1$m^hV zQT(ud0I#z$W_({r-+t>JoH=_+R51s3CQau1$L)+Ws9oEg7NYE+h?O{ z^V`Bd8Fh?E@{4G4+ld#rNhbdLxQWaoiVRS(?aUI+c8@0MqW}yjB{EAwDtSFX9Ces? zH>xVMk+)4;D%&2YRXyMo3L1HvU7@rY+V!G~&?RRq+dvNvqN{ccnoTyrGi zOZsM*R&|qi#IuN7K(>jKIBPw>q`}nXcPV$uDkJjT4EtZnwuD>?xF*kYYYXj<9(37C z-}Sizl!@EKwZ=PElHn`9q)!n1*44}|aM7$JT|PH`3ViX(Rl(cDc(y3K;f+C;9O9%@y zHw35i#X}Jo4TMCC0q)9~_%Co)hQ;)@5o}h?7%Lz-%#xlkAvWi_uX*vj)2^K? zl*12robu5-Y>|C;3>LR2PJ>$OCnr3xQaVS8 zwbGSt_Cx}^sW%fXR)K457qOBd69WI|{J%{y+9oR9hibCOiOpkZoI0X-Ernu|D6$HR z*V5)1`d$3}D_OcfHn(UbD<<1>@-0k_b(pYlvmFn!!UXkkVH+`L`f*y}VWBCTj@FSJ z$RH{Njuxb!YYkNw1{F!LWCkd7X>5<@P<&X241<)gBM_b%niVo8bS9rT^?^}Z?0?Tg zN^Ux_v{+%Xn0`Pp{S^5MJ;|9oeGZ!6)$r}mdOL$o7hDpr3&rk;n4qohk+uu3$rGcK zL|w8DXls>6pY`4K%1&+#cv2qml6d<3xn}8r2d1szU7d2w70vV0(a0d*O}B+E2ae5V zeJ4g7-H`UJ=jTY1*G~JwUELd1K1-u$=9l6}ij=u@o&BKWT z=o}7wj977ir`LrZh{&LKyXt=jrc3GQBd$8hPT9#f)B_&)0fh5`b+hoNbwSDhy3m&+ z9d2TnYSZ2Excr=Vl?O}zk(w=sVzyEwgNjX)mx{}0<P(J{>T zA5Q2TazVF&-ZDZr0Xma+bB?&$HC~N^gpEuywSS%xLpD8dv+-{i%mW862~(YgFhjb3 zNz_eACs{xBdPF0g__|h=K=yFr1l3d5h{|BqUByixI+7{McU0v*dvw?o0nZvtddPdt zyG9+BNez6LoMe|&>%^<(FHEG!C5nMoCt51DS(+v<@Z2QX@2-Vz!I&X^ON#xo3|c6Y zhn!=j)1Z+_vlRJtQ@MC?i=seOH|=WJP8oD4rPZkG9(v}%+Nv9Jt=q)~f>MZGrOA&f zY9;l-g<>^UCdJK0p;QCKEBfKE5;D;X8{q;eFeKGXX_hXYejb|bJoLn-#TCIwc!j)o zRiPQ8eo19;WiYT^X%#1-*Gl*EyPu3*{Fn+OwtLS47oNN3ePN4vO5M|S{mvVAB^J`` zR><=uY1}p3%iL|F;}Zf6|DhPwSKB2 ztn4u%;d#?&gY%Y+Ytj1GXZ*n^gy!w2_a&cCBFjze)d7mRN0GZ!YzBRn^mylnpBBPv zxB`kCI^~s9c9KL-b!K>*5PDV_UgHjr{ULSJ+Jv3*xX27BCd=cTP&7k~@Us6tCA@rC z%!*-w-OTox{cdnc(1651FPG=n;i0!JaokIOx)~5q=$Cxvza%72*#Pfho}WgM#7PpZ z^-T24jMDnI%l8Fc1Zwb9Xtpy1^ptI(6_APDD@ze9T(}T9!#m|kqC%g=(|3z1-3omk z%{vY909y2cS**S;&4)gIn2VRhsrAc&h1kYvXG5W%IVTIKz%jpLE3!3@X+GJ$#$&kN z@_dfg6pJ|Pn)$c<6HC2};A?&3+=nFfrO|+UOf+CN#Q;Z%hKfBit2hE{nXocj-S1v6 zIXM?D!ypIPmO+?xa3hhHlYp#w=HN#>mjzAwl6B0Gz)v7_AvBI+4K6g8eorj0P^ z`;mVmiFe|~K6F$X;Tl{^F>5HY+K?I!ViRjHVPT|zDl#;GkTk{4llt^?wRW1HzSF)( z*a@{Qe>jn6UJl7-ZRNz)vxVZxmEF9=Y-Q9`LxROQO2yZJ@N!k~OH- zAr}-V&C+gZZtK7w-{f5I}3N^nfEt# zRvl}pm+@K8+vK(^iF1~0AltZ2eEoDswJ5Zzqx3}%c47I9UeC?v-j;13t@9QxqCU{7 z*10YZ?v&w?s+fg;g9AIcnVd|~Q5jyo-J^he5eEBsVLbK0@C@2QV%QV~w)jEDF<8kf zD&xeVYw0aYjw~t~Vj=GzQ-A?c1-U>w1zxtI#;GaLt+vLY>zP+MKrW>1DhV>oW^fjs`0EHt0eCJXM0Oup|zy z-MT@i-6L%d^bZ;)%W5fO)46^AV3ws!rhn|s2%|(1y2flHyC#v7Ca+=*#Q?$70fXbM z(kc!3F;xK{@~|AV*&N@8aZUNK({I|Q>&+=I?kZA%po6OOIXmN zcoepqQxb(SsOp(%li{A@(cN^rd>LcpWdZKIujcB5{D@x|ArzR^-bb{PUZ~#>ZR19+wysl*jv^gY zY$`W194XMCBtMzgD>lSa5z@o))jfe2(_BvBe}_qnv@RGOMrra21u*Qu@!>EiHg!%3 zEryDoctO4x-i*`2ZQL~ZM#v(*$K3^lJc-cz4SivH5|Oyssmku4lBiDk@tNCIY90T) z*N%visH@&NO5_x20;9MJepyBE(r`8SnWX(BPOu~dyKU&HC6HMJ-Z?TKThyKPT>+MH z5@-IX?~gs-i|4;NhAT#UJ1oB!#XWJi-!ejN$286zQuBgA?UD)9nkeQ3MUD?s+Ng&U z0#RlTxpaj!OE+;^NnRx)qwV|*9ZZS}@VIv>4#$&)ky;62cSn1Z`u{&Jd6P-7`og9c5 za?#Np*hO=g^HYhC+Wz&lU5F}rzsuO`dlfbKn4Ap+}&~rep@0Xhyu9IX0mK)atx1@Q~n>|r2 z4cMqmXSAunW8;8{r6Fx{oE1zaeD;3);%K7)`Xo=5N%lIi00Mb4LI541m_rn)!Np=8 zoj3iVEG!pdf0Z-9dkB4IR zri;l$j{`{0lP}|{n_ved-#;IC!@Q)%Y2}TDA{ias=5_)?2Vj>y*a8%%T33{Doh65=s_FV6vn>m-|10IieFUwbR z2W9F8eXI+wsO6+YX_*Gj=R%~n9&$MtT^+n%3ml%MN|dl|>dVV2{FYR4mMV(W@*j67-^M3d{f|^8yhXvY=Dwcv3~D zhoADPp1C>Pfno-xfW=(1pRjS+byh&NcB(ApX(*d18V)s z#AjUBg#wL5$l_NLUQ3tjBzpp@fOuf#>#0+bz(A|IB0ecj3f1_xgg)tQ+3cCf9PrBb z%H-4sH%S}lwSkpW#)Ul=P_iHrEFT^Rl2}p?4D|i`cYTad`nOXP&X7$njndk00=xo> z$)iX%6|0jskb_F}+kGYLomLB!32njy(WSypR|t2-Lfj>LP9jqoIzTT0G^#6^ZhBYr zG0-f@ynQk1oiPjRVC9Zx60kdO&M<*SrbwT#p6ZVQhhF)Ri{TS37-*Uv7EgNlD`&RKdq{=SP(0Hr`P_GKwmwg+ zE(mRsmJ#_d79ZFcPmFv1$N0U^yN#We3i>wh3xN?WAC^r13t7!hv^a5QMS%&nHdD+- z{IRi*UOz|T1%uP>(3i#4lJaQWU+2@r*J9^ow;keUeuvxT*Rw>|rHkxNxAb@}*@Efu zjdtGn+uBdRDKf$%i@&~|BuygFWog7ZLqjp?6xl$x3`bk%zgQ$k~-r6HGHe&f>X*8kaTpRA)8{#d=H&b#Vd(5PBF#Rw2(;)-gr z#fh~I^luu0+#M7HA|elr4OkRErfkB5j)Te_5%GdeQZ+h#^MROcz&(TB;dV=U5F+C0 zE8-#=T5cFTdCFF3y4Wf2M4svb!EM)jG9BRG9@u6N)#uzuVe+}389{%&=&Q@A-la%hqR~;ywjvFqtdGGOKwR!BGxdO;`5x-o;$ef1J*c<5t21Nte)ZV zUC0U_6x}6|yy)E1ZL@F9hSYJvYOdO_*)1Fr8#gbD^Ee9zJPf%ii4JqDycFP=@Yc4+pSUZxgsEn|OY<~6|V&%hs_8@jR%kIr|+CsQj zY9G!qPw{0_%;LPp6|s=~TN!wz>DsHMdxd&>pm!5+7YpZbA%k?mIy0f$%xvjEFfLdm6M@ph?~Vd3ymc!m23Qu zE4!3Qp(XB>;9fl*T_~>hK0@wBH9&6+3`$>#ILyPmm3B@p{RO{QcFLzx*$Xw^MNuV? zZTA&@La{tDIr1cF0QTYYQB6|hm~G$;jjwRhdT@+ltv=__Gv@xer`XB9b@Sf)j`=e9 z=_&#z_Jvqjv*&aA$>%fH`1f+-CwEJ=s#DyvK}SgK)XJ%io_#t?} zF`e&opY6CWU=Nycf2CCeSO!h>2d?j!vqiAcLC%}yZiz=C6B3{>EBChi_>59uoW|`O z&kD%1yErX%<}$%vXuXmdzCvFOS?Pb^Rj8i4z^$9+0NunMJt$|a0>HQ~S~B@T#*3bn zvQEB8kihE*L)+sM<m92t6U)bc>=>Vdx@S;@ILNHj_QB*<*RLWl=F&PVjr~Uz3@8 zMG8%vRa+=#6Gc9uVy}9y=9N=>0?VoXC~T*qRmBM|2{2fCo@CJHI5`4rqnGD49~qgu zg`-zMu>515e>t0JFlzPZ4{C!eSWScT>hJdkbOLjHIBi0k1>m+za{#Vb?vpoNr`ROv zfNVCs%k-*u9^Ec&QRpPCa^z^tpdZ-Dg*pnL*aLTzLF{md6>ukht830`^O!Fi^TUak zz!oCJSE7+8d@BR#k$H42=OpQO&-L$-F7kw=Y-m6|{hstZ2bo{g?b1_FC4@`kRRaBi zT|wAf;o3~dP(vyu%%Zf9!=6nZlf{*e9iTN~%Fa1|nre*MEcwuPHOZJnpwx9lXk;hF zz=|OsX{#R4*Xb-ehnxHw^dkBCnpv0pj!)AlZX0-adbx2S*Ey$w+e(dPfVkT%)$f)> z)pAY|9e5`UMO}F6wj`CDrk4cR!3Jdi8jtqGZU3;@y~a-KVypMnE^;v%bp?1qSUyNFE|7w365<-K8!BwusVFT<38D8Rh8m8*WBys zv?hrKKk7aCUCCkzc8_gy+r&v^P?I2)!ZN{4$TqTrn0n@9T89gVELBZ@Cp%pD{@~xQ znAep$EiNoHE=Ab@H8^fsX2IkC8TwRo)58ZGfexz?F;bbC%$Y4Q&bo7f?HDF*1_cTur5^h$z+1wFjwoKjgopa**) zAFO3~Z;5C=UkRdZp;qp@2^n$CVeDF43+a zpy1(kMmD?do8kUX^DwGyxo<3iH?kJ8Q1EC$Q`(R;a1nxlL`V`omS z%1DtdOYudgM-KR(l0sAc6PZp9j+h^CIy5foim(!LNV}z3nT(dP(`vJ&IY>$89RFMhuNSUy=~qF>C*@h|o3 z1&ao#Z!4!Iy*v0@ zqI1J*#+P?u30<}Ty`{0ob6j_B!36U<;+MpOpJO6oKcSdZilk7n7l9ov$9t;;iIOXn zJtRX^ImJ*(zWddUdN?m(?-1vDZ> zht5apV=>wotqQmMgb&46pl_eS@lzf%R!5Jucp8U&^$gTlJ~ey$(jPxR#nFjZ%@$G| zw@j^~Z-#soeKELJk*!2B6!n&=UGJn#*70-c_3o$XTY*4@0&TA{=o0S?dLxkTj*oTM z@*^I3W^6DRIf8w*EgRn%C(C~LqQ?M{*0w;951lwM2xy?hakxgoiL++Y-D-EhBJkNOPm8x zkqk_M(hff6SocQNNIE$Q!6gyUr&0EK*n+i!DqbGc+1DuRJ#&=#0qX)WO!?O_iMFu; z>1k*^>&RGu_1e(NN^=rX3tXKU^cL?fIvX0)sgY?Pie%)eqVicoL7zqE(JSQXa`kyq z4hOMvJCE)o2Lm38HHu9>t-fju{g3IfS^X!CtWmFRcgw{q>-MN#I21Z?Du z6YC@k^aV(C^YLe*X}$g2#iT~(O2Vu{Rl}i7GOX7p3vTp_L=9ayd-$Bklj-5 zansY#Dr~rT)V>1G((|_h5teSX>+xTC>~b-eATLz0p9MLw-n~>@NB4(4 zg2o)BVu&7b6Bs;R1GOYAoP^m&xcQuB#h^=E$YGyHyg`>w6^oRmpdas%vVa!paV7F` zmDwYSS3j#(pkjnP7-iE=*s!b~L;jxF{i@MH`~AXS|Bh^S;^f!^CN9?^ih*AAxl}Bg zNuRq{1DW17aVJnZVx4fdvX0d9R=l#tH{HJ*D!x|z;_$bYO*-@D=68DMZ+>U_B&2Z6 z=N$V<^^Y5YZLcD-gmd5Ls|Y*AZE#k{3k^FVgJr<-yW6K%z32^DvGxK3<8hJIQwKb7 z^>SXAFCO~;fm_2$P}pngEstwiUX0>NnUHv~=WpgZ$7!=>E$Ej#dbJc9K1DSFUDg)p zq|+ zC4L*I%(pJSRr~LaF^}do#w`0@W6b#oog`h+ABAT(2~gu)n6oajiGLvakT4Bz$p*Hh8jpcpBawjr0PP!RtEiO*g|<5_g?hk>F0&|N{@f63=>v(nb1JI zBF)2r79wmLgDPKH^h%3D18uC^VK(%w2B3vDaCz^PoqPkm;~00zhBkiC+64Ryc6w3G zbnzJ8TeA`Dv&(88oYx!m{h_x4Bt{J+yd(U8Bs;MNDmKwTnG^%Akxf)Awmd*jF|=r+ zd!ry>4SPop0NK)|fOJLNY;^&*95j;}8+&OKz1&j2yVKOzcHQm(9s_puD`EQ}_Q|NM zusfmi!Vfr%5hCsn`Yw=9UYg9MG7~`MQA{>PvZz?hax9V*(u0v5plmLSAByX~S@Ye8 zv%7uKfY~#%4=A-Z3AS-Bi*dVMeuon$Xr9)h>Ud}HTV3-@e{)7FsD_c_(FEucbj+@7nyDvcLpLZ9S*QtExndmq@YTY*@qQ+{7?P*5D*6>`h_5El*I62Ci~ zPDpdx4qaD{(x)OS-L$GEX@Mw%-UVaL-x$2hMx3%%gJl#Ktc621!I5RKjN(_JzRooI z_!|LKH1xB*B-4ojRAmC7T@U|BU%XJaC${RDbD)TsZz~{ePR$-qiAX`AjCKxdr zgdJp73`HI?&ogjZSDb|e2!r0iya%9&J>m*YN$FhZD~Tlc5xWzpLB9$s z0ap8S&^#x_tJDwIX!f;1tn}E9CPU2T)yBK5U^1~JatqeWP{=GStg&iN!I_&ixZ ziIkd%_FReqMv_b{>Fpy8fr}vnxtPxN-^pzgVjgl8jmK{VE_J`{drGQReG0u52e?VR zbbmaqQS^tkhwT?*x3&L$Q{RjI%dh_MU%!tP(~JwDsL^Zyr)5=Ia$x0^2qRF0t})xl zE+;llPMRRNhGKv*?0}(DSGMv%Af^xL3CUYH=zh&tVoje$kv*p+Vm-GcDvR`TktbB$ zqSzF81^P^C6zy_-CZ6Gnu86a~C35f_<_G8i{TrvH$tx74e%SmtnbSmv)~;J%9tUERz2v-ke8fW32)+3E z3Mel{m>KFez~d6MfdS&bE;^CPpx1>C$T#tGmDRurSUyXm$cOTd3_n~v#R;GlH)eRI zO;(pqtDCZA>Jq6s32qxNn4X?k8^P=8fya5|+PnwrUq(+$30)bA7R#VM6Tee_jT<*R zi>!sL=o>TaZqYt9Csxn?Q+L|;fMwyb^FKCz!_%mXeq6f+TH>+u)jF|nZnuf%+D0*% z6xl+>8nR(ad=CN-ZI=6iY5kJwsa3+7;9O8LaRLk|;&Mqnd^msFeMt|52JF?|%VgrM z#*di}G8x1UJ3p`ZWyL?1M?Wn*fzk2rxMxi3XKH=&{k5u2dA)CfPYL(9cY+V7mv1fX zjLGH3&(^B8K^|oRw~^F~5`t?2OSs8jUpaTc{i1#O(8r1t1dbg5@_5LCk3)WG&N1t> zUKB88N7SJ|CcjM!-|#)?b1Idnp>8q=IKie_mtFHL^a)6k399<`<8!N3y%@_6cOwqaS@jAF`8oTWOh zQyVo2c!KQ1Zl)h>4%b3{m7E6`Q5J9;s36oPAPxUy?98PS8VN@h%MbXmQBL zx!j$i7FE1!BhmN|=!=f(gFEFw6C~a&IimuWWGFkw){}T|ykMWCMYTBy?>!8{Wv2+( z-(XdAV+IffN$`q|N<3E-vpA$&e;&z-wvmm{@46=vzd@_Iz%26GIeS2UD=-OMzUVj5 zs?y%P6Vu3C;3kRM<i3Ji3Kb6`BDtjS7B+aJ}13Sk3*$HXwWgM_X;N0wJ~KM$ZD75ri49{{D|7ZPF-M zbz<%W;20bs%5o?MI5skZktse*swJ7cT>;mnopvsWW9#D^fef38s?i6sjGdp~_xiTjh!ffWywyV1 zII*yTFwzK2WKhgzifjaqvl*58o-N?n%?S9Yi27T1V2k1!_Y6pyv%Z6}l;|NBEIY+2 z<|fx|&;_@8rbAXzW0NGK&!8=jT1LYtB4Wbe`jzDYMuhxv=ixPEn-e3X)&wCX6a%D# zJE&Of@RPtx1_h#%HxSI#2KxpQ_^_(^CU0G6f#^VBH{JIJmNW16ZgbOtVnC-}nmj2q ziBl1*0a{Y*3{(Idc6aD*DXyMMqVCDM|A8DX>`?R9pSKr}ry9|T{dpFu5nDX2ONSz} zfKl;CgjQB4$h8ADtdRm%NF8g2vH3M_&GjeD!_jO!yiV+xu@Jmo;-Qn2L@oB{4ug0B z#v7MNH6ChULjb-@3}E3Y)qd0oG1Ct0*gAGQvJF-LyG2MhrOR z@(wR^&y3Tqw=HB(W{8%$=TEy9RtxK(D&f&+Y;TGL7gfS~$ySM0mFT}XBrZfZr68aR zh)(f%NmNznA*Du%cWIOd?N*d(c+lgR`w=r``%P9VhWh1}`)``77^g*w1*Op|`zj0# z5OtedaYRWJE{$y;!m(natU-ho!`S>1R*0C;b>TZ#%u`v|s2C?UIxHl)r1(}wbSRbt z47rp z|DU=rs5k;ihzkUced!HTW7QloXc*O|AfIn2uCmSY_y(WTwJH>!au$78Ue`D2aXF>T7aN9z0 zTcbFOe8IkVc>Sc7gZcDZnN_bHAz!@y+3Txb)2dJ%r2A)x9t5BEJPj!VX>L1Zc9j=a z2?o;z=HtERyxZeC%}Ovh?|a=rtyEG*s66s`bdn6ZjONC5+T>B_0W;F8pF}aMD6*W2 zy%I4fYn5E!>eK$a9M?$M3{zI1*X$0*>{>?eF9%-#H}e)kFH}Omg+@WW@Tk z$rS{ga<9*TM>0Zs4lL+_^-LPH>?iabT;YQ;r@ty9zhMWLp7}$0F^?hW&wJ z86;QFguG*(K53!3XkkzhEFGYRqC|*#7@};)$Zl^#l86JUY|Ne#i$`sCr+s%VKk64s z8h_ez^6j^c+T?>Xn|?$NIdQbF-9&SIMlnqkIYGtFKkHk}&Eghu`yp3kz&%SeFk?Gs z{swL~{W)Z}Et!4TXAKmtXVF*X@g$4>O4dlq#Pj0>XMLOaXS_;f>!)t#^vmX7Q_auh zeId`3ltNBQF?2d?mnZty1#49$uNMkxc@KRyC~>%a&XSPZ-WLNZ6uJm(V0G8MmRHKd zGgaIM`Zh^b=6LP_BiC?dPtfP=68>4h{fS^1;})A2gcaPWg3>kF-bPfl{A%w4vXR}Y z%84Cw`%EyHPcgX^6w${bO>?Vc*K}Z_y(a_bOGF}r?a%VuO9Q$=mR8;@__iD6;Ex8Lu+ z);~vpCKvkn)XhM>1JnmWNrV!N%I&nn~0zjOAQMVWdjwuI)(Qccdu+KgQaASgw5mE(i;3czSw~q zsj@}?9%z&n4fEEPlWcaTs1qZi&P2A9QVbM~6+?O~q%}56v8U-?AFa<4Xzp9$w~t#2 z75ZOBVT+e!Qa-Cbc<-#E@Ttbs{d(EyDZ|*QbZeHDkyOZP!_#>@i}uzdPP6L3^g3=pXALr$KXA+1b4SF2jjEe<;3 zcZ65TEB3lDXOZA2G}q85inv;zhrn)|qwIrNLMA6EG}HU8k4BN@{v{+X9Su6{Qy09+ zD-Ra99sEvNG1L|21)U2Tpqu#JklNJ5Uq7{m&H-Z<;uEvCc-smsHbcoUlx&0+%gQOF z^zT&yBYr+Cnfw>B+KG*n0u!ukrkITw{)lasF7wQzd%2x@DuZfHkOdUzv01za(}F6e*7=+E`05YbdfB z8Z{_4x~`J%ld4bpW^vUx< zCf1^3Sd}cVdll#ALqrQw(HI@1bG`gX4T3`Sj3zGs z(V5|DLqxBGkIfd47iH@!S=_CX>*7qe4jK~|a+FPU-mJzsLoTiIp6GneL+CRIDGc%_ z*Wp#Yvd@B(dC0+o4e@LV6xLvPjEYe!Hjh|Es5 zU;60bU<-iR5Lb_fn&O)L;}_k@w#L8K53EEGO1P)h3!~H^l&@ejycm4hW4Ghg8GA z7}CvyHc;JiEY|7ZCrrBW&XJf$b9HYP{$}Zl}w)7(n-KW>s`+| z9n=sw^7xu>X2dR?WJ}dEJnO75Gm6&H9~vIahM%l3L*4tY$Zwm6b6%(SfBL%L%ZQ9dV6}e0 zxfRwH3ONC5{15Zg*x?q}G*!Z6UaMrl-EnIgoKP0XaA-WwGBTvT-(B@>qn-&)c>SNr z5q9fg=gr|IohDl8EX90AktQlOdDePS0r$FR<8)wu47e3|6k4MXKw42YWW3)3&J*nW zk{{4OUlQM#aYu?Z4(}f$>qVDlmIO5c`EDAwN|-~J!PTdH%EY}Ui{707)jJt)>0;J=^W2*~zc{1Ps<1<91D#6!Swl5W$84%p{~r1j zf2{}pv`=0;8pKANv4Wf>)n)B*7D9{h`f0#;s>7T(J;_3K7;-abh>!;iO3uNDTOiu& zqw&w2tj@JNN6u68Vl673y3x)DzW7n>{CADwCX*WYE;%_`$j6D}fnS&il}i-UN)as; z+bgbAv`UTz)JXKK=(rqCiK?RSLOVJgzf!SDvKF#S)Sq(N!_wqg%ArW)EQaL~1P1%v z&qDtw^!DRL>L%g3(0m3O(eK>9#05 zPS7nx@O1{Ph}I8~67}~gdj2+MxnTaFOx-T;Cpawrpv}{O9JiK(_u`@FX-8{eWUc&Q zaosb4Xjz{LPF!kx(Ycqrhx+35<7&EE<9d1Y<2qf3x8T2AS3B;OF z+B|O>pZP4UKRFgOSU>vMue6MY(ZWK%J@M{e&AFG@IO?3(;;A(ePLG+I>Vu1bx2-Fr zMwX(lC#+G{d#2CEmAFPxMyjV`=hTPNW!PqELW$a+GaObae-ERld)w)DyH49_r~kGyomRRy(``EG?Ot3^+z=F8Km#a? zY$}Tj$}Zq8prBDi1&O$ERd7L6_@6_fl1MaP7AAV9_gDG$!1q2m?|IL8&U4JXT1yYJ zgNCI~wf(xJe=q*VGy^mq{7|x#Y<+8>QEdc`VoJV~A_Y|R4fj?c;>YmSXFQk9&Y`zX zedMj)IrY%YL1~iTvTttwK;6O7GK+(+eRlH);D4aN+vBM{^p>eN=tmpYco)7jQn2dIQ-FiAi@{u|T|R&ea(mGloa~rgsmm*HIX9L64?QoE&`BvFjpR zhW^Ds*?)_byD~m3>05g~U2I#d+Zk4G!y?ICQN`3Ta&JvX+7kSj-ZK{aAGdzt*S|8% zHy;+(PbN3m%{NZGx{foF(-@%S4=HkwioPJYAjlUT2+R&UB}nI%4R@}CMG^L82ERmb z7(y-s?odj5fk6`k^AhS;tnn)ms5^NvQDt+oXp~@2AYDEOIenZVm)+7kV1T-0J>lvT zyz0RFlFYeX(h4^eWJfJib$Xaaflt~zElr45ftw6>s8K}_%b%Mevxx`#fxU|l#!teB zACqgnL5i?Twu+PMrA9^6a%C^UUml~Q>1PKXr@eF}XS$H2T!oy$ObId%$-3F4U&0dt^6d)$|G|6}=<9=U&K(XRf+~jtS`M zXNP@kKTpAu@A?Q&;B}jJGXw1A3a8bh_~H4VRz@03)`lEuI@!a{WI6Hbsl~`eZlL4` zDN?O(eU>acORysuFnU{H=>+1(3hb!~sRai6^(IC%(O01_+HDamsalxqiJcR{dSE25 zNqRd%z145Udo7c1hbKc%`7EW59_}mwwb($AA5I(T?c`Ah`&Ydr1~MAUW~e2e=ur4Bt)3K>>Fs z72P^#j}+&W)G20ecNidSMpu74cA zLxMq`Hd!7Yr=qK~QMtW`>5$~m7z94;XFm(7A5y#F#%hF&e&elAX8ZrQ0XMQN8K2yC zVmi%gBQI!(l0TuyBPu#mnF~EU?s+%yN+Jvm6E)tk(>3Ih>_A|$VU}#-4MY{sP-C;1 zWZ4?;?fg>dpuYJG^WZ)GMB2oQqvE-%JmZChqzjmyebPfq;!P96t4fA9C&svKUvl%%>fZIJl!g} z>K;esGO+d|7^_Bk6@&o5%gnKNkFxY4Z>^VP_|v0?o7)(E&*ZyTYhXDxZithgbhAJ%)?nki>%v zJDB%2P3u-_{B{U1tCcK6Dl(o#CA&Dv-p0wuG}vOPj2>j?9irORY#JDPOK@T)jFhFk z%F%Z<#!&M;$WTIGa@6|3wnoFjqXd4I2Zouj-Hb+ol&>-&>U!WwAed4P)APWwwFM}S ztXpu|_@kw7qefjUH${Ae-=Y{^KqFV&p=QABx7UJUH+!!mPne%|>OTMDKnwLP2c^Z# z9m#smZl5)NOPF0yU78Iwh6nujflSUP?>waII2VDlds5U(Cqy2Q4N7yFHFI_`10t=o zNV+bhNwURv7ssK)V!_)tT~j>Y|KEB0qokUhH&gu6q`N;iWtW>Fao>$t#J}-wmU4?% zC$EzD#B1DoXpHOha=W3h8Nyu5DS9X9bYaT5PJ~Up@$6cHXRk2T z((*Zb+%z5o4$%*^1v8EWkNI!x8-;BVF@*3(?1_3OW7^Go!EiRG+KJEr{wD> zlAvz_9K)}A2je&;Dvjciv`5wjek!=Fc`!lQ5?`^#MzO%RHIKLY9^nd)m8K=NW@=HJ zCF^7h`8%ZvTyXJFT3riX-6DRjXm|y9jxZdB0qryYQ<_Q&Chl>lr7G*qPCGMWPHae-%)ATL=IzV%i1jD_+I&A%84cDVk@; zw1oxmV*_ss9%r{p>Q_vc+S0nrc@Ymq*drsknUov|;~FYD4b+rT0;nVcWBlrdz<6N} z-QcPdcheB}2UZn7$!}@M4xr^erelIv%VWKvnb{SuEahJDRZfwsL;Y=AR=IYA#Apu+ ziu!z&sffoLRnulhssjlQ8b#W0DqD?>>q;W3rRu~*< zFY6L*A!9o4!i<<9V>V05L^mNk`aqT%R>O4iFgG3_@;G$442uHV#NE;a2O9{Jp|%t> z=rWny>d;;7?xW*v37sc7CWJan=U#Ik@ck$ zT2!s7UV0bQVegoCMOFsAXgYZVpZ~aFUKWiq-HtUlY=H876|UcKC{Wh?+%hk_MjNoa>%tB@)sIqgDci+6ft92h0OA$I`l=3e7Jd&JxZvwf(%Nop-3tfoy%l{dj6@Yb)sIc9*D>_gx`XGs=3Uw zBc*~)UXLs>Qlm%_-4oV}*LobH&pS5IYd{Z6K#zfa(EnoQ0j+)iTB|AR-)RZlW_bKa z;6_~@>}rX8$EqMT25y^qB|baEm-+ksyZ^T?9ZUxQS z+k)oF*MJ9$3E3_g{)>xR)KABX)lt4KE;=LY(9T@WK4a7Sx{m0;Bw!~Zov|roj}#7{_Ch*rKgko8dfap0O|+1%9&+h` zc6eB!wp*FNU+M9eN@!^XwI)|Nt#s<-G$>=lkJc!(ey4rZ_?zkb;x&@ZQ}QRnb=Wws zJG_sBzxaU1x}V*?JXUDHL@Q}p=;*c8U_RZj>DN1*cu76XHc;F&-g#= zC^GeVoi^-Z#*58m8i=0b3J!z5ZYN%NN>DGZaMMXsqn5hkQm9LISX4h&%04hmtOwlg zLp%BdSs9b@YsY`@Cp?2O`QYq^gJgvh&o*EUMnq{-DES78BvH{=gLBs79>MzA1A!Xl zy$GGOAgoocRdvleKev%IN&c+9@WG{dXFhP;*WY9;*n-34c8Be<0tfZC$&>d84B%Lv zdim#MwG)FQ-v}IAD0vD+Hc-*s^l_+)I?w5Kza?!@R)bC)t|roiSul4Xy$GVUxa4nW0d9Ne4XMTM_)~BlMo#M`N`8SN z=ddAW3EeHs@F?ISd*Ch?WJJlJRI928=@xEcHoKnkEDgF0HDy3KcxS!)$#>>gxSa{h zc~_%Y=-MmN@vxk%o>N3uxUKP9>v2nZXEHRdQtlRQ<5Ux^DxY5w(#vgwn(^a`t1}Bl zJ49W>k3szaeY@w%_p08*trpeYcTc<@!`JazRTXYeqO{V(@N8U-ubUhf)F@1t(IMy& zGzxEu(gG}kq{TRTVPS0E8U;2l9Wp;tot(|z{38Rxeso~-7s&Yuq~B=JZcy@W3bga1 zQGKm;Y94UZ)59Q6fi3}NQJ{+xB;PXJ_d}g_me*E@Srh@WFdfEpn`lq)Fn=UHOb5CC z;*GET(UrXOWdKGsxB)wivM1JluwRSOIo??dtkpn8smVUS~ToJHbqBM$eh6@*VI{ z$INP$bn>1=^-S&tz!+pK^lB%1$($mucDr-rMjt?nG4sMQjs74zA2pAeozpkpbTLd) z1&@CD7ZSrRBjm)$0Uqv%kZ%GdUrmt}SQ>U*(aB5l13e`K1{4@)fn4%fTNfB$%fMo1 zBv|k`w}1V`&RGUX{N(iPLQ*t=G#b6j8cGg*Di2W62O{e@DFK=EB9g_yNL3wYT}ZEc zm+ZJAmJG-SU7qmPa|T`3g>3fAQ10M#%^Gwm@!G*@m7JC0_Eyz;;g+dY0sE#7x}YM) za?c0gYV^7fx}2Z8UWiAtyt`&$3y*E0UGSV5?m^hM&|iGbA8OglfAKZ7XzMkfX~}8x z-_%ju5$6q$MogCV1sO0@+*7oOsdcnAhvQD?30J>9jQBe2$*9d<4MX&VW?eKYv2=wTiTc^yAY*DRuTQO(Q zB~^j{F5|A4)2iC*e+z<1_&Z>Q^_JjP$u-F#;g;a^;3ZtVd(Q{@dwWSqU`oJ)=YJQN zGFt#=(R6IpMeuvENgYs^CvR&sO-QkMfD=2=X3|%+fmK1-99(OuaU(r!xPd^CTalaE zo)eTM#=>F zr`-+S=20X&sysip8hX#BnGH9x7)*;n#LAwq_yx>ddhrMAZ<{i_-xAAWhN%85;D-if z&?Q58QrRY6=@9)oGq3#lVT&;F{OYmW`103oF8_?du(W;YdNwJYKu#HbBkL$R^oXdW zqL-@D0?!Alv0(mTU0z;fIq&apPP%0=>5u4^7C;&c^o(b; zDE3Nv=mH9bZLv$NM$yb`4;fP*g$<6a#iTX&na6SM(tr8DG?Bz+)^Xkx88MS~DiJ_W z4XpI6~$khR<{S;f>7!39)-W2Mnw8ax(=dFft99*9~}^liyAm-Z`Py-3!+VSI^Q zM$_b0N)ArQW-7W67Rc4}KvV{t%6-1*ymrYBz&>st0&O_m)Dl?W{#AdVClcGf+?!Uz zl(1?XZ{5JKfWpYx&AJa`ahG}H(c1I-kLDUc((=6}S)^15g|C+1lj2wJj%%6_L>=o0wmYm8BybZv8=yDXfxwEn6>ini8==tA z8+c8n(tZ&HPCWX_lc44PrJ`==?~QG9)!m^;WoTwvp?_5_)1pY=R*Nr*HYv+|N<}EU z5)+&U`rA$79RjQ811pDTSF-{Lb@Mxa^#62G8#0wQJqO%d6gZ7y`%Ki%8zaaV52+;pGQMMM_6jpD|9HN^ zKbcn*6-GMP#bup%aRce35p0*cl>8Hl+@zvU(dpbwx*(!if-QW`kX|or?5aKly4fdz z^t;5rz`s5y*)2!h1@uv*auh^BWCt$E6ZXi?!BKRbQUc0(1rfsq2Kr6V(&em{qnufH zXgsrwe-mXp3MXNM>Q0Y+kT$}Py~-(wz)Nx=y9q60@j)F>8&w&&Tc~fn*vA1!?HI}M zMQ1EugjEZboEC*%w&PGxzZfM#^7&ikm7FeFO2FR>fC-obB6+>;-JnHeI$K(TxmWsI z_B&|==D)V7KK!u(bF&ueW|H=|26K;%FxNxLZ&0M0irz+5DEi&&#r^KhAte!Q;yk|k zcyNgi#uot}puU_eL+O@bJYjeeukV#~DgOHLpEQbA(GJ0Cc_IkK^&?-+|&kW5%&S-_(3D3-M+$xDkk*Lqh`(%iy9g|`S6lyk0h97|+ zm8pc&m~3x`7FU(@@bmC^g&Y1dY4E96ypF?%?eT8(zT&0sc0;svLQha+XX76MSnuRG zZL7TRcxRNGioH9n-)u&{9mx{d_4$%0hiRm9lmL~ToN7o36;3KqY83IzHjhlQi8%%> zUr$bM2FrCIu!xgMI_U!Mdl8PF@vJs9MrOhOJFL8;@vh^3L4Zm>q(^lEty-Al@qLoQHc0euFr?85Gtvl?V3vzWbM<{|t;u_@ibX|0PH z0z>&{FNywqMfT_W(mE0!QY5Y9U7A@W9H2`bEYCg4SXesDqulDyL#%8G^@ok3eQ&xx zZ>nD}ec9`hXvLf)QUjgW>V-5P7GhG zx+ckGu94mT^LNnMep$4>#Q!8`#T>1V#_R4xcpjN=Z{uJoXiQAO%e!`l5IfM+b_dOP z)8jX>TS3*C{BC4za zGVbOa(OJgeq5S5^gTrJoI}gQ)Il|DSXvAWD9VJho$Z9IOR8cZ9(;Yf_PdgEO!@t69 z;hc@~Lv(}Bm<9eu7z7KA%!oV2X4AiHL6t475uG>TLCmy9Y*a3F?;tro>KpE@bFR9r zVAO45jM!WdToAx!ZD0l0-ZRXrhSEY{=x3W(J;=`II6p{FeBWR$et7*A_jC69(u)&tuH8S*gKU{B03SCtz7g>zuPt zx0cQnFkP|+pLi_uhqR|w)dCHR>cw|lv*=7x$=Tv{k|c2V(>ox;5bOCgRI6$r{kS~= z3%gT;JE5^?C&cnHd{<2@@yVJ}42m>J`mzB}tl;soOY)i{)?>rnhG9*=q#7TuKgYu(nardm;yYyQF{ErM%256v2O=tTO zCx{N>`|D?{vVHQh5g2Aoj`fhx-^tG8eEr=^E$5Yx z^Y6daYGSOI?4bS|HUiw$v$=tjuYFVGe+P# zOv#~QyA~@gH%)5=pE+N2AP~im;*|Aaoo-i{7=EQdqqxM`DJ$nyOvMxcGJ&&{TA18` zu7x^BkXhnkF0oyp<5dR@3DkOlq)v}!4irJ0moJX@YKbSy*3ZD!pz0K0bnW)ZqCW}MYZ&QI z>Zz_kUS*)-=BjL8ROXC5z*$YB*K=`eQGjl0vaAlI=j%w#v^r8k?B!3}9X%GJ<8_yP z_A?8zybb92VdJ<~lH$bk77YK0xLqD4hc2d>*f}R@xE{8K=?Fc~S;AHKLpLqp^g*k? zVSWp4Y~v)$(z*I3)O`_#lV~h-^tQl7)01V1lPkg_IOIU1Rzd<=}@&0$R;j;m&}*<4tOhu7W4 z3NKXiR*Cr2Rm_y+*9CsYlBgbu8rh#~xF`>ag-z`ut;5?vzBG82qv)ku9eUt%b#vkE~w2Rjg5z01a4w9Q0=* zi#VxaoitXyAwhD0vqNexr_}sVS%?yo6%1wmfp$7W|Fdh&o6diVn~?OCOQMCK7Nk`T z%p6t%^2(;~JDfN<$`DvK%%j}s;6v=(4EJ|eo|!@gK+6gtWS7holo=0YP(sy_;v8?&9lg?G9K!_K1kSM*ycgYyF=d(gh%E1VF z_V}eSS5})&D^8moGBZ(VqKboWdZLU1#5J&M4Mw#xXTT29-`0hkPz-?7*a?C^2cm7|^jo%~hWEozFt#$CyDzbxC zA5lI#nI~+M)xcBr&${ni24&r&Wu#){uxJ<4gUI5Y4lFl4w470h2t#-vGrIS z4WAYrvk1KVKAM1vmQ z+$!0cQgS#&agK(u<&b?6_Q5AIJolSw}$t2jFZ zTGeIIQ|}V#PEL>DkkEk;vJ@<%&m@k+%%7`nl`9P1jOy6Gmy?Xqs4Pwl3DC40vFa(J zmXLm6PWd322O)rK;G;B5kCU7OaY!(!JoN?_Jn?jOmc{Y!kn2*oBX9Z^XfwN5ww$*~e%0QzIopsi5-;{eiTSx`p=>!tu5aJ@!S)XiuwdTTXqt6g_mb9wq9npO1;kRMtl%b1KH(DX|n1 zqm7w^j~#xea+zs4h|?}@%oKR&Ic?y83YWSM$hx6V-UU(<*-PJi#) zWBIp#w7VU5%>3Ljimn;;?mvZm+jKs0T40!&A%Gmh!%iU>gE&J_F1yCFONK|UNOV9N zz@vxgwh#Az`RJG88Les~N&oQU@(+LQc)(bWk&(WHgF|irJ$4uoZ=bZ-R4RkbqQ;4% zZf1lv<|EZC8(P-)i>rLI!)kn>fUI4BQ4>5+Ej{3`Zt!`gms`N;rSWhQNrTpt58f-E z;#go=2`WUG0#F`Y;$946SuBH9pC*4AAy(65JsL5m#px$vU8E=vD}}&d;Qe z@elH$&CxcJ?AsHdRmJengjKk$C4;{9Mv4vCFtR}QMv8fZvs*KBaijq?8*-%SWRDYb z)mn@Y(m=@%QluIdyu)n!f(Vq?xjwUtQ5Q28nF5a18==zVesy+*+c8ov1g5j>5a=-> z#T9kCG>YsnFf+6ox&uR)U*j?00bI9`im535ge#mb*&4WU8}#^8FAm0fK>MRk4}-5P7EX{ zKp1`I3l!a>7G>liT2lgxCdR zL`t#(9X5=)<+rjz$GEd6HotG`=e{iE@5He-Grn&=KX!sndVOXSFEy-5+UUJ>f<}Sr zDh)mrz>x3c;r0RlC!#F6l({$cW^ig)YFMMUy{DbgM$HP`IM_ka|CZ))^RsBrgz*Lo zvSe9sJsEN#PP-P`Vq`(qQ1Vq2Sx!Y`Bm^@u!_Cp##Ie(lI<_6&Xd{ouEE#=s_d4>p z`OzP3PU)dS!c7@VPKyXL9HbJ^Fvb2rD8W+_2rQIfRAU>ICPr@exWJ%l8^%}1!Xi8y zS4*(*+?9BC(-{0@hmDU?H~rGbU@e}?k1Zz}C-yhWjjY8EN)Afw+o|X-QaZDN-=ex1 z+$6oK%Am?WSN2)dEPz&q{<@F)K3er*-WS&R)%a*tC{%igz8qK_Jm?ZfiYKr2(5fB= z6i+_wqftN!hV%lH821r}tany=hKY6$3&d>D(+;#7pv|s56 zY>XT=t8p^&Ar9E~A%B``8moJG5$(j`esl3VdMQEoI1AcYp<|2^lxe~wzv19qrV{<* zAt2n=D?541BhL^JC{V7Jt1CkuhocBT%rp09<%*7i4)TyowYXZWj$=w_bY%n;w zxuKUxDm#PY#I6vi`HYC{DvTBgMdmvJivwL$ek9U9?BE$4*(m|E7>2A1JaetBlQchN=}6vDx$W)Tx3T(Vi!a| zpN?#Ru*t3L5dGIbN6r7H0XT8LDZfunykQyan?_)2r{w1;Xe=2W>)j;Hr+Wmu{g+M1 z=NEe|3u+3{DE5l0BbU!Xw)XwdY`T+o->s5HYpGS0L^K4(3;Q6Q5a)T4#DS7(eNZ*; zu6NGVYm!D~ljIOHMB}o2KYcF(=`>YdLy(MI#@{e2p2>0B?r~W9mp}fsMV0#DuTcyB z{M>X=sVH;q{_vLBc;%oAUa3)}eh3XpIEU!7-f7dGdRMqL3NJ94DFeJ~k|h>(aI-Jw zo`dhz8z%S~ZPz^JR-C@+!!zLPgR>hBk`+#Dn80rt;mD*=@(mP8LPw^O#ux*zN{;bF(mHgFS$%W{**R(z0BG(Ws6#Y7RJN9tfjNr?(cAoiOb*P=@|73E5X~mJ5 zQXTxBVoN#j5@PsO@3bg-z=y~lb~D<9C8Akn`q`jY8N>MaQln^_QM(%bwu2=`50GO3=a z4cy7uEof1tL4!IhKx2*%0TAaH+pT+U6OY*ss<9%Pvwr+9Jt$e}4jNq7eW-PRT*O zw1JB5rmN`{L`Uxio+gTSG;+EWTJO6N>%(>j=7;G;mHQwwUNIGq>eHBKNs~0geg8C6 z>&^rDZatdC z9^dM3jb+$3a)w#$So_s5kNwYm=EE6t4RC7t-jXa*#x4}>#JQ)lMnF44$r~tgkc!4i z=pBN5emZxxypo$9c7CqLORK`R0=V_W``P)+f-2u)uRFjnZ}HsDVCj7q^Egzac+Ah@ z7fH`RX6l^s0SI~wM5RMQ9FN8X_XXm5TjRBuU&qh$*&(QK%j4(smvObK%!n9%*Q~|N zBG-q~F4-l4)er*+I22{pLJaQPy(l4{GGOVc$J3K!(8WCPMq7h9ZAg=BHo{UoC0|LA zI4W9w{k>(}430Wei6Ou^SUA}-u2#Q#O8{Z@!)>^6!>ayEru@B^6P8YFLdnhefr*Yk6aIxJ5#+JIPsfaf#KXb*7EG4tR5xNB9v zGF4wN!!SvQ(!zp>Y5^cu#>$xhJrPUN+%ea2&smgHh7UqI(gm_!W`W zV4Lr{S9#qO)r4a);#2=(*-q$&c`rgI-6($Qzw4b+>7YaERap(g^YJj*aKKQQABIuP zu=9Q&e=0w^Mr@d11a}3$CJF4E6({zJLAqmvSDa4Cfnc73UU3h|nPzznC!^O*>=9`E zR)_8JE(}+rDA*?8gF@o7ajk5jvAs*$XF*wA^n2kTZvGEV4&yl)$*sI_e0fcH5N$gQ%|U_picuj+LL%O z)@L-eWzlh{Vu_*`7D&J4Ff2_w z{>LEG1@Bve6*KGhZ5}x@y2BIRfxLFIY+cBP>4yU=!Z9+R$B*GxaC^OyC#s=Zh|?s1 z@}j*_*WWubyHA=hBbgIN4FHv;Pul$1I7&}=L0X;O9O1uq2ak=Id5$4_gXiQ^fA8sr zsmDdA+)J{Y*eL-qjuA`P-IN^EeRk>v1aqb=ku1OzCHDA1n*giU3nW=~CPXLhi9zSc`nM)fQfh>2kZWoP zY=f2Q)9knZUypT-pRR}+wzI>whkrRG*v82LplIdTf-HHsSMPwweq~|!YI&EU1FREF zRGvLF%52-wz?dC7s~}?bY4$s@bLO8`yy?O|S-_wxjXVw0&b}$T0vFCf%`hYj>49FWMsb&3MEZCuW~<{v@R|l#?Km1QEK%~@6Ls|Jm=#LKt%;2* zn{Dt%(q#XZM|QLGN1Qh$MNS*}BlVOVxY<=ybUuHp{2G(OToay^$GI(3=FaV!wHAcf zwW{vm9QrZI6eTGK+-tabeoLTZ>NRGAaE-?fa9qGr30r1s6r~X0?N;UTKY1r*(z<}- z;H;Dhw|!>mEMTKZ`a&`Jvvk%CP(y49O$gG0WJoT*16<}*2oEovbxdw6Dy>G{3;d0~ z^%~pry3^k20FN%JiS&*H|T+t02fJ( zLD}@M3#2k6HmD>V{gr2zT=v@HRT2RaT!{5zXT&V!7Oy?tm~~zfwTi3Gq_2D5^uG7r zfUH-NB5o3_3jtv=_}tpSWnr=2>hm6#CN1a(2RYXJxS~8@|CCs7oWf!zF3>Q$16vdt zEtFTd;qf&30;BbAgl07R+_h2+ z4Azl0ajxiRc0rUGpxJVmmBku2LGrEt5g72dJoWO=$!c~M%ZYQ2po%ymgRq5?r%+@A zW)PYc6XV*3QOKO;u`040ao+1*t1Hja+&<}9SL$p9x4+G3V4S5>hS+QL_f z)&=YeTRNd-xQkgfSsa`aP|nLHSR@!bVVB<~*BnlDkpBDwT7NzRe!e*WRMr!!KcB_H zq0G#n0?8#MAb}fXC-Gfw1(@F9SE$Lt5FQ_`=+%{Y7*}g zEoN$%c;TUl`=WH;2KNSnXP$Z=@b7}}={dob(4kPRYL{p0q%3;1yoPC$9fAoW%g=rT zVf*AusCgHjA!}r-C+i6BmI+l3)$(!*5-tfwW*@ypvzlEMPmd8-GoK_ zToEYxtW`d%M67UIBu`bS%Y#zH>Jq_TP@+3C`HpLTASM*g1sA`kMxB->@m05L%$7*( zS5wI80p}@!p9KM9kTt~=Qd39*_u2EXk3couAv!7w|GizwvC*lZsxfBozL zVEJ=xz0$lTiMlmC_0tvWT_bCitfSU^pi!iVmj%E=e7FB0pf40e%F1M{vyR75Y33*nk$q%gdJ3`)mvxLJ%f zI25dg%chIX8;8*-aW!A`xn=68z9nvk8Mn63?eEV~ROt4PJYlZ{s|VA=F!F~yI1rYq z77v0P!5zBOL)}hLVi`kng>Dd3^wMA*JWz1HK22vo)R$iKzj0G*(A2@gRQ z%`@p%^|BAwa3B@c;A#^(+2PHM4{PSIEr470ZPI8;_BX?l?~*+UdLp~~?k4EJu*HGp`;`tQm)Q1_ZF!BazL(vX z+G&@?6-R%1*tFjn8*AglmC9y1p_K=n0dbV|6Z__ToI{s=Mx#IpUD)hYejFcC9;6fN z>x#DebjcD`m&7MLo`xL(QSJ+}k8Ls*uSV1GS*xLL!Lx0-B=MLQ_}|9XsLTcaMK#=>$i=&S!rz>(~11 zMhE|7WkH;mSXR38{CrdGH>a&zHX}-xL$`RYg?@w za5+&YQ~*7gl_=cjdjQhC4dGg8E;OqeaL=I|TpgYSvL|i>!6X6AE(F>D!TQ{mqanpm8@vqedxl_#Jl5FD(x3lOJipgL=;3o8~Mnm z{)_p=%pSK*GZueKoe3~$|+`^J!6}wQgAqOS3s}3{t1WaVk_l)ZN{G&SdKVs6+*kO zCbb6{5LMh$w20)qHHfM=LeySL4rQuERCIxOwYW#{K=kaVR<+3$yMW++Hm5HtZUS`U zn6O$rIIR>^MOq}wCM@J{r?WXr{ZY@ZnwQHgn-J%@2=)%q*Fe>7x3q3J?5{tX=^kh2 z)t()ZV--To-SoR3egE(z19IBlJ^MS7O#CBR<;12W&&ZT)rsSK@tBURfUAH1; z=e&;>QIK#cV%p>hu5AK!7t`*R@XnA6%ByHV{Jc*FrsQlyfepPaU-jNi&RGKwlSepY zmmNG}kKCvjZ-B>=Wx@4ih@FPD69?`=QFO$-w1$$eqR4V8I-Xn$*D9`sH@RL5&lUYI z)QuTN-2(hL1Rt+^=(h$x{+%fa$!RAQGlV0QM7heTq_4 zDw(nApAVCzPV8x98ch!CDfv2zBv8?JCl<}g1l5Zy2$#X)RgtZ{>YfsiNpJQ`30OoP zlCc_N!@IS7s7(jMvHPsd2G#hp;Y+oPTcyV$C)dj-L{1~^0QO)CGlO^vFWYyEkbpk@VFvnQi(K+ ze#GFCrb${&mkD)KV4U9XPFBm&+p9ihv2c`O|?N9#U)+As`Dn7sg; z&08-+Yvd7mT7%b{9u!IDGzFroQqH?1YEfl>_mT*^U$6Hq6U^T`^UmZ-L6=t%4XqXY z(T5o;%ffoVM~;dEw^>b_RQ18>)86!^8h^#rZ=UcD4@k?~lzGq>aI7$3G>n!2!sv!E z+F=EVao?&=UHzu#V^D_TCO0c&xRXv6P*HU-wbV`pjp1jC7Fjg5-x+XE7wWK@bBos^ zes^fI1m(*ejUG#2VrEWoC}7yZ&|vkaj;KN_kjj%i7wB zkzr$*+Fklh6I;q0lfpW$@N_ zkoXA%B+^C%D>Rfml_DFd=qp~`;ibUEz%*-~Tb^f%7>nOJ!C@%!Y6mT?9>}Zcr2n`g zD}jVO*7J`|?84$VJJQ6y`lZ#y1rctP-xvz<;fHX(ETKg8@0IHztDrU#b~bFk3X`m$=Dhyo+7kC$rqw=dycZ{q|C*@{!e$+5!637q zHc{tMH4$`oDyH_66CSTb;j55f^*IZnZT0i5y|IG4Q}*eq<0uYDAS-2iCF)N?+e6le z?G6O~Qn|91EaPHpu`x^s;E-7e3$$z2j<8ktz5ji+$lEY0{jhOdD@k$Uv@*!~jqs=P zC^>Xg&!nP}vUbvSe>ezMpB4_dYZRz`ljMigSnO)Lh`(hTgz4~L8eK891fu@Pc)ZIS zZW`6Nn1{401)7JABO9s#3=b@e0_XFM$gAst;nIglD@D}G*& z1naIakDu@VDWcX?Sk`HYZf2xnGnIS1TNJ~=+DG1XT#Vo5(CI>TiX=q>#}P8Uq+AJ* zLBYS_mD(6M9M)mMy76Ummw7CVqM$SWBn%&Ta<9U z#>!D%xPIj0%+Bz%Ov}wzBqs2rD zCEq}iB$S8G<3lH5^zYil?XZl1gve!40R(4{iesW+2L&Xs`*hW8N4)RR{$SQ)Fxo>L zaE#R$IB#G!#v^~XDLCGeXqg#|cgdd0TIXyNR?NQPUf~8UIKZ{VOvAJGrAUiIfvvFu zIAD_<#vK4=Gmkvl-kUtd5Fhzk{!f2NHorBbgZ)MZuz-?7l}0udUCrCcNf93-xw2k5 zoqIm`0uUJsMR!F9y;J-u-1ajW!W!;;oRTtxcnF&(+@^rKiihQKE;t$?>~#h zUO$ye_B(MbR%--@qm=v*Me3+%)D-Nd^Mn{_UFZ1_-z>7IQL!Bes$;t77gf&0Uzlz-d zOv3wPKvc07H&(*I;@iyQMt1y)|3rhqx%P|pzmetaW-I6Q=?P?pMx=^1QE~_otfQiv zK{iDvzQzE12`oo_!~tbRc$@hAT=h;)I)nY^aQXLtjxS&;bi=fR8AlljHcz2 zl>8V)4pY$w0yWC3oJHi4sF9>VcVJ+hgABwOQW=;cPxaeR--{S<-yyimUFmUB`Iz8- zx)Arvr!)fn@G>kp^;hYSJ|^AaFV0^NZ-S)Ds;G|81nz;aqVNDd{ZA|ZgltNbwO z1RcIJA$^=48kKXf`*kjZp^}!6PiCb~EArYW8syfEUHf&mxV0R4wq3~%w-=v;e$h0J z@3e_QGm-gTNzRlW`jk8Lus+Diq5C)`0(E-Cx=5&D?~-Awm-dh*P8GM&J8o=b*kJtF zf@7@si>$yguFH2@qrYJ?`pwR|HDsF;$JT0$X2}vtzKbG-RJ6upz@tqZt6Z*BqeeM) zuZCtoEs9fueuCnTYXiCzPrc8154k|=@*tEN(D>~Tl=_@dT=9DFu5K#C;kf&P)tea9 z%0QK_H0ZjdQD8*H?%@&Es4+Vc+Gavb?`DOX@nuPBt;B$uqhI;zMY47Rfw347S=dU+ zfs($Niq-)QFiBW8=cXqH5Y(|;9GBrVG75`>goJ)cd2(_(ABluIUbhq#c&dEcLmZuE zvwAE{U!I-!*^PzME`ZieqJI@_0LjCUKS@dFgg0sdM`Fg0M077DzfF-_RCF4>Q<}hq zE`e##as_%8KPAnQGv+CUh1!mjCs1mB~ldYC_gf4_a4jX zqGk#7Jf7T3wm~J6`l4u1n#W)1anKtZ$F1~e7s1o`SU#I;(YY$0EL-@RA1jJw@m-6E+Cy6FEitG^*;t(JXGyy6>DGV_sXiI3Ad zJ7$V@ZhDu^$>ArLqGh!Bv|7bTl_}DAP|K(iHxW>QUayO;a zn&G0hfS_+StzI0QE^L+*K$&S5mU9`Irx2>Jf79-civ>k z*EhsI5*a2a=|4W#Le@C3?^|f(UZqj;%@o;0MX&I`Kdmw($?sb5Ud5nGy>u^>sh$&e z2=G|F824iM>wPtfTQlb4q3)=?V-tn59J()N1QV1wWS13msn5hW{9AwlA%{L(TShXS z7$LPr2q~rH(Cupn@Rt7hl4x=8*>7Nkuaj3YvrC%ij~shULmZY?LNoLlpDN!<0iG@P z>X9WzYE|h>ljIP6PuvsG;L{gTM)eCagqd@*!WIuJo>$TfOEL8PhB4Y`UeUPm7Y0;}fBLV}i0-XH#bTqC(tSz}vXi%|=%Y%F z63s-_#4VBY&j%IJRleIi=BLq+`%GbUbTx!o61Y8d7gHOAm);?{{s#gJL4F|tw47r2 zdu5v?mqp#7F(3fBola-q#_$2oG6q6^d!TA$xA#%yW>;Ngx@#q@gS3iv0Uobs>XmBz z98?Xdoz}$J9)PMnJ0_{?gNnS4LtG~jKCY98zfLv%i6m!67bs3*b!9+6{J2Whi(_Bu?Ak_cSDlLfBy9yqu2j>R2Zzn z82vuaKIWr2|Gw%SgO%|ARmFU=akOZi^Cm&b9wS?@osxr~+g2*N*XsiF$h*RQg|v&= zF#T}ggW-So%sDCQAbUZKPphgAI-!Ucf-3Pxn?Jhy`OP0J;daR|m99~2B74M`h;%4M zEC$3Vv&FN)bHq>cF7MExFYu%WsH9E#uiublCtekm7{MZol5eF*8Wp|Wqijy8f62U$ zkIcI$>Xt&AzzD5M&v%HM-##y%0ij3WK0FXS5jFkbYl|21v*=iP%A^GD zF_PiC&NV%F12FJ4iY=b$ZvDW?L76fv3~H6+M;1mv_%sxi^AJZs>bp^ydN8SgT!EQO? z3hruoOmHu~#J|8_qv(v%N|#O0e)rsG(?5utPz@TtE8R2-{gay{1036D3Cmvph;N!@ zn>LRHyScz=v($dBshVH|imrYAW--Yf&F^($0|NDBBMe9xB?m{i1nc#&00=`CTm33H zYOEr|8jfL+rYl|p9+$(afe2F>k_IJdkLe9xhQ6mYOvpSH8xS(zpPO#E5hEbjqZyO0A;Et5veY@7ApHsRJ^NqFcFi zmR7Yc1h46kH1N-IH}bd7ZU1?NTiiEu#4E9>yQGIspq9F8Rj25FQb$fGP6nc;OAo!} zOXt5|;kM;V`@W0ouD=$_Z$~8g{q3npe!eKlNV9)sp#HJUGdr)B8@QxYeeY62%VU}h+v#R1*k z%VF4uxnf&Y+c+RMcW-LDV1HyYuU@>1(9BtM+UlSV~V8LC%uSo*CwY3w2Vuz8-lTOK@A~OXzGc&I-2l?%y zYfLvkKkOJ;F2lri30)DQQLOP>#n~a)<2}UdRA4K+J>HGpiIZZ-rsl*#h!{B_7LT(H z7w(Ph{i-Q}&T0KzGt52|Lv9!Jc;GPS@lPtBh*o+m7@#xf)+lQOE4cqBdp(Q~R`bbt zhWE=orj?9NiwiR)mczZ-IVaymsg)HI)99hVimB*lBnfjBNy0wSBL2-;LoPKwPrVT{ zPuy2}47s$soq!^SO+o5y9G!HZdlI-Kiyi4j6LD;-g%lMzJp7Qi%F0 z|0=&*xmrHta-8W1MOJpQtk5|w=q_t=he6U;^3)?rc0esN&}oWEKSTTXjVt&1f)jc1XVmEBOt@$ z+N@%6eNg_ZdI`SzMOlB=vZ1&B`PSZP-mEl+hh^C!$>GF7G&7PMSG@Gb0r@U@eAF&& z6F0b`>#-}OIC#sL6-yh&(lW@PnXq>7t-J4MUpU=lI=8%REu47x&CDQKKyynD!|XBU zH?Pm^Vot$)0{wovWG5#-5e+AUwnS!!>7dT6lg7ink#IO^PGv|Z@3ioWtSbjc zc{r>WujQB+efn?|R?M3b@rUaln`+-Wt(R&}%~m`ptv0Bw@8oCkF9h|H8p!$`rYl3X ziMn8Aa!6Xo0rhns{P!5QjsyBIP#Q%Nv!#ZQ|`n$|S zW3<%Gjl=($Z|aJ$N!>VcfX$4PqN5LTpv^IM4N8d2rdP|e!?MFNp$jtWW$o+7;x$)T z3Zd6L#_p#J#`{h;RZwtR8)8PmWCw^a3?%J9$~kCg{znQ4C7= zxqc$a4ol$bODG&4e7AJiO{NsP53}h5Sy_(pXU6~d50eZ_-nMto{*EL$G10czXc3!9 z$$^Zop`y|A9G0T3aBJq3_@I>ju;>TIV^K>oO$ZI@miV7k9#BFpkfbkSn46DxVey!Q zB}c7<$uMe6ht{6^>|nBEk$Z+IOxR2?PRz441FHFJypuVlUbUk1u&b*1X9bT)s&Fm8 z(fguUtHOVC=nhgWJ_q9K*O+sp-#r~>AH7Z#imyA^{8$evdqU{tXC0KT3;uM~{j`Z! zdJHSIv_Q3CoA{x$O}uM{x=)%2oopNw4}1Bw8;)3x3bT)J$O)G@vBZ>EWQLKW4<8J- z0m_D$VwVipiNGP`td^_GL^nYi7^@4hZReX! z6TfU`7AIc1m`M}s`&$7|wgr~J`GE!E29iP6a-W6W$JlDHfcGdMbljnV4+olT!G#lh z9AXkonJ zmM8b_B83wOC>f97r&myNh)C_FqLI&6!0q%nM(TxqvL->|qyzpLfl%h}l_^_6T2$#? zMUg#Bfp1q-t7>mx6DX|33$>~iCe^Q(u7==Mb4VqF&Yb?PZfTjQNs0miT2*D>R{43c z{w&Iv;GOmwAd6qT$->g?uARn%)qLZ;-@bnj6x=awq|Ii&ao)cSAV2bJkW0z4DYA`< z#&~NHUCT|DC3rSKTc;vAms#k#Q(EY+RpD3T*^xfZ@`+kiF4I7oUAty=^U4FNe7Dnh z7PssL@6pg`6XZvGpq-8#e)9WyrpmQ1=WLwV1F11W#B(*PzOE;|(pYW)HEdP+pMxey&uhl3PldUxVzhoxX z?>K-XZ4)*|*hD));`w9Do5L^37XJE;*GhNQDB}3JF#HF-hb2HaPmALpBOvgZnnzcH58ULnile`BMm^2Dp-L#3Xd6mxUcgJZcS=R4<%N?2W zD895pRWTJ+Yr(bvWgeIJc(z^CF4A(e96Zw|P7b~b^jYE0_bSuq5}y)aQud4c zpxhAqrEZ?G%dI}BfUYB;%N>%hh?&*N)6w@uebAHaSv~*+xfRPnjrIhzw~OI#3G9Jj zO|lH)zfi&W`l*dQ@o;EuRP)ID_^+`)b}^u=;L$JtLSjZknG@3uG)5>(pyaD5 zvO=%*eM$g}=Qi=)DJz%-`Wn^>(|jNs&)11fl9}vb zxmy(%!mt=r1iTEm$3$UA%o4!@EdDao>mZGKi&s+!8Vc>O5^`n8e&rL-1@OPu8wRUK z&qDZ^u^m=_>}zinlHfm^G6|eEX>67ou!x_AZKALx<(bJqK}we0k!_Ug#;%O>*?YEh zn4eugRzJ;KiBVjZaew&CSN_{HfX~KqIPo-NCJ&)c45vY33C!wXTV;(07KUR9$J+O| zD6xe{NqF~I0?JY-=}pF10cqSg`p%q-e$PCOq3`V>=}ycbsQmxzy$f7Z>6t%1M?4{U zG2})tIf9A=m>`Z^3>CG}ZKwBbyX|(n+uhmjb~?1%*>2nE_RRD*(+eUhUbrc!fCiA8 zAc9=HpmG)Oqk^KMcm*Olilcyvpz?p7B#b0OG>0T^#{PFcTTag91@nH-^FHr$`F_p3 zitQ8woY%!vGe(CC)wdxzgiJg>d)qQ_JJ^Vm$KD6MUv~qqJlx9B*m*-d`g_l z#|w2+5`Yynk%u(LNa6k9aU(2CxGZ1C-fyR5aR=Ya~tiiRVMIVcud9?4Q+{{#o=nUbauaFA~^p zoSZO=MZYA0X@%`|Jq_kYQ3*So-Y68Rft`1s;-qhfFlWXkagzTd-+Fqvce3DO@cP%; zYsvvqsn{fg-i^ILm(r-lajT=2$jatFfD8hTZXX!AnVu1p+4$2AFfRVm*Z*vpBXU`1 z#Y)zvL%8Ov$%343T#*`i(X;4cev*G2uZPb0$-_BL^gze_-ff=gF`stcNlx~KDy}cs z7;dsJivP!MB4w|PeK~GsU+O4kA4RHh0o&x$AxtDUB6Os>{pLb&-2C4pA<%8E-Zklabr(MdGTd-oO8$2|>zLLEk1Pxgp4f zvqU{+2)aTs(6xLXR;aK+&khnZEs=NEXT2nsUlLgqQNgQ|_0p&2H46{ZDL`L)F}MK~ zKvF_i2)59z5qE*(?y3|iosN+v!DjCi0p7VSB1h9CcnSpPOM{oqIYu7LPL$yJg9No- zb*feFYc#oD*FxjM+q6Txt)e2{@`)MRYQL^GHu5XAD`)2|SRXv(ckIij=GD{j6CZ}9 z398{a)Qz}HZwFqzVfnskI@QjZ6(P3SMrR>x1ZiXZvZ>FX;}ovKMxDsK47v4xxbbHlR&2c1qn zJ?`(n-!V7c91xM;X{)5 zd0u~H&2+>q!>RE@-t5V>->;qjx(PUL z63s4>@50;i{bo?uK`~$n%c$tYP!xL6&94aADrynlkoN|#ccBEdBBV&l;_b_yew8*!6XMy5cVrca|XbS+na`icN6+INu2Sug! z?TgI5a9cA1*5=*vy!qgEeBxvYCtMDFZ@N9PJ{R_Uti<~Es&XW)-lzzUoTPPgH-cx9 zN1k{kKoWBoZx_D;8qEsn!fCeo8+J_8vq!lZm1oa6`{*CNt9M)GuC1hl9}vu@tetep zgEby5^SlA|1HyP-B}n3Q)1BV|I@oSHk09qnktRn|6OpF8{gWiVfrbzDD061Ow?1>V zN0AS_)cIbx^VsRzs)B$%7S3XG4<>W5P$fph9-{H3FP7jVdtjdvPhBHxK$a&>SvMoo zKQ;(;m+@C)T^}wV+n^AN-5|SVQ2r_6p;5B8Mkxp2C{N{JpaQ|kPiWyocVT!Ka)2rDSCSR6Owb#rkITSkf{ zC~n0vH%s*Mn(97aAQ4Q zzNUfK$JYbVdl3(7W3biinCh@o=htF3xJ|dk!%qCj&BCowKFGd)W{A!VzZO~#nr$gjt#bU|#Ly-|X80PfbRays$jh10rtS3EDuN(9 z$@VecTMcjGARji`{8Au0(5Z%arD71B1EI_K;LA$mJvvoQ=c$MyUddW-ylWhS8{B4Z{Ob|diRLtWuIorX z@I83J+XSnBY}j;;taah7Rk<17b14RjCN@#gyVSjOR#39w=wu8P8~f+6f)_R2`;^0O zgXG$j0tgTnsaRZMt}n#QH!df+H7rPkU0T{C2cdAhShH?4xuR<9 z`?rc-|IruS#_V}x&l||m(x<>8T{*u9A8!&JIaUu%F82aBEZgv08WGv!WB+er}+>C__ zhgDXYiGx3(m_dp>97!bsE0FvFsRbt_C9+S6kU)l@dmh7X58lvE|4g+CgNl`K6ZYa? z5_bsGltt2Ew?bN{DV*6&Ckt>L!(gM)pTUk5jLKm96B56qMAp)4Lp#Et()`)sc6Gb# zh&v=5_5`65ZM+U%kRIwQu=%8ccR-OLL$Am97k1F74A-Y*i$oB&B+GrveXx%OBSLH< zkbI8}01{d1iV!%7T`$|!H6Er5VyqYYh3)vW?!-9bica*lSZu=Iip;COBnht!{`-fzZfJfE{u>* z%@A^pVlGppor*@gpQbdnPq)Ym>GXhZ4Z8*QyO`g~01_#qy}uQL0T(?i2<@^MAY&hn zKraUqUZ_Zg3s)u3vDZSW=vFmbpgBv7Yf2sb!~u#ig|#OFli|3@GPX9ME(!KRb%!N_ zXM~;-?$}%8d75@vkMcY$L@b_+<6ZJ7lO#yC(Ay?gg0Ap!Rj1Dh_m~$}5{(aI4p?bk zV3asv>^17wP0KBLnz{74xUl8wx+`<{Ib|4@7R~wW%tCk_mgA zUm_cnZUs*x3*y}=e4T%WkS)@j3?5W&S0l)V-SEE)UL{Uaxp4%xUI3GEU7q^u?>w?> zm$TB97BiS{S zZE1mmMo< z#p5CIWU;6q{FInY53UDXd+3ulJT@Fl%!jE^wHz20BLb(EVRmH#>U+GsWHay#<9HP&MWI8MeV;`^&7!)GDdrJH9vIWiS@g%SEA5|y@ib%Co1u3W zI|gu<+W}EojOU^@?=h0(kLL^NSotdG*S<@4cwrS!iyZ4j*&A?MA&n*EtG%~H zlLa?;C6Hs*`In0KNIT`QEyn$GtVpLBgd$<^ce)Kqt|)(YbM~+s%Ac|Q3jOqM`Wy{K zt$CV!x_VxA_zLAYx=5K%=Rj-t9eI~Bkq5sF2VUiMK!efaM)h6|P|TEG@) z+v`(iN!HE4Ru?cdeabZDRVDW98`3_QAt)B<>Fh}*R0&n?QSMRVR7Ewr&@<+|JLaTy z^wc%Jf6E?mao4RgBv!=5ug@(NC-Gz7*a!M)ZQ4yf_$hr#HI1Kc3$IA^(l%|T|0bWU zvLyaRaiKCPw8+niKJ>E*9uIh)J?*@sF2D12mOQRpV!AF25i1O@%jWb$IKNtRIjS-E zYQ!~7qqJFgm#p(_oqK@St!dMCh@Z&Mit82Sihc7Qkf(}WvXbyenv>#s*$#3aHQ10q5pvpQnnHr~!gSo5rwz98(%m2#F?kvFH;2siIiY zz&lAA0@#zJfE*-cd~lN$$?oi#d8H9Roj%Xl#NtTwI0!dJ067pn!s4%Xf63(Jwtw?x z0jc6v=itKD=9HOB+(a>8erl=c7N$dWdG5;DhkzrZSzhRIMVS~{rns%n;g6iJ2hM_4 z-UVjyY2(6?>LSUy>29LG`oUt+bVJ9Bv9Io%S6v22!uwg~Z z!`Km-rp(}FN%r%T1*rMR9)>2!0zRliO7iH{eCQHJ(ehoqls8x`zA5%#V?-s9gM8eR zx64ks$6|#8()eXbYQ+~8T#;^qg!XblJ)Pu_3dJ96@j3;i8GR5ALRCOL{lNzx*oidG z_!(tCUT|m^x&*!dEBK;Q)GrS@;dzLzrY|serZmBhr+|)k$D>|)dC&lp64j-|Ya0We z1U4{^bUU8=#dDqT=MLt;i`KW#V}$9buDL(|UwgLHF1++u*-~RVpjE|UJ=9j|UNw?e z??v4<*+nE27JO-}C3c}cI6 zn-j@$vv~a@ig`ehKBR2dsXnW+HrAe1Dj;0i%0sR&>;+4SiWOa-iT_-LT4~mbtcLRH7Ge}pk9jYN! zmnuV63$5}7N|iv{!{ zNtLvo?%H-I@*j8puN-7v5ftUWW1fRR4ppm5%XvMS}G%%Zb(d zhneqLwsn6#HtfRNC@Vea@e^x8llVDrUR!tt*0Wu_E=3U^12?yn4T8qGTj&m9C0(gK zthx-_tf${h_{ya>dwy_#;kp^k!VR(~2I)8?aR9-{Wets`~jui3ts+$#k%2s*af&`|)?~+HG7P}S7_}RQJW{9`Zz1aV6 zq6fV#H}+A8$)w%*-mMB!G?5%K+wfLX%ub3_qDe#QOry5+Npe62IYFH!Uy~2*8t7~q z%WZ*usx~x7Gd!m^Y6DOR4ZDpfqTtE_I~mLcVS|aW8W_7`Q7(l&Ldt!>_8E+z<7ywq zxsF(OLdRYfIG^R@n7FQAp=k9}M3}H6c3+rFwsVX5xNxlSgxM=lPcbzV*+WIgidK`| zk+8a=fClcFjxlY%AL^$M5Zs<+(edP*TtEFl@OGbufF%DUf4#89y@y^Kl&>iOp`Jee zN4FMW>H7^{0}mUaMjmPc@j39XV3mIQ#XshI z*s^Mr%bwna*E=iqr1kVlf^@UR{2GsDp^@ISO?zR%L4Jluv44jU31eeKd!wB1Wq)S2 zxPjxd$DDrfhtt!h%rM!Pxc8>6CRyC=W>ZYa6jh}cRxLMw&#p5RC_Y;Rya{YYM=fUKvFj=y~3Vfu;hW8e6RTj=Ezcg`F6q~q`6pHqBFchJqU>mD~} zVV54b*2)^lFC2($)7}ys^u7#9uY$;OW+|{4Rw~P?{?s|AGn19WX=9E{dt5 zNI4a~SzE7IJsBvv1YJB9dDwQScc?FU6oj)GymsY<1v*u0XnSa)(M#AarSWu)$6-aS7%nq8o_hj}k{b=Y7!mBE6uV|Q&Wne+b1Xi_35GEqv+l@m zh)h{G%Y>Xo*NRq?Vizv@J8XuMJro0sja5LG1zYmPRHt(9!utx?9!Kqvc6d~P+p>ma zi)uWVPU#SioEz|5%qxk!KKGn@h~K4HBC4j6MSYR$_}iq9!V;neL$Z8g{Z|W6>!Um< zIk<=3z%LIP3cIbiIeQhk6*gY0G1i2nQH1>JH@{>5@E`y4v;X~X=`xC0LXnuy>;Sie zF@C`~Pgw`bSfS{jpZ-&~Wu}6QSM0(Yt70>#ZKRlV^wOf&cw7is4$e%LBs;uN1OL$= zk^rHCyHk6@?gVA2FNrahGwg;LF$cos`LoH@>O6lm$6ZaxRrUY9$+p~_)9@)d;=;D2 z+ib$mQ_NY4oTj1|i>?J<1?`4bZO)sQz|Ne}LVcESfDA-#p@+p7qSmR-$Ug;I%3ZUt zCaXU(Mzn6`Y5{sHZQ8R_V?}4h*{WDkIhi@ zOwY7)XHMJG3GTvsE-RhjXQ!6ZPXprSH~B6XU1~B|05Y1A;r0tKC>7lbPyw}stD#|WxhRQeoVfW1r0=iHA#-Sse$*jNfvMTcs`CjbDRu->$(pA88quJe_PR@KK`~q{Qp*M(%0&~ zvS|^nkxk?w*%tWJZCl`8cl_V|5v|%SRMMjNm;U`3ZB`I2qP=9Jx>b$^&G}Gl2!%p) zQQ+E1<@93SD(Co4jsX7H(6#UEU*0`i{I^w1*r@)?X-Ss4xv+n0CDvEQuZL9Q z^6&zJJFQAyn|9;WgAnm8&|LQ{=5N+)_+cWPJ5RclRX)c1{#?oH)^^eONbZ}L4UZTv z8_(r0!s`#{!iWNL_`7)Bp|G<7O*Z>Rj)azVH0}A>liPBWEjjv~zrRe@aLYQlZiXME z=|)?g%@hL+#T%(;BRzSnyM8(r&1~Vd$dNGl zc;|QBQLMDKg_<6z4^BDUlH8 zo(Eeky`@ARgARJobI`Mok9#g7R}PkJ9ad!vQ2Huauw9J>k;u?zpsNM?zdYPE!pgT~ z=6P{3=!!cGRpUjEHH+jZ1Reb9vN>OL(Qo7!#*$crONya@YouP59Hj?ntO|DE9NPAs z+dk8_jpwMltS>WTWC|v{{qC|emW%Nax6g#|0TA1HgJss85or9hcVQ2^ds<1upYv(a#DeD1ok5n3Gw zWP8;Wb4wye{|@7sGumSQ1L*J9M+}5Lf#05@#J}K@xSW0*TnYMj4N(RBn~-q4Djg!b z_$84Sf^((k#Y5zZ;9(Tr(WV{!j84_&Tf;vX(Wg8_|A{SMw3$xgC-Pp6&KTDXe0;EI zEp3??uu@;ytXLa_r3HgtS3z_TwcS|c0Pg`thm!_@UPD&Ou$OI>_l04G30oCf^RI~6`cJ*A(XzhAW%X07RLtaQF!X|& znW&8m@fDq>+^0jh&$|GSJ?w@ux#yLrI#wu4^>?Z#GYU?eaXac+XC89G?Sv^+zv{84 zDZz!~c~+Vdu*Mf97JBIQyerb$P~@~vn}qS1E>JGSe0;lXUHA_5%fb<0fg6;Z2_EyY za~^rlnPPHz0{^Gx9kOApl3>?$*$6cAjdJF$W9f)&I=5go?{9Gge`NQG2X=F-Bb3X7^h0110+6*inn-0hGOTs#!(cZTZH@3A3Ds40~V!AaN`fMvZ127GuQmUOPtK&xs>zOAZD2! z+$_8*yzGNlpSpFCgY>#U<8gEFnlIx~t+p?$qNJBbf;FA0ANUt?y|7ol z#z|}n&%TOrf5Fd&aLN%US6FKfEVxM&hq)R9dt?Bwn>GZ)(^1`+{!pnyrb?qZtTo^A0&G53D zVxWMk0_503>Y!CIg_k8Aq;o*765D5yR`M`ip-2WMqN0!`^ACesT`Y5NQnRAnM?d|> zoF%XsZswmCZ=pBL+%Jk{9*T-1PE5Yb^zqNg^F^iNyK^2yZc-oq>}iK0$RW@%!N`#a z;shP4?D$8@@0g(TQPG~8|H5lw ziv>E>UUkg^V>dU%TU6;}rEj?;MRI*ktM=CS?*Hh_q8nd5_IAOW*yDyBO3mW6p?679 z;Gk+dzfC&~F&<>1$k!~HkAF)Cofw)P+{FJhlGC2ck!!v`gMNak zC-||VO0ZqopeA=sIY4^pKX)pZ3+_OXIS#RglmItYl5cS!4&YFO#5ET=HX4j)ZO4Qy zcXGqc&^H4XOH3vyN4V}9NpxY81bk?tOp=abGAXiwiasgnmR#bOit*?BMbst!R_gsy z;CMx|aUb|x_fOmG-y?3**1(!qsdjw$*2=7m$FNu#%Au#Z0i-VN@HESQSu3g)&59*K z3C!|oy&xm-$hXELCZrg`=xwklT;$a(V7KwNd+bo@r#tvPeBO7%Z3><@l4GxNngZ9| zI_3ZBb+6ebg`JQ^+TdFg)xhfzc943y zN_q^mPeH#Y;vVm=YzzH>==it73OzoJJWkM8xjDN@(53~AAK|Wfn>962{d6`{KsQg( zsWt$`^|kr;Av=M3y)C|1)s;~fXX#XHyl+9-X}8c}?+LfJ;CU;RW7n&Hi@EW?mO9_B zNS}%o1@O(h9(v_mHjk7k3L*EjEuw_l3(~t-Hj0fZg><^+7_feJd%U~Os~H+}yG#Vy z@f7~&FhGoL%*GFsvBz`N+5cHK$=$Lr(@Op08t?lGB#5e2uVipfTkdn2UKZFy_V^wG z2Hht9Azmr};;b5vd(+Eaf?|c~3*h>_I_i6S(KwtIA=fq4k1o;wJ<+rXY4z{1Wbs63 z^ciivQz#~pBCEi9!+X}H+~__OmPMZz*M@F%Kjob^$;q$PnA2brCdPcyxhFYc!gU=J zc8L1d*G!nuT_Ke&rpjy8Y8I`lR}Y1DjM6}jhR0-W=@vm zsbaS~)?^k*3!!`dWh3UrP%*id*5=en=dD9!<-cE<<7R?I$)jI=L}I>Ru+W*oBAH?m zD6*1@ZkHL_>GhEQO`E;+^*E@_c1}j;MKfj-BwjR#6Rul_#IJ5BW?Sk>S`lLGQx-vO zppo?i6>#_76r|gQ1MEx9+O!{1Og}~LQPCYjqgCprvwiwyC7!wb zEi-e~)qaH{teg4piWHl~47!B4?jRxeCZA$SH;v^5^+6qBl@To}wti9%eVt#e+UbjI z0PI@jVo}+`5B4s+K^`f}l)$l|S7bmVQMC%kMJV^7ExX(4Y3 z@OfM#=OF=rWrrn{~I5}==TKHb|Ms>4X zPxsPzKeC3~V*p=l#mtvdPM^kMy+$74WW^|+sQ5!mPuXSD%T~PNUg0*sN3!~0@XqQ4 zCw-w-5-pTHb35`~a{OrMXl9J6D5jhu zrBw9Ms0tsbB?oCG;25VH=m9a3)3wWbqv~Xbpft3C$$~|zR$LsYQ!SZ;KevQuLe*3T zujm_Bq_}#uX{$Y#h${HlR+1qb2nLU|MUts#)Udx31{EL5qred~>^AI{HK{TZdy?_^rSE8M|3*jsb6>(Cr`OhX zd(f}U%F`hD!cr zf|HtDuMXM>ZeuPm)gP~87f~fKDClNO9zZ!$Z@|3}BuZqj`qa{RFhHv3RY!q5yR1q2 z$s60%sGiGWnR<;!kNEmrY=-?yZNDr`s*lK`m ze-CW2E;ig_<2z*iJ-E602lbYTW|#F|tfY=Be737ksN=n3LLP#aU+Ub0e&@w{VYgy2 z@2CdnJl(e!3c0aJZmYCIn6?OdSv!^Wz^gSV-R^XXj&Vc6#;G0eZPp<%Mo@{mS?@pW zY4T{QzxM4@WYt8_-y9uy&!m_Q6#QeOPt9uv`uuZ1o}U5RN&JLM|6v;IVy{RaOYz#u z`5B_V09)Ux@%|H^J;!kpo;~Q~BW{!6vXQ;2*grH_Zh>BrY85LRCJbg{kqU;Mvw*!3 z`Osb-oJXP37G#gQ{iPlLa8uADZRGK-LVzMdFkch5T4~kO5 z;=)IMrYUPeH_HaWa@Es%MQ$XjyLO9)NTQ@$QNYiTb%?V)bgFpo{BY+(!IpU+!+&u4 z{ByHt`sA{YESD9R^**eucbl0`#U7|sV{gs5<#7uDz!uKj=)Mt}PBH7%Ovh90vJ|gc z{}1pc=L5qq!`#u~bYK|&E>5dLN$kW8mdlD2dyEcyD|BT6BXKvKz?3OYykWc7?#O?; zP2apDf1!*$h%62K)eZl*JizOJi=##(n2)*qX z;69J`JYgLJW9(&L^ZU+`FB6jusc_3aPL{i{YYJqOqg>PV6q8Dk6e>DV(&?kqTnD+; zD!)D@)X}6Vk?H~ow>nst4LC1E-jN7i*4 z1c(*h_oWk;OxfbT)MHl^)&YQ9EiCYl3(V7WDfPeycgDBhqja*~r&as(Tibtfb79Xn zP*h-pZ-Z~XC@xShiVyA-@8$JL-@Es<)qj`&)l+Ylef{@YLC&y7Hsi~>W8xHR^k47& z^xrL+Y+n(?>@-`DKBGhK4tsyuoK?zdsynR9r%=+R_}qzBXgNM(kk@@kdcqok#;l8n zt;Z*Qn}yAadxA!C)|uO5`=DaC#j!d-Tt~}V%N`wah{!vz~>`T2u5;VQQU=gn^)UJyAFJl(Z5n^d~6huUIh5cX3H@V3<$ zceL14hWk=X#iBm_UH1z>X_6V6It9)WAF`%KfgNn`#Pl&oa>sn}%f12_#?;fa<)zeP} z7ZzOiWV3llFRC{vk72eqO}WxLJv5t_Xkfd(D({c11SqjZybb(M#EHC5#9KiH&oTDO3FxsCH^#>~q31QpzPA-(lMPytdG(hh zfm;om>&6htRx_+^qL>Vdtf!(6OHuv>u%bLf@yIP-0GboVX=ZJP1&Y`vivk?YC1ytJuB3mG`ov` zPcZU`PE{Ma+SgI$!*bfV1K@LOkL8dK`OL|jOz>MjX@O-f$7L5ID@mpjL5EkNERmPP zUpJH8?%l=cy^>#gNCoH4CP+ABlDQ4S4HB=PTFhAHa$FV?R`NEf{>g%z8683vozFbY zo*>xy$lt~4+(O0BB^vGV;F*_2=Z|a4gUu-m9ID}Jf;AS+=y07WB zByhCClh_wF==ngI8NM%gS-4IG>Po_0{KNDU->q~xjpqxYU9*kZ06WN5)eh3EcuDi} z;(7Q#c+X&FIn9IX_9ZlZ@1IHnO{>ap%bJo%?klsZG?=X_l@zm`BHO6w1Hp&YO}_QO z_!T4C8->x7mB31U*Qc0|n2!@|p;7eb`rIDnJ{KEi~`MBqNndHJpSgsuWtFieqs8ekAJ-4%@4{H z$g?y+bi5tvc>hEJdyM4F(9@+ek=mn=U**>y5F4bQ4%_56@BL!^o7LZC4+Pgh=+@X7 z-z*&PTpBza2%YMxRM(l4`%l(}%)9a&i7r?~;WViTSf_+ST)44a-ESZ?SrR6ADK(5x-Dv zs!eDK(i9wLw7Cg z0hYO#`Op|7!MePizHxG$7mJZ1)S*W?C1-?-ZM3pux=Eu>zCzNA+jxiVb7ZzZSNvFcf??SphxEpwzP~r}k zpK@M1gCXZu?;&zode^5aY8Qxu+>qm)pm`VGsNN?Wbg0<8E!Y`lP@aG2K=4feUDFMF zlGd&ps<4){uJF;*`=MZFgADdrflLfhE79@>g)=(q~WL{K-Gq=!-J+99^u8MVU)7Gkw z^EUWCoVvt)sr${@FKN<0``&Gy{m-6wIS1A^d*X+`_nl+%T;D!DcN;06NKTr4)eRH_ z-r?TSb#0(GnyMM%Zx`rgd+2^(4cV+^v4^W&b|8Z7QVtmk)gA^J(|~NMe~)}n**{&U zS(#JPOrG^yZS=JkX%$Z9+eMzkr|98YN903+WR&>7Yu?~&)z=qIEpkGSh; z`!a+!Ps{Vo^_W-fnLz8)GFJDF%aR~eeE18$sU{5R7ld6QnPasky6_4EGA^SMky|Nd z3q|s&=mB!v3vrstZ&hKSa0xFBNRv?mwiZgHjC51^UWK$?)+=A@jl{l#o_O{;{|tTA zrw@pH@w{_MsqBQdJ?MSzv(}+~R?w{sGodBNKW`l=<(7JIVH|yIhND`F0cF{iicCZAp!%O7-Qo9hEvXZ@Y?E9VKvswr3V}kXn?4u$7`8!io;QU+xzGj+BObClCS4(j1@=m4 z=j3*oopb(zuYS~J?&uixuoDkAq#kUUWil^|t`)5&#V#BWK5S+%_D~EITUAlfH{^T5 zn*t?iR*GtOyMXDP zw&TK<(HwUVCtOTu{KNhdOMX%=OO6ZgtF6$iR?n*n{E)r9=;Ff5Qsj8*i)_5 z-dVEYO}6Q+MBd9plI<`8bK*FZyl|w_-UV(v(`xhEBM%OfCET1@7mniRn5|7~DJGdB z2~;!^E*Pa`z&8aMR}XgMh?X7>8f|f5C0UvS!z1mk}#fxcQns zr5;##wm@6ia?vfNX`4|j!AO1K%*UQUzpOFdl)&`RmqKsI&bcw4DZ?krZ`Z!h+~&q*BW(l4-}SdNEMKT7z7=D$hQx^SJR#t#7zpc9 z-c+CD*G$okL_bv&@@`1Gg-zdCJIR4s$j`ynxB&f}@f>x@I-q5$ z|5|FvYt6-1bzzsp3JW$Wb|A+OP#|9QSuCoIY*hgtL$6Ok_$e{l7;s69?b6td&l;n( zE%FECF5ReZk)!M4WVdBp05aJ_M}cLmaW)23{Hvu>hZTtnC@BSzx(B3ITp5`S4FSa> zD6a(fV5ct%QeO@1;}`lNLr#`|zUVOB9=cb3LW-@#J>p?0p6}zQ@{cNO1o#>6(5VK< zWuIpEEJ>c@i>K}RzDF-ZV|{1#oU{JIjuD7(-4JDIcGMKcNk;0i6AK9nrC4V|CDdJ%@H;Nz(yS2!XMeM`Vl6b-zUEOxsF8PyLTh$%H z2K8otq=ZWKKk2lVX9L3QgRngyeg3nYAUq+)uQJFU?l2c#hpcdiVMR)*=Z-+V;QOG;zKg46PN-0sYpS~lF5iJq5%es_@r3VBx9!XGC z%N}&sgX~>FxKkD<`|velB%hC!@m&4h!GHSAY?EzyuSVNIO2*0~yD%<}n%SFu6a$In zome=54oDWA6~aD}ofqQ<8mnkgmJS%RYsJZeOQL3hk=u0GZ7;|gRegQa^g$3LteXpD zhP$S%jQ}wqnlv}=V(6bwI(%3ArSA%!62i1v^gYH~JTb?&bjseLC zSRCijST{l^TJ^uRYAy2b-J zV(hEdg6SR)f-FqU*dOhLQrO2m(=EgOE*sdj62QlaKO4SyYBB?B3xsmjYol1S6Ya8? zmrm(o^E}%*Z(JX@zQ~NVS*vXM^8@LKUHR)*TS<}&?}@gVS&A%*0ln{ZDjExspcRbQ z5r&jd89^T{78Ogd{T2BIA!eLKulCrf*vvrc2YFu3MxeBn-S7E}=f!4(n7{8Y9@x%1 zI3Z-hEa9CbOX77du89lhkFC(OFP{cfmf?CKGz^AcR`!VRDs-w7q7M2H{bbf%kGut| zUkYK%vq|QFqh}AjgrlGSt{^PJWIM#}3vu`S?gAem_g8B}Hn8ztwY=tXovDtXXz!1K1}(SRli zTjOy}d|i1lc>U`Gq*49Y8}~99LHFRH0dgA{Fwy3;heD%Ja2F`v*q^%M019*S^88PB z!-kV_d2PkN$F&EWAXL#)9!mhvjsaaeBTbHMb-R`X_ z4A+4`ZX|LqW2i-^>VlR+Myk`XVzfh3<{Y zoys1f@kfJTzb~5sJ*SQ8Y+f0u5kQMmdazD~V?sa=?kJnz6LDRdpk;STdwrlp;}ES2 z@AkSOZxVorbDL^FR_KQ}XVKV4-=(|@;lTkJi`(liD7JwC)0DMzZ@m9Fb#hSD^6T*0p}0%49m1R$mt2<; zMn1DQKw&~&*0f*$mZZ6`LsV(Da^+JDNYi9d(YVD%js)n1tEMY?UA$%m8#i@%cme5v z)@6`0Q|z191DT}E`O7>F3Pwe|Y?v$-2@^6^n9g>l72;rXKJVu^(8gE}Wqq9*Yt#L@ z{*chugp;>UzIKLWxG+wD(tT86S24vDQY4Rx?hxkkn}zk79p0@G^>mxI0Lp8(M)mWz zMs$#RdLXh=yEUp+n-P>p7xK;k%Zpxikl(~_(_WPx2Nu<4VUM_w_h9x@r4!L&7g(HF ziOUHr)X?VDTFVHHl|aaG)nyHvM`}Z}0)g+V#-jynC$g8IioSv5Z$0U1Y}ds&UFZz7q!7|>alMh zKZ`C^H1W4a9T!$8$|W_SojzSaT4lUwMBOjZj9!DQAjU5ae_9$iXb1e5rH#L26?`aOr|6h^dZi+GQ{ zkCXJ^bPd$RYI=eOVaa^}B(pUhE=xPL`8}q3k_m=4{_E18$qE-<)nE}DwW_6445UKV zQqgEB`y<(2`Q51}!Fa|6W(ZO6rpTvQ@?dt(jLXu?P^RpZ!Uk)`jhpSXHj3lVbF-aE z{oBJV=?`5t`DBHc(a184C4_82IQFTv%Sx4{Q=Pwjes1z?!^GzvbNUHRn0W0wKYvnX zkKCT?F338h$xECsj{6@RW*VP+ko}?ot5#-$S98g`QM#qPkza?w6uToW&x?1 zNKTo3$4wLic_9v|onwlSRA~3Qp%}hUW|_579ZEdjiT8 z1)2lF3BVJX$lDRPeB!S8H)rpjUKA2P5lZ(Xdp!p|ivkCg12mP`c zJcE47C#LL_?VjGIedK#TqL;oAaYmjM;D)&X36BGlb7X7_naxNv=)l{j7&tp^D7 z&?O5|v$0r$B~eXetE!1q1wh;m7_6WQvqZLUCi0yQyCEk>6Y2G7B8Pyo+vy(e_zm4> zQ|goh>o}o*^@C2@v;;l)q5m-w%WaeG!U)PXTS?bYOcF&_Q_%(C`}ru-nIUHfJa}K;gPeB>@_-Q5)n8idj#Q zRFohn;UUpoD-<#G@e`OM{HwiTgvoB>QsrmbzW^u zcK8h_W4OVqjKnHkkPC-=kskIe`SfF%6I~H*({Je`J8-@&hO)ktjFA$e<_=8`v}|c~ z*#L=^Udw(uo2e&H0+H3WS$Nl{o~{*-b*@cY5wyc}7YGBpZIEFC=^U?@Mq!(~ASKd4 zVy3mIjL-H8dZxBIp8Cf=*gS)g#jM-?ksE9#FY1}-XIg-sFvnI9oeRfYp{{?_l2S@B zAnCgW%6GvtMPka+vOa0K;w0FOB7Q$CC;Rzb&<0XIwN3i~!tlT&$G=FhlmPGT0d}wh z!Dwe{=>d;gagGKn29^nqh#G^PWNYjPjNRB6a{z9D>H6PRp=F;Mmug-Yc4Mt{ut856 zMA8cB?y#B#Kt9+a-^^P%_Y?>tVWUVHd8i&#exgX#=;rPV>N23~j_;mnE9Vsj4!eO~ zkp>lujB25%0nw#QVD@OXMGU)@ExbXD=Wp<`!@HE%`HMx4+cwz`wr3W~a5N@#xS#x& zMMH1@#$<1H|A&{H+#0K1(}h8@!YrXUL@@&txsQ7t4CrGG3vw>B%X*`r*$B&BP%U~> zQ~}5frbHG6LS@p~P$Q?|t&nELChx6k?7wbSED8Epl`*-+8==Ntj!Jg#<6q}@YZ9U| z=NC)*_L%%Y0GhYeIne5YTxO6rz&iAp`RI0E4$Or9+FHXy4`x@-V;j}SU_C_w=!_}t zp_P$|p=Sb;1^5&;)H%*x87zq-jTU#)Wc`T^Z!W3*+X~$rh|Cbiik>Lh)gEZq&}tRZ zXQv+hKL`oi2ny@7=cFqU-PbDLFs)?OGyk}pY<*=yH%H9;?7b8NC44)m==kt1Uccf9 zNmka&o+_#oO}J)>UJfn zdD0e8wI1{u;&(zbRA%_KQ1r^n_`CQ6vMn?g{L9T2Ni9=4KnD3pvj__`ClXGljLiP z0-?gc$~|LV+N5{;V4szzX_p;WWxM0!tvsw_EuoD6b$eX!y6Eu%{vQWZGhUcjwIegR&6D_xmg|;PC1-4gWzF`f#$kKVC`LyBRMs_FS1{z z7d{c3@g1bs@!#w5)%)g1T2#Lu>uRg^j5bYq(if`CO}82QFRP@jbKA6sB)xQ9NWbif zpjBJ+FSTDeqdh0UQ8sJ0l5^@DNj+)RuK)VF@ANK=`Tm84NsHdk_gKZSp|AvvUX~ImW)(%^jlxbHUb{mIY2@>3V_rgA z|5#tJ4O2YUV@^NE&9?l@;s37oF#)3WXFJ~|=@SX4I*f7$iYTUlBDqvF_Cs&-$>71N zAK6Uc_;dp1nT za}BtiB~w=aI8IaIx_;uBa>Zb%2`ekodbg8pF6`VKGXrrA#q6N~*wN(+PlI$(n>L%5 z?7cyBp4TBfDXD~b-Z9ZJQH{qjptgewKShx;WAa|jfae-_h znZPH3*V<((W#unRbAmfOxJ~nT@8$%L*LM77eTHS8%1VM~hayq3(f^Z>n*paIV|_b> z6+YMvpX!fs`UgC{?5R(V=iry_f!6%?sO>(-RSyWBX@k0j&A~aIc#Mrd`=S_j|Dm`10Ex|6OwFZV6RLCmyLzgR&2Q=5Sdjke!eIRNY zBT5Uot~BhBKLDOINMsd&yM{YyXrTwvO+Bp-da6hZ2LhY0n&}(eH+t(P-&Dj?_T$^p zz&63$5~rL$a{~5+FE9FIq-Af4%a-F<={8A}B>6w`?eMBlLWYu5LK=L~6GW!>1f?kl zr3v9hnnOXr($Gmi^n&DMUGM;q6k#G0&1O3@;5i`fQogf7-r@C*v-VW>LxnR6Y;LeyeW7TqJR6($Sn=^X?j zzBw~i@eYwJRbpthUvA_nx>S5a{)za5k-OWK_4Ku{#L&H(WWixYEXagr3qEKL9SZBA zaj;^)RnTaAMOrowg);27L(gW~ewd9~YM%{j_mN(a6aBZv`HS6ec^WJq27-DJ_k)lD zD!oFxZH8=FXsH-6q^FZ-uJ!IzVAi@_?noS1ov`r$#Oh!UyzsM@iSGxREX!}pnvzKF zSj9H3>x2>L$QuaZjM9`vcx-)AsSJ{FY8E zn|DR}iRZG3+dMWwU;P%kGvaDwtb0~aB5#=hYek!dn6kvpHs0AN#F6yEZbhYfr`pkg z84or_nTuzK4#WixWqplW^1=2nOSC5&o@9Z_bJjl5}qDOJswu%1;hbA9=lY4w7^))hdE6soPvl=ktRcXj7jpW zgl$u*Kb+e|_Nbzkz8le{Dx{kP!*1D; zYop$6k>_hL!G1@My6Tc%FNV6l^JKuYQc@C$3hH0L zV!pAU_g!U)yVe9fefYEtve$*7ciwDuJ4P{}h<$*H{!=HkR^C?>^EY{C1N-!yzuN>c zr~7o5-;+SdLi!y8dYV6XDnGoh7@WV!XXTW*saVc=1FE!cMC_6^P1mV*g`e@=FfVUD zW-;5e8|LY~Fe+6VQ7i5duW`=}&z4@FyIr%6UlUqQ#ez&n@;tn=mxfXQbYIaS-XSf} zp5t}eZ;jWu7DVS5GBc|-F0;nTkWo`qzW-sFDz}njzAEgO0ilp*i}1azBuDao z28s1Iq?eyvc$QQ)yTdZGoFCpDc1{olePi{_F4GdFSN6BrQ6l zeKf05+aSo16!I|ByqQ|^^{lVI|Ky(6Z1R*HdM&U`AmA4C$8B$vcb1D+ijZA_YSgKwuC_uatUM|vP$ zRKPDH-B1shrfhv<2Z0tW@!6>>K@?91OeKe%SVd|ljK>90TOQ&D>Wc6dVFd_SUZ2yZ9XVG( z-|%bG-U{0)ZRWp}Z`!u+*8X*#>l|=j0uI{&%Vkvi@8ty838kl3?a`QU zaWwibYe@yS6oU)*<29S1sF7moD6$WGg42|J%3_p7M$s=|`Gm%~8Qw;)AlZ02U(*IOYI=HWc&ln_c)aMe|FGNE@NW1# zK#qH%YARN>pO|`Ls`0bkcGPyrncK-1eX{-o=bu`ID>k->h7xkr&Z zRCG1HSd=Zu;0@Aw3mP;?lh$MCY-;(%g|8f2$Zm^B7HsC_`>=U%-vGgIA@+f1fTAsw zu-P_eN=`sv6#&_O_;P+1Bx?Vt#hC(Wn36$T}7aD!t+YeKL18)&x9D9$Kqrq@ho zGnMH;p?}aX&hw^_J;ofK0%6EEcKg)}3TJlE&Cs8hEMSp^E7`qWywkT>giYqgKQd2~ z9gg&F8{LP(47oCZvnIwPp2GjoYE{lUNZ783RbIx{^kKJ=b=Fqcy~y@B@VWuVKMMYJ zprsZemquO}wnbKy3-e$>GYTW4Eix=@<-B&L0Gh<@UU9gZ23t|VbzFPz&)NLT4=u?B zTsCuMg-D=;-%A(pFGIOtLU`tUkUqglzv@#9ROLFadeBuFb}N*W@za!;!mJfH2DdA* z6#kLo5%g6iFej$wfZXzi$=Ad3HMO*12%}C4)aHO>0pO&TfmcD*5Q%EH@;1wQlqtYa zpBa)pr!lyLZs7IsuF2EMHt!wM9nzO!lx(#_W~-0AM#hHct#8i$Ibi92drEd(c-^uh z(+x7aP)LgOvU(xAVQMYE5}cSqdcW_5nQW2jY4Bn7VKwA4XQp|SDS%$2ZtlyWhjdlw z+>=!LS~3hn|MaMEd^DdT#dO6P&eFS>z8Tk zc#gWy`g=H5?$~9cdLRDYXXO_?ca$v3_X6hw%Mhx8Oy<^{?Yw%$-arV?ATv>Q6p(DG z()cw{+JGc9H7XX%88STp(>-*(qE_4*x^6xc8^gZl^*F{!5@EK&*lanm_~0n$jvj{- zj3=zn{B`aOlN%EE-qh73Ya#)u+EEsyjAFo87a5COmV{&k?Gt`7=MbGtYK6U|VQQ9d z3beW$jlvGiPv#7GB9j#A*Wz8kgFpEp*)?x~w8}U6Y*JqgZUq;vX2#a2L-b*~Tj-os z-vJXd`ulX)F{g=f-JG1}iLCvL&NHj0G1AiD_rw=#vM}U(F`!0;+_IVBXXe#-U=?2> z4Y~{jEs2qduTp$Yig#jzF=|{Qr&jSJX)MO{?IgYgUrO^uB32)zXO-;)~)~QK@(t zsCWQh2(%`2&EGd|v8YYer$_}fl}{_5R=@{Sr;ULkeaHu13VxY~WO=&+yWQJ3|p+nMlm%{jUy6e!eXOG3^8T`wyPuOh4#(dg&C+D49A81(|;j*DWD^(j8dHs>io(1K5Ar^^(n+5)kxREV?6yJfe|U7!v?{fvh^8UJ=FChGNr9vS3~0 z4hfXycn`Z(OK&NQcmo6#=DHQ&U(ZBMv|r~5m?1Aw z(S6E6~LaS%FeWg~1argyJ3xMlxGk1I3XTD{!Kfz09{g@o&W`A7wUq-sj?9W+> zIZcsfD!OE0r*bg`{cS>{Em#^{8F3*bTY5R-)c?oco4_@dp8Mk-@f?yDLqb9@IRgk1 zK_ZSUh6;G&O!s?lm;3AV-rn20w9^h<>^kjDr|oS!xG>-XZlEFxQGp2J0w{{Is)&jU zI5dRu{2@s)o#MHY{PRKwsIA)8{?Q1I~+q1<+I zr>0$`kH);EWe?i5#kvO~6sspcy*^&NZ3MI6nV`|p_ zJ4`-$ZOl%d4aiqf>`IC(1AbZ2S}iimB6;ae$zo;NggG^mZIXwF@HgHm z(Bgw{r~E+&AM&N_b84v-7>W}afD(IQRDz}&ZFm^uPK(DejCdRI+SyTP)(h$sS0YA8)(DC z1@ZwVc>LWgN2ecIg&z5Bjlm>u)VVPvF4;h$nPN{+q!I0&fnLKbkXvXvF>-r~)(bnM(uL_vh5%LJP-DeJ zW6=YeyaJ#|TknsCuQO`f1ii2Y+T5@>{6@e9sZr1+uN5_zJ1>(GD&l)$7ikx1b5vdO z4djBPU4%w)p7>r|b7ZEn7zF7)4Xgv3*MTHCxJ+uNTV>dY@&Sed`G{;-hTFyV@2P?0 z3oc%m8vCtTHI4QeCq7wfHxB=DM|H&5dhz3jefB&5wJzMx zYJuLVTYrYE@9k;tIoRc~ z8V&;2*!c%EWtm=iK5Iprd`p-s{%z11jl4t9rikI*Q!o#>?)zr@j~d2%wS9n#_V`(L zx2+Zb!NS^i?89%b$&|wxMXZ^L{rNh~V6@6EKxjpWI^S3Po8^7Jm!oUJrKyfABe`@{ z&}jL`yg}|FP}@It_05%v|FBLmr>C78L(cM>V%&E{NABB9G1n>fDn+hPiH#%;*c(vw zJtG9PxWRpeSbkCjR>jyF9q%4N5$TOeSAQ-?2d3X+uLc7Ld76Ap6P@H&9|&#O*qrmhF8Je1y!6qQ zghL~-d;}za1nN+f6xVy|kk@f{csG4|a#JK^D_dnL;u}hH(z05Bg&G6ticT8u8=~`p zH`bmyzs}vT$J(#UGlK2OhuyY-hJ$GMm;@QjQq>p5hk^?yw}$p9*Rt=uw_#FwTqC20 zCjDy3L8v)DKdnrFXD>*qW?cb>vm(Jgg2cq>j8S`tV5;Yd@NnF6dD(cpKS%XROucZw z+S#_qX*jSS$Pw=Inpi2{^FO8Sby3|GHNZwaC^EBgKQe1XpAX2wOaGt^0 z$r3N@_pme_U{B<|SZqDO@Z!kgg^rHVFhSzzwjpnRB3eI8T3lGi*Ir#)H zcu=SH-~Yy|uH91sYqQRn)A1)XZIV7K`m8GnL)`2Zft_4O8>MpEyMD#*MNZRcY8Va+p|d zaT3D#;^*Zah;x9%4?hk4Ur^f1Li7hZvm7g>n?r{$FYiceG{1IU~*E zDZD^2=B+PG(b=o?xGg9+bazgJ;&GpMOYGu+jQHwE%O2zdM7ziWCM*22X&+B(h<)PG zIP1);Cmvbhcc!I^pLnc;Cdx+%us=2LjVB%z!MD}f0fo#VQW4xZ>v+)OfEMkRw@beC zhqSL1eQUVX(_CZbL~FwJCx&&L7jUS|AFpY#r>b&W;Bb&nVQ0Q2yn?Q(ZfVFtp$@B9 zFuCKTsl(xsa2OR1KhZfC@j0_X4Z0;0VBfAmRNVdgvDEs^xpycWvnde)~U|m9!JX_pJr?Dx~ zDbc#4P%Zf=;Zed#`Xr5nWV%#wBgxYo3QiZ6vkMZO3@KiydEr84q&Ts?pBHMzzMH!9 z&IGGPX?^p|`y|7SElR14MaiRBm=SZRL|nc=HV1UFfp*ZG$$3ZvRalOuRaOSID8-be zeW+EoAfXEi`?k)^R+UB|M`nwvm~w5F*x?y*2!PlQtF!m`!Q|IFJr~(G3cXUj(Hvb( z$KJjqJpjVBIU%J}tGx2*WFXKifi|>0dA(wO#3n(Lix`WzMw5q8;d+Eij&q~s|K&<$ zTB`vWP}Jg+PfdJg@qVp+9hdqj_*B7W@zpjBq>O@E>=YCrYwQCy0<8D)ZYm8;tK zPHB4m$`Z~@%cXAt4@-@@T)G+bYBMLbXy3zM?`)N!Q1#_F%3*i4G7pc3zTmnh%Q4&3G6O#cqIk7`axLGm>s zJ39A!i@x}9wuQdm$&>zz{!RSe>5HD^+;iOD-62}i-S_L5w?8}DzWmK?<0KpfZSA7D z;_~T#NG0h^-mIk%qTdZf&DkNHF($=QWtGVx| z71TxXB3Xyx{T`xcDko-&yClD}9A72qiS1C_m2Z{}T+^bx@}tyTv^M(4V*Wt zM!g5}$>v%k>G*j0C_09qmp{kZS3LgL>5(B;LsH!Ty>g;|#n{@bZH&tf ziiNJkA}aCT>`PL;|DAx`kZqDKdDWzbnT>2q%+N=(L-2Tx>d#k{)zHg6KLR#R!$!O4 zs^3LOH2;5ui}%u@JT5DjXd>n5a4``^iF zI~c=v`VhYSfxL{z*uSg}Vt-{d9&c~04*bcIrHRHGY zAJOE}eevg&pGN9(X@hrDT;tTo6APjY-UX9QF%Os`pB}+^C76U{=xtjY^Empv(i8@Q zc(EV*uUFojiQ=D?GtLCBfR1Qvb;DyITvZJ!d8QyE^aCA`SBD^Wq zssDN99ZjnQ4LmxGMzQ*;|ssqUQ#*{U{@rAmT+=~Hx`H|py>)z}-6g|Ra@ z&e}z2@&ezO^u`YZe{2QcR<9XT$Yp*rmm4p!pV-V?wwt_lMu3mjY<|Jy4&vnyyjr4DLY`ZzcP} zYt>b=Kq93}o{zms(6jXLih70QK2L<3X51Q!dX`R;t&^p!3Z{ zZ<~4>sLLMG$pNi_x-1sd;)F>F8^{XbDc~RbRA}aU)kk+IuL_b9auP}-k^)ih4HQi~ zR7h3=`jRn6L8lTM6b8slo<NOI_zOUnIZGkpe@Pg&|F z(PrPn;kBAZx(L{sa|F8uDZ<*(m3QPb*e{>JoOcB;YedCGto!Y&?j3Cg&UdVQ$kSY( zK4(Dra}-xpPsh$7NIboI^fq`wV(j1lVq@JD>jdTDquEXJUz-W)fXxK8i(;XUW;>PG zDnqWZl$+M zesDv$GSVb#mE|zI!gZUdDph*qI?r`}c&JGHL3#WMDbjTCr*`20h(jpjmg6nj zGymB+d+`rXfBCdFQ@M>U2tf_7$I>Srcnzw9J(fOu!m};s+bAbtgYVG4^m=X!bfo&| zR{Qeq*CgY_QKjA<)degf*ifu&ZnE><7wmyx8CA+WnZ=U*ENN zOLA@^N7KXgEsXxSlvyrJ1J5^C6>(o{LOYyC0jZE0x+Fc0-?o@fM%#;`ye?sHwqe6 zeLvUCG0$f|q49@naM6jEMA(>^!>^x4<_|o^n*C+aLZ(%$zOZ#j-&fHDwZv zY~=btm;~FPx>QJ1=yFpye2}=-`tbtE*eTy!^y~Mm=4EzGLNsY}WAoB)V_t4k?B^7@ zK_zaH?h8lzVByIJVtC-J1hTmG{$;WH=mmi((MVxt=H_W(FrCSuEnKMKw~9G`?f%Fp$th@O}6QuK#6( zq|u(V&}}3695jR-%Iy=HMIh448UsLzH|TMcnQz{_j%#7^fbAz$xF@2@Yj^a{;L$Q5 zyvBPJ$DEhp7~|D>|NCJB^U2#QDoEaI6S>`QGyCkMSm-O*W+{qUF6>~Qaf9jllm*dQ zq+6O9eM+)PkTj!K-6_fkBa%xSS?sdLBqOStgQdZ`nGU9bzVSwt9}pJS%Qnic_?y`q zly_pQBa12L!jCt|xc!e2n2_gh?F>oS{e~4bp}(qnkF0rZU{h`bo6Qsp-H#hB+KU*L zUGHBzK1=;c%%iC~Y_F~zpX$}`fwC1A9?IOXF8Znfg+3d}))*cBc%nE+YU4^67f(nd zgUa(m?5rbIOGVNDWQ9uVZz}JSz%AKs5!#7}MBHWs9m=Ud&WWPP5Bn z&yxJcS)EUt>kfK>#uPR*&n{j9)T9N#=m3>^s5OwML8||L58S>Z zLBdtk8Nfmqz3?HO1bW2f0m~!T%FK$523X3&pP1YS3bLI5O#H7|z61Co^wHOtC4R>w z)skLyPi&FT_32$SsGK*;F`V4z`|>FA$neE&mtJ<}WjOpd|0zGTkNdi_MNiokkAA#mZ`l znJ}zFxmC~=QxIK1*N{=P9u5z;)zUk4OkeOYem3Loq`#a|SlyJ|iNE{}$>iszxNq+_ zDYJ1?@+lUKNG_FlIrO;bi30hQOM!>IIq({AEiCt05P{B0F8C3}v3-Ij$u8-RDK+EE zY87LxI9ZSqQ4ilVen4|uel!q` z8QhSxlwy}qWFeJU7Yua$@;c7$Wq^Sa5dwh_-p*b!_zwGgA|cw~Nf#B-w6 zT6}ZMkAF_q4b=$k#>yXiZEQj@#X^L(fJ#Kc)g9sU#QTF_o!TzCBI)&R5#*>EN!?_# z7%MJMb5vKST25gqaIGu}S-o6+maKB)v@aAc4QgyIpx8|m(Nl@3Kz=hZDS zhXe-&E;J{@2N1&O>3BPRgBw6se)8AIUt7Vl=Lf!Wa$_j4xNm(YS!mPf`jBFKDRLKR zTHZ7T8KBhyL*{2Bor*@NFo9$YDg&i6b3qsLuGG?ifqJOB!h2DA>We7F_X)ZLjc1uwf zq)pitu}TQ;5PMFL#-{t6mGyfp6LvwxlZn3IbzF5y0wtKOAUV>Zyfxz%sD5*4b2Gz)EZ8T_{qr7 zXJ^rzxXX>cm17!Y|F_Ec5p*~Fuw@zw`7HrkODX36!iw5ayg1z)z2t%6vi;dXR-2?r zTUbNZ4`q|wcmmsJGl6ZTSkS04P>Gq*H>BOd&GgcVJ<=P}20Ay^s4b(rg@;HPy_sGU zxRdN81IJg83{3-lMftJnhSaF7ipe zlN^drVy?(9M79W^?x_bP?J}b?mCH$&q{%Z!b&GEI)uG~90}X0t(t&%@eJm9T-J-3* zIjU2U3t&Q8;ddiSw>+>arcm4}>+?qaXP|vSjl3c;%r(Aq+C>Js0wx{v%&E9Z7l86B z=x5^LMpDlhRmCyOU@~p@Z5F^O{6JkaR`TMKj#Zw%QMW1v_0HbF(4xdtv@Q^K^e?R25k03n-E3)46bg#kO@y5S%EWNs)7u}B~L zeptq1dp2f183Q-=emJ>ZVAe_Gnyrd}cHr>rhsT z3PbVA!cg3=8-;RS{+KdjYw+;9WIOb$t9~=h>e;;2R=9)|@ykfN@w{``X5QILvAZd< zlS+bBj$yKUj5xOSH%E(q(r(&6BzehckTyFek4lqMCDRwZrg6QMKFf~iFOn7#7{TH1y zhYb@v1J-GX-_F6((KTzcrdz@DewDV46!Qa*8=IJ;HsGnG*b0i2Q;EBQ<>#+gltyiu zY_DQL*k<~Ya<8I2x<_zYUQ0KL3f0&At0pCh_efh}jM{clZpdMBPS7N1i790MbjAOF zTC_>jA8Kbm@u-{)zfNln^ln8sFI_;C0}zWJRKG!l~8hvu{&kW%yeOyM7JdzeHnQHV7%7cf&hecGc_r2*wVkGc#ZBS-JBkvp=G2~Ntc`{>~f6#l!z8c1Dal%pG zvNO0MzEG+|eoO;xnK?%O#UAqAzWAUchrwZ@WYCUd{9dAM39c zZMEb+xvD)hyX4)IUtat+()+;qqEPtJM;C=Qv#@bX(G!d9Mb&hJk4v%U2&RVcL7X#* zya0|HzWjN;edL7C3doIpTaK7YUGU=o^kX6NZ?4}h@!uKz#G^-ny&5_U`MGwwu^$Uw zz_7pBh5PI3|7)tR)o=aL1@azQ&d+akV_<+j&7ff028xAdkZdaP3MdfY1rAbN(zK{d zF==e8(xmA3FaaAbwj*AY9#s@mt+E>drzFQ!dQVGK*QHj&I=;LxGGIkqh>d|q$2I5q z&$MDBSNUH>WY;+IvCYe^r&#Fztf3N1RVLAHK~-cbN%hQxs^u9BT(+C|->9%)tN3Q`Ql`ZF7xidNYXQFk2dHp>f{USiQOsupxQzuL}O5FQpzUvLvI2#*Q=%RgSVkLUAoSKPS3 z7qZiXW~Fr$n}uK4L`?kTV)_qxcuX|H40=K$be%vP)@d%(ghT+8&&24`g^0b;n2&~) zHGBaEoObG+o!|Yjy&-u`;K5~ffUl{~_i{8cjMYuP;iua^6FrwYrbAg4xkY78Ut4%_ z4L}RkNEU{+)5xE&ObBt9s-P6{4IpZFA!bHqcAN%`=N>dA^~Uq$dLo2HXY>-W9;Ue~>H}o|RyEL&~^j zRV_)1I0)?kspAfjlyS#Yb2U4q14oS7HL}N{_zXO9iar^zTzH;YJq6;pBxz!^s#$e) z<`a*0(k!?%1m^x~(t6p{rOLwuaeEm-sg+8o~}M4l6LTJi4pmfmiI%C=Xmb ziof}JEn$LHv^^#`Uj1PyRvww_Ns$Hz^Zv+d;zA3}_0}AE^)9TI=KNfICUp0WLL`0t zEQUWTKJDa=QY{Y_UDDn8J#0te$Qc zL5Uh=X-fOpf13m=w~>Lx%zhbAG_jak;?+g3Q{@Zhk!lfAK3!wH1zdbH^ zzxX(}-TASyvTY>ce)s*R0QvIZmC+`O)l<+SooFS2(%@=)SMWB8&Qg^R-HnnQ71rkW zLJdQcpjT$LrWql<1apg6_U(FmI&A#Bc=$(rcF+0urrFblmD{}Dr#xGE8cSy}a~c{r zkek@7@RSi>9f@DOKEXE0(vZ5~YIxThy!TGl)l9k&X^ENEL8a*d&BwB03jcH6%r)$H zZ26*&_+4^6fcrZ-R0*2fvJkxc_S9yHb%I)$efg&(Z5-KR1C{j@yOtuWQG*4N_%|Y! z2BiDk2|{_g!cd*18>UE#f?#(?Azz)vP00w^9*`Mmh;N6< zv|WT<9q5%{CM@6ok|39X3!6b}bGty#B&!>U9x7mv-vxO+Fsj?{fgW_T96gvKpIeYZ zyY1TrI*3S&iaNp-q1dN_#Z)%}E=Y}nLe&#B+#~Oa0)ITVLy1i$ZBTTX>NTgG?upu< z>GsEqVHZ0o@#8FyZGzsAY8n-?xFJ3qo665Ge&N-eHo0KV_FDUHYmNrywale2Uy@#u zrU*??f4EAJs?PAPitM9rkv-D=qV&i^X}NL8 z=CaMCb&6t7Qsfwwc#BliSo%`rgU)(!bbV}&CXIdcqt4ldnsuHJqi&HT@oj0R{2nRu z?-6W`!7TOl>G_%_SO@kfZ_!A6+o7zLr3g{4zB;nWvoE2MEJ-*E-Su6d(VGs^_AStB z527ZTW=S{uY)us9reTaGvB1 zzNDL-IrfVwJ|@SFLB+8qTs~oEaGB&hlcPxz&$&RZF__3kN2Fe1De|7vD%<-e<{;Ci zK;uFO<4#m-#WBSr=U>63HTZodN_^9(-R{`(}l04$zP9_c+IzsT-8bm?nBhTyns@>2r#F3L%kzl~p;a zr6CW13^6&NEOtkn!Myl(1@ia+?5U}+-{1|Om*3D>bnYv0R>-V2DD(@4yH?D)asc^&*eT%it>mC`p)D!rbMwF1dZAy^$gyZ%4?V*-aS&H9Uft0#1J)U&AF_0G9 z5W+mB*glHfAIN6k5H9x0fr|fX-+QsAAWEi7_vxmQHZ=+K5D$YtYl5Z$Jutzv%DUrB zKqJ=gkr{2Gm-sb__QvfCPY#6ob>Uj(44EIXnlXtApr|9|jhab)f*cn3VjyA&x8_2Z z!S$FXVW$EKTWb|Xz-$Iy&pq-)q3dAwtCPg{V(&2-AxAY?p0^>xmL1&|c|g#?=sKe= zN^kkEQ=I_8HN(_`=-qPbp&B(F%$nS(8T>KQeNa6zN7W>|>ZL1W4v|&9Mz#hz{L_8< zJy5$bmwp8)%!Y_QkG;7f`ilm!hI3_7(PSMNb4nkWm#+h?eDS@V-T>3l%MW`5>#!r@` zIjl0!^L-K8Mj&%UU#oauf}$4ZBTSB(Jx| zfc3&X}O)Q|%Am2jmY`Z{*XuWX-jRE_x2;IOdTu$?=Gz(|6MM zUB?bO(y8@D)8k0atz|y@*DqU*)$9sUC#f4pF4;IY%@lirB8|YW@Js~HoD_mmD-+#I zFwfMkLe_UI>W1aAk;T`KlM{5>TSyKM5{%kGBTAm8IudKF`n-EYI{b=a1jYpHn3zUt zmM05K!nP|;NlbJ;%xt;9KBdcpaosII0W}<{tAZZUqM5oRaHdg8mV^~A&u65A zF{F#bagfj8@bYuqH#8St@qVN|uczCVB5$w(`$~#kMv*iq@?y~WStk4_8h@IDiV$r! z$@I}qi0AOZ1YGMVTzNir#Fz7e>)22Hgnwt>BjdKH;OKg>R1xK3tdq|5>Do(Hfd9VTW>oycyEzMdB8|9S1zd$x);@!B+A3R6Hfl$BwQ1 z{Sy0<(&t$f-8aw6RXBR`OP8c`fx_IVH7WLnoC#m+wE-4I2Lm1p5Bl{8wm?Njy3fI& ztKN;?PdswN(|j8z0BQ3(itQ z{!*qxm#NOLL+Sc@;vvV?5v&Y(ql*s~M}6_uNa`!x*Q4U9uRtcNRvGx!N#YdoH95H7 zWs=<*q!j-oaK{@ubfD2OM*^%-!-^y5u$_Ng3=8YyfBUDF1NOus9E=V(*au%YFK)UHX=}&Gb+aGPg|< zrhK&GC-&*^=RFTMwi6slZv?MCI&bn}MJ^<`>!9irIlH&Nl`Je0XO73!JhBs6s%zH! zw<%9TSY)7}^e=P8c~TR>i}E$={Xd*5?onh%chR_F-yfeUE)v&G&Kz%4^~z2V%v$kU ziN64pBY^r-%-d|d;BWCc#SQh?0;P#OOq-O2mxpf#L4Z1DbVtDP0T|tuTz&)Kwi8b4L9cc86OP-O435cVyWoOZ z6u?AVrERdj(vqjS7@Dn0N+^#zMSmU!B1v=(GzC<|9{{1q8-6DOF<*&So|{z*Ed>{) z&GKVRThukE+N_SufVwPfKfuO>vRKQ5o#?{yfVIsW!nNZhv%*t;;>a(5*kd1P<%nmR zdn$1NP|zwzb%oAOz*lbw>tILMpun4Q^yNG2r8@c?UN*&jb7Z~W_}Lr5)|Ad;ws9fJ zapU@jN}KTeHj0I^-pweMWTG3{!*67Sm=xy(3xuDA9}x6=1fd5^C4zy+N zeUsU$(g-s}S$$wVcqotF_(%%}fb%Zus3s?*YSz}6Oy#2Sxll{eAO3NiE9cH(p?bj5 zxCGCZhjOF3_!mV#_^Ng0`St0w-ysLaL38ThcmFZPHc{j_m3UaC7oG=mj&(s5K!cqs zPD^NqAOEwU6z)?J}Ap>45Rp78R%rhqOc% z3w2oLdyruB+AeZ^nyv^$i(wX80z6bb5lyNlMYUwD*d#mbVv!Q_NDdPZu5oP#VT&6N zL)f>m}P`6_xq}lf5uYeJ21jnau$$eg;i$=uT-Z-3vVg zJEa?ATfoemWsXZ8L%+!aVUxUp9yn`Y?vn-4u0qa;%nicFnQ*ax7B_qXRh_>XWA%GA zn|A+^Joa!~ZweGA20P?w6uXEbDOBQtuU?XVctyEw_FYA9)Xq`xGVEWkgJVDBZ7x5^ zJ@!LPHB_}C`_9*_csTUq^PiD}ZXAJYx4}aT#Wqvq1eLgftPrAlKo)D(52y#VPS~(0 zs2L_Ugz0t$ulH}D&w$suefAY4R)Oyf*5O?!Ds6jWy+0m@mZN4lUSAB?moY^?*fQA# z3=sVuyJZK1sw0;KtQ-%D)-h*2b+}{EIVp%K)Etx+#*{Ms9w}Z6gma_wqE7}~gpzR{ zMrgz{($lHQkqyVu$O~3ue)n%DPuVxsa&)6&bQE8%ncT@r|6#xZv}19RV&{Y_?%9gdfdy`2|B zp?6c97TxcG|6!k~QB~r<*sIyf>9i%f!#q)4knDw_C|8(7CZJ_mHz92r(r*{gNnWWy z#0|uPbf0gvphcChDGXg6SQUe=2ew-{uL=!cj#FQ{Gv2oK8N9yU?pw4tU6l57fz|GK zgn7&($!nH0N%B1@^nZvPtKhzEZQimlZ%X61ANP0LfE`DSzt_7tGD+O2(4kKK0l@}P1l$+eBq@#8VO@Uz3}}-F zBE(kNS~+M>Lpct3-NmzS&UU44N74uZU8H=R7LTAdwnb!L;~qj9x#M?&Sqw1Z2lH9@HIk!RD@z?6TJgcd=O|PL-!mH9+)&w{QuD6J3X8YStUf>%xv`bmYy*L{ z6uX)t8C2r^u)L|@8H0jatL$Rj0=YT(b1`n2XA2Ok6#3LK^@^=*Gc<+_+XV0)BS(Q? zH4MC;KinI*fiOg@cTCTPf6VxX6&%-R)ZHXU+!!34HsCl*v7b;NoSbL~>W-_3*9&#+ zA=e^uG`*xK1i0cETuT{fqz$U%EX=GM&eubt&-n22xN&%eqgicP{08q1MTPgk?-uRJfb8h2 z+5z5Y6mI&%75^o|ir@}KlK8Q(JgzCeOxp}@Yc8}t?3A|5G-?~5`XW_*M^-9$JW?8G zj}9u-3}oFO0CU^;d;xAhn*#FE%~ z>8O2w;%j12INB`dv1k@Q67(y z6OCkPNRqfRzTC^G-7dK8eaIWNIrB7i%t>OBwJMiQ86_aWxbid~4kBUXm;K;j@b6!B z^HCmvGQkk;+(MVo5H~eUwP-t_tjs#UC+Z7cx&pdd^vJK8Ucg-NN{LQ!Z4ZZaRCo@# z#qk{Zh`#g*yv)a#+7JKXkM=8Fw~c;qtZJW<^I)Ynur~NWtB-6C4RQ8f*~cN7!;n0D z=}UgW1w3%^f~PqD6faziO(j>p<7HjA{8{(O?@8)V z4ofXh(TjcO%t;l`6`z&#dpwhkfdc)PWu=^qu9w2^*#$2~8838=^__5kvpr|rYZ6T1 zpuT&rP1x!VW(O5ck&pbg#<&1Yj^gE+tr%*{`FPpCK=-Qz>l_t2fAJQw zhhJrq8%I>2ByNyXe3)YEDY735D8Eo>h6QCSn14#<59A+T{O`6Y}6hmg_=8p9dY}G9m={m>})O&Tqo_4T~a+$ z%#;8*!!n-~zvaS{nl5<(s>dk0Ne5HQ93NSj*Evuc3SJ{iJ}x};d4WB|3=QMWTRgalZLESZH}oz86s2sW=%>ug0!Q z+(O#m^hiA@5_i#P)E+2`e4suUo_;C#j$mtGv*#Mt!tiXO@o)DCHcYBl;|H%)WrSRo z7SjcucrV^zc{V=9@(;s*R4)CHw8}QecWRIWs~OySlgw>ZH-FIk+bsL8G>+!4yjf}N z{b0+_7HzfUKDny>?P={na-VD(^fJM7}6DA;X2cUfwn@xp$OHL@E~ z*p=L~h>Fq~V5GAz3?v`6XGk>fQy?n`44asT2ktplKysMo*c-3WAk4c*3k3+j@{nt>8C_5xo-N+mYuAp{K zI~3dkWtATZABQ5Xq&M(h|WoD3v*_tqiM~5B#~hEHI#%S5}XokWbe$VQCDiZNiDE0r!(f90%jCk3z#7W+EEM*ajQ!Y=_i*3 zj-*z~jaS(mwNjg-dz8T2f|WB!JE&_@_L5q4N_4Yd8k-SPuihfnU6h`Xmc<%CPIHuS z;wUyZ?+#?N;bVVE=Zk?3sOe0;FD~n5-Wml24E3p)uhdy zXXxVCc2Py}CSl&Js!8cSEBsnw+LRliBzM)+B@hoj3F3LSXN}rIrgHWl{&YvMC1#WG zHj7Wfdo9m$HkkOM<{5r$ww{i+b2iq0f92N|R@CH9{N-;*=4*qRG8@$7Q!G$$b3K=MRYUy0 zfbMp=$l|bZ;v6_|KBaoxnTS`NXd_wJ?1#k$It#JWVqaYathUPoP}HP9yebIFV4x^L zq>oPZa`xAs>%H&j+jwN-ZG29#Hz;zAN?gnIK+zmj zWNOzk>C;T2lcYd!ScDwO_mn*eheeMe{dX#IuJDP+QJ_+N0&(E{u->S)sLk|s#U_Zb z+*fX%*`qMfiS8Zf^1e>8r;Z6dFUF7q2zF#H6ksZR~DU6!v zkW;?htov(@q*Ord)Z?hTQ92CR=4r|U*2<2Eu?mx8M2CK#!#z42d<^$*@wxBYW0iaV z#cGhgbDa7r`PhvMdT-m9m@5=}nIadc#0&w}>-K%+gAZC|=-KKJ0>GwZRQ1Xb3Dp99 zv>rrTH?pRnOl37)=-(x8qRq+y_sOT;Ncy}zr~?`tJ5)Loje@O~JFz~&$o6|+@$M~8 zEE~Sgw!(jSMXt{i#U4uz>y%`Zpvr5zpa;qkuJ~8^;SCRIqGn@!@o#h1_YMa)&a^6ANlISx@tE}A{k+~jaObvJXO2)D zR-71Qxxt-^jWXoJtrlQ4c#Gdz2`D>8T_aDRb)b&P4Xz9t6@y~_lJi6dAHy)BYk8rA zGGs9O)m9r++Ep@-7~I$fLB-{u63pEc3uSjDRAMbu7j1>0eT{mL6lJKaFzhJop%btlM9Rj)89P*)y*Lb_0XFxnz5vu(m$&rQB1 zkV)SnF7-b*3yK;*WslQL@E$Fm$a$$r;r#(PZs3KSv7e3k*4N1Zay<0MNy<2qZ!*5%R#kU*$HJLI1c#gp|sJu@G1}JjKbow6!DE=$g6VvD;zk?YXBU0p0h6DHvn!+ zzt{KSm_zpck8V5TaP&GZo3dz1jtYq*b)aPec*i#HXPbcTbOXU~P4k9_NnUfZoYH_tR+Wq$C@o{YAoiLO{9NTmM1)*DW*DOc28UcebLQtt^Z=HED2nak3x#P=3ro?L%B_{g|1+8eT>kn7=wCa zI9KfO>eK*PIs{IEJqoCH!W96c&` z6+OzM^5wD1r$5^o=v6tbZJ@WQmQODVwOr8v9O(yzb-|+&dl?R(hJg>>x4aYb^BDUg zY`0zXaFkU)WVVPwP$;28d5$>-od6F7`_(0iU7;!Cvfeo9lRD0*t)*9a<8jN;-OP%J zJns?613CVdhcoO>?G@nosJ!e5wI{ys_M6susbcCMQ^*!Ko|lf;%uD+yb`M2%QHjXi zetmlBOapz7bkj#QCm`l_O{({*XL=_XwZ|mbRjo4 z0cPE53^3-xYTKef;Dl}_CnM-7wD@Dj_6}YM8v9>g%aHxTilF~ln))WWHjaE`(_`FA zv3DubZDHy;Lhg8D2N}xPn$x%Kq6Yy7LCF2=I~xRD@_vs!u`Q7p;OL3POA1&VPxFP> zpklF5^F7%Stb7T)Rzo$#oZJriyX(lLiB~ zAt=u$mBPGT7LXR^3pzD(n&p`GLBI8s1Xj*?bx&-scP^a=QMp~LZe?UXbh{zrVbY9b zbv_WYLpqUd79kS)J?_0}3c}vRKY#Go7Wtn)Ncm}wDplMpy5)(Uwa#rDi}{VP+XjW3 z-?;J9S6y)&zYDqa$No96G}{@hyC7*7wLsjafUW@{=VA&i5PlC(f~YLE8%leVB2J5- zEjAj@wn>Y`u=;|IS?A4bwzdHFMrNj-dz{k_@B;A|(*btkNIH<*xYvZE18LcpHh$aS z)1ffN;a1z%+h(r}`zUhRm%2wm^3{Hj9mJ6RZ7$#cl~Oe4m)6Padi>HUH~_sN0A;X5!>g_lT0NBX|}~SC>Da=OeaVupcTF875qpb?sEhq~tGxKKYu&}3Hw8ZcAm{!LGzxfCD4J-CSe^vD!S;Mb( z)Qv4txs9W~nPMS{zY!hv994# zBU>m1(jtT(2Cu6{w_GY`;h>p%>S^0;xXfI&ME&zmzhs5V*2&CmQtQTM>Y~lyH&HAs z+m2F+4;1rcbH%2his07Bls76M5dk!tTVwWVvOokBJ63Yx(9X#_CqI!tk%OAad4dXV zy6W(aY)(j8berU=GL_5)cLz$fM0dQ?8H1M&-R}~n6^zQcS!Vu;tIGYtd1AoK)|d-k zt;$R#UVbIYs6r}Gr0PT@Wh!?m7RdA9cU8zz-&DRa=n;a%4l^&><%cJJ{?Ly;JLvUg zE2L&uh&oB#P=wNM9Av&^1G8p|JwcI1Dly$BYtm7Ao@aARS_taKUKB5qHHlVDYM@K~ z+d~ZW2LGiYom#95Up1*qQX9NS+Cw(@G%^)bBVz~$`3qo@?e;!F@V@!t?T|pzdzLb% zwfbly+_j7J0H{#NyiK|)6z49CU~ZVyF3KY(Nt3*bP8nwsRZT+HzH`2}K_Y!=NL}#o zqM4(DGky>=U(Jm(dDDzs``$~pjaPGYQ*M^^#NL7bS4b-;z4#32P#g-*rMJ;X6 z@~-sT8)DR6mbQxq_O2>V$>v9V3VkfD#>K!`@wOQQH@f&YKYVN%Z!FbX=O$D1#IwJGCAO) z6zjhFJ@CKHFgvBO`e@xvQW3u=0c97uz#rH93^TON8+$LeTF+fE9GI#Xoc!HbE1DLh zMAVahk0Feb8?UC;+hArX#V(=9LMpK!8uJwSnvea{qAe@*(XukOqu{`}+ukrrd)zi> zLekx3_98nRS~hvHpZhNmW=~iSIsvKL&y=4jSIdw{6I#w<54$qzI^;`o;8l;xbZ`IR zxL@@U-!%7KQ!M@IxyAPM(632J4hNZZazG2T(ya=-6b%gH;x=;((NcH=bs3^ke`vsy z0xP6!!{1$sP7b)G=t*dfJRXDB=cw*Yubk*&i{mifoib6_qndj^RsR#Q)HZWSD7OQtP_5>c zf0giFY?IvJg-zClex=N#gs#AJCRx2F;;;(S;^ww`!D0H8pHW*CH_vyIU_S6=W=1Eg z@mQaCHT^_!%iOPxcR1f{4Bt}Nm=34Gc;G?&;M}iTLD?7lJ>Qt%z7r@Hw`v0{mWyX! zQI-Ol&j07?Df5ISin8(B1+`dCp}94&-M3WPBj}CZ8;d1zWR{O2o`tt89M|v;<`;vBm}4PIYy& zAn(^MGSPRXMNs5Y2W_j)|3Y%S8NRQZ<1tP@%>BBD@L=6{t@uzrImn)omyZYQzW*{P z6CLbd87LMAm^M&}OW!uoMZo)^TM?ZWS46r%O%g+~`@?S$ee}ry(5EdDUxBDCl=$^R zQqeFK6W&I4nP)4MkfA=Uiz(�^Vjia$(Qq6aCRj%4$U(`#wHFo_M?_-vJ==9*mT9 zid{^RR4Nf`Lb2MaJOCs1qr?;QyrFtIAfCTxG>+6>o@5eR0r7s>_&<^*LnQ>=*gzE7 z4Es8Y%|b^h5zAnp{v;|pp%%zfu_zca#aO12qbi)-8d@vtk#;E0%C=27H*1)I&tZ(t z%Ryjk2V7_fY(71=33z|SUsl`CEN;7?;g~B>lL70OOYy|e}Qz$W&dg#)$*XLKD;~}6EmNQut5cLe8nKS z)&a)k9-Q5&*c?{nwQ_t3oe{qzz!Eom9B?ldwE-6T-Gh>~KEdZuZ?4W8IDD0VqT(osVk`#Ca|-J-3*Sk;4-VkHwlpO8xNuI(A|Khoj2+Q|PBzbEi| z`|g`SjlBPPhkY{7ZKolQ#9h9)0W|2a09J?jTXacA{;lVS2J_X97>Z#$cEq>vgX({7 ztXOVO=Ha%8;2`cui7p^jUWTv^wk-C5pdFfvOfgPpV{vF)9KpnBIirI=FfUAuxwGDM z*gnDQwwU0^JVWumasycbM7mqTx5Z?!w>&FV&>&>#Y&_)hO0&T)AGRIBh1p@SyV1qF z{BZGh+r&HHwK^3)F8N6xX&y&z+Bg+$6btpstyo9e4HQb9Q0JGng6$_xx% zhtY4Fhoeu~qqyRKB4yjA>B2=*Em4%!0raxT2ovpJC?c zZ5LHbPP53rT;zi*8w{)F`s7QY4>x##o6+L*IC~B4L_EI`pL|d9!?XW4?glz2F_1K) z&-XGEU=#u?LZ3HwVj=x?Mo1Tgv9MA3#b1BE|BRPj@A&(F>7$%_0l)X}woT4EJf~fn zV6`W$Z=QLdWbiA7b>nO?=*|w}-p!*}=w!_SiGgo-e&g)FFZyoz>}Ok8bC;!V65R?s z38k-%WaW6HcC&ABY;RP%Xj}NYaJ}>aYX~nDxE%YAquqACbSkumc-Yq#XH?lQj9-(? zAICDf%Bxw9h7jE-Fol|~(aU1nynhD}VRQXx?R?Z>I6jh&cQ-a193Ahg0=D`!tD04v znibU9zi*qpBYac}Zr7WI=LXdQRG@))?mkx^IQuP=cA|AU`R0}%|D3FIWA}8g z%>+?QvB2eDU{NY>mElUq!uF!u7Y>@4F`cmJ0{Yv0&3b>_xfgdLBG+e;c5URNgyaBF zEJtzkRvGdX_IoV%tq#dl9`W27<7{_yU;rHifbA?{yEpJCr~c;U7L9qgB>a~_R=8-= z7S@pUuZ{Py&jwsuDHamQ1}bp{xo#HWe18Q=Wq$XI9_14Du&M>vD>IdNBu6!Hfl=F` zEEXV_GLo|lKE5e>KWT~_xXvXmi4)M}65CUh+PBJ{r=5=*;=b#8Fg%03)io3gy_hSg z#Dj0=sBqPVdJj-sm!VnWm7!@tJ%XE(#mXX|uBfxjsJbTnpy5>8^1_Ert^CEWUv*-# z*oX7KH+v*BIc_|ol-tZGPib;i1@2DR30123{@LP{kx1x%OjRw&2stmwi$Q=CiT7!? z#w_;TPwIr7bm{ln1k2w-7L-<5H>sI~CEkytTxx#!K*iRhb*=eHHnvvyS)J8)>einj zEBQIAZfry#Emgk&DGGh8OqpUyB!4=6 zX-JO%=`WG~=9+vjgpP5c6Lh2I0`VHkG!*${069{Z6BX$kA4)BG(EgJ>>lzc*Qa$bPn0JEPDP?VCLTrlqjaVlisy4wn*_zN)wAa( z;4z~X`}(j1!g9qG|2s48#8yM=+2)u9@QfuuwtUF%s`ifea^Vx@6Ax@7$KPd&WlH>w zEbYna`f%iaf1=C{8G7oN6`Ok&+i;LRrr17; z+=rAMP=&7d-={$@18bc3sg6!=l2_2FKw=C5*}~9MaVb`LvUy=9(2O~#Kpjk^y~$Cb zPYd-ct+FoWs?rkC%~5sBZwL)gcJRm#o8@;fCWx@5vAP~*A@s|ahE>N|?lVj^$(EVB zNzF`wSpHk1HnQ7*IA~HfM1OvP5c&4ZamX{ZsROprq~MiOal zpH+fXwvoZ#E!xvsqxKBh6}r{$47oI|MceqJyR%JiHO|K4aGlyD>Z4bX;@Gahok0Fp z>c7gnBK`nk{lQ}<`onkHBWGD5*7Cgt`J{p$#N0THbH)Z@M=16XMGjDj$VI!=Ynf2jNVdjcNVwml zRd#EJu0g%hPj^7m?Y{}89*Fy45h*sK@P|HnfBZHH5?^nU z7LjD((vSrVuI0-E`n*rlx(funCSfn2@PZ(huAt|N%^d}~J{EnAevfmL4D|X5nUhk7 zMWrw*SaJkBcGonW`)7^)V(odF3^zs$$C9i?bxLvy)JDu?Dylx;lL3eghraf9Z)l{u z4vqXs|Hv@z=bqMLzjEm_`}SqG^=vp=l{=IVX@e{`cx7Zdw54_^`@Eq8H6}Os#v2=f zbZ(3I%D2~iec=44p~4|344N5+F{;D5g&PV(csNwYmdOeBSp>I*2S;AO!k2T(d!ASi z1|Fh&*yxD$*5v`PTdyvraV3>c-&Ps}7AkK_?#1d*5vj!g6fC|QSz|!)?Db!CRk;g} zAz!ErLJe*X3kHtd#0x58b|$Z>vM;!OrDhk7vRb4;=$2+iXF%Z-W*E1GLyZFXBUnIQ z210qQvQGlbl)B{;b_Sy^19&Fu;|iiJl(RQUl@mIXUoci9fzAsaJK|H^(7F1q--#Jk zhxSJw&)iB%_&Kz0ykdzwl1;MpOR^(YH33paS(ms_q(inuaizq%vXBwH))jm>y#Dcs#H z?-A@tNd5oV`x3aO(sS=Q;t9!%AsfNu2*?sa5Jwinh}t;LozB*2Z>P6kd%Jvhx=4F# z*SWpv+;+O)f;$lv+|U5ZBFHMDqO!SxIw~M4iW|tHGB^k>3@UujL&8WRM{`KR&DgK= zo5@+;^M;=HpXYhrXZb&qaRMBRlPrQe)~h;e&}Am1hMroDu5YaU^&8ft=sIw~k|jmA zM!3xn9D>z!5^R;bmAPWnVMiq?;BT{jnK`b9rKsRIt<}TdtZ4exSKU^@&|{T!v0Jrx zT}X+`QlQD7Zy@IC;2n{kbuaRqj|;!Z?X+tejbjpB=415w{m!=`oup^6fNH3T-9z~_Vy-dKF(_E){O)zBF4gaSK5BR^)Ux{+k^2D(RK7ror^aRQ7uJ)-vo8&fEB zl2s8&lB>{@I*DHGSM7`u&C^~f{642crCIv_25^uhj!A<9;Lz#kW4_P%wMj&;IV1S1Mmwk zk33K6G#hwbf@Qptkb}Z5L9}R7;Ibefz?_D`iu>vZq(W7}TgLm`bBlAXB0J)-vO|7a z)*;UEeMIv3aqg+(4|?wbuhK#9N?xk;%9J6B)^?y9iBvLw<|#lqfsV>a_~mJ9DiIvm zwq~hB7|bRf=vKZiq$w0rix-N3{x^1f^sE%$>tXxpdl5qx`su5+&k;8u+2?NsMVlqU18^@*xk7I8$`#e}W#k^o6)$tienA^XZdfuGB37Lb2ca*ou?wXq z;yGXYsz)Sh+`$?G-rN=$%91Qmmgug_fTy{MZjd2|vKCX(kpN^AYhJiR;mg&s>j66y zhS|D}88dljgZVC>OLn0Hizk-&MTfjsSvS6ez5^8v6+93<$is}iowxcz5d0*m_J;B) za@&x{o(vsnwtq`|MZ7;UV+!!jsP^)A1GhKism`^V2ICGaQ@LUITD~FA4vV^CSXzVSMs=hT1deC>!*i*Mz zd6!-YJ)gD3${Oe6GqNN*r3V6$?jN|0Xl?X3Eh?piaatsPO&DiSba02n5XoRKJixr{ zkQhegyi&4Xz2)0RIkN2c2k(#4J@^V&||9@(to=n z3Ou^VL3O_qb|*qta&p+bqY=%@H1~s&BpTWgIi*4O#72@GwpX%|Nus-j^S}pW$Y9>$ z{Apw{*0&w%8t1$ytKRK72Yq2NwdQQd7CL_ z14ZJgs0`Tbl==3$4XiM_4{vxBC{L)nY=ll}>g$uoJM>WVw$}475lqrJ~xxw+0@bxXmwXa?beW?q@u8lFhus zbn|$K#Lq2ASl_L@6;=afRgdTNkS}EEP>~C*&rppZ!)7nE1=t>0 z2!v+fFKK~PoGT%BBeYw_$4+gtdwt%>L1G;UUi36KkZ4}b`J;vXIkCxM`GEI585$Or zxSZq_yRCws@8zJn@HnQN?wz&OBZK!~T0BXSET7fQZxhq?jWVsi2~ogb$z+{ma3FCx>j7qD}gO*de|v~zW&X;n{Qk9V;eyCbh{hG;6m<$ zJjO=t31PcB>Jih<@P$y5~ff5wjCPayZDf>*9SN!M#SJvPcN z2QAT@opeK#$@@gRdrtM7<^Emro#D~JiGF(i5>08)W1tiGi#F+vYjDSo0>?T8KL>(V z-r|PflHK`vUv*tm5_ycyo7om#>~j%Zw}b!Kap_rC$RbRzK5f4noG>wF$46&=|E^JZ z1jS7K9y#W~!lT(lc$}x0vlOYOqLQJo{;bC-RkW}OOo zotG;URRgIBm;r<7ctFc)pkzHIEDucJCn)a)vPf*elL zOtejI`+~m+KPPSp-_O%tntgfVE_Dm=L!VOJ@HiUO7^3YV_?KV@rYs3oZ09O-y}Baz zLLh*HgV!QpJx5(~y=5sBxdChLmP4qU zVo?Bk^GkWXuDE-}76VE2Hs{I!?L$&1P9bX|sv%<>nv0`rD4Ca_tfD^)?032;hfHE@ z>}Qz908W_x>7^$6m~)G)4rZ%UStHi$iXuxPGt4+{D5`o*RAcU^;}S4(;FO0P6JeG_ zG0>HK6BYIG#~&jd#9+o0oE``)nf)jf4h~JaA~rNzf)hhd9#-7p<#u<&Ioz-YG{dad z;3ezvG&qxOKU?4UKbhujr@0`+fwQEFOc0VmF{ubQ6m4TNKvX~$ zjN(((=J9o^_TXwop4(=(3f^KK)V_E?UpiDpd)?222wO%j5FB79kPIxXRYweod5RBhE_{F2SQP8yqE!yJ`#r^l|a_5NAFU2XVl zpvIHmLZh@xP~HrWWo`m(vG!9)0X{pWUC=}uE^X2bRNpLhX`;_YrtmxH?%BVv3T0eP zBo5=*_1pL6AN!+uV3LcZa$sG?62?Tfi6+&q;0!32s%H!x0=xMszV{;TK?D=)4f>s8 z2hz=fn+P()TOf05|EnBKWs~1Gr<0MvhtpS#Xt4LMzZGlFJ^Ye@goTy3z#~D49q)1Z zp;^LqX)}bqF+;+hI%@c}u@Dr)KWxXVA;&WRrZR5QH77qjNK&}PA|2SqDKoKg3MeL* zB0H(5EJ>#NPUsZ?g^q{DOmuLTv@oJ868G}ybQ#g^3iwpjtX!+E5+*adlr* zyI9iVovLmNMM6@kkB1=iLc(uP33fyVv!GvIjU4AHSC5ag*6H9!v>E3w>R>RIt@!^Z3)`C~t}Q zi_qA}dQBybvF%T8059up*X7D;=MMV*v^6#$4jF#gtOmvK(^`F%-8fmX`p>E3j3Vad ze_r_;S?<7ozzh?iv58_JV!956MuHNHK%l$~=&vr5W@Rzt@g`76E(6AFgrU3eXsQ)X zIIwzkEuVlL2g7C5z4gOgT3;hJ7JuZmiX?GUKMpKpN=%THM=?-{l1W9~38(>jRt%}? zRr6A3cSj_P=3!&mdX3?vPE#jeGJU18o1BBXYrInq6^oBu^Arywdm;*@eNOAA$4%c0 zRVuKZn$#)Gn$&HVtT0aaVpBnu!^Yx~dzZL%yBZNw{Pp*0$;y`oG0rz#cZTt z6{7aORSycLB(jp^hV7sqDp&J6p+04!XGzF8`2`>Zc+X!+Z4q!GbVl> z-T}319S}}9!tbQxXRh+B*Ia~>0KD5opHS=7F|Ichi>9`L<8Fa-m$W53PLMc$jUZmV z->;K&0%J`x)V0)W>YzPGhw^G5JQc}wQq|}}LPHXacI%eUmB-f5*X1pWzz>$WC=I)Ik2x7f>=WgsPz=Hjv{e}5>434 zgBQ42*~m|Zg6^Y4i#5JY!h6D6_jteg^KI|EVV4QqfWbbGt?&51K7Ecgdoc$VODub_ z10XwMnI3YIr1)0w_D_uQ)Ye0*?JkcpkK}Rlk<_|LxGtpDJH`{_gc6c`lt2_&0Y2ms zZ)}*_<&p0)A0JobmF;arEN!X`Y6PYtP1tO`a&(djaA${7tB%}$*k)dLujgDn1 zz;r+5HXny$qET#Q;w&u^bDe{TF0`|PQ{>U^0vrhZb{&&e3qn28{H25g#cOnj32esA2gk4QBt0=OZibB#E+{%1XO5ads zNjimASLCz4uIJvcWR0IYft_x0iaNZK5jp7`WkkdiwK@PQgaIw^@u$w7%cyEKs8Ln(E2Z6MT3MQ%8-AD#t7*n>cuBqJWK0rqzsQ1l}#w6SGnI% zVq{K>5oKh0&VbzA&&7|SR1w?XpkIEpgbfFHjMnyO03gAu-u=-xjLPJb%>B1Wl>;l2 zRug4XPcgu6SW88ni0q&-2_Q+bKRi{982|=i?IdQIs99N|ES_27a)n0&D4HL7gfN>|Jc;mK=ujUz&}6wPxi5 zvRsaITnGI-WvgA2fIB=(f<>?$fG{1e-5n_&^eQB2MIyayUSW6}%?W3m!SW)b^Evmq z2qR7kX^eBmce}r1gqCt;z7P`N0u6GT%d-Q;!{atLA>17A2o4w>ANB+dTIBVit$q|6s{iFG`rp~t9m zgKWd(sz7K$sKm&W1y~zRggph{i?PNHVXq1DAI&f#EJ^tvS)|B;5e5uKL)>~tDdrGG z%BZNFB1i<;z*`_HrOAW_Tc<#t%Dw3*0PBE@CQV&R@AkoD;lYk@h9>S{IPDC? zgcEEKX8~T=R?faSV*Az?U+0FSe|W6@qo)x^KR)xyIkI`UW*83azS(Djtt^Uxh(Rh9 z)u{YJk?x!p&_(w^6Wt76y(S^7F1$^RELgXu?hQUYHAQkp@&tGyi~K9;+`w$8)lZVd zFgw6KlS?NBULe`7M}sz6;7ob(im;5Z4nQoE+Tb!ffWEza+uATASQLNOZY6nR$tjcf zUq&$pDN>A0_~3VjBrXkLk;MBYhiMN47V!I>uzaxv^oOA-W2O|@-6he}JEmb3^TqLL zG=>28cw$eiOsP&VkYk3w>g0%ZApe1dEZESaPp?`rW1wpQRAd|5q<#txoaZ2mDk0Mp*K7*oQ>A{ zA#dy_c>+pS;0FUgjd;IUrU|^ddexm!z3LwHos4DH&4ligU_{-dQ|L5)++_Rg?ri~l zOTo-#+@Z1Wo;HtPK3n4NxQHh#F-hVCb5ph6zi(Vzrn>g@G=y;~JsL*JWJ!NX6zYEw|ZQ{MO*;n9dLg4~SJ zKrmIVu7O?yIw+CcH6>RJ-XFMTg;D$bF&o=aE z%ixs>?z(2W@8aF1lLE5+TOfjY7iiA5Lu-q5iWpJ}1XqUZb^@5S@UoYCdGbya3R`04|DkC z8G6-12n~H?RF0>oov8~oijow+)b%8HEcw*r7h6g(2Pm?SiYg)(#e1hN0Mf}z!8>MD z@i0%eNr=_$*pRE%r#9qhkS+wz%jp==f`HW76};s^HO_V6@Y7YC@oW-eLdYRVIRRo0 zII0WB5e14$dO^hYurr?3!m^+ac>%MQA2;(5IcKX(Nekcnmo~wR4{WnA+g zsmirJdt92CKBq>eN?1->XDkw|ji@9S#9Ja-rZqFi$Z6R_(M}PLvANoUKEw$WA;0?F zAGH7d;ctKX$6rgAP|PBVM86o~CL?Tv_o1_N|Kq18zViDcQL`c=ut?fRU}L1J^>0=r zGswCS%R~oX6P^jm18D4t%yB=-Lq4h;X^LyJ;;~T2Yl1x$@+ugfxHl;|$PNo{eA!m6 z|Ij6~=`lCi7wo=hl6f0-E@=}EY`CyAT|YPT0*y}Q<&Z-&K*HQaqqc>UBuR_1lI-!s zP6wZhS1L=z;Hk-&f;=3(uB%+`$Zt)vw|Qd)sB8)%^U2nHt1*RdG-aFuoEDN82i_Gy z1EwLnq6CUrOOe$EmU%2_uEVOM1Z82wO1B$|YI!leFPKv|6Z3a7vUAH*C#8YGq71brje)i~ay-6U0YRjZ#=Wa!*@Sdx7 zja#W-g7Ua?&S+GZe7RtjBLggnFQ2mi0XJk^i2QznhY=ZdKixB*Y;s^^>@`6~CdB|F z!geZZ1t|$hk|a&q?f>+sPE#s=NLn@LG~mO3_28>n{)Q2iz;1DdWI~VFN{A{~gP-=8 zPqT8{gbE+~nIKjgp>~7}C#+EZV_p1zV8qJ$zrFP@g3dB`|QSKo&!A>g=oG$YNHkzQ@-A2@HhGBkLox zCqoBJz*xUOI(n0N#p@I#SmDz_uM)M$HV35fOKIpvC~cnJAwR}H z0<#oL+rjCF$$bVd!}=LrUxU8~I9Sej=D4pC)P*-sPx9#ttk*o?CG+at&l#UvFrn8C z%2B*ufN4%5%w#`g?aA{o{fFL9eq-KP^4a*J18Z}vNv@5%Hm=o0dp@K^wos&#WY8_)n2=q- zOCG1ik$V)|7#x`ay2j2bWaL{S%U;OHM{xqlEBo3jyM;!@^7@4hhslcJ3=jvlR3Xu6 zi2B$}F&iinPesMeyfnLRhT*@k!BM8u3)I+X7I={9x*1yoZ%x~+h$9#5Y>=9)qD4~W z%YBK-qjo)E2h6|!G;*xD)BY>9jbU-kcL_GBvtSpR4Be@dB~^+}#fFeN*rm1$j=86h z%>mV79q*h*551`$fde-~oh<31ks&)-T4;a3aD&O0c(J>OjvXf3evsiCXp|vSlK!!V zq;oS)9XQYSh>3LBLou-5$v5!iYIpl=g$8^b(Bph_z+Ko11924QUSmB`Owdg^=3e$g zzk$urwPLqVzf&pyP>}Yfe1Dj6G-hV6az_r10*`9gFZV#Ph@sGJnV^hVRCQ*lSQsg! zo`P%Wk)_}I;&I>IJ8!kO5jg7D19ml{O%sm!Y?(Ru zS*Ixp+2^~@Cov2?y{M=R=bK2g;)|)P1ikd1a@?;f%iQ{WH>q>nn_P0-Yk=q-_1JD{ z-3&V!Io9Ll#cMy35&PHW{@l-qm+b!U?;*(!?CUQzK|nsm0Bckh6*X_(yanTv!w&P) z)JgOOdQm_h^f*eLeL;{V>2+%f$C_5X>Z->|aFNbi>2lE%>Q`lJ1LoChu1k}{GDY*^ z{W5&6gjDmdOFP8{ZXJ+5`w0B3H-oz(=h+qw)`Q?np)r!RW_^ZT4-K;kqE1x*B>kT& zVG^tND|w+gbCko17Zxt4c-~1cElTM}yj*c|*mC7Aj~05T=qhiq>n*$V z<9zN?Sh~>9ox~0|xrq@vI{F768?Dz%-;J40?m4jay4IvY%;GT&bd2evqMAtb#9Y}0 z`jGJ8#01gj;d_A`EFOdfUWXh5`kFFDS}IKTxaZkPT7)}AInrgCYQv&I&C7azcSN(| zzI>7B{Fv2Vs|0PNPIKqIhHriG&BE`se(TEHxC_JY8600NUg~*PR;NjQ-CXRQa{{wxoq7)JlVPDL+s?06EDLHHcZBo>MSE#ixg%;yq`Bh5eB32rF+5bdD)vro zR^GLQ(#Iz9!cxigrQu_fZ+-D3?2x&!`ox>>8l_ZF%+&9ZV-75(noaiD=P3qAfvS-w zq(N3lwCmlff$P9PA+;!AUvQ=x$?N)@T7)Iug%KI_Nv7Wk$E2yNf#43|QvBvJ3R%jK z9%}!qdQ}M|FEeKCCFP-w8Wl4aT6V=|T`ZgCMJR?gU;u zlBF(-1y=~tk9-!`H+>n!v6`AQh`vM<_k$Nh=jjM`5Dl{ve?{bypO@=+P2tdO=Y2S-W zzTt7nT-I0v2~6G|Us+ zfcTN8?i>GHUO!}g8Swrba|UbLxh7nj$x9Xh@0FV4l5%^|!;Q8S6vI7h&ujly>E2-O z^?OOA4U3QPfFhT!qRSPjt{}Fpj5!d z&<(Oo=|In`4EokIEmA=pbAFmVwPJ?dGj5;0dald;^hx_3aWiIrDkHMM{g6f4zk?m&4~usCq75R01xlisji z9@{KlaRHvE(d@7|6mXW{cj)-hI%dbO|AmO}e${Q%81>E9@ZY+D1h_MjLORB4Ex14o zrKg=<*F00yM+6xzb{1Yk7t<0TJk=dTpS15WI}nBwWOFa*ek3qDT&HaMpo%PUU{8Lk zi3@c-#jK-992JGFKVajM8;$}(Tfl2z8e~_*9d7+j3CeC|mZ)XaA~^#q%LN$rg@Wnq zoPaUr-BaJ*W=^K@k~l(Gs5hFSZBY}AwaQhbL*7Lek__H;WpY?-*n?@C)UlEKoJUc* zz|wKRk?BI)`OGhFU6I-v=&9WzOI5>< z7WuTWaRvrpYvifwJ)S^}b5ouPyUiwHIy5M}A6iNGDbgfgDD1EZ9NzR(iT=!y?O!{$ zU#!D22OjY`yCcM?W0og$<&&J3Ca8DPL?e|@3^-Z$Qc>HO7ViXQG*o?L@VfcCeZXUa z;pQX>sHp_y5~v!|X7i8FK*VE<7^vwGFBBz7pe03}s##@l#(}vL3_p=Hb(8A03i}OJ zLV-~R-6(HTU7evHxU$wggTBav9C1mZ*aklf+x!g7e;3yOEcTXf?=}zAe5Lj@EUdML zyz3J08}50eFk~n6suEr1rK)dDv%y+&eCrl2J;zzCAI>fT-MX(G_^R(}x}*tV$w2mX zLU9?CWRaqi-pu4MNb#rV6+zbyy<2%u5_Gy7MLbvt1mlm~`USRh26i9}v#S{sQSyyj zFHT=`TrUpP01ZtR*+Ma!@Q;m32*ZM|9d5-lcl&40`S|#ptI{?_IwULBYi@(nuMa5I zb-eSymw)}6I{O!)8rO~+WB@squDO-t*dY@ixG~GTvx~z9W>~toq|?jDH7OEqVQzf^ zlgvazE`7abtH(9zQo$pFS4zcIk-6g+K#$BMx-#H?Xth0z=Vw-lGagWwJTr^!FR8|z3Q5@b2^@tEA~l8LpM1pqJE?ZUTGecdDbv>;G9I35E9m?V~nN zq!Z{uK860LH>JrgtKDj#s8qYdt;{z;i5aUpUc$KfH-QG>%M*LTVnlE-S}gVyH0%QA z$M?S7Jl(hh{@K|XIi%n#Mxb)W#2%@j7${smKt*BimU>MJDUc+FHNobzl>Y$wK`kAd z_x7@}^&0FZh3BcWW0}OTa(buqK;Y*Q@zC&j>DY}Fjwl!If=J|p$O_=v(5p_npYg2M zWJxyjaE1jxUG?V4Im^bLA=UB@8lM=vVhe0pg`q(#nSa_qXR%{w$F7OrGmq?XQAl6C znBN}Hc)u(cq{m(kp2Ae9A}X^NnZff{&JwUZe+K*94KM!G?RQ48vhPDTIcay`;LUQA zVzEaQ(?gN_R1^kwaUYzZ+|I|AP};wi(l?aJ0!%+&4rRoUa&tc1SJJ4Ib{s}u{yYjyATJ0VjTK6u{wiRXie23an-E$(+}pPtXdEV&Yw zD#e3{!iXClIAf|B^5u>OWkYF5DcvFORknqfgzNqKogR>S2BT2my*nvxB<1S*dDU}J zhhm@bzm(E1M~&Dk%lrQR_^Tc|teu%Lvzvd#rCGVfxg~Ob*|-cJyrfqxnsQN`tEhFi zg|4yu$8U+CF)AI)C$+&ycC9qb)@jVCFHVGiU=&NqRPQ&)>9OQ969eKZ#WYZ)o{GZC zE4`{+ur8!n)eNCpAb+LrctY>x!iYVdN#l%O@*@8S5u2C{`V#-pByFZN1B%E?LprA~ zBv%xv(iSrCBWw~rAW0A+S`&HJqt|t-2O1t2(c9;Y-JE+}H>v-2MS--jK(Zu=hT2aZ zB)%6$q`B3|+vrkZEQ6l+9@nF?J)W_ViLPBIMrp=~FZE8ck*8Ue~%X3&q za~r~{^A{~Or*(APeM?*W4`gGE09(8?_$cXja@gP(*=mCP)fBUeBFm|$cK0Mn8l>rE zxoF!Ian3d4kBrJz`Wb{+0@Y_88ifm-pgJb)^7Os{BO)s2t}Q01W62?t7r&cg3Mi6G zMd|%Ec_t_^9W{S|>~VppFybN*O~dazRox(~_ST{S(5x)+-YnN{^^Nh|#yplC3OW#Y zTDr&cIO&pVH_KvzqGvTh$hs_O8Cf8L=Q4ozphm7&?R|a28@5y#maLA&62pe8#%XmN zSCG6Kdj3;uoO}+vr($vPVUVfay;+&TdoXRg8$_is*PcdZF!>dz#zkcLfgPW`$`2U!kbl^?aEfbm7Kr!_c zIY&h$3(m`)K=S<#dRs&fB_;23xYl{%b@@(sI)0lq7v3X;VBRh9`!O>=NL}ZLSHveNKx=8$9<2I5Td_TRi)m zR(qAW^f}!OuM6vQN(E}&KBo=L4SAnadhlIl@Y{ELoPPCwM1G*}TER1vR6*ej0~?NM}%x>aK_E-YSA+YID} zi<0;MeXe=uZJ7LVh$I1ann__|B#GV@F+ES$1ta51Q#RnXE* z2Q6Y23vNLAa0ad8cZco`UKW(EK13=6H$1RGIgn{U!_`6>ir494>5(bDZf&91R;F6m z0ntuFRmsKh78$lGs|SN7lXr?FdUeo-dGso5bABwm!8T2uErpPoIDGD@El@ zG(t!s`%xt6bYL4;XA)?L9>YALNG}x?>;6Q(F~IQOI?ZwFoTg4wC$HCF8|Rqu>wy@o z75EN*{pg3w=El7KYp{Cddbb4Sub0fo75}aF; z<^S@<+;f^fr{m-ic_NQ>-xyG*+4TCxw=&-O^e5f_+W1D+kJ7(Bc)waYc<<2n{47=dvKZDlx26(dX2x$Z=mJG8`Wd z&g3VC<-*l#Q#%3@!Llx33YZj#!*(i?0+IsonZ}SNVcJU)ZJ@sM$fwubQ0AiA9C)kF z(jpki3a_{n`DnkSZNb(xn32&Xe+=yzZEu}DPgHOaP0yRg=J#gpDgTzMQ8X2Q{k>YU z(t$-&ris2vq?nBqyjMu;yftmLGC`TiThC7*7X%AMSN+?erO4G8nY?TK`-(H3H6e?= zkd%4!)||t7qx!M=y0T5k%~C(V`S49aqlG7qXc{#<_0{Dmf{fB*S4RPq{Jk``@lg|5vzKBZ*{^_# zGT6n5VU_fKMPk?i=Ye!AwaybO@S2t9X6}xBNOsTz^thon!G+i71a7|oG**y`H{LH1 zya>l2()x%t%v3q+lqws@b{B^whShrf%VQF6D;hY7$||=Mn$a{C{SA}=m5=A9$Bxcoe8_1X?J1QbSbD>d_vvl6fHVK zZ*e}MuJX)>kQh=ZBUw~SpW&v2PW|@R3dRT(*GCU7 zkqr(El_C?UWKawg-KS7dE4?;)ULY+2Sj~8E5?1ua^R{|)L3jBTq=hb$^f|2?cSw)| z$KA8~oYLs^UQb6S1whv;JE!iw6i69!yZ84x^Dy>TDxsVu=HAS!gbwlhBM$_|c%6Z6 zMdgacjP}8+E#8-T8z$fIxHr94d_tkcOi$z+%k@iC4pfSth}<(1Hpt?0Or85huw>%> zTIhjxBpLJc-nHKKnmYkCyumArW>u-0=u@iGflb0p-a=7_FxGETFt%H~ zDs3WZ^p^2R;{S-ZQ*`Kq{2Ay-x&J}hcXNMn`^_cq#?GzKY*9Uy;+Pfh?EZ249K#5l z=c>;~IPmg2`QP0$=fBV)Q^#AxYgWd%KMAdKH;l+sZ}ZCo;>!eOqTdSl)ggwz3CVIT zft@vi={cs%^KV(8$Jo&{OiCa1pWpt?55I0i*u%Hp{wX=c%}#b;@5^-)Jl0YS)H-~M z^zhI?0&;#0>3{>m3NIneOa}efwUAs9L*5Uj2`$kKNUnvXTiEN0owW3-0^W)l+Wi4Q zQw4fY+o;4&7J5|zkXt~C5JKX-s6ue?OD*_Zu4Y=l>in3^yhA~~(%7(tz>%q4Gk&*E2Hi!HNs4vikB;wL#613$Seuw z#{)+V{7r?5Ue^w2ql2@?c*b}hP!ImaSaB0CIS=1344N%R-1BW8>+U{_Kd%qITV&p- z?IjUK+e{D_PcdsKqC>xi(z5I@3ds9 z)Wg4+G1e%EG%X)#$pUVI$bl{5BokPyp%@)SR#H)D%R+tLgNR&lifd1VUUiZjl{AxK ze#(}-KWoNVh6!uFVPwyG{A6K{xpHz?UJ({`a{_o|{&GY4$$dpa*u%&}u-m@w@i-`V z&c{dQYzkViz0MVuGF?)j^U? zy=GTHk$9s-PyJntx8d(esA9 zmQIEs7hcCYnDtZEdu<8cK^vaIDMo5J%L!#pQ_#;FyPa}4q9!aNz6C6 za+cv_aP;VXvg0J`!`RQh>XC9FA?7E$VAt{ul7z6kOzVt}0csn)>V*0()8};FtHN{e zGg@?)NeXyKiu|hrljz!zyTO$-jp4DmEgA>DAH5YLkykm+MNR5r3QAp!)e&Bk3j(djhF;Om7&I~j+Yo6Zk4=Mu8 zS{$KQttBhSdj1no8_2`EIpC8#MM?ORkr)Ysuh*I-8-5{&ULE`v?w{P2%wcO`K6{w9 z?>~%KYX0^&zeyU#k}piK)J`!sDbhkkVcnCV4ghPo&jft|<>B|heXl(ev?{0@(l1xJ zx_P*-^(Gh7gL@O(JNL$n(tM!IJ z5hiygLzN|hE6pQGf(cZ56%HJDs*$w2J^ETF7O)`Wd5ydn%-kfpB|P3w8}GL!qLg0c za@+OX%y>U^2O=lEwnYJr8&%qP{$tr=nO3hp>w)AKsp`^@#{*rh&BTp;(KZx$FFH0w zBqy{vuJ^XRC-mfGBifuK>b)e*foX8wP(f^rEeVxWi7mLwp#b71|{FBE-oDjO{v}fubjsV8aN6-$D^i z5bt-_Rg1;1+ub0T0YqX0QP>Rnpt_6Xi}FRt0c0DGwQa#>l#tJbk!3f%3ghnk8sW3} zBd=8?$$`aAi3uR`C?=aCnN-xCnWur$DN7XZw}~M4`D1q^U@=_TJU&lx0$LU0^=-V8 z@Mh&B8Y83`bjFk(sOgKFvP)Sb+~HPBFXmlw!I_qdc7~g&KFeO^3rE=i126oJkqqD_ zaKd-5UTa=h{_GCjfuj>FRq#v0(jg+*?vC+EkR63QypXAG279ZJml4p+z{M>7LQ=@n z;@OpuyAe&$l4E^XQDDCl(vV>?Pce*5l5{I`#W1=Qq|wr_#96(3y^Z&7HbBy#ku?&) zus*|%q`C=CADR>8IIN$Uh5F`%0>hd|6^AEkyA`V#NYX+3JcEt{I=CgSMZ5xf>04MG zQzAGg&-K&0^}3a~RP%S(sR!xMMHnF-hE8FZV|F|Yvw43dY|_^ErAA$(_{YE0k=4VI z96PXN$uZGSNffh%BAciv+_-^ZtVW6OfTVe#ra>-G-1eb8f|{$}Y0LQ7sfPkr@>-O( zu{d75pd&!dizl_g2zJ!`;NbV8%zL)HB&ifEz16#wn4q3Dc`9u zfFydW$Boeb&<=WuCeIGsOV3+03)dg5&gV^I^Ly;NWY*+03iBW}7hU4OjuV!Ub{aU5 zYvj<4L*1y%lqS*LN}XScH&jHm3y!%D_~4r8<;p}=2W?ku+s-*R7eYfsT?ewFulrV(uV`62DdZM{j?Rj-Z`ll_c8{l zkE`>XuaML3P{i#su(I`@cxbcrwEX>Bw(^$GXRDFy;^a^N@~3d~?i$Z38V9y9^d_5^ zIEsm-$OVic1TG3f}G4D;jlgrmgUFAU+H#% zPV&1A8E$p(&$7|3Brz<7B@)5%MYG*V_T?YW#jz*VgCUlo9r*p;^!p3!8PehF1K;XPx0iUX{pGuSXC@(&u9C9Vn!zz62E@Lb~)Uf7+ z?Z7kaOkopR7x@|8VA=iO-$RleI3xpYyoS^qp!z_O^Q(GPX?xzA zhM%YT3d1=+Jm2Bni4l7r!?U=rBat2j8L=%g_^) zSl`$0gv1+I=he(Sf=aJDp_uUz?FFr~23R<-LjcYdEh?L!&61P`u3!es`|$AxN+5nx zWzc)&wpbIZ-u#dswb|kGUaCd+IQ^g<*SB!?o9qAbRqq%Fq^4@*j=^LlO!~lfP+Eig zEKd<9xC_15a>BKnd7CvSp;C6W>z*mt+(~O}GN(>=&IzxR-*>qHp_)pfSM7(!a@w7u zZ47o|Z&6%v(XNRwBm}OQP%=Bi1zjYCiY6gWx7sz0UNK>@7lfD%)8UP@Njss@obfWW znVVECPgB=Js`x$ep}-n>qRYo!k!|WUbrVdagV+&zc$ZOWYMd}jgn15oJbPqKZdgI9 zy-e^US)dEbkgZgp^CsRe<0av6eZ|AKqRkpT=fHkm7CqM_JoA%1Z!CTNFCYK4mhOR? zHca2_<>xAn%BtmtTc3RVXT7S5#+Dm(ye44+6v^ShFK_BX3Kd&rcCnr~RQa39)2F8v zNl&hWqSGcFHm59i*sV7U3HgP&>;JX??TR^rpN3Mj;(z(?C2w5)0EnOe4dUE8qF}&v%6#!f02Av4_9S$vI?3Z}9(*9!i* z;sZ5u%nmQ|$rWRx6n_H~bE#dvBT&&r2B@3ro=ftXZ|8JET^2vw`?TJYK$Xci;s1q1VcydvE^KH#v@ zhoxXGUQ#?WSFi?*qbw2RVdsW7$gt4BK5FE-@6)nXe(n@@xyebKjH&-aNVIv{B$q9^ z1Dh@^36(l#hyN0{Dpj&TugaX%;Ex@>nw6ch)qWK|{XzBEk3iY0?3HenRruU>E%ROO z{s4SE+a^@_?4Yj+n_Y2CHh;t9KBr{gW3KqrJ=d5iILh`QvsBF((^3EO14 zkRty|dR|ZDZC)*%&djURBv4P@+%spz+?ubaL-&b#O{22c6}uG|(+|m3-wv_u-|)bz zvHmsN{0$rau9xGN*!SLw-Bv?VT|1)fn~^V_S2^cAIEg>J}xr_DS)5Qph_dR0TnV`&)y7Y=f; zMvL0ikhmqZNK^mS)lpU8=rxaW(EGE{nT5# z$`kn!3|BE}90*bg@*`MRpg5wR1e;nK{`!c1AerYRZstXXT(IQ8dt#RI!$pE@*SZ;) zbdXLjfK+ZQ0c)n~W^4^a<0g(=uq9EG1#7|q85YcDyCIdlzy8J?8P9s;9al<1U}@+k zZ3D%`Q)CSlRVL0A>m)5SD$nKe9DYZ*_K4u{#2Vq1kkx)i1YdYIdwy&uV`SLH;jsQz zCXo#L^{f}!mE{)`RgyQ1c5KYLZ-1YZIj~#)iU|tNPz>Y@A4fueL(>j}F4-<^=IMYp z{XsyF4EYLm5ce$?W=+=a2OU$&KLBDPCp>z3Qz#~h;D8JuC|n#2D(C?y7M*0U=oG25 zkC9~O9^m-ulCN`rDA(4jz^MIgHM8n!diHPEo}U+ zB`D;Cp5gTd_p7X!d3u8#J;RtZ)@A{q*QQtX5y+zJ)V}f&FEp*oee>W%wnBcEq zpOHrcJQzAiKX2!6k7!fWX^#Hj{+xvWdwAsss)semnd59ze$xIT(808r3B^S z*V}_H&R9sr^LOy_=EOO-y4GuwA%Cecbb$alGm{19Wx0V{0^1-iz8B2@f9N2uGQ+zo zvYc*|?iQ!hu?*%~Ci->JF=N+^UqQ}zHZx1c8Ya1~##RI6vLZ>L`?0_j{xO)mi~idU z<>5K^6?fGK)Qf^IkenB`z5QRoB%=5!{DwX9y?*g~d?U)70-P3-7zdVH+f8tjKrw46 zvYLwOR#Yh#jL#2*l8}rkYxsIU?Ii{ZTOp>r+O5Xz*;Wt3z6hILc8%lWKHT+mIiKSN+z_)F$Y^0)Fj-%%bR&svSfOlrieTQzwR<+$Fy^r9rPW~erdD`8aQT>=C2QPPyWuk+-NhbdoWF=q`$U~Ci`Z;t+Z<5t( zt|>F!ckvDY>ab_}KIsm+aNxKj^iUA?tJx5FLj752h8-y#MvN_k1hN)Ftk0(i)+U*g zb8tb61N&@Ps5~&Lg#RXwYZt782C)Sm$)nj<_~}Ym0*R-iNAKv3?w=h0gHh2eTdtc) zIviNhtQ<2`(LAA;UWz;%Y`xhGVqh7RrlN`3Bt%{fEE_arOtuTwyH(Tm4D^&Z%uk}x z(73NS;16yXaOQ>`3PPrAbn6(pwBF?5=nVQW|0XYuE+9AMr6CxO#HyY|=;GQ#if3LX zr-D0Z3{xgcqGuiE;}dHH%{)VY@@;;GjJYYp&1 zteAlh;Yj3F)~j%u8eub&=%U4M(yMBOI|H{tEDjxy1I6XIbNhuQh~jL~*x&(93hb4Q zJ8zBss&B2aF+G|s$ds%_etrlU$#%*zreFn`ZL@4G5Db>%!h*?cID``}#>{V3f7{b2 zmwtTam2+gX1Iwj-CI)a8#bi(!_(94|T!2-VOpE%HJNV~)BGB1|76Pl{V!gp=6GH-m1a1`({GJx1ga15Px z>6Y*BFt=4%qAdlYc2S42lGeZer-u*%tJhrjSUaVHx0v_WKmGkOX^~zYm0F3xKc3Ss zZt$T|_~fMJ9oJPQl`s6$Z;XQD*38wD$t?%ARhOFVp}(M*2NdZ>QGjVpUDIpd|7=bV z!6Us2>7lxndcRyT>}wvpnxNdsFM%`$$T+|~cC+e=SQqjHNDup+PN*vb5>>I#s6mSx z?hJZ;80>bVgBwGlp{Hn$s6y~0^nR$eTCpx5Zt{xB+O3i4VQA*Ih3a^@-j=&NZ{OeA^erzDm0tuuV;U83cdR1x2D$iwx%12H=3O4K4=GQ?za+OmJMoS8uadP6jJcZJ6BSip7khpo>NZ68nozP9ULv8s}1KUL()a zUL!^jBS_(Q$Onid8)O*bJrmSBYVCQ3zBvou{m@57<(SjEcU+_J@A3COF;COvB6=M6 z{~22X=T85}paNjaTuLL0Ay7OS_Qf6Y<#Nol>zqC+`=2jCY9YW4I_Ar#?0+Cv{pMDP zQKu|V=*lNK+%`fEyjebJqJc^%WiAY^^wc=aT>)7x~{-90)w8sRaPp+-|WI5NxyjPiNtzDXA%S zXL`-8lxOP(99X!pm{f-DK?B^kWuyk!ySrx{aBhYIA)Dji>6>RcvJfaNPuuS@J5b)g z^vF$RlqH&zA08wrFO4oLGm$(66q8Glom7;eYc~=Y#|i$r)dR@GyA@|V`$Kht)pL`_ z&ByvPQV1_)N%nY_@(%`d*j(;d zh^7~79vcJe^| z4=J~`hB6cAaHj5{_p2YfcPKBDj4A2P{Z7dd>msu$D>Hv>2BBv`W)-OZb)K-tJpYml zgdBKt#FDK!n8Z*@7b!Z?-Pht>8dxN`4)UZ{yeSAoO@cC&G{~0Abv!VOw|ZnsOT3?W zUi6%gJ_0Nqw-O4^T`%i!^4zJdy2%&r`@XMHJuUvoYZXa)Y1C7RiF(ST7ziR|Qc>6k z5h?&A9rQAn`Q5;9TuE)Bo%hP0kVKbwL+(`z4P>aU(IUG<9<4*nP;p`vK%RN{q&Ow-Wwdg!m!@9T>0C-9 zN2RtTbQ8<;a+8~m)(l!_mcRt;`F z6O?DZmgSFkDE#zmyXX}1SlZ{bC$dMp&a=;{K=P?)iloFPN1Ec=taxmn zpsp=`fP-(&mY?jaFkFJY!*aMyeDdUDbG|tiTDKl@%B{=;63sH`W6qcqTQJH@1;*vG z1QE7eX-ntggb2so988zr6`D7S`AYduSlYoP@Jm!#{J&<)z=cvtKYqP#PM_1`*Q*Eq zTMz#|GPmF90I=Rdfjvo3n%>7zN9Nj91zCdXq2|QUV@nY_l5y;)e($yK{D*lmoWr(i zU`cP=64>awNK{P4h8`0&D~hRnb+vd6)8Tnd*(6-=b%fultc6mbo8jl=eF|O3W6%8A zU4on8b_IfQkw%1vkt{IBXV~#DoZh=wC|qP-7s6!&?6_i*uvC$Jazj}>=O(1S=v7Uj zaYPH}hvEDrl$7Lz_d7Kzv(!i69NAI(oif$6fk^=uWgT=Xo#Ud#+SN3A3ArOT+{Q|c z{3&`>yC5w@izk^-i-s<`wQi54{Z7c?P#BRxCx>BAz(!>@|M(2ODnSW3czP9bfi4Va z@lH_g6z%e;g)Z~`PPnBi6|eS-_k&CvJUPXSamJH=r&{rLey4MS5<9r*RSy;T*R{~a zGaF?2Q}ACrYY0hI7e*Wm%7ixe{Z2?Mf`3ybuSe#vRU}34zSc0oxUu&CZOJ^cakw&S z2S(W*lkM{kiUC%kZB$eNZv_;j?GM)uEVBss{SE%!7(p?VC26BYSo^A19iNda#+==Q z0r7rdxLt-OB1v>H5Kf(UKRzQ>-RfdDXUmAh*C^;TcJbwVJFdVQAB+S0>R5a*3CeCD z5-J0k(It)ds|sukErk@ley7u3Dg0=!ZhpVhO66YpN)Q>qPY0!K&B`i8nJl( z!4EeypF5FaHd64eqtMS^PWM8(@gveDSjO8e-KMVmdc~ZFX5sHp931-u&` zmu5rDk9x>Pz#LhO2_h|}y?*TM>t!(n*r%||9d-x|V-b&8+4{qc<~}-y-8Hc|50Tuz z!sm($Gzu?_II75jp48fh{AEzRdScpcaO}s9hYJV}k2%AJs6%Wh zU4%rK1ZBnS6=3xm0)}=dSF{jd!w4GlOSXkr@O$rm$GkL=rKYb+`hb**kp)t_V0?1e z9?vtL@qY6$MQA^ya5Tt(RXFJ*7p6Thmi|!exm#MEO<$To}E~vO5C}@ESK{i1|0Y#A=7c4Cm5HyOo zAQ3H!Dg{Lf75?WWK}jT<3km;df78FIODOhHM73h@mZ;#3t3i1nz~h&Cv32j(wefv8tj^O}L^ z2&R3=k%1r31-zb}^1(BbrVY%2(scO0lc5Q6e)3NHO{Ez+!iPPV$kvxELj$lL5|_`Wl&~60qatJ2I(h%J)!rBAyYzB$ zm%b`()FgiMDztB4{pP;cw)j0@zff)wp5sG_~E zNcF00`u1#X708%jBQfg8B(kePG%=2k+>>}B+Ds_QcZ_*^cKAkt470&|_MVQ|V;dR6 zByq;g{zjSS`)0eMdr&)(d_J6Aaoso%S#Lq#(@!ZMQl!^J--C5Q+4N=r#RI4W!Z<7T zpc})c>phdmK_*RjhvWocjS)s)&v>>nW|rFBf%#$CG)$*h#E*ejY|=pKy2^P)V(o6x zR^Z(+-LDnumT+lzAP^Seh{Oza_RMxbVWzr{?wP(}#?uq^P|%d7?()hNB~BUuh1X8; zZqZI9lF*@qQ3BIRw}jyDV05?Kyul5J2Pc(~z*qmf{A;i3c^5{x}CxmAIV?;Wo?!7h&zfl;C= zwO&&m5$%0jzKhOS*b#hc@zIneOHBRBxe}a@x}-aK+kjxb z1}b2j<``;Ud7Mm^~}Dg%op$XN*0_V$0Tj!D9KW% zfx`goV2Tvg{)ls4dQG)w8{0T_i=MuZ}-BDLMduHtg02cbjb%C;U+vznfb!&Ae{w z`LTQz+4;)!*;iRaiV7(ubUfuzk&V!+fuyNuq_83e+S%BsvN|{KJO;luJXssldX=4i(J#GDN@Hr|4c3tms z_ab@|>2q=b(Q0hCK*Q=%r*3{c`a*^6AvT z)qK6Cgsu}{5q+yz=icV~7@F&&f(v1i!2NM<70g|>z-yzPm+nkvbZUMJbU+HN+McVW@e^NW@iEy8T77BK1%SI!<`MgAv9?4Y}1?Ngy1@YnC;px&gr}2@D8ZV9piur!FP0p{r*}LD`GjZ8z$A-() z0XZn_PSql_b}J-qAhlFNAC#;U9h7Lh!%OBDDNd{USZ%McWWGsVW^+i2f0DPu&4u;o zjT8>n_i^;LZ8!`Qj2S<*YcH{$fw+`gxo~>Z#@zJrymHm&Z*63bg_i=$X^8~52}9N` z?h|fgv=@VyL4pyIjFhE<&%cc!siIbet z@i0~P@Mq*RPoMhf+}+aBK&MRNcKjRddX{$F$I+*7 zdQn~XJ$v-#zxYlzJ0;DpUwEG+xNwVep@jpIK`EhAF@=hZ_o(nsWECk_VLx&xB?M2?O#${)8tHb7T1TdaRP5aJ^JxHD%Vea`1OhEo5r>i^LKl)FM$B-{h8hW@bCnGZ6_rVW- zZml!qvI`m@X&Mq6*-9xBC=yRa4oI(wp@wNvDuCeOoRv_eUg=k^ETOLX?ONEz>xRUe z)0Zw5!^r^|7S}m@gA+2S>7u#=Gt5?Gf<%3gWWF-_nIjfBIzTDGiP=j z^$RYFapBS`K>gl+MSMg(bSWC>MxKseGQSa&$+kv-n0G7>NlU*}WYQStX@ynKmMI2$ zixA5zJ81)bEU1U*m8SjLdT1y>Qq&BBrV5uwcUh#1(1q$ri+Dq*_7*P>Dwse!hTlG? zUXD~z6(LE}+SMblR1Sdp6U^KG!X5scpgwNGv5A-d!HlQ}?|khSq{f92bzvsdt@d+r)yFI%Wghx0}fnx(Igc~QGe^1XQpyGwR4V<6 zs#%7^n&h25%h*M2D$Gok=t8mR4O9vY(1N53)#gBF85vi;#tpL&Do5+ zaI_o+8jfb<#eWxH?d?Xg(S;|kJr%U{_zdy`t5bvUkRJY_vWt1se&m87$mekT4z59G?q$~LA z$m4)oTBquT{`_QhAvoJ#%6r3;-5oynJQ>^(pk=7f`{X{3{=x|@)NlH!cQj^uVhEj@ zOiEqYo?NuBC#NXo35pz}BJ14JrKs?o8hpd+-dt>T21g~4HMug{cqYXk=}XX&)t-## zq&F>2QQ^i#(kncnST^~jTIZI->lJGEGCEi(6@{aW<|83qXLNaRWAHqaw&uWuYJMhN zs)myx710Xb_j%}c9F>66DA6<&SvJ3q1ZkYmH10dG9jmOlL2S@%6nbTP?Ol*H-Nf{U zo$ZzLfKs%@{eUlYvMYLJCN=j@g-!Fb=_rvF+Ja{8)TA@ob4-KJ-UUVBZLH&cgdtC8 zbOSQvMh^bUX+T`Jjkun#^0iJqJ*$Z2x_>{hk&ddDBat^M#-Jp~dQUVoMU!D2ovAMO z!PI1~$az!JFi_(H<6$m$+AiBLPMr2AWHhZoE{qu)twEJRhoq>|i>=eznrSWK4o_r` zc_KMvs>2*Z*l+<2*RaZH5cNZ4yUMzQ-i9(At|*Xq3|A8H?sc;uwFI@UFrj111s1*< zCmgenaEa*>VqNr2rd^#Y`U2u!wX^=;?DD_A_}zd1Hd0C}ClHDnuG-aQ8-!|lS9^P# zz1kn29oIyXhpX&&VM}qy!qd&Al-U%JMn^`?t(PAlsHccE+1b$30rhi@Jd}plg|6fo z1m<12qC0M9ne3@4{`;WbyOaKM5lA#*khd1T^a;zMOvYK}gq!Uc##bGUfq2ffwkMpA z1%H}8%M6yq*K^jB+~L6Dx*0M8tDqry?8B52a;f{N$moeF{`um3I?MkHde{6W`KpN- z>g|)ef!(Qtzs@~+;wt`4O(Jtf-6ZdvvufhD#mJ=EL)TA>e?5`eCfXbVOg|t`d;`+2 zf4jjpExNC06#x^O;9j^vVY&^TSPz<2j)T&u!DET1=fHv6A=}^)=sx(lwO*CW261g@ zQ)Q?zISIlsygEA9U0V^jQ;Gad2@q&O;xcebvgwax9doe1%yDD={1M=cgy-MRVMlEv zVVDVWTvO%&g~5nO_Q7u`kbx>^ zi;(Nt?*9>ZC?{v%c_WcMB-BG{*c^|m2y78MrN3v$K-&cp%W1;AHRL8v_{s?+5NfTUV|g9kxlj2Av|f|L1Bzxr zEboq2OGuFd&hV+6ew0a`-UFQ8slZWh0Jd6Ws@H3tUNzkL+01g%V?5tsSQ!x@hBYIm zoz??8(rw$eWEhv0`gyJ2TInh5ce+Q4Do52Hq`l52dmMqX z;v%|;hKL~u-2p!`I4(!Ln`J}%99iUwLZlV)ohd-!!7glsYQr^OrKSuG%Kq}Zv% z(4xa&GtR-d87WMjxsDTV#xLpKyy)v@vopT?*PoGNE^KygSeTtAO4&e>vs5IqWZwdn znKhv~UfI66qM9I}@J5nsU{p|>g3L=(>AnSh3)Y01$heb&v^}pIdScyiPNB-}m_08HQp*>cgr`I$f` zmjx;6rHwr7ki+s{1fJauPY%Y_b1c>1p=Tg8><(_Y`ikt)JZlxTSH$13p}BV6_eQ8m zF48C&nN24_M{GNN-@Ss~@7v94*E7-68|8KKSe_Q+b&YatQ3A(#b4a5+ddB9EfeBdX z*(#_H>0p6BsLvyHdK@1wIuzWms9>__PWixuKRe8RIT~ieVrB$8Z(msSbJ3t5R5a;- z$$D-rZmvrqK!65lNI-ZyrA)%;b>t&YlZs@A=Q&wm^?+@p$2<6&jr{!Da`!%uYoS%f zOk#r=Jd-F<(!$GAQ&dOj&!rc|^>jO|&odznZQ6fJehTWn>qE+YR`4#9>WD6SqkpAe zCbBtbusE}oZe#JPL>9UVCiluJ{rV>yWcrxXK$yESpfnKAJm|k-C~z?tIR0UocdWANt;M2EgpXXr3ALO-Bcu&wgA5CXEJf4r8@!C?>E^0uSi;`D;Mt_S#&{{NZ#vUiDKi|h`sc3@cld^)S25f3Y``%OyB8@NXMQBZ98&+d1qAH|XB#{Mo(Vnyt$1`5R}D z+b$fAS#6=L_9dn4p-49si3^x^ShRFAT{PA%93WK!V?VS3YV)y3s$X%)txSLw45(3+ zsm5xCL>4iG{twtdlqS!RW*S=+4T5&X{%Izj+~yE%o$MxP_MKf6wP1H36i;CkC!5Bd zdl`(j7`Y&@$Pj6DE5x{pLV*b^H^kT>-VCdKyCND@+HH^-#b)mmRUWf-M!h^&f`39K zIjr6jW-Rg?4BwcF65;PEk#?}L;{uH5-pIiR zed5BuZ_SMPY|DrXFPaK1y!Iz-nMHI>a29i$T!2c;Ki^=nwzA``?f-&4k6yDH)R6W; zb$a@Dz7T_~o+jhnD~omn#`2FaX;8ZP$mwzxT)<#4wK#b2t#4<0#`#s+UxKWq&0RJ; zV?%Sjie5!NEcc1#V>6Mqz^he&-faSNJ`86NbY8Z*Z8tJ`Lt(=jUJ<0Aa)F0ICmEO! z?Y)s*{cnfgZCTW@2sHy%&E4joN?+#nfdUHN#wZO7g}$@X{1RD**_9E)!~!!`(y zq}#S(GE8oJ{NA%a7Dt$!;;q>V9Vwnj8ZBPh4ByCu18mt`qd`YM5>_mYsy0|+cX$?fHJ}T6+|3}{t2pyI zu(x2v3YM}M?}*X(^C1SA>1V2BSEU7t&jJl_2cXoh4AM#T?#<#6FU%c1UQb7E@Y~2n zjN5p*Z2!scmENcSzIxcAXRc=+v&%g)X;UYgxp4eRn^lW=tk&md;BiCv$j~~HtIi+! z;#Xp`jfvTI?KfnD3opEPTNu9`lrotjTdBxB(hL0KBGZ2xH4WeGTJ+^NkPovXxLKLQ zY$7Lk$D!yKr>e0eX682HV5dehd?UDqlTD*`&H25%_2SsZs&bddWuEy3%TO14W5nY) zfp~+dsnI02Yhr#I@jSu#pZiJ-g6l3J|M4${pIIjZT-Ld@k)=2QE%CMVU%FYSotk-s zX@FQ7aEtYl&*d1-Lxv6v{dTjaw&y+kEXYZuGf>wTl9hpI9Zf&-~Pw$sCUgXPEho$ACTiNJma)k%sA&M9Au#xF7M@Lzhdim})0|$IW!!b@ndLpwO|Y{&??XZmbTD0+todhTVJtmMji) z0v4sJxgBUdRdI>qx^Sq*#*DjS?m)m!*+4+M90UG(%{~xx{jg;IpY{b_k?fS+U{U+o zASjvN2Io0!ZEZ0^&u(zW(zCa4%AVD)mwxvLX0!9X)6_fU+$&T4aM!{}-Jq1f(%M2r z7DuE>_DBmPeRQiwJ#%kvEq&bgIEc^|N8~c;WSjpclFLK|fARIJQbU+dxbwAJ!YEPy zgnN($h+PmbiRG1eTnJ2)lu+B}mn(PC4FRizHfZk2N~oiuNi$Pb>)ngpyXcF)_vL!c zCeo(Zpi`%J%FlUin;g&I%ajI2Ptj|t1chE$0ZoVWE=aCHs%yF=&RBlfhnr>Og`{LXV1*056U5igqwSq&ae(osO2aNGRShI z*>s<944nFrj`|CSZImZ)ny1F+{`}^i`p9Hoejhqx1QbJLYuJRrIcaL%}!aeeBk+I7Q!{T|l1zaIv z$^DS0&Zf^0Ofn<)Lbswq&?UwFF#&bLYN*Cs?T&ZC{`_Wk9UB|6R{Mx{F>;-C#?BvSH}o_{>W_($VGo7Qf&EYcOycUDtIY{`;Mno7O7TT-+nq zEiVF%rlE!`ky37=$VMtMSA@J(2gvPNV5U}uVp(|w6q4mboH}aK7?>+dAlVOZ%j=xI zbs%b2^go)-(SNu9_f7H%H^<0@aq+1IE-q6_VC+0kMW!y$VuhA26xclNFjpS7JT7R|mVK9-8kAGKeg{ zv3dFtka_~Sg9c%=cQr4IO@`YnK^A+2q&XVh?!I*C(%sS>{^_CE3t=p?!M_Jd`W|^= zys`|kH>NxdFyyP?>}QPT6Li=QuNRoVksW^WcUXV<*N4CGw$>DN*$|NpvD2q~GuRWC zO)m$-VtXX4pMV3m?HObT{+!*n;xt8$`*3b-?#te>aO9n%<3aVrV!5860VMyP*QmI#<_q0_~CvS1@X3mp1{yn>vICek$;ZC#%D8s+#!`;ns z`#1rG`Zu4D2UDK!*S#P{5x?%$M{1(_}SZkUk>d~X!gg73}6&T-(xwDhvYJ+-^3tayW;;jv}SN+ zeEf|^p=O6x@qe{D$X*wY!JM{mQ7b8B8AVE{$c^5$^m|pLe?tF^4oMY(@zpG!R7>B3 z6+dq2HNB*di49skW%bl5(yrOatO(x7lnJg%8#NX0T!+fNCYW>57(A|nd8!szj(W(d zx$T-eUUhlVzd&y~n*j z$Bdvf|BNjp--Qtbnr=g)comeglp=?y$Sv|O#1$$G=HxJWqC_@_*$iwVs8pfXp#DIP z`Yym|FLT992mWj?S?}HIp$o-x+t}@H=}ZG%#m6wt!9dfU(BM?_KNaH{AVSJ(rE$;W zfEX7U6!W&hNMok_kRXr7`n$ zkT$Of>;j+pKtMc@sO{q2nQ~T++Fg3@6R`5d2AT%&tsMldr-SS$kUhPlW4_FnsF zJQ~+66tNYeJM;QVKUgjSeaAHeKhD7%kQ}j znv*xG{z_JIGbb(_qT6XtBSvF!9AH4Nq@8Hu_OH* z97elgF?hikeeTmI{^0j!SRDG9hn#c_$I9u#gCAp&^?FDtdnt0CiY#?Au(c3VEeq9Ft^+_4#Jg*)vzW-P!gm=(D}dby{H)M;35J_n#J{#1W7fX0lecaAJ?zQ9{_*dN zun4tK(CJN(l*~72j2bm=vt^}$%O)2YL*9ST5*9$woN$bu;gBa{=i?l5 z#n^nc{T>d}Ipn$~|KGo9D6yskcG-xo4T4{+Vrmr}S#(a)521Y>lu}fK4mzgmS_K(l zx0PFAS#v^BAzr_rOSmhvbgU}x>_qO6`N9Q}OTF=zCWUnY6_+Xn7hVF{D7P{ZbtcTO z2q~POMQ>0h%ai33!}Slw;79&mB=la5)`T z18>jJTT9Av!H5e-SAb7?NHA|VrQAh zPTDK|NOn(FMId&fxz9ID*K1`vs1Ww#r zawPh}Bwm!L36i{Ry!V>J+STtN%lCU}fsdr`t%ok!RC>GWF0gxlH1jf~e?OYp$JEJ} zQwO2I?wZ#@2?Xu=P@~Eb6Tn>76s@)w!DpBLa(!J@%E6@p6HzChep2xs8$44%D- zJUupe{w9*7G~EGijT=px3i_j&_?sp4L0Dtt^O{3WL*>dxixa;31OlTRvB8mZ*ykVw zIQcrR`!fIhm!Gx$eNh2p!6?Y)me9*Yxso{ko-nAQs+S}2Q01f)FRji!7OF;1kZmxX zY40ga%jE+E9e#BDkCkWXLnXx{ll}5U?uUT+ijOVmLoiaG=&;rrTr1`C( zkZu2(*AB1Kq(O|omMh6vm`Oj7=ZZ>I`QjzW;FeAA7q<$wyOmAPHrDIm#Y^D-o>)%f z5IYz?qcNivkmF=N$E6gVGgxa=y6lR~h7KjtL?TLxy^);^T8;-L2O$qC3M?T{qDCf!Brl*?sn0=mfYuvPp@Sni*gx^c=errnNKCEL&bc}HFeyih`! zO!2Z;dK$pEU^`Ti*NKJL%ThQ!jzEYPdo627RZzNzHjmyM(h8X=NMV)*l>kk3QTQ=M zoji^pS))@r!G|DcR6rSWA4h-XW>_wF(zmU{J~rYdI(`ddAh$`z!aUiA@ErLek9zTq zscpWyX;cd=m3Kfr7+^g3VN>Y95#h-gQ%dQW7Y$FsqVXv=tC-rakcgt0!LOtyOIC%NMVixLkQktk-OnbSbhRioIWo z&EvT5K}3z z6{M4`5~PCwaY<2?S`UpWC+HNWS%yR-t3{RMOZk3SXI~3 z6+nb>A_7U-qC_bS?!|NUnw9c>?zd-|&d#C>dAmH~Lz0wjEZz|-_Hf@pW?RHAk4nEN z(W9C6TQE*ov}Kp<-)}dH`zn{M4mP9f)^|DzNsbGL<7zGNQbH+bD6!0YM@2S zU(~k%_vd(Bgn6)wKJwO8>ECX!dy#j9+@5X!aKGUBI~oly_&K`|bFxfSc~trDCz`RN zF}$xO%Ul>csTSCYrx|{mmC-*33j)3Ts@q?$8#&(s|E&KK1fWV1w?2%;9>^{27_kI>E}629__x1au7l zh60&CP-YvC9UyuI1bptWJy+|E+Gvi!4)A`(b~p-Dz5g;4{L>T z@j7uY7^r~>*IAgdj0{PMEZhbOMr(~)LC7JB9Dbd`7=gB%iT2K>GyPAhwb9=D-Lh%m zBu%Az6t!-b6wM)zq){Ri;OP{Ws&XZGF5bC~S4*4z_$QYq!jC?XxgkJ1Hetvh7r4raIFL8@X2sZwsJCVyAz} z{A~LClpH!0+L^K8GYqx5)jZVm!;T(j zzg5Hi^>#hX;VyLYK2AP^>)PIn>EHM_f3v~J8~9-{(Qz{vF1+-vurL_=DCHgsMAjp* zSsxk_R5d{QnkL*O+w^9e;x3&JiHii@J!Usl;}u6-;BN#;+yr6g^fV|*%~0=E+=GIt zI8k@_b-Iu4pU}mxBkTEnG>En`-QmZ5o$_MXh5Q$I7tVMIChduPY|T~niuk`an9Z=T zp&1BVwGHBQrd4p8AL9|rdk)K-GoHHfp7ue)X+LlC#xv&qhel?$BI3K^-;+3QR>Xzf z4iIS`;&$jLh{Y1a)BJNP@ZqS`0S6l|1=8 zR2He>fAk|^g8z~<_3hdHiV}&z2U>*%P4kVp=;dUY2zSNEyQGi#cX0V|HRQ(Renld4 zN_mT3Jr~xM*)!KF+IRyKGSqGCHZ}?3Yh~_Ga*g-HTd(sNp;fo1TqH zMN6ZoKBj-d-ov2Nb1pHZRf==<0UOP^@7xu6RWl66! zX(}Nh@z)#dU+*j6%#bPwyWpNjQj=~*dVpSosfDQE{I{=4GZt=PN~n0{7N%O=OEA%J zfp;;?AS?7r#snGsy{7la`LDxs-<9DG8s8sIuLCuqiT7*k^!H1pV`_s!4O_SLnokv< z&n1&pvvkOMz(1iRfRwK zqwi16cB}2X-~Kjf=H?^0?i!MOY2hPvP|8~rF;J21(1^Q134&4d4N#W57rqf{#<~?} zy_02qEHoeB4E%{IdGcYkwvxn5>IDveXiCt9CbB91N!}P7zBBbpZ!NZ~n8c$x<$b_@ zfW0i;OfDcb7w%{9G1;CSUU)K0^C-cI*Fu?b^7Ol)3J^6H?+uMKp}8WUnyv|aWYfnx#$dI!(hu2GbWvw1X;aIu+tPAG^91dr( zUcAo=u4^uI9DLbHVS9!3Ax7W&7;j)M>LwR^#{hSDpRik?EmiLc!(B)>{un6x#wxhS zeY~v(#z~vD?anZzIOD%q_Py)>V75gM-uc=uNX;u_i$K+9$a<}TQl6#ACsbs;9NU)h zjW&fc%iWPMKo2Rb54Fa=h46df7~|3E!Z%0YdI8pMbWPA{V3LLAOw2-Hr4P#Q7*FpL zVsdFhJAtQ^sk=y;@bI)8DINtU2nd?AZ7dXCk zBQ83gwJXno33mv-*m~GL<2iqb)uG8MZJ)JBDwnvB3x^VINUG|1_XIk(?7(i4q5jhQ z_AHa23vitP1=4EIIR0`sgAeY95CNRy7{|xzp%1gk86U&k!x_8c*Dn5pbpg7|*8kb4 zjKAU4=W#8x3Tr7Ooe(nCNsKS`Hr61Ep0P5xlkT3=C)BQetsI)5z?fhTs>NM^WE(&?O55R?D{>N-Ip^^o=H++B_;Z`07O$iK_CMw+>Q}Y@yolW8HbuFx zHQ8tp6j(N1*-w#&RAltT)4XW6RTI~c8|-5SDzmA4fSXT_?8A5!s(9a<)FEXA3zO; z6%$X9$IQyml@q(e&rjL+nl3bky{TC}5y?GzWhYd)Hq53mh`HVqukMl$$m`@t_0a>7 zglykqLD}jCco`0d?S&#n>;-vxE>eG>cV;@bRxx<)%77RsJUbS2nc%s2SC^&r`%XdD zpRFsFU#dmFjVk7y>J3W025a0~>15f*qC7_LT@uj7_J-F=%H2x>t~#w@cX+$P!D0=6 z11IgZZLt>8TPsGBHg(~}yp6Q!hf8nG-R!wsl%lS4YvaMJ7Zq#(2RN192mWg%*}`8b zPXVINtk9aEM&32hI&7Aqc^a7Tr==_YCFZT-x3CrrS(c8;u7%zeR{CMyWVjXeXvdNR zn1-47hU^;cA7XWV%+aGeX!C^D^F9AlWI4A9&4qJhI*T>eCQ7-HBI~KhU7@`sdwQGl z^EoXc8IV=cVgMS|!N$rHemVkng5v1~WArWTw~K%Ovh#`9H@f+md+l-Pd#maJbh|m>72=HT2f!7)-f<1d>RdjqrHSdNZS>_yQ z*g5<|UjApkIt386Cz#=)Hs4+K?VA5GTaZ6(T=P1)#m$1auuHtk!X@sdl=mspMMbWe zm@A3#>|xWP-L5X=SkUzly=IxH+WRhj3nIGd($WBfVr#@Hay=xLw_I8D8VY%!fO0MH z0M-dWLOCKTxKo@n?G&7ib563L+`S|CxNjPey_vQ#%h56SGViqN3vs%KNtN|jP#%rJ zp9Vl>imDjm!#UD{ur~IDDwpX|nFN~C)Q5qby4)uTXpoQ}r-zr|Z+U0*aj!{l&&D&i z&Nva+N+&@1YCp8=WPvTjjM`m(fg($7fp`Z1^%>-jbjMJ{clQ6}+N(&S{-O3ynNU}N z#f?S4t_4f>6LbYkR*6uUzJh;3)&Pl%?UOgm0yZS`K-F*!6~>~?1f;Q&O21uU z81$=rZ3E;AFaeRGisk7-O##9TwTY( zrDAy-l)6xD;q*_SZatB06~s>|59?RNM?kIvP>t+Mm|xp7`3QsRnyWptY0%0GFZa=6 zfDhl-uC{C&!)dchldMbMT{d!LqnLenV1AfU zzXQcFaiv^2Zz;A$w9+l&P7iGt=)Yt$n?0B6AqzEFF{G{bZ4bL8KkjB>BZj`04*5O` zqvMd@w{;WSfEaFeZ2$U}#+N>_+9yXT zhgT_tZcCdO*d41qboK#=%=aeV3`MgFqOje^dxM}m(>_5P5n24)- z09kt$Zw@fy=MQ_2ZY1f$2@bh%IH%eI>_wCkT$()~Kmyzs=(Y?hvky8aiIdhd`QoSi z#k`tD6`+_2bw4vv!?!GG4HV<%x);FzaAlRUpfGsrn?=D`J&g_#=G7fuz2Pv3+mBwm z&a@5jMKaYMwHY-hBmcUY6iy^{76Va3DJv;bMnz^UygDl*v|Jb^st^}PY|ykRj{Ej8 zP0D4OdZ3BGeaA&t!f=-}odi2jd$@aJm z<7&EY9Oa?koYvRPAS#*pzpKe^7xr>Kwm^3&r3BvM0}wU#xNKYkH-qn6InUG$en+{^ zqeW^e)d$-Gr8cP$!D;Z>yWoy;V8S_O?*jCU%fyF*i@g(ImPAsOW?2g~tPf1URrIIA zCg8*__lbfEK3s`ElAR%rnM+Nmaxko)3so)$Il+4TxBc&qx2|#KQo!TFQ<{zHXRHt% z(l(5qvI%ru@?=@LJJx!bSu+ODL<$7VdQ|{D;`IeUIDlVjXF(%+yCM#x?=bcVGzWg2 zkl)fGOZxec6jd`@2Tu9G1XCZf1AymXygf0dT(C9Vx!?HnhabNv5_$EdHpU6 zYD0~@dO5Q4W666?z%F{D|1LN5?sG*ka)-Mx!#&@@eVoRB@S-vJv~6$xcE3i^U(G8lb*li>3_=!NMx@DeHwg)KIfy?w6PVcyFvHm<*NP(r`}vYy97Qr&;HD~ z;x2bjWXe?~&RlLk8Bk7e7<}2ETZItpO3dSrF zr4@{c@pAW0ad)_ha2LsTA4~gyOs7p%>>bCeqaX8Q`A&@5SdKqeELmRdysdA2?am!* zl6)@ca9yXJ?6*Ki7Ntz5h>nVc7CZ?SoI{6_6-cPOeI`L4EhJ z_wCtD^DjVRsDM3AFbUa7r;$g}3v{xxYq+=qWzgcfRYHUHJD!PcSeimXEzK zH@wc#2Q?YKMd8VE?U)&iA){#*B!-+nR$tG>o%6R=hId)#!-m#7<_fVJUR$i}n5#`t z)Ow~s+iia|lRm506~~?V!+%?Z4vLBb*$L@y9RhIQ6o_VlY)~$_Z&p1EZ&4 zJ~>5oS)36z);7RySUfv&PTj)|7N1T1sL#_JMEidAjtgY-D^q4wVBu+`Q%b1Z+yM-s zq7}T?Jp`@d1JYj7$zBt*%BmFAyseT}fk9ColEZu?dmM00 zaKZVCF?M6>sedshhUdLICobEm+-{(;lNIwF*z^q2kJxaCRgjHivO2VQwd{SI78i}d z2Dew#WpjV;?E3LxYlSA4^)zfqN9BbV@%AdxfJ^*=XE(DVxKweBzn+PnUd==L)HAcL z$nJ_;$=(H9rfl-vyC56-TuUY&8T;}R{BOu(I0ZW?^^nAg_Qtf%(E#oJ z+0owHr|zE8t|*z`B~1i6h+=Q7%Y}G4_D^3XxW-KdVh3Y*9os6=sv2a+f|8~UOjzZA zo*;D;-puGO?{~ur3~WRzn2P@ox50}O{BJS^Q*$Mcu;)9K&R0ZxaZQxp0E;a|>_f8l`NZ$R#Q=hj)9{15%}kp4kbyw+~1? z(1GmtHBp3M%vIY0HQnpUq44U3u|V%tEsN)uyT|j}cm?!z9?B*om&eVpdkTY3B8#^{ z+GQfdrisPcyXgWNZ=gLEgk;00ZJ44;^J^CruqV~kGT2bwTQLZlD$xQhfLJVMbR^o zf-q(H{GosD^XGh!6OcIHedHJX@zX#0F=l8?m_A`SiJnLx);XlsF_BVkqR2)n5=ekl zK=XZ`tO_xJwNDMk^s2TMx(ZEM{jq$U28hl+D`=a_SG~#IQP` z?yGadvpeZERI7U{5L)R~Z7j6O^0LgWBhIiKoa}?`Ib{TZef}k!;5?4{^|9YtTN9T> zjSVxG{vDJ}-Y-rft-wcCI5U%8Ib}7tyiwHBGw(XJR(rUSupbx0UG2Pm+;H(A?_c-- z&OFP!wddFr@+mh5#D#nF*I2Bt`Y7cCirk|jW67;Bbb1rn<)XD36j1@rs@6HsqQ=Gm zV;=UPX+IHXkea}H`6ihO&}+Uo&D1=A4R9$cT<+kJ{l0L)=Ecd-XBHn0UorP^ zT9qSRPHr;5y@p~gSlOE^$&tpeZ6sInAR?WCAI%`A;5UFUbEdM)i3TBR?II zE4kzLxzA&tyFrhATGbu%Q&g>x(uIt&avs=s>`^#p~ay_|dhW9R8~Lgp#ktf8*ie zythC2K(Bcac!W9ny+--ze{0gfmf}D_f!E+K>NOSO2js&K-v6`qFCPH8Csbmk3IERr zfBK-`D@}NXZVtyX_Q8`Y$%e`QwVnE}54QWK346&g;3G*~WyjqNGMre((mcWt*bj_njF?k*a015opdbC#S86syiodb}@+@$WLMgXVL?jJ)*@arx0}(c zlBVUluc2Nzg^dO=PqD-O+l!2rZL>2>-@v#pe(>?d8D@MO>C#KXZ2OAf@+i{QcOxNR&2Zuk75%8VP$ zsh^dRw21^7tcSRd1(Xt6fp$}oU2HaQ6KPZI@mwMLBCOmmo7XdYcLXqpeZ{DkQv#)O z-At4GtoNEAUC5s0U}@CD`c zFEyU*m|@`lyDz%J_Am_dO1rLK$-MU03Dz9V{Y|q?`BTo} z+vLPVa?`@5=qY6rMH+y`0SZ9dSThkOwhdorcdA!3Uf`@M= z`*(*U1s#}uzZP+>2v0zc-7m!J*wvoe^Rn}@ijdMk@GHW4g|R${9jg+d-`GSxmkYP3 zqC0}?6eq~GP_0q0U4DjivsVQ`+Zntqv`%rCuWe?}OnJya-8kePyBYL9QdGUN<3juM zkDNz~yW?Pag`ERyd$fizJ+7N1fB)v%H4DwqYW(5KOj7d71Z^%@ydfV`%HtF{LPd5@ zE}VxB(79=)!ZtdKUZ&myeUi~2jJIOe2?2hV3bSbZ>;a7?{M37Y{3c^cIp2~4 zO=^5N=A3H+O*7_E=>{d{!JB2!8KYX~wn=UOEI6X4KAN!~(GS{r5vQqhoOk!epH4L+ z<=eY|^`B(B3nK*x{D!Rb@+c+r=x0%pcfb)&QB{PUlbrN99USXjHWgfA8uvm=Q~|8> z8Z{f)PWCR{qVwS1j|5v7XTinB zh}w@L8#i#$?wrp2f3hR1$m&8-fd;f+la!C7T3n5r3g9xmaIb>_@SI`e zh>hpm$+3rC5p3ih`!lOCyCsIusmY}DCCd+7v~X2UQOXk(Ifjy^IXoR7nwQi_6rAa$ z?IjzOhfu=+qtERibP7O*lv1(>>h5HEVXGVh-bhTGLE1q#pp)Lu+Y_cWkTq@y|HM#~ zEl2|0ma8Ex?lls;Lc7bOfyUN2OjjAz3}YqeD=!zV7IlDzffiF^Sg1K7PI8XWf7Tk> zwTJs_VH^6_PagHP7GZJOqADA*Dwh`Cg=*h^#UZ!k$yY<}$o2>HF%TE5QeZ`HWzZpr zETVm^34{cv=_sIL+5pHnr6`e-9`lwc7pTQ_=uxtLayAqbn@&W%iHqV&zYR)lw<2ED z0AZUL+q|L9_6qrhy)7GKzUgsgqZ4B0LVLzYY_w^Ag5irl;C28%5A+BiP; z+U81-#1$+{7TBC_=7eI|WN5t1qfe1KpRt;hIRl2fi5bxqwgL0e#)1axwo8|_OEwxQ zkqm#Adn^wsBDO~0)b%km4QS)snj!O@E7}v5$m&&GV+i@4##Wco+`YxZBs2_NJC=#z2AnzfhH#N@Ia7sLjiuaSiZ33`-rAO6c;k=f&Tzi86`lJ(s3M6R1nCAk*9$aYGZ zgjFYzMr1_IHLb1CDS>DZu#;e%H^Kj`u>oQ4f*3(H59DqC?!A#%`isqT&G7i;xp{j? z0XKNKZaj&cwSY%8r36{zQYsQPm9(Zd$`qp#6ShqzNJ{1>ve#hQ4BiUJoWuGExgRp= zZEV*xXrX+)Fxa??M*XH%K_8v5uo^gii^8gXO9S_i)2b?1Qo*u3Z9!EC*6JOAWqhn; z@U&b}8o3radUM3d#0*)ayzuKN?azOIa8&cI8ACzQvwlF1zhsM=HVdSjrx6zQ1_HJyqT(@mBtkLjYzrdRqE&Wo9Z z6iO+oT^<#I;5@mnnW5e2eu}J=*MrA|N*G)gQ%-}=aQNZ?DchdT8!@YPS+AB~5wFI^ zIy#!aR}setv1L=E97uwi`36lP?@J4-=rW+nDpQYLaGNu1I1K%0PW@^8UqZ&4tNW3$7;9&MzQ>nRdLMH)l5-HPq9eeT-!u;XrH%}F#kT6Q5~@Z>T4 z^!oRo2q-fm9(z1KL;5EW=p7tVBM?g|*HL5*RFgp2eToW~56JkGBV7vr$HkU9=wD^m zlXPll&+z{cPA`({x`2LDq+fs8qb8WtK@p=W@KV9UL~Y?zx{p~q5C6L-3`>fUv8a{# z(z^}hE5``L_5pXOV=z3>9KDwtaASU&|M)Gl!Fa@^{g_k?=eWA?L~+@|x}Bkvrzmm) z+3dFRmWj}lFjc6h@D2nvs;)%Dxs?HtT`DAWnq?qf9%fQaH^|C@q879(9)j>VI-poK zm&l?rc_RCeWC8Bbr8Vt6FGH17yn3ZP#RKqlRo@xXymp$cF>#NV{nTTA+ou0#f3wxe8~9-{(Q%9Ix^53Afl%&{W!*kX38XQ(ROAk?tI|fziJx4R zHUS^Z4OS2T-6_lQEaY|1(Mfywb%LYJq2Nc-9%vS+<|jy2PfV883D&!q&7$n&k1U3xR2pr{5Zl zfI`NqPH&wuO7xgtEWWR(hc+f7Pp7nTo?e6Wb=iR{L`B~B<&XJyrRBmZ;Or% zUJsX{hM--u3?&30VVKBbtqN8m+?>}5M2K1}>n?&7(b&Pn4rn~J{bNWG%Ly9ezti`@ z>(+r5E>p|P7lMhNu{k6I3Zk{ObSf|jRe}0WH^}N75++Dc^M1(^2a^Fk-vn?%(erQR z(4+rz@he{)O|_E?w%2wjtu`>FE0v*ayNCkI~J}$S#7Lj=((*PZETzGqc?!t?w0&EtX@pF z$QK?6C=D!Nv>W_a1Vi14x{E{wSBPtZfNnlTl^#(9_v@LRq{c>$b?ia8HWqx<)j++B zm)8U}135in4zF9XWa#gik_&jY>D&sqT`ccl;70#5p#NpdTyaOv(Aml%--pE$x9RCh z@jphq?1T#sAx84v!yo1RQ^%shuYYjrmj}OjYteg+npXD{yb@}UXNmNltVxriy055H zlm)E16}gf=J+XI5*-!hvcI%(=-+9kzp3P{HGVCxrU?(S}P%VF1?Ptw= z=(39*8!U(Df1nM>k)Gs315hV)pc!bxbX)`#g||Tfx6e}>GpU#04oqijyX2kpBhNiy zJN(l_wf7Xo-Wd6<5Zn`{_-_aWg(%Yw6rRAHI=2cB2cl@SxUsY`4gie#7MyT1?uXJx ze;rNrlMBaIY*at(2}@Jg3G){%T~7UeMR5KiZO^Y`+#QGudq>hf{MdUp$DQ}x8;|O& zHLqRPlGsqOE(L+bbsl*XT8d04agF1bFG`j4E3j#(lg^_IAob8JgB5kRqR+FIr^RrK zX;0pZsd7j8sO=kfOY30V8s+c4=yFbWWBlts@AzYw*#j24ElMZ*xCQuKxESV)g}+=y zDa$Ex80AhnY1HuS6JC+FLd$#+T^^DFsbwtu!!+bzIe$B?ZJJ*-9d*8q9VAWj^TRMM zTZP@vAUV3mU3++XHf$IHXsX9WlITQsx7!tIzhb@n*`Pjly}OC+*#S$f5d&@*3=ZA( zB0#<9_znHBKUim=xx`*v7(zDk)mYfjsLG@(p+X9@-%XOnSgN2c630gjgaKQj7k(Cn z-CX4S^3IL zMmsGgqXbGBPmx$Eau0Nj?UovQJfMta>Porx6xkEj;knLZ45M>{gJeG}2Co>K&-_(4 zZdfnh9_Cru^Q;G*vtcfG+AiDAdYDQw*L~H( z&e!FJn`e;kAKa}YMJ^oBueX>Jj!?=fij)I)4g|k0i!;JZbWTt?JJsL;)5ths8i-O^XaG(R?JGZ)3}S68Sqc0Q3}fHE%tWE6Ei>@$IoG$ELuEvHr1; z^I5tS*}u{~3V1iXnq`+3YVoFt?CHfjzJdo)gb!(%Y{lbW06;GA9)vm9&lmg&!ye`a zZ^_L)q1Jwg%Z5&Dcot~)qNZ$PwYmOnpwf%$j+^s<*o_`DCY2!rVK*#>Tr(D5xxwQ6 zhKv6>(`;Nm{8L*Q*~QI0b>SE*u=5Y8pFKn=fpxo(irnVkMXw3USXd*i17V3B{34)P z*TJF%dey)CK{Axp7D{@;x5xlkMiH%6!3|yxb5C|fdVrL>w<$hj(j>=ytG#o;?wklb z>TM8QfFzQ^r6t#-{_z9XC&}b>_F(DgXC^HM_RUx+^y1 z*mV37dKoN(lV!a!gP`6M^E*KC-N*xt#qg4dqrl*g;j;|VB~ONaAMk3Crl*&b%tvxj zQ{E??$qTQXw>?z5-h1<;PH|0OOUT{u5jwKA$BR1(22F(Zr|sVu$r0PuW0)1(xGl#+ z-t#kCl$GyKSx-`3c%cM|n;|x0FQo)2;2dO2PGCM2#=ryw?a5H?mMPlk58TNr?Jb~D zM&8@Ua|(Frp{8!bQEGN zZt8#yPKX&FFi!NgHKpK76@{}w95~n-4<-fO5zOe^b=TrLxY`qCg1XslV_H_g2`vsl zfxWk|jh2WXf1_)H*_Pye`JaCxQ4>ju#VZ>}DK}7LEfr~sJ6u|b9%80fcTl&qaLSk# zDi8i^?Rv_CCy(K$?Wa796?a{?E-JnDtJX{!uSibV2FFG>y8*n^9M4`^3Gihk1!an= zd8bwP=>|GJ;w*D5z!+yUk!I+5P4f@Rpf(OSD|vMwVBbYwTG%un+}RZM>d^J>DJtaY zj}k?RK!XwlzO@4oWg zJzvWdeF_EMD}o!uNkLpEGrQgkYe!}z+Z^AiKW8OBnr^l+wQp@IAsH_0uhdu=zx|XF z%A4{bu+i5TP|*nFmEs*)z@8?ee(HYxSsSZp^8rlax792h^B69Tmp{;pu%x`;C)< z9CtI;|Fks~B^Nv7!XY~wB$c=3R){zH-&ULve<^CzELWZpAD6W)Xwt+=D#T6799}di zEY#BRl83;-QKhI5w=2GQV>#(#09T-kc1_wUSj|sk@Z4Tvkj3%iW~`k7WR&u=fqD%F zbaEhNiCaj@eHPd{t_2)i(B+$@)@vGhcLL5Y-4S|C1KmqljpZ*U~y(o!jcr@`dH@Rf=7qM;PRW+C05;PHYHLcIY+n;TL##rm3~Lmt^xA zH7EyB&0o*o5>NnyxH@SeuS&6ga+78&=!iU|@#<1_O<*VV$KR9X38O?6VvdLzW*nUa zqW_P*H-T#+%l5}Dv0`#DWFwGN1Byg2h%Jj@MD5titUdemo8P>7v-E#vWR{+pH*e-m z#`o)2?|6I+;Br>6%p1Baqs4ZWrv}4zjfW$jJ62EB_K9XS z?1($FTC~P`Zu*P#%Vq6F`<+&isltdDF!;%DdFA=GN-#g#xC2iSbYHzCitnh?v zkALcft0YyvSJlC}uWSU$Ff40T#=#;~x$Y~dH50D6W5UpO@9rr`Y`8n<6j=JY`T204 z6S2b*EEal*ty9mw-|FMGJ$iQTQ zt>>Jwfz;0`E2_U-BnvGt;r)CzzXx~@SxqXN2W_RV&A6~~ zzq6Xd&uJKrX+`?W@!l%3*oAE~8D?wD8j4AxNIY_>L-vMIm;FM0g>?j!+f|NCHnnLA zTy>t=X|WSt*(RN3LWNHL%UrV8g;4?C@5)UY+iJh8X;7_*HZ*T`F5TtzLbSRPZXf$4C$Ga`$kHc=hUa z5Q1DB1=I}kVo?h$Pv@D{A@%Ae{?3pNC@~Yq?BrzA%XlTc$De=(n32jCn{o7@XJO0mICX+O|U43bf*a&ygOduVn7>Lk3LPdA_rOZNSz!q-NK1!~=`H%B&_Q-ZBu~x5jp9u(_>mie6zy+>aSUK(5{y*&4fs**D> z6N|TwKr*p_I11Mc<>$n)O=>>8br zkqqUiGkzh7gw!^~gHm9GkgGIO7+@2Rb{_^6y1~C@x zu2-+0c88M?dTVBKU=zPacwbTy+$qGU|0CQU?$UAP%2GO)yNG0QW5C&mzZcIb4vq6& z28`Lo+y~;l?w^u9vI^fWpDOwe&lzyC9W&2ZES!KKR+yot_|W?TAMcrWR3wruFOBL_ zX0|vLP|Qw>-^9dm#v7qdgySx-m&YO|R95q}n&0jvv{pN)Ix z=avRIHc>7vY~{CNuG=Pn=%_|;JZGxVLwL?7K0HBGD{YYjlcX~iooCIH&3Jj%AWpdU z%Rhej3rhkpD?D4cae!oaFrM6@JEg7kBR5R5h=o`oLuIs*GUkz*6Y)MTD$)7{4p$3WpZiNggdd0Xb z75ZNFkY3&mR)lV!t#noR$FFJ?t34s^14_wZ5LjkJ8uz+9lVFZTsiXzmXvQ_Wly}l8 z`rwmEu@@--oF{`g;fNJdT-PR5HetHPG6BS8)6A?SdsO?DOfDAYMfP*gjMoL$D_|8t z%KhBHDn$?Y_K=-_RF@AnkaYqwtq#cwvSZ)=G}+yfj)jf>;kp)P2tOGr3ezZN4Mmcu z=t15$9eqlH`srd`l?SFjYqC^GREd0mqw8iE&38r0c$M3IwNMc>ss0zVG zLsq@@6a&I-4Nf)k`?ZH*gb~hij~ojXYCaYwaNes6O`C%y43X)0^z=?PvyDoT)wkTi z7g!06F<(D6#zSPnMg~8vktB>Gkk>h63bYinfgg%%iSI_9~Zz9i1JEj%o=FGC&eUZ@D`ksA3h33OpY9^9{33V!CT!k#+oy0) zPZ`-i7Vvr2DXS@TU5Ond`^BF_P1uS}?cPc9U3jZ=(hT%P6a)23_fye3y?VIu4A#@W z;J5P?O}}T0qLCP;rOIQydV$#V-Zitvh&)4w?j*v|s3 zE!S89dra`kP5&5Y!jHQ3uNt!O1!IWQnPDoKViG8_ii$=i)4>V}S0p=ULlru$0`XNc zi;d{Az%B+$LlpqK0Tg{&7I1iG%am3CL)Z9v1|kkYdMH{Hxlk&h3u+V&3XE#n$D`PiYm=+7vomm)nFXtMVk4dTTid*5mh z#|Wx8M`hrqf`8LI9`bUz1+w2Psfb1H-|mLA9JDqN5K7l}vsJdvWA9ZC9UCB=J^4s(iir$(Y7# zTj|TekK}hF6W#086=aX+T~PauLE+Y+-=|oa&a{IY4cumf-yOfv<=}GOR{D{=ENYv$ zMcyIZC95Vw<)>4T$`;>4klho|c*Ald_H?+4>-t8mS^wW3aKruC$FF`c;I@9sKv+}Y zKL_0IM&9-4_xNzY%|Ii0*wFbI@<^O<$ZqHoZd};3XGOI> zOY3=Pw7yL56}QMLIk46x@ml10+{AzqpEZH~kW-2h1{2P@qo7)A9p{RC-RxFR4bsyY zZz>_%VSoX*Vs3W0j&4+$ipqeYP^XOdH1^Rd8s)iycG+gf;Ex@$Y?BnjV})JomE`7| zmJA72h-Ax^wNlVOAPKw;LR=FAmVmoJQ{tTv{6T}b-vg8mqzX0Aiv$IvT{TKCE&IpH z?(x{$+xiPToMu5>)uu-OBSNP%sXWtkA0?CQ;kHsPtk_hTsnYu>rhp4G< zM^LZMRNavqCZao_UcCz{)*a;)@v>%AlVfy^Gywp z$d^kL%z`9{v$OX|){$Pwp9wBkLd- zM$}P#8@~NZKa)ghog3CbHn{Lc9GLlsIMKIJ%vOqIQPHTeV)D;k<)9T0_12xRf@~x1m%K zkflJ!7I`t9z$=+mAOw2$9yx}_=xAMtlVr}w(c&CQ!wM~9W4nKvqcO>rlTnj-q}zpM zOOlz%Wzkq>fFci&p$P+XQswJ7g&;r-v|d!JSUF3pDB)H6rU131K0GC$U3FJ-lDW)< z!l^*?hgZP2KsS<`Gg7$vaN`J2CI%`z9O1Tbs=!c;5$C0{1P+dc1`&+(FQ$96zlR zd`Eseau8Y-Wh1)43WQ^RviUbvlT08~#mAPAP3)|ME^Na&WVRgUQ%o*J3{-TDtU!oi z>?_F$dX3*o!60wbAf|wHa*4DSHro9jsq#u$jcjne%=gB~90T1a&z(BphRY+ust#JR zp$;%>UQ-(-j|0$U<0rA}CJ)d0mZi7KiUSB)lg`pDoB=nqf4~DIr#(kGJl6EL+4L~k zJqL&MX7#GNuIX!zN0Ewek|D2OTziD99B$j~y5FCGo;*ZxPN$f)6iK0?js8juG6nxq zF^$o`ccEtXbX%dO#m?y9m5Xe@WxDD)PGnCji;9y7uG|ph=vT$TIpH3Lz?0F$ZeO;89Lp)Lz*(% z@yo>XC&Z?ifBrBI?_qr*+5h*ihuN45kD(h2sb+bt_cGot8qEbe=A4Sc-$x_*Ali2k zQzbbkEONggIR}M%)(6+{m+>A3Zd5#QwEtw7$79ojH+$sw=@K7423ae-c;n`GeD@cT%Eyop~td48RGFWIcF;kWY7 zs|)B`@|0KiOy9z}4zB1v5)UZqq}ygCLeO_faC_JdNgB|I?e=Mn*aXQ7t$uO7kFTAj zwbLtubznO)(04tyK>eVF-um!T=$kHC2E|`7y`e@H%PPgeVv<>5>1nC?oED1@V{tn9 z&#DBO$xx!3_@}>+R2R1RL+HT}qevFTKxUc_gDi6($Wp6F4>EA?lcNMfF7fOz|32v> z_3CqUzegcu5M_uiGcB?v{!+NUe|jr0;veN!INb?t_sp}HMSHrkTPHvK@YDs%<;aR0 zZ}iD~^}>{7?-RZ^mFZG6Dx}JHfE(o)yk%Gt7oBvYswuLW0ehaLz4x%20hiTOzL7U@ z&eCbfMpJR&O|TUgV-x>_X8FJF$sWn}DboY=%-;yMX z(>(Ty3!;+e)p45mk9#Na)~oaoyOkz5#E`3mySW)W+z zthDpK*uL&pEe&)o8%1ZuF5q}-|qrh2$dh&Ra@yM$V1x+ zBKbrK9AbaOckW}W5nG~<;E$R(OAr$ObH8EJh^EUj)ebSvw5;&IK`}q%oYbH_pb6qyM-cz#Y~a}|=SBtz+-tW=Ry*XoPaHUeKz#h(6;VkFO zE{s?X;)L_1?@Q~yT$uxy>Vg?jm@judu+|q+oe~4OeXdCEMrsPhx8!NC+iH>BlP(r4 z=!?W%KPC=bfwX*0p{Pcv_1sLhdgj8tpF^D!VA=|G$X5NSnY9ZqmJ8NXqmI1cy;nEC z8*W;jWdB#QiR@svJh^bD$!W9Ysf1z2iJ#^M>$caHFQ)(qJUi)Rhuglj8$+jN$&a9Ov5 zm5J_<=75Rqj5t-kAs`EATu_HSq-qpy<2Z=Wh0Sxfb)wmf>4?7UwnJYhpZxtT(|VIX zg>#3LzBKF2b+h%RhGJj=IzdI(N{xwjNxVg5p+J){8KQ4^4S|q*$IYM@3Npp5L{mN+ z%9w-aIYv-8drx4xBy*Yu9S!?@t9`MjrXjG0TS5-NMk;l};10(?=L)Wg`zK_oibInL zG}c4P_c7omDwzxf<;9@~VV#2eRsl2Gb?{~vxz|d2Nj-zfx>`j`ME+Flu)>{{rfe$! zhg)OVwy^=`;kV`=_@N0>daD21CSr^qn|;uNqK8dx7I(j-l+NLN(CqX6UG6~@M$Mr`E}b$<{s=YUH(t`gBP8`XfNrJ=RoLUs=PqmEp7rnhgKSCU$OSwQHVCR zK|tC3bZzKpJ1}e*%z<540b|TmDr&DKDH&pdJv)V_7;z;4SZWS8AnSqYEc!?{S% z45Qf{azb)NvUKW_5U|>1kkz~<&}#?X-d7|CM=^qrz`zhu=Meb73mw=u-RbEBXC9&W*xjUnf(bVz_{Gq&T1Xdi_KN0E$ z7%gG=lI2Haf~?0NyizE1qodnZ`CdC`X^J*x8Dbj)WdlrX2@XC)^?rTsj z@!Sd3=@<0Kk_A^uuAqtAP1c0s`1syyGF7KTI_aff@8N4clcC4@l5hw2lyEP^4r67p zMWlz1ltl%R%|2Nw{5}hij1~W1F1RJv^g{)vGF3rDnSUdA0LnelnS_S6L&|)z0L{LJ zY2`{xfJxwN_b#SyM7nI)E~`0qU1QlVj&Dd9iQ~bAgTSm%7U71kcV1n@VIjJ+-{0hu zMYr?& zI@(EU6id-I%o6r|F=bDWb-dr7w@d+c*>%iHo$g-JApWRXUa!79R~J<6TNU*wa2xLx zpHQ_cw{n(`yW>^swPIXRR59e@8ECDlRPsoMXWPUjK2DnLCo^I1ygwPiIY+F-9rbS) zSA@UlR^;8L7%i-9JV;xVE9e^E#7`!zKx*c699Ybv&9Jbz&)KJ}U@=C1iTa&o$|jrL zlj|;{#7cta1-}A0=C$OA7qrvJUcb**%h{`}Q|u7Mixw;r=tQ0L0tj071nV;7L7)e+ zC={AdSz07OX7vS5gqqE$8fJMJu>wCifY-_|nS7Cx5bVSWHh;!!r`!B)=bf<{a691+4;foRMhfBLgZ%aZDBRw5VP_E;&veikT0cZvGxhUo?B zgwPVGy3|eY@GMYQ!V%y})+sOZ&j}ngSp;*<3P~fv2D_V=Uf$$EOaf-NoBlLeJ`R}Q zhq`ChQcMa(lBwuspFWSm+4#N=7C}M#XQs+9?P~DG@fs{&w9z|$`iLWlSbTd{k94FF zL~Onnn|I4~ZRBr#|967_Fd;)ASn~sNkzEvo3rmfAW&m%em==mOQqig5`pK1&lv&GW zLP5ab^{=G)9Ti+s-w`hs+>~AOE#>2tBKId(8m4E6`a|+OuwART3bDO|*>v=6s166r z0}H8pvUQRgpw>7gOyKr}=>@I)IEcZ*px{NkmYEnmu_55`AQvQCfLi*=hhe$ydUZOd zGB_r39aIRu3#keTa~k;jxwYeS1CFAL-t-kQ;SQsI9(aBJY(E#f7iRcn%ih*!y+72> zh?ADt-)!C?7v3pZ$rVS_HU@QK=2MHjI=EkMG+-l{I3@;^D~suFQZiYqxE6XRa->?! zv!1=#HTD28yWMFgp`BTOw5*Q71|TjhimX&#LAxzFB4Y$M!nF!aC@6B@=anYW(dT>~ z$&#Zm`KjMylh5Z+gz`psx+EJSAdYc9hKi`Cgkaj`VC{hi@)Dm@Lgy(t?9pN-Je-D> zxZEH9RAO4Vgm;AhNs@*m6?b8wvE6KW(o@VPilk%3jA9xyaE-+xJL%)&A(OJmea(b+ z*z@Fw2O#Jh7$#Oy3l=PhqaJ&uiosdwJO~))&~6|&4Faa`WmQJh6bHsnH$mj}#+*2k z>%!aWDl<47q?r8_*@FZTopb|aqJ0|nP@F5mO)+>Zp@dH4r(rnxslXX=lUxt13BB*B zRm6vbs}wiT$m|96MnyZgsZgk?M%o~TyJ|z@Ju5*|fx|2EGG1#028&~FY0ZEBZo0P#JXc=5_*atR!r<9$2A)ic$)Jdqir)85N$_Gps(jhZ zCdexJjMGmVU(wO^5U`ReKd$=Zm7So7JP7@Kd=hWXglq0S^20MRcy|*N+HQ4z$b-Ew zaSRx&kHQWni{@R+_;NX$>(u``uihru%+$SG|HB08!Mq}O<59;OhKaW$jK^_r)~k~v zYoH!d9tVZi{FM`QVB?#ANmvV{p55eN&~~A*=jt~dBzDYa)LwKsUhJ5Eub&mE6Mke8 zEAmxAZ*6@yBAaOv8@&1WzwC`(lXAGaiApF1RBD-VPWxamZF5`%xgJT)NR8+0bL zXZ1xs_zLd9Vkbbv-&JCm788=_ofcUXaaWkc>*lTC6uD>7Y0oQV=m>-jMH`D_b4Xd7 ziNE^V*W@OItxXQTMAFzH%!Rj6d(04)O)*<2qNk$IOpoJratf(^Q5mAonDwecYPI`t zZFTDN>QpHDcAw;ni$c>xcL#GP4D51GES|Ovr}hKIhQUYBW!_gRx4q~Q4al(oahe|% z)1|@7!;9PvK`jhaWy1n|2O}J&(c1WsSxmHzBRZxVD~)Ds#pEEQiX`sF6+ZO|7J zM8+JB43FE2Qdz100dbML@wf>>A>*P7W^M5-a!>ZwbGE=9Wd&a=St8x(?3a$EFJ_4F zb<8ujJ_@VZa9v@n>H7W&%VH`n8_-~-a7wN8bf{L55^(d?LE28E$>rgwmNhWp#5U1k z9aPS`!V0Wo)?OrI7?bMZ@pbOsS3I!cJze=x6p| zDoCB;EZAo?dMJ{Bybienh)Bom?OJKMQqSr4z!cmzZjXGx4I{{%?M<}}S3{(q*`c#g zHD=7$zUwWYXhO<&xBcc1WTOjPR}PxZeICW+P$Zj**1^7|Q`krJoLmvoKqc|+cy!Q) zDYZ-&1mfNhuN9r84IW9nK2jBx!`T_t?NRY&gSb~%6j~J8>{I3&Be>(c6iC$WMrKTQ z${rY}Z)5;)UPK%#!9v|U7T%*U!9{)Y-NPiqh2aA6P(wt;E{Xw5|28W6#P_ZZ^567iXboY_D>Qxdfo*I_T4$R+9{5okz!_0hB!fXu*1whdgr?dOqls=!G!-IaW0G* zC_Oy{GaD&pJ!*ciy!6iKo`Q5fIw&WL3r5n^l~Xay4!NcvJq>Lhc&JtAdG9+-h?vhM zdw7`N@2oRc@EEf$=YQU?42iPpNgAeDnG5SGRwBDFD6@?26XsEQ6sFi=vLhZ}k(5l( z`ry$h0muXZ7C3SiWQHAg|Lr1<6v_|;hn5bAbS zka{F`Ru{J9Td}huziui}bfimqDwQmK~1ywP#(w zHu>N__s~253T}{7xTWBL&QkTr>mV)xH%lKZ6kzm1t#rHhjYvHQHpzXYjk{2w=Olz8 zd+uJ$J=?`ZE<2A{rdn6Tj$cW@V0m>BzicK`C9T$99b3vydTmcBF!2?PWT zkuRwfvzj7_RP+jh^(ayTS`?7W1R%m7+g+TCoP$9}$9Fo5@51UF+lC@1Yo)_{yTqd3 zUJ;vcA^Yg9I+EzZD@nfDOzS9SBSqFz(JNnRmnC}^2h>Q<3Lku>394O{ac+ef+;?+2 z<@>p9UMT_D+$^7TP8YpLR^f{UApW1O@rVC@_ixd^`qR77|MROq|JR?Q#WdqaC~6p+ z7j-*->t@=7klr5!oFt1}7$JHygsi5RM2f^=f$CbY!|Qnop-6*)wPY^&=q8Q66o41d zHbF4Vo9g5yR?7+X)_2>iLF&ZoKBa(QVIaxlMEGV*QfLfh>k`(S>!(ACx>?l_rQUMl9NLu7xQLPs+ zVL$LN^1(T4jj?(1%}f$QPFO+B_Rg#X$e2H`>zd+gg38aU$DAkWwYJdS$o=a%`N^Ql)a@dN&9 z1@>L9N_xFv-p4+jLC%^HTZQ_NIWRx8{UNjL$!bnqH}33@13yw)f`yIW&V|kHRxsZV z5$%vi;h*OZ26o_)#Lyzl9MwVleOW7g$s6fYa96yLYU4IQB@nIRMtH(28Qj)+Z>ev3)PUPo`hs5)G0^LR-a{3( z%zr1RAhJb1Dps+616G z?a6rz1xqf^*uCS8RJrje29lK@-&e2BLx~1q`o@piCdj7S)W&|R-~2o*R^F!W3p1X7 z`ugMivX9?*{MYH;ETvx#7G3IM)}M+;VvBR#g_>Bg#a;AS9==|Yt!nc~56T4g+4LYC zy-{(8UPO8!Gsk&)+3=FZW>Gf$?M`dGD87_LnS|3r&xdEoBR2vTt|7b81d3Tjkrh<* zQr;l_e5MN5h?IbPqvbxc_}guHk`}j&$|bpp5co_#k{RIt7Ww?$Q+7|84~@Hgn)oHWlF3*pY;<5@8xqVm z?iO?AoKx#a7{*#Z=KL=@5?^#$(Z|FTnLvG}se&lyeZD72#^j~J$4G}%53EW3z*>>5 z%9v!VdZ}3{+u%J|*PQ zNpXy^1EV8Eo|>ZP1LCP$Mx}ujAjbaLOZ{WZh!QLLm5T&rQ#4CB*W5As^NQrYOw;Uh zcjnG%qyPM$|L6zbwo!T9LqCMoqjz29?w3{5k}O?-E-NIgIQ43!2u@6e&|p?3E)beQ z@M6Ho0%vm84+Vtr*haXHI%ZkHb*v`(X9JepNKbEBTzAVwtd!eH1wJ-S43sjck{Jm; zfs7`PlNj)sZxbH_mVtUEyfO4FT}lJHOpAQKcSGQl=1--DY3*TslKpbc8Rc1JjkIl+ zbDYL(Fl1HLcp91`Zn2J`RmFe*iJQshk@wjjJ|HnJY>?HO8Do+tCY~ZIfn9hoQ_mQv zSSrg6{NJdoGB`!HO#g`I(vAVFPyaAmn6dMJJiFF1iolAqz7=v)%BJKfvyQ7yBTrW+Q=uDzxPMWv^_SCVHaM9tR(q8Nt!?pkLHGtjxOLV56>n>9;8%x5^3@8 z=cYsO!=U2bE=wB!Ubi?kJXOBS=bDolct(Ks$G~*T?|}tui=-{t4UsD zy?R3cLMoX_;*^r?h@6PSK?_y&>cZK30?X%J7q`o@xjoP;&2O*Q<+%^U#lP$fs#nKC z5%&bn1@fWh>No1uc+@L=Aim-AxPR045;%6k%()Olu-OMQ>k!MAxkbK=8>SlJ66rPw z(yNgb2y3M}`g1b4&CY?QWWOU5FOpr{CeO7~7eU#KObBDDl^%*ZtJEqmg;dk;i6lzb z+;_UK2rmWbtq*q2cF9)v&l4dt%#nS7c+Oqc@uEDs>cmJ&YPhgqvQkoGnHnlSyO;7$ zDH=nwRr|tWl?q+w87tR%UJz!}9Z=ZzjHJ!umV8-I5%22sLoz!>$8%=nsmts|%~L}- zrQtt*t|_q8iCxw%XGP1!D0@984+0+R-)@A=jfE710N;aJ;OkZU75DwpJd8FNkSg@O z8Rg1r?t6s2k=w+FlxN1*+q-nwJQ?gK-Xt|_{sj9vT9+Ba?6N80f*xB<`t`8CPL*Ru zCho6|?i6PUYRBpwz&I3qSX13#ycv7|BVU{LRj0XHnwl`Y>h-lN$1VBshUhx3j`SGMHyr@uVhvN&p|{bLnqAa*iKX#ab`A>V^7EjQx7Xm zvSga>qhyjj4$K%sgBkZx43wMRNk#AAmeQ9)mZ`6cAGxJXUK~=vdE~Yt_~Tcui;t3( z5C99CuU%4I-4tqV#3E#kgx zAk4uE;fS8N+57F_C(g$5>AKs1-%a_}E8jE$XKc@JKOvPaya+X$A^JSUoTW%L72W57 z9*9`^26;kgt``_%RZt!on^Ht^K!8*0bwm{_&zyE2O4csrAxloW`XrO)Swj~@VnFY4 z(k&h&5UwC42LhXbr)`(4 zT2fr*oobJ1C|l|Y?f&k#*&rTeq@}* zHLQ_b;5U*Yh|xu){I8e`Yy*XT6FZW-PR=iXb0j8Q7xr3PG2tRlXaQXW@u2-4P@Zcr zBQX&Yl#$_U2N%M0A#yu)M!yFZzy;HuqpDIqZ<1^S$QL324xTrnW3GEVa-U|&{%D0Q zyyANe;#=H9?3-1pmAMMNs7%tvy)SOS zGh?a9YTp{Zu^C-m#q=I=-7JWaf_itIz6Q=>!9!l1y5rsAH;=t}Uj2v|l%=Fjeft;t z-q<(qyxQS_aY%rIQ`DBr7rRp|Rg{-xV`D|r%2KVIb&H2wqT2+uKwGR;7| z=fgf1OKCj$aRLyAe|y`(3gN*b3WX2;* z>Oo7-vdac8f#-6_MkR}4Hd91LMep`m=&lp#!v{-@*A2$XXqLM#;Wc?42q}o%!)XSW zq>ip5nZ^!@l(3GODz z@xJE5f0Gq1>>$iG!^t{|Nu|hYD!SA2A30PXlnDGsV!$aRm$yZox9 z+r-`UMTn~1<+E9`(YuF>5jFkvUEdl`n>?N0O_#v651ZxX^X|#kE86+>>L$7jc>4~D zJ4iAAFa+s4l|p}{*s=pQHr=8;_6QPmxsDGxaiQmjw5!?g3>Uwbj-7Z=JEEx6vJLc4@t9;Y^TKNxo=Y6;MZSmdY zSFcVMEeXNQ`!nR2v(qwI0A+|4f`_cb?#_e2)X1{-R4UU+*LA_Tb@G?l@5bQ8VL7ZuuT!tD_l zO@fK;SVGeA%ZmM&F@q2*JlN8i6=udnUj5hk7kxi`LQ>1eIC}__y)E1od`+mS61U2a z5j3-Fas;pxO)zDN2ie54|X6 zPOfO5?^%!kXJs{mmvxx@dY#Yy=jWTd@YdD}LD!Lq1wI=C)-hY@6#`vQMo4DJQmBS; zft*(403ksyr&Y3rn;nuxL&UtW1<1Uv(T#&E3nqZA6~zd8BX7=F#1zP~RqfK`D95ig z#_3@xR?L=}!4tDtcHD_|ygdAMt~}6WQv1t}ibS&IrHKtFGc&*yP|Qw>Xi1*$e3)P?*Iu#HeDdQLsxLIwN$oS3HTxnVgt8Xp2I1o z6F?!}<##r8bHFirLqNS+7nm+d;%<}N2zNZHMvNyT!W;)A(@4Mns_wgAE@Md}R4u+D z=_ZF&sKd6%EB)|Ct9USHr+dNV?ZPXPv%>Yk8Vrp_<@WOAngG;=9S@18=F&clJT;8N z8}9savxlVz_$4u@S@8{T4e#?Y#bDT8|r9HqnI@mNur`V$lgGt0pHKvBTRWigF)oTyV>xL28o{yQ0Hb8 z-mE7N#g-@fcrkx+m-COTuj|8kR$Z3oFY!M<{IB1cfN}J@p9YYY7mRajky$CHyA*SW zBA-&xI@-v9bUEY-lke3d-YnT35g&yHj}{;?XbxLHrNXx?AUpW00p{in_p%Rqp(M-Hot2j>Xdh~EJCf7-egbxF|=+8mf=_QY38^@pZ($P7A8$@>_ z_j7^w3+{*ySR4X{o)oo1#_0FdD$*qdveljsp3WU)^Nu`zSDf+N@~yM}j%YKs_Kzek z)`d6LR^nof!Sz}NW&|!0=o(JIhBgU;yV1gv$Ls^ZW%uU2}`c&#d) zK^t8wy+{C*?-gX{w8t5J_}lTa*g;^j#KMTr-TVp?zFbPp`Pml-nRescHAwl7=|-bK zrtP4^D^D%6RF$@-0rEG4yRIAAMY7aXp%ok%g?4fzh(gaSY zFhB5;v%Ne+X3M7eezJZHxzo9a)|ZoER4%IRD{~8dOqxyIulBu1*1K@T(ta~tJcnW+ zx@`*;eZg-fi4hbFOJ#B1_3C>bYbNBoH_Hs1E#gCg+vrYVfwDl^HtUgF|D0pu+`wXC zCz@608a`4(rjRoKo;f{p+GaU^jd(K8EV1$gHF)%77>6IR!%D-Mcb-y5T-J+hMZt)f z*`i39mB7h&@A6y3tKmakbpw#NBdJ`MUuT%+)35@SR?#7?3GI-sjWW=^z)w*ZdL{HE zS>xSBYWNA9E+4G|>n;p#^f2T^M`wSDENiSFGgg~+q{uQL#Y)^qD=VTI9 zQ?U;Rc1hZ1A+yY=*XFbRjdl1-`OE>?J${$nP|rK@`=_!mtQf6s_&}*fV3Ii;d^i{= zKB~l(qm@Q(J`uJc!u*cWIeSSE@wb$3?X@iD#Ks!y!h2n-#h(oH`YG+IV>BA%Yo%CG zSF6C)$eF6h@SwB>+|E@+hW}n)iOoJEG^wYt8*coL#E*ldg`pPNO%#((k+oFxIiDSL zH*dxG0p$gHsqgZM_3BfQt)DLG<)%*7iL;=sk9;acva^$6JN3besiyUobXrk2^g1W^I zyz(PZh0-(Yjq+T9 zfi4b3EADl1TI7IRJuon96(f6VSicKT-iqfPjp+Tj{aXAFyi7voow7|A$r^Sc1}+>~ zzsn3nTPOxtvo=xDCBfNTS1LnaZ=ehv)Qv3C1!&-;Y5A(yFdE22_+xo9K1=YT-Z|5W;XBj6my;;XQ}A<$ASlm zIpe*xs-?a?kq2Pon@v{`^!UWNchUXi^N+Uk7r&1P)ff#42h1bmrVLRZPg$=~@L=V75ut*bA2D$&+VYw*_0Qme;ZG z|Fb{UlBLgO4J%f7_p(&QG_uGp5){mVnZ3;P064G!vQ6`7%QLmPkBy+Pc|1pTWCe;b z%TN90hqF!c#^YX9D%tPC@&?4qkaYB86a!S-Wl%JX)~fmj!|~Cp+esIK54Xv)mfNOI z56Xn&CA?Jm@^MLl+mvesE5_BUQ{{ybE%Hw%elmTJtWTUlP6N;O5wP%=(#cT?oLxcp z!5X(#un?@gIP$~58{)MBt?E!9Ml~H$YE^jWMPa+FIjj+Sc1q7GM}|Hdk@PHSVeQ(8 zK*b74)XyS)-m#3reo2&MEAiQgmlk=yMD)rB}EEm2l3KD}kk>NpoPFD;rVaOwZ4-p!xgTBL5$n zmK;6R|7~){g_oR9%$A(%6mx|l^;C2rgcThQUKYN`zXvv~4}$lI_sC;s_s*!57DU!5 zfCxW9)a=s=*ba2#DEGg)Fp7hcZA(>h+RIgkO@j4C?!h;Y05w~t7#7_@W~E8OJikVBvu)4+b^k@ zgnMjIeTCWKIUGF?zoc!^DYin!+Hzx8`>-G}Z)CL+jhQ*)SFx5wIbNuIAS)F_kQcN? zjtOHA!|sLY!tND{{c<5? zl1vXhAD$tP+^lyVF3!PaJz->{nH)=?m{k;6K}G-Zk2#!ACHl#A5jEUP!esCJWMrzW zeKTScm9m{fkLb478ZUp-l7rY~x6D>pg>lDn*6XY?U2>8D#orroZ|L7@;@^6+LA(g+ z4Wg$aFES~3G^d@FU>MQ-vAdbwyFRaD_CF_^gvFA-`o$3)JA1ARyAq4cgvAbu0mAzn zDjKWS<_g-lJ;3%isFgH=@x_?x4$DXmCp}0#0eO#5-DLtzU3$ zph9zX&~ebj?FMDW(b!{!uV<+&?BMg<-j}Xk`yajuF|S`-dxWfXVZ|3{x`xL8SzA(r4*DlB0MOtW_Nxt@5<^%-K!%b7#q69axEz zF+cS5oG}urVHe&jTcH}hB3U|h2_$srM8{_pDKyCUhL+p`w^g2%Dm*J!7V$K@+;_Pn z<*=iLgq3M#2PPCARt9o#b8Feb1>;S!kJ&;Ppa z8|z>H_|4P*v7qSR%ii4k?!I}6-&{*o{OI=U3r6WVKK}DfK7%r=zS_!?wih<2Om~vA__X+#|w@5B*Sw}_(n--tK3aVpnZu$56aV9~bZvCr< zEOcQdLuV$zk|`#EBCDwAD-zTy2kWm5hU;n-qeHl7@!;XqCh_(3ZBB1tg^97=RlyQV zDZ$1y!c2y)fnV^Gm-(*`-o;rd(PTipNx!^*PN`~?7d|5dY{8Gy!7|JTOgg^j zd)_7~apl#Eeu#iY}0;u|RDDn%|)(O4h$7`@XcR|G*xs()_s(V)!>1C$=g4vfF$g=Nn*YvcMSix18fr33S5C;S+D#Re|)Zlbl4cyBzn0de+X zW;h(#-u8TWv09qOxZn7jY`O_yuQ%qzk=&O?5LKDs@gT)OV&NVtdf<(JG`wCHfvihM z>1NKhD6Fm0Nw-IwpM6HP4GKf1i8`f?(hN}{mEsY{Yk`E+3ev{d`2<76zQR&t6y&rZ=csJ zKLWJ&_3HJ(HT(zCf z`MZ6LqhW;ISz(QI3$U+j3ON)wzg>L?Dr>J5p^FM+AbfbMOh+3y=cU^NE-`fzb3%S= z4^nMk@9gjS@9*D?dBcRg2i%OGlj4^Kd!SeiF+853n3EK#qN2|R7J3#^d&T))cO@6e zT}i#V2T;{bBMQrSalEv^U7~(kFRl?9lxcoP1@-D2P>OQ>S8=Y9if}~Y(0fzm`sq2K zZWXy3{4+#&q*c`Vwa$oxN)CA6bsxO1W#;@k_0hNQ%`1ERB-AFv;7$ysIXfG*dHW>g zC;-YXad_Yl-}${|mh(%pi?@=jou#Uw6QCw#Oh_CA+CnH4hWl&k6m^g(f_WnL&97Y# zA8ulS3wuApuKU_nmIlu=37YR#{&*)j%udj_up|1inV>mEF~=!#l#1?!BFPIDV5Aee zoR0-9Fwn_?YkakH7G$Zg2HAoIH^XZ=3$y`sp?SQ5$iovsa`-M&cS1Pk3aUEhGM#>AOmT;hYqx^1$dzqQf4#3---DnXq40{;Ddn+?4+Vf zG3W!oCT!vx=n7d8P>Zp`BMUg$zX+M`WP_^{AN=j7mdUU$RP>yc%-yrXE!@Yvq_JMT z@{-B-!afbV77FsjK(|Jm;6cEGk+n3{-5@C%rSyM0P{3v~?e1$`&rbVBMTI3*vdeC| z%FG1F7v#rhJ&HoR+-qcact!5(gYWugko2G$VFCwRDmhhjIV1cO#2K zkrS=}YJIfGAIZu%38DGk1&~soJ2io`FQivqC++e{j>02WBEY8Eu>&E)$qSo4i%lo% z2pKLivi!(zer6dh<+7-;691GtSm`u%0wfg1GiYf!q$*eD3NRoyG2ou8XG)nrm=ZBj zx(NsZbHorObwiv-*YZ>4E7cnmaBZ71PjW@l8x%hk$O8|)syW8#lQ=%K!9j>|cxVGF z#Eh9M`0yX-tLhq+{hDRfEowE2tjyYN7^<7u)I0pqe;da9DYn*`mYB zFb;2Kg^sbaGQZ*dqQ}x@fnJG}U}s256@=Du2+r)H_sWo$1FdYQgvi#> zJ0n(}EZP)OGHas~_~f3J3TzQIsJlEpjKf<#Tr{gZ(zI)SXa6=mDR5!eX04eTdxT<2 zDN;m5H}UVe_e7%45&e_Kvs=9E@$dC{)P`o!#{=(1YLGA?i^j|`ObF|c>cgRyCFlHX z1O325gCREiB{e+bRv7oWS%Qx}2HYwmcY>oSN0siqQ<=^&j)q2;RC($QjLOuGyil24 zxEM~ceX4C_P+_$Ujae8l?jy^{8Y>|wP-j%Kjeaz<9uz?=0v6}~p0f%UKl2IL_TZm6 zgi|iEdho6rk@THT?$0gP9xFmSHN0m|zqC@cI7Elc1EJaD?|NtzSeU6^eM8*DPYuW8 zQo2&q!!4e-oz^M$_-`H+rVHCJV)5oV1r_Flvcib#uHq}1Uwr33O>57AcRi)#mJ6>v zv1X3O&nc#lBKN3hOyWtEUlm@32n5Ij^4FwuGPs@698Nr=!PYkJd3vE>p`bW4J3NbS z<2I@=u4O|&qkK6mAGOjvPPr2MWru5$c*x!Xix`|@UH3;`Ens)8Q)qPbDFr&j4gOG( zDzXzwBM!J3v!OQx?C_2cuZ-Lsv~h}7ffUR7@L0J98+XjP!d#Jns^hO&BER!0Qu@Q^ z@eBN~Na__(a`9TIJ{*c~Lj4p7wFtik#6tU}gWnkSr)}axO7z9A6`4u`i1y05XpK&E zMm!MavaUu}uf^Ek6ufbUn9#S|O@Ep!cVYBpnQ1<2DJF#?$>0H)F<3{_m_nkt{(Vdr zLG!$3i{~!jdb%RHJU4a1sBNiP`G!DhnhP~&p0PuQ_pfL7Sb7+rHvPNq+6y%8Lmktp z6tkKliBxpC@{1~~c%Tm-1P`Ve8Iz1gZ-0$*&1P|RCN?a`wvLseQ29lPNn!}^2>+8L zvGXvx?n+Fyn;}L|F`&JtLl})O2Duc@u2<*tklv?s?so2CPB(pnek9xMwH|71f@K$C zQo8AjL=XF~OmX|@q&%_#&tv_}a-yAa{!{mL$1N*7JS{q07zhxYIV7g#i>eQ;lPZI= zR66?RjAL{ufBqdH0P6{BB)ezDPB<{ym2Uv!PHg$aj81 z=xIeB#4*K2T_d&QowpX*gNVn-aOPTMU7Wj!RryVK6BKk#83 za<03GzcPO@{R`7#^NT|pzfLZ{G>Z-FoQBAlk13{|A}zR;YUelm)J0&Gp?_WEw9M?3 z9#qvu)T{gFT;N}(kHY%WPiKV}QafLPQh4;|ujPAH!I~5!(2_#xj=0dHiT{2F*$w<2 zi2>)QbbGg}Z-uN6ZkdU}$nb*oI%EU7hc>SWrl%hGfo7gY=nnQsRkUXRkOz^x5=Y zQqSCx8>XEV)>*If$@-lcBB2~{XziujapJp)UoKMS!>Vqg6O{%^C{n{jw zmTp=!nKZByNiG~=_s~ov-KH2Iyt|1dJ5MPRAXpd4lJ(Oo`th2_gmy?5UjSKF#;>ph!gDk^ zoORqRx)hk@w|n1+L?=fRxYvtA7YXo-8~=pCPG?BjTuThwjW|aLCG#=N;Hk>_=)Wv! zz+aMBUn`{Hso|}#&NYIU>Zt$jkS2aN{h4oKKqBPMt(mZkZec#>0F!1HDIPp5rLn&2 zTG1upMNY}&cu(BVYZZ5Z{kMk;_QyI-6MqZ0l`f^bqt?Nqy3u?0tQ}nJxLQ~o0Njj^ z?}LOYUYh$OuK_pgj+R68ab1z1klZ6~=U<(^d)5|l3nzio=CLU%JKPdJW^-n7GRn}h z_;l;1t1f#eUvy404BWltwae#h=-@sen@$4$`0^>2gohvx6Ki@c5@1EI(b^Hf!)7w(;*j zHzDNw^vh$&C3XmLVO8k98A5JS46IbmRCKOax&$d|`Uv==fM(S%)=#TA!rLBk)?=yf zPIo94_c`6lZJ3h~{@8VRs~XQ?*%pt^L!7|xX|LdO-XPEbN;z^#-I8R_or zDoh?I_w18sj9H-jeD^BB?e!`6zc#723u6S^0}cmw0`)J9hnb&wyt|Wjc-QhFQ#3vR zwf6zH0%1K#jyj?;%x+P9PU}Q{l72b3s~Ur2A^ZFkr!mYxmvM@@hlBH|9%;F<&%eff z(_{;|Xb;xRbr(mSnZwoVZ-2Y5Heszbd}2B|#13mNY;C+`hP9Iv1Igl*RCESa`P+_p z``&5d=aUvla?vVEI7O3C(>{fX-jScv+vr{NBVMtvSH2rGGiRgeS)cu^Yl7EB`z*_!uu-X9SaMk@meNYM23?Wlh#|EFi`i=x==8?m zbGzsESQT`_+DNnLlQkj{b9mXJE^tbjG84 z-k7X3P1k*yY94-rXUU&tg{CZK=3&Ji{~WP_?)T7O@sA@aG>`l*pu!eRjAq!uW30R+ znD~wK^jns^@J~B8U3l|jh5a3~XR+*YuBdIgjv3Yc@?(VH@8_d;Xz$TGwvOH{KbZW^ z>n6=~=7Mugq{xLssGwrW5UV@4f`(AB0sI@ z7T=t)M|Q`fpWCiBenu&Ow@(+nlXG9%$i+|TeWOAAK%6e9+Xg7jo*=ab9uU?ZRyMe^bD6{yi{q=;KoLzxblZ_AVW_o+*nuFKzQ!Hm=XRK&Iz)NPD<@Ah)_z*1}oG zDP+g9Vb^mUdJ~Je9(tS8kJ$;3fAc@)?rTz6eqKH1JV|$9mF0k$%92YlP}?RG^WBUb zo1GBt-S5#LZUy`5AZy2U6_&8dQq@BifziT>&cEa+4E_fLt#OOCh_rEI=3Eclpop0R zsaMVk`egwYtBQ#;ki!ZrV_xUIa?mp1_vsCf>u%z$h4o_9609DLS=PD`jplK!yaZl5 z#0*%!=5#&9Uf4JTDnrM#4x4oG-*>&}Ze(M@5v^j!oR6iNb)=QA69F%W0Ze>3UVRdb zl1d0Y#z`6_Djx<^(Ytf>snqgOEf-VjnJ?dM*R#0Uo5(m6|mv*>KP z1w!ud&Ro$0sC=O@W(W6sq{??sJLZFY5XhtHV51^Slv!P<@PZ>@_l=Lhb~e=-&;K|__k z_Hd6S_2JWciVORJtq>e8^-klaaaX@~RM6?WSb)(5)xK3x!?)I{&%A!_+Y7(8;_IJ% zCzHDTol9@tp0}Pl@_$l(ecnWd+85SMo8JptDbtws>b|fRc?_?MUKrjRR^;9VmO4B_ z4e5rk+IK)~x9&4?h&@yA*e!XiLo9D~HmxwpAN9$150eab>y`@xt;9@P?V=b+eA`Av z&)3oS$(^soy|L_#%k$!1ubWk;Zs1o3@8G7Wj(T9IK%uD1M;~6V&Wqd@1q95JR{kgQ z1dcwqNA{4Hz-#5FOEyjVM-H`(`o|4%w(2I7&ghLitLpQz8#PZvf%%c6&}2U9$j?&0 z-F!xF!pz#_;7cUUh1aY-W+=#}7)Y4aQ_(Fm2LhY;jq*OCqc{1atM&)#AzI+R>|@{! ziV6kwp+u6g6z(-YQsB6qkxYTuN0?L{#{{ZxSqtIhYd?ZtP6$PfpJA z1@k?5zxR8W=UM&szrSUxrSe>@RV?}{dTA}4J>~NdJr?288Oob4?0~53fpAoWeW2)6 zUZH9gc>NH$$!E_Lv=N=;VtzZY(`U*S11nxF*(`CwbhR*O{Q&eGPit-5t)GV_J zO&)ok&Y9dCpsj-X8dIA-7TN%jnGonRkE(USy|6!T@?IWSH+jHIPagDcm+etM^3$HD zi$qtV++BV#B1||myUyLjjtM)KXiVkX^z@PBl00~O&LaMWo=dIFAUGYJ5S$OpGO4h1 zJ&r1euD6UC=jH;k&V%739XDT=u{>SADmz_gFE*5xO-Bc{)8Qz=BFQY^BGRIV{Gc{f0<^R z8Sq$8u#kITv3)l$A!2XXqKQqB1LBQ-kVNY9P33p zAjUz6*mw|b#s>_~4I-}{E&AthiyN}NvpkX3d$1dFz{(BTMKG{MDkn5IBN~OhKJ%%P z*sHVJ#5+OwC|P1IVrqfrdeD}UESQF3HmE4sGox$D@re(4H`JB#miV=jWJ#^;lJq!z zLtG!;7<1P9(l6U4Nt>d)e_{Jh;v>%L+|N1hbr=mmrG?7LRMO{2o^zm(IN7v-NC@I6jsu?_|e_ z>^q;Vpe%U!&5@qN^n!7A@ZiBwH;|qkQhK?XU{VP>2?#&FaY>p;Tv3*SNZXmOOX`_kz5NSh%r@b%;_861?&*%z89x;gXO%mz5MdGZ5y<$!GG!9getSv71& zaL;RYDA=(EWBza>qtdoR_31d92QMNleW)mYc!nL{bKhU)Yk@M@A#}ww`Kn_+d>DA-3_snE;PJyN&3MRWtmAZv`v^dH;FMs_JOk6J|WPj%=uV# zP?0ycO|VT3r3jG^qY9vuJCoYs-%W4zd+fU)0unyd;+Yt{-6~B9t_ye^kO>OJ9qL?O z)$9~Ox~u7`U5AmUrg=mshQY|w#)l}&rvC7}3m2QUj0&ajJfBMU}elJ~2FbfDe!Rgu@ z_TPxx6`Do(u{Y{+)L|fp!{l^H$f~dERGW!E=w3QZ8u+K=#+l|EHD;je>Dq9-p@CN% zmn_MM#;ZEj5&EhCIX80r?*^F9;DBlTMsRV5nPqOj`{3QK!p7iuuKLMtPk*5(e8)Bc z{UxE;U`bau`GK$>w49+hIVKB=B)7dYl(2mgo}G|SejJfc{yBxWd=_qz&^f}j`XpHh z$`i;O(C%&m3Fi^yDy-NaCn;X2hl#c|jHJqXb53ttSJs+E4_0RvvWJ}5-5Q|WyBd^;4IWeGF!HMS%r z^MO@tmwcgI=UXJfq8La*J&;vL9-r760t8PRCL~RFp-FKb3icKj`_t;oS$1P$&n^C+ z5BC1iwsOH^mmro}1f=3YTVN_~L5iN$k!?|^U!tG15EO2(SK!p_b)gu_F@1f&s?)0z#t)CQ)Q# z%TO)~uZSs<)Q2sdb6I>0m<-MZH3F*^aR1?qplvrR55WT@+W2yR;GRX2H}F4W+F_26 z0LDE%h0a%qeDU3N_><;xBd1yM+^V6NfV^q8(HJgniw7s5N~}UI(%bvONRoq>|Kn-N*jedD$IY3ae}e`uTJC|lApT667Syyy!H zHfnoHPF8I!G3Bqnu&t8gVmv%I-~^okL$lYb2quM~R}vcR!AE}hi_&sR+YT#dxj&?z zibRj2T+bSjaN@-CPQ!tv)n$s~FP0Ehac5#cmo(;}H^@iFuBVY- zzh+hmF}l<3=v;%vaNR^YFcLR&Ax{7G+h4J`rmf$&UO?~g;H~dTD+AC#Fc4L%201ZEG^zy1I zI*(g3s|PzK^;ULi2f>sR!1w|oIV9=NgAOwjDNriaNFo`KNg`qqzsav!n8c(9-}`!Z zY}IVkoynDMT?B`4%1e2AO@ggVI!k*4lKY-7!qJ_P5Z@7NS z0MK>-gHVPtZK5Ft%PzOhDiu3q?p%mBRv_A5cnVy0PO#&xTOj?-3l?mJyk~feUgyCE zsnQBZTL=bZ?Kcyez3QvHY9vlJb)0Ftq#4l}N^`eat@?o9Qo(>1*34#0I>ohA36)7@ zOK?Yo)m}&uG~k7Q9GYJ{AFU&&Mg$8--oVLw*kNHOYdE$ci&`6Gfy%<)OcCed(YlYOum+6QWfcc>F#E%95Z9B4v& zByAIRPE427&fP6tV=G@d+}sR*%@IV|?plX?J^Z^lew7_jukARs@OdwQ)Kfh(jMSoW z88g-d)iSH-#pBw9O|M;&W(8Nq4-5fIB{hKJC8a58sUbozI}Z^0n7b(knbT6aqEfLuNdS zVAc`vxoT2)l~7WWOs$jMjLH$M;Z^xGDv}r@b%@OIU&AwktduV3T2SrWt3IPN*)Uxk z=i7Y~Y%?;qh33qL|9DDoie<})tgZ$@883$_4%-=cW5x;U_QVtN9Enl+d1w=LlDspD zA3pw@IZS}fN95j@+$O+dJ1{@o=94V6*o~LYtldv9x3|yn;MHS`l`Y96n6(6*4%|)Z zYI61X3j)l(nOb--E{4@3*dn4Q6GgYZZw6G&){FjOO4yP7FWdj@ZA+TY#Xt4nI4TRp zI+ou<*C%uaVh`x!fc#+XCZBFd_SI3{yiGyPDs!XYeY!JBPoAZ6#$(+P@aCusL)v_s zfSbp?aN=MLZ8yMfwT2z(yQ==%Vnr5hObntgd9e5K*vewuBACwz`Z}`zwFKOaMdI-% zBpsPlw{WSb2=q{!Ro%QKMvFBm`DCjyLs=}^?mysl)2|3Lf{MVHoQGC4QB1eEBTTG4YwGATK&5)g~^Vsiv`<}{H|NUv^0}nPpEMoQMvI1yl ztPz)r2bn0Jo{n4pT)9wB!%5egS0oy?Xz9oKF#7h3T zxyR{L{xbParX@a2vM}Q6tWM=XSVf$B-Pj>B#K~fS;TZC$TMtWr^@r~UTRhs5fq$!{ zbG|h3)@m#NcN@WMCFmkTlM{VGoD^Lv>sMqbPw?d5|4P3#L8ambDG)3{M)$3my8OX+U zS=}zg1kWdNWpO&y8G4P+4fQUyTP@8Xlw2)f9F(8jY39Mg6vyYKVS-ix}R z*cqb52aL-53gj!U317!Rrx#f=?<%j~t9)L!vRYU}=aKzh`-H3g(&+lI0$x2|FX;Em zmSm}t=k$BklGWneur}ZQqU_Kk6jB-|Q{dR7^34M#%)}wH$#iuKl*^{8PkS3xXUO++ zsLkH{dF|i}R|7j?uJ5zaHWJ9NSsDk<4@HOFWA<&2o$c|M&t9t!v28qM>DbJV?(?~= zNQMGs^HFB#9dbK$S$S^me8HL^Lri}3NwPsyD$b8SNY;urNlubQfzSRJ?fx6Mopfd< zxV?jOcXOHv&o%1)Ge6IpX0bsZd}!Q9Zy84)wEC*=CYYTBy&W3$!dJy^ow~{QHd)W# z7g`gvQ}U4RBR}Rf$D{V>M{gxlr4#0Rcaj^z4$Z@k-W1+hszY8i;Q+rk7C1N6i~P>< zK7M0$=*HmO=)pUUmOHEAdwg%k0oRJRI|CiIFR61NvH#SU1Uk{b83=pL0-bqP!VbEQ zn{({BUkdt?)i{r{u1105M%^(hAxlq3N_uA`fAxfCB2gJYR)UoU$4Y*3g% z<;`8M&IHfumNb#yMkYrZm}MM=?_BxSVEf8g&fEVEu$=wy?7yzJEx~1}bbS)nz;HbB z?iGz<7H`O}@5o0vAZbgE>kB6g(!6MRJ?l%_oz z*GZlT?^7ndU{qp1aJ#TWp(js99#woAf_D{Cy*?GD$f|p$$k}-rGFNVyB2M$-xpt-D z#~1F}@<@7Yvj?=04%vJ^VTRla#gQmY))Jo)-6(9J+d$ak7_~`Kz{`|1$JCO$relqM zHF=zu6>&&<#!VfgPoa!Ea-SN)Ehji3m&p0+ODAn}%N`35mOL`FaDXM-&(o<|Wox2$ zhM80i|4;lR$1vfRX&-h_PM8?`@{9w*=iO+u10mp^7!%LZ3f}|jL=n8mP_~oh;G|TM z<=&>=vFgb5V63zUEiwLyQIWKB3?M@mH#{76YF%@Y8$cdR`pJjSdjWR7q$IR%y7}*x z_`9;ZUf3;0kEOU?JV{hI#qQl-0ZaK#Z9mL-D>~7oA z6pvj}Sc*+DDM$u_rcZ0+BWbhB5P5r|_Bg#m)yq$qs4Y|Mnw~PJj&u{Vy4@M%h77wy zxaA}#WQ@JJvvHQ~!osBv!E^m=;PSHTyJ9D44h2zP0Vqt$%rr?_Y*{wdX%_+eNxr%R z2$Kf9aKWmXHA;0W+~8zhj@*JDn?a}ixUS!JVR^199a*}4s$#BAdm^yWDr@(x89(5K zl26sVRPR=?VRDhA75FR=2cx#yaK(VlRJrXXyGi-V`L<)vJLi2O@d=P1t>l%8EwZ(i zACAA|K4F};r7@* zBs$M(u~Y`YbV@(5-V0g9~9+|lP#6Lc>sLsD?D4oTKSMO!|vL;cP>NYi+XX;(ZB*&r)a_w(<{Dq?Xb?1mz?p;+Jqs!`krU&|5f ze%@uXk4|_H(7^9k6hIYYTU08o%?|B^jNu7VyDO~Aq_mP9ilu8>bVY8{9g3>i=Ft8C zkkF!*LJ5I?P%*tovWPCHI!W`HazVKOk6dbfMuG$zV7N3w!pXdh{omHNm)dT2za-{` zWoz6d#|=v!*{#%Ke<^1DvL!judZ1)7Dl6ma_=T<%2^?xW>>qQ%k^K?ebGGu#mJC}~ zr7sCGSooHDUTmL?WeR3?EE1DqK6+e8;!999~G-G1G?Q{+N<3(y=CN;Af5 z&Gp#W0tVvk1_3Iy?;-C(+HqZ|Ny{BNQK6I)3u>|Z6bD&KxF-C<1wFP_hwZg1@RWRG zY6@lXXMb~~=Pfx~S^c|lr@}l%e z$R~lT94J?H#pgP73Wz4N#7AjU{kwY8}Lk#WXlJrYr7>{84O)t>znxSQ>U%pPGFJ>ckC`^L-%^mZ3}# zRa1BPqxm_S|ILnmny&bSxQxO64YU(`@Z25Znw|fZyeFl3Ycv6Sc%I;2#Mpq}B!8^Qn*}nEfPSw}nQ}NzncWn35^9NDx?!+Ppre>!Tb)hkxq!jJdpTjoa~JN1 zU-|2O+jxn`X75-cAWiZ@AkqRA$nh0C>=4(fK8H=qfY&`ipD!vcxBro7tVeB9_hTi5Zeus|Xj$=;v9aI0pSEtRjn3m>$TbNH*eQC<#&sbT=BLoAT zBKrwV4%H~Hz3CNe3^E0Fqv>>#}Y0I?q+sPj7{o^#lV3b2Xv)Or9{u2wF{9 z%8JNhm9|s9Ox_ckG*z4CTOEv|DjCXssvUu&nDXNm(jS$=BQtE+ZBO)npEyBS?95}| z$H(b@FVD^EY_P&XGQq4M=%s`P8_J<)G*fm&v{G9 zPsQIMu^v7P^g>HwGekwfyS?wt&|K`?44U^ge~hcAz_<&|b3WX=li zBKMP<)%gL&nO*#Wke2v0Avw~45TmkRksP@*#7Mn`CpJztUw&(JzI$9h^3nS4v47-t zEV$=>{YsM6heje#?79Cknnj-YthkB1>+@Kd>-%{qiVW*ihoMKztl_v;a!rn*!aboU z0`OV^h#2?rJ`Zh)FQn?HoE6u~a`<>nr|OD(1Z~0w0TMOpK&7}q-Jy6mqMeFev+o4@ zhps|Tj_nRUf|J+%I{jx`TvVTXhcoD1+_u7=8$YB^S*=Hh2&RFc>j(|7jK~s2DYFkK zw09Ml_B~Ino!Y9zqnzmdY9Ih5yQ5kFN$q5@sEl`?2D#Dwy!x<9(%UnE$)s4cnYl*g zsnen64XaajNvdY+REwrA3`vvhmYSpz+R4;uZ|Qvdd^^PkL9d`%+$nxbr+OS!PVMwN zHy1UrQHj!7tLrKxJ%N<%fs<%@@7jM9%PmMc`o^~|(yKhUh;5q{HZ~IsbZ2cKGz)#M z3$+uxSSGO>TkT#&%>gMeTbbt22*=cuj zHXVM9)ABjyS9iYSZQHT;w3F+>rSdFYe1j~GSN!kJED~w!fC>1T{Ghk9-*JaO{iknq z3@1-N=k7D?I7w}tUTRw{z-3eH!DwNrBQWKCv55k??srHk0wIri#sA}oH9>{a(T0p; zc!4#--2kx}KYjecmA1rKTu|Y`jtL9D*Tc2Z8hmX$1RvMUW zR9SPhgJspZJnhA}RPS1lyY6r=FP~{r9EOOsCt)+8u7CN${~VtDj>Y4s2t2N!JGl9- z9&9)kjU5_Gxlb^63HlD9X(!7RrXpArkKE|r=X;FqnPHf`P~I2`ek)YX=99}HW}l(l z54@mn-Q?#-wh0G9-h%2i>?1DVU8i%R@hk$VG9Z<&?SguyWcpf+HdktdPA|Ox7(*Ft z8)?1`Z^6VOmhwX;077}%wAe!=N-5M)I$1S2_zosY4)AHKdB8M^=_otfSaN)+EmWA@-~}2_hq2W8^1Km)#p~*=}!sf0zsc6G);04Q&FT% zEQmg=$R{xsWN9L9l3O|lG1g*KCQ~`l2@|uFTJvGIq9xqKSABctTE;M0PZo(11WSo# zQ19sm7L1OlCMeyQu*7di%`qZ=nX4O)|58i z0k2x&A#*wRH7K&lrF*D0;i7Q!Z>i|g2wz(k+zV! zAWjd?^=*v2I!oKC+^6mW!qd9}<;tV#d{M5i+eI~;M#2cpjw6q(`Sw+w(*g1$2EA?bW5R@wviFQp|PJ!Qh({~}$qUefS;^%fKSWOET zTv$Em<`?X^kZ&ygT95@73x79pC7tcT@mqrx2FeHqC@cyIO>r0o;P$*|T0QdAdqPo! z7*-Vw?whL)%y-MEp=_@plN=& zpktU-#N}9FM~6L!=GCwZ{oWSD9QsP_U+IM&jF`<`AOV$)=?|RLz<_K zxuq;Sp4V0mZOXX7CsMq&e3bR`MDmPi3XPS zA$?(bf&nswO@t;v)InziosgXIT|BE)Z2hV^zFK^rz6u2FhsgXF_q|vYV!rE;1UsFP z`tbx^81f8%bGXTI;MxexjRT`Qe@no>ojPk9<7EksViV@|Im>77pQKZL>Yoh_ys*lf zMHbPd8o;t>SS|T_r`pG`kV=~EYQeS{-lWB z>A}$W#0nY*2nGrxYY5FfMFS5Bkv4d5@Ee%9S6vcY3|*|-;B>i6N=`&$fzWrW1rM1HB(01Xe z__BHT6v>iB)Lq&77y6VP!eev;zl!V-UJOZz94Xwm<7k)-@z6_8km}q4Cmazo7roW> zya!K`C0!uhb_F`0Yl3pTt9>zzr&CpnQ-Xnp3;pf-DHnkF&!_}WH@ueXyGy3MAbv>g z4J+qqKaaW)*+ya%e}xj{%@2CF%j*1$N*D7k;as{3G0x$F{V7g}89U~cb6=HOERW)E zuQt;u9&C9)cy!3VH=AIfS!_LU{zc#vmw^~cAMYecr07(S!Ye?ZZXthXNEz6qi~d-0 zfZe1AV@kz%JW^}H{^<|B;sip5KJK<-P6#39CK)907KF4%Ps*hCa*LyQuvdH53Qxxf z<}g7YAT%ol%V!&2fX1+F@*`E1`Xb1?m>A&NXPlDn@CW}km48{;O9QLYc@VhOsUP_z zF zha%Nmdod&}4kG=EmEhb!xdPU0SCAFY&gF9a9G`JJm+61TP%gd9ZhDq%TJ-J|i<#;9 zk$f?|nVXsM;BalNm6_R2Fpyp@AvBkiZNdwn4p=ttigG)+wq0cB#QD@U`C@)KxsJ&v zcR)iB(j@HldlX&{TpWLT|IeKhQ^0pVMqi)Nu6PuFi2Td@WiO>rk3gLLe%!&B>^X z^LXy0IG3j3%y|K5=kQG&kxJoi&cGvih8h31$mH=VL8MX^g3KwMbGV zy6p{>h{CfI^2rXc;AYZ^Y_bKa2Qih18f*t*^PxHsyZW%^RhuZ<0>LwE$8e{v#fU#< zSAV{i@z}jZdvzcl%W3oO8#TAQENCft@XHVB1bd}!9&AFO`f*6yC6!>32zoi8=~kQx z-XT~h(6;&Rj#$qe@z%CWYu#K!U>9kQ;n(h?ZavA4-(zQf`vco;Fqc%J2ZM!WgWM+E z#yd|P3O%ec#WwWQmIxC0-8?*Lq@c)cRAo6hMQ_~-5!=CgyXb;nTWrPdpZdz_n;smG zU1rrD_K;wD3Hlytuym4DWE;JiGzd;kUKEfi)K-UWX6nO=0x$YQkw9!4lo522NQeQo z@UnK{vhlqT8iR$9eycdTI`W#O9v;`ow~Vm!1+JzbGqnbegU|FUu&}08wl1t1dVHGU zf%D;Or=Aq!cIdMBDn{peW!#3d@&~c$Bu_LU{ zr_nn%u0o9yh}GD*fnFWi7L^YS&WUsqgCv89Fm0ovLY$$z8++I167L$^^voU%+b^(# zuN_;m*!)@3c?aK#+il$Nayd8O$97lEvK>7NoLg;>(lshlnW4wHJ#Br}Aju9c71s;T zQ_V5AUB8cD`wY0w0^2yQKF?_uJU2A)jqzcJA}!`*Sw`13daFIh*K>b=`k0l8F%S%F zrgssVor>FJx#F|1ga}-{j8vz%SNZ5GxL{N~qg(Lc}#Wnk#4$;~V$~!msgrWoHzvkvbK=ldE>s zT){r9@xhBOLOna!8%$vxnJH`C8)h&ar;`pDsyDE!m%z_;bws2~DaKB7>kw)xdkqZ}#h&(h{G` z+aY}`k1SPGkqx}JOxy?|nP7D~6t1!hET1*Ui63GuSO#>_X?AmA$D<{_|52}MBx)}n zyg0FFi#3X`i;RjYA4BA2QL(fSv?wx^c@U>HllWisUqQJUI5zXl0Y^4Zy7%SGSGT-l z842!?|J41XO zUE;yp-=kK(={|yi1Vkke=KpimcY9$+`;cBAem?x}6r|?4;j>J6HNHt+LZLidmgH{i znm5*dTc_%wDg`a^%@kg_5_Jel`4^2_;Jr`P8EBl@DgXHUi^ny}FG)8Bb%M0Q-Y~=5 zTQP;8XL%v=_QVr%6sdC=b{qo4kQL@xhjwSziId#cnv4H*=qvLq@cH(Ef89p!^I-U# zvx3h@1OtME2Z4H$x*%v_Hp`HQB4b98s2Qp$O>N8=J-h8)Os3KIq}{PP)pf;kx-+1Z zOcj;;Ul45Yei*f$)}Ez%eY6YaY@3CTrw4EFUi;On({A$HW3Xum`s9GtX)-ctR4967 z171BZ8kOtAYe9kRRt&ty7GIz(WpbiTYJ~{Ys*uYNAanv`u7;C+xN*l5CpeA0oUrKE zQ5HCD3eHw%G94Cm^CnAF?E5D1F&RQTM7B*M0&pHbNOC1P=+CgN`}%H zV^rzXm!yc!TJ-_HJ?gD7P5{PsD%}Gh+jrsqniGJ=`2FDZgJUfKTDTBJDrA1gMAXwif>>?q)UnJTX_s*6ih?1%F;5HJQF#fTw4Wf$^D#qr&_o2eJy^u@D#DjxrEKE-M z>JnKo*+28BTBmONuVeF8j;{zrZ?sKV1NqFqT~Qt|74CQO*N0p8r(y*%W`>Zd^W512^-~AlH2|@ z{sUx7e6nP|;Hu9fVGdd1b3)xH?#**7>RvCdDsLvM&rWci8Nc;lY<1apP*|AO}*r%UxG+`+V9A$ zHqvxhFF7M^BQs_oO`s{LJ#m@{aGS#s=N zauQkWps|w|9Q$Z)bKvt1q@EYsC%35Tryl~wgkJjS>}&oh!F7SCGhf8(pMt14$Mn)E zAnt+JFwlP#TA4e2>ZdP@E2Iw6JLzsEK5IS$U1f^(;YVlVRfFJ@h+HOf_OiHMlWu(u zh3``li#S^%s-Zhc9C3tZ%6&j`5y)J_Mcb)7 zE4xM^)00kRgbcxGhg%$n%E+dk>(lH|`PGprF9li_s>jT+Wwg$N7pi?$3sot>fTHIX zLW6br*t3+OJnq-VGc&)L3zW3oAVUU)(Byr3VKkbd4CVIF^AxD*;(FCaCrnBY9`LG~ zT{U})cPhUDSQ3s;+yr`MBeWRyr^j({*j;Yiw)bHNhn)p!%)9YliF@9q8jzU>sR1tt zOr#d5&Y`X2?+z?z>!Xi$vlQ~zr6)a=^Oir8-64WCo9wK+kxVFtlr(@8v05o>OZT_-v} zt&MC1-Iu};NEIKU{`&q0?rY4zkJ&-%!M2_dBw#ur*E{NUaTn_uW!$n6R@vJf2ah)3^0yp-3-hjRk@Dl+SWd_0Z89Nk}f-S&3 zxE)F>ck!~~w)oZrC5RqPbKX)pf-ts~$bs`tgO43yY2PU~*z)OfNfUZ-b32PHK!)-# z*%X=U+bY{7Lv|)@v3E6T(vcKl`_6zDrUiR_uFpuO3Ze_5izItOoe#irnK7rp!{rT} zym!Lissa=iCpLT1yT73`JlKh?uySJa2?j!D*(fWAjqX^dcm*_k(u1$^Hu`4?$|kf4 zb-~-G<&#B{%77zum$Vb**`f6Qk>A#sB9WdvH+L)Vj=*_?;Mhm*)Xby3aBO&|?_xI_ z_M#%yn|?cB>#J}vB_6y3W+`vOy84@bNNj8(Q^yQZk;vU(u{WCz0m$BiZaz1ACKX^C z6!F+?4NI6Kmv?p6hA_jt0$z7)OT59aH)$(hhOfnvv< zWMx1VIUf_rK$IR`8PGbxsA_|gHQ{$aE_-1}DwN4)C<~SOWO}epU7;9Um$aFoci_1P zpoNN!8OP{NLF>b|M<5}FaZVkR7}`F0k@V;k~#RPKFK6GUaDV+#V%Ub%Fg}cvm-X z-%Lkl-}Y1I_S#A8aC;Lc?mmC8=8i3O&X)usEJQo0{6t#MDZ13l9a|Y3`Q)B)5BCXw#(*xm%)YqT+4ziIu=VYq+isA#m=h0nby&93X0H5G zFKbbCp-5s@X77Q$^?dMS3L%Da87fj%_#RN}RQ-zG-b)3iUanFBXH;03d$?z|5ntd0E^@MqZ4CJ(V5&*#2)OJ?6z`=iwFklYV!!q{0Ufkk5o&b@xpW~ z4=ZrN*$m+}ovI|-08!n9mwv5Nod{?YCWE^223fA}R^OKRKmQtnbhP>SFL2yrI(E;h z<2+afLF1kc*Tk>;_4-@D()?!SJM?;cyCe^W#4amH6cP-`)a4QyR8TkpjTl(acpw%_ z8jaK`If{*)k{f3hstY7bqPJMi)lpsGjU-1V`liSmc-LOuE#3CRdSSia&N%mA!Tg0E z5h7e|Glq=Ez9{XyE<0?y<9RwIZAl9Qje{A= z7N9#gIcb3yUVN;|@$Mt@=GFmeq&uNunGNR$ew zADC$G(@?pnht@!y@69N;%44uil7q9)_PE>5v(LUAr*(|rfbdQw_&3Zq~X zEYs>BI0k(Ow}HWSU~b?D3VG=>!UB%RzK@U7{azl+lR$&8;1F;m6U+*NUP@?ikD}G7 z>Zh0vO!UvA2ZJv@eun_@#qFbZ#0d~%Z~eJC#8&i{%NEyz{Ztl-T%<>86JGIe9G?#b zCPfk~Pw6o+-Q1jcorK zn`NisFRJrNJa?AqmEoD@_=~`>d@8;rzKyH{srHm$tk6Hf`xIzHKM8CaamT4cs2emw zTz-xapRE%+-g?DiSQf4R?!VDBUz#oO1uHXjf?$AV=?L%-ON&Czgz3Y!kn7c(Bzbd7 zC-jjOv4_c8h^$sguE~$ksmg9%n_#EZs3`HlvZqGjX>qOMw)mEKhhj@8@(@>mAQXr| zfpDr$RUl0eHOj6E3{cX9=xdnK1ub4EBG#|aQ^*a3lwdm5!N4;zq?6oA8)+zbA@K?n zV=_i385P`ys4vjx6t@2Mj;PpnpUlNFd2rPM%MSXu`jNC%w!&A7LZ_7hK-HzLd=ts? z&qL4+_eod$3#lAFs27pO^~99PldRq)`51vbfuxJLfn_p2}QJf4r8P=>a7sK`W zq$oR;P^@5t%IrR$js6AEgIgwKKtQ<^lsIdIlH4NE<0x&B1OxtU!lbx_h$5*Li~Mz} z4)tBxGI^q?#Uy^*K%b6m2NRQ_)HciO!!7sJ1YL;KPc!gQT@59Y^VO*8krjP8vPfdy zwc*(|vO8)gunU^6pqX0AR(t0V)V^3++;C(coa|8h31$3?j|FN~Uwi!|z2Zyb`hZH| zkR4bS!K@=-el%Ag_WudF)z1jQ{nK>0(!gXXyQVaJFPUl+@08YpqjQMt=A{HT3P*2a z#t}F6C)o{W_5Xf`XG_=g^wQ+P+b9-NCX|0_iBF5|k{3%dV)eI%M0m z&U&+*2%raJiKWLqU!6_XypZI(P`*uIR3-9_Gf^(8P1xsiF!1s8yRrwc6~DwQQ~q@U zv3*`2?-*^IiLD`B(9r(S=T8fW)R+2rpUZRAoBal#aEsk>0hHX5Qk?v7db+J@JB!?T zX-tnB5jM zc^SQhyeL{Zf(|n~Aa&h0z_kq}TNa&C*|P0;Y|fE|S!ZL=Hr^9sm@Vn`SrwcQUJDvF zoyygw%+B0eopOAltsZmx3vQ+^p5EL!)nZY+Bv*OV@@8j7}SWgWmb z*sMA^IRlDgDq=H~Up$=;o9DdfGgNl)LYMo)j-$+;6FlUci*eWXYIO0l1qwYlRKp@r zXjDE5*NYy_(P9=m6=En2f{o*Ksz-CW{IwnAW6eo%E-yog5(?-=@CLFft~@F(Q)|TTN~gbf{h6_PPuu zLjdE7<%iuC{rYQ>zZML7GTPI>rzyCgl>aZB9Rn`@!Q;(`TH#Zo#*dh!w96Jo`dEe=2Q|CPIG(G9FwuRk!(foIF z{LYu_U!DK*(wC3TTl2GLVw^5wIy zoQijWQ5bGQ0KoJwfBjqS|Niv7H-G;t>0*LeK+p+R-+q=C!;CueOPqj0l>X|$)ewsf z`Au0v3Y}{&UEsO9Il9iuu2c}rHiF(tXw0>fxN>~a5wcXE#od2L%t7yWx|K;xhH|eO zxo%M)Xa%)9qI*hRpcd==v9f>Lv~AO1S-P*d4{2?P08|Dn@k`~W1TUoSLH^pM(Bfut zSTJQdJ#zefF# z{!;2ZwP;#`s2UW<=12A@Y85LdfYjw7a=U*!iQ-nbq%F`Ay=dxDS-;ouiEH?fNyI0r z#XZV`;I6oKsBtM0H7XW_4tNzrpA=&TbK#tPAW+5<2(8ZdOqkAh@l5@+njrLykgi+X z12M>zeweyUmTU^rn|=dis@oLwX|zstC2AmQQNTs%PQP<=!I!=mdntC! zfFIV1IS$F;Es`@=hl6(U+w?za2SFAW#HYJF*y~LlJJjn<9Lo$4^dmxZO}+=Fr!EX) z&A>&7M91yqRbDk&A->9Mp_723=cE|xUz+4N!8+ASUMbLW=mVQg`yfbr=ZjYI^5{Z! zGi;;vK;S>%bySfhY6sRA9PT{*XmY|#tiQvUb+)8|Uk8gE4xKDXipD;9J&9^}`e{&u z1=mrW7x!sJ0&TmjQ+#z60_wA{ha?KXxB2dl_&lmU3~$0VdvgGNPfW)60k7&v+=QXB zD_$**?ts*K3$q8pw$^%4G#w4QY#Sl5zKvpz6)Eg$S!`_oK)|o37cix;x2~lno}vWcdj|U103v~Ay?x& zLyb@ey47!!3!M9MfRE@X`N0EQvgR*Io;2HPnN1^@WP)BnXqI|! zmJNio(I3G7+rq%62%$82=77s1g~P2o=4%d`uV)YB;yd*E@E?mU25DK=r@x?+p0j-% zkiHLz4{aovOoCpE4pudp9h?%34w1P4tRnUi6l0yGdIZP`kEVkW}`HkO;{48eW2X!0I(5t^R zW~JQ9tmG05pkyPVIRw;ayQa6xE-UYV@48e}5_>@Q3CJGZm2I19R4kj5D=jB629PY- z?vIqcon+aZ&w<9*FxM5vVfZO?7$w$*V)#K_aP@<}rhQ*dO+8FV=BZ7Fq8W1Hl>yDF zKA#-PfL9&$kj|l!sSOj7rn`UQpUe?A1Uz}n?Fa0R$&puIeQ&nKg81A!oI&sM;GOg- zE1PhLU>XQWKxmeZt2LEKoe=2-brceO7ekF^W@sL1kUS1K5l}~cOlM7OiN7G;6PHis z(uc?;<7)Z(Y4yVM6qZ&kq);a9W|UrF5cKgL^FLB$POv^-MV7?&fe6weX&rSl3h%s4 z>(oo>!wS@fT{>=L5aWg#Yvb{ZS0h3eJ8JAipNUg1hWyaMYS2yeD*mFW>%&V3;9D^N zhf8}l*+C9DWxjwRHEiwZ_4&*Pc`1RBvQ4Nbv11WKTdlGx=}izgtcKZ16O~1*58tIM z<9!4{JCrKBA6G))ALr-CjdO!j=k)w=nj6pc{qD31zBiI)OAp=`u{2v2hLno0#^c}D z$E-iS|F=K<;e(8qTjFnv|MdQR0$#koPXhrt zb!XT%-hqi}FCc5{ZE;Ds^We6RK9@n>{_swpmBhpvpWSTfxmXAHb!dzwgN={LhZ&t}67PYEO|d~bTA6W2uwdhQxOmp>566xLJ1)ss&5w@0M_ZiR?Or(_ z(@VH*k3G2dGvCUVtR`%E{&Q`7XsK_>lb2@u1?h;z3pPPMi(N-RiH7A!wES=^}}P|d)`Gl z(MYThza_?I+*TQ?^%g-Pl-8&`EbR&(EqC?F96FARCy$KE0ViCHU9I}t4BO64_Dv6*l#y_tOMXI+J#b@&7<^$RL6=elgB&m+< z=7GX5B#f@f3#rdRyi=$8@cq9gNeocCttZbfZglVC|Ebgk&gbkGFe?yp@9x+RU?T|O0H zyHKx0yZlw13S}KjCv;6YC0PE#!ilN;jA(qeSlZ`zTR7m=3;o=M>Z4OC0*^wBMyEpW zyF2#FQ8GJb%EzzI&{&{%Kk`osx{X^P&V$oD306SuBA8nQ{TZP#i$fzN*<}dt4tynj zRsruieO|o7ACK?H)kD*DsTj}g@J|WAbItL1_EbEcJ41Dmc)lYZN63SUmiz$Y%ryQ9 zbrFq}I2BOUf@kpFExskOgZBgbuMnT>_e!L<`4|M1k>zo{QATC1w21DAzdCDO0B-sp z1LJFt@aFWp0X;&4z$}q{GV|rRxfC4=oP*N-nC?}%)k_QV>PZm)X3zkGn8+Y`yU69oVNzN=8K@u2swB( zx}yP(r>FkeXE-gLVtaHbNe>ld6YbRrbluIH2f0mYjMIXCB(? zwWw&OWjXrqPiAeU%fB=}&tBzo*33@M~sff*nmcu&!;Lm2DiZIhf>ULo?(LpND`W@yDcJ=G;$7&2d!1N0=9 zlsm;O@%f>NzMY|cK8?Z?>XT&u6jvaPBZeGvXq*9!%`@Nq;Ef2|z@W#jLo7kSk5#p> zxyQt1iR{$N174r{XOn$%3jHf$J7PA5;T5+pH`bpjm&vw%(!H-ZEk0x4(0}|3+YY1W zDrU^mM>I$q-U(44^jQpox{2q-U5cEDdnO&o7RJ1r2P*LlbJz)-!PU1%I9IvIeM)J@0WI3EjfCE$s_1Zgr+X=ae&FSH3%LD>i4oq*{A{asJ$4ylsl!D{i}g&Qv?D5+LiK_^-%i!($U^l7)g$Q!RS`HSda_uA z)l}7yScQiTXqacf=71U+?>aJ}WU4Ei4YnB@5mwj+b-{T~SQ(Q-zYuTBRm8%L+CyKG zV!#&kF8G z#UbPxl>}2k(Ax-&xwl=L43gzo4zrP3B)T(RHvzqjBqk>sxDX4eOlmuWR;Fq;vWewT zK%!U;YWFCKgwN$rASp>zi}41Ky;rY})*7k$@C_3W(t6NXa;YmjGLYE9XT;ceG5|YB z?D?+SdzF9rx&=4~e}4Wmy1|2i(`E&nW`b!V=;MSYPr4LZg>-a^^3?3w&>KFdW*-bp z=4VT^r)KxWX3A3en8|DA?~>`KtrVp4@rD7f7V6aO-II|R4DTw5-W{QxKOrX?K~~DY zqR1xCG26nlb$q;%DJy}X{|=B_f_xDqx~s{fucQEl-u_7y{vg3+Ca}ki6$}f3I4l+J z<1L?se{Th5GESDtwXm{aY8$z2D%^*oG|-lZpix(RMvZ)l1728@fprztXIG^wKmCkFszeG4}@rBuB9C#eNaiodh_ELd16e1rfJjz9eml2U<;&ZseVS5+Ces zDfHV;^~9En@9?X|P-jJM2fYgvFfOFF`|q1sLDlg$LGMc5+%-I-qFi}HxrSFQu5(hj z_*v8J6uz8ar=IoeJh}eLuP(IOl4f#2j0bl^vQTVR%)3ULiUq6S0Bd5jd*L9Te4r?j z7?nwJS#z)=uy0N}=wUqcNsGJr!e*IvzGyksN%qp(YDsr&8<{D~nG6dP3SnU3U!O8l zcto@ds*$hFeB{?C$5Bm>;p&FN@LE`M(~Q40-b$t>>*I+(VnI8<*nHC;7$_V922B!!#iF~<>&gA zQ&*q~A_LSZTV>cWbs2D*3kp6Xp)Psy#dm>0vNlhqS5Jm_`he4dH`Xvd{uA3kvd1Eb zCBpsil}plES*GqC0M-9zWi;SC%O=*4m9Kr&U$G(q` z)BRqw$F3-_ryerx$po{4pqF9}3WW>GCTtUH;Xgg8p9YG!uI$lwYz`csV#mgF)m@yd z#+b2_!u?;g0HSKzAD7U@9=w)7%x(xE_7V)FpLP z(Hvj?Ml*zSp?|7u-aW;s_%3oy$U50a>Qb^Le(;e^~mf@=vguTj`vg2{#3%tn9h-!GUR*86_#Cl<3~{O-5EO}BD$Ry;VA_rS_X z-6WW61l>+(3dqIuU-}iGNlO+1Py8aF^xWXxD8~FZM)RuTHmjFU?-%Bi7wJq{+Dn&# z$rtsrE2vV4+TDz5iNEIG9#BChOAL{X;@jj2+DLWtHu~iVTH<%fN}{jEFBvy@1(Z(V zJy!YMQydP4Bf4Q`az@^z}avCs%NV^dHE#1rqvZVY}PTQp~-U*Gg~RQ8lS zzYCEIBJj6riJwnp1?Tc^`xo#&g;#~tS?@e`A`~mN&p6=C3R{;OHT$NH^(>nW{NEyZ z&huX8xebEAAS&<@hJw?1z4UgfQMf1Wo}yh?7g!VY$3#(i-1QkL$~|%C!`DtV(*on) z1_8EHUZn5GCA%s1Yx&SeGWTrr=8E&&7PRb-0`J;(y**b88cQc!le|@FW(|Ji(-DQz zS$gs~eL*_kuZBkHt!47#^r_ir$ajiGn;8uD-B4sfk9Uy>*soZ>?yi^~7{h7gxaj@a zhG9pJpY~#ZsAcs!^zy1II?sch=6b93Xa~UnL`q>71u}sJ(52P{GHF=;pGfzRSnq%Z z;}^tjLZG=a5m7)Q;EF%eDmDriQtc40H`S82lWnA_${edOGL&T#nuMn4&7!FbBeYoA z!1fE^TuQt{PV$=EwFjS+w>#4nw!((!G834E07&07!Ybd zOlXp+PH~mb((s)i2Dme@A~r*bg`)$(SAnnQ>Z}gMVjoBeVC-gRV208lxGi79!&_0` z!Yt3N#oUoO&oejdJaVaTU2#N6YMO=d^n*!dzfY})F`MqpMi5;a4=m0>c}bSw-h z30Wq`7tYmJIB%is!H@MEcQ`8k+wecOAwQ3e=CB0uuqh!y)G5YQ$RM~Ebx(-2Mo;Inl5l)lY5*K8SARKNhAp-~FlPk8i&B+jlimlJTMm!j9D$ zGb?I)o^2B8X_v->H&iT%C2Xqf0wCzALZ}^u{rI?=$vZ{JI5Sdl3hrM-hb@E?LyzbY z_5~_K@e@B=>Q^qM5grT}7NXeg%stqC9|si^Ob^_M+77ag4ZM2cRi9Gc4*x2mC?qjB z6}IJfV^e~AX)L%oPFD(Y_%(D%w1Mdr7l$tmKTH`FRmAR~HN0c|d+KCrd-NmVW-~I| zT)w$-naeXG{5<{ejyFz++4f_9NpQo`y^VUZyT*6(mW*2yw9osuShdArB{G8mB@M7rydw6BKf1Nc;|o=Ixlo!2CB&T| zn{9-;U_H4l5I5s}zC*PRFxZ?Nnc~~z+X;Eh&Cs>mCOif1PqzZKyE{qq8<~F(J(&aU zQBC-cz&tg!WEFUqb5Vli0w7j*)!_?{1CD*QdSU7h{_|Z6w64!inMz;hwsrF0)oO_q zo7w|{=^^NDLer=?7M2jvZjzZTh;AqM^C}<^dv#WtvIA%V43P(7&-gF(TOV@9KSNmv z3=-{Rha6;eqR&u?q7soo&>^lRGpLiYir8ccj-^xG5pADsRCZ4}9g_z9Xm$LSc-(t^ zD!l_*Qss(@@kx`4LH8QR`h<6NdQ1Grsw2Q{(-N;I??q)O+k~Z4(?Ztx^m{eSyU5%A zjgWB5gY4PGIDGEp|7Y(_;F?UY{c&I64atKc8^PoqP=O$_IIx-1zfOpYa3kPT<$1ss>b`AYGTrk8 zSfb|Z z`FL|TUB$cd>JGm=I@>vm8|Rg-dulQF16mN=q#F$I!hbHJa|5pAX z#V3lR$^`FTIes*L0$z>R4#`jqGtOxP+?6fb$L$g0YrG5Z>6PpLux_$yDgF`lwB6Q+ zn)Y_%EaRwS!teZOo*sqDjWOk9ha(UUjolC5LovA&$);3w(#()_mq9RYfqRi>mvj(9 z)CXVMMo{yhm7l2u1Iv4%i8m|cEP?*1y`imij#v$kDDQ}0owQ4;QRwb33~Q5TfDE8* z7TVT$+{ZknR(r|HEcP4BRQ~+#UybEno?Z6a@9$4cNWj!dYq^(&NK1>w!=3aaa%bVX z|0e-62L7f2oPA-%d)|5tVv=UZGZ_%7jVbooPchJQl}o8cX-Vz{*2)f&B2Ns6 zn6K)Vn-p@;Z`h%pM&@H3lL_pC%8vT42`TbaCqm#ID|$Gu4a)+t z1~rxb%mxL4*6?o-q>M4FbzYi=$2il@jK{w@Q)sL}_<}6=O=uJLzgGTwJCN>W_|(vM zRVUuKwrp&WY8@MxGm07bF=rU7UvuAc`L_45Ogc_JKrnvA zY8*^4?xuBj4LPjhJ(R46dfa@;b$Ox}{yamji}CnzPKhMdw@)@e){~TwT1g)il-*hQ zh~V!L%lVjpk9)-FfO{cqnVtPYyK5};0pjliol^5Vr6)*nV80yy&X4)J&wWAS;mH5< z7Y?n5VgEbcr^!lo838+11xYiQ`8HF`CW>qT!7ArYo$fm-MSMckxKrOcLhUx~4$2JQ zG_t<7y7QT2z_jyMmm1R%+wF#Ej)9p|DCR>7zE_oQz1HoNy8w9mc&lF9NvjjQ+h8v{ zH4t{g2e}2DQPznlXWOjIp8b-|wfEDc%ClqG;t0D~=H1JRqsD|4Y+@NZUM!iQw>aS5 zPJhTLh2$TG$&Xz?9JPUir4rhCal9J(fctV?{SID}BJc4@hM5JgR2S0c+_XXYvxgk; z?Z=CfXJ^xyP$z_Iu0d|>h#tvF`a{SDp`gc*gYM=bhaPD<8I%t>oRnqwr22Lf9Z3&1 z;p7TFv#^wmJ@}r&(A-y+fRn%Bf{$1f4NJP#MKQM)DA)Iu+gph3Z&Oib^YO@=rm1 zB|CJ`vmtV~WRF{n{FF2$Q0v@F(m2SCt_?cmb;z@uKqa$Vo4lVs8<@$B3D+oUWe<2* zs-7p%x}=Lvcvf>x`t7Dq!p& z#*`^oqQlfMRzLnzm!oga*1MPHmoL3X5**0^gXc1XVxS9KL#cLBUEdo1=GC`Qd2$z5 z>-8Y;EWdJsl5E;(ot1i>>sGnf~)xe!rhT zai&MK&Kc8=eVlxm{XS|XzilWmW+P$aMC^FIY=XN4m&u=N{**5n5y(Tb^k&Hit|M}F z``u9CE?0yM1nM$=ImvZPmLM^97OmFR9G@ds=OXP16nJ3dn;XzA?Up`XdUn>?S#apO z=XFn=tm`La*r7+92wU6*?$@Te8DZ&z@c9#A^2O912+xlfU5nH^n)Z;n$>fL~C;l!Q zc)-&XbCM#}pwA+U;}wQhdbTZwDq}hW8ZmFl)g^vCz$Eloa+#x!4cb0?T^Y`> z!%=??D69;Sdf_f^HjRy)_k*$O2SKrXfSh$bPHH#<5>Om#rF)@DqS3iTzGYSu2dR-# z-Oq+*gH9tR;MP$!G~RPDB+5Gz7V$`VvyQUNr`ir5p99Atps;e3Gx*}ye)n}fnhyP7 z{*TFNJ4VwT12i>LOcO;KaibmUVm3pWLp$9@9|~+x#PLS=r@&!g$;%1P0?^*xaq0)GW&@=(76`A=fxDRQ{?9uH+b)L5Z`5+Ha6yxC26Bo54^ zac4f!`4EuZ;#6K3cAnG%)nGf_CvH$w(Cs5sUQmjWD*~(m1ze>PNxxd1Tg`Ay(}1qe zVK(N)q=;jMu9@_=Q*Rp=z1wXf&_scIl^?dVjF7oub40O61IPs7A?&ZR%ymrqEt|Z+ zY*?5a!8Rw@!J_j>po1~n1)Ctqe#1Lrf*S)hcA;8|S1U!y-~pFS@>n1x7;xzqXv1&G zbGRqP)vynR3LhKXpffxycYc|0o6m6Y@Zxmtfq)92%IN3h2=ML}k{tl$u6(UmCXEaA z7WtJ|N3TZyZtEe^Y^;s(9}~Q^8u0Eo{&W9S*5P&>LYr8JV;wT8&ENOM0@INmJLtNr zmH=mDwr5rdn!S?KEVnZ2`osofi;js$v)T~gin;K3I*V8h z0TT|yoVoN@;~B+n!&DP<1opY$o;0c=G)qo6K`y0(DUG-X;RNJJfFV}6IS>TrqIuE0 zT*1G%z1c;p^Jgvm_BZs-R3|XzEe`hf9L|0`7Q_Nr@Qpf_ve1R5QSyiJ?rRluWLaS^m-D@>2}ZMYMo&T%TIQ} zgfMjKLj`|RJ3E>r&%O+D99kYW&>fT=a8HOR3)&~sn$396{;7TX{95#C!e{mA zNmf?0-zCF8|2t}iF`d5MIv*1R`LX17Xp(C!UCHST&Q$h^vy}TGuT}+Z3|o9r=oG2e zQWPoRni7CVU`-@WK^jG=0%#(l7S{pWmM*=bOYdQKpC$SmPmy8rl{KJja_}jfU9!G?in2Rie6Z`5YZ)yFMg%lnVtknQXi zJ@z|glEVh!S3bo+P@F}nZU+uIG&v7B#CpTeupx)7Ur3avyY$MjWEZk)!hERnj!`y5 zu853LV#&)c&`3aDJ}hymfG(AT^P!w2EY|yWV2sjcxHAh}6U0rng4%~Y?|kPqJzN&Q zf1#CV$E%~W+7E@yY%qm6XDXXO>KRFsG%$3cyoPSGXth06)Jy=Vr>87{mT8x? zf5W{?LOq^hQf~g5Y_#K@#XJM-?4X!rifpA+tASK+aA_G04Y^;(-ff6q=MH&a@@{d* zZu$i89{RXPrT6;C1A(~?y!d`BX=R+UJxD+@Sn6b@k$=7W19`kWwatNE8 zV1Svw|7kR(zf#JpxHlz-X!1f#;Lb5X>~Pcc=0*qV<^`xe<@KO zMSZet&;`3tR=_Kmi+~N?0}0Gm3a-2gf*MWCVc$eKPHd3#!yc7;ZmX9feSp#6tOe}4lNE5dtne46J22e~XdAWK;YqPf@|i>&}wb`fkqGoE93dd#}8VFenh zGbs9Nu6oe?@)~)U#IsvC+wmF+5H)7EFr8vF6iGn|C!MZfHfSuXadndbgnC1=_;6h1 zbSzP$j(4+Ui}UhZ?(Lye4qEt}&w8z2Qh*Ax#jlP?;RjH zFU;IhV=%XrQcN*L3MrL-5d)>O8pS5>Pd#w!8#b)zN3J8<1yDrWPnv*10TtwMJ3KzD z2^c+Nq^+(1FgSZ4V)uNUpDuem@yMQA8gUhpwyAW^{L9KV1s3F%ldO<>p)~^;wn%## zwT$y>Jjhsh?B&3}{N{OA4a}aqTB^S1mWxFxNQ-BKRjN7Pt~n+)XB=x^!_Jxf_0Rj? zGOnYx+tL9OmC>i6G8(H|@ex1dk|2U!iyTEO-7HCwKa!0!i=mozr*ezi4R7@c&u*t! zUM!~pDhY;Uhn$Z3M=jj~TIiTtAEELc>9WXF*3ww@)F+PR?Q#biMtGkJo$)O~D}}4( zpl2LcFwr5a0ErcPW?cVPaLDuCoH5{17qrVYI{;&moKz4j#DPP4PuUDirGq$1WJYE-x*3+%ie;= zf6(7F{KSNVS@9%0aKtqhWrgd(u{B$k zMv5ItgMkfBPz(q^R8XpH8lT6)&`vtq<2n$p7tk%7E}&w^DzNk1dmfLaduFHlqNOnQ zyYgy=vMeZ`bZ}RaP23D%*vb=NbO%z2^9nt41uF$u3e-g(1=Z+Ox`%?iVm_&YHk-_b+m z^s=CDlKSxqQ|!1}soMZPtrT;EB3D6L0LVQy2c0LYnI>5t9TToo3vGzZ`QoQP{P2y- zuQk5W_r{8^*1gmJ&EjQ^kvHW1i+X$jPq)Q&9?7#oKL>bo&wJbot^rhCq}O{k$%Y&* zGx)QN`@p5#OQX2*YN0Giw99o+T)gxB zx6HccjIH9hglY147={zvq~F(MzqdK(E&0yP8F#AJnEuYksS2^*&HLtA_s@9f0rKPX zGa5;A<*alS@u1>h;@73*+p0)SIL%nrhT`p;Xw#yW0xo5C}p%!cjh4<$(fO~A;5 z`#=9=2_>sWz= zy1Zof^?yOSgG8Vn-N)OGUu<^O_4G%fVpToj-D{_=#}|!V6MD zYohxd^Z&q>tqqD=yk~ZyGL>E(IP9QXhUE%|9fln~aIIFh(p7$1p6wzMBV(D1j_GqP zx?ga>_s%I}bUas*R3R@aq+I*;7BUG)1KvuUyCDcU+d!+@ z<^A`k2L^0009Y_Wk!{|Zo@CCi^g8z?{W|x=e_Z{T9A{UcV!tbR(rU0=Z=jeuikzcV z70PNi9bF)F1M+J?9kz?cuD~8r!HMOlyPbA8mHQtOBf$pNt{~fzx-B#lm={n@7B(}; z(FhCI0Q9DS;Zr8eRci<*!cT^rhuV(jG@cG$;_55FnS;neY~ zCD($$TWOu^gB}F-FKTAg&}yi^H`^HsB|>)u*27MMx}38Ywz_t?7I1F4TdfMu`siT+ zkqIvSLR0Af$n{QXYrOwulJuNy0Tdayo-B%ihO#tDRkrLlK@Epmey4KuXW25Hu7gH_ z-DO$aHH+)Pt29NX(q|OsIZ2XLfp}#hy_w|OIi99uO=oEXbmTFv=Ps4)PJcK+GS+KCRWtXx;$X`DAuPKSx*>9NzOfqBf^tp)L zYS2PNDKii=EA~J}bQGY%PLS#MGG;=^G8|0du3vrR`DbILZ#EID9fQJzV0i*?>puxD z;2@VMY-`7nWQZ~Q!x|J#iUG+{MUm&QL-%6j$tZV^S_(Xp%Hv|~y!3eo0@b(!PzFLV zg|i*w<*0!Z z*-J4%m$7?P16n7z(JRjtz^bGHwzE;+S)*79?aP%ukc*jf8&=Bd2B^x@&dU`66N+nt zB8x63aIi2;*KyX$?_ab{iEhXikg)<_Y?77vhvpYv&->ON_7Nj34ssB;IxjDk56Ox> zPLp&mT|eo@@RX%iTb!DH^v0ZFW$ZEc+4?1`fMjNd_YYRg(EAeQu9ClyCk{aBOPv3x z`t^UQ{`*fq`^~$56tAI}RTPPuo_yg1Uk~&Tj}0+4Qp|datff@Vk{q%6k!(*Ww3TKD zblT+6p5WKgpE3gvlMiBpD_G%SM)kup&6B6>?Kt&nB4xiPG*5iQZy%@u!EAtAG$W!= zsJ5c33vZU_noMyA#}-q=c!o6tF=Ijrcsbr6HaKP)G5^-p@Fq|1EdsWE>m*rY#|zvH z18*!tLj)hNnt-@*$yx+#u)1P|unN2Ni^w&-L)-tnHT)PcBOo-m@4ULfwIdn!fY+4>(;?P;~Oo;%?$;l-mOK z3Y#U}i?>1xYOZJ#Fyn*w>4f!4v*bJwIA+teL8VM9`82eEqbEJ+3_anr9B;>Te1+ty zFavgBtdMA%GP1{je9TMB;5_CoR>+>Yw*T641zs%RhI2`hB6`loB*lOvSlj->Ugfk zdw`@Uszhp3_P}C+3vK}2bIzMXT7(^Pby*OWf2?(jnKeMvRZvAxMr*t`hS<{IObNf1 z!*kL*S>bmE$6<@oc*~DX?%sYM0AgZ;a5vp2J10hdojnWsg=Ky{M4b}S67;zw7#8AZ zg|vrmWSS*y@;%(H&?&jON$oXyfkk*4z1fC8>4@ch#t5-nM`FU}ZBFN2#+p}kpUVnC zHi)n6lGo2upW*7>G;q&xM+gs4d}Y7u-8p!9g|980hu&Tm*?by;6?Q&k*3BPyJnsle z<7A1}hJ3);>YOP6)*NWX$qUuVNL5I31b8lWezGKuGbG#NmLn+RZ=Aff*DBAX1qiXq z=T0^VD}+!*cfWtYnBo4pTJ4x%y+?WL`yTf_+UZ&+jRF_Y(F;}ws*#i!*SoFYX>P+r zayu01*`}>9i9Kc}v$l24o5swZ8!eX?=$AbXB4r85d|`Hkf#PLMnbJXufmnAR@W_Hb zdQ@PYv_AZ57!;PY0X2RcD690z>go9KR(^$KNTyL_0k3(EcsNw+d_h$1>6tIC;A*rB8hqQ9Y!X8m$GAl+et%D`qD7$ukRQ0?sa+1Du zLmWk+6kwwGE%t_#vdk39-RN!+bW9ItR;n=E&JFSkalt9tDcR*M>gDmlrLOX8ij zm#wwb@YXc)cw?D%CF3m8DZfYmbTg)@w%ZWa1o8A}ZdOOxH0%&7+B^HotEr-L=F^e!_Vk~POW`&WNkGsF!_q=0dG#?L{k%O{s zT1Or_?2sMMMPCcTIRiO~@qK^THiE0$cG+s_w1C7SjM&>kXXe-6cJ#>yT zdG;Z%)0|$=F1{|V4Bg|Vb{AReER? z59TF%uYvxY6`U+@?L4bM&KwAhMknU}G(v+#Z|`_C{>r!?+-_4JCVIbD3O;h`qSrYC zzuqd}LxIO%*C;9_jS+=m#U3bVk}E3nThG9n7dgw%bK6EvmXD~TV>PtRE25`jdW&po z=1r_7j~VNJ@(rQ!LWWJWYR9vWi6u@GgC&h^@)MrGci$}O7lNe5Qx<%vC=o1gb%s_Q ztM~_VY{nKy=8S2LYh3@ay>q^P8v5P7EC&$`|CHthXDbwOEBKFH4wav#5<{!nr+ zcz~2fAO%h~kW(PJziv>guX)&_#`NrC4XivJV^+c!@BeG@?bS(fR}Z&(L2I}SYN8Xo zuLV5_OPn=r9^~_H-%L<^e*JX4*j3u=CeXW*_X=nIg>10n2Ao_2=dhh(U{RYysZI+U zJvtV5Dpzn$DgS<(?Drc8j`c63*DrZAr)}}0#TVSQ&hgSDmpZ4m#hv`=GQF7wfBNuG zlf_vDA=EDg2fzBJ9v*vqI1fn03p3MPHvm&D#hj(cDN2PMe8?i8+ul>Blg+$V`r@4Z zFF}KNo?kw+3m<-2ts@IS;)y{QbrabuYz11hVZXHk^+xYifwhtjXWaW!XMzr3yClx3 zS%^AbaA^all9x%RE487(x|RkVH1wZt9m8*x8gRjF%5^cvG2e_`r6pJKM%w~#gqEy)S z33bi7Zt*fYZC;TNEO*W*l4lpX?UUJxKEl*@%)-*9hOo(TR<35omD0ay!}PfLc1d0u zDP-qr?0CHcWQ}7`bc|v^E?qlHLXjx%CuKsU+CynTAfs=9ey&V<4bUd32S88ZIQafx9_QF|epEqg1`}jm)6yli&I`M35;qKy>k~8o>nK#L>);*=1gC_2TafTe9${EQ4>dYU0q2C%utO`|8{EdJ@=uhP z(RhE9pwx4OQKwe6#TOx)7BaP2aY~?@%mbN^v1hIfYVQ4ku^1eiES(+i2oxK@=yU0~ z%3s_NU-nD`{$MSXFzuzw-3Lfh1QTTnJ2jviEE`bbRj)@ zUZWUXT4-DHG=8T(rCz-H=9(s%V9=bf|E7s&-j^1$FbzC|%NDRPZc#frB1MTu7c zEfs!WJKSq!gPvW%J&S9DdgvJW$MVCWY2pHTtuRe|8d!`{!ZvXWXSMPl%qoa@K(8eI zpn=v9x%#c^K#b5Bd0W~B&A!o&IO4jHH9&=s7_u(p7`Ic|G4cwsBAmUmk;Y{7Gf~?j z9|-4rxm|QDsHlD5TnE1gW=@(ScM+6rOO$7X+*`Pwvzc5Psf(`T*M{`?td`c9vAmv& zc_w?1CpbT;Az4iy_WRhbS+PuQylMSx{gWLB%OfIVRy^ z4t!1CY{L_#;rs6c*Z*j|TVS_4jwbT|y7Wl-5-3~p?3Jqvp~gB&kSe<6{%|oi*26YY z4kwM5Dney-Tt@Asv%DXXI|~csCrA-!@D93k(V(*q?SxT+E+NwVwM$Uu&=w_6<{)GO zNA>iSmC<4vA>&js&16EG|76UE@qz?4CYUlB+)IT+KKg$n8zY+{wc*)e9ZRl@V>#3A zYRmoQE9pC1w8c zoLp`{DfFwAu64xhOP4S!q{@GtV?*R|a&Z1y#}(9Ppc2*?X*J-O3y7!gW8znnTH4GD zSa}ud(;xl*d&a_WY`lscuexEQdQ6gUC&d6q{&r0AmC>z=oe|jEp5UFKL=v812W)#! z@Lnl6=7IAPl09WZV4ngYL#`OyF|ZnNOxw{iNSPkbT4!8)zB}VztN9eOxhn)M(l%)R z05z~=-v{s(vwlggwIKV{BF&~VX|;xP5OT!#rFbHhz6z<`GJ4qWA_3CDZ9bhSy9-xq zB#_y;3N_zl+)g07+sA9=qTY`+;ba+ZOt>R!fd7)q7@G?yG?w})iM6V3&L38$K3swh%UsnWe*a*2$Bd`c^S zH+?&FGwI=V%G%`lT#e#TV4v(`aRIl2#eqLvj2jc}*4db7HQd16D@0=1!Z1|H(1}v2OZ<8u8i6-q zolw#Q1r55Yst3*>4Ojr3t2T~6&@7hOe2NA&<91e_W@hC3o72^LPjfoli${9wIPDTQ zb8N7>awaoOk;jxO+A)o)db@bp#@BXzca3AB90f&txka8mbn%GY>2%1mDR8HeC6S?q1TF+ zHAYrS;(1MxDc&2IsK85f6+iavwl|}`p>=M5yXf_uROLIDzuvw~qiFbb@f&?_Cco1Z z*~-u6-jW|vYLv&nTKKD?*KaS|FsslLf1~x083FhjH!<@i_#3tI_r7@pR4%s>5Ds!~ zeEnhAY1s7KC##fn(QQAy{!YuAjX%&j<6ngItVQ3BcC35@KeP6So2Z^Qeq|X5KX)%S zd%+5X)YW56&ya?h;Ni+p7JDR1V9lX`#i>S#a@IPf6*X2EL;*8y$=vmv2hRDF?!Rep z1eW;%GoIwBU&6`^_PhM}^0|i1Q(1r8adOYZ`WusV=bWy*sy;}1T5PaTCay3RSr8G4>=AMFzPyKuD(=n(8boPt0_30sY7~#5nNnrUowt%UXV16 ziP)$m_^|KbNZ(SXvI+S5-z^k===-r>hR?ejna@ayyvOIr{7=NyoFhuqR7U!!Uiox* zM2iqIfn}|A>e$ksWywa@Rdav)JCbO}(N3YkN-2|KV0$r@QtkKK01eKuoJ!AX$;}X~ zi91a4Z$=(Xx3ulw$C+&oDce^62{93aODp48oJi9<3UAX8Z4SO6n-ZstdW zcfar-dN}QIUc7+Zv|~688NjKVV!9}Dmr|kQfW>UspuA1~zvwwK#O-_bhlIP0(CSGnzt2LrUh*(a|-497MJ~Q59i%`Y25<# zWg#}jwSa&M{!VI4P?ty&xw)bfo_HRff?UD!<*NdF=zZKYsB0~9G8FB6F_zjhv=y)a z-I&;r&CF)Mi$vJA9Ge}AqnM2pSx>2Q1^HB#)6vC+++D8Vm|!z1pPKIZY{Zk8<+CjC zIa9r5d(5L3kIJ``{_Wwi zL*F`0szrFsD&E$Gx5bAzx81tkKa^zp9ujxcRlM6$jKh(%D8+k|b5g)j-*T@Gc@>cJ zR)*s^_}-Ff4oqtnYNjyBS*nO-dZznTdfS$TdOW?>lCy#2+VQM*(f~V$DW-%XMY?8f zlv^Ds3_r~@K{Zt`7pi8KCU~tRgWPMv1n+!mn|}pRKYS93Y(yy3+W&IcSF(!CAf|#HzIpcHoa2NDT{O3q&>pjI*)4e*FN)VquYW0Jb{?JW{GrDsy2B-T z_Bpb{ZI$n*!5hPeWXF`*5lxXO0g%SIA;$)oHu)!^k6k_sUNO5pbVbOvfL)*s^sp^ql~#nUXUQi z1XD?!6oPA)-1)Ep9#G@nNwrCqWTw50Or0$4#nr|Pk}JH3%# z9jF>$){mcESM+j$bA6;{vW8Bbh2+VxoFO-x=HuCZ$%df<`!CPjJ6Juj z_M0!u|F73L#$KE)d>1LvS;fQiiDE838Lb#pIQ5>D$4YWb2w053$UY@RUXSeHB z4*DJ3x?GMWd(f?<=Tx+Hh-xON3?n7oI}7es?DsCK1MToqmwjxabhOAh$Pkr7oaeA(%t#**-6Ns7ExF(N5m$8Q6@=O_WR z_CH*l=$tLVwiKwJ>Js7$tV3^HhWl7>2tOD59FXb^M}^fNQt=&8R{7;g@+4hA7LY2! zPsaT!-K%D(O~ekD#96R)H4%I*Mc#N`iS;qlz?-2YkN;^*^YMb@PD~JhBzV_*?AJ*-h4`o%cK|Q>6oFXyy+II4z z1Thg+en;nkuZWQz55au_jiN?Y03ISoo;lL3G9**4^lhJqol{tKfm_)Vc?xq4YYu*_ z=_$yj|jhoC4rNPVhcM(&%Q8i+Mml zaeu(OMUS+<M(c&nkc4`BK5cz(JOC zJ!&ZyYH1YL=cg#5mae1LFFZd#&u<-dj2j~zbX`X^NUQu~$?ed1xF%74jJFq7Z0o2( zp^o-feSLm2gD)YDZr~)!kvmPdn#a<@iJ|^E-liY(pMUg(FIZuIX8f1^%6WQd1#J7) zNwUU{p_O3(tt5)sLXkL1g(k$pm`?gWY$RqXd!$?321qKs4vb&GS--?ggHHIZvN&lz zFjx#~g-h81qp#$x1Ae2xNd0F#$$DYp=Hmu1DyA6dRoG9duuXT+^+ZS7L8rWd1H~`Fv7EJnd%!3%?10h-SZRaD;-MrN*(8AMm%GV3&kt%w z@3~@mX^H@uCam4NU_Q(ynL#bBp)vu<+W?JQRKBfBv>u51?OOSh9x{u=6~2T0B)5OQ$~>^l&Y0~ z_3b7oyIK`^^=)J)jt_5&+@`FRX@j8giH{u7x-zRf3m?Ihw$r;y*zL5(twEM9IzhU8 zlO=cM=`QR1byqe^KK1Ae1t3Csm2Qli;$n~8;D$bu+L5pg z8!pe%&pfyPziWRdF?PHz$~G|mHj07Lz%8(ujY8JIcC#*KWgGiVGyMEwMKE6vn%6FEJx12Q zFl(IM2G~!gn5`5^pj6eImY}qdEFktSqtB7M!Yco*(A9WOyf0$FB_-s_t4QWsCfww0 z11y^Y7CVfa8pZm@nM+@B_0)so33EDzXxM?njw9Bi2H@CBF~GvN8+RVdKxs`IlYa6M_%TWyZSKd;|^83?Hw}FbBeomBg`&6;=~Yr$f^@K2Qf7rsSt|$X`r4vDdgZZ z<;u)4xY3h8Emmh07|krZGH}OurTc6rfc?f##Kh|N&pA}=*Z+L|r^U-wQEB4Ux;O3c zW^-5`_2(kbM#u_ivf)*$SHd%AM=!<1e#s%8XV#AK}sV2&>vC1~V zJC3J*>Q4UOlraWvrV(-K4{N_`>@DneGBM#9aN@Y=wj(?%q}=OTP`eB&1YESvsHZr{ zz3)5H*@1J-G$MkXXBcBB8^1J-gK~|AO$a zJd4I$46a_==!PXb{O&9&lEslr)7i=K`~hZunLRFVwD+e{EabN5>sR(Zxq4cxm)64GLpS}=~ z-Q#Mv1=QbK^V1)^rB8ay?ETfpql+%iaQG*V1GMb5*1j2rayNMS!lv-CjV zkOOM}V!brB@QoBU<2Fq%|3hZ4@RqD=BsG-fR8tzH(E)C7|D-F%0@eb%>gw!wy2SclU4rhRzCVN74?3E4R z#Iooz=q#!7YZ2a(vfApK(nDbMOicdP{E5!=LvZ^*Sdwwiv)#rnCi?MmQP ze~seKqT|4u+!TrLDkKL*2$NPmURle}Qr=mlaUFEMIOm>+jlp6WP|!zM3IX%(V+WLX zzV(aD=Y1vA%5Pg-!d1tLI%ON2v+0X-ZgbK?;yAUm8Y@2_E$P>29gM)8~R|mEE7KVcsEm- zN^54Xg{W}*?AuU7+%73n^w5QG+z@ATD(G}^8NY|ddg9&mHvd|_E`L!iYJ^^yYEdsR zu_elnE`wHu5r%1?Ap!1kD;tURb2`Q>hA*va8`G8;FFv%}Rr6cUKS?yEJ}{ZsALlQ# zawXz~lujTumNk=MzR5pp(iwcH!%>{o>hOWcyB80 zH+CH0nrOeNlRiy8LrrWNv=k!|!nC^*w3NB1U~AQ>UZ3>S18*Q&C zqCwzvl47bU2&PnM(r%~NplFY+b2LhI8n1&JwX`s-O&;yhJZGeR?~!jSKNZ-DG2hw? z)Y_;+zfTs&Yo}5B4kZwQ)C~{D29^8oc1;Tzko1OSD0{-s5lqnHGANdVY7GOD3*PPY zh+-bzg_kuc1|%9(Dk2t#SmuJm*g=^4-cpdT3Z`cK&-?##-c=8!UtS~cl6X6Y5)^Ka ziNVt;21I~TC{+dj9Es!YUy>Y=y8tS*Ju1DS%N&SF)&`XMU1lzL)^T^!Ih>2ozje#K z6*L!aOABO$vn)0|3&U815f11@|oOmPa$k1Gfhz%GWxk1$HdivTVp9DIjT%?nr}Mo7*MfI!Ea52eB2r7Oy+b zO$*#CxU&e4>#nV#*E+t77q4@)p7kuUk?DunQj@@1I=o8=S2Y3AXftz3)+PKy4SiB^8svQUMPTD`uN>(xw9YY(r`_PX z%Z@p3{2|Nf&Y}wX2y7y?@~?;XhHY_blf8G}G5M<<-|Kt>J|)TPiTL*7lKC59lq~^m zFSB|$tc-utPX?TM2ILbQFpd9^{|rlh{#6=uLWx~;K84k@x_UT_cXiy!DWF@&%8V0y zlV+V_0|T=Lw9-YUVK7bxW#(JC+ut`+?4_Pg;{lH!ke)#{UD+%dbit+vjrYco#967c zGzxsBOJiiuT@8BcS3&L+uN3gL%wuOTl6Bv(GKu{rwbpx4ZBCOFi`Y*K~qq70dc$t$a*HXVP(2VpL)Q zG;`6-Vo0l8dSw59*S(^LMe|#CvPh8~uZWArwlva=)` zit%m=xA|Xl)-dXBrzajQL0v*zL3aw@&7a=}ynjHKgY;ZoG&;e35ow|uVl`am4jo@} zh8;4LH{@v{E16H`?WS*q_VbbhaD(ank^dcVj|K7Ed)({u``ykfTgV6SIjeXBk|OtO zLHYA7oWINk#xpL@>;snJgq_>lGxSE2P!FS+l$*aM8|@fI(D*k7Mms1bnIc;$RmUtS zWWzeRELx3yd}$%M&Ol|Lop;PbeN~w1Q!LxbWGJEkCuR0A4_!QCq0QM?#4>!a`2@?Z zWrdHKOMHf3Hr^62kxRb?6lZ_GCBF>*x+!ujEL$9{R+WZhn9D_pUh~^n{H?|FPG$($~Z5>1VIAuF)*B&$th-Wri?u7c2YN@8tfo zKj!?vc=f|(m(h;51WYK}#R8~H=?fx_0yR4}GHD?N?jucypfiu9w5>34ec)9WRUa z7b{)fuA|-a7xzn>ZhF4VD$-5K3^qk18TxLX>`~Vn8P~7y>j&tWxKovDk{u|uca_F zXXYC38gl{4o%7ul{(4{?=qgxAvhCRQK#|Otj7$l|K%-FsC^*xdU?C9e^|!>)<$ILOX_Ma-rq9a|sauyK$y2RxmlY}Otc(S7O+D##tX$HJ-41U^7wM7no;I?Q z`{cio~T>}z5e1s2#|4rh?D}k`%d26o$NYW?< zNL`Y#gs)eAbaAmqn;gwL?0|czup7YJqF*jyUIyzn)0Ib{g<3l=nO%GHL|^H|=PUuj z#3NegtZA?FQ^g;+|5~jBbyD5df_7eESg}X33>qD``RL_<=&PI@@pcJrFNw-uK)WyKMK%;%?%ur1sLOeo)~=a=7V=l};s`Hdzk0LKesAB#&*~@oZ=g zHpVu*)3)zSJBV?LU8rBw`L!C0gWGMo*M!{n3ThSE#NEYhmub8Y1bnna3te#CkZMuK z$Wvg6f~66-Z`glmzcI&I-REn+vrDVQ&@1lvCs=r9$QJ9Uw5QD;<2m+bg)| zahgPXsB384Ca9C*5CQFVe z_6pURKrn>832MM6s3%5&`f5*jr!rBVG%sspt+kHA|E3rYSqizy{n6CGC-RG!|GZPa zWGZ*K*9EPfvsYFR_U{ah65!;uKJvatB`48&pKOnFg7V?wgMO(ZJa&R?l3(R)WNKYH zmS_}dA?Ix!09kJOshM`NyIFY~N_Fe|zkl94-&TL)CS*#QBF}r|aJ4~2bSbqxV%tll zQ~{Iicfzw5v`}05&{z%sZ)S41*X0)^ZL$+k%aiMQ-Mxm+_IxP0J4bg*Iu5_pQvSy&1TyALG|k@Ht`({jy zjLLE-+18J+gVKlpIrPWpy+AKh*}o_T6r9WGP4cbK`-a+LTCWS9o8(peRz6m^C(82# z*Ewf_iOg)F?#Z6mNzb#!$D|`#=K^;4aQIHwX-9q5F0%bSHCe$f>1)UNyi|k0CZ1wq zDe?iO(mLPuz zwc~mj$nKBXR?4FoAl%KQRO?B#e>Qzhh@3RoZ9nYL@7thgr|US4py}`+EGguGXxO1p zR_8NPj@9GyAh60mNBMA3hy2F;b_p(?Pja&9`@S|@HkpSB(=!YkJF~ggv{(7bC;!LQ zn7qMm^D`#s7z*5rJV)443n8*lZ{Y3~c7|qq>NpvuxENvcC68QdHUe0WJhdB4BVe4Y z-Au0)Jtt*)ACfxvx4$EacDxN;Xy7k0DF&)GQz_MX|3RR{?S#x25a1#^ZUU@SF>N>O z(8}#0U7+~WMR#z^=v#7-I=vw274PD9^O4A*hrh=&(Z;>^X5-{pXE8M(Oma6XoXq&m zzsF`6ZwcA0_b|Z^)e%tSlqkntz|zo>+*^Wod+0`{S%QTqk9@nN>DJr*o#e|OIn4$c z=8R{pYuF*f>)XG7GWmY89Y-N1_LCn%`K(4!6L#;_vw@qu6Xkzj6V&Z?)bGHO7-+|Z z+S|Xa3Bokm2LV0q@P4JBhaU^QZPU;v%?37OeTuc{f?I;IsHKR$>%PrD28vcr_~}@fG0{@w2{nZlBoM% zZilVsJOQ5H>0KIuvCkPr@@#NN_X1OZdZZ^@y?SY!)7{_$eqA&kLcJ{OBv{PFm679^ z_^lNX=Up85<;m6V>^L|vp?0^4cia0zsMP+*NuyZa1a-3mlF5rmtnhsr4`G(wjc=NU z_on^H?|=2jyU)AEL-dz!h*t%+fANMG%A^TOf2`!476SVhFO4an%Y-MxbG^%iRtgu) zvX)sm7=7B!8qqozna065nKUZ*y=?** zq2_caqz+2L&;pNSFhieJulIgw-s@cbLdC(?VI_%v zVUl7XcR6M^CXr$`Q)CmRD)!jOv`bDv+6x#18x$ax3k?@o4~=wz>!xAL#lWoQ;$nK& zG&5#A^f*8N|I_1Q=9BOHkal*su;UPP)y%QD=%bh(ihMFEz^q;yvcdVY;2633R5-Ls zxKv9T6q%sKv70_fwgjD3)Nuw})cd$+gs?7?CCVSTsQ3F-`V6{2ElCrnF#kMx1SRA@f##sogo@Y4P5(~iak1a@0mVxm~!G)edB0u7lC z?gN+dCEG*vbUe)xUE?4g8g__xErU9OqlzNWVF&zgI%Xt`jBenrZ1aXyq&wpa*P4T8 z=v|X?SIJ+<69-lU*zrD2iouk%kz&?UWG$sar66p@!j{0T4ECN+U-S7>i<^g@r`J!{ zt6cuo4C9oV-MS1DxxO@x)_Es=(`yemm)lR)@Y*=*IQhJEuWZhFX|&^q9;XFs92+9@ zB|GM&(I@Gna~mTyipGdb)3HW2&+NuigYAva>Y1l5{jx~lp?4ZTK0l+8B-?R99(HfX zM6S6MlTDEfN|gv({=og`e}EesRH3Y(Yv|!On!n%tv$J1rd2`1*_g=g8y)5n>a+)rO z+L3lyj%SCyX$grep%32k}{`WxW+}jcXl*6>c0jev=5y7 z!+;{TO}@cd3s%9aQ+-o?M+#2!mt`tb1ovLjx@8d<0`^s|a?-tq9dr@ndXnRV!Cx=8 zU9wW}scao+vF!K6|63(in3|zR4QBjv>IP(GDEnPx{8Y2)!yoDqBoJ)*Cb?wC?xxQG z+MN^wWPmM{s!!G{uaF!RohJQ@?z+Utn<8UHeX>qjy0}T171%R7AM&_SbCc%uy^;m{ z#y#XV=rwfHO_A%|ZUy6YRX*1+V)0%pXo~EiDC< ze^HMQvfegD-W20?wV{`tvI4W62i?{RCU$hTc?66fiE-ag_?)nM1gKXebt$2GqzRpt z?IQc_7-{DWkXAu4T8bQ|RFI^Fe$WQRliWW7SIF~?N6WwSh>?36=(`K!sPy!-PfbZj+ znxUEO`0EpcV?qmke$plo`h0ZDU$BDE4DH3d@Bdj3p?7YbC#08MuE&m3!|?{Wo>eoM zsF~ymrOM%C(JP%+%&wr6mq!&qkh?c=CkpBno&;H8O`%ls> z!4F*U>_HLUKl<5tWbf>^AeWC-%E93-l;_uW;61$Ob!hWVr^8Y&*ql z!-oJW{cnaO$}^zl9rP8w)Ws6qi30AXR=(D&l>q0Y;AY5LsNBF#BddgQ=C}yrUh&d9j@9qY*+5bq3B(m+R%`nxCZ8fXlr@1?W6Yh@33ca^I- z)`%jW{YzPW)#FB(wkd&Q^~l<9SdhNHFXjt+#QgY9eF`~j$5Ty{0hCWsObtbjQz~rl zf$bwjvhM?^5Ul@Z`5S3(uKr>1GVB$CO`@Cbz4F4aN=}=+d`Y2D4Whjxq=oUbi#$iR zBpQSB1drYBEYc`Y6C_8OJo`gm^~11s8D4?n0Jw*Q7eJ~)cQ!)_(g9eQg*U_o4RULx zU7&rGA2!Ifa4NF`%E!n4n4`}=TazDQh4L9W1q&*l_on$0V14@MEqU39j9aqr(PjCP zIC2?;&eu6&b5T>|X>w;#I`=HU{f!;(W^vP;PQJ9q83~(CzLf0yzmII(DLn<8Om6Hc zT)_$_RAS&OZy6VL+U+)li89S}CAI<$$qqT;R@9(dH-EaeE{)A(7AA~unuhnRFfnsD zGS$&|B4QKN+HV4sn3%SjID^njQ01oq>dAH*%WjaF0~^&!K=LC)2@_76JW9|XhPfaN zb*o5)|FA_-4URs9J)I-&C7=2XZ@>%~Cvek~Gnu7*W zO*X|q6G}RzI^aIw(iC|HfRHG!pnG`R{2u}zWhL(t@0NSgoKHfvULTN3*ee}-D8P&|YSh*eQ-!}$)+gR!9x$2KLp?=jOFN4Wx#q6|@Hu)v@ z43~a-73?VVJC$)(kyuVsq}F*8lO_grp?NpLH;Hye+B$F74e0ZCR!JxE!7M(T4IVQx|mQ)pN>S+#~aGNs{=z z8<3<+9Z8%`x9!bdyO=hUZO)jUKkohU+pGVz8WGBbo4ki)*_>F>vAKorRlxqV%e8=W z%l#@x`|7zlNjxC3q!0U+(dGUJJ@*Q6azWUwfK3ZnUVLNLjaheHw9ZYOW4s}UwR2X@85J8r(X=`@ zl#oU7h+atg4m;eI<^ZSeVTfd)j{$laaCb!AEJ+i!@^uXfg`t(6&|Cw(6UyGO?0~h- zd4g-gk4T&xm-Pm-q1o=pM8smZ^}iYF^)5N;2$tHC<&e?G#ZAFwyO$ zRIU7Wx=&o?mr19J%An{aQ?P-HEhi(a4my>POn^=t4>k;PMaKzJ({*rrL1uNIERNR> zy8vob57#>5jfG(ty@HjYDo4j9WMIdq-<*Z7tO^7Re~0mIGwqtjA^w=T>&rRE{^RnTa8Td#2;f+TYf67{@j*NW^O*>M3x4C|OA! zKy_3$_ZF0n9(Ajsdwuf5GDYo{uYN5x)&I}lm%ufZo_l)+PDox1*$55M@;q6&Fy{C~j<` zI`=-dH<3T3?#q^8XG+bx5EP--5_#T%vr@#*qF zXpMK#$}D z6=bsM4X*W|BZLX(R*(C03jE^9{W)o(D8UxLqrhi6UUZB$Gv>~*LXx>bY;gWj?Dtmw zb4Be@QVMI{_JkedteafNTp_s=%>|;LM@t}KgBgG3kSFU~|E@L!MB0O=CX)l~5;=Ap zwZCX&F6t--DB2H$nkwm|8x;C~C(84D*M42AXiyYFBKZQRi%tx#=Xdd-{E^OBP_2p- z^z(G;zWKf4MEOp)GrspFI(5CY-?xjt^=>0n*Zs9s{trw5l+G*i{hMav2Uo=%vMg?Q zs7{SU-q@~Dx=^$1HC;G#*};b%;bZZ2;4RYMR2crrriF21sx`I*O<%DfS$64}*m?Sf z!k0Pf|!=BKY?b;9tx$G&zo{FA@OasE7f zTci3@K+rk0Ef=P7;pFM*wb~hHxBsX!mESEg`nqOO3@|;VQ4zbj*gqOIc^fZE(!zfb zs^y$iY*x1LGr1eQcL|rhe(z@w!kdK$>7LnTQz0NRHBk`l*eUIqeNUcA@0q*I(Uz|l z-pAPAu+?7Xr~|Qb8twOMH|a0e*ZUimr-IJ>D3blcEKi4x@LobO@CC@HBDRruC@LxN z(5X*;aQfflspxNO6?NpW=pcVbSQ2MDz5Mmh-?|yzBYXt1XKi3#j*<>p5~qFI0p%(2 z1Nx-msQ551P6R8AQ@%&(8#;9k=LpGJP{TXmyA>2!n?UT_y5ESge*vCag+>MPSWm9~ z+4UycfSjK1`qYsqJI?fgqR@z)&030CLy=f20>xpDkk6!%^HK8dqVRz4WAQ1{XG^LN zbFbU7X*Tz??as5ZbCWt9m-Yl2K(QjRYY*AU&Yals=D5xXLgf^*pCTpLgkQn$6lPD| z$z3OCmn2KDM-{aaHTbVoda(fGD%HHjQLrjw?3Kv@f3 zPL!jLQDJDNN(bEnnu@^T4+){UBHUxG4o-upk5#OTWoUcKfPCc%YhbWEP5h>_0Y87y zoc|+-t8NuKL#eme%5h?&o*dm2U;pPQ_sKaq2)Hhv7S}|nO@GRj}W|;%U#ORLPWMre#ftYqgCLO4}uC+ zozl~uI`u};ZlG7F@r)C7N_V6yR%Hq>BYzE`p%jTG9kg=sLdFo!6n zih}n2h*Z(zXD0;@y0AR&Ul%Ezw zi?oUu=L$MTfXN|Ply-{a-0_}Bf#$OCvM}AH8+QAtbosQM+|Ce`;cSCyC(HsY7wibp zDzI+tfO6+J8(J+wj5(X})#t~BFY|~Q#js2&pY*qHnHH^>srFdovmVmAP!^Wwsmb?1 z-N`=J3M3+)tI3|aR5~Fk+%Mk@o45bvLlbbo>h0UFz20`_{QGYhj7xYazk^i2Fgv4b zMy9BNVt}u#4x5I~lP(V|j{HZr9Ic2}aaX

    LVcw-7Rm1-OdpwG$NhSIQI+{7#%Eb zKvwx3$~;MvWdAFAhUY;C6r|`SrSpbFfQ6bk;Qz*S^zpFbc{e$oFz{vJ4)Fn{z60sF z?5xM_fO?q*&6hrNiG{mpvA&RCq0AUsui1PxzS<5w>ngkN(|1l&OI!??Dt+^vGh`Jz zOxbT(5?E(O25eF&W)lUo5YZqUaM-G@6&<1T!?Ga@)XUV0){@~*6EDKSHc7{}Bo*K^nSlsm7SdwOm=jmJ*}oq_XZ zsDe8~b(v&AASI1gIuBi%A>CHkUj^M>u+$gx*{~!xf`lBd#+fSIt>ZZ8-gMHn;mw?D zVof#O3*}bFA>JFuS>rQ)^$V5|dX^P=a(x`|!AH;j@;4{`xv~R2ob^8Iy)#8s{GFtY zRPl3X)XVz3Hcw9z9riz|h!e&5XDlcb>L4%B$l0p8CSJ|#5Z9=#OsRLNhD7!bLA3v} zIV~h5xKNnSjdnagE1$c}aU1zObiGdpuybzmyE#rD*>pm!h#b?Q$Mu;Pb>?lpVQE@; zVdEjPlHJl|$ARGOMoUvN#cZU=dMcui+t16U>!npZ2>srkvxl?t<)u@4p{76GTay^P zlyk+oO@5c(p&EZ3oy8Nw79182JkA@-5525--YGZYU%Mv$Q!O~62Z3h;IiI&eY;8R` z7g!CpVLP4T897DkljPZ`PUDr$Ij6>0T(YE&6q98xTAvzTrf1t6o4P1$=l;o>VcQl~ zR@*u1z=9uFn97H<@nP)P!!;u;-m|ds)%NggPR8_FI^q3QZ&!!g6eecB$rj+p?6bDK z;IaBn`F9KpOU&=e9*~ocdwB#!+k4c7=ZuNm*2@X2Haa z{4QJR_}Np+?u%`=BEmD;6dO##cWf+<9q*dVL?JQUf{bGNtxsX-ZMUITfhNfn;d-Bm zHa-@k``LiB?j8THkTAo?vfc-)TP_MO0E@vT?niZGXGn$$vp}m{CpK3BvtX}=f@8DcRQbkIlD4}rqc5Uar{ zQl#kRMNXNBjjhGAZtV7q_}1Fykd^cE(hqL`<89M6RW^%>9s3w&8d#63v?@Jppy;ha zh}0R~L47W5^7TGiMbh-aprJ4=hG&OcT_*xfEUd^#JlO_9Mx%c7^rpWl4OV1k>i;B= zEIYO$z!5m2xVD&Lpp|0}6|wjAYvL+?3m@gKA_Wz~bIeixX7?IizPw6x*gr1t9Md7& zuc{F0)CC^f2;Nm8Ojlu9RGa)E(JGcZMltoWWv+MWJ)9WFM(JMP<&Hg2gn)8cc*e@E z=@?DRNMIO_jb?J(A6S_d>ew7Nl__w{%+Cr?>B$R6mWez`J^%8wxcLcC29l&K<9EnP z=hbkJkwqu@MZ86AoO9CUf>;P|AEnEA-R_{Y6WSV_rMMK(&wE7518>eL@x?I~3+~fZ z{94Xo|6HFIK3=uz3|Vh}tQ+8BH$Qf}?~Td#jWlg*w%f*dvyIA`9;f~Ln^KkpHfnJ2C6Z@VA~7Xonw~fS;^D`_jV&h3!JaYP2Hd`Y*_8J6?@; z7_CNIC?=UA8=>t7OYZ%y0gZ8r;(=`O%bMsJ$Dz_~uk0?8cqPiC1Zg5Dl3N>qI`K$8 zYgOZAZc16J+I1uK(s<%$cD&^=6F<{$cS`3Smvwj| zX9g}l6Jk?LztNTe$n=?s_#7(x;KZ*CtIA*3#Joao+VQHg%t#TlmtsKuw@c4Z{MVZ# z9z?1O>GD90p5h#7AoY4LxuWI4Xj!so4Hg1mM!48h>jNuJgIgs@;KE`1v|iHS)-JDw zEe~GTKhP)NB?O*yWFH-LNZ?{CSr)x)=$2lBG7rU`>-EoeN$+v8XRessOK!*y^0Ypy zgEe>}Ew~w!TodK`UwL=x7oLq$AVnSe1K`Puzy|4M|BW*W=B}8$ZnCC@#x+pS$)G`P zti_<6ok{CY|LrEh4-L{>k=@E*va?1ZZ|^%SGC(h#zoD7L*)jA$gKxz4T1zo06xl>Y z)JqDzA_H21;s7^s`Jh3SNyo!xu5bQM=M}yk((ArO(>E(uF5F7TaqfB)iYLhSSYmCA z_skPvYar6YCcs9OIo@dYH2~yyyKB~vZ7)puQneAB3Mpm}MRrmVF}&kpG4q>+$TXZy z-*rmzujlu@3Gxl|n;1=z(_JU*9UDFYT*}TZ&YE~$H}JOLg*Z+Nzs#@F8+*^PxVK1( zGw_CW%dhe8L8Q;>TQ2q(Ge+E#Pu9iWmeYR!{u2Xg9^cs|Ay-H1UbW+zl*dNcxI-~e zS#gt!=mL@P7Fc^8lD)o2gK(NOfuSk)-m5yUIwjugTT3r@{8iCAmm`)g%zL-)yxOeWb-MQeh=Cm8J|yQu4)YApHNU>{M_a#J^ji6Qx8F$m7kE|&B~wGs zuKwAL_v+p^oIOHdwB{du@y_ixHoVyt{^h;MDOB#gBCHo^8gyz9cjT9FS58g^7ro!N zR$LSo{WAXTI`v|wX8GO8>zuWUqjU%83hz~=bNh{rqV-_)1Yl!58VOAErt7o+{*@!o z0Jxb+?;j=0UKle4!d4?zy$uwzjw10?1Qsita$2SaqDv@QfL=8^H2`E{NWRB%!EJfs zq1JnCXY#V%})Xo?B32su<74IkMxr1+)6{E985zm|t{q z8pQaHg4WrFS?$xhWko_wrNVvgn*~!Q1(*Ol3Tk? z1ehtJ?c-`8^nXZoR-)0V2IlOidnP8FYdS%e%(3Y+6Y<%+aTuke`K9X_7tVj%;AQ;2 zm%2DBg)~8_RM>oZ3 z;&y&Ngn62UH{=+3S?l^(+yZyz3O<7WVT)=!FWtL^pP?%DeC(M?LkAFF1NqUR=et9( zofShy37~bBrK%+GOoKET_&T!bYR=WLc&CHP9C3e;p%*PvmFcl=@)hSR!WQ5O>IiC( z9&z0=O@p;p*ujtV-}>P`GGB;g2^h!fu3GY(IbiQU{oDUGO{xIIgS&0h2e9s{g1$AoXlOjE zdNkIAg1KoO%XwCIiP}By``<7vz_eT7m?^1@B7Iy`cF-IS?IKmOrA%Z%h46@1954Q5 ztIoJ3vtcPJEE&t1_pn06ON)A&cLf+$tDE1tT|jc!tyXp%o~bojtx74Tm?DK##9{w6 z+{?n^X}yvvI!<(z-Y&W0k;G|}-;|zpTkQ(mu72&)s`v#S38JL|#Z3FOE}(Tes?6qI z6E~_a|0`1DCdd~HDnNPjj8_+w`CtY-o6cADa88BjM~e(75LpwBmLkoPWyP9aaercY zkZE&~-C7hgT~7z+7OHc7`+37ZAyqW*zGoH4Ow|KX;aU2MI7y{dw7pqD_w!Ks8#|l| z;c7qcNoSLukXt{(lTN0;X!gB}y-a#a=3|SV^g0cOr1~Ba>=YNx+N?Ynt|#a|3i)!a z;(#)G%2G}~_$Y_D`S81tk{KDBPjH%pYU$5@sL5Yw5wNU^7>i(K*>;PSZKBpriBmrB zoT=?}2DrfxeQA?lao!c8srKp!xhX{|7mFs+stW^8vT@p{)yb6^mVPht?G(Qhk0K$+_ zGXERKc7>Ipq4pmC!$qmV&>a8fyI09tN2t>unP%BaF{u>UOhrV3@^%4~6zJ5Qu1CCf zx*ZX;@GnT(CEHy;SM-pjQj`qnq7C}oO9CRP2eO2)I@qk*=8HJ;?Tm3QMh;<_TdioDZ3(tl=1t zYZIiJ4$2C-FD?J|rtKaEE7JIr;*UrYJ9ov7(?=yn#xRFsfX#aw6>)!BH+MBqvY%8G zdNs@YT(DFw0qUymI_1vLpfkKf&>&4%h#$3zPHCU~K2WfppHc`EsTkhTDwew3;@i|c z9s{yQILW4Xu-<346-heNRHK<3UFzh*&}@P6wD|xP}6wLIRTp;<}bMr2MhZ@Jx z)DX;=GFV2AMjIds+RoHfxN4#pb6F%Zmn;x1s7^74#P$>%eyK@Xi zdUD{kTAS!{G{ z1I|eZpMQ663!!%z-!|aDtEcx4thkgxD}i_x{?>;Q>8-w z!UJyCc(IeaxtXfI`A6xL`4{C{MW!GJsJg8IKDM|T;i{THv;yMH)&V!(gDU!pukk*TVnksCcqkfP|MQxv#O zKIZhvM_uB?;FW5;53u!m9;ZyzMPi&k!9J$HZt=k8qe57wQt9{7hyOG{=B550OegKnnb}!l)baL!V!9~uITcX}3B13x%KwhB z!Mh$=DlNB#bn%MBH|8EBRjN470NLidcNQ9x<&H(-B3aD*)nNzuF^;I}+ZlAwAI-~? z^YzkfXG}tZr_9_Co)tE_rMTeR~}GXcPrb2kTT6(GN) zA648MR5`apwcW#%Rm6H&G;g^qykZ1P^H|*U@ACUW2Hd6lWNaY2*!j$MT>fyv$govV z%mIq*qar%}E^&5{QrC}5IPIS4q6WnuhR^78N~b;R`NNm=HC098#<@CB5$K|unD}{_ z!F$E&qJCbZdff-2kG^#hgRrqfge*UMm)bF=YD?i-$hE+6$ zZ38VMl#J=a{@LD{ra9`TH|6#lQ~*umk?Y@=TAv%`SoeTD+_>hIIpI8@%iJ zTE*3n=$GsH8@#RI7F#lN79!`#l48lAR=DTbpYMFzl+k^!(f2T$VxXlfor<^`c5Bv7 zag2HuxhU+S_rP+#8~d02;{Dh76oXUosdBRtnAGQ01fCGxn!TIb%uR6rbV{qR%i}Vn z{A1iQgKhh=S-;{5dfDc`oTIp(_WOOSog}E4W$;fHHs{2WTsz)M9W{dReu{w%Q4tk! zl)fs$+Wkjlvoi4weYx9O@JaFWLqqx75tkBfJO8kxUvyEVQ==GP421BouA!fo?4lmhv{sD6U)Cr4H=8TujEq54l>?2-I)q#N$s4&?WQEWc8d=)w7NT;W8~N5 z<-E;KOL&EfMkv-g2-W;Lbp>y4*b?5_(0h z-x_WU{JV-iC%qu+gK+<1zs8U%8kl;ulm@5>yL`7gCa^s9BUU%ph zz}t*5`ffSMM7Ox^3)nfMg{wi8R{TE*KnLA~AW)0qQ$?yTGIpF0YI-G=P$sd;^#ZUJ zw(!^Z+?e~xC!b`~Nz*%24bnW=L{|t0z(L*~+%4}8t))@r2y1IHRXM&${u@JmrW}}4 zCCl|~hrM~F|2}aZ2NR%}GT-bUEB&Zcd|RHz$>UrPMwP!f4hmsp@^mw?C-Pa9Tb6*& z+`=@@^IwVJ3tg+`xf>wX^vZ>g$@&)tVg*JJ%b*x&yw-vioFG?_Grh@IgLB&?!Rk9r zHmwDSAQ%!tE9e?tIZp%J3iN8`fa;)fk6;ftI78*VHYW`q4@|ytp2vCd<@I?VdKzHz zqtlblk>uyh0p4c>hFppPTB!^w;?S!o*_XqKcFgw5q|-pn=rTdQXZZflUFNuh9+;EP zJ;_h@TTGRUdP6VDlLgBio8|X7_576JOhMVy*vWCAKap%pmZ|YT!j_vU^UK!lfBZJg z)YoN`BC_L*q8Tr^=asZz-0Of^I1;~P&phw6P29}Y$r2}P+Uf34B$Uq&kDL$MHW@06 z#5F1=q}60>*cqp-u;6i@1HW0wHJB3zKj@8!X1D6tZ&CxgL`LSE6DejbMb<#gmFz*# z4#8C@dBMiHV}cSTt~QT7@2N`NYq%5b4_k~23+B?Ax0HSyuQrtrU}HJ#7#F*YaFIqa zKv{~a#yWuC9AG9xQ{BUhV(GRNM|K!0pV%}|h z^%aa9g@f;zbHZVbWOMHwC8pj3&)? z`N>vjc|owU>2EEs{hPrv{q9NXEppb5i&E|x`4p`bbDbhrsE9Lk8n1Knylc zl}g7lSDZJ@)GAPB1UYZ_C?Vtl#66j+EgpTY7zn{_3r07ZVf%YW+7`YTLLLLaoP4Eo9eL+&ZQ) z1cYYX;(3{ntHyKMxqH2A>HLffS!QuD{tK>~zu|3~dwxN%VkRqHs@%z4GsH=_<*i1* z#yNElq!M-N-q0Fufvihf>9@fzkE>I!bx!eq>>0-t$?kJ^gy2=3`nqfhX_cnZ+0Huk zq0k$1ONWk=#AK_(MmWZ0;53KUf1<<3f-*do+FkO*;4P zo4*V*tZ5swWoe}Ng^9K_8m)jwDCQ7Fs;Gz~biSz5bBUn!oAqdG=GO;y(e<)EFQ`!9 zqy>K=1BsgOU399lOMXJAU*J-Fu%#N9lV`O-0w_|D#lO3D^r4yQc*Y{}z-(DTHwn<;k2h;B6jvW3EkrGz75mp=rT;oP_BHAocu6x{PeK zW9O*K2pR`QFt|yL?dKsxTBHblfPYdr5aj1Y&qav3naR zgBHpL9q`#)Q42p6SXUv5m?&?PB{MC>pW#0cjQ$4GmEL4 zP#3(6caG_z??Xi&Mv+tLo6^VgJEs)Mdc`<8visuaDp~Z*y`mF}WcM<^#j5d{EZd-K z{414tbk*OxBy=_)=`Wh|e&8bV(c0E!m)hY5cRe)zE z^_cH*#aV+ACK%xXcc)E`fgLQ-ZG`%&EGxDaS%Mc9Ahu*UtKDNB#9I_+-gPyYjGta1 zi%7x?vwZ@Wc!cFkrx;MMPJsa8v=w1Fsx!(wUbJJUut>IBkU=N=EOXQHDGZK`Q|s4JrDjy^!eiSiAhhgDkD0q8{8M=!?;%)|bPLu2QP@V5*)WU8*inwBWP zOxmDi50ej9RVcS!LC1mK*7ab`Zf+U}m5f_xlw8M+S_|C`n#LfL1+~ACg1eZ=fP{yp zjdReyh2KtB(YxGX1XZRgPqK&GCO_hJ!~>%N!~Ye74%p$M$)xc=0Uv@;&-gKE1W>De zFptT7lFxqQXuzMk{bLPTY{y|7U?Cr|L?uv697R@95m*F{G_kn?g8_nr8nQpcMux}i zHy`9RVYgs`^u$+;V4F+k=YKSt)SUF$f5{A(NS*rY-;qQ+#zdhJCNe1oiab)Ohz4n* z{E+81SL7=j>OxAT&w30cF%d1X++x{Q1{W~29hkgFU7Ew89Wta7&J}HSvMqKj#K~yi zf7<}WFe*Eoto)Ft4m99oMPk<;vh#&;hU<(_RZcPcDN=$xl1L(^Q)9_@6a2eghT`Y< z0#5R)_ye+bKB{ZQflN7;kyi*?A>V$$w>eO!e#F@!xB>U4ina$I@ZIZ+&ks1%^Q&k) zu9posTv|{?FC7v%?}mQvl;C=qDe2hQCw?4g8x4V0bDUz&&WnED*Ph$Fb%8_iAdC=I z3U>)}1xU?cZT3S8Cf~wYwqPXd-onbJycF@0|Avji>8+c)|tt3X;;-#aF{>xanM^G}NgJIFUSz-DZd%bGC)*)Tuy; zU%+XV;<-#wyKTcZ76NNjQ(;{w8IPu{>G+x{XV42$3}J?0sEgMiOqN_0UZ+u{F$)SK z%3i%XR1~38H%(0wM+r^~AIk4|w($3o9;Y}?C9eXytUpzK%Fl!}^)}Kj>iPOlCHvjTO0|YyzbEM(7PUy0S9E&oD0GZ zSzhRkSxro%I@hIKre=->UN&KtGA<{*+0^st1o2`(8 zjvnXvFsz-Al@&><;d6_FV?FhE)$p>N@oX$F8Sc3*?xIVf)1g(~OQM_>yW{hlB)5Z_ zVo~k1~vzI&K!*;tXKa(YYg%~hc;E;BbEPu{oKd^v}*pF|dnDrD%pdwJk zGf|!nwV+zX>adjgnpigo2g#9%sxb7p`(1vgZ5J$4yTHZsDz8xX33Tdo z@93G>KDSGlN@q?$=MEdJk8oy8zA@JS_%CF}O)ta3($CbbAX;|Ljs52332;@7SaNnz z3~=OaN3M_*K|54TRROQZ8XpY3Ae+Nc@nN3kgfN@#AyD(lF9FdtEYZ{GJnx5I5ys5X zL{3SOuacKYf#IP^f`)T@aIZ_xoCMfXS*7C2bZW;4m=V~)L(@^m|9R%^<9DVQ0Mou8 z=o(332N*lfPn8-0W(UP=r${Ey!-F7SFFByN#6W%uw74>9O)m7bGy$6?w#eL5Y9a-j zmHJ-XK9^5DI{oenHLF|?tJbbW&F?>Wdx__nX625 zmUf%qHj|(Hf=Lq}S1p@!0KC*}XRYFf{2&jfR;R%3z6NOqcu<(ZMlVWJ3Lg^XBOk$A zE|Wgjv|!EfX;8H{jFmfMSBOrLD9;h2mS_@FK5fuJ>$5tzipGYh-Ljn_?fg%HCVW7) z&q-6@5ywdh?(!&+Kr(0fMW)CjnSWe$ z3Ur<@qqUWMYI|Y~Vq9Z@M{28gS z<1~1Sk(<*D~#=;vjHAWJ}zkHJP z+EsCfOwW3M1vX`Iob%)_n#&6>13@ev_Xr=0uX_yNb8TUL_$PN3#<40Cv^r3pK<4Qq z79e7^dQk)Su08fPz-sBoQ({RfJ73$5SFds-5bma!Jc{IC);!to9yl%7*BdD~C+!4U z`gqO-Wx6O|w1JDT+9;%+TFm?`x?Xx?u5A!9JfY)2 z%J2|2`T6#bx9@n~E4LC}+2@7TSy-Amw9Rhe$2#Xg?N*z-Rd#M3RDR}#clsem#baoT z!wRk};F`{&`{w6(oOid?UVM1E>~-c@twEDK{Nncib8)NbIVbXXg~|-o9=CJuo%Cn& zjbD!;S4DmE^FmXV)sjN5XvY}N2A3jn33nT}K)%e;wl>6osIdTQ7>u_3S(Tw~`wxSg z()OMA-Xl$R?EQXWw2a)Qn41)7ry}~~$$Vt0xxrlw(Gkrl6^8DxN5P4_Oeq;i%kYXV1 zypxJ3q&s9a-_R;>kJ=DeC8>gfxmJ0(cN}LsR48rcHmdOz*3yaIo4M!IEquJDRg`e^ zLhm}YiThmJzz5Yad6HyFCtab;@+p{CFs~hS2fnacnC!`$829Vezx5{rTIqbl4xknUE#8b>_imU{?C+u@+;HQb8%d$+>McT!&(9LVBb>gcg&mu^C)c`iSW*!pr z)_r3!&#=18OnUz)S!TyOBv4r%G5s4TW*tT1^%^^gP&Ak*zdJRNy8%g$t3mjpnzPzn zJHH@cLWBuMOq@j^u+i!@;?H>M&^!o6$xpvDUlLF~$p8%1wvs=Seg|Mu82SEJQ_M<= zETri~*GTW8dh~Ss;A?GKDan$D}7szU6BS$CK z*`R*q+2+ooaG2FiAUku%k6lu-z+hDxzrQq-l)f-|hx;^U?2T?_ zOwU>Zvr+H)cEL}5_y#mBys+^QS;-DfcHE@`Jq9Co+sPEOks|A<2&~=C3&monyinv% zYm;k|{W9P~m85-I3)Jg&K--BtL37kYnmN@iMoY1V-rWfN}4 zyD2lJnl1bW5PzK&Qhgw%cd3wR0z*8I(518ZcaAzR&SgRj4_aXUda|`LqaORZwAlVKLxP2}Uy~_C8LI)f+ z!0*|n9B|mr?Uq+g&g2d_Y-G}y^TJNoRKEmK*;K8f4LFf>>iE~LiTA6@SXF5bA%P{V zOl^wApV`80?rVS0g&J(m_eyr8k;3O}D-LD7BP`1yiUDf-a?q}wy49y&9tG}kYQSnH zEjWjR&GGl=#FJaIyXBFB0f+l@R!Z_=3x*nUs~PMX)~S(>eE7+{&^~VD{L*==eU=3D z_^qB1&%5K-z-gCl4BMf+F3zLzi9?=eT&g*>o_m>iURLly|4wP96`G4}gcY{Xc~*`e;4zQ@b^7VlorYeYziV#3svEV}r#zfpo6nc8Zk)3=>QtJZ{PLHv zIAAqv_PY^Dy#L_2r+F=d4a#;pPf{FE5Zp!Dz`v~@D&*eoTRp2wdO>&#lJW!Ihto{zjhE(OCvqoy%Yl(hTT*|m1MD?BK(HBEj-#Ye%=XL zZ%|(7j<84MGucv*PRdf%kq-4?xPCM|S#nIUB%ln6IF7>S0$2q{K7pFKx!m5MIDh15 zC|j63AFo>5*%=c$OjkqvY)xpuSmtfbeZTMh!8Gk{CcE78O8Ybi0%HDde3S!fqsEC3pedO8P1EKqPZT6Hr38%^8b zo|@nqP#w1NOjvCqtUMFuCt?&|(|-4%KU@CBGE;)!rfu0QAII8w6GtGce~~2f?{aC+!+y)O79S8Rh48{Xsv7o^uzC--^2_$Knw=b*Pxi= zoGpoiWe?*sSQ(S!dw$B+X<9{+{|afjv(|OloIwX{T*T}85j3dlgaO-JL9U=t(P*WR z9xDiq=A(>u&I-t41)-OYNB^N}JiTc4yT%eTy=WP#V$Vi}R+S4~s93X*DDRT)_lxDN z_c`X&DX-uTqePSL-NMJ4_40y^id<(63ag+HR5b^i`eU4JwPhMVwyXf8QHH2^{hPmd z*I;mN&8hyJ9J6DC(_v(AE>g^SikyWkkSmsVrUj#Nu|Yr*n^?C)fy(*7j0H)mn{!U7 zke8tW0x?L`KsU?LU5^o*g{m0P)~cmZY^yE3jTb0pm7@iX^`e{(g1I;AkNSA)hZIWS|5A>b8vS9D z1Q)I$)q`ZtIl(zWr?5tR9XMPg=isMxlY2rJ;Q_38c)o()DQv{t`5HMoQ8nTUsAON% z;@skF74(f0Xrr!W)&ShM{`#NaG;OoC+nVz~Kg(ABt+`ik`81mw!8YgF*^{u}XQ)h>>+Kd4X1M6vX(V4Bbhs7Vz&WUl zVbF-KUs)6uAS5 zCSLZ;A*Q?RnbneX?onBx5ICT6IBlF-U?q7(jsd@v_-_^Te}kV2O6-Gk(z!#|^Ma4d z@Bw@tPOO(z&K-KPg5L&_oJLNfd}%;2aGMuO%BR8RYd6RlH%PS#C|@=b;2rw(q5hP4thnHR&hb{FbuC% z^0K(CHfYCu+Gvc)G(A0x)z7dpP1O6`Z?$~OfRTMa^!*X3vtx|hGQvm`#poz}^Q)OEZzS)8b7RvV~XWO}yE$@R&iPfbBK zl5}!CSf@S;L<6bx9=B}-`s>sOlsAGN1oe_0vQ(PlgU_~kbqcdNmEI{nIK~>Er4W9^ zaj>4Tz~iteS%PCGQCCXD~uYsI;GY9~s zLXJI6f#)qS9A@7(s|#!)!)owx+-x2)J5FtXihta0O^q2w@sk0NT*wMOBZShu$bei3 z4I_(yW|!=YTfMAop5`LC<2UF~AS&~V_Rj+l>s#1&V2;NCOqI zPQF->N$>Ms>Z<7o!a}qID(HaY-Xh>1lY1@@yW8hoNlp??mo!^g6phya`id_+T8rje3Qaj2uIH@hq!U!KO9g5s?CR8@#$7g+aX|1OJvb|P(4k45>RTT?l#<=?9&3{} zQRsOlTug2-kDk$N)&Kn8rH331h*GzItRai-7*VN4h)STCIEt*IB6?jK0xSI=$~33| z4AtuQLwiCo|IT_!P2ME?j}~Hr{lMdVf*mHx7k$&$l+uw+tAib}ae3T3&XV$VI!7{#7gC|L+2Ty4e)40eylpZ@cL=>}ub^CRgpvek}_ z1yBNv$Xpgs3>q>C-73HR6xxSN+DzcA@(`x=kz@}l4s#FH^mK}d0d?h2(HK8IA!F9u$y8_ z?q+38sJ*Q#i=OutW+Uf;ESk3)6uEAS%LDg`BLziX6I*XQJTDfRhT)qh@Fy!wOyW*0 z+vH?0B(-0w`UhER$A)C9ks(=6F$oljqasjQ6q)U@oifS)K7>+Ei`w}s=A|)J)^4VG z_B>hyi-}$?;Lm0S3o7c@U2Ud$?-#`JFq0)eF1ySbDjQW{s#DW5y9Q!1N#vpQo(rfj zfIHi$Xb)MUPHRL@_^>6h>Ix+>npVD8!m zaIuc(!WIk`+@JV=BF>oV7nqR*I7(6kF#^p+Ve_;Yfqq*sk&JXt{$niq?mfAE!j4#( zcKh8DzS~%E+_bOXZo~R!I`vcO9@!T4Jzo^!NpQdFn@v}6SH4o@y9GM1b?S|xd|Bb- zEo#(DIwjKXoJ|`94p{bIUT=Ln~Fr*Ek1yzY{Z^7jbgS?BpIr8 zy^Dm41*btBs~6k~Ji@NoTKc}rUV@*17tduUr~{rBHcD^5T<(TP6M_=f5SbtsmK{W1 z-u2msrZv1~N@d$=>=@Eu`Sj3154j65>g=gYr4#bLM}C_uefJ}unTTUn@7{jbm4AGc z|D9liYbyJ1%@(rTj<>icjqFAx#lUu@l!`z};R<@AsD*!U?j@!XI)l=9$d%Rlqs8Be zdHc`VAMSs*^0h|`NZ`k@W2Lf9kk4Z0-*)z0r!e6Z((Rj+P)x8eC^@3$y4cz!gQDY z(C4QARZ!}n9qK1Z~d>+_mHzer!7Ee7NjiDNoV}`GSf0AyLD*Hlr$lE z_!bXTsf-kylXl7z!gT6vWz>xO5Rz}tct!oi_M&|c#Tir%67t^9>F$e z*jb%`uZM@2GPK$)Jj}4MqI?O;Tt^9#r4S^+IGPR^%ld%}V#3fcJeQWDV)*`v`pF6v zlT?R(5EWqXW^R7#b^*y@7ihC%CZbv+-=~yfiYZb^MU>K=!Z^;EX@`|s#RIyAH{ek1 zmJVIijgkV75^k>~ftx+GTUH&mLRt#_0xh&&z2YoHLjJ#k)7zTagVh-1U9S1*$C=$&MJ9ZqT z-DZTHM2cBUku_ArGGK*|g^J4J0DWJ*4pgYi{M#i>k|h6n{zRJ~i*aGaIk#k}%Y&ndq-3-*NjtV2=Zt2*mSSos5Y~>sy8Xq1e2;75gs>u)3Mgz`GOI(?2DSDm z0@y}$vK5Y}MVa(@;XS%DXr-f`apBf1tj1sNm`ZnoS6a=9aqQ+|5@?0vl36L92mQ|n z*1JZL%N)Fa_~{ltj-2JwEvxX^1U)7={`eVPEyl}8dtuqJC9p7ikNFkrALQ46Fb(V5 ztD1jSdv#N*A>`k0fWedO#B-2|x zP%#oHpsDDEIpS;;mP=xbVulJ^6uadcXGB2@+iva&VYeKT)Qt+P(uV?mDEl6C$WXP( zD}|Vu0){<#t^gUVTKIdE%b`F!SA+@Fh5?Y>>u+#-&z) zdQY9g8*__2(?pxyH*xSr;J7_1#tR*^-L_g6Y&>$@G;6@7*u{?BVl&APtUS;vP-Rg}#so*Cqub(W4o0Tuf?{@si;QF4Mb!ig0XvfLvhelzR zTNHDHBCS-!jk%jV?|DXh4mfP}Ocl5DH_aPxNc3)C@JLf7L8i*cfc9zG(>@i}ku^Rb zjgZJKnO7w171y|BdXzg?`fpa85uUKH!O+37^sW~cVB0f(B|M?#Q1Gg)$2ls&DN+z;I( zL5T&;VM)DolV2BUkb>?GzXcp={HwaeaiTtX0?({9vJ6)y2E?i#M@OW2T#dF9-y-^M zjm&_x)TzJz9Z9rftQ8tzEt6uP_&=43z$Ty1Ly>SdMX&*g4yig@*L!AG zPhe<|B|5A#^%ji63LTSvBsnpEJPAlUc5loiATyzwB^BZ)`sfPkNNO4t*ed^s#t+dm zt~hHdxD|Y@Dn_uHL0JO*{`#hLe{chTs}J0NQjz9_F+vQFKx74~f@@?4l^Aasz8U>u ztJx-&NPFf=WdVpRx@DQ`s|%zC%X0ji?_MQq?bx#HF|sUMDF&L}H&YQYju)Y~BQ3a= zzCQiN+)IM9{`*{$C!hDcH7hOnhI+Z<#eg+FxuUyHcjYGoQUh{l7D`shPpD!Xtv0#l z-}Km~-TZjgy28r9Q03F#Nbxp+Wa-CKVo9naNQ#dvr`}C5c@)W^B9i@(Y=4XQ?r$Mo zB#`1%di2SMxU-R5yL?)SJbR{Au~e#c9duYJYvUC7#glHC&L6sUHc4&=p)9Z_GdSI) z54E@w<@-^-%R;7pR_|tP6Z7=flGQJ@;AU16W4~d}uP?jRE;1k{oxh=(#M!Y~0pyP8 z4A4?c3Pm;tlmI{JT?5OA)-Nv6V^JJ9W}Xv^oLi z1Odbzxfa0A3b(LAE_EX5-39+#JQ`TDi!{0C+|#)=&RZr|p~S4bmNZGaNSP{cB4WHF zXW4QXjeN@X$1i_e{Ar-Ua;!+~+Cz5QvE`^UvK-|Uv!5a*=#Zf7*2Wp_z~x;>Zq0rm zyC?+b#G_UI*9S5r(YPYqA67?JOnxA%2CB)MQZyerHEJQXKzYk9_b*_7l*Vh7x5~T4 zdU28q;n&1}X_e!Mz`y|?c7?afPlKK)4r3k5V>u$8UIxvau?k$+8J0Gm&Gn{~8f;RB zcDxxfL-BNKp+0Zj$T=vwAkL)QRgDy93yRrE3!Jl^hlPyD$sNww;=#@Dyf z{h%A#>8Dc{hqXBCpGi_?GY_W>IBZs&40t$2tLXPkc4+|{r2l)yf-;LxV#c6Ye0mJH zVP|LF5~?r%mjP2htNkFH++$ZEW5>0dYm6#n7Qe&{P~;I6ffCDAk^xyYukn?E8S6O( zoI*N=b5LB*$CkA_LHl5-YUErEYg6Qh@1j$c`LIf5bE=?bGevRi>-WUnWWb?F9L@B5 zZ}Hd`dQx#p*+m!92Z1oCN_NSAzp9MW3*60VoE*+>?n6irq|yVx_Kb~l_0p|Oy|fm> zhfjVZ<~s923+Y>)ZBWaS>zhGu<24H#)k&Tu+_TWNtN$wp9Lk-Oea=rg2WnpUY`v_{ zt3vqL6PskRIT>_r{V3--eMyFSP_keJQ!8kdHR3#+S zju*T`Mho6viYcJTZYl!#qt?%>2rQ9rmTK1f^hypWFVD&qU=C8Nz>At}FYG^rAC>-D zT>JsSTlkUdb=3n%c%R&O6Ig<7hCuPda?UBpc76eLRM$OHp$`3@Aim3LQF}@3CV#_J z4?EjvdcKsMuJ%gu&*!X^K#e07-D(x_PEA4xr_-8MP}&&dtcl~D_i7F9q%9F#u*%oK z@-O2P3u={L{ER`^O#!gd6X!oh{Ngtezx}Tdf4S(7;$;-Ggd&j>%Vs&sf6~9C_V4_; z!KVDSX}c$BwPTyI*vN(HrkJ}F>7*hGJWvkyvJl42p zS}3T~rJ-~ji+~HsV{xK<_#$vc(fLk;4oIQfOrRDQQkVKO7#qw`?f2`D#XD*6H(-Nh zhU(UAO&vKit%b(JK?fZ1$yi#SB+o?o;SG>U(9@R~TDw;leS5SHUsz=?Zv-Q;|_NVeVYC|w5IFszu_>~vbFL$1NTt4?^%vq)SotM;n$zd&-P zH~HR_#)0Niop6n)kS=v8bxC*00xQ%>)fvSlsi9!J$BplJ~xxiR~K$97@7a2eZg zefCp7|D%hiDGjT=dThiDv1+R5*6cO%-hW&E+LNP3HGVGqw+-JYUpVl+b0C3~3Z$u3 z$|ym8Si0)w927Cer*!IC`Y5eGt`X~h+Qwj5A?ot%=^V=)>{cZIx!YYCg9GF8sOJ*d zXvYqW8EGUGt4<2r4R)i=>xkEo8hVqYORnv zjeJfMbB|s#?>aOzLiKpUD+K}A7g;iOCC5P5TuUQi^PaF{5YmbEAN1{#X4A=%L5EWD zZ8?UTb45*(P09*o2fgk4QSV#{$C4oI>0SH%RWx0Ef6+P zxN`!o=tU30+uyhnj(=3H0E_+DEm-4+6^ktr3robx0KHVd$20Ap%fUgq8Biczm<$fB zrX9*eQRzri2vQaZKR8UAmdGvhJt)r=K`$QkMPUi@`6+u|*6dK`Nsti<_S#?8C9}hf z#nWgFa9RAzSloEw`dgau(1zP_{@hFlot{P>`{RIpv;id3jzVgwPke}hPMIo62SjfY z<$ZL7hZDgG70{sysx0I(5GOr%n7Jd&SF7)+y(Q8vL8@mF!3( zh3xX6c8r$>qj1h4iUHZ2aw?+OvrWE|9QMx!3Pk9-_XeT8HqJr+Ve&<6g-UZ?9E|yg zdZ#^tJwV$y6v}H3Jj^MZQw0>D*v*ejCebqo9rSKfgY1LItJG69OoBR=t zLpMp2RTM?rjevvScM%yy)pnR|CmC%sF#Sy==%7a;GaSj;)Ds&;F2 zL(L?e`ntF(Ob65@ah&r+tLmbWn6aJT&udgeJDt46dByzdg`Kl=IBS6c3AOspskI8c z_o(csbqM=x0g0z4nEe@6UJms;56@qE7y#7xlj4s^k{tu6#0Wq+6q7}fZBzvEw5Ex$ zc_mF0IX`21nn-^J8L?C8)PU`-XJE&b&uR5&lIWmJqL*OFe53%Rtzjj?()j|9cAM(! zj0-F5Az<+ntgtev$0e;oVZe&I?!z*YJ{q~29S8O*jqsCCF~DuVgNnc{5cd8fUDhLV z9w_xrPk<#4C=MD39m$lG1rv#ei&|lcq_2KCm75 z0`>*yUEni98&tNB3r8c?ABzEE;f>~TbW$GM_B8`$n!a%>ixkc)$l*oJ9E71P>U9> zSF&RMfGo=A3K##cRSY_GN!Pipn+%l0*tLNz(^|%BNwe#5Qi}ym8u=QNP&FI%%Onde$&jzs}Ae z1$<+p*FRUg1Pa%-aXP`%t`CGd3~sr1W8i~Oo5XCMF}D^W=ow?!;_`vpD_%1-6fa0h z+>BW{HRTkvHk=RX0g1>=8oTl!!5w_s7Q)5PnlOt%VVhEjul&=39bZwI8VWNFNriNI z;7}eg-{T7OZZ}Bv>|*~5Uom@qne1$|-Ij!U{m17&@Hcp~1)cd(Bzv@tv>n$}9X8r1 zmHa>Uz6GwS^vv5MoVciY`~_uWSKCfI zZPVH9b~=dS9Rvjx&;ZIsE?&5Zq7pzwybBI0idReo$H6KnGI-&8o+OMU63u~xFS_6O z8#y_*;CcV&d7t;W{QneNL6K5Q{n>rl72bCFr%#|M^UAd5nJ0P8K6;;{$^`!o=wa!S zT%FuOc2By;yAxXN(E&XxrOF;AU6CEOZ&sy;mbYX^f?xB*YGsRRJFk~?@s~14tc5BE zNQ~4;o33)?+h8Bezh3SFBx)9SsqBl%Kgcf%TIMr&yIEx_sR6bb`F?)otc%lD^2atSRF1%sLoqaf zFm7-e@@(4}I~WcqPp>Tc+NR~kjgFFyLM<_{nK`rm&4%dgA)2Ogc(&mr&OjGeZh_Ym<6 z*9}Kpp8JjS^UdHW4>&0!?JgXzORxx?JfPTn6uFC`6A&0QGK8NaC7zmkCRwPdmF0Ru zAVjM`zH%d@D+2h5Pd>!#a(tm_AZOBkka8-9Ubyy%B<3meD~31P=pyloPzbAFUIF_b zuE^U28x)VAX}p@<2vzC0Ka0Ng8glAGeB~HhJ*hl;xsRq>_BpiNeHrp(#(ebVvc(5@ zk0b7bjt|uFXqA^lpeqJUZhcRIDdq9jN2>Wd-@meS&Jq=ODXHZPLT~E<;=h?`Yt~U z;%gh+D@30MH1S%cg)li`+%8VITT-q{3cntKl{dP;2cx01V6=XmG2;xIImA6}W9G-c zwacs-*SM^%T^Knw_}Lm~Re3EGZWmrwej&mZZ&+=UZ1*OREU1jh5pR)JOuHv75;uYz z-A>8sZ|I>QeYawRPm+@w5iACaGklEp!{QcB`S8ignV-F8p1y*9T|JMiA4e)Je48y4 zn@5pNl=`@6ow$xUMGR6cG`C_w!AhTO-+O*Z3@lEfR+A*=YD`v8hBT3Ffj6`{T+iUV zf(>k1MTP939~LDz!hy8;9ESIdO@`X$O&xH+_A5Gsr{lVB-I9sn$^K@j7-mOYA(<}h z=~P)@xR_$WVb7z~re)5fH}8($BWn}HP1G`v`L`nKnETUfl@PD>uI1n36-Vt8EsacJ z+f_|bN5W2qwF!ErZJ#yxJSf*wIuFSsfiZyD0e~I)7TjQb=z3f$Wk$>{9}zb0!b<=Jd4j^XEIO4PofO&PXcSg@PBaQ5+5hPBvidFH6^EyzK0rxco&Cxi$gT zCM1OB(O5tS>H~BSSdC+l07{d8IqmalcY-&#FPK>7@4Sh{VlcS_#Nu+Nec%L;F?!|g z!|vu;rs2)nzmP>P?2~|u!Ju;9REkZZND`&SexaIBSl33Q#15K;Zu#2D?MjSfj1m~0 zj;6h^czVU?Y;l6cSl+q(Nb88L%g!}6LdgbWKZvPU1eoN_?QE#Rhru(W>fc3>o4vt{ ziybXb!(MOxW#9^An3t&eWxz3h#$tJ!@$?8d6zRiB{5w$Q-OEOnkBy^s284gkgp~h*T;f565TaO|b^X&4Q zBRz-7LKmK0ax7+-H58jnkwi+JGr1l3pY!Pizny;F(aT?30i7#1NLv&~+_9(FRQX;c zObr~p!LV}>2Hr-eJ#_0Ea~cHKjSgHWsf@SwNNqTN*^!q-iILkU>7gmzK({~{e3BT& zSo0(IhmIZ|%;R7$EUbsk*`Hq#Sp28g4~jj^Q^}t+XMazYa;wL8;qF!FDI2t`Sx>Q0 zaIl(E7l^m>KrtF>O%h|vruIsgOw9uxC}`zC?_;36dNg5fFT#zM&)KjZbpH!8d#kLe z#9pdAt_{Mls_$NrU}4);Nw!2MIvmp*p_lgO|{a2cl-k44JmDUK+z^ zxVKTTCuU#N3eYWo;#SYA47t*Rp> z+>D9q&Yt9`g)!Mjv5?5Aq|{$d?Vi0`(#^Z%Q6)A^+5_bBODXJL2R2xCFGNf-q({KD z+A|S~CcQJHI^JGI9+aSz%3vRmdH1^&iILBxk?xJS2E`;xsS5W5Am_zpQ?4?D`9hYe z#CvL5rO1crsIa{QfMp3Y2jzC=M5g`3#|)ppuiJQztlhXfwq^9sWj2 zjXEZA)OGovP!;8jpkBOhYF5x0K?46l%zgn{sC8m2jl31sD0={0r=_w6de>Wv<{CaY z`OQz9V|1RHiD9GU=`EbH^>?9}6RgWeo}HFl*e|hBB!V?<$To;AT`67=BZL z=pBAv%se85#!n26VN*^Pv{+WXISM0gJyiB0!1_98f3}T)Axc=r9`pb38`f)Jmkqz$ zSe)n4TR_?#^jpzhAXOLYk*CQ^eJcV+9~bt*!a)v#-Du(liy7{PZL`gONAm}Z@<`P< za?WDHK0>jFC~^Q9Ca`qAoz@8wS!}$rg-sg#Xy5jcs4%7Z%eYxddo%cvzvl zGPcjFmd5Begu=08G?||>xlK?WovTD%bmXYOlu3PPu>{DD3h4AH?MmbZL|^c>cfr)n zyprf`Ts#gR?H;DVCvqAx*Ig(?_c0==8B3zOqK`;2Hy_r8 z{TAR@85B>+rr3=X$uN}{8MA?i_sUr(B{w9ADovY5n}?Bx;F`Q3K(iM(0#FiQg%>1; z?wUDK|F=4@`_I1>W=$~Yva^s4%D_|N>hLDT+UN=yu21r#3ZtIxIjYQdPk?TT)8PyF zPu$i^dn2B>R;yDbO=G$fWF`*sT;PqSUv1kPijMoXHPz%c;F2Z zM;%N79Tx_CO@cnJv(kLzNVz8OHBz-N^6v90ppj!m7qSVox$%#~aW-OrnP*i8V#|1| zpSJBgEipqz_SrklWR(jih+)hJfgp=wp^i1(LvI`{~?&qjeZKLD|`s25( z>!6>FtGRHD)kbY}`P-=JvV`9W(hZ=O37VIk@>03Mzrpt!)5qH;s`Ggqlr?2RY=!S> zdYdR~N^f+>n+>4K*`m4^y&jmBYU$FItB= z3Mhl_V;eK4>dSva%vPk#E&Bvn;==O`bc+p|XVy_{Dn(K#b*-Sxw?(x?SVIOAY|lI>}HB&Q)-OTw5XcG`$Z7!30tn(5AyV#&|9um8vIRLw}V&7FG6eP6kGuC=SHe`F6ypNHb$#Kx>a$jN! z#FDNa86(8$1$JG}Z~nz~cjuZd%ZyJxZzI*?NUOyQe1c*jUEe^d*ZSv(F9(%GB5Ngv zsDQ2k`7if}{H2d)T1%M8TD6M$=)Fc7hj?M3*dB^_vOS_x)E!+Dn#APOH3TU$vwc7c2w85h>CTM3$hBTA?Bj@1 zW*C@r1fHHE%k~ym%yGh#>&7)3Q$Mb?=5^y@!Cd!WMzSnkixi4YqR0wLjUunTo=uRe zFe(w{cy`NCHpnT)@cE#!58BVahck|@9GhHX-KX>HTF7;6q%9gKEF~zQ6CYQ);gu})7CKA2Z67xmaW@3e7|k$_Lnc0eHdb%fTm`BmO}Df z*tI=qVFD^Cwwxk6Ou;qG)hwBDKT4wmOc*9%DO!_v+0<I1$340ir#v188(8gnq z!X8yN1Y?pXgF$}j*6>G?L>3~hnB6SlT?_@mvr4~aXz8dSn%xkeOLZI<0Y{L0jTT!Q zjzWuVkPNXV8k7H{wezhtUbxIOE^KdXD8isaj^28A^b(*uH)+&i-9i>!5s*da5YR_V z1Wk`ZNt+-K#H9v!_<=GNPB#Fpz6Q%CcYCIa_W^4c=EwWpibHPAK+Q4`7mu4UaH9)s zx4hEtl;3qMjZsL0>wFz6HrHEFWlx3U>_B1g&Hbc0rtDy z@;ePwT}y?9ii6^UiA#kjt$+-f&Tx_)f*G!Wu;cu1C+<$MCi1hvjfWzW`)0IC@63QU z-&KLdA&u^^*v7KOZTysQ6Z?z~5_Cz>Qda4aHr^SJ@!5aTFdpo)H*vxqZf=9ivR~J~ z*K_G*FS4zaZDLLSF2oN@XiOhs&TyS$X9fE(WG`IpyPMs(z#fOe@DQW$D-VrKo7MF;l%He2k?NV8S#5$n%5Gi>uh;X5tczYPTp-*O{TM0#6~~1q#kurtzq@R+ zY8_MTR~l0x!M%_9TYwwzvg9#;0h1qi*YgaJ%AzJ8ySBu%;P))>(03pj27_v0AhGUZQoV4P2-PF zXlKgI)+TG>uYN<)TsZ6xQT{>Uv^G6WAu z7f37L@L0~k(Po7qDv{k5m?_^T-{uddBIver^}t>}Obp?)jQt5yw#`~R&Sl#>Y{#h@CPY-?KhN;4SHw>0_k($Y)i0M*)4vT!g0~ON%<9O)%Lt$n1q4( zoy^5=U|V~ZxEp3BQ}D1vY78BMS__b8bV4?=&r_2Wo*sG>%on!E?^L#XIkwo}PLvJd zg$%LFfe1{@(XW|c#!c%R=RPK>+;HQih2I->vOC5-bT~qc%ZxdAmQJ7Y7igysIAb|GitxvLkxaHfMSi__=IdTZZ&3Gb9 ziNR_dT}UU0A{qqkvPW;(eE+zD2v{dNIUxkr%U;J~sEG20ANnvsa61H*V$$H+!|Ku1E@vX0>e2=VGnmB+)VJ z$ri69=CCYz?nmva?Q=iEeODygWjQ>P`PG!py)>m=l{2|mqLVS@}Q`Dlm2A)*9Vs8x6(^W{?hudx7d??uh%k%f=KKveIU6$*z%P1R#wnf4_ z?if=^@Od1e(QM{n5kbv(t*_1Z`uUG!1V`;0?t@Gj*v!H!uqm-<$W*}SYPkY$rf z7Y@NRTNsQ(6blQbT1t(?qZo$4*b5M36almDroiRXoBjV9g)>7l)K+enVcU92G*Zn# zlRFgA8F>U#ymW!s-O?sVA^Q{+0S26N474E_yI@MBShoT@a>!~v(gL;7t}$i|TzMB}k~!Dq zOZXMO%a~0*=}L_Ek6^fmJs@`BWY~8f{)M)2qIo+a$r=+c+k}%vA6KLXVoe(+KY=#X z)b;fL004H4|7g>KakUd;@3wh<{xACvI&JpfUYiO)4aZ$c5SWA;Xvz z{Fx5GxIkGUDGg2zJrva<%aE4KZip5JH>;|@x#e4R@9%zp(cBi*MymD0Gb)sOZ=n0= zdw!{tQ+ap=?$;{TdR9n=gE1G@fr?eJ?!ro&V{z;njumIR8jwD?}nOfO<8AYJSpI2 zYg{-*088aTL7H6@3k~yo6;k(I_A zaAt>L84jX`zUE){zPImxm!v7Ej76`B(cp>*S3)CN(G2}e?b>SxAndx2qfdGe^VZ*6 zi+b2l!@$}7I^7x?&#ru3IN)ys$H*`W?4lU?Fh>1N1q=I>nnI?R!ydQ3-eppeEUSd=muJl8DQ6xkWF-?2a#t;l|YVhz*AX*`=Eav#|LgSB7-h>+NSLXK7<$Ox-Z_Z{MIK&n%z^`e`_|9?XQe`dcwjzt)W=p zAF86%xyBqg3TA@`dW_Z42ZU+kA*RyjnHhUnfwBD0HbPtKHBvCO7U*s@@xrp;WS|zR z_J|kWjKxkjOa#JiEU$yeY@<7JEoeTEN#;AErZa)S*9ENdhx{)l0B2!oON$D_EvEe^h(X#z zFN{rQo1+f#4)XeVN&GbV#n9%cWcG>MU!%5o;m>T}6$F3YoAQ|72-wT^z30~+wPjW^ zJ0cLlKsyabBLNXyyZ#II;xsdk=3XmYNs74P$aS4Fa@Yb#dnk4nMJg!u(J7d6N^Z?<>_qlYpB?%!Q5gU-7CtBycWKPJ{qZM6BJHq z4KkgtRrHW2vNrkvNQ>zedS*Y&UA-Pqo#NH!r8xmopI8yz6;m@AN?fod7HVgVwJ#TD zX9XXQIm7CKQ^6ox7o!tZdkoxhKdB8h(1#To?7ccJDww)3sL30CrKzXsYCbmEV=7at zh!bXs@f+^VfO^nE#gl|r#%=*$=AZyy8pW=r$SO+R6TB}pi_WJtxwG?VD1oNI9 znWx<7i=_bZ!uE(H#_V4f!;&mr(HYT6AC(=IReh~Yqv?;z3hL$=;7wlPg(HA+H9(5l z%w*Bklk(|CnQ7+8RhEg*N}IgbM%M;+z;JDtkT!k;v^B4d-obp*@Vl!Za%qBg}fQpoHNNrJoU=vFy0`5S{vTY%^XSy__# zC~31bsxBar?WD1|q1@O?AJL-9;T>1(^S;JE2Mv(> zWCkGp-a|ieD|>AZT}AH=ahzlHJlik~pg-?+4m-9X>N|6+g*jceG{T0|CR(ZCS1nKWtDJaIee|46cHLP=27 ztR2wxq*d(nJE|-R>Wk2zBaI)vl6HCi?=~aVPY%ZZ*zD;1-3jVj7d^6M0P zl_HlYb^G5Sw@0(zO8zm@GyQ%szN3uXbZS6_0mmJY^^R9 z=F>M+r=c^TfIbBk6SpTE66T1js0*|8vL4Z)fM!+4-(2|F&bPk$-l88=ynRNsg}y)^ z^sAyirw=o^QEPd4j~10SaGPicPcM%P>jbgj`p`TjZf&9Qdpo?hO~9JbjNp9&9gY14 zhan(vm>hE7?A0TB$UC7Zv7?!1xMpaC?$iry7AG@APBf$J`^EqM3$nq5QMShd zWhE3_NRfO>jdYgXvQvHuWMhay&?M-0OA&7jOAXX2zT`a)x(AEoFJ(8U?+?WwbPKa2 zcoB2SLz5}*nsSJ2=k)|9zTWRv7JOWhCf`5Fd0cBFhRX{Fet0Kvg3Fi#f`8m^U6A!` zLezzQTpO|`abbo@82?F=Co#8!7ch_gsy(z`=OF%!A{Cn@JM1tB!+ZWiU&R3}L*Kwj zdu^lTlXpUXXDu}DvRZjbGZ zGK$g`iIIg@(@3<%NlNd1K5RJb}cX~tkBn!AmGayagGrek7dG!8C@aUVo)8tsGV1Nlc z)5BQ8y(7Tce&evm!l@&{wX4Hri89Arqh?z(#&bcL3(r0__~qLKb>y19ykOBp5Z1*h?3fv$1IN~^y z7HK+TnNN@WE5C!_<>o8v#g2U=hK+w~%jDofp2*;zo`2aHR#Dx&RDM)hD_g{Bio~_j zC9l;$BH)^GsfS|+(GFW=?>O6WD~Fz9+X)$>Mrh20yFd8Do8~#E_1ib|Nd>n=k_(%Y z(-u?bL5i)Th>lXv>+tTIkTC8nD8YjAZ_MIx@uY{&CoQUz(CBn`eCe!QI-XpmacO?Y zeLL@5;3jgK-V@WJ$_(7jTNabVELPRhS;1e1Z1jE1yE(l@rH3-|B5|p2ldKzLA?3%DC09rR(es90 zehzwI;n;O{D~n7-Ey^b|5@RxBHJAk7C9UIE#o$DAC2*HAUI=XlK=_Uca7_k-x5via zTImyT%aWLCk2@IR3)vmEb;1Fj2EPdJkt9xu*14lK$eCHY!}@r) zgHtBN^YKW}bhI%5Dak6oav*a!PkKG8Vt^@NT)F77??c{MkS#dnw|HE&N5Z(Q;Ip$H z^3Kd^_E}99j~l5qvN9TuGfq}F9^psZGveR2{jG#=Hd|BEe|Q+4iOV|vB-dg-TT8KP zD3VO6TS3h}3(OYO>J-q}(9}R<9W~aLg3*X95;}O4UoaRPK=6V)Irgw^5DeuO&F%gk zV@-nPvhf-lG}c(Tm8(P2gV8ZB(@6-3y?vZ|a$*62`vK-XkKbif++EsN~T-b0+FU;O0eI)6alE%9T zOPuori}MQ~^X^4AO6TQFz8cx0+AlsQ!IS|`N+7Hqp6P!j zrkA8Mb&OuVUtaGI6*ElVj15z&kYdPLg9(ChQwmdyw?htk`!?N1WB`9=hK;*#C-TiGbR9P=CN0icO|SBBg%t!M=Bj zzNLrem2=DqMWG~zMv}}s9(P7RGhlDGn-pz_VN4H4f{5F;M3q0`}1{}&|87h zp@Hs@REOX3y(7b9{|RU}ec-PX9UfIOI8JXGr>8QqlR4RJ*UgsHyH~hdtJH6?c+)Z{ zb{$1hDfJd${Vt{-^D!fGN?cAiDUh41MO91}c;xevLwC%^Giu2wPVJn);ZP^;`QUj) zz#;uOg0-GaTz1i7V}3wVk6S(~VDSodFc)~3doeP6%d1Q7gP~^tc(dr(ULmm#NSEp4?C(m20b~m{%!1Kl5HpHoY>dlv<1Vv5aE3 zQ_$U^-p;F~H&OTJ@M;!mv)in`{ac6I_+F32-}OLecD; z$zO(afRnZ0o%7$kH0Qt4|9PV?vQ_ShiRbHQoD4Y~lFdIW?F=q-e3k78Am&uO{b|mM zPvvINtV;Zny>i#d5PDp(FbtV68|WR*CUzhG&_81k9J+&(b)gn! z^1pBGnQ&R>x^N=Kh99%~8yl6Wyj2iMixcV@jRxgJvc@B4R_3U$bnV8FLlN@go0T`O zee$xW^|8+0$ai3(MML4d6#wcdthnDgbDLDx_El>};mg*+71Pv^^Q1!`lfI3Vohh#XZl*Adh!6tm=)- zu%7ArCeE8z04{KvM}fJLq$ykn0$y0IrVGctd!p{hoVS?gfGi82!~Efe2pR4PqCa%z zdzrEF^GjqNN#TYS7oJ+cPaTAn9E#OaB$HBSFl&N01)g;O0_fF`2y10!A(~eCK4lll z66-{(0yS~M>PcEf4HV8|cl%aeXH)@={q4@f#=zJO6C(qcJMCwF_UsEPGe!)N6EnzO zZWwW0A`f!j0wc#L7Wl;KDfKzhC~FGe;aeYl-(y+KHqrJ8O@gJ=(U{6vY4S~houXoT zSu`}6wmJWD4>(XxwDrEG+=GyC?jZ~ zbCv7FMdE(9ChukiR+psk68v_`7I`;%*6|yHnGysBAR1#)H7%yhmzd3J#8!o;(eIem}khp~0PHd~mtw$}TT zFUFCj7WM0W6#IYz?Lsvwog$r+i60G1v?^zvlz>zrQlYg<>!xUs=w2^>NH+;K1p<#7 z(vfNv$e>imv`6$x+q}}kb^rnTcHSkFlfp`L)7|jIWPZZy=VoIep@|^QbZ@Pqg?E?V zE8PSW0~Da*7vTH-kl7|Gk8TrWdn}0TV1NlO+dYeZ$U7qFiYbq-W3rWB@#;f?6K>!D zlA9if+Oc+J&19%zy&nN;XV9pAP2Nj1ScBBT6wul3tAT*|DQ}d8o%XC!AeBY|-OW4gX_Op4?b#EJL>UDWwTe$pj9&`M zuebc#=t{rTz)wbF4zm|d7Ea120BVP>zT!vbIqTS*@E?%VE*!nRW8vqvQS23pTm)_& zsFg1#$v~3aDlI3P65f%SX>y$2fE$;t4o9j3$O_>~+Q>GFj9^$0ofER9o|o#NDcKJ!`YYh+3CaXX81I` zS^F2Vh#P!d_r)NaEZ~z$u_+WuqSPj13G>{BiN*e3tKVQox`fx_=Xm?~+)=R~7SFwn z!w>%Lu75mZJrTVkQ5zfcsex`*oECvUgI>^p{xnXCnj~?PcdJw{PYnci4pb~}6Rh+= zfs%a_Z%4rLiv=vLU>@=ylw1-G8Waq=S;*AfOHM##mZK46KV-f(7DI117&MzNy)XIS zXi56rws>8HP)^NE4N~Xco6$g*1?xt&j2K7KZdf>Y*k5!z->~gJd)dP+SACQD*bin9 zw97gp>R>@qB)-9GA&HTWez{-tH0>Ve7vIRiXK-?-$Ncyk9T!Drw_{o6)&C@kF1(s9 zws0giQfvlA)=_F)wdT@&v4y@>@b?`OPx?Ijc#qyR7O%jbyFj^P+Jo9LqlFZ^?ZCjK z;9wlsx*aEcjQw=^U%oLC1}7H|$JteToG9gN2RfepBk%7mk zUAkSiYg)gX4x%(C-5bL-sUWC^>NF$5Cuh(NfX|Mfwg6?%{<23unm!UuJ{MlyoUkxT zUu*K^DRp#3K({D_VHL*4=J26P(@0KYgPp6meY<}3u+ zgOcZ+)3>vu)NLRl_th`{hs3$?L^ZO6V^?Gc znRj9KN##N4KPnGtQDF;9{CNEg;Kqc4&p!EOrJaE021W}zI?NCDVb~CM?jCD~m*>Nt z3uGNPc)6}gB|9zPRY0*o8<+#|3eKW8`nJ<26n*l0P(RfqxZ|D+eQ8TQp19reIqP2z z0K%*_E}62#PmD{aW#L;VcLFpp%DFLc)`PAHBL zT<+Lu8iqL@_6KjDF(AYJM*FVopZSx7x6LLcVa*RdAT_UyNx5WUQch9qaf%$F)JR}# zs%X)q$=d`5*>b4dzArhUK*=%w>C{WIk$?uJqZ=%M_MJ8-+!#kz0^6!fMp& zNo#x_c$<_*xf`kxn4V`a_S_xz0qN_r#qL$U77F5q*%9gO8fT1BJ45WH_j zInyRM@YWT{XV>LrbMDIuA;*So8+SdsAmg@7{&~P^uNGA=Nn}?AR>G^EFb+??71k)g z%bQh?VmtlMs80E9VHS@wiG~e)0haWeZXHKmm;dR$tl1lHe%o8mbORMr2~Vro7`S;@ z@%~X6imxBA{TrN|SJyT9-v6MHTKD3=R6N->I__70uWrs238JwIklv!&#S~7Eej-~kyBk26AC^b23e1O$Gd9Jcz#S&4(9pD~(&RciPgnz6%?w6-775q*fLgHq z2(kA}#~|F^Z?peA+X!#DeY?@mJhA=k)R;3Q!-d0oyDg@wB8mkjxm-%UKv?Rl1Bf)M z4u0n<{5+#-QB}y!in{1EK6&&`5O_G_y>{wKelN&CT$QAvV1u|m>;bv$TkUa|u87I? zyzQImpEbz|fEXB}kpW}iE{^|c8yG|IW>aI&d~4cBDjQrl8D*nJ3aeZU^y3I)z_^)7 zljlN@fnJ^;xeGiOC*ozuk+vH|LtgE?ozs7LX3PIBzU*V{1>T{k78!Q9t&N0!svKqi z3{#ajRvK#zv|hf4z8Ag5=SrX^Rg7&@JEo2TTs|!dwR?7n>#= zl;9d>6h`0lw7FtnV2qso)Am%OIdT8tWHUMzeeA!IWN}+uyRJ)4sx5G{onoN}sgP3R zOtQu6a7dfruhAz+p7J4mbNWGkHvcPWLM%3Vp7dB9iN}}5CdSlEz5^wc&8n7nAHTcg zXNhz3e){E4Yu?`Ye${)|J}7?|8HS*9I@~$%;)*2;vts(-6j+@rxtjl~nf{z46^dSd zod5j!1Fba<0M`(B8_6!S0ks(c{(O4ZK(nwq5wKzGxIafAEWhwg9uk`_!G$x4y%6ul=HCgxmaN^~Qr#T}KFg+T$U zR?#gxg-UXzzE!iTJ&q7;gufU4Inm($rtoHk<0o#rp7h|i)W)`>A3moy)pZ@ZKmKB# zI>ro`bp^6)Qt85R{bma!AEMX;6se`u7*x9$x+5fmxdBmqLsa3E)}T^Z`S=Zrn$RS_ z&gqB!c7){fig*i{qskld{UBL@pz%oW9{moykIT&n^7bNhp%~*qc z1g4Iid6;9F*Y9QyS)hbipJ%h;2Jf2Es3+FsUBc9mogq6zE_v+m=y!t{MI_2%YcRNh z8Q)cbU-`Al>-adW%^%pQ_3HC7?Oeqh`6T9$QF``$9x){pr%`gxh7>}^j_ECsr%q&3qF1uzA&uUL$x$T;x#vSsrw*blqnT>t2=O6}hBJ2NeE6Wi zn)RE@n#hH*VMF>mHLzsPDqi=LHGJcIg>#WvUag)jLP}oOtTW!^94TP1Ar|^s;XHJhIz@d*SI55{thp*Fm|DSG8k;k@fEoC< zez)w%WpezrcJR79)7kC3&qZ-Ut)e;nezf@)t7i^};IGswJ?Y=?mKc*63zfis8Zx58 z^$i%peH!tuq22;cZ@}1Z*8e%kdYSZW@udq-LN;paFdMD85Vm2~Az?PFm*qeKa;+?3 zDx#!Fj6S+fbe%@7Jxv`y4nL^Si} z;pPokp!K_^eYl>jnd?K70|H6Git3}_ubCJ?vqlc zmBmvuwbB;Fu2}{2d7zhf%xpYjlniBLoQjemaz|r!{nyo*)+;2J%^lfT^W@Rhy!c?l zB&Z&E{UQTN;sSb?_nPth`8m*G(jB|j3nJ4_pbLA>SaHL~b8h6|L$)o*P}umy>w~V@ zW~USB1z_DGQ@ z_#p+IrWVLB?+_imVan1-=u(n&5D+1l`p6SgBh3n|VAwFFKSZm*<3(b{2D*sX$7`Ug=j0R3RbNB|Xd&@S=fuT=YKg)S zmB@mB3*!vsxv%%TohQAXnjEs4Z<^bjtQ}`Jv|iyUT(g3Jg1(2^&r@v7S~&0D~7@daYr@ zx@l&Z%)M5)k`xV9%_H2>e0`7UeywB&3;1l={ z3I8*K?}8V6C?JWyW9oL^aZxU`Ot(QnMmG2$)suEih0|K4s{*@xTct>LhacrxL5&c- zOq1ia-JqRQHLFT&n1pn^>u0VBKETtDoMCrF5V3~LpwBOYz>ALCo|AY#{pLaIw1CSx zDmDWBr$w82JN%X_Px;ln9w$s<%9zVc&G_xWGqlY6gwwe>o_Y56gW#Dh> zZJJuYek3F?E^KdXkicAn)k~Y;(a+oGV7_RR5@}bFh@qC&AVVwW)iEoM)D71_=;o07 z{5G&q%g1Lyp?{wju4f$2(LERt_Ty@hM&r^Q2<|Fg`v-?9YZ*q&9o8rPq zS!U5@@{nSCDRQ4uV{cjpKOy!3uZCQaG$~YBQ>#80C;I zlGH%dECv=U>y?AbwdA)8pd zWg!HIk&4J5I6ncRfM0l?nRNpcTZ)9&eJg#o`t->)8T|CA39(zCfboh1kz6S0k4g-T z6T$)fm7(7WZq!EOjc`hYU!Z_apVF@EVDOiKMZ54av(g8NDu7^6N8`VxjdocvIq~F6 zoz|=kTozC+oKmpC=YTBniEMVHvF-syC@Mge5Ft_O+bBcYhPYV0d?~m$xD>CMxgktVP_|h-LxUh$$!hLR0;gG26a=5~ zfF?EYGE*3o8F?*yapZB)6B*>6+9Q&{s3kLvbffGpzXmXLDf~FdtC#trcfQ17l>NJQMVz3%hpZh`~=^EM#`7lht za>AaYP&Ei(&)KvedwxOJ{a<#40_?~ym|8#=Ol_4LWNG7#=><$Fw6G4h4ljDL_Kx$5 z?&Ij=+#G3r^-@2X+16xD{MBzrnhPf@%Pp)&9>qdiTNb6hz%H1u90)j*n6so*Ud^_s z(xbacj<}Z8`RL{Kb2LldYMfIgzT^AEZ2_g}g%9VM>)zEKpJ{f_ZJyI9>UU~0ioG}) z@+EgPI0oI16HchhZQHu6>3(gn>Q>Hr^7D4os~iSk(BpCbCbt0~t{V49%G&jgwaPUY7uSUsST@AE6Zp82LhbEbCALh@&r|kB zT#H!lvno72WUYUD)Y{lwx{WRp8@sVg@}tm^<-S?cM#l>aC9tfjifR*VFh-42qPqq4 zveT2+vl#D74ZOiyAUs9trEpR$of3nxwMF8L30I(yIGMlUZB0#Rn*74-FZ?(9G)HLg z!uBX=tt*afr;8&i+%slr7O0L-viwy{ziS&^8`}N`6zk9(P>YdH&XSX)nvca>ASXU> z$n=|N6?Ib{Nk9le7jF86Tz7r-<$>mr=o!24`U<#22F0DV6q`wr^^_WyQk8!DL%Mn0 zyovynY3g^w+`>-M2%UZgc>~F!&x>@>)z_?WRCSFVwvdC7HBWSQ+|&W5R6qRun$+yy zi0+C$BFWs=QZAgO+GYWeY>I{M+YCy*608%3Vv>EE*k4=OmhQqQuBdVPa$!UEn#tBPu_#BdS~05<1-Sm!TfE?R~?6hr>fJL;QMo zKcv19X0{+5e{-vh6mqj5F6`elSXhuMimjwbIi*JVi|f97c-=q3zQl?Ff~64 zEr8O!bur11?|c-y!$Z?5zZZ>#MLEG!Tr|E2?YlN5T!t17BkDum_yAFg0Fj1OJ zw2DseW!^@8jReWUiHQ?{F9`n|sr?xic7|MVS?wGG4_tPA^^LXT=a{kc{rVs6Ap5vs z$A#UK3l`WpPO(QQa)?sL%|H-fU_YC`PtYyT6s`d+F3cOOdp({HQb6}+m`G=`#93mk z;w)JVDzqA`3|_$B6ATVZNHvs>qA+b-Y-_k)j-`D0q*$Tpm3Bi7U|QG>(1yDi*~DNS zvR?i;;(o*tk__oD9c^@F_xarSMz-;L(aqu2lQ;$Lo(`XAi%DOnhlPBy7JgYV0# zs8V@+Aa*h!moy|rWxBvLIo64u@hb$);x@tlusZ%74-6I6(jDIG!6eMUJ&1HF9aW{IA0SX+yuA>k;`m|G0$V*Afe zSbNH@8kQn9eBU&AHVm_e{Qpv5-ogo@yS@_g@g3~@46e2B{+UQ+7IxC~$ z2%|jzRd>))PVgA}4{sd({%_3o<+n{gpGv+QN0wXEC_JLruPD+(sgZ`V2$~8P#vWGe z3$B@LpbfzqP{CE3h}x z@^%lIVj<`zg{+sJQew{<{@g6t8#HiNRD;FF0_2Z)$m=5b9UA<-u&3>$e+~}`Y5Ltv zElgSkawHmg!m&^Szy5Mim+vZ1AcHmpZ*bq>V~`q;)=CYNOdJ;tbi(Uv(77p~L0A;3 z16?+V#%+{cHW2-ttdL@b8Gfo`AMPVLFIlB&jm4B#O0f{(DaJ}utpYi|j6oefwlhFi z9&!vDpl=0MXY@Xg78Vxqm=P^6iCJsQb@6|%(q+iIg~blRl*cOCtR^7`0GvdgNE4AuH52l z+>Hr}T;)13UfSrc6)Y8QpQLA?{IFnZc4VG%8=W7BRTTIY`$M~B1%YYfpKe2uGu9_2 zQyGRRi~am`8_l84*mPZFUi4|%v!ZsSgU4=ZKB9YXp#WyWSYdkP*wshb($Ot zz2b!%eLDf6P)|da5^U8f683rCi;jy0rj89zEC#?88>_)kun+^1Nz55Ghcqz;0i>r) zlQ0%C#=~U>_^C0)4+bvAD=sS`GNlC6=P3(=;)HnUlBi7pT1DNEjow*78PYEPQUK9e=>p+j{gP+o zf%QN(g1G3enA7y)86%B7+cR>*f{i^l9^pB*Vbj`>`l*lEj{JSy#&cwi3)_)W3p=uz zVzVi-kx~~>2mbZ=J4cl#C0BxOPQw^H8)+9}~kugrlN9&q0lCL=0g_sA%7*PyEd&G0cv*LNZ+#B~=zEDW=#h6v?C1 zh>wLJFQ`@AkJ;>1$LIsrlP9kqh5oEQc^wFDH#4=;eE~_~eY|>Ei{Bz|)S0_CqlMWR zV!EP@#u}gnOntQTP%})l47jz=8SjxD;)a%su>qH@YY1I7d}yP%(bP_}`<+z~ZNkE1 zbUceZG`JYjY?-xo@=7+(scpX2`56;(*1C+#WszZ{Fe@?Qkno0Q z74?Bs6l^8JY40ha)CSaySVnW^@`9d!XU zlWXY~)xdesqvDfJw)$N6{o{7}v0uIPG}MUV%BL{sP{1MI65hJkTU47#6?{%7rZalp zFidI_;4Sl@t1byb=W#-u)c~PP18w?VxI=8O=edwak|(lU0&DT$Qt*g4&S%__HZVVN z{mBiKwvnb?*DB|k-50lDw|J89%9Lkpws3aRD0VePR$*@__6(b{FPbky4!|siA>$K@ zT50lYPKXFUf28b3%=7Q#=%X91{IbiM_U5IsHa5s@vYDkmd68w}6<&>kPM{UVikTMG zE&d^&d(rEqxUVeeuTE`-u^uV=5db#A`eG-3wrw+({q+60ubFK|(66iKk@YWG+!kv5 z1{GFpp;+*AH&JR7dQ0O~E2hpMTvK z`PnctWnyS^V)%^{RGyttoCX%#Q2Ar9*z;FrsN9^j>UDCH+w|hXm7_~6Qc+(~Y!5}c zDK!LXK^jtnbId&%Bqo(zWC>p<+8?UH${PHrm*;rynS#x=ouE8tnvS}op*)7zNv z%8o=m5Q7Xp9%rtBw0ECp5~B;RAzc$rdSa7#P3R?2HvEp!D1(F`zDRUXXxPGT6KSvt zs9~~x3U1$U)6aRlf;Y~LJ(b~O z4O!sAkr@d44>Cz96q`hm6_h%gDFIUKC$cukT;sZFUY0mtvQ#*aqhIQPH{VX%q^g;xM- zxA#0BLrG{7vnK{uK$?1Z#rMQOVHb-ekZp9Xa!+tsa5Az)>gXh9%;wgwR^LaN?ramHaP6nw5S?daa{PDaa(I zE?`w)1vioa7-l;$_3TW>eGey4al+KtsTcp^XH9nftgGz8b?i0>uuU@)P64K3!G5gYvW3R$YeFB$kZmbTTteqcc8LsB|BH?Wv%`iI$Rxi1OKiak#L(DYPmJh?3lxCl2MNq5HKJ`I%tAVw6@Q3fu^$6NTw zYuvZZi2Aog3n>!W&YJVOyaT*ij9y+GFcRP z%V#qbLm6|=dU+MFl-JCRi-i)19Iw*hdJg$ z+{v-KxVb8;>ehE!3q{#bSeiF)o+&Ifuh(;R;JnlHB5!=x41#k^mAH=-zmXf6C0-rf zC|EoBG`veKsx}&r$-0NWOxS0WWO%ltq=Xm z`;=dwJcC`tUnj0(PKKNgNsG(}_hdtyje^>x|e~ zt|LVrL`bk$E>E~B5WJ5#VU`$+Sg;r^W5P)?s-tc9P&>~K3<(z7#$l*%$=@w}Z?AQT z)n$QUBb^QQ5V84hL-cr1?oiC zow*L<1S8YpYtY|AnG474%R64U$(rleWv3V$?7dm^Ru2<PoLM- zNF?{WAxTsnov=E1ey*~BE*3UzH9u~=<`nNBkezkU zzBM~-JSU3ikpW~7R_xof8$Fz6r7;ghZr_)iZO^)t(DP)i3zw=@SXh&MiiK57Hl;>k z(m!{K^fQWMnpLO#w)(8(1W{b=DKfl zlvC%YkpRW+Q6I*MoPa_Vtt}T%HUnkR$Nnox7Pm>tg&mt}3s`NZSZKX3q|}>~iR?MD zJvK9T5ewC^6O94}HO%H9t)hhA=2ax#9NFhp1PpjO-ljmH42=s*hg!&$5Z+CM19}Di zkAd8{uoGlEZ?zH)l{0`g5`2Aadxn9Z7hO`?F!?i;88L>)i5X6_Cvk9@X9 zJ@VR4F^9C!N=D?v9a+YFbr{=ll>VZ-!bh^e{V*U*zeb4 zNaHJGf1qrCP<`Z8ioHaU3zYgAvn6QRENtX>K*~i)z^sa8EabuHAxJ{iI8{``XNy&6E~Cj zNQ!6>ls}i8W>f%a1ETF)+ii8M}<;Bcu_hrwnDTQrhgleBAge{6%|?Pch*)EYV6 zx5pljsrJ|#a&a1N0f{@Dd0x~6l~}LY(ACNov2nuuklWtaN2iGwx@=oD zx2bXNZ2FLg8GwJ%oc%pn%FS-MaIzD6kp@kT>nS#kBC9EA*+z2d@@QP%*UZ#3D>`XJ zOcC$IHvoM&P2w5Dah8T@2)Nm%VSdRlE?eFDk73qD5iT1ovQe#ZSb}ns*u|*P2@=5y zWCM|N&CCw(LNGjip1N?PaDZHPn!FE~7qDXyS6#=Ttq}6soC}~hK4fkKKiDbZ_(dE~ z=VU>~^1L+XtP|bO<|bU&PS{9io0_Vd{T5E!EZM=+H2bB_YzjX!3tObIV{sG>1P-75 zSm8L>BRt(URtl#7{#Vu>u?<(U(YuS*iA;ZEaaD_|T$JL48cH3Ka{A)LY)P+Vt$%4u zElGkB#)CXVaEof0{7ZR-?4TcV8?>nIjxU|HAuvUmue<@JkLN@S!{UW$%qnrA<1^9r zkHp|N_Sx+bKKN%Yw>=U=WSd=gxzV+yOvIQSlHVNZIZPJ5G7d?Og+sE2Vv{M7NU4*= zWxfU)#Uob5HW=6G8r1hGk7{>jj_Miw7HvHM% z=xy{_QAJQXyU#sEjA`sVVT+<%^u!H+W=0}wYC1&LPWgT5qZ8o1RepCtoZ@Cwqw{s* z>@^;}g>OdSZQrEu%8-n} zZ1)>8GyStB6&f2cKwemq#dOiv{IVojbcLct1u}h8SBX#4rwBG`Y=PqL1#GA2#>{f@ zE=Cu1O0o_ul@p(<`J*s0e6QAWL1V6t%656qbu48)eA+bOKnJ3nlY0CVP^Y zVzQui)aQwUYt0Jaf%3Z>J*>o|y@T?9*?SkbrqVNi+#{Yt@?yx1U~&W%i6Dq07ehs+ z(V4c>?)19-b+_AX_t$MJ?bg|LJKas~c6vd_8!8tSR6s)zA%Z9(ctHh$L24=Jl+#SE2KIkLs}WUYGOGHd3Su#uQ0#Rs_H>K1qc}h zITD>`VYG=uubc(Zwlu}fSwKRDeRj=stK!kbei=+ZB}n0$)Je+5q_w zWFJmb)QOVFUl#rE#_4XGEU38A5%a2Zu-=z#2x|Pnk+;fXYG7}L)X+!_J?dL3umpA) zBU={3h3B~$qg}q&mrgbtqi^r{#jnX`H_k~OwD98#DK?)Xc_{IS;>!;p&6fn<#vTYZ;ZAwaPQT%{-0C`lRqzG38TSw1U|M7uzm4 zUU3O%xWVOm4}Gh7P4`1J=Q-i_g+wSY?s#U@f@l}Qm4x|o`HP1sRd)0Dl);TV588@!?w82M*#pDHI)P&d5$RveX<%dQuV88<< zWzknQC7gBRM=y6|ADV_(7}tX6>FDXd5I(V1faQV@_l*n?8+x%T{5yHcylcQ{sZ*@* zzdQaeeNAFeSAFZ61hZWq@-I!=8&nNv_AvXrKIE5s)zb08jp7>KD0nfPL1M(q)46on zL9!s@!&hD2GeLGb9Q)=si96;r%PvgE+RFR#TXUbVMe+wRZBbt^txOkP zirq^=S|BbhvXGw{h=r5dz=rTYYE|2LF>DVu1YsDtk zpAY$0g-<+m?3u7#yeA$@*iBMhDE{0byb67D+oF$7%nDA5{cP58+R5Uw4gzh+vh?h5 zC{RXvd`*6Tz?#j2%PQo?B`r31KJp;g(j1Wj*@`ldCfC!*)?Oe^WTB9<&-*B}-xdfP zAd&(+?nmY4rtFh%WigJ2<(d$KK^f6Cv0WfyQ6O~LvNTexj1&{Lp8v`A?09PmDK1!X zV~^Gb;nWFv1)ZsAgTPk#lvU(tsGe>TX;6sIsDGrVwMKa%6emK3zHO?TvsRAEhRGp> z*qT-+o?rLY-bKn~&M*@DCM$ZkY&a&xG}k9|)G&FFf+Wc(k=8g6cm z8{3}(3oLG?*mU&C;`aFE`eAPy?wpqhkqez~@$5jhguL3dh{&&$Rx@JEb?4F+C zZ{!N5;YV@)+5UT#%5R$C@j?EAc8L<%+Yu$ILm|s1RX~C#+q*}BldicgIi<{)fR|9~<)IM?T4fI)X21iR z+jsh^lB=?4a#D_2IYPkhI@5zWvNW-V^q;gor?;4ILv~bQ6-X{b8gP6 z5G`nsRfn417bO8PI?Q2b#uY4FO^x;a_QHyj!34&@6x+rx4UT6zODA;CvxgV=0!vIQG%kR)!4k{fTJpo@PG%+q|}X@0O4i!W~hSF$bn@@!Jw|CA1;xR@hZwwKAwtPPTncm!er%FjgRe9t(Hq5ERbR~O5x@YZr;6nDvcL8SHbj@h%_}LanAtPw3In(5 z=@NPQ6ddiW;F4%}%=u~E6IN0NQ0`nIyruX^wHTS{v2gxah@L*J)OjKq37jg4stU+e z)qzy`HOW5zt16`ZsH1h!>!JPL$T8h5*cmWf!HpB4_B{CPKJz@>I6-YpWAjgG$_%x~ zK9A3kCmx&zaATi&vjs4dD0VePR#I_TNQI>jKp5O8Z4#vhj^YMz$ZXjIpCPx8+To?Y z^am@=fKZ?O>0y%1Z4>3j3CtrFfGDQeT@=}Yd?Dv$U5d3a2KBA5IzbZi$p0=xaM0m} z;Ev>`zb-6EoGLDvv+m1hX7-ROf332grwiK=lLtA5I`6UxCNaNT{A0f5UbiR53)k@b zVMFE$7s(#Xywq~I07P>xEx-+Xl3N`aYgbfX%TPPxDTaI z_%~-QAGZ>+ll`L?0XEI1{lIX+?7v{hi(|f|{2#M1`9mVH8z zhg4k8ES)6La}~KSui@pif9-_L**f3*^7FFKX_M`W4|@n3y25!a3MAY!s0V+E zz@ZPnYH4jVHxTZfloYYnSSW(+&`rOj*kYO56lXF?0<7sj?ekPi-B(2Qm{U#`k$5!Q5{jacQV4JAe=}-~w}BqeMsK7($$Z%@>c+IpuHT zty%6~lP$Lm{=3Z~e_hl%@3&o|0ZF-6JhkZ2yx%4<7f7?9IOP+^bAJx+4MW*-rfhWllH+DyK7R$yuicO}-8Y<3Mvx5e~q~EsE%g8^5nQE7L z7`CRv-f6b=ysOT0^E{qxAhWFrhh7t7VS_g*`D^#*t@*(x0qcB_Kt7FkmRwTo4yj_! zgR~}sn5#PoiHhgXS+gs+F?p(^RVvhUYoT;a7&(3G=&8~$U&}Dh*RD7aA zW_seoux`5l-}BzW@o!DsDqRYSJGrWdz*hrK^3##=3Mlaa@70>vrIQ?v2HOk2!QC16 zpN;|aXN^t2keaQG=&tCuB*l%b3`F^dYi1zpTsWb`vS zpoD5KDGh*ZfE@SY7*fF+n`53X+~N$G2Ztto$n&l`HsY&4@gH=g%Er_60$_s+pfIn!fRMVqK0?li1eSq-Ybh=G2;DI$XP$w=z z>$HO|2J9p;Si6V%Iw%R?B0);-m>F*6v7B#byrL8jJghmKUlT{h1{HOE)FW{kuO*_E zZdb1uS3Xtmr3WIX)Ig+nI6xcJ2UUj|Jd53sD0Q;gvkjbbr*eRZPs;mv{y`U=ypXq|ly^4QL{ zEDoaP`9%@e*W%f(p0MIX=Zqi=SE9XuHoj5RAlK_NqB}(bQsc%lDY4i@ez3 zrjR>6h8xBC=Q6wt?^=7duas(y4gYl+vxeUld=U!0O~nmaVqhf2%7}Pj7c|deCdVjw z)y`W&;$t-!XEBm~Ak#C-P&}1~)YGYf6+CQ<#c12(i2D&;;IKHiz=bZPQo8zO>>ygfd#4xkyI7)^Y%!#LLqa^0h2mfoG zPIlW8I2-&g8=1wzWqd3}fr^1;(Bs<9i;p#tE+kHf2mRq@pL(7Kg@)4$ zetO%RRnStFFusqJL^lgg$v>TSG&(61102ZDUM~Y>d%9bYHhzf^+rheM>;%)p{?{ah zj%{OIblP|a42H{Q!(yViPYt*4zuT5q{?F9T+t&SQHX6IKr<{@C-vZ&nr~lpmdd~N6 z|KRD%g{~?j8!|r*0%FJrt~ulc2>0D?R=>UK=KnQYnbz-p``e_^jZ@EGSeUrO`jez?U*qmK5Gb5fB`0_es9;^o8G%fDCEs1LXPP}#fuG?o0_@n#3yqJ)n(fHa6=6ro05?7c_eDH} z^@f_mWQtuwkpv*t3{D{s0tVJ+#Vv(ab@|OvM~(>u_TXm};PTvCp1y+STp1ncrk`>~)-*o5+dp=DUtq?DPTPa$kl?_H*91I|LVlG!b#RS&f6^L~HEx)gup8GVlv^OZh+?54 zHy_$qVo-67xAJ#Vsl0EN%zHS`ReK&Mz}o@vzWfo}@cx+o?QUztxUIKp z10BtRe!(M&fw>CpV>d%gQ4$ct4%{2kKBp|G95%NO%6oYyx zKQ*v7c=NPg(F3Ti9K6pptLm$Vj^T{QRK<~>Of%b<-k->pldaqmN^Wd@DlM$f9*QlZ zNFfz>U6AUXE#50J#Js=l&AQO;hq4K@5@aJw>!L={-NgYOFU`Yn`J&#{`rA* zbIZev#XHDF{_|&sLZ{599wNr*l9_;mKzS6~II+av5c8PdpssrN;ryH5{zJz<{GsYy z-B+5v+VOucJ9im#7IaL$8#j2o=g+`z-9cgfHBy2^L4 zQ&jjxuKA0maRk?FdoRPSPv_Pz{mHA&;oNK_<-HtsCA3+zSF)Y{I4qynB3}&BHhauh zJ`O{Ium*JwImJ6H!sGMmV!y4@tbmim>8py>+&OG1tj2NnOKvz>@wM)Wue#@Js*Xq! zV>KmG@}?;a%00mqF+j+&O;sIIFRS-SF4>!}O!S<(Q z7i1kAMvSGkaW-Jaa6-(O4neNm+wA$4eeKA-ice6XaP77-3^bmZVCEh+G@5{|}mKVNuwH0|bnH_HbFaY6@Ge&FrP z-#4Q}u{!u&a@LK}(Pe>-YZMD?3zv}EdnMB!TM~Ui(C57?x%LP zy-$+I{0wh{8p|6tN&6y>8%rFT1lB(T(CF-r5Nd$tM36Sr$mc zWiUrWH;UV$QuugBVdR0Zbf|_v3L*PY$`N6fR>!edc43nn!rm4yeBZk9-fcZ$8?E%F zHUQk{7zx6#O8^UgD`-?c>7X^}6_-ToB5|Esh%xI9r?;F9H!}_aWVnHxe2x=9#(rh- z`hN~Io1nshAMPVMH#R{~tUF|Z+D)-g9$rAjHA-*#cl#NbLeNax7QQD;b6C|4oV#Zv z*uAJZN9%xO57ucMtFIZ2Q&`2j^-eP;u}r?Rt!p zb)gXAXPHNx?H5#28z^Eak;8b{GQ0y#z1wZs*h+cgttje z&5BK;WpWWk!jNR84nlQuxhT!#v7t(Td84U{s` z2k7a$(Ot3u5By^{8b^YxVQkS$PuT@4xQP?8#(r~~{} z1KP4fEKV)Og6>uo71zQTlnWLVstl@PahbQ}D^q$1#R01PyZQCXLTQ)i0x6Z0d9UFu zVf}`cn>{!CrjU!UtlbV?4h7{G$tGzoD0SuW z_Cll6+L&7U1ju~-wZ-w6j{Vqpsqq;021hZ&$-+=6=g;%4wPoGb&9NbRyPSV9q+tH# zdF$S(`aTkPfIEye!bxJCXD6u$U*q2>z5M2AS`JRoaCDv>c#ab^#ypJMdBcYGDJT1OBkbiaI%Nw~bqAq!z=IspJOmOWmB+MLVaU>&4T<2_a9-!;8ms z##K%T88d(NG1>3T%h1Y%q`9PPxT-2Qc0yKLDB3)x*nWyUpyKiw9kyMo%IHjCX+Vzx zB5I^70yx|=g=q>%Kaw_jFZix)((57l%AAlo`Hk>8x%nonKQ;I_fo}TY(B+YuyhwCn zw94vfjnWH~b_W=HkWg}O5BrebrPu{TsU@PNkp*IGKQa}PgP)ClGK#F{($I`MBQNrV z3W^oe6xRW=`Wc%&A4cTx3ZwHP_k&m5M-tgq1yal7uXk?lI^M@qQ+dgO$LY_Y;-^lY zsf0?%DMS7hnEhx9uj6$=qo_`dmFu~5o3uN|ZPjdSV=_7BZxd#j@%MhYx{4IKu^V*4 z!VNl1v1Jt5hhd!pAboFVcZoEp0=YK22PAGj4yu%D;)Pk_X1bdkRvDKwyroByN4F}< z=)2yF`PTw9?QGfHQcv7GrSY_rjnLB5NDpBQ2^|eB1WaUwB{A3&Tm>YG_&A3>lx>6J znE85Y$7~$fem{G!*oMk53GuOM|Cz8_V2029iyJ>8tK1kq+b!UeL9rVtk_zHmywZSs z(>iH1Hf{6+=$$Za+woYdJf&P7*)6Gr4#-h-f4G6e-l-k=ecZtD)8-@Mm($F+@!DpC zW@e)p*N(yULQ~J2r_nqdjJioQmC`5jE5I^vdVVr*o1~2ZEySK+T>ciB5)N?SQXFwZ zj-?eDIX-NUU>iBNWIw-U?IS-=M&!mQu|YjEn2mz%c5uC*jb3;~v@Hq~3JppP2L9s1 z3MAW9EfE6|B@hk-i7(6~jDq9rAVO?Ut&^_UMu_y(qEkNRMds&CWC2;rtt!}!yW5}$ zct~I;n_{&T$)w^Mlqq~5u40g|tWlbvx@}a`(%kn-pVpzOmwisqBgXD}B$`Q3X|78O z!S^*dKN0oZ+`21zcoVKwk27OL-S)nQEPib;l4XIBwG^8~k<}nTtSFyy$8)c$lPL|* zAWI@>1R_g~OP81D@PH@m#Kf@Ay7J_I+>7&k!;Fct8Gl?sc5uUl8+Q|YY=MbG6nlUo zd#Sj^;X5HW*$g|Cwft5_|Muo-B>|Q4B<3!E7l^wa4Q&=&gplo{*se*BgUs_{Fb3Ao1MLH^@8v7t&_i#?Tu_X1s##rZ!xa1Pu*b%XjVL~DxBctX%{}iSUe2jQuYT#XO z%_ixs2vn0e5_%-GNz^1dLK>89a+C)+N8b(}B5#{LwOIg3r@6UwdT0^v0WrW`dHaTzu z-)vlFr2pe_vi!9%E+7s&#JFsr*!2`iq2l6&OT1 zDrN+e?8qc=0fr;vJNisW-H+9b85sZdaqn^RAva*SvCooi0gUw&n?jL9*z5RiWHze% zqATgHS%2yzAX3gNi>VdgA)PL7HhFH6>_x?M54!k-?t(1fH)d44we#pC@~In#Syx(Q zg!(D=0Y&a1IrAx{K43qy!ONJ4!5AwNXx7Wp6xbWmpw!bnP^+tp%x4l=&HC|BzH-MC zS-YF*0`bn-raGsip*ti49+)7)h7(P`a4+O;fb%Z4+i#!0e#R9hdn(?j!DrGG?K}fn z8QIP&l-3KnNu3NX^nCML;C>ZETafDRlrm#NSyaJvkYNX%NJ%}EKHT#zgROM|tWSkN zX>n^J@_<5%3u>^}3km|`g^(@)f`Ck5Dyj%aQJs9Dp2jlLsu1{vE5hrHZ=5e|7PLma zg6W`PV*Fzk{4U`8)>~`0t#7cgS#E#ln&i(difU5DKOT}r-|;(6m&EK2`Tu>}r-sXR z_-zWHY9VQ^;QHdU)45AMOgE(rG8yMVpKuYHONXB}+#)-V12 z%?sF1&W<@xGPt=lZoG4X$jcCaw}4{x6xn7Xt2QFW$kMudMx*qB92T@R1?B>JVo~s> zLV#=6!c=jOq$j#;?vmiPDXof4bJhoc#M3GlRL@`Zr$u|_fnrjL$iTWHYvdfTI13`r z#`*dm;%2MOU}=k-oIwt81B)B`z?Uq*a*|?!e(NX|hgrk)(0!r-a4p&dsdHA&jGw$a zq%gXlSMF6Fo;nAQ`guwG%s{-F$GZtq#Mf!l(c$1DQw?f8UC*o|D-}2bDr_}`w=%sX zi*B7g5QKMhsLN=)*3ZieMBb3j*-i8bRflS;Y8?gKpu7u^>b}S%g*v_>ItNWQ4bwAt za6EeRf0}jE&1muvtM-%J;VP`$7&9MPU}i7H?x9E#6}N2M4aNSDR!P6l*bKB|5nh}yC!&Mn{J;2i+iuUF9`8@H?%}sV5|u|UCnfXBr{vKoBqR7Py;%5#JeMwl zGX>%l{>~@}byWpC_Sa8d3e{$&)_8n5enp>meYoRa zMBD#hr%ZzF;T^tWdlH86J>A!JtXLZThG^4<`+*Oj}Mks-xlAiVTJ zPDpWIr?HiJLu=h_{d~=f8?Q+=IY$Z)llMD^p8 z>ce+7y$yBZ)cGHke6{1<&Uvmda>HPVJDP_X#095pqq(~DtN&)5A8^~4i;al?v4A8# zBos3v&-!OZCV*FA5+B!jm+(x%yiuBc^TcrgfeWVE_K>zgFx=voK#%=(h}oN+k@dH= zB=5D^8$nOokd5>{iY=kYZYoa4d@4^>!LEkiCCiPx2y}Iy%uoH7^Xi90$7K6%^T}1^ z2i@gE05+ggc2s_n^uXr1Gi8crxAnjf^= zt3%~v6mo!>&ApE4f^En+a;2qQv2eT@IGz8|^cRva4rC>Ue&XpA3npegM(Qt#P-F)B zA)xE$Hg*yRHiQ@Qu>)e!B0b$UWwa}eIDV*(d_E36!|l84w!Gjux3xdBc1PTHwXxxP ztfiK})iiI>NAuGZcz8YN9tKpv{8^o1-{R}66?``0R zBXaSZZ(}6qp5}yH>XYh;OI~$BBivf)#6{RCBy1O+qtDOP-160lbA`=5YkeTaw{yZr z(9o4dKZHHk84326p^ESwx+(zfTQ>cKRI79)WE=p>v-QKd*svIl6Hvx3_*F-TbW20wp$ATfCFX#V5`i7`L_-dLf*V3jD*HHsM(0%2^I}Y&s+5^I8hRE?F zt^oe!{$F>A%&bqUA@ZzZQ<@ajqo_y>(%?G$HxaGhue?~fYyKx>T;w_S;B%y^FiiuENo6qq=G zG;mZ>XcHuDvS>{)vr1YOaD>s^@#>S})z2ZvFj}h==NWfLmv-z=zOo=X(riEE|Et+T zin!Si_f12P(-!vQ2*n zhBAdE#>QtOQ35Kge-_&W2C7K}gh6X&t(v|~h3r9^vbndwbw(|)vbk6?1scI1PKxE8 z=E{F8gTf-qrK~}o1{@t5Cav(rpd4ZbrHru$0BSg4v!=N>r6gdtZ<`bgtVXf{b5|IR z@j5W+@2k6YkS{(>u}x*1v2t5y#>VoKMSnWyV_^e`fuDxL>9Q|rI)QTgnzU?gXVe|f zU9zINAhktqOw=If%2rs4&Umhvdj-0MD(OC{W;g7|Z-~x@w+Zx8t@uhrb*KjKY4JNf z`-CdpbG2f>3|67j%H=WX)9NDd?t7|zfg7Y}nURrkiI=TGc1`>XZsN8JU-A8A zE3KRk=`qDMRLk4;Go&PzOA)m2CyT zl3wP2Rg%x-q1r`QW@Lq^g{}_W6Vd{1^|L=C2GcO}=@fhoGww)FUDur9?5@m7q(VP!*RV5NM`gf$bk zN4mmBJo~UmF!j#{alsL%)reZ&wz1%!&0DAMpQhd+pSW>9+g*#zPz%Lgr$`gfhRM(R zqXshQWy`nIE4+-%-dS|2Tw^K$EfC}WNUJ(1hw{%Seg= zUK^bsl&^de+9*9JPmpP@2VM_^L;4;lmL#FR@)^k)VB6KxPrx0nguL zP)C5Wm$^N5T(=ERPAD65>R0{&zcr)mz)yW-IAPOcghMiS{>_WJ)#~-&|=RB84 zp?eZog%p_}f>Rw@OJn1IMfgWNjO0KR5;pOgbD|txn&K#PRE5JFnX?%LHIS#i9ymLZ z$i6;&z(dEs5ss`Ox=0*gK~Bgyx`yP0z^?_r-~tFkq0{w5o)_6go8m=4(TJS+1;S57 zn?p6Z5aq58-QWWamGCcU7VMgE*fU<(CBomi00K}t2Hm(;h1)h+HkM~L2+inwzhwL$ z$r?AdaRnAOZZpNEqx%-Ok-0CA7h?B`R=G;v%50o|JftN07Js!@z2K-cHLyjN7g#p8 zLA~PN9mp)%{+hq|3RmOw;?bOUm6LIzR&V*viT`7UhuHr&V@Mr0F_Lh{TUwlN&9wa$Jp`xyMcea`!O`hG-ONi1%)nB~V?Mqk_JEsOs48Lc#9^Lz4F1Uq#=BpzPMW~O6_9o;fp*>U zGU&ejW8ot^Oa696^+p#63n|P28q^ZZu-5n!JA+Dopuy4oM1c5x7^Y9ec5xIU`xF}bMJZUf>0ZEmpejsMIqH` zz63cml-SaASP6`TXh}*vjg(^hK!ZAaDtdwFDUL!IA3h9+=HX=@q7$xjng{pYfUFbW z|HXIB-plVge%DS;xv_Kmsf9msg<_!#`vR0{LAPcKzev6_N|Oo->AU7F7B0N2I7xO! zeG;(F2Tv8kDI{^$RMKl>`h8aT-Vn7by6C2Gcofjw>{HLvVD73++%^R(2Jx~cU!@I9 zj4X-96DxdE_(1lnpP`FfA}kBQ8jZt(Uiz8@*;`L3?+MQNf`_8&ft)H3MQr?1Hz-$b z0mZ>e`KJ*YlpMSP&lZTQn9kYzgY2sCa0vIq*deZ`UiQ`ChWl~B|8gnZ46lT=?%ibP zYvUZBv;bo{#U7-{eqaEYSvgayJV_FufMVIW9(gt}!fBOT{3^na)3;}1FH5sv*|_Xq zG<|3JxJr5{2vwf<%obyAJvA_ee_LjZ`hfJ?ZXPHTLUag^v`XB6=QGd7s}24zxRyTd zG}%+9qwQrdvh(`3|C7XwghkDL{WfSG9@?|>ltm3=H%I0e%Q7&CqG^Cu;v1?0ae;7j z*4v>IV>c>>J~Rr)v&YAuw{EJm(Yx49SMlq83~V5<%VnMxt*Wl%?oPj!Yh2STkVC{Mb8cHVAuzt~rGI@oQad*#M$ zEwmXhz26T!NtO%;828=P5uF8K)=_LSMb=Pp^-z7@2D!uaip4^#(FEe=F6eB@98o={ zWAkDsC>$Gn1W(xp#l(-co_p0>;xI+HM2I>7#;I?V_L0**qw4~h=Ekn+AHg&_`l#*c zABMzX%-u7we|*zyCmMdZG?$c(BNr{cdLL8lF^YUd#nnh!!M%uwLg+5VEvUKAi!>)o z8yL)u=0TtV{ryHDtIY}n_aD^0XQX-QrS-DB$aA3EiRA^5YekyMKIzsed6Cf0niP)G zW_TCW7hs&OdRmjH6)F+xKv5>2-4LKLaa2L4tfHEaH18&#$FT{wYsy}1!e4gPp;_ZL zWp3NMpPcxib|m`dZoC@V&^M2tjhX+xh<@nWM)X3h@9cOXk{D_fuz}q31I{Hw@ zg7hwHGP2jCeA5P}TJ@Z5QDxrQ;=!K{>U+}6$aeL4b(%saY!j^EZwT7yn=RfJ)yV23 zdnFkpUD+QtsC-^3X_(`ffwOmpI3UH^;Mw~Y4&B!t@mOvh7I~$}$ZW)Liv2E%9!lf+ zPdv`@o_K74zum;3&ZZwJmh$!VL#bXXdeJvI@iMvZW{o@_^xBW@2$^QntfOG$)6r3*{@A% z=7@z0UQDsOD6)f!yUZ@0P)XMbw92#oSEWd`+$KGt(yH#$S3;MC9%cF_Y*&3u@S0Ys z^ZY|TwS)RYg^xkKL)a{OEL(8a|I^u(`~|7em2!hRpV>-271as4L~Hn^PUn-c4`J4e z%P_Bp#fjr5kDs2r$(qpH1{1c94=$YYdQ2H4NuS8q#+V`v6}0BKeA~3s%6k|$1^hGAyr*(vVR!Rd?czWajP ztB%ild05tce}7wHS=?rWr^;qVZu4Fud?Y^XdpWFx?iWD&KrhoK)iEdwuwiPQC~I0) za1yY2@A1>qySyO~TNQAW>5BM>R|kkTuqX_wRoPW7z#T*OE@RJdI^2;X@B6Z6Y@_>I zf-&#D>a}izJ<}E))2LxoI<4Q=*fYCLg%u}hymIBjg)UGQ8Rp}6#Lf*fjMGoqM#3-- z&6oqS;MKo0J2by--0n|WxH&X#oc~#D(IwJDv3Drag)|QJP?v;RAZWQLR_WSY+9tO#G>+r!&SZq9@;TY)@6 zb6cjLkv`3ByF;cN>f2#mVd=JeD;uSsWic0l{RJ%!sGZI3i7tUQHKY1kkD}jaPYAYA zYbrvvg8w!6XH<4wuZLNx68-FCjXqEW&RE7jm+KgJ9+2T z)nUn#lSarc)*#`V7?d;#k<#qE*&O+r5najXgjzgWZXGI({ToarQ!EILFtXbB;oUo7WC2GaZ+`D;i4H@C-)_gA2)HKgq2AjR&dNGTQ9Mz=xWph1}mifjd91dyp!0qPl$ zrwV`~F9C@apfYQe_68?T*D4V|(A98CSrQH8H1xJNt5mtPS<|oICxvf1g%4xQ09#^< z{hGWqwM?Irg3B&|nxp8lc+UIRzs!=F!6UjW`YlO$ZNLLDzaik!QS26qWKeOs$R<%U zgaA>l7Nu&9;U=ukM^gsuj6O!gWRild2cg*pWg#^x4Rek`#9}sG^5W|KUw>rXe8|Or zapP^34cUbKAft9QXl=|R?|9)8k5gilX3N?Rff zO6;H4tntMM4}nSnHn1fzU9z5N7g&}Nq39W=oP?cc!`MdA_J_f5`I=pm|5LT)B3VC< z6kDw6ITQ=Y;w@AhTIV*eNs+7vi-pi6I*rP@8wdeaqC<&O1$qMPb2%eIk4{A%R-}4TdTYu zQKCY>Xk@{^5q=?_*`nu0=8d9imu}I^RphVOK~# zvv2NM|6?kHG*6YyLlI%DX+RPv+!IwXN#RM%1yU5XmJ|bxW~Hjm^SJUPYfxX8oKT&S zR58guy8^L_vstj!H(hmHiQH&UfC+kqmwv`0@BK5|r14>yfg?lA&}nrtdWH@?qT`*H z_=zKl^182uwbek^Bsv(gpI0i<;JOQy7~6fWFe5E7cX$fyhlRs$;n0H%lm1ol-)5sC zmi{<~+<9$`N|HrRGJY)km?HfYs61X%^m}jdD~zlMnzsk!9560G!4>a%kU{+rY9017 z>8eu21<`gIj}}}IHK><<;|lcTqDW4uUq)c5Lg%@iZl6;k-lJ%Jqe0#D4}({)ia+;1 z7f*8+mSMk{kqA$LTR}JQ8 zbLquw|LoQ`w$t7GHSBG_QaQ@?4E{EF=QZPs@(K*P{RA6FW6R^o+A!%P*Vd9(&P zWW;%qEq?cA>u1z?SMiUCc8O3D1A7pFvkce@KjO6lpW;Ii%Df|w$Z8+WAyCMOpA82s zen_6TCiW7nl33WbYl05wg^lLVm`XPlr((L@(N)joNOH)OjT!S!)OW^MH^99nkw6;? za$U0iL1}}+fZf7IDYjoZI;{?V>c+?Hg2YJPz`>hhoLUr;}Qs zVHzu8G<(E{gV7coh`kYhG;|{oOEejdf6?vw(CwOf8w-eC zZ{Xlf+yV&SZG3ply1UJ7BgHoQ)e3~glNVwL2`lt^nbg1mk5pcDNCgnvK8{%F<$|c) zi>8JHE?zW@)30&EMOlbpf}hzzZur^0g(RJugY3rM2_&nBBp~xCHjg4XR2=e=t@gpP zc&)O0TDp25 zLUru%n9oA;JV(vZ;5y+>w;47Dyn1jXNdez}^hjhrx<|0c&Ni*T~V1j1EG$!UK$gjR|$`X-@DM^Y(A<)>=0~a@oDOah;TnmPtfN z`D5%n_D z&3bxwz<@`YOPqV!k(0Y$`?8QgKP&y(TfJwQpPx+B#(qv&p|ayk2>6 zW-Glm=){!0;=ST_7jtV^f2bVh+WJwKzUE{%sH@`c$uDQ-aAWIWgPEfZe2bKc8mzR* zq6?*vG7f;w)=67@7oJf)lBBb%N0(cS16W4KLp4LtHdy}c-My4Gzp>kf#cVJZWBSCj z06@21FS$cEgUZw)Pi)>Br3K;HH{D(YJbPkvj%*`fxN55>|MU6Rtd*qQ78OMn3rRM` zLM|nfiaP^o+B;!+sv2H5ovGZ!-eq%xS9lxLE4&{=CiqU+{-A`|&ZrG=rq^@l)YgEr zBD`xf0AnvkES%9%fC(pzP(8cGEQv5X8`*)|HjrIz>};H}z(@tf9-_zrD()iDGdjO} zkkr2;FZa?ZTOg&{FT5g*2lcz-@>;r7UgM9i4QkxQAbEGT_vfIgfD&?9^oOuzsS&Q_ z8`N2WyP$9l3zqJ9_CoBmO96BXAQaOEeI9siZM1>aha7bl%c(;EvOtsd3+JGV8-PCj zcn_F_!}fKTVnDeIqrs>4=QVr zJ$=myZJa?J)=hR_mGOKVQZE9gT(pG$U5Z{{5h{YaI7O^yY^j0xS6 z0gsK$V&P!_Y~&8%Rpw}PH{rw|Gm0oQ!qHwRe(?`w-xx`Gr5l%d+i0Rs=3SHMc`Z<+ zou}GGUk<$s>)GbOz2iR&Ug^?yqhV*nURVtKunSLK_X#@aWi~f|(p>m&C~|jOeqRXH z4vnyFqS!QwtfS&^{V9s=iYlaA1C9k~+MxCX+m*|`*2MPvR2r*M46LhCy~E9oy_0QP zPlg-8HRm|3L}R~Rv+qBx3r@I%YTejk*eFl==q(K0orr0jJrLBRI6S)#_ExFPPJV6l zeM#{sr3#C#l>#6gS4j9t#7rjbb5SqXj2fuu8or z>_}*aS{I7?cc|frzQ`4lsWhF3ez-`38s+Jd!W+VC>0<#&{5C-kS;Keve%x-DJoR2j z$LxIPC;yQVW`;@UcW&(=`L7L3YAs+=MzQ-SQi7SD9G_-Ez2~8jK44ca;w1;3@Vhg) zV0tLh|U!7~F=R_pgC1H%hkN@w_tm+?1@~!8@UM3}~eG$-5=EOWLMFQM-0< zV{Xf;7!;b%oV3DOTX3VPa1=XMP;xq;8rKDU-MTBD%hKb{q^2A1!?0o$PbsRs-q1IN^N4^pgIJrC$+8(6^{fd zu3VM=d_ZzlX0|dXzWJ@IWF0pvWC_2M?4NlW!Zck_wqRoJ!40m<7a~U+cw7k&)7bj@4^IS}K~nqHx-zoOjW;>f z7SmozvCy)=i;BZI8rI9B{!MD&U2jdN>|iAJBiAE!W&G@u?0xz6$R0%x(0tvFXbR7l z=;?m%0S}BP;H zQ?cg6y32oQ?a^BY+}t(=X(NR8EHRCAJ{L&0UxBzmc}~PBBI>N~ue}rPtoO5b98N}M z?2q0#;62W~7OC6b*O0|-ycU7SF{GqzEyX5LWHlAnqX1SOdWjHA)__iOFIx_L82g<> z^k?tKZoJPvOmG^~DK~S%#F$6#1=H3dcwDwSZtTR`(DE}?yX302L97;?R3xG_7eQ6f zWy)Qm_2I2bh~&0OF_PO1ndyASSjUfCrbp%^@sXag-&>OzX&^{AnWkt2A(ExB8oaJm zA`2m&MqNPY`@xR0TvY`yo;I@0bc$yqJZ$oh95-b4A@4b7hRX?uka)qPC&Ri(|COQ% zvr&p)3>m!!Wr3GQKchmF!cUvk?~OYk7jJuqodP?je%NPSdD6Dgd2?m)TYq1D=k)@t ztOUY1{p5b=Fzxq#FsXm`Xy$+hC)z%+4BkC@KmVL{Z;UlLoZGsmHdyO0gnFLcDaj0M zpR+{THB0Y#mWL!v@d4#tU9+Bm$W*^~5_3H&9@@Q*k)>h1%n`p@=C04ZXx9fr?3?%J zmIIdqz_yVXrjC+Y{FzU(b-RY!qQpi6MjxrAJD3t)j!&&Pj|M_E-#yGHv_aVs)$9er zn9(&rt*?cexQim!qhfG+Iih3mF_-`3hK_$4f8gQYmjK@0nWw{=1v|a;yhg<%-<6?P z6=hT-yE3#@-T@K#v*e7bG8ls-XZ`C0NstRn;3tG0RQ>fLgPcbwT%}cJ1pnnEX`OwM z=_3bKABJTFFQG2FpdN)I!k#lk<`Ygw+`B(jgjoBhugPA=hUaS1sX{8HwqQ-~?6u;0 zS!L+M2p5x{3_Edlf@0WbU3oI9`u0qLIVdv!;>M52DmM;_Y_|xCWKirziljpJSA|AP7XIngz>3Q77TFGD%z!jyuotDWm6w+Bw}%zx%G7;-O=6T4ly|mNeXx zn86|c+gBP$vKxbArv*5&D0VYN(y@YMIkXj*dm-~tE~tSWWm=WnpuluNf~*F(&rnkp zW9r)ozQ)X5&v!<}8R3c$gv!vFbk#8@givq(HToALNdmiZxqyu%FcK4(`1Sj}vFTis zAuR;v#yhBxndeym+OBw_29m%%y!CV8@mjm=1fg&EmRv0zsUsJISV8@Ol2Ar-I^eC)Y1Y9T5_VO@wu8<_6@=f%Q( z{@TEW8S0a4Pc(=u06o%>^O`;Xq*bovX#?@x4PY3}qIUvW!AJ8}`>dLC%+tB6lEWZ) zx)wb+|6GfF}k)EjtLqjr)* zL7gBVxs*u;P0Q^*`Wd~^dX<6chWt-)OsDKFXdNQ&pH_)iT)b#G4l8GZ==rhIm4ZU+ z_^R86No+86?g$yI70Q=1N>>PPDO|r+JpW0;8G9yq*auyFlAG=LTsn8_m(9z~Pi~&e zBnQV07hrSWW($Lbg*ZX6Q1x0v#jPKIU(z)zhj)d^XC6RgZN<1!MZ(qu*E*4*;k3;9_aXK^hxZpAW0ofa*r+cBX@W|B6K%^r+uO5M=u?Jw! zmLIN1j-${vz}n^Ch+)nhJy`4i;?>2J-8%YmQz$jGhX!Xa+F zgd|vqbv>lmK8oB2@){tU(WFJTK`;16ye@k2JJ%#SzE-)Dw2}c26hF~LA}dX$pqAD# z&4SCkZtqJZ3)pP7M-~FF+v0anANeG~*yN zi)$LFY>1KrbLs6+9$hJKWj_n418Jsu84jvR4NPS5k_PE+HnDB=HWlJHW5OBuS2Qqq zFR*IxN&`^&{g?J)Y^Vb>8ppOB;@P?4ug@~0@BMOh6)7AJeQq4CKVe~c4pVFyMfOo~ z-EVfNO8u%OrM`Vp`=?%0JS6LdrGBWbRw(VCwJtI-rjp(ZVY7B9gEXk|+;%#HBu-9K zv`cj0=Jb+QD2O{Q-{SX>432e@PMm2_55C$1A}x?iglC*)%A5@!XW1^>@HsWBH*c&N zK1)}G9fdE2?VE>*)VQ%(+hPH%WQtuwkpwDE7uh6A1j_}Q&~#Q0y=J^YUFF-V zIIXOdYL)qZYsNQ;%3Z(LF`|hBbO*f+t|)ZqeooM(F5Q*?gLPZ*E7hN6qgA-i`xYOw zc&&=^Db-V;;s|vy@_>lZ;J;Mn+~Sg#&yxL6dHGn*yu&tBmL&xI^OwvnM&mbb=8@9j zJXAM!B!J6vNULxS#a2+=WF731uy1pT5JLRDZ?wTq zrUDp~c7z!9R<1CmK@G~bs6-YSXd9(jVj$^EAqU@pcH;}-YoqJvgHgzA0W28#Of$XD zzXSI>_a!~i^|G#rHb@)X@k~>!o{tO|jytaQ!)VapaQf*mvTVaBeO~(FKbT?k{g$(Y z^l*#ey73lht;GiX!?A4qSn`;PLr&`?rYfK>;zq<7&wJ6`^nLkOX?mzmoGV0H9Idh_ zwh_|J$ZFjTLUH}P{nElne4<&fK72ry3*E=Z0xkk!1wyG+@kEXsZs_%Hnc5{fE881X zOPfZk;eRywmh`mfj(i7*iD7NsmZ`U9UxAHP17q6ZBWED4j+fuWXyGH+U_F_B&wG`; zl{pu_l>fv7_tXvH=ViT$qoJ2X*CmB?d{}1SnF&XMvjy3flf>s`_}f15*c-Am48#qm zl*H5vYWQ0Nvw|0x7{h8*O_P$skBl!6CeLx(jdSb&CvKgxVQP+_j@dFTOANFp;W{RHPD0FW z|2$P9;BKp8muOvZq976KS#*)@>J{TYCdGa|yj~`Se>I>^fUJStlC;PS=4t>wbA(YaPX0-C`^cjQ3oV9XSOn$21W-pR97WU5SOcg)i5*iMmQ=$6*6cikTX4LLdk?3 z!mVtc>Lzrvng-EeNDo=p4~3M!sa4(}XkK!BZu#z)YVf%&sxzuwx=VJ*bK{px3cC$p z)hg)3n9SHTg(gc}7~Rin6LiRLM%F_>QtzENtxjGdGLl~75{=y2CZ?$dni+c{VocFwGu#IC>ukuxSyGB;L7cWAaA{D1xJaJD zFP+v#S5Es8^BGemj_+*pu|lS)*Xx z>;}m*p8<|+Pp-4BZ~_Un>15zP`kB3uhM(DtTQ<#wSyMFd$Un# zM{|m~5pji}n9k)lE9vAw=mdl`Fs}_dN{ zYfyr)Hm^rH!a|uE59JpeLFd~-KGUdQSj zh$yBhu+Ir1hsc_XZ-yl6aMD*#r@)SEz+<;BQg5#i?y#qEcBHs}_6^{6WQ}v{*W7Tw zsI)XW)C{jZUByd?o?8gfeZM&5s0A4JQ!I2r7E^Hs%ToNaVtIPeC}sPpXM z<*M?7+60#-?2OVNDlpoN;p<{%)f`Q^R~CI2f{8s4)G^(&&p&JYV}24}gJkV`dh>+D znOJYE*%8@6?}+Rpr=6(d>SUZ*!^Pp(IX~DP{#7$_=9LLLNR=BKrfU{hsHfO76gf%7 zRZ8#6zmVrbRm|p)J@X!rYe7KQF!A0jt#ZYO6~8*Pn_fw-icuUU zA*Rp!TIhoqgZjGUB;6Kuzf5>1L{H(fuW z1{CZZfr{O;I1)_R{U#3Fm^r`VtN&`YJVN36?~#k+pg41=``kgXtrWR|9qt)SB3lyO z2+Xn-fjR~YF8k?w{!bKFg;>mqX212@nh(Qtz|({QQ5}Pmp;c}o$%^|C$3l|CyCd~< zF1>8}N|iBjSucAeJ`8>W-cTXx5?un7%px8r6xPd-gSOu%k=-{9-zH1EVd^e$cRB>S zFy5+FVQlvfxybbUU}p-_2cZMBjn?{9@NUimUmhJBEV)GT|9)?z{?C!@1|qu`O!G*l zX!!3qx3jipkJH!QeSL^SYspmW!Y{Yo&D*F*ON-P3p7$&6O41ac)2GS3X;n-fuh=i! zyDz%R1ypplAYnRhH(*~pjMJ}j0xT6)l6c;lDDyR0r){0BaSm39@rD@sprD3qkf~v;=mo7_B=iY8CJ+o;~pLONI#?5WVUv;aIMW0e4;eW2G zmFa?24DwvqQDAB(f5C#}n$Ok#nRdi*q(ycqP=2Zs_SxH_SNdL>QYWkAZ>KxoIrFdk-uhg4kDpl%l2_v(q(&v@jGdW$zC z8z6#H=UER$VCUyvg!jS{uS(TXhzdS?C+;hgUC?>9!Z>=w8D6$+Pvq|p6ihPjkJ{dh zxJEL^5$GTql3v(BvA}DZOU302cSsI}oFqNa&6Y)%FgcOQ(EEjTO>6u&ihC3#0U3-= z*eXvayCys!mGu6g6A<=D^U^~DeFq3a70%lVL5yQK29i2z@<0DT?Bo~FnNDh}m(KlM&fbisbTBK(qQEx%u( zK>;%??a?Z6SAHe*1k^^EE)958%>Q7~pBMd2gQdrN6=>OUh zp`L~sD;07|bjkV^SQuIcbzuAaQ92l70a%VT;Bkc%2(dYH9oVx0j{>otegw7&Mf=*O zEDFG5xXu}pZdF>6*)OqGeuy|YZ zl)k7UJWY`idoQ?x7cYEB7l<1bNY6V89Wi#J!SY-0G87CBY#R;3tde7cS7fTjn-`B-#uYc3)y}vK25o9XDSldZbHw7ybG)25qBIFD4LHG3Umwln>_B!08tBJ8CXQ!R>PWH%iBtwoU(j@z+X zdb`i@kOQ8Ts-)41ofi&U`y8)H(9K3i_dfqC(`qEf7LN{9 z3RwyiFvWh=)AAzE&()xuJS2m7pMxeQHmlUj`h6C1^+unA%ciQRtO=mjOk#@#T zZ`-d;o%VL-qJq0DDlQ-a6^I~;2rj6E#Rc3!QBlMNg6Q?)MNmXk_?{;TN@AjUAz`BZ zdViHS?{We;|2*e8&+`A1lP>IY+%fYxu2OQ4Tx~?fk$yUNb~W?@t(7h0CviG}jVhZi zif;MVi4Sl~3Jz*Nw5Vw5%2|e*x@=I4=$AE0R;X%(15$O3uuhf--fA^817WvOsZXn7 zL4%}%YaI1j=!U7!yl}h)rv*hI2Am>El+=VkD%?wr`Jf8!wa|Xg%OUyhMbUU~zWW0F z6_7Y>A?%G)IfWpH11F9DyDh8XG${&1CtuGbFe13Tv^%3Z8itNn4sDwnD-P|re}AO# zv@}01n?bf=P>@tJf!`^>XbF}sGy(m;UiPU6k@T)Auc#1zZ*%Ho6!#5xw-de^2x(&W_OG z$Bn9r@B;w{{4=3d`l1RF3o_&DeKC4z1Ft8#pVmW|6{U~LM2)kYxnmvyR$i_!R_54I zrx;;{QvEUJhjNoEQozs|zmZiQQbhw(V$#%;z)|$j zuhbt|4UIk4#giI%6^x+R3Rg~a ze+PEn-S?geDM7!fm`66cu+zKGZ0@&Ha?m~43L1<+&wyU$;+X?dEh{^C2R$&eTmb|O z;Te#ggqnja?^e1rdQCW{Rffnmv8)ziDG-)L22Dok7RKM{QB6MMceDQiMn+}ahvWVf zV5#ByYJ|mwi&JbUwN`L*rzQF9^C?Fc#Yi2GO&rdJg!M_X6DZancH&vvK)L?;wL5RS zhHE?64$RfQ*cBsB6;?5Jf@LspO_i{kN{On5+#qb}^^Vw=N``KFrHD6d+w-KN6JEsb2eI@e0uB(xuIu^?kQ|n`bj+v!!#tTKrO5e9VE4<_DhI&Yyfm7k3@Hq#?@$QAy^Kc8I zgC4+6K|Z^V=S<^ZF*VF)+u~s-zO-#CtgJQE`Y5-cUz_a5$0^I+CwE+UBeTSe)v%9} zL)%(672g+~&fXuF6*lyKNU)%w9j2#aLXor3+-lf1674~1Nt4m_FS zPT%{Jug*TJye0T_W>(nl5RC#WQaV(IQd+u(bDVTb^TENnCCCp;mt_X-^wLW&hg5U) z(!S_YpT{vZyl0`M5bm#z!6zD3o2k2BzdC0zb!+avITvAnyp4T4sywDvh*y#25KD72 z14{!o`s9bzaPXb;!(KjJ{f(5lC%>}y!#`d4w~AkE`C8M53)HFKTyRm9A5{>AWN639 zwdq@wFTl+VdM{TfXB`cpnHlu{>BxWF@-snCxq5je$zry2xv&Ft$ZYA_Ny!T-qNU7P%{K~fE%)JTTbpmf>3UvzxAdX9m}H(Pj=Q|-Tu|CuZo=uijUlF1zi zy;X4W8I1x%FI$xjEMP7ix-nh$nXDZoE1clnj2Z<_@@^Q}m~mD0p9m%lVW$1}YO>IU zF_UeEnKVkCOp%okr&F#2X$dSksHXb_P(|G=ZV{{te6boh_-po>;q1f2YX)%AB}SMS z|Ls5A`H7`F{En2g*eI!N7HcObO0aMMc)>ZBfk?HFbkG>TSsl4}+7&@62QfC5fU*xs zPBI~{L1EorcK-2SP4?p1{as>m-G%MNGc$W}pOW9B$Q>&Fx^D@6N7Oe-6WB_IAA^(n zk6V6}@uTMNfk#1AeHk-`dxG}{&AaY9SUmFd6fOxR<&%9iMMfZDS*;HwI$r zVvey(=yQB6x0;3HriKn5y-C#x?k&dF#2~Uz%5&z97{TxIl(R* zzgcOvTYFB)f&ck26@Q0S#7YC7W$nZk$Sv-Q>W#+so?N9C%3q#GJc>wwUJxv|g0@8o zR`7NKZ4|1)bpg4D`k-=)vLXtlGjNz*k_g?3#;zf}y^mmnQI7i^0wluce2k3FS%KRk z4=RC}DvfjsuraSea*F79hO#cC48hZF5s=^FY-LqMtqZ(5+Yq}p-novwkfja5_f)t0 z^${IA%WKapd^rgG^yxAj1PALQhS3XHc;ZmViuV$Id!&0<*ID?d)zX=PF3UcjQu=?6 zTFRokY@hsjf7v#^iA)Dkh)Hl zZ}{R?`AQ%4=YrnoRyv6Tx+;b|)}^TV^POH_HR7eq(TK1Zj}bt|S5}pN>yIXY{9El0 z=8y->mM#~L-=vr=T?@y{2PpECiZ2wl@;1lnr7NN68%0ZUSQ2HK|a`51gT)ea9Jf^%>BfG2Y&bvee?@RVc!>ve>-(fZ@@$)T z_3Tw#o#CDKOaloQ5KJgyol)etV@08sE(2yPjiQvU2yapy3&k(S4i@}o;iu$!gz6HgZzwW3PXkx>LZoTLh;^FdWf*uI_rb?MvRuAiO_jQEU6zd{Es3)( za6sBDUakbW*?s^?p|Fj9>?$}E!L{{{_dV7!TdY2 z%eO&Ck8aTQp!+^*)J9qByMm~X0%zv>fL{Kw&>bSDFKXt~%m6Uv4?6kP0pCA%M+|v2 zfw}9+u6Jh9I&J0y=_om5Ar4USd%TenrUP`oD!7Z}y)h-SyvW7;6G2x5+dQ#dWk1w? z7O^)>MM9VjQ|m(XGF)12k<3YHoNb=R$t}_lJ_z?h+Cjb>RNI?WdiOHHBX*q>lNfqd zE&Xy-JZojEK9V_Q03bWpn34PzuRb$%(XHRan*j7U{LfO->cVa@sJ;!Ot?8!ZpHt*B zbc?aOph|fLJPSQ!q)MZW;^((yOF5@V?~FzX9wFV7Mp5Pu?I}Yy6oGrOmz?0^L+Z`$ zg`jqg9Zed=JUn~B|8C?w^>)u|tkP&)qJY?Lh`94KbZ$5KJOH+#evb;jLLqkb54z#D zKAY}=V$v#UqFjx=bjiYEb`=CB@f|wp7SuD8o%L0(i`9i^#pFj;$d`H*vvXMO^lEv( zD9$?@N;36B+OAP@Pns73~RY_jx$;ENrZsqo0W$&b&IiNwuAo3|nf96(jF; zZ8XNN@yaMX9(%xKeI>Ey?8rg)2ezPhG-M$zoR-*P2Bx)?Je4A=sd%g-%<^6rh=d_6 zf$H7d4y9i96r@$gcB5^WGJ8=m^7$RvlixltgJVfd=Bf)Ms6P$&ebXuV8j7S)@sJBQ zctz-UVrF0!a0?+*@1Pt0H;S+Fh(1z#XL_tx6lPx1bNP>hMpK0D!jWJbMd+B;zT;iM z9{Sm+(g_|DV;62MUBE63!mCjJ30oO-RkprY?OzZy7U;4YQNtII5n-3vL)uJP%xHRR zU3f*Y(OZkuutxc;Wds7Z7@lhum;UJfKd4uFVxqD(v`?H*SK_sD(LG?Q($NovI^L3* zcV>dZdB|SjGOkk%(uk1q0yWHLpUK^OlZeNUOpfgL&rn|_7hE{P^1y7BX`|%8=-fiZ z-vWw5sIG$S|0S;zvrdq1(IS3^q?7KJXaZM9qLj%cFCG179UFfvqSU(Z?BI`g1U>u0 zeMzRQd&-$P*xRyTtqe8{D4iuqak&KQu#L=|$rYRjtW5;%7=mAkR(0C7BoR3XWO(41kfWf_zanh5+ zPYX^?dg$>SNPXugL+%N0cQEg!Lt$eB0wX!i3^so)*!$xzm|)ZVwY&MGjM*;Rg_p7O zW*|FC$*U=%qvE%PESj1ws|7e*hypb~Z2MmsnJ6)mIS;xm^+J)(M~9zD~QzuDNz7=*%saUGHz}rY|2J3{jb2)E+rGgB)T8BNz6VL9=vN&fyd#ucJUq zGX5~HIsEqY98q@2YR?BA=C2x6(8(vg4OOXkzSZ<$;*a*t`Rq&0Kf9>9%<85$h1T;5 zgO-Kv^y>EjO+YYM72!qESgqPmuj1U5r$y(82Hb}3z~gr5d5?*X6UJUUSPE6TFV;T;JE+vtfXGCF8qccGkxwndPyFTW_F>}X{bP01uz2#>NJ+=Bv=RbV#!OCAYt3Lmy+uyu8XX#g;eD}czr{+9l?})lF zy+nA6Uru!fZxMbjX;$qI+dKU(ysn5Uqw%Ow^~s<9I^g!npZ@!z3UiJH{32F=oQ_W5 zYbWoGI?g`Ib`Wsv1B3ZG^7>bffeRxahbr{?;DIG$jmw7lY_QX;h}jJa6vn;LO3!97 z);nOB&p{K-WLoS41(OjSb1x%MjQ@Jkre!lt_T(2AX6+;;E^JTEnc0(SO0J{GAu9en zJ)~)5tbkCb0~NIK5GC3klRFaB=215)&I?R;^hB4(8rjY&xCcZj;N;$qXbCsaw(11c zEU5GJ?Dx!rf2%@4s}zcC)ETT}q1RZ|EG)Y01$PAt_C_7Yu@6J9Hzn4ujt)!AF!Xn- z+h3e(!qChG7g|ZhJLB~>n<4HDB|k}#TCB9G;8uy-q4IVi>!h+T;#Nduz-mwMT|pu} z0}8L2R9)O2Su6LdTrblrx&(PK8bw0b-5L6s9m*zE5xtFlC`e29MQjSW8It4Iq^g>I zcScPp7GC8Ao>cZo_kx^MD{p!1ir9txCe_fBppflb8PG~svnpBZ*xhu$>?j9E9I5eh z9{Nm}93cq9yz{4zW*6ncn{gYvsC9wo11a8g? zwvk>I-lAi-h+1jQWcW{^-V)HtTL;xz7}x3dI1#4SvDXC}pTljpcG_n5rxEy6Vw`5O z8pWQoA+of`8~F-QVF?3D$bSveEn&#`h3yn0HEK>r(+GD`NBVwU`nOdvoJ~Ka8zcaL zX*rQbGNsj#J<_JoP4Z59q_$#&UuFBbw?_O1BYbf3Nx(Mm#;7U?|K}?kqYevj^O3Kt2pF3Wc?4g$5q?(89%JM@xNgKN z>&W~v?$AzKed!Nm*#s*!oG!f9a(|M^>z&TI$d z!d9fz4A*&-9NLt!hk6Y(3M7Zs(rMB4G?q~!ABp;kpoDeS{a}%pn*^NtMR+sZt?!!gtTS;h{N?nd`6ir{`k$4Oj&}wpNoGAMk16>>iaem= zHOhn12FX4jgWhXoPNcd=wpxyIuLX*M&_dxs>2e8dcph+e@mIgM+v~jGv?62DrjQlW z+N0`}2ZI|VW$qcV>Lr1fNCLl6fjkne^b={J5H~JRVABX}YtQ16rtapZbFzS*PZwV4 ztuBhJ4A92jVU@Y3%N_?G4XW@D?hn8wdd+|6Qm}$lpS4MqcqwIk$ZW-EiNPO|MfzT`A<}3xbN~48eIc z$g_>fQCu;>`R~5-J_$D=Z+Uw6ZnD#b<9w&gY)%CwKS+`NR6Gc$fBXm<)o_2apPd;~ z6n%TvQ?f*Gjdet+jfLK5Xo$<0w10Yk+!@75<)$e{!EWrA&hmaFy%CfLJ=xGM7rr(e zXdoq5#kUM{oStvi8$u6G}F! zIZ@oa9jNy*#C>A*<){{JM^sh_Y>d|XW_j(4JuBKjc?C<26pg?pGIdiZY?zL_w~LFV zXWz#P;2zl#c9Y_9jPqKS?3(-6EDsKuXe)y-vL54uS56D~XVdDV|NC!$n=~`?dtA7! z?TOi1b&ryR7GXOT-w1vIb|RFCkc=u*aZ|WK-UGFm?NX?4=#eH$(qfRnBL_NLiy*L+ z1L9zY?pw@ct(G?`8iD@`8~&Tc*F`M?J@nHS1Re7|=37rnApwd-o@z)y_%$lp#XW4? z%3r1%!t)^dRKPk;>LNb{cw*Tnkd5+^dFmUYu9?ZK{)t#1+Nh}TD+}5re@Y;~Ww`0S z1gZuXhjr6Pe9|t}vl=Abl7qny=tblPF#7A*jS8y(A1yXt16%tqS;1!8l)wT@6(1&b zGp-wxBKyqH_CoV#O91flP49xtP#s+>I!^Y*K9Ss=foFhCU8EDN_ia=yl0TTZGV&UK zB?}n>`{Qawc+X`)E&W)w&b>9#sVm!Ppkif)9d>H_iTM)Cnuu44W?XoCZlg}(SZER5 zL7x{C3RB&i1)X#eW&F>nlWcuf?3;S)&pPw0?G-HWQ_r8h?fr1=Kzs-rNa?~E6- z$8256rR2~~t3fZOURo+xE;&moS>XJ(2oA+V+leY? zt7`xe1d)}G)FpFy{q-a2}Z zbXQ<$^m19vlp#T<3hp+Mp0$@$2B@3Gs78vL9bmD7lZQeoW@{8l!8NhYN=S~_hP<>W zRvbs>V1Oi3#!h}`FCtc^%05h3dC-g7Ls3u3&rswfNaYy9ZAh(pz`a>~b@ooSx`Nv&yCOz%umX`5 zs@T+h9u48BH)#mHVScd{6w@nbsZmV56oUN;{A^(Pq9xi2xk!gs+6;So5*_JQQ3l^FRE zGgL8KmRb&c;r7o>kowjQ_a1U>0s%VrVU1p&QS!SKxlP5RGIRnzKjgM*AwMPXTmC?fLkwpS#XMMf}qrORJrMVRwFGHv-*5M;1=)Ko>r~YG zfa|R7a3@5(AidV4$|cFdQ&H>4r`}q6xnw81Gpb3I3DQV9dYflLSc7E1?YRH*Aie+H z83S(DNWW*jJLV5DJYNs((_M6v%J|OBV$2Ngm$ZvZy^e#(x9!b0DE~sv}7`t?GQs`h%K08Bv44{9Fb&A}b@kpA; zY2`uLDI~saTCF!i!TuXY1ZMk(AMu^G!Mr7+Krq{ckH#M-=8+PvL6-y@rh?{NS`5B^=2PO0 zq#_CY1EN7Uq=`=!-WIor8zs8H4$p4GGuY7IEY{2FM8`OTZa^4IuaDdtvMr_*Dr)fQ zHdd8%JKg1tBUX(9btA%(^>xwOStnqr_}cpaYq{&Pv0Xt|vx0j#bV($V=;qN6U~-(b zNqzlaYS+j7`XQWh)%I7`T5Q#28Ces*9kux%O!h_S^P6#`o>|b>g>y&u%`DGNO5Q?| zt5kf0q`@l{^3i!p+=3f=jP6Sghe8?w%k7MvlL&vktel4ZR&cvt*%WZMBv<7-7`J> zp)ojSWD}_$p6Z)yxm|MEM1qZdoR$qjwq$t>Zp~M5H-LgpA5?9k{>-Ny(D%yQ+7c2Zp^&S8ocfe zMtF@GWwTYdGJqyg zCrAoC7SgCnrS{D!_};xa7om@!UZUl~eew>~9`8P|Q&=~5{_od+t>+s{=f1q!1eD?$ zWqM?3$i9&6^og(n_JG^XfD-su>tE_qBJ78N_4AyKbUzs1 zHD7sm1|+*|2P19H<;YR>I@9&aHJk5m>%uSRoYsz}?#_i>WgB&O#u`sdF0|2^ap?SN zgY|Ot6LC_UF1(f30hXsnhOIC2jzwo0)KKQ#muRP9lZoNRMf^mMySz7huC_k$H=Z9Y zRNnXoBfQ~4`oCnn?WNBgxC0csFbm@fTNy}rFO?V*FPFp{%$5HfDHs)Xyb8llr$DC&7(0pgbqj{E+1up9Z zL)FwU7v+TrLCU|NBzAImMfe*3jU+j)N%bryN7OGm8RKOC!^(u&i3%%EI`T-H zL|)Smmheq}@9#?|{uf#0!X*@iX2xVQCEtW@cYKqgPrT1(9c*wK!`H{Pf`lllt>y8P zSxaYThGDHlx~xj76X4Y`GcQbW>rimMW-vy$9ruSj9*-kU@R0slokfb7ZHrtN_^<^U zzA{x(^1~D<2l`&9w_D2532K1y19`*MTiDBE2HkQvi^$^OBe0b=Sj|3yY4Jfoxz zcAr5>7VeA@BS1z5Y@D1$o@z`wV(K8v`xbl9z-d=6sp42|AY=B~+P-^KXs|u(sIS8rCHGIPv6 zV|(YB*{Gwx+qMO)v$;>V^*u)HvO=om&|gdge=zk8jV$^R{S@*f|Q z1Z%ZaE*z%Sn0dLWlzcTsR)CiaWjg!4E|Kn-{oti8;{Pu(hTGT7mPST)oXZ zYuWSYvL?kw7vs(-?1jC}UdUe>iEO#q;CtMcptLx6t=l|HSZ&_ei=bZfMJ$_3^wogJ zVc6!N!)_RRm5{6B;{mt77Dcy$V+vdvsPcDSuqib2yMu0PXRMv!_@Z;`vpE7RSs%#3 zXP0&!jJ6aRV6xkB;dHYNNrD8bP1ZUG$;ztedR~`zR@h!)vJf}I4?L1Xw})i;b?0NO_jF%u)=%W!+g|+8vfnUPsB(D3VOY zZ(uFtV;^lUgvWqPER|%iuuHZ>S{{2pLOlkPJN!eof1-y!<@{qt?(4YP&|UqOi_5EF zdlydc*jPf)8(kNu#-!lGXyDn^EADySj_Q=C>)lt)STVIxQWM%OIYqSmBRTDb8Gevwrzuzc*wZ-{&Ut0Pm{g)MGKin(%u<+QEX-} zwoq~ix@J=G<^EbeB4@d8PBb)%T>|DPtVC&$V6auAc*HFT0yesI8P-l!vL3Js`BlK( zF&f!kv*~%w3%e0DY_S;&XSDqAzo%Q)3fM@$cS`Wng!(n?z$@`SE9zyf3(Tb#^3|aK zrECsg6P^KiPfU#dbr0zb&J{t^jp9Lwf%i{KALh36ocT2#;RkH|xL-4IqrJrJE9avi$v04B9TkuAwz&|PhW6_YXsyQ2xgh9#cr)ly8b0F+zh}@sR|qPS z&Rki{K3Jxos@cO%e8KFS>axp>UzPNSZzkGtVMAeqXs1yzuQFgD^pwwW|9U}s*k-T0 zw0=(3m*%aV*uvd8sZ@!~b4Y-Kge&TKRh)gJpggf3WlX_raF*;Je#Cd$2J=5xiWOfB6k>-Xr3==L~8_BvHDW<)45`rS^A zwx`aQ{mQaE=hb+s3)>PKy+o*JcggD-t1ReX@CvAThO)(OQI+^$FtY3;0nja=CTWmh zw_m!fSlTbz5rSujwCEuhEG_d#W#(t%)g(U>R?>O*!s?}aNtypVhp(lh0>=+r%slrRJc3l8T2ivkNYQc(q*f-6~KRmJ8FZ#Pr7WqXFaXsbCoqjq*2IVr>E$jax;=#=eAu@>HPStlm+Uwgl4naHt`8@VAHm{RDUEZ-$~KC?q* zJU$A_ZyhS64o#M?3#@|D(i^fptSVAR-*Q-~=o_YmIY8bppo1-Ng=i*ZMLzzo(GzZckV$#AjHY?59`GVNz$eBvo| zO%)*pE%1cMR(YWVpiwM|td~}BA3;wGD2$SNX+KbY!m4Cszt;&KNi!ogij^!ZD|Flh zyz)VfZbn`i&S57Q&4}yfRR<0P1X_0JFsWB_-M@_3U^%E0C3^OfF3_C^Uf-2|9O#Ry7(PIg^g;ZLO#?eWeeM7wFD>={2`uL#M;L0 z6u>UXzj3UAVJ|G4x;eIjF1L9*cy)LJu^3ojFSW{%MYO1_RFX;l18C~Ze|)J!xCnEES?t_DqyHNonf z$mPJ$heh&Z31_=tJmi2n8@t2Ce+`U3{@KaDtz@5(9KpBoiLZL8}c=F;?&0|sc;vZq(M```f zb>>-S-$<8DuN3}Ume5iz-DOt~8`9`lw`Azj#u6uFbZiDUF+0qOB)m3(Wgl9{pe@eg zH)Mv^aaqaYe=(VfnG4&elNM%$AuhayeQIXyIw|?56zQPiGZp=GimZTj+x?27SC%;` z4GKcE{>i{|$9gpR2z@-{o(!1t{98Sb5~RS>PAit~pMFcUK4=-$C%zo0rBBAFZ^`zD zU}CzObtSr90$jnKK)VMW;ds7EdCW_VFOpd)frmnN5_kbZDXeXt8bxh@eu}z)bsR(! z@k9-r$_mhk8@-YzsW-7YB9d8HVUL-ITO`A?f_v4og^eQ@vs<8z2UA2k`Yg!_IPS5M zlmOAiEt2J#%(9kguoInTTN)?4u)yZHaUQ?^K z`(G&d+Y*MVmo5%&^UkK1L37C?$@(CCy?KiIy67|BP4C)Zq`}Xhi@hi6BD#s}66iuQ zkhW&q%#F(b z8pOn@1bKwHJYi29!SxOFcaCKC?k)CNGzW^llDyQWE zxiNCS>1CxJ6&`oIcLOvqB!or&sQ;R&*aP~_-NEUc`@s#Y?b8-cdl-x~Pbd+VrP?I` z)kJ!y2uHX*>t-O{SHZm*coPKZ5s|v^OkW_eRo;xm%C5bnN0!dkOorXyv{f_Ck#_MY z>b;x~s8_IJ*NK*e&9!0^pac&DACXiSu725N29X>}4tviGDn6TThc9*~{V{0MrA7D1 z5D18u&wVl{n4YK!dDNao+ju$|Ie0yJl__H!<_(|K~4Tzi8Q&=(1Zp z8|{(^hxzkSa~$h%(c?{KT|w`>L4q!Kv-l~w1e-oQtx==|KJl49AL(hE#oJkMCUm_d zlRMveX@imCWXJ;<1r%QUC^MY=R+yu*RB&NZ$>73X_+B%^qow2!M%+Thw+PC-5BMAZ zZB#A(;lmHR=jgwTWp`W1Gf|1KBD|mO=3EV19-|FDI;}xb5ndF%-nRpM<@LUuzUw{8 zoZ1RtZ|rLbMP@%=i;K}-6qo#bhGl%&N@l`kgI>0x$eR^CF`1z)+`axttYY*_3=2_` zPd_`2{rMPYs2G@)(atEN=WUynD`)<_#0%YhXK*-y9%h0e%WB{F)W@Fg%&N}=_+=+-cm%zC2 zN@Ak~rbkx9sTE$GeaSbCqldUkA-iGDV^j(g9-iC~mE?2DcWhH94CdciQ*Cuz+ot1Q z(tnh`xmnePBg;0LRna{)wl$T~72)c$(^qjyWEj+K7MDu?=B#^890{@~Qn>{(pc;jpo1L0*hS{48%^s_v1kmN)Q9{ShdILPJ%R zku{@`-z?kbbIu31zzeWbrCwSpz_5O!SDG+U(m*e8-fCnwQikk``A?@|hM##=WLc;G zR_$4ARO6@5*ao6rY80O}ar&X}*$TPB0;@Sdr+vUUu>F(o)L-M3!EG-qKBfP-S7eHx zNBT2DVlkLI|=}xoIMm8niOp#4gd=9IK)grh%;~X_^B8yD9L8>oJ>WdgFZ6~+fG*nY%PqV^pzAT|VTQQ~ialM;A80x2w?()hYV zSs4K41otlqz+gQeS!uJO;AzYXy_ti_3COWBWKnbBmn?(nHmaQQD~Zm21s8q1(&$_z z@Gs^mb#w`w&D5cbt8c-59l$zX}K|@aj5NyZEr+DBJlfiV<`SfNOLooQGAjkr@G( z`j?gm=iYYT13Ue&bbHVZ?b;bdm7#K5y?ELc0VeuOyvzKdrxb}lwe)ciAx2ZTV%Avr zos2OuMo3|Vk#S4bg;8A7_VzF8OMf8C-kEwcVDuZNrLvKdr&DAN6|V(mQ zc&CH70R%`Ip)a680;P`pwXsP)V`XXJ>&*;EX7<3TFIrYgpIDX!F>y;>I3{AFh$x4X z@$(y^CRO+BPS5oL{UQyim#lJcQk6)b2kjE9A?J7+MVB<0wT7GohDZ%r$c026Fk5TL zbe0 z{hL(fP_o*}`<#9v8Ia=L_c%S#8U@x0te(_K_potvt$%$0j*Hh1&&;Ebiu8(XdY5!P zM=$IBg+{S} zhA-PLk!Y#o2GO?*-90CT$vo6%qoh!2knqp$Z`rmt5FO{54fXWl&5S$@qjAv zfHW;yM`!t^Pb?IkR-9w+nYKJe{X8glcDFRxbqK-*YcrL3j5vFmxsz~H21_0ITTSf%YVNTaYUBYgDV ze)f;^Ee(>5Z!%V180n18(*XH9tZG&wbpB!n?F6sGQ~i{+gMSBjs;|y2P$ja8`RaDM zioQlLV17&X0PakhrcvAwWrwKoj%wCX(Q5fq8rAEwgR_Ixxdi)Q?niVe@v<(w$tO|r zh-1 zjr07DI99*={RdzA(6kuNDd%;N>UU-_ykX{8HBj=i6gfr3>%!AGR|Gw>RB{=5db&kh zJXgi(kpf1CCD6!X3JRltDG5Ds%xhh!*m1Mz-?)=Y}W;3bU^W&ILgo z)Gb*%ta;~ru8Yt4oQEdq?Vfw#8*x7T3Mf?RWxFOFCkycT64r6|T&0%Y!Ef@ZVAsRG zWt-?w&;e0Z=+emEnQ6i*X`-BA`P3MGH5gg8^3y-;`K<|4ziar#H1e6X+7}mgqgI%? zQO_v(6N>aw@#v6f+1d0ec2e+xuw-&4>g1#+K1Vqye6(7ANLdqt8UiNIC5L4c7RaY{ zf=-z_S=b0b!q_uP3M`kPY8hZoD?UYXfWH}yqv2F3yIx{^*+xGGqK9lDdPNPA9sWD~ zYeMjyG|Fr`lfQ^;1;4G9#tknprJ`ZRf%URPiN+I76hI5VkvXxIu7;Xm(*<3~R=JjL z2Pw5|Au=jL69IoWKPR$5QWKKHLOxyyi|4Sg;tNTo??_zMld^4pdpG>~yK_zOo4MdZ zE2(hd*41V+BXx$7pQK1F6@PB-4bflQWV$(zq#41BNG@F=G&)aR-sn6f@T*xF!JuN{ zUCp`|aZ;&KCQW$reMRhUx}I0Vsf+2CWynjUhhsaWdhoRC*;fQhCN!y%ysom7CcJvb zk_jVp)|_vk7}=#4H`~UJl{yRR<40eZATi?7TT>P&m{R>z3|^c)j|k(LhiL4x*<>d0iV_~_wrbc;xYRY2x@IrsSk%H z#Gw>2j*%XATc*Cq11cotG37&nM!8zP%Ih#sM>mHbC-|I}z7DK1_^f^k21g6QzNw+1 zAq>5;q1q^Tuc+1XW*PKe8hIp+yKBByNjgm1ado-q9`Jniy*PpK*maVJ(Jt#ig?-2C z^C>3u_5Mt7`rQ~S)9tx^HlE>Z}T){(AUZeWH*D_{rWsU^JtT< zfiUCXY~JA30^s+E@%3eGNA^m_Urkg}M8Vl}`W)bFWdW7-7wVrnpJq)@n&zay#>xWsmHwd^W- z2}y+@=*}p0J{;BvFN+F=T|B(pAjt@RKriGQa~TJfa0|Z68qa4U@c0Sz^(IClGJcVDR9DcJw zHzYdyT(DA>6NwF62bExV?Sh)QFE=KyY(ve;?rhxJ+Yx`7Zd%6X-YQr{3g5C|Q=QqG zdytZY#%viCpTsGargC;hmD9NZ5;c~I;4<6`c{VkANonsN@P8)RIY|o%yktNaipZ*m z4swZh8eUB0oEAWo>7Wwz%rRq!FvC<^nRI*bDtVswQL>Dmz;}?Lb3C9NW5R3$iWm3S z-&m@&xvcHlX9koP@^SSNEf;s>we%(e!qowH`9~lg1j&ET$I@f&18y6HRh-jNt3$Jc zi`W~c7E3F|-BPSTs+fK+?4sl{)TL>;4g!Gn=^70ztdHoJv$kPj#m-Qt`BRsTCUfh$ zJ{_@|Abtw6gT@V+Q_*5|ZtaAS)n}b~^3z9u*lEdo@3Qv81|k1F zuNDE+ZYb1;mEB}1vet zOapG1j*=e103`S~g%F zx`TfEat3So~QM$56CUOTaci(A^wzMn@qp-XFI2(d69PQZ+9gjjJ=$IrhpJ=w>EnxCB;cadZ; z%O|+7(E&O0VNF?ulw3;z*=Bt2*KdeEX_M7Jn*3_mQFbdYQ?b^4S*Tum4A{UgN_GY2 zEB8p7R9NVWN6F!M*FNrJ0{RA`X8BI9Egp+ycV`@RqKxDyp=H%@IO1g2^N;R_n7}eE z`;#=1KY<)I`;_-la%gzkO~vo#V%2V1bdwz9$t$5t5^Xwi`R7B~OE%rZ%Y#hj%BfrB zY7CZDaC_-Hu;8zpx^gP2Dnf|~JU{;)XDtZy!4OYiD`7ae1`(}iAOHpB$8meY9M||V z%1`{2Z?z*nZ5)Hym)T{rmckpqU-&x{aF#7kjwPKgyf&>g%R@e=hzi%8m>Q2HZBgZ}!qC4+VA6oe^sAYQ!a>Edn*NouvfAx`gCt z)pVagjj|F0(lS;XD`DpK7`>t}2#UP~MbLQhOn6TCSlA}4qBl-D9ivW;+Xc+OcRV(Q zY7_}!?S92^dt~Z-<=q*21qh;e_DD~IWXosLyE6veAPPP8BxE}8N0ch7=o>-BkdaK_ z_eSp|dNSbF<=rJ6aBJsP!ta_ASnXf!Kj79T>GQyM$qwE$WiUdWAK6FFk}^dfC&72n ztsRstFo&5m_1y^gj<0>PJ62}G--fi%D`fpUv&xm3+24FhzKw$5O#Cjvp2$L|yIId( z9h}2C>IYIftfUD~rX2EGGGSTd-k9@2>jMt48vM^npwOOIz?!F>)+)%OAByMAI~#SD z-z2M_IM0c0_UP=bqpa(z<3#-No6{{h9ADjOxo}+022-Ok(X-yO9)h^gLW*hx{T|Es zM}<)KnJ&8-lMVTTd%&-i1&grZIwm)brwfH=p+2G${`(umgYEVAlj1+dgqlpsp3ag* zL_2{TGy5*?r{wVEETQ5{=Io-gLmop@i_m22JJxD-n)EP!Ig#c6y1ZU8=^dow1qV;mm#yYy@2vmj|_eP6*xDm}q;jWMiOI6gdac186g8yR`H>j^C6kSy+Qb1wJ>xtYrw5udXqQcwm59)-*xN* zB9yJHm*6?$tF81d`hbW@Gb&?TnSqS)3%lX>1Km!DcrG%ma=KWv7SU6dSH z-nLWmSm)b2Qh}bUm$dUTJaD_+?{N-#M(d>yI4DZnq{2kPVphG! zesL$A6S-<)D-WN}gM}=es})x`E2uYm2-rQVjBaGKOi@!Q<>fFFVD5b9-X2o$&KROv zGkBF#a;Tgsg(_xN7lA5IUWFgh7&NIalbxbcdJU=bS?}A((7+737+q189QdeV&KAiZW|$RbY>3Dh5d9*>>bh`l{=-)`?_dAn!qoo z?-2EN8Vg5vPAY}|-i^U(UHB@W3~20J9)!#pH6a;d_1cM@g4?nh4pw(Ii)%c01*$=B z(z|+grf)@54Fm<(2y}stgN{RizqkB1S#GUZ(}g`?NIMP7Lv5nu zu-RKn#iQe!?SI%0xfQf5sD{W9>Vd|jo<^C-qnyv7&$ivUK^My}EHeaHKJC2Mwil_N z{`7*ErI5xuVpMEMY^1t314lq1A1c4`A1#e1;IQAL29V@9Yhc%G+7B4JhI80WKiK*{ z`)y|{2HA3YFE^drFN1zEkAqW?I0ba1;Y>bTO@x)u;JBNg_{CVtXt--W4w~ z%T@PL@<$ZursA8$D084ult=9dQLheM9JWr`91c<`Q~JGX{D5hxJVp<~)kWevGoN{t zN2#-gkNqpSsqU%n<^JVSdhAt_wRu0BxhFW46!MX+7_`uJbha>4@zC%3ETjNQi!Kr) z6?Qo+K`sEbYC8g&sGw3HoeoU5XdQxN!z3uun||E zlopAxt^IINm$+WC1Zueog*!R7**dyZpswQ8!jE2{u6RlVCrX<)Dh#Nb6zQ@>{DaV} zq`u|4+-;1YH|~dNKRs@lrhcokk2aFg-SlyKpLAWoCi&Ib>RkGUs6g4n&V}Yl^$z|X z?~bUm0M+{u%Vn9sR8|FaBRh!}7~dAN_1-CgnB8)+V>1G~^DT~B?q?jZP=`q#911NxoGITE`%U4Jjk2QI-K8 zNwc_XW)W3HA&iPB=NlA@`Lq`=7LPgoRh^cBZmG`#RmUmiFW5w@*I z!nJSy-mxsMctf-ZBsdm_?fpCBQ7aEIlFaJzPT=bVeO#Sj!&IH1&6x#~ z7W2mt2^LQ~@1<=dSaD4$*#mvhS52%ofIH|~&dMMp<&@WHU(OvWSn)O(l6AQNl9NYGd#ic!WX53^}rmVpCTHx^R?m$cbzst0UC~ ztc`5kQ{57@Np$Qi?>xE+8tRxVa11f{(iV;6hyi5X{^j|1Voj*>cvP28_P=Eb!;5Cv zsiowQV6LX(?|H#ChzI;Z>9SkUMGX{sYZB~QmR-2_9s>`lkHgsBz#Dh4C^ybMs ze1TFb%UcWd&PZ96=Z&ONBb-hKpYX6-nSfx}V@^M`4FW3**SHIQ-??oWQ?U`cK<3t` zw2@%z9Jv^~G58J?x^4?uG*zR>jJz0oO5O>ji(1stktIehot!9X7#lx=$xIAcY^P5H zyaA(`aNW|@b3bbh4>YYpMT0-yM{*_*ota_ZP04psq>zd~7Mjeu1fa+a-1Mad`{#a? z2{fPsZdG#@T%7aK3KbszWkKrPk8bepP9Jdl^MaB&ANA7@eenFB7o48+(IWnJ(OsYG z!Y*$-k@LYvxU3fhbxA9@e_2rS!AA|g?K3ju^^(zl&qo58A?Up3=crNjjzH$`1EYRr znf`Iv<;F&;=U6D}EA165kIjKL@6(F2;Mm@wwNviUhgeOjC8|r2$o!4fxf(^kq)1%B zP4emGZ4~!KXcX7^tEP<}q_0KO$PxV7VMl$D8G`$~mwdxA{llc}$%TvMY$S_t$E272 zRR>$7PI?(vZPdUljo#;TjvN!iUJ4UJ-QK&oMby>V&qRlz3);jlR77FA7&lu7qaMIz zX8~;F$TC^RQDeyZ0RQ&!g9`zsbtv)ozN<*~JF~~GFk4QGDLJrB6@U_f&*i8VZlUmC zFl=G4H@P_&#TRC;*ws)?DAx1fpR$Okt>7gfiJ zCa^I))oXcBqiT)e3<~Qtsc!Jjg%(5TrG_*@eSVvF0V{{oEUTAPK~CD?*5ie>AEg<2 zF^;1TY@?-fZAQDLhK0-exi(ZOkd?b5DuEAV2?mV^F*G462i@}cDS=SZ0xA+kK5!`y znmjfUTzHZ|T;i6pYq{|)a>n{10%Am5vz}Y+?y*_IhKM%Yd9o;F}LVy>k zkd%Y^;W3cl?UY;s?wPdr_k=+yo27jZ38fT6ItpVE{`~nEIL2fG%IIElccw|QWlHQcRkj{?+WGeH zuBo43TE0~6zgLrm)|M|9&NhJr*RZmwG)kUKk(F2moWRFTk%7@Ao38M~WKkiXp=6{Z zK6Y1~dE1NBZ>RHqZdnt>#7wv_T5L3=9afb2-vJTG`;uPZ=UXw$NXA(J)pmLrl9n0O zCY|TVyoxhBQDfW(z52M*&NOycf5Ea3+C~F$2VE0_>M-j97X`QbX_N=0uqmw)FXg1m z8WmM^Pjq>#9%vGE^b%EH^uu6aOwWm2E-Q0SmtndOR{I_9odO)7#2aOkao>S=;y{CL z-QJns#o{FV&tdbz=tF-^RCnl-qh`VOSJhgE#$~AxV>ab{&rJNSb^ zkI5ZL_mn18k8CjLGOL^J^1V#g@s>_lHUWfc!q<@USO=Sj*G?`YB)xVd$6lNm`MsXE z-FBd*!2g3_WS${}3uIwX%x3VNj($)m)J_|WNan2K4!SMk7xKqyXUk~D$IwS*W;>FK zn+}a8m*~R#92>br>?6g1*<+}L>tUDqKb(U9E{W8$cJQ~#A!rAbQfVALYqk8e^i$y2 z?4r}4=DS6trBNcQ%e!8xgSfvlZOaIdHRNqOv+r!&`K71SY*U=XE!b@lNpj&^h!a))Ue1Q54V;0R7KjygvqFS_LiR!oeSgQni($6Qu0$2siWeNO+)X!C>RyYZbcjn?w9SLOJI|6 zcqVMZku4dEt+BAS_x*OiMg^=xq*)9ldlHSOq2yY_+8ffSXjG^xXKjwe`_)N4{cMcD zL94n(0UH?5t}c!0pyy`?;m9y5GQYzQV~4YPl@H#hXBEfgD)q8OWGkJ=UQBwTnUOcx z4XBp}W5h>HP{;^SU^4BD?`4LQgGmK9ErTgG zq7b)drN))2YK22T8{zmdxvl#3rFnP1e(onF|9W*!``lk&RHe%bf>Hv@1l_&f% z2ACTpoe}L~tUqYw=>!+$xK$duI~@|t>t>~)z%)GeSfb8THp#IHYNr>bhq`!%95Du( z@lGs@f;9@0N2}Pz^bvF>?euDsV$FlPSX!VsD#FnAHg-`o1m={L1m1s0Ir_e2mDf(M zW&AV55W4I4tmFVqYjnG`9`cVmct^u8P{8UVIyRQ*^?MpiYlgQaIO^+Q!;&_ z;fLPRST-fuiH#vU;dl_(7;)F;FC2N>*}2Ps>j^94(j!|0*Puy$CAk-o9NsJ|kruFe zS?9sx>)6}rdO^EiKKm}w%dUlPnA)Vmm7tn+D`E>9uV@shW}E9MIY`vhQ1O+(d5=T}T7PtF zVSR5>yf|!yb-yy8CER#rK_d%U4iflKGbFwwUhLT!k*9SA^aI3hTDGV6Zu0{jQR+C%T5-UgfMK2R$&zU7FjQAR&w!zRzx;RvRW(TusS z_5R_*o*R};12(GdKLuH%qB(7{zuqC3=*?4RA{8aJu4jl(0C%7ok#S&dLR_KnVe~@} zy{v1F@wHmVz9h~!kbj}yio{J7A0))7bA+9A-Q54btD@ejvW6MQzq;x_OvWel*YCYY zZZm5$abf>9(QFUaOUZjE(gp4KVV49Y-gP01$YI_MQHRHAP|`~Rk=XS?8^!IU239&P zx6fl+bcbq>ti`7nR>Cw$fVatZ(`m{k6+XR*wVm~Z6h&9FdP!;Ypr9`1KxmG(eSf^TO!MmSA@5D zchkvVy&3B9;+eWEb{3UiR)SO0a>>zHf58}B)tHi4w~ z`+=v(B4!|Q;VgKL8A#Sp@)U}!qT;VX+ebG^g3gc={HyG4uq3!C7`w`=Ve@7mD26>S zMu%%I6(4%r`;;(_drW%XFN1fDXs0zo9@0Q9RrkHkU#^}5@!V;}(4c?vgJ&X$88<0z zLwB{InFz((mvVvh2Oh`7=SwoF?1`{lOyYpM4k2fZJ{OZ#-h~fa=3E!G5@h-=hLl!5 zWJjS!M>}nm*E!x5fo4{z|9qh-`Q)FeYVKVA2q&kvQ4rbSjg1>?pK!mY%ZHCx6z@7kpn|M=?<$yEGdX)~2P_fP-x>v<2pa{IgIK6Ius{i_pc7bZ*=;nk;|cZLxrsDwj1bt038EB)ll zjU?rr@o<4fYZ!vEDfwoKY@*^Z<&25%eLhFnAS2QaC0BhOy)zmW#)j?P)6~a8i)bYG zz=T5~zs&QY$5_}LCU|#* zaa!~u6)y(KFGmNl)D!Zg577rx&Kic;IMMZHn$${|_eA2z`Dtj?}ww+0}hB9NeVjEfl#* z#b^07Li)N(lI@2{?M&a@F*{~zX_5nSG!6mGS)j4O&@)dR0hoOuG>qCy{UmA8dfR9kw^e)X#K$a~&^^5~tb$7M z=@c%XPy;&4CT&A_x-xrOnfR*mQ;5YoM>mX=f!UEq{!|}M*m&x0*EMj%Mn&81Z`!7= z*vSyVOtV^M%cR6t537+I8|+;G8X7q{E$e zz0+Q^EK3zL|7!&)bmL{|eXC{ZAjR&d$X+T|A6*hs>6PodO;#G&D(exRBS#`Pt4cx) zqBTlPw8j#S7>~12)UHCSX(EgEmL1+_XpF%#mq@~_7A%WYJ3WtH5mXhL4wW6d$RlYL zZ|4jQ{w4; zut6&f2L58z>f1hj;A&2g&eVT>AFoVY@0lslWkaO@sxM)BEtZve?3Wt=fx=l# zjt0Yr$6}t%Q^Ik!oiLA%4Z&%38Gj+KFnqp6e#~CG@Y2-(%mHaqGu_q{UjEcPq zwXw2=-O7GtJ97-8hVuiD(~Y5dynf{xh=1G7+s7-N+$QLvPek0Ik);7yOc7a0lC?e< z>Xhht3!)3Mg$IB!rVSXs)1d%*X2=cs0pPXU6LnjLwT4@?`(#Doc(+kjG07k^e_kSs z*Kv#_D1-zN3NX@0@yZc>gX*p0za0bHQ@s+k-@81;f~Rj6{^HkU%S(f&16Ft{qF7L0 z@~GHiNw#DYITJPbyGh&0-|xSMzd8K6wpj})D@yZu(Tp^bElCmmxoE~w|D1W*l0WA} z?DT16v!plV@ldxMakk9aL}Y{WGkUlTj^y0m-&u1$=tYk?fKq9pMj5K5<|U{VYO0O& zwI0>eb!3u9^#LQincIr=wO{>87iEzk8w`ppQtrlE;U+7YaFk--qeu-@j)!62OB2WhlnW(6djp^JYcjQU^O6JF1nVa!YO^KFm||tSqDO`jB$Iwf zu&$TEKUaqe?JHgWgaW4uAYCaN397?cCI$ z|7C%ha&^!*$Z2jGj~hpPcUl4P3dI8D(RnJ?WPt&_cd76u_!kYrK53y^mkT@+kf}Lk z*ZhMTi2ZF+oe^Uo9tPN=j*NZ5meZp!PEA)n5MLAPDkON=5n-5F$p`8PBMtWBnb5^Z z58ius5>nXcDxrWtqa3rjb#-JT)XFx}6_99fj!6pES%%Hg9Q9l^2O^Kr23;uX099&T zP-hqvUeDE-yWn?7om?4FsJ0Kk|4!@yBJINOt>2IUTaPyvre8M}ad!OU_0oE&K~@S{ zagCs0M!WI>?9#4lkzb?*em;Qd#r9Om(vTA7GE)VDu1TI1e32=Q?25`~Hb`;;*YJ1H zdAu@Ji)wYaQPA&k#^*F`6jZ6Z2nKDnXpex%?9lYcki+db>4`kaUf{h?&d@E|55&v) zr#0QMKg*hSR9FkPqz92(rX&ZX`1C1Zr${eYkQlHeWWjCzQJ)1p(`y-Q!}y+kz*}M; zTZ1bYn99kofA_ZTPe1+b&;Ixu>2iu)LXo)P<&hIx?PAMv;a_X1w{7e0xUI+APK}); zasvu2VQrKaC~5-LQOXr*(^_R0<^kzUW(a1fBV87*hy{{^nnV`qWk4zz>3kJBtV%-* zRVG6c&<97lN`rTThJDRNT%2C@i=fvbic<{a&y*pZ)`bAR`5%Gd&91VJ@-`~(L8*9)5wAH z`4BES-~<#ZSuyQLwr))>E0-JluIxBLk1I;W@6U-7^?TeBHqULNfwi$!h6!@j!B?b7 zz*UkKFqWJ&9QpZN0-50lJ;iMHe8r+R2mZEl0omxr+x&7XMR*6r=2B!E6}xOgvtMOs zD!ZPy>6Po+rGZHjtPAtt3o^F zANsm%=@}hdTn89#aA{NDziw;yaa*|9vGLT%HZZuk&?PawvUQW|_&}^w>|5lwZ*m?) zT^ZYDRxU@Kp^pT2^9DRhe2p-4_A~<+lDg#8)8ax3q$2-4Dy9FNqUo9J-aN zm#)N0rI*B=-eXl)1Ln?M;h}$IJa6mwA%M^8f{T@3>JE``y{egK%2QRuO@i0 zce1=jxYg&ncB@ZKFp^*|^MS0qz~fQPEM{Hd-FR=CAeCx@N|w2Fho$E|J3W_#WC6QK z1>A@C480=#^OITO{x;{SR$zdt4-w8!udCRtr}i#X{ar7A{<-l2BK{x{v7! zb<$3{?hTCM(qYswhJLN_`i!m)&i92p??UMYCLYotnp8U{chZ0F7U-FH+kj8K-p#mlr)M>rAQJLi)7R2DY*Sg!VI7Y(rb1r zc1EDTVYT-dYP^;Sb!ZuLtrcfjXnCd^mVC_ui}BsR{Foe>K(1Lm&t{4}Ly^-|Y?`EE zE+iyQ#d7qu(?1r%*0hvd5*uMJ(XRYk3F!S@$yRyaaQmF2Ryd6gVjBZd|g^-6n zo|$iWpBy_OQ?Y@rpFH4!sd}Z6Sx|fDs3(Vcb$CK-UeJ-A4*T?nY2>!6b=z{cUyF_S zvTfZ1x3%insbqlZ|2diiLAXmE@MxY}uD~r(frtajhMC}zH8}{^kqthIySTyijek#$ z8qGe+jg@9Q`=}L6rPtrCNSn0nZ?phocC)sc?4n(#eb}E#hX7>%c9&iG&lSIV*H(dj zn(oexwGTULYEeuw56x^~k5!fV8^P?_0#Q&VJ;ZEgX=TltEY`d!C8ZDt2=Pz2#5H){ zktTpy)}<(M2o84EI+tMq2DUFx%Kr2A%0F2&m0ukDehj(CO@Vb|*^z9u;f^2A_EY3D zD)xP*TiGBxsMbSZ)xjvE0ArZ0YfnT0+df^YG5WW`mN{$M{^&aCJ|1#>rtt48@En|1 z!WQi{W#zO4(jcn|J*c`Da!LS+GC=!j?p5!(KN`5mVZ6%FJ|Ox`0PgR8kF}(a?)NCA z%|Fd&7eX!uSB9Qs*6{nFFxpz^SLsy)J>o=o$DJv0BD``UsuZqOdSUtH*3e!^^fX_8 zAYRW)2ric3r(^|Rmh^jU3f}26^gM4`cjS|E3~fc&j|QD~Ks)+o_+F2`b>z4$U9j6~ zHFgMq%l%TA2GFU28V#k*VRfdY^~+5_rRq|7Gpx*TaGHm`*OiU-h0rkJRpTz#Ev8~D zg6Z2wez1oea^uj53s!r(4=5J6C60n^k3LJfeDlK3z}^hJL^XnAbm_$1kS3SUq{=UN zn;P%{8@d&S0Epc+(3XRXBwe{rZj=>K&=6R77?FZ?adCtY4nnSW2glu zIE{aMQQFlPUB(z_EG*cx&5?4x-;m76#igEJfpa0yJ z#nWvU9y`38=u^C}K#e#xvRip5czJlY#m5MhC9(E#opcw7ALOQKqs=?*F~Q47#K!P* z>fBat2-&dTr@}Ts?Xg}0rnQpx6#H2E&1y0aN=|YH&1XxcNC+C7`=xAT~Xk~=0ZSEz3 ziXBVddpvYQqH5EiSQoT7C|y}7D)FsmHjORAk!cbfM8(hv8&Dw0$m<@$dA65gODM9N zip>(WD=sT5D1+qcv?ZbnYROE4dXHu~Kb7Ag?}Sw%Kd@(ZkMJ?5D4WBZ=az~u`(7dq zp60v2P6FpG+B67EcrdGLdeMxWdEJqlfkp)TZkTcW8L5g^cyRb+1b`k2?kC3%f1+a@ zIVmR8pO-EDjcwqi9VRiP4znRy~C!F|o=wUR-}6uX8ZE2&tciJ4QE&Z|_JNjm>K z+=&1`M}R$Ql#XD4_v5}b- zl||*yt)WL`?Q_;ZNMw9;fYua5Vr~% zdpv{pq=#2d&0rfK2jBy+yd+7QJr^t}_RZy`Y0RJHnBwj@zMiNgqdam5lbrBH&Aw8m zvUNn;@$DibC#sg$9Lt2d|gXJ?vN&WPGk9Bh@y%uB!CG!eq0F#Dd zy=1%hXR}U|5#er>g2yh_hF*9IEeK-ZidWB8N|a?p)c=kr!jI7zX`DRP90JxR>L`xd)1QUR@-h=qfVrfL>#^kPvyq=F=P zL1;pgs*|n&@*pGd{bn#VbZXSv&>}hR+s?fB6 z0*O(UNykOw|0Zt{<_{o8RmP-xvT%Die&S(u6>SI_QOQ83;$x^#v^sVa6;Jwcl8@uQ z_46Z*3oU3W^{-cu4sK$|eKk9>%qkG?A;tDk2e)$|z zG|d@0i{B~kj>L-Nb4~9J`H1GThrfx5ODx zc^5vqJ!0n!w@r4tm>nJUZ42_=c(40M;lD}qev50%P#jSj)$<>B&6ASQWkftpYF1u-P7tvJ|%eeoN&^ z{45F9c?V2GCvk(YmT#axl~sktYd@Iks9|z!#2R-Q$4RV>r+t0?G06h1*4NIxL(;g3 zHTSi#6L5YF*>&VlEV#+_RP1dZGZe8xE|Bh3fnhps@@8;;?N!_bVoxYnnB^58j%=Pq z3`X>IXsV|HK4tZ!%=MfRBL~%!b(7|;aO@wWk|IaWF?ot_7!T>4-iRq2I&`W48&S{rp<+)3kA%++sU)=TmN z%*N?Dhzl;AekiEYt6cW6@PLVXZl_`~6T4Nm+It};5Vy*f%~%?;u!w4B4hN&(3`?_D z&w+ELnbfM?|EPZx(6ALzxMmsD`MjkNd4aZeu2JHS&*f;5<@hIY>&bhSd3!WuMeb`d zC3eV)u6+55G@iPqybIf^7VWC&z0w73ByLKh>gKF6#q~Mo#Cg1w;8JxtY|Qf13r1wS z#$Xf5u4e`2L~!aW8uPFpSd8lpqUC9=zt58i_=8KsIlVH8gqEUHkIeK6$RQ zZF|3~&N(eJFY@w;LT1@s(EF>EgVEz&!1fna55nQkh zVDm5pt_%7AbD7MismdMDaN2Lc1JaR{*ewS*esL{{i|mOkkldZE>j+C?^1>6@q8Y}i z-HJW1S9XCQ;R>Un`pJ`H=Yz>UjK2NT*s9SG$G9<=>=4Ja%JSLu{(~%uy1a>P6Lpxb zf;x$-nC9)ysN;aKJA_Lvy!GRbcH3|Xx83R3iQ_mIa7WrE=;t4g>W zfCkvHVyxYAemjHS?-p$_Z@IAF;|OycJbyXlmin}&d@i=36VVy7IP$5oow_0CPMA4v zQ{}d~2U!#Nhi%s&w-q*aR+Rz?npn|}`q*!oUr_`mpPf?G$a={sZ)ZS|6IvX_E2lx7 z(3uln$MJi=`*qutr5B1j#!h}!P4J5FPi0*~9hzZl>0U9USzRPo$!+g)KP;@(BfJJV z`p~I_In3V^k|%YY4P?liI*OSg5^PK|tn6dv`#XJqX)88fl0-*4LIwfTtGp!6n6yn} zB4Ie{i4hHX;pr1yn(8_PCQl9>2F#ON#^mUmw%RJa2Afr>d5*)5P zOm>Hwc1X6qD*)IL4Kx}yYG$`X@-ez%ak{_}Bk7_yPrj$Q6NFssb;R7O*mqy(HgXS2 z_H`?ZeSrl?l@~tXvBUc!lgMU;qjhv}hr(r}>hPRcy~QIwadF4&$LZdkV!SQe+hsi~PFF_`4M;d|j634YE(B!!VXHS7;nH$2ts1r*#ZC z6;Noq2)$^!D|%!GIbBSC;&Q*tiI>E?B9_VTge*FuU zMX2>_H}goD8^&-0Bd=w!H-^Pk2g7A#oHF-B+GeZ>39F!2@e~tbVjFFW z(nRY&u9)dc+&5wEn$d_Ehha35#cvci;{>De|3~+w7>NZ&*@E>~NfI}|u^amWi>yGW zr`XLD*$95)uU?V1Xp^bs!sLLZOebA1|KaRjsG40sIusb+i>-$QH^?KwdR_|j8VyD` zOe1$VaTt6QT|<{*7|vMNOM?9#VH< z#}BH=A<%BXliVmzSGEZ-8n#N@%r*)ja48yyQ#IxdK9oK{ck&%M1}t>|nM}7bk=?DR ze+{%7tXP2cuCb0aW~{``#!<{8=vDa;2W8MK96U`|)=tk?A&rs_C0;#L8qzfiD!q^r z6P2UV$X40e$)!r2xq^1O^7iZ@-CJb`UenckF4vlZJI%_JUNxw^FI)&QR_Mn}SEhKj ziu;6F61R=^58ZdrXRbxH5#B$ZPWHL6h79JSAtBbsC>DY%YpK{xq$R4>FF)WT!AH}e z0HpYxO(a{gIG|71u4q*K!S-sCwpp95Jnh#Y7}^cZJXLg=v}sO@7EMKK_?4lB||DoAWsc0j=_32Qj~C=b*}w6p=8|R%l}?SSy1xW`*A(! z3!DaUW82*pD@-L(>}rZ6VARwHV2FnG$-H#wcK&2c97(p*=n#BtZyJO1`d|G1*Apy& z&>G*-k@%NpZ=Goch*XMAqR470wuaRCoe9jAtkU+#nzV+0Eqi0#Pe!wRjC_*Z6%*?} z`pZIFb}B9|MK`v$*x|A&n$Z%~CP)ar2g>J1;WeNTp9;wjG%ycl9i!L2HgeTsWY};B z7pQ@ZZ1@r1wW>Jt$FnT5;+;xuEh&0wWW@<9S#gMBD=1P<#TEoY@lNrzuwzgQg4IY8JZm6zD4$8X3*xr**kT3qU{afqIOOiH zcTp?^5$&L2yI@P-qOFnjEA*->f}5{aNbdUXppkR%L*@vtKvJqkD!KQ`TAxb(1E8VG zXWj?)^y_mvmH7}*06glzkb)I^52}|<-c5=%E}Afh=f&;|=$eV)G%xPEi2uDa{0gKQm2Su<5ZiB1DDSRa6LzaU@=v7Qt zHb5;btT3}As5BBQiQW&31G>IT;sa6X6B8(itN}Bg^EX1HN51)$a5nzS8|F{3)o9(; z{%=P|?c?1E>LS(CbU3Nz0wOnO>Cgj!b2-+l4~Lm{_GCVDJWeuh+?{{h8EtDc<04Aj z*c@WV#MTmN5cMe$J|=%sq4c1-!LyS-Bq*N)98qV<9OVVztA%R@9UD6`nh8ZqWL#+Pp5j|V05nOY^>3wxy zoh0n+7X(u+5YZ&AtR`EzZH?VnCpl<^^qmw7{%Hdhi^Z%U?02j0LE!maEXn}ITUQ)& zgWQvLKonFrZ)XHFr0|bG(;buLUUymAugI4g=sqE;BuI1^FM=G_nP}F>aAH@y1|x#K zjc{>=nPq-N@8S>(UT(Z~yM*Kqm!s^ycqhlKP*Fj#*7}QEX5}fmWr<>eT z{$XfKllJzP7j64~;`}=?=f6=NbNmN+U#|Nqx&t%m$I^u0oO$u0<8+PUtoX4M2WZir zW4h?U`m5*EeL}3mnkB_?ody=CAY=$)oFyjh3&tsv_ABHT?40=Kw=R>lZtUsXV}+A# z6bn}QEmZ7j!CJ^ceLyevyT@N}jJ`)-3BCwS(W|}MmB{Z0&1^nY1g=tNur1n#P$Ov+ z93^>lqrhqNZg-wW2nf65IqQObKnxS{ME$8@U7;=4j@#~d?6Bt~`*nIBmX^$0CdJst ze$q?o)o0Z2(H|?A1-aHq)%IV4#qT9|}kA}`-Iz`B@fA_ZTPe1+b z&;Ixu>2iu)LXo&3i;V+^L&kH~;m>mx{gGG}pc0R)Q)I;il4~XBH&AREMN+9)bbMfW zfc7^AN&j24@tVhcED2B)vkf9zUHN)G{ReY+VV}O)g)Q7(Shvk#j|i*UVO#0yg-TYo zQ_m^~NC9-=>udZq6|bt*6hFBUG+s)-}=TcEepu+T6g%7Yi_)Q ziMQfQ@21$h6zQa5GZ+j_M(Xpnu+oT3db3D}hKw$GhU_+fU1%oVq+0T7r?6~#pN}qQ zUM}4w$L3rJEzt)gh9kMP?g3dUzds8DUrp&`t+MNLP$9v54^ysbFR3FZGXd_6Sox~Wty5xgnW`*Nt-Jf-mn1)>K{kJOmni5o;eA6g;6^^sT zvrL@jS>;_CnE-|t-QefpU3$}(9`HC3o+byHcMG~5N1X!-iBr0mA;<}J;~L%#%wa94 z6MpvK0@=XLM(4)H{=HV(mw{q)D3VRZR;f!RSCw5+Ik>C>({>Yv0f^3$~mdWGG*eCZyb6*Ejp5XFhsIgv_iRKR&!067ack_&XuoE>U#B@hi%+;$fEe11Nupef_ zAT6Ic%xLJ(xY@x0J!b8X~Vyk%WRe8 zmqgw`V#ko2p|up7Op!Ha^do&trUYwYSMlECK{-jIv?g$j^=UwngV=U1a2*)!+^B376C{%-wx^ilx9CaDpRjAg0;kwu>Q<4R& z+t{-}VhV-9MA%Nc(hD3uamc4n8(YIHAM_PZqk1G^(Eig99#J&7748*djtEV zm%I)IRZp|-)%xw*^9Q~+v~Rj{Q*eXe!n|zB5>e%}wgo3+vZYmmrU-MwYE#5rsBqmT z_{jH)G!r)3-O?ucJ@~zXf0DjA%Mt7hH+?S0({LZ^KpQ7KQRjY=Hq$odl8aK~#@lT> z5t_(*XH*!xwo9{wx59K7<^@U6>}SkXk`~~a_?BnPn&T*X#vm@ZbU&r}E!)kt+b&^t zw#(-sysu(zi}t>JJ?|vsBx6Fsr+#((<3Rn?Cb%rgrTab3PI)*PiU>_k)3j)JE3zRP zJCR*61$!K$OXjScxZ{;w;(m|ykt+lHJu=ma-lft?uTJ@PX?e&cuaQa)JJaPj6rQ&t z+4+DGHggjwo+~^K3M>oM{Bs-LBdcB-``r$!@__hTc2`iK4lFs1* z>fCruS#9NJcucW<6nRL+wll{dQUM}}6^)^XgLQ_PO{%k#Yh-u9bAAJA03$Ig2$rbL z(O9&wPl%K&8BC@mg&!xnM(4xL^-{g+F5Twc?}2HI>B@M~LA6oP2o@u3)$Qh860exu z&AUcF^xh63S$gkt>a&8gf+}7$X#%c=&uAQ6SIb~+P8?mY>hrFV8K<+%4$jIy1Rn=R>)wJWU8vnB0{CP+=+9Dd!TsxA-zb3T>y)}Jr`M|sQ=YOrtVt39&o z!lmX5#vvEXgW@2C)Svr3E{n1ysgV^_>8mgbwBTLlR{RFmt>@I{HvCTg9#772uoe?* z3&!kj_1#711;mM(c?}qlAA(1z zS$D+LNsAtk3&%ZV8*`BJ)$+Schjg zF0e+m%5lC~b@cJQe=c00$r0^DCLdP1vKxrvvEq1(YIF3%sQ=l7SrAnHo!cS*T#0W< zk~32)r=l_gn)I{}{dIdujX>8ESrxk78yw<0g1|6MuMfWzs;l(c8gM@hRiWjx&G{;> zhK=LW<0(Ccvqa!>L;cS?rrDM?;$p&a)+$d+KrZOuJ) zD5JJ}7c%Qr>R51Ez;^EhXIp9PJV2{v$7b2;mN7Zuq(+ZlxAyO6Yy-&M z7875X=@#P9-1#-t5iht*jOu2n9B#FI(DBklmPzTZ`6xqq8cW$SEuNdRKUHPs;AMR)VM z=&e3ILUc{v@UMZ?H6y)m(n*rX+bce%xFy^{7e{4I&*L4Za{^C*d+H-;n|N;^Ud^OC zAv$Aqz)8Oj@s4>A{wchxZ1zlI`on6ylKskj4@;d4DDJ=+B37&poeQC4y7zy-=%s6x zLU{K{a%L*1tw)%Hng+!l24h`TdlyI&C+$%~^vx&Y)yn#xIj)OKYTX=OhPgCwXt&rKvVNS8mq;FrthV|P?|2D#UD71G;E2tgHN z5gaTg2W0FcLyYr)VTF>RCnG`(Cw*p|EVy@~EoJ7@32|<$wb&uX>LOs-%1?2!TJ>DhHUq@MnW5w+Nd zRLIGy4Yr$SF3Xo2CnDI{YGduIR$0m`IwS?@wh$#5$GrX;%%fvy8PSA~>Za6>%lM3C zi}d^V9~>u3hVyj0@yusiZG_fQYzjpZq0pKs?ctL6Go}VolRCnZn0ivC>gC5xcGmmv z_+xi$?p;9)^z83wUsBxK>-)&IOtjmYRP2<5?&g*G_e^htOxso3PI@6m0h^2Al|~*` zr_9(s8*MCyX6xrH4%y>X;598lQ6VbqGn)1Un87W-HjCgpbAPyM?;^s`8f zpPU}oOfn{r{Z=nx0mT|9k^>x0ld{6=$t{Q~f=#95B2?lSq{_h3K&6VL;5I?GA_=ya zhoG46Cel8;-@_=#qpyjZk%kH6juluwqS#9mxj@Cb*+%PqdU#c@exOQX z(A-rcNQnMSRzt5MICcRKibM&H1m6QF;wbp3x%6Y+Ce=!LUHA=YmRF0`9O#5Ouh`?R zZxvHIaVga4#G-ktJ=Z!$`#d^`4M7`+Lr0+Zl0fWodD!^Ta1Oa~B%d9^MAU3e$+dUF zE`{n<+j$jp_o|nMEIdheM$~H_N!w`Ln(M%;C~6DS4bmebHR3`HUg#oiVckmP8!+dB zt^lj?NcYyyz?rjP|BR0q(Ixdi-u}ODSVYnzUiME&l^bWLT(pvv^%Q%OBFCv%n0;Vs zhQJiVl%;ZAukgXF!_zOpZl^%hgyr*^=VsDiswt3Q*zX#zZeH`;T@kIad?{oSl=_wW z=~Wos1e_zi73?N>@;=Y`Y?hJxUi#vLzMNZ;sNxsi5w9 zT^F7dH$f(YndbQ+xgFBV=JPHD7t^0aR8BL%b}iqhgsGuBpllXWpzMh{EylB6@Ln!~ z%t^dz1nO5kP{3wS>sOR3D*3lVj`Av})iRkO#ggujC8G`Yb_hoHrOGp|j08MS?)m$J zf`_)j+jde<*G;_Sh3pE~e0ydCO&xNwA0_q8Xfy;Cc*T{M1KettkJ6(xQTFNDV$u)=wa1 zR)RW@VnK3bQL$)GsQBg;>ATmUl3JPPnkZmB_UIucb1uGGOm~M}g?z1@b8gOB$|M7! z*B15a`N_0|(HKwBher5Umo-aUSXVCm#;ULpj^bwr$G z|FY=s#h$uHKDjh*X!M%=Da(R^{JULc?pjCi{!F0#)>TN1Q%@7DCW!YH#lUT?vTnuZ za3`gLi~{F&;P0Gju=a7jJ@lTaL9al9lslDE54c!x4F9q_^q7ag z)upZWAM-H2JNGqp7yc=I&Wlb_maEw;LLNqQ$UcOf^m$i-dEQCyIVTi3gjVOg=iI&* zm%Z!wc>4SA+A1S%8;M{?OS0YK>oo_mlI*qMNhocu-p-=x=qrJ~+T(?+Nb-ua7$C z$Uzx29JZd_#u-f3H`s@kBs%}TtzzP~^2UzFgzU))lxo{G&1$1wa`zTJv6kOb&K zP77!gAeW^P(m!tb8t7G`A{kolkv$4lsjxg#k-PJ&ll#=vufr zxFkI-KCtC!I5@54wS70gC$%gxihq2mi6nDdWZZXOPj*@@GMN;+g(91<=C9ePgP9pd zT2~|6BiaKY+PK6lFa(=+gVKs zC|!vRYY7U008iak)LUolOkzKg(D%lU|L+MH#5)?zk6Xu&=gTfsLAOxyY zSsPU-)dQnW5jFVFnZjpOuz3O?^O5b`kK&vYY{tnK6n{J#E=czccOXz|ZHSh-onoO* z(N-!plm3iO<(K*&4n~(v9=%U;P1#2Bz-_(9r+E@s-}y& z^Gjp_Nu5A;TfLsy6bp$4o2l4V8Ro}*0J&=!va_T&EGhU%a1j_854={YXcOeCR*SkJ zOrt}Q;)%r$3Pk4qt{2H8Gcb+C?vvY8xNShQ&+b>h^NIx{fxoR>KsLTKmeg`9sPCXy zpcCCj#jXhNq-!Vb_Nn0;Woc4p$26+u%2bon$EC*i2TM-UgZDvTP zctH9|#3P{&&GEo;KQqCnp12DAT!fFUCTlu>N^Cf%;l1DW%#5)><=aPou!kISV>RT0 z)#QIbvF}sls5#>fP3Sr-7LSGEvFvZFY|l){=tgw}6E+qiky<{}OV&Y^;C_W+W`PJD zdrhh~uZ6e;y3X6`YlPoWd?j6ZRhbV}UeNBkuq&!V6YqZ&YBQr%9(UPElJl}oI?gMB z8v5^ASc|H1>cWLPylWZMcOWU48=uOk0WvIVhHvBo9Zn)~T;}Ubzw{*woOVuU?vWa9 zlE;lL^p~x`*FdqSC~|^|ZN$3MvPOA>cwH!3>P+Q*Y6QnU$lWd}qpKiTUR-Ne$f~xFId28&###)=w^zK8o(~d(1x+6z>>e3D@sAe1C3Fb@(w{ z(aPzyc3+D`$z1Q5zp}`iqu;(0K-%4S`C4pMzp95~?@>Sl6uX;OLt`031M^@OmJDn6 zKkDD0xW{i-?hd&+8ISrs%IH?L`Dn?j=HE4RKVNS-HqzSyjs~sgmxnxn2uZ9O_1NzS zlLszjy{bE`Pq6rdssW?B2y7D;DPz1TRcAtC51(o3TCvLs1?2m|?^`{p_Fk~l;9JT=C1-k|*3_bzKJ z=sFSm_jROX0%@?CmctZVO_4)XtWmNrx*Fm|ARE>9OyDh;q)hq)RsCyS^jgv=evfVt zXQ?{GWx;*o9APQFm^`3MeeVgE`Bh0Xr{D3*gYw(GBq?;ogtR$Hla5OhCLD+Q2)Xn* z@pk@Na!6GrFkk7BBur@0;;6@^HS|XUqu?n?|C~Q*n9Htt`onw>$0kl9%6)aE35}65 zTaN>mOeFXH%Yb%gSpUR=THrE79kI}iW9eS-G5Pu$%|IO{U9)IOG%|9-WLHb9%s*AKUg>?VyWfMO5&yt^&$1I;vk;3>1 zsGmU3nMd5#8ZVEo4;eW6A}MO4gv99GO!NcvosyX-Ku zpNJCcBNt)ZKD2l5_5F*j6nUZ0Vu#u?9jvPPfqK>LSGLKPGDbn0U@cic88_FfQ#Dg_NGR3Z;$jZUZ83y?tRjws*qAIb|CZ5woIe-#Q<2kFdeUxnFmq|k`JK!5{-7X>d zZfr|CW;Od26bk{Fr8xWffu;<0<5Y~MMD7FKRbrS4VU?haT=&`<1*ri|jPBiaP-dmY zuNG?3EoDtUyBkcMA}c&yi7W^WbOL3reu~k0Hy8|~YMVSVef)NAN-t7_HinohfRnd05^ZPwiv6Atk%{&l^OXduEO0kaAzboG@WKGg&;N6csBr>B zo-^4ayUYzDJAd&1UYlWoh=)Y8kK|4u?^(^nUWx@4q}^03DuFfhGLb%@DYT6&n*dvg zsEfh9la^0t6KwVA&@_rS1?Ou&(CmV#Y0;jIcr4EswFxk)ajP#b`f;LWd8aZpDuur( zxIu7ta#r}o;4YFRK?b*OrL!l=3DaYKkw-QQcJJaeJ?`sNeO$A@apK^z^nWhBOIEtE zXn{(LLlV?BQ7mZW>j2OKtbw`KtBxrK*WR)TU9We&lBwPhwS2;TO$z^vrq?HrzTxj| zVA@3nT;K%SXAR`!YurG);+=2*^cxmzluml|Q*z9Wv0=2r#yN^@qDUiFTI^J!jYNkf zvLQG?IpEPZ5!RjtF#KgoDyI&3U7cxW-}!*K8n%JCBK?3DFDm!bYtBG6z!@R(AhICq zozjW_Mb@}6vI?w_wS{6gp*RO+7|Oy4F}1#mJ{{br#K2EW#RV)uhc3ZB?~lUusshoN zrNwd+JO~3v=XaF-IKgAwM<;{}Bo=sN3)WvHNp1`es0TJ=QPNZFW{Pa2V)uAAh%={~ ze>Q3NeRcV-a;ZgzpPrq+@rxN>T7;E-TC}J?-F)>SxQyeOb7QlU9*O)I<=i`TJ12lp z`i;K14rYtwXv+PdP-K@hJseZk7cQ)bIv0Ue*O1k-Ky=r$UAb&VNl^O4|9MF{&4K$m zfCm20Z|2w*BXnESZKaj;_`JN}5`LB>Mbz(6FKdBHGf8Z&WT~>>V=)^iS`w5Q)Wz>m z?3E-%r%IBTQ>uD(BAdcrEZPgk+*aiWs*FiSL6c9aI)CpOy0;80He%?ZlXB>l8s*~Yoq2`6h5l=|1u?aD6s)~S6;-6e6YKd$uV(*chs z$L7GGs?ZwE5ud^9dKKOd+NXg|_HOh}V&c49Bz2rHFD_9HIL(Xun#Xhg<4+zhdKuh9 z!N!P07Bi=j8$x$aeNULL$(V#&nK7p9fH&V^ygF4%Jby55cr9LS@O#m1{8&xWAiElB z&M@3LsZ>)Ql`T2R9Q9A*W$`~&w2?}$Ps6rR`fn_Qyu-feF6ELStU8&gIz}&MvpkoG zG8l}x{p&Sl3V(MDl8V6|Jgi0A$2&&n&_$6g+S4(|f}FSVOn$8a5T2pvNt z##rIQK059u&tLGO+t9k^t(Wb9B=pO{@l3P4MVrACNtcA&gY@=Jc^8j4g>v7PchWwNqU z`GBsWmlG)2D%Y8*Sb)GVc#o(~S|z9^y2I*xQHHD!_$hT|VBOaTB!=H0In#kF>N8%k zZ@qNDqbR12_vj0~Bwv&Z+zMIYIPNO#VF}z(AVE?W^X&s3+0(#b8n8VaNmEC;kK>BD zA(H61hoccy|MKp=wYH^C-PV?3r_5~z*2;ryhB1u=K&8-SO5?m-X(Q(6$9q+p4Ym7*Z>{ezA4@}=QeIP86)itTwRjta( zh)bc!j_M?(*7;x<4FsOtYkw^oMgWcfk8k~^AG z#iq~R7hOPah*~yb@CZq5o3-gm40XsA?Gm4(3!^^!QO~csVt{`!D4r?;_2oRQUtKhA zf{I}O^N;(A%>iZKdptk+%_aX?9#j3*`ZpJyBr7L62`ryakJC`|`H0Rr;Dj0~!bktF zKeC`^{kvcJ5;^>(|Z*6Sv)MelCz~U|7@^pNq^* zx>r~+7XrC~VFt?qB!wdX9fV7@E6b+$`P>ug(8Ue(9Q<>y!L~s zH{|&YS|zJO*Z3XKJn|XvFwmQ(cWBJ^NgTD7NeaC?DMOY(VRpUWILy z0jUy1Y`CQ@A&B z5>n%3XJ)nARy%WBtq5wp5AjbMC>AoAv#Hp2=5TPU?2%wKM(n05ulY90A(CHt8w5oQ zgUZZ+hdH``6;nCYOn->X&JRff_$=DI+Fa$i7@WJ&AZ2h2SN!INl2b{@gPXqN#@4_6R)VL1Vj+Aphl<5GDO2%)PWehX5|W!Z zxki#pw}!SWk(?!2Sn6M5@=(FTW|SRN*Gtm^ntZkdB=a-rWd0rb%~{XVI-Z^uE-l=l zWutp=`7jQBFJ{+mu}xOwlKSt)V6l_NXu`)OuU!$n!p6`YG`i1`=Sf$^>wo~xDyQis zj!gmwTsStMlY0H=LUx|5DVPf`+&4U(*a^NI_|_F^Z&)pCo8m-g#gC;1rpEs z;&;%P4Adpq>2IQhT)0$t6F7FzCfg>!<1s6p$$6?>5>4|oKDa3#k*{{Q+Llp#+QI0? z(g7&GhU{&&QfvlAHc+vJU%w*#$2H|*>OLeeHHJ3NZKv~iH{|jBY8j@XJ?6i&Uf9Kd z%>M(nVi?_+NC@tj8JqBI`kVRrVyvKCR+!2%Cx!O$GGsmA5waNjvBJb)3_tP;f}XmTzlroK>VP=4Cvq*` ziobwya1JJY99AFDSo}>_#n+Wt;qjESP&E?Z*eImdr%%g`5w+L{&Z+;qfUzxX^|WAd zW8gr&nIVE@1I4CMB$bNAWam8IP0}5i7JyLjX{O}3_=wtP2;JTs;kwg;FKmy!7b}E@G5)7#G$n$UiDXZteZ11ulQhKX8*Ae`yS@*R83z!enO}<=6Xa;2!i$-UKL)=TdG7Ad3tegnN0T_cBSmqmm$?2`C%@F-iy>>J+(FgI*XSDn;itDbY*C1 zq)~=r}M-9ZG(#o-8hoDI9yj(Ta0xeDA%4)LJjU~}RE3oXO zSg^1esMyMgPDuNR6P@6vMpi*M6owp}W*(}~`mC6PiRw3IW0f);*4A6^-9V=WU=a|U z<{V8`Xg@?G98~QRmw9&c64|r>M<>*wr#rOM=n*h}E8E1m*E8AX|Y&-`=Dh8f4z_206Y-a4$%jZ1kOecJ~NTdO8!f@<&yq{XACH(gEBpeHsdbeZ3V$$E`mbtPb*=AX~u z%4tsA*SosguPNTPuoM^npc@2)sldz=C(`MoOF|5A zfK`EY^^>!wroIMrgVH7y1b`DvR2%TYzB|0o4oEnML zFI#>4=$v`A8Za1F(Wu{E^X-|PFCDE9ii>=5Oo{!x*QkWmnF}=^|G}c=Dsz=0@(H&M zlpBL6(W+$U0L4C{$frG&>e}9 zE8P%^TkKor54C_Bq1At;SvQz?=}#hwp|7hNSA zz5qEUu>D3a)N$Sxwm^hzB4dsthd>1Hx(o3#?9j=vEqjA3qT<*aYb!{O8;gp=Ruf%D zvBebG1)?INgD&yy6k^s#yK?#bco7J32!5RV5&Z~?GwaId)=PWIDR1nj`-{$$N8)#C zYK9EDWyp$n271M89oqab-J>V6M|f&#Cw-l_)mPUk+{2beHo&RV_TY{J!C|>S?6h)k zI1zFv9Mihv!V3O;0{KHQP2PrcA04{3f1dQ?QcW3z6TJ5;#ByPM@ zvSZ@|4pc!RYu+BUD-A*IET$b+@!DX!bSOAOHr9+U?uw%~x(|x}xdOE}sX7E*bpG6= zxrf8o3p0W?&c~>S1De6sELEwtGb|(MNX$dm*KZt@4jkZNT6zZWbM^0t71sm4Y|&a{ zDg+&*mYa#ieJzmWij|JrNU`-4IZ4H0URkaGW53%zht-v^_&_vfyr@883e4O!DGP!u zGofK&R(Mu8ax5o!={CKxesZNSTVl?LE(uBj%Sz=`$YDgY$eyV8nUicWu(_V0?*!>p znBdwf+os(P4BR>#;Hn(>8f57q*Oj^k&+XDSFtOa6wGbWEXdz!2cuG|xEA~ATsyiEj z7C#O}NVsCkTBqTB<}eg{>i>7X(t@cCsln&T`r%@D-FO>ZW`&$QiiJ44EGqV<;*M|o z^c^!ZUVn0gmA)U(&U-!O^*l9RIj+K=eYE*^0hB}A&8sn)a$Q;b+cY&Jhl9g2JED&E zHJR%DhKfZNpj`O=ve~3(0$FQi7+pG^jUP|?so2CX6@RB+xrU#uspAhlZUW2Z-oOWB zvG?HBeh)nBnO@6aU)+E+`CuKG4eSv`mAVUTpLx8yzS#S);35M|w@km%e6&k(#%FMB zy~^BHrl|9~IjfF;IJikZIA#)a9oPr+nF`GszwO>hz?*wqkrjNAY0}>2AN9!Xfv|IFaL2?c#C#H zuI5AEgX#vw7SDo!PWn^e_h|P&3ND38udC#U!wWZ&HF zkez))enW2N`dHv}Tj0n=n)T{7BDwX9n}l1Ascm!Yxgf@kwI(~Y1?nfS=j)CH7lD%~ z9%^TyrgBES9SnXCK%iDnyA+Bgp7g4tniko@_dxY25OGjsp7Ergb7@2XdCr*5xIE$X zYjL&}fw=(0jYW!`TEVyz-$JFn+4Hr>F@4JA!oBKDNsG3K&R`6h4#mFcL>6n&bVcPe z1LCtl^&9u|8)W~qJ|!kR6$C7mr|<{wcDfY#d_qTrmM47{x+rK9uxvZz?OT`OHzWH1ME$bNJQJgMr&*RpO-R4A`=z_+<0WcrKQ`&+GFpRcv6J z=bocYUMc8*grCG5n%yf)_Bu&!g_Wpl6fP<>> z*w#_WMgMi<4U!$U5Cqf0qQD2@4Ko(=m&`mbHWT_jjM}EvtBSpfy~boe^!W^PfzjtT zjlsdp8Kr-*Eik|Z7;fySg#5}Os_tfr-AKVK#3Gec3-cb|9P8LYcZjp5>Q%^eiYjgu z#A9BcxmdK-ceFvgTqchTIzU6^GW4kWuK9EHHCwS^=Y7LCMji6GnH??`PSI<=YpKj1 z*Txv=e5vd9k;D9rT%PDKx4W!~lh}4&sT;g8;g7x+4YlcKw+H8#!OoRlt6mup zyRtSW7py$JKXl#O$O$X%YoYqXABsBPwqPY;Wzt;I>BhU6)mHAV#}wN~k%v@li-`yb zGVnEM={Tc)7zK{R=vEeEZ{ud8?1R?N?XSA`c&Gi$?&o|>lH>+?EX8yGaKv-d4HK_+?T_@K=2i=mUlt<)- zBT@5!EGbHhZ+EB)B$Whx;z74wnqXC-TelI-@8cQIsapo%luaPtI#h4l34|irCSWZ>*QW)@7J=qFb{4oJlAW_}y}1dJ2pJ+RQpupR(LS%= zb1fGf8IyOk9Kybx^d7kAV#GmcKF<)9a|?w!bv)xBW)TNr%>sdz)?;+~6ZG2q^AH8G^=fv6R0T-xOMz=4k(d2>X%LO)%j{Yjg%;Z zZawmNre@kDVE$bbeLZZKY-faV*FE&&FuhehAiEZ5@I6A#5d0nV)A4k?7*Fg(p&W6( zph0>B-8WZOUq1zO(;r){^x0NDa zM77Ba=RJ_UB0KMYo4xZN!ps273dl|+g)dp*`AG|lR6#L7owlEf(bH$>CQS*QIwjTr z7RjD_Gh!#Vkb5KSAP3(qXwvKntO(6icY4)`v(=r@=#Le}J3XF6g9;^`5PgIT;Bim=%UYT{5c?Vv0?a^3jk)T3Y&B>;&ysNDc8YB(U6>s!Ju6SeP6kw7u0Oekn zJBDhY)0-(~5;(VJAc+i6GPNo2z&3F^r%iZkMnxzbY9+B#b;_ahGU$vDJQr=E1HC&( z!^HN)v-`w0Oh%|~pFnlU{xB9l$#p$IV#7~Dju#{fFv=e5={SKNu!7xHgKiiE-lFKB zxA^wU5BtEXwAUv&KnHAI_xKq>E9Hv>_yJz$Ko~tcxnl#?v*S7H2zS!wv#k3KY!vE$ zTu5iAy5&$VD=7D<=Wh_z^Y=kGBq8*o3LV3*=^RV{?wyA&vB-UUx~ z27S!$lJIhH5iA^?QMIB>Wt*^3dLW`p<`C0veJjRxt4-G3dQ3;WXY~xbZg_R-ksSqZ zn;n~<-a4C34ooCX7QVu9im9bQd;%Dxbjp3)yTVP3o^BB)2WT&N=coyIt(Me z7#*}&fys5FYP#R<`GIxBhfNm3b^kJ3SqKv!4a$$|yq3*Mkt3Zj?$;su_X#OdAmLh^ z027(oc6qt)@>%0ksx=yzj}Ex(uFCgWb8wQ`*xdY8%g1EpNV_N(_E$jkaabYNCW?W= z)b&(Mwcn=TBF-v_DFuz2<08&QBcVO|C)K>eKFh=V<)gG&>~697Y*q*u)i;dpBHM;; z@cnPDx5kEz+r5n{hwj0CuWJxb$$-bL&-18cK$auQ4L{69Mp&J4p?>NTCuDVO{>`jz zxAV$up85T3-|i7M8m{}?ZTaE3Wb2afS7cqWQ9#}-IWTk34YaBGC3NA89Q7J*x8k!Y zO`xPPAlvVLP}#1;((-d$JZIeO3OkgKUvcMxUwk85VqR8+_k_PAi7p%+F0@!eGAU*Y zMbfAk9a$^C3qo9VvJ&A|Nj80F)+d3Ngz1x$ct}AHef^xp9BgR3F{?v8baY%Mp6#Tg zSNx0}a>4;trj1JBM$Z2iGemyZyxoVivCB2Nu5Bd1Vlm7o>7ke|igZG;>P&+#_T?uJi};M8r==SORB5bicS!S`*wg^%JPTLzeVRCYcMw_Ksy(N8G)Q{h$TXcYI_1Tfa1;Zl}krH^{Y#z&J8|y1FQ)lOmr{ zF_rVKitG7Z3LQ6l?(L9kK&0CzuA`ISW@DuE^4uzVd-%b?TvbJAm!u%P9kv+AHL^WC zS5*aZ7?keAE8^Wh6C&#fwj?%bR`8BPM&};85T%18 zyta<62u*>)BU1_r{py~`vN`#0?3}#PJ4Xz(OOs2~CKgfrtRv!>JBW8X7EVv6cg(d&g)zbgLYIngCeH{D9Qd9i|%fg7QUc=^P%fR!R0#IRRP zY}9Pv;Jcp4^BQMOyVYE=M}gGR@*kE-*!K4nzD1fzo9fLT?o>4(# zXs05ym)3LR2%3LwmVYaBQDqa9^V|B47MWO02|yIAr~qGcNBE{`&YBPQ6JX^)j{5|R z5Y}srQkbS7uXo*uu--Lovk^*##{4lUwmTI8{RTjEIGyibUkIC_o(g?2Wqm!E!=(tQV7b((6 z#q1HMk$6p(XoCn%;c?%KvPGO4RVTekT`xQDbIdO`#NeABSRt$s)0NOA zv>={|4Y?M$kEHr{$@WIIlcdNy5F~i0It)RFM$V$irfXwE^2JYprgrJXSV7#R3U01@ zrKEvgD>^Z4*~G%hNKU$ooA0;b4SnQoi1749HUOv(s;Yt?h2iU3;O}qY4gg)GPFd|! z9%|q};cb+}N%WBrf8y;4Y|`i>kNE7LTQoNn_;;%K#R}86w$k`_B+E+`*XYxNCQTN1 z73m{4X6&8b${!lX1mN+-=7@l`K4`>jZE9SbHOaBd65!Y%G~OGv)4PG+4gsZ3`aX9@ zM7wOK7?n7hG?|?3AT+m&yD~a~Q|{3Me0|r%+rroQ)C&tIS25XiA-x(HlUSIT=RW5) zJ{Hx{`l*YymcSz>sU^h>#@J+dyk)oXu9v&0~-3e+onBAjyuIH^G`7WFetFxHs0~7Lul zh8l!rbK2#Vv$NH};|WO$Q>_jDOp#wxU|ygER0=WAkVUXcEH)ZTZW^T6zXypjZfrD4 z!Cw>O8Qsz>vW*K7n}r*_Ant=79Fgq_EaIM?S?Y&}$8K=!hJ&@yfByS11A~=KnQ)>h zqv=1*;1NmwA&PXlFnAIyQeUyJF;6JcPsLyjdfMdEL=THkjv7M?^|JLrnc*q&J*vyo zvq`21`Y}}{Ld=CVDm&?uzL$Z|0zdEbfJ9j~*&zm$`L%i_gci~XoDz_#tN~)BCSW5T zbSnh9r8b$i(mz{$6SiP_XxE8n5S|(IVgg+;5S~~}@&Zlt#ktVe2%4JvIQr?ub4*CA zQT1>eqy?NtZ@omP#N=0lG(KQ)2sVr63D9RskuUc^N4wLU8mshw5VjpEE)PT@FGq?T z_JRB^<5#zOWs?*+YzRV&Rc#=D<|GI{AE0)8$_|3__Vs1|zzo3?^FqHv&W^ObbK%7c z60pNEvaJ*YdLdV+7*k&{6tmMvblxD#0MIU z64v)UlXwq<$|7_UgG8(6WKS#i$OUdMo$^jtOQ@bcrQ8^##kyR4+brn~Dh$x3aLbiA zrcT)yni`ECw9?Jc?S;&PP0B<;ju&daVzM?xen<&TWbnk8ev8W4o5O)w1LU%_n6pK) zm-9@KBK`SljDJnwJ6CyBegnVUBa2=uf5=UITU!zdU*dO{KrTONX8PND8i9hXfp7>_gQ3oK z%2J=v`eWk&rRO2W=4ZB{BjxUED^9$S4F~UyCbsyvZ`1nV4Z<(>v{5Q99 zNh!M!xC=*{&RRSIM<}M6A{A6jE58c3yp~8S6&R@eH0sujDv6$hyxLXtF|vz$it}ky zj5#J|a55mU$TsZb=cJvvw{{uV3?sK7w>YwNA`li1H$f>B zvyLKbfz29RgObQDKKLEE;d%&`#WTkAK{oV|#|rYtS-qRZ4IS$}Vs$+$^WwU$=>Lmf z|0}5(7stQ#{-VvpOOj} z-tu3unD)~YbAlpuR7?iF(*xBUnl!7(ZgnevyZc3GTCJly_n{1)TZEOtt|e6@U^R2xg3gYw=rKn8|Es3El)ZGxyye3XYRu%+Hz!cEf#!^-9~ zOWNgl^p#h*KN)v(m513D{YiV_Kgo)T1jv_%?WHzSObSKTQ88GCXZfElV3QBVIFbW8 zrTy|s#rU?|u@ZuR_7QX55mpbc>ppwGO+Ty^o1r25%e#$a?L-0sEyJLZK{1;tvIz<) z%xNN{{6=uPV8HFD-#&>Jqrk>TuuoFyzjN|o?)_O1=DO{@Qknn}k5L(HcC$7*^Um%& zMsqeRgt+czdUe&~AB?4t*@ZXL5I!0P^e+kJE|PBPO~t2@R{jm>AHl|)3#69ro3Bm! z5$4)aAlRgDRZFj6mP8M@-4xyOFZTet9TB#OTo-miyNEM^KRU#ly_nHJ+0JYJ>yJym zZN1L0aZy|tDmIqOfoaKHJx34T?9%8A@flGDeMF2cJ@_uwvz9IsrFuFO3C~On8)!Up zu@evYwY3Lw%BL(MdGQL>!dNXIrOnOPh9R>Kw-SUBr;o_ zFDP(iXxPpOeOjm;Z!*;Ja)!rz`7gBqpQpgHsr*|23b;cUrrxB=I-bwG{?DS6M zCW!I_KNX;Ia~3z=-D#%6PPo|I=sFG{mLsu(3-$1=sS!;4RbD% zC?cSvvN?a=4e<^GAL?;t4F)=G>Md{huy=N1V%Qbq^O7AVIzAOo znm1%Zj{aabIm9j)>$;9Nxom-;lN1B=>c^;OqT_jS^7iQm%p6T|AT~m@>@ezC}}) z`h4d3(FV>ULBCf8wr^pc=Hhxumph2U`(2cj z`!;DRxQEp`H6A+}d70a&S@QncuciL&)%jO{cwTc>SP$Ajmz^NdYdL45#);)H4u2Q* z?$JAwmgs1 zM_TzCIc2Jz$jtDSQ16lpwRjF#6Gx4o=Ng!?!|7)l?VA5xx|+A77og{9?fj(3diu23 zNHl;2!=U{RmBfQ?#nWpQePP!@kisGJZr=>CMTdQ(j{Pj{`b+0TX6us0PrXhO*jX1B zc4`VNtc#9f(kZeLm|DPkY!AO~WH4=(0F^b?Ubf2(vL#Y&aZs_UUET>)rWvA2vyB-i zkgx>^T8*(!|BFX>^P4~1^_m$VOO}Tm1+E|BvfWI;#Wid}SxYf%D6*1@=>jd9HQc2^ zx#87*;I!h(;Q)jG;(KR(hCOp^ea(2hXSENSAoqX!&RE)(T-b57(Y92@KN_qDamppp zM|?7sHR68pAM&U^;FSHtS$bS~DiI9!fXL)hj0bd{e#b{&YpFMddpN2n_?h1tD|BtGXpymC}6CURS1Xrw-Z9EWJ{+PL@xEd%&n!jaqIaDD*f{V z@stG49ciX#3Do%-_zRqDXY2=qwU_yH;+*lywncfZ())zk43pzA|FfPHyKrQ!-U1Re z6jMo&LsZOZfCS3lYImrLfN*Fd7sU=5e3ogn7yNbz8u-wP5nbz}pPJ>HEo$Xg216N1 zeAHc^n&3s_1-34=)vL_cG9(VFPaU{7zvVT`S(` zWAN!x)I#Do57Y|NqnCu=&@9-jyei%my>iZNpPp&mybjq)d5Qnpfc{wv8i8v9&$&+O zL%ZpRAV^gtKd#Wzg+tZ~>sMH&QQ2 z2I1btK?81Eyq|C%5^X)d4a&iVe18Pz{EY-7Ad_8OtuLih;7Jn5V z<8FDUJWcY6pqG;zP{r&KmxtB|BN5)of3E!dP4C>9x9Wox6Pq+x7>%=cf7U8V5vOqS zBKKVGl^K^fr-azje8w>%Z^Lez1E$CH6yqfbUFBTpM^?;l7?AeW)K_HF1q^bMXRq=Vv(6z`Siq|{XN;wSw4i5Rf*H*GkV+M{Pd}Xg{Aj}-f($BLbkIa)~BIsB+k zyUE*D$QxR#Hhs5_EPKhqgxMBmD3xN?QzQ{`3}3%0{IWu1xGY5K+evS0hj?C95%*p~cTyjnV<*RTN_1 z^J(SHIpvA89P|}I$c~lN`)+10qu`5Q|Cz*&e?^+CW!B;dj}b3_3O%|6lcbIg{=Iee?Bno)~{2tj4?!ziea-Q2HQM{!*?=5Hgsj776&g8Sv?Bt7Deum)+5EFGkq zw5Ughtt7hW;*fknqf+~s=lMCvd;mLMv9^hm8jbN3biA~e=boRF!aEbJZGz;{5!oqa zrn{*UH+gn8-3*xk2kw9!i<^-#Fq)(6aHDP*Ra9Li$`0n2RKQmd^W4xV`oHeNQ3)HADt5OaGg^w5C=Yg&%X<7?l@ZW=a1&N zPi^1v5v;oFrbrg^Hzj(T{hIGpZ@Ea;Pb5Va?!z{U$)v~@DyD$jP1o?RaQ3KmUODQ# zz_X#C-DD(-)hUs)&&0r25Lv>h2(6S<@w$OMNvFIuqur?-1Ev{eixTs??>IQv80i&3 zNoR4h;@4)B1pnss*U1eR_I{UG@PhYJObFTMul>P`-~E zfKQ`7gKiDdvmiZUBI}NwwSH2k@JjH8H*{((iZfNIFA4KSxpbnjmIcLZjpT_y5Uwic zEv$FnCeBtLFptzVVvfQS0ZR#*@+0M*j!Ap9aVk^+fkX zc1?l!LN2}3Cst7T`v3HEAHN0SPZ8_qXuGB)f-+s@>qq8w1`S4J1e&zK!HLQUTnamfV=x<5nE5Zi;D_zt(|GStvuwZyoF!BIjAVuPklZT5bvrT&9_2e`eaN*VxV5S;Y9iBijt0}UAin&L($*aMu zx-g|ijO~Zx=2cnF%ZsPj^8E37$7=O_ZTYf|1J)!RHn=dl>2{f3l_hGIWk>WxmWr+4 zH)^i=7jf%lH#FygI}~460iom$a*?-3yzcD==wL@pG|b$Wik&uRFP@d>M=zW;ORQ$a zbv<9s>pxAg7Rq^c!|cLU1vbQUOq7%PqAqvDXcAXH6&nNw+%o9%p*^zepaZ`Hcwche z??X}Tfv5s5csI&z;@#Y;pg}iOl23|!M4#gwk@d**pgx(vDd5C2nc>`&H zQXlnpWD&wxk#6-II1Ko%6?TjnbOneUj}?R3I1y zf`*WUKP_h6k#B4xDE%jcXRNn4Hg*z*p^l29kq6bS^lGp1jf)T64}0$M(7EIJ)Al_c z!G^o8&!<{a`12S0tS)RmZ1}9#_zyrAy-2jh_l)QQaQ6&&!e(*cX-=j#bepYJ&uCs6^HUc`JO4 z$57d9x%ZCnickm3VvRgoqeaBXqdD%VZA6S@`@Waz(erQSoz1N|Yu_NZTzF@*+#=Nf z1;zAIq=$-GDPJjX4eO-KOm!yL#04~NS$oNV?7lFOH|SMw9CtV_0X%)C((8T?n=CO$p+o}NIVk{g$(gg$Q9PghwV00wHQ!NQC>^5 znXq5hDO(`|ZyILM4N1njr3Zr2g6`2<6#1gG$yoDWq1wurjzE!rKX<8jJP4ISl??o< zM$w4)4~l<^t^Li6HjUF7mPm5o8Cr${yxvw~da)^i&pK!t|iO;D4jo4yDc0O$qy z^8;D4=D#1vP(>D<=Pin7PKl%NoC0VXz>^%fxD7Wxqcv+!7rf!4IqvnIpT=$#n!&O% z{o22h_*VujJ1oGmg<{euvVn>@@82$4quvwPq{&gQVOplQL+zb@dM(84E6&kGZ;W!99ccd*-{_aQCAv!iB|1MlAWh0^r6 zIdI+Zfh{pDEUm>l6eB-JaZn=gpCSVUayft#ycgiH*?sePR$8)zhV=|O?Y*J!izubp z=g^$^*&&kU!kZ&t8ygnfFQS-23P=ZIAZyJ_lN50E#;x>LPzY(456JerAB0{fQ&CY# zWHkrR>XxPfVOWm33AQHj;SJtB+)C1>xIb%=<|JH|0hRk{68yJ-t5bgT(MJwnT3Djs zsl#dY%F%FOwV|HS^L__C7C*#=&4~>^WJ6TDa<@l^+GtLa1Nz8j_p{!M!UtqYQJbfA z(vQZYfqnR#v%%5u>zw^sQg*&ZZnh>Hl7cUhRCYTU7e+>@h3(0u7^v>gq+-fFE|EUZ z40>5W25pD{rgkMTgxv$mp=6QKZB2??7JzJspy5bsF>tb1w$-S0f1UJu9SzXgbDM$Rr{l9W}x)`RJx38Wd{@&c37({fKp5`yC_mX#iYN}EzJTq zq#Fp>TlpDaGu8$y;Uq+-c_KVYg{}O4Pb5~&Q9q*bXfX9Su27i=4G0(!;-nlDi_422GQx#p-+Z>?m;W|!l_`(GQnXNSQLK?PXg zU#H7`u_HUh18D*sS*m330;+gx-6%{GV<+4Yf4?0%`Ms> z$!%07@>1j|-o1Hp{0yD)tZ(sb18oQz_5^g$75sPx18Pl3SqWQa5K|2St`Kax06Hh7 zb{jWE-sg#|Un|13hk}=dV=~Ct=y93p0g}rvPt?=L+zpZvaDv8co}7nm`*A$dne4C~ zak%1bYb_BQA{*{1K5nrBAEFp2v)fC>7@(4F zIj=&fi*8cpfj7K#V!!7VVVP&Scp1>Jo~8Rp5hw<9&Mc(YktWR%k~H-g>7F~_mI=?n zVY+?ti{$Bg;K%m|X4voO8^tIAp=6;-ga*37V9Ss^c8R zQ}BYt)-g}Sw>764tlLA__`|OIU;xAW@JQd6imxf*JLrK~NZ^q{7f3cK7e(iCGd=eB z-2-;+eG>ev?scgB=jI5ud0}HO9%<67@y>v5!W&R+wRGaLiEZAwblvOy&e%Aargf}v zaSS+~9p!tK{a(>#YZJJ5^$xOsBDr93pX(^*C`D>uvpVH?6l_tsMa%PWBH}7w5jwiIRJBxd;AK4G9E&}PAgt| z-jv#hmgkM&oX>vm_QSt@*){X@(~+}hwO6_C39>Uno8^@pFs2o1w0GK`IUeoUW4X?@ zMHs0jx3m6Ne%8^ISHw!#h{c%rSR0jxlqWpb2X%&NGkJ&jJ@PMP`686-)ne!cg&47v z7un=c2)q>{FlUAEp{YQ8(5+K`NrE-DhlCJGMD<=1B73;^JhTZsBqqxUiW8inTRjRW zJU9J=0&1)%wFhv|bC+y>HWswOwM+bq=UKNBy;MGkjlL&bMhrAICRfl~1InGQ^KD}q zUOeqKE_Cu?+jfRCw`%@8^Zw@BM}M%J9AXzuabeHnvW4eyl43rg$T1L*2O+4ENRy3# z0(@u$(9@Zs61qHe&<$GwQsj5vN)RQH4hY^3x~+_QAnpP7N!0qz7h%~;rU(o7G3{~N zyIqdL(DytB-Si+gjy(aDz<#+L^!GZU;u1|u8Yf9q5`khm_3j3c1dg3F=$0=kk!_LS zD5EwrF9w=T3u4y?qs8MNzdC*C>qFb+-b33zaA_ju4aO!-L1NKH$^YUhtQP`~5 zF8g9eFLc}+1w29J?kp5+H)*PQS=>&#+yjqpp4=wWLv}-_tWfO{joOOXp1QGtX2`_Y z9)Gl-?38u>B`me#274E$V+M=L@byc2oZG_;V#5iAC;BzSnxt)I@GTdmFv$5ys0BsU^2 zqImYoNegen+YTC+%S>J?2wf#}k*m2;*{EDUsl_8jzTCsi%xduMBX^|w=^gYd(w3Gp zJ56gX81Ra0zjj)zZ5b2;ZuTZBW)Cph0bX`ogC2bc_$3FJSp;nknh6wX6#SfJzZ;{^OPl#rBDg3 zT^ZS>Hb|gUEGcq_>cC8tQOOr!bw>h6KmCllha1m01$y?wYsjM+>8IoIw!~xJht}j# zHaMEH{5Jqw*ERo=$TH6zbdygvedNtS&hpoDGI=_(E4q1XsXD+d!TCHNRB;=6;xmQ`utzK z6Kq~!iaU$qh<+l7nhq}p*+((Zi&13KyiMSADe}4H9`*d?yiTb9siO0N*mS926)B+nrAv#HvL#IwD1x=1^;&!kw8T4bQ!PluR&pjqLon?hptYzzzb?ceIjuUY zx**AgToJ}$F?x$q<~7qUK_)r{F4`c9hmNCakiOUgvPO+clc~a%f=j}R&@X&i`BiYW zBfSy_A|1i+Ftpo7x^j^!!MdQuMvVw&>#&=v9ay~1DX;|)-M*pSq-iy9E*onZTJkFc|F-RE+7Phsz5^NQ3w^Wujj*@Zi#i!GMQY>I)#A{`ZTlY@FwvA`k* zb|f+`$l%*9ze#6LTi6knz=;EK@FcH_0H|8g(>nyM{H4J;bS|w6C=`~0baTFklZhAE z;n!j{bQo4g8i*BsCotFl-?tQIm^A);??ST4g<-PK0w#GB1Ga7(6|*;LMev6^y?Y`v z1+5;)1^kQCJr&B)S zCaaCzL}sf5M0KD>+Xy<1*e91FKcqw>q}`z^kPK6EM3Eb8hl~>}iCzQrcdI64(4UL1 zgO8sgN1{ai^bI1=wF$>~i*|=#E0Z8XMS7%KZ1T>AyhKSvEkQS_D!5~22c4@%ot!Jc zzqmutBE&X6tw}H^mA6EiPv7HV50&jjeT=BIFlnQDV+}|1`>*QXJuu0Pvzxza`IxL^ zhcnk*lnIc{4a3;yj*R|f9~B-G2+2Vt;IPajk*ksA2va{ZiG!9MwB zsq^g9vlGDz8_!-)ibw@en^@_i$iYA$O1O*Twv;G($vyxeAWh(3cDd!m_c8G2tcA}xpxIWk+(>Y z8*r9hHcO|(k(`AJRtOmB@plpi*sWPB=LPMu4&=TfDMK6a=r6_5m-_Srw@@`FPSEdJ zKowApoD%XFsHJc(OT{I3qZR231kl_{H4?m{Oi%##~1)Lyp?6x;9Y*jKXtjaoySxb>MRLq%o zj6~WAp%0+)`)0l&j;Xd2LUxSg*i#n1@$*7!(hZk27&h2IOf3H; zktn%~%R~0`Wud6=i`Dk6{2i*Qpq@x=vZ!oMNo4!9Qz4zSlOwd|8M1TAEgTKUPVUH~ zv9cNyzI#jm`pdqeV4pISPXkkrPFW@Eon@d;Dccml+JjZNNXUac7`Ee8Bj3g`9c71& zF?|)Qi`ijgM%9@~jCn=zeAIiHY;fUi%pQw1LQgT-6v+bahJbE*7x(_GCQh&LvG^D{ zrG(fg|MFa&@)L63H`%{QbCs_naU=^en8>zoQX}g3yr*i5(kV}KuQ*xUo*h!g&zP`6 z3Ps&rcQ_zGQ{{7HffNxZ=4#ZPKh0EE3et&J_2F zYiCy(m3=Qvd8+I?;Fhbt;n~AIG;2KOpT(>{UBSjWrFLA;ZV?mwdiMXbF28*?_~yb{ zz%3TISxYf%D6*1@$rnHe+GhUwIT<2oQ>~zzlr0|PfePyhvI`~F7mdSv+b9W1I~p~X zkctaq#YRX4WkHG*r^Q;VEea^L&hmGYWtdBlxx zZutJs%m(EBtL{DI%0$v{@#K9@F}EmkgNkX8no90_LTX{6Q4 zQOAbQ7_0Tlb$!F187fu;nBi71`1gBB=EM=!Ti5;lNrlC;v72Im1tFh`F?IShOR$Go zi+w9O>U#PNeMgGj%TK}{gu(6zSF8ppvUYUMS}o6|D^zgUAc+mh7grK2#ML5rV+F_p zWjf#l`{XEm567k>Ufag^j`?Rr)|mn}tCtH0-fZL*TE1Bi(gf)N*Z3W>rBk{{8CBww zIOmRHao{z6lcwjrtKv9bExnVplZ#{@#9NPGbT_P)Ua(W#>5~e3suF)k5wedswPV82 zh$A}Y;JY9E@e}L9NgEZ1hxiZ4tr;bepao@Qtl90YQ=-d|0c|<=Wfvq6<`!V;Z`(MM z0R*RPcY+w_fXxw4EhyY7GdrpoQ~&MvBxNFj;^kqvlN^cxHlz$H=I(q{6U^h>mn@OS zd37o3fEfXK60V7}xTa%Qg4acwjzLi>2tBk2v;52G`;t0UhSPc>hW;{EH~8tTc6#iS z)vDsUPujo6`ejL)f(Z`C_D6oKW*uhbzSLeQQk42N@PRH{pk2c)6b_N7{}r(EVw|8E8!vP`{T|ym z`Ng}rzkJ)gnEdqC*>rN?m5I_fS)k=O#ne*d2o;kOZa@M+U@^V{bj6vx9xnD4Z~xAn zuqWVv+zD%-kIO3kOZ|G}+A99`?=(nJnioV6)xaUxBR9|{u|$)kycTIpk@f?JOOfyL zJ;%Ks4HcN}a)S&hMd6YrW}BjVZfUI zW`;xb7aTE^h$;00igs>IFsOhZ2sfXN?a6DrZ%W(cMT&Y^9;Xj9$`Cvykt=wWq|=dZ zm}7xsKQLZ&=J)q4So^Zq>!0zd=Vypc2vVo_0IzkMqA{XLb5glDWH)SIjO?TEIWiPv z&|4IE&NwYk?5F%iljWGtZ7)e9gl?#FkAMEdmz{s-XmA0yo9>hXk5|ORNhM`}dtz0VP3;5YHuJKd64PX3Ew zG~}9EUgzw0)3|Yczc&Np-y420oqYbv0ArO!>&9b>`GO+7R7^H#AEbbb+2XN{YXIu= z47%JG*(37>gKh@kjl>#>Yj6g#q5{`a9|-^C3vO^@qxJNs;^Q*ybr5FA0Jf%wK?aGQ zuJF^|3UA<)M6ULVpA6iGi>F@#r6J^Lf;v8v7-B0wN8MpmvBHdV9&D~Kjiqf%VW5`=4+g?Xk~}e1WTZ*@B&bD{FE9zBq4dn48@R;ec+}oGEkeAmQ}{sK0W?m) zBt>g0xJUq1&5395yO8?Qq*kNt_v-g*bzP-~`4y@EZr!r+QiWn{G-fn%)=K-mDuiux z5(R#ayp|vqaNCT#9Na*Uul=##{ETl@#`Y?9`@4UBBgkx+YUi&jBiZZ>lMCkpYb;i* zQi>^}NFf>~m5h4~^qo$6H&8XjX3B^zlCVgab3h0W&Ak;WhGv2!PPZnoyGb=hFQ z4N_ecxZLfhO_Ll|>*-r>YB30+hyC)B5bX_l4KGie$lJt}k7H{W%gG+?+_ApSw)Gfc z?>ym`_nW)8W{ct$B6o-65Ku9d zMAndM9%}p;A{rbSPqS<0>_(AYqdD%RhktozHOGvL-+$6uM;5zqb0AO{58LgmrF z#8WY+)yTnCIa`a&7|rq;RXI<46Kd^mM&ysLMmtQ4tIjhkYw=p;TQ%Q$*;_`pLnyv* zUYmRo^#JJFbUq#Gk}3OUJdW<10wO%JSV4twRrH$48)hVho$;v7W;`%G zw{~M>xYOXcN8aX0tT00TcJeo8T1#EJtg}&NVL6H^2KtN(sF+?dAj2JF!8}OxVfWcF zzXaZDFKyxE9&YcOPo#rxySUdmizpP=HCmWP&ZcRDVM{mza?BXspO&LWhMx7ItzmJ3 zJTdG=v7T*x=w+EhGd4J5qXNxv0Dbw|=KPQ916H7!P|`j40pGmbymN8GVY14Fmz(Vt zph=^c4HQYHVqyizDQuvtBnID#&|R|AGhyDb*P}{yFtA--?i(lA5>%&hI0BwCJuJcT zoRJ*+5i4-K)^q&8=?~4|$b48kk=%J@eAG1-;kW^cc|?&uAR3+0q`3qEwF>&4eA6`G z+z~bjmrfk%=(358n){M^*&cF&Nss=`sUH{5Z~uqbpRJs~=X-nSt)~j!y*qCkb$ed< zS0B8q`{wDNY@-%^ZR6LAe|&e|Z*wg6-7y_jdkh`+L{tmw(Xn!SZ+R{Uq)y z)eC;}A1fxF2N?kP-T3Xm^>W51kDJOZqYlsfT->Nx{oO}D&iJR;58B>2uh}5l#jOQf z+dtW5IaHSQrK?ZpS?3d9sxYmMvdVUd@xq9a&KTTS-~wIwr7`rHBmQ&eC_{U7W^E}+KQR@j39lmF_Z$+j`9 zZ48VMV|(r2zrXszJ7)i7)<@@BNx2I#x zAN&E`iVv4k|7m*H@!o$dXpmlu=%g2PXx3A;KOcUI>7D76T=$KVPP#){C|uB}Il@~K zzMI=FKc(y?C=Qgs*%O#Z>STT7y6O)d?=@-m$ch!mNVOlHSVyOaZ>RUp8UDUPmB!%3 z=QVaA(|$aCxhop;3mXu!ILr!9RM*VoKd@eST-MImSY-M=+rkFJIw3!sDY`ea@UJ6? zYL6#kF;C<6h7}-Qqc_k@YnGkLsxjoJL^buxk++%T%+% z)Uc&zEZ=F*wT53#yclLiNmf91Dk+>uPFmdF3W@=ZrTtXQZQe2Oi>iJ)PI3&CAop>b zH0kd>e!tw~c1Sam$L)yNqPiPV?tdk?F`|{m?uf$4hKNPsy{G%i*9vj=loiJQy*<-I9qc2>Us7 zN~}ac_NIH|K_)pmM86o3V?Dz*qC;5o48C|iABSUG!Pw{*4FQFS88p>;|mzFgs<-9wVOCEjV1p_Vm+ zKB+vejt{S*7dA*MLQ%&weez+Sg78ZJ;wjo1RfF`tu#?6^9X@4qu1|+?bJVS%f2vdW zL(h>VT^Q?0wGV4nE2MpIS^qA>FHm(pKKUQn0wnpF<(w)|bGGnYFjcnz6}c z7cv`EQf-R+vyM+*7ld&)ticA<2WVX$1eV%DNws)t;TWES8EbRaZ-@Nk=hk7| zR|FPQ2N)^2L2Wt{Vd&Hu`>2 z!qL6>M9@ut2JP-p*&^(egA$JdTHm>C@)L9=Z+uZ;*ZiBTfO99@YINJHOwHpc`O6;Q zNAhzcv$biVFt+qyRa~qlty7M#vA}24PJ?0fI6MEXZG+(+8MxKDwcTZ%N*m4UIqF9= zc6p`9FT%2s$H|1QUgJ@MB85b>OX&P`*+Ook`=!{ z^njdl;aZg27Q2lWin&aYi%43IBweli^^-dIJ(0k^1^WqX^z48=Y$L}coUl@wz!`L_ zrBUkC^jW^g3JUir2D&HmP8f#IP*b7YqmS%{{2yKg{o$edZ;n1bzbF zE{nwiUr%=^V|+CXMmArv!1m*Q3jn z$?8rOvJdZ3Wr)^=UJcM5fG!Y+Hw@dn=}$kf8x~J58JAbAU@>9uj2~~GWroEs&duIQ zie8y8?HLOYRZ|Q|EgYm`cKD=m>Sf8?wE^|+&}2$F)Y^LY3SoTI0s442s48U81xBG4 zof5P~xJc-9L*5>pH?28Pr`$o`0*w~~qyYw1<-RHAKB5w^K(P8Hx_8qJlaGkYJ&bxV z*X8+wer~-q2P(VE=yLAZqjh#8j;-+-BQDQ=C*zyHF=Ob@Ygf!9H%HRoaN%^&GK+wI zAH_VNNH-OO;n%}lltsYqAuUoEBnKRtdH^(MJ7i~kw1;7K@l&kC)ne9Ybub#D)9$C; zv(*)$i=x|=ri&kvHC$~M-6CvKo>Fe7cTLU@?3OOrsH#x4@-t{dgppkmc~bg;N#Vkj zLAPX)L6+*7t*&?Pq)}H$+o(jYl&j(uf`=fViHESi?U?(j*|4}lXA=$lFUs`r>!C=j z0*tD_z(C&+$7{5^xGlnZX)UdrdO8r(4=Hl2xv2+9x2s|^zQ{)@os8hv; z90AArVBlTxMv@%RNq-uc6g9@op&d{CXf2K%Lyzec+khCswVrUU?Ayz%HG*6gBsTPb z?tT5LcsZ|%ZdVl1tr6v+b)?i!ui7p-7DM5G#fPmjXGw!M1V zE6(xKZu2VQv?)N?Z0?{NKH*4wfHAfn?s1^ba>_)QZ_1eF!=4*E%5zxFhwJ(d3zy}& zTeGvftZlKu#BL7$LL0parkCelh`>CMGhsP0YU~8-$QL{DC@ZkO*7Qd~g0)lfQU!i( zI4TC2VM?3yWZ)S!?w*nZuu$Q;upu;;Tcj{~pBp)4&_@42)+0a4sZ-U@V*&lC>@j36QBA z7NX0b7$_~+1Y8>-DEyQr`9uID1)?qmCX-u1Znc5QrB6TsT7_zbpoEjB&~eU4(LU^V zzKXJOlI=vua0B6NaF~8CHiaeBd;j>u9pOU=`J;9V*)@@zvY3cUiaA7qz#{}NPC~FD z)7>OTs8hChK(=nLPrp~9ky5QAtU-!e7Eph>aN!n3rU0S{aX|)OlTJpy=mg2~Zs{#AeM$HT)drS!JmM)sAQ^RH0yQ5QX0dCrxoB|F+9)KVJo_NLBdXYylQNt(g zxv$4EqpTif*Y#JkKL4uTn$L#~1YI~?ZG-)%C$enL4f*EKqhT$f&5{*?MQ=c?wpESM zJm?h(SPOx(d{MsO|JZ*fOmOEi3tsjGNKbzfcnNxc?h04S<0tP{4Z0zVI?ixr_@J9E zpm=(6K=(TsY>g9i^LP4<*SwFyu8qZ9*?lcLySCy`Y0rzdHm+;_h>fjHqq38}5FQJ* z0AuSgA#pMI{fU>1d=njp1}Y7ZG8`tI!A%_x9lS_JKe z!01l24V-_jn)ns#n8+)##@L9U98)Dgy9cuLYfFUt)XkC$Q%XiF6E?cp7`n6UjLX&-Ez90M&RU6;ti^#0ME-du027l~KE1zDSTK$1JK& znMkxKJ+LNv$rP*;=_88-i9BQP$_CCN0WfHhe9=L*QDdNrY*D3pqB<=EDx|${c26@k zo)5Z}iEqoXXl=9mS?@tNBn8|VQ4PXzhh|~&47+#6@<`h^0nc?)ZGSk1`Lg}Ej-#T6A=4#M{VR*fvPYb zJEATT9#DS*l)Lc(-BHKYI`vrK|0P15@f>_P_!wTKciy>fy*RyW0ejOXJp<%l?nq;) zk50eWq(oQNA@j!W*4j7A?8bD&*{rPHgzHysbXfOyxGW}Y^l})W&nsVmYiBFGLQ3^q zXcaY;XQ%uX%Cqf5hHD1}kK|w)<-pmlvAdmJ+|yGScyql{PK4C~bO3G50BQ zkBa%YOR(pM;NN*R+8o3`J_{bC&FMiIVPDWtQp4;VgnxwsG#Gl34Ayd z*8PO6lp4#cQsgT*CG;0E=ubqwUFH;7KM9nt)VWXxWo%r`j=UeKQ`V?r1)aje$vXA5 zNEk#1$#(s86s~GwdStsobjtOUI)xp^8g;l1%_0jl+V;L2{R#eFHhq_Mi_71$=B%`N z@=N8y+0y?os>MqZIH#xRykMuDMb`09dh9y6B&p`@lZ^A~W0KfBVmlg6@M{q%BPxyN8wi4b2IE!b^uKtv1Y>sC{?M9E^1p1lR!6(d# zl|<*lsYytH4NFZHQp^sDY=_h&)g~{ThuR|hm4m{%sKUwZ%tlp_;*3uXuOM>3tud@r zmKdBMx*W1mQtpuv^gtcUTjSHtG->Y1i$(jDccFB5v%4WG0mLNE(s|D6jvqS&9Aaml z8N~KdHNxsc75L00j2SSVk9seY4X;dJ!5#~E=_v-<@$fywsA!MXmO{@K&PsCz(1iGh2D-v@02=ls@zplP#Y zmsh%%wjwk+U@fdc+D2(|z-LpI`rP)>R!+YoZ4oy)Raf3sXP%WsnUHz6;EOlRIB^rH z_mLcS7R7}le}^qlwTEJW&1Dx5dWO{V&r5+AN_@*FGvtbTgQ&@8iL{(|bJo7V#UYuz z9+^HMGbA&(Nz+Gaf|u}6uwzTeZq?0Mk4SmwI?|-c<0QIo5Z#ov%eVOM_x+sTue{8~ zQDHQPSLKZfWy8G>$C(!6NWl^Fxj!)D=ES_v?~t=DY+ry$Y?$lPN-{)a!Cjk;e*JJofm3ooQ3Vf@hlIku6H&_i~a0 zjGOLudB0~~;BB95(FNdb#-*uKnlD=8-6`ye>>_!A_f&0BDIOK{b+z`wluG~hS^Ky; zvNoVuo+4io-6@5BBizGp_d+h;lftvUXMN$DnkCodCxv<5eG(7?y}|A0K7K1-Q~*(e zYR;PHcfL6cHp3xg_iNi&V`ZFPt2?1p&lz%e9@M3feJ%{L^A;ehrI;fWsitE3JdZ*I z5&RR}$+QSTYh;KjDxMj1`&2w2yD+(xe}gF$=Frf+5N!+rnPG?0B&KZ>s{dXO8<71e zLo@(-<$azFq0q;~FY(KuZ@-S7&Y&B9j5{nqqI}uwMN{=bpLqg;+U122D`BHHS|`Zq zNU{x<=YGNtN$>g;-+S5nw9(UTRGt+=p%s`Ubajm_20K)RKs`O)5}zz!@!V;2^oQ(V z(YE~2Es@#4Wbsq4lZ1(&hd6w(&{0e}MK)3~#fm=9CQU1FPNIfGr*KeQ&8bo!1w&sO z`9Rz#t&J`q%V9OR#!m>371WLw(-(_*Fq>~j-|>;hV>J)1>qO2>ir?|FGf`DZiWHBC zE?^hwh|sAs=u5&n*)jJ%aAb7>PkibHH(3+|eUs@B9R?bB9-zon! z;q?%+!*cU)Zx@q17Y?V?TFiPG#q6a>2^F)`yF`=9ZI=!GY}BCa@)^)3iPvO_j`{uJ z+GK-ImZ-tlN>D+UiW_G)X=+8?^Z?n%y{O6(J>fNK?*8M}`N`isuNgYFj#QDMqkl;G z$<29o>MPncMWaQ-Fk3PFrCr}?&q=ISON#qx=sjyq=$ESd9+=yQS&{V=lSq+xDh7ip z*jEbyIs<*%8zZ|~jHqBAilf(#7tM;@NO{pf4*uLWQZ~PS`&ZVDOKesh*WLKrYH3>G zUItY$OMP~$)`xBmNDs)S&nfWLX;7^>8eA`J(tv`9Y``rS{%z7UaxTw(EKY_-r*XiN z-S8Q1PR4<);SbYRfAwXLq3G$2LEWT^u7@JadVZ@{xks$vRzy8&%IMVX@`LUh-l}&8 z7tFy=-gAayH`~t{$iW}kHWnikv`l!TWAFC`W|!o{l1cwX*1R(5=X{F=XEViYLTd+X zzS!RXFsfdv3s^R58>yu~jp~<|3Qv(zVY_=Bgn(|pUhaK!)?@E9`-pTMF@Nm9G#(T$ zroXCi|G+lu_&l5S#f7)RHqs0`px>@+PBC*4QVDHRo$^Uo5l5Rgd4=Gf=Wbx=-J=51 zN<&0E@S^2UFOgKy=jV)z&0z}=Tw)vn%omOEos7iqzwA9GC?{qrYJ=b~mfM;XBfuj) z2qer{=d{|(X>s2XZZNyK9$^3{9b&bzymtG~t6#I`8g$w2BO4@$nSx5GPEz@1t?22i zMooqAA-SRX&82sCe&hI8_s;tbZh-Nu{6PG-pUw}gf_iCeJ*jn+n()Xo$_^+NUZ+!y z11q3V?{DpYUuAYdrvHET-UP0x^xPl!i06>Jn4Ms9ByI!{WJVUlh}z&RUG7ZVX?xq- z-Y)-cru}t#+y17#sr^hhR9r_{6jVT2WRpb{a2eFFxT1~%4vZ+S2%>{H2*^-W_&rY& z91@AAQffX&EF&-q#n zLf1~Z`^}`q6<@9X(XPmKlLi#2>ci95Pp)EWWu5Nl#itd!6;%$d#u&@^SdNxshl|_z zxNOmMaK%G+YXV#@&WQ_$Ty4*Oecj*ZgjXdv!}R9sU#JPmmh##V3|sl1C6=e=_13Ww)RI;=2lWV1Jhv-wl@-G0?< zYaa13eoV?;I0Adw!U&$D*fSKVrgVnL+r-2gQ6_1fVeqJvUwm=~_|65Wb_&Cb(fD41 zv(R=%kE*9AVv;Wc7w0^k&QDafDo>I&KoIz|n3V94hN5d@kt3cQbTg8ofoNi?8aIT{ z5km3`ykvgJh&$weZ*m2sHIro?L1BC(qGEs)e5RW z*s40BUYq{XE$xF}RleT7IR6JFUqAlUPv35j{B!@?kH6gYc2{Jhwms;U_MdM(tMFW4|tX@S9fvZjIVHuf4w8mNik#m$_)kgn_X{bv2?x2=UdxmXI< z{mY0A+0Qm8Z9b%lVZe9AN@t^MyKa8Fk)*@KoHwo*w-?_yw~PX&qrT-YnG>?VZuUy< z%&Yp8oEnKT$c2Z$&0?asNUO}8zNE`>Mw)3X0u?o;=oI$!mq1F z-bLf3_dSnm(nEKJ#qoDA*`RuZ;gdLi3tc}SYM`t5>*nk8MIabk#qXetBfACbfO4^u zZuZ2LG;TV=R&~8syE1pqK~0?;8sj?|B1&QH4V*(*Jdey)$eID(3Ir~TrgZrt0Z<%2 zNs|MSWPBgHa_6`p$X*ZS`R$$$EZY|HyR)jvi|+BZ(HK_gRz$;!c!+@t+2)M_7F+~B zq|dM^3vpqLC8k*#W6ESb(+K>kn08w(fT#`1)nSDhL{@;)UKk}p@5PwnfW!+e%3Hj0 zx=m9t3#b^NheT!KT;yocH$iP|zYqgHEzFU?X2BWKI~QX)zzZ|$K?q0lHGc4UK4gZs zunnJ)@*3P1|F*zdywQdpB5Gl#s&6utie{NUW7e=2!>^ zP;Qe$`)0^UG2ILegb5nFtk0d34xBpoLZH^P75H+_k-h4udDvYp-3yf`c+@Jxvx!Pn z$AdA%j?nU;+F>?5;MUKJ;b+sJzlfEmSQuSIA5gvzz3a1iT@mL2P#{#$CoEfl5}tbe zfzrb$I=mAqU2^HO>Mp6?bl61>Yk>BrG#ow?9V2-G^$Y(E=!MLn(B-<@aNG!#!r^s4L zXW9V3tceudqC!qNj(&?{#={;&IOd(W^}-1e6XSQjGe6jjh_#7bd&n*q4k5#?=CBB! zfnpC*P{yn4=k1er(dphjs!mcXs8+Q}2HcVYdsI;(L+~Ay!DG4TDi8>n&Z9KnfZJW= zmM`ToCyZIbQKCcORn8hPANyPTVeV!VlGRJpYDuyx3*u85eIy43dq#p^5 zthwBt+?ZS22Z=Gi4d)IxL1N-t`x;NbX*MaPQy%@4oE$0s#!-q>j>E z_b#Vb_H!eB;+J}p)>AL}Tf#?|n>NI>k<2~)U^vcvrSu!(<#P`L&Cvh(9@)TL% zBAsHvD^A8(v~l;rL_Rj?mJ>Q;6RH>_jz_bAGWWPnZjvPh=8KA_7miJq(w-*^ZALZd zcJ@7%X10yW2oa-+D*_TH{l*NDz~4@tN^Ws8G%h^3MO)}0^ipg$MLH?nW#%D-0?!BE z57PJZ?)le}QcWd2!0#lGs;)jmj!RmIzF!!_^z$;M{Zo$e^mwIN0O1vw5zo`{5!d{7 zPtgx3YJ+>idxUyq!nww~O~b6Thu1>Nq5S_KeK#=E6K1DZQ4qJ6-aE5W5e@ZK8N!3U zAa13(GdIV*IYb{Xxe=VIPL`eYUorrO517uwR!?7&?4JhJoB5I!WySnbF-RpWInTS` zvqTTu&3s^wP&ASvaZh;sltDMt0&1gk!JyU4GvG}I-7bK7vY-U;7SCk^o4@W@ zm~CD6Z=<|E_btq9`Hap6x$&!{mUmISlitniC1)9YmYvg&0WD>%phH}&sL|XAS;og{ zppSeSeim}?+7vaalk^egBW0y*K(SZRCyiMcFTtJ7)sxa-2OsEsm1`zd({&D8?cjHJ zRJOsE(=irfV0AAi+u*uB#NWTYQ8vwNv|1O1Tq7xyKnHpFin)MdK_fPs(xpr}sxna) z?T4&z$OOlTXblff;|XK3FQyA4UFA_#re~gdAP{v>4$`PTf+3gHp{qktPGF74SltAx z!Fcj=%xHPilVcA!p~ZD4g1HZld%Wl}VA-0x3$I9#Z#)-RHI31K%;a(wBA#yeTad?|kI#_-7Y+|^voLKj6#G6!)=;|Tq7wf&sA+-f1CzY_{{R|b zFzo<@lZO1+-EsqjY;Uofj~Nh!gWumz(kBrc1xNKvqB6y&i|*o|Uyv=i56LUXC0)LG zzGp~Jc%~*w)a89Z(lu`nL^V)1rcoOuik)^(_1D{E3)2UCZ>v2rG+pyzm>LamZ7Y+Q z)){-nbHZ3t+Pd6DE1*6sjYQT%7T?A?_dc^ z?#y~MY($#Z;0LC!W5(CYaKP}WusGVAY-4z3X!;MQnk|Xn?+i=G)=6Z)g$3V9v7qd@ zgVG^aIwla~V;xXrLv;w|tDrLqekoL}H$L|F8bPfBL(h|LM2VRTR5|B2k}Phi4p(_C4*n&-T+FA;#glp6Sni zaA4UCGgLGi)>e@1E*zgbWP#yb6boE3xxjtru>@A%@>Uulx*})`gXFN)0clV$nI2lB zYSgwUW0;#EC!x@Cmq!i{1WFhaWBJh8W`Q9%R@J4vCOyilb(Sm=2Qy*=()PfTdmBzb zrq0Y87@*B~>HUt+8M54k@se(VmyHw~OOf@Iu7WYBvox5%2Z|<%%H6;svtQLMy(=8w zc?E~cPEc?k)iK_|2^14v)lW?lnn6+a`{##z=1r$gXwKJ0a1xkeBIsR&DU=asrWqt>Jv} z5c5yXtpBirY6iXc=rvK~)sUH+Z#kiF-38OeZ{GqR{`M0eh9s7B1UsspSQC#=0K*5V1Ff-CAG|4b( z26`8bkfJjFJ_i8G9;A*Pciem6gw%;2e7m6{((HD8J3ge?QxrKt z=~8`aW!ZF`GM%3opo1j29x&8PY;s}7Y0*>1(=E{7P2M_C|xG$g;LzLUcJg}h_l`xm73Ck zJrHNz&5NIMIwDnb$PYDv%jiF&%e^V4H;{feL@ z*1<#m{6Vo76wmL;u`M`(V#1Fe&W(K03xtrVumuvS`-NBiQq)lk-x&;mAL9-P`xeRL zOZH-d7B@&-k^IN+mYC<3$clxbFU+!sH|sO>Bk|EZJEd5 z#^2uW8y27X+I`*OKp=JKBOBOybss@WOk=AWQY+8>`Nk#7zGb`)wTQ|(V7III^470l zwMLI`s``R>WuSr8mj~=%vo)w!x>mUsMCuO*W{}KpZ>5%&{riR2ptA9M*B}F9+!%XA!+As zz`(qBjQ_2nM*27OWw6(14f!)p|82H?jV0&m$5J}v!gNJ8N{2qDS51Rm=tMf%^ zv|$F^*gB&T)Z4n`Jv z&l2xepzSz8n%F@%s8-CC?B)UQ5EO&7DNJTg|B&}k)&k}*cfR=fygr9-H_RuskipTk^TgDE7XZ1e%Yn=QhYHrVxD=#`= z_WWqq4^Dn7&EK@BhPAk2uR3AYMdhA=qncDrt?ZV^uK9Q7o(PL(a>9VQUVGIKl-LpX;5R!-aH(cvss(JuyI?!59HAD-F5xcP%T{Yq|*whKEwnHKga zkzzMeB$m=`bN^%-Rxss@3Mu^eskk@18DcCL0ajxtW?mE*Fv)_xkS^ihm);-=KB+!%W2Us2#=|bMEdWvW$M=?EaIhFMJn{Yv zO$+nU%sSq`WC;uKW9_7?W;QBL5NvY-Dq4HVfZ`GVc=$be?i}Rxi}Q+RmXVp?u;|mSTe${5rTFAl!whmT*0}5B_ zK>6*TETu>i$6X2#qrC!kCm zxEUC1&Cb9DX)e6PvcctGiru8DO9F1vrdkIKLSwpE;^Z|HF4CI?4?McVR|ADGo{a+L z&T*N^Mq_!LiX?OALBHCOA84MwN;-;{lU#0YjSEBMxW#05fMP*zu$a=-D)umSUb!Gx z2UaJ+=fbNysK;+yid;7}dRhzczC?-Qr*sP1G=pv*s*|T|_sI-B2o#$7emZMWm1Zv# z%QVUM%zWPy(!1UVj_D2TS+XlkpQ=GNp=F|c2{s;k%x3t=j|MrJ*MNIHINC` zyP8ku{kkwdZ0vN_$zvp#blM6$rU-~~x;%Kj%E)tC9-Il3QKmTwL&&*vDrQ{)h7AZq zqlgO$P6}mgbkoM1^nVb$RdcK(;VkFFRr?aMY++e1}JP%3M#`9+eR)+FUnXe^y0Ff;BV8 z7sRKrK`?@Jy} zriN(}Y>L?#PB1}3ARH^`hx&2!Z{DW`Z_XZy+DJ49NcI=Ty*X-OH}_F&2}O3B*vz7Z zcLe$p_92}Z+5*X6sD6NM@X|z{LEQwqn#Su`1qBOya$cOt+e!Dl_24xiuCETb5t2*y zdlvHabxZ^7lVZMU6>}pb)#sr9eaKNRr47Lj+^rugZY+$X19$7|zVp3a>$=&eog3G+ zlEg+WEpClNE}v*8ounz2BnS;2dES+x6{2IX3)vc$KI^LFuHw$z0pK}o7F=5NAfiD~ zBZ(5-SolDCk?9f-D7xgSJ{v-h1fc5qE^AJmoj+$bScO;bwC;gRw*>en(1RCCU_SJ z|5Y9(#d9U1Wy^-)VR9?L8f)05U6shD( zV))3dt3MsFf5B}Aa~;wo$=WmHo$_ZtE^H2=Q!d!X#g{kSU+rslReFCcUqyDf@J0eK z>JDR%FQM4o6e*;1ar|y+qVn$IKB7m~OJpZ9%0V`22Ii%DUI=KC#fh>#^#~ACq^=RF zn)=9Dpt3{_CCHh9+E0D|!UP`X)vWhQ39VJ+gyw{z!k2^1i2VSu@-d!z=~$fD2FM7@ zvYWSe`wPooykWz$K_$MZMymzWwZ$)Zc0Q=;hzcsUkWrruIy|0%WJ z5_o#{abZ~42>c&j&_-$kOI{xOU9T+{e@1Qro%cz);7hl?OJ2_R!|Nv{IS|{!s9v74 z5@algZ8Siz?7=y&Y}=VdD;{5zn6Zl}|v zMWHPOB;uxRBsaaY7A4D0i9QW8%!>6Xp4O;60IhD%s{&G?&3*@YtF+s2kJ#D&jRp`7 zT-nE_n%j+mr|xv_3ELPL$%U=) zZJ#bSTbZ>f*MChmd|_rHC@2{=6K$v1WQuH|bS09*P%Kcbz+6idyI&`Y5+QHHcJB<{ z26rGwoVsGpMKS8lICua>xhkT^CyAX?%hRiBShzi8ri{gCTOyv+TQg zRqkHcG@iSK+@SQV!FTMn!{2=CAKqqg)c<_{J0!`4r=J5B!<|R5ITXo6u7(!e5e|-F zj`349eX?%Il8p*#7U)g6`Q^bKa=0KX2q;tMd+67DedN^)HSqD_37QVSIDQ9RrY;gg zuBLq&tfPAT+(!SY31(F9Y0piZ49Ub_$G&ma+K@cGtK-7C*fz{Xhj6W+NwzN_O=1+J z>JT2DhXLDMx>kcdZ=kV5;(kNld@l!sEi8Ahvn9W(3H?B{yLHbWJGwEX}SFnv_LT(Di5}oQbKP8 z*Bhx@QM)9U#`D>ld`Wq5y&9#_V<7m4cfwA!_t~(f;78J2x|bXeyeRGfVQRbs>yYrS zdZs*hjn~bPWTujCrpv`RYd$o}{iLd+LmSDUTX}Gzvfr~_jeiv~)wrx)Yh?6)trD4G z_-^UszmWCZVCcd-10ZBG%p2ZDu}S!0b*(dw2R3R~PT9-cWCDkl@onr$IOULO8cp{&6svlll4-A0!BBSCuWu73tem=fxnG zK=5jl?25FJ*CZ>LmO6dVEsg$A4F+yj?X0_iJv}zcmn8UPYsw)-@0~S}6~5j3G?0my z${*WkO#g?{$69$M|D6J9AE{yv1)P&^fQ+l*PC8czJ>^zA_QVrRa+($=el>qv?<_Om zqTijio}^78P@6XFi@ckHO&rh<(QTg9_@zUBrl0kYbJPe~N6jQ^dNV1dD@ix+@Vr!Y z+*HG?N==8nUK{7tA;&h=>U92og@NA%`@`$VrNFqUmB9QJ$19K(L^$W`?tp=L@-Q8F zz-eGy*Ur!X)#dfpI>21AFwjb zaSnW+1vxvR{j4sWa_#v3)~7OJo?e=}@W|W9d+C{b3rfpSdB>>SSOeRN0%YCqW;M?YTCY|bki=1;| z$L6ku9k@!dO%%`_>I!*X^k#XnWSw>yb6T?}VAZ_Mq5A^T-OD|pN;~uj#CzAX)#_W? z7T8zXMOsL$;y(W~f-88_T`kJ)kbdDVX@_bvZ->m_(dltw<__6bSvJ&kKV%9-t%|+U zy^2B*(Y`LY>K!GjRAdPU+$sX=Wj!)HHh5eYWG%WuOjjDUABqb>`Y+wPM^&rXAJik< z#T!t7NQ%0cw@R4H^xB)}IT|ooAdB-0E7aI{X&X!nL5R7?(3S%4DF|#efM$VVnww5yt{f z+5;0?_d4^uZJ7M&lRG8W^gdkHzplHox5ZL~T1vSfgOd~pT5OWsfaSBR={oud1pfCb zj;RKe>-kj?@qEm++rl2>H);#R`+4mlxgJf*qAA_HC7<8WOk>6CYNkb*A?#KT-Q1{! zIAADtv6~me3@DeRs_TLcGaLX%d%!iK0ea@ivBOSsbFtcbCz}g|TzDaEV~<Vy zKFpQ~eecVyGf)pQ!N(|Fj^+om!t+IX(=P`Mygft29US)bI`%dPQh#*s4YOwwSte*B zRW9tAT(g*7>L?b7&dyM}JnxIjl!ei3JAX~!4jSm z?Xcj!urwSu2lsp3dFjsFtB4`=?*w{~Lc(iZPVhVe~WsxQ45P_Wm-U$ERWg1V*4DrUjg+ zO@3chuU$E5lV3X9>wZ~$oLM!=)Ve`C;1(rHlNKviPQwCFnzbt@Ny}WpNtfWDUf7_PXW(u|Wj1ieCYl6N!sa%=iM+MS~YB*ehKV_}R3_{5bzhKoE3)@?~+d4-)?#;5BNu zl2$?4g1vq_G);@IO4}42u(bqwSyVY)38w6WSxz!iV+YKTJ#p-BZh)DjP4auu={{PN z4{4-9&G#!8FLz%yMc?V(ELbx;RhAQaWcOZg5>SFbB^;tgG6&!usvYBePA3JJ3;Z>L3DHjPN0}rr2cI2i@x2wo0rbSs_rV+ zgvK!K{1aYxNYH^V#+Xu611mG$7b=jLk;F!U%6BG$a>!m6c6`rUEQgL!Y!yYyDcwp{e8ipr(@)IxOjKe)b(CmVM4vP(>?Ch5Q~;np zal-sn^Y#E~ZiL*jBc4X=m#k@yhDEfSo#kXE*P7e!UaYg{Zsq!Vxtu|7)|V5*J-`d=CVoP!UZcf z*5II8B<&7IS&{+82kwVprF9|fiU-KobkNA*u0KOEX2Hsv6p0ciFNOc|H1#Tjsx-Vy zZW6kIgqb-04WL?eQc<=8aAgA-V>|tU`N=xglb4?&`1o$cza8j2peA@e5OR)@y;E+Z5YMk!DJ_WKEz!SSH-) zb(vXGuT6ThNW-f=k+||skJ1QE2{GUuSc<6zDPL`#?TC@M(`V1d( zPS$6__rh+z^1n-dv1~zs=$fD;XwWT>N#~pD4~@w#xpb1Ri5BET@guJe8p(x>zHLcR ztE!pK0b11#8s*BN&Zt0yH>awRU&K^fa+&0d9(nDXSp*f*4h5uRjmdDNjO}>FH(z6? zo9C@>6#VixWSa{wr4L%nM1>TaN0A&#hr`<}xIg)>bUC?BBR}fkH?JA_Q3u=_J*w!# z0zB9F=FquBUTr{o*R3q)Rq?aw&EbZy z@?b+`EYFnn-6zYXPq;e)MWcYs2o9*@fx`_lNtF6cKQmJD2b%nc}`I#eADFma4{_T2|xE&pL#$$sw*S@J zKP`BDf8-zE{bQr{?n@uN^ubI_+m2zXW;SZC%07@f7CMeWjS)x=g$^gwPzM@hJ2Yn0 zoYwtg6DfJYn7wN)kaUD%D=2apcvb^@MW)L4qyWRzWmEMhX2$RnmB^#9cV;)%E1r!w ziwqtjBkY;iD$&kHjq5Sq5y1su zn`-p=Xy$3nIcc-tv@&-=A1R~b_{+&MQ8W|l<5;8eoDt#zjpuab(95==5x#J$zcq)H z%Z6BNutFin4ia@4d3=(Yf`CF<93Kl?OT!QPn@+>`|Ac$1tTe1`Ty!E}@VG$1WCk2M z;x;?84TTZ%-X`uXy4PWCIk;>bx^Q)o4V&_Wprb6TS#VC6P3x~p2NVW&dqiv42DVAo zrigPN^AuupbVj(q4M-z&;G|Q{@`7{LDiKdlDlQBF8_E%fG)Dq1i?hPAMeD)O0Ci22 z+-wUw;hv)I3*S4Fv;N1}9zTcCAKM#oyxX=JIr_EIJZpvo8$<*soB)|*==$MSZJoT3 zskh)ounj(Y9D+htfH)853Zf=(2JG3`Ome z{lr?7$Yo874Vj@P*-DUd09T`1@gT$`iI}KFM|9B5r2Jh=_Y3vO{zxv`;F~>8OPEpM z!wz#W_9mS0F=5l60=In496+1c{j0m=xC@6}Z(0P)8YmW0YHKK6oxEGQfyHoF8$>oR zrnQ>CJ>dL;lu$j+N{PyPb(~QX@t(X52y>UF(OvRnSv@F8pj14B#IiN5^if``Y`Lh! zKaQAw!f6mgqY-YoG?LHh>*d`c7$vJzQ`xD~Jz3AZty zL0=%+!9tn|zg%qp!g>_A8k(j?_uEd$al_TIum1CIZ<=vcI_1$%2~;IMz5H?E#97E~ z8W!TeK(X}{siSoJNVhT%K!AmbiBR6Y9o&~r?;2HgM2+f#*eJ~kb%^Xk21(khn48`` z^Y-vsl&ITU%$#u7_mK_kJvpY=Vh^UKQNXoa8lI?Z7OV`+rY}!f=LPYY{i?LdPIEfU8xrjZ8NW)cu!$Q zR@(Ic`U6RHVT)C2fud}R1?lKCN>?lB^iJh93o2VJAru_Jut5D^e7O zH-bJ|Ez}uy5+uO4aO zavz1z>qw(MYNS1DrW@&vIG%gGjTb97W9Rf+U%yH=a>I@bhxK5mdf2?SgJL20w2jhX zSPz%zg)<9h!t9evr)u&=o$mWpIQ+neio1rFKuWqriEA{}B|+~f$FaF_BahN(K{9f0 zjys!mZt3q|bY4qLdj!1(O)n`&aTti#-1DykE*B#JFv%tU<&vOOHPX!KafcaGS#Emc zT}{e-k&}bHqaAoK4vrh;5qjb@4^1?@C2an(*_M3vm9PGs9N}h5T$teHx`iz{Pq8%= z`4D{MmmW)NBv@}4D;a*=sLi6c2V5WrG$w`xRJ6FkRs+jIlqh!kWB&D7wc=It8nyqD ze5#5P-Dk@pc0kKSC005nv3=rl^=Bki9V_VthK6eTfp1-K5#6xh%uCmx;xWm6pC7*C zTA&GmB^umuL);-g;0Jz|I^Depk|ajO<~#-nK5L4yx;6sFo^$GtfBH+=56qxSZ5KTz z4PO{gbz6X{m13JIat*ZZ-$4H*oqs9lav;cPg>3}-jiRt}7-U>b;p;Tt*y#!LGeME~ zL*_1vq^K*E2 zfDS%`Jd!5ATJh=$vgb#sulGcz`Df7`;-TxeAep$2?Db1xAFB3*Cd)nzAAo`ttk;bO zQqEnxjp^$9nBQed!EOKH$FJCf}Qc@vw?R0;r$*@GtPc-g)AZQ zlgMt10m`6QDD6n0beLa{Gsby3ogW_#(Oc2+z(U|cxEc}Xj_xJ~>~IZx&OJ|xjJTcd zpD9|$B~~}8hx_d3;*hJH<`LIjj=T}_cjiUcHOCy8mHasVN)?1un%I0vWdsBO>*W2y zZe9nSy*L(<6tOsDOO*Q92y%5S>SSOvVfx3 z7*)O<7RQH_yaoEheksac{~{ApufE+26SC4^=`~Jl;iCbQ1F-VkE*h5y$gL}^Z4@%& z;v@^ox{Q$z)7Q{BTZ7N}(Dppa=XBwe%iO+~E}Jd!_0vD{@HhJ;CtuxIMzX#zzV8tW z*J>}t7E`2%(ya*E0$UDyykjI?P)}W{XqGh#vNa!hV<-lBG5dw4TIp2v`Z>j#awa`A zKVSz?#neNMbCaxJ7{fF}D7isiNgF&G=g!QKNv#dT?CkbbMIJ!E@CPN>6C@cHPC4G*99`F zz6Xr2Nw|9A&r=NeX+J1p-`oc^5y;p9fZkc#lmWJru2~C3Z_y};=Nh6B5335*QCoqC}|n>KygogT$1c0;`iau z;|TVl$!2&LPC5QG;#+Rkl>siBcV#2j|3O5iXQ^bDbenH4aND+prE22l+!ZEhI$-o4 z!nIT6{J6w#o(vF|!G3bnI2>@ZY}tq3PkPZA|8IKNhqVb>1xS>P;m>JH-tdOv$}fD)ZFxiTRyOuEb=Kfmm*n| zZhd&g4E*0DTV5)zRb+y>*s3xFZE#oJmA(# zax}Lf=5~0V6Y;|VEJO3fu-~o%mZKkktCnvD%bcWdSCdsP3>Hv+80J}Qrr1psiKBE| zRi*St-iF}Tuw|l^{#|_B{VSzc1Xa?zcpJ$oXM$j)VYD9@Bi-%1bKI=RcVD@;+nRjC z2FJy1c`j2iD^-1R#+4aYJoW+6K(hcvj#~wfL2TnBq_HOXrmD|=qx`k{H`3nT`r4() zM(q*bn<2+}_a|@p(n-2w+P&%6HdWn4YE;Ac!Z)sOg`cS-+KEqz(Ma!c^>-j06Ey}gZ6kt=x zrGckWi#s(j!S%tw8b(?mgpv|^BiJcueuBg?!1_ra&N#3QYiaoJ)z;;UPe;*Qce4_* zZ->uB@e~_Fk@roz*PZSgsl7Ax4GW6G(&)4x$kjgU>D=)jY3%I8#7MU~^_*=??0>EO zw$kj0Y>5xNNH%|AA|iV&(2_&3P!*j{>GtwYYaRlFM!Gstlo@(UyX4WFEzAQ&X6T){ zNQ#}Rz7%+YS;nA}YZfFoJ(^P&w#u_ob`laMOPE4lijxI^?Szok{AhgXR0yTk3nK!C zT-LOI#*-Wuc3F>E4E=tJgufm`D@?cXc z4F1$_T8z5S%|@Y^Y)vhlCV^yuEC|@w(onNeCtn#@8rDjqh~f?9YEeNLCSh*(c3fiR z7@#l`JRFu-Y1^8N5DcF1yC3@2FlONV;Z$!mSus*Ikqi4LpdvqP{j-T;<0!I$((UHO z`s{_o+s)JVdEE~>D>zC2x$(^vL4U4`?C>A&na6IVIMqZv-MMX~=s)mXJlTwj+yCdv z-^p4RHYX4r8HRu)iiK$KCQ5fhxkH1s6(GbdXr-_Fwa&=ol`SZmQp|vHz=VbsLFtl5 zk>dykyK&(B`N9GWwsA0$#dsy+L6J3`;M0Dp3tJ2u6oV)b|98_COQW{Ld(f|0a7U3T z?NG)?oFTD5Xxim{nMs>@ct$6^2fUAq;kEL@$bMmm|D{E#>g_YCSmaN?=#3Po*Ook% zw#myS29K?>Z2zj64#yZbB#k)TI0a+2k@UNmW2yyaiz0kbok$M2utjOGuqY=f7Ce?J z5Ypu3@G2EW;%rf+V&$9%bC2^=e9|Q4;tY_nYz@QLar`Jzxwyrvcj3XDA?LL{I z7sY7{P3;X~31j4xa};Vuz>|eN;rhZhYDTa>lxD$ier_!T&LykSg;&rvM8mT+ht+#$ zHq*;R6|+86cX{6yLI!Puq-;T|I(1=5P-%EB9V3BL%l%ttBr2~7R%x-aXz4}LC(P2c z7;|a+K`CMx|C*1%BU#ln&EfV4N8AzU;))!5+W*kqYvGH|!+_-b&4PY9Yf+V^4YQGo z{dX${6u4sQR>m;;O7L&fA*(eNNEucJ4*D633NcGMRf8kHZmNORSIUxQr$jlSrW*GF zg;9wroxg@)2^tp3r_oUhTa=|@Jsu<~ull8^^~kkgqVCo=L&ZUt_xo~vzh`DBzR8h* zZXPCHcM%M9mkWA?cH+trNEyKfJ>!%;H~|E6G9%g`-g{HcGn(J;3`@vXZm}5`&WD4d z=V2c8PKpI}`W?7Hub8io5^W2>QSXq)Oo7CEiT=8GIelvWP485nk~y%(mZZ@SX(M?D z${r>7=vVUVyyEz~!VYVY6+R6DHFm{dsovZrPrk;i;&;nabXqvp1b`cZC`(mYQ@cjl;4$xT2n}Iu|Y1p z$g)8=6%~{$>t2XKxkhb)^b8G*37|~WOODIXEj})32sU`Ec&S!#pZ_n_v2hd7vmZ_7 zImS_Nc}_pm{l0r*EK`gNZiX!kk)^@H`1@Pg~Tw+MM(05 zqEkraljd{hI7+0c&)%J zwRu{qdjFh05;a~~RNMi=VU)+vbp9{D@~an}rv(`WaW|k=9>XB<4$=({hSY{O$$G-e zgY_i=t-zi7f%^?0Ww|Yf1+}_i!B+p?x!JG)s`ujZEg#Ku=05P_Ez|dgUqHNfRI6I|7tWVW+ z(6@NYM12HXU=@6Mzz!CPUQ&H-EbN<+s{TOOsOS^k6(-9z27?ckI|pcjppOjA;5>Zu zLAOdUh)AHH=Y_&+xpN+Q?F!SIT%bzfQ6P1|6S#L_i~`|@$?%F}YUwg*yYg3~`yScn z+Ki=D*GE5``du>se_8yiep2Vc`%|A<7|$CNdzB(hlrD#d+?T1myNmkR4skqN?^Pet zrbt!Sy4R?7YEF@rlQM+e${N+GNy+{Vk?Uaz{%B#xv@%Jj_iCm^`B_92GvL-a?ao|- zN0;wa5xBfTDL#$bjqaJDdlh}sTKQwYmYI+J3PTMnF0-m&zY34>J(>l}+0DEIvVVni zOv4RRkf4SP=FLLvdzJRQdra^#>2*LVZk?VW`O_18>O%(1Xtu|A|Bs-?4nxez`ci-U z-mELum1{N%#6Eo`MuLG=VG?R{DI%2>tP+F?&Pj2GOjVoiYFO~`f-SSFp#tg{+2?nXJ`MCF zXXXsJRWP;uGjr@QDqPUV3I5NThq%0{9c>H#+P=Y>!^LHH7Hp8qAbpu}Yssh)t}OPx z;@l9Pt;cH57~8tnndiBgIOV;+&@VcPcd>V~0Ke;%wBE!#lsMN2kTyk)>R!l|;GOgp z@k6E@Dq@G|?-J*NoKw43DS^FfQMZ$HOU#384^C{E0N23@rGIa;^`0b`y(iaQOA;G9 zn1=&bg&Osw*L$s%H+!xM-LHbAnHmC)t!kyJFwD_&#|HF^Vu7f#kka+O+&beXY|R`n%GDhVLsF!wnJvoO@)PdG9>pF==!f55 zRQgDV1SP`^Wb#WQWnmbSZ<3Yy-B_sKJZ-yIX*f!Zpzh&4=^7El&mBG#9Gm^;$!i3 zK@Y!*|8CN2^^t$BTU;CY$7Cjn9G3PA_XYIR2mCg$*9Dj8V<5HkiKcc@qqdhZ1jA|$ zGOo($kHwFr*98~6Pb#{7vo#l(>p%q4s6{c8WTupEQ)Ft&1*z)gq*Q#2cT8F?h>;An zg5qMhXVhO6M|~x1eP2fVRawDg_tHZK=sP}V$Z~F2a@}tU)VB=}1#YC+Sc1>cmLV3S5-}a#ElA^H0S@tWPfy9E56FSp z0=BqN$fX8i5Z7Q}mX6vpY*gct9j`-T-3nNcq|u)Up^g?*&X|+*{TW@}SQ&|vYM1;{(1)rF zNhYsWuy!&qHu!7^)dRPsI30>cvZ3cAG;9~=(nkWex#zf}NHx^xcFFZ^A$9WAUg=CB zXe2Fj-xaoLvc8MI240lRUkiuwBLPr>`Q(IQhOy6M#B>{~(xj^K?es3Y!_4g)5$gY+ zZZGTm6eBsSE`;S$eE2Qi3jcaF)+nvi+zuLa!w?iw5)~^q zeyIiC8r!V!$6njLT9lwC0n*lkZYUm$tM7|I5PG7=82?uUK0wRuc6-AVWJ3d1px?|EbR zoJTo|%XXq-6g`ja7EY+3{+jToA6fHIxhyJda899o3~UG;kzA1O3%E2#p8=c0k5$<| zjt?w9Z)EJm#5gaqIo3K(G&KG9+j_GbbS7dJpL9(kaTeaj@`>yqMIKQ)WNS)Q_sD7$ z6%@+$l!|LqjoO&t1fLCT@7ylmkEb<2-Esz1uSHf{5Sare_jObTwVrCwLfVk#oHW6w zS%8tzh6Nu$8FR91ji_GxkSP$|(ss;k)3m>_{?$KMP?sVbwDoc1x;rG(a~ZhEiqzwNaVVshg&RoH6_ zK6eJ5XI&g~ozwi_y5aSYcJ99YqQ{6a%>n;^=kuf(CSQ`M#5~&)u!Q4;R5PwP0M;=` zw6hm(fIXa(o8xAlybAyIo4=B%ktohwIG+JjAcvLX$5HGCiman_t+Yw_@uI9U@L@=k ztk-|T?D5412bkD7vJP!#8z$!t`p=(khQv1te)$`+&4t4{2Q7xbkYYhGCx_CN3pRk7 zVxqEM-2qwrry(yG^oo4+n9iTzGXOFnnAD$3SMkr!&~NwZppDUyI6j6bO{9r2P=$@_ zaea=d+&GkwY;t1X)h0PW!(%Eyy}?n@b0-J7)rrhTE2{6q`tqjYGD@ zm~@vf>IqMyyOsLnDLGJTi&Y$T^4vM);^&7+95`~^faJh8eoj|EJGR&g{_qjy5Stu?}@AoF-)!d5mFiG{J@^Q7_qh=4g-UYjxsK_|Im7MA)3#@h5g0Nj_z1r-TW+3BPQUf_t7PLO z;8!1RhIddbM9#NSIzw=}azJ5-$PK?P&EWO1wE__9;CC%Do$CuZrdqMElGJOrFjcCJ zq(HSB%JGj$8?@tKqU{+A>o41QB&%^63)h{PztS0d+*)kYhDv3LcPm{GP$)}OwuK#h zS$|Zs7D{*Z_3C2pvIVt@Qt^P|6v(VYoiQAjhNG_f$I}MgnxHt``AdY+p#*T_)bA_D z{T0`przao&ZOng}P05`Fv9rk?ZjPzz{$*sf#s1a5D7Kd(-B6b*z%-RMIw`OhP~HlO zGEt&W#qTQz-O^|zvTUW#kdK3N=>ky<6D`;0840D5{nKbYmboC|bD}auodEI!sp=Ny ziiiH15{VO=1xS-vNvDS{8M?Z|r(O;G;WRilSEa|NEzzf{kN6gdfO_1mPQKH#-?Jic zhffFWYgO{Qr8_+nl{dY6=AE0i#zTKad@v%L&LWr=k*xtvDvSY{5-9Y?AU~~xzRXlX zZbh>HeHxN3FxMxB>GZ}wshhWVW{+<(_}sCcb)GTM$Yr;!+#~y+qQe4B#eNmi zsp#WKwE`*wl<`ycLg^fy8VBFXVhHSog|nwjiw^}aIKg7#TNf8-tatyotd(6jL}#O( zq(ym)7qu`x_#rPBsNo^qrJmX8c^QKG=}_icF@Kz4WH*Y&8Ag}x*1YwiuaAeS+}pkK zftCmSg^J0k-s3#J#&4%xgKYgh9*5W5Ohx?TdV_T$2^Ul0!cMJ?d>CB489YjXN(EU? zF_#w1g^A~M=p|N=Y?ZZ#w2fb3lwD}DSmKWZl;ACze|qo!%jQK-*_^+uCIuty>9{Zo zPFVzd4pA&9ithtWG({ZWyhHU~was?V&D~tQ#Bh{2h|>N z$A|@BMttE26xp^ozbGmA#5!p2vdb15VfZTFJxuI^!q9mBmEgkAWT>q`?$b(s?wn@9 zCodbkhF`0so9P?IT*~p7kw%3FM-2FL`ml|M5hAqI8(-@G{U6L0=Kt3HVh*|M!ucob zECMBuDE40z>7{i2%FjTZeA&Xn@Oo`gq+!MdCMK{!yD7LunaR5a40Q*-(-ryW_BU=v z{<-qo4f0-PyyO;W(C?rgMm~uA^LrTRKfsF#+|BD(uA17f-otCu{!6xU;d*|851#vs zk3HX(_k|$mZW_Iw9jU)s@uuJhP=wc{eJG5NXjkTHcF3;H?BQLVKj4-Ks$Bi*OUC0< zWV2+4Z0OZPVSR7`a2=kXQ7i9sFJKHFO^{2R#|(W+>@`(A^toD~z-j1HyKJn48~o-@ z{Pk;UGyJq?etMW>aEt1>Zd998SWJDz6bnSd1(a@Aw9 zTxM^{n_EF#fRF<3<*UGU)A35*!D6ltt zi=Z&U_u#A(j43$6d}KTftUF=HmOVqqw`!3{Gm>@Gp!!m#|x z`xJYRBJGr}PTnlQ;>0$(gU*5o4LC+vf{nRQ$k%~tYru6xXZZs8Z7cmi+VZuih1qo3 z0=&@8%b3+cV_0~Ff2um2U$)>D?}MOw{-28VSb1`+zDp5ENR*WX^4hzH+mDpLipmR9Mk^?a}G7`C+eKOVly6D~iYrRl&Sy#%&((8!t z6|r&K3~Q1)=VeRvQS(qLt1PUGAkR50#F}KgW;gTn*dTX$wA}f*3||mB^NDL8ta|vO z<0F^u7p8~8mmG)xins#`R1XCd^T62jdzOP6f;IsefQ{GlMdOqhF&c~T)q1He8$_B(m4yiRVx( z1mDvsT`BC?0Vx6`#OlwBlLC)O2Hmi@G(O^HM3hLM$je(~I&4uIW_&Emp=&_V46D&| z=~~$>4=glv>L_r3_Mgwl4t>~Go!Nf&Bdm&DcL(O`sjgS76RNlbj9qxWWFs%EHEcD% zQG17EEqXk+i@u^dCCUO)n|keo?_ZNPYO{GCNgA-GS8@_|o{GecT2!$BX@K#lPa7T9 zt$^u_m+RjDnb2%a%D(*OdGfvs$8?~CXxNe`g<`i-@LY9$LevZ~X@K;K6C?%Fe6IwF zz967Xogn#`6iAQE0B;q`5%)M=)ZWIJa2{_P{W$tGC)+k5_sCmPz8M{JlD=I{R=F@b zvMkWCnPN9lB#zRdKEEC^ms2@s#dE@k4^bL&qguzF2sl$-{I|)aSfF za+x0ADrS74VLvh)8zXxjsyHEI;x>PS)Oz*8#p`h4h?|Xd&8FZ=|7t~V_KcvVh49J?Qok0!;Xy`DX_MW^A@?J4LM!%1AE0K#XQJL>o474!! zAnuwbNe|WUSM__w@t6B!s%$NtCq2!N53i-W1iNM$Gj^lAj!)BL=6Z^6v!{Lw$k&6m z{lA&+y~*xk^~b}r=^DxAX@_7nZ3@s{A)u&bJku-A^5qUh?uhz)wZ`d)BhVVtN_M~K zJ_<&|4}y0B1c4c3BR>m>pFs1jRaQX&2srS0(s8mI#)r>os5uVTSekR2iGDkU_tg5= zW#jNRRoTP8_iUTVZkNKE2U&0i7 z4=AhYO2s&!g%e1g7=W=1zkT;|0tt0+ZN(yMk_49pi4EF=Iyq)mnscO1kvJp0*lu1r zV^Y3s4v8bT$v7b5$@p-B#*@38f4~VE6Ka!Vf(2$*W$}eAN69)DUgPYva6XbLb_+!k zC>>7yxCd+W6^eXPl>5DkRr-oqKuOx-z1+WrITHAhq#AfmJDkI$GQya00tecS5&Ce( zDNf*+$X|A2{ZunJ{C;OxLbi^yN#nY~VY1%>Fgqz0R>?ak-A(U)&+DN;*d7cCl=MTS zRZi95K0=f1<1lbWT9oNC>Q$z+Nj<xwM%6|uZQS$6@i*50piMa!a6X`1jO^Lm7ZYO}1`Ze)+Z%Db z+cpM9+Bj0ye0X4~*@RsB?#g+j$A#yYjTY&Y(G%Ha6UhLj!z`nE^#(SN>4W;xR#}v& zFub2v8lI@cdeqBP)_J9>(Rvs>>L67yosVVKJ>fk<6jJLKqFxkEL0Os-uM%0NR}%#B zF{=PA$<2_HfVsAi7zSr6Os359Lls3-Q^bU(1RoR+zvo}a++->h_3}1NC2%pI$-?(* z3%Si;2&MD0fYxKU72hg82Sx zslXgx71{;sa$uQ%4Jlgd1UqIN^7~YZ(VP+*7w=z;w8{zcV?Ak$%`lv|^rL2C_`*bF znk?qHvlM%pA}1-`T4jndQTZ9brI5mZ@4WL4GP`5>S*rS-caV+!3`B;}p)RCMjgbjB z6)FYtG>N=A-_r1y;7%j27^wce^UgB&Q8B=&2q|R1IdNgTsAmS&(rCASVn0uh_6SV2v<=^jR7ywy5Njb4y-i#bDBre zo8H(Q80O`WeTjJw&2-c}a}Th|g!|i00T)*wjWo+R15FLz+4zBV0t*+oxbS|XjdT}; zL>g=r7-l5+=ueU3fmKYY&>+0+y+;hIb*zTP%srE9;}j^2FpQkhF+v~CICb=&Kid1h zOO~}xj)XfH5Z*`byaZc}=*DJg&MoHPS~_SX>;;8`-p=lBlKDac<@yi5o;mvE7Kb-2VCM&qDfnC^D zwV^=u_uFJoK+C*Azms&UAO|YN4oKFx?~vh{afggucmaQ6oM7_gZGJc5n@7izK<&a7 z#Kv~>5M_(el3-FoyH_c>B8FK6JxnMVRICHiga=ZSw+ig)NOYSg#cCjDPcQa@D;L9R|f)|326q}n2?aFAjTP*9=)nLwE8gKqsX=`zut|Hs~&z%`Yg z`{N$*gyfAO8^PoVDiT2uM+8Cz9Mmq;-syh(>vlW$_BZYDzpcH!ZEu@S+v&YU1$Ph> zR6qmB64_)GL}eWica}kA#1%ob4lba`pyKa&lIV~~oScyGkM?#xm6Njt&ig&j`@GNc z{rc8((w%RRJRwwlp#DL@vJvj%FI5U^FI zTaf650ag8RnB;5dc2Op`!lT6_(aUzB2m^7KQv7=%cD8^nfIosTIgDtPRO8f@J z1_fs0F#Q7U!Wnb2#JWb|p+!|;S*k;jAIz9j3JQv8;E^WEyXi!^mX4kA)Cm(MR~5Mu zJhI!pEU4aotvr?uznKXRcH0XK>?S>F>Y=~A>UFXD+UXwU2_`ih3r*Df<#*{U5RIx` zqCN=%%9Csz*Z{*`AQN-da<4!z?gy*(fd6^qHt5IU^hBn&I^$HD{Zm=ph zesS@?$%bk0>W$5=CsPcRy=|i+vBd!c2i5$1-?9*FZ-BxI3DCj!C~=`XEbVs9fhwR$ zTnDgws5W|H7TkuFVRzj3EY7*>XpjLLBIjk9q?p}|jC{-)aV?Uce&earDdjcyA}&gSW2>?*U(*F=b|@b~#ODlu z#I?mENs$mzMP~zP0hG@zx(?~TWZyJ8;SKemq|`r2kuj%3c1C_yZUgJX?oAkX8X9+} z1z>;aDeYeGyg>f=TXG;bsSaO6~uufgmrNc&p?$#LLHv_>OCQb93g6v$tJA<-ytuEg?d zkVcUXf}Tp7Okbm)CLVPHWut5E8FROIq(KkpS;14MB6+Js-=qZC<)<8<6NNh*fAXbCG$Fj1w=O8h*62+_ZzAp=QU|4KB$Cxd(5OqYmtn zTsLw_Iw|H1McS#zY!xz8?4>nc9RUTr0$#UkkK%sthzsgvtFaT}!JZG~YTn zStLufAJi5Kcy-{)AW3MPI8~`G0OsGq86R>7p}t;!qk+&|&d{QaxfR+yi%dG%NM4*C(q3gS5ez%jnGFxf@oN9!X=t_frBL1c!w6!BN=g72jA4hIqGW( zQeHY}_JYw%ztn<*tYAd_&-z8>rp1vC>)enxX_W7Y;8Q1>J><0cG+=?%($M)(Kh#joeJK zLwqA>kb97Gp9`!jE}anF5}5_Eme zq+}PY{nz=-siqBpuSvehOuJyVYBP5yvrlx*9izVeqCs#-G|GgK8X8HR(&+Wjq)|YB z2*oUil>0k+#ctn!hQ#sb+g@_stp4eAqY6{f;C$un${mJ3Nzo&(qCQ<%wy=7VI%cdG2wObo z&%Fn$fpFY7%U;jUGo}%IHa>{s{$<2WyuVw}yc~f($~zj`NLpPwX7$q>LRSa&gNR)n zJq$vsz4EodyK>H}PiAK$^BKt4qWrTf?Ys4Z%AUb*84SpQAA7$;POvLTaNJT0a?{8T zc2P_xMb1!>dR7wDQtg+GDE7@5m>1*M>?&UPLMnV1 zgX{H5?`*h?#{R*RFa~44&f;xWo{{5LtbM7W1 z!0YZ!MpT4PC}xBr52?s*dT=oqp90Pf&K9p$nHGZnsJPHC0)v;Wx&lqg)xu-*@Pt~D z=6)(je-J8<=XH6v%JzUN490SL6oZRF+K_kD1CM`pG+%Pa_fAkQrv$P_`Y%}p8KpXT zo_8;N5qw3e*8qK&=utOQoQ)?GIb{XVi322_UOOXy5Y&uca)z00} zci#a!o5cfkGE>AI2*rjLy-@2@r)<@laQwB@_^nzg64DhwW^9XB1z#N}J}v4Lbtz8I zdObJ@r#esl`9)6y{+=)=HV_Rvf82qWYln@@)jo;=#nru3R{+kp+%W+Q_&N->WpGE7CLgVJ#yT`<2yoKKGiA5xtO zI|pUmw|LDuUQ*;oEepY=&>_`&N~h)weI`Znb3qsR8s)|%>#0J10&^6S-fMzKmunYD;Vdm(7rOHH@*(=q`b1=|OHS z(J1dp%0-1iE8nXKOCT5d@6}0rINPRQ?Qlyu>3j#;w>gZWRA**ITmcW9)2O^8tKH0Zp~aj zUmX>msCpWtQFVL6=10mX*Yiv5z#yd^z~PCQ?M{(4REmk07z*%R+L@vzGw-G|H38X7Pw3#Xel8 z#)HiYJkLF0<9pcS(5McJ7EPmhVQvB=FP;`;Rbjt-s5*>Dyh>198U4 zgDf{Jy1~WNE;y*t3q#;`2KOE`Mkj={$ua|DLr3VWrS&RY;3DWY256e<3edEz!VbT(sQt&YDxey6^N0s9qDmtwW^L!b-6Z(mR0 zZagB9M?o+khua0*09Szw1Ln7M4ibG|AZ61p2E=iqWus14$vXHqes%?fVY<(bsBXVdZTyFZP1b) z^4aIr!O=54TRC*>9;p6ocEfGWH&}NKD^oLdR-G_&n!(gW^!`pwRylC!IL*j-Bv4Eo zMK)29Mht1FpL9FGG^L_UP-EvE#(|})b677b5d@VzY{n!)Dq0Br|e*NCg-3tts;x8Zc z9433&IVBDpwmV{ER}N6jeu|V*ktlAKq-cOb^bu~WtO^=eP=;m@$^tj@t9f`7*X1CM zMnB-MgIccb;k6`b`YC0dbT6H%#3WS|zhO?Tdq2=04U^2kOi2$Xj-%yooLSA}32V3~ zNMVpofoCbGUg~5R4w?gqHTxGnq8dyq)E#!EW2SU|yUT{qLjEb}RVoV8ELjzH*ST;; zz3gk$42K*e8x)-paRGRwTs|x*hq=e5 zbH_`bG9F@ufX}&?Ef1SFGbIa8|BGodkxkQ!1N$px(u_!$i1~?Dmm|)}?3GC zY6N$oeq}v1nRsCZ8CwG4^Y69(!J9ubUGlO4h69@uGYjNQ$szAhm>jLLZ?OtHG&jz>uAjLVU+Em9G1rV4v?l0 zFg@IAFh865>46_;l+{Y;ZqY~kw@vQ{78!JOq4fl9R3nKC7=Sx60A z@0D5`o8f9*$3q^fEAIV@g22Bo-sPLHJdT4+oxo}V&9TwE5tn4unn`5|Woa1Mj-&PG zn>UQ<(Or#;4Ynr3D{DL1_uAN+lSa0thGIYl=KvL1>fTHD(b(`mAWM;UD2K@jk}+5P zj8S~GFjulNP1k{z3_*Vlzguum(kB}Ps__=O!Xu7@=|wO{cs>wXsq%euIVZ^4hyuEN z-qwKk@qc|Z($}AI1v`^GFJE|xgH?eOU3J^hx>%1ZxH7COa~@*=kjtZ?^JK>~0$GzW zuA`P>AWo4%MRtLXVVXZEIXt2_xD6@LxzTByT>$2vR%hf|co;fBfSKJxql)8PAorl) zQVWsMIB_q~Umo^vSVk|5gR5(hgoH}+0+2E&@hPBsY3z8#)i($Lwt=(TD?RMDRwv7a z-H!+@``AqzpvxRUCba**X7BsetDa}>gg^^0wXue~&i0w=3lW!3zP z0jSA`d$-SxQJrAYoiBxELT6aww=cr^1&~mF?2*BZbv~(B|G!;dk$-yk*T4ObND0k2 z5sDgTolgBQLEddj>THHP8S6ulSfERa!~&R8YY({RuSVL+WMBfs*h%!Pb%E+61v1mA za{^BhJpGvW$0GHI%kC_*tD$G{;=Zt+H22ckVn4eV*J1lexq*Lf5I+mh3btP*aj#8) zw!kP%sG*otitGega)^bD$`3g2mq!EV;{D)*z?X`l70BR>xW?-wlCxEg*5KS0IR3nQbj;09dDYjtsDITeuMyo`*uzol?3*Z0I4B=+1OA*ca!~})c2vji zlE%z0@@FMt^?6V4<;SfVb1-r*zo`w5va)ScyT2;BVA@~Irq0cQqi1Hik-MZDIqIn3 za!H>cn?@n&c;_w=RxGKpe^5I+8b}=MIw_irF)JLITx-vdAEgKV;#DVQ>K4S28z5uUao-~#DV(6I~_|DOc@!Y1jJ zz&zlON0RAHvYoI896&>7{td|5^Wr0e)2r{+P$%#q^wJk4c?(tpeWsq3_n;5f>7tK)Rq;%V02_sDG#yUd)S22p+n7E@_SI3- z)ixU+$boD9%>)hhx>kjidzUP`B}H;(hCXD8bT4Q9vAYM7Y4qiqhX^`3Lv)nH_Pc9FbC5*&DrD>5=085EVnB2~=UIVyhRdJ=Yp{|O1>6A|bbnILWvYWejX zy*N!iR79PYq)UcfS_P?rHKH?2`rIQfDIWWoyL2&U3%`+vSKSBh+$-KE2=2?L*M#F; zMidFW&WIi2BV-gNs{N8X%4!DJYqkKD4IpHMC*22gxL~oL6@;j7zq=%YYk-iGkJD-r zIyqYIikuiJIh38f!UbF0}WPmIav9J&^`G$uC~gTrJoE?A7IEly>Hiz$Epjf-ra z0T)B>$kviQ?3P3h9N4QfGDM{m1I|tX6*=HDpg_&0{l4cH=*p}6pt%ee$=%9U=~gI$ zL%LHquv(ZW+{()abioIWYOm`BPArqItBAwO2bjQ6xdhhQ5DJ7)5R$$1*=h-7z=6dO zd(ps{?y&{3p5LRjq3;;q0Ghw3&;FjQX9pSwUI{^m;+S>xPKrsQ$W|(H-LyLS`f0`7 z9>w(toPcF?bfA8_K==&j?_-!^G`{xgP{0%nH82=9T*&82z?xyKHW5%2xa7h$r zsqQI|mRR?SA>%N%DMH2uUn*Nn-j@H7EFP%kH?_f;tjxLNTH`mne|^|=#q*j3Pt2^3 z(mCmYz_~)VLQL(P%DKeBtL^qGut2@(jAAQ4cHDPl>YH;!E`}MV>ML)zlTFhIEIY@T z=2VK=Nx`oliF~fP!s9B9N=sKk#QX@a8rA}!r~(ufaRGHIWVWy8+*+*V+wCd7@F%c* ztS?+|uN|z6iQ~@FUZ1wSWtww*O^gYI?Z<#Tk79t;DT|8C^FHFaU(zGjM`R%o|E>2w zcot~wjKD_cwadpI`s>C44G&i~1(AEtoj;T1`$*G8iC zjpyb|AJC1wJHm7H337AM0NwoM{NCOfkSwnny)q7(V)qLPnkF)f*JcUwR7*%*O!!Y3qZj4`lNOU zvBHMAfwaSZR@k6^x$^FfR~;J|EyXBl0k!f1IH1VvRXykY+*Ypw3jbSK!Y|?DhRw}` z+0kQ#4SQ9~jq}^F!p78}jQ)AQ%rMm)|N1vBk*y9q)f5{|HG3!qJd9mbyTgLV4-z($)!KbD|Fnd?@{8G`a8WZ}^$Z%Bq`9PmF1JWNOFb~pQjg(XOyd6VM| zt8I6&0tr>}fA6d_)%n3undi;+MbwKNa#l z407Y6iTS(BnXoSDx>;o;MH(eM6E+mm4J11~it>49!ZPMw39R?*3%VgY;%pc6SP7VC zz|sA*EmZV(V+EM0!C!v!mT7~K!ve-k%TJfIa`r|D1!DUeI-@lKEui?+Lf@8Y7bI}C zvq!(IUC_d_BEz(^A@$0?lh|n_-^jY} zrkE6p?4Tlhfo>1gP_DV(4~G6K(3y0`LD5zGU>>y+)VX>Tr)Pn%HKt5ECPyxX{V+_9 z3&3g^95;60$N%O@Q(h)B^hbm82aA!=tTW>NoO%utyXGz(;*==1zIolHnx8K06>R2p z3!;FHq7ll4N_@H(SMzrP_eAfUVRA8`mD>*(IxE#E@6hXffT2QCKvzM4zs$E4lJIfj z7LOrPZ(B%N105s#KIVfh-(*(Maoiwg?z&xTOvAG;QcF56^vwlfN5T`i^`1{WkXu*1 z)%CGNw~-LCN_dLs<$19|wOf!wZwOPTbMnM#^hss6AR*+sGqhf~#q;naBmq<-7ZY|J z^(d<6Jd{5KO&A>+DYhWtA>7>@5;I?;Ean5%&oNTsj!DFZc^?ClDb}^%jA&hXlr5n@ z<}EPZwBnsG4}I6$f4^-iH)ck`tB8A1G`j4d3guBSw6&kq36c~aefPe7d)>$Vb1i(( z<6djeo#r3(ICKP#`;Ll!@QZ(Y)l2fDgahs<6n-NZq?FyCc(jmS=?#d~-Uup_c19$V z4ItfK7GCL%tPS0C9q31u$#42jhL>w0cqjbUvjsXU@KUo5KCF1vqi#rA+!>LrN~85< z)m@5*q?vxNi^=64@l5lLb47LC28FNL;I(53_HK7<5LBc5hZd5)*ZD&nWFFNginfYY zMVy?YMh`z*RYT)jY^M+t*-C09V!l#=@(1ck-;~^g48z zVuxwXTYEsq-M&yT85uUnFuByOAD3|M&H9pILMoicxku{Q72rFrt4=N%O<%1PbBZFz zsmL^OnpmUU?|Z=iazG;wS+6xpY-HceM0+3dZJO0IO9P5heT(Bb1EIN`V%}DNy@F7h zSGMpd-4%dgrzFJ#exJN547(Mtf<6b-ih9QJk`!rPct)EHKh;jxxOXVqMY!_b>srsj zJ8J1vuPFbc;C!cWfL}o}7_!x8kQ>Xaw1zB~4WulZv&KJ(HT;DCULo)^K&rIAWHr&g zHjrvEf>b5NfMik$6`4f?cZN$;_$A47uAg*B#`J;wJb#%%bXQ#T%yK! zwH6eV)S0d&oNMmMOeaSt^n#QyQNk4IpaM_RC@bBI!nF`a*uskk1*IXyZJ)UC6i^?q zoxQR#a#$n8+y%k&K0YQjAV+jh^q(Z&fxY8>Mp#IvnB5dfp(2mZsPRn!s*AD^6vxej z)~^mFFqW=B~0yi=!svHCl& zI`2=ms+dz46whmu;Ti~w;Hv0VLd-PBaSo}}7yNe2-{jh)%HiH52$BY0*oi6+wdC3E zQ-Gwxpr=lFDsuCv%jErZ8t}8^d0h=qqtpW?E^mMyVxFX$d3IGb9~B<1imod@+ow@t zmR>56&lz1MTaZt6! zznetO#@3iRI!4&B_?!q^?2i!K+otST{8aZD3vf_cr%ry(t~v4v&1N<;V=vz5wsy*2 z)rQ&ZMCe=|8DQrJIWSh@jR?(FPi3A`>C}vvng3Y|E!gIo6;a<82 zgk{!FgXl8TOn(yG$gAd~bVO%FuHHoP(|L`;R%djckzTbRq=$2n54KO*8&tq+oO6f1 zwYW1P%KzHp)1(==%J6J-t3Tupa$~09z4y|eecRe;9lB<@a?gF@^=@VKsME;8v!aWN zQK#*58$Cu917t0<)AiC1$$ru*|J9A3rcu|vT>sq%%c{OsyP{&*N6Q<(apxPq(kRi4 zf<>E_CN7j@i0gShax~GXfEgRO9{!fkE=f{2ta;1#Ign+#@^{#R95ZWqR1i;9miohl zvMr!V5;H&A9~EWl{JOlc@B@WtZ08%l_#VHohmFB`;bJ@PVCPAF|2G9y^9|<7Nvx_Q z*{_Xx`q0QV-%l~XN>&7|Z`>A=AU-PZ79ee1i{N8m7rG}n!L-3N))|4~M2TJn(xH&T z8D%Q$tW8p!gasS=MiAeI{m_FeXr}f1K!^{nY$1=`a==Av0e{Ta5M+b0@mKDKy*5VN zwucM6|FNhKE2WKda50CAa%fk<9K?VmDKJTK*S%GMMaSiygP=^lp4#i$4sdCeUkzxL z>)M>qOvP~?LN)Rg2~HkEBwD7~+n z9>wjsqr5E;dTe*w!cX)n;SNH21luxKGi!u#oE#7a+{ZgEwH{fv2h(#@;fB@7s!;sl z4QIm?*8GLK{~^)rrZC4PxgdLtrm#ecNuWp^6?vVTC&YjeR3(e9(YO3txgW|kMqKyJIKMOvSj^N}L`!#q z#+H`O7j|;e=(`ZwiG~%`Q>PfeV&7a<%8Yp4QF;XE7S-8wj330Jzu2+t*0L;BuH>oH zYT-KJEeHYGB`RqtWQ^>?3m>dThS>|RzW*;`H-jblUj6QKWSav=o=c2uLng(fQ)D+X zJKXZ$FDZ135+ZG8o;R{e;nXtf)Z%eTI_iWJi1~C0oukU&VpIQBMWr;I_vx~S#Qv7U z%5gV4UbZ^GYRN?XYsO#wOuJjy7#0VfZ_M<<^ns26DwOv_?ePc@w5l=Vi4osy*a_?L zLr~DO$yGflFP{hcC~dNI=e8jGqshwp9b19M{C&Q@J$H)SU`Enr|Hp4gk^`HOG9xpR zO)=ner0Lmsv1z|{2_`JJ2k!U9m2sVP?UFlRC=-=Hwz`_x9g2aMG7!2r2$XfZX2dQ; zQS=OMzul?UtTZGOa5Bx0_Shq=49Qf^cYj**s+SRGt5Su%5c`0pgre{+rYxid$Wwvx ze~(u;9Tkj4eWSrzNJ>FjU#)1LE|AhC#i9ti+XrhgGIl)evCRr2Q$%0CSs*mnlHZk0 z{}b8FZgJzlUTmI`<=I6s$(ZYi9F%wZ)w=i7dqKZsnBe@flT+mjK4}n6KG^xcnTrWg zU{m$6JBXeEfGwDwU1#r~tv~8q`l_qG{vGP^qJ$^8^_*-P6>JKYp*BR7s!`PuG(tDp zgmjz5I9NW^EN*IpQ(4&!>R%K-%U^Y3hf?=mx(?`}qJ%>s8YRle?A5VfHiZljHO8|t zL(5cZjj~}LM&}0=UDAsp(9DAB8oS#C%RyqiY_$OnUU~~wAenmac>ZCfVRDH$@zcX3 zgWcrfz$=+rqsgU&VxUkfpNiZ`rGKe?#pbt?zjA*W`YRe`^TO6Z^({B7-c(<4?^87| zEDLE|1WqiCWiuM3#&yK?`GFqAo;m$XpE?~@4G>&J$8oN^#4;VAtz~n>S&J}ZyxHF4 ztxQY4Y)Z)Nu&HD-Owsl9uxP|(3p8qC{4|>Bl!Lnc|P|RkEY@j0BfL8hT;)?-Y;Io|L z0Lgfs(B@Xak|DLkQdn{yTc7i@jP|V|21I|Ft@fKMq$6sRmRD%bbCC6Ip6ZCMT-Pt+%LjY(KKsqspNe*PrIezFZJp44#x^-JN> zF?;ZYMCflqS;tjE2)h-#VN>08Zh`dvoHEaqn>Ym=IKL*KhKm;`O}DCrKsWG~pvZJ? zVfWcOd=(A-=o3?I2#1|-%&0vS%wO-0t&mM5ebZwQbb;*=e_#Yv+W)_35n>(+|8w4T5Mcg=XwntBJUBDq_ zG1SqvaEl0@Tg^Z06Ade(8o`a=CQq%jRQj<3_iq8&VZ7E>&dmf+VvLz*pKJsd^TsFX zM{Ac&q{Z5SJ(>MR$kI|w7DY0s$f)2FP8U!j#)ri&{)kv>vtJid|9i?X-oe-L(0*wI3^+knawKOb@&ohI-Jf~fNWH6IwhFD_5<`^({!{2e<)=ng@ z;=q2cnYhY(-Tc9jRa)km4wRquoEiZVRpkr!Fu(;9hT)Jj zpL_IbXhg#{Eo6l0k>@Hi#3#V1iRRS^1|)cw_mE2d282*LBH99KIA8%11OdkqMP^ORiwDf%!?OJ&D`tN~U|16^KeyvUvXR~5*@02C*T}C) zp_m;MNu(m1Y3!wLl*BSf0i;n5fM^jkWQ1K-9P&j&^FS5PJ4WK{Ob%o;I_$t-fK zx@r<4)R!4^arN? z^`c9#!G8Sgqa}r;WEyEPdhzNh1`5>=QIY3XTv@T}?W@b4{p^frBo9f|m;R8wJn4HE zmUTuX{ae<{|MPy#iq430oB_IU z$syluZ@qsPPApuK;d_7i`+q^vx{>eNoB^SbG7fq@@Gg^O!!!cZqEIzfg+-2rS+$@7}&OPp-gZg=GCIObd! zw6a4P%itxbhu5h*=G;KmlcebrbVZ&4&i1H&cIS!xWQFP}t2bmM{L%m(zhBLqNv=CE zc%qG{5{4*dkRo@fNUV=-lj&>eHOgAQM6YYI6!2V;CuqdwGPo*wTFk!1Z8F@4Jvl4e zWEkJ4#gd4_Hd&XnOg=1KiQUVHxJ;^A*(U2ztR{CLJFwZU zY(|oz98d`9!Z==RXq&7i04WvK7l5s*O@>o+k79#tV1;T6j!j2Kz6PblUlMXpnk$$qVZ0eb!0mnAL2#~wYTSanJGyZt1QyO9h_ zAgv%v^+JkPY#IVx9N_*L@Z2H($o=N*jtK2f_x`B)OIN>m1}Hanh&v+&L1`M4ss(3c zPk0TaH1PV8|LP9X?|_7Ocv;B(kWVC?5g^S3>I$ubb)fhOJ=9LueAB%~oifN{30}B< z+Fn=v#T^k>A*y_n?Bz7mBMWzO(&n9s=!__bFDYKSktF*yf?vOvGwOtY!Cu#+zDGQV z%<($f-uAt8jvwdX_%_l!1pm2trsuB>a9dH>G>iP*f$iLSBRe-jF%KznpNcFD%9j+$ z<9SeKwJiYYeUlWI=H+?sqINBB{%Y$om<3x!+X6=ApGsrG6TP18&RE{M4ANFh8+mFx zj3s)*iw9gl;p5x4md7lw_};*>t6zKc_Pu3WmS6m0&WgHkHvKsM+wT`a595{fWzx_~IeoY6jZ%lN4j`3yt|`*`Dnx;-2&`^Te`pB+M${bc>oGGiLlfr9Xy? z#{2T8PI&h!I@deXRlA^Fl*!f7uZgXDQr{Ln+W@|A=KuVcWY;ueMxVZfiG$=Jiubk( z;y67d$9L_tG^pY4Upx@p8F5gweO3>E$KGszV&5H@mjUVPOi`ue|kN89W%|N4)UQE*&`zW6HVt!2;P()9gea+qCE$$={} z&KpgdCn@F_MUGIB*wh4J4*IfWw=a|cVrNr>Z?3QpfdAC#1j(42E5wXD?vL{CorC?j zw|%gf9gl<0rhu%+jB+4~jrLbJh3xUVO0=H3gh@3Y@*+CPx;*a^IIBl-O8J0(uqXlY zt!)9-bDlb#g!Kb9+i8?L=C9_J>b}vs7iXex#W=H^=`O2YlS-$b@9{Rk={H48jkmf|b#$taIKx$O;sJV_PygaUMJ$sZ*d@i%^p)1ZVB9gXEjGse z3afUSx0t_gYSyeYpr*|8w4C%gu)`E%HMJEKHH}Ei_`is}Eiq%8Zh`*`5~&cjDg@2N!k}LHK03|! zhNO&c1CrV{!Nbr&ZnS&^0?_M zYUuO+g_9t+JZ4DPKx52hcHd+L4XWa6s!KBsQ`8L8vXhJ_V2KD*5BTdc4_FM`Hwf43N61SNokZ#FnLorK2uY%}ZM>`L3d{Q4plq6Sjeq<`c(>3)sMm;vwh! zrO;te4{xSXY9&f|AxM8Fp3_Kb1o6D#kQVV4$g|%M>6N3v%Ez2`+b^#z1JpQ^kI61u zBW=-1&X)rXu#x{+y_*zGBOvNG=Jl+ln8OsQ0xmJdVNvzm`-@V&@_sb%_K9T)ycX3F zvdVb?iVu#`eF8i-?9vvnm#!s;MR$Vsy2dl-xvkuepi{~r(iu@BsFgNxj{27R@0i~Q zRpk0J@VG{a*S5-x;@A5c-EysO5eE zf8?G`87$Bf*C(gRsFTCiX_H+>rZJ9UHc@0f6^T7n6+kX>&jr^+y^}|!Xl#f+e~69G zpR7-;EZvm(Cr_qv41oB}v7w`6&3Nvu12^Sm7=dIP#l%x2mWnJBbxG62>LJk(h_A1T zzmZm?J2TeR7Bhq;V{I{To_R}_+|&l=XuS^p)Wu*ls=o4e zJK6NwEPry1%tjo@qF2(IT;2VIfe zQ+%k>hD{bYSd+-xVBjsf#o*t<^Gp}V4x7(0v)s-7dYl*%|8#U<(xMR;bvJ#>EfdPu zAkK;sw8Ji2JnFzuWJ-ECUD6(fK5~NBv~kWry<7DhxJK90yC7zMDLAf=XsC``P|xXt z5~xQuS0-U?CZ3s#7k^qbz?Pn5etsEe%{8^RGL*dPyuZ!MFH0&H_DYk{m7X*-%*rlFzQU*A^>~P$Bc$ z{{E`7hL^Z^3$RhCjt1S4tEAH{i>JnZr4sjsc^YM%s>${Nv6l@V8(_TbKDIu_3K&xt z6z>T$UB|p8fe|w+oI|Qlq0wWFxZbmxkCd9~BcxVdrqa>_y2~2BZ&@Ng#;q6V6axdVp&~n!?V^4fX&ydu%kwUvP}OEN8R4R` zjT&<|khtRjI|1}2Rr9fqwoV=mx#LsH$GoNxdjrAZfj4zEK4;)2dYGN@_|L&l)7%X( z>3FwdB}rzto^jyt2()^S0aY%=0MTnEf~rHQJ{!<0MKqQA)P$-_I7ghdbd*qE;5Ool zWLHSmsYX_sB5pd^nxWtXaWCB=yGEymKMWlTKF_gZEX8VM8IU$n1ev=&Hu+%9t6rmD zCb|IYmde>1IXYrQteR4TZV6P(M+bJp`UP!Dk^)r(@DMC}KBRkQA98c(SO)s9v1OZ`rf>8Z6`k`y4Yc7z24hnQ(dM({5z0N-lrCf31gEM;}_EPG54Hi#( zU4a@23mmYdAAbdXnM0YVTB0uXZ5ItAQFX>#y$C{?sA1mgQKLV#*5@aiOxtN+6Box! z^Q~T@D@lawuvMrw5{J^;4${q+1Zn)MTyRBx*lMOVCF0}K9Ti^Tf%~xH* zTHk+w<+d6nuB^b1#lU}(Vzsag!oB+;#i&M}*dE~AhzdWvASxXEiS&?dp_70DHsfyW zHQ0CHSuNG4T>fa+`H7?+9eAm3CiU1R>lQqt$b!aXWNb@P#CQ}2y}w$RD*>@LVeQN# z&IQZ1aM$aouIyB~Z3F`Auw--#yL<>b@}$&MSLMZ~5C^VKFr%Il_ts^}zGc@GWy>DO zqx|cG)(Ew}b@C^?YW@Z=Fj-3nLyjm8yM4NFJGUPalTUaf+(B+<#7^lG?*qy@&nGr1 zX?w|lSqBF+GB3HQ4Gx(%Er0&*mlI7ZabFWZ#7sePSx9;y66%#gb{eLoHeliCg$9u% z#Sx%zLG{t3=|JnEqfJN+M@f~TP>pM@WHm4>z#NstZFR=0Yv3myNDZ&&Y>>tCbUjKE zb)7VgPV?W)ujXNh24!5bK;+aW-mdl7GM>bYKTRB5wITeKrd_LMnnTeWz_{5hkU#@o zlrmAJs!h-XisM~UU06&{zGIW8W#Qk&@}XLI*NLBI{)HW9t?9T+#lR95^{bwx#sw|D z)DUx9Ej*%o`Aaw(^^zC{8kfmjwy;9%#DQq>9atfD%1u$^F_9tBv?2A%zmwS4Cef5{ zwEW#oF)0+;0o0r%MLMWZAD2dj<7!u(;W7xi7$!_E;XND7se)>oRk(zjWJ9^{(X%;=qaj&%RPFAm` z<4z{OnEr1kO}mvHww%&TXEIVTtl{(kLj!u9tAwrcZZar;usC0MiF41Tc5Y`xH@(-j zUU$Z663@fLAJ)x10trZDc5bHkFKzPMztoO1jzuUob|S}ovLBLJp_r=w%XfcdT13fa zUE{z@V;~?L6JOm!F<=;XQIW{Zw%U2q{7?KdTnd>KV2{(%L!`+sE?|V)4C&`G-|G>{ zs^qyHb8u0!C*mWQJq&K!5lYX%U>&5MU2VUgP8-vLOfAZbQ4$A^l$o(IZ8G##uPuJM z_%6xy#v%t;=;Epss|T<)DjP_^kq{$S(ktzTQ?=L;h%8Tco%=(9@0SJwOJs@fP}&O} z)=`Gl`|+m!!+jYV~f`a7(8s#-8WvQ0nIvxEidtu@w zFS=!LdC3jzbf_shuH{wF)$~H;5awvQqQ(?KI$q+nP5MnKD?+&{lHj@%~II9loY^z|Rxg>4_I-DyKESBBO99Nx{4!abDXq5FHPdw~9zQ%*Z zYP5|nwd*HVQx5h0rXQx5wgEXTHlV*_%!>9ob%8EPggoikRvOPk#y+fNt`S|MCnL9S zG@2Hp!ssgde_(}*sb3D8-sfxxyElKK?te%$yRf?huVMBWVJMMe5-1W!MFP8@EDl;W z4BUTMezu9TPoyW3oNN?X3?e(11DL{6@fV=;Gr*$MDg6{#=fK@zW>{}iq%|U3ps4eG z_XQd9au{vR6W(=3O@wxl-4`oHkJa+A26xivO8b6bXE6HD%>0emfQ$^m_Nye$fss*Q zWCAo4lS+}DR3ug%JmMyTB0Un^>qY8Odqw?-+bfL<#vM9s!;7MBL6V|jPXFZ0F026Z zY*}kxZ0PRJ3LsPdwA8QYRnPE04}Rzqi0Z45XwmCr0_hDzLhkV=)Bj{eXEF(Rc>Wcv zfH5`4drF$AtKqN?s5$2$G)0=OTurjMmnFZeUw#4Bn*Vj*Q^)MhAw{ zULz=_P|OaBBvO&r7N4Dkvfb)R#V%Q^^IckvpYnyJ?k7VVTr0W5(wcxuV6UIVMJ`)# zSToSpT*C?+Q}_Jah61hu98Nw?t4WLlgCo-j97zp zLDFZ9M%Vgl*x^EW@n_%q4+Ac4hHaToZn6uNI~sw0y^H!Bu-rFj**imi%U45ppq+Dxmp-c4!!}?PB%TcI7xwI5Y_x{5;YsL za`Jn;T*)q}8fytO%GAKC0cE}wz9Zp@T)oJsURVR=LRS_65k!)%-~gKVuqimv3-609 zG};9I{7(|P}V&#BOqQxUMFrO}=4-0ntZk(P5%ajI40X_3f>@IF?JkV^QG?|43 zA8h>B{r_0tC=9yD*YOzU3r{eq;WuPg17OXWq=@ER^H z#12lzS2q2iC-`u1x1h#1g~_JT7r8HJ5p+hJQXV6{-`TYskK*1TpV;}(<|(*JD#MPD zX1_$QL*9=i*)%SWj}W|%?WQKSco@$e8MDvW5>4O}t9ggYTG{0>5wZXWUgnq~3qbDS zLefKOME#;`i<{{}z7}d2v*;XdlA^=yCh3-rI^6`iBmDQ0x9#xq@*!o7rkCH&W=Ai1 z{BqSqs5TuKA7(_lihu&^uc*LuD72AVBg%5A5pD6pXlrj!8}PWIrfILdk$X6-g8z8Z zisu~_9-IC&yLPEz^6B_tbT+AS;K}El(d2WCVvbPcLn<;`)ybsM-Q<*VM4_(dWka&J z6PCgyoGwOX0?~rHQkJTfCzD6TLLB*u~WH2%WXG9;y>`7oFr> zg&2M&H^%R(qLb4PDPnc~tP3J6wW;&#gD5}pNhc|?RlD8mJ)d}BKXHR^^#o+GS)*gj zykpU(mF9%iR5W!~zz*89s_ivd0-39Ug91p6vRqlIxb3rhZZQ|KKe&7@=J%=W5kqt9PP~Qmd zBRz^;f-_;ddKIi@=@yK7pHyCwRE1S?VFjdJutS_D?olRsfl&djBiC(i7Z$(<&UsYGEVtaSZ~T3ftlkY`U#L#dNB7dtr>t9UXk6^9(lh~(OcVz%y`B|Frf&Ik%GQab z{~S2@W+wfoYyU%Bp^jwzJals=f5yQPvM&j6az{*u*1``Um8if+fW5m!|$NAWO#>#J@ z4t!Jb_HqMig5UqBn^aF5C+qCMA-FChB%Pv|;}mJ8B2&Fe-I1FhRna@ASI{dR;>L_9*3C0vZ#PwL-~@>k%1qy9FBpV?*`(bG>gaYK9t1Jk}Y3@sK!{s-D$x3tC>vqS%@`l(oL>9rKD~eU0vH7kk5bi77hRB-^S)=m>>ReY{%4< zjdOk*U@$Qok_L)Np#x7$CydNVHN_mHNF~%dD(ZO$=^=Rai$m}EYi8V(*Yl!;{q%il zBlq&G4^<6*Hy1Sm6W>7SV@YR35w{M^MjDN@uyFyyZk-Xx{u%4F#=lz-7qBmAYv8Cn znUe$31Gv9QzF$(iWLIG7tQufYD&;iTj2+n`{FpZmFCMWGl33w~`hrGSYHCtmlc25{ z19Vh^74SN`j5y9+Iz6y5tTpg|4bW%k6E>`z8)PskBIjk9q?p|*$$_ISr;Q9y9mP~r z0a^Tl5(zeOqyU5LU`QdP7;*A~-i>aE<@M9?3!tMT zoij3BT?<7KdZUCAo=7uWz^V276joOFxkoX`&ETp#ls7nAWYMzC{DhE1h$naOCmIE| zxea=5l|F-mHCvbNT=)6MDu}wyT}=iB8=(;`R|(aY=+ zJ)?r_=vHWSMmE|KfDxXt-oEOPY?1QZO2CFS2(a{;OK(o;HswZmu@}sN=NmJ82^I1^ zj4r~j&nIhC2j$1efaL!NC&aRmw{*O0cb<7C_V5pX{C`6Y-pLNFJe^cH@Penq$lo|Z zF&|Q-j*9%4q(PrXRaj>PzUogrnm9KkZ5|pWGS=U9PY68|b`C0;x&?L8Q8_AMB+q-g zc!awT7D#u4I|4c*dIje=ot*0s)?dS0CEN%Rlv4NO^4ht(oXbW1U^3FhpWWq(WQVwj z+qfu$Tg1itf{OJ-nK>*2%~&I0>Gs3`WF9o*temN&Gh46v7=YBgVrv!2dTrd?1|!&& zQ%ng(_EC|YZ*G7FdpLY`a1Wjs+}GtH-b)( z+UZ(PEq#>~$y?pj*bb1bLT9)l6mN6H6C?x+Nef*l&2(L<1&k%xw6DabAe$JjjIJI~ z?ar1yTOHjK{U?c^2G4eE@+F;Oc2gvUid-!`DgXN-Z~*bCp|F)k*}SkdP**#Gb)x4m5U6!Yp9jvo(4QcKN{S1$JCF@XPvzMwdTxGbs|=I6`Jr0=C%!jG6m`Q63~kn-<^|JlbL6l>jpHU;_D#!v<4uEMdFN(( zDmm!Dj!mbL#XC+h%@k<_i^uB;9^xc=p<9!=g%R}jQVftH@1Y_GWGT`_FQB}J>dhp@F=q(VRd_rOFXjV*%YC5IZw`r> z4kcvJqcT}DbiVXhn#g$?BTWFN;UEnlH--~5Bl=Yli_g-n2^qwn=;$WMA z%8acVZ?{<#nHI2vC!*+!HCI%0##tNE1&q8(=a9Mjs~z6r)?& z^uxRD+%3N`0OhxBKU+XPVFwfko~Sk&QH?#Om`^D(1gmi6paQAFu#Sa#P zNq|OL2>CYoT@AP)>ro(uLzTF9PLd*?l=|PFQ|77T1MUE2%Rw$4MCD&K295{0<(~KF zX_TM%jt1w6LAcI;$R~l9H@iip%Vyy>=@;&>1$^(lbRK^_h5uDaZYhv2Pv0DsKs3r; zSncP60PE)PGXZrx?E>{0|8zaImgNgQtCs3qyaQ3=_u4_Q1 zd;L<4QpYck0RcP%|I%3I|C<>6#AwCu#<=l`Vy>!g9?cV7Ujn{yC4)-Q$;>IwzeUDcV z(oV0Ew)n1{mPWVApKz0v(9tlhj=s6*n&Js}?X-U+hv|#dE}^>tGac_%tR%^=4Q48h zFq2C$ITXpHA`koQlb!=-_c5G+yq>Z>dV|BMGhTL)St?;={#{f5BMSDXWG+u zu&fEhoIy5zAU6HPYEA68>y*ds*YBEAUcM$ZSZ2sDk&Fyn zykoh1uaE6fVlfWP**X1wnTh zAp>MJGs@fIRZH*|RsWB@cY$jvJ@dys;t9!%Avc1_8E^z52;#^sRK&r#Y^UvP@4NeN zcemYcx6*FiZrkbpv(wJ>f{u6xLBR`Xf^ri?l$(ljRk?Tv2L(k`E)kVsR8VA4;rBdA zR1y=(frQ5Cj-Qc}a|@pLd!F}spUd~V@CFbtU-wuhxX(ZDjxlUppF@jNhzhGUu(KB{ zHf%J_+6Zi;p=cA}eg2Q~Ur)rfb>P*w1=qHlj-9h9s9#)tvV*VH=w?;=>1Cvh|) zv>h26@ufcrICM1PuY(B29T^rE_kH+SWE)r2(VLH;=pO|PmO$#8j68Rt4b<5IYi5E zd+VBbw;TGcpS^YB{dnl0#zrNLPrBQ{&)XK`=S)Gm+eq|v*zS({!;-&zD#DZRzk5}8 zP=9jeTX+~xEq$w9cj8EJkub@-aq)-8-um_PTh`GJ7k^ldd+?@pbl2h!&o2J;IGtkl zI2!GDShGDou(89D_>Zt`@36A*pa(d_OE#0N{ki?_OGeP_4iv+4-QW5>>ni3pMn+^JnyP zd+7pM*}|=oo?sPTes<<#@vu{yaCKOpILW8NzYj75ct2GGA4Hd2vn7v$S1>KIYI;zp zLarMPjkM#eQD}esjc;k@{pMN8>jp%9YtPSqNw$A!5Cwggqt;S+6q7@dEGoJ?xKXO1 zFA|9Uhn$CMV_+-73>AiYkz((@>l3+dY1Jwja@{-Q7`F(V<=8nVC8}&mJZC3&^V~cN zz2_##ooj(ak2MmkvJ*Cdi&fvl_N4iH(?2t%$J=G}oSzA^W3y^IaD`QKP)O(tpEjSG zfWC+vc!a%^FYsyY^JM!xCVD1QB;AkNb<>}mR=0SDGWA_ z`VSIp&DF~5phu#DUg?(RqB=FJ%CA)d>J}Kd=@s^e$8(^w4%j*$yC(VQa@^fI-7Bc8 z5G-wiRl8bI#Y^&T;iwuFYPdZdB1h~`uS_;l)Idn^;smgXr2|Hn$@MaLsT(g zvAlk6o=$S7g~^3NZ1`o&+AnKg7i;{Y@t$Mu)%tPZu75J^m2%W~f>`K_s*`zHve@nWE#43q1j&Y@dv5+`+n>mqX(Zd| zrQJp`0Fq5q^pI=4Y=?I*guSxo4!J>v3A`QE9OU0_1v$EPSfvE7>|_pwRYUOf4YKz6 z+2A}@L*TgeAxr=G@5f9D^c>b+SfIrFb0-uX49%;e3xHq~mog3fyB=!pJ-S?UkFKKA zxz(IZ8W%Q8uscr8J<981mb64wM|E{&j+#6A%7ffO&gFnkk4GF^Bb^nRpQ75_ zPa7b3^exz-dDnLbPJY8+Py%A-{ePs^frB+&Mwad(#X$6?fr{Ske_UKoYPl!rgf~ZT zw?x&^Pr|a^On5WPf76>al6pl;#A=W=AEN6OiPE$;bA8ncU6h8-@deazOT>MEv26kO zJ@<+m-7^*~7hDzJ^6UoZqh@BT=M&EqAl|#exv`*~Y*QXzoW&jDwnTl(RJe8Ut35Sh za8c-f@j+3n?z{~?UwC%X>2BHd8n#VzHUpM9uo>+e4V$sI*#RqNWgpix+x=pvw*&ja z7Bb+~i}Hm}|6hu|el;_&ezVy#j#-+ato4cq<=0%n07s?Y#F_Tu%pT9Bxo;tvM5eNV zE}ylYi&JlgQlr~W^10`V*@&e}@xJDL)Uij$+!?k-lDTgf-#*Jo`Zz)SM^pXk7g6KD zIc+E%9~FmrrfD4`02L)KAR})K8kS{>R7pN@%n@ZgrxoacOBdz~Z9O@?d|X)I;^jB8 z@lngT7|uBT{@+$XRayuZRITG|@+a_gp#`d0>IK_!c>1LMhv>$HX{Rg?leL3yZn4yN4>U!&)VxGax zoxAOO^88~x%!uvaoyu+a}ibpCJ3 zK>+^XcZGlPG63TVb7~DyJMj9a%xI|dDFy`Bc2m)x2dfpyaum%PmZ7p*iR>=jD8;6o zcoAlMVghr;*sX@!WxjVMRlcb}D)We51#9LFK23@y#cG!Z*J@t*ER-1@2)7e5CTdCS z#Z1M{l6>*2zED>KU`oIG{&}*_fdK-&k`MG?qHu?tk6M~3qDAB)x#;9rnkA^ zdUVhgi}+P~9VNA5Sa!%659NyXh(T%qgkYg{r3DO6j`$=GayEF$DYC{Fc!hd}nDN02 z*31|P$NyIYe`W_=A=OIiK;M7;u$~sOXx264EcvShSg|b}M(;=v5^srOt3msjGr6*%3}Xrxe(t zce~|syW}xc1zjo37wReIF#(UYd>DaS&+O-@l{sRa@C7$7{0fMD{ZW7NjRi4+25G!4 z!@{;jnzc9@`@Aj9U%zRZ7+}+l7zFVZ3nT)wX)P1*ouNu1!%WUD^h?I z7}AlX)D@7WSmArbACI3=9Hn!VEm27ew)<>?vMXe7Jga;x-Xo3?G>IQWbT@xSEV09M zXdBRs4Fa}jB)CY*O=XH5c5!V%cv#=Ajyj?T7FK&zdZX;US`p7_aJ{jhA!sCc#3@<6 zle;G33rU~rkU@r zh_jXUi}R=+-? z&rNq=`*xF}mqvokR(V?ZrU?5Dy17Ts5(wsQVV8ZDfnX*&H>D$;c^G1FZe@Q^?I8Oc zcujrA2pp9Z15ue$DmsBzuSnz7(fd7Gq7DVtE6~{()K_3HE)(W(YN7hAOc)zb<6rF6 z5|u}H@Noscoa&VS`2i3t=r7s^>VCa6UY6qb$y*O2)QbM_Bfh6*qq<+0ygO8_Y(UnHg#z-XNprgJyAH9~vKhpm!6k9ATTxwf{4lP$pX zGMP8^&wKWnrUo2#X0j0M@AMdQyB>lnN*5LmgvSU*oGvUpAj8W2rQM;(T4u|tg0TmL z4G6}*o9)hEWmBfE`Prd(D{C#0Hm56W#8=}|O0V+M^0Vj+`XIWrl?DA zUjv`ye=og#6@HZQJLMOmlI6>%eCd|ckQVhBF0worf zMI7*4AyvbA%cfMm6U3M?>THIvDN{CHs9I?nU15V7$F-cG&NkZ9eb%YC|Fu<0{J$lt z3(QRi9UoNUTSp&VbcHkQg!`{R6E)j$!41iko1!)6+bx1rg9j@ecLE>8j zg}e>2E4Ga38;|Kk;bMF@+kN_dUB#>)8ex8GK0}&rLDjHnE`BjUBc#&>kb^P88(8?U5e?Y$Zg=5T2Kjd&k_Fu?s>Y4quLhG zz>nvY1*`;8GDv0jxF7W|M?gV(00?{&XB2xa7wquBfvSVW^P#{-#UB4WiE6$3%8&*a zZ{&|D3U5^8BMCiBUJwMzR6e4CLOdo?t+?u&s<`Kx>!Xz{7o>@FoWGdNzUY36xy;=Z z)De6>sFjIh@SEVzQeTmUHJz2-LvE@n z}T5?^7Q#lfOQ3@jY2qSFsuf$8E)E$rXFN z>NcZQg2`*FZGj?Q{Lw1G71cBA&T3pIbh33|ozPc|>tc3b-Er;E>wjl(T^5(}yGb=W zKiP574m1s;msO1v1IpN^^a%tl{Xh&1E>NMD%q>?WhCfxQP~oy#9a$Jy?y_%2BVU&; z=y6AYfgjQY0@*t}u;~En^}6JEDeOK)c6x(b6Q;kHT7eQP+4KNW>9qv#Zn}?1<*>o$ zI8T)45eOoJTEUQj==0WC~7#fGqctq(8OGh_ah4|o?g7TVX6UBD^>;8 zfWzjn)FlLD3`eDzHc-rZimaicuW*WeJ3ZD%4!L0tp(02l`YV_ic;;4&F^iG0-Kngs z%#@G*5c8d18xV2mCvH;G4~HAb2yk0=Hf+z<3rWXKQJCOjMg$~SF79WToP&;8sh z9PAQMq0~s1{HC;CTm!3M%rfd%w$QVNOLLdpIjD`@0s`rUH#!$Xah|71+ry=ns8W_x07Uf734VaR$9rn*^N?6sPE#O7iM~3 zvBgSROkjg^J-r2#c+`p>UM6HkF?X&BFBSC)N1U#U^XR7w^AyQ`sPx0ycjTaSqG)m(|L$t$6ny^CO)H`?nQp%L1QCk30H8$k3@@3j>uI zl0+aFt7sKUu_NbDmi2|<6E+W4SgX8h_9s1YzObB)GHf=PI|(Q2Ht>Dz+Fro z-1Wq3>pX2ZE3yR7Xds!{wiX&=pJy37T@5ogni>{1ISB`Lo-LT10!}Lv6E^6ILV-}w zNe=-}LxOXwaL5(nc>1E7$_4Da*s%xTrDw21%}Z}((}UALjgK?!inGvshL!czUZ_fT zT-+W$>{Kc1q{|eihl{zz*>VqbcSH#T=qY2^F15@9;>LWB(bJK&UcYj|m!rdVyPh z$PKv)b;WtPqIz)}2iMFhq;|n>>k6-PeCX7#@NW~U)S#QEmRAyRR&o?u2~yIRE-dy{ zb-SQK?T5JV&QxZ4?2^F!(?gP^_a{n9&l1*V<`v7k#(1M%dYBcgr<{76)Ay?Li64Z5 znth8qi@8D;!ZRKacpPX$!(DwJ@^KmhjZn0F8H&sB~eg3J<0FA#iu#~v`tQklb1NeU zc42dkrmgK1vkm=$Xq2k1CmZIfD*Sgb*#3%D3h)tIRnlFS>Uqd?kzcG9PV$))h_HnR z&O#F$d8;nB438ae{UzyF26)_BxM2>t<-i`yDx;9z7ZfuM+LSH+p!6TT|+ zYqXNM0F}nCkH%VpB3w=3DnTX#WewH<%m-ToQz}?5RVof4LB3nOUDS-yYF3)FD!RK z+4Wq}$r*Z~1SHj~pb^T;I62~8`YvrC$r=yNEPFh*If6j6j59NC{gg!=Kf9$gVDDt~ zpEr@hX~23pI*xyoVk#+8rjsp!%97pmX79o9Qv?g3mX!E!ox8!O+N(mi1P6J56hVBW z(7zLgVZc9r%aO9NuCDG9iuH%CdGhahDJa6#026~2K5@K<&FzU3(2EUjIWCB35*xr2cmwgv(sty zwaTk`-LeE{4ECd-8|Wdkdb%(EyT;fvf#DeA&J#P#Za5q^?W|bx-e*n*oaBA+%a2LS zm!|AOZ3OoOiixAhS}I!Y(;(3c=RgtY0BK=RRL8y;Hjgx`kZt}_yY9{W$&+80a(5mu zdX~>9GA{>ogWi0s@+Q=&s6CqL2cgHB{Uoe`pT)gJ4g%W` z5H>(0c}SjU=cTVz&)AG%*6p(6Q~r`zj}=ZgT8Y7W6gs7yA*-g5Oe3glp_n9!BvR2> zL#{~X;P%SP3Bw33K@1pg6XYUA;WQ-Ln^obXR^eA4BU7RXuMW(LzsM!kqLl@TT>ZQNvE>zo`a=t%u~cD>mF24%*c~jt}m7+}j7} z`nx6{*-$Dqn39u!|IMppqXVyx_Z!)`T@<6C$aX3^P0=g7>U*4fT-i3W*zG=a&L;A% z`VNDl&UG=Yups5sFTcib6W)^b($_Q3ofrV$qy3-HS_rnw2DtyrQvL%|v2b*WJ zr16Fg(Zp(UoN~G*@E;Z#VAJ%26`7>efnn2Z1e+5S1I$FnsOUcFVLuGVX{KI3?UDLwmL@jF+qS_oG4uY z)Oz%-`OA4&@bHin300cu$#VS;Ofldo*m8+87=E=XOIhSG{v@VNv1E*_2~AI|ur&3~ z*5JdhI`POo`HF=36;A-4+10_BKDLa|IwwHt(Y&II10AzI+t&&PtlUg4&oM*hPHh#Tw? z?L1_n)=t;Z&73~ahg=)=1S%LNK8DYq+E`nxMtiKA+3+~KS!c=#?hMl+I}3Gd z)n479H>I_(8rk8!9|9G?wsG6_!b0pQg*Ogmi}OMb1>&wr8wOTm0B@5GAM;CZ#cB*3 zHyAhk-Ah}3YnXXfuZdqsdL4Knw9crC>IuaRQRE>NjWta19N+-r^@}n2f0LsH4aW)^ zE5p@_Wv>CXht{hDYH2$KL?!$MfK=siU(( zjt1n7!M7`23^#sy6SYzciOF?t&2Zxak3OiD(kpC0)SE1chdKwm1~_MF2pG+- zhR@ldLtab1B(0IZn{&=N)PO&s^Wt=}-+}Q5{rjU@wW}zmoC5B&=siGO)#6#s&7|X) zl12KohN?UC=q%`EZ3@~PSp`~7C|uDhNs;Q390l@|%D9kiOg4>tLrYsFgX9dfEEulL z6&wyd5V7^mB{7?$1Up;c8Al7fb6_xZVN>TKlx9?xlw6B?sBsR?7~| zi$`AawPPM)Wm~6wFz1KgGG!HEv#N4n=NW>?qX3skG4T{xM@9Do>0TY=#?VuSo?8s- zV9C>@&B@OP&^iP?KYy~eWG4@Wn&u8HgyYEjtwW7{vM}%}%f#aN7;(%kRW)&U=5&ynfTE#bM!L zVY0yP9bH(q0r*d`a(4WQzL!7RF^*b;h%s(qm&2?OF;(vM_;u5gNj6i91FupnRDPZW z;jKZ}#-I!(hESmWzF(FszvsF%E94mW>2erW$nf9f7#C}B^mOhp{Im;hOmFIZEaLA? zOD5SM!*PR)mMTBBG%|`~y>O<4Xb&3%&_fiQf^flUI7T1*vMn`v~@$4nY4vo3MOQ7xRb% z40uRBwu*up#No5=O8_1f#c+-I-*Zu;ng9PUtL zA!tcgdM}?|N_EM;n0F$ul-l-Mwtr&y5WSh3EMLdDD9z(;nbQ)r?X|@4Y=8aUecWb2 zy8Za(#%Uj(fx_m7!Ei%Xt_ZbZ_)RyiA?EJn=d_%}I&jQA-DusKOfefNvVn?5hLvPF z@-JhKMU@zi`I>g1diWMgq)cf&$`Z4zie zd{hBcah)g?Jldl;Pp9#)4e7X8tsIy?;-vo*4+N%hOj-!$5_{Z{XBC={5nluIkBj@b z?XTT)#i^;-cjv4Ok*%?2J(QkV6-yiDF)_})mZ6lfqTyFRFqo6e|5EgbG&!(2=`%7X zT@=$!k*ienDWX;a6PRLeXt_&lKqlSDFYb%10dAvZGO z9`Mvk62UvsrE_Bgib8ein5Wl{I5kSQD;`5mR|7n#x;1PWdV@9o)smK|0})A(??pC`hol6`p;k>>J*_3GCLj&+{%O2I&~S_A9Q7~rDu9NO zhop<6KYm*AN#N1hmxR4^gS3v`>5Z;ay}w$K;-{e-LEHFqX{t|}V6PZTcNL`&?adUd z=N8Bwf@&9j-%FrlaquPuQx?D7@&Q(uqNe}o?AIoeI(Fc>&_e1M^P3u~3sf=-PyiWbSpBr=EVuwXrm><^vM-|%SB7b{J@3x{3BSa5rq z6xg++^B|La;+UAg3ZX%JW|GKCM?Fxh$9dE>ld;LpT>R*ljqjL-P}t;m9N1h~2)JPK zPYZ4CNJiQve*{}++*IW*7?49?X=#k0#Q9_^K?Sl!(I{?(4AU9Y+d)&Lon^#>2R?h|ARj`FW?I_R@xS$H*X zC->I;Vp)9z?jM#HD>gc3DbESAU8*@vQAtqbc8ngD6>^U9KJzc~hn{8#rj+q3=uY`n zk^*gpJ3R2*7(q4XLKKEj@KZ}vf^&-Ggh1c)v3%WsM56Yiuh$x=|1sudh9AKK5vMHqX-p2{5e!AWJHWHr{d9?+NEkCh#n@?U=r3JEvZk$*V2 zCyf+2aK5S0$TA$Gm@0~tg9fbPgET>x?5g~zxPkw>Mn#kK)_nLIN^tQ{lIsv%Jg?RZ zOLCi}@j)fNb#z8(GcP0bAg4ckGq*`vCVt?GTVO4o-A1iA`R%J9YJ&6N6W%gjfvjvH z`cwKezz+YX`zIRRVTB?~bBMY9FF}YkhghMAI=XZ1vcIj+KRV8>5?w_vW*ZAQYZufy z_u8*m#`2h1hak(>+v~_L33JBNU|D{2ZpsCc;=q;#G_XeX2If)>&?;n5(MZyS{OTp7 ziVwAwD8V``)6)5KHvtipszvcwx)&J;J3w3dupcsQVzvUR8y-vGa1#et_8R(qD*x$_mX1E##=m0#=4#Gja|z*=s>{tw`3-jblEv4`$2`Eju8Yy@g$Nezd*g zd#`$C0GrW^gazD2X$M~`TT7lQ*6U<_u^R;cpHw;)M;&c;W+VQC_-#`RW^r`Jq?-ec1FuV}jX+aGF+l0Emx}J->j{pLJQ6u2cX()l zOKFEkyBrwnfbJNU!L5>qGSzKak`kp2*G577bqG-P*nMwk2fZxn^ek8aS4g#z$GApD zKI9g~AVf-%18>O3NcFKDJzJp$LB#gEUv|D_Z>q4pJHgapwFogYwJ;8hCkuWz`XNn< zR>(_ZuZYbJ$IFIc)Uh#iNFUDYASTNi%veH&V<7imcZ!U{sj3 zZxi;qjX0qcPzQ~xn#q5rV?Is2Ri4ZTsrxR=@12=f$#H%C9lHaVZyG>xE3EnsIl*of z=gR<{eBf1n-K%29Vn+6F!3cYsuHny6lOgIhu4z4clmcnAGAHOTJq&M6zD=q<5%(X4z$ zOx%}oeq`~Y&0qTaubXCS95xedq3vQf=(=D_p)UKPS3uhwQsU2Mcvw8Kmz`?eD8K9u z_BqDR5DpAZx$&y=>m2bf;Hni@<`pi`(po>&_8D>W{t9@oZjEc7F@O18*&(9r4`1ZD z8D^mWXd_EV;w!eefd%lW74c4rfsTz-iJz1c*5n0TCsgbrME_u z7NR<%==9h|cEX?!2s?#a<}Bx_Fngm`v?zze@T@9+?wnTnkn1yfRy9b>Y3&w88z(TH zBgFm?BwmyNeZe_`Rns2K@1(K&IES;9 zn=PXItg3)>DseocOf{OEXU5#x_PP+_y}P$Xu#3)KdR7~xv`+mPpw$8!uRvwwxA`0 zuBv!WwO6@IH@{H6cSaFpce3X~(Y+)~SxIzxoObyjM9!8R4e1vzX%lup`8|plrOY^s z`YbuVXG!&lZouz2 z;?$vQ64U>_NK*)=j<;NpBVG;kJ>^D`H`aK2?#9@D>nYgaCp)}7@wi%Osw3sF)072) zD_A!{)8MBdm8v=bdb>NlD;LD`K<2cEwf1TKD36P2n{Dr16F-pk%aM(CkKhYgicf>=>>I;Q8gVAw z?zvZ@@yk-~#qiJtK8gg=aElAa<1^IM|w!iTuEaJL-aR`8ggepgG>9szD^X0P!Z zcG7K$mgqP3`Q;0b2kN&wIoo~K&eLxh=lk5%HNM|g9*}}_ zvCC&)WVs7w$gxeQC|n=rgOW)6t*n}i{yM|>Y_u$Huo_0k4T?1W=laZv6l6MZ8H|O3 z%>CcUQD)LfK1n`9a@@kYs7&zCd+FGKcqM#4E|YV)W#kwwt%o^8XNmt$3BCd*~|LhrH@Vdq8xT@Wq|=|tMZhv z5u{BEVA4EAYJk}Q7w>)avBmkc#=lQKSqmz|*jNvYu?N$ZXRrbWbzJcN4_pm)rSz-s zpC{|4ksPDfH{-9%{Nv~gGNqyt_YY0BJNFJJN+ox zO?SLgElIRb+}-RmvVIVB@1@y$*kdQVK_HI$DxUoA_V9NM=!o6)-5-!j2cC7>j9`0? zVop=!1Qo5NG5CTK$mgN!IFC9-*7$bNdT|)kDa1bY4(L@*aBdamQ5Xa~6nMh<0n9jw zz|6GHtwtFa0m-M`ZdW)+1Ep`8%cG7hzDaN@>)`jxi^VnJN0e3Ixu`C{?$4!ocjS7< ztFySbNRjM>b7FV{{e`RnxShr~GqHraHS^Gm=ZxcBU+sbj^}TB$+&T1dLtWuW`&a}n>%7|nDXJW=?ZY%g`Op*T&&tOtje(+O}{dq zWS>SiR5#^A8#cCov*M*rAgI(~d)`+MZ_G%{%PhTf(asHW}$)lncns^bhwlBcVZ zl$%{U_*Kxul`F#NJlJUL-9etcWcd)NqQ^1&Ad&{PEAT0Shs-6#Uj6c9Ig(slUQ`8@ zCB&ph7(a08(My1A~BNE0YV9yKvj|O3MW_83bR9!Pg)3kM-142er|$5 z)h$bK-Vu=%GBka}DF*(>Ke*aCqCD!qN3sW5F|vRjt^yW6c;iRGYhX=;{S2_MfzF0( z*n3H61_oA*(u)S(#(Q2IH+HYN!+JhbaukK8ZPSkGrCDl(juEuElz7!TYb9yUjiD;6 z{69_*T1)d~=iSX<>c!Et-5%@lYP;{W=1#2eI<;9mob{^r3ZU-lGtJB)?%D+!dWgF> zbdvFaV`CjAtQkeyp2H3kzq(wi@ixp$D?XUHo@m(hCOL5Aw%llX-A6Hb6hsF=mm&>T z{qdaZ5TIKb_OS+PuXJq9r7(d#mD|LEhc)t4m=Tj9Ls3EBWp+UJW5nrW^qqTn=Y=O> zR@ITcbkLn1_rr1pwg;ef_!@Icv@N*cZXY*%eW7Wr-(e>g3(FQ$c%Dd#q4~W$wXHLZ1cQ47R@S2z}U z-Mi!~q-yYMQ4|6)kIC}bIVYhs3O`N~KAsP(wMcn>#pG;RM#}W2oI|E9XpbZ=!BMvcE0co`<8P5W|Kt0n(+C?xIB-+6g-IuklgEho=KLw>V%PiccEhcEfoJ&J{WHG3i@qxEmi79z zMBNF^;bgylC$wwv6LI0~yX+M6KBG(ydfs znSQGz$!Acas#3O)4mx&Dl|;3hlNE9_;9sw{+9tNA8%Ow2D$0LDZZ6cN4FvlQp^L2^Z|dq6q5uUe3VJXVx<@gBOB1u@kd~- zh54jRfyJ0QX`cbVxH&3J2LeMe@SpF8S!W;AWnzbPo+LxQFS16~AXC+P#q+v@<2kDY zJun67rzL%gM}LxT)vU^9M+#ZY++InufMrl?EI>m#M4mZeS zB+KJP2vm@Yp;JY`v)uvFmUt0j^yx0W)UygTiEeTlocp;29>rehC#ljyZi2ea7s_+O z)$Tu^Ak2)7sSQ7|GM`hv@muvOcLVhPp?XI%+2p{`D=@N1=@bJjS39U^oeES4;2}K( z3*OeqkIvSOG-ihob^yafh^k;431SeFL?ccgeTWD|DB9KSod`-s*)*GPvO~#<2hlmE zxg85!vMs=afIV5LJXpY~bzZtv(i5Bw75PPA7ADPwG(0cspT5l>@JnyV?&&*h%5Pd? z{#;WR$6=$g7Tg&vojhO9-T{o0$uMPMd%)v>T)}-%g4g*qMIh{c2vW?sf*agrf?gVT zVqFuWyha9XjW+{p-Rx&XtjDm~N#vzCoX9~|7@ivX4b7hz!#v~qXy7v0GL4iNne%Ll zfil*eSS(aU-<^3*+86OyTn05_AcR~g%ojE|qjVD{M_MI)5vmMjmwdNd1yE@8hVFOK zhz<&lxa)|k(eT*Qy|gO0Wp*fZJ16qUUw$pS0dZKPlGgKU71w|iCqJj zrT3&bS1hpf+y=4*#yHO3@FU*8j7>J@z!JcE)PWe2|(f-2fj75;Ufe^64oFkWOohr}>C<3a*pVjP3Z>JrOVWnJE}ou&eo z7IY$^FO-9?7`husKA>|qklxk6PLL%pp47}5tB1|%aZT)G%MaU3tk!YMc7C>{VbHYM z>&5y3$Nl|5W^Qy)ZZpMfqDTT2-A$hsV$JS}*LE@MJYlXtf(=w0)akuoajo6W#Atl6 zLGkG8?7eLn#rKPT{|!@B4~JdhT2Syv%`aj0}-Dr7;?SNZ-%mC9&xVbRV%v418MKf(~1OMxp$ec!hbD;dNH3%5Bm4fLy=jc4*r#ZrqFGQZvOVy z_k)Hzq9>CsiuiLkh;BoE0QNhdcgJYTDnYp|ix6ysrl;t#9jYv&>D1b9Uos_hdZmzK zf#j+3pRS3UqSnZddzSH+lY8_v@r9^7P?TzkDvD;&(ux7d~Y?LPXWXz5A zeCXFpuO`a`HuiqlA?m4FG5^z+@S-2w|MqVUxLLg>ej({KSBT)i&h0v*jPnzU8KTHT z=&pBbfW&XEi|V#VKO~IrgtwFSAV>&ro87HU41WYZE=VW4^oPemv9+p7qMg+!#S-j0 zUx^oCs##0d1Y8#%Sg6ki>!d<}^nb)je=Oa9`J4`ZTToia>W~#7stPVv^9;FN52^6) zBN{r3)|KSmQ?3X(2el5@A=)<(u>wgo@N~7J&E-jOlFy0|XkrCHXI;ZvJXpEaA@LmS zk**^JA^5$gNEhf^^~(x@DI9N}L6UsB6gBdVz}>MTq=m#Wy)G4;Gvpdj6g5g)C5djY zVA|Q|o%dFoikvyD4{brvtlYbszt<&6iPc<#!M9{Ryju4m*D7B2+&1nS(2hwHq%n_q zo4H8mz0Gsj>AdrD$!_OnIxifz%fh;)!%mgFEN&;Tf#q;&Nf!5!@Dk^Y@FgT*<2C^{ zvP0wkeoGfxp4rSf(W}bS_I}G?wcb9~|Ig$kyV=fx?b%JE+3qsMG*hIBiq2nrP5g1E z{K(=5vZBx-c?w?@Lv_mQ>8;FlU+t{B5ad%e`)4_q&uW!FcE{bSGPvSTo$?yE;yA&R zQebJ{OApCad+0=>Ke-?0zVr~ZZ@8mL0UwLP(?XJUch=n%zY8iQ!n))~<;_08+mOpm z^68}yFRq|F31}lInx7+Iw$~B{4g24G?PTT*1Grk> zX#RjCnVS+Fc!dW3(I`In42ppU5;f{B_VKnsAyb$95skuEV3<7adzQO={`!5-Y7V9m z9=Pu249l9L*8Wht7=*Ik$@=?yUtRXSgMVk2<-GAIj28^Xcd`8^%XTS$<@e>LVL^wT zxh#YLPm#T$t&(+aDt+%g(CH!xJSx>>D62W>8;$i6+w_UE0f9}wc#NA_27#H$X=?TI zYd;?`WHSD#{rgkMV<(5rr;`*TTd|&E)=*?66}{87AIJkS?2_cObm^od4__PuE8y_r zMU$|H6&$8`24){NCDr|sq^1Xq!1IiP_kLI%ozBG-ww7P%y)bE@Om=HN=qKTQ z^0+OKskk3rJ`0PiP_j`k;aKZc7+44_B_mE)nVT#}8dm*f@f;u**~?_oXw|yp=%nC1 zAz}bDCdwLj6k^o!%U$}px8@Ja%9$Mz)f`NMR4%aW!B_#zm-!zPKV|{s^L~AD^{Y;& zfaP`{Vx4%eP)`@8=Xw8dQYRwa^R;r6EsTZzcC#}-|NDtsfvF_PD-|`)HZn5XC5=)?JWJLk3Na4C-=kg!z9ASO1v1}e*59W^v;&?@{4O{ z;$C(3xYdHZ@ZEGmWFA!Hu922NdTEcKiar;RKdVXF5~UaYsOEI=H6mngkLRQXRMFWI zZD`@#8*}q#*&810FfnRkFdal2f3}PXGafA!ye#jhrtEPbcR6a2nLsgd6j_T@VQNKV zQ28uffn!f#gG|fsR@(b1n|nA`J$Z8%+Ib(VC-1m%s+iz&v&Sx}bKKt_GDxHG`%J&} zsDGpMz7SWw;2{=;uXEev-XJZ7j1WY|uzRId(l1L5fhEyl<=)U-Q8v9d6ppk?un`UR z+Hs|16`)615R*;2!P+vQZ#TAwiw%J^*agEoFexKW4g3|HDq7v>fiBgkYrAjFxpiasUd=I=&yPF)+r7V;9;TRs z6e*^nk&X}oaH3&ZHuMry2CNLr7a}vzh<}=BFnsL-0E%B%WU5>%!Id}C#0(H*?M6k% zevlG58nAt4K7CJq7XHriX$;!N(azc`dE(o_-|Lp`)fl8!>~uXmJJXiB=C-v*qp&i; z&8qho{+rMMo;9i0e?j7=kv&FFeFw!rD1HkS-4xobOy}NMaQM6Ji+lZ=qRRNU!@9yQ zL;)+4^k_ikTPNSey`{8%?<9B?x_>(B2Rp$BX6@yy-eJd0*)GphJ}@PtbJ$#>1$sI3 zcvoXD40;r~?nxrc@pEfOcUZn z9S%xIt%JhE&w6L?CyT6<*No+% zypHOcZze1*`p5wwTqtY24Au)mm{&7&f?dKln4a+nNtk6R?tJ8{<8K1v)Gc-9w|6TYyUXZ2*V(IWX z{v}~~0PacQH#pbP`?v>vkMqvcz0yxTTB7hQ%z~wWTJA~+6yz*Au1pJre;2*db*K0h ztR7Em`BzsD0}PvfdSEHp=D=|tSXGb8nB`DR7Ddvj=)ZKz4?~SE5`C`crHVfH*eFQm z9;P$tG{sKtYRGMGpHUXDRZ{FxEF0GKL~Y=1mTcg*L?Kp(xOMbK(#QMEo^p*7v!!F( z#POcN%9c{!Z{P6WrXKZ|qz=o1|DEEef$rTn2AWpFi$WovrDL9XeT5V{z4djU=t*_Z zdYP=rTCR-)P2&X!tLxID|8OVTU`rkb{Xs@L*ai3;*rSdyn(X>0<_<+ZqoNyR7-POF zu9>N0Hv^`S;3uB3o_V1oPKl%(qRW*;D=QDm;r2<>#aK-T`=NCT+NwZ}3+n37XTzqc z4*s^N6Fdy~snh|RrMR)$WyGn8vya;*M0!osVrv0TO|0!2aY80FJVyhao(a6wf)4)k zZy9kq#?{yDX=zNiCWB~oQF!*;V_X$(w2@pvG7u9~@#}ztqZjB45aSi}v1h*`{$lsw z%S~1Q;{e@w{68zMF~($*r017P1L#gh|9KNBWCuD2_OKd^rn{pQQ%R9BDjLaLZ$!li zvI1`_4=Ve)*PwN!ioWW5jeo(pB`SkH$gQEziJl*e5v+9G%6#gjKi)%sLLcU8W~YFL zbvEbuS-0jr1ixxPnaoY$>(AWjT`cT_z{6_JwtxY7zbzzF>;ocWK#b?p>|c9t^fG|x z33F-J=)pGrx=(cc2m*4oEy4#Pp$Zr*XL30ayg)YUMo8X9faxJyzq6Lvalom zdx3AqSr7EVx2918)XGjRrC{VJmGFOiD+XIFJ1>Q|#p-d*_g= z>JG_<*4;{i6O27T{UZFv0rMAcVuw8~-}qJpZ7?VU-}gC1mOJpW3TU23Evq(C%m#|A z2X-T12giQKR>|#|)pFEF3Pb_Z$r{?xqhjSSkG^D5w%J*;|NbxUw@nGsz9j1>3p{UU zm5;>+Wk;LFSX$G7FJfk=$4*fx)yUr|+OKSi()d7FZ1JZMn(Kq`ov!!# zqX)7$CV_WjL6IjEmP06TfONq;@>%2qF)onoiUd|eZ?X!pClXkh5K6vy@V2R=%Ep8^ z@M_0`zkM3kAkEN-k9woV<8nm6+x9`b9jlUlAIx`Wl^W?-j-r)5?GAB@r8E&WZAgLJUj%U zh8aibyf~fgciz0yoNZ%#HSkEngNd8U&sq8;u-V!*|7 zu^f;A<{yl_>CrYi3`DmP?Aq6y-c9GBpHHqYFvmmG-EdIPT>6s z_F{U~P>x>h)g`|vov`X73%HofL}PuN7@jP{#Y{ct)O(&ETs9?4J7zRe&zQ!r2=)fI z9~TjWl|Cj5TA8ef;?skFjOhKK^j-v_8rJt0!(x&O&(=fljnhQLlK4 zSx1r8RP=~{Or+|5_=Eg+AuSdEXP*IBmM=@iq_^b z+0&ra18wV(9OHhhX27Yp{ipu%KMj^->XRSNCfaG_6Qg3H2Ncsskvmj0I)J$%RN@_S z>!x?ljrT-3)ouReuPu8`)#4tFv4W>hdf0ET_t`geAI>P!oVB8X zH?{Qr*-%;3=$-&&tApfoDXNkH&nUQik}mn}Hy#pIl27&Q&s?!~?WXiF@F~a6*%TDV zY;bFIgQ6syDX>*`3$ukubWLL+DnC%YNUffEH#1vScXFQAxw3R*#doVAd71W5p7JuCBDrW}@6M|`lA>V0?J z+f85F{+?&JM?!1qC&hx4y6Vg(B z*@^4<&M^!X>%*eyAxlj~L|@DmI509SD4qgMA>bmNtL10Ww`X<= z>lJCdY*6}ZiK++T*z~!3Ux(Xc-w4 z?iAI^F7R@ICj!FK{1t%@px81-a7erXve^Ukw<+UAtEQFl3uGDe>S=1l2@)@AiF)GO z4XH=Km@Gz6A;fbv^ol?mV`c&%d2TUuTWlaO_PLfpVrIUi;!2daOzWb)B(Yx$RaCoX zec`D(;d@D_y39Q~Th%LFBfZB_b%!=Wy2{278EdethtODevhDGe-wyotMCzg(c(rYz zF6w~iHD8pw>;#AZns4!JNNQnt771ix1lL1Kyox4MtBD=KjBUsM<#*eing7eT4D(Fb zho5$k@-K}S+iEl`ouQbM6se=4)7_eV&btBei+?p|^!5ime|K|n^0zmA?RUuNoa2kf zKlfV+J`8@U^{R#R(*4=XIA3@khc?S5cRa3E)RS&mQTTbcE_u2eQhQ$TX@+_1x;PK| zz7s^N+y_Vtqd(g01AQN&>mhYC(mB@zWGQdW$9s&=L|JcXOy(x*KfW9|7LNY=QvYF7 z#Q_V%{u-Z5K_akZ-=$S)A*oOXtMS<#nkl#yqC%~Y826f)<9jp~ezUQ5vw5Op-OHB8 zSxx_rYpC7RFa6pS7%wh;960Z20mT^TFYyQEtg3+SH;{XCuM030?D8pHh(}`QYzz25 z6vOLa8O2|7`hK%%k&DBIbuCoiTzK!A_|Kj4Wfb;E1CKM^&5swYkr#+A2X0f=lg-e_ z-s#a4+8^E$wc>~WRq~zH-^zUVzp{U_@?R3%`?y3#Z>E?{ z6iJ|>GnBakpuf}7U1Te_JX9-zH8GN(J%T6Xf&*KTexsq*QVi6`T?1xbpwHC{!Jennc&bKevK)J(kMnl+bC<<10ZzjA62T^5q$gOzxU49Pa zgRsf9oCjs(sXjS^n7~|d6B3kH_&*4SG-r;WRbI_Qt7gviVS}6(jgO7?m_ZsV12pyD z4*qJ*QUlJ4yw1ud&M%&m4434XKjF&eiijU*$B&GUeAL%jBQ=-!2=;v3517=3#xaeH zhg|#lD79QcuU+sdu)r-BBnLhQnh2EfevVj&!0B_iKSS)lf2;aNqyc%ppRC(M4mq%e zyJ%$L>M5p%B1fs{KHd;NlkSpl1R?Y_5yMV>JpE6Ns9T7idnBi3m%ldbR20?qMzhDT zQ?hrO0)N(oH+yLOwn{SnTB0_R)y~asF=D(=n&OCmhU>6XDZPh-=k21e^YQv+f~!FD z@|n0>sF{uXn@POy8jk)B&qw;%qqTm+p5ULFK(H;_eA!wL(TC%udi_o2Iw~@ z2I|Qt4s3uxns`(>%qEISphz4QjRa-gq3hhVP~wmMvrvg8EA|5SsM207u$ITiT3cfI z4)!_13Kdh+zv_P2G^Ar87LqGi?}B7q6++cc?+*SIp!tdkTmRSSB(Uz$+voAV?5^w{ zy~C>iw(Pi^W=fxFfu*v=3Vh(rX0lB&R*m_^bw~x>M+v$$@+jK`OIZ;QQ+w-hS+-t>Zns4QE? z#}$6Y`)v?%yAXAi&ZTpGQ~WaM?LH}f_vr14>yXhY<3A)>${Vis!Gy_xjt3$px3J5{ z+yDGH<6jLB@!Q!BmB zSQL&&&Uy9%Hx>R|>s~EUEB3mqlb4VNnJP({32tintnyh}ezz>ir_k>Ts1-NJjzP&C zo`7w*&{~^Gmo9vMTC(oivVfH#s>E>o5hK{{j}oI5!VQp{M5SZZUDC`!ItNT4u9-0q zKH}6UjTglTx;XlJ3XR8^P)tlhZd*$~2-ZSY(2{)`4|t5M%fvsJ1s|))Ys%@|S&4sJ zO-Bey%=V-LnH@q6S{PHLDbmEhLIm1M2$>ybd+scp^H0lz4fbcJPsV1FZ;olofr}VU z8`-1^iaA1&LsaxtuWER$RiBVubUY74sxy_yka;yC!5O1(@th8RY(OS`Vg?pEJfbt* zkgo#@t@G!<0=PcrwzyXTA=Pry&qwO0wF^L$ z5-DUR9-^@lCamqrL?L9^(v*jP^%ti4kuSCkIIt&UK_xPq9w1e+$DX;|V&Sk8RKh57 zx%Z&YC{x+N$Eoe?8&7!1K(;EN!OnfrqsERcJVsq=*G?7P`C*$Wpmp3R*1+b|YuyYp z&wsR$B_xsE;>dCTWn`bxoV1f-)D%gjqP4Wv5A{#69Y9Afa|U`1`=EwSL-)%&==))H z^yZKf{}CrV0Vk$@aGdW=2ICMLs6U^3%%an#J0;G|aerGp?}M%bf@Jyiz$T`OTorF| zy(=prmpFGpb92x&i@SmcLS>wr;vzbM*Fj=lzY)B_=L=7@VuL3ZHf6w~ddT%tQG#<7 zzaSJ^S(W*;ofg;Q1I%;iQ1%JpPVo=?tk$$R<@F!B2(?8z?^L5SIA1M16FV?DuaEUdD?F-0OZ{ayGC_wrP3+{e-t! za*p#!V9{%Bo;yXw;wpL&geCqTdv5|4Rd(f%`^5LCx>(dkVbyC;krKKPTNaCm*kyN; z=}BjwBr};wI@6h?$5whA<~sL%>hqHSX% zDAK6#KZh!mDtTHDs`!`Ym;MCadLO>~ednHg?m6FYau{B@5y<1q7M|A~{^liFVq~49 zM&WS4*$1cLhGXPyM}w;J-#a|MZUIf2XzY8Q^nK^SgWF!mh9)>EVw68XV{zDrg{ z?xv9pES{;7UlLyu?;=L#w7Lq$i@h)I3ji4!!D671->NR9ffB*zlI-ItZBgBVIJPk2 zw7Oh=&9gS7QP+lg0RBsYarD{bdqshoi8<0LCW{6-J~)f_Uy|)2ZNyBg5FfHwSUpHW zR_}cto_*XeURa~7@Pjm5K|XWby(<#br_RCRN85u!VM|r`C+w_d!kW(}MJ92@W z1zCAt{QA!%hMzfd;lT^s++jFOpx8KytfHcGq^KQ{PSyd97|?c$bHw93lN;(ydjK?a z?Rfsq`=394W~+O}r^wNTA;F=N6)TJtA{G3uaD9$A_Qj>asnSfD{=qK{>dz%T5r=?t z6^R*uca_$E9-hr~$+j~)WNi#aaP?cK#LY$tdxwujkGk9*pjg=ptDQ&fIC|^rvpxRo zVX+p%VjS@DcTYeA`gn1$FxvB?x!M@1XyP4XL|NVAJ2k%QIJsWVgQPas{tg=B^8 z@Y*jQUz5QL0FLhEC$z^s)TCv9{l~V@;A2BvIKgbz+uxOL4s#OdL`2v+#D6R$-%v%ZfcG9h0gSFQ;yqAk_fCD@h z3&aZ=R)dLMJEI0Iw>=rbb)7If?V~JP8{@JpKpa*hQ=1ooJnlxip26&-GRba9Rv3su z;7*6+>OAMADUIAqyr4016HYnG3mOx@^!5KOw{>A%78)E*DMnP2RoiCl6=bnXnS8oM zmGn{_6p=wfKa!p)29=d5n3R)f^(V5pguv7tc%_+ zcrg1M0aA9~2f`wc6HG=p#^g3k-Q}GweIV?2Tcs=1ZSt*EZ`W7RoI5QPC&0_ehMmMpQ>Hn+wVPgA?V>X{}!UZfWkl@_x4( z0q~JsRN?%&7gpBGX6*1Xs2dc!{S98_LH%yclJprLh>m++4mb_vXood5qJps9{{3#) zWR0j3NL68u9sRcRoJxQcnZDkzV>sW`Ooy%_4w z!nfakyH-)M;9e*M-m=0@lQY`;vTC|;;pYn-_u+HAZoXYUG{$f==di=PhMoyO-O%lB z8{Ov!ixNct(tt~mY0^ybvPp%0h0^mX{w9OF|Lj16{YHcPF}%<4GbUfT^Oq=F0Y#T}UdpVn z^H36TgW@ib?{o~NQ|pnR9aB*XBZ__FBDX}n59E`Q81eNv2Ltpa9_J#EtUYMDn zc==DU&-yaaHJ|Q?zTi^9HlHkQf%22^LS2V12nsRBprUlSI5#}aW533c*5au>#0MHr zZOnle@PY#G>jj)e4 zY7SCHO-V?))X4PANhkTzN>U+QM~XpQ=A3YUBwlk0c)bj2Tt3<}AH+j~km0F%-YK>| z92b_7W+VwvY099k6g4pp77`i5y3u3flLK+{B(<3l)}Eeo;1kd%hp z3{7)S40Wu9fA|LZe9w>dgZaS2wj8;^>w9+Hq@m3-UjJW##n89~LaA=73mcjpR)!{- zV%Je*EoQ})%|qqzw=ppesSH{n*Qp-r!s>yY$yxkNoLI6Su$)aeQIxW9!2};T^{;t;q~fk z;dt4A8+g5_ZC^X}8m-5~eoTAH6MQglDRA_oPJNqdYZ$WdqY^-RRC&k_ud;dZ!a|Au zichY%O;rj)(d|IdQRQ9<#&oB4t443$Y*Y2Cx~CM(>moe_$m&i`0kVCde!rp|eX_W)ZBq`zUws>N2lc1bc_2l;hDBk}57jAC zN~W(={lb1Ma+pph9p=WO)U9iS?Ye-=&Z9UMEU+vUBYgU7rk7MOsAS~SENo75VORHa z-tD{-+}%HdM>Fx(8^1q3(K19WT^?LT2Hbcx;KD`<$e@OK&TA=l4MkQ`(a3L$&e<`t zYid3Nrol-ZgHP*WJMiY(9>w-jM5 z_vp0k{>c!k%Lktv!*|_-VLKpQfmq9x&}tfkcty%idaKWXrpoV_&^#((aAv^mWTcr+ z8P}&0i9x<6sudQeGQ}t8izvF*1|q?w^dhloy1}C==om55$CNqZf?%kg^~Xz)T_RV! zooxbDiF);84YZJQZI!n{jentu#=!}0P*etW3$O`1+9AccduTkL*#SEGON`TCCsEbFKL+VZ^#6>iUO|cvb5=f&%bs_)(hp(c8sl0!vTY7&4?a*&z|C)7KD5DthVGqnls7bO;X$U%JPp;zFj*X|e)AR-*~7n$4PXJdXmr?z1%zK9ER zc5ukPZYFKYSZ$WLRkZ4B9f8?{Hsex2CcXQWYouhx?Kw`c7fy!OO{ zK#MJ@es$eJlEH6^8gm}oaWru9uKf3J{%C>C_pY2Jq` z42of>GMg#(*vXui)ImE9uU!HG!s~PPDNjV6B1N*;N$H;5D!lWc>iV2AA>QAp+bA76 zdR|gZU-r48GAUb0q3qLF)(RSRXTNeuhQms;5ywWE)WI66|WAm7i7^U@jtQ|)m;N|;J zUyl3DfmbXjnfu;Ht)ztC5a+^Cz>8LxI!&=BDN;>EV`6!-$~dihVYdkB+A->OSc@h9 z5Rs{es-z=9buijtnt3ld?FBr|<=S`%&zK-KcrLsdc-!$mjN{x4Z3de8mEM`a90IKI zeV!-?dsR?M-xBT?x3TvaJ(jzeSN%87Y}a0tul6!%`@y6_qX|q6SpFpc9pJ)`0<5D% z-(zFnlMnF6=DWXp|Eo`DXm(-La4<9<6`16ybZYPl51dh9C$Cj?sOr@S6%4{&96Xyz zmxIeQitgn9qVW#CAQ;_K_}|5Cg0@yReC%s6H*cJs&_Gi7nH$#~Zh;4W_z;#uu||qy zP|x?SFD zeX|30OSZpI;dhr51>iN+G@i=|z@ODpgBs5{W&9Z(2&~-;XTX4atQl#xrVj2oKEq;C zmcBc64M}z3Nl}TFk=R4AP-C4v#7@;JGSRzdWzp@*_1+!yWmz7*Zzd#Rp~v}2q=`Nl zc@EN2lLaN}RvJSDskABl6hXQweW}7EKcu}6cMS%WYal+n!Wq2pjzEl~1H*_wUWl0> z|Fa@Ct7}VGbOppc-ity3~#X9!_NE+=|9&lglRqGYc=y!rt-v?x+Vi1RY zbjE@TE!M!ag5dP%t&(GaWNnIu?Ty+h@dDCBPhINjXT4CRR*@uV6_v}&BT7D!u>bwj*TcNKC6P zJoh#M$e>t{FLI`=QFZzO;B$X86S zAZ=AD4XzDbi%!hd+3n;4)LJ&`E-cup$y?YF*r+QaOPMbEVtBhYn^_0esCfQuUwetB z?CwTBa2jSoM%=Xfp>|xz4X0ItH;=q%fsUJ0yPsq}H*+N5;v5zeE~3~13dC+e*anuj z&C0Jv8uPqs6{++#pLWnAxhm*Z>8oiBt>t*9$WM_!QHA3;`_Peya^xa;;hEoiJnM5OloCg`Mt@;? zA*@c;f+La`IT*slG|OaQ2tjhzHFH+I2!aA(IZ~v3Mltd|P?%jjPyc~e zUfRf(%|ND+V&!($@dKEHK)?yI(VpNoD@+^W7;Pty=00A)nDFk)!YQ^|uPzG=SXdl3s;r~f zwG>%{B~3=U&(okrLjtu#DPUOA!ai`1b07zyy&YjEDC}*{Vb}2k#jk1)ud*$nby-kw z)b_Rq?~}BO_G{v$m&59H^@!M|*_ziZe(XC9Dc>VUT{tk;Vr4@o*9?7K7l4YLYvi=oW<#F-2TbW9?$TkUZ`x=8qO10h? zky#e9KEy=VD$w?z6RS6cXUx3D^m*2LUlr_^g7EKl~i;;F?iJrlLTv_pw>j6nVK$$nc6MTpPat} ziU+zQ2Iovh+BoaJ{0uSIO4ormq{mc;tuwdIyef!^S{!Jaj-}oQA^X!(p53TU^eXi) z@Ih0w+yjX(@_qLCt)MdL4O7n%Q=pMv65K@}^1V8HNpOMB3Xd^VVA=_#5siyI2S*FE z2hH>TWJ@vgOjY=E(9R&`%-`-Pv!F_{H8g?f&wqa{6rteefE%9Mvk>#6jx4-JZYoe( zOTWac1895)+%N_7?{}1y(sQy-QQQ=LbBGZl#f7l+@R7WG@DN^w>BVI}eX74W$}4g> zK1K@&oBOzjtN*>49$~v2@Hlc@IOfYCdD1!ctfCyU=XyiqgE}DC(nV(pDwv&gMgXV= z2Q7kJ@cz-{yV<-~9=m*?!K2wRU&;-Qd!?tB`B=uM`u{3^n{0MrUl=3-hRvt5DHd?C zgNkl~B&S-DInA*~StyB@mP)RGKZ0&;l7FpeJB&oFH2Nv^YP_~!VV`H7JXeD$l)0MS z;Tc-wpmCIW7^$=P;Nr0%WTXZibc`DpBLw9p{L99Pf3W3<=fl71!Ul$eQNHHOSm02( z@V&piV^CvdLWwZJI~|rKs|5LwBwqB&J6$xA_~Opdh3|lpjx2v1QZh#gCkKIsyP2$k zyT5vRLaqzLfFmKdRg^2wm6v$5(n;W%ToaaRvY^9{%sMj5NS_YQ4ZjRzcSiaw9UpSv zp-ijXZyL_<1;G;=bKpsCn-V^Mo6MGS^*OO090aW!nG)&pNip9o`RdLuBR?Y2_pS<9 zF=snwAj$slfvdgpsAG>uG=e)Ryj!4`y6AFPDQH?`Eh$eZw2; zsUacTt5kC1$Gg5#ug2V|V&AfPPU!k4cADL|cw#G#KiRqR8y9TzwOuyk&yj(>M7U%A z)p;GNIAs~e_sF#b+s~i140y+6|QB@G0y9z7IHInmc!|d_udAoaj$OJgPb4P5E zdt4d8SEUR~ryjE9@;G8Oeb`O0&@Vfw=rpLqYJvKzRQenPVJXk+b9$KD5v3vhsuC#0 zD)CI0RkHd&x;L{7)K|o>)xsy7IcfSnv z&%@s%zSc1xyXJ4UC9Ka>jD(}C^%fmJ`>b?n!0nl7!Zy_s@iB2NyOUlG zbT73~SzX1Xg|(0l+5MS#W~HLl%X}_Ru<=DF;dmI}MhWR*EjVYB8`A2yiMwstE}s*3 zgo7J1UKB6VC(p@&U>atxnP`aA*N-!Q*kCW)2Nvf5;z_O6{7muJ-&rDP|5N+3S>!W* zD?Bdjp{}wblDJQ?Ur?lnie4d31?s_URVF=18+#x;$5QSaUfX8m(v2z;0zM$fQd}acp}iu0`kgXVIxZvIx?S+ga1}47yKn|4V61 zjj)*hN8#**6}(n&85TjODo zi*#8AiLV5g$X*y)VEk>#TVxYI__^@1bFr1bwv%ErDYBi4#)Uvw4+2tq{dWI@^O1=P z6$i6rgM}b%s%x?|j}nhf%t4qyA(^KBvaEQ1pV#HEa(aiR5dt_k3+I)f@TAXwgmd;; zs)ufw?ZyZ#I_A(7x1F!gv_Pfj4aG9DogY+OIB-{P1(gDd-9wQ)D!N9rC#qgo;;|>H zL06-^B}|6$qIhW+tV*9Ds9jwnzv^L7CyMt3_bZJIUQw@GJQ=SxsEY!MKw=4nWBWW$ zDDNn{Rd^oeXwx3}I>9v>BWNrjatu(0w=(UQi@*2un63+35ssK{i|T6VUQ*z5Gw>5o z$g0z{LoQt*T}T6c-339HXSW~&yd8bntZb0eC?rnk@Zm3cXq!I>hVFFkZ*GV9yVQz* zu`MTZ**FYG=@2Frm^}?lIx#^W!V33;^V?MS=)JS_o$kr1U2G=Gs<6!-x#AvzGAl=d z4B9@=Z9cWYnxIFEQma7j|9B}dx2Q^{xBIS!Dtb&?NR_5a&B_E0gb;t2*p4y7_ZJQP-5^tw5hg2TAo8lOQ5FZ61#R-Z zJJq1xJnOA^-<@8NuQ01y0JH5zb-r}-3?qHq-GK&Vr0(K{oZx`aNxrGTj}d_ApLu> z*hqg)clkhbI2W#%+9bj0egROKndEmOiu`s<3L~!g;J{bPyX1cKV6#G=Jv$>|o%2{m z!Qz{q7DyfZr?1zM)h?V4m2EZY*+Q|KD0r{Yg@Z{e@gax8(`2>keac?hVqrDi7ZvM` zx9NeUG^Ed~0}7W-^j&gI zQ3$dUgYgyQGX#Q|DT}}=rIX8H$tr9tMqKtniAO)KFQE@0otjF}*!`fXy z9Tm?QJ=S^Mmtl(Td!-@CKr-D+3L_e3CI~aNjjTa!z6>;MjPzQ?Qbjw&X-YymXy+9g z@x%NG7Un6~9qnnKR~p8Yg6qOy;Yh(v(Vb!ki5EH)D;3GXN3?GnX@k~WjffuWQEjr| z?M#TYYz`_18~HiF$w|VtR(&=)fLJx*l%3kROE;f&(mC^L4z4$6`YZ*FH2q=C8QBhC z%s3`&W3Fpb%#2?K|INOrx@n%;pxTvjR+9hQwTeban>z${%WAu_$Cj$9-JZlE-Zqz)yip#r`XjLS%HGWInrcRjr&s1 z>!N{x7?1H_DjlJB>;S|FcRBwMKR|r-=Vxez1rR5{@y$=jIv1V;@3jKNc8X1<$kw40 zg)F*Ga$W)@g^$7?Inu*gWMI;_z_g|+;;^=y?v>|ATSbtlAUPPhUF-BYu^k|J%tYV; zXTyXaNd9G;{|;NL!6)|S!j1@sjX?pFR#6hvAnPq^!A^(DNL_9Zd`7y@*~ea+E&csJ zR&f|=%oKtG%`%cEe7|4W@3xLK&AvuJT1T3(;1g*{NV{^!ziybCNgoL+56SZ_c1G5~ z2)$|#Kx_=m2(38&(EOYK6X0$cR;s^L{x`CeU-pUX&fp;_X?Q3niDDBe5>G|vh&$ce zRPhW*i)&%TX;Uo{Uh*GY6~qaQG$|%ovns zQQz{&qcD<@e(zL1JgG zerGI|gDwmbj>u0fP>!P>*vFQJ@nKWU&+MrDmmjhPXaQOy> zE{*hJaYv|0j_xb=c>0W1(J_L(Z|1}Y-FUOuVN6txv|jx%H4?+L$*MM{GN@_xSx7kP z@i{H3S5yW;v|qU%SpSk$wF-lJx#!4JOkJKI7G zJ`-gZuHNA&(E!;gd4h1C@-~BB?v2F7b}lCj6se1UkXNz|@vQx%5fsw@2VGB<>oAqMqD z`5_Qgej9JYA?P@)A{yjiq(J%101gmpJ{YGhRpe^2WQSgYfz4c9R2q^h?Uk1Vl>}Xs zXM6X`E0}#V`@FGbREn}%ExtKW}ojl;HhVP=9y`Q#f)i_uD*_M6|4 zV*=PzcCv%*sLs-9Xlgqiu34eg;O;1hD?v@KXXXYB=V8f z*IY%hl@uwXqDw>0K&@1x0~`F$`8&gv85u4-VLNgoPB_X7w$v+;8@AYzh`Fp&%s~s5 zBZk`K6h(>xldMOf`QxxounQ6ndz$?X@q)y}NtgV;)o)4hZ2tD7uakRj{2FjsUPZFm zYW!SHu`4JNOGT%_%()xLm^KHELtYC=#@95*`o@ZJ9?xM;f;$tTq7!s#v?OR5*; z7`@MSHI}w&gRS1vzjz=+%b$31;6=uiN`Yz}1m1^trers_~U)@zTm3->L z!Mqh#!Mr|-y-ShXRCKO*OVohddetH^rr{^6aAr|1zwdik8_yh^kCO@gMrM)trUIFG zV4=|zj+_Ii19iY}ov=G%j{qpsVQ-yqg*aJ-!pjg6OcWUWQ$!c#z}ACTrO5LHIIl^O z?~oZlpbU9>AVdhUJb!5Mp5R_+vji9J@!B3p&IFFaB>#Abyk&cDb+1)iQ=|bE6fAW@ zgbz-YhF~p`en(h~1e6{?;~5{H7IxNmz^zu1p=}K$7kKV9scv#1z7^dZc(l+Ca>*%n^ewq9U8A)cmbakOqp} zxI&3j-mEY^7F=TrRt1CBiT^a0mrbzO?2I()<@F$*7xVK&)(QVLKhk2PHX4;_r1-f> zn5egMWR6oTh{l&w(J9RO$;}>zwF$xmA+BoQRJ>nGUs3gWVsb>ubPWB2QaDyB=15cI zABYc4JL;>y?6b|MFd|uXRggbTkJYPC^&EM}BQ$p`{9+jP zrtDelywhL(!w2x{1J_-CDf_{Pt1-*>XsI!3u!Ew5co_*R#!vcWvm>W@vAC9|AAbc|HqIOY*qa@(_29-Qb1d z34Z=x{_zxxCAs+0xpzqtKX=7-KLb)=We_tc7N*_?D%vF7AJt56n3bh1P+of}4dm_m z1iJ+V$}UAMIR^2VtAefS2ePx$SZSlKlic#etsG5`=A+1~&X=GWEvOs|p~qXJ&dkut zEijSXlDtC_TzKle+p6Q!D0Uk~Qcy8IUV55|b#E1&TzE-lP{&YLRPWzWwu;Vza!skG zPYh}xLFMvA;#TllDulJ(G2%PpQb{nXPJg_Q9DanCC86TK^%41o1tNdWK6IU&7%6MP zg(GuUtzgqgu?-ZdqoT341YK9m>jlt=voy^i2G){tm8?pPD5dk{XC$9N!h#;BKZOxE zSME}@sV^)jSKgj!P;NY>+@i^?B3pOYWjz`WKO75NQ4d~U7PKO0>y)_JpxXga7~sw}$xY!s^kUNI*)3WH z+TG*kiSq@59ewDKOSpkB!W#I**+2S@%9eng&su>CXTfmLwWHU2h29cX6|n_GiOQ8r z+3R7o3}&%(3p!|%a-ScrN5;&>>+ZrHUTZEw>4N�XNKv$&upGZq<>waco)`#sbmU zm_dryI243@f%YWR;t+&c8kzkQ-&3}-@EknnpVEzy8T2Jty)NV1P2fH>=uUxtNKU{e z-*u!7baGD7@lF-ce(ayk=bb)wvl9>S`gmP;_K+4rCB0_H{-QihdNZ^+L?7qBBp9PXP@xua8?-*sXAHPCOOpI| zcp2#n3-o)YoDRlmHi37+h$Y-wUAcP@=1l0S4`Rop~yN0Zb7-W%DYjQsof~8VDgn`NsA(v#6sa)gW{ahII9FiqkBDj=3veD zagrC&2Q&5}Whb2{Z1oz`nB^ScnPF%LzBn%_+Zg+uA6X2MSiJtb?pQ&o zonqT4(n3YodZQYRX;3W>W$p~H=#wOc`bI37O;(xbSQrL`+0@{g0xUPj1n0;y0%C&K zcz-AvRFt@_G)*526=IETX~=Oulc-gMl4#u#2j{1St%O>2TxY6PV53GliK$f#US6v} z?K||$O{!|3;KhRjZpi%&T7AI(k3JPj+?0!|Y0M@D^kJHCVublQts-3fesWE(Gm&O( z_Bi%9uy*tt{&K4{0uJv}&x7R!Ol?>2idMj%Sy7bqkKn?V`JcT?_X4 zq{`~&q2k#oi4n}tF+n`LI-ri;EoszUQhgHHCEPG|!;H0p9Z~&m8^GV^7_YOvnu|+#XI^?oib%Z#63D!69+LJ+#7Jq)?$sF*^EHs=0kOR__gp_1&T2` zc#btvAB`3yBemtoW80UPoVAVT@R_o^u%+RM{?s!taDcvjzVx`*2-PX8L3J_yh24@9 z%ASZ8k_{8`Ze>%zxEA{I1BWBK|H*A-8hL-uue`!^|sTwb{|53*Su=tODn(j=D%57l<%FUULzmzb5UH_8NOxZ zqFkZa%M@v*qGMqlDnSTgV@%l_%FJt10b7WXKBn9yt`I%|1}rRqGtpH+Eg_Igg{hxd zPKCk0!5g!oj2jOc)JKB0MfNM#dmCUrb;*CfR{tT>68eQQJ_IElk+h^cw9m5zxjIiO zo0!6gEh>}zu5zu{THu}s4(U{Sn~&MUMdfScWV0-TEDkhH$9zwuAJ?~gyp8e|KwJWU zA-cj`?ha2(a(J_2IQ$TW)*Cfn6<)Wk+2%7mxo||8qr&?db6d1u9^)}6a&(QsDHcY2 z?t9{4nB3IRXuxpQ;u^CpGApdW2Y`icOa&7!y~cp#wBm%UN{AW3=0tY$G|34rj-xPR z-Q#25Gn$UA@UmE@pEZ-WNTLhJ3qc!tSbcUn#X^nK7K|6-$|V?7 zeX(*oRLjEBBYxIq%7+RDKWr#mc@N1sozCVvmR=uw-*UIWYxJ7%=Ox$Af7^nQ1OE~5 z206uV^m1Xla@`717b&)pA`MjZE{`Kny>C!|GH-<#11~8+XOt3kJFvn}zssYC&ZL)- z1mV?rHIj}wt6o?-wOg>wJIzC%1&9A!GQH396xk55ms#U|FLcqAN{u<{4B{8P`OD3t|W8xw=hlMN%~U$`)a_E}*l zi(+?BBn`CrG+hwDJ4H;&T)Ka52mSWjZ`Z4mp<9iDzNl=U^Qyv#9C40#jR#QPn7qm+ zCr^eeI#lNYYP-UF3D!Jdwf@_VSe+Q*)jI$VBeXj9L)@k*IN}ZWnHHe*yaB72+g*65 zDz^ef0mZ`Dl?P>1ng;^&qV@GT4_#wi;wpu8SRBQY1Z%vDNR4ub0O^lFKd@fi=zA*y z+eB_e_dFg`(@uu34m5NJ(Yfo@jN7*M10EoqGjPeiQxQ!biBV}3m{4(@a z9}8&e|Eu_IvYFr5<-#GG{Z<#5^!aNvj`UF)GiNgDkDVh_o(i^q< zZiem*yX=!H?N`=BoD%+ht)NkN+2;x{kfaM*pzy^=o8;}`25rH@wE{av6pjvh>fx~= z>#5C;`hrt&Z?rEM!paYqW|D&=jYKYth;vqm_<&*|5$7lsUC*?N3VgByY9xtMvS?h< zY>{kcVx!76hc(N{%E%3#B`>rqTS&U#9I5atiCiqarl=&fA-I~E9RMoT?qA5yLoHp4 zAWe!Zmo1V9zKuGpy?bf#S`z9@uE34moGCnF600&k7;lG6qDA4KYvi zUPcX=3EZ~m2lq})_{Wl2YJ+@J%<0mjhVT(>7a2o;kG@CjDbg83>!ZI_SnMa6WHBx} z)4O`Ih~Gfu!VYk%mDx$8*m#PprlRZUG^m<0XtC~k(^os7WEUNr|36S+m-!fd*N*zm zX|sa%*z(CdCjmPS_L&Yvp9%wUkToo5S0=c>jdk2kn&CXLTkL|w_$?AZ-tpuOTxe|YELdq^kUN>^&K#drkIro!3r{NC()sZN;{JbZy8ZEz3M zBTX4xmXRm^m&u0(mLbF~&}}h^WJ zkz!WU&CrtR&5}crJLIqiHQP7dUm>4G!$Zj0=LibZFM5*$r z=*f@tUi%(*;J) za?c0CMjfuZUG}*oYaw{{{KD<@O2`QKbm|pJmroOuIHl0Rg=coZ@~8Ru4whl%Wl>yr z#wdHWvg#iz>#5VD(<6^bo~*ydprUIO zMvoi9pNjLslGI8I&5>s7n-{m!xVwkzdul1&ny1Tf8?+ zSwFecr%{J{JH@A_?h$7~Z7%LzkY@yJ2~Ly0)ri+#5afF8mtPVT%JJ+4L7jIgv%@C_ z>b7z3ysB)*ijdWedG7=FEJ5tljDWXN1NnP$qeRrxZK?GYyb-539PRnv7DWAd?TVMk zH5ZQLFS8;<>!nyo|L&rqQ4DmSJ8Bi^k-!<3Y>@iw0^62YDt%n3A@ zOtpS;hr~=a)Cy_hr6RbZRslhMI%7cvq_&g=L6K!t3+U|VF_Rk;Rj`sJO$rChSuCGu zYeFz6*XmU(fD%meqhdx$_&#RM|p9oC@&6&vDI3~Vy&}Js{!2}r8Em-E& z3^d2)M$q2f@mTI%Mshs#i*z4OFB6u{SnV~~ZfS@?`<&3BP49U#QeuJL$}N|FN#gh! zGZzlSt=PV&a1FipJ8%xG>X|8{(&okZeSe0td=!CL>LxTXt!0;Wu+N%P(H?V(1x zTz*NmPuVN0)GQWO(}@8f>gu7#;tN#7N>-Vas|c3ZcZ603!LmPm>rh%BBW{*}MTnPT zRY)->vr%PuOleA^=m2Z0|HQq~yn6FJygX~y z{Tk2wb>=SHN;;P<{pYBSL;m_5!2Ynos|Er~%e`SGNQl;?2#QOp>9_Vo9hmLO_L!rfx}%U*)7RqYTVJCGQr)( zK_Qx91`Fo(qgm*#58oWU|$db6F^HD8ixlQYy-k zqAQb1$A}LEpC`SZm|E4Q!ra|kU^xpV{cdIR3MFIc@P|Eb*9KpB*wv^5F5+bZCnV+1 z^Ry-IHnAa4E~#a&6EpeGy1$1 zMfG`>EW~sHy;-_((<`{MUx38D2O+8YukW2uE}Bvc{w^wK?H}a2+#$$|$P=uI)c<9X z_#RN~-d*_lLZ~A^HON}UDrFmJ?xG~gD0-Q1e?4)8;ylT}19|nA>ke5L{xJRZITqLW z-4a~|$>Zl5yKs2E64Pk@}D5@ZSbzuBi+C)`Z*rJ@c=j`s~=$pScYgLLKIgp*ZJR0s7VTohSf9`;LKRaz1`zU@ z+M>Urw zrepy>T-pODPA7}s30^4t3$yyC&$@(HuV-z@{%NZi{omrGl~SZHN)EoN(zmKVRTukW zS=-(h!716{mCs!9xv*e-@@GcV-@H0*loxZ4HX|i>tRMK}5!PaEJn!{fARAqH>7mHV zfEXz@gCgn3?3pGk_C;59!;DmE0~BUMRpVXK<%1zXEE}kU^~^yc&JvH5LH5rY<&pWR zQdDj-&bk767`bA8-#Wjp84Xe%8U%ii;&nAIq)ZSR&g4$CAZ6+D;3}XX1GIqLbAdZ8B zvsHxbW*a>BK~+c^(+WOBiLgSrUtZ#IH?o}00yo2;zCLH45=9mob%mN5#R*L+-S76H zx?0*rrvX0jY@aF*OGji|{4?o7@IGS2xYn_nR0-pw5~U6WjXhv^Xgr>>;Ye)ovIW%l zgXNcH7Gt9P`&a78+DT-W)mNBGv0Ew7+>727-t194|BMXkmUo3$_#K9I;a2)Mff!sX zxQfRJDqnTbD7H0N;A|E`a_B1``8AEMBSvn^kwuA`I4&x(O9Y+Ej+TzCe z{osHE+nnIXqi!yog2j>fo9+ptjeBkQR`;z?54#kI6|s`BRRmHh5y_Js-S1E9F1vB@ zw1ymfo*Nfu6UrBT!(vW`R1I9#SGj6sIvy$*n6)hSF1R1nF%@XXqB`mBC^Q(D9$`=? zkT$Nd$A1XL|N#~=>Pe+Q&SdoE(mNlH9Q z{`nkJE;osc9y?^bkZ3oIMzAP+5Ai`7H;hJ@LQ}82@y94zou+5X36FO+3w${ zOCLPGLv}a-2|kXKB28WdlG!)vy6EOtF3A>~b)WKmw#y+eOSaY{$=_Mv83kOd$0sKN zCa9w8TejXVA8X>maN+QS_r7{*Fj5mA(&bmELq8=!xLdPTlP}#MXmfC})zdr8Zd^RQ zDTiLM?cXFdZ(7XAH!rz&lO`9A>i1b0hR-M#NXV|@M0pzo@N3--{%1&y{3?{ynM4nS zE!yMc=CqZOE668OAe2{srg{KM0v8rs^}H>|t#bN-@Hi>-UGl;Tr~)_!STP^z5( zKBm?!Voqw)rk9Jb?W&+!_q{&p+BT>!X&pS$p-Kd@4t&Nj-x^7MXv&oRv->)tdw9~ecD734tF=Rr%D(h$&akL0vVb_2-T zoH%Tc3w7KM%t#*a_b%>ewKJO6OXApSsJu9-Tp4vm@iFL0^e`*kJAGQ|tnh5_li_tx z19SjmVce@TUbGk_H>q|%$>itPxUk!M z)XIDnQ7q_{?tz*)QmAcI8`Kx&#oqb8z`c5pt`uCI2gGHdRslTWRYAq`4eCyx&F=cx z7ncU;d*x*_^h*^3p`cs}Ou+4-@eDf4_vL3`p$&B_l2y5MK7<1Y>F1MGNYTq_R(L~j z6!3qN4dVTLt{Z;U^OJv;T5yvt+R#GcU3ehNvqFM_Vz*Fa6BT{Hvr*Zh*vl3KpOh!* zs-{AwLV|l2{qfWrzWQd5-QliY#_vfTURgWF5t>rN|l}A%G;=5+PRIJy$ zYttd5QsBU(J4$h-48RO%|QpEzzqqxFGIcKvr@xs(79<- z-BF%_s7tE!jBFrOe>zoAM@R+KIe}o+ra|?u|ZT6bkeBg2)sB3A=a(% z!pp>!RCShZRM%x!mpJ0IL&6l8>IC&k$*O8PfvEvyH7sN_LFotN)B5c5GpKXMh5f_a z9lYUT-HT4!;|GszqOC`tbrI4%3t<(Ax&vHDG2)(xB>yXl56JDPa=J!&PLuArRh<)% zD~k_l4mlZ_BrOE>wM?iA*$GkFJh<=Z!aGVldUf7rU_ZPBL6~eNj)i+gG2+9KeNo$d zu#EP0gh}~DXdHL2gm-W9MvwK#W4p)u`CK>fLJxI%pk}Xakk@663rE;?hioqtPD8nn zk3K1%`Gj%^#;dwoChw(_8^{j^TNf1*Taf_Jr=Sm7Y_dKwQ^~; zQ!G%bZl$8T7u;B2fZ~i5;uYcorr0A*cp?Aa@aPX z0;Ur3WKPgWMnM(Cv*+xg*|jw%xX<@7ys!VCA0O|QTRd6GEy+71;kgOC?Y06*8pT2y zcM27qEGSVYF-Z(gq77l!qV&z=8j~PIsvnc6U0DdKu9&vFGwg7H{=$O$vI7zIAx?5` zt$T}y)fomA>o%ORcll1zQCsKrIax&HaCH}d^^&X?=EpUjy`g)=OPQR2ezz*7h{fM) z1UUhy`)Z`GDEkzf~vB8g$Nv1z$8gv=|QS_mLaoc2TD3%H_Yc;N;a^Ra42QE*vviVMSus zN3nM)a+``?H4h1_uzt+owQTOkQwvr4nBW>k6SMI}eI9)$^j>H|U;QMt?kD!fd_GltV`QHU32sr_3{g(?%B|vP2QupOS*)uDH2HY;G zPKLE$MXP(+j60#{*$Th18BlCg*bT1qfSYN0Rgn1#)H;6v0NJ{+w-RM*7O_vOOc+=;UME zwr7N-go#Nhhn3HI-SgR>UVr1>!p|Yx(m`MU8g7GiV7Eh##AR9i)Cyq-aA9_b9{1bf zHGVj;3mp%KBj?X>hCs&)9n`?qZ@u~}%V;(6!N1QWCKn#9mROBecPO@tBA-*y6@CwV zQ5G*|ZUI{eR4b@UkFy?34FK&t%m0ww6bXHbeX+0zsiw@@i5tHJnf#U(21grI1x@$t zRvAHYu~aZblzLu*4Rp|0{fhtL+5vK-8r1t{B2$(gS%HkRwy3&hCc~}Cs@=*u&}mE% zZf7!mV4y)_*X1C6{IPdFbCFG)^0^GwzmY`<>8~~iVi*XU>kW+$GN>1eGiCe5FjD3F z;6KEC=YZSISNiDZWT5)!=InPw7TmpZZsSq1%7t;a)5>#Bq1cTSNur{$BNvM=EC4SE zRb<=gAxd3?`Y0%+q2g>QjU-V2&nWRSN)uX5^IiIl%4cvnsjhROKBk2R2tNGNi`iisWzRq|4#GRuYsAorJW^^bA{68 zj=?0)LGaOV$h|c*j*aPea^$)zw)s+gVmYo`4-9Id!26;gOEP?nI{%XZ)A@NSIP64UQZ}1x+g##Vf{VNY*jfBcVI%DV?Oz(y<+}7K0dxN zuOVpS^qZ&8J?s8Shon2=26(%%zDfSLxTw!oWzt9%(Fzfi_fi6H(Tl`6(gc`2y*ED4 zfP(d8ijZHw^FR9kee<{f^@snKEu+{a6p3;2Xz(=1KBDBSK5I3GR8VXwMGjNZCLr?O6`rCPOwPtUm?UX7Om6@=Nz!f^PH=L7Ny-^fyaKCTOUD6pViPG6PerHs92XbrnwU?gwuEAmuiiKf^8DRC4o8<|jQ#(lJ{a+w-GCYKZs#5R z?WHe8+OAW(tY5;h^4t-)jX5jrcZ(74lAV>VBRwxS>TXT0jojp$Nq52&B~{k%RvFpv zw%R{~u2q+6y2v=mQQF<VUBT}brFs>BEo zoC?5BfFLMxohQMPVX`W1##Z-S4RD~}qk975UL4#=pN=6>WUWO8bc~E;2N#}fYW}9f_Q(P=gQNU!U36lO%8Li$sQL0B`shwdEZH+Ppwz10XK#b~{A=W=gFEOIq?8`izA>-gp<%BAD@ zSq{AlJKz1#<>U67-@A8No`lz5{^zT~1&V?gv90(wJnj6CPK%6_6e3IhTHnY#W;ni{QQS8Hx?c~b&@bP=&vx~MW4v(7+*PU+> z4rPaWb%GET#p0QE8YbmA(h6a^pb%KV+vZ~M9MfBK#1BGq#ZA-TbSqDKhO=W1u|KKzMu&$Xd_1Y;v7O_#CDi{l=xi(7pOYXOM;&~owv z!17=lsFbyenk7xl9T`ZAYX_<02cto)qSzo52TyRn2up{l;PnnJ-B|91#N1a5YA5h= zj15Uch;n?8?Tv}=F0ozmKl#t%g-fV3pXzz_;Y;yA`Im9GMwGw~?xw!_=xi?a$6n9VqWbVG zKvcV^oc_qCRB+#Sv-VTfK`N2GKMOE7Rz4>OVdx>49&L=x8hu)Iq1Lwi{W-zV2Ubgd zhbb1gl#8h7E$Srg3iktk>tBAfRj(_RAazTl?ieWw*rM*xtW~|$sN3w0`EHq!D@7Ms z1Jm#JR>I5uZU$`vtQ}R;59VgQobYm%X8p@m@)PRyFZY2o=NAid0u1UskPwIKhuHy+ za(f((p0R=E;oaJWtK4m;L@pa<09g56FH$xUK&^ zQaJ&aJ`iK{D>J3$bB_AHpW1cyB4cP+cxrGA$h zl$gz=OgTgJI6cO^rLuYB1|M^8+YcfKp~KE*{@Q*u|Cbhnar|31{Yl$MscSCmtuD5z z0lH1Gw7A8p>pGsW|7`xg1`)S1sd+H26I2G-Fhl-U9J zr{Xhhv1>oLhHR9Re%pYK+d^%9_1>MaxV|nNy5n$t_b3yDCb~^^h`t@M58`tA%d$N1 zed?hy4O1ajdsTrx-f?Z8cQu2xXpl~EitGv>+$>TqRa6C4)B0S^Zb*(x4c4bZj%1Gb zs$h|LaiD2BHf>&Y!jd%oCNHdTOH%Chw#A7}_dIpiX}Egolb?LSR#WwVPPG^<(}K`T zWQz-XIR~x0oLq{9>ak2J`h={9lu97RqK4#rTtLhOzJ-f&Q~}S_W(V}pC*`OlR2u@} zUF1G(r87Y70oMysr5!M;p|AY81j4@ba;J&3j=BGz+>N6?%WYRi@`?9NFZBA71uYl9 zc0G&ib78d9S)t_^#a2+H6d0Z$iy8yC4}bzYEv!w2!C#A+C><>H>Odf2dbPRc5Lc28 zYvWj$ROLwbC^0Bps@cwBrK=^oE8nM81Pi9JayW%Gr60^IQMai|reE=fOGRySKbVKf zU8od{xEs@aY|Jov(r7hm5b;g&U$v!guPu4*<7){n>?v^(>P|l^-4j)yx=S|9dLZng z2S7>W?^jguA?GF4bcOFaQmd|2Z__-wq*1pL_zM?72rxbb_2xfQc1|q-xryE3JrSva z+cXbC&r8mVjUE*YZdJpitwWOW{cW0|>&7}NnS0y`pc(^|9^J}qo6be{w|ZIZ(9fF5 zTO`qi?a&@8JCshbK+ds+iY}M;E7QQrbSgFoiu_J$+ci6x&6*ZA?M;Ik_u`ofzk1ys zL9RSkUQK_&o(@iX)A^)8`1VJIhezG!v@LEtj4)PFl=_>?o^?)3EEGapFt}`l1wnn2 zV}g+b;R|_M*mzS~^1y}DATVOXyl^q$@cfnUykM~%0lzJIi)?aX+fi&~J9bhm1p2m9 z(dS@g@yILJ2x#?cK^!w=9r^4#<~>Lk2S212=vxw}EEZm$-5uKJnJQi8S;XXnr1<*D z?V=Kowewd_zQa#ObHw-lgoC3pAr{}{3C)iA6fdYyFRk;9AIsF*g|i_zrq)<+hz2Kl z4iq@|dqYL5qC4WEyiJ93BuIrR_QiSyOt88PuJb)QohV_%p3TU!9Uiw!!w<#dvp1h?=7rfcxs`u`k$+_?8XIG15P@gUn|tuu1<3-0)H}Z z7kg2jKn}e$m{HWCXj7vp&KgIWq@kB!=i3~*eB6HX@&H`d1h2XJ^Ug^YFIH!IS5FqX zuonx|>cgtm6Dc;HBCDzBD?S@$Y?s$ahPb@;LnLn8CpCk>vkR=_2GK_k;sw@;*%y!9 zv1Mj**)$douBCmlPM<5G)$~4PuWb3G`>KvPD<)y2zC;@nfl=B<-PL(UZD~jw(=DhA z84uch`vGF*Og`#FXKwKV2z6_fKxMnw!?C8)$U-7(u`e$2>ywn1{8y^BsPq+6n(&XTlXUflsAz?Qn(NknAWn`+(@V&9m=lzSlnUBGi`7xSzSX<5 z>#Mcul#zC2(aOd=PpEct1QJxP&j6jJV1ndlFh zmQY|94$hI83)-*G8B}9gEZz@XBnX#V)*eFrz7zjiYB+Otlw z-X7jS_jz>!%YKX)2vin}QQL4ly4$uN?CO5oo5o?EyZc92$C)rAKPZr~3`4*Bpyvcx z^4zd_rCW`r>nS#YB5^=|4EY35n21&d7l3fx=ZeqmxvNzjq2mrCyD$X9+c?0+>JfJ@ zzt;8>3o3qccP?-J2!pGuA3@K!z{UEf}YL z9==jBnBLp2JUAcew<|Tc0BKP72~H^Wr2WXSjpp zq-`ts@G?OYyT1CZB3peTmz`L1Xc3|3u?-AMflqAEKmZoF4!Gg5Qu*OpySJ#CzCOfQLpi30nM3M7UbnK)`kSp6VwbX0bq)b>h zZlEvwED{$+JP>RK*QZU@C+GpkrPs4Ygz36Qk3NBctq~oH%oA*U5d;LGEpAq~itc!z$Vj6ujVX-4dp@4J_XYbtI36F0qud?4zOXmvo?dMmgW(8k z6ivQF8s_f~$DCy}^0-Q4@Lx7}6Wb~3pwBLBCbfzUUN!U~ZSrLOAmudbakq}O!rYF3 z#wcH|9W9UT5?)rwb>qJJCZGR!)=399GK<8qLagS=0M581f5_*rSL0&F4R7;QMsH@@ zuxird>C9FFV+zRp#)u0gOBnMN-4Wx0DxQEE9A@l7#27DnLAKzkEwL?!*gAfY zb%O*-18z6G??+*};DFm+k9IAps^B*MSK@)$VFq zjTqFI{P$}Yi&su{c0S4IaN_)rkrz%}*B4p-Pusk{W|}4mFtnf*t?BCNeQJ7}nO^4C z)6+6NZ7BBi^4<;Re&4z0p5^-`)eCk7Hfw9?Qty>>23?LRQhip=$pDI`3kI&9RgKX z(s2Nu-A;$)GQUn!ye`;j<8QNk-Z3R+8B*lQez&Bpk~rxOZWn|ZTeJrP z&xlcyqD714Mj7abB-_kpq*BZViX>7|s0^`n#x8D#=n*$%N*g4Ru~q$3uWSug$L|jtVKGT+wSUvR4yUOLz>mqFg#u3*J2y^9vJ6s5ywky2)h=zY}6v<^5vGl=J@n#7gD&^ZqcvaE-7;Xk_b{V5^33 zu?@h_fi+P_)&ZRP&uhNtW?Dr))Lr~NiDl=yIIkHbxn_&uCW-+`%k@+gZua^~tE8Wt z_R!&;4+G5^bg#T-Zn-d5JOExy2ff7b-^3uuHqe-P2NQyuWvB1_Ok;w^iKxG9AZ6^J z;lzfd(F_`gD5i!Y2SHX-ev}-53gT4vVs{-@9yf%e0zd{3mOY|%one?HO_@?m;UBCa zE)pLIY?b8DT_8r2qsj5c*Lfh^8n4c!%YoYk#C_ebiVwqx$$qh9DM)DT5nuGFnPKGn z-YeVA6bc61GhxJWBW(;IGDDQL7vn{msO)R)mTL~15U&%1$jWMD%m{Qqf4p%S8g#)w zZy=BX9{^0#flVQ0Vxo#G@`^FfSRc71e&nc#lBE3}9M$UEdb@3`* zy5zW^jh{~~IkymNrgX{NeuC>!9-S#NP{id^MgC7{`#W&=`>bX#~UgH|40>3XVM_TMvMY?-|V# zF|^WPN*gNxH%x(!obvD#`R6i_nhw?F(kn@!=zs=Arh#XQPT-};Yb8qs@YaoprGiA> zkV}qct2gKykmIm;8k!|;!p)$?_2oMY_9}|}6M4E%7T#HaUHwJ=|78>QtQj2#p3nH* z@t;_3j@}S^WM%VzMR-M+=UY2Pm*|@43OdflBCE69V)dY`Bdl3brI=9L#0D7SH8gDQ z4^AAsYMG~gadYIv8AmIEdD~``acX5xdHEj4=vvuPx-?YSvztxSQmT?>dZ0vs#m7YOP+yL&-(*XnLU|MdGL^BtgU?MBt#n_HA5&9gd5Ub1{5u3s#1& zoyeb0)affNHE_0@eU-@+lSGjODyqV`L1@vIaTBuz@>_R(|80~@7zZWH~b+VXuG zoPmk_GU2mWJxw6S)n)O)wDK~=;g3UHI$^VqAvPNIXV zmFfvtRh|LXW{4Y|-e>nY`k?9I?g~pD!`I5wutK_sJ0VnnMam+*GJz8(y&jTBpYfS! z(}q?A66WORzy6Kxe}DX|pZxw8;*}J$oFdW2iGKFCqbK^s@vU38+RphrcavvP`K7O& zBWs+vz_-9`f!jyD5I?zdogpjBlRO+4uYx*EzDEM*BH8oRmIVo% z7SIC$QC`0+kCd<1eFb0V2(mo%U%B^1kgwPNanWDTi}2!o3(x-J^NV-=%QeHTIf9iR z+<>D_?NDgBFnLDZ!X;%3e*JvNqx6dRmXzQb92CNT3zs}v@ar+IVq?YDFvec}b8JAf zj;+C@pYSbs;R<$L4>^@aBdpNR3{#zco+dB!zKbPTd}7x?8f^vcKFDxM*+@3SfB zjulxOvJHEw=yKXD{534c*r7+8F(?eU=F@E+S0@$#G);_pliR_E1b2MRtOc73h*h z3l2oA@!6zU$7G5JT~hrwL0e`u-NEma-QykcO!ddJCl&bjJ`xu(=yFF~POn$Q3I<)C z%v;>h>-Vze!MQ1Z6n4dUx@`F3Guf*~ejO;v49XvqE2{P3 zH|3Y-pH&=-Xx83)zvqJu?;rol{vT|3zwK*J7qnNZUcZ5bI15dA(v(GBon=#8R=8@mjna9mIC|pQ`C}UaSC? zCJc*I(>UUg%S{l2FbeY;Kc_q#>&fg8SsUwIJ$rzy9K7)5vpm6SuAH|N#ii`O+`LT= zOY=|mE+%R0c1BLTT2+{>Rs|FTeQa4&)G}9;c-1Shb_VJy$sYghp7$i{xb<8k7i+Gj z4BovU?S@jTMUApmigWZi2{6VZ7wcAUI9CALIEHhMvQy)RJWD9sdbwh<|9~as_iN?9 zSRwXq6F!jKr>_HhLYgvHbKCtsy~Qs*;2{4IanuZ3|Hf^cd+P%@=pF0kZYW}sHrQlA z?)=leGEy)Zv`0p6c`7LeJjZhA7!9oqyA46Jed1PNs*Ule1Ud(eocr{KJ+&%eI&(a$Wg^P)v@-rwIE(2X7xD&K4Y z#+DLcktAixVq|--j6`BAY!^e~b~xE2ZB|0t8pPVl!z+DbK}RP+ls!L=4{zmYT9{t$ zZeOE{&+{Sk=-jY^urf}{l-TjNyAGq#-asE7cf3EX13g)~_}(;=-T7wGPk&CfyfL23 zelzn~Ofit*&ZVLng{j;fp7n~OqMf`RUK-Qmz0Vc6tC(In9u06m4LGa##d_-D*MD(A z3z-Ve8S-Eb3WPzYh|gU`qvW>RCq7TaXUTougE@B{wc{Q?kj4v{(eJOTR(V_YPg`j` z&JnET8e5Ka1)P48t*PU5Iz}pO^k4P)S=+wM=hsivk@cr?l;y{H19I)jqMAu26Qaqg z_yc+BGRp7o#LLfSGeZ$iF{>%E3X(sH(uizmQ0 zI^H%eo=0lAtDev~D6irz6L;yki4J-QJHsH;4Qj}i@d0) zJu>71jHb*T4n&N2)xtw6NK3TwtA(4@7b2_N%7xdwby;d-r#*g*@m?74D3D8)d3VsY zwDB_&M0l-Uc{K!7)S9J-RPf^tCpoe@B$-PZ)A36CDi5TLRFMv}lv(-ky^XlFoVL zldJ(A?+7f1dQ7f~EBsru55hZSJLhhipt;5gV>#aro#T;%vvPK*57%#-6lk&@K^cEd zBDqdnFjsG85B5?FXdLXKq7DV_66_M7r($4HIZk@SJyN}2ygJF} zFE73rIrZZ^W(Ax{!YWBtxCuDlsVK@M4wjb>-x1Gbsts45a%-b2rU$wKI#9Q(*Wh*7 z!dA+;1TAbjyaKo9LoS#Z!yTIrsn$`17)kEN1}*!d=tW$?nK8kM6^h=P^QX{Ul(KMFiyTqFgWv(d>yxY z#uG)W>rwJRGU$@@PRfiymz|1x!O1hq7Hp@Z1@+uZk^`Iv^m-`h-0YJvy^DVO_9fEK zq|Vqa8DY{R`i-zA{Mj7Aj56$DoMZEkWwDEx%b`{cHHN8gvPHmdx(QkF=S$0-2p{6O_ zk{HuzAz_8;5*6XlTdibDgemfSo`r&Dkne?XZbw)=gFKbsuVR^SjC8;Q*)XC0#X!)2 z*Dnh|ZoHEuJyMs+gOodX%evn1-td0+eyEl)oX->O1P$TDpvp+y30b)>dTToL<5J}J zVYnxfJ}9TrzN*T=biprk9lN&d$7Idb|r^Ik}24~NQdR>n3?B}vB zc|Er%0?ky5wsw9A7s(t`_UPVyxgoL1LXEQCynNL0zsm|H)R!K-ANh_6 zCY3>dSVfB1xkt{s8z9HcKy`p(_EBUH6@~G7tk`P@jqOH|VABHyXSuI|_%?$^HFU5# z?z#q5G1wQp))*_&p{2u@IrIt@6tZF~MFXS+YGL27AL!!;e0!!F`sW5wDi@Oyr0{7{?Zy{kh#(}bSBA4yk|?M`fDAgD9K$dpkGP>2)*_nhpU z=OI$78k9Zx0=luqGzdY`d71HO2mK2SLAlx>$Jh|aiH_+yGnuPK-YLb+T$ z13YVx96BW|=A6)+5x2@y=MB#4;D0h_z53-&?3liM`@mpKAN|ss*qNN){ZrcKmgP7v zRtG!pjuMQ>NFY5cy4m=J(_x z#E5w2pz5Fs9S)5Cor^ffZSz4?@bFiHu|q@#UqjX@?{eE6sKgCyUHadE^97yBp%%tsj&b2mXO%RZIY%8Rkm zjbf~MN)WApGU3arP9TcukR@<1o!KgBmm8?`GC6e|lr@0zEOn|!3)2I{TzDz&g}O?}#yR(v-}$rG;|9gEk?eb5-4-1vm>g+IRpqpiKr7@^MvV#-U@HZ(0gRaG zm8HP+j&uXD)M6t~K;8TVNNO)O{Gd}unvyVe4Jl4{zE1)EmUPLnWxDrC-!l2%$A;w6q8JmB+v_mLQLct zJ4fGl$3V+dlEBen-Q*$eO^$AGR>KU(nTXM5jukRS8^A&5*dgP;v!_qDLrRT<|iP-EA(S(%Kt zeyaF=%C}96iF{4K*T`8XUQD{ofN`B-u2AG-*f2vzuSI){Q$Ky904x2sc?<<6aJr-o zs-@vs>P8^@SpRN|Hl8_3qcU|HKOKa);(-q7LgXcSUvLiirbToqeTgoLK#q!o(^3RI z(ljVYXiypN-9w<^TDm^~N2?XwB$>Q;rYWRZi80{=u4Uc>@;dN(3Iq?rb6tC&3lm$h z@j4W>{+Lt;LYC3HK~?T~Brp>kARNUupt3S1BV3bl9NL6LRtTdqKU!5|NjPMMsi#?) z0}5%ne)k4Rp(u|Y@G#_OL8VZM72C)^bvpREAM_K+n_MxwchU3s-0F~MK^5E83)Y%AbtM@zVjkeW)++-<@W<_l*gU<2Bvg{;Y zJLjCRQQ8lM9u;q6n{yw*?y)Z6)6ib%)4CsgnxD@3@c)VzK&nppG<1OLFq1L-t=RbO zzF@qE?|0NsR;I&wAFz`y8~ZH#QQi>KVWoSeGb~rU4%WsL?i2B+0Vq4a%yp?(3b#>x z#P^b9fZGtR2N&fS&}2agmJjgc&#sJo5MIS=(e}O5F&)W<^vYcx>*Q_p8Is7`$vq`J zD$4R`(H21~CUVEv`H2`S{$F??#@1e7_3@eih4NFD49;wT*jz5QE6xBZAd*?HuS}%A@Nz1!6m6!0IHh28Y#=?E8jpS={!2Hf@h$|L5RC za++OI$a!5Ya?fltb(LZ+Q{)mARVdmYFx*{zBp?-vHX$;Hno64__d?rfEKKNxvdU8L zazIM8aLDC2DHmc{W&&?e3f<9%RChSop*`S{$}1FP(4DevxqLL<~B-k*lI(sH(rIZ1uYD|S1v5&VjErtT{S&Ms`off8l_$IQthef zwjq=?zFxFCY|q#P#J`mEzW=&chIc|F7y8$#copJ(U)C!(KqxMUw!d=3tG`d%Ci&Il z*!v1r9@blb_#Ja2z+`fczPG-TWRIq! zg3D%QzobHrWtolAr66r?n<=nqc5Q=_ttP>ykF@C`>rgt8efyWH;jB}x0jh#xc2lI3ih9DoDqiNgYsw=ZpyznEjXtc=Ys%@pk-glD;Ln!OxqjWGk>8?S z{{D4w9;Z-*;kc8e0Wh>-N-2FGRS7VW9`u8U7`ZiZg-sd&3Dd`n=v{QOVq z7R1@v1GXckY{H7I7yY5(Tj`cT7pKh)SP7%-k#*BYd=q%RvKEzYnE-zvag`%04Xl2> zw$8KF0UY#(b=xsombv{}LE0Bh_GHIQ&V5qr#EF4xW)`B6Vop=!1QqqsE%`?ep>A#` z_u#Z%ZgWWLyk0J@ClBc>suijnL)rkt{?I@K<#8y3eOHjgJd~wrDyfIkF2ycksi@HZ zx}eG}ZN>?DfUczWM#M`pyzdGULN-=)|T&I-`poki%vOz19@@F?tho8oKd;lPK8_ufkqgWQoc1z2}h>Uf>Tq zswSziQo>k^mjsPVc%iXf?LOTlMGm|ac?$#WYPiF~H#O&zZ973S~VU2hJ0<46bT-4G>?HIK(Ab(dIU+#=pE=`GU^D`r)21JHcE zA&WNIYPKI+M=@(D5=TWrluxgznWFcO7Mvowy^0#%S%o8K-r8*0ga~UlJMjG1z5hsA zMq=0)4kwP`S&7>$TFzS^o~gdAUDPCf2+PVf|HcrcT#FVgcYH;L#Z20S3S2cT?sN3Z zpMB)I%W_-91{F^1hgj2Nfm~rNT|`$4FThs0Ru;qPI>YwA^Z!Fm$wo{#8UinjUy{Ch zon@!I)5dJAbj7112^rlf(E{`gJAvyb-RmN`2|JBj-Y1#MkOqnd^#RQI>dIvg-P*{2 z96ArfJ0yLa6TF>XdpsSy-?wq@&HYHzs~0BBjwe_-32&X?^(ieY|Jhg%Ctg#m^a-T; zUlY|#*)1Rb*`h_|&kVmTk9yvAzkL%{(_)%(+hJr!jMy;LkM?i;ft$(w{7`rC_axSd z1N@-mIwGgMiDFVHvYv`c=in023zYmPyy|?PD%zmky-;wQdtD5{kbd`rVR&*PJxtpt zdt&_vb~fYt-+fo|x--8Q1r|$?H5p^XuoJB{Y)gx|J#>s3`wW5GDNN5`)3SqCk&#a# zD|AdCKv1w4yr7P-3j4-1t!jE8?0;xb!+?Q zpZ;{Xs=#f|EONt%5i)3okZy{B{=-kGs5a#VS%okb-S2*d!;WeN&9kFXFq?kO4 zWCO1yxW(BTU~6`ZCD+B7JalWYEGnMC^m8#)Of>;xGZySOD{p+T_PtdLG4+dexv_51 z!3`X&St+NhLPoP0cAB{{z-WZcvC}Ux$}ls&S!jcRuoK5jtOy8q0ToyR=OQ`0KzH(o z2^?mLAvs160^=Fq(=z(?P@xh#bEm4|`hxX);50Y(8?9bGgnS)Y_DWOOa6*Vv~ znX797=JsWR{b888PY<|{W%ihbi3MLbiL3WZ=j%`<7AY$$MIF=g=)9nvGEC<}Ubh9f zwKoy$&`+5%FN=)>(0OY4@H;Boo8E9Q^~lgWBBG1Zv2n5A!#K`Kh`*C-p9citO4 zzal{{eOs&d+v=I_b@%ON&z_lGibO#TIiX3Lx!!G~-Ji6Y<$qFbC=S({`#q+RsN@+tk~v+&cTRSwY(C<4-Z-G#bp z{DkG4gb=(l(+xg{+zG4aUY@@@0+eEe1)N+Oe>4288iBn7i2?DBXh9zb196W-e0O9DO13j=10JH7qC_#TYoRr?BM@ zia&h81jE*^+|4Ds*;y&){g;t*W>%`6V(KVT4Q}xluZv3-+=7kqX1|^DTA8&|%6->* zp{~&#ai;h}WVdTCcen3@nfN!9IZo~gNDC|rh9+44fMn(5Hh#3=K*ULMcear#zKy?S z=9ZbLKypi70bB7DbrC%{$9Pqd_y$m6X46}g+rjlNo7)#YZij0O*6d{qc%-W^yWkEVW9x)C<$vXCYPtq|I&*rI&bp z!sHnV^D^kiboZPMA;sZE5mjy|T5^cw)2I1mU}iHp)gWQA)jN4c2^WEbBY{Z<8CwuN zajkeeTCl?Nf^;qK>i8ShoE1iWx@(_X_DQqJVmmQOgB884Ti$yJdPY0MXBAzF3JHo( zj@wGu{@gyl(iTWOzk1@1ejYVzn}=zE`p>Im5lMPu7ATM(84 z=Ehma0#oFh{XThTgHIbj1D2R;;x>Xjc4$RzlC!`frB_xdON8ZtjT1tDX$wZ4Eh)nX z69APPB@3%8xm}#LOv(yd3o1|VjXWdH5DmGsO3DHfp_V#BloIje3qvl*%aKPTGe;*q zfyRN|bWgxMn?8cwK77HA&_$MOiqnF_$|`}O+|JMhPPwdKqB{U;Q+Addw|VYtf`rYZ z+3!MDka+9vFVFqL)3m01r*7*dvVq;2;>2zSkZ_DxQ?e)q^pCevQ8kjhz*b2j?-ED% zn6r7_pvx_K^SrAec>(3&*g%%5-V>4CMSZ%&e^!{HXkCQJ>l zCL@2^4?d%gwCmn?WS0FsPHQ-uzLsCN=y4s9+;sFGs&X}D}vNI+JC+(1hnvBUdzwA^}^2QjGQ)b4b znqm%6WFHlEo7)v|c-nS9OeWx}bRZBq2RY5kF1jA5=8U3>37j^5A3^TH?D;F_CJlgn}g-GJ&P21*u52KLbyT!fG+KMAR4qlF1I(lMSER52$Ia0i9QPUz)sJ% zZQs7LPByDWhg!_Z`KD#Fn3XOo!-Hn@1BP7s-H`+oTXQG$OE*r9ZC|}@(`e7DSbg=* zYi-Z>J^fD~lOxhGKlD1;JegFQt>8rz1JQ>ZDymD^8xFh4aD#JL9*!iNP>bllM>FKo z#4q< z>^8Wo0XVrjQd@FBCqL7sE)a~GD7cH0-TZUAe4E>yF46YZi{ix zTcd*f9h6eWGe*j;PUs7%3UPpK82RgyIU^5;Ona`yd(2LJRiA)1g5=6(iRmy&^av74o>7u6qUv>}p*}d{CU`$x+hShi- z0Fq2~S->f%#Lbw#bly2SZ~A68$9dp5DdRE_!?WsupX`95uTKBP|C)gE)1%*AK<+y+ zP}Z4Mfi8WEfd;k5RMZ`^$3I1$rS5lM%IN~t_+0fNRjd4r&o*e-*u<#}!~%gveizVV zHgo-Irlb-NmmbQomoc5Rg1=i#LYCEaOdtt6k5z(Kcl$mK*t7qGzH45@NUmvln0 z!Fg}I@SJ;9NVl6_Su?c>P-y&Q19AIK&~HWd!F7=hvNN7-Tw{6fkV}IqM^gzm8!rWc z8Ss5Du)AAUugKv(QSEZ;ggxd>`AxDOd?MtEFVJMr&EjrX3>6zcu7eMIP=kRSa=~BX z%_sq#^iSfNKdVg`l>Je+mF%2MfWLTzO{k$5@Fps$s1p&r@|&ScdHvAKS1Vf)R2-fl ziu1gzy29Bw>l~Ep7tzNeE-)?H{?K;$s>w)pUdp{6T<9O?87-(5;+?URJGJ?Ax}bqq z;nc-Jz6X9%CG3Zc0mgX$*_D$ zZ+N1hc`k0-fn}S!90WLffO;@7@_{7Lpc8qN-pS1pw>v7HJbrZB1OKx=Qm1oJ zT_Bh4nbQ!yS%A6&t&&bzqN`q+37dGkAdDq=hF8s({|FZ6oHIV>===7s>9+|YFSP^4H|-4= z>kt{mKv3Z~Tf~+*OEw;g6Fa+BQm>twR%VB1yDWxxoFs@6MQ6lW9{KbRXmVIPZ$Ipb zpYjr*O*0e9r>=kFx_Fi5vg(Go7?jVB0_U9}2qfDr!#fYD?6Sa&I{o9s)~I9J<78aq zl-MxSsv~q=kVSTlrkUr&m^fp$>eNz96-D+_QD+tPTxgCBX^L3mx?Yp1>=HIbJQW@x zwUQ2g0+^I|zvaQ_d~;znIz9b~e6we&S4IH#nzhQ0(%6U(r28stNQ9PKIL3SPff6wx zB-QIiSc?{4Z<6E$w9zO`gxBw$+0C~Ldd3NqQO%P*fMVSUef2-y&9>YUIqf=SWs`#q zLK|l-d^=!yU?@j zw^`mZGnVR+>iy!?dQCiYNOVbTENw`c5=)_2)SI%eJ7@xoLv83RV^Z$oH$px*UX{Kr&T4*O$j&__jK2US(jS5h^jD8hm6%b1|@Y+W36 zmhEzq6;!BidG#!?q?mGAb7F-wDv!S9-4(nYvcCo5MBYQU-tbypUQo8CQF;Z+GL!u> zpkA{OLNcwAbJHg*7RMSMcA5NF55fu_Z|%}}g#5;2Nn+y?B1zX96X9QDRwMD0Vg@Mk zh>A*)xAEf^HY<_er&+1H&Ak9ZWhwGFpgqh2wSZb0wGMB>t~pLx&#Q7Xx;DltVD!D$ zlT;rg`OkLEb#Xj{4Q}|l3#unFMTG)zg5XAVCa0J)RCjk0o~4jbI?bZ>ZN*RZ7Sh9IZ7_TnS91Q?jvt+;SX~%i(mP;*g)uFjS9fEQp^7K|fd`D+26K$i@CC3GMbVX#Ny327b`s0Pxm~dmmP2BGn zPgR+gwz96$Wh8GhX)yCeD<}pkE=#GXOdeLnrTWx!Yk0ZbTf#>EAt-=)610q$DX64Q zb1SJdQY!BazZBR*uldn+aiOS;^FY$b&-bVSF2x2F-dRg;6vgn+e|Ug7+!rQ;OW9qOSSw73I-T9;IyL<_CZyTgS^WXeE_#u5n_dO>{AA&HKV1gg*e@ zm&%AbNug*H>7r2vX35TL4Kd%j1K!oGu3ehVv!Cy z2fCih!S%uMz*TCOqi@vgnIyuGT7uz0tR8;&?&t4<#tyVpHG7_9u(&-=uEPWdzOiNI zuE5k8%Y!$IkcShvIo+0Is;`3k^|>r56IOG0vP`1IqNP=G60a7d=4{ zSfQB*L7ruTt3t3Wn58oEC$_=rwKd>a;6&!Z0R=YV$gtEn3^bz*9J8X4XUT)e#_@Gx zWLRNM#70U(H78d2W^3-w);)1g_M0dVcf=If28j_jOweI*5>-c=9FM+l`<*1qoY?UI zBEu1n|9Xm9N0GHu6qaopinei*4fQZ}zU}fm(j%@DKGC*KvtzmIMnev(738g&>X|#H znLzQ)qM!bpY+;u*aNeyyl-iAiP%*_6P$ZX%s+A1N?un1_`^8(ro3$<4dg!!F;2m|p zpe+&~2z;uzI~(~vTD1Gax@A>v{lZNlyCW(@g@J{EZTtrA3Z^`)gTI3d^t>>J!$4;% zxU6kqMgxXD&h7iTChK)CX+WA6EZ;KpT-T}!W`f#-aGf7+NR2fxNEXu{dNUMav;oD` z5!Vh$ohS6~+iyc+Z6?M#>(<7z*Lij(CF|7kV}D;xX%nb2c1UtG9pHmLb3qKNQ>&zh z^h;u-6YG_-!jZi|V#7E;Sf3-XBolJlEo_b1^6^Yq4T?aS)xg${iC);sCh|~55d;@S z$Z^;xDG?fS!k}2sJq^*WcFt9xY*_1-sm`UVpfqZSTcHTM=OIRd(O7KBUgNRGBZuB$ ze}ODkD6qDWz@x$!|xS*>+XWz>6I&a#|2nizir0zce~#1rf<*P%`Jb=Ve9zt7cu>MytfU%+&Dy-5JoxbiZI#>Jh;5>L3AWT% zQrThc{CmNb6b{_1E&EE<&pul?_?3qXD!;P*y)>%%Yn#7#L2Hl4WUNoZOAM41~zUcSjl@zm_BGGoYh-{y3;P@Z8wOL0o zYbg>(MIGjLfd5fMmtq9|i91Tl*kj7r{(3eA#aIt&hZ|T~yGf6AlRuwrLd0*Yv~{G| ziQ^?F%n)&qVxYoluTfnnduoMpms|9lEG`sA`1eB3nDI;}SrM)zbYE>+dvck?-Znye^*%Iw|5&1RkMM>7dJoDY<@CZa7k^|5{!yB$2Cl zeZVFj z8FPN-YJ&ZYSWRn?nZ{fowhmzJZL0*y$v~sJP*gSlaVYXmSJN%Zni)C_$g~3u8j_JA z(OCxlu@H-Fd*$Wfg#!ETt?dd0dsrRQ80?HsUqagpbx%$U94nPfnC3g;o4_lF-F7cj zH0h9rWv^^IbI%iqJ5X9A4&p8tz(f6_337DU!^4gkc*#(#;PKYnUp<^O9yTB+u86Y2 z24qw#uZ05owD+6%ShkN$-%xR@SDsR)do?SO^t*|V7eKLooP4t!S-zK?4(_B&y?6U! z<^AfponVmgbO*h}SeI*Tziwd~fdvS=;EaFl1V#2xd+8p@ZjqdfJGXyrDOB=?Y_@aF zmat6}lR}a8$k2FJu|MFf;uut)4RDXrmGq|px3%d~JYVIu#sh!aY0_Wvh>v)S4KzmF z=J3OR`4fG^-Q@FDe(7uH$Qmc!GC?|UM2dMc#cZPBdyPW9DkPRW?)r#SF{?b6pp96k zC>O@4b?1ax+&mf!60t9~idiy2v(9F|jYh!XQ*HPFR;$jWe-4cddff{U(BhxbJQ1gZ zh{|JHH(Z8bUmw)A$+Hm@whd;-t5|{Jt)-2gzQHDwap#}zm5~B=+hiv$R61&AU@9qQ zFGb3!s71Asb$*{KFpRQv>Yz-w&TlV}E@P>58(k<|(G7mDzjE*jrjWMq8F z(X>EdvW=fDDnWw6Hdv5KIQO7m?V3N_vJaXEF|%zPdceb8sI(DjuW~;|zHb-2J^AXN zcfRh+4Kh#W1!by@OUq4int&wdO_J<)6Vw&K8HK90Wr9r6gcAI)oo~BLrVWFzvMiJ4 z{>S^5U-yVUmb+HDZ6it0HfSV(njnwEVm53*gvA{b^O6-J-YPHHHTkQiwMMn#?SCOh zMr-$R;$&dE*@|+JV$M^fk%~&4UZ*%oR!!@n>lEurlXAn9RQFo4o)m{i3lex|MLoQ} zcUNjl89W-`m(y9I2cl+es?Q$JUGLvqu;inzfD2Fvw`y9fV2}Tz?Q|BsN3qCQ7`Lc& z+6qM-zFHpqWO|oyQG(*O+X}@VML)s1xkX!)7o>xH!v#IOr;4Q%Dn2d`&YpibvUOU5 zqFY`_>fqYrI57T}i)?|f4PYC&l-TFk2E1Aa+kbsC^*?@R0+u{SE+C&xCh=yZ1w$0` zgd(3)QAmTb)q9n?UD?LR@{b(46PW&Pb9ZsJU?UCc=2i!8XRy_2nILakj5Jf7sm=i| z(h}Jg&o1fq$RU?&;(kek>S3s^*SC$|C56r|2#RHEaOFoi2&24N3{;-zm4mWC9?yk`1vg=}G-d@hF7Ut7~ylP>KHctLGP`??16v6F? zOPV%*iEM}aS&=PP#D)jjhD%u5A{$1u&n2v2OFb+udvD6{ChNbi{+Yy$w(vReRw~bI z;Y*_!NM>xHqVheiaB^s)>C2n~(vAh3lhZ3B@9|Df@1GYdDED;?!{SS(kR4uLGLW6$ zvW}OsPu2ItCTz(5`o(6lZZg?n_H}1a3=pWNQBf%0+@{VDG;``nqpanfJi3$HqRru^ z&NxTk0aBcYnmxiS|8({Fh!$-^$bI)3?}>R(*?s8-f5FaQ@94u>|I$ZUTb$SJET8wI zo0grYYy$pHY*(!Gsoo3Ttlk=tt8SG%2(Jz)bOTx^tm~c_G@D_`HfWpOG9fSi`{|xv zShAoxZGzbf<7pfJEdR7JpDyz`tvm6}Iv7=z3X|$0g{CM`szE+%o8lV4y08LSaEC6br|J z-NT`=+ifh!8;6w|9c81PeFWi3X=%e(&uadyO_l@BTVJK#SH*wTl1|QPeHSY{cb#EpG?%l@>)yb)eBLo- zChy>!qx4!{Io$<9GTUaLsO&x75zjwg7Jtk+Ev%Ya8Ce;*+kZ>o%718HpetQ?O$=+3 z8z^v+LT-uIQr8c4+25|^B$7+eQFlW;!2MH~e2Z5vcPz`tOJ>u4n0d*EjNvVInE7tS zRMqRQL6bKv$piaUp#TyXy>ytMFjZ?Kov~i|m{TIeEVaD~4s4hdw&1W~NV{CZ3LKNB zeee&}mVsHPU5cy(VV`*F{n8-GHz-SrsFAFL1l5pB{1@Y<^b;V-2yK-=@x(KBwhb?x zD^}Qe=}7i{lN~lzdHlcqmK4eR%r+8*6a#aTjZEl=0&@gwxt&m^tXrXqo|i+TTnBD8 zPXK}AZeLxKClK>Vq4f)TwpMF5LuJX3i?N2eN1UQQ6j)7Xs?RDMI8ChPPUyIobCk9g;nZYQWRJ$x#hiETp5`I zPe7rdRWjhNyG3uFcQqsrl<-<5sAM$ck{)m$a_`B0SWJbRYX`y|05&efKL@DYkM8|? z@duU(AE#Y&tYo5^q-g4lO&uxngDNDA)fs(}nz_(2By5%7p8@y9xCUj2`XyDze_Ezd*|~qL5%xdpudKGLMVulJ}D{sp2En~REKtP;9ukQz9J^+Z@;%u4ecFYEWR%ahc8b+A>AddU zh7VeA3Qe$x+kE3^Bz`g}GK0ldib&^9%Vb)abCCG}HUM67{sc3Lnm!WL?qwy&qXJPFm5bJjOX9 zOLafvKK#=d55E`O$j39lQx7_XsqRk!HO4DI;1sT&2m{R;BM22c<6*Sn-Vlt$_aqKX zH`x$Pd|VCL%Ff|%;w0b!GgG*OVt_0m4@osTXzUGHMh2iE72J+y&OsFl*Cq0b7evp~ zZ3t-#Efj5J4hDe}o5gLA!H>(_9`}6ca$ilN1&|N|r8}f$**L3MjS)IWJO*Qey`>jB z8g9pVZeRXy>tFXBaG7_%dmG&-x$OpO3SsL4fB_Erl-o%3G-OI?(H04cR4EaIF8ShB zid6rMfCP>sX86U47{i>uIEVwzurqbS`a{j*u_(^Fo&r(RNT2(eMQIdPg&c))l23f< z#rbrd9DnMSdQAuXY|*AMPh|;jS5F&sDdnzoHy*8idk?wK!{c&;(B;QM%o*yH1Kd4g zM`3LY=rXq_4nP(T`LO$({vFc_R2lS#RiudB3gpCifl{RrGzkYN267O4sHnf(lDBB} z@0=9odmN){WrHp|+*bIMKrc`&u%@q(*OGn8THmLVi8ig0|2>b-w)!*SOJRjwQQ! zXj``aN#t!pPxDXqE+%PCOq)?*hKB-*0n(H#kc4{wy0}@}?cK;<1H~lRPgt)mr<3Hh zl3LAC8rP*J<-O24PI|yLx=Mt{c~DTW+dqvtMDl4pcQ@$a9&)H4-MD7oL3RU&Ww$a- z-TgIRFB4W;-?{W#lKjS48qgLSVQI1{2BK|xDyrYDT6l!S_*V;oo?muHnn&M*_*%it zR{s`lGY8*o1Ky?*IwSCNAg;t($rSlo&EwE+y3W^8>(Dbej1^SR25`{9_typP^fbZb zJ9S$xkqzu%;=BPPPt(dJ;)aBrEdVrr#ZvYqMW1m&iGU!u4DR{<5hkGRy zao8z4^NtR=PIaZ6R_2!XgE>oi`vY{?e(RX+&-1ArD@q(41xWax^3cC8j?yjE@h`7g zxaTW+<#G8Qe^4Ntyjq*4O5nso^{Gf!cFw&jq}vTQ+x9R+GDm^HwmdKwIym-5F6A_YXL;P6eU$!OQ4f*n3{iPt0q0>vmd8=pOSNb(xNZ~G z@wRa@fOK#z^BB&zXraWPo8^(fdq|&F7H}?-tw80U`|b^|Wc&1^Il|D41jks9t%IsM z;l@czjSx1)PfiRiD^e#nL-z-4n=vTs4_`g^J`gOIa*nv_H3$5gwYR*B1TET)oTXrL zjxJco?W3=Y`w6~(n|sZETWRZ)I5zlLAJ1OzNS~;uEize|VXbecEqfX=vnbDWy&GM# z#I2GQvtojg2wRu$u|J>_LjDgVJ7(yW+x;Fzpeu`s(c@l;KpEC5IZNxx!;ucIm@@up zRwj$0=ONvk-UCWwwt7Fmw>-QWasrSjFxoAXE>YNKx8Zx2QhJS{$f{kQ6y69%48_5r zldMosD8d(p#x$gk$q<1)xGWlT_()jtUf{3G5S0b@@>?ZHv8&q^xR${y zS|v{vIesI(@Ufw2#5!i*!)GA0ZiGhZ7kO*TuPgp#Je4p`Y?!Q7!uS?Lc~lNrAG8j7 zk8XIc4{Dcxw3g`*Z+_Q4F3+=>c=ds;7<%tJ-l&x~p<2DRLc?f9>^uE08!fPK>(kf0S!fSLatlRy zbh2Nj8dei5gfnO%VhpBS1{zf$c2`T|#(Bal0=wy+kkShS3fQeC4Ou_@&{7TL4RLd= zXouXVxA+|+NkI^S@hI2i*7abI5k_)xaLo;#UiVw|PM~#Cs##<$IyV2lzVp zW!~5q53wCL>}^T*Yv*9gdj`ELuwGH;w;SR)8Nl5PX-a$36S0MvbsyzrHiG?oe;J6+ znq_i*T|}A+lH;j`p_-1fJ`{pExYGpA@J21H5U9q#`y}&fJKz~%CYpZat zqEN6S%s%2(#@VqS5)dQoXWZHytPIYibKm**+cuW$=Fx?MX652$2!h5EoQWOWuCQih zE?qh=Z1&uUg8g-^FDY6a=AH*x zmY#frkdR)9)V+HZm2?XOft0<9W~DC4quLLJ08->eoEiJsbS<73&$&GpyR3Bs9w%hoHKfz@5y}6ML2njpT8I=wUG%GcZg$wR?!&Nr z=T*>})LgRkFP2?tY}{Wb#+DWJ&JNIc&?^mfzJjWVM_9xgEf{c5ksk+L&TMf9X_jKJ z51V4J*KxNjKOm8}ZAKfv9d=yNg0oPbr^EioPWq;Ny9O(MAtDD763l63F%;b72Yf2c z3)L5HJPRzZgXal8puEJ7pIlvpn*GDxj4VzDDH0=%(}Ob<%vzkP8MekQf^`9EQ#A73j0a z0?JxmUQnw9%!4a^^ks4X+^v zV%u8iXT`Rt3z4_ok|9=#100%pRhUQLosD!}212hUX;S!oI-6cbdikq?JggAZ=PxME zdmDqS1MVY3tni&VVy;*|_RlUKIfZYCqqOKm?4TtnHyexP#Nl5n;#v2388h$k*0>=7 zVsZEYH&zfo@8tYJmy)2Yse>-ZWWD?XuxE|Zlk;D_4vhrFc#c_ud^{h;>fTgAz~5$? z3{u~Zq$|nx(bC>d9PF(%vz}!X1ASS=R8%&7MTO}pw1@5TEv_9Ai_r%{6NGEiQ^noc zdS&;VlCYvcUGBmGcT7ih&}jTxCFMld4#EY-@Zo^FA%M6ERL!^ht)AWN^I4dkQNt1% z&urC;2drlPl^0vLf8W1KzHM16;j~#yD|Hds^HaH%)DF*Oe(5B8eqLZ+U>iRLiiS%T z+>(FvP<9d)G`$ka$^)SwTmz)3(puhW(jh6(lmXpXGba-yEC%IOyf*$P;0A%$Ejvq2 zl68LlBq5|!Q{lgq(+q$;`qWrWi0eRa8{1WKfnOZ*;|ynL(L>^yt6>ba>D!O61++^%EmU zDl(yhH-(ap<+5Cj-mfgE%1!UL3i?>mwI>>-zXa6cA@!k#Sy zg2bhq)pPOZj<6vY%wSk}P_|5v*)FP-lp0ge9<<-z~x}75nU&UNKo?!ptki zw8z`lG2``zMB(38mjcqvq^K`MUNRK&4b0o>y)-;aja`|S=tKNNq>k6DMZ)hhWNElw zIeb1_(=NZDwa@zWk{KNnfWWN0WJtTb^}~Pg>UuRtvlF|(RydklB?mP<1bKJY`Q@q& zwOSjdbO}$A9bs+sgE?#Lrp#&MEU=k>8;7*ZW!4RdPsHQu*S!#;V|s%MDti4&gh^gi zGxA+8!IpTlUk^PLc#%#Jbx9jk2_ePYwGj?~{GZPi8z4CT@gMPfu?mayKm0eIWxu4A zw!~WgQP_9t{W4vb3aVgR+AXXX>+VPkV2`jN#Hf##M^{L7$&u9|{hplPZ{7ba&k@aVDlx>CcnrOctZvW#n(|+fhML+#H*}^U( z<-GqgvfpfXR7^1tl*^@}P)r2l(8f;4O!d&z0_AdOYT2ei{;b7|SAnn|CNLo%1=}3V z)gKc#@E!wsh+bo)P;BD21HUsS^ouD68K?20-ciPB`4`LHvxJM4@3|(<{$|&O@9J>^Rgd6RZ;413hnD0kjiUbCFvxo@eH7D6k#5L`f;7k~s9ZP&n#G#}s{=bhwZruW-ONCTZowV^R#?*FluL_gGQH(fS(<;Y;iOov9!%zusJlyyL%uZrf6JaB+vU4w4tVOY{`+ZY zDF=+me$5W3xEgZ7p2!QX9n(*eZ8LOFLm!6X-u!&-Z_OO^_Z%<(AJhbYlihoDJ0 z9gJ$9jUdvudafaH*`nMBE=Y%DAB{xveN1n-gZ<_4O#O=!X1_D+jE($fktP3VTEBjM zA^qFrFuSmh6C0bGX2#|`#WYgnG!U#w5ihW+{nLF;UeHD=VPVHY{Ws2j?U*ck=7Z^{ zl>>6!7O(iXo1_?Txc}ue0>Ytf;G)CY*37}Q_>QoCkTz~s{>>Fm;q@2}N|9gr;Qp6^ z%Z9*~uE0Ny7&mo!L1=i;h~b4fv@R*!Xza2z)gi`@sdB@&u$;M*MuWy?lQ$j=vRd!( zVipkn$*RO!jR{pJqW-dhl(9pV6Z^~nq7me3hbX3oA_u9cQ-Wd$Lp==bFclhu9)_{~ zU$49^Kg4TL-3)1!>~Kru=~^X6AbmOHQp-=_Rw;XBXBD~XDy8l)w+m|ho_OvEOC&nu zwrq_@gQiuo#-mUa&m0PDl0Nb5fOQLlUKr>q6y?yyj~J&5?m!T+fXs`R{$H13pT7x2 z|Fg4x9m#TH*SgLO7UdKJgd;nss0{id>G950pA#BL9k3f=SZ}FOdVB62_h>;o(dDWW zc^CjF6m+{9SG#u3A=Q9K-|RFHrG?%)T=()l^h&J8LG^U#!VI?i@@79Ayf~NRIm2o} zdn-_If#+jFO~?Gu>twSNqo&dfHANJYPmvrd>Wc4O-+a2u_nPmD$YyEM1mOQA^@SwoX}?VKCplLV{&>UrD!^nQ&JHj4?nIT|Ng*nJp_t68Ci`tyHo zduq9Tx3Wt%iew!LOc3o4$fI-UMVIGq4P2x{ItF}Nv9%3bsIfbtUA{tbO$?3zrhy8f z(*$)rF)f78KWf~7mV#{Cj7HEA)J0D*G&HnIKGE!zLe}*EWA9r4no6(yz2XhYhaoQl zNhT;r1VJ1T3?u3Uop##pwA!oB!2w@7;UX_!VqF!sauS zBj53kPL&ygzwvZ=EB}mtJs%uabruaeNRYNYOs-FXP?lWxy$t!3J+siOGzYVVqE?ZK-Vq^tnwzAox-zs#8FVF6&#m;di0693vb z-#V-HGKEqkQRHJ-FNsmUFLuF}xesEsWpo+rs-TvKgKn|?hU?j#b|U-P10!^_p(pU~ zUv;M_PyL9-o)$Ebt$eIE|3Y>Y+8P>#_0lHb6pa@o0`Fsm`0!+drZQshguQ_V%`WjG zUIjQv22H7aMOeJtD9Z|QWtpg*^Uq|GT!sc#mW@g-skvk;73H?cLk0~ibD{?Qt-GL5 zi5ppq1e=r{w0RSwiGvAah^NMGk3v&XENk?|D=fUIVb9gDFf!~7#X$tskM8dVogj-|8wV96 z6b6B5HKkZZkwhwPn_#m7v%Tr^riit(YGhTwHPS5qUl6sm7eWqTVVIA`U;g(=Tni{X zLOd3crR+fAz6k+P^%)F`bV{+BBCDu4Y>!(M(iV^|KL|t^^XFrXzDbEZ#CYB|GA(gW zXN!$n^7Nq*IOzPl#dH3)qFHoEAu)(PtBC@oE(|ELyhdh$v6@i(AYt{$h@8}a=fY=y zTOx6W7=7r4*{9X{+-1`TAE(O;)R}(J+Rs10&k{F4dU(Bmx!4f10(9dly;}XV#pPms z=&JEgfZOQl^O{`l^%=M4EG8Wx22CA(fNz}Mt{!;Tc`&mNJxrnSbJv}QtT9#J`tMhr z#1l&sTD`i3eG~MPQ@Ocwu~rUy?@en+p9gL#r*d^crb?(*ehW8SlsgwhD1*AA;f{NB zp9j8+QW_^n7q@hL>3E%LBmeBo`B+M&pNyN(&P<*;7YR-S0?Xcs2`kcfy^cM$n56H1 z6WaBgIL(qOI@d?rsK}!)D9(jqg?B9n8D5=-Zy9eNX*~yE7dD0 z^1Yfm!vcsb`M(#EUF>2R_uX3prq02!%n?d)kRsJo96GCo0;H!(^(zl7Gdr%ueoM(7 zZrM0sh7mwGlOK5@{N8k(s$RA@M4JJ%ckR=#{#&Q&k!%h)O;UJ`F^$|xF9R3Y76;x; zkm4Jy{KC+}(4G99pihZ68@N{>sryA(W3ZV^1`FvNF;VvX9G-JX?7m<>{B7_2=zqRe zDI37i_+S2pq`Pq*t;`A;d6Z%^MY57=@ycyvFS}LAjWcXOAw4MVdW=#Wp~yih?zrqCP&(%U zcfPsFNPBmB(s*sMXiG$|yp?V?Ri$XLKxMgLF_f#M%e7y4=K&ow$|4x$TcGX=TQ8r; zL29R(K$E#2G%(>wvZz9x=-(&=Zd;JJOj2Y=YMWF>P|59#IZc36`myg}fHB2VgDlRUYzho_CJ&>!EbSXZGqC>ANxqjEp||HV+SX}s#NO{rRbr^ z11jzY_$dqD==DLmye#@5NfjL!zkQv`+679&)qbJCYqhtBysaaMCI~!g1UF-4(4L{@j)jMrp)E`&R z*ch2U){zj2{e-1cg#5o>|5p1yKmOHE{`(i=Wt3tuMdF8Em{^ZK{0CzFZ`Ru~ZoVdk zD-2wm>5=8)rWsp(DrS}kpNU;?V$Nl8yDXpE6>ZQU@oyifeSs7((-t{Clk9innENxU#pML0I7X2p(2S;99G=aq_~8|CHYYhE zTfP(~w}w+kYGxO6@&6C1Dbbm^;I?=PQoR@3Bb^gS;u;X)}T5-w@Fr(fU<+WP$>!RHvS>ZsO<<=Q z+X`^)!To}ai7wCCSWO3>*bSBEo^_UsfpTd5CBR}2zFSbg81qflV#okO)l`h}lTNYKyDfqUftGQ5g?VCsrb0n;Y-V{*DTO@$JM?HkN zX+f@aQ0Wz|rZTQ=&hr(@vFD0Gc&>U$JcB=p`q08d`1=C?4 zfb5uYC*N{=V%F5Zt$Nxr;L14s-Q+ocxgXmn+YdeR+Q;5V9W*jI607O64-~^mvRZu% z>yr<@zs+dd+5DQsxeN{3tLI)3|M{A{HTHpQL-d((_1w>d1)Q`wo9T{mtGJsXJh>_{ zXs(K@pjGu~R3Z2HI~QX&(-~o_xbe~JxmzLCzjj=Hcp+KHSt(yG*cN5b6mWJbGdWi! z9nuT4nEL+mC(e`L^26PK>p9tv7yt>CYr#D4{Jk?(ET9c`X@~0{p8r zRNOMnIbnWScI0)dJW{Q?X-{Z0>IwQ35jfHPO`#pCNK^eBo!KpDdbS}3j`Vcpt z&K_%G%5RnwDAop=9bat&e>14l>LKCsnGnewuoeF^j#(3@-=hJdq&A6(Lw|304ea53 zs?^rAeNELX`@r43xYxIQa^d{ z*i}LU=Zbh!C^Yn&w0HVFaE~8#Oii1hxP95A>f%1|s-mxm3pAf9YdOiHlfnckYLJ_j zR(80w*AO{+->6ZCCI1~_%Qo(|u@eIgdbu()=rDJute$sReQ8{0;Nbv+rj>8Lnk?EK zitp-Ftv>a)wKX_m-lAZEif zL<|LGn7{t$Ez3gI{EeIWq?}#)#f_J^Myn<6Fr@%uswyh3gxe=;kyeGRlk|C1#x#1~ zRib;;%3m|~`m_hKE%XPgsf9oM-CD`nh`YcyUKo|3F7{iVzmq(xhR&X_MEZyz^;x<2uJRyB@l6P>0i9&5puqbbTsa*D9kYG0 zK@rPSZs#Z8e$_KJHDi0of$0}Hwengygi%ORR5P~MyMx9=oe_#%@*qhAmr)D-`iu^F z%~&JosT<|_Q6tdnJp74gIg>1JXJu_h|MKYZYTIP5+b&KF>DwZZ6fJ?k%Am;%`O{wU zPT6|0ihD-n;18SS*KWGqK1>W}WXGOig$YX7HvSj3r1A`G>;tkVm_f&1-XcMc?^Y$y z;3oR(=?k-Q^ET1Ts6uM?5t|PCCgGXgb?mFP|7li};J!s@7k9S$y-Qn0a|{Wskol#>ae}yR=R6iy1dH7{oW&~o=_$15JIETGZsCpZbM>tjHdcM zw)&X1ph;3BEIC-!Kb)dTJK@3>!_UxcILkhl|9}D%@z6n&~D&Q(W&$3wc~U_O%bi6m7IVA$F;NC)kuu^?E2QI z{h@fZNs}`H+vW$(&uaQSp53J*tKV8`w!JNX&1@(h`dRrtWhcYJWR7JtvLTy^ml1Th(wh6jPEuA;H z&ts!!qi0&sB`=-ohB(^`8G*H_T>W?*e}$kDPU|C5Jn`ua%~~&Qo46);fAC{55RGE7 zd#X=WP---OC`N2MhKG9c?I1r!;s#`$XJZwk-{oKe?HvjusA8v2-Q zBltzt%6|A?7%#Zzb9WeNczb?H4EM14;WG^3lFRIV_-@uYS`c?r3h2V!0p*!c3$cP!(bwd+rAZ3Z=+deBJus7iS|hi3d&IlM1{rda zBh^s9M-wL<@(TB(p>nH1x+Vg-ovY|YU%wHX2d@pWutgOBqJdB{UNN%_w!8MnI@egf z960v*zHpaQQT6i`A9qi*0H^y$(q&{bJK(tSK0%EYaCTA(;8882;_r687yk4E#IeezdPp9SotRIo7sEgpZ=%M z)^l~+y#R)UG!iAF%Eb+6SG^HcAclZ81^C6{r(Y88i$TU^tZIOof=>SSKqvp^8S%k- z1z{Y?xo@z7?WjNB4EpW_%X+iqgK>!@%Z-zHl~yZA38g5aNC6esA?*c?#zH|_khV6^ z2-2TNJwXerHt>mmKKCX;Ev(f(kHw%%EY{*~Mgb=quru(sSu6|5WG>9sA{kjtSo@pJ z@?~S4p9f}*7={JJxp)Bs%qCPBeJbyx-~VlSD4jGmEvCv32+c}I*WoNr`5=1BY$GUL z-;ZvS^-OC4cels80VHXYMLFJG@;qfti~%HU|J>9Ry8IFI0<4*!w>tq_LkQ>N{g9k`pNG3B;&PlYxh_gr2bU>Kfi_C@-fPbqI3`2_zH>W+9Qv|t)`6FQh4$aU&i1=)dN$+j;LOjl@<@dnyD{gj!2AiNI7*R2 zR9q)-?YMSzySf5;;%|^^kT*UnenfVJGF@-dV5pl#Cn^fmRhGLt71yNtXnZy&sM71I zs6L_%*m;=V;54-RpD^QFMZTRzk+hzl|1qDOI&ciT8)rgZlI%x?|c60?%bfCdgZT8n`{=N${50 z(1#z=DMl7&cnkA#b)X2Sr<=kyM&9?nM&p)tp87iMGjCBN1t&ymSs`t*HabZG62#m@ zU-JhcokI}^KDCOz<*OgR1jH))J#fc(Td0{$tmKRB$ zPkeZx_=3-M_#@5Ygl^qFzz%)mX00|#EX$klp77Ts#f_IYaLooaYG+f5O%%xlK@vdF z9nuJO2)BT26?Or%rJQC(D}7Db9#t+*@oX0AR8W+FC0GYTPY6z!L?uU}yV70*WwRo= z&@`EcW(Sc@Nzsm0iFVTIAiS`fS+{k%^S<}sbK7zmHmM{x-m7A$`pH%#$nQ%|lby_^WN-uG_jx3_ZHZAaY(U zEQV4zP8vrah?@+LyX+D5iGZ~zGQ0BGRek$gQd<{zd9SezGiQbIM^ptUFCw8^js0WWiSoc6~I8il&FjQxBzu6Yhq(Y3O%P|Tr4+PR=tdz=AV|F@S-USMdZ6nG#!xAuqrqs2E z%z&xi`LY_Z>zOGgyY`)gRq-?k30I0-ztA8FiLqzBZec zd#&JBNGZVg-2&9$V0VtuPyCk&FLD~`9`b0~B@jZ%Q(uH_F^sb>C>DYcL3ZRBe}krp z)2_ZRX_r;Z)T#2LDrVM8u6t#0+5*ZUFXzJKvSZM47(vVf%u%XsHQL-Z(t$q7L8-ZQ zlp>8HsZ?B+C{GN`ErIvEpl&7|bSNqVOcaUOkdO9d?})BAEGq(;XU6I8urnjG!oDoD zP1>?akGQe>!jR%^k2>m!NqJK%8x}cWrg00LLP|ef-XcdqCh&&MDPFWI=p2s({r519 zhtmOMJ7DCOzcd6}R+`_G99~6o-PlzDT9-k_ql{AQpvZPAu1U2sWIa?7Bll{u2o*sN zn(8o`p}?X;{uxwZn><*9s#Mve!gMq4H|>($rO(l2V_QgNjP`<}%e2b$LPmPMUvu!H z5S=PVc?;YcXRjy2#KC|c^6b@iA{{(uWmrc2^4on?U$da**Jm@oOAZZ1is!~@k`^l@ zeM%|nDRK&XDsS=DL}*iZ4bl#29<-`9N91xJkv#egSqAhl+CQUR!M(jMSr*+QUeCqj z8yqYSZt&gfZICUCSQF6(@;7=qJJRI&wnc53c!SgH-QatjyGN8h7Iq`@OuG>{63_&z zc7_8yp+7ZdNhA;eG)SQiURoBt(x+ZpFU8BNLy};VbkKgHy%X&;q&>sj3y$plt6r3< zD=zz9)vLb4ea62U^s%l8iv-$x-uJw*J-LJad>rx&=>uDR%yANKQ6a0@-f(P9#&Ek? zvfHmmqTTMB5mp&tl-;9?>15HK@B?p}9|Nkot4eH7O6DX;4@01Y(Ns?RRae5^E9s^gP=>o`OaXbTQ^+}2>i^dIc92e>}P~8hJYPW#) z*M^DP{LZM29N6F%=mQ6y$3JfVR*+>fX@G4}yKOTyE1Nbt=ks*Rb}z|o-Drm0p(06@ zu#W5xE#)2CUduV4(U#3C_DfXk7`eV0mVWez?H0Vq2n`jEFMVHn z=~Yh--vOdvg9ge^`5QgY2z6W?SMPO%b1d{qSS_cGUm1`=azqKUwtAk2I^ibGqwgnt z=ft~bH4oG&Jbhq}sDN7tDz%5X2Lg6WRz=+BI@>K8SUwzHv<3!p@+&J-Lan0LzGIsl zX2`c@L5eKHf4?`&b%f~EuKL1<_@3C!!4Lzu z=n}g*aNC-N1wVLqyu#u;dq3*FK-Rf&#I?)HPu5e4EfmS2;#&C!=UfqEpj`&WWBtrr z@&4fZ(+l8AJ-mBKYF4>>&K7x9;+SEJ!Kp|G{pO%`OY<#_?B`6|Qzr3Q&wJq~gwm^~zCF6uW=5 z7#m$vq8M$Vpa}}|z^QGPU;+pCrLfZpq9q{eRALGT?$RZlIMFLl@VR%h zq($lcDBfv(*v`hSTF?+blqJg)t6AoI)BNkj07twd#N(No7 zZephN#uQ&R1U7i<){x>LE!Gj^$LLgz5eqTwG0OLQu6t{Vw1IZ4>Cz#?V`GPGPaVMu zE5kGTcfS*89HEMKW5dHh6^%@chqy*?yAJcPl{W!$FXkv4(}~DG+6#qOZTyGC9I@ll zw?eRH=2CwYAuiz5K}iL4!iYc=y97jy)8%)^bxeZ+q+Qdn`>UGh)Tz;r=r%qO@&Ooc zOKUmHL3p^EBuE{%PA&tTBZz}%PfUqDGs$9%JVfd}B#+(t<;L3&hpbH3E=mE~U?o&s z7id!>7%;VWd)g6Bj~|v=X=^zf#~mf0K<%Fdy|>4wcE}p(+mK=gv4V6tZgH)hkR>|n zk2_zj{2h_0(Vr@h1|EftiYsD02R6ttlh`I(;smHz&8RbRV>PgYCz_=be)n4oZdN2D z&m0YN`$0`|&P;pchG|z36fH}$cd@yxu%)as3 zBfh6dHt3e?R0R{8gX^W671{#MQSzlYjkNL4OY}f;e_U2Q;oLYqr%H&o-JDXUhIa4( z?N$!1nTWFo(-!)ra(lTgWY5e#kHciutR37NUxl5 zH@{UEvTBx5-Y3&@jt1@wuk^Z1+%|7`{@sqxZKc-PY@51qRXKw&J5mAaRA2fZ=B=HT zHo>4N6kQ@clFJ|+wPz;2*CjtNU8geHq@XAM8~jpvu8iKEnU7&usAmtm@JN4h?Eyy$ zh~0Q)V<;dlqbote6k4Wo$Z8&Pl@tmbcU*=uLv}%8I0HNQjP!ROe%CfwciZg)hVc7p z%mdjf>i6rZ`tRiuSOjDst3$7VCaf&+t*_So!~NLx@1LM=PF+5xi`%Yl{K}25w#;2I z2FWB6VzPN(PA~lD#JNBd?A0Sr6D0EXOj#J+#jSIRW~E&qu#KN=&mElbp&n-yW>IT9 z$8?Ui_&H0KhaVyR9&VepB+x=RsPlrDd%I@XB?TO!RG6? z2N5=pjL0b~M2yz_V%~=w3nG5=N%t|b*p1USpng0E5o;(#3PqBrxE+EWg5Cc4ky`9r z%Et!p>_}WW3cW@i8up;VWEMsQ3(r1;9V!~nyruNDpyGSAo6eIp?BXFePREp5VJMeU z0Gq@nDz1%w2$AtwX>riSDLZFqFA10M(jm*Tb4J2gO#WUForfU1DWW+7=p!-0!GDlU zr!BBG*wyC8ZcL59dh-lXqeexD){M2S8FyO;grN$2Pxvv;iI}X3oqWviU+12sv%E@` zXZb0hJ7CaMMyw{4US)JAZ-r+k-{4)tKc&iw?&RxKoqqSZ*L+LJk+2Q&%i@#HW?dcp zH4l6GU2JDqnH0+V?%x}}ZZRoGe|qr?a(Ji$J~s|iTCEI66QzL7gws@9gA~&c56Kzz zGVi@&ZJuBi&;;PWIu%~3;n&d%;SBc)>Y-4$1Iq0ZWl0Jwr^1GptRTGesxTuA`=6R6 z4Za;xZI(z!p{$QiW$uT;YN=*DqX$^g7i zr@Bj*(seWP=%#U}!daRo$Dn3weqMAw7~n5IW`^48<@dh$4-2fOFKnAaF0osw+!%0= ztN_mXNe4HYOyl9jYj^H z)8yiK%uZtypIL3HmETKmnO&_nNw*rnMJLSW zXNcN8>L-yc6JPam8(h`y%Fl$|kQRVAH5N&rP!cXs+hBFk17B8?Di35Z%%-SE(iM$5 zr7OI!*&F%PolZk*pSs`Wgv|aP=+A@Z)N&CC;;>Ex((jYD3wDR?V zUC|AaFMw30jsxupJ*1U3Uu+Iq2u%prXE3b_&YdjNXzY;WfAZ61k~A3K}z z)WLi@>8?76oG?|uv%q6=#&?d9W$fVL#@onH1w05IYbnJVilk6+^R6nZ==P{hq2=Os zKt&nIEPW)Wt+s}>+nmYQ)^!&F|-p&ujcaEv>-(K4MS{Mu>1>71_3|tV5 z_cjI`79Sb6)Ia5IgQl0eOWaAXCK+M({ms#9t$#j!7mBRP<|NKN^G>b*6aP$**S*NO zKBI+XsQbBfj|&s7%F2UHJ9BaIc4s1n<*&g5L? z-1ENZXCN2-K!Wy?uuInK9SG~UPug$Ojg?c>g`ddA*0(2 z%*FOE6%NiETi7`GadtmG?Y|1%bHuxKW5dGW-RkMJKJ}nc2-NZTPp*$i&+E9XF65f> z{`7`%HyG3Bc7LgCZe#-tHsd;aZ-XcObx(_TTlZGYU&sjy)C2pKuWk{B^ z0ffq+#mpo%fI`HZlq-FXsL@!g^f58rS8+4_s<@v3pU23!ATYspgXw96!V(q3y#3@i zB(<-4;Ii9)uXmcLP_!?kgH97w%-o{hF?lO@D>r_638@L$9);dS8pssn)5y}YV=~s< zV^@Yj)gx(A8C9Jzr(N0=J&ehJiUOB%bFi``qt1Q)KcCqOhO%)>-PrSBkQc>G{0=%f ztT^i6#7nRm6^M`XP*tfrBxzDM4{0#(2b2Zk_Wr5x!ysFcMK`Iine2cHd!pMxLj(DH zHz`p|W4YiKG|IZhePwT4p5e=KG)(jH;>TjR{mMdwP|(G0EH3dm$D8ZP~QyNWhs+;$d#DWz|WKnTdk-7 z(xmz1t9$3_-n|ohR%6gSnwB6(8gAq-T`0gWzRv?${4bJDa+W>@jaq#k+XUUDdU`AW zYUJL)3U9OypDVj$hPNA_tkDQn_rsnfC(Ofh6Epb2i@$B=cUk?;-8cQc{#0(YZTFAc zR-G{P^O#H7Ks1X^3u=&NLz(spFX!q_GQ+^?9lNk<^~A`Wu)^x-AB6o!q|9Prvc~`N zHzb{1O5csW+cGOdnMWxgX39c-@YCaRxL8OD(x&|NlDrTj)SnpT$HYf?+H^Ty&<7R@ z4u(~FWpYY^xK8hdo-Y=o)XREZt&wgQss@|1r^l{E*QC^6QjVn5+}L+vNU5QrS*|YS z7ERJap(OCKG)p>TjGQ{UQnB3kXp}*QLSE>U412X>J!34v!Fv4RKCq@T>ZBum9&T(z z82UW000kqUEoxJ5sh&1VE+S0^C{09H%rwdkstvPlPrHOtjK7~O>RlVm(^b)Z| z$x{R3(ntxQbv^k%d~Fl2ZtHz8q*70i?NOLT*J4{S@-^Y~&xiKG83%xN08B)oU?iOp zc2hmvpEvvG*!dto-1_xETOEK`YD14f8$eGl5}0Yf3=nuVD>h3&BjB!Qx*Vgfy<&`; z93yRFIWerz!g6d!Zf8bIb>@Z-e{Atizjej4i(Gtdyi?#$8B~e%1*N!2k?T}k>YK=` zjY&A9V7eN4USd@3R1^efPbuJT2157*pX}&K^ zUj|2YXH2~W%+cv_&B0iIzdv+4MDymGYJ{tTwn?_ZP?npTICo)hC=aT#%{Rn{H%Ni8 z2e?wqH}rcT$5g#+O-Li15vJYlcZl0a?+U#z8AyQkMQ&E36l#(p-^9(_0__>N0of(* z@vi41tJILzkQET0txwK98UnG+$G%$g%I^Su9qDpWS+kSQ&|+C1vgQ&x~2m787PXPbh3P4<8o61Kbjmq>L`rMg~nYg)XZLx}B~ zLwW@+oh2CDE&#duj~HV7->3C$9b<8FHEkbg$->vh$<4BIa#JZqGDTK`Xd1Yzt^5jg zlYgauBmKbJ6gM}ylKfzZuibRXE=;&OF*3v-*Zb9@{eNp2wc;pqYp5yJiaW%nelKVV zm2dYeQzOZ3f^QzwH-aWsyr3w$mss+tgtck5H7j;@#WZm7Fkx69RAuClFV*Zabs*0Sz$v*DK=1KJ*eq%E(aGyuA6i$ zG+n;j3##cwOTrIMIVXMTX31soVn}r#^=y+Mja4b<{77msJ^Rz``7WQ`J#t6vVgR=dTLXf1cWfyq zo=?QPv1MRTeYqZcMf}%ma)V||^d3&5XDi<TXH*zA*_ICt*df`4%fJqaGQ3^# z25FNDMF?Bu2LnJ>fo_D7MXgJ*G1UCpi+2rlT%i3Ek1`wh{qzbl%3Dv#zi8z zRz62MrC3doRa6`jN@vko;ZQ2SFT|wYgB_mfz-g5(zd3b#)W|}^UR=1?Y&-|CM^nDr z9$~>nLVD*8vi-He#R)51R8oq46xl5=7Pqa4o<#A-9%xfnYVlI{avKg;PJc+BOm%x=d}yK~?C;pD5Y zy8LYogjEH>tv(Ih6J%9TpUj|H>c51yD;RX0)Sq&9j$7@s)PF>m5;o9qIUWXm06S=W z1(d99LtMAD8w{~GECJBVQL0P}GFQG=39q@R31-mOJk+5*kQnMYS6^cViBXT1eO4|2@i&il-;lgA6DLhtASw|pA|kvC;gZBFSZm%Zfi0aNQ;(03l;_@iv$M)KzA~k zgH`JN9#s+0D+9uk+E#uyl$}6**hB8pu&cps0S5y>>|l#p2YOjqbTu@0woUE#xEQu> zQj74qJUg;I3P@c&^F2G{FdF7jm-`k1Q;8Puz=Q0_dr$+0cbG~ zNosUGoz2s3pAB3x zqLLz@X6>7v4w;VAq4&Si65B%#lPtPOf_%2C++DStY*)BTNDl- z`L7Rb>5$ykVlhx3nVKkCNC5}{UZYcZ&5~YkQ;^(3Zuua+eF`+(VEqX0`XZHMw#Y8K z@pP((doV2M4)?=YJjV{Y*?-(|S!3~e+9JnilKo@IXI8Go2}*H{B1fpWT%U!UmGV;A zfthFgkE?a+9%9fS9p*;QGr}D2JE|I4S?KQ2EAo8N^2id|!Kz;D9#TO+ z4y)%?@-ii#2(Hh_P#@r5fr67AZ>;h^E4?$qVHXP+vjCyvw=%<+r~3A|*%mv~^n)dN zq~f(n^Paadm!D9IqZBzr#kEH*@xICL_t+M??@d!XL?!Q#dRu79Rihy!6ZvY?^2QW$g~>ZP;T%i&+7RFWN1W z*F+SG8ik2dJLn^__*v*NBZaZnJSdW7lutYuFzn@!1XUy-3!-ZZ`Nh;WtRw+K0mj1pvmg@4=2wL z;({ytRzOIRTVE{(HD7oG+&s>1Rjj@)coQ zW|xY3VZAZZWgLU%Jf{=3AnW-kKUWlO&}<08caI8s!)r)>R6*2%kcDc4=AL&ay-Qrn zJI^!9js_NRI(Y}bk!gm0Jgs1*Tf{u?2caZ&gIz8l^r1l&DwFhd@sy?EWYr4 zJWH5!MU2~6f7O2RK6}@F7?`rfm_6sLQJ0vL-0g_Xabrv|*qmI^IO>aow=@TNH~$X!T*srr-Aalv^P_!AVlY3#x=V6-b1J zFXGjNohMbo4xygf`;SjzpZsI)ckleu`gdA|yMoby=%8EVU&=sD%k*5mtR7UZUGe?1 z|FVbp?XmtUU;Y3qcfoysc`KA%-?rVaeog$Cd@Ja$rxfWFSxv>Y0=v@ZZ$2bgK%7Mv z`=R54qJPN|D3^Ofx-l|!qzs7}Wo!`sa$Gv^j(`4_-|x4rs(e1R;=ZY9!caRoAhE7J z8i)kcuzgX=X;xq*{yryEEwTIl?Vfvf!#aF7vvn)(ZTWz*EHY1go}4Cq9z(=OZd^OL z(aJ0&Q;L-oSwY1$OHAPm#x~Gk5Zo}b6)sOF%nm?2y>}!Jm;sTO@!rW-y&SGpcwd6M z#s*D>Dx1D1FAT2bERU?A^VQcu{wC3Huc}{oOJa0j8-Vp(*#U$>jt=YLhy4IE5QZ=V z)H}W#KUZ6<#qqemtRXvJn=rWE%8neQ6yRzf0MD8{fP%RaUb*=Cj1`JCqRZm-en=Jk z8RuAxPIZdAUh*YWTHctR0h%i1;&NFJr1a;V9`}gbuHLGgcU?2DL0Tcc1smQ)9R1`r zNeT}|;dYT?2z%ybgUqD)nWbUEW=^W_qwq zG*S9H3y9=-asj!+4n%Ie7$uD!+*sOADIQbgODYcAJ1_-*h z==gES3zN)Q71ZNZC2WX*Q8E;d#Ybd^#`k%|hi~`kGf7I_knS_J!WKm%=aSZ-Dpe+n zHu_ahz*Fd-=5|2c25vuK4T)Zz6}$re7d*iD(RB!XDMVj*!#8Jt@NaKg5Lhw!56j6m zH?GqI2JS&zUi&G!~PsvcL*0QNVm;CUzmS+Meny}nS4Y zt_1B-YmxQKDDUEbtU(Sivs73Lr=G@Ah%{MQbQ%C8$xquE)2=pfN`a9Z6xr$Wi4SFM zG%7_oXp77i!xeaN%obVxE>hXA){U`#nBffVF)oykn|tULCzdsl`IRQsrs!M+;EnsabLqtjM2C$2!$s zakFQ$q)t{!YqLTcqy=I<-Ro_XuOL!K@EE#K&~eqN?v8Ku zSv$FtZi#p>%^3^EFyLa34H?F;4nM+--<2^3&-z(>k>)qge?ZdM#ZvCOlR}EqR2=L#%TK8`h9)Uec*UU`Lr;e`h9<|Xonp{bg>F_92#Pp`+-{-|%%eA|s>71~ zVf#zDA^aeJe`xgtr#P`MnoWBEG*o2rqA{KEUQ}u4?jQ?Hiu!-Bn`Dm}BE93rfpwJ? z7m(Bfhw3;>}G`9LpuPXSl$y28%PGU_A!2Ww~}_@!q)>$UNre z{D@SIAs4M?;WVW34RcGo4I72s1Eq5&8KD2o#OP!LP4RRNyVP=&pC5sJ!3=B&zR?Eh~KdL z?%z=xYCf^$zh)D6yYW)RpqILI%#&{{n|tFYS>Ina=1&X$*h!z9dqo^iU6X(K05WmA zMyRCDf-g4cekmlKc*C-_`BwrhcIJuVL;}$bwKTaetYMR2aDJ|sQb5?ebs#^dMM1)I z<957(nl7+5eFnxYGvpo=Ul)o{y-De&MN1m;Reqh8(0i!%q$lNkLiqKW~0}ayghM}1uz~W^&XPv#_6|1R^Zr0 zDZr~Oq2hYEyTq-0ln9F#?1z1%wc}QktdKUoL9;68`ixE*%O4ld=?GaIZh(IwyUu^Y zgW9f$PWp_fDRR~LG zFT^XEe%D#`(prDJY|e&(5H`aP`(AbknRNBXU;nkmvh4k_kCfbUW6P3Ywbl5DQuI*d z0Trj4wKIH!{GqHmtO)w#Q6Na0#@z?$`zr4J=r-97Q?R2`^?S5{Xp0u#*rLuB?F09^ zLAr>7?V%a-v7Iv42l=G3=*0vZFiq{UcuyV(vh3h!kt3oUrG`*>2)H8pJ=Sp+3Mv&% zs(tD_WrBQHsJ24X7VzB2&vu=VEOq{Ff1?~@f}Nbpob1R1xoPt$U5;aHQ3IiXv@x_Y zs!4h#_}tiY&bD%nR#S>q6iKAw%r$#0P>`TSlXY!c|FrGINL(e){i4~m6GATlV`)e!4Jvnd%6Yg`*#o6}Wzp;XcX*eJ`#kD}HBhQ*2rZ4lv&-Ug zWes@HnSKRoBssgQtO8~7$IuUvr+!4=g*8hrX^Xn6OpQ58&I{XRNc*6Nx|kkG^@K~} z#92p4h2Ih0{phBMqvYn)!#rnMDm(FG|BB-zG_C(qy~DN^{58qNFjR>j4P3&xNk5*| z=dn9xyYC53gJR3%O0V7ix+%_@Yx}3hPE^=Gpp)+Ws4i-zt>S~*dLRrM4(CXM6w9Id zJsx`}jC&Z`?}21x$oWzbyhOTLa@cFrSeIn@8Rx*x**4n%#!;OA1}isYbVHEwueQ-M z1K-YZXh!?cl;kvMlBS*DoTsybcXF=MH^dl}=5ue7tv;y{*elwjxFKE_km2R z*T4|AaI-~=!;53J<*|klyU%0l<1OI zy$o0uPmg;nhWrRB-{eP*l;s$hUVBhsS_KCl7@^m$H^1{AwmLCx>xD9C!hCpxq((`qTcPlJ<3G7c{X-IX)LhPLC>nT8w#wkZ2vRT_iV}H*sLgSY(E%i+?DFI3D~ev6no_ViV6^$D{hQgVyk-Uwl;)8$247Ds=VZ-#RiN+{#x`n zv8TOyhRcisEAIprK zogpbiyGC9hKTO)`4_is^A-;j9k( zO=ffQi|D;$!Ytb@J3C4j5&cj#YVMnw0rKU+n<{%KML9)Ev3&Xw6x=TJKFT%D(y1EX zPT-emZin3CWlcjG$cA2J^ zE(~tc9Q{_$`?{}me5Lgp7h=z9oC_hxk@FNt!$!$a4>BX?XE$=*nrcDL2bG#yQuK-; z=eQMe4p52;itMK1%IJ8(J!nDCneb^~CvOF(j4q14A6*rpQ{{6DxeGbR)sKBI(KR&G zX8>7u&IF^po!=ka%3n7996_$YX2}1Ri|;CLh}-#{P};VY{_730Bh)y(KNyDd#PV^> zrkx{SOwIIj7Koe(A2W`I!oo1FF~g^+Do6acrFnSOuOsBTrR1o3!h>mfG?w&3i!yk= z>0_OR1bZjRZp7F-pp))1&RzVcmlsXD@q)xqH2vu9ZR6I!-n?ZCX>oWpq)@S9zQA`U z=kctLamQp^Cf*SD__Y$G`oAhZMmPGVN45gV!O}6S|J}B@j7{+4#!zA?MVpuMHvC_P=dR#onpHFUJm&Rz zB!;BC-RH5?zdFqLrfYy%Y}Kr{Mkjb8pACem!YU?K=w&zFRQo(ZAk^-8N?-rCmqmQRjZR|Gj%Zz4-lt zIe#tx+V0pDRKc92_xJwS`LERc-Hoqx#qOP}d-qPPHtGGcca!Gaj%|2vLCSY?zF9Np z*pK00_m67k7=N;$X3oa1t^ZaQ_1S+daKZ$!_|~8MeKKEgd|H@)rbA^p-L@V5S3d~g zzv|^8m}xh1>VUK22+%8iR1}T;OG|i0d4u0!@u@NTzsvkqD$h6~-B+V6=Ax zR-mQkUVV4cWQ$?>@FU{^vULmrRf|EDeS0YdONRfkE!53y1IiE{l$f;ptH!L3UkmES}B0GTmb$oN7BQA#SeekYX^gEDZQ?}kfv`<6+Du88+`{?T;A>PF<;776QrBCbVc-d2nC~O+hsY~ z1CP10X~R{y;RAM=uTg#1Bai@@ee_MVEZHe=?`oFNN0z z`6V1V4|*o=;^>8f6ykGsY$2J4^e07r>LyIY0sCqK36zZGwbhaln zbNZ(R0jV6lN)F_6P%$te3knm;0zcLOmwm$69-<3Mid+=Z7SJix)5p~?cZ&oE13G9d z$u)a>xNV#f_aECI(vQ-s+4oLT^TsY6V#+xGk+6p&XmdRN4a;n@2|E)VuFT^|qv2uaq@m zAmTD8iiH*$?IzHp1mU<)aCp%VLV@f;(Oo{QB3J}>ChSXC5I*-lI}^CriXf~?Qk(_F z(n1jkO^LJ}QmDxXahDd<%>`cWL@zykZk#qddQFTo<1Bm7@j}C4{if59!)!`^Q8p(f z*y8&9rsVJ{lKa}YKDAb1cp0VGK|%U9u0)mse#RwXZJ@T&>wZ94V52aXYj$Y*J@ldy z?y9*uRf*pU0leM7H^2Q6lBE_4`#pAsoK&G>lB=wp-S2^U-F2`k!CE9V_@hLFb5^9M z^X@!6Jsr|12RrJ2e%&^fVzU6bu~)+od0qTB9o6*R1v)C}owIZDsM@cr`TByDU(5a3 z$8!(<^X|D{&fOhbGbitdPM-c3%@sRnyl5onzIQX~TEA`40UKzzv2Vh_XmKEH8OSdo zuU38}_MRv6`lMQ<)M`@Q3a^*#BISX|=k|Zt7$v>$x5l=^{COwEjS;}mM~=0xW+5G& zs!)Jrd?Ps7gMPkt*n~lMkK7@%`@ucBtIUx|#*G&ehDc_x>k4vSOT2ILN7Obvopn3F zG6GrIr+?1VgEzcC7^Ad!q)k8BJ)dN}VjUS!jyb4|s(?~}OOZ>(E%MKtSvaZBqhzM} z^#DJ60-nbUvM21}Hb&nL(M_ohEQ{7p#>{RT|FJwPV7u5TyDzz|JUFFJ1geyw2VF{r zWfvl>Ba~-nPv_712h8TEYMq+Jm8$JZ-$7l7|`-{|k27zHiz2 z-{Q+gWd4?iD0+2YyL_hr96JkDk@tDqdcgxbvZ`*1z zx$RCMgUS->isCwP%lAn11_7qVYWQ{ZYOlkbRH&1#iP*2In28l9*&>WYAn7Hn;^Jlw z?ywNg$me_#Hk-o zLd{_)T&#*{r?bauv5mMaI$hoo(izg{u~?DH=?tm&gh&VSU>zYJPs*afbx|R&qqY(x zigl{Z0UfgoMXmf!dYR^sf1k(7Nm(ITA*s=ipsJ-YpgY>aSB|@lX6a(AU05lv_dE=y zD_w3-Em2$&8e~`oQ_Cq7+#!XaZ#%z1x;zp-7HZ4$lv>Q$?e<4`++OVMb58 z7%c|ok#S;N@CG{r^sNsKS++aGY*K0N>!dNYnk?mEQC!1BZMwX0&Vmm?V+WMWwWwWi znAF6Saysbsa~3S3v}^qPJqG@ZCgTJ-1$IY^3>xSFX&WY%1=dPZVc$6`9OSmI$RsZorwp46x3kZf)W;Lj@mBxN5D&XYPx!ihwo}e*g ztLJ%OwzwG96m^t)c*MwE&9i48GED1?XcZt$-v=aKb8<=fmi zDZkswncYe$06v?kxQoJz!bJku`$sNEBkcZvB1j9;Vx)xClV%x|>KWQR8pY5LMsA2~ zrE&eJktA{plu14ZLvesaFV|Td7VsJHRR+JZ00g^Rnc=hP=QIA(5vdb5c1RecPTJ^Z zsEstyz?#cXP%$S#zQOe|^mB7&CtRL;&ea#jkCNBeQaQ5mQr+0HFc3d3iA;uu=yZ9r zq%#J(yzWmwsoF5I8@vN^Y9A&B?jE^UL%{d$M_Set*&nr=NXZy-%F0YvQwpG%si5NW z>1AXo&s>z#Mr+Y7FA@}pp{gsBlNs`1{H!FX@a>>8Luz8oadplF*^SV&?f z)w=Bk=`D}~%%dBPeB{BFkM)^{IP?=N~1~~!DFNEE&k^aI_9e=Z9#rXLl79&$P zcXb8X;>JCJ2d%7MIi)D2NHIvMeeH_)Jg1XR^FGfr$`X0Yrb7#atVz{M7IU(_u1~3# z8Z@Vfo>RcdhQ$5;8EG@BLw944O{5IQE+SqX!_DRS|f{|6*95tasKXWu~M| zw#qBb>A)B^5Ns`<<8yB{0xOe8Ei5UE^R~dI;@$Te$x1hdO@S3`Hc*Q76wE@LS^Kj~ zu0=iS&y`)WHrW!XDJh;V&zPJYna6EcLjoUnxJRtGggrW}CedXltk5wk{+qwevnB3h z#G7yBKzu+sfadTJL|3-1FHMWzgW&``-2Zg>Urw zluyhGFlcVgz9xJaW0a*$NS;-~&7$whp7{6Ei^Cs4r6o>KnFGq-$34g zJUX8kD*~ltF5+OSU1>Sv}uF{o?SN-H>?R&0reRDc6}HF;rIXM>DIw z^S9NW*GBK*8s)nu=YViZMC&x23b-i1&~%3+P0~K(=0IIpuK2jjMWM-1Gt3T1Lyh2? zYx|W^owhAlZX2#Jv|H6deRHRHGp8nOdE{*=Rt$~c#15JkJLY)M-6MC%&RE>zC^~{H z`LLq?AM7UCZk!LRvT`YRPzoT&Dx~6!%2K)$Le(e1tt5W@?(lAC^RI&N^dQ*^^qab% zridb5p9j98=0KmvhVWKW#5*v)i)+xdg7V`I-?M%<#Mxeb9%pBsr)wqklEbdi>9Yee z&#x>tS6l%kR*<2-Ncs3@w(Om30%13fV;Q(U&9%m9LAddVb-m~M{cnt>r%r$BWqu=S|X6$8q~X&PbnaV&8FhEiT8(^S+HPV+EVCT z6kExy0ows(&RpcJYK}M%Y?NP^P#*-zj(W*;uT3%ek^LU9i8tN_%n(*{U_TseObMP^ zjpW=DW~*|_5c><;MTfz*cL6n7eZ&pvuHbn`13wcMao0^s=FBr_(&bIQTfydc2y?|o zPBJ7Qjq=Uvm7F?JpGT`tiJ(cdSa4QTz&T1vxY?XSZZ}Efm8SBJ#(-{leCBWwYqGaiWaD&owbT;tF%i7dvxCm_*_k5bXi>v%5#` zXxCrb)o%{$af12V#DfDP|X6z~>-n~TC5I0SS!Cv&psg5X}C zYO=_)A?gNn1ezTc=U6>^@bJ{0*>~U+(6BOeqi_6DdDE6L*liFHOdC)e|q)Jj$3A^R_K?M_=gU>;yT>41$VGRc%PJzHn z24RJeQUCK%#8lgC>+>mi_gyhTyk{_Kwor;3is(>PHeG%?^gL|(ZlRw9x5?JUw34M` z;sr^Oc)V;(gJ+Ll59yO#jnQ+jh&TFmfqz>rdoay)dVPbYZ8*?D8#8DKmmD%fC;OW> zfAl)Sq+D4gOo{E9z1VA^AT9a{H1a;=t_aKJEe&e|IoOBX{=iPM-}kJh&!cVXt-vZE zEa~%DFIeyYAS!j7a~7HxOysKvlk^L-mPA{;mGAA@mQBjowUfKCooTSLGlwWe4Mi%c zxHIow5tl$&3FO%LW%LTp3Hs(#q;KA&M6Ty}K`jRrLO1%=sIhJ+o0l1s#$6RuNAH(C%%1z~*rBao}z-UhV_yEZnD*5KdW~AE6jFeIesFdGE#m&nI>z1Dp=)Gz=E%H5~ z-BVZl6w4ZYuPIXmKny;uWm>z;s4!@<1r1S&yiWQr*W`<+2eKvM*`9cdL9;6uZ#oma zO}uBy)~Iq>QS|j0OMNf6pjnh1G8j#TQ_wJY5N6+HhW?L5wn-JY4T2cbBp5C0VQDF% z7A$B|b82}@;hzYy5+WrVCX69lBJ@! z#DJ7V0f+@u07dvIRj*uI7QIisJqq4MTANHsO-P}rNmU=wJ+)#clH6p@1Yv?SSy^lCeYf7Yt&Z7%J^sO?WGQja&AUhGnXgC<( za1c=J5cG>5{z+wPQ`jVG-FSD9!2;=mGW@Y=F)awG*LuCr&?8>UW}0Dp@Z%s9hV}S| z_cfpMEuP7o^Xm?gk6)Y6XRDRZl1VAnQ6!CuyQ;h*u8Bzi0fJMiWKk;j4z^K**3pJJ zpDQ~;nlw9OZ-pm`&T1TvQ|zA)d%*F`fY?8-qj%>0we6CR1swl*iOeIZuMId#tiX{& zDS(@F0~NQ+A}4&n)NXN5abMCAvfa0Wdc@5Um5MG$p`dVAbVo?9sTi_@9yv*2b{Mgs z#Zds*@%)5e<+j+?6)ZOw(eKa z7&25vB>Ha{x6$*wutpLuFe;K1)jSvfB8T-2o4#l}z9t85VYS4&Zwy(qertj)A+Xze zuMAAZ)nWPE?8urBq||E(F)QA<7?nc<#~!Q>{iJKJvcl@9e}5zK53jm(Z0`8ar}ufB z^ZjH73LIgD*0mKn>p$N;>_Dm^Pm4x?tL!`S+ zuvwv>Y?QZ%&xa*?RmPwY@$G3_Bk`76PP4q%+r@Ah22MuAw_t^n(f@jDk_uG|(9W1- z^)kX5A|8ti1&(L3hWf5vcFnFie%T-{c%Pk>$olVZ)y%T2Cc%qWZXG9kW=XclM2DtRAwycBC^fns)T$SeDq3q$l`8f09#C>^k{%30 z;g>AkhMRsI^ z6p4|rsv6%Z^(ytkR|l3byMSlo$Jl>5d_T*L|x+f$g}({?G?!>9|xwK&7`o02bif3-h}*6`3?2q*j5kZRbJqw!)3t}}_TMer{D1rqgE7Hs^%=F>`|*Fj>ah$c zy$cdSkriv+^VBFEzaSmDY`WxC{9e%V?UU^V@d>zSz7KbNcl%%G#?R6gLTV3BUAp(n z-Wm*ADh5xilaJZq-&VRD>?z)X})_yKd4Va6a0=@wnv+mC{%} zkHC2_Z9`-u-2+M{=YX65ws=uD$3zOPKH`gAMCF0Nx#p^6$01}P$efI*f#H->?(DI( zDQvLfzS&~JkWERCEEhM;*e33zcR-Q!hu7pEK9IFVReGILWkHthDt~uOb@+wZZTwDh zRg6`Ki{)D-IimBS=XvJ$T!j$xj5vUkfhl##YsE}Op~B*>dOzyEK-Rrt^=G@RoKiid zfM(zv>i@I%CU8xi>HfG!JRx~8WFwdp0YxGhq{?EbU>j{`I_-3(v){S@-@UWlPMO8I zvrnd-&PziQtKag`u@>Q@sF>^qX*y6=43T65(hQJz0fTAkFu9DqLY(4=N> zAm$Qf(cr(oI^D6z80=?mqx2YTs6(%<`Gxl_xmkJ~`qsD4lT}Xag)gv>6q_g|m^T}! zsAkt~;+sLO!cH$}P6DMfjoLl$VD|b=`Vw=Ko=+~)w-wM9O|X=AM6-=<(k|g$kT%Tu zOfmu}uyNT3=&BCXCWPbpIGbfb*+eam+i3!K0eMir_^4O3y~m zFJ5l+>;e-em0l7MD>=hAJ)a5tOOY}aQeDe|q!$dG;KpFnxGVBjAs9Tu@`?KcPmAp< zZ-#^gD?wvEwJR*Ri5BngAHFlzjGQaKIrnd}Yz)b^SfKTk65MO6vFbKnU`!#-roni` z$`s=Pwq$LEdThAX>AF|7!*}GWQaCSxU0*c&WSo}3c_SZ77av#f%-Hzj!8>(i!5AvXH*UGnOs z-+NE5O7U4D!rtWbCK!Y*{50TZs`WYLoxm@5f9P_bxjAjUAlZAT@}WzQXNPYsi4DFr z?Y$1)U9%oW9DLQjTFOAQ4i)i^&{78ey?!zM?;lwg_dZvI6D%dXhZGfbC%sd#M~xJW zM^ruInzZ@Q!gG&*Y%o$UW4V#a-0AP^e4!guI@?Ou6`H9%9f-Fn9_-hjL zoHB9$~OoPll>#%Dtr9OxxnD@D90bM!(cF{*-yXOD7B8uI4v|-Vlaig zX5~3?hljC0WGkB3m?gFcDtWJa)~L$`y87U<$W6RVFCAXXcEdkcdEA#KKuIswheMkI zOcZ}Y9Y(I~*6VU8=&w$({U5Ph#u0Y>w z<;++?u~^q7O$tq73cYstj!ZiBkO1Nkp<_Ae*XkSIw|0{}S4ItsPv`Kvby3%6pL#9r zTNz*PqOXDOg7j3MKuvF<*CyFvvJ(nU7xK=z)hTW(dPtrW8_;LcJETQ3myLC}9y=s3 zfPAoAI{Tlo??=Cw7a#t8vw@RF^?r*Sb>hg)=N7xkc}m$#ktQmtN_c?Aodi=9Q)xq3 z1`v;Q(3t0LprdD)K2weTZgdUu;_zM%6BP;WIDNoyl`k`QzHVdg(b2QYBK4|badG%3 zU>cnUCfKsbN&*rmL$k)Co9hj#zB%b7l_ww<}m)epyKU6TWC^|M?em3~n!8C{(JnHY~cdUHg`CfBH(IU)* zEfyE@b*c12#ZpoYalvdl6LC@Ey(-8c!~ROxT?u&V7fo3hG$NbN&=B(M6&~&roDed4 zMSxmq-HOj?4JDT5dVAD`{EFaYc`_)dS)r)j?(psJCY~#D_`YMcU^e5zYMi4_x#6O3 ziQ7q6^VW3W<*JX!LMPsupdNTYQb8i6OrS_C6}85vNxe)yUsNj7wJYvHKdco@&E(VK za<|xVDdR>YWn{pb+6)Va>mvjHQgFx#7Nh=>_svDb42vR{v`@*B7p6Qb(*i_mDP z5~!#oCWp5uaKCz)=N;w%o$kJSvR?IB*z%dWbK<2lHhbKf(G@YG62^8MIbO3Jh@^Ml zpY}aZGctZyz3~)T{lXxl&;l8oDJ8V|+(zS75_4XWBoFN5HfE(K>cYizc`ye591R&RD6!B@2QH}tj+g5XzUgj8N!c53A15oE zcw5P_Soai4xsHPO8ihqMvAl(XZR4(o?3dNcR|@icdihIu_2Ptay-b3j5w7e|-HoUb zI(n5KFA&G|emu~@N1S#SYDv=yVx5oXw8|(;qFj5}M)i8q;#SJbmTc!IF>8D_dS-$k z$d_&Kx(W(vWjvO*L=DIi-4+E#?gpJV$6oS9cHA{|S-=hlB{=)1pi$)JrU++@N zTNJrTMVZ_jiBJ`q77D(TUXPfFmUkbchC+R zQ|~dBXnyXX zi^KcW9rWD8WSM`tTdHK4e{nd{?6eBw$tm7O&!U+R6`Axrp7Bf5C$$P2Jlj;68UqcD zXuXel@782!bZ1HS)HuIXdfBWlFL-WqXg7a{@2$z@ZU%bw*xmF>Py9FJ^XJZ;`zwwz z%wpSe{em+O1p3mx^1z!|N(685E{QFSE@qKj<|rqkO! zh%@jfZ*k0dy)8yN?u46M%1CMHws!S%i61&~svAoX08XGEIn8xlWANr`5 zf3?d0^t_#F^w^h4uQSHnUZzgEUog^c7hFi&Fk$y4f98bkzyI<0ua@Z8EYrJ*h&y=__k9mw&u# z?r=I!J8^~zOXq1*=TFRbGHI*2R7s|ETUh4|@Z-k2HjHynqjI(%tZnkcw2`pQ;fU-1 zS@^QcSIrp<$x7%=%MNH$Vko#lZi@F}c=&`eRx%=yzsGCotjKsHqopp)Y%GgiXKyA2 zgJn)TZ}Ay9YTwsn}n?5?ubeu_NXpEi9Coa}5*6LLSQ_&fV`D;ibTS4y+FW@!ul9*mP}jTZQvQH50+rkqW$A@q%0lVz2=6;ZO%o4M+_Dqq~}iTq-6%GLF(ALF8?> z?r?mjosJh=52{z}B$!240WEB?7rZe=Gn=k;HOSg&8>O6sK=)Kw+JeK^8$1qLD>N*c z0K7clyfh|0qX(=a*hnXWYtjk{s?~Ilo$I=v@z>|_MxQa2U2n7hj0edIa$bFKYR~C| zK4zI?cqQZlNpa%nP??1k$)l9e7AKR6YNsoNm1K{mMx7_i15-wa9@}QwM$J);k#j9m zgC5UTI@%*g+5yr7YU%T3*yDUvaIy@Xtr;2vFi+!e9RG7noPZ0AS$^$yg>V04oM1d^ z>aN=-X){{x{J`%pna^#9a^BwofdKP>gqBs5GL9lEsi^dc=@WtNA}Bu;l;B>Ec)?2k zHZdAYj)6)4vJCcVBZLMU>d#$vMf1(LNuy{K2}B5*H7>6Prma{h&I;-QlV95fxg>8FW3e;LwJE`Bb;C*S#H5 zY4IAQ%FYw)4#1RYw5nQZ)IIxLPKK=ybkw%fznC`4j=^@G{wW;tshV@}7sKHmcj8bE z3-|bfuV0X6Lr3y5O@%Nc{DybE@QS=REM`I;?^6;Tv{TV8)~kBm3wT>6ogFzvn=3pV zf`vgQeZ_a$-!*1^<51K`t4Wa)Yp3-V8bl4Hgkr$GR8)f;^BL0wSSz6`aBJ5XyNh?o zmqFQ|@eu9JOCbmP3o5|N((YXrc@=U6S|CqA51eb*cmNs8O;vy?ZgaDN33Pp6imE|g z4<>InKSrtB;o2&^Ln_3TQ!~x|{BXqZJSkjZF_`#Zzc$Ev+jxoFX7ovXgJ|ip z*htdBO=)%F&1i*18p{Jp*-MeTRMbNMJ?c|XuF*o0J#_d7I_f?p1~7KE(d%6o${V1P z1HwV_D)6qD9yTVpr15ekxF@ajOat=2J?ipT(ja*Rz||hos=B6%4Sw)a&NRI$me)#G zh%eE3qVqiDCB|yRfpbAcP-J#ws_$X)P>~j@S0#jB<~Na;P+bxeD=C#dWU3U4rgq3b z^D^PPLYN((Yhijw+q42vbWpCem2MTMxy5@{Md&UX>)8$xy$9Zazq0EwZ}_X;fZx<+ zz5)IXi^y^UE*o{R#A%aR{_(rNy!X84^*6(+oi{OH?j2XAD>9G7E3I@=S&S2+v!!w0U2~FNpj1OBnvyhU4coCtt45t4}Qj2 z40tj0m}HUEAY2u!yBNC0Cx5JNi^p+j)Rx4oi#!&X?sYm)chPI37f$dHdXHyMNbqQb zzQN_=v*=VvOgrjx)JLz{;`NDNpGzA96@`*ZUb-`L7R@ONIHJN?uKAQG;|1CD)#-W_ zb_j$vJZs#x&+27Nrx}`}NL>zZK~NVBo1WWoafYNg9P7o}m?fU^0%WS|b4i@DVUAu^ zB}kl<1hMF5S*h%hK$ql(U3kt(oA`Y$hax)YyJWk-X@Ae$mKplZ5VyZZ{rdM&fB4(a zfBW8FrHd)$0*XYBaLvh+zkrhO2>vzAye*bZ{a^{nb7I}sS%oG z-HMb@-Jx;u0^CH7k$U+hCW(JmTn3txE&%}e2|u1!59D_h>Yb7lH{er9Vro1Njcak! zCHRzwJyaw_7!RPpB!|~SVg={ODstaDgFXUvu(=Y`S)50M$5JfsV^T*vK-(wDI2xYo6wRl}}+8hb7e z*Clw`^W2ulX{FBZ?w$XObw?Oym5_v`$BO|pASB!XdkzgTj9SNwGCdB^2+uYK`_c8f zmrPAJwv_9NXxAJjTf@=>6~Sv}q|$3!GwwBNt9*8aHvga8m5RZstR$8fE9nHWmrdV+ zCP4K-Hw9LiUKQu}rFRt#Ipa{(7B>qMCkp)9d~1E{gUsBv|<04p%cHAkqMQ~b8C4A=WKUh;;K1nF$ z#1U;4`paIjI`U3loA`U6e@=&^<3vlFMY-6oJ|MucI+*;4fj^o__ z+C2Yz{T}5%?!8SuePKMx*DchTbCmK7MNUysCxZ=;QnhVK%7R;1&)% zFWaPrV8{FKqdAF|CN%b(`K+iaM+w>wL3?~!gU#Z6N}%GtbuKZN@VOHUC6+AXBCkxZ z?X#}Qj|oqUugo~AS@OLrGqAjMo~X+!fA)Q8hSWf+WE)=3o{$yRHRGXTBRqf2=e~43 zkd~Frxi39UdfZd#Os{=X{Y;3Tg{_&DL2r*d_-Yb!E9%OO-SC8NucixF$x?&%D~?Z0 zj9lscEO1z?2y-)z}JBo1tY7YsfT*O~4EM;g;eF z<~``0ykGv7Y;a=5cDKdmm`^Fey_kh`W-Ywq6VLl@m0osVD9GVu@snkHJP#}P%QnIK zE*smV{oJEURztaTgu;(q?6IaNCSr;;3zP z*Z{C;Ik`W?Y3HPhR@`Zqnc*V7DgFbAA49fTtV$ZC1RW-sirV11_SG`gB`>UcyBJzU zr%IqA0HZJ!@as#Jz$r%i=DT7mhvM+HQ?M}5)?X9%f98M@LtBXf6FBIQ(@Hq64l{M% z`F{tPA#&igRb?c@i6u&n1$K8(N?^*|Mn%;`ZWA}lur3L!UXz%^Bu3c-?vAqr7{?;f z2N&nQP2A^_Ksv-}q4zyur@T|N+iD7fe0yZZo<#CD}qoUFTEy2jqk4kN>q>VB6 z6xkn`52Y;0<8mcfa9u$|n~_#}zG$KQRY*I8xMLG27e;*si}zFMUXKh-HhtfHp7sdD zD|6j+t3(xoPT;G{rYnLo=_-(ttwOBHToi~rd8ohix*LEaFCY~(Ce(P)5u7I~@;Xjq zPz^mkaNEQMzyW@C!_^>A8|;#;@Z*N7(#Ti8XB~HP+6)<%7!xL}ZhyP^yZ7c8vt%%O ziUI8#a?I!g_@srej^MDcw!Ucg=eR+GeI9$?;RcPL{x#;C)(xV$SliCK2eb9LPJuL- z@`>|BX%figMD;6Il1gK{zBI=-x3OMpTQOuk&e7*SJNin8Q$y#(s9>p~yY~9_9~b;+ z@6WqumwcyW_WakceB<(4OJB>FbN-zhvp;)h{hat8^u0d+dne|+fB9E8zxlqS3?!>{ zwG|$ZcM_`!9CW%ft+&}mr|kkB_K}Am8T_~OjT3K_y){Db&g!3T!>hC!F*2u|q_y+@ ztZrv7?)qwxm_f012u6Fb@fDF)W^b#pmW1Ep`evr|8ZnGA|fpZ88FPRJMfzqbX9T(n`EWu2~lE?c1f zHb#kz@}yGJ0V68K+r&qBEg>e44baDDYI6M1c~dVw1qCnA!2q}wTj_#O?6_unsGsrd zS%S)jt@E)0V&9{IuzBK`ee3+i50~8^&Tizy6c8-^jaEP}lYCwd?*N@mZ_zes(|8$V zrRU1A$K+MAde>8Qrz1;4o>&=9yU`${9d{`{BJD5MzI&(jgs=n^4n?3>tH-@bdj$CN zl9-HFVwB~2$R&CwnqU+lfU4ONrSA^{&7&>fWns`18jmh%C<?xRivH2@~EPez&$3>i{( zf)3j>fBVg}eZw(xoY=%>F>_2N>$b3Ea^pUPMeljOdC(>rLOE$XY7LECZ?{a(KEunpo%p=jdP4;Qw43 z-sIo|9ekCO*mm9@)aY7$`ghHM_*46zE|Vi;$TbT8$PBR^1Y>tXasTDNdZPKFG8;=^bXI?oGiAQKf)X^P^^RoUYZW^B`cSEKr zyau#!5noO>-M!$yecT;hqF=e&HTO-j67^XgwDp#p;%$=M0&f_eH)*%f{c|TP>*Yuw zrB~tIeE!-gHe=NWc(Ez~L%CrqDmlT+d2LRwGq-*rF~ciexTb}~I8f7j1Hl_b{%d;l!KC;ul|5}-Qqa2?s1849lUKb>l=V*3F zKUXZAb&@;~;AbMK_U)dvEFw*CMcOXRq?@#7g3bme^FCKZi@L|{EsC1_x2yW`W;x^k`||h5y5}q` zU1A|kwopniqBj9qxUhQS#QlC%o9^Q&yG@fyO4afOeCNPv?ee5Fl8*x|GG>pp+AhRO@#onrx! zi8rA=40&2ix%4LI8$zfBn!{sl_EDcS(OgX5->KfIKIfJNZUqyqvtBiK?uKzOK(XmK zku*#Y*#wB8JoNgCp6^*(M$c6=k;PiN2JOTDeTk`>eOu9_&493BDN`b>5!TDI-D*g+ zV9}Tjs!gblc*QY`$Dr@uaZ5jNS%v_Efx{hk%q~lQ%y0jbr+M4?VfDsSWc3&V1y}=I zPmG z!OH|U-8W8*7C8Rp58=h59Cz3Y@1L0c#Sa#Caa!iM*wRX^by9-U>aHva2wlLe6UQVm zt-{05um%E&(bR247k!DhG_X5-y=FbhR-aO*OPYflwX5F9`&!bQH~**W`?^G`YxW6k zmgeg8-dP*HmV~a6I1bE11aZGY(vLa2e{uuy;-9>I&AJPK(_)3CJ;8&}4*J;aIDS*W zhG4zw$~W4gx@OP&MbF!PE~}Y&qV0-!C?f20*#z20Hl65~8lppjrs8m{1xsQKuUCHy z+IUPU1nU1rUK_7sVv0{+XicClH6)eJqBlxbk3Hws&Fr0_S1pqt1;&})S&95qx=Vg- z=2-@77E|f_fk}c+AeYi_xB7RM>wXTV|~tv#6DYUcfX_Mz2&NMf**3JgE>aiWL>c(Qxv!p!)i;f*85{4-e9Jar4k_gVrh;t<__BCsl z42$Q(K=dS9v{tf@UlnnHF7(NVLjea=bYf- zyuS6^Kfn8>b=~p{VrR2dIadazNm5nKV%_@5eIc7Y9=ddd6}aK&IdP)jW)EbWE0sBz zSk3~J+z@Ut)S)Mv{@p_V(QG#Tso~wJRwON!h9KOXcu>M(X} zN~Mw90O)KLx(dFYw^NNx>P@Xpdfc%MC@>KQ=X&a)?xW8IgP3KiZQv75l$<0t-B$u7 zPzD%?d1@qpzobZ(=?Wqa@(wFILNe)6$?j>Fr{ax+iVhl^2K2df$&+OroumP9rnx#z_TyD3uXkk|)Y%z&hrp+Is+X&D>itw zXRoL33LQTseoCLqqA9C`k*&c1J&$VCNYY_^@|?I;n8aX){6xZ!JRoJg4W;A_35FD^lqWu%q;<8&cbs&*t~Z`XDC& zu+C!t8|=E0c7X{V3F9`bh zE@{#>@~Q&wgje(G-KzyXAuS;Xgk|1bAbcRlO>P0@#A{_IB zk)d^`Itwh9kV`+F0v3a!j&PDBh{ku%xf0mFmCl)Cc79yp9Qf{S4E)|u`TguCy zP)v6T3e@ed)(7|bmql#SG-}bXyfJ>!tctJ#)KmRi$q6lH-&XK9dX>8s(`$U%0SyJ} z^U|x+(;+korTD6N!Hx0x6Z-ref(yhv7xQE8o` zjIIexVs6u)s*)zxLvYX`z%j4ou<*kH4_qH*$Is5o2Upt|MO@*rUz6?DEL)_x5(GtT zTf>@VD?IvKu)i1b4eJskD}tkga;3%U)xj75h7$EJ#48wM`=%U7g=_=|RU%Y(IBWri z_4eVH%=N9>`aJvFAI#Kr+T@tUWj~F5w#&?Ui}QfLf1s~q9i>d7$SNx84yh1hk5FT2 zYX)pHx`U+9s}(4Ic8BRUdld36d7XJ> zB<3=|BLEy`&}jk_{U5rt0SN~_2Wu?{sp4B9sU@E|zW#ft3vaIlW}AmyiVU_J{=lX2 z#&F0NoLI83kS{C^tQ8eXy68KQ`u$H+_GYVir*yZbo4>`QPQ64>rE!=BHQ1VQ!O>vj z9d&Ke_J3{_ngz(5Q)@pV%elpdomf%XVj*slDdk#2X%>{M^+PvtwM5miJC@L zqNYNuiu{%5VEoEtctIPlwS0gg{8y1*!BjT&gC!)7n_zNcedeHrVA?||cTr>~6}2ZKQL@;*4jKx@`F-hK?VT#k zq&t0@v^l(m%*|=FBo*#%@cZ2TM!*4j*Q|DJRm2*f#qOUeD?D0-b>3@y9=cR}W0`$2 z6yI0#utMdEw`qK;B>vS}@^~y~X=AxG1J3M>o&hrt@9^!hZv^no=y3^fnNJpRLyr^d zJew`hlSC<3Q6!Fv!it3^MaIY%OqWick0-)FBLnKHJ5FzY=Q*aUiw` ztj(k6DhR}41a36o(%rShAKYdJr;# zlEEejJkm}bxaL+!7XGRuT1@^s+XxHRf6oPU^<@|R}W_M zQm9g_lkOK}Lp6P$OKD`L#z3d@mqiqZcSY!$nOMQLaf_xL(x7QGTx4;B%Ol)8`@m}X zxt_AaWe~^QsJ*3md!P4Ir|7ZSbd9=PkTkYSxYajl?44U)6{0n~V{!+5 z!y9uUmoX{KVlA*TY6_tD5Rh{SGc^_T4&QF5KP~Xse)5bgM9w$hQezWJN z-+s-!72W5j|AbUH@m2&Khz4ZaAElIsDN;v8oe8?(wf3duq*Rt4`W~i0RC~5DRmM&s z2BBWnN-Bc;?{8CV;?*hQcyUnag?EiiuVF)wLAC(u*}KrQp5P8U zPJp8RdDK%k$qcB4fA(2PQePMw85ExZ@Y_x)fi)-xsf4kV5B;u}`Pg9a8IG|urr(50v)JEWfmUG?5J zW!X!ZH@DNR%ibkp_JIP5nrAI8xBJ{gO~i^b-9r1NZuDN|`}H zt93JddMxCJAphTHps&`eKA~?&9}0F!q2;A1{%@+Gm@nEGP^Sh_2da#KbUjFX>)KZejQb6-B|GjarCpaVMK|Nr_P4}5Hx^BF0WQ_ z#>7uqK1FBZz{sB3>v0h}x*}a_p_hpU!NC|e0dHtX=LFgzJ-quA&;8G4VUl(i2s3Ut zu{OKXf`sKErMyRx9-wBIEfFAtA$nTAROB!lJ@u+XvNYF5DF4Z%%Oa8H)+FR~NW5Zp z!|=Q4&!Lqs5@GbYn2e$-@YpuXj?p?~Jwc+}D$Qvzj_h-3rz?ap%3KLLjIoh8nACja zx~L?31NFdjabEgRkt=DIT@EQAC2lFq$#H4W`}LAnLijx(U5N=bb>&+jdH@c<>p}Gj zy(*Tc%N2FHRs?;k+L&f#ekj^mP&17QHA#?kQz}pN1J8gWEwr1Dv@iHyx8juF z4!4DFPD{hWO?XcKt8S@vJ)_e)`B*9x5fo{Hr6fPRm#;e+Y>**uH|p&eUg{y{8@Up^ zH3Uz~kbc;l1WtGa~ne|9s) z;d=vXCUzJ*if3p_yz6DU49(fVwNtP;xDF8&=hx*mckbLP@*G}5_-O)TPy+SHT0Lsi zAH^|oAg1h#F3uQwQcLTDzo#YS(e10g=F8~l#Q>*mN#$XRCaxg+naY<%ji?zd3&u1-nm#tPMlcD z!c%_=3PiR8S3UB3H)*d2)$+?HRr8i>*L!2VNNG?ZvuqYr9w}mH4LqTxr4yb7O?kkBm+JNnaRGFyn5(wG3MQ5Z9?E_vgU;e4(+hmd$K4cWK^b6QMVLW zHc}FK%)3Gy%c~F{^M2@}S03}$?ISG&V*|&$&&y(Yz2iRN-yt^w>Jj5h#cm`y#_?woRY?*Iw%=qSG2NEOEg`RjdG7-i6!_Nee9&Z}z_?LMUX4 zun*Aw%QI!qE4V9$7 zf(QM~-9|QAj9uw!CZ7{0flY${wK3%rkz3UMaIpF%;^|NjGRQP`zr+n8E51psvyL-A zY5F>`BEk}T#!e3R=!M|1+b`*woaVMZv>~_x;u1R*oe)>f6Rmmkf^^ZCqS<-=X#Xr8 z^U&oB)=&8>T}v>@tD*AJ40+JU>YcA9v|Bm zPwZg&==h;mt&5;ps)x7pv*>y7Ry#=*MDg`}Y#-f97b*_Won8qMUwZEwfhI)*Ukitq zJ#d)AkJ(@LAoiy7{v`kWt6QIYm^Ynw&yn{?!V9C%0vq9gEh?Q-f`OkxMd3zN;@u#} zs1G!E39qM<+_FJ|)nR`;RANj?xqAZ_!|iN(uUivjNsUN`CI_fL79sZ6TAujntkOjP z*qRQEi)HM@zF3wPSpCU1$aP*h6U&4!f|yOO@u-K|kWSYmX0Hk(g{aQviZ1iJnGWN< z;_%1!mO7573RY-nSTV!qi50fr1?}jFts1jDITZELYEtyV$dh^tc~V0uD=D%U6SJCS z#pIkAs@wBrSb$tP6DhMX(wxLJ$Xgi9h6K{|fW_{RUlA?BN1ztoRAQe^SMx9g)GR9i z9{$s$$6bdRmwMGr5PEpMJES7Pd{H9XB`SRfi8f(AJp z7NW+rc>T`WWzR+6IB^0Li-SK2QV*cT)>-jc@fV~`b=~t4eb)nFG`s~K$|~3hilIzB ztTF$Slyu*^L#xv|R9L#PHpnXjv-rAf-)#D3c!xaKP4}tll2;lphnFTmWns z#w)jpfyTCi#)A(>00A~Gxz*2nSgUE%U`{&r=m@Frb!*qG(*lIW*_tLOi@eOAPoiht znpo`HL;BC!cr}4>e&?mBlC;ol`tr>75Cgq?8am!`MIH2c-j$$M$Uv!6I2bHL0}A5C zru823L!5vz>aT%s{lr?5ydc{Pi`=OYM+=TYUrZD#opcL>{5{~2)gBit4O}?>xNl5g zg?OX(;pF~14u(>Uk#(B2(@GK+-is5Fm)%Mfo0LXoU9@19idy98 zb3N+16;#a;TH*e6vVEodM~uuTC&4kgnEJzI>n1Kw`d^(``($a=VoHD1t1tx?a*r_o z{{M%xdt7nba56I382ByhI61ersmeN&#A#&(ONxsr6V@mxnq?T`GblRbBcXT>e8qVD z{4M_=&fZbpolyy+%|4KYOM>=8=W2HR3^IWe3y_T#K9D#{xsoEws3=o%E9Olbvl219 zH^Li?2}!o##~O1ZasJZ_FMs2CuTe)KQ5Lx*6jRj=^n6kLcqkDr^gNHBMgLsE3JZ0i(sreYKLL_eGk-Kuw{uh*D+ zw-o#Af$Zt-kX~qGxg_)yG!4q0TFr}zhzA=vCL(2)5o1FoDzyb7R)+3SPC28Bn}pc< z#eM(hT|;dZt`DpiA{``F4|jQH(M4%C{lNR`^pTbyHsgZx$_{S;CtOg6Tz~BK zTl3Zv_`C7r$>lL*p+$+z9ZK0vkxoq2KqjMdH!Ov@?sSGDoZxhfdr??ObS{B)HrPn2INdkeGrnH z54l5?lgciBf`Hw82OSCg0}Q zDMQp}nFBd#p^E}Dppr0RhdHjbdWcnlEZ(GbyVHH-aE zIiYv-uZsWol)wyEaK}Y=#V@R z>(g^Zy8itIo22QI1F$@>RoF{syi%=cr*SHP7qVv99X~?`j0Jk^AM)V_m@WLUH=lQf zr0hEvq)l4nflKv0OcrV8^Y%uTxHV}DUim^^9(;5fgp4Pg^x4PHi)f|KM{FjaMWpjf z`Bxl}VgJN0erOlYIn?x;s%=I;iA7j@AL$?pDYL{fspWVajpcLYlAZnyB4BR4_PY4l(p@% znt@=#0XP_Z1-Xip!Dq4mRZi#_^}GCYiLPcv<-p5TACZMGOakO43$-SZQYKI&mWncJ zE1mQ`URhWd!4@4DL8_uF#8rw#Q%4Lv=5@CjME1wYQ)m3$udn@wbp?jgh6!0pLmn_E zBIk>uMPRzLE8<;|Ce}#Pw$HOmo=tc2DP3Ciq6kbm<2}a zg#Y~`N#Z6joYyBvfLU~a3ni0Mf&@vWq6X&D>k=hpktxhw1%L|W#&*TgX?-qOe1YZ> z=1PMp)CS2)O=KIeWmlCjeq0Jc5y2WRBc{6Chu@9g-%`tYkKvYE)oue^Qucgw-vV)?Y zkl4U|f>j=C=Oi&pjFF(Va}2_3LR}AuQML*(aCmTB{*+A3SyB{vvirD33brDvvuWeCEhujOk;CZ z-QzETd>KWZqFn*JEK5S+LH9ICiv}4uEZg}y5^_&LBipSB8s3w9~9dTflcQPm(%V%nH{&|udfTRNpX{I1VB;3fra z_@Tm-J$2iyiNpYPEqZgB2xMPXlJc-}$dX#(+2>L)C6+hWc6PvDc1sEq#QNLEb`*UmmuhW|8EFRr1 z2m*8~Zinm+D-A0S-@z*Zdcd3RH{H*^bWFo}S-HM$>*sBSh4nZ`pL2u7Yt{6j;qW3j zv4X?Gi-6>BTSAf6d^7(*z+yNu=53pN$8)h$VjEKtTt0Eplp7(S?IJBpHt4OU#Ln!g zWfA+N`(DeHV0|l4v*c*@y@u4rZNMIusYwy7oibkl>Ec+iv^xOP^4dchoa?ToEx4O#lCilXNJum58kOG3%Erqo%aU; z&9MeXD_2v>c#6bOQKpnj1ez|D&J%4WCuMq7uBg?m)y)w*$y1lsRt!CL5(nID+~U97 zx-g!l)_k)^A@sn(njq|Cc;*!(IU26MHne|{O)qn3Qx9ttw;!a_7Es)h{ysItEGg3c zGS-l7PQ1Y!vH0y*P)g{+xr>UrIPMHM@=JjyAC%Cl*|!z{T0s39dN3)VTn^l*|hV2#i&Z?lZH06_;w6&y%mq+jEfg`covuSj4 zz)c8G;rQXo6`PQ4y{Fi6J}g{A{wF76r8V=ypT%YXElauhYZAk4t8rqJ8R`KBB*tu{ zl*trXYwGcedJ3xBJAh;oHyxyI!OF@uP-A-CcdB;&Sd)rrj!>C+hKCIcwS2))3T7 zD`?#*@uI+Z!Pc-IlB>yuz$0=(>drzj{BCGU*XI(Wj2D=c6I~V^5Bb}8AKPzl)Ew35 z(2OhlY3!V?*Cg%LFpbMZ)@%PLWYAe-(}F9$sucjb2a!LRfx zh)0#p)~iaqR|jJ#vD4?6JXexU-*@kzlg2(%JaI;hEwaakXP$)|PV#5;?B8TI&otvJ zmHAmN+3Cdi`pg1f`za-uOqEnr0c>6Q(Bh>K_O1m%K!2g)& zGp}axd|9tM))y_EbC}e_GmV}poCq5-ESKgK5z1`As z2oQA1Z}AVjbaPsxc9Huf#inToUfLmThXRCH-Y~Nxh9Y@vz=-jYX949I_c`H+di_G^ z^sg41=}JE4Y8q=FFY}SC7DJZ9cgsFwRR+CnqbM45tRpX`&icuN;ef)4l^Yg{+hy{T z!6wpURB2j;pYR`eFY?YG*>Y*qlG+9eM=t;1bHDYwv9DQki9J`jE-d_Fr$}X>ksKNuR(ga&q%VrWZk!zgw%43@>RQR(%% zHSIAcJKJeJNQ8wX`N(Tn7F{i9@QfDKL#8GY7L~cj24hCc1z{(%S-Qu61*zs?2Jm)8 zIzN3aVa!g`x} zCIC4-z`i5BMk)4K_Xc8Vs z)A(xNCMe-R{tLaT*0+tpD|UY47O!Dguis)O`<}4B{y`K1=S}kI__nujyxF1V|NHXy z$T}x>sFhgAku8)G(q=YMQC)(S%FX^)N&ly*$|#jS#k;NOX0Sa&hqy`G&5!fzqVYtXoF#O^7T9cmlV%HOqYi7lG7-W~c zdYNR%TFVs~OL)MYu3PD5kRfMsHeKL5lEH6|0?u*;9eBnG9HZP$$i`YnmboahPOQDK z#Gx@c)=0B)Bcxb^W=A z96k;Y{VD&~-sX+yC&xyeAj#Y!%g(#SLej!O^*5JNW>dgq8HFTKC&RO+UUt1H$l+zt zC&O2{9wc4T%i4eaePiIspmg{1baVLLsb6}X46marf@{bYdV}A!pmMjH#!p}x#b9B zbZe=5Zy$UuiK$h`N>&GBqins(VXt%GKV~SecHnqN9di=<&MUQd&n>yBGK&-K;h*m% z=`V~psk9I$g_Lp|Me?YqPriX3xD>yKiXv#!-6dG#)2{i9E^}WgF!|lq@=VuCB1(C+ z#%m|Em6JN@3iTTI>(Z+rT#iXVTc#74)PPuCDs2D`?*|SMEkl9+6Pu3%@bB|m(bwF} znE6n5@~>p66Wc;r7AROpDU&F&ii$$6ID;Z()?)XSZsh{hQS_QFBYAsD=uu$EzDFn9 zpPlq+Yr$o)pEi%Z?{SiIqyC-s-X|}ap)sc=XC=wyCQh7JP?9)a0@%FFgAaGuRc;dHqhtTk zW?k#U#b9<~dxxbm$e@Vx*&#LkX>!qh&a8d4)2j}8lQckgl1{H&ez{w{u*vQG%vGcv z;zVoQcYsQ>irk)nZo4%;#WaqHp3tO?<((uO)Wx$m(etPUB$j7)mEC5L4I(!NUa~jZ z_|@H;pLgC2Xi+1mhJu|cy4-CC_}WtGBdT4HYzw-l#*W*8jj!J3m9}v{dp*WYoJ`*_ zvrT4JP{cRIe<1PP6ci^GCt%(Th#saT|YBBUtjDl2A|<^`40sh1Ay}M!(pI@o49%Re_y*c9C72sc`Gd9rc1aLitbId z9s9^0sP^b)kS8~Z*#ShU2D(PwAlG5FvtD(XsS(T<<$%l4?rO;(3p^x14Kk0tFEzP- z_qUVH;^xjztX?sXBQzhK%d5O0_a0^{3x-MNa=`e7yALCyHz9*a`)xRC$ z(#@Ju%_ZKl2nS$gNYHpf!q^)aPpdh+d8c~zaQrMzELT|kELd%3Vm*NBvXj9E85Dd? z$G%kHM8ObXt_1I52ge$9zHfm@SE1gdYy)RZ1+Y(MPu<{p&8N==Su%S)u+bTe!UtA~ z=V^}s(ZWj4c;z;6mF(oW9;jfm6U6KTtARMP55)GYFPdUa#lym3V`6%&_1!oTeOTCB zFbCRLqEG6~D_i+}E;WO$_&36?d?Z#Wl#WHQ8okSJwHY8;X^6Rgh>~ zDQ;B73+yBXtPsI|YBPq}r*Oy}b_{R&?DRwHnjbE{Ip+mDVW|eHlOo$rI|RRVX}q*h z6N%4=dv;-&Z3Bx#@bJ`0|MpqT8`c>ITp;1ZdM-=iL?dqtojz$T4?2KIG0_)zM*UWK zoZ=mP^+o{F#58G-i1T?VjA3Fdu$5-fm1JZDi){!Qs9lUqpP9~|Po8C6%Hgz89+u(^ ztOu=E6pP!K@`+dg1{D~3O`dvNSbNw;VC2#zjN9RB%y~0_q|nbQ3PM`0q@c?{0O{z( zAHuSmF!0o$_3&@9?;?ZbfscM|?T@~0-EH{=5hEYK;!kE-vvgT( zB9tep)SEqkE<2=G)~?tj)744wi#`Y4 zt0iQf;8#Pznyun#IM1p~(fN{|zq*DO8u5O`UkZ7QfOnxb2-Xya}6 zXqJ`Abf{}L%gVxfWaVyaTytPc91)R)GKDLYhAf9T4m`Zvh{zXCESzF2 zgi{}-yibuYsi-^Tq2d5|?Kij@B32Rf7}tw811}eHpqu(S7r)e|$n`Ia*rYi=(cry6 z^j^!PdyvPlp1gNT-Aga>$#~_We>XoLIKJvsdQ}p$)$@p|AfiBBOX}UL1>5-Lf>z;Y z^csE+znIPsKS*xz*Q@uzH=Jj#D((xbVDCFfiWF7C{LuT}9rU%3A~n`d*OL9=T|gJt zHf@vaKD1KPt9E#;4qn02le~!4!5jSc@%INFBL+a`{y;DRA&7;uvbODuk>s9BZSgnFJ){Ux0P^Y)^ z&xsr4&~HW%Edoz9W9-m}o}zB?L-Be4K4}7z?v=y9mlk<}5lAZd-HP^*SYD58A8?7c z@y@xiQDhDQi2WkgexHM2_bs>P(bgetF4~I|2dY?t-IyiQq{V9LPk&bOT5NEM+g);i z-b8JA>+tLmS7B_0#ULx^&`Z)A@6WcbhjrTM6ibC`qI;X-%a^`X?ojkdx8N2e+)BzN zc$ULE?!VRZwmg=10E$VnArHI>T!FQorTmnLGl4Z=$keK%fvUYuQ5>GcEQj(Dy(*6` zgrbwn?v8+nttfb^Bya#8*mtQxR4(f05sh(_dFOiI@!%+V=;FM*gc~gOqF73~f+9`X|`yhjqaDZfsvjX*Z|0OdEGQbKA~ z3X~9&qF-kXp>3=!HhnbrJdg2Kw^McI{;#YIuCNN>L-YJkuc%vReKj zA zbc5%P$s2gkD;{LY9`)9dNSPYavO7d{Sb~Ss)%0sAx#Vscx;)v>jcZG2A zjB>XOyz-Mu%`uWJ%%E!mi`DgV@VZ;fbPrn9RM1yu+#_3+*oY{DHhquYZ=ax-ndhqa zYtrryyy#UyAtC_NFo20`&RxY+4fn@K<9z?nzV-$XQX(P!}~PojO3yuL2yh11yA6f zDGF$ZLZ|{hu*eG!d3AajghfGzWR|$FSd3wu*a2U|_;&W8;on!pSnE7bI)tA;&1Nep z;$a3K>NLRqG9D@2*G}0kfTRVwi1*ON^x35Ce`dSjgz~Gc*|1BR@uHg9)}}QVsp7;b z5G;l``ebV+pN6WcLNDZhcZ^J6|5cjrJX>fya|(w%ki3%ewD!+EDKwmT*;#yc2Ktg$ zgS^+{Vkl;YLL`ZR`HT*D1D&JN6$NZi-}j!|9bOiR#P9drk?kGQ&NG0iES1ie#qz3s z^Z9G1GzcGayhlS!3gc1s4xv698Mdk(_5}#L5aA`x+V<7TqiUxsgfYqtjnSufPTT>M zxMj>kMTREYt&QpC$Mfpxn23~F4syv1x~?2yGw2+SxcI$~KiX%_(BQPbIu_0aBxP-u zEp)#M`AnD{ZlHIA$yDlfn(3s^iQ9NJfqB42H(!v>$MnqQ+PfrvBrIy~$#W<<9I?pO zbF#~dLHzrp?iT+h;CXkvED1&W!7`|J)nRBSTC_pElvL37Up93VX=4tP;}fy0Nq2T! z7L6g^X4&PLn1$N?K==}?8xBWX0`gQtrvZUBpXu;<--<{i{0b=^JE<1)yG=ET+! zNX!AjqNS8_5k=-xQ8#8EeCHVp`1IE}+`CVM$8hH!&UkisjCK1>k&T(SO^`Zo4*I8) zRglkRG#jF4l|^FN@}ep01Hl%pbsf0{c-q?@p>Q&XemXL(-Zf0B9o$y?d zy**}+!Dt(b2N7}ROOI!=?-3^yk7_LZ!&PfFmWxx~i8HlXl-^2mK^jdxbeTtC30IA} zGO$4Oh>R}LFJ7>gcR~6w-b-`60?Kh4gDR(PkNT*z_@%pKJHJ$PND~vw#R!jT=Q&${z;_opGq*QI^_J&$sogoSUTO4j)Ertim6(bUzjP0-~f4hl^lDt5Ux zX-(Jn&aRgqfYuxpKxKT2w{_|<{$;l=;LN_mGnhy2o{fp~*tD*HuHxCIIgt7gLv|}Y zGc^llVaal1@Bz9(SnO+>yKHthvv9_v*J z;T817aku451GjqaB&&l_Jwe`>V`RS!^pFaEyHQ%=Bj^oY?>%3%J4{!o!J@M~DJn*o z6W-^747-VbhA^l*QeF21c+Ldx-)UJdgd#x^f_OJhj$ejzWc*+Dz67qRGhM%@ctY~U zkd0t+3Mddk5G#veL@j9d&eB;r+r2Y)nYl8H_s-0nJCp0&b{5=G5kWx(ltqEaDvKLc z5L6IrRiL0KqJl)U6pJj1DDZz@5-W*Bavob*6?e?}9yC>@PZ`Qfo;D-RPG= zVpYRqLEaLYXE6~G7qHE-fr)ln5dg7VEv*|=GkaHXMgVdVBL%dU){z~>kaSf8_{q?C ziX&@Qw)WeMvbMu{z~w>Y@QA=&rDd@92)sJZXV zGBD;2X2I*&Eo*N_)&?1Guh{J0pTCmKqk`fES2z{Y1ZFMpSDhxe=ADytb5Vr?1Wy!q zd0nL3t&&vHo$}4xtXaDEWs`z1vk>)U71F)3AaYC%MGgjRHExf#S-WAP;W^toM>=S? z)vv`r*$ZPN*+Q`6mqZ1e7W(p{$u=KmPqe8SvC?r?rsJjh!9_px(|a=O4EfXx-gR4?{IleiC-~=N9sb3y_NjN)g|4bxo_8pG=$ud z{pB-pLkOq`NVS6VE*BxU`-T6eVC)gfgnUD(K;s3}$^NH?c@QF;Zh(>z2EeX5Z-h!& z)E`I}?v+?*KFyvcz5hX4n$tN@21XHchXt6n4%-=8PdbEalA-$0_zryFgY7 zb7$c(WPMGhQA^|TnS(CX5_P6ipR^qO(g8)e+s!$tAj2`ieUkx1E&L5I7<(dbt{(ZK zXd+d-HoUzu!G@S6?zE2ydPpBtVk1BaAFE(si?e@bztdgC11aDKc1s;}1mAiIE8_`c z3l8J^t@ZxX&w~C%thXzvf^E$tYP21;4co;W1M87QF^Lq}L4|byQFOIT16A8uZdaX4 z_{ns(G+TOITI_gDR2Q=M<&@{$73yKd29}&?-3AG_i%jWR&9J-)wr7 z?E3W3yT&yS&sE=ui2@BwcH`mzDZf~cR%MaAPio1Wm_3a7gn~){!0&&+Y2_&BDY6dT5Q<6 zSZ5HteLykyDAJA0;Bi31&^532n_qk^fl0;=U|jI4sBV*qrch5I)9h5IH3MVBbS*up-{GdOYS;RTC1ZAF6xn-(Tbb{AidW{pZPwtl>^-eBn zCD?~PN#23(3(4kbTlp(&FudsnVU!|H+w~Y{q{+TMwtR){{{6&+e8t_V8t+pDJ(g6I8|GF1pZpa!}0%&U)Km06S~F zF%4(CH{bSg(=S{D%$fBhX|#pQh7oeqz>CSF7+~PtONAxV3DVO-9pC##Qo%dww1W&u zj!UWqQQ|F5B@U_La^||{A%bg1Dt8l>`EaU$oVQAOhG~>;bj%W;C)KhJy2-CdeqEpy zSP4Jl0F2R^Vjaks2F&!Yt)ySk1IF_=B}>T;b`G%(`#=W`ptP4_ppIrY6_(F$gRoo4 zjE0~F=xa-NOb1O7HFiO2fazTqe^W>EFi(`GI_QGA`kW zGcgcZ2S7OL8v}p*pSgOd{JtcloaESW&D19blYf+AiYaoC3XAut7VLGb@=goPb=t-) zrc#8tlC(u#^yg3#{+AnyQc!WYH@9nML&$bxd&T?Z6DXGx-K z2Qb*<@$G(iO%x}ZcaeM{zdg5vALWt`e9Yy+F%}b(ofsgD3=F=t5JpT}mC?Ma@{>tZ z1N88b|DSpn**DtKWW)C4tO2BsQ4FkA#Z(yTJ|e|7@Nr9^PzvS7;}qQxoWaVJZ|-+X zBcI5%QcMSE=xBS~6RAIQzBxzNFu8}Toe`(F8jQtz>TXb6N>go~)^kGxp zkR4(NVttKL)C;tJP-e6VGz%=F-Du@w4}NeX&tGp;Qa#ePkvtn-lRh;tAf*&@gd&Hj zutMiPl1}#qVb1T6cNbkPx+KZ~UABghE?y*5xK+_5JoJA1m6y34bc+3Y*GjU}qfU;E z1$|_n{20_{b}FtZYUl_>C0Y8wzb0_4V>J&Yinfc9@Bwepc&Bpl*j?YV7IJ>gg2X6` z&od5L1T3cQ$7rik(`{|i-&a6*pS*qnE;rb0G~D#tNmeZ^<7s9WK^t7G$BuwIylC4=Rgb$K)Wlo13EF)!UhuyMOTosTj?9vEhC7 z4FlU#PcgL=IZuV{_5fc+{y7xLMFcd2?DnX1*h3-$3KgKoCcn!|btn^TRf8 zg0`i5qBL-g2V6>ATLiJ(&7d)>otrzm8?L!*Z{wYk&DLR> z*cCMs4c!a{d-#P;^ys3uxmQEC>Luso84V#=-xB3_SQ+D)q&y((Uep35En4}ZIa+XJ zAd>`)%vI1Hkfg-&hZg#p;)r*g0{7Ns+)sm(lhdNyH7q*os(ST zbU~sKM_4Qh^@r6kfWE!|zi4?Pg0x*Rg zfqt-!j#{7#!9oKK)EBH2u9@|T6#CxQ_zd`GO0-hUB51~DTrdR0MBcFl`9wa8&Ak=V zej3QrqhogbcdN+SDJ0F{v2LT7trUr-!jR&+iLWl;G&0flji5W#?@-~BLt`cM8M5A5 zfPf|c2@T~BtOr~TiY^}wk6*}-Kbo}JmSU4sbSE+b3-GbsinkM~ev zDbNvnK^p5&NfM+Hu+c@930yS~NOGM{Fc-a3LFPDt49v-uJOWA1HN4$Sr*o$q)6(aK z8a@{3A>)KKR>;RNWdZr8m&wP2SmC#y=3n!EqsPjU+zMCnxebSMBMj>6`YGl2EOngXhI7w6REvcVeh%BCM~MZz z)cuF34;DQ(R6kmsPKwzXD%&-|}k_x>IiF7r=cP?aP(OE-RXN(WYuV~VjE6$L*e*{uC}cd6HSp#RG5awnL>F9b@`$V zK7uU0g6gNxP!KWwq?Oq(GiA{_VZ> z(Zvo`kFPe3u2aVsssFy}YZ`!yvr&4U9*#$-oA`NwYHW&PX-6JE*$MnM6D$+>gz??_ zQBwWX_w}pR;eYb{DLKQ=FScPX=eEHb*GMr96sZR&yUyG{ z6LnO;>fGd_(vD_!tOrs@qV&O9zjj`_U#l{lA3>$7(1^4MniU^I2URlNAsQ~>!?#)D zQokCfZm!Ob-Bdu=6G|ayX0LN9g}URC85Q)}Wh9g1-4K$%iH2(vyspz*U(p6`lb+|S^6i2n?eiiT9HAiu&*Qri{^R30L4$S; zaI}&XiSwDYFrFDO8Dqk%grYG%b?nn8?hMqr>QSU!^TkgEF3ixI2 z(gTvS%1oeUOctLB?t(tLB09pp!J&oD^XQa5B3gI0YMhvy|a{LcPn=r1e>&#x`lTIVI{g;7X24Y&TgU$&J7j_F^xGZ6x z^fQmRspE<8>1V}~NEwNm$FE0BKC73ekuu6Y{pB}uUY`9cJ$haq`0-4lePNcVRR-CR zUW)0W$d^>uhl6&=)wXw-8eIbc;|4~JdK~9{9w?T5G-!t@_9%{qzXH@1YGi|Qj3c94 zeQ^HpW$nDJe#>fPn}T)Xba?@ReUrP(83hv6i1N>*-IDVJvnDW9max-4PI1D;W)mT4c`Lp* zRgb${ziRpmSwDqn42(lO#Xxb`Rw}I3xq$AX%jh-2ByK%(#CwhFQLkg}xEE>&flUaV z#ZBUB_!eCS#`waq`s$A{mL*TK`s&-RGf_|e$=;Yki%mzB?f%P%339D)-=nWzlVCdy z)Q|=q5Fc~z^~ZN8+Y;p$3ze#;Nv{lR*IJo;&gAgH0x((Oe0bfmB$`c+#T?pThV;4( z@1RT+HtM=Gk+uFZx67e`Zid1O^n>b@NIH$=iUs0$cl@2Ms-}ye>u$)d7Z};nRb{+Y zP;y+SjGKzb@a~UQn0Lo_*07B1k@48A8S&fn1dscSX$yGaS}^g{FF3#3cXBh?Gli5J zEI36JlTQH|2=Ec@a*lCq1+|m{QIz-#ZnJx}`z9tOpi!!IJT|o<1mj+r!W3Q_{h4#W z@QSb~2)%$9#{wY21L9byB}cN@J@hJJ4yP#ShI6JcT!0wvUt#XA&w_=fUnr&!|;A;!M z3ER;JvBFVyu>I%<2Y3HaZ#}+umU^3f_QGtg?;04v8x(V$B282nZbVxZKovn>r0>g) zco%Syc4Uo9G=qBXePk1(m3MPX>3YEh0y5g%OF&Vflkh=>9^85-K{6a;JEg!jjn|~p zWzut`d=b`T$Gm#o8(ArIB|ceVTz0Q}qpC}d>?VwQk92~RW(TNz7X>}$&cQ?#{w+w8 zgIqe;S$f`Or~O0!;UlvzbC1t_I=p8*9yE^{Fh#(b0m?>PVc1{0~ve%#(a>bFPO;Gn;v!IXV{se)Tl#yo_C-i#pf6D zPFLvBly?F%jb0Ht(kYM5fX?rDpDbbSv~s8pKI(d8dZ8F-IJbIXTp6!E!F}M;VO=ei z7Xm0RxBmOO%83MEY`E^&L;$8s5(i!W*aQ^8R~LgS9I~uIL7^`Iim%4~^m@oSV5!6= zCQ)%=T84FD^w=_t2TqUgwdU`?T`5~IQ@`Zg`>AX#+0D+evE2FE_=t0Mv!hT8Je2Xt!;{nZR9*U&_^j6i&6 zcSGhPwU*8jL-QD<&aM6J8u8D1^3t3K|K#e)d&Ta-+pMnodxyzyjrpwDB$RE~^Dsfn zRVAr_9^WRAz(9RMl+r-vSL@^g$AF_b*tV+mF!A@DUG}Y*H}z&Bv{=wa%3m0_ zPSOjz;3U;yCRA{zsCIg-kQyM~aSEP|dr&ulyhS!b5;+lskBlm>$TNA)}#^FEiiCbZ;Rck^L zRtw=MpjDu6&sAeMJ5GU-tro$MavQ(cZ5)r}8GqH-r_6qyV-ID~D_E^Rwp;J9WUxAs z=uJ(bUCLRqj-9Eo;q?d93r6gGw^K|kMPjJ1Oir_6BS;t^!9+$tEtKA-tFR+L-2hu* zUCb7$a3>GuW&>PjBrFQzXAQRS*6;pgtYU1ldm0n+%6dKqRkC=kM}=?f9L)C(Q!cbS zA$2dL@PI5N94cbqa09fvu5lT%i zkvr?McLo&b7CB=JNvy}GqH6o%1vh6~1U4)YVb7!!ZcmmP@@zyjG^0F4FYY_7U-GP27UsHa7d4aJkrBE zy9Gy7&)=* zTW9{hHpsX(VP}Vy&Qx{^kZ7oszCiLs7Fk=^eOJu`1-mgVadpgpEZbz9ud!L*#6$*W zn`1R}$v~nF$T@%)o>rq4Xi*S$^fUnH#w5?Sc{7c!h9&0D{4v=1965=;YmIU7T=lA& zzzPz&oc6pVLaf9oK3x9c^1p!C3B7*tHOb#@DE`_jKSQj`foL#c=3`}Yw=AZy68ql9 zM4{f%_SCTTiO;k%)K`^ab zeo2Hx)ZOxI`U(biCNHmGtH7!KT1cO9TIOl9VZ+|0i5$`c$mF#GVIe4CB|te^gip)V zNnTcTcEg-`H?T01^v1No_}%*NY9|sFvtg{52#bBSE3M>0H!a?^@_0#7>Jn!-$$h?IeosH65?pC3?jRq=anYx%>A*@r@iqvIvpA?Dxac4SN z6L02pe`N+HYj0wD-58~V@TGShm%A^~JD-oZsXry9Hrxbp#lW3BM==mtucE^C%ljlJ zxhS{@J*LtX!Uytnx|&4t`stJ0>|o?7egsvhCH$+NoBg81cO@~xO1E668f6)s6I8LT<3z&Dih~w8DTGrJG_pDbfMl+b+3*SLa6vhiOhQC#{R3W2=-}$FjVn;3QQH=@6-b=M?&^Q2eSz(Ek?JFQ(IHnDsEkQY@;!JvTQH;;_|% zD?U5zm#U-0ZSwVs<1CWo@w;<+EO8p{T@Ro=E>lLGY{Apsdv?!a*vOWj(etQj_Eldfhmc=QsXdy{irmAonvMUrH_0rY4G?8SFC=S#$RC?p$ zh(P_AZe&)=@Xr}Obp=FGbR+Zz?}2I={KN21I>{?{tMJK|4B6Fjj{2+(z&f#G^Nkaf z7Q)=K@M!Ki191M)q2njMdd%H=zrB!TK4(f3rwou)Ofd&3QbdJea3)O!wfnd`L#eRK zi#kB-QC$bK(+zGp>JUI7zjs2&8Sr8B+iXt1KD00zB-Z7szMLC#>4)1RH0Xb51S>72L`|+DK%VQ5? z@heOlhS6d^+unWnnlWiI8)IR^2}cuj$|;fo&t!3_2!C%N$>O?@hLA$91D>f=)RHxe ze;-5r&kkzCPufGP-n{ra_8q4!+QgLj_0!#?A-G?#UxM+)%`O9u2b|Wx zw1$utx<#;ucb7pa}aX1R-4q2lGz|0q1j*Bg=#)ffIiBG_%P zZMd}qq+>@^*Y2R0IErk+ZBP+3bV1K-mS=$i%P7@HRb2$jYKJwdO1+yzr+ogw75CCh z#W&91G>+KWtido5gVV~Ft#5t*Zfx?08($C3h5j?G-QY5_!xl84?fLFn+e2S#Y$wckMfJ+ zWH}{)W(e{g=gGHvp$_pw?$Fn^dfk#id#ZBC8&t8%1clBRP&@AZDFlT#0}0`f9V+L= zO;z9GtQ3}W$`?fe%RVw9KmNZ))*^`DKwO8P@09PPE_b~Ga>rP9aa9GNSEZ{mJ@0wN z@Uo_#SH*g4Vs!eV_^~5Hah8MKn7B0Og0Ha44tBr0c7UF*hh4JbU$e-8DFoy-M=Ssp z6my&+$FR5W6!?zFKt0^l1-ZUvlF4bK3!L}6L6kK+0H_}6qdxcs%!FEv!akU{D}sic zVYP*12;gUlkq{E&qIg>e$agd-u>vM;Y7Uh{p~4srnhwR4H8S0+7_#tn@x$e!Fp`{MyghfBQcHise4KL1ul6lafQ z{nYiIpL<>)Ad)66k`Kybg0n$?{bop^bDu*)$dl_2ann?IPq^<%9^TWh7_`&9&JFAb zI^Il9CBdu#G**Jp^Odmyhak0E6`O6JO5m-=cl`WqmW6e5Q{6j2^bspZhdr&U!F?or zUO%0-s9c4W4{9u&$eXspUUzen$EE;OQ$r%&D8Cx}e$c@#1(}$KM0d++(z6f*W3WWx zApx>_0a86=dG7Mq6M&wby21zF>c+#d_BtZ{>L;pA0cEs0)g7qK9lB`UP$hi4G;Us| zvINM?l2k`L;gWa;(jY-wgNOG5eWcbKmAN7N28wT}UYiU{J(zBbqtBd`(TrZ(gq_I^W78!LF%1U7`9+*=-36Ney~>8apDwMID>UK@OGGn z8IKP$CF7P4MTj#&5)ws?t-i$@KS)fQ7fYO(+1~e0V>dF`FceHQG7N9(Q9TV;tQLBs z<47;p5)aPIG0d5FGl#P1Wu`r}cZ=SC+goooW+(qGhGeia8#bJXIB8%u4pIyd+vQVX zxlW%edZgC{`{&fo82P$xnC|qvOSK>y7*cD1>;oMbq&bCHw782-3rrQq2nTp|Az!|^ z_ua#9Lcu9V>vGXMRa{2!*yhD+-i!UN#f4JLfs`4`16*QJM7{LhzrNq(tOrir&kiml z@ix5KKV$%fOp1X3;T|ds6WCZTlMG%8mRFr2^^nj`3}_%~tcXZbV!Lhaf@-_95NHPL51CM&GP&I^fimDc8aJ%7rvm#u$#^s#r3aY~kBm@I+ zlokP!Y?gYT<2Hm;xwMJ5Iqa9TDhd?cK}B;?#rQXr@Spq*>|+lT#-NqVlT+ZWX~q`l z1=+8guvhnk%jhd}5HnXL)uJt=P1VKQ#nA*+60NjNpcM@88bZQ__s9;HOb=wmy99gi z0#33-ci;8k9WG~;?N;s1nGjaKvM!c?v5v&QZQPY-vk46oeRji6?q-h%E@kvxhr13% zL9Gf@6fG55ZS^rZ59Us@$pNf%g4MG0@{|F;7|_!>4aCD+3M=7f)5TCtj`=fv_AE`c@!Ro>Gv88UT5L5(D@doe zxelZBs8g@3BB)J{b;FH{0(t*b2;rTSv?@>V@&c2T`TRDOdKXprmRhqwE4|>V>)F%L z{f0bFs?n6?p zJjsoADhsv(JVu$Bap)e8I*R2Fn#PRm&W8JA)9Tr-y9Ke}k%s0g5iBkAyCsXk!GNtd za`iS!v2^I;z#+Rna@MhIGN^e+Od}gxGU7I?zch`Drmq!!-?%02X}e*==|&TcZ%7Rf zjoVNQnxx$5c$Xi^&4cPLtVGvUsZW-r&`+)Tz_2>J$WhS!r?!5PA_<3*U`rvx#fLfudbT~zC=d2AmFWxWFoh>7k5Wy|+!|ThS z66^Nd{qp^8HG;FsC^3?JP8?n417>u8#Kcj9$sf-f(|{Qzt3;Jcx9^&&2hpuxHT{LG zw_y-z3_ui5G2nD>rNSDe*n!Y1KPOqi8E`2T-GG8#lx*z=!Wj@=1m2`&81|jS2ro-) zjAxq1bq_0Syu69D-+fh&jpEsVT1T?kWgKmIO*(CWq$3n_m?8&sQ===@NA;V$FC0Ap}I7P5%4uqAbdbQj7Wa{5^Ral zH(X(`up+2~)s++`^Lm=5Q2au)=ZckgTv#(mhnx?KKt1(o%HDWpnTUw zvh{^QPk{k?_D~EYn|1+uggPn_)VRqK_6QD1q2wV`m`vYrR-+J`?s&6gn z1YkL2Ij2!-@m?6?$rwLMo*K!r2flmtR;rnbSqC8Ad1{zQrCZSoOg>SZM2`dBozi=A zTPB%N3Lb&E-#W`at+7XAW#MgiGuhIZ{72*d>=&dg!9+LrA>VD>R{8!phu#18BJXP8 zN{>|WkmslevBpl<_W$JckS7$fD6sBMS8*3Rf9<^1c|(?L<$sJlbCGYbqCEBH@o1L& zy|##O1NG~T$Ubxr$!E9T*l^?xz)&YGSkO{OZ)&V9Q<9Q)X`y`g1e%a)x7h;v35$5ve&;F;<_kJj+~%9UU{M?*B|0y z>V4f54UKo8z-I$h!oNYqDT?V#&sxzQ?pYANPg3Gt?2%&)m>7a)_#JEHuv%lO?+yIY z&A1KRM7Qlvgh6FAQb-54aJ{(0|Q-ZXem+q&(AKP{t zrP)e#qO#Bzy2RiuiZbTUhV!}(-cI}baxH$`a@_K^Kvj1V(Bs{NZZkNvPma6+Sn7@U zs1p_V(f8%cm)FROg76Q)-t2M?)Y!^1RB%1!BWq-sN8IJS#-}I{D|A`}|EQ5+{RLok z!@p$Y4o#$b&xS)hCaU*N4GV1`&A@s^ihYq=+|*^ubjL>i1aaEQlBXSqA!RkigN)(6pkxM79DbcaSB>n)vzrN$QXr2ShB~ zs5Cl=4!^r3c!qwFQf*p)jO==0wp>RH*4kW(0SPb-750#mEnGzS%EzD?2=#c<+P zyF-emea1!6ukGR#aM()eyud2?M}u~M9<=)_H1ty-v;ZOuM?+6=TLj6pHn1Ti*{h3= zd1L)Mt6y&jwXj<@JW=Dp%Ttc==uL{SG@dQgOF*#YI ztam90E`tryeb-N28bXQ}?DyR1eTx2Y`CqR}+T^PU_9tDHd?CPYrQs7DUVl$`af<%G zX%lJ3x8abEiFSM}rdkQbR1G0T^bVK3-r2!1i~Ak!lT)0F9vXf>_Yx;V6)xQFuoCL- z)-cIl?Y<|ugS@KAD61NuQJIjvFzaT~biaMm^^4HAvj5}PWakUB2z_j@2<1>rCPgxE z!}pNx=iUJh`zomPu5`NyGU$ihYXmw9&^uoD$?9Qlwi12<2iadMIH}x&3wm6j&W7LN zRq3X?ZjU4Kw^_rS3{Ttm;PPY@GHdR~anY}?w>i4&;j;Sou8|~p3KZIpjIZvam>h~^ zQenuflF9ivuv30f)MH;P?uY)K&s|ISk6bPVYNbt*PRQNenbkGXeRiDPr3j~X8OI3DEKx4a;sn3-$OyBcfqydb$K z6AC^9yhGepMaSIzlA9qBbHb7} zg?XYfx>R&c0TdqZUy~%LE=Y4F`SkjQZK}Vl^8MQ_(#knO56bJAExuNBRYzEXaY1*) z{nmQP3cA$p|C&Bas7J`}i>CgGL{0(0fRPjNm4y8ruNnwQ+$T|-a<67THTFI?N(;TB zNuQ!gl%(7gP{%0>S}n`xpL4xAe==qOpCId5u-0QY8kf~9*scND?A-2Zr-w(*gI|9{ z!fn`qBpKi~nqoFlWCPZGN52BCL~ZhDAig~P%0?F?x|}S0j)a9-SRHx8B)w*b)mx>n zdp+;HB+dRGd-W@@ge9FmOFohJLQuYhpYhe}B;`T=eQ<2UeY@y(-)ec1@;{TmoAKtZ z#s66svO;*)cL!4+Qo{ewSgQQ3{Y&7OR)9wuLe7G!BrxFnpLa?0ST(x<$e@a;L%ugS z@vn5zslfiAk*1DUD)})M%mR@qC-#}gAAiF5ezO7*Mc!KRzs7n@Pj{->a8S0=z}6h5 zm;)3kq{9AuL$PnTxwRPL_61(iKFRc|g$PS|%$vP$^n{iz zzWvLHcMgW8|K#XTS5bFDtKZqMLv<5c;Ev9-7HXbM^f>YJWB@C@+V$WktBoVKHp|mtA{N`oAk*$`rqT`A zI%IvMOc2A1onJs#_-u4Mr$W6q>(B4V>0vR~BX6_%Yty!3w6@o_S;0Ru&H~viGEAhW zka;XG^oHU?VzG24sYcw}b8w$S;m!#39gPRNS< z@jF<FN{00P(zU>>ldn1;+45Q=lPWF_-mNNk=@iAQibHkOv=!27S+~4OepOr} zZV1`OEu%9zX^So|isW4dvCnPaXno`6`?o{uhbe3=I9ZmMq9=1_xyfQTMK)W^9rEY@ zPW_rb;L`Z+?F_R2g(*>}HP|wpq!?&zD5b)Z>FablH<^CSu($@i*qER$l?GJ2w1PHy z(OX~qed(GlAp|GPD9s`Nm9n)_5 zX9}YrsUFSEmYxxw5CV6js7To2Sn6F#_Hi=<3pwj2&Ol|^Z-eE&7eC)$b!?#?MxLuS zW|PA<9Jjk*0E{Y%f&AlfDl9HQOYenVqvJkF$_jb4{A%!4uN%&#e&wn{Ul4Vp)yd+N z0HBXwxSA zWP+y$8;ms5JLQ=`)e^(2mMskz9-M#IXK9i$e&&5rIOEhD3ob-s|LfU2|FK82+!d^z zzwK^6kN$hlN3--c#!jp{MAB{8#+)#)F$XB7kRto2u)5)9CG0WklEeiJGk!mkYUoRx zE?I(93*v*rs+4FOuv;m7A@VmDi4LjS0(W`0zY;MEn935QDKpYkI=+EKWwtjePsS;Z zDzW%Kf^X5@5H^5*Y$TrX+hT}j1?ZPrcU|^!)Z^yoO=KyFnKIhG%Vu3Pz|F|~Xez}( z;wcgIvB-T=ApgR9Q%Lv)zm;0Wq}^WezbE6Gf4YRM^g$DfU>9 zqb_i6;;WH~s)a_{X7wcC!*GJkMT_BG)_+$T9C_E6_0j~9WDzu*<}%$u7pAoWeG(L% z=~`Yj^dVWhZ(_hCKT>las0Hhn%$g0;U%yf8vF(1NG?l;jm2qRL%?9X9w3_C?#^o+- zQ7c?4T-Q6Ef&^f-yht=;hy8_WEf9vGYciRzH6OF!V3>tA%Ur|`v)yk+95e0;GSN<> zp>HZSF_EevJ6)wJ>OkN_2jhXl5>aB*@HpzT79>;e%`I@Qfy`X9U$6gURFbfoizoh` zWghr5MqxDB8_FRm1Iy3UrC)B`H)?fa55n9)GU~XiKlL}`T?%+P`7$skG=BCVWR2=%WzsrFygQc%442XaZg^BXqmStB z#bb}}-$PE(XO$Ts6EHOIx^pt!Avzs`<77~2P9-xw2?5TT8CCM}EobHd|A`eG$E$ID zNlXKON#&m|J@=_)>6%&bJ|G=)2!bFw0naGEtxid*94g+Y45$yj3>@Edj_X;oEBhVi zM~v$>(-1Wz_Ig-N9%s82gl)mbCi*G8wSUE z1Fg1)6w^nMUMg&#e4}FpGabSLAov6r8va5Y)?$ zpIPISp*qIP;NB1+ft0$Px7DvG==coKhCs(W)+1SbLbYCzHv7JpHgH&h9@i`!7n!DD zjC5fn3N{;n zF+r4Y4l1M|P~Q-8oRovE2-2=(aaVJapq#(Y;jEy@`=IwyU9nWByveUf9?SiBes=Iu zynZiqzIKvK&n-^9@)l7A=Yh*fo^_m|iQ#w{)fU6<$bneln96#6@m#TfVM`TkYbH_8 z+1?5$A4Wv*lPD&UB0H$CPhP(Y*`9Tf_t{HF^5Wg^dPee~f`AvnDdBG+5fIbM3n~^D zE+`d5^X^PvKN&t%w=J*j29}eGplIGl%Nu{b8Oobc%e^OPhqT7NUoj}CBuUe6&us%`_e&&A)y~_< z$4Z4pziqB<;&R@3!5+yG&q2F3Rru_Broeg7uFR{*HD(%gQUmKrLr5kE?pJADI}}(o zX_0R>`#CXOl16)HiJ3AzC!=g%Y_~Gu!mmUtp7(uK3kjcI>2L_vpUaERsn*W>%o&l0 zJ-Ck+cDN=f+vK~YcO8zoTLh2hVTT=R%^TAaSFuBFoa2<4#^GKzsb$*@WSdGYw{W{i z1Rt1q`?&?24fFALZeY&;4^db15M&8Vup47<&+HsyX_2RU%4|1WO?DeBkFgXJLy;(K zeMTL{w)fgX)oWZnfgsFo56mzoDK|`w_nAb?F+3aQO+GAb?B;I5SDQ8;zM80|#@+91 zOoa^t#ze0?a@yjy9J0<>=61og=pViE1_1k3qy4WYvU=DsBuwNyF$s>zPN+)C7v>9f zCgmSF&k?A&rEExP5AKm;{~bId1-(*gN_tzH>WpHDyJVtu^^qOUPQ)lf0-b?Zp2;`=p2)Vu z5MvraHXe!X?%9b6(!o^sL3upnhE@_JH{LJlbguM1ys#mp)GyKn)uskK_mCR1KN_e$|;2)Qn4^T&NN-j8ZV7`5HzuwPQ^x5C1)+VLaj zDeGcBEX=s_%)-+OW7-Ux4bq#S$k;#?Ev^oYq;7`PimGL0-fb#SE93QnW^;D%VNN;c z&>XEanJ!*%lG_EfDYbMT!Q-WV_hhX~JXYjJDI z|6BdXfBVhP{_vlYwG^|8BH@N#`SA=q!F$+Y;yWP4rK-M5l02X7d&NIo4uvtj4M1m)#b zNrLoq&&$k;X<760p%nZgC&~w#a{9Ta=m>i}SHi#O__-%a7*3j5VF7|+`(oXxe=@ph zTk;#XIO<*f=H}(@UwlC-rjQ#3yMlU(0eXP*R9KA+iKLT3K1m(rcSCfF)G5no^}z*+ z(BFNGs1Ldv1g{#k-AyY6IAM#q>NMOd%LUU ztGtG}AaOUE>Dk9W=Dx=f8!c9OL1AN@0y{1sy1{`uM-82)$a7V1ats%i!frMSh)uCd z38x<0i4MukjDN75D#P_)eD9u2i0N0Rr)m^ez;>;6;fD|2Hr_Ow*lnh(5<#eIxSgg( zmMPHyCwWm2LTH!?00j<@dutV^Kwl?Wd`y%Ftv8v%7PuaFn`Tb7>Bl|RneoG9tdS>n z{sn=_%5SX5HEuXI(Zidb~U%?97 zFH8SL;ufs;a<*&aDdgZ3QfJ`hR8kD^P?S+&8ajSvoMIinMs|ke1s(vVz6c8Eq}@fQ zYOA4-dfjm9cSsQ<;RIGr*SXYzFc4@p;3aqIX4M@TR!rkzb-PO+i4dM4jUY4xUQ36p z&Z#J9-!u&kCDdfdF4^Oz1KO-89z9d=1S=;UnuO)5=QXV=$v8KNEXI;;qEE`p{b>A9^>`$Z+}BTjY-Hxc3oNG!Hk1P;(0*8%!Q)=#%mi$BW*O zmBg+P#25UStPo`L2;at89hPOQgx%E%|D8RgkTBN&cWUwPSL6; zro$JesfJrohP-jB8ZPV>z|F~YqOeJH9E5ig7gi}t_!_TP2G`hwF6Uk9l;B3L6lO|3 zf!cC)I4FH)&>ga5`jQB`Ph>qo`m`oH*nMR&t@&aY#qMcu1Q-|OK3Cx#6V=1L@^0Am z-{s|y+jCkugNnTYEp)9@x+>GN#;*ieP^+M)HJvU}6@#oy$&97Bv$_aW{3EI0{&%lQ zGPns+Bm+oNCV5m#aTm}fISP)FR(>7?TK92vWA3!y>C~Yp4?ab2Tzt(ktf|(3i6yL+ z>E*YqboEcC>h0F8Up4)OtheD_HjRPL7Edv-57H`4=JP91=K(SyVj|U9)O10 z&59EDU8-}cdvn(-U}wk4fku)6&o;r*r7=Md>CZ(8j>jd{g5IFGxwq$@qOVJ^H8DJB z2Z+AbaCSNMfDq}BXEZE3cvZS8R@D}Q_a%CC(J8_nX(kkyUvf^KQQ;HiSH`VH!FuUw zVFp(_qfT~Qp%ruot#+z)x<|^*+38usFpa&qjcvMqGQYXVK@YLwx4vFWHrlZH0pZ>e zN^$iu@Tq zb=LDe_EcHq3U&slqvLyfm*~;)h@1LTQp(N%+3l!ttqd4J03{|8!$NN+ZumVtxEtKdxchF6cg;^==cTnSPEi{${HcAJ5!d)Sq)2i4N zoDqOXW*?GZ4ZpE%$>>+G1Z7l*G0=FBtjpRZftw66 zqbLGcCRiyv;gaQ)AjQ7@K9VIq=u!%yMfDD8yZ;H7J&q;(8rLqe)*(|l=o1SWKkGV8 z6Tk_}nP+$T?EiUf=y_k3*6`Nw)&<};7z%#7WKEI_0a@@>zv-&J5U|2^jZ2pBf~z`9 zh+B5{UvPGh^Q)`b&-YiiTJ;61W$C5VfBTL15A;^#OsL;?NUaS!Eu98-p@m|uQRE7G zx|y6kk^=fJ?-X6>b`ctoQr$J2yU-SW#IuYB>M2f@cLlcq!ZfINR>E%uujVR*ZaU}& zr+OI_d5S8ys7nNzdUFQl8LC$2eZcCm*&~y4%PE(Ocb=kGIzEH~z$h`kYZ2h8Gx9Sq zJs0bdp}IY{gFa1q9Cms1f)ZN_r2A3!WDu4;yklukPzATwe`%6qKU4&0IQQm`X_452 zZ1lB-?a{Fi$_ivu%-qw<6?%|ukMX%gwy|5PY_~RmfM)p!KRSbAfGakI3hVQa0zHA7 z+)}!exd?03Fz5GY+3F)Gw$3(QK{sbBFo`m_a$}87^$G z?hbz*6FI+6Hs|}syL6k)kDAzb*U0LW7<-Kp7dm5Tc)-5^0MPDqc@em)HM0Ho zaSBvDuY;zRO-#D#j@Me)JY!r>-K~g%u<*xTaf)^)O#P!{TrDpZ-5FNqgS&A+Eu9w_ z&A~R8G}SP3bCx&`G$z`eK7JLh(CJHHeEIg=p+Fo8$%L-L1|K5ArQks8NRdimm@M%g z#{$j`=L#Q?S=P{(7mc-YP6RY;?T^v*&%_ifTlCT|);?%87V5KEh?$Vy+v`}oV87?z zu1ebEI+doqjxe*N@mRw$wVa=w%Eyi3|^>HG3wma;EGP*$ZkQ=+8)UU~}8iK#bxo2g|Gd@`1LSq&>#(hzC$#S7F z5tYp@I3`G^RxkWga%}!BT78K7rD(`IMXYWhyQKq)b+d--&qvhj|RwI7jWSJIjx%;7qH*08!BJdOleRSQ)#e)1|dr5 zX1e%Bm)mx61!wIPD_WFsLF+58b%OYL*6rVJ`JW%_!IIQ293b^C3|P7iz@nv?7K&V> z!d3{&U5e--&+Xzq|5imBeFzxEwSwXWr+_~3yf|FgrYi9pv@4?9z4>=&rg8SDm8!{JR_mc)BC__no5mVX=@mC^C_;AUZLeeFZr7 zFP?=lIJe{N@+}j|F4%C$#zb}ji!FeIL|1uHBTIE}5wr;ElxVBab%s`W%diY1Xvi#mhvqW(0{Q zDees+pMWYdBrkc*K$n-|9;XQRJ;dE6PMJ)qJe)f7fiba(W_veprXo(MU#LRP{M%8I z`ke8v9y3^u3MeLzBH2_}+M@J;7Jeo-ZBee%yoa!x0|OV&?sAw@3j{u+KmNT+iZv^kvQC2_T0*03GsBjXvX2is{k<9gHH@n1q3 zYK`~QHoH|au}l6$r^N_aCGIZgOrVZP@LEYg&R)ESizNe1iYNmJ4F+rtJ_N#%g zNj|?#rRxNR(A=V)&qG9?(>?atqtC3Z@ zX$8kfjX+CxFT`|@zD@$~Pg3ILScG-YYtt-{30`LLz|1%RGx~aCVfe5D%u9#=*K3!t z>?fOUC>!4IKt_5*iaL&Bwoqg*$S5;lX$DzjMQ% zx=eA}K1-MdU?~+%vd_UHcxFuqj50PWRJ{D&k2K#lmL`9?an^=CR1X(skRs0~I9GwCkqa&^LrxbIV zB9&CwS!F*k7_FoS!WltYzB2mM*#Ee`BM2#yRAA@dK;4EN0f#-Iak$cL7$Y5~rK z8Yn1Ao?51=aC%&yhU_(C6{wngz-G>x`5xd?SDO6jXP#g;Yc{LQD7zER7?b!uSL=@n zVqY!Ys@OlX-)XoPUq=^%(gKs`w0;F1^8ozS<&*ZpG=MLPH}U_z;%OMaA8zx|2p=tM zl&*JGmq{PUPc8hCw8^9G|H0v_1Eo*uPUg4zrRTZTn}%g@z`xhO1?&SZf}e`kxae9k zV9E#O2VJr}R?F0=0p0J|s;rXK$dXjGqRWfQe6qyozjn(uP19YEs?Td62onu_J}B->sD z>iPv2VP96i0A6C9IQCc;i6GhF@(c4U&{ON8)eo4qK%?XhUz&QJ`KG%beXD=(8cC9$ zv!a9&121hK#X!MsCKaY*Qmp;X-OxY&6}g7nxywsMRlzOvIWqWe5@_#ZATEhSAySG9ILaSvOevi|n7G%Phgo6b<=6iGUuH@5hDCH&^ji}B zoGnFA4mTpan?fW(rnL|(LGSjee|KNjG<6O|w`)V7mpfy1<6&FzCebBpFHsIMs zZ-8D^q%qvJ(lR0(2TRW*)%$$m&AO>m~Y-@ z7Znpk_&51Q;tg}NgG<~s{7i@r#sx%>1L9Ke4UAUW&(+W$do_e)i8JUn!DVijL!tb< zXa(o`q%1lt@u0ICY$kS`*P6z|D2@m9xlfthZ2cnh;YY1U$zB`wUMme2l*1HrfFgxd z*gknQZ%~dDPsPe~*qug^TY=ehtOp*66<;8cOg1Qk0`a(C=d=WAl4=iCqC^@5+`t|c zl!CBQJ|`t;Yp{;EcE@QFvBaCPj6hR>eWs1GNn^?PBur@pzaFNm5>3w2vft zU_fX8%uc}p@8feWdNeyXgtUO_ONzKv@k#JX-%8TYZGs(DzoN_e3KaRGL_%9gGwBIh zK*Zmnsg+){Gk8LJx{wsY4oMWIJ@?KC`@G9SG(H`!4 z*GG~}PKI$ouqAQ#7=PpYZ3sVBxT6@i_ih<8FXS3nhGdG_Ns)LeEDN-Dmet#9XPhCa zpr$fpr=hQqC~=4A=6rROxZO7~pj32&F5#B}he}aUt70=KP)vqLF|LU)d}da8$7&*M zw_M=$Gu=mxbJ}c5b#0hQ(?rS4In^Uc0*H9jLNxecU_TdwF!8W}U0E_2a5luEncz0O zX%gPmbbsv^#&Q-WbRbrG)j+%XO@};gqjaNVmbe^ZR?!d@)e$Q71nu&e+{bF>BSl+0 z2@mqgN3gONw!00jq`g$e$Y5hHY}XQ+5;{5HeR($Uut-05J;8k4> zbAtjPr%AK|OK_@`DNuP`Ko=<^d79ZJyz8Fb+{wZM3p1y4JD*r0tG+UA=489S`90(9 zGaJicyEYL@yGA1CtIhT&!G*YkA>q?YMHxVVS`xg@B^P99s{l=<-d$v+v|8R9JZM)E zyw)+re$Z|uljGGR9<)ngGF1IQzH!$#Uvk|W->(8%Ui_U$e>o%B3wsR8d8ynlya(r5 z6mP6S^^6^jB~f;NZBENX79AUgjw!E-B-1w(4Izc{%Zs)F;psN$Gg`?EDw4f4+$c`H z?3lAfCw|PyVu6e?$Fk%FZ|%B$AV|OHd~hf`g%sIvUR3w5Rj=Vwzo#=Gh-oB>q+jsBW+biwu zeC=(g{nF{3`PxAg@qvH}DkuRJh&&bI3ze6GI65jgFrugkf@mEa1VsiFzO_hjNaSb^ zNobtj&aZNE9y^fr&)RFRz1IKFJ`IhN*`1n@D*E0GBZ@i~hqRJyi) zC@FqO_b6&4P+BFZi@*eMrM;3p)+5LYm^|ZAb_7-FPW<{sFF?2*R4;5(91Peyw=HD# zzg>kU#M?m^7h2;9p7BvW^O8mOf944GxyWw+aXbjMBW>V^LZb&+dST0Pvetp;A#e_3 z=Am?o*+P*NDmqj0fRBQnBW&vTlsCQ7g+RG4G&&1ZRD`YtEVrC2_6!a!Wur5QeX#J< zsGPttCHCik-5+HHhwJ^5spKFxnn4Gic$!S0R7Wv26sb0d%X9;w#to=%`e?>lD5XWl z4g6QD#PB38FlJ=`s7qcJg*+TkO^cPZ*Ss>kQu*adr1(d)<+}jKyFQ}Ay;Jtc1MWc? zzfRx~G}PXqOej>=A}>3ZiUUru)ORPfjoL`zC{~oqS>%*PF*kvqPXWrp6Q7W8Pt_Qa zbUJdjfb=@Bn@Kj=M~t1y3{&J072QF{G1!>i3{gf#fL3{H!9Mw8!E#6$qpB5NHL&>W z=zdZg*a5cfhk{AGt5Q}6bd`9NL*sFlbf0_x3S3ZN|3jwgI@RD6Zc;uE7|rt_-kTAvMq@!k<8S9`;10!V~M;P>K` zhgrF6V5xch!({*mc4sW*nla&*77y|pJvNf-UY%j=tUcz=F1F0<#iW$Uaw3bM(T4xE1%s$S9VBNbu|7lKh$yDWgs$$*Kg(TQ4}vJpQT?ih#7prX>kj4K~TegEWk?4Zdl12aM=yPXj{l~!4?py zM=94#`crwn_y(9cx&B0$h2rqXzxM)z3`ST*CQJqO= zh1Suh$K3y}rex`w$a0r=HJAVOK{!6u=sr3gUV9M!NVu6_?(z;C7cn?elX?WV6Lm%7 zcqf~kLO5}9MO0M#{&}w$VfXV}=eLo=<0XL|cq^&dWDYw;F?AHF!L-hD(dAc8(iOpH z$e~3su#@;$kPhu@JtWUB&U@{AsJ&$J{EEd$)ZZt(AC&1|7;yn0+2nj^QI@11YAcg` zS1RtlT<((Ydul-@6pdB@`F_Ni`7OaX_6ZUr z>W6xXvN@S^tZ;Gm1=wevhwBkLzycS|KkjS!phs+2ZEgS&buJxk|)K!0ocq|Vjel&f6vmx zOJf$LDz5r%g9RA2sO;fq@j8WFawL<$vn$-XLsAv#{x|qBi@NDvuXXNraz)m|%LElR zd-2^BX%EenXILa#E{n6nkR)!{sYMnycV*D9Q?dNKN2}tK@HR5+R4q8}jsG8Zx=R-T z1=R*PcbXK1;rBgXw9cv`0!#kd34g$r3~t3e+`hjK%Q_V}FFMtKG+udiV;RZj7OyyP zSbWT68rw%PB@`*9qW}J@DtX+zee&J@LF>T@gAEynF*ucY&x?7=;J;*Qx-tI zVT8>$Tb=vJWe1LD22FtQF~!`X$PL)5^nk^)YqNY+@TOqk@JR8^o>wEi8ky{zBFUtY z?B@!>YO2*#gFCL-i^2?vKn;pbr1-99mf_W9TBTu(U3`7SzB#a0*%y%lTp-!Ox~O*H zfUD++swhn7k`=n_fOjE(%Uob)H8;6F8<`Y2EqZPmR7vB-->Ha^Z6`h21ytTwN5?L6J7tNgp9h+XEHl z6~YRTLwYN2S`xp6zrugjv?^H@z2C2OzCI!oR2So?;SJ~MT<2R0js>os_Qz&*uBtY0 zyXSpipbzfvJd5cT4}&L5V5XozXKMW5IV(@b$aq6wzE(VGbNDv5-f%i5p;Uyc3cG)O%QH z+)E7o-Z?PE9h-BSB>7HWP1uN0xIlx&igr0YQ~&$z=6OGd=_$xyYtEVB1QWe{L)4r!w zKK^Rn_D@0@I(SK~jRPr*W@*a`p^kHe0X zQ@*p3<{JHplSFlpiTQHqT2CG^$(NR|MVqW-|!O=JB$i=y>h)Ot^Lgo{aO%@Mu!+rklJs^S4N9pzh-E z{9exwlRt1)@@~_DRjEaI~{!*0J4P?4DzsjmIhPMt$%eZS*Gn-}609VyBTTlds`M zib3@jzfsu@wZEEPX=>#F;D_bldVZBRm4*=KqkK;1;hsa90sz>&FyQeMB)IQluW63mTR8B63vQ zfO%1aL2x>N>fl3^z-ngP-(_}c8I(*i5RvG83v8IXyMmK}HUwZ6sTHGZ1}r3=?gzo9q=OB}4lJFe2I^kH3ez56y1UfnePic6;qq#qTP zyl&`i!8>=;@gP`YvxSbW0r8Y&CwPLB!=z?9{i5YXXPCimlX`iUTe%Ag4G!|-d3nG! zvtQmrYp`7RBz@jPQ}0|iU$bnd?@r%knnuth$pa<(U2g3FjZC^MYS;-SNT41fbl9o? z<<41JRX1IxN&?NnR%vFy2x9;iN9^PsQE8QH6j{!CAFS^jZFI2$Q?^Etsi&DBgy1nw zD4II^^{*--jZSEbPM$>$IIt6HG;u;FC8n;%7q$3`eZoZ zfP1PUMb#a!A-Gteqfg0qs#B-K9CFhuMWsQZ92C1ThRVBXt#4y+p&(D12)D+FG$`!V zMXzUyW}khz(fx=D_lh;2%G2g_L4ix$+)cq+h_f}Pe|BZb+g9{S*%wF%tBHY$8<5Ul z{O{v08G+>YyYgjZE4MJof%l#dn80o~#Q-x;r2(P%l~b`xbwQYe_(JY%19BT4kGB1JAqGpgtcC@wi9wU>=qtzfC~SqHGw1$M9U zZ>RpS#+(!0VY4 zv$816qD73A*02-}whBTl{dL*VFkZ6vPw8J4NQ`sL4&mlDlIXy}3TQTtSzK!=W*bGe zQqe~PSB1uTUzKLknL(PZvfCjUk-4G!$v&T3^E+Vi425Yk?|&V-)fDA|L4K9TWMz>i z01!)UY}I|-0CN7Xvm4A^2$upl2hP>7IGj$X>(3WmV)_&nVjVQHb%zv&VKdUbS)CyK z*d2no8C>w77eY1Klu2T}cT_xmP_r=O$3_>Ty)Al7nx>IHlW%G##k5nT6&gncCp_W> zRWwNY!eXF8d{meburf5!8&(h3z3zoy8EFYDi>#JuwuPSaz7ujK_?lPUl9)vuki}Zz zuhEMy^IBw?5)Iyb`jy_mVW$#U?Bc6Z|Iz>$r7vXAb!BJ+y-uYCqRYyK!%n;DGM`Vp z&nT~ZwJ0(rIZ}KNO*ykNv_-Mb4U%saK=OS7Mva?$iEmi%j%1`2SHxP?YLxzM*I6IC|Dot=)60ku2>C^0Af*DgXj(Hn# z_49L6nn*ggWwrw&?vM%M@+n3~k!)xt;2jIBgZ+UU()G~chW~==ET7frlI(){7}PSx zJPtB8YIeIpQ!TF|;C@K50RPqO^SUkv6rKewEe*CFSq_b-jlzD$1~6hD+=?p;=XZBl z(+}XlU}5P8fJ87|EbaohU~k8kxv|jsufgVPeFX5w8hV)e?1X}G4aE%_;|^h$ljnc= zMx=SV&S4kZED1O)@7~S_O-SA@sh)qz3-%U;_anCRuSst#>{jR&-xbsb*1I>WR|IY0 zrNLIzCBHIA#wGcF>Rs+KH20L(b(+1!$0i>-H{+WIR5tkqSO#IAUdCBADs5CB|6Z%KNj(7V(ig-lf;N&hq9!IC zI8+lokNR&URrFTbX|XP%QC;J`g?EqMNE(zE6(v#o;F%$(BK`$pdTkHgE2x63E?&jE zO|LeYxO4pFBo1G%+t^4Wvu~#{P;0HP1ZF4N|>92Y+p0 z)n2SK1kwm(ewU3Np?>H@|hel;m++^f+#sfB+TVm@IJ_#Xzj1M}ZV9zB&+f7L?sEF*LfH zT|NtASUqSKm(POK64p4xK*dBNuZ`&DVQu+GGlrdTE&KtLI#wvyjTnZ4mW0h%=?MLh?gxJ0xy!y4jk=5cjFlEkxMbV zDCmBOPUY7J+@)g)#=GTjV7$BCcNJ-J|Afx;ISr9-tgO-nR0tGxB!`AtP4%wOWRIia zK;RnLtiG$*Y46Z#;-=?d#|hlZJ^_ulEUNi-*4!7JGI8Upt&zy-o1Huz|qN70Vqg86X|C~pqIB#GIs6WF(DfBL{ zKDc91lNux?g*mD&Vf?hMQ8Dt>)2aYk&E6Nqb#xXl4l1P%%(IOrnK*t}=bZ_l<>^Pb z;pg!W{?hVoqoesl*B@?>560UeaNz0Z?`^I_8nWLnQ-arz`kxb_;7O{&^lqSzg8RFGs-=!jnxUsVL53xf*W+anRe9vy z#~)OrGl;h3s>gzZ{Js1R1&X8$Ib9RSc?>yS54k8Fazgo&au+Ngz7(hfq2zl0ZvP=5 ziaE$12s)*Z61HA@^3$4|08t@G2bs(E*ahE$(sPbykH5 z@b*#+?2YFew4KVpuqm%vKJ4{ zX$E0c$xpW8e#Qm1Sk zxgBzm)Os8uYa(&9frvGc7W@E7#J zMGuex8Ej$)Z&P-?66aqmex#`MnnbHxXl~3V%G|I#pyX~B9_M%PG&kMzc%8tid^Fr3WpJ6ak=szM zHsD$h1i8n7(@BGbPB>^in6q9!Qsfrr-73`4Hx$E8DNz-{-QpVyD*}26*16z%MDvjc zmU&mvFn0*n!E#0GHt427U!LHBGYH;ySD?A$uAA4Qzzv_%q{CwmaNx$x>)@sMK4bfb zn>lUc(8rF#W?oL_blhD(<&=frecc%Nyn4LvC*+g^FOqJWgff>X<^n|;vGZR)y>TJ# z8twDiJgJdwtG}MtB0ldvs01=Epl4r*5Do> z&O>eFk72v2mt+D%Cy>eqcEcW4x@QNlcy1)QP$>MdH2&3Gejn^h^#WH5Zh`V@VNI6)V1kFRI)B~_6nyib?BjUqf4QLmMEKqX-LtTeJ_ z(K8(2(?*!snTbMbV((`8Q~XsA&c5i}BMmU6)Qo8D*Apq6CBgYM60ON^2pvVn?)$Z&*t)H2%}} zJG*H5rK{$(o(|i70CXQ?l6hMwCWRs!sOV-$lOYvMmY^&OlX|ERi9P-I-4pnQ)F$sj zn{OOG+a8wM8m5nD4`k!3oczTUS26foH9a zZ3$oJR;$ViO;r>~j|9|498tj{TF?uWJ$N1##=h%AQC7H(Z1e|CW#^Abo3sy1RTPHQ zP&r-=?gzXFJ>o^U|7(x$CQA>S$$317Lpv8;gRKX(DO9uTXc}$3fkD~BeyNWhS_HIDWC{1hC5dT;WO`DS9B{j zK;d2ih#zUc@{t&CuAn>YRK#ZkB-;VU9Fa_YqMd%%eENC1)xd#650)tjX@N7m+Jy$a zy)G#3?4*ms`Xeys*djwtO?!AZc7t+23;Q5;xWexDHvRSAUZafD&ttd8XUUKg`_1F* zb2~6*wwuhNi4?PrBCDzB!XG+!BHQU^jh z6w59L_7Xz_00_Dobj-JU-Uz|#`TTxT$RCjP&>Ml%`827OOaj&?g7nFll(OG)KKG)R z&+Z@_1-R(g6O=C+cFIJD?&y6tW-rRq%KAzZ7?ob0@{$uiru??@r)#~8_-Oo(18cf~Gu*0q&<=SWA0e zJV22CD__#%yp9(WR7tvN6re!r-C?KP(DR^WUkQZUF*EZe$V68endH01U9*pOore;8 z>lr(X^h^Xp&sJ>XDahy0YPQCOK56UO}YJWsh6BTZ#M}fkIl@4VN9N4j#^H zTBSa6r_7o_#X4|&29t4*tZ+!2+zu6=<&$AeNbA6Ehb5#)7%tt^IQp+RgVKV zqOxRGQhYnZ)-y5TH~4+RwF|Y%JRorybe$9snAyo9Oqe}0883eoxTMCM9oJ!#sw_;n zSYci6qV;W_wM(iQT5`Y}H;)VXTjmZsfo=?}cd!R%IAHBU%@rwLIsg^iDZV+XdT2R* zK=k5`&ZR)?Qy|*O41)eqjHp|T`bRgxwHyg3U)%_?HG^*Fyfy7~AG8alLQP|fEDxG= zAW`4Q8y%~H)(Tf6uVq(AA}?0-K7>Qvz?iE87Q5RaT{MS^HH(lk?p!mWlNN)F({wat zYTn|7=9*Za6JOI{5(gcpm@0~tQ_+18#MD6}dyTYuPPuX$1SXnfK~dNM!3^VSvRq^! z%!BrXZgIP?O4XrU2?Dr7@`~VQ@0ISGfa_y`ltuNrHbZ=T!239VbQ}=wgPRq{`Lz)^ z=&%!PE-O0}cwEA3VGJR*l`hSGn6nsOld#~HO9v)4x%H_R)8G)?U}K{Fjrc1*mX9Gsv>HIADp?Cr`e&pwcxRLj_-)7%%E!< z?{@fP+IZx=)&Rb|~{*Gp^XJ)+kOnaP-NdI5iBcxbcE?S)~Ak{1$b} zkwf&NI)NYOe;yXT$V%FwKt2ww63?A5QYsaDE_c4j>lL3M{oxxV_>3Ln56fN1cb14ex&d_k5d%1#<}Wc9E!fsQg8Rz`N5Y`7*rOs5?CAj&Z6-!8b$dC|oh8lZyB zgm1Qym#U}~7lWL0D(v}U{Rm{^b@TE>m^nqWXM&2|eo`d_X0LZYl&l62*wOeCb19<` zu%Gccxs)jfE2`h%8(oT%zf&xUn+7$6W83;uDP|)@lBnqSU)KBJMn$Geum7nzANpvJ z_ct-PBYY3BbG_hrAC~1G|9DT^?;N*}-(g!aQesNKW!~M#(nNPG00mP(RSCYDxU->I z?wJ*eOdlMoENo;COF?0yx3hQ{PM~nyBvR#k^$zo0Ifu=IugKKF_3`7U9d|kIvU=Lz?O&*7DSSsK zk2qNW$6fEf^t)aEzQh=Kyk7gSd&yC5YZ=G=%g7~@ z_~`?Rd7mODu*)aIOCMJ4qoZR*$T*?F?YNP)9^~RVByDld3LQO_swnX~3q_5ZT%ap% z7^z*X5Lbw+WK}X~eTCjBkn+tMa%;-#9 zP#VQRn6`izdVt7+2}sX=KUp}6gY6k9zu{Y zREZ~-Khp%b`<$?;(NIogA5XS}92DAh}lJ@aWlUzup z=FMsG>r#JqS9fHcq?f)Pk~SybuL35RlfVgF$Zz)UrDH^;^FH;if#xq9ZpvF+tQBiaAArXm@lc(6`)+02RSP z8TREsfcItIj@ikAxVf8PV>!SK(X)I;G@@p9W7{2~OnQ@dpB2(N*5wpx($%VJ5Bj z!vAZP*F%OwVuCtnX_b9YSOIjnDtM?s)Z?;ESps`W%kIkuWSznqL00Ipc$bUfPT_zo zD^zn=&`aJY$v~%{6^f_V`sB}r)48g$q^3AC)?udm4pT{A)rgWT&Ecz7y*uAZO9 zV54IqPqWQ;$LwF(8cl3YitBlHA2m+N<&E57nwBlfbW!EcbW+!kSun=L0Fo6L|` zDF&v@W-1yOmEDpBr%6#5c6hAxJFC((xaUA)NI%^kQWT~sa?$Z~=v5?NvYyGB{$?|< zXB!@YaJ`;h=k{h9Q!F+%5n)M1gL`FQE)7dJ_#+Ldg6Zmy9p%)i$wwhGXnUW!S zzT}kWMRE5t`)2>KaAfy+-e=vuWKL=8uq!zh`dTdY9to*5$RVYNx&|!Mh4dh{oVJ50 zlbu^E*X(BhtSr64&DNHVYW;njT+h^*0af|`WI02~{I-QG7hppO?t7gknPt)G9rjg!(rz@65ddw%UbdPl_7vRRlJ>f;MrrsHO z8+u^bbx(NSLkF#@%{*mQ@wK!4dV&=U|FZZ9MmXF9Bs{BnBwgn=?6hh66_?#~k*6M_zLi033QXqX@}f;~McfV7OTDms zp$nFlEzE(}fv;eX90_%XO`vrc)26(>b5{mwu6uRSSayXwofW|`A`J>Jw~ur! z-E)1c&@@FIkYn4DO(siL`&xx6)={m$@YVNtapr8@PcMuexbTXFle@#Kg}LsP7~B{# zEUpaNtE!P+3%e#>59)NA`9&TF`GaohyvC3&$f~V|$_Zrk+2yg`vxnXbrCQw~gR*j$ zhat;7?c`!c%;u2z4 zA=g#G_!71ff#@`%EJ~{^2q+Hg3Tbdh835SW+3og;G)sUdQx#{OaW8X~8aTK~8i6tk zym*3OB|S=*ops*ua%qrZgkh)MZs+M!^2;L?7mZ5f<-OzIz(cOXPNLCy>+P}CcucSd71BRAh`~_ zA6{dEq5~8I6VhHPdeF5MfYGSTlC(h(uYmO~Drd0-wq%g3ko4UD!phXBur=C}>NNSZIHX zZes_@IP32zSAP6xxp~)|!>*oK+WAKMYPa&r0$OC7L-z-0PB?eUaNTd;fMi@_SPF`9 zAGYI3PN0~2U+%1a(G_ZHXVlKnDlulwCP=A$jy?yTLpSfLv@~i}c&*Q%;*wjT;4ajA zwgi_3SvoUVYa&bW@nr6>_6;q3o*O>Czitn0?lidM`5iZzM4sVd3=;7MVfZ!$)-Oy- z>AATDErXExQ}%x?37oGrUkEwuOv1A0K^aZVA|DC3JYP32e&&$VX4idEta}>>xg{TR zs`h#8o$0@lR|30($sQ-dhMcy_2116Ml6+V4j`~;5E_cBv?deP`v;GtJVR~Rk5O8uJ z)KCBP^Vi-shO3u;5EoAFjkgeT;6?LBlakJrQ<>#c$q*I2U%cD@s&uQj9ug)^?xwF& z6-aNYr<1&U6|K@5etKlXj3)OXSU_jeTjs_Jn%qBuMf7QsosFkhD}$na_%sdG6&mIwDkpXC5#1GmcR#YcsB*UEpvyH*6@T}XRo7!5Zj>u9`x z&^5y=UXZzMc(nq2qEx;@usooSKH~fl z^z3!Yl1I)y^6mrLq+)*dVx;85-xz=Q7qGhX!`t&`i;OUQOF#WDB*B4KP5CCPrtK7y zhTo6qi_Cy*WoV-JC*-&amOzP;s=#<4#q3a=l-3GBe-?w|PI*R9mueE}F}ZswJG!PZ zLv!*{Q?I&&N5AM{Fz$WgZs>MlQ&gs;M%t^|=2<1s%|rfzN@G-pU%ZzzxoWrh;Sc=__d++qc3K^tuf3u3d9h4}!H>*W9In|J#eUyeH ziw$l5pq*cHn?f#J5qnNCMxd%)2meAIJ2@=D8t|o#?PFU{F>5IjPesRyG9}os1GU!q zA{!^!EjQNl@nHE_w!D)Q9;R$~^@od8Mm(H~{@W%}%B||aftSqnCg?dvF~AaY)SwRo ztsDFXNvH6!;JnBAfU_i7a89zuf57#OFl+W>ZxrioR&S8348pqx6noolhY*Ts5cU zpwkQF&G{k?YN6$cj`K^Fpq32Qt6-yQy)cni?X%OjgWM0n^=~0%z1hv|V=EMx4QusF zPL5~lZ~nvO)=Z=0dA;C2e?zuUBZp1Cv4s?qN0D4A`ue|ieCP4f50f%n%Mb8Q#fk=@zqx#tRy8Dtz!wV+wtKFhKlud*78~Dwb{=NnvQQV`JuMuU)} zf!&iOZDcWWp2V)>8rtS3zr8*y&Rj9VVS`r|4G2U3@KIqu#6;H>niO9wT1yth65N3s zQBBEb0j1x9Ev?zVpd0ywa_$)60!?ls0M zUHfX~*QDml94?;3fnyMsMG$ zGmOie&cz|EWSax;_?DT>Cj}G(61zE6G?q7m-#8ao?s~u*Wwl0XtslOUs^}2{!AhI> zh62UDTV(6p@+Ik>r&QOZXG!Jkk-H8<`=o(13b}W-22=#t!}`{Gm=XKi&gf%C;^txA zx|uY?$%vN1Pk-|+iQ$%wbKp8RkiQzku9-|R>nXC9iZ+n+03AYaJ{w>SZ=qJj){Ppu>9XCJWz>x@~{Koi+Z4|SW0xLw{`&u(dvNo$%`1g`K z?%RA@6`28@-na&alCZFwQazmlMDBYd&rKr7mu2X963Ma|)8^+@-)!|YpG6#YMZ+>L zAV)4F*;`~oimRd#ngpyX`v*)6mKoXRHhk`YoC1rfv0q+#-^J)K-qC#YC$jnltAEWk zaSK~1CY2%^kbuUgqoSyu>ZtU=;~pX{-NugQdIp%LnxQy@29 zI&X*npc}Z9blHGxQiRvIfMGJH33f1yllGW;x6uFB=4?uzle7m5_Y!R5%4;KZ_uCK982x(UgJgx3yky1TjlC% zQuw(EM1iY+OdxucVjw$mfQo)BxE-Q%>0i)0r&*oLKM4ttRm$Zq#XwZBZP9Y~^~`ej z!}L9vgSPqZg@9G7ELoBW`*b_p>Rnd^?U+3z?-+TFbnA({UiwogDtSmws1kWMLe?&9 zR(QnRu^eZEL{<~$jH@W_~fe-ZOjDb%d z#q?05n~L5^RecpG2!JZ5#A~N4F%+9`?#ugNo3~w950p7rM}iGuTIIcn@>v~y6wKM>OES{8NV$-O9SwTF*)Zu7-E>p?cWj?UqaKC{i&@Wh}y!f!vcks+S} zj&M?$CBWaMU06)#`u31|VGqIU8zS`)C5kpxyYPzm78Erlf_e!yn;np+K;cWVXB_F0 zmqo#EORwZ5!8XGcF_OY#>&Me!maS38j>l(Sc?zsQh~My`xzlpkh=;`u848tJWGD@} zDik#uYev&5mw7ER;M7U-)dS^4hvKY9Q&H3RyHbEH-V`#il~>)bVE%3|>W?;cun z88-1yY6Ha>3bU)U}5e4CHw``O2{)fH~v_yOUc?tNu6&S^^j9y!h} z&T(Lu(Po0D4=D!9uFp`>>Hb)s35#t=;_2waGe6|Vc@I0CpB3j_%{x33)P=};riLKV z^~ilF(2y@0c0w^KiyN%W|;3Svuxp>7hMRrLT{dztGdT;R`2(y_AcZ%s<-gg`FF-{?2@s z`kbV3Smv}G)k8YV`D)|UZOpV&Z=_h!q7UpbjhKhX7%-uIy#+qSCOVl7M!0|Bi*y;l)@mw zXf@U*03(a$A8YZo;L@ihFFG%eVR?NAT6nM?B?p+YF-$3P)A>{c42lc|O&VN#C-CF^ zb3G>vI9>~FrbSgX7lUM)+RZAZY2Ygmg>H15;(JGoJEwZvTg$6-@+EJ{Be ziflS&Q9(eeVv~nf)hW}QB|2Qc6o_^LH7J%;=1Wk37gs#PP8lFJ0mX1oC)|9yjbFFTir)`3^mES>Kiie+bjo_1L^^qQ`5KNCFJ?!ss#sJ$S+ z`R_ky{_h*V`;R~UAL%NJSwWGQu~@M;P{tjJLx9)PW%e8~sObWe;$4!JhVS3&CIo_IY2P6bb6 z$GRbK(JE6^DXIhW4!{V59z8<10aKl@ z6;Pk<9dzJmgGCj!MEdsy9nNQ8ZdS)Ff-RmP?fg96{RO>V&FW)rnRD`Z%i1Qlrj83* zCSvu8KEsZdY5cZpQ;kk#WqeRA8Fpa8lnfK6l0-2H6j?(>m$_+9N;?&GkZ*;4jVvB2 zPW>aM4vVZmQ4?D69J}?$tA|?ue?B>A-e=4u>~LT&!otJUrzjL0oZl(a`ZfmF%N~bx zLv224D`O&k(s5)Fh+sys0uK{<^cTmM{?&ZR!UYx%9F4Lpa2jS8%i;ua0+h(ps@5|o zqz+AANLyV;qrTVyS*mDAj<;Qx!-k6n89Xw(89;o z3~e}Uo*R2Q&T!y1E*2hZRJH92xyh?{ucO-(HM|(ek!|-ZU8r-xj`L1`*ha1h{>bCr zEMpal*U828&YGLhI9j@J&Gc&CO%PYP8J2Iygqfvo;|UN<4=l7V_uJU#9`%p^Tl$(Y z@+g|ayGyFLrQ;pgp&(QBt?LV%J9l1`Wt$(Ibe9mZNF*y0^_GSJ{- zYccXN<m1R-41;bAIh^)}0FW>k^!D~7? zQKD&dKNff(>eAx4xn(1r@EWXQLANft!9A1S_Jx83OU#O`fg|N7=uj|y@^+)L8#ebe9k7*?=Qp|Fum~20@45B1 z<}d;~EWR@L_m8cKW*s<$VToo@!H!?5ftY){MhfC)A`r0fnUXiBEHQ6bci8rc8QUktoUI4G4h!F$y#pfV+WpnfOlyO#?mQf3q?|>=#(F{N(+}DC+IO!EvT1Y5MNrH=^r0j zzWCOHee!%iJnAJa%p?Z4+@Qo65fgr3)$Ytco73j7@XfST{SnI`v7&<_$adiL%Y*4{8IE*UfHy4D?545n0vcB5RP*i%F3*RoFI^g13#GNG z)2*4hEc`NzosPI1PxQMmrTFrpA#>((hmB=fDAh3+qQUak7MYd@3dR+|`6F5yjY?Sa zwlKX&--qR`2cimq)DgPVUExY5eNkKibfLL)IbR>qMW;^Jph$K;Rz-#fXaK5~RZZU~iDQ;EtMP23M<2gDxL&+Zx^327=CNQ?)NN_6 zAP$oD=!+@?6M0p_d&2V`_)u+Njd#AN8(PB#UE>!u`D&FtP=|j_isTD9syx4H-*evm zFL%zuUiNIzUhhyO2A`xKdk+aJ#JRrLL)xG_zL950-0X1crQ=;LT0*H{8`yDiMR+eO z+MZlshuw#-{+Hf-{lp~?z|NDI`|%k^SBeB_!I_~v!5JyE6pTIFX?=Sg*-YH8E$X`ABd z2^y?F_|9KtI|W9^G%xMDC&?-YUUg@iIHAoHvxy?fKuybk>{k}Go34PWgMRqB-wQ92 zc6)E;SIM(oOZ}54siEtMEi5nNRT#qxyy$2N?;i-z2rP+(ip~m`(bZQ5q4e3Qv7GN`yN9&+lF7BQWzhXb==jfY$7vAmpJl5a#lNxV>R`#uL zUI9L+CN(q?dd7-Uc}WtT%Yf^yh;hv{-bZAUa~$k%llQZ95OxRkqYH^0=CjlbwLi@= zbFC9)1{aI5O3t7smg1{fCn|>KutM1>!BOZb{a;`pc-VpQUi;xALB?6?=C^K_k~{|v zU+PTeoid6!KoLD|8szyMkUbDwCcRSBGAm^s(lz`Z`u0oZ9z8S8f$zs2_Kexx;zAzo z4(R5!DPly`K$Q-C$Wc30&Cs3-+6&OX4xPwA+0>$lb#4gj9eF~l#Lb5Mxy50&L=p!F z^86XWQtx2Fee5_HFT#0mn4acl#Lh3TkY!}Ucx!S8PWSCG!P5?kfx@V5Q0NbhK{=vI z@rQ!M$aCI#u+rAtTF~TKEx7F0Gec7&tq_9rx^U1H)RMbG8r(k(-{h@bl<#6qmd*m( z;4KryjfKNc_%?Rj$d<~F{?UjLMUFy5KH(MxJMf}|8IrD9)Gk zLD515|4OiaA(ZXri!%L_1zO)U?-bwj;tC)e?3Aa=1{K3j`I1f3uee~l+Il7{wBC7_ zG}FIlgdR{Iaor2add`1y*r`vE$E%}t^EwpUopYUG-yDDFK&SD9$1%wT>9A8Fzn9)X zy2TCd(7OgDn`Lw}^w{j-UlI2yu%>*_ttPZ46z@P0092&GO!Od~y|`L+G61&Y`D>;_ zR|~pFBo%^0RA3#cQt6FsRo6pc7xpE5eAubM9l7zam%m&50Jx!-ouXF;E&J6(vH5%l zXWbqDv#EBPc|RkUw2lL>!&tf>d+Bul!%NnOHmY%(GnTq9AMotsH-XII`_Q_0QQQqI zg**5IuC=_lY45;A*FJtVkQrWkY1K5X5=o$&)w`kCuLCZl+SyV6%GuSj1i>XHQ@U!} za^8JNG1)qf3Jgw@m{LO4tsv;(Q)ATMOKB`rs;?QG3qQIuvysjR}Gj-;~*x77sgZcmDefd-jvCn#(d!us&|L3!FeP z_1Ygk{Eay=%L}#aW}#m>qB<6kAl@UY3@r44sA20jK^z)*a*qU*@HKUTDZaLx9G-7V z%dj#MMLd5n8{WZ=mG1p3Qaz1Pam&jW-XbaEt(+V<_AND8Ib~DK4vJ{0=;M53K+UAf zXW?|xsO%uk@_flnV2jiwc+`$m^i|2~-HS*ExgrKyRAfe8sX!Fj@WkqxZ>~&&}fLwh4f)P|PKYT%e-M7k|c&dz$Q+ovKKe zX;mo_a5BB_pg01oayw=pll1ZP$&Go)ZQ7yOJ5PgwEv}DJ71{GNDbOT~)t`6V)8=$R zEVEqD3Or&ts&;6FJM0ey8;yd4G7NHb^B%dTd!B_&_hF|tf@-ff_{&B32Cc!zBsHW% zfocTZz%7(UjzQ;}vN3q`A_IAEJDj}c*Ey?9(O?Oq_r#HB21HIT$AOyJse_H*|G>kD zukTl9Tp*jKkrETb?4lTm6*8#kc9#x$s-k!a_5eXl$U`ZQJ@BeSQOHk}47l!-?}nX} ziv$b1R|>FHs*sNhhYCw-1M>#*bwoMHuMf4H3!BgnRGLQbI7^CK}9pqpaBkq+*%b1 zfneRUJ#)WYurkgBVFenV{U|4_Os)Et(5@G~-m^+r>eVhxQgnl$OfJ2Ol)K>ZefeeQ zJfNdRGUQCZCZ^nFm#T@mwD_K+Td^UsS>2}?_WM{^%|pJ*d*R6G&?bItKL&6s;bK7v z-LqlAc9j!cs2{)gm+RAvp^3WlEe%=j!26P!CNa|nibswt*|B1fp`V^RGJ z%A!^>#bH^Ysfse6yYxXxo>GIH?1qMa6QGL=o{&@B@Y4l#nG|Pb{{b>iPy=1k9=Z9RVWL+XoDn?cTiR{>wRL7yK4(sK8|{! zb)WOIp5}y&XANZIYd?KIv)_Cp@N+WzuxuM*8%gED5pLvB1uEo5wdb0c&1M#Yg6(m; zo&RfX{L4#?lTP^CAGVWnZj+7!Zw0lOOgd*M2DCQoAYnYBAXXqf5}*&O_UTaMiyi`V z3o`!T9MPuOsd}eX+Te~iK$748E}h}EN1iNrATJ6-Rk0l7Yb-Qsk*!f|g9Y{j@H>}z zf1lA70@qcB`?bnV{;~d@vv8aa#d+6t-kNQS&RM`*h7Cw>zKJqsv7)Y!L|(k;wqoQI zmpREoJh4nN98YsVU+wgyul(AGrxUN=@g*JH@Z`AJW)f>sR(FqL?o#9q6+IBq5uVGx zMeygNKs}IpR7c|LM)mEWZJ{^RxnFzugS4f`e$@JE|96_d^tLAX8_nucX!E$C#wVJ` z+|{JU+O-Xl?IdQVR@uI&SzhjPXwelH91R78W53iY-R2w1G|t!^R`328c!yFII7T6V z%iJF5WRAO+6s*s-$$z}4e$^w=ceC#*ZsyE~CFsD=u zmF4OGL5L9^JAAV@lj6_KJoSMIm?|jd2t^K2(HFkmDs5Evc-0HvGc>AY&Ph^aO6%ww zF1^zI;(B?bdcU{|=q${_r=wM9sJbL&(4?fj*-&P8w+g$Nv?1k3q7`D4ya zPtj5gC|GT!qTd(9x{WYv=_7zy0fX zxP@n0(IS96?HrytrY$ZWeQQSWi(asDH|UHoTXow5NpeSMG7Xgfn8_b>tp#e1xViQ> z@EBuovGth8?0(289GQAQ^!>I4qtR-)x|$R^aI^|}lri*$M=1tW$q!J`*w3||!E(oH zNWWvfU5l&}G9wK_*!4*k6op~wBUUjQM9eNj9ixt4@7^ibLwij;vCnhv;9B z2aKx*QZwyLtdcbJgPkTtVbB}G>r!oVu@T6)z~^a?G2R*5Ad3@xrW`Ai{*QSu;;@x) zEa3&pGr)civc;$0FYdtE-7R0<$VknzMNOd>QmA!F*LKlQd+vk&CJVJgc$(6Oi* zX{_j&+kgyWmi+-`Q3;HpaMDJT2nSPJ2q6w5+UC|(Oxr2*idinpSqIMBv-F={1_sl5 zVIprxQ8~K@LKcinss#-mF`@$LZq@rFm*3+xq(~6#ruWHfyzvNAak>2M0a|6cf76l# zCV|PZD;2zjSb5@cOb%=Vi*ZNE_+-ykl`&X775%qOq?Fq_*MSRF>PJ zlDxz0qwFUYKvMwWY&Lyep$B<0O-wkZMbMj!5E9;WKMdsXF{0dn5_ydP)K>fz22O6uwvm&G|WMm-zIc`>*sP#&P~0|!j9a|`Ll6i-&oM({a`pf zAPGCRgr@S#m6w=41y(?!8WnV!>gaUYWSM^CBj;j&a~s+7HM`$PI2ZAO(ioknPrq@L z?BEuiIPjDM!N{1zTM5M!Q>1{3?xd?A^?uj&h~FpDc)|Oky8Dk~Rm57iWi5`Gx_bqy}Cv2(VMzt1O2P!W7Rq*>X-BuZ?KG%jcw zC){{)%GStg`hTZiG1r=L*!&rbdJMJ$q87vh(iYqp(oAXt3#5hA)BoArsO-kzW^VRz zY<-2@d5qKlJZ1mie({AUqeF4Me=?OE9FG&jffv9{CT^pSVrnQ-O+}Y_4Y)R|t7n(b z&I-ILImiDfpbB(CHoL|#2YC-b8KjnXGT=avZeIMfUiULWY3kL}w#r?8%l@IgU6Mt1!X+;oIwZUoz)w zVxd6FQKbvw7Cqo^;A=;upKdH{!VS5Ev}I^JluJICo;>bNV1nBkIZQscxUfA6(MlJtUwFhwSlQ6|M~ zr$`zVjnY6mUJ0NSbMrZLjs)^+p@XiU(6>XdV!khA(6wIJEyixbUUJTRvQ*}Mm|=7=3O(SzUhaUO-u7pne7R*nZ72j1`j z@`Eucl@yBEK#@e;@JZkoi-$Zql!iTa~kXgK$l|(p{6SYLOKvHL#!=*(}Uf=_4>@k8CTphmOVI7}rxg zIcamGu!Ccq_05$1=XV5{r}J4d+%ARr_#9tT=Yz|tJ0XZR`$Xr*{M$7w9nOj0MjVRd^i_Ir24FNOIc)WPCxs zIm`Jb$%B^l@>v?gE)O)4%afh$UqxUv4`;xbnALtS*Z~tiW z_x|l|Z>P7n{fG9}PM2w$&fHFKi@0wH3NEMt6o??AfQlQN8;&bDs8L*zh|4g8;KHE7 z|Gr635;>R?67JEzV@uBZ1 zRYgz@!_lU3hMa??a(oSVyNtewUo$@$l4U2bCsDeVHN`aQzrnez1aP zM<6mw17p&e7(G(JS(%qIRG7|mQd_6M)HN*g*d#6u&?Db63eb0ZeWFN8==MsNcF4+N zn^ic1?WXU>0)O1N%fS^sS@c8t>bTS6^gH5D$m-P0+@kfH z$+RmB!AoFSVD-2yz_pPo+@j12$fA!~1bqrZuvH{)IyT0j!P_jYeMA;01_aX3!RA?j z4Fn{1OVyuud@T5Sox%pPBj5PuC9;g0_2O?}l5Dku3zxKxD9@V;vsl+&;Du7J;oH$Y-9P+COo z1Dra`CR||3$lX|w*A(rY+^kBUvS`ZX;Brg)LbWJEwNO_Yv0Z7Tk><($hQ&GXKRd8E z9C7=*;Q!Y*6CC9;Uy{XU7Gcc>C|KPF`f?k+7b@H7OB2t&eu{1ddDcr4(^Yj)!n!p0 zQb0%4rvXj0CF&$k6|Rw{yFWdypPqCHBmK1F?nBd0{b8Y_0?f0sZ5}%wEIKbpvyVx1LT{0dryMe)9k^|B%z)Je+{P9rh4D|x_kk@-ac};mvat!o2GgQ{b zE!Fk=!crK_>@z|2Nuzwfus8tk2Wv*V(_=}}|eaQa18a7c@-qj$e%&>(GsK8Gv= zt&chyEQd9#ckbkLttASGlk_HfUFhL3{YRuEpk7fjtxNFe3!MaO1R8)|MUQNfAmlLC z{q72fEU1BOg)4Y?&OO}Va;>d9!ZA?%lDIc4A>@qcZL$Vff86M|D}Gx@EqzJWqRba` zfZKan)(!4W3UdbJl2f?G%zTc~A0M*CK>naAlNGMTC{l(3@iz_=rf<{{WmC^ z!wnGwS(sg-a5(Y29WQYj7mhc2cz;jxlbIKnbLvk-vxP>L~hF%Ang$kd8()Mu%Se39)X`m=>hZ4y&04-+vDDc7=G+2y) zQZL2i*1xn{|CIQPcYmJv!7qOI$9EHDwAzbMRMD-t_ZM?5 zs7zhA3 z<9vn1(KkaVvm{{KSbY__EGq;tH-jZPUcZvLEISc)DhvoJ8s!%y@PTrG)!_wT&&{e@ zQpDVcLNm_ZyJm#u#$gbRF1&2p3&2XZMWr=A+ft-Xm;MrO0Lk^ zB4Fm|(V1iqw^UdD`cUMYU2NOLrT`1NYwbufWiK|JpAGCJvgD{~% z)2-a;T@?6H2r_~8cp=#r62vuVIs|y-o-h;?Or|c&n#U$Z;N%_hWy3a;PSPn~=2zi^ zM=&e9mBz2Yfa^(eN7&6 zX+zV^$u*ymkNS(J@_2}8vBS(cN_~bRA5)1BnN5(zXqZ?C3oVDno`W(|;4)U6iCv<` z&t?8ggA2#E=xztzozxy$O*f3H5uS-PPN)%{(C&(B)-;LkPO2H32kHBZC#}&G$8>v@ zk7^HXP;8tqb3)@NgQiPxF|tLsMsg<{??!c}HdS{JUW=`)lv?k?t7_@JAz4zqqgZ0n z-DmDVjQ0%W1~ zrAWPnB5GJ7vG(HNp^6Q zzKMp_;{@YqOQs4WtpUkk3#|WSfGKikuDP3$!vPxuU&@Z0RIkZ(l+7wV_IrGSbbCR~ zL*K>K6iU5_BJ-)lWbsGdKpS00m%+>v!d{Q2ae1i*{slRlb>x&&rH@JjXcUjS$<~J)>ri=O(!|-*Fv{KqGTlsZqzH9L>DC&n0vC0 zSXBDRqATggs_kQoVQys+w}eTyzTV#|V`tE$#eN*ntxSuv5+u&{Yo(jOwZ-C~b-^2= zQPlyzLS6~ABqow1Zk1pTKYmq@SGBxbx%5YM-k%2^5FYx$QUNx~P7UNnuswwxx2=BR z$O&x2#J~ITZ{M+jZCX`AG`TvQ^w{<39ZLNvMQ&4xnew^9bU_V7F-ipHQ2l0gCXl7n ziRyqdxKdIWvU*%k)CFecc!OpxFr9DkyBBqHVvkodRO^(ej%f3vp%MqVRgkZ#QgRJa z&C-K$9IOhu2a7ocF{O+}rsZ)I?jDM1RQGrRrG^+U)i?O3NiRy~Pqs7wX%X;^c-x%7 zOR@@|i;~@O8)VJQ$*>*DLTL?Sh8i$r8-iccQg*yRaUdEQh&Iu!l3Mse>u=mfm(Vv9 z%^(!BOx!irA-uwauI&YaT^DoeTK>A8L7Ptf+)-eLw`mV7Vlv3Va8We}3OX=JXG=n( z8nuC9ssdK2S3yx)owAwWThgf7D#-9L$}5EK9qH^~(Pzg6 zN{<9NZ<9(MkAN5KtmO$x4Y~h^sl;7SE(BqLdIc^s9hi25{sk%>Yi^pcRPjK;>eg^a3$Xs3O(!YIzH7h7r|;;EoG`>NQ=QKDk1V z6QG8L*5}M~tV4TArtwf%(#LZzqSUZvo)7%HbS21Kb^7j#Pm3*wis)VOdYptJO$@p> zkSu}cQz9RPo`W&4hA9-I4qAg^U-&Ui=16px^0Zm>tb``Zl7@$j*Wkg$6*kUY{~ez) zkzkl3;};`^H+cXsuZM z@cp)FPiq!$P3v2=xN7>))h%*Hhg&hg-q;5!@v*^J68=WUp`5Yrz*9h}^C_}{N^BA( zMHI@6^jzh2sB2A0$n@>eWJhPw?EyVr2LzdMpgKS5w4zRtE?63jBa(4Kb@;=m4#B4A z3{{Kvu6*~aw_kO^PhszZXU=fON?%9jnbzEQ;I&Dw85}D~csng(sTx7rf0J&pupn?2 zOu##o*C!i-n*h~xYfTbwn?yF2hqSxMW&)bQ}1DE6qgWXwI!Z`lC9Ofe{S-3oIP{F2#>v2mbo@& zbDDI?Z$A7|`NWyAwe$hu`d~wFoqW#l?cv+PSHGIAo;!T5u$j3c+dF#kh{N=ngmbX2 zf!E_299{|ZFOVzIss_zgrz_x3++xy28RgRS?FH}AaOW8~*hdBaiY|);^1@gP1I)T$fcP_pk1NbaXoN~g3 z-P8T`s-w;xUta>kqN!IHk{rB3(5%eW;)3{|e{T}a37RWg`ZQ5=Zsxwk* zAcM(41-@da6faTkjkm5A9T2pEYLh{O648)pu6;-wM&4AWNau&=PTfqJ!RI+j4uXzj zv*yGoqZai7JUiaHfU90(F?jF{(8A0SH%128kTUzdQHx0yH>B`52vcE)lr5AR(jN+_ z#4I|WE|ye`Nfow&zE=@Ue@}q~(!&VIz@jT6_4~r`ASPxt%G*g*5HP|--if^;vmP&= zauSvj7psqG&&aRJR)iQdN3<2}PU7jg$6l3J==))?!h`t1tykXG9 zN}~+4g(Hk%cjX@G%mLm&r^{vsYQq!fI1M?&e$+iGVw%m6eCyy3x01d66~Xb?r#)|H zE{;*^k0^48N<1{yEXs!UoZiq$9*w~heJ$VztBhJ`3t7npf`iN&v6;>c{cvVDy0DnE zi4u+w*=<%8io1N15;9bmLe_vjBN8w_@`rqx2Bstcvil(G6AjKa$iQk<)x`C9&GSe9 zHCH>&-$*xWAjuQdxMG*XdQ)ZO4j&h$>F%&4PmBut24@tqL+pnKBL47}4Necf{FPsj zeLPN^x?%^NQ22QVo1{pufJh%|Ngq{rd)<(g(KB@` zrPXld8qF%uJj$o*Mb~w;7SSv`Q!uVEq*anMsxGpXPKuZd=`N?_1tBIlQzyFa+o3#6 zDtxT1ADGk*Y7+&_g19rIT9oVM58~R%?rBAG>skCwmZ&Ryw!1>wJTw5}jGtlb0O}`d zLpA;UftO=g@7drlkDXtZ1kTwB2Lf}Uin~Eq{q0MTw}859t2O!B>%MpBwZWO8xpXqg zq8|V_7LQy#u9^(d5XT)f+yUUq9qgd_)3#&ZV{BIE4OlI-e>g3p2o#m7|l_E=^ z(tpgsaFcEqy)XO}ofg-uLSeKl=}B^6h|+fZ_K1@R=zGhMY_lVvKeKT(x#N~2e+Ucl zM#EPZlFD%x{v-2?8^31$Qp2>(|1m`Gn!{sr;=w!IF*LiJ9z1{NmcLhYhrQ_D$!C)L zv4zsKSQAvI7$@utTQ)sSpg%{_<0~Uu=o^zWN6tsmLvg+6hN6AkF$vbX>;fm%dPaZc z&82Tt_+0qT(l>g%4u%)e>A}h3k~nPVm?TYbjJXP7!Zi9MBx~cp-Yx5=kv0JT z>h4hF^6T-!dj7V_`U;=juW`?qasut_PAiu)9A6SnXdiZJ*1WL6A|h^KW2)%z3|@={==WNG6JEVBF z=ednDnu!-yI&vlQ7A7q0hgcLl1LrmsV=Bfh_iv1RKSPxkTM&{RwAOo5m>$Oppbd&{ zlppf;FeaY#EN5^$xzn>7oWSu)*xW5YebK>@r!5F+lh5_t6O$Kk`_=VwOyu4q+Y{C# zIykKoYNlHy)pR~2BMcQXpnti6!?S?s+d3z3P={wFU379HjOvz6B4FvpP&kUOS4|nh z$NltAbmqZ7y?W@jI6Zj&diJ}2lXTe2=7$!2{+ka;(o5rq8tfdDR7$;sA`7WRi=qr# z6JVy-7K*P6=E5Z9e*@aiAloAgz|VO}KD;({W2_AlKPcanP0Dy2L#?;#9al?(QS0W)Do6ee9Ki?2`I$v`Pb0n=#-uha<3<;VpB)$KE8EMkmMGLI6#`WsBL zxFxDyf%Wi~uW$)gPT}HN50%rA=Rly})0`~cu-XrOMo+b&sNvsc=aF4JM$uV26n#Xg z4^d=4m3Wqc47Ruhs*}=U=B_+R++r(4o*hytIUH9nGSeLib&AI!?Q~l}ov0KRZcMri z6&|^&`W(tH&Pb}2`$QKQ>y;+m;kd`b?4U!C&VL=QJRZ@en(bc^Y|^zsZe*wYkZ-=X zNtYjOrkfdDGlcbNG16DL1R`4xdK63jqQtp_oUlZF618lqmM zOK7Qxvi=*)b9C;na=6ObhwpHk$9FkBeEzyFm5Hw%a6|=fr@^dnwYVeoiV7dpz$uHx z0q9?FwR;7)!FBGNhsq?jac9ArE59a7cs%ZGvKx2SQR+;JtfUfAF4r>evVfyz?Bl>D zScO^Sr^iY_44y;kd=0ZF?1U^sFh2oZ4;NC3?4GeRJI2XmIrqYH+&f5XM{UW zfw>f<^4A^`mL=^}VTY_)g)BO2d_a1u3-}C&d?>^ob_PF#LxO8}FFH@Z>!_y61ra=6 zihhL8(s#t}ir)bI2i0_?B%c&Z^cab+^Tv725JJQmU_3P{u6432 zcWwMZ^#WoWN=m)5kCS=aVk0~@6?t~RSV^hVDUyaf@?`|(xmQ)4#1wcja!26j^2$ii zZ2&^(y69q1q#5EMAQxn~(!;Lav*6I$osOaoymdWU1RS#HO!-G*T(Y;O&x7KN;D9hg zHE(n~4Jjb8<)aq@BZ`HrtwFOyeTwcR`lBQ#2zgQs8mvj)B*wF-#A}grtPO`cLF5bH z9=1KqEW#-)_e>D>9^rz8r`CdXi~Wl@p5f+xoEr9nW=E+R-Xe!ZRwhqd6gw+qcJza| zEy@RRnABW6t(eXVD}W`tnxFzuCc5Zj(iIB!P0NRj)z+x?$<3-FL7iwtP`B5tkUU5S zxgBvYuG?#VbPCh$wJu`CsD0rEO&eVf3>O8=A+nL)l2EGZ_L?ca65Z`}TJag(?KL}O zNi?L!X@1{P?@GMFj;T$x8FQ6TgSfvWR4{1f zd+&-zXf4#0LOoe#VB;uctu$y*ej+7u3sfDJL8S*&whAgDt)FNjm%|u2TWwNS_*mIF zk|Hilw0@I8^FM3ER<-ipFGEJMbkO#zg2o1RvW66u_LvpnE@?!4E7yZK82& z;P=>lkCtCr2HSC`cy;i;G1*rSNswBk{3p%nH>!O1m^?z}SI42(&V9P$Sj2la-!}U}?QrrbH{X`Oe;HZ)O5Y$| zH>G|^kxnYnDi5+=Yh-Rtd_XEC`@*}ukoclf(kkipst!ME+p1I6j7=vsV>=R1RK%o1 zQISHyY{no!RV0({vZ|ChAKq>e8^Qan*W&El=+mIf39`wMKvJFc%xPUWXf?K2_72BY zO75$ys!8p_`3!znQFMhG?=A~N1*LAUI^`W<5h%3b`6Q4?!QU)}G5Xw!Dh=BiRIS_z zZz}#y*48T}HB5yM?pyzY8s-5>i>eB0>y1fsdynkBR?qBNU%tX4`sI#JFK=BM7B@Ca zdUMiJRq9(;r#;(l6eLe__==`nRZ0D4l z-}{=o2zxxN0!Q+)jkW}llGLT29foLO(O1!E3d}3})j`W72c;rlP)IKsPO;GD@ z`g$E?w?IZph$+r$TvC`Tk_AF|krlOcmQSwl!>E}fJ&hI@aAGwH&Ud*>S8~J2FK>(- zGZ-099*1eN>`;|PsZ%Jjh)V1Lmdb5DvwcoMW)m>UVT`!N75x{NI>+uN=MZFf%aH7Z zn}2ddY_lEjm5um-EatK8D75S4wUl}_dRK|*!7xwuJqZGLZK77%NLv;Ku(WiocfR*b z-7!!BLOGY6u4(_c*bUBgcpt~X#qZ&I8z+-7toY+!4v(9&WQiNS*g&HCUi2@qe8 zKMV2o1&S0{?dj#6HtD*2O`sfBGxnlSg-^LM1ywa6GJn_iF;LuW*S1L9XzDq4QJ>y= zc9UyIuwT!qi}*Xg$UMZPInJD(_o*?)u{~a9dijVHDgGwLv5oT2B#S`w>b|z%HJcb^4YSxko2i!9 zd0TISiy1xE?Kg?AP5h=`x%X{(_Wg172>aU-@+I*XLv6;Z_Kjt`NG`XjJCC8c&&~#K zqtrm|UQ8u6Xn+q)io#TnRdq^0Pll>8GA+bNZw)YL41U$&Jzm)iXhJYW(GS4tB1$@9 zFUOsD%>X=MCq|tZ1?%gowF;;=I7G6U1%7p+1%7RCYq_8kVCP2Di^k7=fMPeLuAHg- zrQ)ZKWhgAw9-k|lM9bd1Ec-huM5FR^lPHI-klulq#|A+SlSiM5JO{}u_ZcYpffZyV zq{a910V$I-GiK2P((Tnc=>gDiZD&d>EKAj*(ttdA1@yz*$qz^-gUNFR7ENiqH-$MN zen58kJb3O3o)dq!`yRjjo9o0^4|oat``jPH4sSWU{~3E;8gDZ`zV{Aikn-W=wB1`< zOQ{b~q=rhwH9RXf1&kvnnFnF(<(VVTKsG^>$Y?21FBC&LWP0#Pa9}e>=4+R}fv4w- zo9HCqsD2QZudRaQsUEKe=D2@u%ptO0kON@DR&%wMMBa0x)3*tTR!~j4j>cvzD#IOu zoFL4{TN1Y>VK8sEr;udB)qrRj=p-j3@z>a&nBA4_C_KX303eIV$aSB)^gYGi&=Pu| zFpcaBgJJ6DCfyZTJ6#K^HI1qVq?W$&ogLHizi-mjN$$#5zgpm5${eO^Wo{mH`oW8f z@a?B1_a5Q|U+O*0<61{=o3|)o@vK`QfdSb&%3^l}rU&QImb5=32wyFa*JsL$1b2Kv zLC+!#&?R{Eg|b*9WZGnpDU@c7(f4?vPW@^aTn(DDVfo%YAyq`bQ`YF69i6AeN^&Dz zD9#uGmnk}wYoeiI5~~-El1h5pID=+i*qTulG5f;N8UjO*A7m&S=^Ca|utbf*!jPvJ z+d^ZO51LslvfK~Et36(*3(!O#iB4BkjInZzv`P%YNHtVN?m;F{3ytJdg}{b|Eu;vo z=^6E!iZOU+8Jq`y{6+Mv`x%Lc{rypS%!?j{-cFXNk*X{!_%PiVsmI-75E96vu}_;- znDGZ$8A#cSRNv^FE%iUgpk8p*6y)Odtg&*@P~kA@Ud!ar8Ye^b%CC!OfAUqEWAsYL zuRb9Mc^q-PX6Lb;rqrh>a*|546V~OMXq=Z2;_oT^(NU# zun%T>MxM4oV~$y(MggFrXx3y13fs%YELfwPiEw?w5d?QsQ6=ExJo=}L-0 z;L6_hcXrPKKWw|4a5e0Yfy!8g4Od6L@y$zQ8MoO!e*Hk3OVsx#m3c;Vz`he* zgOq(6p4c0`{@(s z=#wnAQAc=o1nccm7vRx*t4BW-U3Y)vxI)UZw1}}(D??GC7zjHW<-kE|Nzr_NXddCy z=iv0ie0s}}Y;iK*{GEOWzL`5+V;fg=M}NMTN+@+PMK*!IB5kL)2(p=) zu~k8b#%}Q2BwG!$o{p$4Up$KLgGrYIs#(3~VOqje1>JsiuCGb=kl7@@6p|K~s#+mM z?SKr`Rd9uO0LzFcfbKMy2l6#M0B=rUrcUmoe&y(kJiGMu;&~hxNNiNi8NY8FN}HH; zS<>w2YI;RTj`YgNWB*{l1dpgA6C5T38>ijvWvj6Z^Q$8 z?7#{9G5M0?VDl_U_-hy;ZL4o8*OR0h^d~G~=O?5Q*(30u=P}^eeNJM+@ivd|vty!h z+uQGsvJE!osZp278XgZeyX^XZ6Qzcv`8+D|ueYXe`T5%E?KBFKZ<%&YwP)HrWmm$w zfO*m7FwTIKI2`|4qum!fsC5 zrc5P6CGhFOg0+WG;?OKRXqKk7ra7v4@^(>xMV}M-{UP{jNo<)DcuAHlK0g&=Wyk}7 z(e;_)+^IJwT?)~gwB3{InbVN}mn^o()|(-F43rq-as;gkPuUUZiNTo?hpu$vE;}mv zF(tzmmH*dkjTEP%??CQ*aN>5!wF{!Ztgy_>$!$!0h` zd;WTOf4TYf0LN$9s1$n=JU(g^+Wh|-8MV3tOn5v#ZiqS-1m z<3T&o7dp=4w$}scd$;Gdx#3oSZR;f;o3r{i{h2?L`7cexJ>Sl?T}7!gD6))7Y!Yqz z>Sfs>^6;(8vQ|}E92U8?!y@luVYhO{r~`uQs#8O&0?P$Ro`cP^Eyp~Z^x6NhLF50H z{Hlx8^LQTfnO$dJrPP-w(nuw)l;TqG?$DEg4Vq@=p5QPPB>i?Z_1pWC+Q(g=@<>)o z;~p%mNRp>C=I_L6p&e}W_y5RgqF9i4a@qkl|Ic#@71=GW~LTWj?K!8J{^JuBfu};qD&Qb zl16!Lzyqa)Afi;d$nTDCRnU@=gUR)DvNK!-gpK0+`kNG8{mbF)el`p9%Zub4 zk~W-dv3qHAC^e`(tbrOd&@4oPr31oN#rj| zq>Q&h)_tqy3}llz7-pCMt^saztt$p-&`qbfbp>zpypz|SzUMd!@z#*Aj5yii&9dC7 zxmrwUTdm2XH~Jk0-{OMgu}qJI2`LPco9zqBX7&pXGOa*GrB4rW5k=-xiATXnZj_gXrb1E5L#EiB1)5yCO?@+E+8)?x0Roqad$LmeZ>3IG5sqo`;d~pmWZ5N7%VAy-o zgU<&>AFJ@>F3)dp1IBw(sb4x4fxJ+Q#Vmy()(WNhKKiQg6s(fd=(D3URMw+CURW;M zrP>QLt`%dlqPS50?n!<>iwBiQbLC6Qpg$zmq4Lo4#}g-Bpycq$8PUq zFFf1r?AD^cDeuh8w_gphd8B{Xa$qUhz%8cFV+&Ga=f9RwY6z=rrV^{d?~ARZ492ma zfe#DJ2Uc25WE(#sPf9Rob|%zCW99snfSDGVJ)8>Fd1uir+S-Zw5}y(u7@3-(QlkS{ z!B>axOh9(lF2B#9DA&y+l{1r7naAb*UpDy1%jh- zS>xJ8ExOjgT{7zxsPK18f~?-V;_>QbFqmbkaACJ|m)5xI=a z2APBc%?d$|)EH)Ma}WS?ADA4$h5N1pEI0lS;Q||&yh6Q_$(;V;`aG`0*kA{l3`)I> zB1@@6A{0|r#bhH^Dr3HJI(#~;xcU+ z_K=x<;6+bVv99Jc$}dV<{q&zIlF1#qL!|E>e}U-^NgJteR+$39F*!>*RW|=*D3Yx> z!Q;s=z`3_M-@tAq`f+Ms`LFK@|M>^oi1Nq!Urr#OaC1+1Jk~6<<8XdNsXwR4eJT-4 z)X(_e9aARR6_11ld&B2O*D2O$mPG4uvq8}jm8-?7QQ$e4712(g@XiWJ78|ulA_M$o ze#O8QbCwvH(zuei`}EC822CeSukS?_MeFmlt3umBj>*WRFj;gKsFf6gDsg?}5RCeY~=#1_5>Xfe=11x|O z<^*Fm2#@{somJ$H8T4j>%kZogg$TFdRQ&=%noI*`V zP?sM74r4Z3lqX)BaSG(8cLZLGGHD>PM?a3z<7!e$WDVx`B3mnpsFyGcr+8#Qbj`Fq zv2m6S&UVX*8_sMlP`>DV-^fE}9nVm?zFx6!9Igwd2Un3y`2nAGBRUdV>FN=?_o~E*|4To!6h%-a8~sBZ$2bRJWj5I^oTxu->H;(2}KrCiC7zk zbj%G9qeH@rn-iNEkG$s3>nLX+@w_%Xah?+-Uis>ui+bL+LE`kp^TWuw{^ra)4$eKW zgUL-weVroLsKl*4En0-i!BO|)^w$Na>6LLv`lLTT0(H#hi;t2l8kUCO|2V$v4?I7? zB2|sIBt@9V0H0olDjT#~s>qSJ8v)xUgiehfFNTbALAO_<+!O|mI3}0(0EzLL$VPdyDnnof z?oE@r7{bj>K21J7Uf41gLmD(JGkj<8XFuq;f3H9Mvg0pndS}ZB8`f_8*Tw%K3wUgP zjCQzMO{pPbu$)T7C=8I@%I1vU5}`+W&K6~=H#nwr?$jDBlJzvfDm6=w_4l*}e~cRr z2LFA|w!n^q{$^T#J@eg2$AU91t}c%Q!7Sxz)$#)1eqBgrig7~HBWKGA;ZZS;1o~vdzwlW?ULh)T>B&J?#Qx>Tvc%`y7r5dA=$px&&0C1X8PC`GEd zve-lK){jO>}cyqWk3VV(>XE>5)386#%Qm@%K?=$#J|e0GLt;2mx_z1sPW?>mNy zUy>1ur2?`^v?;D!P~cxJ&xNI?!}Q`2rvObR-L2O$)h#i`2^GRxdfxB~;r-b8!#Q>` zw@2UiYOe6;d#@+ABj5X;d7RDK+<#j!m#pVzZFoGzsrIZ><8Hyl(mu!w{lrQqL zn1kYx*}`35X$qw%RSlyP=GZjK4b1ki8R*|+sP<@4uSJicTuc`)^v{%+3-Br!%#I4hZXMpZf@VOw;YJj_mHe|2jy8qmW!xa!6Vgr}0Zp?GS*Co0Y?Zt- z%B0&2`^CVNQVNtPnerprdC@uA3lsmpe;GMB3jCcq{QS+oEOw0OKAXGC<3%PGRmU@t zW_g8A0h296&*He|T7VwoH3vuG8c>fH=IL4$=zF{>gIoNnhe~JoLH^6~vq85scnP@-Q=%2oSU$a3u)-IGzwqpG=_xwbcaz+t zTcG+xwMtt?uau@Fc-k_$f*{La!`k=mIdQOFamKL5QmKAl*gRobz+K;KFvQd=8s$l1gXXa8Ao!*j zqObCB^sH~6a0Ex+TRpnX2^_D~uKMb`BW)9$#@EiiN7A`XaQN$Hky5(}PA;X+p@@M> z%#B$r!~jT(=vqKSOf_A=)X9n^Clx0H5l5S46_J~OG5;U$e0W;1VnQpOB5j3W?QQ>j z&~vC0G{ksztlkAvnlmDieyrWO9hX{ z-2qa`2Ho8^uTDFyD+xfEwq$X(CRyA`E|I#(M(>T%b>7`xD}t^_PeH<6%FwJqyVB_% z?QL7*oGb}t+!p?wAR9zxzc*?z$?9*-na9(b3Om?rq0~hbDWDQ@=(#YlbV`P*QGPeJ zGSae$jNy7~@@`q|r%_cw_o8N$FpcV5?e3Tvo5btY%@BjdY0pK;{KS^LjfxKZaII(g+V$+V;=jthUwNf>beR`%ko54F z8Uy@|K5ze0O1+pO3#dek5dEI85`lhRpb=7tj%ik@-9HXIzyF=;{pUC1_AA(Xzn^LC zE5ZlMK4xr)_`^r{50hChn7;wjkv=F{PN`EVvV=-}9QBVbp!36}TazX~+Mu;&&Z0ip zkb}i3q=4h#b|^fjL3X4he7QZ;aq!@6u!CiwSdq{@9z?qYhTt0JNO-CA2CT5vhIbAv zm8i7`oCCmfddmZ+*#XdxsTlUxx8IHXjm=bq|Mt~a$@Sr6wq2#neM;Rykv5RVVLqEu zHDZVGwr~;nY0Z$&gKj+d{2i(bsvC+&+RtJNn7yG7LV#TliomxBW=Zu|RY%8{2(GGX z!!JY`G>|GN-4xJ7SBq~?yrpV_jMUmtT>WUJZ%5QAw~a*(E%?|Vs2UH^1xCL)I63ro z#L2)qWqG_=zFwkQueZ+0}Rqm5YsN}g6VDTSJWId?9 zU71`u(m3I`W~I2D=!@W+;L?$Myjx*~|2|ZqpA1|Xb!bF)WRKUfZ>;*>r|%s3sm+u) z*T4Nt6FhdvhC43H9sG3#mwbQkpB$|bm%oh1(KQy^RUXl%%o7(&()>y!ounD&%w4`| zArJp$yPg`siB}U#18vsm?UTb!lg!~{r=2fQNU0%FITs10GegU!ZT3NFrQE3)m)tvT z#f1OPk7*LE4hI5QW*K=bJRnOKm~=J%&8ow6i>QjJ6CNU1T?5j{p;`2;$r$GJoaZn| z3-a`_xOt1y2*Ka?EGyv9-=b|`QU14n9ofPyM#AHG+i^S0S5fM{6xl^3W?2Y{n?x;& zw78B~)c0=C=+^|d)7fJ(1lviqXa^8*cL*>U5tZunm0|1En67wD(dpZ*#3V-K={^~r z9fb7$Cq^ZWvSuCVGehrym^jK^qAYGL(YpP|UdRAL^O2-MlmqMN{F zhrDX6Ew&!OOvfIt{iCl#kycjdIk`c*C*(e3me(s%q~!^Vg-b`>{l=s3HmYibt*UeK z9xqJ7E);(zS!7AjhScj$c}FZJJGG5IAk>#=p>ml(D;i9Vv?NB>E4GJq%FUvqP&Kdb zQe}b&xYbW;RpgM25s)~!dR&Vw-5Twkb7QwFNMx}&&K`COvKwW8E_l%?Off^{;Itd0 zECw^%ipH0~nlO}cG)1M78zhZ7_oj1QAoa$ zBNo}+Tcc_~d1%F`Jo-{ddjKr{C~rqJstW|=f<19~%%J&LfM-t8X^{G~d&;i(gYr5^ zJ}8Rq5|~Agh1H_cfDS=UP%FKgz6-PZ6M`)XM^v7`9tVW68w>XV>q$#Z$U(Z>@#{p8U)Y;Zz=!@Zu| zWw$X8l2`91HYcXkEBiQ^Hyq?g`-a$8QtEUH>~&(U7E1GwOc+($8zUDBnpL~#JHFTa z4ZwI)60mKoC)Vc&cq&{T_5jVe{S^Dd?kCF1-zeONd)pT})?@H?O@O6P1d~cI>Eytq zJHiu!a*$q6i>;$qjhsEIl-Va-DJ_(i23(1GNXo-1!-mXGaKSB(6K)46OM{$nOZn7( zTN7$C9JOyO+eLDDoZGg~&Z2Fj)R0zPOeG@YM6$STGE6@TrI?9yF#IZ(V6}N?sI0U} zjcS7irMUITj?t=Eps-eE^>`hLYo$AV>myM;wNTukfm(E5{T2aojAEt&z)ycLya?#y(&BnzlEuK_-w4Ms^$D$8 zhUz4E!kqMJLGbs|R9FNPU)+UfQ2N`vcypWYUXTZn*oRjWC&2zqn^J#9VNjUeF zC64Qj&(4o{ye7c1syK_Rr&FM;P(R!ErpPRBlmi6;{Ff}oeJD%r{ci|7AJ>4wx!*b8 zzz&o%k9_hSRi(HDTKMbOvuRM7X{%z49y9J9$^K`lM}j2J?ILHO^4vx|Z~>$E+gmSs z+2DC~2N0>DzpCFL$WYyS?IZ6SvTXEjjlTKnR)s-RDZ1;cUnblYqrVjLDbTQ?Zvh(h zhhGJvIM0Ssoay&xdX>oMlGvIbT^ax>?LASY0g%wN z7{=-y@k5@|41cC`5B|yX7_tpc51zl~wtdT~6p3xdvtZ4YUy~&~p7Ct5oAIop)R`1n ziTQm+%t~p&xW2-7q47{CB7O&_B?g@FZ8l+>AlbTSyS#d#wZ`=TKFQF44tEm@+@j<$#i;v`{h zKu=gJL^x)c$8-|B9vifN$JQ|8ghti%DHUT1rG;XMl;i|$lH~|m6*I+Uy80h=e_7w5 z>hfy>(mqR$2!w_$lTh424x3DQ8m!I^Y}TAJKz9A;ww-|h!KoklYk+?H?~>QY+5q-g zeRKgayfk?Vko(l97@(L^gZ&!8GjSBjRXTcQ>ViY0ZB zRV+vv{(y_w|2|IPU1fyd+Z}7FVolgdsqG+kbXQDp*!z(jKBjD965W1`&?;48{JN^$!DI zbk(4bB?q7)#H_4V=K5x++I$X;T`ylH&kHr_R>|Ax3tu+@eVChH@PWI=na=JV=LT-c zy=OTMd(_{*Hg~SCZFJf7*4IywMLc$I3+%>=HI#Z41-+1n@$w8+vkFsR!R|N8Yndhu zB)8%Cad#a4#NuOX(0!;N!UY``l(-iJtgUfE$1CFR-22=yy@0m?dDgrKbYvSW>w35S z3&jseE3ERh%QB*kbPY34e0zxFWE_ye+KZkz!4Vm2ekOX+`IV5&vl(<(cEq2AP}%07 zvY?sFB49_!W>O;eh|)q@=^V(Iy*8Bi-~bM=8lHeb|7~&t2URoj@KHx`7A_trkJoBg zq+Bwh%|LxpAk&wI9@F3qe4+o+$e|i*`gWO90P1_g5N+Qr{neWx}#Lu?8RrXWsS=2`#TCEyd;K$MLr=#dNW~R{Ql7$DhO`GG^?_`JE3gCTA#ID*#<@h zrFu|92fdgBlX3)w(o+(Q1)}HT@sv#$Vde3$ss4Z7o;KVzZPA(E)03Gzp0;G!Ot&8Z%nOwV!a^2n zrmKPDb8YlmC;-)4`Pb5eF(1T8qwGsKc2Dkb|sVYD(bJ=jW}Qs-T*yg%8~1CLY4 zc{_f=aGFv&D8dLe06o}XG4xOvLWNqF18p8UH%PyFB)zx(66i85O4MJTGD^>+Ri ztz4Id9SpDyFb%)hF@vmrX;xjz?S`WQN)0;g8?4#XR^3?>b2|RA&t;z-ffd3>vh!0< z$8YvG(&m^`k`02C$ZY0@Vs~^_*b={OW7kHfNpS&k|LC%Sec^@Tj1iuMz01Z1%}#$- z!cE6M{Pl>B9Me``sDXzil@&AL>n+P~yI@Vl5L_(TASWx`ob&etM_-wQdK6jiG0oVUgMs z3&$`|0js~8LGs4^ee29?>Qw(b(~z65VekKbhQn9>@*lsO7hxN9ZhZT8DJghqMx9!_ zQD+yW-a(NvDzQUVz|_*WfqD!r36?=vxJ`BhrU&CvZkDuL3E~Hk?yRqjM1}xZ*1;l) zEIKb>t9YyUw$O6_`KE7T zPLN4=BW4*{MsA3*#vUc}b#=;jPSH1phA68~A36hpp|lXazZyFP`teVxwU0Ngk=p#z z9MOuaB!$OecTlqE<24v4^%{z-q7rY>rnozFGBEt66T0iR7%3$?{_wKpftxBwGU^hj%I z&-4$QMG^@IND`8Hum4W>9&9Jb$HhMnQpx7qqlov(c7J8i4-s$nCuSFSx0f-$s zRRzEKmcll+9QnpKFOg-#$yPfPzMfLUQtDbNu_E$jAXYO~0xKreFA4Jn*Cc0|cKWpX zu&5fU?>BvQ+w@Aw(%@>kQ*I3nl>O+VuX=t>23{Q3p8Ip(xItgQ{@nW;=LuhZz5lBJ z(eE9{mlw*Gu;c<@dIt)OH$dtKrlw+|szEbw-PNEm(QgCJ0l42=J$jHGz=w-wH^1m9 z>`3*D2|g=E?o2?>W4m(i*dnG$bV*t+K=#8+(t7D*W|rWNaOftm&LQE+_~Fss@7p*# z68bY`E5>ABa%6&j)?)BDeY$j4yp8zVNaw1|P!>@r zt_rslML?Wv^lXLa;lmL&z4+NxWORPkjPN){nrnxf6_k28MN$DbV-CV_vq!xoz!dwb z9OMHuJCx}`xiF)-syaW_pcx8;^ywGpjYprGhG3H&EM2q0zvdVzWeG@H(jjVN=ZxR# z10-^5y=RWzp+tg9NRs$h8B%X2vOnp5UQ7NC+u7qfq8uZoTvlWG`>AS{}N2y`x$)XZ*Drr`(Nto*kOtuE?M!yRT62@2wB{8y~ z@6Fw7l%Jof|1|1iBx;s5Fh)96c$V?7V>p=ZaO?ATXpB>L@Hbu^+vqQJbXB4e!hJ*TN-zVBcsV`IHLT_3grsGvgD#m0>v(?qYSz?gkMA5NUMMo@>?$|VYqF4Bt z)K#Q3OpgWC`$5JxS$twtHj+&xi)ZSNjlpU)E0uc{!KRi<_xkUYRYDd1l4yO8SD|!w zXr&}oSQT#h8K7^XDrEld-f8Y4yGvCR4qvJb^ z-dOd$)4FRw0=`*LC9GD?ffUAeSxsEKMJw5)YnL@@&7v0NM*8l!8VKa!gy~>-XVidJ z&6Tbi1ZwO|p)21#&^_#^`E+yTnlZK!OuJ-36H3siDIM@_xgWK zf3Ci#cyCACtO=zN*)r>AHThHTnFifyU4{ymzgLIXi5xG(B>GBVI9Vl66QZ1!+d*XM zNH-v9AV9c$&NpE8PaGG2xVXYR#tX}`P@cBJ2Xf&>N#Z?m)mo5GgPizkV8Y4@ZKoR) z*}xy#F00n2fLW;$o(#0E6?p~|4xQ!-L>!v%$ZmqyFMs1W5b-t+!!qck2@JuRq72p2 zH_njq0D~q|-lakZeV_EKtO-^J@fggERBJ$qZ=P^n7{=r*GQv=o7qwf9KOgx+flY^? zecZxVXOCLvpX~#Fxp+SObT>F$!GYaq)4StlG_Vgk{l%SsIy>Z>j`9M$Ma^2fp(TY< zFQUkND$$x_q6aR4m?f|ZQ8mTGPAW@>IRglmTRnQ16Chssd00mM^J^%099m(ip|G-5 z+tLe8kZqbyVbUmf+XQF&(HWR{vdEk5cVvgjS8h-IiX(G87w3b=U}52zuZV0>qHAIB zI~&%l>Q?UbE)7j34Gg5mPB^B?9GM-xK!xKDl5ti|87dSxg`D0|1wYxk+RewmRNwg8 zYqrVEt_dH^Bb$czH`MSp7n6KsH?HiV)H^A%ol2YonnkU2p|nGABj9YXzA7BEqtk>R z3-pJ$XCt4 zmJL9!SHoaV97_CQ8}1mkIAf$QURbvW4;XgqqDY(BF&L^e9#3spsM0q2?TX(PGI#p> zrPD6ST9o;Mj#!g!qhAVhhBSTi$~2P>x9&?G$;zj{|F;z8gWii8bl0H7=Wo|kf4isj zU;*G@AvAcf^xs0l4}SK#WAX2^nM*t#Y*@r$anZ-J#Df$R`kJwi0}saT(HVXyXD1`tLP2A7!^Ae&(Q1M=yo9GZ_}Y1J}5lM0>w@8P}X5Y7s}I?y3$qdiM><B~f%L46a-7QQvD6a}B_$kIJ|-#5!SGEoi@x(w zaylMuMZKcS?@DwvPE$bQ*~3nU3ox|OC9stcncGjguB00#9WO0 zusyDTsS3In`S**GCQdSN8&WU%HX}SqVXNEC6Fn*Df+o7D-MsB z^RDYS@oX@J8%}Qj=m%HhZSfM{dxtYf`Af6-d)h9NQcI~1P^5-RH2KufSLn;KLfIYF zF2BN%4pqITnVtvcWxYHzv~=170*iZ*^P=z2S)rMsdo24PX<|EXe8YMk{TV!8WI2B} z0Iy#gT>eEo`-!SL+LV-KOv(<;=X39Irb>G1mu)*MtHK<}e!TBBNd&8>Y zceReW*DNW{50rQ5yV9PJ-cRdl`zNZEueSvp1{L>uI)!NqC=|dXLzJdi71*LRj$S%y z8%#uS-93HE+{w+8dqPaQT3C6-mbQis?veKWGx~13;ZgT(&TZSAUUB}$q~1t5^G8RG zdESmuEPC?0wVOc)QQsj*6?Q9&Vz)wAei6;eZMX4W9o zw+ONsT->=&PSJOK&qSJ$qh8fE`Xk{D2Ha#^2)L#!9(g4ec1@a2Ic81XVBRrlFqrpF zk5?vxgVn51Ot0KG?wkZudoxt?eDK@wd-V9DD$h&z@q$p+V}Q4V<#KpfcPg{M&>_oB>lQK7Ul zvO)7JC!zYB;mZA|-&)$a*pZ-|h1t8pCxsk<bQO3OM)vIlkTqt)Y|w5 zajVEC>aUB%xl<2@S7=STi}XX^>M(=$r1Y9}BYiq>et3zrEVhPeC!QI(@QohiL-v?a z;A9W@J9?GwJ~8!My@qOk$4|%+9v8UZu=9n_Q|hx6X`m9%!Wz!J;RTb{PP!1@PUa84 zPi8BCn9Cn%r-Lv{_QAB@KMLV*8%&6-B}MM-+_|Aq_Yq}6n?%iVg}Kfgw&32f%ZE7U8Q%;B-Q z*_8|91q^eyxf^&aiMu>IowVtrOhmBXPFvF|+O1f@Pok;7CX##U082F1Q` z+{zc+Bzko4(PeLgr5cl3-=MiZ#i)Hq>+8HfQ(Tn5MuVmv)PAc-D?IauG$y5)<*7o0 z2HtVtYBX4;AU$4ra1TyKpa`)e3g7sevBfcYFu{S^$9>~2N{ka8`C=)17JUvXyzu1x zNjX6sf+cDf`j2~0^!5n1Pp&=0>9ywXEcVjtVOJc<0(l#>$+AQG6Ollal?8qnq|-?C zZj|UzC8G%#V}{69^g)i3VC{3q&}?!8tMuQ#Rjalc8Q+KZ&y$thVmADpIpWrMI-VL>P;%6SYbX%Ig#c&F9Ki)wZ~g15CPOl9Wii61bm; zaROvKNWm(`t{CJZ!1GYxB(D0ic+YEkQ0G@3`BUh(ZAjVqlaRN`(U%4(H|>zpNU2Q} zIZY*I2u^^C&N)dv4Gj5;0vI8azM=o*jk;<2WxrfQeUL+ej6(bgS?0*~+O+8~*MkOJ=$y~BEERW7)V7?vH7}KIn z6}Agkt1HIz#AJmSG_%Nhx=U~h+A!%3hT}D$&F*u3N|j%scx@QAogvr+*M0aw*|g_d z!9k#_Kg2nE?77H`u6VG-N!qYYVMX$0)kAvsYx=%KE;tJva(*`FBX!Sx0`9$l+bF=> zocgaM9yv7DW^^8_k1ikv9#5V3+L@7JN)73m8>z(0sy$(lr>`vySUEl!9PGo&N(pe2 zTKt$i?Ye+G@nXSY`hc)9a&ce_y(YR2WV-ZSe%GRUqK<2dfIehR0;K!n;sN^f2y+)) z>*u4(J(l75-^29^Zb%*<9JJi$!wieK*E<*&aIur8p5-cGiSH3K6k^3~^Q z3qqRw>IM2TNOsDSqT^Xcdb}DmC!v}H@5PIj`(F-zzR~6SC-+`I$S&z^#_gY+x`e+8 zuN#k)T@ADum$y$2J54fqY+Rt0sE>L}A*D7_K>w6jufS+YJ}lhj(@6D{3vz%3u=a?F zc`4x6{_SJwhn4g-)dk4~2~wT`1yPr8nFK!CP-o6gfjBHqmGVJ5+izea})h zzXvF#dc06{?g6aGh#yjz#sn4dAg9(09b9LsvGBeH?OI zxE#{0ER2Mn?auz6w^Np&|64tJl-sZr7_=U=(k_CC54AXGiEa(irqL(yKx}GW~hBPn`mYgDH5n755+E1dw49BT8mDIUzFn zRQTM1^y&s(<$qW;05_A8xN<==IUn&rTIPQ!4Dx*CXXTg36`47vSdgzRRj!gJ%S}3? z{HWy`Pb*2f(j`xN$o)O>q(x3$!rzej*Tn~O9J!xf5+{U(DGK8G(G9 z#}k_&PrY2aJl>KwSuR}{us~tiJU?|sLWPfsS&{HQF0sN43MC5OZ=unhJsn>iYS18y zRG}0zCf`pHwyMs_PsbO>yzjoDZBXEU@)kVVcmm$RZamuht1mgauDl(BSQPXc)tmiS zYbqo6%2o(A(#^^vTFi0DqKlPRL@Cib!R=f-u57GHSHRqrVBSJDQe2U4DG^HXii|EwACkj1aQ|e`4FQli+v2L0ZU+>J7Yt1bIc;YjgTjk!dwVn zIiB-e$TdJcu^Y~BaD*NwKn;s8Fc*%u#bdmr+H#UNob0y?WNfF@P{g{0N<5^krmsjJ z&^1h^B2$qSvL?Z#YhaQlm*6U#X8-6<6_}w9xv+snu|~M6i(Gvb*V)Hr0buTXCi zUt&uAOXZc3XJde<8x4_J4taU>BE=$jM=r@cHuD-h=${8)*Us<%VZ%H}?Fbgph4f&w z>>meoE7RgI-U*=%BfWccS)e6H30M2vZ7Bx)lej$50UL7jNls>jzd>_q^v<6Vn;9wf z%05o!@p$bh&(1Kcq}1sYNuv_iBplXs!1}UfCHh*x-N{&1mgBp794JYQHE2*V`d^R* z_;Z7-#BY9fj13tQ?`a#zRvshcgdH-fDK(UR@1_#Zt7nchXyy#xAgBl`o$$^(nM{&6 zYwX@Jslw`M#dKDfNtYx(6?ywr2+Tp|(caN_gmZ`Ev8q>_M48jy|Ho41vMgI@(lv>e z2Ji5}0>vf~PLyv>S_+OcUVn=21dq$w{^O7p!j$NxU_S>9H!fDsc}I5e3y_>3M7=7Q zu*fl}`;x?ASVF)+IRtqvkQ1{jX1(^JWOtnYT0qJFWA9DinmW_{agTUH@U>0R=74 zc%ks@`|p0!Ip;;!X+XKD9Qw91Dotr2H^n_k`6(Y)hV9&^V?0?4yLpT|?R$vR@G|L( zL!Y<3WAso8T*Bs%YcI_dc*w+`xlJ*>6wq6u+J*6?8`73w?u%N?Idlf_`TCWh{;@nT zLy`mBsjVv9ZRR22B}s|oll^`=Sr&~;k1&(#SKJE*X)6E1P(2ZWY!qZcRBjKymfoob zkuzcGENw3D7}+4n4oUah!Ossb4R2QIWc~bpe%Y+Mv!Q5_&j6%G_W^Oxpr7`Msurqb zt&+#{;5R|=i9uS;$FiQ_Yo0Z9RaiX~E60*a@jT3wJLHcGq|h|{&}LYm$fl7PTC^Z# zs`)EmAwapfKGt1phWF1x==(2v>XAO*$WR?(4YLMBbYWpgC(HQ%&Crv+=46H@H*{+M z;$}2MmEZVo6InBbQ4J^ae{~AhVf6SzJ5q* zG#-kN-e@N49XQHTU}DwMDF!MKZK0wL!0OW}CKY%EiIU359*;OzEEa+VV^LPQ1-MQb zAsDl7hf-GMUIqN#(222aT%l|CS&Z?G-~`=CrDvMoH=p{vBtaAwK|3VN?cnPa+6=)B z{xTB1aJTe;e~N5l5a%W0aho7ZAvEqU;@H5VWqWU%C$X|*e-8U+NO~b72+Jjuk~T3A zfm&q=WdAF-ACB0&0P+vRvs& z$ZfP3iIyV6>^iFfY}Ssgcew#r(A1u(HaagSqyDjplsK^S(qQ7e9HN+m6sbZPh105B zs3p@X!%5eTc{k<(BaYUl)73Nak4~7v193~pVNDPmQMXOUeD_=tUIK!WLAj8}W7yjT z(Y&0wXwFv269l_x%w{{_UjSi$9aJ((Q+)y|HE4H56>u{qxY>Y$U^R%IHgG0K#tSM= zAeyXswPf8iBZz$eP`!kF%q@!Nz!^j3CXm}jF|dHW9f>Thvh_ZFScn4Z+t69=dbQJ9 z*^r{#D{f(yK6|Ko5r)UK12DE?#n&W$H3%AE*0na;^}T2i{wQ-!iL{rM8aJ@Uiy)T! zb950ue$g4am>NTGnY8%ZKVST!gCvVij%d+zy>fxcRrf1afO>CkXcL|2yi5I% zS0Gv%(xOS0UE-xVU+`&#D!++7Et>NS?$9N^_A&^gI?y_hjB3D6o18#Gy;}CCp9zhF z&TAKv50ljn?70B_aa78D3dJN-B$0~Bp>ZaKUJGUq!%7R?&R@M?ScNmNE3YrFEa)rX1bvPB8FkT|GGIYH*jV`rAZC?K*UCUnbBUp1HVc}e1NEW$r1 ztDhfU3o;5&A;2wqVVY|Hj7*}%sGOnp#IO^JDD3q;ABwV@hDcDFYGhswBB{OfG78l@ zfF3n>zuE>&uMl`w*a+72f~1z%GfFQ6#dF}VLOSQ(L|s0yzN=0*yXok z$}T!~3LdM9EO*IN9|59nmbwzSFZLR$tb*2v8Np9BY>e~x&;J%8H7`!%u-1g7+zc%6 zfH(^3W<2r$?oy9eW#moh%BnDf3T}flmT!p`6)Vn?_j2ep%Jb8Eyxs$0e)Yk*C{)=d zEu5}JzS2X`a;}=2tAaGeB;~;XgAm6GcO=$!$U#>J-5kSV9IG&~1^KpDrSP5clE&#! zd&X_QVwJo3F#k2u!5|nye8IlImuHy>B4?~J~g?}Ht5-cFZUkY445%vcUx?$RE zJ7}yW`edoIJcpg`Ds(epwiaF)X<$Z`hMxsZYHdA@Rc9Qw+zcm6HR)Hstp8sNjMJuc z<-g^Ty)R9y?o*RV)=`Q%M3I9dm61yPPOET|g`V;S|4g^7sytt$uEm6`1&GCaeeMsM zPgx+Y+6g5`z7Us&CxHdS*iJfqVk#CrIx$s`omyduCuksIp~wkTQ2=4?@miu@N+_;U{XoF_HJq|MHc|7bngLk&;xnTX{vIs zbk}tesSC%`S5`kxOFG%=#>a6ZC(|-n5_3KMMQ2u~!$N!YOl`W)4YwmwgXB-1s9B|* zcvx713H$jd?4aR>iI7lf;=dS&8$| z!r@^TjebsP_%7!=KmtWd^pI1uD4tguUZu_s(Q~xBd70krLL3IJ1CXvuH~VU{Lz=`m zXyA%KClD(zn;*nlWE^lj>!(m4lz{!i*wPQ7T57`_{rz3^r#VDS7K>wsEG|r zne0<`3z23%c-U^A1#H}q<}r75mwA4)!+OBdL)S51 zh}VEKF~L^Ia$vCZpq&nL14emzNFHT0L66-ZpCLm|4oll1K*t=lP!LZsYbX*&Md@g? zI+%IZDjN!jo<6}q80p9717hULiT&gRh)KF1C9XA(=Dkph11u4~4$nG4LS(Pk$R`%W z#5_`7)nIU`XKj4HP5Gy?g?yVazRX?&f;RI@R%*_LGw4+x_!(0l~WHoZjk*a_oLo5 zPY`fe-<5^(A1_%cs1ZW}Yvxw4q4sJ%HQ`c22IN?VR?AUlt1CI7b#hS0zfFJ9)mz#j zT570Zrc>nkuJVZCt%u^gYXxPhU3Bw|YC(KZnY2rYDjO+N&wvU`o47EvP*FLx#1qTN z_e|RknaMk(Mf@SBF2z}|d%*{w`uTB~6{FHiHgKx4)tUTW8Yk!t@}(dYgx_Tox5hFE zaWikWfDtDMQR^n8Z+OjIadE;zCX>OVVQ;4;2h*)>+8#L928 z{5EbV*}w49-7os+lQ6fN)GThN1Z-B9Y-hzG`4G~s1nZsw*`(l~Dl$GwdL1n^z#nrkBKt?!0nb)vVd~di(1;-@#cl1XvX{KJ|3f ztbAU((9q=n-KlRmedaO$+jJ4Hmd35ZsY_w)do5EmUh`uajJShjMM|;QRTqt+`~jl9J_~DJrPJ5d+C6*>)78-f zV!ga+r_!)6+{{CHmQLjj{vEf23ra!mRKHYQ!aUT6^>)$muneh#3lbvllRIw1PP^RG zR8SzxrUmwFkEikBXyk5oJROeP`h~f$>Cv5D$i-t7N_^V;!|y&0Kf$y4m={U2os$m1)(v%XVlsl z-xMLRv}eS~-&^gP4ypioA|yTyxFNZ&Q&~Opfnq|Kz0i%^(P4|ee|{TY{}esP#psoM zs6GE*ByI}HHW}zYrWmNawvmb&e*LPnMRQ7YLC^_gucAemO^kc7Op1DWpiXf`8f$ai z{b&9IxIEKmHe}RJ+_4#RaU2s8y_A z*s08&-8vEZ?4x_x60q1J3O4#XE!k!VOQUA!hPlJRGOlA~z#RIK+xh9MdD>gk_X2se zkbEZAHi>tD7{xJXtZPs_-ngSB|C*6&EPvYg%t!9B|0ny?9>W)PT=kjyU(A^|btDG= z%7yny;uHdP#6}SWcTx=Sf6}O^%i^7(NcX{9ESTmQ3)QxQY&7b&X`6XF)pwMMK>uhH z$Ay8^bRWILJ!dXvU|p20o)%+YLLoPpm=4tT1PdpaIBrD1JH6|cxjwDKPFYxlbAc$} zUn)*h-6ul|ZEx^JD0umh-sXd~t4=YO`1j|Rst^Y0CijcZd-wzKx8Vu!RCifCR3BZ4F%-9+b4;OlEyFSESpc0A0@hPb`G zE8bitl1n(3Vm8GIp z%2iOPUO=Dc5x<5r5_PPkrsmfs|X6d(7sF1aB5 zgxYY(LiiLC=|?0^$2HNm402}1#KPV2AToC4kk5-A6)|Xx0-fHV5YQ^aT43YqEi0d> zMV)MA+vB;7n_XdIdU7L2LaO(UCy@#VW*Gsuf7H~go??zrqz(>gk*}nip^Tde7o*-^qGIYpCO$ z%L6KjPOo=-imdY0UWw4rl@Ujz>2$s<`8&J)wb#OrL+;LDey3{5D*2I+W_Q%FH3&7~ z7cLcj;bk2TEkcro=7Hrhs{q9bNs~6L%sTd>=j_~nMK`nDE72!G026jlSJ4X-7_yEh zYQ^?RPg#tLCwEND1}9WZKJvz$TVkVqS(SS2H)Ndy+n3!Y_GKHzq);T8iUP@L=e2@b zLC)M(Sp{h7eDJ{shEZsTdoi!W^U@;B@m{Uyin&kGY*xcd1i&ahd{Y zs`%>k!|DV<16eB=oPg<&H7EhVSk?1fZsLRz>gJmZer8T4a@YVH3q2?)L`M%U+~(O8 zTpiS;bpbc$Pl)-@cNRx5_|XedJrrr{o2Ty;)d>!RwCKc{lM#qo0`7?zWJZ3^ z$fuyaC)!*j)M1Aa7THe>iNPR3>!lfS`S5XYF;uGE<{cx~9t`*x3dQA#V&rM=Nv>Kv zlZ%CIb)eaJ$iG^k&DP(lgF>|lf-H5Ho%|0AyML5l`qbt~kJZ>6a|FsRyL6cmdmT_+d;`4S2pet>le z^CXErIS^<4U;-Qrj(u+Id|UMDv)kY_q)htXZw5V;HT9w)WWcR4sM_V!%zZOT!_m*w zmPDQdwPvjIhU~b~aOAgxjaVk?pdYvwyA(qy?g;>hh0DB8tb$?V*_~^Xvz!nz`JcX5 z^>_1m%hM#)7jJfgqewHXo%AA)p(J^vV+uFcv=2udi`5V7BNN?t!4829~fr7?qsx>Z+^pZ5s z%X9G#lyzOwqN#Q{7kL4yl`h#ASmdz;6ziw0bH5x=k_OZ7EpH42y-k)}x9Y zEMpLCsklXR-ZB>fcuD3mEb;)Uq9Q)35>?CqrgpY>uXn5Lu)0;2#D`rxereHdAa5Ir zh&DKX>2*X<_}Y3tuH!;ggag18$(w&-=CD(%Y^h+s9~3Bs#o=m~g8}zN*GVo9@;OB- zd1ruLV#Be`)=0A+Tc&q54v6E-3mLy8*eNwJRaq3Xog(Q}RKIc^17)?roHT&B(bo8H zXS(PHc`qpU=?@r6j_RGCW>u`;x)9{d#w<*5F7Yq%$7}35q>TqBraf!hz2ATI_uv`E zLF)Uv-}xQcI+oCg1E=8aH$g)_#lS*NHWjr4RMD0PW=OU=KMd|v{(cMfd#Ef5`I|UI z)yhtR6*(3f!z<>sGM5(h(Rd!fAh=HIz|%>m6UYVL$Tf!h>;jJQfMR5_>zl9Xq?8iWST-+S_2uff1fH@v>o zzuX^cH36ehr#L~5h4w4)Xd2&eWd%P!{E=6@00$;p0>&0N8NrJ!7m3+%GL}2DJpTH7 z-bPas^ZxX;B;A4i;c64JvWH?oJUW+(LbV9Y?rfEiN+`+XNIWYO*(H)SuB zBW;l9xYW{>l48$pD1_ZPYpv@=P%z%3sFSq|AH19U@3yolEFEhda9A3%;bC^*e7p0% zcbLmvaq)B=IJnOugoU-Tme0K?#yUF2r$g5I)GHa4D$F0n_K~Xw^Wo)63xnRbI zka%uW!Gl4vpBe|8AAUl8lYBasZ|lJQf5*gpbWqGSinNVXW*cGB>xY+idcR+ia=p(H zf%Y;OlI##HeuDVe5T5Ic^JQ(G=(h7NI@7yNyx!+aFAVoC6GcOHp34ErbKss%Wu4%x zm-adz{yI2qzn}Js^mE~Hg|?Qyz&j4}DHQ0#MWaL?Q0vP8A>n1BrJB<-VVSy8+5_Kr zD4(x44j}9FahlKnc zkseoAdpi}ffv);gsI^N)eV!OqMs*B?XQp~M4Cr%^6g2EqBiG3uc@%^KPLRgkDtja^ z;A?SO-{$g(3hzR-$CFbV-Vkt+Y<6G=xy-~i zWmC)!ieykxHH!v7M9=VdlJWwtPtTGpiQLbxCHQxl^!~iZPyx4QQ5{UM8pOrUIpOCt zIDJC#y?nPZje;c)&$4T*?|RGHxArP_^D@>6T_l2Oz}L9hS+j;++JQ^IG9Kd69pC$BhLontnyKpp%ykhNsjsTGSNO zGwqzb&?j46V$T?#_0YoM%Ua+(|Gvnq1&QVgIu0AMVNuJ$8FsEOE+KV-LR);$Uf+6p z$$&C%&H?}GMHqouVrRQy`k1+NwdtMq-Q#9OW~>_csm3_m=t5?skOK}p+*~vnZcb6m zaf%$JqE=4%a^8u+T2b7Tje*-_U0$iG{@FKR&YUd4+KqkkVR?hxPz}3hS_{*qc}O-e z)j@r*DgfyoKI=kSH0Zu&@N(uJm&Uj!OFF~KNd-jnDj_O!RdiK^N4vu?tW)T(^QsQY z08z_iNiN82EGJt62EseTb}p(8&G9=jjs$Tj zTrlClj2YIX!-6@jvVCqy({ z(<>j~UzuI)vL!(0x0=NBj{l^5aj$&2ph3|F>92KyOnwT3>fE!F~LeW=#Q8DBc3=s?{#bn9Z`) zyk<>5$$%P*5HzG4gqR(J`%Uy--*>Hsp{W70`CiMR;-ddq>kKVSFXS8{bxJ>SEd%K< zh|Q%-vPA2B5`7G3Ceq#EgbZt~jTQHu{rqr&x!$D1&ahd8Fyp_8CZ~hmuD&9zn5UC% zow{zGp-v#Erjlq;jrfjWhctOkrdz+VQ#q*YBMtK1bTM5c?y$eh%cz4%u2C8t}s z&5$DA&Rsj(IF8-_g?t6s&TSlXU~jnAWQJ5iF?%SIPeq}p)31nunoqs7At^<>j12M* zx~6$*4f7^ki`wY^2z(sOVHw#&s^@LzVW%Sm@BBxg+(f8h(=5GYe-1Lt&kZf+Cao}k+$Uu#nR+m!@5Ic*gs2qCB>@GDz zCrO#6!bPIe@FXb0p5$s!d8?u0dFO+4GM&B%*wCCq{B-)X>TdX5x2k{@A?-A#(pGq! zC(SBMtZs%Dvgw0*(y!ru8;@9y3Y^UcQ@0C^zjRh)gg?}I2MWaZvjh`Jm`nf(;_LgjWhLgp2T!Z~h{{G*Y`?4%# zlXyvL$hpW`pqX?kk4ujV^sXSPVdsYr@|TIQ{^Nw90{fB+GHl!T&u;A{K}Oigb)n`V zJBRgUS%PqAzzjBRx!_>HKA_U+Xf2XpGJsZqe9Kl@r}B$M)lfY$o@aAUf(<5IfNO9i z*jlpfac;mpyZGP!uX(EDOA@kUk!VSgr^s9Q+4O*H9n-A3CBHd8O=TBP!h&wK5JMKU zWZO$NRbHB6KAvzvg#(A^Scb*J{04dKl-AJu5i6$bq_csFaDpT#Z+hQW>{)nK7|&~W zpU_c}%dlwMu76Hz9S7g0%{Aa&5@fI_#wY@DRTtPDH~^Lnz)WRkL;V%Ab* z70LzgR1dlzBpbpX(mG+6;(+VKzF;E{!h%ORa_NMAavR2pqxy8iEB5a)M+6sp;ka`y zHem1e&quKyoPFz4hT6ov6FhbpaB2at?MYK(eObQq+u!XpCp$i!1mwW|&O&KCpe_h) zry-3D3r}rz-s+L2f&wdHd7@z_q(39Ix>dIS6>U|(wum7`O;EaIKxQa$zmCZW*{OzP z52#g)zNp;{p+%5*GH5)#!P=+eJi%`3^8ea&V~%k^agwMjNY+>>9}etp9yS^M_EHS2 z*X*I99tYnGJ``|LeMXurDN+}B+?92BcJjJH4eZBE^$|fWh_hf!y-Slb>#l4ysiZ0b zJ3N~;9ly-`PW7Api_d9tBzyS4T@y72)zdN4;()cZq(xQid6XB=voU>gyto;Kq;Vp} z^mcaK6usGTeVP$BzJI7*LOyn2E@Zh0Ms`sQ%u%+ZRtx4uw7VabWrJQzsaMSm;4wy) z`6f#+<>m}Y5Y$C%5Fk?^S<=tX^!}8oan}j!=mC(3JMOi4dWQEg`+B6~frV*D+IEi} z7Gsg`^J0JWz?@IbLR;Rc){p4TswOckoMV}SB;^5E!-8-WN^Ip*`IOyk_ZW|e)AJwW zetR9_^!y!npx^i2hrju;(T2QJ{O%xWe8I+^FHG#pb&9z{kya||KDjDgMs+BQJ-5pe zJTm#0=Vhuh7>+%_i6GuyY#Z9j#J6n4aaa`jzdnSg&fqDU#-sN3KNd zAlIas{Pghcfd6PwW02u^p8{=Fs?S$PkgJ5>u#2wpJLr9zZWXs^Mq2!;y>_}*SmIf% zcmP#lR!;fqZrr*<=aIN6mb_S;UWieH(zrG)c#GIy2=%%5zF{7of7%&#;6-1SkbQJm zBh(4Y5OjeTqoW^?bh?EB0c6ZN%>ju7*cmC-tCMw202k8GNmk&)g5DkfCG7a{=l$-E zISZeKzMdz#7E-Io^M%0(74@O6EgcPW4@=&g@&5}GMQ3b8a$e)#T)TJ=TomH^a(>iEHx#hZU# z>Y2@;{?i7*b{THl5)or|qO~v?voRYUoHO+1`TtofJB{>UE*#Q8ZbBeC3giw?RNvyG z@ANFb_>Gfq#0G33+Fah5sYk<6Vicz|u>nQA)VJ#v_bZD0SA?ZP`6_fz@WMS)+lB39 zyKE4&kuDihfHDebnI`j=ic(ceMUC=<0r|7qg@)^} z2y&Jhiq?4@1o<9327#u_Bv16n<02Vw!*9W2S&i~+dY^ZVa2=yvJ*^tjI4AKM$B!9OL5HIIttkQaEiyG)IdQ zX=J^j1gcG)ShLrgZPzk$#CEs+j<51bjxd+pc{-lrxG_hM{lq>C}-)n{H5Nnj{rbaK+?)EPPUPE+Pm8xu(j&20gqwS*m!A+W&amNaH z*in4xDX!XKSd_37LpBA71&v+3)Al?+)* z05-_)2=36B)xAYkM4y)D`DUwcxV2~sNILJVq&jG?v`lfIXRnuVtbdE84^{s%j@5vj zj&XZ29hN%l`rm7gzUcb;C@+u&>aj`649_hA(W1S)Vz3IGygduG(P6c8x@3i0JDn#< zr@K8@FGy46^ONTcJM}B$fiT%1+>5?Ot1?L$NANVBY=_8*Js{2r$|i%$s2w}NnG=*J zO+B^ld((|Bt8P*7Rg&t!!MI8jCv-Q(KsIw0iecya9u1EVK$#MZe{Gw&dj6W%lII-v z&*fzSb~3#?mF2QH5i(Z#6^W7U!VQvgS-EVRCO>>rP!^pKxmmJ7k^@)R11E+q9w%Dt zfD_~I;xvPOJUQtn5+hnNg`2OFcy1FM2acjZagR|U{WOY6rO3yabCRSytxB1?c42!k zn3lT=-AsL^d>&|5SNUN8r8Xi5{uG1ZK@IB(N`SyhxB3n_wTY`2rKu12 z4m({ZOP%8-Ng;SPOVX*_D(e9-;Z?YXh+6r>PVtgKcT~M7_RN`kKYYk(&^^)Tt2T=j zXF;HAy$@8PQl3*jag6?mD81v=~mnnnKa(xNj!e_Z3fD&jgx^y#B@{IkL} z?q}2$zDFR&RyXT1NTTf)wF`?p&P~0{i=B(aeuIjn{RXFz2X1A*kKAszPfjikb!FoR zzZuWEvjY!OEbGoFI$IuzT3qow%nZ^i9qdt5%v-x~1FSTowK46+s7fRalhLV+FB- z1Q<+Um9I~pB15+}Pm&@#0R_HMoGnS&uY91;L9DHWse#aKjkrI&GO|&qqx%=)P2ffd zKbJ=@Y~-B-MkLnAtf37OZpD5Nz!Axz`}y7S7KOG?niOL2nlnij2ybD6;s@~G_K*)g z=H$$NEF!M6LN;g|n0X3B&p5ytO6)__AI~4`_?6Kmesrr)My_yki5+-2dSv1f z-=dfvigZ&^KzD(<8#S~xUJ%c#3A%>md`^%)>7Kwm(R+<@gQOo`@Lq;sDd_9tW$nSG zq1v4!RoORt*uNkY(UjlwG5 zaYb3%;?)*}QC&0+I?xs!b~+l64=F-RMQButCvri<(%v?_srK*Xk#Arhj>cGMoK&Uz z?p3v9g(a+a1320`MRLRn@nv2wq}8|6$q}CicZMcFaf&4ZbBZMw-%3*ACnk5| z=gs*?1Si#>2d7T24n@Iu9Eg+WECXk>&1LT_`%W{AkB0lRotUS~=~Z>yH*He*YeVMp z3FRgbd7@HKAgvB6R$hlN*ty6HbZTHm*pMPi(hOm*8=5WC3Z$ul=QM+E>zL@k+nyUD zTA6bO=&=4QOVZ(4ey8dIDHUUW z$N)iYQA4CZR*)yclJH1_YL)FF>pU71%l+D+@?rJNJW071#%t_8W?Bo0arCxDx3WWG z4A*w@Z`?onjd_Zq!@`6`P7-S&(GbK#s>?e3V}F~~e{<0t zhaE^*)PVLV5(EXIyXdeAAGCF93Gb`B?zkQF#=NjP zUOU-H%A|w*I^Gel263r>ssD`yP!H9=$s;MGOp3>Bgg%QwW$I}>V{a5P#<9r`s#~I! zADKHK4m;McxD*>*_9(~tW1~*x^4u*00GAeRb1gdn#)@;iGvkNMi=M}EQk|x*QrF7T zRH%i%DX7m4Y7D!52Ia6fOg$Nq;bNt#CpM^`FdfAM=uuPcz&cI0Da7II?i+mx{kG6`?0P z?1Xt57;`V1)enTTL`h}jK=Acoh@rR2ZigAlH{m5H!Hl}hJJsc01~S~R(@r&t`t&O+ z)q57^`59`2^(rsQQh#7mWQ`L{CMX*N6ILH$2h%6--2dr|&cH{_rWOsjvOMIJrSOpL zc%9S<){wL09=}lB{L7Yqil=H9SG(Yuo@s{t4-p$S?ftR3$67>;)s`)fu_I!P!HCk( zzi@xiJrtBmf}mHt1c3r04G;r+JO9&2X_X(KoD z%yf_nrytTgLw1WWZ_4hk-dO)+OJ2vZ?z8tXPEX!(BhpFssqnAc$P%*Qr3s}% z_485bq?r^0P*0_z&XW~EK)=5omH@TQ(d)p{Pbe^qGHt*HKyH15Z~)AVPM-pgq4-R) z(T)!72vjYBx{1SK*crLqZl8)O>7)z!Ue(ucI(!$8s=oxCeM){J#EfO9O6PUHQ;lv5!US zu9ognpjK5q6prqV*ya9(Y8lh=wdEl0d`~fQCR%jnm0VK6+oC=%-fZ7GNlzYw6_57G zz4qMT^sYK?Fx!>-)qweGzr&6$Z0q;xUgfsHVc+U$SnBJ+;_dVmX_ulGpqozjD)HD> z|NV2RC&rb2IglG87wrr(x~Qv?`ie+_125v8GI2VqDP})KDyXPd@GuW9 z&(6@+37u?Lxd9=V%a9%L7@9hjb2obH1(o87uHJdT<$g+Xg0)ljWQI{!iB4 z5gPIH+J)r9WVHk12jrASO?-7KNh68 zAtx1UpHGZo!9L^&4jXj4WleAb$E2FzD_hLTsV|8O%R*(Xr!i>{*zLhj1g9^P1CTA; zuheGfWvC4@Hc75=)C4%L`cCDDB`_?OVPU*i^&Y=hILl}{9y6y_k+hd4I#Oj~XbLF? zYQpWJqH1{E@&cDFv$a@lCrK|bFz9yM^R^;QRpEQvvq1=%5&gk~ZeJ>Tm3ROLt{O2^ zO(^%$K?pxPM4Q3O=T$_esgjh@fsOK7MeJNV*d*?N8EY^aFFc;UfARm>^@^tvFwMUz ze~)b81{epP5m%T1CYNHeDYAo#>gSiux;uNB2xh}j*kJLXTQtbDm4P~dwlAU!@^JL& z!h`hAMRkyDh7mQrBb4HoPH*$PqQ;b*T;F#0#&IUkJlzvt9T_)PzG}**TgM7p|L~&o z1&Uqtc~a=d;zDj=c)enk5|?wkp^zlXKy=Y3WhH@Sv(i)-1lpS+jl2!vHS|t(hkjWY z{jbw-2HE-!G@coBKQyb%U6t+rE>0HR+8Ogan>A|FL1)dejK2feH7V? z%Au%2lcsve&*WFnH00TU#v_g^F}~|U@ zOuNx85A1^&etBeR_$Q=+-$$eBUUXm{$O5hs#0#`|H6qj|W6}5#_4H_D3)8}msK4|t zm6@;ny(Dff%L3-1$emEb81+E0qH4ON+w+2tO*)@zM_B?Eu5H@l`r_A?y-khyw(gJ5 zkRc~_{ER^{a9~$utBI+Jrb#K8_*f7fsGE5#bmfd{mv~--0=Kq8W>p!+XeI zE_y8Z&ygcZ%1`Od{9G6^HoG-r?gI~e%#a3QHPj>?a4T}nloUwwBQhY5BGqT)=Iqei z&?>*BYT!t__0bukEkJc#E~@j&gL^ycqg*K2{|vQX9Hkrkm9&F1NQJuWSJ zByIR1E*!PJc2yeBJ5T^v&+G_b#WW@3_Xg{a^n5kLHRaPrE3N`}+xt z3W*LvB1y`e8Mw*^wIGv}g`xMu4Mp9wx1Dvuf*A#&kQ1Jy?4osaE(lCtSLV<;Gh#rL z2IFX2E(U@#)GWsI7*>SymvOj(9nsk$3} z7=(wg6!AV-C_dm{K(vOPdoFEcljwp17i{j(%Rt2DQ%RoW20wb?5osD!6@?g^4w%%{ zu-c3y$29c{5Cg{lZHb!kfaEFMZ~(#cE@cPF7{kuwyN`GL>;Ei58`hzPGEb1=mFNSk zSTsY?fi+^>iKC7{QOSyhSQNWil_$FGoDJ-PUb&6h`-+#jqHC(*$h$RutbcA9A?B!hypo zECW=#uwvdPWRq*2Z-b(TKR{YE88AKC_{LRfwmOr4&HIMiqgU_G-yk@v&gS1FCGYev zPWe_Q^}%}TgB{XNsIY=+YYp;x`hfpI?-tE9X}hq*6R*Uxz4S$xBH0vM76a$0(dg*| zw#3t)YGPKISES_v90zuGL6l$=m#Khapn`D@6_vlZL;2Tkf{J^vWSmZ~WePls79h#( zwh+7(+=&yFhU<4S!Z$_SanTlgo{=JhKTU;8#o6>GNx92qP>VP4pls{O7(3TCc(El? z3~l{AC@j%@9nfKYTb4DyekBSmwtZteecL&2jvfd_9{v1vOq!~6R-rhZP7h2E)V2$; z&=wY98dB^68OC%F#!6FdA*HhhmAO!{0M~Uu#xp$-)pD>NX+{iZoME zJE&?9el>`CYV$l&WVL>UVm#O**c600>1&zN@Dv#s9z1wsUc0bTRS2Y8@N&>I&QuSF zVFkHbIt4@#&^<=*;){ zxwZ6$2;|`Q(#U-IQWm=qLp3(BJYzifaU)=iHf?c;lQEk7ksx5Hd6kr>qj?VO8MD+v zIRr`uUwUE9z}*r1cqbLxH9Fyjsmnl7*j{*G?>h^T!rp+L&aflp$31`F7ib)!y5H_8 zA-N8`q*QM*=u}cnIYmmb{8F!8DH_xNiuo%bm$(#i1RuF2`ZNk*5CCE%W((%}BJnAn zSI5g$moOzv{K5^Acpfq&JA#LUqg{5YPfqOx$L9ur4+sn&3fM*OhP*<(wl64Yv@zbllaxDcHma$^6t8Afa-Ba zUe{*^>Cg$1*7 z)hSc6L;J!8g8STBG##G1MIAr5`|Uc=qdiVmEU*KhEC-F5RWRALJ%YG_=I;$|@A?=) zlRx~=Czh4=aRjAv^e`;gGbfC>8fi+b(O-)XHkq6E@3!R7 zebQQ)yyg&x{YE|D3XlA2)WWE zS8c~stWwGub&~f~H!Esiw$&Dk?gxU}`DkneK4UnTtAZpQv)BAibd; zcG{!ZI{Td0kfL0%9SUh+Rj*;Ebw0}jYw2@dP=gvoGlrdV)wOgi3C!|Qm5`TbDrrylH1zp`DpCvt;wNO8y?CFhX&+pjz! zSmqA2y=K)q#!xRNNjdCf*xEXE-MnX-rg0$Y88*klQ_p}N`yeNLn;$pNh_TnM=dLAr zFO4-iZi2r36a#V#WmME&f{GibNU`FqXpbNt77T7IDC3tYcKPMbZuRXDAE!%r)h<{@ z`xLnz6we!S!aEKLjz?gQicSd9K(bQ=7Z%bbrTzy6+58Ui9s%CfrcCk%F1beu?|8%w zTh7^S3#(_~gyX5N1@)|RF@onq?fL&AaSrS)gSN&fZ~0@2NutO`^p-o7It9}H^)LM? zCUs%yya!|tzrvuTkS9s-$mH8)K0Wr+=GNI`w_%T?>^5YKWuwUji)W}rMrgcW>iQS5 zmYZF1;86`IccXx`m14G_OBRLX7QH$PXql#|^7+YgaK?oL)Lv36)b4U`qERsc*;U$I zcB3(v^*lE);Xq^7fSq=g3%2YSPx7k+i*m3ezaEzk@Cx-ckwMO>RffVD%_?o5G${nt zd)9hPpoM`x<$w!2Az5uY|AR}be z1nLfs^o50}7`c4HK9h%*Rhv(l9hL>hG6`O)KIZm`s#ccFs}iP58&$31JK{3wkkdx8 zenu?_g&dNcmDG}SzcOhp&=NKR_x~=vk;EuhDbeN3<@MP)Ki}|_=K!{Y21!16SpIoViGB`fr{$z)P*+6vw$>PLvMIx->h8epc~A46MayM zvcM%za$7i|Lkkxi*%dfujbE3hd?&()jN}|e1}T4OkkM>{jAImYm?E`+3_brpNx8%I zxa+5&7N=k7+2Mh)|D6N{zp7^*@qHvkN_D4F8|#I2aiJVGMnS4)KJa@a)#7O_da$tE zXt*k!uH)rN+8~J$mJcvWgRONc+oW+I8kSC1Na{WFeClQa(;K42!UX45agysZBPM3O zG-{U%kjAn~791P(ZI*A|{Of1Ti(V8N`EL#4O+iic0Ku+;l2CPv;&~HlC=5^BvR<~l zcWiYfr(tK(mhb=hKg|OnT-;U%p3bm@O&TCHi;ANcc-k%GntoaJ7?$N$h==&|>xO42C%b$}YQ%Jjs!EUCQMv9!FqLO(_ zMX`b!h)(87njym`OPwmx0lQFpY*xJB_KYI}oe-Dux2eylS4=+#>-d-`qlKk6P^}nt zYLpukU^|s~B2~15-xJsXF?Ah%G$3CuhOyZy>WFujOqETKXDiR$~ zH|muyY9J}6M%+KUZq~3>^^*1i)n5M<$g6grvI{W?^Eg<4m)!$G%<@A6tj&YDOA+PS<4q3nV-vx6SWeZ z@Y15s+>QoZ4+0?gmIQw0nL{6t+yFN2GS^!Aym*}kD}bbi*c}SkN!GxDd*^b$gB=cI zMCz$mKL02G7d^oPXE2yLxKCB(w{x~Zis#azRUSEWk*B}RPd5{@86?QA%k|ZU{I@eYWj$$D^@4JLMhK+agA$O$uy%9;WT>=h;B-M+NocuOWn`h) z0NMu-;Ym|vd-r;;=eDrS9fI5)u#r2sAoPt(Z~SaL>e-I#nGqKC>{eNoI?)F)(I_v4 zRJ3xhX4QlYQ|8@b2_($#nTY+bx43>bo+V=kh6Kx!af5TbPxVX)zO=ivyXg5$HFTDA zGrwA}SNfTU9yj}D%Ce6K6)2w;)>Ph;*^v;FnG$)*g1oca)daC&)c@b8weppT~d)3PKs~ zY!@~MCj{J3UiXX+e59-j8@37mz#62CAcpnR22f&mZ_-x{oHNf=b=c_;OZMq{pA12B z(CP)dMg5ApSuL7e`-%h$(a!!MF#3vLMTmozp2@9?Xr*#L-w{ z&6R_$S`=V0>_CmWM4ub{rJ_6uC=^0DDOlqAB(Rsh!Cyw!Lm}o1{@Ko2OawyRo|+)M z30ZP`AwlD`s$OzWYS4ItyRgl2s6(TRumLi;n4uAK^!(59BXddNFE2}t6CcG#Z#0wj z+~RKzJSBnz_)$}$bc%s8Q(LI0VK75jEHw=RW5x6~j|>;gj|WBD*Z&BKh|+rALC<0O z6j|@nE=>2?;a(M%H8BI0H3%7h|3ANlf$xR9r8OetRD`#H^f_=IM#ALLeZFimGfa_3 zR8*f}vs)g2AoxJUs)&C)hHSx7x}H9%zAIZ5(W0pis;5ger@U}`BdMpiYswUc<9&j9 zy3BL)T-+{@=TpAKH)=;^6S=+NHKB9|TRt;!0S_Gmz(Je@wRf|)b4RqNLD zwL&r85iROcAN00fG{c^qpnv(H|YMIpHBLLxjk~&nL3Mwse#;hL#~h( zWus6`>LFd^(xnE3Hp)@IFlFi)0!6y;7*vf9S_^s}IrPV}o?xsc0fnUZ0UHcmZ7p7r z4zGn?st*0;7S#Ya=LZ)6fvj$}wni8y+NrJys^NFR^)+-K2y2eV#@H9!&&2ODPjUnI z!j&Iu6-IlMKI1?BM3NlXu_`q&lvxx5#YfVqs3tl?fMyYM+rLsS)armFi1V~O3C_}> z6it|3${SasA!!~J4Btc5!Za1Aao~S8m|%_7G2_R{Sgnr#7$=-e+Oc6#+t-XZ`Nhqq zRI;Dj6wPt1C}}Z4)k%s0ecmHf6pE~Oc&=TDI(9Wdc_IuR>XQf1^BHtc_^LekX|bL} zks!F|a?hndyfplXv{R*B?V1jR2Auq(z{7QUr@n-6!5l4)&_jw1kU$SQy^=i8Z~$rd zVwYkUpi@-&ZFX_N#gZ~ zNu+@rm>hV%f6D|+*C?iqBA2MBb-_zTtvtiu&9F>`OI4-ftEx32Et+n!MSXb&d`e3p zA8zyXTGFgZ=53yNkte=LjR<;(C)x*wit(R33W{P&Ex5nn5?K-E8}Jfh7<8>1((R zFd#5MT>dNSH-CuwCJ`K8KutZ;97U^@Pigw|K@cwWt_JD{<^tl5mj8^yf$WOEI(+YrC*z)); zdr1JYf^NptF{c24e$$x1^vy&z`ucDYrij-4PScMHSzDk)29M2XFCfU>q>*<_X zOO`asF~KrVlrt+!oe}V*SGy3UsS4?A9}EmZ-2xzVpeq&6>k%CDXq02DqBK06ekAP( zhVUfV$|Plrs!d!5;%J8JEgFZOk3=ip0@GpVslDfarzKZ(_DpM{jqthuL!VP*nFG(j zGE4xmkz&?UWGxk?gGhFxJfDXmy_lO-0}s;UI{+&%mI?>GqQgrU`NRPV8c0F&^O%G&mV>glQ-utg>yEtCx8~qu=bSQ zU^`$l{C)Vv>37W6<5?E6vqPH1*OhDB4a$Hz@JZsxQjj?Xl{eK<#f^EU=RU+xJ*G{b z5OoDN$~U;z_@FzJq>Kw0ax!cUxHSmlc^M!ORUUbYxx`=Zl?vM!n!(0x^z3c?;Gsio z^Xxgj zN)VLzeazEhQWQ@0dOYgr0if_0lBIMsYEM8Ex?kDhc{gmYx|;{{{Z?5iX@iufEOlvk zzoJ{dKSYZe@6GzTVx>ek6MgA?uZrot!Glm60jj~hw(RSW0Q`FLoa6qm&f2L=P-gPm zRW&d?Hitfze=OQbZ~giO|E(UeWSf7wU%DR>9JRF|ho390o|z}P?}C+hvVawvq%22La8S^E!FaE^*XppiN`7!gTsF+3H*-hMGR!?Ls`0=hLaeNaap-9j_BA`Z(+w ze*Lzi<=-{}aB}~Y0jI z)UB5W#f0flDnr?;M7LZ#Ip^23Nm!724|XYz}SFARjU*q*}7^8(D8W_PRO$ zqH+ZkkBb&HhaTe};Vcb0Zo|grAekTMXzw@~HmaxX&-q?PT>awgq;n+2fpN9Z1Xp& z5FmB*!DV>)_xpA7_9)K0f)x7?Y~;I{b(#&pV%DNN=)6Xpk;vI(QjiN(L7WNb#B$ zjbXN?6Rsv8Nxtwny+SicZ}#n@PriK_#+z>W3c+Joi^x*9!iADIpIrOP7d@{JXHI(h)^7iN8S=T> zq2P|)Xgbe$<`#DHGg`3EDNg2N@{iw*T5ety&S6IvmO5{pyfzm-B&tAy9)=R2_fSlk0hNo7=N`Ik zh$xAS5BMluGNefJxgUOzq|R4NLSWc|qqx?ulcmAT$xt_8cSyU?HWH*A( z&)yg}7;hvj@S1C!1bz6|&MLBtTPVv?i~KU5jw9Q?*BPCIfZna zJo8Hw19=k7RMfFq1s)Ii75wNKUu`8RHO! zx_NpROiXVq_}d2`VJ{6i-4_4d&|R7Q^W=_OyYPNGc<%T~5k76<^L< zD;T6(gD~%^G7_?+cw3$81bD1Es85g|-bptvD0g`PG`vjjAw{YW{))BCR{zW3hF19< z^tQ0$Y}_l)e$Dk~N1pq$KRJydj{Ek;zc$!!UR>cN8QfS(FzEsaz`3%{(e=0FC71-Z|bSfjX}i z&0gicd56`jlt9@EFD2K>d9ZXf!XlSOcbis!W9d38QNq%i?G7J|`_bU9jVOuTw#=Jc z<%SXmc9tKTpyUgR`J5s*K`v4rFNr7J5Mn_G83}$#%5D&k><(Qi>Vf<#ZNG9I)8m2J zgY81RFpD-Q7(lXTB7Y=1r-nWo09m29n1JPO?+KK_rHW2m{%_2i32g%qqr zoCO1}b(f4_)-^5U=hKad(=7_VN>aHY&T+?P0*um8p7m~ufvJBM6?Ga|NrsH`gS-Z) zqT5HG7U#@;LWG9cUFep}ovg6t7D&EST%)$H9CzStcJ|a; z+@?q`6iWa4RcQ`yrJxsF%369gK^c)U#eg)^t&fi371K5TDIf!pA97Mr?Q+0BcKRuh z1x-^WMy>^C^bFm=zbGCEjt@wuyX6^frTj~aw)tbu+A)HY?RefT`H)kEEI*`1e2%GA z#LlgDDTcg*Ht`|BC#o9%K9Z)ov?$%LMm!)bfXK^79m#$_#d6q^eM7|&XwVL*9#XSiWQ}<$j55dH2zEFx646xr6c&^ zQ@rAFb=UJO0GIP-J;=!hIIid5xmcBEzQiRZg8%r{ucQ7$pFl=4PK0vcykizQOG6GJ z8X)w9cdBuKvY!uQhKnK?iYNT+-rgOmp~Q2aW|}mcPJoDg+gJ8-W{HXW0w{k3RuVJWS17%5i*tlgWW#a z4<_uu7=unheN)8SWMN)smKw7cFf%_%3CX>l7a(O9h`mY5vf0S?#s8Bfd6IM}R@DV= z<9HlLE>Fjbe&Eu))z@b4mwFiwehB$C zKR1!Lgi_&-M#1cU?otZ7uk>z(NI@^X7=;EBL|8*usO*b)_WcLSR=N*y?*z1X|Bt!Q4?Qb$K?(l5_ucXA zA2_kOu&{rCEV}AQ!_0%_mT6%P5*1d})zet))Dzklp;h38q{xRL;@SXyYrF86EZt+s zqukdjFLcZ|(7KaW+mGj*Q|tP-^{)B9bvtD4k~-~vr3GKPA4ueIMWK~Hq~oG>1%%}ZpX6T75)OuUI~ipikJHYzrUTQj3UewWu7)46aT`gmKJRo)%k z%lsxdjwdxUmQ0821&ID{^Xu@56AVPIo8Hg50^7;~7=4>+_1{}pNbCwP z6sy|Z%YxxY23_NW!bT8R!?<)FaEoZ=$)f7{ZG0_*QqlI>AaMY0tOLdu+(o^~Kj~w{ zjdK0!Ip$M`+x58MD{% zpZs3fr0^da@e+`A6|_*C*tHojLBI`)(NRQ0#U{Ub_01MhUC>@}iac9+S-hUvrsxsg z3BAwHqpyNnj3xgUg|$GBR4-dMRh3Ip+$(_`G>_gqE#0pp1e0&HoLc%Z?=f#TPk+l* zvM+4WE;?a$wdnkUvmDj!;AGcg?|k~6>muM`2R$0m9jRWTK9h4NG?QuJK5@Ady7A2_ z`Kf8ZjTF)q(j|v|0-ylw7gvc;aAr_)3s{zQbSFgoRad;0042^HalGK6`>AOOoU-{D z^!6y#D$Xe~ws{&icq~m%Y|f8?HC7JNYk8g}vHx6jUAS$Jdp@5Cv)pl)sUK)>s<`FC zbxexjuL0^e5D^LMp_c|22)U~1bt3CPhgJ?CZL`kvHsL z2NVT{R)(q?j3DJHai?|A_x$X5DxSh@c_d5hl3;#1~xBiwW7XR^Yw-16~#^) z8M06#jL|6* zQ7B$2h!cSLQ6F6^Yz$f}U(VU!w=FD5q@Vu=BU5JV<_@~6c7`|1!ZKrhoy_3ObYCy|^^)WtXuw5C_}c_3cb^h@)CJ%pfWN3;bt)| zY>>gC+i~B@3K_3`;~g)d`QG*`Qq^E_C%luhJ+MpCFY62%b}8ke{zxv}FUt^RhD|7d zaHOpT(YR5;vV)(`P3`)?xCjgS&Fim|>n|CF-*S`O!~u%wr${dqJLG;u7AIKgUOE2) zuo=|2wfhXhu2Uob0+3NW zhv3#U(??(?JH%`7Jpv)bGSLCxn(YQfz4OeHskm#3Vw4TOg+Sql_iXlG$Ezbrz$1E< zpoU+K8~zqZ($%8OoYCaIm80{~%8&BP z?|xAamSOUhzOReTgV7eFxV3sMSXhYwhuVRN%K4|j=>#E6kXr%m%hQl{*#n6zX#%|$ zimoF@efP{(mCpn%*kahI26~m~@|jST-6nnJgrvxEEnUN)KWoRmYTJim%vbeAPnXp- z%R@0nIf3(r`mb}&54#!XwZEv&{E5Unac~j1`$uJAeM~Va6xm3{p5!KzTLQzs2Kl@c z`J?C_i1QiPs*&m1!3M(=I&3g&+S0L~SYej({%Op^e=biLscA8nTu(`%IP^}s;_O$` z^;ETE2YnqnTGoQdp})i!ov;26*?Uur7URahU;Gux z48n9tkS-STgHz;KnWMrm?KxmY-_1QCEceBz@kBzzT5Q-WxH1L|#yquu!@qxtYs7|2 zfXh;{f*m%TIHfPg1RE(7vymbjKuAVdBSxapx_JhafF7^#0_Zg$NIDb~I_toQSdEJ@ zN6&F57b~CCo9FPoB0&`k8Gku}&7$_e?yz=R%K-w~6#0_qVhQelZkzfDGG^4=4SwCS z7H?H11bR_mR};M#c5Jpyy&-N4vZMDr=8#&AlIJ|=F}t(dN%gnnao;dnkG~e~yGf3) zvmQ=7vs^c^9-S0(nj-DUoK+9H^r_Js%A-mf5gUzgi#KqKW;t~B8V^#e8F5?hk-on>{GTg3>nyO{{Qa6-4K%}@F})E3 zk#~9ZG*q_8^O-&%;i*-ujz+Qhfv7fqITV=Z(ucSgg8IlIZl8Mt=XPiZ$ki8e^5~)H zN8-bxzL^d}&3N!)FeSDk;>GRyYf39#c7fgjPO@K@q+6Egsyae8%)U?GpR2;Cpp~W1 zFMdGmu3A0J?7AKI-K;_=9Q3ACyuJMnD{0Q6jMu){ZuS+W@46+ zhJVKMo=+ZKJs)INLe_b=y22_Lw{{w!V5Nc6Dy`?KG;VM%UD*MeDjkX)9$4s7B)ALP zIjY8)7ln}O0yp)EQ^NLu%>X(Iw_V}-*s>PLL2=5-pw9`5M9XHM6B_y(rwOqdLsq*O zBR7r@M(i;31O7MvYM$NlimaI|7TyHDqXwC3cnG9g2@W`lvE^K#qx>36K_}DiZX%X840Babzo9pd2T_ zMqprnppEoAzMmhw8?f(XpObQ7U6UainETn6a9b%sq>fC0w9E4oQS4;?@U0tW^Hz@k z%*vijy4IxP|8t4PUkAq)SxIU{aRLq5NIGIpD3&P8#iwJ+#ni}< zuq2>f*&|PiiucS5>zCaR1wB>e4tkZ_=aMa=dgXEOhz|4eymGilWDh+1xmr#$eQ#z1 zC!brcXosC^JJ}O}jCEKWnUB8kr`{L%m#|Aqhg!s z`k;rQn|P@}JlYiA#pp?2G4x`;!M1jEtOSL%-P!HF>r?;sdv~LoTm99qx0AK6Ou(+t z1Qcl$^DzaZ5Q|wWw>|fTgJ+8=x|ohqK%+8-szZ?(wi}(n@qZvlVt5 z{}*>O`^~R!Hm8Ge+VpG-1TSe}S3_3JDg$kGGB9_m5^6rkXE690kJS^KQGid{+HrpI zsM!B1D>vk|&-Z-H4Kmt{vYyhVq+l`>Dv!3nRTNWBky0wQM>=HKY}Lt|Nozw$^6#x2LAV0Kd=`iIxNa_J3ns4~SDHJ2EKB;uiF%dnDY9lO4G?kGcXaEWzn-3BM9y1R3)hh%cC(EWyC6`k zKgtE!M=>CITTaF96l61)fUDX?=RsCohF_ilDyL@`a&iK(ZfsDJAUY@Mkd%3;4v|Zo zs0=6EWRmlXL#X1K#JL#=h&7;reKkim^1yZJQs zeFrGYUT?we?3}NOjo^{}<0~Cx!(>uo^3*dZW*bGeQn7{7B@Fh0@am z9@MU5?OvK-WrPZIg^IW*-!!DG+U@zh;P18R(Y~NRdtG8>UYvK2_6O-H3Fa#t3rmA8 zNk!oDV9*T_?gG|}>+?o9ohAYWW}{{uD9jp7z(-b~a9+#&F?Con*Jw6AtchtL#q8!B z=Y=~tW@7ygPz-1-RZ_7SZOaTRa#Lx9haq?hC5E1m4X~S>XHc>hE6s-7k4@be08F_x zJ|mUTsvK^;q)oappvGsZ0R1u~wF18CLSin*wV{3VA zF)qN6;Or!MK$__lQ%@-yfq|-nQ#Nn)l!3^jf~vqO8SW%*_HXjNJLR%?4M<8}P{sw{ zoSW}jD$_ej!seEH9Vkz0cbK+u_J1Iyz-M)}ce-~3_PR&z|e zBGG#b2-4{@u-7qlGtNv?%R7}%BKFY(!ik!2Mvtyl2pZjBf{s&~e)2Q9(N1hh3OYwN zzcRiFgbznqm3)eUX)=y_O)Ikg&kt%n`sBx4M+RY0*ck&Mxwy-wZQ(Kg*&O6?Uf9u`E__6L@77o+xh zGl5bcG#+R~{vQ|*>mXsTfwsP%zpnNxd);U|{C`)ohW6X*P-ZvQe#*i0+%R~M1q*K3ial* z-1j*us}%lZ>z=Wt_s(ts%I6+FyBXeTqxQ=^SD)}RLZx{4`;{cqiQQ4id>l1#?4cM4 z5f>rv=>TLS0nyUbRj$jYrSgZk2_QP%z&oW(1&YFFtX?&4S;}NT!{^nW1E5Hr4}7Mm z)mO-AA}1C1>6!%(p)3RQ*BoL^#))M;$e#nE9Zs;qGWF#E@BQ z!mIL_VumR4fQnr*wM}|Mx^<@N5rnvEzzxlycfl48Ccy$x0`~~#k|0Y2f*}t`8|O5w zl~>4?OS?n!!y1E-!3((ir0E{2OOn0dci?(B1y<3AC9S-C*sCd^i@C`5lnME_De|-8 zK@u;B2a>{cKK!Icz6*;!P>r~bljE;XxXt8TimGut4Dz`B;Op!U1ed3lUg@zm3^FT$ zY9ohUCa7^M@?Fb;U5;A%#v5zFXlJ>9285g%5R}X0U=bT$xgy*e)EK1NAYBg2>#adh z@CbJLWbm4K;2qfme-c zC)3J%$x>dGWb?E`+|HO}ze3L0z-^uy{u;85SFY%WS}?7?GSzT>RahHe!^iu3r1kVR z*K`sWTm;oXRowkiN9ZSxaE@8P;02~(yo3IHdsfDT`m&meHW!j*vwCvk9i(Cth^13Z z8bv;)VsTps$-dGRZ6P&2D)jMl#iu-VM8j2S<#qGaLRA}I?c3>n7xsltDn1Ty?6P`f z0BwTPNMi^7^k4g)zm;n#14*z7bDFn%rE;)DHxJmF67@8p$f%E$PQxxc1B)dC0qbX} zZg?ewxGZuu)y+%bsfFkf6x(6D&EnIr=@~zJc`f!k-V3Z=S?7J@LIi*Kjd@1pD`HD5 z2%bzuf=tAN}Q$xLamuf>rSeoHHbjQmy>*g@xmR9D5v%g47osnjnMS z2Ua+uq5{ar=HgivRHra+6>A_Ml7QSM7{X`_1Xl8N5F_XiY2{^Jbu&~$9+;hN*lEdw zvI(vHfTG-22i%jxE;VitrQl=bXTLn3Tg2Tub2#*x7bcGA#e|-?AUFCGmtEnjrI@eO z0ji1^3&HS2M!w3_fIvZN>*huVf=V2utNzQdL#uLwGV zZh2^@8|SQV?fm5jWZPu2&tz&SrWk-rJ{4O5cBUB^kkE8szR0}@*s{awkAXM#g<~-m zve^#1TqRIE4%#dror01mda;#(fF#~QUW;-u3Y2YSKx1K$ob+%+u!}WZ#@Ur_= zc6fdJ(UksQ8m&l)Tf{7K#fd%NCnn~tmtuM-atj#Z-nk4>*3H~o@??Ql*5qr@8{Yy_ zFiT+zvPCo;tesQj-x;&#?aN|h+Rt*Y=db1T1Qv6*%))!q6??vRf1#n(wa{i7H=q)U z7W(PuqPv7`)iMqLAm^~;hyppln%}xC-XEB*%y&fwIc&Z2%ZB!61KT|-2p(g~W(wT5 zYvDEdLEw$j$qaYb24*u`L`Pi>ztt0$J${BGWbW@HSHv2A2CWMp)0egya$}lC`IiS*9&!Ym zhkbxF78J&J&N7fr`ee3*8tqh;U+!kI%ZcsO2@^Y2OEIANxffF;(gZsC)4+O2sJj;d zft>iTZ@YL)BmglU`_>cmvErVr@^xjFn zH~AX#(zRD4E^MI!FVVG2Sm|9PC<#|3`?b>t_?xF?D9&>yZUOjggsp_ovj!7%lyF+{ zzR+k@R;OM0HCaDaafuUG80|E%G}|dAl_FcHSe>lg_cm`+2uek(P&Yb>(-_nmlEiu9 zQYFWuDmf6`qOXIwz7vHAD^W1oQcMUAEE|%~XaBlbX2eIv^xyu0q`Wfu-)a-&X5i=2 zn{)5FuMSFAU<3&-s9bzIc|c_KsBuN^=H(JwC&VJ^CwOb8}efx&179Et*C z=a#`RmS38*<-~dOFs{=A#6o--8Q%ux)*#(V=+#hFmt@7P<-z(a$wBu|A-@(Cd2#Ov ziA7QJqf62%ESsO@-Y;M0hTwV_{@ItH0M-5UdrK1}D%fJE(lMY;u`-@uJg0#X>pa0nmYu2jSDwh3gp5 z!VJMWL^Ti}qZ~{ECzTlj?k0`fHHeDD33`MHyjpIDB2QTyd6c6Cc0osCL_f&31I1w& zG5lVvA{*`8js|bzD(6QnlTMRVcB>pG4%O~8u|P!>1M+3LRBV~o6VP!O5N0y<^dWNH zO&5Naal>y5s6R^}%ObrD)>an-vUCcV(P_PD)r*gv! zy{kPAi&7=Kxu1pC%-G;pIt>;>^dLbFkBi<}qAsS%7wnZ>D_fy_JVz_rs;m|61rXhW zT&WuoiJoe>XEXPcg`F{1#Bssb7yfzo!jA9*LS49qf0y0~iZk8(YhE2=@BgqI)$u{d z)c1M%iRJGw*6rWgz2@zgeJdK%C@{Jf@1{@QM|UNWNVt1D@^r6&AW8P&AC{ZP$U>~*U{?8ov^`N z>0JWrY`t=WCe+|754&jinDPq&Ilow}SED zE{ z^m^m#YVX?V>53j%G(Brk=Bm;X&De> z#Oo;aowAIVJzkd&_4YuAj9WVebd_{dNR_lts#RQ`(GF~ueX<+mlkhd5g_yw6=YjUi zw$5CPN>{k$lfY3&V%mB?S4XEz!JN_a4#@@`H|L1nY&Zfmtl%;!XVSs<&541WHV9*Z z>c_xpPyw9&yQ8}pL!sFO*o)@}&l(Hzd@IMFVFiiTstYaj%s!>pC#KYss;`Y2`eJsx{JK1H6mj4}H-@z4V= z{HPt|1d3Tpk$7aO%;8qghbVY4tRi`@Oq}hY z2~;X5ri>!Hsn~vbs^6)peGz?hYCs;hfP z=SQKz9)|^sOW!k3X|s?ehVs`)Ijxmw^Khk#VfTr|1uL@MFL-h7dYK(C_;YXn z;VVX9l+5JZCH1e&`sI=dj9Mw?Bn3)AvFbTJ0a+r$zbW!wwjHA&-y#ui1NYD;WbMN0`PXM-c_8|^g^DLZ1+y=T zJGsC`Dr(@|o(7H*zk>uXz+dPozJ~VHi6I( z3JB%XkcitDh&#$X!qrh{fK$%iODg-IOmBO>05|L||LY>&!#^13uRpf^VmA5Ii5;6Y zCIs7$DCTpD3{bHd^b;3MBS3}bK{+H6q$$>it(8|vRXGsS!NAY{z!lSKg*)j|8r>X- z&)0@v-SI>E-pmHh2EWg}Pl@5QVaqQ+tV%lUau@p4LN%ePL)sXW8McxKANV_?w6G(H zh);*_9r{wZ-WcTn=grQPQtw?%PVj=2Zv zirb;U!aC9q2AmO$fnLM>;=9qGW;k&S`) z^#h=5{A%!PPxAZT(eHk7MVS*DCkqv22GuGk3Gv(PwmVu?5ZDutDltrkIKkm%TU-`) z@Ajy>*_fYE%^ZGcd6N6Vx?|)^M#I$g-pzbc?!<o>Y5zM zGOFg>0U^B8OtPFvc#`}41cWO3I6iN+2No<+M* zD;$u=1*iG-xt}7J1nYoCqh6xpYdPu7V-GPAH z4GYw|BvsSba8#$~?+VxNRmVl#7xxHDrmBtdSHmtCVa^cUQdUZlr13a`Z7Aq6RZ`3~ zT+x$Jb#OLK=>Zn$wH}Ruk3bf-Bpfs8Fu;K4RZXrsNv3NXoU{ptj>6t(oZ5oJ1@C%m zl}4j?w1YhbrgU-+0A zmdM)^)fu^ZzG_d@?r60){7Ch#0D<5F`sQ3nzcRGGBgG`F4e}nj>O9jgzre|wfd!E( zcz2|kK<2wBm);Y#NYyUP;pzaDJ#-o85>z%N``HK|!_YbwM7F@k&?Y?UY>hd`j|D27 zl5kj&(8o#EjQxsT!d`w4kf$2bTKlBwe(S?B!?51eaK<3QPy)}dVD=-=lJE&G zkaNjvLy82~xl16nayu03$C9P@BGuvDu>YJ(cgvDU53G|)!W(?Y^|VJn5Bop)(Y@P# z%xdaz-glMvozSWvqZKOaDP2kmoY*xwY+_fcC?Jb2H<`4y~yFfRRYz$$b0S+l(b^V2MvN-6J0W<<)EQ<~N$&+5m6z`Jhr7+xC$G zd(vuHH%6n~!WSLeZ}?b^k@FTFFZs^*GCYjn`Nc)Dh$OvaWl5lfH7cDmi(()*JPnA& z^jc^7JYEbY7toK~(>WD^sw_SEAZ}?~mFLmb^AY|1@`R8*foFL;f#$)!bIw?Bvk}l^ zb!FFcmI3X#@KqtOR8+xx*_ zHtR(R9?}Q7pyvoV85{goOItluJHxil>{eEg^$hAd?g_)H9x7PBM?~?2?d6PP`<3 zWP;l}6my#*H-N1r>hn4Af*o{*Uyesv)cxS9n0P@mJwR4XPvxH#cE{lL&C@yqFK}wy z@VIjJ4jS(>oWuLNfzzcdYS<0WSMm*K@Xp)3Ld6am|23QgO&yWpK117bQMa-q=H53d z|D}In4we7*rMK6;eLBW)ui?%@kXlPuWb%f^H8bu9@1UQxD{(ss>OA;+r~@kln+P7& z%)qu5j2tU_G&1AZ^{43@F!G7jn?U`&;n0@x2m?Fss{|g*Xs;%PVm4A_188}|j&{GS z6Sk)@g;13^D<`6dE|ph{2P4xJ@nI?p!#(nN?fjDkitPN&FYYdB6}{|b&mq>yw!EoY z>v8yLItTz^RihQQhkI-(P3$!Z@vSYr(fd zW!cgHt58xY*$;Pj`ez>SgfyyD$x#91)`;!}+zXf>lfV)l6Xsm8no%b8AODYEO);q@@$oE4Qd5mpFOaBSR|_S z&QlipYdqQ@rOAh79|41VF>|!X+8$l$8Oi)uv zF~Hcqhl=fi?Cx|$EB{mRF`tvc3El^$7jjlj&IUH-4e}IjZ8%fs=vr_q*t zuVMRHvWcB7apD=M)WnwLPz(@tY{zIm?ij{JYvtGB-?-@N`Gz7|RkirGeCVqhex9;h zhQa*n@D?nOo2Hkf#&Y`A(;kU&dvAin7dTALI?C*(lNrVTXfY>BcG^{~1?uAs+)u?- za>KtJF{fiXV;a0Sk{(ED=!n_E!Q(asekvEM>C-U@9B?@obSulf8pY+FgW**(KoEj# z3F(Zv1lPBS4soBjq;NO-#|f4RHgH?KH;LNmT|#KX>sxR}cun}k<*cy4KKwvdLTIe{ z!{Rx6SP)haq84|PUzroLJ1vAPP_1i7-1GtR@$6G%>+Azct#6KhtK_;{rbu6lbb+Il zKLifKLqJ_xC55cm4o(86M|hQ>{&bvRe^iV2mRacv%*{-h;;`vvykM~hFvjYG6!IUG zEqx8tg>HneXKsXJf}3iE_eq~r$soZU%ede}yj2i%A9hLLr7JGFRRAsWa?V9JY}zGh zp^@aMpPMzK%&XAfK!Tehze1Nrr7Ko@4#1x0&AD-I@zgk7+VP^uqTLy1q=glVCbj%x zcE!uift&H(MJSX1$2GYcWS2W*wgv7D((rFhE1a|0z1Y1vDjNzq3c01Co`8Fj22N{8 z58Xx%afhN4q8(;op1$2Sg5+s?M;)*N3FRidwZNQo+iB+|3k2I3pU3F@HTe@6)(=4b zG;^G^`y2=WwN$LdH(bJ_^NfbZTlAttpGf+}?ZS0f!chQ4v)#(Au<@6!P}4ObIB8)z z1u_R=LCA`rVSmFFJ+07DQtDO9MYUfI9|e26X>4SBriG1IdV!rWyJRESSYc_>kDp{@ zoAcEEWVD!{0U&!>X$7%2Uc+Q_nBB0kQlXv6fRQOP&F>_0!Vi`pRpp$alP* z!J)GOA2yCSyJoEL@tUmqwSMyi7B+Q?P8{L4knU0flUE!R$F}HaFAcwJe!7Q_?v7|8 zrC|liw5gARE^xNY>Y39^U!PYKwLG{0=CifrFfSkI+#5w|x}Kb0aM8P!Uk`TXG;@UC zDt+vaeI#&NWyoxF$v-KoGse&|nY+R7SX2-F!1FLKNxEX%X8&!{J7RTUAIX+nXEI+0 zktA9jsH3Zy&X`=~1E^=rf2#(lzjXmmTpr7pdUVEM&#kh%y#0Zt+%h0gf)pBhn<5SB z9n->2&rjjD&9s1a8{lbXpT_%fY}o!dFyx;3y8RUaoP|~XVQ`Pm8QI_mNV#u6$g6e6 z7ccIT98mU=3{jq-PS~MvI2Wt!PnOMdVt+5Y zxdW13iQAuIfBn1I-~Z=NKUnl9@k)wWMv=IQ$|1G*_kAV#CwYsFYjw5nDH-X0Wwx4D zm{8F?pqP6UxeK}^@-E4cdjX9Sun)O?Kuoj`mfs*)BU5#UCU7tWqzxV*!!D;ocM4P( z@KHw|B?+P%oH{71(($`x5DG|<-{2rU@zYbeVLJt7UUveZHXkTyG<1JxihQYH$m2Mp zlRfs|CQ+rh6-YFSni;B*Yw60m8ABd-dBZMkQe=rxC5ZZH6t_rMTo(66faFD<5=)0O zfE{TM)b)3GCW)3y+ra9iq+BI>d?2L*dD16@fBWcf>59jpL)<|Ez2O1S zaoPR?jdxaiar>j5e)%tZt`(MU-NK@2E{dQ z9ihX*iiuMgU_0OzZG)XJu(Fs`+VB7Vv$;mY^kGd*11V-_n4H+#K4ubfJU}tk6se?Q zH8k#aWrlUe+>q`A%7tF9UK#Rbb;J~YZ5t$ecgCPjbspU(u9qZ)bc8h1c*haC4N^UC z&Mo8&f_bV9%m)AXFi3+I%r@Ku{Eu)~hjip#$PFg-_M;PK3$_>AgmDHKM*((=&wifR z2p`!$zS2Q9IB^WR!~_r-6ax|#Td7#g?`;gM;epb0NjR#Z4Z9SIk3n2{x9s%%A^C%- zR$!P|8KJ7DPkZl{AzR19{B%>#Mk6Vt1YF~JYd>^Q0*7$(W*W@}8 zPyOkjuzCDX8%@9gt5?){IR#VyYuESQGR`g$AAO=BHSC-dCr-WUGO;u#DF(ciW-9h^ zSS|l^*dtGhxhy#eQui0+i~HrD18sd{NK)j|m?8HB&bFzC-cSunhCCjJ;e`%`YTvAE zCWqS`fPT&4&~8}p=o=nY?xOSk^62bP*nluzNSP8(#n10nu86D+StV+fUVzT)WeJ>k zs(`)+@rJwfQt~;c8o2Tek%z}&w!BYlUW+gCR4iV%;Whf;y)XSf*odba-@aW&3Z2+n zX*NMfHN{j?L`}th4qBfHB6UPqq0Oh zB#HCyNvf%{fmxyg5US4gPvGncTOU?W56tbG2?FIHbKVUi=P6+Z1@cvbb-XR2gpi9N zgHg7Ox}^cJD;g|)i4Esj8A)pQruV|lv14I~`grl2bgxY14NjhND7w&Vo!fQpb}w}# zR9c6&NxLMy(+}}fhLdel1J&+YWdg6C+aFmyztF1(SP^cypPoPLk}1eiHt-I57gN}! z;i#T&1+7;+H_r5h(ND~-Lu9|z#S7{{#>(e*+UVcCW$$c$&1iR)uL?d4Q7=~DbKV&m zlFCPW%o`|X9Yt273LxriP8@r?VPY1}Qw${KbWpK1OqNJ1(-L*$8WJa{ z6#_e&`05O;tjYHYPIjW>oc_osF83rCgymv9Q^-B!UE@~h-N3{D)scgtTKQ?_F}OZg z#Ko}N{TV+K*b|?)teL;;^(QVB;`pFtv#Z5fqKp6?e=zj1e-B+av&+Ap9+Y&(q{wx$ zjX)fYzX3?3ztJKB?zWJ0#hNJ@Fr_~BKffR&pkB6O_K-Z4Vd4I;hL9j#jD3x0yZ5MmJwg zgUdsfapEE_MPcvQUuTTrY`WqIoy9#vYTR`Ed^$ssrA+oK^TIn=*cx-pSO-0pCg$nw z=HF)hXNlXJe^VI26m$Hi2T0Z{118`CA5}|LN-?`AvXhE^z|G+;_fK^{7jx^4Oa9kn z%L2~9goCm?=b0r_dm}PLnPDp?kKDC#a;fZ7Sdnl1*VOOSz1=&lMcNqytaxEtW-Wzn z2h2dt^TM;^fQH4vvI}M`_HO%y*5h-QylceF?@y=xZ*q_wW}Mg~zG{M*PbdZ$=1)?w z?Q|xuLy@lBA9Wl8c9eUnyqH@0`VBC4 zs@vj$sAa%ntZ;d0W`4TERo! z8HfsF4GcmD1?ZLqd{oFWpbg?jvT`Vw`RKN%VYEL&Okk4BM~*NzqXx`&jEw5U0f(P| zXXCvu8!gX@P5y24Q^woTd%yV$wP+n!KR1!7E4=WxdJ9`Li) z(>L@o6iJbtOa{F~kU=-YN~!}a!WPaFL2AGiI0tE2D8r!2d^;|BkW|l4or>%C=6B}1g140-IDdLNCSWK&2X6p#gVsb z4iwL~bM%=-$LB4ya+6z6n2XEX+K}B-u}^#0d&AJDQ@6GtW zhiz+C3tZ=YrCNKoE^#+bKh~!MXhNC7!zn@}&OQM2`JvcqvuY%m1vx83eYn5O%QK5L7WPpIwieE=aF)z*Uve$i? zTDD)A1S+&oT>9u^kd23QcQs)DmId#EqjPYi;XnT2;8Zn)>C*ht6d5ynqck)yHP7pn zui;!0AE6IeVq~`JNt@neJ03N9 z1

    _-Z^dM0$$mT%|#F;to`vYyf@iBDZ2{mEbmnL2OucupboLFv9+O*tTGWkx5sM z7F;nWbbduvHK0~9DysW2#iUSVBNdAR#d*p+FQmmf>VHnCLiScu1E>{m|GM6$UEL`akWV2ldDW3OQ=nr zDy-{TDya0{9Xu$RI5u2%4|d`_S$1!z(TFVu7%)6C{+H@z{vwu+0n2nDlyLK`G{tyXz47w?4qK3KDN#?(S( z{VDNqXpujjJ*g<>CXx-%raQD&xWTVr_I}qiKfDtZSfGdX%(c+r&?eGDUkowyf-PFe zUEX+&j12&P=9IkXXq<+H6~HMM;kq#MI%_slkQ1+lER>7ud=3Y7#-Mb%n)5_?1Iq#%?N@R3;&EjUrx*8Uud9Fi*CfI>8k)DxZaGNS zuv;!WZZA6I>jYGDWpdSrU$MLRFEUR&fHPj(1KQ23oXu zw+RdzvIU2|2LFrt>G?2mhPkeQ(=Ku>Xb1p1j2CVIcggdW8MN9fL9|qG$v;0BDn5d? z%=(mnSB{!_&cI?xp%|#jFfDGk z?9>~o13@dpAVbdUUc?5!2212Tju;G9$e6&*16IhG^gG{GzcNo|a$2ilAw8*?#&nS+ zazIh;3$zwX1&toKx>B`CHT*Wdj_!}%?~06g{c=nj(Nka4xFKa-l0H2<11gKId8t;* z)8t28VYhb*^sv_huR|a9wKeQ8^CHL8zJOwfnHgIi4VtfaEG#QD{5ZjBX>H&dW)KJi z5=29CZA7}FgxeF|8B@q9^V$idU#T!w8FY_uKTPiV+?yaEjU*wQH#A>*lcUAC)k7TBK(HWcdjuWRZS*V@sl2nIhE8C zAoVU?65a~-r-YEPT&+rUZ2y&h& zR!!b1$YwHupL2S29h1_(|F6i|-0#xCLczB%V*D72C!I#It@SDKt^~&d#dZlw~+(!DsbLm`- zutMBJ*DH_H7lIlT7??Ru27o&FLQo&M<-S_35nc_s9K_~LA%2x??hQqW5PB*yZYv1bUBZSmqNXEqMOyrer^g&lVp2CdEz3#~(6()M) zDRrb)4kBO=WU4H(k@o=dCAuW4Ue^?Pk)Q=>uA?ERQ$u63er~z1o~1aGHwa~St03t& zk*C_jy})^T1rN%+;CDDDEaj*pu~j^UaU_dYaw5Su7AP$Nvg&ljAOE@PtPKgZ1#~;e zWx%3L4Lb)GN6pexv&VUB5Mht+VcUn-{S2R%UEKiLb#s+P0!#?)lcq{AmwHLe7EwPR zxnhr#)cLRq&zGxGo2lG!54RDw>vDO zU<_OHTGCg)Rq-XG7u@yU&3sZm8P+GG7r%!nrhy{0BVwJO36Y&x!@m=%&ty9sbb(-D zX)XMRCxDz?vP-x&2vcR!%8{5Q^nhUGghohNerXCAec4 zc#YPAXp*6EeposX78QxGYR3+FH>^g>!b>^!>G2I5tI={^lmGWG9lSBi2qPDfvWn!r zGK;B$CT~Lp#gtLN%NctMtskZvpaZY z%(l500qT}5;jGmwVAKaS`Id^RsijOp2%byiJ?nF)Jvi~~2aD0Rku3#T=qY0~OmO*gIpZaw{46*%8y?ofdXICi&f+ zzr6ny3nD9+tu&&EzI~y@^|t23SKsY;lv^c8MgAf!n>7i zilOL1vFa?RkKRbK0{`og_`ioBF;GqK58S`-k+^!{5Lbn*8oVES?+;7~8&6S}V{rEr zgUVJH5Uds67`U`*I5^{SVxo4QYV02t*PI0+#Y{>}83 zIZ2^-2CsC)WY8JGg`7oKezp9q@-OS&T2$d*;g;c-0~$=mvrGS%{#>3Q$K!O2ZEGvb z17$bNSpEu|u6RHC^aGjE`eaQ1?H@=AJL}`bd!lL+>yt+@*%Zm3VsWz{SJ@>V1-^wG zPzCb>-ZVI_2k-yVch0<33{hJo3o(>kV?Y^oOFBF^0?(x5)BbG6iA|$#eWR8YXBPV# zaq@@Vhc=L$R|Y4rJ{&bmsVQa;MM|hxq*p}B7!@iSV4eu>R3QgvkpOg-5kYli)0A#L zP%kb}bttroUe{8ui*ArBhKs%wxq$&DMGmBltNk&S2#hk3w>`6Ca)4XL&OovmHBTp< zX9w0};*0OR|H0(%nG0I5nUS2>x>%6EsF&^cFOD4Ic1dtoXxQbP{D|U!Kj?K2xu?kY z2G=WxT`+eL)yP)yfU*%v=j=Vzj5#c}AY{yr?0oXgnE&i}*)`hFFQ}(;X^al3BQ^Zn zp7??r!g?B`))nT=#0enBhNy)XI9-yI;OoGB2dbwq6QOCq?9y&W?lsvm{EjgRrubpC zRC;aaaM};dHQL#j6X%UU5etg$BRkwE#6IM%gWYJ@`rhu9BEJg7-0&e;R6Rcrh!XSY zI*@Y58K~S>l@zI=F$}3H<{q5~M6L9u5KSnCXCU4~*KiX;uFhN!)xF@_@L}S@rg0Gu zWI8y}$pFz6IzO3_n?PS~e6!T}F!v07V;-5ph}EYDnm_%SGvdk|9XLXkjg^(-#8}BP zaacA{Od>_rQ?Uk9lI#a5>NvRDy^2I?Sq7a9Q_h4;F{1~;+OZ$qW+IL)+Z#&$ry_Gg zA*W4Dw;<;;NKSfm)5u7e?FShsShawZVWpfk^RW1Of>B^C2ArGeXx@kKoiHzRVl(GB zu?uUVC<^hTqi=g+SVeU~ctHq^-XTRI%=ty}VjZN<-<>G8)nF5>M-t8z6JbXT-@z*# zcrc!XawiVBSx6`^^(rR$%2U&_=musfZ2TC9XG;K3*X(i0lY_2!hPTq?Z`(!frq1AK*7IS@a_N znJRYJ8PnGoebR2=+IDy4U(JcioOZskz@T=Uhb)DvL3cdZ>)t6_8KRCXpV>n%#-z$U zv*7@vD5sQ<#HrE=N^h`XyzMroFY3s$=P^cN!=yK+|9g@-=cUu4!vc$Cjz=lyn%DUS zs!Z1#q)PHgJfiz-znA+(4}A1BD?u^3$%GuU0>x|J`sp7(Fi*&JT2NS&GOC+#9}*9c zTpJwJi^60-M>8dlIg(Z(V$7C~J;@3YlPa1%Ofk=ldqtenJd;K47K%xxND>vhD|{eQ zAA&ENe|mnUEStefO+CXauo*&-Bu+4~t6i&bG~z%w7CI)q|3IQGeA$HgFD0TBU|B%i@k3qE=&W#KKtlbG)dm z7Onrj#27*q-WC3yB(mELcH-bFDA0`B4bG$(aAZ=c*rdpZbQ?d1n<8HmlLox|mtvmun=!81f=c; zed0IlQpZ90mecchg(G*d3WJ-uO7sF8Hu>BBm`7g8@jlS8+q1I%&Krx2Zv9iiKNr(j z1pvJnV)Y39vm6YOG;ko<1ci$516B>)kzW4Bkh_EJ$0%5_#Q3O|jy*i%R=b%oniBVi z1Lw&WcC(H1!X9Y4N6$6|6q8GlEGia7r}hOHLhuDp`;|PuHsrJLNABs0=Ac2z9cjVr zb)F}L`C;ust=H0k6!}iNl+N+!l2idT4X&(*96al=M9PQ>#Gej@%EiZ=Kaa;RaboY* zf?uNHuOUZ3d3A_R11Wms>&Ot*b9M>iCuh*hfj9L5Fs2@sWP=9ds>v+@%wlEjK-vR`1_S4^OZ3swI;@755+)^S`ihi;SYWFqo##<#b`PD+=zKIZE~-C5+5(h-ZKsj?s(}7nk{PDwyE7CmYH0MyAo~$Z zrNprYc z72Uv;(Et*5NX@c5`m_HLyP3qwr>3GWy!WAb5g3~otP{^U7D~pJ2rB137BA|ES@iJz z^0({1adTm3Oo?YEeV5lvAN+FWch~>u#=_jUT7ETh^)!S`YC<~QG9U%|fM-ikrP~A2 z;{AYB_#2eRPY5xa`WmN@tCf8|=bprH9p)~$qeF_-xh+7>fS<S0K@a7Xa5E>PB_O92MPHVk-@-H4gctd)|X0 zJ9|9Z=l(i-JnQE=wZjQ^lY`T0NR)r~ebb%t_s5h>NVj%Z8 zmx`^Iv`G!Tjs|{WNEbW<*wDfT|B(A8&&z(rRENUAkJRji42|iEPDmnMoI%(4s3URP z2Ll&CPeD5@yUxE-+vC?}KsTD(<7f?n4Gda2-9O`!eF1}hMZ@{!CA?T!R?&QhXw@52b= zqUIN$%qtjG3eJn$_}kq-$^a1YJCCjX3F9D7%$K$||MF!o>?sRKBx~hWQYh-Fo0qQWm+unVe|{Mm_ZR%M2C6S;?yyU&p!(X{ zmgSSp&Bya3>P{SDwHWZpQw~KJdLdQ%6W|+-k2&rAsr(^H_gf!!MLgt@I#m~*;nxiI zZatIk*BOK9koT2|yxrV-@PBG)oxGVO@%DH*5L(s%=()+|=?RD7=LJ~_caL@S#u)t- zUD)(gvY%Z_lM}<}oQZdHoMMhp7$V}H^cvY(0F(v$A`10fU|U9GY~(T^quDF&i!`B-bxN)Ndwa5BS+pi&eCf8yNYsWxe)ceg=0s-E9W zw@E?TLz$<U1BpQsNZg{BZi;BB*k=-l80^0+#@ZeT_{RkoiIYL3^@+<7F|LSE zh)T6DAfH|jHI!>TwDP}Z(;K*;N3P!$0FKek9<6j9=tn&cy&tMCF+&o`HYt*M6+k0A zoe#N^SQyab-X?`8^1^HKf83YpJ2u3gPbkzf4ZlZ-l%>EiXqYA;66Mk6S<9~r>Jc^u z7W?iIBtvy3HZB7>i8d)->k-0lyh*-9PzXt!&s#Y=TPsuBW&Nj7Ff$xxee1HzTSl-& zeDsNi)Ht!T(`5pmlN57|BF$86>eMr&Q!$7Y$=!4k=rR7BAMxut|}qNR+1pO3}|{$cT-Zxkye35M!6EF`)=~4hNP`u$s>%9r<

    G3ra-U=3omsx$K0Ic{EvN|k(rxz;&$2! ziD8{dpZ&W(kmNUJmsD!BT;@@8+#P+Bi|(o5x5TkZ zLy_bVL%hY36{*1&-R}jpf{W7rYyfn5wA^lqo@^8KnpR+iI=+rFk&WU9pp>hY4-yT3 zvn(UH+#U5Vx@jG+9SAegsf#;6lqWr)T2kPnrS}965xiC!stH7%m=&mcfm?83XOy20 zxW+2wqfTa{yc5JOlLNY>D^{$C4$$!Re=YtJv~}I|W?3VYd9K*w(g2Kln%7Z~vf+a- z9W%{9zag6@Gx_Q)suJieb&Jz#j~@D-7=gDumUB;BFHZ8dwHsq%MJzX9CNJ9f^yIr| zZ+^pIl1gVi`2{)s#&|VaBd_K%C2OQeJr&hSa_BZGcrvH_;snS&3EFSa{fP9FN8c{w zl`X9F2frp8cvzFAMeL8lRQNBTz=Yk@N-7m-#VR?6L=QMf`dcT-qvHg?035xjDPr4V zDEjstb?5^Q&Uk6E6!+{%6_Oh=%wCR{j{9wNsa4>%4%6Pq|BF|FNWyW{At?fl7@D;h zmm-!bNjwm0X$a7AWR2NvP3q7xNP1ry{M7;jvX-{xZz2WbEp~RCV>)AmpQDrv)KmX1k?O|oG212u7D_JJnmw)yLZSZ`9A^GcvofTxycutBP@5oOX8JEM9>>x!-sHm;dmAs*afM!9nyp{h% zc*;?$6X_mxsPM;bG-z@}I%VvU!EKc_ld2O2$oQVDweay00q(yi?cw?wpmTcJmLnva z+0Mz1J>TO-pemze#T3~`MSToy|Ckdv4iu|-uG!$NrubHHW5@&gntx&VJ{ zIU{$&9v!JC(t9EWXzy(Fdq3hG^22)5kL4+|_DPP+S}jdu6+x-YL&-_&^9NZAP*y{d zVW77!T(~)zG=lx6<%yXzf=*8-cTUe7HOAOLUUndK)M3cEJ`g$S4$a=cQXP};ewqBH`O%x;6#)Rp3b%nycrUKbQ!uj2=ejW9@U;OsH{}R$N z2SVAg0RbNWF)nWgCEG=j6e{Y*mq8?3dN?41^9YQ@Hjk6+eG=pcKCtAlR|cmV8UqTv zV}WukMf4HxB)d%VX;6uu%|U=4{zQOaI%&2zH}CWw_@AGB@Q=0n|`~Gtq!t@0BSr>#@e7dn2{ngU46`$?RxRH@El$a9nCvDN0l3l zX^#gW?zsyTg0Wb!23TniiJ@SzQ`!@vs^aeh{XI+_TGJ^s^It#F57*3aR=aHe+d9q` z)qdCU{(t|EPgw@X_75Koo*}E4IktAZTgo(Am$p)}1d41%y41D2J5p7A*e$mf(LVAi zUvzf{+^-nN+P4RP6VYf|j%&9Cc4q2?V-+#1iimQjA!yFXVCAzi zIF~#=RaAj~Ud5te_EE(R>xJAc`uHuL>#3kI9rG76GGx?W&Xu&i?ozUtk;?|13Y25S z5{7M|13=S`e0iwu*cFy9#M;V6|FqzGh{*Q>BgoX)1qKLt=01#>N824TLdbN@oNEEj zh85+=SH96eHZogL?0BaG#HV9+I;oUw2L*E$^|w;L)vi4tITiaBcAquS>t>YlYe7T$ zA~1aHp)U(BI6tJ%%MUtDp(Hv}EZGh{#v?EFZcUXM(6K$y{|eb=$H|;BqxsIIWI&6N z0kqi=$|&Gr9f4ZDFGR~7Bz4?F5Za37K{XS2%jtBp?4EBuYY)B4wGcdLd}a@5(qT9i z<0CeVL5b`l7lcpe4~BlQmHOZm+0T9uaGv`? z6LmH6gIfQ#z-A?=rovb7x=&*GQ%;I=BZUe2L+tJI&iHjolY{ZiBKjuOl5KHm=Ozcg zcv3ANmL|=)%T4#S-mFc0f?iD3m?Psa*d)#vO_2SX=_9{A_n)u31}nC7bW4sz=)$Sj zNHWw1U_l!&2s-ZskzQ0HM{>ZA`B+(`KUW==L%EVu^rARHH;7WN_70pR)?Otqg1(#wUU2=dUsX_VhD%4lSKb zIg0&m_KBHt6jV{&7qVWA%NhiEDinPL`}4H`Gj6{7nSLu&KdSVIf`8% zFOK}RXG3SJnFO18U7UK?kc^k`~+= zxGLa+$EZUxq2uVW0TSJ;?wZ%+VNi~u#gZR=OIz3#P? zk9;5bLXlINE@~BIlS|%?Xnb6J{+GTn{AA)iNMNWTNp!y$qRq~)puP@ z69G(sV#kTYy+%Mur(|gq*-1qe@^;Ixnhn~6<|mNV(BP*PBPw@M1aN!AqOZKK&F6- zN*CN-7%xryQfuUkqh{qva?s~AeTe;Z!DSK;b(F|4wr}20=taRwUY(?!v~y2EsZ>LIy))c=^J=uE1)1tmMZtVBx?x%$ z1Sd;7+|a^obXgOO&%*Io`AzZr!=!$^jzl~5YVR9akTyzoogyu;RrYxTlshVo-)>m~XKx6uNO-B3 z)dKutdI`JhLP#n?DjTo0)A>S34mqp#d0$=9N2e_7j=aWK-TQh?q&Bj{4Sx;;3lNG* z_lQrhE1haYO%c_cy!p6h;S!eQ{lFQhUy+xrkk?J)3-$6Yb{bUV+;)eHYQ-_2;y{7D zt$|x1cDN>RAHN*}K8d70Qgut5&r%J$47;?`7)pFau^$86y|5luI(iv8&ExKP%@Wm2 z1IBLsN9*6n`Zs3%$}>XWPD%#3v#nHAA9t@;^33JSmt!&#Sv0ZdHQA7aT)sTbv6QU^ z9=jM`fwgx?=4OaFYRrvf-Rs|(UFpO%U}Nd!?I*~FHwGJfjIfbH$+lA@iHhnG4@+tdjYezN)NX;M?UlyRWu>pOBfkV+;;6<5>IhV4qi| zX@u8IXf!vthwfs>km$v_o^gEaN~wNmnwhyXGEXwgzz z`*R#;JoDCi952CsHxS=XR`r<@`sW&b?{-kKWQuH|qV#H-S{kVd^jp+U=}PKI#C4xk z;KxUXz$%VSHk{*p`npR@pkZzdTU>ubpb;(k(Nj~d4ZE#uF~h6z*A@R=z`IB{&>5UY zxmri&_5{jVqIrXmU5@py^-C9I1UJw_LhH9A=!#` zoU`Hi!lyyHQ{?TfbBqDz%U;LLunJobOvZC>1!x%$Xr;5hJ)5q)?x7VeY&@|RvjK>9 z)QSwwkq9t1^3qUjw+2o)7i*aZ3qNwpr`o~+t@Ic(E25IQ%`hbeGeeQTfouThJ{w3JtCn@f?{L6FNu#osy_X=H#fZZtkM+8r)$|^p!vRf7ThET= z(_v!Lt@N>*8-~$z*l$@7@6NklooiTa27fMILv}M;ZS42Ij8q$~HYJn{$RCQRsHUJo zUbAwm<2wIN$u<8ip`DH!-rgH_d*NAyS^-T_J|{VS!X)owKKF%NXFe33rVp};={9}` z&{g8Sp`bqJ9MBJ4>3ErSL~!P<;l*k>-d$_O2C3(en=tl`z6#HWv)aL#td^UvyBywS zQJ1uVzOJj!KNZp}DfPmdx2M7;aW@x*>N_Lj1n>{*&Re^Is`4_cj~?bTot7MR43{~Z)=}DRD8)==B$mg14CQU@a*c2(s4XaYCYUlLL{`gd zL%StV-v+Hium^tX*)IPSA|j~VQRDU0eei9qX9~nzDm_8KJXyNM<%tkcaw!Cpc||rT zQLz*Rri+Kkxh3<;=ox1AExLKy_vRU#S_gsR5XoWY)Y`EbIbr0~9;9R?6mU^Sb+V58 zCrd}avF=;hKkJSxT-Lhm>C%nMRB@~xegWk9(&^RWXx=7H97`uSULd$g>cl(!anE=D zr`MOYzrW*)kN@TVFWSGc=WACZp)#deo-D29)^cG3?xU7hg{1{sb*;4$%5^{kqYzsx=si+;iYx9BI1qvcDONG?g=>oMpo}(4D3Y)D@W&8@zSPbGOkz)rv*OaF@_HT#yBLsfv8shfAp;{hZ@$F9~>&oAf?PUzjmC_sW)1QPEayP@l;Y#8Jto@tJ^A|t}9rO zJ~z#|*P~dBLHKT<-OnWPg7mrT!;9D>4qM*YHs_r4h{Mj%{Q=dSbIxbDBM$1f)sj-> zIp=iW%jA%vTN%SEbIWB93P&6^NTL*?ZslWGOa(cZZY-Gv=f@9u%&D zvPZmf(@)jyq|yKlWU_SY%yY8dS(|6|(DmX1!BN;xzoPd1#UID`XwsiDeQ@!L@pDJpiJ1~@f13cOCeBGO%I9I&RFXnEnWH6 zWqO-)kH@+h$a;DK0=h+40a z)m@sGO(7E?y=)w7--~%`%kRHna9~=#b~~4py)h0>(1*kLJZfL+Vke z(ZZuL7$qxoh~%-Fl=-YvSg;nIAXWo3<0HY)G|FV`8&LAHdeg4NN8MrhDsVL-{z{qAR|0uMFKX3+( z>-~#A6&t$pN+s!ZKlcI~*=~v=huKfQkO#fK1%i_cA3*G1RSTRYYI%d_VCdB)&{>XS zZ3xq`(*jveH?-ZglL8(NhofVn&pKan8+`1O{ls$sIp&#v{lodNz+yr!0CI^UO9OYjQT$+dIi-HV)S zSa*WZlRih^Uf4%&3f1g9Rt>9JSs6GW+T&6+XT$-IGM#(fYJddlw6MS@Q-G`0v%7fo z;_iUxGudus5>(fn@{1D;JD-uP_o)_DD^Alm_7Iq56hJfnJBr;#9QJc?>~Volm4>Ssf%W;JNKZk-v&8lIUL zinsbD2cSJ6lMRj4d%$^z=&4T+o$yXK$k+_BszqAvK@#ts@XkG7ttg*$ z%e~j5Sy{u$flXY6Z`B-pW^dTpu

    oK;<`d$hlXN;g7+D0q3g$Od&JgbUnLfRX9Bx z)Cxb%!|Bn|(T`tu)?Ks*Siz%~*LYRUIq6>G-RQc70C`l5Xux@d6YpK=gclFP0|mS@ z!k#HghE4zsR+#En4`Ch*b z3F9f608GYuD8_?X@r)5nrhTb*-H%PBqM5`D?RU2cit}SZ^;|?6gAxg>I{Kibm(Fql zmiQuiAhZ(5T%K19UkZf9hQ|HZ#Tx7 z)@Ir*D$LYjmXcD~&q5?ps(Qvnr^1y6J4cX$aB8t(Auezcu^9%&8<-I;re_D7|ACjm z?|3RZx1Ok(Wg6_bPoTnRLD@&iK)z-V6@_$j9d27^c5@RPTliYoC#(9|8{JV3R)zI9 z$c(C|2D{^~x&xabR8i9a!yEvO5A4gFB*jIxj^rqYm*`O$66jHTS!|#5=Jf0Rg}6})m^mC2X6TMC-)f}z0K_V zY_~PE13|z4q0C@toF5NfCELeq4zT0k$w4D?l1It1DUwM=-J~~=i-J*~XFs7ujjUs7 zxk+JvyG0H_;-PcVsE@vT8gG;dx6UbZdkV>db*xf)&7x9zouf_Ht;xXJ2!~caF%RqS zJvr4IX+X+9)&A%JshB~o8h!X5Q!+?no}!{+XIz_I9KIKFy}61m=NQi5!ZkCxl<}d( z5zR`i@4=8;!crQah=cfgAE%!GVCIv#+4K!zqq2M1wa6R7CS|koIH}>B6Mrl&@VV@^ zHn>^26TD6wtywt~dgiUv#g8CwT4MjKS zv&eq#@->{xV2{vSM#V<0C?nV=p3As+V+`_fN3s6JOYc`Ln@p*&9S2m*lnQHM|AdOc zh=X!>9k)BWz4`{sT=s1e!_$T7w7Pnb!_G+ppl0s`Nl!S?($R}cLpRUL)lt}~PLd?= z?D^UAANl5B1$z$kf7XlJr1(1S({{PV`j>`)M&A&@&h8qPhKyh|DFnOm9Va}(&>DkR$fU!Y3PP7ihDA9)M2Ajr#yRc zqAPq0-d5%FlBFBHH!P@h+T)WKfn%v|1@whC(2s#~98LhaoQ?C|w=wq2g7Q^2zFO1w zRnxYq&u1m=coj3#N`*~z`y~k+WWE70`Y zX+2j}7>r6t>2#5(o(s7qSg=&ZEG$Ay46UP&hN$q3RRKFERf_ehDYFrcCfYbaIjeh}*IP9Du0PQ{wo<9FImhKHFg`n|J^{Q(Y zqRKpi#@Px^yVaTdu4%4;Ns`iz*D5oai8ujz)%`#UjfshSL9KyU{j6^ZSIaG(eG4b! zm5*7BqF0V&-3RXf?R(BH25a-{R`MQ6e9e|5V9y-0BxO=EHAPaXs3b*8U^^^BWpjsJ zL6otb-AC3rjXLBgK3lYpk2YYJTRS(_KZow7wRC~tq2js%EAgfP78WDH$WB;y*2f#o zygixmsc+#wmR8NE!>B{_qJ}x2%sDUbaLb4F3Ansr--Ahyj9{#qLEmKx!18>e7@_9* zTed%_n$zfG>eAY64%>|Ty2s^6M43x(~lQ+Eog>@xWQ07FLT@!ZfpFEfdMlpjd{_=hq3Fv z9rZPX2Qt0?{kx>rj`uqqM$_Fy$)I)eB1Yc&AsCBcGcEm4*dN*mdllf^s}_w&u7@W$ zqMl}jKi0&dBs4T`gZkMWSA7o?gi(OQ*=66Xt4l7C)L@8iu{O&xg4^j+j#?4$Y4TM= zWV5WDiz*X2;He-liT>9uPhqXo^@TUeRr@7}7UX>l>JS?bk@Xd|V-Y)@(gEm`%NM4J zROsiVh;qUk10ROM>v4jI5+oF83&KbrYC2jgNo515S&F8YK9~N3pOpS%F+Nv1H3sgS zdnvFsG?fPiMW3&<)-=ELvo^hH8755Mvi*6&<-G5_XD}#{Rf~g3n;kDyqedp;0VTUf zkvkM{5_Pdx!u%$1P713fNPmUviLQJTwcs|(H2e;_mpo?UGWA4UDZ{ESSWu#Qs{<$Gut90s+9Es5QrO4_?sX{Gp5p4@T=m(qc<9^uvw%TJ5 zIHfPTS2`8DT@apfJVj2>_k_i)Q;rZdM!i^Eq>^WjI-mjT5sye9vzaU{6c)H90kfx$ zXc|Mv!>+wtNEW%&Dli$L!h<}gcId4uFg8k-p=e?Y7W21Wpp)4m<`OW6YsyFR`9zW( zFI8q(M+*IlM5jo;5IU8Rbg4pK=7r++c{Dbnl!kQEm|?QqV|nrMPV|TwTqgQkn7?Np zE_F-Aqo!0?cDwU7!_lR8WbcU^q4u;X5_C#qe6^kh0#v5%q5oDS((6!y`s?O!lvGp8 z^X9h_Ej>JQ9fkkzmz45X@}K|$pku3@6f$PVb!1{#_a(fV-A`}DJ$#JKM+3|XVx28;a)dGtvM7Di;U zH`})0^TmuZhtrF9ZF%&ztJ~KX8l0OyRw`>qksUiXXN{)3f|4DfNIC9*)N)){aM#=6 zHn0em7nG+c;I(tHQw<%K2L4%zZg-43q8!LgeyTV+8_QOJ8gN#b7jA6|cq+&{hHRFV z@~^Q5`GfqG%9ToVK;##}RYD;QSg``SfL72dCWkp7#<4Lbm#hI4X7;6I*3J&omKD2Q zjLbBl+~I5u#72{{xvD)b*K|>sYz3e}mu?O&cgK#$ByXKscCs{|H^6Oh%>zWNbJBH1 z)UmJd(Q#4BZ?T>@1JAHqrEO|rtOFQpK*V?qMu3@qY)ku>`38HlKK16mlg)N)Z=gwO zOpb9EB}<{mb}H)pJGa=4VSmh1)bXpypY&(V%4BISy%M%RsD;x>H-_!!W*uv>~~Q zOg0wHb?`GFKHN*UPpyh96KGf;0Om(z293`??0f%p7Z=W>E1Y0yK}H~yF;b-lUj$ug z+zJo7+ziGB>!~pohB;kXAE!H8UI-hPoX?=X($BvC2fH~!H^yh z2W7c^;f?Y`V#q<=5U!VYNZaWQ5F&4rwtF9TNeq1qp3G%ZC)uiFzaV=T@!;J6C9cNtd^ddw<0gO!^pr)JN;W$oPz;JMW6llLlVs_b7aT0a!?#K#*;~) zWSc3n0d(4=wV_b3YbcO|IIJ2fFDC|pm&}3%w=4pX1>;!z5&rO*MZ^G!5{HcQWUU>W zmK>u=-cHGqD3XZV<9_JMKyrpkr>%kQLD>pj0|yq^>g(jJw4=53v$trj4Q9d8F|F|& zBV#fB@Hck9_bmf5ZZE32Pd=JKx{W5_3MIQlkxy^}lBHVdX>v?~Qo*Wg{A+xuOyt&! zhy3^Pu}~FR4GILv&Xxr? z-^fK*Ld)%zslw616o`w!2QbUW#1w zN1)6fk>474HoT6X#n#K6=h12fmdgy02EVqTl^ir&b)v2*={%Yc)GdVTWF9MKDnp1x$O6_-b=Q>^8^@{uM6*Tfq;d*l3MZC zzZ6AIffF#HPj(9Udl(rJ`%RF1%=+?wnlhv^F(P*C&6?p(ZRcWne^KcA@C@fqgKAhg z^d=`fIxO75Is#HZ5VXC--N(PgP38R+!uo2(Ux!G6pcvetQ{;+BEyo8wga&IgLIr|N zuK%Ma&B=h;gcx4zytAT=nNYHT;@Ot3lrQ~?PwX4;(uLA0-w##3;>cSD-^S-ZE8inK z?AYrqH*&i6P%?0XcT-V??rDCz#GjG}bQ<(9ZjoDKAf@@`9J6p%Mpl8~x3v3Ov(ifl63|R;iMq^szVe>#~Sg`!;WEzL;cTpvI zMsV3l$sjqg6{-#*JEToIS>iZWhJUJehTDMqEnzn|OQGTR(4AgSeeled=EQVHEkuoe zF)@AFhUa08PoBQ+l&D=mS5*Wv9!pACH~DQr3I6*z__Hf)DrZxL+q7D?gi@f(9K^Vj(wTC7(ZKIT`->`qF(E=WkX9q&$JjVPNQQL+JwJfNbQA}%8bm%N_cM?YEI%&JJ;@O`{{{-q-(zA-q!8ms7va?@n)Q$1z%P{)mknAN| z-5`&aHb*`ibi_eF9RBW=y!-C>$F`FZ2jt0ocD%#=YQVGOJ~HZGCn*oWXR_FJT)e)K zbh2XoGv{=3Kb8;8vRlF?M%IYBvhD93rW$^ByGof67tEZK$GXC=hji~p@8tR+m3t4J zIVaxxBM_2N%iE;86$R`;GBP)beTrZ$Z60(%pOI&9@O}ZvqL=-uCNdvLR10~#Wk-E8 z*$<>~fy3;3zLidS^zpe@AT`k`-SzIUbH0!Dd?5=h(aQrPFYa2+GBHEX;tS`0_K%fb zZV{dIM8?5vMGXrpB-@?ZrzB?oxe2xeK%aYQ3NHMm<1BgIm1eMKG|9Vno==@66aT_6yf}J3fQ=3qS&r#n z?CqRuN}Fr9>y8;BTr4ocIO|=%#96xXY8}5nR2-!Sw*?L^+zD-c-9d4z%8=H;DeM1Y zj1C*2!E{(g=$Mw-@ym~YY4C>^t<)|ct#<4WKQ@Y=^-!`;DJYwWis2!#4)E3%2-4{_ z-Yp^7f!R=^hJ*>r^A%X3e?-0qD%sXUp_-mjaQSk5XV`Jiy&(mHUPmNhSnYh9o9Fx( zs6>~eHW8{DmrHib@P*|iz8}jmXWT<0qk^vbLWK2}wW6EB%k?x1nEye|Mzwrzi0Y=> zE;p>L%bUGgoF*S;r^%OZnAz!-quAx9BQUI{TcpFzz*Z{W;j;V?EOa|S+f&a$w_Q-; zmPmT|$%}4=80ly+#@h>nWQ(Kcuw#a|pOk;M;_C*y`NuB!F*(JoOwNwGRy&Q*d5Mxi zwPqa^g`9*fB24^k^SB918t%xU-V&(9fJO{411dQCC0cq!GUPnC@DxG3rqhq1HdEi8 z+z6b!yA@d9(+IUb_r$uo9;~^nrW3-psS2XXImny|K8!q8s_r%H4>%v-(c<;{`>7Gn644=d7RR=A#JKufXy`jPI zEddCZeCQ*>>DQIop)t~%hAfv3HxRuMfi^pKuxBfvl^tot`Uq5mfsC>B_Dth_*pry} z@rJY7HAePl`s3gJW5v=+8MF`umkB$ic>Z_Qy6?tKQR9$q z8ks@EYBT@Jffzx9qW&=V&!&k-yLFV!q!RPKc|%wkczDiXuXL|-1PG)e%AJ~({h@Uc zdtJ*UBM@#$n^iHvI{jCE+~YpL2pi*$V4I7~u(9aV`}!D9pF~$>IY}cU;S5l}@TSP9qR7zqdQ$J!qKT z7oz2=<;Uk@2lXY7LJw7iPYq{>N0zJVsZSPpz)2B-T*PMYRRLIa+#}AR*9IqfgU22V z59L58`zFnqgnd7!mjCE)7r$NOmy>E!3@wm&V zS%P)j%MX%%DblFlq>IBr+8O3#5X3q_s;%68*d>9ZZxVu*b>LCgaC>Pe0MDZfdHRa* zU2_hwHGZx90-(Fi06w|A`KSFJ_-p)j%hKu7e#6pz@QbS1XFXvOdIEOK9Ukd1y9NWiMOvym0x{!*3I^W<5|24raA`AhkFgf{5&Afnzv7xm<*^t1&N3JgE z5vvmBWi0|V8#+UP3HB9vg8gj0R$w$w!`G{zw@8n>42f1)4J?LW|!#1kikaVAY{Nr<5`JGa9YSeNJCO(%sZx!n7(Ei1F z^EK>f9tIl=1ZaCK_iW70kJV6O><*dzeU{!ZzqWqn_ub>Kt^>AX{Ftc&Mv6r29>CfM z^x@E6{QuM>YZ)jkol5IHXC4%boIB3E?u-qYl$Hj`Q4rG`cF};BVURo5kvL=lUFH;5 zgj)sz{ny@lQ~c_Y%!6Q@a?k0idasYAhSg?!qW=}L&5l!6WkxGYE+vEd&kXQ6zI_AQ z=3->;?VLbS5x&{y=+!&Om24lnYs6!#$!^iA3{#0G?$Ye)*^n;M$ zWL(9`%$DiweeQhxiLmG@quH|G)xCT0|NI}*l)>k_l6PUsp z66HEpd8+ChHFTT*dTE<*n-7W#Ob(Y8fn>}>jBq^WJzF1|$MN0zmG7H6u1wlw?Rc|n z#-%;w7Y8|Pq+UkS%p9nvNS5AmyX6L@3Rc*- zdNIbGlcgIZovbXE7SSPSh1Jrx;Oaf#hX=P81IeoBnZO9YmUl~h2WU|19Faq$*W)5Y zk8z8Q^*D8oYWX@S0BomYcy*GJ2&BnGdT+I&i`T{LR&0j-cAaGZV*QpJxl7_$SXVJg z-ZRT_^tr|Hl@}%hrmw$yl$vXBUj~0JUPE>>i^SRSid1dng_Te;VD=~i9uyj6>_ zARmk6nw3|`J>NolWKjxlyI{lOd-U2Fc-9s6M7Sy-Lr}i-_or#Cq(+p;xlCUHmZ*x^ zn`SnOD`qG2?)l=Y*vf$~teerSyeY)KfweR6`nnmX;ix>KmUYH&-3)x{p6?pxW@Q<_ zo!bX@&d^VFZ6majBH6Qbfbjl~ke6r~}Gntz(tar@TO0Y93ISHpA3S{k2?7~j1W9MzpZYzn_=Z@{8jmKveS+uG$4aDCL)ke$zZ#bMMWX` zdkhbo-(aahZJHKQ?#%X?%|NAw0_Ab6fw|jU>gYy!4k)YDLAh|hbQdg0;K5KM4%rGU z>cvV08=_@0SYcwln4*UfR;CqZDH=^l;F);K_G>PQ8Or#bBvTL%JujV-0a340rr_o* zwS3#0Zhkf0tlZBo@QM}o(fCXS_Y=Vg!~-)0L!??#${q+UVWoQG6D6#2rv|!dR+%I> zBtOJP5StvlZ~H%p4=aQR}&Nl03RLbi@HA1J23%=2kk@bNl9EQfkDZ5xm)c zb|GD?<8r_2hyH7<`_(BiC=&v&O<}7aoDq1ZeG;&u&ylPz|1{)!$9Y(kYULt*rghUZSCI2=kStp_)L{ z{l$OO@&XXlgy?%7L`7;?*}#m7B@S7B*t&3PMbrM=`+0sa{#;q{Dx=k8`tRwF zfAYHTZP5yV7cp3T6elF8jTSqsGyWp0@G>(yTyc=`hhrBza&r5{q|>~PM|ti@1lxh#d{}3XaWmB zShXQ=B?s!}R)j(hqYiPb1Wq-qMlI4lAzrDPvnND#oa_q6)N!Zu zmN?$&98nd!b;^6hMZsq!c)wfHDO2UMdZgz_r)wMZ`;<${CTEA3+m`3cj|B%7L1x^C z7~yF8@BcXCH>Oo?c8eo36>WunYWef7N0jG9&Y&1xl|Q6#K9eK|*LbA6PbrrwGiZ!E zi#EB)2pZF#{L?+RuNhp|>HY8DCAG{7QS3NH(_yqoHBmCC557o6#Q~9!hDQ0-CQw=` zVs%PyFL(s8`7Q5sD^~H2P!9r*Pz64RSZ7!j{>6^D?A;cza==;cVjetCddrqsn)arRO0l zkWIG);``8X%humn^HpZcB{LLSV3W*0!U(N|*CKq~<7-#Qkn;nW>48OgAm6(;BvbK} zGZX|YmaGRtR7}ewdqakTZbKTck%cM}^^U{RktyMFpU|u`n!kxX%*ggooFDD~$dtYA z4T-Or;krZqrxsD{yww566njHjL>r-?5=-Kc0P!)fs%Eg_SXfz*Jd>FfnAmLT-(rZR z$^0rt@1B=c-7uwW`Fv@<9j`xTXk60#DkIY&`9DOOmAAy5P+QmFc~ZE4@mXnOP=?c8 z?mZy?I^Z^S^Kap1>Dju#Xbva$uz70|Q?I@nVDMUQefv%c$)7<^8!h)oC|Nm0N~x&5 zuGvd61=Zw_TG2N4jRl>O4B?-iU29aH_~MSAXS}bvx-@QS^p~1IsvUWk2A)4zO80S! zK!*|6`)Zi4KF-bXgopuAjXNZ8dE3}~T{9MBFUb&|m*0_EZN?aWA)if**`Je}gcom{ zpN`k6((%sOpEmx^fSubB@$<-SI}X;aHDZbXjFJsfq#s+_)bh2#xxtlADB1_@8F-?n zQHA~QDbfZ7CQ07NC-X1}yEstUxtdPpT_$m?W_MI8(t6hVxAJSe5;#3#CwCt6;Bw z6qYxn316xF*mKk&)<1{FD{I~J6l-~fA*uhDy?23YDm(Xv-Qo_(!jK!mWE)TQIJTsvjojwL>_uX#Sg_6F-hWf3>Tv0LP1SniFRZ zvB+#RK}rbLj@=!3f_IXM_0ZJ@VMSYXis7(;#=U2(Gh7VAnGlJYmMFV$9yH-G+Xzxe zI>kW!52@m3YdWCJZAWN>yjh0AuMW>Va-ea2ZW$eL#BN+4|BK4iHZc5#c*!hbWb}sH zA?c_}oadkG*X(~Mz(^Yd95%ml7%jWd!l6G$H9hekXa8hN!hSqVfx7wU9pgTS&8 zatM)~g0i~k?W8EPHC6O-_Ztw++wJo-XqgxI5va@k{1Lr6Ho!QwTeaKG;C3ux*;~1g za@Oyz>rrer)mI;IX@VN*i6Pxko6iyUW`vj+8Z8bmBJ8ux5U%{VkJ>hP5(|P#Try6P z6;6yGDDpaJYjPvSBvWK972V~Eo<@e?I)AaKU5UO(DX2Hq%ZkLfo3!5ZimxNa+y0^9 zfDHRPbwEpQ$hfvSFy(cxd(>b`WBzz~2pW%5!E1bV-Ex#P-AgYI!L?JfEEntbRl1c5 zn`MS@gFx5is@HTX^_mCndNrDRz{`5J{6wegL%_YM9iF2gb76`&FK)B zaX`<|He`sa4gk+MZs=LF@vihgEOteitrU^FPHa~atXAR!6!VxOUr^B*yaFC3!#|cB z@XQu1h5A=`MX%oBdUGbGccg%9>s9|UzqS64WT~1YK}*O15d6XFS_jl8-PcbllT5^mZlgPxaEd{wsJFfP$XOyBE}`-SEA# zZ=9TmrPz|*!L#*W&;B}|9gD!T8}lx_Z@znKYG**BcI(d?esfNHjNSu%wkj(FYlF6X zEaqX$-9LTub!?d)(4;L<^|=pj-=tjvp7@FhRe_!IOcFQY4E;=rel!p`1Sq*&G69$q z0;%_Ae)tF5G<>Js7-vbWANWcpovXo$C5uCDDGhGx-1^+0PMe}r`MIov^m~Bdua7QQ zR27mhy3DKA)OyBI>nFCj6;3LJhq5(iLBcG3;!?TZBV}S7)g9TPX@`K!nZP*5$)SE} zSsVnKFZ*-U0VmLm|LQOL{$R_dcrl3M#N=-lh6M(>OU0mHx7T!FlDf%eC+w}+r*X`ZRO^xzaYo;Z@>8b_*puWPe4Snno9iEQB*`}D&dFhm@YGg|4aFZ@et{#JHD zXu-$)3mXrURow95#71SO6+Y4^W+O$Csc0;Bg}dG;k&0D}OyS@xm}AlHi5_-c~G91x=N)P2ZIs21kEHTshfXOx&- z)fZlQM_=GapAgxmA33eds<#^VB#KF($SNwj#wRtXn72xXI{wYF7ICugXa;6qG6MDt z@=G2cwOi~S#t`wj@eSh#5^Z&%oE8!tR*?9dV$M_KEESER`#9J85#6vttq%kGx1^g! zW#vw~1?G<3{7byMpcDo>&$ISKJ$1og%3WeUPU6*ru-G2gy9AHS*YHoMJ1xR?wNT^{ zST>q|ca-afoo`qvz4NfP}a7Kpk^+?2-b!4{Y0?aSNR4@FMc8h7;lIe#B-nUt2U2(BdqE}S}WVvl936%yA_ zOgu$aQqg8_1k##dC4||IrQsc8=bH=~M*HwH=z%dhJYMwPU;eiR5&GMrC*<4&pri~A zx*I8`jUrd5Xbi1lAl0nkhbdv`=%pyH&)(!!uc&}h9H_z!2_Seeo8I8PJgi<}uCmdg zxGcr09Qe>Bco8$gPO0&YJp=oAI_u}n&00WUE>_eXSt%cI=_OdcL3d7mm)=Pm+%8Mo zXmg`f<`^WK6nUF_td^~SGFV7}jNb<>#W%I;~{H$X3c zZAWmIb+|dbpU%rQk&vIrZ1s1Yc2)y_cF?ZVI*M6Ck$5T^=M_jUosz&PuC$x^g@#`~ZY$X|3vW6wG zethR$U;b;%5uw&HB^kq}q>=q^%*y$5E5mYcp`g=06*?x4*sdo14Cs z{*%S;7yhzIoA>Q2(#6y@<;M>cpKH%)o3toM_c*H4y9xGx@bA0fhG~gW`H@Z9E6Nhy zvnx5g-kJH4N69*}ieQWU$ekX0)R_D9RFUgxZr2gMHSkhUqv>ub-J!r1>=WcrZ^EE! zcD@wDo+1Z*6sN=Yde)o@wV)}}H){ha8mi)z^G=-PGb_wgQOtgd?4hDj5j9VO`66d&w&!-4l_lDUH}gx_}@D zq5N!3vZga|7rElwEGr_n#&B;C$Jah!nC6c$fWz~T-~^0u|6F(IgV&v1ha1G0$7m_4 zWER%0jVKInqi=g+#nv%i@o+%K=n-$91LDggPRJPl&#J?;7a*cay8qdvq)*Ajja(dvpP$v1Raj zBG%7&A}CgLMjnGe>%EB$aNQxZ^_Txtn*%N*V(anQbmh=fP9x&HE1>l3A6|dmiA+Y} z4*DKlDQTZI07N(p3SO3MlU-A`0b+aM40!+i^Q1$I_c1Xu&b2QzF=##Ysis47DX2EE zjdTlg-o;*^~g62W!b|{juGj3?E`$J%^?aad^i^PeSKrAfq zde`kz-1E`xou3ay&~*E-0Ki98bT++rj$XZIN@+yGB*!rS?HgK-II^!(2eis;yfR;A znR&$b#D5}5PRz`OO5cNGuNf2r8#`%KbQ@jaRUT-7$(V>(nXDFvQpS zkQD+AFqa6o4P!(%R2Rl*7nWTpQ?=>UOmu{8PcG zj_eCV!eOXicy=@|u-l3uf@$M^Av{<4zQumLd${MPu3qcptBK(_1|U z{BKD6CLJXxrcq6wAhkj0%rt8AMJpy;5@)*_CO2#Ees%vp_kAsH!hq$8jc*+ftfp&t zeUr9&*2}S72D3iGy7k?!9{%dxdAomh@ZH<6yZnpz zKRyFjVXZg0HG#xgWqD~y!Kykb>X@|T{@kzVkSEOB1~9Q*}%wO#@2 z8tsB=I*;E0oVIN5r!E!Vmx8)UHK_w_=FZS2t?l({ayzt!cRTd@>?SS#bB;{C?tX?o z2{jwdz0`qt5qo-K91Db%Z%&)V#pCe$qx zB>Q3_wk{7efO4i6t4>R^>1CkgI+kH#j>gJv=CM8HdXOvDcTPTu>r6{WC#gp5m zhLA4r(uhalIrI^ME>@IEAXt|NRmN{CtN3kn4k&<}6=y;Y{#eT)!^FCu|iG36L2bS?vb zs6o9o@WISlfx*ov1TIu~nc|3Nwxg6ghXuBmdeHV5mRGm=e$uM{=XKv{0=0O^JUc2d z2zq%&St1BFG|Ea;XXv$&?I3X2th_I*17{_}wS?ECh57^ZvH4e|#j0&G49uW2T`#Pd zQ>N&XmIhV?HfsBSc228T*UNK8BfZrzFnIN_a&5sr0}WwwDB0Az%f4c_<-SPiVl(C@Mur)AIP#Uo>AP4rQ23)Y8hooBtI#{$W9=zvl%_F*I?yA{1 z^__VSBosG;=u9VU@MqCI!d0_J=tnySbVC~>u0R{yexPmfC%?DA=%Y1pZ<8BN?A$E3 z3K{fK%ma#aV~M>gx?*+{QzSO4Dm1EfNXcxYF}np*A$7P(SpjZuMwsr2*Y3dO?x)n% zn2dDZ`?l9npAJQmD|kU&u8=|vo8=9&uTMwch-nbk%KG`cA>F{#$xe?mG(MLzWvLu! zp4hkwat~redUcg=LS(_#0Ma8?trG7m~U)$i=*(rOpiAsU@-o} zkiSm;$5P=$K3Bi{2}sf^r41pbe7I!aD#z$o4SuO#eQcjOdUZdJZt>(hpS8PM%+ay0 zRR4`EV{p79D2*`6Qa}wJRGVeTrL|DZZIkSFC~CN3F7*g~Q+oz;SST3-L%|6p;{;bey=rSm zUaN2~OL(tD^%P1bWeCp8_AB;IYSOj|H@K&I_|&D&3;$dyj}gTSp1S;f z2A#J`V&R+wJBY0~x3Bv%Z%+$42g_;fXp zsj5H?1a&o8oQ+|O;ozCV0mnlRH~Sb46)qvpIvIEwcb&G)Y~#Whc4Cj0C4;|Su_Qo0 zX@v&U_M6CZcT@zyve+#68+!B|bJadv^wM117h-2lezPox&J#gEA%|YAT%~B| z=`gQwAn&N1)`Y`+(m7s`19qaS|j=Vzx{<|a+T%7bL<8?!iubm~D< zY$3(uQ6vW>tRgUpAYF1qP~)@B*C;~|88oDPOp=G&eCxd0=#qdWqHFLphHvv-AE4_D zt&!~j3EoF2HRYpMchNOIDazfvG+7@=-8o1q+y-4-k|tSaL)$(balSproB5LkFMEFJ zAt$$;IJUdeDz^KGV)`iZfQrUtayoq${M)sW5E`xU%JamkB~KL?;Wg>Nw33ay#Ud1# z#~Vup=1NE1s?~B`eqeztR+KA+2agbQ{jG#a*S+;nT1kiUF?S|E;_scaHpB=vbeq-$ zDy)xXehii7%SpK+oj&K6C#m5jOv(>DBGEO-^C4o}4wbO6I7+q#Ws!gdM(g@K)3j+q zgBuDiJtQ}0EPJaON}M*xQ#C?)#NSycyx$aee*9@gs3;oYxK$7RRPR-c84XN(f3|A!@7W4E1HvhEY)cex-G= zZzg_aHx6$QvFhGZg~cSLPx-_DB`MsdFeheWm0C?<*%Si--*nuZ>J+b^_&~aH;##tU z4+?DIt0ChB<>H&PSEZQf(#yN-zcw;w`idzPisWf^2S&jPQ)MPi!AfhM2V+ zjtuJ#Z*QvquC4K4dFzaH!laleow<-b?CDi_CHv+{Y87LdbZGhV_WqjI|JKnzXFjT*`a{b+ zqg);ET~g=7u5YK+%yETcE>Yxj$hK8qp1s$k&{^ZBGq~ZVKt8<}drE;?} zhwdf1%d=a+RV@V}MC3c%o*egLC@<$>I!agXS_}?Vni4g{?=|RiuEfkZAwlKO` z_Jsm@DzJLNf`3aqIzzF%F`Obl0cFZeTa~zO!Glq?BwvosFyQg!v!i*y2_EAwPTkWO zXo1J{^iRO|dt)5kBUWIlpcwFccT>@uK@u^8y8F$Ck%xqB!d^(q?Ta`va~tW0oNiRR zyB%C_n!#d3y>R_#IBc5;OH>)E8ui+c&&AK4OrDc3NcZ2xFNo;p_w#Sg>w zDC@EHz!G=!) zlM&WN$BNP=-KtJ{(V|61KxQ{Z*#Wt2#_&apQj|vn%Y4jy%Z#x8Z{*NxNgP!vJ2JH@ zB;KnkWIc3M=HnO*8S9a>14u0X#o)`oTWpL=fJ-b{#?8h!@qD!1%EqKn%sPs! z!KE@fr3KVUQW(D4qkt-)a0AgCbRMgf@$=W+ZWKL#WK52J_s73pdFvT2B$`T+%?%e$ zJog;7!bLg7fGX84D*8ZB{6r*eE779p)$tRnAzdg}s@q9Fs|bvGWjXYw0FY>C@Jte* ztovTNK-S<_;T7+82C~?CDum_{gi>94^@!8XY)0K~+zg&`*p6-b z9?9vSZoXA*%Xi|kz;R;Su#{3k%VEw|OQ#o$FiA0C((b^5d0QQo?95>}>_i2J9v##4 zD;fU3vYlmE%rp9U)j_#_$#d{6ObE-Nw@gZyp~HZ9H?S>6!=Zc0ztPSg6kn{D^y2ti zoF;kaoul{v?Rt!@lq!p2X)?bNqT71kdYKM+L5UJbpa9!i$nOjX!X)y3ASKiY8zH_>wAJn-mZpt05~N+o0fkIM6hPqR{ z!TSh*Yw%_NB7PP*NiwH((gm*VibwC;r_Oqp5rtS&TBWs zRQ8nDUAujYI?sPQ-$-wRx$zLG4gwyKB1W{GZ>kIb^5Wo}hTP7^91v?~kB(`^39;jU z|GR)0ue*rXO64KH-kIHg%`#LYg*9xFz(}XbPKfGV&(MA3jvUQeiV}HcxX-IY?X1#> z3vk*A`LCD*t7`(Es96T3c~!djsHNU3!gdfHw%4n(G?=KhmgrF9G$YJJjm-vnBUH7T zI^cq(1e#?_gDXG=7q5VZatdU2J%ZW-xMvvecGyHm0}gJJdUF(}{pU&t!kcBtzy;~K z&7#|r4W65$p0~s!g8*-z);yLr%hKc!iNt43$%}l@z^o;iKQ2IvnBa?<71^|XK>@%pwG#Mp=ujB@7xP@j0RJvcTfz( zSTd;S(}DTlYng`@s!6+x|JXB!x0Ki7)<epjY3Y*se(WLF2qO;qG^@NS6eo z?&;b{6z}MyQQ~Km9kbyBg=K^tbeSC}VO^DHY=xY-MCF_~`obdnl&vY|UEpCpb4U15 z*Cas>eRJl3OIkoWeJlWHqDSwXoYvz9dF&H)8$9aib|pq`alr z67zSEVu1QO`Y{d5Z~@ZjIwaP>Z1CGoMnX?=Vra0?lTfp`mABJ>CCI24fl^c&f!0Nr zBiY5rB~Cg-`*pdzdeQ}Y{l{tDolq2^K#F9FvO(UeTq(z$a*Xrc6U2xP%&BmDs=$Jc zU38m=t|J`#awN5a6K-rB+ZWEQ$9^PdColZR3|pVXX=fW2FQW=7nIRiiuP!9*WSj5a zIk28dQRXrmU^R`GK>j2}30uZ6+=;vzA5bU8=?R&R1*7d}41e55GODkBmfOTL>D$*l zY%{l<7A5Ic%bB$llSq-(R5Z%x<6FHW43Es4f62aM>V;T_#V$;|dI%lff*U4Ul8aky zsY*_Z2^J0&hBmM$Z<1@XxZgbkC{$+RbCVj(hxC~aEqWUViIILF?V#h_962kn4xwET zv7HuJdvn|kP7oRQqko@sLS^w+8-G=?kfd_+SDiR!0_C3vrJd(e3^=gcLD)c8EknK9 zbtILEBbOyR>4eauqIN}GP`l!?Bv+F*=}-{d3%c(G;->tb(7m(~YR8}kTt-m6>l)u0 zpEOMen_Q_whH+TPdWJNIe1oAoNeHX>rw|KPR;G0CCWYLv;=~T@Nh_=vCu`^{B%mq;Bo6L=r9Q+sQ?EYE@8NgSy`p%iBy&M=m+q4toHpR%z`WxT;pZ8S zUif=xOmMl%yZ)P}!4~}7`2Ot@lFJQ0PHbzAS>b0d#Z*wFl#1?s`xsp^XPw6~uQVvX z+Q>urv@4HA_Je9pVo0+%mtRdbd+&}c2fNct&eG|Ao!)i)wcbxan7PmW&XiKG!YKx~ zO8yr24c>)QnzX$nR&b22lP~jHt?c1ntmkZlD8fal!zT1N^Zp1P#* zGF11WvQRZ77Cv>sYmdl!SDeY7x?B;~1?}RuD;|TqcaoqYvQNC;>#0j6@0Jn;x~u7W zAxeG2!d$V>H$A`zh&(N=59y}Uc$cN+WJ&M_?{d#a@7|qPCow_~m|$f7MyZoCm{l{% ziNWneP&@C50DH?0*aqJX{aW%{!=meL$82&p+9)&`#nWZ#D2d%@k+;x6u0Iumfpt@(C1gp*Di?%U|k_=uB zd|V6U%Pb+=U|+l{q=UZamQ7d7QYOC4p#9I`1~By!=x}qpdqQ43h|85mq$s<<;fAEfK9T`~ELC!&3~w&HuP{zE3x8eKBnj}b zHjf6ruAk1DTca@-p57?AEm$9-ml*-_$fbyZs$~swTq^b`HitnUFGH5?;EpNMY(M-Z z=j=#ZF|+l@zpLEx@Xvt9j|tSdX=$B5~2A9!TfM zCr}-+TbdFU=Xou(#}B zgrDhd6h4fsrn_lu4Jm~19j&1{reJ*8>}g{JL(?yoxmzH&_Z#0mLsmQSbP0kTgCYW( zDP|J|ziafasK@RxBHb3@E%zhj{;YIKF|V0H$?lHuCfMx8UD33Fdvp(*ZT`=#O^5X}J9pIEVY13ra z#dTh2K*kCV&xXftSUh`RjQ-{X3+mhd{{2o{k)Ri8drrLEXORK|?QGQzX+`82`W(n7 zY?vM|z?w<*P?lP+K1pgNu;`a$N**fjsA{3~^kq@}^d@b#21{1s)lPAj;s`k;tO~?5 z$x%a=-4J4#1x5=|>=0VO{JHgCEOU%VwEhR=!cfk#6Gs=JQ1GC6r-NeJDbh+sr}47s zD)CcAo}}LceTH;;v#2j(iEKOnVWciefE?pH@)mI>L{z6t0P7zO0n7lT=LIh zg@`qRS5Y8T6;J7Kj$16slRP`e&qs>sv-6mVz;QNL2gXXn$Oh|&!q!kf!Y~zC(*rKJ z2pojlks^%+fWy4}9JlO9JN)q159Ai4efqukE|Yam?9J@9Le(~kfnC)tRCJ5DO_-w0 z^xUZe${xzVWC7jNL@MtHt#>U6yaw(Jax5ym+Lbv05Vd}$v^)ANonZsw`Phxh5j!A; zaB9Z=_Lnz5m}ddW`?WvbO%6CQNG@7|?~0q10H&xcy3F-}5R(;-NZXa|yhA}Nz3)e0=5|9!N${t#0hbKH z-Z`IYjs@Ha#TXGJ_xr?(+R1=Rg;xphGH(?RQyuOrpbR=x58Up7JG$-4>(T`6O8Fp* z!EJPh4JSJ-2)A$ZyfrsC{r<;GYo}Y_^x;SC2gpt*hSO0iIPIaBa*C9|Oa^q}t*SO* zy|9vRj>nXS1BvH6)HXt-ZCBMmRT0FQZkGZBM1|pd)kWemzkW|$0sq8oWD6!j5@)h+ zC+x26os%{FVL-duq?xkTqskYJm|hLq@2EiaOI6>n7?y|M{MN6?AtyF0uyHZSMLt6@brkuGitd#6gENckp)3vVL74m)OlyOr zHp5ih(@2x!pSm>JvY?o#j&Qh+#oDgVz9Yx>SeOkQoOa-o>eXxK-<|iwt2ZJK{NFrD zvM+8%Kx$lr96epEex60IQEsF@``(55c!o{0G^;&QCN6~k?c#qXMx}~&C8h%6XJU|! z1I)O>#Lh{IwVA!x;%0JwDG95zrIfw6ba&#l7z@qJEJ9MOs#oYr#g&@dp69(wfvu{; zm@cfGK;0KzqpA=y$!rDeN5EnzPvo!(iFLGw2^-e_9Cd>oHmO$@UbGbje?w*^7P;_d z*|N7%gLJq=#~k3Xt5#w&nD$`B=8-WuozxU^(AP2(6%72Of@C=HOjKnx6YZuLkip0U zDi8fYfg6%-LWlrIo}Q5@xi4Ej%P_4&*dKLH9y>9WTzBu5H)##i@*{J3AN~2S*JtF1 zWBNmrHiOr!Y!lvdFA|@VCx)yHN)0Ucf&5HOx_^hz@dm^m5U@@f&u=&!80-dPh($Ja zpy#OZZ3|@fPXF%}B;Sc418eF*&dYv^*+Y?XDmvaRSA|*JrT`4)L_K z&1{|4DK6msUAn~okjI+I%@DPNTwm1y5Fk>?QJ>4orGhqbxio`Uz{}R`pchmIc0&>K zTwb0iPqcMfmv^1KlV0HX$;t#pdr@y~R0f}MN3=apWMo#Kim=Q<|6G}$K}wx?4r;KP zgAP+nHAM_mG%}7a@{TGxGzK^Pd%z!LYe2)VM1@?^Y9OCq;3Wn%X>Z7Ld5@zyz00Is zsx5(^i&I_e9h|V1zyq7((Je-fP8#3*xQbaTD$ZpAxn~ z)#Nl_x?K0<(c#Q#qT~MF%yX1YOsH!zW%<3=w=L~Vvu3vtL=SKSyr(fQA=OL-*A2TLeAaUdW zw0ulfI;mnVR=^QX>7WmP$ss4uj=M53YvVsw^u!d=ZKn-IIk{kh#3bTqm{&&h6#@9^(E`?}}{eKSUD&HBQa+J~4iGag@lKX$|{m~pJbRy^|! ziH5MqYF-05)=#b}t4!G8!6FTR%%ve0R)&~P{n5w%*xs(2kT+-z5dS_yPhe(Y8qr4}>H%yBWHHO=7 zsO(4Am&Rr!2X;epY0bSwwn4Ku#OAQX{xH~nTv{u*!NaWRMBa9f|C!LAU4@bC9!Cc@ zoX>l7z}A^{+FUXg$2BDktLW_X*gL1p|HDP)^Dj#~Af>D)qDgxTT$85?jG&z-$LNhe zN*rx%<}XKvvkCa}$lJp?Kp2d1Qm zj+2W42LhV3N8oOuXS$?lQo}U+js8f-jU!@4`f+wY&yJYP_fD+wu{bMFn3F4s-iZV5 z2dvzWB8thU$W9En8$mlLM${W|4N61k>Rt636n?J}oc2zX9Gslskq$Z=rr6q5NYTK| zo%Y)w`nif$V1}RT}_2`wd~`fw~4cinW>*%Km>dBlwp+owgF+ zFHSX1i~trPZ(O@vmY$ZP!2jjQ^#Lp8^&Y3x8Ll^giguW9l=bjOF)MkfH<^nm8S44b zy2frw+W(Cb{MIth9DVjOk6xk3d8dx*&0mXFwq@LBMt>3)iDDms>Pybnj4spoFvd; z(IT)yi>Vs~@92u=-GG`*aHGQwN+6;3Pexe!MBwD3Oj(L@ZA2Bl+M`yJ?xOngDD3(uk|!DWK8_`6h*e206YO zP?wtvAcvgTM(Luj;%8Ek1$G~lPW*2Y|C%Mp=UKTtTPP+KHzlI8H9bnBNhS)haS~L_ zvPJQH82<)2?h&;?y+;_LEDZwi-q@)Ajs&3$5|#2)`je>Z-%{eUG%+75M|JJ#<+nDFe(lh4}P4$CX$h!h)xGUTb% zh!6LK9VtUBlg9PESM&FgYz;ZFlgv_i7*o_gs-mwcjU*W=*WZ~H=b0zD2FhUn1JPQ0 zF*IB(4EMQzX`B-<%VJZ0eRg&sDdy${J24>vvXTd7`qWSiWQ*;mqL&1(0wKa2$o1On z+fC=H9+IanxvHgMSgo;1iwe2Rwdr#VZmE6`$$_BwiLjF>s~2vZp;v#VDS}O;WfNAy zrq1#Sdi7Fy1DxkeH~S`$?SXoAn@6q+;(VGG?>bU1KQN^t7+XSRcew6|dCgu74QX-M zE)EwsgPvRcU*Bz(STK|++|WuAxnanOy_o_l4CyImGetImbgIxue<|<;^U;*0avg}O zfFIMt&y`kqc+n9X;M^<|;2P;VKbFKqnLHgVQ{1iub|tni5;w^6B#9xo>HIlJLoYI~xR98MdIuO! z-Wd+PTq4zq8@z+l9)vFH2))6>q#jVEg(nwUjP1P6omNRxf@{M(`Eai_e_ z3)`mgszKx|2^P@>yyNN&K~)Guu1zJp78wLvCfO+R4hL;)R%nFRt>7B?_3PtpV}4G% z`N$IAD^^v@G6ehm)9F&57)a~FA}ATW8bKz+Igs#lO{pJkij&oE%4sES{p9GLaq@ww z&a?CWE%ez;<(svE6glzU(q~qtrHW$qQ)CYa#`6-G^RiS;p=Y~dC%uYPsaHZl@qYT; zoYR`E!F#4%R~`~vmEHq&og>l&rc@m-03yn)R&nyQ>$5wRO^{F4Fz-)I+G=5ms$81O zFP1D3)eHAUZVkTd-%Xc(^@{Xw*OZlkyHzP+&w6vn+%+23!V%Eo2B0T@ik8{(`(LZ@ zOtBTHwo^<7MYdw}ZnMZpmx5v?#^(lHu7x(KtHe9#1V&fG!{vF5sA$r~xdSc_!I)fX7SvU`Ng`eCK4UEoMf0@|OolCO3b_i5FD|t?*M!F`&_t z5804V?6VR~PN%$g=6wb7qrQ;eocYw{K+y8AM){$jRf=|AdH{-kRSR-JNw=DQ-hQb7 zbyGVbh53qqt9+YG?^`lmuf}!`F+4USIBYZxc2R8lbj0aDUG(2+E7ruKf3#83O0I&j z2CE;nHLy*XCR@Y52W~+YeMQ=+UH;V;;3u8a7OSoSEwPrjB)Ak_?9uosU-)Ma=eK>K zlTo<9{-zEQD0AMIFRAd-k8Sq`oYt{r*$2QJGTf*(g!IE^QMTq0R0IJOtB{b35~Wok zMxkNqn4r}HG)|x$1#q%;$qlqS@4WZ3cP&%SV}9n(NtF|OwwJ6Xk<%0dCH;<5(VqpE z`W&MVc|0V^%9{{x+eUreMn23$Uk>~rs(*%#Ux+UI_qe~!MT>QNj7rCQzX zS-W|wCNK5A{jT}>8{aSaX~{SA>Qw6Lyp!|NWN}zfiF~T*&|rVqX21o@_{Dji1{#e4 zw7D8PGvGL9gOMl;>4T)HF#Fxa?FOBmt5xrsBt?}?zh;k zqF*admt|#b_H?n@J?tUj7ImAj5lV(+fhMUAuVNjn;f?!Ce)^Z);s%$O_2k%l>~Pr@ zHn)Sa*qSFEPfkMsoxSl8GprNuByF*>Oo_G0TK?XRgOr7_Qq7CwUU|h|3km zNKjqxhw^%95Si(u`<1X4@R@qEvfHnX=plT+(({0@iY^sHK@U~G$39JA_`-!))e-e? z(W5C*A@hbS=p2EkR{{j({GCZy`MC23%NKk((YOzt(9UfQ=}O~ zcG)Tun{&HIG4GoBaFh3T&o<$~>Dj=UX;)S#fQe~lgFlYap>TEwy~Xos5GwSyLS6yV zE8CU3-8wXl@)YHH?^Uw!UI7&Scd*_%KeaOp6rb*!dI`J9{iaiWsKjctM76mEP(d#F6igAZvoyTKzfNOl7cmV^i zLB2omw6slVt_QJx;s*B)`jR*+90%P{Oan(pxk$DVcUMtXzY@(khN zj?bE?`kU=Gfzw86ShfTpnSh{K0UjO*O=f_?3#$I)N_A_#dKvbNJLE;;(ukbtIJsd{ z-4^eo1XE$}!Cc4wZjW>v%+C)SC64)c)?T=wohz4@s%`7+n&O1fKM=3D|ZyVr5_zZo66R4={UElbj?cS5qnjV(jq{mTZ z{%aun;~KA8(Zs(ZJP+)alVrsdEPQ`Lv`N-Y$4*@8U#h)l+VEIC|FUVrqe+`NtxSC0 zJBJq|x}Zv@pGXaEu@iH6d47A{_re)hJPbV=)>j>Z8S5?`a>fla6Lv2Dcq9ZqCtlgJ z5co1c;k{M)1<2#>_E@9*oaqL_re3q(zebR2qE2Bk=o+7c(@;p}3`pgU26l!RXzp+t zqLWd$^iMzfBz~&J^VKA*JV>@q04nWZ2eXi3ATcP1icW#jn4hUJiVEB{UMIY{z)8c@ zGSflKrZ;p!1ODi8O%#X-@C0^R8LPf30xw>%LR@sU`t@UM#wbYok@#LlP)*m&n6Xs%JeDCh2_> z!cr!V3OLzd#SOGQ|fx5u8Q9i`jp)A-s76BF=sJBJ^HY+nf-tdr1Yi(eHUXVSnf4e)bD<9cA}sZ z3z|Vntoa&dzS_`h9WhSa5HpmCGhK7USF!4Y8)C}7J8`BhVqV-{aN;Zi78JEBSBO4W zH-UfyRz-yHiaEYkuIMtYlG>G>;kr(sVIzc)Qgt)1C1kUDv1q%r9n$`H5raV2Cu{R4 zRcvH*cg3rfKn?~KddIseuOEs&y!fNc*%ljjTD{7(6my6o2cIX!qdhmc zT@Wt|DvdBOoj^`0R<*c6{`u@CrbygI-}bx!HFPm!yqxJb)5@xo}E)$5Q5K3h`S0Yruv(h9@3;eMkA!FLOxd| zO!`dhc+2t%T-l56S9Ig>+mnUA_Ws0ZDW~R^G+|z3BQaV#pUq{b%%V{}}zJ|N7UTVipD9E=T? z?${x-?EbL{w8d`p{?PX%iG9sjyO8ZX$d;_5m^BoMr=o!?2~0@6<{+%CVhO4P>D9|Y z5opmEAz{!!*a?b3PmIkWJ1D9?$xsJAn@rL_Sxa_squx4kknD&RTq-CAiWTjqqDxd; zRX1lI5%lnnfMj7uxL&jCYpBFq5ovHg;JMGYAR<4~;D*HP8bPJ((hMwBlp%?k|8Ul0 z;Re3JZ8aQ~s*3pM@zw)DqC`>T3nYnL zs5n!@%Yog(CI&JBKt{Kce^;EMeC)Z|D~EpUsh=_6f~B8#dQ=4>W#SU43h98>J<2>u zo@l@Y%SRhwmcyrFL`#E_zm0w2|FJyrNI~U57+K?MgrK4>tgrZ^1xD`{)_Rj$PVCIa zSxGbZQ_Mq(+{ZAENvaT%JF4jY8l=PQacx&3Z6@36wqSSUCg9z#jmY%}0o`^O)G8Y1 zt3!dsECZ095$8#&BRfGXrboDHwq9M#(;XohVYnaAEkEj0 z%o}j&A{R{imR0-?I?ipt<#W>KUdtPpg;_3bw29(_{r`2*U(zO>C%QKJBGaQxf`r>n zx_F+`3N*7%bVD$`$1Qy4qmOO%6WN3ZeNVeDf&Zr{yWi>buH_vM>{6W)pO)4`Wc{>z zHoZQiJ*r8I1>dXbI?3``2Dd`~W3L|m4f!o7Hd7?k%c?!NAav)okXT+bvGk9&JZ#}-8qw?mD@U1s}uQwyo^~IJa9wVV&Ixi25 zu+T426mMTRQUjrIMGE8?r6_BB+Lb18=ti|(Q=-C&Ot(GkBGR9+u4^4ZHCxD5;E z_H;bLHs@W-nC^$Y zxXnfPuFY8~w7_Kkg^h>FDsKL?6WfuUR-j6wn2i)k1}!{bMww+W&xaUgn4@$R_Pj8V zQy#c9Scezy1M_bEoKfJ`u?#O)D6w?o@XKc&^t9Q^f3xUjgI*MPZ;;m6OS-@e;~R*- zq-is8lwZUb{@Hf_oG&~%x)(UTna;aV4c_?aPrNK9}%-8G$^C0gb6F zD!LG|e|F8grrc#J6`nxgEKcTIxSc}1x>ldOY;qBu89Aiz1R-=mkaCc`UHFfx!mN8T$xh+~7I zRoOdim87^V+??1gVbPeX@JgS!RIc}^nS04!uZB3CdI!*Rkmg-V7kU~(Vm$R@yKT({ z8XQM=M9q)X{p`0Ri2ytC)&WZdI0*!nb9gHRpu}vdKY_X#df$4PUfski2X-Hx_tG&U zyoM>xn0%;LubA*a@rBp3YjG2Hd!C!r39Wj8lD^ja@RT->64144(x$qW`>&YL9grFb z1a;^EOPaSb=HqkP5l?VjVD#Lcy!vQF7=ni%&YGS2Chl6J$Nh}lU`js2YOv1f@-#5Ko-U~p@0oU-F5&5(uQ7EPdIxm@^)5vv?8TTDo}h%GsPjTu6bNwlgyN

    Hz6F0(g#^2F=UNL5#;dWdDl(J#hUFu zwm|CSypSJ|Gu&KZC!V;vtlo?^in&6OOH_37_pV5H&~ZSHFi%{0{G~k9q9n5`5^!y;N~laS~FQGF9DZ*3%GLCv_P%9PZHJPPfk*H0{_ z4Q^$MBam#iLmDHhBlpIJ^KUnH2N`^hFdSBMPS~YNvVU@Xyv2uEx+1ui47fNi2^scNI8s%fk)@9it1~Ht_dKVdm@iND?IS zx>WUwDsd0L%;&ye50oi4OI)6TXfDQRdtgb|?{PJ3z~!+gvh}vAKJ!X{fbnQ z!~+sTtMcH)93Q><+@yM0rx*ix11?zK5jO#ry_E{3v)qcp5At>e54hlBWtXBx)=5|K z`rQjjJ9!8inz~xq)vz_bM+CZhxVHgPI(0`~jj|KuKCp{-04UpmDY`$fn7>PL`fXjC zFbe_}Wj;_^i3Yt^^PyQ2xK)WIySUU^@tiLKLlE9-Qe zVrnUJh>A{|1nXr%sUm553FwIEeV=&Ua@XO7G^nsx)OAk zG5tlCFY2F%2V7BMF&3i_4z5Z`mezheShF&9Z)X+)VG3*90~QAwB;h{}7ac z+|I8Tmdx&rNDDw-OtK_bouZ5hPKFs3GthQ=oT2;3sZlCQ?GO+QZ=7HH6sJjZ++_EV z3%|51qu;yY+Cwf)ApKS&^LL85Ns;S7e^e$gxIfe_-xq+nrnt<+a#FftUNZwim_LLX zR+z(qr`WCzRMfkEsM{VG`&NVJW^Y6Ip`bL`)@kbkAg{4q31vk1O%STb9-ux1#BBTB z>u7v7Yi>4Vz7~3CW4-YnXxJu12h?1R2KzA+FR*+Xd}4|cZ)R(nm`^7mR|9*fgS}#1 zFYgAgoFp*2`B)a>Ag^9v{!&I6B*`jLl=s|+3WhjjR5`rpEKOe+VQ!=9w2}LeMVZMy z7PviOPOc<+ZcGX1b@IpotGA|zV)7}n6D2e;+IPsWMT}9qg}7j96D|v?4Meq~jIbJr z^dW3_`}Vo(^21X7FtW3-U1^xQLUdcX5S3%IU~Pr6BnvBLN2WsRr{av~Fz2loQ1$$R zgK+=6LkAr#*Z03a&jOhDYk$0(9B|?=-$g43d`2-RC~}mF&I&J^QOPfpKKnaGxy7|c z5HIM5IctM^<7{)9Z%SCbuv<_NfrOf3WvrlHks;j?o*RB3U@RV@1zGs_U>$DycOSGQDM`x7h^lGgZvjx3%) z(yi?NT8c@e$Z9IuNMl4VPgJbRncf?*M3Csws!W>f__=$?-_PozT|cJP6JvA2>D_c* z->+%aG?}gJs?*LYEJCUWAQhzm@+KjEB0MBG9{DJAy`WB%<5t5nxFLu3QRp6Z4G&{0 zV+UZnVEf#o7#qyOr*MKT^{*a(E&dma#Sn?s|A1WJw&ro-DdmBc;p?E7c8au8(WgL< zCSB6yT_-nHxIoDvbEQ}e+Izpu7s!GA&k>`RJGqg)(in3@fXr!4vc?*#vX z;=22FP&>*3Pi4DDGt))3$eOh0x6PscChcYa>c}E~TEJhA@QW3H4?IR^ z)8~LNJSt<&IszO(z}m-X++YX65Xn~5?Bi$0&#=r|^ILP{N!|o<+{)bVqnJvHlu^-N zF!@3Eq?wWr6116;qdtGWI?3pjDQQsu$Fm!aTC5yW=T$GXX@S-nC~8uSfgF=`(@#ky z|3YA<MAO>%7(>(Etcqp>t;ha$*UHf?3YlfCOwhFwM7npcrVDDVhg--X5q^+!5L! z$ISr;&uVtV>6Hf2fmnjBet7v8YkzJ5&AAyD$C1xpGnH}BbREROyh$C0-5hAUQa#B?N=R11{Le5WY!vS&H53)oB5}zzfHZ$roJ)z1AGqWQAQ@ z-k~5|VWrc@rB%?mxow==LMTRR7UWCh)w_0q3g3EflP-SF^tAuLXzs z_3{M^7C6|#3^5dgFL8j&;KxS&c=0os_O$RHI?yam~4Wo(X8%wdXDQ_;OM zKfJEZm!|qbBFekV=I^8vCEZ{%3f(%D@iRB8o3uyCEI0D8clE zoaqCBD@5s%G%en#mtA*@5fw!A^Ub|s4|#9>89Tl2-c2s?+LZ-inA}n!$2D>F-xib>{uv?O$mQ;&kcK+t~79f4JChl!=4_ z1$bHLGbcoDj>1e9;FM$i$iFQXWq_zS%7`O#DC=EaFTGgQN78{3Q5~5iI6NPw8f;h? zo~428bVyGyWiuhee5Gu}BFBauV)&gw^)XU|cpnSP*L&U1tn3spSfjA*rzTPF$;Ay67#LT)O&5)i7-r#P| z3Dm2vsdO>3J`3<;RU2KmtC6a%`_yFiAqvmvgBzAf>cz_=0gBDDoNzN<^RMndu-#-}iQAivDe^{Bl>M-fJrsCYbQlsF zb@*4Ca0du_SBcH$6daOVQ38_@7Av~%{zTbH1N9BZ0b8u`U5GIjJiR!#nub$Hjtc%5 z4jUPpUv{LYoJPiZ_X$5-Ib|8Kn3@uoj8kOA8{_yu-Q7WkX(PoXQ)DeJ?&G7*g#c3> zrLwpA7KP`Da-}H~b;sz-ynIk2*cTp0IU2+-0~;rRzpNt%-}{l*@%;-e$SCzWtsor} z$TBNm@*&0Cr^r1jdQU_mZ$}u?kYH2!G-;e-5FouKMx?ufp2op;dw1Wb4$W{*PgwcyDBpcz>W_t}fX(n;%O+4I;AC z=c!8;NXu;CUuCv4I&9so>Q|lH7NnIMTB22H*m5M7hAKp@R2DB?% zAqu$E{ie{Q70?8!Gq|Zy0k2YF_)w`4BYO26hvpG|jTau~gywO7o%}7et$6i|`&v#+ zuwap>MmG6}x_$3nnz!h$&*DcUl;~lUf!XD z?y|#0KfCeU-ju%VQF+?bE+_s2IxBszd=2-17S&>7dgx@oM9kB8gS(az^Dvc3;)Zpvofz*0B zjiSGprIG?FTHB@N(%pfmL(>Q8za60&yaHZ>=MhWg(#78g#r8>0qA*pt9dhR;R*4<7E*6owa2 z4f0CtLgTr15@;{m@ZTGD$k4eNSA9Xqnh4n((dA;kmIQ-x1fA?auCSfs~$@$=5+8` zOvCWdXE*-supz_y217r!rXsZbXBLAJklETxnz%)Poi~m|dadl@4T>>Rq>YL$R>li< z2Be9uhFt(b&(xrLVXG3;xl3Tr=`euC;0DT|ywo6bNibCLY;mg!>G!CdVU#_47W4kQ zL8?)Ac*-_kY>+S7#`Jqsg$!<#Cuvqf5$Q+}%@A})>aO$8!7TXyvG*l#O=Vf$^~Co` zE{1FblL#mfK@dx}(4tn*>aOmd>YAGFuAb@nrkCpJF20$vd%CBlJMHw2`wD`J3SvNI z5k!#11tkcGf?6sdXiD5b5G}=0a6we~&Pjrj6w$np(9!yS{;GV-z4yg;|L@#$&+9!xKrx!9*FsFaS+v$EK7Q?_1&CrT#C8l znm0wKYU5%1YmuGGY~_vVi4c;JDBA$iu8@QGFNSL5P|clpa?VGK=45jUg^hGl^oWVp z9f>51FtOss(ULzB^9qL+N&|i>K;ogz{x>5pFi7BVYMPc?MxEr8QRl!*krdUW(*C^s z>j`r<{A}YlOW%I_PP6K|BvpI@*s5}w(?2^7E{beU9%uFVe%WeymKVmQl=^0Ku!RFE zQQGT$Tk#56HF-^xq{Y{NcYP#%-?5(nyB`OGz z5?vNmKz&ubYI!kXg%fKUv%|PGBig6U&tg74ZEn|L6JhMcijV~>P?rR`r=fl{7OPw2 z(G^x5^c2c7o`26f$90xgMUi1&P^3A_!2#4Hb_2!*dqY@#f6}EcTva+ph@lF6f-wN>Un*$PQo{NK6unbvz738_GLN*3&RY`3Leg@ zuC?s{=g-Y+!8z@^vejg7yN;44QY0P=_+ZgTLxk^bNxb}NSP>Xdfb49&{i8sKL4wI8 ztKn+W%kd|y;5z2de|+a7^W1C;>CPRHc8d8Cdi4!6#WsZR3DQ>;L7MteUQeuk`?fL( z+Bi|{-oDeO>_1xJPMdE`oK{#^*lOhQ7l#*d)QiI(M12vVHgNsq*-$RoMkEaBW+hmM z^cbB(c3`~~aX9MVji%Hmw+qDN(pd7uL{8tP zF#11J*(JLP^J&2w*!|=IOKOP>T@4;6V&$xbGqOf2MpoRb!{pGVvedUv)O3dw~Ul5-@-vhgl_^#L1c3D#JQU7*|m0 zmC7?il<7R$|IB*C+W*-v8j%%ZsUIaD>-xwjiDC}T3MZ}XoI1`6c`{%kQ14Rm+Z4G; z#Ud~MP42o-ty`~qL#Uz5q&}~$d1gUG4!x8A05V=Nfo;GA6QfqUog2S>qWT`ETXGI~ z?hMIOtKFV@-^(YkGVZ=Wn@ zDjMtz_Y;Iyy&zz#dFEcJfovE{7bY+ma|7&rZzum7HiZm|czHKesf!OsqQ7oZu2f@@ zx`q7hLM?`c=7;N;=9v$okp0B)B-2m8dW1Soni@wo0j%Hea!-30G57O}F%WStGDm3)tf~m@ku@oD=zS736b}^i5=?5k>A8TL z1F@V4K-2j&d5Ly9u@NJ7db;Id{M=nz%onWJ#GG$|Y!5RrardoHqb(0c_9FE@2nNUW z+IW3|ZpWix$b7H?#*l6fKg{aNI z>e0r_p-+PcaicI^Ud8lqw)%YrkxHFBB>A<|@?$4t55c`1U=2gCLN?V!T8^1}h@94{ zXTf!~kiUaTV36^!i{2Ro!M4m{Za%$7lr!=85QnV|ML?r~H`&*3(Bxuw_!ufRi zsrP^j){Z+G1Vm`OUA|ba_>T?;4_nX-3J?20!}2JWLGv@s7{9Mpz<&-=GRhKDV89xeYsXU?SQv`$^F3=o$Q=pya(2DnKIY8fwLe@*$C)&yhd;a;cJFzXw>!tm(3{+}u3HVUu(A9_g0&DX}Uw8DhD?J*_qrO16M zwg#*-6?B~E5g%XyPKIQ}RT4N@B|&BgOqr zP6)-4;dOKat?rZ*2-NM71p=+xb;$t;2&tD;a#LORO#_lRs9Qex&|?ff9DZMq^t5-M z$CYsW?dVwRsSMuESlU8Z~XE5S*Y5sd#3{B2sdcHgG~8js3gr=?%(U*{onu)AbhcsVibM{4Uu+k}p-PqV%syJh+Z|NY}v-fq~$pxODX&qi^P z7f{;^7x3#q4f`VMBtakM0lENa@sa{|dzDIejNXopEj+BB;E}at2ak&1&Ddw|-E`XB zy#;?IayQ(V*3PIq^c?UxG}h5p*NHl0wFH@FMnP5#nKBl*7}Cq}hpcchW|rsu?XNm6 zK6`HmrHeTlQ~&dGL5-g0<{EO||1$q?_sN>?r%pR(iOa_^bHM@^wr$^R{MoECCT;I>(KfYo?GJPONh*GtprmQSt{A>7ioVy^4TPN~d}#TQ4h$#{JJ7w}-O5vx}lr zRL5ZVqf;%Id}CT2S;g#t^0Kw$GDOPVoN#LVMf&r|?2ruZS@C^o50qCf9KYWWZ4xE) zawaq2X}|*29#JJ-At>iv7Ht6wMK7r#mBCpw-pB3N`Pu1#6@q3ta5hM4>AqR#n9AU- zoaAsUY}rQQCU(HxHU6DoeK;7TQR0%A0gRB(a0E~=%scqR$)S7Sl0>j?nA)9jp?-}X}N5$_p*)9i0==X=b`&z9uhwH zL*eUualcEcZ;z}*pU&lgSRjjdVh34^!`S@>JIH?GxkqL0(X!w;+9U?@FZ!|wOu}Hk zXjbKmOT2&({_jz(#=3WJ_gDMU0a?9!=d~K&`u6R&zGYNwj{f4pEmHG}6{l=7F|agI z@&<~WqGAp9FetUK#Ji8Pkl*Lt1=hJtWp7w(*bcf|wBA3@O}#y+VeWk7!raL}EOJFb<|>rCOKF0h}Scx&}^bw|WQk|(@KnkQYJ702HWaTq$t zJ6IIBKcq#Bw9$}vwIdvztSw^Xr5|u9iAsjZ^EyS9xKfD@*d@Uw5$F6DvP|i-?i(1f zhGH1trO1EY!16b6to6H>tYn_*kyZKT2J~_Eh`MQ12m4)6%{)p@O}iw%E6}Mramr_d54yb4LoP)pgzoiR68y-s*HfdoFTtXXi=k?SMsbH!(Djl&@S0bXI_VvfCeEkNiqAqy&n@2P;LmIX&-jul)sr=f z*02MC2AV;=rZLrVwvrL!#Xj)((kQP9UOGPcboxXiUjA|CFFqz|PP{EYU;^ZPN}fxR zY{*&q)@3mHxAC;md!zFwY27lT&PK$^(gPDh&G4-)Uh zaM@`Bmo1b$l_DFdSTOTO)Jr;P$Ye8Au7IKyA=e7X<_C-VVsKb=Jk|J6v_A!+6!>EOBZm0B|AU-?}_DXvK`X}f4Uh)Fg zJKvp9R4#{~!v|9U#FOm_z$>+M96ye)Q#DEU1mL||I!%xUMns)zC;zmt1Wb@6Ob72N zw^^lA?Fs009~`ZjkE1p7oB99S%SU zBikanxywNZNC;gJT0r08TnetGYdCl4<+I}Ll{jSezuxNCZ}o`wX=wR_eWf--lFb>p zvy9RuV9~0bWWN(jm$N3)rH+yxp-45DVW{6PozpC7kJKoZg==Robzk}I%Q5Y;zdVfW zrdPasS@g%NvLjK|5YU_&_?LB|Y5tY73pmxhCe>fg$0Si@-(EfU{@WM7wKoO@$tlp2 zdqj8NSvseVE}4v>tP7Z35jxd_sKv}uca7rNqtHXtN^%)&631^0y*Xix|6;~Q)!1r` z4<|>gx@H^Z*!#Cn5A6=YSH^sy=GtEnu2~9_dhEdE|hJ)7iqC9PwqFATr{ozI*`YyT4l% zI=HD$?%qvyy*3azVFD2?B?lAEJ}UMz((8_!q&=jPmn`d%?hZPx)F_v8TRCX@&t=kF zABPoku8T6<`lU(HXxrIBHGZQ>RUFgDdGfVpRTkYMTh3h+vL3>4(Sm(qZVh*ve>(jT z;?uE5ld6CEE^d*rwT%ncmnP08{Jdxd`_dC^yU_Ib!*BZ;@w507pVcIt9e$iRxmRm~ zpF&Cw*}l0{Yy+5Yzz7Y<(Lwbb1I?hiOxQ6AYm;Uw_j)%%l(#Gk+))psamRf^uJcy! z46kB3f#k0TZaj>Rmp3UkyA8P1kcA;NXkCy zS>UT`;Fb7dytlE2v7tt>KGb(Tv%s|)LhVKs%2uLSzi^BKc%MDOGKvH0Toc`m3ryKN z?>3TE>=qd3jrD@XWT;JVGbP_d!Tf-a0IHQ721jU%7zwX|NhuGkdYi*fdzVD5@>t}N z%mEuPQY|6d>?om!4Q>ot$Fr$*=)sXMI1i$Y=-8Me%^-V+3xILreN2-H298ql!xX8c zV(Y=0TN%78{3vOG$lG0{m{}3HmgGd`M4g7D#x}B+WQf*{+b&w-y_ne%ehsn|JESWD zGejM{PAQ(3a&~xBbMYuW5ORURR&zbFnAu4ukv+oWlU7c{J4haZF33H?N`2@?0#h2j zC8$+)es=Xpg%W$Xyhv3YdT(TCvMgpQxdFFFvQ>6sF=Jt?oEn%$ZSlea2kL{$Y7jH& zLd-ourx9+CKnelLJ_w&Uegbm{R4@qK^c@v>V;guz1`KO@|Hq%_{_D$C6ek9P1r_C{ zCzi%iCkPv$zy;7)!1R9x?w}Oa=8$%v+I7h0(mhe;B{i8}GTzFZ41hCQbjocOFS` zV)xHJ6D1^b-;IKLNf&*^3#!TU*Z5x+We7Ej1ryYr(n6Wmt#Rffg<-Ww z2}T)G5wu5$BxjhOa$V9T`wVvL{f@+tbyyjyKMXx~9Bx=)Wz6kI$0m&#w4p5vIYI_p zoR;~9fSr9vAi^3-zM3K{sMxbUD@hOMqMQ1WxMNZd-O2g8$PVj>cO5tHmPclXh_3q| z{;zptW`6^>xXuAV*^Q@Afc)2kh6ygoH=A3MEjaZ77HP_MP8-8cfr{LjH^J% znkR%@ABea$aPHL7yS;G3+oVY39-5KBEfI{C!pj;>FF}F5ZqMg^EHr1fV}lRp^)Of> z=s-uW7uEmpVA_R{Mf0j)kc?Cwc*t)#Q!hC$OA8qY=y%D2c!7aYC?);;&9j>A=iMBB z#A-F77LNbyUGo}87K#HQBdeYd_o(k;NuNid^wgVJI|@_Ekre^!Mj*IMd+b_ zKsHWUI2w*(R#WlpLv#2{u#01MUb#&#`DBNAya0TaZoiHFkN=AO_3u9X_@m#7mQwOX z6p0&g==Vq`g;)q=SQm;W_1^c7&CL_;^+qSny3ho!M$yGq$dB{~LVqjaYKJGW8-kVO zc!N3f>4$E{RppQBv%e+FUmLeD7~h6iST<4eWQweXU~FSvn4!E;K84CkyR1dr=bp!J zg0e>#X)s#atA1v!1d`*ny%z9wE$m^UxjFtdQDH3<9>%hUhRBl5vK!Nq$3carJpOwB zQALAcGOWXdg^6?E2&?uu|M5tDV|;F(Tra^OH|z*05-=08x82>Nzr>DKK=9H6cln^HlBe6hGiM#Q1|21I7PbAyHM&Bz| zW1-#*rV0`zI=wf8in`x-hf2LR5Sb6En4;)p4k#*X1FM)8@w(6~sE&dxk!$=Pd8_dt zU6?NH6>B`yN20RmB~wy@3&%I`fQjAkl;Xfo6=$b)h&#g)m=mPQvx&hzIdr2SCD_m% z>S?44Ync_2A*Ttm>OJ7Jjh80IY8Y5^<6zJ(qRvzzc{@H@qv#{86WYKJRU)XQarE;Z zwTv1Ms^CF!A^SM+qZDfcvy@ItvA);W|N0!Gyo#KErj2Na3yO5!9Y1L?kzpq(`EiQW zQn6U>8)}I2;^nF0d`=$(Ae{9)HUrJESnzuuGVs@kvS=(Vep}Km(_gF(Y870Q8R~;) zgja+naxVn;d0=^!BywN0a>8PtbjZ}0R}r*LzB_1XWZLAVk+4Fx$ewss==bb>?md&s zJ?C{u`(#+>F-xff-czLF1a`w7>E>{(+{SJVwTpavCg@&#z(x#OmTzx%UwZp@Mm+tZ z_6ITKF1x+96RXv0OvrW@jFIM%Gp;Zhl57@2@%u9uwN~lX!L4Sh9NU@3GR;u`R*#G2Atq#I=_-ssfJHjj# zJSI5qQBo@H6>aff5Z*|)`#7ykyZ9(%;j11alTK?VVMZ9*Yt${`rzC-4D?HeI&=%WT zDS<5Zchn8b5@;9=>kau&*F=saF~o`2KMRQ=7@-LSQRsbKIU#?F-rblv^9cdTK zLN5QUG0V(Z1=+|ACpP3*;2Xq{+7@xUtcW@43S5I&lL8}bHHwqGgb7%G0(94Q83tfr z3$0s);3}4GRt)C4!rt5ccQ`vD>{QUS1xY3hC=zunHZT z&>qw09yd#!Cq#R;Jl;dK&x7V}CTaC5THvq*_DP2l$_3KEk+9B{%4 zBvhE@QpT$uSGirN1?mf^Cc+(X!6<%YMFD?KxT6xlUNmV|qT)qEIOHBHRE+t=?{S5< zQC0cz={K55DmzuhiG4kwnGA`m%%kKG?4L=+-VvRKoaW2CcK2K+C8*uq;DGrIoQbD} z9lWKgb|@2%M@bWkLo(zYJgwXL*{6jor)Bu-RQYte+amW*6o)tmyT4uwYI>~LWDL%zp6VUxn%YKRpRVgW9Cx39YH zfuYuVSIc@~+00C3T;yh1HxwSq0r#p-uFj7>1XQ2#;aa!C2}p~B{unhTZsiGY%ypm? z(9e^V)Ba^qVBcB((rBz!nKvH&pEZ6i##JW&(Z_!xaZVgjr!iSK5-E8+MOIR=S~nB} zLpHe>nD6N6|I&Mh3sym5xQ-6K#tIT+l%>31M;amV^D{Gck)qc|@i=V)pejmUK|x)# z*d=4Nf_C3ZNL4HJxZ|;OY~QRcUMI+d>1{kDl}#6FVXdizSP+csxXC^0xn%58^4Ps! ziZ#-<&1~W<9owW@Pd)lsM@-3_ec$i+!P$44RbX-*i%jU(ptxS&tW~pey|C4iA1$J; ze{%~}0gUPPK&VxS`?3gn@fACu@#Y7;-L6Iq)qbt=uVk?ki9K$%v;Y?-k3i6yBez}+76S)W(anoCG2Tol6eO5m)&|!?5MGjm3-7U zftw5_)=IkF_b6CaK`_WWrKi0Odtrmm#-a2}Uz#$j@q(h@OT8U;!7^U{FmL80zENa+ zQZnv;$?Df;*?~%%L-xCAlspAD=dmC4h@euge3KYzxX5nFbgF|vXMJwTa_FnlP2%1# zo$9#s0+8V6`7a9DXlvaS)-%ZpIHL%jC-dj>ziEWWubWf9PY$z#hZB2lT20__hLYD) zjHFiot3FE*D51f3+t*tJnP|cU>s* zWvI&{vS5dGmWKsl)fJqLLcH@Z5Zm9DDX16Y_Q$a-d)uQj;Wg{5Iu9fdk4tzi3X_K;uQ@*!$knsg-%= z72Vl{4>+;Fu~6$h3lbOGcztx{>?&m^4Nl$;2*gYXO#tdyjS@6q%yVDG?+~W|@!JDI z=H!wnyr~gul;9)IQf~2z<7Wq54Asax=o*5z)_H1_>SLtH3q8;6GVLTx-a!ws`oWtW zlR9V|&`=I@ulHe&x`|mK+bn}eGUdV05reclq*(1;Cm*!#Sifgt1FIK_6D#jojlQOj zL*E*Sn&ZUEu?01!Ly9CY0H>27UG#E-#Z$Ax8wG%w7V(lP%atwS?Lj)A*gr1Qg7e7w z8MgYB57&@AJcb*^{@2b-j@|ovqnh*mtEUO+VW;Le@q(0OlHs#xj67}(c}m5m1lRM5 zytaw6=tNmB(W#aSE-)8^o`#*DouVwC+XoBPlCgEnQE9()i|5j@{Vq$!9{%CI7{gJ9 zpVn_f)Zq2I?`-?0`A3bn3WCytp5IzYDikYbJwLAwsp8&oTgWep9_-xZy+EEZ`T0Yz z{}r$M(hljyXv4!+lC8q7>Gz-g48s`r!|wA2$JjnGeX^nL7yNUObMzD}L->PntcK75 zL3OaJjoi@kkyLUIo2#qM*%G=ON5}eq41I3_*?5;GFd^1HhQL)M4$VgT1 z+C?8Dn-zE65BY(c=rOIQ{0Uqd3DGY;B{F>q$3tnUFRHeqB6`>hp3&K#Ti%!t6W8*_ zA4DHTf3^IL7{iZTxDfKB#v%*`-{UBO#$=tb5REjr=oL8d=KT4Xrm@!>wN)&|Fn>PgDy7rw z0&7GX<^1_46$!#_dKtF@WIUE`FpOi+Tn%k;+=B7vXKNJm=i5inScwBG6UyTF8< zBssF`MC7Ypo!Q`eJP0H%E#g%k#st7DdMU{h7D^MjkjCD`B+FJv&gyM08pYr-n2ce6 z1H7f^xZdxsldc!$D)h;O?XvRV8ZPd-F8l9OsimFS4LUb6A6 z6~L8n)eF`&Eik>0AYsk{P%?v^ZOCE=rPha1VotvON_iJ75OOC7ulQf_&jebIb{P_k zVJLR5`-AD2+Mep&;)9!M1M$#4y`k;|*%fgwGF7s}qmi!iZI@k_9Cfvi$gwbsBY}^F zA?0Plw3;3QqBi731>U)DW3(U}+sp49Noy?)dPG@Po=#9!JoygWZwAef`Xe z&<06SV4)PcE#x=4E`riC%iZjSosq!AOq>`_?E(@A+hAqZOY^?_Pv%JguauOrkQjip zKMj)9z;wM85Q#HNBC34VmC7T4uw4;}+-*#j*SsT9yCP00pNRAe9i*U%(KDcRD~Nay z1zvp59Xo~09*b3tjII@JdEEV}8T;nLlR2v{*w(>Xz36_P&FxD69 z+^mR_@kYS3ym{^uvVJTm1VgQ0nUov|ur*X{uX{siTx0_e{kae^i;yTNaIP5hN z=e85CI~JVVar`cU&U>+Bxf>+SgHHkMy21o*C%s*CFrZHcbj-^CK9D zuFHelrGu@2SVP$Bz9r;RzyjvD68k$0M^?hcOi_7$YCQnhwfvut=U+6>9eGWJh=oj( zYvX{R_Qv$(U_j}S>Vo_Akqr);_ZQ5DRgic=e@ES#@nK-1dGU4&RnO1O&3yORyO&}T zsbpE5B7?hAlrmYP*f^!v{l@e)vP$||WT!GcpjW)!uT>TYC2bo)gB_JuZPmMd{&ne( z9Q8hjj>}3bbzV{YD8$YCAB>xq;3d=kg&ZAA+D$euXDRs^iquoF56ETF0_rM6M(6T( z(0ij(f=j2BQJJ%M5?yo;r&*HbTE$(>?W1#puX}0zvN^zH5}{MI@v``Jq@L3hk<0wy zs@I>IRAn*0uZ-z;IgEcbLrsQ#(x!+iW;y>l$qgtF^t%j>vrl{nIEAaYi}(wAgB(?=G-7XQK0=WSG*FKPPz-ka*yqlKaTz;UI_db z+$=sk5NlpFo{etfyz-Y=s4+h@F#G{I$K*{p7}Q8To_-o)@^Yp=h^nO*Q1jQmz3}a= zb9AcJvs%NJlgHD$WNAUQbal+m0G$e5(y6i5i^CVZ{o!wo3@gKrX^*Qjrg?JA{mxEH z3uVE_jNDxH+kipQsl$L7+)NHALk3fhVVT>ZS;{S;Mav} zM*}W>9(#iJ1YLB?hE)%>Vf2)B(t|xTiWkaS>rrMwV21VhTEN&kzODUEq%;X> zqwsm~eg6|=p%aHHWS9t=wUm4fMOIU>0j&^nr_sXrD z=Ha?J^qOV4FibeY8;h?jYcfx1w2%wfAh{Y=3;T&ySss{lJLq{=!*q%!MWaAn@0!P7 z7uXdxuR-#h&1rPl+^n1h6B+E#*KZu0)hsx#CV2jz-y1W-C^OP!Kg}b1oLFXn;dzMG zdW4c!Q{)g8n<_rW@1*19STrhSGSnIqKygJdtgZJ~ua{;)Vu;}scgJUaE&?qS2K&RN z85y=ukga~73UX&Zep92s@a-}o?ynEcP;c^S<6WGZMR!Sh;Q#8a?maRb=<%Bcf*R?* zncI}$Y{HoDJpNsG%}AIHZ2`Cm+FyKP^S8hb;6EJ~U-;GXP8aS-?kV>OQz^aXo*-|L@12=WKOmLC1A>P>-Q0eca$dTq-=#WQ z8=4;2@zwKA|KaBkYrkd`J1yV7kxfe3xdWWnG(Z|ngj5YBhitYA$atM|S#%zt(e715 z_mcITUed<9E-Il*{5JUXyHwJrg$RUh$#$?S9g4XuTE<1O*2df9eLiF{aO@V#_b^2? z-tSaZ^76ftC*s!YM8tYB5Sh#=4Osc6PSqn_IRmS+>QrcS+Bfr{>&h8VME9j5){q?~ zqihyV#*UX_{`9fAa^tkdT?<;*9#X^{h|#YmNko5;5#B9X2XXxaF5SFczRBaD`XL&0 zGnMIddPsW6h-cG^w{FvfTQTlPo3L7x-l%*mb1^SwW}ylgmZ{YR4_dyU%H1aM&9F{d zoe+veIjVz_-EzdBG9*J~-FSH;e8G9b7I7~*>EZC%uxTnlCm89gep!EZQ{l7)@BX9t z%*ENpm8j{5i?c`>yOqd^Z7JtWR-$8+9CFnTL(v%thHVtZ*NFw z2knYbSHmIZ#jXr35X?iqu_gtEscYRDcqP8+^x_ckL6c+|vIBwB+I71K^G0%&9lX!R zXF;*({FpVIKi9!fB~e$sJ{3U1(e9w-{B-)f+>l6Hz*XbOspFUPTf`3nZTos`{Ulzj zDVyAV8TyC4e^~yrkIk7m*x2QqSgWzX<53=YZN^qQi&H4nsruZP%mU&v2w*tJ^ts=b z^tta1>31n+7J>2L4n{ag6Gn?Dkv%|$olm=8{NS_ljQbTXbTKSnw^P-=;Mb>rp}C>Ojz08p~~Vi@-=Tn2b0&21Vm^ zu>yuoQ*2u&%fMJYcH)oc7&Vp1`DfaQ)`?AZEhf6hNlJd4BDGZP9#I{$mb8Qp{@tVx z<6lcswW2Fe9H2%}* zozZCdvrz!C!j6erVeN^%u``TkfD#<};8*6Q1e|vBY@s;7K+u`!WZ62;hmwVS9fMnD z{L?K;4$tAN^DOep7lxlZWQDRaT90gh!Uzr{7?$@ z6EWTjM=F)pK+xrY%e8U%{{~48jXvsp%0}P}nN|m3Xvh<6JG3l+hQZJqe<@7P> zR6BQV8Od~FnFQW~A?|}xN?t^f0xGs+(vjdKk7UjT*w&=e{n9%V6FqdQy)$)+Uhy%~ z0AXBLM0aI7=!de8E(CYO>5<@MS&q1dQ^C6~>y)0CFBsp!tMaa-n^YT_(?V_NGVdim z+R$cITu7STsa@;_8{5!fu@@`YP&!en5ZG*|0EK1~dzK;QcNl|gKoS+Oj)+VW-PZ5n*22i+^N_$0>bTPA!Q_al>+V7DI z4~x@aykR3sUN-EA@3TWm#;)UU{n&_-=4t2OAfJsT4^1ZH1|`2nk*gTE_DEb2+5~a3 z81N3XUC8*B0}=A8h}KP`=n@Qwkoipw773uZ!0@043|_Viwa)9 z^cAlK{F10D1~R^&!HS6~;J&D04A#Sq%mO|}4`wN|ywY4h_elihtcYomt2@IW`cH_% zFJ{*GCwWx(qiTj4+@auX4nA?6Tf3qnbg-NHJg3{ehtmZL9gd}sg-@2@@F$ch!7Ifn z!Gprq3BDiGZ_M~L)t>(Kzx}%rcu#H@h{>g~m-hCd4TpTW0mZijyP`H>ZTgXScA^3$&3iutNmBF2FCU7;1 zb>q5#Yur@kIn_)Jk$J2no%7mfe7?Cf={rWi{VDgr4N~XCx)-eOLv$}4C2yv{!5)hl zbqG$ifwmLEb-D!CW~`42P<>c%n4Tb5cF6By zXqNIZ-8}OgtuBi`2TeL46%X=d+GQ=`K6kvHGI=0OUs}wt!e=Qr^PA`9&<7&a$ug{= zToQr%$N`reVIikFs6vpd$^cB`_^aJo#NDJ^X^QM2U@|-WJ7C`I7Yu*O3iD(9;{W;` z^P;9e-8*Ckw^8zRilkAod7i1@K1CZ$GDoY}DY`1Hmz046&#G@s*zT3FvOvg zt`Q}38YbiXj7k`&ZF4dF`?Kd&R&(LJ@+Y#{{UR~SpJJDclVsU&DH2ZX3d=G<^hQd) zo+3$9ESfU+Nbk_c_}z3aQyj98ThF_%yrjH1D`A>O(J$R^cU!=V!Nmf`FCNIwcP+!X z@cnmxZcfGcW$&01Bf|pWV)yI}*;;Z-w%l_~=v{gtNe|5RO#+g=3`Mo*V9+U1f-p^S zijR4Lry_EF2NhO3zxFQ;#{wS)#q~=gIOc+7eAFDF`+jAV95-gJnL=)`+oL%#a+aCw z(H>Fq2Nda{Vj&HRw}84VYEm8i$IGH-)fumZfGd8FB&E_WSyFVt_#Tc1c2)UNOU7a< zDz?5RIPJZ1TDu(U*=BJri}1KvrBmU(0thQ?Qau5CRkP}%tZGIvGuSJRzr&+02*=Fk z>=Au=L>vu!;OK*+9ng=4cCW$TRl(Wfg}<#|S_5qQFaB){NjY!u*9}EHYgC$Us_c6< z1mK%~;ae;-VsB$o@MmPhSW;@T5M)#G?G(vC;)kpB7m5MFg3xx)de=Lk%Ix<%BHkv} zsSrG^h6I)|;nE)Z3lBZE5g^FD`E>m_?+&C%6cdN1n^x8|Je(9lcIlk9s zJdV)1r6^8AP)DY+np8o}l+4+?JU7Zm2NLYipvQ%+6U?fgWi$-8(nWEq+&D&rxCFW^ zBumB;P&9@ZRgx+BT8gZpVlR+(*^clYs6}%^zIhe|vG0VuIIR1rb;HV^qx06Ey$b7J z|FbTma=_~KJFgw(=AV4?%t-)Vsok=LRR2z8rL>Kw4egYo8McG>l-nW3GEFDQ5#cB! z*m@qFQ339dhgin&k9u3a7h)6_*WbTcOmdxgK>?<`A#$vYlJBKR2~bM{$yKGeU#d~0 z2HvG}V4IrZw}8J!mN|Jjzm4t)ZloV835Hix)x28WU zguJ2tj8ik;C|%C~U`Zso$Ib-j#JileCMLMWW8@3QkbWu_Lwu5D&l&hy#Cw9yPhG~% z4)2$?LY8e6+HUwz8(q=FobuVqIYgQig`A55^euKO5Bar-b&8AnsH1c`H=spaMOQ$S z(vm5Q1J3xa^{5nIXAFDmc8?CBU$y@z%bWjN!(1yH&Gxu zVzGEey0D#5$N6lQt&;7NsF4H-x7`CS=$9_0JD^G&`maYxo`Gy(Ez+!!M|NoQVdF(t zb96XuV7rC+HbfTQ2HxMzv+5K#J>&RL4I4>KFi3sCWf9p1cIArDPP5nFLl;;()~lAU-5RbsMSWKnY1Ri{(2 z2~aAu72+kK9(uA2m-Z}W2aVgRgUaflF1pX-AjG97FltQYR^#@li$-Q~EH7CfS~klyr_IOH7o=hm`ytMeb6ut^Zm%XUmTs z{RCI)4v4mGkgVolIC-lq$5%fGS;{S57)zE3(d#8ZYuO^ksv61MEh;R9q>ke!FeOo^ zNDgdb)h)71Oao^%XG!3I%l2@LmAx0$A;pNwY$$Tw?gQ;{raR~?8WWk4Wtj7}7?eIe ze?v{V)Tgh68TI+#ay9ORa&SL<*Ay*%jL3})F7*LA=YDBj2aI7BL zWnLS7p7WUi@?apZSEDS5SZQMFAIStin`wI<7~#Qcf}Pj&_s^-2&Spg3AW=0No&hJ` zg<0U+PGGPMa?aG%o+X0A91I-K4LBI^M3m>YoHE==bA24PlV9w0el`LX^FX@2irnfY z*y4$>+|#k9!w~@9E|SaAAjkf>_2D=pQm+5!#lMghugyj&$7JP8q2!?AuY-*eM7FOB z%>r4fUhm(o$W&s`?|@4_=h%#7*>;a2U?8{`d49HI;b1Zy?66_d)uAT~7Jla^=42EW z_#JSc39PJ`SNklgAM*~Kj z4t5x&zuDn+>c72JW2H~J8-f8EX)s)&;%TTy1eOdvt8kx359~YUJ6uK#d4*QLev}kL z`Z@ZL)$4a&#q{_%iu0;l-kwf7N9uX?ye>f;e~;TP&#l5s(Y+*f{3+5!7kg=in9Z$G zEK{btjUI4W!FCjZ%L=yCTOUNZnzMCUAjLEY3qxWd>`7At(K69VCvtK3Ts=vnn2#Qr z!KA`bnh2J&FlrKn)hsx#@c89>|FCl;hFRy;qKF0dJS5t$;2?oW$+#XF9C6da)9Gwl zy?9olYY}9`^+2L8D&647o^pXH{fZlA3*X4ZLw#>om3yWwL> z-b<1DkWQvjwjcHvH%+xQI0;A~ooS?yf6&gGZzu~zEXD?-+@9G?I^i1I*}<}KkV^pr0d>HD zbk}>6INcxVIUY>!o(xgpaJNs^3A~~8?!94IAMTAg!c7ux3E9b1Mr+-&m7TndN$G*T zVcqUKMU8@bK+7GkEg^e^ZM?r;#bh-|ncO$)#`MKL%fc7WS`DG+8{yX9uF7g+?n~FW zt@nQ}Rcx9u$B81Fe$KN_oN?>EW^R@Fa_FxU>q8bSH9La}B5rzO@gX&aIkd|@6K6wy zV~JNo=zw5^bNTG$S$kmh{<3Eu&cKVU?ZViaW#Mqxm z$>S-ql8QY`cDuC!YX=raP}h*E$u&?Md$fUq^J*1X>;wQd!z}syzoxxygv63{-~Ayu zVgxwK<7C~MFbWD z3<9o_bxPp>K|L}^NQ3got+M2hYce(VYV+6++~ghL>PGK(zstJNwlH*h$A<$+95PY0 z0X=b-hq_F3Q-;iKEwVoMEamM$>}%FI>70A1X5b55k!5Q6%ZtR7m9`L ziYOIz@~}tpxaQt6_hO71 zS@m@C0>MIl;>-s`eMOW{uZ&LSoMRG#bLehz2!v0U1j&s{#L!qiL%a!EkRH( zcZ7&~nS`?WJ}X)EM#KjnE*NXXpQ_^%HCZqgg5!p!93@fm1d6PJ@?7w~dSsh@+9yK@ zel!Hv^RvUNM#P@?h${DLy@4R~tAI0Iqa3dz9e^75Bg-$G!PMU!A5G60C zNEwi)86z&*nR+RPha<(NmQyLNr1k3!XeSya1PVg^YeAQ+29ko|q+NC|>K>Q|v5%fM z594Brq(xHo6VD2|CG?5t^RVKeDkhJwous}MR2__V=RAHMzt*pPGX7_Wm{6;rZG&tW zDgN0(`@>Zq959cma@v>(3sF*SJd7~K2)ZnY1HbZ4x^MZUWEtj2c1B!}I6UzZ|M0|Z zBQ>9}nif{rU^Vi{8neO%)qLZ_Pt5}(*r?Y|Y$Ue8InXB9;GPp@_&;RZLg`a&Xojpe zEFl=8z@kQbFK#u8Mg|Jjy#@EqM!o8pwQal#!BzUQUn8Rr1ZlA&o_~KD0n}} zksaad{WXe8>0T7BeLoX za*Unv&WR;Qr->Hx86`hUkuy~6GViqk26DU1cP|2&91_bl(0!2!p46E% z6WJkfa29*@OLx#o%pF++T`I+L%Beu~dy%e-J}B+sB!u3KObXBdBl#^Lxx;ZZik)KQ zSi?dM8pWW?5ubupK9Zs@N5?@8UW^ykC@O@XN-JR{#I|vfm!h9~CqczJbfdD7U^XFf zC@R?;v%u|l#=pDgRo7USc`x_;EAkh1&aE_qSiHgE znzEbi;YM)GHCA$mde3)aRG3ljZ1vCFKnk3AKLiHhA$%bfl>8t?_CuHil;g;zbr5w9 zOyC%bo*;}5&jVNBR=4{PZt;G1JMU3TAMzye|asI5s zgdHc_%AtNBsC+NhC@vm`{y|FGUK?={XW~KVrsQ8x4I$ z6!g0+;MYs>=$BYuvnMKnG5nrieq#6~_>64M8cqgx(VWYo|Iw*R-n}eJ2(9NGh&J@Z z`gc`Kk=Hiy63_D-jEOg02CV~~wJX_(+L^%d>lJZ3#~VS{^5(fu$a*JMa*9n>olHu; zm4f&Oa6@)O7C#m@YmnR%+y+yfAxZpR6gVMUNrR+CtoKJ^ROmq^Zg~s|dC9U;sGVUt zPbM$N9w)X>ygi<7c??#n%p36&XMZ-)2$O%@`HPQ98avtK#Ou`o6Hw(-a^RiGred#p zb%Mh>(X|DtFy+utFA6yq6F?uv6s<&VnrlJCec+l`Bcl}_wDFeumq%s;k~1fJw2a?>ZT02k^-QKwR; zisLU1$z{5^-CS+xM{CJiQbJ#ecuX)p9Ap`=d+Jn+LY@zDJyILGmS{tF(1vIIad%K2 zXLk@3C!=$eciqziQpJXDwhAkkPyfJC|InN_{x#8+EzsYiW}F4#og1et z1VtuYm=SL4oZ9%xjoqw3#K<0CWAC3v{H|hxQM?5Fm-ZvF$%&Pgy(R)<2PKCpN!zH{ zM1Gcsf%K-5R$m}1$F=*S4wFu|Dlyy+3bra?ZNao`tS)FMtJsTC42*Z0)2CK6^aR&we8M zYgeF7wF5Tb3CvmYC-vc2OVZC)WImH#BR#MHU_bJ~hvs$MER;q|4&M=eFaR@PPYHAa zovMRZ03{>V_*eN>2zHTnvTJ+^L^35Xw*&Fa(u=V0J8Ztp`o-G(Bz7HHee%xx#YTr7 z&Ey&to2e<^t0PODSh2}8kuDo3`8tXuLh&2$K_V9(expS)tsL`Qlf#>4w99UPvneKl z87-(90c1U5*&%MEt=S>=yNa;ydK!W8LDiOXWF5N=wG-Q4icDa%oswrzP!>#I=>Q#; zxRd~k7HB!AJllY@@o0ojs;6sgXKLw1(~4;Q1&r?8CH^9!l^hs_*c$6GWwwrt6f`gC z|0CbUtIX?Zy(X%Rg-Tm(yoKUi@2$ek;TWUR6_^vWfwLXz=7L8-*hqKqmXFma7E3f9 zqbPLPf#W4JYRB8}r_R{qJ!q3V{YEoMWhX+Mch!VIhoRo+JW39+y_pb36|SAV2!g;d z30muhX4tKM?XphkX>UxAK0vp~`#2hTI$c3umwYB@0AfT`r?$FR@=Aif5a~@Xdhz2x z@UR1mnKW@A4cFbbL#CKZ5(}RomQT1BSqlsZ=^?;`_?I)d0}%GQuk|m9%AptXS4eho zx@vn!%R)GYaeE>k9Q zZ*j^z+NF>uoyr&v@uI<{IDZKjjc?aDakKQvA*hL9J;&4DJ7|o1f=El@Dad`RlrEWS zV5iO%rHO$nx?hS6(C7)C*XC9qyn)jO9MYBa_UIC&`nseYD5f7qUtsF!2KwIQDn^~k zl!=b|o&$R?60G4Ryp_mpm&Wlkz-5`oS**XJE{W=qwvxE0O`r@_G3m<3WG9`(O_BFQ z#@t7DWhu%!h|4+x>2OCpyCi44AT}9xi8$74lDhVR&)&P{>zLD`&%(-tbg&2Ldr_s6 z`#cTQv2DB)qytu#~xjgWY-xBXj zB1ei;wldUgZF->`9LY^~GBoCw*WaIRRC+$qs;WpnJEg~o_f5x5l%8@*UPh6OLimfrU_!Y6;=hkHo|JNP*78cr%q;Io^OLkZhFEb3KH9{@1}o`jp+b)isRQqnKQ z-E}FXIYLH%8@+p~{-7naBy^_&h1h^ggXB3A0IGukE4o1ND5BK0i`FqsQ2W?EFnKmF za9}6M3?0%g$B!>RUTq%P>9os}g_Q3$PB+wQQ=c>7AcsyHA3v=mqLa4&EWYfOTRF=w z8^-SU*{won@5`r&jbcP_SMY0+_}Yk(0uwQkLCGPcA(e`4hXo=-i0Tcf%}5iI6k5Q| zp>Odtip&1{(6fC}aDwcL$l&G)G2Pmr?uHjg~u-8{yJ5~aQyeAP#S^JxRGQFkMvHe{)@FMAGrtzpf+2PAr1zO|Vf-$q!MaoQgfJJSDo~)kdy(Bgv{pafXvg zQn*jOA4JvCgDn+tREsVmSKm7O?xe>Iai zlAJdu-f&yUdBbeT&Im)uW^TYWQM?MQI4+O8+r6O(vM=tW9I^d*!K|_Y(hG*M`>nla zLQ}q43?QG=y6C$!reOf|E@rli@&c-Z?tmqD#1ez$3|Wl}mgA1J`QpBs4BEKv zJovu<39^u#2IIsc140>wq{*(O_WEZ-1*|;cWBZxL1nX+d{w`5_;GR0*zdaG75x~-lwl^C$8n(T1tCtDfdf4^|n|fK+Lw~R={zoHGES%U% zZ-H8DXHdR)AwO57zT=uKLy|~`NDmdehI4mZCcTiL%6ai4198+set}>Sw~o7AoacKg z@ElaG>6fNJ?Bpiz3{f(-o%FkudNo8NLr{aDOi=1oEn38_nB5Ukz_}K75L%b|t{%TS z>Qlc5k@dWd!WvG$%UZI*XCePI1f_R|WiSb|3@xvV4C7szRt!=4hLNz^b{nrwdKD_D z=Lt9aq)+Oi@wezy-Sj5$Lh+t}YOogOK-uj|9`?J(!%^{R6`?12#nMD>K`{PX66ajj ztIxM3#nNmO#hI0~8s1o9zpF#!S_fyt>TesdH>T(R-Xcd{8|+;*QEHnhc_T$mQ?Yla z^-M;}%yi*{=z4S<(wDqeF=>)~K-NQPn@0k3Us5E@2ih_82V^kik?pctuVx^|TgYF? z&zrK%Pp8V|Uk$_B>4ui2lkf1l^g%WnMVDXzkaAs;%OnUhn1uu)_rO?JOY)$8QxlU6 zc-0}Wyf40?-2-MP$g;+caw7Drb?C_iP{X*;mI7Wv3fEZTSm#K zVhawena7c~v8)Y6JbwEV^IDRx6hbT%zU-tcc=7T);g;y^aL|j&k~1|UyoXcj+W`igEzDYokX-JDbRU>tkS9cHsV?vhBydXvXxq8xzLj9{ z8enXO3V;xi3q@gXgf}o~^Lg-Qvh0#*T_|pU1_Q-0y3v`ayF~AX5b}I?{xXzv#>O1f+4pO%4 z*Q^pkEf5qL>hc9`@T_qK*5RU^J)=j9am#KRH>0Mg8r3FdoB~*!vNP7F>g+Rf=zdOL zSpL>{{OpscXnQU?E}RQI_GD{pEXfn*I;@05v(XL+$0X!Y@K`g0ddc0;NRI(~{8=!) zOZ?Cm+ZjPJ5O|6zqaOnMcQ3SqpZ6()>)q<4IAcm2g6Uyl zWgZz)Swk@C1eHQ04AT@Fy_SR)fysPx4|DKwj(V><4{|af{Qa6u zo8I0YXhp>%=2$YNe`Q<|&}1ADzTHkRz;(8jNYKYsgmoxtr>+3T!hC2B>>~GsY?xX) z#Z2q-g>)sI;#U!-TR)|ctO*$OJ_RA%6u(}mqP6^IfHG_}LFIxKbB2$(JN3G>?1i#7#0!0J^d)j;=VY5%U;EMkVb%WykTwM~gJ*wGr zvm9N{ArEYG$MD%|pNDd@G%u!$^+73-b>c50%=w+_;4^L}d2(v%-7)9ZF*%vMiSa&b zX4$(YENgnXf3pfph!#omWFTCJJ56I$Tewj)n zB-y~Ba*L`Xk@`umUh!Je6r_C05S;@i|09|_@TQJ2?E!9CGDbleXW))9hC2>%0xpqM zW=IOLt~pKbtlmv+nKVjRg~y@P{Wj}R6~O@Q;C3Rxq)HVuXg3Ep3KGL|Lhh0Kq}cY? zDZLIIU%hk&bo({N^pS~@QgRhj=&#qBru0I_vQbnBlsgaQ{q#Q2QZ$gQ(#py0@fotD zFoVJrg`Lk|kT)V+jxf$pvh_utymK|2*rD`+^2f7gTVeB)Pv&f+%6aUnoU(yPBf$Vg z#$F;}2UH;rGxGoeMn^cZVRX}Ldp-vL+(uYiY|2ecQ!n~@P+s=492=M4x0 zl}a`-nSxA6(DX;|jxCYx34wqRlK+g%=jd+uV(bm;p78n$ltpeDV@A$F$W?fG0-UEW zoPrQLUPiGg6E>|*u6W&wArG0hmMY{ih7Q~*YoYi?MSsjO|vNJ7~~*2JEamq=XddN}&{9YTbogSqfJ zPKcQJgSh+)gw@tO@_BR;da?Lh83}zFBZB#<1hbN$mV@veu-YzpJv-DKncbl}`3ASI zwyU1K1AuVVM{yd58z4-F2m9=a;&^K!Sg7AjsTUJodjY6V*xghmNNm;RKB_PV-o9;|7hbR~G-yzP^CB&|oQis0qMGM`n=D zpf!-C60CFVr*$1bCs{h~dm4vPF_x)y;vP<@n6U3^O{=HX$!PjY!(XT*9&c!%ziWh( zkx4Ka1eHo898zpjY$Y-B1tHZ>1^1xU_k{Z1>=Wu{<#HEMP&+;!uLLY6%8DjTweQs7 zt+8O~%E68bOYf(dJ!PvPRZu71>D8~g5Uj(6Ybt>mRxk~h|v2eOo0 z^55v+FnLKpWprKmDqqa$>D3w0d1S4u#TVPObQl}!kFJ%aGe-ih_z!qnUe$#|m6L%q zP1CDWBJuvms>L(UNgD-wfaS@>M(D-38O_{`e$17yie8%ad;93uD}o!ASXrTD1hF4v zil-(i|2I}P0#b`t|D$-?@XC^sTu*J}R`|-_#1fPZjC5nS5e#rz7@$#JYLZ>hu2Zz?hM>GR+cP&g#TaXcR?8=w3Ke$^n--?%c^>LBZWgc&2hHMaUn_zAd)D0pb zE7D}%#$bwW$fHdFnP%0#(4`o`uQr=A$MHj*|sL{k(jTt_8T|b>F$PpOGTeDZvk0S0ym`GDdz2Jbz29QppX!Ip%Oh-qekel;Lcu6%c zdqDP&SLj2)xX~V8F6#}c5T=U`i?Ty&Vj2a_;cKT9$l9XXEraj_hj*vNtXJh~lm9Sjzrta~dN zqPC#0aTUq}%#s^BHLI8o1;%L(Qn{LHjc!xW?$|+}OOZyA<**LdKBVu$B%)ruDd>Jo z)?_E(pW$R(CjRT8N&D?-n0One1JsOw(Gz|sfR0!nw4T{w4(YcjtEo(2 zZ_J497h~IKhXTt{@cdp4#QjlgB{L{Ps1JH1&k!yPFlaNN^U(ksWnCt4MGj8JRcfUV zww>1dyS|asWvFqniQ$Br3Hv_$?sx3j(|L;;y$x#82_}W0RzP6^__QX)Ezk3F8_2Yv z1Hx`Wv&so`F!uRy1QKH(ee8E~gTyz#r+=ihu180|{kNY|t6!PjHn7l*Fgu$F24oF4 zpa$K2s#LLE+6^5WJ7E88jQKe9SZI6v8EL!37;_@*gim|CF($*upas4_#t^$E_=B~< zpU?eGUWnUa<_M%lTbITT7RTe5B|o{17j5k?{}ON%5Kyj8qBFQ_#HiHGaR+vKAFNbl9G^?`N}@wM3pMS2)-aabQP_SD2Sz&l-HSU0e(rwU#Tx%Y{RzF+^V%!+ujJkozQ9afZQ+xy<*riVGe39!~*Zf z?9}9lEiyt`W7Q$#XgeG+QTA&V$B!7(<^RwfmgWq929M(}Ec_W2{sYtnLDCdU>BA|i zgKi=X#GN1a%uAZGED{@-^2p-g+@QncgSRdSmdq>)N)=r7UMf09nVVI^lSGZ8liD+i z&DsZ1c!!%-ys-er7C>Xd!wYY^v|+`dzty6&{kZ0RDx1e^TCI(ZC?*&vsoYE?AhQ^f z0iKS%>IDpAC)BC3KC)PlFER&a3Zb1aM=;>s0x}GzW4A|DMt1;-AaJVe)a3egD0)=s zp1Q`so~Uce!=7%M{ISrEjmP9q;IwO+cqXTRy*;@oZ#U9FtukUCltnPB2`Y_9u#{9{ zWDUg?aQQ%1bYw&~0b7^5=AZ3r=&)5_8^BG+IpJc$KvB@SNmd&YZ~8z-E#&5v@OYyH zo97WI${?6jf?7!=bOGl+$}8^^C&uN3EF`YKc|hoFgZkK}(jf$mZD^;9~$Psi(_r#jd@^-eD{*G~N$-J^i(Zx;KXqgQ#Y^1_n-Uh-7v zmo$2`g<8W*qo~{q{~HGzG&VrF5;~)g6|esLJsPWPIy3L1462C7+uMUS&Snk4K&Qt} z3+EE@-6Q>NV%$>#T#RhwO3VyOPZ@ECJl2FCf<|!AKY*4VXF2ip}pVcuP`v?=Nvn4) zq()|+mG+2t#N7d6;Zx9`pwFjxR1OR)uT$37)2ZW zuD`j0O7^@KbsMCOx00)2y;(cuAY>Tx!%{p8gsZ8_fc(%;Wj*nW!ybI?qWm!w<0i|R z#aqdGW}WvES?83_DJun=v<+klbO1DpiUozj+vFB<*R<2hyuf2rd;Dszb&8YIF~~bO zD~n#{)d+l*1JTJbk3F``eLSPz@1PJRam$450qyY(;=x&Y)BnQ7{h2j0Lf8)p0er_=2v1@$E#!G4?7@(>=J}fDQyDy6r zLkYGmVrIaSS-z=6jnu@j5_zhXi3tT6dsMq)yXpRjL2t}6>gvL?=r2{qX(&8qxh*Sl z(7QyEi)n>?*`=rlWQAhs>;<)ogVQ$nrOdiTofYhq?38qYUT;xk9$6G=O=DjZG=jVd zj=MuuJ4>(Tw%Crv5Pa?=$pNNIsu*DY%Om!$AX8>tAGu6g! znQ0o9JbWR)9IiXAMhU}S^1!^rxc3)HEJ8r;Z&e%XoR2n7t~|!cJ({q!;`rzGa?89$ z2#d5bAcV^IS{Ruq(Sv+C>bInV-AAU&EP4Q__=jI#2!^r}S3c_Jjl~rn&l}5?=QzQG z_@qS}^RidsHcJ~MNP2uzSs_dIy%5|>CXPpd<=IJd1QO5Q<^Cg_ATi-^#y|hXeof&L z!{f2P%(4Rg`3hZU?m$DAvw>V9$)}HN28GqKE^()gR zIP{f$Ay^Prr|FdHipbruWs$qW3W4cFuRi)_6}d0?mg0c#w%7AW<1`&!TOvC}9a8j? zA3pZT^~(_DYR+obGhYTkL801wI}{*-{l{AuNv_V(?V!#xB{GzkZ4{jnTdvj}4onrT zV|1ya{K!hxnvmXzepC>wRAEtVWwc>tieC}9i9;Sup!{WXV~cW&)}*MX?vVr0cjf1( z0q;Z6_)#egjy~j37mjQRB{G!DMY`->5}T&+(<`HRy8+>51$ayJxcC3e@B4Qv`2KzG zhOblSCs6}78<5WkriY-qiG+S=8SR0&-4eADs$mySEtlVb=3&cL0)2;4GC3ww)vGjm zoq^J`ArI^$yCA(Fy-aV?)=$>^e?jHRdX;Uy7ICL*%3?u@497!xCyTUGo1kJ=k!)F{ zg03aIjKw#yOH$+I|qE{zM zR|)ZVwA|SnVyVT&>z?D&jKxIQn~br)XE!38z{cOLZD-_f%Iw?QUMjtKmX^0Ba)ajv z-)7Zq(P!Eu@TDKhnpKZPse)VnRjL8`R#1q~5Nz>Ay8m7>N${9X4Sfg^m=%ytUnk3o zY!vjzEfm!&EO+53&8o)0YT3S!`svTXDmz-9d&Wf&f9_|D`A{~Ri*`0MIr>IatX zw(v9Iw_$lkRipO!!%^+3dxFfMKP*TjG9F1y?5?Ngq+bUHYDW5v%4R#gW- zo&B=pTR)uRZN<;-uYIePTFK))da(^6atUTV0nar7{by{YMj_r^o;Q_-nK_CyFK8n{ z9dL6QI67Sx58L$%%jnOR^L&p!jLVO*FE^v95~X{epKl-6dDd~^uj56rbX{Y@+^o9U zD@_$xnD1QiCFed{4#9-`uuk8>4wE;26W3^;C}%0E-17#?gCVss1m@AgrZ6VOVajS} zivp-KamTp>Aa%hdzlN4& zdqzYSD#4A=jG#HPCief==hPu? zduSe~s;=0;@if5z2ihqhi-0z_Ho*?aQt8#VVO2VzzARiER~g+XxJO-sLM%N0-z)z4 z_s4(w-^j>-3)kYfjsB4BARA#fT?Ck0fY$(}3>0hOX`}Q}@VeKJXi}K`$ZqU)hF&r)l6Y`8!vdPGCyW`Y4J#*IXRF?_9X^;>OG z%R`^OYK#AT)+e)e1?ENqLn8FyZv$^Td%h{PQB9M29*ZonA(M03596EG&spL;jUJ~a1I1z%Xgbr7Btl#=DJ^mK$j zly`>Tj;^Nar|V8mc_^WZrJ{(`}+!1#Uzz z8G?C~&5>PVi`d}~6;|8ki?E(DUr;8vMRvli?dlVdZ~1URx1fn^R$i2+3fh$WL!r7- zQA?xC*Q{I^hRqej&M!^=)QRB5i`TQ2BPVe3ejc%OKoWbO1c+3uK4 zTBBE=Xnro;3+OD1E{(aX{S=rrw*j$TFIf{)7JWDRw6Ybn2fmP>b(?ktE(}=&Mg9Dj zzt;WFkACsv-~P9JDZwlzsKgg+n}<)kYObFhhkd~s#ZVD{{OT|MuYG=rw`JulxuhGu zW&)8xa?9LCutkvQD3L7x|D$r)F-SZ+W$r%02@(@uFRYsuYxQfq?;gpbYI*F}oVM|6 zng|9uy&H*yfrxG>1+D{@%sT0y@JLKsd|CAA`1W|<4T)I#b{!vHGKC%}I9~Tl==v&}jps;)UV6Lc zovYvOndh7r!eyQuLdR2c$7N`zu1$K3&}464AJfJO|sQgZ$Q2~_rh#56#GLSeK8pA+NZ7x99MC8#~$kNSIvD7(SBr3E(U+A-YdK>{{v+; z7!@A*C#YqUs6v}>+B$;CB&ZC~S*22#L7yIIC*1CLelA2F7D@VIEKO#XRR&&D3#3r;DZlR1^21n{N^|FF+!0l+abh$tqRp0 zA;j+IE`)xqEczUEGsHwa1a_Ybg3CapawP%_JMa@)Vf}Z-D$fNw!?xvxKRM!Lw97eR zhrh1e&V!;)9mGy<9CTi3`(m(r!c1w3|76jIkZxv3jmg>}59AN4pPn0(OPi);(b+RU z*IXk{Vd&F9*2XqRT8h^yGxH#tqfN)C$6Sq_ zQ9oo12RUJfzbn!oR{tevs@3j%r}RI6PHo_EKo(lWMmVKq1Ot1RA|fFtXg&%^A5vAz z4@F)M2evOkmH&SEDz9{4GDB6NF7YDjHhG2aqi+W%O1s3LPg(1o8h%MN7>DX<+ho0O zE|;{$Z&XytPsh8$-i-z)r*S;mpe{JZ3CD!bcM@;g?}u47l08s2___CXDK-(eD4~os zY!lN@_oJk$nZgOBJs`5C(j`jz{N44`?H&;KNw6EhO>a3(0)Hdd%O81t7GO0dKR7kv zG?l|+W3tP}n3NC07IaGydB-E)YXYofOe)$T?eRW@q`RA zwhMGhjdRNg{4ZV53`8>}5^P>RMs17Q?w2(=N7+kWgTfojPH&60o8AiyGFV)b!t|2) zk&pkjCO{vwoXC$}1A9W30L@uY*+R_n&J{poH;v!Un{fF}D}X*Q#y3)Blc=LMbFz}%auUN5ZMED`A`_OZ)%nVdy~>% zL$~*aPn&PIuOTp>>|hp2F3NAshSD~RGSOa*aqf^umw3zEwZd*r zu+PZL_5Iy^wD@JG3U3U26wskq>0=@>^^1)`mQFgiESRoR%pB7y-)wfteC-7YZ-wKgbBViS2+-b3n8gteXNB25b8yfong z)Sb|w)W8yGV@vn!kYd3<5Nk+uxTLvcp&5Oe9R<&5%Zxhl44z=;-!Xn8Df*w{$UoH~v(q&7!lBU~&j*9g%>mhna~5MW1o++!v!m zVrJYj=f*7s`u0VV^&x#!Lul~~r1&2nsB?rl;@y4N9x?Lwn8JSh4GouK6CUG(WnWgP zHIXgKV!44F^u82xRd`X}W|lVURW6rcI3-uICfr0~lS+f4*>i{LI z@v<%U<7LiNqwp9$EY&Cma+NX*I)}=II`m^crtWxmdKE#6%0)BHVKN*91&0w`bLmfM zqsSSWN&aT8Kx1uW{B0A#fRnp{NWgwGi%8nSsim57xejTgGlPzbRtZaFRWl%DgQlaK zMiO7hN=UEH!Te}$eYR(=78`~#N7z=T_Fy_Y?pS*+lAP=e8s;fs!EQLkq%JfE0GL( zrwZyr^CK~4QXE_yY=9B6CSy!ruSTc01$y1}P(6X1gh;SbKPNRDug#MUD2$;-h_c)e zl}Pq#Q<%j-RhT{FizsKTkYitoXY7k(2b}o!h>pIapKirc(g#ykQF)Unqs=7jAQ<3Z zD@Kj>GgFM}e8}7nL@yClYCeYQ^sP}Dh*V|y6L8Q_bG$d zW_f!&MBu0G2^j!-o1&;go_TRiWcNJFbV1jY9~PZ!duxjE!|J;g;Q6|V?sEl&2I}eo|d+lGQSOH<08*!1!9W8;w zU%!ijl=X-zw^D-HNR`0)GCe>ol%T>*6NxS|e44fwD4~*|)Z%RP!vF~Dl?5UJY!`)Zi~hcG9u`*VgYrF# zgx9HDNmJbAh=rnB;a;pq#vrn#mJqmUz(vo4Sy!r&!Hd%h4@l&fU%?%Ta>GW+hK!*Ib)d9E)s2*`>f7$jD}+d={W zm6nAaqJ}*9>!ED4S*p?qW)(pt6A9OZy&&jose%L_#m%Jy>{H~(?6cqPzBBBzKZ=7o zu_z#;{WU8hf`4J0Pp#(`h~e=j2^yM4xH?-22AbhE5eY|O&uTS|IE#9Jiu;ODyOdbMwHctNImCmrgrkUGXJEFRcIg>2o>%j63)KU`mM}w zU7R=kvpxPGb)P!@&hJ|0ZGU^^H=&jBhhP3<={GLSd+f37>-BH1dl#D<^T?dIojz%F zdwg4X8l4x?M1Hj3qxN{DeAzZcO8qElsSr(6$63tH`ijOd{BD^HZP`<_-hgAgv0$Qeys26y7xP+vd! zUzeLwMHaoaQ?u85{*xDN@paPs6!NB9j8mfI3}wCom@>#dVWkW%E(tatX_BF=F_P;S zyGpHd%` z68X?~Ta*qXWqBcZa~)J0rO~}KDs$keW@UrOd>XsFaKM^?HNHT04hjYkeZUX9uR@C4 z7G-60zbY4me06Q=6y^d*YFK3FhCF&BOp1KbRbd@OLwacp&aHfV&|5d?1FfearfEIq zzX4i*vdlvPmqSg!LPpK1qbf8U^YF)0Rnq{L^@0G!eS3e1BX!TdKN~6Fv*<6OViVW7|rqzh}jiEZ`RtsAg^<93Gcu-n2p6C4%WBsPn_^ zMhzq;qbzjqSN!kH#tIrUXDNykb<@~$f|>x8(O;@c<*9y^eyDZ@D=&7uT3S-{>g*Xq z5k_Dl=?C6-3ky8N{t_T(W^dm z1F#>pL3-1zn=ogl{spt=)M<9RGs-@GVsH8PKl8ELqMvk9^QjDOwur|o-wqpFR6sC5 z>YfW-C^4VR*c|(KhACui@Onjy`pDFLd1ZXs)CMRPJC$%z{^u)n8E~Jb1@)08RikK7 zyo*YX>3!3!Sh7s`*bwCuSRA@JXR=8M(6FSR^LHy<1{?`!BMSVdt>%a!Jd#)0XJSV`W^~xR<_4q zlGnr{feogi_3Bb`m1y{eWZ!%AZ(I+Ru_5D`b;F_A=5i6ct@(M~udD6jKfGOlSQ1aD zT+;%%gh+b(yDbZh2@1 z)2;ea)dh82b$~JhSSZ|puZrt~uBQg5v(l><%S}`FXV)E|R&eu!`5WV;pkQ=FSw#-PtRtvQ zA_3LbJH56?f!qwRe(L(8cgJ?qSA~gjrJ52+qGTlyPcMx>4Y6~Fk>KdeuwKUzO046! z`h;J{hktRfVdX3otF*^AORr1dff8pm1+p)s2En~IS5q~y{le4A8)41jBZ9-KV<4NV zSLaLenTMja!UF9_k##NL3GFWL7;VIdBhw`ru46j1G3L%f!t0oP_VZZ;`%|C#5FPXVZ%S$ zpzbP*J{8a`Misx`HHuQ0wNpL^C*+X`nwDfOknKazRLR@spmI?kxhUu}`e0;V#3NC) zyinFaxo0UI17Pf(Z?^#Gn>kU(Lal(Q=q+DF8Ac28^Vnb=l^i$tja19?zo^f{}s@GE0~3UW)~{fFHCQQ99m%Cx5isfg{4QWF)%x{T(Cl+ zLuT#6)MFJiDwRkwgIqeb4o!!H^Y74DuDpnS{znU^(OWN0nP#;n+LYvaY9o&WwZPOq zB0#o{U`hy3-Ay>-d*9~)Y+o4VzB%b@sez^`9;FZxhu zA9)paL5Uy;yaQ?rEXV6KE0_VuALoSJCKpOl!!7rmpbUaTp)H;Rev8A_IjQYEVmd~? zbQ&^%p^@Xc>NPjWeCE4#z&^zGQl$-90%be;_0rWfqcNiE+sM`5)r(%aHs zQt2<5UkkM=BUYz;f`K;c91s?nem7zf(0b@nVe4{%0;+DGLGNC&GaUOL_3E9Hosu%p z2g!#-Ggc{J9i)prRhFq76GT|Xa^+!mh?EvB+-}d){fb0USk$<1=Ysj?-GLXOWEKlv zSNh;i2H#_n_@Osy&bWMU^W5p=h@0n*<;s)(GaufxR|eT+^PS8jm<)nSg)(RVO^Oah z?bH3nSMK#SO*DhW1`h2#m}6vomAqCXPfm7n!hS`>K6|mnSHy>5k!HlkkhGvA!BX#PK{*MGV>e#w7h`+S&4^Bj z_ih$~Me(TCB!xlIE*7;l`gTKQ9CiYw3XGz|QOl<83#p%u4ZhA}h-H$Ug%-=OP9H8R zE-6%5*RH&&|L<2+)+@7iRod95LV^KTEssbzA9exAm60uIP`ElAs1rq-m@3r(lxi6i zrl=h1(X4Iqt>jjpBbp7G%hMN2b`#mu#aUJU)$&2lR$sRP#5$8>#);ENc>Vx&a|LTW zhNE&{{`^BKk;hRv=_%*4*RD@W{iL%elVShi`XtBN{_W|oM|pl)@qGFfTWp+4;PS~R=qbPN{g zd)nRT=C|Bnv36N(oIM>7Z(+eg@N?JuQ%N#(4R4|b>CW(@P_KmD7|4X0Eig^%5@WSM zW!zclN{Trr$h|j$^>> z({A$t>=+m=_}4nArqZ6klFO!+$3z({)Rx$gQxlL9nIEaEknIa;v38y!mMr@4VyS zxiP?GT=p;A?A!N^nP0beOkWWPgvE_*Q|q-?h5hsrZL;LP_<)(XAHrvdDin-yGivNt zHOEn8f2Vuj`@a(!_Die(vFk@ZO6mr;06LF1y2&=os}BfffS~ST8GTFmF7I|#va(j8 ztBE}cY+HNY&~5fH%>xe@zVz_n|fTG<5wme1R0`?DaNSfpa}C=l@U`5}a-FIPRiPv%>7d~V^qyq3mN@SQ*EY!9$x$0bXYB>G^x(k$H<`y61lH)D;>NmFd ze(Y;S-}f3fo~71sLm!W~Mddb5{APjy;n|Hu!fyEu+EN4AO?S|P-etl*vRi=j*e&Sv z(qU51FtaCOK+!C%h%F*hn9o(%USA?Hk#6sZb=*b`JM!bQ&`k*a{=xKDG32%*?j9(o zg74exxmk{JH#30>Ru`1byDChB6${l>p+$O;WZZ)J+%e$Z^=Wu?c5g?SvRiN8dCpr0 zn588iqaw&7Y8IAi6QBpY)~A`wh;Altd7>hsUY+YTo*oM)7!1$RIQ(RX!Dvypk`gSX`9*YQ z(EQ;G=ND@8$ooK7;a=2m!invw;a0S8!U^#&J;ncOAD!hbPFUitm*gm~+Y@n3P$z8^ z3`JDUsE)iG4i1HJ8tQW!X1W&$_Or|`w&!juM7XsG_NumHxxAi^`cV&;{qNIn0aDK+zJ-!;Ov+Y z{&x3!_S7a^qAEPzJ+ly@Sky{-NE4~62sy4!m0313hc!275dC=n{Rcq1cFXhAz>LX_ zvOKb#(WQ&}<*12TC%%Q}-3X+yz~>2$+?e$_yG;J$zpjE-3)O5P2@#VP<%$^@F?lm? zi?9azinNJ3A?{Y^gVszQc}5K4wu>YNa-nDs2-v{27IZy|07Wye*sIA;0B0RPzhixJvy9lP501CDQY&6~#wn5XZ?5APz*{ePH zj#sg_I(BW``-48K zywa(pNXUjp>NZDaD#2?8Mq^jm9;c!6)aY1Wxduz8JAUB-B-DIY$8J3d@9|1TP4Yta1L z;a1a}JuZ?{hg9P{d$fK)*a)QuH_CfJhDLb5WdF|ZigcRvLlXk-Y$ z+QTjgW?!a@p;fLv5;TLd!@Xve`pFQxRr|Q6YcK>mdpap*o_~G9sqy<~?^Q#>P z)2PkdmL?uEfi~KF4Jru+yx(m^LYwacYT4}6*Pi^0-$bl>r|A2SJ@iD%-&DN&=5}ASAG;NQnm&DYY zabBM6v1+AHFZoAgF|GF>1mAU|Ki*B8?<5j61fo#GK;q9;!;zUVO81kbukf&t*(W|Ml=*J%AG0Aj4{A?i0WANu{ z%JQ>MWcgY7>#a*c_3T1xJ@Q=qjCzt2sEC_wMHlB=K{c;FE|R*;Z58A1 z`b-VkK=3xffQH=-BB7s7Van!Rp$`II6E+N)C1Gno4Q2-%TtwUjV*D20I&lh={#xS5kMP%}LM> z#bHk!>Nyr_i^wWZP^Q7eIeyB1O&!@s)lWx`-P_T&FdiR-vO;&{2^*xAuDGysjumrx z^bbp@Y91T1Pi#2kkETTeF;B5llxUt-2(Xl@zt0Y?>1%6>xmE3loZ62;eZ`_6qhd!PC zB5t7S-v6ZA|rZ z-ao8Km6fB?wzJS0F}bcq*@%IieEI43lrx>7Q06giSSXYSyjKSmO0b-wFQg(&*A4|# z_+pFpivJE_z5w}*(mgw*x<*kUiBgMT7>$2#FzI#H3Jc_`F@1_UoM$l{p1vEGH zJ-GxhPJ&~!+oQT(oBk?RT;1}-l%`&NK&Wd|_D3`dpmn+ps7|4ShRm8g|hh#{jdyK3XEuMW* zzHjQP$&GHHMdgmCk$&rw*DeRw58g}o-B+y6OUHZHi>RteRIAOZe~@4x5LZVe3<|5E zuIY1n-CK`6^8N1wJoe}lWrtdR*2Ln57yZga$=(>GUV1d^EcMvKq}mho*uxlp@{PO@ zeEcWnX+d8G^pOu>jN`$NJ+h|v!j?GYwG&}^A%}!HK{!g&EECB1Y+{ax4^!Amm(Lt@ z+S43E`2Wc>f9?xsCxrd`JW3Rrzp>hehrk$Xh0$1Ko_V}~XOVfv9)A_fpU99w{IT7Y#^v?B4H=w58C5b`Q0=tTt3}-H<0UO?eQ1s3c+e>xd8Xz zJ4Dy1tpdY1)d!DZ&K;OB=P%)A-u8W!{B4vKAyUtIo2i}LcGCQ{kJL#UJT(xEk)ZYv z313ZrZ2Lg6eER*1`f;9^bie;ifU`Me0#zP*KqD9c&%8acGFIN?5 zvnJcdf!qW-M}YavH6xK~8XEZZWGm>d{jBrPR5CZv@z`z|Z0uGx!9bRH4YX^58p&zc zvfrn!0X2Ldc`fR+a)9ckK@B=4NAa=bQ2_SGId5i0KIP}X=7f#s59QLc$3n%Q{$DFL zzI%fBI`s)RZ1C71-L}ET6@r0mYbVrafQ85j`EZ{UNXpNIn-n=Af7~bCqP_m6&U|!{ zq9yvOCF^9DBV_{L7xoKDXnoDl3?`< zQy3_A8uY<8H34|Se5g%rQeZblQz+{5=`KR=aH`;FOfQLJVjS@n2?VaCl1+?Z<|^Nj zABrdHgoEA<-WfvUG_%cY6m-xfk~Z~&Q6o*9Q1*hH8=8f?9#|~5aR;{TUEaLigZQr-5G|dcLKIVCIug+N+l`G zV~>{b)4**ayMW_vuQ zdHDC@^5viSt>7PQqu#R??!4S_%<>?%|(?T$6TrnPNccMR~iI{y5} z@!{j|#pU_)*ZAK4*;^?uyGmz{z@S<5`g$rg4n;=v>eB(mLZdb>77qx$tug1NsJ*z+KQ$cn0dNd=85hTa{AAUzec>j2a{@9^ z{ww{j?eW4Qo(kDq@`9jXI>=#$8ogFfMZ$7_Jv6?aQv!STJ?(tN6t$LYov8z3AYm&NV2M`-5ZAqLkp#{V>VtB`x5PjZQ10I)z7V?Zbv)aqhElvy>bReM_|HAU z5iz#Q(F@)^_k&+>k{vOl*uDviGauHy?80fdc?D*OQcW53i>`>*)d|yrN;OcNL)Ho* zIO>-ONgdr5a-TRc?tRQ78OXYDtnu$5%}i&T{Tt_6XbpjmWY9aw`?|P9HvH=Sb~SdF z4*H~vz-mP|LMwP3iI)roJ4-9!MgYA4BF{g~4L~7(`}IdJ`^I@YSroL0N|a=PxLzyN zw&;-C9J*3E6ni!0P_zjF{|?3a*hIJaQO`^WS8O~pfSV3*!p6iezWIZn+K1hE8&qKl zq-~G77maWL6+A55*CA;hW>DVmQ|<{tUuA$gscF?<+PQ(;s4;q-pSw|Go7Q#9TyRB* zb;WVZ?=o^wZU|ZYm)Gvwmwa=H4e&T%&!Tv_zUuO&B0^)St=y;WgG!Zs zQ6^Qd5*meY6NGHMgWebA7o>ZjpmMLqI2Q{r`(jMMIp43^5}Y-8T$D99B4NCY%5wsS zZ%zJ+zr6+F?Ha>k6B+{_1(@|!F3~2si9$(VEMkIMfM%#dY7Erj(fmjgcp)W{aWV_s zk>DmHU^gtc^<{2*o!76eo8;xmsj*Zqk2lFHZ3^HX5zHV#-6s-Slt|8eLpY!SRihKI zZ0Oa?B6Fdzr0mIS1Gz7xL3_w|3-lo-N^;{`rl8!O37W3M3#O-wu;8tp`q}Gl<-9@fM0m$r@_0`8o3N_s2gz|(`~@#wSxW8VRvpD-|K^NM zNaqN_93rUwM8co1&~@O_w8uA)dGeh;8~ry&cZu^o8_4S5`PI^9Rkq-<>ai+?>5prV z-wsSQ`$Yx*Ey`x`O|n?!7Ej81U~K{)vsz5Z4cfH!B!xl+73{y5+Q8#A4w&9XRJ)ZC47fx^L;^ZCdbLU1BG#cNoEd~|bhm+9Mt4=70>Rx} zV3lsvWB{EfaG>P4X7%tGT`O-gq<1`FL z8-N`eqeOfsu1Glk6UGV`?+5qJQ|m@cD)BezP62`E2nZPn1`JEVa8e0U1m4Z!Zb7dU zwU^8luem{eAm?>kR4m`1Xi-`i%Pn}TRhkq#gblE+AZe$&U}K#!XV5d+th?$MJ9>&r z^6Ks7>$wDHc$}DI5xa*l-)G*ZW6w;fq|NCB3v0(X?YKrxsAFImdB^y?{&?dbE=a7# z!b2`ZDG;R!U8_MdLZI* zgoRqHh}L0h`Y?4(p6`_#iCjOIf>VS#b5|IU1AHOHw80iFD&)EwL7b5@3?lofcJX_5 zf(Cs>6Y)qCHIxaIn(<-v`C zS#)u1YHaIU?eRz@*CjpLf!&7k3g z6r$n4e=oFWzvQxc;c?`Zg)8$$SO-%qG>MUtw~(xrRmE)7oS{y~9tNlOGw&PF2!!vD zNA(zUUU&>%BNKjc@FB$}MSHwy+M&>H`T>Mb@cxl^Y$b=UZ&PMLQ@cs5S8tZ53WmqW z*3^5HvzT-Obx-j(+N&p)W^6C&T0bz@vtRPo;;`^pA}v9=aH+Rm{Xno?+D+HQoM+Z4 zuK1U|o+`@@MafUBjZ6*5oSq>vdTsD*6s-~sl8Z!J6{#{~_#MTTxod@O@uiY=f+jNG zKTUWVB63gfaCYYH=f!16upijT_uTAH`C#F(+18!%Pd=HmjVkByP8nFkM}*WG38s#q z_7Vw}0$R-HZ31qL3?Ya+r^@=|8;F)~Kq?Re_ZD%SjOg09YyO=xQKVXzOefR1bRRS* z>x1q^>nefgqE|UcZ=rIa8!BBiNP`|1QYvTBYkk_8qkekza-uJ$M_VF8mc%~!Cc$Ot z6{VB9DK5j%C>F&T7;(bTgyZjiEyS62W*%>}S=yOv0`ex`i8f9f^0*US7oHh(Sc?p) z`H_}%v#Tx5i)PYc6uoF57oYz^qI~$8)wTrx!Z@E=|H@>QYiumhR)Q%cs7*uyvNEon z{y7aY7J~D0k3!KxdpsUnBsn8378D8Rx6L^xEebj|qltVFvCywN*nFr=yPDbz8RI6I zh1b!I_;DC2qZkx-`pv4sx4xWWg^Foz#6>E161Cgr8J7|awCNNMCs5Kwy})$Sr_50x zG5-RLXmvxm>s3%#?53OihcpXX#Fc>^ihCg`uP?AD=UEt)EYVm`H`u%@&X{>f-X5L2`R1uHe+k-X+uZCbi^?WChkSQxx zZx6!Y{mz*{#dHBk*C7GeO6H5Y=|>TFBl08ng)RkYKRDQ;+!xvlI!sWKiEaHu9vI_C zBGf@|R29|rkcHY}=pN~lXMr{yxVFms)S*zIhjU@9#yKcFv+lS61ePH<*}jP-*V8V% zV+GDbVZjfmI&QuTkCVz5Y+!bhU_j^a2$7(ll?b&h%_`jvVW+f3i8Y!X3KR)|`UI4{ zM!>Z}ml@*bEVUzU&Gc^hU9UXyCPW9#+J|sbTQJ=KNlM+vR5?`3L49R-w|aQo%IIzy zDF#t={puT_cY~j>SqQXs;xeL7jFsb;{5QgF0B_kDMW1-t>`h}(rF1d;p24H_0B2Yl z>fTMFtbnTU$Ui|Xy5%Uv_$#n~d^chrBJSM3xy;2y$&Y%p#&> z$YrWF)=eJOsLz|zT#q_}dk%6lHSex{YojyX439&=EZ$55*&d%wVeqa$dY!CU)gs>J z+ZjF(1H|;23aG2QOxJ3T$Pa-spye9uf~_OlqV`FbPzAyy<%S^lqKp%+Y%PR)0A(F@ zuUD>zGa|-3UX)lwjB)+xrZVTGfuL@)YH`Q~pgA>qwJ5hj0^lx?G3(V=g-s!N=1%mu ziIQ+mQjXV#>0$Fn{r33a>mH6`neoq{TLg;p4?%H$A-R-VNIX_m%o`N8$9IXB1#A<3 zDmw-npkA`puR5?)uu0)wL-6F=%-te9xz(*NIIT2<_v*#-?1@@lk$p1@g$puwpcMCZ z>5kwm`WB=F!M!~WxoM6mW#n(;MCVN?h`t?e-mtT)@F&H-x31 zu~^b2KK1rRxj}I6&1+GsseWNw6uuuM?*tTR4~iNEg^~=vEBwS-a_vDZS75 z0mA&PcN`UNHWGHj|AhAQv`58DRY=P+_a-38z|h$uNkvFHlBCxJU|n3TuyV>caAw#( zWQXV8_BOY_V|P84$V^G3)mGg5ZqPAm5jR`G8RR#8v z*brYCZGlUso42W>PO)Qf8TD4T9u&7$&$DOnI&Skc>j(x|CU;rr)=cmmQ4K|RNsdYN zA~HD^#3-|XkFSu#=I%Sudi9ckjw$9&Ok^CmufqR>fi?iW@Y>*f`O?|Ao>-|NNVmXpaT6BF%!@Z! zZn+H6wZ|&R;Ar(~Bzl1)P2y|=SlmmIsjnTQ4#|oIdo^q5yQ-z}C&HXJm@v)GK#1LT zI*A{+{a^eWK6}}Ff0hMw2BZlu%G=@-e}ZarC~4O>ca_AX$na^F8bFwCP<}MV^4hI- zV&rsio$!(Oy7M)sMUp7{-MsXdoy%bHS1-!{a)oY>Zxj_kb9uA4M3^MdL!fo_yBFmj zU7`0=jiPl5fqhdPfJB$f{KUcptUMA{z;jYC3o<xyt5wuffg3Ur~4$6}aC{n|( zsM;~$gBR(wx+G)iu~7)+!E8YmF?9!ma=VnL(|2@&{D~_5;hr=8Tk(KuCjoFzVQ6EHYVFdq}tVIl#;b$WHR3MjV}$x5hn zY=OBqDUJYKiiB7E^CjElTjqAq-Spw84yJaNNl_TNC8|JX^!iku1Fi3<%6g2-2B$YW zv{{@gE04MmY#{MSv$z*>MZ5hDi9mDJdy`;GoAgh=vcl=tEkBt_ea_>OrWH1jR$Wow9$eyt+s#JcAfPLs)Ch zJvNaQ{w$?di)@ zAC%~2(dz^YC56^8mV^!k>rkt}g3p80C%!1r33d%D`CbvyLi*AlC-__8*Y@L@_o-|i z!>`uHaupLy5kYMx5|Abg8x za}c^CvuHT~rJ^nzNiSE-!J^)}QDst}{W5bu;kxS&dKS z>y`Fw|6HsFkGEGWIsnM)ohRF^-VXBdS#(cSKj^}@CGluseq? zBN-Ofxh9|Lz>JRtQSR!k9Qo@mE&0~}$KIE~HI-#~`^5K1E{1Fbk~B~t0c24T3?*pA zs_w4to?fSWmYV5xx>}~E?X7xxGJapxbhY~qf`SWZ0NDf)1>8^xiy)ROC~6c@K_a-| z0u)lH@ST%Hi$tP%kTB8y%3sOLTW&D-pL5SW%m4rOXgcDZxR%aJM?98b_q*pa4}3eP zKb5VN)X~QvI=u0n`iKGP(=iYI*^G;!lQss&lh0NsjivzCiE&}202k#wtH56^qf7l8 zibEamJs{MMMCn*oohtr}e+mfGv6Kh>kKJzV%C4A=G3rWs zw_CP>HDl1FkL-3U5MrkC zu{oeXh+`Tb9CT@rpn5}cK)n3a#Ohe2e8G3C4lH*=al9t+1J^+pydOi)pSkXg0!evt zkwFIue>*G*(njcfmPiI&%4ZkwcSdPB+n5~hL6?4ygYb(M2)ka0?y*DCP`Abb$3uY1 z%Apx|`ltT0Ef*`N-5Xn3esYwJk`_`KqhnT2-Xpvi0CUdhr>&l>QS6#^GG?szHa6p8 z$efMEAuC)^Kjga3v6RJoLmXWzB6?SW4i&A5QT@eGZ>Wvf7SO;Oke16)0wxy{3~i7Z zuM90_j=DlHwPH-<&uqriP_r@yfV%jDGLP3is+Q$Z%-rxcsC%?U<x`CG6u8{=2Ig!j6c+?bEm7ye9EII$Z8QTQQ^?wcri3PsjoRw$iL<6~Y2 zbJCk#SIhQG)Ll?EoJ${*Y9S?<>3%S9ECIsG0+`R9Ltn9j#+3Axq5C+Nz+hED=};(G49*T21=XQ>bQd|S zRHJ%kn+&%VXe^L`>Hj0cfDu;67&mRY?y#j+0~;^YiIX%|R2(whyJb08fmH3$2ePW^ zbT@Rp7&IG_I9H`@iU*`5e1MCL2%DI)5U9`|;A*CH%MQ;#4LS$g9#-qnP;b=1X`H9r z!PXy|oQ@2t|2yO~yOuL24y@fZb5X8R@+OMtsJNY?o9^2^27~TLCkI>=-I}&dd_&es zQ2RrtLam29dI4Xj`pornX_M@vPu^R2Ka}zEO1uYyZq6)>tl{FB0qNnObUKet2Jy}+ z=8UplvXD8hyy006_iB_Id^>>g7VkU!o^H0WYcN<8ds2VM60RLo3~=!Rhm~4_qg{1Z)E5%=k`V~;J4Uo-5Mo;hA|tX!p-xCSpkUp z#_#;1TurN4-P@I)k|j>Oh2CbiJ|$D~B#I=URA#y`M+6q6h2;4|9pg%HVw%KASXxMp ziAMbM&mInJhsDsR9DmFT7US3Mnfnv52^NBffzz zPiA%cY7`ggThg<19k~v5{GD{+oNbU_-W{Cdoj$9M{&+33h(fWvPT#fmG9@paecJ&w z#H74*MEhLieP>(g>mHHKR$>gkQDKO`BU3F@ofue}#->VX=tS5$%|o9Vi(__nWAo|M zdzQ-yn=P~xS1egseDpL{DxFePikD744qi(kb<|aFV2U5}rN{QgXDom-^gWKgyC7}6 zX*pT4JiL|+y0F8Q)AGX;SUrY>_19AJH56G%#bK7G$ZMf%Cynf)&@dn1xUJ04PulX> z2D4)M$e6s|mN-)~!Gs8v{!ePMZ~{p;`zVtrc>+aNQ*qg_*nJ(U>{n57ODF91 zNRm~OMIOs0RMMT(yJloW;+czN1GLy2=IwNAlbsO)M*=5-TSsTptpq7oN>y2J#+A6oDyoIb!m_oi94b07^2_EWL=Jb=et+QIwy{F>^8F`+bB7#jax9%fvLU< zL6HFVEDQJvOe>8?TSf~jwC7s2KNFF=DIC0Ln)(kNFDfu3X6jN~z2*~Kb#pd{-Sh>PguJkfNcho5%3?dGB`dM)O-onL zL8;Z!=o{YHSDXs`+07yyXMx)&wJy!_be6H?|w}Ll1>37VOu@BNbB@&S%U-#3XBbZMv{TjC_`f)mc!>Lw*u1u z&{sg|ehWzG0aUc~f*E>1kxssVpDxrWsseX}lo@VNqwd}S_+z)dyQODHt+?N_j)5S* zvG8eeIOKt0DBKLH7rE;Fwz!@J#zJIC&?t`jmxO~cs&}vAcEZ?L)L^4OXJ@dyV18aa zVH$D{#gPR2w7dBd&924-i}^0)Hn`%x(q z2VK33hqxAEh&?H@PnT_c={VaKxPNp>^gwBB&r`xCxx$7!} zuM&x=uLPEJ)mmP>{47Y0s?&uPf^=FV1|I;KjpDs>X#@91y#&8_TfMV9G>YVaW;gY! ziEx{uf+_a{-9XrCJ*G2+$D|LuA4svWvn;CA-!N8Q*w#tuQ2MGie7|`OkhVq1TXiBaBsE6%Wm8Y?(Cud`4;=gMEXro}95^AC@aWlp@9Wt%N#8z}jD ziX=iqnW>R17jTQUl-WuYgX)gP1XvMOb&WGsOApE24UA?PEWlU(BiFrB5U!!Gh;Wi z>MKip^XP7ly2i6h(jvGi!yJG{(HDKwedVl`p1ran=~+)rKzwAGP!E>yYETCO@ppT! zrnbZ86_yMyj}j_FA7=%dag87&;X%J+=iuFCD-^ z7jM@dnQGbVW2L2L0sj%%sM@G{UeJsjB#<$JzbknOOl{~E@d`J)e(7HEZ?@$#{qo!R zioxu7nbmjFc|%)&I2IFO$$4vqEVe+18t&?5L2^KrM;^3#FW@)2{cn8FY=VRxr`pz` zSV6*h{m}#eu`yj`ay#^qZ>5q$PV9DEGIKjlQ1W9Gsm0DfEsey+i}?F|mik-Us(8jwH7$n#aFA$+% zOq=Xb$WC%7qnKt~#ia&gxaJu-Z99$JT52t)73@nD(97aJ*m!c@LFE(Aol&i{ zA@!531Zq&yDL~3sEiY499;L=z585Ee5TOOT19YS1en@%^@!lGeF03J20>SQJ5U4a} zQTT2*NMv>ju%gz;b(}+2iaSR_6E{kD*lK%NLx-KM`RKzFO9D(_DjfV^AIX?Nv}SX* zo00>mL;)3t5xCRhEs8YRX>v8R4thb`n7-(TqRX?lhU}p8+}c%LG}27h^A5&n1vlM~ z#Vp`!1&ylwZ>;*(7e6TeM*O?`e|27Uis;3MLYDe!1Ch9jUy%>$7E;Ys(z`aK{YAEU8|26-!nQDUn;FMb;CojX-vMELSOa zYnLVh;h)3M@my7J4?i+^twxs!HP4OLVFu z%!QD&fV-T<{GFaj+%rmtLBagnYOwD>BSn7mhmX|%^^4#C^3T5&Eu-X%DY9T#v*AF) z@xm|0&XNSG7VVm4vLr!^SMMMPoY<0FFta4bC^=NhRa0?V-ileRyalk@qy?M+qPIl; zV*W0EqiQ`hKaHB7NOgW~+24Kfv!n2Bo}cmMYgn0FHUwSdmb-0<-OpqS(}VlHjzCe| z=hN$@jjH+xWB@!GrIRk=_D(}##THP7Ns%Q8ac{jj@DWLvl?qJ+=ls`s)I+loN?DB< ze5189RxtPc^a~iV@dP_4`Tp@||NMIsl-}D}>q~AA*G1yQA-NT1e7F6S{1HWZAS4&H zH6$Z?eGG{0REMChw3hA)E(wR&FH&qk`$E7K!4*LP9|J9eF0IpdMh!?CB-x_no_ey} z^K-ZEXe`d{a@YA7NH>g1tp)`&lp;C^rCsViad-3=fopu81!Iiw^!#1=J+?#6q>9N@ zc0s*nJ)JLtwk(uT!w$$3BwlX1QwIy~xc}h;;MP>DN#^C?qm1=WP@Cn!a@_m+Wz3J9gI0 zj@3}|DhfhPaXaWbP&2z1T*x^P-XTE+7>y!#(jK8M8qe(K=zKN^OMF{-xtz6}(yv|< zIG|3oi(|aMPS{23n0`)?!1x+loQed!(OW_qpr<~O+xqS{x?Pzm zTsf;zmGb?SKRmA*vF&K9gcNPme1_;+#FgS(I$OJ7eN`!Bho+{$19_s zigHCzwj21sbu{zls3|k7J0qPhc#hqaIc=rqhYQwFQ%zVZ3;RFI$qpwrNg%Q~q?`H> zC5K(xUMlX0d<}OozfGFyJ{UCgxKVXpl_J|K+UeOO*YlERuBDJUEFFN^He)bIr$Pp@ z6j>e63&ndR$jhiNpKVk<{pNOR(^qfJS@`4IKR6HllKb2qffnd!fNYL6JHy(1@KL;C z2kY^lW*q*i2|Y{K|NRG~!imw-Y=)jQlpGkRKBMB&0=npAPNSj|2&C$zpy-O-<3$6+h+bAlpa)7DvQ3^%wlfWqCNX+9>P)hZ+a&G{ zj-RYXd5L6YQv@Wngx%9Miamk-D0|iKktBTQ{Q8{5c+rf1LlY-dD0_sd^AUq(hRcCi4oOkhNz>I{0K!VskoanlcMW^^-yh0o3_*C@?GK0Gm|+; zmDwz~!sww^15==l3XQTNbdNuH>3OhSYgd-L>7}`@c;YEHf!oUa94af20}tC{uxYaq zmTx2yDWtGm4ga5K(^>vrM0!>Dy z_otF&WZMMz-VXIGN+>yKkQJh$Gc=k(ThkfZz$~)Q2kHYWd38|1a0U+#ho*}4pdD1j zY>Y{ST81Z~Dcr44T8e@%Y7}ayV%Bh46}J@y!gB#?RJhY95;;4>g-o)fC)$4Q#7IE% zqG5l2#;*8zLqM}kur5pXY-LIcy+k&=F?n=o-5ugr=TdSIx5+^4cgu>RdZUev+lJ)@ zUt-ly#Z2{a=@sbkZIGnMaK%B5rR^SPX^@-Ff%aYC$$(x@X`k4Eh16_~b{M7PPXPt?A)ZRqFwl4DyS@sh_KcBzkX3AbcJ z6cO(4e)(bC?|$){xqlYXau-55?}7=|V5o&yN6FVxWDOO!_WdjW_{Dqk(&@tJ)w7;T zdfYcaiSeP3V$OW{Z)_1@%ha1JfWxMFxPh&&e*fwbfCa*8`rqcWNG8FTUxvH6}L_HqV$LSVa0c$VQ_!@&Y74|F4k>V2kS( z&phzLE;0^WOJL?$0LWq`actuBQQMr>*IkihH|Gk2`zbR;ZM8w7&uMTzk(UTriN*>U`zWdFVlzd7K?VVXvT)7)2w03b$3YLb}5oC&izJg@|MS}Z~NjxFf`54!g zm9jLMI-fz_2vnf&lypbmlVB1nU5Hc^9mH|w#}=%-u%LW-!eP+y@!fJpYH~l)-}>Dj zNs1FkT1(AlKZlY7-*P(8vJ*#Z~$sW_EMfDLE19Slf|C4(HaxUnofRdhTyA-&jFuv~i1?G)F z$@r}kwHiw%4^}X9+6*}q?+xAiKIi>-BwG%7nr=?!B%P|*dz<*8_a#A^qU)V|KCQe9 z;yzOQ-Zha%p%wJXE;0$BIu-UKT=30f3W841N!%wQZOAHl9lhIIr%H&}8&pTn-Qc?) zx^0ieVAg1^oo&;ILH6ZBWID31ZMyxfPyc91#p|@j#R>yCR;6VtO+uXM^hw~PYEu;a zZ}0}#Hp}*%L0d*+x2^eb!MG1b(@*WhPOO!FYNXOb5}`I(xnI5vWr+%@Tv8-JOsTii z>nAq5A@|EE#irn~lZ~(qOhfF0El9A#JvTnQ+AcKtt(m+HEhNE-{nkRWwL?S6(t^;mYx0B z^vSk|mL=X!yQWwv=LX&mVCr7sd1Ly9Nh@QMK;3yuU>#It_H(fOL!ALm>UHmYXp+~` z=|PG7z0wD?`ntE4#sj2S#R3$(vxcO&<%ts*H5P(qD{Dk;vIGXpQ2N{|fG82Un{$+f z(Klz-1(yV$RJ2b|V1Q25=ZW|1Df$Xw&~2i0+YqZ;17A%pv|>Y~0v!f7g>T=XIw#JSiw-u8Xji{}^4vJr{9c))Lvnsg06d zoD8OhTQ%0|2et#lzEYr6EghEN)($LzFy<{1;4XA~7 z&FZ0#eI?235y8ESMsY%L-bWvt!#@K2@M+&F|N5%0I)3@G9VKR#;#HG8>g%jfGH!0# z2bHh;e#jU&X(!d;>P^92qycI}I%Su{j)%Ytb7dPQUUTTIk_J&}?O8=Qo(I>tTff|bG;p%d?KKHWNV%c%!RsKWn-C50|eDGWP-PNCw z+R#nDowP>T54}wZ0lT4)8vmdhoJ3A3bPDW*$`={-T?1wVSb493jn<$`w`fNQQjjGFY?`?~*wAhbZE}MyXoYr!6h##P!J?kdjnqfQ z%aKh8%NNw>2j$T>rVmaZblD`nGaY+7v=Nm`?83%t__Ld{$G-=V1SREqzba+{|CGm3 z|5F|#29+b>WQ5#Ez%*`L+%H2vu&iuy+8wx+TBR;-r8Ld!NvKAFz4XZXv4($?vvKkw zF6fe+R8)tc&Q?1OQ&}-{&;?oGL8IPhiDZv|H$Y`nR^rtk`p75#)nkw54t9=B7xQjL zkSUV)$6YmRNtP2w@}T>9NR4zUB?l3)C{x!j0Pqa&$2?zxLF@^E?L$sydi6ql`4l%=9kauAO^9u1=Rc^IM$QTW)=u{ zhz^A;=Wp{<_mTl=s_KM%U+ihoeZgv3K|n6O-lxeIj1w+JXq+5pkVc6r3oO|0<+-qb zs{O9Xy4;#kb)S63uDsTH868QxnJv0R$uCgk92IwQ&Nb0OV7x0aFy0MFOPCc*X*9BW zR!Y}+K{c)HhD1k1x(wKP|56uRz_~edvvSps7tAv9%OFoc zu4}Qh+^ydI^JzL&AFy~J=-Q(+%G>WQn1zfpP;L{1L|Dxc^}s=m1GkYJ8hAehwM-(v zBpg-GAIh2|y8Zh-=FYwK9{#i$jHATXFlOuJr)))^bl0twmSLLLs-E9UDjRc8rO`ct zcrRdN>6WHNstfoh{Zb~{JYlwefUmsI#+iTRIQG85`dYQ>{fF}{sd=5&>9xWcT<+cj z44VDn9b_?6EPX`k=&StW9v!4nmGh0=-^Q-i0U(sEl2q!t?2|6%@}O7^4GLpcq#cygc_Kli#vpWkM|JbmCpnYBM&n zrh!o_aVnhjh}<2FwfF5V<#`IpSgT z+|8lK>Fr?)!{!=}^1^zf=cdTCh8^Boj|Puk*^Gk41Ws#|taJrD>otN7{XB>PW68gs#@u8pG~zcZ61gZej5I8%30LwN+MDFo#(#cDCvlOYSuInQWCZeq{yO z`o`?B4w>z-c2aVv-OHom_IPJ8PetufYh@cJSM$tYHL4!}YuB9PUq#lbMo$djq(^Uq zi1M?$Q|0xL3r(jJg!uHMm};J*@HI+wIjpMd!}0qnJxqA{u$jyyNyD*kI&lFfkb(?} z0%lTj=!j3F;`aGuagqaafEYoe#AX7l-qR?KN$cH@x^5S3Yg~z?pTj1O?8NawD_)V_uNL}HP+Cb#xARJ5sZ0rTm3i_m zrmRY%&kC|Q%lR8;K);-*1ZeB8Pujp~<+aV|j=nRy!57GP{m*$9PH-S&Hj($2L z_npLdazj$zsTChnq`p%S1ozbmf!FGtd%=zbhjAW92oXDg;N?Rs{P#Hv-ZMeu2`BTX zM9U5$PVDqtF@wlyN`8VO$EdhPd~A?3(sjV{fi3yD9H2dGcEb`0^-Wndw-sgGGTpmn zpG!|eGaPnGl}0aeZHO{Hr9SGPz)g`E$fEGay#V#~n46r1`~u+;HD$>nfRu9H55PJ^|wQ;Lf|-NJ#S)uYDY++co9EeG!MV-JNQ?bEw1{Q6%@NK z{K!61i>GXxd5ed};H7mRIIlvm@O9@A+U2!^zfoSpRoA;_L8&L?oBY)2LR?m~6216| z=d)g-G5DYjZ($n>OiRUB{ZjBg)=@CrPN=Wy;QN*hw^lk;QzG|>8fR|zfaqyZ!MpSK zy?a&EF3sh1N9$C#q55C|_NfWX1;`to`*{L0R$oh2vw)K^Ml-=`7Mypv&i~ETEX#gM zD{YIrq!?OPuaC%8o>KJ7di+i)AdaC4C<)7r+)Pq=dGz4)1a4pSP4|_vHWT#|Pc4lX zkx>*csMBO!f%`do1a3?S{WI5{QHFyQS@L`Eu9ci5PS*F63_3aBw?2fX$t9AF zLVRC*|13|eei(FVcT?~4soF}N*=}1${Z_07pfgpJ$b<{&U%%E!);e*N7zDY8%yc>>-$IeiRGem3N%(2k z%^_g+O2VH&;|j1UZ+UA6#M%taEnSjs*fVE|+8DjK-t|6IRh0WVys{i8!S*>Hx(Q}A zeIvS=)p|PqPxnlUjM7oD=rDm;1HIFc5>N z>Bd+JmL`-&=Y{pSmnxB&w;$><9fzBdfn^AAMg)adKg13!bG^U!;a5!NBaWX2f~j;csvJ(NRksfER0jo!IZN z!uF3vd>}}J4tt6WE3J;v4G{^T(FdV6UQb{ZB(!f#Zy|?*o`@Flp}LAzWBqum$4W^h zuV6B=a5@Oj#_#%QQ2lCaG4eaCqx#>!9*-gbg z3&|f+?gVe;F)^?jVRYsM*%{R&UhSbaZV($4@p4Fu>!sLbsR`KZV^ldpNu*PX9Z&+8 z7ly@J1Ja}z6qwxZafNxNfHZHO{aT_%1F0bk*z?1YgUI|z&8PqJJ4*{=j?F=<_)y%Iqx_5{LTSNcQf{=wa3VaneB_F1x;orQ7UMV927 zS(r3RzKH@b#Z`H>$$EHac-65hpb5J=_PFx2$Vi|0+j^*9sE++@qog|Yw^e~F_;Bpd z1?f5RMga2X&pPtp-~4y{_;nX&%%PVN6rffYd36P5MCQUv5Qq-LWdx#ghm`ShEb7Le zAU2(?JVCTH-nk_(!!;ukl^t6IHza#WC2xyYffAP`H3~JR20x@1@^P?x5`1dq=wel^mZyI?Sy1Wl9e6)QwaemTfdhT4c+i=%baV6`Yqh@IYj2)*)A9Sug}I zn#66goum!4c8$>t*k*uCy)vJvqxXo)VsmIi|0#s{QI^lJRY5j}eh+N~-jAV=0$|@w z4p4VtABixLi|zDFLv!fmk;@}dnjXwXzegfBTZuaKIS?R$`VwS?fp0yCzrywh{iJ3A za&OwQeAqZEY`?=Vzu$(LWc9^%-py*=zOY~Z)&#Ck)~=XJI@ozQPHb_OnYB*!QS#3z z(gi_bD9_zaXE0}JB+{IFYf2k4w={VU0Z%vFA1Q_$J~1EEG|597kii(= z371bTJ@*k=$(w8V1@f7to=>MkLN7qQ)H4CpHgK1c%&Ct!ORto{DW^@KF+X+r$lnCL zr}7gkiH05YoY=WOY-XN|DESVGYzJ0iS-Z!v^_cE$EgO9qJqjp8p_pdv2iTm{Rh z6*h&Rg#xTb)hB7+T@>h4smw)jbxb*k*k2VjbNWQv=?#;5XFLgOW6nzkq@^)~v`%#~s)w$e zohmG%FNe0P@;Ero2H#5|CCm*;Ve}Iw%Z_r$HX-HZKGx8yc0-MIqzn^xqW0@f-epYa z`QvB3$H?Ld1cVfaB&F6<@*#t;9+ z2OWHd)rarAF7eL}MSpI|F#d*Y!n4i5y@8Uir${0dht1lk>{i9J1)d5|nFw3BHd#ib zMq#XoNtwtl7BgCy*@$n00U7Okv3WjL|Lu*V;R$wL$V21xP^-6@k^>>ZdMXZrG7-=s zBGd|UiTd$78{8{tJTwFbi~_S-!3_u2KD;O5|7>`H?4Y%{L zWjiIH8A_LLyPuCm*@JYUf&bxw>;75&GJVwkS!ihTgU-j6gd3%K@Ce-C6j?tfI|@(X zdl`7DOQj$o7o&NrEXrV`md{8tJTfw8nI)EnIxMyxj|gqc2I|zK$vQbPmaJr*cJR|6 z%u)gc?tKInAz3fc$+f(Gag}(d*AdP}W+C`qCGvXSSZ<40qGLoeJ(7ot#s`1;p~=!L z-MVNJxi*13HT#O*r{s4ja*K-VkhB5S&rYufURPjgG%o5W360(8cvhoWAABuBy-Jm) z$n--^o8*91s&WWD!Y@v=toU`9{$8S;<|!LR3UlOn_KuNFvL)yc8ihz3a^gS9XN(gUu0ysmp0 zmk}f#GKQ&JX{<%p3bcYF%EpMjqyjqA23;Uj7P4;`RtJ_-W$Txh?R_uICM$>7c_XPu zBEI&~>&`ZfsmdH>T`=-)>Qql=)pEM%$KSjLx$QMx_Y4(SCEqIl_5F9xy{mbDHP|4m z$wF?u1EN%x?XEAN;MFHair-g{Ji0qr;g|aLZNJ}Gb~(Kv`z9;RS}3D&H{=Khg93(9 ztR^ugwi2L}7|3A5(<0HVVZ!a0sCLZa1gya{I*i!-^arhTwpr%3U#sY?l?*o0)MI(L zS`Q9u0&@=*K)tMvjHRpcYz}Ok`e(O~*-z_sE%x5yKj_jXLs?>t!XT!D8yVzb$O~&!?3aucE^NfYP#RXiz%{T?DqOrJ|}A>b@|NP zb)F{`b2r93B05!#w2SWYsPSoyTp4pP>Vib4!aCf!cvvOvl-OIB+S>1$LJ{(tKYXPA zuV4KBmw*24^UrW794wBljAY-_tk6OI%=M%7ue(xhU-Vts0{)XQo$8Ww-8)+X_d>Sk z4A5o9Ltv%P9bYZt4+QRuIWfCSlEYsr7+I);^(?VBAFSVEL=Uh7$hn^+1iy(H<5MvW z61}fR*&^E=-W8)!d@d~-D`1${6V#>cq+}Wq){QFjDd! z2tTIigB7Ti-V%xBe#O9hh+=PRB}*iCJ+(Zn?HeQO!oCG|Y+y5P++V-t_w$)1hcsRG z^8&Khi5=2&W)5jBC9kGP1!irKLg9utZbR_saL@zsK@cE8(P%?OX9LW3ygWg80|W(f zvHsJjl8(Ab1^mafx(Y;;IwVJ34PI?JFwZ4LYw5#5hl7x46{7bT(r=JpU1$&QtY>mS zy6{vW+*>Jb;S7dCa;Vj<)osK`vc;rWxF)P#juc7OP0DbN>sR^f|6nP-^;+$At;lg< zX>U73(hRl4SY*7$twEB@xkWPQ?H;LOY+}TiE@YEQd?enCdVj6F%t?6+wm7Q$#1)gY zf&^i`q>`(TYE#s6E(x+cG>SH8LmM$^O(OdTmzC`5Qn@^%DIlLO&A@5 zMv9MDKX;(+#@!#9R;diC|2yQg6R%Q`iy9I(xk|~KD59g{dK4YrEn&@qDt`!@!5X7( zA=o994fV~JNIIPZd<*y*x+z=WRF203NB|k1F>)wlSEhQCZw}o@wtJvp?E}~Bu%$j~ zya$Cx6Cxf3Ut$`-AYm#Q6hy%eDRzgH_+~4CGt;<)JIBqidlz z0o#}`sDm=2d2|({1zsXsXdXrf8CENq`OU9Djdi0l%(zxb9ro!Z{ zs80OiFv(5YEU~_;h7WZ2oPAADt6egGKd3C zU;5`Cov;)+a@uvqil9+L%rRcQ;Ng@`di!@Cfyn1Ma6glLFOa=4J)HBBPFYe6HuY!} z^}Gk(hhz7-)o?Bfia292`Rr!hWCt8D%a%j3T63r`GsBTrj zM|uV0pRuqRUw&6vO!b!!jlmHs>oGp=sB@pdeF>96jE|1 zyw9cL*7__CZ;H{WZbU&V4>{^yC!AYJ23(f=)C=(Eme|$Nm5?E9in%NMjNi}c<`{3( zsZK)vu;5$Qpzrjd&pn??(nYR2ELX?^LS|m*^A`?+lXVDvrt$u#(JVI3Tlx*8gF}P2 z&lj6KAj;08u{o^Mpx%oWu*MDpq?&)?zH3%z)J;*mYP-i-10Mx82y{q#=rJ%xBZH5F zFf(M>h0`wIwXA=2+Qr67t!uaJ5y_@g19GAoRR<<)dOr{T-}-e&@lXu;@LoK$ZJJ+v z&e3PATZv)f?9>mKx6&=8AKs8SyA=fs>_RGc!?J{QI?r7XbSL_#TUB z6G@vt6cqsf$pLqrbXmk5Zcc=HSwxW_Az~qrQl-;3XI>yh0(FO^1^C;MO|DyZU4#Rp z;`E>k>P#PoI)uZa>h9>WSfdy^DubfJ70Pb{=^Xzo{SY)h@zLW_EfncLm~n8by1}!R z*C|1n+1-=RMh&`@3ibmVT{@8JKvhIi%wSY2oieR8$dEn4PR4riO}DKvxsih|U6O;r z1$^Z100xgDugkM@={w=b{{t4!HmE*dzQ*Wa+_)wGRCnA)PVP#>za~N0ng$Jo$`oW{n_W6%YK>TYLBrgSp`6{c|j}Mc$BA&x#UBg9J6AQ)I}x4ZX9l`T@OL0Y6pT zCpM5HBI#)vWE1;@pZt}5|Ykfl&90>josjf?+67v3N)I*#&`fEi!WJPFsL{4Ozv`4(tGck4pr`2t< za=3Gis`7{flMeW&MWP_RPL(1vzTQmQq+d)~!|kX4d(V``vTIXHr|k4>fXyKKG$>nY zUwGLB9P>rR+AF()XF}w>bsw30lNBoyVo9eH?}b*Iksv>%;d_sQld&@kG=dkqfk>^}ItNOMM%eR7Ou%0$FW7Ft>DbvZ9;G8ZTsr z$)op6@Gg7`@93W1?_M!;OJw`>1g2=>y5MxhzSzgi4t^^yL$-b*qGfS#B{(zb^f~W~ zq+VL&gAA+6X=_{^g1tw`l2PQf(6tl7a0`Ix3E46RT~L)1UD_sDKX7yBD0On&yf=F0 ziIF@ApRupGF}D$;GCt-*+4U7GsEj}U*Oli-{MbswqZi~=XU zQ}uqO#E_sU6&#ahLpM)9WPTbA&QG@TU~qA0vjF10{qAc6zVIpaKj*E+2%NE$#vwqm z9WK`VS;qj06_}u`_Bwid2X75er7%zC6s= z(6&=HqQbVp?06R|RE+ce!KW82DOui-NK2uaE32X8X%yK+#dS(nLcKP$6kZjbhLrHu zNgIEHOt{Gb*r1BxmnQKunp{&`y+e$O<&4`1Co>CUIff%%|3)t?B&OYSiA%;QvfPPx z&mc%WB)GPbk|$FniHb`O=%mXa=5a&T3IVo#J`Z?6XuN<@FaLVBNKYVfIK~1V?B?1L zvpW2ktW3xFtPi5oL!Tj|vv?87W0&)B;>}8}8IJZ-a!{Ksrs5h^e?712_9~fK#Uy{_ zDOab$p93y^f}PMkw2)gNeiYNe!Sm&AS4GA%%Y7E{pK{w(*qEp@Jgb*h(E~1e=|kuq z+9bv&8&!LtR8ueAB(`&97CW4onU)t`43C(Ro&E2KmHB}taX>X$vm%G^+Wi{=qa z(6I4ToH)g71!6f$EsYv?sJgnrqaFl5)LEPq85Y@LK^I)f0|>2)xE7%9@hkP$fWS?; z+u++pQ5pr(C=9xwjN?Y(Hszx^-Mm$xQKeHnabKl!M2Eb6AO=R~Wq29EwM##nTJf&Q z(#)8Du9cKKv84f?jUnF4DN25vB6Ua)1(A5~93fVrsL_?V2_g18x<-n}KnAs!#LJTd zD)|>fdVyMQxBo!sQXlYfXagHmasZ-elW!}pA}ojA>bFw4QVGL=pUZ6p3`I9823-0* z;=NiGmq2tL1q_iHJ_oBPC)P%{OR-L{E_&(Y)o#7fi^DEKjnw7YBWkO*8$w=KnTDQV z4-r;IXxw+FhYY;#vHU*wyY97OD3L?{4CGZo9Ocnx=wxLMofKV0uaV!7p8Es};G#r|i1 zbbIrJw@u4Z)3v@TuM}d7IRa^=i?&~5< zOCMKil*dT260fY0Zwb*TZii??bgEtk3k~Web##+#Hz&m{Ph1X_4acBDM$a1!y37G! zXRMhIK8jbH?xk)UZ!$tlmWS7Z1Ojo|guqraBea&1uc63F>=7_X@B$T`M-8crDP}rp zNAe3={j_bGW2;X)?gBe|vY>zBw_o>4se0aiPM>?9ds<|rbZJ;F)M%enwDNZFxBBhi zX8|owE~r1W0aH6(*9wZIhhx=OW7-sWQ112_(LmMZY2Qqax<=XTyT%jM;d`R{#Hq~O zw16FSr(i*N4qY3XOQ*^=LG4R!$bd_wbcwv&tx|dh1og5+yEzq;vOF3h;yvbe%-A24 z$enxGHA$Ej*%FZ=>xO$OrAWCt=+XhJUq@8GdjfM~`r!0LE+jzPWbrX+v$PQvVr@tw zw^DjQba+M%t==AaC|`X?3#zCf$3rWZt}ad&y3( zQgDmxw>VibJvQ01VH_bV!`bb0)9Ipa%WM=IkH>jC_K4No6*|gI;&y4GEBeX3(Tn&Q z(BNte2_6cuKMnKCCyNC#UOteWude(3y)BU@Z$|om)mzCfC*D$>GV^9CDEVQElu>bq zLXZL;SxnW~c(&EM5qgJ`Ip@GB)k6~^{>gyVqh4An&_N%yA%KJI%9q0oeULSxOwLLW zEW$t8%2r+@tTXs$M@XT#+)WRaR`_I+*B<{yMVq2Dx=D;QyLPZ8EHE*Q&9HWA^Z;S! z{doQK@nK6=5G!;V*fE_)>li~XAVxYM&hfxCMRqmj@QmEZDo~WdxEg-$2C=Cvu3J1? z6}MquF}hu=EuYAU=fal3M{y6Uxp3Zv{KDI*jnhnK=1=9SDpEKc(~$H2{seLgLy`=K zDR~)1_EB*~QF>mzYnG@a_@rVHe=7)047wzdPQf~lC!$59pMwN4z?vY!PHc2jF??Jd za!|Pw0x=8tor2P+Dt~Q6y0A|?59|J~M`;;n!9RTF%cuK)WFNW~jdGtq1YtTUASBt<@>;?}#a2yK(@k=FPakFJES zi#8Sk%Ni1Fh8%>VSi z|2n_nUA#K78Gq??*ar?TeTxxYv~GH+6Wn>fGy&zwy&^HW%FgsS@tXF;3_|xP`5lUM zP;sXe3BoSAgF(%;Q$!Et5vwCJP$bDvpCd$}r0&2w+ydcpeg!C46h`-RGF(fT4sfxJ zdvR2$IO?hwFK}By<+>J2mxeBwRYCUpX+t}hD$h3Pj%#+)D4HW4Mi0o;Sd-kr6bO;m zsNS`Je^XWsetQF5&CQrNFM(-^Xd!t4z-5B_dwfHUqEXU4{R;308{Zl?|BcK_*+PB? zW4w;rJrv=|ggU$90jPrQiM2aA`lv}{~qvsZUwz**t;H1=g?h13ZT zhc1iUC!i~Sn|DN1riQFIeUM3sg z7)!L*%o62MazIKZ6pkvMPXBzmVZV;~%;bOr9vzJOjt}I@MLLc~u~Kp&V$cN)L>QW; z1_^MA`*xAju16rIT*%otd4-#!c({xcY&%khu?h~sh;h)sUu{3^YC=lg+m)Y^CG3#m z#NobeW=Kh<W-*JMMo6Y=MB1G>-}PGmwN^5!PR>O zi}?x6g76YydT@F$-fMIN23?GM>-Oml5@RYUk1mz=i5*ds#R3JxScm7Q>@)R4MzQYd z4#V6z?$htDJ~o=xY9}@&R$8l1g(LUoLOxd3q|;sA#cuTy$1Wp7Cdcg1CIp#1F*Yxb z{E4V6OZ2SNwyl+Ci1fVVnICtM-67d@r|c|no8~K1y>jVNpY8P3DQATJvUWRGjlA*` zu<1i}+_%r_@5kzE>AZ1?$%9pk{iSY;jzVW~R%jAy! zs_q}>kcY2XtZl6syU@b%@&SrGrs8%&|Iz~LhU_n&OSgiu)xkOY#O=!cv-Uy6WO?KU zp-y#zY!K?9^5m+hUeLvbYpM-_0_ukXZ$>nrM>s0$^-I{iD z=4R#Ycl$X{-qxueko=%}-ocoDjz-=x{b29~&mO-TZeG}S`XUg$KJh#loXgoJ)~VXV z^O+K94+klb@V99po0W;&^q*E4#*GhH9IpAH@i?E!p>?VqqC+8zXJm4Ed|t!4nagW_ z+%THWvlII(6zi3{1(S?jIdjXpRdTz4=k;WyNrC z0!>SWph$qsWZkj?ERm}6-yB-ZtP&3s9kth|Is7=b&i(LDJMOG?Cu~B}*K=QYSqmin zULt8QXc8EN?HhQd{!QY|p<_P(wrT}9Zlvk({xehIZ*pv&%1^8$n%B&+ zf%>T-NyH*bzJns$skmeRv~La;A9T|DV%ua@F>3?rC3k>SzSW}^{DyR@>{}(@yEA7y zo$}*@|9t-4C0y*%uMKVUe;TCcJ#?>_WciGP@MWtNdA?-4a9{-mqdaXLL_a_HozU-? z0P{)y!CT}QyTGgy=XP$IK}ko+&r{?q71uj8BQhI!E2{%jWXN6lh_o>)+zd2TCl#q) zYdnz}Gm~_4;^hqztj^o#)9;aJ%GxzZ(&-$2g+Ps>p(S(r-51Qls@{D*=eS*lQkzo2 z9w;BXNYOL%_k-;Jtn!){IMj_C5f@;b{by9SWGFey_TSKrv3VNXEyP+!}Gl_53 zE!zYgEm$UKGek!Zx))IR(i65~QJepdN#8R;>%Al0|4NPzXN#QJjqNZ4&}B+~ks^&y z+zGW(P2$xaYIJ7v+_12!6^g0~sdV~4_@^BNBmaPp92!!?aEOIghJ+C|Fd>O+azMZP zO%9US6jHY&w}6M!B;r!VV4k{3yj0l{Zs_O5#<@It5&s7L$U`qqj@3q#`(Y3edhkSr z6!IV8+h)CWdj z3DB1W*M_!4oTb&rmAR2w9Ply|U4hRZzQD^=Nz7?*KNs+Kk~Y#1Q61PUXp`y1m0}br zLP}l8`@?76N*k{Lvz-t)%HXSgZ+5u@^gHb|4K{qR>~66>4a4~$PTPAHb)jIKo5>)3 ztiJGPvcicg%X7_)&n8NqLXmY;T#Bra+bydVZ-bBy*a(BpO@RZubV(gU!;;fd#UBnBvUD;^bXBTERq)H!+K;wW&$ z3PR&EuhoBD{0u@51%Duk!?`=on}$P zRm3$>z6xn2Gzz?L>m-|ZH6xmq5kbOeGDh@%wmxJX62m!ApOp5uSa!g%ske0EEtQpi zxdbN5<00%;`lh!`S8wL_MkmUpm1X<=Gs z^^r8+htP)&#wkIFJG(p(%WjU)DDDeZ%Mv2CL?#IbUD}xHP>ljN@9TZ{%%}~;FwF5t zkiczJWH1;$Ji^IQ)&pZX4sZ&JH*vfIA(E;An7U%}nWdADE7#6ylO+)i>=)MwMg=E_ zA;|zMD=@L)^y&12U8A9ucH;P;6io?2;V&)=rW!&-Unbr>+&ILj|hY>Q|v?g;hA~QKleGUh!QDO)wrxLk4_$A&) z-E#xfOMRaDqykk3eEYr->;2ZemG_n&Xjzx~z?H+n+v%;rD+4J#6c@kN~X z>)TI$G{w_|yML(KdXcPmVqd!0Y|+Z12aa{eSiPH2_=ud@y%b7 zY9~g?bu*NlrQ~NQa*~Qm0^Y@RIzMPTg!t0K*L&^=F-ojKofTAD?G@#{h5B9n?n@{4 zLM(fC$g&wa)pq78SvIjc6dmXFV9QFFs|@lCgJBHAD~+ml=`b_H_vw=kkN_ z8d$!`+x#>NysA@uAzlhpMn`~n?Wq4H0iL@j$qU1>oDpDfgrG8WjbHuZ)gZFGgB7SK zm&pM~Eh{fxt9l%;K0|!yb(DN9Mb=Pp$Znwl&A`(lXi%A_H>}J1e2^qy1XFGK5wm-3 zD^S>tI{Le<17*$6-_iWcWNChOXv;UqWhb^YeP#f-P08CS(uM@pDWIltL-s!%6C$<+oC;T0g3KV2 zW@S<3iVI>*KuK6Sy*$$35vIwy0wIGgE0$_tdsnX56ui&v5p3hN91XJ<^d=h>_>g+- zd&`3_kcUvx`XX*^1=g#r%&6{o0g2XuHJn#AA=$m_byo~~o)@ndtdGD_PeaN(moAH~ zjRv)az?{JTua^2_HX;RT13@g+#7B%p8l_Ui#Z7q*T`6AcUL%F7;bd10>YaDo+F z#{V_))`i!dAT&1;>C^W47&~mObkN)3lYP|?HX!0vLpbC*D@2T&B@@rLBxrQn#!D-d zi&#FK!>=O{5Xk3uOZS782Bsb6%|jAdq^L%Q`#idpmkvzjXd%W3G&Ue)n6<+ppfDSh z6+*@j{NEE>TulB<;TOO8ge+heLU!ULgvQMKNu=cQ6j^1=p98Z!Zm&@TTa8@`DHC6G zAajBupNhfwN&wE zeB^T0D3aYfgY%{=p)dJ2$<_uukhjT~a8FN#(1hW&MsayIByr?VM4gf|LHW#42v|U$ zBZv=g{ic@g=AgA-5t;9eMSG=!C6V}dF~`1N-%(7E<*b>1!lQh}YJ!}%1m&ai?nk^$ zYfj@Y_st`l*{wOw`!6H=&DNZJN)GjXSyWu5IDQsAYZs#xYEOlr!cvO%dC9Gfe z_@3ewM}16K`RSQ)=Sk`Wa=`3EE}-PV0gz3_wL?bckP-{lz{KbCo5Y6dls@-{i2ZM^ zV$|nji=r@LR5249rZN4eZjtpwYawB*Mt)eVf@|fq$sWksWR6+^M)fHhP0PzZtPx*h zWmufo<$HR<^_P}x<4(KXvBDp|>fLLiL~b4Yty7gSi}-tiIB}VKziio8w1Q(X$3Sz( zzH8m9Fv9|}uNum(*I7Z9`iIZ1cYB!3$cN2jE=h9Yuq_Cd4vE`mQgYbIq)~Ad;68fj^CWH&QMsx6O&BSA|J!HDYIbgk^V&s{Z-%HeO1_DLIWws7q)aqY*Bkk%kQN$K z0C{gMk>oRdq>4E$!fLuP0Fh}1Sz>9}d3VIg@{s02OCATOU0bX$GoWxJ&Z=6YJP2X5 zCUFtDEl{J-XbssMQbU>pAo`A_Zz(d=uhsHEloV-F6G#VW)b5NzQgKJJVzQk3A-2SD z=!j`IjLFK5jJs1GGsQA*{D#!^TFEFIbIH$3#qlL}lsL+Ze%UbM;p=)<#qiYXMA)jtboc< z>g4HkyxhnIFt>MFSuEz1u|{F;{2D0qtz!1Z+?v%WF7Yl?uJLIRZ3vn_SMOH~n`u-t z$Y8Lgww`lIP!Q0{!(R@S1y6^Z4Y>wyujiF16NScKpw;7LE1QjAvb00?dHoU?TDM1& z1cHxV_Y`2evSN}!hxD;viR85J2I$Fu*(AN>a~j)%<5Q&`b^*h`kV>r@3oluc(mqldh1c-AF^$=1h`8tZMrQ!;Lie>BGxhhU{-6p>w z)$=gI-AQ8{{MNKO+v!j2G#xhM!cOmJ>o9K!E`D|HBY(?^wS#83D5T^NjmV|qT6sv1 zli^wkSU{Cbq&CP8N)G6v)4YJQ83h~4R>sJFtD<4;}U-Lpa1c7#V2 zF1k?MDY!reTyD-xk`1_A72WqaFL?l*_en82RbJ2+lB+?L^xYYU{Ysc>Q0=Qw_Hg#b z+)(B!&-(XqfdyYx>HR>mU)nA$jnVV6L=S*+8FG%?h0Jl~=hLb~`ljE4tkYRfon-Eo z0G(>Cep=pJDoKW7 zXfBAYyc6FWavS{$!?Z}m3j z8KE~EIKfK|Est9I;MAL82i$eiWf6M6KC+sRwcHE%dC=Lj6Y`Ojux@rrp7j)Ex_3i= zksg?&@##H*{m{$SAnEtWi0lB8`Q!jCy(6SDxQe?wW*4~{)a8xuv%n3rURn;8|5Pzr zj`4TMq1z3uOP%!mq5s)t#QqmRWdo$fjSH6Llv;}TWt)B9H&F8R6iGBHc0%lC%2n}0 z!G2DQ48P^wvTM9UF^~`BR{7iZLD=+L*>aQZpQtUP+3y}!j)?PG{aZCzU$X>;(|Xlb z5R3Al%eh)XbtpE^tM5&3j>w}oMD7iNNN-+P#iZtlF%?|19kH+`*bk7e9ElZT$7Rp> z&A(Y{_Bkyutf=smQI=;oFL3s3;@v&D&t@W>KNNGTPfzp6cUITcCh&9 z{zY%4$^Shb_vw04GJ({aEhN>H9Cp8lfw{nIkN+ZmuOMS$qoPfrQ|z9s29FS=Jv0i` zRc!@oTbX-rXKrWu&6nv+mnrTWASkGyhPXfkL1ncl zVQ~YsDkv(7D-h9Itb$UC3g7c2u_cjc4oUc;y)(Z`a+Vi8@BcjS^FGV}FW)&q2Q6G! z;GZKhv9N54%ZtJ&38r)Vy^n|M16IU#8bu`Zz+AA*Av#LqYsBgmw)XL!^8K0b=rb;0 zH&<>;#ryt`uW-Lw-DSx0(jBLxkrZ(;Z!>o}Y=uVK0NIV~XMoTpr>Cqcnq@6r@q+An zY=~Nv%kI$FT97F>QC*?0f}bGti(LrIc-J7vw?NR3?7WM4$cbQLPe|tO6q=hDnn60} zno0}gn>t`WaD`~70svgj}K_#&Ux`}fJmd`SACm(mjoBS?$sna2njO5K-G=gVw zXf9Mv0b8L$jp`G)LsnOXCBm1@&T)=LWk^e-JDn3)cFkJq zIHLaL(}|6LF#~AX^3++R%Z+RQS6XoDJ))R?ihM>TW^>vii=xwb$4F6hX-t9OI(NPR zatW8b_JML<7mZ=FR8Gwl5Gil;tmc0Xj4paORRmSLABtN&OJlTANWX!31yk(fn((^3 zK!mhW>L%$r&U}I8DRs4S3zN#}65@5*C`?kok`{tvsJE%KQKx7vFO{Q>vb;;9*edRT z;9gDaMrg}~f%|<@;*ebvBMYYATOcZp!A9GCN{veE{c$8P1JV~nT{Li%r}2uQKMVgI z4G@07kzWvQz5Ja8?b64d^!I;onXDd1c3ZspTPP-nA{(j1T{3(MP@s+LLq?E#(WIqO zm>18Z;jtGWll%IRZc-hQ?61bYHtcXViHBfQn`_++2i|VSU*V^Wz{*p0-xiI3nf^rM*@Lu_PNp8rl_*Q5cHmvvq2BMT8&q$ul!7- z!0U18%a-&*r>|*Y>gE>(VuBsd-2uK?`)Op@1Li4l1A9-|0X0^Jam+N`V1c!KHmpfQ zO8(O+CY2&9p$~D&{Yc!dsLd(9h^;;g1bbZCoo?fwYu}@{akC2#vwQSz8@+j37@9Q0 zY;xYxC)AO`7iQ;m)B@U76jMo&3S>syrpo8k!v{l;^7Q zgHH!Gan+^51KdU__Nlbd2W34dHz`Wlgj1EwIW%NU!DzQb2YIbIZ_}{A6BLsH_q|9_jccH8q6@>Cg}vos1&N^b&P$csO^{OwQYWRW&Lj@}bf$JwVmA#RT} z77QXaRffryz%E*i1rpk*rM`G>@atK`cmP=L!w>TA#jNsD;~#zGASw1cA%=Kj7HN`p z$Sa{jOkEnYG)kj_A8527@_0Ee4dhb_B`^j^JI(EhNt(I-2Q72&&gqE(bB9Dr8r6%j z10UOYv3KwrX5h`K$QL9Q3y{L~9~+g6cw@eGr{#0S4i2(!L%jP{6q8Jmh1GnomA@d*=Pg zyIg<7_L3LT{C>T4AqAVvmK%p~Y?NK>6dL3O)7Em6L z!Fo(*+~D>5_ujTnNU=eL`$n;ejr3J!l!3-hegi+3r(vM`)9;Yy7Sa#ZKgYRq2y^*kp!EWn3r_|(sY)yt{gLAA`c*F0c=oA!pnSM4WR0}^N*Aq}q zrk*|}COu#GO;c&KX^1=cROD0CmlU!yuHsD!?u4g3_FASF8TZw`--fo z6m2jTmWx}3xltXmJ>ki`4p|w5+%x5-YVt<@9xn73@kav>Urq}PWPEwkD13TBknvK< zK$)GHf=;OT9P~KsiHVg{o(m)Zz z*AChu8%8#(UECYeI^PcPVy^JPv!Qo~te0T5Mm6ZA4*~<4es;UC8CvG%9pZNK=i!oh zRop18hhR(&{TbO7oXS}ipAzS?*TF8#Ke zA+O>mGuN3*pk91Qct@_j&Q#8-r&schhp%VUs4&_pxfcT=vxBlG-~CFgejgS7ypdwW zSy-?&&WRuY=SS1b5nB(DY7fa}7qN9;Q%Me3#NEm%ri3E9s6-U0I12KMxH-{#nPjW? z%Ji`spL0s+bV7a<6uU8PS4`WYw6H!>5?2zXZlm!|Y@C2TU4EO-?MRbUB+5lXYa>dX zplX^EEzL-A^M!G-_+SeK?2qpYUaGXVEpFS)WW!o$IVMJ4Ou>NS9)E*xJuipuPQb-om!*q8O;X%cm02nk<)Z0HV%rxp7K5Y?sCe-q}gJC&JQ%5SRiI2#blu`o@k&!A|2D*7{kpD0V*39Ph>k#U^pQL0lWGQDAG6x_HiM zQ9cK|Q3^zVISPMyrrhWz8MUNxXb=&bt4 z`KWzj_2k`myOFzqKH*67zZ>TXY&8FCf*K`l{BB-WXn%Z~yvVB?vIWb7JEgm1o4NJ$ zzVFt}*+OT9t^#s&G(`EFvp;`)Bw276EY0F^dx8)0%2l1hCF4%feXi<)vkUi6U3*;f zD%7m{ZQtAG)#$>{7tbX3*g4N`44BmxKGLGG%>1!r0O-<1`$79JpQGiWFv_N&tx>na z8~J(MJAPF&OTimSiKvcl(B^Nn(3|RHx+D85zzh&c^6R!U- z@4MN*%KvV~oE5Jep7U_t>0h<}^5G9Z{6)naOhuuWqxGv(ob9mWz**kCj*X zm;DYpS%KM>H5q}OvNe=bkLSPF`2^9C+vXi?IAPs!55-lgTBy}10P)=G~_e0AMvgg((o##uan I5DV(zky19|-sASSbh$pAN zR1%II)=g3fF^|`bM)-IHb_2W}tNC}q*Oo{A-~5%?ztP+hJR)b@c()B5a6>vq3>4Ew zkt>)h=#f{=#EE_sc|Q^&erbXZF*XsM7k1H?z$qz>!Pt$u_az;@7c`t6!T!6M(?(;N zcnxQ3@Sul@C2*Sxw}4=MpyLfgJUf(ylFrCf4$7kUlWP-A%_!HoSHRb3jL<+Zs1cZh z@j8enl_=Y3J@;r_pZB0g8AuZ8=tjwPFXRt2MjekUFu8^Cr1QdJqdFXp+$z!0y91xj zsy{O1Jk?dU|Kw8;$IcinTJ+i?>kQVD-AZm8QM8e#x&|ycNZEiDHc8amxpUP!#Jvz| zKKsg{SFr^%Yig!k?}Y{C%cw7-a!+}z&N#&$GS2v|&mO}zWY$a0eQv!ke-a*UyqU7G z`#$);u85lx?)WwG7m#iA74eyb!a42oZr(CZnd~I#gXkS@mJUKE|D)iRgmPiNY8NLp zqJ&f7RS%T*cm_)`N~iARej@us<}&GZ_JFZAAJu%p>lJ9q3-bN>TeQ|i3sxLA%v=RDjyQKZ{wXq+= zD&~~%r{l=7Mw)ZHqOiobGhQ;|r!wsCOUVv5_Ir+4AYvcIfJ|jMmH6;$Il<|2)1NH~ zCj|JZCE+uY4(b+Xq>cOm*%mM_J+eKD=7e>elcG|(3HYcli)#rUTg;qENaeI793TaRT=($cQPJ^|0N?hYTMEbmstq>`ZwlUXBOCO zt=u<2PHgPWk!YeAxjt(`dj+Vkt(&|^>J$sXae{^Y#j@Ck6UA-2Tej@2g9gi^4VI*Lq(h z>7-+h(`&8FGdDs|3^5qafWVVu*apQg4#=412A_@)v-?{A%IZq8&5hmH{TA+P8O0P+ zWG9u_Kvze!MwEn`n%=RG4Qi%jJ@QnbdOogbf>h&)NsD57>2pvyRWhgF`!po~a)Oq6 zuAk77p!dQqw|w6c4t_rb@?084vH6;;&|HDjR~pA*V?xlA6Ytm&R1No> zy_TO$3fver4HmH3OEG&WQbr}}r|bYpi<;2&4Dgb9_r$E_R#FAQWz+k;*FrA0&1&1WQCr#tXrxq{oHHx`RkyZ?1wMv!> zZpqapen_XB3UwF;KIp!JZ(#2As1z9jdSoa43!^>^U%`2Z$`dB~Lc9jkIGOSlq#8WQ zx~aQ4*a%YK1=RFVJ_4^f1h!@$1+Y8 z4)%Yu4auCL_qE>}nKt1fydqP+OppVuC0H(u6!*q{lVyU_PooiHF_lAn31@w@4F^_J z>AsuL-=Cbj*_saI1&O%WU^Thr`&d6tjm(~yRC>3_3z#txmB{jms2oLe#SY-M0qN{B zg)3t(Of*}c;Q!XoBkRWjlk(6+@pg)VXS~HEFkB{n#M{H&A>OCFNQ!u^5-oU@dhgAm zbIN`wwr@!&miNSza2q&T978}Bvzyx}yUWvip%8GtcMG%C$E8o4-7~f{9?pKQL&sxf zMBFz^cYSi@hrc&N zQas3yE`i}q$4#x|mq8^6smqZ!02qT0g+4yLICL=BL?(P3Dmxpa&wHlx9*T#1$sBn% zEZ-Cdz72gT%%5$)8}Br}*Ta9j+d8-I09nBp<9F3O~-T*vVLeio1!zvkKSw#dG_{9{E*8>?W1jbk@78%?Qfdx z#fQK5k&;{N?1lS!?_{|}ufYR~>7&SfDlvzC8!O}rWBR!oRX2^IV}l;p7?98D^VTRn z{l?unEec?+ulcd@5``|LgTdC^>WGaKj>uqvLUsTwfiTc`7?Vz|l23&zL0T@GQ%)LX zOF@qiZyNNtOTXR0X_W0JwV`+bEAN0mzgn3M9G+(sZ zurJh@HKpCQ#lS!Fr;8jR5pd(x$_9x*HMB)7<6se1W>j;ehUmPIbOcIZrtjjG`9W{3 zceyZEuz>0GA3XzRkJTFDzl@sI`gYv|iZAN@YO5gmV|{NkAjy@F5#dN!KA_+XlC0fD~vfzaL=O`fdoBNa5g~mS)JnXH75l+ zuR6{yUITqn0Hs$j&cg6JuUrwA0Oi)*r{2>*mvHcI$4rsENE%{d?EUBv#BTT+%D-CD z1+`3u=lwV>@1Ot_a6)fRSqPpC)@oKwc65$MFcJ2`Vgy4wbjyO?koY+BsL}&r>G}b_+n|I8)w98=%_4KtqNDsX#jE8;}dZ(8>*Hl!VAn4^TID$QOsIt0OYy z75blu0Aat~8$|C8AmN|QYWUo;f_PP;3YQfT} zANskgtDUO`ehcl=B(&pkxT zG(od<1Klnv6YC?od4<3*k|{5x@je6T6z!Bf@^48vMwU+QhL|E=*Fe^K7kW3x_R^iI zWB!YoWua&J<#0{$jIA>s5PigI^2mRmGL=aZto7b0T>xai$iZnwV-f52%ZD_8fDNzR?XcX zs=guJ8)d#4O-Hit4S4N~{TgMa{WPXHOJkPtYKT)4-OhdDBY3-Q++sV@GF(ZO+e$}}I)AI!bvh=tD(JomU52ZW|&}{(0TTCb# zBQph}3{mAQ4B)Quy@D){Up8M4dLTzE76{dkwS$ z|NZNSyv#^@M}79MWZ7`#7jC@O%Co@JdWy-U$Z9I_?)att?_!_#q90v_1UQmgEgqM{ zoD>~awvY#3TQcqu_vQ>|8lEA~bl8uq05I$bt~qGiFirWxgkMfFL&GpT@(S6&&Iq|N zXeup0vx8!wb|RNbTulsA`=}?$dc^bXz zz$1D<8~8O-vZfA3!k#~$Qw!oOn6Ek)eF^l=&`s(iNrFY;RiW7Wk4yoO-hKde@5gFdWwc0Y3yEGh)w)kwztaajvU@Srz_WwpWtgbPJM|HrkVr@I7`@kqsB01J^k?YKRjIlOCPr`& zHIgj1!CxnzgX)$@*;h|Dz-T3Yeco!fO_YC|DnIwUH|aG=O)d)xB^qF!(qpP2iqs>A zVtOpVFAzQS?ZO5tg_$rSX(rN7R13Qxwv0lXEsX1|4!q!w7#^2d9sGN-$!?Ab`p5M8 zua?e>sT2nr`%F~l>RS_g_@&B5sk&PVU7;ZCQmS0-(+mItC0vxs&7lh=_X0`-3q;q5 zHVPzSVB-gh4n5o);PirCV$`Zz5$klUIfAdD>zuQ(7I{qaH%AcT#(Oav%0RuLHK8AZ zVyl5p3O58m1r>u`58Cl>V1+(zt#G9=N-YCfU)VkVDc?LuR%sL|po@$p$hf0A#A}kS zgTm59@yWc}&|H`Fa5)7r&V~z&cOdZZ?^-*~ZX5Nn;R?^w$AS2^SBKxKphe+&UN^nb zbEB``D_`+ZP)oB`bXnsWNA*1PymJGF!D7BZt6KvQN2q5*Bj- zKUS7MI>7azsCPN2)Caf>Lqq0%L~t5%zpFl`rT%q^wQu`eRi)b?rC8~+o3n@8A-^2l z?RSgU=!q6ay(Xe}hDK4T*a<eM zfi}L*`a~X_Z~}-NK*~_Z)#|h3AhL~=VRG$MdtUWD>vhP60k4(Rs`isxkvcCd)CSdS z8WpVbIVIr>c=!0Is+;h?nsVB)Hj`CQYo^MO2oV+rkw70V$=m?QClTx6GsWN7U;4Iz!rCm z&KKJN{^&RxUfEzNtqVRSHkRu_-d4>N$F%AVc_4}mpOkY60 zPK6QL7U*`u_}*DxR5VNrtDIHG%TsAXm--?{Ms>vF->gP9u4QDtpu#_uQ!}Lrf;4_L zeyNaDWKoA<7mTb{8%Ms{3qsp48fHBj6Zng-SNNHs)%@=jZ;-6v%#9mY69YBUkV3+I zih)$+Rw{8&nG~-T?NF+b${v|adS5aGB+bN5mQ>(9$b}#rl$~Qj+CXbXU35}>gA8R7 zFj=Ujh{%LruVF4bZ#4Qu@QA^6CUDp$~u@{G9b&A z9D{Puc^FKCj?3shs(B0JF{HAKdy9A27Xq{~wX(%w8>f~Awj?0wTH5ccCU;M5l${Jl z61Gy#_NaZq*COZhT!9e>(P1()tmppOH`qqUFxG`Cn(^vO)|+a#MU0INbQ5Ui8YBh3 z%6hGAu13-FYPWYS1WXL{9nvYQoMkGNff$P^O7{_X>OxRzMkhppLVBvUwT$;f~k;i)?&hsvv;Fe#in|JiW>zp7MW<|w5#aNr3GVuV%_M;eba`^k&t?>BTON#Xa%J?VtW9ZguGa+qPxZ=I=k4 z{n(xT<)U0t>Bc*ya~2>tLNSLaasb<}F#+5o&y-_Pi+UTu{gK}5hO{&qvxSf!=I>L% zwy74%gb%6kn^pj9lP5>LdF@@tX5tpMXs6ueosDtwI^{f=4Z(;LJ|2DhN?cB!}+ju838m(kv1)_Hf(y%LLkhT&Q$GD$~Lk;1h|w7HMo+D31ac6dMYrAO6us4_nA7fu^f4`PWCZYW9z>pIsn}zNoh2&1 zm0ICQ)u$Uri*3*&Wkwn3q6wu?+FNSXBrg$!%u$DU@>yoRHT%uT31&l-8 ze$_rzlToo}2Yq|uE^=hzDSB7@#mVYo`B4>CiS|RIRfX4PRlj!@An%^{BYC;(4#fF7 zIG4O~dEHa5#P!0Stw#pzeCsD%kJJImTNHIs>e5O3uz)luS8$5f1rB%^K%UG~ zLz+zt9S*3&*b|v4Zvq1ME8-lb+g5!hCjR}D#B9!lcZF|~v~fUUKXg;0r5I>*%cc_B z$ZFCr@8s9h2SJ;<8x#WZ8)OBc)b$>HcRZAS^B?iBarcqwMCLf^`K7_Rrio7EXX_&3!kb}I{~>AkSp9RCe?XPM7wuy=Xlo72^ogl=nd zSiKix{y6xq{;NfijTsFj@&oF4?2I!>J{x=D+Q&n$0rI+IFlkh5D_Uc=eiq&{8vpy3 z_fxi67m3=a#q9TiP#Z7;_xp5$*NaQ34 zYK#u!wKyf%{(+OCQMl&Gvwh6NtZy!P>a1Q@_w`1jv<<7D_l?_9-hiwLx|n+6FOv1( zfW5Duc$oZCKe3YDBWg*g7Oo0P3eWRw7xqPV({RqDd}%tAtvV}t4-b`-kNH>R3sg5| zW(0LWOL;Ghb6C}q@PM;J(8hll-!-Y0UMw*EmQ`LEL0YIt-o-sY56EhGcj&v`SL8T) zIWY~phdY9o!3jto2K?5C*uM;{kWSq_Fuvp)X1{1oCBKu@jRV>Cp{{ci#hjqXF)9(w zcr{p0nxgFH26ryugyLW@}&zm|J60Q%Y>2^m)!cP(JI z*C9?1SB>R)=Y<2Zs(=dEl(#S_0+JJ06@Z(adWaiV(>0;jgzeH!pccS@&iwdI%1x+j zd0CJgwkoK?3+S)#Ym+opgms{}??%ZImOJm!nL5){vY9|uQ|G=F5C53>>o`Y(0q*-R zv&EH+dO0bO0z+PmL63cNZjiP}6g|5^fD3@ZqMarnJMMxuy%V1IV$^-!pLlo3SMjbu zRr~6flLQ}%A4M*j)CNw{KQ{s`yLyAXdn#^7Fa9q3)wi+6C{w;IqF=5)8T#Pm?1W>? zkyyO~f8{M+8mG}%kkkRMz#-YGfNcK9-e+dr@=X$`PboJ+!v=Z+K$P(Dq=(`*_~jR7 zqwu|%xBygQH45A_oSC&VT8)2TN)Nq*K@WV{PYKt7ck$SB8}vvT4{}V^QiG()w}!Z_ zV}5#2m~Ty2>9!gDEf%IWgJRMtl1e3-Tsg?ur;5f>i6!VuXB^)VHIzbT=zn@ca{t3h$h%>JX!ieBl=xzZu(8- z??T3#5pv@{FTGEeyK#s}XEEnl6ayx14V8G1R~NWVbqK5lt^?|Vz{`HOCYAw@`bFVI zVDrkB^hpY-LTVIE2)o&4GhdEf%g%ZvOsEi8D}u1ew7Tz((w3+^#u74u^j_(pBZT2V zm-7nPnofHFWb0m69v^=ta{xtnUfoPHXkI%VTSSW8*zq`N0fRbh7OQT6UJpi-QW)Te z3<4I3BNLGeshVCm5x9pcWLubRD$_>vtna-5Q=6|wfjXN;al1953NMqO(Dm36*MONP z%{i#BMhSj*3>EAgl$zi8OtIeez96fSjXiQcCl8FpsQ_IdrX4kk1?02{Icw2bae?u8 z`efP*qNk7I{IlFOpLSa}PqBf88?Q$;+N#!a%cdU|ob+GBLj#c!l)^lSTn=&Q2lQ?w zkcx5|WVK+R&I@a%v?M_PviWnWNIRuw%KgX%q#u|FGM*+T?eHYoiIuX^G!j!& z!Y9?NMP(An+dne=X>Ln1&ijM8zS53=5lK4?YzQ!<}TcB#lW= z|M-wKk$~FmAcV$s~COSR!O;XQ8@Hy;&!NwpB~c_ zb9O?9qQF5BsNo*_lgCB`kSB+6`Vm$D87rPv^9ILkaQ=Lxw}C8lV}rBV!r-i-m^6x{ z09C9}bS&#TSH$(A%i?my0&pqwa&)TjJvMS%21mtWZH!wAThb@9Z@_@)RZ? zGTNE)Jm_FCGXY>Dx4KlA9@7XpIp7k<57)UH1dD+d59;0ULOirL8fakk0i`iI`miU2 z=647>K^&-wUkgG%kggLLB>mp%1;A8s*b`PIOvmko<;iihjmV}9yRKLqd=@yKd7c$| z#*A5WfwArYe?c5x8~x&?bJBR#b9T}>%9ezl$bP8lUO4migf_1t->T@9@^zv;iUE%$ z34JjG9(jVUxSWup=(@l**m$3ntdkUp*K(UcE_btPaAqykXzo*{Gb1X7dFCv#FTk@v zk?D&azlHrJ;ji53ux_GqTQ|iQkd@F#nnP7 za4A&(;ijmNGSQr5h2{!6jk9ihP1DKP!noPg>GYeGIF6>&R42cS-5-L#G210Fj)>=sGG*hx|W22sfLgn;!?Uv(MC>+YKh~EkL_7;2rIUSb@&V`tWe8dU)VCk?CQRy zPpBh>?D7I`Tzvo=?;-2~RTKl&%N11OJz04$)*65~UOkOvN^5-@X*>W`E&0%&hcR8G zQtFT|qt25IPBy1Lqyd7+u*GVU!fr}lOYfS63eN>1+>CEjpy)ylT_2)u_rq%EZO|_T z{Y`Ynta9a`a3s6zVIbfT*oIlK@X#>7hV>z@wX=Ml_Y8pnYC4Oev!%u|Q4F|bi1HO> zQ;m_89w<)3wFl*+aBVVi9O0_AnXKj+peF6?q*5qxL%+J;XRoRVI>L?gNE^T=Z3g{? z>Ift}F%FGq=O=s=0JUlbqP-!k7_)|-7+0Xe84e9A6EtRQg8z@K_f^joBgGL5L$i-! zKBP!FmH6P>SH$nOL$)y|I9swlZZ%mHTrF&f0|jq!YW!AEB2D9U^DYBfMH_#0ph*q< ztgu#w*C2;^Rp^qatNeY+t>O;Oym`sK2H8d>e%}yODLxq1DEml}Jpr$&2i1*6XK5lj zA5p_2m+ev8i2BX1-~Eeqe#vc}9UB=VB&NZ#mb$n|PmBy7_B2RJqdT3je`fit z?S%l#VI98QHUx%YpcuP)uvH;6`!&lqT>TG{G7h+QhdKnCC?=aC>!?Ir|FLF&yH7f) zg0@5hjl_3pf_(7}AU@nopa1}}cm)Cseq@e!rAs;V8QD7xUo7rJZ+78lR@30V-r+Z2 zja_S9BxR$zs1maO%RIM8+U0FPCR`69O5g_o+t6qsVaR0K2Z|v#jM7I|pm5*R?Z6M_ zl|1ij$wg8f0bSUBWgw3OnMHI}JLKmjwV_Vii?G0yeRy@!UWCoxi%sNp?56Fx|9IZ> zg*o(D((Q+|G_4S(&kr_BbT#oa?7gYlP)GtrSI(;Dcf)4o#>_pz8bu@b zys(Y`F$k9y&S{rd&bcp3jW|v=Od6dvaRjJw3=VeQZ5uWJxc~Ei4KTZ-s+8q5WYcrT z6}`{GcP^rs0*dIU#K#oT`=_F4@j;m$q#KO|DZrl*cZ73{8I06>Er#NhTpE>+F`tM5 z6|C7e)=TC?2MGG!?U5~ttHSfl8*jM6=!^g_miER)M#nZ@UVq(lF=fWfBcDen$bg62 z$}SM->KYPgO{JKX6j??knsPq_(+*8}CIcN8Qd>(!TVb-Ge#tLfex+%zk!^ zK)QO+I!^DlVGbLS_iV{Bpz68KT^N$hXnfKHdj#sD2?w}oymK5is#<6ix1}ox_89lb zu5r}Mfjt@zo0vg#VadLw6R=Sb^2i|{hC?s1L2^EP1MrXOc|hoC2tcKDU09|ZgJMwY z27G$XSdE-xJFLu!6JW#&K4TAM{!6)aMH4F@*nO>!jru3d&FBM8cn*3rb4rvYN)#2w zep4qUWLV9hofu*@rZX<)fBZ9*^~Qz`D%^J`4-G&=*CY+aY@o<`D)9m{f5Kf}&QxR> zGeCVwp?_I0lDkv|7IM19nNcV!z1V+M(3XjHf%Wu%<#9KdWo`e5^nlQ@BQ-4mQ^8G;llfE`3N3;`o|FT22f$nsj z-@`#>cHK#@zb7H~AUhC(z+-@^aTF~AyeYzftpIQ>~Ifo%c zllAlM-_*|S{#C6C?`ZViDZ{faSf89@b)1j;F=I|m^!(my=5=WCrUg^Tm2u>eg}MKf zVm_hBO)9YmG!xtSP|g?A7Ku#$8w60qWui($(z6_Tg8)OH$Hm^pe+c zsAbkcJsBAKbSN5t&I*|U6c#h(pTW<}@yjUOf&#l45~Xt2`1X6hZ5lgMz5-l{9xm1u z8W$HJBjzD9fq`DnsZ=75Bn)Yw+xTmv%9VI1m)=IM#Lb_ogMvjk1zZMxb+b22(nn() z<>rif`fU_vd%VC@a~nO4LKO7d7E^Aok$PeYaKVA0Q~rV&sg*ywm2ACou`yEaTR%Z; z6ibu}7YK4d6cYM_uq*Lxd`sFO=eYMJ3)~HzYgc`?7u&A-{kF!v|AW(iecsFSb3!U) z=Qz5s6XH~GUQ?KBk$XdK&R7Lvkk@B+st%K7a4j3DyB$?zyBlwvK!a<@*6BlvDW^yY zmH46e9Z%4)0cz2$;(fv0^nO_BK4_O?jjKr_yHr^DzPI+znb_nee+nXYq6{FwyZud+(aTe_vik%(V{qWcy ztt+YBHuG$w#2edoG>ZMOS;~a=tww$-tQwdQH0GFVrd*1^3(zk?(rIV(tlKjUBd`Q^ z4eRhltne}Ry`Kr*dESMpF(g|Ktm8m{elntLYFaq(?34owx;hDD9NIv~F9&GhniXwO zEm;gS|Ep(Z`9s$-Cl>;{HPgChtzygb&vJ_FHYC>02YzSyyVzXL3Mga#I`8GN3FZaH z5H&fQ>~rG<=DfuMbBtmdC~}BOY@DSNZ5EYHf2$&{aaL(Ulj8q8K7viknevnVjr_-_ zkeardE)zG$Le~H&+pT^%!@HG1MZ2}$k9f@q**_}zq5eP0=5GDL-*&w6x7+jDzj1HQ z{Wt5LQlM2!kd*st6rzXcoc#nX;T)R}USMF%1lgQOV} zmv`xFDhyMO1TA(?q$9Ah`yBjlPAFu#QImMJvWEYjV z4;qbcz>fJw^h%#DsJ5+%E%qw*!Z>t=f3q<)hLjrJaY(?C2@;zHVgwVqKmFc^d3s)x zwBIKO<%BqEj9yPN!qq9L?@)o+tuX$C z+cZ@Vwnd}Ba&H~IK4kO?ELougO@%8VF=Xt2?0J2RHKBpq?wD6 z2`HDlk+gvTv0l*!`t&D7{lJjEo*;XA<*YJj2kMy91Pncw#Rh%`2i=s@^s31N9&K^U zMXM&K#~mhpfy+WyP1Z+riWY|BIO%cVjs&&^8~EAY&a)@1@nErhJn?17j5_v_onzaY zy=V4Vvu#mC2mtSN{wmk!+|B?I(j*<&^m9LSxkmK zh+r|C(=KI&h_OdP-_%uMMaX}fABK|eP<@t|IoTo z-3GUaS?U;dQ4W#|vv)=hdK^}!iaMhYdyXEN#pA!aJ@2E@4nMwOlzy^%AKkVlF5{=) z{e>fLsT*%uY`CS_9OR!p#?&dHqJWbc+6F_Do&@oTK=s8tq?EG%hHnPe5f1 z?4ok@9X@PmHcx5e_xn8bEkcIlX2n5*y!z$~bad9#P8D(l)$rhIr0-9UTL99f&P@}s z1kDr9s>J~(LTc-L!!N&D9YF1V_hy!aY@Dh_7ojOH>AmGI z`!i-c;{Bla0$Il{_~FK@3=}L4+0N)FW*bE|Q;GF4_r(RGEAoBHjT3sf8r6GQp{t;{ zEeqN|^a0o8Eu1vp7XO_1mIREbw(+sRX@|Hl>Zq)q{wH1vbO-*fM37xbv9J(BzPJ!K z+*NsN1!knYRXYCfB>9Cw3eaB+LCQvo$wF8sYN2X8?bRz{jiUaQTGiIzS~@%Ay^m*R zhwPlQC>|V?2z~Tg?~|{ja%xqzqh?fC!^4G;7&`QY-pQh`*1i_Gm#&DlykswpYDvQ7 zssEqCBKB0C44tS^`}~WJiHD!}8Z?Z9Uk*M?t8YX%D{e$%Vph-V_G?z;(KQgubLBcP z%skou?hkXb3lFi{agTZTXAi$2H7_JbzyE{FWVIWYj_kHr8n#eO4h8*Ji5I6}y{YNX z=7i(FUh>Tq|BAf!nexr#p;#ZWS6Ig@@?9PC;5RkDu7#`{?myOgH;;ydt^FVw;x~*! zHiZ>PC}rlv1CH!(-8f8YV}Dx-+ytLVYGG6QS=>(HZ7C*jQT0pRqR=SLMi?Y@ft}Gv zD6GE4TN8n8IJg%c^w=e5!L+JNfuC*{ho4>{);2WZ4Ug zXAb1#3yx~z_gG6#z96u$AyC<@NaH<~_L#Y-Xw~+?|~i12y>DsYG0hH1U}IHL6qY5~YZ$ymD1W ziMcDm*wEE2tq4P1(o?W5;Z^D@!KSF|DyU@a1C4G(5*{1$SRm*D_KdF*HRLbW#J)dh zcCmhO?WbeNLl0KokQ)OASoep7PLe5RIYpL0kd5=PaJjV2=WV1yc^m&X>Z_5x%~@>U ze#fuRekt`ozgnS^&d6LrjqoICl;#SsUB|`M%QmOTjupk`y{m>`P782pDF*CMHkF8VKXsf3k;{So z>jC$TG~dIp95qU7dE4owUOHgr%AxPePSNRcpzmnGslB!Y|VI$pE{4; zD*f+<2fS{hDOUF3(z)-0Ty}< zmAD)#e;}h5r3q@3G{kKP%8bh6-Xv%$kg=$Xt_n5Cie)ENEeY!-xgpRO6WQ&3KEM!= zLmOQ0yRrns7b}koabO(U_QF$q?A_R}R{c*OuK@nypA!&IjovZyY&j@nTj`OesZn!;6`4g@5E_ z)1S==YrntiTlePZe%1D!X6fV4gFinuXTSqF9J=5o#^+raV@k;v^4fe#16vaCb#COR zi>fAnE=>o`fB}!~MlRL{x@xkMmwC2N{EK(>i4TvTm2-aDepQFzZFb)}&;OWzalUng z%tR6P1WM(u=1p3Rh#k^{6BX zXg7^i!A;WrO7KcU^SGUmz(rld%CQxQw47#z*40ACBbs-kd2{YOTaU)d#ElIq zUVPgT78*C+2HRkvG0^!6)G7zvgQD1y@K%X&a|1s(z0lN+YAHrMwb9v}TDlrE?h6F{ z-Y9UcJ{vvYVfw!d?aIhuh1jwNgY&+>Uv>6^wFZU_sR^V^Zk6PV;jPutI|N243E=J) z>mbyfDaX7z@`T@m;^7;9=YpVp30`6qvW1%-ej9My<*HhkL67PPU5K_t|~3<|UJhwpjOJL8@fPCNzy= zQYf;5O2id@=zpkP@02NK`!?7E3AST8<9b$*a9<1k?UHLt93i%HU;mugAhyExs*ga5 zsX%bodl4^Feww?DEcS2XKk`4S(x`I%2b8zP1G0NFdt~KtN2azU;Q0#K7UnLeIpIe1 zo>^PW*I?ohsiQxh^bpF~3Pd-eFU&rZ;OuwA9{3M$_pFXO1M@@1xc0j{dTVXj7i9Ni zLx~pEfbMu27t&VAQo${`D=Nu`JYgFj?E!?1TU~lMC@`$Zx|D&j*#@>#nZQmr@F_4olVdoAFycgeP-7q00yg&W>Yl(mQ z{eR8-tC(gy2<65m!A4tzsW(QW*yy>@_Y-e5hE6J@yNTMg!5w9|*asD~2u{YrGlsJb zm52{Ms8X53Wk(bLxrUUuak#9}V&k-*VrnQ-g(d3`=rY;%kY0{KVvwDkj9qwJR65RG z-f?1pUjOsLJk=ox{MQnF@GjYA?tsTuaWM_eQ!zOq&A!K&Pxu}3M(K5+dv50Fy%q>A zi*Ncj_;-p7{71awpe@mo@Oac??#J_ct_K10l5x}xki!8RAN@>J^voryqu~1@o6txe5t1& z&CGi_?d3e>nwJkr8Wd|@?x!0RrV9&zauG`(bLjkF6uBse_R%DPKH{eTBVJ2FHU|}K zcPck3wc>^06>%4sb&|D{vb|wH#8ms7fIZJOlBMdHlgHf!V7|z?9*f_b;CnyUbWIz- z!GGs3wrhBiEMo(14&In3-`7pUN>HgQ!UVa8}T?0zTf%|*kJp)l@n92U3 z-b8k>gPHqAiO6vanAK1Wbmvr3iJf!`gOR^}Z=gKV1TCF>*7r!Pj@}XgGHuwPRZpjJ z8fi6p!nG0uzrz2VGS5>V00ey1QiG(1&$&S`o1n&KFC4SVOCxcVBgusbvcq`UzNCz+eB*hSa03NOY|#Qfs}DV~Vr_2- zGGiGrfc?P76&{v12Q zYEc{05Ek{p6tlHi^wy+glJmk?8@+|K*+nsh6v?L&F9}nkE(t-F^f)w!wlG^v*`n^i zW(78i0s|WujX@7A(5;5P&n=Q0GxZAfrYR^Y)rBYsDoS!bYaVjuswwN|cVqrtV8Z=)t`wGX!+f)%+CD&MOe@4Z$AD z(MssUKZyN<9p)*x;1IhP*KKJ&`?CJZH-WBmb+vPjw+?TZ^Jeq!v z9ORbC`owMgyW=y-Dp8~K5qP!>n3F=SXJJe+5UH82X-U{Ax+pveo2L>E>Y98g{!ok? zuHCcOj4zFOhCezYM6t*6Fb?X7e)H3BVufX@ea(OW<#{L2$>cUF@6roM75|F3IpM&s zuZWR8XFsXq;hAI7kE8{`=lqsNwIpOI@>JKo@6sEC_PqL;Gq%@HpNf$fJOs(JMs@r= zJCIb)PV;@6LR`VHch?Id<8NXARC%AwbczvQ)@J`X(7`02!i<>Mjf$3GNW7njc4Fi9W1KeSeL$ag!vgO@Cr zAMz=kD(aG!#4eesQKW$NX#*}4U07;Fdn7u$1(P#^ZqBH81=*ur5F2JSvOMH`j3s&f zPG>Dz{et8ZZAe_3+;H_#RhH*U`5q}mRZyjRf#AaI6Ci!jAgdNa^aP7At0VSEku+AF z7ljlk1)@4mtK^Y?Zje#LO&7LKv?}Ox|616VUlKmxTo+sybVADzY@})>sO#pm&9-x( zGZaY9fzj6r*N0fQ6+W3=a^nD;jpo5d8T6Ja_6Go|bY9d}A1LlC5Pd2>z%{_$c+dkX zutP8acY|(}Eet`iwLH}*aYc>{2+o=8r;TEJnK~L;9B(ah;I_LT8?psJ#usU%JVLV2 z4*A9?b-ws0Pn{}CB{!l?QMbpTHb+CKPkYD~=h!2`;&zvPv5kabQb=P$He8?WNO6-J zC$4Q2H?>OACK`+QPmnT2A1`Ur8*lvoC~vY4J}yIv&4HN2p~ISA_C2n8)T!%J#8O0zVs$+PNtJjhy;o2CYeL?hS`x< z$ObnqTBx)Lw(p>r?G%)^C7$GVD6l5>GxD)_3S-Q`;r62ms{64sqZxL?`N6kCmUwFv z%e>TTWQ8@=K&qP+NVm7l*ac={r*_p^!*>3^M8X%JoILZU;n*!LOWS`HYrU;;TgSu3 zMkVRhwm%&B&PTspO%48B`}L#Wf_5+iy+E)}SVYq}!W(5}fv4y` zQVU*bSzv*x0L%$?o~ZXIyP@$Fa|`M=uQo4ru?nQ7$$5z>Bhx61X8v=bcoI zYCMb0%j=|vYu7NWX25;pQ0+(lbcttPgTk_Y(m<9BM}^_WUI<9!4Cz8zOEGIGl7!f;aH(!%n$#0>PL9_;^y?sQ3EWKlcty>?#!9*SRJ~E!Lnaih(Ff1(k@L!5T%ytfcURP%@wG-A(t-xgzc(JH$m~ zpYo{eDzLaUD73r`PKW%!9QCX#&_nOFd2YF%`b6nQmtVGq5tC!BzeAshm4Z<01EAa#ouBPGW*ImTb{q{Bck zKYgd;!D3cN{>8ST8OEg@JNhHSBaO;};fB`{o%g15~!MCU8; zH)1Ba+i$m*dVcT$e_fcKSLO*M`;t8p^+i$*zfK>ZhyD#+82(QDZJ=nYfD7r36H=zn z7t9yj<+t%`czN7*p{drT7x=)2#jMi-Xxm2FJTn|jZ3A$Y{DbrWGuTuq%WKFcb}OA5 z!*HJk42viR6l!%;V(&|a0ApVVcJCl95Oj{o&dlnUZzGxV3V*~(Mi7wHe(Zf_mX1C< zsZ_4kh3%51ffdvQwUIUm4r@i=G~b4@l}hL!a$aP2=*&37~B{~z|vcGBY;W*GZtdYGGCc*wTR7;e#uKX@t98W+#iK5DCl zwaBEH)f8DpB^r#tP#Yx2!}S43cn&3Mf&xKalyj3hcFyDThn3o;o7${9cic8bXrra* zDC~q<5-!UR^D?|Qi#Bs>RXgdtAasK3V~;V-(b!y!8ln67pOLiDD6Ti6#eOW9T?IQa z=ICXt><0CJwaZ&)n*ER0PRACJ;^CZKH{MU5w6Gg>6a(!&`>4c+@m-THi!$Z-Gh4F3 zdq8$g{<#cU4xlm#*bkt5TC(Uzjesn?qET9p>1F;=zdL>gD62lF?DtOQ^@G~+L2_4q zeR`vG0Gfi%ak^<7bv1be&PZ!?9g`_XRUnN539xJFKBsWFu*Q{5Pjtu$M?nSm+s4(g zcV1uryc5&r&;>v}+eQB~*;hMpr%>I-$4y2T{l2M1-PDu}C9*rj`;>S1UG%%jzJng` zB>N`&c1hRxx;9bZ5F(r~^R|Ytjfi19lrbICmA|v*8+}1)@NLkhZlh0wwW$&A6s9l` z{e@2T`cNHxp0m{l09GpO_wJ$}(CRHgm9sD$a|5dHZ$#hoegyj+U=?hawaZQ^PboDD ztkTbmIwgcX4`vQ0pPT@^;2yH`5j>Sh?5MptB+AC6_d6662B&Cm+-3Je8K0kYw@KqNi&Q6>f?>Unq2|)(+50G z1@y%)6AXCthp6$Wj@F8hOAhZd-L*?z3bhW{jE28@O~mFYy8<+d`swQW>7}wt3QJ=er73Y6z`QvZRBTW)ydnAG zsG~r!xPM3Ye}C_Jr$uO#-hBxfxi%`x=l(5i?#F&zA=|_T(vna$^CILC3nEIwJAu47 zKln59#~fk$87b%TH=v`C~q(F!VBn^Owsv$Tf;8el>kYrVlt2YLK|11>Q*@ zu$Vw+d||Z)QAd82J@6N^p;@*(br$JjXK36wAFy)l(CVv46w^L8ujoJsC z-Pi(a%1Z!mSdVN&6f)IpQxyoXiKW6{D}uh+T-7GO(&)OtiyZyzOnDXq@P}qObj6bd zUBYW5pVRM?N2^U7X89Z~uanLZ?dF(9xI?R}m49#&TEDPc&ie4PZ7+3L*WR8+q z9_bz~c^6CH>xuyl?Vv&D_mVFXJ7J)6ppt20iLH zpOQ!Zy)gqG75;bVZK}@58=OYrnh4xj7)x{H95clqlFHnE@?j`xbN_F zVc8ROVcpVQTrCIPVJ)W+Rm1`|YB zb5{uao-U}DI_2A;xEQ)_+NjNk{mX;N?|b5Lx#$3^S#aN^Vdl%PpZJ4$-FmZiyFY1v z&epB@7VB0I#oVDt7nQh>Y@t(OvtcIx*%8%A?})+`1qc}W<1@IpjWX7uADq*}*C@^l zL0kkDhdL<4YL%3NkBoY4>Vx7A`P!&*<)sK6NQT39EREMC&7qqWdtci764O4RA3n(n1_+#ddbhmO0M?%}ujbjYufa-yyj83GQ@jdk(r zTNCxZ8{(U!75<>v2Q{m}<$RsHLC~Vu5Dy9{*CA$vA2*ZLyqZ{GBGH9(FnX^`5qD!C zn*s;g{o4JW!-eYK-^rd}&HLcCPV`fl99kttE{A(Etl~RJP_hObY1n#;H9=XR*4hat zM$J;%xnO;?#3A>J z?E?J;{_OhjB%16!^5BE<0oGJhHW;4@zI{dfUb`HmL3}%W3uAV~RdYA_<*7WA!q-uixW6wc*-N%vMHA^?#q7aN`vECl+?& zGR3q~q=ib<^5zTjAW7cFUmTiC=K(KaTVxthBlYrpK`xvq5Mi-6W~6l5|w0#1D@~6`@P@0JkJRNQUk5>`@RW+1?t0- zHErB3vMLa%$~9ZLhlO|tmMOyqTd-?pGS5_{0h|KU?s50>&ht-zpE3<^X-9L@u1E;Z zr2D|*S~?Sd7Df$WZrCE0+%$7ih#zCs@T zY*BB2IrD2B>f>t7`oIeD-f8P2-+OoU>kH@K{jYa_n|2cNE&BQ?~~NiZyVI* zBoV3!TYS5wUiLaZ3o8pR&%`5V8&Rhr=4sPEjL%phhVs1Rcfoe2?zV`r6GEu<-QatP z-lRGw{@BMzcSbmbFzf>ErE)v`B`0Gbn$_MR93tov2l^i@p`3$%T@ME^#_2?9_4= zhL!?T%@F>jGv|UWIGp*rpY|+P;22Xi^HN8I1sqv{*{eyB8-wGp6*#IW2AXSkQgLnk z$)!_*;QcVie5~ z%quG(IgW!7k8J!~$<=2))St)6g;+;=WFvpabR)!6vlyKqOO+ug4=|I4=)Fd>+Y+b5 zU1yt}+1S(d+pqs)F+Ma&%{7o(sD$a37c0sDK^iO^$HTK;MNuXp*P|qrXHHx`lr>2o zi4ACs5aC-Ts+g4^0EvpLkxiUE%CZ0_twUDdb5`j3@(N_WV)aDloc#U6fFfHRb2eQm zZd`t4NA(;NzeQ0!F<5o0IVd~BLDIZ~l9Dj|0QjN1<;V(V=ULsflwODgbYK z>!B8SgzJ}6Kx7A8BY~5o^QF287H2o;mz8gmwc~(wY~Y6DV~WY4$R;YTn~Pt$6wY@4 z%<&8T&xtklF}ESJTO?dLzK&PGJunlJ@DrB#ps2Ebnl7@N3zSk=u#K-5U#un#^r)83 zbj4TH{02t&jz{tupL#E=Wx##El>wDMUSzAe$Hv~caVlDCh4@s8Sw@kiR9w4!5f2hW zH8N-<2+-3vV>+ej<3~F(htP5i*-r(bOFq0@DxYH;q-O&O_sxhBJ0blO6Ys|sD{j)L z>v`af^FR}U%ARG4Q$n5ip+~_aop>RshqB(zSlvIf+!4<5)30bF^r zmh*s4PjcA{a>4bXs{%75cqv1AgGrm8%k7ToCEc>^GS{ty)?aqVtj7T7!uocXovpr8 z*I)||J6}622GUPM$6^n33083rcw$^3i9$Y0%%k;j>!6%x$@t!pRqwF-r?P#o?4CaS z(a(v$@zwc>b7_lf+Vf`MA(G(6aaf&|2eE=;mQy5|iaSMmy%&ZqicXg&^Hv2ue69Tz zt+Mvz_fuaQt?&L56gi02CpV1J5i7J(ub!)V*_JHEZOa1e(7%+bv%L=pGU*IS2j8I1 zhZshRXiqe>Omi!}s)P-(4Y8MDFLhY3nLaQ#FSK}8CcQbP(yNhQ%efYz_c}>30&m5f z6Q2_#Pcx{i!gZX*pumauRZqDA#_)7#9gtYHK;zIRCNrqoX|!h!;%C4zGQ^>i@Wl>4 z7Xzm!{?cMpUMk!-k$mRHMkUFrn!k@??oi|wHb!Hh19f*-1#b3fmE?)v!F{%=oy(^& zWKVGQlul_SSA*sI7+Aw`7s*mtJI7SDkMydWx%I)@UMN**vI$Z*e5UA@6?67`LO7(2 zkHVfgbO$IDSVAhu2ptbOX)IS+EZ6`IYfaL%l4}t-1Ooz^0>UWdB zUh@6DZ(N!8?s0V{eJ2)klZ}#6i2G*J1u>v(so={tSiX{LU*v$SV)2mJZ5<`r@bVH|4|OUO^2W7tbwdt4)+avsxvbW z_*e|ePtTIKNtzq`LR+m&L>9$B%yvB$*CfTVeIxx)ra|@=Bo0T12;wg3ri7w}i3)w1 zHad-9ju0wwN24gpLGXUbNp==U?LCJTc*iUnb7cDeS}<~T*76s~)n}|DcCnQYeV1Z- zDAG;EA%Pt>gguD39f9l!D1UP#Z1ogKETVth1Zo#XiTP5V7zsK1Jr?qse6`9#5IRMM zp)P5L1bbgWu^NS{v8gp(eglf9G!XdC9QI5xMP5uOp^FVm`a$u!$V z_o3GXdeKe)mP}Dy-~y$l*{_qNox9C8De1QHAUkPBusOIz*YuDZG?MtXN|52q)$>2d zbU6lzL2h*MC039a>wo9$-=&r%>)`8eoFgmT*sTFh%>l*TneX}bn4tEN7xPUAUbz$>fGFrm2>w=8kC0EBth!jKDt4f zCe_W&4+3c}2)ks`c|itsAGciGMQ@*V%RhTgq3=1torw?pGrSFIB!|67)~Sxo%jNcQ z4eC#sDxWrfseiY)C~Bc;#PYzq{+H;@K6+q+1!4ix4xBSh;YXrU=MLvv<@i@)FyrLl zwJ$?v$SaG4#;?w^VCq}@zPF9+8AncAO~zr0f%xfuEI6v1ToPLcp-q(AgPla>WR#*` z%DGD7qnFP5NP)3YBkVv5{Yq69xwCvhpHGVt`P|GOrt-SQsFh}5VB1j(X`ot8sVo=f z>K=sOx5ajOe@uV&l>;12DoElWjdLp545@PzOhays6w6c0f5C93!NFoB2j91yVMD(3 zK?Xc0CQ5zKBiB~AiA@s0jl;lpWS%gilnN}o8>ejJYS7;~L-f;j1)Eb!M*b;ac*ztY zuY*>(F(k!lqNRgCXy7IG=4RmCt~;i`s15m7i@6c`|8fjzbYu7Tx|JQgKrv@2u#G0_ z=WaO)LS9lJ8RlhfxBN1<)5JOkx}Zq4oG!;;N-0G5oA{*xC=qFLiSZr?{EM3SrZM-u zG4KLn7ZT%SY*IqzU!)SMrF(%`2&jT6bNJ@b_3H6jg;sezc#*e>xhj1IUdU191zitr z;$(_2@RO&~P1zLKEjKWiXLiW;!Is@fpW$@KHo?)5_L4nl>=;0nmxqEIJKOljm*ak7 z8<=5Zg520kvJ+-L=HJL~kM5Kvg`SgFP_0Z-XuGT zhd!sk9?C|3*QgZu4GU32!NCci?-(C-%yxst1|x2~Oxf8d*3dUYFlL({crGX{7UHi) z5N<+_htcJ|GtZsFuyDduvw1|QWah-Nmenb7ao9en*L2&>AlYcOIxVM|WQr`I;!NQd z@NEQlza%DG8up0NzmIV+SjQp^#G z9K_6|*>|X-u}e{d+~4^x6#6|f3&1#tTRc;;31pQ(%?+r%QD4G5UaLe;2K(EtzOsE@ z=J)D3+Gm~6<;29#<@Iz|K;ne+ z{Hw$WJt@5;C42!di+jw!UX(@di;HAwLu zqn8IZN+BE%HU&k?p*BD03*e1Mx!4<^3WBk8t+G6%Zu(vSy}XAWSQeKJ4IR2E`@LJ# zE2+iv|B(Gk_WVDzz1sFN_SP5B_{rUG?Dt$XYb~dX{`-ev3#eu~E!HVW83`h60P^_s zX*2EUXY4EuPdDb~NP?|y?AY1~wq{6*=|7(r_d-Eb=lt{H`{FEBPt1j&vZ;Rur5lb@ zpyB9LIZDrg_;CE0&KPOvYbjJ(79jPZxAu@Mb_rs8XvMTZ1d{(z?@-P%nngGbO&9Fz80-ju2+_XRfd#{f%GYlv%zzN z?-6C6T+>G5=_;A&B!!>iEP~4?v+69v^W=%nIcn+s%b^$xX8v`@79A;l#=@~pR>=5- zVrnQ-NyTNT4>4FjkJ2!jPHBRtak8FH5P+tPScAM^TIC7f12gsXuHa3Rp|bun@8h$| zJTa+!Vq*R@&E`qhqqURv@<0u1Hlpd2@FZu6RC7hTD-r?J`(kfQGSld}Sqwsr(13!R z%>5o8`!oX`8E%qB7+-M~RvxcV0}e)l8vBdXThh0u*+yXPgf8)Q_W4|h#4bJvN?}0i zM)W^Jmrv%!{#zbo)8W1*J9X2QpZZx0OY2uozDHIKW?0#bH^= zQBRQn2BlUIs4~;vY7o`PJ6Nh{m;4Re?!fu60#L zU^$)u&;>{KF}=f2{K-gqg;p~9}He9^9`HhzvOnF}yNrqe4>I-E_CeJ{#y#h@Q#zTq>~u1Q!qw>dJ| z*L>w%Y?Bm0byI$bU>77(dtf#4A)-|+W7t8<`S=-#C7WXwfU<9we*HSRM=Z!H@z5P1 zi`}@a5(rxdbh50Wm{k->qvG`RG2q+iBU`>w5Od4BiGVJE@N;=e_$gJXY@?*szi*;z zS0merV1WYLTU~l?9|ePWT-43#aSL9uAmcta>qn$&Fq`AXA(k^%XgNwTz+2To#r1&L z=HF2VQU@y7+he<9`sfPkilWe0A5uYWkG(q`&b`-7>E<4nX3xnI={X+?@F{0U1_28KB4h zG}8@=uzLESDsgtd$DYtdCKi~p!Zx4)-z%&1IT`ZKEtzq02Np-Q(I{4(ADBx+lg&d} zT5P^3Lz*G2Cip-k(@KgrB9$;%aTEi#QPcBJFfpXB8wHt z;;+6)QpS-2tGBPEnDrD{OU12t<-E9tUP@9pHx$Tq*~o8{J_zasH?%}?(>o=6tE`^J z<-bhUE50IamygnXu>UHuc$p&@%4#0m*S(Bb|JBcV7Jy7!_nii^$c;BnP{uG|Ll66zFyuVP^$eL*#)a<6q>qTAk%Aga%(KRtFkYR5U zU2=V5`m0N9n}zK3-{`1yue5#j!`Tf$EZoM=6)lcz05QT$dd&+7yh^VVUt2i;_eYck z&{A^~ij%JK_WM3~EqQ*&k8{5D(Bsij#cNNFH~rg%H>}SdeVX!bcmL(dCx)LMdkfDh zUp;GjQpOLu=Hahf`fAxi!A2C&fd_0%fpy0cGli)yZmanUK=TkfAYZR!BO(i zgVn#xjW@^}th|{Nius5lNmN|wo0`1&$=};PFX83o)Rnh}hst39#^x3WaBXwP=$xMI z`re1*EPzlO-_wu<&rQ`-rWGL4C?A5n3cqD+xi*($+C4KjV@<0q69Sv)TdO>1j1zGQZ%z~C&F+w+;&%Dl*LdyGZ#Ar&;^ns${r~9N4h5$*b+xTVSiP8^2x{`@a4Ph8<4O6&?vFh1997gj2(rK zg}jBleBTTSHjRNa5F$g%DGzB;qI0C#ACdw+%EgMaP`Phn@Q1Rc%E>rrS)>kRh%PD0 zNihfMaIpyP9n278b{7L|kl{lP-6|no_;a`ORv%~K#My|kkCVZ&HVZV{mVC$JrRb=> z*U3?KdmJ}5LZ4gtFXt$xl_CbLW_burhn?WC05}jK>2l09RY6RsTYgZbIigH~NSoCq zbGGm?+cHA3jm{Grp}@b-yWay90zMBhO7fW10o}4~UK#_lNrK`sM%lXAnvMum%5H~b z^Cc*3&r)0i2%;)EJ`RalI_ST2^Z%w*CV7DbKvg&rfSWH5{1cJ@_)lLVoe#$+z#QV} z%eHF3mG}dvgwMKhGS5HK^ynw;_r&U1Y#-VxyA#_<7tbrAGefRUzeDmA$>TePTIH?D zr2(cwgH{RFoM@2w3mcJ89Ic%!@Y))-in|=j46aUZhnA#)|NrzE9`I6!JajS|hWvew z4q#`Tvb}s>;aZH7N3cf%S;)>fxp71e_EH1#i|G`zf+EYQIM^;FkUp+{5~z*enT$M7 zd%ZS}-ye{|yFXj=x4zNiXE9AED-<~j0Vj;p_4kgqwz-M#XL`R1{JS5PRvVxQp_|ymncI_a#N@taN#obk|HRIq7%8 zF9pQL1-w)1rCuhlMDo;pQn1ns)S$woy7Zaa%VoPO%fe+cxh!lt|`+oR0oyT@KW0)ToZL(HzJ!sp%FunGJD_=IR2^P+mV?FIj&Nvi zc&PB#7U%UhO+TGRK6hj1b*Yui>V1m2OOYPz#mP}X;k3M-wDGsl6-)=+EbJwkRX~8S zUABqA2-$Tha={f)h>K2WFisI&QeX@}m&O|kqfENKW|fW|T+9=c#ZWCO#UgHL-^%4|3nnpwj9fDS|Qo zm6tT{A1exnr3*meK^M#S0}E#L6znvt^unW~;{FI%?VRjc35U&+y$87Hx?sbPS43LO zlI%~Kjb!UHR^S5k`~#vd)fBUbA{A6zDP0+Ej@aWDxmA)Lb{SGOxVJ{Tk|4+vBOSs) zi2bkNB?&Z_xm7gsMs0&_I1X3@b_os3D$xnhv}lr|M%CS^$w64|mBLHG4nr_N)@M$E zl*2d~;Oq`{=nUxC-+Ax7U2xpCy#LE&+io1iw^PL5K`&!=t1vF3$%nl4t(Z8@oJ9tX!UZ6w^zQ z+o0jdsSEA%N%lP`tD@J1R8Ckhp^bl;n?T}cYn3PrR6}p(7Q}o4wR5`hKp>X~z`zW0 zzXz)MYMS_)0?`vj8E~W^O!v}DPJ!|ml5V3}K_|o{O|~jV$Ku_w4u3 zMds33WCc)dV5&@mIkQ!wJY^fdOPT=?%A}W0MPh_K)3T>&bUdId109>vm|gU(7}KTf zIXQH`s45&>Bv=X0aQe8JL0TmutltBl0Gm7`y@OjSJi|HanJHpv`9=>?>`WRvNM%OX zmH%To(bs(QVP~e<%x#oi4A*pt(<86>ZbsoEpw1*%Yr_%;>}JPyG#eDz-s{TqeO@Qi zZM){(*3{T(=WpZJ&;{IuOcNg{<^(%Qi?T7implmCOt8*szXF-v7l!80TLdMFYEGxm zJNSRDpM5q{K47VFhj$n3yz@nCCD$UF zr0C65(dIZLRuV4^ZHp)cMkiAfRvX_$0HoR-R6}D~dN(&~O1JEoe?eq7w}ywFwYl3x z(=EFlkuE>sn=P(U?E|)G3^QhmI;2Iy12ap3cU(Vhy}WCx`4)86@%NcKyb>p-%khuW zDwoBspRF+!Gt)gJQF`AOBAlkmQgFDdK)-KyP`sdeLbGo(@PTI1gKoTN7LSIq=Q z)NK%w)2eFy@dKCR!)l&cfW_)QSD)Ai%WH4GdRSty9>SZ#Uy)RH4zC+S2Bd8UY?yQu z1O2rbR9uVi3eq0^MGPjWS8;P8yVNING+|{pmV|yxA~aE=O! zSb*oTy>T^g@c9itjGp)n3q(Q}PX9jHKaR9pz1tHM^C?9dsklcHn3tk<2|oiDv{a}; zJ#O4cl!C%C2E_Y4PLW>kY!wDOy5$SPcRWp0zN`;kT4R#)u#^@r=mrI4%5f7xd zxM8>ybpw`!QXyXG_juH(VA7Pw{NX2byWf@&AgINYj6nEpjHZTPPn(;OFt4;Ka2b=S zI06jp4$tOrX5{hI4CBDj*|h(49;70{eG`KDZQs~nTN3KF&a9mx&o=(Tub&q`l$C~T zhn;6epgyEBTqn+jf=}3hO}`z}41%C7>eKV<-fsTc>3N6$YwOp_-`qZL=}Tu{-u=pn zH;%t8Y8!?Vq{ys;Cv%)d;<2&Ei3oUkN$@&ZbO_sT_ zqr1h*KyIX%42rCw;&hD6?*qV~g z;sl~+w{3Fcf7w1O%d~@HfZ3{;ibGF(RUpLNJLtPI$N&^`TA8DsyC4~WQT04khTpc3 zHh!@vL9&8M{PyX2IW%Smpv+i5El<@P(xqDO_mR(%SX13GY8n^I_VR9fb;oG*)9!os zd9Mo07v&3dGHtKGrQ2Pj#jK1?AMa8Su$*@*qcHUB4Sxiy zWpT{R8Lxg|OP1)i8)rMjh{%Vyg!7&@uu<|`^D*|(V(DE$%yGXoUKjnz*Um$mles!U z8`u(#Lk;TEfXg!%&F|yhfAOC{`QuN1`J4ZZ6Vr?bp{PMZKh&@OQ*_Oif7NY)Vuu*E zG^ztiQy#OO^~p2%Ju@n1VKY;ocZ+X9h(>ckTtI6B*FofOiG00$0WSy4%DSl?uxr7_ z0i%pX$dCP3cqAspm1!b&CMD_@-@9&0=l$gbksG_Mc1XLiH3k27l$0rZdGVvH)c6=p zLxbe;&`~|G4-&n8wcuG_W3Y%RUeF0?gsSik+LRB#!fg}JZl%{^-bNrwyrj_JwQCcS zUeI7G!3AhVyrd`-ckq#Nj!m?K^~4Pg3M>D^=7jaHbNB}PpcrJ08MAlttQ!+7E^E*) zE8iw-*==at*rsf^a&SMU80cWyM8!4oKNqLPZk44%VU;E&yh2a`g$suT`2j7;RGwys zxwj2B#;Q4ypmm))S-e0GMl(cQAn zD9!$WVu-OUV{mVT?Itcm3D|=X0mWf^!hYbC*IFgEQm$-(;>I?m%gUxa7I|&-+$@HY zxK8O3DY`3~26833Qn^&N-Y=bdf{&52evb>0`7fNFquKV2?ejJTwum*G0*j(7qu!p9|W5&IesfU55jG$ZD$uZ7T8wXJ|||#PHiq` z5%)x!s%rF5RgVSaTO6;KaU38+j&umwLq5g%172RN(8fFp(C*lxBc+4UIJj|;zR7C+ z`GjI>C{jtqtq(*pGHqZR*-!AOB?>aboHS|ac%A2MX_j|yOp{-yw2M@O&%1bBMlfTOwX%qjWO*%^Z4vrf{roTPE5gkAKK`RBzST#_H3cUxvqAMnJ-p5ZK< z);;^ZY`S7r8~Lpv_EyXx2mp`J#c>EL!&)~+h9CP@FeYM}zb$>09o{EQ*$?%<5xzfm z|4KG*o86N~HroE%9&Fd@wrkEDrY_PpKln@q;_PGr(<6m8=p7=1x}IJa+9Tg5>Yi??mn*v8+OFvw!56{v$epjpd53_9}{BD~?wiC0jRP;eVikW4lfND_ccqnh( zqP(OqMxg-KV&6+^*`^GOACwDFP9CWl@O6y6$aon0c`8|kzH_#vjzVRSs&c)Qfgc=E@x zL&w-zn|>{}7!BFqUuhxB2Mfo$u~$)OWk@n92I|<>K_fMI6n#DyB5NR-m@hydqg{TF zP6Z~wWNwkL1?UYjp)U&+#H+Y?=E7o`^m1u(6MteJTBae%34MJCPIx7}z!{ zlD$rcbxAJ(!9|`zlP*t->6@J?Iw~{>HCLjNy{<&%3xFGCvAL>Q9GiOJGff}m=`@d@xBo1Vp85^I!oN>Q}d z#jo4;JvsjRZSQpLIjh(2zU4a6v;O7KKh}O0Lt$y*MZ5%osnr;L6o(w;z)?i7e@K|Z*y&Ik#n!xrC{v(NiWMmEei5V-+LidfLnLcTySzK1&%1hN3a?bS)5j>;8E`i6 zY+(9$S7u-w$BLEJcyu!4oDX*ZE3Ax}dvR{{cP&`?S@F;NNYiuE!1sj}vMy2#$p5rb zaoH-gFu9ST~o`QpGBswUF)=k;(joahu2{n>693-T#4aiblpYQ+*?`}=4O)@5I*e_Xs{B;5~7AW*C&k3aF{(r*IC<=>i@>6jbbku&)Ct zq!A+W5i(i-915v|F?kRn(?I%I`jU2-f5Dd2XEcEk@~wPVvf)TB87 zb5@3D?A4#I`(lR0@Vr;4t|bM}jo~?HWq9^b4EV|0QP;Oyjua=|@&uUZ z$YkcJ^mGo5-J7tZdPH`DbZXteiH}+-AXDdgg`loVX-u)=sMsW}i7ah;@TDABNQOB> zBgV=wfA3JYz4f!*S<@`0=7T@C?;#(vGd1p8=}q=qnVOvx13Wqj z{xG;RqAR+SHp&XcmlO@0)l(Yzjr`S9ibdt(4S~g?>wY)sA`>4MO5@;?mMc!--0;*Y zH%JytFsM%iBnVb<*G)A)vYy_`JsZAYf>wFMsY~C8p)=3|4TT35pMQ0%MCfmU&X10b zIZiU38|dt`f=)ig=qZwot5mxDBx&X}!rFAgcYz==s8^OKZR77y_sVXGR{EXi>%fpC z1}*S42d?pQ!rS=$L7j93y+CkXx*Ny^mN6P# zWV&3(tAN^v?f#3RwW=m*jw+dpl`p$h?{v%7N5235`@;s>$Y3OmTVYjT9-a1PsIxpTZX@PH* z@Hk^omkW#Nq@a&s?MdV;4?M=b5M)qy(R*G#Fa9`akFZMEDL+b5dB>R@+{@e)P7SAr zJn+A%sPk#$Rl)}SpuCc|%=0s$L4A$4-*-dk3e|qntmu*&)EOLnemD2R2y@ylMAC3U zWV+8^HKFHO-*{gD{p=R?`T0nje_UO{(Ru2`>Ad#QQDw1y`;JYs^*y6;#%kkDefrM* zZ`$Sz*zBC#c$LbsGA}DBCY2(~s5neBX_dK=wE-G)0c9usNULVF3T_|6z%dNLg7Wx| z(K&cmcR+7j>*TfxX*=CGd5Tl=PT4sL@ErM|-tCUBeB^a?dgbKeNlSd{!OLj#X$%>* zw=kS(aCovudA`{hjC(u3`&VM|M@l?&N62C~PA36L-vCE-4aKaYNE#KV4=CXjg|w(Q zgyeFUb3c9c57%Blu5P0<{EiZg*y|#X`Tr9c0C9H6$o+Ozo2@v8+Zqi!svTzt2G%lR z-?34l!CbUiS$EWED4y6A2d%}E8%F7f6(+_Wm4;+J>kCP~C^xc=#7~Nc{m4V0FRJBb zL1Y!K2$u2Nyz-{D2Hl$aPjkWATr8RJ#kuv@1>NWj%d60-Hte`tRl8 z5KW!vq&NpkURr#=coFNN3PQ?&^K&mxbC`1x1k)G%Ji~9REQ5c8@0^IegT`g>NQV8Z zBQ`;-jLVqsUVih;NU}a|ymzsa^)a#WwMuqK;-Lm#JG&$d<=r*6VtQh7p!m8~vPacL z%;_Hd09#WsMdlRH$3D$80t=tuN)G%HB5TA5viAU^|NMV{*4@^7b60ckN*2zph&sht z!UwxEtyYptqY}2}hG+f@2?7J>pX(6Yb!g0~(}ih4mUZaYTlX#}o5w*Ea^Sw9lwyEE zrjUy31%1yJb-bWhfpMtgYJ<8@oIq~U1*BNk8_`4S#CQyhT9Mbp#j-SBJ-v9`r}Wy1 z*MpMA#S2Oy=GYlqFY1Eq+mc|NI9{-Pl0F9i+c{F&!$D9#!Y|?@BR|NKM|O#oiFe=N z^-8b5cnK`1d9QT*A4#$sdoTG`aNa;M>+r>zYTvqm#kt=D30`}>duAZtbOD9`Z={XE z*LW*IO|n%|9?~lrHN(OdwyXu}cVCln{iU*>*%I+RQ(b~~*d-0hUP-qs4HO%(&1WGG z#EhkRsvLS-NS{xcuyS&qBAKf(PVObB*OW)zGf2$_9iw1nNhtoy|{I z0^B~RSB$+|Fw?TMWXgY>Ps*P$B8yM0mYaPPQ%8|%D(-KWMfuClun(3)Rdp79g3ES_=46(*(P8JuvtU8*L-OkqxZuoDL6d z_Hnf8t>1S)>vTEjkZ6@yC2OQLSRK;n*(@|D*UzrvE@N7hrvB6IvQ5lrYAdrw!tf@R z^Z?W9N8 zF2nQ7-cQb3U_Xbllg+`EQ?gWhp;qJ&St73st@J{|m6M!3o@t^s{<_djk}smnBWg^- z#z}%*kWfzH7}P!Fp{xz`K+cgnv7K}e+*3J4?^PPo2)3w5Ryk!$tXA1K`y>Z{LyfFt znCVB>__Jego^qKr_V9*n55m4<%Yxvx`Je4p7Fd0W+ahYH>T3C6S>81#cR>qIGTI`#Oef{w23c+t_;^S3Rgn&R}QVon$pI< zG4Yr*Th+$z6oMT3W)MeS52mF-+#pW$Y4J@HWzxA$Ewmmuu_Hy6t7=q|>m)hV z)N87st*D%=4}^OnvAwX3MyzSlV)8}l@-$}Gc7EFL>%%RUXi0k4Hc~j)#@u}aP~?!6y{V*_T@=|t#hoI( z-XC(><%NEws@t-ABvZ6eu{mZx=?K#?uq&tgxGA8){&4CodY!6Wwx2BbF96eWTXjxl zP-Eoyf*5ZvSxicb74J(ll;%P`kV=pj&=ZtW{#uM+&C_DrZ_a7v-ntw9#jXej0MRETB?c zl(BRIIGh9r>&Ms!#~^!jYUk4akH%X7bLD@|{*^3Y=lZzudIard1K_oeVqmAZl8S4l zb-X6(pVyCz@4r={L_P)ezRj#9P)6O`SigT4kARum26Y zj)yh|$E};%p~9$56?iY7bF)OCGREnWZixc^4sn*~tQdQN(9?bN6pY>+7A}TGup_FQ zYaUYr3{*z`1Ms+U^v$p%e1rN?k?tIMB}u{X|AtEw7Ic8kVYuGUQ%?zGubN84w5L zVtTfbU(2}`q4#1t&n#d6H9`NJWH;zQO-Ho^7c7)sWMWwPf{&jJZ;GqX5 ztU8q^!gajTm|D)B&=%!oZagq7Tu>OuvdGdHT_j{b>0Zef{BB5bUX<$tTt~2AnP=8O zd2B0O3zWC}#_X~Mira>R?M%H9Y9`U!HO1yKf-lfz0j0pJv5Ya17vgE7Cn`C_LrrC| z49azpztB7mRt-rODGP9=y%_$~0|s0;`>ed&vHx}Ehl;5dPp0Qb(nVyG8*ht1B6)z{ zT0${fDN;biB?vIf45&bUOXyLbRROL?x$FiFWGiOvr$TOL6c%sq3oW1*@>-P5!fSJ~ z=PU=Jo=g!m&7*b$Cb9EHC!ppkPuw4sAc%)BvGdKxh(Pn$b~qnEtUxp7y8N??wtcH^ zYhCPgqNdNlTI}nRFT`gj=BipHt0T)p*3WLE`~DBAzZ^t`3ysF$k7k96v9JHf&(ojv zk_Tk81Tz8!nPjiHE~H6{`h!b=QMuCV0MYtr+Mtfbp!^7;>&~|ggU`V*BC_@#PoQ|_C!_xTp=%s;v26S8z1ghL1V0vQyu zh9awgzeJYE)Id7!T+n)mT4AGk6IX*V_L|Vj$$KW1fQ|2oNf78n&U~X!e+ew#_tPKd z#82(N??J+$?i5624DNg%hDV*rQm%ckC<4%YQZfkqNn&mrU~Ac%Rc?W!rX9 z4AixhQE>-V^}&7OWswGDpZ5XJc)_`#>%k|59sEXFsW6j{7wEkKCEhyGS>dX{V$Ocg zJWemsgC=Sfw}U=Cr$$!kwU1j2-St&;H@7U{77&9$4KP*?V9A2{f#%0J4^qPh&l4C8 zJQ@rK!@9-_JY)L3PL__ffG2Tr*gk0UWcL7Q_#s}ui~H|i#{KH|Z~fx!--#Dd%!d?- zcV!&Fq4z>IT0w0&#UxW?2^EL72dU&bB2qYu1=r-G=j~em*$&{x`i4#xr#2O`+%-AHxPTLbjmvLbn7FpIIY6Kfr)eg90QDp0}pi4 zkn3~v=T5em3)QkE)nua^uQj`^3_u~pK(UgZid)FT=0vFb?~bXMr3F_yU$i1TL!wz3 z-WFjJq`Svyf_|dRuzn9Sm3|Z7XeupeD4+ zd!u4w*v=`dB^x50xBfe6iXKhZz;`^tV^&k-zIoNQDCRZFVrw4wJUB`oJ~yrjC@>F* z_NP$HM-)k-;!HFgSS_bjUh`e}!okt&jZhwj6|>S}?wPn%@&%f7IR>`K zz9B*Knbp#9hGI@pPply=1LUqi?dbvg8ROOQHisYdC;GR zw_*<>{t6A=+cN{1pfdbhCq7mOxfQZMppu8mzDs<2plxKG>JWqEQ07NhaW^VP^#^2fp17T|cS3lj2@+MsN;S&sihd%e^JZgbmvXp;mb$sz1V@-W_rjwtZz# z=#oPh`7Q8VE4dcon)T2LfMEdx$A?BrVoGd##@yDavC{>X?4^q|8;N%Lhn)1VZaFHE zHA;$v@a6CGX#r(T4R#P>6YHpjVcAc(LlSYvgD(<|zgtJejgLKdRX%9V`XEpM-@)|;RDl>ymhq~s={*?J~6c$f;giAhR z1q$eN~qr z2m+OMJ&mN~t^caDm=X1%xAu@McAl&oBdOZT=oC{-5k9d+77R(553V%gX#}Rn90|J^lih(dk1}FqV)pCvsQY|Q-jrE4D zlA{xKP@LKe5~9^pZby{H#7|r=|46xk+lq5zxgR0Fg$D0q*mIQ27CjMr~fnyyO1+L5p_T56-^)$!m45 zEO@z@?)8QpCN|SGNwIA8q8ycL&7@thc_2oU(}*XgFD(BRY6Pv^x<^{mqBC$DUds0sBSzGm9qXu4--;$ z*8+PVSOaG0v*!d-(Z4J4v&|&hUDgKqvbga|WhWt;_%8=pNr zh2zkxNoF>LEE=a()=j+>(G6+OX?5czfhc{uJp>D(-LJ+ zBT?pc$?A!VqPyt7T#{FSGqqW@$@l*3&WWs{ZVzPV}JWSjHA0I`Cf&g|ivp&J62lrmp}o&L zg;TAl2N(0Ad`F0Re2OSZT_|2JTtNzAuqpU#cxlM?kOt);2J4|ekKQxcpvL<@nxDei zLoW|BsJC)!z>z43>EqVZ=6gp*Uf}FWu(CsJ-~52_tn85cep}nGd?zm2f~B8+I!Gw z-nb*P&*pi~8f$;{2XRiudN26Xn;sUcQgHVdetSVS3@~K4yxxD%xph|nHE%ihV_RhzHV&^FL&A=HqFExZ3i?zp$xV4ftcOY; z54%7UC3n<*=s_Dz#XLul@x(hC6kIO3xcO=!dDa;aI_IAkpO@E>THcP?S7+$GYC;!B z8r16pyXc~6nD2BOY#JOC*pJ)F8H z`rb%95;rcPvg45)7GkhbPv;BX!RGj0a>VEV#wBqG5+j}8VcbuCB3;C@IH}XteW!se zdTyN5Y%3>qHN~u?NGcU~hF}Jx1)Nl1%U0G(OwzsPiiR$6dgLXAGcCc;GvW|1%%))I zhdFV?C#R}k6Kadnt|bdEyb*-$XY5+2U2$!U#m1Fb9;rY%o0@tT_Z~g zuaTV)c8WLnqVQc2-6pK34a(I{We0f5znI0Fb-t;6+TX>-<9FYn*-yXs#tMH6Kz?*= z%yE*z&T_c#zl`j(!g)T$00U_@6}P~zMs`wI9t6@KOouE*R3+4VeGzkn?4paH!M)38 zNvPKEDmW%BKoXcD>h*3D?t?s4d1QIe+j{!)S586hSxIoTrV|+X$-Iy0 zm7a~jL3$z}L4Z7>`2iiWL|(c)KJ2)E9`|E9DKvXdF_dlZ2&xdJ^33UnN26T>nvo&~ z@G~N0*t>cDj_h}ybv5`VDH60jDyhbBs!49)67b}-%aeg8y(Hw)s0PSbO}4$6alv&~ zTi~%t*~`Lhx%J()jMfg1{kRCUhm5@v*#r95_oqbmO^254wN>; zhwXnABtPlxu?5^UFJPVqq!B?CDwT(A>~|xUflAbqCzQ#0#i!?N3^@r46NFbRcbyGy z;X287u)>e^%3|ltIbdRiAFBNI#SI_HoTT*m&_qD$o8EO1%<*RQ=oPPj2# zZdt+QGQ|Kv`UT+17jEXJ%hwC)CL2LS^?5t{xs!lq$3_kRGAp2sIVpXu(^d%j%Z1)< z9E7u@T8(053q$vX2d85NA7~#hiu#a$+9F;^HMXCTHEkcmso zKJl!3#K)jm0qLcROi2qJ?}j2*rR6>dI_7^&p!e#IXyl*u@0rmx)u7gah^6_{MWFAp zoLe1g?id0?1Ef`W7bKG;34%{lpF)J!e5F_c`mBDZ6nIDVls+GW+N9WO8huwDH=QQYaVlf2Xu^H|;L)zu*LJ!jCL~CBq@$N=nmjYLV<YCWvVJkJ`EhgrXZlc@b1M-jqT;Rt_h?}_u zK#6?DZ>cZR^=D0Kl41^NHo;&Yv_Egj`y(-G+0Kgn^qONBF5aa z(Zx&RWCk1}Iou0$l`soLdiRrFf}^{+dwJ04B-_hvnYop8O+6=g;IoWD0^J+L$loE^ zJ9(k+17LM&<72xSN&;un8eZ+cfx>69MxA3prpp~;?O zV)y#fA@h_^S)qygrvJ~9Ua+9aL!{b4a-Oj+qfe|5R!%V`6xm9}9sTd~P|RB{F7$0> zHcoDj?hjfkX_P|#3FgW5SCOFonLg;h`+LM~4EP=KF9=c42e>7sE2^yUgWc|g_7 zU*Oe1TaV3zqM_lVX7Gt~8kPos{IIXu$I|~8CciPuf*bGK4e4Zu8(X2{Rv4+Ln7tIK zh0YLWkFYaXAJQ(Xpb`RGnViTvrd>Ap)fP1r*ueiD1U1K=7NS~fr_i82FR!3CD)M|w zIOgjg`&3VP53{I$*aQE0=yBsEI9vh$`K>`+9Gn{3Nt?$t@;imMV%AEI`L{E}$CA@f zGq@3Q0cr-@PMyo;DMBqsD(NgvAbK}O(mpFB?VuRYT`Z>JP}R4dPUSShK1`6zJs{W- zg4;cVuer=(8Iv2Sxd@u!>2f4w)q~Jcx|d#(C+PQRl9mLg64c?utza&#L8YS<(ZU%I zzM@s`lU#_@1H0W4X${w@{%q@n=(fC-5(M=f_q5m zn;^&&L#pbA=N@Gl?DT=|FfFE$-W4?pWF}VgJTcqBN z?TOLKo}8qZ7K$`cajQsoOu-ASOaiZ5+%C)Ve!zRLCpIaxT~_G(p&(176aV(hBhuw7 zJ#mOR7W@Drc$lP0k8I>uh8M(O%aipZa-;I1_5?2g@48}ESG?W_BysbGpaW!vWsGiPaRtG>G&wfv%BokT``aM1lS~9CTbSH?F80p3SdR1u*DkMFT zxxqeLtTDtGHHIZojU19#@Yta^ALmozF(xse6 zS(WgCJOPqY7sLg08($mPMmCGvIim+BHdq1+@wDl8>SsZBXoO!EWn;ZlOyj_50+b4*ZMX;+jcL7=DigV_B$<~;wgS$B7jjjrZ{3(zPmnfj{^)`J(x!ZHMXTAs}oFGAmc~txha24&K z9r4y#g3Im-#O{Nkak%S$-Z*T_%jmY*2|I+5sNnce)+ukM(?B%1Nm{@$_$NwJ!D~MC z(uH|{`23}_-~y+`A|<$)m9|^V;iUu|GdOOWr1F;~*+K z(5z=s475kDr{WAuk?@)?h|}y9chZ-+3j~@b&T;yhZ@YZEY!jpDm9+Wf3*h@x&AUM- z&hGQ>q~pCb#a_i;U-J1z{BN{*z_6ym^kuVokdseYO@;fiv-IqYx@yZi#jk#&`TbkJ{K@a$eXNl> z=n~*$SU&A>>>wNa?&|k!tp^+5){VUvJ4Vn5u;~v$v2|3f*3(*58(k$eO78ojU_lDG z0#vg`8Vw_&tzKbPw;l~-u|m`%m^}>&PWmk1yQM1In&L0_CAje}$4=#NvAmK=5nYI^ zr!f{2FW4Qs-+Ko)gL8w<2);=ta*F9nuZ-Yq(~(w*MHa{WQ#h&QB})iUM` zNta`u$;iJ30;Z-slUbV9ymsX&HLc+(;f2vR_~2+6=~fBqa%r;23LeVSYT~0C`H7tE z{+Qr;z)Rt2ma0p_vPjYkFuIA~X3{>-6gBZR*-&i}FSzP&E^IvE*Bb$%O!b^zNzL$5 z=FY&F4dlLbY6Tt#PeNL`f3+}k-hFpOm)5;@bgpIB{G+SQ>&Y&5 zQ9C#Gb_`aV|AQ0*T=aXXxKcU^Ov`OqH-yqwPuv}QE@%Pgpz33vY^Es0pza|LWwj6p zXi(~SSmc*OUzFeSejtD7aZ%nQyc1hMXQ}$5ANcpgT$G=lqg5VMUF5BrxhNWqm_d!t zHu6`^L@E>W-R6<;KnAGV#W_0-P9PS(v4kLQ{@NfQR z^qgP4YOyWzD)=3wmYr>J<306xD>yV!40t?;u=xfew>*p}Zuej9(@5_LH7Nvb_wVyb zLff5DXQyaUQ*G?@T*= z-ACJBU(;*aTl<>!{?nOudRxF3j`C7aK}~og4<8UvQG$RdK4t_Q7#Q)%Q^!FO6p>N* zua!iHAkiF1xEb4zA0;_wpMCJ`?^%28wb%N7=U_Ve-2j+j<3!oYolq!cl9GkGhOjDM zvp-xQ$Y3zOQy@MZc$AwjxJ&dx^nj(S=6@`#;v7%%vA9Fy&a9SS2S2DX6nwqo?Dn6lBB*xbxUnC1T|;b6EKQ{OvKkVf=QoDk7qfs&&X zQ%Mm6rLLN1P_!@^5>q)#Bgf?Q<_mgVJLItu+n25j?E#|BE|3(~hvz^xUOzXM$)=BS zP4lmmro^|6#&l7%NKIf59KMS&YRp*TWz?09{b(I8SvMDHe`dBYi=sLgkPA+1VLr34 zFg+9l66)QQdIK@IRYD<48n1j&Exnd^1US~p-naxQPFemB{PAqIV2==67~D)}i;QQl ziqpg;;s^c*#kYbhgiQ>VwPZtp1+P2;!q{5nVgE`x&0P;=F+KEJRY}mv;9-|&rYvyx zTwTD4h%R@7TOqej)DHe~tmF(gfjjJy=3Nv~C(;F6TX2``2MLxQ;jjy4QS=D&ILASb zB~q})^9U!?zfyKkY;fy^rF}0cf_b!n#+E3?ZOvq{#Foudka>=}%UmM~= zN?>4>WQXfn&uWm^X$rhb*3B5AZ~G)Ij^o74u!aLG!!VJ<6NA)5?!@jIWp&2D*op-! zg#noE-!I!2aguYIDGSpi%TK!Q4jXYXIhW{nCK5cd5)}uM{5t7GQXL3-U69$0?!za- zI8LNn47GCQ6=4@hTIaeoDiV14epf;UNoJ_Q4YDSKwmxsq8DwIQ-0C~Z?zujpHLJJS zd0mAY`QKNK-GOl0xUrQz1r$$-A(zA&z;$4n>s693%B8Qp2n(SeVSo6;gPJnBFuX@P zB-eZl(@Go3o!!b{V~ye4mtW~zo-gX+Hkq=nvLM{k4qdIxrq=`+80aTofP#1WaMX0q z_Qp;+jCsW9j%zmYhUDl$l?NpG*}Q-njVwDue0=PWzM+ku*+IN5pf1nbJW(~ixql@| zVK-4Z?>j;cSWHxT6q7@dOiI0P{x0q=n3OcP=pO0rFwLH@f$*K)8r&5%uGjx18`A0$ zc;>U&a1a9k$?{IvK*!2p%+SlFD?)6`y|Z!9ESgOoSr*;d?aLBv_XlJ1_1NrtI&mb> zN_L-#b+l-9A@#6o0Wv{5017)?v2pZwz6N((~*nZRE#Z<@oQ`r=7ET z$@4wUQ_T0Ob~cmEPQ3OmvY0e7DJG2~J1KPreRmcX_xH=o7p>++2A>Av>Re8ryff@H zIVP%?)D^{TiC806qABsBy}F=7`KQXb}8_y^u?^di?fR8TKWX= zhF%u0=Ja#>7Y?~MhGi{F^2=H_WfK}(qvddGWxj!Rqa}+`dH!RGiJ^gPl~mD#LQR?Y zx>3clLfB4Y9g zmYtuSFVKnX8EG76+*kl*O1+}5+!&}GJHa^XWJs(`CTW~P&Hz+em-9baln#jsSRmI%MRcb4 zus@U)z`qB@7&~bMW@+Snm@-Tl?E(u+FpeE@+#%~;#yCr(>6{<_PwiM`Tc~*)Wtks| z)SISkW3s&|%|2f+CGn`o9+gc`{;{hZe}L7Kcix$KPh#iKrkmF)tJeh|BO~mRGM)E- z2GN$$$d0F&jTBjri~fFyyPDJ2ljRk{L~3eni5Z)=0Ne2}e*6x1cl<}|m&}N0{ra^W zQZj>_wRn^@6jMc!3QB#9yPMpVVfshg!UUp0^7bqxhOZNN8r(QL4D))=!Yauw47E|{U)qot`KoUKN!||I*H1G`_s?rSq^-ERATFnObQTcwLL?SkP|Ia`9z(k!Up!!wnBx z9N-SRo7&jZ_S-x}wpbE7ju@*Ua^8%wOF!4%8jBq^3nM4qEwu6q=;)6~pZ76f>$$i5 z&hldBq6lP>{pw+x0<*oF1J8Q4@r#$=Tn?Mb#yS(s9dGx(y{KMTyhMXRf2@mz426*| zHOxIY8x0tmE~BnoE?pb6Z+@p_&GUm2!yLPE&vD<91%;e#_fs~OsJCBbvMRFO`O(rSPAEJ$UmcTlQ;_s zz^;2D^TMQSea3yVPd#W0B%a!d0~)Y`#Ps6oS;1d)NaTbbAzEb&(cGqQ8w-H0N^eh$ z6l0RZVGAUtj;QkU{|qZgOuO*Tcj6YCClj6gC;6n*i6@f}Ehdv=6jM!+Bb55&VssxH zr0tN9f(b8K%8C%3NP`3$J<|13W0oRn_9gjg@;HN}f?KAn;yqY+2eM8Yq(dGXJxx>` z*v$|(RRYaW!A~!1SaKnF*H#JXL8kXsiOCOLJh5!1)%a+86ikZ*o z^_~NeX@D7Yn00+S{5GT(oE1WS8C@CF{X$M?H5A`tM#{$^W;^1d4{zh&Soj&ls4x28 z@IfU(Lx=`{hfH98$1fu;rr!&Zoi92djTD@aMR{RB8e>cip0P}&q=ME9i@2Dehw<#Q zbQ&*CWNMd7NBL?J!;Jzf0e$p#g~fKJz3^zGReNf)hd{NC$8{woe|pP|y_ofJOGpp9 z$`2c)^QOVegzwNMY;v}6w$J$*P5$Z8c^ zIZ;6JtP>4NTD>}E*dy?s-91y$y5g`G+DJ!?3Vl=XZjf|2Q%!-)R!<#4bFPrHnE@n-S%+R&X_A=Ghe4SuGV4R3EGBnE2IoWC$cnsq zt^W1W3}(gDY^U2gjg9t-AC0#IKm4Hj)f|}_9J<-R`8`Qy2M#Bmtn?P($fg*`wa`&& zSiZq-1lGLWUUUyAS}z54EvTh;GPjr*0Zu=6P|#Gqj@JY0wq{|stdrjW{&Ru2L*BEn zZ;_*o&_r-z#Eb(mf?cb2Zn`oy$oP!xfLjSke&XJ=1Z6jj+uOK!x4kQ$MFiior;%vu|O5?4b-+Jsona(>u)+1!yjXZ1>v6Knh?u zuGByB%%P7EV?45kRBo(#nlSmejW%toPaJI%#?fd4Wc~Am(e{hWZ~tm+ zrHRvSX<4Zq8RFK5_KA~2lR=)ZK6H;z6Dep~vMm4!%y8M7t*j5tB3L5Q7*Zb!u8b>0 zBDp2L)gT~PKo>#mXyU_a%X6Q^(A(1egxj$5T3VX40rrLKJmb=o(Pa`B}F`;@tX=ayB{!~UINAn)eB#;4imjC&qO7Z~f2#ybI&FQwvy zDH3;^0G=j{5jGcY-}K+xzUZ|`cj-Q1z56Zt;m=0Z{_mB3?=~z*VYgheO7H+&RUMtW zbe~7O+_6?;)RSjB;G-^e=toutgnCEvM&sD}0xOC6S@iovqnkM_yCDT}2vLk6=AViA zqeqg}gLmHLV(C{@;$NAkl%QWf_Z;bR;^@d~i%{DD#q?98mr|oTQoaC~ldph|NrQB+ z40S6y=`8rRqSm`#zCu&UuLWue^gE%|5#b)84k%yt`!p(0{Y0Y+NZ_S$kUP&*jWR6L zoF?gWQKur6lN!+ea^%tkUTQ$?@<<>-ECjM0J86VyY zH(a$!2>s)xt6YPdTl;;E@alYSxgVJe+|MXBi2~c8NJ)B-B)TB9jlU)cDXKbQ8y@d& z>K#imVkl70@ZKQRoCo2JHonsa+St7jw|;x`ywD81_1iE1lEls+1r|=gPKp60_pOu~ zpXRA~@RSlDVqwm$#Z-ozxjl2E1&3aMGv)rhlBRh$s`;W*^QL50U^^q=!eO)lanxsa z!{D^E9kSii1IC7JtwcxnaxVxQq_8HPsr62DkA@`#IegE(e zPmPV8JiguR#1UC5@su;-E5co_&A^h{qH2^TQNK;3T44Qy8$gi1=iVX9T2>?)qSIh4 zbdZxS%Jb=vp%jHykt#UJLyUot)R^Q#siNBXUMRN)6RWB zyghmwFpgq2P$Y&@r+odAI5M~~v`&;8J_NMxjvZsG1NHR5w|c2VPgo!Pakc;{d2+-1 z|5%DQ(E_Mh0ueq_?q(Z5IrI?yX;5-#r^ovdCgvYa8=c|TDQT5pLsPQ`X+?-ufpp{9 z$~=Dz{wL2gof&ay^Een>@7hCSksB6}7s+aP5cnMoKj@3(lWP!ixzC~ zczAnU6Cp+U5M=SO#_`h;XvO%jJj@oZBex>%E;JjNbib@EWDmOlt@FkS$om#1se)n- zQ{*5f7h`%l$XS0ZY=E^Fgo+jqghvYcXiOmULOVn7f3Fr~(X-c|_`A+6NWhB+-9*pG%_oMxljMz=W5 zC+=CwHf|p==9`;p7;-_K3u>>=LXZv09NfDW56coGvc#RFE@%jJZOfKG2rj&WKFxF} zQYnL58fE%^(7i?Wse2N4km$q9=XWiDnB62nX9sjynY<>4u^qX$rB9dFa^eM)6?v^l zK_#z1oC9>ICEQ&cEz=p$qUt5Jbh{K&TJkviB5n#VIlkz2)Znl}z^LmS{n5I8`E9+T z-{YZuxj*+gRzW3$LGs_{+ z4S8m$DUMrr*Lk_~vD)mAGI{1C-zpoP4Xelg`735a&y#Cn{jra;T5(>-P!{`V{@6@P zr%hh5k{ekstPE^iZ2A|YEID)(xS<$p>7W}I?+m>5Vu>t)G*3}(fe}{hP&+v!vbmZS zYN?<7KIF5p;CMVb=e*!1Rse%rby#Uz?A=Lc2G)Wm>}f$EXCrs-oDR-qalh<>@K8jO z-|?vo)qe2+)AGW9-W|Kc!zP&T#IUilQMKCr8viKJ?ZuENLB8lH6g;9_?Egg;zl~}9 zY%GI)+8%4oNS#dXwG)HFO769ZM+i0U(Lz8PT%z~u6-di2jPzv-aI6Q#&a8ayKEWsZ0K!Ty&0dkIlM$_Q;suQr(W*9n-K&@k zxyz$4JXN_WcxY+5s6};_9wbB3W1=Funzt%Ak)sn2yF?07{f?7JK}xt|v0)yXi9p24 zsM__gI(>`&i_YCp5LgJ&_?_OvE?7Pfw!r6BIMx&&3~o@=OZ!3UFfJs97wa`_tbuQp zT=kB99x7B(Z5)I4x%1n-S6(3pJyxzneY;3Oj}XhBS018ohi~${zR-4!c^gJ%A~>)x z7#3e_xS@?rSh=jz@~3OBJ!f9?`v2Cjf^2i*gt+|{i}76)lTDEf$eIR(RPy`8cfzxl zWx$d>ZiEW=7S)PkZiBR*A1OE-5x4Xbf8d2mzK$N2)zbOgLEcsA z75^wNBnuwk)bo3UCjx7Hh9U12*VNZu9N=Wrr%AkDAtXSYrBCwvB3e|}NFB5|OQ$P4 zIIB37{9^A8`Ce`>eF1df&eC1}C+9&94QDSmg;Oi4<99&ILw>35Evmg-JRisD^~N)J zud{S2Q!mYRYf(i8PRc}`~}Rw^~`)H;NQ$3{I8;443UNz z=4)3nx9PQ@O#_MCTRiG%WVt`et(RiDbPSai zzWE6u{fXKW3%_LIU#!{~cJ{<+vz8uw<*5H0Gscv$>yMC~PV6~B_QEKlh20bb!brK4 z`fy;Uyp|v}6+&Al!t_h6VqK_Ki4=RtGi5A&g*sP1t>U_86z@>L$|~OnK5g`M&+0$} zZ)F_08isGsMmPMbRZMoJu7nP`UXRnrOaCwKcuzKP3=VG4M;JWYO26n7p zzQBRLxM|2I3aQBf=IK}at=Ql(*2lIY=DUfc1an z_(5>Kzo|{FC#ILE zo02sE&MyVSWu}rdDtL&pcP^7tH!Tjxc1~*d-7~Ck- z6hJxRPrN}H*|$9ubL$d#Sh6`tKJZAMxia0eLs13}7;Zd(!;}-+1~fv+a-gPr_z?(U zQsyOx76>kKFqk~zf}x>Aj|#tJV4zEuZ<#wp>z^BT*#kL!244R>yfsW}IK!EK8%)hg z@{HQx?zmg@1bcMoiH;BT%~l=|)w#!&dj(x6S$`b^vzvwEe4NwPkn?TP$- znY-Jq z%p(@)ITC`*$3VHIO;{?v5OPdbN(>=qUOFDuK>h-v^mwqaEj1ig}MB>nQcmQb@&gA9%4D!nxz5ptONQOf zMws!iHCL8K_On|^IkDMjw7|o0iaAP=N=kiEyw$x|mKOBzXQQg^)$^}wsn}N=P+Ys znXITEVTsUnAom_w7{|+eL#sHzEfJ>tQ~hYv^Jg{nJMYqD_gC5#zGma{fH}RMXr0*6 zEw?Zidng7NC3aEjLHC9bWC!nctrM*gY3@sRd#&{=Uz80_=PkN9Fl(L$xjweb`{w1+ zsBSpof&xxEL0e;p&RSNbL|%>r-bq(r%QSIyJEjE1E@ojA0`2vY9g=^XGV91}Ghn`1 z@L#_oJDeCWhb#b7NHKX7$$@gEM?qjnT*r+F-2K4pi8NUK;w}gjm0)U?DMV+~<;I9z zF34rT8B{emvRE(Fq>1!Gts;Hy<UXDmAmLW;{I@qhk?KY-2w!qWCd-!+lY+Qq#snGk+8PcHJZf3XOjvI@D`&fzZKa2gq>ph4}gsc*HaITX9O9q z%@65(+iWw={o9&F$>d(1 z8yVawK11?-H-NlE7q^*{!0lRGFX?haMI}=^48%s0Ec!#yuuJqE;QguU`Qq<3$7Fs3n6; zhTj?HI5_}Bd4mhDEYvDKP!w|4d)9$a`@??AU^k-zG$k-3eA4S~L?wtgHIrOUPHgh3A-l*eHqsiJ%0YG7~}BsGH1BnN?>xd!NAHgfj}TU1*`d$|do z>D;@JRC-_j2^{J1DT=5hyEv=-_XcOv$%6foIv!r%A&(83G%TzQ-(=xpt>4KWX9X9^ zP4y-JvDN#}$c~Yfy8gXftj_I_|8E6dChGMD4hR*N=T_2#l3K_`ERx3r-H}ddB$6GH zCW{Z-um4--{>x)kDA{1cd1I2qiasTl<5tq=y+E2z)k#MR2Ew(9&p04P@S%6t0!`0Hto@Zwx<0N4e=lr(qJi zBg|WR)ZIxH?kkHC>b-Z61G>Z4T=Y!-` z8r07{?%FB2;0E;;=wr13Ej1q4J@?ED_DB|GXN@d8ohN*;-Rf9t$!cDke#fsf#LbMK zzi2-C9}+WOLYEV7Wan65Y8%BQQ)CmRJ_>ONEZpsuCAcXdxL{Ii~C12G9OG2)(N&C;7;_krT7MU|}_mNohqe46I<0UP#^ zJZWye2wBuCBNBM6k`|~lKkR?s5&YO};AGw0*dJ8ymw3N5mWae@*X347M-n}<>DB)E zzF0h?Ror9F33fpx%ogqth!PKy3Vs4Fj`Wk1f0B+&n`Yg5jF)y-x5Jprj=;afbB12dpnGP;bN@K(vWjfm7)-OG)y-Cx=>3xAnJUI!txyy&zE0wl`7jo58XoRhLD@Q1m}e=Ch38Ou35Aza3<=vo;ZOvO)A<`;<7wV9TZ@?WaZ5;# z6MHoqEVjfSP|OfT?ow)#5Y&0leM<{9Nt&PMRB?>zZw=C9U`%UYh(rTA`h7(ke;q_$ zE9h<+(kgY*odFQ41sQA_qj(qnZ}@!T-7hnS_K>!~bj1NlVW=MVs=H;qpxTxT`&h|x z4BMo^(Qa8B2l)u_FI47bDWiG0k|e(#qv%|PusR6n%tL`B!6Y8nq9}q8@(@?c^mF^X zyJ6|pFHV*xaO!CdA~cPIY;VVW4?tXKwN&c?%5V5zf`WAVGoZ2VkSF=w0(Y$#nBoj> zP8$PeHQ7w7P-k=oKAf#|e@`SiGf0ia+q|D*fI@E{r9Q_*d3~a+@@@By6AimWGxwL~ zL0GMNUi-pf7j!`N!UM{!Uy9@D!}rL}dKSa@pex(RsUahQAHmqCLQLestDh}ImXWK{ zJdUYF3I|hvZ_;-b+Sw-9J`IdBGgurN=e3>0Ay(t#ynbo((R+V%H3O&iOO<~kt7iZQ z*63`(B#KF-NF1e3G1PdWx@k=$ z)rq~IgBI2;pJE_YB8yUQ_v?ddm2Uosf0w+P_kfq`ex6iC=zUWC@F=9&%t~6E!@Wk3 zpt_CUv1r7<=f!Si10Pqusb2ZZqBy6N$HP9g_w+Y=xEeoLFJ7fYx4$DWpgq%C2;|VJda1L?`N$fFqlw+~|hYF34Z4DG+2Z zrn4v=hC*P_B41PmUhI1Lc8Ci;kT)&S6oawaj+ySt9#_ z3wryv)W5|E%!v7ee&#!5;|v0XQ=>S7c2G;=?|oq*@Pw}iayukA8(i37 zRDP_N-7;!BY8C5euJ^y?ZvO{p<$uohxmxMQ?uXgEw@w?+q8_||^HnoEcF*J7B9%@& z`v6PeDBRXl%m)-XMXA$a7ZSIaH+UeYQo9@$!AO^87G!GVB+K7Z6$c*=j59I`ZdVis z=L;GY8eD(jmt^_%f4{t3gKd)KwXat#*EG4Ng9+*9rq9I{Z<^~>qG|H#b&U~3F53N_ zk6wket(P(&p&j0gnI4kJ zbL7*0tZ!D2-+7(B;OPBtzGBA3n$6$-PjZBvJL|lmA#%Y2-gOiMB`+r_btObt4N9H= zWzh`?4911T!TvWEvkXE>60V}sIFZ4n{#$10Ra@pB_DqBHt8Bps0LuiaYtzT!Zeuk);z!3zu9kp#2O#1;;7&# zD9_cxy1khbO|rQkicp78t4I@NPmte{#f!mua+=Hw#lQILjIr5PY!YgmH>zY!5bAPZ zw4f+r4XKRS&8&HO(^uO<-y>^*vA8&(+XpF&G)0_Nrd(M$X&K4KjF8m>Y{6l*BRiis z@XH@*gl2QIe*5KLlGyR`Ae`4+65tdWC zRhuj~rmq;uY0E?v@TV23V4FF8KpC9lcSf{6;sMVPvWXOiACu|kMb55t-!iKc0({&3 zI%P5PyAjp=Vq1p>L6sx zuMk~X7zKq2n8KxJyJRvC@dU8Oc3K)AOiwxbQ)+n(W41W%pADQNTb+1$U20)naw!IK zUD7G_ySS!?N|LX}1F1g7{wwGjaRvWKP>(RlZ^Q*1b{*ZQ*z-~id3YXzbol~YJMZyI z<8?cjDrHh*|CI5Y%xUX(W}L|G^xBW8 zh_B^!$~5)-16(BD#q7^#nFhywi-V_XHO(3wlNoXr*Rn##wDi9< zPk$DRlBxrb1Qq(iZa8WijJV_jErd?g1M=!4E?A?4wMvd5JhL`Ztl(m;69*i(4wwI} zEqQyaQ|q+hHY*M+27J1>QNUj&KDqdC;2Cih2T62RB)BEGtuRsPXg0caDDnmG&jF}R zQT7}wEIe{295DD}y0ngkaVC~&Gcs3S_{U1Q?1ap>d;t`jDJ}?gfmIScXkEOoXb73A zDQDDh*#@Ce7fjJd(Z7E3#j=vV0lYw%XxS>c0PgGm0VHhODV`!6 zjK0mO*}uQ>ve^T@&rSaksd#3BvF9!P)iV?Wb@wNL%Q>u=i<)gHP>S1iNq)M;hhEsm zY!&HcTEFsu1CZN!(LXUlM-PE~U>sn;{C9o?qUF3a^+Xqw{K+2FW(39kujv!xs z`T*Dhh^O~r{}vw)4g7I3WkF612rFel=(L&&KaApSW#a2V{oWwOQh*Nm3GuNB=#5Nf zSZo1?**r|GI{R|5#~l7%96UMV=mDr8CPUY4r4NPeVwgr&e*svyEFS>jj-9 z<&nAZejj*M&D$=}AmiO5kxG+8g9)`r0$Iz0D=z6eC=3I&Y{?~YPH2|ATQ)$VJ+z7mIss34UQa%4(!)Y z&Hje$&zhZN?=$1TLZaKKfa?*80l9%PN?i>V%PlI5$e$1DBR%xxxkcPv^l|#LFPrMV zklITp(g$LOL5u1P$fdRM>mh$l7ye1`EqaGvOZeIFf|u96VQ>Sr!dDG1eP)y`F}Q8y zt_jd856L>@A1Ep%2P8Ffv_Perr_A)Q4?1na+Y{}{V<+u`zj^cdq07w(`sT5J-$%-w zI0%2v0u%33%t?wIr_`t$+bV(Vk6@#G0Fr!xFu-Bu?yxh;`(pgM!9&x=N3~dtv~|-! zDRwx}uq;^~2U|5Y;$*G?yc4(}mf7WzEH^0hzU{J2A(}(b!zPdWP%NmU_j^`w_j|%F zPco=uV^7^s9heIjV{21QV~m^eG`c+#uxSne$YY4X18h;l&LFk!&ieLPf7fXzDJvfA zQEoc7F=We9@JRqZ=YbjZuWz^h0Lnsh=ck3XO3?SPk4>;`C|Ci*wl3|`iXAYzUe4Yo zGTWJS{+25wj@?Y<#3?FJwJ~a@(oziQs%@jx<%$yEbT-MB*Q;Djt!4ZV{=-|VTCko4G$?RHRyYw z#I#ni*6*+ae-4HpoE-~HsleC_61+tW*)M3htvX0z>-+ak=m`i3eYG0Em3Dx$W{qs76}9q+Ga~ zmn^>w$IAJpV^`!S{XYgqy2^<4OCZ<^*BU|!$6?U1-$2>St1_OuVYgr7v5u?*sQd@t z43(It8sRPB??}QkGu7;|m}=4}W+z2bDK(1nAlEGl^5hG)MquqgGH<79#0CEt+^&!Y zkIp$MLH$8{eUfM8E{Xoy&R3?y1hKR}Mg%=J(hfNA-RRf;-`KRr$L)y|`y5ts?Ykid zwo%bTm(gvCe9;A<*f5olKz&$)XJ=Td>NImm*#p^WiC!hX(SkfNoVfNPYDGi~pymXZ z)zzYAaiap0Y2j-OP@!D(->%S%?rOxvM4j6wP7cin3AP;9BAE`-Z+AxIgr@QvrPV>F zWJl)Wkya)H#3m+?&N30)jAwqHc8o2wtQ(zMky|%-n2pZQFOU@^aRvcG*-?&RI>kWg z)pk?q_zAiqq=skUX+dVWUfM5DVa_eXy|&FWi|A}vSJq1_c`d3GWxcG)=eTdlywwxT z5@q)voX}%ucZ&(Nw*J`1N&j%(3Yk6cG&e0UW8|ePc^gT-6USOkSb+Bs#T=kW38hW} zo%*4ryWRTadu0a}U{z>JP!m6zml{ycZwBr9Dx&vg?5TbnOga=l*$yJw4-Ekx&h+E~!Rp~V?=Ea~=FN7g85O5q|HehKiJh2q1>B<@&9y?CXx<*U#{VRkhaCA?4?ZsUl2NgGk4a`1|VP3r73A%rt zT~IVZpo|5M#rH>}J&6z2jpDoiz4oJjEa|J7S1yU)Z3kJ#Vouj$y-$hwGDr?-xpjiQ zAjj7OLXOF>fISYX`fmCh_d&B#A?yk4kYAOa35$A!@tM_T`uL@?!&DLaVV$jEA zixV%BODya|4#fb)OB$udZ2NZa+lw&YUh|1^zf`M8mM3!?=HUq}3B0lJLWm~ayBV3D$}j)16c>kOxtu!KJ=D~Z zW0Jw}H9Uh`n&7zaMfnN(ny`mPsiHzqo7zL?_-R%o`JIv*c&KBA`c*4ZIXY3kC@$p8 ztTVGfdt}ZQUJe9;@si9y%*`E?WXv(L25NMoQgIB#4=@=si(vYGud9weDN2;@UAo2X z2cCrk2mtvN&qpynBx3nQj{1AhDo;=;sh8?R zM}(R@L0kxOg5zxrZVm8?Y!59AJ)ztb0RO=}L0YJ$mUddU?Gr=BI$L&sChMvAPb)To1qRqT)%rDKwmR{8(? z^s&~+X?LToD7q$xA}z&PVLd3EwWwBPbF;Z=!37a3sysX7P10yyT5uFq5ZtJGw;;G{ z@%o6Jfx|BFv5Hp_h9Cd8AXpc8fU`5ONo8=u=IP>VJ`XQ}lTm^#?&fyc+^Zxlc&&Sr zDw<@`_!)1UJG*aDqSqRqf$$V%0`H>r%3nEx;PImRCeq2gNdJp$~foIp9j8oJ2#2 zVGa^}AIIG~g;sgkzmjgF`U3*L&`jt!$9yDU(EHOWHt)qfr5=RYtdN? zIeta*9TX7Bo}CZbG{?dZD@XiCTs|Y1vYX%;6>7d?hrEJ++kGX{M`;!Muqxm49K$kA z9%m337c00)e%QRnDu3Aj@$kSvk3T}J%!c!3HwX<>U62`N#XUvQB$u5%bK*$$F^iYw z0L8!xvIsUcTt|dMz}>x@bdYxMCTX*5NRIn<_dqeQjelV2kbG$APHz;^=%bsY-TZ8@ zCTWl$(7$j90!uwYjZNqpw6yN-CKs=efa3`KxK-*|ks4#_M0P>qD zHJS+{n*)|VR?W-9YDwJgGS*qPN+3TNSKfN4iJU4lu|m=m+K-w~7nhB7K%WtoDJ$+O zE;~M#N7Gc#y8uK}`5ee^T`0$uFzTXIW#! zXm%Vw)qJorXHaHd-s)!m=JzCdymTz*U7iyV1|4N3vMB}>Aa#`5M9+aCuT=pDgK?{~ zjo&FN@vQ_x)D8}cvqlO~Hb#RnJ1D}2Qy>FdL6`XU%cBK%Wsdb%CyE9~8xQMq&$s*| zX8N0P^839t@g(z^34>HwAV^O!knFjeQlFqJWyOK}Bhp33NxCS37t7#A#Quo>@Ra9L zB_{=Wp2eKqVnawh|4>9~Ko(sP(X|-%2E6;|df{zJG_ONeMx`islRD9DNlcKzZGS|` zf-ZLqyxM(dB3@|R*8zt9h{HlJ;JNf4%q4b zSiVPCEGiGc5c5aEO4$LfeimvEN4s{;xj^b&S8Q0C=e@%IV?SmJnkW>;bYu6g)=kUT z$}=^dX4CS$s-4Yb^LSH`6N9A4!hmE_3~15sggEMq6}ovbGa4cPE?M5rsgw=4#sT}_ zE&gWzVUXc$nV(JTW!ZF9oi~5`)^F72a_XEn z2Y>PlGfIBfy32>OKQj|hl!c~!KgIM?q=!=PlI-?Ockc;E7nM=1Oqw8F)KBVzngbIA zWz-t46p|f!jO=wcJGMm2KPlo6?sEs;7pY=Q!d_E{8u!b8c_-NL3&?$;r znkb6noCRh8sGX8$g>B>wLHu_EC(X4*)dAV+Y0MF_2PUw%kUd_dp0!X?Y3eVZYj8u( z`D$KaI8K5kU%v$T?s#=v$e^b&H8o@#_gv6f<)G&+_j>N*@SdpE8Z~*@&zW>*;X7n9{C9uQ6B(i1Y&S41xuGqN;k!1~@Q z8FG(Ys!8QUz7)CieMPF*xezQZ+@YvY0KVW?h8!t5T$GVilu|;x^J3mgmw6apsrV+4DI7^=a>hm{~UIV?Nq+Oa6nvxcWbq@{86 z;kol8gY(6So+vv#xp0lGkeSeYG70Edhs=1!R_Hpf$T3KR;k*%- zVgE$0;pfjo0@ajT92jAn5HtxSO!U&fEiU;l8xx|^yl?wt!)u2%inWVRGmT-7A8Ax& zQgyGNUf%rU^dFbJ)Vt!XbgJsLv{x%%zxK^5KR`*$Qh!qeOCy2%75TYkI|Ji*?cR-HEvn1j<@`(H)ASK|-why728B#w zouBEq!~8f?j}L#-rwzGH!!x0``zaSaW!zZ3;Lf|;UA9PgeJn}mGh%tHPOIQm&SozDiCBq5?~Uy5h`ga;9sp@ zV@M%rS?~9$gXBS+C%1=Q38@iR2V%cBml=Rss~@xrc6;fS*zpxIF$T-Wz&6huXjEFp6rsiURp~VLb_#xZonAe z8wW(V zOYrCk;QK4~uBGvF2d5SagHIUe?0d5PZg*L1U)YGnr%bChpSBsoZ{554=@-4M^8-a6 zWQ$h_~)lt8KSgRKRqolo1s7GXTC!=I`LvX-@>r$pqLbNBGp;+heQ`R zEb9RerOG$Y9|*)nYu$%kdKclp_k8YzX%&Yf;{9rPRlW`e#pi>_79O8p>DW)~;NiXG z>sys(c$`xIZ8Ir$VtCYBz@wUCj!>kWQe(mldM_BrYnMZ%(FI|PqR5!uuT|{yJ^{hL zn_-2^cY14(9xz{&3l&R{30}p)cns=vY1YfP%M-b|v$fEZMhk_!Iyz6DFWQZZxj6=q zE-;=j`AB~o;@3A~83dDAZViXxeer@BZ!S^q^_EHA7)AWGSyIw{yD>7%#*2g+`2nN{j}ZrOct z;mV?fH1Ip={-p-DE#ajQOlXH-WgO=t@)ymCA5X--IT~d~uuTAJ)=?%a zjVyl7Y?l0gYgj?HIk8!SJ*rXpTe~PGn<5#M8jFeIIF6B%ur2Vyd?u=yqvXlNZ4-`fn8#y9t`Bq=ipaBGelkvxiljP6WIec8JLHjW{CPF@ZyK|58q{99D#m?*ES zP>-rr#0X9=D-3Vu91FrFLW`;oqWVDhDDEUNuDwtM*`msa3hoMr=HyOhNF15d&u&PZ zmb>9f+#SQ%bYiEC)L2OnUN7$x--7jTi)yQ=hI38C(C($DB zX4P~iayofHsjqzNlK8Lf@`~kmpx_lP+_0>egFeW5&jDCN<<3r8d}MAfJ?xSxt`H`X zF87O~df`s*%y3}4@w&bcp@nVqK2<&+kxX$nzfrXnLOXfNA$iB5jzxDDb;zqCur$0$ zrB$qET4d`&4a?ev9lj6zhqyX_+~I6dVSonjhZ15}r7fzSh+1HG+eC6_-}Xsb+^uW| zHuEaZ?Roflmk0Ve1@!ue7|%RUyf=h=7O!12Bwzd7;KDwL9OunT;r9*{q!IlL(qJ*G9i^B`iWn$0 z5~S&ZK`je|?D}xny@DvpPS1kC2l8kSWEHv?u{pq4k#Gn+E0f{8`7f`z{_x;f-{=|H=(6JCT_I7f=aw~r%;~U80iExU3;+dm5yXGv{jjt~ z(?(+xd}r`fu5Z4eI%wD>G8p9!D}?Pd?pRF@(-uH_{LADZ|9F4)I`VS+g4e(35#^6& zT?0+mzMP_X29TsR7yCwToI*D5#FdE;BXU%aUSw^vr`| zhwA>|Fim{u?KQ~b_ruECQHD=G?zb^V1GyKla4AcPd^1|52Dh9sH7XwxIqSF`ijAR} zoNQ$o1n%@gV9D-3++HcZ~ILoz_fP@maCNcOSPy-sgReF}QX3 z9tQ#BweFid3~udorbmnFeIpHR1_Ko5ZhN_DoQ&tHeA~S{!(y3yQLk&VeEY0*0>|d* z*k(Sg%vh6y!xOJ&g%xVrwfDb1+q|a!X2E~`itKo1GI9=Cga!&JCXXUHl=?oem|F;o z`$G{!@@@!mXL1X}@!QVegLC!aZTv(kQjq1|8CD6I7N;T(2Wpk5M|B(m0_oi2?w5rI z@Q$!85c#<~%t%cNT!6VJB4iRLwV(Dp&o#rvCBP+`ta0K%AP}XG+VM!Hm`xOkr_{LN zjqYq&Snup)?!JhVoUZv;o!RSp(scr&>XUeyc&W`JJ<0p9y_}U9nO1k@+rJl^aUuKr ztBoXn1{B+me*8L$0aCFPN^P|4|9WmY3O6EWr%|`slw4_|5JGNFNLWbYbj!+?WGSz` z=+Mx^qvvAtK#Vr&kFIvy2UY{&yu3eJS?^C?Fe9XF{yXbPffHMflNQiDOfjIXTnfpY z0kNSNC4<7JWI2Ax0XkVz0yhesYR(EXx%+)8gSuZx;N^r?2PMlZXq^bt$~7%OS`8Hs zp_MXI5;@jTWeYZPkySG%RIBXbMuEQqC7G+{e=MvBIi$1~=qCj0(GY+9vt6*x3f9wv z4d2XjHKV8YOO<~ktJ%3J&ih_KR>WvCnM5&(6p5qMI3r!>Z4OD6$A#p3YN2`n!n9e+ zd;yl%In3{{GPE{>$jbGOI$WQ#k{VlO@QlRrt<)rxEBi>QG*x=RZJ(Rwo@A>iQqUp4 z3^J)x4+@)I7e|f0wSE&k0uT9`jl~1z^m?Lo;;E+G;`QG{F$EOaMX4(#!?JFeBTSCb zNFbE!4*~%>=3yeEVvk&NLwcH|&m9TGJkRr_a5=On!hepzv{Ss?So5D5if$Bg=2ZtJ zNpFYS-)z`y7#^QRMjx>@k`q48Zg%<4#%qGHbiGa+FtI|-i>sQ|?$=;J+X|8^y5M*1 zKiIa6h7EwQr5n4qVF!rIZdG+-39g(L5LU>nKI0S!uFbtR7c{SZQv$J0D36mZ$n+>* zw3?}r3M0Jy00ZA&=vv zg*HfI1ldN;A-tfFDx`Acj*P>zN3n%vVFOH7ZCRms+WK$B(O>kiY9j{|$#L15=~2XK z_s0B&4*Aj9dhhC>bIYdM8?hM|tVVV6jalJhdg%+Qm&OLoti+~z7Q~TgGQ>T%Y&W+> zg(S#^h1<`^*k}Mw}y^tCN z^*d0YJS2`+;k7NQBH06fQ_Hh-tcU%E(&j<3cxSDCnf&OOdtfz4&RfE@|K--1jCl%j z|7_qK*~%_x=Dhzi2*!?{uyQFTiz4Zi8fkc2RO{uR%1dN};u?|wnQmoEdc8jOE)Uxv z-#{>p`(nguDeUb5S#;L2B4A7$;6(fBcxOlgZ^(0Yf!ER4(}- zcvzMg;dp>N83&G^?W}>xx&;VX|0{vWJh7znw_G7{PQ11$v@if#irG$qQBYU$285LX zwe-#KB)qWdD|q8$ANJi*n`2^@Y(mc zbuf%qe>&r@XVS(dC2B3kA%S9IDe@krjuT;g8Y-qZ7u;GUriA7xde)Xt$d)H>dBxOx zWA)^n*DAmAhvh#RTfM&D0uqmk#Pz~x$X17qj6%8~uuy^wqxk~h4>inrPo>EPj(*G` z?h_}6WuvzdS zuQcK|_kaxR_v+otyzp#`3RC%ZNY03RXV=WZmJt39?4wH~4#@iG54b%-Q$^=4pLMgF zR1N^n@pwUOwanZ>Tb0{eR@23(pvO5OS5`X!mX z>Yxv~&Ee;kT?I1#^JJr2D(5mCDQxuI1xwo$MK^z!#|NB?5!V)A>nH&xwGJ)4!xVF> zc?pmveJUc>>(Ju7SmaQ|L&>P@AAxunV<5V6J#wr#5nI<3$gpr)^ZEQ zY^F#8rCtN0+#~&b^@E8dgL7~ut>U!Gk%8F(cF(%SVgn58z1aWEyNBq^x6HtZSs%BA z^f>YIW`hOW!2^mJqR3rJy(P3*bXM5s4%vo~Vb}qCFu2dV+wH377C)_GZD@;+<_fv) z0jVMZ=Q%f?OY|CXso)NIbVCRqh$Cec!V)OBxD{M4yEeN8Bt(#v8`-(Dfw;YlHgJb! zns|@SIY^=VP+l!xR3z(xePygj?3{Bf2)B*aJij>v@82$qa@AZEZw>*KY51z6ch9Qe z?w$q0(phvmfdx56oV&vMq?Pj?0YkH;33`&%NIQxJO|U)9Q)G`o;TyqMWj4Hve_9X=bfaCi%>8Z{y(!i!X*X2;CMNrkKpIhRat!#nvtxh_YiFsZt!5sa3Ak#G+gE+(d z`|bmf&VYS%FO1`$<_Y-3j&O@@0?bp4ha=X54gY&-^<|oR!mLci0)FiEwa(hXyIQkCR~;o+_kr$sPd~E@9S|rUsyO)aTrM%?_{F1^Cz0 z#=aAd@in&(VxDR-ST}g<6c4(J_$`Y*Ao}#>cwi{g zn0_Ea>RmI%NWF|~EZ7uPLmXgDMk8k8I2qlG{Xeguei=BnHr;8nu&q>^H%Pl7V-lre zvEt99)0jLnhFXb2B1xjMCAstfE`UuN9Z*APaBBeORV4g3@f{y_#mn==NI?gJjP|w> z#!;I%07iKz_8H#2wnUVOGoiA-MRlBX%KCUEpt5jz?wQ~RK&^OD96uX>o*<^5gW+`|{GLJghvU!O zu=Vg=7x4EA`T@@oZ|O5JGR_;1xOVKXoG&_|QYU@g6RC4FQGzG|2E9mypAwrj|SH^kc^sFMtywLi32{c0?D)w8+6yc=!OHM zB=~p41@kLb;yPxX;JVzQqd)$@*@lVnuXW^6c9?i?)sL@-nf(u;>+(#pZ^k&eQBJ(v zJ!9c4RZNZsB4b zoCep!#hiiu&d+fRa2=!= zNZ~IAR(AP1UW+P|YnZ!lajI9EYlX*}pkslxbfu(|U&<+>w@DuO-=WV3tqJ;A*v!H6 z2DkT~D|{Z4G`m#@OPtD1h>!6)Wm_ZG3eR(nkimsl7Pj&C1#WRK5jT2nVxeqj!@$}T zR5sks<|a)Ium6lTLucT-ey2(Fc-F^>!?S4?0NF$_@f6vJbW!(%?l0X5f~yTd|x$~3dM}&Kf$)^yJ+31!FzVHz;`FUuZ zakddkPnnR$>3EWdjs^aGJ~3O7VwbcJ$T}xZeE~_SQ8}SoDJF>`i9nq~=Y!JACeom& z6KJ_jOr6g~3DVu8=0?0bihCxoXyaTa6!)V}Q5z`EHO$8Dv)(zExMV;Q)3f2j(O0R>$!YO5J zC?*#e`6PGI4boC%{ZbS@pX|<9(SnRQhh!S#@(9lMkt1_QTn-1G5m#|`gns0i>UBz~^@|O~ zco1eXU|rUTi|K!(Rh0Vf5w-!*ShD=0e}{Z8_pZ?~Vu#bGxHI-UJx%-8o8m9JS$cQa z1yYkpY-qClq-$(wi)zK_uaCa8C}KmW$}+JYZabNsr**^5EFJE&OJuF`TDoD@RXE-X zsRVuVdT3K?P@a)BhLp&51#|%!NGGIuUlLbBE8K3@Dt3jH$YLY5FU93j0x2yld z8aq!i2Tub8i}U|Kd+!3*WP0U~`-*Q!9t^n=OkTkei7){ixrK^2IG0{_yY2qB?Qgf+ zZnv}T$hMu{w@s&==>;!1I0Ax-7tjE5i6DxAsE8mSH}Bw}qId&A9F-XaL`22^IY}5v z%FhZyXECT+8((8myl(mI%g!b;Qh(d?XrtD&GH11#Q`ha zF0J|c+g}ed*_jOQtPNxrH}}VJ^X3G4fsfjHS5QnDMfOusIN#^Rhd?eYjebD)hgAgX zR2AxN(D>3Amw7<9Lv&_XQQ}iiFKKrvb^F6kX_Ncq5WHu}^+<~&I@NhH=n8G;eQIUf zWG#_+*QiUdt0vw!@x^a~0Ae##+|V+{)juEZ@Ul!rb1^m!9KK@_t=#LG$ZMoA(i_Lu zrb)_OcZtt9XGp4e@iTIOv~3VxZZ-->i}Ibp6$SY7=k*diNj!w85nhF+#ne> zhCN>X`#T?0Se7xqP=OhiVnyr@)~7re+N$gcg>WpeTA;a4t1>y@y3%eO>u1alhv|RD zD7HVv?!L4Kt^Mg#)AF<8qkvl2u(4k}#`fKTt;bfgbAD2zRyR$yi98|{4bmp}D!wagJZETNVqADAbJ4y~YLMQL;kW9$~UPZll8r|d$E zefE4j`p+F|}#hvWu~-D%NfGSm7Yz9aq>N#wSDap36fF0=I`onl}?Nui=p z8}2XB5qXj)?cBAvrNKRKvnrm^)7X9k@zp4=q#Gew5YK!TGzrk+0;ngu5mp@7?8L5r z{kq4CzMyo4Y*ixxa=o{nUaicgivqGVIm%sjRstP!3ap#vF>kf&K~8pKO5397mD5MU zt3SQ;7qZHM!>b^4J}Qv0nPOn+TSrCp%2Q-m5(*8uKz{<_p(!&?LU;5E`buDfs?%*y zrc?Dm2y7A_{gH{a4jUs6|HxhT|IG~>Nh>!LEjD4}o3;PCn;hh}rZ{j&?1CARj#CT> zb=Of*>4N;Qb%AMg0(5Gsm1G3~?^ShKdR&!FCz5i%mdJ9yY`RgrE-;-s|CN-tR=QsH z`GkMpv)uI& zE-$X2_jsfTKXEUgT}NsKTV|~HPLr&3Ju!WQ9~k)WiC*%hiu=*)>e2K zA1c`IN{;!7V7>`2zb%ifBKZ!yG#xcVz(I;BrN~|?Dob;~7x}NTF4_dpu1KT%gvbim z5L}`*R&s#8g|JJJtkiY_JGyjXD{PO`Xv9IXvQgMAMmpRn&(ne062%Jndhd9qonI_& z0ADtiN}5$3(j?cZK3-squ;I^fS`3RhuoWOinG}mp7~fk?tJl=~|4f%!R{gvr@%S_| zuq9DU0!3C+QCOCYQd+G`#q1oJAmAA-ukGGcqD|bO($hCWZ^DYxtm;su^R9;E zXyTd4F%O&(GD)FWP6(m)Z}`;h@2iC^mhiO7PK!7D4Z1cNxFm-b7?s~A2REaiaVs%2 z`j$y~{r;cdaI&P1IIP`ZA%kG(9n?gf+g)P)`emimPWS!1epz$mgYT!kUH10jxB3?^ z`^K5bT{CV4Rm%FI01zldF6DCuKxeC6m_=7HtHN`%jdMy^^#r)zZ?wh zfuq1a$@#HMo4-zVL!Qg~OkTx1?3wA-C?DeEUMkkbHaJ62*cfXz!5I$n!;UNQ zzcTHU_rK#NC%2}NRc4I$2NW|%k-Josu|f3>UOq2biJC@OOOvhXQzCss+Z*1cxd__5 z>jG<~c_I{E%#$DmKH4Rg>Z0*oGB8y^b0T69IKdm3HJSSME$<{Q|xl$b{AonQ^9VW161mtrRs1Dgb7YIclSzMqv`5K(wB)b*VMR%x> zs-FO5>MKDm5`VrNGQ*DhgwHv+VyUG|lwr0%T}LsA6p4p8y>O@3&Jbv=i)9jN^lq4T zte!w}fbE`AD<3OsA7(3uv)N^w77oWn#XA|A-&zJ=9M&ab38!?z>Qqd$dVeJ9V^$fX zulA=djsYd>z&gf_c0R-ntOfrjUTC?!eo4F>md$vnq;oc^q&y0`8}!&or>dR{3=pN_mu4nCByJXmTcBqt|c!CLRgkWqnPdA<`(yQ_Rw0~ zA~ngo#SdMh!Hw4zx)j=I7i29$%q_<=-;JK=ald9?hl<|$?cHCS_C~*H`msN`{nF%| zKQ`OqJ*1fX6d9nRF1}$XRW9IP5~J`nrjs$Yb07qERfrIT(~ukVPNlI(7~FB(bVK2> z^MfF5mEiPjq-E<=n4~OL49Xq=2oa}9uQ>rn&(Y~_INX45{Huo6M0t`P=uFiWvPYUc zJ)c@WZ#Vz0v@Z-*w9&19?1T!*?Eu!a07!v$yPhRUVSQmevI}z>z2VQA z!|w6^^yT0Fz*26VMW_7c{59U0(%m2-{mApv`S?YghJSUkvL*7kI3FJVrM%V3c3~0o zfG$(V%(}?i`08m;ylII<)1ar@g>@k-csU`p&I9u+>1BdGzP$qbHn0KXe$JnDBfAY4 zqgujonMtpGK5MsS$*RL{oLGub(*?MO8D-QVGIO0@0kYXWH1?kwB*kITp-u9NKx|rmc)qaNsT+v%8TgJg{-bcCTc%t_+c@>WjMsnUZ_d2V#+og_V|aUf(u zGjHw9oDf1~U3M3`ncSZrUm{CLk^@typh@f~5?uzx=qQp(MYYP-c=vi}S1DI155wx1 ztV|UF)%PMPP+*q3f{yXLIjb!QT6SR6w-lU?0&;=DUANsg)8oL1$rrYEw#RoJCyY#~ znpL-iZ^DR^uhTLTinG{Z|(|6kU2&b>!al`s2kNoR=OT|ep0XoO^y9tXvWizi){Hdsct`TGg$9d*?w+l1r zytFIdS(-o20m{+J^boUVWaRscyU-2@8Y~uz!k5ckBN&Gmb>f=PM=f9nfK2D zDbI}1J@=#q5}x;#lPRPW?^OTVvS^Wu#c|-G5tf>HRPAdQZV0Q@KqqHKQaFZ&uP^vt z>6B_6gq}Cywm(00>ap9F&7+>q2srLC%GT)$1u5?-mx=~uT{I>N((O0_0F*UfpLG7Pxi5u(joQK9JaL z%bRt7BY$71#bv&jd^Xg@qRePB6fqtm`d98j@zCVDEHG`>;^p}>Y}+e-_|=G5E93LV ztz|Q>>mQ;c21yne>G*F&M;y@X3`r!%NFRTYU@kXNkTIuSh>vHr8mf_f`(0aQ+4R=g zSG=!yCr`&s(G7^!e=c6Fw3lABwO{o!9!}eS$EMNPAI&l1^Nzb}d%bh?Ct)T(_dEM` zq?6*8CZX75<`o~Nm`aM2L;FS7J`h?+R^C;d^1H0;@#p~qlA_t?(?QCCE1tEeTG}Y= zAV}8jQ0#PHF1pRnkkomdQ^)f5hUfAVc%{-B{{&u|?~^-odCPgcfw?hJYCgh9GZ3p?EmVhx!8HC%!G;6~Um zVcs*AY6g0;i7mB5TBhy_YLX{Nj00tmtHIHtUimurdvZu1SMbxFi+B^}%Ge1f6W)Wf z8%>=`_8*pc!WYT~X2~6P!qRE*S+K()B!Cw7I41;d2-Rtj1JJ@~cl&qBDgyVqX#14$ z3^EzYG@Z(Npc`mWx^%B+nWm2qG;}-m#+x`+CTu>&9Dp4w{~7dw&x;(II+g&h=D&@0@#=APzC=) zKo-44h1a&jefU>s`xHr{i=;z|&LD)pt?;fVj-j!ZrRN*`SH>h0YXL59{kCB}rwP9r%l+iE_80*p~ye-9gl1zSnSh3=WvlSCz_OBo7ja_DH ztPkRNjN42RN2LW+?9D!HnS6YD8FSq20AWc;mT9Whhx~R!qM&bXl^3*YO#`WIaA`=T zUm>r~*%}+c5fIP1oCAoTHH^*Av4i-}yRRfqhNDBaLkI(K`%UuR|>%CWN5(Ikx za#%=f6y+R5(@F{r@n=_=ka!Jw=>q+Afs9OfJALb3iq+ePr-ROxzmem7m@kw#zkNmUsc z1QSE*7~r+VA;9b2nv}>FT}>uN79DV0)h;X#zpc!5FXCMbUhcbL&Y*PAbr-*y)bNL$ z+Qr>&-9gtTQ9hC5e1Bm%?EQk1Wun4kt`&<+mg%>})Bi}~9C!%>y}D7{yDb#68L1jk z7oBsJe;Zb`MCN%HN}`p$9#@0&!|Le*K1RITl!N5->>6pWN4H|Dd%BHkOrAf99N{tH zAk6dsBI6a@;PKYUDRGv4oszCk!jtfIPKp1<{#Hke}JFyUn~8$yKf~)wg$9B z#zm|Pta#(f;`jgZ{+~ZqbU`!d%HklpJ1O&R zeIoLShhT&0v|~Z#m0l_`gnyIyL#11M7>s+x);IUI1L}Z$vF1AyaAh}48lu>|a| z;SN`(j|3IsYCnjNfO0pl+7GLrPpF!B#o{b-n19Oiksv{UN@Kv%8&>p&gQ(e@j=A>1 zh1qP47bIqb*x?N~ko>&la@j-*3LH2yz*0~U9eSEh7qkm6i7)c7R35AK(KG?8vyj)y zBs#ZFVx$G;p9_xc02kIFRnzyZn8^Ovfsw(oKQ>BCfpm9HFbd`wRr$KbePKG)A@MG8 zva&#M}`VY)sPD*npS7ch@YPNGAyg4*IinlIWy6efz@nvh@)3M|xF*V(E;! zkQN1!75eyhTw!@aF4a2UYQJpF0~$}?b?FO>2hl|>daj+aKB%2d3VT3Ts8Nf`I(L-y z;bb|%5ietTt}G)@>~P+apH8(TDmbhw!or}y*-r30s;DDFGY<39B*p66>cjkxrd7da z1f|39d@Pj7qa-hMa9%hNTE%Mz4b>jJNp4iE;%=P-nYb(}lyulg5!{*?;)o7whP25wjNts*7hOm`)w`F@_vwV9wCfAH zl_yD}KGrj*xh2MSA*x zO%O4K+bZ?OT;HV{(<*f|>aXia(Mz*RHJYtb)f59^zJpk^J?PpJxpYQ~>=-XY@`plY zBCko%tV@o9;ylnpDQMHS)><*_&JkuA=R;sl=FW%r^};+Dt` z<^9D8ymn!wUzacklsHj5q$P4qU_WHK*LV*?U%wmj4&bjQ@NNqe0?R;veB!vXJ>(1^ zoG|XTsD7e)8LMmedTtCiYRdP9!z<6Ab7GpJ3=ez z6+C2F^?F?7X*Yr_eK80nme0-sWuY2^d?_S+PaFqNO_*(f@zk&rKEQ5^+P=Q2jWO9B z*9QX^$VLaw50sc$5m$6nWLx;VTg zvYD4+#krCfo7N9q;~(Ya5W6dbCkk#+FCJrH9pD4yO3jXGj`Ogv+Q^x5l?d`sydhmHEOhykTZ zQoJ+VcB?b_tKWF+Q%T<=%LJ)!p0iD4mfLJt2LNtk+Tv<%E6Xb9e|zf(CL3~Q-i0aT z{7bV#zGr4#KBbuJ6uCCCL)KA`7XRU+H;*mWskY8O;4kmUI0_?{(=((QDPZ?PoK z?=+nR^o3ka%8aw>a+gosS1!QU5T}8N9ma1GoqOp<=v-dzlIO069WiDaE(JF7yA4OP zVH6x~kepPXR^xTXQPEK; zJ7vWoUGy`iiv7zP_ZbeeKkm;M&yDPMDf8c6IW^B@gWhP%jU#z4jSXrrvq1+a24o)g zP*DwvD_|H}BFn>6Ij!!UbZpUn6K=c8d!m3L+}$LD@!00Y8ScGW8(( zu~m&nZ8WP-f1jg|jAzbA-*2zi?51Rl5DWE#!2KSUNdSlSds)Kz9sgLh_=>bSvRU5o zkInGvOk_)Bub@kjJ)=j_DSQM=&jE2Ue>;5_xR~c;Wnd1{A_^txq6cJ^vX68q%6K`> z{o+lswm`T~n&x|6)++se#6Q=s$%tC^+(f^zgGG$$G;jot(QLa#ue!?GB6)A z(#s?AgUgw|u$IVZ>Q_3|Npe4^+Ar4qT2Q+%I-piip*|pmp6ALkb%|nyTu(og@0`&b znF&<_XPw)8&}=O;zYQ;Ka_N5^?VbY0R#tBi>s(_z= z^n58Z2M+17NSU<@(^cJKop-vk-X&9dXig7kAJi$GN{=dlZ9G%tBw1#@W^JnGk-)feUMIVaLynUF~0+2rJxyQN+RuzcamX z;Q)=kP{}`CHh4n2>ErS|A5bClPVrswwRcv1z~_Y4d{dtR_}Y3#aLSYE7~7L&k9#yU`eHlXPb~ojub8 z7N1|1FCm*87%Y3uz_OiUptfim6_pd*ZIFq_^bdc?ZO9Gz{a2*eQo32CgZ2RpbhqNN zOuN_fl;<^%R2QgH{Y;L<_gI;aqCXwHOD^`R1;81c7*r-qwq*qoC#X!l!uw#=i@u+3 za!(XsnsHdM-x+I0dVnyeUC%Fbu2v@q462w|Bi*4arK5q+T_sCa#*%#54z+fWV1;#> zWKgC9A@dl|ZaPnb10eMY@YKRwBEX`fQ)F;~s)ufO10^WKA+K9*dPp3^Ll4-A8KVh; z=qU>{>HsT1tU196L{pCZ@Lp@gh`m|9dI#D6g5|Eyn%STRim9bYH5H|UT7W%XIu#lm ztx?DuHr-ebp9~f7WuUlf5FEp1phnFwZ6_TgNECF-A=6zw_mX&rD8{dz+>p0{2&ndI z@IlocFYQ6UZn_k@+vZ79WJg5tGom9NJ7HrBsFl>9vTLpsD<%zj@9k~}d>aEF2jTGJ z!kHc6^##(bfs^Gm?C{F^e(mzNO;Fl7mv@I$P9vAiW~Y&2K%o05tPs#Xd6)Pzc^U{G zxJ@d-|zpB#XX z_hTm{CnSYc&}H!S>s{8mJ(OZCd_95W&*)~(hoKZ2FCUoSJ9oR=MP4?I+CM0ia0k|# z)zG_{>r_p~EE%TDavGlVo_&_aa5@~fB=yM0A2>UiAeDdrXCIPiZW$@Z{g)AFB04HI zm`E}46#1BnI{Ah+0VHQ`1*8TW#aZ)2cP4cgH3BTFkUetoB>wJbzrD}J1c?u{XMaax zxk18#Yix7OAhC&Jk}0y5iYg9E2rMAjz5pa;7VYFi#2C9#BhmX>P>X6^V4f&%(nRiX zrmcd;$Raa|(0I5|@=#)eMuu=h8%f{>4F|Rs`DW12QA{dDHc?S!8l47fsnLDTlIGB@ zf!LR#12WUN;Rd03VK?0|H+}|q$D$$EOaH=kc^l{CaRSJAmbY>KU0<11nELk>rCX}0 zb3LJ|lx^f46sAd=R7t#bl3}NlvQzN$9-TWk&9{o0|#E}%*?@lih(Xfxr|JdLKIt&cHLQ3L@hT z+?{0h>EHy0krjpgH~R;1Jj2O-p5p%COpee5oj1;HJWM|3w(dA^@PE4*&{8O7BSqFz zQEhX&J+X9pFl>!yoO7<%3VCDT300P~B@)XLyTLa}mz?5XwN(+UHM7SJO4f{Mn`=5; ze>BsQZ0WG97nZIO$;vqXE^(Vu4*``lI#mRL-4+$-%|T=JJ7L}ONmTy9IpHwVwwXdJ zN8w~NrvCIles#n$yUay#!GZTNEP3bMuok5Sty1=bc-`>Km9E-HkSA#7Efb|lAWDNr zix#Dwm)>qB_v6Qx$P$v|z@BS?nM095F%Sn!rJ@GN3DpoZBkBx6qCQ02OO?Y0 z5wT+=V-CUqg6hI9g|;-jMq1@{$wk{MO$K4yL-tH48|RDjteY%@y=Hr4TyCq4xNF$y zMOWs=>caElB*BmpGRGat8*~Pp%};z)rz-RLTsm1s;ES&xVsqqYzxs{#_wW4jN5A{I zbS1?sr+_v7)PEy4Oww;|u;Bl-;M z*QZX?0`2WP=_WbucXQ~i9+*wardNf>y1?dmy?2v5hc2Iah%~v!Qcdy-Xc=HH(2dvT zd;vZ?fo*Q;*GgaX%GD*6^wNj{X%(-~qfv9$^&p7sw2>1bCG+exne2067_jMJ~rs+GXs{P)XOBx(A^Qvi+RJF~4~{Xe}yO2e8u# zZm4+wyZ2XGvRt{WBMuz!L@lo zOXN9TFOAyR<$?KOmqYF|ae_k0P(X%jI%M=xy+;p`6OtQp5X4+tBJa5N(w_hmXzS`~E5_t*A7ZZ4AN#UE97jpud z{b2Pp{XPmB(26b`!9Kt&Ib??LuUd%}OWOMjTQ*>hOYpwf&(>Uzwy;Z#?+mGSB0^cG&ed>tFO@o+f##3$cM+q%G)x?~wAKaH~4TuU3r`>-(jK`;8Km8_0Suj z*R|3w3q*bJ)>ZBUN=^<6CsaK#KI3}D33AUsmFSbEa1*LF>J{l^?@QwnLl4$br4@A) zbC@EPR8*|r)j8=*aTp3Zo34s8K{hcqFu+5Nn@p1$bUOB#ihiaei2 z1?qMrKyQ&X-c?>_p^K_(IOPwPgd%@H+#0q>iyPxYkNqp&#jr;WY`5wTPZWf2d_dM z+9ffZz2>n)l;~fiy!qN4Fl>7j+n78F=ChZHphPW`Um`)N6+`d{!-+dY+n9YmH3B_- zTX<7&(miFy=i=4MLkrFU6$-ci$;unj*1*dPVny2B?$|&rfoE(1kPx`jy+^D)JF8fU zC2%`LcX)b#WL7|;qMgog?NdH;FCDjaFuxOxrHZrfYeJc;~`~ z6*oe&<_?la$}KCo!_U_UiIh$#nIZ?a@gx>_i+k{jkN%#Yv!; z)f9=PqIRpXd;B#8ILEEBK1FJy+pTsz&J7SX|FPQV=O#dWxF+^Ba>Idn zl`G9$l|hQROOakG3cZC+Wj$$?ouYH0BPVhzQ+)3Le-?!ix}jjcTHPdv6~-9k0quk6 zfIKOh4$LC-Dckv({Krlo_Jx%$Jg(XW7H6I4oT4iTuU{pvfxr&_59{|oz_#`16eWtX z-GI)Wto$%3EY~N6*GbOGvZM$7bReJrjs{vTq)S}$_^=}Iz-!6VwL8?U$^);}`lYJi zwjSpXb*ielwSGOam66)xytW`53bP;?WG#43#t+1uW0s~0L<7LHmf;_U<&FtZ&+o8~ z3Mc4I^*eCu(L{V|2Tppi_|!%r7;saQAts51kL|%Y&%NPb%!XWZ_}gT7Ae%m} zN||9)dwJ~CEbj|41`?`xtuoBKR{G`6Jvr-?23UK|yaY*;9Ft{8?7Il!hxLH$K4qTI zp72jRn%paZSUkjUfKs;vUaenSAn1b_yIU9X>YOi$+l6{M8tM(slR@yD;~D&fZsmCy zQhlnR$;I~IL2`{wjo1);2}MFG)Z3VL8mXy|o$#sBsSxP61mlyG$?t^EsZ^}^LU_gs zEBSl}*tS;~tna)>`!Bxem5>*CpNFmYempbF|7Pe~|H?(svm~@jGstAq8K7UNQ{lU% zyrY^f#Wq<>WP=Jj-}i^zmhO<=kT!V^@YX{*D>*nx0Q)(m-XCBk?RR|y`~FU0t*le{ zX>j_SI@_u;*b)r3Tthk8E$aV!ZT)v%bZ@(O@h&<|J*+75=oD^}b%x*~^GI4rJKZ16>z9>6JEUyQBk3ktw(s(Qjl3+)7Uw(m(!{qX&1_YNiY=R2o{S6+{Nu>V#aL9P$By4l3vMsA9) z&+8Npb4*Y?^|(Igz>9~JGM~^pQ%x3S#YX|P}Z}dQFb|P&FtIdgP$`fr!{7Zs;!~YGI9En&~sQ)s<+kEesS(4B%Zmg zMxoUTe!=WCx=r&5zb(SB_@Dq{CcjB&D7wz48%XKG3VVW z=(VJaGBn5B`}sq3zh@mJ?LYPG_k1LXi>TBzL#^Dtd7ZLq(i~DRwOcE|iNj)KB5nJd z-8B8O;9H;lvk5T!zvuG}3Q!~J{QcMd)&VVMkv^cCUY#9iPp=xunI3aL5 zH1oS5*V767z98&Vp+)kS4w%V>9@`*lg0Yo$=%+O3x^G?z#8t4?vRMVzb*tz51kcg5Lw)E{MLR zi#C%wQHtOm^yfS-KH@wW)TS)*O(eB`H3a#fK;MI7$;vk69dH}$xzboM`|SRL#f;hV zEvE&A`i`2MwUm8e=?h*DB|eq1N=-eDTZcTK3UNbtL-^$O0Wy72)_wQ3{8p_Si`#ea zuoRTf-@E@O%dSRG2WB1Emtg61)CI+rsBnmTi3(UhwghW&$C49V<%MicLxvfkeNlW- z{MZTE4BNa*7h=Gp9m-uczFr&vfe%zIvtuaVI%H42VPQx1cO+Y*d`vFsuM29{kW2@r z(o~r_nZ*>dn<6`*X9d3_G#h9Hm&7Xs$;wpkN*Okj*x@%omJB;BV_IHsd2QH9=Xv7w z%rB++#_~3}URZQae2mlxkPVJGX50vugkN=!nU=LEpC89;QLmh)Qx*EQ2ibfHnkYg> zuw$J?v;Hb}gp9F~p+W{P7Qgs{0G9&qglWqKx7=?q9}}nrkfaA|wPoQ)1da0d-?#b9 zKWi?ypyFBM*!mnNR7`z0WA0lEOtvIV`NKR?!fku&z+0J5%q&kW#Z*(|5N>btL@VSv zH|?XKLS8(RtHgK~Zgmr9>QpO?3~?;wk0llKH8`nF1B%pM&wRd4^Vlif`35=Sb6%_+ zQlKhc1a4% zIf#k`T(@(!c+@0to$gV$+Vv$TBQvFIsHs?PGBQX1=^K~HT5d+hfzvY@-+m zs%@d7md>aWG`UBMu7OraCcjZ=5Nc4CGdJYDb6X-GyuN`SAK4d{?^@@&kH6gw&(_l? z-_+*+r~Q{Ou8=ZW54*w%2mjy|7cY?sB)=`5{znowR$|A2p^#?=g)J1b8UNU*o;R<6 z?#EEXR*#3CdY5Wa1?0%hevQJt;dNfQ{AkePMcqWq1Xfv{&7JUMa0S7HKhVkn*zF3n zc>j!7P4IA%X!en8ZZ^b$9Th039c4pGD5i)a1ymIJ9YbzikpHisvFh-a04Q(;ihJSR zz@x^-h$mF}eB`@n_o;iKk7Bo>u6xsrsUh`;@54l~1RA2?P zw*rS|Ncm?&y<;gi<(OwpB3DE`&nbD*=kDVPT3@}s}-Llgr^ z{8D4XmrlrfV8KQjy+t+TTB)gZ*Qw5hK-pjy5a3l!!eTEh&A`T#SsGN+H*oBM`!0-U zJ`&wRcM;Dm6IIWKmq=g4j+UUho+lY}9he`_z;*>=A%+H+dp!qb?QWQ}omkHYTYvSV zr-sW{ZFw#G!#u|R)Nx~5pY85BK}-lLa!NlzK6+^^4|JFwWqCGI%zBC>A#o0@M4fY| ztUz({HEp6`SaHy=j~~x8Lj%}Ws7opU3X=UDl_zlG1n?)f+5J`6%0GY8a!qpBO%KZ| zbKbM?6|AYLpzFNqJW`lbKy>SsuJ=Y6_?F0{&=T($DbpO`;qeoyyL5lpClGhp_{I>w zSlyslI^&k0C31jQ!0!#u?IcFgj%jj_SW)KR~i zv*DPqtw5%c9uow{xG+=6K<1$zHP zIFhYtl7qUG+mPD@ey^)xqlR7WjdDI%kGxFO!qX-zi)NK5t0lJdRICJU`|%=b zjQb9MUislfI-fak@QtPOSt@AJ^>|e9w$54}5FOSVjv~4R3!_ElGy5QbQXi2I4K`QE zhhJUcX3O~>ntj_8PNI!jZg>VTE0f?o15DW)@}NU~}Y zcG7(jOJK({uU59)_r3StOMz4&y1bV)i2|JlI~?`U?QU?QRC-I9KyHM@`04#0x>ke4 zVg(fX4Dox(p*cDgrmNv1PT=B*kFl+Wl}EWCiqjG`<>0sfTV>gpf(u91ftM(jHWml{ zRt7H<>FE^_*^oDcXkmw{DV&`$x6dLI#{Mcz&p{Iuxox&ziH;iWv`aw7vwNxPZ48kjy!%x%s zo<$3>$+XowcMjOH(hxoGGV9Q&cZ>SAW%&*lqvOEXVySV){kE~6tF}UX4S2#<&!KhS zsI^yFg9vNKvgLIfm#5@eDuHuBgaa=_Y#QN^Pf7}Fq=#Jpx&}0MkmUqa7R*`|`5uSD zz#7E~Rl4&eu0_^Bp__$}zGs65Nl1h#_E1u2QBdJCu3uI?dTxx1c4#74Ll- zZhM_)KcVJlxk2TLMHzWzKAP>`t&ffs{$j+TUAX2|@+mim)`6FpkIZy1?^DbGMf#{H za431Fpjcdciqy&m!(s&GF6p8=ugl5?`crUD(M8RNTCiPoj<1p*Mt*#F zYQ&NhUYaCNk`Q=u*2!5qRm1EJf*ima-U6kj!0I2AZJVtv@&E@oO;RF_g{X0fqKD2R zNVe#4PNR=Vl9UAz8(a>NPHCD1ug-?uJfak&Gvy%ZYbcDzP@%S3eMi_X)Oi>AWB-H0 zqQjz_%A8>v}>yj`XY5eUb#^QJ8PpnN~^3qkn?n8WlCDhUSUiGM;6Coz}k&FJ7_Ql9i9gc($0>r-44 z_b9ge8v3Xl&^&bC8Bz?ZZUQgU?GA{LWk%PI(?zH3w>M@X~{ml}HN1L#qVW z6n%XBO#BlPf%sF%`*=aLODyHElHp4q{X^D^&bh)pdb(hcl&Mcjv0--d+(vMPI)z<| zN!_y>ZG^Q}iHk+C#kHKQ(v+{tyNRW8uETP!SX6SCxn)YL$#(jqX_?Xv-Vx_+W$d&Z z{vOqR&l(_SCcEZ3;dhAib2vY>>&veDXzyzGXEGkJ=+Eg)HA2eTAYktn+&g? zIqJkua>5C9rLgn9rTO6ED?9KepM~Iw#%7rabt}{8v;YwEe?S-0<$+CdTw3f68q4Xm z8bp>i+WEczQKwj`$*EEN?d@i=W*Vp)jNay^Q4DA*ZKk5I`BZ9fTi_9|I9Oy`RPAn; z#H-|wosN)RDGKaixDB&8kDX4EN^k`x%cNWV)joMZ?fj)0hv)QsI&R|bL zcsm_qw!gVYG5r*|Lq#>p6HHNFWJWEW*&&5qi3Q3E{${^CQ986pM9rUb-U-lWF>@T*sQJ9q`-jd(_<%8Q%;v;i}HCp=iGA3nA0x28ob}D)D;kHbSYx7%Ws#w|!Lf;VIX*VDbI#0Nb+>6b)9tofqd!uU=^!)-1JJo)@WibYo zZ+i48N=OGh3fQN8mE1rvjAIVX(bGpYT4aUk=|`T~-dZGw6_YMQYONx$3({;|&xs`d zFCuOPTx0%bN8F9v&NfThpTq8_ScrU=eUieXMf;^Cnj5?0VG-9$7idxiN ziZY*Vvib$bnC1{<85b|^g1FyfpXQJr8k>%Q%zMB9sDcbdJd_@z4+(OCSlkks;#;t| zTmrEiMU`Mk7%fT>HHX+9X4WI=nX8q}5yg(Ap~XKBdd=h}mHPkjBeKJR-J~OC8^1D& zfh<%BD7%AHRITDP@Hao5ZxlHk{$bnVR#|WOI`CGGLIViojIU94D?gL33+#s1c;0wr z)!cHRY(ZG}+f|Eigz8k-adKF(-?<{N#d8HDIP*j($eSyLj?>Wo$aaXa5mH8AWHZFD zBV`PmGv!Rs^2Qgvb`BjFXo)D+-V&?}MCP!eFuw{K8`*BQt(+cfL1E=^HoJ@yD5m~f z)wi!&Mtr$M@*FqRKv?3#JJk6ytVYPv+;T_tGHm?XGF_rf%bFy1i7!c zt7ul87Heyu`UM+NUeV@GU~U@5H!|6!Tg_i%rY%Kor{95kkl7UkE+yc zIod%nzzWUAO3M9nmI*pwF=-P&^33I33-0vo@yL|c(+%LD=p_aG&%@$p9u)5n=vF3$ zw}}VBE-WhHwL~TaHV4*2F3usG2tWYi00JCXkOA6Y^ON`wc&QOkry3nl zA-*r`1%dWzXv=HQrXDZe*FVPU@3?l_YeD(0*(S@PiC=YyY;|C#7ML@mNP9ae24wE^ z&#OBUo;_*Ym7D(ts-aC_T>@zmuRH zldSyPTBtw3O6!x-lTti*jBEh5<>z(?&5i}Xalm}6VQhZvlIwf@mQ3EK*Bb{$0SjL@ zTa!kYLiqKXfu5@em8$&36*RU4fRx}dvNPnQ=Am@8n-&?h*Ok=*5Xr~`lu`qwD$1)L~Git)VhayFoXFf(^JXd;`sj!mnuzQM3`^XbZ@-dG+a{8WcyKtrV z!O%K_0l!j6U}|r9VmE@NB7^+=ZZ9;(#bHTq72e6P6ecSZ1PQ$L-a5?zO@;6fWUY6G z93$0gXg)w|JE5lv{{8rwsB4e;^yq*Z0e&{R-+fAldZWaqhZVyVXZE9ih3rbRc#XNe=pAU$AIVS@`+|D?{7F z8JdBxXc0ENPgZVsyU1JRKA?>8ye!?!yQgS!&hu^;*76$EIu%}x=dv;2L)Q^(R@|w$ zB;5^C5~n;Hh1;mRi!B3urI`>`bE$VjRKK>0*M%&=J2VCf*n%hCeVF`f` z_}3IIk=dGdVG^@}U!^&x&XDXR9RwAU>Ln;&Ki+wN&a_M%7SH*Bah_&3K;iFr{l!xD zmy3Vo!259)k>F?%l76u_WJl=1&{Xf@unn$vrf-m67F||CKgb|x$VFe|rFuX9(h4R? za0+0j_pc0Z0fN^GUZcEE*%FDAnKlB;WEe{GHOousoAPY>`n-; zN)#AHiXK4)`A4DJJW;eL)#bhSRs<(_Lcd|YK{daM9t4I}v3MWo$YyE4(dr8;U055| zP2YmNMX@4UbQyTk3Bm;7W2Z{LkAjU!arADQX4F80;E1p{Jbn6vaAh+Pjj?jE0mPc) zZP8Sr!~~)Y;f6Mn;K0G-d^1?+C?=I6o2V$P^hWO%i{tWKP#O(`$Jhu#=iMm73d$j9 zi;xBmIAlZOh4I1(&7?Kos^OY(Ou5x#Pvk=Om1vBx!qm_jVaAKjz}w@~1HvK?=roBo z!Jv3n>(>To`CPnOxg~gigk6G4&zK<&z<9MgQ#`jgo@753mOAuWo@F%=_o`BET*np(Xo)v}7oVUfX7UWHPx z#bH>8U?i=e)N{6b3S=3#QG4Btk&<>A&nHix3|h-L5HjK^jNEG1-`s4+2ejteMC7m> zc+0cb%t!e`4hwh417Kg$1E32^c)u)Nby@VW><+z)zCi|ETOy0Ydg-!n0`GV6HRa## zDpGir^wIE6-CtG8-O6;W= z;M(q{qCQ)E?3=$$bB`4r6y|!ISN>1TG(56$8n(2*70@U=HzV_vb9A-PcuAvN@3O@? z)~}R$@cL#_C|N1!;0-$!D5_+=^bUGBG-leqfNLHrL$QSMllmJ&H$=Gc4fW2{-{6!xaIT4N81{ zvu6x5>{LwSZ<|`0XUrC7d^}?;Tb}CukAAsj;Ebh06m#k6!Fev)Bc#ydnnzs3kQ+$s zNR7hVX>_^WMRm;o-5kDA^Be8_j+0q;-0a4+p3qmPo7S5fKfUx9vg)PD+vv^Qn#~jg zVV`xzyiKbt!}Y*igG6G9dPM-}%pLGeR$f@tEpMksL{=xuWLW3Kj`COR4D2VvfB!+% zUtTd`a72DkF>G7s_Q|Ah zhRZ1A7+;Pa%a*SneDg>3d=p%pe4Un&7zf_-WSRjWnPS#bWDOOCeFL_7AiFJF(+KXS z{*@JSEmnNwiFW&=0*CRxNwN(b);`lL+j2hH$Ktg1O^J@LO|guQu!I%Xdv`!d_@M=$ z-iiHpbx;F7Nnf@3Kh~Nr-u#ZqdB^FCcU-6AryG9pjdx6z;u}|-`^d$y%1<3QDmG+h zJ#JIXrxdx49r{rFwgQ?>93pW*tWBerf;Lp8?_STFu4kYx3FTYxs(Tiw%LlBbUx)u|FB$iKy+*6Yxj0(rOCO<=EkH+`MI zRMZVDlD(eDibZRs_w9E@4XFwmuTBqc7oq|~nq*k9kH5wFLC{6Njj!H--SR>jYDT?l1SS0UE(3bNSxMkuHc%8 ziQ$4oc5uc@YVHX=?Y`&S71Z4YU;J1~}PLo`i)98(*8u1LSSx8gCevWoVz&NIU zQua>lW@D^1=?Aw~*IJqqmS-?c^6C6F%EW)Yy74OsN4yCc{XQe>zuf@Goo(Sl(-Jtu_P2&~G;d=UNnD&}N2i>VF0{g@AgRxdyn+sLi zxl-Ho#WDv`GDXPGe)Sve@89|5kAC-a=}L-OPLXJ?K8V?vws`XlaG&^NsJCh9$$$La zy(Imm3FcOqEmXTH1}tSB6_qjP>Wqg$$L8&tu|~3wztnlHs5E3S>>701Yfz*CneC+X zJg;Q>(DVkyGLiryP-*l7s5somA9TIOt5g)L_p00Jc43jnU{Hs0h`&n#T8H+@BaXvT zJ_R&efX8x-cxmf4%cP;hQfFDRnbFS-7hR?<zS!amhZK89l*dSK!1zD~F`BXHx9tSnOB%gvk&HPQiy=48uIA{=@PfemF z;RcRp&ZkXJasx;1;Hq{@){?`5gN28Ln_n~y~>l(T+ZcA==^)`w=e0kgIWkRh&kK+5 z6tuyy{Sj4OEs;4ODRw5Z%P`Ua8CKvpEs;1TUKwo~0q-|ngJa;QJ-Oy@#t-3Y-^KyL zF&DDu9^?qt>&YS)OQOOUUJj&Ph|BMye)7wxU;X}_pDp>_7n?keA7}H+2@81vV{Pd` zj(rkq(_)e+Sy||D33_vKoxTbF-PX_k1V4!5?d&G#^#_?dBTVMTXZh+KWWNJPXV01$ zkOqpWrARdul`gsvaDl!mJSBPZsu_AhANlXD#pO$CzLE6hqVH|_%JH``KDa_2Es7KL z3i6?O&{g`8G*`9W_0haKuX2|L-cD%>$l~BQsD|7Ud5+gh8;|Ds6whl?8Aor4?4|3W zK}15}GC@6*kL7sZ)U248W~KG8d9GP_>H>~PF}s`{Hed;~?4Y zz)-3)gVKJADWOOa71aeE4a%e=!#;U_o(NmhKX!@{6u=6igEaF>@}ZGfkS9TDb-Wab zQ=8{@X|grDXb@lmGCtB85_rX+l3^@OEm8E)$@4pu-66v^*I&0Obj&7yVsAJ>XR3GX zql|Bvpp#Aw{S!Gkjodbyq01BlVp=U!RIFcoy%`te|g|qf{%L|Jnqq=-WmP;UP#n!QP;{E<-K&l;x)W0(sY>J9+y;A zp(IU>tonPT#ODNbcB!KuyuMV_sHhMFKSYGoNc_~_R{rx9X!-JoL*J}|U;hKeqhCjR z&?r0*ZoK#+l(JXxV%#tAQ)Zkb)+|CcoyVt)0LuX@rkK+_I&S8Gs{dxBrTg=eq^VgP zqe|J&w@^P{N&Tr=ZhWz8@%yLVY<~OF;`iTwe;A~}F7o#H)YBhW15PV&@|;V?0q^IG zXq(gC{+B!5Oor#jm&g*5G>sIP8J-M^(NQFoio)_wgM==w4NbxhP;_PyV@9~YKYl=ESh2j+(>zFcF;e|SmOC>9#W z*;y?vRrE4)McN#B_CK#k&qS_~3`oy=^wRgFTfI-y^@8R|n0P5TRl+m!8}yk-dz(_@ z)QkfmEOTnx!@lqST1A;G(IdAm>9$#Y2Asio=hI%fs$oA6WAs1ao$1y&|Df-#*$@Kp zK1DwBZQ>1(k6tNZiUKZ9EYR`nugjY6pWWC#d+dpvW&i%ic={Z74^;lWgugH*u(&=L zxIi|J#rbyJzaM&Ujz+egVu1gUfoj^F!Y)M(KPCJQkT9Oi*O3OY*FVkW(aa(c+KTqS>_6<(3uo&{v1F$-&1G%_27_PRqA<0t^MsG)q--D49&%Pc)lgF z)-PGvF2vWB^q~74_XJ4}bfHg>6#1?KYXp59R9ItTCy-A7-KK@~xmIFa7qFX_+h1~W z`TNpZFM0clv_#qxzFU(i1&S~KQedqD<8P&In>Cv?jo@;XYA%OY`{6>>O&hQ5@o1on z!*{EjLGEA2Z&0NRl9l~)tKe>oF+-xRYU??& zZW<{x^9M31CY>T%(I04o26GtiUkbf5qD57_D&8^Z)m@>wCC>IZ4|>!c%4$+2sP=1# zM797b=hQF0NJ_oz^puWzgvR;xjvB!}FWDd17iu~B&(qOcTudnWKzsIgB-VjbkU3`X z-9$0T6j@6}9amkP2iq_RAm>Svm3g8Oq7@3Px6^r||KsnP9sy36o!U_;``L@m-~B@_ zmGIAhxctMtiS`m7b*_(>*+sV^Y~C}A5%9@^EP@@Zx{Q<2`W_843W3YWJSbx#J?hm+yed% zyoKInW^>Xh21H3xaLL@jPvmv`W=J}Pk02Gk*8Nt{ea{*}C9q&(rZvg&{mN-qltZpt zJrV-rrel*c`yj|-`o{w(iy`dtc9CFro@JQ`msq_6=Y&{_O1ePdBc4f-<@58PWuo4H zz4v3MH2OHuU5u^gCIt>p=F)m#Jh^8wUKL&a_9;uX>=%k>xy{T%tf!bHiX>1`=#Lob z-r3TNeldQl-Hwsdv-R`=-{{GgE3$U-Cu71#e+K(9vorVRH!Nl49F}U!A`~~0d&Mq4 z291n5uLS6sX_U*r7KMYZr3>5XWd_x-Sm*%tUaWKJLe%Biv|yd@VLphtsIGYL(-?0+ z-5^x1y+e1)J0Z#vjW<~QHx8R(Ei^{E9Nb@zo^Aac72LlzgvzuViJe-wh97NYZUQy;q(HBOq6lP3}5PCV!6< zj~dTk^JosKWk@w(E5)j}OmgMebEkWCa}l z(#@y~(g7gcuLd#jCtmkl>t_IglT(-P6MwLh^7w+0U5 z{!j}#Fb{_@d={#>J?`t?hNy{dVaOGd>Aa~zZAxB82Y#;(ual_P#%BXF0 zvBSq)QdYqA>11ip>OYH}n=f|26ke+|sZdbg8x3`#vP+k#em(w5t!n7j{v!za5+7wt!qja){U+dBWO4`{RtVu;R)C5+KXCHf) z@BNvO&C6~R#qjJieY^+(qnrz~hxWh*qql#O_B(YgrldSfW)%lEA7&_gw9-=k9S?mC zGlplg{B)ATBs+B1tWth*L?1ViQ$V*1wd@HrEzF=~573O==WkbY(%&|$UCz$FIEh?v z;I+#`quJsXCB06OYg9CrsF#Ed5Ui-e-5O2A0Uuzft_;9EI;ef-9`HHuH6n1;7m=dS zL^O%vF2>-8#Cx3~IuES|>D+~lisPgt3>P*keUG^$Rnp6eXWjh5wi$VlK)rElO<0$t z5!h945inFe&Z)r6UlIm~YDJNtWswSLarY^p^IrEndR6GEP*oQFIIw{CctN>~s%UNj zUG1+zh8N`b&xc*bIyP1#U;)?q$jk+*L-GXJU|cWlevZ?f8G6jxqVeA_0TDAp^v=&jd(=<~9tuibfVo2bw~Rr1?ZYU#aSUjJ_MH#dFxw_4?HlEO|1F7mGc#@V|8 zkiGlb4y8tksYk7>PVVrVNUed%(HK0@i4P0x=dUmveA9`m14F-s<^D#HS&MjI;C4)7|{LAsX;AvRMfNTuH@-CU|v7J{7BF zna-J^Yh7Bvx9JWU6ls;&KoNFbxwt{JMOGh};MEUXMfwNy&JQ*yh9vQI9@wzieeq`M zuZh%Oci{gowi0dQgpw(=#yGbQZOou_^o`34X@g(f|DS%x$?~yE1|4{51Qh|JieWcU z(nN}^qoON3clow(b+oDqqHdVhq-yki;?+s7wO65z`N6UjDCT>x&mHqX+4Zf!wVno0 zw7yfdgrqz0Vrh>N017E7MCZ4|Vrju`cZ{~6;vUlMSI$2>UsdA|yT$ZDPLbf4xQ@mq zYRN&4DxP;gqH0!8xC&`7K$wixC8>lZbe9B|D^{9^7mt~xxERASj9hy0Cs^0QJTAuA z6rA)c<(c*F2CN)=>syz}1_#DUxe->jQPM057z?6xG=}0-sQYx^N8{TGr-QCbVtB<7 zgY52*M%Wb020!J#=staardswuT*_a9dxdq}au-`B4U9Eh)`MiM-fVY`5lAKnuhLC6 z70Y}@LUU#`Karkvm#+$!(FmsrFX?$Y(w5-oY1@hXIsJcU9x>elblBj&nN2-pJwhs4 zl-Dhe(5ptNKB#s1dz&cM_0|N5f;}BHO9t97lRSN7LVhwb6qAnB-8x_ji&w-@m;uqq z%=6gwegp(X+W9GdjV!(FB}VcmIL79yu? z^ev#*NTE(53WS!I*BZ-G|1Zra|!Z0V86kLCH! z_Q%7h@niyxQIFaEu&?g5ofc9f9TQ(-0qrG5mvHUaBQpj)GOGP6qEO67z=@EBHVuFf1G1khHVe)-l%tG;OfkY?KBpOM2;h|cJn zKTkeWOXl>GY}lT?B7~8HYrQn~{EusreNhj@O;8G^RTf8GlbnUKI{6wY zGVT;Aw$bT|LRJC2GwiJUY56+dA!b=dwjYu)$7$pulhI*hnjF^)n19~;Cog+7@^aq3 zIgn$7LHs7sN4}~(u3a)!85^=v=u4zX(BY2R_FFww4Rms3k)R7Qpz7!i!T02x;?qlu` zVyH!hv}+Zfo!mP9HM)UT-IJi=iyc7n`9o?cgg(C)+u!~1&u&$iGCw+Oh_=RPm`f=s zRH5&nqH)u=O>{tB%TeXAZ<2hFl~XMzH^F=WJN-yL8wjL^tp}wC*#AM=`CJ9s8#QaU zyq@HTsn5${QeJFB-w?1SLI6kwa88iV$Vb_{g_hzJUx%iXeq$1?R5kIsQujA*Xem?vQ-; zE?{tPaqA8_Pw+@Xo*vak-u1+5Spi*=64C`vZWiSQH2StGKj!v{PkUY9H;WI2)r4zU zXO)@=NS^XI2BDk+Q0i?K-{Wj?-o@VmYs_834AF6T_7nO}h}Cw>M&L7xO^c8=Mx!6S zQ@QAo!PEB6+Wd$NFe?Xg;Gi#<(NRe&@su=n7kt-dZVYdKqOz-Wgofe`;|q_uMY*Vri8p zWGjNUd88|@aMh%Moz1xpq$xwq1RG-$^W2$5c(Ab1M;}~yI38v$ z2X;HmFmq+mwLThRNuVVWL@` z+FMRaUYUvJxY0yYOG$w_at}^4TZ0~v9P$`;fYgfp{#dY)&Fd0hmK+c^b9?ENZ(V{d z=H=8iD10whd_=B*jp5~{9>TH%sX9<3un_52af{n2CYw>@tGhN-7;pnYkt(n(L>Sw5) za>g&g><+B#J*9XoY!j_@n-JUl^bcqBjXr&70)8@!TMkj)Y85%;Z34$y^J-?LyA^U;2JETie>yzkmoT#2s z#$v(x9Zrs-HtfTwL^fo=E#nQqc5nk7!`tAjmTjEc?xKN&+q&Q+=Qh3y-4rZL#`tZv zpc}lIMh*sb@47E1*FkM`JP}()2>(1=G4?&H-vwp}UvThmKBk!pjtY8WCimc;SB&rw z{gA${#Hj0V*ai|W*(J;canXEsA4u5N2IKO#|BZdDt~sX#$q+5YWigf!ZdDG=JVWn> z#G94M1H$w4<5{OzN1&ji(lg)rh62h{{ju*9SW3gOSSNZ&7UKnL3j>YE)E)pb{simJ z7y*dddG)lIWw1nlINE=NtZ-mUlw)LxHd4|gima!iQ$r7sBjQt#j55g14Qo{veCw=o zz4WoL-?r;oi-yBOR9MuPEgLaIg`Z2y^j{iKQQ{K*I=RZs@pa&((2x;^dMIf(MQ&5k zc?#q(Nb*kdPMe)J`^MK}7O31L3k9LWcPnHf-?3SP#1ha z-cJv**Sa++u6eFo&<&Uqak$q)9oofK$ zd$r0&QtZGWJZ5CcYA7i%<5p48aZ`rG1+0}*3Rzb@uLz%nG$^hJ9}CMC+y~aVMqf3k zP9Nkn1>Tyy+aDD$i+CG9hq6L3ki}+C$szM%daa*=p)!1Rtt{_&?x1g^y1 zblfwGe)PpF!cs^I{pSM^x-94S3u^-EXPo90xHPg(PFoSwH7A~TT7cJe;`5#p6)j{8 z4_l4@3;JOOj}t%tX&z;O#}l_Fryy^HQ3np}Rc$c>QvxMjOObdWbN1v5+rc?t-GK+-@#K$oi2B7-I9L>wfmJdlM; zpFx24b&&tI38X%{WD~+c{NitSY`4+gU-B#RZe>@r{pL-B)p*R#d7IQRTZ=hx3hyN& zt8t2w!mh*-Dte=28Lv>12kZ)k;4ovzu-ro}Q*B}`<2BGLA;nYAZV8#-&HRnQd7$yF z%3YX8L*_Us9FYC~yiPB$UOcr&b3R>V6A65vR7nLU+x_}$v~V(hoY{8gs!(N^`60MkNRhm9GU5sJo$ zFv>(fqO$}W0@jD7PDOg7h8%SSzRagua0!YM+h0{4R+WD$JaRaAMtX9Ok*n1MP zdfb@~B*XU1?BVx+p7|d$4Q8kRZTFRAn*;l@4Mv8fl9HBDq!@QI;{$US;?59i&WqvHiI) z?kDe=Qo%Uv{9}gPWu!E`hWil+kF8^u^7qS63G0H_c^!~vv+_fky9(p`k}V(T@qG@{ zry1Ff$-P~i?D4StJFx9A!}6aRs@H(H!0%({DYR0xtb)}hRdtYkp?%_y0{h9Mk(BVR z2~5he97-ee4pT%}-Tp^+>$Zsv-bmK;cmF_A9oQSu7Oxc1Mz!`T?zD@=u@3Y58qw7BEml6|KPh9f`P?XY@O zEqRcc5&2j6Nv^3#ro+xHW&|-GxYqDn!)}PH=WCT4f7Z#Z;SZ43ux55^*xlD2yK9x{ zie~qXRN^A_+f`9rB#%axx~-ly+(N|;c;c$Cj=j~h5IERxl0yGdSS7XdHn?R2xj%;Y zTf=OPqH*A6G=^+mK)?t$l>FMSPmd?}*MWm9W^#WG)Tr^Kf~C6Sfvl&+l&!I0(fgfB z5%Qbg|55dqpZ)HgKmS&^l9H~VNQ_;vVae~6yuAHwp<%`m|LaSwBykEUF*58~loZ0= z=~VQfYZAnj%BF4eYve#QDj-L&dVyLN9~v8q9MvdXi==OPhKNML^jPU_AfyrXmflJiP@Im%=VVh*N*A-q9z6aEdsv*fBm8#|F@99KI91Q1P6?TN|3xLu|Hl{b!aBqhfETow&KFiGbu2 z)}gSQ?kV$A{E|3oS^wNN>BJ+$0#u-{G9gG9{UOXy@y8h6*ChsTW>a$D1(G(F8i)hW zNV|+&uY5|nodO1+=*0=ayZkj%3ue}PYL$2t6O_uiC&9CX;A%xFJJI6;RKM<^6Qmnw ztaew+1{Z9VXVJQ_{b2Mvxi{_XiCG99V`o=?HhQdWerz5*aTzsRUv`PR2JRu(J?tIw z-B3!<%eg4Xqi-vA@e{oYSjCa`k-Nn<%v-Z)WEh~pqW)~v$UF+h&;y}7)R}IkRrf-| zdH0=PM8Es9-z@&Kkd`_T%7LS>W@_u(M7ToJ=YOk^`F{{{ao0=B#HZ#!(g9}J?35Ra z)G{p0YZL8Y>4R~qeRKBV{fvNnG-efO_?n0u0-h?+EmqG*2~41UFQChPQ>WT;&SC(i zv5boGHCy2b`Y{is@5;;n)6}3aStU6zpn%YQRAJmEN}58EWGWi7@Gtu(LqhFMVYW-F zJEmM0&?jY^15}qt7jSPbZTpoiG4Y~dWQ3U)b!Nl+e>t)JrnkW-(Jc(QLNXmVrVqK* zqsDqCC51hmyr*R&D`s8br%LWYCcauW>=d&gW`QQ6OR{x_8Z!2;`0Y^)J3WNzvjVzc zRvr5^Yyb~CVNOw+TQhw#tl9Zy*iq;4R5HG3NEnU3v4z017j`5NVkRpT1=wVPfruQv8*3jOYmbg2*&ZCC?VWcWUwFnd;H&%BxZ?` z4vdIWBWsgQNdY1mu-ef;{YHi;o>wGDCOEmYxa+}jHvsYwrfPgug{%RBj4yg}+`9pI zaRN|d2}GW^EcP=kPf=C#vwLx-sg6vX562BYo6`(wm37H#U3Qa=z#XBLmO{ObR;t3C z)1s?@Jba$8Yl=f&_OD^X(4IcNkZ$v499dKnDK+ova; zB^k^Ni35jC_8J+IB1#J6bGcM>FXYA^of8*O?BCDVaj!%!o2e)LM_N*?@-b@zuiGtc zhFX@#TgR?vt@6re9|0nmtHJ{k9k)E7GW5Rdc~%Ushqj{|Ydu)(>a|4--Cx}LVdixSRSJb{=|0tC-0Hy??4PE#`4p zl$a^cX;t3g9P(&!E?$_;*(n?p4*MSnJge-b*O1eK9L3G>Yy5gn{fu>ywXP?LPP8(7 zx9`pH$3YITy9n|XIqXt#xwWa(Rs+fuU}n#)LmneQQA;Ja zCF9AscVOSeOve36^2oI`s@6v>d;Eplu4{qzK(AesHlvgu6TX2A1Zg7Dl}9=8fyY^g zeCx?3mp(|JTgObr#>%0yB$w$qxLEgDOrF~8GwnO3p?fpY@C?yW9#TOdD@lqU2C!5& z==IVRKTy@cSUAvB&}-eUvytEgV%(Y$`jX>NVwJ>dh6Ja6kTz|Vt(WF0R{9~`#2Lt4 zEE3>)w2plPB44(hC2aT#Gx<7S*p|^N?6_fw-_HAOxv8t`u#=S;=QS;Kw;afDs^_cf zf(P9SXSVrghMx9PReI{^Peu8m_KdrZ0TU*G9piDkU19{-$>9V4^Ol=o3G|D0vX~^l zWXS|j9XM*ClS4_>6v?Eb^S_NkY3-6V;$FHfOzV1vKInEzdeOkDF4#+X_>RnEBfEb=t^7Jmj=fUK%zc z*@NUnS3QxF3%62|NEbvtu_&U*rI;$FP_F0SIIL;@Yae6h#BpPouO;r`nkFSNiOxE( z?`kGp>CAUJY0QqTmA1;N-n#HrkP!OmkkfhinM6JL>LIf7dszW((ru7!(8AUD*18N# z(})kgslRe|e%b=tx=+vEW5bbEW@g0RJlB;?FZ}MGKl#`+r}6pJ69& zDuH#FH4Nd^4dftu^CJD3odB=|lThMa(U z@wzTadE`;nacM1`>1qbG6D2$^8iAnmXPkTv2R@@D=8{&u%R zq@BLUO%1A1Yr9O{G;Gc6Dq?w|HC23IVr3jR7zEqKqZiF-lyoCSlBj5$ zlh#Q=xMI*PiSxiEI~4y_%T|NT-LMlT9a#6)TRWzf!U*huwI5;CbJzaqr~Qf{X7cc> zUwelfcxASYuNnd43?)5H0o8wWn*V0eUSSNcI>_hqeUKBL7hwR^_xJ3~CK_g-a$sxIVg#xWDJf8V)l$*@b96qapP4FI<8fzUnoG71 z>`EXmx>yGSa}}v01l4V#M%LYs-Qpe^IY*(v0H}rZNj_xjL_3B1$;M4D$F$$hOXy>2i{ha)W``->SSecvOxl>6Br;x)&uV*zSt)hsAiXIYI1r@Ru z*ZQnsRR#6K_nqih<+AuNeU3jcQm5TZR`M^g9!4G*QKvi-hP8 zAos%Q2aU_abSH_KUKP|V&Yp3|HAA$L(Z?%ejAsRXUfOJ#2*w}uKG4d;AyQAg=n@+JIKMdMeJCX znw_v98Q2N83maLe0BkLAYz7c>c9S_aMxJ11A8XxzB{SWPe5qUf~leo6|7{=2E@Puw4fABd|JPk*)o3V^zpfbd<3P=9N2>&TUr zp^IytUcolFLaz-zH76$Mpw|;liN}@MyM#b}%zgrs6|SnD2v5yvVqIf3h>!ErXFT>w z6dZ}P4UViriaBedziuTqj2|G_j7zH;eKDw8|Qf^PaJs3h`kY7hC1*F3Iw#Wx#~iBueJk z1}nxRWH=iifX3|7CKt^E=*NLyIXS~%YGU8>UPrQ+#cdqeE3P#%Ol6d`m?DK#G_Juf zF6^3vHLZ_GF5T?DJ+wiLDOEs@?RnPyytskh7j_WHzc%vD0>|Zc(htd|T>!B?;sn@} zH~{~qvsSw&aYiWhSljJ1iH#3SBi6?z*fH%ScsR)Q-wQFQE7!?AU{YCdg=`&K3NQ#pYJ~q;{xoM;*@Zg zuN|zA4G;scF&*#mpTDd4{c_Wlx5I8vnpsP)m&OETd*1gNavG%Z-x7~(Iz8~dSFOuN z=lH-bv0Bz4xyL%@v(;y{N57=NrIV|f-34lwYS}VnU*u^)lOSit)4N*b0kVU}GgJ;n z5j7R^n3)JwrbgVRzl!mI(K0_D_wDEMvLqQT(5;3KmGNKK^y$jyz!a_qJsKs z8%VmM&`aNP1FZP*YAh#JQawLGuz~cr^twX`$-PBpoxz*{{22qw!n5*U)du<+5cNcQ zd^J(OGKi`%LR2Xwg?;iJRCJc$*4#wNo-f9K7nOul9Yj^uLcsogInzWh`ir zFXvT6Vo5hdjz^xtec2-*xraR*6zYFkhldvq&-elJ!jG`3!Hv0}-hIyinCyp#r;xi2 zJngMB(y1S!q>m^vKt*?Zl`GDZE^#NfkewlV;93xIoK@|%j?>3Z^y`Jht2zh;@13#N zx0^o<1##IBe;Jxt^6_$#udZbp*4lY7uB z&Py#*EB5=Jfta&axz)du-r$@MQlC9A=Ut!OqYn~t>#QY%U~USSf{FPjSog${D`D}| z4fg5VJAe5bvc-Yz(>|jqs+f`%QY4>>-VcSi4S^>~hA2gl?U^1}OQ28rY-KfY z^=>dK8pS}ocz|G(Xumwp`+(a$*s#L(RbBUO_hvZ_vO+x+~s7erv@fHJ*?JC zpM+F^UbKEKhOJw{_F%BFP}m_k0L&v9vX0x~ZCiQf9C9+7zh+mGCMr zm8#l~6zv6OL-s-TXcneAxbk_JzU%2rfLgDj)MpwCdt`5nE!9&3Z65nVlVaUAedm6iiMzXZd4r=@GasMS`2*k4T?5*1ZLm z-Rl$&LhLp>nG6kMc$hqIuT%5zSihww{C5UZ@@7esH@W@FOib}ctFl2#`j8^`sc7U` z+ON<_4ny!I(GOB6W;Xh&R)codhp;UU1rN$oY*pTf*g&#C6NUr*GKN; zReD~RIIODHPj>u$-@n%;*d_iYew(OEa+-aRQ%|<}fqF#;DX_%7Uj?-~r!A8-8eIyvFM3#;2ke#|9G3eNC3D!S353v&4~c>@tEM8G>jG|KJlhmon0 zb?hN=Aqy?YveyRPhMlk~YLC2^PGRE{-4GJOXJU9OoI$mZ)$Z~zQiYXRsZimO?~K_P zs#+IxdB9JW;V0Dk)cPP99MoQ6#T798z-}Ar9#pW_%tK|{PP#_XOH|i9JGr}r)${Xc zWTnQvl|ff*hH|7^rmdfC5yy{ZfHC9fT6@_7e3_ry#@KhCJn+ArSzk85E2^4%gETUm z?;O~vzG4K%7D{@GBFCxd`oQfG2gTW;sZ*D91g}zT7u3#EosivU=ges5t5PKmqD`#( zGaDf@zBf__j5xUF`s520Glxd%`igiudPysY@zl`?BL#ri8*Y|mf@ZePErFF1Qb5SZYi9oOq^ki_)nEDM8M4-aO%{ZHMs3k%Qqs*7e6P{B=)JzNoC3O6n6J1s zw~oE-HKd)`%x)H4cfoyBwXEE2ukUdZH;dVrJ^PlJA2A$n<{x0?6O1el#R}iO?`21a zK16dj1gEHS77EAK^Gl(4JUJ4k6W$&k76F4R19O-i335zEiy=BDEYX7zIwmjt>DE=I zfkB7OA2bu{OO$2{b=;JBS@ca(E?3L)*}49`*H^SB5*TjwDQEC$N6>LBm5A|MG$uDV!j$(<0b68lV)c6(#LU6RJgeD*D{L5F<% zLiWoq%^m`R9Hh`)6`G^CH2Z;A2jRhk(3kbLi7A>GIy5HOwfvM3ny9n)&;G}&80={d ztaVvKN~c|o$c8oVNYJNS+2a0*GJ}=kdWc*G;`0^|1zX21<*%PG@mj`^v3%6W)zUm< zPQKUk>+wiAIdEQ&87U{s^~z>ln|{t)t3*}w`(TW6*z29+rlqmF#NA%|S!rN~(gT~h zn*>{&&v-P@PwZ6kU^1kP!2<@zqzk*8{;BoBA58Pv&7_TQAQh2q!OP}j`*nE}Ngm*)FUKLN{`8wQpN6@Dm)N>wGkEDc9*OG>y7zX-;6b1xEUiqR`*t4Gyk@Zv3zyh7x4dW=LH2D)>LUn@G5fUpc4J&}9Y%zP-=>uSneWQ~b=hYXn zon0KEp?5jg2Ji9ARWz~&0t;myOY&${k8>8iP6`}do=I;EJN8c##!%1)KhnR7Z zAXU=Cg*_|e=x+!1{sMZl=uXHD8!+|T5ZI&fZZ)8vKFA2zlbp)KhD}NEU#e{?GsO7W zoSN`dNewqkcyVD?z-g}|k^Pa){EUEpP++lDY46i-{KY>l#>$JkvS|zRSP5~zbMgNf z>`FjRM?cXri`O}D;zYlZ6}m}Dbrk7@#0mO5taP?1x+R*3gy42qH|>G2Tn^FE8>imk zXxOSkp-$8$f>7{4!~-tsNaa9#yttfK02;W1ZqOD|<`0s%*JFcJE$(>=(6AIDvsMvr zv)9F-VW&FoL-#IL9z>v<0$bd3$$t4MVJ8NjoXwHh>tBTDF`A_v(TlkrVu4RvrHueINc)9g^=dBe!cY2uGVE zTOW@B-*^8{du*BkYHf4QzeiGD8K?ng&ZwYYE+qw~Y&8{~Dv9Up4cg^f>etIfzJFws z#EOgEWY@^fBwo{pesFK}X)OkXa^bpdy|Xcf!&;b((UnavUHhnK{|^i>IUW`KJ#xl@ zJ)It-f$yZGut{)Q02AGLo+RgmoHakLZm0r@8q=l8Bi{RRHb#mHHI4u zX`&{yOFIzqPuXD(=yq)O6f#dv^1Dx?&L-g&LoLoI6urH|!2E$c&$s*jzyAc2)2 zXb4R7E1;2VM}>E9ge0)A;ATlO85Hi6$AdV9S@>BC8(SF|b67J!#f{h z+n356GDBl~`b$@U@8nWszwo>~-x;ZzKxbKT7^b!x0I*F~^Cy}6)?aYlaG@iL*o`iTGarB;&Yz}Bb4$ogba(k&E8r=oTAPT5Y_!Un%$1E&+@%+7ghBq}7x z)Ituss_u9s0G*}T`CvR>yy>&n0K~Kpdz>=@$mDkExBOfT0C`_^?$0EinV;*pK`1C{ z7!9Y*loW~wHc-(;zH5QzCO_aI9lM|zRK-q_T-PLa?1GPlWnvBcFs*@og9!o9)4^u~ zq^HkJ#Lv?wOBTI~`$s69tj}L5V;}OZ_1W~sYN&Br!NSa8q_r;ej14_UYo>OPS{u== zSuk2mK!6jD1zp;tiFs>sr)JfWNe0WI*uLj)dxJT9OUI2Sh%YidiE$inWCXJl#fOPK_ke1y^Cj>XZdgWvJP9(zW0Vv)+YeIF59wZh9{UI>M#AJIx6j(+9*hTdGo_Xm%rW-HRP1~Kl|QH{MKbSTJ_buAL6}KiT*K6s?1nW z?v@BsRcmmfBrD)B{r9Eyi}Wm_CG+;nKaE;?>`kq*WL}QECTeLDJX^*d{CfL;F8#lI z+~d;2psu=Ku2r^$$F$V(p-NJuk7E0y~1W z${XxdNhkLnr;eXTAN9WgcB;|0G)xnLH0LRPH32siIg0%XJKH84JL^DV#i%oUq$YKC z|6GH`d9PO4NQz$>i*wA#;?z*mYKl}*(a70|ROid62jcF4GOzjIBv zhSkVvlpSZSmh25obIqbrf*H9R|9(w!0Fv*r0bR8-8d((q`7|ol^^z@MIc|nGvTDOl zF4{)l3A!*(tE>pvB3}{IKdwYO<8ibOKA&S!jJvn`>>G!&Ibs8RvZlZL2a-Bgg{=dx zgf&JW$)lv(C?MvGPV_oE|0qe4^s$#uEnr=r*UUdVzYj_Q)UtZARe5&)Jx;3;E-XhOhmDdNz z5~C^-Ajw~fAJi>kqFswe9T$t6hQxcF%ch+p_gt|Ue;MbBpG5^mmW_qER#sxiTz59R zu1#9^hG~uC^E){XTvn82WJx}w^UPHg(Dn27Plc7ST2U8#IB-A&`-gS>B@-~##)D(= z$QwUq_cQZW9Aa#@qDIGewXM!*AABgv`_bV`f|Nl=T@BS?MZ}0r>kMBkcX{i&Ts4-$?lYcg6 zz0@=#j>**Gz$V2^uG}&4B`A)~4%M3$RTiYkt%0JFL3T-4F}37#_WbSgDqCz;CN45N z@^^A--_I~5e{opjVTR}hNfTmtSprGNmy?)G66|T%MNi3%a}#y zH+@LfykyIm9Y&K+1|{7@krd$Y2uhV)V0H76yD-*&wd9!SHWU?fNruFokYC*hDUA8- za^5kKy=R(HgUk#tqaL*Lp?Schy)FHslK~vXpZ?|p663&$0%{{5BvI1!6j?(>pYu36 zPro%Yl6Ho~y$1vnIO?Abo<+m|S?h^AFhBfb_#Bg#)c4--Gd zfumq%7=Eq+1ErpcCka^KFOV!6Nm5Hl7t}+vaGS+fg+1W0f)+sm{U9WRg^7-sKL4>C z6X{!nYkf2kjf%=3q~TE=@=XB#cWiP7~WOnHVP)(b0FVflIs2u)6!#_#vSYE6HuXiE#J}S8+g_0&yB!P-Q>3{!9j6`m;C679aeAKFojVV|k4Z zo7MGp_TA>X$Z9GYS0*6QcG*8U@{_Q~t}!6UHgRTT z)QDPwl2Oek?9e<)=6v^^@60wJVo^uoI#M)+fI`Bk^}s$#3fm;RsOWpHy>y~q+^j3Y zLZG)V*U(U-eC)>bDnb9VS;i=+z#|g@U6baH?+JzHS z4{U6VEF5OzhbG{Rk*S;9|ML}ROm}#gSQH0#UCitUVNMc;Vo_dby|f$ZxURE1CL%Td zxv{qhLC-y9{|iGa*5v+sVIp0gCq-O*LR5)c9tOAd`;A-AlZ_4>I;$}9r?*qmY>I59qA|kTCd#6-f>f(L`X#5l9}Dk?$9Ruq z8L9nB`MZG6{MKAmtLzNFNYLl5>Jz8TZVK^{I|W<&d+(Ob83fgf9x(T#>&`_+wH1( ztbBKE{r@Qp*5g?8KR1%fv3Benc-yzx$f6ver1cc3p`uZZ<(j8nN~JCs_gB?2)SGJs zdWk!njlrF?7IWip`QkoChS{YJiRe5!Wpp<p0a^ zI%yV{EBZpzvbC&p?n6!uq9Mo>+|Ictjp3mN^7VO_d22+^+U5EuupWBkc^(lqak}Yt ze(l0*VE@YlHrp%D*phGBoMqNuAOE*HRu7EcGslfVz8f=*|CRwT9~ADpMUF755qDt! z<+>5>wUqQMMb1#s4cxo{U2nt123NojKGpkNg?rJBUJc-ja^;uEZUgpoa*l(M4aTQ7S)0UD8P&fh_i zG8cFl(O}fccMFzOd1-;9ad`l0Rw9R|T85WZ=Yn;jY)%dc#nQ%QixhG`};OG|dmgFkZBD!fTGoFxi zOyj|{5JAQdO~BcL^mpDd#e~DAs+k$(EJ2=d+l=xVBNV{6+Py*Uc-SP)!E^vD8)c@w zj=w!4J2CmdpT8xVW$-uN|Ddae>|o}~I;A7l&Y&)1ShR;zN^mo9(pX4F|I6m6iJxiP%`krx-9l{U~Fl25@= zMFu~Oyb8FEI=N+Ry)xP63s$hwL|a0SgM{EpWo6JIk`A&> zWebi;E=_L@$#&5yG3MLJZGwc9LXdgdDO|T8+ochhV|&7TL}}h}Ub~#re1M=2BEa_` zJL`%6c7lC9>ml3uCgh_)w3t&1eOzIeNlK%J#PI}W7WAL_tnCeDQ{WzUxA%%5oj7;l z%J4%(qzRKM_A1AN$}}wgllAxU~Ax;z5t!5d#P-ow84oRj&*PA&q}jFmDqjO`%9K zEX75Il41aYR(U7rAUh#4U(rop_U{9U9b8Fg%5F?7BQ+Buzh|~%0*>aaNe*^1AmbP9 zWHCv8Wsp&3gp3?Y3PtXjKppDcKo9T}{Th7_g{_(n^qwv$iY70e*P7YceyNfhL7M|H zC;NUxU2uc%n%Oz7I_?=LO||J8qaH(LwN4I=OeJ$MDv125Ef0nft{5 z5-V;N7$?zj`qtd>85~<+!vrc%PsarrM$0bh!JAHxml%+w@jWFbH(nXT6KCW*KcuAh zDRK|CsDaPzj>lRz)dsi6P-;HxbaY;=tS5Zm9Mx^tv-6=i{hUV^u%u%xI)nyOCFPNE zUa$$!$}08Clw1ckuSfJq{*?yM)y?mV7{Lb`$_FKbuD$Ghx}L!I z-zR(`JWZNJkX|FpgG)Wmzy+qrcUH?{S?w;vPN-IxM>mL|Ddy5-hz22JrhrbLc|!pV za*lNdkX#PeLxXhg!bSz|d8w*}cO(U~Dm+gKv5$D@nyGEG zqcwYK)$~6Z+BQbkY|;tuly915^O;GIJLml|6j*4L_k_EJ+msrwy6O0`XVE~4~xHkhF{-!n=|_Q9XAA5nlDWMgTdDPvE>)D$j8iPIS1Z^SYtHHJ*K3eQly`X z-Y8iHRdaEH!~O~6X4nm2S?MHdzdhej^}1(>28ap;GLMLtO)Hn9m5B*g`&G^=RAfw3 zZDK9sVRGNeUx6vk zcY)dhttG4vig?|09Wbz};nUDg)&=8Sh}>An#t32o7QaQtCduw+vh#nhSzv&l|BAIc z$zEpgbKDn;oHLpjk5JMkiW~rX!?3OL9!ae9=})ZEJ`D3!Uqm4rQEO*NY zP%pX?Rm;u@IQ`~{x4I>*%G0n6_&_fRa8}854jgSP?FyyQAH?uD?7yciF%1+vpRfM% zg#&P%Uq4mtsSgRL?sLz0w948*KOa_KA z5SVm6j<)83Fjn!($uDltF|Dw3Sa6uBru!H0e6VnFtr~!w2_(0}Xzhq>NI^ia@Ucfp z7^YRII!M8+R7qOIZh>01n>289W^{6KnTt`sGRQYoVap^AUdBt!5HZ{=x)Xv|hr}I{ zA|5Vw`#~4)vM>z-&X{mMUP5TbAex|ghY^S-X;#q^Q({$z1(6vN)k;qtMDa&<3ow-5 z3Jgd2bTOs>Z=!6<)7Rc&SUi1Tf_^fC#e=|&@3IWu?H`WzA0aCocyA%c$dlbjNs}nD zo{H8<4ihMmQ{*Y~JXP7DdNm@wFEY`fusF45g23?E$1p<0f|QP zKPG9QufqB#hFU?o16GFb_pjrE-QhPtIOP)fx`*jPUL0pQyc<&dZorOIEl9Ou){b2a zk#ztwwniq{u*qo=wY0(802$rFkSipU8Dtzdpby&3qYThaN(u|#bUR{7#~JwHbIAJC$s+e<8?nYpylFEG>=H@6w+(-HFr_c4vJi+qSM$V z9$n(R&_0qQNGI$9~=Buq{XZRXYn zZvdg>W^sjJn4b-*+L?aF-M}_(SWUn(g;sf#w^`In7ruqs+tu@T`My|N6Hpyqafl6NB#WA!mFN7+h705(rt?pSf4kt zws@P3MvKXb>7a}{XEa?nZVdg$l|JW8vt^zS;X1JCF_YSpI=6_o6=oSsq&q1a#57^G z;-pODi@VWT6D_W0IfzFgYyu9=gZQKKSFgY9vU>W(PP~gM#czN-38?^)PAKvIJY!?0 zuiUcXx6_AzPPgWVf2?5bTyguImz@m)iJ7&^K6WLm$FpYsL+=fwMRuC^Aw48cSdc8p z5@_Ta`Sl6aEV2w6hDqyl#?sce7-56r9xv2=!Qid5ef?HG+2z2?SCd9|rRRzTtzkV7O*zCq$i~{*Y{^!6GrKiRt=RA109!!?^eSE! zT`SWn@haqtOA1;4`s+Vb@V^tX>$C5{t^4eq(CvBoA>R^@TfA!Op0JXg?tPJU1z7Rd z@pPh$fLzF5JM9%0h#z9TZDIr?CN7Y^t<~sa1|weYmE`fHsyMEnMa-nCG;kX@H^VDg zy_^BzE&fT_kkeKFAyPmWN9+xHEbRYW*VyVoG5?_DFf!kVJ??ZI)@j@g_T|`H-?~gT zOd;h)ukkiY3M>s${HgP*URYUMh9rd+WdWC#csLxt1;K z=C2GNbgh|_POt#K%d^7%#fO%EyT)&*J-*S=eKLDV9abo9Pt&yhrj$g_uVoxK9mR~! z{4!pViz+6lNQgNp-I6=N@{N^`1rSWj0&cR8#Xa7!{_8_kopfIaZUwZ;)QVjCre`Ay z&yev-&%@pfGV;x%yY52m4%bPN80U)e)>Fd#7 zASL6c6Xn@|ntOgoEF0$MpR@NB^Y&q^1@O;)_VJ*RAKmllEPNhCscHc`!PVq``19&?|~TWaIhHf;Fs@xEuBDd z`;fTE1N8)k#E>Fe<<%-n;=nE-M z#jl&b&fe--B-SeK~w13w!7h$D1O0J=zZ!KKrd zIUgsh{Z$TITF3}@lm6aw_Q1=oiBrzL#2RF4Xff>9q;Dii08?Nv=T{@ux z7E0bB+Ni)wBM# zitKdYdF-gs*zc#LdnvM;ioQ5|7rlc{mZq^gxkKW|U$`yKn6?Ruj;{h!VI6SHCrrET zin&ae=f(5R!4j(6zu7!+>RCaw(0jqy5_NEhI3F=`1WJ&;s z)v77m=xsiS==dqNvp?g)gOTYOIk6csnAw>ek0tS@HHc=)`tlTa17brHymYiCB8Jyb zvIJX1Bis78$KS`^<$HgoTJ}Jk8L9&}i=j&IZ@QJsTTPn}Jc$~GP+o4wDB#>sILKwdvKXOn3S;ww_! zY^FM~#U0`};BDg)9;=*Vy}Rk#;TvZ4dX^EC>`ImVW9bKVG@b%|)YciX?ibkyW+JnO z?TpXZ*1~G2kKLWkuActYiWR0ch)gizz}t#uDi#~PHjplQRcJ;)Kg0mF$^mkKGy=6A zUdf`j$a3lB2}ax`VKIZgx*#?ShD2m!uB%1gd2t+;(N+Q*-R=(T*NiX6S%PaP-4= zJbvphk2VVp=n((wORXf)fxTNu)E_n8S(Fqc*wR7o5sF5u>8+mn`qYj^N9V;mXNWcg zt1yElRf3eOSig#Mi6#OG2kk#ojEorLF`uAeXXF`83OMolUwjM*QLJBGPqw}?_5jZnKSnVjFhzE1NwzE zQe&n{R=M;zXM^+>ZcXo(r})`r7P1CidguDtzxFUJM+RO$6K22*-#KbKDSu_~0)&{O z@X|m@Ybml1h`=B)oC^sHAj^i;?3lZN;;@H+t7+INi_Q=ol+ag?UU4mBgaB=1ahH*9<2l)Np`FA!!4+hC7?=QNskek7WZuDN|sHy7;FKdXJC4E zG+tQc{f|$FsYWCdf7XG6b7rJ2QK_ql_ax-5JI++OibKA!a5U(e1*E$B6$vb4E9`d9 z7h*83O;jjU%Npe^+$!-lX|HFw3lcxIiDH030T~rEK!s=*i!;dv#YE$U#@Nr-MBQ&N z6@%|oXs>2Zo2fdp^)HoxMCK{lp5sb%ea;9n6G(7LeQ;v4)F(&&paO#3}{ z`!0t`=nf<#d0%CySr=onu8YvABy1M`zwQBR=WJ{9IS zwQOVX`RU+Nzn&i0CFx)-t#GMu(aMfZLk>FhtlM2m!!X_mL@*Gk>0=|E;Sz?ba25gT z*<>@GBSy&k>PP>##dITpiIH*OD2thGgH%1!s0KE_v0^fb8vb6QAMQdAE?CYv7^RC^ z=~_1b)SQ%gE5keZUE%q3IY@Lh1>)UIKM2r-*KltDUv5`;SNOwl{gb2HchbFdKe;Bs z=WJ%VY>gg95c!^&vRMAo~q$gaxy5}CImMw#2FJm4XHCRdpPCIJ#6Uy z%J=aH=7uJ5hMku4jyiYKOP4NvMN%TNI`;qZM+WSv@9>_Gvkr{CKBLJ^M@c&=a)pX6 z`mZa(7{K1&(H3=yt6dNAQb@Pg3fL&i28*=Cxouvp3+`zlr_eoa?!vC{I?%4rDjVoS z3OswDxU5L!bnpvVxx(Aw;7m)?1ozpw3m4+<)mZ^C`gQ6GuKj5&`R1vMK9;+waOuL>NLAc zmNGAwZiCl%w{IyQEsF}5Hb@iKB|&zd>PTQjmz@qA1e!RYbh+$*7vxIL&MysvT&7x| zDlgQ$x$SD8E5wY30w{{kcFqH}1AWi@-ZDKL!@uWi(8V>)5{b%Wvep>=n3c z<|VUkvKzQ(e3Lvh;%($aWTS6s*!sYP;MJUau6^QeNv9}=^8#~V!H`?tgb`>h?{?fR z8G)Ai$A9&F*~MU6-dCObGl_TLIV|7E=xwH?sTA2jMSr#_g&tBTz1D5mX=voXHqq0% z6`j38H)9UhD*=tkb5inFdl{hd_UTDyNrnSM0}`@E#gL0AX#oYbv(dM~wbm+A$R5^x zQLbXoY#j(DotnLlQz5CQHd9cvJ;H&L0aew7kb!ktR0f$`TZ1n9_dpR$9!!{fwG&p9At;?E`*nc0!y(lfs@jV(W4MCrh5bu$s?WNdy>xNZHOX06 zv}cH#1mFO3Pmv_f=J^FaDBT82B++$9{o5pH5_C(_Kqn`OQwHtdN58t5ekk8E^Dte& zUdFQzlEw$35mea(G^U-H2hr@iC;mV7z67qRGu_)$JRx~8kcD8*F{ns@f>_zY6?LMW z>2#LP+?hM;{mpdld^1zNxzkxY_okgYEAAU0D!8F6$|49Tn-(Q(MWL>ss3_tB1hf{5 zfJmvr_dH3kBofUD2{$@l{Z&rRdCv*v{pWd~_gVgbt$A$uNx(R9m4t-|vc3$^P&W7( z8Gt5n3TJ#Ko63bK(HN-eUR5s1g0P9^if1RZoY>WTSci}& z)|_!q*b_IF5i(AH<&?_1ns2;m^iwX(yE27beqk~a4@|t8+Z5YLk#;J2m7tHGD@JhT zQAqrLR~E-U3cVltb{^Fz+u)h#`)8Cs8gaotZ9N;vOEg&7rR`~d9?}unnzuK#swoy|TaQBAh%H*UpZBT!x+GVe74~pZ_sjze zuS+(NYTj~oGe6xuRa^lvKV&=UrPJL%QC7)uOqy#WNSbWqf5hNX6n)boxw`*3lSCWs%)obiPm<+My!TEsu_>D=b|Xb%N5e8X5)IM@Y?;>T zW{~&%-|&U41OV6hoA7wc02t5FDf;%xbuSq~v2soDF(5Z~T2dvl+XNIFDK?fO>!|4c z9?6R|m^&(#@0NmmCwM87VF8|)DQgh%#7&cMf*T_Cd^m8}oVv-E~Yom)VhuM=es4U_R%`wBS?FnQJo~m#0|2`ywjpbf+lr|LbA(-xKo6 zIhT0xL8pWh6Ks9fv|57|i~i)Z9zKEVoUlSQH{>RmN9ZgBy#^yYecFWRv-LoiJcvGr z4!LiN8sguFSn2^j{zL}v)4;`#8QB4HH(<4UzW)c|_a${u;nb`yl|LYL^qQ$%;m3XM zOYGUbYVv7KXkPRaX7V_@z2G)4PHVy+D1H#M)VSP4ym!8hXr0&xX)#%DPE+g&iquij zEmACtNm0h~R|RB+Rr_d)f=@`-$y%g`rJ5Lz7>{C73{s6c8jFFk-1jQ4Y$4W)qq<4C zaDYB9&f#Gmpi)#I(k%B#6?fC8csE0@f+$;)vO~Fh4lZCix`WZq?vK=9rC+zGbHVAT z4&_zRRZ%Kb3GR-{(u+T~NU^8Ew$*OpK(}a~KEGfC{5b*rr9b=K&iR57PjB8nmrRau zi_#gRA5te?y}~ z4sX>|!{clDHNw?X>!4f)MS*3m;bj1Q8xQnj)tBMsMFsv`{ZkD%_)DK*7LDI z;ydt)ZgvuLG_a7bZ$>%H|3rj!ec3b?`C{hU=2^4pQ{^z#Ha*|gn>bmSDSO@&dzyS`jX%$>Ki zi&*gBv8NzWT((f(N7(>*Bk*`;t>vI>y!jb*4_HIx6cg*D;EvPCU<|jwpP&BCW z-Lm<`k=495VZU^}I){G}SmvsEcS4)h`#qo@dSOLKvl_?VB-hAJ>2aTUpjEiza0*15 zKVbWRrY2A1^!;^S$FJe&jhi!!F0KD>wQrH_PP|1rWa6amqu5M}?4hC`_$5IqFhw=Q zuV&ABcgf2Z#<4f~Y2ucU&Uw{rfoN0I2=pFx$@fT)1+|LO=iHJVf$k-|dO-dlFhz+a z#SV2;Y{oE{3lq@U`2{x&KkV4>FJCl5W!08%|BzHVF;rSjpmLUCfm`cCDmqhz-PqXU zkqaGg$6a;OF4+Dy%78~x80!Wb4k75n1&#O@(`URg`FYc0-Oj;k0|X(wnjnL-dzH6I za2T3Uu&)1rpj6zsK-0ua6=OeK(c;X-*>kl5Y_Q9wYvo$cOM%$#qUn<#TA0Kj;qZ`W zmGqKG)4`NMJq0$VR0(rmwgGz+L6n6xXq?mB;l>G3Q({WWzWcnF?$$%UZT|FPRX1Jk zlLu?3PL4s}{z#pCqwlBCjF3uyHUV`T?lZut72Af#+k6WrXizg|t(s{rQ(%GLzeSo9 zT&`-7cF7Z5o59>)io6tw?jr6#kv-wQ`!eufw6o96?w-6FYOY^Un?F{LH*%T<=dE4* z)5mX~nqypxR=zVcmZVN4z_~G|fcgN%=21Wd6+IZa*S|v4OE}KQtZZ_C! zD`tr+V(enCt@*vU>@%I$(LGr5>o=cw_U9tSpnJ1Aea`Lq#oot#4n^gAo#ZtLa}-*) zLs5gGE#7#XqRdgOgvw0N#DSh!P^R{{7`{g37|4!AkB#sh`v`{|b250;KaR0KGOt>B zva-o}?IN*I$)b;>mWvvS+Q7Io_O%ojR)#&~Ffa zEa~G{l7qnS(aFz~BHxClXl5x>BfccL;jf$BBa8zEAc)!@^y;5oCCHvzs5m3hT=eQv zRn9r~>a8WCV|PmP_z=UB_d?IviZ?!lB7^?OjczDkKJ2nf`A{4}dQ5J;bJ0;Amm1IeBY0W9+>0U+NNhdmCgD5*NP4Y*>nI?5Qb>{p1-&(o!!fR>&SF>Yj!q*PHdO@AUT;lc8JAxX7 z>11bs;f`W@GeO!21Zx%0KyM~DqPDm$=WTJV_TERYm|Nza?w%$&t-NEe>TN5H*EpLt zOQ-B%y14RmY78V37TBO%d zLRWcm{TuyETL&QffP%|YX6^pR@#j72e?vIrzK#L?KS`JHopbKmkRG?K?tdJ6wn_cq z$NOJB{`Iq8Jn+NaOS_h!RT&+aPdZ2yjqPD*P4F4bYP2|`*X!vj`6Ee&dk$z>BNtHj z{71mDmoL$V4ePM^Q? z+UN~OAI$inarL?KqvFp<(+jiufNJ2F*mMWQ-lWJiOh+OQJ@P`gdA=X~Y4`<^F8m1W zx-p-qIVEagdR$ZKp+%hw>H>~I>EzllPzJ;=Eo(1(3*Wiu{c31O3}R`Bp(T-6jOSp(w3eif2_265kN3 zZn|&rS#MO;)|}@b3)<_wd!_~@zEbG{`9AuTD32E}ZVJ&PxITm$;LE%-GtWmI;8#7P z^M%W|#Y)>`{jjHem#S~g7n%3ya|!1-@!rjX_U&jb)g7-~k0vGLt1+OXpY0~)IZ>_{ zr6G|ixH}ZtWky^^`CiZRvE&M?uz=Z7X5u}(t$-b{*`D#jLT*gkk@r`R_vlG|JfH?M;g=r7PLP75? zD*7@>rE~nTwjo71DAmm`54hla$+wQy%|FLBt1;?=b%M2m{JBj~J5#hUANUzC-{jB< zH%yal({A4>9`9+KoG^;_yEk3@j360!Q?{Dyabl2EnSi8-VxfONkBTmpw6G_mQvG^` z=)*jcG`rV(KM?ndhh0u9@!GP%p~FE;z#7Wf)b zkvH`bFr9v;(oP`Se*0 z0(0A8`r7=l6MoIfkW3MD-d*~Hqnu*PC~}C3-a)kP{nBn;ndj$6O=?U5orRLh zV!B-sD=72nkT)s~DWDQb6VwNz#C5T>Rnzp0&)#^M3pA{IqJ2iazB=xnxxBs8R>D~jtjCoCWv6t)SV-Mn6=j5< zo{r*GK&IUq(!lSgm#I(9?g5Rd`;Kj39y=pk<{J&c*n1s$YpY z>D)xvq=MGf5P$sw{2TT4?B-%xKU>ye!usdgYa};JeDgv6m(5kCUXaBEv}DHYe0ESQ zSi&t-^jfmns}usIl{4ZzlbB@R9VA^`M&=C!PX(k=wMiU ziv2HS6qodwO9@#}>}imfFDwB!ZqqA8!5e)m75Sc;CRt5TT38Y@dg&CuT$Jybqj=&% z5pSg|mClRQD+l^I?3>K#Q~unWdwgX7&)4uoPTyDOt;0!K-d?}N7!>*DvF{%wrEFkV8?|m`~^3_INjY zp*G~*G>1(3 zoO)~K?s+OBoK8gleG4h#miBetFA8Ze0bdQpR#HF+26d4Dk?)s%Eyisvg~yFv34%db zr*;Nl3HFDiSE7j*w<}WVELwxxW0V8!lN^PlAd_4nNrF zo_S`r(ZHzU)>V?-PP|?nH8C0a6brO4*;F)|h;G>^K!ej!tp3~(U}4p903J9k zNK;O~H&A1q9DXut@5IZFg=R1;#>aLr*iswF!BEZKCdXrfQ)CrPCiot<+-?~re>(ls zkbkadJ@!vj1|Tr(sHTcpc;G&O#$BJ5FW6JqmN+MMI=RdoLJ5BTxC zYM+TZ7*@=iRqz?dxLEP%2@UzLX2pVkt_lKdFV#1&8LE{u2#O>Is~RWWtXK<*3C*w# zH_tg8{rBg+77g1ITckCCS^*TRZt#TN#VO_II}pg`;hsa^Xq>|vBpIq|UK~3aD~q*r zZe|N?m_=^4SYAiQi`CzbD!Dijv*E;!hy_*_bVPLW+z>F~(U-nl`M$Yu)RR7_6X$m=h)L;Y9}CjWFPhmVtoQCvCeOSl=u&+~ zvgk+N!}4k#3U(XLHAiaQE<(lhJt$Y~;3YyX_=M_=q#88f)92JV)M>RE5oT6J|8m0f zZutS0jmo<}e@i-RG#`v&4 zWoV4E-c0%F`DNbb;SdW^hb!!|>7CHKfb28zyo;oe(L#URM@+x8GPqybwBV5d&vr0p zm&AiA&ShSgtZGTStXYj`2Sg7gb@czd_t$vdpKd6wNxI}&{Juy>OwEL+#X(cUXvcQ~N%Dei7(RLb*)~yj}C0e9I?t|_Pa?IJK(#>oPWbfW97w&?x)|f@_ z6^pL&@C-7GYEWGcsku8D*b`lxS3V8!DU&!hzrsiXIbQUjMRYyYxLFWK;znW z1T6zXQSsw6v={_OQ6jTLuz4FjY2ay;ACemsRc~lLIXNPjM+~B_34nIw%s*}IWqfu5KlmbV2(*1$CVtls_$V(5ouU8iZADnw| zZi=$WD=#pev@ntb#|yi*Q^hE7-2@3($9V=%`BGZ*#ed4}4!gwZi|@SN zfX{b-`b%@lP^VqFEO42kFmilIuMoa^G1Kr*9%cGxa)E0NVvXBnhoN)42XjK~l+TJT z{E`?q%|$M0r^y;_rLoQnc>>y&V_G)0QEVbbHc`>p^atKo=8m|uE6&kr0&M^33NMj# z!D^K2G34G2iZ1^jM~JKM{OW^mzhOkiUviGzCiUa#FgWr2cbd$9GsRw@$T=$d&%-WP zc_sekpkwxzVV8mVWebt5Gl}W&%k(P30WxWqF;~Ux!^bzIxR4A}ROJe_4aix)wFWpg;n0Ddi}WJq*wL+ptR|PYY`A zC>^xH6LcAb7Z#?9|32)3#Shqye1$(OU+eadVV6#(Mt~*NxgOR)k%J+6R`|~P>*HA- zvvZaWQR>p2ZV^Uk{rvobd{XGd>BO@pFsi0lVBk7RMcbJJ`j^oJz*lc)krHmcjPoW)Vci<* z%hXaV5LH)D(f61*_8f=_HHG)mxZ6GeRKut`->mMHof9pmO2rj{>mi?x+4cQVIlOJIrfX@`78a6Rw1JVm)WB9kta#6y%I4Y;pr`S@8MY9#k)6O4%WVEar~B5qvCt^E!tI8{fZg~-I^jOL(W_>079VVa z+ga}%{#O20|Cr!p-x9~tlaGSQ7VM7R>&Tznu={FEBkN;?$RqaTI+F6jEKcPnfGVKa z{S?^;tZ;r0NCR81>|jvH9rrZv2iGfCO~oEmJhOUg6Es1unwmj>EXLRQZjbn@r(RI! z@Q#y*K_wEyXuP^F{Al3nsS{joEJ48*czwRq;E|;#Pwf0%_P4jq3C_7VxXx=Wi3K|J zfv;VcG^rCAt=k7s5qLqJx1^n^0*@v|Su*dqPt2@lbq7S^cLv;}duI0Y^CI_0HK~_- z-k6>iyg%xK8mQ9cX~FwIU~waG?(Yby<);Pb1TW)VaD>?E#6ja2;3WU%?BC2;HeL|W zSRf0*wy|M(Hv=y3;!^*_MX8ePNaU7m4@;9YtNTLNGF4>QC2>(v6#hIC*{J{7xuD;# zilhdlD(l3CdmLn4&<0ElEEBULPrLit+u!)Td0U8ueu^AknxKt_?2t+y7Ryj|fwDyt zwEjBqW3%w9Tlr~PxZ823Ik^MQ8-4ioD(|!3Hu@=Re_wu|oOWVY_Ku0?(Mqwv#d3*? zJ{NVt8@OYc(VtB~nSyMz-GcXhVj&g1)gAd~QGzfj44m4KJg^wazLx?!yeH-Pp9CBJqBrqd0wGI5V0&2_WKG7l6#wC+n}CGbsslZm~u z>f6Ez_*B5fjlSk_JEzS=T8PA*kYD9tfEFusGgPST^}k^VVREd2gjI8B=Ruqx@zRui zKYz=dXp)Q5;=IO`SfHgu!3BLc>u5W3iI1jUGC(SzR1=Z_SZvay)UzBzh}SynFyR@r z21};aWWsN0)dSqutYQ`!7bO1`8}^g;Uzm;aMU&;Lo??$tqz1La2Hl(01#}q@r04J( zA*ZxM)!^A9Z&nYya!J%BA0Tz&HWD+dlw9WZDDZp}$XDKy4L}#;IU2_%hYj)X24vE$ zqB_v=z9vo%+xBwC;yCu*tT}y=g<@^U5S<*>BR@yyOOC@)S`g5Y$x+|3ND&J=_T_>V zizb3P?ocuX)>B_u1)A{(oqBhnua^-@->u$#iEMFV-=@$69D6AiT=U%!rUR!Uo3|pk zNdI$<~&pV$2R$?Nd6EG*B)0v_y7A=RqBIqg7g~zeYN^STF!HCPTpv$5{ zUIxjgnlXNGxBR~i^CUsRvu@L{?Perd+ z@0r~pA97#e@lG+F2E3CUKv-KP{Ns9c8r#C+m2~k6k0y1$G&!tOzkBMJHV9MwE(xv# zmnd7<%j(Z>N|Su9>oJjaWBl2%2{TVFA>%)M0%tj4hKl^=a8ZyEGezBn%SrZlw^~M1+4>`*nMUO37KdIZ$1eDET01+R-yPnpt`S~^WMl4Eu1gwZkNoY9B^zM# zIe^9-na%oyUvh&@`Y-F>SC8&?+rLz@>BTP|K$6oht&K-=CyLp zie_S=`fZh<3kn{G?SwZII&d z1~v|zcWK??my90F$~D2qfK+z85RB8VnUG%|8;{#av9T0cM@1VdQBW&cFU(LH_q?>7NNTaNgcY!uY!r&4# z`H-;&FW@sf9w27#@Y}auQH2}rgvfQtUUJZhH^^s9Y(f>q0@>*ih|vqPXMr+Ya98kQ zW)~1L9}PSsX%JoJwFy&a7l|)JNTpewP3yutm?N_Wr6Zw3{CKEU>7U&Y(jBM^-|KnJ zzgc}=+TfWHcGf#4c&}%(`X7{KSkoDZE;+)v9Kz49UQfMc|Zlg#d6}=^- zEi@Z^T1~D;C;y&%airdRO{KN7v*{jJog7MCmV2Nx>o|6H^LgbACG!V3==DqPBPYxo z;hlD8W1+P@U(&9q=WPXgxb4!xP@VL=sFruj{RCK*Q_2KE-4d*58zd0;k<`vbji?Gy zHLV z*c)JnwpPx-c!qr$y*yF8aI-PR+kbHA2Syw7ozv9oQZY!@30p!8dNbJz`}mu}c1cr|n*%-p6IKO9 z*m`FQdWv>UubhKS@mRr`yl5r+lD!SkAYEV60C(%Y7W>SDI3~T82d2R|xd1dR&T3(Y^AHP|>YU-hd z1Eh=J>9%@m64URu6$-!i(J8(S(k$Op=uqp8(gChkL}dBK3wlWhD5&J{R!em9F8Lnm zuIaUcTz>8ReyF;$8?J1Gk44*I3slCt=MQt*R{X&TpSA1a7m{u#-f?YsX{-bMh+>B* z@_>rYC&Tg+@aON>C7RlAJbZ0uKn4v2xAbsCqaTKO%N7=ZLQyS=bps(^8b@r3EcX5g z3MreXU04XrbVZ63p&0gM47hh0$kbZiSY|!~EizK-^nLmq1fOuFnSGys0Dw zu;%e}Qha*6*l=g8SMhYb3wL?@=+Z!>VQk{vkYJ1;E$oEc@GGH&Y*AENq~_?YsE11@+Q3L4IeGwI7b^n$OFgoUNB;9Qe{v;-V|H1iPK<2vYrt!GuzEBs#f zEPCBS+*RS#TyZt8*GD(|EY$4wLyex@mY2t--A?Pw$9K8>4SIP zPWO?X|M>KG#&zcXpL)v39d7$GCyqz2Gx5_tqu4=;+($pHn?@a`n!w5+XtEV0`f3Ul zXC&B!4PH)yAQg(w+kx@7o7S`aViDOPkf7ZZRY~r-SI*HpXO$$&6%N-;8ziTcJEyPp zY7edPYZWCiLrj4P15F0%V1406t~k~U735CHE9iQ9J6jX@5Q0FOO_3|+UKXKzO9elP zDJ5G%x?vYE;u05SP|48X?}nOXUEwLJq9w~bv~GqA7&*`5Z+jVIR2>jB*I+zpjaLTH ziQ@0Gdrl=mw&4+%m|)%P64#vzA+B!Cm&oE?&Vd*HxG?tN^ImbfO(pFLDf-dEUnO zTN1m}-X56UfbeVs$!P>9YfAZU*dF(e(P*VTsGCaeI&oAW*2D}AQ|xCH8Kk0@vx%;| zWQnBHy+2~x%k_ae;oZ<;5OTu$zY0;hdmmFdBhGUlT{-7|s3u-~j8`9shpmJ0XeMG75_%~t7ium{V)KAyxZ_ec|Kj{C$547}zWcrn3w zVOqAEZUx3Yq;>3;9S?7o@1yZHO{=I)hT>=ap zh~?lV)YFY2C^HBSP5Ht@ZYlg-(;td^-B(OD(=?zP@ZV&a$SSiZ4N<91BksA1|x8ACf4$BYl z(I+lc+;fLrkc0-=*9`-wD19Wf@ugT>kO!L2yD7edE|YUdb~{d4=G) z#tMq~JEh0Og>((%yKeHe{>yk9pzW-Wc_hH6wh7Vwy)HRCFII5lHN(?v@NqZ*Y683I zr$J}@Yn%W@{i5?b#KQ=vcZ#R|iNrcFpmI$BwTohRU|2pHvu`cZbu!q6qoh}(ELDsN zvo_cjWvX)gD?}Gq2)#7GPNz-S=70ppxbwgr9^*d7?k73HEE;E-;A(GqlVrfXPnq-d~7n&*C@7yBF$8Ey+p^<5zOFYjm4^K1yaIg$S12Jc>=?dTAnW6*@xON5d z;Z)JZM4x-d0>L({hE5b?ULJ)Du>t_8Eex6ZV$oF|NGgkqBYR0X#Dg?FuE}8yP$iuS zzg>g8STy3&B1J-sJnESXmDP}$z@jwXtu-jb)ibni7i5>@Ltbkg1MQ0sYv$2P3|aFiyNy< z(+py!po$AiQU}vBeKTYQuKONV^+{Sn)-lNtwvP{KnVzCTOO?deDwCKKejs}4TQv8O zcctL0cUIVG(&aX)oZ2W$SM@5;Q$nFseAjsKW$8^(Cxhi^=!9p}DAP~`)rOlrA4>W_ z3o(aRr!R}|h7z=DpGV%=b050rdSIp#Ti>;AyM0fn(s+4LPnQjp$j$H@;O{=3;wV<5DB)BHlaSMzl`6h=EfyrqJv(#X_lb9qz3RNYRYynHq`_H5Z|$ zB13h@>%{cU0my<+;@DjAA@7SJIw|-salR{pv1TxfuEc%{%%|hkCgnMBWl)2mg8^;> zU^hMn+$|6@#xm6e$W)=Ht%>!*Qp%#abx;CZqFCXYCn%155Sk>=feUndK4z#m8@9

    8LAaGW>V8o?N0&>7Z4E~yf?#HBBx5JBK1si8Jy}pr;Bkl$?^7K(pES=RjZ3$QDxwKQ$jCShQf42UO ztmC$^ciwLn$u=>!J17>aM7B`TS4CGv29h{5qh;ay1EHloQB@&oht{7&)$TbOLu7gK zmNAYdmaC<-dAwwim%lk@pap6^LzT6@M-S47VOUj}#OP?$4a6^Yp#n7D_6Ht$W!U8d z?{n@4-PX+GlD_74@v(bP2$2z(%`y&#>$x$K-E(@XT$}M4H?f4rV+?nviNGL7CPTAav9rnW25DgxA%C&W4ov_TV@%Cc4{gEYM%@hvggEyG`Kd%GU)S)&#%}v$2LuXb<)hE6K~^N*mN5w zXi)!s?u{!`jRr=odq+c-ji){4#5mbWGnb&`%XpRvRE8m$JOH#+fg!8)Pu;`2pIv$|dJPL%mqfwOmoCr#) z3?UXB-3Jo786-YrnV@DitQ&@P<(eu%&=0-2HM5~lMHR@6w(^(uiNX{ zA8}af5JXvrm{Hqe>8D-6qJNlW#3(-_T`oq%2(f+7}g=L<2ZlCKVIxljvd=;zkG1i7S92tf; z_x56z zd~=i-JNVu7FPg)Hi}`Tk6~#hB3hEbvZ-Pz={6GS@4M5}vP70LXVYw#efI#+zREi3q z9%_={#1%}A3mKcj^ySwJyrYZ`$F^*F8ad>|aginy7vea@LMXNhTiU9my4f8}1)a!W z<`K&t1olqcuiWCri!smEB`gAB#*Mz8de;gL1`fLLxMevnDJK#T^=B zSDB|Bn@%;?SC+lh@o4rne?R~7TSg!)sazCBZoV*{>xc>L?osSrirk^1tE9_3vgiZ? z+nsy#<2DrJ@ZyFU^oY4oY(STB6v6e(o3Lt{L; zMcE+uo90>sB!=&uR;Ib$gCb!F09J&Q@|-sR{FOW6E9Q;bFUZ;k(f%>JrL7dZg(3+C z#)m|44%8LY1ZoAE1}|Os(ZE8mXSp6#s$*W0wwxJl2?x`~)%r|m{BIVXy-AG5sK_Pl zG+8rV;Khku8&K~aGyU5rHjyHm4AYO5DtdWlY)wOd2HDTBqb*^E=D4eFQtNj)%>QIL zGoeBg5i{-6BmDr-I-3cCn~lUMfyK}R z6wkvn-`GEsv@v=NI;-LR_>Rdr_*B!&PB-GB<&{hCkVGfmAwen1m~|tAVxes@g^E55 zL?4fvVGG5lywk-e$wpt4-PQ@WFKkwqOY$Wf#WBzbJH+pjRfzk5&kh^WHiz7Ecq60L z6K}rNaq#?evTM|ZH~qTIeNm@P4O{Ro4|xxHCJ3-jpcuIAfiNsZm7&u68b<<>nB%Ti zG{D<)PPhQWp6~mtXK@0?OE<}Du1Fw#plxv^ zd@c3%q^=e(1TdzQ1-eU?7#Egr9{c`5QtreH%N3J_r(365CvOL))a+U9PyyMj#;e0FI-z#XunUe)bwf3d^{X2MTrp9<&owasqmm0%y1yaM^8FWohs@o}LOUUT>R56Z64ZBn^j<*?tT;OEB z23hm^S@y)zbxwAM`s(Z7j*}SSB>%^kn#jiS_PoyPe}k~bSl=a;Vs}ww2Sj&49OVPP zPFfa)o$xmyxRn9);2jLU+UnUrR|tFPym@aOZw#aMjw)1$=lf4lIWWt!07&9 zXs>&&7~9qKQV2-aTLRgGk-+m&Z@--;#wT0~Z*{8)G?=xg`B=6?$(jZ1S!13+(Bl)9 zjZtg=H;$PTdUG*GP8_@~Fo9nh#e(3=PAd9w;C4Iz#QYMvL9`{LMcO4#fC6i*d@^v1 zX)=I#s*{h>r@&}-+;uV}nr#Hn<4JbZ!C`Isj}=Bc)bz7MZ<8J4*&!!hl@6N#A%|k2 zBzrFveM7o}!T$}FF2$k-`ANN4b9dAuS(;=}((0YhuOjh~#qLq)0)RGw`Isq@*LuhE zs(prCvQ(Y!ef;tWo84V2rprb|80>@<&$o4mWpna>{(bOcbMMz_!!H&*T^*!oD;K~X zI|&+^vC9+%4K=5O*DpX8DLpMFo=H(Q0=F&*J)q3PWMuqoM1!r9Z1sJX(J)Tp;iah; z?i7B_Xib7v&HDj4?!?>e4in39kz&tNqydBnL-WH22^JpYFqw3&7>k=S>2;v{G2()< zB#5AkGY`(tOD5q0(@mEPD`;fo)g%b8+8{%9$Qv&p{iTjBW6S8Zt|`j9-UlK7A1^NT z*YwKHxNQzg@r?`9>{9m7J5=?goOfT2+LiA^cOo)%w$bQG;ZvZ~%iF0FCiaTmNp>k1VdJJo5{q70P#uc$wVbmSuEe zune2ntR9N(rpQNB^h)>JLR4kPpZYay&8!CbL7pZ<)k-RZ3Pjnw48Iyu6=Xb(T)#+A zQz0tWH?(eiylXPz8$%!{(l_ut?kYZt*z8@$y!P{;V#-LP+ zyv?(bp9Nd73Ob$#;*ZQu_LdAoGbzedLB|4yT(&~MiIB&9y6kbI11 z{|SHA2XAHT^s^sh_YrPGUZv)$$4jsws>6QOCrr4@BB)d*PUt3)Z(;=u)w`?K%M}p?d1Ww*-6Z9MXQ@8Yfkr^)xPBs9YAQA4BbV& z{CoWKK3ca-?jN960D@a)J$y*vSp#{RyL+x;$8&dv#zV4S3 z*-ksL3u0kAU8EQ!rJ{WS2jmT2b&3Lhv${X>guGY21Dx7@bg5_{V%TMv+=G6lqyB|V zmI|ETsKTfUevy3GZ*L?zz8igG<>wr!tu!HMQFg$Pg-29pr2PArjsEI`*T3>JQscxy zy&EPT=sAi#OOaDlbO$r!`H;zzCV~Zt=OwttdZqEo1F_hA8LuXAdqBOkD;%`f=w`X$ z7~6y57qtPO@`ofXtW&*aD&Cs}iOE7BgI(rU01_u?a&l)?3(iM$@gNVo#@05P%fy6I4%Hg1q zz*XKUN*toWmw5W*7ixT$DK$5NEh>pY$qH;G!{TzxU=Ifsgxl9|vLncz0vYa?uJ1hd z?GU3aT9?#&kmNhDEdnvfF(ndOiakP+!&G!pz&d8NS49X^g+q}}Vn zF9Y+7p_CYGR7egF@2R8f?U=f?R-fp$8>4J$8sOI&AJfEiMLG_hz#3=nu#vy zKsy$|#$zrx9_Mz4ow96NzPGD7Bh=`!r2A%UB?V6GvYaw;St=;@C`H~kRCV5mR)T71 zNx>@cJV@X5G3P`ld29 zr!;SlF1et#hY_~o-()((nor*Jj>&n&>63R}qZ_a#f8R`_ozN`~ zy-t$3l^-~-6HdxZtjd0h-A9p3Dth3RBqn9Sb~c@C7U(N~4XP*0J;1&gD4~Y^OCZC% z$G1$Zn|*gihccaLj{C$5u+#|+kOn1E@B|hNJLWrK4d}->A>$o7w{nUFKpX<^6?W$sB1utR1wwu-`#zw^q^pD#vr_{S1Sr{xiVvyu zm4$a@siEtbEV`V(muv@uh7up_8yVH*ef#YeX+MdZdl7;!Z@<0KcSxb>V?O2OijT-P zgf@GxW3Vh38JO(>5jz0lNs!n9BAl$?6qm@?>&(T$bmZ?$ic?Jvv{ zbjSol`zRKOul7*UX<^7&2#zflvgP^Y`QVCD0e&qWzlYSAO%OWFqKoMQAPO>^-a0o$ zmFrO%^u9#Xu0W0I!w%#fvl>`W8WlXl1uLA!-uvS}{$sWgDyq13m1H+Js5q}5Mj*5@ z#@ysnEO6~+W3gnb=txv1lt^MJZl-EjzA_kDx{n5~0?n$E$O|a!JtqymhJ8gNT<;^Mm5{M$lGC!yu6-)%vOea?>Z}PPX?%;s{j8Z?;ju01?mv(Mf9 zrET~ye~SHIbHc|<~xAFx4P#`I^C0)8+1POyB=859gsoajVSSHQhywHi){OD>`yL#xmn#2QR;JxWcY2N z)+}xOO4=JJUxH8SWX?%@>D%*fgb&Qi=WmHj5Dd(#P;QBYJ=D@qmaL9I=Cf;2gODj& z2LbKBw=;Q5?#qv;&e^<^vK1g^c(MG_MkH~vI@BI_o%i$JJA(c3P0B2~5~6WfJu}I~ znUCkm3Pd~}JDCTZ5HZD78*$6kxNoZam#V*$l}@~G+GDbFPNdjP6p6>|q27BnP$zWw z4S8u$O|gx> zFMo+0p~yojI^~r+!42{%X_ac9XMNxxsh8h}vXz5C*Zk*p<^xy{n$@^4mB{ahc8k*3 z+w%`9+Jvj7rn&C++6T%HI{CJ!)l*NBf%*IC2DU-IB4EYTv#O?9Ily)?n^~znYWptPLw6a08650Vu{S9y2CWN;a=K5BMhUM3p@}SjW_!_x}DnvlHNxfJ3 zvqxWsTHdJ{i__gN1;z?iPfZ0Xnr7%Zh!>BZ+XMt^bx0HfNsMgzm{+&#ydp(eAxaBt zR$pgUhjlP{{QdOpc_~W6o8rIpQ>oL6if{rUb@9Ky@BO?d5whu4C}wGn9Co<}#roS= z4Jv50EBYgH9NIBIt~IgNG>v<+15a{7#gq%1XMW9`L6J*YloONNS>#rP9H#1obo+um zzN@@T8O@q7NLBN(6i?HnOca+bL?Re$3dstqqAS3?P8EOPkK<;?V_$BL|l zq-DQ(0#}{_=~i{{_sxmLUl5R4;HNi)cRF}AK}ZPPBJE&$X0>^0lEZF>>t|FzlnEmrc*{$m0!3i&c;yV*ZNV&G8|1f+40RDAD%;KS@x>Ymyy=zYk5Tb8A)~ zfKH(@7?Z>dG3ijzStH1XvzMXp7tgiQHyu>XX%z%4yBm1LAz0wF?oqA(h&t_O1kAvj zvejgd69cBo1TaMudw?Q&*v6Q`D-CQ?Vtc&S4U3|HkQ8GkNet3Azwe4Q;f+2|1~_C>#pS9hS~K}J+saS=DW8#JE^+(hoz_kL z0sET=jM0v4+Z1$}Y;|HgQet99vM3fRqtbvPUf3wq(S3{hB9X-*mAHMs)$%ThCxG;JelZXnA&IXlMbmr#s?vhvV~A0D=wcKvk3BWxQ$H}qnwNb zvBLo<<9Mcy!wxqCa{2OCf9+*-H@;iF`x4pm!o*U6rgTg{;$Dgcl9t_6bdFaM@2t3( zet%{nNt<2f-lOP|_xnA9L{kd?Ea;VYhYq{owGzo;UWYt~UnSo_s(IDC&IJwxlMxO! zV#;w~GCukL+&q6|BDokRj<{RM#bog}3AVjlADE%S9{DW(D%j=x`H_E@ygOhumADj7 zI35mT=9LRF#@y!c!_*b0zBLi%btkSHvB12Z6t-nrtJ^J7>YwJJ6UGYkm21dJip68i z%0j4PMUH>X=U0Ryk@xW(yz5iaE-4mTZR-ynTvV7I5y4f%O1 zd5dh~=7u=28G%&I7&j!HVu7zLnTpP)cTJC*hYjjjdZy>mzyqwOM_p&V@*T!eRs_5s zyi(RkX9`l-$_2Y7VK-!>^%y&y4msX-`Wrd_TuaTJi1GlfTd{hr2eQ^y1|=~^X07tp z^owc&w@icHj<7@C$ZMe45|T-$s8Z?k3gk=IkI7WwUa1XAm2)JHY;3m?ET6AB4uQ&) zDPJj&pA{PoO1g0CO%ngY7?eB{gOWnA$rRa6MWd``rmC4u2F0LtATIO)ze8EDBo@f^ zZvy8*qs*X0gmINTI6FKUR0lY>+yUYsM3_BxwpYXMH;e%JOU{wor2d5gNT&%vnkn`I zMb1&thTa8WEWR6B49yUayl>Cfx(&-Sp$Y)G^)&~kHwsa47s}68vIXHuOxm1XuKNR_ zO#;eEbaWk!yzV)wboV~U31UH}LBKZ`@P?Yw!!B5Y66d=j80j{OBU_}|vsTLV{De1| z{>X}u46@O)z!$l1)9oTT>Bog`jytqIH#;^RnL)_X%b>qc z3twb}mj8+k`^o!L$wiYnsi)Xu6se)2V=zpE#<(G*8wCF=l) zk)|lINAE#YFKH7(BVD-W#PkM;dLe1WhznlAd5wvvlIKEEdLsRuQtr8D_PT%y<}iZoKusq~sKO^WZDS!uJg>GeF#8ZUHp^aZg= zpksC)NELv@r-Qi~wC!d5U#~ik!kXE0V}ff1N5B$+9(_06C-0Nj1K~#VQvHZNc|33D zbj_KWpYbs+tjST;O5(-6G%`(xcUosu|pjN3E6HxB=TSg>Tz_Uf!07q_#YrXs)y_}$YNpa-)QM=kFKkJRbKKmEHS)2>5;IQ~0 zJ72gw_Z!q#jNmx_i>n`#T5jNQ;&^YH2{@W4wvi%dsOTa0>ykKj$K18NCS^UyJsbw! zLG7IVQAx}Ix#xa;USDK3kg$}@yE$jQx?kF)-uC+P|Fh*yq-|;9?NQ;qIC>xbZVnay zukT*_k3&mVP`g~$2CeYuAO+Ju3|@(+tG>8`GQ7+Vto2?=?Wf*_meSyZs=Swhjal3T zdONsSYX$ug*CZbc)A?P(cwW7{4(c_ZX=lbFfGeylm=LQ@kA2@~9si**=fLxxt!|_{ z=kE19G_7EEn{c0hxu{9K&wrW61$7H^M$!<%X_}t(aanQQ>Y29US@s;l$^D`}d-Z2O zF(*W@Kwb<@y!4|`C?J9;87Q@2IU@G-)=Q2FCi}CudTVRX57wGzar*F`w=MJ7H@_7X zZd`;!u1ofkgD=b?bjD;6s-jpeMUGI>r-kt|mXo2-RZ~ZgR!@aee#jKWiwo#1Up$-T z+riwDGq67T+1Vh`FlOVg2?{=ia7>e`vSa<`5W!T?+?^&Bzvch(<|-7yP9efYfwzF z2Pu*dA>ZKad6{8l)CaGmi#y1DSs~M;-pbFOS3^NyGpK!T*hif(!* zPe&Kbh1`LTen4_0skC->t~fyeZADDN+_Hswo)i?I>Gwl|cX# zeqD_V?raJwpaB(#AZ{QkDys^(1BHqrD#)f*ECf^v72Y{XC`sgME+kIMdd08IUBAHn z{`Y+6JKs6~|DIsX_}t)m!;iyQXD7Vb07<;Wb}X))7501U0+8p~RGb()EJY#6UD+&K z>jC|hFY0l91e}GCsQZ{c0K}HVF3)~AbfN7Jb<8U{B4Erm9C47{;{3_&NS1XC*=ypx zuvDC78M9$xy)zgs3NWA;I4+IhV3anq3zjj9=+aQ2gKKBbdvD`sg|B9+NtRpMOw($m zgR#UBK%-2BBVhBa&6JJW`+sX5n1bSF{}ZX^HZVD{qtk9NFkPaQpHQTZis^X=3rLnK zyWBT{q|C!P2D*UXOJ@nwX5{fR!ZX4z&Ih_umvrTFC@}%rN^$|nO2NIE+3Y{$GICXy zqBT^fMh?s6;R&jTbB0_@?Lku3+`WoNQMeG>#|*jT@v2Fz_aoZiQOj!+uKv3l(r!qu z!1H$nYu!d#!^K-ui+C4DAdVkcqcCRE<}=XPvbA{b>fcQyB;&-R7E4H`3%uP6{B?6O zp!TFU0ER5|d6JZz@XXz`4>mtgj+E+kRxQQyU%eD zvi>G+`@bHLS4{*$PHdZ4Ae2Tfs_KGvdz!?qc1a;xt=OWzA;qMdrs#BDub>Njp@w&B z{td~W9+UN?SFllb0aT$ogz4@kad|K zXbvoBR!$H+rY>~^gt&oa?5l(0L;h&3K>wO}K`eUpCT}L??HaH^?Yue#CP3)vB*7gx zg1LR=Gg?B+=K)<9q<_{ba7+8}%?=;TXULK?8fjf(=MH-GxKhCd;dt3}%VohMyL9d#1%tYk;5+UVx4?rWD(!|6kTE<1jEUYtk;~_kK{nM{V|> zqtE>O7}@5;91P_a9`SxkxrYK2#N^Ve=PwayF*8%AJ}hl^Mcs&2X?KXWB4Cd(`m$s? zCWd23r;WyKZ7VRH;x5Ks^2vs&aG!gV0y90eE8Uh0@+E@~NhKc#3ZB`6(OY)u`)7@? zPs7KE4NrRG@?JN0GiFM@`F9Ou?G%z{(KQ*Aax(>gEYR)T5+CBPbJHRR!aATa#n%yL zvZ^B09e3?Uk0ubSD<^t7{v{;i&wka3Z*lH~iGQU{EpY1ur_JM>5%O-wmz`~{OS97} z(XEHB0H0wsgB6chy;LQ<#v4+cm*RR-o#MV=3Gb<@DY`;Y$!mig%w3{~WQF(|Z>wjr zB$Lhz(Wy&Ccf|cOD|vetq`5ZG7)*A+Si>33qs_<{+cp7|(?y)nO!;dywboN5r)7*{ znfx^JlICyZBVTD_^uC{EQGJVwzjZNMr!I^t3`aU31D))Dd|s7D-@+vCjeKNZACRq4 zeg-mU$&%$EBgZjrtWzhsmTHa#cR|J-N~IkNZ1ztGO_3DMYZD%xp%2ul8=>A{IIO}x zYg6`t&=*)TZg9d1%f;Lv)bZ(`{{3Igqtd?~-SRDR)rmcs0gK`C4yEj(NIO(YP9Nk0 zfBH7honBAF_Ve->1FS8fpvUfrt_WC9$e~w|Tsnzpb72G5oWPaI(O}V2GIKb@z$8bd zPSwU~xEnn|6!v zBDht0pn%t6ms3uADf4 z!=e*g?v?1bV(u2%PJTMC(5o4wtQObH7q|QF1Y1=>Uwe5C!01Oh;cX6BnecaUX@DIo z8n=JEZ(WP&wBr~{btNj3=ZlPaG&mt@mgPx7_M?x%s;WkH1CT$&Pm5h}goZdS#!zZ_ z1G4uV;MiDq2nR5XvH;et*ngiL3}elmuHE>T&qU@?>iwdrePan6`yd zrlI{a)?)V0EewEB3+wT&F07yvK~e7(eK#Z{99GD6>g3QAO%)_k=;`gzuK-900q$V3 zKXArsaq3EX+9WePG&>Ici9B_2+KC=eJB-fDjHi@qD6*1@$)+&>p_&XSU>!SO^eSw_ z7?0A19v#Jip=Zxk6oiRE#vc`klv0Jk5QBKIEX8G3gr<+HZ@9+B6Z^#xWo+cl$7-0%1 zC4`xCsTizUL?H+R6oCxDvhmg#`}n1h3H68+^Y+u)RHiqoL#9(#M>HvhT^_4deqA#jxLBGy`~5DXTjB(l$tAG^XRX;)Uz7Di7KH$; zvok4aYO@#S&?iWlu$+vD1?c?hWa-n|)h7r_R&^+Dfu`btnb`~kWVU*h@w@0-%wfl_ zC{6@cqhVn0?Q7dR`(b_LcAN0rTIQA#K=N~dDB&1h#ZQ7+MM$fb|j zMdus0kg^0dz!UwdNe{UfkmY_S4EO&0;a|4Xt-J)D<42!ycDlX$___8t*%IpVB#*yJ z%@9%i_3Mo!ajeo4C(bz7Z2_B1O1Xt1X;cj6?SC4WJPpb0;%1$e;oyS&b zi}C@;f=lu=%-OF<7SzzV-8io)+7ML(lr3$fg06tfxmZz=cp1-hMPH=j?G5zgcubD1 z0GP;iDi3j$HL)rSvGKBR+>o{c{oCq*A(xvWgZvY)T3f&`4Q-r&{DYQgOl)4yQr5g4muFjMyeqad&%3h zJut10-xb~Y&L`jAGi`}$Tu=_g_V0Mx+02Z`pjgg_M)oFRKAyk#Ncn>YbIh=Lzcjj% z6gaWtdfEaA$0#M}c^`yzW0=x5%aH0Wi+3-q1?UQKe`r>}`>;zr?;=c8ifGKd#tI@V zL`!87{mxCRoS7j4uDo_8i&wzQq>CYYt|e4w3#|or;kscO? z>^5PxA0X@t3{K{7Qs(z~yR3N^oOZ>Jg-M}SiprR%F0NhYmKBcvtDRr|=U=y9F;0k>Og$(eerAJGF#MB)B+H2nN|}X0*+(g1j+0NtWCfK0z2?!t(uGwsTOa`2 zq%NP4I`zHx(0BlUm%sq!Kv&=jV@cvpk1|L}!o&HZ0;+)eTv6(SrSzsKO1>mtWT5li z@QnRYWea3R%)<*l?T#RBkcs@(9b(P*?X*UOMJxi$2b8PVVXbqcx?PPE-xcb6@_1zz zT}%%6LjI8HY!QL-uhnwQ7y|AVOl>-^c_M8CZmYErj71_{yNxi`y*hEQShr}OeQsd! ziue9dX|^x!j~`woo4EP8&bxR;iY=^>o>FF0WE&Mz6^IL)7;){2K$hV08D~hk&`9%( z9F}K@P7|+0s^uObk}4xt>guTjz<{(YWE1b8N*~~01;&Zs%037hyOn=Woxg7)#HUUi zien)@#T-Wz2N+T$M-33|YWby5B+LavbIE(3@0RJiywNTlfn_244@G150p~ypOEY#q zd@f$zWK9p{v||nn;Zut^#czj)o<1bgmd|()wgWP!(`F8bLEcHMy{E4)ef~BZftR*q z-wQYmE|ZeQvrb!^43>8dOCpM;8KR{i_zrouso9E%g5jM~$7_}mJh*HAp@O~;LJs=ZAi{hrVGKT`& zrGp+9{(alqE2bbHQ!ky!?1t0=1Je{8H>E!^(XWB7b{{!cA-u(`;Wf}YlE|cKc6(L| zilv_mJ7=E^(#ybIQFkb|`{;e}Zu`C)HXETA%!SOS>}rM145s^9=gYICp!Z2j)fm6p z=JVdN;43iqye~Z<-W~B?8^4#N_^on#wfTfx9;jWP*h?CFO)Scx!~>J8EsE1RAuKim z#e1=eCYvsWe~sCc6|x~kVepF39M5e5paFw37o>kHfrVo2wi#_gG%I~f7fgYTB|OQI zW%N0Z0_wf@`Xf6OeS8CbiL3}c;0sgfdz#}A$3?uFZ(WDgoF&vPrVy5Mkcer}*G?U8 zOFwV}hb=AI=NeAnpe`)zsBkkoDj#Vt{gJGi0(pa@NBGT@GL<6hsh9z>ilF%5(tr*h z-~teJ@|Tj>1v>S0{|t5P0uxu?D}4Uk7T)diY;9|WQx7AVYzVi%lMr%zisH%ahl(V% z=wNlj3=PMeROXaxhmlVQb@j@GJ#+!J3VxQ6?R0{m3-+KS)t;{aHyiVkA;>oH+H@5= z!dIU7iDDvoNlpwWmb@gBhk|LZouJEV45VRzHDPuFEDs(H1Sp;JfU3Y!pS_CnZ|>(+ zyP6d8>x`)r)x8M`q%9zFc^lQDl|9&(z$b32zDaNHvhb{Q3ei!seI zW91eK9Ip3x0yH^B8bC~O)Tz^X_uh$J&=HX@>84{BK+Ff7TFgDT`3{tRr!w2WX%zQG z%3_pJ#9NZ5p&N@lOK^FK^e)pTe0HML?V>V4l0#$FxlUaiI3Sw5Z4cJ(b#SiAKUtn z3yy4S(LPsw@ArSKux4Fy+6fN}9aOy>lT3ybhh5QXWQ8Bq)T6$qQ0JGXiUo01bU=`o zZI!*|x;8v|F8yt{xvnk07nj$5(vvygYqmxQmj;~JZm?i!r^hv(ab>TS-1WeqN{loF=XrBuvyW`}>${A$HzdMKz%wchil-xkmE8T)*Pf(pIQy}k8o zDSjOh@j#h`Sp-Iwqj=9`L7sXjXpnzSrPHAIa$L1D7qnq(VZuzG~mX5Mz?< zx&`~9_RddcZu;recpZ?ZV8IZYu!+sCCyF8qNLlTSA48KS{ozllebxyET#SwrqlhIP zp;?yX+AP-A3wvB)4&sWsP>P=xKT3U$^V@j_`2|&@{NbE5O_yT(L}tXc^p4Fiu`&#f zc3fJnCpB9e52u~&u$Tj=_JsAlRpbVcQDZ_HIH&qxuCNglrW&biy4y?J;2J7(*$ep7o(SwTTDEZU0;qktywRM6)gj8$eQpr@he=3w-p{E z!}3JBoXfqO;NiSsgB8DQUf^#YNM0uu)T*~{NMj*)q$wIjw0FHUk!=w%CP)lJbmJ*r**`r-L^5_xGsiF}|lhtwscODdz2JgXz{nlxUA z&*!QK5%vy)aE15?)Svt0f?pf1Wrz4XyY3ZP2U(rgDP;+pBAx4LpmFY?wfjJ)DLL{n zjXZ(4(u*0~d7Aj(OT@I8`^pAI&PcGK7wvHgI}*lPx+}dC-DvHHaEW&~agqRwkA2JQ z=ClX$RUr6POg4mG2pR;A;Ihy>sl6?Q4Si`FgluTY&X;`R>;0~`d7$Z77fi&8>*U=Hx6YK{b8G;kPX7qbLy!b)$fdDCgMZFHNk1(YQoNR7LtZjj;=37cx9 z%*8tWgDNa^N)lkzMV>T4;JE50+aB|C7W&x%J3ro$<~Y5n&O4O}>}tIHWtYLYz;9P? z_bcTi$!t6%3zda-Veku+(yjsLY>MB0=Co#4V1aBKSucQ)NGhY7cffa7ppk+$0cd%y z@Q{K0>O!3wr38juuz}rf8@*$Jz!4iZY6(>Yb-mdwEe$;^jTIrctsNjgVQfD`xy9B9 zM#uZf3EPu?{@#j@zwCSz&9W{bz9)BzRy z+_WYB+7!P5f;mW7EWR4x5`jT{Tn%pLA=$uE(Mr)>(B;t{fn-TLCq`Rz)3|YB9(*hs zv*S)k&F|+3%s2^7`%x`fF&2xb6Qcn*B}Td58!6=miX>4nS>Y|>_;8~`Vj6KqtO%~j z*-dEJ=8oXd{WiC0j|7S0CgeI7 zMYc0ds{4`JdKe=*=zDVHH!@YcCrGrVZv8$*z766Xe3YdI2&M2K$=RqdYrLsQoE$YA z(ab=Fmqe#-C0jkTy~5qTpw*Q_SI7(#S8?M277N$(WjhlE3_DoH3ZO`jBwn^=LHbga z=CCj$A*haOssfQ@4YCqw7m~4GWK-R^+<&3h=s(?hXUaNXLyAKqL1% z>ht?k@mV%)4tL!U>E6HkX& z3P6{;-xg|jP^Ib<(st&uU#pjP(6dKaFW31M25b8i7gRX^)ix`)%kB#bz3{>vWUJ?( z)R9wSV~0)vB6bAC^G&fst2p#s~4_RK~<@1 zhYZ;iio~^w)T#Eg8P9ix^$`w`vA)%z`=5P%{tCG{lydsJ-@8uMPa*p(oWJdqGLs@( zs2E%@X_h?aaX4X$UTE_ix+t$t8+nwkgoJ`#0e_r;B|8O-AH#L7#@iAHM#1s{< zL^H}Zq)^IaiX;FbWAqv2F6q(0y^16OjCD|GhqNT9uhlcx4mkkaI>a_u;ET?!-HBe! z4HgG)bf2@PoM7Q3PoISw>*!xKDa`W%HXk9rauy;v++y=P%b(i${gT3%3E+el)1t8-ruIYXRUYg znmDK|>L_~OEV{rwm!Ic5NN(Bqb-!K*Gt zUn8cR`gQz0;*oorq8oX*|G6TIcQUBl5gAV71lVY>Ib`rUn0&hGyJo}k=e#4gN$nKU zY4QFxQOb)HX`o{EdtK%Cx_3c6LN7U{DfQ8*YZciuC)TvAJ=o-We132_%*T+K<( z9MiP8NtFy-tQ7(zqACN)E1W7H0!irDkbRzbbjYO==EjW9m2Z}zXh#sVO zlXg<4xGz9;D6*M~!EpUuMz6U(4{1A1S=Xq1vP)DiZxgPVo1!UN zln6}HChn1ZN5n$LdiLDAe5_XNe~go{n-u+tB_rvw@2LXqs*e{93XOqoVakIJ z1Bb<4eikoNk`)fQvXIdPlO{O2Wm^2~s=%exCy>}?npi<$VFsvUEER#!TMk_&9#Z5> z4ytm)wSCN>N1O1JtAXhX!%G{1{|LxMO`8UKq&3s`923GXfWO5-Riiowz^BJbsB~Ve z2nE2JL!U?+)e!$N{=$V!VTdu;0aWEYkbbE+@M`d|OCv9pfoyM`dQZf|$fcq@DJ+>h z=BH?yW!+x?J`R+*P5PG~O+*ULiA@%Z6x?3L#rcEao!5t=uu!HX zN2>CYk>==<*I&0~7Y6Zdx417e+D8*yY^>%?$n z|AB6##X|f{ zyEcJzN8p?d{T(AE7n%TLw76bwO73>lXg0C_H`swRaLAO1t&h3Dh^-Ar9RJ-KX8IJfs}tSv zzLqR;Vpj(;Cq_ktlPP5aMb=U=CnJw)kR7UjE_8S!Wa_swgPt|KSK7a(zOk*B9np_3 zZj+m_2>Gb{!!P@?Vky$`WpF=0gm&?1tiX0J_eaYhoWg7CE|M=h9 z|N8mwfBDDXN>@ z9dV!64(d%5hf|gL75_0`gnL3!1_c8hYW#UfvGVx{J-H?p%ilCL@I`zrO4B)QQmIZEC zW^0aW%IBd242?FapSX90)yp@iP!797SOc|e4@nPw6H1-;qCopD(RP@>*UK^aUz;3> zx^!)Hbs(}YRPu~hpu>y=#d;c2EKDtwiEja!Lkd6`zOzY<3NYzB^s6fdZ zygo;>ldhMeR;F+3KY=-D}j8PLY(%2Xz<`jn6&57D+V+ciz9<#P0Ea_K9wK?ZrWb9TV@))lCsb~ex6i|iCn8h zOGJ504xP;ZEW$wR{ETkV$Sns|T@gA>0kB?XN`@3ht$NH|%nD!4V7=rlyWF6f( z&!|y`%clvUTf∋|mkojI%MuHk%KP1Dl1vxEf_ms{Wc}YOzqT-+a4k(fb!vRf1y4 z9|pn_f>usR3p%`@F>F}8YRdn&J|+mw{6Hb~G20kT{OWSD)rsd+M=h+zZb}JD+j=VI zazu+;zNBhq3+Tn;HaH~6Xxc_+2K0iW7%Vkvw2AJW%7ZY6k5gTj4!bm|52`+wq5Xi_ z%yE9HPYP)j=S!MpPgS?)aGF`${vz7LF4+EUwqMCU{*1BcIO#9fvdX_&i;lMxIv|CD zMS58;8CFi-c(a*Yc*8JN#s$*YJ>Nr7M*$xzrS9&bkWcadcv=8!&rM+}%=xx?&hurAfV0v|;M7O9y@AO>Lc`NL(V_?Wv8O z+abI%EeF=5t9|#w+T{oZG>R(6D>2gwv<)So^?ou4Q=hT~1Ejzx{*IC4$KVQK{1%hF zAO^g6XTTDHL2d|JytphdBYbfKeJ^0xrBn&@XK-;kFTqF?UqP>doOkVJ(kzP=>C~5b z2|>dy*ur_wHon&MeeH_Y8J|Pp0~XWZsbQC9*_SZI>UNX+oItZD9Gv;rWRlBmw(Z2> znG+U%^Fc}poS6GyEn?;&U{>BeedMz-dK=j<$q4U^{`_a<@7!K=X>s4;2aC>sFY)b* z(W%Tf@xFfVGo*s7kGw2zw6_7Y-N3zU z8f{-wcSdve8xsMJ6B{!Y;9&i22UO)}(q~~Jbu0*$5IXrwM8|>*xPnF{Z1nq{bq zIUESq=(>M~TKffcYPKdTTnCAz7|hYrRe?t|Pdv4UWG8$!%kGAma@S4dn|5=m><1hZ zuL`*y#UMqQPQ5riAd_AqT3kaHi`VLsZ44DuLWSD5j4vQ51(ek9n`{J1Y-_qczYbn%zg z%2!T1?W(lELlLEfHSXP1OtXAhNI6i`HAQEG%UT;)5d^iKLHYiJ`~*QevxF+6hg_;4 z)?ZCNE(52;lq`|rml0kIf&Elw4ao>^iY|y8s#E+hp(0mvdk$V_cf5=rA(mEQ z+)!Y-ksTpp#Qdl&2bQIUn=Omt|7f?8y-pk>K4*cNa!PrO0?nNmBoS&8o);HIc8NQL zU5ez;)dE;J@+{=(dF3*b2XqF%-2u%ny?!^wC2%5tpP;GGr~EX?m1gztC1WsJ^a*5pUPyUWX=K}iDR z_1c@#0(#J+BIxSELB4i{s71WZ{m4w6I>An}-4}GE%@FxwbTDnaph>%&;slXN*^_&! ze`XG)h(sISC6}Ezl=8qLl+r~h+bPltY$4Lzw`xg>Ch4tbTQ&64xw&s8y_Fld;jNRh zTJ?sv20>rZbXC4TvU-5d1myVV`8P$Il4|Zv?^jim4n-&b5Py(9=^Y1qt4Nbu9HiK! zL*V`sWf>5LRH`ziLoR8mqwbHq&q4}Lv8vs>DY`~-9}+otcwl%xQ?kN+i{Gd2_X37o zmNWO?%nYcZ55vUh8c<-hk~B>$_2I|4{6P@6ZzXlEHFSOGxrn5io)CTz=tRI*96h&g0E0D_Q{L)k-B_sJFs4+p7 z87Qp{JTHw6IUy+E4>HU|p>t|~sNBWwY{GtYF?jSk*qfBoki-uYW=XgKYt!UBV%ff;^f!hV>xbcQ_$Yl@Cn zZk9duzvF(FsV0YJcgP;emc7av0dDIx4(@r2=GkE|#xO#C`oTX;|D)MzNM!#UMS8fU zA3guZayAv2V)uvws8FTc(I!!N~|Bxt%o z*+BSuk0;W6(LqJZjH?Tc7)1TPhjX67)OpzDqYr!K>9P{_9`OKy5(WejDpX?&J85Ku zLEab;LdS_VbS=GSR$&P6LcsMSqO&3c6lx-K1agzSC=z2~IxKg%jdRaNJlnS{zt7e^ zb^~`___Y((b8M#_=UC>`Fb|S`^3JANwSgxidr7e*-90WyizSiGG9+T%;h_&c<$p!2 zJq<*xaY6e6&#Ln@CwaqjyCNJ|n69;fUkxI6nJ- z&qrJ5vpDT&!jhqq7E~uJcp8W?83L z7GR{e#2|mE+s<(GNsv+#CMhUufwYzmhR(Cjf-O!svbfW+J3pQL!?)ed^WTzh{#^rE z%Pj=s#39{0i;K|EuLMWc6!<4!s3B(kLa2n;S;Zk?)x<16b%mmgbx~ z%k`|0ITcBFO@X>sH>-gj<%SKrLx8nGL+rM1f3@m4kVyA)De`0>f%1}RS+424q_-1! zH>6pD8^T0zUQf#EfCY7>AYZb@wZYRspM~s%tQk$wH>H~-ZNg0uxYemU`AZ4jbBK>j zQh1#M#vyK?v2-TBUA8@NFdg;_pA)yjU2>}37 z2Xyd@dcgM1e^D#;KgaH~x(L?ut>-bXiGRv6>p@DN4WVs7hyfJraX_0Kuhi2Q=zPhs z;MfJGC6@vkOFnRU<@OvPJnVoB{T|n2>I0D5;^+a$b_6{$7UnJc!)8>Azy4j)Z-dPK zNTFf>Qlg&%w6ddxU=C5r0~Fa$#h^-*PJI-XtxygISD{QXNn@ZApZ+l*K}S2@B5o6= zg3JTzf@qO1c7)pmX-$CL6i3OJik7y_LEyS>^Ic;wIZ0Uj9j z)F!ma{?d*y(RAK<;{BQ1W-c@%r|~Dtb4Up{Pu7Vme=b|R0H-KrHARk7F_*kQWqKI{ z(-d75wHC4$E|6++IWXOAo1_PpjQN-{E|G-LSkc4CTFDcp)Gcpedqh+86*AzipWdS=bt@A>yCAou z2Nq_m7`Q|RJy%Xhiq_t*gvEO@ftTWG-?C$&W-P5v?cGOqkF{p(#E3a#ftXTCd4wW| zsF><++>jP5x&?@-02Px8I!+Yl**&*TSS6^9Y7JW=Nc4*fPFRpE$&?fYRzk#Xt7n#5 zQ}p6J(~h}j0?FMn(OwZ=vPYaFs#UZy`A}|JV1 zmS1C69yh!8#;B(`(M|T=_)-mHiP9hBTTi+xETB_JDS@)00F>~8FbAY58XGzLdaYxySw*Ls|Xwfuo;P0?+_Ei<=3)mJAU`ScI*^CkM= zle|{V4*niG*UwI`?dT4*4`@cWXrFCPpmAP%e`dv-9RX&0Q}fRH5|Zu2MQK$QXg)wG zAys7$6_dxiE^UEmN)Pb9#VfDzw#_K!?T1vPI>mhf2z$0OiEd}e4v%3M{6X;}tO!E$ zm?belEZ20r132T+aX<2=X@9@leh{y*|76NOc($VxZ+;y(;d%0f;cxg@2Y;VmcXnRb zgLJ^?G4+cO@BSa$kS?Ln&4E=$NfPgJ;6^^C^VSQO`X%ssyz=>_ZgmqN5ae!qIN^nR z^W(mQ9WV5+idtOECZ*u>-~5@xI-2WI|UIvNReJD>Y2M#1i>U+ zaWvj?)ac|70Q0FS5yrGuSpqkYe~_{bT^4g=Zad?|7dyK4m3-c4ZCqZHXfTT@dPvH| zYdmokx?Wx;&zGQzo&)^sm-mz{V0n2f_P=7F{TG+5jSCk@IB}W+O9FozGh2PPHpW|4}-z)Xk`(w!wSJd}Ap*hOS2z49e#XyA+A9@t~18u_CP1Xjh}u zFv=|E`477cV2uaxMkVkneGB-TW<8vv&Er**SW$gwtx%`F<@K0XFPxZXYzvaO*%^zg z#|<&x-GYbKOHE(4l({xxi?~XV>-V|pgrG}NG9RYwbc4@;tcp|$P^n}mjgzA!a+%ja zH-s(`;uji?hN(G?w|hXt5utpqv@cJ^HH#* zf?gJ^O=WIJ?C~jysso{rc3z8{v0NUh!}QZXi>QzdDGn%3sE|hPWY9*za`_?PpSvLi)@XVS>7Q8%O!U>R*La&|4ZG~{Tob*T`IOP(hUr&Sm)N!IQ8iL=OUC!S=jv?zG|oKildNFNnbFaP)!Qv?M4=mVRC(QiWy+K%bD z0e4mXbAWQ85#rXE*t-=tm;3nJW}u5~T#FrK(uH?b?P~l#UxI4p^{(JGr^`->4yxLO z%jMbu60gLfNo=W;SsKvcgI8q35p$Nida0gJzKuF1a!n9hSF9Bo&J{deyM%Tdi zw79+Zz};!l#L4_k`Xubg@!y-V_s;H<-sBFq^-||u2_~y7kUB^yA5o;Aia~-r(_9ao ztsG5`mll1n(_b5O?-Q<>XDq zj(n5Rmn;k1u5?`5|PlpoK z3PpwDy6<&Jnt5dtDz4*+UA!9`8sTICoOj0b)5oiZt*L*V)`Mpu^Btu7Rky^4_}erm z1eM^*)XAUvr!pu`fM>EqS;{AZjr{T%Jy2ZPCaih;h7^T!&d}uog64 zew57-Cr8zg{o&PQ3Gb=O{wU`PXBJ1u27q$AgcF>p22u7-Yl#A!(!4eD?jL-$?z|z}vbX-}}bJ=%(lsq}cx!z0~7Dr0K@S z=m!dv3miFi*Ymg#uS;bfDEgRZ-?V0GCb&#^PeJ4^NGGch?gby_GkGt)jC6rqU~b5L z?CY4tbeh{Fd9?JA@!-DSJd-YPrS-G8S{h;qc3_#|dN<@6!dh88gT>B?nz0 z8($mG8ZvK3EwtrQ${iHR0?(RW&15jC%p(N^F-_^VOGUlH1azTs=>vsLAr={`&DNy& zJyLW)F7&EdI#~h_R!DT3L73G$)P_wARxbm`_|O0J2Y*mmd%>?s%$voTH&Gom%g&Q> z@g8ycRGpV`c7qvi<@2=LG)Fa<%WaY{c&W7*|AV){C{7% zm~8%%C7~X)Fcv#0B~jwLJS42V%9jpN_|b%Lz6;kOUm)=s4Q z+j*yggr)j>#E|qxtppk}W1L2(-W`EeAQ&ugd;yJBM{~i9)vXTQ{oU`Y>a97eSg4`S zYu34)*BHW#BX%mkRIOd-ws*lD_f$|ZPZIP=QTQHghd$jGypm(AI5}^`wEK)wG;DrU|JeHDN#qGPwqNJ{%SgJ#AheEB zuBONeDyEgJQMB`b?u9+?VHtI`N>vk{8+%S_e}N zq&2)R*aJo0$osn`{HZ{D(RG{TSkSScyPlw%<%s-uwvMv}0nh9iTPyZI&hBc|uML`Q z%~i$fmyBTqoftJNELOL?)-5o=(zZ5sE-qMNDqu-OFD!%OP+0AnqbZ~JJJun!>R(Qv zvAWlhyN5PcEEzJ7A}v3d@^|F1i_=oBku-~OVJ)RxMUglvW{;x5vpgu1{(SCx@BMdV z3S zg9%3`e{jL7@HSz+d@nx>O8WYMpShiZC1b2hujF+ZFUEOEo$H3slhT2()-Wuh&7`jY zmu{oFM!He{nHWt2b+tvX9!3a6h|9PnKeo#fi-en4EL=!MR{8%aTRQCVq|C+Uqd z7QJUHQY>zI1v~6WS=IixfAeM6Ick_+CyWj0m(>aNvRvT(&3Jc(d*k%Hkj0CO#ff}K ziVMU&oqBM=u8AGg$+sGhO6geE0AkagC+OSI&uiNVtv+H7~+KLJX0$|LhIiX_GPoLbHKFz%J z(){M-_esi^Z0V)YVs?{FDFHeE+M3|dqErks5q&GhC zt@K5aO-Mg7s*P`w+*4K1m%Phov`CxyjSF&(;$GnIHXX~9GzC}F>3#)~m%)22_D|&FxkSH%ZkSEI-|M_|8n|qB{AzP_KZ05JXg(m(2)N9Av}qtuk|en2SFF*txwX03+PyYDZMMs}ql2v_`yS_n zj>+{u*!-Q>F~g*a>;$nCphbe(wQAo4K~iubaDlXeV(=l!*1!fDwZu;bp|bP^k>0J6 zmoLc-=mm6TftC7Pen3$r+!f_eAaRG!Gn-*wz*x4%4j<9we=GW$d7z5h@a|7Y`4rM( zF&A7V(j& zROZ$qEa*Gn+s8nXAlAu0)36(M(bHMsBMk1lMAvy{icg&y8{86}3=BZ^a_uu31}I^q zvm{t<1L9}MP?zXuyX7BwBopbRXWeDP*2ovy_{rW$V|Wx<@Jj4mFlcT=f z`(0~Tusr><@7w?)?m}J#T_@bP5T#^*yOh_rAeaB}m9EDcAD*pG-;qyxthPBledne6 zs`>}0nlz6;E&(n}N!)9bQx7ThqjKs~Ddl>KBw|iIa#rC=Bc|QK;JI0L7s$lRLQ$lk z&h-^r7hBPAxsEA9D>wJZNf~t1C&CL z?&hqx7$}=kA&b73{#3fmf3>{C1BF~2?UmSwlV^5fWXr*D;e-?F>0f`{INfYHzQ5~N zzad+kc&q_Qkx`bTfKtNxO)kt&B1(N~=x)A&Kge_mo1zZ|;{Ilt!F$Errs!(T(Lg-D zV(yU3M(_SN@iQAJQDUd#)R?Kx6YQ_y02m2h{;L>ydUY&KC8<$!P z=MDD|5JwpWEndA)-imk}#60bGqV;gHVbKdc)Kz!QGII@SWjD{Z9rVKU$@4TF`#p8ksqJMOH}dyJODNR^5A? z9yflh7ze8_`qthXZke(F^r0CmSN|byF6n=5u(IBQKyvwH<&w!{h>A&@(Jw2Wy>8B$ zS->>_G`1ZPE2l#cmidgn3sI7$=nh4R?`JAph*}=LpPv~}9nqvH_J4L)E71od4^l4> zNM7b0U$lqL41{p)tWuv6a#*_dhiBg|_`3dE`@g>YZFHuwj0h`K>lA~&S>dD4rZPv3 z_pX?`mRU+3f;H=qx3I;47nIoS;|~~o^QM2iEzK9U9r?}8ECcbfh+Nw zb42)kP0`k8H%SV;@UP48!~3^P>y{=7y7=29jS*d9r=^no`kzx*Ph@4tiA@&E%21*5 zF};_6R=iqXtjs1D6|RVC;$fo0-UVkJo47KD4YAogjd8Cd5B>fBP26D}w|l-a$cY^r zmRKGt+nRK3jk&nZvJM3l2-V96J#L0q7+>%vrAd}Vr!I=9^hHtoW*Jg?54+?ViIXdN zW&BkBakzHaC0>X>j+kfJjEj+(_*kC~00{dOYtlM?#`nzw+B?Vleo9U|@$9?X!kxKF zDM2K*k%~cW*Hq?ultF$Bs+i$_KO_g<4-z$vYD}Sob!6n!&X**F*2^HcA-^47&(qUS+~M|;2vjI3kaY$R0Z+qc;YA{q(*xr0M=9Ci_7PMhYmUFqx~QoLG%LG#jdBmfF?I*=La$5 zwzdBGYZ7i|k#R@D7RNUD|joZ%suI5C7xL- z>W|bSKM0E1oKRE+UK3vv0~il8pLh)^lB434jp|$)udD>P0PVY^)W=?foZW%;1ym-s z@vk$hb+p-9J^V40?k#mycqdW+Ur)v~oHR|QpNGfwFv_sYbVs0>I@oxAX0)uki8?=FvBg^O;Qax{t z^0Pn#T^@uJdWEuPhijJwUlCXG`XfILEaPLs`RHMZ zYp>e|yKxUyTz{Qs*Z)ZQ3yB#>+k_iiNdh+@IdQNUGW|x)n{<>igCd)$m^NXdmw`zJ z&4vzPmjbA!0`uV~TfSA^5_%BQGgbr?N_Ox|1xFQe-g=)No9m;WJ<@S)aR_3+WB6tAg4Uahmh9R|Q>EZIhskv_V|$J|NxwZE$S{629 zPyDhsl0Dq)i4#w^p_q4+Jt?D34yfuyRzsY?X!Vl4FeOWn^wMzx1KkpO zd(Hs`UUW4GNgq(B2qr=}IpD;_a z6F$tz(l~Fq^!??l+GdW-i~dEvf^6qzX`I)Fg1bA~(iBok5JxMZVjAYB@(1VM3abG% zkVO8c%#gwW3!e$ROJs>)kbhTD?U~2RrH={k1$0JdNh`?dhz8j$`6V*sa)N(RZrXYv z7&-RPwM&EdW(GLm4VWnQP4;BMP=Bt0Kh^y_-+GzAX>06R0{^#_*$Zzly>u0DMxZ9} z7QfONl{^D){ack02Nk7`bDf@hn#KdRclFr!JL)JWGe^x+clm$WX%9`DWwD}y$U*+H zfWskBwWb5NWIuCS;(xMQvI9rug# zhh4DDus*cZ2c5^eo*jx!P^%09`C92ofk`V_i}4JsX-<%sS}G>DXY14t=M-rmgM~gq zpAHANaS-$byA%l$)5+b_ilDjjd0WCUVxgx|O5Ri$jqEce^B)saV^~PvEJFuXr$&h6 zhF1k?3ndRDv2~sLv?k4UEptc?IS(e@ngq#%u*YFoTn(kx@JqKd>jYY-O@iar&rVy7 z=23Oh0hb&)PK9(+?M&5o*1f$C$`UTg@Yu4D z^WJf)G~fX5rB4Ih(Guzwb9~VQ1*WR6oYE_J3P0)YJ3MwvJKa~z-51y&SroWx%E_Q~ z8Hxg5gS641S<5}p4eDg7;E1u7hUn)TymR?0=2nxv^NUotH8JaFt10_p4rKergzOt9 zh*OEj{@^;rjHKv}_qAjRH+R{IC+3+JxJ#y#2^3jN#jJBn;BDrW1a6Hi4>}A+s{_>1 z^F`PO&-Q)Y0q?B7IP&kP(2oC$M z6JVon{k4EbYqkjX2=Q z2R|xl|E(Dze@h?;(QB4U%yn6$-0znD3 z7>TK)7zEcsNROY58wOa}J_}2t`rs5lOw_>z?qL@!i9lLUgZzR@PbWygi9Z&w(W8#; zo~y;kR=sB~hz0k`w*kF*GZ2C1NVSJUwgg`HPma=dCaoW%CoB67SE6xVU0mf#mCfS`FzxNJq`-#~!UHw_n zW76ovu5qu$aMVsITPboK7qv$fE^PNJnSaT?2sBOP5qy>TYLZ346~rP+5beU zxs8C%8$c)R7NERDDLZEqP8*75nPTCTs!vf!R*8%R#Onj}G+G;uJ!ISSdhu1Z_413`aKu?o#%9ui zDL1yvG8-EgiRKW=apExz^0G#)hZj>y&{)}v;hqZ3Nlz3L)~R#oT&UKqfc2_lfmg)q zJRgx_|57(RcMig%*Mg9BydeCEbU9FQRn5F2#`E2By$lOl@G7ikdDdDBknPwTTDIZF zhOx-b;NgTDDk}EvUp+MoCM+(!_+2vS0;x`;yJrohTuG7TR7@tVcR%I2I|9@^NB?K< z=lY_@!ReX5s3AL_$n2iXN!-&XdQ%JnUWOtZSKB36xR7@UB50NY8 zy!5>Epymwe<8PiHt4QD>XLqhNR#Yl@7^%%BC%_ljsyVOe3QOSiDJrARhznu$qARSM zS2!PMqZnH=?WbrO)hQaCM@5jS;bE61b&I%5)5lMT`m;X81yzb)YiLHeiI!%y{1Wof zr!x3^oPxki1AX6p*yR+IXBuTf)8(cgD3!b;PWQ+JYqCeYW28P4R#^y6jLTuMq)9y> z+cGVFHdMw%G^)EJEaYtK9NKB;yqpl|ys?#(?Q52O+l;{K zUtRl@oN(e{tj)qsHd4xZikzon1|n-DxDHY2Tk5uj8ibkHDbfoGizuN}OO~ii=LSGq zFay#Z@#b@sIUmz;tW3|QPeH=isd@TfW%=I)|oDC!xG$n>vsj}_rD{B9IzFB5;R+Tl^>UderhpOeM4VX$MngsVa3gm5F&tTDhTfam!&-ieY--8D~?`p{rCUy;dj4U zgvP}T#KJsl6{T*)Ov!xQYL@jzHL8tk`DLL=g8S~JN@U{Eo>N!Idg%;xm*e@pu!qJ6 zkr(u>olbq{J6C4;n1?G({OWSD^|cA{K#KCHIofVYnNJZtERFK>yimmuWi26~ml@Iq zdB(?tO(Z@X?8Sm%mjeE&d8te)zZ9l$SUi&>O5o*&<88REa9w&G)>9!a8>Dmv~ix@thpVYyF)sGo44S!IX$x!uC>+`HLpo6FiT-j6QHzNmh=X! zWu`L8k$L{71WTbJ+*ry7kzlta;oZC@^?om`T*u_MZU`N9!CX3-{%5nnA2 zHi2|UAaT+`73u(KF(#J{-Ujk1jl}3O>8Gk;m*u<`HzcoV2U0%B(rH)U^)&L_U@}9t zMjyO?Y8+hU7|=|}w3v*{#B2!5-6@?viTG+!u%T0eJfTRPU+3OAuP$^r%rSk=^2kRH z+X9J^YhUH()FX%9e7VrN6MLiwE#f^pDP;~two@_qnx``O`U3={n8l=1ALqC8x)j@C zNmb`@InY?rg(*^|l&Jy=e;U<#SgI_C%wha-BvNbv#4~@xYRiELVUO*sJC^?!>uJrG zEl;XUJp4EH5Wz5rJLfg3VL97WC;dyyq~#0E5!l9` zEYBe&PQ1VjJk+CN@TVvxFyP zsoOsIH=8b%WQnqrPXrtJrfVh!Bd*xs;KWP_F`PUZN+;czLd;%qp-a{|;85V!fYUCS z6PR6(GBlegWeP=-shD%>UTK$R#oRN>r5-~rXO#N_m+>}9+Jvjf4e23%^Z#S-P2i%+ zuKaOd<9k#+ENY{m>NTK930}7G9L9cp z^D&_jW`CHQ37Gry*cg@BhRn`-Z#|vw!YgawmlZ`cy6UzbpaRJG9; zVw&Z}yu|2cSp$&E6jFb?sd+bLW(seO$G!juqG-avv49gh5V+pq2Arys-VZ}YC!e0O z1iHY5fpf$HI8_7#nJ8sMT$3D3gsNDI3*=liW+NHP8lRN4VsR=~VQbe%HA4#Z2h+C) zYU3lb#7Us6oJHyw?I$36iR7)IchDk3x&|yzX$2mx_5QVxsJ(g~mY_HcJB~#TVJ;x! zij$|?b3)D(+1;2r>i{p8^^psw^|R!yARiGj6J;ume+M>8Yds837UF(-l<+t|Jp=<(MMXqh&WwjMc143Q2300MOBVoD-wOU& zb)W1K=5-$>X^O;!YMi37728xODTw*M$=)4gQfP9_=9pbjI|QOYfvKK% z=Gv|C_Uw_h2{IN_($go;{tfos{ZKVczQ7Eb?;QE(J#@`f`m)8lK1DFc3Hm4z*BH?5 zab{t*pkIzLPc2exBc%^=C3edWKsdccRv$7b+$}Ew)e>!yx(B2p(tR=mK@(z^B1v#_Eb4L&44XyX=xPZ~c;Z>m?J*%A`ZEeCjHCU0{!D z#Z)Zv!-#p0;*#_*ofO(9+dE6AsS^x%!WIpAj?Cs41EFWTN2YS+)Jt>Pd8?*=7_h;~ z8H*D@iXG!``y-r?;<{1LcNeetG{|gte!u(hI(i4Uxypt8mqrWwSwS#+2)cxbd*`uR zw(k`P;8uu`Dh*|MGL`Q>c0)R&9C`yVwTy}({PnRL3L&(}^3*yg6VHLw)HW3+>tcf% zwNbbfeP$G$JL8|ji}&y6Im|saB5?xXx3Yx}bB~SxuBLCCykRZt^@4c4tRi5wp$L_B zSU$t*2sOa!>8^;q62}g>E%wJoP}tIyL)%=9*|j0SY)T$6Cll$+7siy-SeTMhf&qHG zok&M=N|_RPRk0%|Ip!!Vzjynk`WM6i*;$b=CD5eWj#bX1H@3Q$M)WAKh_wo~0Zd|+ zM)t+~(&e!iBCr(BDMF_<0>;{6Se`o+DyGhkc>Lj{^**%A26n3~AW%p!`2@X#h)a2W ztLkWg>F-vJiIMe5{iB(oH7Ixfs*t0?Kc#r;R9kCf zJk7-F!ftGih2hyqFv$eHmWacI{#p-^q7B0pdi^|@b?^rkraBY_wip0gkzz|q@ z6jA$Soy_d3ocPwauhJV_IN%TKmN7d;+X)6Dvs;O{jnmpaF!>FIv$gHO$GpR0spTOfK`B(}!gW0D0f^xjklA7f#_-t&%T_>G8VSe6@LB&|oE9m?`fr}I zUgfx~f5NhEX$UEtbzj=&3F)?|*jFLfE0LD0RXOQK!nSa|^D*!gHrTDiIN94T)qL{h zNI$dXc|FTRamK{0^&f|fP6U4{sz)9CGe-ah=d8f-w32c(riMr$mgSJZtU+aU-f=i+n zQV$I7gHgvw%%s(6T7RPZeiG5VFx8klovAGGZeu=D+?C(uHOY0FZb=LINOF=fwW^0! zs3ZYtP;l4LDB*;J&n(7f!AT869t%05sZo5`Px_T_u=lo@SQ4?Z1Z3nmUf_;<}qqbgjj}?1Y8XN$SDnPJV-qgMu_I$HIR!23hGlxxm)@C3tlovO z#1gB|{`xh^yPb-IkZOBC1dE}7lXMS2<}l9?(E_v7Hs8DQsyPGF{#jMLjOZe9hiISV zDxEIPp&v@t^KXh8soP;X4W8)(9tB+z?fZg3>4qTxJrA>qc}IKUkMt@R4wyh_eoVk* z3&Erk^afLgB+ezqa2YaR7@27?khV{?Y>xJD_~G#NP#@@&J(zL7Z?1M@+)kWuk=<^L zGp)SziPSysi&cYaQYl6r1YF#tQ%SJJhn%!|TUqd8qh>iq7d#E#pnIaN1OegAC`AvUzH^+p9dsv>6()a z=nnaw57cJ!D`#JxucHdctCOE8V5@m?f(3bIm+a={s3GMS){d;pMqF8(R**E~BfVo% zY`W)7VB9YwD^!)Lo#c|>PszRVoxxDf;a=x;LY?h?P_^8Dr}^F{v36}j*f9AA0n`8^>k}GV=h5{7j|S>SefUicIw*-#Z|jXVP&%t%WKcUPEPuy`m0*PP2*lbb$d1-jl@FeNyC)-L}o> zoN*L%A?jzI>uYKHt-4f{KFH97F zhs8=Mm0&gy^g1FAF>nv`@GenWQ>LqS?aL+-K&&fU<#48yXAc+$Ts*rY2jA`O`F)@@ zlOzj?c)_=>Nh+f+Gn=KKM%2#S8n{h$C8RZC{Hr?6G4koQnabhVT5)Scuk+G7IJ{>I zV?27|;MeS2wXtA0NOa@nrXb)idys0LI)WkVwVHt$`k4)Q>79YopSHkLJ|Duz7qDbr2Z z6mO9gLnyd*MoM4@i429Xz&35Kcgs*gI8&MEod;~{$k2g8AleQA_L>L$V0DnUjaKUb zls92n>#)}WsEOjbfiTf;WAq!b;pb<#x8>m7Pw%%`$K_rS2Zkkz*8<|s29VZllcp+G z3U4d4SoGWh$rH!j8+ga551>*B^rM{atUMVUZqRu0oU@Ns{PmyyZT4u$_Fa7qz0-w5 zwm>U5CS<#xV4z&BoQT8P!!FQR>~+^G9+1mJj?vlSu=4}9Ayau(vN6;^RzzWSPA;^m z@h}Jg45blB4`(7HZIccLuLa449$BIcJef>&0mz0oNO#N6ysB;SUdz;pZqLaGcO+c6 zf#Lbqz)>LKG>1;{WWJJR?b&>;l4e=lFyypORkTSDs>*mqJeZgS`=XoOlljPQfZ12g z?ih5@7E(v)VXwN-75p-&3`?Rg9ZQQ&aJ*2j$PV2Li7d`BG7f-{qflef4?BdY-+%1~ zU#v8zD^R+Q#e&E{d4i8D$m@Of24FZJ7R)9N4ZP?afLM&m=m|H}j6UW3pKhBqZc*l0 z=%1=~Wz!Yhyj~X$(1F(Km^teMf&t3XTI9wppOe6E4eNp$`d;!R)N!PUE9Ui(*98}S z&x-Q^A(hdac-Q7$@YgBViFKL==@Ma?Z>pliV-S)>muT)nS`_LEucAxIEY(`(T##O# z&6tWu3VQZkI`Ai92I&r7b&OFux}K_?VVsAS`(2fwjA5$c(=R27v&40x0-zn; zrfQXy(0Y0$f19cxph$QqWW)`hFCekpSQ{@q60(D*#pRGezD%Y?CumW;ut&5AyDRbl zLsaS49I98KT{6fk$l6e?L7wR~>;*rXL)$~)iR|zMX?#GX-wB^0P)WQ%;|_&3&kK1T zP%wLk&vO<^P6CWg$a)s4*e9!Tb`hrhr6hW-^(x9`_nlbQMOeR{3)NvY%Dtcrq(#pL z3YGE4ZSNlHrUF$!GL?%LPukLmGc=wxo({an4I1;_yqfSEv*o!Jv+fo87PtA!g)7Te zzBG3Jx=%2J1l>=>p#{PE_5@*}_mEWEuSlXgfGHaj@G^wj{#orF)yh)QIT6r=8ORP2 z4@TTDdB%jW9iSYML=|`}Ufcp@CdvFHK8A4~h1~_F1f-S$s{~#X^xO!$;*As*NF!ux zJ^Fb$GY`$czB82=0~&GbgKdZoL7%72?~bRoDDXz;q8;HWfepc>qOI;*y-oiTZ3+;y z9#kh$i)uX@)JK3szK%?yPSS}%$bo+jSt z+_&DmY+ZoFCAj0l*kdVI!ep#WB_OMnLLnlMA;J|GCTvZjD{Znhwt~WemdU2BUVnN1 zNo&F9r`P5#>^8H=I+yznd1~+c{PaKF`Ii;^V4&IusT4h; zT94|0PyLz&2cl2XN5pr-?g}t9vp?#SunKjix=>h*%r1~{;Uv7+iVF6R-=;z4+=zWr zs|t)i%{NbAv(vucM6YmR*E!e1b>2iU8wom@h(j&!!B-74T9t-~wM?a-k@@TfI6M6y zw0l@`m_KPwjd5VOfPnKzPT&D{5RB*ae%*Wb@2pesUXa;^B{?r85aEN`cUn|}ul4Bm zKFm*5#0wudZ&Q_RIBmrc+w;!4e&{XL_KC#uT-fJeDOo(IJ_H+ux1ge=PBkp-pVb(k zEfU_K3j(xwwM~kgxr?{@terb)YbRT=F=mo+791@4jakrp{dMzH^46_$8T0`co=S8U zQ^^T}IY!WjiMS5I#%VV~H%`-ToR&l_g_R13Ohr5v?j-lmS=}SF}S}mN~dXripQ)dp!!v|h_aO(^hi*fBHd@Z`bmpI%5;n=#MsL{dSB3W z3FLOpPE#+P)9U?z#?oz_rcIF)`iNh}OQLRiwtDyTHj5jnt>Np%?c@#7Q97SjN9i;L zQbWY2CIdLd}I6-F0dvBF~_xt8)Dw`Po7JcT0nWjFrn5M20OdCP#P%E!gbTDX> ztSoxb3HcIXpG>dJrs_j5g5M|arc1oLCE1{jR^ojj;s_Kp9rV8F)92F&75Yh_I+7y3 zDC!SeCtIY`B+ssjDGGR>%2RKB8FnyR7`O)8+knGl?;NOT$s&h&tD&Z)!>fgt57pW! zfdx>!P>jNAz!sv@tl-x}+TX^7J)%wiRlFfjtftawjz%n_c6uFkPg6I^*HP3El)kKDuJz%Cy%z}jBD_%@y5!t1Yc3&`XW3`|{_L|kTU`5Y`SX_NNR zHR=|bc8h<9AW@J**^te0n7{<>r%s{aLkkf-b?c3g4r|Fi5jd`8@s&gmm@M;t4fv<;9+ek5e8 zRJ#dOUC;XgA9HLQv)3I#*cAo}7_dZlF1GB(721gjA?Mi-H~lK@SdiJE*u(zs!V9ZL z3-j|C`~Ov8j<}l8(}Nc6l2 zZ?h!{`Sr^$(>JHm%PjIs1_`F0pnHk9e#Ke_*AlnA)1{DKf-8*T|eox4jwdRqtU-es`r#X=$6 zB18ug5}{HR*-;y#N8C13=%QjC%LRc^4{4apo!_V~^;zN$^+VH!XuJsdCj3(_gCxs48uH>#wpT60UJNaw3=$;caT-OIEI(@`jV75sMFK+&TUOg35JI3a`r4!6% zf=(ID%^ilk*JOSIeUsNOSUoQx*f1kc-Aq={`1i0Ux{{MPL2v@2pOy1T?Q!GoPw-n#0C8VwDs#^OA)_QrNxnHa&ev)dSC%^s$B&jgBJ>LC|sg z0-iX$;~iOnctWQ%FZZC>Y*%swn{LoaE*ypdCe1NLDwzb6LC{+;qpDejWSUrJpwpC7 zWxi>uKKVMo!@P>9CRLYa5Jc$fc;!?)zc?ySJZYK|+=qNZBWd%)oQxh(n*Ng*Yhk0$ zRf;By*b!>qp_Lelx-8tsT!fv%HGYrz{qoiGl0qxI(T%zm1R*;zkGDs44ob8Q(q#U} zip{S-<~ISQbB@M2omWNKmO0K(T*qXkKh4*VY zA8Ec!RitY?kn9WT`Tsf`Tj+a9u|-v@JU|XaRmPl=G)r%KWci+s*gLCNFeL48vb-<8 z8av;v)86W6$LvOXymV@pSQ5%sGh!*Vf{G|$=<}$OyXE3^qw?67aQZH}SR4hFAVGi<=m)2XuhAd6Fo5n@0H~8-fV`}oh+F?^{_E-V zC%%h5J3t=3N0IrjC%>MrN`HN)rb&r^50g#Ge#xSuAlOBQ2^5)FvEeXzZpLBB5&z|W z$*=1)!@PZxJ~Az&S6zH|O1;)70T&uZu%p?$8O904thRz>-Jmj_XzT5)&<`8>m+wTlrj+jegm>K zkC82*rz4VIe>NN)k1{LJv7bH@j@(f8^*m&WwHzmGPk`>eq}`94*a_(?D5757HI4~#U) zv-7}-YYccc8lE{M);Aps9qZohKj#J={iU*^i7>djaIlw!!8HS_;-QE?CMEEiKo2%& zpZqW+d-uz)39c~No(=?vO+#oiifro5agE-O2wH6I8owZ$04#Cey}tcX8|WcPQs^-6 z71Ex6hpM>{9Wzd?>SYP{K-Jt43IH1&6B5{cjWnT;5V$d z5ML00u6>> zJTO@Iyjz-oZ6YFCuA5XrvxsQrN-AgTXJsl&>5oLlbR6W0(SwkUUXiCR48rv6Ye78_ z`LCQEA6Wwu1^QXT-dd=oosC6C$h*7BLwlLm=h+WwIIS|IZE*&UxB+I2Nf~=+J=zX9 zZ#X;MA%%L~+& z2Y2~7V_2U%0=A6%b31WD19oTi>pPdX%rzro;f=yIbdd|MoSQ60{{X?jHsM|(u8TA= zahoy-c9CV^cGiVLxlo+|$P2+3V6eH;o!Xi)iFq<*+^XoHAoYLjozJ(OU|+dC}VBG^Nt;b8FrwIXD9m7mrRqI z!6UjO`W>Cjt=`Cm=buswsAUrjSe|quE}z`yTRuRt-2)Osx9|+o7Fm*5 zuf$pC1AZwK{zBPdp{lX&aRW2A(A!yH@~M-xk1E*=b{P= zh~yE>c7o0t&EzvmJg)Jm<@ZHDl+>u(MB4aBtb)uULFf_ditrwqshji8f*mSXb&{XR zF#7AC%CvavN5KNpV>+~In>DGPD;H=?{42M;2TstK;*t9Hqls`vxv(R`!Wm_NO89so z77n4es#ny{OX63Ea@9~cFUIrQ8GB*nJDE`8f+S~_6k22K{mUP3)tWt!lhHHzbl>Mp zn?Bj1iYehGW`v+05^+m}&GHYRUa3{n{*6ZXd0wN_#0!tnEoAGwR`b#3FKrQ(ivDsz zbpLhSIX*=f1vJi0^J^jZ`sy^b9{VM2AW)wk_&)W1IF`iYqlvO)egXNR7uKni`)>3) z=4I-1^Ow3K9|d8HOyySTRo`9Gde6O*Bjhn~am%S5(WlDml3~d@jj6>M)hh2K${@H1 z)r_P6sAIrL)#24CZy}FDN_QU#Ohm;{Xtfh|j2_PZ)bA|W9GVuQr_PEVyQK$SpwGOT zM6C-Q;PsHB|A0(owYVF^B0q-}O|$R6+b1-G@a>9ef1%fKbAw&jKowcI!CMI?4L7Ue z(tWOl7Jw)cRMp)f??vh(E23(kbk!KNJ4RyWC6L_px@Q3eNEX>WNzq(RHp;Z-wl*hD zIZrPPS50Jl(1lk?EZc**YU2tCS3)Q{&?hT^g3MvhvmimxE9jYQHcZ=b;gsQL?X)=X zy{}v6@w24HABh6mEl5dilYZp?I97{lFm2MR>8mO2)#zmRnvitRc#n_ORz$T)(?XD- z0ISK6=Ashl2tZhwdmdUK4eimB(BdBDW(u@qv$)D8FIaBF+iJ-9}&M7T1%hg=z6r`pb#_SA+I=>iOabw|P`3SV0& zY?H1F?DI7U>Oxy&YrIh)>VSIf^CX`>d$$}8xC{e}`{9%rIJb}1WlKDcis$I(n=#Y+ zugmi2YHsUW7tU|IXaW7>1Ou71AHe#SPF8gJbqH37u27#w-4A;#&kipXrc3*Qpu5(i zJ#-DT2ZD1qH7lpyQ8df0NsiK3yn4@jy{ux|A;A?YW#KyBW4C5m5_J@gR!u!d?xUAM z_S5$W5WJBC3(<_nDaBo})gAaX%Lh zpQd%w8*%mjt1x3}^ZJlW^rjaEOM5M_lut0QI?g8Idfku6@&B!wbzj^4wdO@9fA}d> zg`kvpVGx!(LjcUZ4P=Qrf>L>_nQqx($y%UBfOKoRd`_=>p)HJ()-RdMx&1)l@>{sK zEoBxQw{dmc`{iqr;n;1ymm&-hHgAAhn||OHI3Q}|mv}b|c6!y%)M<8-X>`HdG<7wx z9f~Y~D&MzP-bA*LnY@o8ANk#v?Ug$-oNoV1hVPI2yR-QTP7{>tn(>1_&iSpias-P& zgQ<3=Q%Eq<5hWCa@pPuTKE0Pl3dS3%29ihRz(=B>av+Vx6telIGqbGbvj3D=sNkX%5IMcr}p; z^c@mHv6!wnqKc0#TQKYyA3J%k*7mn*v<41*Z_n5P*pBYw8~}GVHbal>~mK@_fui@*G(du>IxD%vowFXhb(i_0%P( zN8c25k38p|Med0DSanI!rMWB52(R_HJtrf4Z$Lj(Cl~rI3$KXE2&|dDX4;{UW>J@4 ziT5yB1G}I_5+rLF?Xg>iDr-9Ck^+zH0x)MAGuFSh@z@^7N+a~If3Y?RTVy$2kd{yG z4A-voMs`&!0D;<&OeN|(Ot!2u2VB@Qymmih_;OjMbpoZ!&OIzCk!Qsfq#m{#(t{J| zK@z3PhG=A3Zlv&90WX2hg{yXb&UU_64x>G3UIzZBaGQlb`EhdvVV;E^c|AG2xVC~0R~CuKnd*Z<9gymd72Id& zWd3UTPRJcD;@t?_z2F{=&oz+9QuEF|x`M2r1s+LMlf2XS zM%aO1e9Y7$Q`saRjKUh7MxeU+TQdLe6Q}~`JUkX-_52gt2gI@^`|ylY$?3Xr`=p@Z zWb2|ImfD&-)5>CyQ96snr1)iTHk*^;%c-l9vgjd6+N>dG@03{$K`KGh|NXoF(f;pG ze*4ou{#vqvV3rbe{J3Lo`J{8tUJz*H7nE+d<{oufpOu9ab(6G&PM~v)i2eA#FkZ5g z#OjJ<-Y(duJWXXQOGT}kv;MUnON80V^S}l;NU!7FR7}F;W1nGjE^@JVZs^qvUb;CE zk#E-xC)1S{uI^_d-=kazZIJMdTyH>Sj1{PLBudcT2(9$PM7aA>AV|Q=Ii5pN2UYbm z*D~4RI3q$mnW9MORJwipfHi70od+<{nJ-;=#k|gO6RWG}ys41rGd3S}AHjgA1xVp2J>sp&1HH7Q(JFl_j*EUM6dS6rZ$^M44kh?}@$e(HC$0>Cm^&?Mrsq zHRqhpcVGRJ88<)ZJWJF4E}US!-lCXr=}S!fOY|cmu8$utTs^OTW_Cz}e2veZa6`oM zsYz6`yy%q*UVeDLV#U-fP_uk2*Hb;fd!d8uqeo#C>YJKHWxlD34guPn<;2(rRucW7 z2%kH@K4ib-kP1`Zko~)d?3XvmyVSj+JV=s=7bb$R;R*EvBi-#`pfkyofY{8|tEI)^>i@$Nt*xtO>4m9EcYHQ{(3{DqYH zm#lY>ST<>r_(>F!SVPWko7DN*qpUW=7Q4jyuoG`^vP-U;A^+p`-}<(79=FT-$1Dlv z2B;vdpesZP!eLKLA;;5vS>wWzcHOz)FBsZ=D9N2)BLaqEQJOlzE19klod)5kvkSTe ztAx40hkXoK<9hsY{gh6XtJVdU`QK1no{#c#U(BKBWKoFU|E=O(Yk8?mi&a%J!K@|d z)kNGe@+yTiou}yx|BkR^y2xV{;oMiu@?qHW(Xl-0)XSXSyz53TLcjm(R6p~A>8-O< z&eQ2GykOdIfz={{f!*g^BCa)JKza)#KaQ%B#OoAU!NV2jw z@~*rmY{k@8&GM;5l48jhs;JiY3;y?oaPbyY7;WW~qO_q-atO zMh$yv2ju76Ya$QRp9(LM+AQ)USxp+Glbp-IVep+d(&O~ZZdbOyF*r?Po^a&veWjIN z=fV?CiN%DIMKF-kl19Ygdbdql7Say%SbKO^DD7PczzswmBR^1oD1}gd+U#1-91vyc zA=^XsiYCbs{^S(8b3ntBiDjJ8bMz&4G>j8Hduh=c`LC^`XP+y>!4m2#61FO}yX9wI z)vlhF1v!toUTY^oPx0xov?m2IIzro+c+eHG4X8Z(Xy7*+6%R!M#*f{# z_@iR(J$g5 zsDQ)`7bA^zuDa51Xoj<)%@NR(25ojg-~UBLh{lW&L-_P``rr#wx^c+@JtqkUWEqYS zaVeh7a=pAQbo6Jd1}O|S&}qy7FwAZ8?+`;KCDli+3en52LeRAscJ@*gRaAe}K8ao) zKRt(cfxhFJsqEt$?}XzM8G%?fw#DaEbdP9N$VuvGKsBM47X=*+z#2)NCTHejc}5`q zjy23y)fH;cFOGwnyC-8g%KxMv-WYpiX#xV*i2V5%8S8pGRKdI>i)xv(wDvOv>% zf=MFi_mM@vj?@9gg(+QSnJg1_g|ox$A6Z-8^7w6R!o>KGI`!&b#phNnFk|97NB(&a zUE{)wn#&fWe2QR>6ZBCc4l`u($Yc9G6A$c0W=Hc!O2kU(F&Qq*Ca!5)GU1zmISo>{fdf! zGI>iVP#cGKLNv~lt8-e}Al<{m1@sQUR8f##a+t5DD(u&+v3G7HbSXD^D_C#P|| zErKZ6^oQ@7;k5rpL2uC~UlEg>_0+9bt5D%CHf!ox&zVQbD|x*rPB}1I1rHtWG_b8`g7FvnD4AzJ$=+4+V%7Q zrnho4F)r*QAFwbn#ROAG(D_KQe*qN6QWb5|l2?qvbvZLT1UraYr3oz`2Ju_9Dy?AQ}s(r;f)WgTwKIj0FkS+XgI=_<9gx z!CYHyN}x`T<;b=tcB>fN){Agr$r=GjMJAP1r6-_gUkCsxGiV26}IA zLqv~kSg{r=v+>69dD_aDGm<5mQ%>JJFu}?ed{|8j6T+jd55LoNLgsJwLk#o7uF)AU zj2{9E;W1(ky9frT5c5o--Nsq9N({IZkd+WZNu8DwXoy(Lq|54kd&pdM+b>SPQRGqU zvFeS}Z(s;mufUqcLdjNXAvHwjf*(^G+9Y4)#I=qcXc-G0TL+(WyW7>>C#`C{o8EM%0A zg04^}RoT;=>OM_u_&JZN%@1&LlBRs)-%4q3vuXLK#%&krjoeI&3kTiGEKJJ|g2^W6 zZA9E^O_OAmu*pbE3kmz7HL4Our|I=c4AN;HDn5cF|5mC-H3WMMw_fYz^~gT-MYZ_L z^Ajhbpx#bQ*^CW)8y@;EKjG&+IW9|l>$P=?nuRCcd~#&kf6gM)V^0ejR0+be=zLyx zBxL9q+fI&;YR8z{JB*y3<77vOS0m~cPDI1qg421l@?EO zJdq&1N!g@qlo|t;DZraP0$cujR4xANeUijDf$xO>-=O+vpmi1C=Ina^+K3UiZdrpG zAMGHoODbdV`axA0PrH-6FV6-KV37VK7$O%@1D^O5A9bJ5SNr@S82tk6`Dy=Xy&&J8 zV8?}8u^CA@f=xH*q|aH#Ww8b9GYJL^$`+)hED{!wX|uZ2CP`$IN}4Iv|DL>BY?y(H z?ptRy%gf~T^D-t&3xGSC?MEn=C)v?_rSn*ij~O6suU>qcPT{uHap9>47>CA0QE~|; zhoCcwxQh`7Az@@4zdLXXgN9N&L|4aLjA-}BR9>K0gkBB03JYjNP!io8(LsV-7%yig zn%K3`&c_jV0GZ7q7c|(>R`|~9KYp=R#2=Bm=-kgp5mUZ<)kuoChdd=K3d&dC^aPF` za%*4~xm8&sNT5xm$-N{bcuzvErJX3TW8AqtqWbG^?#~$YW7j6u)7xGcV{*{Kn3NDq z5kVI~NS0hd@0KU?wJZG3heE{O6qoo(p=!Rn0^9t~=9tciV+i&ZES#QTffqJ|A51GpSGdJ<{@(#g; zfNW}y8jrIjkAmt%P0@X_9;idDBU>YCM3{eTYS|rmo~fsMfT(9Rou-Z#)`?2I_411l z=+(=IdFL4iES4Mv2up<6iGneYvHL0GW!^@)fypdFb*+9_ih`27Y z5OZ0pCASqf=t2t1Pg`V|`thN65%?hv80#^w_qm6*)egiEgCKF9Ec#I=%6M`Iy4cj_9rN8z6tZA2?u2Xj5jT;s! zrgCvjaAou+UWo+7c$y&lek3d#a-XxI0HiAnuVpH`M4P3}vSVbkJc+9HsGKvX=n)N& z_aML8AaF89Y=VR3EHWtz8g_$nrhA#cb?og0NjqT)A)0uPTVzP~2UF1r`4tM?Sd9C1 z2r!6fp_3)D>$;I32Y_50`d{sLr0%P zlHL)wff*G6d%O)sdgZJ+5T(Q}(wP>)#O68sF*YvnEHBz-UB+!ClrGA%UJbD4Bmw3d& zovk7mn4ijsI8+cRnBD81r%nns6-DgE)9(hNB6=1t# zPes>y?2av0Xi?iaHCzuXdw~FmELVb&aw!w6;%nnBNU2LsAFA+Yyi+ zYcM8=IO|L~kmyljwIJuciL?IV-l zK)KcGn7u*o{<(C)RY{i|x#fTz+kc<;<@paK$G_bJ{B$QE0*E4=r3LZ9_MaR@^Mlr{6jLQGnUWj|(rgYb-2JDZxP2_D&-1plZ2)yQs%Yd)xb{yFqHI3(6uhK>B0EEr;H~-=!*+ zRz$T)Aqd})7ZSOi& zJJ=0v!SrPg);D_QFx!HoC-cQKyK?Bwjd%WC^M)B2quTB+Yb)9<#`-h0_Nv%wB3Y=H zl#=;@%cqV#Lq6h^pFVs;r@5#4fR{pSU5I30JH3jdkOB;)vXZD?UlinQ7Ssr`BrVEj zSw+-di1I=;vAk39h^N!6;6LK^k;fL|q9&VP6Pc<&vL5_hj}!@B1F-N0#$)r9Z1FY8 z-hdTT?K4%}F!BT~9FFY54kuefXceda#X7mgWf8?vz>&?b@LA(s#Vd|#Ag@Bv2Civ7 z6&0wmV0*dGGbON6dJh;-3aLi^`(AB|hu(OesXU{8I0?$0cA~_tk>~OdHwlLj05Im@O>~@oPm3r)UKIS4Mw=5Ah%k8r(?4xKWmYy~^?7zaU zJM37xop7hc+P8fgH!i#oX7P7#`<(+`?RLqY0Bt4gLg0$V(SvQ%NZ5i1n>uoEBX&e| z`hMf9QnLpmx+D4>ojhJrwhISkOD(*bY=QxZq!V%Rb5a%8Bnd>P;+?y41L(iQl9$W~ z+()PI;)(q;Z}5(h@9m%21+&l)eckt(&t}#B71c(-wlTwR^kUu|wql z{pmk?XPY7N&b#^=dMCH#j|+q6s0DcT6U;t>E(Z=}h_1iWsW`CUt~@n7pLdJiNj{A2 z@xMI3k<=?1p#socKj$$3I$6LtHXlT-e62Oi;_G-cVGj^pat(U2m+II!GbSTdhYn zkS9F3vtsILWeaRG_xhieZJT-4aTJly8fbd}^{oCJb+3HjUoL*$Gi*UzJOub?VV?!d z!_K&uE7JvM)hj$JRDF<>VYk!mAyY8EH8m zld+&wd6%j*!kDCmXoB)Ea!3R`O?pL^c$xoTR7UI#dYS*q@O%oS!cUM)~K>vaX> z8J)U3vKdz6x*rb2kyTg!wY@ylTEx<2oe>sw%{|@&GcY;Nv}ZTshJPm8Y#3jwErV?J z{wL*>lkIqE{j>_5wba}TVmnww=x%$XniTBdltvf?rgRh70Pc0KQSJ?xbST*ZB9oDr z6zqKVOB%Md#(lB!u7M#1IEWhr$H<0|mEkEdb)k^>Yf8pO0S^oQz^Dy~Rqlq>&e+Y%rVM0%R7I3tkva_^m8sTKnC5?nzl3UZ@p$t2OsuI=6?bl9+iGKc}s-a{hm4q+b1*qk_}LJy12_jyDz}#aX^w$ z!R&1JmWaj3_h*onL;wpM)hG7Mpe53IzI` zMQrvk!88(d9T8`USjkTbT>2mPf{|=gi){JbaL^r*sUGHOk!dIov|T#oSU)u4b}!hp zv0M>_H7ut@opOWVeXqqwLW;?hdAVd;=oMP8fIUU>qAK70u%vcIzNA2{HKpicC48yR z68B{9CE>LmiPMH?Y-INog#$lf&^=?KdLFQzb^pO5X2jDBqw24I^Bn#D3j-riSs0TE znL#jH2pEMp9fKq*SOu1t3^9|Eh++Z>tT75= zqQ|J8r`_RIMs@lQ0Cxg#?Lh5&x=%j%slY7;FZoO0ES~HWAE5f=&iaF{r<25mW>;@-LFj z;foU$iHbf+D#)pRthy%A)0xUY!~Z7BC~%%M>>Ot9k8uLVlzmmjoz}qM5^8Z_PlW|W z8$&z7cB$6$Dtkw%Zr(P?*RQr}G<5S!*jios;Ck z!_T6ifVDS;5`(l&y0}P;bW3%-b%DE7+B|g`)y8y?Xi}2h+h8pW(ULy;K(OPvmc|(t z+fguPdKsr5M_*z$CaG2b`P*sc8KwJI?SG>aUl`xCz+(PMBN$leZ$xbcB%*~mNY)`h z;y;tx)3LcGidrL`SD~cqJCkTXxS__;*kJE(=3Bpt{k%s@Y66O>Myil1WOOjX+h1-B zSj*%=aI}t$7p{PuqAc-l@))@t^5^eH-;Zvj@*@f(wz{A5-Y%(vlH!m2A4&?TqoOkoV; z;*+>=&IyZ8a*V7IT$5~5VJb-ja2A_<6r=K_vAn#4L|=CDVr8zoES~Q6^SqbMrX%RL zwTtL2EjncE`%DSE!26) zYF!=@AB(bI7;Y~hL30I}nEDm%9&L)j;5Nm^*;!=m3|Ivthkg>20s?QCNN;2vL#}g- znn2##IiT533Dgb<&UvQ<>WwTG#wNHmZMWY$5ga{16#LOAF1Y^Wgrk@KanJ2CYaG2G zu{9Rd>0qCxF0_ZbslWwtrt*mYCFBdlRnx|4?H(P1KK?4|B%RF1pLL-*BxEUIn?n-Z z2|Vj(ZjaqmN#lhOgV8Cg1GHET)Jk>n3TDH~8-QpKbjiEOB5?sw@2_R>SHUkq>AQO% zcV{=~oVpDNSmX+}&)R?Vy?1^zkr=HDyEQB^T4X4^$-{69QcRkZfO<&WR4xOmrq;=3 z&9L1ZeKuy3a`5Y}^mjDoiOLW@J)J)2!V}dci;3za!88%{5%g*D)N7bMsuhtPf(j_` zI!6w{R?a7H8B4v4MS^8Oj9wX?Ol=Apkm@5xf?C|GVoaXR=+)X#ypA~oB@sF51in5J z5(B&k6+67}*wp@%GMT^O%hx2E{67|VM|Y_Ufo5}X=bxviANJSL3sCE=rVpr;COAYt`M@M%)OlTJ2dOk1ST5R z;dBg;Si86T20M^auKuk)%nX{Gz}!uAsSBH+Qx*WJBN*V?-A}|V!gSbpVNFb>$D$s^ z`e>bI`PY#cWu17?{X%GEOj77!b(bcKyiVo-~e*RpTB6X=~3&`qaiL=9vYI{XuBW06;9L;cX00e9wAueFC zz1NAiWCu&=q0KJ~%$8;0#myhkt6kW#fcpBFxNbVZY$oUwBCbt}Ds-?s#xl|YWw-B* z*_D2{GBze`cPRI$HU?t`WJ=(qfq@GU&^Anbh;8eE9SGxCjVUMpdgZsD_o~w)A$jU7 z@G=lI`;x6To9foUw-G;W~1@$x>c{z z-BalmFOB^+hY02_LHC)`$*{7f6&411MccO$zpO>^;Z^(-(LH49kM&NjhidWu{+IB*qa-AckXWJxYKzxWuOdMzeMXrWI`o)-H?v6k-|$ zz-|IDHH=VngYrRw5VN5nR?#9$5+hOht=Eo`rZnk(c}6(y?ci@jNsc`cEdmVPqy5m{ zmtzPHZyiz%d*W|dJ+BL5Drq5m1KNIg`kR;kEmvJ6T#TJV0SA_v!f(hYlfh-h_+NJT zL3L&HFKixRYnQ;u=bT2^b@%ojY4~y02~#gfw2vi~3TLind4y7;DHH9f}1 z$srghhR(pf>jWWcF@TSgMLvMy3`~1iCTjr@WQlvpHM|4B3TGDaK0RmkysFqP5;hf% zx;s5r*B&GoZJm?hz;5u?E&TVT));ZwGBXxDY>_sL66ieMHA$_cw39G&J` z$OqJU!O?lA=X@$z61y(&D1Bu1y1*{Y={Z;D?9^7e#=63t)G*!Inn9^79+TS;Xc0uNE>8pR&X>QG3 z7uZLdx~bxIX-Y+1vUfX0I?Z+8PSHClvnyum;FmNq@9H#r$qxf`njFvKs3o#K+9m^L zE4UoFN*e{xvwlifu21~Db11Bc%n|G5)kL$XgwFBI3Mz+01Dz%#@S37hUdu0$)B?pI zaKHKHh)X1Uqm~3GdN;_+RVkiTpeA)!R!x*ha>UIs1JXLdc1gNVIc3v;KI^59e{bRt z7{A|%UghSv{B6PYZ&aHT3;VMMi%IGt!L$-|3#2@N&{L5Z6Y`9dPi=~;R15E*DpTDe>k*kq;qoLs zvP@||4-1|l_q$Dz8J1Y4Lq62z*Cgqc9rt%LHF5jQSm zbhdsV4x6Fapk#HOk>~NP`Jz|^ZZj74$#Upyez~Gsk{4~X;-DwZ-a;4v@_r z_B5qequ_0i@w_nT6JWByN4w!z0+ISWI z1>{3dBOzd~JLo-y6a^)EH_I_d+%MOK*l*gLJOoxR*7j3QHp+DuAeyZ_s4qe6cD z^2_v17v76pX0hKfNHF~b-D}DS?vdfFr4EZl4z5ufu?E zg7AWWg+d!2iS?2#0$?GF?To(U4tJ9yXtvqF2BZNBj6+Q6lA@VBk^std$hWxjg@ z;FtA5o>r=2ASwqoGLT;JEL|u;BZz-i5%@rs7`dFGhuVntP}2=_v4fMp2)k?QdH_uu%w*&t;T!{4IMjJKKK!miQB79hS#Fl_{_!%$zF^nI@) zaWWr1|03Z5$RW=TA935kYnI>SZBlvs zq*rvwN}*h{87h4j!Pz?agJ;8@c-F$;@#Xn=yhLa|Hd!9_Pt5kO!7jVs}bIyu<=}Sp>6{pwo!BL#j1Ah}-}cUdVgtWbS@ohSV8$w(-xL9t1kX(X+ z7<48P*T=s+r^NfJur{;>YPJWW>Owd9_xNw}+$-tyJ)uVNJxngI4ef$5JiOB5-$8Z= z@LGd&%kFSY?SqN)W6~dtoU8e|874=6e&u8OFgKXE@Eq1*0h3mOX(8w{MBK2VRD{&4 ziLwM?6|BdNcI1To3e_V@2|TJwf=X^&QI5D_x>1?p1eERKo$_E5D#@V-Jagy*au2W1 zvyp$sAI}>21+z!oE_ffG*8#rq@>j}(_XcDtSIhNOmps|MP+cFA1j$8>u;9cc=dkBF z&uZl-GRPwmB3YicA*Ab7te{E_#~bXE)VP+xURXVEN_zVKxWAtD;%|4pCi64H>aDX= z&eQ4KVCBNrXTJrkiU2nSp735aM}2wlUkKLp~M20*wZxNH{6T z88eG+8n*gPK9UgV`1}%k9rB*e}!F7a55XjoUg1p;9+(_-qBw8BfW~7 z3*y3_Es*n$N%h=9FsTH+0i!b*Rovv&j6_8!Zd3qu+8B2=kVBp=p{D5LWSS37K(Y%G z_C1vU^=ohVo1tNtA9jt-aN+%eYKy_%MKDmzlt;utfq`HxW8|bt5m(2gi2I?SVTA@< z4+B{`JD=>5Lj=ydSD-cFp(3hC*z2CePoiK`Q(eca6YQ3oI?ypri<58~|50ON@zcKX zecn@?49S#V^nCZLGtFpO_O|~TI*XejapAR0tp&Dr6HGBd7ZP!&L^l-o)pm<=mjY^T#;yFRios8?<}Zx;0W;*oQ# z&cyNMob(E&_WFib&4{U<{g;*WE*G{h$1N~$kYHeeyASpwWhS{iFevYP9Z_T4A2?iU zXSB$mY7#izN8E6)5ek|5Wa(0vZ;-RdxHl-Ti9w2pjMx>P#z_e4qIQxxxPbfUus%ZJ ztQOfIT_?(h2tP)2*GJVkBm(CEJXlyCB(C=6$3%jgi%aZ@sD#%!81m|R#ZHr3#Z)Tgo>BmzwlrB2(wKQxi8p3?{Q(9a>l}@ zG!P6-GzZW@g>T(wZ$PaFW+!0IQ~f-!BT(UalXsD-4oIRZqB;ayeb&x3=0F{YfZVDs z*feZa;${wdw1;N3i!ffcF7Pm4TL>9hdFsj-)DB0PG!W|~kv`4D&zTP!lcrxVam3jQ zp{K3KGp|e-R@@*|n;aKwUB1L6kmACQ3`?<7YWPR$kDyRD6~g!zqyyq5v6;&I(u1mV zxpZTit)ixz+031fCjOT60{ zNF4WEDtvpGT*JQwNdeiCKa9QJs_D{H%Vp?os`pks%p3njreS{(Y-8VT{Xq zFD(3a%ffG{8u`n@FEh!rlj&vQ==`GiSx=ZjTK9i~ucNkw+w7CO`SrV>cOhvE%$HMF zsT|J(voe(hWL;>NLl!QhZ`9_WwoD&3cjmx`oFFl! z1pdJ+v~|MMqDzXRXl<`_Q2v4F%G^A4anvvmztsxS2RsAt`=lz41|;#Fz3i}MJXl;o z2fqEMyW%z;E}KvJwNty6TQ7-OR-OjXbm;a?QA57-JvnTkL5B2@w0>q^^cgyJ+QcX6 zr(f|iuGo%>XLRC(TQ3MIe)sy%DVcdmeBxW*zDjT47MOM69iBZF3+U|xlSR;5fh*G+ zSF|XDrZpuT7{~^4jYlmX%_HU(tq@(In&o>apn4gIY=^WZbX$?A_Y2GB5&O$}_s_dO znlPdy1pBxYx-|~ zPp3|$D=e&W9>KtHWWg4RV0$1c6=soHb6{6vm(WzZ05UI-X`Dp$`lf{pD*AbQ0|sER zfTYHI1Ns&CNIjJn;y=1q#Xtrj(#}5pj(HkNhfQ&PZU>H=B=)migog zm}fe|l4)&v@Es5aX@CpnLa=52E5oy(231?;dqy(i)+0-lL9Eno#BDE>U=7ioq9M=G zCXTmyT<~FS=q%31A7;C+vg2c%9M_lrAW%Pi(~Oyy`hW2;eZ++^(`kX3^8|B_pwAL< zn0y8{!X!w9Sw}#via+#uSXE26LgmTf$%^Hl@~pU7djvH%&|iS01sgsUlU ze*4oofo5D4kNj&To&CZrk?Sl_xQAe%aIgsXn2mV{1+(i!$$U(`M||K;a4IiVQ9tu; zSeG29HP~Rj%|B|SzI;I9LV1VRh+BiyFrx{Ax|p(pP4h@>S|z>1=KvgAhte?sv-P*x z-fs8KzYbgLjIc;|^efgf!(LtFB^rYa@KXEb>xCn#k_ez`E)t?j5K?GEwL^snHKRbn zObp-j4I1iwAzpzY3gp^=QqES$N<<1lQ>*3b?m0aUj;lN7Z5uyd+h3&(ExCQpE*qfj z`ex+=>w@@yq97zj#2}3<7F<=O)bk(^a8fel*~$UlWxr z0bLs7rf~svG@xQ`j(9+N>$Sb|M|^zvPmk%`h)ui($d?;|X4KE8gzQo1J?)KL8L!mkPosoK!Yg{CV7kU4XoDWo>g z$Bb=kvSAMVPT1zUr>@)ZnN0cX9~Q*CZl1{Ax^*stJ}};L%Y`GlI*aM*1i>65=)*)@ zgLEaoLr~3hfHR${XjL|lr$mP(6*E&|pQ~9m%+q2*lZim#dURG0?qz}f4A3GqehUr;ltyT`6OZ_4Q$H2G4^f8?{Is|Yh6kwGbITm2(Cam*-~%*NM(UVs3+cKUkpa5?J|%-3aR;7k#*J=zD2DJN=DE3p>BN z$R_LY)EL-!VImN+@mJW@?9x@Avu~TNPWC-uzWZc6tK-5!o;4O^Q;!Me0YMKDaV9pq zOlqGB4F)i|+~h3{MxPAHpNyIqS+Id#KpvPrFypk(z>L-2umqzHg%prWXMe!cDeuwR z$MS)w+8HQYtoJP=wbvjo0#E8g_DdeS-HzNt<@0XQJG~C7>SsQ7yEXS7y)?2}fY&b1 z-=?YwuAiAi6?i;$D~D3)6`l{mHu5)4E1cC9RNw)8c*n^9kG(GeYbwju^@t}V8$%|7 z$q`T_f*@8g7)r!KYv``qud3&vUia<$`c`+*{o3x^-PQNyR^6)G-Ds(WmLMvqpoFLl zf(SA>GCSc66e^0SAQ3D?5gZT|-dahtNaRQkB)sT-U0=f)_SwO+|E#_C+H3uP0h2WS zKFJSHuL2sH_xxIR;#!BQ!iCGA3Le zVxPN)h#pb&n}rQOGoj>1C#i3fueh!99XPS~wi%|bQp^>KG*MCIAVvBpqSGr?ax!p@ zOHxFWBw3y;&romX*T@%ncPbkqTZEroRhG}M5$+3bpf3ho2*{^5OUs1yp-E(~>t;y? zAF5VD`}lE!l_I?isXerFioLUVhe@;2#yzAd5_w1HVu%o#E3! zpQlTw9hq~SsSiyTRSBUl<0Uzy8sCNB_-)vNv`d@*-bPATeh1OCQDDr^p>7YpS4Y z6d55^0VM&;LePE1Fu;dhUY)@oYLHrRCl8~`rJ<#QO!`_dFq^8jN?UmYPWN8XPu=B- zC5pwQ$Q#le>UjkoST=zDBr2~4HvqK{#BOlfy1=iRj#Wq)_DfV^vqa?qewBYm;KmUA zXajreMxg#i{Ya>u58o6HwUHHcPjH`W4JlN|0q>9TkBntN1Q6qv?eaSiZ0u3S@h-!# zqtfUqp9h}(PWZ=Z)$qr}NxJDLee9TYfUiM^JW;6^zJPPf5SGMoJsHvOw%xPbt>@yh zJMhYoWpDqAN2mC%6pOYWxvmQ-_Sf8*TIgL5NvDroN4ioQ<_rFu+g978=k#HZTNAc~ z{rF|vg>9pOVo?q(GAOiu#1AYFs;X$Nht$H-Ek>{>utzzjAh3G;4U>*NMtbNz+5MzP ze<<5C-ZWb+`6BQTED{_xdJ10h;Jx@&6tj{d%c&^jfWZvrVzL$z&3ho8KAi6`^jjM3 zRVzk+v>!2|=h;VpsJZQfZQZq&nLRI*lrc*hQf}B0#d3)FwW$|_!>UysQq(B!heQ5z zpKA`ko7V9fR88XD{9Xtcqa5IX(_-iQ3arZ2!JVEr|=d7;wlR z1BBb?OCDeHPO5f>*xwwD5I$DJV5AK{O+P@r=m{IBdW{uLF{F&Xs_c~a!J6loIyS^e zw_M4K4m>FM%6+H*MrHSOY?n7_Gk@T%BvKxbsKj%fb34I%weIC)lXHeM^2ziP!{&KF z1CuL`o|sB+f~52w;R?uE$9;IDo=@;WR%Xm3i*Z@F5c%N`s4>$+lqu2yX6!2OTl_(6*ByJJn z*lN|~)`@Ca{c?wIXf-T`@5-hvIKg6k#`hlwS*~*&78Wd-rMQAYvr!ISN{evq#5N(O zZ#|fw@}^d`ORzqjaM;fxofp-X?ifIsW z-D--7qsW(-vxSNx5ZJ>x0@DbAPpDIQUS@yy!tBMviIteJe9Z3WoG>w7|K|Q(mfY{p zu5=vOPi0|v$0X!MRKKDj@?ttuPuID(D)IP8*j?#ya!7J)heJkkGl5|n77;{!CXk)5NFX^TqzIj=STfT+#8W+rZsq`+NwthyL_!?xK zoZ~e_V#XcrEqBwm-nc9+b=e6&TZA=2>~X2+$rIb4jHX6`SugGK`pCo(fnSkvII{4?7ChH{gOJ>w6u2 z%UK^Bz4C@>-g?B}{7X{yg5~U8G@E2kQOt3QRKsQvXw~#6k?VQDzYMg|wW=!rS~(D~ zVo*&Nf>oOkM#GQ`DmP-K-dQ=5xJea+g+mARi!X)8JEKiWQ(u1-a$}HF zGf#2qm5)ERJClg{$Z^1n`Ga;o8(-L6Z#e-utee9!r63FI(iuqs2A*at#exJZ(go@P zopd3{r?o*Y)c>DS07vY$VF=HB{N$3GpZ&r#0f|Iwe?ZPT@Ur@@*<{m3F|8D7rlOXD zKha6wrWccI{OAby{`ZoTZY$kSh*5T+Q*0wNHMcQ05|?zhNX5jIY3Cre zs~5H@pb8~XiMLyX*Z`EwF(DhxNjt9`>JdzBazMKRY30krn7+4vq6YQ*P?Z}CITJ(l z!U8Dp#amiczDL#cQ|_mszFbRvyqWqqDWC#oz+0e23W<1hI$bL)^v1vHn5)*clQBBe}>4JJe6V`>j?>4V0I}t6@=0 z!4zVx>YiWh4BVj1q*pTK(=>Pyw;n(s6bj(d)Oy&r%!U;`ni@^LDqa1M-adQROs#4e zsj=ZsEFG=H8ed55Om?|?Q6;TY7IOW2o1MoxOBD+Z1CQPSw5>V??MT@e@>R0x7 zCr2J=AgXW4A6>b_UbltOx^{}R{K8w|Qy zcfvNv&ht)+AKQbI=s9C%J&x?7smAuFE&Bb@9sj+uwRJqq&|W6k02=~2$iE!>`THgp^};U- zACh_pUcPjjIi0N((@c>|=yc+44w5cp(pP+Y_?qpa&FX!A8~n7Ya$W*I#UJINF>0nc zAxHoU8$A80fB5iH%OVAK;hI*%VUhl1b+U zYE?J<_xs{f2tHy9gQ2wDbCI)Xm&-QZAz6)Ft6J)MRiPK2&3nz~ZID;@W+Cvtsahj#Z{mYpz&w7|(R{z1Sg(PVL z*=;tOSrh}}y<4d$tVuo}b|fsDd~;MEnM7s0*ZG$27gyiTc;h-mx)PNrHH5@c`HT*h z3n)RfL6+{P3uuVMd>MOwrAIX+EU(iRmpKiI+!naCx{t^SXTmKjSD($^b>Q)5sb)=(tPQFY zLsX}?(Cw>9I=`mh~uup73?PJe2iWtvwOG-s_O+1w&*4xE8dZ8kmb zqZpVki%{LX6YRoQ9{q~d;gE6?N7_T%f@|dm_>Y6{27@YSlQ=e{NvsQraX}j3dSEE8 z^FAunoB%1&&C^qy@KPpy zKy9NjwR!}s9(=2Nv`s^Ho3du|*KuN#O?&)gMB#c{cxn=fN?HmspL3C)U6=Lp^LS>{u8g zi-r2&uZo$GCVc2arX-X74vdfUX81TxF)*zhqM}w#+6LZcb5Ir!w{wrs)wGe{vWuKy z9t(0OEpk5XxiO^Q>5dG4mIm#aw#R#8$hMIEfFtC@Y@wF~mO|(jTVg+WN2@|o!G_3c z*a~hDcGGRa?Q&qa7q9X`9>iXx%g*nKMgAfj!P>dV(tJeh-Tp zwg%kL^vAh-Yc0|A>?GvCJ}nEvzJ8@iT;|_R^fKf#1K-xj2Zl6ZJ4h=j_TLkzLE_Ac zfJR@G3M%#=a5_bhtE@(L%|#O@Srqh0a(-&RQ~&g6(OOBWU#cHImk!BOT^?PEj=;O& zt>B_G%BtKB0Ebz;@(QS$@0E5K>hI%%u=D#hcn6!;7qB4Z-@omPhRJH-K?X>VAnUmWi^IVi~G zpN1MZh!FD6Kq)O0+pF*J+epkTt!nYaQ~ui|TGbMH15Ap>v(0Wx+)6f0oKG0pE&reWTrwKcX9xC>SxBF26l-A}lo|*E3;LOW(+#OrFnUx5uY|YB26z*{7TToc#Ve zOUvP~ss1dspi|rvmP(&c;SNw5T?6#AdRp(dnCcPsxdOGE;TlkeEfVa5{iFe>4#8R= z>%oNV0jIN}_E#&k1{5p~tzE$Iq*HcK3}F(+UHJQdq|P*1j!!>pEg{<+c+GU!%u4K` zm|}_)K$he*edKE8p1`Bdoyue3y{=~>dxK#%@kyoA zMSJP{q#BCv<6s{;kKfGeSA6Ahm@J*Jo_AM~!*B3hHUZl{X=78TV-(PN22oFj-4?$& zL5Dhf>2ZR{1f7oxCw@X!Ixuvy&7iZ1Vv=y5Eef-vGij83(+d|4J^`1alWq#?n0J$h zcQJuas~V%(X9V!D#pn;)2Pb%p`-f)tp_g$_2z^pmHDjReJP4S#smrI?)Hn<~45kLw zpv2m~Y}t+-C3lv3DJ*jh9M&meNrLF|s1c^~7lV4-Wodn636LtBiG)321EEVd{p9Xu z!(F6MI1`C1{5PaE@(0sv#A);mV8eBXxi#)ktOCVVE#g2<1{1WQBKm1eO z|1?c5+78iUa)#Ufq2p#zkRG$CNKY{>6uC@AU7J%hGZzd>o?BsL1>LON;MWbchA6Rh zVrF4vL*&kh4_|Kx?-t~QT?Xdd%R;UCre7@&?`c)H1*ssVUe80Z!c5)9t8dO!B<zE3^dBxM}^8K=p4y!3lv(O zj@OEJS+1NN)+Vv6l98nlYcWvsIx+Ctn2|VZZ3M8r0#@ zWMK_#^})mRvmo`g)xR`NOlM|Y7)Q=K@Tv!-Z3dMUUZo~YmV{)cQ0tYh z#@v-Opj^?^(KGlJOulOxqyRo(N?#T*%x6i0K+ypVZc2m>2m&J!bOrO=$cZ%pm1BS>JU7i zMl$I*pxG=4zZ=-f>){vhE;~9Ttr)RMj#T!k5o65WrlL_A4})=J41v-AAKy zE_1sV7A?|j_s@|b#Z>tm6x*tm8*LC)BecLq5%Rc0f?nFpwL$#~Jslm9%){in=!kfM z(X2#+Hdm*j9g5)$va{F(-SFKI0RlBH0GDg6ffgRKfl&txrP`M(H>xztmCKch$~~?L zuzLx#R@v?el3sG0Y@UYcqpeDmx6rCKPfe3(u1IsBkm1hkeNg@YuJ`mlP^--1=|ESv zk4~FcsZOWSHmA}rhep@<3%<=l%feYMPL%_1WU|!%UWDqBJjH(js~Is&XDB?7Nsl#O zIGPD74H9>YGRh|&%sE{AEz{(7|J(0;K&l+rkpcm7)u*a?bi)WOmio78&rvTE53qvct zJKRyB0?iVadCW?f*u*j7l3l~iW4c+KUF&25RnCLoe@3Dw5UttR#ZycyMOGk#yYUU; zG+}UAQU76#nc@szA?x1!D8FqZc%0jtcUYw|r;CfCEie!q?u$hF zsr8|4>J@VzOiuMl7sZn}9wb)rwo37;R)xgB)o=~)d_EaY4j6j!$e8@rkNe=hr5qoN!d-<-4@=e-;hu;}@6~){`fG8k5 zsgh#eMF_9eIWLCdQm8XUQHU0Lz2pd;3|D%? zAQU3N-=!`{zrNG&e)!?gUOVn7jubGG1P{sJ>s&5{_mYDo zE~sZxnXoMq+b||Sv9>$e3Zn))@ou17HGB0@4fn1 zUPmvTvBX-=f>C}?%q^`|XV!hS=!oi1Mzf{mz|IWImKGMXT$q;zD?DVHiDh&lMkuLwWurI(!(*U??z@Zb}@vV-uggGj!zP>~|LB;DYN z%ahG~Eja=ymJN~df^1KnbmzQV@`lI)NzbHZUisnuP90&j;XA`SWDSu?yqz;Dp=Kl} zEL}1>{a@~bVAVHq)KlF{GMSC{`uT+6b=({{o(u9&gCc!t6a(ueEfuwPVw1~OzFuKa ztI(__oxsMM5trpIaq zO|ha26lVAPwi^%IfHJ-w_#o;bCwaGA)8e)#Fj3j4tQ53|)~nh>@iEO}KmF7+sG7s% zs~Q@Vij^9?j&bNVI)mQI-zd`b&?UlJxh-fSY=Eqx%;%73;c~Hb-n&}M4JC&S>alDQ z?f1{;8&CiO({sy$9tY?1Fv-C_CgQm-*Sb-DZWngB$!H1)-R^)oU-jzojZ0IPhuJ-}(fR7?AGTLPhn>+UC1;uJLDmKhGjQh2e8p*kVvvT# zHQvCBdtk!K3rw1BHw=ts{1Fir!)CKnPV>$9kB_hUvDFzb81vMur%tHbAtAI!UX&N<(cC|6m21RS;qv~6dfr~wIK6+z_4Rvl`d5XTfFX9Czm@ECU=wo6 z46zjyQ%aG2R8-x2sFDZKQK&KUYh|!d1d-dSL~g`tRi#XW{>OIzTuja}R2ylUAy?OU zp(G$W;+FVJW#YsErwy`>U}J_ahF+@tv19BNs3T1M*0@#*)`b-NW3QwC8gRijtBHB` z#;76B3ESf?t%|tmX@bx%Pmeo8lDVyT9QVsX_Lu=Kn_@t|FO7=As0n7vX|OB>i*eTk z>Sb-T2CL}0PSK)FZ^RZBmaOwiaW8fG3A(0 zhU_5q;t8E4S%|}~d05EeJLP@Kv+hUdfU|WGBXHIMHlrnR-W}MI^6Ycw}auET})yb_?k;nB|U}$jZE16Lq&5~m7z*b zq7n6V%!13){||oa{>r;<+@gE882~IyTVesgo^gD z%-|Jqy=sg=U_Nx#K*Icqu{q-eiSc=V{A-$J&XU7If+ar*{m7Hm4c?A?l8?M_q$$)0#FX^qEhqFxslmFBL(J{!_M}td$R^6Lro-j z&7Oea1e0;!|Ln#+OPFxk*>m8WB^FTWY2@laL5f;op*Ke8!H21r_XJ~pZG$Qw3gWj; zNu6G--pOxO<|$6Sf;7EQ?u;Hwt=!HQY2=`?C#1%Wdp|(iX6d*-s~h0J`*|$R>{9Sj zSBLd0_BdD2+5Ch2F1lpORrF|_I;WBb?JZpz+2^V&YQ~S>3w=MGqST@{|3dsrV2`=}3jP*3+npn5azzO?&davh>s&#OA`t^V2;JKZT;wC4! zjB9pRPPWXW=CaOl;EfxWtZQ7(q|uAXR}$3R)T-hjyrv7$^ttLJs3&nutyLTA2ays! zUJxCa4Sdem=3wPU1+76|!#mR}LDgs#IA#xxlT=2CJ~Wm=7T#LkW#viL0EB;$rxiAX zvV8FXfeLo$^o+#DGPUNAM7&11(X<_JX&Qc>lybF@xs{F$g+JTa4ZRkDV(sbj-*(!#J? zs$OZk{0K0`p(~&Gs!obB6pPe)VT*7*uU%RiRIBQDY7bol!QaEKsq{gqUf2fdlg6I* zyC=9~wdG~u^;zxm-rzWxC3VsYs2f1Bq>Vm#{1{OjuYjkMo*Svg8V;^~!NdhV=55_S z5-zX|;y$~^c3?xq64<>bPLOPM!z{fP7frM%RZ^m^u^@hTM8#*BuUvk9=@*qisvj)WoB_tfF9Crq^Db8=W0n1!J!aZdFdt*Ykj z9ifHNOkNr+hrgIm_O^b0%mkZ;&dgsk%b8X7!M?%!A^$RNsQBv#8yhUMPFRvVzF~`i z#8WytuxvEb>X3iBbr-F8@z!-@<5rwryyH&H@3yb`_3)y<4(yAvi2m-7Zgy=|JQ60p zewLi^jDOwoydK1(?~`khf4Kdvp6`GCod@5|czfS3F1>O6JB{ym%>U|prEjJMii zxc_)|#_RE~XZR<-en?iWN`5^X#HEp(FqKBpkUoB~w43gu?|5Pg!(sB|Q-H~vcWq8H zNmBnKgWty=p}DdG`p;RbSaWM+-wAf;&!}>W4>9>9ivQH4kUVZ{9tRHIo;34d$|VPCclV%{#PE_7k7#?VpnL@#RZcjd=jn>069GA#L471&^HNc(w)1j$^{O$?#Um%>a?X3 zhM4gjckxKJnmIs;6tjjRtEi}p;)~*?g3Kv@+u`z+1Vpp2MkrmqQ3Si0VOT{rHtx=o z@wFZ+cFnzf^5;DL&5o57e`y@>H{oJt>OWSK3~tLP2X=7|njvNv#Q>f7PAV#9f{yn& z>>KJrwCZb-Urg98+QQ(TH*PU%)vLoE0DysldjF;{d54ottkiK1G8~4yym6D{?Ub=Tptu ze=f@0cPdEL8&FP@QLbWmwp1 z;{ko%+vhge&+Eyy_t@>n5R(Y?)|MZfwM>-a;{Q7EEX0ylg>-{hC4j#dVg(tL@Piy# zM#vZ^jKZi}3nAMAWcc1TWvoAGNu0sLPw}t0S)OwMrDl!g{acDSr{*W8q&m!7CR^M+Ox8@@3@1I&6}7|dE|ua|G_EB?mxS{@Dm zx<=B1)3?-d7n!`hKls3sQr%%!Ff1hL74+TdC@Q2^WJp%{_B$={HU2)SdUAf}RlNRS z`eo?~YdP>of(f%{I0M)0Y5SdU0xtFA>#E5wy616q{$!U7@8es_(2>tcbtDw432hun-j=q>^GGLsjel9%NiRg^ro@R0qbJ(S zseX1OxI*LE9_({v^D2K&$ev z5w8u>9F(Th*lG#WpeS|0r;um?`_bSJT?ls82P)r?$*t&9mIbvLN-=ka7dhueXif%p z((wyc@dtp-_BOppgk5fx#xjuBU_8`?>6Ajwm7(g|;3pSaIpOcytHrg7K}p0xM%zb7X66*%cmk}n)M@CJ0n zgOYLAQ%nLyR#Q<(UtJis+O5n7x9U)}IZ?TaTIQ;M6$?ack#xfCtIxsIaO-%KtvPwW z<0QX$quLS~FI46}3tF&b5es?@f*u`$wL!=!XsGZ(_P5L_OXQkBs%br;D+M*t99b7e8cDHrydHZm7`bO zN#jNQdlFrc;Ie>siFa{UA*@h#^KXe4i-6P8b>DQTU+i;DRGyL!K!N^4-gc3uET{*x zJ-WQgf*L%R$UiL?Y;wOdRZ}j=_jA~V45#U8T*AF+8(uR(ukoL*XON;7EP<)c%>Ept zm`aM2QBlSmeys|J*WlB}NowU-$lkA59lnpg6NYsAPs--g)lK3Sp^?NR8*-d9Ey8-$ zlPUn9eS#%yUC<%{QW#}GAyu{%A6(_5$x{@mRz>u=evAr<$o+!Q)B^u$qHG-GKwZEEMZq!Z!61FReiZMDs}A?6FvMV9K>Q5H3rWCr-?v zulpB?FOpW}q4~GwW58zZf?M+koYqe46t@DaQ0Y|MTL2iU?YI~pIkb${W_XZ{ti{Lw zNSJL|jOMT@3oPYj-SRT}sz;l`pz?w{L2_TaG($J_GRh3|fxw zrI=k5DWszA5%ha8H6~5n$Bz@V2y29$;(kGOc&W=dUN?P(-Y)7CYE@X=n8APS+b<}V z6!8mq@d4%2`UO>jb0Cm;Apj(NJr;=`DAMU8^w+cZI2TE~JW!ctdHCXxYsxB{=4e}s zDmFV~cHMTM8Yj>5yk;5meb%Bl@Ma{7w6~t_0JRrbe&dR%1((>TNP>?YoetJ9v2x^D zpSRZyc6;)V-Pb$iCVO)1-S;k$H4eP`*=c4WwonWJViOew8BKZf?{qE5@ zvSM{RZ#&&Td+CIgKS+D$03W$@$L?rN6`BY7B z5*I;yZkbyl$@D%M3M>lnNHghh)2cueG38-Lypha!?4kCDj}SZ}}ciFT>Dr ztFm3*%r6KWa6)A?{1mib(V{FUoXwIN)L9F~Jz>{hT^)Woe3SoP$%T0bAc(UktUgF5 zcraV5>h`D+XmH;Xy(k^;u1ilCy3mGTK%mY!hWf4DT;ZVr%i-EMPWT&t{iEuKb4>V4 zRsOGRQapitWo8QwQ4G|tmjnF)ggzfdbb4J@u9qAXZ1Jh4H%6RS)(0J-SMsww>LUvz zC4N1@eXb~rYgDqt>)p~;7@xd|Mkrfvh)m`k3cwgYSP!_@8+Kl~A|hFo&+nPOY(jnH z*N{z#rSMB9VE=gc%-nNyC2!e;Gm)cpU`Jstp8xhoX-qf`m*cva=PpcNX_*eeC56v% z|7Cy_vM= zYu^?7sPJwrct_A~-L4<^PBht;_W#}VH?o{tM9*=*0tAG52SxOfCmi@Sk4)y|$ryNqx{I z$n`D{h@CXx)WxiHU&+_YTIlkCe))OjfD=mi8Us6^J>t9-n8Pq_X27YQhZ%f_osq69 zA4)##h$5@N!s0>MGpDojVCs98;ax5U#DOOlmSC`v)U6goH}w$K!-yVc+hwU+gplF~ z+eGO=41tM#EpD}Z*lb={BfI8O4{dv8EpB!M2%3SBgT&@W#H_>pzxr~#B|#hu4_l>o zQb0a$g-o+wy(1LYC||m1cFjE{9s80CPVdv~pKJFMPAeG4_4B`#Ke=|k$ua%zp`Y#~ zB@P_*zhLHSo}iec6gf;qt>-NgEfrweEUd-5<;gM>dDkrV(}Vi^eFatyY0CKZs=JC# zC0tEZmI$vZ3n7PP3516SoR$V4t8%ocQ>;}z4!#>)7^VxszVwDCVZ-qoWAn7TRQiNU z=Uf{aN7_RTB!zfT_?wd0#t&GEl*jUWp zrSN!XEJ1}_aSYesJW}M{qbv+Vs#5!^IzKx)*2Ci2XYF}~6D-CT2p>+egaymn|FA+> z;E^S25nfV%E&h8MjRm~D=s4Lt&CZjz|eIRO!rsFB7#$(288QmC! zd%&%*Y}pDSV@)rKW$=N625Np|+vtdq%psQWH9xl=%tK8agRYr(JL)q|revJxWPiWJ zgqh94b*dQrT6uW(D54&At8vHr zuA)boF52pLDD2E!JJ6~RJ&xQ#Gju<;zRC_7U;lL)OVzR$Dgll~$t+EM{nhQj(~|({ z)v+N^V3w$CqcaxtxvmDsWTVKyS6<*358N%KQ^#tWWXX zc#K=Q_`*Jg$rVkV^2a}u#23sJ1!nv~Tes;H18eeBDr&J{v!qq=(WTIsfK2~d)nAS0 z^#&(2QCS+aXWAa`NQDdYSHJ)8-=)ha=5vZf zkCEkJ{|EP#w7nW(LXYp~Uv4LR95_gF)(qCw6my6oRaDf{fC1kflL~=h14%M;A^u*$JYSmJ#%&g zk5AI{?V?M(eXvgzLp7+7{(VPirFUW2BX^9rqcyfLMyE z(mOX4`1#*lxghnel?xvHHjz5^qf2ir`qvfjep(-yK;51Hc>aHUiV=%yl}@l-8W)6_ z9xcLNLl;`rIiCifblcKg+j4k@edin+a2g)Rjajw*v-2uT;^Y@9q{2dvY~WJCoo?Lm zHWU^%$}q&DIi^l>{!+PH4zuVj@t4Z-X-%+&2-%HV6+UA7(5xk-MvQ06j%<9+j^|vh z-=C(KOv`t-|K|5((@T@!xX;W&=1>d-=rX9N1fNbweZ4jZ3zs!m*p3D6i=pOyzzHX; zo4{F%li7e1mc2K*7=m`Y6p$2)wX2yx)1#jX#0jm+dckg4f=`{;E<1}O2GY~X2@Y%w zBJA$HN2`8onX<%^>yyPhE50hu6|d%3$TqoafyEpPvx>bplY!}boEsunPP!0~ORowl z2wBH(06lFz(Gn;N8w>rWk-c3zyiSWQ*}tYkc)pH1?QHrbYN)D_)P(>Udh)s;llvmJaWu5YLI(AZTShpWY{R)je8$P35p2h1?oc6@!+Qp;* zbOyM~agsFpvNXjXQnq1sj%9QqCBCi7hcxa%+TTMqj>!nHVBKJjvK2cPhM0iIXMgPb zxn%*!v&$<7p66J~Ja&0v?%Wc&)(z71!9C8P%f}2E<^yOIPRt(}lQVXlyqf>PHcOgr zM~zSdIpyGB_8N*=MUj z|4(HHz)*%FlXvN1sL3r6JI~)rb`Is1IIcraPMHCvf?`T3vX6>t5ynNFBAs5*qV=;@ z2Pcqgu4UqLkf4pyR4ACYZd#4-6r-0NbHg_Mf--S+_z`;jtP*Im)3?*DK~OJc{ zbEtaHQqM*iZL>|#`t?_H zc947r-o`v_29`>S0UnV3RMb@uV>wZyvSW6ma^EW&#LXpOTRG~CD_YGqx-GOw8Yj8x z0SoK~!-yJ1t}^Ka9xk;{Lw2TS?ZhS*tX{-R3(pq(bPpb+OJO`YzcshYrmFrVE z-H_3VAN6qbhRlAbuo?hEHsoZU#=rAf^iM2xULAJwVo`3zLbG~RC!NLXB{)xIOX|d3 zU~E9j0c9niaBZb~PZ)}HqHJf4{0Q9+UXIr7oPWpUEi#*UEkllk;o`_)!rrGX?r?(1 zxEM*=5r31ZnVI^J)g*)4Eat!~?Sp29X&1#n3fE4^uYe3+t?KIq?ekBnfbO5tPla-N zRXiv{pblcGOS(GUqeIXkNTv1Q7UfWIM~;&jdQXK4^qz94#iAa7VGh+tr_HNWFBWwR zcX&hAFi5R&nSQJq?UBRGa)z?%DnK>Ps$y5j%qT4!fA>*K|xlT zNmvBZWej`>=YUUnpC2+MKq0HqtRP3GiAq-|z|wgrh!vMZR@t4Yh29$Ut&JR&23nsA z;Sn0|f|_SrFzRbwmUhifr60^jF8Iadvb07~0;XzpSer23>xcj*#8v_;v)42gbd3

    z`O=5OXOva9~r!LQuCqyvr+D*dlBe z=MYT`*&^$M+^se_?2%tp?DSt3kudF6Xcn(6IF7ee`uP-&px`!C-?Ziqrf;6L=h3J4 zCDSZvp&ZtsVIg{|{%$2~5ypkA@@a@f4d0xwB4v@HJm9kQru;O_L0Z*Hey(l8EmjT% zr_r{uPg^wOWICuNH#gjvYMPAHvCGRziUUtZ`^_e!9TWrBA(M(K30MY_dPoDNiJ!3* zCJii5xlgXo8gN<`f|<-PO(>RnUIH@Y>#u4yhBSand`!J0!jRTn9?M9T+biW_W3$7@+nz_r!Tw zEQ+5D3!+w7Tz>(v-l_mhSI2rLBeNhL#qqYf-4pO9$kf^ggdtnhoyvNZ7DA4gjBoU&jHOmqMHp>@uO~RO`nLs`hB)Pfukk(aHT!y- z@b#EEzMN=ZnpIVa85(mbW;;c;QBjBe<0QBW(yCf#C-^*ySn9d3UA_+t+K21Bu6you z`LMw6geuPG!c47d8!wJ`oYcE7T)5e_M_KA}B+9&S4SWdI)zE#e-O@x57d-~iylnnfH{>+bbSRM;D2LzRxd>u} zYbRpjjO`)7dHZ2#D>AzA9_ndMCT83}A1Dd5=$0UGl!aBORZYhl^F~=>;58uRMB>eS zUL(^tap6|~6_IEXvn7A;5e_(E=m{&mT44ud|NIB;tZTm)w&Y{iN zh^`Ka?Rn>mk65Y?vk3ZPO#fWiBhP(5s08ZwD#0fQn_dn!|welh->)^Ptvd69;KMXRNt>uZ`JIRg- zy_ zv}l>D)+c2e#Mj6{K`Re&)exD-?Qs`Bpy zp()JOZ9^9p7r2FCNDWfy-l|+!DBi6;>s6@KJedNFy#5P!OiF?b-|hZJy5)r^9gf=w z=oJ@=HTPb@c?BI_gHwZzP@X=)XCbb>3;YYkeG@f33S2dH(hG53STSewv=(6&9A%4) z>KIzpDgJ)160gFrs~*^5Uno@h;sGqaX#8{Q9E5&!Fx`TzC0X)gNZ^}3B@ zp99ZD4Q6xEF^U1<>w{Dj=2JwAx;=JFzdlFqaSB?@FPVWQwe(qI4dIMJpjt)Dc*t_+ml|wfD_UKj~NyGr<@{!p(Ke6vVsn zq;SlVdaT38pc!UJl-d4b;tx|U`kG9}pYsl`B3rqc49AUFl1ejt6jBVZv+SUv4uB@t z+G%ybRljzcCfYrf&U8J*TOWQ--9=-%#@*o3shWcXYge%Jw?UQffmJL(G8u-#Th$;+!k9M4uv3Yi=hk4 zD?CR-=jXtlFAJR?mPp|iYQC~rtrzZ|rH5pT$H5R;5n}dNoaZ(-B(Yup%cH+-{q^nn zNJfj-^s+ww5&Fi0K7M^<<`g)d-^V`_X*WAHT5t^o3|n0LKW;q(?k2-L>7uKdEFK_M1qdBI90OTuOV_@s7~7Iu_rK&*ARKu z>xS~2IMbyeGDCgMwTE9a+inneI$R^f!qeyNb--?DW`8_Aex?Z}pMKU_Lbkm$%ca9+ z_}N1-#S|%^qHx87e4nsCeu7^Dv;s|1&1xvO13F~bDu?XyZEkfm&Pw&F9{ve2rfZ>g z$Lg?$G*X2klV~Qr9UhMsT@+(tm2nBRG~kdw(D=DvKB{du6-Eduwu#GT7?XXLFw2s; z`XyO2u@G>VPTL;dB0LD`J}04Yp_9HCP#>D?w}`qA1gMyLH}J+~DYnt7&d5*ur>IY< zGw6!&c!9B9+5G$DP(Vd^ihsNy*)NSonzlQ#8ezBKv7v>n&6CY=V>84UJF0j2PrZl9 zKncbBU_R8}|b= zf7_($bB&IGIV?wZOjZyGTLfX~2t%qWk}9<-A-|Mpu6pD{cBxinYK=2sS$F#B|r zC@~*PTVlq%=h2;c3Ag{d<(~8l#SF``^L#p>52O;JMF)Y69E}E6l-P7Yt$pp*47{~{ z*|Ht`eDuc!DTe|~^HD)({$i5JZ9a0`Tw@?8AG{~Mn_`M6l21kH=@Y8tNgaX~VT${9 z|DG^og&o)s%rWn9*DG)dtf#Mf^toLLser&-ssxH@TIf^GsgeOFB(BEdBP^ZL7z;Ep z&0)X^Q*z719f2F6px35(dDdX+IkSw#E!&~0y7N)_i%$KCnYAlO9IwkGn_ucuBRnlB zq-#~RsucekMWy#9=S}YMXI*TLCEdtN;-0XO zlOQQXy{cZ72-PHo#OSI(84DKjVZ|l#qvi3N#OQ3G5lK|u3A^Kpw&IFM8U4B6o&WSWP@_x$z-IW*L2*@0)9t7fy! z8H$0z&eO<6{piFOTkKED{KR_X!4bqK#m`rJ^#0+G;5zV?L*0G}Ee zsu#8hPyhDXyQwf;wFIX?U=DRu2tKZ-E5M0`A^_0X-RB2k-(%_{epi6>26YW}A^5xo zQ)BWXP6n=+dsPWDThjD)`5nVtb~v=myK0{1jI%KoTu$0J*ftxWXDCFmU9%cq?6%4C z)1{QagsYiJKddIpxcN5@yi7_nL)qcVpa~f5-|HwKS=@l(z=0U34H!f|vX^3ZQKXQH z+V8Jb=gz;XESY~-QOQdOr)Cj2v9}=0t|78SxXr&vaawdrc6;ItX{pNwbs>F_|5#8( z3PP^Ui3vy!`~r#q@Jf;5pl424HC@0foK>UPJulTS2}tmBWZ59exrSugl!DKC?66so zLDxqI7qh3w&kJ^vny{nz$6NJe)da|99Q@K#DP|KzlJMJz@pDk-!)%mCu6M%DlASIK z>v?(_sk&?Bcvq{s>9@=kQk76xAcwL!KdRZ^ZaNv+s~xH*+g@Y0?ymx`Dci@d8*jyfDFES|xW|qxVl#%xX-PvqWqpVJR42bh!r*JD(3KV=u--`Ok zuy1rq{m9^Zq|&JncT1LQ7K`cwHie*&rLA*t&9F8y2tTn1W?k9zgcF3vksCj67n+c= z;N1EHWQ7AGWt$mNk|}0AMG_#SqFCnIPBf)1#RM}5j6zM11#yA_ry^&hby+cIhu5fC zMK}!-M{scJdo+zXfn$8hy#Kg4-2{$%zmzW{TO1f16=vWlpqM<0&l2FUK=N^7F0lNzUk*j-i(9b{#Li#(1yrE0WHE^bDwROEx~}ku`RZ= zLqvunXr7o7i{GPx$I#Dm0u5F6#e@GG4YQ&Hd$ughirYo&A#3h1eRrFqYoZsnZ z2Wa8Ijvih^!(+5p-&t0swp34kNn&*@BG9SyR~}!1-`yxb2@-Z0Y)OTzj6Tk6 zj2LiQHTRIBNAWpGk3y-$!L>-8$I-x+kfh1|PS+K^u8qv5xkm&0otDkk z%g*^XGOY^CX1OElcRJ_asyOW3N|Hb`u-_?P*(=^G>31rY?r=XEc!BA6Y9)_^6~c91 zheiEPM+1+G@mi3erQwU+mMcGM4&LWiZD8g2tMN*GWbr$frP0(?<$s(3=BHIY z-Sh^}3xTz&3xPT*2(3rz{fa}s^xEKg-%BUuJpR^?%jl3|{pWCP$qgx$pZsv$A4~@5 zvsE##koE~=+4#ZczL#R4w5%JY`*!-A@m%Y!sq}vAdDeYp#Cb7>FyaMCYR#U&Ua}*! z&(*+umpStyFW0cke}5`6CpJfP%!VzN`+mFTXv#pL;^g$6FwDol?{{*#1||3&MeKCZ z;6iwp=VB2SA3_0|kufh_jVT1BE}ee4p^!_-*JRE#-q-pxNAyljRIYUI^utfd<2NgB z^S<(41KP(u$`T-r?{h2TKMv04X_|odB~ghKQjqi%3iWV!9^Na{AX7B{TT1-Or(v)W z31V;|1~b#=fjY_m=epwxJ#OvyoT#~MDY?wWVRGP|C#Yr`6fk@$(2V+XZ8YvQ;cpF6 z)E3SVSds-4r-JZY974HKa`1g1TKE|uYYanqzwa;JjW<9hy z$nNbG%{yt3=&RQYCuj5l|bn`5I}M5#j)hVS2LdZBD=# zUvTS##1~y8Ll3KH%#cDZdGvFzU>dfW`rLFOI;cDXHCv#RA8=Y7rl)IRsa+;42Zoq> z71C0#Qg$jecq3Z^Ppoz;^8qzNNSx8C>@fXsK+N;Zhi%~ky>UX!xYd6yzhlXtT3|N6 zr%?=)0%RBiGdNV>YTU@*rD$VTj&;HUqp1oqKG7@KmG9H z%p1Yj;6ttIIJ21hT9w<28Tn{%`y8&VIgN(nhTiwAkG`TbEu_}34meNNaZ8tR;P4-$ zg%9#(Gbm;&MRdsXT;Q>q!3AC%ZyVi3f62=M?v7>x+NWwm@?LN|uhv~3ipibv&PLf= z4F4eqdLhW-j@nguKGGZ&?~2@%KU$L9Ijn2KLR;4&T;*9PE_FHTe4jvCIIt~m@T;I( z=uOUb5a(_FAo_R5=cB%cR&~$gq|nX-{S*eANBrrN_C4Tc;ePg=_ikF&AaPk1JFp>P zsZKH#i=cCB)F0IRzbTfra{tl3U>U0Scj~0{yHA32S^KV&YHo9j14l}(nawQ?6my0m zbyQT3GL}K@iz^@~StdrY^;)^n)y?N+OX37&{1a1afugEO+BdOP*#(>)UEtF0np+lh zoU9Am33|u1at!{}$|1vhPCKv4zgDiN@k*My2A)GnxmvnP0DM>ScE~l%7G=^2KJC0k zVOvAuc%9gH89=6x2u0akLh6+1rR0P7_Q=#fMlUqe3XWR5Vk9e4SLqUk`+Qh)~ zx^X6GskiR_ggkcQWUd_8+H5p~RxHJ=pvV_g6sGl;&jE_~&93{Vj~&OI#Q+SC2a5;B z=(KqHWT9oY-Agk2u;k_$3r379YTR4ET!!*F15TFzjg=Pz13f%A23vqJLI!+H^To|| zmh+Iqx~?oUON+1&Sf3J=o2T_Fnw7=C)APlIPQT5P1Z`{B#7&s$5wi0igrV&c}`^hp(yW_CXVzDGm$}O|D`5J#B%?5}~ zfcSco@Fw{M*1;cS))-vUM25(xW4? zCB{>&s?K$(r#7M>uu*mgg1YvD4o9e18V}pzV^D9Lpfc{=s(*brnp9244KNaxRL#Xq zy7Hq#VQ1!My8IQ7O_qm`i7`N0=mk5DKu8rW+8=IgbeQ1p)shmMR-z3Wux}1W3aM{; zu+P!vE0f26XtF2s%OgTa^H8c07JLxUG&cB!Gp6;)wrI&oYOn*H!=}vkfL=+bg6kI^| zO;$nNmDL58s6bFrLoi@f(rk0stQAuTvHdSn9`ZaZ<@P<_kHep&wG~VksVZD zryscVLaqT7l-`uJ~whmlY-QS2tWyfMjiQG!cxR76e`=et1Xb>VyzQ0=4Nmt?3=e7$>i4^-Ur zs&^~moJJh3d-N!>fGoO8gufD3^r^qdcmw+WYJw`4trr;dEcJi!eyPr@Cg{{n(`riO2#$TGKDR@gUzD7UeCHR*)Y^B(raq+_;oKlr<6+`0rfd1a`oo$3Ns zh4c%u1xPSdO5dK90}H53x{Wm*F_YJ1)g9hguO{A?@U(^NJ=-GlmVT-kcKF13o#S#$G@lL~b{J%v)Z1yitVUg=X^YJEXr%WF`XFJN z?eR!?k4|803L4%>8Dk!}$9eK+cp-&KJuY0JFwS0?bAIDA8TwAcY;%%$4&L>YkX;_BqRsP;tFgK)!?#?5t%n1G zQv{0X9p4YfGi{M5wVDhQ{C!EP`xU=Bx?Hi}-PS`f;lX(3i?qf`euF{mHJFj(r`q6R zv?l-FL6(ywJGLg!FgvCaA)8_$qnJiT0T-$$;kB!>dSSMpUs0jX@NErhi`>av3OE+p zra0mLAiPRl=#&f7OpTy4qmHnQ*kg ze&~E|3HC>$GC9ynNq`0lc>kLqlcz1E>xC^|$kK4zZLJGb|Kn098P+~Mp}SJ#9>LJ*ew){Lev_P7Sf~GzXas2Nc5axher{;voXE6YQ}%W&1j3nVC|FLsc=(p zNO+^;sd2Q;fuj12pUf~isFAuqY010%il6MbLM77#9Z3|MK#{f3Pz}M+HnP&WBj_Fw zSf(gHQ6OjXRN-LkC|ZPxu{TW7cK?sIw3(~PKUX~M=_Vt;o?_QgWVK=Bv3UcD5l{y4 zzh~4fvISE#(x$(L+u_)@?O3t7b78k0Cx^R`DLoK6;!q;aVLt-i<2c7;O&5?qm8w@R zy5m|bw#|CO^!=LbS@iMieFc--Y`(({f^p*N)0@^E%At(Y$77erXUMR_IG5S6ZwWM+ zV`3f&6uXupacC;~!-~a20F>Ir+d{M@;{T5M;d!(a*NS8L#b;82=aWe`8 z>zwL=Rq$bOuX4Ba#^OyrwN8Wb`wRQ#Vr^>$bC<4QGBgJvPt+EfAnjF_g>IZVqFyD& z0+j}Jb$FF#g=(KOtMnL_kIs zfDdGB`dkq>4eB8}Z)v?IKD^m=hXN%-h86o9aUoqu_qnFZku;+ewzEQ3Fda^~&@K^Q zVYJ24vx0n0wO6sUCpdv=mp8Z!%B{BAvOY-ofr0C4ZhKVo`jxMk`#(Q%J9D_%FM3=Q-=GcrL94mA_m>76rh+mZ?)U~sT9WJ2M(^GwEl5!a7ZgF zzWltGF&1f#tLo*(q)d$XtTYzLP~x0`qOP8~OuAvMge9j2S(rGdILZdpDB}znKP*gS zUKr<>T5yOP3**I6zIH*j+0!`p{PgU!b0qaSn|t<|%ss^vTS$>Sa9LdA+^$1YaVFjD zmMH9n_F#S#)+d7!M zDx7*Wwt~|{0A%X!T7ZTLT*3{Y4^$sJiANoks+oTw>+HDpsMrJ;+bDJ`PJvO_6OJV6 zSWA*A?N>Gf%MTvQ(A*3xx@c>V$6LhQfHZ7|I>5l6K!wxHyIyiEl3_ux5o%g;a z!w#(>amrzbB8|>#*rCF2c>b`%aZ(Aq>s#b!;I4h47a)b!;jw(!Vfh(2dY_yE;kzsH zOG|5()se};ek?Yv{0WSG^rW`iqnE#F&hz&4>ezm?0ve%3P#+~YGLjBr2BYG=4H@|ESQ<~bQ%nYXp){|x*i+!v4IBHOWM(MN|SG?n(c$H1a(b4q4Uh3zmM4+}wIdeao_DGM*e6QjM@>5z4r ztg_>P3GkkbS(9z1*kp<%K_pF-<7?mytx@+8ZHk~;g%S9wFqBbE_SN=;7J5(Rh+8<; zQx|YG;)vfw@V8kJRFn}J4}$)xAYJ@s5~qHN zfEPq9hyhIg5(7pXfBi4Je2t(h9r@uNlEn{ncI^Dtm>91LiUs-dVq|zvbuXZi_u_`2 zhAyWwp)GogJm0@Sb8F#oMY>0U> z0C53VlsR&|Z_v3p$R@_l8j!KYgtpF0@rrqJnU7mz$B^MlJJaV}k;350f^}@A=03B@ z@0j;N#{yZtM?8z-3|T-i`zT=bl&p|Orxx(V$#4JgAKL%?`R{-6=l?5PNwM!yB$~r4 z*a8OQKl8i0f2dL$?aBPieL-~4B%aK647Q#cX{_K;RgyFXtRWaq+ii5*yWw&SmOfTT#F=EfNmwK z>WnztWD4n2mmE!NP_eXCg*<5gqh@`#jIC*ad0mYv&tJ{qf zn@Ev(Dk^1Ru^2)Z`<#=d+FXH70fFkHhkBmHQdCIKvAS_8Z4a zf9>4A^_&IJ>=o}k!c6gf;q9ST?pt;IY2-`CSZX3~YuSi6FD>k``! za;#@TH*BwKH`5T1p<3aZ;i9bx+PmZ?Q%m~2-$Z2BNX~(EtW&IW?+MPNvjT8JTJLLk zs?T0mrI#LYFq}wHwyF?9r(k{bSQ+DzPh+hMilc%Jtx@N&SisRP?^m=1 zXj@fK+D|YcBzW}hK1{W?*wD?HbZsO^+3orw=Pu{u=IdRxFAcqDPm1HR)e9r z>Q>@jg0wZLTDWJ)8JRvLj5Xp%nd49bk*3fxuquzVv zK=D)~$&p^+Fx+rpJgoWNiYQn2D#~w#gBTAYpRH)a>wr#57%+QYJUL zE|PnaKafOz8A&^Kb<0fvltr!`Gp9jl})Q|q{B_l-k&;QFRQZj>_FnPBJDHgiG zc2iNQ{`cqyur$T0eh}*lJtT>7)OkU|Ci4DEI_$`9VK2)gE}%?yF;oZHnk-O3 z25k%IZLguTJO(^lL+=8`K@qdv<+{gl#ov~{Gax7l?PbcyCP5!vv21w_JSEFx`;tT} z=_O3?wLd)&pB?$UH)_#&Uf;3(W_cgHarTSmnS-aRdhOVZaHKi9=mY+(&|BXwA9BXr z60!qfoF-pW3d^EBo{3Bo-5Xi~#zY(Auk&gJg1Q2_mW((Qi@SVp2Htam7*7*@09x*K zaF>m&%jB_b5mb3!#Eot0`nfkejK-$*7kl0$TlvjL_8a{sdrfR+5ye8qZ!Q&uN$QlD z_`hwSg1E;Mm3UW)Z!5b*1#}|VQmlg7A?T2Rwy#IJ+kV+?C1@qWY8$WCx7I@2Y~m~% z0Z2cxlR^=1Y*>B+`;)cX%xQ4#)}7@bwmn74Lfhrgyr`qAB+=qbD5k@;Gsa3AQ=FZ8gTL^mn_o`O)(pluM8zwL+0!CEW>Ol^f3b?jfxD zKewbj)IdX*?A!Y_JW}p^R}R59c`<}o_Akj$VNoX@Pgxi{Z$+SXw|ZY-3w_5wL)8Qv zKYBsuM{YF`)HM{5=L&G$odoUI+f>owTb`3PC%iHAnfLU}t;r&bABGxNEvYomW81BT z;YhIW@hk|*&}a{862Q%dezYD1QjQs{%A|*x=Wtkz3ZBoM?B)DWu_2lLqPb;xu3TD< z6hJFzFDFa)FTrZO-UatqZJAS(=!y(WvM~p#J?2qJZ=GvcuOS0iCf(?-&C_5SpxG6D zVM9&c0e`r*AfzA!l|vA8UBIEJ#RXtDxGv4EV0>Ug8*$hU${}dbx&(cbZG*KKSI@L9 zHldSYKDgW`U7+yI0Z&=z%aTE_mzY?}`V>{v|gAsQMib zESTDCvO#EYho{Wy@P&XihS!p3aGRTLU(Bujv2n`T^!L|ZCFkvU${8@3a&A!Ub&6c0 zqFR}~fyX>nM?eD^RC(V95#4H^Om)3{JCi7^4qESdOnezimR7oSNUD8y18d#tx#vlS zQ(NQ?@G+nqPjxk@+UFR_VUIwyc$cccw=MFZ`i!iI*&47HEXPr%+UKtOuDQUz3)fbJ zS3tAUeQ;q)nKQD@Qp0cFPVW$GWd|h()zwRZ(jmN&{8tf`@U8#4^0Ph5-lL9$WU=^H ztiVHcUwV&92|Oifb6P=_Q2%vWl@19=BSfrIW~is!$D1hy2){yV-qxZ+*thSBIs{@*F7sF58B17WV^Zl zm7{0)kHYAkw*AR#6ztc?|1LX824&l6yr$AWF(7tE9c(v}V)1=auUf}$Q?H!SNaKAvQI8^lxeG+|wZZ~H zqFWKuB)AB{-c?S>pV;mF*sm?JkG>XuRrdCE(zIlag;Fm72-@F;L#*OWl$e~yezuHDcVV>!*;C5gW)Wob{ zc7-Fc(0NInqDM0zZw`ugX^Y&gxJzGz77FNnf-b0Bmv(l*`SDb6;Y|!u+)%LAng8Se z@)FG`NTU9+k?gc%$EV3;*c&LemI8&=s8siE739sCGwK%gL$ZOS2i{uP775EG`U2BO zUuSgk_MmE~1A^-wLv+1o3!M}W4h}n_Y!7OU{Aj@k3-)=Zh3xj*@RfltpZM;kZ&XC> zS=JU=C+d+VYm#O)%xY!}HMfD9^{yk{b1rg!c#dX}%WGuil6?AHWRkGSb>fq~MgML-FWB_!wCM2@ zzKGvtlDcJY`__N6puhDX{TmN0b&s{j5DBZv0(c2?|2;B zKGW!F-26?)-^przi(NZ57N8k1#$nn@u~22Pk&42&KC;}6VY^Q?@ z^3=hD4>kZIrUPihy?+dRM>Lr_emf3?aMba)hc_u2WYtdX;YD5zvLWYmk4K*Uf-8Q= z7Zxo(BDf&G5qxQB8xYL3GWT3cLXQU2%BKV;JRzbtBsN6GlSik0?LU4q&uBgzq?)}X zpP%`#9JoOx-)0rSa7IRiCO1}g)? z1-Q}T9P4>zcJ%B=(7Jk`Kofnd3TtAZfzfb1)_AwWy0~!83M==TO$0Zl>##9jM-E#m&2@sH~6<3FvQNjZcgi4Rh?gjY0U1b+k4s zBvIHe>jLKwc_E8b>jYIq&+Aypw6li-DjlJ{yh~I_zbic8|7Ibbq0vEu4vq*eAkfeo zRVlW+TxJgVM@NeRKdO|I%L+<%qdI5FYsaiUd10EgTgISYDf_(bv%;txJq} zbx@fPtY{3rEcTc+i>kb%#VIqVWX?gr@kO!~Kx)w=xZ^i}F8^SaxzoZi+h1aP$i6x6 zi`OjpZ0j=;eb6SHyV}3l&u~Q}gw%3o)lUs6GT*n3Cc>iEHpz>)-}-ndd%KOA4G#XT zU1gknB2WGN5XrXV+zq5^#;js0DHgz70#eb!ei2d)HUtbfCkc;{Ph1cA-(|bmYqA7E zfACJ`3KJU;9kFuuxu8_hr@`^zc&;t-v%Bhz*}+5eqa(6?u7%j*<&O6c8*|LE zcR%KaDca=q?(I0lu;!grvy6ts@AuWq$rgUAA3Oe+?J+Ss+bI^}Y&lfafTEY~0C{h0 zBFj(>%4>Zffp=VX6mrHLLD7N5vJBN;0SY%K=sCx#oub89t>5H&V{uhjw-T?kZON5o zh<{>NOgC(Z{(mjsd@I-pm2BUTo^w9mY+K_=d$u&Y*5zM2$(OyM-TXBAjoe``&c zxm=#zwv9EKKw>||f;p+8qJ|=p1=`xhX;5mLCDnOBQGQy)an%{eK6lQPoYz#vdpa&G|)a(^wIrayDCeN4#=BCeXhG4 zKXBV5%@&~ENVQK}6wg_r9^eRtKj;>^pg z?bhmW)Pw72B#1MRm<~B>^$da#=|igb5eE$3riT}V=*z_qI&S`gby2`Mbo}AL`6#QN z;02HAC1I;mB}T)tI_=u8NxU5!mJ$=gl1{Oy6xnRx7pkHA=N^LM)@|xdf_OGjxNB~M z7qo78>O3DhZ+EF@(4eex`$$l@V2a?uJMcCo@gG1MFMLq1`%^;m>@r7Mw%K*ar3NIU zQk1!Zi^@ZDQpAs5J{o}Rk5HEkoi$g&KMk&*nq-^tgW~mxTJP z`x;#)tAF{F{`rHFu8=Ho7JJ>J zNwVXq=kI-G%U7?N{07~!)vsS&b}sU?s?Qykao}ck(4CSqq>er({qkfE;Dz}yT~zhXpPC8<(Gi?RAO4O z$9Df}hSAlC)cr|I-kmW{cHDkvMrbM;i;pCVO`ynHD#}oo;@hn%Vg`vpRHha55~m0U zW4~O>FfsOysoCbn#481VcuM@mZoLf-IhXve>hn2u%RW&+)2Oh_r^=^ycD3ZO-*LK` zZU{i7g*wGPSs`g4TJKyY~xv=}pvIS=3wgKW|!Q z6XaNijc2boY=IVTY>ZZwbtdR zra*9uU_FZ7(x3o%m&=>=-fapj)@)UEL*O1L`bVkOM+JneQn2}~5uY%kp7l^GoEvB# z_>i&Vi%Et@#YyjM#2UYN%|UtVJd2eB2ictiiAvhCSq_nxH$OgVH8v<`PGhg=e&IzL5;D!m4E`D=bUuT%ypOZP^g$w~J*9at>A>zLu& z1`_jaY_n@^zw8puGKzJSI%GE{iSjRp976G=W)7vFVU6S*@m zLxqLht*#9L*M04WqF#9C2N(9o;N*idT1|8E=yXB>JL~LR^&D)f2)R0 zU@)Vd?7NFD55-^ACrb%w2Yn(e^_`H+*^~fh0#=I~IOEKF)B4JO>TJ%9%coD#j=gFQ zCf#UpHgy;*jzO5w&`gp^1M%rpEwILnsf8FCbH~(dzH?9dee)#r(-9dvp6NIe(+|ErJ4Q9GW2fO2q5Irlk=ZK23nDyowd&QFjUddt$?GE^wZ5F?Hj_tL`- z_@AM}9p{X8`4!m}nc--R7-ucR_R)Y+dO#8nLcrKs)-JD@KjKgmgh8Fd-~v4$Rgz-q z0SNrz6HC;ka*$$;%uu0gjZM-h(2D)@*b#;LunEio=%q(p*9=u0#doDP5e%6C%F-R4 zz~jcyc#~1OZcmnZAjfVO?HoZH9sQ)8NJkf`(9s!kdq)0_=L_X=xn-E(@jxqG$_o?I zs}Fx)_diR;7wti=;6b<+J1k@8)oDf?as}8%iKb(eya{=gcl-4KS-UomuJAKNjBhf< zVA0yxT#x#ty);^q)vmCj=W9$X8fVMU^=z|Z(J$h8QGwIm-10&C!btP0qNMHgNDWjWmkf0eb>AM&qP36@lRi4m!^vE`;T0{Vsu{~ zy!QGpNP``_FV{@mmllcz*@yRmLRs}V_!H3dMqQ~>B-`gEC^=-PwyLXVy*mp;<>4F( zI%Lv1kS5cDSed9BG<}akcmxFff*PcHS$Rw=>{a*CJv1^94!NLKVp7;H?-7R^i<6WF zlY;*XdRCz8nm9|_t!(q`2+}v$8=wi9t5U&<#d(@U$9B-B-zz`jg-Z7P49)~W^=#8J z{#jng;svVd!@qsB<9Szp?s6KC?+CoHc#|{M0JgJ5f<9qsSf}y~$?(X6sb;70G1P-! zlXZ#8UxhmFpvwV$1Q<*e#ePkSD)k<9jIfa|rwe?W6!(N5YF5~)jA!zI;e`W>FXjap zs`V>7zB?Jm#*QN|T=^_m?C2$C^g5ya%2>^6@dMDJ?`B{k!w^g%-AQ&Rn%q!~qzIVY z@dVP%=|h3N%8T*>sIn^!!^6c+wwmuIk1pFJ#Q&us&i|XnS?kuK`j5#8JD#bEXq;I_Gah?(zsyd{9%WdB~2pSs&?w|G;HyizN~ojx3b-Ndh_#umOw73ea#L7j-NQ$6PvvN^VIG^$4@rs{usfDO`#uIFFir);i z^!kNzbF$>8qri6DC(A*Joc;P$SzF{z25J2}m6w)Y3cR3Bc0Nuw${Iv*5%E)v7Yo4x z2EjI<;Cd=IScd-S|8qa%RP@IkN7j>EJD!T_O{StMimjkX85M>7uNkUC0li=>3TRl` z+z`~zTiAXPw6@~HsnNd?WHz&1ZZ6s>*VYTWl^guyLA~{eaOeDNI03clSg0Q7q`mFd ztprgTkii>qz?zENZfjxNsy4@Rgp4<9m>%OA?TH-cM#wnP-RY9Y+Ao_Ifb&T;+Htyx zqg*^6*v|^-L#Q6UMNl0&B*Tjow#K5CpI;GZ&Ur4cQglL9W1P5eqZ?KrRk#0+DW)dds_nyy(`D2`Frd`-Cm+!AEw zDE{?ZG2->Y616s;-ZcAm;3wgOVH^F6fkihKsvPRWvF=m5P4!3yp3f+ayRAUQpVH6_ zRo1jICeIRI_4uZFm5bf3dO0du3})q^I(BxAWb53K;23cUU5PT!=*70nfnOUtjrx#p z0jmydH^^h(>;&w{*xPNk&C8BVOW(HkCq70yvf@v(*O5$qcEpbT)oK$vvV&qvDN=+? z%lRI?u0zT^Sw4MYaUq@K1th>I8h%hy8<4;KXSvBQw^;HrU30$M(35H z#JbG*@U#eRr*f}4-8VM`t6yvprkw!3%xsEn{F<97dFRb%9=kOt9CJy#YlX*2Xq(TY z4@n-$`XE$#O|26>2yYN|NV0t{FQ}wnvoW-or|&mu|gdiu$5?2-FGPVC~}|>9Eu?> zlWXBks9j?uCyh$#TciLHPhVRb#ne_XrBJhCn>DYjy>dBj)$)IgKi}U2zB%$d4 z_pg8GY{bw0uYA3QtbJkdQ)GglG>Y9q!6-yw&oolkRfS>LP1_GmC6J4+d?}ShRo%A84J2jZeR}(EpD>D5%;d=o1d4aKU}PRY_DRU5=O^|D~tNqE;WMQNyIeA6~{sbl{26Mna2 zo?(kk?vHewaF+e9BKrmo?R?%V-q5Xxo}H##tAt3&(Xc*(dLNJsJgix*yfGhLjY9$H z;jOBEL6Iy0h_px4)v(aLF+ZP%iZ@BMQ!|t=p}I}C@*#bDR@}VFyPV?!Zo)_&V=u;i z@FWlNLh^LS?Dmc3UMq*YhZ)!!dcUe%vJn^xA#|1Oi+!9af@|raiZ$JQPD) ziL<82+j}txBJJ1f+wtq~t@?)1L3yX>z%6ooyzYAY{rkxc6E~!dV$V^eg^B_#cb97+ zbx>Pf69huh&~6MJ^nz>Q?eZQ)qGPWsv}gc~u^}u)c_}bUwgM{PaHh=m=@MZL_vkU~ zmmhLYV6sB6_PSe%1__(i?#j=3)(cy_uooVqZ*`(Z2>sRsHM_;Lb@I-L{V@HGIK+o! zg&6iwtoq~7N6!6j2q{7I~ixE4lz2`(;PwV~uvT)1lC{ z$~t;*s#sq6Lx&FlCVuTVPH$*4pP8Pk6*I>)+6~ia6;lv$Zb?hf3gP{QXlIgrwT03h zf^5NRUZN6&+E3-NpbwG56F}1 zg484WzlKXk9Pox=MWqyr5l*p{{sv}_6kyN*{|vj(irwPfTj)W8St86zmdH}&$)Mqe z7pEvU1f{}_T4ddTZfD#<Zf@|n_o&TWc= zD(yNqykNcXbRgV?jn{DPk~~c}q&SD1v%T`9sSyMI+EpR(;d?ys*%_+Ss^^f}WxVy@ ztmOAQ%$cd|HmStHHPu7Mv&GV6-xjY}&-1fydm>8@MEbWmemEP$`dHPX?enN`LJkJ) zpmTFj7mb9#Hd;-MwF~@bB!j;)_8GR==C(%TU}XBTU)F_~S8Bf?rIj2t+y>$X3_JBJ z+5*4;y$=dSy{e1g>=uiU%X+7bPSr5@JYZxCv~WJN{%7~^nO9bF6f4$yKlVg+TJ3G& z#vpVQv}Xf`lueGgz*>ux1|tr6nmwMIT$*4luEl1oDe)M1yx1oT3C_1+#S{F#PrDW7 z^lPf|dEOg;bf8F83W=fG#airWu4igMQ3SKiQ!ss)Bg#^^n%`&ZeSUDg@c1_^3ZuiD zIp=qOBq?_6@K%|ufb%IfhXO*~sO>KMm+bOOefh1lSFXr<6a|8QSO9N#0U9`R?TyZ5 znC`*t`%vU^{N=a)mcaaF^s?5-?I4PFF4C6v%~G5^v1wyBYzP)!IH68mSS~kDG;`!` zF=iDbE|P)!9wWZHXv6_sz4Fl2Q#?=OtF!Rk+wd)q|7c$C-hR_Ui`kSkb5Oh>z9FdF5|1C`2_KNMVwrqUK$vC6Dtv&x|5@*LVN`c9YvV~$( zD6)ZyDv=%3SG}hC--Fzg_G-|%xv5ZzVyv`=TG=j9T|m!EwwyLh4Tpt;Jvw!m-fY7? zeuiwNs4~FU2#wN_AMPPpFHCU0#so+e6kA4-Vk&BR7IT~irohl4?=Ddv9T!sMQS0+b zaID+oVBmRYZn`3GdxmQH1#+3e;Ou(OzNKxEIU3xHnHd+dJVVtS6z|doGEsH%3PHEB z)$J6UYlEsco&XyKn)y$wLrq-%Q~J9`$YfDN-ymn~*qVG~0)i_P+fI=-Dr)1*4yOUH zDsS|I8pwcGy|=bQav&VJxbYP7I$?BUx8^1@;(#Tw$RybP5^8&6c5uW2^L%(^h6ZXM zyXYgbcxWfR&m{XId2%70=s4oA0*F#>&#DmI_m7{a%~i+D1Deaw-QHM~U*h*jiN6Cm zpR|=sI}3LR4~93ntrv_qWRr)%SYg{n;zP7&1xJJ_%KoLdms+x(vJzy>A@q6sp9?X5 zHfhnlB~@-lfPKF{{Q}u&#{jD|0az}@f}4^~MK#iiOcRg*BBGG(*g#WS?Gzi34>go{ ztPqEoO_`}h0X8h1&I-YHk6PPzqD~YgpF7Nxynr7{V*bCarRFpU&sEltgBk(t4%P-V zsB0zV5@2C(mq$aECsp3B)K*C@Gq|fYsQW^c5RC`0QZA`xn?!pBSjT`mzX?*j7bhoZ zLet-#;#4TFh7v04`bQH*(l`cXlGri5of}E_=Z5ZCXtXVVs*bEDr7w)XbJD~T9imw1 zh2BF&MF%E+=_<(7=tA2f(^N%*{Vr&1M%3r^4P+<$E<%CACTXuLii<}FrZPEH=4&^X zZKO7Rt8ZD{Yn#5h@h2#3a|E7I?KJYrRhcpCCqPm0Vriw~ENDsK47nR!!t@6g1NVt!}oJ2#1k(eSkXV$Yjos~tNvdreG95yiryJeP`Uc8wKcLuw&?me!&l z+&wpcE=*03{nyV*MXF(a1g6Uu+Q?Frn60W2-KDjC;mEX(Nr5DC!m%urKf}8D+vM7x zwB6u(@Fb6OPe|h=l&5tS9z1QHP`2Bm1CA8&CAME#EH$v~8`8yHqO1^finU3zt3%h= z-pw@`G9JJ%xzE-+{D5&LIH7kk(Oo;P-{pwz_J<`yIgw%3*%Y!{4c_n+)uuc(@)knq zsrzhwU?9h1k)v^DY5YRnKgbGxF0CDd0+NYi>T8oJHi;q$P)^63cg+fc&SlIx8vs$Y zpc7gZkek+uzJx_1XCW+XM$g0bN2+c%{X}4dg@eCC42d1DqTh}i3xQo{%;2X`>;{Ug z2kv#|p0Y*KF?SGpML5;oYiIVHfa1ZgVftV35C*LFWCYphRKOf-<+|%XMBy z1Avp%V;JaE@JOXcm*^~Ma*g%et+X+sEH)ocEl;>UeCr(d8-6Ejx)CNTRs|j%!FJ5A0*%NkvXcs%HJVmj;~ty?HGopU9XUd!aCkjB!LZQfwkc;;ASss_GjTO+3GP+yT1<^V3#ao{S(9UD#6flS(hOAsqqXfUZTai{>?<{H zP2%N!{;j+|QTr`gJAeIOADY+6+wIK4QNVxKqe=A7tFqR}3*W}_)kI;RFb=9*vq8is zj>#97OZLg`1vZJA6eVE^ ze8;l=)WK!T2g0Jo>0X6icj?`VYNwBsmEn(tTh%KglU~xv)(MV_w$0ukX=FZ>tP{lg z-=!;N`uHaor)e#-Be@V0F z%1FI%93mH|7MY4F`LJbhG)C5(9_CQ;5Zb%-T z;qn&#-{cIIy&{5_9hydFJAL}R*U=tRCqP>;61PQ*n?%VZQ=Sm98bYUgmD^lVm_2sh z$*Jir8~u{`Kx6ctsrbnY8qwsBm-kN!!ld22F-t~6*8hItu-nOyj5=i*mb<~M2 zx&wtmyL^zG2IW+&K^S=0 zY~o5Err0`)R0A20V7IIfKKdq6wD=&Hl7{f=pkasVpffKGI}8BjgO-_x|m0Y?RMqG5%2Agl!xYmmw?46 zG2$#KK;e;dlAr^mwzJd=~k8sbfRYXuUZfz>=$&A=sCC#Qc{a>I#|j7Q6Tn?);U3_BS120{6u4roxv9y#Ef>O6YnxaX$Xxq{I>ynj%(-DPw?T6}(X z4t-UYNgr3gD=wfvyAsdf(RG2%BD`z#jx5#aah=F;b*6u&e=GcL?=8M4!Mp02O2Nt* zCjI!9f^UpHvEk-tc;y8?>id8HbGLaNgx#7Xj>-n0z;*3a-c~}@0`O7lYjCrj`<0DQ zYTd5@E`HUO@QwZ;<}B?1!R~aQ)qWYOCf6Zo3wa2ue_I=^3B%;G5w*z{FPKdGPG0Gc z%yScbTpc@hm^m^Tb@F!VR%ng}F?88)SkV?)tGTqaTY--B5lzQz!@1ETwnot?1bCxq z^hO*1Y)L6R?P;9DetLG=Ig-k6L1o9T@IDh)xR_!KDUwG;ArV!#vWG_6L5vI^r$2&( zZHlskZ1UOSnLE3QMs2$cPuzyn)oj0Fr$+*WdUP~$&mV&2eT=xAd?={~)?BOhemog! z-e4JTetPmW>up4A^zAgS$+X)!hok=hB(LFy%^Mp36ElqSOr-8l zTJo+Puaz=QW{xC^O`ynHDoW35SQ~I(Y^d4S7SQP)d75~3s(gsg2FJqj|Lle-+TtGn zaY{v}zcF;-&*u3^KEA0P8x#%}AVX8(Nq4B<9OYD~3cKsRYc4RGeMWLNB__mx2n&HU zYDK09M$@MK`f9^3oT=P4G7|C$W#GiL?8^gHPu8NMJ0e^&(X%h#-Mz9mA)X`9{pfl@cy2Vv5s z(#D7D^#OW9Gc-Gwy(=!H5BVR^JaW^%uS#`a<(40k?L0)|6&b!jAClnp>hG^DI&fJd)FjrVuacoK=JfXWK~7NoPX z#8LZiv25H1?a3K_mTmP{UjLSPr!_~5GS=bgHFQ&ySW2h20R~YaOJ3KLzt0*QmcDyy z{4PGv;Pvk9w-oBs;P+QjMr-rf$8OeAoyOW{>5@$3H4&-!rH3 zcp4w}J3$gI!VZ_C8l=l@P*)3K<(w$oProZJmLfT3@BdOZjKY<3<=qqmn@bV!S3RY-6}By=nfY zV$959_IX=#m)lf~6Azx|u+ig?d6gYUF5dCp2A+eHXx#fwhi%{f-DPN|Ioa5ceg#$Z%=0@94R> zFU1BNRcATgB8!_1S2RQJEHG~EqCvAy}%>@*K)NNlId#mG?+iE7uL}077on>)MaS& z8B4ehPoOyThy%(c;T`&9_g=v|X#K~!D-;~vuPT?s`0KnN>gV6gAW?IPx(@o?@PD;; z9c$>duzS zZv|#8+|%B1-SqN$fwq&}b{SEl0!qFn-vjfQcAJOH{DtbDbB(^y{hzw7Bsnv{4;o7Z zwUc5&aJiI>ZM*qvCiv{RHfqx2F*RXUze2i3QW3Hm!U;go0hL(D%ei-fPM*LV2!EG3FJJAJFVhyzFL%GR zv_sto7RLIrMo$pS1N0n|<&#HQb>sZ0AHO=wh!(%!S1%`9_@Tv)o6`1}pk+J7!U8LY zimFq@&fhKFExoaLw`8^QvFi%q{e^vV+aeFCZ!L&nqk-%!Lv>zVItWCJD^&3{}RL8xySPa>65|Ys3vHKdoES2!c!Pg zytPnkxWf8EG?RgWi3u?q#>wvTn2p1|TM>XkG@ZFql# zO>K%{&?g3|$pYr{@0YFpn#4kG-%{^VZ(IgEAg#e8lD5d`z)UI7(7lx5k{6nxs#7FM z*GDwc2k3Gjyxh(hj(00>y5>#6T47?Zm-_toGs!D>eX;gCq5r1i=f?>v@^A>GPD$$`GR?qWG&eo$PpF9Ffx|jz}%V0`u5@Dymez z#q&XUgDB1Y5d_c6B?W?81dm->TI{#l?Klw7rOHo{M*&q#x3WGQ&z1na-CYo7{rh!b z;LHc!x`)9Vh3_(JVgCHUZMSzmy$gJoo&LN1Mjz7_`6ytu-%XM?_rsSDe63@dRiJ4x z-b}pQXYWiBYWmx2WerM)XS~QZU^p=nlL(9<3&H~8T`QUtS4FboaP72JO?8*%|rG^`XDnya1g+*(o^g4b3%HA#}(LC#rnM%Dt1;h!iD`rTh>mC3^tc9tOg z*%JqzHGQTv_P7!L7ngrYo@ulrH-FRdce2`!+faeUY0R{-m14oq-Uzyi5O0UfzUpys zC3D9WvrgJQiX0X@HriOM&BlKBDXk9XhXZSjuw~ES#=$u2*Xc)dUslUTUD$uV+Dg{j zF*eFfu#riz&_%eFic0qFr8DUdKp!zFT#Hgyc^ZhJb}MT=hLqVp*Fv-(I-gro2%Lf= z4%pU?XQu=y{17r~Gb|hB$y~(`AvtgNI+?RSy&xeH4qm7_dQhICyukF)d&0}yx)t@p zXmO3CIy76l&u5bWsRqx13)>z#Jh#|8*%634#MQHGFws0c#1jG3(@(Mfkq56+KIY+( z7X%_4!IWgiK-_-Ez00Z0t;(}zP7$+#WT^Hm8StDED2yFMOJOngo~hV=#YvL(@5Z?$ zHa*6NT(#r5<*~`!@-fAJM3Gxm6c)P~DAQNX$(wuAwNQ@CK3k|YU%D+ckW>H}26`dc zFstAtAE^%m{YkY~OVGVQ?M>HiWg(L{w<@elgh|5N!UEu$%ct)IcYuqE6^1Fw&C=4a zApu@B;G86^QWR)v0Z35HSnaje2?#cT^99b~-!SCdOWzROTG-}!FRa-WGt5}7hs>3R zCt;$6+ZI>wH~NOg7HMq7*!rq1rgJ!>JHTo4Q21WIdufx<+twQ4;BiEMks9 zwL=!vb{Eq3NhNd4u^9|m8W^620CSTv(ShfOHVX{VME?-8`+l-K;tXw^&hCHD_Y{fY z7pt}7^4TmC{A{4u^%Pl$jF@`npi`tQ6f0x(9EShP9b^KFSHJf-%REN=bU@CIhn^## z+%E4Vh2F?y2>lu#OV%pyy5AL9>kBt|gP*fwku~_7XV~I2_uSL*vF6hA?t~l)SQ&yu zma(3fKtiTVw9YjFSO%NuBTTe7LzPLFLRTZIHcTNRECRxl*}xVaKxQlv#32SUk0$WL{vv}sD67q%bP1*{4gR=|2EMVZQE(hq5! zs7J9w@c;^Dp(o58u0&#<6lJGUXSnJN5Jke}&8|Z(H9^<)JFTua^gnrmksqL*AZr2} z;{%85C_g}L{@3oK=Ec@_yX4tp0;nhD(_7^E{?X!nFs^7-w8tEE@=O{P$VNpMHwGP* zq2m!dZ)?Da0~#lt9C1;*c+AB-ho9^9k9T}>P#N!u);Y!tD${ju$N$dUfwfzxaJZz{ zb{ZX63?cr}Zu&KfZe~g+qS2wV3>2eRPu)*`pvV<}tH!*=-EKj_(azo&7CY~etV{Hd zxBl^mBBlUVv>z?_U_oQp4n?&e@JYVDYi1?&-jM-Jxt!^eh5@R+<%}MFx*{IkrdFNY?ungcM6Lx~i=c zHG=zbNzmtfP}86;2r-CA-(Pqb{9y1<^n4+p=L}8h`E*riy*I>bkwPR_ulsulN-MRM zOgq~l>2=Z8dt=o&Xwx7KWuB&4)~-0k)bLT^wHTqg}`r+D_iY$U(fyzYvvSXQdp5+t~zEO(8PhmQths#+qS7bek0zp4qiM6+vmOheU^{t^AH`epa z?6=?g2kv$%qs6%Uj$!vfc&z7R(2CgQx8vov{xSYXL0qM|at)CS>_w#X*Q269kc?t7ov<*{n^9;eOH8_ore z`Tm=n*8(zHnL>Fyi^rPDO{mm9?X}6dde*Q*qi|4m5AMc$sx@20K&m6N6ay9xiSaTk zAco&BwD9T+^j|kkA4SZnz@v~+=QX|AZ*?(AH$lvLid{#M)l?KtMHrMpUOc2vx#Kj2 z8dhUQ#S%n}xqQkt^B(^4(^rj%D4WaNBXuu~4Y^{1h$f1K#-Ni_)T$XpKzR{2qXn8~ zSIvnI%z&bm3$j63pR!kZRdk6t{}Rx`NNNMl%&wB+l_f8wLG~HBd@-MV)3pn<#&x1< zFT5|8G}1c+8LErSO;^w-7wz=e>7f&KiB>C*cn&KHVXc%9R43Po5}4h7-HhQk*GS)2 z#LZ}p{6x{D*i9ag8_)_pAuD7Fq@FR!SafP4IBeQJi#Dg?d9D`U9K1mp5jz9*bYg-N zi&K7 z-WzB3k#1!Fstf3ODHBv(*SoJ-jH<85;*+5oly|vk6ND*(A(w8|2T(!>54F^oSPmeI zOvUG4WFg?V0W{8ndsw1qT*m31zEWH$;UGEY1qK z5xE?F-hsLGOg;_dv`{xQBuHS|BC)d(s~yg2PLM|s!5O{Xx*rEmoARv5=b3Qkd-BGY zGV-7Irn57S5cYRIFgHWB@0Cl->Y&?Hn;`9@hg=Fk%D2j+Xl}b7>H*hCQe7}p{m6NW zEZ;=NobUKek#1e640_NG*m`cDOf-Mz2FY{Sk+LWjxzM$Fn;WXCN}CMe6GsL8d>p zL!)9ns(Mcfd={bTS^mW{&P@hZyihdl?U|=U=Ik1L^3HagVFZYcu|rS90rrGvOSh6T zw=%cHB|~(!)O^1+vN~v=&pz)oaAnU$wne5W4f{=!t(sI)9W+FPXg6d9yPWpBG(!n6 z_80Avwb~+SZ5@bB22A7MTbcNU_nnQVXa85e-a^*evFRx?fk7I@ZlU14Miom_l-pE~ zWckqYbplp221(RXD%+<+Qb!j{FR}gL!`6f5p~Zo*@n05?qkPAk$D=28DL<2$(e@l*zti8-)B$e z`VTgy{n$L@$`PwAw@p-4V zpgQ7u@4P-u`<<*m-O~{mXq-`Qe(Uy5Qe?+7N~6h)vY%q1$D<0JS>#FRS5&BD18@r! zmoVXDU^YMIogmx_bra35*btQNb9eSxS16oqRb}|L27Tc6f!ols7WxiIJi%@Au@Cu? zTUltz%s6V5TOsH(cZn+0ost_$tC|vipnitGsz1dAQU1qyfqL44knf%s7_sBv?+`;` zXOLWzAxfdx4HQ{VMLi_Bf^xdR7gV=k!tavw%`K$6L~TrI$UWy;pGhqT`2Q=XS^WAZ z@a)Ol!t3?hZyqeiVelMl#D(+2`^XO7}hVRb#T~OzY-)BKczaX9+aflOt;F~hjs^8^PgJ&^P#`{K}dYbhP z@Is3H`gJG#ue6$nUSAN8l_OAl8>C$ep$MUSZvNZ>|8_a}u15n(=~{Bw6Kk@uwH2c+ zQ)m(SLEL%}{=6q$PVc&Jj*aInERP}AWZ!PcPA7$T(bqi&7PYvYg*9%96166G`qw~X zkd4)N%w99E(Kfr+mb<(R+VqR;`G9X3XPA9I_WLP0WyfCYEt9FQonqT4a*m416&!QM zV)_PY6-iLexYoBH4~E%c->D-gth zsMDs|sm_MORm-#`Zx{~tk8P0q{OPrhk5fIR6NNT2Y=;%(|`xDzY!afN(W^4UKA zDUH$=uOm=v-ALmf(&Du`yg4XEuxY+d)B;UN7TP0T=o)Xz-&{9%N|XjWFIUHdhOpD|v2 zq}^6vkrET*kxsFx6xmEg-E}OU7b7l0c{xd zRdb-n9SjBXicVc#hi6shym^bgKOF3;BtMf3z4u6#fqyCR?VDF%a%8LA?Wi&k4w zPr$G|dv>xIGta(|8&UJ~zIEeyH!MfxIl!2qQ*4dUz89D!Mb3lsZUbIbk}Dob?lBNk zMh1kbnOO|OzzYV?9yzO?<%Yp{hNYt5%R=)Y%m2^bo4_@7X8q${@rLBVkc~j@EucUI zL98r>DPp5trtd7j&ibF}%sbm}I%Vb^J6mTab!Ix9!F>f$!38vcvIqhSxS_08#idoC zsJP&QiD)er!G%(V|2aue5{c$Q!tdI5>Zj(ePvU*P_nh;b<$Jg|;CL)@>^R_729R{ zU`mI-&Si4QO6lyOSO`LtQt@jAb0iBy4c^Hj-Tu%FnUSd@dQF}%0}|r8Ce#M>xnaf{ zS}wUEDM7l!q!rdiO=?U_Nfy-8Wid^V{+2B2bIXubON*r&12+a94BILPD#9wdL}1Qd zIpprlummSj^JHD)lx0u9ys|H9sy+@sHk?cp9*Y$_nJ6bfloZam!px!WsM3QLic08p z?6s+<6o=sx`wotzK!N#j95VfrIWnK${obx9e~U~pPLI4u(s}H-Ew_>%TPYS;^mC~A zMrxtw!l}!co4)Hs%a~4jfnQ4ma8{_3WOK&dm*1c9hhEiYf(PIK+rRRm7N=R8quC$rH2_Wo?j6tjuYWIih-~71+Er##B{yI7L zjA@|Htq^pBVy{ueNX4I=Ruh7yin_DY&rUBA76}`?tDuj%E=^GrqPseFp&a>V``k`N z&r0*z9;%z~^|>F=znOyIx-!+tX{YH~AA0C2JnHFPe#qvOI#yT7!iAM=cVtqSt}yE4 zG-&3LrYIElK;8<{SDLR|Ox^cB`BD#)=97c9%L5Oe2FpW24!h{y_xL=p+AH zdD+(@fSP`}YZh70t>~1;#_?_|(UecI;L+Pm#aD;zk#F!xoUB*ZsOGbfMvKMN3E@>z zUDo*H@`HlI5Gksm>&2TCrt^5_mWiExYiOYYtBmv-qoNwp?ObN?ID)Jv2?)nd-y6F~CXY>ui_r za6#}0AlxVVNq2A)FHJtb8gJVrh_@QYPTwD7MoOdpw1(=Mk^19BZ&ti=W9Igs7-!`D zX!)yWXUyLHZOgf~*UaZ;oz&JVI^0i?dgY=qmEIlhi^nu;3uIg9W_63SVp2vhRH_RL zg=;4^Yx~b*%&~tvA-rO$GbD@i@bd^s&WIjx!VmSQ*s=w-c#hNwQbt)a!ySn%pX1!VjFbH$SF{jSQz!_^M}2O-pd`G|ck zG2Z#jv(BBf1^7NQP2HY$1{6z?Ca0O54=Ywo#@1nnmlhAS7*d(Jq(@jC*yA-qmV<+H z{!nn>#z;&&C$TZ=Wa$5F{k>(;`%=+CfAR^B7rl!|4@`dTq1byAxlP3*UquSg$lU;T z@*?ok>&l{gn7i~Gi4Hq%%#o}T^)U6y8j&uKB>LphnX)s|m7*I{w+rBWg?Nr6pUI(* zvFUO3>`C`5S(h*?a5?*Nz(?WxWzbmJz*Gm;LeHSI5F9awUPyEmVFyHYlaon3i`OoV z-7o8N>s0E45~KE;Dh+j5ZiroX_i2(vOU8DStIU1~H=0Mt#Swrs5E85~jydoWT|rPT z-NO`1FGVK_bQvLM6#X~j*frp*?f>%(I%fybS)-=({&=x%^TWptTpq^~?Q~@>QB_P! zQ%NEVlGtqv@Zx!yzL7;KlwjK&M5oRYE$&79CwRks{2aKzsUT- zVn;j%$0jQmT?)mnq(~AKe|lP`{GINYPnkdCRj&E1*IIhI(4&jm$ z_F=*N*X43A3oPDjA#ahDJXToStRRp@v9Kaer{aOKobFOV0lRv>vVqoRMx`=u8JPCy zw{+RyD*ME@OVu6?t`V#&MDc+n8F}9tP7ZZtX|D`=SLS zKfT_VPWJK`BT!5>fNbd~#R5sf0V>|WZWmzh)K>BOuU&+8b(fj5(@*#vjy*&aQ3qDbf`gfeP-vYpS*HW zzCNs8yv6HqXpQC=T|;;HLPK{UZUZr;dyNLkM(@w~7%J8Z>5tSkiaW8(HN((RLkE*B zxCYxFy4>LeCaN@z`RB9laxpKHw|lf{bob-7dteRgtbI_?feG8Qu#y#7_MIK#dBOy7 zM8*?Fa_%Wk$Qb?3tEG>%bN`aKTO$Ku5xuUcQWwG<9%aMCS-cG09uA)20M&CyHZlBHx?NGg2 zMH{`)Lwe)6O^SI`4qZM?ug(g*6m^;83yx~^nk3n|h+87OGm&kHxD_)Dp*`4Ba^z_b zCZZodHrVJxKgQ{GfCV zO7~99pbLb@$n`1h^uLN#$;5e8U0f!|AxLl;&`H;Ff&>*;{C~#UvNZ7)5_agLR*!A* zDDo-Oq%-TGvDc~B&ip{vBHbH{n_XY4odMG0tXH8(w^xJNsp#>aqy2Ea4%;M{8R9BD zs=>4BIvL^+QXW@x{YRXHi^qrIgp^U={&B9yvo1VxG+m65&ckjySih4Cj7zSZ_{Sf4 z{}G2E(SPv>esY6^w07?ILoF_nNg4lMNpiU*#`4#i0_Vv(%tpY-Q6ZG8=KI!)@bccN&06HIYZW8APC@Xl z@D1ubu*fen-2|^h-R4v=<&)b(XA4TmuqO!=8jB&sFBT414 z+ZBYu0Jm!n#X`6ygNnaPu-kEscniHhwlZSn^JzgftsbId(3QgZzljjMtE ze1TDX%QM%1f#0HWizNrcjEZ~2)d1mw2`ic4OaKfrmVHbN;-07cBNv`wEI3)aGVBal zGgxGg$GNRLtgw?yu|S%eNyTrWJ4Jis36cxJEfHxU+4PQpgE8sY9I!R@>=gS{Md&bh$>m9}Lk5otNrTL~lMdMj&yTkR?^IZTBE2R314-tw zKq&z9^0{Y*bf>#f&0-TQ{MlHfJ$Y|dmS^X zXgC8N<7O>VQ>I4Mi)$c;+^kI#bc%Z2%IIFV%7|=$S)1w$azpfy_ULfzZ_1AGeZ99xJGGR-pKZVvka!4ry$SswCNJ?-Id%&lYJr zq|dG*4Nx_B!mke2!Pk^4yqY3(`(iu%?=g!4ZiUx_-pUBU(@7DFqv};v(!|(n%8foc zGYci2fF=#TowWY>T@kuM#b(k-pJP&)1WCKJ&#lzAQkbUbV(xoF({AiKwqWA?a9xJ) zN6{b9SD;W4i&cl@)^Q<11^|XTSU733(SL1!y-i@j(;tuAJxu0t6G=SQW?5F~TTQXa z6iKAw&#Er6x#6v1BgqcG$Sw*w?e(#Dx;i(!Q{_}0eO9O_qSyk8hUtbdb=$=MR0v`t2$YfBHSK) zFbw5VIlThi;BmNSoX!$^9NAk3B}khd8Kc6k35OspQ? z>sI~TA#Ygx=0I^itsY%!Wif4ZJEWyt6=G|2-98QS(dy9MQ5J)F_om=ZwP15FMuE`Bk9jP*2IV;g zo^|Q$4{_p67~nXUoG#}iOhzrAH}?2%Eiegt`?=@H)h~@OnQv9le3xRoDAGa2ADwYg z{_iWyff;vHxpXpVjVz1S!Tt%@$9iF1>x^sUj%Nw8M}E$|n$$~IN9ZmyZIKPiMUoo~ zX4vZR7@p16U!q2>S@Xg>>Ajj9U)@|un$kiZmKnGwyj*?*tZoByO>($9G z-VAZ646_K9A&D|hLTP9hf2jD%tG0Z|Trk38UyB{~WV5AX)?PaEf2mA^5?|-bjJeRP zHnfW=r0~xnccUVMZc@*hHR~L#uxeBbCSl2|`D7tgtZMW|?iM(0``@rYH6*kQFnyl1 zIW(M@?|OB~#Q(7+>EZ2y$qwyKSxhQ}y!qy-22-Ygj)yg4Lwc$XPVYl z8K3yREeW6<_PiX3r|t&h0Gr}u&>KVBRQrNADe7K0#x_Ok)oCIBZWH$i`+}RIJ4TLA z&A~}{_(~1u!E(8m(Tr9cON-4Shn72mTB81q)4*a-n+06tr?Y zte{6ui1iB=N8#Fk{YZUs5@w^dFXq1@v8_dKBI9 z{Wb`)NfuJbf1V4UNwR!FHz|uj(pd}zAC?~!6;l7}WuM@`E}wbU^YpU<>mWCvU5Zvv z3aJg~f_~bYh4tdRfO?4MG|;%x=$vpXJj>H4Zr0+J+27w7`l+hJzm$34qMj{}gn^@C zv*Np4m$b99`H1K4+T&0LG|)a-MX@U=vYd)HHy+HRd%QBEzUfpudVxu@!Jc>^tvq=^T2PVe|JD5o$qVj zKOMBnYvC9jwlc(ua4r{MPQ@O0sFA7=Kl-TgA98 z;UR)w^JFwSh@5^~#+jc_1R*!!6Y%wrQ??G3FNyH6<25w%M`A`oGkaT=M`Mz`x%X+4 z`T#NSCrFY6P3kAl{}U&H<4KNK;yC)yFV70i_9%f@$H{bfyxg_pN@dGv$e&7@Pg!#&?GOLi!rCE{po zCKL$uM%H>XYgfk9)A%MK&f5qFUA zDM5|WtHL>wR`n<9TslQ_^@T2BB5>IBcy5_AXVUEv$^Gm9XLEX{{Wp!kCpVeTTOa;6 z|M-KKMq9+o{6*mhp%8x%#uXmNhJkQrfKzS-#V)7F5-J{(rGRaq8?;gBG@ZGr`pB<( zMvZa|z9WEFo3F4B-*>0BMA`OdveWWtV{C4eF*=b&j@FK;YgrvSozjC0>=%NAkan)S zJ@u$&z56m|w~K1AZT?NT>)7*tBxMH~!X=07KksMib^G?1Z+v~CMO?VaG`mR-H*18) zR_zB?f@CMfLh;WwDjxgvtW(^G%%O|e%MqvMx_ z=m@uD)b`Ls;bWonT+!w*M75a?euj2%4o=fc81~c5VBn1 z|L8?{eCy$$PT2S&^ZSQwlaqM6Mz)h~)J9i{6WN=}vtC!-F^ZBEc%GnFed#L?zI}cM zE_8HNuO>z8iD`-$xg<8u6Yhk`bYuu_@@~}C=4M-+!&`XRQHICZv%<~tUR5e{Ote|P zL$w9UXAXt||B<|ol#NZ8is!mig+iyk?@yizZs2(GSk64l2^>^Y;#uj-7M-(T^>^MN z)!cLrk5|AgRyyYd#ePVUBUF59)B{nStZ>{Kf1?&yQQ_S2WHj22MY} z+dy-1YTP~lX6<3WHd&A7`p^b86V|_V^d9#kFJ^{a4Eva^@@j;5Z?pD+XgBb^U1aK2 zDvv;-P_- z8~z^GKn@%k5-dlj$!Gt@f~Z5U-wYxKZiwQs6*JeWxb!x~-lE7&Dt?==McPAWPp^fl z^c6w#O$kdFD}+d4p#UlcCSmbrfuxMCjL;n>XQFi%1ZFo1RGPXkrXKzIhpyB&bNr^6 z8)oXz>d%nrax~}=-V>fTDRaCol}RTB6Lt7?c+Ff1T!1@xlYueJfZGhrJ^ZMnF+B!2 z{h=dSBD<29zrw&QeIbwjI9g}^u?3Q3Ne+I zUz6seSEC)a$Fo@l{e5{`TDovq5N*qx@wnrc$7^jnY#N8vnIeo6WdjnbUw{}9c#C4w z>p zM0bo7L5VEJ0d=RMk+8(Mj<}~yAQ!AWZ7gS=w~v(-V}G%GtOXa>{-@=AvV_M;l6h7b zSWmHFPOa`I?8=n2)7kV~(&L#BXTm`;=|X;+U4GS~6Mpx-Vbi5@V&Hsn!2zm;f%m#{ z+dd8kF-bYM&?Hb-sZ6taeb{h`@t-#DTroAk&GNLNhjsqlOj>e0jji@D<-VDPAzf{yA()mmsMV=0Uby zT1Wy_sBme|)M3EIgq&ev;t4lr)EwU9U?u9Dh)Yo^!k)MkXfxXx+pJ9qTp4^ocvv-$ zWC`yxS;97H4bZ*b1CGAkfnBN-ewCVvNzK~yplrF3tObgz`%LwJNcrOb_RHV@>38vR znsp--HHet!@B4RY(ZBMDMY5E-Wqw2!@z^$j;^_hQ=~{|Sp~y-qewQCQM0=nk6go@L zxr(J6g;XbLP@W>&JzSJiYxf1@AQtS7;ffQSER)f_T>*)`>C^3!62`ZxwTrd38);-UqxP!-jH9L-bYkKld>37L~g;?(O87sN*kDrpd3l2D3M+6cV5yS+N`Y=<F zE7DJ18-9ICFNC7cLu;^1_bP$;&PMmS3g<7Chbd0_<;q}FIjm!x#0mBF-@O`SOC)QD zfz`nDFtvhCd1_QybQ`@{crUik4Ve~^J+nA^3Dk%~=~e#$%(27+Bl30*Z7PQKlpVU6 z(^T+RFmIjL#Mt^xcx#>5A;CZSbyM@?uqpCE52E>`miinIh8&A9D7J)P$fld6BYz41N#tp zc=v~G@Tmxc1+!;Y%o^dNtNPEbm0gq@<$87h-Q|?|%C+o+;5xehTd>7W9b}i7zo9~0 zJe;W=;+*wM{il37BwYr*h})d;w&3IE%tomJ7fP#C*#>h@448bZ!}2L^@n zkPf;Z{@q_AEZ|v^+PR$+aRU#3H8DA61)fTZ1()t_Dn3bedqSDRnn7ellK{E9bFTaKcUHcr3~_)XmI@lb2i+JrP4W? zX134mU|1pD?O88Q5?~kJ1pjmfD`cQ!{Be>JCFpSVVMwdaKBAvcX>PBfh{eBk%^&;yhv}w{p{&g=F zJ%L#4iq4)sH*E3&#g|?M;KpfEH>pp^*7)xWtyf)+=wh6^3UDfJZ03GAyCEd+<>xJO zr+m`i7Ll#o#7p$!pkTTik~mSfRkDf2{x6v#0F{w82+)xKX>eH|-lA}zOZ>yqNdsnboSVIo@)k{AOtB1AP2K8}Cy|i4`roHA} zB&=7}sPK+l_j=W0rp-kiGr1+|4rrvoR6ZZh2 zd9T|i{s$lj)+o-UFUeBI)zR2zvV-IYr;NMlcZ@8U)CM+XtGIFe0r{M9TP^h+^F{fB zj0u_0=yZ;xQ=Ackq`8LzPCMPT=Cl{o$q)QXvb4(n^;doqV9`mcS$P;OD*ZA-B-Vj>*%eB9Die`BXZ%#n+V`Dnw05>En=c)eu zCktQ}FG-q4+Ib9^Wmc6%4=A>WBKN5H{h@bBmDFUKK);c3^OaCign5+-;rY>hZnYCC z)t`8toQ72Vx_bA`AR05mGs1O;>2++S$3o9*1k<8&z?qH9DLjgmBiNL#CE`|$UIS(_ zE}$R+4wj2SPzQEU0m;$i2k#7piX*z&Z@#iwjiygqWRd_I;hCFEWkZw#qv%*I^qG9F z=o&dKT@Y3Y#Gr;dhxzj|CIw2lu(<05fWWN0Ml++u zL=v*D4-i~;WQ2o9c~Wbn4oiV5CMCeTg;A|akCi?<0@_sN(<~|3uF)OvOosQ2U6+N+ z?e*oYkJ*np*8HsZ_)URcYq@lR?1~Doo9~ei`B#M$>Xj0~Id-mOZ!{c6mbSw$t~_ z2&OUk-)@MYeRGzkZ>p|?pu)r4tK&1BWH>PUPrTn%C+*MvASvmi1)CP#w9U=?lK6J*Wb&;Lt_xWt z!X%F^^qRot=q_Qga8XQ#yh*!Oux;8J|I>6?%qntH3vNf|w4#wNmR7{R)ug==z0&Kp z;-t1FwgDWzC8}-H=8mnS-^v7+CEi^}_xNYZosC2XFgB=T_pyUc23y093~`WU;ak}E z#x9b{W5)}y;tp8KZl~BHiWE@sA4+j4^+0|S$bov@KAP}xK(n?>wf4m&%%OnF@p|>K z*iP`-o%rfSdA)dV>>NqG>U>xysfC0F^zdDue2**&DvMd?R~b>t+*UNbu*>5farXS1 zrp*~3aPs_{e-|e?Aq{CrwC$hBWdI)gO3JJNmP4_dD3U?N*Q?5;xGYZyHv(tj=L%C3 zD{N(SC9;mjkYleKrjC~ZQ9Q7|Bnj$OI|8r}!N6>AH^!!pbqSar0$~`adN_tFHYV-4 z{DUZqXi@!Lw~=h)=2PQwsQM!-oK#aR#1P7<_--~uxFX1uqg*NaK(jS8c~YMn+Bbb} z#_`yRdY`GO08(KZ=@eP{G^}jWb;XoVL-?3@;FoI83(BNT%DW*=%AKLQCbdC*Yd!(?|NN`D`&1ZTTay5N8zq9%0-Zjnbkf0?q_Aua zb~)^G%L9TRWKlp$jrCy-5hl03`OAj3{&v|s44jl9%n2t$l)ta-^phSbw#}`$C@%i~ z%RosDw2U6M!rCeU^0Dd{l0;?F9#7o@zgwQwf$Jc*?jNqzNP|GkQD_fxt4j|}{Eg=Y z+fp~)uA}V~uVJ0qGOqi%e|5G>wvsUT=YV%-`a^IB@;N|13@x`xgN?bobW;Uti8HRZNZ23*e@zb z7LV~!ZH12#iY=zdRw_PMc%AH+FkjFJdHgq(z=kSl2OiR7(LZ!0GdHtasCP47#>9Ut z3CLwz#fzdUr|Z>Oo_i-GMU=j9O3^{8=vMJ6Ae+1GSLsnhi# z9ttj=bPqSo>3cHY2!ZSJd?B zQQHhTF0zBiix)fjbQNI>qmaH-uf|n;pIfH;dd=Nfow>Q5fq~8SnpP+&F@Oi;4C(fS zbhg5n9$}K8Eap5U*MN~DoF`fD4y>nouO6>Mp1DCLepJa>P-2NuI&5WvA#*G=$DSn_ z^koft@Or$k5Ji_6)gn!Ve7IERgwUY=#D9xdbzo*xCUl8L9eJK)&>Pg3HBibz5~EBx zf0&(T1m)STY^1w^sRr9(G$`?xU};?cZ$YmCw#sNwHmUpEutEX()6BmZzlgW_bZ?!Q z@;BRtuXZ}6n!KQBZ+473X6m$^4fI__!glA>R&aW~9DJKOjp>YwIoS>TRr^%;ujTF* zwN>|(ntze`&lp20`1c1Wrxc1^Ns%OE$E>0sMBa(SMt_jq1VX_xGb&>&Y;N#Locy2U z1k)>EAH%!ElXbtcz@o@w>O^vB4C%Fc(j64rPLWTj_$`XAn97JAy8p9DTN%;6ogQ^n z`*sR7=jFXW+%fa*Hf^WK=FIY!%6`4++et6I-Rt)DqE~WXtNm)`%(r!>&5PeUITH`3 zP#=GD!OV-_x-kRx+O%bUhXTxlSO#DI(z2IozqEmn5F~h|?;_traAU|^{Qf@oTwh*s))C$ zLYK1bG&*%!#r5J^NaWrD!3_W=9D!}s0l&}v4v3t-Hh)YXI;eC9EXz%skB#H(1MC>) zTeeC5y!G|kN!w2k>V&xSGIC96U~WX#$v+jR2Ysln(bQxh9# zn^{GlkdzAWI=%XUaBFN;#9p}WMr5s%x&EzYYKYTsfrHghPTk=q=|0@`-HhP`i}-7_ z*$Wm`lMno`?>J^I-0(&t9$E{zxn5j6KFJOt=<(ZJeq{Rdb9TE~?3%*QfBPOu;IYpa zw8Q|@Hko1*DYCSm_#M~PSo2d&n$)vKnord{PkfzZPGVxz-@mjPszbC zvLa(*51HyZ%a@mv4ofdS`w_`q)?~q^d z?-V@{LA*xTKBY}_UV2GfMlYQD=})hbW^KKyHsBM@C*q6pJSYzA0xGh1=e>%#k)Hq! zR`HB0OxcV(s=1;%IxXZDy};W%Al|gWeS>#0In7lgI5d5>Dt<`&++J%=)5l+{FXYs3 zzvWCfaB?47#`EB%ct!K-VjCH=vf$1}OKQ>wY7=UUiS`3_e6~YbUu!R{DJA^gQ@S zKXLFahh@d!x4Z1R>MPJjcTO@F-O1oe=d+!JYMqv!}owS&nAk6 zyvPmM@a(7x(yC3xW?SX^LJbT$_PdzUki39)8q={`0XG=h75b#oO+aY zR8bn95t`_=BVZ)O04D|V_!PSA2sdn;KD{c}me`$(0^zSC#~wdDCJ7ms1!MMk-19gg zyCGjVrq^x3nD*(_ij3erUUMYo{e2#VV;`N}X|-24c3YhJTt-L@y%3m&?tve$n{AP9 z3GSp9f8(NjVbne7S^MZ4ptRZZu&R|Lj9VDBKkSAOQm=G5s2 z;%pZ=ycINd7BpSqm7YuD(iFfUmk%vwa==@xS0fQ}-lQc=y4oPv56v_6UZ+Emwd;hR znOO>x%k&!TdesV+Y(AsQ7j%;vRf94|vRvqzP>xeDF!ZUmdnY&X^V*qry=|+2o+*wM zJC#A09Jw&6UZun8ltEEYB3`mFl;awvVo-{!YBj8DcxZs?91Prp= z9rg2Fe{ZwR*5@rM?Bw5rrvv(GVWQ)9k2a02*~3JPoo?zNKk^sG;!a(=GVBal^Ncy*cUZYtb14?WN|{tV zaw7oi@ub|S`!g%S-F|;&DpM#t@&Z<5^tm;wFOwTyeQsEAXzm@O4@!*6pg&Z1D)Z>0 z8kZ{~4sfuNE3Pg+wJ%i~t0PY+EI^UolKz1t4+az-uX~}ybbv2DlVX7dE{%$>olqt( z@r4`&uiKtLBduN%x0qB+?Qqwtcfxvjn=DZVJ=^4kG}2n1gI;R2bhq=%Ne6q+hXRqo z#&PyB`w%(vqu0e^3q)p~Ui$%A%B^US$4X|46?oDpb}dCxsCcuR9ePRa(cV`z%37#%accw%oR3|?sV-|)#h(a_P2n;I6d+rN&nJ-qudG{ zTPdhiCplDn88m1(MzCe`GOuiUj-)`gO$Ft6G_HLSI5lC5qCQu2dfqgZ=2b=P2;M%i zK+@xxOy)=3RHApkQ@%-nSGmqq;fj`#Qbu2Yb=|^e-DQTn#~E~z4BCkSVUzE1xoMq? zJ3HLgd_iPgsA9tDY*gn~1B}KMSEKnG1qT-K}WYHyYg7n>0 zKu3Jdd(Gsmz)Ml}imM(_jxMi>T{7)%j9z_$+!7@QB}kxgX3)IwT7^+@Rk+djc1%@- zUfm*ppmH{32F~=b;4^R_CvV$_&%T#`HtvrWRrIHZHz$!#xv3%^rz0)3Y6WwjVn3(I zU2{6pIYp*>sh|g%0oQ77Dysk=SZQv=&UKfeC>(VY_Jjaq289r&DPea*Q{R|NzPO9k z>N_U1n}`Cjg$wcoBDY{dT$eB_u$|s9;ig}ca(@^e!S;A%5nCZvtwU!1tHMGBA`@8y zS4I?z%r{``8BgRH7Tjd33t$b__%v@TX!g<8~Ue zVN5?60}I#S1hlMxpJ)D&WijTlb2PlIaBNdHzBb(gzos|l=a6z9XE2?%l1fJ?_7Fur zpyK<Bx&DZ9^byY%aDKs3hHY~CA zNaqc{H4IT^Uu+Gg^?njP;figU>El|D#|tYv#iZy5PG*aNNV$g22&tGlpEXG(G>!D? z3>gTC{NP2vYzq04raqe)vM=bH%8?-N86s|mfe6QLwU3)EU;W!8jq9O9C`(G;B3}rdRLPrvE81{+Ut`q1bjx?T|xLWL^Rn*9mf#D<*LS?eeiwO zqN6;&xO zl|c`oUJbGKMkLc5k(>dj_*v8Uo^;x2+H3 zZL}H^_6I}_9ugU+39{)c&_8c^+!^Q*ugeIj_H9tQL_^qp=#~?XVd80{IR88+OpJc> z>+!L+c`aOQ6dqf+cCuh<1J;L?2$m{zo4rb)5j=SKt_iRLzSi3?taZ&$X3{YXfdm=K z;KR7*jqjgb{ymEo!Ym8<206jwl``;04UiNUDfT=?&QS3^o;&4TF(tBO!J&YG+o{lt zqgJ$B)C2B|HJ(i`G;6W0J~gUdTp6*76aw31Uo>Ri1{zca!T-(`cEN%4pq=s}?{jQ@ z*gVNO_OvKdo*SJXT@f-z)2ziJ>iaWhFPw=5)aJ{v)AQNs_vINc*UxAVF;`$e5JArn zRmOxgajz;*U|^C2H$%#0XP_DzE4HyZ`#s(ImoLqDy83b$GyRIkIiF8O6DPSt{qvn` zo}m_mP0IN9N|MXX&gHQy@PHK_cTp@fli7}Z8#%uB_6nO=<-qO&a#D4zPo#}xS zi(InU<*pWV&@r+i=;VY>pALGbO0V7~%oJSmxuY~pEuCICJP94kB!{*VDiD zsaJ4QRLK!!k8Sz{Zx>B=5+b@{QZ#p;H;&)sXIg5b?&Lti``m0_*4nFa4p{O+ORL3{lW$prs@thl^l>U*zGPmW*-@cuBU~2 zSX9hEbf^A87IX8P@E9|>Rtjky#R3z~Dk{EHo*I?P91!maKBqVV&fcQvS|E8sr`iVh z9a5vBS-t;1RnLgTj_BSVTWHvMrbxA@Ymq|sv_7sh9Mht!U+8n|Ql3%V@Vh;wHo&|; zGSb=f5^;DDL7ltoy?bn6`!C`kky-u*(MJC*EaK}HsS&nd@(;-&Zc>BCwur$>YMi3j zk10}5#UCQT14tuZa;;yxUlO=Qb_XYV?iZK%Zc^lGvV@Dn%0QOPC3(U!$XwbC{z+6)va_@7)gxxkeL4v8@E_18dn*_&kVQ&Q291pwU`<0h1 z0%yVM@4P{(dAt&Cu@WICDE32&9Kn_FMqhksF^wPdX}8LuNgark1z}~;`$DmGYLXyL zQ0ki`!y^W!ovs(2k#6=f#vYlVEA`Fs?O`whcMf%h*#=zgSReq-*(IbBO=eULT@Nw91dqkktOjpvJ-NoK-c-zk6yAFE zESGQ?4?dZkpaHUA@KKz9lp8euCi(Bb+0t`xu}^rsT((2pfeocnnFL9iqKg4GY`wZT z`ff}Hoe{1(szMV*ukMb#42=-;c*-=gUmd$G*L)m9ed2qiQKJ(s$&hn4p2F9N$IjMpDV4SNsGR9lp+vwGmA=r*@wYs5?aC!eEKSs5qvLkv z72n%a6M!8%NzmtpIS4iMkrz^?RYhdbW~e2ya2-gOmCx@9NYfr;_0Z1lpm!If_!r1B zWl)Nm3RM0$NIShaY>79RLBbqg;I)KA1~jw#+)4!J*uBvibme$mlVJaZ7U>mV^MwVH zlpt>M$#KLD6f{HJ<^WadN1ImoTGmHRzuYy8tmpCiXtxzo@+lU~+s#ybIbA7)YG?Ew zmXSU;{8I+GkB0)1L{MSaN*_=iV@m{!UYZU6&6VtphWy4H4KAPaCfpO=rB{0w$gp2` zF5Rx=HmMFjXJ^d5!(YbXLi_S&Fs~xz`^tOII?uNGbw_rVE@o-OMae~wA4Yn2Xs2n- zoXUJCh3u5%=UmaE!_K^L#D<+woPU@bHY%e3_OWd(5EuQ#W5=$Y%Auo>XWy*N_bwG| z7Gi(21WBIY?vxQWWK7fU5Q3Z!kf)C5w(0Bfo)+Cw{?%_Zl4U&BE%{cuC7oi|QSe+L z%ihGK(V*<{(%k~fL9Z@lzo3yX8|;T{O|77fzAEfu_Gyv?i#IIwuVw?q9;e7LDt<2eu~)mq{C5-Nm^W)r$SNZmpn7+u zpoU;$$`yVcP`ZFEP?H25z6HWck9zTGX`V-~8~C@NiRFBOUZ@x5Ap;XrDlp%@L$!=- z6JY{jubcVEHp%%g^L|>$Txk1MDKiWF4qp`b36fgVoksBk(NGnDqf=-7PIKJjGG~@+ zIZYjZHTYlNPLkQSXaADeCU%;+W2mPFLOspeJ%J708$w&f2Ek=%{j{^7zs`p>MK?uv z&^2_aDoa=^s#SM`$E_!>ip~;NLNN$t*;WQ_jZKiOne^{>|L!U|Hf+of0nd)y686Tc zA4@F)Xi55|-;hLZ845ghv~0DKqZ=tUjUsEQ_#58fQNf7NC!Qy#nSCm{Q~p>;I9Jl9 zN`}ZXdPAVZ`GrI7E-cuvnVccu!)656?CSR{EgTM^3Xkz&hfw8cY&%^hJuk?RwWB8t z%e=7Tp@~oqDNxN>TP6e{Q1H?w^$Fm2Lo98dXm-fBN{qlH4*@9yRHtiDf}{TSmw(;y ztXF!Son8Zr9c0>_S zAzB@Q9u?OCV0iQIx_BWc>uL0O=H;K-%9Up-)7p;oSi&rUc(Q>p1hh+QWADc72-qx4 zm!wP9lhe`rz|%e(86ow+40s+Gn2hx7*~x->dOg`JEQ>+!|Lof7m^0_P_K3qokIUd= zA3c8>Kf1s+LTD%GmaVx6Njq}{b!1QYVKsO!g6=|+`$wddF^XOHLfR3jjy`ESx4Z1H z{U<$$6P&-Rod3yPbk#Qbgv%<2$021qSxiRY=-)ocL~DUPgph?+XJGD!ol|tiu8q~J z@t}(WI)j7JQG}b#kHN=q#ZkYaGbzLA=fGdk6FdDJYDHPXM?M~;uK=oAWgQfU?ib~G z=+&Pq_5u0D!l@%D1o;Y%KL-AA&u}6)JYLY)iQItH5Sq!lfN(vmhtXk`N$W^??+3o% z4q{^9EhDl$V?y7L-mo}9&i;4V7o_PK>k-sx}85vpyJO?H>eu?JIRgb7J4>m zAN;^LGx5zEGmd|A&ri00O?T%PpUpU_O=Z%fPEN~ozea9(RtMIAX_`Uj2OpPXPVC{( z%aM31QgGoXo&&cTgp#WR??h&bihWX<^Q0=G|7+k5)nWB#(f2)j;2ylTR4`u&&Nksr zNV`VX$yTxX$byOY<++e|i=63??yeQ}G8g2ZiWAv0noUpbk>?QVc(V&P?89L&bNgSB>L1H2k|;~O<}yhdOcL=}2NhaL zB0a^XQ)C^Kg@q=`)`x9}hB19^SA5%|cR&_CrkhnoR14a@&MA64vALZdLW(0=0?8dT zE=0o!)b`gW82nFL(Il;bG?p-oBuY)Ju~S0@^j2RQb#wAd#faD1dUpl zA`XxlFrlvBociutw{6`dyj3oC{5M@pzMu}W=&|Yo$*6Vn1uY)f=W>@{<_I$saF}S< z&#l!r*q051D3np-?roVh#iCH&snphzLT>Rr9?OfPR;s3gVu8MG7ZtxgtX^C%-X2>R z(<(-C*pI?1qf=&XqSuG50MgIe(3PT^?``^V<5%_StiX>0Af-*vBh<6oggweaMWshy zWRmC>voCbHPcNfarvzP5LPrHbp&%&&S<<)44|^HM<7&IqcR|>}P^XC=4pPM-E7l4z z!`S)Z_U(;97WfqQy|IgAa)S?#y=+xh@YznWz^+q3#bYf_e@`IbX}}!(JQ^a#80NYE zV~|PNnoQuuDDgE3D6_kMF*M3dnA!kK9gGMf?ce<$cd53AVitbByg-ttNgeAnsPxaK zvmnzyp0hsN<?#bawhCcG<)^{(AMExE7D|;~w}mYfU*QsuN^^s6euX zUN&)u>~zo;x-V??q~@SjG2UIJzCHz7lR@1&)|nw;`dW`p;w-c{4JG~R%1PqIzV< z&Vz2-`~xllAs#QP?eNkrB@G}Y%s&23&l+f8xW((dz?_e;S%{5)z}2hQ;DM@$b4;_} zd?j`R*KN?O6=0cJJ6%qL*ALjgTo*PPHjbQx8{1)i_}7HrMOn0z>hHRZWZRcUOMPUe zrK%~mf+FQqJhH!m^idq9o3zCm6}S})bBJ4Irj7;T%~sYy|a7phus!U>MTA^5>`j{5uPRu1{MrkzD6-xyK@_{;Ho->|Kg< zQSpuRHD$iwq;k!qc6z1H9aXDWeJB=PdK5@bkz*5}R5X>r4$--EB8xtLw2lE0(B=~31;r{DpMQ!+kTM@Q*ClN9 z+%@gEJde%^nl%fH>oNRrA-GKSUXSqpSE_^`2!Y;3b;1uX1leb%;;y2_kxa*xnoGC)oZ_sVpmXPITfD`VVE+k^2_qvJ7E^m_hBkq zHPAsi=k?_9?l}4;a~RSomvVddgs)yT&HH1se_#Q{u^AEHA&uNr8jsiN9ah+GrC6wu zIp1HcrdQu1Wl~JMOJz>T%IF5DdoPSy5rmz350T|@^ZijonJe4$Y6y5r(&N(DdL?oK zU6kMR?-Vsav5i4@A|D+ykB$4fJjm(8fz#u#xo;sgSAvd! zcDg`@H{&H&g(vCF!c@k5RqEJ6suPMP3#oGo#~J`T`mQ`uYEIdD>id*5<>x+b>ywl6CRJpTM<;lEtm40Djj>+nPma<3~7aT1E73j%Xuhr`#wW7%Ve zuLD!Vu~v4mMh4OE-qzj;S19qk<;TX(1Kj9RW!Z zbHR0dO}RJf?DUL~8r6bH8`Mi;7ffvN*b=ob7%yC>K$f66B4qf$VgqF1$fMD6X=cHo zIT%y#jwAPv+vip<-YwJzou5!8#oKYI*AutF$KZQgkXxot#!2ZSD@BYZqKrwwlUKQo}2rnkVz{Sq^7oDEJu8 z_c?m+%fFi=wyn9hQ`(N4Cv9}ETmNT^v{gJ;f=zj5l?4_{4a{6vf^>jG1rCgqY0UYt zI{I9Oc^xO=&tKX7gVvf$w!LHQG+kLew#DNEA)1Gqg+`$KGc9YAMMcp~fCntL=PguxKF~ZJmo#hF_}^A6ht4mpVl+T2z{9)4{je%6v0hc{dk|xpJKlLE?pt*zs5!S&;h_a#olxciZ}s=`r&o$?`=Z~ zeDx6#JHdeughqvlY7p1L8FZ6cXFk%bZjtKN#bn5i0YY%aoI!7u+*X}Z0KWXQgh_&K z1uRP{LQpVWluN}&IG9uVQXE{JFEaX9?+ z?G50fJd4Wgc_g(}J2r*>o3lc`p$DGf8kTY^w{~8#)Z)K_!K=vtS;*5u89wsoqj9 zf5~FC+!tp3lvE8?IKW@5gFAnqC3Bo&A)V$h6>mQPdeb6qt9)PR2f}tbgZ_fX4!zI`;)LH`4XosL zD-Mj`6Vnu-OOUin=aEfxk|>9+Q2^5&mMiS?!|w*|mg$Zf>zbg*<0CN+)&^W3TnW#H z39`YVH~7&<2y$S{KFGSFD&1_0XnAYZ*eQ!eIu2aBW6T&*IB`~?>;y758^qW%=0Mbe zC~VI>l8Tcf5FD67=UieR2!q6*M<+f1n_t@&-}4q6cFO7R3L2D!f~|B3_yw2yRYf#G zzz=AZ&kJe=H@uJguYqiZkpu>Zf&+Z-oipnWjlvBYLHqttuCgeijB&sFBS{@Y%B*C2 z4#k3Tn?c26hDoRIrKm!Zt!eV^f^1oA{bLlbo0baB*bDOOe(TsSaiL#LY_FC-Ls<8r_eWMd1g5 z`<}m^+KpD|SV6JNDYAr$zZuyQaeFF6i__!s=)ECW_UdXmW;^kweMe@0&>LNR(EjWX zW`-;;=_s*n*UVcp#7=AG^r*`~7`MbfGiFwg=S6v@2<#Az4oTS-lK{3V+8GGaW{%OmB%UfjFVYSVT%A>o4m1^Byh&Un>DgJ2{CxKzQAgnJq z2fVb&q9jq00P|;WhBOFTy&8nN^CWLlW>kZ+hE9WYS{JiXas$gF*z|7=I@3-#$O$Z? zT6X?KWD6E9l_NYhQtUwM?DR~{;@~>p!kA?5b-*c}5MB>;t~r`sm3ccUVuMc|{Z>-! zn!vZ3wZ}snLr*JCDl5bpz!JVJ=+m%XxBan|(0J&8>-=(jg;vDHvqD=dMyi#u^nPX%IT{wEKlMe|-+TT?X7*Is3LP;}aLn!ecRF zhabwEW}nJbhwYJX@JXBu2`M{O9l%@(eVU_n)zV_%&qjXbhb=N&X)GFYJ?1$f{UPLC zkNIs4)5NxXguJy>?Jx|^4=xPWtIMb5gyzz^OL7do7`>{;j^wh%q68eA<&jYt1HQ$+ z_84ScJL-e)r+?=)i>KocC)2)1KHyf#!(*`l#DoJ9mOiG~dJ1w<;<0;fBb_PA_wH7t zLg{i25U}49&XwrL=`h@8n2<9e$5(exSTVKUeX|@(0`(A+!!EkW%@47iCIO-jZ`~2F zd)!h1@b zMw{=|;aEsd#z8{MTK=zc)+;7m^Y>NhYD{1UhGn(f5oMJa}AES{9Gl@l> zWzlUTUe4(7G(TlpoPC;;s2Tlg-5cX=Efn5@!j8?6ASsL~gC-rR%olXC-~;fF91$1< zNaFK3d&(c3?X%Fx?h@_@z`EmN&m~?X<92lzA_FCWv+*#3{ILaw@7X<<7 zcr+I>Mdi~TL;{s@lwO@je+G#lWzvdZP%YP|^hKLC?mll|T7?O5hgGYjRY9O~^qN*= z151*%sy~6IFa?rZp1Ry9G%wC6I;O5=``lIpE!LDt=fjZD7ZHvYNI2LI4TOlaP6<)5ZSix{th z8q*;hvNgDJGWp_~)YuKN#G}O36=U4yeV{aO?mXL$#!ZB%r{?&ET2L}60@eyhzC1- zJ3!swEt!x_dzozXf|#>`DV|zFt^oro_6onLtQ9ngb}K5v5OjTR1ur>O#0(uw1A%4U zwgtp+_U(h|yT8g_Y}@V5P6N4r%2u6$Z9-KIj_Tq?LAz+*)|Wt#V9@6%vk z_T6#Ud|M_Vcf{(k`ymy+htUVE_bZ&37m!R{bnf%EdM3ks_*P>%bI1O}AI!tb+lW!l zryIB0rZjP}Qh1z;U?&0UQ)aDVj$jj*qJ@y(yCin?*sGDv+RM!D&|6I2q*B@SDNYq0 zPsJPufIoFCXI`*#`s#cC*P?u;&M{6VEnk|Tz!z3(shwg!p-3BaYKArdJIRiKo67Q; zU1BI~$A-*0?6iBWLk*z0# zHbx%{OH>DOVRuyJB=3+&6e*9}-yP_YZ!Y^mv=%m0LrLqg!Fk z2j`HO=Z1GJl&|S3!f^SIta_EvM@2fUBS0nCDxRR;b?F|`- zXbyOLTzxtF$N<28^qV)@BQ1E#3d&wXioY~?`_Kw+RTR6AB72}oP4tSOY|mU8nR9mr zpfTO!c~0?w(V0&mA@|Mz^EO78(CR^h2rX?aQPlOD-H44;rcTyLn=i)WcqTgI}EzbfM6;B+*WlUyde&E{w^b3xruv_GvVgewM{F zsT-y6!nZ}AiiX$hbH#$VIdQWL&>^H;)-FB%ypgUFc9IHM&FJAJo`vP2SkZ^!yWFS@ zIyXx9sG|@%L~vjoE{}?c0YpH7teRk9A2{`D1>NK$FZhmYQ2dh@kVE&A1xKe z*nPtXf9QiAw^S6aCa=Z$6)Mw16re2a8%Y!2NP z-J!rz(oXs`lz^DIIqpNgOqWov`798{odDVXJ7?eZD<4ZVQbbl2|k3J;G;0+`&WQu9=`_5V|U8a{VSt4 zh98Mg>70OH->k*IFEeMS_o|8}UHtydSIqZh|6jej;U`u9-7*8O#O(F+0yO0kB#q;{ zBflKnyHuF}Pp^epP*~bt0u{$RM&Us#O}Cq3cTl8+ivIw*-0x6z%eQz{2d)tw78q1n zo|T%6;3fsofUu{56Xrr}H)$1@Fa}jY@I1-7(0j25gj-{q|9|$r1TL!V%DbQNimHo6 zZ4^{J0R>915L*_D5wVNjl1#QqW-^(~moG_AcVv>7$isFW-fPz|pY=Vfa zDk_`pLc4;Z;(`hm;?md%iZm*G=TwzSm6X%P;cYeeYf9cmMC)bI(2J|5xV_ zn7P~&Ln_5tq9o2)Wh(bBcWXE>%?#^hEpQatWM+F}%xq{_pY&;q16GEWI$t62m}Epw zz^|(3k@e0XSTsBzW(&o@DrqwnQ|VQsNSoSEwt$3hrSO#SoUjcRMMqzcW4eP*(p}O| zT@G+=kes<0epS;5ZvcHl5Tp$XzQQcN2Du`&Bevvgk&5LI~|CylZa@W_CSDssR0 z2-yS9VGX?%Jzq2ll@N$ zabKc#dc5}rA^smc@g`}G3iF5eDVwB8kTt9+5VV6B0PbAaJrRTLs6TPh?KRxSep%VJ z3A>Bm|LA8%6LRp|xBZEZoe6Q^ZHzb*-})}a+@Z*ADh4?`x6q(K!NYt&OzG$&uqfZ; zg{R0;md-&okZRZB@D6@?L_ScNW%AIr0NM;<5lOsWaf&>RREK0j0%~!1Uql{V7l}Li zpSo-dPoPjkWs_Hmyq#a~b5WQ}8-9-_J8Xc?nOm#WT)ynC1r-P0JPOrtB>&;QmOUl3cO+Xgg$DJK0&Ju z4~7AM?IiCX{>TWs(=*P$M$S2t`z9v;2E|;X$W>IN$)WF)6j&Zy64k=yWwUyBw5IgU zr2#GKev;=>>UD_T8?7mRvn6_I0IJpKc?;qG4xePz0`(!HzyJ1|t#JA26VOHGTNjN_ z90z#gSMUjh!MJP zm(5|1QP~_2gPqk}S;I-*3>f||(y!OpV=5Fvg9E9!2$N9*s z#|z{sIgh+@R9gQ=UR`7h#A+^b)`2=xCCPSgblYj0N<`~`!frAeVcc0aS+;>c`%8KC z%O2tBBlFsT380>nD5{|^LiWHa{|!EAOrzu)Fwb1|+A6?S!yR>exOqFNhE0q-(UP#j z$*|CxQSUqG)E<@We<&h;T6T5FS1%g6R67Ra|QfQO(bCNOF z#&sIDM44sGu)i|#sP1xvankzU&aIiG+<{}tO(ql6Clpgh0U1_IJ6*>=L@$+pCLfgT zrY|rXJXiDbl?Q-@{8HqN8B67;ds_fu9bC@rQO1MX?GciuYEiG9n8?`@-Y+bfRsn82 zCN&SZm2)2m4hWjnr(Lo{?KDK21AvX*HCQ@>%VdW9zpp=4e;1+=bmN`ftEv^Y=!3z{y~o}P-Cn1JE%0& ziUA#YSWbABC>_v=li>pDy7ahswY>TDdc|p`0K(_{g6nzRuGrV0la3suuj>DXL8k@k zCPl4Zwzx*v5IN{n&&w1)J=>_<6;wrSoKq_AkI=~s&!&-`+$^^*n3hO9&oIgSFdga? zijZIY`nQ@ty!Wg3fB#GIVv1QvkysNn+4IEYj~RAi8G={pH~q^*Biye4y!CHnsRJ9W zd=qG;Q4EAj*HSSJicFpk>N3tVYXi3d=kx&gB%}iw5_moV`BBUu8egSMR;Fm!INAHq zG7d&?geJVEESN4bVk3*6+C~x`7#l#`IIJdJOEDWLvYv`L?3qua%L3X4P%@z>*J_et z!0?n8Dz1Z`J2}hA=afToN~+J!s0vy$)@jJ9F=Pc0qs6eQ<1yaE3Lvljk(3>I*^AL@ zLQ>>tJZkh5CdE+FWytm$7Z{k1mUW;oeP%q)*n#3*Y12{jy3$u9o@b#7wL>-_-zPdN zMk1&xx>J>=EDOHuuemPHbj=JKkniN`(mwd8%5XsgoR249L5gOEcbP(z|1xaM2NCF>GRL70cZZs4yOb>9e;ymSbYmBP zEQ6|IoFH8aS-;!4nZT3{#60c*$FL!%0e9Te?T{aWa_MC0-3Tj?Ha^HQ2iuF0wq&5{ zt+7mV4jd)4FwJcXN0ex;$@lTrxIYx@D=1KV8@;FMDVm$Esk3oAD@A?>a@zoInGtIv zp^(Brt8iZiQy5;<>v9_IzyJ`Y)t7S4a-Q5c;BL5R1vFkbNOr-_3xDg2Z}`sTuhp2# zD?Cd->cHDu7PKHrh1&)BbpG@{PO^VW816ry3L(y1I_3dqlSig&9o?dySML3gf53O6 zPm(@wqtkGmqRZ8{0~R|s zt5@bK2bv$Ge4x=>@Ka;}l+-5Xy@n*UgvtY`k&@p6VX75FuP3YIee$PAuq_11U`NOSD6qp; z*aYGV)1=6oq+Rk`a;;)@M1e`_r2XFc z)1IwH#nVT|<%kt39M}4WZkt+co}y!+)(J~>Z+T$dXY4EtlmW?!o&)I-XlH1n%7T&4 zZOAEy?q;?FRmn#F={fDd;e~BjiqHX6SVK-2@G#KA=BjeN9=x^jO|0cbm?it$DzNr= z_^n2c1&cDQ`Ei#v`T2?<=k|wpGtA}P*?1(5Ybl8ZG5922+E*`&2W2hl*f)QzRh*sL z67lGtaGf>ib!O@b@725$kc9zz>+*hr>#eiAO7TtkY93z878gPlVE@-Hi}N{4cwNBI zu_Dr5+o3kVy-jd#-iiD9pSqgobXiD-iudYN#ru@Z-Os)6vj>VL7xLn!;n~8dRJRuO zR_~)^9-f|`dxvROuj8!uJ|EnqD5q0>>PR`=!$Y^GJOZ<5dU!o>-%cpBMGlHQ)qZZ3 z=N?--F*bPNc5z}@85+m^98a!VuQhjE*cckeoqR3%U&wzt=(H?sXP5>V8a|WVnE8dI z$8CEsu;`BW>c?tun}6UcZA<2){I zkllq~Xh--p`7U@W*?%Qa@^{G2kRC6*`*38B@M6TE(|+%I|Kpym^a9RdRkpZ43zleeV5|L8mlocM^xkn+(xK}>jevH+6tRq6u zj*W9~=mz|auqYb(!48t?z!{!3CIekUG219oNX7KZ3j;P!s`oua9~WOyrcUXWcdC;} zC%H$j@^4Y^ol^|NHodZY(oWK;J{DX>XS=t+&shJza8R7Skps&JJK;XX5_cebvjnT-QRKYlogJ5ex7!JeLPW_c>4zKkgqNyfUj^y&Ycd zi@Id>>K)fp{QLK2e_&;F1u?Pvbx=<41L}r{pq#Fg*Mls=hN#|%syX{1Arhn#=q+@e zY!TBdE8=2=RwqXe$5=t0607#prVfQxPbrVUm`Yv99$-oXice&fsc>N_@%5s?u@*Ph zIUiR1HtTO;g&W76yH!>F(Jy;}0jj~a3afedBFhP|$S#Hj5o*){H-C;wCvW0!_CFT+ z#mxDJJSb$TESWYQaTKd8-ZRLx9)Y91lNE9&eE+A{3(YCl9hM@=0v9}py+dfHJiG&F zgbnI3OE|~L6X{gg>%;gPtW|bIcejwQCr9LsbEPwJRb6dn(=W;8)ijl zFalT>cvP{)6?I`WJ>rzG{o<0yHm|BVCEPqn)E_4Rv9}_&0>$(Du?(48;kVwF8!g7V z)j{V->MIjS16JW-6|{L2lTDFKD&{e8^!fhH96Ze;YoOqIW#on^O*xR{p9d)+y(krC z_?LQB_-eXc_9`lTfvLSvutac6ZcjY1htTl3$F`Hta8Moa`t1LFxpt0oVNbJ3iu`_? zyil$KGWalI3zzWYWKN#Ge``_k^oj8~id?YgO>^_%unQUs2BDq*sptL3MGW#ut#Rq& z+>N**)G;NTYS%jv<+78$iBR&DE!J_c*5gw6nm2Of3k64r&SSqgL7;<-B|C$`+Mzcw z1NsyDLcsEzGSW`cowBd~x0`XEsrt${PLkyg?3osr%p4miW<3R?5YzX1$CMQLU$JVi znbRL}&24EE<~?k5Iq8wBY7}O3HR)2!jnFFYyFcLkFY&@Je*fO@V&4Cs_hSC^{;z)f zKQUsOaUv8o0y-xAsWS5?lo1_|+#ekWjvQ9-a$xs$qY0K0DP}oE5~!FQ)edizl9_K< z;QWJ@IM%Fj>}y3V^uCeaz0N!%iA|oG<7VK2(#3GM^odaF2LFS+EYbPkgLD;@ul!t^ zCE6x#3|c7I;d-AOqT5}|xZ6Aq`r`E#^&%-Sk3n_=9~73M!LFJLI6$MGf-H=#t(rgl zm!rAnYg>o)J}j(bao^`rXesy@u#PR@ApLYAr!wFZt_H>1hQc!fmP$2eLOLKG)}m+? z+S%DKj?5?q+*V`y+CL<%Sda@(?zlr9f~ zsz?=ID*9uW`vp+SxlT$vPJ3Po&-cm_^?2Up_JGl7lytjvs*zFdGG7b%_5CD)RD)l5$&Xgns4r=Vej$8j&Un} z1Ies=sF++;3j_REzU}UDg1ZqAGH;R=aS}X^12=X@xaNu%3Kik?@`%pR*x*8OPhA8v4%c(2z=f-&THXn7y-;?O^lB(v#`Ke; z!UWeb<&?0&i<#Mb@#z>~Wf?Cc#nG;XCj8Dkzle>4;=qQ-LN-#CUq0ugTXjgc>p)a_ zgyE<{j4WKSg7S!ZFiI%mJitu?s;>2t+`xAJL0^;)Eey~qZmMqbb4Qd&8pD8@Jvs(l zne}AbQ~xceVtGbuGd=BJ8-N#x71SIUDLE##DV1W@QY49rDUv|Vr}t&~VOgzTtN?Pj zN$liCS&gs@>m=Um1B-)=%~YeM-upcz8b#OT;p! z+~x0-Eby$JQtgTwH1}riQy%L~g@ z^+K_^M(bV|ltJgwxD3)NR{4M7-6`8F85ay#JL;x3i{%iG;&xUxZG!YWJHGL@qsgCM%`5hdCzyhSsq_~lOXV9N zHSr?QgPkLt5Mu0=uVk8S?}ETkjt(-W0J2B8kQHQJo40FyuX&Pz!)_m2NE?U??3j`m zk}lDzl0~q6+)rw_SgzI1H}Ix+(#V+)K@y}-H)zJB`t-YH`c?aBVkeH>AA0U*J$jTp zx7#saWQURu{jQw#AJ4Qroo65OE>YT|XlNaAeF~=x!n2Pypw`>)+aGLQ)vj=M2 zV8b@oE_vU~Tsl1vE;XyWCaid~TXtHlRrE(365$OP*TkPSbb~C;WjojKfL3u%*hC)) zItNOnkFU0o0?r}&FE#X_)8A_7X6f2DYasHr(tlxaXYeiFL66SxdF!U(&F%c1z`9ik z(qLQOs$sjQV#UartsO>t%5H#Jjhy4|h{!_YzWuHdQqfiXPE!BMg!C_)d>Ks?bDSba zsF)295Kbd!l_1p2-$%C4_hy#B*U<-p2O!`=V?ty*Ki)YXBogyMtEG&ykN5Fn=N5IY zs*eL4fSh(d-tx>1gR(Cq@ll2TH)f5?v$V%&TD-y^&cYhI=O zU);05fz{mOxU0u4>fj-vaqd~V;mXg*N_KOP14mD`n#?_$C?=gE>!=t5XVOV12gGEh zv>+(o$kQthwu*~=dSxZ9ARM>RbL9-oQCS?;D<4zZ(zCd+&|(eD&7CR>UD>qx!YVgG zxDg*hm*{L#;yglPljHhU1RRcG+4Xf4Q%#ZGR16a4ZwcSW+e`-JE$Vt+F5MeC;C3b8 z2s!231fDWF8e0EG2|5~G^3{=vA*qri(0actM(4baMi(3VI4oZ}r;k%9E|U*S|X!YINz6X2ZT9(rA5z#7KrcqFa*|yWz%9Szr^rbE1f+rY1b4WeCs&C(1}jW7c+wMmJo``GVCHUS9?sLgR=3y|R7#U?gC;RHw6 z@m-_&N%+m)d*m3q>B)f&6VPoBGfb@%bDkn+shCexT2)u@W+sn&gJg57{cg>=$VrzR zf&8Y+;(E^6=p&>OfY$ux$0!%L$*WX$f^^YqC*Fg`WdEzuJpZaW2T6H&z3)X%_1qTq z1rR{p6Yx3c0VL0<3F#B&Oa-}+@J<@rdS&rZHT)!A1HBoZYEgH{^5~vfT1AgI83IL{ zB^ipUNS%D%y0_*XVM+s=)y;3lQt6UU1*Nh>^y=5kf=~0Sr<_(>1z4M-(cFZMaBdA? z{fZXVPc-7|`p;YcMwY%Z_{ulIR~p5D&0R~yWclgjrb#+s0YUX;GhM)&%4E zxoFLT7z`eF1zI(#*1X{t-@^_YFYeD)ckJq~`AM`98{at!VJ1u0dgxHQ`%uFS{>O8thTs67)N5GQskc!Ezl;f z4QUV60jIg9MbYDx$eG_HE#PE=aC;Y><<=`lx+PGBBIRmLzhuCDVQ{9X9TNRIA*TTe znc%TP5S7gd_ovwbWCTlKWfNQAu$szV`@_<-RP$mF61IC@`30c5~Wary@M_Xf=H;BFBBQlr=C338Jfts zAVH;WI~kaV*@y)6DrzV40F($DD(-y>vU@H&{1 z+WE19L0LZM1W8xr(buJWl-t0CKJTve$_VHosMuQWo8g+}*F|GbrO}uXkw9u_LuSM_ zqJw)9Lz)$hP*;L?Y*yuSbkM_nWg`fr#RlinN%8{DN~Qx=bxk2PoSl;WQ)>9vT+w3! zbs1>wqU-2daJuj+R;ersE)+Zp9SOE$$DEZ0>Up<~07F)oqj)OC_jeuvp;PE z-{HNmg7Ry=s|I(Q?>#xJ|7&4W40n$)Qr9U{VfnNk4=M7qq$5zPz?E~S3~J!Vb-sgPvkDbx z7{)Thqj;7TDqfrJ%PaYl)(F>!gr4%Vu}kmkGwU_k}@w`oNI$pr5$uHWJo-qPw{%_D%B-D z#luc+KX`1393&sZYvsY(pqnzJ*hSWeL8n^3){tAgav!bY5NQVm5!67!zgce$Q{De~)5ogBM5;JZut-&Ujb*{RH~Eg!$JHqe~w z)B;K9a`$*ot>T(|9}kE$a35&Azj(vPXvLcx#aBD_M_7ILj{Di_R7|3|CdWm08E7n(&SK! zHz$i?Km}$46;tbXFSJJ4Lt2FivtZ?OV`NMm^djI#o6fdTj zg%pV$2^YiKYzv5ydfT`DojG%izOLK7h7_{XnVdHH0uNEleu~ti>}&@V5jP6&O48u1 z0_SvKi~0!Zn+ZF3lFP99AMdOa54dlk3j=iGNAfuL#m>1<32>|dZWvOPz;%$@j#NpYqL64S@4M`iqK!fAbcL2?NV$QMSAqp z|LD^5N58!}`dwm4B)7+>UMd%6ns(k{mcu#=?~Jh?5YMp)98J{I97Sysph? zy~W*}EN*IKd{=5nJV7?6cG34AIoiE*BQ44h2HiK{ssp z8L|_Goc1X*^vagELvvI$AxPb^)u$|Y$jOl4Yp=-NsD|Ifwy_^tYo1_b>0bLUw~s!h zjbQ2fH~%9fj$IhWffK_sO`x@gVpdUP1#So22D-`&kj!1d)f5N{MOVTX1$5E>h*go5 z5VB^}ZF&zU!R64G%c+f{uF8v|??GMK(XVE`*~{%zReUerIrW>R?<{uiBlk%I-6K6G zf85gkcB$89NY<g}JAEr)OL+Z%y z3}?uHhv^5HI#NUH=;qm*ls9Hp$dbs}ajL-|1wL$QEG%`F^7q${&oIs`Z?+YzAcgE^ z76*>BHJHpSyD0|p=qgasvOtkfrv){IXo~$e$d^euf~eXm(0SGP?3QKFNb8&{D7Kya z%qYNPW@t=LErEe;`{tjue_|etb=W;a3xU@Z$dPSz)m#)_6kd?vwn3e}nVv>|%hrtX z$WPk0m6fA9p(p0UK(W!B$Uc6jnXGc)Rn0aN1Cc>7u(Ol~!5yHK!@6aiTraS`RaqRK z#uSK)=;JPn)tY2c)tnra@fylb*jwZF$?#f(5Yz7*&z?MPw*yuPd9CZO3%q}7M94PR zS<}dsk!lzncv=}WK}|QsbW!9c6;to~$om#=ZHT@|8&6Q&5?8=j*ouU5%iT2xf{JLo zv`4iq9Cu?eX{AT1^Rru@<)Q41Fw+}m_TkP#eg_(DGEqn9J&cuFOf_GYa1cUA6*gj zS^7iULi2%dK_7+5pWdWc#A|{rAmqr@G=}xf%%E2=Nd%AjW~QkcNFQCIYz)&XPS05% z_|Zon83sR}zDnPBA2#?;U-u}EY>u2!{0$>LVi`GayY(G>*@JW!Ns4^HeI+DE=0S=H zvZZFw59lU-xvx(CfW8~iC`pkc`6=#4pqdB*N2@4wMWccBBVcDh>)kH7l#d5vY`}Qj zi|v0}2F3{65fe_G{?@9OJ%cWer<-+xe@?hGpcLvp^65*FiQdV6Ykl^QuON|rD6K?= z?T7sZJ+eXt_08|>33%D5mbbfaa6d`ch#voJR-b<}`-6(N9{li=e?ASN6%?`R;_dQX z!9!h{N^zR1i=zXgL);+U=8?pS6?DzKAAKDvelqAs;$%|q+Z$!8NyTKatc1*ST@RBE zZFIs8GCzI)PuCBD26W=70Vql$}PQ|pi*3)t1vbb4&=Es-Cr`2l& zmq-@(Tu6&LPnF?%lJ207hoe?=yz_N=r>w`Vm0s*@7f6_o;AkLX-iJNT*+E3G`J?Zf zyR>YQo?bp@x1GPnJ3*kiF3uDm(bq(4E=XWWh29OW!rRC#cSvo8B)a18zK9!PO&qLl z&8PRvs@=~jQzi~M#q%=gy+P>`O-2-u3e|HS?@T6uc*N?6t;#q-0kCn}Y=dbuID4*L zve8QZLFn6I<_z5q8?&>(x!oq;soYH6{P*qur-v?qD#MQr+d})MUJczj1J**EzL_1e z>W~6%FZZ+*sg@T;HNe{EVE6^b?lxw$<8BX|GQ3ZA_GOVSp!99y6coSqyFVbcugtbU ztI33Qf?|$Rji?(y|tt{WYB4u zx?H_pf1FFNlJ|S|a_w%K3}eL}AXbm@m~Q;nkpKC_oOst^C#K6LL;i$Hw?uhCa%-06 z^YDjm*Jf+b2;|JonX7wU_qygvqk-ur?C>pBftN zOABmpZWVOWxnWsu1CU-?N!Es7Y?p;|!yY9jCW7^ewa{TTjlI_Rna>89(dcAM{-58F z6bD{Om78EHmtr%pNw5V2LvcKkv_&|hC-W-w|Rqa|hZ4r~gtyaM5 z3hQWVgjZ>MrDnpojsZPrVMc`vYaX(Up1tX^i{CU_l)n`0zCjus*rHrBu_!GR1LX`S zshBvDL0<)87SIR{s8KFsTGUG>-=t5;FGcndY->>u0H6EasRL2>_{*b^>s`mmR4ocP zJ?AXMYc@|Q;a-fu)&fqc*R9Z4!9u~H(~X(Aq7M41v?$^sCp%)$38)&FM*g5vqp&ox zj-*Q(`5FF$P8Hl+5*?|dE9t!+cz=$#l0K$rQKxY#xDO$#2H5XFxpb?bQpU;|8Y7HY zO;*;x=kYz3VKl=20VUn+U20D7!X~EUz#g&%o|+}js{=ln-4?olm*BjGb9ZVz1Z--! z$c?#6aozVIsr8B--+c!*KzcmL`+0IagG*qt?1O({Ywv5j-o5dSs;}dx zR~~T$>chL}CC-LEVuP={T=i-WPjJ>MzVNok4iKBy{hW`~<|DxFyLH(5(GS9Q{Cctx zHQ(O)(_fH{4m?-wHi2^y#S~B^4|5!Hy{<%N1SI=oGRiHF?peEE2f{;bR9Ubg$wf~9 zU7#2Y+MvjX7$4NU;zl+kcuwt9l|*Jk)e)|`Fx#euvDYTH4RCC83}c6^a6Cc1^1lk+ zGs5LtmtA_vMRpMy2i{WvYL8(Nn$IZa21TwxkqKx?chLQ$RFzL3BcG5{8 z>DMMN0Mm->A5aUG$SW73`d=O0tV>yCL@`KX60T zQjc6Q_};L2Q61hIV(7dkq!E4r7E<96;=7ky;kz#=jad?yPv?afgqK15@e`hoGlH*f z18_da_F8@E8GwI1w)m@lM!3DKq-AUVx2G+Rs}yX^~I0>vfO)ygFX~bXLyqGcR_8 z23uiZgs#Wb#L7TXRXv@*6&MloVfn;AkrfU+pB0*zoQ)KdhW>d>F1?ui9hJ2nMs@it z5H!l#$<@$9^bJ9lTmFo#;_I{S@D9;8efEocyt>#eg2pfqtl=?+-^J!pcJLVVR0g{k z;c@7n>i$j^ITHw>4Kw`76tkKliBt?)Y$U|EASszvM0HELCCSpe5ttmC;A-=a??wN` z*1r80^<(SntlqxkPX2ADX8M??<-H=yY6~=Yu}~)YiM~uSkDD%ucTWFK<_}N2z1X?^ z?QzGpwGlpBWA}yqrMCa)F>|d63nC5qbZf|+2+dZdLC6lXVRu8`EK-u#eeqv{Vk)@@sOMKBU0nQKIughZ|FG19HB@9@QJ~Y#|f73HYqiF=KYWr-nhQJ zJM|K&;SV|C~6cHNZA?E4Wd1pH#cn(fD)g+yp^BohZtgyOgiNTMajz zRLB)`ecT9294J+*7)}pZc91uzpuLPB9Lm&N;7Pv*4j>aa92o7*8nLNcvNv&H=)r!52u z1!`w!Lh*oYab-1X%nin~=QhB~(=OPdX4>(0Gd+z{(S_H~en^rXcq%G0nTm2K24W{# zDhA7;FF;j%MzCfd7g9E1i4Q7yT7}lXIi!e(S#=sr>(==m1R1@i5GbCwo`@^n;fe#8nUuMuDG&My}ORXXRa6Pj422Yu8G$Vihi*7I3;tXP$!yU(6 z6a)E%>c#TXiTLkG4T57zJeSp)yH#fER4Gg6S<3lS{$&RLoF#x}+`;wT!lk_b98P7ob#YNf@#&TM$j4#Ts5*}7Y9gvz?r zLFY&+yW|!JCL{tfm|?5zJc`MtNG27dlMlFM=z}hr&QR2U8E}i8g%rG%-iK8h)C*{m z7IBdA1L%HaMZQJ8)vg_~5f13yYEeP=9 zIzzA9ks^;3>=I|tt0331MN#VY3HR=lA*YR=D`y~~Tq0+cf4^&uP_vITx?kmFyTkhi zJs_kMydZ0ZFW7!x@0h;be@K4)2#L{-`{|WU@7@08s%fm_r$t=w%J6N043=h(>d1KS z8`8`2B_5feQB~!Z5LD^Ao}`ciK^gRYw?^rGVn1k`e75XyZt}R@PQG$B;agnetm72u z6i4D;nOO%WnPIa|3dO9U$SNp$HPC5p3s2%9Kf{ny22>*6m7RfNv_gSBP2GrNXgw}Q zeB5r0zdZiYPu+~TsQSt`PLkyg?6^WA_OOxOKr!nnc&{;+W+wkC`F$yEn7E(=Ul zID09kn<8CQ%s^B!SF?6vt1ILSK+Jy72^mj!%sCFE8k+uyL!trs-dVu(gGKNdx49HK zAU_bXN44{<66K)N0Jkp!e`@`gOvax*stum$e2w7;27C1J-ZLJb&b~Vfs`b01OQm}Q z3dBQB{hsBrbcp_L@LbKiHLC=IM_Fz?^r^QliQ9p@Y{;nr^lhr<SM%1Su;2Hfs$1+_Lj8FIpZ@zQd^6NK6tVULkE+MkaR zY8hc8Iq4JM*?f5{)r*dsKni6@!zaM?6q7=cH7HSnRYg01#|2f&a|I3bg^+u$y%B9e zOXZqg*@mzVab3WZQGWbZ?f8=wknH$(n0K9&c;fW0jRtAgd+t(l%YhBjQWJypfMWV7 za*v9MomCdNo`ZEgae@Ii2+^HqdZCn~+hskcN|_0@J3Bplrj~N+Bs;lh#VA0J_M+W= zS1{74UJ@Vh(cJXvm34&g<}8(Jw$i0gpEeYV2~=%lle>P8DW8+YT?+e@4Rj)BYfzP_ zV~S3G(07p(3qrud!pW|C)^ez4L%g*pT3xGMKXt+I0p@Jw)9WR-Ld(O!`)YzJq%OJ& ztB?B0G51>3W z5K?mDmBr2|T#45{8zmE&`84pnb5>K=Q^)3cUE1VY^nlCnU-le_df&7lC{&Q(&S;7p z=bk#*BBq&xn@TvV)Ce(FGvJ0?FsRY^xp1S$V31aUL@{=EUq(6b_Fk~FES(Ox`txGG z(ZhN3?7C0LvX^YdzQx4FO{bW36iKFH>gZ*p+Y5MHWL>})Q|5KYO(&0)w5aR&Rc;4@ zIwbeWybX#fw^+$My94F%=gbx`kDs&mPs>h8>Ms^+qmAJB_JQx0ki8D<{+>4h!cmHW z9NvSNva9E&u=P69>!CCg~wgKayXz^ApK+SlxEfxvEBbz^w)tH8kmg8T2Jp5)X+n;1prf z4n&wMr@;eFQj8X6MgS|Oy-+aP!GB^*SPiN5H?uM)uho6yE0d-gXQ#d&Nf(pN?6xQ! zcpX${GMAN64A_?InV#%B|1_5Y%ab4T=+_QQAef5j(JUuqZEhf~k=XtcQ-Zch3Fk zdk@U#plA6q9T*xGW*P{Xh?>K@=^pwlu&(~)s(c@m(Ju^chkx7o$!^tdYk?7MfJRPT zoos;43Ew6@Lf@EaFAFnlI7Vk|hIL@4L(9fyPu0ikzFfHv4T?;jrqm0N;;)bBqDHI^ zg2YG!kAiyR0Fz;(%mOjPp0fYpZ)bjb?PV`Ef^`rCNGg0G#*TGEDe}Q-7dhon+ze%l za1x+XptlMk*Xr=YJiYAW4sYBPXcZQRXS-wVqn@R%)T3Rs?2B>g`ESl&_~L>>{>r$X|$|D zgr$e=*v<|S=ii_AP5QCp`fvV6NSp(QRWeOn(KQsaiXtnhnC+e=oPM{@+!m-8sKI#b z1~$Llfx9Pb{l|Bzz%jK76n0OPdKO*mKv9tv+c6ehEC+U3Er`IzJI4yDLvDC4c5YUm z1`r+LCA-~~R8a{*IvKXK^H02USv=s`D60fw?|A2-Nd<}o!D8nJp-bhNOp*ZDJGB0d zk{dHi={9+`GHx0K{=>8UT7?aC6TL6!Bv{!YF1NBB1M){F<1%@dU>A)oD;%4gEXYIvoS|y9EaEV z-mM&E)}u7m2UyKC6TbhatUt~&V&a2Sv$v5_c5yQY4#S@?K~O!#)KFv(6|**^JrqN- z`I8q(^$_kZ55PxJ>+*xsdmj) zw!nT77DDMSFLGbx-Y~UcDsJ%E2X*DH=>u%k$*x>!%k9Z}Kscv0972;f4>t09JkfMkdkgdl~a>SjY zV%Z~j@(5Wnxrcsy$3D>wziychG|V>xm`;c)z01|*q5@72@6QWKSKu8tt>TinC9*xT zMO_`z!+YetN2&G3a*zThZqlIBt>C3{C_wQ}b;ExfCS?T1!(L&Vyr1;Y_-E@t`L&0a z=nW~~@JEb*(y^h-++2Y08C@e8r(Z4Kwf(zB#3lS@?>%zNff0Ah1aYkt1G!LV zvCN~L?h&Vi9aCgNx_zauo>V5C)B7*ITr5w@6lAIDIq8xWk!Qu5B@f?BmwYOK?1Ms4 zqUiB0DCe0^XL;y=Eda9rb5)zX_5+I)iicSyVM=897XAaMvJ-?0#LpwRBRL7SzGN$jS1nBrBN=Wxb>v2&%HVMG<(L6%O-1 z3SZA6%Cvks(Z_Jd-5N%NC>Bq{L2S9eW%LR4sGeniDglVwT1oI@M#T9Ws zosE}p8w#JlAkoQp$~Q_&Cqp%gYdchn6nLTVY(6MXXNGB3M;ufpf}9K{^CeBP0yd_D z_4v&GV0!$U*!^S$rU?N7NA{RUdDuh%9k_zcLQDwJm8+_P!B6C1MG&HF(YThZJswZ1 z5cK%war_zm-iq~<(ZD=%e{`G-I*l;TIq(X7qlwW;q?qLtNuXllrzNN~i#=*W=3~|^ z)~o*SPUJ9HSvDN+{Ov5)-w24Jp&#rZna%{lW5Wm%N+@O9ObWm?qilmVT5ncad+rZqlh^#>;a@ee@S!5xn>|2mjVQxsr_^;=r?!g|y2$ zfOcwBM^ye~O(N%}TemPbQgeDvaX64YlqeT5xK&l}`^XywKbO11Itex)Y|qs>BJM^D z^$|O>VPnXl-QDJF3y#XtX^B5!C#+W2hMb+e71ko%@wVUKopc8x#tR{qP*w(z)245?Yb2w>6A8J}{wzQhA#BtuTexCtJM zU;|s9gNXz|8x$+U8iC5aQC2QQ-$}D6GA|4@gLU!;bai;OAq^9=H8C0YJ~;A;ocjb7 zk1{#6L5Z9$VQg@rxC?SPGs8}JYhq^|2HqA(`UjPoy+Mn^@*xifR3{bzmm}87e#&3q zG3123#R~dBbE3ujPDFV`iagmrLv)>#3ok;2T)$i2%yJb7RSe29=-8-A(kRo(bKnRZ zJW;t2AJO1H1~H6RgKc3#P)(;uYIJrEfAd?H$QlQZ$be${Fvoo}#lSjwBNekSxQ8zH zO%NQ2I19z_56C@P-wZ=~)goSsJZBKB-!^?PoqXoNAX#QY2=kC)zMx1S3X@~ltQWLva7P=BjXvq2 zou8|!q4y}`eRf72R~;b-fZ`?7rN(_wme0||3W_3tD2jtg4?Q%p2eo@Fo2|iv?Nc5? zuHEfGOv%BLqp~@9uw-1xV8{Kgi5w)D)hbpqQ0{nF1~+)r@DoAvxr`|Y*dClVbqkFe z(3<_QS=PrvL)PSyN3Ud-aE?QvZXOWRe&G!Tvk5++yg{FXmjSJno=-A*U zspfIT6X@vxeK4kFAZduf9D_9K=-~Ghi1s>S1-}VD`7rk9<~E9rgYCFWO=7{KY5n6T zC5G&jtmYvuTGXe#SMYSQJkgS{#S^bYER$ENw2FAn3SNskha`sN`yQq%=v5w9!oS28 zt%jcQGh(z+WlovL$P#A)jP1kbi**!}Op({XdxuDS za6z(D8XvR`yw%ka51@p!IwXTWsyGwUA-Nz~pgua~y0n0Ek<^4V!CmFyP11|xak(?z zSHM}xpym}mR~~_>ReB1;!14&}0oTT;a)GinxvE^Rhf(p~O;X%4(#eqps!4i5f>qE+ z>y_@A#OtD$yB~nBUF*LxOjE}{L}xJ=q`odo|_?$3@{UBUTQWm%p%?#2h$E zWT99F9jjVlv9L{!J_&4@9`;-li1L5qYFzbZ(n@lZ`A;F~)S{W=@;%8W`F_!9Qdl%c|-rv#}bN({H)*-ZAGddqn_YLD9_+;W;4c znHncZBZpOBClW;&B2?VKIdPTi5ds{&=$pmP%BPcf$Qh%P_e4H~<<(|;*WFoZH%3?{ z*%vKmL;m9%PDYoe=!;+cg~YNeW^iD;p*67^Nffh^BFm^44E)6kI#soP`h2QpkJX}V z_wk=dtgO4V@92-%`*OFQ56q{MXY-03*o&|*VRQgjO92OhA+@tL2C<5N#Hc&$gkgD& z_TJy3KQz~Ibl5GYttManCW=X?$T}*fMbR(8>`MLU=cs@KS-8r-*mEVUd1`|&z}6a) zABZJPsbQ7MQK!R=_&q(Rk^QXjoX%``$ubs3$iJA-wBYPgbA$1UxEdBriGi}C#I;j} z<^KTBnY=)qOUy$D?hL$0yw(lo(d@M#tp1pWhtf1KQxavSq zlS?hHO_-%BrT5t_LB~q3VgxtS*kmU}z~5!jA`EQ(Q<%97tiw((7W8Q^NR|k0$u&^a z5xGhE@Xc!13|jvaN?tK7|1M8Nwnoidus)$;{QA>c<@VMcl7n#cAq1v%oVgsFB~nFAgsXt7a-;C;DJG3OAN{HP3#; zO9t(%KR*&5ackgTQ`ZI`^xZh8RE}rO>ZW<+-#;In@vl{HFMYc=`n3H5tYd|?EnqxS zAC{rr{#O4<^<$?dY;ro;!!Azq6>1apkF{>>rn zWVL)(@WmjlSAo19(hycd_DiR%8M38X)I}0VS(hCr3nC2ELzqJ~(o=Qx-pMPV>bp^T zh#rVK%W3DYmKVtTX6~J=Rdo5ZlLD_UX^Xl>xtS>xJPO?^xHx(A)^w~mGoR_q&PRff zRc)GzHrkvE@!24Y17oJ%1T*Cn0|^M*sF()t66HZC&>nDWkX4Dgy-p}@$-j`@ljXUr zmz)njI|m)&7Imqtd?MbtfOA;2f+^u(8F#NdTf8S=y`(>MzZhrSOuvI}$Q^MT3TA8l zT0=&U0ZRj7BWzgee$?%jZO}ruk~tis4f@TeeGOzGI~(M{LAxvyJgudeB#Nx0V)iL( zgN`Xs<4>z-C$W_)(0*T?_W zGvkW6K!n5E4hxbGTf;DDGvIzD4C+8YRD?!avKr7XO5`PSuo*%-nBWG%AK#9dnk(WW z8nfVvIE}J9ysPwmcYM5%+efOWXcc(Bq5KU0wKdXNq1kL5VsiR?aI(J23eB%gYPgd% z+32Evd+SesK{m2;Q5|^6v)g2PDx#PI3M$-UGC@x-GYlDa&jq&zoTOI=q`ED32FY@| zL|MdZcdvs2q6hTdS%>I(%hdaTbXS430PQq#xU}=TAP@7pR0m7xbMj(goxSYm*uljb zH H@AE|RFPBCuvi(2?0`3|vp49^hxmdTRLH`CTKIV^a78@H_ST&wDy6I={4}Se+ zr$NLp?9w+ci>qC4f}{+J!NZpi;n)_nF8ZU?H=BU68@b97IS)m9JqD*H1GSw_{{KEY z7mZi2t{}_pb49zv*3-<>5n?m=CKHU|(VxbW!REMYH)0`!4TKlL zYyi(L+K}5I`u&e#bGCA?*~Z|orn@!2IlRN1&&gpOR|^bH$lYH@gC?ddm5X)D$+OcX zhgJ472A1EeRc6BS<955k&P+7E_vqcP82!@wU;WzqWS;|9o?bEWG*40t5OW^|?k1i= z;2l2wBV=2+R)N&+Nce_{jir<48}_xK9F6Ia93-%<9V@7vwAkl#NEf)Ed7wts?eYnL zshutqA>mgIzne~QosX1-nPGTCuBroU2?khqcw;qu%EbAFBHZvU7Z!-;$MZ7i9%;Iy zgDy}MidK4d(eshg8>^f!`5|Q@8*o|$su3ogQ8lrEAS+P4_D=nV0Q1PJ!x|b3A>jp{ zCE%VkDBD7T88z$xH=e~v_vyLLkRLo7c({nOJnq&~u#nR$-`r2#|OwX2gRB^s9a2_L85gv`aIyx>s3E}aLCb(ikci$B8^Iy__?>35>LmjQuwxDbM+U_} zdVLxdQz)o=C;AaPuxN^HgYuUC3If&e@~HO%$ZS>I61@ z3|nY`^+bv?F@=Iyfxcl$Sh7E+_8T{0aOW4)KmvEdYFBg5ZGiweP3k!(r`UdX*YMIu zH~uDo7-96r-o(n*O?Z^8T5i6ywLsm`3~}oF^A-#_rDA%-G}?F8!09YvlL?L=+%)x zmj|Svu|VQpplau5&>i6**%6X1(ID~X4)1GR44$sw+RL6i3r}lN{=}v{+l5_PJd_0= zhZ^CL<)4#EiXAxqe$-^ZYbXX-+jmhhIE5rlzXesC8jNi$m^9$7xiK?O^_gr>z>w1w zFTB2EdbwA5L>AF0bew%bphXrOuc~%k>kai`ctg4GWvI(3kLV1=%TOigwaE+n8gi=V zed3)VYI_6X@kN}B0IlMJWV7mys|{jjtRe9P8_zu$GfY?+z6o!1R{X2n2%vSVgU*pu z2L@1u34rn_CYvIeRLqCf%45NcXDtlA!r;$SUccMmzlJ;@J-j6N7kP&@AGYZ)EeuZ6 z-#WrA_M14z2$V4c!cqsUfI_AFt@&as6OjXFpjnuR){taK2v{9aML!Hmp53YHh5b7q zJaWsJ+M+%Xlp;T>NcSw8+s?nkdzsuV3q%AHIkMW`ris3bSm^u3&MZzL4{~-P>%l2Qmps2$$Nl_!<)qy~OxJ6-N^P}$3E%#w9 zWv&clk?mlF*pUMp918+RnM{Ey#SfR-?Q~|CRyA+lynK46v_o<$xDh~7yU@=Z)Feudy;K6TTPue}5WnoT$Ei;QJ+HFS`Y`1Dh4lR~@zy zbd+L%wBsO1{0dUJmqe)%6|!nqq$&PPxtfYC#!ME&Lra{I1)*1dncwf0A-W`L5+yh{tFL|g)DJtp zQSny!SMI*G>CM#d?trIGLb$M5-T3bP?_G-iC!M-rS-Ywgd&6Pg?Y^AWv0MS< zr!(jcdN~7BI?n;7JguToFyz!hZ&enD>y120f9;raT#WKx^HEm-PhcA;)xF4XHeH5vsVzTI~U;*w*&Ir>v{Y<8~hMxen z=A96HUE(uu&}mRs#+i3SXs|CQ=XCpA_ZxI-lkVkcqjKnVqDzrw5OIrRisXY%xuQ03 zb_Shx&0WP!;o{W-PPtdz+&I$4*-Q^Q-In#p8f0;#H=;$oMv$rMp>yav%-C$msFCrE z$#|w=OxR&HnNk;>hhm-+8jmfpJu6f}?zj$)QLV6;mTT&0Lo6obfQK%O};X zLL4io5iW~50>Wl(@@m(eVW*h}euDG58LLCjsp|j@8KN4Xtlc>yo7?4+9+(3OJYz$` zi-v0qNO@7$W4g``DLqXqDXtMIPJvEwB>t6|mtgmLSiCNUV%AV(6)0iOZX|=UK5|@e zpP)}u69QI6cw=U>v{Ytq%se}=R-@?INA23;x=0y279}3X_2-ENB_2rn1Nr3(l5&zV z5eV|rrgo~Zx?@nb%jLE-e|n?KNuj1p*#%;Cknn&TMHl49VQpCMwMULd-uggxdB{^*}9 z5Pa@kA+8W#n;pko@yql(LOQ}T{W^V8eLDD$c)O!(=mzgp{iIawiY-XBbj`I#+$enn zDNysC-ipGy7lV2Tc6Bl6khq>R&qkBd7y$7sMD#7s8)oaTv73GV+}!?`a3g$#F45Vf z#DU>++yp*#6jM!+-BipLcdcS2lL_er7?ZsqS>$#D;zAEN#X_W=!|Xa8^MG@NE&|wK zYm>D3|Ficda80FI+PGJ|A$c)mBba*&ERhJZSjA#!Ks#F1%h%J}^z!d*rh967T4tv0 zwZ5LT-CfmHh~Unq-~yU}Y_h80h7trz#idlRKygJ8#HFwlTqsK6J0}TBBGKHCFwy>( zzml7~1@HUZ^PcxC&!f!Yl}J#6)zH{3Opj{kABLvH4yH?S(lkYlOCU zlz!?K=X&-sRZ(b`v`UA7$ng+zOd}B|2#u3mj*ofSn^W6`&B|_rBujy+H~`z5kSS&U zc~~%~x#Re!`e%8q^!`}9+WQW0Z#g4~IKmrl2R=r)i{p23!pHdEO%xud{dE59dQ}%~;^4renhr)k zrI+7AC(2tD8tj0_K=!a)CGzw*s{-5(Xj#r$j>i?;al>BegMTbywZc*gj5Geml`Bd)1i)R|0JTrOI2CyajCnY@mWZ9Z3hR!{5_VjUBonr(9*k~&2LNIZSaTc~<0*zARbZ!6zk5KsLcPk97U zmVD30f&sV1Z$BhANOtsqTLSfv>)nTZkc|5e6uAGy~*HlOD;`e!83#k_ziPkHxh2VjvIfHHm)Rj4S zVzct$H*a|-&MNR9bVGuVTKb^k)RZit1}WxrbXkuq1 z)KgP*w0?@FSXxB)dsfcr_k1K>>S0PuWiklY{Xu#G#+;3}toAz#Ue#u=3%pz2MXx8? z1YA2{_pFuTy2l1^zF+rYtR+S1D^lZah0t^h|DqV18?$GvnT+eDB9Hmaz^2|KOaYMv zOciu6MS&gc7=?auh&W=yhEa9QDSx{0yH9O&dfRDa9Y-OHzNiClrp(InG%%1Cg{<^G z@aCb&%hIaIoQP#iHV^{mt6JePb1r1KLuKd+>yR12 zFpcvK^%h$exx7?{sniTw8!0x6B5SEQ6me)(=kV%eb*cus!n18ghaxKwdfDrFbr9Us zDzO*7mUk=saLldnA`d%E#?bTT@{>DexZK_TbFE|M;cn@|=_WJdCQAx$Em^sEU4y3yH`QPJ<|(pb0s(!RlsaR$m?{0ob32i>~m%j83;=`osno}$cFa6i3b zY_>Az6|S_5t-RW?*;C>?ekQoMz04))mgtX_jZo1N9|{K1t?2cgzKL=zL}SJzYWQS; zZN-LR?s#&Q^EVQv_-j%TXpzc*9I|gfmGJdPC-s3a+7VJ6Cq)QS*>wtbCUbHtR1&Eez zPJSE!lfL{g{b^90B9YZ9Pfxz(>(Dor(-SwFvY-9n$Z_B_)wr&kw;*KqJL65jNn8?o zkPNy(5BYFMB$Z-UQe+twXO!!e- z5f8)usvzwy965;}KOCw(6#E%PJ_VmSxLtUFpBPXY*(2zo<7XnhI*L^m@J~wW*;90j zd;o|iFH7=0*GdlXb0Q817K(C|D}vYYxA4zFYwFVArZ9-9xNq}0L3VmP^6lk60RHWr z9-DcI0b6EWfL7-VynFmd41NbRvIcEpjr$V<>r@B5AI4nZt@7UK(W;IQt(nSIfB!&Y~=;NMq#TupLB&^i*AFUOIcK6zyPz9ES_1+TNsrWu_yA1 zxPi|0KOnsvIdtU+merB=@mY&$>z-#{5>9Uebv-8k)axeDeW&zSza|@87<7Bg>|+_l z!pf}>`_V4&P{+`qQ+5F)?vQa;7l;AaOuhz#btnOVv9%f%y19lXkvnXHXg>p+`A}Y| zRqc{x3fiG1sVd4zevKpG^6X%beh(+OjBij5u6Wr6XgcZp@-`_H?J}CRyf*12-WGnp z=lr4nx_Q?irL;e|C}fFuGFzZ(gz)W|Nj-`@k6jZnmwsQ4XX4K?-h31T0QFt`?=W2< z?qH5b8a~V)O#m0i)DdVI*@TP^F`Upc?)#qaMOrSLTsHA&B>{~cVg|!ofoC4O`46VXY-|HDG$|NG znVD@q7d=Wu_Hy;V2q@1v?yA@hyZepLEG1=Lk@%Yx0h+>yBD#>?M>KupxEEH@LMFNu z;%>Q9hdP^)9JSr6iJv9gqR3V0BQY+J61D}h&G^YlKI{;7Yz4|l9>~x=obmI`oUeIW zmhii*zhR|Z|0KUd*&|;Kz9%eHun?gT=C;U%KS;iaMruWRnYu50`=yaG|9ES2dYdI4%}YHs=z(Wm{cMnEMLm=&-l zSD=*;2{_(c6bP!DK)-??2HjRO3q?m2`=)l%F zCfgdje%8EJ^%j0FA2G0UVx3^uj80iM{W$z2^vIr@{Yd&ad>hh!3-dt8sBF$ISsk4^ z3ps=I0z2&A*Z}GoTlK;noB&1rp`>z-rGT2t&URKL(GCW$;GKS>MetB@h{TH?g`A>O z!VdYgDUqhIW7e*q&Ae^=Bce`OO4x}=I|llNPkh7^xWZ$^n>b-7H+U@1KeO4A=h6y6 zWD2NnUzg@cHA~4Cp6%h~kq{QW3;gY$D+*|IK*sFJ8y*^m$M1xPZ2gv;9>42Wg8bQI zdW2;N*NRlPQ8{c+U@hGM5~*i-z2X6ILQ{C)KNLn3sBnHZa=@W`u+Jw&Fd)|yMr13K zBXp9Sh;|T>#`vi|@~A+oa+Z)Z2eepm0IZMrXpV3~i|YntLXOvd)lv)K**aX;{r!m* z#($${dmfzH?nq5LeN?sHeFd{it|?X(0O3ZP6q82T&MO(o&Npnwg?;%BW4>`|jy2AH~D}V!BC&BXYTKbfGf~+Yl*$bFGjITu>bk8Pc7Xo9t0k;b+Up$%?{}f2^bv1(ii&Pi$Z{(4l3_*B1%M?=^`2l z3vo*B7WYEm^gYi_GED)j=kJh)IhUlG9VCOFl?@Jpagx@s4M3!u_ zjb_l=PqF(bQbWb%eOC`b!AZG%t$$^Ro^6JbhIneh={fi309+18(c4P(Y65~7EYI9A zcfkj#{9Jywe9$f3tJWXNAA&m-S|!|~Sehb85CO+7t;qpF*T2=$nG{sB0`FGkrGVoz2P5mUVYZ|_=)&7erSu};arrN2W`7L_A;5S1O#3g*2gcrm1V9(h(re!;i*?Xv4%ZPUXx z8EU(}z|M=TKkN~jNnH0`tGlnbY1wwqC2i@tUK$kL4X^%ON3od{Sw+R6mi^yQ3n?WG zyBkkR^kGfl1jUO^NKcCDd0PaB=}M<7T@42v7le;$=`;LGZP?!ZoqIY<*-kFdaN+f~ z6+utzr2dHRn50#f20=+?CX>l82q*yUh$^vOs;LdbWBIC18Tz+P-laj%X+mSg`zV*n zHjIVcwRX1Ldo<@HXe1%KSiR)TMYhDPFtR5gJ1fjut-UDQeRepy?%NDaSbnaOG zJoLpc{yxj(YXmG>QA&1mYfp6H8RfLujB=P_4^m`56}Q2COXvl9Q4udgJ6iaTUE91oRqUoqb)t zUC}4~%(pr+kAGj0F3|ZGi16;M>K^(dU<_X@Xc?Wno-I&%{#w}j(a}Kf2Q?kR6HHK2 zcfPA33tf0MtTO{gI>n|?WCiNw4SuO<>M1%$&@H|e{kg0#!l3FfrarWB3~U?a(TtaU z*K+fuweS4yB#uJ`=ps!+0O9 z@^<)RspbUPuIyD@2*~yrZR5yOJ27hLSfJ|ZokzFTuT=ifa#{Rr2+f6E6)TJ1N}oQ@ zR&_tq0gZ>4|7#~&{{wLT#om>?N}sdL<9`B*IXX#nhL;lDqSh)KU)XRG4Jbqc!cE4Umt1O25B~2)i;3~V5X&{Ry7@6&KbpNbu zf0R8+;cX5`t z7&|`N4n6k23$Iyvyj+&kE(|U!{&9!rUe7wv4oC{gnR>|6pfWO}Z;a~2&Bw(iOqgFX zCZD#-mVRLwym#3WN-J4~M#bvDVs?$6rrV>Icbv3|2i=;LP|fFIC=$nl?J?6&G>n-| zWBue!ACqA={$5;O_HmhYp>tmL*46l#NDYH^^qNfQD%nXti{W-{|M_Y$w(O|RkQv-r-w|xOOyCTXW;76%wtprco41@}Uk!2A_ z6@|2>o`==?=NZ(q+89$9u~V^$)gbG~F8*g?lvl*!f5$CsaQHX^>4~`+Y2#Tq*uTd* zklK##{IibWpK3E<%z#=J=-|4GW?qDuBH7PhMtk@5Q7hY){ zE8FI7UvDYO`cf^CH<@82gJRPul0wDx{!<&2x>Z2DteajZJE`2NnE#g>!_F6*sY;c{3{3|;Q)NjrUBw(IrfJPlH;qCE2$ zWO^+|%T@%8ogZTS1+61sBsW9O|G$cty{@rI-sFwlrUejj?}v~Bu`6K9PQ5n=`e%;H5oKF|SMOCO1@02v`HF90VukH25i7Iqb?4DddhIzdrLRg?j6 zs41@n>h=Utu@vryn@FpN+%2O^GTRKIkz9`#?Xv(Zmyg^)bnf%uzx}}EoiHl`ze!HI za5}KtY<6j*Sm;(gOU3EKPP`ExdcgOwpr7AB$5XX*r*LD?jW@R|9t&F4w*-g0KQQjv zE-mIABA`LmDSRyGQWx_Ip)Mz#PcYN+BI%xdf1o+hbW|wN8;0}G)(}uN|QxRqrfiK1y zCxlTS|LnIxUp67^=QmDelRXnitJy#vrPu}v$W!C~)+Db5Hm&%tb<5kK>|H0h0+Gzx zu+!e@QML5(iAzPe_cQSd=Dn>RU8;-Xc42}jQ?OCpO=nJg9~U3E3+ka2dapA7j zdHkB0>n0byj`!*1?*n$#zbzD@j$9tI+;CO=%p!W-WLzm1^ZH0FAHM*L=Iti-cK(>X zczi}WgVUI~Zk5^T!KmziFa`49Q}hmSnvj*bme%?=hqV)IC^<&fNCpEQ#?--b8_iiO zsFGio_Dwpf%5g`r=#x%MO+ST~onU`@11D`oe!useB{#Fn`pZ^Gqrnu0A#>&ZH0GRnmm4{UQUoxIp{ev>?t`N%iVt5`V@Y!bAt4NDMx zj|0l3-M-E0OJCpj!xnXmuvSvc>)_p+_E=C4!Y^II>ozHwa(y-q85*u%YQy=QhQ@W{ zWAdsm^;-&yxU8|UqIiNuw?;Ng4KB|I6a^B)23F(_x~=d!sKVVS!s|$z@qCPMhVAoj z=dh!kU_1W*rvGC3KUOXG>B(FDb_AUZ(Hsme1JHCS(%fxW>o{z9p4`Y5{5&~%3^%g> z#j_#MG{-c&wW^xrbE}|pT}udc+ry{I?G(F>A{A8J-XN{2m{&kQ_T3xMM_Sc)nUt_f z-X)kiugkG}zg~8n><_*|_LF+SLh35B1k}*h1msBT=)K+x-7{hv!>VKJfdpzZd`|Vp zE2?8tB-ODye2=Oe4lP??W6f(BzTeS^x^{GBtEC}%HZ|bF<&{=6E;?bwy)W9(_JMQE zm^A*3G&Z)3`$*S0d!O~`W>v?J{u*jBB-g)ttCAGEaOe&KJj1-`YKq-Skt(A!4m1va z1d^+*>I%Bh{|Y$Gt0g-BJTIN3m%qV1L6jGzRaF4VTrbJ;-{zCaM@|qFVQm*y&`-~9 zRX-pD3S1^v&;>Le!=mk>dpXcb*n*SccHsHjjzf=ioQ#k=pl1ItX2NLVD6ShJA~5F+ zLs}li=1|af7MBu+X@vbjk7v}+hCDzKAm!X_T_6_8A!h;x1#6T@F|NVb;(1xTNb}&$ zo@wW0OGLM9<;$>lH0{Tfz29itb@6|W_=@Fn*kzYCR@UitinX$K;nrWZN$V6j%0^*I z*dmY1(nX!I?^Y{eRi zT}_cRDlT8uE`(Opp*$hDCm2^NpPm?cF!%t$;uU9%zts`r0t~A+al!#kz!=~7_Nkke z4y?->6D!W@d6`a9Al1~UdPsa|tr%vJebHkI1=A?nga{{$_=|4I4H4-<6~DBsjB{B; zSgC?*7k)ISg4P8=HwFC#2*B^2y4!Evq>p3v2kSi!^6C`VJaP3~>0?aMU@rO+b4ZZk znJcNL?#XuX%DpR9xe{k$$R?}|n?cMrUo_yOIL-+x)F%rAkGzVNaW`f>Bv%0^)skzT zCuG&qc@1=RP!c&sXYne;&VaD{>glYDmg|^jqarWgYNh$>{hsl& z?s|4WZS^&N{H#poQb0D8EzK{0bxInqkxulyA;hS?CSO&oik}6or(NMGyj1@lAE!5c z*o+c0OJexoM2tL-6>v&k)fvm44o*!WuIs^C5ii4h@RGN5bhgOQ421GA$S;{Dcr-_Y zA`a(eCrJJ5u}kJ~F5Gu2Qik7r;+ncP%dTXXbwaFkAh*O;MCv{ICRHi02nq)pmMrzM zmRS7wzY$gCWFm%-hLg9gPb4D*$;N*-^?&ZX>`U1v@~9a7YKM~7W+iZ?R47h)7pQ7! zDg|2;iMeUj{wXm7yH-D%w*bv zPetwaY*lDhhn<>|1U+Hc0C0sAh1Bz^6utZo=BBvbuObpCiuF^nyt2Gd7m8m-CkdJi zI%#;e5ro*}elJR^#Nd4y)#dTIV*Ru(6&~xN5BY6ScEhCrVG94)Y4IG6y=UNPAO4p7 z`-!HPy*miSFtOF6S*cZN{TpSNh%wTa+c{YFJxaq>HjeWszQMjrIoX);U4ML^#F%W% z+9E{`**Q|%sS8_@W;082h+;pYpno|Il@URiu3FMLWvA~kk~-Xw{)e`p3mrC~rK`IEcfJ65g4J2&@*E6QUrXBX-$_RR^>*0S+D~fhA0jccsU%yA< zUzy2DYX-42icO}-aw-nH*Kc_vwFXQ#pvK@==+IJW+egVJu-JY(hn>0kJ+a2i1d0E+ zNam6B2?Uq~hPj!!6blbGn~J-_e6DC0t`BY$qB;i1rdP+VXB+6G>Fa|vD6=!9i-ReZ zzYn;5@QFmLL@^!bhbC-#xFfxG4ttCv--w%;ciG6)CtqE0i8f*6!4Le8k%S2ZDlvv( zWEI7xQe-6+mmtzdVqYedW$5T~k4N4*dLyh#){kwL)i4@1f@0*6bK+i3pm5zu{dG+R z-%^^;WrJ;2G!qR%4xM!7n;2z-38I7f-1h)ftR_d)@-z8MXN|@?;Eo_00gHBw`^`sj z1t-9cKNK^e*V4yuSzK808jL*@#ZvGX*2!*#S4DP#+c~yiFpif^kQl|#+IuM{NQ^r( z@lWr@ndXm&VSiPSb{CGZ#hWR1_fYI-6#0~j!e-6(vWwDIzUE#Owvk<7`UoD%oH&2}d|dcoWm+$@ zeVSH@T4A;P?bEhG5uoJ!u}NOf%Zb3A!Axd{A`v)w;yqwDeq{zYS?N&_nWHa? zJD8o441YKPm#a1b*X=!VGOUpB_9*dLN-vA40uSyAQwWmYhSPGrI+PjyI{K`*o|mVp z6(+M?;jf5PcfsPtewN9gXVb|pTx4%0LDVf@1`3oVBCt&>n5`kEYI{SPC||W#)lFAN zUJDsZYh{y-dg^@JYjQYWXWd5Kc)Fd7nTo;bK`NDEWd$e!W0Vy` zvhiaf)ADo>xuNFib7{eQ4^&&`XlO@U2jXeOda*7iq(= z7Q1W?3LFl7!S1Prs;#q)=jy|j0_y?rA4FY}G^?xWOA?g-cmUJp&1rhLV~g4rPmCMB zo|>moZP@_M@BOy!^?xke^EADK;n(fc`n?V9q!^hS&hcW)pxW?!Z8?P-Hf*`0W6rr; z_l5SGCfidr>Cro+ft&4d-CzOfG_yTtC>BxzO;lX}tPFlmgboC7uQB@t8cbxPc_;&^ z?tW$&lMN~9g@QJJO^+PuDDj+jGU%2Qfn{@e3i`>b!!W+IB%+@ga7&Luu8rgv2tTDq z<*S;Nwe)pR%*9iUcPhf9KyOraEcS#LrBpT86$ET#{dE59dKFHLDD{os1rdQ>MRo+{ z3T#9tH)M^<)|f7~4S8>c{JQ&P?=RdHdU@&^el?}{Ecb?tqTdSM8vZfCJ(+P&CbQ_P z^B#hBVG$4OaJ9;lzKud;!tVwhs5)V#PpjH^gLYw?v{O0&t?XHo+QQNVEwN{4y~h^* z8EGFuUVR%!+Lq_;STxx3jkcX*-7_Acbc#Cm&ioI4W3o!`r7nGgT%SM|n=O+2DE2-@ zdVm|0NoL!HmF#0C!E3{$7-)paGe-Kg%x0CqSK~K=H z)R37$YMf3;)^*EI(a3a@J5`g%?}Xmar0GXNa~q1D;Pu427tp&y>r`50SyVqiXX1S1 zCnyp1duB&m0jivS&-0O+Sl}-%qA?foXbxr~j9&ro`E_|QuSt%AE;YeuY_6z4U3)p` z&Yzo|9n&;%(5*zY)I*a4vaThfl^$RN+N5`8rtq--;0h^*PIpLbLz2;e6UUCQBkX}O zHppp&U3c2}Q^Wtf{aX|CZp5ZeCpSh)e7o=(eu>$8@;Svmphz!FdtS9+#`mE_bP{Gf z&90yx`jp2hkKKO839y&ZJb0sX$~NBe=py>W46JmigQb1eKd*c%^((!6LvJf)&yDYh zF%5)Sc>83Y+BlTzJ=3M%>vtQRL&yqe* zoT@ueX#AG5}g7 zdReRbo@cgyxA=~tkEHOf&s@mE^IF(0#Rm87&S;>s6=Ki#%N&m}_1d4_m~EPRntzg5 zK&oGvIKUaR8T1In9-_!cR2(`fMlS}byywqHlHfe3TR+AocqIg4d0;glL6P=6hoiXW&olt6eJ#j8U#<> z=?i^%kiILHrU*1g$TlA%KgfLhL{1!oCC4MJFHawC=|WraO7$MtWY#wI6fYGVyG}@Y zg9lSh=hF*CB@$SZ(ctAJ%KJQ_brF3>6ZQX)|JY8W3sg_{i8YVr-ttCw6E45(qoX&; z(8Fo=PL~Y2p`Km4Q0reAf}>YNXEQz1Q1xy=ksf0>`4bYt2v-UI?xe)|00Rnx+jXjJ4+* zfHbv%zCb#d1aggE8lc(2zbHn^7_IUp^h!sVB3yU*GxN~CZ&|v~E?Wv>#kV)o0F>|( zd57sH`f!lWL<}%|4~msP4EhM!v{$uWh2nnQ5Jmzw z`*t|$SFDNH5j5zQLas-)&q?Lq66C7%kp(LKltx%nJD*2yRN!$EaPYR)!87Za=kF~U zB&^64;p*j*bafa!>pte9xL2{9Xj*w^rd{EuPa5-8jqOj`iH~)JTU&q9Bg{9h8!pUR z=AI!n*$>5sUuh<(E?j3;Ze}ZV6uW^U>!~Ox6qq#dnZo zHj~-TYc;UkIWS`}0ya#$Y=`bsAJ+Qa9Cw5pbakJ8pJh!5@%-Y!IkJ}9%F2bSJ3!%n zSV4Xf#R4H(E)}=gt2%ajK<%6b$LCyy^SgO+KC5Jf~HSTMxvKq}6j5 zETmF+e;GQfS$!3}Z5Q@{K*@G# z75^wSb>EMII9|Z*zz*i5p)efoPiM-po_9K4!hTN2&{q7j=t|EBBFP|00zJd9i?H528<3n0QfXvpEfs>xScgBk?VqV6s8#&UZCrp$pp}otX_v zr`QyVtf1nMR03*vFq&fQ^Edu&zge(4=r1~88(3d-7snsx2G&(QmDQGtM=w>Ej1?86 z!_pPJ(*{~3NLru`#FDLxP#1GwdLyEVR~vSXzd;!$mM{DEpiX#CWah>U%IhWuAPIXW$8q6T|HQHif`YWLK1 z!9|av5bVT*P%Tp%bW!}pD4dj$2WZ${t{8fXCfpi!BL{E(7o*MkqiJqgx-2D@bi44} zvcjw??=i*pQ{*8Pm&HqFQ6eD2U#lt=Rr>S-1>=BQuVR^ezztN4CbzzEEoA7N`S}ph zFBKhu0DXza{H$5`0-^j%a!HlsizVr&#CiOOfjuNmfO%9yy8~X1rwtvx8L}co^D%T| z?h7fPHC1yk%h^YaUxIxSCrKxyt}lYlHP*HQFD%rn_A)gyAzuyr_3iXYg2&FwdSXsb zzU7NvTn9MI4}dl+Kg4uXJb24Dd6MD#F{7Rj{iJv%{(6`JT`20D@@nKj$9*YWTp}{z z@7=12e<3R;5Gc+Y77^P(vFkB>7iY-CU?p$6uvUDPuA{GpSBataE|W=*s)8O`{S=*K zK@qJ5#xXo=Y+MUF-fIWH!6VVHd?h zN_ra=cU|7WY-5h966H-Gsijq3l2*rlK^{a^1vSdCK(Cg*6o5)O-QgX)dDt)0KwtK$ zr46SU~EdW)%1s!pYO04#EcB z#=1HDp19y&EmP7iTkB#aSH6t7uP6yv2HjOy(T?k+s;FDuIA21olmnllO}DZY0yf>u zfhSpqz(_OV;cqA8`kM?&+2BuhlAM>!pwyZfl&usCdFv98Riaa%s;-;Xt5W@2)fb>u zs!CBvH_98sdPoQ8dzMQViPB>G>1t*`QAGF1_XL(rKTZ~_moukmJS?ylf zO$+SQ{*}?t$VVE(D(Eg~j_H#*Ww@U_UyBVcFtE7PsUNMwU<7M2uJWWmYiUimbQ8F) zgHEj28SEL&Wb$X{(2Hl@ll4WVNP6gFyk*Sp;5$A8Os)#q+OT>qFXnC(^u6^)-FP!S zW_e6)aEG!lDuZ0})Oz&;U34bDf98Dbj_Zot7z738j>jGsfIVwrp6qcr*na-AMY}B_ zV0dOX7Ew5C%B)Bom@d$vSa|8$+jsX0!ivkogT+>;~GAQ=g^iYEz@1n-r5 zV88eU@da`7P@lIG24&P(al-F}x8#JC@w}so#2}HJVCJ_TXC3* zYjn?(8i>nqFfNKgl(U(K(a66oct43~bp8)Wf>(xrGP^qL$}DWQ&yjRP-kVz5p>)-!T^FQzW?w5c1jdU@^E}}@h zO#rs#B%48QaiVq1f9Uc4FMnZz(J%IH_&PbqE#T|I0P8aY*iDMGOpj%cT zQf{{kjdVsW(+Yrj1s7m<=ytC`x4o);q|iNu*DcP8$YihwuZ`IoguJC%l_rz93#>}W ziZtlfBRAea>knkeP=&4sNBc}rp}VG2mIj-+JCi{mV>l5%L6&Lw$8#)@oPyHfb>M5l z$rgX9l(72Q#thWZubCn-ehdEUpnHfvorky6)5tv9DO>y4H&()T_Hete+8#(x#&z3d(izOWd7fiTtjv?2~0IjCj zG>RlsaXEs0b1y$^JtUle3a5Tyuq_OdwFC zKdg8`OR?D$S%+?9Esc!5hK1Di&__}O$5WL;KV`-A^RjBcBIJbz#ZY8lYmIsY3ocmQ zI{*E%@yS}ws3Q-P1NT~Y5Jw2Fy6!X_fAPeBzU;|GT<-LHU5==uk**%%GMl|l(M5EE z2&*%lk5F?^*^F27%bfm+6JE!Se>W}6QkRE|kKw|zi4}RF4kac8I+zYqvAT4 zN5RKP3lzh)Ms4-bJa`j>Y`Y{_(XwLtCLp-04cfsNNTKeypClEa0zIt5 z7F63Qb{j=1unl^N?_hxO?`Ab}H|c0h6W&+!GDUt%#dV}da6+v2D2lo@yGM}ge-896 z2Hd)18lVuNM%XC)RB=U|5#6O~R%d+$Kewp0%8cm6GdpGu1~`1q8Xat&+7Fwsz~Llr zuu0u@t<|#U%wTUiUJ+%#-$b=qq`xFRDRO8LBxi>wx{s zQL=DichO z#(l7wRK7CF;YKr1?Wfp%6se)&N=2KXI3^QfCt4-eI)OS)CbKj4gtWv^NQXfcgD%Ex zdB0Z|EQUa5GM#iOFms&eQ$IU}r$MQv&Ab#|2lEBpPV&6A@vZ__^10a>%=a7A-%)Vl-pqIupeU2!#COXZ0=&rn zVq^tQoK@hj*~UB1LQ5*C!9s&3`89reIN(t~b5BR?;5Ie5 zY^d~X^mo%X}1WCJ&Id(6yD8O6eys}Sg>gqy+DDVuYZ$%LO3bfJG5 zyK+{YsyPygq56Q#yGwZ0r#Z3?WGk=8&jfWS+Fjw+^C zhUz58mHiCVHtzJrklh0w6cJYjrjg~Z=~*n4g!Om^KN$qa9=usgKMubiKIm30y~W^p zz|$bYTH!{|T{0d0v2v(DE-APQ@RZ};Cws-G98Pv{+_G=~=<}C74qWqgn{*49kIu+e zbtSKkUQY7(jl#`ddD42`9eNvmj2x7dMQst>_1qYAQCbGPsC$EIBJX+*xZ%;7z{m#hykL65!u) z-lI3WGe>-kS#6~IFL!RX4B5R@1awxycm=8gpF~hbY=w3WgJWV`H?L>GKHT(diES5N zAz717kh`JBg&L!7qA?a@yv(_<;0%VRQ4=#zIr@Y3X=l;FMb`u-GZM1?`-jP5Zf3-V zqbc~VbnEVjX)-v zCRiK1k95;ZsWAY7jidf#nlNrSf8ecKiIKWRK5f^=&M{%+I|qNXmDIdso$BYzaPu+6 z9-+t~DlS8^67qCcNUc9=0jR2>arZ8-CCx($CM5 zLHjTECRIy1rxdH&g~)pgQxa0p;!!+HgQw^5w@-s=TTDIUHyA5i@S+^vR!Bvg>Rfik z4!WhepBL}+EfJO5RY7GR#4Hf^;*Xp$X&qv_>*Bs^Nj+(W&C#gobYRYcZEsmti7fo_ z1*ivURd3m?*9i0;s{<0Drp4itpQE5}9mMt>$xRmyT870tS7O$gLl=0zmOVCXHANeM#%)-^NRg`L7edZ#zQMVL8p{wVlRYU2);nkrfVHTo)m;)Kz1DkdXs^(5;o%5?dl#z=0>q-T@kZ9S0B% zzs%{|f~`|GSn7hfto5*>^kGciA?aI==gsgsMIx)o_1rV94yG3TX{hJVC94Ej&W1u- z!{vssu77!Sw=;!4Mps$lLAWQ*M7n;iYtw=a=Z6UbJxNqrN=0?EtyE?I+&#q#Xg!dWNCzf%c?HuG}+A41sdoLv?!g8^@Qy}N8SyzMh>61%^x6# z_t0X^V16G*f8sP2u4{`E^nT2Ln?N%D@lU6b&Ix3ZnMTWfitVAuXTUo>JDJso_0vhg z`l*Gmuxy6fj-&D{k|J0nD&{Suu9M<`3h{n|`|t5fBbSPddp6T2eUpOw=>zg zRN3@{XI*BX_-89z-X!~(z0Yk@qG(eaW?p$}_ zRF|-z_M0Z$RZV*I4ry@VC|{?U)jUJ7%@k>Z%6@ua=2_;jGHJr4=nC=T37Jg7SHAda z|E!HZ#k?gGn&e9-q)3{>&W9})bO79nearj{Ik!J_gEZ`lr$}Cxz z5bui;A(>3OFkXZbo-MIw=-r`*m0D%Z>%9=t(V{VY&wcw)wOZ8}%b&#y+L8IbS7c@kp20f;tq1N(26u^MS$ee*- zXn)WHQY&_@8)P;}T)|;>u@gUV0>}7im*iV5YZ;y`d2wB{2|f11qcTq_77BbjW~RoL zMQ!D!3r?zbK-c}y?3NX{dhSE#=sQ>+RolU0{T|LZVjV1lzs>%mC6DB@z;InZ4s@r6 z1LKKkD~N+YzzvA}Cu*9!JCsFqmmbPPb#-*SrLdjL&Lmb8 z_l}YB$YS0hQt$WJ*CZaURiYY0u@subP=Bfr*h4P>Gyg$FIm^vouzOr=gTipf+3v66 zbfI-9jF6NVm%n!IFDy-l%fiEoc`!y|z+s%e8bnkfXJ~A1A0r$L9~|2-G5m_L+2(|a z@f$yzk?^u(!q}S(`t*B)PDs-P0FkIJun|rWZh6rN+J*@yO6ywr1-W5D-*@dzaSQ$g7?B)cFac`weJZ=dL zE6;x=ze3y&Irxlv;}2OOf5cItk?ESQm2^>S9{e+aw?R=1Lw1obu%L0y&6p8lh%K|pa20&U>tLmmvd!f*Mn@EehP$)Yx1V{{?Ann~x#@fJhoY-~ClsX1 zg`=xUX6c596nl>%cd59Yv9)2X-kBhHksM)605*YQQKt-ZPuNXd%dZKzBg23gk{UyX zU<$A_HOWgPTm7KQZ3{n#2YJN9${Ya-9icvv=9G9nta#?%^+=Iuw)|>6$?#wJ@11j! zzIvYBB)coPEx+-m<`kVH$nd}M&28WAoMYS#UU3mU7@i`~^pRw?1d4VI<(tT7iE-O1 zMaEmjJQU(+RBYwJL9E2X&XMA2=L0@cX;3E#e*sjILGi2kIT1M#`*`{(Lpe>{M$WQV zLkxa=b@1J{Evpt>*4eRAhmbq9h&SL?8MqGqu8P_&Ma{bu*scWvmN9j6@ZGo7#Mosg zMsX1*S7-d(k0cWEkZ+&kcyk%DQJu@lbWo3ewejCA6*{<>4i|QYtq6^@3tN1fgpKlt zq>o&R$s^Ziw#XCZ@uHQ2R&@rykL*-{BADv^3PqxPgTH=iJ#TC#C5~`HD>5pW*m`kI zf^Y0-+Q?iOAXeJQ8iM=9d0zKs?vg=4LC7Q73G(NvIq`3`@@`HZ8z>AL8e35@?2<9r zoO-76Ctoowfy$@x?vlEZT7+CT20$*Et&|%n_Bch3QgNRvN+c^iaw59v6A@{Xcggkz z?*{ov@G`Jg1ZxI*8Ei!@q%~NWpCU=8vLmiQY#&SXZ-;lvFuIGX-I^@kc^Ot)YLzI; z3Wp*!w`HG6FHSC0X}ZFfK#XaBaC~U37?mzNWqq<9X*X2qXEHr8oiZa)pz+qX1M%y* zrxa}k&J%CM-VKg$vpzc=`SQmLrkNo1;OFwiWYYvvXEsWe6bl`JWmMe9%0;2&(yKm) zc^H>iGxbwu0OSpt)f<8e`KkOI!AEnjLE@}BM^J6p!d{^K*}0OX>>7Sb*eX#2eV9hk z>@ECG;Tc}Lu$Gq?aEdN)MwHJnUljryWJPfI;=bl^l(>rMXH+F8RVMhA@XQe z_vE#c>v{1&&XVe1@3(gPLBYPMb&8df9(}1%)F{HUuqE~|eahPbkhBYNFPyqYi??y( zV)e~R%TB~+3z%G(YQsvuAu<_b5u_pAh*4fJH1r+!{#nqRu-*NR`z7fK@PG6%RVvN# zNG*$%vdEKdqNHr*H36MJ(k(OumVgvyIlnuySy@P9zr6#%U^Z+HLx$O1NB_aPEqXVx zJNzF@#IBd$LN@`=b)vjgF&~9GH~m}v*FKx0RXQ3BFCI1<;qu}eIrxkSQB;m)1fNU1 z#f43bm3Tigr=OSYc|(If4AOR_g;ax387_;sFKb3S8aBdWjAAQK|25l_)@Z#=KFSa4 zH=ZG@M_Nz2t{+Z7b81*WdmhDt&gI4-&5w)Ue9Gp4>K^yzUVa<79EOxWI_50+>oA6f zX7x>QTw2slwi-5L;Tiwp>9{##%leEm?YEaAB1|Yz{8h7&Y~kj~x^O%O!Y{+{vyWnH zDCk>@!{jc;RCJO8siqW?yO_gs{=Omqe*)>!`Grd2g$xi_~a@Z#`wv(W4U47 zevA;>8K2{KbB>$Irj&jD>-R|fD`QiDEo_)gNu$_giY%w%5IM(%DX$rWut=d_HWu7q z&qBZk0F6zK$KqG>d-pH&Xna%`c2lhQsEa~-=*^Ma6*tHYaxFShknP{cHbk#wvnExJ zQTlJ#5ZD3>oD_y#GA5h%4qq;@OclFq@sX9}?=of?gL;;|OkzZeFXVO7c#ZT*552(9 z!AVa@_Hv zZ9aGDHfgiE?)ygZ!fbz#5))qY+#aNlNntNV>jGb-btc8GqDU$gw@}om z_yC{r@o!x+tP5M!>%=<$tzPB)UWjFn!RtBvIoSR{hhOFVZBDPM>%M8fIQ)~t|5)OK zI>{wJ)Ktjy$d_aZv7#C4uu(|?tC~T25X-SqP(hOk^VE46?5Pzdvn#!?%R3mYGSg$3 zuqtXv1Xk&5dgPDfT9WFI+6>r=yq?8W;vx4=Syg109aRKYU~Bt3U_N14|77KhIPU(R zHx@6pbX;GmER7ZS6r&(1k|z01U!z9?u?}Ra#c}O&c%tP+&)nw0f6Wn8@%iagsddbBED3usrt`PB|!eZ_4+6VRAoyv3JAQ$+-!n z&&>U}NwHlN>7e3pQJ__J%lnyZkNH?{r0HiW#5wNQNe7c0lRZnbI?PC{xE}0uAJJf6 zZ!bT0Dh3nRlbybJrsk+BIl@R%j)$)*wnBv%Bno(WfVGeGR@jpN0X|`mgtGhX5FK)0~y{Gnj}?6I9bE%eB*23lxv@ zk>~L}{M<*5d%>L0<)LYqlg#!Cmq(XKQ1>_`tVamv4fC698zt-3+p$NU4!HGKdW1-p z>t@FKKbkDJY|nV9rX?#)7*Jjpo3Fw~i)Q6g5p-}ZC0HO(M8|`cPqT6XPFr~Nh*MWM zPvHVvT#;;cu@gU7NAlzDQsrwVOBC=s{XDYHg_{X?nwiMW6k9-%O;lVLjh%fveKEV6 z$)NuA#vo+LzZIGtQ5kdhW08PC;E(JGzTslyc( z76#2}z_1RBkqlhkMD5L&UAk1O+DFa;17#+IX}UW;g(@xok>@?(S@D3|BhPjc|9Y1? zcWS$EaAvD|5QwdfXB5#x=inBeHEi~B^yrKZEROzdp5BiWSg2Hu{0BiMyAq=NFr5^- zuw6M|W>f>Q4qF;bLPQ+Tv5{7N{`;!DrK&CW)4$}rKJ7+NBF9PU^YjSe_w zHs|T6J79|&aJKK+ueFib81t|>69?V!RfZRuz7-1ch2ZB7x}kC`ve=)ORgsIJ(}iNy zxEk5+2a?odq3~zlSUBKezu*>}#=>>$HQe1lTW?9;W`!edGkt7!m12zq#2tOGP38(e zN75S(i=8In#jw+ptC&GI?4e1Tz7M+g9s5FB9}CX#uzo9t9(hIJ(e(2pzxtzT@>#kp zC6;u%@Z__?j3MJO#r9L=Ar*H*mg9a&bX|U9M$&{jx=~mjStacNb=Cp5HB(Qy>jG1F zJ> zxpY5YtDJ|on+I4AAA29VlT-Z@{NNCOquKP!U| zeW`t(O$Ndq2pB?Pr4~qQfoblc;vobcAyRx^wl%~cK!JR8kZ0HvxF-@Qd9YghqPSg{ zATpf36U-MxlqE%UzgI1UKMh5bwP9(3y^=>Ew|%;3JRhkz;MTQ#j92Ek_tW{a;Y^I6 zZiF*RMU_59bQZL}a>J}$=(Eo*jSMp@dnI2?&2jj>iH{!C@Fg+%2l%;{JMADR^D_RQ!jzXRQ;{wU2PKXzX)D}MmwuAwA~=zK6b9UX+U2Li=mg?X$Tnc#@t{rE^+bAH%+t7yLxp!DVqSa z1H zxrTD~R`p@JM_%lM)HugOdlk!L2i$b@fT9*0hyk}0b>`Gnfw)!V*Y0)X%X*K6q6a*z zhieyBI*c~Q_E8!dtrz?p@tJeq$!U~aw>oBzPr}iOCiqWx!Nd*x!bcw`UC`{$Qkx(;h^`=!YdnveJxq%HC+qVGjvZb)PNaR3&z2c!pswW-cH9{!Qr9TyRFx#i~dqYwZL{&A|p^6n%xa83) z_q;LamLYksep+eJIu@C^6GWiqkR%w=a{%5`Ebmwq)&_+ZgKh=TE_Q_}hJ_TaYpNnU z>AgYa#3AmHL(ujN2A`tF-YvzCmZwWi_)&cLm1dIq%HXHm3_m)G-9VA`R2(+7F9e;a zZn{ReP}D~f0z0Q%i-vZ0MKV-aHp#Pi$-vltExH93(K%78W1KF+YZ^NH5MuAR+kTzW zv%Kpci~68e+``s|VVwhXltNWjNKORSYW{B&-Ps5V+egzuXP*7qe?55F`-hP6x>UMr z1}2ef=G4-M_y_n#M)VG*0zQMF;JjmVa;3Gx=14p?MsBH%Lq2RSaL|~(;5MB65~_W{ zXQ7sg4K5oqu_7hWEZ-lzP_&h7;a>+0$p^euq(^WH>Rf7gwO;#!8-vFP5JTf)J1~Z> z7@MEmfI&&;AF`w{=8_L`;bm{JnZuDyvFj)pv$%B9EnemINQxmi%^|-|VUJwv4<*E^ zWcHDFd88)8ABiK48t!8SDciARvNK}?rE!0GurtLn+4pRDtqWtoN&>JzrGqSRD@f8} zq6TYipd~fcze?Qd+hVA@y%=^kN>dntNWe43OnZG8<0*!Ia%uJ?9j#9{ASwQ&;J43X}Ewj##l}zQ(cx6(A z{QCEQ)cnsozx(B%ej{B>v5P1YKg?Bqvgv#aGVB{3eEX;W_OfdYRz+np4;9T}Yi1cj z=E(1u;eXe2{(PqdM~6MUXEtq~GM>58xjQ&ni*XIPoxf3-tcC9N-~NGQx^P($lu!?Q z{|hJHsj=pQ7~;d7bYCb&wg5T_eWDqI8h}p+ebFKFit=s zZWvC=DHiH@i;N-jUd2W6o|r^=6AeLcU`2`Pg!FM5&`}S%wF^;-3%ergWL>IH=iGkn z#+zDYo3x0h!8#$3G?iiNSNddqWM9-dndej$mA=@pl$D+aWvb5>Tr=t>vvnL?S0kat}mKQqt zb7K z;a*)}r|_|$kF+r$_2JPdJR`ftw5Yq(CnfqYy$9YKogF*}zpsJLVetF+_-A?Ai7lHU z@^sGOxMd%ftlN~4R>F5@t$W9^qlb%~apCxymBu7v^wi+lB#0Ki4rZn{ER!h;ITN{8 z^@;Z&@Wf$64f!E4&u=Fa46~M9ni+ElVpxX%c5yuI3mEl;k4r_>*i$Pw;o)bW}1s$pxm^XuJn08 zumkBRvt?HA>`T#FWxVJ?)aRhzv_rZB+FIJ_e(#$gz>deYAV*Rgc45XgdW#~Lzfw>- zdEPwKvc?{Q8u26PM(-StY^G;gFI?mB_mkr|ST*T(JnH9OVtpJ&GV3negK%ffucrOX z1e!H}|N2+R=?Mg+sfR5^uTkt(id?4Ru*3?5Mzktpq1Em1a?$~d3e-hIj*&Yv?}2Jb zKVO4wrBzWWVNJ>ke+{-EckuQHBcoc1B!yQGt;hQ&?)NH|Z-tP@3F$H>n}G(Gm4Y5R z9S#>pbU@V>(wS&lRRc3I?Q&jrod!nnD+Kt(4Ri&lmVzlb5Y0y7pJpDO)fL_(?^R@j ziZ-@ErwCGbdsSNHZFymY2BWJz(o7&Nujg4O>T(p&tRnD52du!26VP0DnQ=4jyBWbI zpjCEnOCUwuLj0~<4NeZ4y)U~cb~^<^3~?97AM(rfS~8(4a%0e1cFU}#6Pnc*NPSGD z&tSmcc27-aYlThjyJB0^18yt4PNCtLxXH1uBk*J!w6qzfz1e%3;U-@iq z$GV#yBO98#qiKU zDOsrgc#5M2|JlcGv4<-3wa^ZrvO|~Uz_4j^4_PsK*(nT{9d6Hs3 zroe2*E#h@B==vA&Qh24J?injQlKvli-vZXumF@2lPe?Wmc@fAt1r-R;fL0!c5pB`7 zGo6{<&ZGC<`OnBHwNNhRxE>+7t{zQ&9t15gkkm@JZ>wysHZ=l(u=s&a2`#D`95$jiFQPviZ88I$>c%OC(;oz?7@{$x^vpal;WRCU5?j zFkt0F+@La4{mZN7Sq@x0!H#Q^5eqpH9kX|n5=FA_vdGdIr3-FNU%{+@;h?G|vSW6f z2y0CTLJ|U-*bDABGT^!{cpxM`V%co1Z*%b3IdLMpLJC%R%&4*w9b>&M^U8l{%&Gbv zR^c2pF-rDQEOhkP3E6PKpIj)qt>|MCcpn18K~nf|oZ7(oTot_A1*&;A$gYLt(T006 z{O-^xQ;*YH-(oRx*7Xl;*9vWJ^1E3G(`KM2aa^tnuwV)irqnxZ6IcTknmkpBv?3{Dl1TD`J4~c80fKwn{Ll*7ak%heC&&h zwC@j`nij0#BY#qZGn9HjlMm)>#2|xyW#oScuXUPVvTqV!>x-wh5-i3=9tUbh2o*g1 zUqsF1b*?4fGLKg}ta7rDKGh^gCgeitK9!D+5n+;cvocrRr9hZ;Lc3BV__rKZuyLs>^>HCCcB{#Jxx^JFrAqX2MQ6K(V(e(o01l!P;)Gc47QH zQY%H%Ejik1p`NS~s@LO4EiW_l zG}{NgC02)G8wo>Sizc}t2G$!bG*&!^NAs^gtYc{wE6*HCVtckZFP6vw<)v-n)4 zOO;R;fHl;H0Go+m;{-i3L1gRGktpN_J?}SerI;5YS*Rq<@)=TK_`8Iz6Qs*N3m*v2 z_8W4#IlX*(M!-^D+eCMBvHtfT`RmyKnA7j?xLeanhi-q>#VC{B*L?aX66e4&DbK{p z*+8*S!?Tu(TI-TSpJzM6iust}ag^+KX`~CoOCpLrJ_CN9t@K&x<%!MTv>vmzHdY>w zdv2Hwt@zt5b9KeVALF<|5Mn{O3p*mxI6u=oV>@@l>Bdl|?|SGhfK(!NzYKZWTH zD4o$NE1$NKr@82C~U=Vt1rmCK4k7j9H81Rms|O7P@tpONhbbxd$&rT-OjSglpx z^^5nfbm{ZkGh-`6rfvj)(+_M`^zh}WugqG^+cee0PPN_YSP6~CyuY^lE+c2M42{te z5M!Fee>7+P;iA19H?fRZ=rN8omD&8F@LqYkOsm%9gKVjy`^CwUkJyrcb{h9rnk2U> z-xQ$G)`ClCg2c{P;52gC?RBjBZMd`fmgpG)h=rZqQT07(j3`~+O)nSdAY#zryi=S; z%EB)~5q=fD%FRyl#L@y<4-88)*k^q4(A4wh#s?Q*II!Jo!7RyDqgw|#e+&&=p@SZj zBup?Ih8NQ+OxS1HA3N(`n`HaUi3J_jt($3Le5|F|H57@bqG|&3{qD+oU6Hs2I=W(0 zEIgbbHDL1dTD8m*yEit#kJEcDRs8A4PrL8L@WQ7&2YcNP=CZs2V?ge!*#hXKy3Ob4 zi-u<36DbL-goeTTGy#=;PdHhhlS`Ksy~h|gE9;X2Pm$Cq5cVIrC&;7NY>H%3QO!Jz zvUMmAGxbmsvg=>;5Zz4(Zl1Fpvd_|j^x`Wb%pq#0Z$gQnUS1*2Wm*DJGmv&?(MYfXTXNU^IavXY8Q_U(YCIl#vDFhYWnIXkcZ zsB35yBu0JGzJuIeQHQNid*{`kO3c-Y1<&_Zx<;nsVIQ~-=5zXpvWcfTLsA$6Hy;j_ zyMIb15+YVX1vtf!*eRp~#}G=uoWp;ncHl}{{;@}S;~dvnc<_ekJ@YI!3&~=;Uppr) zVfwskASZbs1drNXN+dmsdRdPmdPY5rw?vjQWwhZs9+_Tg4t;o~Ikeq*t6q`tq)l`(h2#OIEd98~8 z$(F*PxeNvJ0n5fpxHERNSPR#&f5OY4e%FYKxZmx+O^!G)Dz2HJqK#rt zQ{*IeRQWWxQ})0mf0m9$ii8ekWyDJGg%^q*xEz*YA0AAP#r8V6o{coBN*&!s@EQ<1 z;6eVZIO-G`fHvatR+&x(*uwO)1adj7miH0KRp7ZMkDqng(M1!+Mi_XAp1tP0jj9H7XQH9`` z2cL{NPWYPqA1}ZDMt~7tjjyb&AUV%WZEBqf8h27`2}O#L6cw`&4@T%@SYx_;%K3;4 z|7PAX>SqhQiiy@74jUqN^mqe?~_wP~QJz9?|3LU*weTZBVAM;8Nh#~uh z@ob)RiDw6x%=r8eV{@fzE08@IHr(#$TH@R?oIOj zh|VR2(TkIS-67dEUeKYi`O-grL97SHB6pXkWSPj z$D31_k6txi{_NFT^ELRGR*4C`NH7LzBr_g3VYOkUOONc@Y)pNw^E2$ZAq(w}+i}S9 zM@=}~2qeq3kHMrg9i8l(9{Cabz{&8X_NljpH9%>-22&4)?_A|}09twHg<%@@69P-t zqh`zs%Ho5MLmD@owP?wY3(R-k4jUM>;1WQN<->|IJR>ooj@GKt97%z4=0T`Utnxl4 zoKT6v1sRW9Nn2mzhKyGIJ5qCN#$l0R!9GFzw+#MEc`lBM5p{(m@Rp0Nq1-_Wbx|CJ zc_Xo2v0gpSv0mt|!!A^ardTfx)|VlQ?gZ5PYc@_0VC4=#TEuq5+Hk0ZXeZV&l2t8G)uNIC$7k13*$l_;Hl|UYdT3&+*>{=xbU~V~9OaT$L4w6q3oo%WIo{bmQVdIEL%S+!hXI6ex zk~r}0#{z$H(^t<)Ke)*J?KTwqq|C^YG{9ws@;l8$jhmrSCZ3# ziO&xyHoISc8QFb~iF;j7@)80Y;m+l*-8N_V;f^Q^MvaZZ^($w#+&l*Duy(D5cyk_q z3n;W~KEwr5LwBh(ov>lXZtU20T}MxJ>MMtpw((j%emcvtW3&+H55Mwl`UggNqP-@% zPfl=?CypCKC4DCHq?2OXDRPdA%8(#w#yO^47%l2?(JIlwpGPBKv@w^VMmD6td>nN4 z7l=0d40A){NvrJW)J&-AL!JdF0+dt(@i;UDs0SZTH;9uy=7EzT^Kx@UUg#HHYzjqN z*luxOcn4E0E{3{#KyRg}i>^~)nR_#yX3(d^Knn!yhDAKRypcYrs1kRJA>e>^onbgZ zf`dzWYkaryyCL*oQ9p7(r3Jw=zF|w){KlF7H>MhK*7;J%IkM4#y?PZU$SI`QtrW?n zqP9sh*(A1IhyX#Jjm*$XOo4xuIM3J65-ovu$+ac&qlgT@UXNS+12Vnb@C2UqLf4RL zpW7iROaep%EH0m=pU>*?ajw7BpTqS=PEv_FGx?Lqr=5@GgfJ~KEfU`M3`wpOSmC`j zC`i&>vT(%4W9PEvMNZh5e7x}u7xQ*`7P``*akz^d5;X;% zn0sHkYo_^mb7YnH4!HzIWh0Ql8ty6ZANfS1BoB-XEXm9Fx7SO{aakUf&$e8P5p!`` z7RTL?{^{P0_2zhSSj*QEkcF*mU-%M{PBr9oCa9IIbV(<>r01pBMG-lYq0^OT2L14! z3lIQkr1uN)9=w8oGXi?)oAf3nKK)N1M@ebciBQxi7BMyF>qB3A+G9ht-f4j)yp=Ky zM)un2Rv9v3;C2bON*MBcj0HJD0@&=3i-@z?hgvZaClNP!)0&@OGxxnbRjxS;9+os8 z-ONF{g3`?`=jG3=pbGrUd1>mCs#~&=S9Zc3UtVd6{KI1Ez5jVH`8#LkFWfu-y${}N zd82v$A6p_*r*;Zcee^;-{2R-Q7i7@IfV)iPfUPPTxM1l~T{+r_6+p4#7Ry-rhtK?i zr`?*37QvpBNdgaAzWRf6(s)s)ESsMeh{KI@K4SlH^{2hx(R}vGU-rNHM|-Ikvo+w3 z1G5S2bZQv~qiGvq^LGNC_7Y|6(|eOXC*7>760h;qlm@I|mbq6!I_3l+VZ=gO35pR< zOw1uCP)wfpqrB7JMrX-=_M;U*BCML)r6{o!>D6=J#E{zeGNsLav$s<_yK2qR6LI)H>=f$6mQG z|Ezxv9Y3!*WRN-Gbz-h@5Q8Fkp*bYeWh;%%b$e&Gljg`Z)S<7peg}I?U0{|AAQi+J z`OP=aIO`uhLrV&w@lf*IcrPS4N|${E^(L5krd1+oTJ`g*-BOtK68L}*`do^^fp7Rw zt#XgDIXH1{n!17DC+-hO5@23ix&M~u69xE{@Js&X?)+Spb23$2i8W4+}ss_)L6Ys8A{?Sm<# z+<_xNCrnnqkz(s9QcFdZ^FI?~anEkA)uEZ8n%$uDviP_BYM~UVEF24ynw1dIYw%7G zpiR6_byWHp>4WWWbzqw~fd`>Jt@_^2yXI?^CA?Pl7`Z8k5v_8AAfh4p8{?8U+zkJ-r(CeKG*)54(!+Kf=Y|Q_y+@~*XIpZjg>qQio>)j4 z)jMZPvFQgEg=`X@n2X2JqQxG_<-a2$)(fRny<$0tqyg6^XhL-%1X(3>)R@wXk;p!N zK9Cn9ISu59Ly{;yU*v`VUYcWb99ji3D+G+K@%EdZ+aG!ul~U`AC;vc_xfw7HOxgk2 zb0Z9x9EydCYb_Opl|UGg)@+%+S8-2@!GQvi=E88zY6{aot77hU**z(cPM?(?R$#`r zrc7~7&U8@vL7I&wkkmlOJ6jW6_iZOvr#?Jyi3D4A!8vL1{EN(~`L`97E=QF4Ku&9E?hJ50ZGe23!|IOl*uLX| z$CHK2*g?ZdRZ(}}@Ou51Mldye>xM7sa9}XSm^75RMX@(2a)XM>h32YF^4P$#Fsx`z z4?sfn5Bc{)Zii^v#O=ZqrbgHZdZpKM74X(-I(UurwHX^jZ-OaY6SxMrj0auyKzxS?^3!r+Arm>O>yzJ#BLIU9QCZn}cLpvI^0LZu6yU7G<(!f=M|vRAWU#y7bq3Jjlb z`0j&lz48R!qD700vd+G+donN|y~6-iPOzJ_+l7#TDTHjy0@3+UOwiOSF*dcv_e1{e5WPZ^ zOUDMD@mL8@MvFf2tm4-Sa_Q(9g;H$em_cVquFTRH8W^pMu%#JpW9W=sM43Iw4LT=U zzgcJ=q~c;+J8+V)1=d|;RyLT*0I3var*|LI!BkGaqj=Y@_K(>r4WC*Iq0yeT=UAN2 zPyfw4Nae7*a0}71=otl4WY}L08laUypAB?DQzE;^`LH}SGBh6G9aQNRH^CVkEm{$w zDGY6cTlB&z@o6^8HAWOYwZOP_x90}lS_nL3V^`S%vG;{AqG@QCtVhf4&X&OYUJ=d zjts}M27Vjk7 zo9THA?&uaLvO`YQa9&L3`6jXrWQWN$8!M=BfNTWb@NmpHAN4`!-tte(^SmF;=5=7B z!vZf*lYFN$?vv4z4?Ei#nxj+iNjsR@2#r2CiKh=PXW}L_eSy>ZkGfL!8MG{9Brh#z z6h^g^G3__MBPkB7cFIiDPAl?3VOgY=zD>U;ju65aDglqAWM#1 z{^evNRF*EEaezEW zH&5UJCqzv44#-~fv{$QcoOe)pO57;}_ozs-o!Xi3#7?#&7MA0>wIFPOAosW5l(oS^sOCpc9`2B4(o-t zU{zo;WG|`q(X_eb``6L~VxXZc6x~vMDg%~;f3}e^$SO{-n7sM#$IhDj+8q`a7QE+M zfzEih*Pz>ZC}q+rlfu#Fu(5B#IyLZ2XqHL;J$w2Ib2pj8Lc)UM%aEmE*n@O1NS0Y3 z`b@lv*&mb>wAv#YQaWDtqkrJ6ongz zf$l@MkILZl5hcvpmm9omy|&VEfjKXgNCqLn?V9cTc|2TZZV-A@&ODsQ)_*NS=+1ZZ z|M;|`Abw8rf&_h;UMJO;zV$uX93zl27-;aZ%YlLPi3yMnP%K!y)l^ipD0-S+cwX2gK1kp8 z!=pm)mdI0NvGZDJzp%=85EApZ(0zXECH4Fvr~1$V=@RmZWXLH-v_mxHq!%6)ErGPy zA*aQnT>8B5tmlwZs{buniRV)PA*UMpwr~x9w{*xUc4{7fo&O#<$_c+M+$c%)$p}~q z9ni)JBOAC>Ea2(!FnQutZa9j&A*?p<0{@I8C0S@@U+S6fzfToA^@M!LDcx%q(92;# zSxaQ0bT54?90_u3WSznUUXSM*-H2OXiw8S)Qs40t=28Lx;uBX^iJ|KMtEd1zC+IQ5lJUQ{f9}|9R z)IqZscg`ej+(N<*yjl9(L}PVRED+3eL0K7;nqHB&3s=G(8tL|XJoUn@ZUW_em zKcBZ_!Mh*)wo|skKiQ{7R-t5hr5 zZpdH2WTqZZEx%nj%<>N(pjBR))6LwK?*wb>ytEg>2B!n}Ps z?XVfJKl%RoQdcA1D!%r{ak6T(Ew}?GF~N>|M54h)irqlLZyi+y4Kg=|-j(%2)1hqX z4nD9}Ve=}C6dX||ND`e}#Tx8mhlvOiC7L*6z-+DTb!u4}y&isT-E1Q|e*VdvZKQ;o zG;(0v95ca9EyY$*WFOWg7KNjmw}V#~m;!>O&Mz+%x%oe*arE70$Vuo68FcMniumYO+f0y>L9v@C zl14>U&)F1;JvC!QgRYRcuoY;^xOm zY&iJhal0K@mNJv%e|mL>5h0hqeXW$_JMg|6YT-xdn+l5EO_4Gx>U;!N3YWPab%hj` z4yJ-GmsCCvv^N;C~Ufm%!rs)i$%$`yyc-y!Z?p~S*M~)R53SKg*2Njk$BcF zgt~S(D|2s9;gy!iGX#424HI$ZNtTBWDbhWT+KMUJY&~s&7_%8|bIJ)Z)YZEUH_a_x zF5Vdj)>Ia3;WOkCzerZ7(qL*-X+R`5VI6QJE1=26Za;et%S=H z7nVDgo;amVZg^|y_uigolqdd6Ruz&xPnosysfmm^NU;YfAb^U(4RE^AJ19X(H@2fHr@NmR?~&C_BmBuaNM_n8xOHI>tEK`}Y9#To{%?c}-%t?9&{&FWv@F2Awm zNxA*{4qI;1^q*Z7zcI>zWmm)ncd$vpvpQR=t|p6+Nad~~ zm+l2NIot(leOHIBj=(asd+tDvT&3CqJZ@FNsSB{VYn|W45b)w;hT_2hIYV?o*aj<6 z&?|`J-H_{ohM&0rJU(4?|Ev^dpBjC;n$;|9sK`3^GeIfOX;v^R81Vc`Go_n(H>Dqu zZDDW+kjgz*6>OLX=ZK@?C{%msM(hen;PokfVf9Nfx>B;+D7Tm5VBHU{{+AJZ|Fv(^ z>*Vw^6E_$z(QsEOwu>Si&_;zR7PmrM*IbEakl8;QOIr87s7dzi3|q|}R-EDCZU|#d zfC4O4`qcf{>{Tx-3j{EO3y)m}I&~ zybI!1tHh_F2@52lLo-llVI8(E02`MB;O~LQIos_fC{zqY;4TfixXtMpK5BC*4#TOB zEZtcP|)AHzg+n1v}_J`-N){mXX))zVHEo%O~f$iqV zct%8t1sruSu~QCtEcLIIEt`@<=Ml*5oO0BEssAQf>x^YnVo08yVhNVk*->%yw%hv~ z%{v%5tjx$Uu~AYfb{$0$sVL-T?*Jl(a>*Xo7Db`znqXaUhjJ0}x3|i4)0WC@?QF5M z$cAwaOKoH3rYC$aE^qwg!vXKie9dU;Zu8^aBsJVjUB~rSlCvf-(o^gaiX5h*klHp@ zkO(w%=s`fsw8d+wyjht`8{kz%ujExuN7v38-(JsVC7Qq~Oq~Sp+)nPOFwD7&)Ce^O zj$CLSFvu)(eb~=LQzyCS3T6MpoFM_*!iE${5iz1mb2Qt)cm=K@0bHpkH4!De;i!XF zZ7_1%A(q88y@Cq>$GNe3JmSo*e!rscFXj#%E|_#+NLlbpCHvll4wFdt09$9Qw8hMa zgVF;6y?DUh$>C-{zT^6(Di*!A#hmNZVO5X?_EL1Tba)?k-pFj3j-AfR74_hW-07Jv z%;pz`pYj}Z$IcKbu&qrEvd^aEk>%#F;6@=Z$`p1P;%2+P`IZ0aH}6FAR236g>Nhjo z8=_r^jkG_)_IrvVFPVJwyLS7#V%pa%3wY)0+X3rFU9MtpM z7e(VxA$Ebx!$~beXZOqh>1ECjw9iD*Z=+avdv#P4IvM`wBJ+>i;8PgzXo)MOV3yC|pYXO&j=Wr7==-m0gL1T%lV-;3RmvKVQyhOhyzcB8O zf7`m%i(GWzAmn0`AY>oK_E4l7f{?+@$}_x`ii@7jVX*mJA-_r%HY@eP8FXggMG>^Z z6IDazOM{?Nuuk+Dtx0qqbn6Pq7ZpKyNCK}c{8X4`K$b@z_wJ2YINXKRK(4KGSjd7Cceq+vT>fkxu04p(|0bk$zi8;~fwc0tARWZVo6G=R%yP zPL=&~;>(?31Fn}z8(H{lgdZm>*5-V;$cVZ!@1qLR_00HWVofl5hhlG0AY!6GA`XD+9K!TAb;P~#8J!Hf}UaMBwr}Pe&Ch+B8 zlstM~!LZM!ia+SqtjwdUA#WIY!Mj1&^(k5cG1XJAC==HRmWVLkiCp1&Ql>sDJutOT zkqBcL-P{PE)hl|OGbC6WqgCGbiuEcA!+G!%GX)odb@Wk>0a65UP4E@Tn&fe!l}u^S zsvyi2L#mXdsg+?X&5HD_l6McyJNy1Mk)zFD{lWj^8Ikw9L;VNI5(k!d(9d&3ver6^ zO{B{iI2Kzx>O1 z(jgq!p=Ti-0z(U}vhrywdD}g<_!^GTc7`LFUSN$@X%2Xo%PJyjd5{~^O@9)0 zgtRM9c^wGcrC2K8K4&1LRW|6lGO`YHN?nUWq031<>4WB4_x(Q$d*F0LdC@bSmkEyO z4qls!CSQ73fcHOeLf`WSuk7#}@=}RGW?_OMCRw8tG)PI5ll!5`A4;%mV$Fwq6q}%* zp`jxW^Ko=gXO5W=MU;hPY%8H@SK%`n)E7+o$vX2SV~2GUT1Y4^mFy4Lr`kp`=wdoU zrBxmy>1rU{BdPqN7qBbiAyK(1pI7X$*!gr|rr`8T838ql*dQ$F*U6HG6;0yoEZE3( z!2;uzK_;EOGIN(bwrjS5!DtIPzA;V`idr!BPqF4@bX<0B4lJQ86x~(OIW#o1fm~$R zzhQb{yAU~PDqVJwo!(lIEVpKN12MSX9Hyzz6Ljx%TI>9DOD&`hQ z>ZYTh!2j&v4scX_t-trleDm~DE{Jj9y}pHHRK0Vybhy{~t?>Tvi=N0UtX1~R))b1i zu<`EY{53&M?#Kdg4HVK@|0Gaqtq{k6Ht3KQNbm7!hr)+^USg1DN5mmsj$c)9gLi@; z6Z#GV$a2-_)y72AN)|gILkw&`^d-h@Kh#**dsb7$uVf1(#U8zW%SF2t(DPUql)@Zx zhU``B!V)d&7GoRYeBJ;^%~Rs4U<}IU(T2{-mF(J3O&^oM!>H<)Lm0nZHuXesy6}*5 zwI{?=v6)z+Ak(=Ds*(qY!|n!k39kg2cT;~xuxFw7dXv0Wc2|)P#!i7G-sMIJhP|=o zM$;}#3)Bl+Wk-qDGgA%GkB^ndfHky(fwV9M$g5^vf}}~VqaTD^o1s@2I#%BZ(d31e zdUitHXPR=cXc#{g(raMy9HZAuAdNW)QDB?^FLVr-CNB%^V=CqrgQy#mn6r%_d%QU{ z7L~c|y}BB6ehLee5BTPM?y6N|x%LTp$BV!QJtHpkr0-5hJwW3lA-FlXkuHqb!$ZRH zGFe^Vn0_hF{}*HUX*qw!SRUYJgE(v@MfYZHq5s28{r37bB-?>E^|dCO`Z9{$L6L1# z6z-Ke!`g*CQ;*4O=o8Lo)rHb!f(<}6+9fQ7_Mry?^B^O5hx-YSE3+B_ya~beKucWa zRiVTumI@l_%|6}YWrAbsQl=>wE$?p7XS$$lH`_ye3~(MzEQ|%Qk9~$4oO7hKx6Ftd z@lEk>N#YbzY_d3+6boU6bYy?k!uDo>tY){l-1k4>RZq6k_238oh^+B#cRL2s1}Pt^ zRD+^=f=3ztgA>!pU@R-~*q_iw7rCv3!wQ!93l42~+S}6>iZ;#z`?U)Av9KNr!=f8S z5D=>wZtb>}ULKl9bO9Y~2ZMx3nwr4np*6x%&pq@Ye~gBN^%xopd<~Ms`p+24ot)r1 z>4pEg??1&TNFqD`pdpJLSp8&}NVp`5O`ymsV8r+BfU>51o+gXmEiVg?Bj=>Znq_<4 z@xMlT#d z7(l1=Ty+bp$s%h7DYHJES_ZKM&3466*u1R{-5UT-4|LEpL+sHGRG_v4flVbM5@nK=T#{wt|Q`vNQ)#W6Qca2iz^uNc>Cbym$eYDmj<6`M#7Se=< zAWAN1cT1HWSH(`*N-y>ACs$@In*th!ZVOJAMT?HlMH=gU5%qjvmgWzHXZyFjyvpSu z|3F~1r&d|y-!E>7ER9Ht=!YDN4PKdSc4)lUmWa5(RG$QXU&wj!3eaW;0u3X`OpPN4 z`PH7+pgbJ2g5ud7kkymI;2b$L&XPR0UV3HbT}4W0NyIuyp%f#Z)q%h%5R$~?(5u-* z0p8mZd0WxojQ6(%=eTVPY9JS(;v3In*jutBCee8-jo+u3Kgh43m(9L13*7(*0u8ei zs0-8>esx&K@Iv`-6u&jXaL-TNduxT!BOb%Nbq zS|!A#4+Fg@hNRNvnjU9llgOZ3*mOA(v0->%Gk~fG0_d8RbM^vHMW0`rOByjS&f?bW znpeeKok-LFV(0V|&V|x_DkLbag8P9cV7NRhM?Kt-8+rRvm=e!6;OjU1*NlOAO#cAh z$>}NlJcxSTnqD9&5cN2pA=?CTWSMU_T>z;e_&rKIk;niOMRatz99+-2>dl_`alP_% zSwHjv7>+O;^V+M1gLdFqACR@X;fcfIiIbk3^z$QsbvCcSuu#;nmEHy39zzHTIp@JZ z>0mymPXUWWSvVHT9Ty)4Z~gvhDWYD#n=>@m+>umT12m1pt7czaGZrOdG+J${RLm#i zwi*uW^^sql@q6>~K8H=vuu!n~{xaWW-{FScC>(Yxwy^j`EH!+mBGLVT+K|zeF1I%(IAh_Vgs^r|$9SJ*UG&4Ecg{U$lp>dZ)%JI? z!h!cpIuk^sQ7jlv>yUu9U3h6BZcW?7Sw1b1AB!=UswjMruT|$yYm+W^$L0$5Ui*xK91i^@rbk+Ix2o`6qc`MrQ_%OlUg#B7FzSS(=r&i|wXAg`ITO z#KxhxK*P>xFqxhcG$y-l+5b(-C^qi9-9HLd;10XvBG9pCL>wW3VpmZlj*7w@IIJ+L zd2ZdTtXUhq-hKDoiIfo!e|@U}@$lve{L2Xtlf*A(oPN~^h}d=C{XVIlLfTAz_hS_M zF+~nhQ43?IocFw~xE;QWzii59qLcJ>)zIvmm%6b)i-1pm_B~#4B)Ed_xW8Rg}jTN@q!8}mERZMp|}l=NAJ$OBC4Q{&a8APpV1vq zL7nhA37NQMp8I$$kxS?GO0hd~OXOBMTJ&*H2h*X{sxeNe^*l-6rt8PFajdx7PwteKC^tmgRgC&(*g&=TOx5+v%5afCjgE9J>}MW0 zeaO$^WA895yoJ<#pmGe1o|j9fx^&8rX}_3UgFR1{W&T#9^{?xIzGs6IQVU0LZ8C;BW56JQ>W zKop0+4AkP4uzs?C+D!=l9Z@y`yLmc$3{q4L$Jo-jB@zjySF_83CM7wjT%x1R!q`~AJn=

    xCJ0XQkJLHNpFo_oR*VvIXzPP``!zI#2T} ze{G4Z^~1s2vy%Az;f?fP-~D?P{oYw=m!cbdvJY?nh@A9I_JOO-kwu}E>@ag1;0bFh zPehIvK(>?=CNO%^l0V@(CzV89I(X$ZbM7REjS529!U$t`4aLS&WCaz~4120pFm`u( zpJL)7_d)=ByleY}pV-1>D-dDfN&60SLd2w~pEhXo)1pj)0gi+^TjDdb^xGL58&r&;AzO|U<526Qitb6JGMMOyWQL4!ly z!7xOuUfc2B_*Pla%;$}Q!~dPig=B*RnKisHi&C0Iyh#Ez@(MKz_4I zGb{iyfj|b6{56@OZQ|9TO>*R&8Xl;B@N)#8E?QKY%vlkq=3I-D?fhXf# zWed-9(lL`y?)+8Sc*Ka~IuC)scBGB+MW}EG*eCB_go@XM;0@6E_Oi5$Zh}_pJ)S!i z1Jdn+Cgr6$rvvN8exq#I3RhXIok$+_3P7Ofes@85p=p75|@x)J!ufpl=nv{cE!6nXz19qda2g zQ$(@Qb3KoWO6Qe&Hu+&`2xdTPj?dK#6Tm-KC_2l#7}5fcT)lIXpGNQ8L&HBU$`Z)e zJ~8(s4JJzwuPV4od=bDFBf^r%cTwlXx!5BRX&Z3a3{ziB^tU6sj|3Ru()h~S3X=28 zz@^RvE;}g}QksgXs1DB(dLZONh@piFq;+eRMm`s0Yw2KGp@c-M#0D4|Yy*h+XcG1~ z-*m0^sdVX8uaN6yhBJdHtSJje&K79FGE53%OMwK7WE4WXF$1P2EE;R(DfptNfaP!S#0=C2K%EPf|WnRXqx1MZuudp&!Hqriqt#X z%lb80zOewrWCA;#{P{r{{~e=Td1c#yspKj*x#GY8i8HAk8l>1e6uE^oz$r{pc)mX} zf#IeY2%`236V_$WpAziFbR0OK6L>%#hRs3oI%*p%(fOc?zQ}X~<112t9T#hAd0jx> zTO(|w3qUO)k2NOB7==_OR0$35$8a7)c%Vt1Ly(oR*aP_S!eU=o>W_reI8}kmaN!Zw z=g!|mcf2I?sUHd0XGN8b;*9a0va8M;4k z8PKii;OhX}#0@#bLSe5Dr3PaGpShW2kMmi=DOKkFTrSedfgLm!yuZ1kcz#tdmRy5b zgW;mQ&_xp>f6T;(RfsTqW+F~IPKA=Eof{hUir%@7)C4|oLM}lpTDzmL)G^q;zp7R1 zY&SoP<^MNsd!>@M%Kq+TR8mEEfBiQS{mdvS$bcHLS4yPV)f8DtMHy7YHhyt8#;So;uHARX^B7tCoZk{zAUQ*@GdvQ`PQ*|7#ZSTIH~r$F2{_;+xEZD+JX&4 zIjci8)udUO9&nT1>D??x3fZ0Balj5>f4p!3$%wbl?l9p5635*^`dtq@V_t;LMaMWU z{2_Wc66?tnn?#WWs2YU84^{vTc$7i8ah4DMO%$XD^wayr_J^|>F0Dnk*<*G;RWUz| ze%kja*b{VxJOPqcGw7mlZ0oUTK=z4mmtyhEf3C9;L7b%6WZ9<3>)(3lNc*twKgeMR zPAIr+^7@~q*pn1#rlJPIPtr@ohJTyo7vwv|P4ph=SC3s$SZHQfv_l^f03z&l8uFHUppuL$8aCi6Ok>EDNF`^5(vKa!g7A6u$FhvYsX7? zrd5J^6m|w2;MFKDxgy121L+GHbS>id0u?IW*$h(W4oRIl;F{(7nG#5sgn!&dI!G>U zc;o=D8v+8@Rm)14S}R0GsK;?0S|@)vfs2~*FYkAbC!yMbHIapc>OQ7Y(j3-9A5|S! zwF~bku1?)0-8r+;r%U(|G##>X@vU15EAFxyMOMsZ%Uf1Yc_T$&6goj^ZyhAdri@mi z?6^gzIVOTAm15UXB$0}``P_AB47mx-XL7*h-W=KS$^{TMz=MLqrJ58^Y%wyI=$)0-SK0Y)vw0?p|@D(aGSb!4aPn)@2xT0cxS zDUd|ZexN)Hd~O-E&ifA7GrfW?MRj1IcM))2WXy;K7F_6&7}UeGx73{S?-|)fTVO6k9&YTIjsz>%1vwK)AotI1iNsLxK=`sqtk54IICfNe^hH3A zBm+3Q`=n)FHTKtRV};2R7J1y4AJEe_D~;kN@}r;ZCs_`>fvPqUASDzFB*}$9z&;$n zH?;1&&uEYW6ncsAVgqyOR{0T<2OP6NLQ1OW9iknOytGeM`H}`1c{(7*kGa{+@&f-V zag6g~56#(#MCbj1`vVIk==LA4Y5g<}=rOP5?uDV%wyha{ z;`(qv#uMkVL+x3z-Op6eL|xo!=>O3bSnXM&ERB$6z`UBp;z-YLcJ#-0XRV zB+so1j(3j@+@ad#o<~>lw|nH$1(N&zse&_s&|(oADK^PBDF^wRAZ2&tY&!ao|0(e? zwH`XEKX5X1P(&hzwVZombAX_^JmPlF5ubInFTMCPBdG4(D3+14+(6~P7V$k3P~D)| zYZSRmMa4}yszPcOB(>J@E=W`T&ji*gd+F=^F7-!rwcrrxCoyh!ppWMe1PnTuI*6av z1)lT2GIO8s9y8>0P;!s?)H_C$01X2ANw(WIx4J;B5-neyD&_?(KY>@zUl**CeMCN^ zuZ3XGz$=Pm=cD3wIwAP(H!shx4uTGhbo9J(58YhgC2~6p332;pH`49oOrVypS$uoeC`jdI_Sm7)T=c|lom%L6R5x8?a}#IupOky9wxxM& zJ$PYSrnA_rD*2Tf$rUWzTJ$WAJ%Q_%_Rr9%q*tHF$y%upoO zUNY|>xOA>USO6mNO;C)~Dbvvn&X7A>BTQk?B8OJ6O6(n6>5>3o>jV<+7*Vw7z92!M z*{|*;y)>o+H~+GoMoNCn{Z00@5g2VRS8M~iH-kP6d)h;o*t8*q=>m(nTWsi6q`^`& z^y=AY^{oG`tzBp9XSB_ooJ7jxbGdGt&B+uTR@+#hN5B?U&{w0K?h8Rf_Q;IS=^;f@ zMCsf-dO)_CZ3a50K{p$@bgf(?Yr$c(4Xm7X{I_ue$E1K)weOgRnk+vD^(5&G=;QD7z6B9Xuw+|h zOI@!BvA$KE16jtJ8agSWlP(S_3r}IzLZdW5B|2ux!mEO@;T7DEKx$DALebh`wXyY- z>AVF%H3&&J?o~8$d=vnUbDOvwjNyx(I-m-vp;wYldY1%0eN2e7IrvPF`0&IM#4L}0 z!|TlhPy23JuP76@%JSJ%Z%tQ70#wR((_fYgK5&W=Ef-y5-k(4Nm)l}nUL0Fr;UsA$ zeWHKsO>-u8E-^?4-Yr|;ZQuQGnmzOMp}48VxI)WBH=HuAw-xZQH8QN4Z|bRa4Xjc1 zxPIP$nylxhdK}o53uLw<;!HY<1v5Ph={J&?{o*QdH-mwJM8Q^n!&xuz{A~2?BNtpg z2j^;)_+Ui8a~1DKNSQizO7(LC;jw~!sur)d;AK%i~QDoYfc2*Cy(jqs`fTEHg8zO9-18$Nm`=<* zJ&BxpW(*kMX&GU_T%p)Y6uC%6RkGDSr`g3GCrLHLkQ?Tgyte!8)eA1XmiFyqUMFQ| zNsnhS|J|Bz>md|e1Z63PmYIe$1gttK=bwPi5r;&*bW7yfhykLbx4odF(;?uK&a3%$ z3p{n){kSSoQ1k6`aNH8vF06q&^-z;i&p$#sW!W;Q~{x>ZU8aKl6Hw6=i>9n!wcT#4>901T%a6 zj9Back$f51^2}hZ#sq7n6bse&MbLdI&8I(NArkj=)5v{zAg~h}j%V_UhAZZ(=sOW- z;AofsWyn~_Rbx|4An6IkzO1^iLGMyF-+!O~J&=uh#Z5t_%f`?P!B^Ew;faL*JDGg> z8DG_!SI;`EU)VxDY@Ht#3slc}*y?@c-^aYimfxiDT(MR#Kl_yyJ)889q zN!*Hrxulz$EOB5@%qkPYvHKJ|NRc~0WC+!Yl@VR)+z8lNrTKNKQ<$^7K5*`})0%eS zhy1WNsvRogG?>YUgk@DUe%>*6trDL{n#4iZ6efDcGBBO?3(LfZczXH9P)&|{ zpX#VI-aS|SIc*ryx#sz`3p0S3Orw=`$_^`VrZ$(n(9W>cY#O_rucPmK<@!Ew!ubu4 zrv+VC!8ZkVjL+$Iw>>j7eaxU+SI9Px5BUzuId;IQsNbAR$N~vZuQVkT-~Kok@XllG zxEU5;!e4uMC9Qh%$MWiq^BHb$zQaZ^i{Cr(r~g~dVJwS7FVDMJs-UB@`SzgFeKNSL z!S$18vHcw%KPVr-8V2YqhO)o?Zqg7(`QG>g9Oc76z;9#2H1IUoW z?nEZ6y14BIDk)yB8o~#le6X`f7o>GSDu))PM#-4@obAM+HDEB?u{`e7W~ksM9rh2chIq}8f1Wd@yv8%3S8Ap>D6FM)>Le=8RR2aZiU~1$sag%xo9o}8RF%z9_Rf^>{b52>Bx+gbGqmn z#nli^K5u|@(VbulR|R(}qebn?t4d83n5?Lo25)0V%;RLyldp{qlM$nw zq|l_=r5lvyElV7B?`@$8NhkdwAN35785fGSu+`#Xu`cMIE4B(a!>i>LNLrz+pqsXR z2x`QyZhFWO8>SE2@r)BTCb#ILJFQIQ>;x8S*ji!*$Blp{kmKF*N=RG+p}r4#kG0BH zx!F$rTk-pF`Teb!&Stl8`~4ku8}yC8H+(~8R7&ET;@^@)2Ubc@Z#_aOWl}5BzCQvb|2@7hfO)l9KX6gdhN{fdpDkR6dM$yZj; zUFvAjI(O(Yo)}ay7p<2ZHJU0XRqGrtwoC0)toBF4_<_^5u;tUX zy-*~>bSzyEnj~>9aTIp8>7sk(>%dzF=2NnyVlEmmI1mk&A6i(@Fcd5-xFO56k0}cu z?hJrt)ddDkDGtybGYCx(Y4yqIh#O^e`w4c7IUB2#2!8XkpGW=XC%<0!Cn?Q35z2u* zzZR%tkIx+xUtsQn3l}}98l7yP3L6-~_FJp03_C1%;MB@s1zHTrqI1F6K^vzCaz3pw zR7d$4xeSdJ3S2&8EcaTLM@(G7pUn#%EmR8b^Zrzv0|D65pki?v3kKv3`ATx*`S@uX zbelIT(WBkL92USvLq}&yI=maerqKtt1wU|VaE7dXRbD7=T{#da9?`T8dOrQ8nr^h)ZKi4EBP-U3X^a!~UMbTPMUbR!82DV*@Qs zm1+x{DCh4lXp)$DrLSV@B9BCxq$U{1f3FmDC4w%O&s%Vh$# zg<}H`7Z{A~ULWPY%VMLXS+Vi_FUe{L-bEFfNR`bLn@*ASR8&eRX6b2_-2pmPmZVA8 zN%zfbjx7Jz-ud^xcK(MaB9FeJ zVZ{t1J}!UzS}DmNO%pk=pnyEk5j*b+irr0-GAgPXlI9ju*fe->UQ)!SX**nQ@&@_0 zftzK$q@I6cTBo8cJk>Q_Rsu@M9+(IGZi8#u-l!b<1dfIitq%1D5eL`ML9NniuGEQ9q7r zpAie?{05STb`m`;DLff`j`1Et47Rk>Dw_k-0+F{~gRRYNWQ1E@lMz=q!e+#i_C5UX z?{;i9-$Zi($J4jH#{F5e2m?i!Xjuj&u1)f0Xij2Of96|C>qKrP74M(3pxbXge8nsVA)`fo3!p(u_zQ`>)Z z{%FXlm%93^>%Z#w)xh_Dca7TpgAO?C`PJE9^-`U`x(LU+e{coPhn%)hNpGz9+K^NE z52|0yfWK?0Kd+%W-!imBSpWLoA9l{a@Y?2oy|&=u_y3&vqrz8zcjxOCP&-DbJ%pSM z{*I2%hB?UFwDb8Mga}Rwbd5ITvp{wfv>|}F5IW)vUE(gNClVOe zB@Y|5z&ID1)8m|wF=#!Lu7Q#V?BeMN!A=1zT71-(XZUqcL2c%u9x|u`3 z=2ADaf#7({i#Gz=X{7u{PVi2CmFm<)FgI`#A77XZoYs3xH**3BbuQ|?f0#Fgby$#C zX!DBB^{s9xj4r4v1nVj){5u$o4krrI)W~RuHDulLzVHW5$L60BSNlNJP;E25%EzwJ zm@x9#8Etcs6GkSV%XsnrEaML750#O%q=?%N$brq$!zMeR{S;e4k=<03A;;f90;9?D zSTOJ5`$8~TIIscUa<8_p|D6#&zi;|^ z5V`Ka?){Y}1!?yv_AW*Gfe9u6+ZlJ!yU0#&$SUbzDyR4HPtZ%{Es6(D$7y7EH(Z1& zu@t64dX1@`vnljAeIr1ttmS1&K8H+N0|$V?#zDHVOw5vLlD7jk@8bzzU=NL9t%_ED{KVVS=%y zXwfFs$I2GjzWJr3JK*Tld(!>GM_OeITP)U8x}r%v;JOj4?P^jMen{9G0SznZaYZ%9 zgQQ2~)u;nw!`rnjZ}+H;NID$#KkG=T18XR~2{!5|7J9zzM-7F=v#_*azk;9dBq`ya zn0pdBBmrA*P>lkQhq=yULw@&oF(SiL z1(5OmG4Ddib=A=9T=hoLQV3h&X5Y}30@KCED}C$+i?N~Mi>J70uD6%WC@`WqLD+ty!JIK8Uz+k95*J|=WeS6v=>kc$Pg}5oj}ITi9Tz@=ci9nNV?~xh z1KA2kkIwU-D}4TL-k{b(XV+X>FD_xSBn_}-eRyQ(@SDO^2Ji9N<6SFg^NJhp@Ow5a zE>Ndxf?5x3gi&MvaNt;f-$#ED+m3L1a~-xm@+X@Eekd}Els}X`_a2FNVC$v8L`ZC+ z*fiX+MIDalrt4q10F~27pjo2GQyZu>xAQ-So_npbRv8+ena_=bUXwd^xZ;-kCyeb0 zPPm=?Z;QO+&FvB{`F9Syqqbnh{#zIyv4(p%> zR~O#nxlz)n$b_x+F8||nBfU|%(Ee=nTru^0>>$rq^?oe3AX@r$_mLty|(z9<=iiC$TX}9~8aSoNJ1U*l=L)iiKuNTY*A%HJcfV z+!BjEBHO}_s_#jU07J_vw_&oNOsK6aRvdR<`I=@0vw}h1`3hw&w6f_?;&A?(#owy> z>YX<=7%8uibWYCI?1Jo&&z06hA}r{6O7;XYb6 z?ZA4)f~Pi_q$z92A8yG~`R8VKC^DsgG>o93eUg_Js26^DDcQ9{QQ?|K23^lecadau zw`u~+?ePJ^Myo%{G?pb#cDeS?9r;JlVE%rP{AC!T4NeVc@3-jy!VqBD@}vYKt@A5>`e(zn9* zgf|DlF0Xzfv6gcqpjOP~~!-9|J2qM8 zTarn--IB>{^ki}$I64%-M)Oi8-4BgIdga76lC|!7@hw^X^g+6U+A3@F+9loQej4hhaYmhFAbh{Dm~Mp- zbOv4K*-NLsfl(ryxkK1M--}t1ND)Y${^A0*zXVw1O-y zjUph@xbQ!xs!*z|(z;N^l+M5X+1|SMy>G!i&pq#X&+;6h7PdnhIK_go-;_5SzvJlK zCw`;fUQgopYvjXG<>5M&)@$_MY_bekLRJP=atEo=u?>=XvJr|dH*g!I`vV8uiy%dB zAiUCNtNUJPnbJ&YsjE=@atUZc{&Y>0FMKny?g-E1ZE5D2Rd0w1WF-#^H&YKJNs*)f z8iJGOf%YTcZ`d6;A&gOty`ts?I@px;LHCnJc{#ZW%xImeUf4&b`k#|L;W}_-8;61$ zW9!(guUHRc$}~m(MwM0n?xqSli(TZ=iS6a1CP9V01OtxYVj_C*?M|O;P-Mg84&^o~ zg}N{6ljJCi$sFG#rj|OuX@)LHxzOJD5~!#Bkplpt9zhZVjdS$(wDUH|9zZrtJv9X8 zaxM3q_{zf5rHdFh5@g-g7#4Ftm6Chk5 z7@(&*4?uvj$bs-q=}_=x!6lCY8n@iAqHynq?H6Bx6L^xd$P+*GP^(U_~*hO6DGx&SSAh%Sk!cO8+ngTeJwTFV{uH+%!; zSfx_=A2qnI2&`Fjaq;uH%HdP-j8*iCzzR6BT_f2GDbF$U(xX@4nq-?zRu?Bng3k+Pv(6F z*=QIib~0M&N|+7tjYj12g!)Va2Q)gE`+;+yXVS>{o{7g+SUkI7k`B(?Td=^~_w2M0 zFDu?;>^F*Z67ft-AinlbT`>6h)TJpB|Bt)zr!E6x{C_RC(Gv_H{|27@Uig;?%2uI| zef?iDo!u_6GmUTmI8ySr#&t)w-v0yoJiB$riRtuuOxB%i1ap<3F9VxB=m_BNr7}g^ zIj4L(z)gpxb4`L%(g(riO7(S7A(={5Q}?+~U9hY7Qa_&w`GiQkwu*~Qmh<7WHL%$UOo4 zM58~9MvjB_(baRXIN0FR%#0|99zr%CdOUa>i_P&JzZ$mZ=d6A(kt}p4_61qVLT~)5 zYoaUU39=J*IDOm{Ixb-E{9Fan$<|1-+|LV#ITx5(@mc!1=)n9Y|H?V3aAgQwh9}7! zS8SHt$*Z1IC9px{uYUMeMob7RR_3(F=~yv`nh0N-6Kf+Yd}+f`TcD=56Ur1yy`VjU zUSA+t1@;%j?YbjT@S?<6;Nr!b9e76gpug+q##L(7&e%EhHFm3%6Wg~>OjfCT1oI_9 z-@;WYo~a7zChG#X%&;0Hu|tK34DKdCCFM2q!f`|^YmRTk@Eho1iLCXO0eJs zxh<;5b9C)OOvBLU2yUL<>h_S{=XV-dV`IUgb8%+L*B}k3JbYt#18|@uP@mCBvK|sE z@-S~Ofy!|$mEsLcmqHK@k_@rOT|V^f*y#ngcpcT|BtP#>&otwHz0O(4wzapPN{Sll%hz;Dg_?rih@xYL@>R(9BB z{z-JVG|5$6%Wd*$C+~QIP!2=aCu~}#ZV+q^!fFTOA?%)nHB_-6N4Yj4UwF+gn>-qs z#@CN3AWKPgi?~&!*sPpw-9hOe!b9THJ!@uQilaVs^5nDWV2p%kc z6T6eFtm>&#l^!3tk9u>y^+6N8ej1%;@}j2^%yt5PS*WFjY`%F|;2OhMV=C_!UnsE# zZkl}Ies6f!hB<@bm?M)*-U~H(bJ+d-tVXjI0;}1_KbVyeoLA@1JpF#uM64Vqwur4* zIav!2EAVf!tey8bs%PfycM^U3IY?ptkX{)wS+E`jp3U$cy#|H^hrq zW}^JE2quG|HAM7hKuE6MO z(epDO0d_@fMGeK$8YSk9W0u~I8S(S~cRpTLi;VMbIRD=VAAj-fIhqq#AKwPL`7}uz;UHea?Yi>_ON32HsIO`QTx$q7^D7K;cw@?@c2|(sCr6(J0q@ z)G2Z6jjzDPP_TN}dvwW;adc|*ZB#9o11t}&>1)@l&d1H_i0R#U&8HbXn`M$aTJsvGP-h<+J7-a}vW z#XXyPi?SsMUHV9UgiQuVTu{~Yx)z1@gXY59JLDN2RY6q(^?vWC(v1=6{*OIzbwz`1mP~|go+6_En;kQJ2B|l2qSAs=4Iz61T#UbDt(Soxvc!%%nfw96R%X2CcxN3 zFogu{?4zp#S_RODsTu6nCUS^Y<8l;BEL%=2)Bl%C9`bJycZZ&E#aK|I-dC3_L)Td` zBxf1o)*HhcC8q_65YxW0uvd0Qj#p|G{Sy?rnPlvV0kf$m^pV|il>Ed!_bnr0T!hL4 zboMm*n2BKa5e#fO_7KrX>Rb-alZ-j70`+maI^a0nH6H|Y#p9}*8mY!QsH0FznXZ7% z29oM#%xUL6mF|HoUHFitZ1fy<*9xzRv|NomW$s9DBBumu-SBch2iSe_rA(eavDb3! zbF-&-;C&rC^@NVuiN|T#YgB)Gwv065X7F!)Pt&oM`VKm;X+jZ+muu%C; z@H$!#n5R&&v2mT(K>$@NV{Ke5!R#dHR3f@Xd|=i(*c5k?T5d&TvaE(|kPgv!bSJO@ zGAq&YMW#EACTdWDNC6i z<>E>oH6Bz`aF9yXkTuF4vWv6AKVR4&!6bi7k}LDy4%Fb#LdT=g0}`Ed@_xtV{*MFl z&-9QZ@!xu;C|qV#Cfha!T&A~L+S59*8M@a*^W_i>9m6>@ibG=wxtu#_-Mx!)W3h8){raw2eWpaPs?T>~Cy zjXwN+BT#OI)!w5&V;945Vx`h;0-?(UbCIAwC!)0>tpezPKH8@1NWf}1(au9+G-D?? z%(83|W1gHo0fC0mt3!?kR0Y*2x;Ry2nON`BMdYF1zC~Q}7P{iGXIwHgwdht>&jFV& zUdG5e#MN5`ha%QVJ19KWBE~)z>UOd}7$TVbMoBu*!gh03@Kaoi1tZY#DT%u=yb7GX zk415AHU<|bf;tmz#NvZ3Xg&1HdofZYw$f+)>Nj+<6T9%sP0*4}Fi>KaPDJPP(P4Ip zd?>-c8hHy-|W_(NPqpx`?w=%K0pK$ zki0Dd9QPRO`o@AcW$Ii1{PWkIK3oe?JFH*EE+;(#o$8vb9x|v4!yClSK^@Sqq`_6A zOqy3F!|Ri_Ghz#mam$+von&Wd(P$ACAxgx5qxe z%7``I>*#uN&9rjzv|uM)u8N!X)Md>yoCoiWn`U27u|>{HRt}cfc=>R4$Z7if1-};< zk+U}K`Y-AD*USnoFhSE!f=MOlZA5e~xy3hCb(%?ZS9dWLk?Nc|J#*lnE@mGjLsmqV zhocV$+ekvH(k`Dv(7V*W=&&rKabd>tRCYaU9W#*+p45EPh>um9|N0;3>S=VV$s<2c zFlPw*Ga`D2{5(ArzKZ+1opVY!|2g{U;}f&<{La&L;?KEV?!WubWEoZ=S3|J>ssMv7 zO(c{~z`rZzG)M+%+zX#lHi#ca^pQ|N7%(&=iM?ZITta%u+imRGOpr-~EZG#;zNp1L5ZT1(Jl zAlo}9>X4lj)lGDR<~R_UV1o6BjD_EoOe8ngiCu$Ma&sR@(?W9RRdQaV}2!O z(4Crb*`YP_mIY%YDlFgZz_XcT+d=ap5vL7GTd7{j_sd#1Og^21)T$?DU!p66YPlK0 zB2i^fANTR198N!QCS2$BPy^z)fGdGTo}1_$ik5em{r8=~e$ygj2RvsR<^H14ANu+!t@?108Ta>KC=fOzqyNjSp_5L0~rv82SjG}UR< zhLvK|z0x7kaZW)*b-)h!IevNgdAdRJRCXv-BhM6-a3*7aJXWkW;KHCJ#@;ju$Lw%X zle}V%xr3jLrRv1H9V_90HPf;^uZQGPC#7-I(#e$}{hW+BePrpPVlrK^26j9hR6V(y z+{1xRT`(pgD%Jh0{0TRK+eB6g^8Fl*GMiClz5qEGSwC1rg#W;}D?a#t{C-NGo(60t zW4Dtn1OwH67l`OqvQ2Q6-V)i)Yb86vwA^;y4Ou%~4oo9S^kKiXe$A4jb8iG^b5fM2 zl+CI;+!LPXgli)%t2WVbvP{km*(qf$_keis?58e<^C^lObH|=j z88in~awpxOu^q%V3s##!8oZz%pixnyaEtStH^-ltN?JOuW)%Z4ub*ep_ z8uB`n8s6cqcUv_LJ0euo=1wnM^9aoI)i)+ezJ z2}{0Y=BLb&zcwyTalbk8fIjQQ;l(>9t7a>~Tp{R7MD&_ziQa>P6VPV4Eab5)mHL8T zD(sM+4P5P8L$(GTnTM<@I#s975Pc95OpeNmMg8GRmS~|78;)O2VbaMX2t!uSN%L)* z4UX1;Tkl}Xokrdx!ToeDiD|yYGnTui1Jz}b@G8}!2EQg&n?r za_Q}VHt!PXv}(vo-@l`fW~@^!!4@U$yb{Hd74u4k!{mq`WV(=uiXQTm@}jtKQIEiI zWv6F>;;G9MSu8ZPEm@E{qtXqp8?N0H4b8x?)(QI0RV04c8ii%_ILtO!jFE)ZLpwvq z>i4j{zz#aU`oEi7MMl*m{qOgh=|m@1O+_ZEDV<=TM|KLvs)s$aA+bObsIKAFlbD9t zA;XGj^$}Rr(y?8{Al(by-!w`{s~gS1vmX+yNobr1u{NFE&a*-c49S>Wg_;i1hbi- zlZa^KKTIc=^Fg|u(9HS91p?6{N}E$#7=vBM_X{! zeCN!@fA{+BsrN~H$+}?ZJgZNUT6ToGyP!{!D&Hx)x1fo9tX$^!b!RLFSfFU^Z4N(v zFP8bn>t4Hxi6mAdI`{@P(Y_G-kyp5Ik7e8ZYqI7S~15k zH`xQn7GOBxQ><`2C1z^mzndE9j|t{0f*vHGzByS|Bsw?$#O$LAi12VQVtk!Z)SP(tp(kRJ;+!%E`&ro<>1LXqE z@+59`K#?d}hQ?GoZ-;!PM-7?n+0Q}JQ+2L*qtNiL+ zz4Z6wws!-Kw$05S+$p2;ER`lY@%9L~*vC*4RS?X6f-WbbuL`aTn&rr?+6?W*OZX-H zLMj)y)f0l+1$9(fNbN$fdv+mu{_ZjJu%MOb8}yuK&;#VR$*grJ>ahr+6{ z-Q5B4ai2b~^^zKL=~C2Tc|gjJ%-!kaWdY{EtPVQ?R77AF3DNDK@3)scT_e8{8;NJ zgJ3iSorcBGRiJrV#H-%gr-=7m5m+7`%ipKyf%a5L!#(Ui1q8F3ptFhScAl2M*|lGK-utZl9&S86 z8aX+nPIc-B*P+u#;i4O|0dWJbMp@^%o7D1qpj7FS0-Uvq+lnehq2D0fF(B6R2kBZn zJyJ`*{{$bir77)l&iWf{VISqZFQxeV1-E}{M9YQwSEkUHoj3yZ&;-V}2&RjmZxGQv zQ1OQCsD|h-piww3H@Mi~nYGiQCfDDk!yH0ZY$_rWCB;~ia$oOEQ>6Oz~91} zr*VQW0Sgsmb*ROchSf_jWq$QvrkK+^zgG5+6%uER!5hO_*fcTMZIj1i5%w4dZd42+ zV90BWup$zBMwHBh@F(V1)RRe6gMm{MBl7 zk4J`kzih2%w&*$2ZcNm1kp>dAk@07TeAs;0k%R@V@LcT@S?R51fpH~L3$xlkWgVzJ_g3;ubf zd|8Uc>Y~r~!mGr}>g{&kmvhO5+BO%X(Ome|zkfoBsKTlL#iBpw|I8 z=*&&9+e6p9dK)l+=YpTU2ciP61mddz!TP$5$Ywq7L~po%cgB|gFhb(pqPm&%ZFZ1w zVlQ8uNfyp9!N4|bfQasfjFn_rqHpc&?U7yc)FxwWqV~|zTaBNMeXEu z13T{;?gqCG*%7xRZnxdl>B5~pL(q>L5@6kj=+>FoV;mcjW3K_Ik^~BX8(tgSFkJ@> zXv0A^|2FW;8ePAqAyZjsjn=&5Say(C2Ae8l8@40U5~gBslQt63;c<*E<_r# zVcR4f^5noHO6XRA7n*$5N}!Qb?z~;f9@%n!pV!veO+NeR8s3OY-Oo{h>ysBtsY?`o z!H7$b?6TNtqfx7G|G3!A2);k5FaDm6bK*^Wj){iZPB6&?y$RqHSmxF4Q73MYJn_@1 zHimaAQ@&OS%V4Q6Ux>lRlVn-Qq^cVmHI06WoT&C-b3dy!lsNL>O22s=%4t(Zt;C$x zgkTG#onFHp^#MbW+L1$6(-_OnB8NFmphe;!wp}jWDefV$E1MR~q?2T!crw7k1|!dG zDK^3$=U7%4nete0=?@cOQgvSMCT)dDbxE3UpG@u?O^jcbDDaFBw zmBSGeOQe`!pc%z(B07iMO*JWpJdt$w+I-+;57D~iaZ=<*6h%yfV4tW6tlD<4NE&!s zA=s#;`ehZOJF?t)K&_|D^XrCa+LxYHaP0={m>$v_gB`RGjR)Ohk?tt0TSxZ~xBt_V zsYbx8ToYJFkGNPLY{A{<#M@?QAw4Fvok%bn2zo8BRzX1%l=#JQ4+g6DM`m*N&bGmM z;{d&56ILAXy|0=w|F6EUGpAR5F(Tx|a<@XFin)vJyhdIp*&mey)~lf+1lvH|l{9cJ z3$REZ%SpPW6_Hr{wk5JE=ti(ko)gl=X>)7?M-YikXfRb3hEJ~&=LD}?R|eOl9;FPom8khs0JyH&)}?ub`iC+u~yDeT4dbFc#f6{V!-n4){$XhnV33Qo^$?(#&zbm z-M{Uk&#?2wIkCxd+hm=&N-&oR`XUkC1?f5uXKtc3baT*>^&XeVbl=wa)Me5Q7vd$6rQDoom(oE>YoK4vA_|cVac)^(wHdBVu*># zn!7UU9?-Hv<>VEnA1Ll1%>`UBptUMPE4dCJov{-F#>mb-hpg1ml)1mz`I9N5c*#8Q zG5y2^Xf|lK-$(!QKcavAhoAlXlHZF~6U+*Nj+vD0=gXb}qBj`0mbcbxe zbR;z0FNR?gx(ug1fxW^l}Pc6dnq+RsDh8}YEXEpG#z|C;TaPnuxl z5WzrR)LtUG1l9~KzmvBCdXk@1)F@ArFP`bspZoNNta8x<>1klC(y97^#t;`BTv+1h zlhQmY8=Bs)g&@Fjx}Eoz8?vJ?6p86m{qK{v!;(DL+ONyNs{v*Ex`89@1uyG>(tPyQ zj_F45-2AuJPwBNzy!hmr040TBK<#ZNq7&auoe{_1$kF&dm8vzqsk~;-zPVZmCSHfG zd86qU`EHY_WZ2=rUQ&MbJl1hw!S0T;-*4%Ow>uK zRyc`IeDrs>?a7ahp1glqJ$dK#(H{C<`VU|2U~%4`9}4cr+9S`J)`R;*`8cX)W&v5j zXyof6Az>Rcsc(^GGck9nMcnI(Cr4Z`{VG|egM2GBMksn9I5F&=Jl(NKw++-5mllU2 z+&ZX3_WYpSoMrh%N1PLHjID4f-vmyU+oUcc50aKN@~m*J=#0E4LJL&yu>0{}uIy5! z%XI_QG~%htl2zX6u=%MXYiP#;1f@JqkT?_|`k%{c0&RxtZ*TEN4cANCj?wGb z`Bj|Qj|=rsW46hu1hb8xw-C|D6tuyuGc=yr7~X(!&CTy6yqoYYWCvj;ClYtuBnt`r z<9NIn1 zjhHEO(VnH(IPqpC+XT$p2xbdGZv^^7??&&sfD-8A3=NkD->Hba&jlvVVRxsm45LdyUB}!8vXjW(r;)krbSK$ zIzT+^0b*6AcD>oevN zq_Xe#-r{@woeWN!tP`xJ^WI;G&wDoojkvVY4M7gZ3wxYQjy6&5=qNQ`YOLt!qEHNe zK@H49=|gyYBQBb#vOt~u8V>?xDRUpVRgvd~tAzPP% zb+a@XyEWfwgJSRN{`QCFW-1#^;>7MXD@H94p(5cwRwF}$dF&dDe6?K*cFgU7I+)6Z zkAfYQfno_pEY}#zQ#j=8uls*~(%ZO!`RVy77wA-W9uX%-#zB+qPCmgvFei(MKA`9$ zH+Y^AXvDxJD1pl3!M8ho&bY=v)%s@tp5T62fB03pC%6xId%8HoQR%`sY>P4Y5URjMG2HnwVmGWkw9If3l}Zx zSoG^|Rr#Vn4lMp$h^O*7?LRpC-q=B_zq#z-Vix%kZc zW5;Ub6^oV~S@i1(p_47*o@`BOh{a48qXv1*ukk>Kn*Ddt^592Tq=@3;a_1ZgIf7j-qD*f$Q%1$v(mpEhHf^`ejc%L>Tm)smQ zI?|ASfVNU=Py+{}SMf7SAj=6k=hQm+5OkGKmMtmdTn#8&uy*y1H2jR`u+2nI+F z3bAWif#|T}VZ<)gU-|(aYD7f`bu>UDKgQ{x_C&QX>TSXzx`Wn+bc!+gAkp_Yw?ozw z3XFnz5dJCT42kiwMxFqy{5@nlSxXsOhTDDX#S%0xQ6v)w%Geq1eEH2@^DUE=y@n?0 z;>^=bmU8f&zPUP8nO6sMh-_iv8EuG8g-HZh=dfm4NocBf+_YQo)I$~spGTX z!yhA;u}^WF$Ly>Mr*-YFdGDX=Uw8Ujtku$~x)$sVJm#ID)NwX&3ug9{!=4FLuUEOK zcJZ>KixR0sD9X}tPK2(co?j}3t?;{esav&&gPAwQ0_@;0I@9kurM28)-hXz;Y9sJ` zseclvr#^a+Zs6?pU&G%Ka$FP-*_SwBv%X^Nny623)ATM{R;mAKkjy)tSu?ATjNuQ4cP%LL%L%+{E$hx9FTt4i;njzh-x78&!JNO} zO)fAla_O>v$*1?RTjZSB=>FVfk*gz^qXZ;>MepIX3b1qx+t?-1D?^~xESBHLE1KN{ z{$E4hTp@9A4t9OPb*wymh(7DlLoS2=upMre5*zAu$W{Uk^WEUqAdI+G1trUFQa8wl zFs|1%Y2KZ{w2&jbT+U7fe!<3Y9g_saBf!46V@CXZEVH)Nq<&Gt4o2OJUuff*->9X& z=4t*;yPvmGf6ylT3IeF7JfWo7b-TEaO!(n7(fB20%=$${_@7mQ0F54@qXtlceQYr?k$ zOr~p%wVCXD-nHi^SIu*C*#xkic;&H@VvJiN=t~8>{Fr|U$iT2vD|Dq%=kt4Ak^f(R zzL~=96qH0qI!iJm*1hT|j}o)f9Wom?G%; zC;MW?=UhbP|1MIvhHs)BeLwIdufqqUNEet(g1r%kln1HFUzuz~g?&YC!)$-EGT@fE z?eSurkQ3_*EA~h}{}`v4+U%bxOkvi`?g`e*dU=yCbr66pg113Vb{}Hh9{E;5#=GXe zJEt{At$5=y=Ct#4s(qrp^SkLvpK`%@Z}iR`QEu_=gEW>4-I zx%r0dgse2<#D*eEk(akg01128u`HL&omb0ka#*t|s{*wdLRM#R%!RCyTOUrOi;xq0 zPpoti%2w*YaoI0FJ<}fkJ(`H~eEH~FOu&yreieF~* z*kAe4sxW$hoyXpZt&~kBO+8jlWtLB+pAylT?%TZ^gGL1V=pnZikkfrYydo-Dw%4mq z0?}vaw^|XYQ>By04VNtIlN|C-osmPvGu^>jYL{=zjFzBdvv&DnURI6xsdTRh`?HpL z-Stn4TFDduahz6&q)r--6SQIeepv#QN~H^PXE)37?iBizvg7TFNC;`G4*tv5@8x{^ zf~wLjmPr?utFSEySV`ot$1dM(Osa1+)d7j3`14kKE;8TCMTRnST4@)_7Yvq`Yw9d*AnKBNs29v~f}(X*O1 zWMC!BcDUxao@Pq@6M}kYq>434uzk*Z@AS-Z#RyV5siPsBmqy_r?h~D zR(fvDyyD2Ro?&;Eoy8;gN$Q4*}6VdI?k7{&@3Q)jJwLU4@Bn~T3oEWOH!=@&9hLd?kovx%S+iRfdCipg}v zQ)xxSE+zVGI%FrME9m9?lVpvweNuYBjFp?sxEOoWBpkEC1tH^Rdx(v=5Zo91nofFS z%pM@18I!B2B^V&?PbH$0ISq>YWGsD6)U2xf*)`Dxl}?4!$9R5+?`k^V?~>OV$Q4VG z=P0|}v&bF3pFFgdtYTMj+6<_%>yBN20?U)R-CMQId>{Q{9KeY;MpkyqI(eyA?d+6* zQfi%8jeVK3l<`b5CoXDImlyprw^j6(1?0H1)^$lH`G*nnkfGDssa85eHSl^M2@ach z4~A_7?iHOJ+v+0cz9U?%FPD;y2zdD<4!CL^0Tx_tQ;Qxy`q3QYYO?ZoGdIxb>{b&e z4klKbtR{O12Kr;=5z$S8Ob({Kt@2KDQ*ZIbY=dU+N^f-3q9?bO91Q>T@y1?4 zn10+MZWZY`hVxgUFXdO#J+dvnqvIR`J{$2hc4>Lx*a7gGlKAQ2zSrF@&F61$)5xEA z-+Cv{>oN%P5}C?7GAkJ=^&Qu~!>TZBm$cO>9CCr3CMjGxkY*kLV&lAbVt1UCaM5LE zO-SX!Q_6D?`b+d}=jD*G+%8U@Unj59r%RPeUHhwRq8I`?!2?BlNQHF2UR=W6&S&hz^tVUG?1HC^E649lc)$STOxJG#ilzQ0X zdPpU=J9MYQunK8IP*FiTwrt5+`g+I>aMr53mFK-<0y}+{sIivOa6R7*>W*Mg-RuM6 ziilcj$%u<##w2dPC|QQtbv+Nh-oXFD~tk z5<_JcR&`DAsf_Pux8+%mZ{~^J#Y+7AQ=6{nTZD zM3LudNL$!WFDE8OsPQTk*@9ndtH!J5p2&4p_$Ah#k8LqGYpr-SF?|`r97z-=EpPG7 z^ia1$FTo-v*FD20OZfkXXAMGQ_4u7veH{DlzdkW9jdxnNiIrmaO1B=#utSdWoTrUy>i3dmM#Ohgl~xtA4er~c$WAG?@$W^h_`Sn0x0 zb|{8 zEi4I%El3Q@jHQ|EdyW-Ih=hN>c-TA#`wiI{TgljNlkHM=aj;+`EA+(d;~`%$Nz~n7 z+`RnK&QGB*Z^7Io9O+Km2P)u z#qc+9hbK@jYE}N@gP6HJV|Ds*Uubx+^gFM+yWWuRnk*Zns{~l_3z@N5z>ZP9U=llW zj0(V3IggH?tbbXllBt9LdSjZvxS3g-cKw%hJUa)86Fb=oOtwBd2_}`Gw?Qvd;jMQP zWpUqo@GXtJE38`?2Mh>zp?}CH%T|B$(g%0mt6g;C$B^~rHzeLNr=I+;Wt+Y~877Ra zXfn|glK`pS!)k>&i|K0>`L)7qismUQ4(NzCJROusvaBkoS>L?_N=7?L!-+l8b8kaB z^<>B&M9CfklOTWmxvZKNu7R7M{@SQK4*twTLf@H2uQkaueMB%r1pR=BZgndTZ4sYv z1$HBkD&M8;ye+=yVStVokSVCn^SkM#k!zHP{W@fNuCz_un_jh4I}aIn@bubQ{lFih z?scu6Q$<36p1)7JQj+gmL0)0@dlz%#XQ=P`H&R(7PJ@@vD_6UNAfcGczdRh%pY_*s z$t+o?cu-lI<0PTuic@1cQq+|><ee=Yq-X1sLv<4?1W%nSEX#h~fs!EX z18SUy(p}=(xs{Nlf-E~#KrfgNS^G80I?tueR6A_BwQkEp+5|TwsE!iWQh&ciPPM4CR*kc!PFDNNQDmTY6<0#?jlq6P+k zfLQ{!T+k<~C*ky!`EYu~V)Vnd!u(!L^O_)FzX_Et<0UTT}uAt zY$KkYFsIkj8cWK_iTC42Ou%1EFt9=2O+?2|&jr`rCBfG4<-XJ7g*JC8{vIgi<3*&^G2cw~a3HJP#V{e&AUia;1We^oUHL z+8~LyO_EDylj>aI9$2gGrG@OFcuWi*)+)yG4Topkst7Pbr>v(mmd;ot!tvfMK@$RRpFwCVXZ;s(tWer1RY^Xo*A4BUg~ z;H`w?d``A^N0>&AtP2iB%($RpTuJPT4YSEuL4}xJb5&vM*+wTO=@ zD!C&r2mFq64c;{5C8(oD!q9oK-ov3!-Sn4bCj--a9r=ls44l^?=IeH0qj}w`)o7N5 zrOJ7KGYB4x-2!bQm_&l!0E{jkrPK|tD+{3z4Ga?O>bg7%U3la3x=jdTwqkoiUowuH zn?9@z{H}3v30O7%$MlIeX3x`M0*Z?S15Eu5hA7h!SW=J#`RLqMAU%e_cN?87LklID zQz=)kpjRn(kv9XAxJ8Qn3hXF8Fn3kJh@Z~8l9M8@73YwBVaMN9S1EQe4bl>7J@5gc z7jndmr>N zvcavATON)Bkb+q|odrN`1(wOKX96E=#aQ~_r(4arS8_}wvYlX(33?L|jegfQXkM8* z0~onTbq&Oo)5$a7w*_}DMg#F0izj01PiTD{yJu>B9(!MTLl&a{`1smf^M%NY2FE~e zCo~zVbVElKq&JF2LI=WP9BT_&7MpFLuzas0&$0rA^SUv=apT(|HrAG52b4??Y;e*P zIgAAN)e9!Gk=%1(+60N`!zb;C6(pS3tn1F=|8$noHusP;_ zlbzUqR&Js)vIz#df2R}C1^jKm_?bWzQkVpa+}4otCSEU%XA*s(tsPw|D}{J4CXr&2 zN@4gg2g}c}+~a|?-)_+tYv8dNz~=hG+JtsG&kDdqs_2K^<~1Q~{3Fh5ThUespGx>8 z{5Hww^Rl7QLlv1Wd@4nbp3`)b&t$7Wn^OwxgbdrKS(}+HnQqh`H~+2mQ+llvYmZzL z?U6z-5J}ohM0YFuIX#dcIqa^JpOsg-Wz1<6K$?GtY_~_1LXCaI+Id6n`zEp14@-?< zySOGel9k4ox~J(v>Qo~(R;~%Gqeonvw%aQWY>#8oL=p*R13|APqBr}W_wE$c_^J0v zd&x5YBwBrTGJXEfXmddXTrTTY?Y)<_!w(n*gEk9$Fai4)IERG{?_-68sk=2 z?LGRl*KAAMZK6dk6U;?|{+x&|qRXiBl0t5k;ABLNbi<6joFVrH9-gU@=FPn!I7ya) z{hCc~ow--kAidAsCqz#BUb2)t;$J=IzWW~TeaLbIUX{Q>+2|+3cpR50!uIg%=~`+J z_as>lh0miiT@9$7(+l034WIW)*VE_uxo$cYZjzCD1c>Uob$FeusysxpKF(!hHSKn(LJHJRN z-sNgsqw2q2^A~!h6YqGyJu=3s-9j)M2|9s@M(X0Re}Q3E(BpxnKgrYCtOBoojy8<2 zRejhnm0gEehkyn3F?IT#+AQ-jDWiH7XZpmT4h`*#BE zNTP4ofW4m|BAprtLxtF@rPm?eI>)NyJ3Wu$szZ7-Cg5Iw{w3 zvxWC%X}$?mrYKjO=XaISy4y*bw#vxDdurvBeMH4dM$W6dXD2N5FmE$vr9W3EiQ9{O zJ~p#4{4sDyj()O}z-<+e?ZB{gM*ECo1q|nn@!kK+$mRcC{Wf-cAm3#1;$<-e5>Bdj zdm!WFz~VLE#MWQujf|6cPM?CmIDz{3+z-nC$?>qa9qA?-&pvQ|$t;J`9)IX{# z=%)CM{N|(Y7{PMlqpM%i$DJ4~?IvJpCYVNoK1W2ajH-ol3S_rH6E+zr0wH^qT*to{ zn(3y#E2)6JOulfB6tObmf_dD)xFYIx+vM?BWLP?Z$7evGf$l9B+17x|zeaor$W%vM zI-$1tkpDoqKCMn)2#=9)Os;R1tG5U%B8P(8f-8ZYW-avHZ4o1N^N0(2`V8bh>0}zv z0kKQPn<%WF;mei-7Pw;ttEnIEyk2ciGU&8^T`RPKwcJL}82(8`Tkt+%j%$1H5a^d= zSZ+#XmCQ+Lxd2-#*NCr)4pIZ+eWLT;aRImp)sy=dt53aqZ4ri_4Qo;!b&b~v?J)QH zwFMvYuOCNs{Hx?-C(yj=Ygk`r1sa0t4lkZ&)J3Y!->K>4POOX4O?1&lf=M9g^+Yr( z6sYE$-62~k(Rg4cE~)~@8zV1@#wJX>c$>q|u)@UD4}VZAfI z%rTqQ+yh_ zwB%AXWDhWVt(wspl0&Ro{Kw72qtK2ptR-F#EIVZjp$?FnqQu)Ai3nSTZ_&^&6PP_e zvQi+#kNtjCu*A5iEUJzQq1)b=f}IhQCFnlE+$HEcMD(B3jgn!H80dR@Qkp=Oha+i4 zH`%qI-XE#qhQb>oV(Bb$@asTukqoqh{a`OA%b<1g+LhxfWB1jTVR%0@`fQ#XV0zBaNyd^7w7OKbX{m!Ss}X^kq#dVlp= z_&M;WXc4E31|}zjP>Zmg=t%?4ll^}Bbw4BKezWIzBAxlhV6N5#bL9jBH10)2^bM~@ zNjEH7L*ZTSYok&nJ+N`Yf~_vjuzRU1?B5YC>a(6%q2**{P?j>oquL7!-CM-zLOoH& z4%Z!?X?*n=sx7!hyxyah>gB=Z8gVzY)vS@;4z{PBvny)ssMrR-J^y|4G9RbiGg+xW zI_R#G0#}%#RggS=>82okXOi=fI_^L}j5P_^DrRfj{?=> z*5lvEGkkAl&?dSfasrCQR)CFfO&LG@tOIP(yw2L!y+4RyT^Z%_$7SWI^q?zp9qKKF ze13HRFg;(Gb8F82h)aAtzcKs~jLD^zO+eN*Nn_lAuxa-353&vj3vufy(Q8spn)l#+ zN%Pud*-o!kx1|lPnWCFs@oz0ng3PpJ*}(-Dq?K;Vx!0g8E!qXivNU;FAR2jKbL!cM zB_;EAb9STH+JHr^0DQ9Pu-2~%C=azincBDoWA^HfUj&x#WG5J>O_fgY`i_D&n!9B# z+OzbUX*AH-jL9?EMlf3ldLt2C6;vNEB)%)jowpVc0W<=4JQ{)b3{NiYhVAg5cY;~y z==XGtFxVku%#98{WQ7dp^>O+9+x;cxqy|oF(pq6#z*4dm-0k9aNW^KQzo6qI&d%Hf zop1rQz}%NIw~6eXOvJLwkaI9hO{QmsSt9sy`j=CTN<*1>;A8rU6W4;InP{_kf>}q< zYl!F#!1QN``ss`6mp%V)5>*&Ga#<6JIral9JWMV5bAH_mWI%JhTF?_Z9 zsdQt6!H+qq&)o2v+GbdoK9$M%#OldAuVE4KZ`~c{q2f2B_Si~bxsxr^N*D(2vwlRc77vO5@(UZL07|GVXrS^F%K(9 zO#NtR<7eg-Tx_ff=e3i7OM0wMc~arUWSAb_L#~=;erENwX4M6iPL(V}@~Lco zrpImfYN|hQAKfazRG3u%ql*f~N%UpM5mVXz3O&Qx-l#A- zJLjeU_eMI$i8Gszn=Bdo2?pxOONeNU$FK2TM3q&0-5`15cP*fnx(V%22FX|;D!3+E$!+Il%(=}g;%pc9 z&pqOqWv{~2QG`tt(a+&MI-`C5Wrt|>)!4stHv*^PTOTyh>!;B`iaBPRkVY`u3HW8P z8*F$5IKTRXhuveMDk62zMkr0bcedI<$lWD7%EjH~QIBHZNn7993c)^)_|>zp5`y#c z)E!Rwh+KVnl`UxZ)P!}i)#lW`){7L#YTZZ5Y z;#wnbkV0omtgV_9`;;wm4eH|6$eZX0R>+b&jxlh4_6X21hH0Xh z65OJ1R57W1*6gq`QHk4rKRawB*6`C__r}&()Hy`=Nm7AY&A>~1#}nzzcg!gA)W|V0 z3`yAT(Un}a8hY>zhR5(@z*q86vC8|dB!*a4v*^Jh2QG%}fb!hpaI`XE1(Yct+^BUi zUuD>MN1S*4q^+z;h8#mIBub4%l-Mjox@bc{4!YqC zxvzF_C+~P3n6(ZXuY9s>SeDAWN_WU~RFhAo+m(e)J{4pyqc3#$oUN|MbP0K>r|LGxeb_oHcvmzRJ!38f2p_$)T!fDT;qU| zxoUd(=W!xz{~v$+f6d)(Z1jv1!^lc%85Rp0x^Y0XFP-d{8R!UL=S^EhtXOMb9aWz;a8A+u8H!|XJS6W2#vHqkJr2?jXA>WJtA;yqDY-Euuw zO&dKjcnJ+Y!gQfFq(_iNCeAE=t0fRC8{()M|0~SKh>PL@_ZWVapnPE+wcP)DXoG7# znHGrDD0mO}7{L#m=G!T~I%{AqItGyXb9Knx*(26$ILnsusQ@s zj91ODboIVmrC7Q;4jdV7=}Y*FCt*a>6OSk7fNRZZJ%{uT6MQ5P%zA>3Bcj&;b*!F{ zm7URO-|80ev-)s!#$c)?rd_5=piM%u48zLUK6KKu62^mbJHpxOR7s`Z35qZ( zmTkFGExq4~9kk6R8t(+bfC{W3qO+A%hFb>3?>(K6P59b*F7(bu$c+z;xp9#`d5IwtLk`8=_f511d4rjAv zc4AmqsT+^wVtf1jZV%}-vtj~!{PQ?l-8J$ePc4}ivd`~O$dVra5~Y@31amId za671aatn2CMqe0&l)$T!$Ig6Rc?W8;bOVz25Quhlb0H`xn= zuhD7K=n9iHp@3j^6HwY2t>vGa-yuCG&q@r>oJ8u@B&F9`{pxCvS9Tg^j@R@!MH7D~M!6Jfu1^@0R$Ikuqfb%nBUrAD3rI6%Lrj>(%uwayY7;}tdQ z`&ePad3O&xvVOhI#_YHP>XNvVhs|uZd-rj>g^lu)-rIp;^)%hdI}VL&YCSt-=NGM@ zd;D?4YSmD9x3C0yy&@|}r&oG-gW#dKQNEntwP3H~@WT==FX(C3II!^sR&a6Nwf@eh zf8?1bIXP`czLo5&6~No4Q@z#qE?doQ`;pNs`Ka_4!*t^tE zH7wAnS_SpwO;ICh_;iPR!V`}w-EPlp^0`mm^}po%C|IZ3U`FI5zi*?8b!mEOg(lr64)CsWWo`NU3fV?%|7F@GL0e-ffL^SH64q7MO=eJF$@hGEC(+NRYo`;nX zC4v?+I@Jkf4S7OwFsd$~+6(M<>2OpXbyyik9dz7U`b-+*#KN=t9d%$`I|=@9EHiu* zB?k($bh#6I^MD$EOqTF5f&rHEN+KE?!|7DGziopiw)uX#h@CziVY@v>{C1O{sN3=7mmHh0!j z?yZaHmOi3Celow@dlOh4kLbRs>`tfppOYBwdw$eAHEUrBL?9l~C&>}NZb=@ML!x#= ze_cDT+`Cjb;@9YUAhJtkIIf7;rL2(kOSii373Fa9Im@Y2N*rsGavBG|Ua5RuHNY11 zvBTHbZ~gjv-!tOtle|N>=#%UWYUho|(p@HK)Dg@Df^Gsch*uAZcKx)LIpB933X?7i zHgbn$>MR;_?{wa4Wg{+)l4MzL;6oY*q|oXsOh16+yf_ACm86u;bg$bQ$!OZPwMYHL#8{{Rqe!QqsjC4K(05F@mt|E1nZ z?|EaOb=Cx0)dX{d0PlEoi};>ky$qXvH%Ln;AhaHK-|1Bzj@XIgo{G?)tO`01*~07+ z-{&F()_7L&t^zgoh|4x=IUn|GKIk9K0dIIelwSj5@@)Y&ySgdqfEECFe|h*pat&Xd z$tess%wj7_UM4aJA?RgaY16fD-}~;&*WEhmmtmU30nRdrF?GKC-6bCxSDzxcu(#;z zPHZ?oHCa!338sgjZxhjAD;!>gnGeI{fcFW`an99%=FoIuxuQ9=O*RA@^xLotFAyD8 z6wlCd%NJZ0^m9rSCqi2o4Fo~%c|Vmkxt=7ud0OQK3645C7f0+9YUD^QP&ua{qLhi{ zmqx4v+kA9<48JMp&ev{-J@zb*xCidWeS$={Ch39so#NrB)ATTO{?@6`83^o#{>ags zOAWp=dM~!GDfQYVzTwpjd%lR?h z3Ub7wG}VUxFhcI#qPm&%ZA*?rC$^a5rjB(J4HL{mf*v5Eu>u!06c!}pi6C_q$(c_n z)5&db#exA{0ksPHf`vPQX(2TMYeELYHVSK`D;fRCRz)9JwUyjwC-DT9aiY0h67d;T z>aVT~C?xSq&|vE1bxYB!+#psTRiN3Nv$#*9@r@5rBWb2aUd2PB87=9wnd|9P@uT3i zvg;!7YRiu*w2)bgRuzu?%r%{C2)z%>>VtPKyQ-78SOBu``5QRdWEL<&>{O_$0}hII%!r?_k>`Ze&NlQXg6psr0&|MK2PZCzYP7b?1?^-k zKV8^E#<|52mK2PIcY7SCCWwl0e4}l*SVx5gpW~FbCI6zF$iB>ZW5=|WeOWtCJ3j&9 z9T$TzyQ|~<>rmO!tXeBu<6T4_f#VCRG~WU7m!ia3862#3**fz};TicZ&x7>Z2%T!1 zaM=Bt=(8CQBe1RTcJBk6tuuROb}LWN+WC2~C>KN&h9?DJqJK7-?v@3)tThvNZ|em= zLA+VeANFa^M>Am)qDXUM1X)QyJtx{9(IHIr&jph2L|G|x=t~URBCIE~LW>2C(>GS$ zep_K-<#vaiCc6L1f2wf_TDc~$4$4xUHe(F}qGN5HM1t8s&})h40!49XEI&`wLoQvq z^awEKXykEjlehPJCL0@N{A~PW9k5$!S}y+VQ?3ya^HToiB)!^+jS{dy#so>X63k|T zPQvc<)eDkk`Ov#NS%w{0k#zAH(CCgXY&33*!vQ=rqXl$IpPV)hanrmjSI z#<e zfl?t!QA+5#2^U?|x7=|j3(Ow3cH>8Kf2YBk7Y3*Uq zPYg8ACEi_-pu9j*-=V8(9NZtHAZ8 z@vzZ>74T#Qu5sD3*1b2H70Gc6@IjJysAKcxid5{g-V3A!WB{o-7|FXWUCP6(qmd&r zPgLRzQL+wcSNO^aS^x~j`K-Z%6;t!fTSkPZXO3u_i+{Y^=55{*<*>;n3%y#n8fAzJ zA$t2+2x!U&E}MCWd>o{of|Z^c1v1QBpY=$b0|McYbZ}ikio>sYT!Prz(!c{^jpC|Y zE5hp)!CE_t&5vmQpO_WfO!hFNu(BuAw(?20`NqX+-q{WN$;$Br`sIe~Fw!Vy14WV{ z>`5;p`F@Y&x$-PJk=G$rpHwb$O$o~KO%1FMR_msuaO31FIci)j#%d?Zeh!BD!t8L0 z)f_mkN%q`3VY@lUh{Ku^3*;w;`pyKA4z?>;!SFbwkdyf8%2!X(sX!`!n68px<3KjO z4QfJnd31PhgI0pO;l(nnWl0Fux$G2V39vo1YhuwXJaEi?g3kHG_jrtoy`I9pl<&izFp$-K%-Cy?n{iN1(-g;0u=*& zUAapeUya+86nV0!Y_?j*E1R7HZfJ3Ez5+bP9?;0vC~om`LBh9Y3Y^GPwhoj#p&N7emj!sEf49HV1Ulf4wIE=)Ua68-21=#lDDZ z0Xooz@9>O)A40vt*5<`BL|B5>@<2Ae!V0t$FYwo24K#W)MID7PBr_H3(XL#+Lr zTRZLljK~@H_&=tSHg?EyTzg5Dn3T)hr%@ERHGTbdR$|ebv_SecyE?qhp=g0hkta?Cg01R zrPI^=|M~#n_eE5Pf4x^Hx$K$dznFRe9ilNzvU8u)iO|wHt*_yOIQUXey-&6>ovY*R zh{Ahv18;cfeeg;ma^eSPj$w9B(b6wq>D4erHdOwc#5?BtqZVX7F|laKG^>++TNTZcl!zYF-`EpW+xN)2 z-?Du_!*HXs@>Lvn!EkH5@8e-KNe%DrSx8ccs}FVH)MlxPtCUMIP|}`3sdUWZN%e|! zpA8TRDDkd`P2VQpUe^lxF)z*a=8QUEe#Ce{MKGR^7E}hF6!kgbGkTZGz#Cy&x>j5q zP$#veGh_}T>ws-xeK8-*hOaFHTj%=K@4x7^H0^XvK&5bp5CR2;O}~lbV8gHRIYU`9 z`_CD|E|)FO{xB=bIPa`0$IXdxEbz}{b8nI+Pd#@VUCGPlw(<2-P_OO!tYXg+@iAU4 zPanK&X63vqQ8(0NpM`-to%4;gO%xqkCp`eIRRT)JB0mW zu)zS4Yrn>SQP_|Ep z4}_k9Cr!{>X$!#Z^7@xG~+}l$F+f$@~b7^xGb*V*YSH`d)mUk7+fr?<)-a@2>Z5?J=2;Z2VZ{?2Gr7CRdiEi1t-=|OSgp;$?kDCd1eQ2}t=``8Wp@oI-gd5-6NfNtKVRRH(GX9hKZk0vhM*_ z8{}6oT~HIC6=7tp9RSjRl{rVqJ=Z4p4k)fUNkainm)ua^v%&?*tdQmrTlTx;XGt9t zm#Y&*_og&Jjk$f0Gg7z=u>~Up#Iqk`1($J`!#69ujVse*=2#rjIPjhs=%J*?I_*M}LX%xoVKvH@TQu=DKcXMdF)U@qMK-)ZI*Hn-itFq99n3@K9^q?cv(!IM>p8M4L^9d99I`7m zL0%Tc0B?tuQWeuD-1WgNT=ir54Z2IV#DCF#ZYz-braBLB2vr2fyGIA*iBo}b6A}3+ zuxQ$TZnmLLx61@4 zTPX${pUsr&EVSYmL|v2L8DQ6IpV#Y}z`5gf5-L@aMez}<{;kBjh2QqZmUkNG9Rksd zUT(7QHEEip*iI{#$sCOgBql@H<;DHOkK_s?NFtB@w49_n@WNDSGVO&Fvx6e}lxqF7 zvvg1RQfx7j7YHtr<>T>H4Cj&ux-u&!)yYy+3G?IUpL#QX{*t$vBDHkc?7MUq1a8Y_ zYh6+Uih`OVyJw{Pba78jshyJJyL-a{eo2ZTd7*YY%p6$wW@^$W@yu+{K9g64#{-_+$PHrxz z)%&Dqz3Y;o;>fs&(;kQD)xP#X(-RbRi<9$c0kzzwPt!7(XLtl-)HsLY5kBkxs%n;)N$)pIuEPU zziULrZ`*%+oqWR1F>zoA_EQsZU!a(C6gh+7&Z4pD5@L2xl9f&>XTjd6LD8x}{*D`T z=2SIqa9Zer&6DALXT%9Iz{!gJC8y{fQZK~|F@ltcco%pYSwDapm3N&RJ*Ot%&P3EF zxEQQqGL?A(Ev=uTpQ64BimjO8PLbc75sRD<^Z+?X5(sl-Ljw!$u|V1j_)}L-(ldF2 z4!VL{6VOh#%TVqec+#F@v#kD=n5|!g z4%wNnpqQl;iJ??+|GoXK=Sh3dJZ6@hvdjz(!VMgyZ)oIkGHY#zRQNjzQ)!M^sO4Wg9LBcv0${9`%QuyMX7 z{L~A&#Kb!x>bIcNqlzw;B?QJ$i-GN-nuFBbxztKZ|Fid--uxh*djA6ThcEu+gM8}v zJ9-zaZMZ8>oqCtoNvgdYA~n=z>hq|3QH|6x>f?EfHqJXU@BLqFq<*oCI{vLif4%+H z57c<2>N|_p&+Gi==TWBDtZuKhBbF>c(K?77&HBIpq>MJ{vD@Plq|brrRDtAh$byqV zF{>!Df>M3@&HkBP+&!K;=Nm-5N!~g2@3ZJB8l6~PtUd#;ax zb{CUMS@b5wCPk8$K5Xf$dR`~pCrgR|8HrADN@#_!nzKQKH)R7xdbt;a)f+_70z82) z>ZIC%i~UYWMx#C|yTCcDI0iZ>(SjT$+&LiD<$i+fXEF#?2S}%crAB8aQs$fmlcy zVA)l@biZJ~081nKouHL<(wCG*nLs`M!IWe7kYtmMU{n0aI#@PUA$YXXtJRTAM zU9Qo0aq@SHAxjH8QWbeXIQyE%y_pb< z!asXog97P?bTPPbDRT4*`<)87M?m5M84t7Q+aY;^OnS4TcaHk7OiLe@9adCJ)X5wz zoV!cMkls0&;x+<9w9pR%mU_krc92hnC_L5ggb{W1UZCVYuBej4&WY!#QCtjYbAiVR zC%uJ+jc{n5s3u^q5(4v?bUA;qpoip%fDjV|K;t5glPh5zz?|6clpr!ZibPQAGhr58 z2&-q)%q_It%i#gRYBn7=j-K2k-(zK3atYZ=C-B-K-{RZky?WxVfQ5Jl_bu&X*;qO- z9af`a_{p=|*_Q+realN8npb*1-6!q1sD?_pp_VdsnEkc&pDY zA9bE!Kx}VJT@gf|wE(soQY{Q@8Nlk@A^-k?aY2%2$pz$d2X<6eo6ybmQ_LfZd_k!$ ziP!jEmACkQ?1pq?uxg?jM7?x3>4L7qPZUrRCtFF{==;!9fjhwQ84jR`lNK5BIBjXXl32_S_x{|_+!*FrKz{_Cq#!lIZ*RUkhAcX<% ziCUluqBgo!-tM_dv1fLhq@6DGDDeg{(bY^Qq@!^2-Rs%`Bjkzj#x8k^e7Oto9KuaN zVGrcRdu4Ibey3BANH=_Jd}sW~#h|f_X>N78YnL1o>h;nJ$Vt5%jB~I2wfT=(FFIpX z5-7t~&@k2dDOlrB?w@SudMHc3^Aw7Z_kR6b_3wZBt9O6*3-L0FSwfL$6R*j_qjo)Q z+5QaE@%YMWDoC=>Xp@S)Psm97OOw=IYGT12P|Q7w+@(~y(2GJ)3Pups&PO5NW(jhO zFYK9EENh&77Ayzuv_UiW0;%R;Ix5%W0a2q^&Y*wZ3Zge?f^wBt}vQptK9NLI27m$<_oq0xoR@x*qbHSXAMNKVSL~Ufd7Idct89|R|$6CEkr|mTWa8H zRii#oo(OMsSj(mCkNn)+L$cs>py1+wMi$OOCPeH(m$^Z4YV1`tt#vx=`U4+sa8^Hh z#|^7zUQ8%{(X*Fn9E58kZzF>xdOaamL-NJ@0y-mB$@lp^f&z#-VbVloQmb^Uk_==4 zmIJFpO(2B(P!7F?J`mn22g4*Qn;jigA&d^%7yuFq=yESyJO|tTpI=Ki96&6bv(||= zS=B7Ru<$PIPqV}Gk3alm-e`D-9QOxdiFc?CJRcB)!`jLd%17cBSRPin^*bRSA7+>O zovzC7g`?w&${qN6Y_?}3fD?OcjNl2)XYUDLbfL&|^23S@uDV*%I2%HU7lj>e3z6pi zd4;tj__m(&h_#daoMCK!nVrp;J8SOuJ&lgwIj6B(`$?ffEnqVq%UBj6%KmXD^$dy-?k zD?Y8VGAfm{@q6hOekpe)Nmc5^iLQhM7_!(w$|zxCW8J$y+2RVVc?a@7>Ark+lO)ZXSjO++$EUN?-% z;&bcW>S$ETN(z5SV=)F~i{M^-25PHhJ*0wr3NLb zE(acj!0*Uvb8HHA8`>sTP=DnQ)N4O}(dDyHb4H`sDJy{;(Dhjdl~;x7$~$x+kgVo% zis==cI_E?5;)#%I;-8TeaEobd*6(vFraO6U{B~KNQx$N37>>3o%X|$-<-jA3M|VgC z^uLzPu6A#gr*XC_`@^(;i9CE4AY%p9o5sjI?%&k$-&*vy6W?iwj02j^KVOr7cwdIJ ztjlI%ze5ZEz6|pV_{>Asq=@a}>#l`#wr>@zqgUn8kf$7w@rb+-cwu@rw{@mQVYs6v z@UW~Qa`(KX2)y2C90#8nU}_r+_*p>ZGfqQFpH)St_dCI;oR{kalR={SJv_dUmqd6!UoSZNO$|;>%-Fq&-6LrQEEWXj9M-} z1p7?=6G13-R zBCGXW92)Ow7zHSXJo{utpn1HK-0`U6Z*w~g&XGQ1spm~;r}z+E$FJe_$}58PVXaJ` z(@yRxj?VcY=@2Gxjw=ed;8}(Cy6PmYOmrZaD8L{^>*d3E32VP8piF`=WV>-oXD6)xCH8b1YTV=;M&5UMN#?*ZA#))xq zJkm~Wb?y#78MTHJ=h763uQQdaTw3Kl;d+=_eR|Fn0Dyd^miw z+=!74NrC6c`tbyGV~0#dHpM_aunt3WxBc7beF0blvm!8tca2jedjMs?KaQerJlSdjlw^uYqDTUzY7c7!H}(qT`ymHV8JNv&p>?4KuA6_Bi4Vf-C4q&T zuuyBRd*G<+A+exjp!LOyKdkoCg{H_cH95c>Q!ho40t`jO2~q4`RlH-o zn!r&{^q|$S@~_pnuyQQ^~*6x$!O zn-7QG=)7J#?vOd9C!1Zm<9b~{a5U5x|B?W-8G86DeeQ>x@OmUxW8HcM9UHm{Qr}Hv zbto39Yuu^=v0fAYAma5?uuS|ozY6q;fxNmER=Hl+UEB+ts)!~Io-(L{*(8f$3e%B8 z-}JX&I_l_OeZTgtX!Ftv3pEKTU$6Pmr6`T!G&hr^0)fY=;ASRYkWFutJoanh4_-}o z-8=OZu--LA9tb}%agQw7r&)d%grv85eLC|PsfA#P9an+TeaoZ&HZ6_tk{FoU%e~4M zJ(7YACPuQ{>@OvKGd+wklZJQqEF`HfP5EA_2_$kU1_&ZDC>2sMUiQlZjjr2L zq*{bJa1@2af{B{hDBx8gG+b6!0QF+qlspk$-V9u>M*SvCdRB9=C&Z>D%G}6FVk;R9NLJF?jHb;Y0UUxTIC#gC!;gL%vy?B zLy>q&b(1dgYL+0;l)4bIJl8zdO=#yQIye8bZ|Y0aVi_pmi~sAcxnKH{_#77Ufi(eH z0*uz;ilT0uy(6rJHqg>Q*`7vu4T@-@XXXMOxM9sn3a^0DiPFg{?H%BOxftI*d>(x{uLMq0RsyiSpJXzyuW=;8QeNPp3mWQ&xjUObMRR#slnGacW*}so2PsdTf-;h zbX>`cPTE^9%?{NlGY4`wsOM+%$+PZDfBqC(H;%3Evb?Sg=UEg~szN3i9T2}?=@*cV z4(y;pdGQbjbsNQGQIM@xWpH;uVVZVox##A996FJg!`-YnL$HMH_T+n>eNOjXFA3rV zeNNrN_1qPKg(7`$FWu*KBq*1w^)H5&*E(6Rw_QQ>e5gG2!~gPxt>H0rJXTPlnx_8z zcjk#3hfNAwsCL19%;u@;4X+X#u^?K~)I6J4M6NDLcRbU0e&dn6fd>Z>G| zv)vt943KdFnOknnRHp=8w`0V2G(Z8w44qq>9?5Lh?r=)K^v=77eN&hRM$Xr_%ZKR zNHwQQc?`-Osw91~1kM@q31_eW=77CX`uUflFwC-(TLt8Yf2o|;5Oo%4zp9kD%p|(U z`StttgdpMkN|NsSxuQx|&B@@N61It}{0>92|E-Wi;!VDHJT^)$1lCD&T-Lf2ajIb# zwwrWM-{Q3{yh@30p5xLnBSVbWmqq4#r7HC<>s?#vGLa#gvkF+zG>UsuAM-AX@6gL4 zZ*aH!d``y;o@;;Epv>Iu8O`hG!R@>I-ud4}MwFfZx1}@5z2O{X2gcG`lb)y8am?az zqz{CQed_o-xmQ7f;HNEBc&v&8_7&M)o;1J2yn054`39p*u;8 zpxL?4X|u_h&CYn&A%ahz3f>cnA-`OpqW%mRI@A1fxEEd8Bd^bD9k{;Urw2e$!^0;s z!mH?Q9K(CI3ro3%vsJX=bFAX5_tB7iXgBTyk&FXeyl-1rwYTZ$=`@ahLi`BE!6z<7 zq;c?xN1TlCC!ablBAd<(%?QUU9YTChhb6tUylDUJ*P=Re4l9S<*cX_XFAc?PqR2){ zwQOb_5Ur;9XL0+SuF!fBMte(StH=rQNf35R^N%5ha|xVipnE;}e4FvS->!An*;^j# z#vSclmccNL2S=S-^qYKNqn#?~|Ir?zdudj~3KKZ(qL}Rz$%BxhxI)+!etnicqFtFf z2}6lh^mgv{SvP0YNgw;|6$7yuRsJg{ z9fNhUR8pwC%zNN|ie77<)F*p@J-wif7EHrE-<+A3m=J3zgY)<(jNz(c&=|U+?QwFxzSC%b?-5f zyV%*5M_0>lPB7Y*8~=ClFC^}zu`R%THY8n~RPuuTA*rfgg#E6X|C*5(f+<}u*kc%2J+Z!k* znIcJ)sy`rlj=DT5dQKuw5C0nW+xK0%!Yc{LvQWrWKbBL?s-VGsGyC+&tgOZ>SC)VM zFaKEbeoI6xd@XQZHSs#=3nWia&cFHEk%_QBF24DsSii{5`VyrExfbC2z7CsxDdO9NthU@y$29gls!n1qfI#z zhUHd^>;{2hr;r6MhJD^%CplA7ip`63pH|s&;MlE&Qe3U*=8QG+qw_CEJvnNK%$&FQ zYmJe}oRBsL`%Pm6g~0E116aQdcZ~%mSRmnvX|UJy4;k08=bE~xgZH9yH@5-b8wgPe zpKkP<(x&^hNoZx1K}l!XZML9RH)RKB0uF^ zKC@k_b8e9=#1c@f_4}M)onNnOF|7~ATcC#;+O5O!WrL!gQ{ z_gFAOI@b|*sF+y0XTKPk0sMW%EfLQMq-m+&KTMXf1Cj%~PMIb^T2C?SD3VC2^brpP zO^Q0{AsUM(I%cd8An7L107Hif_MnV~lQd+mtX~d?JTwM}@BA|HsE-jCvG04u6RiVB zb@e7wy^~@f|C~#ymIReYEur?$Yl=JvoD2Cu_at@vbn#73UC>tdRAstPJH+&_hoH)_1lR=;tXv|h zx$(74zcnIeMO?yc(!tJ@IPjXZ%ETplOfkI_c|fVYkmZShlP;}@PAZqV zCV&%khrb$H`um-5eY_}M4|2A9WLud%vrmZg1Pe8DieQwsIeHh=1jCa96cQ-h(d&vI z4c}f%KQum%M_Lh%g4{c4bMEcs09hvYORX z%n6Ddqf~#$r4qjLhf6=%18Z7VsE%IanL0_w#2QYv4V*kDY;rN2%%}d)!cUu+4df5t z+Hl+X3l7upPx9=F;C;@?zNTaK2IYjCjC0);Prw)&MQY$9ag#(3Yh^=Z%o`8BTJ_d} zcg{yOMqcx10B&(~z_l=N>Gbu}tcMvplvrHzti510qS&o>*Gg)$%u^H&TaafV=}^SE zG(cbnYYz0KVB3_Q*TUaUT1nY#9j`e23(3+skEUIhL#IWq5^utdIyO;hAli(#;kF-z zGp>Z%uft;0aIqQy>~3TQmvJkb>#zLOXo9|d*}02cc)`ku!RHy0xB84?Zc^krWHX_( z>ceaDeV}*H6j|b)Fuf-5st`PD_eSYn;fd+3vh*pBraP62hK4ny+=Rxny&Zpcpist!E#_JkIenczL22_Kv5pMPHo?NbBM7)w^e8)X5RrFyQ zs$PJ;0=$zY8#ZbPT5GE;-Dd%rmLb^oy)#KujBrz~j;kbF*m>g)9H8810>teUlSh#p ztl>EWjSWz}?ot7^ELK|j8nDm48tbP%CCuPrWN4{-98aCZ&6(Um$4dJ{4oKF5x){cZ zfHvO^7_lUsyl4SNeu~3$AXI4y3l^rr%F{MLl9lho?zTUjt*d&`IVZ6D8#Xy@18v}W zf(C`X*%OEVimjUB;UBf@SxE|Vn(sw-;ozwQ3X`8W07Fx>J1 z^n36Qgy&9f^fZic=iM#y8sAv;)5gdK=}xW=_TyV6l?316DoKXVxPH!Q4{LtiR!yWO zi#^iwRzQH&L^^JSmh-(|$C`JvSZExKGMaW!oE?5v*aKwvacq6|B>{;12k8^cd7{}QSsgf{Xo2n;^O)&mEpMCi zUC_HgYLmo>K3SSHOXEIgsfBZgG^#Fo^EwS zBG|FVys0G|(VJTFQ1stukAEx}Z*;*qJW=A=z*RUk$9*y=i2HP?f`czaFkFypdX^2TY3 zR{m}@xdI2?9b4#Us+YC{GjwrOHus_oD#EbVOFJx0+&BQe-)$>J+Dh8o61Kzv}O? zD8SeH{24a??#JdeJ6rL?4Lg2nE+gx(aefQ>wx?(vuR)1q) zE`-F7$H&e$W}YPXW)?t^R>7-axt#HeM%e=tGn)f4>1*;X@h*=>5aVu!`>XgVv+sxZ zIbGuDm@;~aAc1qmqe@cbQ7-(9f7Z9tqkKYBc@y3=PPc$N^`L(m}lVk@jhA1+@L?*>RQ>BJdEl6N)xE~YhLNQ;1EeqRd ztn@{lP`%3?x>LHCgROITwS_*&)AKM3Xejkgusb3Bh$nueGiv2?tkxcC&A*n+`*$Nu zjztB3kDPK~nA|aeNejh55zPfkmE=|eyX49Vr@_7Hk+is>AS}k%GL?J%_tF<3f{d*Q z>aEN*S*J8j(m{U$9tQreL6Iyfn+@{XiO|X26<#*G2oi`7>5a42`S&^XzrN0Y*{lxw z7}F5~HF*KcW)4Q(41b#?YrwE9?1D1DtHGNAIfaNUFs;SGr0#eunLS85b1 zK(w}#v~rR;TRaW~=8OBBwhu7W_PXtbZ1;1V_7Q{3I!p76yPl0URv>$2OZ1iN-bRo; zW{$-X%}cYI!QwTfUSbEufO5|^O0{}&mTQi8o9s;Ria#G)nu^)TP3Xt1d=Ewfg-CY)s6{$vNpE^g2ljDg>?ssyp76#Ms9}! zf|U`W-v01Ws>Fy1;a%acN#aXmM0S`UNJlYSD3V60pzg^Hm`3PiAm2>&IZEeFF9>aM zgIqgV&yD5Pidz-f@lZi$2`Ys;UL{%YwSDq)){~)M|HzL=Vm3zpZq`4!XZkyj%_Y1X zHi%pa31bQ zWVqqkmVx5!n!xgzS@i9RC8Au;F{X2B!kjc$6uC$cC5SqtSLG+fb-_@8#qSiSa@wY( z$TLWvtVnjx*4D;a;5=i$YP`e>7tg(!9dI@{AD`@Jv?ssWS-Xa0u=7wIINw}tVttA! zW*0@aQ>w*+Y`RXi1ZqFEK`Y${RGQEb>0P?X15zc6Wm=(xsn_*5(3C(!BEMY*d=gls zjR_&VtWg~F)=%ko!e$eUmZZq@Lsz=zLomhmIcl3>?7666cf(sLJHGOwXQDBIgY_#K z#osXz{rpTcyNR*c-K?z0IIsWd%$Z@dBJb-Xt4YC2V?~adSdnsyDWk|9N@eJfMmv&2 z>qP29{te=LGeH`zo?q0V;zbKJx<#BsQ@vHZvzCYue~Bm;Flg z6;~sC%D(o^Q)Jb6U`ZGnq})U?8!7nssstuA;O-D#`|AkpM0=wU^!^6aI4#fc3 zOFE@G{nlkMQ zp3MNwrE1Qt5c_V7*g|-S0k95>BfN_hgsI!&$;)2!RxzZ$K$CVuTukSBoU@}^=rgC} z`G9!l<955q3J~Mse>vx0BaNn`Ep&1k*~iW=ap01ab0&u77{wf>$N@@qRH_etKXYzZ z_y>4|OJY|z^p&|bfKfTi>7-La>-e3#0#F>;1Pb2{srz$p&c6L661nT=y7NsA}+%F~eL1LD7+yamONs7|U;#B-PW=hJz9I1-Qq zNi>S<@*3XVpyT{Lr*t55xHW03%4AU! zdpBSCy3u4E{P~5?NbO5wvRX_`Rs+S=x&8Q_1_O)4rJAOP7o56(2k~ znuN4OsPe2941Z>lHn(E(+~e2M`jW{n{I)F8|?h_G{<7 zMa>|_V+Hh{fM6^^_~c!mfAsUUjcHbHu`a)PS^Vc~@}@`~2OZYMg3p1YB7s*TMr|tS zlIAQ4x-aXJC%RWfu=;AA{~;~k!S-vlJd$m%u={O0>=tXouVX{aYggFxvpR4T*g~aC zDR<4ZljMM8FMTgueKa&#I`$1~MkHDXZQH`un$ZrHeLZqCm~h}8XA79XtrvPVNF7xW zesg*&N1Z2VdtSMtHQy!MdmI%LB%;0fMK4w{L~iS)De}K8T1>6sqM)_9S#r?XCLvZ< zGi}{8gU88g)Y0F}YE^s1`>)@9<$a^as(Wx~Jh?p_f0zR=5%DG@s(lpmkRm;lD&D1s zAVo}^^vnQ%_qqsmm+TsJmBB&0OCqO^E)LHV+;P4VfwZWlqC4`9jQRq}nbyE7mB0lW{F#%tH4ey8Y}sZ{^h5=7U4uO4Yu@jbGXCj)!sMgE^C zFx}Y3zX46J9h|Gu_ArgIoF7AwoD4pZ16lLZY?(k#c1WqVmSQ$jB$ZOFBnQ2BDX+U0iZs4;l5Flx zg4>=tNimS>9wRqlx81eUdO9}# ztpocq7IY&A=r<0Fk$?pkIMX%|!>h4@gmHG*Aj66Ywb?D5-8tW$X|x?bJ3VVVDST;c z$4L|0QB5%*R=tl>ot>|rzx2(fsKLKYk+o3xQuft96i3CrT^9vJ%|AK*t%k2&iRyD& z{?`5wk#;+oHWdUymW>?X&g;w7ekw9UBZiF)&shjRH|or#&x?rt#H*t&4@m zR~=7FFBas=v*=wAlvAhq--VZ1^fHns+9|`vGL+rx6a7Z!T7@GFGZfFpCLnHqG2R?e=X_bE%KEW|RMQOZ!- z5kaHC&nvHl`h4(?dE)YEH8=-$p}{Q~;-@4~%qohkKy4h2qH7Svm5J_nTmbnIpxBmdo{E9F(?HUi6afwNP(628hN;RFdAe&StrMh+s+fbY z0jd@3fcEM8vW}3BkZOrNL1Yb9EV>_~z{2w$Jvw7nzrcu<_lqa|k;JpZiUT_&z^FaM zdTgecRIJfZHO|LY_$m}jlGS^4LghBd+%yqj-In2BJgu&Ws0qeD@&xw8#E`jR504=a z+wt(V52w9vu1Wl~x8cCGFBX)CE8o64P~DIjmM&^$N~YhFBn0aOZTxs%k?i!?V`MLIwyZ|Yz`8Qr{MzT-f30u(t9j<sVvJIsm{Af9&Rd;GHZ3V7Q&9=!cu$HTN#q#M)mxzavYq#`hC$xiE=$UEbrm zNpOf>8Cevh4_+6Hul|W516;d3QoPJN*IyD`Sm<-F0S3@jzMlowE0RgIcNb_|ZV0^L zu_xf9GLMWkDy()a#@M#J^6}IyoqXe7b>7(x`^n1Tss|j{vDjv^BTb{24HQYHR2|Yd z>A=de+a<+CHwC3mn#y8dTQ`g4XcRg2fawbE|gh!7g z(YY9MZHeS8CobX)*}(ZihSbEXnKSNj5$k7ZY5f#@0>qSo8aq#fRZ10fwoivoTtvzQ z<_Bxnl&mVyq;?(%2qy z+^rp`YAfJ%A*Y$CCFke%J0*zPWmxQ{Q6Sn3_p4X?E(wZv$)OF!ZZU%8f?Ke!>USy= z-;$r_cDiCo3YMl|;q?g1F71LfGboM(dcjA7^IVLG`cQr5cVxwQ0{M&~LFYhR43E9VdHwavXkuRrUrgICz)IY%2!+bEUo^}VDp_?J*W+CwEX)W zW&`TDKgoOIOCSBl2#w#?|7;rh%z+!(SDF;XKBAZ}DAG--a6vlg9V@LQZ8UV1f^tKa za+6^Hxn&PkZ zA5aX#5!&dDUWKxBpDUpQ9U1ZNH>7(*b)tA~4sfL#*aeHj@fN*HX()tfq6Hm7G*^{G zZ%A`FH35dV>6B{eHS!dB75~teZ#i&7 zzF;d;ri}O8O;*AGce!X3*`YOo1)PfD7qBFDxZj0JaqgTsN0Cp5eUujTeD)BaaZDTuUVNOQ? z8Ve)Y_OxZtNZ5HB?+T)2WKo^C3Ba(gZx&i|gk{e(b>DWyA(5bQ#07D-clVi&O z7|uTl_{zk^<~t`g0U^hA{4H$&(PIXB47%K>{*fs+Il>Q6G2dpsKyr-Z;3sGrgqksaCw+IIRW#RSZW z2BHC7a%?~zOAN3=z!+wtEt`_-uPv+cH}0wr&0kwaG97qVU1JgnDxsJ{itM0N7z0Au z*;42itqNQ`F~iLejN9v9A?%jlh7j!vmu>WJvKrWQ;^%b8bLbcW)HO2wPKFAmc4+Ku zmMjh>s7ANo$I0FsupMc0EMol)n8Uss^8iTQh^0$7cT7q@5 zRg$i$)!zF+chex_e#;%HNbPEhMh7e#qiM)+IUg>Ke$lg#yPz_5N7OZWG<9Fr6d5xi zC!mbdiVNjCCO#mA^0i+nS#5s~&zYC1c)k;Ugs#zK028{mD zx`|}vA2@C_R{x`MOq3B^Zy)&gU8J0y4_YZH_1s)BuwseD&ZyuR>_u5+vWzZWcOSjb319IS_UlB1(HpoUBc{8 z>_Oi#vD~>%b~U&TR>31AL9|J+Iuwsj1)mcxsFM{7m(Gdz?2h;}1PD^$QXy0aA0xZP z3z{NZgpZ~7T+fcQa?k$13hRCa=68K9y>FAk2&c$nKP@Ne4!nJ;GyzB<#efQ3KBemP z=#aJWACh92@^;xnQtDUB>yRf!)C6o+td~X$YQ>eLDq^W8Xf)G{0Z5C%32Fg@s}H{E znI)p?Y(j^i!`|F^*`ZxikMlpZVG*D0t>OjO-#i+o)M~HR&IVPDLxq z<{qX8xw#=alSmRtio7f=KS2FZSS9PE&yH!_mlcji28j_pH6yJzz{hA!V&C_QC))8u zZ}OwvNimRJ$psE0Su3Ym(k@GMRwFM&Ti9ymxB~RgK`gu9Y2j7*z3>Wi6+doB)%45kOcUU`P&5-z96D?sbx@Icv25eJzLkXz?El#kyE z$#EG=L|KKTL33vZ#8?_KV*ziCd3%||+6xPf>Lt!8@@+JdcH7!3lx8k+$-p0QQ7P~(Q2l8?#fw^M2wYc z-0J8RR2Ia3^8BjovW2lY+g9TUtcJxw>~P`texV(sVMB3XWLRKB!IbjjkozI(t;{u9 z9rv6thi-=~GS-*3W2_NsWHF<-$Fq`m(sQg96*lk~SXb--h2e($#dhC!&E>^j5@*AL z7#XMl$)g2V+>?0cgdKDaH->EX-63pvT*7#e|O{N_Ft>FoF(gqYgKh%a{m6%bKHV|WO$)bWZ%1g93{o05Hj za%-@tz|Jr+nOIi#!{kZ(zW7am=K6~s>cE~J)X2zouIGczknQsu+xoO8Gi22)Ke>I( zjx1Y-;mVLYb(`KZC;D{Q?W_grPOYdCvTHfq%i=8{JD|pW>{$NXznocX5b@>BV|c&{ z5#v_O-j@?-^fHP%3S&r)11IDTn0OSW6jMTxLP~{2h&m36XKeA%g|>wmTJQRu(DTd_ zb%bQYeq=SXPpN}8WTbWOcUnAA7upQuj6lVdAnK&+oD(>8&Ub+IGey47dBf|_XAEU( zc_M7Bh?BOs*_fTl8st3XU|2d~Lu9c-j<8}*+SiQ8dGOZP-z7EU$rY0cIz=%+I`9dl z(g`x>w#pv!2LGnW3;fnX&o`C?Vu8q@RA|+J!d^Ttb8fT%g+jN9HV3rI>J^YY7i^St z(pTPYfWk6N&Yz6J!ZM_|sQr&ce|bW3F{l~(p?bdxRrj_TRrj{dzXZ(kNPf{Rd?bD% z8v4@UZ;o^yCQF9f^f>P42lefto1S$P zlSq-($TwX<_sMcN8pUFP4$_R5`L`r-5qLcIYnPP}W&u9i+-e!ZAN*PW4Rg1JP0+-F zH^~;9oG#f)j(&=|l>51Gkh5XzfdbQN4OVP3`|L*PI6VIF^X-HKZ*DEvi4t);zXiH- zk_d1W><#-Os@HW5SAEO<@8RWIGxq~46}zb$_1%{3gm%VFC-d05!)|XZMAJKGwucqX zZk)Y>s+X?yhYijVC=)yaRfg?h*If$eBeV31RM5Ct<6Aw|uI&*f(_}RiM&P-ayb;S6 zmZ31*?svg2^1muEIv2vb!e5ib7tFbU`okd+RUO5^>XAmNQsi4a9tHNvHHt27dq}(N z5PgEQhjhtmW|fHR_!k0?kprRzV8^K8H%7L~t0PdsZA?}Z*02HiutdO!Znq2@x+G*X z&j^uesoy_Lma!{$cHk8!(*!*0DP|o-645ioFlq|W5hEEf?1sgg1DYXWqfu<kKf*TY?9=H85?*i|FceK+>cuS!D69VW51CVH9r>u1P!g}BkO_AT1e#v*#ps1Jb z++H@S35xV6*cn_obef#E)!E z`83pQSZ|>QouQ$BhmK_1WB1>seqp{zc39`v!nU`at_kQS6+$)sqYM2z>7hIBw%4-b zGK{}Ft~hG_!55vY0HXAhW2Nflq1n&?p`|a2fxJM}1+K0-Q@Ic7Lvs8#%}E6U3;$Lo zMSjJj8}7LCdL6Lt4cvpukA|v$5Wxmz?D9Y)DafV=*aI>AZ(ILnbf#gbnXs)R%nmlQ z-rMkt2&1dBAxEYodtR_@QiF-7dyr!GQ>2PgweaIzj-yByKaU&B(QtD(-816D^*mi* z3xAjLF0a6^8)(|Hq##sF$BsrA+os?$!`f)JU6#Ss_Yw7{ ze=Mc+=7`VaOM+r1d?wdHOJGyvak9y)Q~}NFwB8TlfW9>PH^`>4Q&u9@+k^=F`F_L* zu=D&uZE(l`9+*EpY%n6@9F;+KjVC8ee!W!`qo>GTN|pDG%i=#@lXr=mA_sp|ZL+<- z(9tf!n$tSzW5319Bwi_}%ymyhCZrXy+&wXHOMphv;&OTVdM?NycNa+ggeNCF4{|l`P6tl!@A=fVQX)<_K%K37jU#5=Of(#1e!yx{I3#8rUm9 z@OqDDHdIu1(2Eu=vIRJwjiV7l_t}Hm;3_NVj=MQw{51KXKa&)Aj;v=F({tc0G}ONg z2?%FX47330kb4PyTd0f5QZ@mFSAIZ&L>&imq|r0W6(!znbUi>!y_0)Uh{2qJ7*GzK zGcDb{RURXN(2@KWxYzas*9c){FoHIKkEJ0d^J~68(Fm00*Ur9Ak{!4fs>o#OGbsi# ziW*8)3I#GTP$2Um*2mQGYj}qvPKaxG)zIfx=&@CT)WS>0Ykcb@v7ADBt#h{Wq-R6q z%IU34W8|{&_MCM5(nmb;mS)UG=ULek>h)TF!4Hfuk+1UqCOOH@o;dDWN;*tHb(vx= zP~;q?y7gw)H&gzjh2IMF)Ii$fUCWCN#YT--=^=W-=hE{G5SHxZYLu101lTGsm7FAZ zc?Tq0CAZ$(_I6Wb0|&XAG|E0%L*#Kqrx+=NFfiUH`}8%v`&J3&yYVXU?hkqL@!O4& z?UC4z+RKd%jpZGpiwCZ16usZQ`QLZvw?!?&AbM;leyk@SYPn~Q(r}{{U>-3ZpYq4p zoOnKH|Mj-VulVv5;Ck?oY~ibn;#TXu&o4nF{*m?zvD zXRDDdy+5QFYh;Vx_unR^=4K%kCMe z%A9F>-j^K#Hl=B>DQcb%2CMJ*)r9|CZ%!O&K_x8BHz@)`<=x>az{%GI$}sqW}xF8`*uU&l${ym)Wfi4x2|3tybJPLJfYuf2W->@Yvf8e>D84;`xb@i%5V8NpKU z#d{x-Xm((6;2o331T2XZvzj6+!EJ^@%f441%e!Qsi`A&l-Yi-Ds=cOt=vQriY!xKT z9~y(#jeDa0Gsp;u8{fHAL~@5q3OKOiamZxi%P3|KMT#jE>M%jr#l2OrW8(R_sc*)) zXMsM;VTjqP4V5uh#K4%3u87kO5NC>a*Kw*qxaE9cJHHK@N*WaER!&V|v%J?8*`umt zH36vObbjt0S+Y+h?+#R4w98I<+A_^#6+(uw6&6m}3_F$=${)=8Ldj$ zv_CE<`3@X$0ea9O!M1%A1CGaTN(B^gQ`>>-0=v*qhpr9!%Ajfm=`VJJY+aXpHIpa8 zQ~gdwBG9^|D}8_~0|ir&UT%#q2s779KjB;r0Xar^9P3xGNE6!E#3`Zi?z>5e_j30x zF4`KKnGqIvuwXXKPum6~mTkx|If++}eDIss&6@xmHuP_y;D6<;3U05fdW~-mT`OMA zs4G420IK`;cqZG;@3b&k)}zA00Cqaf4i)-8{#U=3(WbP`4Y^D zu^irq#s_yZ9duVnueD6tEN9Ppw3yq9FJH5KN9%59g_d!t8*Rwj`9FV;Y-YCs zcHn)`UK5K`Kry)#2!^RHdtRDlD2Q7Uq!Sr>>9gsjo@iv+!q&}JV>D#%#4{;1mbf6U8R zUJ@f>p&DoB#8lElc1-M?x}f@-$D$UfW2L)XKm}QvO3siRx|5VE)%XGxns<{beiv+l zu^6-8=@xHYaHg__uf9sML0VZ0)tGygkKDJ1mB_aFtLr(9fU{&@!zu8ABifZ0NS+`% z=vqjLa~gETjWQ@RdI*_AYVh^wpcJy}?sfAze+vcjNOJ)K0WR?%)UjcDnHy+S3yy;D zaplbD@N)l3pB$IcIqBGKPA5w*DeHo7c-(;d6N1aVYXTk0({vij;BH`y_H zg6`WddOZ^g0V2&kWaO|9a*GG2N3hRdR`?yaYy^I0nk`vKWMT)xn8Z6~>EvQIuuF?AF< zPN^~g+r~HpGHvv zs03q$eb$L21DX>OIBR?{BVZ8YR!0Zri4#QlZ!4#lt6?AunFm>eYn%?xQqDc-*v5f+ zB$YmRafOQ-RpFz9Hu+u%>~&+ylsslgW7%G08EMs3A8j$uB)(9}+!peM=nr?3HW^l4 zl>$Dr@hM(>gRfA-!4E~+cbANPr`sJd9xMq#}t7?BbJYGko!QB#_v(|b>E zGs#T<`q!E6$n>;5ot|FmXHU`#E-0vM3M!y13KSrVENWEtxEm1^m53V%qA@6fTTtPD zPF0~)$-{b3MOU1E{6t>eckeBH_xsL0_blJDjcsP`jAf}*eu26?#mK0-b0g#P3yIS$ zgWnu>6XS@=dl<4?wTvzIGB;{s+WaKVhui2_I|~qgKrz!!cr85N&isIKCt%Tq z0F%}dmDs(L6gaT&@VJ?VQbDm`+TTNkZJoAL_&|tS?k2%9Ns%f}R1mD|RwgUbXoyxy zn^;Ph(%|k%lNGq@t^^r-Lj*!L1C9C_`sv{=kyKPb!L$d$okF0nIZLj4BxvfLkIl`2 zQVm#>{m7zN;nrzer)6qx%+}TV-J(B=Wk!9}8U=pL!1VkFzKAjJ&kH}}WZzo%@0KU- z=k*iEwJj2ktfB<(FFo#v(&l2qP^$X$)SdL*koZYQNw@qko&B%)6Aa8ir2NF^g5aEB z7gUz+k=J;&Lg3sO1Tl;hI_3*)`%m1^**NY0s=s9tGvj;y;|p?d0=a7THlLx`QxrK# zg<*;wiu6oP3&|CL(iaYs$S;L#Pvu zduT1aqTK#f&xJ-dy7ZvJNz0sbf_4f3Z@7b*;uO#sZD?&!V%UOyM!ON zwHksW?Z^+hf0+8=>z08H91#IvJqto=tyThdGoaOp55QOmc@o z$&u&1QrV=vpt(LL&b>#q(+@{M@oVKi-@760svhBS`>{=qa8;g4kC8t07mee!DvqlU zbUWw1|G$e4Gf&dxyxSA4YQ64VVko_h!M$LIfqKy>a++BCOPayh;(_nM!|Z}; zyP7nZy>EZ#6j|%Q2PKGb8WgylLb01Ec&=fX)lsTy6?aN|LUhfFi_%WWdPL%|D*rm? z4%LIuLg_%6)G=Fh9_ww8VRm0=(--iPXXEG0*}m7(8Ox2xGEGd4-VdaN+`C7BKcIkXsdmq6((38k!8t&m0W=&yFDF{9xdR2S zY%X>9L(B|sPe0fYU52UV&Td-vze_ts(+`H!_Bmyc^&-q4LkiPF&h3iQV63@p^Q9kt z$*c#)@blREu5jfiJzsTyPnIUvr`4@N*v!}n<&`pB@P=vIc`VHM8L6B= zfbwY2X`zl{!OwDp3d6Xw3Kxtw!_c2n&sHf$vS)$teIkkx8RE}Yd6&oz1MY+@S;Ak*82P$cuH*2Z%~$2cpLngC<62<< z<$7N8n7?@)?(oTjnh|>x2d1W~_dz;Pv|FSE(0B?kj5gXP!I^o96QSYA_q)Ya9_e@) zK4<&8=ih{kMKG=48-w+YaDoA}-=RB|Na|ni-wyG1x)o#_B!U=9SX>uiTplp7kP~n~ zw;uFk1@&hy!i!;!J@`TWZ@(Ou>TE*P;jdNyjjWtNwwWzl0>#EtBnEQ0eA*QTr)-|= zHpG@iPp((?3em-rMnAH1&d2lra;u;I^F!_0!td!jET7XyKd5nXF~Qpy5oLf^dM&vbcysbLMUxhFWV~t>cTG775iUn1eX4k$ogQ1KbxPwQ z2;qRRi#{p50@UjxnXoxx$WA&nINlWU5}P(FR!eRg;vkH4%Q(oW=B!jU(2>+?_eXR# z+o`M~=|R^8d6G!VsQ#DI6#=CoaUz_@K+l}1x#^C@PB6wwPp4||YpVS_UTaf>V^tLS zRw>=6jC`wFff^L{84kO)fZ&k0w5D+wym}P86aBggXg|AgDuwJDN;WyL{lCc!PRA$~ zc*PD-VVDw-C+T&?{~=q%JqN7+tzu&$0GMfJbVDq693;kP1HERcM-}O$F9!5G^*imJ zR|ouJXvFDv8W?|3`beNNFnqKDTTr{WLxpEgoU@Ion!6|#4BySq8X%bW8Scx6xyLQFG+#dXizO$W!-n;8@Xr|`k~ljL2wZfD zOtRpiOC=E6HU~TgHlV6$7Xzw6Ow_CHh=5S4jmaR7<@llB=^v%ky6_!WM3v3=zXMe*MnB@0FSa zQB=y6UyWY!0tu*R5wi5|fJ!q_74W5!--J!VX* zhRskh{6yTC(N28wH_9Y8p13`!BYjRoY(EZsUV@55gJekz#jd4DBo$WY-YNu6!Zzlj zv@jTD2!GeZ@Q>8GyVd~0x|wX-iyILC`MdvewiFl+tE6%WhjRa9MJGfcV6cB8Q|xik zrQZoFA|lOTji6QBOJ^>q(R4tM7oAIO`P8k!g4qegmR~;1T-;a~LK2J*mo&yrGzpCB z|9SDRB{O*VgcP7sIWBpQ+2C&vtOlOWq`qAF~4M{E&^CNQq(0gyOkT< zb=%mhstn=S$q~MI8D@!XdW#z`*Sk}`H`Rm|Th+4|BPQudpnrhsDeD56K=(~DBf zGK9MHnC}!mvl3!(k#hk|yd)g>H15euLvS1FfRd}a7Um%tkUqvTzkN<|&dqRs#@qZ- zllUwtUnH4HA>#)=UuSjS*MIS!>qAU<`N^Ig>7?|f!ApY~UOu5%2&=0AGcLHL%e=7G zGZt;AaKT!4r6J8=^hleD(G-ZJHs+cze17=+-80Ij_qyw_&|-r3Cz>cGbH?7_EZEcx zQT~a{Jr_6>Py2Z0b*O5kLnnCg!fOzm2T+WZq|ue)Gt=Wl$N;m;?;-GTpk-D!sw0#Q zP&5RvMh~yn#5GebsR$ibW8vUPTu&ZMiG5|o%y7wrDLFL$9s5c*dE$B4bls#iwp9wx zD!N5E^fu23Ph*d6;I%B(#=bJDrG4QWJBqb`;oPI?z`Yj@p}SB{Cq4QVCV3;hBmIEH zy)^Qs&`jQ>Q|wlXBvWCDLG5(AFEaDDF$Y6Zz@CyQtWX=$au3s}=3uym@wV|+9%R$o z7fQzE9uc5MgA^;re&7}#{U0WXe9)ZZNv^&$5Lq^Uu;SZ8v0W4tPz(bTzH0-73f4Ps zm){CBFuALB$l+Ax9OIv=jt;63Bn!%%BRtQNV&GHOVciJh$2(wg$()g^u5g77OxCQp zIR|zoNOBF16X7(FW)O-D%x#eB4Rp*cL_Z^uX;5M~Vj{B@4rPX^8r5;2WcRwp2#{3B z$k}|$&7@>{LGUCX5JmoST*nZkSNHQ-A(jmIbUFb_?SuNrmZYU5hcq z;iT}cuw0Cl%*R5KJojQ-v&hd48iNL2d)4?iv-;oty-5(IKR7&r+W!!)|LeZpl)R+o$4JOgiZtyL={uBgL;{PJj!drO8Tm) ztF9Gj6W0j~rI`z|U}T!CQVK*B(QL0<;UtWrUN`?LFn#M9)lk~T{U(F6XK7kVFObP( z1@+QfX0}2&dm;m>H+|7%mzSxN;c__G5F>ELgWs$MHVd(?ke16LbJ z8s1a*wtn$ji3xzeFP->jvhFKp=z$aZAVbdP~i5un|mOoreO2pb`|#pCPR z?-UPd1nu;bz)moJbirHNCV+y;!8KmB{yNMqxTS>H0^c-)eYV7-oyI}6(=SR5q%wNC z@706TbbCXMWH9aYNzYE@9w_GUVV@HosP(2YG#~XjRn3hNA%8TKEb0hO)pXLgrH>Vl zfmUYSJM|0st}fq2;{D1E4B)xC+>12CA;@uEIUc{t?zSX~|4O+kISSYmkP<<%B0=_0 ze$DkF)918Zj(_hETpVkUHo(eC2=fHs=(}@6_(v!AkNdT08}k3nQenQa7a{AV0+TscyakjlsV!{btuM z{_yNA8|kk<#5f*Ln)78HV%kwR@q1w%Huve_S;eKVdTidBxux1SO-6B=pYtJlfqh52XCg1^BtLC#+CtiRZPX%f{EGcFf{Ar?F@&~IB9fy@YZQ#w-h`U!6-Fs^^{*Q55GllSg8^Cw$~SywF6!f zQYwzB0+?BZL0Ra0z|!9J~`wu9_6lLE5x z_fyu9GzUHsRhVf;MHHJ)kz6Xwm(5&k3(KNY1I2phKxrdKK(MM<(3&Gh`j1c!2swK-!wB?o)m2+^o1Q zUCY#w2>*kA^@?jQx^kan(Ej0zF(-6?cYOce8p zunoM2MJh;1L#Aq+xm&4+VmXZg8|Oo~R<`pGtL>X6XOY~ zSGZP_Fs(_@3HD+hy19*DH4Neg-?Cl|@q<;!tv&y;$ONl7OHa3wiV39IZ1ql1>@kWQ zrozz2+^pC+Pj?ssl?qj--0Q`_7GtbKwOnyWpsVzQx-o{-RNXET%_7NV8r@1ZvDtK* zZ<cFQKtD*awG^qM!i;6y2Fh!pw-Uo_n$*P}1(QlO$t0Xf(u7Nn$R2|S5aEkH zLc9y?d5~j^CAhlior0?#?ZE?mTjt-BV->Ab?|NX;S}xIp8zW284vA77bm5}?P~ZSP zBYOIf^P0Il@%$Xh>;~}XzKLH0FS#`C=f@xYU&|PEhYhFTh(|AQNzfoKTCUHw$!*M9 z@08b&Yvj@EeNKtfI(_?`w!L3nGz!+We!o>I$KVGAPj{7D`8~NX@?qbueJ$Wem8T5Ng?_0 zqqcIA!!HQSfk}%FnW--yQ!GTLmr!BZ!YzJ>NRjYn;Ac`yk2|71&L%SPBKV{_6|hni z&2Dkt;(kfq<*`{0Pj=p8ir% z2?w@T@9Q&dx@qx&&fqiAP}jDE(cf49RcKk7?YP-9e_8>p3i6jbXLZpvf)c@Qm-=Z* zn!QvCwU;WJb;_r}y-^$Sy`yiYP)j#`mP1>_g7iAHmY|Xd1rRHN`RNJN zZyvL!`}A2@`Iw*fAFB)Q$2>$}h~p}I-p&!>^G(`{&x*A>$j1(BJ~?Y99*$7#L5kE; zVLM?j{QFg=Nn1@vNDc^3O3G9xJx@wn#oeMv!EyS=?8V!K2jCA7C3+qS?Q_DbHIw_L zM?&+1j;TuhHY>s84zdQASuuz!mQ{htP(HOYt)gyin z17rMgaJ}-0DB1On2tSYf-LMklL#ZFuw2kbZkk`Rvc6^uGI z2*?_*Hs-qWkTV>No&=5@2P0o7t4Li(JpG6VF8a~`?D8;4m&RZ0UP6-iorN425_`;K zQ7**-0q}MztPPl`dR=$=MM}WN5<3IA^UrxUPtolRjqo(^R%0u&ULtqIWgo~nJ1N{Z z_wMXW&2G=jKB*drmPwP{bnhnR8av@=^ulJua2R%AUGnXFKQv(_@^|I;$cdqHM;usr zy=jJsixhjFBIl^EYI+Un@~9VIlb-Z!6}JXH@##|~GWY2|23wl6Idmpm_X?l*7>S4U z^nKE##mtf$v&&R{%qcjk)r20VFF|3P3#2l@*q$E}6$(lGvNBbwFo&)e<58{G^|2>@ zHqOv09+W(2X8N z*6dt^`V5PmXE(XGtV7XvziwH6iKA|jG2s+FyU4we>Ye0isD{-l{$$aEMOn1bb)ugE zD)NQ^OzEm3hS~v@0arbCkvwo1&8oxji0ylS7+WeX)))2TKb4I^3!-7Sik8u15q2rwov2}6^54Rzg%UOQ}IL^_;{hDg z7q)uT2yO<}OCKn(wN`vZdNZ&O=8chEl;?+}s@nub^EWFV0-1e}aPdAmSCH*o4%tU| zmEhe6DROvqNs$M|!%Bj>APeH6w9jdksMvH>?;In*w#7jAxv*4|^4g_^d*BCXY2c*4N&fA=h_Ey$cCboR z1?kxsyH`8!o;-(o^zOztmmXTE+aa%qq<%b{;ao9-d#XBltoN6m)!<~(1JOu2%!9lE zr=XHhztbkcWqJO>tBlV0OXdpl7sd*b{9~cc*`eT-f+Tg;@WqDyfw&pA_9;JMw(M8A zb>B6i?4#U$H^>ow!pwoSpKE5AY@*mR6gdUSSE2|}gL|9^BpAd>b-oQO^@jMu_NfmE zx;8OYJQY|TYniK)uoCWl-{@Dd8tzW71|>4dqz9!3WoY(mFw8SquvXD2K28dP(|j6~ zP?rV5OEc&!;c}8K>Ugle8FwZ2KNU_Ls`Vl!i^yn z!Ms*x*vG@?rWH%fXF8kpEDbl}mW-e*e7R()JX%QTw3#K({3tt^* zS?{vU*gF3k;g(yJWkOMUE}9sG(1#v5?m0R0;;6XNE*Fj{@oCauBjFw$J*LgZ}~QA zAN}*)ub=yV!}~WEoza4_?OZ0c3JTUN%aDCWfuSMivsL5IKTldR&m@do?;S}bd-y3n z4vZAga|SUS9j4d=6se)YF1)hrm88k(z=hSob}{D{Bxy<(W($jCjoQ_3+?-h?JELvV z-V_~j|4sj!8{bU+cFUV3-#YNk%Wr)04zNYU2p%f7v3*XzN&C^pZ>@dfi#IcVdHw6q z{+l>z&rkDz-2c{^H+TN{_(E*i@kZL$)82~u_xhi-y!FL<=NGmw`b~w)Z+d^3_EzdU z+kUYA?dtE=y!D%rU~UvhdMaIBcrgNaIaP_jWf>;Q5ntJ?*!LPHx9ySXu6mToJH$v} zm<}9?Kt!acw@iwjHIfH2g5TLVgwGQl@t2HXAh&$~N?DfMu5NqXGVtPg<{}3^L2(3Q zT%pq;7qgu<=Ct5B2${5yr#wNHt6uKgK6RaYq<`M4pFkX9B6HFc&*BCZ8jeIzg$0Yh z<;j{Miw=iuR@@8>mvl;FfJ$#*;(n*2q}ZcDSs#$9K@!k`c3p)_xf&P-RF&SNic{i) z`xi=#=Q@l4@TMW$xL@}AmGz8ix7rl%{~6iHZ?|$_FMFBUZk0{3PzEX;e5$}FiwERH zX2qNxbTM7+4^|@Jccf#ztERL;4r9VrtkUUp!tho^s8>CFwIpww*`56z2s-~TnBx&}{!szJ3biM1tz*KdS=>D8C)eSgmVW0EjB&hQ{<(KPT zA{dxCjXuva3~V$|c}}ch>y9{e!B`OB3*VtV0LdEz4Gklwun8Zz~SDlJ2BGZM5xQalnDG$`@QasCKV5BN18Ie4PZVVQ+_A#YsHD>MGt--Nu9 z&f?`nKY>83@Sub59*PCsqL>Q1|62Q0tfZ0)WU|@v7%-2-GKtJN(in6aDxiGsS|#Wa zc4*oJ7hO=~UzF|(&7#x6SD8lVha3z!GY{`1GOH(SoKhXUaay4?5(wLn23ikgRklJg z(lc7yiVrJA!|+>y@vco}2CoHjT>Hq|vrqidGOU9mdZIbFW&Qv;UmBh3b5GU7926Y! zJ3Rm6`L=v zQ%vw_Ul4eiq&P5q04;;yvx8!DCT}MvKLK zw))5mE#p=thrg*Yq2*ZE-!_mE2S!W18Cq&7wu%B~+pu*mF@nv4y?$7^2WsrxbWd^b zR4(m+Y8z>EqZ$<3dV1e$X><<|`d<@nqf3FRDNR-Zg*0^WWVvKta3{UAjfn{;n%X9S z63x;|@1^^lw+p*Gnmq4@oN`_cs#&Qy@hIt)+7z#C&JOelXSu7{OxRz$qI%g=st#eC6501)#D?DUBrKyD^|`|a^2fUI0S z^C0PWB2WNs&^p9Y>^h1>QDL9J@eoRJ%+(obKDN7!)swb(P}V$niy>q3k=KKFTx)xO z(87-_PtDJ}q#XC`#&yJp;Nr;;FxgWAa-Yee;gN<0Z20kD=Nhl3lDn- z6&6liWm*?vB1$)%LET)qdQtq_&EH7&Ef-hFa@^{L)tV-4{GrpP8LtOF9AAVQoZcpHD-QXVH6)AV%G^bSqB0BNXqtG2QD z6H9WB6-4;qW9%UE7q?>xU-dYbYml$IXV$Wi9^qkmvTwy36|&pnjjr{IgR-$&-DmHu zmFxU$z!+We(}EuO-#hw>Nost5$NRq~Tlh&02Ua%rnMsX&iiKKJ*;H7G2dWsmNE;B+ zpOsdtjym5{>FGKWC&FSHP1=M0Sm&{qNoF%3bfT5k15YX5#lE%T3t-a6eibe$lj22o zL!Fi2vQVj?eY6+WzWC0}cuRs{hYgnFp!|I(J>=CYUh#DlB{yf5sUFHP&@y?_6{gkY zVMv=Q*Xx?=VftgQ4!TB^<91tJCM=wETivAnRMxIa7qyCaeEYI|wFt*Ca`=Hi80Dy3_jY=KERX?%BO7TNpKC@1I4l#`P?t?>rrx zLzTVE-KnXXb?iY!=5*aD>2@K;N0$gL2|5F76csLqf}2#j4A&lDEzA|vPQMCd8J)^3 z2>eJ@Co)H!+eb;UXU~T}?9+Y0OoW7mLG)CQOppfaLu>colsWk{L1NUY>WFXZm z3Z-_Aofr*#Ux2_--R1?~@xKZ?^u#g@n~y`;an&`#5lwEa$h0x2L|y)jEHN#`B*^us|Tf79J~NKX&a;VFH!s8U+XGmUh_&(Cb2HH$ok z-k7Jq8^67y=**8gEJI434`Xv+QNR&6azT0lYW7u4O!d%VhF~I7B&-IVH_Nka&fef1 zeyD!vP~`?DISJJH%aiPxlnwFL_7@jc!RGmc$`UX77$*J$=|8*uQrKbT( z{dAb!FSwuf5RoH}`}SSE-1=)v>L-WYP&mk(!X;b$A_;PMr-Ri43g`vK5gi7Hpztsr zQy?b;KNR~uAZ1=lTEMjNsd~3(LGTjef9SGFntmPZB3YVr|0jX>;P4Uy8PJym9jcW+ zSHzK2cW4KwCIAeBjv8?Znohr3k}Ee@^tzm@{FAlu6I1O~wr04cu?`c%&0wE< zyf88Df?MD(EUl|wDZQ1$6da{I8!$jM@rkfq5k6(#YnVLx2ulbXOWijs+Eqyq1K$u( zG1Wi@*6W(c^r@Z%0u$cdIk%;2nN&@ms^1Btm9Pa>E*;Jr;du%0cZgh)L+&Yr6?7>Q zNQKK2X1UuAx_;EEuXVVxdbfGjag=c5#?`SIf#KdJ#WuhH-Q6UepSQ`sKX+w7J+Nr&eKz=h3WgG$@nbxb0}UrVkubi8!ETwP${&@V z@Q}M9d9toKx{kmYL3U6yoBwM2)E?DN7rU5fZ3CuJB4>!Aqu1!&|tdwSGa_GC$Gc=WcHR@C~W}%M-Q21iRu-n)L73PyV|LM{v(@Ypy z`THsBNSXsX^eW6&y@+BV^E{Udivr@4<&Zm6WXK)b;hEsRmoAf+$*;{V_Kkz03=tD* zL~Gr)Ov={az58S%lzd%FZU(NJ(4>9Vf`w^{J+>)gT@KhOs&PaZc&2vXmHFb?=o7a> zGKB4i`nLLwe-WAxWwM>s(HIt#`w6S9!91 zz8Kw-&8}sWX6*86h4|@eLl0m`79{jsc?> z@?}-JRl1!}=L>b%bjvr^d$-Gd=_8?`DZ(?6!G=@x>Q``-u}(~>U%j%L-V8ZDSb8iR zOe|MD9@!-Nnvd{!x{JK%VlZq+c+HLW)?=dYPc{L_=eHG$$z}&W;$YI)DnkNY`Um?PJ?^xf~|hnm91j^E8AxFse0Y7D_6rB*J&QhD_m|(J~ydR{#f2C z1fF!L1+#Y12Nf=b3lqL>Hzfm>Y*erq;gVZ3j2jk19d7 zb&qH>BJ6&OeN2%rq0B2V?pOQgPDc}b7s>U(>zgxEf(+5v?W$6`1DLH_Aqu+_>e=@& zP>cZ#^Y=;0{8g?Pwbt*15#bl53Etb?5}-mo#5%PAvGoB_gA&8Gs})et!x@b%X>^LD zS9r^JYiM>*)wGLHpsG*xu`?zmKvd&M5L)4}L>f#~T}4Gu?2vl7`3NsM5nvfl{_wBtfk!3W+2bKzGu zC~N)F1z$lz>|Hk&f6tO^^m#v$1DobJ2uPzsS4et7(j@~w8?_YyA5Z((D<$Zf77IBf zGRDsdu5GGAUUBm>7eulJ?n$J=HEv#;ssT)O7_VIhtFn|2vhR2>SzLL~YEMp7(%1|L*iBq6(KvC=!!9z1ineAYK~l_)A`>v1ZG% zQJ=tL$6gzaH_?H2ISy~)5t0&wn**x6+l7$HQ3~4(Ha-$G5h$B*(@jrzIe*U7tKvhN z*fyp|m@#eR{Df(JPN+#^36We)@}xB1W+q0k-R%w(e`ptX*p3bR*}C$A45#k=Y(86E zojd>gN#8fAKW}`}^Kays18d6H%{1ln6bl)JjZ|2(Vx{X1v9Z!dJDpDr{5b*|4yI{f z-H!%wDcIif;mlP9QTh03i@H~MihjJXG^A0T3dA}Pg4+d3L_RPu<;jtxcFo(m17u$? z&Ts(AjpM9MYK&w+ooV1ZO2BoEv~M*F)p&UTmfkm3st0(sIXRXgnt&j z&ZU#qoegML0DhpZJXC`z@jw_XZC0#x%adUV@`7O9ZdJmxo9^cqq^j+o^f6L|Si90A z?Z*ot_K}a|C$t@wMfBL2!u|hk5*U$DG4n_#KY`)E z9)-1LR=y_`+e?uLnC;Z2#!Ax|sGvjFf@v2i5?%x^gC45-CO~!F9wtT7PCp2(o02Zc zp|jkt1a*Vm?t%d7-Ie&pL6if&Ajc*dw3})70`G_nfxcCulb#)cRnkH!8t3|*u%e!^ zp4u(QE;#@t>6#VKhC@tV#Z>(EJjvE+{Z8SM)spMVn@qoxURET;fXy=J=ox3>yAmPk zzM9^n$PdDh;-u;Qfe=I75tt^!Z|V&EOrWdwUmb+r^mauo1>@lWSLOfEV3mZLp!1tq zhqdwPHvK%=(!>2#%M!qUH038OSA5pwYPn6l zBvB6>yKSb%pu(8ya2|4QSCj^Kj3ry`z{)u)fP&afS(Is;dgAt^j`TSZIBX752xBOA zEkz=!Frx(584J2#rQ3hNdkBxyiYGbn(8u&AuUFA=clWnHT$%V&6Cfm#4L>C3hVnr= zu;u=q8G1V?7K(wlP+^Z>>n7(H)M>JXtL8j@EyeXSZ8z!nA82XR#{HyaA$o#Je0TW_ z^hLiA+GMeRg$wpa!`?b}U{sT~hZ(rOBHukPQqcPJJ5tr%IMB`gnDFgSQ41wshM~>UIxw{rEjeS2`*b_4B`)0*! z@tmd2Rt>|wXWXyF-<_9yY8hF>5v-Ep-AyM?sw4MQk6b^W^ThLt^ozhIZKwJ(@Wb8? zZqjabu2-%BeXe}Y1;J^r2H{psou(E{=uO%+%q|UjmAAq&{PikR2tnF00b86Oy5`CH z;5IasLlY8+ZJcsi8ZYP*9SN->g|^uu?6D@pJ!@<4aj(e`mQ%-l)nkLMCs>+v`KU9F zD})J$0T@%N6PXls$xQGOG>DBy98-QZoSeWnkS6PQ+U^k^lrp6vxYj@1C6XFeYFLYy z7xWtu&|fg8Ep9K182=qht`0s3abS}!2jvKc^9^{PbZvrK|7OKCd9+rCo+FI%gQNK& z=T6&*H>*~NCp4^@(-xgRfBWY?Q%!n~COWE$Y<+3$GZ5f8NPjP&SYXc8qyC;J0pgjB zQ=0q=A#~xSe}yZOJYybp=8SWm*mfoK%si~wkSBpmMsQf;z50MS2zMCh5g(H4vl7bI z+a=8v*Oez;hEQQ_fsW_r2>tfor&;2~VeKg#2&k6idhDf+OP7ldc~whV$iQ_Y=)I`> zCi&NkZ%#hswN41-`Lf*1{(F~V?@;6x)QIkoaVtbcL+II=ejPioGRVhHlggZRu!Iyy9EK4evA zX)wXM!X-TxSjm&l|G&XQ6HmB41?PQK_;O)tJnUlMf7KNBq0 zAtiGgbDO>wsB47E^@+^lZaLnCd(F%d`iWI!8s>ABSA?z&n^SPvw8^9xGHD~ zfWA8b{?NA-3Qo5xdW6xe?%4!aJr+Bx**8_alJPUaZbsVQ<4N`ea=>gq*iEse6xm6I zU1qxJM{i!1$BDW{k%HrN6%g|rlwFaRyMGyYW7Y$*&AU^&oNXbySU44X9H%ikpp|rs zI#he->1q8dP1^06R`QT+Cs>{XS*+uIc6#&(55eNzaqYFWd>FM@c)C^CL>fGugr8V& zSbxvI;X<4Sc9qL=$Qs#}BBl{_O@WL8on++JF5An^y zRo@)g!rs5|1v%irAychpLavcw8z@o-X->>xg6!@+Of+lMb+H~GN)QRKG~hyB!u(X`Xo2D{DMQa0s&o zC{Lx_C@%Qoig1~>H<|N7%qd9^N@S9up2Tj?Ubi?Ac&lc#F z7}@I9AUsZD?NT<+COF~?5-a+1qv1 z084Hf^9A7Kl;IlYQ~{4EY1F>10XP z33aHBToZ=R&y$siOT=v$l~oylnTffQyHndBud)a77!S{?52_;d&PyMXXtvg`lSaTn zDr4w5i7r>t3i=NWTkhLFHBVL#cH?Wp4mgoAiF`*-25cLA!q!wmQ;xz11Y7w47Bj@c&*8EZ=x-p{?K#qS`)|hY_Hk; zZ?DTtV95|~Y#}iYyg304-XH}wm10vUvY862cis+J{E$De-K|(~)#LmEa6B5r??8^E zYO*vt1sf(_bU986f-#9884~I32-v4^84+@yj<(+gFXWD&u;vY!rL~U3%!Q=-sp|cj zf?(w5xaq#uE%hH#%GD~FVZZ0gJu|;2@302kO^a_X3^2)zsKoA_q`-k?#&I*5Q9-d# zEPM|Y7A~m~NBxvzJR zX1nDvGOW~jPgSO_cSYAyna75Jlft{;#Y8h)nrugCEB&P+V*UY5sYflT5oLiJ3CG$< z-i%13Jhv$hZr}zhpF6yOP1WgtQSzDzKYM5Vc{SPb(wJC|nxSn!#e(9siwgVjv`409 zx94qgN~-(lw1-}jAUq&02>xg%oVAMu&1|LyGe2~w^J113QoZ%Mma3{<IhP38!qjih!pYL#WVtdzMC)_mw^?eX`uLU^CUliLk)$o=M4y zCiMy6F-l}sDw|ZN49r4ZwBa0XbRM9UrqZtzLiB6_1@o1#5irarwZSKDA@qxd8>%d; zxxOU9c^tK0tHp-_9QWy68l;T6r9{URkmf;7e|ivxJ5&Nq1GZrBLYiz}FoZ57GN?7% z127hAJ5u~uOk?wpRsZtspIC-)I;=&7BPesbFcDZ^N~H!7k~HHwt#>W+>32HM_AuE& zM)@)ZD?DgvcrEszarVCC78XMo+s6Im&Fg`dg?>3I4c{ zvb``K{yACw^oO6u=688LeaCgHM9+V?$;{Jp}ZL3xT! z7j;09Ty(3Zs*&LUf?Zi19tN8gJ7*YbjO9tXoMQrvJ@Br5pSbl8c(Ad-aBYw0`gBfS z8;j$9DcY@x!IO^TQh z>)B54rFPG(S1os2HKBV>llF%4^6PEPjahAihzWay`~6l;sBnpx@X!^nnzUbtTWvy~ z{(^_bXDyxykQYp9gWLSp!eI^cum9KA9$S_!;;8wv!;qnc1=%ra3(Np$dxWf;;3ie3 z=EiLM?jp~vnDw}LZm3-wc;Uivx0R#u^YvaP;qin$9z|058QvY&jZMnUWMCo1?x08x z6^1UI<)R*8C%s~33)8N`z@>Vn-lbl=o}8a~h|UvMz;&zmDg-`+Lus{rkfL8B$`C#z zkaHr;4>|gSO8SC-H>npJN7?Cpupal%q@nqhec*q(qI#Lg1eh5~KRQBI@dJ$G9^E1O zW$>oCkzzMcB#sIz*L0KBqH4NQie&>Ja$l`m=TafU3=WJPSnFmd1aPb$4`6W2V4vpR zIq&`}V*mAHMt;C>Tp0;rv4eq;O|h_Zrc+^PDKC)M%{pX=tKTxIOlmBbr$ZZG zB7-csh9EYyzQ%eNPU#NaqtqRN4-Mk`v*U%Apf0pMp~C|x=7Ix1*#{$rSr2_w-Jmc5 zMS4g20g2-W6bF_dg=RoWr`W9&Nv6U|d^Sc(Fzbkr98Pl>>|06(3kuLwnw#>%c(ZDBWgQ z_ZV4>Al4iAk&~VWNHbG6C5_%Av!^5&@~K<9_CpS~Vt<-oZKQIZ7kKc1slA8`3 zWE5qVr~ZgyA5!EV@?fAJ-k7#`j((tov~JasGS?(re&%>uGdAfxYn5)CkW;PfYI7J#@uW9!`}fd*YM1;MCMk z5K0i+k4g@*lcuU;Aiq5^s5Aus36LgV8IZ^f%!(LwN=6iRWTtCO-IZULhaP7572 zU7Z_gLpl7a9v+ifX{fybl(6BFs>%3{cY<^`5au^&`sQqR%auQ(4WzQoicPN`fi#8~ z1Kn%AGjhP2?;1e|* zpa=OSwo+^|MK)1k5fjqrP9_aX{XSj<7Df8L>EwKz+?d^@#lJU9JWQ9;z4UF#xiMpZ z;tg{f`yHQ&QE( z7i3J`r9Gm$ACg9Ys$S--r@Q6rJe#zb>X|<6kn5eVH9?{{(!+5|3cjNKb zj`?aM#m|8Juq9>&&&7?O>Ai2Aw#@ul4#IBF_i**+X>#NePY0(-G#;A2W-5y zZuy@dXYUs7wj1yNL0pT>$Drj#`ExJ%5IAU=s&N%v7xR zm!8!$a0?_&zYZ(D+NDM?z@@g|uYiozvYQ^X^{fAEWya!)TaX<5(e!eONeTUZ>BK*i zb^MH`4y>L6|M?*GbPL5M;SoJ-lVF)7LX_a0Crefwm9z&}Ow*yJdJ}ZgJeMV?sWz!| zeJ>8HWN<|9=Y8|C9Jq`H)$|(D)SAtyag1>7Lj^d{L2a&2y((;B!wAS_c|n|n^0EsXOo66O1})7R`?c5WmqZ@_?^ zm!Df?455rtckV<6u_iI%`sn_7vdMux#JkJ{gPvj`{wxE$DUjEZCre?vrx(4S1nkzm z?nc`3EbtK{Dqc>SU1G4{T?1N-~W2=34n-L?$u0ch7-G{MH9vrsRQlD2u z9C%yd&_yv5sA?|Q#*o?wYqEpk!)W$^F)UH?DqUC!< zxBlN(eY?qV+dHq!XBo9LYtbcnhYCx$HEDC)VxV%tuJ2x27$a!X;`M#$V|nZMmi|wS z;Liiy8nro#kn_!+_~4C+XZnU+#RxYhhT3j6EIR*R{-*7wq_cQA(eu+<9oScM&}?(s zL$MG8TTF%3vBgX^G5+1C%^=0Hl%Q+cynns-uNS{w_tTBX@#pRML3)lmz$pOz+G0tmS)Qt*fgE41g z)P23^)9t|~{IQB7NjJ?V%x@qRoI*G0~%ucor>mHh;9cz zb_Ld(4eqB&wMaihx196}J5+Jb4U-PYM~x60g2D_Uqus4|37_!?b#A$qX@3rTCgR9P z>|($Rqka;2FR%%SgIl0PLE^*$dCa76Ng5O~=ngIS&xM@lBkCGOsey>PL3vJ834w*9 zhQN!LYGfdJ@w}t$!3!kgb|kkeEwhIC7-JpS633BaY)rv0IC(JcC_N}fmM%zDS4ywa zZOjq*L6LovE}mOnJ~(=AupJw2`7||IQX9P_$lxGCS{(#&(Lj@sO&iz|3=ku+b% z%Ya=2Geq0y#<^SlZpBRW}#DUE!9IWUM zgpm>~EPymBsp|6!9?Nx~yPlcX5ey-$7^_w_32I_M@d6!YaU@Q}zpq1ar%cH7FlKh7 zxGX7mNmb<8uO>2ou$U>Y@WDQ4nOgtd?^&k%Ijo+_k-*nM!(n8AEzhufBr@piHD>Sh zI~nb;Sc0NRm@X*u>R}QmMp9$ovl-!{+h?T*&B{+p{Z87O6D@Pf9Tp-SS>pXrzNC#= zPO#_+l)*7xl!hcSmwZyx;q%8>>*TX_3|cn(U-ARSJF>oivWz!--b(MlhZK$&v_{Vg z&l>unRCm*}U6HBT=X+7!G?tp@({;8=emw0xb{E_*nSADF5&yfG!h4u@D2an4ZwFO+ zdYz~nm?*B&Ic}AHy3eG=f>yfD`TT-@r%R!6&gK551I}{aHs+%{t{DVhsQYyK+lo=%Y*lRq8na1l@Zq2S z?fu&Cnn3eW?!Fu32tUv`@Ok)}8EBd)_6$W%QDLhm=)KNJBPZn0t0$aaa4$4+LaTV= zlrxa6UMl+>3O{AbRtG&It0$}>mFlSZkrS>%L8u$E<6e)OSE9NrY}DfDIFddS^VK)X zI_XaqX0Uko4AjIx|G*VGlisFy;?eDat>BDyF858<>{G?dR?ch{?}lpX>7wwNX2+ zATZjP4epn{bpus@T4{_!7=y#mz=P+x*aO4t{*zm&8N#_d{_)!6b^p8a$SVRiy0+23 zkJP5gtH~dXpZ@VyIG$)iAz*hT;TOa{`b*KzOuN#VS?9-*&xXp|a^SPk1GBy72F1c= zbd?HAqkEWU#m1mrpz@;fvcY`^Tpf@XO==Lg8w;Vd0cTE}Xf2~_RvM-;C?;=u?RY>gMrej*`q=P+AkDq zN1Yh=wp4YB3-F#`rFOj+vU1?f7LrDPCanx8Qq{eNw>_5oVl5{uK#H|9@+8-UTRp%| zgdteBrE3{1urZ`MJpx2ov+wZgGZaL*<Bhv4`_5av4g^AY;6yjGmzjXhhK^9#<;Jm;6-dWbx`Hug9{ z^forW+i)?4dZr_8<6ev*HVenqFr_Z;Kenv1!%?6H-?4lFW^>0t2&_q;z(hU#zgt}3 z1uyJAcr_O{sgW^h7g>z;G~hjl<7={wQDM1kqdeDQ&2lO*zf%j7 z>=qxA>!YVzg<159Md9=H^jiON(1LUg?pR;FGz9ra?}q4_nHW)e5E79=VL_K-7uy}g ztbz^?t;ALk`dvMh`HBfT-{0~6ugR9Lm}U(1NC%~iVq{ZcrNE(Oh*~-1jAq+n zfes^_jK}Y-)B2sZ3vU9eKZf>4F$Rt~3|y$D4aL9o^d6=0Zl*>LS*h56Nab|2i(m^- zjP`dJYCQK&>C*GwZ!HNh_-sCot9(IK+rdD3N{!K|{pgJLk|qmE+cs%eimqv6sij+- z^SrSkRgF#2Zg{=b zjJhq>|Gi90wgWy`ao}BvgBt;Z84MM@YP^#Civ?>GOKi7+@K_G3#0ro3Y}IvNer2I$ zC=^EwQG#smyt7lE#78zNRp~^}^ZF?!_&3b;zh_J+@$RW(yyMtUe_ITRA{C2NN6+j zfE=d7C6&@9?G^bZ8Mf_#(j2|QJfBUk_9{0k4ug;UNNAHbS8z&NIuU!OsxSI=lkFoc z)-cK!gZq_vc4z1}@&&E&o$p92D_`^3j~p0Lm1c-~S`)isYAThguG2i0H@?^OqZsPI z!ozeCxCEQD#?NOK_Bk1^PA=>f#sKmvrY4$3^f}?^7+E(Qy|L{CurK3GcF|#M0ht%9 z{)le>{l@yOqX~gnBkNykqRqlYKytMmi6 zPWp=c_LQ5CBXl8fndpvIMJ62!{_J84yyhE#y9I%dPHyehzIdfwZ@$n zAOEo>jRGe#0fq@c#3M$sVZ5X6?3A5ex?({ydr{iyhA~Dpf=VH$*KEZ9BJP3KuYqZs z=AO!?cX+LV23l+3Sxun-sx50P2?1Z{nfpt|NlYN8l z1wq%GyCJ8f+l6SrPx6maLQp`DDh2HJ#y^>w%7BfFVuiZpl5TM)jeN!SfDk{dJUfb* zUD+NgJS%;;%vt`irE`$Ot9F!h%afHr!>6Co%x+WM5iMEgxnK1Uuvl69z)h>L`l3nfaGxJm z|5_gPtsj_dYyaKx-`B{eFO6;Ob2CfG1&aNQB4?qhATUT~y7vC$2;B=7vpvEX|1xQt z^Ch47(8KhGcbl{=Fv+y{3-q0E@!PbxN zzjIr9Kcqv8Q?`m%i;haLNY#f+|7Yp{^rC-+o^la@KB zs)sa>30sKe*9{MGtH5ZsKiTZVun^XUb^Twp30~Uc@0F7be(-W!EuB=E z!K;{JAqjQ|71j=}<0qc_c~~h~FIh1wReec*a7qqcFFh`;n^HHWT^yy{#kQ#%lxLhX zWRMi?c34p7chLX7%UMrjg@!B*CQIX3;}trhhhP0+mw@x8O)$eb`^azPJYK%el17e$ z9P~Wf0|{9dr3v0kFy*F=+2yx#W}K+YS+A&9U00?<*js!^6T4(d#Tzi;* zOw^R|J1mVTFNqPK!xq#1AC^hhhBz9<>wEmIG(!;~y8YuwR+H$B^-y_doc$God+5yov?|2ae$clG_ z-h;4rJsrg$FI%SQSSa2@_guMjmtYz6y%Y+Sx}JXC3w4UqUH16dWdV3T>mhe|J3MNB z=TgV@gG7|QnKj3xtNilx+yYX}Pgil=b00ZrrmNIYEW8T)sjv^1K0Qfcltrp~CEKK` z^@|at(YZd&iY7r9yVV0L07WU!2J}09WFW#d7H8k>iKNOAudX3&f{QNwPM8{)Alt0CJG};C@M8o`L*cn{lb4fat&{2BOd40djDIO$A!y_o>5j9@-L|`JR%H41Ib{ovi2Iy2OzeQa zzfg5eDS>b&=+GmK2Gh%ONji`(Bu-q?pzM|>26cc$#Xk58+)~EK`0vs-gVZx{4SmHE za@>&m)VcBM*G!OV{??6bQufm5!l%r@c7S3*W>r#Q8{O)G?j_G9pUS7s`xiqXtHwV}cXSBMnfdHuJAYFs9l}5F1Ne3Ox;v|Q`ckzj^TGhsnmgBi)wj;ov`z}5W z&j0eS`z*;A9Mzlz#@oSJA6qFlnIfC0uwJGRilN?^t*h~R;t3e=s}SjmJ;2Nlytq)i z4C{99_Uv*t+8XQW+NrkQkcNEVBU$kwC$e!Pw@4WxsF6b-ikXZyKF8Q`hC3$lyh)`X;cs*Q<1o1cjj6Co;>tYNtX{ z;&MqUl2k6>gfF=WW`rZ8O|^h92+P$jv~=i7{=%{sq-|w zUt6+d$$BZ4EnBAjbZVZgTCs*2lP%}j8d$&f&&G_!75C<|V!G~qXA>$8f35m&WTgWa zj@f30fCP$-r~W_oz67kPG)vnn-b?af$VM=^MNlMwK`dDeEojH8-lw~3wyyqrdZw4@ z9_huJUTS*M_FlN4QnD(zpt2~7ATEF#Dw~2>QYt7ail`tFETu$1q)_2MhlG+uqPdXp zNBggO9=UrU-+Rw@zO%e9iK8MiHBcpoUgJS!VraKw2d_9Bv@dUY^vpvM=Kl|YA%t0m zz$pI5Xi0{z*fs8VWQn=luIY>hMEtA^4=a zEh_wxCvH?A)5?UI7;9iN!soPKSlJAVtlxn;^I7CI@j)%jIq6KNOQ%b2XidQNux#EQ zQM53{r35nKeQd2@!r(AG3f976`1*uQ`uiog2J^JFczRjdQs3}8?hL8-EZ zVmDDR3K1D(9lyk7ePByas_Ju@)H-(%^N?i?@gn1@HrE$B0VfP4)}mwBa7+|BVwNoY znYk>G!^XfY2>rBty8?=2FyjJ^oa~kb)%4Y9*zwIHIWt|K!^^HQr1KBy8TdSqu)$wGPn<(1hSoxLZ+X40AfD)J3+4n^i zUXO#KAv%?%c}$<5QK5|YuJB5S0OvyE*sP8{UMI6;&M)hD0fc zY8DKesj8!@J)&eqCsdW|cB%IOMz4YB z!m?>~frFr51H{Q#*U%`v1rn}0%rc|!{CVK-$LHpWJE6Uid!E@8zT5<5tgcICfnSAN&%)Xan(Cm zW_&ZG14;}E{n}uf!XC_0o?-i`@~E#4IV_O|2+J`0$QpulJyq{is1rd%x=y*%?L5qJ zZP0hL>J2QtwBa({_VD3|6APF3e(Sph;~`>kU>jtCh$YGU0I3Q+3X+znR>N`6X(W&IC;8UMJ?W8NwG zvT4AYG$hE<&pRO5%vRHSuNL11ZK|e>+)#YXGzaa4t4^u5s#^q^^j6gxVJ>sX9y(>V z`L@USmA>G%*TmadxN^x8^Hie4uG}r84Ie3zdG*uF{0~EMk{+Cp$D}9}YwF78+xX3Y z0UU-;SqazS%O~t7C%8^d?THC9XV`XFR9N85E^)oA&YxxYw_bY@B%AAi7cho@irFnm z@rx8bvB4ToHfE#U-0|+gnjNLrro%zPKRGQ*P zdE~$pa=~N~Qbn=HDQH%Y!0s$?6G3{XTd`_>H?J+EndErEm*;zpNfTFi?PGMgk{r6< z6?Nlq7j)x5oHvbb)(kqKtjO@vHg#tEOdLoZigdSyY*P;rl%)bcdI#Knj)pSVR8400x*QHn3Z(W&Esrx_|Z>{Yv6$XJL7^hj$38N66$R9WjIj|tM>=J<6<&X6shHw1dYqp)kt1E2zk z<8=pZCK=+-*bTfmrYBS{Pj_A?yv3BuZ%SGO7iE1g8$R_lG>X&;I_yE}!VON(fMvR6 zI~WZ=>4R_n;@%XapQ6>juOrJGILDc0;y@)*Y#c?_ju0BcGAWGt=g`Z+W1T=_G5l>> zH_F3TPux$-Mq@N*sPs2IE6mA}pPykIcr|2!Fu7F`^~Mv4P6vF7shT=jDJ*4BQwn3R zSg1RJb#yrL)?vl){!biI;PS-%@rIbciw|F;#NL(KB{ zk7g@wob_VAYn(r776tw@IX{JTnwaSpioHRRYg9x}D2TfW?h{>=e~X}tJ|JEKT3&)U zrjEHtqhcV6&t4+%avt;Rq|5lVG6;skzdF4%iI>7OFE{`tOjy$v%Yq;-#2;p4X$%KH z4Qlj-l>&{ChGfO{z8YW*m;P)}5ka(L%HI2_|(BczHT=0%ofOsjeklp2$RzMIALOb4^V6&MfO62tD;GSy5MF0xgH%d)CtK`=c$iqZiIohqTy7UxJdNbygV_o zEey=;=PeUv)4CX!4iGO!XDMcCt_Q>ruDzyXDc_SBadY}x-@Z=Pjn>8Jz@`XzUq%#;Z>LyTRBfdqs{Es-L<*}I z?3Ta4$5v?EUNyy#gsNgL@T*`VJ+E$skTj^TYLQtlKh&nFBunEmHYiwnn(dBTHZECj z+~S*ygga~|(t>(<*>`V98ngrQLT0DiDfJrvR`~|sYxrle2U@(m>p^~CEq!;E1eW^n}QtJ z$kdq_nG+NX0lX5(bSbjvHSEpfcpQOe@*EO<_q2ID`l zFc93C^N1t}>tJn%pLQ)Y;frJPYHWeIm-+W5pD~u5?=Si8HZLQJp0YL3B=xmH(J>Pg z?V(s8g4qdERcQ27OeR;;e_UF*sKYZJ7~k>O6aOLxtGlY{KQH}jJoCiAL0c{@r~kC{ z1KfdHFgYHo@_!U|G9(Q;Wj@dqytSQP5fDk~PW^B@jSm_>Or8zl7>(5EfO+=Vac}*4 z{p6|$BVZl{{!va^xBjh;qyLnA6`*>ZhY4jCQswzr( z70K=j;0BHKazKa709uRE>(3xdERxOOPdC?ZVpK+XAC-FrSKl_Es6x_ zSJ4&vVeDuVQ>@m-iFfcyW~Qp@r$;F_%+?EV;dKSXaxjm1D`>NiE=rkAUz}DeMMhSn z1sE#A$f2LOcJoe4hm_4Xg-1ePcrEXwT9-ewY38SXJu`~Ddgk@JV%c%eyl2mV2BcnZ}NmoE`?0&D04|F*kw2^K9rMZJlXv}oNHfoR3rztf`^hmCzrQxA?B$I z*mO2h*voV;n9vu0Xhf`h`z?;1y-sjK1@%eG&vUd!$3-7JJ%t?Q=D0ZUdg+RZr&~j@ zrzlcIMPPm;&li=yYh^`H9CIq{s{alNs()KP11q5C?#iaCA$O8Sx65i}am>|)X?{qe z*`U45bb%<2`DdW^xIq(gm!!^lpT|-6c<-@e)=KPg@G49$9WykE-;@L9?BB16M`D4* z9Ent)1{E-yRY(Sagw8qiF1mo;7mOLN7T|0?p}7DZ2%CM{G%W&D5zq_T%cy0w8_HgF~#;#<79_T$!wO(+iFQjm&Q7eAIQxQoUn?kE`osI*iF=(T8Ua$P>eKU66Px zaarqDrO9#G3Vu#DoPZLlz~b;T&?`IWgx&qa#}fF4>)>(s`1knb^oMVhuzTgZLtxp1 zv;hb+Bo-KY=ys+gMAr-rm>sfm8fl_W%TLL-s^WQ%zYy&^bj3wDoF_c4Hat{k_<_IS zYWMhXgBEF?7}dG<0!zQ1z6~jiKJ0FXTyGiKJ^#+!SCAbbc(wObbJ^k7#8PFO%zK+C zHkl&pkkH_yGp3vhnB=KznL#Hcu|S=30MR!0U215zTu>NbPeWvGDL5G>^PBCsziRh) ztIcVgxj@5#V>uQ`pVP$fKQv0{X$Z?4({z#RVFU8IX_XS)r$HU`W4a@#UwTei;!-NQ zMW0ie#768m4xI2|@L%ls+dO;pVLC1ZdH0Oi?G z+H30@^ZxV&bB;6%~1a z{cE+?7is!jPD@+I^&Bg|`N<;lo?0#wj03N^Ep*^!E^L*j3yZ0Jnqn#)I&VKDyWNYa z-I7)FZo9tdT)nnITlbY4l0V;4L{bms7h!j|q$Cg(EAEGs%*@bK@^f5(Eqvxdc?sW^ zF9-KGz+?e!x?>9BaqT_)-I0stLXQqx1Oc*pBW9C*6uX-uc~}*5)8i2-Qs=m|KpZ)Fi%VL;)SNuqvmvHNJnDM*4;Gp;RynMVvA{uRWS7fhrkxJ!pl?Xh)Cs^`hWn{1+}q1bhkDK@kS)%FlHW+- zE;`;@w@ip^M41Y`w9XTn_)(q2W}dDs?9DT9;MRyy9mggodo$^e!hfuvXx^E2DrvE776ei+c*OAX-bNI+Djw4fWRIvx)FN0e-0Y7pLAih1 zn@JEOKEZ2cYzb$8k?jy|j>+-ub3*jwn&xQwe~iFUWGRH?KDS7W<1P_NtVwcVkYb-u zE`><})vaxy-DWtIMYqv< zx>ACHeW=WGD+-8~oh7;iKF))NOTZ5Evgy?6271qa4TSLd0btny9D_r8yg&J%z|}B$(kqG_8M+RY1xXY)BnS0l>9Q9 zdr8Awz8MlJgEpGp|7ouWq9ht1Lg0VmT)k>Op&IK4=@a zDvCknHjQoomSNS}?7SEvEU>Isy8OrL=)OYq=0BL9;`00m(bvfM%Ud57s>*9U`n2 zMX9+q8Wb)uB~}2wCRUV*C2(1IAsq1a~z2`xFZ#pUe`^rRMIvraw0but< zYZdjXLoyvEtP-gbm*VipBu;!Y7>O#Dd7Xw)U(GKS9rv$SO@J*M8$_NTQd=MSL;5>E znr^g4@9g@oUy`k(RarQ2MEQt`&B>=&$fIOa5eLW#CH5=aP^2)ozjg!oK5u&TvN^Ea z|I23oc;0T$1b!R%7snRiUW2w1F4^o~E6CK820nori7e#_&;Z#qX5v_wDBy zti>u}lL#%0y)f~DA+-)JFL;2RPSvdutTlIXpQo;I;GwtR_&#GRLFtx3CscCze=zLU zt;X#%wl93FWjk>*{L_Qxg`N((?y*q1d4(WZLobu${n7jqRqFH<(01zd+y>=%4O&>y z(&hY|*{E1uHakDGcW#SYz3jL@qzNPyUQI4(e%Eqcj$3I*LTsJwA~qX$~syOWH{beNcHZyoS|bJZ5Kb zo)`ljZ9#ENKhe*fFnhxlOE!ed3m?^#+?lL2dK^DEH|Zitap2(ZVG|!Bk79EuVBU<_ z%ljC5A8~0}uR_Y<$Mk?xs{#pR2Ao#P3KeHP2b>B$F!ptdz6R7xHxy@7gFypM@v=0! zSGAW{2qIyT!fkf*Lty~84*y6?VyisTXgJ_xN+zd1F3I_g5iGy0`{`_QkDGhyz^3Fw zlYs9Niv5fty@r4<2qZZlQ|$-#YKVYf*>1Apk^EWB?oez+Pe(3jRjhZv0b*?}f;iF+ z73fGYQwt>37`BC&j9!YuZ>985*cW5L$%@jzdR2y|3kbUgoiHkl+eiG86mBf;=~`3a$dHkEucwK>wM0jgfzWD zd%*X$cCq2q6B%mR^t-Ng%M;eA8nlM}s3}#v0=kA&iL=$EfvbRW%5Z$0bGGK3Fvq2W zL60Vf7th2oogSx^9dwt+nAIuAXERJK$f$dx83Zeiak3eX8`|(^zEJY6(U|>L!O!|h z9XDg?j)0>rjk_|I=c?~#S@{Cdr_#Bp;BpY0>h3)cc z(4Gk!zUn5(N!=j_T^dx&Tn3zw?l{{j{tCOlDxB}~l^O`&CI?V!yxgouP zHRu@WHAHEk@Eph}FNLJh*^q{uFee(naV|q`#R%E#5>8fe@{!vAJ#H?6;IP23AZoB& zh@2nb*%%U+%Y_ZLE|aisaI8Uubx*SGRh$qp=~w$QPQ3EU-GbYo+#M+_*Ob#+ozIC< zp?2V!sCH&CwSP|Y%#)x@x8bd&fBAK}Cfc`u!4nBE8EGcMc31I01wz z6inv_8vT{%;n0O^66brefLsUea`+lFTOtB;b8n}@MsFl(6zNw#q ziR6x<;*^3#{jN8#PZC!`NNnCH#T0TG$qv>*L8)80|De;#;5cOqjp@lqPaXcixvJLr z>cVoU8Xbxp0>MDMcdBYzSaz^2FgGSlJqI$g1EZpZ+d<2iYW;nskI;yp_lu_ffy6j) zq$ke=D_bdcGrHSDr4AM9ZVtJI)f>3f8?Y1&H9 z6v0j!^UgqOHnf+=>m4d&hF%uo<1hBPoML()FwhpFm&Ysie7R;(npX`Okgo*l2&j{z zJEeO>NWlo=$;_vI{R_5v;8PO#%Y_AD8)xM*eGuANCcMUsynbA^g)@@O?8l3C#|$I5 zkrR@rC8-w`0wa>%y1elO`HGYH1Bjdeyvbm zg;kmjP$|_Y-R7Yyn_nbK1IXpLjtd@KhL9_ixIBA2*B-zA$EOk_IOHFExt_#3u$h5| zxe;-cG>Y9yk%d%2Fe!CYAP?D1U1$$U&&e1G4ddAWnbE{!deHk^|5l2Q@U3mGCM2gA=w zPC8Ej1k8uYI!Kt`W&eFnkeIyQL;Y{&rSD_?@FW`>Mk zgKGDLuyT%V%RZhO|FG9*Y#Azy9XED^3aEz{m%!BU=<-vVbf=x zbZu0$Dn3(|K)Z52ZMddcaY&8*^y}WfAvq?vrNCT>;hxK)Hu-+9Zu^mkjBQk&y_xnJ zA5NpN@7d7*iUT7gDqlIQ27vhi{ z9=XstL$T1;Urj||iRKC>XLc`x-o*p{;wG0yMKN?G0!hPcWD8lrpghILq65Cw^m|=? z-IDjP2MAk+prJ^8K;8yoDksSv?^Y(;xpe{Fb9H8|_lhZ)ID$~EFja;7EBO@+)`9kF zuFhcKNa+r!(KefpP8!^yXL)oC(}P&Y3RTD&k`-Xh4*49**nck?o4 zt`=?$Ua~Tvm)RXs;guvV4$q-a%iEMX)B(ViYms8fk{!U0`6->_*W-A{T>f?-MI^uc^>20m`{Q5zs4eyW#zRGy>@G_kDjrYNn7oCID)r*anJR1pOK&oj+W$Rz#NsVv93$$YXvV z#VfaY7+T_y{;Y$Bao5wis!FnNb|u;G1E`5y&_cH=_WNK78#0#yp|0y){`$aV#d%Sl zuv^~m4iijbD4K~HAW{WsO8H^uJaJ-Z16zlR0Xv6u*wSb{6wIZnu7C;!HZ!)k)$wit z-`z3aPG|~917CS@Y zcDLlIq&Bck{=w2eclb4EPiw9&+^Q~I)ZzCMxQ!fQ+iY74VB36}^-r-3;GIQ3OEb?d zT1es{?{bH%d-fysR-XoK)VyQv1E4FCGy6uMUVvq1$%-WJPQ_|AWm<=9!oGQHzwzPM z-umxsd{k~BxcVEGlS!&8`Yo)Cy5Sh>U8{t(Ac_>WVk4@@;KbJ~B=^j8u+THM8fQ2_mG zBNEJ;DHZ~Y>#2xp8VM8YV9DJf!{CR`KvA}w#IlhKib9hW$7g^HC^ccu5my9E%;@8% zSe|kg#wy>WjM#YU_Vg?naI$>yD87{AT1Aky9*LSbid{>QHB>~0tRFn*jf?MxJb_At z2`&tQR zxra3B{q{QeU^e@ixs&0r`5g;x!$y#eZdGK$$fk*l)ceCOc--=+55d@b8r>!+UUW!` z`!#eKzl$#Pn5cP!!-yM(-d;m&_FPW$0M+%a-V%S~L{ivM0PS!NT(MSVGBq5c*!>hK z00j}Kmv2y_Y)G=ALAw}%f~EG0Q{DQ3rwhB{jS_pjl}sw|HI)krJsPwd#mT{V#R*;p z@3!civX{|I@nAsl5FVxao%`4~*B^5>=d$^SS95Wy5Ur$Fs`~Ih%!VpEj5L z%rF_C^%R>xkyt9CRnZd)L2jg_=m<(6D}hC=UvirN0D81-`Mp~8w_E$=TlFv-Uc~9k zcU-6cJ5B5R&FPl86tg+7V`_n%d6)<{${$H^>jXeW5GOA5*ao}@MUaJTl;%udDI0Xs z)3AkkrxF*eo@^1sv*{Ad3Skjc9*9jF>ZvP(D}#@OL_yO}qcla@C(WnwZMzb(7XPD| zn&%GL5dM?WdMgIZSG%tXa4f8YZvsZpWsM6=6JPVX3K5GRV5-&hrlVrs*pv;DBwL#6G z6iBv{dB7;9(&3L}#c4=RkGo*u#b5vMS#Iz!T8rVE?fA3*zpoviYxH5?FVR+zd?G>4HP&#-Qt)_B>#sXdImpxMNcGEiymqdDCJqtYC3<>(o;~L)1ZA679Eb` zS4|f#!m(LF@aP851Mbyy&s%mY;2Gzzz#wsb)vUXo+psvSFwuVPirrR9SsD2GZg?Nk zZ-T3$+H&aB$`Di*splfw}5N>&t$^)pZrp=^E}zet3O$aW^5 zDGkgM<_Vul&$wmN_9%ZJKVtCBwhjecuHu9k$6YL(p0ha1y!G5+WBC@E$+5OYFKClv zMK>x7#fdBV+r5i`e)8emMBbp&utJ;tW}D}L!wt31-EOx7oKQR2t?qmD8^-C!_g5v0 z$tDMOQx2NUi#sVciz3^ph!kl&?3>g>qPZk}vY^8gW<*6}V7uf)(lNI|Tjif7E*0g6 zwg?^rw_}PXgO|B*jqr@%wre`GlWDUvH`h3jVmE_=59WrHGUo6)^VW2SopCHQnX6$6es(DG!UV3F(7JWS|!THj( zB5iB9NS~P+puH z3sTW=(s1xT)VpMX6)O%eWaF4dY3#J!Z1Pk{spI2f+{@`&!BKUT=jPd`r?u0k;m$aw zPueJ5FN2GaU=aAcNGUIttycjD`eR;EDA4$L^+P3{^@eQ>C^Yqexb0aD+^>G=dUvo9 zSMopVwvoNuaOJptCFG0=uF5F(7)6Syh?20pkoT_y#83Yp+-lGkiS~$20{QhZL7Jpa zwbi*s)F;8yt2;gP5H6}zMon3(II2ckcO-FGP9D$~#O3sM>4T{S;kcg$o_3q- z$Fe-%6u$@bR__!)=*idAy5o^*Yd1sFgjnr#Q&a*FNTv zw>)SzIm2lNqka^k+u>)lG}UjfD<+w*jiotmVrlkMYym~~P!X%AbjuIP^JiWYoe-{> zvc-K{*nY1TK`*S$mr=JAM;1MVw)_1+a*qx6rQlF^(E0opw*jZdz`}q5r+j|Ti~*<1 z5^%c%DrOBh70B`1M4o~cJ?vp4ZZH081`N!_1=Kfds##OW>fE#xV)D`z1Po+oRDwEPJ zUAk14N$sKroN}ow!)}&wH`}nA4Z9^Sxm1o}HwSi+c__uFDD1uXp{FJoo(jAiZsk%C zl5s5up_4A3ed--<=g1TH_cEyWR#5ocbWMKG@Ox|@Is(5L9D(1dL7Ny_`}q?WJFEr_ zH*@sQlYaFJ^Nf?j;>toQtk^Y%e|TmuSt&y%+X{XRe}iuU2)|$R?;>eFCzRQ|J7kme z9B?(nuqn(9!2`*4aAFRI^@-Li6ZpF&8LqAJC%}L^IPZk=nB7|0$8A}xwiVniivPtXeslT{f8vP!aXo2lPelhi~N1>ZB{Yx_J*=vEd%4 zpZr-l(zBi!yaax&ydUxjMWXJ|J987mK=p-gk)KKHIuWx z=jl_R^|FH5LGaG*&^5w#*ku#Nh-e}nd(AeTW+8K@&?@64%?HfUQF_a%BjI*@w1WXvp5 zA5=GJ8d~7hR4X~af8aye-4i{ubgmL15#qLu-Z$p`4q`DXg%oZ zYREX0k@MlTf&uvupK_^sJ8NhJ*bsa`t-H;;NZ<4*9m<#&g|;d-F~w>flAfKGMlOJI zkB>-!2MSB;umx)CwAeW|N6&Z=GZIE4N81+LoDehV&g6kb=4>3V6sMB~>X3a4?}OJ+ z!(@mH#7UZypxAIuSu4QG;f>---d1OQSe!UlnJ&rksudjf2i|@N3fUYYqYdX+U^Cit zY<Eq*CXIbE*;b_b|wT<%sUI3yi#S}&|2k4YA|-kW_c`1g?u!1#?f4La@fJuX2B z6RcK0sYd43hoQ-HJ9*X6KZVv11Za~LCG%IxvNdPO&5#zblsQ1^R~(+I`b?^yG3azi zS~&wvUm6XHNjD{#HCJ#&2IIodKJ!B(7a6#o$Pnp zD?j^Yp7@Ag2ffBlJ)t&%fZ6cbV%M@k8jZE%v!_9;%<11>lSU{DY9#3K)G6zfpfvH6 z-_OH7VjwDqgM&^;_dMuyPkv`^lK00VOsM7ndHe+B2c=ll@9A0 zTVSe01EIU?o+n(=p-Si1Lx*fO7Ci5jPf(xRv+tP0C_lSt0=BqWjf_A1e7~!4nkoM3 zH_wx`4m{1|noKiWD0UMCqYx1V;!~(FzdBfVF}yc4MVimM@>bm2C&?eLd~LOG`wShj z`yxf}UH7$ax(V@d$3PrrhuZd0(8dWJlUDxsp2PEu==kY{`MXJh1Eb@d2|6k$wv>V< zn+Ux;+Hb$)sqd2o=fqbRZc~@jrJ~V})oTm=4h&s|SD@Z^vchm3-uV=q(lokVhSh+0 z9-1dZw`<~g>8=C(PCCyw+C36!1gnnuloX08Y@0dx8ftukCaOoyAL-1)j@hpLSUfT!VM(k@2oh zww=8t*9(e668PvqXTfSJU%E@J%cAuhDzZ!m)hm(6mnvTUGW!&v~{;bj%WppX0l=*tW47A zL~W%k)=h^yiJ>Qb_t4qiweo7wHHF?URn<(c1?_OdzNza#T7%crtFj?f2gzrTyWxKUbt8o?QtLG1Q*9G zvwU~b{6*&U-CS1H4jlTkpx}5^6URWfPp_yZjZzdt8g!bVb+Z}ftU|WgJrl8;`rH5A zW8Nag1rm-6c4DESY8|LcN0CA%#qXHlVdzC!B`?MA$f7;Gj=XWNz^YJ)R=KavJ3jjrzx&P}>VQn=*}92hBw zO&sMsiUlI8Oe&&}Ef#HZvb#a>C-UWr=ofXlI0d_;oIKVOl8ynza)JH7*_rFPHMzEaz z*0-;dbq>5v+GjHS+bI^pbz7;3I@vXnGdn%FEo7S-i-VW(k3b{TEswU4{SwrU#(ug? z4YrmHI+gkF6AlyL*=?|&ebZwb_GkCoYm1YgH2IEG@on=)y62PCj!RQOpu%B9rYMzS zw@_pg6_KOK8A=Kv{oyHIhX{!ikHNw^Pq=NS?tr`#bg>Qxq)0y!#(STcR^hWnI01-( z5g!9+?1dAP=UN8Mx#NFsH=lAGHpXvZ5=nfsFyJG_0pHbA4lvigxIS4vPJRB5hPeo)9w1DGV}%ypHYu@)EgTy^QrrEa~U z4IuuSF4pC0uwpmKy9w6mO)dw#R=Kv5qR{31Zs>f(`KTV4pi`JMF?M+mE5hnl@|Tl- z-fH2l5TMV=Wo`vk(;2Saiaar9u+b)!xWq9PUK@a?WYDQwfjyL+(j+KmFvJ3`O{^DZdmt`c7anMbJS^tu9bR4dcuQsUCeU`=Ntzm?e`X=QE! zch@NZ0bXfGGc;l}#(5(!#)(l| z^W$SA!+|HpGLwn1fMTH_bQcv-3^jNgT=z?`OJo_d2Rc)V=InOA!y>a*8uYg;cZ-)D z_s1Yj2Yu4(GARkH;Ae<;0}vtuQVG_HHfWJ}pjKwMK5jvx__)74jM*4W*Yneu9V{3( z)5X6!x%*Z3yzi-@MVG#z05(aFbFu@z{k-e*ABCY3_GX`c*K~0SWHgZdwHZhQ8l@e~ zZ3Rwv#|68>(xAKwULRekoC~5431OwhAwy3&%lXCSHVA_v^H0rkhYxHVzm_ z?PxZhjouj0V8;dfh~JBg0hv_r?TFj2IxWd1&o*Vw>=J$seZ=pOR5w%RiOpUUpcg++MI>|{mR>_}MkZ?moV$+M9#Y>fpc!)}TBvuwF<*=mdu z;+Y(p@-)P}zLN{k9C!+{Pz|~g79pvsKG(XyzM$ry5|>)%9g?1TQL~_BE{Bc*XEsuZ z;UAy{m^gtcu2`}qUS9O9i{F3pOY@x5E9Hh*$U{YGmV+;fG<--wld5V^^}Fio0!zU4 zXjOEC3be zG-ITs@y*NclcXu6&?F#{NwGkbkV-|I30UQgRg;;TBYu}WixfkZmoYB-5S-ty`|7SP zycpgnZIbNq)HN^{!_(-J8R!_cDh8dBWNoy48$7suzr*wV@J~B@vBpPQel16t?HxDo zo1Qdxk+}r%4ilIpQ|vm5!~=n0;Mq6R1&0adTGqQ4FgbKK#I7DFKBXbIV^gh7Yevm# zTv+ok8(zZ+7mjOz|EL>uCPwp7=#+kjtbT1Oq=2@5M5V(JmC5I`Sxqj+S*uG3Z4SE>VVUyMo>yc(P!0k#W{rvB%^!T=$xp3QS;&oV%dD{z( zMc8Ci3WQR*zR8Ns>W0}LF^9DE+Pt@~E^7K=!PgEhTDLgnt%@&Sdw2K0t$iC?l=kv2 zu%&K#=#k$^f2bYs)(fDgUfUCT!AK9aQQWGC0qN^!=TcSi-k&LL=I3&N&+ugV!k^Cr zVQfcSKAaj!8x2eEzk1h@<=hO51Fw72O$^IAijAj83>6U>P{nMQ>ZRBKjg-&ZH91TT zsS?|4sI4Ce>p)@s^K5nHTmPwh$~S^y_U3;%MOHa*S`V}=M$8`@D0V$X5~zrep_6?l z4OJsEPrBBSc<&0YJt05{H@i0!stuSbJ4;tQHxkwjHNNuC-D=kZ-29^-eD>y3^Cj*p z6$i1fkj(`)lokP2-vWb_Mpr(yRkhEh%}|i^nLLXwpNdsYxeNOz&grloFot=;t|%Gt zaLa)4O8a4;xz%_c9FFT_LtXx;mZVgDg}v;ZtVk5E3@Gy$1mEKdko%NE-BM}b0a7lQ z;EBU}V2l{G3D~p@jIYd`ANQ)~V6KPtfvec>n#f)XJm6l%RJd%Iu?)n5AAPCcHI7MP z_VP+aw}2`>GN8}{B?MAcy`pk?BnWKn=9SuAf82UZm^qbp#)V}}j1rZh-udZQ|7f0r zaoDm)3t5#Cms|hz=qs2@!ORNs4PsUoa(3sG>-gYq%7+SBi$by1b)pOhXLQ&Z6~m9` zgpSF-{K+!mY@<{9r$4tIBRd_~DLrX2l^mv6=ol=7wTG}IFpggh9D|#rS$_ARk3P?* zQ&Fl;VebZPmuAw3hL|_b1e_9FVa@<|#!Zi_l1)S}*s8928&zvjv-T6;TM!Z%{6Zp> zIszxM;<{wH_ul#I1FOUO1*O1)U{h3ahK`wi8M!+=?6&yH2|Co*uf6-j@n}vsZoC*s zH%9s_pKDH_I(eq1!>?75gw@KvSNi>M^Fb{ORs@kM6#k{;T{nr7h;r6fweY`FWVAiwp>0PNc#E> zXQRcb{zCcR$VvycI3Q9!Vi}x7v56FkqaxZNF&E`urQR8Q)VD?O(4{*RSio@n)18yN zZE7m4c&)7-Xe*v&%S$*}lSxU9SFf78v#-=T$AZB9Zplfn82(8w6nezUL_?f4KeV5R z1tP7AY!Hwr@+goOO-)sybcumjzt*{f&UP-4+Z#((V`t=)gWI+Q$|%pWjGf}_G>?Tw z^HKl3m06^C$|zp11J5m&O-#y16boL>2_OoC6>tVw$KN6>49KCc@^pt?lf)a@jclT8 zqAL_DLW@=wUE*R8d^89>cFVT{?=3Ro#<-*hNBJW~(?dxjK@mtD24ArGE8SVQky$3J zlcfWs3|Gg2ib>|e3JsQ{oc3qs-$muY4?(=N$mCk92Vt9r>^51ZVTQL+x#C%ty;=#F*b`=lvUN zANG41ZH#_l&<(PMTl&UdkyL6&$T>} zjN*YzzItJk)Vwy0%jDv~E{uf|JM7bHgC#9e^gH8=(kEn&t zp_dD@yjQ|;T>Q4t2grV})xoQS?G7>PL1k(_Mu&p!j)vW8{Ri{8#$m&=7G@9Rvdj}U ziCP3W^w?j6-ND^3ys>PpbFK4wzmA|(Rh9;?ST5|AZwanfp(k>Ce*U8P@D9JCQ0%P0 z^|Sq8FxK%mqQQ0pZuTTjMrG2Kv)c>Ibs4xA6$dUXv7q9RM#Iv^FE$t_rZjPxC>tVb zdRkX2YLUi!BUx8TU?gbXwb04IKwlB8>v!vS8+6Ljl+ydXx)qJm0*~{7Ln1`m!m@)= z&0*{d9J7%#*&Qdm3@9^sqHT8M)BZcg$LPxT{y?^hZ0F|6I`Ax1ZsN)oQtVy|2%aM9 z0%P57dGyRH^Vi+-*d?s@tl;;~I}Dms1MH>>P8?fviR95@=H_Y0b(7y6j+V9YO0*64Df?L)P>o zn%&UcWm^-0I2;p_O}$v$vpJ+WA$ih+rl%F=`Y*4E>taDI2KC6$HtGsIARvpgOq~oV zjct#Rmkq6z2zuFLY<@mZ+Wk-Fl=@sS;kf3JSWpH48YQTXM(I_I+vAd_4tmVnf{^4t zCg5REy7O(IGde%xmPfO&!M&1yd|^A7qQ^nYrZs3ka^59uh3HHJYdA9eB*a*0U<~Lh z9tn%&*9W1f>uF7!x>(caZg`R{5N9QZp0_(MI${e%h2(VnqtRJ8{C(dakQxWhC*Lvg zq8cf-fg%^Fh%2PewMDRjx5huowIr|~dd^a$^&y+pHw5(|olyT)5|R+U#=nJ*3(MBj z^6G(XtDeba{_^_|bbtK)2lY%Qd8kR?uXkSK-zv`&cE}3Em&yJ)#nfiaW%XuNp6{n` zJdySAc7;^=9~Bov@~0A7+FArCYo8F_$bQCdej{qu-GKU#a`>oR=5N1`q;wzr{?EEU z{r-bpA;;A{(po{8|3>jcS%+US)#B1B-{(=}(M6~DWdcW-MQGUtFE7H9**Qzl;xubf z{M~crn@fUou}}`YezG7S9w%;{cZhf0x67k}_ke#;zMaL~BU-AZ7RTDscRk+Y)?#S9 zpTqV3Q>B8RnKKtyAOnb3CMk|f$~>^pb!&Js@YSVjp8Bp)tPMq`^3srM`naS)n?s+Y z%LRLRadT=stNi;V=f|wI!{JY{_S-dmu{eCd7%%1ahB~a)%kO@o{?l8=iD}-_3oWE% z3TZU4|7R!`y78*12yDG92VT}%y5F@N8Zk>;awSiLbXlOu1UdZKS+Ho1SLl`tKXM*) z+RJMa;fX!d?gkBUyVf)J)U`4_-9qnBSBv{4vB5d6**=3#b+Tqfid#LzT^|GCUL1q^ zvwD#A*5wKHbYwuDBu`wyqzA_a12ajIcf0HeEN}3-Sa$2QF)d@ph9^^Wy?l>5s5s%t zaSgU+o4bBIp*#n6mMw(xb^;M#FTuc3x%7ZIQxhlN!PCoARW+oN-w%~jcjk7`D?L{S zLrJfsS#e06!mRn1__vBeGI$-pvx7Zp#fviD>VG?(s{isI(y5AXtb(RByw&y-nT@c| z9o#RuqwQzQ_L@`nJ8YQ70zE$tZOhP_X~1cdG?D*A`4Nc|cggmWY;agx70W2yy6-+B z`z4(o3GN+qwEg6+Oup$+ZzG%5gh|UaQ-uYjLr8sA!O3j?@x>T`>K*# z4xNa%6gTOIF!Q9V4og-9WclrxRV%nEj#5U4*&jEi18y~nOmDOM-p_tD*Uj8=JYTWx zxc(Hez_Gu6YLg4rW*dXIn0Us{Gst7j==F=8RWz+bhUMF-sxwm$PhT!PKBI%409!J~ zG=SYv@FEq4KOddf;A>m~{bp}vJlVl5`Ru?RYlX=|u83k``M;ZrIK&jY?v`9s9t_J- zS15D6*M**+aW$Y-Zs~Zv_THQea}IlL3BIkp2$IjcL9wnobgTOOj6=K>k`5)6*b}{9 zQg4U0Myu`0i`Qf}N7#AZTeQfCn&>Sze@zIoGly(3Lbe-k;wy4eoe5S@UQMBw*;GokPRh$?vX;dV- z#k!@52jquAGCYmm0#8^@wu3#XrW2T2dMG2=sOWcXR9y1h_r{n)KF^QN(2+4g$FiBR z@G0C6pEipQc1BeLzO~dKZvKcgtU;z=?&W5Afk#} z<#d$rwB}w=kIzoGd+LFC15Qy>@YqB7HuqIidVG3(ZqFZZdUn;51s$IKydIxZbRmP+ z;=bv%$Nb7>o)hM{R4{lAZl(EM9lu(KjbZdTz_B_M@-5*v3=ebIJ+^LdY(C zY^~s7*q)G5(BkRxxh_oSADUAi)F!VX-MkzcLoKH@n|<_1!y$}vuZKH;)JJexecrv$ z_rAoAGtu~l-*m`qxyNbwO#SfZr#t`MXmfg-LFFNc-HbjB8^wgmOaKqd=&21{8QHtj0 zV-QC~wjeMBSrX_k@zT|+I%ur=ZW2{^=`g8K%ew*{DRHDdB#Ul?=IFiCGI*)pg|que zc5n{e%_~vq_Dt*YNeV9w+$g^0Q=-BPau-(eQv|we{u^gCi*74)D4cm&o$ahc@#)Bb zJ9F{OSDi&C7P~an<3*#V^IZWj@bZ3%7CEBx`Oq!{Pt-^zzK>f?uhqpoLmt?9azM z12(Yb33XNe`SfGnKA7?{HLcKQ((l^tk)+Y#3arx^+TWlBA1a7HA_X3Ga0mpc=uYRM zGSD3Qq%(G>6fhs!A*N%aT4A#>wZX3CnXK_^zX?zq?bGbE4-!fCD@N#0X<~&AQY1ubv2f|9`e&<5=m?}K5x-Fu|vBs(NGLe82C(w}!bL8nhcc z54c}l_(T%*H6%Q{wdjT#e+_3o0OZPMJE z7~|LZ^4quq;pI=X%@vk`Fj_DF<13y2XR6WEwEv>%uO!-m*YG(eW@R(Q0*u#F5k@6= zERRU?Mh|3%WFJot1&7(eD0b85S}WIs4?JPW;F{31&A40h6mHm9{K41tPDX6xfA-5i zlgQU*eFZI^BYHj)C^nWNA5sxTa$F=~-)x-!T|NkEiaREvT>1H-v&OhQf0zA_?El;5 zFPOKwIqY0wp?MAIIk3DKRxubjOM?|@i%Y(JMK4Dm8EZy)0*a=d`{lTq3H2_|qi-6` zMDgrDtR}l&8#D2diJ3S`v4<&g02=XSC%xkM7>c=1&d+F)6o(HuA&GbO99^U^UU`We zAp2Zez%MaYu_r4GOXW1lK9@`qTq6OGk42bFkNo)AxA4f}mqC7l=9 zY}&F8YtOr*mXACKn=u{O-Wt1Zh7moQ*yu8{&4C?|qb7LRO|c-?okK++yWyUY-cVy8 zByrk)uTJN3-T|*A7#PvRVs}s-xnRJloxTrsFOe>5s95h60lAvU1(`lY5ay3oY=LEP z*?dTQM^44!2)no{wm`s`hWI5nak4g(;(h<8`sYTxoP6g~Z_?_(cv(Jqq?^%0u@5Nn zDHYMbAc?0t@4Q@^s)}Rcy>%FdHSG3Dflk6HYSv2M;_y4cx<`YR99365|E`gczYGc1W;u|T0pT1l1mG8yM)`>WCf~{{Iz5T zcJ}L#|3im2l?HBgejN0{(!XIT#n5k&t5{Bz%(%+ya=oiOIX#R1RD#jnWkNi~z z%Y};3{gQ3!`uWd-H$`-QX!L^3p8WxEAG85t7$-w~l`fjILf9(bzwFf{N&XIobWR&XI)BnVT6A*}XX`jo5PXTDd+v0^wu&il8$tDim+=Eq>cF<_ znh8wKQfv)HPEirZG+m^V|Sd}s@ zRfP{Lm&Z&yu3pVA6&?52ot;({8Z~RPf2yiZP~=tQacZHi9jLN$X4lHf`C}+ro*;-k z2Z5Iz;AEvHf0le$ZC;jap`Q0BP`cpA=ZTK9{1xzDbCqu5r3eQ^}1RxBbAs;V@2!?6eA_rt-I-T$pM2!N6IGkZ zM0JH?FHz(I5<%xrUo|g_-YC|~H%a?KJ3Ko*a|3odUj|QQz^Rztr0n#}^jRYuaJmFa zT=(Rio>8-!Tsr0sI3>z@=y-9NKW-nC?h&Q=6ltTw5Rb69B7Orx|&)cV7 zWtntjlu;fNw#Mo(kMjAJ4spU36{LDI-MlHnVXcpaHis&vBdC8WP)v78l7pK>opg*c zS)p42CFybes|zEAJ9xUp&;~Y7j6?@0R>0YXwe(%T==fOh@uKJ1>KrG0OkU~Py8Pv3 zHx9hMvQT!D3tE32^j7Ds?j52|d5Oz85pKmXIrLq(8YtV^;dapFuKTfhRsP*TYE^V3Iw27dSxd_jRi!eHpW)4F@26@5Y1Jc`HqKi zva0{d!6js~15Z|mOeU*bip{3T4k`kvP%=Zy{0~b`z~UOaWDLP%@L>#G3CqCKL7h>IGujChFcjm&h3R10aL+O>MoE89Iu$iI)!g;v(sAXu zPQ9fbV63yk#$T+o((NTzVNx>i*h3L%Jyc+=c4-R&=R_3moils2OFaY@56M8q&3-^K zj{R8iWJBnT`lz`xe#n}$*mFUM15ZpAIPZ_MmqfjD4FA?c?{0&(NR-X%_iE67tm+7> zP$q`;N$}rtU{q?*c6#U`Ko_mJr`RvARK_tCUIR|qn%l13yknf4*-^iIiwDPw7URrh z@wwwYu{xJYVvTc;>*L<5WFxmQu>*H-f+FdNl8GFO1^Md?Dq@##%K}}Q|32a8%`sSv zo+s>gZPw_pv>4f8F&7G@#kla2pfAv<`~~u>wz)G4;bTZ zF5m0Zny+^(EFmI=WpA_u?GC|`>NL7|s3aS8S|N{|tSI--^(iA={5;>4Y=gWMoWfl| zh6kEc3x|<=WBf8-cpE2I(Qz&Fp8x*a zqfq0V^PNMx(n--2QfKlCo}gIZp)0|{735}al$OeLkpU?TW_BazVFP0+{h=I{IZ?|K zPUT5%(U05!M#tu@WX};`-o?#pOx`S{av0h@MVAI9E4o9A!^?rU4$Zqk1sF{|ruGJ% zpd$!6gpRB0-4n%wPSLW;;Bk48wm{KHM44>A5GOA_`I?~U7pZ&in^*R6nU5Uz{U8?V z01E?7L$~i{e`rv7BEfFon?Ms9DYT0Td)-1B_s75b$?x8itfJTz6p1uJ z=>Lm4z?iBPpMAw>Sr!!w+DL^1M^bK>n7lfQJxh@qD&pg}Z%B}5Gir)neoN8FHnOXx z?4*0i##zhxS76pGr4PtE>8L5EXB5rv2`zbJ)f5P)dDjQ#^Y?olAl-`6z!pI{T_4y7 z!Fwbp$0l#6d;xXQcKQ^97>Zxe0|g6lVw^>{I-i4y7O#(G(}BGcTPRbeodt&E1M>EH zD*{@G;YniwnOzOwFxFlS)KSlJGJsU{XWwua8Zq^L(bPYX7zf4_l=qGBWVcf6W?ark zd@4)w?x6ot3M~dhl~YgLvq2Be(D08l;x5TiHD(~6LZ_^qe5&V1f;&8(e}+AdaDvCA zwm+;`Yc4tZe3PvM2UWJ4KqiS|6Dbl$MI89*gGH;pr2h)|<#e+mTJa%}In>H>T;dfQ zXN|=ejBBKi}$|k-l{kysp8ufu$CcXdd3Q1o8B-%d;cCp z*1zfw$ia7REc*SycbcIrHkY}|+bFK&-J1(U;&YbIspK6O)VsHWjVqHM1@BZyors1N-WMFwY?bAVHlDJRWLgV%@^QP&~2{I6z z0b+l?@F56t0?4HH@S5v0hi8=cJ!423w-k@#&Xo`$9J$=tOR*qZkP9gudSG6vs((QZ zIYoEL637)r)RZi$`ezwmTmGe0QyR1=x0N0oGd0(vLsl-VnORJA_}vj!@r%NaO^=z{ zpsgWy=62ENz%R=OzibcqW%dCacYK-JBiOTxNs5e8RlH$DjPI{X7L!d5yyQJ-f`OeB zn?;fBR0MYXU>hTHDv)q>r7YDA>Fr{Jv!`~4W>0OIiVDMP$TDF) z@Z*()oaUE#*m)s5p24@vs2d;3&A3#Y@H{ik2$jY+FTYQcrjSCD5y_-jkT*-EA|4ZL zb%WQOd17MVMwgIm9ac)EU<18oB zG5Mo?bF{>0ItrcA&ydv)Jj=kOG9s|Lkz$i5l88($cjh8NJn)OTVmA@C*+&X>O`;_4 z9d0@FA?XBFP7jZ!b;vMy5ySUP#7|Dhn6ymg;$lt!Wr1f27?dyaIO^YCxH&8wxk{ACWjoUu?_ zeM*(iPY%AwBG249vWD5D?BYKNcq}iX+a+y+;;`+&Ft^if6L2+eRVRLV71TIxl4kiK z=iD-CX(Y9sMjE*eKRdnw93gMEiWv>8wmHTekN(kn#wq5{@oU~B?cA2W4qPq0%7m$* zk79u@yo-v!D(NLCJJCgBKrKyNDtbWQfL?bjkjB#dG`iaF!#8dRA;DO7@Ocpy(4cfh zT;SquGZ%M2{UcJ>>CULirG`huF`1z~p?i6qaH>(!4n*M{@aRZkwIAx64zaJE3Fx5P zp{o|O3gn4EuUjWO5YW!F3751pS;9?xB)zSP}HQwZUg6@yZC&@z%qikWHt0?-A|y zg4uO^&@)4#4R_DY-SSiOa!CtyhS|Z@i=g(jg>F?;L*fU6UwY{^rh);PS@bpSybv{i z)wq14=J(ojpOX*axJ#iQJ#C`@#b}e#ss3+~vkvU`-ZOD~uTw0jCO1$KT^{>f8l_Jb z45{Fh!umazsRr*CnN}`;<^QtxC2&n;_x@h-hU8$#MlkmdL6Ha(#2FStMV(-$uhZpq zUtasVzG;W{jcuo$PV={Kx}oDbDxiW2Y6!9kqJkib%Ibp4s9-@+L? zmxPf-qPZdAMPJ9yn0wDX_uP2TH^1{czvcVWs~*fM@R~Po-gT%XfVx@f%>1b8Db0}d zv+d>cYM@t04($N?AXAB)I(fy^{HSE+fHK1;S2Yl#yRJlnOzfhAS~&0%#|D&TOGyIL z#Xw*mc0;qyPToaGKHCwB9KOKPYY<#l>NV#`pLZIa8&oMetG*iz=|>e%v$=}bF(XY< z9XMYFUkm=Tu@rJaz7s{okfDbE{cuD6^4i4fBS{f*V+F}ZF8x_>E4_aOiqE10y_d-s zvbtuL6?t<~BB*oG{HbTL^YXv=Jn~=zklr_S?$K_6O?)!VwTxVv0%H~h+*Ts%u2RN zX7H?w?58tlS80soUud$sKvqE|fM&0`^*qQu)ok)>gp5JVEJ%|q5nXpdIdZE>;|h|& z#ldPAr=M~H$=F5e^-(G_NVLa)zMrJKF-R&bKvGDt+bOaY3WP$QSD6=9GAfF?;&*#G zRxXO?K_Ut&8<-rnOdbzGZy0U+$OF;R23LSY2)Gj~W^9bmVM>Z#g*1v~URd|0C<@s# zU19KZ6f=X^F@OFqW9HT6Kg|y|W2WZwt4m2Hx6Qcws`RAV0%64zTS$@ZRBV<8!PLvN z@v1{wp*+CJSGxVXpw1TLG$Ht!WL0#ZShw7#jWKy3SBuL7fz=B6enA5^!4Z=a+n9cz z%-LB`)(gkbt4{F`3UsYrtzIL_PqBi7lR)Nlti2|G|L8Z(u=zvhA3DfK&slAWk1fD) zfnuT9#2G3UstrjO1YV3h98ffEE&l)=C%~xxyufV8YD=U4>9<_I3F2Tr*0h9mXqSvj zhOmUp*#`DLzg}Asy_0`9U@bqM7cWWjPM54ErBq%}y5w*`8)P=^3fkp`InZ_TPZfER zlqsE>tGqP&oVW_=nKZIc;Rwp$;MGB63$znw06E|MTR$3VS9#QNAihN4N_A>4C z&fxN}o8ir2UG$Dw$^3>1wXimIL(RYPu)LRbm;Fo!edGq%i`$F>Nw4bQ*GtO;Z4fbR zh*-uv@@Rwl6nXr`Z`4cShI(lc?8=RdQnc=}G@r)%7DMlYf_~;OsU}AhYa@24A9*y% zFiU!2XqljgHa&JLjV}ET@5AIcIWeh*p#T1n2WH=)Agos7;P|W)@b&JIA~8@6t&=n{ zr^P8D81PlEGFc^!6`XKH2;~6d!p>l1S4<>lAW|@!UhcCj+HGxrw$V0BDDv3S!1p@5 z%?K?0;#W_T<--X$H(vd6Ed<;Kid|2E8~`OeiI1xnu)JVl$gW7-ob@v6d8O)w%q>{P zU>Qu}pOmzzT%j`!FA6xzr{RZjdaG^32;Sb)|Gaz2&(gmw<$3lFVjBg+$PH>@!&ef#%yQ#>-I@2uQf~UR8{7GSQ)gtz`d<22gbs@wV|mC*8a<=PAkfPPVi?@_+xj2vLr#lB zx6yM>RY=D7V^Cxe32(7&bPv$eaF+gni2(V@&QjPQxbR8Wh%YtEoE<0z{@%%+R5C$jBtv{EIs=rC+`tgAFB0Og7T(4{BqIr}X=d5_;@I%x4k= z4n_qqPmv-yVRKKsX5Pj8Ah&iTd2&;oc$i?GYVH=Il zGTpIk{0Jb)9cDuz_pD>bAvt#07bE^}Wx>|-&WAR!SnDc{{$N(_l%}A1U!;WU4e$5v z5m8)5Sg10Xb|M^%jql31yVTvqF;{awjzga7VnL2)E*=~(Yd^4ak`5X zV<#ZSaacU%PQj#?TdnswoHQO_ULs!a#s2ztvG4rp=WoCD-e)vvPkHF5Tml;0VRuD) zgVFzTF*;84P^rVsQIe!|<^oamq*0DQ#+P6(Dn{jn4SC4wabIqK5mlDHs1EOCkb~Ro zwLP*kJf2si=#%NvC$|bs=92Wu7;U>zb`t6#xl-t1Rcgw`e-K?Z)2Z)fV35g=5+IW61y&TPuGUbOeIpk{PV`Q7P<^R?6Ci@sf-w z=+?a{ulKtTYf94070O2KV!^|p0R>X48oZ16*yAuc&F_oCevMkB9z3XN_CHPzNowUa zw6TwLU(kR@UQoGsz$2bltAGkcp?8@9k2T{9>9k-ZxiYzrZ^Bk9gWst*%R`fHqjtrF zn_#Iu%S+{zi#xmrJj$ow-$|syyHQ&%&I^j4mFivOzYl!gH-j^b&lo!T?I~3}(=+io z10MU-i~I*X4oBVyAMi+d$-r&_v(I7gyr37urA!rneavDXGYo&xOx_}^UKq`JhsDyB zPO*B5Y{2>>2Hr8INShY~dn-W}y=K<&*?R)EXiMoNO*bSs{VjD?qZS)(zLX^>omoS# zri#Riy+=WmaTQ`FVdfM$_0&Or^XjEc>oQhc)EhSrNU%}#s#;wjDiC2V8O9+%^g*fg zI+HW?fp=$kE@bv61b(V)VyXmf3~tg#ZJFbakEc$n(=L8t&o`DR2Y276#x#)B7e^PxHL27bj%;E_9TE0o!g9Vw>6097V9Qhi;x#&higN)pwAf#CbbHa0uN6-1N>M0vnn+%n)atonq8Q9>Ikm6_`_|I zkFCo!xUCjyqcj6%pkDXIxdwWR^ZXuf3|1-^pCNUVN6l*S+3eUg(XQQme)e649@t(* zhSNdc3Cn)Xnq!@dsBmLF#70S&TJi0uw1{0n82Ow-&ks1}vtaxU=AL4e0PN{u23|hz zBh~zf9@?O)cBMgM=#<&3l005JEbn&R8S`HE*FR@gJet0IEt?d(v0LY~g(`87Vyh@p zPQ|X{%@IpViUTl(`pS_K*yDnH0W{g?dr`Z;HQu9FC3Q2ZG+Js~V zMU3vMJQq@vP1U?{n9A{OsnRk$eRQBWDXIzq33Y6_o(eQb@rOab&a6WPbKHQ{CYRmO6Tl22ipf@?=dvr#g2;Mq|lBI;yg&`x@+@!!Z zRV$6NG`A+E-#bopO{qKV>C7zfr@RCQ=HMx#Iqwp;IdI#c42g2y3cgv!cm#VaAd9$d zZrs=(lVKr|k|}mIMG~mkCRw$*OrR@Ku8fAvn_^%@$nY@+OWx%r1s3|99KCf1!N>*P z&I9kWM@qZ)?mQ1OF7iKp`!6K!g~0{5ZikrG6Df8jMLwir+kx<+Y*N8gSS-*(mrVmn zk8+`W^tUl~LgLwT=;R*94*b6hBXRG$u^q&Qd;b7GosT+6nOAm1uki>qk*%WDU#$OP zNzA1fy=oy43V?kT`7V-;B_2k}&T1z{th5m)17eurzVz`QJx3Be?Zz0f5j<^j&NM=m z53^Q!fZr}dN=gi4)m6%t2uu-8{od^|QvkqCc`N@2P^Q2V)b9hSfl+C*TKNyW!BD?X zqt{@8%yt?F$piul^OzUNejusn;KxC#D99we&m#|rj+)dXO^&G4G_eMDQSeHr%In}h z+%Yg&Ve1*E!(io^O|Km;Tx7ilxvka2=Bi>M>oRH>b@C=CkNCf_rr4=yJabB%-+`MR zEvb)RXRWTfEhKCxpeuzdg^xUpnczU~mf=&#yW-bETEa}l#ztG;u=~gF{HcHHh@5aS zCb)LY{^#9P(l-53(Ei|x$pg$nK?-j}U}f-mHbr$xk~E`JQw23Dv8*=sITRHSgyfCd zg@P(UAJC%bOaIJFiGj{VKKoLd*8cOIlV$|6Uqi3uW)|a1(5j!q&f)VBOt1rF(yc0~`C6}c0sVht= z_vT$6r(wnp9B}Ev4KqJc*DQM8BaHgU{^^zU z0@4B;d^He-n8ItHbX#}|!p!$g)^F-i(<l#_TGEep{&nU zEQyVP)fG{V(Q%>e(mZ2^L^8iteG8y-8Ca0&wP`=RByG@E(wk-LR0dM&lNYo+@J2`* zy(A(vI1Om`jzca-y|y_lf#0BYA*edGMBv^>m6(exQtQx09NZl&A7V_=9#qd)de9uL;x zzv&Wv=STD^mPPAY#86Ts5b681K{o|bx{3Pr`w!`a$ZR9@d3MXV!o!(4Rz}rf+#CGZ+lZmz^~1*e;e1g>ey4b z-^J-S>b}~{((k-^%{o8OM#^0h7&Wpqwe%?(B{A|nBbg^|B1a7jpG}jUsQ9e^XdS#D zsQ5+6*vodJ;<|61s2&JybWKb@tmN^MoPbszz3M(WBO7&C*og|)S-5P#j1wx{*Onl^ zoBXbIpvse$e)pAi#76K+k31D(mY|Z#R&XktNJVo;H$2vi8@upp17ogiwvFGrcV4|a zl3ECEtb5w1g@9b1joJ-?J@g~>89GJP&nyEP$CQv8uWbsw7<7rr_FOM3`<5QcDQ^me zM2(jkwa8zRK$*@;qc>|Xz~KTAqv!HcC*;lEChhP(Bg@eA(8#i6>g80yX&0yl_bSi$ zbSB8AznQn$n(L2?&DD+NjE$O2`$HB-EK**S;$9iat}7-thviHy6D$zrN|#bcp>k@8 zvR;)L@LjT(cMvH?XVogKPLP!y2Ozeai#*^ow9 z5*p7d=k)_U7LsQbX%m4icEF?G`-~(_x;`u=WWZzX)H_kvlmi~ig0tzJ%yw~Fa6E*y z42}dNC_ARLF{Myq0R#3MwG003k`%|^&CicI4z-<5k*i<_?)P45eB%CzIbQpxrUeb& z->5xDiquQSr3O~gZ2=F12EX0XarH*_N$Ym71e(8NNp#AT2BpDsJ+GC<=iu}D`K!Yk zwW}lxLyM;@GzVGPi?e4eOU}G(grRMP=gRjM?S9v6yfxgb8AooqvGI0=1qbN>#onh# z9~Ene*y(pFq@QV+xjixuxcTCFb>L_|<6rO+W+vCrD12J!7Lv;|Kt4!wn4T>Ypj?7H zkcbTtI$XR^W!9@Y6`LUkAYNix(DHf4+*h#B-X&Po1I@jUJWOSJYl2O+P9KIZ^nc{h z00rx66$QM4ubhw3wJ^Dm4V%pD`O=~}I)h+I@S5PR@Ls5xStl=1uJF=p`n~Z>;uqE( z0s6_LKx9vy55I@(H%DK54{QE^1-KF^hG z8l4?s>atauDmyBwmmiKi3HD0txJRnXljK0;!z%x>$N|NwXb6^QVoH>s%F4q!>6I+r ziBUYdeX7F&^KS4zW)25HJcG&K&_y>x(mvKHKY$TGG^w^VX{l@Lv$<#dLi8q;+PTjg+!B<~{^Kq3(j_;(nhpuX@!f zCE5xP2gD0NkjM+XuuokWR*l1#n8Gdz9t}Omp(V#2-DHG)Y%e*(1V_2=-Y4gO;zE94 z7EkAYRCJ%zbBo$?W8JOC!WG&^u`Lw2NX72->m!ZYQ&1?pUHDLNg;%LqPO5mw^W8;n z2()gm;&seeAUGItN!S%!Ll>z#W;}WC4gvPMtElzv_wFSF9v?HMnks>QLZkK*`DQjj z((QMI#te#6q8zU(-Vu5y@3yd-V6Op>9RfV>QeI=Rt%k0Z=XmK=w`E0>KUH4gU7fz2 z-aKhhaH8bX@H9!0`k6mChkh+iP{*G2W16u1t>WiBjxJTktqn_nzWAx3XF>QcK%9{DgjnQxe+D~H(PGr(A3O7K4`a2-L#NGJbUlj!Jm zPNH(G?E*l`X5geTKg%P##*u$o%+Dc; zt)@r?RCN$``(Y)8!OwbaWkju_Lu(kD^LF7+*Swwaz2vX2dRzbYrnd{eU-;Abx3k}F zfJ)t1-vKzFzH~{t01}x@hT{HN=@L|tiluA4djt;!S&?=s%p03BU_pJGN#&u59<-UqT(b2ZO)g4=fTVk$%_}o37kYw1_eBw_cn21yf6d7W)-|CpKxuVo2O|F5o!;`mq=r{>Z5a z=LYCqUu!;Py%Bd?dH@?1g01R&S={vfs^Wk&Ai2Z9vvM(JlQgP&nQl6RkHzC_g`2_3 zSTXrdcv9dk?;N0kcb&5Pcsd=!$m9E6b;JoHV~!nayC^rymSg|^wToo+xM2(l?i*|Z zYVXj<5E>fNqPmwY1M%{v@$O7aklj7@Cf%wFb{WUc=$-vgiT3P$a@P;y6+ zF43!6nDevCCjAQ%Z~pe@v44K^cYpY8tdwRw2t^GeB&c6B{b)&m872?ew zw*b|4iiME#EmSPhv1ar0__>igc}r&<22VsXgXb#*#SrUwT3ibS_!frVm{a*A1$bF0pt&_X1xT}4sUc;*|=GI}vm=l`bFaJ}3b@YbYDkwJM zF)}siR|2n5GSC;@r)#Fgg;uMZWIfTfqs0mPcp~gxSsvdqDo3_gmSM!m*rt!F-j|xi zh~gh#s3!?-yxMHD5DaM)yOAO(RP37IQ_6?RtU%pK|8#zwe=lE$gz+UaPXbSbE`=&k z*U(4E&QNT8H^B}0mTPoGejVt=XcR_F=Ar77@{4}`Fw*L@q%v8 zI(aA12VlN`e#8S`qoWGnt+nu0r+YC#44R%hgy$hWxcw(J&BPk_*o2K#|K zRVkcn9kgyE4*VFTl~so1PAOI#RG!o%PA!U=FB*lP?vwu(d*0=fce%cW(~sGGe~ZU| zylu-ovy>UEBW|~BX?mLv_Bti$WvV83O@1&7@4GCFk2p=Q^Lq46yjDr~ zduOTFgVS%(xB*3$q<40f=(u4S zk#fDT9W&+q3FdK?hpkZN(C24^Gz3glhIGE-7@iyfh&5oI zaqbYFxUI*({95lm+6FPc+6KyZ{=EDRYgdvDkIk*|`Uwz`8_^k#Hmz1+GSlvJD_n=+ zNqzKL(!!Jm=vA#gcRh6&&fUT^s+u7TJD$g77vb3ZuqNxG*SYr8hmJ=mwSq%l+3YKb*18JKpJq zE{1!@<~I8wfJ?+v26FNx+k&J~{Ub{k29@C-S5!0Z>r0zy-eAolX z4LK!Q7Xto(SL?MIWIt@=69P-T8njpC1$@jLDdZjY%#&Q5zB?#U@Tsy*W$ItAeX9aI zjwR7IC+y^(6})9?ZP4zYo<%!mX&Y{Oxnajz#dzX;xJNkDW*a-XzaAsEHmbO-Mq`LRJQ_Q$+cIVq;1-s z5{L0%dq}5Tv5kk}tUYHxu59>~dB^ik(-tMU!fnUnzV1~r-y-Y3mtwmq(nZB$kPPyo z19R}2@y%X!!lHmSW@7{}fmB1-y^$%vgo_7rT`DHv2-yOn9Ks=$t=HQ=CAYq zzGK2e=|AveDQur~o%B%xj)fjMRwbxbW1)h*z$I*~SYU|AqL)RVfYdqNR?#}%L8!)2 zVI-zVV7eH%2{%c#VS{~Ouv`AUiiVjt!c7(GkmNExEW@vAHuTNXY=Xxl5%wm(R-uWX zxYNjRp=$(T7$4Rz?vtU@0{`Ul+}3p(t9ie}(~P>B&sF}NEOcX+(q@aDa}vd_qDVXy z+ZmZFMWU>w#;EQlS&0%?r~jR@f_6t>Iz_^87=)8B8uR}D_B;&RVmW}8gWE00Q zT?}~I$-BsFRNjhmo)W_4)hAPgy#3A}bbtE!@7{dxH`2uv`vFDbEEEwA!#S-hCrFGf zT+`b=)+{L&E(twE9(iyYz>U|ZO%`$`fnrxsWEmB^klh|>3`;<>En3*}{uTW-6nrsB zr!yy-5%Kk{zkHi)ZzK-cYk`P-iUpH?HWho0^a1%=Y0xEM0zdENR$;sFhBrq0 z_w#p2G1#)7e+alhn`8zFkgS<|-HWb-EVfU?z2R%d7t*P+Pongyj43x=$KmI&I&2n5 zL#No8$1^wmY(8U#%eqw|=g3+&RyB$(0FX_wP}JA}WiDWoqdP;E$*=fbV@th{L6TAH z9K9+}gpo>mO(E~Y8LjjczmLRw!C;R;E;UiTQKd8Y24lPls)MdwfJQL&&YbBd#^SV4 zjI9%VzsK4P$A#+CjXi@l40P@ER_P8!qCkgrw~VYFrh@5@Jla9J3@wr3TKo)|IId6{ za;r-Z->0g7?q!xD@9WOIN0ts3wB*K1OSXk%Sx>Q`9j>NgE6F`_0{)K+-9>ttQpq{U zV9WyR%0VDZ>!%k4K2VOTeU@7SiA!hPYH-`K#Ha9Ato5el1(6#zwm9uH$_5CTtP^ey zJmX)4DI%^cBx*Gkk5BEB7LWUn*3q9|dwaIREG^O|{`U7I*^PA$pmrLP&zwcEVDCty zV$tx5jNoPqRkhk^(CTMeL7B|K+Mj29kAl_#Aw^Jq^PoVtjM)H&T5$0Jd)Ox5vPjn! z9kg%eE%v&PPh1EC+hSrEB(KbOBVG;wM&5S9Wvg7^Xt&97;aM;>VaoTD@HB)$Cf$%B1#NXwkcEA)$( zw5f1_YDvAyL<3x}LMtg`inTGhyfmO+-5K)$)~Iy;T1iuwbJUX&Vd^ouh8{TuX|^%- zr`><)ectH@F>D%HY|}z@edL4><^|40BIkS24*@e+)Te9`MC{+EF9zLI>bgDA6;v)x z2*hGPCS$9MO|Bz_g;Ovv%pkU5F`THmI4vdLxXLY_S*`4a0kk0eSqu;0BXQ>w+;gu3uvZ&Q- zg`h!sH~`!7B;B6zA^WEfwr!`g1C4iir-bZ=k)0~0FZKR@ zmRbD``e6B1vd4|LNoOon-6Irxh$7Wg?8WG7)7r>wdfBWVx@_`(ak=_KwvSXs@0ol( z^db~B*{IH$s8?N6Bq-y%d zVx=i8`O@=FAhu1qDk^biC7BPAZ22+ekhjsO-8~0H4eMl)V8;yEi3&SLa&8xHsCca) z7IMw&5_0 z)pqQrv>OXPZ-&SNe)>;Hxf?^|f(1lQQY?6fj!>~RQFo**=6Y+%%np93_aWJOSyr%K z)fe6p-l)xhmh-bSzhELp-san;IIFG_v?-7p8BEa7vnrrbi(DQlA+^H0;tYL~?g*~^;xb|wPH{=mf^U|q#&ELKF9}U_}O?lY)*$yqb4g$)C zuo=GD5zKJ{6!o7Te|y*34dAvL9vl7%3~;NE;ondh5jd8@+UYy;jS(PXOp6rS=O9x8 znf>eJm==*kcQ8PpyEiyLrdOy}eK@0?u9LSy5k{j++msJd5TL^w@oUbQQ{+4qyUpv4 zu>f*wcp}kl^SVW2upAVt!UYclvzPnK!-Sk_QU%F~2L+j$9(fm?#b4t)uie;HuNqM7 z@kD*u-SE|N(X-E*I!wRBH1m)uD5HvkklS;^XJgx?niT>DsU>0zR1wdjDA~^}t z@OMdfNztPQBxCA%b@F`PUD6FStq{3aDrui|5T4&jX9887t`@#PJg-Zh1}9G`q~!uY z4lmG#Z+!xYgAITB;BxkBW{Cad+UX5sFE@y}aim0}1;ma~Eab%=gd`%^1E$d%L=f0e zt*(>r)8Hvq9Y!c&G8txvxhcl_ zaO}n(&KItG-s=G@8kaeHEq@Wj?Ch9%Njjj&;h89#2U`!xE=RNq&xc+J!GZzZetojh znSHzG{IRCs*d%hi%eIghW*bg@^Q#B4UNP$>rBnaBglu)=Z8#9#4)HGTqu4zZ*+s=( z7M@kdi7F;%dcS*7R48dv{PA&1y>{ErSAPC@%)zf6{%YzskG|dmq#Vi2LV28Mg}~Gq zD8}iB(xoApv*(MB(3OfxMJA-AVZEcv!pzw%#OPF*xmi;#J`#P)X^sU4(DN8a!yR!% ziLLJ3&_jOuv)Dj0dh#FrpoAE>X|-;=aVodKS^>pExsp5++wwvla4X1>S{oTV(&$SV zNsdbos-Kt`ZPaDaRCpND8|z@18q{X6)XgX2<;-z%dXjDj|06M8!dy4!i!3Jr%oa`@ zZlC+wE1RuZf85qgV}s#m%bZKnMs1oToyz6y^xG49M73_(?rEvPZ82pMePDf9U(|!o zrHYn}YgU(p;?*PcY2SEWs%YuBdM(~N;E`9S!wt{MifcXOHCE9?Cy#fB$(;A-P2tNB`4r`-J(ty-*en zn6KpzgVMcwXsDe?ANA}D@6h5=yu`9!KOgVP2C6Va#4^VDy2X%rvUl=bsIm_EHjfXv zNwyz-ZdSLSS&yxMvt%wwapPUkZi^K!mtr9SI)jQuT@=eZoA~2Cg|OQ8$dzoKj>zzF z#a`g?&%%7<*i1YzjME#q&4kOS#fc_k_n*l_k6~;tZoK)0Vp2nt>Xj7x zAw`x@v5*&UqS1g4R5f%iU0Q6R%O3FBy4RKMw!vFIVWqcqrFgg9^V+BipF{5t#yY(P z>UxMw=v5Se@uM9U6Qj2VGwQJ^i(V9hE@X5JFZY>?6&XfBp4kJ4(bhi@Xtn_{oHXdG zzES;uYZH?sDV^CQtMO}+6;Cl%MqMDv0Uv)Sc#20|u-J^#wAFM5WPLc_KV2L%V( zj(_>aH(tbobjC&JE-5z_WH$00H_Y4~dENJaflt!jsdr%ZMs$nq#ivCV9bg^G@3z{A zjj;X8liMHw`yTo5XPg|zGUJkf~u=H}S6hiHrWP4vhg|kQob?w>hVh}cEle#@WoVHc1Tjf(8n9Srr#>akD zG1P#;9JA!{k32Ai5*6Mu_I`L8m<64(|FIiVHddhFn(YUU?Vh$F_48l<_?|TvDVN=j z`!Wpy+4WG9+Gi}Kne^$I7}NGCkYO}xD`Yj2l9`((o#wp-@dd`3?KKiOaS9x4Psq?K z+(0w*PS+lP<&AX@MT3&1X#Aha3T~lBZVZh)3utVl*cAM-vF(bB(pqu-jKR-F2wcmi z=R+VQ>Mm_`{ftfOYp)h763N-o3uxd54;N~&R%koaCH*h!;i-hM`Fsx%n zFMlQQqg+$op4LjAl%AAU@!FUpWb12pLiH*PE*m|0z;V84t?0OU6uwJbw)Qc#-En>VqAd3 zIZ=<_8P_9w=br1hBkgZnl(Yx@X@RxQVPmo0FFdA!f@<{O=f4)~r%a3OH$I&3c3U_9 z)-V4uWP({GZF=SGyClhtqlR``D4CfQ3lc+5#r8yJ`4}QhJJfP|)8qm^ay#N5)J9tA zJV}#2FbPg? zxl0Kd9+Lbqp%s${;=?ZaAJ8Ow z?NI3IRrRV{g6*P}FXc&6rl8{#LP66&Ni+>`l~?SEAx2d^teJ#PuU;V(G^pa;CkEql z7AdQEsq$3$eX{bUqJV8)CxjmhQe|8H=TAJQD5p&$8y#rf)4Qh^`1Z*A2$But&>6m4 z{RcvJ`WHo2@vw|zlWe2s{E50;oSNSnvmPYFfQoy>#Va$xJyP88yl#>p=T-A;<!f<{aIY@kUYFI#@heK1te~4NwMKAjAzs68?#M+50-4- zF0RxxYEMbZ!Y)b+qx-QlpvaOW zl4dX0i3#f|=7gouGslj3ts?QGnP#Qe`_7SMvfGXKMhzC)cn!rKq(~JNn?il`??=A) zXZ$$*#wNc7g0|_Gn0@Mw83T${(RYEq=m^;@IT*8I&SLKtaohB^=@(%2=p!3|*61>8 zU>^lxHS)^Il}PNds0>tp#wVZ+S0 z(W+W**ECp&ZI{o0@d4bX!ENogS?_cN{?05jKELgdKl%8DSy`5j9lEmgQ|vv8+@@md zydvOR+!mSc-OH3H8+m%o-x3%Ips$lx@t_oDBogVp4=pHxs(lv9ANpQJt74b9 zf|n7bSMA~z!Ud$~%b~9*^s2*xG6;jxnZ`XoyC2B1b9i@Ojh7gBdexQiy9D_EkbWUI zs8*chyF3I8Fh~$-dQP1jt&w;EHfH$vt6%_aVH#D`r9KEXNPflrH)QdHa@oo1gFim~yl1ED z18ZfuJRseXffA}$-(aK4HQs1Tue{F5mwP6muD9z#L@LGlU0v=3? z537{5BPsApB{R2Z?!wQk(DQI)G5@rt&ligqwhB`NuSTXxZ+TxQ_or_in(n7l#SG{)#Up076z!?c(sC7VGli@cTAJ$xdN)XZ1YNj8ZVXf zDarZS10FqsTeRu3NsOO{$wl;XpOczS&En|;9<8KDuxLWMW;N;ZFHk=|uT|WG^R(c7 zYQ4sJcy|oeHkXyBW0k{IZm_;zlD391t3vBmg`6X6$B|--T}n2^0@acM%EC?S)PMtp z*$gJ~YU#ykXJ3Qbr;ug?i8XMX0l|oUq)}U{*&xBz8hR)6YSbS02GQ?XE3S)n-m4$> zZ8$z3!w%=PMsD-rwqmBGbRf*y>g%>*#)chtksuLN)>Q3rwp(^kV5-yIMH@-`GNngn zv_Xj69?vRXuI4u1#bAAmF0QbBe5Y%VIKg(z=f1J$ymg$V+oHopsKC~9j5ULw z+XBP}wcrsZnSYvI>%H#f1%eAuCOJ!ky!I^&7Ignt-q@fx;)IK_QTxaJJ7v~C9{M~y zNd`PP4dBK`{*4xDR6NBlr^r$&7W2v4m|dh+QRdaCycIP{Vq|c720x!IRt9^H?q}ol z!0O+c0dZ|+!pr2^bCyWB#KHpiDaGEUNH0$NO15FbEeH}L6T$QzG(Ql8gR;8d-YErIGRU672l$*)C2G z+H&`)D<*GJ1CkuAf-b!!cXGaDC3a!!aN6TEeeDz(^((H_@v;4PP0JPRNbwWaj+0ba8Vo#kg=D zj6W5JC&TiG=JYdeli{|Ol=k23EBKRHr2MkxdokqJIFewYG|V5%4p8JinB1bTO}p&3 z%}cLpjW{0JsQpBk8JR9QsA=}^6Ssl4Wea_BLO0t>R`5R+rt@134)L2ODYQ z=>z+jW^ox*s;uzS^YeKH0UOng+FnJU3|k>t-Pn48XQZ&p;rpCtdw~G~J%}p_k;9F>i;)Z#m>w-05SmrQKgp?l0 zCD#i4a=Fr_UWM?-)q>SRB62~v#f>gLSv>xJc9dB>eRua(11WN2ji=5+<2gXFl@uwX zVk=3f=I-aOk~q=A&|C<6SWV7p%ffmDmm}|!OF>s7Pf1FqwhJE$Dj_z!QF}&)J@%>F z6>I%(u`L8rnPltPTkL>GC0*t<;L)zQ1C={Y`u7M2JT6KTCLSi2C!Y*W3vL(Uy*RGv z9s_%Zj2uE-aKS;OjTklDB;PNWog=_r}@-HzTDqywL?DV@CycqZ=vqKG`BcFdeCdv_r*fH#gyKsYuK@?i^yz__w zi){$dVRuCqcxCuB$#%Kr+Tb zME=liZ4t#n49qqvwnW+QlPExj*fJRj)Fc*q{pgOpN1fA=iD1#20vZ zqB@{cDVvlgL8BJlhW2|SH(+zvtted!(?{aN>g0RCwB_XKvv+cyEzq4iWt=9*eQoVG z^B$gi-V^eVK)`0i@z$1ZQz&lKe)CCY;nLFh; zI7$)%7xK!ycFtJ#uY~;AH-G!{*gwDdyFdIkR!XxTgrbHq=25fuD}AhU1)eLi7aREl zn6r=6UC+hZW4Z?IH-DPPGjhs-4yjoh3^nrPj5~Uval^vXM zjoc?gYuQ`n6utPxpDx=jBIr%ox4Zha)%~nP-RY;RFtffBB=2p7;GQ+E)yE5!}QH6l48OQ|ffJs|Aj4(ldQ)E7cNEp1|;aw##GLPX&cEkyiM&GAXh%OV-K$^*#@m=Diz&l?BMbyAM zPP%BUrM}$fkVdDMwae<{ZOn<#4*mktEBtUqqM(wj^{yawzPd(LGSlN%;a4X|eFZt! zF}E-m5*GTsD?>V8NdxBfrL$VZBS4&^NU}zjt#i9zGfXFLNIJb}eB77J3d-SMUbsOH zy0L=NYN4RiQ!IFcPExVylUpJA4Qp*8bzTm=Ti#8VD0`<|i^!poD-W;a^Gwp`6seV? z`~hyT1{SNrbcs7=fRCD)!Z0dC?ebe+842RGNavCj4^;-mZUc2s%TQlQSop^}+f z5ZVW6tOzUI9py~wiJsf<-5HKok*F_AlN;1dH>&hd_>=2G=xSIEf!i?PvE9T8R%5@R z*Wa*))r;Xs2lB?D#_7*M*a13nHDoDaPaf(}P0{S6kGMs2@$8ze}go3YNnoqtJ~&A%^p`3}p`i63^h zoY=@o`cPl~O8F1JXI|8PS@^4dQa4;kx%&p|kWVaRObf+cq(~DL8!uVyTNQXpIry2( zESyy*`!jDv`6Ywbm9=n5%MK@}VV@#`pIyRpwGko-LR4Xsy z)yt~_Dkh^(yk6TSzsode(}HnydHc5U*8|ML=plQ2 z8PU73hP~fH@NB18U{Tsa#qQu^rgyQTT5yeABONnZg=ayTdEXeW9NguPF{BqlYM^+@ z_3+f-@|aD28PYZ$o@LRuqc#O%q1WQ*fBNZF8z=PgT^CtCGQb!Gh;tBPH3~Ptto=&3 zuXRxQ3!*#Qh>b_vauForUgGay-XBo(hBs<2ftLG$0K)I4Wx_@*(YK9-#VWj0u}-qn z4_kHz?bDo6R@k+;Sj`5vQ%5ny?J3-Q_x;DZ567Ct%EBd~hsdKBCXaHHg%C=h*cB96 zM#b**Ylm$;MC18%dsc$&Fi$kkg^Q2v8LfYom{PFOk3r7;lpOubD# z)O9n@ElFW)ONQz_KeOnlf3svRN#VwZ=f)eS-4>!Fmtr9~AOrX|)kVr?Wx8bWvtD~j zxgnxM+n_zm0}9{C9o}VO`B8&c@A>BjVpX5Pa|5_0`e*gIuz>Uk(_dVOtbe48k%XGD za`hY6camH;)?RBYX1|nTODIwVb**J_0oBswykdn(i`^y3lH6jB2AvZ(Y7M+gi~*4F z)-`5@SFs|Ow-OYMTwW%9=9R;wN?8pmNZG`6O%0t1(WBjTqxMtp``&M@gyg16kk)@F zH<9SFu@63?UtnoOx0JISn7&6xr`b-kh#5V<&CmZ+}mc-55JX7TC$6*v%A4 zqhgOmw=0$f7fWjuw*LP{#as^WVraZtIczhA9)s#~zZPokFo- zB3eVmE{onSZj!BI=f*?Ptt`K8S%WekV(u_%uSAK#cm~ft*$8}-!_UgoTDc+VX(KuJ z+Iw@9zp*Ye?zUz&8zsflBo*Q`Nx68x@RD?sIw|lLJs+w`e5%L=`mt5f*%7D3K$W{I z=uD&`!c=|SsL)m_>g4FPY-DTYHT3N%RY2|mrdjYA#Pf7eklyig*H$XhO<_y!~6l8D`En_Esl3yEb7A@;?6b4fA!ibi73(<0zVh>QH z5^71)nHq@GXLB*F2hT>VpBpEqnuT*4(FYTu55? zspF$-({Tv80Cq!>=r9D2{;Xw#rx{2!pR4>kS@@i}lt7&svJxdx>?(@HQ?U<~*Iv~J zZWkj{QIWcb-1$<*caHui_#ovuqnjO#Gh^@MJ3NaH-e_Cqv}OIJVUigZnw85c$R=*` z#f`Vp`z!#njbfo5ZVt#7VXkWJN&=W6bf{gX;$@nJ`WnaY%O=Hcyd^qqfq;V)3wfF4uq9F;qaB7?4e}!b_gF}AfbL_zUPZfNph;-o6kXx^!1}O!K{2?KZ>aMHZH)1*swv6L z3SZ3AtOi{;O=5_^bq5RIz!HRUTV0SQS06SqwRpt9ut15`$p~O#Tf_|GQygVoy@!xT#FmR`vEsU7;jpvMy)p0gWD*SmS2ZtB|z}kMyb{@ovrB zxpR#@>g32}R`7ZS>5-=sbQ~b1A&!tIfVm7rlw9f zNt$GPU%}3hEXULk9Yzl4NjmAJUb_8I)T9)4Z+Vif(tYY|k_2^_^C|per^Y)HlvoKH zJ5TKc4ma_Wu;l8l=bbhYX&Vw>n!8(ea1y$W()nHTTcfY>!tB@u3HxTzsl7NsVr-Y^ z@87g8HSD%_8XLufcPJ7Cm<`!Sj+6AsNE~sWRMM-WkNLJog2g3qYRA+zrW7Pgk6*n~ zXAqRm?3mgrY?#pPxm0w^8)ShR@5v8oj~tnf^0aBR3oTC*4Nn_<1lKsBWlU;N-5Bdq z!(8+iH)bQtv{=H{QtTRvBvP?O;v(FtXM%RxnXDb&N zjQ|y#Z@nC_uAFDqR(^lVdMvEz~+4u|G z`-+{bG3T#m>jt}Dr>y~8)XwQw>ApXz<9GJH`+v)dyN1a{hHlVM69s<(t`Et8B2R*w z<93h~Mt?!GxSL*~OonuUo8i?{n!%l9OybX>4H0d;>JX^KXNYL0AM%m*P*>-F6^7dG zt*cdBn2;w~4|#NH^asIvLXBKVJ9!s*3*~h*q%<|j`p5|%r1~~_B2Zg);KyY5@8M}H z)ibYg|Iw`c>1(fz#9ZgTKM`WfTnEuo2mvD*2r{uIPH%=BHE;lA(-j0ZKDWK=A<_yf z;Z*VJAzuxfVMmlFDN+QpIJ@$&XrSy^t!Q8Wt9ch^ao?vxUu9R6?>kYKLJ_CG5GrLcNB}p89X=nE*`hZ zw-flI&#KeCo5Svkd+5O~+5A)>7}~7)fE4jkLTZHtv%CDa_{7iZrhCEq($Cx#Lyb@< zp;RfWq&NB>CUMgr%7#^F0?Sb9sjZJJi?I4;8HthE>m7 zFS{NdA9hbv1hp8^xW|;9=K5_g{YP?DNiD zmdvb*0w#~Jc;2p{8oE|kE{+rJfpGqK-T}}_U^`p{K}~6)F4@%Bc*$Hu2F$vRQJmk* zNo0)q{cn%W5}7sNcZQfvyw-Nb$j3T(w>G@z`E*y02E9M5#{l68`-AR`VSF$C_4nDev zSNVVq<^F56P=~KcRy|2)h*$|^L}&a9UK%YN+38q(tVTOuqSzSmn@Zo4&%3MevgD$q zSo-dVq+GRZmO=VQ(|Nu2dT6bpLwn;JYroZ?#S}8!DGzq))HJc%eA^V2B%7|0K!g+h zP`cK;*kx&DoF&O)ESQdL%-=|xIY|=gzuwM#JKVg`qz7iMCEH(^2+NNwmZfrv1sm=j zBwMKRZDTqk^_n$rY=%_#GeErnY<7aR{+*f@pad$4Du(Jbxxfv72O{zCdTQ`p$WFud zM(w@mO0eFWdM}>y4yrmhz@d;PX6U7j>ZRlelpw{c&Jjr?hMP6~oV2u^`gqAoYtASx zn;SRIF4}1UpiGK|-Hskp0DuubB3^kn3LQfWL!0=>(Ie)4h8cyj%piMnJKM zg)q?JgPv1#{>DO9b*i zRVo&;xp(b&k;I^7fy<=^xJM9>40fAZvc+7i2lH!RCzql^m#5(+OUca=JL6m$_i}1>}kP zy}KYYK3`ThWiR*y(&$3Zx+$g%)}%lvJnm|c*$F8ZX%@D@ul?Ki%YSKBTDN&kf0A3y83Ef3emGqT{auG!otJ8JNT7O zc zZF+`Y)k3l(GJIe#NPvY$LAILc{AA;8_;gflo^5gA4i_u*IC%OjT-la1!)%?#6#aVk z6>Am6#@e0(QvzjnZ-y6XkddN?z9_9z{%dW2JRjWNJa{{a*L$(Q{$1=lfBN~`Z@nj7 zOtBwOByRM)%a6ag#Zi3NtBR5M3Eg->vO)cX5yHleO08#xbVn$ZXxu1ln3*O?_9_zq z%O&O_7DaZBQjtC5TyPaG78B^glY5?DJ^#E{ly4QHN3arj1Caw8qfipT?UKX0M;1oZ z3~9aX{kU~h#2nX2#|6ZNbN$)Soa7pH;e7MN*Ujq6qBY<80jYRl>@3X|%FQW?1#jU| za1KI_>Ndzm_1>zk3`Ty8(_#=I^*pqKB+OVU!looiX(UvFnUE^e&8N!gMZQZ#kJJXq z-^*#;GG+tR_=I>Uh~Y)&-$6ka-6b!Xs_XZ8;0xn^Bt8gvLyy#O*zdDfql*j8lO81L zK5eR`Kpb9|rTN%vo3xky1d{HILBcwd?u2Cmxa`9m!JmUl3@7O3ee=-IUN$4jL!#MD zvfNk!K45{7T@(ur!5v7KU?OPg_Dp7q#oHo5LoJ?CJf(UP@~RlA*Q5&o@f z3R)1H%%JiMZbFka4XXwxGdpAE&8Glrr85tbQlojdoW?tnne}s=C#0N=gwG_&NDnwk zqA}l&Y5kUUPz#rxwHp&b*~pg1{E_52m!#SJT@X-tIIAs-r zg2TSs`0HhVeJI5;M^yoh+AThKR64UbXrp@f8@j?d7p0wwu3+b($#T|5j4jLIoOZTm z%iAI!vy}NoGkJ@wa$_m8!$QiWQ!MP+Hc+vLLdwGSOx_c)E)0X5;-MJOHeUms5^`G^ z&r1v5!8h=_73;#*ljMl{84m>*XQEgAZQkCPcU(2upUsY=v;NrtE;`^OQK&27my4`f z+TC>`>$UA)8~F17XYXAAno7^~agTUH@?pr0K+YLZkq8sSk&B@M zHqJ~t)634ZyVLEz-L|{kwzAuGyY05qbhe$g3wT8ZK|uvHfN~KO6i^Wr1O>%AFrZPq zB8bZ11-u|C{NFc;qX|TFAmJCM+o_VA%NNY|ocDXb_wqb{NTV{p-!)ec)a!apFMp+9 z550;sMd~&9_qo6l`gU05)NFOFARpAJZjf_9SZG|;Yp@0~$+H?X>h97l@+#Fb z;P3AXU&`c9PUlra$Hyjhm9R&E7UnWUZ}aI>5caJRWb;2~)_(PvpjM#cX9@PrYJ#}z zydJ^xgZ5pdIDqL@n;ctlV$*fj!Q^#jme=bpCbLD95oMr2Mzhic^NzGm+xW_R8Vj=|yL|Aa&Y`3{qXLCZ(>t<@tb<~hp#Jzj{>$q3 zEQ9)7Ha<=q!L^c`I5x34ygBUL94w;5pZEuPQt*HFH5;#(KP~`z#ZY!V{p;|{AHVLJ zF^zI8cf&Xem<&-6&E{03M?3FXb1r4uSWb7I1`?vGfxL1?k8Q)-+PaYi$ph3IQPB+iq zpt>=0ha!ob;f>9@G6Hzm2>>r2h7&xfTiaE;Cz$p;nvUOUNi4UdkP~|(nP$78WQt9s z$Z{%r(VI)vK;_aXM;ZetckwzmryWGU#|AWJQ)e53%x)QzH{bnAUXaFwiH^|8>0~cA zOgM3P{;U}$j!|qqMe3+%k3Ar(3o3V0E2plQSmuk2=!fY=6Li!k+0-X%sRi%m zzO#4&iZZ7M+=bG9y~fZh(gru-LCjj?;c6f*tr29>Wxg%%=r!g1s05e4d<1}c)C}mQ=^QYPWb?wB5e)t z6x6Pu%*;atWEmP@Z7U1j0{1r2S%PLIR>tJg6_JS&G-fD-)z4qYt59iA2Y1L&GyuwN z@K|X;5~(Fku5qB#h#R{`Pb|za{Ep2^XG1il-F7VHT0`5-5M%l>R85=FfXpEPQA2FiKtQ)VH zp6?Y|isOArV!&2p^R}qA%CMO}M{TI(#STKKx{JiX-s&;$L9-b!BbxS+-!%32r~YD@ z8gN=~$4aojOwk5n4IN=P%@@So9yPvcfewB8t$p3LOtQ6m9eLckRTwT?^zh@q_j5D3 zG=I{b{+KKsE+6g0i%Oo^qOy)+Qz^0vWx`RMF3ktIRq`h`P>C+XoFtmifCbU#C=z{@{`3pv zIGA%#Mx5~!Yq1a38a{pOwO4$=yJprdNOL}B=U5yRdp;7 zxU3z|))3RrFCKPByx?X)ay!5KU+0YaNXdnBS4Y#1YVVaZFAy8NdH8+E|5Yoyna1tKX2+ zFU>jwrIIhKGr1HCo7qe%8YK=F33}wv9)<{S;*KUi`&sv9Qqw8wc@LK@|n z^rKnD;z9|^q&9#MbD=S|?^tM!6jcTT^db!G{JMMJhhP6M6HeYKs+&Y^I&sU!QnT=C zKgB+z$UUQnfn0!%!Ie|An;^YY7TzYz3eBF{Cd5+NT(zD}4#z-gLIhM!8rdNmfG53E z(awUPSSphhdL42AKOMqOg|^=vnZOD~gD%*ruT#cM%3uuSO@)#@Z|!+Y zr#w8j-#sR(6#niNE}O006ke@Z;dPyt6>9ugBtb|HuO)lJfNG)5qZzoPI_Ng~G6Z6~ zAsKZ^cvk>DY;}QKX>3o+QMWT7a|R5$+QV?Z0aBA4a8((30)0ps%9tnv&lB=2qf-DAbs!hONBvk9^`T>0krR$EdeJFP=(h4eU| zMw&rnKG0$=8f}k}N&SuxyHfduKz7^>t;<-}01tlr#6E)893%1m`SbzR8C9OEwpuoz zSmD{rzbj25#o}H$+=!;FDj3h6C+!S&Gq3boHXBal(`BSnJ?a$os8%Zr-`VV~wYj$O zI5+b$?N3F23^SP*vFqGTWIH$W;>6W4ARzyO4^&ODl@!@SMaNOGqCTL-&ZN%(7sCZ| zQM5u(LG_8VCA}nD(hS{cPzT{x3}u?#KCx`pw5G@g;Trdfh>c*2y2YJgI{<_W-ERmV zu)X2fRnSeBzx`C2Dd`JOzkpcnJ1ooKbmwnYk&)Eoyp*K4CNNReH=VZn4S-6y?VBWXS@U!G)yKU zct*VtsHak>{O~`-uZ7ym3iU9Qtfwyz=VW@vwRC4rnS1Ze&TFR?dJQdW_)r!M(yW*a#i>aVe z+}jm*rT1h@11qTe^4{=v#Tc$I+`wUnS+{WrZs7R+cfa**%l`PGwMEFfEzbduNj!u5wI}d^qOt_ zZt)#OF`draBi=u;iQXbB7MBH9`xZ>gb6qrHkNB4Na%S-ahu~~9_Ixq zw#w6lnD5nWI73!XASGsB^(Kmi-Q@--p5LWxH4Q%wC z*W|y3pDmf!tl9A6yYKYOZP7Ty3=DzP2!K0e3`ZQ`2He}T`b3r`lUyQeP8^)IQcRil z%}daw*e_ovixoA9<0Gz1JB8SpSO~`(_-Dn1o<}3b20^CT=K>^00LI#wZofRU*|HCD z=(fg*qi9xo6FcZ;Ws|s@z9mm!pN5q1wAeUSz-yK5RGph{lnmNWioD9y-E-51>QLX9*YHN4J{>%4gR2Kh#M0kLbCjSD=U`5RV$ zTAS+;9Jg*~_FwV-re(C=N|+mapLEJ9s6yNc5v@W=D^OWhJ$@1MAC{nS+P$z9v_2qJK*0lCJJ2hb6>99x`9HLpJqVo8I__h7gU<&h z@ASs3n%m?Ew_UFj$J9E_eA%-UdzvDjfqE`;n4k&+GC*M+lA)}iKx90GYs+p|U_PKR zyK=k; z@K`fca5gL3A(z;$ECdR=cBY8m9d=UHq-s{?(ff&ZzZlE$kTo@HA_vMK+w7**6=?(~ z%m8Qg&7HLKsjoZ7V7ZZI165V^8YDsJp|{Ul_qALasW$YQE8kfwNtcx=c9J8aoN0Ou zQeuCuS{#}#taLlZo`X^q$2zj3n0l+zW{(q`j1cu%wEG|awitb4#d>AY>;-JBsAEdA zPo+<%PetU$NXKGj_<6MnLBrqb(4%g;yQ^RKoH14fv*bEP|+++RmJGZ3xCz_GV=Qo2!0I=Y<{!EcR<>%6uy#jjL0V z07ofNoFdzzT3KMyR<;8jfjv-U6Dqha9CpA9C%ZEK{dN9@tjW1?fBfJaS?k21x1DD0 zN2aL@&7A-72Ps@kCr zWmnJx=7QU4uf7mRd&i?fiX9MPZ3u2i`RmbNFPEB-qWIshHItPSNRiouWK!&Uill-3 zpyGZ=Ixj_1piB}NNrtf`s7ze0K0T=nDp*UrcldVD<@{d7{@ICwGO=UZDW3oHM=%l3 zZ+74-ZWH0OVf;V;^cRaQOAMTL_iUw{01FvYnf&Qb-19=)fk`J`4y-%u8qmv50==@1 zF)Oq;q0M&0IuM@249yL(tv{FMS~{;z3k)j+lPIT}KmDS*AsEAdI^|s^fgLNmWj+Bm zVZ!{DF?q)g6TFZXt>tnuG;r_4u@ft6A1c~5xGspoZd8n;v@`Y6Iss6V_RxD(C#47Z zo22c$%^p<{QN|Pb(4%D!Z}gQjY!hlM+|^eOYKIG~qvpr*>T#AjbX>B2&KqVXR#fYH z!KS1#opb`*9a1B!3|`}&>9-&Hys*$=-{ch&YoPNTdDfAQec!C}kPXVGQ-iKCxYgND z8svBmTZi!Y{@KVD?XBtoh3)f9(poPg*pUZX4hJe~?7qK#@+ae6fBfWAV zrs*J{oexB@VmdNIdZ^jPLtuhYrpHQ{*#ve>}%z_Sn+XDp+Sp8etT6Vd~G*# zR+b0Bw~R4BpxAFvK!hZjhEC_0cTW0z0AN0K0w*&S4^h7{l~pQFHII~FWXHo|A< zQZph@*nX4dKVz~j8NHl#BW#7EYl~U&# z|1{w8>2=c2w9jgzPbg2S8eOy1oS?vEdWK%h4M;=pchr$T{yH?;l26BJmm(_+Iosd4 z1O+Bpyu{!*kTY!%c8fQt+L_D#z5M*JehArD)7v61Nw)x9#n??axA`7B`cY1XgZkF8 zs`)b0awNVh{xwN@%}9!i&6cPvirqi~V{J5wC=`VOYaqBS8@$0I#m)M9VN0a849;9( zVnndE00Vw1b35#bJ6@;7syD~Oha+sBEijI<4quFG9X76ASF2z5zR_5(OEcp9+7->3 zyYHR)X^Up3KPb?J=6WN|Rto5{r}GMUr9p85BxTy+oi5x=A0gX3RtNXESGqL_``jP- z)H=5NY=p2dz~hTKbifPiuo%Xn!J6bU8d#CduRx{gWHOBoa@NqR%CX_c)D=pj|Y{jv^W<&gqBj=?t#Hc%?uM`DTDfqgdFgtu`uL z>NTKM_3F;dPo&Z8~Z_+qiM-_*Zyq|Vlx z+rDGM#_oUe{|Pz94I55uEN+-#qm^Ry6lsAb9vaO>rQ7P@I%%9J>CI(tY7azI2JZ$o z^?;;p$f;c!RBMobN{6?TNr zutRk(?`)ZDT9`iE@{9i<>zz39v&U?y*-Ehxn9ZZ2W2mc484r6!u{Ctzq$QycH#@0* zBJJ{A4n+2_jVh!}PT zM|qHwQ5o0y`!nnm6I|2@@%zaJZhnjt7e($h16L8n7E&aiiaswsFaApvNYq|sZh>IO z5mCGxdx8gDKJ#pkLM%0U=0U||wX9pdW;XQt=+sF1XFNFQ(kMT!O!r#hd0JI0E*7kx z<|wLsJD_WnI2!);*sS+MgH3pePwm-Gik#TI95X{eCB=fhDW{@S_zNc%iV{LQ=}qdT znrq5BH!QTf#-G=#&@84&Nv4^Liq(`5-hk+7-zThMB99no4xtr&N zJFH1Y0v)Zfbxm8sYh`=L{;67QZbMcWZQ8WKkjg2W_!W^_qd@j((PMQwM++;f1CR2= zkK?+pTE_UDHlJ)I`B3>rr@A*>r`$7H=eOQ)QO;Fv-7889sf$dU2fXs@LYA}l%m2}spQ2b$*EvE(BR>r)OeO~J^gm4BkJ#j)&7`-x>68`f6+Si7GDMZZ zoeG`u4y1aaH>-|Uufo$y+|G*O?G8n)55C1~bJeK~KI+&m-Ys5Ap~s0;t=c25eIZ%A zT7rZ(385LG0}Os9ZgcRq0cM3jo5ic)t?)!{n^Vx8gANAgNgmC_J$s^U`8oXbSU<{X z;+!{l>i(l2$5~S5e@Wb%&1MfqI>oM~ND799b;>Tm>R^N1hPEsm-4z`)I`5x7=#t`> zz@nvsEqkjh>5cjJTBNMQ$Z%!^7=Oix_Bnm-+K($HGGL^ z;}p;|tl^=3I|^a^KCd!`wm{ty`5?Ruq*(H&XLuzGv`KFs1eVFM=ye_e5YV>SXSQE5 zigjaGmh$xh%T6FG%{9n9lA{JKHt|*9Be}>!@!IOK<`XddBW?PcU-ET0>Xh}bd6=Ri z=iOLd|6S@tKa-cTb?^tfNY(@b0y!^KJ8h>}=pHYGazuU~IXL&eVnI*|{B6`P=me&# zZ9#eJ8#AAV+z-JvP^7;@w9M-XNK;)3oj2ez4{KeE6$|C_G6OQ{wQt4=<`pZF!{=qv zw*pGT8tH|hoE?pqPpVCjvBl4@^TWysGR~V&IaMBZ^dC%+$)fuIj+}hW%56V4!}mps zZKa5with46stoN_??!pQ$7PVZ%~4kcXG(IV+D5NhX&mY2cPf%x2VGhsi^Fo%djb|t z+&>%q?+#fLhkmw=<{a-gO^z`4rl)7l-PUaiY%!3q8Pv?RU=z)su|SQ$DS-0x32h{#m<| z?C@=)+hI@lm_`yg91pSv7y*XfBp}(=De+JayffA|;5!>I@shc5C|q9hxvx6K$#{-? z_q!iuj;5E$c^7O*VZOLJucO#himaldaZ?0MBfpv_EDSvE=fxLXVRrxW?b$-3hYmNLr%JRd4XFkaSFGhjIWd zQY__qZw1YM+M*7C*m0=8@ztO*O+dGjwNDtd1fXRCKGXhVFxwX6%x_uP7owxPeK;zL3Pg zR#`<@pUhBIx!j{p-5It)wcKNVzk6*c{&Dy^V>{Hp@J9_V>d{de2s@1``p5#nl4MoRG zsYcj5uHY{U*x}nC2cbyl8NMi89bD;#*XxuRfGG<{vI~rVo(Gl!6xPV2jd$1Z_Dzq8 z!j_zN;B3WYGDoL%lPd1bRHo6>m@okdc*g}`J*$zKyj7MZLCEyG#|uH_oIt}) zFSBnJDxB}~(J8Y+wOEzk28fJ{ zUOB%*h;OejcX_5u(??2{YR^YvcuPA3?KT~#x|%n0ylFjKxFo0!!jDedi3=*NFUCn$ zQtS$f#6uUF_q|yc#8A@R&ZPM1l=HC)=KqD~V$&SkG%NPKh@07K{mGV-mc%Sp_)l`x z4Nw;i&3>rSWF%Yr^pbP~iq(l%`n3TQSBI=smM-g)jV^%6{X}s^5+DHQ9@&?6G*Fn?UxNJ(k533lg=PajR*_v|tP0enB4ps`#n&3@?v{64g4d zCUK5**K}nRwxFK8)V#o+WjFBb&Jx(GL;C0T--(LNDA{@3fVFWHyLB zOukIVk<0X6*uf3JUHfNq5SAMbInR&T4cjl^{Q2l3f5i>k!O1NTy`Sy5f3|BrN#ho! zbK(jqAf0)kE~9{AAr-UX}W9OVa%8d)O;&r9zN51usF(tB~*M6(sQsu3y& zI%JpWEJNy{m!Cv#%!E8MX#T?1U(d#gu0ux%^swuc+uwB5{91m=oafte00+J2WJ$)2 z`}t2*vrU#HllhlIvU39HjlTE@)=}&MitMMN|9q9fnkc=dnqEAs7N`pL`ZrJB<<=!Q zAzVe4f)K(TvNe25#AAZ*iriAvcSx*P<&ov!fXVo~|M)k(<(M>Gp>3YGyzvAVn zCX&;1Id7ce;rE+%DxTqJ@~?hFQk}S~vCItMxfBbD&rB+MK~T}$KkS{`9JWua*IW!+ zA*hgUrtitt`)zhTLvIZ~BPj5>D@}W&Bw)QtuPK4FcrU+{mm0byqMoi+l&VID&sTzI zVWY6jHy2#lThb&*O0-WyhXm<>W6ifehZhFt`qyX zR=gbCLZ$Kx=4cTf2l?@g*4UJGomU=-tf-Z4OL=8Y~h+KB+-f2o2_Q+jgDe7D6)=k~6&pMkPXK8$dUb$fO~kXQ;ymSJ(C>l2n{`t$Ofv%YtBXgx|J?+TLm%Gq zBkdDN-1rwA+j|symm;^Q=w^9uNH2p4!9q!)D4$Le+;;No4`XUthCx>H9_qu!! z6td1cBkmMtO0q&9l5X+51Qz(T_-*1kAk^0@lRc{7+Q;-f!?{#ui5uprwfI7>#0!zP zSf^e}9U&#&P%MjE8vNLUe2fVX1jbB)hRDh(`0y`!^C*N%PVw87pG{AZ;JT8gS>p!x z+!Ho=UJcPJJ3xFHqbgU_+QT66kjf-UmUwRCBbVyDBfcvHg`#HVye9gtyMcc{PP9bS z1@(N0pApRNE8Jf@oaQ>a&THFm#)plwtjQhf-#D)yZ>{q7I0y%y6U2!&LAO0J1YwPT zGP8`JsBku^h7#N|#n_l9GzIWKE;u&cF*bi&w^G0Sdh0i%O%~~4@E;YV&54(+7&BAZ zL$S9h@;Qd#9?)laM>SoOPa@V+yth!jaHg9luy>171hxJtJg!3tRYN+kUd;z5ZBOhBWK zK%21T{Y%otf_4ZmpxG^-(E*JnJ2YoVwxmH(B~BL^KRMB}%>zHP(hVC9x?%s81-c_y zl6pwB6w?C=l$bA+BnEF&)X4JG+4M@LL2(zlZ1J~k0y{bkbh!Y`4kpmb5JN$09bm&P zgWoIm`pB}PhKm>N#FolRVbB6Wzif*LCZBEfULdNLU6LLMsuA3AixE|c_o^06TKlFB z*fgsEA_bzX&}*6{6S8=XycS4Fqy}Y6`h%(!E4BDX8Nm z@|I2*1}H@!Xt& zX>sh?N$COk{M?b2wAk{+4!MRC0*2h}xC7Sh%+G&$aI=l=v`Z(%u~ou;S=#j4sRJ&J z;tFZc;zAs2*?9=EoS5WC=C|1j0$7BBqmq!lESTQ1vq zTdYW>S=?tE#X=ycfQmM@Pj}LJ-^9Is4$~o9xs}Svb;H#fZAaonS;W>6~d;>JCh14#|3QL5vlX=#bD*^1z9emMkz%o6)wy z_=2%sY{1tkF}xHb0**4Y7E9d>EI`IL$oh@#68WHNjRa<6VlYNF>9#-aBfXN-LOA{d z<^4FzR_w^M};D7qbHK>_J+60Iy`O? zQ&ln+Dnn@$nwJK}mEZ|vzKw+ zr?2+Gg@3f%%Q)@YVueWK5_885`sUi`Uj7z(U{*IsJ14S7>6;?Ze+hkdY%HN`8)i)w z!OpmR*>IdNJ5DA4uxhT!UVK>h{q1BQx4n!LM_bRCnZ4r_dxRo~fKO`HdEQxZBi-)0 zL;T|FW{9diR2}5s{l>1jEt*tD=T|ztUJ>hs*I}{dc;y5> z^Hs^FcP|6j%gzm4yH*R1A$2PbXMHrQ^3y%UF35S5;2JoBxwfH({flgwb$hZJlyCm5mbo}1Zs`@<#ex0 z!Nxm`q|J#24OrMWr7}negv4R!@{|T75$wJ~jjIs@!Z<08gT>1x{nh8V*`$HbazFpN z2}E-%gq@^jI3_zM4zOJ^16Cu&0vp*eqi}SGqJuW>i?+^YYo{V`L{(-fi}{5(3O2VLElm+>H=k ziTAI!Pf?2_W@-czGvpR-Gv&0|rC-F-36`ACPHT=b&6cxdicO@*aw-}ZCk*BHM4(_+ z2k>E5PI+h_^bMKzmmjbRtS=wJA!j*(b^LcWyUkRY+@9%~|C>zmhI4zIc*9j|hLc?s z3tcPQK`mfPU0{-+>)lIIB%+2=24O|SM(C5pEmfwZd3Iv(xu8Raa=co8vj$%$^6q$J zd0s8QMYE11M(H&@5qG>%x9u^(`>|CEJZ9ljLAtu!-tj7%z%-njGUS+jF=ZW05r6&g zyZ>zh&eHhANYXtVaGW#h7AIx)CdToqhKs)Fx9A!uW8f=5+wE7JjW8F|=&S>jgW zHt13j3Hx?1U+|x*5~web_gQ<@E0^9+T4g9_JLrPHNE4kUKt8W5NP?wDZg;D6%k|g} zw{mT0QAJZci;fPv~_G4e|9WX|*U1#G6z^BM$r%?agJZ~^g74x6nJn2}LnXat{T`z@cO z1jSnR?hZ|3Iq}ZkN~3Wukjs~ab;vU5dQv5<6qHMMN)JRX5?u1ng{^%BEOb|y1Cc4} zT&SK%n|6^mph%ak3Pu{h_{dn%C4b%_V!s9doHKoqPeoWRq`3A6 z?)1fR4@4S?Y|(xsL}Wy1pZ{;!j1o!O)DgjI6lk&8qS%g+=T};{IG@*S`$+K2RZ9DZ zda`%|$uaXj)==zfiX>6djdHZO#+cqRQWuyHk0#o3qy~N@s|p^`H}?E@bZm@x2NPxG|uoF(zV6wu9Z=j*JetY6n3k5G zoEq0c(zzu>o!2&!-DWFfA;kh%a;WHTc|W7QC|$>*cLH6cs{^{BxfE;raQ_3_tz5c> zmn>Nr)F|J?Ulh1NdDCzH6VIZsZW>Yw_&-c5gouC6qPbD zDp}GFrbn;IRYL^_^K9Jys3%gqGn4KX?}|jZ(+!?m>_SvxKMcJ9C_BJuWy_PgKof>a zyG!Co{+9+rb!HeUr&y>0ETN){qQ=U;_H``YCYV``& z24S;igZCwAEOnJx=zm$bG%#n*0cnHqLU5g;k0b^!BCYIUx<%6&0%4c1?LfJdLl@KS zJkaT#dDhjw%a|RJV+XuEAL7y{UA~se7fzc*u%cyQ>=)<@UjXxO+@zr}*SR_8Cu@(u zZVlsmt+K=!iJ_VPNg^!@&Gstvfn(vQjBwVUPmhCMaoedy_kG{P#(GfXAFOoSqACt+ zR^rNk#}(@cbX@8J?DbE!?Ypm+BIHV0al zkN``IT}=m6C6YMV74bEuN8Sz5J)K`dDD**H4Z#e49<6;G63>87<9z{=1NWuf;x5lJ z(y7*-2}5Stg2@-5d-U|APM;ilJ1;{t;DSs9$Tz#*b-m}c>4Ppgva{fmJMqKsbY(*xyvPo@vJ>`>efK?;$*swy%5Y-i>znQ_ZImtO`&%{XYUqMiV| z(oTQ9W&y8&cj@i>ig_8zdo$CNIZ%HzuP$(H#6!XzlePfrh27FfUvdVUY+TE-zpXu0 zwXoD1=AGoyiITX9ZQ|47ry(W0H1^z_Ynnw9D*1g@8)P#t-%YQIXjxA!dHej=V*XjDyh|_g=^-`11d`!b|NQ}%mGAGG zJK&N{4Y>RkFZx|J)h0YQXB)|*ksN)IPtz3Cd%@yxM;a8{Q8A2<;XrIyN5wE(A8Pf_ zbnar4pBtZX`QJ$bH$T^jb4wr>^upF>1I4COWGxk48gN{BozA9n`AKhTv99H~IC)Yf zq>dV)Yw8Nw104v+aCRtc$z*Mxtj*8fiKF4Cek5obekA)IZ94q$!>^d|(fZ9Bd1QwZ z$7Pz#X1tbSYba7hMdyR4CFYb*i5`kFLg$_KjT0sEuvX~`Q?A1PkxoTB3p}(%US;ak z&;tKEip(j8R67)#1F*LN@0^!U*M&dwX$LBam43V2=HVVGfz9VNd!8gW#kdW+EG^)b zdMB_LX2GJV8#50>n5M+LTDEC)I`1O^#xMqF*h8a%3O6(JgP$a8EPYrmeC(E{V0rl=aFA^|mO`ddEOo2aPS68zS;T$I_SKh@{bggVVixm-lp9(#(8G zuwjLi2J5V$;TCp88)emsLjMkWlSeVEI~VC{7%R+hn#}rBduFimarf7fr~K*71<8zMXd<@O0Qy`31IvPUn?P z+#-7FGj@nD;=^WKn8V58@9MA01HX(rUu?)gu|7j39C9zWL>^yV0k<5s* zPQG>LV^*i=e_nT5y9(cYQL7Ayr3?jq4R24IXprdFL2I>MO; z0q1!!qB1d_%7fhV7bo+8!oDK1EF5G@{WG8}{Li2C0!z(tWxCgZ%LesH^;z*5-hzPL zL4z)T`Q)CWW6C~t4{2u>1=gtw)ahP2Wx0Q?pf9Ay9XDj!1_id?Ve``V+38+61c$1e za*q#Kc~0!h95wT0_EBsFMRrlqkjCqw_d_`y7R~83 zcf*(XWVs?ye2;v`v~C*q)tjNU_BMS+F;Ay#6JH62zFPU7fO>krV25`W3ltJzP`2rH zk|eM(LHmRik-4y0*D3X$sD7Bvs|&2+XM%mqmmMRw>8=p_Siq5*ISZ@z;#)4(%<8q_ zcaz^A%_iE3(*Q5oL>rZj(E}=wYE#tMW_?%HL-OgOH?_!Dl%paq@Pko)1o}``fB*GPur0u6WSFRd&eYnGB_NY_VgQXxji}7ff&(r0Ii} z`<61ZTmo)R?1oqorA?IdfG1i-m&`r>p-!3c{@OR|L47uV`V;qJuUOaX?^e;p;+|P8 zK6=e2au_(zF!|F9<^EaJinm)dj)snHIC;^XFumq9>)bE?5AnI>e_O3YrILLCZNkdn z_=sZhW;&k9c2ADbDVLM3kZQxb2V|kVFCxVcT89|pRk#D$8Xg~!XeeaqA(y3nq>aQ( z?j^@ObbeW`OD5cySu4O|+$-X%@+ARx=n@bVTsk4&ZIg5p+?nZrF3=wM8Zm~3fz=m> zxI<>h9go=m&I>nfdA?V^C5x1C+wwT^X1mdB%X5%ot0@qx0Itpr-pHNU29?)2z@XVd z1Lt8XgA|IH20~4AS}WjH8IXxXkOx~#Fd@(?I|V!1K||#=64Y;>u1%2?C|3u!g(Qdf zhA;P6?Riy!J<=$VnGgMGNS8S1@-)OiR}QtjcIcmW2#O4F8UhSX2NwO4#8+-Y(b2#A z;38S|n$ZbuH^a(CiUk6*^;9&*S+ElzR&dGBK+s z7iyX0)2O0`c|&`cjF->A2%+%uLG5tXy7&9L3pahFCXgup_iN2$_Pu9P?0Sl% zL5d|Xm;TIC2b85(AvU2?B59@e3TTce2^!_ya_q23WvwZkE69RM6N))B*TR4$q|52(+m)?@R{&LaE*8R~%O9h6ISX zt{AnCkpxk@86+1`xfmghyuNV(9aFHkCnhSD**+caME;*nwGJdfQ6IgGt_(Ib*|*Ai zBAQ`6dhx%}Jpv;%DPOoFFM06_{J#7&Cv=bhuiDIbOVM;LyC>&;Jct#E^fHQIqxW}*8Hp%xXCv{mvBa7wer!1J*9@@<`F zi`yLjmXp;Ow>{@ax8657C%kVAL}!U1rdW2RPVTLa zq|EG)Wrb>YD(=voipOM&^sHEm=kurU@T`;c@>`jHcl;CUipr%Z=yuURIYL_(xE1(c zuFXhu%@7$-SSZ4j`26`;L{}}$Rc{T~`8CMu0sF{T;R zzwjJxyNvqRcipy9S2?u~*^0rJtWdU};(*-sk>C7UfC*KH-&s{ba=tXEI$(yX9TW>> zW5uWtg$T%1XNWrJ`8B?2fxuEK+e%dh7m5l++Y~3?N)#pv;XQw)-xk#~w$?`ZR+#!N zDkEDf{tW@)WJ8VnGh0NZ-u>>ls$?jk9tDL>w#u~GQ`>|(WiNk+@4ac-n2BRhi#<|T=WRny3gg^_(3td2E6bqraA}YGT`z*5nit{^Q+q-&V z3Ed#C1_6>bVU3{CGe)#tSuRT=XPNppD&3IqYte)(x1MP`eHTx7ex{$FD9BT1^Sgn~ zqjc^=Qt6f@D1o2$-}j6ZH5M+1waFjxaZaeAW~Qf=N0?BvHeZoNb~!O>n$1vih++>? zq?(FeADJaBqX+$;{nTsD@OmJovW#GBb0NRdtwDH7QW@N+JV@vH?2zJhJEU>!1*Um6 z{=JCR11Wcms51D9I9t7i-mbz>3&!B{C31C>&Ly017tBO9buGW_C1J4wpy z#j|!oYARMeI&j$uNEWyVH(%C146x~ZxWsvaCsD9{CO^{jP4O+26PD1 zRJHs&lkV}iL>RBX3)hbhKQEi)QDW<5Lyzj5Dm&5Vb&tke2cy%WSP-x$Acoo(sMp+A z^fG$_KHX=R%rUob*T!k)HmDu0=4L_`Z<$wWSuMvUhU>)97%R1Pn?0&js7tcw&7}~X z)naY+#fW{g8og?zW8H^{vuz`caA~mvIBXwc$L^1R{JJ|SSv+)08x&RI)56QrD%Btq zBJ_szyFUacrP8gPhipv9|CP#YrW+Kyg7!~ErY7T=#j}<(y`+LlBPC3OIA2wzLP?tC zya(dLq$E_YDdayO>%D3PdgTRiJ>5vR2^Y#^L@OWxRl*!1n+-aBqg-P+Kw~&ZXQqtt61cM+JFX8 zq$B|bWOE=tgbi#@WPP%{&^#dW1ev3Yz88J30NEkN-!lEvVIMZ=l1XFZYn8BH*3N9@ z$B`;}y$V9bm!Y6$1&f^{7`uy?C-Ojl1j2wg1ghSy_ez+N7m9?0quIkd4tNGk4u>h5 zj{V@+zHeE2Fkh0z%1ZVLb5E_Z5^wExw@y&8f+8tn*mo;6O50-7_ic8Ot)MWwX$;={ z{G;qYSb8w8RS4ILpK>0Ga58C>J=-0$UZqp^$vR|OERiy%g$hM>4$;42s%)Kpe8OKb zlwD6-UyMHfa9hw1O$(2%OZ1eqIB}Sz*K8^2px8EwTzV#V{w#Y_t=J;E$opJ%kiXae zw$!k!U}Jouq-)l8UOIGsRl4DczK|;+IwcAPw=?a&hhSmAtRU3I^g@UZFMaVcECxSC z?t{!u)HKn3P8N(`6c*y z2`sF3RnW(V)ZodKw*H4{@4duOalzC}yqOW6vJOM*m>MPsJU+aR`jN_nvuDzvwPw@# zh1zC{UpKUVqTm>=@twfmm?apH4ML&@*7kG{tZ33HQ81{T$zV2lG|MkVZ1OlawZTtY zrP>JjIlc0Lst1%0@$NjjA~M$Nf;cNQD|Ck#@9Aa7C&BeoY zUa*7M@QV@MFABf0PBUMfPj3-jVb0B|kH91()^zqVH^?cU3&D3GYYA%{%Ftxej|~!y z4+(B{9FH2?Kr&?Yu_qpe9V=)<<3qBn9=9%v=e@CZW8mcEZ znQXfjbJ~ihA(msf19n82bv&*5=pSdi?sW{YUa6riY(`Mlto?zzrA?yPH?GUeAaQ(v z?3-LAUg_6EAC)D0bu-Bkx5H16M~Ym!N4g|*z@-{Az;A|>ss_NLUJPuxcnn3t@oJs2 z>y3>v`x_)XA;gZ@Fb-uMA;WBZsP`XNePk&Z?X>0HR>YyJ17nd;2b_?8cRkc3bO{c5 zwy@^}4Ukv_G9JJCAqztr=>eDJY8|^*g|DguSNPzo0v?36`T3iB;k!KJ$7NXJEURsCPd)Y}pA%cDFm+ zaP%)#c`jcytz193(Ud{w=&1}Ia9KqLTzUj(d73rVaK%yePXET>`-(QwqIp0+m7be(5AInr zX{YDs!p<&&Ynx%Blk^q3Kea zBq$2&Bk>dah)$irHi%aTYmu(3o<7g(b2qlK-3n;r?RLjlUKw<+X>l05KFcj_S}N1- z+B8L{Z1$=2IWe(K*ymmm34`XUf$#zJ>T}c$!aWfe+;;k+W^98Z4H{jZ-Hu_WRM3!J z-%K0p)m@4QI~Gjtz@LPB0r+ zzGi&=>#i%%ByaXyJS%Hzwd#zjjm}p+RIG8=YmP>A19*A_6_L%Fdf1ZnyPuy5KDG}| z@@Cmhra$Np>s42#E)e}8>E9g72r??93_Hj6ImihqRMEZU^}!~C6rb9&ofL7)e>?Gx z4dk+3DAKE>*gX^}r=qJt>@`yrBdQX|2_DQee~12+@ko~UJ-`(5hv1b^aSR%NhdeRH zvwm6yV~A)&eNI1<4qh4hT6m^Ic!t-`AiteXS*=(!9nX$%wHpG3QK8Muyo_qK8}fTS zTHbKmQh&{9y=*IrYMGJ&#co$)>KbZ(1&e&T<lZ&+lat*nW@BkQMx5aT^pnoCFMjej~CI^oCbNZX@Sd%!lTx^(;^(dh8CGkB5!k zd_IkZT}JM!W~HG`4ph*CJL#>|v;VjQ$Lxta)}zeeZQQHRYgQp2|D`MDz%j@-RH8Q=Sg16Re)t1gt}P4R+Gz zsAI!7g?2-`LW3|yv^V0$jE5xMPcO?>*9!9876?iL3p`7C`MfK^C}@JIhifoiF1=<_ zF`WX*hv(zrCayu4CCXx-2-fgViSpfQpuZB>dU@5n6CiJ!NskbR8~`=E5qag&k%5aF zpt42(o%*)P)Xe(ivo=!6&G&KQbXKdGVLCyv!25fcif$7olIvmEcb%(lpij}-e)mh# zi_$yZ3w;~p3*XUVji0e8ublx+C69JyK#?3#B&rcS#S)7gLlIzww2LYAF7U?-4%1kR zfz-Z(E)5`(+#7-wto`n{Nt)_P6qb3UsWyXR29yTjY3%PSQymbl^l0?R3q24xGKFx* zqYvOY0suZa^wU$8UEWUX3t8#g?ja3Ow%I1^5LQmbo~L?nOKPDt1zVgFgI9X?0UJ{y z?^ZyG;*`2U4y8Bl#eBV{2JUDRu5sTXO;P9hHPLTk zAdFzHUwR8SGZb*@tKV=lnV~;vPk&674wo8pVq=qMW^C3`Y$`=oQPJ7z8#7U6Jx-J^ z*IxC`p;s~wNL_d$JoVl3ex^X04~pg^T!3Es^;KA=c1)}UVyqkIz9pntBoGpQ>mBNX@-V?+ysQXoi&r|?i& zcnxo9V4R>$h+8M^Ch~xH9UK@ug;Yk>AkVeP=b}`LL!2fVp=~sl>@4>IwfDlER_eey_JkFs6JPg@*cR1RNN3%csofmf%jAXD_#)qgc6D$a7$u#$ zLt5=`Ee`B}#~g-y{VdQhnbmO<_nld6De1J)Yz;`E*kp<%Qqh||pzssQZqc#CXH%;+ z)&)bG$qwMq982rZ>DzbA^jjI#`ZGApx9_~hKK$X+_eG|~>bGSR|A(yLww-d`e;Fw> zL-%@$O~bf%^Z`|)d`(2I$WSket4%6%-#u28qfPr6(p<3vi)m zsrmT@GLIxXagt!0+2WB+u{w%mU@WeKzAwf6l6GHUH+_vC6BRE%9lR^@BuKgKpN#@( zUCbjz)J2u}o^n*)Io(#>p( zO65Z~8p68iyo29J672j2QtwwlHnTSzLcNB2&XY?_FTXh~mg)pl&`s!G0b4)YK0UTf&90>>W zv5T}(@4B17QSsilo5*r*;BaE2QD6p+42oSx!7PBsyH~Yn(niwH(jm5!BfA z$Tvw31Zp427DY8fT`#5tI>s(7#0?!|hmU{y#la%Wq@a~NTdsQ7Y^45&qB1uWroz^H zjJMu_ru%22AY=Hvaob~>KNF69$7%oIy!qB&d?wv#8Ax;5G6O52G2ELVGgPw@9{i3e zh0uReV9*TEDfds!@YCY11rm4Ov2d^haI9yG#bbeCuI6A2cD;sYf)jMd%f5C=Z;22t zp;jkeTCCtHSCj<=;fGtcLCrLs?Thb0Ih@bkHTAzAfHATc=JP z3l?#hIcLSwFaPyckhjVCX#Ux*`6P{-^WnsM-EuSMqkv*z+1x}$pHXGIKam%CrGqu8 zSHybVh2889aXXXF>xRJmr&pO6>b_z@P&~Ye1!m{J+=s%qLVoVs+FQ}uinsM3HX8#? zbDygQW;x0ZjR-3*ghU+WyN1o_O8CABE63&r|2=7P;$Z6?v)ON>*h>^SPeor=otuGw znl)A8hvb^3MWff$!h9~)^o4iL7;xDL(T#19dd*=;tY{Tk$rSqE|LUW!c9O%;>bR3E z1oH4xK8GcFQ+vZpfHu4zN=nLRBt`CuxCrMjOD{&W8;aucWqInlz&2qSy^8chJEigO zWdvI)ugjMP#z$lYUX%>D?25eSgSEM7f%kkWgR{vKDIP~IXq~cYO06_LGK;OI_sj0u zk_;F%Y+i(r&97hWsc~AesCTo4Pkv(p*?%|wd^-8ui9xo^tg7h=#Xh3Q11dU0Sqkh9 z+7j=I$Xe+&{%M%a-Lp>wff8(Z26*R~=0UZSJ#T4aW@RWJ&Fq^ACEw5-0_)R3bv&?} zVU<&*dljASnF%bK`O~vWFaJbPX%Mz67I|);T^5cVC4(+W(<>qiC8^9~U`(r`Pm@Y_ zykpQMO@l%?eFWdtl0B1q`3WqNaplvs(oFU^{3eBxOuAgMQ<^$)IS-VY^65jauwj6N z!y#Ag3FTGqHV|t{QDZ)P8~?aC*BeKIHq2b_8sDcrZFHU#O8JoW5nmpdg7M-1H}HEU zPPuW_kNcYTQjcZrg46ocR;m%O&<(p~F?)oR0d2Igr74x^ArB)Ee^@lPRrZ-j#}r4c zVMELUS9lB=z(EH%S*P)$UhUsaFj*%}$8WVH)`_hXNEW{knoFkGM2ak@qOs#_iJ(dh zv6fpQ7lN;P-PRknnsWZNtR#>&NO9H%#34K}?CC2@_iff3S;W4Og4%#Ex>6 z8BVrSEbOfcF-nJR_d4}n|7vJRs`aVy+byl(0TCy^DtLv@8NWmx(w^xx7kLG|bUbdT zmc{Z{y=KX*B*9t8ELPE2ZJSR&fOe$BsC37)yV(Y#m#-9#fR=R_eNbF8Xt^)9vY~Ag zw#d&Ya0b#yg*y^1TzDt(?LU;xy~jW6{pW|X%6UbeYxoMF{mX9$-xfALisMA z#e#)$P#0Y&$4Zx08T!LxwEl3%t%HDSF)SzCjuU)0v;SwNHRs8#Vi~zOfjluYRxw z4NdI%^a?=_jiMvST&GQuR7B!Pr2)s4)xu2461?Rw)R`2DAc%{q-tS%tS^>v!9}T;SM~Xs8V(>PyDi}^S%FmNR z5eNx*-tkKCg3X7XJqmO#IBKEftapN!SW%$|h96bn4pHB@vw zlcqiwSVG5pK47ZE^%0iuQyCP8Tjz5ws8Lz2#3n|}LSC2O@!l^eVUB|QTcQ8CIW0i+ zar~W2(hR>gvTpj)sAP|UX_*0uyd=*{m6=#6x~qveEAFAsKnH&D+^b9_lyDEYU?I0o zxn#oL3BB(efA{=9r2Sp`2c7R`!JoQs#(c2uy<&LP_wJSN@B1MCqjTSj5kZ&Rgc@F_ z@E8zijSeeTXKeH!vO4srPFM#~UFx=P{-??Ge7bV!TjaVEo1VpH5y3u+y-$%ID*8{_ zLeF#=*6w0)6o&UIB&f39&h&V%bZ?d)2uuzy@b8+n(!Ghk6<}Np4QlFmd6!^yaD^dm z*eYuhA6014X-Wm2Ssb$)mQF3i2{bX^tz?y$U5OJeSSBKI?!gnRJU`rN*The<%F#IFR**aNy`+Ho67a+TGS+Hq7>5 z{8#qmJ+TyRv7&MXDX2?G8mo)M9r7Imk|^9q)L?;q z(PFw1>7r3P;hCOH;lr91JXy6iHAc01~M)~8gZW_fn``s4>4*FLE;|{(%MHfmicGpQK z2?`84enw#%n38_?Ogck^OFrB`AESuIQ%D3BGpop}3>fS3>2(2k7W2tT0{q-REp^a;Zj~1d?I=a+CmAo%NsWvI;h;+;fv!JFt_BL zrla$#_`v#08w(h4)J8dOu)iET4CnO&2mZhF6{BfcapKZZD=jP4fd$jBm1!%5xHYPN z0X~6}97sx_R;Ps5%I4EauB~HIfNU@Twwk49yWO!ePi-5=g^WfL&Us-^peOrttOaWx$)LO_4W&J2vL)l&yutV+DlgpHnsIXSXeP#6!cnPHYvd z?0YxyJLvsWPfx0JTc;}V&YlW|!k}Q9$jf)z|3-pGrJK%U@vN~00w<&Je1aVQ%DNx^ z!?k_-eb&F`VIr8$YiOrGYS{Z~d*QslzqL47 zHu?YA`x3aO&UAf`cn-;jAsfNu6ci+YELIjnMJ?9OzOVPr|K2-uXOmevbKB{io9UhD z+!hgB0Z~B(HK1&=D&mI9S`nA3KtWN&1&N>*tDs0x;s3r#P!fseK*B^j^;bE2FyC|D z@BQB8dDilKedA}S(TTuiO|C?J*Q3Pq<}1r*sIQ0AGl=kFx80t$ChvFx7+~QfzkmBr z>i_xqfBo{$e|Vfy9&#F<`uX@y+isW#a{HQxVW#%tg_=vu#B%ftUMOrp-RY}tDWohY z-7Qb}PviC&XwSgJ!QUmAat(5cxZAPkVTLlX1tYLSwcNTL%pZ1y0{Ks#@KNyiqK1Rh z@pv%DV`J;zzw)YSB_$VF*l&u7m?;R|H?LoD)-T%;`{d8K7KG)A_fKz*$f0urFpyKP z%#aMokC48wm2+zRFNelXX|_Jy_J{+^0U<*(&WO*k_CZxx=6{>g*S;hOF{3(CC%?mw z5+0hn*D1~O1nC2D#{%i+>^&&Ct93gYR;;=)egO>KU~)i)wFm#1&%5-Cbv=yjqaWyH2`h-lnh38F0|> zO8kI@Ex0*iHxI8nCE5~LE~;X7@GGX{70nU-G;mNutLdgG5B=K}RpC$WZjR8()94&# ztEzkU7LN+w49P*C?JkY3?N;GS=I~k$8SLl(eC7*h*-1`AWykaP2b<$f^Lch_UDA!f zy_RBEQzVXxEEgpLB_K-LrFahpt94Fji}d)EwZ7wPRqzG2a20DA#;kW0C$vti*xZ>B zZZIqxa^&e`uN~JiHX4EA2*rLxkxDAEEWCt{CHlV`BTjtzlH{DOjlNFs$Oq|+Q77@P zcNJ4E+A!mY=SBZYdE6A}Uh=+A^hdGtP)~e}R}MfY;!&`9c|5RzUU*Nr=o;vIruf$h zRw#rAg08>lz|iMk&4VZCX8&`aXgn(1m@l;hi>x)!l0 z)#gK2CavGndW$t8`Z zg{v=A1kFs9Ervs|9J)o?t|+H7T-4Qm2@LW_X_Q!iJidt-%O@qU6qb&hRjxB{OE%IU zF<&gcp9in6=e5i0=61*r`k?a2<|%8$dA#ehPm?>)t8)*m#W_*M^j@Mr_Rw|cnlAUF z%r4$NdX)mt+tQwJHzLDk9;@JG@;##woxp^;5So| zSY)IRyAL{4`0tl=1S6+Or91{$QqW@PrFhp&+3i{8gsB_sq(IJfTaLHk+aBGpk33Y? zJZj!}jI>n#$-jjyR1A5vYy8iV_1p-=>^Q>*+}OiPL9!_pIFQq+$YjMKCFZ1|MsBUJ z93pL9BE4p_`g1up;b1pRy)09eL+1;TyztzVEV_a~%}Bw59>K~v$ZKj#bd3&H28&`% z9GN^z)s%3~&j6Hwv=7&i?3V^8&|W$WD0?Zkgd#hs$S7g0ZzI&tw+U_qr%IRm@0N6c zeqFxsB6BUcn8{QXGrGr=E>WsUMEy z&~yX}yG$G}7P|c!gIa>(*^gjc)d$_Lv*%q}dE-+ZFq$DZd97`-%WVp}{hP=KrVTeQ z$y&I`2u|q~yM-dDROGTrJ@aFHPK!VFTRsV?VbbV_l667d&gsf(NlC=Ipv_)&vRZoC zq`}wj&bsKcetHePe3C8aoW~%rS3uS6i*zxqX13d4teN`c7HJ8&8>S~lj}rC>;&{u2 z>K(jRQL-X-#=0PMNj5p*pK%#!-H>w}?#D2(^#(WlvC-v2vFSR8i~X=;7uL+GXn$a` zFo*6Yhg9kWaiQd5XgkyDlnfD-9LPl>Kb%H+XxafkB#gw0=lbcfz7-_S$=V*rNHE|C zsUg2(q(@ox*p_UG!&pk*?YREPOv!tTH1QQk`9SvrzZcBT4d~+W1vwmK0n66(2JGwgPv!Mr} zLu@=u2B$dekii(2i0q+rfAIKX84a8~>SdMkG%+eWXrLNxC+{C&Gc3v%kDH|!_CXov zyHlqaRpe$EIPI(UK&SVz*MzG!>c7FQpd)1KbwZd_Q%1NR|rKdvfRu zp1NJJoPUKk=#UX|o1A4p+6T-+0)w=(X>>My(n)*Si+YAWtlV(bllBv-+la zKlQ8sv7*p4zxzTZ@XTa!F~*Kv5xP0DqBfiNmi310Nsq`c#NcXq$? z_{<$JsNCUUbtH`VK=bf;_uYVZeGOKmcH!DmlF2QYYR6?$)kgMZH^mlHq=f*e zfO2L@KtHWj)RGu}vSPhU9bGvY$?%s8>-?G{x@7gP*Jma>9rDd2NQi}@*0y;`P#|zc z(MJkqWVnDB)|6Cf5->9%k=1&SY}E<~$y$Aa;tnVylk<3IL{yC7plVCwuNN9%^N_do zr=*-4Z0vYix?lvGQxtoGB1fr6)J(=y4YG!|NOuP4+Oi(^XK9oVcuA@(x{dC1i=Tn4 zsDlnAVeJY%QNp9(d%@`;Lme17^z~WlXkV=dau{GVzF3|Oy19Dl0(B>YmAC~=mg)h} zwIIm>GA3X!AH;Ye|}af3UuM4!$-Xd0X65<0PCJ7gx3 zUk~+e<-CWI9$x1BYEr4HBv`|HMX}E>m-*nm58obeNCGbOZNRg!!l#*8;d5uq>9Q>pn@W)lROBwVJ`fWwn$p8t9MTK*(#S%X6R_Lm zuJd{SvrL)OeX<6QW)k2%oL?@my3IBNe$H>6ndeyhuzBBLge<-NzGgi2(Z#PC@KGA@ zzspGeOEV)KGlKa+irr6MY@5l@Qxp;1;X&rD-{f=eT=zB@l-{4pmXd9DTrE&(WFmG^><)?)P?2{eTg68_OR1FU zJtQlnF(QS!{$tJ8wtsK=*P0@ZkY4{%mr^RpIfc~uH%6rWWck-_E@=7s{x9EK&=k?n z1Jd81UCdsmTKe|X)3W`_HT)w4)fTq9?6(DrhIK3hhmp7NQu6>AX?h%dXZLFX2GIQN z{q{p-`%5zm9WjE~eu~{gkzG_|vl7#NG0xWosSziT+BuQ`w{rR_%p$i1I_-sArWS}Q zG+w&`_K-?}x>8y%LoRA4T7m0PMM9$-kneX?XOeu$K|c-f$sJHVbVai1d-5&?Jfc~- z+c9;n)nYi-0ovm!2sSK4(sM48|L&N}Hx1A!ney=Gq}Gn@jn)VZXDJq1eCnx4)B)K7 zCG;_n1@3dk6i+gcz1GRIR9oDzTowflZU)xL@r)V>Nkwa29!gT!haeXS6b09K=#H$M z1H2YFbid2-DH>&#s*YY9@v(Bw9uLcV8DJXE2z}$tX=Y z(e$bpV?fqB$9j*FC3cJ~NCFI7)2*Y}M2f^yk(KkWIcG|eRD(X(XQt6-nR9^^{vatS z%=!nAo>+(q%ZAd5S8+lGwUrKepfaH1c;pA`$S!VEmHjRSNu3dfswuXD0MR~@tXMLh&=e~dFSOIGQk@JJz^JD`z5ZQ4t3%EDK3{einLMPc) zDpKFoQOvBIg9M&~4hKOOBbxN{8t7F}NR5hDS^y0)eM07X&SsrLyar16ql9(RYfcr@ zTWw5iKL(^&Hsqtc>f!x?tERPac3U`UrrNF8t<_15#VQ39@&{`n+@C|=cFbkcAvKV| zq=#sf81OHFr57GA5f$^&Cu3oyZIjYCBBLGlQI4j?w{e=2CjLotbEj#t+D!g8hquMg zAR~d<-}I1VMJdosHv?z1P8trGq>*)JNiUrD@EmWjF8q%!9tj#jvO~s%C3cg_cq~94DOfx~!4ZBST9WZp# zCfMs86__VE=(-1vSYdtg>~XPV+#g$mVH?kW>e$b6dhI9t;_W|nxfoFLt5&jztg&O1 zK>U9gO14rgB#1UqkuB1GXP9fkk`5KR@(f$-a>5Ap-u>fhX9Gq`zw*rnvdWGzl52#KO%%J4f>D6H zau29K>g5-4J)5jJN;UyX)Q!wO-e;nd$^>z=FkOr)QZ~07LnF!+9Ygoo`X?uJOxV=g z@D0;>hD(u?9aqhnnU%IW_b9Lu;cgh#BJ7oKW7XLH6X&E2jc2icc@-%LQD1?0KMDv{ zF$VEZs6wzkxVU2K@x)|fP@TKfdHh=j8}fecf$QX`9ovv=MmD6GVw)(^Kt-nMI(zg= zvKTcPbm$TRV*w=ilDyC>$#N^D+Xy80_RWjqub!$=o*;+${fcBo5uNRoMIS}BO8EN@ zzOW|jpen}+^GBFFIz|#9RhR~C06J=X6o2oT^3WA+)KFz5 zbRzmzOh-RRUk=eb>ygazxXh@*@)$MTHY~*$Cj&I`)Zzd4;6)F@70@@F_xLP*vvC0| zgu@T>&WiGdY2NqgLxPO3YOm=9M{BL2{`cG_}h-Fm!xUSpe-!$K{}oN3T$90`u1`crfmg9TZS&a@9R=~go>0#@_&A#$hz|ZJ0jkdrF zzkr!Xn8)olMf) z@}MNIME+9G0Y!)Wv^&r!^gGwd`#mwMbyU*Hz@i)+TE(fU>tQx37hR?CM5D41DyOh- z*4&(6=F7G4@Yt}Sn=35*ByN|P$HoY?niGGys-xVL%I76njhi9<8B!7}RSd{;=%OiY z^eTVQ(#;cV!JojHv&gG=)@^A_h;3jq(qM9g(nyc6$+55BI5+i+l_Ww>N!Qd^5_Z!y zPdHwT8&8d%C5U+Hrg1p%?$^50zG|3W-fl|$A^C{gvdE6><*pb_MxRhD6wsZdB2oIg zY{5Nwa|G&*)CsyoMWHKS+3vDO+&ilqDAG$kbA-3)olL$&D^28Ynz9CHknx@aa1WN{ zYUQiFVi-s&gk6VfxyjgEhGaM_v*=6y zOWd;qvRs?lm9NB1t)xE{+)=E21v(`SXSmxZOHlR9MT})9IpZ+mU%%ZQWI)x5n7r~maW)>vmCK{1>d=x&I>4!CQnwc z3ESz{1__Kra^Tf2dbyJ|++^F}hcjHBy5UsJ;*F*Xmv&oLXGYBuXP#tWU+93qa5=qh za;sAp{g0_ouxON@8Ux!O;O5w-onRBbZm=m6dw%mNsj=gQ&{ZQV(L}Kg6ewgxHUjZM zmFFY(7U?O~LkPWx9RFv>k(Ym^H|uH7fia z&?|=WEah#8IpOaBykt4hpi~&jDg9{?t}L5jEV{~4L~_E_gy6mR_;LfT(x(0K9g=Lv zxB}+vVY5#b#ez6$8WlN2!hv}pTq^6*+AyE^B=gFYNFS7}xCriu8ga0Txv0x{1GzR< z)wD=4=EB|euyiosy0PGd!^d#Glp9Wz`?tSiDp_M!MIQ$EJ8BUV6`9JkI|(b~d@o&C#{vtkT+v!%a=hM0W8+GC)P1 z5H1m>G7WR<1cxQ%kifV=s{Mf3hf4C&Lg7QjoC2zV!t6}~=?rayq(&E9QMsU#Xr*{2 zjUJHa2@{359p{SIXq076_%zgf*NYb=h_AZov)*{sI(N-H{8>P?NHLqFQFin0@ILj+ z^hQdw3!oQ|)x>x{Pn@g5xiJeTH8@iecGd04lgopmbe~!;(~AcEp+(b~6));+^u0360K<37BC1HiO9R8> zMl;?aiY=wcUMdoY4hI_AlAvHbXIhNxp#(!b=jR*dzIw=H>Z8Rtw%EC(uOHRYQP7)> zwZnM4APh;;nnCRb|KRl5CeYINB$$lL6Jkaak6X*!SsgxBGHnmnwt9FP;PZ$*v4UtO zL1*Fc*T0BjftF}H6&d4`>UD=rR%~>?DqHW|=Uf!FBD?`#!H8NqosTKxdtnbn%iLFs zx6}9J$-EN!u15zDo!*qz$vRYx5gWvZA?I67)`cvcz0MGBU&`#MAM=0k4 z$(pUx+km%7OCN@?(P2J@k}{zHC>ja^wBp{^we($>^RzTB>}u&vQ_DiD{o9o&bfd3_ z*Lba)tVVJ@_&pu#L(xL2+XVOntyu|ex3d?gmtx|M2h^9R8L;=A{9pf`Y<|HQf+1iy zYzPRD(mqL3e=%;l^f>fhircFM6X2-k5(<%_4(ITp#g zreN3*!GgiUis|=)AG+Jfs*Y_bhyIS0%;-~(F%O}u>wmMvRK?zmSUi$*r;Fowm~e(f z(QEb7Q3y|+C5+=;4X%X}rHVN=`(mH^7r_B8lh{c2+v=G4FMFhfwB7bye*1$r{*`Ai zBo4j~(Ij>f0SUWd?`AT^uBFIoDiWQ^D(^C-I%ev!kS@A`?jvicD+`uJK+x@|{D|YY z!h#DfYz2W4pUVjs_Pg@+xqR!ZUIttg4E|&aU{ zX~ctEETZ}4^@xui)4^X|ZvByIE#FHL8#JS5u!C1g?+xz>E0h#Uw(FeKDB&Ss!io~^ zmtd2fdK**;>sah@4;^crON?C)Sng_DgE9$UBEx@6__dsj$KK(Zp6j&;X2j2MU@clqs+FL%;CTL0j@-{M2w z!w)|VtR_c1_xRn?eX2TY@_X;S_xu@eETDUiO?uX8^KXISOzxs@u6famaj@VaQ?=K< z5!#K=89~p)dU5^deucv>uxONrT|Zu%=1t2;i~7?yf8=M{<@EHlWWNDyVx}bv+bqvW z&w!75n15)l7K%h`{Ppy&*uonnY?7t0Q4spwJS7__eIL0W=5Lc~L+^(5LtgR(sS!W_ zEg#c(S_GaEU-mJ5>~n8651!legQ@Qu7Sqe$IdqSl94T6B$3VGhv}A6j*z*)QOGPe! z<@1>bpi8$yzQuKkZ|tPy{+aXZg9b%iqAjjVC#`tnlB8(CRmF$*;52k#MpVi-NZSOL z1t*w(=e0gj!b*9ja~i!Syn>(2yDHkpYmUIqOZ_Z!edc|#H?UW?l~XbhNXK-Qc|43XSyk=5_DfTd z9lNb2HlzM?QVBUd*^^G`H1R!okD}SL5GJD*DG1!VrvzLHy6CwV3OSFDhoaKx@G+)Q zxBO;K_?WPH!1aT%2o>9LWuF120K)acwlt z;IIf6I#+faz%cA18a_{zn=(4QB)Blc|4=XMcPS4-p`2p*AyDkoDACVogR1H-C~?{} z59=v(v_F_S&JVey=y$GiDrcVde?8IqdxIf|?gSF5e^45cv3~7cm?!oR+c^v-h((rt*$l<`z2+#F`O<{1{pZ@c+gLlq^C@hc$}r zJ_8`KxAyfqsHTpBex>G!bl?*z5`9c=IG=##$aT~Ahz1-!CKdh-^d*IUJ4@O?>$j@r zHqb@3a8@`N5_2;Q93Ku2>umeSd|C4)LyW%Ao4G?O?YNr%qES4)j$%(z zl@#+DBX07KxZZXC-2EsVX^OyzD^^sudp?5JS-f+9V3(pf0+*y0}gTXI^l|@2n@Re-6AupY05evc?Nz)2_a5)h+6m)|7ugBA-fH^n$efQdT;0h z(fWehs%uXC$4pEXl_puo5O8}6Cqzt~|Cf8qoD4?hZ|XCDCd=*E$YdKCnT-?+d62bK zShmJX&0lGHe2|_%lXxAt$6zKRo!2Hq)i>(`6?w zUf`^PDFZ^S;vrdtsE-m&&kS7N`)rADRa=ctq}7GkAiHocVOiII?O&_Pcmy_MeXjBMY!;(X+> zt71<1HTV_D8(eMX*H{juVdlnY(>8co@TEQf@tVOGY5Dr~Y_fY2$U6-0{5njrRTL?w zB9pu@@>DwaT7Wh*(;Jxjkd&!VNA;Eld9XaOUJc@yBWA}sLQ|#4K=H_@bZ!PL5@Oxe zdhVsnfD4Rzr&9-vlAf<((4kzWm3~IYK^2*<0l8Fi6AFI1M9uCoP~%bN0kJM5L?7GG zjQ#sJ8Vuu|F$|6Q7kPy84g0m`KmXTUkFm6K*m1DMOruz;@7*`z*hbeLUKU-ViVEE1 z)k~V$L#iaNZhE6M%ePU6f8Ts%sdID0Wk~^zeaZ=9Jdek#fZ*jUli;7`i&gwPcl=Vdg>Ns@`b%T|qvl~Dg4K~@4b-5KG&}sGJV8B3+4-*q)co)y^;PnT9j~45 z7~$Y5#a^aJD-~Je-}&oHlBKg(DJn=2jWUM-yoG&R_`sQ=`CXc6h6yZP&$?FH`0(kff?|R2suX1+QRyBd z_*nF)&Z75EKjYY_M3ykD&d;F}f-3wsvU@>#E1uQSvTAAMAxk7NA?WQi%{{?_?~Yl` z&t$El9rIeSgduQK2$G6a`>W59KIelF8>;s%@x;YIw6IlV4L(In5cFJw^7I8`#?-=2 z^*Tobf@)u@{D8!~GzbFf)nN`#62-2eNP=FLI#YGP7rO^_OS>UHU*t=8{#-KVZ`|ks z7Q@2mF5BkpomDRxe@0M)G7mdm#QDACevt z%TMHk#1RGpbkUhU=Njq0X)BpQhm*=mov4@kJg?h1$t#|%b4;T*%!uZf&}E?Og@4+d z+CWwk|5Wima@TKRNu9o75%18z2;r@`IyfDL2bv z$1%T5Bfzbv*mV?1g!B^#&FT<^gkLR+Dn8P@E#+(F7}K*6>~P@f7QyTDf5Y zZK_9Xx9KD^3d5Jf3!Sp)^;3~fr4iOWslIpU8?$qn4;MvI0}g**^x@!apxp&+i8y}Q z^c<%;(Ov0bkAAvcUP_fY=i0V6jKjD;WiG5g+E1Ry$$U^@g}b%}7|h4t-furdwof8Q zjNZHb6bs3|T~uTWDOR-0moxqJR&gwQTA1Th#f)?eOOTQk>l~N+x63d1uMOK9-Y!4l zS?IdVKV7mn{9f4?S<}R{hBrVNu1vQ$u6sz_7=Td6U}+ z*)4ILfAE_N78;=Qkhk@xq}-03mkUOiK1H!0U~`m;j1p$Q-v8Qi;iAVI$l&{EcJ}Lu zuV?$Ey-4(07K6k%L zXShIZuY&taFcw>62Jj;t=8Sx9NdL~Qw>?a4k=@Q%W{hN(D%YpnZyS3WqH&$F10fpa z-LNwLN=3EeT13a}C{;8$1l5?JM*$qSrF06b=W@U5))LyGYVc@|=$%zM7njjznbNtt zeO5A{IH0>uf5l3smoBz$Y{r_<;f#-G+`tVw?)U%YOQyZeW?G0*QxlV!hy0g@L<_4u zAoJKD%HUzt0uS}@FhL1HjlFYENmhGRd5=#W)2N2w8DF2ZPH`Fr`;BwGdEu+~O$#y2 z)H~v;LRa&MEL{4UNA3pM3SYk^h_J%PaDD_QMA+}@a>AJ#zc*bTJzYa($B9NWOQtmM z_0CPQv+|*zC&+5ZN?-amO5I;oyniktibByxt)R>U6N+sD>=ig4T0*ZSHix+-khY57 z!4cHvA^r8%nI}yv?aXw2CWzzN^bnZA$*tKMuXev&Rq7O!^KOJfmIPZfB0cpFw)m@l z>Q37pF#lDLQ1fEHZ{J@s>0rk9j%OZe{H1CY{*mKay4Pf6*TDr$n!NG1#ju~kQ z9DmdjJ1WsCY6&7l9W^_hL0;Ql(F5K|WdZMbgZZ5Al-p@8$l&(mv0cp#8QEJt5Su2K z?G_nk(#85Ec#dy7tFCkGqBqThlHnX0tBMoEQxretMteK0Ty^2&%!|vUSiUvS!NlSw~+%I$ZT?mwFeiG|i_jM7^C!V7lE_$p#&= zLy&tAL6pe93T17FSWQP_e#cx+2RLdnpIsiAVF!IK4v-NOb{M@nbnKv;btokSZe5=AqWiUI( zBR^P2cG>Y#s?NxeR8wpP1)`ymX*8k+M_;`wVExpdutz?b^Ydo+@UXnQM6_s8zTYBH zwFikHS%&0Z@I663y&}9_mzdbdD+}Eoa?l4$CDnDZR;MGL34E-aQfJd!+}aiWiYf*d zTIC|hoQpt;8mlbp9C68ZP=$UquDixyQmhA}k!G>cLx~%RzVf$ky(u1=y9Dd6kOVsh zQGpSNG!(muA{(hlWQ{{XkRExnzq%2s^fRG;zk$93eFIoc*$ASD*>nYg3|6!duNkMp z%FzI1^uuoXotywN;nw%w-~OTlq+W&<*%*!iDi)A&1+``;jE-c%|20f{2*&pNoHfdF z-fFLWzva)3lvw^^=|?p;Ru)1-=TDe>hAobBg2u#SDMy}CgW*EIV87`u zb2>5Wrxys@1UCcg1R4B0pa@egYIiH9J0fxd(nB7(V<+L>@Iu!U%%+e52kibCa5(M0 zDdY(G1WHIt!ZwAB!ArGdxVgZ}*jKgW>7&2qmA{@Ani4PA?XuVmBg8=!R9?#}{Ij8b zbkJcVmHwrMh58PSGoiY^4JPf^Ti$@`Uatg3y$@I(u)@)>i}IJao5-*5*B`<=k`QE0}{~|8jCTE?&xmiHV1W{iS{xT0(wd2pdCgo?pBHg zM*dAy=@x}E6(Q0}mG%Rf1Et&kkB#v8H-;Vv1JR^*3rq~qxWg~aF^+6TB=GAEF zB}p4ZVz-AL^utPq*hx$IeWEVG(n&|@2G<0p0qXn50~w=&2d6O}!^Peo6-{tAz~iT< zCp3{%Zt$?GT)eC zKRwLbE$@>Q%S-%_y`>;z7ZW9{n0_=oia#jP(p|x4Mc3wxn+-uIae>BYpf}oklKxa$ z9%is3KiZd{PD<=JepYW}89t&|;8`!DB6mQ|^o|8sFx^Y4ly!o`y!`MUc^SW4N6C{A zRIEyLtqgCDI7&L?ZGzM8b%JhwLD+id-IAQpt8OV%OT)K8COGAlbheT%_DzHcee9%Y zX_Bhmu}QXyoD(gbgiWn!kk=U!RxNhh2Gl<**|9^23oeW;#WVLt3x~(|adK#>V?UVY zWtu{Mx{Ae)T@o{C)qumYX}#g+fFL)Mj$@JOf8FGp!Oao5 z%p;#YqV>~j>2zhZut-!3XLkanSVwSg_!?#v6tn0l#*oyjk8Dy_xi1!`TQ>_}C79@} z&zP|>>;dLsGD1{^ntb~mT56cKM0Z4QlSDh7wu+3Vt#pcoTL09c_0F*A&3OR{aDBiHb(=Ip?hullN&@1 zhP~bIY=B7VSH9UmR*j_QZO4_Bxkj+rM6nww7@0`S66)&mkr6HlbhGhiyQ0ivv$uK| z#By)C?hniffc4KU+57K(u=$PgfXrwRGnNpnr6@Cxn3c>oeo7nA(fdQM6C`>PNjLh! z*HY|iip1f<6W5tozdGpmN%#p;7^>&59*-~9xG%c z^niT3%a!00%q>*)i(MwBQsA`7Evh_iq$+ z6(qTO?cV#U)??YUZdy;g)X3~1G5jog z38UxV@07JGsyuUj#s}t{g0JIgbiDR``z_P3ncdDUWawkzVpH_gMswf)Y`^82#A?nz$#&Zj}g=(YeE zWVP~sMXEsKwQO3dw1B6PYNfXw%Y&8)w|S%mYLv(jWwT|E<$y6V9K&|3juqVlxdF!M zKfk-c*8rH>g=E>@$9p+boKzDNl z?yVwaPFV69&E^X?u={3erL_wN94>r$z@hcWnckSyyAKKCp{wxxc9(u0KJC8ssg}kL z8$E)o4IWOgp{kbLT5C#X{dBU=exp24OfuZ|Jf^I!5EY58y4{_vs}#YmE+934Lr53} zO@O<66Ri)3XALA5usrKI);?z*EdTlYulD}XU{4gQe7{Lfj}&pY<1BEO5il-MED+M3 zqaq8W8Gffgr8`Z$+Uwe!Zn~PE0dh}WqCuVI=_Tc&Gmd8*k4kcdTU;xlRvH?*q*{4~ zPo-*)>*wx=`ITUcQkhJ83F&uEU@GZD0V(rp{Cj|47InL~@+zVFAe!7KH-dNh?VnyQ z%5lOHH>_spQe=tS6)E2LNu6}HA|+t6tdbsZSi-+6{ggN0a6WK_Vg+b<4~i3~YCQ%V z(%BUM+=w*a1Pj%?xgc!V)HRk(i@<0eVP$XldH>#kF#kWi@(Q^&62k1*^u$aYZhCqt zwud6Ou@N~faNXo`(NS0?X+dYDV*0&cb(F9m40WF><#AIm`_L*n5S}ew25hwl{mSNp z|Km~PuU>?=ElO70^{!&_pc@(L?^{L7XP1HUa}AWjm5cIx)V<`WykhztXm9T&gAV#` zvmRblAYRxdg3|-Q%3tjd<_lv&jU-V$!Tr#lSQpe96gy+Ns@*e@ zuWt9;$-Ct~C{lMZ*A#;e7bW-PJz?uCcou=yJn#n39L{^ud7b)#uLKV`;KT-fJ|Os!$bfth@b^^A8FaYmnkTdd zeTqfzx8WiVDB$wgF<)yQ1tTT1j{UK0v%kS4?dmFyCOLNOMAsNiaQi5BH${rE0dle+0+`HxOknjwQDUoAm@4S{g}}*`~jShLzR4Q`Oi!< zOJ)){NP?6n&Y_dM(nIjIT@HkLTDp!dpz_3P!kXD^3kddx|Dt2w<}w0f{xA*?;4}jE z>nHsBV&ymd3`Qj&?ZY)B+m4OOVI!lmmtspOvXhEz2ea{iR~1-$eP-%Pf{Sns@3e3S zAM3B1BR>1mCCM%ZkLdTe(@pHPIZOB{sytpVIqcd=A0x$X1K?U5BS(4194qFuv&|7} z9kW&VG_*|C4H9xe^mB{~2fRN2Qcj4bHhlKi{j&_fdABU0iWIywZ6U{vV0MUNODVEf zKX-Ki|5vi&iUKGUreQA`mTIfDbWuRGuht`3vD>=~Nh}jcXJ{>r<}-&bn_n23C&pTy zJSQEC+#Y$1EKAkQs`JDLe3KRbw~_ZZ5P%NV(kOhG6QDcQB3&lD3EuA5foDaq4aeg1 z(e#`vs*~UUnQ2CwOUBZEEh;f1G?A%l<8_nePH*Y^i4Z?;RSIh93!==>!YL)HGACSi zVR>DyWDt6$$3FJY83NtuQQ*Up$By}$&j0%Lgx3w`<)_yhHjx7_jd^J{GB3v|ww5A? zab=D!u|`?#{Si+KWq^YYouOx>z#E6VouS#2+a{~)nI`bgbVAH@8?O!sEs@V<&;g?} z$%@V1o4wH_H4`ALtnyA)e5AYN6Sw$~w>p`2#Vy%NqW&Z_O1N@PvH~f?z*IFd3H)fG zM%B-IOaBAR(xBg6Y5L?*^jwbuOv8a|VPrRaBA2C4&aRal8U3*7l}F)O110_Ra3#SP|N zoMrMWC*CpGn?Ke6Du8^-&EDAYe6-Ri68Ml}Kc`496$u3q^gT%rcq41VmriX8N`z*B z`$!5U?Us$zPJwBK20lu*yNH1q)vxabo3n8+#! z%M_yzMR*Aor%rnKUXJl}K4s zPG<=-T(U!|Jn`0c#ct&t2q&sPmzD<&$dMb^ZsURCm_Pj5+W^1bpURe!ZFam2gI>sC z8TDNh3sejRRAkTm)9wjO;_J=IB2nz5;d@Ib=|vz*=tRX~{x+~ics5;}CB!|{X+;e^ zJUet8<$n=~nZVOv1B*2p`}FkeGztUsoLkJ>rx5~z)Fk(h^Ir5K@}rVoa+}ss|KkRb z>4IN8Ue2%4Xjp~`qg%#hhZ80ye)lVrUpLjU;UABu@5&d8XJyqB=XCdIruIpnDzyouiLIMAb7hNz@@rJGV<_a7K=Q>)RK}T5i6u9ovwdMy|^?icO=)W-78C`scf7NKQ?UkyQj_ zfx<&BB&DJE5fw+0A=;O$$W;#bYg7m1>a8Hvipk8wa=gdpVBrdqA!A}|kX-y(2=j&k zB(vZ9q>YqKA|M+z%rQJku@FhEr6O0mpNBH-7;y7+mbqS*tf=$Y&7LCtt|tfxbKv(_ z3W3xqb>O)92X@=S@~a6~ zzu{rPlPZ2i1=(W9D>s6e9Odag=-EUz|`=;yNFbD&5$yUg81_&dld)@10{VtawP_ZVgjy^Ca zZaf)|xs3qV?{W0ka~c8rjdy6kY3vFzl@k6y*#KY3_YcKC~1+da#3rb+vyr_ zu@ELx(Yhc|!P??>>J9ahz-}5z?aG7Fe38c?o~`sRn{Oi%^8|w2p#EfzeDVa_Z1s4P znQt4WoBcm`k&&At)py#lC%eKZs{4Rq`zUgcitLo6JKc1>=Cl$Rqtul&l1A`|12 zb;beus#Mx!(8tcA5<`E6YSu9 zEYi|i-!@tcQCRSa@vasi4a%pBX~-VwYdq1ZI<87_)D>}-&P@=vD6o(-JER(fN|?i> zCm1U``<+(`qaed|hHP-sU3s4EQS9-_p|ilhUUYqCJF`gtTU{XSigfo6}N)R z=>+)43_9S4on^X$tD*3@O;p2l$xgd>1Vah(XfgH#ZMNAmZ%$Qzn*ZfF2K1#V{xgs4 zv14cF6C-D*hGL;Xrh(T`{LbSJaDzy>Fp%Ryhw#BlMzu zc<)`vazZH&{njW_ltb3U#+VuU7|uVq%+Q0kYVl_wm3fchVxfUU1q!O^fu!Jr}$UGH1QdTvnDI_ ziEPx{O9)!$3U13<(c*}YJ<36N;vw$hw34aH2x)eYk=fh_9tA6e&LOeIKpQ=i(_}O8 z{zmyW`H<5x?T_z}WNwp<9k=5_9r3VuP8P+2EOiXujtjuvz;4UV?CEaRo`{T>KWNH zl0z>j)Wz~gA#3LriFSKl2yPJRw;Je6QuQ`g>!Dtucr;g+`A{ST;X^@l;Jx>@h_TOd ziMUMwPoIg4?i_lP_-+`S%Awm->XpH1!R3O_8MXd*OPjn!vELCdXj7$l*9Dh=uEEpt zkXFRb(`e#y#01!!rm+co9e=a;i|Pbx>9?>G;Ry8MzJ+CWxL(J#b;TSjn;DZ4GT!Be zBjbB*xi@hE`){Vmuv;hC%z_8WBQ(lhDEZY3lBkiLCR2sQkLu4vs}xzFFl}qcWcV0y z8tdUV+IX9j&6xNHm$k=ClTCJu3^Sdb>2AAc_J%_`AYXV_eg@>#FjU+P3!-aIYRt_e zGeBD4IAVhnG{%F&*#J3C(3tSu=`C-Y_7<2a*I#r~aBWVj$nc{vVzt*^_e+u-dVNS! zM3byeFd)xij*vRhCBYT;v>5kxyKS&?^WliE6}OQXnyE*8lvNLM8VUPdUjF-cDtf+a z@NM`1o6k?liI-;KeBHmoAEO>T% z$1s~6H@og(is{CP{DsBe+r98FDR2B`;M@BbE&qB`#KF)kH@ves;tI6?Ka?DEjAgMi zZpGvt-pV;eQx-dJlva73_1o+^;E>~#A*p>e)_uSMODa;OMWCV5MsMUD5AUKcyA?Vm z%>}hC*HkHf=4tmqNj!Vd@nC494ChDu42=?_LYCRQ(Y+)?FUzQYTJ$m%UuXXAXh2u( zYn3067&~4$L0n~6pe2c7*H9z@l?L*JjY?=0$q6{5>L!?5Okl1#qps^X2Wa6dw|Jl} z8XoIi#K{^?IPBp$;m-zGh^7A#PHs;k2}U{NsEO<&iu41s9VAj)fJZ82o_Zrw>IVco z(h^Td_`(6ms?epdyu_X#Jc{QB9a?1df;8VWU;WvVFcc1dKt~C&lVi{!4c4BmPFRjN z=y2PyL-a^o4&{I9C5o~jjaQu%TR!k+plRh@obv!yn)uL?FuWX-%XkDIiK<5kgbM%t zl6FPW6b*1!4La0=JFAr@i0frXB{lq;uDG-vbg1_GzkZ%NS%LkuNIQcy;IPP^_E{)4 z$R*8PZ zUj_uC!|hlq#crTT5*4Yh!KwBuPM#q44Rd8_tGVbh&5%x(+nTUA`+6 zXUS^!O(8KM#k@Q{170J8g{{%TyJ7I892y-dtdYRGK($tISfcLW-3A`Ec10W?>Z_z` z?4Z^4&nAesglC7GC3SL*5~o$nJwyq0nR2Y5e_m{m+siO2PuR>&&wg{vzYKp#{QTFZ z@f6g zb>A7j0~8Qy{rg2_{1)jxX>&xyTt+r4SGfT5GSIPLk`}f0)M%(t z5E6^{4feTQbd_#hxc<%0rQJ|(kKD)4bJm{zr*Jtky5Z0d50}s3dM&5nuwU!ExoyXV z7oFm_L5LAMy<_~@{4##j>~y9_k-*^gNoB4i-YZ8n=&;KT#ZGhC^`2*-|3uvrRytRs zY-b-Zzz5E@0V#c%2k|YLztI-bI z_uN0WJ{+;}_&!^2aKpx@v31)_6^QNDMKL3djFX^Y3e@%RVq`;K+;3H5V9CH)3JOb} zWzB0ifnp;2bN3&684O0j;7|6FbZ)_U`wbqFawGGylVS@gl1D}Egbw=;uPUk++=DD} zHZRdN(RGDSIz;3a3!5W0`|bCyQnd+gOx+`i@Zw zt1&1}buYM*-s7_-Y_Sm97HuG)YAL`>XCIS&Yhp|HtT|nQAHC>y1q1hlh zE>}OC(L6sE$XrTa8+S8eX^6w>!}z=ha zobCy`0*W(e8q`g5&w@vLZF~q;pNyg9#^T96xfXULMG2={dv}^DlX+wH|$BA1?}`sr}9+Vd%>465*+FnxMF&$p+9F;gQFJf3`Nj;IIDe{q{p-`$#EaJ4V3~Bd>Hn z#qOcVE-Dhrkz`1JiUo`~qx3mr7#HlHfnR!B94jsiRiE%cv56A8 zEU1`SIY*5_LqnlIGF&2a9ajD8m29m@RL314!|lw|+arwis1BJoHz%_WecO~L_a$*r zfWB~8N^cFtCQxJ*EOm5@F2!!@nyEF=5LPckrP1-Pq=a#_3@xJ?@G;%A?8>+InpVxe zB$zN$V~fQh?YcruXj7fDIYfOENPBXbY{z2x_#?wISb(t@(;#!d`JK=eoj)+lGU-(R zx5z0w4x@f*^zvV%Sm57mMl#d%5UspmhQ88J|92--z#Nt%X=x5!9@G(>;+N0tAobqr z6Xg1=<*lejs7s~qq} zrcM1HJcQhxO$$*}JzBWM4T_Mk3^!kRNC{#BDE-+fDS-(qn|GZU&SX=Y-$;042PuQh z)pOk&iy`y$6f>qH&jXS9d1r)7;Dk>XT>DL=0buuo{wgPJb_}p6qk8dfiv5ftpHh*? z-P9(i196CY(2(evhw=~VO6h<+Ph7`4CrYEwiOPdeu%QFIm>$JjCzV&-(7gerhlopp zHT-r)x0LDu>4zA$SzZ}*P}!o$;6Y3LWyxo+C4lqNOYj7Aka(7P;PrT9$GjF6Qbf8E z^Sp%_q!*4REAV4h0Tmct1;0a`W4+)$aNDIpvprVd;^((W6JJ5Xr~)c#b}vcuEoRQb zuY$jjmcHty4UK0rT#A{iZY|RD^9v|r9R8y~Hv(dfaL5UC6N~>k@1dIkbdA5<`xZ%= zMD`gOk6em{`nL=!a@mw3(JkM4CU4d;4~_EUKon-^cist1OPl=8`W1z01N z+aAAma4wUbK5}o4=usS+9m8MgdD25GZIPzhw8L~vLt^wLT6c%jkl1hZ{P(dd+Dx0E zxj07l`!6GATBf(ti|R!iRefRioEM!`7Qj;bvSg1B@|`eoB(cm zvkmtT{m-{@O-T!15@eX6IIMo-k_0rf2&xkIiLNO=n$x;dhoT$XeVJF`jA?VUo7kfU_J zVk28jSC5&Hm=lghm~hM<8Y9MjJ25xYwDW@tj_f$bW2S*6js95KLH7fXR24L_f%Ah0 zbR}J=huwiK0W^Ba+VLxG;0lg0j=9CRhpS5|#^9>FbFB9$S;8$XZO7o)YUFXPqu4}> z#8Z)L!piCLAEiz?19{1u0Pv_LzZE6y5~YdzY?Cf{)=*fCif28<2IoG0v~Ko`UfqPH znzwwb_}#Npf$0JI*rt?BIrt_-bA=DxD_)6rX_4aO@J~krVuoC11M1KJSvdQPrDX(b zZYY;*SCj%lHr531b>B*|fDJj{@2>p7tB1))D)ntgUDO3tq-~OOeCi#y*m4Z>DxP|X z4Gw(i^FRIgD?{;?uYT>9q}q<-EtidEhz5#1O_7hO$jfexpz2$$>ud*Zxk@@g{HYxM z%m?I*^f0M&2DVA(G*B8*_r8jWAB>40#avFnZO3*utj1jv)(l+kcvB_aCcyKZP+Z#y z3K%6}UBOjSlWRGAbDeu##O%olE& zr%q&xM337G+7(M?#sXizvR&Cm-*heaYZZYQbg}|Dv!iBrgzlDHb=w3TVLiMGrf$ru zlV=XkSYYzZ2aoY2H<&DU{Hx=u2AGrv{O>Z7&&`*$<50>mBTyZr*!>jQLq+yGAC*@N za-5PC9ibqWw%`+SABhS)%un;l4{wfG0vhB<>w+vD?QX^NfJ26)ooSR+`OfyE;|7Oah6k1YP& z=Uq5`$sfDkHJ~T`e(fZ3Yovk)`|V{RaYosL0g8P$lt{4Z>WE|FBWkln#+ z6D_{(2K{_*F(U-V}Q%{mau zjsa+<&E%3#JAxwcn&JW!TE#+H-Z5eP6r{^6^VmPFkCe_uJ&AuV#GbVfTi1Ku$FzIE zZd*#s^aBhrJ0MeXgD8WStcY{n;=085ks{h93w-Qo_Ri#0(2#^LGfGlIrB=`iSrLC= zc8z}>XjE%l&id6)?W95ICY9a5)buQvzT|M z#*s8`-jE#|nldAAXa~g>P$ZX%)UVPHEV#EIQ=%p6KIi(NC}Fpo8kt$rys(2Q$qP8o zn8DdtyLUaLUYF5E4YxsuC$zoGfl^K1v2XzRY{g@)hsK_Ip#msb8fR6CG5gs8D&aL4>}>Vj+?p zBDGn1$O?s)J|^DchI);66x-Max{t)USIWDf^a5&CymkegpWh{mcDyggv-|wIA&7I1 z^g+z$EQqaj0mUJ*Hf7W6f`H!JAY&cJ+wM~1e0cw^^tfso^8~X&LuE9m3Dl_dY3xrRRDcXWjvec*dpqY=Lri#lP5#>0^X1P zQd`{MH`S%T@pIH2PU;RC&1xIWT8+xBs%|Jw$3IUahRd)%J(pPR$WsX6Hmr7Qi9(%A ze`QKcW4D&b3}eoAm(sbreaaWScjAqUk`8$`uLropx4R@TXUMJpK)%p3q*wG%QoP{3 z4-=UG9lE43;wtom|Lv;cZ};R)5l;i}sSzLLfiVJjl%M+i+c*K3QeF4D{>4hNW~nyJ zh~{fSMO5FThb71wC8k@|-QY~s%diBji!Ox%W7~npWEi<($>esM?f!}U;K9V9rTwzN z8X)T71TQ;Y+HWzMd{!Y3V1p#o|fUFQTE8siIzH_7F=M)@%y#pTYchM zjMsPita$7Q%OCsA9{%FK{PnTa_}Q_!FjM2F5B(Iop*+HwsXIN@dUtnZm)FL4vJ~7l z{xzO@Ja6Ccyd;K6ME>EI8p&!qj`jhg(6BXG8pQ&wY6|8yo1sVP5kJ*A)_>{jDA?kV= z(wt+39O{SE%WA?js(2W^!cbgUCoAs2suf%DHOgv_dhbDp%aQ^*mC?87=ZUek9_cbE*S&8U22rv<-3u4_ymG zQONqPNdKGXiT@e6d&;a>^+1@QKiD*I_jF3!j%(1(1oP?~GoElqBtwnYX2%;-)d!VD zkXJyitOMK=24ni(IoYc*eG<2u&D$%dQ=R>$GJsslnC;l7F{5UP&E23In>;HI%5vU< zGC}Ze-ZGHKBe!dmutx!Dtr7yUKG0oLp^fcxUJL3K9BW3o4Uwgk6T_yuC!1V9iqrLP z{_*qgEiz1B3o7P?kSmkOpphSOhhlG0*oM|yKDAm;$ zFLVUg%Wwv3pi>|K+Y{XH90eraTgh5}vf@1SPwT3wa5`&OTmv#HEUH7*k#h_@?2vz+ z_)x&LnegNzehFfYQs053#(GeE>yZf_=ROehf-rGga0yIr`le}A=SpR|`OBFTqz6LF zx?7&;iULjN%E|3Okk+HX>lO&bYUf73@;YT2*B zug(8miHiYqziK6m$QnC#s&^Q@3|lEyLy=8XWE?-qD+~Ca_fNkg&TWwhXZ0yEoeaRTqe_}=c$c?OL5ELoLL z_S-Q=&KP0jD8<%Lq?(FM5O?d^BYJ0bd&9&OkUHfQ1R)z4NM(Twp9`|OD0i+;PTqFZ zr=rkwfp3mRaldo3{8~Vsc$Gg`L#zeMp>I1DyJ1IWn;>)k(%D&bFWnxzR$&LPI!eoN=77VxTJ5Re7p3S%FEep{ZgQ(r zy8?5k&B|1lIM*&&zv3hiwT&JbmJHn(&|}HtM}3WX^!#1>{@Fhmkn;ZO<*$;?m!=wU zsZmIxk7Dmpq#HvLSR{cwH68NP?)r#eE|a&gmOe`B^bgtaItLVQaff1Jnf}BTkjy8a`6#3_1+SzvVBy|D|o;f(ik*UbhvpX!2R8dXM}z^s3e~ z-OyB4B-bT54wF?-IbTk93-HbcT_6hsD99{_<9L|?#^=6s z_Wmyo(0h2RNJ=hFA`mqi7SFpyu^`~rNk#VRs2?7@zTKxz)+Xp6(XXOJUhJgmZ!OwN zU0b;7Uk)!Qbo&47eF7Bo?_{QYHkc(jQ4k$>33F63MsHlxI({9h1?)KUq+wQbe z*>>!)JKIfXJDs+O_W^m@s)E zVWa(bK9x7`@dWaG^Lu{J@A!VNi2I_>%MVOlJE@Mdd_paq8<`hb5Y^Ak2+f-Pi85(2 zg#RYrnyFV;a~mODc#d<Ew+WV0LT%-34J1yx>LgRF_4 z5s~M+Y(k!@L5vMF%O~LYPv^$+8|?-&W%&!4UfDMvHk@$x&36cvH#J&Ejo8En+<3jT z6Hlt}TkCm)yTvzOwANDxS#qG6Gls^o=qH}3o4TTX3pTxPB1v=G$$3tNeeTFySacXLos_5kb55r3jpc^_fI05d3QrMxhZTwog zB@CYgZSr{qQTQ7aIN0dQ24O6hEqk{+gvq^&-V>@!2s;|{w+*DsjW>5q7MQ80m>P;4 zpklBWVynL~S2Ltg-w6Qf1wZJhWV<(#0^ai~qOa2ry@wRiZi<`zvVluwNF}XD*5Z>S zs*r7CH0Uq^KEld{hh=d<=dQu}?UI*IMQV1?=jeNru+iu4mzu&1vrf?+s?U|GY%iG( zefT4f5ZIe@^Yae=3N~+JwX%+@n|}VTbtjBsjL&I47O2!I`(QCf3JU~mw{yQ z3ZgcIApPgkK(*Z|S#q&0VC4}hJSX*Gc5AQ1Vt%bZ#CS4-bt5I>JUYQrqZB2{#sTjEsNai>CO1Q4 zJX@tWGxvsP#^fX*RVaYKmhoIc#8DX@p{>gdJ{Z;Qg%-~s>kI#DUeHduQPe)W%ywjI zq<~|yo_*21ws6V{aMZE!t6M`&d$60|xm`x`+;|VxXt4*Yq?ihdlv6Rqb5po&Kx~VG zGI<=F;WkAvWc3aIw&V4}uYdXr!-5tyRt(&YN)EjH^U|*jeD939MT$f!x*3nX@c!M( zO*7Vkkhm5Sw{g7lGxA?gZcQvcaglr>PX*mgho=ZWGWEwt@!Z*n!71-;QF)iSTe}s7ayW z-{zTCn;B`}JxrFfiz&NtB_br_&9Zc<6tjUMiB!x}21Dpx#^ShRq%aamR5ZsGwNRH) zGsjk6u|pr_$Zm2J80dJ6>>Wn-)SbW!Kb~p=#+y5T^$ywW#%{WT4m^_N)8p-1f zkj2Jq=TE2j|ZzTBAVB)=+ zTN_;)UCLSMxyThoVr&3l-}@ZdBoFr#yVd1EV%MY~6I}kYxM3aH&d!dwac-*40$k-3 zQ$mqKDrR7Men6&jv0zJJBb`iccq)bqNtiah55jx9Wtqxs`6ba+-`zo{=oJ(2N-lTh1m`bK&cT@T?N8U2@1Ivb zU41hDorBFsvQUhmLV4uK+ebMDvplPqg_%V{pvMe;xnz5B|JqmdO)&w@O_{3SJmb_&K;nAJTUbh)IEKyI>;IDxHSt(M1wnIm2=S- zG2Wjbx*D;RhfnM#j{=}FN{0N_c=YhIqdQP1H5iZ?Y#g&LATP*im~$Zi}z+iU@o1d3Thk(E@;(dbUt#U^=wJS~~K4yp_O zFAxGV`o$n|*0i%5$3933y5xWSoe2=P=B|60+?qgESag4UN-@0@>7in<@Oen_4>{)> zX_O|=(fd`qzn}1Gi4xNM7fHS#ozp4oiCW=YvJXqbQ+P(H5Rakhd6#(d+%8QSR})EkMk41lVtZ#G;x$ zGq*Ol0sfdR(Dw~4%bs~_dF!TV6~JFnH$Q=+`TtZD>`=z`?$E!V{K@mqz>l3vxIST% zdo?g9SI$MU&;bu**)s<2(}VAQ_PXC|(4$$}0{WRK7LqR2+yrVqY(MINuTq?#9JA3b z;;Hyz(`4a;*nYuwh8?)yzGD(2R|Vy$f>#7;7m@!oCxw2-Q>Gt#^C05tF2kk|8? z>FsbG8P5~KN@Wj8BaLTrxrh0##)uCZb#kjCT@o$6T` zkm_lWfTlM%RM_|mLQ>ez-7ZHD3Dpi;6yWkS$uEc@&dNk?mB>p3tqL9+(uh-=nf!BT4 z@&;rlCilxf4R0fQb+@paKEppL0v6GE55QrmCp+i^K6RWb&wBD{_@{mi4)&&wX#O0+ zicMDqPmbs<9eQX_aCv}rd7Im2dxj}k8;Z*wSP&1yYsX$M^~FU3Lqn)$87pu_Rg(Qr zm(5h$AI14sSh=L*Klmu@BG+U@Jc2#q$TBzHxNNsD3n>({o+9h07<3juj^&g;$W;I& zG?kHU{L?c|%r5{p0tG9_cF|yg3+GS%Fe9?VMYrbSIqR(o8-K%%7Z*F*-$B*J=!FLV z6{;c{M3<$vy$dBLroeI{zC*8)@8IkVdm`1VlLNb_?~?V2?@fyL{xBpz@)W%!bTkBJ z!_1fYB^H>nSX|7n+e*TgdB z5&ew1*Tux(VhR~tJ0cHFgyc~RXQyBrlOjJ0RZpdywgo*PrW+pt^sW%`WfQJr zgDopNGm@9~p!Cf7f9;Mw8Dhd#S!Zb+(Xqpp8+*OL_-ht=-9s_EC{jwrWN~_cS2LY+ zj&6#`p=+jH3d{37FHe_r@GoCcVyhV_kaUqsnxA7seuhK^s0>3)2rcdkO zADEgVsFQR}>?3!ipzfi_QDxGnp%$Wd7S!=ePx?92Y&jJ<&&>bGPy94GVUo3y!Lu3{ zZj28*dIc#F<;n^uh(VV8CaDhi33b32he@DVftJ+D@SI-VD(Ux1;2;%ir*Kg0%=xxE zBO?KV-NBrA!MV-5W35k)Du=F|3u4>!Ug$Zj7S@4;JCttKgyeFU z%s4MR6wt}J0Yfdq#BoBx@7}{vV4nA zhbk*VuRiJ5#82R42Bk{|WH-2H=mzoC$eYr3pj@hz?G;yf4)E3m>ecaqOG2@(IhS7M z*8y^#hn0Qd*TS1!ndBecoQ&ktcI+8ebK<_X{o3!>{mwd;=eEIrJJC5Lkwg6U%a^~@ z67g=_>-x8!yqXeI_U45J{|)%1`y_gGyS$py?}c|%d2R{498?O4!`+-x7qiGanlrXx zL>xomi0@(t-*3F}ozRy}CdEUf+Cy?CfWp0Jz!B$Qn_Mf;iq+NC~WPMacbi3?OK%viykSwuI*%OsLA4hn?yD4piWDc7Br(-$W zeHfi}F`Kp#PlW9U*-ekzilKLXy#Kz!gd6p-w+@i4&zX7!(B+!#j7lj6T;ZKm%nrJl zUmj36xm*=1(9y>gn}Rj?LJQU0`~N+5p5~HpUo`Hi_kJB-#Lhb`)6uXUha$@%!k%nZ zE6~(~k{3!rVPo23AKWTy+9L}=v3F#yi)|Q^Ex7iOOgO=V_i+DS1W;G9FnR7K`QXcgDOR;ad3-ZQfr_E5kC2WwUqr+`!Lb4YnE zWUXgaq)vGR)WfQIO_Lz;1lb6DYW>7s4&Ie6IU@K(w%>U@WY0|LC=KN^<2mcZ%fGtO zAv2*RbL#K@lccaiiyJdG0U3fBT5>1`I$blV7}#V%6(w4jc3_TMguKgb{5#>eN+xhH z6o_;!HPd!P9t_cfMbgnItb-!DSc%Cd5I3SJ8aYlz^mYzE^t)Ayv#blE+}6*vQ{vPk zzsPwI{wREv{PQEDD~0EB=z1~A;G)Bn7gRItK{ym_gNRo$w`xA>zT-U|vTYJ)yHMs+ zF!I@lHO71(Cm-8iZES3ysSl?9SNmv+Y~9$V*eSAY?7~tXAM#U@^G|8%kox)^p z7H^~I`hqmdu%P>^&=@qkLsd@>%c?xjgYa7VJk&p1EV#&NQ6G`-4r=^L_t(#;b5*xy zVV2RQmS!|iaV9jZ$Ld(}YsPw^v16gPF$RBh^_r#EQf#x8f~tbb#C?bxSj z>q*z1W3|Y;Z>ImtUhl|9)6L<=-iMuT4kY7;8r@w1JwPf6g$A1^$Itx#$Ts`3Xjfe9 zvJRuU?dHW!$jm6`n=;Axvw*_?b`z*yg6=;^7b`mD=Ov&hl_Ebgw$e%zus)keLpTavC0%B98zBfoc;1O4r@}yBQhs7xRf^;jeZ>jKWQ4D&l^~{QB(9CG8{O2)}I~EzI8eNoi~KvH6bL6>i;%5!EQt2#tz9R7TCE=F|8EQ zQ!&eU%|3}d1F##kDDbDT$8-msEUKJ~1rN7l5NlPM1d&13DUD~c zBGNhd&_z&KtQIEnmW9+nO-<88jQkG0Z8fmDB=Rt_TSJ;A;$;UMrjbGKnRSiziD$uP zc}CZHYWHaBx!o9Gc8F?Q6!A=pqK1@58Ke0I*;47EK62dq^WxkwFh!rw0vlv}{*>#F zYX5`(=>(JMQ5)XXki~9X7ME#ZOp+)jfg)=lUjw}xu*$T{uZi3E>&Z&aWmO_?Q2{lk z6Ff9k4$iXcr`75m>@PhwqVwa+lfU)4H4CZRZiejeidMX-U+|}jH!lIrxmK}KeZ~)r zy%#87(tEgPq(c`!f%P!07#USvkSj)Ta+Yj8= zPk`tylg36h9ML&Ya}YjkmIo^!jO-;V_)rsna_p<4*%!I-Vq<4tglkHx1XmXg=!8RE ze!n8lqIdoAB}zz)LN-pC!uT$vV35Wz9yWfwy(q)rdil zelNY>QYro)T~bt3g%xk}#LXbMSvS2(aZQeb4TB!*xV^Ids=`REqRX>@vzD8{x$V;~ zUnbDasAis%pr>g|0L=rr8`Bc-TUL{rx|w`r#ds$?9z7 zr(Poo?DpDj9H%L;ut!>o$)Lz4D&{tCgOSD;MaG^oYHJQD6GWxHy8@tIvMAEzLE{hH zcRR4tW&gojKjE$b`O3iH5l<69-q)P{BU$;JEla@GWL9jjiDFVHvL5K0fS|TjQW@C| zzU`-egP^gyD!PM?XZEY=l?HhuFwu_5w0{l*HURkiDc2v_UtZQ<`QBHpd7PeYGjwDB z)eg54dMr1%TYM|#^oa9CyQU+V_gbbV1hYnGNs(|YqPAx9$^sZ>Pr3fkJ}~}g?;EGA zYagF0VA!dLM0OVJH`WFg33mlRToj#FJlIOs^NfzI2AJMemxGXQ19`yuy)=E2o~Q)g zfK-z>u~mrWP5Yof(N!J?3#{1oNL>UQ`&f|KeIEoCfO;@}s24anFLT>YxX% zLAw77R$505h*Rd8#UO74r|P{;2ITEaWFbkK0R5uon}Mwq0~KN!NNR)oyu)A3{pTkO zAl~HJ0U496)7tpWQuGDesI>cThpuq9&VfNvlHB?S0ZtVR)G>z znoW}1;n*gD9Ol(rJ?9afFKU&$ppJ2v8E|N_pLzyLtMcav^Ycp*q>ZK@XIM0?q88Yw{a&{`TH^=>Kw_yUGeAt$pstcMGI6&JYOUTb|&#qzOFCRD|1u;8_y&YoYV>WM%< z6w5?Lfoc%2+&49a+b`GTi*lj71Zr6~LTg_-C}L^@<0Ejq{i=8dK&6}7C(-)WOwvG8 zF{E_P64

    %?FaW^k5zL9OtmCk$zuee8ZcZj__t_oM46Ewj7>pmR56fA=D3J--Lv* zZU%IY%&1p3OZNj5^9A@G@l493MT@NB|0BZA=O}XKkp< zg&X&L!PzMtevidZX(*iMUs!y&!@qC@62p(+nyaiJF}|z1Xz}+=#z*A$`*Ebnjg8M8 z3*&Q@VlGkS0tl+Z#;2N#?B#8IXNF~36w#^X^WR=w#|8ss-d zv+)>IfDJM*SuLDq>G9CjbCABUJZeK&J}!O@q&f^w){qThcmdwGC1f+yzSRhezz9|g zuMeLjas*<|G%e<14F^tE5F7Wq;&Am_CWyUpWnvGx=*AH1w}99u6myFrH>j9yX}Scx zl|tAj<%>`d_amq*fW}h1fK3&~Zj4@GvS=-XvJRih@|YqxS9=-=|2& zy6?}&)nKh%QdAS&?*)50=mWsZLp~U;f2hk_9k$kQA5_cPiU<8)!p#hN&mfN-+@g2< zSFv@qliMzGcIubXzHvp2PO@g`Nl7o zgSqMu<*FLEW3mK}1R&4MrVCE60toffiN{*3`#;$1$lN#^ZKqoV)aMn8fWQgpobpAP z^jXsFn;UUztZQ6w;vGbg3otVLkUJ*6>Ewm91o+rNf)Vx7Sv84JxCB;4AJb2z^)7| zj4YNtj7F}}E`rW(@w_tmLy|h<9ND9)@?y)snbQE} zC{WCNhG7pso9R&=*awP@eCW9EuPS|eya_x@SA-rUgC1_1A_X;e^PKrQidjpMRa6W% zUL^AFd6h@$=>00qCU4hOwGBHrjv!*#0bF#E9U^YgH>X(Fu(+*fYNsk?AyR@~m-j(# zAUmpv;n1vWdhgW{m;=E>n1 zv2C!OH383z!)iv{*TH|^ci{!=z3mIKb+NM(#y(RWw?__+YbM>4od95`)YtbpH$%oN;1gICsW74?o7Q8aBRSEz|OBRlfTI+@7Wk zq-MMfbbDkX536Bo2n1JiMEB(Jj6rgUToD`i!0@J3t@o@6K@qubaZ1Ecmnnv6+vSC% zo$MCXMJMsLd)tb84gZlkdUa%wFtdY>z7fm&?3=#d2&R9@ntj%7{SiBSvX#pGpi+_X z-xhTzZ*Ra_UPig1 z0R_u>W_7!LP}tbYP{;rGpT8exbXe6oJ!<ByjuK7wF=NZvT)We& zo%S?t^6r$iB$HiFkQ)bSt1M7gOflf}<^k#!ycb7xK%k{veqmY@qyslhy5u=j(z`4q z2Wl{pw6r>`gFY3#BCthTACfz*b6OG)Zo%vxK5@zOu}>SlSpAW|j(&PHS}@s|bLZg8 z#=xUIweRDSfM6418edyqNwVG8@_?Z+YgOMxF`z$JM8!Pi1H_#>W`(P~&nY$7i`qtG zPWN_blML8(Py`djZec&PZ^1*ZrdZZVpMPx~cf}8}0t#jy`%g6m+#_>RowtFZ&6GkzbDJ`&G#LkUV7 zfBEbj`g-^vM1NMR6F9pBIfA=C-}YL@!tMnv>Wsi%@{r^RY95g6fl*yPHBW>nXAh++Dvec~(># z{q(0^{XhnC>w(+7KlZBOIDIO=U4G8foL1 zDo^9}1WuQ)@r;YHR8bp0A#CqFtpXLha)vK{xBPn?c{TF6u$~-8-p%oQ z?9YX*ud&+}34Qx@n)>1zNjLUM?9@nBe8p&A+I>suZhDhnRoHfMDIF`QcK&Yn3o+v$ zguiecXWw8S!vFhI`2%ZIxNQ>A4g#={CJ(3-4o*2ZrAf9S7rO@kg4ekc8!Cc4-Uv7Rso4T;PB5bNLTXK(+|QY24s6DVkg;cAT$TY zNg(*DR&Ha`LK?)Cz;At2R!#SFGlQVgrF>eEusjN>tp`0CWSAYO=GOazV%;R`|BxdL zdY6a(H#)#Mj-O{N(-)nvh6JnEGv9rvood3*n>&B?4%zI+n_mEc*@mWoV)7`GOT}O# z;Z3qTC`JCkdw+%=zhW*36GCh1U+@*ORYS4cfDBs#cM3i`x5hi)b0vlPgm{Ner33x3 zc6l0P0GmNq8{7JvTleNn0I-_jFSwc2IyCMr@%QSj2`1gvJF!C8Qe&|tMWrY;#-rh~g*1bM2MwFS@=?uC$`2wS(lOd&Cq};5$Chrr1+3~mr;;E~m zm(SKJ3g_IK)hUEJe^tBuo>$GZI@Bv#~CX3SVXqfB$=}|NXpIQers1N3NSuD?@gI zor24pQ$DBYJWh#smFK0fCT1r+!CSBO61^L{FTED-%T0(kyq0nPigZmoKZLfG@<<4ys;exJQq*RO=%t{Rr-(|45=Yl)sJXH zhZkjBE--*R$Rs0mK{dY%{M`7!C80)6h1;PS@D{cFM!4R?jSa=5(I*p&gr}g7S+gBT zgKFj^@XCdn1YTA|wGi(Hml?}mTA+|-k95!j^)9m_@HecRk6G(}uLMrkq+#GX8mQSY zGOSP8Ln$k$jca}VqZZ-QomR%xUz7L=1d`8Y4(S$(NvFt0DhAogjr6#{S8m{g%+e(x zXr~)FRgwd1C4G|Pihg-Q*ma*`D7_!%qqFy`kL3JFreP%S!}b|g)8M|X76r9Yzqf7` zVdL<)?_XxGi3HiRks259jr;AC&>KLBQZBs6X#?sn)N(c&oF>_lU?dZ&nKS6o4&F{S z(5~pF9#D0Y#2H(MqB^cy=thEhGyB38$`0S3x9aZ~ti9uBYvtTHYG=m>-YM7yK5z{& z&=17m)aVsCbgcv}+$#Ahd5XL&ctcplT!TdKw@SY3m!BCbWZ30_w1!d1YrP~#g5Wq!Amo$^Zh ziuhCMty!*HW-P$qFq%HI7Z#7Z?8Vy`KJogE$$=gJT?wQFJ~-L^MR;_N|jf5zY6SrMT&;79zb1#Gz3_GYdpod!Tn9 zWfpSF!yUV!>8TOqF%E<*k7$whgzNn-g(Y&D;4jczgJR%jX^SF9WuP&|@gMAw^ckdF zdcevqjhmeGWT$ll(sSkO*l9DuB`ll1&AZKg6t3x%LLd)FF>&Xe9Ekja{az=5{yoXN zI?zZ$xEn|-F)4LbTtK6Y=AnQDUb3wd^Y=PHcuXqB{?s&_qiEy4vIb>&sj(I zQS|+oaUKGbrFyq~(*KgR?2?*pyp_tgh!||9m^55=V_GH4US0~e1%o&c4n!7A-PD~* z1C7MqnuAg8a=qWZC{2efapIQf+=#JR$wmMiXJKR5alW!`^PAQpMQ)pfvZJ2V#?SJ- z?+1OV%p=;se*&Hg+Jh8%Ii2MT%?W{5_y&GEEEMa6MS>Kd;W!~~0_!mr^FD%EcXjdQ z5E|^83meWbHSRa>{EIa;_j9%A*rDr2w*k#WX6VqKxm9vhwRGNCz{2_)=@3k;9~y(# z_F)qE!8wujs$$0+=Ya?Ikd#Md(zjp2FuW#>DdzUbPKPWqYR9#6>LEU{Fm1-UnG16S zH@vLr?OpPBI`oM@{wHqZi|M{0tshC&MMs#N4%weITS&33a)0;z{mF3) z*P@1EpjWPvim9bTUvLXs7 zWdFe)X{{n2n1A}c%aq4MFLH)hqVTbDVU;)y8j9j0u#GM@wA-t2a+-3d5^PeO0G$*k zSfr7`{TcjQov=Xeh+f!-`_7MA_k82&2D|4!J|PF)c!Pc20xf4K<`hLvB3cT?ZICf6 z;ndQ4b?+;wGj-uI(A)SDG+d7CF@lipW$M(z}(&d{riW#y44NQV!y9bS&!PHseAn19Hx zhpSibnF;qQTVLHqcTQOcuWKa(vb&Lwy~-myAhG#GTuU$aiWTf7EOF%;3|T>H=p{QS z*~gF#lS!>Q<`MTR6Rrw9XTMCYx-qT>EO6CDF`X3on2IU&y)M_QKc@B3_2S!-59yC- zkQ@;jWS#yQ(R%fX$n1zzXo^9JUi`O6IN*T|12?<|Jg$O}?|{eNkofQckNltx6?n?z zt|HlIz@u8YOV-9;$w~F957DdhxkaxgaBd6hIKVg)S<3V=7da~+itP!#AlZA zmhn&w2{+Rq5iLSu=5~2mXdcSawkRNF2|^~CR!B?biQ@!1P?(41ORK_a(o#+jaN(iw zJW}7?7sdfyo>oyQzAcBWlg(+j^^rSx&S{8o@ckUUjTK_X1y9|)+PVc`6 zm%W@8GNdx!DuKLq(?o-$RgxUoCqaqv0;(n?Rsc#@*W{IvAV9Li+gM^#!0nS@lo-O% z1yKHXle5B42Yh&)!UPUJR6uR@&rqDAH=}=)8Kk4PNWjAaZF0OcdNer=Kb8i?QJC1h zj}?B#Kf1jn;Cb)DjuT{xG_{gydJ}V(YvebBmsu|r36W!Utn->IaA9{2U2%aGF2=Rp zFAuZUdvsg7Vn@O8U{qlwb{WJ8AmERUy%l~5oDR%EL9&M~qSI7+lm}*Sp7O!M83|}~TED&O58@_PZQQyE0A@}NkxR^4z$4|VUoB(f;)c|gsE!b?~Oeavx8j7s^ zoZPHca$!E!RsH$BF#yZ6v+4+Zo_)&oXS}L~9bYp6;tyxizehft0GgWS`Q+0S(?mhQ zBnDl;0f!Kda>H}Qq!p92ijTd{%msZiY&^qyT)Zt?b=wEuqDMAVQQIsv z$Qwx;|ADwC3iH=kA-69QdjpW0`CQ~B=xk#*tBy>E+1Ge9hpgt1)tAwI<7A_p+IH0*>2ToRqv3aw@@V5E1>rVdq z)Z1rAQ}^q}WmkJGjAlN?=qQp+#Vng(fC85zoc+q$kmVCr@SXryZwj|b8ArAPT3Xb3 z-)VuuEMyE{xIXk8X_6fbN#|_jG(kVp8WlcvFa%0K75kM7U4|it!D7b>x(Y4>srN#y zwMkwOBT``jmd_+s(#WBJ%kuSt3eOr+z+D@iC$EFrDsAA3smal+AoSP5$tDS5S-g1W zsH%7#2**wYU6trkkyf$IzjOW*f)cSVg9%RaXrSUMnBcf6GjHVnV^KaB%lIW_c4ODZaPreeCqq-Bk9E5Se&>WcENPqmAR?#fg zVAn~W&nj}V}tcHkLZC{`TqGnCTx7a zZp%5c;f28lhy|EoV>`uUQJ~!kj51h;`;=5DMg9;9VKfH1a&CFluIXS*vQ_(4@l3rE zj02YcL-vr)jm(QIgGKho-*W}E*$*<&%rYBh(H z2_gf(E}$dwru5K6*Ft6FI03w{nt66N|LV6jeb$w^c8XnfG=n6L!afaBe>?`3ldtrcDnqD!elzs$KE-^prj`An}UlJBYqUYNx$r5P} zgv2woVKuOiz7u{atQR^aaRjHK$SKei9`kN+m^V)|!KOskLReHOzQMqYh zR4!1=If}G^y8M(vuh^-1flq(7s0V*|;GY_Q82{SxA7eBgr6J1cEmL+uIc=Xr55h#9 zQ(Q3jIY||b@U0*e*4=WdYZuXzUJ&7$x1ga*3Gp5(k6-lvgGxk5HQ35)$gT2hL=`Jy}xFsvHBNoC=mOuzc8suk6FXMjiLKpO<9+{1p>4Drfxf3bNCUp>f0l8v7{* zl7YLa80>tCgK7v6GT_|@9xGUk+W1E&qoqL9K&H>0211Pvxl&;{KCEMppM zvd>oad!JCAQsU{e{MB=|`_|3Rji?WakI)=fWKFE*o|WX%J=}beu{z3mTWOZ)v0)gV zIpq{gu|v=Kv+pdo4*R=pA)K8+{+b^gRdqmR>{8A;=!#De9hQ{~yGaue?q-6ctBdS+ zrmCU?bHoA}4vgl!TUa53YJBIi*YjSizf)NRv6OwUfCy@6?0~C*Tj{A z3KMLRt^iOJ2e$Wcl%ZH9Jwfz}OF~VlES_1(s{s+X%t=Oq)QjXm zNGE-R8#@oOi}lJ*uqd&icg6Xlbk0T3N^!#sEIr3n>?0_s+)9wRY*#?Lyqdc!AS+@O zu*>7!10FrxV+7cR6M4pmG#f=9a@#qHGs@tO<+Dqai#a-uL2`D=*(oJ2mAnK*i@oH) z>=OhR%1j_h%NJqN%J_DT!J6`!4@oh1j{x*jxLFZd5vD=Qe0K#P{hFTOAGNSWk;r>N z2+C$pnmRqygua{Kxm`x`*sX4E94TzHFkqDw167*kR16X)>eUI%3A#y^$0_qIknE%f zX179tW_s|+1y!QHsPpo45G}td-pV}=E7>*qJx&upBO*tT=i4bA@VFv=EY9;iKO}E( zkCRBw%aeHRoCMAxf_Jua&cl+J?$k`3gLclwDrMu86D{6dx!}9ja&2~0(LfObcn?;| z*U2*!JHaFP>viAv_sld%QfI8?u?UGe<(GH(<8;cqIX(!hWy5{HcZV{A4AxDHFG!iU zofbzFmHptlycDQdbGhk~!*om7*|27Qm1mpqAg4)E9-vqM^}756yt-R_FzU?Q^60x` zp$F+O&@Ie_IC~0MIh_f#UC3^ zJF8i83=~!mjls+Nb-%hMGC3k!`Ki}Pf~~})8&7?Kg}b4pm<)<+qGHZ}SF`pz;NxBn z%21TNp;`L%*m-z}>Ux77kG`fU|EB3;BaMlKl@$Kx`X$jZ7+4qveBp8+*k1(eUd#B) zXnI22xW>+o45$Hi*X2ZmB@0;Nok>5UJB262jWabVvrtrJjO_;-$haIQW^ZtLLF8Zl z$z(Yme(meOB=v5*fxc>CB~DSyNs4?##oU9o(oA4O$W*QXI#fOPGHCPd6`vPY0$bhV zd7V?PbM{1|P{0~e2OEt$Q}6rfxo3sd!o$!-*6+2&w@r8_eA$E^S#Q(^(SGIf2`%ae zUb|;jiYi4nyy`d)qkH6M)Gbin&<#zBr$sx(v9Q6Z<(K*Hq~pCSq6;N^qOU^*)=t`T zlr!oP7$P<>dB(mte8O!09XzyeS#0F1sWpo__11)$YBJ;M{|6&*_|PvN>;` zo>NFlCjf(>dBC)eVyY>!pNe@BUdqWCQYbUZ^k%Cr0@pnvE}&M@Bu(|!gUi@B^T?OunQrkmP6lYU=Bcg(Rnw_JLV2I;4k(Xmn%E?*mF*AG zBmqf(C$Rft)j{k$ELj=BSJ&!;V*TMe_U*NL7xph%eUshSV-)=8Ut*tkDH{Wg<^D() zw#Zmnc_~cGv`Tg>7hyQq8BJp>;K2f~EJhvut*r1mz9^fc@iLi~%76OSDYC|m-5C&} zHcJv^P|PL@W+BEfL?LsVw;`<03mO=(lL6>1rJA9CS~#eLb_X1+9r+OP&qV3 zEYM-HV-5{F>YLc19Hy_;4{mRvxK}9DJG6M;aLK7$0 z;)@v{gMS|+htefbdWng~W+~>EP-0Jm>1eG26Ujr2gDLVB1sWP-p1Na5LQT`c$x&_b z)x(1w;my)G0TPNirJ#;Ig2rY+WQCw{D;DWOtTR4tiy%8W9Bef!jKsFGLN3gAiu`hP zA3?IIln5+Qz>=suRr=&1lGEbJIrO^M(kGuKMrGLR;h3c?r>nxCcnZ^bSrNtyx*5yJ z7RWE3A|OhUB2S*kHwzlNrp6^xt$5XS~dqE5w+H#sd!SA-sf z{5nYc5YIox{QCDXfB2ub-dXrZ@p6h;LXlY4#}L2ZSB55M^Qm4(F>5KZ3TV1PSj|LS zx?hFXWsURNfxh5ZvbgNX99ujvMlV>Iit&=g;|HzXSUcV#R=Re8Fm6&*PgKgkVLkLC zdda-Q3eBBxg9Hm)&rZ?NB~bY~hDAn>OoI99K1OrE(q6mmcknf4eUvq!pxfFDI~0LS zr>uxTX4?4}WB<_5~kfE^x;Xt3%jh0obv^s`2+?6Ez*5{focPwT_9f z8NiLhuE3jMwpC81m?VlMP%%s2Y~pvQ20T6uz5t3KCseuI<+FE3UJ2UHZ5d-`Czvir zK{Tcaf;aoG`#<@O+~i(tObR(qQrT6hxpD4qmjx_xDF&=f7F4Nu>bNC9!J{ek%%pL% zGepr-Ge@JHR~9-%&wntaU?|G6*Gn%0oy1rHc0yM|X~BL~0>|jKy3R9o6e#9Pq}j2{ z@H6pGn^#!J<=u8&v6C>kCjS&Dciw;R&sa&jaO*T3=W@^)>B1Iu^}=JXA6hW@_8Cmk zK)3QWvUp;0;B{%){DS#fMLXw=bkDqgdAt0EI+^-Y`GR{320S{0TbVz#!NG;lKb?Ue z>!?2!7+*CQe{6eI$1P;aWcRuKP|U;SF~UN(|%$ z+>-eQ{!MW?-K@Z;c0mnwirn(4^`s!`jPwZ)N7gEg<2q#tatzzHZHB$G&Hz6vY*Tx_ zX84LVNtoNNO~cTB;hv{e0%{&j((<;d5>q{ z7ewu#vmzi=$c0|`sLbF#63^s9Jx#f9(c~U(C)Cs3qx-qF;7Rv$?*yOpbIL+wdx}_s z#cD;eJ*G3x?n$~FAT;?bt1_dK%-S)$;nkQ=0^w)r64rV>qx=ZieEy?+D^`!Q`uyE@vs2OA`z>oR zXEsq-Hx9R5w=grGNk8utWX!|F6H*^*j;ge(9@$}8rf9dUiw333YGr0%j^McBp4Tb5 zLp2EUyaiGHTw}poJJ4thX@g@$K2mO+4Kqm4NIiW-t0>?enS+G}*TV-b)?`q>+v~ift~{tQ%RAErL-c7WTY!Y%o@mf3Ql4m2{o)K{ki21Qt1=Z#A z1vS$yg*hp+7`Hc3A*7P7W1Y0BbA704sC`6;lja z-2<|EQpeM(vOrV!a!>;Z#J6zn$xD5=N$SZNsa}l)x|^q0%Dbi%LYHBd?|oOgx}VM4 zh|u%d)2=(nncUz%-Gq_epGlXKZ4*eX#fMu)G4N>@;J3D#-mU5YeOfF(*cS{Mgu!LO ziKG+C4>Sk(r$GTbk*C3ZDXt560)O_>r$w4dI6(G@BH@;4r-TL6HIfIKraJCcxCq+E zP8V$&(Wh<8lsox@ALhF%CcNMM^6i-44VC`Tj0d6I*DHvB?Q8%1Qxj+ef(`#d&apEr z?)#SkIg5GLxr1WbDbNFr!4ONmGDonMi^5aj^E68{f#oS*gw?cIEL%&jCI)&ppxFTM ze?$&}82IMNn>d3W==Kb8qLxEwre+QhslX^&MNPr3R&Iosc~Gg+^sKkr1AEs9e@bZ2u^ zdS-~+^sJA8uBC5Evm%pDM?63a7)0#HJZLNAjx+%R!vNW1J}$oi-}zm*|rQy$2&#A|u&vaMcEcsHeuu#{y2 z$Jpm1Ion8$5M^e7GaWh^We>|XpZhlxZi2U7>m_dEAmJv(7%0+4#bA#|?!;1= zP7=>_(t9Kg{#wNmQUOH%nRJJ&a&8fQR=7*7RdkUh^R{~Jo4!fX7mn>kIdt}Xy*kln zD`W&>1-((IY1S9jp*|1Y_V}D$y&LidAO=9EtIDGehAf$(=ah18&h7}m!Z!dx3(6RE z`Brm~zOqt$TmJNYaQr&XO`k&Uz{?wF8u%Tc?|UtLF;A;FPBNw>h_XN#E1t;$0nY-z z`|z&a()-+D@#Zw}JPW5UJTn~LSeYoQBKy|K5EInOI!oh-&W)i4WT$2YnR_S(O7Tjm zm|k*49LL<6RZn)%Z6t|@1WwB)>)l9bY@ z?yzjaPQMf4)S&G7C#Cg#yf9oov|agE_*1OExHnqMQ?t!QP|fB8>$CZ}@ippx^jvcUxb9VdZiq}Z_g96Qil z?2gU|Gnt!JDcvQc@P+Y+j#(I|DvH@hkv&w*F5ml-1AG*k&wkA)H2=5ja=p5Nq;uk! zBDxFeZrWd|4md@hk=D|?rl(FTqLa7-9`}MaN$$g9iI3~J&paMG`*QTHSq)>yDRvUtEEY2J@ke_6yVv4-t>yUK*67&P3B-_%Mn`X+96wWo z1$nn$+U5lE57wWlW0TB!*ik>g%A|}F%A!|U7hkz8BJQw;0FGhf>G#{l04_Sk3LE38Pw1abF)cTS`QcYc z2D|0PeG~o!@|b46@J@=^L6ICPW+P{@z#vKFL4X)j9Jrr(O0{4VI5NYta))ueOQG6yBu(_0jpTM^A1Ea)3{2Q<b*wJX<-4^x(O-yWP0}t=TMJGx2#fx%AvanzPcS+ z7?c}`betWk7WGEXJ;i#`p=w@mlZ=Qc(~-%txMxI&vKYZNFUmG-Uv1qm_kxrQ+Ufhl z&~l0#h3~icVs2YgG53TxUyvN=5)WtCS#uB+!;ayKi)#}&kc(x|Qja_0ptPj_H z!4C!}1xxhupbrX^Vo~!WPrOHxz*#z01?W#_%)!|QujJ_yWAk&EMdmmA8iP#E%6}F& ztRvgq*jcHwa8}AGri3DekRpi67l4F#^@8`}s95iPK)kxB#Ir+H$;H~=LjegqsQT;$ z-sTcdpnUac5GMpegMMYey^u>{tn^&0UZv71@&$K?REVnSQcklJ>pwAq(*ZRyPBkKq zBjk%+S7%~p7=+j`6Mj3@KeLYXy6tAlPV5%D+%o9_StkT+yFAlWH^JduPj1aBQ7)gI z5SA*rF=t8WU}R>{h0vAiB9!$6kQk5q5Vp zrWnvJ(gO2FWF1ImqK=`C&Jz!MAY1cOhN=~wSi4pdy$tm3D*RSKt$({5UxBJGqnA1P zn``TX8tH7<8q1mI*nOMbmX>PA-6_wM=RIqU+_>D+j`nn`#30ku1w8RXK>%vl4`I8>_EK|L@Oex8`P@0Al1u`;(=3aX5O@zKs&afY8{HqbG7G{X<*q^sQP z-~ATfw8Ff4ZsUh!wHy1j5ZyOh`qC+8BSn&_m}NX9)`ehyHdJ9l++Ig(RXF9X5?l!J zzjk?*KSs^QX0Bms*iN^TX7dY&v2Q7CsELf*RQLNXUlTZ5ep#`Iq&;ViZ+k2NlSeVR z6xsf?@hvMNk=FyN&e#xutu)nPNba5@M?I^Z&|*~ib-m0w-`ycZ=Hpz8yhL*9GuJ?s>S18%xFK{17wFS^Wnzz-G?m9$6kurO~Ljnrs9{! z7BunqMYr*{bE~E$aH@ERB3DdZ@ii=^OqzUaR(kM`*Umwrub5j4GQ7xdxH$MOv`;tE zX(2`Q3D5Nr10JX6eyZs7KJ7i=abJ=j^f0=fKj6_#b}8^EGdMHYsep4FP8N2==8F+y z$Hs&9;WRCyum8&?==|)~sSL93g{fK4TY%~)#ei5q0~K>z0eR!VG`|v`#svkS8rP~f7|b6 z>H_88;Kq?2s0+vInq(OBZ;J(*61Xw<)dC5;!`5lozYoT5YAa|rjdWM#Oqkik-U!3N zcY@#dv`z-PZ4}N8C(p_=5tc22ifZOCES;>q{K z+4L5uTsRb|m${I)%=nES@vPf02+llZ-?-RVtHz!0OPFe%_hGY9cH`iRos7^j9&D8` z@;~&&2S`E1U4bM}v9H%F-13d{je`DTEXq{_3@7b{s(A`J9(U(FL<0Q4vE$7@({75nOOa2gm^RYQYm%)d<6Q1qa@^OT3XeDS84yPkaVv2uc5y<du*6^o=>6zhioEHm&1iF%_O!A;YMISb}IJG@L;Zocs*G1&FCX(5u@QZ25vp z$Y-W`#3xwecEs<*?m>3AeXrt`P-~(FHg1d?$NTN%h;TP^gIi5s2fmIjKI~;aQlwAb zuc}uT%a(fGf<2G(X>Ep`KsGoUb|5ETO#k%pRqIlO7X%r03Ky_GHaFswP_vROpIt87 zf*ajS!Ztqoym$>Gxmf=KJp#HJ$adE6{fMB7@j*@+R6IO_aBsZ#kW62bY(ub-{iQ)x z<$X@6+3M8;X(i}Z)B zSa6QB-e;}qG1N)a2J8jT>o)I@&mK;mB9&VzZ->_Tgs>K$9i05YHh#a?p@5GdF>R#s z)hb$ix~D(u8;y1jES?`ZP%MUV`itPpWWn>!7oW_{g%olk0Wr-MSU`$J@gmLHfa?*+ zsg?`!HTzT=>@VsgNk9XI^c?L>We0$g%xwRBmSRp*q=|~bjyob5qc#ddGqjGNL;b7R3c zU6Ch*T@P=Bloq~YZXjeKUTPQj%kzR*8K%*Jjm-+h4x0AC_Q7Aj^O`jcupR!~!@%SO zGXPuVwajK^wyIrz_}f>+ZJu#~Nx**6?3Wgj%iRo3UU;SUjVof$`)U{z#` zS(Z>T-aZnbkLZn9KJ?bByThLM8i+pfRHEPxRQK;!mX5h3sOdv@WRh9Vk5g`EWzEKW zR?p^Gv){RGyv`1v-ICCi-?}1x9~oOxBKC@txqYOKpB{XE0Y;A-fTtxrIQ>=Qag}F! z@ZXJxpD2@fdw}*CX`nml({w5~kDDHRHslzm&M#ZsAwuFHqx_$9ZH6PbanLPs3SOQL z!U{LkiW5t2J6QRj&WU|Bc3vWH?^j~y)zV4P+qo(7oAg#Xhnx5^bfVA8o(}vfNzq36 zi+%L{s6i!k0X0c(cy0kMTUTyW;7=dLY&`woT=J9EY`AY!G`Z5?0GK* zdS?j!P6T&^zhznpw|ADNcUI-x_{nWRhL{6NZtKI+A^7K9Gh;;MDB%6M4RQg%?ZbQS zt5rW4P1m#g`sbiOY>uV%6tj*ZYpIx6!P$rp{Z4s7)5!E=2I z9fQjk4B~>z*UJ9un-ffCL~VFiLl#dUnHHaB62&A?WDOOQ5|JsoHEW%`n_m6ukW=uN z2Qn-g8T(Juu@BRQ@h&?PVgI}LTy1w3fB%CT>&_5+4H$l%UblVnq0n{6@FLGiIk_Z+ zaw6z6#0*Pvpg~eD>ywPp{tza{L2$U#2;0I3D{#1PQQ3{v(|+e~G8n4(RW)Rb8@CPZ zx3FY|6a#e$Ix6P=^g2$p5EaSOLYnx-SDnIpK)#J((-pie%B&D5Pf%b9_fl!|kRoh? zsG7dW!Fxf>5gO>ViWR&gf_9GUC^T3D`Ey7ck8GjP3FlY=nYy|2NAGx<@KX6t-#SIs zxG`SxEbx*+F`Foug&0K1HTgDZZ!d_bVlw32)Aj0D!DeNapD##c)s8->0p)8*4&5479<^&aNC>N1B_%WTpf7`6 zXZ=9`UZ+YT)pX`06j2*1--<1o#>%%+2KLukvzxNPhx@Lhunw8;S<)y5f`c2V7?cgt z1*A+G5~|V7SPNVk4`f(#@*%h1E0eDB&llwjT(hOM^ZjOr4LgIl;B3!%uFGA877Lc(q#XFB05v4ljN$h=cmp&@=6~` zoOqlZ4AF5;xg^qIF?H;~@VQ4~F^CIZ*k5)w@BQuLf2^>_!V5Dmh#t>dJ$K<1P)*wy zlIoSiEdw>=OxMH1dS)Gki}iD^ejV}`&eZRkywWVH|J&q*t>~#6uO6RRtR9ysrj;Uk z>i@C#C2&op=iZ)y6OtE0HUh~RP%!}nafU^xh&RmAnJ%~M^xpQ~+upX*k+$P)r)xU> zX1=z#ptyjbfC?HwS!59eR76?ba2FU<6jv6}VNe7`1{J>NNrI9{BnJ|{Xm96NIXUM& zCy@7_=Y8I1`Try8gw+BaD;Mm@t3Vi}6+)@%RPS8D8Cs1>FW4b?)}ehyqq5Hb5HN>A z0=F$_$E-vAPmw)zSzxR?HbO&(StIA1%e@(=#dZE2{9-vu(qos9mNgJ69(VgJ^Bb~5 zw%xR1fY!Dlch4!L2vMdl%4Y`5=r?ZY&7u6cjE&+|oZ0;5xd=n_V zq$`D)!4GM?D?yPG_*_rNTo^sIH&qp>t{DO zAMl0VSWX?USzb$n{-y=m)XYK=dowd$@I1M1Z~sR#j8SB>bp(9agM>a^5OGVE921oLZV6Rm-EFR@+_#q-4f`IQoa++76|dXwuy%p;Je?;% zQqxVI4@i`6WLPHbuj^z>y|GHEl5Ut^$B*S;ue7d;zZ?iICz$uP5-t!)Pv&?$2d3Ae zQIZc+{%3m9;(xA^Hsn&*rX@m;aZWHjo-j2S>aoJWCI5$`z5${~7$cDmMYkp>KRg=4B$0Z6taOy9)23;n|R z7FLfN`wuYt10DNO79GLr3u?P-{eRA6{DSmejL;!`8`){Y0sYeken~aOK-I}UQ0G*h zU{b;|f^#@nC7YE&UoYtaXDN_ zn!%g-%onV~XHfczysIcKU8le6v!FBX(T>s7qD4(gop$h|Ma80WZ{(=!PtW|hMj2&aFTE;m7Z1v?zhgNx`IpXl98d&^j;AceB4q`*jT2yo zx3%Z}t=>5F%Vu|NyVlhd$#b=&FdRu4>iA#q8-#Vz74~)fQtmZjXF#+)3U?+3Vz0?I za-E>#jQu1l?3d5o=~K^-wzskaGmerAy>&92RLj4j|NG(r<3g4fDnnqR>gMj$qWMcm zI(=1qLV8k~>xH}=n-yqQHY?L<5HaRHAX&jToUA_aPp6EXrFeRuwU10MOt#6KftS;7f0GpS54ILd6u<^pvj}Q}EcVsW3sR z(@#!0?{Pn+Omj%{t>ma1gxFJbz-I-o8ZvbdwpHd+lWH~0M#dzAgB>!Wm#1>Rr?(f; zzbn5_PQ5f6-cAGSa+zW-Qsg`pu_k2c?7MUU{fCQw%}fsWkNCQbdiXC%+>8b3qFk>d zf<{#}Tu=4cuS^%E1l~~X`ew@yAH2QituNnb`s&s1B!8nrl|D0@Gko=bbeC7WvFEK6 z>iAdFf6)_O^Ua5Elzsh9`2TdMioKBlZ}>esRGWSJ-mD3~Do*1qjC%Jf92ADP$_v8p z%aC19_Z!vmlV&!mbXQyW=Q(C6BC22Knkdlz&P-VHOUvk_N0{V@deiFkV z?`fe7(npQT7XJR=9YGlKtAhSijk1K33E{k4DErnpcLbzyt_2M^T@PM_cdc-VVPc_i z_3WEv$^zjdmrSVLOq6%J>=DB+m;(=9;Zg}|S@EGAblLnX^9qEG%0?yrJxR_@4E{w= zqkjK?aAX>X!w6$%)m%S+r{YCtBrI~f?1*J{9kMuwNWo>t-5#K}8`37z4MAim1 z3>yim0krcwWRte=v4zL*^jZ5SD|o!J>+L@~zUXvNtLBxc4$Ra2+o(d#@DgFQR~-a$ zZ;}D|84*@wc7?7YD>>C97N}rW%6sG`bQ+`KMhYIywl-kc;^c_|urg3s;e=wY{G@EU ze)0Nt-UlC&&7;-g*>K2qpMeXMPchI&l0`*SI4)sw<_y{`VLJU%6<6Fc1*vpLa5AT4 zPMa*(>oSevhjHLn*8x?fj%X;2S;K2or7zpxtiad7lF7ZOFzAAw8C? zg(TAudTa9shXVC*k^LX_7E;L0m$u=S_)`Wzs-l>3ij+|i$DqhX$IN|6u0HIGhlLL7 z=C_c8-ba<{4soFa=+Bc_mt64yfAwb1!@l|8!bHtF3Bq~kejE$l;&xB?B{(;cbL0Fj z!QEfHs*Vcjgfg29?g`N)d5%(D&99uXRsUj%V3%hcRCi9;dQB8tPt29k!9+3l=sTAa zjR`GncAsv7qVkBx4wqFVO4!Ti-0WcM+VIuRfEoZn7L9jIcygaV z3FDvo=RbF%S52FhQNn!G!apBcYTWC}Ce3KW_2VWwW1A(p0(CK{&0v@~&8dW2B`gre ziq4bhK#g(-w<}0X#Q{3Aa<%S?u4=DMuDzD$lP^)%^S3J0=tN^_Rh3nU!N!j+14}Yd zFgHC5wwgx)Rw9D2J_MQMQ`URX6x`>E=#H7c$A9X0s%S$(7i zcUt!at1qnWe&5vjSO3j8i^Rr~*s%9&BImT}>sNuWG~W3^Xfo%fxWa!iFP(c+Tw*;> z)-(>y!?x-Bt###>9j}K<^q#EnuJ9ugXHNkPc{M#GEuW5=fqd?h1h4AKl%^>QdpjOLoY;6q*ELn6$uHQ6aZP&+vXlux6r>JhOphWGKgeIWbtd3^er_ z*Ff3qPS`}dJcc+r=yu523^;Ys*z|+?Wg6uYK>_(3Shf3|a^SenJ7!jINZEYk^DYtY zaZe{`S*&hQ#yjy7o7S=KX0<#{N%`-CyT+FDwc)^;2{~V^n8esv3`n1riP~qh$**{& z0B@=q%?&O%sHoZw_0`aqhBjoe>uxV>QqvKtCd!XFK-A>qT99qXb*<1i z{lmt^wPCYkA~94PKH${LUj?$O?LeasHUlYbTj(VMq_xEs*T}iLC9Iy)&L6VFsM>PX za%@q=(0)fiEC+9B2C3*GHwzjU!|5LfxQwif;qxr@s@{F>QscHnn>8pV`XzDQSuD`2 z)*+o-NeHqf)cLEI2)4LQD!CI(3yYby)Fhhq99Cvz%C3&A6~<=7W^rM{FeGv-l)WH7 z)A{Ohr))ZVc7h_G(*0L21<^Tq=Z`htxfTu?W7#f`;VbzR?o)cC$&wX)UY zfQykq(7iNk?#D5R9b6uV?p-L=Z=sL>>$k3ubuY~pdZ)n_dMm{MVasM9J^jXYD1>ZM z-IJujPAS^AjM@}jM%B<2PMb)H&mNyc+-`ci!+APCq)CN6WZS(L^LEM35UT;{=^T#_ zQcv%<*3oaj_SbpFQG1)UDJJ6Z+KgVRC{ujc_p&fSv8agC%(Ot|Vy<8kCf0Cb%$;V# zyR83;mF0M4+R^!+zNWVvWwZafg5lDErHkQ{QA^ctoO3YnTzJwSx4JB#BVLn%`Qh=#8tOZ`Tk9e&C-?u~h$ zew_b77PmLVQqVIOQKK$Iqn)%6*i0kpYqzHV++B~JpPZR;mL$J4>khzZM4wSE#Q?); z1{HDgOUwFjZc^tB-iZ@;-N33G9;zij3gr&$hW! zGn$E?;JN3{s%g7FO61FS16R!h$-T~cg#F?&SxDm9EmF4omytq)FG@PaXeg3GMIbj% z0x*ARg8QB7{jnGJn$LOnTY>cu^~Wgx$)7KOGsClmu3)n1INrUGdrl?7qm$AQGA^h; z7o?WCXrJCwJ$}#sEB-h)FnLx7Pi4T;R1A zJr9WGcOa;FUsflJ72TULWVa(IiWkQNjUYcQQ)!8?SpuZjsUk=hx5*#UD+Hausr`IU z#u5|71|cK7A4|OTbDzNuA-_6!E&CgKr^7dD&JW2E8}4ClGjJo$QOsu)sneAyJs>5* z%cMcL$@c`oF#a0%SdRL#a2c!}NT05K*&7lKl4*Dy2^O%Jv(mqScY*+UCY|kcf@I9r zD2ME>h(Y-yq=ml7^vcmWiG${2yhr`$4a^jz(j~wp2$=yz@$BA^T+vBI6O_H2AidlH zc`|>sbESW(NBInm@^hyfj_Svx%B^0C$wd};94D|wV2sV-bFpM_ksYwU{?3oRjXf2c zU6f4t9y9ud_`pZLFc%af@0pBaR-kz0R{pQQ^P&s3 z*FzORyCgGsCujZC%Z?er*GbO*4{1~LGRTtw0ps^z2Z?((R`QH13)t*vYfrrGf+*|^Qm^SDben_V~?eoO;lH1bk*~kM`Dpj}8d)yoSlftTH z`@PiYq-4{`Ce|+Nk^+pf=`@EF#Z8X-7%2kR136ldTne~YSa={$=7bG6*f<>qW1qa6 z`3I~_%#@*@|5RyAiSd${m=XivWm3#m3WTI1_PED!>S%DzNp(<|+BI8pZL zqP|chzK*g_6+Jz^DaNz4^ig`JPm`+Ht=|cSuC_TB1f5c%hHI34hwQrla{E;8CRHxC zk0dj7Qha;eduZPnjom8ymG-*d3d*@X9(%E}TH^+mfju(s1sjCTjNG=4hcqBNm^VnCdCI~CC_=~7yMk(}k@bS39ah1Fae4x)f);TxO@u91o zPX|BbK5}7YBFvv&c2FA4Xqf+AY(~B!+@|r>1E#pE2zbYB*f9n_cmxDWDF%pDil~Sz zZnv9;tKojJpr7dy#xwhZ&vQBga)X*wMY0q{CS4(lcYQ$O2qvq~LpHnHZHwQ4e6RR) zV4bWbP^x9^Mf&B6M~+=#pXSm#*V63!6UV|6e6X47F~^!VE2Ct-sj7v=k^Xw* ztW50LNp`$6rsjkJ&MGJdYQ0OTh~AJmL8_ukIQ$c6zNlSqZF%ov_?hqT2}feNM0ul2 zDjnsr*{9~K4@jinJxLw^vHK8>3_eY&YvLu&Yo#?nqOi&L5XhOODz-VNiLkP&kXub3 z@re}VhYSE!?>Ubpf>?`1fh>qN9qh)$k$D4VKWO>*Ej@ms)_?Cuq{@aZ&t(HdoS_)d zh&hJ*+dB95vap4}M34i8IqEcC*?jbPQ|ar$*TGZ8IsF|()JU;3}u!`M-=2(1g1Dyt2 zu~o7*&LUJ3zbq^W+6KH4n>>dCN$Wls(nOxzxdfj#a{Dxx*$rpsxe?%zgL`S-7i|8n4st77~PA0iLP#`!DfBTeL5 z_p`h~JB{-Sd6TNfeH~dpHOUDnB2($Kz%!x}8jq{#RixXkj=w+j5O=-*W<>#agDBR* zwEPyFpW%c>ENvAVN&ecofcDuu03Y?;r+j`Q=d44?s5SNbPH1O?-2_<&4^=T{0Lj50@!Pd~3{ zan7DyBrBQPEYB6-L}D+yw$!B~017nar>CWJt4J4p*?%{w=Bq0l>l|_gYr=MhvfDVb znkZu%&-OKoIW;{|qqwNH`-MFi{=k#bqsr-F-$k;)hJByi25`@&m<$Steoc&ds!d)oBT@kT=}ziMK_d4YA7z}eQUYb1wK;}1ImVFFXJm*PXsrxAv_4N?CbgpR znHYpwV3Orfb`bud-;!S&=eS>}U?0@xj@YxuQA`X)R-t-lADBOFivTvlce+>?sWyi>{viS?KGWY;FDqD!%+<505VP8m| zIM%;NRwpgeK5@uS_a7@OW**?!uA!rkXM@m?tLH+we3glA$yKo9tEDxHM*nlJC>Al>wVW&y1;Bf#efIN57ER3G#z89N|2U7N? z8P^80L67bFYQzMi5bmXrw`@@+NX3BY3|--mEDk6}vZ&kP?3=jfYNy-1Ve{22!3v#4 zC`fPpC2qV3VKdv~8S7p4YsXjg%hs|Leuv4Bo#{1h6mQ;k!v_TTT}SvhYbj<8MOIP~ z{laQlHOZnE2Y<3l+&C{k{GO-f3;~NNFb5GV9%r$k+y86TX=CjnHa5eC9hqDML~N#* zO$gtJ)y{{52b@s0YSla~5flcHk!QVG0##kj5)kzad*GcDj7iM)ph*IYIq)!aFD!AeUqp|9I{% zHugn!AL-c7v(yMy2G@3%K;swv!t7C9#})^iuf0&a{B46clKddv!~&W#F`U2x1^z!W;9{S%KMfR$xg; zuk5CDA5`02C%B8bFI}Q#W>z0}ZI?jOQ+dpxCv^1~1P^1)tXWtwbUe&`HP%}{_qu6o z^Bd_$4aUV%Y#de_-oltDtvVyfcC6tXRi^QDm9=?vO4#*)Hu()zhS$~5LHVFvFSnR; zD)gE#+fk#u$Xy2M@-<;;yhfKE&q2FLUN?Ql{fcD2*Bx;+An#$oV%H;NabRD7w$4FW zD66orlWRdtJr^TvFK&M4w;G<_{>bc}Pv|4-_Q!)B`4PS-@nn>HV#1U05C^6vP)NDqv3e$sUk z%g$M`VUvq)=od6>)*|J2j~WWyKtbs>+Ur-k4Vrx;k(fuUb+aj1`*2oZ*U^ zLf&@gqdt=aDmHM^EkhFliTU@^Tbz~{d$T6oz-r$dPfTw>#1XS;Lv}@OE&R!qAk+U{ z=6#sve+ZTu$?CglyZeCJ_h(hP412RRTh5d9qj|G7Y#NFTOv5&cftL9#RK#}YviZBb za^JWj?vNqHWRq&Ub1df^>HTFar%`oTU!aF)`h<_f1>v!rzYY(zx|J{w($6}mmV&OS z=a`0c*_oBe#%$UzROx^ThV3rtfINzatS!)^il%Of0Ciw?axX#Lq{X?OtJV3~5me)G zSG)_dzF3@chF0$n-Vvlxo(_H_Zh@k!iW$kEu7>|v2RY+~%Lo9On(+ypHw~9B7FGQD zpR2F6C2$o1qIMoI0_hI5+NUM5NZ@tBy>!|9okH~$uiNguKy|7s^d7R)$u?xt)e>t# zM?b$4LC}OCGRDnU-fVn_(l0)bogbejgLb1iwvg^1j(>{y)o&v{`t#2|T=*yPa*A0> zk;qAH`ycI30Uq^{+nZR5Sxu2>*xvXI&>0JAg0X*jxpUg&Zf2hTqGM0{)3K9zWZG~X z`A}JGyoIsZ$v3enS?+A6OOzc1V6d~foSItTzZII;?0Og!`-PXE#&*RLfT;(afW zBpVKkL9^+Iuviww03%P@Fh8@7Hl>{ZnFy!@fnhmU(Cv^YM=F(Od87clv@}XoTUTRy z)^1rhL_@oRb~w1O^!l6P*$c~`xS|?VW2=``Us$G zp_pWfY@j00(QFVNm+9&~asd?WP?yjJqL(+gS_!dUIa0*6$x*o_EAW4p7{P~s`q>{N z-v7=eY?sfd@^1GZ& zL4mMv{#9{$@F8BdWAdc(9-D)tCu`0mLaF#`zic;dzj{fqVWL?pMNuSAW7=f#ybHoM z&ROa8fW^Ef)n0XdxO@>G#VS<4xs>!j7Z24S7_(wzOW&tHSqbmduTGBD>Ca}MyV zyBx7X$`i;~f5+qO*d}BG;j$g{scs=wA`z(j@xfxCk4of5{K9@K*r(N?Y_8YW&Kx*MhNnh2x-T6keJ#@>=IZS8F z-ou|TW+ng&Ly#D~ny9T|wG>gxzis;uWAfsc#N0sMW(4HcQOsJ3tkLO3<%hRI|4cDw zqx&8AUE)H}9lJ4a!hm2uappmY{c|UJ$1j$@PaB8ZZPsHoK@3?mJ<7dgrtaTHRi*F& zxuH6%ss%-$b!4qP*DIRbcdCW8W$UXo9uPsJLG_f69Hwn@JY#Jzm^brAmH}T*bZW=A zpPlizwSRVuv5t?;;>3h9&r0bb$kydkSglya*#xrdb+Qs+K9wpebls|Gp_?Vqfy<=% z6oyNZrO8qx`m;W9C{EM3FfzvcTH-P*jJ#s^-T(UFcX}`7_w~P+O}?;UFJ;v$BbSUv z6!RrT`j9;XyL=OYr#RDXn~!GZ&6&ke)vB(R49ZqcErMp=$T!rOnprFFk^>?Y}-ctW$?c$9O1{Ocf-GwM1|o->f!>dFTg=oUfjkH97|Xy)6Xys zod5JvB^wi2*z7uHg0up4P)s~UVzH?>SD-!YbLy+2Yg7(O^GWgqJ=m-e|QB!d{5cUthR-nBp$I^pKR9=%=qPu-_Ot+qTra7Rb-1j*>7Oj zcTfy4$z&tZc%B%)W@`Liq?2_Jed-IT=bQ!8u9JBKP22cT?B&K9_DpaMON3Y4w9JaH zM$f{p47iR=S|IFNZ@rG8Gphi~47`l_RkH?+SmDKXSN#K-ZITx~gtA1C3aVgg_+O*4 z9-0%g>0WLO1YNCy$LC?mT&zCt8P>SQ3aeA%xsU#=)SD31iJz5|bauXs?HWe{#S0@` z+#-r8phzB++j1|F79~q~Yawc+ChnZu zxsf)ei2&Vh)IX#x&<*EB6RWmDv z*-+kuKh^z?+86MzdNyd0Ya2%c{sEHwGU zk72j|{LsCn%DC>DjV-ZVXUi0+=Q^?%xVC_7uveBaf57Rw|Dr_~7g&1f7@myh{4@_g z&v}5AE}fOWlK9WHx-3kiv$$~%hr|WkZNZhG^R<+>E~J`Y>{dSe6#uYq6`5RgmtIRHJZ(zJf|70w6mX+zY-Es_WL)R+fl#b0-3Yf?> zWkMTymmD*C&64_QH)(CD5Hf}zrF$TTiCS-a#gn@0WnmTovX+A6*w0}F+9}7sJ2=Z- z@1p+X%#^bv*@j)zy#_97F2w*?GN_1C%6oHrK*2AS-sJ&(e<+9z9E-9Ip$Hd^vKfxm zIpSDeHjp=}Z$gU0e_H_I;z6BHmPmBXp_?RYKOyASTSQt=5puo#%Q;TL1i{5VFi z0?R8)GXmF`sRy{{+t;++Q!R9^C{doq!4}MF$rTCudhsEtG>UYlabo?egVVVg3wlFx zxRspFfTN*Jsuq5HNN-3QC&vXUJ?Lys-|L%5x@f!Ph(|S$i(K+u7oy=>e1mC$`O(kQ zI40pax3HQY+g*Wf?eTK^gWk%dJ*c%Ow{6(UtTD*>4N}ZQiu4b+rQ@a+rT?OMi0A~x zs&}@D+khr1mV-obkZo&+qI|SFXoYSEUSr|RHn#L0DR@eMlal9eB!@jYiRqwP* zP?@e&w8af+L@_y7;0h(Yq+*6PvxslG0fxo%+BkL6&NnqmO#5MaQC-Mg>Cz`3zEcD7 z;?vV~J9Nk;_CQu-zYqRZZBRFSK?qe=miuN58#2rMWcAIM@f{id6 z5Y<6L1Z6RApEvx-72F}}(+;tU$=wctwK5kI&oOLPxx@|=No&hqcnXC5Am7#Lvf5dV z=}TQ*iW;p?rwd3nv<5UOQ4te89JEgDlcm{UESR=%K9e@yhStE&Y};%Z|<3 z3=_OLD58`rK&HN4KGF^2N^uzK9+5mco8CHOpKNlC2Mc5v&7E~$vO@;7a^dY)_3n(< zZz~p(jW+Dg>@nChZKoK3%2p~u$IYfu-k0^tvx3i&b{Z9O!D?g>1W790D?cr6@W;A` zBCfg#(6Y(1BUpD=3;($5LmDe)KeirxSU_c@9T`5h8eUjIWy;rAF8Y}>wv!cr6E?@rNMmf&rre|15F$DXON@i6u?tTEzmfBd}E*avw@%!dh&MO$$lId@Ii094pv z|ESir)e5rzukzRki2dr~`qjqSADi_~O{Arcl4ys1(l&QxXq{6H?>dPG8kCjuquiJJ zwax8!(kNHXKM1wZ**;gi;vq12O~QlNAnG zc_7qpD&N0joc%G8WN8QD0H~8jCFdxfwmKGrx2Un0#k~c=S0uB^5m(Hx^5;I{BV#!Q zBkr{BA*(5{-I@%$6@OB=>D`US_xCI!n{0R^Q)=Kq=|#CtOhF3Moy13S)%Kuz{>7z+;-W3T)k4BB9HaYb8s5LslT3QvdIZu1(WprF}ubRgz-ESSd5WN*={Pw{a#Fag8ntO;N5;XcWaW z_RPbWSvH;Iyi^5VMij3D_9vCF18$Y)in{qwQ+?hi%b{XMTHqe{!y!949Rb?-`6<6u z^E&9n(0+-Ptl&h^Vy#>=JGA`yCrAF%c-3L!xZ1AGgzk}%8<|ZM18&!PDq=^F)=5Rh zr7dnqL#x#zUJv27E*b?g>ZG}%l^~H}CD@Fz-myf|C{M7?spc>C$4+EZW4n1yurVXC zv4LU|C=yRaV9@~xb2;^qK^f+?F=>bTeywf^Gfye2Vbz*ERN$PlK!$Z7VB)Ds>%Z}> z_w;Lwe6{yC$(fg?`=ZML8CNOh5=Aah5x1rB+%hm28bz)Mnk=9SRvyFTaBAsQDx_s`{vGvFlEs+frP!;#~_}p^q~t)l0qEQ=_c(-vgZ( z+CW>b06ILKqp91{1Vt*mO?*f^SEJYo?Hkd8T+urJWFWR3vP+-Y!pG*wL$U-#e8@#6 zn~vixA;qN28;UiwS@J47mnHSR-)!URp<(B3w}eF5Ff>3aVFWZ1DP|o-)*@fQfKxY6LaLE} z4fDabURR?p(hPf`Q&yKoFrGUWxG=ues;}AMB70rs?_P9VJRn^{jOBC$+~i|`21E+Fpe+o` zl@?hCSga6a!k-uw0uwxG8bJfiAB7n=cA3zvh@87Uuuj%TO5D_WJ_SJ-leNyLIL04q z)`vXCGpuph^oKkeDW1);DP8t1$>8Yyjo%&XJ4%+?u)hI}Y9suO^%N6Fkr+Kc(s8Kt z)+m#`v1SDN_>f%&OU@>DOUq(YYYj|3=b7x_y4U&DMn63&Zhoh;nB=hA0^4xltkwWS zWfTLYNhL_dk`at}(W=Zw%{u7*fK)iv@k^nb{H8RU?ju0gng8 zhs~w0_x=4RFS>AK3xDA{zjn#OY9NuQ1F`D03iWMi4X0fa$GhxbK3lzlbDDI|-N3AN zJ`cXhTE&;{W5y!be1xXPj&I6MJf$2a0SG{rhG-ux}a@2;GoErw_qKRV8 zQlx>3SlB+}uy2Z@ht8yvrtSBMn%2VK#;xOTga+1T$-R(DC|LyRGmtvRF0U3ojygdO zi;j3a<~2&&Wsi9)1bIO~Rw1h4#ENtR3iy;W!FAGmA(4VQ$qmOQ6?Pl12}u=Q5ua4X z@lH>>p_N&9%-bc~?tO}&98`+@)|@)&R;HgcscuV)+_p+~x@=^UB`3(*X(copOEW%b zjcP~6KFI=kte`~&M-Apq|*xZsnvth;+ zanI|6vTI^wi@zYup8d#a6J&E6W>oMhg?J4oqn5@yBK^we!b0Ga%xr_&p!U#Xp2(nA z!>JG)^DJf_(l}b9#HTi?9zv5}*37P%Y$oElpN*ww>A7#l;v&0QvRP+1_M3-4e$j(0 zt!@L(*I-**>A%|t192d)Fu+|k5C7c}6g}Dc*X9G31==hyloep7q+PfrHs<59*)lg1 z3|p!6No5N^L4iGg%Q%-EfgPuXzkX^5AG*2je|0Tajro^YaudW$)o~8Z(BTf)K?*G_ zP}okdpL*GGSXM(@5UJ}yEONVSaoR!iF*D{$V*#W*eFQ7ayb?Wgj;paPc}Z#yOosiD z3PD}tvB|fHvr4k4Bj^aI3@%ECLVB!)TJ1ck@gr6!dOFEgKl;3F>2>2?3=>TZ30#bV zBKMU>iAT8tU?zg{GW?izU*{Dr$Z-YTypSu!{0bJHWz(AAjNm5azM#aZxbN(jv?yAv zXZKls&ZB%`t@VMk`?lJwOR_s9rgtJ{#D>dMOqh}FZ(J2ekz!7=_a}$Ag}g)D8}ph} ziEy%*gU1&0rOgAL`S3RHd6pXWQ`fwVe=c0?^k|))!#(VN);-B_w@;I*Fzk_d1Cz$I zc-Omk&z)pgm|whPsR=jpIaiZ2DC2nkv-v<9j?bEi@uMyilrIX``)eK5M&P|hMK&aM zob0JLd+yDgGK&o}z2w<{)Nybk#f~V zC0prOsJx#MqsE77p4WvA~FPMO-+Cuh+WY+e|cNKGVSo>+$GiQDpPDY9&*KC9jl|5QxqgGBQS~Bq`DcD3uX0bU_R<( zJ2{Omy<83Co1q(eUY8v6!HMz}_UByg%}7zy@wktrs__7EApBQf7z+y8Y<9+%7oMirh}LQ?4p(ODG` zt@zYTte@Fz;EltyOaOON68(SsV38h5B_5|`q=OxlYVoc{dg*Z9IhDRd`bZN4*}{2PFLhFw5gaMN#Pgzh(15x(qr{_) z^Dtns>!M50a;Uqj5h|gE@6pz3Vpy{E*>=(3;2_2Uwp(kfvw+`lGMtk=Ok7TUlh+OM|=y10!~bVrnQ-iO9p1@PJPl zogB7So+!t~N><$)+fBqe*7{wOj`eGVGa%6;)-vUk^a;Y2@)Vak2G z=;H7Jr?UBn;DUBsCAZu+SA?n7_%{ZeU>wRbV$bR(&)WgWM7RihR2GfD-NfoJ8mASz z6S@9-#+sHk>mr#D+w2V~o8KYZ`Re`9gItXg$pNuP09_^cmh1T8p=wIFR)GT#axtAg z;Iz{vW5FI!YORwk^=?$QDRJ}GENK^4TZ{H-;s_Y$6pwKmJKxFmThzxdI@5JKSpVS= zjT)WlW(n3s4O{#%J}*mNHV+e)d`I)1#SRm{o6*CQ>6fC^>A(3MNo2Pa+3?Q2#9%4P zq!`$xrBV@fvcj;&z*yInp=(0(xdTpvAZfH)mZB(`)5)#$Mp21gZmqaO)+y~S^5EB7o%QD97njLG63@07}U0Vxhq45)o<<97OCA)xx2IA6;l7b!qL23*iOSqnc4@=uWHhM2Nw5(*;B zMU=7euq=L_ecBuKK0gf7V?^i>zK!gBX)toy03+2DQ$dk^R75B5h)aR%VyLs#z|v7C z>xQbBM0uW1uN>v%)m8I02R2KPXi(k8IpEtuFA-$ZSwT5LsA{B+^ve~u&E-SgVZT_#%)HRg zJk~kib?TvW#0PmOdUeUSF(?y=;_gH9Q%*2;sG(^o7vkLqNMq2Epk`)Jp6=8u&jgX4 z1D+Rzy|ON!34_6SiYx^V z5sC76a=>Q|7wm{biKNm0^MGPdVS*(F3FELg8c$RS3qZ}T))y4lai82rYr?X*ND>AV z9NHiY?xfM>!P0Xmwn}!652>W_1m2kllwh%<^0}8ds2rre=C;Lc%ghEL_S{wace-~; zpXS=3nvjMJRDP}cAqN#{3`m+BH(ndfEXec1kuU(D7^kXmWYm^dBrE{zmgfl;bY04T{2|TQ@LiI9|hpu|#(fr!^Kad0) zUX+Rrj7J8=K>4Nyl{spFb$`GqmQzdLNO>;l}RJKvUqIycAg>~h|+ zi94xd&7wIbWi*70_1uYGX9trNYyR(#jDxi=RLg(~slr70QAkg#OGpW1m_d@a(YZ9cT^|v^rmC>rtcpm%h~OuZMZfw!@$raQOtUZ#8DAx zft3)VZK1I=T)Q*IE=!FvCNNjjL9=l@Ei*}GLI$miW!})tr>s?oV_$Ts?poj?y&}FK zT<6gm)IhiMK?Ist`R%LXO70`~q?vnzie;%YO6VS_wc8k6La*~knprh-6HEbaVk;aA zg4+4&&!^SM)q9xE$!vl3Kp;YW7oQA|6G~CYRRCiaDJLNk=I7natr9Y zp_z2w>)Qg?xc3Ld`nJpJgxSCqmBWb(Sy&4(+B3o?6_z9+-&iI8o|Bb57V}YMY-=U} z88dF((Y1QIvBD{v-ZUG|81FG~LZ0ZM)(h+Sw@8xHdAe%q8kZ8MOhK;4pk1=4)OXPC z7TKy=%sJrGMQ4Q6!J4(kr8S@yBt33RvuA6Rx5%T}C9-QlUDjNcSd{h}BgM*oO7fC# zePY}*0%|)W*g}&jW&=eMs0i%6*famM82D@#Ey@Bpf=X^8H#!V8KZpP85)O&$Bu7Ih zD^SJ$3uwl-#{xsxnRS~DOm6(gMt|eYjLlZBnb@8zp$h0!U_D4v9uTAYMY<@JE}+YO zvmAgSMwVgiCBhg?<|EelPV2sAg;?qXudjUYfqs{}!*T8`a&5F^pAAuiYZi)ec zm(QsP%-ODSxxuZZPdlzdigZ~w(7#2w^u3-vUt6=<;=I>y#oYTrz)?9Dk~FoV9y-P~ zmJ>7k92EP=?YS^m@$Y@3E* zX6M$`Z|NZywf=iQB2_jVlf7&JuQL>Lk|M{bi0h=mrNJkA_KNA)6LLXV5Bup<`XQ8{ zoN-T--+->MO8V&-jS@#Z<_&O@C57BHh|^Yb*DCOCjj~Uk5~!oeye7});JFINYd*z( zx~tH0LN~ehz_X!=60XHCE2bkg=V_k?A1r#%jmn-~!5g20JPGJ|f(SMTrqMInCcERl zILA|usJ@>_my@mRLUlImC085Zub5(>QYs%pb-bm5L}0$f&ZLXJMRc=kvwwAPm$Xh; zLKlhB1ZC6%k8U}3HTKXC-?}PJ^%@X#s0z791&yk!;v4h&=+B_vsL0_|&{ERnTp(Nu zxrOrS+1B7lvkb)8fXBSC%U&byv-P0)G5ECiQp|3O6jKqLg9SKG587P{Nf))r zPILE9E1#CbIRXkW=jq#SsouLeMRZp1t=B=8Pm)01oAV+JGABFZO_U2`PfCj2>qaD z%XzZih9iN{L^2|mwvA%aD6)l$h;qLxE^<39zAL^Xtfkle`&KB>?T~eOH>nn~%|=d!>~>J~~H_k~gmB ztRcBCjjM9h08{%Yrj#PPP?+gKKq9!zi4eg{;6SZ7bJ{O0B#Uk#{ZPlb!?6@9$J01B zr3J3oW1B%=^Rrd*7up(d9ekZSX)Pqo zvJ_d|)&PxCH{4P?uyI1|sc|trGRDj}U%z_nj{QtMc$WQP+8UB-!zKnA8b_FzLW;?! zNDdW|!%5?z*hQ>pfX-M@6Rc4puTZWC>to`bD;1hxHRcVdV&RG?9#{G(< z!Vzf?u?;1G+;FNCo(B@NG!8btwa}HqR*2>GJ5~w@?J%Q>KP$AuY(3jbV8a>Mm zhkXaalEa#mMQ+)%`$CpM@C2{b!L=T6hOdg_T$Wj6Gi-8JCjuXn54FGr)9|Ujd+~ss zer?MC^21L_Bs<5(hSLZ@-!Nh&i=&trimXD3VeH9AN}(LFI>xn~J~SCz4Wmq=SzsCE zK5L))LC!l}6M=*cZ=X#-0s*1j9_RVwOIkpJI#)1>d4Uo0X{(S}-0`T?IHF>+{)mYf zNe3hYni$ZEh55~<^8_61d{ z+vymlW%8m29=g(JD4?3|CXs$A^7NUQt=nIj&>en-f~ z1*MWERW0;EmUvtzjmnA{4G^ER^d`t??#B-+qX${|<&8hj`kis?-b54W0iP=0oq_~K zr%Mkhhs|fF3r54O{JD+#=bH6lj(VqcPqO-s+U^p~_5bMC#u8aJyV95ty*dCYiEZ*~ zSuNQYu0H z;reCe+lT*sCn>k#DB4AXh2kW|9HXGHFCxdY*(*if1G(BZIr7?`SHyBq?S3PcVXGRY zXw(LIYZWJz$*!eS(>VFCg?&Iupbzb^FY1>JpF0gi@DP#}9hHIG3irhN>+ag^Sjfe4 z%m(KJz9<|~$h|7gQI!IT_N6z5skEznV}TxiVEnBLEaCJNSY{rLKnSZtc5vF2c~Sbx za8P!I-(g@BXY~NvHKC9Z8oBXZOEGIGvXY8e!&J>YH*M#u>SFJe^5rg-K(ju1HfY35 znuDJa!zb;qtT5p#-_#?b#P!k7NbP8KP&VwHwj0d*d5USINIhf+r|p`9|4)=(<)85_ zgltWcQzMA^rYP=9H<2SA^-827EcadSbBmw` zxjbmKyoj!F%u`f*6oGMC3Ckp|hTGp%-+uFYxY<}V2R~+-8ncF(3qaH6<5x?U&Y!18 zR;v7`xn%cf$g*LJ@|gj$4pR)!8&+W`O&c}A@X5WoQL{Gr0{7`9PP63Pv=X_x(6yeA z)dUzz*$!oJ?R39Wy+70;OkXBdqZ|OHj-aOqY5S3tMPd3o-D1HF##K9Z5)}48{=`Nag=1QT0TvWqxJ;> zS4oOJDKnUwJc@yW*i0&7Iq;A|W)7OXDrR)fRpZM>5X@-de(A2pnxbgOG)T&wo2K65 zj-^J(z@I2T1|g3jJERV5)K*6#c_8uzs?`?H)8QX&kqI#V^eph&^aP9+?y%Vofo@(Z z=N&y-9=!Xt_sM~mCJuAW0Q3zM1ND|CsffGc-hgZ$v@K~+v9N<%sXP+Y3;i*5(p*uK zDxF&$9A)1xY<0WgwA{W}t1Xj7UzhLX)=6$SR`6muHPGcBWxw2hxz<@8{aV|o_aSS< z5fWv;e%dO6CrofjvtBxJq5v>=LfMXEWgMyB z{7YbmFu)l#FMn3%R;$I2%CS_&HwO^4Kdy;4H&D}yV z$rRZ@MV$K2j_^uhpL~}`+5CqD>q#mcFAFOcv_ga=LD56!dQ}QB#!}%}Ewf-%$QYjh z>#yn<&$HBJcGF<9e$0P-z0F;$$A|16Z#R;)HoWBQFaT~U#Q+=gCMu$b-VVFnzqiT@ z!|#K7#0vWzj*x5oezZNbLc8H{b@)>1X!u_az>gDeoK^KhRptL$ke~fA;{D(JEaI>4 z|MvI4i4fC_9igaETyW~seTzE(OOFtNVEy;Vc^gK^eFKDaP)r*|u2T`~1eLrR&kAJ= zUsqs>nu&d+QrH%>+PTN6gp)W`9n0%>sF%V)6t4qH;XbXTTVdOVeQk$7s8HsLu6ec3 z$fmJt3<~*ULaz&N$a6(CoV38=@Wp~0PQO#+Tns7iamOw)z3_yt3^Uqs(K60uN8AGz zxm|WtS2*TEi}-!|lyb9cPN4dR`~YvA^CK~2$xo1#Ui+peOJLvF24friS5AdCwiSZB zAZQ-GWy?r#}}|YZ4{(U`R`4?$^7T4jTt0Wz8Jhg!~qL5G%cQ*6yN5V zMbMm8_L-X|;Y8EB$t$^f<74M-w*+>9qanv;msbKc3?u5~5-DaKMb=UgwKOWt6!7Bx zGlT0ungqVm6^>a{{8SC(-$8wxnT+-6$^4l!`@`GaCpS&PfoUJ>N8b-lGEOkr?AFIb z$_T&o$(*|o4%7q#E1yrrj7gpJC-Y~<>^~VhnFpq4e>4-|_}z{Fm^ExEM8Tyb^Q8&M z9W<~ZyD0{A1PZB$i_Buz9pY%!s#k}E=Y`uPpUd-SEm4$FC3Di4Hd&)8f$IAHnzvWK znf)J!-`x{_R&`Ta0!gUM;4(UvbD6%&uXkwSM+z!lO#>fim8#fp)%=8cC92&&$$P6J z_*#&qv1kU$hE2@#e_9xVTg%%uKhfjn?3{~J$OU$|vEdr+2L`ygMKL!h(h3ne$$)%N zrq(Dnxq~YG42?pk4X%#oY83qhud46L3tfvjS2#m&ZzAf=o@ox%G*)i5$usCR3!nwMGr{Po zhx?tHC20=D9Gy$9?v)QXXV974y?$7a)+azlw037Gz)<#j)_ACqicdH4fKR;?8*g=& zN6S%J3~SFv;#gOoL2BdlzZSf!$J$3{liwo;*d4PF>!3CL<_`n@OK#>irwFz1WQ^+$Cw0XH_Ri5~L%B z-|0%|HqiU(<)+dl9K3Q^)Z?j9_K?lv5OkX)R6T>5F;BeqOBY`~s*1acKmm^(s%+ST z9X3E!DaAmIa1oRQgE-$Fcl4m4{}GvHu(cpjp7=VZ$&04vQ}`dI$h8^tYDqnRtD>E5 zR5mK39o41tv6nJ#b{tt2*cyQUVt%&LztN>rj>+2rr(8i7oo`7A70V)LOgxx9ply(7#3N=G$byev#m zfI%q_?jb99mBItQmiC#AzsTdZBgPN1@aso;r}AEOnvX&*mLx<9wu$!xp~pRXao8jA zF^4FoNu~Q)%Ehbw0k>qe^ogJj{t2MW%?gAidm+@uj8j>j%o*aVFF zI9B+0r8?`aUl^w%O=SEM6kYTUC?|ovB-TKJ7l`y3E&K$t4!uSY@A`mLhA;K22*-+n2B&mU zuRO_VmpG2Wx`Er_neSA7SH0w|p748LZTjFMgio7P1>vpof83WvlZ)IsZevia=&@Hr z=tW^B9Zfp;dEy&$`pGAM`A3rqo1H)XOJR6k(6yir)fTr7eyr#yUB#>8-=2QY9qae= zLQChgOSW-4`1SskoB~;c3o^N%m9{wVnAHN@Pxy-_aMQVEbj8dj72b7>Byj)lUp{RR z;(Ilzvb67s7gAi~d97}ZAYgoo<6e|6w8}bFGptBcER`I_Tb~jTtaMi zN4EQy5fjWqeXq68z{p)@@I$&tRz6?jdP~;AzvSCP>i7jdo59BBhhW#;X`*pUgM33H zCxZ-MZ{cJ4ZXx%iu$}*ycbb4Nr5$rkj4E)tXVm)-3A2H8_gv6p%X!vjn&7blx=*fK z$~OOfY3;24(AzFer{FO;Yfpe*c|^vzonl%ja+QiG3cWiOh~S`tcHh^qel{KIQ($lM zm5shRUISsVVTFDPqGM#8%Ubs@;CwzEEpGwgv;$;2z1DqM*a6_WEeky0o3S7_=pOf) z3`aMq@Kq)EGMzL#EBGE4p9LhyaHsD6RPR_$!pzvPZj$4*(YIT+gk0xrh1^jb?=n{> z^Oqmes5<0c9$F0*z8g5J6<_$)EAu5?!c=i3zgrgRcgVe(oK?lSRELhW>RNh&p3Kx( z2V>0SOR`2b9sJ7YZhD+GzQ1P?*<`~tTtJaH;tP;NFd6y3&kDq^_{)}0Gggv&q4>>AsLQoN+d(ANDnA27qD`tD^J?7B{9C0@GP6;V z>75{295$ATn%EpXbGP|mdFEEDo-hp-FZsRNA-xj!zh8fE3VFn?soREq9VmbwVRF_` z%u0$Zry>sj=ZSyY#x0w_QMq#JB*WwB9FGN*rz0o#C}?)mTmPb8Q+{%*A%*N?mv*#a zI5Zi+;W)+AQsfX7vB{^||DJn^$75cD|6*Q(qMeUD+?#z4g+J}#uB9J@UR+Qu8??*y z%JCeu`$E<|4HTj{QL`%?F$A%G>IF_MtWlT1g+>*)wTgAl?Xn~%pqzmE;B_Qg-XlE| zd`1-IUJ{bzG-!usqTG|5PG8bl_jsI2a9VVX1*WmwrRQHCKL9^Ed*XeS9!%|l z)04@58+K4G7{KEM#Q@9mVJZT5Mwm{<;_b8ZPcUj!CCL+PWnx9;UfmMx_f2=&2jvGK zLdo0?K#JdtU~plkZv)nVAv-8=n19V}DUi#j1fFr<_uv`?8`DU9RcGx5y0jg@urG5(Z<=~ZczTEJI^z9)`1juYB!rolu2 z6Ma|6F%&_$Oobz+BTGWCnqdNSJMM3%nV=Z=^T&H5J5Wp)9`GEg5x=UjwZk}CJ z#tR5D?=y`XhwjL?joS^_g!62eWyC}yLKE;-B+65PGf`)7+G+J>*O)m6fKR*&NG-;_ zkeE3YM#IcUjM0r&e(5#i!Y40!fkVBtS%P$CI@Pw1KfX>9by34Z*#n7yX1@q_qSz|GNr($f-n6?z7xgCe1;vEEfC-x5| zQ~yfVu-g&Y@Y<7W;PP&!m`%8Sizo=cNiINRB3O+QIJ_;5<1HG-RI8hg**8~^4K?5Q z9jnMB>1&MxOk*R#0wY%a_#3Y+(!-;~QRZmJVb^;c_!`Y>S@N6~ou-Y| zS1RwxzpqDL8a42*}!wG9$MzB2IMLPP^^2 z-T$`Tc6WcX?Y5QQc4oKT|8zH<-FCahJKne`ctaCVE^-mYAc%5x@U93BDvCESh>pXI zf+B+mzvoHfNCLqeNNBX%`4n=_dCwa>@Ao|K`#zWNHxkFy#NZVVyR4`DL_S0i&Ic8V zP`DNqm`;;z^FbceE8Pp=cu}B_PGi@o*QlYJ8fk~%5U3o()@;o@P+p-A1>tg9gI%5` zdm=OVcrqX6(>2p86*>@sg9`3@K$L`n} zCT{)s;#_HkA zeB0BXquOY5+{M$(GgHeOmp&w^F6{S(vfCk1qAW_0Nf9j-*B{$Cr1#rNZlybK(nXC+?|CL&Rm^hptE* zHNdv|;bhpT$KrQ>^s*DzKazK|opTd}8R8otHGM#SABa}bbKfFe@!BcrxV-kZ9`dN? zIUr-yejI&~6EY@uzEzfCo&D*uu6`S7sYTP4Fl|tY2rNRRlf5kMgk91(8c2E@y@B^f zrd1Be54z)esdRq5#JI7-J$9sp~^z5lY% zEbtAVlE=qtl~?4ML=zvIs=DK|{xwkODU`+wb>duzNotIgv>?i3BCN&jJA|cx$MIrV zJAia!SXEfQe-23X?3LHQfz_9=e#017=Ik~BHo;xh^_XOLT(%9m;opTFb_q0z(0zL( zmR({v11h~=jkMHBA$}QO%FMIZsdvA7oowLd@Nr>x@ji>WZYQMx?U?OU+%qR)6XbtZ z(Iz8?)%?g|QbXShO%y+z^*H!rUN<}FhK|NGb{FqnXoKA8^W0c#FC1;XhWD5M zsaCo9wG*HS+afjk(&B|j#9Ad@&>;1AHoYPQ$T-|_c0et9$5M~mJ2pcXnzlYWa>O`| zjq4iageN~J_A`U#sp9l%qUAQ-xNyc*g$4GDCk1Sj{#cKI}`p9XH?SB2H_kVCj03J#L_TYMjR?o_nTddMpYSg9TWK_L8D zsTy>a7Tox8??|*i?@Jt6@_p(VV7#8zlL2lSn=TxX0_N8td+|g{v6dnURGeP46c)3Z zcDjwe=dAx|#p_ZzPSI1#w|O=q&(|O(}L# zL`TKx=sq4g>hbrUNKL9g1|t*54N1QEzF_?{aN}e7b`?(_i4vx;r1dL;c&2e81%9kf z+PUs$`*AXSnsg>!K%jO0+h&~f^D^HfRa3}Siw9g!DNa-51PEWn>{0BLw$l$J&p&Ey$gCHwxP6ZrFNat5ha;P9m>6+;lG2<1Eq3yJPF$I2PFqY56%& zqXc_JX?*D2xjCwC9^TQcUdCYBlIh&txp)V*)+#~%2m(7gIxi60RRmSf?vU>sTbcKG zMq>EbjC^J+BgSdz;<^+3ldj)?)7=cH<8Rjd4_WTQeqOK~L*|xLN|8d5MBp-ma`I+% z4ZT8mdtP%Fr9+(``_zwmu>~_BPOhXq9M`>Vp%!(RBvXP4zsjvd;@M6`Gi<5R*pL zt@Ctin346nSy}9(RT{`O5SUsnYoybrI(7zfq@5cNPOR_cgquljSCZdk%;x0ZPWGH2 z%cc+r#teCDH&TjZima#NuEXNXBs1+)!!N&WMn%Yq*MCfox$qFeVY?nUZuZ-FkoW@px>mN-_ zjq0ekCW}sDvsKwX+o{AwnLjXkfPk1tN4F8|W6*6IlNQm(8kCj%HqlX7OLZ+8aBB+N z6uZy+!n6UmGB$Tk-puRs2Hd(A^sr}@w-f_zn>i69+Ku>;S4`Z_a(GKPA%5~ZN#_=i zC$ZCYT}s48VrQbbhHem4Pj7@ZP8IB!u7$M;8U@+(X+<+^ik{GYUe%DN(=OD@WR}-v5x@P(YLl8WB=yS@8o1kCUt+WWw-T4 z=oQI?vaxM?mJg1AGN=39^EORUPDG21F|!*)Htu)U>5Qt6Y3l?z8<}N|vNpjb(Tb2t z5A;$Y&$>Rk61WC689-!w(tQsq1Wn+=*6t^2d>@_l{D$|-?X`5-%*oT=zSm$KJ$^;( zg^j{jT-54lU8pWpb5(RzgjvC7z$KgUXHZA66WhbV=j1aFy2`6O%^RKacfNC;taagy z4#>g|+32KGiY*j8R|spbmOT#2o_Sex3B0>kc$zF#E?po*R?iGjVACr9@KNg9>B5a+ zNbfm8_RP-l9^U; zXN7P_8Av-(F-i~4xx@(-lb83eO#89fX8hnR^-c20NHst%yi9D}xWlC|8BA-xkC{Dec-AbPQPGdg-6y^aA zS14}3-mJa^M8s(k4T2iFFgnw-(|4J$MRAFr0gf zKPNITvLa{+B(j_oZjpA;gNrfhghWIC_0gt9&xwYxuHedoeo6d0ok7j&YY>}CWG<3z zi<;H=yGPV7x#{z)_2sue`kzF*)zl{(b)!LT$hGWgbKad~#+GVV*?*F!+_G6*7+dKU z*ji61)=*>>RMz^|&j3lS(ByY=czeYYvS-IOAmB0H(LvJ;YBc4AGDg0P>w2Ge? zc3E_s{`8-ISG90QaI?CXtXJf?7lUv`2DM`h!dLbW=L?}S+Wchy*KlsRD}zb3_PW_{ zS9L&;_OPm*Jq{Yp)gnv=(wrp8kSvQXtBHOE?S7t|pUC)bu)P3ysmfB%yv2SVVYVRh z|I+LrdtVs~a@N9vR8tDbuq~(JFzR=nZlAki%Ck=^f$hFoeL;Qv#j)fxLzrgJ`H>@^-Ea_)HgQB^y&e(gpj3P34x&H=TjVWS9n)g>G8;O zo%=(&L6{i(+1wL!#Z1S%GIqZRqZt^>YvX$Z%y-ZD+CZ6ghONu)wQc0gE*0(%N+ysM z>7hXm!W)YmGh|$!uN_0owM&Pz=Va?9WjCkavSyicSwPqjK-(7d0H`XK61}L7r>8GT zJK2<&W_9XY4}nxWi!P$u1fP7P8O+2w;aPx9AAL=f>3PEL?ZtQo#SV;EzT|i`o-efX zUQQUH%HG-aZ@%Vf=e={2E|6_5?7lr{F_Gj`id>3hQE}=18T{Q|)w4@u^wE2~_ei#? zt_!>9Bd`tq?_2BzpxE9;${2$@kGF@Uy?YJ7@`qb&6NpTo^UV}P12yQVw22N#dc&Uv zI?LRQ2`oe5W9Q&=ZN?2O?RzFqv1W~OS%YGOA!=I?t}U`u_q;Kh11oAwWQvd4G`kK` zj;{N)&s`15H|J!isn{)V4mnRlQqz@?4nbjbA1_OVr%&>idTG!u6-MWa4eYTHowzk* zHG7`k0Ez%`Aqy@vDU?L zw8}lxKo?zkO{9ao7R_a1{_b3QtNJ9rZf+)TDR5X-feR7;NSGsY#d~>BAnn)YrP(Dh zFql(?zdt}fW96&_4+y@F8+k)f=(x|5I3aJ+Kh7_lBQTq(MVB^zOxC6~$U-kzB|peWy2CTqAU z;RvPS9XRPs;@@4JGajZ?7k0+kU`pNZ)9cwL=n(8@YoNNUmGm%)OsyZ*9F_&vO&}sn z18O%eoHYKnx?L>W{q}6L(OCZBjCDlk!fTsK3p22nQWR1o4|x8G-Vas9(c@#>5^0rq z3duG$zoie=-1LFzOsbDIuzShvIgl}Tl9bQKL#;9;Ixo;DGGokL=m93eTPz&owD*Eo zkF5tf%6Ypn@S-VXlnC}(*@_n-1g%NYe0x^Y@0GIOp5x)CY^^FF2C1Y{CAQt1x8?KipdsA zkw%dXR2;s;=S8RF#=@GTOoov_VatqSxsmk-DRL&5PoLWeyzs#}tO2JHaNQJvONqSK zrDkY|9*BNJl3m!66j@+4gHr6E$TlkOf~=S1@VaB~Ns8qu5az+p1_?RaE;aRE{_~I4*286Y(>Ban4ZSh~a#MJa ziLW{wh?yvK#gZ7O zm=_J7gRa}QB=0hBhFc4ly;S(HA!b%jNxLAa#@4mnlEZ>9uBUa-5;W7 z3~X&o4(~4cFptTh_k;0SHO18AaB#wu-*&T^p1slMVNx5~>)7ze8BH(RCl9|qCVbgO zlTLP|)Y^M898E6F!ehguvfJw_Xx?Q+z{blYIDc$ zHS%%*KS(9CpII(LdZA-}$9xRT1zARjR(bXHr+lr$DIpw$D2GR}Z$?l4)xu|9x1)c% zk5stuUipf}Uil2AI7yLXR9v3VWwuSQV#-#~QC&IZG`TzXJe|SM7Z*ql2NzAt;BTMS z4XlCqvoCg!M;;UZuBLdAo;e}ci*o3$D7-bpqbuNm-^wZHXRVGoA-}|6mnWvL6CQ&b zugCO~F8UJ0t_v65V*mI^em$@z=J_q#m1V%tfX%lB#Z!;T*ybqiOP=AA&{=r zvu$!lNb#eBzxZL1^)lII_tiF*#F+e42BF<7RjDtExge);A!VvTD)p^ z{2gOyksJHP(y8Y8>CUgN{gJHZmVDs4UTOjfR6`14w^9l)lpCqIlk+NpcXG8XON9$y zBU>}%paeYfsG_@mg!X8aaObQEt$*Z(jj@dK*zWn^>Exe^&Dh8kY-%NmF1&Vukn)hx z)=~u6Ptae2Y< z33D5Aqb=b04P$H0ZMF&J-O-JwGRTED*rgV5`l>RhMcU}w=zBC|AE-=$7-y+3%5&7u z=m$ccMp+6-dLUn`2& z72h8ZR9x8eYXd5Kc~5=s&IJV`QX@!%{S(SpmC4$rUk0T|)O%l;f6iA6p1)jRgKAbE z4LQo30GG{4@`P6gLNtjC3JYX7H@q|+sMs~;V{CL}Ue@m>o2Q-St3r>GK{uDBjUf^bb~}2kl2(q*TAOwJKp1A*VDCUjP2^k!Hqc0Gm~F0EnaHvu65a6 zuMOAive3=U1EVhLp2)RcM^)wG_E{Mb=t42DiA-_OUU?_o#Gu0-giT3{^hm&EQH^wo z@K|u7822lUU=~v6SB32al69>TdvMM)Vh>`T&ln?qaxhj*-El(9q;FmP@Qk&e4mMq znXim5t=hr_6;ldGt1O`6kaHB5x<<~??(jYc5O&COrC@w!XuUwQvLa+9Y?P5@P}B29 zM#OTN)=M94I(g4KQ4B41dx3hTGAE)va=jv#PKj=y&qQbYJ3i+8V$6(ePKKRzFqUTc z zpyE7{(~X_TFqsFZU|{4G~pWeruCQ0B;DTnoH zFJ$O_nbkzQVUtwvswgWmmtHEwsyA}7u?a2{@l4e00{ zaoV{|TJ_@_pUyEaRDb^mLj~FG!n@yN78AxnN^yW9`>D9XgV&9#mrEyj-wN-evli4s zsAl!lUXnb0N3fCP-~rf{d~r6tI_50d6X~c~hck4BT8oiBkB$htuX2RW`=<~5Wl6Cw zncf|JHuU!ETLbz$GsK#!q$&(@m*_h_7|?5wLjpl9b0oM?hVqCl(#}BildP6qgtS^m zA;;!WjulciJ9N;6*HgA^3O3u9Il4ckkepX$BV1=;jY=s6tXlS={Hfj#|F2c9^UxGU z_5cZsG5E8KB)g}v=fo?ZUJR3}uPoRRj7LWPlj94sB<&#YQ_Qr$-uF->Y=^N}_E1oZ zG}*mHr0Mg@Sb;N=8G>^Tg3Y8!xm}mfTJuM`?5f2Eu_Y>t0g0b) zoo^nKryK~m#qw!Wih8)1nZHocc88BbM`T(Nr$>SdGgo&&Gw{l@F%4t!-dx? z;G7?_!QDqG_E4by6}R6fU5M^7Bv-10nWLS)NDk2tqe}w%yus;SAuydRnv+dqXb|PV za)F3ntNh$oQy{@*%b**Uap0#C_q4fdy^u=X34pQ}GB#|1`LyE!#0fH!CcXaqENdw_ zm$f7|#Np7hYV@6A{1@4=w90E?529iDsx{^=9w9X$=E>&nT&G}Rgi#y=JVxln@uxV! zWAb8cvqj0E`rak=+*T+q>_+*_!v0*R6fG2Krs5t->v(N~3$jb{8v1f%yR?&j zpy=YS3QC`r!XJRlX}rGPe_3d}utN@IKD;Zud!co_!}9IkMZ8SaVHIXpZwvZ8)@FPr zKMqwur&L;1X-sbLG0@|Epx8iG1z{KZn07%jzlGgEI^T0s%l>yjc3xMsR6kxt`sklqQnV#3edLT4d8go~XEL_GUhV^+i2-Zwb zpVq9tO4_Bj=XJ8>;>#e^P(TJ|_jzWD_botJ#S43w?$~B^vUr_u2@HCtWV`nZ2mVC^ zVB`9@@nWOj@t*$scfJ{9Jzov4NOfVeV`C1xBQvmA?3M0+MDcKD5->6?k7!nB(f8RE zKzDr^49{o5`11t4ZmP4!eAw7<1;MZmoOHtW9Cq&7yJx4F%}L7}mp&w^+{}p!2ZZ)p zn3F6@0dYeuFsR3-@<1YJuRIy5MYaB0eUto`GmoIiCyiYV?6;7q(!`{(<&oe6Wb2`J zP^-+2$n@M7bCNl&C=PNml(+)u`K00`8Aqmv7Nw5{Aal58Uc+bApa%ZRwcAgk|r5xjfNN|?)io0M>n zOXmxlln*?d{9al7i|rYJuU;&RJDq#LX#iZ;@ak`-`v;n*lz-b>mqd2GGE+*e#gw9_ z6#FQ$hl)E4t`iIiwMc=TVCHg}29r>9;%rHWJlVZ*TAjpL42b2)deK#8sSsrljbxIj zPlsHSJ5^0!ks4(=5$hFMz-67renBevgKino8)L?o5jhO8oWTFpTQK5>VSSJ1MB&G7 z|G=72vpFT?GTFopH7>k@EwMmN4y6Es;tVQo%|DK)!u=A@z6N71*hqd`gTOiH>a5Ghr0bk4OudU z=qz5m6iSguk+oD@q8Mvl?uDjDREciUCnQJt52NCRPJW$6d^UDH`VsGR?nzFM-gPrY ze3fV${2jae+v~z2GM*-D~kvhuFPQ1hksgpkYahSjL+%eo4 z=DL6e4auP~z*I_+LXku&4g>iK9vId^UbbDl^7$D0Z=*5VKjC)m#ZWP=CY^|i5xcex zl{&@y2gkGabYZ~QSbKsmgxLz5X*#-_1-gM;dY$YzUA(Y})&*C9+x#fg8IEeCnhZsH zL>7bxHwK;a)u4t>8;uCV#L17G2rPRcG~%|zaakk7U=-f*pxe#7$SC~s{XdX+7tR;Z zTG)zYO0k|IYp6J5`q18KD30DvQdOtq`ND|`@mfAly9U|vmI*nsZ6`+ZgS|8F&I9X$ zV3%EE*eLPcGYwM?KUS?~3xW$Ffi+*)CD{Ppar9Q8?$*T_dbVAV%G)1oP$o>6tIjTb z47C6gf)4QCaw?dS_}BdMWj7gzRY!tRP#Iz)dBThcv>*e{$Pjwr0NMqK7xv=#2Am)< zSy1{93FG0?b6wXj0g>V%Ici0e0xZsMDh}Dt2BR{07)7gG@FYAdK$FgP1!SwL=xU}C z!UAb*Z}e@?)v-uZf$=y9zD-S_x?(z1iR4bCDVe^FHb9zd36nr2IEm@VE_e;KIKw|X z9$V1k!^IaD{L;Ma*y9oP8oBAh%Z>pH2;8R>ofP@>8L2>l;Nh9?y=s6MtNuINxt@OC>v(+6b)pS8f{%drZWALt%z7}Tcqlu{_iDJ9~ z>v`}zh-gF{2{Ce2rx{ZntAW3uMs?KtqblB!5c7hzP0&u)(5Fcc&?q&8T?<1}B!gcv zU(>_1OB-cvg6H30{tMe^_+lO^QeHEDy=h^L08V5qSHcbe(QsIe@yt*(amYMZ-9;%fC}3NP+f84WBnb6tt@12@@`0y;y{%p!yCn9yq+c-L zmQ82G?h~CRN5y+X^MwO$dCW0#F!U3@2fTKN74i%nEz8ei#FS+xPB{9>yR@G*f6{P& ziVKHkY;Y=N@Q;MFl2m_SZ;W2)X^46pg&eOBqkE!{0ZDb?f&oZD+%C8Rxgr z1h3(l-R?Ruw!KPn=ivSKqfGZZXR|y;Tnx1)PC&(IOXL&X-|DiK&gQh9c3s9c8$#RF zq9UNNK#$4CGk5xfvn3-U0}_nP2e>lA7N~n{6tu&V)Fx#oIA3a+b-a9WH`ITns&L|Q z-Y$LIFmeirCQZ7%w*0BJ*cTVDxUk`|Ar021+)baIy25v#s7;U{OmOc|UnKacFuHk8 z`TS<}=hHWYImeXP7(y=4u+fE6PXArU%(lNQdXp9@rj{7lVap}=*!@1A2ekvoCyJg= zC?aq;?L~!ChJKVzOXRbL73U0JkGSyu*+#N+qYSsm?LRGD2-$NN!`h`L_Gt~y9vKnW zz*wX~9Zy#bP8`mgCafO3y&!nL+OX=tNr$#KItkZ*9&Ei_by;7njiqVD*KSBw@tf7z z^rjhlxUMd=O@Mv z$0bqkO|zy&d8wu)8&oMMg_03rCKbS{iA=JAk0b?00C~5+488C-EoVz+~vp9pf@7*TAE7=a-H963<3>n?=f3$5{ zzH|Szx$l^f^Z1+J`Z+n`!cnE07AQGSDS*5B3>8-uwlOR|v<~7(U6NG)dr|wCLAUiW z>9J|724gdLqMNOV(kj=;4Io@wIqRY4Q&v+Qazm2HREdxq_NFA4M!rzgSxI9%=p(`T ze%F0aS3y%1yi)-k)c}Eh22;r73(wJ-UIL*dIKWNV67`u`n$w~rS$**NU`!=O^^HNd z=l^k9SvwFwLu}BzW{8hvyfi?h+ci_6t5|Gn$)zi&222me0sJ#w1cR?CHv zbJxOnwNQ#?id>-LZqgZ&8bM}ozvN+5Gti3`()FZ^pG~h(U!7UYKJ{&o-u7r#cljK2 z|8zzzl(Kj63!{&VAIZDfO*1ylxa*5on$>4`6{0@RhoHrvr}6s3nYUzRyrC^h*d@U6 z(yZP<@cL5Wb;(^{%ll3JKk~$$R)d?lp0ts@vB$`liq-S+udp9}jz!nV)@#UFy8%{q z?)_hX+l;P|6|eu89D8NZWw1ckMN07rMH-Nor9p7hr;0}AbglBDs78)bQ2OXx$g=1G zg4;_#mw{~%6_9fw*%Fi?8F1SpISjg&eSl0|P`jX>m{cqg0}rE)gCJS13Pq*xK13r% z!4D8U^aI4X;`=@)=&G;`_abIf;D!0^@)}rtrA4$0P?ot{zC7?z_>qt{dNrHQ*1L12 zESorB*^Mbk5n&te72I>YEZ)1bcV;O#-*)Ikfed#(Zc8PWsGkN%Wh~bEB#DZ!}gL> z0i87JcIA5Lqfz(kW+X;|@IR!C!~{x6SEQ3?C3GQ-{BnzVL5 z@gXz$$R8a$4BXyZmvx!=ZTR^uh1sBZKIyqaHjm`;ap9sQAle>cHgYLN7DY0txG&}T zViPqj&?6uhp+0(<7ecue2on0daJR16p{iHq(8)xD2{k%#g#bvp3xp5l7hwfbBtH-C zpAQ`9FSdl=xM6MCox_f7!+PHuukTuuv~bytyD)}qaLwrG^$M(dd;q`GLT|G-qn8J2 zmA8D_1RJMz2s-F8>BLSy95CZBJlNAw+Q9?K5kkx}zr+7aZaV_6(=v+EircetVG_ z7*QX6(ncz|0mFq&2Jn>*0mfNMaf%|xsW?<8-0%;YmHwEgSvGSGqv`c5gXm)~I7?2F zG+t5UY8iNwn}MkZHR>v7>BNJLfX{J7*UJAM{KGa%RqG{Luz!~t>9 zv?cDXY@g@$;KK_|SFu=N00?RxAEz<^!f!?ls^1KK|Xci&`rYRp_`XJO7WN?52?6DS*NHmrUXG5D6gKn6ehua=Kfevwz(ju*;dy&UC zmDv<=AlOvLeVezG>6^G^nX}~p|>}9EeyUW8d&-}%s?wIf4+KX*-vh7=fFl!0k;pGG_Y#?mN z@}B<24GH+kW|YW#C5`U;d`?5Hy^c=eJ$Z9Als4rtwE=obE_0K0su8dprWb=_=;UE} zuQQKqgQNVw71a#0v1nfqeuJdD@S>*N!Y~w23P46Sa>b;vPyYJMqFdp|fnMeFpb{u> z)GEt@w<@Y<_jw(aZt!YWFN1QEvx)CwdV;KwHk>d)8=~<1TkSJp z3lvuc)Antx|Hs;PaDj&lS1sDGG>uSn)+wum+`3Z%5coYBQb?Z)XpnW%<#fC-fqYc# zqYE-9`@Q!>TC&AC@wMg(j)^ig;_%u#zcH^+5B|(cM(&N2zUIOZSZ(2odqOFCDe{Pl z!|BAp-kh~fSutDF|BH@=P`cw+KL1W+vHVVCo7q+uig12i+RnffrNO;GWDssoLxQ z+4Lfjrdhe0Nn}3xudvDClM6YJr&26U)07`;2rnQd(K>^8Ph zrgzsWKmDdrBQJw_34L+Z|2*sZUsg(7I~L;OlBdMa=tA8A$#CHJjn+rj`Tt{pez?U%1&WW%_4rB1Kc6#B`|Fy^2zRM(sf=uAP2NO6S)@-5)$! zBtF0c{C_TeO_b`dDG*}i=nDP{eloD$J`AXZ0%Q!~X_EOSLX0MmSU36J8)f<8ijYnk z>n}8`gxe$1J@BU{F$^h6waWcLTbLxTM#zI5{}6GW5llnyqMugJk^3RHwWrJGWz&tn zRDahDpu17EUyzgBT>38T!R@ek4=z)Rixl|;3x_Zp4rd^cyu(6kQ%V~wfRMhSO^~S? zjH-c>>W$u=V1$eiUrKKCP;+8q7$&jQ1Qm(8qq9`C%tmjGfqg>vNWbOIGqCuMoTq_0*z}DU>C!5flGQ)8(|v+dNZ^VgtXDp9Y+}+k$lTF~6t$or*g7fZMj92gaj&%!X)P5YS$! zo){|{_wufZjw1<(ppLgK2JgKiTjzb=8@X4Py0_2jpL3h-qVd~m=oP{`9{QA<)dq-M z#slpPPzLx8uuhrkh}mjGPh6 znQI=Hv5N&J#<@M0#&s~3P9fc1vm$}&^J-EyDUp!|&pZt;X3`??Hn4m3?)C0<1eIRM9BZAx-sH#5|p z44INDC+&@%i~*zO^N&hb;B#|LYr)Or9316K(*1nwHn#u2pr1_ko;2iE@|305l_A zlF#WBULSASoXZZ1fpLP*vx#E#-wpwIZGCyYzZpJ-gFh)H86$ZbUD$e5S%738rGNs0 zd{f0zM zV7HqugoWcV_jaj{&WV_)r)hi)Tpp>zpW7%M+I|;D2t2#)>-E6v*0c!oTqFN4%?`46 z3OQ?GzN;w($fK2GdQ&rLCPna?WqNkA`QF#UfY;xvoHm_Br!O9#7wzJmRh|TwM`Xmpqu9H2);YFM4n78F^+-pi-0u-n7^n?hg@G+yJh9;W(%{ZHE$is zcVQRs2@BJAh*FeMq=bsA5ge!QkOP5BU>fZ6YWBV$%K`Umz63brrDd{s4_v)=Nzx)p zeX&j+3qr8e1f8}xl@>E==C#oksu~)Z$V}|>$UzR(D-fYZFP$mO8P9>R@P$J&1~`0W z-iCj8;n&&SfRojmeC*e$ducOxdVb)4nk?nElXl&Smwi>X7&#{_|th9fu{7j?N3*8gejpB4!dh zN8Qp!Ou*1{G91(guRVBsz8NE)k4~hKG8e|k1q+NErxd`aS4+k17gq$eva9&F{(5WB z!|0q#<*S5*092r(U`8>^Y^Kq|IlRwchkUOo+5 zG^=~QeM2%J-xkzE*YS=Bn!F9(mqZ5wAARFfpU$BA>7`PAz;SxjlvWlKvrKYbB|h5( z2~)<0w=p5f5?8j4#)UU-NP0Wvt+Up(23*#+E<6R;cL4D$1*6O6!v(c!e1YWnps$M1=>=ERw z+?|`?;RHAA=wo9yC`Rwb;aBHnmOuEqdFI&@z&s!|+?K2^oI!luVxDTC6yPa31@n|o zJAHtZ__Yau*F|7by+7jhG`v7~b*3gy)i2H#$A_l-w}xSQKy}F05PdXe8sq@+7$ipR z_PWYE4e#TnGO0{Coy2Mk%#je-O!IC*>)HiHq7N`zK?jKmHFSd@Uwmm=MbtUajWd1E z{&}mXEQlGRg4uEkhvr4&%)w2_M2%WGxo7XC)7>`<+YN@MrV(*>-IDxY8JvBLL~ zZ{v)1-VFgB|F7}pm}TU$*_%zPi)kz3`u1{jtx{CL2~G&!nH=$-8K)? zfA%F#+WQKQ-j1W;7`-zGUj1P6my4}s*oN21y6}u+Lu{>Cu~Yg?I98K9CmW`RD!NWm zB&vt}#};Xyr=|jmShUJ#f_`aJC&b`!gn@CkaNua$?8^3rC%0?fn~Tje%fjl|2-5n> zELH|BCddbr0@8c$QE`8|#V)1pkQ@kC0Ub9|QtgTCC7Kjo?#xbF>u*e*#1bUTJLqF; znG_yY$n~)ry}V7ZGR&CwO;-q(2z$vYD4t7tdtY#$R|*gAEQr_|0Np^Eqz1!U52M@V z#`BnES4Y~Vc*2;tsr6qq^NIpOXkvqV36%17(jQgv@DG5$Ll3H5I(msvQwN#orrxyv zAGMNf)quP}f;xhSz~-Ts`J<~)CLSO91-TQtQvoZTk5FD4|7z6wIui1c%N9A6HoW<( z56qa;`_;=yhYP!AR#+6xJ*E^7De?gMtJ2uBp-KMh73QJ=^yNQa)DF7c<0UfDsljoB*6{QXGRGVh!&0E9hp`uRH2kscHGK*j{yolNVK-qgu(l<5VLzC){3bXsB zC;8V%Q^n=;jfuDMQG0o)3s~aU&FZ6dkkCD}^ zx%Az+rY|?0!Jh`U-}BRddUm|dyDO*!veG|uFA^OK8QS)l*&K^y6}he`EEP2*}9=O*Rh;8g#3z(BsRinrd0^dTcm zF;?0#+N;&@H7HzY?~otKNrL$QvEl~UlX1~*GFwv_CvPZRe|0Q zr-_d~>JNM3o-aNLB`M2D>+47*@ifAD@)J0~Tre`s4mb@V@ON=>0g*`2Q?jQ=GBoLUWaNTH&^EIw--ZBgv*FsAJNTSV4Yj%vXhF9 zq(AhRVFDNa)}XdhwlK+G58=pMzg+r3bf;_#%>E-jEIUuyJJx6`?CjDZ&A8bM`TwLZ zpJ!g^>e!#=lLJ$r`gCa2|0tz6LXm1J?u)PAz&bzXIIXAm#{O}oXCG^bT0wRMo9eUr zV%tEj5H74+WV*CIrU;ZX4+m>iyLebay2~qf&Z>yY1(hC|s#O1)kjsi*$iRCNzM2Jx zASs)_%d67kaPW#5yS#4s4EY8Fi?W7A(^imY3%xNtCN9G@%)Gzg#F+jd^G%$F%XRb6 zdb9o{8;^_6b^m3=hKnyD#30`k*aqJID@`-Fti5>)kJVapLmWWOqz; zg5+}gUTA}SWkfd%S^Z~~HPUY0sR@?8;;>ldHd5nl&B>Zg{#o0bKeKWBDBiKk`z_ z4YhLdlly+=$>*u!^lGBz7FKfM6)V(24VkKnClc0B zyr90V#&f{%9GfCO5`y(1I{FfWYntVOg&^*k$kYNSXRWsZRAL)J)E=3jwaSc$^%0+X zUT{`|#`xhi4h)TQ>yaP-*geP$nB#A6C?{F33}9e`KLju(l%kj-MJOw=EcDW>k3(+K z+m+qX&FT$gS!fr%cp0e>m`+rNEeoymzrwsM?iIRZcDCQxxhk<}u1j0mykw*P+ouWmEVSovPT zFW)EIxnaeHm$8Q|uu@1V@+gu+#q~z#0xfKlcNbkE2N^5g4M`Ko1?TXRc!qbbfs)?^ zb+bC%KU>(WybtM6db*2$j9|Xe8qx(!hFhdhVPSh-q>U|+4>-?LGcLF|(F`3eNp2Vq zAD9d0VcOsW>k4X-_5`he4Hhi=%vf-7nBl|4u$WBV3qsVqFZ$$>w9X`5%OZ+RLCCV1X1=$w>ie~j5@7+O# zo^?>bahxt@H@tP*bJOewkV(?fh0%RHM{it47{Rewj}ba@@Ttowe|5=vFEf16>cUfs zjr~o**FK65HINJ_Lq+b|d~sr!&OZ@CdFc9>fbq5%Cc6jU;x6YO*j}$b68hg!n{7vX z#LR8vunTW-FI!lH)06`EG>%en=Ts{_O|%zAg6%Z+KD_XnPetugBQKDDhWkND3rIvf z@Lc7+FL-6t`k3@sO)Dfa_j^~-YshiB*H4o?J$X8kYG0kR6-ZW4{sFlsfoQf%ruAPR zp;c9rj0hye?SSo(rYh{As7g>j)S?-)m!8z`lc<-Y#K1lQ4A;liJDh^Ze+odRCa#5_9;rzt%OkYPdheVF zO?t$I`RNgbaJWDC7Cq>e%ts0|5X49nuaCe(OqolblL?=VT!dxPx>=aYFzANu?}esE zTqNlc`QmozHYU^aP{3(&VZN~l}5B>dO+HsLIOE>@W^uj*&;H)&Zm90>95lkO!m#_5P7jr$ZU$6;EORg-~ zBtAH+L%o~czwp^zc*j}gEgy`BZW133d>EChYE~bdRVuwMtOP>mhf(oDyu)C;L#s?B zn*t5;XZM_n>7+M_>s1efaOXvh_ws-}uFV^}^c5S`{g+jju6s zU!V#IkkSU$&`=5H{i%aUhwXoz?0%{3@6sVHIjuNccOIZm_>1Ib+oYTMum48UUK!h@ zx0vCwDFwh?N5!F6CYg^M4V7Vf(D$nn?U(07z;EU#(>rbRYsfcM2#S#T;v85{7X@zz zC-J@M$Nbjuszeu@UQ}c^+^wf9hk)94C!g>A?ay9zWoqOv#4=Vb)XAfpzd%_4RYG-= zOTNaaBhqf8s(pz&Rw82IqUI`;pI!`Wm!^k21xH>uCJ`2Uldt%~S@=r<&J3Pk&d#$_@n#SV&WqvCd|t0CyCIY8F>9w7Tbg=Nrf zW7tv2ir2P=Xh8fK=-~QzX{@Ogel~!_%7i#n$2Jn%Q~x7t!6KI} z80xgZ?pG2-xpVqG^O&Y@Ccz>Tm3}gLoid~M-|Xf=alvl+Ip03d4k%|&7ebNp6Z%tC zDF}SxMyVZ$+cRKoo5q@0d|;m8Zv>{p>r8=gzt8@ls+hE?AUuj0F?gX|sgLgSzAwPa zdY`qj8d`JH2jw75H=`gVGE!qnHQugOK2lv*Y0!)f$Pc=s1E}A#Sbh#_{?lHh7BDvG zKeIsN`aS9^zzzCmfAp(nPqPs!f9E^r$=a7pKmvl1Ln7Jflwu17BNO+G?q#hP==N!q zs6+!gG$>j;R89|PvI()1#u4L;m{G^saTjePrsCVb{rO}wI+m{rJql{NF3WC2V5%H~ zgd|F_jv}k6IFkqwcz+FS^}IyBegX&FVu+FeQ)a3V?Dm|E!4Rf!yN+FCJif zF!REhWWJ;=*G;HDQ{ zg*1PhKB^!ad$sIhm_FKg84kh3psCSx%TETx3lE4jxJAYd^==>yY6>glua?!5cwsIS zt@RL1Jt+YWhWHsq0`DC$Ip7_343Zpz9ZPU=#>q%GaKg^y?=1%Oy60mG2!{9;0cLRq>Jg9WJ4ya8+0bovq4){HlE6N8t~XZJ1~2Q+LBltVCM zv-h!I;Dj0Kt=#I()Cr{7z~~bF zIOd3h6qf8AK@K>ww@Zh#Tl2kNthO$Ew86=b%7i;+pC)I~Xrt!$4xbMCzI!vw1m{A4 zC9s+g#KodEvQyf}UXKAOZjW^y_^{2%k!HoZb~*5S0F(O07)>$wF?weWx^*(Gvb4|& z6{J$uGW%W|bUP)#YFvZd^64S>;T7LXUH*3M!ZIMYL?6mnUvmyaFvP5L_&CM^;q)rI zu5VAe>DE74+nARsiO+^5xuRGy4FgAO7Tu5}G8aiPTd8PP=kq}P8;{=B-2C2H5jiHn zaDnldzF>?u*@p3FkHkNyoxS(YKXb-fyVYfPLpIb{F>cs04aye{?1cbLInedw zi?4!M2(C2nWTW7|yO9;Q+ZX|u82oZE66P6hjITYn+c@EN(m!jDe#hEf%VpW>!c`?U zoVyhvxIRK=vMO1-wBH-}ktYT)R>Noyeynbpki+`9p~82}uD8eE{MOIO5f^s7-L!DM zou?G%C~^kHWV_jR2-`sXrcstJ?DOm)RU&UT^2NqR802}DDv30d zS*jLk7pa~N%jNs-NQvA|qvNjB7sGy8s&(Gj=rZK?wn(omfNX|F!J2vCfT`lu^0VnS zK_a;miHYK$%l0n>ws_=BO=D45UUS)RE=@G4AX>5bEDXeF11w;zUAu(zt-@qW6GzNLUpjd2>@9|zH zG{`eO+XTqH-A>nF^m+_M1e}JC6XBc%YPft2=eF;r2fZ&bTOawKzTQNVT-f^Tu^4(C zrPxl9t&pXQ0c4apgH1e!64A>)=#~SU`C6tXq~neC3a~ZN$iHA>TW=r7j+@H^9+N5I z@&#jT%gvO0-S@}uyl+NG$iGaVPHwv}LY7N(RSVM;NcIkMK#B1 z-PCMV1HD9jW>&sgw{1+ov#>diKR~Q<0^X$GEZ7`jExqQlJ1HAtY8lk;;;B+;hbl!} zDm5N7LCyVg*>T!$eoMz>_9 zqFGt&vs~6lXTvh*q3@xXXMJJ;|CP|IA#H+Bco$>?@=kV}^fbu}NPz;)P2ai!L5P)} zN&d^3o05%TU7pzTf~?4Y$E@n0I!@)e6ACE~c8wFHs0(M8teKwRq1oZTE27`KkF6O)UEKr$iUX8PO*sK-Qs|>2q1GBC zFO`SbhAzoD@C@%~YM@*f8C&ytl^$5?)G2BZ?cx2cvLm<>-)seGrxm|@c*#R4{+|839z_4EwE*6$a8d#$zNhz+TZZh3pm z$pFkX(&$BHs*5x9tOnDXu$*9`pQ+ui-jILlM0~XU>WvUEaNV~l*zH37T=Sab!%B56 zDRkjA$tjC9Nd=`Sr${Lkho#TBJHg-J0lZ33b)-=iADS=GDr;0ny)}vaMDVpd4Sy7# z#$J=Gi$$HhMyOyKbi=0DFkcMb!ss>LM$n6yP%}MlUVqu=TL%5jQ0m z#z7D1YW7HgR(V;}2$Ts;%n3+LOZCrV_Ds`5sq9WEy3-QHU3BA&OprqCA;%({p&~gE zIGB&Q=YUXnrs`180I=dd@T{KQA>ZlXYBC27kOPM0c^8}p$aN!?!56<+HQS7w^aB)U#Z?SSE5CQ$q2iFD7mEszTQx28 zHfxj`Zk4quug?P!SV;25iwK;xLd|AHCk?!-b84AjIW5ymQAaPj$ zncLb2o8hhA{mpyJC^L|rdOfWt18$sblnVn1f>lF`-x4XsT8bo4aTpMXe9}U{LO)a8 zE_|xQY#tuMm=*@9W3;t9r3@`(t=RMgTEa|)&wWNEn~xswQ=%KdFszxUm15O(F8zgQ zEek^71{T-xdeKqo7Y?Vfu>%J7Sp&x*z}OQfBQWWo(|^9`FDw1wQsgxcee_OPpWF=I zABvRng;J!I*U`GrXY}9Q(I_s7EwfcsG^W(xm1Or8QG$n)jn3F%WwRJr?)rG22PK3ou&VBho15GNYhGMki7eOLGvMrxLYW{9yt8VDTM z&?(Uvm*$waY1dGU&}b~E8KE18pR$db_HWF*XB}7^ju02#2-yf9BKK+zYQX6+TXrut%BpSnNl`%ZEi)@`7Uk~V2q+eaLcDDQXgIRdb@Pf zoOO`NbTs7l>jjZe7|ot~W7|BKE`a?h0~{i?Ocif1>PWEhW*|2WZw<3=t{BUB*@KHw z=8ds@rCs;f#>EJ6P>TNbP4Dr@nz(K}gV>NYF(^)hvbRn=AV=cfWd5eWZT?G`Q>u7j zK}>vT-$cre<9ON+BfCc5VU0XJS5CKPP<2@&VuKU4B@FkzxHN5}*Ttstwn)zb8*(1A zFJ_<5e&#w@jGh=Q1!$6?Lp8&_$}2~e$@}z;N^i`ybe`!j4U+vMZ)HXtHpU}&#La-j zx&8Rq6!Vr@-TtA5EP2W3=s?qEh?6dbQY2DjEfu$qcNH=efS;w*w+95Jfv~WbRPiz* zoQ;a%V`3*xhTrJqGq!QU%iTTw`({iW{QJQ7$mx+V;li7iyB3&ep%l#&xj@CCt!M$~ zt1i^YONj22UAzW)GdNTAiNF5gW)!eBVk`9K)yNP zmdL~l2iz{sSOdDS*kI7DQC1*19Nh0+BVP_7Ci~XQcB0G51da314rJO!*XM@W53C6= zU3O-&L3Y`SqK&YpF^xKul~gE8{Itqa-#p2nTS_##YOewRPa1n<)`S@xu1K(B5FOIQ zHWEfy&`$c^eAQ2_9VT3o5?pwrY{Qk(4z)#TEQ;V@(d#K`JKZTj#e0Z5&)FDO;%Ceh zZIRxO(v(0dU9nGzpM92Lyo1LTPez+E-Y>P|{;OMG>tA4=otl2KJe!nrOHOd%dFztJ zymgXN9HYp`RGfh|5ab$ZW%Izb`}Bv)fovQ~BijW=jjf)!7zaA&*(~3pxaHFV97GL* z<*@d_R#@31#g?QlZMihK6aZR zQ*|e*69Ngf%sQT?Xj)gm^}sG46Q^swuyYKxAuy(6!VJiSo*~^ zvY4byArRypvY^bQ6k3X;Q*lkoMuFZhOI5!J3HveJ)+R`azBtXGY@Y@acp_K!Uk~Z`k)b^u*X6u0#KX5j*2%_+6IfxIckphMO^x>7TWU6umO` z=8Of@t0)D;Bo9(?7(*`g%@;y(D;6ALA;?l;o&=KRvY<|_K)A_cZ!8!KbnBxZ2I(~O z#U)-3Jr9s(g{B+kGPEtdV27ZG1E4@$2Qke%5o%Gg_B;H=RJ`!bc^ub%C?U5P$iYUCkDKJCQ$HTtH z2_NHlIt4>^BgJ;ZEuXjki}j{yc$CP6kz!*bRRiPPtZoq8_bHM0O8R}vz_4Vhx}v^R zR7akitqTIT9%w^uAPI~kM%VW99eRZ$28Q11)X%o(({FzGtrON$l*?|ZZA?0MWJjax z-cI9HDuEFQYngRm82Wfeqc6{6fYL#of(kDbzYm6h|(5F69 zXPidDbsb+z!pX6hy)tdTPp@aQ8p(p5$QO60Dm^|4NO~Q4f%2H6vlBx4XoLKrv_{@0 zJsMK#@fhq0HqD$pAYTS_9gjgJeQ}MvSI_~{>J^g8utesxXS-AQy~hY;%V}vOERORW zCn!@9Q|Py4<{i_icfWg`Y#1p3<-)7oeHQzoos2dBaxTvJjNJe`pZ)SK zS!LKc@n>waI@4#lcY->VJrR6F^3d}VD52w(Iw;f4ol_=v?8GzTP>u~aBX;G`leU5L zS7$z6Gt&%`?-l&=eX`w!lf4gF0HTmmz1C^Qq3wc1@nvbA&}kqU8#JDq z6o=wrNN2X8F~Sm;Ql}<0PBX)$<&8@plGG_=zs0CzQ3~Mi&{A;)^gz^Q7JOWu=ZI$A z)OH|*OZD%CpZUUd9-0&&pahpMRNJq3ZEvV1jV**oFsOmd(}F7C1TC1iTALRE~M}*vVt@z4ayy}3n{#>&�PNgwfoEG_xNfLf-I z!hV{Stz?H++W%wkTi}|?uJt_wCnN_$UIdd2C`bfB9C;Wj;zVty(@r~m-`jTjy0@*| zw*K2r+j~>n=?hU5d>|;GpoTXQ1eJ#YTYEfSQ(9L-4x zzvJ9H{zgvD*=GmO{?6KKuf5jyjV~c|q*hg+Dv3NWDc4`Ke@ch2J1mKl{op-y|MKgeez*6lm&GLzi)~YPdDrmT6{X^4evW9Z;IU6z;3vOtO@OwC5hI0L0lo#{{|-;mZ1VyHoi7nE>&$X$>!-mE+&D3k3Bc*4_iPXWbrt5?qy5OIT6 z+47puwa>0UKVxj!rq19Pq8dfR{q^v=WjjSl(J3NFr!doqk3QU94JBJzJlcO${dwGZ z%JJ1r$NyW^wZezgb@drw7RsNqVpeego^2ErLNXTaa2Gj6XV&_YXNUpoojBnvy~&B5 zRX5SH{)Db@!@G>^AuGMt_@JY(FR*U%7P2^?R+uN=8?coubGpM|@my?w))qH9`6KP%)24W2hk%5U$w;j!js%Y`I&FqBbZO1l8_x z4M6l=!0nCf7Nx3;yGY%tPk=9Px$6$W4nZRSp!??W51_#Lf%y8YgW+**6%*UkrE@Qf z|8Y%z0QR|&*X)$fk*IS)Ene;Nd*l^iBJZ9@FOYn9N@98Eg$qV!Cb1pZHhU$b0(oY@ z{tEYx(QlgY6Y#s51=NPuECgI-hN&%t0yxQXiI{j-!_TlfNiSO;waitkTpai*nItku z<6UpX6v6i04^5Y&a+)4KAHDmf zUN;j|8oyBgPiiTPO--&DEK&(Y3PB|iF(^xI1H|Z)l;6|Lb4)n^02n^aoW`HnU9t_WsKt;6D=VI|v0- zB^47fn|-^1KesXwGhjhD_d{6jp3i>DJdAF&TgglZJkR@{mWeL<# z!Cu*CzHMrhzHOdcOQuY?t2j)R&rOM1H-3G{1Jxzp^WIKyRBSavFZXRn|3&DZF-_x0 z-5QG_LsPdFC(%n1)U_Zaa#*!DDBk<=5Lt79_>yX;u%1NXX0(utBlE&{3X^$9IETj~ zjR6MM4up(89IW*CcrdEYW!R`Z`pw^{SY}3r#XE~oAvI$CoJ=TI6VwVK#>jSydxjmt z2ES|G&HSvd%s?GCj>;z= zj!F&BtH#6TQ6K3;ysFbtxX$d=U;dqq-?CJuXR@)!;^~A&$G(YwTf95OSSFw9-vzYn zH)pKj=_R+kI>c>UWMYmJXjPrzD=0gwlqK<{X^=?jCx6bU1(i`nrDJv(-SnOWJ& z9h~ty6K>|09-2sf!ptRMF}p&7nVRqbp?FA8_mR%MmR#m>15!)|zDA{g! zW_8rYXiaS>5_2SRt^x_eCgo%C(Fr9}7vvF1Uul>-+;YKFm-r$2j%?m_?qc9}Z{YND zb0$NHbTwJ(-cJ>ahFe1IGp3FvPhBx}VY9T3swYcZn7p(o;XDFj9)*b zESyhehu44(ACN%9?}8jb7pY=U?CkeI)(3-hVCcRF;u9eTvi03HTdUA>)eN_eei%1S@clx)rKBE{IgQ2OqfrKIRkjGry@4AF z@sZ4FB9TTytIVOUglqoK8#WrguNUbKJF;+QacG5R;XF8R>lf3Y1h&9?xd4e$k;-MP z(J=JvZ7~#%%}(~+{U3vu+{c?{sT;rU_y?89Y?fj%By?uJPzIrZh2J_N2Gt(A#p3@Z#}le25T43 zG^1SsX|1XL`iRv-3cOEUE(-G_Dj}~H_2Lcw+l2Q#HcT{-e;>o)W;EJ_T?1#=O^m$T zH&l6ZGCwrgjXyO1Vj6YZm0Dru7Ct5vj|i%dh-sE~^7ZmYDkZ8Sas#J2D4SElK|<@^ z$cl&tKTWqR*=?m~{Dcw+VbzA_sA?imsC_+WF?d`f|5VGd1uCT?J0ytB@fU(70td60 zlc*aJD?JaXuoQim4+ZGBfHHftcPW0yTF=MgHmJiuhjFRLP5wf35IQJqze^hHzHnhD zzgb$sspA!LuL*mBi>j6^7F-EOan?=V`1@YxKAO2rc*qU7RTSWh*7z1r-3CnDRXz(* zQyx`AwoT9~)8vm7+2Ts^G1Y%D8|7zjx6=<8jWT<^!1>1}=fCbraeZ&;eb*3w{?}IV z9#tL2a)^vLs{|LlPT-Ru6)eSh!?O2;`AyhqjIs1~jB zfNhu6dUg{dU=znaZg+RY6J~MDkH!-oQl1xO!Jj&z9z5ZCS^Tsyob{fKlTEvd&s&e( z(elVUi(*VO+QZPl%BW5j$N6H-rrI7t@hL&wM$N?JH!;~B8I(2gyiCbaL7E&{7PbDZ zp(VUS!D;R~;HPYsYPJg2b2L~wlqJdzuXO8#fS`eFm!F)NMec$yVWRYM=u?+eQT1$e zlxqAAlbM_oA(@h7-iZ)gM`=L@yP0!_>?Aw+sq&sFUBJp*PxeO|gct zoKkMBGMV?;jNyAU*kym%)iABHU4q3&+oM6deAlEE98I%VC*L3~AX}9gimoA*(5Eg& z;l~o9UD6>mc$&f0d3K2T96S=FjO@@Gchq{3rOq0YI1P)JUo@(0DQ_Wv>Pc}DxaGiS z=+z_Lz^#eU`mcsPtP{#4(P~9D94&WUp~iLzu2-qt=yLG;>2Ck`cy*2=dsXc9=pc7$_h4#?6~N8`@G^rodAsicpR3b`D4h3#Y0(&EKT}0#sjWfi zi>Jx^gsCw1Rrx@o@KqpMHD1yxQF{&dyY&2SS(=)B82@T6l)JWKId zZ(-@XPPaC8gTzr3yyE@?F^dOYcS;6h+QU#yK#w>r+*lrhTSpm+trHLUT$nZ_J6|23 z=@9367AXgV*L?|vNH9c|>R+HLa7IS%^T%%Y$b*@9-plA8F>`Jg-fQ@kW&D_lF=4TD zOA~G0{k_X#sANzylLZlZU@7V*XDj!MPkXPBGy`2(Y{-*_EDpN7IV~Y80nL&yRreDq!u~TIDIxeaP}Y$LS%{!ka+;w^PBi5&pe7$~jtsi!`OUz;`>Eg@*)pVUg|zPuj-`!H+>Rl) zv1iAcF`9$M&2~%t8Q%nudFR$2pjNzQ)z4eZK$A%*))Q1J5!1=f_Ra{tA#V05gCuYu zmXe>N>VmaOr1+_Eej=uLyhKd3Hr^ppFM| z5bg9-f9>>&NQ~to-4c-W9Fa4Aj8HTZ)IlPq=1tuC$qv6rVMc5M zRZBkpLV}IB$6vT3JvJe6LJ9Hw58}vF zf{#^fC%rHu2V*Qc^09|LqEq?2X_vglck>kZM(Ei1XCeG&m(MfWw#J$@8U_fHpUl9d zT%Y)-w@hHFp7xjJ)K+FGdMu8~9W{gAUP7^(pmq{5C}5yfqJ#k^?5y+zC9KtoRnb?& z{=Qg{%+nx?P&K(&FyK)v0wDxavp=YYw-K^<7Gh=)ZWwfeY8KvF3|^2Xhg!=^nNaE1 zE-CYF3Dqo_hHM}=XF$aGVEDmsxV_B1%zd+?1gg%%8WUu_*mi22*LS>4=xP0V)j}%6 z6&RUDzVStbqL6?x!x(&e_^^g1fxW8Ih{I5JjLBGO@^zt|5c`7H9_dw3rVQPSHFAR- z6c15iCN_HYghG&u(R`w+k~(jyaQGp=_;W=I6i?-FYeIKV-tG5kM9t(){+oj? z2DgrS8Pki%vJ<{9b~MUo7{QmgxyNsnC6DFjB&uOZ5?fP5b9m_4~*GgZHW}_KLACidG zDMUhW75}goNTWs8OwI)1`6m9Js81tSQb;m~_9T_N255}3z3=<D6wYAU&Yl%=#F0WqqBaIY^nb~mBuBB-n2V*4HnZc}GN z{X;h&m)CtWawh9QgY}YNh2*+N*UYU^T4i1E21ToS<#$eha7L|Fw)rfTwy8e@?cXBK zK~M%*GYttW^|B;Rt9lE0O$K=@UYjL#!Cf;A(lhEq!M)svU_-FK!-6hZe^jb}HF!fh z&J}t7oOEtAu~Yns{Hinuwws!Ogmt)`@tc>2V=qR!A-wBkP)iWoH0 zs`dwA->QZz8Q$Ot&&qEdG4u*{74oVdn- zsu5Ab0zv`&t69LpJ0YDr=u#$nAj9R=pvx{!sTZEP>wbiD%^T00Ceyh^T-=oCn~|YH zDD9`V0IHTs*GNYVJ~Sh0H>zm*8?DpF-q8L&%4A^H>txwf6|;~YdqW6Rs~JuX6AGXt zsV8DezHUTMM$;z$N?$>O4`>YvUjt~maB;HTN z9E&nQxkJH6D)+vw-BBOQO(>qsdIi)|83Das8f?A;o^TG0`D zFEWcPBpXRA3vN|njd4X}HYA1K3Tcv6`n5{;2QLZ92F03NAvlj|l6lQigS0oYSkxpn zka>_Ej?J7jnK_OmNAN;_@0`V7^ir5ikH=?y7pV2j<}em#@a;4+N;*OTLM%B%%m8<% zxQBc){T$F29R|LkCVs47FD%$rs8JuiSvly^OJysv0`>eh^?*lHXb)LI_VDz4&^=OS z`0VwlA3b;3>aaL~6s!G>*2y#S!I_rDwlr0$Sh{$XYE@nZmUa|Q#Og#Ry;m>)*V*zc zU*3aruF~6YE3q8*n&h9ye&?PgFXxLv0MHvTmvgn^6t_#6t=Jfz z?*U?-RD}>Z&xTgptxCNFWjnJY*F`;bDHh#z)4$oQhy%^fF@U%&fPT?gbqwP zG$L*MS+AqBbY!k?v$QsJl@Fe4oSsY_ot3~l7IkzMiXZJbPcae|z+suUXcwz39NQc$%T9QUeAVw|3!Emn;&= zV>h#58f#0Qc_90`bs=<_QPI$N&eW`i}3}GsC>aB=f3|4!jfenBSr= zTkK~FqC8O?Po%WWf+*~LL8yIZE2M3NVk<#yL6T@JIe}d8X|=%MG2pS;JApU=yz>SU z%4Of|lIN3sWMiNPJ)4{1&7690l-Kx!uOh=J9G=x=0t1YW)JGUE8A$tb!(TC+qXMN@ ze1|=kYxk`=_HQP@{Icx72B>CN3b>y}Y|ULI6qg99orozUlc{^8CLegSi#Yw%9qABD zMn&XG&m|z~4m?f#yW}b^IEPyA#e%x=de~{{k!DJeLPIkUoytucB4O&1?{Htm)#!cu zA;mvg0qpVAwuuM4bmSgVQ{#@C$~lv}w_;<2B?F15qAHXo?BVNu=mwHF}3JyqzPl++qE^Ds$gc6SUl@Bo4qOr2t_?X)etdAOA7R7 zk^ztFZs(QVuwexB4=ItKkr#n9zHagZpexgp`oP-IVkkPm1bD3qHX9*+0~<&wkw6J* zv}$ScMwkpYhICSWGxkk76pWAu(t98$n)cY^NQ9ltEZJ>*kE&pyHrS6MDHu7h>Cl0`Clp#)Yuqm$0je z?N8LM?(M#bk=`VX$COg#oC1-a%!?d! zIl(OrZdI@6?2oDmg-qb-XVh_mQ?QTFAK4mykBp=C@f)cY2tOfzW&t-Hb{=k!39f^( zq|j>->633#`E!nNZb(0YJGvmV<^Z`XC?R~A>*&5Aw!Z)`b+GZG-C(zN;Hur!GB|76~>8!e!A1$ zwBq?lbN0Wf1Z!z@>|L=@P{2FFcx)gPX#};Fh`|!-c1fmmDAcxBwOWDe7^GnCno>Ch z!*=**bl-n#@80usjI6`++njzx50{^Z8NwY=mSyo`IL{0;&y-~`I}#aG+XXZAXmLL1Oj=-%L93>))A7#<08}Yv&Uh!M?@BHF@XU2Qr*J zG)3|P*ug4|=%+N>C**LOgzG}nxHZZ`$i&8DO)o@6P_Penxx#upG-ti;`=ty%7)gWZE(C}cuX zZwQxLujby%@ZH>Y-n1mb;@Q_R0dxC)D%pzsYl?xruu(QbXk_KJ%8A8`OI4I(JY8R24R4B z*L;s&>T%e8$ud_Y^v1;;q;t=??tzJaccceZ9TRgUcO6T`HhXPf@xgX$WAt8M(cP$B zrk|gzM3#yEcOz$f*<_m@zWbG*Q~Q|(a9O-cx?(nGwGfI^1obfygUPs;-)SN5N8T21 zkG@aFL$!eUnPslVI)R~--61s$sO1*zn^Yoc;y(!(bV&%lKC1~Oux#z${Tc99EvTM@Q%T8i6nJawqnpFee!NjI-K1WS)^(L#);mreAOw@pi71- zS8~!Gs#m5EeC0CAY6IrToK^h3UT6YU;>Ih#qE@rGC=&J>NBF~=2}LGBttVpEcs6?` zaR$a0l07rANMy6OrXmt&6FdRYxP1_UMir6)CD)DbaMvX9R(UmgS5Cpz*m0_tTj#VWr5(^RHy3um#+F<7 z>-kzsu+Ti^F4?BprCzz@Ve;nOK88RjbPu{OSsR)D z_5D9sHUH+^`k|zgS~A_Eb(Ry~WyhYfEziroHjG|i_D=daKltjxub3>$;a^_7O&w(M z%+qORD_RK!Y&oAq5zG|8Vw!vf)kx+Nd)~|ZqERJ%s4!sNc!MxI9L)((*)>Za1Gx+k z#X!{%>ZlJTTA>ztKh-R4Rbtz?K)vrOFJm{&poFtpkpjD;8ci9eT~P{hrYLRu)a9|4 zu>*|p;2kJso+iJ`*Gn{-A{FYQS43`?A(CdHke0K_gJ7B%IPJy3gUwc z$V*YlZWt=oDp9o$QpPVRdgWCa`l=2S=$C6sbK6 z+Y(T5JK$-6xPK@Aju;5a;ZI>W#B)H^t6g&6b185FB4M<#w{?O0hq5x;C6}U>N;R3B z1R#{I5Vgw(pjbCezSFlh7}x1Hq#0AXRamg_8JRKVX&A7b<5!l$SK~*mQs+@P1swB( zP^R7@JSNj1Jp{&)?oK)mUV4%U6FrCT&?-Np&PtxT9N_9D*Szn~ycCt^scBOU&dl@N zYL_gZ`5=O|yAxlixpU6a8Ty>SkH$@^^1059jYfpwP?m94%tGAl83U5w z7i^7f7VL%FoPMH~m-`#bqzNXVU~j;IE}cT>cZ{rrEh^w*^eX2V$nAmx!AhzD3K=p3 zv&eEz1LwLk!%Q!ZsBO?NIx;Wb?YuL&8M%#?X?isIZCil;0JUfm+!wC&JfzB3T?KX( zh^C?pr%23`T{aBY9W&MYmn@)7^I^I$ox4CD(LDW zr)EwTIHZujfR@}qmN?4qKm4$mJn~n1kV6fIdv@$6BY%Ru%hnsKHr%od%Fx6)>fcAP zMJ#~dKB1X^h+_!DE#x-!b?#+xKXnk|>cIO3yiC1xWkilWaI1GSM2FQroN`eWQ()5-6ZEtPiivA!91jZY{m{7xwQAfA7Y zlLUmzIg>MeYJ8u%U=(docGz#Jv|e=xV)(|{=woH(gz{O5<6#K{vNpiPq4&M!ox|NR zXWjS#`JRXx-=T4==Qolkl-X{^dxka;b<^+{LOGjFZlA%L1Gf&Jy6B#Gbj?YxlU|+N zPAGN6#8J%?pB}PrM)hpSrg-dea$-6U*3K!BMFFkK7GWNBmD+8}Iv$n0W9Oin9<8=s zVTS6XjSEf&n!wXIe{D6D|GCMW*lz~4Np!%I9s@pue(L#~U!REN^`C z_X$@W;q+ziN`Ru_^gX81s&Gk&`NK8thh$AayLTyK3*LvrVF z#RYDSd)?##2tb_hS{;?dK^5ge*={$qB=KHP!VJQfI8+V|rCsP6>18?e?esUl(VqW~ z+we^Eh5CO|OId7Ea?MOiDxpXrs3an0NWBp2;&jtsk5Qh-#fb%3zmdVgSp!PZcmRD^l{3(JNaA4+tW`^-|g4$aeY=nIHc%#4Z7@@b1MXjlz81CdY@|IzNDHs z7lhK4e?H?g0Ew=~Q*r)rk{S<#4}L!7g90n9>NDzFKu_P@ zZ&})z7t?Fl8wH?fj3O3MUh!TM2p+DYR8a#ZZBN5$eAB~s25B}%_eI7DvPE!kL>VW@ z@z^uPv9Yyz0GNZo=B^!TNN*HxCH(k%mJ6QG$%Z4{@Mw!lm24YX|am(K{Ij{l0t64y8 zV6g*RW#+(cAru7!l}p56bpJZn0M-IXoi~PG4Mhy(Q~km&c|WXIp1Q0Hgsm>9*6osG zG}_oS*I&~?C2=q&yf#EHN%eOME28nK+cWqhW)GU7bNn>D<*<@c^~R<@e|P3A6I8O~ zKP#qox>BdjM&uBofcRA%5z_@heVm+2JQfGGMl^7`DD%&)>f^si{n7C+Z=HYZomKOz z=H7qj+<%>cvXpb29`ck=6TgBiiL9I+H@(3pU)&9`jpzMom8clo087uD$%C>^KDOTE zUlD|7QIqj2)2jy&uJ^)x`$8zG4*ZGyoQ~fevey2nGEd`s0QejE#_VJZSedy z%$0zu3D!*;vmo%s4)<*lwcO3#OQxL%x-*=0y5z`Azd5QfTyqg9sbd98L-qx42E9m> zUC=c1n`BqScSF+Tk0kMs5u-VxY=9Z4HYih(C%Ecvm$Zc514Lpuijn&FuwPL? zNhGRoAEyq9N~tdDsY|IMQ?X05+57Co3i-mUK)eyZ6*tBkIJMkrRRa%|0!KR9HUiAT zk8;+>=Xj+Dm^IJ%H(AGi^|}|N#Jgq(pA2aNdP754&&&lqZ`Y0A9o+)Vh&yFPoGSN+ z)B+R$#lQ?)yFYzF7BD$$l^gw?>6IAA!)}Z`A5NznF~SJ(>%R-@-ZdGf-=E3+E_Hy# zhUu!=#B`caG!fJZBIe`ym&K?>UBlZlyOFvWjMBoL)F1nyM7x5_c+>dtPx~hIax*1s zxK$wrev!A9b4nWT+OJp=u*`MPCEm3rw2Cr>8INNHyLj3+jmLUfKDjo~czl6Gn$8z+ELzjh67=0ZWB)I>)N5a=yCGJHyfYKFB2;m20cM(F!O2 zc1^x*=z7&Rn&4wfRGYe*d@RPwmRP~Lpj>s1$ZikaCX_L?JR{oM3$@u_{_&H^CM+%e zfU!%7mD5m+TYNa%PBOvGsx7h7qh*>j_0FL8TBe<)9CFm)ybW;}nqDN+i>)5k3ik z+^F|1gPAzz|FLS+2mhsm!vGmC?Ps(v-u&6B`)`?$Q9bQ1%c-p__JEI?jrCqa0c2Y{ zjSOZT!i$^-VhqRW$n5YTrfA(z^&JqERV%j%F)_1d&SzmDAnM*omU(x$Yn2-&?gCTM zBh8@-K$rzq*;wFTEUNN3<%3NZ)`nsiss381ywfyGi$!~&TyK=D1_L~dm@!^{I2zcX z_h787vB%B&&XPwFCiDp1=H^j5Sd5;NX6UIW6foKBA!4v334Jb0@&I{74{2hUz7@JaU}Z^DF# zC!SAEQiCqkCNo4N5sH-rl`!NNLYe_Y;?v~KK-ZJ_?$}Q1!{ccKAcmWc=}$(0VDFa9 zVCAY$Evu#4=O z9PXH5*ap~|IUHk!T}H4aX6#k{k0q-MYtcc&)6xp%&cGN}q)7LJl3g7TP_-&?`$j+d zOVn2lqOG8KRX?MBoe?O;c}f2=&5;s&_RfxUio{qx7&~*AVWc-wt1RTDYP2sCNsW(S+g3X%h(ASQCVa+2wEZUq^4bxD1S10 z(K7#SLaatUfz1-7Cx*hRo%|=^_jtPn5T891jCv(YJ#O-K9_gWrWos2Rp(FE=93QVy z0L?Q5IsV%`?|~6$#;tAm)ntwdG#?)AJ3=ja%~C6K%z(6xP$U!7Y9a>G=E!1h`@hFV zLoQsGJm)QI-lb|`vmZ#CkboVTH#^*&Jk+Lf zvSY-ZGQyUf(i;scn|W`1`)hyt4@-P7DeGXdA)ygcJ2mgJI2P1HOCW$Qu9a>IM08?uKNEkY(qfY?-PsqS~{9 z+opazZ9QkLV6otmpbdoWAIMtByTBcCURW+m=f+RyRzYE;Hxjwrkgk&Ln%<_~FuOKb zr^#;mtG3WX|A_7NC7 zfFxL}TtjqxaQ*97=a+pi>r46blD_+7RMa1x0wG3VvEKNDUv2)0<&;HZ>(O(-5IE|! zbO=A=fMn)&?oO(1GDf^>Lr;WYY-dl3a^dTz6TAH9@VwGR0!pJH#U9UOAL1wo8z_)Xs^+AyCyzG8^cYwko{G& zncB|cfc*)xk**^Y`v|I*!zQq%UZVF!Z6Um~S=zwaaiUYW(MQs9~h3mwdRbRq;YAQ6}w({R#Lrq754~4QtN_ui%U2ak=Y{L= zB4Rd2elp`)SR=XIwU4|nNpNlT>LYK?xGNuY+3Y*$a+yEq(jN(>#XdKrI!@mVt+J0S z3b-%Hvp4*~WGrokl(}hP@(uR2?OaiqU|GNQT2V^URB2(L9yj_hWnd`zQAZ;8TN^Mg zK@u9Ic8{4qY8)3P<8R+c+IJVdNy!=C`1bEi=3{>8p^4Nd%*+RigJ=n6L@Wb@;vqrZ zAI@7Vi5%cUF2UK*F3$eoq#>>otukps#RN=e!&Pq%SJME=)WDTK6IC%miJo)P4d<$M z`TfY+P;g!L1uqNND&u`Fi93Xyd>wgERs$NAn$+2~p(wvuN3|-AS-4MKP#DwrC0Iv> z9`i$LX}Dg}B}e35;O?KIg={)h2sVU){x68vgTN=Qu673P4$&k;FO6Ojm=xOYH7L`d zq9@RG)RXao!+ym93*fhoPJw_j>|+g|GyY;&jHUxsVt56Zz|!0_Irr1D4|A5kV*=mY zYJNA>@VNosWwW`fnNUDT_c#&L?3K(LR&+&@eHdme3e+l3aS!r_^`W)OGqaB?jMC6` zvo?i8P1t5{49egg`KsE`W+{?~*C;B4Wn2i+ELMLEx=WaC*(GmP>L8wr)rbi0r!E^L zx56;mt0QYeyUFsv<_Q?(1xcG?Q8zTd7!0bhVFSPo#B`lDW|(4V@jm+6mQ+jH9vZpk zGrZd~Qh1NWOE~>uUGiHYr+H_1O9Bs5cPS)6$`mXKY*QzR;LYbHajtvp54uC%8{g|7 z>9lSC8oM5@ZCyLmihdrkvLVb~&vM!HmH%l;K*-vBOBzZ+9hu76r-J-S|2ko@-zBgQ znnswT@X)9ogtLzpBJ*!Ba2A!_aaB8GP|LhZOWez9VqrD`ja055xCPtl0aF8zjK zM%T=z0l?7@2@Of1+k!R)Rf-pLj;T7xlN0s_-wJEsob@^vROnXY-tTcUJdxK(VjDd_ zbygoZDOcsmQ&)h@J@(N8H34mEWcF$CY6(i_T?e)V9cOeg$Pi~Q^I;ew>E+FTdQo4S*K z07wX-h>$Gw*ZMa}>R}4$l~?#}m*sF%COo7v1LFux!!1x<79%&jR*9{!gTvf|kUo_b z-o$SXZIUE$j>x(rw5n6{uguML7-#mv$QrcF&N+xPdKi8B!xKMrH^Jy5&DsB^60E_9 z#nz_K3`QFWMH)e^Mef{NUc~||0V(sY^gAN%gA^OgcDl|5l)xqe7MzZ5!N)6xh!Hei z(bs5RriVt{x$i!)i~!RFRrO>&P$2D9Rg)!>G;WPj6BnrW&4LdVkw{&$S2dPv3V1q9 z9`QQ^h{<|~#pZ;jUg)DmSi6oD>p;VuL#^c@x3vbb zePMEwzti0lGh8tMx7mHp{!sA!FYqiiQ&}4#p;1A_Og7Y3hRmg(*<@HFASIDYQwpgR@dlHN{ArUMyN03QfA*HT{xK|BdboH%OLBF~PHtECmVX1a~LQ zG7J!4=0#Y17_=8Vy_FQbGVBa}uFLCmnzeB~>;c>M&QusNZj+Qfb2XcOjf2rWsEClPgbVi zYt1(=J+P$yW345j;Up{!*Q$od5bp6bb<>)Fmy!xZ=Ygoe_GmsqltiF?HnqQ`6eCG(Y>@^FU!K5)FGe4bBtRtvo zV=d2+;8?q4*Q6C3jWKZ64U>o_8<>eP^>!>x3nQ>t8&xNrqX)~6=dQ`N44S=|Hp$`u z2u*}9&$Enk&HKWf+R*Iq72~r25Kn_@e38KVj9ZTAe&m>ysxvnuPQrloZt0Qn!zSvd zj<}>O#te;1+AIYkdthr?`{kT@^}Gt1rj2uEwmw=nY2k%AXJ!wn!{w`P&d{XE+d!^v zA&Szi^gIOGV#A4%3n4LpS{?2zi^IseSeuiqL%`v)1=lCuecczgTRcl5&v6<_UF0cI zB`k9r$@G9rqI%eu%LmURQ&I*NWZii9dq_IWStD#Y6pTmQvI8f5rneea_63Oj`)2Sh z_v~2fq0;Qn#rqgMwnZcY_dwk&jZU;3q%4h6$l1PWTu{DfhaKIe(y_t?3@bBeI0*+< zx@827H%`r$b1fGyFS;Kr&I+Je&>RgZ;#LM^C^FPXA(QMJX^24OrB*ebzvjK2o6Ttg zRVIvnw95xPR&nD4^`t)VjCu?J#0)sY(}Sq!8KH#D+S3H zBh~{IgaQVll!!s5=2KL8NQHXQqD6`F#39L&EKzniijY7h!_;Gv3cu6dJya7UW`KH4 zx2ZLh7;vz5vj`G6sZ~{XJ%JK z>c~Tp>Mtkav=&^8(PaPAnPRtvTY~@8w*^S8n01)0bY0` zo}FO$YS6y`U&;N^W0A=c zZ=cZ2?*w|FEbkK$nyT3+#jp+%YA`vnn!KvgR0w0CD2zD{7!D6Mw+}jYd=bN243ECB zVVnsds=Qr)p`N&~midQ*5gsw*s|m#lf?7_*6v{Qo#C{PrXi~VzyuUMjtrLR*-`l9o+tiPj5UB4=0No+>>xWGX*PiB*`zV@z`2kbH}1(VK|#V~HnJxD z_CqS|b2G10m>HP@LXk^QStvfLN9QH5ZgTCA@MQc{U^&NK-7a~YXAg-N&?mnm9q_;^ zFHIJCUf94nNmVNQd9f40cXujQ?8OOU#+g{*|7hO3mWhn4jik_|CT4T;Rr{$1*t%CD;{5??#;YHrNL!3IKR)@vLakhG|Gfy&_WQf|9N{1#H7JH>MlgtTalBhf|nfg?? z61dt7u$Ipf>4B#W|CG#Kt?1>in58Ka>&O8Q<4Me6!x<({aAA(KVP+$!M$RwBV`m}? z`kwM$)iq0vKpH8&GETdqjiaeiejJkO-}U{DA1(P>F7er1?fVbr=Bc1Y5^8RB2hRE)3J+v$se=%IbVtwu5$Qfxi?=W>n`YBWhwW&+p zQ&mZ!x}Y4Wz)K0(;eBR8lf>E6$I$38gUFanAI9H{cJZoSHu3)dE>=*Cdlix$Ld}l9tswm(@$Z~K_P=0A)c6g#ZU$rL?H)xz2 z>KGno#`pl74BzX_pErJc^kd5m0Vei?#rUAf9jJs80WO2pM?ME365b?jBo`#kKET;GyUKmRpv!_!$rDtYvP@pD zT9BhU$V1x0#86}#diBWKH!#Dd-tM2TeDUd*6tC4(|7iK0bC%mUtc|O+3C+gLCC)vMWYKc318BqV_RcyMioNQp6opX}By1m1!+ zStnT=1eqkALrexvj49g>ALA71xIwYr13i4c8gXNsp9v~Y6vq=OEsIG9Ktp3hLdiBl z0V|*_L`;)xU0@cO6*5%Mi+N?e+*(cxc@OqxkcbDlck)$T^2hQ)>NqD$pt`=-ec zaf(H?p>b}DCzMiMlm*MRtS$c3-@u9|vCZ21wjN{XU z#xcXSScH%CJk{@kuWU*ra^C6u>d8X4o1iOTB#?2&!=`y6%*KGG6DOQzGzRQ7+;eU! z|25NOO@48DRw-4+Vrz2B%$hV1idur&OT?f`VJ~-MwAORk%-ETIQCd%Auj!K4z*%6l zZ{?)PSGm{ldgc3OYLz`vkHrO$5t}B*+-{h9l0-cu-qWUp9+cQn7ucn!h+OHp-*>TJ z57`eq;d*%^wOWxK4t2<=+fyD?9)x8d4!;J>z-{}Y%u&2E19xfko8%-DhQ6`&m%pJl zu^2;p%rI0!C}6|85chrZRJHsZ4?Vvh0Nd0VY4&DiQ7g<8o*om5M+DUeB0HdbqnE@9FbNWBvTEF$ zr6rtL0ji&$r8b6dF&1>05+IwUX>u*n^Kmx^nW8T#k+(uNh9`qU6BbFfORyif)C#$| z$AJa%%xoxWMs=Mu(2Us_zMOYqPCdECA4A9cAamcS!~-qh-ZXi=WU-)FRO4%a?Td{Q zDOHU5TvX{o28v$*km<)(@-p^T~Yyo#O>T&0*&6c zpK6!%%2J?Y023*4IgrU&ExsjBqFAf(;H97PXmbIYF+P8_R zma-W%dK0+fEaRMmwdz*!-XL@zvd3%2!nj+SW1F75s&x^(&U4{1;~$7JC{r;r;u5*O)UYFyQDu|zCPI&54#OVhrl=|TAw?I4A`%d6#uU)gtXC;ou1*!o`3Y}n7j6h%j zp=X9<4u&7J zmmtzn#Ee9b^<(=mMUNOOlhGUP;yj_Z2}P|xuUbfDd~SkuyUZ|GL@0nvFpr2i6uE5P zM~m{l`B}tOUroNM%4blP5`{9^iR1ezs1<@pTfSL?9n(IG>6EtwGv3A@TR&hkL5GESgH*o;lFnSBs>)c`wwJ!H|M!uck1@|c_R z6RMWQlhQ>q196g294DwFM9jf&UKZ!ixhL+DRTBxpyEv`thac$Y-T05K|8_>*4K;=a z@{yt`xQKI@I#1!&VIlV>q%BkgZ2>0@Ux%WPX=Az-)eE@rL77Gx(h*6iozmf{5t?-s5GwQ>yz@v`Q?5!Hq>MPpRZ& zx*q{i#08CHRY;%wP%!?zpiR9?QWD-Is}(+y#78gl>=5eteKT?!`V#}9z(8pt#BEy3t+e4W98XUyq4>mD21!I&2&Gq z6@<(1zKq~RDUwIo21@5+@C9Vr1j>0!i!v-ZkZGufu~0)tJ_&meromK#_^BDe8dUS^ zggiE2AIAXdo(Lx{OkVi+Fnp#j-0!RtjNWVZx&qa|`=97PnhZ|%!$w!?4vS~7Rc8Cq zgM{KUg6aqPTHiy#ZR(ZpTo!+HO@3Yi#S&aWm@fRhu$gQj^|DVq z66MQW|G0ksgRdU>!G(8r&%HCZ_-kk1-}R-{^X|;O{*A)-GJm*wUh;RJelhObXXe$< z>;K-__XqxS!xuB&I}g{l&QJKp;x8A?$F>*d^?&UTMekM3{llL+h1u>Sf6r4smn0V{ z+~V#b|9&U%R!9>ce~~uzc5Xe8o?_5)_*|CpEW09IoX&p zw4Y`ZjI(Sf_R9z*<4y;C>BB&inQENBwwlUk<`=OzjJV&-FRCOI0J3dF45ru|ll98= zB*LbGJn2;vdeZ9v7rDcsmIW(W;(acO7uCXaZ6Hwy6G+D~*#;AF)_{hU<|;^pU@v)8 zAEc$%lXx06?67b_O`h-uZDpmATvYU@P}7_rA;mRf1Un|_e`(}j-UR0>aeve zPZm4PU1s1oODIkgR1*{T`o8d@ODQxb0KbOYnG5RuX<;PQ1@26Kz9*O+#Y6k-Y%2 z-rMM%9p5cGIn`u;Ttuo}Q~`^<>jP%CW+$NlBy1;Qs-n*LWcXzIKMi<()T%zCMhKvF zc>t5teF@4HO+mc<+l6Awls(ou#WcBus0~pp!oL&ThpG|sRYhYa z%sg`RY=Xqd8^-9E5hTVPujXftH<^+fzwY=4mFP<8%-+}xLID_AN5rIilyD6*x>bK% zH-0TuPyMM4-qk{KF>E2C(Z(xvyt_~VJ=*Uc{d2Qv2%f)!^c{`z6-GnA-Z}fDlnIXo zCT#q%V*FpIRm@gL?ERNf#byxSL?|-wmyM|d8W1E?`sjhIO^vd=TgYRo+hKX)y#ec| zmdPGcha+x>8PBbsI#xJmi1IN@>HE`P`pBP};PKPFo8F}^u-FUfHyiLz2t_wRbzzAB z#J*%{@)ltpcYvD-F@6wYPm?2qU8SF62 z;~a&>B${NZRf#=y$p?Y>Rb#^-kacV1(S^7^W0#L;?av1mXVPh_i@YkM~+bDu{csgLz!3O z9zXR~$iRQDgyrxNa)UxgR>CHV2Di+sd`|hmcKXHO^;0qROOqXrx>>qudsL6TO2=Ru z#coKwh=Nf)bM%AEzmO)2)c0Ng<5V1rEfN&gjv$sNcI`bqW%kf6)jdJ1tE;3dmAi5#V2BI^Zp;D~Lixnx4-&&V)vdFU!MlIuCUIG;dV zza;WV&_SR5_NKjzcs@)IZ5(H&-8$T!k<}pHZ;x)3o2HWWDZ%HdbZa4g7JC(yX46X{ zp@15MY{-j|U!1gf5>Q~Ou7nSzU2Me)Y2X@C=6Hfj(aC+}196oi4(8@zx_Cuoz6wj4 zpjdXZH@c;|NzRmnwh=tfjfDA;!yuvSfF3-%e|#cmq6sMaIboNnjn;s|;-zr48Bn$o z3Mh3eAY#&BKc~u2UZ~3#qU- z5MPA>Fa&&T1)Mw{tdgTM2dsMyQkaK)dzQZTiZ%-gJ!!w=aVs)RdiGSfnL=oq* zq|Ilk6t=F(-tY!L{PI^iMGgYb+-NcSL}#Fk=LZ(v{F-G(-fLC-M3X*PEP%9^4)>w} z)IP)M3iT#-xOYRU_%4jttegr}md-0{dhww%1Hg;-JL?SnJf%AMgA0~QX-u}3S?uc2 z2)=d6|4~aGn^6hjk}i22@03p;WZY`1rhPK=!kjja4v2f9)>21y%aYy33_HwlgO7s!-a^aGpckhW7W=t0TR!F9NSC=@aaRtaXjAq=J;_c# z13xLWYRY-Ykv$G{nwP}qy_=+mxS2Ml>yH8on1WywzS5>H=%<&UO5RMK$+4Z5@tq^o zGG>k~i-D7G=Gdkaigg5)jMN_m-pKxiG~cM9SI73{5^c-5puQ`0y^6;#3*@YGA%C8h4kJYxh-P)K%K+qr$!le6Qx?N;L1s??5e zVSWq!#Ij|MYt%xM{iyIeDWkesyr&m$R!Q`bP~0b|dqm8-@g45S-ChxS%L_Pkkwd{i z-jQO2)wrBZGO;lviLMH-K_>oM&&OiTMF{Bc52_`L#81KiC~<;VpqW}KZG!MbGu&$+ z|A7vw2FrN5m65l>B;>;dw^e4Qv?iGYr|JZL5Vg@b`=n+0n$s=YJ~=W z(>`jcR7c|9cwu$aYM>*j<$pd(h{mO6{`iw0n~)cnbG46ZvliK9@w&LrY<{{yDD(u? zNyJ=G^g^~kn!FF{u4~CkKM0kLUwACa06g)Pe)p#gx~x(hlH^a(Y~WM})sv~B>e?L@Vd5;SP9Yku|>7r%pSJ+%s!8N(WXgAia_EHJnge5+|}YF_(r4 zaV;-F5GTkIrHw~`t>x*+4sjbdPLS?y05|A_=NPS@pFMd7D0p_ySRBx!z)H&YIN^6@ z{>(C>%UWd6L~pV7{p`dF5c)*gDwwkNsd}hle=HoU6<#CP_-q{G^cDut7@OdV33rU3 z@y3awZSPvf>{ts8nkZbqM{_8`V<_hdt$!PEC+IzsIHuHoD16nCTGd9mUV2lGsmuMhT@QG2#cT_(F7)FkYbtnrTz zTpb?onnG>%JpehkS)u|#op70J3XpLe#oBnji7z- zG2&>Heg&)D(8K31-&;K1Qj?F#1^|oK=QIj_#`wNgnde!?x#o>V=)xS$74IxEHX2K# zk4a*qHI37cDDldGhX&8#(g>4b>ZlGOO37se?{L2w-8dcgk9||zS-kQ`T>jQ7n<~gN3utM3^Zb1cZ%7^txd)hRmY;Us_$s6ZS|HlH@zp+gy5Amm zbS?BXD#%@!18PGXL(BaRK&-WfE?!vfXWbGfOU=q-C_MeZYM5YWR)W(Gw*?u2BeGwa zNlI-PI44bjQ%5n?6xoj*KT)!eNr5m4w!Rxf_J~j;2ScNA`~qP)kT7=&DqKc5e;ww$ ze^r8I*%-Aw`MTj=_m469bq4S<@#fa?Jrv|h9i(@@zG za91$}F=cvrD?{iDfK3rh>>%{re_i;OcZ?8EpGALbUxzshN4y0;G5=sRq{}> z#l9SfqVboh2D=4>N+CZy1y3&vaIO%rtc}X*^=_=Si+lasL4StPe zgu@EJHXOc5H8FUx6cbI6C`!E>;(O=FCqHR@wN+31bT)Lxooc;>vA}*!(GpGW@i+hX zj(WVb37c4q4W}f{2%#i+BHP?eUJhjUdIgw{hOPJrXY|hu!@~kx7}u)F&F^HN61@Ja z&&LP++2F#4Erpr%K$T;W^IFea1cEbOgQO408k)G<+&&Sb`UL8gw}o~H8i1KKQ?<)E zT+y)9Tt;}ZpSZc_EJw&=r^eh+Rz1JvJ3CMPeIX?KwW~lncsOvKeZ4A4ygoGDch~$Z zCd}_{@F$W3s&#(Z^ftG)&^wF!=({i8VRCrKXNaE2RuM9C<$`J|aef=NA!g$IG2(J4 z0qQja&@xoJbcS6K%OiGHT}Pfh*us#PPk!0! zQjk&`s|rvkT+<0flqf;~2U=xu(15Y;edc1b@R_q)G@d<4*1Ch$ess2YWcIeHEHO50 zG0ZJ7+|A5Ko@r3PSS-TA-2Q+e$Gv2a`=NQeJfYkPGtG&LR;YADNt;h)n;DGzYmSq1 z{_*02^YT}?ZoqMc&eC~_9Nfoixzjs~hfthKJAL8SpG{Zx%mmcN!c>2n5Z zpf$ZvKm8yeN^zLPfClUcXY@j|yh5ZY_wI6i=Yz(F%7EcBcnGurvaJj2^9Xl9EGVjSq3X82JlpbIUf`NdH=nRF#;flJkXsV#z50O zcCn>-W;)e<7hh!yGo3EK4xkFX0j0jhD{m-zA> z6RmC*ycvsI-qM}@`S@~fHt~NOHacbsy>T=3KIVuuwZ6bnr>J#Vo#6q*bz2o{gq!Df z2s#Czfi>v38ml%zA#3%X1(5g7Rt+iIV1v~zg#``eh;J$y`0a{3$&?Eoo}4QyLFLJT zr+B70)Y*P~3h~KR$cD{`nM7g5oO+*A;t1Da&I!K{UmoV5tmiPN#P^oVwV=H`E8Br* z&Vj|V|ICrBdxQCfr&0F3e{eD`Ja1kmt4aKGv+$IdEIb($0|Fu`l={%hbyU+z_$KLEsV$P5`7_UM>4 zK$r|%a@;xVxeV3;r6gv3^o@ro;Q&MpSj2Y2WytATAULPu8Jc**r_E;iP3C8sUzlH7 z{_~wWZ&%~e@{?1u&XVMb=n`z$<@ms40m-Kr&_T+g)O~amluh1qJSb`lP3K}gKgyHs z@NE_p`|R{dmuQs_Ufo5#x1D+~o4WAB^i_Ea+P!YcazNHP!ne`qpm(hYokK8GcOdYXnPqPuW_M`y~TQqX}3lDF*nq%P94EpA7L4_dE10(jbjh zn7(QaJum8#A7^?3j?%YYd9d=2m;d3ws|LZle2-|SeWY&{Wq7=g zJ^~Wd7`TOgBpB4HD4HsAMiGQRt%)M@$obYgn|^6$L{4buyBe~Fod;{fo8UAPOvO`7 z3`I79QX;VRBzO)vo`T}YROcZHY;n_rw}wo^ZuVI-XdzmjHH_7-FZ}cQGmDIv`14MV9QHMyI-T2x7>spa|b=x=XLg ztcd{)OeBHA_Awx_1y24E&H+_XkXDJ+D_E<)38J*0%CaEQm@hW8p^b1-DJvQe3&Akf zLOkWzdYCY8IwnX0%@)7$+a{F}I>*%SZzbi=jX7yH0a^pa)KcUirQYP!sLG{Vp_-`? zc+^{DJu>K4Tf=Sesdqo<4XGLI>_A#Dp!ovAuXLbh8R4V>d3wI^6zI*%HgUuJYDqN% zSd(2yS7Jfli|z&{$8U zXRX`7uMsu)X6U=?Z}Qdvc{1c1_qh%^h6%Bo+KNVKqIsJT zkj&dEKGEdyGuEPvJGKIhK1_K1t?|SPU#ggXh6zfxQcN60HdE>vI))#^z3KS5e2sgn zEFXl)!udzV54eLfRH{FAYAl@lv5(Biv3aNFOzHOmUWr0s^ca(;8+ProtZy5{ z*5`*5^EpKZC^Zra?ptvi0G5V>UiJJT zS)&+RlKUhHP+eCiP2yr_z%i-eM!hT42P{IDMUhMDz=ggc&k=73)LfJ7etfTfKvyJ= z>Lii&qi=bSzV{N22_i`sL`VTqxwun`x8e;^&I3T3csDpqn4_w2c@M-296{RXOGvOQ z`Ekmd0Y>;`cxG)Q#q69P+x=pZk4&s>4aFRw$On|VP#Wv7kr@7N4aM4(4ibjN-4I6x zS+yle+>0yjh3o-Jht{Q!_-&!P>2>^Ahe)q9?-*{g?2%Uwjn{Yh4k!?<^-vXal*SRZ zkiy`4pIx5&{16&D=N{%a_@0K2Su5FR3s=A~_V-zb*0Il>?|l0o$5%A4DSokG%Vwt5 zp#Ycw@K2pI#qEx&Rf%Q87z^ASk^<@hR+?^$C&yx3SUjS2Zv6NM;U}avPdYi=A zaO|(x#8jkH3@{`lqY8bs!&OOq$W_pwxwvBgiqnn<<*lJ0!U=lw_el$VHng03lpa>> zSg_qSU2>S0F6p(_*Zt{}VDY4nu>nsX(K=Vz`DEvAhbFlgA=3J@eQ%K+&&}FXX#%J` ziUA7dOw3sxdIbnomiM?d0c=F8RD_hh_nm!*mMHb5lGc?C>u1n{L0f z2fU>W$9}~r*Y$QMNbdCID**(PLDplCSqDTDj&ktczrQtpJ7u%9mS*;OSlPeTzfpqy zs!$dAnJ6D9$I#_*gZFd{_2I=dbV}#GG|GxZJ z5;akL#&&;y0@YGuf{r^V2A0jOl=`Y-ctO7VeZ}goP{Tp2Yg_15MLBn?7|LNn^W8Ub z?<2s=4DWhRwRFW~@WDD#pLXwb!MIe(2gAkn5mz zzRNpGp9s7cFwDWTVZv#o+?nu$wsf}jE8K)5OwUR4)?uPFM3?kZ)*Hsv232QGh-U-wB7!vTRiw(?O#ygTB=`Y)L15A3UBy z=se9A#|Jkhk#zm9K>!d$+>6agIub!{@hA4z60Z-#by! zC)l{OPcSTx2Q`6eP#n?ZF4_+@Hjz&$w_*h|3%n_ox0uB!E11oC<>&A2Uu=Zgs_S{t zB;ST%)?@;+{S*VilRcCg;~#tEgOcsi?tl)#$Ix%uA;5FDWIDxX!t->O;Fhc{w3~m# z`Ge)3`3`fA21hy!b9RvQIVV>gA!|bWL;9DedB=N|x+FRrn~!Zq4<)(M&LG2;GCIdo ztNajX9jv%+Sr;~_v9PKQbFRw{icW_Nb9RL^EIlQLu0>v=N2J# z6z_mHT*vQ*UnX+-I+ z1U-6bNHokDTAUP$V|@vekQ1$xU%CAcyT-Fs*{m03hJOlcfHygf*7+iNS-Uju1=QU{ z?)P>nrV2C27CzPmpK~jHL9aeQ?l~F+Kefv9q|-ZT;aWbjZlegB6;0Jx8!~mXHFh-X z9kVi9v)vz#eD77G*(zGVy+i8QWfN?8Gj_#fg>0sn6BLB7ut19&B|GPzsQAzxqJJpw ziFwEqbF%4L(kd$<_*1Kl^$h3UllLg}r4Us?HFSYbHc*C)o*v;Ocp}|=ul%b6rDN%EqAlnz%#9n5x#DOlyAi1e<4TMLCS`sLOK&jouimD6lq2c;x%)NjMN^7 zf!gCZ*$a(y5q8MyzelpoEt@{bjSJaLUl(PlvKB-N3+Q#rPy4UoueW>myfoSsRj^Na z{q}^{zDCDAN;iQo zkLTArwuKrPu#f_24fmJ|-w_ETSlK{Z1ypeL-*7r#U zN%n3Qot52m%yO?Kmlk({YhEEwoquBf1(%ORtmcf(7j*o+Z0CFY^V9X3)k;S7&Hv9g z{=RZp*}`a^QTzt``sK9e0{>N+)t<21v{H)*zJ6>ye;&a zOH@cNIVdQUPP+lIXyPYuGORGSPaT4t;rNA{E^{h9xHg>pHq(QP>RDPPa^L}V~Z0@u*!B!YuwKYtFxOk zE+x7dA-qZR)`#{w?#95% zz*c1(KU;O48z$^^I0T}85vo+_S>TpOf&(y1x%DGq9JT^|Zj zuw2kR-0gG;GNI{#*aLvIcgO&)+V714v`B}U*uqQ>3{T9R)lZlQ!vw+p*)E?SxID`U z5LIU7AIT%m1h?6+IRT7}F(;cTW+O#5Q0fZsatsaCAKJ5wiCS?+o{14Fz89OJ%wy!e zANel*``XAPEo}19D(mE6S2_fkDZ;?}%*z_V^gRPu$l&2QKm}}fHkY1#zfE zjsbQonL`%LG;w-hNRFt_~b{OFpYEDq^qZUey3@MiB`l_3m4h!dc#=Qzw%ggK+yE|Q9G$NyF(H}RE z0(Q%b4VMCZXo8~y6a$32dnt8<-E~E!k4`e=ltrz#(<*m6eV{x_cMHyu1d?u#oV$6> z<=oMGuuN$^l-R~_*ZF<%WD*s#YP}s)X~I*;LYEHr*Ur(pXqBVS^}3z1BB?tYtUbBV zJaNVhqS>F$dUs2#j}bkK(*71ta_q=qlaGBL#X#3+DW(1hsxMBtHVYbgWgzTyAgE~R z?U!0ZTYh@?o4Oxg`Rc`$XMvESKX{!2ul12@AhO&nJ4&a!+;vU%&Vsca?{5oT$4yk= zU0P+B@S$%72y&lxe&~B?RZFOqV^Y@p#MytUPkVyZx7&8VoZBPg-|{d5C;VNvXp&~b z3sSWSqDv?SLTP!FI@$ZL2Spp~%BY+l_pkh81~u}-f$x<4Y|Sf?U{o?l4-|I~zrOAV z_5YF$PRZ@XI_|InxtOaRlD*$Q2u;xVWG6oN$6a7e{y6;l@s*Kwx%7b5qpMbf^3!lR ziEHej{J!5?Z;$UjdQN)M%=9WHcy?9Qau z!xBl8z{KPp`ORjr%+F{|Zhg1Aoa8+>PRvmg+fzj``zWB>Q>T5e;Xf|9d=`@vlY_~GPkq{* zF=r}v#fCS`X6y<|jU+&mkS14L!b=2&&3gAkyf&rg;}`Gn_sMrLNX`=X!vD7`vu1gI z<)qVcqh0Y@x2b@9V8dR@ITK6KL@^*4(LkvU;zBj_UUy_(+2+Ud4<9EeKya9vF`hHO53GrV*DQp?&Fj^25O{jyecBT&qkC$qCF|R1NA(#d9yN&?%ZEjo@ST1eH2u zg5-6+SDB*5r;iwlZOV9e{6;t}GSpH)BCBlLW}DRZe6!~wZN%Te_dSo3wG$P#*f7d~ z2YyVlU<<{>QY4yEW1(dH+&%8t-c!32BmV7DOcz+wb$OO0FkZLJP+2&b6>l?dsD5_J zE@OPVhMBI2Jk@ERo4iz)y16j0jx7s8>Vh?LV576P}0Rz~O7dh30p1Y1TCzw=Han!M7Y-hpIgq`7V8wZBrU)+ zSI5s?gdJFg^py~u60dyfjcG(ga-%Q$MFZfj;9b*iKv=`qfaIqD&pMY_!Iu&(bh`a( z!L4t2ll|~0lfA(fepUR z@=J3KdL?P%X8AgA@VJ}7OkS0M_D!@iN)dpBJ=}}M{w>a0)n-N$=9e$|1e#fI^4iFt zwALd^cHJMuArH&b12H8YCj3C6{}tkcq3M+_0}tk=qyEGVZdX7$w~(9Vk?ab>vU=u{ z^uRQ_dVWPPdPJSDN`lTLkV8xod}EC1^c=DR->ia9-uz~W5qwFx@^rFqBCp1VOCVZJ z%+?W#IYf~<6f(gX!?Kx(h2ap^!ER`cMw2h>l+<|{UM9}bpEyPDb8Yt5s*r|8f4}P$ zXl&i;T;PM1;2OUFUZm^0n9sjotEH^3$SBUb3_T zx}qC>kI8H3>j7(Ehny@emgC+xn~oss{Qmsj`c<*8DUan%-*>?>$eUf7rtLeh>+H6+ z@jEG-^_9(S@f`9TFtgM}gD@84-1oTebKT~a>Wor+LryjR@xg^sd_9u|-z?{Q+U9rs zz^qK%tPR5rHRI92CU%6W8JA&TZP6+#+$+QvteRsQKTD>sZ52x^x*x8V7@b)8-(GDcv38`$yImPRSo2sL<+@3Y2Pf1rs{d@k22 zKbL>tp(%7ta_bOe(^p{Adzaqk20cVr0&|$)-_Y#rR6FO0Pq!kG-#f1k8W-(b{bRVX zklMzE0INEolQ3{I;)JpT%^P!0y|>T^8je_1NwV1`A#FI20&$Zu>7l(818glNl)6QR zp%2ut!VpRv9|J4VygI0J=%L@*?xs_wdPMWGR9Hw}$op_~ksi?sB!NZF$HSc&U_2tc;Mtf2?^2R=rZo`XE zjfsUQr5NzB^C`8?rHt+am-!P(rey1!gWk0sHKL}F*3cv_o~!l9l!OUuL|8M2ukm>F zzOK+rNg16fIp7tuJQ;ZX5*1Kzz|-<_r>F7AauAqZd8WnbX0-%S_wVc*ri_qz(ro1?TZE<>l#~2zX+Cm?v{>a&HGq=?zWp{R(@x=(aQ*k z-;^AVC7JAe65HLik$MwcS5QnT1si)cP+-ej=tSUCi4j-O7nyC2P4v0hc+?V94snql zQLS%(aI^o0d7VT{qC#?1w_hp@EcP7WmVg++QMy3VC9mZ5L5Iv~QM0UvHjGg&FZQ(l zvSbPP&6}IYSFDM9{aE4WKTwQM{IQ9s*{~;TCMCEXIBIJN_?_O6`O6pHmIAM_} zX`=? zLatdd@W%Ke7dAk$Ve@0ABxbi$)v`TaAAkP}n2{V_e{fqU6hM-F-?{L$7$7&s;{nk_ zN&74B{Vj(3$I+WwLs6*WthKxB#!u4ZQ8Ioc>%RV4W&iB)L{Dr`V!O*P#9_udUt?&Y z5*1zaX6AsZ!KYJk6pErWNHKR%)ko4iW4K1;`>k^t`0E|D;Hv+P zxu}DrRSrrg+X67*M{e%qHR75&x4%jDsrO~%QVtps}i8-`Yh?e0~GlHh@)JtN-Bc8Xl(xIq#J;^r4|awq1F-GlCZrM-Oy`+Nc&L~ z1SEy+()}-L`XyD%Abl*#QXNvo@ylGQmH|0ojsFl23)By&;Kr~yn#+=*IYV^Tih32+ zG#s02kZ~K~d^CSh`j*9R>l4s1oR}Pd#;;2gy*BfP7M?!$Uc=2VyPEi5aHs1I-vI?S zv|_j|#j}KmyvnGV)*(ptzC)+FU^`Km@T{!HXCs68@IE^I8&^S5?d)_(2|fP&Cq46z zN0^=iR?#5pm6m^r`MS~2ynQzL`{dAbV`#3M7@9K_(@c>Qm%^7ej6Z<-)n#MOta~e55MEKxPgYCY@WC3LZg= zgaz4$<&fM|^@IIuSpe@~>B={03{GTlEIDT(WZzt{Y=wgnRe#o;`yGk2;dDTb2~M_C zOd>_LARR6Wmt~8u6Md<|T4A~vCN@E@M->mdptZvNL6ADMbgC@;lINUE8s161lA3>83Du885KAT2QlI5g~|BhLACdQ1$YEo=BbanpK zsE@wvdBqml1!x6IRA3vlzGuF{r$u%Ex_h+BERYMk03|GNp+?pzedvrm#D*4Va?p-kG%@iX zQ4BD`9ih}w-@XdDlhs@8BYk!9et9@|Pe`iI*=6VH9?>SK{8(KkukblaufFJa$ydj{ z09vK`vu&XXo~!RJsOL7(Ep&#*WfA_|OTzi9Yy6{K4L7dN<6d0x5c+SToU`FAuu@={ z6Q$S$YNl<<%c4%fz)Kk(l2UK-f1&wuC}<#lVx?U&+#9$*bH!%WAzWteUY!D80FO-o4( z{Av{+?tf(}86(>@jpi~&YlKa7oxD!o>w=6xr%1D`gm;G9$OFAQ$vssBe}^hoe4f9X zTL$+$biPk!$gr>n$aLEg!}#^T9|{y1tw~hMwf`cU?MQ*i*D#f0k|~lzsUQ5qRY@4M z;(=2{zkNXr|QQt{mz5J46LO!<XQ7#vyizFOXbrt0X_lSRmvZ1f z<QIN&iIjhVf8x&dP4J{lb+kMvSUv{yw81Y)CSlMCTSy{5ZR=Lk1 zLsji?Pjy9gOLmBd1ct!ub(Bs~4tVrRFM7<>+QDYR^;W~e37hI|tjxyjpMQ6QL}1+1 zt~#G|h-|ds3@|9;jtP1tQ%n*?5-btklz_0ctNw0~LD@Fjezjix^?6F}0o;F#r4pR(t7}g*MHWHR1 z$0*Lcrp2|zwb8c}tr5=nNw*9zPnSi-L6~WVO4&ppmE7vD-xT2<38mq+${4O8Hdo_c zr&a_$N`Ng9G4uKe(uyFhns-?CG#+d=5JjC;iX%vbSD{=5Xwi@w8(Q>td! zkn;xNTI#)um9e0l*XvRN)euqqJB}s1exUy9kcROm4;YgfwLW@GhM)8cR_LMj$9}JT zd~1l!qQ^`>&?VnK(kzHx9`2U%?^s`Rf8`l*9v8LGYJIOtu&wZh%L8tbIK{JY<$Hg6 z8!02>_zf%H`}5mYf`YlZnmk0zjcA?o>>%Q}c=4{;MgtSR!LJeg3RVNyuv?vKVvb@d zCYmBqNFj*3EPXH+`^+$QH_b?ycwQ~R&%{qo@7cHi_C3wljEK1Z^>6%)G_Wg2vE8%< zxn_ch(-d=xA|FABj(-zMDsyr}^K z?W$(Z`;>$8lk7bJ%q-Oq={d~L;~pU;ymam*-+JX9xt51^tHP&Q2L7&lv!KEEI1}xS z*DeHgN_%L0%X)5iKplN`)!VJ12VSqCy916&>cFgZtg4~UhH91gnl&&j0Tl}`$Ly_a zA-K))_q?AOLG|#{Vkx;|!|w1y6PSETG2IloMX6D}p%SQ?G$<8)`=z3w5f1K{kk&kh z?ptz9QL?;QdUn|`r(`*>Aw>9w!KEB}r}Cj>nDbC^-34FuExEEvgBreu7X{9(&Z!<6 zono8YKro&j;an%(PPuf-*Rai_S+v!^H89U*>zsQMxT$*n$EtmBO9plM&3h|Hf1MG| zZl}|56iQ~ngMQtXCi1%U{DF!-A{sbZbbnLTT zo+zQ3DF~Ux*5eDNAY-*G&31YJZ`Ak{rOmoNW|EFrtzG7F+P__jTOmU!H*1Sv`XNs4 zl_ijv+{jaViX9~4k6%oc8Ew$9Z+-g;*jeo;}%cAtaeqMxg z6d&7*A>+rzT`y{IBuhUOVjUWiMb^Hs*$Ml+FqooM&cN(gG8d1(L5~ln_BmVfG^@F= zT_0h?w?dM~ce;Ym_?UfdHpT3sNE)mMP%_u+(kwVH%5`X!m97|cjN#@w)CqEU4g7>< zHU6zKT~IB*%=(D!WQBU`h;Bv8JIvZGDeKSYW&@RGOA79L9Zpq zka|+m?Ee<7kk|ujh)_<@Jvr%b!DGv5m4hd~FRZ==vlXew^ItQ9ZFc`JJ|T@Z47M94 zU^`1O5M4e+sU!HP4UEet%0-=BRve@!4Xg3l1-hC@pj$)7Dte$o0$0~kpY=;}>0>I$ zlz!}ZcG(U24f$!G(~vuWFtbjf^?c;|F{pf^rYE||VSbJL{YcMHMW@Lt9_(GBiCctr318wq3@M1h7Z!`Uz+U#bl!D?F+R!p?JA+-^n(~!dl1_q} zSAFE5YJ?L38gOxZ!0tX*XogDQ0mvZ+p6a4iW{_)v4L&vWqkww>nl$e`?jV6*kgnmS z2bS=FA^<3mzHk&~7Ie=ve2<~X4s`8p+g2_yg6_>TON&UE4TBB>L}OgPdWr$I-vgAo zgjYQGAQbs@1tqBVhrn{gbO`oB(n6~oeGOCVxPL)5-RxKoI>9mGc4&0E?2mHTwcI*Y zqcf&v+d`AMjlPCE)^Il|s)G_dW5h$eOvy>NJN(`B0iOuB z9wTA$2dpqerG5R7lyAh)q8;CBBI|7!Ls=#m+D0*3DH2Di6CHA;fFSuUsZLS3OuM|? zZNXY%>) zrkHGs?4r~dC99mbk&E=CpUd;bn6uXVy9RbA@Gw;mf#LtUqAQ?9_8IAP0&gPK`Jibt?vNuBI=vV6AC*jXco&(|& zV^*yMiixL445dEw%2mnD&=k+&;GME9?zO%jNb;P^x%pmAboop*=6nGrOVRzs6Ekz< z;P!IUVk0b8UC)aq`Rrg}!wJzQ6Ikr0m`aN5q10VLsXm)z+ntbv8AAfO^WvR)6c?mm z4D+3nrQLjJuuBkEEnDlYgD^puaFf$k=XQTh*z&4n1j9krNqehJcR|$Qt(V{7@u`^@<2P0SoW^J5j;_eeWj$x zhW(P`CUB^xnEe!~L`Jh#WgH*t0Z|ni>$MD(TzbyU7FoVnPwEXi)cVc*r!ou;8SdD@ z2A-j#(1NRv0U{^eWl0*n zMpzM?;F$|?L_N8Do+6v>QJ(Y1(!oOB5bT~hoLYD}wC0OD0VHiecGTyxNd`R8Ewa;e zIv=#U#24n41ZC3*rHWtx-CALuqO= zn$Y|_{Wmea$_`UK5%RU;2?1<26KaO3VBJ@*N~)O*u})q^r3*8}J636+Wm}UT>>fJ!*4mmj{yTc4Wll zt9^%JKBY)EcnUOr|MHv%rRWdTb93o>Sno?0fdkOujJqP_Ox1t)A*za{2A%Q|4&Iu- za3Dl;BWO@j3w+%v%PT~96K=jnI0Zh1(qRs^>{kREc2pRuZKmVw+x2%DB53`JKG}Zn zA;&KICJ(sGplt@C#9Af(D$qWI6r8wJZ3x^D*s8>h8N9zHhprLSz?IQJTCx1J^G0&d zv7fic{qBMx`1z@K3jin|7=p)!*>DrRY#PXI zH-Ajb6hW`G2az4eJ>WIvd41~L7J8jL;t%nT(pt|Fg6CU9*S_}gPcE)}^v`Ekodr=a z%yAgb$5TD8e*DAH^L6|L55xJS*9_;{LR)2t3M@U-D&xp`EAHg1{3S2}@>E_dYW|;p z86W*)gB}~+qM3;mwmPSKZ1d16H-@GtH!eQqwafE<2=q#zuY)f2j1-=XnK2f?81vNh zAKq&|+w$MWrEK=2e_cR2+3nqJ*sQEG$$s6Vm_CYp1~n*x%4NG$=gBt5(ihU9I=;xE z3ierPKAWBTmk)EcI7B*EIqj8KiPOA4k@vclL+WamszcD?P$F&%tq$oCbwh39U~rZD zMe#7_@X{TjXB^YKi-?YVtlA^$5R|*Lg`RMGB-`Qp!SZ%RqGAI#K46<$C2tT^37Z8y zqEw$oUKyS573JKw;ErFDV3^aQO!XObd_dMa;+16SaWHeA2p&mpF76PN@amwOV|{4W zf&w6mIZ7vm+=ukp>Rm4$U3{9>(W{3!tG7$Dy+@CMQAVVrTbGXEuw;Qdt1P|G800aa zi~84W{L;JFetmR4>5}!)mtC7-ChKVY`{d#)(mSGIPP)ropd8)d+bl@)#+m>BE;bSV z@^?S|t@>xb`ltY8?afhY}hX4o9q;JQp^tgW!0y? zhP^s<@^hk{_FKGidBt>xXs@JU!Hs~pWhY+lQX!?^QM%EkOV!{zQ&?i{)Ia`yPxHa8 zfEWA5C#{}rV7Flqn8~mm3#sJYcf;n0I(krwi$(|OofsGNg!J~-r-|m{j+b~b7Y!H`nH(UBCRJ@TWR8XTBu9JT)$R{rCrGoxSEy~X)<^j z?|_cZu{D4855BJUHR7ebr)(|BeQxm5Xo8nYirGVvGD;mI#=^gkmqbekUg}$*j}>BL z6{;#=nP0U-FR7s;?W>k0Dxfk?acn`O?_pBw+XwlBbZ#MrS>UeS(Du>CI}=jl-%PiK zfL!hs$wPq-5?=YjFo#IWnh>%EavmeeiqI+j!DqkyrV%+6b07Yc0ME*kev1v;lTH(S zoTnJjNN%CjXbjq=o17p4j;tiPbcvq^X#fvHm2wVU2KB>@+>s@ic_`)Pi|gnca>zV= z=ygGq>4cH%JPBwe8$#unNrm1Z5~mm)2;zTGaeoPvI)K!LpW$&0=`@C%;`lAlPSgdZ z_*Ybr6vIRomXAVBsp7f{YXu=milxt3W?JgAR=5*%oh-IfR>jX_AU*BCGLV^fbN=(x z;@^&EIe!vawj19z$7PP1U0JGppeo`g@5c7js@rrqX1jvH6jF$vs;|JR?bj%2dVSX!3TrA2L_S&~NcL;i9{lN!) z4P_YdA-OaXG0hBGj1$ZTG*-mKGl%+Ee0$MU_zP?p8fN$lI>FH`bJ^pLX-wo|D))|3 z40F~yV`AVCFG_JxkOt1}5qGRGz8hTQk73|pPLt~qRmt)WCyfpkqSfT=@ys>=;<-9d{LVOAprbakO zjmxq;mJO($wg6dh?U_Itd&7ZmZK@)pPBJI`1lcf=t7F5^ftu$rTc;$7NuWqPaLkcJ zk9sCcRnI-d({VK&4jm5Arh@BItFi~2owqP)j)j3UWLntI|JYfYzGL$%(Hl1wc)1vn z(fYG}Z;>4~T(SiPpJPyvM=`M2WKwEuT)HgkancMM7nUg5Iba-BSjjzf2AxIso=L}b4*la#Ux z<=SwH5o*iFtS@yGQ%#Zml)7J0;oc#L@ZBF0;k$l`*0Wi9B&bb(2Ly@kd@u7GXG7!X z-Y3mc9ZBaOT?y$&*eGEQ1Kx+?tF56MzP{J#8-{DrcBd=T8EBDDT zXY_2R(_eIe5@yE zk|<&QK5QSz``u@M7+-;Dvpof7YAT!P8h@-g=%cG0;<)!@J+23ocjhO1?j?8T!*)kC zV~e8U!?%2@*-qBf?q{_$QlEeNs9#}Rfs*2VFOY2$#Yt?}zk+_qvFl6@#el3&I;F;3 zavO9mL_1VN$69=F4UN1sX>_y$W@NBdE{2;Es8wDf(GES(qtK;FS)MQMhfbt4Iss^G zrWs6@gJ=B8F*R^J>uK}giE{bT*QU~aZNs59Gu_wS&baZ?2UU;QYbr!QB#N{iT4k21 zgV#4RH7k$jiv?gFKQj{t-7j7KHM7x_( zuUk~)4uvRp=j)W%8dl-bw4x$-4N%_YK-LbZtCe^jJ(*MFp>r~~Uyh|J&`gO04(Wm0 zoSLC^e>3noYBHQ^q4SzWy$P#{vJ@~+wHU14(XyL6oqU78>|31M%td-PUFr8Zv(Y}r zKNebawV=DxqhB|A0_vqxGyj%-j$`h!e5%JhGy{j`z0?VkIMk(uWv`D12Ac?p4ZA32 zcFGt@M(;wa8kTD?512*+(PI4EJ?hkpHvafx}z%lHYY zVYkHCaOmuk$r5vtVt_lQiBcCys{P{@?et80`HQ2r(4+LDB{?t0y`1Bd_Hv`NNr~6< zy|hZ~|4XCuyzq`8`tZ!}n z(=M7o#j+RYE8O3H$q1--t3&Hap$(T795aE)L5iuOfO$`yMxP^nF44{(iE;u@V?t8L zZSWoOg1Crh40t8&{tqQhl16^EFw+TXG&Mc+Wq-q^A*WNWBb?ET17!Q64ndJquDfPP zcu9z?uIb`yLFLl*OGY@jvBqK}YgNEa10XD$@n_t{YN49_jhM(Gr4c@%$A5Z|WY`hd zbB;-)lu-{kC~jGqoOf=;G;!dRb>`p7A+P z;xenLv0Y35Uhc}z#y1Z=S$J*3E{vHLA*74B$y*~#@YF%PO;2SuNH#MDwsEa$KUVPP z(ntYn;IBi))nVUasCC*)>I528*vatd5a31RXSSB183#0LfP37KR(WY2+&|JC{G2x0 zn1Szm9w%#Uc!^3kF@9SpCYB=6lp6VeCi;IeuEZ2SZp#2L#W$J65@GXJXc5BLbpr% z=n6X4b=1u@oIkj-1~@NTmx8=0?|J~fs-E8#nn|{KT@tk}y|^NRF^tz18s-=8)VehM zg`#kI8H3b<3MTZ6ytq2DeL3K`t?iE`qwsKX(D{m;OSKgCPqbak@WSQF;O~Mvg_y{YL zMa>`i`J(Yvm~4VNw(HRm*oTdA`3op!H$}23^>!&#b%S}T5uIhOD%utI$x(WBwf}L^ z9?4}<^@=*WMzp%x!FUE`tMq5UK+vaNy(+ouc%Q5;S+>n>$Sn<2E8*VhfY{WqWyG9? zcwuv!XQ_X|RO;+(*nc)tXLkVPbx?*$-xQmus9sSNG%QEv-(32jcgfri0h0V8?Jw3S z=Pm-?*TLYMj#108ZlW91R!}ht+oB&j7fLsVl&Z8!Rl%k+gfv}cj)1#Mm2SO9clqYeb4{*YBEN6SdF{a}rtHo$m<2%C_ zR~oVLoyLDIB?q4y|M{W`l0Kpspvyc$sYgJW0O+Eij4D%dz^f^wS@sCR#5)$nI+Qxa z_z$Y;R7YvUm7(Pk&P^dV9lJt%_0PokU-3N(Yg4n}9(~C-!ne`K&>1l(Z4Djx#v|a; z%#`T3r-Zo<^<30Bz@xU%Ymz!?oeH1I_o@*^IS(kX9JqwvL{CCgBbKOAq)S`uNyOB0l5j`402DP{{rVxeAhQK|FUxi@)Xz%^aQ>xU}4ZGn4z zFp+6(au!QIZHqBr$zWEz$2HMl?s6*K&f)_E54k`@jr;;ka@H^E0g?$jg@GH!(F zdC3YP)K^^Y|Kb-$guLCp+m+m~VT3@8c+48rPcgj|=>Y*iP(#=yZRABjk=;Q-rMUWq zYc7@IFk$Tr2L(Oe7gu0!=5><0D1QO&yemkg!}=vDp2>m=d50j2u6`j6{@>@CZhwp1 zful=sXNSOWN102Geq+`pix1H$vgmer2fXhBga&J9JZpIFprHB%w10S}B6#eb8|S2f zwn{CxdcOWQfTF;CuAAqU`dsxo>4;)0J+N&*Vvo{!W!wxpo41zq3-ZNzl3W_H!V~0u zCxN=D*`SG^ege9zP&aG!XU_TKvt4GAN&}>gIZBrZvxPcxJzxzK-fgg}o`07&O<8~aGQ4Vxm6uN`BGN+||buY5|KEsW-s z`?PpyP!td=X)vfW!b#$;5tcw*us+0teFw;u_*NnB!==djQxt?md3YM4g3Y`OSlyEZ zl$u?@9aIsVO{2)RwF}TK!}7#me$%to1vx7$&)(@i`~lYpncsXk&_veRFk~`JAhVTX z;wZ8in1d8Z=AmiusTCEAQaw-@2x;po~or*`u*Th28J|;r#fPIGb&% zHPhsV0uD9eHOqC1E0PY#F5%zVv`+CLsB5MhYAqLWdiF$G_YPLDpxT(Pdc0#a7Lidg z%SaEqXs`_@n>Lv^#*Zjwh$8nX^}vgr^V0)6>98daS80_7&dmMZAjJ%Aa{7&Ve{j_@ ztrDX-Z6K(WOCOe3JDv9#;Y2B-6o(+w1WG0R8c_pija~IP&CTFNg=nsmqJ>2ZvAYGA zxDG)!UA3&)9%V}o^MUxR(YKJs)$9YmM4_Jg70Ees7d>=r@QLB(agU42mnHfggtV2u zukN%cleb$Hw+Kgvu;apTuMrL&9}DS$d>7U@Bk8@SUD4(mE*q4Fd-Z|itwAYG@Y&~A z1Y_S`vwa+Ygj23K27Wg765DJU#kxzcW{u}!wOO-NXtG#oDJF#?+bQ)8m!i4NAT|S1 zk+_4|C*Q>&QL|?NF`HD5=9ay_0d&^oKy)ET^aOm zy8OBosDXp^@K+TLq{Y8Zo;6*`rimcK4z&}FFg@qkq4w~JpB)O!kvZoA+%#|U2Q zTOy3JPZFnz@4R?(c?GWzs$n9@18`z`7bjE66iTLDlO%W^P!@}FL$W6btxe19umcU6 zwQ2b^o*2|TXxjTDJr)|R2S==`B-s-&&)6_P4w+b#y%bYUkrGOMN0K4FEB(x^TcLA* z1YLC5;yo^n+$Ns_2`cm<*UUP|gSUl#4!t^BrQz92qGyZ@1_rul22XYATxkJPy2_@Q2xne9xXF7CXdX>i#P1f z@0ki)gblM3nPH1ST0|s*#{^NOSFa0-JZjLV&hzSpqE4LxrELaCoEQ_?s1T|-;vU0K zRHTcW1{m{ZMSo=)AoP#Fps6pr0%(NW7P3bY z!|iq1_`-;&DWpL1fEdoz(Wwvy>2*2f+7wbew=*b*w`RsTV_DDmG|cWqLz-VvCa9gE z8vb(S-rtu&Nm1Th{5h;H&GHISyEHN|T&9V4J2Y?OoE^TXCxPX{1;Sknh!M-8U28;p zLe9A41jg|(>8(|w!UR+muUm9M)UK%K8{{Oc?W^c#l{LI3uq_k8W2)!2>psgGA5>xE zme}wLWF|Ir)%`B7M|MkwStPA8e_?TurXVmq5IYF4^|nP;$jw$gpu3k;KujnGbl7gm z9*}d=R&EyQpf9e-rLPk$6i($kjBs)SQJn+;WC+Ig&ifETt+<#Wd%WhTruOIv+GsL)OPRf<1lYJBInQb_M1bTgA zqHsqj1~$=kl)A;aMo>#MmqeGq6^)V=@p^*Z!m6qGV9jB&onW65B!u#~Tb<7Upz_7} zLTu?z^(f^Tl1nfpsA8{E3mL3chSzd{k(R?kR#qw%k6Q> zU6kmME5!h`R*6GRzVv||xXf7%^XDfEJoDgEn<|PX#=z9iGCwXkGronE4OVP;i)5zZ7Pc^=9tF0GWYe(23Es?9 z1a;Ax4$vgN?0{U12C?&LL&B4D#ts`#K0jS&9!C6!@+Zc{=j`H(v&aP-&L!VBS!ixk z%uR~ifc9!Sj^85#iTW(q!@<8J*&m%X!ZaNj}l+Sumt17L9Nke6-5rB+3hw z#W99oB40DtTJGpHOw$A(I{Q*fPcIzpaqI3`-58@wnDx;4SFKM zQfZaLzEC^{7Vm;zigLXp+y^;L?{dEgegq#X`|#O(F>;IG)*dz24UG&qScQC03cpdq zg3BO~htKHCKe4T|c2PEszi60W=X_1O^c?KyE|5XTR@Z9Rc4;yBSOg@2_1x|yy+Jlx zuF2|4F}tavJ>+@RHAW{IJ5vA&H(V}@E^+Tk;_TZ+`QmmxgJbdBiJH{dTOf;{U}Hip z9@Gjqnn%m4QAr2K)0{t9aAm_jv>7sU{pB!a-e3tTe?=T@=x{Jk!l9mIK2I zv&~`%^T4>WHS)EoushoBLQKrCJ0kmQE{$z*AcG-7{?}Y^&C$#1UvSVF>0}Fp*#CrA zQBdQ)J`Dxh8J;PMz6Ccy5G)_aXFqoQRECNgIw;e+$n-1vfmC@jgU{d+GuoJoR26If zGiFsbw9w(x_yn$CBb+v6jsn{hQ0f<-N>ptB1r$evUygZH$d4!^py2`81)?3ao|^-) zNur)RV#Hy7`H~9WAlw+kH6%1}gVyRCFAkTr z&_Hl}iMtt;fYazSKWl9pEN{TFc<08#fIaQj1D|yMYJAP#muhV@Q`y%EAt3O0v@gah z?!!!%EN^EFTl*P|2pK80LlNDXhb^2 zq*5fAQe(1HPlb*iP^%KcKCQ|g*j_a$>!sBWpGh=*lEgp^C-g(8N2dsvX~C#;(vweY zS~{Z^g69ddVd?mjy>Ih|H`~VdZGWlkhM8XOP7sSjAiyRXJ0x-QRPG%|dZ1Zx#w~}X zb3fTRoa}Sv4M*~y-~8=7qv7~Y!Owq5cG`h7`PgqGsQY&!W$BPcMrF3*W}<@M#Oh z%|x_wp6~?O=XalMuxk-jDstT;?Iv#otb(PnxH3JSG_qJ>$#!cLK1%-TSDlSm`ayl_ zd9u}pL&s$%AkU;2@M2Rb^}Zi$eXVn4E?p@*PHNm6RsdBaXivrmHv^Rv1bWdP?pvf& z8Y;3hNIH@)ObY3scg`ztn>;WrL(Zh(YuWIVzKa!nXTMhayN{k7t)ijA> z5-1W+srP;Z8{c;L++iS^@zOO(257@I`Rw&teQ80m;2gn|(?UycZrWI(#oQp%aNayx zE}g!1_^U?yv9e0gMe6OyRg(|9g<^mb{y3%HV25vTYv2a!eGLEYmJc7F^+`h<7h- zmSdai>eYK(wt3u@Z{vjtHwABn-u`fYk~H0+54@o^#m&&`+!EdiQ8}mxHVaS|rJQ$y z?sfTp?0pMdROgwuM?AyKi^JR)oH+sIGJ+sRE)Es3qeZ@9eTJdy0vG=_iUxZl>$hI>=2DdBS9z#K~n z7w(s0XKf9H)v7#vLr@;wAuDiCl?^+UK;rYR*B5eprBn5VuZEgp)KA4youn5Gx`jnP zr(VQe^&OX`_S;xn3wh31w(x<98NW<@!DxT{{;PTk*)j!0BS#ziofHGse>+l#towG& ztDTE>()YqE_*m3}A@k+DwZQ* zs3iv#nLO^c!ovRt-e%2b%;9Y~-};=GmG5u)=Ph3 z9(FRkakQu(GFg?Xjlo~YvF0L|ZsZThI#ywva*{I09pQfM20z`SNe=ac@&qWffWFV_ur>4T6Rfh*@EROC&NkJe2Ob{J`psmc z#fn)KbOh=qiNg}7K}KLyieMeZ#8V^=I;o%}v=cb;SNe8%4Unt!Mt8fep{$>)1&b@| zv-Z5eX>pzWjT7$s&6QXj77;9}D@cH7Fybpj{r(@qlA|3USi-j|G%XS;kf9SH&%d3i zo!<>k?x5?iQ*T74x{u!hH51s`doVz!zAFBdSxp&e#O>tsT7Z|Q?d9wO8%}tc)bK{* z@o=LBd2QdWOj15p3zY+}DorL< zpL&YTW~?@+SBY*ZwU?zsimg<>#{<`7-=%95@pF;jvq_G*N<)-zxd?kOfd*8d?S!J1 z0!dv&DNkD$fweDviZ1zqP)Kc_5P;SKu!XkD^5_!p@x#sr$a*63#tS=kWVH{ywm@V= z(VOK{-y-qP7>Wu^P_&I=wqo=nszo(GPP*;&Y!iZ*mn=3AI$^Ma7~SUkio>Klc$Z2$ zAZv4Lm0_-IsU5f1qvmW}@EG;Hy$;#o(fL1rnfW~&2$K|LuOX_*s$x)ICBo$cem&KLtaaG4&1m7 zk8X6v*y_a0Ip-OlpgHH{f=vF)SKpC38C{UVFMszAiGFTe5S@wrNTQeoimaic^abF! zHAaHHdhgD#F8NXCiGhsiVqg(~Odpw$SN}Y3D9f4{z5_2Ql_rqbMKM59pF>66Sae0& z$KT_b^e<;{&$-K$q$?Ok;F_5aSIQhxZ= zui{?*lNLFSZq8{3qN_e7iY40zOBSJpZ4+ZSZx{6ZtF-hLb1me@E502V6D<13yF_KM zKUt$_=d}qhi4R27(U*Wte1%A-{y=$PB8l|$vu5$^+hvZ~2W4Mu+MN0pf77_I{Nm=B zG;-j%Sy)<3tjKYSfp)aRRMZzC7&}E3ZO8$Z18pAAYevtj^uFzRD`=f3s%%62R+d4Z zavyeTBhfS3l{GXFgNXHtVB%;IxU6DSz^zk$Go-cOadf*gBleTWDhhg z?cf)Nqr0PL+G~kt$eZrfPHOolqHwagIkSlG;{XGHyGzncIz`nv~995Klr}J2%?VA8L8ypSW$ilF0T931Xw31<`_keP*LlY>0ax+ zE6CYkQqG=!U@A5l6_C^$jp9$XKS4VG_U=a!ks)y(M#Z5os zdN=$y)ZSq$r1m3{1pUwXerb_0&=(1flI_ZJayL9qKJ1h#!_hJ%8y91xo)%M3*F84r zUqp-Y0Ct#58aLGJhOnHu$|iklP4 zo4yp>=F%6w^@T)98SgR=Iqni7(js-LX4Pp=dwt+z{D-;C!59PB=@h3qaNOPLt?zvM zC+3~AT*CVf?9H$=0QWd!^aU$z^JIyjOuK0|Fp|So3AEF<2KFnr&KwSbCf`P9?8P2- z!e(qt2O%wBX9&=uK};`CU-S|and3WdcjRfH7{<#=rgNYg5dtQdo*Vb_Ep5 zhMaH~FwOuWNlbm9m9D1;#9DL|>jEF- zv6+pkXsAbnXA4A$yq%zVcZOaM^ip{Luv483n5}DqQ+d%rRr4C1JAmQ3COC0g`LvW! zNa$s&J7gD_<9sWDrs{IZLZju$u1L*;E6|Y6J zX?h`l+13w}8v1sej^Q#7WGX)8k2o z0|(ryO)#;WVhSmej~&cNq8JClHi?2nUOf+M?hgfEI?}pP8@SH* zl(Zxiidg8qo_O;X*+9q+AU=(eHwL%Lup`kf7GvD#7_~sMOt4j_+0ijZx@VHl!ln-b zjEI??@&0;}>%fRPY=W4*6a!|ZgotaQ9p}=CuyE$ih za54Qke-Hn*s#c)$-J^&WJ+Ox=D9dzkoNtys z|FP;&KpzmL?x)K@C#`lis5Vx6#kuz@*ShqF8&s0==*&><=b$c_tJ%WXJg* zdrB-t#n?~V@9fIbvwt+__t4wC8SY)m(^qklU7S*9-UmX~QbPp|RiRq~!QCO@O@&*%*84KW|_dcF5vv#>O z|Am(Z%sG)A)+=RUIgWKdL8@KWEr^?1N>@Mw*J;m0UJ^OaI~o?l%MCl?ob1y;Uw>)Y za_SE)n%zLsjrW(*4fNW@%cAj|;mHO%?WJY!|G_TEuypoE9MG|T+GG0Jjl~!tG|Kmr zl8?+C4Tr@Fi|c?KTm_;GNy^j}qPXB9ACRvw1BRCX~NPn;#VXo*>Q+8Fk@Uh~ag9yjOpf^?7D1e^GB7^qgz_wZGF<1Q-4f82d| z9ogZ)sh4^aBT-HuOR)F=3Xss(O|)I23_ls zK{r6XNV>-rX+7vBGy-;F1GC9i-Vt6OKSn;_+M;Qme`EeiO`LzUvQBs;u*tdB4u#QH zqGrT+JQbZ>Z{UO)>e^DtEo*wo9e5#PA$U9vj7z&0U00Sbx~FK-w23REss4B9D)DAX zFI@}Q-oCCJm{zWs5ck9|x7?7#6%XUTKjjya(}6~dQr2C%g5+^?QXF_&bHv2{?4ua) zT1u&?JQ{gUfUCMf(L-Ma1{CmNQAh*VCw;%CYLEBe)Md+->Ao8sIY6+}2T!-Z3a9dD zgDg&qD!7w~`vU&L)LQ+>kS3a-Q;Y{DArT zNqMs8&&>HwSSUp5{_%=5SDnSrp*KxW;je*wSBw*ziVfy3&b zgnn{nnYj{@#6Uw%{|vDak1AKuY-OMU}-xF26i?0(I4}x#Dn5b_#-+;(IAG&Zo=&mcVE!?tAJpYXWdF0n6;|Kt7ETi!-1 z@{l9 zCCgRsmv(#JWLlZHNGOSfvlyYt;TO~UeKz`KL}p3u%1VLXt!rK}Z6}~v0;SPz)8ohX zK~qcr>#v!&8$Txw3JX(82OS-W?Xf4wN)2#J=7I`cvbtTlB-S&FELn=cDD^-#NjpaE zOsA4FG)y12+pD9i|9IWJOXoR(21`#*2{f|yINwurhX8vLA^{J3oR26P6~+^IIP8Rf zkP98VW}_puNO6pWN>IZ{*Ak=>#Xi^QpaS5<$PBfELE(&;$E^vRu|05x{`G@j_#0QB zhL<)}l57W#;~p|uPxequDMgB@D4e>D{^*0|1?2O!C{AmTj)PbUT)M?S=!`R~Qyrt5 zg7-{|nce~VDn|kv<=R2lcI6T8eX=w39oJ>oJ$8vQLve^iNdswf)8b)t&;i@R!8SlJ zpu?7MV0Tc)s7#rZ@$aYq$-ML8@nS~@c4Jw(F@|Pc1944F$ z1)>R%X);bPXT;bN7Ec(99WgmC{_E?1U&0Ex5Z5c$=zy^D)(at3EhHvX$sd_bY@%f*|zjXTLYL|^a-1BP3%NJg{u*eR6#$!`4u33L< z6x*L*w;p4xH`I%lUdynCd&_}?vMk(Ndlk^#MJM>y`V6{(x=e$pg!fYFnNvH7)jm(73siP3ocy7V`Lq{)Ge$88hiagAawQ>2xOD)3&*OO*7B zvml;Pr-+@M&#M!}3&2^W>tyY`W4v0Q8nI55<=O{}$Eq38LA`WB$Wc+dVkhi^^Pt&p zIj;=_M|(iGC5OgQ4BHsP)l<@TW$Lsm!7UoRx7K5TWUH}l>$sxUCljLipy$Ku1WD5z zbqfRflx|IYBu>Fm+LbZ#J_S%6$Q$U9-={|W0NyWk+I}C)D*8E~#&of<=9L8)<7Rw* zpT4iwJmBK6RZ%P<6oU@wakWl;TAmqNMB1SyvQx2LmcSsxHomrJ24{>Z;R1~@9=F>m zPSBX#93|Op-eb%q;NrkLB$kfkkqCda3m)wXfI?g74n+Z%89+>MVc`z9Qr>lsVJBn} z9a5nD%dpd?@I8tn>Vs-*H$_nx>`7khR_&EC)lMD|7jTVjYV3or%NzezWS;$TSWCi^ z>1kK$+y;RJCy|HK*bF@UcACVeOo0V+{FD*wbD5I?nRIw#e(YSM|MF(Frk)f|A;(SJ zger;wHNU-7)RDje0wrX^ZgCxa^9(LPC3RupR#k=M#{6oR-Fy`9J05gAsL1^kNfKm8 z3g~sgccC)Ha4suLiyWn@y<& zvo!V>kmanc=8V%UIc}o<|NJcS2j&Ka<$If@PWDX_?2xwe>SzdXT@okz+H97&%!(z5 z;4+@A?&X9C$6dF-k@lZIwS6bwwwzrrynQjU!>HcPpA!u11orB9~-!)$}51wND~cG!3{WF!kQW zVXH*7B!;)ncMJcx?? z2rTEdE9)are5HF%4}D{P1H|+9DU-pQ?k5fO3Q;l0*$#vluGt?PTo7T`)xr#4?1&gE zivN0Mo%-mF zPu@63oqXfu8;9R$c%uI4m|igZ^06Cd>@2gIb@w*|k2Gg3}_8sgUotCb(H;(3ZCU z*L2iJ#AOCXJ!bbqPBY-Rn}N>`CtjUnG$o6#<;RnPv3ScJ_xC3tX*_BXIzTbNwX+AQ zU@@$^3{;!uv3TE067c_aote1c1%?ddLagRhL0sh5WjC$ zoe?+5Tf8e=w4ES~j8$z|5mzT{ppmaH$+;B<1O7I35r{(BF5J%r>81vT<&}-T1MEn@ zd`?I&FoMVR-mw(2&wLi zb&o-}tDzVwJn&*S{S^&jk~(!MZ_Rv^vu%~-@;hM@H0)$BoCaz2P8!DQA^CotiX><4 zTDK$WR@tF|YQfcr<)Q*fl3*CBR0c>rZ$u@`8VuPSHKt%1H-^|zGloYqdDTkU+GpLP zDTAbNGq00>YR++a0+S@j3B4^WRbJ%9&h8Yp^Ulg!H09wi=MYgwo#Ses3iXABgZ%8U zfvKOlV%1WTd-k-4%6h23UeY)7vU-VqmRTF%W!#u}(u17vLN)Pj*DN(!o^tO~3ex$E zEnBgZM=x9VDdsLk?od$~bOApZ_2Gm-{?{f<_0I-&9-Ru4eLz;eS9Zbw8ovu>xVKtj*Zam%4)AN*{2xj zLn@&CL2dNNAaJUdToPa8-IMl(H>)~i`bUuuc@+c%+vqRkn-)WKurGW)um>R#nKobj zUb;tlc(QLEog~mU$+Pu*?7QevU{Jo{s>OC*AYusYe#lr$YXEw?xU+(Bl!<)}9emE^8F@C^7?EG=q^f!8eqL1bDT9j_1ev z+s6uxVF|cGVT>W{a*7=aV=QEpAn~s;=3J$$B%gO&^XO=iwl5q5@29360eMPWM_hT<~l%)tBJt3_=>kW7W( z$_kIq{Fe2T*y%_eq*Gm3nBjj+Q6t6>3kGO9XoJxDda?|+96A*~n@#f7H4tmjsqT7} zdzXjj(d7!;mK=-&2_~=@{bJnHKE{KbmYPXTH_!jlT%7Vbu`4XHm8~*e;3m1wP3H!x zq&_;;t)R3}E!OGdK?_u-V2R5uWwxYz&QX%)veb~+zzYTwWZw>WYfv=?7;JVb&p%8p zHQE=&`>!;Ub=)X49M=N}n#$4fmJEv7Mv<*lR0j=3e0?E_;|gh|3JSARl*tQfg3Cca?90FFZnJmW8~v@(M0Q}Ol|>V|e{SW19LWA`6rERS zuaPSeS0X-Cx2t-^dtGAWB`$yvARS!Rsh1Q_^95yO_c*s}N>m)I38;W|ZZPtKTNjX_ zt`4>B$~5XrZ;d(8FHob#vgP4Nzy8JFe2wO$aQJ6?NhUXQ;=mP9H74eygkr$H6i`v8 zJ@fgS_?!HWg7|F#zg>Cvm50&+*GAz^x5UUEx-Sw~l=VMbG}sNc16X)+eFt2Z^YVE8 za|e~F%r(7i<=zE@k>%l=T=z)NfOBk@9g@|M0o>RUDkEdfKmX@$b2?-$o{9rQhJ`#C z8V8lJQ!YTSdmVi*BqOLZDAVgs_{u5GnjIu|isALWkd;$fG>1urU+k1R;NG|>&h)}_ zXEiHGKb;HR3@fI7t~~3xI&fWN{@g)V`02UQwsaf%39;aq%{#(*t7!A$YITt38 zPshpzIWRWvo0yZE6mx?j*Qu!8{A#C!Ju zgHHyX@0ZNSl+01@PsJ&Mo*C6HcvKUxojIwx7+4bk;c4dr5vmZ~oZCQuL^esX)hnEv z!g}e?)JJ&nkpm$*6*N^twlh)ksVvjys4I?QXpMZ50kaXJMsKq$KOQkUWBh!={_3Np zSFQ1Y9N0%@@qi%p7nZ;jhfJj-}3|0OpUx}Ut+?1lB}6(Zbh8M+CcgoQDt!O~OK!^qNi;&$JEyf=TZHBBTAY_eFINHA~* z`n#M0@1u&@uU;`>*CDy8|?yT5-KezpnEPrNpu{>xy&N6GC=W23vD zV8WlJW{Xe$`P(h##By9bVh3JtSZL^)d9KBitK?dAoJ@ zU(5}O!={{BI&HDfFpu}n2-^Wbz+OBDPyXmFnd2vqV4tg;EZpSlo4@$3xd`U7WC#h1 zFy@ja8T2KWGjt9;Qz)x;`Ox!>t9G4x(JX9`U*{Xc(_&ibh~flkhsJ_@zZ5|_ zP+ICAUQKG~LD%b8ZhO_EP6k_RaCNY*7E8l7N9e0&F$+297Vq9cZ{;NkVq9+uv;DC| zQ2%QrO0ZKUk4}>`D0HeANO5XYlp|WM7@N^zD&hobcGvC{ZJf%Q$Fv@itWdS=0t;?K zQhIC6yXYO(Z(-?+?}vhweRDcwOX}!GA?hFHh<3QGXZrYgVT0a@yn62?6@Fb1h{GH} zm$aO2QmvVqLGPPW=5f?x^BkQDZ@0^?*Z7%RHTNv@W{cCDtPv%;var;=WrT}0a^Tns zOUrDbX9C!f%c3>10g}M92@ih{BLIoP)xrCH_IsxYu8AKgYdu;tnb26Y+8x;tb*i zq?ck~Is2T7>IZayF5=sI#Y7T40~@Ay(!2SYyj=dpIXCCtkrwd#Nu&Is->NBXaD5>B z@Z7X%4+2Lw?XQN!6WS@m^|Wcg;Vs5y{!UQx96j!4$yXs81w&4Wyp57=&L2mV(uV|t z{1U;)=tc7{1@scU$8_`|Cw%Xs`6JJnUTx8IN1#g5=-&?C{W0uX4LZvft*72)j=Q%z zd{Y0-2s;0)Ymk0%;EKI|lPpvh#dJ`l4Qpg{s%TLLbi7undkGT!bkIc%s6u2&4wC~f zf?Kj%aq7iV-cjdar))?EVA8t5zgY^p5Hr4jyb$UbK@$>2a-eN?Hdg%(J7Md|qsv$j z7#(zkr|k|o6qqPU2-V`N51|he-V|yqp!ccTtJ|f9camHYe|=Lc(1T~t)n4sNd~?Ga zVF;*QIqbC4?IJHT6xx%pJ`jhmU$|yIi+s5wfX5s-_?pvtIO*FLefj3%eGQ1JWv$+Hp);D#?Tbq0U968Dm=&PSB~}n z;Cjxg`QZ9y9PZ~dAC6n;6uv(FuQH=e65kR3fh0MwO#;2VQSJSi6a%rsR4VEQENv~C zPR}}~hW|iX9l}E`1tP|G4W3Rft%zQH~-mz6H!oYGs?)cPz-QIoIwte z8hRwO(m}T?cX-r;5XA;@jjtBiUzL@MvPm*88CaU5MH?5V%CKpnkNGs$5h`Z^( zedf{a-yQNP)JxzI_#&XH zX4l%gj(@B9ih1|SA@aqPhhe_E90D^R1Z7Ivz|d@0Z>QJ6<_FK| z&Bit5A+JLEni7@naT|j{VLPLOI~qo;g0=JglfB3)T5Vr-& zfdhh9O%|v#6myy)AH&XEan!pnyi~GP{pbkmKvR@g=444$_zptf!6w({`39D}l+Z?D zb#SksFuX;R=3C7axgYi3K#t3sR48Op<-c4F?0O6K`0Zc3Z2RKN(oRJ#zYlKRG`+(8 z!ommAegEF=-=c|I3=v1(Ir6r)^rei8I9Ae(IBfP}hY6@sC}smi)=^PeWPj9S6evUL3Uv%pH8K`fTtu7F0G2(g49jhP^925LCwTx2%=h>pE zwBb$92`cWM;>VkBksNkqVcGP2Djrg#LUs5dFl1zFdu&4Uzf8p*X>I(kDkZeZ!nf#$nOH!VZIJqB8#^5(mXGOPAgjuLZrmQ;%o* zp7^QR|DVq~=M*NVFfj%7RJgK1PRbB7nLzt6GDDVu!qJzv zcf5AF`53*{m^Y`#lZ+{(+Qe(!O)-TOM6IHb4YE(Ug;xo^JNVVava&VGHOef0w+dUp z^5|>igLz$01g954R3G$95}+OLQx-^KXC0Sy1E)_SuT=(nP`A3k0#S#38kr|qIx$c1 zF-PXura!rTTOD?ncsir(Tjuq_&q-(&h{8v0*YYU_DDihtQEkG($R^=IXC$WTQ)&lX zE5v!;dGwv|Zn?Hgp3gf>I%&hTVW&sb&e|LNbdOFNBX&sQh~Znix!~49-ljMoHP{y) z$DAJ`?%QmZh-mqUUT>>fRcdskoyQe7i*e+Wk$l~88^)kCn@~F>s^r6Lb zKl-5wIA%9@KUx|3TWd(|95}GgLTcBn!eujWhTdc8q-%r;3{=giPmnC1y`K7#?Pw8F z3N?8)c^(NY3Ds&R0!0=hXY@)r0m!L8_pJnTfE^38?S{hdmn=Z>Ih>7QnMAexI{Nj#MeyWtx|MzFTy`)b$=++6YX_ikhy}_*ss&Vsqjm}tutHrkH4v+@_LXL}# zwn#(3I(r{>60mYEYSaB*WD*biZ$XK0nJb%TKHK5VugDOG!WV!@!FPRkHsznNzn zxpLe7-&+$pbKs2)OXzH?I*C{0-ma*mE-U+h0rexn3O8&nX``<30(=?h6B_3>_hluoFvPzGUisruJ%M%!28%dF#k-2 z&RDVtr~_Q?Sv2EzShY{J3%c>Yaat+6)p^LVKW0(ooDw%M8EsdKbZTc4x932@kL8s z!Y-bA&wJq?(wWLw(1WfLe&Uh%!VOpZ%<_|_!a@W*X$ZSqV@JRkyA~>L^Y=cJ82!^M z;ihXOk(+<&z+MQXB}YX=bQF_Dku8QiWx7X|dv53huiC)5#CM%6y@MAV;`mXQ&(tHEM;`!Hz&oUkKJt;QSLjUfF zwCQ1j0=Qa3LViL%OXGim(?@Fd-?ryVPTx?+{escOKky1RnvS@X-V##u+?b9NCZ?mB zVh&Jb9~Cv=TDbu8Fg1&oy=UaA{X$-~XxTf<^wxvd?FRMF$VS*rSJEqkwYSL`XzM~* z%bMU$C{o)tEnyDYnA;>%6fekx`jk!@Hblz>2mFVfZV5I9V;)#*xQ2b|<>7lhkI%F= zG0f>}U}9gM`ogMj$86)` zAxAm#JM;=N=o%fl3_Af)9=h9xWivKLj#X$G{qXxy*Tm*)(sL3zVOd&RWwEnjg1(j= z&kaR}Z8SP5K{u3Iv?+zu|1PbQCP{wWH6z?;Q(oJ*E0dHzW5oVVCYIwvim9WZ;4Uf| zX!K)34gWT4a$s8vs#MUp)j+3(p7zY4-z*o$z=>@$@pLM$g~U?D;uf-s%JBcw@vtv| z!zYgy6PgyfUDmAG|5C-TTQqsR`y?iG$SIF^Sb#s*@^-_|4F9tlL!u9P==$ewRd2Vt zV#UT*)jE zr+EvZ|*Y#hUtTHG*Z>DcLuHNmf+NyT}`M8g@d0(rh(0C$4r` zS|@CPoet{YA)p3csC&V{U}`VoCrc6ZXt^*yvJbY{5%l`Whd=rIQa8KfxyJ)HC|9Ie z{7-{hGX8vlIETELcDuI(hCxfmAT!lNe>q(BfkvQ$j_R@>&@bWNMY^L-sq|@@C_bUlwqkP9h%>;L3!MLrf$zwj=_{>Y!VR zdmq0*gj(9#qhSN#c_4eLZ{Sj$oz^JJ^DYn_3QS-i!c`~h@IqzNI<>XIOp_0a(|uj!1AMD6 zKIX$moF+-FVN`Np#IP_b;m&B6XsfD%G%ALjRzeg%*7E|;?IDA3hNMQEO<$3w2juf| z0Y|Z(m!;Z9h|}lM$JJUx40oUVIijb`Nt@Z`hNL+OP%Js@)S^NHmK=4Os0`?cQW%1BUnD8Y2FNtE@sGhf*FZO+zJwbM_g>EeQBfFIQ)?SgJ!V$NIx!}A z*iOFeoN>X=xPJX{_u+M9#}ra;vVN6QObG>w?NLqg7HDud7u+GkOv>rNLFG|rJX$kf zdv-xt&`DLDEQ8Jn0;Xb|watq!Edn`ivN0I%#2xLh6VBXqk9Tz{JXIjVEL^qMuoJqW zOPAVoFKg3}o9(k3$9YbRm*ak6H@z3~BXdStE{m4~7gMtE*R~12m=O(m}QUPXDUC$0^%KRr9fhi>BN+|7l7d$8jCxClenAi_3x8zUogQawiv_{oK0j0 z@$KTFBd;0#5Q*o1O(Km`$QLFH_*IIzM3Du&n`58bT-YZaA-lBw(qcYC)>3 zTrmJE1!~hhg!B#G+oEZsPfb5Hy-{|A4FSm*ntr$zH``K*mv+pRpb z#B{OG+8A+n$K53StMj{ME=GvGtv&Z=vYMN#>$p)#0<24;T-_}clR}XVR8(R3AV1mn zk_$48tq`38QC|$CpmE5=-vkf^bX5v(-JD65o2yyn51MlZq_$koww{ zgX#`t16U>_$=}jJI(uO~R3sn*YT2~AZ1pLcj>BcH?UO1t{e4DWak6HUOVS(E&PJor z@Qu3nNQ?uk_-r>Z1IZM#o+62;;*&Q6D3er8!XD>y)6122Tx)$!sUJx1&Rc6&jnt7( z+~lXFurPUO0$y-~#qZvV%2yjv_?!yt0?n0#jGJGTFC`cbSKrezycU!5BV4 zp`Ieme?1JL3SV61+DOYekIyb%Kg?wcWuzl5%c$avW8h&P|4b~@1A;-;?vL2IR z-~ER$b2TW3U3FOWov`INL(=}@t_Yo~YR2B69nyB*(5x7@7R`kq5c~$GdyA|ND(8lr z?&#A#I@Qk5Qo2!oI^e1lmL#7+xU)@YxVKxJ$g6fa7?2Tw>|Q$64StF5jgWgGXXspr z-(tgaXUNE?u_TASurQlGLl1xobfvNh=y?pkMVA~@rr@_|(d?oJWC?B@NmK0uma(D5 z1wx$c6SeI~LaRAppu<9og#xhd9}WAwgZh2mSTA z0kR(G#^^azF!dWds~9#oSOb*-7DJ0cek_2}JI0NQa&f!zD>_r67kM!3ezYb0i#)s~ zTfNsaf$5MPCFv2lpJt#z}ct16{j0GVP^8(@2vVK^F|_;-kc;sH;oH< zG&Q7Hxp-*D}kcTQ%w`$%9vuS0f-?1W6tF-4? zSe=Ej-DH`~BB!Z&{ANy*<+xFpZ^GuvOHN71G)=oRMajJ z2ydVi<)e;%o$8u0ABYn=Xf2dQk(%HZV5xm{8U0^dwxK=_55t*pf|qSLo1Nzb-buaf zue(0$LM>YZ8-=<29*^jtG~bJX^{PY(kVf){zW2dvI#s%RComcJ&+Vo^bB$9@l;^^s z2*F^vY&QE(8;BD+sEYR6!`756IIe9ZEF}vceCvvIvt+=vQ5fU5$rtp`)P4M;t^;Jb z=onq6I6;m`ZU&^vVT%OWhGxxg= z0eRyzK(<{f$G}_RzfXxaZ&u_|`{cmE|4frbY6Hcrqewgzg;jk}*Ow!T=VSYsUU#8M z{=cn_qrg$I&G)UJ8hw!yi-Lbl&N#3y(ql5sZ4`5bA{VKsK4k*aD?i}hCPc0wY)v>6 zkiU5O56WJ;7#JG}N{A`Sv&=DJY+$>xhQ^xNYvSwjRsOf;mGHECJx_araPk02VDRC( z;DbP)@x^zyeH)ycbe|Y`yuNBTDIyWJRN5wY@an~V;rM1-WHoe!B-It(pp9&oRlD>l z`ydwB5tc*4shWW8Oxf&Wag7`vLxOGGUgghh_0twe41;B>Z}72~QX4(>2_y}+u_b1v zg~Q8HNHIC&v|LT@i)##i)=B1&glB1Ac;$j4;qA&4ezmGok>p$>zD_TF!t|MZX104_ zF)D2LmKKe~4HaMBe0`sJ#_>5>k67}PD}o++l(<|YEvg5u(Wpb#29()bLB=rIcQ0C@ zQkT*Rt;)3+M`$rDLV#^f$66=ZYDrI8{`i!Jv?uBdRG>LFwW3?aM%2^LumJ2mqRA~6G` zNe&{IinLHzd$5*wq{MfXppIVW+blm6SR07)+s!Ia#DirEMr&4`_8fFO?b!xJ5*^Yy zzZIhBpdB7NJQ8``^lBI2qAHLKy5-Sv$~5=|ABOaWuMhtULQmM5k{OES!_lG|F^sL! zUWAQ5q_4!+Lt0evE;~a}vgV<3god+xH_(k?RXFx+qViC_^dYE79|}N?GM%bL zrBiK@K{S$dh2;TaYajuU>{~QHb{bxXl5yoB!3MVnQoIklLi44_7_n_;rn63+00P4` z!SGH(oiF}dLu2VN791mvYlL@Qo8(P^actAUt|lDu(Ky^TY7>YoPA2c?r++!`S@(nv zL!JR=b>q}h`i>w=eQ^%9vBmJ}Cz!$?KX?o0nbREE;$}{s(4=pwf7ku22ZU?I#bSf3 zs`ji)Aun2#$O9Rn&^FK*Iz{$C%^h<4w=3~BovK?rEJdXmo2Qw^nmY?IVzy|EHIi-5 zal**thM(+xAUCc_$G`uNm&pchK9K{5{Gdy7REl9c#bi)q8x^${LVnAk$hl0P{Qit- zBXwlCpjNhjdNSD~t@2;4Zqal}Z$=u(_%91$0??P0tO18LoE>XIAgsg>Hs6}sTNs*_IP78%oS;-ph@0FqsI$@oxPqz ziuDoTbz;1wldg6_ciHfmPPGi3mJSw^p~8W%gN$Qa;{JQzP}Mn4KHn| zB-suOrb8xR+Cwp=6e*^n&beO+>?IE%D;X`ykhExyt4~21;SQa&sGq(Pc#&5DZC5Qk zJZjdI`B(T|{n4F8XEiOF%@V`)d|nRtHh9;Cg^%tUd>QYuxt*~UY&hfMDGzXh4fV~W z)H!pEuvvU9Kb{m!A;(N+=K#fk^2;8>t``Ze%lvC1PD4FJ2MzI6`H*6l^k6_@@U@xn z9Z&b&5&FQrC`|tif8dTHsGw}=jKnd?lF9{m*RT`ro|7f{K$ToS^K)07YJGUVpMLiY zd>b=cWxZe_Y&BW7nxn73-B0=KExpWXjvRM?`qx({r3V?o)Ahq!WhDQ(@p52w8|CFx zQp{e8lv7a$poDnUj3K8)<|?y;H{^6?Mi+m_f+43*JnkuuL(OO++2ps*cgU$a;wp11 zXvpaUzw1i8Z|m%_2z(p7>Xr~bgJ%!9}wr|tCapiO>5PFtpL_t_@L=QcZM z_}dEp2uGWK1uVt41^4y%78jWBt)G*q4U6y%Mg$8aJc4Z#q@0EP8s{1Qt|` zOpJx77`c7IeriwfnnR5iMe$c{I@!(5qByRtBqvQQN-f1yQNZ0AwMukLslBi0fD%yf zov)F+*(u6*x*_1UTLGSM4L^ZdPilabwGil95BPUOyg8BgDa7`)(LvWkV97<6Ac#Su zI!lM1J`Z!3;5L`4zv2(N#k+qjZdRRvP5gkY&8(U$2}T4y;1k z^8b4-UOIxH_g`ry>$oAvfp^|TCJ4%)m~9lX9$IIj3C-Qi|%Kc`K(hYqIGcj}EAEIIHP|*6^(}heIkU zyjDpS_|=F9Tx!I4S2-OmLiuRJU9~r;Tjl zhT!tnfHDv~!L45geU9{c7?z@TWwlo)l*r@KhaGo@-n{kUjdYwmk!RhqG%_b+{^%XrADxnd!JH}P6xJCu_m_aKE>Rn$Q>%GdLgw)Vu+4z5#+Ef(}S-my<^4ujms@cZfEdAT?qgy*^?X z^pWtnYM0vi(UE#7^8u0pB*HpXx%eV4i);{N3h;G#I3lrG)ffz9f!Iq^GF#jKVuu&D z-rxYc0)Sx+bC$@{n4+vw?eQ)OgHsT))C=EX*fI|qaG*rKjmBZxm5IE}P|T*TCwMnp zK&oO4al(Dh*FzfU9JS--$vJJWZkF5$rHsgX==SgwY+fCfqX+U2qj&3x6tk8ht5L?h zc0L4#;h&MBQac(Kc1vK1*tB$%EFd8~n?>CUAV>!D)KniA-xpc*lUu6>CM)B4d~q zO|CkNpS++ZxK(*!;i>6y?)}O$bYf(x{j_Z@0oRjU8+*ZP@jJhM$DB8dive=r`gImY ztq%mpTsvT$7;(XpahcMKUzx}Ko+&;xmgpV| z=@{g3wJ5!_T66-a_Hpg8HqgTIIX>yMrE~wJG3;@TeZg6J;;+>&7#A16|EgX>ws2cq z9N1bw;$YO`vXf$<+;}?`Wgt4jGMh~2rm&?@OsYns0WRfLk3MC2IA+>nXKfAC;+_f} z=3yt`aFZ>?X3YXgn=nV+2`+NxRGacykDhc(LBzVaj2_H}$HCp->^3Kbe%ydK@ctR< zWkvyFJ;fwaWGxldEiYk!QFOt98L^()6(ZamTOXwz*XM58%#Ukms~%!EA2+3`sn#SS z9T+VviO2!hCQzClQshXWTIL9EcpefStC%Z7%Gs^P%)N zRi3~ck=+&#I~@ru1SjRTc&+jTIpd{oZP)|;3g~$;sngCu5zaBi@@XXjHe9?cLlvtv zdF+B!P#X2{pIaVk|GxG$w_mv;{o8e@3jSD?B&dKFEt*Z!FEA%SIH)`vSC+zXoo{n+ zAv7Qi%d2MOh_Y#j3Tv=4Z!53E>l8^6K;9Cn%?tT0mJx-H`nWAUv>Gm!z7fa!s0vYW zu+dYAOX)2kMN!;(Hk^l%~g=P3*)N{L9+pKC= z!Dp|^1qlnR*2a60aBMwLdZUV3q3od?tJ^*zr5@;3f~s1LyhC;XiumFCe^7lWU{`3t zxT@bC`|n%*!9Vs*#S@Rc4d)Z=)_;s>h2t9Z-&}n-4Tk-mLh`QAN677#{U5uOg*&2Q0U z^747T5j`GN^U8T!Rox+1rG<38>?G-?o81co>>gPeFG#peIrqDNdTOBMSugp`pfQ7z z7(B>75)Pc2;JQ``_xoTvXunT7@R+BB;!jj6M4CikFjwrCmV-!PrgIsuLUG9LxMICz zzw}68Oi({Bo4!x*RY5>OK!)FiK$`<&yg)Et5xGzGlb_(UG*Z|9>o1JC;&r8o>E1;# zJ1LSwMJ0qB6}2lWsmoz~;hXrTubQFo8jG)yH)ser)ACyrz5i*N7x z$=zq2Xgtlg449@ed7ngN@w@eE7&=u~&{4NDz}q^pvkbFi=|tOlBA+%MH$&!|cjM4x zqrHe(6?B9QJ8<1Vx`_o?M=|jfiNjJR%;Z4lnVyI0mV0kR%G3#mgQbXg+Vt7sIwwR- z{$+-c{K1Hb{l9dRlUvWt3KC~h@$?nN3{d1A71bM|Q+EhKGuuD_f>m!#^4M7*J}fL( zv_bJ@5tCzRiQLRvF6x!1f}&|N#DA)0M@M4JAOD>Aw`GQd-iUnOwcy67I#mNNPOd!$ zJPs95AO+$2PDu+iFSmS7>JV58G57Hc!!c>7FMPr{3)Wv2 zh^~fW-DP5M`%J8B(Pl%85vvn0;;&QTh6@#PQ49(hrM1{O zJX^gipx*5h8BNujvXO5|r>YFwu12{%`O zSd8wkJ&MN;?1FE0ba%c{A0RXWX7TyWACfiC7+@e`JPI(W6tkHk$yC$_g1r&=w^_5! zH-TIM4W&(?7&guZ!O@x~!3s(3u7FnAffuoK>5!Lw ztjCyB>!_&t;rD9aGAD|6Smi7h>i0`xRC+B8Kj(gR8dSjI&J|L9>LczGIKRe8w{V)r zgor&)vKxmn%3mg@AN`J8W;7V$JK{f(ByI-7fgKaj(i&wjGARavxT#cBoA5wbrnGtf z3Gon!Y^TjQ3{j9>5gSAm;yU`YMCZE`^!3`px}@dwjgW)9>WE6O2@*1~5Kt!mhdqHb zGWLaKgYuV+5Q*LR4YVr5GU){TC z{-TP9_%fHW{V?vwVlYLi*=6f{Dk0=_UH8q;thl1DB@;$t4W)Nx&Zq7X{STBhQ z>ZX%@KjhyJNfE3LZPp;InCVz&HTV51!;%D5(ssq?3v)u3FX(XwRd)A>A$5v#;%517 z{vO2uZ1S$tTb;QZvhm=?1Vom-7=eSuKiLh@mNwx}%`5YtlTa&5&3+RoP80*dN{8&W z=WQsyE^*OTF8X95ec6#|vkWaGcTe2U&(`}zT2sj8z~zoCM5tfLlYQSotq|aP5_Z!| zfs|cb0fHq5)lG6N?Zxh}45)J&bRBkT0TMigO7nf0rk;A4F~VksZspK19=QSot8uQ zKw@#&DN&LWn&mo3x4E?|+r1xrLw`%1x;)Hq`qX!_NvjN>9(H=WlTKg;NQa_ZKB(O3 z1}v&UhXsZFL8Z1`Ii$#-bLf@iy+Z2!bPqk%0$v#)`fI)NJgD9qb^^MpL1i6X7(U3a z6XeY{oYwX@>wJqL$lN5F zXzVNuucTMQLVr~X3AKQm9{VKNTmvOaa@Gay1mFE2I9M=Qv zUcwmL702Ce%*}ZBQ*&WQE-sDZ{?7<-9E@_G(^L5Jm;X_li_v_% ztv&Z=vf6?3fw?C0y@g^@D6)ZyYLa88U|8C{V*}Xf1Sg9&>8R^pLd>DT#c()q#Fhm}&8i%_mgtL4>gZcRjfx&9E$9O; zI7N9-T@Nq`l?d@1HPBr&`)3tXPcPTphwrg`S5CyOfcn zs}HzMp`@%yo~=d|#*e0T)9b@8k&LPLNjcP->EpmUbt!bEqHo(O({r4*@q0T95O6h~i-(+pyDK&)o{BU{sg!F7w)m_A}SB3$lS& zqi5!%^AKw#67Wh1nBe!}ZeS9mkGX)7as_t}RPr z$C`h+z36}5+a+Bss3rR2lmA>t7fr}+d)mP0c z_*n|&8=ZGZTV=7cVuFA{qE&`zyoqxZNB#dT`Y1=;KT(I=Ou57Mfc@yF9>4qh(tfRU zgU0lO0i|FX+Ig2;pyr@B%s}U*r)p@Gr3BqjP9W4R0{A1}WTNmrY4WNSu}?2^bKuPk zOPQOYH3Fqlfnhr^Ix;QvT5y)@ftjd>I_%UWKMBdf40^c;1+5As*x&&r^W5{ z8PC-E3_GC;Y_zCbTpreGw@P}Buy_QJM@M#n3cKAq9rxQ`f8Xe?zLS68COPK7o1Ysd z4rmL-oTbPa44LHf8bEF=H}p^-Iu9skhMdB6{Poa9h-Y=G`=pDY{y2nI6g|KgI_P@K zy-(2+m`8Vbq4WB&DowQ$`V32XYv$)eQQRQNsrHj01#%8!O@5y;MxMza3=1S!3sp}K z0J(oVoYlD@!6ZJ5SMzA3kX!9i;F|_pq(Rr+)0%{bLEo3tM%xB3dIYGa9C5)LJB-E% zB2RiX-sNVo;JC zat2HL=jN+Ukd+$np9^KUL(G?sw5QrK$YNtt!^N6B@!rIu!U}83wjDUc!cw-K$twh! zNu6qIK%6@Jg&4?}pK+}Y?s2Jh0reqiqoPSz?Mxx$eM@@N$zXmOhH2Q48}(J_B9S za=JPgb$&DGbPt1xXyKh*aDhSBHW#vuwWKf<&vYwR%{Sx~k_7m=P1xwVPj(R$(eP}N z;L6l1Q$KUXv(cg?UL&3E6C=m?ZaD;hAsQ^Xun=OsUveo9(}6iU>2-F&d3fJF&=P-k!KxW~kUubVK)0v{-H=i-a|{t0PB?ms4{XKd z$56@c&WzsBSaz*Ge2d`klI7IcI6oxm(xc81Or=yT<&HrN2>s9P9lR?w-l)p@g% zeU~lEQP*I_*0gl@L8+C=PJ80*F%WS$@vGb|4d&`;ToT(3>?E^jt>vf>%|v<)bZgL4 zMpkxxda_Aa?p^JIu1z^Gw5Pb_$r5?m%+O2Xgiy$80!>D62Qbp%`3J68WHId2%xi+0 zxxL;g(q&ms=apd48mC2cHW^tj4EW%B`!Ku|J<$W=%4~= zigM7k5uVyf*YNV_exlE)_9?nx*u2>~)j;@~`S(KfYRZNXmVvM%m6zdhG7QtDi4shv zwktnZ+?;zYWVs-L$<*^gM2mX7c1M;eA(fjO3edGD{H!p4l+&zfv!322>A)akX_Lf7 z5)+&H`iWcaNIV3&aim=C2Zsvf^j3&^g+<<0S``jobZ+$%FsEF`w&q^;74MC@fCedahnjfVZtuy zZbe+gI^SKQ!=alO9GE}kbXz<$GoC->l<`vYt3yuN;g!-%w;`v?b4oqRU&!YjCGp`~ zBM$}HR)tzW-?p+T)_`ZMXWt{YcbHQIIcx_d3+d15DZBZZydKieTMny5+Vl*s7EK+L zm)D371r~>Qxu?Oz7s@vHe+eZwu~WLjHjv}$cHnaC1GbmD^EQQoqoXRCR<5?=*37!; zA8X=lbB^7fjNw634Hw^AWG)xruwY>kHbbcBRA2Fm#i)yT_P2W$cl`wV;}-4v(cVQU zW!5Ec(O{6LUU12CbzrGvknRvZ1X?gueC?8#2kZr~tR?NdL1ijqH^5j27yB(o+~D%U zymw-z8MpbZFP?vsBtJKyLx>cPN>^l43@Bsis3FA(c z()h?hH;fXj5Mi7^zju118;ib=-QZ-f72^k$ZEOwOC_h#HpwYY$#bK=qONR`KuSSb9 zB-JiE+)8=ZJ#bT`cW1E#xLk~zn|?8rPv}(GOH&{!nPun69{-fHO_MFpbFw3oe$wsn zxjE02!vchb0P91@l^e=7M{-PU!e&+T|6}i4;F?O$^gZG^Bp-&{2;`gr1c@MsBZ6T> z9Gsc%wA*&uZFhFt?QVB_lkL`>*|yteW_EhR`yKIu7tjF8MG#RyK~xYN0WYJ1qM&#K z0(co51VjcE{_mRvB^iP_kg(DIJHJj&&gCPM@5%eU-+Or;m0>v=6EF;#K&wr`pl8P7 zkRL8SI9M7IZd_m9E8LSz@*OygTy3&qd`fYp6e)&!3U|~6$&%)(p_+kQk!w{)gRs0N zoz|E4>TL*y@eDH23y`8T&$CBfC3)alNhi4Ipb!F96KLT_jp@rF&GOUA61oS1GAr5T zSeJtlk9Zch)M@HWe+(vTw#$)y4aXQ|9?EJIjUebT8np33(fDtFqhbqZM3L*`?yF=A zKQG0BH_iu5Fqc7bkP1(sqPu~+tS(CX`Xc9K?hLszqlAuS5(1DEwHo+&mrlrlwuZdG zCOR>6*#uOfXr?>pMiMg#&d%-*ucZID)4shdqruf^aPy*PSN+f9_srKJJ|Pwdj?7tD zwcA57d*ahRi|9{moAZA78_82##1^M}PGM}t}Wh}@( z6AO|;alovah@m%Rmnn`|9opiYI(MUV6-n^OgPxO^NjF8!qP?^4%xIh5PPfbksyqGF zb99z8Z-~U=i+*wZuaC%f@c%QspLl<3;f$A^smv(xyfvVMR70stNyr-DZpDAb+7MV> zEya79M*9Fi*W>k@6MMdEG#rP1OVt`S`|KU|VVie9ps;93(5ouar3ihN2%UuEi4>Ea@)IBeOh#?Tb$ zjj$qW*{szpigoRl$IVHR=J*v+=cca_Zl6;V)v6cj$_qdWv)15;%3Nu(xCYK)a`_Qi zqKFZ%5nc_-rN01{hYXjxsFcZlN@Rez7u>4nebCKXHmk+8!gwcghvS`lB8G*;aMJ?U z?24Ed{O#c`Tee<{{xyDrIs5x7;`& zQ@2UR0y0)X#C-X%6D$S~OE;86n9HQei17WN{XF{ppZse6?_@ORL?{RLVs@IWK$|FT zJw;-{i;-ggec9Yt$Rq+ASUanO2Kx$7>Wqiiv6%3|U$WJcKlq_BIJUf)^e#O4$z-Er zvgEhkYek*twHjl@s_^-!*k{x*nD8&z$=f1-t0}xCMb(+C(RdD;OU&nC`V z!(z4=8Qwk{#r6jgZ~gBs^X4^&tvImIj@BBsBQQ%`1C{r~{bSGBHJ=w=vu*yc9pCfv zMaKW2ZCRna(T2QNwxgD8H*WLv z`Y(TEM9Jnqef?|XGC!0!u*ceEf|3@BYof?4DtblG8Kz8KIXmvn=SOu>mGskD>2Jop znI5?D&2srkRb9jiaon4|AS`b<%JK&dUH}p@IMhYOhIIW${Wlj0E9uqjPO{y3>9qR+ zCxm6bby2;7gRqK~()mz6c|&#w8iqCt)(A5uuZ+AQ-Rzy; z3-(FtqOLJ5vl|HX4pAR1q|ybw0>e#>qPtU)raToI9zzAfCVH1!20gqc$|!(l3cHcK z8wRQk0l)ju(+IShpA{@1+a{1glNrjSIG`=v1$w9KNj*=B=Dz$3*>&%9b-&Z6f@auU zd=53uC^%Iw-$UmDVGlC?)XVb%@&fyuuw1PbrU&}J1S=!Z1laKhJ-RP>m~Uy6m-wv^ z4$EaW-Yr-&no=hR-Y;1wbpqxXb_tgQ?SJUbYOvoBD=|Tc&kahj>*be8V(0@Q9K?7h z0_i{OpCGS|+iki!WQn4G<{}_3vgZ;uI>6hn?d-#!xQ3etTpZQ`wh%(W4*gU$^gcHz zFmJ2_jt*q5gRaprD7H5@IMz?J`IGj2&(E?I6`rTfBe$w-~#C?QFr*SKIP1%k3ZT2-cJmNbKg(l@BOhLebxZQK(X&`PDM z*ONFQ5+oqyoxUc!hn#hTBteC9{6tN$OXIwCuqvU!GVtNq03$X4#1KdsJw|vLoN-Kv zPp7$r)>ku#4!l{iAe5EKq|?=+M-jM#u9CMq-xIC|P$kii>HTi&*m8N5B6IEu;jzgN zXD$xj!ZffCAT0MF`0-ejP>u{ZqX))|Mz9Q=|M=BAKQ-sB`zlBrIPPMBp$-W*3{7O> zcIPHfY%J8OI!NQpPPU%S5|1Grf_`pz0OLh-Y18YLf#Lb)8t3mB7oQcsKKhWH9jXG( zfp<0cOctc;6nB*(wT7mlW<9Hvfio>jPLkVX+Eq7S*&DVK$m(jb$jV^IzO z+sy6>H-z>&-sK-4QZw+?lZ(mN2bqeMCRdTR{O-jYoeT=rwz27yXiArF7rY*MIQJXVr3e| zsTZKpGQN&IzNj`|og8+TWML)Jyr>q) z$>Isngf%$OzLDbMD6*D{Zg)R+STbM}7A z>&tgs)83Mzo+2@#VrAm3Uy`-_@&k_hUq<$rfO`kUB~WAw6|J+X?Fr2Hed3x=pp)YQ(65v+C4yFarW4Gkir-|LKW)GFylmQd zk#|_Gc`MmVwcUZN@}M;H28vrpku^}N0u(3^4Z#?Fu}da%8G7N-LX?EqzAExXQ)w+w zUNnxauUiJnY}eGp9~jq@e>h8hom`wi9+-U3O%!*FBG>VIZg-An@6SyC{-bZ;SNxc+ z@XZU>oFWI{bfc(LiLuN5?p;0$bRd9fxJyseSnPsae9aIAO`7;r7%Ni^~Lzm9*1*b?Zu-+nJcCTAW2r-1Z+XP zpjd@Gr1(ct)dgPQ@fd=rWkB;-#$*9sPO;Adyb5abP)2*tz1Q`Vq1Oj%C+?XwxBbD8 zb8dO!hHx1jcWtlCTk)fpJuPJHVntofbLuA`;kDXLgIi;R#L>{O8gfo;angFc+TyQS zKL#(nj{iYk&==;;$}5r^SU5eAwENT$Og6*G#1m`pbwbX+4ni7VkyF+3b3L`H4PEr_HrMRi8l#XT3InOWFe%yOzQ^M~T>aZm#|MW(|f0@-ZC_9M!i_munE4 z$pYiC7i9JH%Ma41H~OQRFqT(#MC|jhXU&NjkofL_zw)@a+B)nW*h1|r zdKOq(*XOj_tsEAL)j&6lMM?GYd{V<87Lb`7??6*`hn)8+4O$-hrCh5<4H^4-BfL%9 zb7$FZbB+V^Gi@inbNXlI&WD9nyBg@vn&k#^c)fyhEHrXHrq@O!f<`wo0B6uo{=W<^ z-x+Y+?Mi;nd*ok?n_{tegvBn+qGOXwMOjntdYq2fD?2D_R<;NV1-h7?D}4kD0epu4p`x|3X6>iIxeJEhh5Jd#%@NulY_{~BW$dzTx^yW(A`>VrhFh`Jp?)B&_imLA0suYd}yH8s$LAp7b4>YH;mBCykOMPU$*QS`5(O15NR|( z{)<=dA&2-ab&k7;lglP{^d!ZVQ{*@ffSz} zS$GL=!b^p0#xe(2Gc~YYBbC{|i|r(?c*k9VbpubSj%XAxMpd;u+Ht z-_<118Ln3(>!JE;iR&E^j)lR(jACT*WQ^_3% zF4&uQGpPKkUR%+!N30 zRXPn@bA4&lZW`szFi)N9dBC?B)?Vll5ogL`g_`4mu?*H-A+mXkx~kRIqOwzIIcV=(m`3hHq2fo|PT?B(lxBRJnB89TFF`(D`amT~q_RT-YeO z7o4oR%AD}(rgP{L$t5li7=}tDS^ih&?g%VZ9s+*&oS>VsPC*_MA9{(q#37e+mum;T z?Qd_&@)*NiOQ+RJ`kXF)9nX|Qvx4F8^f_S-K%;2$H}L#Eq2Vk~nDK*(*`mYiEkD%o zGKSP*wr5$85mOsINq__hV$%Sf>z4uV*Z>9eX>X|d#~zn+gu{)Hq>{g}qP z&#SAz+d43%T@=qI`sGKZlYNmH@@CMn=4nh`Ff(}BMyRpkFY&vDA8Nkddhe6(8&Q)? z^?sY2n?OD{nVcIGS5J{TD*8GIW!6RAU>;1|G96n$8b$4%>1wS&3qj%K!cui9T{yj2 zzIoaj_?gFs7_Q}bJR%P!R@0@VW%eQGQgtje2h=b(B6{>^les6r%26D#lC4tW4}6K@ z3;HZUXW8)daZ;)tcp5^}uv;pEM*B>r-E)Z|JG7WeuK zPPvWdk;e*R-F)(!M~BVU7sjgpW}FdV>Rp8&lc!D&`&mJ!V(@~tmf}`XWI2{UuJy#5NfMi~Yxge+o-OW;Z6+sfID#cTO6c5=; z@|X_c3g&>aM}{t9c6guD?yv(&e8n_)cKC79PS-_MfnmaX3I%vA+rOE%gF5fPH^^&N zhAv<9ZNe z*shQs*CGf$HPfYPq+x;v%|~Rh8xjnJrB1;fnNHx8*6Cm4*+XPy1Paj_`YPdW%$gu^ zaK97QS{pu_;Zr3~T(Ad9=q`j@A=g9VSwrgf7-SeV7&PDdHw(qp3M@ZDX}EqT!$-qv z?i9~^?76@T5cf)_p%5Fx;z{&>jP{2y#BaqWr+ED^#{J}-|A`xK#NU$T!R4giftRlx zCd=14id#dGl~nY$Dg85xeRjK*1Zv7bzZz89BFA#wdS=>KCc{Z#(=oqZhu_3Wcp~o3= zI^q$r3(Ty6(TgmFJ%<2uLNV>XLaMEuuO_n)TtGY(7B(q-4*f%o<@TxkgGtistn@{l(bHlcvV> z7B&`Lk!W_iU7#VMl|=7=?x2iGz_VsXQELOF*#cKz#og$U_U3Hr&S=o%z^2FodP?Pe z%C?!ugVsh=kh7|6L9HairB7KGg|Aqord1uEqGP+k*}tLIv1x91U27DzB+aeStz2Fg z)u}x1b$rTW!Q+`L**cXDNM~_VP{W)ccm>b4)BB-K>;U*$Hh_(Bf^CeR(Zj|#!haTi zSNpF<+Y^w|0F`(z*>VNh`9VSE+Y|@M-9{?ss&BLH-6+1m;o`}{s2 zf|RWDD|UILT<4b`j{dt=RTtGrpOc+~ya?vXCs@ypu} z_w`zHjTbbm8PG=eScb;mKm6|xT?ag%cfWOktmfwdIc|{vDEkbyPl*(_m4f++z71W+ zx5IF&TrF;AGX=0qhlbIa$dTOYrKwinl?7M;vQK&Lbvt)$3`{RC+!)Nmz@v73{Tg+* zH`s`d+uyyHM>1a-bX1z4qln@PD3XtfuFq7xZZ)bJ@NPQjV?h|!)E9+eROO=YI>9nY z7BINobKe0hkpP`)s5@#?Uvl6EL|-79zu zjFJWqdDta}&!mn3JHsq6w(EJ^*!li9R7UJXo%zX8lJb)IH^)p6l1p(p6!76hH;M{@ zWb02&?7n#q!8y4|Zx&<<8rZ9I_quOWR|)RTT<^c%Gd{FR)F>%uvLJU3{L^%?OGeF)cm^IDA>-M2wmPu98vSPSFAkdLAU0FP z<#*A){2=gDL191nzPrguc7hh|JG zhT3Cf27`z!mTtuwX~L=$Ewejh@u2@yFW0drW#w?2-Mp94Lt~_0+wjj>hQ?3@iY0>2 z-<>$HJpJeOKarIah=oqwW(beNY#2BekImM!3R0vQ5TkFV@jEYL@G97RJ25&B^&ic1 z3WgfOF6a17fy1VZ!*vf=n>QSPb3x~v7NzCQMbxJcUUqhtHn#{{~m$r=v~g6LU0RITbRd9riPSa4M9CyF<622ZaY zj|2K5WOIxwjO)Xb@ubj!v)q?V9M(#TJ3)~$4Eblvn>>(g3u$dW73PF#*1MJX83?%y zZ7=;!F_EcL64_>-4EnTkXA}_NV6#jqy*?r~Q`Ud{d(kGGbs$`HQF zxU>J7d&$Fyr25xue@iy;+e$mIrO7kFT`I)^C9W1XO1S6gl2?fyL-{ytn6{GG$O`)L zo7kh3L7&&BG1Jv4f_4|^=5e{oVKoGH+?4qqvpXg6h(W~zAS1bT)_e3XUo*nQ|AXTB zWb077W(SVCK^SFFR$w>9rBP%j6@6h&FW3Tu7}+UbPx%eM~rTj3TI8rY;iJgh48itqE(V z6aBt=ZMR#MpeF3L0?U={2bj^I#h!q%jF$E(XOGS@qUGloXYVDs4(!RCH$g@j#X)Jz z5h^-XTCG4zB@B+-bB}?Z6O65vDv$VUu7_wr(+T+?HCfV3HBy!$H3yoOGu%oLF901= ztP$>aD)W8nl}opVE%84GqKRuoYej~#;Z+l%HJtt;3|oP*N&>Qh-b=kuOwz#VG^$w4 zDDm$jV@UoR{f;cV3LB;u2Dms}<3{7=S zU~mbl38@dkO3`elX=eXSRQM?Yz!-?nuLkGSy`oV;gwG`MhZCRCM}BX^hJXJ4bR$6C zYRFtevK-jRoHUv8!xRTns|Tp)HducC(8R`27~QW04b`RGEm?}-qUyYKDX<@HVesg~ zOsX?{5B1@ts6MA^x-;BxrJ5c%Kk#}fsyJkYb1!U~ItA5qmwe#9F8T92M!!n&hrT{XWGaC)cLrT0Z*jrw z)on#4@Z}VEVc)?ES34WQ$tqiAdeiWTvka%9oSrq`jnvAF5L5o;Yc*t@149fd#s_Vw zk|=IFMYds|T$~Wyl_plR*=2>ePr1}9cM^i5*arhJnthR3($#MIFkhCmS^mTwZ??UC z`O3763?W|`#^xt@A!OVqVScA)8WED>m%5qc@NeSFyNp>i+j@%OEPfwA4xLFW4=ek!Rqnp+5-$6eBmYJM+29GyEeoCAI9!^~pQgog7 z%7owh4QY({sc^O1IrXO5%VrgSFGsmQuqEoCY>{7a$SqOaoC`j{oT@~!k|es{-==`E z5(}nAz`{G*pujK+uLX&Uxi(c|PBQw6*b@u%qZqKmf~!o~f@M*wqVWIVw|25?L2{~< zUdv%EFPy;rRS&@%@&C)k_t?K$V)hj^?~WH3offA6r$uC`1G}}mOx%!oirYYubwHE_ zRLEFQyU!y_j7deTdpRzxh3Z?RNw%qBu;RyTwe7HCEL+~f%XW-gKpze<&r>^W38sb2 zGiHTif}4G|J8yTt7F4F%A}tQd4v!Ic`*!;-5?0dNxSgcTHxrtUGG!TbuWN$GPIZz` zMc{#mi=K8zjnyzQTUKldiJ`{g1rv(8-j(xfqxtwv_0Ok~&mGu&tTLg0ctUYsQly)T zPJ>qJqd{00qgCaEtyI)VGp5C}OM}V-4WW*VY1f$Ixy|x8h{3I(@NR=Ar#xQO@8_()kEIRFXDuI0u9_{m}7A+F@x+P}A4`V%5AzJ8IOO|3;P6NsC-A$kN*bva~ zbdap&8r`Y_8zDz!Xg4?p-A|hC*ONoj)Vo8|^($ZDbRAn30e=jW7q15U!}#BQ5*qAe zgkScTzxo4-abWmqO;)csid##ORaA5hklh{wfyH*`)6fQ>RUHjF;A?m5GvqJd3d;YX zp0oEkUKVNm2kPGK=EdNAoL&b`PFkpGKd5Y{7l~VDXH(cta8i{F3J`{DHCE1O&b{7D zuW%X5l6Icdd-lf9^YNSZJZ?1a)OT2?#zOBsC{005pg%Asa%X72Q-N13G?(GO*GA|g z^w`wgOQ+5)gN33&(!!pOXrr+eyV&J6lN{P2s3X0ywV^-(cuZL6#UKrPZ{=L5FXr#W5nd|fH?Th}NwbyxBmT@7Kfk(y0#pbv96F(+vCIBzX z;6VF!ira=!+~_K0zGSm=rV4+W*qxj*IRgX@4#*na@?fRU7OrqfU{(N&bDS_=I%c+> z?<`pREspj+UbEo1=J~DfpV-J5;o!y|*@QYn%`MPp}Io%M8(PNy|?Lk?f0 zJo{l94@Qd!OV7U|NO|{X<6_NIi54<;xWCvWEt>O%WHp-?1f=gkmjW9H%{77*_SH~o zwA<|(gBfBh@5goj*lfGOV$igV2qA->x7Xn-f{@LxAF42CwR=UJSPL9@%}|wK5L*E= zs2wvLe+7ru0Q)NsjlqeJf75(ZVO(8KfBWxlkPSmsc{^|h5D3Hvl_cz>xFm{fr=m~F zkfH@8aDa0o10pLJU%_gYwwcZHoBHO(?W#7({h1F%N%Td}4!VMdRLX={yEqVz3}t(pDPdsX*HTFVC9%W2Gn^wa#xq4az{O zph0q<{IFJg6gct5%Mf8XqsjfeV@}w5J{OT;jjReV^c%=P}Z* zj0F<7ojz@IPl-C9My`u|DJm4yMU{&3JW)s?$){a;Jg6pW8&&hIvVU9h)}^Qm^g;jI z_Tf|=Fc_l&*1#C{c)hyeYt2fdX-S&&{;x^=1fWqG{0mN_xSbS9qM~ad%KgVC_Jp1` zr#!raY#=M3{JG!1+IhccH?zsFF6#PQAGf`=!*h{PH!D|JN#hOu{-?xeNMT@{9SCv; ze<7pzf(H*{^J8AWT*v+4Ddd6kM0!T7#Q$EaJvZ`~V#g?sBRXy>7RF;74DW((BerA0&E> zNkdw(kw*6SVjulxAGIbi-is6sAej?dCclOep?LP1M;VlL-&P!kacW7f;1Q`K`km7< zer!Rb!26JE*6=5-|5@n*Fr61K3yDwh|z5FYoi+kTjzkNx_^h~ zvXC5!{r5U%6JsqZ%pSA*TVALbzh!#%Pv0^cn^_-SY$U}FY;5XHjLlh!J57;FD*6h! z$}RNR4{K%&)x`ekp>kE^K~Q;G1aCfhsB8|qqr5Xit6DnYA$=`qY0zz1vdhtcW0Q;6 zJ2Q?6&wDMK0AdkPmOX#FN8ap>6R*2qXGaTr94Z-hGR0x1JhG?aBl9bz9rW`N3_i|w z)f3smfLvxhNJNzc?&RO*|*opuwKLiv-=a- zA3n~X=gjT9aum~mc@@Pfy< z?|t*fNiREV;a)H_M`bC}wNZ_dgR-vhUFrkMha%|DfNogB`8f9!!LnHgebE%OM77Sm zKPg7MM%YVtg*PitFfkwxd4E!u7nVVd#86l@JtKw7fc>!QkBs0ZUT~p0CzW(gFs@Ef z&A-)|d5GFajdq z_neZn^6Ncy;KbolldA1U6xT(O2UK*S{E1wr5A>(fivu($%vS^TG3iiC3$cG>UcODI z&}o9WsT!mkt04OX-2tBp>zHlM3k?12%|Onc0Xnn?d{gCc?Z1fNOn zOxH1*KlC~M5%t#FU`T*lY@=b*c}UsrYtv4Our&Y0t+eYvjU_=-9@%YI~2z}fxu5g@ZX)lnAc_rHA3(YouAOV^#;*%_APLKqCdo!1MNcUH4x7`8SQ` z>Bo02B$C4p?CaKr6(Ubpg~ZBrIpBc-BBhz%bJ z%oUU>mvEXrKsmHKv?OGcUz|{bd>g0)T0-v%tdw>t*N?2LZKUwvV@*Z^AmXz8Um`@HT~J)8t*?nN*^xE9S3TXT~zv zqpDSP%uQzdlo?EUP&VBSLgF9=3K{K28uwrJvqx6i9ZQ~yeT)u_8dmd08*6k%wyY1j zLN>oL8}9=qu1h+_0sTTU6}?orPZ=lN8wMRJ1zwx{H26-dTEbj+!R~}3svL>t4xQ(X z9SSSRGQVC|C}k*;-D9EKEaX8rHm=yUZgX@%u^Udt{khfuA7&VV^4sF5GLk)10k`8; z?hwdF5AtM>QXKGh7f{hXk`~vhkOy=!eNwfMX>qBgkBgR080ze@38=`5ES7scJ{2q! z-&fSQ>)hKUyXiu~V-T9|jjWUh@cZO&%qWpp6< zs*$l3Y%F^*M~+U4GVdAUVg{3_G5!E@lwjm*#9w4x2l&z~A(x@}?}& zZ~i*JQgx~FXeihVK~-QX#7>WjZpvz+(tmJMb}6bZDv#Ohzg526t4dNzwoNIuZ^^A; zenv3sU?yxex0c7TY%zuiolqVVb!W|WjQCg)2Tl`MP(4DuR*+{^U`Id_y@F&(2W1H| zxy`

    HRKH^ireBlJ>f_sV;|Tnmk*auc)-sQ`M{8fPq(8FejdkmBs`gorN*IVxLvw zY&qmEVGYyV_3Ts@N_Ti_%BGf0y-KbUP5$)!>B}cBpQvefz7m`_1LJ?t*2|38)(=%G^ zEjg!tB14^t)!?0s&9XCiGOPr};77*fSnBriMQ0-@D&HvmGg&f}kL$oj1UeK3`M8@X zZaqa}LBa|oe6^~QkS0$ZY;!u~TB88%4eEt|uBa38`QvkZhGT#|7s!Yv*)qwMfV2OEjnD z!0K>pa<3bTgJT&1_Iixf%EA*#W8v58@flV~p1fGR*MVJ$ToYrmi{iky*g-`n$SVW3 zGDU%H5P9!)jU!7!i(Qt4<_8)S!=a4_1*(wH15?EM(0r3;*DMUbj@Xa1`uENo3j>~| z)qi1RH(ItTTm3pFnTJMRDv4(c0gx)^U9wa)6xbpPQs<`5)xEBJ9cm8H_spR8i1!JS znQn4Uy#sJ_&=JKZ}X?^_ulgNJ=8kyu(hbA zzpl+9M(;GwDfui}&Tl8{z>!tZTpi@gZlSnM6j=|;Prw8Es3d`Df;y%@H?eVr{g3d6P!&Z}Sk`rF+)f(LA z^dPuOd1q2mKrFLAU@LSSmx6NHUZNu;3@{D>P8VtxgSvW09WxzoMZWeg_ za=SSN51#S_Dt9CSHLgJeRyGd=G#yOe%d+yhmOa zbxYR9Vxhnm>89}G5Lr$3-0Qwre1^&Ny&QOf&ScJe*~{%3HD2ri78rAe$NnjA7)?z5 z-`+_l2OKzbcEQAGoS?Weij+{%t#pY93C>$)BZFI?a(zU8_(q^QNTW49KwQ5uq>av; zdje`m`jm&9vC}+Ny~cgZ)RTdyl}lxdL7z_36<+1sM&l_>zEbO#H!WG13AND3lc81R zONd1M-2=UlVbYj*&i9f0YB%P`?t=! z0Y(^AzP+J{r1D#-95}>s+yrz7C=OySIaKrs*{(@=6JR?>huVcYM{;-}~TrCJ=yrWbhqavk@F_nUurdB=R){MD$N<9>c(VRJkH;<72&tOE=f zAO$mZV+b_wVfyEWY-33L#4#$5*29h;jBSe4VF$Ae<8j{(a30OJ$bm6qVOxY0V0&lh z1sBcP3jLh=qWU(va^5$6t966jwE`aSYJHb1oS&A*+jU6`>IQG za>5Le|1ZLxfGm)$7a5#Qeqb{7RtCKr78oPB_|8dT&OD=k@}2U3+E0%1^G_UjTXfaL zKlzN}K-Z&!iar!v?9-q=C_AkzF)}Y{RoZt}x_7h1F5QyUDJ7DFgn>;Ig^JQnqof+r~cPh_&#fIQMyw52=upqF{sYbCXszGu{mFUsubSomsAI}_^ z(hl@32~e%S-MJaGK=4Tvk1CwA%dOZYhCCv>+(x`0+XzmBkY#!}5-8y_-e-Pa_}A5w zTqeBW(;#RNo%bpORpq775y5WZnvkngbdvKET(W5f?2ss`ORqI@AO)r)IYk|@UnKM#BiCL2c z;)FM41zxxnwudh=_^TiC*gBjH`KVn_TE@xsOFt3++UUl-tZJCoB&lK82KoiWJ&D|p2Z-~dg+p2aK#&G)JDTZ z#M%fP-f#o%nN6UU-UViJaE1HE;TOHd*9}a_!p*HFHYWl=a~dbM z(ACLRetM~HgN%$lt8qtQRi4`y!_g=ZZ23{k*q-+JM~ltj;;?Is1!y42I+VZd17VVl z&MSg+%H^|`dR5Lo!_|agJzNjO;o?K9M9Jdw;xZ;(a8TApA17tPWO0fb?`-m{fw0Pc z!8PU)iL+Zv>ZqY%d3Hy2#WFO8@KmXdnd1}w!?;oU@A}4 z=-$Xp0^m>I3zLo&*^9&p`odrXImW;S2179MEbdh5ye?rjX_p|LMRj1!SyikwiM}br z>g^;c=9$-WKrrW$AukqY&@o~S@~bpMy&VuHKyQd)y914?gP%Z4gm$P}RHeKnYsU>w zjiPfVDvXq>%iWL7-X6NbrQZn&>hv|_a64{=;I>?*o@=D%A7k3wG-wU6lPyMkQc&6)=P9I_UzWjv8y6RwVDurybx`CB zDq2stJQ0)?j|bj1k~?8wu!`<=MY$1x(_?~{>g8AGKAoA%=;=mxQN<8aT*0Mz#W7eY}67j5$TI8K2s%l z>||EKDEjNE>g3Sdz&>>Zp4X~U)h*6hE(`GX4Q@|l3sNRGig0}W=*V}M0sO3o)eEW> zYuxcebtn?#@CgeA2PF9EkYOf<`e^yzta&Sbq9%@U*j;?C@=k!jh&!hMr$uC`16w;F z`59FC8BcK=D6)=&;0H!ryuG*Fo7{b6?9mDnd(=yDk0{bbMW1I{W@B+)2ByRz=Z#{Y*xD8oj6e1= zrX�Rx8%}Ee=MX8yjnrq5q~+hCxG+!uDGgb-{J1*JdBA>{9FM7R?y=bntn! z@;G6m>aMC2RM+d}jebc1hJQLDTGht`s}&%aqJq3BjEg!d8Nh&sH1<_=k32uTlKE7P z>3tLF3#*|z$|cdd$DjWWDyEe07aGAgZQJ)xl4TBTlu}KM(q@X=NRc>$`~;F3F7*N` z!+4i{N@$h?mZ4K7hj6k4~1QJ?uFGkONY@^JK8W0?#W zWcoumIksOG4wt=?QH*G%7ND*N^j;-HYM#K=|pz+BWKmMb6hKf&Dl;eic ziG_R?=ARGvLV3=yphGg0|EQO54m}vC*CfX>9L>gwZ4%sZWAd|WRt4!(4hqNEdbXaA zH!?guf8Jh4=ig1dYu<_Auy({ki$p2CiZs&)rMM*IGD-ebAjOFNNt)}jtZCPY;qu`y zyW5QCbHs19pFd~sBVN{H+}z$({~K=fPu?rslT7mYZKEC7dQ_WuE}v3dDMgB@Xxzf; z6InTLb&6Ka)?hd-jb0g{!A2!0u$hq;5C{9xekW8O!B81SgAE-{C&^I|T&qzW3O?oh zP>By}uv`YSxet|D(j%~e&2rH+$U12q8_P6`+9cp0HTz^q(GS8xF#KmU5aowCKHd!6 zI|^Wan)Jg%KQf|d^Pj%{HFB9BiX6BE9R#xm`M)g`*F=$9RCFHXJr{~Ie40h)VR->h zMwsDqM492U7{VzxMeXcL_e0P@3FUv1PBu330ys5wQ3L0DASIgNb62uI;)rs$+n$KJ zsNTqB!d-$Y5$e|9lgCASBFe)%2o|XCotolyLYU!#e*tTfb37J0H?XVS&Ow_P-oI6T zZ{}_~M_J@i7o}qhlxrgQN8spZ$T~mh8x|x>AJ98I@yu<-2^lH}_PQ^Bkpzxifi^5I zhezR0-(Q~@Vg%aC_>TQ#uLB2O&X_=_nBpLJT1Z9jaBo#FW@_B)qDmxPQ*cMLEb4;u zUG^rh19VMU$Se)&_fMnK1qt#R_vI5XOL~9O%7~}_Se{<0J_Ax`Cn2lV=B@KZt?5eo zNo1S%(g`W9oh)irW7xk@gilZcx#kY$j6)@!ZXmlcvN3XX!m)O|^TR@IC}sQznDVmRDQA z(jb&R|8nO3(9_CelObt&T8ZgYtOmu{@NHVtP8J6o3dV|M9jo7ug?G7vcLnqmy8;&h zQFtY6#V&@O_b(3X1Q$Dr))&ObO7X9vSG`iPSI`DKdNe;6#!7ZZRj;lo_;0PMRlRK1 zZhGNsy>5`{N)O%ThE<52ES5#L4a@RqbC?ZlY|JhW14CXg8`pK+<%+KnW>2{@D~Z;D zVRqC6W;qnMha$Uyw^h_Rw^)dQVeprx6${hAskt+~A`sb}lbyAy_1re~8kYnnUEKg$ zeFZMNnPktaOobSmrj}3mT(M$;R<%i*0#theCM~$ zzcB)4#md+@q>Y~g<-l(EYLjf%Q;O@Q$Ri9O?hUJ#uXfXyY8Yftkir7Rk1t5F#Qje3 z?6#mP(E;D`AWUA}3&!I#HFAif)D9Z4BAygOj?yb zh6ub%ni9!6KdqpW0dG7kequJ2O`TLF3o-tl5O7okm1`)uz8I9jFh7PvmwRhs#K&HX z6QPa1dE!Pl;39bx=n(&|j16oX2R8 zZ3ByZN&^mn0z34}pn_-vYp6)mv5(=Wo+Mor1*%L~Vw0-Yu|;&gui;)Os1fIb?~^d+ zBGmG%WXpW}omMNhx|Jethb@RZM!(dVpu_mae0KT2=e5+0E2@clV9tHQ1BC3H*L_5~ZxEr$D#itx$PKu^$3 z{FP@@ zG*Xy68^BHn{9qFBgTH^@+X$29xifE)M1FfR2lk0UeQ=Odw1?t=&^rzKd*qw^+UWJ+ zSmJz$*($#%DRzm4{S@X|p=lV4o>NIC?7>Rt4Eb%?;~>T$3kwDJx0yv^9eW?D_!q$c zZZoA}_zn(j4YrYD>noPW#u$4nXTJ8f5iL&yDL*D9FIl0_H4{vnqc{j(pTsZ_)@jB< z-vdVa(oxbDBZ1Wl+#jchw$Z6{gBpT-@obzBYtXkb%|7u^UfwpdS&kK>&GaLp!Oa%# z+#Bg$H{64-<*@Ll*DXzrqu|cCQs1s&&~QLxc_fI;zjR|$@JLJSw;HEoKzDC9xuR5p-XYz8b$`O}Z7FVG8_x+r`C+aFTZHS#>q zhl-s{8+~+E54k|M|J6n5cmbJu+Ot3RZ;kHD_|E_MoRmB8o(M`X2Dz`7DDDD9&QsBg z{8FSlqwY^ilFlD=SQk}EKb@8SX55?UfeCN!jykEDKff3Bw%1LXpXIMr?Gs>^XeMl& zD(PPOg3k$Ax%X19xHq9F1G*X2o%yklS}ym_79mtIZuF7SO;ver7pRvXB{ z4#;SZtu(OwaMdLD`#bacGKaN87AnRHf**md&4urkzmxp66%$T+ZSso=?wfMMvyx6` z&SHL3`Uu+FmjIDm8{0Y4a9_r>c4y3Aoe=KytCBy8SUTaB$2QeLa!MS_ESrFG4)q~< z!8q#Ww?4#wZJ6Kk_n*DBVgicZ7RqnCmIoEMJQ{r`yuDClC0H!p#t%iC^Zrw6o}O}8 z6j?|(^|Hl5T2(q-G$&PkRnXyz<%Vg3Gh~O)v48^kTJE@ey`oUrs%#de(Hldy`{`z# zRGs0lrmNVe-Fg0A*^RIY@%;H2ffc^~tE( z1CL2^BuL-|W$wr-S1-pVF`(9u6E@RX($u+WG=}|_FSOu?Dbxy%NlwANbdbo_L=+q*rc)?{n`M~+gYevt<|AXTBWa}#v<|{DqlXg=a zB&c>$(fXPNL#PEyE3jv)-{~Y>5)#kq7}Ou?^2zsw>=f2&^*iCI_!&q~h(*zc2e8Y+ zE|4F9uE5f`Jr@J5YRW6njuCw1w-WzH6u3vNT9O)QfWRyh3#6geY{Ps-$0Q4;+` zR_%S;Zd)o#6J<3nEDd0%^IP7UKGB>4&SBjX3*2<2To=7;)+PaNvYTmS0)R!uj=yn> zQ(@I4TYTDn7x=9z6Pj=QkJz}X{5F5$$7Ibbv#Mm7tSZ|nZX4RQ=tbhSQ^6U*4AOv9 zm5zN1$_)AnKSQ}p#g}Sq2NbzaMPtt#W^6v4w9^lEE}iUJu0s|dTHu8|Rai3%Jy8(PZKjW?v!sg_9`UF8fZblV!R|eCNegp-D9ms8;vB{ zjUwdgYNatlogZE=UlN(-XJFGp7Hn)~z&~H4Z(+c(J{Rd&OoreGO7t!c0|ys+Z(1P* ziSF$_Wv}eIi)L3S4r(CoTJ>5nhm-}T>!Ec(GPN#tdS`-1UMtl&Hh9A~T#c|RdiPrw z$m*dI*ADEu0ypWPXmBFMZKYslqII+S{S)M;)g5%rzu%O_P)+PY>Y);7DD}1E&=VK9 zO;sk}Do+bAomG>s(#CIiBF4sl-STjj5#uyBI`C!J&^#X4O|N62PTxQrpeggdCP@Q^ z65#QJs?4L%CSV}XzpXe7)l^Hp4zlS2Em)~TuBni6*2FSpzS~H?F9?`LARDA+op2k$ z3{*2+C_u43VCMtDOnpz*71HBsQ1CnJqQS`B_R#gxQLS=3A;Z(YScb%lu2_c5P4oaPKy4j@bgfVTNIDW1wc`RLp?@j|~-nz{(q z!(MioYon;gwK@dlh75UJMOR%9}{0sEAFN&{Ew5=ESFP+QFz zveLhi#?LwMDK!Oi20l=~6Us=Z(NMsZ%_J}@#BsuWsiCl|QM5G3VH;L>!Ec<~H+qUc zG_G;~&~%=VPJX*F2VVWwn>b>N$8#~`$x|vinOgc*>f8CA>EJjO$R7zC>2;C&1<1cs z7qy$-HvhwuZ{fu=+*JWy#u~2+^aDUz2ffL!Lj4fjv+bdq{OY0(l2TDTi+DQ~oGGZ4 zK;G|?ENlxzc2Vub%J3RRV`QCjncoo={Fy=bpu|Y8JF2Obit;?q ztE&{7rCkC5tZ@{r3VpFwC|=F;)Ujt(-2^3LwW>@eFL;r#ChC4bHT+~6-OA0}ef}d==$8FhufsRd(J1l13m{-HP7*Y6%=JM~z3VtYb;3mIx6CZ3V z#l=%(0~HOeQ?NdkFuktXf(|+(FcUh7VOxMDgsJMo0T^Q%)#CV~Md~Zpd99{jdD?y_ ze&G7{Yvxs4UoG}?U;tPsIqMYYR6C&ke1(e!X;cy;7l}dKC;;L_TG0ye7zKd_rj5ts z4?Hp^ua?)cz?O`H2&2W=lA%l{1r8h%s4=kt6%_X=MM|mY>+CAvSI?l^ovRfGfv3IK z4YjeVXzaDr;6<(K3@KFOE%B_O0du!FCbZb4)AzAllMH6Rkln68e_N}@GGug&62;9l z5`W?2*yDi3qjjq55>0$4mXG4iNpy#2u1ghoVWYD!JV3+K_lDa?0WZs-8Os0u`yVxb zU``U}uo(&qv~Np=*qeD@Q3NIDAOifUAVqNVJ+11ba-l~SbmO8xs|G93Q`H#zL1Lz3 z!m-%ej|d^7n0ad-<%N)OGo~;%%t_+-M5G;fZL&lu7nC_gtEvdAQvzuh7E&ZZFL$aM z>8w?XKKP%d(}ndLJT=$S^~2>V4*P4JIN=Gg5Y7$AqZwKpRmq&+{-3#D<-rl zO6gPXN0iU+ST!+;KJRr+VjmRoM9fet)Tq%h)Brc%4c+!C29*zzWx`8yY9aoT=(TOK zPFY0d1m0km33X6>zCG|zP!ZMSxkh-&qiPK07I|V|sOhrHIm_!$NB_sGUv^?+oo6oH zL9dF+kJ{>0tvV98)ca__vB}qE+2JLUlMx4*TpHgaL0?@|nJ@*a%&SCMf+dXBUPMTP z=8z{M2CTxM5$tnl84*MHtkjvD3tyU7TRE)HYN39rY;KMOTHGM#1Yuj?4xw+)S|jM6 z)d_*y72+!Fm9@{hF!cQK!o|=-*ySuQT#WmL5B)pN=#98O?!HR4yfWU%K@)EzgW^ED zBZZ3o)GK{j7qmO+AjfoybjV8SYVQtG26=@|leP)59m;U6JiM0KCd_0``CpM_1hxfq z1>`^lFftt37iGL~HDGNFuXe%Hz@RPtN{OElDcSu$EFj6R3{px=kg}iRVAGsMML!m# zNQ;rhm^}x6$+e(H(HC#e?|J*CEZMWmD-OgMD}a|Ij;j*XgkdpfqX^Gq>2FO~JM`Gq z026Wn^taq%AI?alTSXw(GkK3)i1%8cS1PElWF*T}zd+L~Wt{Y^)o>z8hvZV2BFS|ngn55J7At>sZB()91_@U>y zeHiA*CaB_tp7F8C|JE>?^q&JqQ7okY?t&%wrvn%<#h4^l9fZ+F4PhSs-5g&;-@3m!WjPN{2|BPu=mCuZifk*lbE&NP}(>G5q zH}6q%*sXAZ2|k{6oo)EqO&J8q=By%J@t z)I}B1vCLW31J^j=-pQY-bD-pI)mt}Zi=cwjuEv5tRGzOd&*s$*2fUy{o%{3t56$y9 zU#%o?;Qj#%*&6ICX%UprO>9Ec8E)64`?Ip7aV~K#AZOu*wcJQ1G(5Gut*3?;UTpox zhZ)Z@UjF!}<&ms$nc1>F=nC09RH?TEXYvl1n3i;kgRomN6^%;oT2*Dh;()}+CERUU zwU?$@xh+x$f%zp0%^vYi?luD`@whq{^%Ary)E8XpbwdUGWMyHSsK!cg8zB2>Vdb6NENs}gCBWQ$XB|R%cwW19~RIu+UlSzgflccc> zlsQ%_un6UHh=B;P$#d|?;;$&&; zmn(9^*AvVI=VF(IA$z=&xekc1cM2{klKgTc?X$YWcLb&hVwud~#gQkWp1Dzk0n$w7 zFTH}p^Y-{3dP}RSQk)OG64or&&OYdCjSaEs+zgtgVGpgCLtejs$Bjd{o^Sg14~!6d zdvCcnxy#QmIqrWMSz%J<-%D|iDAGkmSCGZQm%TrSUkX;}7yGXV>(e*20gC+Z2`k(? zmB(i{iVC4na;w({ay7Wm=^P00ZsAht9@#>5IutxThOO3R@}(?B+~>4$;?b}&dTYo9 zx*6*9=O@vn-m7bDbsQ>;!;vcCrRs~@KctO zM87=GUO^>|@8EYE{-v-8J^;|1lg+Ltye>TFn)pBY4Gr@H8~Az6KjlZgxP;q*H(nM> zxOFTNLL+-`u}^_=C#R`_xklEQ7+{UD#^?=vLAH3cdh%`Y$9Rlk*?`XaPUUg)9Hhg# z(iXCcr4ez$_V1O=Yn_Q2?J0sbp!wGHvN;kgwFR1xkNazmVZ|8=EQH!R!UN;d%MHUr2n-Nq_m&a69ngWPu_W6$`Lz+Zg15AP|r^lm{Vk zo(7jKB%i34U!A*GpuYpDAc|B5VmIAAO~f4(kKqxf+Q;Xi#NV_qbj6%hUR!+ETiY?fDEr@sr%EJz{$)k zc59=L1)ZcTA-J|SRI^=mHh5=LzEXoQu!f@$SG{!WL&VN^6r8Gx!_w|dQmu`_d?&6b;;p~Mz?B#eO8obkQw5@ zYIG0rtUXV0y`Rqhm2tHR`mfhsBQ5;gS_g*35|ijmH^qTETssw=$zW&5{6+E@Uyxjz zkHL@mNdd=YnFc z!7v-ZPH*p@FVem2#o+bwHuj4AxkUqk^d2%GrKZ8;WTU82kS{r@w1q^mRTIH`x~&>@ z^f&P`YvUBnt3EVuLvq-_u7$=T!%B~72>#Op~3ea~x0 zRDx)sc%k^7R}#F_U=Dn>8zu$p>fiE4iitlldPGwD?13-~`|nxgpaWO; zUNl)b$|>$RMUGL?2Ab6sOg^oTts&1{GmT{tI`)QtqADg*^Td5`Sg-3s@dFTo+#9Cp zChc?zI8K^1AR!vd7*y7EEL=&af%v5eC_MD{ol`%Nt#$*SRI9==J`^L+$kCjTaHG9H ze*-7yyyt|;`_J<}@3Z{>N*&!lw`V5s|A@dn2OX4FaWf>fvASzDa2?iiie_g{9*Zkm zFg+6i4!rn_qaW40Vm33&R)!sekQzIDxorHA?6NRQ8z{v(imayM^Fs1Lx*V+#vOeR^ z%|7!m(MhlY5fgn2yY63_w_+#9jEG;K=&vJ7U3lREUGp(pmMxTGBSqGOCN+Ij1;v|s zx-4q93S$r>w7P2q`4Xpk9^-!AHq5hWc^LO@=bdErHJ?)V6aADIeOGhM=c@P^cOTJl zugeojKX7cMMD|1SX)A1Ei`5;Ha{hY7Bk!vrlV}fY2FD0QoCgN$C$joOA8Kq?4~~vIV(AA zJzw~m9hLx4<2}R)=PDX|f3GwzC~@^aIZ85JctNSKnD0VLQ9zMA=#B>}ZP;U=Vtto% zP_9*C_0o3oo~fwqv9B(0ghdb;z<`r`WK5Q-3gYLPGZMmTc=_Bx`Fkyr<6gV`VNt>y zW=bTMdEfV}nBm|E3TqTF?M)nv4^|dsO5(1$8?2RuUlNOQ-2!1xbcpjp3Zk$bEn}Jn z^=`DvJi(edc_As&7B9xUcb=e-8Y^7g$aTi++O~g=v+}Jty7)& z0$w!}CL=l%ypQ<}!r;Y=fdbnHej#VO4!Gy>Ng3{&BJfEmTmv*JTqf9$kT203R_%ko zWl7w}{)-nc_1e$P;9v>g$;f1Bx@bV6$>AJ`NeZumDm=zIJO`qX8K2L*GQJ_}`9+;b}+Wg@%Xdg+_S7K%`5Y#lpmdl@8Y_dDkk6Mj(2C& zad)W59`w2vSxIBXTb|$IWEsjL4G@&2$Wac3wu)Q9HFju|(KDvq7A#qQ-JkI=2VQdj zcj~ltvD?#)(JqVuhRVEdNixxaU?WmqC#y=Jh%s027?uLf(aFUZ82gTQ9We20 z66WC&@Hg45xu|u8TH~D_PTfbdN5fcbIB@!{9>r`~x+9n5OhlT`g%QLc%~vPxC*>nz z(;8fbS|mt;Jh{E}=Gn0gEml)G2`~NbeD`+~sc3WIT`@yN+eoi#JGX~GF@4}Fuo1`9 zAlXNtj#KPZdok{0+6D_S9OK^Yypx$U`ISA=9)n%z)wx1 z;?W>6lR=+RBrvfH3Jv}(N>OJ2l4IMBemdKZqo9chKL4VS>-4Z&vG6h{0VpKV8AaNjD>2KY`H?4OunJ)rGG0c8Nli~o4ku}T zgV&beGtee}NZ>fYvxc1iVIXG(=_&vE-DB@tuRtyvCu3M`(&>GQbzf{+SV`x_ zXq7$m8K_@f8rBX!waSN{n_|v-pAFgQoi)ABf3>_+f=2Lf@*dvJs2XmXWS76=5HbD~ z+KrR(ALp>MtZ+hkCw6?>TIA*F_=xLTNMf+CI!~lrj&Td=Jz^s#O|UdLMZRW!J*f@> z-fPd>5eAQ(nJqz;bdm_i7(9+~_lR3VddOwyxTq(`L`Vjsj~WD;^)2qJLXEP^p_|p& z+-@J|;K^5zfCvC zTorGMY4BVcmI%cMJy1ozm%b@2iy54EJRn~UqH}O2mzzkgjT}{Rs=!+H$y2#izAIl_ zB{)kPq4hCSq>Ic1SM6XBUNQRY^DYWo8vXf|Dy0Jv<^WiI4loC9uzxzE;fL02$S;(+ zVcdzamb-P1a1!U zG}Y3l|Mlv^B#r~k!?UKD9dgi0JnLzWyY`mys_?t!b?4`W|2{|>U3f2a&tfHMqZC&t z(gMoJs$!o#;v+%WWrf?Bzb0{-MmW2QIF~}VglD_sMUdZ!N%T4dHLiGekGP7{D(;5C zh8P|960qCh**=m9iJdG^o!S;_I@2`=$E=8J6W4G|XYNUpp{H?0T$fk3kHG`)Y7p+> z*3m`a=;Vc-b>BMY9O!Z6mC9^D z9(&Y`Szqivd_KCE?ag?B0PPX)_pND^+1MZ#j`%YWEI0Ct1Frj|s#gbPsnSJ@vs9r|H2P0!xHv!<`H_+s)b$mopN2#V3+llGe(D4JH?{;oB>m&@862Er~Bf@zLh5w}c!O;kY&Ad%6| z-yYh)&sxx{&f%gKgu&A&+pa1RpYqA$^vd`iij1;K(iE5Yr3VYIf4TFkZEv27b13E}474#2bR^~mRS_jG`udX= zoKElZsSe2zKaS9BnbzvDjZ+rVs~%;mapcGKyl2@s&Ch#`!>_ZlCsVGfU^i;ECqHW? z?~+YZ$$pD3VkV`~Qe+1eU%9Y=P7fI%%XlSTeV)fSbs-m~?|E(i{1!!X+|sbEoV{K} z(W|HRtM_mXGp+@d%zS?XXYkG4c6N8O`dDn7iZS0SJY@A2b`54V6|NiMD)`oqN~~+J zUDiio=m={g`q?RR+!KMvDog{aixh6g4A{ZowkC;MQ`*xKGO;U{YH=b)PEC~h)u83g9km!Y18o$_gyhs)!A?71YIB!uGt*syPA6)-_P5*s}B&)@Fe?bGgMlxtO|{W5`f0kfv#ctLC- zYGq;_^l}^&(~W-)FoW)~qJA~eP9;YzmaPIx0fG6wRQyI>mlT;vV7ppO8^TA5V4FBC z(MR0R6UYUR(B7Aav=Zj=;IecRKM&;pY9jpWE*$%1;9qZ)wn%z>)=D+U)H`6gL+n?A5V%pk zL*DJaO`$P(+!SYcd<5TpJ-sa$Fy1Ri^uv~lYgFn~IDWkX+DrD!kke;?)B^?TnA=+< z=7vRVGki|`&K3kZ2>;d=$CkH#=9r5!~`En9VvQ%ENW46m+?Y~(kD zCrb1oMp&fF#CPQS+3RC^x%u24IyYb~WbeA^cCY5RETR)%Q6W(*HuB_i%P1pBiZg^a z$L;4HomWeno&Z_Gpd`*QUd4=Y7H@k$Ia}w*`nxv%YkU8Sr|)AmN3Oe_dUert%G#p5 zBxy*7tU(88#~bA(GY7pjs5NhtBNuV5cbde3pnV!HYZOOJGAfI zI=h1$cH!huvxTR9l2X)Cqy{BhJA!-VN9Q4VSC$HEyfx*hMFxDNdb-=)D1S&JQ9(we zR=G*ttk^h9vnwV$5_LOqNQ(q{I)>fwDVSnvk*CND1MHP+ zvQ$8B$xEkK5$KghO6R2T(|)J@QatpME+{p(4_Y?DV|+_z+gZC%#LD!&vc9UjaUvwh zF5K?QK!S{nlP!|9UKs$0C4y{v1wls2wBY;j>NA;+23=`h*9bF5JM>6T8mAn4IN5~` z);Im-k8xkI7XM(g1i9|wN*FXumQO7R*%>k<+vUAtYF9*cXrf|C@Kt#R@4Vzwzs;mf zwNa!~b*gITXq78pd+ggB*H4-ixZ|mmbnr`klQ`=X1O#}+P5Ii-NwhY!)kF;N<$ScWNc zP^>;ddc?S~9(HSymxq^vR9%YP+|agvI#l4PF=}!oM5`*|T;U+4m_{$!&;3l+1;N?f zfkqCNLp4DV=Dgp5SBKq>dOrYpdm4II{Bs2f9;+xvSQ{2_v$QDBdSw^DEN|i;oYy+b z3?Mg=>JZUQB_COQ@&_qJ5d|5^c%4e;qYW&YtyP|gHYkx&8W=>8!8$pxO`0hJrgNiQ zr^*RJwhLr7ND9@_DCVtyZJAVK^hpdZN?K#`2<-J3 zE&qo`Upe#xAAP6;ezW?1yRI8}_ZxqTwB`wWy3xdi%?kr}SToeHq{wT;2LD?1@>vHw z@*VQ+XO9~j;LHA@4!DCEzWzUMNhM}WQskC#imYU}6uB2j z#&X&bGC^d`enPchAuJHKNRU(A5o?F7(`Ex^K3`32eW(M*-QPPfbs{`JF1*)a;Q2W& zuamqBYITt4#m2@Q?;=jNI3ujd+vwdtS$#%NCeCJ9Jh@|X4wT)$mj8VTFBnNl(4DYa z8XI1Z^0A%us<562Li`abJ~^-p$OFqks4!QAr4E|EUYB27c#~lM7g+{D zVyqorcJtxIia3M6jxLMUDqF(0c*EFq5uR#?{4%^S^`Q)U-iXFJ7_D-rd$xO*T+<}2 z4sGW@@aR=%yK6A<)h4_lhf+>XIlpg0fMbiH&zB(^jwS#&&%SSVO3fHL`Sw3vA)Cje z5_4fM^?(IZc2f!<8{A37XGE@sP>QC9v?v}%zyZ$hc2cYcRp3&{E4N6}>GUuS5@?$` zj*z=@74<;6%0au?XH1FxPFMS?hI7)gq}5G4M2|;Jhg?@7T>Z^L*I>59WtC@Cq|Le)9`i^vej3L);ZW zN}4M$%6Gco_Jht1(Rs2WT!ZyTcX)ST!CB|sB`t_rE!ESx0tb{A*)&<~;bGHbCp?H5 z9&;Nqe_|b(e@PNK3~~H+Iytb19^z@|47q(cztv;N?LMzS(CU#Md~0suWHcLC%%;tV zvAE0mFIv*R`)6x`2A9p)Fi0-!pPwD9Ro#d#jXf>a(;MA;XzVer4#A(P9;*cPAcoc@ zJ)qFj`q^8&KOvfv(2JT1watU_%bb#+8iF?_2sA~(A1g6BjN$hVey2=7d%sK{GVE5# zOPocyuf0cWHVB1@JXLP#CC)++m>-C6sBmu+mv8UJol2=8wQ%XBA;7x^JlcqNcFo( zs>xoc3-96W4AFCLLWFVW+}+|;Q=8%jzSjG#;5ew9NI&UzINT| z_Q-Bj*$xo9M&7=qJTwuUxbUWk0VhUIEo`5A<;cU|B{e`<_4-IntFWD~r*S9L!Pl&r zmQG_?6o&YdM78sK=Nmj~)vF+1n8>SlGE(fuiIdypwtT&8c)ody9JF+89{F&*EwT%T zV9#4@k?SZ0@T68#@k<1CbfxTUYy-cBTM0)uq960dJJ%G~Inu_ABQf~0Dl}_>Y24@c zHOGCdtQB43KZYjz?5Lcm%79{ShWlm^ZpQb}DiZ`}W4nDC`1oEm+$7F%vY*=}ygD($ zxd{N~xvP_17R7uM^K$i@@9dsqO+M$cE)N6I9F|M2SKN)zRC+!DiiAd~R=IuVa0IF` zYc_b)lYQJS&`xIIbxh==#(UYpV50A4`~A!?7{_aQ<@Lm+7px`j*mx~2yc#hG;wAAi zB2O*q3WjKG{=%-{9L|7ehWkyDD>_54(zX~%WTAl82juS^{aBCe7h z>=FPjyk?bItXX-KVlM?r=J@panVxInYQ%d)y4eQblDXEeo8roapOPD}Ue(g0V{(H= zk7`BPP#cA7Sk8Ejq&9u#&mS~Fv^@U}>>TmuHIhG1H>0KH^$Wix$?VYL!tq@wzZ#<_ zl0_*%)LDxf&MEQ@Z=NTbylMw=1D)<(0ZI~Dx zJ$b*s?R{~zbqypN=fs6g2SX`mN@NFJ9+IZ$Ay5m8tv<$p;{nTOtx{>yI9U0LS>;FK zB+;3WQWG z+nx!6evjt3#Nd99689U?sh&FSP9;($=Wq&N>zLa^=P2qI=skPY*CSSXV=K=ddMPxu z?-O0-G{+qXIw8nVR`QE{u8@Xs(=(gnI=pJ3`6mgkXZq+v4h!%4)Y8WU&2gsd<>Bjm zP4^|oE)6>nl%c#Ls-;&e_J^f=F99d0OgQ>neQ=Ms6BzAGqt^r-^*$@aaRWdCZhF>` z_pLemxl0B6MD+{W`CB=rv;E?J@ooN1X*<6;u2Ry;-{_s{mk?HDTKOCbs<8l@z%gQz z)xL^~ZS|Az%@AC4Ve3a^jSE8%oZK;fOd6%wN|9tLzDJxAnIbRrDG5OT1=lH4z^)nc z3R+b=eK;7M8)qiX5W9G8)7>2Gf{8ZGk9I{=D6j)AI~fhcJLqbFzDMg_YQ>S4hGfIE2rWWE$0rT1)Af!VvvQfIxL5q zMdweb!s(e)$qp`08;EBhK^}RG&2Py7exII~RHBj?Gr>a1> zgm)D>?%Tw@^bys*sB}@cN2%|%s6>HYrRkAs1J_4Zhh%U*R;EpF^u88#c%;Y8uHBp6 zfO^_mJ@bq$AR0U7;b81NYX^pn(Q#owF}OP;+bub^E>=YEj9fh0&xoCJIbw}|_9M?6 zajxJ7H(~x{6zqRFklTZqg+9?gl&G+eHA_1t>=%hv8XQzUT1^h1r&P zKJ33lwz{w_0lxk*wnR@UvM7>C#TQIp;+~)`6n4sXOOBH+;jp+ZVhL1uHvr?Lp1!o8 zISwN&Ba*=CZQOMFAzd1~Zhn`1lUJcotL)`wI+ELNoX^>=xi`6?<2=oAr&t*i>c`bL zzho_L<+9!igItyVwR=(d0r>$BWT(Z+vE|qffONek_J$V4`RVO^ohsME(Qd~xri>jn zp79_DUt)!gSKc{L_akezWHt+p3#XVENS}dmP*O}Z#n1;2>F5u|8@U;g8JrY3)+ZDT zO+TBIXTiQ`4oQ+YDbs+%)-0bT)z9wZpY?81A{(?GSY1B<=WyV$gBSDG*TLwT68wjH z-FM8(j(lzK*U6a|tSqz3Vkx>xDImUbk&18UBcE4?Y#Y=+LEKi@?RSCh^ZN^eT8MS?eQI;Fjrr$JgvcR*n$j% z>(6}Rxo>IH0A|ZEPJu7AOO$%n$Gl!O{j_v3**=wky8D>jNDifdn(IBFV#Gfbvnq5T z>e}>lx`4KP)fBgVQRC0getGbR*u?`nn2N2Q%SD?#x;;{7{0-RJ(t@*SJ*R=6>eL;{ zgr>x%57)jy?54zJl^`~MaBTXkW~>DLm*HKq&4strr4}IGODTX1YB!YX0!v3Xka6US z(m0^vAcoF>D98eRtlY@E50~=c(p@QLi!{qPtsWSsEA{P=WJE%Fw~oFR)geg~U`6+P z4l*K7<7q+|2HW!V6CHDg6^36KO5RR!%uxCDiT*mWbUZiMgUe<(y4ZW-C%`dYA(TF(YN#Kf8HYM_L#H8rXrH3w3Jd5Biic z!@Y`=B)Z92#@P^<4b90A1(V$P-+%_o!ZX+X#%;Y^FBFxM;r3k$xX(Q&WKb=(Is)Q=gGVU@v^=v4xplRH;O zztXnai7}4p=&eru&1@&eaa5=5?ECEp)g<}s3v8a=2&Pxy^&{> zJI7zfCfe9bG)J7=bxBWdhv z;#?RaA6l$Wxs(FDtSl-%F}A}$IS@5v20b;W#izyX{LMZELTz}L@>DpG54NkCl(*&- zt1G7M_tE5Vu1b(Fror7X|IsX?{E@E?2!x6}waQ)c58a)G8hgR=#Hct0DpRJsfBh!s z>t?VNPk;0iQtQI*Nr%PsU!W9C6ltX5lOT1hr(rLvQymRm5t+o%do(GHN+7LtPmw1F zHbv(6q)*c-v1$>ogPlmHo8U4wEVe})4jz;r;@+B@!~v>(xQq-PDe^=?wLc7NChH?> zXBcO$B1;5%dbNCqvYn4=MUXd|woIz^8g@G*sqsg}h)((-DGAVEB?H#NfRd}MBw&D? z@~Si^jBH2Ngt77jG&UXBj4XDJjsLf`6)(EuYqc##-Ki#JlXBQiPah3V^=ejtssrYr z%VMDmz;1~tnLYJ?R07seW=Hqz-a;L$WK7^+S(03GP1Oeo^?bFyHL| zo>HaT4^u*Vq_nIi>8I*F$gRJOK8k64_%Os$|{| zJH<~BEDPI9zgzQaulkg7x5{>_Jdyv)k$c%q!Gzz=<_EBw0+)^AOPc%de97#|?3>BC zPb%5Dy)I18amB*#X`mFRC~}gDzZ8}s7^E{rO9e(*rsrdx_3IS*7WY-5MzA?p-hy8N zOhcwbHt>sqvMwKbX?Fx}CkJM&t63ktCNW;|_|Y&~Q1SHe87 zcF)ik#b!v!{`#dRvf%}b_V2R*PCBL7Ns;YTJeF*u@DWr)UFNhv3xEdw<&>}+a`0!N z`#wR?B`j8NpV_H`dOP&2C&{<54+m^#mHl{T3;AxfxX-$>|0RjuG4ug6DN&wP54w4| zf_uPR*sOr^Y>eA=zgACrx#hwJcj&?&VI@N!2bEu!&)OcU*%8?mz9#Zeq>~d$_Cd&G zK-dGr_B(#^yC%My*_7me_Mab+1a_;F3kPMi7Di?xrC3jqHKXbgpZK8Sg9iH(k=%O{ z-K{5cV-qBvjGnxMIq$qa-O~(-vafvoELrQqp_d$sN!~#zwo&l8#^Y)M3Yv=404O%V zh&=XeJfu&Hha-mFE-k2}Hw2mlH9IFMDFoAJ(|nI{876DSnOBkV@^7|Lb0 z$Djf-CU3ZjQY2AiEi!_QXfn1)KJp!QLlcl9M?Jg$0V7}=CMF9BpT7?~OsxM#{eN3W z_FWbe46%CbFGj0j3avrruLI&ITm7`E<&fOdBus+{>UzbX*CTIuRq8wFnIT#e4PK~D z)vQ?Mv3PMoRC!<$xfwOurtFNKWZS?p#)6D{`1vF|>WCd!t{?EPe$m-)6T)hsm-OJck`Le(=M0ysaf<7({y# zdB%C|{11lQ{$!jt5RnvWl=sRyVouDg6<+`zEjYQ9V>+!R8v-}QOpa-{;|prbhdALe ztY*P=<4s@PGviHbl`@xIW*8L7+WFgwKBUYumAhFD`%z*fI=VRyD@%r;W*{lNDI7X# zfUT*X9204ked_Y?UT!@x_)h|W?LaRqFcSvnb05GCy(hJ|yIO`=|IvAQ@sPYGM( z(db!<9NYOciGm*fKG-@#uc>p9gtKiYnkE7io9=Bce!OyqIS#Vyw|?tL`b!fUGg$14 z_fv{|isVr7*eZGi!lXJC)-^W4mRNHKDjhWsIR(NhdO)Zt7Y;V*d)SM!{26+ZveSkZY z-D)HG*CluS&DKPNwZHt!_ zt2~^Xej77+tf4UGUS}RMLt(ryyszKC{MEWW!)4*Y&~S0nr%U*$pXsNG@C55XphA`Z zso!SOs5~w1laxT5?U>(k-%tHawI4e}YUvN&`!g1HEG5`+j>N?`qPO`zcG8-}8ZJ*; z52v7p6ho4(5yOo4Fw*p()UQmm!ODk^?7v~HAR8-=EQk#m-c z=gpQ4fOy^`oN}ENAg1il$DXxjHDRCw>5Ll5mSJe&M9@Xyu-lcG{nM|+BnM9Z=WoMj z@3f`*S(C=^Nq4?y}R zNt6(_BQV3K&sQ7R4)sJg=M6YqQH}{xHp2Y#m4@j-4nCj$t;Kg=^!=?VjoS&NTD|fC zk8&Zf3UorhCbDc{k4#lSQSdhCxS2%T#%{J9Jf+Wg94n(Xv{0OCh-1|W)9_c>1=%f66$&J~XGCCiW&CmPcZ{tkeOlX!W1`d#l=wY`ePCoZU zbc!5D_NvQbjnGq>FBu?5ls9HCzD$hJ-jFZJ;FZNzi!Okh2q&{tAnXPY(JB`M{h}yC zy#=Vw91_B0H%Xs=GoN|%j1S+|OYA1eWqrP{-5fSr?{;63SUkhtc)+uNE(TYdOkL*9 zoO7}H>OpQZwD2LpJ;da(comz5Yj~@rlVpzAquf@UPHbG{hF`rYG8>mnerg*@dclHc z5MLith^D0!J1DY^ibrPYD?V#E2G66fuj91=i(?hL;sA)m34@Z0z1xw5f|{2FUE~->xGo*yU$VOw&h1VK}j2DNv(-LZqbO0o$$k0;4wn7`O8ATu#j!1C_bep%X{&sr~AOnATg>CfVS^^^a6 z_xEC2;YKJIwg?QW*tIm;g5|tUx*Ujtb5v=PL3yz{4cI5T=v9;s6ao_kS>c0nc;&IU zlfKCTHk-hr**Q>qv~^ZpNU>T^Zw~8_YzbdQISnp0Vd5D+vm@qzyk{`4EMJTJrY79H zvaC+&IY9QkG%HKJ#mZu!6cDFBgp`dBNyRj+PpaQFC_ys3) zl1{bDCmqzhHmMuvGxS}l5z0o498DH6DqDq}5slK7f;;lXsa_5A&8UyKP!%n%4BFwL zLCtMFjofl6^3R{XIL8s)&J#t}_@EmPE9}mxrgdL=#f+b2E5nYFVHb|P@3KI|21>Dx zBCC;w5v6l3a|{cBv<6Dwq3_aZ7UEf-t}S!?toxmHjoEsPm%2-xqx{}P_PVYcKPC)& z-A)?xxkML$LlW?V?YlX!?bRyLK8#S6Xq8C(nj%L6%1H~o*#ag@qu?y~SU%T%9{;Tu zJ;;i^GkdsLN@!}S87UrH>vc@kA~DKKWyv!%=cBOqfAXvd%h|S^5+}Z5WjbCtq9~vF zo_R63^xft2NS_P0W^T4nQeXCpV#zCHh>G74*hAlY^Qu@6v5)PFTKX{E0%bv{iM}Hc zY0Eoh-F|0A#6hb=^XOXo#8*H4)+U}2M&tQJGRQ3p$`Ccjog&9V^_;WNwv{3WO?IfJ zKNPh4RdiQ69qO7+TsNC%$3@L#jYkQBBM+EON^@MOMCG;W*$X z9}`pAN-2^lvWbcxAcy9Fs7h!9-za;;PYp^Clnal9EcYuHCX*YU<-!Z>DMTh@I#{4( zLhr-&W@fbfCilo2FZ$wz)rdNfxNeHO`p;Lz=i-d?%E(H(o=$?svh{Ni561&?gC=k3 zVp$>2b}}b0o*4(@_`TE*Pc(;HOBlaUaTf-m;?(&=v!EfAm(LxPYnFN5;_r*nnD}6| z$~(L*;g!KUI!pBt*Eo4mI2MLuWKo#_23Xz73>f7j`k{3T`P1NV;b<&FTX`kDh9q&i zeM~%74bsO!D|ub==D0_%@9@Oh#><@Lvj-w>$+jzU#HhyA&QBL@CTB!C8rxU9=nu;2 zv%!WLoqXrn8@6Dvazj`FfX!fLu*_Ysx8J%E^d-p_GqjE3R=LzSBXYGoMZQFE&lg)n zwaUuSm61E<*C|&~*m;9dnqfE08soXv@cZgX*-P68A2vX+Z*1(`rzz?< zw+3X}@x1BW63%1arYOAAbY*|+718opcfy*Ybb^ggqpTCu5xkQAw1}8}u(PsV4tQw| zPWj>0`tO@z_nlMJSIH+Y4eS7AV|Go~DaAF4v;xby`;zI)r()TDE_WG6%S{itBDwE5 z$X%y+$XPKpJ=iGE=bjC}C~S^95>iPo7o>-r5u6sE5ma&RL|hZXf8t%JLC4t9*6Ab?p39S+>~}mI^(vA z_H%doZ5EylSRab@;X`f*Lh@%FCpV=#{d7K60f&N)sBV!nzJu`BB)XsRz3*9JgErd# z5X!7U_0;R16v=vH%_#GHP?tguxo{xpoCR)bDFp};R-wcx66WZj{pU_tgA8?W2R)HE zWpSowuUrR=U0Et@B*i+w6gg&`Iw1mccVXYpFD*RyjmKXp3jP=(KZ$~)$`b5U&I`$x z^arn>r8)fi;n$)5uL%^!keo-W?0~$dxf1Y(;%3xMPND!)RQViCQca*d(COHCYSwK# z!R$AUvn!iY)A5Bn)^t2sMu3ds7<;viQlwC1GZmkux-UmfAX9bSNH0k))LVV% z*(^KYfho!a!K64&PiB)9ke=M-{DY*Wf~D5baM>O+hRMdd;pM^WBXdO`EvgM(<)Npw zs!A|8NuoagMt(c+pI6wZT6&o@nG}%Q~T!TGLkiwR9k%A zN+?Ak1=>0B_vy8PE$Vjub>VHlO2`X+q*@JJO+{ftZYds1psc`T?R#ZCJ_kIs%GGgu zyo&rTgQr;|P6|BW(Hz$X&F9N~hum7!cfkp`Lbs`plLK_4y3sHBHHUAC+04$foJX5J zau$PF&5rBtT(+lGsjb7qY!(|AF3Mt{Q^1m}Y}I)J8PRFLrn0zXW(B%9ix=zZJV6Kl zBBYYEN_1iNMdeP8x}rT0XMt3^hej1H`Sq{NCgoP_hFRnmyN#0zV`HVoe*H5_(NB?H zD!x&gB5zWb3(qNxP|BMDJN6{b)u3{|0YW>Q)Hk5*Y73slzqWJrdR9-oqK zS*ypv0L^()Khv~+0S)EKNTU+veRL|F4?bYn?P?Iz`GGpnS$Y7v*RVJkWP=8|YohTd zM5u2IwaU`j?SZCIaLwS46@W;#K2oHSB5#ypW7G-veNos}gHOdjX#Yv}KHvQSE57k*KEtgs>XHNu; zQ95P&%;AV)wGmk9@gKSrZ6t@25LQR8;C0AAdk4SN$s8l6npB1kyWz*zAvqL_wb?5p z4GWy>UE1^gezKI={q24JjjWb1*WFc|iAej_f0@nHoBNLWk=s+rDvR&;Af@<}B7Iam zwiL&%h-6&!VrIP@OP*Bo^_(P&#r64Z~Uqc5Qv>u{ndpcbb&?qfd<2P9Yfs+v# z_g=XsFA|HUYDghScWEAZchSAvl7POMed4c=AiJ0NNI!h4bM8o9oZtSJHX z!SVzdJWVne6%()?M>9A-nTzkNQ{@EV-;yh8k#t5RfzoezAaH+G1pti`tywzzqA*2X z$*-l)1R0{VO57Fz_k3st!z^A(rm)cTEMG@QdCi- zoQgLoueM0aybA&~#{v#X%3`X3|DhJh>+S|u(r4*xIz9MkXe%(3waGELzD?2yNd;uU zgu2tP>k+lEDZ|u7k|-6#$jZc<)MtYVro*NlG+VJ6wUTZdDQwp?gd?kFT5uwjFdxf7 zZ_Sxp6Sr2c{D*UT^qT|Vdg&Qpp{DAGX1FXuISZ-kvV z5QL!9oTV!Dg zIC%Tu>7!qMSXIuifL0>g1E9`>$|z(ke%p*6R`zMiy2LMf{e8*IoAgL>fvKj8zQJ80 zFuONsQxG*edU@zd?}v0#c5iy!q3y7+d7hpcXcwTKdWz#uRD0HMng~Q( zH+l})l(8Oh2Bm;aQyLKC$*OpR{C57qx2}rQI9UtUD-7X-(EIQ>Xn-`wnXY0D;^ojP z<)M%akJj*GvdKV*UCiE&2gK7)yx+Q}%9`iQWfPkW%wtO|d#xFiFM7_85hy=g0A9*sU`EC-r-YFS^DV*3)#z zupSf{|A*a>pq-7s;J7)k2@{TcJDZ=x3KOqHb^UbuH%8;WANBq-ImvEk5B{ZLF)dR_@qj!S|5{5@*@oj9d-N5J#RgZ0FtY?y+8QFwnZBPKl8=L zU%c2q_c&>TOIYHJZs1|U@EGE42n6n<1^L_$!wUSH13v*Kq&-~2jK>ilbFx*p z=62C|Wb0~X6BI4{*a^O{5sbo9zW>+V46fQYD*r;3jn`1;!m+yD77$9N6q_iLM8#hu z*yY<2{z%-UETFRkHQ4{VOse(5ZI!K)X5(l0++j9P{&Sz?*a6IT==QpQ`ZsG>xU8ST z0D%!L5X=wX1TW<#Pzf)YkC-U>@iLHb|6d50pu`M;aY{B{`E!i>cdRQVp6-=$-7FSi zsE5Gfg+cj%=L*lPk%YQtzwBHrE)ds!^pH|V-xf=N8-G~$m5zmQv7Maa809Yy=q~uV9*n`&!{rrPNU8|vcQ(pbw4gvWB$IA z&VpBW-tJg9vd7*zef^x`s7{QyBnRe-ut1_iw)G3EM>4HOx#l1l^ePX(?t?ev3Jm^` z-~CL6dD=$!OHkARo^m?mY)EooqZE2EMn-8>iyNhFppVhcKkave{{p5?TX7@ggKm`W;pVH0)%#_qU;kfLA&Y5X^~t-gZ}?X5 zKNx14ecSgxFdij)Uz&~6aSN~aLrQUwB1I@Kel!?O`~weE^}G_)99KLIHcjC4o|E2# zz0FEry$WmXsyO8#Yk>R2C{F>3#twd`tT`@MR0V6=nz(A{5pL(Fa94vSXum4w^&z;1 zXYu+nDAaMN=fu|bJZn>A0pSV0I94?K|5}UOFlfnO?6ViF`q)OYdCFS4*hJ$egqT!oX0-Pk;kkojq3(7 zmiFZm>wXnBwJ@&xe?}O(TaXv&vG1+9IX)Mm0%nic$T_JxEw17WaxYG=4nbz3B%5O+ z4gnM!ggfL7ZG4E2#|M7aES@gwwleT?9#O4V49Sit3xvhMmA5^v18Oog7%$U7Dj84X zsXdXmcpuJB4umq%eNn4C&WCqMmc)JRuW1wK3Tk)*kR>=ArRkNfpMDtVJ2hCpy^2#z z*OPrwLo!rk)6ZV>szH$pX~uHUopyjFWP^xjtt*z-Y@GII3}a>UrtJLzx6OLn!?3%+ z(vDp|*bIG$drpe$Hm*tANYeZ+dY{K8`GBxXcvITWM+yfAe7?{4UmSCZ?JSJB)v1T9 zX2Ess^RcGzcde<7T-LlW&=?&638cSVm;dDfWRo+w`LhZ|=VC9=_XHQfV{MSu()+nR zvTQoXr)#prC|L4dfq`%;RL4A=9curOyC}n&Z=8)=;<{Tq!oWnnm9s=(DxcCU=M@ES zBrAb2cxB{>M1iTPUsKLa4onVgf)4(b9?if7i|NV5$p5-N5+h_eoE~Y9^rkq&X~Ljv zEu!W6GxBN-A{=-44>Ii!t(7cX)|4=4Q`FJr;e*_R;vTx4B)nz_&xe-XOj0O1#97B{ z5H?O851MIOY@ce!O`|POVrNFqEk8YIEgNPQSi$YBQex=}=w)>(;A@14Y1uh_%wk@nLY z5i$SK_(RH8L%2#(<%fSG{Fh?VYf<2hd=D3Z$cl6DQk&2zeAGyJUpXS zW-Gd+-E<+m+eEFD({M35!FK*;WtFBJmM(pbV@A=hPxRN3rR=Pd>+S(brUmA

    PKd zSx?1h&Va^!exIa)e=o`~;|jei;3`|L@LN_S}2=o*|^tCjmPf{`m{bc;FTYiCViYt3-+ZQ*yb7^7Tn_Iuy@`GZi zC^;8r>VU-6tDIg3UQe>;UdcZ1K1p*NJ_#c^U38_a%JEqM+W=+lcRc-ig7|sr_Ak8l zmw&ddnPUSR7ml_tRM=t54^|uL!I1mm3eO^*6qi?M=fvRmbChs9Gf=hPG5-gNID z=a~JDMQu6jNbXcpXJNDtQ;I_rDWT%;c$@4rH2z(PI-=4li)rjU9w}r**S}7g!E5K2 z3v*}fn|@P@bz|!lP0EqpK;V3lSB30`ish00foUGJO0@k6f}+`4uL~i-^C77x`IPCO zgIoaHKlP`-fa9NW%w#ZI_wni)T~?j?!kj6A*5Q3NHpPXv(hSJ}v?s_)qoWO6WKzh6 z-=iGkNX3dYnMj}NApF{bHg*`d<%y0N!~5OWrdcpq*`zZ{qXKR3G{Lj~+K*%zvP$6^6BQi?W;T%qENd~||}**TEYDJGi4;12(6Zki+`5`uBD1L~!c4Blt+ zfP2nEb4{EPnI>7mEAd(Co-WFW?2sLVe)z>0pwZJiM1AUtP*c3_*4$cpFDHr9gOy2ubP(vGVXekO@t&9*&L!Yi8&Ky3)#aIEY6ATCofnvQ5GQLN>3l?N80Qq3ikXy279d9*f zo%=IB73RN--4**jU3Twb&uCVTjO$u-jq2)kmDv`ZjQ`6PQslxh+y)ElQcWo;C~}mF zFIFFup@g+Y2VQkKU9Y$V#k22%!cNe;K)%tdM!9p0*n*5Pjf{(0I zNgGGV8Q;wfDi^++&)qspPb1woD!pO2zjr=RxIp3{OJxLxORy>!|3_-PQeTh`AZ@Qs zD0pKZx^1--_Tbud4=Z#}+3?#RpL*R4qO#e4TuJiSEoiR$^&lrKAa{gPfU46$Dt_hE z9$u$x)zlT9$Q$3z$E@~{+oSnieg=;RvR?T~*Md5tu8G^ocGX5+`=Z-EUBOzVj+8ktGwxul*$*zO zI}11TSO3zCo)0#xdY#;4haMMphgMk7fDTZK2NdZs(fp?RZ3Y%zB*WEgE))IijL81EWw9dxh6B;Szc~~CgQQZ881joq$cuR>q~k5;uLtg_A=!r@ppINr z+Q99;Pz2g2ZS<}nkGxHn+CeGGD8D&RgAu^$kdw0hxmpNMbb*t;UXc;mt4@ZPMqbdp zH}=hf_h^*XkZkuZy3rdSUaZDUsF_RxbC?})cMT~IKNYqnx?*JDU8#O{jaY+<8k*&y zDBpn`r@;N&Ne{|bym~j-WxLuoz4NDE{eA7!PAU85UH@(^T}gMzS9|x%o4o6x-{j=C zZ!K&iP2qU{M)c=bCc(cOF$Sznla&c^*2_8PX9ii9Ua{H9yKv~3p-}6NH2H&oq+|sQXQ5!6=Vo8hAGD_3eWB8m4CJ~-`IVp@ z%7povHd5`SK^1}txwqqda<-kaY2IynltV```^DbFA5uPIpmH<|8m2de|qo#0Uc&JFj^ur!>7PauT>&3=FZ4XB2CB0%j%h} zLL3Z3)25$yL^~1l3{C_h%uk*86;=?LvcKj=_m|CvWnme=lT?kz1mwb-psNNo&Gr~p)GUM#O+rBuDYsQpYuv;Qo?!t@J z9t-@WP>RhI*#Jp--&NqL;?5T*waM3j``%L7ZXiBH(r=q*&&Ch)Im1|B;5kon>}6&Q zjI$g~Dc|$wpZl0`(e%^O#boOE9EVh3ebpyF4cIVkI<+r2&v z!;@$Qw8}Lcga1gAQn&jq&Sv+IWw+-Z3vB@Q9?f+hyv@MP5$jc54k+ejoOTw-8pl*P zFQ++&?E1xv?(kw!)S74r#et);JJ8^f0~V!NsKFYZ>-48y9egCAG3hyB6%~-vaL$2t z?DARLLFopNfp?&*gS<8S=eK}rm%(2T1G=FM+}72DapJ8mOcV!Cz+v^=iN0iIVP2{B z$cv%P7N-9@f%PQOg;$;o3!AZ-Qf#2eI>=^&^RXIqf;A^q+uhg74@otJJ|NQs$s$M3 z`E%#SR!}_mK@J|k4isDCzf)ixO>x;>5JTj|tOn2}zbL^MPP!f6@?7X(Y&O#JQW!d@U{;+ok5pa?R_yIus0ji7iQwj6#vrEzUfgxS-~49rR; z1?={+E?kuIv4vk+PAQeg9(AEK={d2xfwuPH-%m# zsV=-#DY1Y=Hl={5Sq2r4d0ah>9K@GE?DY~?bIoU$4>C=r)6EJq`^v4kP&(HH37ztA ztbD!1t)8jD{*($b#H)*0<>5q|GQp{z(DXj*W>#LwD{FULd*519k1cF}_@E5^^7lv))kLx7kyiX`eF z2mKJ*>-x!viPEV`06T9H|v%HJZX0W8o|23B!d}+V}vWjEC za*R?`Q=|flYyZ7xp{d8xw0u=WxAW71d#*AVO0i=P1Cw&@tb(XZ3$#$ds2QZQ7F4P9 zbTR3KIJ~ArTo$W?OdxXn;~fcteDxsLq|Ld8l+Q$6;0sZ$9#wvqg*sRT69ntL4+rb$ zRJ8#zk%-5g6D+x%Fpil0n+XBw3o8P0|GsJ++c+h%8^U(_s{BYuj&~98{bZ}O+)Dbi z_>5pLeQ7~USZ>g2Z!Glhn$taLv0K*on9!`-e)HFU$Nkv4U5Jer_)PSjWb1-hmyOFZl(?FuWYE^5N^|X;-4aOR(u1D3C$g8L%l?m>#Z=q> zCU)G(Y7aE!r074s9b)!UW~cvk6UlzTs+4Lhe34R0QB08oRD2qzkzX9315!5>Dl3Ru zjk*8*vW>ibQ3C{fe?AuOrSWvHET6j=yD}Qw_lVo+>pp$+`h(Zc8mTziq|OP#mNAr% z?vnPZb43`n(<)a>&xbqKp=AN4QH#O++YZn$-NFo}us;S}uqO3+x{}(37Z>nZ$2hSi zl%kLVS>AZ$0D*ZX3G^aX2n|k{hC7> zz_j7`VYg0?#n|wbGHuum_geUDt#XCuO^`CPa*iBolg0pz{h=iUnZY$)Y217NacG5` z8F2ZZ{pSZH;icJzGstlj`S0)@ax3yjQC9qYz@t-THMN!>sIy+#0w8C-_p_cG^YQdW z)@{#EFI_Gi-DPNQJ{Fobvp7bp9OSl!BXM_%{38(UP35i%J;ZGjW_td1JC**`-mf40 zYTl22cS_kxV}-U>SxFB_s(3^4(wOb4tOfXl9L{mFpPRwM29c7OU65;fdW@q0WeY}F z8H{I5Jd110My5RM7qe!X4NCtz(iLR)R05o|V|!2x>v6@Djmi zG945G;Hoj~c3SmF+#*@)HSC6-X9qZ<{jh1Rh}N`^8_1-F#wy&Ar8~5Xpt}B z9S+u*3{wi%Fyg-?2j2Gt&j$xzo>?Pqk@t|JDzmO@iQe%4CfC7WE>MFvAR#`0xS3UEqxQ1NGD8~Dl5ovLhZ-^@avcK$ZWKtzrzZTcoo zy4c{c-1i_9M|Hr6=D2n9+agXvQgT`F9%rQ!E#}3pS+{tCb6&7AbX3`Y z)nycGPUYT^Zh{mFa^6*2k$PqN?*%00kUO<<{Y6d;4Nfi(KSzCpx<~bzv zYZrK)=e{@qPprT*r6uC~r+#fVFu!U1*=+JRc3UDBuEtzrLEQ3)QhY{{elSNgwuaOI zd%G#^h+HSf7h$8&NE;&bM7tMpOqYOG4*lNi^2(SpdZ!Wvz%u}hP-=s&nl`X#c9!aR zKoV~|(d_ktVjEza8+J3@aTc^BD#+?-5MYFqW)?(?G;Q)eb$PfJ(iEoceQS6FcYVx` z`NM8l&|?~0tj>r;slM!yR!Y38NvB7&`yf^@8sRRdduJZ?hUX0sg!@5q=hcG*N`XF+ zd?M)uy;QBzRG^F^7=vExytRQYOWtxJw032T8Gc*!vJ6t{!s}a;#rk%fQhY>#w-4dv z0n!x#Ou?r1q7?bZphUh;ykl+$-8XY34;!c(Jli86`q$uD#VZ!p$vUI7sw{etdlr({ z!*1A8adUWCXTZvhn><}aeu8BSygmoAo zL=o8mTec3#1pFeW*s&*e&VGZ&%8tEKV+b_&grrdLv}vh{@Z?>cmK3?-Oba{9oO9|5{Bv;3?x*|%mIbHJ7EocsJMcz zM(=h$=A|(IFv#s!kJM!$v3a5Z_*!cqq>DFf2|p%=@~mAk9Z~BQdb(4#(S0!zeUT*Bk8ZuZY&#Ut(h;TAtVtwg>j&h1ue? zUhY&`3&pyui_M^5d*8RjCz0s5S3ysGyYdR1OXmhvaaKlF(%GsTKyirGHG9PQ>NS$Ppn(Uogw)ZhZVVqDCwf^4z6V?ncY|3t2*okJ~i@|U)hA@$kHHnkpeS_Of+93%l zcq~8TKjIx@^8C_*(|%xwj;eY{CUZC+=Ea%5YP*~nH39r7{{+m8x-u)E0Wt<`_U3mAS zv)KJ?r4-2&*+j+Xh2%pMQjWM-c$rh{eo1y2JdI>g4ydC5!G3s|Z9UJ%t?C=r4I+Iw}a6Y}UabdrLp{}zhxFAYT9}Lm-1b0i( ztYEGaqgjv#Y8C3~ZNZa+67!U^0n4PIWc2T1^e*2mN=o`K^OExBzGHsmHoJJP3-5(j zS;TV(DaEH0>7(K?TDDbGD7vSHYAI-+zRZE*yCWf5)z(>cP_A%NRxv|U9g@V!;p_-Z z@XO?$_5&R+(Jr6WQcW%7nlgA9K7GE&qk83abRz`psyIhN_CZDplMJ8n&eNK9ex+n9 zXuej`4SY?W_((`EH*?0YTjmTsXqQ0gxqAU;IAov&jZSuU)LX^yL-v)%JRpShOCbAJojG~M8#4Q!AcQ(tgSkC45e zwX;8WR$hhce*L~8j{TFhyW_I%FoSC|nr}pv2+Z=lhrLM^(eI(YtEryrJ*S(5F+K__ z#E!Yo*@vtU`^wvgW-AnCTjKe!{}S2CZpCq75BZ>lNz_vcz*Q!S-LK@83#&s>^aI64 zvgp#;`xK8nlLMiQ{F*qQE>ji=2RzGSvs7hKM&-Ncww#;ZAX!VQI8ELch0CC8`&}mj zjTNrOn!k}_=b=k|Hb2Tb*78E_m>A+OwRHNl{nIz{>c4nlq0R^S%8Jz+y&E7$)60eK zm1oP|IpO1GG5=0@cjiYj+m&yKHoX1cW*hS1PrRh$&Q!A6Vz2R#QVdY!0jQJ#Q8j1> zt48&DH26i!6=65U9p0Akka2~k#@ezZZ?=Fbet58>)QGDFXa3_%8Wjt`@%isz)Q1&0Z06pJyRv6 zFi_+$6_3u(pw}7bM`>_B<7@E16$&@DIpQSPA|(hkOT#imrqfy_boO(0Do+cZT*PfQ z>Z<|Vs{#t5)+r8f+Uf03HN25Gf>`|0q3uIUc!HPcrMV;jOmV>Hp5Mglr<=Vs6f>rMF3d%*xjpFtHaI7N%nY z2w=LEc_A(SP@Cgn#?1Sg^S>vnUNFpnKERk2Xd9(Sp~z-v+7)3R!kU0n{-6#J&@3AS zCF`Y}He!^Q$|_@;IcNR*)y_Zw(+sdGgJFD#6D}}Aai{D*g*9-sb?7&_1zI_Su{UVHx!|Zl)-I)&W&7*bAeq%eO&RC>o@5 zLT~yQW;E~*$23I&1t-7JuZ}cuo1$_!A6)l&zbUSA;UA7H9CACd@DG*nYvk3rxIS(k z=Jqod zQ&TWKgO?oGAx8wJsw;z=eGOdbFUW>ME_}PA{2c;#zKIzHTT`oeI?&a=F3$@>ohG=L z3x5IZQStCpSQoupwQ11?F7)Pf(fJF@frqV3si}d3bP*ijruaJWUaVMjfU`r;BJYes zL8#AVR{n<(Iv$)yQ0uqiov3MM%-#G&>z~PLc9?Tv7YcYB$1I22DFwvPw@~qCzJfKO zH={O_x)9(`iffQQ4#HQM48^2qF4y4E-~{&=u|-xz0`L^bxn9qisBfW79b^tR;C4g&`62-yhcoPrgJAcz&gFd{bE&el8Yoz9)@-rhUExmRX> zzi>qqn@^EkDz+kA&-5@?r7Qh!d#xwN z$<69gTF?BQ*(h7zeikz1?3>F1^S=}~*789)nT z+XWcKf5Y%cIKJm6?CJ2a9Qm1>2SHplP58>2?H8>Hr(N}LB{nFeF`!Z!vLPJBlr*QN z9C%sF*S)Mk4y+G?s$~OiSPMG9>=Gx3H^^HcM`09#--7q(l=o(p&ug33MHhfXTP58i z)Lil@8l=N;0-nc;WD~RL~?jjyReZ4_qmdv8v$bIc>mI$_VXVK7AJ$)TNYKJtnl0qjpQ}cDwOd816{*g z+(x$hAd^>WR1S2F-{5akR?=Du`ZpDOcv`Q;rBO&RnxINh-3+@KmO(ESEfp>Q=UiZk z(51PUD9`9Wu-D7zldmWK3@CjSKgYA?vXDxjbHmhWIrwvFWT zAQ9%BD_I$#L2~mBX0Ni*V;Pt-TZY5_X&nWC>@gqq{{c4cyth1(SZ|`UL`;?hvh_xRGg?wjCCVeezrHDm+4?kEbK^EQcovRP4Xk5mm!(bAUBaE{+3akbi4DysqZ!!76Z_?}^FUM}* zF6Wd_4>v%ZjM=b7Yxo$>`24c;?Vm=O&C?I}7wSnVH}mAe`>zHI@AxRi0*7S<6`KZf zWBLfx(?-VOLeUi-Y~BkbSm}nRd@pX4It;vl+pmRG8u9t<@?1) zU=`YtDVj`@DnPxuTQ4O}*O>apo0Lu9Ie~#Q$#t6wT2lob(D;I#Ut?K9M~k9ifU^B@ z4Dh;Dwm8u`rT&7{>exuOqcR7Qp5NeiL4l4D(n@LbD#Q1M^vPO1_ju;f8GffD9YV-v z@2qfwhP{J2;J(l!J8rw14UF3OwuUU_W?)>{|H-g0FsT%qOp!HEyF(ughW_C2j+jN269Y{?Jak+`N9~ z^XPiHE^x&h&87J}rW{l}kS@ZWI4plelIFn&oVxIEfypq#F!<2{Xs|g1Czwol1eCl1eSA5zR^ob*Gzlwl zFR|^)JIb3NKU+nh_bem0JhDL4MSm(s0YS|M|G|tuhDdMBz}f*Iknt}HLHc@#(udu! z*3+{2?w@^OEU2+Y5gb4J7utCTC)7;*;FY9H|5%x1Ba}r!>IQeVLyc)oJR`k{R~};O z@Wr!mLn-~~?27PDXPXm3$eea0q(T6u39`>--P0zZ9;#Oug;i5eE0ZM{j4hjYMyi(# zxMAe}xQ9`;TXE(Us8|`h{c$UGau2Y`8`g}KB(GVPI@Ns~!)k33R9 z$tSZ;9xva*>!x#5%VV;n9lQazML8<07;h%K#6!Ym^@?gp95?&lj;dGmkw$SAAKOG+ zHZjZ%f8A`$`q$0)OW5$wKOt4zg0(K3A#Anqo6b@!luLa;#b!=z7iyVK8VN3dA|9}W z8|yOv9Oy{WDI3L5abS|mE%RR)0Yy|O#$Kk_!D^Y6o)w<^ND_+#p{Uu7rNpR=_o;j( z2rT3C7y?QmODAb6Lw2y}WTAljbw7OQAT-t8rh%@)yH{ueFYdIg;PQ{=#9`ETmgCq(3}JxM;`&4j^2Ow?%{tFJt&5g=X05d?heYt zfx~V9{r9?GE%}p9xnnCDHhTeI)+z&!`5LGt}z$JNI?0x7IDj#gZ^Yu@aHkX>Ur@+M{J>7JL&&Pvd4D;AMW z&l$HvnT1P|N3lSVwgV!5F~|H3Aab&d+=kBOCUwUfHNX$@{&nV@x>=n@G9?L! zq0hrJ_0?{_6Et>~C-b`fOx@*Ir1jz+K_{@` zp$BJ(YF1z1_0ZagPE|!XXcGCSPsTA4$rVWsd}~YTM{X72kKERiI$-3(5#wHV-0VGj zLPq{_+Boa539~&RBYYiQw;Ai&>erTAqlZgM+I9cSU`ssQ&>W)J0~Fau#op%?ikj6c z1e;_C?yZ5T->IaRQWtpLut7edC=G!$loEe7tFQT0(s)j%yyka-mn`W4U6oBxHhDbY zqFATw=YJZU<+BU;X9`7ik{%kK0Y(}8R4TsWml2%dKy?IH{Sv&DN8ftGyckXF{pH_D z?Q`b0T(`i;If`wh$XP0ORj`h?D*6=B1s0G_d7WhQH?B$zyp@73rjf1@obfK0mgsvY z%A{$qQ`+U9&fg*1B+ChIRu}s{3@V@3&#$A)0bQul*R1Ym_VBL+rAFhq{gR58ntiVX z0nd@DF~q#~IqHtK0Hx6}@TkNMEoljym4@!~T7v;GfcS}(_~jds3o zd*|QI;T(3G?fD;J37zos_-{T@Pse4Ya^a1$jp{p-gfS*6w@=mJF1m;+`d>*&*##hn z!NUH57QgQ`kB(mnt&GBGWi-G%?4(vC?Av@b0JjHJ{wp(Fa}>uuOJ1I zL5kgGsD6>GKJ}FB>9C&2^CrRF*=IfN|8-eYzccGWQEJ;8R&GMX7+{`MsJ&xnW|%g zb%9k=E5-Qf5hxgJQx*he@r};!Cp1khcR1M(iLu;S)tNrs&kT^JUz9B-o4EnRg*QX{ zEdY{Fu~5F5Ma3?qkaDwKcvaXZIOJO$od?^SjiP?YJ2$Jh2I^k9IOUOBcGO4H9>n0! z3n9(X<&tXpvZ71TtlmQE=B6@LviKNhdNkazGRzK*1q{!4_8a}5#Ce;sQvUUCHIg+h zT;BjpXTxG{TPb!E1@ASsgZX%72QR_Bd-B#;os;6QzVn6}3i|z$s>hgKVfEuMoi2wf7U#CQ_uGwQ_~!uuC=Pd+P;6BQ9CC^F2*rF-U>PH9l8)l^fI^MeZuHF;3n2qLVK z>wcQkAXkIjm6{`x3sR)$c_2lJ5h&W;E6$K;a6l-s1a=jy-PVkKxf&xzhQY<+U!Ddq=0W;64jtz;2NePPVZZVNM$NwE-3+X`j|n6gpbaVb?vYle!^4D`X5d;Bz5 zu(o;HAd?cg?Oio>$>q?E(^0f2m2`=nF8nxZ^o$uFPaoj@sG2D2$^tF{Ru{HCHfj!V zHCgA=;Ddr!SkXOp%kcC!Y~$*)0b%1oC!c=J|6qBj*=Uq>6)z(O7vAF5T9|?T6btT0 zF%^q7?#YsR@#nGw(%WHJu3xQKJL!?zZoxVB#@ywC&B_M4T{!%DLg+vcMqp9k`ve^q zlVk1*!sENM>lLTSy;)teYh=47_40Q`Ryw#)<=I%USTde=;81kfURp-*YN_i6)qu>5 z9Pw@O??{R}0fnexTM9kJZl}l=Q`LGCaMBs!mG{Yxd&kYm0o`a*vmdr3nE0!-US_6F zQ%+<&zNn*`35UM(kq6^66RsO!_i$FXiTAmA zc&3>*9-y^=jJ*i9cxW6BY$Nz@Uljc5CuX~G>0e9llO`8lae6EaS_j3pQRE60i>(p8 z%;7mG$f1ug0*My%ks4vYYa&v@-caD~6lt>fNW+Olp_yd8Kyys6*RP$%tLp-fL4zsQ zLl(*kA^F?wfkPb+NETq0Z@>+i*ek>J5qO`0E)+E>fi^v07o8gjbjSFMkopfhSxs%I znlxTb^yj2hTpL;lB%<}Jg7#F{jFjkG35R3%1=#15uF zblV&DrfcO|=1@p6ANu*)X`GoRh+mA)cYGAHVDrs6@&dP6a9K5w?^oUZrS;O}vXNRF z%g_U;vp2xD1}Q)Xo7a({IAQw9am?flAOiso0aU1|j5mu(xl(4W-H}#WlPPY28en5@}&I_g*nh!45&B zLIC;JFrzjL_o?^DpgUEVGBQ1dTw+K8v@!xn_Hn;`NVPLEBUsZ< zXV2d!)-t$Wt%08D#Yl48NgL_Xs3bN=wU$3xyJ{!$#;_obF1hf*l;Vi7MB`i-OE#i$ zr}=dxpP$BG7nl`Ys)BA(D3eC1@jlN=aR-#VHiTD9X_0jVqzI5N<4)8a?^e&;abcPp z6-LIvM!j;i*FDB;Z-SPuDI^C*Dy?>5B%HUfDkmrwA~Z*+*wu5ZLMz4lBCDZS1jI1w z70ZKBR8{ZMB`@+#Vl^9~zN8|&UU3_EpAx2DnBT#>4n8k#`!ngbkk6w*ay~4Z80eLv zo6#!F3%w(ub#g};%o`oU1B6|qdmK5_e z2ZPaPG<8;t!Dryq*MV!V?G7_Fay0WBtA@o za;X@K(J6DGSv4_qnE)isFnoioa9PqrL5D&r=C_d}q#D!|9c^^Dq4_bapZqW?%-Vj4 z6PhRZiUrlx0xDL|!$eY(FVrqf%9LzYeiCw4)dgv% zP7td5(~?hLySyNMM%q+p2?4RG8|szr=(kpeSF5mBz7E8Ad*l^xH8rY1UIP*!2l-cL zrpzmGhTyzoAY*&#M|*`6WT@3~A>GzR-&|^{TzKnkqeT3oxB-BYLN4&i=j{#3fz1C5zjhy~%Omxwo;&(SGm2H%|XxYv9_SmHs&IjgQ|t zxA5;FN5QZ(t1eVk&H6t$+Qz3Pp&~$XB>=;r+8Q*pO0RfY+iqi^JiZrTU~hlDYsaE z3Mn>^A~{%=TS+&1mqQ386N0!Yf`$b}GEFC(IvrLKS+#H-ABuC0veRL!=XTPCAp>rG z-sSV~-V7QWPT(>&LDvC~yD`%7Zi>T@;b=^l#^bc~O!&mD@EvQ#h!7dR9zStpvgUgdhwy?WL~ox6HiEbb$DKxXBTkx>{T4|?-DzCc^bXB zaSoV>ojJ7NE_AowBh75P zFe{+V^NQpS|He#|9ofk)s%F2K*`X+>_Rc&Wut-N;RCg@6v|!OW`BlZDe^3KbbM1@9 zsb&Kb|MujyB;$oKAQct{WH-eE%|<>I+shbf48Ck(bw1guk7q$Kpib59b0EwBbWd7O zgDjb+!N|&5U$8AeoDFt`*(N{A@8v_%s151|4|#MzB_V|P4N%Bw@^!js$M+p^wJKJQ z?c+m7TbYXEul&fGHI9pC;=)Ug4L-R9L4p8tiIBuWhZYluN2axk57Mn-4RXSv3f_$w z5U|ri@oxFNx+uI$r^M6c%99|C2VQZ3$P}fni&_y_=HCs(>R68YaQ5BV1)>|kBZmRL z991)`N%g4=%u(H^Pfl9zQ8HJjtc$`+ST)?|jla2D2oXbNSI9m^B9v(v=u*zU+FyY=5bMG* zev7I@kWp@t-3v~d?&MX*>Sc47YO5!md&q4owEMkRlfG*P#%o7=e@0HeFiX@;3m9FZ zSSaRfg03onj#-`xixY5RDiC4h7q}Zbm^#0>7+kXqbiM3`=l1B_2u%yq?V)83EB6LK zt$Vw0PsqxM+z2hx!R(&Y0JM`Dle^pot}(7`cv5qT?PYKY(=zv9d%hQHK#eq%Zt4Y1 zO5I?*cy9nUJ$yOHfEx;d0+LY>2g@ojk-bcWJQ8-GngbM{ZcjK}eKw2@i-X7H|G2P^ zY{UOq=l@8tBXGZJBd<_;gMW_hRqPcvtJ}|}R|bDU5c;Q+{>d2s|vEXH`J^qc9{0ue6Tl2T@|COsB| zz>*~ly&3latUd~K@Ag%D+&o7+}yQvKNqBzy3Slk8MBiJWhkv>wO zuWPETbAXJZ#jE}DV62n>rmPE~2hCSE3g4JnFfR%erYJfqBXHV;ap znqDaliRY~g+!oOq-f6zOUW{k|TpD8VysI+Za@Dud_W|<>c*4oy+hsc?yQXJIYy3M^ zjR(fu(M0Qv}ZdRk3crtG}uZGvl z&z$;b>Lb7ADC4Z=z(23>)+sLpUKk%v2RqG-h3{dXp%~j;-}`mZU#v;{T{bjggNFZB zSPPraukgGg%9L~{v`l|Y8oyn5fh2}z&wr%oVv@sq$TjKV;A8$!y(D`CcXaUTXuNs^ zI#o?FeGN)yEhG5|jm4U{u%BYXniTuB3oFI9!nQ_0 zNkTW>0EPcaEC$mw#h%5UupI3Nxw{Isf-MyTl@XG1;AYzKwFuUeI9fbekK^d$U;#g4 zECu^6?3&mp*l$=+NmnaMLMl9)d{ZT>d<`@7-k;3QlJ0RvSI0A^%T820V-&|<{@vMi z%b)kk0`xcsx#&tGJ7I%bN%wnSX9}g=LQSt?n@6sP)5@two(?;wdT__5?+0sn~sxU#o*Genv>EI1A{PAR&iM8_4_w zxq4s)gmm3e{#k!mEH+QW48E4ZBK!|LG$*6G=s3|weg;`uU^mGRs#omcJ#tHqNtJv$ zJ1GKKoV=lKse{2=bV@v(3|f|{(ePQ41SJQuo@m@RH!Jb?VkOeHNl7zD2SYAsdhC{P z{V)o^aYIvL#ZmtAt|^qGY6mR!GTZ&W@IMCSONk5PV#?=X{4`e@C;DP`QuIBbr`$)X zov7hpJy9IdVg0z%jya)&x_XHJi*U28S*7jSLw0i83cIlLddk8wRZuL5CGMwUami|s zU!PqE+=W$hZ+k8e>!Fctu0@vW^SJ`ABCBB))HvSuY>=b%h!b^*JL%h=Nx~!{njWk) zLfsHZFsmMbM#LJQ+n#BG#ZX$X#-~=bCHOW3Y6~Q$(o9EfpE#oJ*;dFw`1{2{kLuUV z_<6w3e2Y{*XVfZ}El_fXVgWWMsMs!f|6oxX7M4^)RT-9+z`_OPwb7{i(7dqUdp)dY z7@0>w5Dng668xbOu~A2#BB!U+hOP^Q<*0s&2G_d(S@PE;iDoyyRg91S&yqhSN!mzS z;DB3$yp8#A%5hjCwanlsRiRp@%CDb+m9fF&9&d#g(3EF-_cHB14SXCO%W>^SmBV54 z3}D%dD%(ct^GW}7nr{Y_Td>6^c=Cl;Mo+9f2YTvmk)eA5t*UK;A zW%28UsX-^F?33jAe5R}t*ge(uzWQg}XE$P=F^;1zv&^EN=Y6-dTd{wp0i0V6PPcYi zJyan@r+6F=Jsb*OcI1TV{D0{^z>RI=Fhclk!o9<9w-Kl%S zQnEDWbl3@D-|V*8IW+d*d??0>`F5dESUxWYRCaI5fnFuF^A#wQ78ZaAm0r@UjuXX) z<gSo`-@vaQ7_ZYehM{f3ymsh;BiS3zlC#sEX zR`q-1U-PNlK-ch6CGlb1bK(UDUrzAvRNsk;7j()iB?sv{QTx^Fhbt$I6?`@@Im**R z(Bo*TsCvbk-&$vtxTL6E*tfBfgSy1tlXfb)X7|i#RzpJ=KNIqgP3mjkZ38BGv}u>Z zZg`hU9=R3CwER^}M(`%tilFtPyVIK0`MjpNyNPyI^2~UpGrM5NhO94<;OI|Xoi{73 z#eTV9#f5VQprtp=4^E-jB#Nx2VzvKT^4jTtz3|oEH#4Yf6QywoTbK@q?Oy&$c1Ay_bnjNMX{eyq?3w8DHPoEnk7+C(hH0H z&D^=EQI+&f`9auSqx{+$Ai`ZLLIIi%zhb%ox=;&5SV5N!4Ig?Q#&IfX9R7kG#Avbv$o|#iM z*ZtCiG9}k0btno%+myu~8u+uKuni?g+CxAM$MP|~=hBFqJk3pB+PuneolkYBHKy&t z+bEYRPLEk(@GB?(uiC?mx4N%X|CPjZ!C9n`F0UW3&&6bQ3~z z=#@Yoe>}1o8bwBzL9lwE+z#!91-DVhes}f%m-f@|nPG9@XF+d~QzM1^T-aCn*kbZq zD7Kj*=Rk)XLUvdOqf_bvo5I^kHo+kAIr(Q04n|%2YFU+6kzi{~Jlm|U4NaDm__fVm z3N~Yvd?&vv5~4SXrf?9TfDsz!os*}`C}D7%B0+-td-+uI_uk8RZ8_EFT@dt0a6lT5 zC#qjxPVJ6)=zom=NRaNa&G&});OQ&>QMO-JY1`Pe=7hwIu1sNi8%cHtUee(XgN|ai zQe+bq+wcAP<(~;e2Q0aSyxRnbn`l z`EWGvobxXk#RInA`3PYZ*Yy=<{OV61zHA1_cMJdP*JK;FRmp`76C|jHImSg4n@^Ek zD)tdc4#fXAs~^O4P44$z0hH1!Kr9IuVRrMY>1yxz2<(17Jg1a*XHFedbuJA!>C>z} zBhBZndh@e?|9AlsCHC_72ITw2J5By(G^n>QB%?sIJ&)8SoY?-n$BkQ=p83U!vdM<1 zI{J24akL?-SzR`{nrwy=3q#Z~Cw02pe7p8fwap=%a)Fb9qqcZ?)>&_`UADy6#+JEB z*$Dm232Y9X$~3UB3Mx9B7XW_RoY@13r#WdIU+8+J-%n*h}@)DBSqS& zSk#)?j#YOcf`;2;lvC97GD&QY{9tgq@FE0JDruCu$4gN2R}JJE)q{EfK);cr+U|!{ zqW4TZ4Q2lQJ}HCzWqo9&=oT|5d<2;aKp0d_Ko`@M=oY5Iw~|J_vK8UC1R5-;L6M|B zg8Z5|GEzREbnXp4?Al~|r}YE9rlx}_htusq1hSi44Fqc6!O>Cuu@p*GQTwqs3jgqt zS}Dg0H7n$e_V5YB{^Tb$zwtCf?S0MpcS#~QsJXCV%C&&nCW_TkWIYv&O#?_zS43g_ zugw1_U(4L}%z*R(;Hnw;%l6B)UQRPPPfUQFLE)hZAK`gtow{1KqRE<(*#?zyJKaVM zGm_Z2u zW1~x{Q?Y9XEGH&j!-K)+Czmqdd$fwXeEPkd*A+eT%-WBak)Ld-orUxHg-<2|=`a#q!`Q(%Q+jp|_(~ z`fE_o8%kmYxgH%cwV_)XqpVrECPI@HvTXW*Tb|F6kYv!OIusHQ7G#zDm_TC)$On$7 zK^bsE0FLN{3YBL4{Q6btU#>GHz*+pctU=j5 zCresP7m>H~d}?^_>=qP}8*|^@EvVt$nETEnHyyt;q+V`tUPS)!e2x+zkDqtehpeFA zUzu-?mp1(*K8KXM6X38PmR$dUVu8Q$C>4uLCg<2rD4##?yN?t@9vuq!@_FsTexMw$ zQFVfFdL@l2+AXrBf`h?JP{=7Y>OPGMw+3GJGz=;mW~2o!$x&5IDVN@2OwUz@mq^z5 zG>S{TmfYmsQ5dIS0ij8)r3*sSsZq@=1j~ipbb+Kts*R}kND+*dA7zx2@U$h(vm)`d}0V1bft6bo)^Iy9m(7iGH@kkZ3UTO)1Cv!SF1rhu^|7Wv~UqT?e9UM_N4 z9g+1ka)yWX08TpQ1dj=cZ|Jg?nBj5hUlQh!JI`6o?0SnfL-Kjkzk#SL0g!<#Uz%8bJ<747UsMug3qj8lDj{;lJbzO{-IgRKZwi)0hLxac;MC^!(dXu$2{By^LU7EE`*1USVy*q(rqVyIwxiRZhq zzxr+L@BaJezh3mNbOpsOr%2p*M0=n5pa0d@H#bZ)8?E@2VYOs{Tb#y)qj=jbR={-> zyOtuWs8~I3qojwfg66~`Fj59m^1o9}XHRnAsWa!Si;EJEEqdO)oPCOF(lqyNgMT%7 z$Mm^btz~lQrO;P_^-CRe)7I zyF@l}D^qe|g8|wO!=eWUiUp~~Oe(gTz7=-%)zSd$5<-u;S$%%;UPYfYIlNf4MOLcl z3~QC1RBiA&PddQ^xhXBB_s=xSy67uGdT3hN;MLDKHyywoSC5?%OCUJ-=!5^rYa9zw zT)0rh22u(`c0~@BsbtU2R7}}u zKi}7v?``a#>_NoXe+%~!#J%79+E>l1%+6`N+oalsLoHV<){T0KJx!65R4mr2<)~n5 z`owyRcw+$Ed=$?rar=VMY% z9dK)rogJ*RMnzTA1B2E4sBM@KS{s6upZPpf)iZR!BlSup=$T<1H~u99ZigVqeITkS zoO5>49*{gtL!C2{Yp54M3H-*rBQ-js+PJ_TC7-?|E0>_1;;E&{a~vQ_2tKJd6G*7JJ4yi`r&IF9jm9E+bpEA|vr*njD= z{zA-HykHkLJ2t%FBOw`*Dj?GZhc<`KkW`A(0x|vhz#FMhOM+8KE1VfDtvejlEj#0J zX}$)pUlbSf8WybMw=wdK(PU;%_@Op@;2QrJjVhD zZ5@QP>OHz4>s&F#G2bWq|9ZRMr~N}ZTdR1)W)|pWT^|g!WkBx zNGio9Q)CSlYtjkGBoCzZveJM8PcRJC5Fr_Fqynt$1FJ*u$7v95U|n-5?)*R2qH@Da zJIhpGIk;s8atv>@bx@;?rP-+G3(HNRXa|dgnHb_NRUA_85<|PZuv1YWX@L$>r{c{% zT-n+jae@k`o88o^-w4gtW8sC3N6BgzwjOyF)?*9BZlp*W6Qq%3$)C+zzczxw?ra1Bh7hd%js1PznSZF4}aE~pE;S4iT-Wh_Ovd`b=5>A zsk8vc9*PCwvI1cA0zVT49SS7Lg2Bdev=YTHYZ|8Q5}QOGfGEl^qg0_eH?K@#Iy2yg zq7E&xH9neLI)(HRjKvm%oK=S85g3a@swCF&#)2zkY}pD&hhe!v#_NpozI7|I%Rkq@5nBP7Dk^Y*O-H2%VE)-1?Rwj#Um-qr@!F>(YN6m}!Q!36T*p6(Jq2Nr0WWddo8(b!;h&(C4NWGKxMPr7Bqn92tjBG<= zq*5yNn}66qnLXNvJ`c~3M{X{=cPF6CJglT3nPS&aB+(R^$LKto3*@zgP`X2d>2`iU z?F59KqdD$2PJozLyz}M{tviKW)_&ON{XxcS6ezp(QsQ*YjzCnS%8+ab-mkjE_AnL)JicRM)3(S-(pV6Vn^nNI?Jkl~7JQ71YfC3rABo9FYRjonlBby0jY;velOg01H9;D z-g$Cpf#&kU)&G^{ zimat#3qy)z6`pAlunja;taN)ER_+DQcznDZ3l!J`p>gtZEdH*Uz881O&7e5>?eAP6 z>z}jkvON}4zk^~SSGW!MMtO}sN1$@#VNj>AHnc4qdcPDGcspbt)1C5jlj>#nmtdHZ`4}4KfK%$BrE^v9=JS z@)Ehih8s+(-&=FW8YW!WR$MreZzBn$3*7AABC899T}3~0c+P&=1F1P(i0p4BlIkOp z3y>vDd8u;Y73n4*R!E8fhU-cQ@fL`ji>uLs3XsJWC!Fqo^M^UD)}E;imrkduB3GH@ z@Q2JYpK4Vd{n5WY`0CDud){#FiyQI3#=f^X;-gMI$?YrZvO(_GPv^e!ycejUyg?1! zC+_z?Dbrw#HG|HLa9SefX{dT3z&i2N@*RKkGVfMy{@>RBA*)=t?8soTGuceB;4*J8 z(bKocx;>T)PDJzq+k}x`fUn=Wf^$A_ zvNaQz{&Z7rfZ5hOWKXRkx{;!0E}WA&Y++G$QEVYa@~Bu{V7&|j?HVJ*+mB2caJxpY z_5B_tW#>Y3%E3k)e2wb9|jq zgJ`(@Z~~5OgXph^_!F#)37#xmbKxb(Mu9;)iFV|i*Bt526)i2SRcH?ea!2IeY3_LAoh(r*1K#2_Q9 zl6gLu{NJ^(X~B|{uU#2d0bq*M?tgvHH_-zbO8*xvSn{`I-XDi*0GtaX?)Z6ZS;h*H zr#}B*-`w(7>-txh4aD21bv+oYWgvzlZJ&*MpDubMZ>i`c6n(sZo%x@;ihL%AUPdaW z^n33Q>7kKL5=1IhH^bKW)B?%UhvY-jMX#HeOCMa|JOqqk_7Mh8ox!nOv~4{`2+>ar zuAky&UH|%mSQH!8vWJD0^j_&+#Rc)D&WnnGy9~~(R$@F(Prn- z)41w;=JX|Y-BsOq?n)jp?|@3&^ry*6ZqXJO&g|q^?2R^3Y#K#UQHaAtdeQ`93Y({O zs?vSxpNZsBC^SaW3*)!bDc-AF!?*E z*m!xIXssZbmn_MqOTzNJb;{FWTY*ZVfv%$$>3q`Viz+-9L6M6ie1)`5r>`{tToA}RD!&eCmo~eQd(l(-x&@v}g zI@P7nV*0HA1#uo-49Ce(+}uMtc(tMRAm4*`B5M+c=kJpw7H!UDvNdq%nIj~Rz9wD; z#rIggk`%o)CdC6e&q16Wyy5&Hd>(BRN{WpU@s30ImsmG_8XMxdAzYF0{i$wdAQgS_ z>pzn?ccQbHuM~<+qR48X3jj*Jv;J3u2HYxPCLYc%<~hEZ+phm~d)~ZV13CC6PH*0I zW8)P+Z|r;CsVy*>TpE=xH2_Pb<^xg;e2`jZi|iEH&r|?;FmRe6%Sl>b;y4IBnNemB zYKL0_2ZPh*aGX#(al@J>#=0oQ202lyv|vUSzuWIKQZW5sF!tY~y7U;WuFe09r(Lk? z)jw?jCtc(8>Rs1hw0wBFLTp|`R&BlZE0W~G&IsiBhILA8r`RnN*=V9WL8H+k%i?RA zcv=Rv$r3<5@OVhF#6TlI!Ae1;_(&)+6gVfTHv-7)0`735;k4+OrfJZ{l!O_e2)UWnpj=#jP>ZZIOsBjSQW6Gf*?c9I zlWG|xzN;<^%`)?R&_8KCpUOPb?Nlg)3sija*1(D=j;Y~iU2o9NLp1z zP7~;nyB3)uNR2YePKPZc{rm@CfqPM=DVJVB*7%eI&$vkjM-wMnCMu3L3Ulc!29^yR zFI%OPRn6_9H8^l3zlvnge~;AYBd~>{U6>J^3>)Wbkp_B;Op_W_N%#1{*k7=EAhTaF zr$v@8Jt;7CadDqri>Y?X0nRhG*>%}H(!V!;_IsWgcEA7N-U+h&g^AE*S{Shn6q`bk zBr3LETn!ZosB5|}li|0`w_Vvz^iz6*OTzviE=-V1 zNc`JN*75u&3shX!2_rV5_K-i8qq+)~)~@x;i`Yyw%S4xaZg`+WZu)0jAz-&%faT5U z#$eblu)XLh|Jdziy^gqS&4G=TpgTGf$|Op zr+dH+6Tey@6K*ym<1^hHH%k+vrZ&AUGkc%n+v4Am6c-Ma z?6UAD^%M)8v zUGm~$zjk3-V1cL8sY|#*j>~aWOPJCwx#X-~K^ijM;<)Enibe z4shGRx^77oId5TbPEc$uMUGIhhsAnG=(Y=yUpNlhy=tM|>lBTJ5M}d<{dUOi$m)g2 z`k2DzK)GEub6nJ+$mbRNRjW`y`C&O9El-M3PhL7#)I zQ92Yx=JWpi-urj>2UQ7zt$`_`d?@IzS8SQo1$oZHzNx%)d9h^m80a!M1MjhAvg^YT zjyV5Z{nyp6ykUmY#NJ>2oz#xh#pA+&xo!cNa}?W1k+W26yYPx=wXB>v7Lh5r!fccr z;~yuN=wtrn)K>q^Q#%xRMyEu+;9LMqz4(r7{j25Fe%a=i@_bH&td2QF>V-ADjWbhc zpA#$(J`%Ed>Pfns%J)y_KY|=U8oA+(khwSOtSn#DF5J&QCr4sQLu7wIv)Ulp&DSz` zehcWFZ}%$?*d(g)Z=@S#cn>z5aW!{NM$eZJixL0^0*<*tH_NjaLc|OOW66P-}+)+#}lYd&M2P?Ljv@y4NsO_@FvCJ5b z#1u!^JoWapYYV4ZN8wyH)?y>#h8hUBeAfU|wPs}mvgzznWP>U#h>{v%yWAps=)cyS z0D9`HaM(p|bM3HS`^z_FG20YX(yA)5{e`h9hb(N$PKqs{hyk)W z^3tgFWO>->uu}D^=z8x>vfK2YkZmEw3rs|@rt{0gx+NbejfzdO6J$?F!Z%FU3Z*?X z+7{?3DhF?Ai!4jZb=i65{F!dxhORFcn=e1v`HMK-=Kb!kj?6IQ<^4Z*942}0Ib-oaC2L+|o>eW0=fvDTVs8~P4Q9IzV zzwLbMg4=hxz4cIRmTt(`jU$}(L4=Uju5!8lO zhHsKR47xL|hIfK)pix2kaB$VsO0hG;@*J>Yy~yCP(;&kME3TV-JKoREwoWU$>;}k2 zqOb)hRPdi7s)#c2?j(k0OHf%i-g)h=wt!_XUTr<^tn-}kI&tOASgoDj?2;T+QB;?_ zpV2az-aT~V>-%9J(4)|Y^)q^9M?O;#W-FEA13s6kp)wK-5@D%_S|{a zehHUC1=syAvsLnR0RaD(>&&7(@Du6bn$;z|-Se(UklJ7k*)2Fh*Ewr9JRLlJ{WAIKPPdXtC&?__9~{< zLV;8A%)3!qFUOfFyXL_DskUo4huy@hr&A#X4abX2g&7lw-6wzG(S)Dsk zccQ98E9Pt3nEfI3@;Vy$RG>v6Ic8Njbc_Ja?arxZUcqpZ(eb?^_Csuh#nIxF!_YN2 z(1o4vwLkBkYrC+Y&YoYRLgL6A6^h4cP=N=7_s-l?jXV){PxQ!-JMAndTQae5VyK6| z*_OO@cEUNbg_|vLVfW*Jg)J$dSRg#kreZ(vC=~60+S=^-JAKhU7-l^124L|^Lz*J@ zMt|g|7wn(eJE>%DYEV;T9uyzwc}2Vqrk|sicF=PxOCYKEem@wbl@~( zP1yRI-)?^1Yo83k9c+s%-uq*r=0jfF>>O1ybf|3)ygM5&paOzn#&OcZ;&vF97)Rv< zo-rw-nDJ+f=$Ol#jLC%P=)@n7WiR8paUWZInG@0mc@eJ*d=4X!VswddWr9jvWrwmr z(jrTkzBzQ90A!wSE;w=wLhKu5Rpjl9V_9Zg7#TK}nFhI*Im_=*mV{iVEBJ#{*PRd@ z{#1_KTpE+07hc;M1MQ~;k`AVKcE9(1UNaeRs|_tOD}JR%w1UU#td)ifScVoCi=Xz* zJ?Fb>zdtEMY4%Cfr+$8zWWF#yNtK09QcST>#9c_mu5>r>6Wt9^v$h7xi*(BS!fn72 zbyJ?tTj?GrGVnXSO;@{s3G$X8G4vDqzHkS9Eh5pqS)B#cc01^9dIwz-kfCfNn~0Vq zx;N7Ca%bRX%$Pv{aYn?mAL!h7d+{q~%#_dg<4RJ<4KprmU_P+G%pr;ei?ff4T_!@_ z0~ChUpe`+_f(;Tff$4y{hGOP8(Loi2PT3G%F{RHN3{D4wda}6dEu~9oO(kC+fvhX} zyv?EYLY=bRr-7fu9)-e7Xbiw{;m0P~t=XNlX{Y57FgXN07AD9sV2bmay*SW}o}z)D zlo9=NR%u&lfss8FyOSaXRBW}(D7^QBN;)6%0}1RcDds0z!|VC&F$wNY6;G2A!2O`zH89oh+~!U-Mfkstmstd7I7#&g0&w6?5>KLx5oYg?0om ztjBc3v2FkN8x`FvtR)s+khr}K@dZ-fZ9 zKZ=+7#2C&vVf*_zLXoQLuJIWm@2|2}S8&;~0vmb>sLRqKt6}I4NwSOt+M-2+vlP`NLKfB?Glgv0tmFrWx2nJ0p*KK=V5m zw~ZqX?3@3g-Rrna{opULZ9yi42t%!@WS=W!LxPfC_oR4RK;{fB$hMQv|tF%3P$WCq%H5bkd zpR#~W1;v7p%zli*?xJ((3`rsn*+umdWHQf?Z1TS@$DYv)>CHh6%0%}j^*MF3x`S6o zBhPsoW0+AZ$(<I}y?~>topBrhD&|HX=i&`kLQWsn*jVJO1MxPyp2suI0ChlUN|T zgs~iDnYbfs$OPH==^)GW^E{6eK4$#HeKRsQ z;$~!%(m?N*ckqB=HvHZ!XvYCEQP7b@-qJ+UD{mL3M)d*#`!SFMb6&z`*vyP>)`tz? zq>r{2n~^r;2M)b*+uB`q*?q4K2ejb(Y4a+>t5v=H!>ShcNN5tfT9FsxsBbryX|f9w zTn2Q|y__&XedFg}n{HjG#$^lR!j7zshK#}HbM$7hm+00U%^6Z5=p&zr*Dz;E?RYc8 z+6AJ)mBT^kuydGM_viof*UZbwf^uOesp001xbV*Qs>R~dK(S{ia*B#=W_JPK`Tm!o zL%Bs(5}ZN`M1xs59qf5Jm>g9--3lZYngjtB7Zga6CZXa793rJC9)7W4jnC@2kogKc;NBB=(=S6HTVuetdJo$JXj9PwszQ7&$hiU~WqFp6kVxfy+d3 zB7g+gm+{>RKb{1;CjRl=WXCjEaG#?B~zju5@ zvVR3@pi=}IgUSGOaQQsFAcsz$yp$IQWDDTXIdLhMn*ZVjDsdDvD-BS=1gdHV9k4yKN7q?dzkp-zdvXi)0|ODub&z={lG zuL~>#+Jty{1HBmK9g28ybFM_=AC^HIBdf@!Nk!8WJ&SA}q4=BzU0=)zYR-6I*-14?_XZJ> z8Ha%LV!$nZawhOvJoGnlG>yaFdbA)Jab_KO(l$u^)j!x_y^DT9c1$+*+=v#Pve<8L zfRSFj!S{1{F|SqJmN&UrOY3^iwNYuD5COtedJ6?W5_N=qwiwVsd0|o z*^djSL14sz_r`BD&KU~ee>8U|DSlxhE@v%FNDak8{p=wsHiM26#l;kc=Sc#1yQTuhMSbApF4FAgw~DiXN_Mc5tS8zi%b@GScTCapGw9WzV4D=N zHDbW6-1jI{KJA{=Ak%yB?_8l!Nw`4*kT^WWe!Fz&Y`Pirhkndc=lt+mU(&V z`4RD9&&fk;-udMOGlo=I`~OHDy4m()M#vDj@Gfer1%{F+b~Qy-V#)~nk_SV&7>i5lf_(SI(ZjcY%$tM;s@-oFj0(6(@p2|OQegf#vG?<{`#Jcx95FzW*2fjSR%h`ik*l)%s+d!Cl5 z7S_?(QQE0F(zU*Y^eXs*4&88+pCY&j5`j2ty-{5>1W6NS#EJ3QP6Z6wAn*6d2(A+E z68G}g1@@78k2=`7^fFn}J$xgLg(s-Pf%C0XE_vn`*p7L5>KAde7oRneFUXg2@3KF% zJ?~mjP0A8KAU0^=RfQgq=FvBK)w1$=So7^n-_!;^?1ih1$DMYX8(e?%Ewbak7bI&eY zk6!Um+z!hn22v7Hp@K)+PWG(eqFGPxOv3F9rn*>tb)-Ym@K>IKl4 zA;fObb|Gen?@D*i9qd>)vFo6y7&u=!s_lMw2aXUIhIb9#vkIKB0!TArN(qO;K978m z7mA}Qru2K~!mmmpz|nmhPJf}ufkwg#NDZV23M46loskfFF%|5t@qw(ttOm#ubjdY+ z-VIRMJLp4WzS zAMGKBUzlz8B@1YNNU_k`d>o(&{UGQ>x5&;y@<%Ud7v5mo=}sEFpsv|Q8TkGT0ui;j zHP{oML4T+$^RI+<0g5_QGXw2lefNunFJ?qrpR(4E@FvCglC(U-U z+l6h{X$ycjTE79@t7Fh@Lg{(4sX~0RJ3<&>rgOHF;xi#Y4yp$Ow zOhzz*3HO0$?i{gkcMc|=HRI2|&dnfA{cP?f>rQ8vb&72CEXoCh5YL_$d%`x)SgGvzC8M~76JV&zEpPnRy7HS#?#YEa2W-@d*M@e>Zp}e6We7_2 z#U}S6|0IA?UsR&IPMH_H3j`amK|Y7hCR#=ho_G(vR^TKMS;Q(=G1Eupt1>lN$bg;!1x%W@F7^3J9G4s`tTn5rnD3qM@rcH^cTTf zBRUmr;hTBJDd#7bGI(D1x=vl`n*g%=B(!P?CMw`0-{XgxvaxKokhF<6y_E-db~ z475&_LgFJ#oot%w(EMo_k1q3f3UT4skzt03g8^cDZJMUO^saR@=LK1ZY{Y&NMMY5= z^jfwX3b$9QOTyN&MIk3uI#s*!6KAHOSNWP0A*v zrc!(ztdgxDuHCP46gS$9icyR`mPg)Y7RNAV=XPUi=tYvwE#&LM8)nd79+u9`rPwTr z=&9HezdjJ@)THqOe!N}k^+&SSg=74n zo;=JVY@^uCXue`Qm>c{`8kr9>iBWdzrNrr&!@-g+BV8Qa#M46Lzci{>fkl|(1`~VX zF=Pgu35nrjbArdjqK{^(zh#C;>7)lgCw1<`XfX>HD7J|r4MWw7?Ly4AL9h}sX()OD z*)mL>)GI2*J${vtsl~`cv0ov4XYWIQ*G;)TLenn916b!Fl@nU~}Mb?76 zIG7l05rJH=PKmh3hQ)WItNDj5_vvIiYpJJBoL%Nz99~CGxKmv(zbZ}UohNZ$ z*TgMsUXbkMLNJ7uYqcQ(YU1=hu_+dC(XAQk+OH&@wpA-xb}HA382iI7(bw zfPDDDGvI3XWp223e`E7rYZa;&B!*IG0k-WFyM-bfsaV`BWln8}T9pPl!cFJX?}JU# zm8pZ_lWj&A?mMrLt(s=KEpKv`rLP!KT@EETs>uu!A8fkbdY1r#Y1_|8e9MPh<^Ny0x?ebc{@cfC-# zPtHB}oaa1;Qy-+`ck$JiWlQ|a*q!2T@QpDRkAx+425Unw-iNq-#d57a4coTQOIq}zNeV-C*D49N%S+%h))3XQx$u*(;#ADspdV^oYZ zD5D%ZhY5y6@CoYg%!rFKFHt^^kEfAp4_>0qTUd<-N_K=I^(NL_KD}DHP2NDGTwxnY zn1b2mB6?^}WlSS@xhXM?u)$IT=a#N!8@TIbr^xNug@SX&!p75*JF}BmP<^#pTp zbVMZq@$#84#a9=cCSB=`R9%hypnPu-jyLR83xWP!R0{jCcP6c!ik;7(mWpzQMdvOg znLsvnJTY3W-e@9ke#=^8>I<@3*-@V|K-|DcXuvxA^)iF#7-=5=>@&=-LjW2(4#xau z>zyAB{MKw-vYtRN=Dsaa7Y|loST|whZf=N@J*7xL6>lmH$Tc#zk;~mhFG5OO0c*Xi zS&l6r4AKUIwC>87EQJAT0Qw2$oA*v1@Ei7O5@*13Sc6~yn}kuGWKNT~OWYeX?A1z; zg3tdtd_>}2A~qG_*aKuHp#~(hL3MGwIv_lORkY z3k2ZXxR7Cp=>T|X!4F& z3PL^~VSt^)DOPOrhpy)Vj>o<~2ZrkaII6As{1-ZSwBjEQ{};y$shR0NI8Ii0u$K*W zxFhmbnC2?jKNeh8rRu} znc12LQ~##l+G&2R*uHjL<8AX=M4J`+!qoSpu7c#TUDSiaBBPKM{Y#`a+nC`jD! zy63nv=|c7%zb^1^Z?V_Opsk;YM5&|q1)Q()9|E#K2-axBkxK z&w&x_L&1i3GGX@#gYGBhrR4HYE1!~P4_-=oES8dXO4dq|YZ&Im9us=y&>W3CiDgu{ zX_j}?4P!nYK`S*jFIqe>Jijc3%1PcK`LWhP^Uin}n5- z8u?MtP7&_sKuWxvE}Pm8kM8mAP%h;?iW${!9s|st!_E1hqri<3%-*>1fluUQGt3P0 zqpp)}ui4fevR5O#m?BCBEtGPp_(q`RX0RR-9gU(#TUj+>U38;l?<|eHHf;IiMoBkk zujELevHHZkf?<|?z}MJL*!Y}|-$05XLf9-H6$Ycs#H(&OkDyomhOynZkD#yr{kb(? zE@4(vR_f4PW5+F{YJ`cgGrmqR&ZC298QYXQ@m7T3*llY?XOE4&*imn;j#xvQ{GJJ~ zOD{zmM9E*dF2rj4WEqe*=*rjupAPO7a$VX(pZ~^nVex{S(&0JS^s{3;nld^#?N&>Q z6Z7fdtM-kA4R3}DUa0wtwc+qskl4{^u8KP&ggWPd^zf$0t9zuD0|%Ti#Z%q-Pnk|HYP4SPW)tXb>@9p4@rBmU*A z78!6!bu>PPibB{*2}AYnaz+paW3x5SO^*wKWHr=*4Vw>tXm(Qf|H4;H?tWn+wW}>M zPfsZss9inA%#%qq(UcK_yv;54jsW1uNOp*ff&ZmE+yY@IMlIyJjg2gza#FG>v@dl1 zj5d$~KM)F+_eqc&u|J{>QcPQ9M40wmc*O-F9o!1Y{s1?9TBGD(+>oRT_D*{F zRsPOkkO)Nz4tBc0hcxnwzUzTsfuf{p!_&7)VdF!(lfZqFz^V%qr>xDTlChn$&aCP8;U`yLAU0 zk6pX$bj#TY63-~V33dchO1pj2r1h+7|NOY#r~#is`mit9B(yjM+T zdm>2X~w@Tgndp_`@1!7`b>R}~rnFb6SDg<` zf2)%mBYEu2ld~dE0 znw=cUU%uK*HZXH?Jh+mDwQNy)ZTBpr$+Bsl*@Q-N*Fkqo208PEnQhmVCQ@`?#+ zkUK8`cw`DrOzM(CwYB#+EJ|ZS$T&dcYd^nuVe#{nz>k;DA&;0L#DiCvjTUZj;sn`} z31o3+Zkew`$X)kY8auHX!Y=;&49 zrfaLbs{%2{V32Hi2TLNfk;$Cp;kon){}cXMkS1u3+&r%t8kOD#&Dp9z)M11gLD1Z< zi8H;2f6+R>EQldLr4LTIAEl9>4JqX}`Dv7>$yo&+mc|$LI&tCb4CoBdfG*O*NytFO z|3;7XhtGE&5969)`0m9mACooAVCcDZJOpwFBa#Ujlxzz{QmObpa$T4JOIZT-SfW?u z2W8D#LpovGvQzXFmNXqJLsTg_OqwFc!8Sc#cbM6z=Wn_H=eLHUZ~VI%91{lremXJO z$|rg7TD8o=5T zHS;PWnz-%2`)m=Vhc8MJT;bsbBy~9twE%tY*J<=-SlFE`Y`!9l$m} z!rA#&eb5uKFGPJK08L$Q46yk?Od@+4nvhI^!S~Rc>H;3{EFkW<((laMSdk1lPXXE{ zHFl)zqU%{Lq6|^LXc6W)@wee0u!pyYcQ_1{OFAIBg#RV#wXB4Q$54)8GNPEI@AiJ- zrZYp=^8k=Lf0~eMe^>VVrj)f6U&g+%CT^#7mz= zpXVE*55U6oFd!us3#iojbSgMccoR=^L5jLpauG^sfK!SI4%5Xpy3}W1NGH29;InsH zBF7$IbT??mYP{@&#zv-M!s_dxOTJBL~>~0@~<4mO%uNmKz3!R<$+5N4l?qh5s%LdX^^4kk3t<`R z9G4KG)ranwnX!6l98Q^;8Exr@iBrsG=DS7z@*iZo2d`HLEDT36B`csv9x{b-uahQy zh$*d0q~GT`b=V~DYRr*hl|p)WDenruoZTDK1|C!~yNB)-mO?#@v3DS9zb=Kclo!5- zi>JylxQzPkcky3j8jcZM-Uz7j`foooTz*u)<09Ggg;|x#Ex@&ll7Xzr4l2G^QY+mT zHUOE4ES6r?29{??k`lWXSb^(h`SdZOSG7l%%)nHIcG`+az3PHUJFSDem3N4^e-}1J zOt}R|BL}UmE$FjGiN|KD?9BV|Z(oOianAhB>>3L7Hw!~{gScW=co{p3bzS&yPE|yo zw2Kdj;0{0zYiICU(NRG!8?#zp-g#$aOQ_q`KlZhj?6jkGiP0(Or4&~oOPH5S*Ti8V zh>p$(Y?D1Cmg9~K2VpS|;dx7>F@lBXhD5L44*hrQ%?lHfmgfQZamI@|COHJX=ZOU@wb!Lk<5uiYw>B8Q8G|S zD?lRC0G%_xUb#HHG8(EqjTxK_R{gsP^O8B~UrCr}s?Zqr!XPJ-ol;^8u;p2nq9a

    @*O; zjig?6oL=|luD70$EZIuP2qTZAUeqW$0)futIk#sGNpb~yyqA&(fT?Zriionej*1S) z=~ap1exEk(V@a>{tg=xI>Mr5i{3@fnBxUp5@$Gd9X6!?9%y$@J#&ee*YInh(U-yuH zOi3Q<_>_-1F&;DPV1%wB6Y90*VflLz0uq zGs1e-DbmYc!(9UbmODP@0pOVE#)fV^^g-@mxM`fE$TEqJJ0!Ul(i4!uS~;^A)IM(d zm64<2i4ji1-X#dPX)?zQZsUv05MJ;>bWfmpx%%1JH_np`5016%w^&CDDOo;6a&R53 z4Af0AaE}PJ-kT@y^TXwF%TK#N84Y;25==irNF6rLZQOd!I?m>ha=`$d1=eyouUTQ> zwgKC~^n{au7;V~|1IuV5J9m#6SQ`F2=LgneuO1siv7@(k(bTucGgfhG(%$oh(jYg_Yfn1ydeqZmH&&`nv{6Y znL%z$1gg@=t;9o0)c>Icb(_K-2IX-A{CpM>t0~6*Eiu>ign) zR;jOHY67p(S6xRBakP;`oEyTbK-9CV2!(djS)l8LQc{n627!?e?r}bSVe+20;P4^n zx2%xdntL?7NT?owF0f^Dj)oTsAoYuD1r$8byvsQfc$b3$RB->oB%K%zTBQ&8Yo(vY zX0x{|204$F*Oc}1;SXt*_LG~ETJy4WZqicGhTtlo1(+IEW2f3f)2>EhS$C3P)d*?D zG0m*ab-^VUU%-LU%zExtb?^K6fA%&*?j!Yu-;q_!qW_+|FMu@C$ZyV8N|r{Eja2;I zz;k?%TjZSrK{eFOhq@8eX*|tO5hO;e_s1LVN*KKWkHrBagj>An-jnUK->Lhx85+N7 z`d}uxXNzy;!2noe!MA!w$v&e<9~Ivu9+-#il~B+%3#h?eASY7D8;rVd)XY;Ob+{zD zpBx7gCa!HcN>H0>2BzgrNC;s~0ty%9$MwvESOPXCg^Z;4ra&Xr82fE#(s8x$k&+ z&D^`3Ccj>2;a(`v%TG)i3|XsG5C7u|gNTx23>xfRHV!i&QW zO@qH1t0mU?c`W18z8U@d)w_OREqL{^u94^NF~PMSX@6dbU^R)0Stlh+ptVOhNp4^dr%8S;c4aV<91?i@A)b>Q(?<3RlVc1a z1=CwXI>nvP9e1284Rh$Y9wxR%u`z%LdnG$8+|><~Y#l{bQ}I~7Ve0h-dkz%q?H+Fr)$~Pm z#AMj|O1t}-AmYdWY`snKSd(FAOL8){Dz1}-B^$fxi=X6sjz(%gtyF{ps%!7c$_^i_mCgQo1=iV7GkIrZ!JlAQParE!vtd*OY zxT7Ato3W$mjBNy#^7O2PIav@W%Qv<*|7d)$Yza(!9JfOJeV#JM%{ zG%2zTP%rAxw8v3f{clBSy;pF6h|ga&2Lqr>w1d5G{H_H=%w*ob-_M@9{EWA5@6l{H~+F~5lQ!8!c(=y zmbQSBK}vWR6^{!HHkB|5#pCu?y=~rc`M{f?HaGxzv@~g% zyghrBH?XcWN>d|I1Dz@;kQWJS!%(^RAs2`USS=0RN+COhiceV60K}GNc{}W}(xfe- zw1`{MG2y_GIdT>&j{J_XJkx!u@q~|gsjB(u!(}+(<062GAhPe% z+VWlc*Dn`ai)g&eC-7hsVMkaai+2S$NByGV5cCrUg$8aOT|-w3+UE?1=v8=@31>A`TRw^&gsDH-rlN~ri!$vMBLq%USbifSwCIH|#1H0B=;D(fI)Q5Cl*Y?p7c zbTa@2Nh;We3$99bf=1mZ@;b;NfoltT@pqys!k)zD(*r(3k~VNikfY*y@HmSaGl*Oc z9{Z>{m~&^MHQnH4Bjdqgemks_m5Mxqj%Oyl4Pv+3usk3a>Ev+|N89W2^UD^Pp zx2^|+tyvrcKx_@;@@w`1a_&2S8vaMKSqaT14Y`X_${m@yd=^Y z-6PfnuUDWBY?h){X;eY(g?0=rq7wEdo;nkHH( zbYVMXJ>m=zwkA`f@m!q)w1A{)0bx7{_NTCD?InR{4dRabKN=_X(x+knz?3q4`&ri6Wkox22SYs ze1ok%d`BQ*^~$)MGJ=HXQiAHZS9^oa)~9&*Csibi*#_Ek|I3KZ!iesnWWdZQ#L^Jl zD&3Pn#fK@_i&bx@*ndqLFX(eFk;3*dE99aJhH68wgX5 zAm0IV=kjmE|Hb~>E$d{G$L6o?B#tWS^tV>ajiKumtEh_5(~$mNJ87e`TDo1F z_9l2(Y0|BtyHS|)PU1kdnp-q#Ud$iVNRt2f@Hgt;|Kh)X`MduVuApSgD3aiSDZxjc zkK+2D!{H@Auojl{Si53J@vb(k98%0ECkU&_CB7~)Pf-Z!q^-*F+W(B07Dw&Jh*!qt zkdghEp!#g?ALp1Iln*|gTSCg2IVhg%jgzw$4oW>G1COPaibqCOK3&8c@V%j^!yM02 zUPWkWFjNO&H<6{h1YTJ=unn*y5xkB0A2s0)L2{0+fa zV2Y~CIZEt<1AYD)=sJlGnvWWw#~+L|c^lX$lMiG1*SXcqQN%fDYDPZh6f!G+ogHf3 zg6|8GQMA*r&p?~HBMlo>=#p6Z-YudEK>>teFtJz@cQE83lpQ{$%ZLUP>aGV@kC%ZO z3q+indk0_s)2R=aylpl!H8cOXk`&nr?t1WU=Y)kZsikC~1hkKe$CAM;MWMIZdSK2N zLOLkugT~}o5S|Wu>1Cz-yMDOIft$u8wD?eXnsmFO04zqCf0uXwWJ6M7bo515DIYS= zGgH`_V58B&W}#`)N=dsM6+{-fBm*@TjEuAuHjf>|!Dx@!U$~xT9hz+|5oJe-3B9`< z8oI!&c{3B5sq{*7lm|jJN`vSSQ5(BCaZ)0+-Z?XuLoKLwDw5UIyM50Y1Fbs(fL$v_ zG0Dp4`XKjt6i$EsqkV`@+_V4rzgAJWh~;4qtobc3X9PSrk7S3%@_@UHSHRxPe*Uvr zrH$M_ZGX@<#Vw=CZ<^F}UKLy7mmi%FaU^h%(-_w-NheLf-NlIYUSYFat5clK zqu2Wvh25H*_9iM!X_RV{RIEl`A2bNRMK53L3%?Xqc=O>G-jW{XVB8Lq65vx`h&DZq zuw!CLoq~|fg=y4_uR_j?`}QGZ%QpV;?t~bX8CC|xlnhckkzBHvy;GFzI7N<9@z`(U zu53{DNrXwg{U|F-VG@$uO|RgbRpLFI!j{Mic%qnnOO?UO;+URYF1ri#>`ScMvogVR zxJvL@XviSQg`9B#uQ2FZKxIsk@U*fN`2Oud2@#pluMv)HkAVA=rW%R_T5p_4Bk()1~sAIf#NhZCETx1mr*7@Iu_3Jo9 z+S)_~eif`EY;10Kd-h(*MgHaRi?G7&R-OPC$JlK+W-GfQpdh4zRpos&JcV-!bn2E* zTrsg(Rq{POWS!cg21CxPTwz=s1^1}#i1p3!1~)Y~_OH%aOIf~FdnY@BmG`7u-pPSJ zn?sT=z|)4nO8&NBl_{;fB#1(!_3H-nfaF#lIOvj=pz<$T}8<%DY6H(pbfBbO`fr4hWap606k?jaqs$7 zPg%n<^7*cSiu3>JXzU)Xz5%y~yw<2HB8I$D-p1q784%S)9aN1RpST%N#7hsWh*$xS znEI!{E*nSg8e`{F8xz6Ad{{3kw$5$*iZNZXZ$4~olN0+U{A`99I9>u}HOcj0;6M}T z5%og*C|LzX%Bc9|z~^2ChC;1posZi;wNPNJB^!`taC_O=(=fhlN)BfDpCLfhG(|;{ z!q4Fj2U_Yp51ZiadLRP5s^KZb^YQHSdziS^Mc$eV!yQ!ikX4g*kS24g((DM3o+cK`LW#@m8945G1EAt&DXzxB&v*w0)LWKV13s<#N%%aT|df^#4<_?c+=JdLt8#I;WDt3Q5+XWyNC zTpNF;eG~GdfBXwI-HeOApNUtHogTbY)mdPmf|8X{q!^?XXbha72>X%uYGP}ZtEdZp zho)sI(i8^Z-yC@lnNrxq2sES=YP<;YL7e=Vg@OcLk9d>c6EP;L@%44yS`H?+6Xvwb zb^Nq9U2i-X0A#c!#l{pncga3LZ232*fBDBQsTnj|Qo=8h&0m<1-ChfbTnJs+yGw4bGwAY(&FrrhpxP3j17_vBd0*S}ddLyU|HrBC|ip4WXPf{ss6 z&#hN=03}98x6Af`p2r=ZU5X^u4Pm$RvGNQ_=B%8!4UmEgyX!csCSr%!L{6HbU50)8 zput{b61z7lPnivc0j1G5_=l#|hCFbBLSy$U?G$KUtW9=@AM*`np!v`9+gY!BO=CtV zG!+UcrSpOlc<04k(v(=eY9Ff{ycA^m?-CZ0q{t)eqS%X~RFI=f0qz!dy52nb2*l@b z2V6aGKnk)A5m|2R=IJC_#*Btny~qqL-fNy#eYyM=9nf;}=?);Q>*y^rmO{X#i!KT) zjYhEulL|;CL_!XOf;0Zllxs4PIxAxec~4!JSh78lV~3UP@mzKFV)Ol<#F!o2A5|A+ zkxCEV6#-j!L;~q!N(S2`Eo7O6`-B5@Y1l1xHG2SzNhXaU{WR%)D1t8W(?$-^8-h3b z4s!IWGIkvZ8Q$d_^TT`VI0HcIG)T0O1>oaklX5{Nys1}dz0ue#y8~4FIFa72TsF-nYrPVML(EkMp5E#C$A%!wqh=x z%MS$Q!52Gzjjz-6+xe6hvc{ek(cjR%4{|kg+khu0WBEVV%BY zA4O5W35~E;_3+sJk{z9ngW)R_?NVa_)EVI!&<*MYRR&B5bc!E=88UWQJsMsNtcU@h zoUj8^PWY!uQLzKx%coO=`$4?y3M+fsu-CQNNu=~VGWdn4wPm^BpU*ng=!R>idq;{%0+oxBGfjq>MsY$Q# zGD>FM7h_5aZ$R#dezu^Ws5b?=EiU6cww$&yHV0^2^oc=EcTn+tZyTnjNprz4+>|1p0aepZ%c}wl{W3)BWzBL_TO06goP@gZUGyoEJq;m| z#@3EQI(v-sZgZ|&d6Ah-`NtnnKXOoS%v6@Xl!x66)u=0p`YCQO*02099Qx#6Ig0Ba zVD`xqPhIVA1by?uM6=;g8GfrKOFY1Xedn_SpN|7=d-dx6fYbc-@76kfDMusfh)(s46h?O^kw|`#SDc%rR8B z!{xMx2h(+CsQBmO-;z;gOEKs>c!msl+3;IEm}auw!g?f8vb7XhMaAzU>t#?1?~U-N zq_ytZvW`9xc=y#HdG#3Xc)~tNuAe&gku@MZb(te}CjW^9svMgFSE$sjAhnK+qj9$m zj!pjfTLVA z3z640E?&?Q5)xbP3Tpx+Abxav7?7^sHhvuQ z^`Lu%X^PwPQX&R?R&zU47+LLvhMk}IYeWTLW74EmaVXk>X~23`1KjK4*ZK|m;&m6< z6qqz7NEywTnD03T0|%b6j}%*GWtv!cKi0f1ZONBpktz>fmzpisrDK!~>d5M-_ymwU z)`HwIbh87wmLW+#v@6g2(ejDs<1W%Ab8pXHF;N$`MwZIHJ^PY$`NX_$u9yg&9-xQa zwum+mO}-&4rZ2O5d@@8Qc;x{*MR+G8w2v+pl?R-QETV_z=v94WNU{vxOq|&E746hV z?9^#3(i`O>Al?GQb5oi(yCkIvxA#&9mXq>wN9 z%$k|);s#(?d?xFV#3(l+ZSvk`i=$wCB`^{?$+zJ94v*s`^ zY`p8vOP#X6n`?~=k9B11%rq8wrAagV+d;?}s!lNn|wf$8Jfhxny!WS#l|}3 zCO;i_y{s=ZlRg-(HU)7ZHWOZ~cob750^!`uHw}_RF&qH@Ky-?nnXCZ`Rs**~qBcb5 zOh?bUNsPWVc5Cf6Hc9$apphrBusu^T`qeW^*5a1Kzi_sW2bf_ zM{#%79{y@+8~3armqz``X5cBU7T+PwieBL*b~niMR!DED7G`+&z1=RYG*(eL}2uNK+sY`3h7!oFbbAi_iRW+21~1p!Htdz95_WbuM**=tZUcDiLl* zuaKiLJw9D@+V_^fyUcGu+N{DOy$ZO6;s*Mb>K^+d=Nc%SA16mRYb5Q!E=;GEemn8q z!ml*YX$pJ;djvrVIQ$)l!^fn@=Fq%3WVT0pi1}nQnm5n2*}qK>R)4v6%2~1}qB^by z_-MJjqoNy%Ho7-zDNm#9q0datW+$=wMbH~QnS;VBYCZVPX|s`0Gwjt9+W~?L`K*0` ztDJRBMn#)bfFE^a=WhNbp-65W^Y_>tk)8N{!kmuiy~5rYjeL*yI(9y6gY+_WBFg}9 zz=*GVC-&u!pTShW{IVObFtTlLBMUPMDOo;6a;W$X zpgFpEa<;M@Sg+|yy{aTSFCv3=Yc7;)KweLgC0kFCHB@{h zA7z^sXVMAt?u!R}vILcGQMY*Y^f)yAuO7zbPcU0@sHfVKJo7f_x0P@Hk*s4Dh4Wy` zQfL97?UXDX|Je9GP@;sYWPYyvOc+*i)pKfNH~KZu&2x{i6L?z$N1#+NY2IFWmas0! zjn>0*-W>wdh=s-SZTFo$ZU08Jb&-XgS_X_-nj}Kh=i{m*ZQSw&10b`8I%svE9nvB~ z;s>-9H3_Pr{9+OB8cUrToEqHh?s5jEAMfX%wEI3<9GTH~)pJQer3Y_spJ_HCAN{HQ zAldD~Y=dJKW??@i+eeWK6Qdq_&5Mx*pjYK6>sd9i_0S9GDZLJ~FMxN#>X{C%1(=RH zWtaG*@{CXynXYUUV~RATSlF$dxFX$0_^R!jGd$Tq2W6>=# zmztBYWBPM+KRf5S~F6x|x+G#YVl?Lhc71 z0cC6~OD~O(rfhKaS$;xq%t|r|a;RIhS+7dWU0o zJ7j}6=ecVvv9ljSuWAo(53iVorNJ6`AM7Qpmpyic zvC~VrR#_6AqeM;OwZ4n6&g;=^ld82ETi-sJ+B&Z^7<;v@Q9+Z1(|nYKyb3M0-M{`g zH0d_J{H~i`+aobsZ&uWtd|CNGRn} zGLV|iq~cBO0-IS$f+F5dS&|@E*&(TkLq1W#^cIM)=P3?{BUuQoh{=^Wpg>vlGn2b$ zt*c6Q?9SpCaAGnlc1Lv04Q4ng{`uE`^5tqmiQU4Dy==1~DBmKwFnP~g|1Vq*hl(~_ zg9v7$Pbl(; zioeQVuei$3Qm9d$u~C#wdZoR>UTK=NRazTX4)UIW<$Su8tn|JaRms1?TFT25fXqrC zzmLB}wM3Oar6F_|DEehWI!Aqlw2}f=JCt}Gi`Gkee6%uSOSzud0@g`oo?_Swza=_8 zHYaS9==eo63XH1r6c>ph6qO*2f?dt5boi{!VBbiq*=a5FVBkE1btC@C8@!tYMAj_2N>HX z=4NJ0Z+t;)Qv|fS=bLTQ>TNgvjjZ=z+XOAbM)DSn-o*x%c$g*y?KRR1rfPhk%NV%hq=&!^xZnHmhyI zk|&dKfWnY+DGqR)i{04s*tba=cj=pFjrX;YJ)BjO@+NH+nV#59&j3}1BJn#JU?0T?@sr2)Et-7eV-0>!4E%^;kSCfzh?Rd~B3+ovmV z881tyjob*`2vNXTM{oAo3(3UcXwyBsv+3{h*u61MLt?}%SRFbC7OTNrc*+bcULW81 zv(UUGN&fQHX0l--DY3BpnUoB~Z_=^X3AW0~X0M;5F*

    eD$i!vL*g$ipsF6u;t-L zfl=Sc-^K2aN(w(3o=^9=GDkAX2Q-F>7-clq-C{No9;@)z{ZFbNE;b|N@{gC#A&)%R zY1wG8p-P+}TQY$RQSnPfDZFfG>!jnGMC?~^219!2+uykXB`B!9lOEm(TUkHIf z18oR3>7ggF?uY95LtYn2IU9dAO1e21EFAJGA?+ZAhI_B&euGiR$f0SEqlUaH>5T-8 zA3SkhHRM%Cm%au387v$xC`^6D2Ins$sp&&807!FXTmoy*(h zi*I}}wpP@l=4YkWy)*5!s9Y7EGry6$d0G>9#hi9Y4V5*k)W0|8Q@UNE^{$G|7L*9@ zNL#>^;WJO)DfP!YHPmjAHrjNYM>n$jMY_nQNqYh=N*;jP3TB5JpgtMt(T}}N%Z=%M zgD9EPD!n}4z(x9-t|02{reLf~$L&nSI znd?A-AG=rL(q{TyTZ~j({B29Tk6Y*>jKKuH?4`Ks$_NYk1u=oeGydjah}i`Gy6ngX zvWppo!h;jx^%iS=B_-QK!TKMMHTqafohCI1`ra`>;VbUtu*-a^pe7CiL`%tl&vhXL zo!0su3`ddQ=b<4}_i=S49}}$561FI;jeay|o7|XOJ@7UrKwKPTbtU*mVAtuxfB5+u z;@8~|ZxS~{L#ktw)XO+0=3-Pm>8(b^6=LcRzOW;zl;0s)h}6@|^H)nDD%mZ?TMO6u zzqqk?`qfA^K2hp_dA>W`ahFWL)06%30E}$k8y{~T`kl3A?iYk>lWk#QHc_%zlg%j;Qrie!?gP_Oh(K=?Z)#Di`cj4bb(0 z1<)q#1aGUTht|A*-RSw(#O)2av|#aHt_eFN7%tGOHc#H?cX|FZVg5gKf#-e|;#Enk z3*;m9)rCv}$plkQxn2x$S9&QtKw!pHHg|I?N%G+R95fah5ro!IvTYREO2u#d>UE(u z_H0CU@Ji_O+$nAc6@km)dKGA|(w)LA;W6@TUe{!3BSiO$+F3h;na&-( zbZuA(!B*f8gdd`;cbf6NF#@!#J?UQ zUD#!&UwW0_!L5+)rU!h_Lem&*0gD{5TxCJXT16eG3YGYO%vN_qC9w|9`Is(aAM;V? z&qR&~_B%R5Qep~1dL=2mA+JM$2kG6kCg36@O^ew*H2$roMm6}HI*WzBG>fH?U!GsW z-c2u8>SKl^YE0esk;(v#d=c(W*N`odZQSGZVBn+a^{i85L$LZa-28vGM4J5NBL3u0 zDdeRF!=?)R)wFR_gA2XK$zWNU9Y@TUgm0Nj$x9u$K5PD`bk@S4C8QO|C6>+2jscC(7$>A z#2+Hfuvneevxk&0+h2RIMFF+75$M)ZvI7*UrsA8#ZS>RV$^bQ%x1pevi7NYq6bg*h z4f%93)ajOZ7YdLy3-Mhb9Rh#QWM;bPtC7oPxpXz8M<6AEQdEiak|$*W`43#wb2PF% zpdSiUPAS1tu8gS)%qBM!E|73`^lM2J!Z6?xVjOI}`gJQx4OP8X9l@0$Vp zge88P0)a(}4HY4lzlYr}EtedOOB4@BfwW%Ok-)>SIRrNjSw8bzO^hV zTQ%RhOY3U|AUnNSGwCM;lpeXeAbggi+~Jqyouxpqpv8fEb3Tzb0bp_i+vhY*f<^=Y zF|BucV66}E@&zmA6eu(Np^Aom%Ez4y9w*V_5@gsK#6DWc1vas!7c)r`c(5U|!;;Pn z-sO8uc#m5q$!4Elpl*y$58uGih3qS%5cOV?}JuiN!VMz_EDUeJA1n znDx3hR69KVDa87-6j-8ji`_N8jg3B^nO9Ho=g*Gk(I~dz?e83 zZ1xe7N#B{18FY&5@@@2~7LAWtd9@9(u{{|FW5Ng{)bCEMd;GeySlXboOb&fr*sQAg z-gV)5)oD~htk$J)WK7+F^!ZBnKgZBxK1+{qtq&RuOCffO7 zVI$Xl)@4j!`5ZrPgAFsVd@xw+>x^u-2Ro=BY&IgEl26G%%Q>5huV>wjI;qqsmrtyt z3xO@XoQLVV6wW1nBllrIZP@b3T5nS}?x-LmG#gf^J^*5C)Dj4lY!%&&TH)rHq7%4s z8>VpJ+y1wCv&>7+yR8N5NFlSxn+GH1xCK%UP_k-@?8WklAxVGi9o7->Z9f*D?DF&&DGWKWUy^=P105-|_bp8~u zO;CaNfF6>r`P?+UIC34H{&mNcuR6oW_1x-pVuz25WP~O$ z@^KF=SP#vJ))3wFEq|?FP_@8FNjNcAE59`}f!9TUl=j{LkPHABwem}>eQ#cez-~PY zkFQ7G^xrB|chI+Hn$n;hbP1{?eUAd~p^M)eb0|ECmCtGt zmqC_TUB_*JBJT9*9m*zS-!xQ0tfaM(H-uGz8KSz-HE(I58E2EYM~r8s{2@u2qAW0N z@@;xW@M><6@TO#Aq&k_G3*3q{X_{iwn^(R2eQE@TsoTH~Wzze7^pX{kY7}Fv09~~M z(1~C@6y2Cc01Jh`I8AzrltgRfLz4ZIOoP^j4fs9es&}xuA#$AWo#Bh`Yvi}-ipi!C zwoG0!1=Henq3TxtO>sV$F#Pk-%2oQ6Nz~2a!{IpWYsh^1F6Zjecg^^lv-s0CqGje2 zc`!0tEPSF*DA`GhG*Iy={(FOlW^~eOl-_KbJP7@43}Dlc*pMzO5URD_1Hz)P0%1OQ zJ;Pr5z;56>WC>2jcF`9}qohucWe+qO5MW7hf*bHRjGBc8kc!k{Bnq3a5e@Ff1p}ooD}uFnonVjh!LUJ28Ft6U7?e&AyGaj&IylX}uF7;ILP^UY4Uw5@;j0 z8ATrPDWJVz6{G>wLJI|zk{!^V0SyN-&hPNSm!T6Ty974O0AYJAvA*gGpfK{a-=Mbd zJU`V8ALaVhhsX|FzP1Njk6H^zl~6LsVCS2B?K8?}LR>oreAa?*jOs|m3vNn_7d$rh zGf3v>!gk8A$xCgjjy96*cCf8{U*%;Fmm+WJx{e#;d zSu?GetV15`8{6Sv4a`ergLi8XEt`3Xk5&qu8|)eRXhZT%FLr#j5Rm zx*Fm{+0)v%XpbsmkiCb%dca=`J=dX@1ewF?HZD5CSmB+k%%ppzP2wYxEJect=K<$5 zw5)BD>D*~_>A!yRPhYOh)p`F$(W9xALB@+JgZYhWC- zUd{4dVM}M2enSh$I}Ur5v)0UQgZTBZ7nYEZX=}yQ|FI38S6v+oxC8--zHIXX<`wF- zlq~mP|7VxQVwFb8Hd15*>R6lh+Aw{y(Vn1%X`@Zn#qDjRUi#1-wj3|L>D~+W7bP29 z_X!v8o!;+l#>GeK3%?_)m~FZ}II)#yfugOHER7-?sraMeC?2czHl@jv1Qh}I=>k@@ z=*ZL)BnfgA8-1S%onKcN%(25j0=vLqSf?LtO`FNFPC&fO+VWtZ*G^Kxw2j(M74o{h z7hzJngI-h;jl1TxzR8@lH%;*q{I?0haJYeXK3?L$WH5ojixt9wah-j{zCH1~b@AV& zX4|nPCHxZEJP|U-Be$-3lniWnmdSLZ6wj7-DkC<9UXR2g=6Y#8G`&Zys&=W4M&V`* zY*q5Bl?OxcD4*8L56brjC3w5Nqx6_Qd*dWH`{|DUzWvYMmM!;KO}^_D>0h%Z?lIX7 zd9aPKL-T7AV@`YUml}CNz&_>TTxGGSh#s1=o6cetur5F~QrWx~kQh(kVejF?6j1}sm1J(Jv1^xf}!CbR3$&~)QknHnd2lZ16D|eKV9j3@3s8;h^FU#<6 zG6tVx~ZrsSsNT-Z6-5JMq-Jquer)U$de zi;jjD)5x9Z5)ViV1t^dJO;khgNK%!RF=Y}gc*D0Bp@11Cp48?BpT$ZqS(C_LE3d_lkU1s>4=Q`e8DXWQOPs1yC&#Xr^+LD5 zgW)yM4_mzv)Ic|e_KVtl6L`3rbpi+srfLkBVKDmGAGME}we|B31)1&4pzPFYqVZrm zbI`)hlu|MP$8IVfeI~uUJfKu^G$x5v45ev$6|Ofmu(8p|H&1KiB0n!pdX~T4@Aj;G z!8*TU>|-C~2aLFVScnK?7n_k4Ck zeRXk+0c_9!J2M31y2D<1f|UxSET~f-xdTZ;xdndPzU8{r3k>+=J%waTIf zzD>^9$~$>5#O_&GuxpeIs5N?Q%W+0%3hFNp&!szz5nZ5$<|=mtU_5U_a8CeGKQ;0k zP^ivlw?<{KK1${sUXU;ktMuEY8LXplj35N>sUzwTzZ8nJakt(RuvQFhh&rO0#7P3w zC(4>s0p=022g}0~SsQ}8XkGC7SanZqhZ4D5n1DzPM(UVGj$p`DHiA>GksI3xWzuc( zLP7q_W;t@fhP@KrT(oHPP=NCm&&mKYI31hCtGd6U`Wt3|eRw|Ohvehe3}CGm0Q;1Z zHB#ggDqiz8CIwPr8tHcF69Q<;A!sV{XAby$8m*mrHbff@ZnvKFi)zDC0Tr3h*(qV3 zMqU(#;r~=HdFe6)Lds-#5MA=!^i9?--+X#&=&;v)SV++oF<5F4(ht09s;1q5C6R=E)iZbY7Zf zcU+j~{b7wW`Ed_UJJ`vOBMTJkt+2I;mted^RR@;0#E!~f@x}wzCKD|;D?8F#POXN7GqnKC@ zmrNzILF_L&N85h(eP@(7J=lxhVd3O#pk(VPvYLu-5f#9KbA_dy4>XwppDb5sS23Fn z)6^pzx|ECORlG`4P2s_E`y6}|>U1Vx7?2S9W%R^ zwZBr>2=l%1pEG9vjkUCy$Cg{#kuiJhvjmixk==%T+)c44+kGKYy@b~ja1|YALuJ1@I6H)npcB>k;pq#x*XI{t>>zS%v~SMu&k zkAcE}mG>EPjkj(l=BGYkFJmP|HcE1*R>fg2!FK6(MT+qaAmQp02nJP@v`U&>FH)g#X@x+SWl% zXY}o8ePE)vnDa4I{6C7(Megz)7c()INq5gQTaLgFM?eQ%1Mj;)sM&Hd zo}Y~cES4tX`ADw)$p|cO{MTT*)LKKH$pYlTn_oLp_)Xl(pxwrxyQ%2e!Vn$U2LKOLu|aYF{9VL_Cfv4AJpdP1MMDlGCx+O&f(Q z)1)aePw8@Lt8y^#_H0w@*>qML_W}Jlst(owY!3{LW?4nNOS04KCD1#lo9+^ygq;Yw z0msNLwsu;*w43vQz8mGDmB7(Y-tsFu+L>!6hSAs8bKO71gqE(Kn*9_WZ`1e5MP?Nv z9&A@0TX;1cl&qa1tyDa+2b#osQ0l;yC`-`^*~g)fdHzDY?tnx-YV@U#%IO3@LQVYrd|Ku=~!QxCT$em z0zO1DxL_JNasq+4VDWzzUp3c%VMhBhjec+BSFK{R>cbDsNZbDlUopA+g+bbC3+bb$ zl&qg3kFnhaMrYHc+69aMuNMeu0DF=3wE3fi`(nSlMq9{xHx*yq&U|*qu=K|5(^AO#y(R*vL|X9O6R3 zN&idOM8i~P*$bqu$HEj4rfHNQjxfbw?85^D!(RC3;ddlS&_koF684d+ zRN}Z50%-Xo-K^Lb_LNqy^DAey$aK(?{3^f8_bL2Q$Sm_%>ec0hyXn^clO4wc|HJP_ zY=g@AmC!EYa5ylKaNclYdSWfUY9rz8relY{&2IYr_J(D&{CVyoQTD$VFMPRjN+vJ^ z2Nc#MQyD%O7S?*1x>b4@{5aEAZah)$Fv4-Eu?I&tT>s(Z_e*K>de!&C;8SF&2S;$S zEPS1flxzb<)=}}O$xvuyPGV!EyY&x~R!M~RXhb)q^+qAWGIm>QiRJNhTf=q@8GFpidh z89WNRimt!zx(q$@w~Cg9n*ME8omc5q*QKe!z`C8kebU2#@mv`YKn~-A5!#%8m=P|h zb@>{Wb;+!qT2o}dn}yyo=?lPh&QqM@LxZ~C%O{Q8g!n^9Ee*yg`Pkss8^g{>9ta`1F$KKl~qhM9{S>f>%#xNDSZf9cGrdZQ)(A{ zD!3u6o@Ed{o7XjYW#CzQmv4>WVZgEI7T{JVv2FqH8raq3JdIcKl(%Q6%e!0^mVSAz z$7&v*L+#}eU2|pP)9-{jBgE^$Xt5*2+ed~ZdQ~00hCB>tlQ^CP?{cstWbnjEmAk~H$}E_TlvLAoy_SF zr^MnDl`&_DLDV9u6uEha?F>#_00+B6uuTNR<}wullYZTqNb} z3vqkA-2(+@5b(k}*qz6*-(cp+e9&3d^S;@xOz8dBd*tXB#;)A7uq)>&**S`wrQ&Y{ z>LHhd8X-D%Q{*r|lZJliO=4XnR_qUZ?NXSkrM2FLq>Fzz z4BJ#*jqHvY_QLz^@~eETiQNYOe>hwtUp_<2VS-nepzsA8UwPm#EVI9n{fRThL!P@! zBzB62aI@7Sx;1x5k^-fP5a%uAxj!OcV?G>0gpK#yed&KDKIQ$~Y-!rR|K0DBme;H@ z{xb`Ea+i{IQpA8N@;XqID&-?B0XYpVA|&S;)fTX?%~^Iks4i>(WxlK4`+|&WRVML| z!-9*#0rpyD>NG4nQyZrC6TGCER}s-7!oE7{lgjmqL*eWF_k^j@ROBd8{iBz?hmEP~ zHE$(O#>VPkX$FAAml=T<&@7?!70NAuIE^=Cp4+82qcCBO!i>-%%vi33e^;vbQA}NU z$J^jML1Q`~0#?Rg2>O^!ZkI!tdZqDj`hk}qYJaVpH+=gy(Pp&qCoR}T_IPma=!^xj z>L{6(A_u7Wi|?)a*6H_6|LRpo$P-fY)!!dqaPZxvZ$itjpYQqF)^FcfFyyuJ8-{<- zFTf^XnRLdyJ-!>D6>f9n25Dw!1O1o9m2l@GtA{R`TN{@A?&8yMS~52~><{G&7JmZA zW$c0Xul{iH|8#Rt$r|XE=-M#7swFx*%&EPUJK^#Y1{hqh4;LF1Ow_EeY3_c-458(l zzW)<)$b%tt)dE6iDH(7OPf+pTbnr%Io76?Z+OTH%Ibfv>d$n=%={{hrU`cd4(0@^D zs7z8lYuM{!w(0Ctghu|5Zu5occc-{Etc2jtE3A5ufEo6J?o2)=ol&@L8pU{kZI?hj zCKBtRT14&A+DTjfO6#33N&zw8b{6_f%)#YOM2#>@46tN+lo6sP-1*-9XJRv=PJa74 z*T}}#Yzf?BfwG;HER!PJsrXIOB|M}Erzwz?ht*6&l1v&yI@^MK=zE+*U-cz|r@hb| zH%Wm0Ff_&hB_ZZOU^+u%1P;?*!}xjo!1?nZwx6D22FZ7e{^dW&c3U9v;F#b63y>63 zvI2_aQSn;@ho`08xwJ0S43~!YHr9}w zFU$rWL|{j>ZQ4u8$|+I`YK>63lEg`ZC8v-#7^Ox$mH7?Ls^>hKg-pqh*=ilTo|8)# zg**eVTf<$&Q{M^5r!NHER2YB++$F}f4^?B;24MOEJ6Nnf#4q3#2WjMkQQH(ip>%Lk z9C^%HK)pZ~9ysy~9en1`=}l#Cn_*Kk^N%Y@kq0|BCoF(aOUd?AWFHlu8jLL=wcbe4 z)yT10Mcw7!3$6>Q&Q+31Vy=p6o?H*<`r?=Yb^;HBI=v9MK?i2xtn=|23JQ%@Y^VcxGXy3rtaE;%n>s!0k znGW0@yn5N`zVd*?JtC{7wew z2IGYSRz_$y-7L-s&7n8F)hfL;SFb7!-V6bvOz*w)r$KsEr6BK}^W38%SDG8+yt@ix zj6w8@7tSR9;2rCjpvUGW>_qvR#Jxdj(r#A%%nP7Umrpa*9(=_|uK-`77j!dw=*#m@(gUEjx<$5%)vVgWx+mT8 zPKIc+&tBnql}27e@1NGiA0CfBISv5F!sjrZxDq0^$FdKc$c-nt)+x;|2-(I?0@N_| zXt-XLJ#9#W@!R8)F8;3R1JVt_iL6a;;=g#NSAlx!%$osgzjb;3W#vw~e3D*Oz)Flb z5US;%P;XKsKDFHEro`pzi+KXahNKbTvJC{MuQq%(J8`<%we9(XY)9n5si!&% zN4SEL0fn`gitm8DcoxT0Kn4xafuRcJVOR`?bcvOI$($Z>rr@!p7pN1T1{h1ussb}Z zlR2mPb+P5FWRBK5i}QcI5I)ueZ3~OaK<*6uXyZo7#prPwG%xVuG)^oAAZJfU`I|x0 z^7h5wl2i}g4^>#eAcvA=Q$$0>pJitgRMxJDI4R0e?jxo)kJ$NSwYZIbG^dQ6#X@N8 z3o7M*A}^JYU#dBC{lt{FnZwevs9Xw0C&-uIK^ue;vR zX?kDu4IyNOAWc&z8j_#}g+{(idH~|}%VuSm+o04a*99g-6oORhvRNHUw_6UJA%+k! zE)t{6tjlgOf(Z4K>s1qGm=_r@fwG$9P9z^&*z$dp42*jj6`#P{4(ZwtHS%tHFrb@r zaXKgr218E{kRH4M0dPp-a#|q*n!v-bR-Q7Oy?OHOK)njNOK9SgA~yur2VH=Y_nmLm z2c>zJ0UZ~Kw`uGif;l)BPuf_f=fy&9ap1CDxM}M-W=k?N{RhX%3J-Qyb1Y2UW=gh+ zBFQ7$Pa9j5pHV&&K37YH==)%l%VO@1ee{k@^>b6@*odxqijfVQ(D=_+-~Yf2jO9C) zP9xWCWsf~q4uT9?faL)tyGN1RRD7Buk#*HOgLQ(pETTHBGNxKotXS9)T@iOaZlQ6j z+ayL>UD>=l;(nsO5NQyhUP_u2p2??8Pu=l3A9wxD>u+Lt@p0smPc^MKdI} zY-QCz_sC(d0{I3X47?yy7udd8EYl54Fx7zwNe*5{)?XX+z`}RcyM50=&S^(LWpsTI z(*KH#J?}M2Z2E;(Or0b!^~t)yza+!|4voAv!1Epr*$sKW4AT0+qp;R-a^d`HPSqz>ir*nVQ*CPEfy z^R%1%sz7w8ZU`I2*aRs{-o@_zV`m2(ht|rvBjD`k{5ag^5e!!8rjt(V1(!Y(YjSc4Ip?H9rU$b4PgCW(~{D^{Omm z+sRMlXTmhfQu=Thpo&!-_$UVSq{NS0)O%)>c^)gMMj6s2xBfr&z67qR^xWGcof>wAoUzg0STeC*O&1p zlpQ*B9?AFvS^2DablmFs%f7!aw!>_?V$o3!@*P$)dqR6c55As7j`%f6P7zS)q>pje zNe;fQRq6!Maif(a>zCYs!YIw1IbE`C%pEX1%bO&f!Pvyo!Rv(uX|?~TPc{_nb}IU? zr|t2mK!NFe(>Y02=aKMw2~>z{9uJ))yW*!ife&ky=>Fz9HA{~8*M%KY4ewLet^U#c1ajH)E;PjTa%=F`l~9h=Btbb+ z=&dSVGVF{J?M)K9Wjp*{*zOL-V4Hq(Vz7a%vf~Xpq>skfq0N*$o+2?+gua^z8ykzl zA1&A-+%L_Pw*pgT^}?yKIb%(Qh4yFcBU5s0-u{f^^-eG9$UXS7qoVL1pdY__AvBj6 z0(7@SA3ASvf!DIbsl#b%kzpY!a5`*al_gCMWjI@azdpKsKP4>FQbDJcX znI4u6gkh{j&#b$5$^9i-WjUuzpc#CF`^x2x{nd87ep|>kN+ZA947^CC1+(#9D%|=h=^mT(bmdtPn;eeEuKJroM|A$8iNe z&gn3(UDT{@d$%1X^_+T#*B-$sm>`S3FH8RJ7T;sxmv8ZS*1D11mv6VJ>37@L{)iat zPKiVMCuHq-F&H}z&}AFhqOFuXks=9H#0}8~X3N}4rz&s#k)DV3yvI%19y%@vuT3qK zErctx$yz(xi4k2>I=AX5SVQ~qwv&kguE{_R^{ zq>CLs?06AcX(V5BkCNY|$So=&h3N_H4ejPt!a}r?UmV`gO@;#f1ny^VRL*Huwy5z6 zW|8ukYSoDJ_E&nOUy}5>HGbFU;oUi$-UTQihCi<@=uuTHYF6(gYt%iebLvVbe6}^T zM}_5{_rm(gvSoOsVo{|No~>Av=9DqFZpjA+mwx%~)o&!e)2!a%SGOebow9d+iw#Q~ z_^m;}I1|zugug#Buy?_&1(im6HzMjY@O9yLqW0L9D-rZvdO z7_x_*Sz9*g3uDpxX5oK?m@>-0BAM!)MrJ9Al5e6&92K#bGej!rBmOseu|d1UeeUJV zJ#h{+m6R&Z3QA`hP1@Km%0|Z)`3q(!~mYk&B%M6bW?6>eAV6QwZG{CMvEfSRB~$|=H$#qfAy6wFPibC4u7pvVcyq0)pEax7UX7Z%E)0xngdRqAs@xit0=X^@u=PqvD6wDPTD)Gf;e zl8qL*1`jbcITTh8JWz$Mz`D^S(d&vP$qqQTNWqy#F8U)-f>#$-#9(pVLua5-L~gns zNG<@A{Sm(u@hP&-hpijU9wSe{!uEg_h}ikZpDG&NOrySR42~W9$7W*0RX`h_3;V4O zq6NZ^Qo0myXp1=i3n$Go0KIS+tKYEc&E{Uy=nET2*l}ZknYhf>*Bbotg?d@*e11A` zKSSa%J0!z-m~M33DN3Usxlhr;qn}L6kfFCCR-gXo4EXbcFY_i-Uunn4FjHT-`8yv6 zmU?1Y3#O-`mId(;&%MBC^qmDsGVCOQ^VR;tvKrsjB-Jm`OMh#Ps?OE6byRF%V!Gr! zXO<^@$w8|6p=r~ih;V=Tlb=TX@<;D2`@M*kI}pl_SB^bKTbOi84h=;qR0Oi!?svyX zU5k4Zlnm)L7eVBen-GK|i@K&Fpn{5p&|D_$X9uKHlzp5C5GFm#77SqC71{A??}euH z8+IGVHAAS8t@@Hiwe>!5UI&7Kt(%wrqrw5zn|0s zMbi!Lao2YLoxJrtA)vz>QTP-jX!G@8suwb4ia*KIvj1$JC{_*@L1@9Wd zbfs<&e?;o-I1}A%WIirX@@9%Op>A{%5O$p+n$=vbDl*`507^&a`QZMfQ*m!Wx^UE? zOV%x`Cn@4&|5N0?2=+4h!fKMo2Ziad);T4C2BL4IW5DzPh?pnIqRARB&1dofSfF5M z)66Xw?{v+kHCR@+D&QjM2BZa!I&5)i1PX7|I7t`gJE44N2d^izOa!(}1)Ctl-wStR zNi#J_6K=`EHECoS!OYSRFUA@xWKFy9>9$c<1G2tXn|gt4eq{o{z-&3D0&^!N2RYeP zDxzIB;+^kVAig;lH6TlvX4fv)n^39Sqt>5qTiBz{_uRk~hd$6?uC?VLTp_KoWD zRz=4IfiSKWu_h*18JFq*@aE@*eg@+bkoLC(k~4!GGkPocQ*w~1Dgs`9(lob*SL2`R zSLc7hz0`e|BNWHu#~OMA|Cs2e-x2?n{9$LUGRvnhRI_aF;`NKNeb+2lc9y>7o2_b* zCj;;3RZ9zhb|bC)Gxlv6zy>K72P7QL(6E(Jse$uC62h ziXzYFf>#F!%UlAg)XF*WTu@84dCdUZ;yX$=jN&IU8%u2hDT*2a+cpX1-2X~0bHf4BXGth3|YQ?3yTk|{a( zS(~W{Y~&h-oK1|pnjC{kywaG_i~uf%8R_3CSh8jYDC5nzwIu(?Yi0~^GUjTP7yO2u6UjB`LL73AgM2jb zHo)Ha#v6@d>?+9tLQKdysvAUgTs(I#EY8}LS5Lvg4qVR~lcytD_bBk+1Qn(nL9Ym0 z%rFYo`J)asXk9_g`3kC2b_f~?I(Ym1%ltdY5GVs!b=&je$+8?6FCNR9cU-&i&O4@s zbauN#Hd7k6n+DBjNC#izWGkXP)`mb&;vmr}?BT&_P}S2Qf#I43+r9FI9lXz_i7E}= zx0h4Lj4r^85bUOp_?L$5h8$6Ccmfyt8F4JUdKN-)Rdip3R@o>mV({KPQ7ktX?8{1i zBwtshaEhD{`ef2la(7yoI26;W*a5ffXpG|3*dpO~Wp17AZT21~_G&_q?luD!E5hyZz!JH-R2C zvVlZX-SnPN>#lMZ!pssgV_l^EkNW3-U*(r;)aw_0zCizHv${&~fLsH5v3feqd!28- zPs#i}g7x#TwhkFzt&4%@OdMNyJZC7YUJ9KVbUwha*nId=*I}}2203nI(+^PceH1C7 zB2rO$LvncbI>|oy5r3qByF_vW@J#NH)_lKN-RBw0oKufD9RK>&uSYFyR-c@+dC?Jn zJX88_+26jmq*oRqc`A*O01r3_xgOOwk;*PXiYag2*lr5&x?dF;O%tNiNay$ zyUu0a!>)JSHU)t{n@v}fJC1ijP_K#CP3zoZ4CrAk@3nxF zQAr|$F_Ql6?Pd};gA^J~MH(gFL6KxC;xJPz$>B8c+GXYQy2ZGiZQvbuJ|LY5>ejw~7z>6>~JPOs(eu=QGeR&?;_})Nq$)^Bdrf98RnF9~Tj531_Ex(M2AgSEMN(z-D8UQG!yak8BZNvoGeWC6#JX8*(z9SWE++k} zSPj4q7anORBNrNMN1E*aS639n;rG=r$z?-eYO-w3tS9I97dM<|U_PbYe&w!|XBM%kI zcez|*@?YBnas$K8yZI#wl&a39wah-K6>BtfS8A0~F%ftVgZ#R<&mMDlSb-6gvS8L5Qu78BgRA66Zms7d_g3){ze=tKDb_Jca*%El=*lUh zNf`|hS{@Wq+FaFj5*ERNxt{nftWYw2edz~ZGv&2>wguLX-4HV*oX4OU9*Y&KNuelO ztx1jZ@voYvg5moe{2f zU*~$p9dbJuX2t$PBvR0f_#3LQ8=+|XmiV@)%_B!;^S}w4<;BQ>u-?RvKazPkjML~k zEp7km`!9PvQ=+h95hjy>4-UYByOdJ!D)Xl(sbn+@me~lSN2cThD^N^d9WX2t8n!YS zye(HrEW53Y9j`M5Mq3#zCErewZB#@vFt9+6B5<3c>Y`pBWYnRUX_fbcwh8)NMjfun zj&L`*;5HYvM5bu*u?$XQ&C8UaX!n=`t~N=s^-Hn5t)16L zH%StK(J4u`4$e#wARbSz<-mCS#MJz3ODI@7%V0TtepR`QY#WbB+m7wSek1#^i;}}; zHxs!O+GU6Qnw45rwyGdZ>(wZRT5YYWgNJL*C(2F*a3`SU(5He2=VNzAt8Aa&X91cH zmji*VvTm0%U_@+(63YQG)>=F{wi!CiFFc=xJN@Sj160)A?`z1486?f<+fSh6u@u=r zMF2Bjz(G|!LG>(shCP=qn!579F<*{FU-Ovzr|QuB*F2unXSYQBi&xLxYl?|iB+hOI z+Im+*Q|Po4f}M|o?+3TaFeYOA?aMRBv4LyjqV-0+b3zDxqQL$~l_La1$0h z!Y~1Lt7yp1PA;au({0+TYNp9^K-L8rjPAEmm*sHM9UEpf_+1UIf#%KO#rLHNWPr27 z@52xNoJkjmThv!&<$lA93!S34dCU&Kk3LvIWdZ;iC1?Cv)Cq16NQ&C3t#ak zUctBr?BC!QKEY~M>^Dy~(!WsgvL|sO`PH-1=zI|V0M0P#5{RxwJ0XGG5x-7BJ%O`R zywhO;*%!jhnm4e5?DU@q00`6-Gt;D)J#x0_A? zRyo)-DK%9=t#4idRYXk(FI&|Ek=ylO2dX)y&n4?^%|YKBU|GOYgMP`SrJG$fgY*vN zu(Rfd>k95cUwEe7r%TyE-*By!RXV{(B(O`8HA&V&sY^cp3_$YQ!fb^mp4$^}NSVjv zzzN{>ZltdPf2zLlAMYG>xZ#TBRqQ+~7VsO_Sh6|E2A*H`g?s$c0I2JW;$9=y*#XLq z12AiiT&OQ7`5;9Gs0cJY{oH7#<*O@Q)^LYi8l+j?yXi~w?u%f$Pm#jVe13`qnvc_9 zjqQ`5;O1WG3O5bPNgsmXUKPE{SA&|INMwjudjQO2Kt9;1Xaf1p8ebqhgns+FFs*X^ z_x0kPNW`o6%*s_+s;ZzHv!Y+q^~>Ur6;L=;MR!Y2$%}nKLiQ-9ONn~s_d%xb0*s?s z?H{?=aDFGJTiWK-B&jCL(H(0PYz9S~JZ6ZXDh{?r4l>YLJw#&Uo4y*!KOkB^E6M}M zvRfa@<@LBXOu6aUY|`v^0Vigd>^8mKF8Xt)>>@BRw5ZeNT}(Y)18Zba$U5#3UK*VX z#Ht;I*^(ulMab;nZfx-csZ);^Q-*U%gZ_E!12asThw=D znN}A>vYN?aV~;GjcZo3QE0ZCTvSZkoA(Cq36@^Fgs|4?#UDU|?-Q!Ov!j~kw0K0`q z{@F!$6i`9AI73(lW-N{Fl{E4m`5fcifooNQ7U+*XC+`cd7e2ld$0)F7J(OYItb`d0 zP{8h%{p}yV7iL)IzE{3Gos_;Z%UqL@-*l9c*HENVA1S~Z@K$k}P&2?eM@LCoq|g@N zthwWesqj*0M*I?%!ZcweXOBR$+phq!RY0Cu>C`9L?WY%@hR`(b+3RVn_(f0XV5nx0 zw8$}?g;og}$+{uk)v3q>-jEgiVvj8ztwGh`>rXC1Z6y?q?=!Ijj?vv4_f~%WH3Nzq zgsO6qJzi46j$NXoMp)TL$w56}FBMSg`mZ9B~zW*{1jvCZ1pLohvYVatvEww?2_r-uqT~lCvQM7|$1Rnn+^?VRfU29DqIB*? zSAA9snTkdou*4#RY=RPAEVKY0w@cQrvEBnCvIVC_^JCqzQlOA)&p4btG9Mv`F0nXtC)|vkoP8NHByl_aX-0<_w zX@0*ln3p#T>paNkGf1@2=69HqKcL7xd@p0<*V5Ntn1O!i47@u{u)|ot5Uvlt-aY3M>5|>0SCVU-0Up%OHAygytyLCq`sq&Q zj63=Q-M(n?4SvH)-|jhg9FgN$s~i@^gdkN^Za@M-Pa<`WCL!!gx_`XEb%M>uV^pLnruu*cNI^x&M+zQ4eCQXQx-awYHejzp>Wx~G#w(ZRf zaxJW1h^&YB4}(($oD&PtR@m+C*ZgwvQe|3ye2ZK(E!1GMiMQjm%}fc$UbhSmz84)N z6LM&fP{U|Y3e*0*VMzjKRbcd6nl&Ehf>M|g#UuA@FAd7JAv&*mU-drVcK`(aI$WAK zeYEY27JEDObj5-rtJtLR2Fd^HjtxA6omzTf>rt}ajuWuEj2xR3O1_mMiB!aS$A_-f zvtl4my8iVjNwM?{DGG@RsiD`-K!Xzn3T${adWP+0_nB{&-E=?mxXn(mGH_JfS5_@l z7{Jl|ll{v{vK@m1+SbP8D)K0K4n=lS5qiy&G=`ws){aV)-R0Y_X)CStu4#D*kQ%UEa*Pi3M{O!GVPCt{_^@<16HEm_t;3% z>=-MRMp)TP$-(K(qas$%h~(a(SA+}(pK)I^qlP{kbWso+l;x9XGoJ{ks{GQ%hfg}XR?T~)q=tb zG1Q8qpPukCAf{mS-}jSrJ4Q^E5n_rcIk02pgYV8DGnt-EFG}7bJ>+*K30?regAi&%*!3G)T#<0bCcE~vS*ExUk zHW-t^A4=AcOglCvkPR5)s+3T2;AtqJB9gy{R8p0pTbyH=-Qitic=1C~YraVqq{m@ZKPi@-RQ1tqLBo>DImzCm!4-@8 zAf}YWK_(X{nShMJFn!Oxk=FTvM&gRf&ovV+$r$pS>Q z1vPRa`FhIJ9d0r6u82PoSNk7;y!J)dL17cO4U7!X_a{+=y!Y$hYX0z}U;X6wzYwjV zG0eV|F49&$)E;N6B*=2>LF`ythc1O8_d z$Hg$I7UN_*!(mNen2dE>D0K}n;6(Z-O)A;@%HZS^Bb-!I^1~FVpdzr8ssqUVbuA0f z=Z!9_CYJPuRt3d7wnE`J(nC}_b;-V@w=LZ4n+aWWdl$FL^%Ympae()uE+}>#+;PP9 zNUJ<4-nKB)OM^QvJ?Ssf{Z={^a{53nAeNb2y`41>G#;vS=S<)*fsg0+{*k=w)H1vI z+ZXG}nL_!SOII%hEh~_UO_D(sCVe^l0=(C_+ij3SBW{hjigyQ$Mh{&D0Ttw!L4xmG&s_R$ zSRV)ipCYIff~7Y-s_MBKY~me|o#3u?(JJ%#-SD8*U}hP3%o>AdE?5gs?7(yD=pVm7 z8R-B!c4^H>2Q<0H$cKnlhA26s`+u3^04GNNh%`yk=(P(`6Q}P&!tVd#;y09Ayih#rcswdk72g8x3Y6|D}$-G!J z$xrT{x&v10%e1K1Q({e7RPEL|F~gX8$vbgjl2au_TcIlmYi(P_=(3`a#FPWWB2<8H zF?D#*|0Z^**n0TX+84*>?KskECN_`tYkSgZfd_^9rP{x zmZ$oCWGTXJzpq+|&(CvBGrwla!eWKj7WAmne1^n{b3c0n8#=GgkAsb5dPoaMu%Z+K zc32IG>-^D(uXTy5zsA;=eEi|;~eqUDy#hNTdASSWcDqC%riSg zqZ1}WTIHkp7jvArgZ)B=A9i)%WiN`Y;>C0CI5jZ)L}j98b;~>L->G}6d+9l~R+&Nf zkS6)=@Qtn=yg@;(;xp(ey05yds$+)S`se{pH9_ix&5kRX&5l4EChi7a4e(`W+r~6n z0m@c^uIV7`pd9drge!kv864<8gFqXkYx4O(m2!ggi*C4XfFK$2<8|<`sVAPCAlVSD zhkBJJNwzA_XEis2vkpwC;UU}LNk5O1LX{~LtoXS}+w@Ck!v?qFYww;V8`x!V?AK;O zA^zCy@ODbRje=Q-C<(tQyQ9$m*{sg{db_9qe63~Y)Gg{b_dCGdQtQ5n90{)>u~WB6 zW{VCh0^qqrouIlNOs2Wrj$^52n%glyhwQ;QAxM>g9XvZ-bPcjy%n|?e5G2cZ6x>Z; zVvsxvw$56WUS?v8OCwM!fZ3_@zu}rE-{J{Zn;A4USV;Nh+gjy5rkE}yT_j7TXLLf# zWF3pc7ClB@*Z3cviwa9GcK=eboNur#@0ZT}AF}b4u`T&Vwq*w;Pewx+u?4Kj-#TT* zIu5ABg*~c09AtDlDZL4_ohwv(msHUQL?aF#^-7Sm$2znB!r5RCj~5PUl^ZU7w>Q}| zWWpx!ZpTF?W`Ze4Ige!{z~e~v*Vm1pIJoUANXD5l%W*Nz!!|r6sm*r|H<*pBxzcp9 z-;QJM%|?dcI3+(yks2zZ*8P$DuydMIWl%pio?A+9B9Gj^gl$ZbPp5Db8T{&Tp!C~B z@UGSVgMx>mdkZ4@`m6n1{ndW%*sEzmUr-iSNnAtnEU zB7;juD^kqBdloBpo@3L(=;+pxDYvF@sm%d_H zrCPsnJ%^OpvAc8D$lWMf!*|tS;sv$!q$4Hs0&CD zA{XTG2T32)$m-8`$*zdAWTnu6lwL*}_QplKhYR11ZSs0mz%y_wr`9EFz zxw`>IKRh$-97&l$fUj>%K}bF&&!tG#V`m485yzCdW3{`!{JnqzUUqn)Q^Z4JEI-sM zw{MVdkavT6djSYh40Dc!B+ax|_+q^8X_9j?-jG%~$7)XO*N*R;adTk00W48#1MA4B z!z=PHBdJE1kE7%pDY6bRza5HYPf9N_j~A)`Cyt3F9Y0n9D%(LGcvB}PQ#EVHc>pt2 zvyIY1&RzKl{x)$x?CV;bYUvER9W#^fUWNJ#Oa~zC_DXU^clOXX@Z7zy9H8g9{bsvp zwfv6aLHO~|D$rfRv$jIhBn_d>9G&RzemgYSU_mmxvbK;tGswqAuU-`;hj0D>6){9E zDE9gE@O$_d6c7b<+{EADwTd6hj5x%;hCeHvHaX$XJMh(((x)9SD7s`>Avfl&hdaIy zHTaDy#iBcoJ?hFJyk~{uD(+^V5r+a!AvAgxh&zG1IE!wg?^tUpCjHGn z;}x)dC&#~^m7%uZ*YZ#DKjh3e0ONZgeM{^-Gnc!QGJ?$D)aOj|OI6W#9CD)+On;xR0{RcBMd5iFZxtJ)^%JYYv z^i_5kiM}|0xhXBX-PZUQ7-8gzT79FmR8Y+%`)iO?8}+5n3QkHJ{L*MmH!L4`HNmY6 zx(uOa>bN7M45p~|e?mrN($oLq0oVZ~+d=%Lsl1il0_2DhK*}h2F-7)J5xtUwA&2~m zIhqP#_Z;L_FXB7^k0ejgCpqlDaVA(26w#{-yFeqyS3$^0sRnPq%yjUgsRDZQ%r>V& zw^CJ&jsd%bxy*b5bn$v*;d$#?nhC+?84uh#zQUs>n+(iq7VLM+{7zmvW6E0liX@b?jh3jblsu6l2~{g7+Oh%_PmQw3h1@GP@>KgahDpL=$q@9-w#!M#jp@Mv&6w|ed_;N`goDs($tCXB!d zAcPfYCouj*A7KUB>DRiaJ#;l-<9oHK7szHi#s+MF$6#Y8B~PayaU5~StSPikJ_qesv zke19CJ8W1oqIK?Jg^g+3`Xh2pv0=uDXREqF?RbqddWon(2qK1ueeQ@O`N$X@?WE~f z6gjm5qwCa}aduNMb>)ZbrT}`Fh|}*Qe*UY7U;p7p?=AcN6QXqM;9=sUoE5zPo+;_h zE3)-9L(!89dd<0XdPoMn%eh^Y$4_?bqie+NqN5Y{jwk%}TL|t6k3Z1|k53Q#vCQCe zY)$aLNVeFq&jC5#F+N8QC5I-lbRd-dU)67=LwjhM=KzG@tAcuH&1KQ%pl-UH(ChoP|^W)ZrL=G|_pr;5l}7 zt-&>CmTAtHjrp+Om>@{JkA>Bf4DntNYk9oU>5*oNv5@4TZxV#fKK$dK@;L?2iP``u zK%|N(kwf2-;IcxiydryOBc-`9GdfX3jTyowhfW#E*M4g-CeiC+7m>bKrXXp9QPbNa zN){8U<}p8~RH+(W_JPDENLbIJEj6o=(NR1ta27F$7y->L4v$0$931!{Jt)}9?}wV?Do{w&In{3WfxZKUHTAz*Hzq!mB!jvKUIIt! zb#+z^{G~@7>Y?swkinr2s}jBX>0#$8=r~Ar*Q5!x%4BkNmfd!}O<#HVcM=2sPJZ)S zm&vAA27g6HHY$^nLmSZ!Dx%U!>(vMhgt5Rw7|GY{BW1vv&`qOLzuT`sj8fu2PuL&c z1B?lcydmcn_j_UKb9XpKQ?^|3bjHPAjnusVJpP}ig(=TAklFEaWu{oA&cD{RONK$@ zlE9N14M8^V-lW6~}H;#xJd1zZ~4gmcU!yX1US(=_1)V%jo zlvKRKw2t3iH+95JWqz|VMh-mZ$^JOkWuC=!6LSty43YfX{yC~+Iu$yIx*&m54q3m@ z(5l_3=v1TufwiuhB}>(%>#KeMBtA%Ns<8oH*kfa?Yy9~5nW>%h3G>((r)_MS$IYQ( zQ-fmW9W7CG)5u_)yn(m!wV8sP&MR}=aVt_N{D8U^sFn^lBLnjN*jFZ z&wf1;=vU^@M`xo93VN6s z&I7XQSurk}y>3nLKBIUos!7&J5BVVF!6i2+I_rXN&sDw$0&ht!>Euu%7e^BfzR=Mm zaNnQ{-KBnA)8A39eBe<7nMm9g zZE-FG(cTM8KaFRbLn?zdk%F)hhaG-{!1_`d)H7L57gKav5B#PhPww;A_S}AJGNE4k zHKoK%sP{2XB{IcD@*8;?JV3GLDJ!3RI%Afi;^`ApbM)N{4^m$CQ132ahdV_E=NEAL z>EoQMVf~~L!WvjpeLuJnh$m6$2hZ=K&wCGr)eDQm(|q=T#4?m%GhGmQvEITr?wnaV z_0O(+;|^!FldM)l>Q~?O`_IV`n%c4XFhgj1+%uNj4^ltSLx_d?DdJOPoey*d6)VmL zY4VsMQp*fGgR8}f2KUKU>OLm3Js~WajA4t937@~6YTCE)iX;Neba@Ptf#6|hWER8b zz+uiF2*S3=Th!@HY|zPo`Q{I zP)$`(Za|eF4>CWAUI}6OLd-PxS*cx^#XHMpu2~E_>04PXH`5<{^Pf3?UnHh!bX4H5 z3#6&b{rc%1I-U!$xWoQyLa^f)Lqj(9QcUKT1&U0bwC%^ZU%RX^rDI~V(Ae*;-W;pc zanC3TXjsw*p@>#f<(DaM6|d#rkZJaE+JLzrNjB`#!ONu&EFAsHHhDIkBCewU{JEmf z_m;aRZecEcj0-=a`42c>a8|qa(Y7IuV*!?N78U&|Tfip6d;V{O41tn$N&Q8n@RbRa zoH7cOR8sPT6e*`7>ipNvNE6n#X*7;u)M9;96NwNlhjDTe&Fx~80H&0~1kMwVEHyR zkXDH#labV?LJd;w<_0`;ZdR5918!T8CO4qV^O4V4L8C;2_N49|7XwJmxrAc1Lhquu-0^Sk@~!8F~w% zXU9E9WVVNqrMNNjF>OWx`4n{Kar4fL3>^;SN`j&W2zPYTP#4=K zz+LS_=c|IIAWctjEQ5RAKKg?UvWcrzB8OWCP*NWDFZ0yZFjOy$<(ArnvtWYQusKxB zV`IE>r1}#JUNcSp+O3~zCY^g#mJ4;P39pv~YEJOW{gPw}VQF+JvrknxrESH!*V9r| zJbjm^=7<$4rnmg>^MyD!lZ?g4Yr=`>IH zGL2`z$I=U1kCOE>#xXAT`}ZfijKG{i$+uD@k%~wPIVUgmI0UlKRf1jgT5b(pBV8R( zK^KGpIW>?_ujQWLA99Vebz#6*FtS7P*e7j!Y#zx{*P~6QO1O5rq?8%~A)AuJb|#IA z$W)iea)4W}SkWm!s_adkz#q(Mpf`Db#@8x0LQ?k&=@-ELSxxrPE$Tr=e?K-sJ*c#hCiTO_V&2A{(g)%)gEI ze_JD9oOjzYP>l1SjZd)x#q{63<+Iw&0E*VvFT76@?HCj#Mxe-|s7-s%(fPLz7>wZ$A)N^aSJ89B4bXIdCk4 z$rzhqIJ7ZJSb9d$2j-ohZGg$QcmM1?vcryH0vWh5wxfWOLtR-8Dx~E3M1{n<>l^UW zLol0v+;RQFq?ya(1N!J&k|Y@>S|D25%dF%MaAM?IWz^!^0GB}!J>D%^o~24431N+p z0AAiDyI>oqWTJ2x3m{uShm}>CwxwjpKbiWguSn$Gj37dE&n5b3UBaNRr+8nk}vIT6$8}rADm+uZU*ply+?@lMBuS_zx$;k2? zrQ|gfsiY#(=zLBO6rlZOKsM@7%&cFi%km-JXEZ5hnq1?#-Snsf){tbYQWYCq>QpQF zDUv=ohr9bES-WUJ6y<_2ijiYyIr5u7R(aB7%m>sUT{1fI4nN*(~Cz}Dhw!U{>lF3BzXqZU5qvSd6XQOICoMJ=}e7y zM7mwMg1^zdO3r=X~#F57MvYrutP0j`-)&+r&45duZ%M z)Zib7RsGO*c+eYlcB(wjz`_&D)BrWasKZ(=K9I;HGM7l=qG9I}zfu7#Mv05?G26g- zQb5+53-&0r;0abho;If;gg2QM3p=iDG}B_S-yKP(f!_Wqxd97JZoolRVJK7uKq)od zB*_=nsOntxr>3OJNH?Rbv19=btZ~bck-{*_u<|@|d32hLu+x^bV{}06*_fSEEG6GS zk!UJnEvVgY{mMl_AEXFRIBKV+?fu!QdNClLeP$}o{P>-ZhfD*o&(;ap@wA%>$`(5n zJ3&%6N!F}H9vh&Y#2?6+l+TA#N}cQ3E=oh@BS;lPz8JUENDkU1tM?y*5)|7kGWt1v zF&duZe%cTg=9ikTjK%BUGZ+s4U(A|CuGw)ACCZ3WYmkx;P~?uDQY&BB!P9z`1Z;tV z)Ed~kX}*X}4Tj=0P@2_3U=sc5J%3 zjaIA+l)RZDO{gEI7ta81*DF0l(pL59~ z6Is2w+|e}9XScaWGoib6vYWEC+;$+Rh~wOp6^d?4R(fR!3&oG-9plt13c^;noe|ul zHxbKb%A&9BMK`gV?iY<_z3Z&Z4kes(jAyFMVYlEgBN@^Rd_1L|H(Wug1x*K%KAnR0 z9V9fVSERy10(3t30SW%J%1uFc{f?2?pa;I$ur6vIy6QGj*JMYyC};{}r1;*2svKmR z)GGCRu3p6jFP$`QF37>a2CSWTKBSB;4Lbwe2^ZX9kSx`0!3J5K8~*x<#PZl-={W}B zxtG{uj1`uq&HnK6zPW}aEb4uajU;UbL{`V{R`ybI;PuL*B1WVo%o_Dj*cOi-VVUUK z!duI3EZMs(;om=7vdE?ZIsk-hv3_;r_U^H7u^WF%V?DiG~REUH{S{q!~R>h;0aa}W4}Q^-@J=Xrq<^b$w8U1RT$PoHK{cEW94m+ zjX+U0>d?%9*CR;*>oG*Y4MN8UeWM+YWdf|~1(VcG-;_iyu5^latXKr(8t&z)5r@*C z-AbK4v;LmtJ44V{TJ7J#D^c`FD;8A+6-#qen23ny7KY*$u~odm1$Km3JiO26eCR|B z6+qJpuf%h$6kN361$g$BTmSapJ=0`^-L7P2vLL6t8>RJhvtloEThL8+Lsdi{xUk!n zUZ3Bh&YgXI{-vd*kQ+JWolUn&3c`jtrGj!*wy+G&U1e_ay1dio7ExXp2J*JOHC?bo6L0o>RH_erm6n(+GkIN7a;mL*S*n$^ezmqw>7{)AWK zw~}8Rp2O)~04|vtB2jn#srcV+jBoPPy^W<_*e$Z@T>0R3Mj)(_+Vw z8Z$b-`X~r)p*m%^B)BuubnvhQB8PLG^hpN7_L4p*lc<7FN#_)Fpy?-yC2%I_RTzW9 zl25-n?Yhd8jOSU)VaI*TAPF-j!IVVFH&G;xifDAK0y&r=ak*|6oec%xzn}B?ZYdkz!;UvDW*pYYz%QijqVCWZ^&Ypwjwhwna~pXpUu#k4i62S2=}fOW zroqo<02y!Y*nnod=WTUP_#yv2(><--#y`yLQd3AW>5^7ZmqTkq#~wGUdklNk3(RV_ zEu7n-m2)xtn>$6c8Eod$A-_U`3_5q{4i`<0xQaI{-YY%*x@}8a#?Aw4c#M76hR3YnF}*GPY_h4y zhTX!$jQYoVfJKV<1iyoa!ad+T1m`j7(CZP)EUTZFDehtpxpslFQw?3ecwljhxJj##B~PEo*A_#zVmjTOORZYcAh$<189Vfy3L@Y=vS&{Z^l^Ehjl9fwO& zjlv~ylzbya)=?2C9+*a_1!~TFl*%584*TTUl(x;>lvuP)Hv6y*FE9sMdzKjxv7~x& z2)W9x-p!5?F=~W}yOjJEMQ%_Ln`JBb`l^y{`UChM=6Dl@+xce&xDCcKOjKOfDsMXP z5uEe}`X>Az=5)16WDLv~Ryz05SrCNnSHwB%1?RFKbPb4yWzm%!4K{|RhUbJ#g@reI@hmm>Mh{KFhgZacGiC3VstuN?J7jhWIb_Er z-xrN+%PC6UK#@A++(?okHnB*y+ZS{zs|0aCPFv#FiO3yxsS)oA8FI!uP*xe-m~5yA z+4@=ow5xT}Hv+fAgUDp8RpM=3Ofjw5?~bQ)J#`^ExCFEebS;Cyx@;9P-ohxTmX{=3 zuipOn_YBVO_r^NfiDd&Uh57S5;1^z-z-z$@=F_&!_WTE~0ZjqP|I|R%*fE;2jL@`& zl5eI+JQcBj(P9hHIi8~0}8k`Mj_lik7l=z~9H!?a^uwgYM@H%u9k9BW`qV3H^D z2rFPrSN!_xou&o;cDubcQ?g&8=#pQBLhK}2^^$hcN1d`$WS^*mw}M~nanJh^)XT2+ zuY?-yeLl-d>C2*X(vx4kFKS+x25V@SY}ok$LH4k0ewDuyx@(-10*5LQ9wthvwiO=?UhH>O5KD zIO5PipY=SV>Q|isZqzhi6n|y2JhMjD3+>BuZeRg1R>-2h_Wk{fmKaRVx9k43h#Y3; zO51T#;4>ow^)V$sL6PHBL?@u{FP*XukY=tI?V`(N8v%R8zB~MOGnqgm(4yYLImPXg z4lhpRq%fO(E0mv*0Zy^+1G3q-MO`ZG;O&;@Kyh97oFTGRxY>8RS2vU~K7EH`8|M#<5?r?wBP4o_XF@C(N6lhQGyZH7)UHlYg{h)R^g1z~X73 zYk4D^M)rNp7t(BKSHo7wRD~vsi@F|dq85%O>y1iotN3=Pu1(}~=W|e|T{$O`554+% zOp4oSqA7JN0Sa8v5}qGNy&i_F%|n7Tg4Bg8f-(#7h(@KhFCO-s%(g@Yzj(&mTM^ThaDP(V-Bj| z1G0A~$yW09q}^=X7xvRJ36!w^Wh|a(cK(^O#DJ2M-~85PvWXo^?AV0?QSLG2beWVK zlx23n9xHt3B2a&Di3yJh-x#)+y9@ZC6TOh(qZIx-$~ohcG!u&cLH;92wr)lZ-N-8y zte+7%6$$T&frQ0uPw>{a_y5n0$(Ry5UVO}$5`Cy8NfyhEWp=p5%)`))vKt!TadqjV z^W+}Q8GjF?$&P&=ecwqa?#FBC>dHT$4HI!X8O0l7Z!Kwymu99R6Ir(}qc zH!C$-AT7J>bIaYfNN3#v^3-#+iTQXs=s(>(8*jP~w%dT6nLRJkxa5Y_kOXdIV2J|z zr|(OL$Q^Q1*3IjYmPr%1NK0qy8GibmvYhR)58C*cm1U!v=YQ&K%06d?whe1YP#rXf z(+SdFI))MqpF?~dJ7_S}zQIK^IR6o}1n1b4jA_gl$bRh}^RNvMSxtidx^qz_zjbK z{nllm5p;Bl$c0;+MuTDWp@;tlsR}Pj>ia(5uK{ z-VS9Zo$MdSIj3$>AM&dVO@zW4ECyZ4zwEQiCEj1FL>AVw5W#DK?&np3TS0fA3F=q0 z%G5b=vTbfr9;+5NC?QQSuSk_1(y1r{=HY^{^mzkeU$`T;UaJfnpxC@KwtPjNd4*WK zGiUj)iA>93UXfs~nW~;xCeAAnm@!a8pqgA_Y8g!}01O$y*i0yk@7VmUmZHM^2)4QM z!K$x+=VdR4Nuzt9+bl+&I#*9xm@mA;gGlSyukI0C3C%^?C!~+d4alK4&W8_}(Zgbv zE?FsE8I(+bQv7jP7W4oNdWss9VARtj4%1n0)WTK???#SytnuPH9+D`nj{dr&ll>; zpz?gy&-)CLIzaM`2ukyb^ZLTQRh+3>9=V_k&gw`9Y0V0^bXG+PEazl$IAQtASlrGE zC(}LuW&Q@G!Ou~j`q5#M@yhr))kc0!F(uzafpBC*o_mJyxMu}*Ph3H@%HtOeaN@lB z>Fkg#K3h1sVGxI5)|LGIW>dv=!znMfOf9ISU;=cet z+sLd<4q*m!!~!{%-WnlY^Iw4m%=CWeMhVGdhZ#GL;?)~rrh=01r${Lk@rkliQR@fH zEy%UoMv#%?m}swC69@|IlS1vb0QJOy$puZ-ho1_Ep-enY*e$Ck7=%E1ttwtMVCDI!-+8)UatzW8L(m>o$sG)v$dWbA#90M9Ml0GlzhnZZepmX<^w$j7 zaS*D?N%l)tZ;EVS@B9ZJo7yAu;gVyI* zV8m1M7>cZ?BC=Hb-BIso)M3aOyV-JQM@clADTV`1pkWL)mZhoeGE3<+httt-R=olMxG(3lXCu^gIR3LsH_0(OZZGOEn)qf)4%&;S zsR#_Dea7i`KBNK;IjGQw9Z4F!CMZW}a9ry=>ih?!MNUvcxh4n+E6;F%q*Sb_;^0B5 zVzqm|5P3k7WZRV~;?+pNKBVJ&PgRskE8rPTGZ$FuAjVrWQ`eV|<6yuqpC2XBy5O(H z2p~@%#><9SAyU0xl|?yU2M-5>dMrGci2~(W=va2fQs9{PO2(_y+4ZxLznGF1vavpP zyp=XXXV}7=TU6k_Ib2gNOjUr$2WOp3^SNR`dT!C4(5vKXkc|txOs0_yicB80>Bl>r z0v}UVMmD&xV-NO_k#Q=dC*XK(#x_VqDEec^d6^0ovlg`Mg`so9C0|~K1fESUy^p-8UMYU4#7EztsERScp2BO z*a$V{$r{Crx5svj3Nzc|JLHk5a@J`h`1F|2Ny>zvLmo{C$-q#D2G3^yf}lB>%Ykd2)v? zVos4buOX(eh`L#@?Q@{6R{v8k0DC*#4K`1Ai zbJwLyUPE6MZR7OQ{oyu%(F8Cvj`gt;P{s_iDegU|$=DP-4(yw;DOf@PRR%f+IcQEq z9$w_+t!2)Kehyl#z`Tk+P)-P{DqaOSG8=$l%=iEJa02*v{HRTTrqBMMc`~~6c8m`* zy7ao-F-p#*(K#U+Y+{`n1ei>tC4ew_WJ*4YJ?Abot&cRjVIF6fWXDd3nQF>xRh|#l zooJO&3`ny!Njih;AvTW{O;1DeB@5Th9Cg6kR`T;i5Uoe=_OIaFVjZt70b~hI##u@9 zr)-EA=fksjrXtSIGA!FM8q5eg55Arr(j~jbsiN=5uabh01A)4R_S89*AjgZ6m6)~J z?ly(N6Sjdg8&10M(B^kH&Am;RP0I=Fb{A}>K%mm8nzRF#9G1l<&&7Ri(-K?0=*Hje zm`|2ifHAjE*`fLAA7@#Zeph(mz{{=%(k7_&)YDyb(2)yJ<6AGsohRBM*5!3qzm^Q` zTv$#5)oSQyXq8g|8A}lI%o;HTIAMp7f6d%?^To; zI38zMb@{QH;0Zx!45nE9G8Q+RhtO}9NBv}$AzJ41tIB0$n;jRPK(+jsilbeWJewk! zR0LMR=L>Uzz%4zb58|PlgX%;r;Di=AjXHb?irNCLDoyy&h(n4vn%~Wf<`=%!;q&Ja zhg`bdr^qiW0L;t|@BL2QwsSyVK3I&i_Du8e`1rk!y6&OFSb0Wx*75rMBd{Rl^VcjW@@rAI%93P}f!F7EEvP47h^xt2 zx|?Uy^=UHjVPPa#-)i2f?Ar5IL#V;32ppH}Bt`6E@OB&zIc;QvYA87nOdg~nuFb@V z*||kq7Vnp4%5~8ZEU`yQ5Z#AWf&Cz}dRqVv4hk{U_^%Do=vZ864AoaUwJ5hQ)++Nv zqoKR`mzhxq?B3BJv3vF0Hc(p3_3e`5b8SI)!!+2`0pvST}BhRbO$rv{?z56B%(pX991s(^T=f?BO=l{c$Xe{k`w?6+6_ zYvSx97+cdnuT`FLzbvkF z!g{6H#VO7;2F;fpj@{Nx)l3Ww;}{4yWkoDJry^+_Z-y%S_10N0dx^n-EJ40bRzVF3 z@;F;K>GBGyP;}h$0s!Qk8oQ3WrEC57iQ<_IVY(>VKbmP2)Q~h`^THxkhVb;f4B#1?4B-*cZ#|gYuXoUV#k``G4INZhn?>qD){Nk zUg6fk!>lhx?U6Dr(W{n0^3gWYN$DjfOO?+*JR3qVu3Nua8aOr47o8{PiQutnD`ftJh>a#%?7d8koht1*e%4cNh)zj|AefaK_H+pCc?3m zk|$Cmfr_{)%LJbBEgof}GyFzDf?GP!ewK+2FFFH-8hg17bct1s^owVQ#e4eV!K`=( zD`eR34q;ElEQ$E>_W7>h*Cd|Z_Suf}g+P)r2H@$G9CS8QsEB<)vQ!UkfX&Ky?v>CE z-jEnIsrrIzAtL}C>G53bA*=Uoa&L0iDj&J8b}tCEz5PRAvOxM1^Jh)R#N5lV?cG1S zPe!rBe!)-7D0bBOckl*5APDP@uz*dg+(YMh9r9c4kx7Q3_3$W-3PI^moWHDFx;4De z=@Dm{R{2G6On8mBhSuajL*i=JN5MA&u@ML#ixuwjwWe=tn5qdO<|%dzN9N{z;^!}% zExlq&3uv!PBQZntclKLlo&z$i5~xi=SMs5Fxl*%4B()(*~eDnVO~I$?RH+ z41Cp}JTwIdtdKD+c=Utwrd-%|iwrX)-AN(cpt+(azsIUGg92t6{jn0c{L=)a^W^{HYg3|bGl?}mX-@!)Stca z={q@`y~1X7pR{*D{o(@h*&82!?T?9X;mLDq>v|V#fn)F_#`|GaW5WtKROvtdWb@xw zDLtnBlALnIz>0xl%q#dS_!+KUN@x*8xn(p*)p>lf!k=Mk)F=P&ADsDB1jtu(D-(H995=D;8BcAz!5C-s_EGke!2N zuZHdk&Ivi<|9JD9=kuv}0M@$5z$QtJbalXp10J<-?s9k0hom2e>ggdz9ZExsCl*Ya z2%yHZP0u@L1~bn+?Z0@McaQ;4>yrA5NTD6CRDg&v;gL#84%KkwKq|}c;C&(OgPjio zpoXp!X3UKhVy3TwZbBm>J?n~$6;Nl(PyR-`C^z7t^8lyJQ>z?aJRpk>c@VyOM#X=A z_RYArcTodNV&9DZ*1(cxb(sLD@0Hbaw|kvauK_K(tE4d~jXp5fTFaKz5%Ju4w)`Ty z1?z<3*Z*Z&h0Ugw!;TG-nVRigbfH@*G<>DeQGuYLi=OfnElMVHXc2%+o|%e|H$Tmp z?`<$CgFlq4A(`W`h1zk5t;WdAlu&ZuIVhkakmEiV__uD&UKgI?i!{Vr!uA2PY&Y+7 zUW*#fBI&{7^PRHm-?{<;xuM`kKL3}!H-T#^OVfwF;tk2ckd0t+3n-BYvRJYhTEs@H zs=IpT>t(vSr@LpVnVw$gnYL%Dr)TC)^}n~l1w{o!1sBkO3Pg}ajH0NlrGlce6e^0i z1Q09=3qg^>h3|cnV3A0yTu7XhJ-_~SCArH9<~;X3?|GN!X^lb0Io^9fsFU<9D0k}< zuTy8c_bph#%k{hGZZCr}ZfO+i_1b}zqATm_w9+3K3W4ZKToyh9J1 z!+2R$R0;C7wE?lgJ+}hNo>;LDNK7cQ6F_hn>mZq++F<%`&()+^5~grjfSlM4S)o@s z^o>@qD?l>KY~?41Zc)UOTJUGnVZ(h>n#VgOs8VQDCl=~>d({=tH@H!m6;dItvu}}; zmD!#YqO6Q%&%+Ow{8O_fC9%^6>#UFrm-(hp#_mz%lO1&_1?My?-1uJarhgd{$>XQQ z_Ng9!m*d}aGad_rdJ-)wB)J$5Cr%}wHgjT*Qp{nB9He4fp=zcgG{tiT^n$SWg^i%~ zTM?>@=%$fT2YZv3KlTJl9U+>I<>>8nDp4O8>6-c!YZoQWg4@H_R6Fs;`Gy%7 znkWY9^y{&%U{Hyb1;B#ursdz2SMyQ?Y@yo;VZH>Z@nNd-A&rP0ZI?&o%;n3M_bgn& zs|jq9Yg9X38^!0yIg+i$0n(Hjzk|WmbIVoD^xXg;&_@=hHrf41=yW&Iahu zT$iJ1DdBmAMrdv42({ItuvBWeQ61vX>I0Vo+m%fzoXVu+hL zIzw;TPLJ)JzUNm@R?Z+dGBVm4AFk%~o!@T}}$2sVh~!`Z0NiC2oUB=&jeZ2Z?5 zFSG5p=QN~UF6Q>zJ8dCc^2U-~e_jr5OW3IR)2{&G68#`@)PFa>CHxXeSz0!~C4AIB zWhs6h_0I$iplTqz-Ysoa@1u{ZT0&n2@@<28_*lJqav}E@|70&dYwJY6coh4e zEW03+!gNLEi8y(ysbIlB(~TP2{>~E(&gr*z-gJD|4^n>PW-=YWV=w(DNth_B z?7V+J$u%=4+bIU{vYCoS-ij9Qa&NZRyIHKI3+G)8grFIUtD_KDs$Y&e=jGNH*ob2j zt|vmAV~$zJ@vO9YKbmQRM$f;r|BMMx~=L!3_YGtD@r;M(bu{0@q{uHPw!pmMmVc$nK@o%#=K%hs&Lp7J-c) zdD1;liQNW_)tMp1uF1R}dA=yY4J9`+!e4osnW1Zd32oYknrc6pa3;lw>*a=C{HO0tYCjkRth1>?$UU zDUPmLlBQe>q-<%*D*7o3rF%gwilB&2pq3lOzsebts;Nd*Gq+PYc2|zN5gOG8K{cK2 zlAihm8K2?$3TJ8ay0G9jHRDhb{W1R%aN6y#C7sZq3F$=aV?;l zJoUmp`Cw$9e4QKvYK?R&a0enkIu>2q9Wqv*dX8AKI*z?h{OZD{(y8c_I&pyCicTrE zgc-xFpUC5)b&+Yx1HlJ^Aug{4qR2EQRwJ=nLiJ(y+_Tl4%Ka)`C>CVz^3n^FNl%0> zq8e%?Fu4Wwv|4(le1l^owK?Kt0$p0QzDC2QF z+G#q9ckkto(D%LdVTEwdY5|@-zYOoOzm2g6HXN~8vc4{rcf1>9StIGRJ7_Cak0*q2 zqEo!SFeq{qcGFOAPqSA;?k((u(zPb>C(t++Cu&qc2kHBBQ&?O|^yj|Ub<-={lBy&J2iZzxk^x@O;=|5VKG<0bP}%p9I+FTmIdCZC%YJFPv| zVKUKfdWp|lzLs^@T-+BYj^xqRl^*~3AwVEt`*zECGYT&&_jocNY$DSHh zVFca>m)|03N_>k4%OqX1)06}9bQ!ofgNlrB!`Ts!PF1$LBJ^6&wt4vfN!lE{o0!jy zjp^OwYz`+AGp%yf&q5vDiI!3w-3&Z)oWizU=BZ&=I~By>{UrpSt{Mmlr!A zza_p6b=yhddb-p*eO3WgK;p;r2{?d* zgE2Qbz`yGC+ZI>_HW zCzDQO*2f$OgbMlP%W2>$O>PyyZKSxI)f=DOnJoE4r7PKQR;FN8fyO&}g- zD?%T8;!qW#oo{aBvv^C6I*~aL*b|Y&^hCso4C9(AIPt=F#Ow*kqZmkG@1#QC3>C;bkQ2}{pP_mzZHcM)R{MAC z&HiKl0eV+PGg5X4DvOR#IJ^MtcFS(HpTm>mytLdmqaf1QyeR?6Itr zs%I#7kjuP&agkr9ge~Ty>Cr`=S5+CxA6{marD4u(N0A*NV#81~{y)J(%dia>TsW~~ zY$eFkEZ)lpc8;+2&@6z&Rx0VOXA<37gLTPKW9kW-sO3%cbzs6}v5cld+(9RMe-e0& zRP(ySdU!^kD`DPt9}xELqi^$91)Ni*&l;3t^`_R{o|T2RL&qjB$p%o^cFERVnq`(I z#mcjY>}?wL0DUkZg|MdslR;a6S zwu@rYDYA`<#m#4&s4(h~wA!;%`9LwK{GSj0^gaet5A#aseI5mopgXB<1y&p#od8d3 z{<05#_x|MfmA2t}(%VlPhoAIiobchi`I(lp@4j!zq5Y+5E3(3?-6Y3_!nk@Enu9Rb zGUDE$%nf%qWGyje>zZMCpQArphj7IY&T1{?fL;+tB;U;V?WC9;_;5mXZW4o{QT1f^ zoVe)HU?AxMP7UBQ?NmOZA1Vx8L21f+bRY02ytFmaR^WW@YB(A)#_>P9=M&*=iio_| zFzq5qpFxhA{jT{G17*&(j$Dl=82#xD&D(9c}tDyfYN0jd-XX&%3}*uct3h0#UxT>9Tl79+Cv~#CSt2XkE_ZSqal0gE(gwxEpu%D z8T{fAx!E=G@?Ui=H6h~Lr+#pNoR~qbn*H$SDCP`BPE)b_JT$5#CMz6Ciz*i-xUpx| zN#TiZovIG6`wK7m=#-Fr(+jt_J%Qp8lnCrpK2;obO_p>ks(F_}yJWaDH9%{kmNs0R z=$7YoadD|4Tdh%dfhsnvQ$-OKp^4sCm*h)Iy&Jr>uYU@hS?p*;roUF*N8`8M0@WwH zubS7PXbzmrwA(Ukmg^7a&tE)A&t1W3)||Hpq)%O3Z^^pyimXOfI9e)!crq=#QJDA^ z?Bx@dl}n4D^R~e|!yqIw8H7Fu4GsrvMmL8s?Q!W>*2ZoBrpe0uKJVBaQuoS~B-}PL zmaP;6wK2_9>`KvY_hWOxeBnz9Gz4ga5BoRMwY2Flq;|keB2AfIsORg%bqdJ(r1P@D zW4R~nTv$oE6rD7CT3to|t^x`fG9^zHD+9V_WBe6~HXxIorfia<*1I;i97vxaMV_X_ z+M6_GsW;|}4BGMAWxZj0$q-1xcPaYhdOBHf9@s$Y_&b^Pi{Nh?5LS8>`q~!1wHaER zt>p`DumO92I;AB+CTJb(D_KdjuMD(KnL+Ca#T=qY2~rd$G2OIA9nYZTAxaVH0Ss%x z8x_dle}66%Mzz3<=SfEmU2GtF*6*68JfT8%e-`;TfQuuzMb#LZ3_>+IG;%cP#G@|2 z2J`GPyKceliyRvuVdEufTR*yH+1U7{ip^PRUtB4|i~~I0V0#9Ym;f?V*_C?V=HFSM z6GP?FCW388$08t|lf*ReJ3%C(Ca9b4f?h@~n3pT!Gzb;+(~erbrVI!hCz$JH)_t(x zp7$k;$;h}r8oElhIdP53VKcL#r5Gs8-A%=oElroPtCo$p$3?R`(hk&3XbV_W5Ynhn zONhNjDniSa4hq2*K!sG3XA)B-%VZKnP}e;GK2EVBlg<#?%s8U9I@#>RNP%9Faha+;6q8Ajom4Dx9vfZ3XGqy*h0U{^WY+OZ}>sU29E~|?u#?&REcc_UN-*j6J9oH(=p*__B+T8 zB>(X@$-b6sKCj3w$qF~`5IO0c!@nW@jZp#QKU!nTqps3FKc_Bo*eYXfa&3l?wE-M- zoVRcEhcPD0(G~uBIyvsdmg9<<<*28aI*Od4V&g;~`>$D)zMxj}`8%j?6h9=lV@lsj zd~@Tn5C5X_&5IzLxc{pKL3Jv;Hs2p>Qffh~p?qdN*#V*58=-BXb>uvA3L4Z3f*u7L zj~4SYly~Nz_Ps2trLzRXq#&q@E{Nam%V&E5)lNyfy|ycqy)JV#wuoL4~z3#K+>%ovtfhhux+|UFzK! z+No+F87fF3X@e_5Q~lax4fI{PHn>xX`6b6$|0z$&q<{1m-^gi7oHuA!ytY?56~Qtm zP9a$lEGq_W_j;n`^?**y&rklv^>^Zac>3!8?PHDv@6VtUpMokfY8J z9fWk zA+!Pcu>hbh<3IW;US<$Nix(t%qB<)iIL}62iNtc-5&q+^us8j?W;eNNpskIK6IH&} z1OXnbxx&CO@G&=dUkGdtYzD_jqdM;YsTB6|+TapJs$cEw`q@sB{_a@uO%M!!sTE@ay8M^-?$S|+WtKcNsOKs~qCeSW|OjBvW-rhA{;Kbm0E z`Q2}So3wF*jT2j%&&*)cOEEnZ>7rtf!8C93sPL<$8-+*sm>ue(y8;aPp*-nX!E%fU z8>tVMqc->fw{A6LlCFubfolZb6F5i9x6EvJL*d^lNJH)9_jt97k42>^lX(s59xuo; z;5Q`A>hUt%uTh;3c^r+2st*N45s#ze=ds8Va4l%WBSo-YR>C_UqFo&CJ?etH(QLIb zKd~JDv^Lp@dm=wR0ROxW@W=+0mAt+Fcy!d|i0tHS=;@u6G#~Zao(oOe3vHj{Zj$dT zu(-4^y3?|o1UK#AD^CdTwSg+Ee(m>BUdM6scMj2Zrd#bo8MY{LXzR7|DeycaD7h;Kv zap+-kk|cU=ihM-l)tHYomcQImgB(*PeM{XU9}KG@Pn3O7=KhLoY@D_n_fej>W-6L{ zP8=(;A}Wo2vHJsyBHCoI7sriR;hYO^7@mziej1gLb#ug>EqCjtvCvhc8dR>9Bek!? zjTdpzN2`7sGdAo&dbmG7vusn||lE|c(ofO$Y#rB3Y3UOH);ol8t5NX}2 zd64LQ`@uIa#%z;}9oGx-s$57moRSRk9SVYHkMktZ@a$ekzOjym3A`Q3`)9kJ5EDWU zx@4UvYiE!gvx(S7F{u>U1R|NT-i2|Zoq;KyyMe?~3q+O45a|Fp2yi`hfq8T59dD1# zem4#iV0NFQKX97>r}aZV>Z|LTYeL4*d-648&npwD0W!OBmgXSEfMZgCJ)UXG=EWo8 z4553vUcej0Q+YfI2t-VbW2keh?OD1@0=ZB#=`CCgY3$CnGDnVtns*+c(>h;VA z?MvKBq2N@7W0&xW|GBpPniJpa$fKMriSv4UTh8TbzHYK458wI5Pe{!SU^5*52sTqp z14Yguu_qE-e^Yw3R1E(Lr+g}h5v!QbL}E!c1^ zmfxU=W~)l)J}kq;Aqu%BF~y1tZ)iAbhS&zG2_8xAhd5!76R4aw+`jT?WzK(?K&8x4 zipZx^9#%EYibF9AZqIpCE+ZUCeP2H~b=&$Vt>#K`$;yuF2&qkaTe zeEtS&+{Ro>=9afBli>K$8>AdJ$faD?^!a-WUJcVxcnQE>cqC{pxMftO)K;Dq04{;}&0T9U3lpRIIWKM#5)$2)b;RI4FF9gw!59on_g&qoIjtERx!g1%a) zzCj9um7**OAP2j5K?IOC;7aeNcPnr-73<`YV`o!mF`RYe%4tQp-)G(< zvp(cE%HwA=NXjH1Njj8kX0*l}ncpSDxL_BZOVUL*q>rVyl=pa_DEfW6fWIb-*C?+A zJ(lu0JLWzloxDDKN=Z&0Rug&u7TB1=$Bt{kSKl##=lzT6-y#gmdwivuL#shR*H7< zHoF=}%vPg1X&)?qD`)PebD@SAx21R=cPGuhIGG4LhGU!&W8M5r5Ja9H{n~-o-!P%7 zEcmx;$$oAo$$58~1Q>b7nWW>Gqwb^sX5K%W&86$sRzN8`)^-EDW zV=};+V8$JRwF$n4^P$UOfBB1Vn9y_jXV*Ri@@Y@raoq^YQ+vc4JsGP(3;djJ*Ok@GLP6(%G-+8@>Z_U`U9|fYZ4_yxVo`ft3t^giV4cYGN6zrAkq1EbfXxa&C&rJUnJcWNQAMU$ zks?6v8@VV9-4OY}7N!mE2C5o4%JN)r+Z>)1o~HcF`=nsM^mt%J;Cd8@x&{JK7;w!2 z)t*je8<5?wyXN4km^L$kS`l+@b}pE!U31dhv@*S_d``bl1C%Hl$jPADq8dul`s5#Z z)sS4!m}&%rnLs__KUYio99&=4b7J}IJokrK9C-ZwkFwoN0Q((#=|4#VH-I_uMh?VQ z#;tnWDF!GbHe-UU-xWm)HL7CoG@pY$n<5+Ofv8?@s1nQZ!s?w~s4M1d2Q7Y9E+08T zgUi5EytXCfhrLr#26JAkN~|b@-B1qD4R7C&*3t<<+Q9p;QyYxD%B%>eQt$Jq3&bJ; zjS6+!%cw*1OC&8Z7i09o9Q8+DPZgV@j?BLvl;?Sn*T+i$hLj#@mYZIDikA?b6otFA z8(_3>96TOBV@FV<@H0g*?C)@lZBbKcKp&fAPAIap64pPD4L|AEw>$6N`K8Ig1pVr@ z*U0Tx#=xvLONb0n%pgVjsaRtvYm>Z<*(19rXoDqZC1q;88wN72Qg1xbD6bN*`0=qP z1|C_Il0%t{d`V5>uhv0|1+g_pp}W_|$IIp{~Fv%>WrO?);_ zlrO<0uX=8|Dk&->nmwj2jQHaUINB`krxQ`Pb^Rhekg^mYStfpell&wspqP4uzY#Kv z;4w;e1hO@Ofn#^Z|v|U>I`mVNlxea{?(U^pqk=0yedbxPKD3X%>!?!J(7SL8ic(k+9h` zaPrBWcROAF=Byun)dZ~f?lfnRV@@1?YBfXrS&FHp$Vp7vW~=jfwRDx@s1!&%VApR% zFIF)h@;oIuJciu*=mIKR-QhKM%dv=5egTy>GXo;s z`Cz+_x~EF+LW;LB^dz|QTBu4Wj4-qqXN9x5bP7bv(VI?{-;Ydio4l*#aFpAOYx4`a zp}a}?xAk8$foW-(u$xqKTOOU*$+}?%y(WseK#_VXwotlTl0{!-#$MBuH|N}T2fZox zmPN(@2omHD2vFfB!>=@Ix8&-QENN~;Uc`xj6~OmXw)l|$??2f5)^%yO0>dD!F7&9SFRqtl*a@8aUPTbDI-u zPO**0IZPubzruOl0&Vo4g8sasN;e}LJ@Ef6F+J~8{Po2jZ+LUVo2@b3F)cA2VWVLw zf{PFXYLTPXRZGmdpCKhjgSpl0UAhFZ?TEaXC z;|cp+*-NskOG;c*=M+FS^HT*{j|Os2FzRwTqE&&B>ni$g*w6cAY5Yg-EulXLkZDvq z)IbVAvq1IZivy+c+02=Gx&_*dF8M$^LlrPTr;yx$UZqxjeNj!|smMC`q*0+&f&^W6 z#A#RCu{dc^*|_jpUS=!AzB4U)=^WF3ef4jA5=rJOvtKVa+piZ>OaVpmK$~QymZyu{ z>VA7sqqy8nN9!W{yi1}oRk^O)ftsT^wD+}KX&m&qbc-*0J#}ddX%;?pxhzfz0ybWH zyDTT*u=KD*Pq!(n1y+8bjjw0TT;3zkGl;q2PTO|dKjmK-2l(RIf3^hJ<@;8j~W+;sAUioN~L4e9SXlr>8qC@Pk&qK+@y@RMusTC%h? zCVs{Uf4_ebo$I$|M!IZx7Vr=%w@LJ*CA2T%fZJY=L#iXnLEjqiaq3j@GcxHBekQ%& zf6a^$d7KCnYEU>0{~h7i(yM0N3VW(R-qzJK_Q5A~ol^v>XWWh%e`d8$`H~va5>W%m z``sQ#l&vwwPoK%+qxC`@|M6Fj@GFGcfTE=pG^SQtWAb>~fMiKVbe*D+e?^=w8jc)} z%#7}aamT)6h_piUC5~+zvjsv;>y#UzUeamPrswMZGvo&*6xJ^d{Vr+dhC(Oyn)}U` zD?P$mDSi-GSo;p>MA<{+m&(hMl&@;<6i}wOiOLAzu_ckAmfkvtis-*j8 z0VWCNg;6l6o}@z~BlHxm6xDbPgHtr>(!(p@jd+ZBK+<>Ys@sctydd-CV!V3P1#7;I zIo=!6BqmvcExxyuYel`Vg~x(fgUgM7mw`W@k7?m*P|P%T0_*7VQ2d@RSXwz4!DrUc zK5t%Z4)uR}JI<0YoXghCdBa=QsKl#;%^vsM8{IehH4q&HKJN)2cq41#)qfRAic}w~ zj5-vxbp2QL@ctxeC-|=3?+YgQ#XqxcXD0^|+s2;CRSTH5f1ipvs1tioR@6bU)ZWMz zfW3)&s87Y>L?qaKxg3awS+f-uQ<>t)+*|dFg#WT+=WtruODnV?D3F$|#sp=8Td8-8 z>bM$9kp34Y*;YS!G7~kmd#wA(6E*ML{9%W1DpiI~?6p{_GBlEmU>68Fx=4k90pD%& z>H>lBe<-pxphc>S)C=#1fxZB}#YYd^2uQbpjsx2@dzhd?K)$4B;UjW4tUdG*uLwjv zk)0058S%hzQ1POJM*5%;j}}#n3a`OhH`FOe3WowkEj0S;h1g?;r>hjl{YPDp%M+u} zs1|q(=thpc@zg~J6+b${Vn_f2K{AL>h$=9Y5-Em^zS&8oCLsQYF$O?L;XuYW5C6n`x{p#@g#ef)h z@k@NW!@nf*&nq)k>W!@I&{H#SOX#3-c}3{)zzfn7LA#=$NVwB$Wq{E#bt_U_p@Fc= zP=R68%X#*2J8F8q7$Xxu=#Kh(#>QQlxA}mdX^&kn`lC{k<-`@F6=u8U0~7-km-$p| zH(eo2V)7+Ml4ZRLgT1V=uO^+BAW9MFB$!k>DbT1#UAnM?6syHiTKqC8;aAQ%lmKKlWQq`E{}&uz<#t{|?ImxH&k-ntCMvYJ$w2#T)05J5CIlwPxYI z&nRYyB7=|@P>d*%RvcW1Ou7sB%=+lK=(5Ge%KPj~+X5$+IFf5&DwPt45T@_6p+}sOQrL8xM@SU`IuP%ya^~=r^pQGkkdG7grtMAA{l=0b%4v68R>-T) zM)lFfvzmlR#+?tdiM5~xj}-b?wuF|uRr@~i)Gh{-ACr~pS`v1HWXvFCX6wg(irGg2Id3dJ<9cBwlxkyb z-yr{xFVG;)SjpV_YK3q_RxN9Zsr%-CZ}{597$iZ#!+M}VsiL#VTDOj{G-aV%nMV>l z)jD9VaF`!4wMlWDJOrmnao$+zuRqNGb)X3?!I}TFiR3!*ns(9*#-$Vk?C%Gt*kRJA zJmfnFG5a&`Vmt>$Y!ZJ^>b4K|v0bth0lwQu*N{5JO2N>A))?dYU&T{+VlDNn7I{t3 z=bwI?KqWEA_=9JmB_jNAV3uozhy4IPRYW_=DdL9chwFdYqcj0$+oqr^WUCVc=a3n2 zaw!JV=UG(jK`(?2y2w@ZdeTK7mOS;}H0zKrBnW{72zE-{^aFzOR=NCcHCwDGj2I?Z z^NCTE9J>9Bb3apC1EvC$seSW{@8$%Q={3K-`gAJQT236Du~Mzo1gj)&#f)NXqb{Aw zG6Pv11_`%>>cg6RdT4`eTdmhPPzeBDYsy&M&7sUAytud%+ zTP;A%gwOA=-?C&wH1FA@+wf*805HFolewY(IVLA;o@rfjk*JT5oQV`|oYx^EADXRI zhbablREw$D>rr>+Z=C}QCaOmZZqJ8?OZP)S#f*B#<$_$g3K%Z1rScX(3pl7CFA;fC z-R^TtJqTbyG0}_=J@f^2C{6>P!9M7n&*B*yF}s9MKq5Jbu_KHOlV=yxf+#bG!jwN* z$I_-NfB8R_YHiPRraJLrWJOx;*591}9-IA*Z4^+aD)Z*wcJJ^Sf!fepN~30MfnZSC z3_EeCDZfQ}J@mpVzzcdrJ@RAWCxmgLs~!b{%9#a%X8JDFk)uPsUzD==a7058riz+r z_VY0dNhc`~fFRtU;z5{kB+Oss`e6X5M0r-78(tGsC99$#z=L0bzWIO@Nx<(%wgBg1 z<=3~rvxlnw-o1aqjc9L3MtGy50@aj}I12bK)VKMAO3)kif*hIQGR)OKBsy`Ad|)nn z4j9I9g-uf)4EPcTeZGF?+x3i!VUu(Wnd3IN;Sgf;xez1_Sne zG+u+{D>dLM*!7n+d-A6EGnj6|pM*U3+Y8_$}E z;eg*4D%Mkov6pk31S5tI6n)APMWcK#n4kxW)8quN!sCJB6X7kgNs=6d-%$}-zO;wj zm1~2KL*N?f-CIC+2T14Opq9px4Ww~^1#|pYT?^Va&%jsLCPP*pwock4A0nqB3nQxN zVF)!G2fC+Ck*R83a-5?tbHe4-Za7Q zF@N`aq{4~A%-76Tt_u`XPk|Ic>}EHON4lT!zj?m5-CIDS2W5ItQefD91@E#Lm@@b^ z0P@zRM~Vv=W_2*6|q9D~@K~H5$gsRo`|$P1508 z)75GG&HfuDhbJxLG&zjfV0^|dnH1-xKP(-9BKLY;qF@C2`c4y!D#nS9Lpf@f;+S7%K$oyJW~1NjMFTVv2R?kO zFJM*3AK6vQ`T$3`__fAd05u*^T_7b<>*sd~Pm|2KMG*~>5xB2Tl?7Dt8GgyUy;7|# z&;J^ex%@{FSgx>9rjIuSd-5jQ0&H z@W)nH*XmOI5~uJ_&1)_b+S-o$RbQ zVj{*KZ^NtE&ujY!>(BTEkvr#oY~4p^Zw@dSq}n$(myzrlq{i$?K14CV@>oR0-V|N~ zQTA+bSr3!L{x^lgGW>KA8oM+6;zX5mGUsj*G=en4dVf7ROj2Rrgv{Go-c4aP(23U3 zc)T?R|BE6>$Zy>kbjmPJzw2btVRgRj1skhhFv%0I2yA}&4+(*m6+f>?!?2b5pnd}4 zB5ryiaw#5lZ%|hOhObLA1M11G(5FC8kjJZ1*V7*_ys50Cn>}ktZ&+PaHLp*(DXNwp ziPqUSocKKWCXe^$$Fk><@2&Xo^p|~$wR?_Ep%?av2P6AbPzts+5<4ww>FYeC5-_&# zVT~9HnP5MK#vyGJ6HLlv=6r%-?0$%ojhlA7J2QE@X$4xnHsloa8*m!HiQCF|nJuIn zDJGF3>!{dAybNGR2JHd}iZg98tSYtp`e)0e+JYZj#SA_tqfZF zu{(-NvdJKZ)fItd3RlD>FCzjoZdz=aX_Het?k>l_<@BIWJNfR>1(t1kU#hYkE3JE} zetnQCVCzZq>>~29nAQ00@IDxDPPI~`rE7!m?D9-{1+d`s0SiuwV0pG0D`uMHf4m{h zf~KiH`G^uPTaFtvJ%0cUXs5CqTFTFR*qP&t1B}eJSQC7(E4Hk|=;&L~)R*0qq=oO5 z<`Yz+M4hT5{PkpqI!oHAcpx8)tRgRa<**q8cAeob9?UuhCfbib9P-gxvT8YPC5;uH ztTV2|VYm4ebh#jXZa#kq%7s%Q;#2K=4a$aBL&5miIgM(7;598jZhbeIDqe=yB0*YP z=t;p46gwn?N+}YgK`{j|&fbnG;2TE)HKo8i^BaU$#CI2hl~MF8+^s$d(YRuM6)0UB zM@v}-+E(z5IL6qOJqu3?9(oLc5Qy=bqo!*#sc>(W5Nk8YCa^4%KV)x^-PzUc zKNZi`iPPm)Jljk!%h3CLCy)3%j+Y~H$uVLCnl0E~HY~xeB`4j_> zVm1|vAt@s%mBv5@g8Wdx_@Sj!d8trTgtghm6NYUn_NbYzNK+0f^F>|I(Kto};;5f{ zvcK}PAN;xB$36!*`I62XfBNRK!x8=_u>89CGA0MNk-lef$IJj9KZzG-W?@RgSuJyJbagiOQIQ z9O`5yf)frV^97T<$2txsh^`R}0oJIw{3`GWcE>7Z*izE6Nr($&sS*&VoKMl~?Lb420eovGa7qdu8-Kjh@_u}GFm#a%!c^4Ot`C>+Ie$=JWb*~h^ zH{~@0tzWfgC-?0WC&To)d&T)3SNp`D6Zez9GnpEZXv=rW?Vj_IFN+5>s$t(A zFCc6Xbt!8}0`MamDLK-VNMMI_93Q)1T%15@>3q>@xyFMn1wK3t@JDM@#xEFk**!a3 zvIhE}F8Nf>F%}A>Dbcmis6G{MP=c?F3iOz~>4P1a@uq4XstdG%Q!?t3$_Fib`cE5` zFI4A#SyNSus&(fD^A+nPzr!6U8J`WCInub8e<2e%?6`woKZq zyvnbcyVv80tl)nG!!`gIjL}QMtl5`X2XIAM!=Ee*6u9^|P8{C1Qo1leUxuy@(2VMe z$naYuC=Izkx1QVqUcDAxmC*PZiD&v3W=gWc4@tYBWPg|>F#~iV+*Y)-L)o)1!>?Uj zLO0Vu37|CI?Ff5m8dAFtE>u<%j;}uqBu4^U)S328I=WSH19b#M;*{kZMRP zCrWqo_ZXxQFmZf~jL1((_Ii}Nl`2x_07Hw{-GC0J5FFrJ$~aM{s$5V$8(C-^fr8KH zYzlbzd=Lk`vu<4O%$5?%QWh@8#fg_9D-<$@Dkj-Seumg{gXp-#@TS8f)q|}TeCW;Y z3~?+*%%_(VV$5%K=%<5q->Z?C91-z7@%tp@l?k^LnYk!g6a!@k=~Qe@U}aDNBrA|Y zhRxz-g&PuE7-}i?#!VS}Bsx*B(MO|dcPol8YLMUIUd)ss-c(+ifpH1A(bgS65Zp~9*pIYV3hViYhBNK#dr@g6K_Osb0fT)w!RgzsZ2()jGn}Z$<&81tEX?#ab(j=zE-1l?PBypM+=k@yjKKc9aS!%w$ zBEE< z>B4fi?Tfe0$15FW>tGUXhJ?A1bKJJp1T5?3E?f8Svqh#A=Qp~Uza@!I?2^FBH*P81 zNijRnGy`RPmm#B*GH?AnHcJv0-9u;3Wh)mY3v^=W?#feZ13DQLF&S~ki15oX6qA91 zUCn@%!K_1JB8McWZq0DC2_F8dHtZ)yXOK%~vrtDdrzmJuk4*%{ksi4o()LH^?1G)a zQ7H--b<@V$H#X5NMKDDA>8x;MLO!nUlC{Y;c(6V4JYK$JZFnZVHk{qVbj>dH?hb2` zqueu_AX*{HrPq)@md_q>&xOLK6k^Eit`OzG(R`rt0@Wm%?V?dV5psLgzB__Sc+#wL7@UvSXMq`dGjrx|nJ=)$~XtH~QCD3t~Ob}iY@ZGY^5BsXh%z0>;y{>-J`s@WXP;%w0T00o!ENiw z`hnLzo(j={^F9~E3ef=)b^~<>5BalmXwXo@52%C!A54l|=m8BXu{yX|ap8?}x5I&l z1KZtl!$Hc+P&aHJNK8CMQv%AwL)zu&!8^l$;bJl~1)u%$_ayF>F*6!7Gm}CwNfcTC zx#Hx_2&kqIHj5J%u#fm4)aS_ApKz*dg~Wt6JMiE;Pgg#kYl6hkd-648&qOI==iTL# zDzk|{NHN6}DWGCkh1`1wxhP8%hyB;iIHWk~jZBgWGiqu4bkRU>ngA+sH>KMoYiG20 zZjxx-5@wVr#?B8d+$6cZ2(LI8fFq23b^vf}g~kh9Cpcw)j9B-rg7RM5W4YisZDp;M zC8SVV9(q5rOOY-+BkK36qwfh&$_V;R)%!f^0=4utwNb!kfNp&2h7=RThZGMWHC(&w zhV-)EIkf|TFn$Ur4~p@l*#EtCP)x8zqK^E+Gs}|o-3ncLzigvlwQn6=5jw)J1rEMk z`ZoV=!0nipm>SX;iswe)^FEKE1s7v(2s_=4*QY5DiS@z(y2~I{-y+8|>CyKCHbr$r zmAm0PqlCTjYP-(a%?0kD@#_dcON}WiWljH zccnRW$E=#5d{NRo_EUMD>S+*^-eRRE7Sd+Y2Y7n}lU{>5Ol6K5Wr|tk-?$BQsC}@T z!akASg}u-Y&#nr-BE~Y^+NE|y{#0XmW zGw)Fsluv>l-Zq&gXsaA*|B>Ju2W*plBsi(q7#t^JImzhQgu?MAWs2`@#cqoD(l5U3 zLPu91TvkC3N7hRI8~>BfFX5emEWjgP5y+C+dCmC3-__=c{=$*$djThNL(TrmhOd3u zVPO=y8g&8PItI9N=H{s&Y9_%>k3I*NJWeokw!&h9TOE3c6D+3fee=DhKQEr0cG#F@ zg|nXv3;(Ya=-cM2Ius~Y2LG2gbszq#$uGIHc;h^BX9ihorttF_#SBqokc$1?Q-d(Z z4LjfcHm`J!!6*st~br!M(^kAd!U=Un!I`w9^dWxu9}6icsnv$zb_ zhfK*+g*JFwR0=3|Jfa^*kGlNfsY|ane3P`5&*nhzy#kVq$LRaeh=#R|X){M%&_gKo zMlO^~KHc6)v$VVq`5IN8>ZHWj`^Rb}#r!5YejU`pU1bK9IAGGOCV6{s4(+t^KHMI9 zr`=p1dF^X|Kg)!@wl^;ShNL=iWdrC5kE?9RrkLFn(NMAR-uI*>UKgbIq@Rjw=~M6R zp`ZA7EBZjR<6+dg#bt{R`Pb4p%dShi6{`pyMd5~8`m=XFSV8p#ING~)wEeyMktydB3g!Q|IxhU&;FCx!{2VjN7iP)rI%k_i zl&2Y>IH6+tRqe@dylz6pxA*_-mt?0CqvDtuDhenjk0QBL?0Hp{Y*5h(VKrm1G;!#qGB8PovKZyHke=4k+TbkLw3!aFCX+TI z&#a2sEZgk^C%8;o+x*P~Did5{>VJHK>~>d6}SO_7Z>)XyP@)POWua?@)Cm=+9xp!WAg_+O{;77yxEF3ma-oJoT$1KmK_?l8P! z#2tM#l6;_<u4R(ZKvJ1lO4KEn)$c;1CELDP@FYR{XXs#7WBOM5ntqliL0tN`y zC1mnoHwl_&8%jjSA~i_e(&3dX=mRbL6xTK}n@oBlG^$>~me4j?s^74T)hUK$$OMAx zmo8E-#0$DAX z&{7f~o+`O9zbRt31PGu4x+6Xk{LSfSCuDC763JvU+=;LywyGi9|rZE*T?95cYpxQR*y!yQjkE6y(ZX| zz=taq#u((G^S2l?_`CTbW6aAg)2q<3)w9Jr9v&e zUv!FB7YO-g5&I!uCq^qXNnJ_Z_pH%+rmz{EDy`qzMFaVlg+r^MaKC4a&$GvR1Tfb}m zJRPV-+#-2AcAv*Jo`KZ#_0^ya@G{nzNdd&c;K1QI8IkFUKQ8^i%Vb1aep0%e?3h80 zm_4g`6q8Gly;N*@=!G}H`M0pX;DBmtRN6g_KCn zdGD5#yRDr03CMdi%8f^>=tkiw-foBP9VayMrSrD8LHVWr{i8_>aV2&$(`-@5s&t#7Sec1ebuQU@2U zn~9s`d%T7~|usLAF(FrV$9C)j<9>c;If+64lMw8=3fX* z@WeX{_}k_oacMtPd6jtOi_WXgKtOKnTd>#Tz;`sN^M>nI3h?gM7))NC6BqDs9Ha85yZAaZ_Hp${na+y{`K-J7v47k?C6g@+`OWe9@NB1lO{~+a&ptP8IeNvCYtlbdo^PgI|#tDRHZP z`ru3MfgV6{^vCj@!i$U3Ajnfb+we{3W9@W(1k@xN)y}u@6V_#6zpa+m3+u`G*~_t# z_$q(Ey(VzyTUAjX1MlIkIelVu?~?^p>Kt{B7kuewvXwx3KLSPQ>lfi`VFZi+yk_Su z2V6b+58EsyX1L@BoEUvp1Zmo2?b0Hj587mf^T67K+*U5v(zyNVraw5wA5>~ohg{Q? z*Z^XqU9s&cf6-N3Ez^s}vgf(?9>2cavc~ci!G@KJOe32SZp8L_wDJZc%N8R)$A~-D z2PgAdR1Mz7D_T^2z@2s}j)zRXJw8sR4kGW@ps@0K(P?}+mKTccWH1_Vz87h}rZT4FX6e7Bvf znC0NfpX}5-IAzFfGfw?@ncrWDO`guWjGO;Vl3p24XTO;%y^CVfDY6av80DK_+l8$S z}zO2X_BL2(k`zg1}mF%5kR|*VLx3&7mX1?U8Ip6 zm0drx(-)T^JZO-dn#W?7?1`XsSr?tBV)InxZb$fW(fN`QexDe;(i8qC{1ZL9Xrv%} zgp7KFQa$~VY*1=C-=b<|io{KFESXvL581)-(8kXVi<}WPwl;BpB0O$WDM(YAk@U4CoZzKYXT)DWrdO%_dad1YI<1A?k1ZBZL;`z zJA&9#K?MdiL_>on(!!2a@PcWC3W@FrFnkd6`2C=gP37<97ewm$dtw}wv$7p5R#zGa zKxNvr-?zJTTK3RCpG$CJTd~K?R-{tQCW<6ev3a6aRf!iej@^vd5tydb3qO7HhExj` z*?ljaZNX+fZime93kJ2m5;pnI_giXOT)utk2M5RrZi|Z(+mNeffH_AoK%0A-ip>PF z_KMImubq@`@Ybl}y&-6;rMqTt@+pr1&-Bb|eb9|v_y)V%eNb8IedaZTUN{;JykOMj zmU538uer_7^9S0Ne9@XywPS-Xu(s9juI1^E3X$3otd|sRN#e3;(=|yVI1LJs$x1 z<)<|A>|K;!R-lMkYs?DXMLF;$1Nmix_$0p>lJWgsCl;5wo#eO1AVo~KVzrdeBI#m|T;sG!ARAh!|J<-uerE#J< z#R&hFa<%&%sNC+LGosr=wE?3pgJC1?_~~Uhuahzl_L|ujjAMNrnqWsZ?KeLd_;f1K zIVW~ztVHJ?$Ws<~DzF;-p-;RwmPw2qsrGJ-xy&1&I~DoBF9nGFB`x?SEQ+ zv^s~BIdM(*WwRa48Hxek=MTr|SoV1UQxt3|umA*i6;R!srmPayLV;+$v;tJB)_)zj zDeoG}TZ~C#Am3;fV_8bFWPo1(b&YD1pc6`(Fq2$CcY#gm2pbJENapB;jjnjXQ^iqN zB*X1rxQ*$9LKvWHKwP2nP9~kpF-HIS6x)K1F(Wdm*U!I-(>qSaW%@|kjpBtSEdBJO zMTMlqiLumRhNWtXfd%R~6`RcDN2Sh5jz-XERF}m4iX!oT*(!32m&~h^o)R>Ox4O5+ zJodh)PUr2HB{N+Ty>P4h1@|228K`sZP+ptg8j}vSN~lD0PIX_{AWjkF@G*;l^;7N2 z^@?sq+2S2MZ82Mpdtvs6^X9}4Zzu*vkopwAUlFK z0(a)EuoVJanTAPoAa);&24153!Uh($#tbSNylduOfnLc$1&*wfrYX;=HBeT&DjM%v zH*?S~tJWlT*I(=a*#%Kc&z@WTQxk&XcdeX5Zg4}86K4jVm?7v>ih;g^J5+3o>Pgsx zuuf$%NXu9IKJjjo0eeff8XcKD&tX}osv3%~tAXrwkiUD*M#1@zX8LXbi}h1Yo+*Nc zpb_O$1>33`mZ5Kai41tLJ6(5r0%MRTawLwrV4gBVbudJu!fTN^5$N!W6$b*5R;o?b ztw?cYFD@>M(9?Lws7t;CJ^3ATyVbP@`aHIt?3|nBT1m>6JOZgU>`W-}YIg&5wo>m3 z;eGc}7o<)z*3M|@e%CXuQ05-jcANFZP@C!%J4iGPE4gGlRErLLzDC3(E#%RXp(^L>q~OSaBGbxxJ6 zN)6YDT4V5Y9h9u}d*!M%ka67Qr5BzLIXu6LXw(Vhsf(UoBgm!G!r^*Q%;-{FU1C4` z!Ua1v>^jT=FojRnv7-u1^ZAVlJ6R8FXOO!SVaJIhY>8&w&`&Am5k-cvM<_?V%L`TH zp)mtG4sR((JU}@Q)mHi)bWIIE}tZId)*18I~Svqy(qA29fT)I8Y` z)JNkz+0rhOtxlF`=^P+}?^GR_R|zf31=L}Z#OQ@1boP>JbvKRNytt#iu1*nXAeV(? z^1EQ03zBNk4SjKL4wUjHP@p&c*fU$bQiR2IP4b)mU9+)$30ohJsS6`gJXeI{`^TPz z5xPjM;lt}Rs_w8J-Zi2P{*-*|Rwx~$VVFT>5>u=wjQA4jqin9q8~CaTfbZRD&LGEL znGj8@nOQqaF<|FTB7K#SShq@s38bX(tne%#OGJS@449yH&Zr9(iKCDnQe~b{4**9* zjvA9r*gTJ}5nkUUzXpJe6J^4972=J4Msb}Ml`*@Jt^Q1545lEnb2X$yQ~fk5l=T^P znIN9R&A8av&RLj3Zhy=8tN&%=58ryt1SbD~EB`s!?!?(kSmegBo9&|*aFF&;v01zV zUM78K!O5k`@S4xR2_?Nlu*<^yBib~r`mzig)Ea@w9T<3JDT|-X2igSBEH_AS3#)m# z{KHcFt!U$aUDNxiO#TGJ+3k?^MQ(z^JnH+@+Vhq=!d3)$?}GqHnzD$__1gtmI%Jn? z3I&RblLXVhU5d&%55l0(z~gcto61Y&rG0J0JuVtDf6CJ&7hMkYesr#U zUq!RV6V-FeRUL|SS)Mw5R@ZE#ow*;`rHBhTDacbJ<&oZ_iO;6pX{)MT%y~VQBo4!r7Ca8g(UE8PGKwi;S-O_YihKzHTNEdEr^7 zb)X9V;k^fz3f5erLQaglOtW=x6U8J`WCIn8^v_x0`ED$V1f#)ildV(M%(bhzv~BIN z&8XTomOZZu`bZFF$t`b%lpR$b`(1Ip+U~R0)6fs3r9bq(8f;X398w>Ly^*7oVJ7}J z+xDwITa+dq!~sV+Emh8&uG@U@+!`O#Qq{F2>;}njVpq7#%mnPG7=Uq(aq%gm*Z8q0 zo~0M2`lSclgO9KhAy1jH?F{!bC^gz7Pv+Iq=T&XuLFM}S#w|altdZ*u^=%wwH#oxh z8J?fLaR7vINNzCRyKUa^+a|OW&gI=BRWrzSvq@>B7${mfOU14ZPxi*{Wvm21mK9+B z0%j-pq2V}t!Dl!Bf#QM`^~_);uLvA~y4_zY6|}MKY`2{|oqfHXJDbd%v9opNW;(Z>xr2(kYziu%0c8_Jki`XMb;GJv zpr|NrED^26DkxH1_&-k)S`uEo;ywCFe()EHpvYyaEnc>PC zQ2vj%VQA!(cSaZ}asymvc-%Ah`4m3k1;-iYHgis^C98_N+Goju$3h;LtFED2d8ecCTiqRAC_+x= zHraN%l+N}j<#l?s$Ev>@6-ksgfvak3+PwTBipNtz*kHspd5c%@zyzTB;Y&GW|z zDpo9o&A*7_h0h68T7J1;#+w>R+x;H^73-R>UzraV=F|Jt*>CRQSBTR{wHpkH*Oc%tijjPJ zJn$PEuB@0wYlk>uY2@YXxcXj~^h3+_=sAh5uq>;%Wx|CMl1@`GM+!@73a@rff}mV{ zCjvU+FacHV_DSG15CzSro5jZz+D`%x1f?p|03!KxUU)%x83gmj1Cx%@IC1L%i@@V0(bj1zA#G7NubAIKA64|jw0|1jsyRST1AR9hy+FE@ zH-ZoGW2qV%-yaax3$kVv@c%yi)^T8fxf>52>@@XOhhaA=znNJ+-?A)+i&1f646zjK zZ1nDwKT_m_UD!eLd37>W9X{%v7_wYodgr^DSEkH^B-%0cE`JcShWDA0#Uwit1+24q z9pubQd-wzHkk5uz&1345z^GZkEAzsuiVt1#=EUcuXm<)<+q+(Jem7r0uXX+}R7K3*=>fPRjp6N0a?gDOAj9izHE>&9y zR>T+_;WO?B@C=}-y{EP`P=C;I+IQe3zYZ3pTEfWGu(#maOAuUz|AKTxw=tvMDBuBAcn0 zHra0PV^9%{yCT>RVc`Q1%umDZ&kiqp8@C)fLJaJvBaZ; zu8Zgp-UMB(7S+3U`SHcu-)&f_adTcL@Ilu*2P|3fSm@-iIJ<^! zn%@(!hkqq1lRoLL#p0bp$puma!sBgBmMfI-i1JD$Bx za>9r6+UO16zxZP}lU;dFd*&}>g|)5(=fyVx@zW8ty&EYel_G1gph2g?8iqcsrfa5? zfnK*yxy0j;6y6>p#wDlkX;MCOvvc-B>66D{$MNJp<7B+P}h1use>-P41*lh;8 zt+3`3RZlwj6-$gfC-$SnY|+_5F*=H5P%-;myHqQJRzd-EtmqzTbnT%FqQzvFkDES^MQlQ?2U zUrRBG6j=jYgA%<=AC3h~m_5Z7E?~S>raMHV$o%1rdS~0y9Ce$+4{-Z_I&D2$%%sGx zStkCt7!D`)R#{SlpN3aY@AMk<;!AKzs@-sl4%)p? zgjFfMVNg*k6ghZdvJLRV3z6kd$Kig_+Q0K(H<^#1rL%rOj#%5hII$zzWoGNnP|QaZ zX{2JX3vTI)>q1b@SDR(%P2L|^%HJ{lsqC%@YU1F zkwr2*3A>zSnnR>NK-(R@IdFh{=+RB1@vQKL8a^!NtAx!yXbj-~=!(~703XDLuZ+0;}wKKNL|$;S6KppM;Mrmz3X zJD-kMaF1!T;rv z4AHs}Z6W`khEy^{YULQdMF-ean+=Tsl6c8tQKqB^a?IL6D0u@8B7)2jhISUL=&lMn z1v%(?-eJWKCUvqF>3i$wcd?@-pGHoY62-}vQYViq=gARnpW7arUW@~bVMq7%xSIe} z`IR@C$!cx@a^k&qff;}@C}txCvjEix$fS(qxv(TfYw<;DKqG|X8WrmTy4`T=gL%-J zz+IB35b8Bf9anUWW5PMz8XQ9o2b$0^GxMLxB+rQ*mP2Oqy_;gnDYBD_F^bZE!5EgS z2GTFs-WUS1l@k551A$GTy;x%qsmGHAP|UU?eBr{9DIG$cU)juDdXWE6@B!%s+GV5@ z)2@0U&Zmzb4~!SLL4E6%uv=}@F4(pw&lL(!Fc7x=QX8(~wl4j=y6<|32{Hg}2z%C|$D{WY-{|;!!(U zi(Bui`LPj+b5@A{sg0%1Gv$$8O5ALNUaYiGgi)Oozj!8b&e9k6g(FIbT=qdVCTO0J zx_~|a=em$ycRLkA*%@sHFxmj7Ctfn`jn`jLned}I{`m8Qq9o-J%5oxZ~ZtY6!?nj5k;N2K6!MGeYs--|q5(((E!{ogNeXeV~ zLP>kjN+pU4_t5Vm|D3&7F}q>ri74Q9!B>{v{%=d#UM^_^Coa2XA@t1xb%|6ZE^yV7 z-NF4z5OfS}hY)SLs3m-_Yn?iQXYaXx{ETow#^VRF^VuUIpALGPtjSxaCY&be*3tw{ zJpX&mY(^o)z&1OFirF{6n9dBWqkDj8^*)Wo_JtB8aKW9+hr&9jmc1tJ7Pe|mh>M_8 zNH5$1Yfy?h&#!~Vs}e$*L(W0(zMYm5>vP5d5>`AJ>tovED7!o3vTfG1#BJwFtY@CaU9I-u*~wvMY*zeyN2dI`Q^shnWq@ zq!^HZO`~GAy>eMvJik-<-dzQxg;01tB3oKY8`ZCNyWJzHVJ&kYD}bkpr!-A(rn~3` zPhnXZ8+6$KXTRgzpnGIy#{0iFA>`E^hkeM$-27N4-UqEPqp=*Mn0pktgLL5S%B^Z6 z7ebeEPn1%RN6-6#>1Bp4JCICw#jUX}XOimPR(^A(tKeQ}73CF$YM_ zyj*Bh*c=6!=^Gv>t#w8=Kq{l5cLCqP57fJ=3ThTZU3#h#dl!!>RzgyI$fW>2!vbt! zOjUM=4}XvaxuDOrB%m*{gHGmOQgo>XrSV~x{NuxR19g@*-fu}T@?T}r_aQA2A2#ID z3N>2!f$O5Q#;8!AvMk^P$Or-hCP7v#J@iE6&vaTQ%Xj|N@m0&tqR|cXP7E=Y{vmwy z$eNI&zCIh+HBzC8s6d6S3SG*I8G}%Lrx$iBfL9@QF3#q?fC8wOFdSMgx}mgVsK%O6 z8JpR%+rHV_M=~5^H(NsQRX^}GnW%@%@s&i!%|tozRvV-qN2CynDF!-Uwox%X;=V|o zYQcc4FETA?LABeNu#bWpW|ew<8hp*YQGRt!p<5A!yjR#ZrIVfU+4;iowMl|LPxCt;~*ZJ9GCpW-2@jmOM z8DQ!t20EnoQ87dR_KACk@FZyly5o;$uAX@(;ueU0*6`~ExpXY`ZqBRiZzR2OXZ~*4 zC+m6U#-UY6_miTkdoYQdd|R$i-S39lVklQZda;{E;)pq;LG z%K3T7wp`(rE=1XZcCsfr+b* zK2y>%r||U)UrKuY$Qzg5+$ekKkDa3rrAQc`3sEy{LD#8(ooPGCmU%%*g`UQwH)N%diV@W+kd9G%DC|(! zzo@4ZwugM%vRsCxD#f_tfRa~3<*cC{dK-=H(D)n+qzjvb<-RvwSn;Brr6V5mb>?hA z9y69bPjRv!&by=d{@>2j{cACEbO_6s`}A4TsHme`G|1PI3e;j-WjBH|!cJ>I-A!FC z-ZK44SevX=*{*KYY!|)V{-v^4%YKvlt+%yrFL=A!?d|T@_kCs2_jY`(>^DiTznvNI zud{5f@4vd=(vm!Dg?GsC)FvBrFXDeNCy&2XhPmc_^qt7sIh(>39*fe+P=f6uD4W@T zT3PH5R};rFjNK1&GjSg_+*vT59wq0sk%Xm3Y0LBuA(EWt(Y+A{v6++V2U4Jb#^X7l zBcBgK_g#j)G1f9bY_E2XQ7^z)fTL4A480rbK$LOC)$n%4_|xui`g}mDJ^l}tdmS!b zs}m!IWy{v5?3M2eF!Ik}5!D`L{!E=;OGqxt6l`Vs6wqBAnjp}w3+WK93+M^$19BNt zEDsNLkgc*oID`>iy!L}BNdg1)NS`7j%-&Mh<3o)ZhCZKhd%Vc;{YmaTQ;?2 z(uq?`XU)PUA5hE@3do5d6@Yzag`%tcMIvk@OYzfg;4KoJ60Y??$tk2&D3q)V8H92t zv=kq^pN_^U#?Ho0uO>MPeI`fN(6vEZWrr2b)6#|Iu)mKLEzG4Wqmv^Wp?fWXhq@S` zT?Em3DBuL8&)%qRgG9-xpwE;#^*-PLUqOxC=-3e*W>7J`88cA0tyNRgHpneGfVfy7 zC*C)+P#N9uEEJ*SiS{TIiSPD?Voh9GctRPQvZ4(dLtCsGGvpEJd zJ#kEXT;Lk}-cP5QY>hg8Wi8n}i9oUah|u{CiYcUkH6;eM<4uM2sNkDV*MwXRZ=TjG zPnT(zyTu9OUdr{Z=NY%t@D2CB)^S`k_{&SZj1jwlAfCc0MufRfPEIlY^F#$74xC$ zPDC$#e%=n|fqci5bfHd_=oiPUhVDqC6Wl5xWx3=y?bexP5y3+3J+hLy{k$W!|`fu9=+KsM_tc5C*B<8 zK#XyQmDz$hX4E|KMoRtO%&1|Yhj_}SDDJ_LC;%4fY&2M zp`=-i!Mz8bJHj{nWru11{>a669C!(=doCu!Q}o$y-zTx$43879P(W5U!pTjdn0Sh;qGGTU`i3X2Q#z=h zH2xWj1sFGqY(e(8f5{jIm>Dp)_sMKiPF&u^!ZuYS23jDfIx66$^DscCQ#D67Dv;j^ zT1$W)Gf9BO-_>q+6@AJa`YbP#z5uNb_Y7Bn{e$ASTgpDz6O{xPig2|89SB*SpqIwg z2C0Nn!y1J>;w5vBLfb>X>r&S|*F)a1qPAeq-ln1KrOipLAxHhjOmX^mh7G0 z2W8}V#YJ(4FrPl8sP;G;n5*9G)eYT}$Tw0bDum*hPWqTyr(TE+l&Q)=exd&v&&$$J z0+Xj?(uMpSx|)weRJ)-9aybZ}rpvBKyOit0E{0%*&4EDVJF_)N_4I%@3{u+wxb+*C zE=@7PtL>$;?~s&dto;pi4Mr?jITW*nB04Gtb!Yae8^e3(M#XJGtf(q_w_6=u&8&hF z@YJwGNxg5b*Pho_e(eeb8e28{pq{sfKINS*Jo3Yxj&F<{D;R8>(y?E~ZiSAgHbh!d z!E;FyIPt>ALK2@EHV6(nHhipMuFkn7ZlD{!dIj_kYN!pb^#tft_rtr@92cLlPYmA4 z1vKOPP2BFz{p(%o-fmbPeOMWtT*=3kbc*8IIwd29Y%#ggEEbSnRGSCI!jY&sS0 ztj^Cn%_O~0J-sFT7PO6b2oJ%HS7$W>b^MN~IR8z4#p0vF>gnCW<8&?LojT}rkn60L zq{}{lcof!P!A=Z5+xdCspRL7k)bbOb)|5xqsSkTXb*PsfM!Y`T+#rt;V4ow+jxWs2 zCOg2a*i5SLl<+4{_IsT;P{ZQ)4s%(Rhu8aJZRch$DF4Tj^)ul+;E+y*tpJc~*hlw^ zwcVi!yf#_6VnC+F5HW_7&F9+e0gwO9c%U=>-_AY)J9PHW{>IkJ%AxCV3p^POA}ZhpcE%M6=b3cCHZi81+@^H zpc7_&7}ieK1`N4egIEx@Qg$hOeY;-j5bg?Wo}Bwq)l2Wb3r*7`KI}Zx2X!K;%4@tV zsL?`xt3r|St~$XTSqcB57>d0_t2|eE=BQ&u+5}#Yc#WqOh>izX_OnN$0mcq2D>i9D zgI`DKzYXuj|NVL%so)msabicR*$kS8CclW&NsC^Y zUdrzf_9?3+4PIH`bZwA4fV#f2`S-{c^&vsNTOa?l{D$C&=XyzPz$2G*zgF2Jmz;V1 z%1188RSCS6QPtCpuXZ!t>PIenqdgM`XRw95DQ9 zX;Ag_%m55sj{`(Fp^gKREpTDK<`4huo?t1!K^ z4P2fhf@*=Skz4lX!|{yG)68+)@m|hpk(${2O1it{p5AFe!Lolh5`-AsmQ*DQ-7Xdt zp!g%UcLHk*I3duYf}RfYJ7_TrfldnM4Rop%ZjJ?x%_yG1xxNu>GZ)G&_Z$eTEU=w^*35UkHLOgC_MxWyXq0x6c?#O?A z`#X1EGWoZaGyk@nYwby>p2b~M%x>=uFCGolK62UZovt~~;7{a%z9G== z3&;f#rTetD724ev3HAjT8d}zcAPvPLk>M`a&)y@^srIXo zckx5iQtI`p#a(-oe9&W`dPn%UvBL!qR;&u^q46Sxn@M@=lh;aSo6z%4wWgjFJvY0f zV`f;ZqL@kwgqvcHLh9+L>Z<0pbc^JU?3N+{=&UXaPX(3m6L@DxyAo-mvm{3qx4dov zKVKnqSiYytdu1z~0F`GY{5m>Tv?=iLtN0$Nf;)ueP?uKihI&3lk+Gt3U%l{}G=WzM z6{ULNLwnVv+aGM>g_-ew9e?A`EEPwc)^B0a5UuC6hi<1!p*g8e5ywCilNv*OKtPwO zY*(V@B?R`6aVj-z!;6a{m{Y>f;^CE!Uw+0q>-L#j_E9*Qn2GQH_pb!6noP_C{+72$ zjT0|W7tD;tNs0mT`J+H!7J4A4UA~@wlvhdJe|bH>Vpfah?!Pa2L;tPY^Edsx=6lOu zIj!lXR|2!=BC1na4e16bw0foERlQA+dTHF@C@=t zLy;CxR$=iKblytxWvdnCk=TdZAxsU!Yw&220MBEHCU$P+ShKlf1r+OIHR=RAXdZWl z6QHPDt2_TR9yi#D^A9Zh?R6oCK?(!f8EW`ve9D#8(+VZ%?>rT0Y>Rob5keMsIsR?o zxBmPu*0xZm*zaEi1$4-nlnK6WCcVdf#fuw#K%BIQPGg`$E}+;?iyhp1qGN*#rG36y zEI39>fb1}D&6Jg%yMuKqJUir46@FpwqftxK=j;?_km?G(90 z#Uzmy)q=8s0fY7hWUy-JxG->kuz(ZuPpNMCGJ}Rn5r_=pWsp4&%?iUJP=lj{a2s@A zK@J2a!1`3+zhIT3EU-sj&0k=I6zJl%$PWZP^zNnmXIDmR3q`kFaTp~01S$c^rv%g# zrmf*?FZ=6d2b70>kPR3@MUNDE;XTh)z-iL1Y$xd+DBr%ubNF*syDey!uJ=bDd_jWb zp+7e0SVN;@VP+c`t<`}rZ*G|Kuhs3Khbm!=1-8kSz8Dv5qN}ksWn=&D+IFW+kg#ni zyI$oJ_{Mj>Z?Zhc=ZAcYG&`}Qe9O$1bWqG?id>*#j2twXw1J;sr(3tWP}Cc>AqvRD z9|Y}nZ&fS~-U<2ckKNYK8E`)=Yjn+)cF-W-lkTO`Iv?+><(pEm+UJgVf7y?2Tay+TUUxvkEy-_EL7Q@W-utARl9f({t zN;<_eu!!yUZk$s)?cucY$cNJ~8Z_jBYaX6%lf}7T7e3aZvI#ae8U*f_u_KL>^E{z% zk=toY=h@4o^v8Wz8OOi zmZP0wE>Wb7ia9^;Zd8R7+v`gB1Ei6EPF(s{FWD3bi-xq9J_dE{|1=b@@0n%1@be#n%(56n4zHK6`(7 zuGbntjmP_myg_&D(z_zv$Uhc!e%{h47iMiy7f(6lxz6*RA~WE!bW`B|`H8$lezBs@ za9vHv8t(@|<-pVMNYUtiB)lBXKGEai2L6#&Xne1q^tIsxeyZSWML)BYWqwX7FIYsJ z3q?ShbAh1Lk4~kVTNYgHhQ0`RCg8!eGTKyS@=LfSGqM&m$ss%qYhRN*M_s^cfV!Sl z8spgn3Eqs**+;iAI#mL&(K&$Wz@9LB5_GtnPdPY;wiLpAPJqKAk=Z0~lQr-LWIEMx zvSqrlp1n^w6c`I?%L$S)fJmFHPmwIr)_fy&Zj#{0ci{wV-WDQ9Ha2|W9RsoskH_xC zxlF<6YsU+g(IeXD$WX{nw_4KtIqlw@g}4vRMOk1c-5t!|kZm=eO$|tTOc6h#Hnsht--X{&gX(5`xKmBD&VRCjdr~dAbB-M$XopLj0 zCzoQjQY4d#!44E;U(=~D@|LP>lNl*+iYU|^#50(#(y0zd?pGH@u8dkN!q{99b=BjB zhrQXTN6*ffaDu+d=%MU-gd0v?`o*tL%rxQTz4yDS$hPMuiv?8IBV6dc6a)16WmF6@ z<+rG`#o;Sqfx(6+RK;vj9ac0d4Anfi9n=Ep4P=oX^6PyeW|*qn?71CiCb!d#%nFD~ zH!79{w8=Kj$dqVHrZfS=2_)Rp0@8q%KzmT#1^Lzc^lFc@cFVet8EEW|v0VW7!`Pg! zSkmV^?J5KvPb0ukPBA+vvI7MNA1S(&mC-i@gGz0qYY)AN+Uy0n@O*l)D1U04Tnhv? zzSVBA!5P3=okH5hxiq#h^gu2Gz;aDqz#H`FhQd<=XlnRKY@<`*75LZAHprehunPc8 zgK#o96J6%$Hd-n(k7gxs;^a7sTC<);=LK!SkPC8~Uw~veWFG`q`AY({sEmsd-urZK z)B`D)l-=Gfup|2a-I<^LIp!C?`&rE2e)0Q1{VqmIGcJUpBukdQXxSY8oNR?y+QyN2 zy+#<%oC(p(8pZ2?>#tjpJQpW!w5%~(taefi*qlNtW+#OAia-$l9F&c= zYR-zg6~N>N{C}~c{ZV&jEaG7a=q*_aIZk#-Gkj`1S~VLaw|q{?;Tm8BJSV;fys1kk zZIkYmHq&v+Bfh(QmrZhr2^tS%hMgA6>9)uFx|(w@s7#R29RFDr*}@GnPP{7Bnn9+N zVqn|7or=k%k%lBeGF+vFq>)XV(TM*B+QLV86BV^?&X6uB zA>P0o65tis=Yaa!yGWgIJjy*Io~`9edtktFOnbcN=HYr5Ed0K^2{Dyld83)E=4NG_ zcws6q!&e5yY@}e$V)jnI3q23jZq*?<;Rl#5MJlwzKxP$EMkfg9 zyo+vF@QW<^g(3*ewD54vN#O1Ft`(f}Mh&qcm($U80jHyPD+-{$AuCL~O3}s3rysZ* z{Nxe^p4kctQl{qdaz&EhWB*QhY($m-$Z?G0lql9f?4}y>2Fk#UIw`<+xZXv2o zEy$-Ep&j(FV!YAv_3unq@N~8WWff=|qFf7!oaAgTCGtP`JbR-gXgm zQt?l_4nocNF+~jx=J@osT z5>$q0Q8kNm)q3#6x&f`vV11UAqMQl;h|B#?p$UGkoZWDctg==%=)|l8+sxLzbc)$P zkrXNh$t%(rz`-G4rtS^MkwE!!^|Z}Vc|fPN#p{L=r4^y9&u)?65#MOGms~9n+c@?< z&hA2ezwUuz)-X8Q3)Yar=f+(+Y6hEq6tjl{5u%twVKqKgz{iIB`5VEF!jFA+NmtHA zLZgN`$C;M!Tv4$&iS){mZ?ccSOM-N#z4VG#j)PWs|Loow#`;f$RtULRD)5Iq+W^j%32(;HYknMdTnt>!RS;l zOUGaC4eEoF_*)i>l6|T{3IL523S4QCIwAX2B(1Coso`&)j-7-(^ysZ2cMv|_?2Pg3 zC_J-+a8)TImYV=~?Cbw=iL7;EKN~WjBZ5U+DJGL5o2VFLmlN)}jH-7S5X22U9_f(} zDx1Y=K|?NBN?HQjtQ|11+We`5{HKx48Y3V^O`yFGumi%1RhszL(g`~(iI_%bK%Lkd zW+7+7imnt%Wi)QTcEZNXSiOn=;tGd}tV}u+`#E)L*lt5G2pwhPMOZ(YNw=tmi$d=y zpauh79U}y{i^{y9uN=5nP)HpAj5Zcyo6|8Oirtco?PH9(iycv4u3K^<#AIStruOb6 zJDk|W04d4{6H`qw`zW%9iaE?1iokgE4bRoE^EPVATZ{>I+*_B^<%T5+!$#QqsqOMw9Dq!L4X~KtO4dzl@fy6tE^x6sRscn#CVmVD z*s)^G`g}RtH96RXpKEX4C?f^W&33E73~iMZvzsF2R7@5RJ!#yBFKClpos&s_6A z9JOG@3+Lw{&Z?&sL+e}?FJA)GUJ^W-Pa6ogp|}v(Y@_Z$j|5(Ki^dka0^Sjl#Y@v< z1Sc4Zmoeb8Qoe)VGk&JH&*$FGNc();@xEAoKjNB~2|F#n*u9XXIkD}5z4i#(Q$R5w zi<3>oq$*K88oTQZHCq@&?f?}k?M*obf(>n7$ZS#*2@yLC{ekxj77Aq?`uKV}M~(b~ zy;0gyw^Fxv4gcB2wv3z^d(HmHG41i1lbxCH*7;MfTK12#v}s?J#%n(CT?6L7QBf|w z7Ty9Ilcg^nliB^rKKUuw_UxZLj=ito^z5CtOkw`^&UY!30eR^0@FaQU; zKoTfsHAPlXF?t#WJq%Q{_yVgT{vQCQ*fogS1EpVqO*bhQMhd$biWLB7n$@46}GV8zfWVQj1y&>Ei>E7*f782zl0k? zmPs}(dDbJmMt`z_Hlz|W>1B`+tmkzp4|8*Wo!BE?VwTvxOEJK+_9+!(tj*J@Fo$g@^~H>G$=sfB^otCc z=bNE=85TEOBh(E0i23A3)XckV^{>M*v|U+%eC@Z1_d-KCo6;RgtP(@6iegb3=2(a?d-F z)zh$SBL|Y!#$VhQY5b_hp5HneKekSV*J&}r+ZozGW9>WshE(8gdjK@aBP||h+?g3D zyTADylbmoj(S4U4X>`PBstH`F8Y9(+fzX?AmdO>#Viti>Rll{%wNi6OS^cd zMEP#@AV@y?80@d;RV0CT%c~G7!WxCfYw$g?796Fk0*1$cw!XkdIX3Ru-G8zd$UaAh zonG8bTCaQ7-JMe9G4*Zwvb5BzQ@P3SguF$w?#Gv5s+!3 zn3EJaPQ_Ha?IIPv7{%4%!lzSh^*cox<*PIq5P5;sElbh^l~sMAr}%gHE9FBjSloA& ze_0BpeNFOA`kc50A~Jj61jcB9i6VbyRv0d7d*CF}-r|B+?bfP7P73T7PLM$DSW0va zRJo&K>_>uLugd5I36fS6(|OS5H^^Tq&GgqXBUde(Krt2sj%`7+>EnEUkyq0DotB(P zqZ3X}Tz1SNXnpxVE=vn|ODARV7Hpfi^ZO@$vf;Pc-&;26oh)7#jeM_vJ{;7?f0{bs z90Bn81UdQzy9xUKPui+1q2aXqZx$F}JIequbjwvi|5L08JW)M474(^MWAq^iDi)KE z#rE%4dCZ(~gTZ44vhyi+7+6UZP58S=^ZwhWmFWG%6)%x%lgKi&9n1j5+@(k_6_X%Y zKYhIq&^TvGvcj6=slXW?CvSwR0pM`PK3A+xNmU|Y1NPd3+79UIXmK;#;nBoToxFm| z5|l?3i#tMl=*6T@iKOA$POmj{+e5Lypa3#AmB8StO%UMd88bRO!1CwONsu%KwyQF5 zI}9}_Sm=#yZ?I-`2z#OUq!Nl`bbj$+10)v27GkNpikot*S%BQ~X}4TRLPLg8ov%J6 zc_`3=AhciwFNe;Wz6U;$wx6tbN98eOj)k4p%p}1Iw-r>c9ESr=IBqjMHXA>dySUA; z(}rGuu=)E3EbE22^hh|d6U0(UY%Hb0^dzPx0l^p$y69%0#ldC>EWd#=n>cWDo{BRK zmWMOO?QE^shWpyS=s!;~*(Hts9W7a8t*Fn5y%{L88sP^gQ%nLyR#P!$UKwFs%I>w0qr+4qcLqSXu&$0!DB z84dwg&3s@SmJ~{`iFX@xao32ef|iBV1m;2d8p44g&2$e?KN<@c5$CL2O7$A=Wnqm9Am5VoM`A5ss|H_{1*}v)^v?A=9CU~* z^SCJPg+5Q?cC2J>n^nOC8zXiqRZ_5|H85;J|iKD?VM*ms>AeQx4YGvwqOTD)nbbc0M z*|0V_vJrYLvDdPOUN`xoTcISyFIU|NTAMl*)~Q#j44n?xZ(zuBDYbo(xEdz&d%c!{ zqE&3<#^^W(YDMuQ6^a_g*`P9Ct{AY%7TOb9Pe9%mr)|LC8}1pQj0sxq_xsP24cwsR z#Fgg2*f+ve&Zn3hifkDckGUzwWi3^S066er$S511k4V~Xjf zNEa1z(QTJ67K`Ig;1x(!V*h@5IF{cW4>V7FnF#M|wzc*I2 zlUMJ1(G3Oi4k?BtYTq7oH36jI-|GI4 z#97<4JMroSEX5;WkU}xZ6iJ|BjJ)<33WsmLv9qZmpdlbrg1oU>2QE9-X2|x*w|2MV zjxzvg6~28+ zY?dq(K@JMt6?9QPAcuqSCiF-O-5k7XaDebAhrwn%j^RRX5LW;5l>|#p5vN583rh&L z(&o}#%0+_9e%czKzwZ$0>AO;_`@}ts7G2f-YD|5v6+iG?q&YDSsIK+}z?mF%m(oBW z5s3Io4oR0m2vDa+7dD>B3PTSSt~v_c_eOvvxweVzVkpN0o>6Er9T)>vM%}^`H6@CA8T8%)K}{aNYN}4PeA4l6)O{oK zhdp0iF{uvZnX9Ju(CGms^9Sb^crBl_;;U6(y7tY!`70(t3zzJc>w{NXH96|as4iuh z*LM0*gih7;T4nh5K)Zm)bA&vKiCCyi#m^PJ|u zc`bD@|3{zwYh{{p)%T%$BwM<7Zi?SQ?{+BTEuM0Io_=ByG}AEUJiV+hbkIFpS_;jUogi&#^ky?>+!pNO zmC{fXL1U<`h`%N(i&q5I582W}V9`sK#0Bqy3PVsQbjy}57nNFJof`jK*aaEm{}`i2 zV7HAwzgF{>izyUW^x1FUC$Uak@}M)rM-s)vQ)CqtljvS7&Qas$2KGUqIvLptVogtR zMg076+cf>3U;Xr6uus30DA0tjZ!f9(oe2`xqZ3~w*Pom4{c^J^=FcdmpCWx!jLt93 z+mPPDm~Dz*hp_pzL;tz@_4@g^3oi$<>H)GJG<<-@3X+Yt13ROJTrkmdYf2vEqK`A% z1F>Kji;yu{0~>6^Uc7_fOH%xXT!se+Y9ksm((z#j`5i)I*=Ig|iZ;;kYnKX^0&5>u zK_l(a3O8*sKc5EacFaJbNDL^f!iO~S8zUE6E}hKBDrfB#ugzYk_{H;UAQSpXp)KSm zLkh`o2D@Tll&PTcORza6#c$v6RZd&Rd?#_j%g=hx_s0=>ADt>)c1+ztH_*L;OVYpQ z((8Saq5R>5>X^DKqGZk`DX=`Nmqk1^BW06i8sVutE!)J&GEt|zJ~{GTlVy^4|9%2# zoSu42aRX+nW&AYvkSRW3J9QrDM#L&pU24yzFT2d(4 z?3W$34|+D>dL&Om0(2Z8#cv@70kD22O9K2zP-=spa3Ovw;LfLOrw>9i#^cnfxB_wn zHtq0|9p1U=83Umvh;8x9Sx<^5krQTfQ$sQPDY6&nGT*o?{YR&=+@Q&n9#ADb&8!Vh zmt{*g%0N_|M!hCH7Z;EXC3CoGY1Pz-Pct3D4}~Y)yVcchdHfo2w_2~*C_BxpjH;v` z!gWi6%e+p4KU42r09zNm0>{ke7e(SI_OTWYc=GIzlfxNqrod@aRlCw&{OYsrP#=}* zl?@~hoQ@=-&Ei2n$F#@goUl6aONEgy`kHLYL+1EOqT{w~Id8Cv zRGFE^Vv5;Lk!?V=7uF-b1)T=#-HY7M@_Ol2nzR>E)YxL(B+O#YGifhmF%aC*tQB3K zy*p})q|@txcb#GlgTg^IfxTpl`=_cp#SyYeR;OqRsh(yZ%QtG~ZGzWhmzz;T+U3CH zpa1cjlTCoR_M40UAS<1CQ=V@Ifi#MNbn7}Q2H7!D90lq9v~31{ngj`!KOr3_7HjBK zouNaakMj}i@+`j?gyU>~aifD@aWf|CUj0sn%!Cc`ZSfyS(j*{h8Ci~$MKPNxk`8@k z%9B3$w?%VWgFjQ1EuJ5$20@IbeOmRjECGl!sS+e-e99pz)+oQ_dOP?-p@Y6@=8ur$ zOc=g_<|8=f9jBRaUaS1|syW$~!i8MY*iKyR%OZ2An_Cu~I(yI5RS@shg4O}a6(#x^ zgqO;K3xI?V!-9)NP|{07y)P(kcn-i`X}Kq=>Ov9jpnFw}_ET|xmeV#Zwl!ZlI3OU-s(m`FoU`l@!7ZQS-a&ij8xK!k8a zvG)y%xkeE^6*D+Dg`cX_D>CLTbJs2=8KOJ_vXUVkz}R@hbBDA_t;IUe?SVy*jZ6j2 z?F2z;@KvRu9uLz6Nb^wPy8@*CuwS>))xf$~#P2n*Dq>tOk4EW(A(wXL02%bi0%5sS zrFON)VJPvdo|Z1xZjfv>Sgk~5kFpSk!9qXiF`1b@n1N3NUPBhGS74#g07qUq){1gJ-?g>MH*!2`GvADw5k-Mky z&Hbn;v)chq26IC7!wVapb>V^*5GO!HXcvi+qO+%FPfeZN;&G6VeI|tx<4^4h(yCfP zK5*5k4v>e?7zynW(3A#7BA%%Y-vmTR3A_x)<;2H_75j*xSOXY0tb92xDewi;8s`7| z>IGz@H80qSlSsSGR;q0j1Kl%QshABw*ZP?vmtJbp?b7Dad#7WEVI94UEQeeW)_H1m zeuGLRc*RZ8`$pE_)j`>isMKBv+VCPKfgEOS#tR=u_0+J#=bbmsxY($1(I!Lc=8|a- zf>!$LKwbEVt6qi{eDrE-4Z(IjX4`YN>#wwFF!tyCO!>ec^wxaGvaZ)@drw#jj>})W zEUlB(sq5%QVWEGG_-NE^L5;X568|FgbBpF^)K&M>nx}|f+fX~~XxaogyIjNxwbZK( zIWnbbaoLa@bdIcd;+TGg*+P;>G1(N!qGB+3jC?QHON1Qr8v}aD0%VdJ@VxDb)tG}G zH6ew5m^fTD19Y}@s)5k<(Aw$o9@?G!3%u>3Jcm?W4$mexpbP`X^mBL+2mIs&l!?Og zd$S`6iIb9Ru@XUq#ZWd@#lim9Q<{$c4_WMBZY4kU;fAea1$!ut3nbNeG#Dl?Sr z;L&PNsT<{ZI89SBcMv%7^1`!bp;rQ7+C71IA~yJt0E>H9P!BwFp%n$S0}G}u4cev9 zM`idd2d+fmyvm=s+zs0%&}$jLWaPMTWdSp@x8uYbJFdzPWO-TckGW(3oY==@>8?G` z^eM4H5_mpdMAh1b*i4I&5>VGhsbiG!10II>Fe6PZG6$C^PQC=`g5f)_&vs;k?Km+s zPNcBkE>7$G-HRq`<04V-A-PVxEFCnnEW0SCj3OmaBnR~Z%}^}cKyPu!8Y=uZ4;bJZ z=*z;-W;^jR@cohrZWSV3wo6HMTbojXB&iU8p^I0y**zsOJp)8@03se8(G6G3E;$_C^R!Y znj?5jTi>#Vl?|V1&kxQaq9IsokRd$3N7aR#Mf>h&E1Bz|% z_`;Cb6$6hMEWE7cu4U;87n|Y4UI|MD4#w?wdoT6A1OjhRd5|fo5hLM8c_g|ccvxE_ zNDEso$b=nJJ^-nkMoF-D9gh^Q5HW$}F_UYbQ(t-SFF&+oU~t+H3Ja$Kru~d*nzW!I z!A@kK%LqFOxjD$aL1Fz#PryQlTL+}x)^nZJ+~+?F3F&Ox&Xpq!2_4$o0TYImz&HE2|KDc_HsTMkl!ZL%{yI!4h)Odh&pFy5ha=2fROs^l$eduej-E!v#*DCb;6}x^*hae6Ya??wsEBSP*Su2%1hC=`QZ zd?xRL_I_+S!1n)o-hS7^f)aifd=PY@PWo(tp2mwInAOAq+KJQlFR~k~ooRp6Sk`oN zaekZ_n=BRJZ8B8IHfo$|vG%J*oFpg|4Z6RJZMI*;XP$z)W3nO~t&^kDilc1W_Fue~ zKf`35`rnc-BU`OmCnt7)>&$jXWfW6Fks>N)yQoffS*mRm?&4j78lw+{y;0hg%9YAg zXuh%iSHC5joH%l~ z&kP_%6jMNvJSwIYtUkztNkB4G&>0HvJnBFOvQVU_7or4-@dUKK!(lDRAj^^^d0_{F zviX(Kk6bW2sTX1o3EsWXaFatE9eY57!;IT}DLXW*1nefZFJ1cb6cb$9UOM{@N#W)k zJF)eFDDa2^%^Zr^LJ=K=g+V$r6-YKwJg4x*a^O?AJ{zevVF~i+QJo3Vwy2OR5qTSG z_#hXk?xB%LHV;x$Jr2!N;tY=&R5rt7Y#08m``TU`E4(vlefS-5JcB&teW3SUDB2x; z5^BB+MTDSiA25eqBZ+*SN2}(9swZj{S;KT%dF4te9|7`Ywt)K}g6_*D+&bK!!O zk>04`wIl4BX-rO7p}sV+>9zkcxgo!5_}+YS+ljr_M6;mFqKV8SirlASuve@k`m%I) zaK92nLO{0smg~|AXG~1YJgMg+AA(LD4f~AnubuBxNx65!D41%xLpZ z^n0X80cMUw<>qNg5M9;Ng>DY3^f#X#PM|TaY32hs=p8#~CY|+3vTPCI62*05=&&@8 zWI?B5F;H2d*g7@}8u-QD>s0>ZF6gNPQ+OWG86x*LrWmkk9+>Yysv3j(`Jm?azk__p zdmjl*ko3wOWRAuFA`AD(F(^9PUFj+{tul&#eyN2dIXS_#$=K90I3Tqg_ah~>9j$kYwaA}(`alN+nhUQo=1+#ZaA#?Arlq* zqE!JVgfzUmwvyy<3&}ZgxzqtOJXKIkDFrMwG54VdphLJ4q-Axg)841O%Qc1mxCXUq zIzt=i5w<3B=;5I}zzqwyJyS?3WMAy2FpEP*gV4e|U^L z5&3Gno0qYH%)++7+mQgV$LpOx|7MtFm7UWDVp!_tjzs~VW%wFFIrbsCw`%Sx>iM{O zbkN9D+abKG$mZ{z`$*BEY7DXa#Ij8S*9m7E$KJ;|SrzKtm%=?Q#r3$jC{AouSY-DG zMSU~sWOspI^DJGh9uy&qT%3Hsz1FLiJT>qPPr9uj8ovH%{q(;<|HD#O!fCArizay{ zeL^+pp_e5IviUc_(?roUZ4N=Lt6owC6)nKA4fT_4GHgs6kj2Rj6dWm%0#$6pVM*@{ zoq@zVRs<#7d?R57C}w@(xD2@*3Of{LPg)vV!7zH!uzJL9Sm9(*CjPz4yV6ou!fBzy zq9M^wE=yynPGyIIt+$ijs=g#m5*&jf+zM%3(2%%QljW8cRwvU7FG`Cd7hDNk>Rl3j zN`emWK;$XOjyb0!5IU3K4cPsdFKGZGjxw(W_O~x~!i?RpGCU9`%uvD|i+{s6`L-?r zE{h?2>9mx+1iD*CB#=`nW-Ub$sTkzrM<)HO>4lPb;QEhez-dwE)A=*=T<^Hsf98ju zwDq4h=^Jw5GB}pLp;*ygX{@M6+{L@-rab}#Ex@)BR>RK= zd)ff9x#+-QP+=RmStFWi$(hZ?Wp!duut?SzS?~%)w}emutA}Wcsfha^mKVma&o1k$5NGtFS~s%Dk2Y>Z+O@v^PN5mz;$5*Y8=ns$49I6UXOS99xXKLIcPh1@3}whHe*K;tho+2zuSu z&V-UlJXohlmZXNQ_h_VnW$-x2bYTAa5W#;hitqEZ*LbK~-yv*t?PBip2KhQBV=f4a zjGs5bV%qJ6CVNNJz8?1BQGv--otgISqhuMkFsu_}CdbUNT~9IVD3V0Q<=LXmU0@mN!4h-w}JL>4K|LpaZAQLdkdP*0Qd?((+ z95$QkJrn~RKBZL5sh~k+pRy_fQ!_&@cOrrJ9Vjqv2y#QTonCnIhG1PttSFN%1OjWh~DW6yKCd|{$ztM9@`E~Ee0 z-AiGcWUf$R8^_+qSHJz{#z`i#qS3#jC5xV$^(NEI2qjZY0!3C+F~;2v^fG@yO1--2 ztjUEElnZ-mNcds~ZNtPD*FUXS@&B{(pk=s>i@|W>9WYC9%~&m<-9nOqz-y~tW%Nnc z3}^-g)+Xap^>o2u8!im1%+rE|;rDpHcwv$yYv41r4Q62yZ1w~P!|P^1i|Uj&_84Dv z?+~Uj%L9;h@|M>v#ZkHmG{NDhs*diU6GCcbHv<~z1s$ZzcR{hD+0(&w`W)ae%JXm} zKG?mGe=Psw3`_0n1xH@3u9V%O~bP+AW!fhTGcn?EMc~8+W zaD#y(C*FK)b{JT5M&sM$aq~_7#y1Xse<^zfx0lgFm1YlgCVb%vP zYx|T5yfYzPKt8Ugw@kxw<01%=ip!C?R|>+z^#6DfzZnHM( z`)k7i^VzWV9k*F?+H6@-#kXI#tdVwFOT$tRy)1x_3` zX)rT{l@zm^0x_u=vsd(lJ9eIE=Qv(pSaTPX&%zniF-iB~?I ze;Qi8ZUnb61^g~WrTcpSJF-sD`{m?%v>=6iubtzN) zfaJ~hkqamm&RHSCpY^iS(VgNC$Q^ziUC6JK?ULwJdHf2)Cfi|tjLm3aS#<0I4J(7N zqveNdAH8Y0VdfHXb7CB^Y_GF;UA*oifsL~0n#DNq`QSpyyleJ=}LWKTebsw^1QlG|h$ zaXBrlhcM1Yq8D~6+GLI5vglfAzLI zAqshPt3ksYnb@&P_x(kp&6;8{QaN=g^|SkgJwSn4D8j#nZ=ua!4VLA}5r^^X#_g-| z!m{&^o^{4IL$MOJk2c9`fclJHW+qNV~k}1eXZ`qE8B5HZ5asU2o1H z`?%$qo!G-^HCv63Q4EA}4guqvaI@F`@U3J}+3eX2!rtBT{oxhTZec5Lm%3GR%6q40 zjU-moB5ic92lLe7(ZtWATQzAxjlv4WfV7wH2j6HhG>ODgcNHDtUD5*H`FUyTRj(xS z(t>~lNQLCI`Mgpf07{=y70^T9oqCY2m=T9GwQ{{ME$9g8mah?{1&w3nw*&MZW30xG z!b?X!SUu4Mrnu$7he4XwX;*5p*$f_u6tjjRE2)?^*+tk;YmqvoOL(>2hJe?d)TsFzl$_Q-4bfn>1Cjm*iQ1ovONl2fPHlhIHO1Dhu(OXW8pG3q!>)d*kNG9owGgSCRlo^T$Ugw&Kt4#wYZ+t z&Hiw53izuBoa>r6InG;#UA^aZ#Y6}D zfo$yPG9k@DAbT(=0Lx0 zaL=b#kj-rLZq*d}uI7Pu5)?oq!}KLDe21YaeAnewO83&MNOj0+Qtbw-ym-L9QL*90 zM=l+1ogkG`BQ~7}S8qNU;%uVy;QC!Xmbu~S<$m!X@hoL)QP&JsbLV$GCC zE_E>KEw4u5x`29KiJ+c$bBe3LB(z=GDWdr_`&$ z7fQab$qRT*@|*H>Uaoo#AG1+IE-0$lB+sW0%s3~;-I4KP@Og|%QYa*2NG%jf21sRe z4jnreOZbpiVSAvqP=W_iC$HvVaq%{fLVl_;@0HlONUe=WP%JG~xy60ojIBlZH>!sv3~#RM_;0V-|-Gx!|AJ$tZPE>xH{@{A0xg{y+P2n0=8( ze^c1M9CLHmyk}Wp#Zt3$QCvN*fcKuZ>}T40->sp0LhEH`JeSX@o_2>{6S&H50+pjb zy|p78p0u^w9d?ZUNn6Q)JFgj1%;z8algSLNSeXz_dM1(8W)%4kDQ1u&_o$c*pX%^x zw?TQVXb(RnddG|X;ty2?YG^4)_UKaG2VtHbb{m@E)WuNRVJP zE?uhlBoGPh@mqe<2N{Y%Z!F(EZ$=M&Ct}bY$qvANAm2eXB%JStUV$r4`mBen z01|WHr$G@zKn(~kW^RhDali!j*&N*{%;WEZ{-j%8MNr2NS{e+n7AptmZuU9_)REYLm@+*}kVBsZ-`x1Ff!;s= z4^|d``#h0xO$yA6eLltLQJ#~9?f%8$|9#7}u$`WDegZk?#9asX%nZwQin&UWPNZr_ z7r9Ng?DSxEM=Vb|cluLr^Xk@`_49ka@);-W(YgaB8G|(rd`=gG+DqO-J zASv*-z_W~(M{n~uOX|JPymS=eREs^XOu;LzNQ=M_4FzQh7Ek^Z&f#yo=`}dFjEC2a z17utv!_}&cXL)8{fd3%dI>WL&XEd0cIN8Nggi{{4E~G3ViD*}QoK$ODnXOPLk)z%h z0HycinBOsdrT;(n-UP0x{M;M&h$kdJ4A}_goB;&+qv?%lB(FD8_X6v0r2v6mJy#c$LcN=p0dhvWAquGzpGo z6St^=Vu3GlAEiD=2HX#kK6z_IFW84{-WK{ESs&cv-2+XUeMAfXdm-=dfBJ7tEcK_q zA5&hNahtAI-r*NVv_LK$;eE;QxWU5kytreD@St#SKKx z8(<)yNDXTmnFER76JPS0#Ev{N@d6pxtb=E<*JmZWK;GUS~ z8S_g|BSd~szwsyFGm?|gF|?ikMMP`N2ET;h z(=k~AiyC?5c}ZaN>!ibs`MSqxlLQvl zI9PVEC3s=*od78REeOVDIfF#KouOz2fUHET6?d}lDV)$c;qN6m=gd30ofZ@pTFsXV zNM=~a&3 z;$}1^-wce8e7bn4EY2nsqt)U$v1w>AS!wO3*gA?-Q|jg3Eupt*G!4t&x}-M^5Tq9# zn7L4tt?UXr%lnLf)EhsiL+fL=N7V)L4tGkuLOe_t*?Q%gFlpYoVFmMil+z)wvBv!N} z+V(fj;v*k%fD<-GTiUw*d+U+#6-aqRY!sr+jXBLb`Z^>*rHZbAHIBQ>bO?p}zNobO;EDH2Kqd zreTLck9>vvPUJ)AKe#0?62-bf)shNt*E_2sx&oR9G%cQ;!XA0AE7Ddav6)J&Ld?o;h_-T!hp`_01`gV5+2)MqYfXqR?)s{?3dt1Ndm*kIs7C+J2U8B8eI;ZevWm+H~_``Rd*0X47cS3l!-sBIQ<86 zamJDUjT4s?Sdd`E%rfSc^~Jgvl4*lgb^^qC(twWP~(gB^{ulp5z2#oxctN)lfhPiCvhOq4~j8z_=SsrLh0T`fVS&I<7bUIvphP0Os66{#Q_ zAjS&!JmtI8j_`9W90aZXezN7cW)u&0;*W3P%D;ROWi%w~3KZF7mlGS3Rue;VkYe{!q>fTw5a&iL6fG3BDl`3BLp8YI zy(`asL#xoFdn^)_c<)rK8lYM2+2*}OUM=XJRq3B30I5kj z4|`DCSP10?Lr^ZPLafm;jeDDIU#&v+Pt&Z&6|EKbj#=<>fH{S5ACuFk_9Xm5Q5HBB}Wzq7SL-wnt`-d)9boV ziIf^ybQ`PL=ie(sIR`9&*I*|ED7Zpb1l~g*+c4;U_e5-W)1(67c7h-Qh8QN5Egm{q zJPU2f*iXF6d(flT73UP2Tv5l4)ffi~F5@&-&Kt*?T(L~@Tcb64Yir|F^0^aRqc{^% znn8+%p7}eJx)8cdk>q8O2+_HSO_$wq*T>&lAa4Vt-morime&DDGnGlf1MmWz6Kk1l z*utp)Df+hi0z-@a14WSpHA1nY9p4f#S;%CC?olgr)UV^A!Ala%Z;|8jxkfxcIWmi0 z40-`WF33HvUxgh}7|4<`ReSw$=b5P$42tQ>3h3>xK~CsI;HkhfvV0nEct9G@_Rx41 zmUFF;pCFn6*(RSnklMw+F;jVxe~16&z47JL|525)mxGaJ;y05#>%ZzX-5qY*-I|3p z;=O@od=Ry6_E`Gb68|sWs1}S%x0s3jves0ZTr>_FoJ{4!jNi5W(VPv4OZBev+FG*9 z#G87?9b{n8Xz&=2p^3tLH2$|JKwZM9jhh%T8{W+k8`lfO6y8re&ACLxakzgl=%4HL%oTAv16gdVO zv}9|I-p#H_0-SV^LM;_$t4t;OYA0O#Xe5jVC+sQhXt#Vn9}nY9ctmv; zGEspop_^fsrcavUw;8Gv9=TV_GzWbX1R!KPYdfHDr9#WYIc&(KShR`lh9ZU@x+BzZ zY&-t~?;a2Qc%+P3FS`-eMWckJj$7i>UZ8pr=|oq=Ba9E z)I@|eL@HA?=Y~K#E5#2KzuYwWOs$tD8Tj>@=7NNZa3uiK&e03|VTKDLUVLtB;~$#M zl`bquMBsa9DAfmQ_Z;#m3Rd9)%<%G+PH7Pu7g?yBhQx?=-ukm5cSh>X_TrYv*?wocHSL`qLP8;R#(j2p(EcPINE`y}_HEenj`sPH|ZAYI(R z;lzADOHCH-_bK)cMQ&5-RUVBhyOM;^-SLH2CcHd zK}RnS2X~`QiOkR7d9=t=_|2}Fuz54OL#j!e*3KTG5Bg?FZbrkrlq9$fsjfl(qptx) z^iiK85wuB_hdhej?q;OT!=-#3y^I|7(aDa!SrSz3c8@+Dayv+ae0RgIe|C9~7@6$O zM1X!z;2EOH4oCO9+HD{*(Oav?R2I%$<$cuWcF=LQ2@lZ0b=M?1F9a#7Ku8AOXPW~hqPvgD-et8XKySHZkL@UwLzN0q%o)= zuti?zU*dh(-%vM<7h_+q=8zKcog_f#C^i~G1!qEwBzpqTi^8&Z{EG~KO{YF2q-Dx@ z2VM1|h8Wd)p^fBb*{pGztvlyMcF=42)gxX;e6@bCYXM1f;s9f%31SK<79`ghnVO_yyRSnX5un zZK7T{h%b7W7LVIF`WEkN>j6%_+l1vcAHFuMpRX zJv4HV;#1${ogaQm{4xFyIescgE3aMFI}0dGw}Hh@Wzf0C7%OCz9@4<1TB(WV$BT_aYu#JS@hpS|r=fLT|71~W^Y3k0?mO?l3@QhQyXDW8 z``C-prNZb?dcZaQMz2nxL8qfjFM^7Az3jEo3z4etV?n`&qSxp{?^S-(6`lyjkrGVryUb);rexA%)o>;%VpA#dDKe)Sg;}tc z42@{BEQe$&v0%Af);JZzMQ5hk9{?ks{IeaLpz-V~$A7pr>Cy*s;}mw}y>DM6pH3p% zO`ho{ip`?P21?y9y+hm;Q9kGN97r?jv7;@6EJsS5t{a#HNXKJj4_^zNlt z;4|2X7NH*~oiUQRNK}X%^2LF=h;FY$L62O&Xn|6~fcV#-ln^si$Sz+KS}28u{?oap z2PBn|ARdwsSc8ZL<_gml)}q&c0Z(lYgolznYW$w2!0dr`+h9Ec=Y}12KszT`Px!Sw zbMU7|^!#-1hPTQ2muAI2V1l*JDYlCuTEtq5C%XF@%sc1OG-gF zY-LOXv_>orSroh-jBqb0^e>`opm{ifx7lNZ@>363J5>9e6g~hCzZGnBrw4$vS+Oew zer@qM1-?G65iW(Eo~~s|_}%h7pzwm1W-9j!&a?O%kba5T8wfE4)krWq1ez8Q`<&e) z!2D7CkLiCb)m_Y}>Idsg1R@I-V|Gttjd;qV?XU_Fqh09WeecZ}y7P~fbgzg6MKE4G zi=n+{F=Relch8T|4;3=O3R{B0q|n+>b1Tva_Rm+E;>FYTev0teG4Aj3Mh zn`pL1Bni?3K3OQz`*W*6w#=g@v?8Jlrp6jNpMM=Z>^A;V?{p#7grJ^EtH%S`(&*Tb zY}f5j4qFIn3D@{F^lEmywXAMdf$BN-f%_rr(75D}$EUa$=d_Xszy2$UE%y6(w)&4>Dnh_E%5>+M1dubp6MfAfK z**ht=oFb*b2u819G9V?C5CLqp`;k4$W4h)|0nPxObOFo*qnq)cg zioM#zyp&RG2?YumY9vR>RHnxqWi`lHgzG3sP55b{{%DuCzEKprLK-9Mrzg7CLs+j! zblXjXTqjvTA6dySp!@l_D#9}{u>rhC5Go$IlHoiO<`>upG3Ix*eVY?vCRR6ezGp6m zz@?1YiOYQ~XgYKX?}2aom#d6UkwsTa9t&=S)$>k@%6T`!nm}_C%~T#;AC=*kGwlhK zOtgO^fp>+!IAqIP3D$bAV;{q@VruN)Xx)_;Ppq${jCSR*$Kw;^iHq|VumN4kFt;Uv zVpmb%c&N*~I@~}x$Dp5rL7DMpEjK)BtA^ORQL*b;+|0)BeZr2+G6Le~pUvJ%%3d0; z7b@6?!Ka>Lp-X))rG64t?`8-JLtp4dz)~{7)$)QK`5=?xhfRoyOb-Jq>ITVKsg~)A zh!tfht9>Apd_;9K429`qqpRj%)6_${Lfjz9U~aq;_nM}Ze@u)AGMK*URUjv`%ljzF znWmrSHUm#a0IWPP}A>{Maz0 zBvWhxMOGnEQVt(An$qPO-47tZf~^M_Eg5%#K&z>dY#PQfTx{9ayz|#-e=ye|b=p-? zo5}beq1Yyh96*irjnkIMq4ZO46>8}I>H9pl2lqhb=m3FMPtQ9($DnDlOL5ZshIbG2 zM?o)(_bG7^F!vRS?(xu)RQsSB@qK97JmZO^TPvq0N9~ef5ks|mMP#*m0iCa^C3{rZ z?RM2L1_c@y+pALD3S`*$i~BPeoe%a*VdYkS zI07kCnd5bXv28lBFkbe6jD=gwO9rjKQxf~{CUJ$fHk@nS| zCxZr=3aBqV8T5eZkF1)5V(wV7RYZEk17TO73^jp=nRE}l-i}LfdmwGNTVoTvhFx4U zOLu;X5j*X#pZ$oWIC1a=Lb=1xkV~<^_>)PgE29S)EcnqEdwN`kuC7er#rSv;aJBW& zSR)LLbzn#Ig{r|7vX4K$!eDz0#`3D64cH-;<=G>)Vc?(L!x102nV4Un6^P6gO1VT_ zoY*0@pl6DjEgQYM=m7!DOddPH`GFdPR`69c$dO~1;fh7uBAXN|Aa77Cx*ol9&i+>& zizPh{#s(-yK*;o-hr5TSxf(4=)7R=gAxpT;LQY(`xyi)tq)==!MG`1Awkf6f#j`p5 z6;VSjrvHrt5KJGxwXiU~%JEx2i+t2(uG-_Yiz5qiLg=evV?I_V>F7#GDeK#ouq+AN z6UqaUz)4LaOTy!V76~*7lB8hV&@x(VBiDxS49rxn=T!ku?-50-v_)6}#9QfXi*N~4 z{bnl1;k6vwNI2x-o7|M!NH}eze2?trzJDw=7VIX$evnKREKY!J z1k`W`O+{oMES7W;8-3uOdjfDFv;b+quu)~dcaa3$>^zkrS!rMmTpE;2aL;~ktkXLblUtZ+bi+_sn`!TxtHNtX-yw&Aw(*!{GkT-Cc%^yd?j>0xS>Qaz zG*TYD)onL)eM8VB7W)2?jucCK`=mJ@;QT;Ac650Nmi-QT)bOFFjqZa!@FADIs;v?F zcFZ+QA(Kxhk#)QWioGh(YR1sA-Dhc=z%T$|k-Iuz!Zaw)Db8fGbmf3!D> zp3@%g_ML(MYwoYSBy*4jPe%KCoM`3D3ZPHRllDQ0&rU_AY|XUKUw`7#=vyf(aL=FK zDa-<2bjYP9utEZaVEjXhYPYiyD`&TZui8gWNNcCrG0s2ASlfi6QSN82U7oj2tIU-K zxnRYK4UYv~LWAKkR$s=7dc+la$%Zay1a5_;R6ZRSG;UKAjv^Oi*lX0SxzBeV@B7d^ zlB_fej(^#eW1#SYxX%^#W7_{ZOy zt1FE3N1Qm-V?l8tPh|iLF!38(s4+8)EI66UE6_Q;#QnO^P@I5n23AI)?I_SUfGlWJ zc8EK~7>juvR>srUbYsAygr|iHy9G$_iV+b^J{Qmpfr$)cq8A6B6XO~wk->@?Y+%b& z+7K4D9$dqn601Lt8Ge?*^}(rAsUI4bJ%3G%d!2MU@v>)`3CG?5#onbzpJCa9BxUWg zrLO@i9ZGR@>gy!4Bzq!&A>TkKv0+O5jP0WBq7oh!(+#Qxc<}W|}juCpl@A~~U za*SJ+(uu9x=O*j(3lw{fB4;QywzUm5=hCsmsOk`mmYqjSg zuxAOpdf)wkv3@>KqdXB^flMpf;2&wqG}H22Rg)ipjR$C zdod#jSzKjlOvZ{5OZT(a{{8>;o30?E5i0K~TSN-D z86hV|Nu!Aos-)N*6e*+Bso~gki#@iu`pQ(U<}VFMbbCBw=xZ(0>J(exsHR2Othg-g z4KY%NJepUhSQ^|Kx;DBQ=C~qJ8pye54hS~+RruBeA<5}E+amBrjghD4w969x4gh21 z3v33L(HkprMmHQ>zWMp*DmwpiUSdQ}j&SW2lHkOi6SPJS3m9ioYz9Tv8#rPsXXs>) zXPhNF6ax^*yGNGD+hURh=fpjY1q@|KIFqLcHl=&{tdPL zS8IN39y5MPf;twW$;cLq4g?Z`YNro+YA{D`;COrF39byIK`QL7h>d|jhOMJ97XkIa zPkZW+lo|^C0->|6kT*cOz|*dXjA!v4HY~|Tg#&khjc!eB=|vdr59=(-tNfnPresrt z?~xOe$W4=1_aeo%Q>2Yj<4PEpvpdNS-&&u6$a~Dbpk?7`)X~X9kFrpO$1;_*(?H`o za(j3I$P8+E4MAOQ$he?oYJ>Olk%ggx03QRB%JlPh_}-bG&-Tb4doT6g9+(HdClW=Y z69~+|LL`Po^+FtBWS&B@DuaApr?N;g=yBX9ehRuEgC2t*wNev`5*aIJA5*4J*#J(d zl{13JneQKM97gWr^y)kB%Eadt=@0+00?kkLf~wutgf>bs$^(KVM}c^xlg8*H zZdwV=k)RmM9xebM%l*fC&723Xo&U$0t#G|vPY;OobbA=5(MOLk-dGiN7D{U#%|ph# zu84!Kkg%c?cu#{w8m!CNC%gcKrmG;h*a<}z=ky^+AXHP`cC$a4xIo6-x((YND>5uz zcfKS2((j%Ia9VyIOl~@{2fWfGcKC>5A5i2Tr53Xx!s5s@5tGYf@vp6JM za(H&Nvf8ajepe2YX(XOw(1S(ZT75vKUs@7b&pS$*eTG~x3%({4NoZDnVd@#pD5ED2ZT>s%Mm}E@Im{xPutRK=ftIg7JBXedX+h+XG&@&s|ADLl3kdA z0@R&kqi2_@S-*liz%NzYao;LUg@%Yv!9rEJ7w{H~T4%S#tZ|JC(kd`2bcyK{!mUc- zR-NLV{)P*SrFfy?bo-ewJDIMhR?MO);(8$e!Vf;Ng%00|v15S_f97#quXK(pC1y#HPFH5d(kU9{94TXTj2QP zZQ5dDLn`>!W}mXwt>-d6_Q4F7Z?x{*>TN5&|Bpq7pZR(XG-_l?KdY`T1!g~ts1Vb`q@dL0<-K+tzVC@UgU8QhKT52hCM zX-nX*@Itl6MIsc7S*1#VDnX!R#2^Efr5fa<9uggib3Ngz$(5$4azTqc3;10x`e!KN z%mwJ(R0h@nJx-?(!#`LplBv9@*y|7d56JZh1&4C2auBF0GL=U~`yr>>;@TsRLsj4f z3$BoY;8X_tERanWGiuq;d>ku^BQ^B0m?zVq^@t;mo=!gw|IIDR@aWWGp?QneSE>la zLO)p+{g97pD-Av<71ZFPqcH^4;DZac1|OvHOP3ubEBssJ6_Kkvj!)Ljv^u<1%+V@FBb=DP;P9I2sa=pSs_l28j83+6My!K zb+Ve!U2ghq@Gd_MyG)82<8O02ZJEspnGoc{eVs(?#P3nRw-z+ zr7}-DPugXLphl?k86Y^@U-ZA_t-*+Frm}R(CO>32!?u(PF$#@fYdK`QB%1xNZt}Yo zu*}`S1%l}*eSGc!yz|MCh5p#9V#|1O9|F_Y`UP9heaz20$nAA@+Ai|nu)paO8VC^m^A@szrR-v)YNC&74q{AI4l|7PM2@cf_q3CM!{4&8_Tu&V0=;qJ>6O z%saboKUv8Q6;3?gY%xJaI>oM|ND8HH|CidgAgB0t*4y{U6OgopL?Om+H?db0hVY$c zkqD!6*^?o9HaF{mykQ$doj}uQe7J_Yc!csm~j`h%3=-I5O z5%-9fvUzke(*l)VHM~K0Ew96^F=$=XeS%eBv7)_E?Nx!DgVo>i8cWCk8k>^+L~L&Z2#Cv4I;`eTn&awHbRXgX=!Stmn{3k?6a z%XX?BhGdcnnrfw`0$KydfsXv6UYNql&<8-6K%YE^2LhJO!shv`?2?NABnwE2MxN0+yn zry!j+Zf8Mb8;uBv>WHsFolA{4(GNJTpg*`sR3yUgBkW7ijjy-D#n6m22aemrZGE3w z@ky91m7PxP)>xnrXqJPzg>sW$Y4jqIPFm*=@!u}!>&=2T%3Ku^g4z=F8N;wTD~{4m zzdhev+tz7~#zqqtHj!diQ)C6Du4hvDKs}>_j)ByWYm5fX^*FhzCP=Vuj7=^Xmkmy~ zV&WSSOaEX?YorrclnT_z3V9c!_IW1r8c4kWv~@trYaw-&DTi9P1W85I zLw=5=8VW+Qz3zBppGp$zcwg5#ObuTOjZgUP=DX!T4>LNtIexio$u=i0YB^?NH)<$$ zFGY4!>NU(lQ6UiH?QjL^S=mx=4Qdy63hO2>_0Cj2QS5dt2IunxzXO~3+hrYm&3U$; zK`E6DuAfgeiW{dgN$fT;rhkw~U2wNM^gOkk{cQ$ zBt~Gwe21;Jh?AL_u)97n&Rp8of+j2)c76GM0dT9LM-KkJYlZIxf)URW*;7LArycXH zkbuK84XL4`*S1cag&Nf4?&a`yJ6d6`uRE2 z41}(?0$QNKBtEJQ_O|j0Xyo%uWtO|ev#}w$5o`H7W^~2e`BodaHh;egT_H;+-I`nf zZrnSyZzaxK_Ez(I%O*W$7I_$EuWK>OCmnVz7G=>JUTI;Gp07rqm`s2#xG2GSGau@~ zlOutZ8osC=LM`RIi@eX>PtPgh+UR5lteH<|7Ph$|&@$F;q`iO9Twst(FxZKg)LABo zdnP>yu^9;GX%0yn1qDzst1q1@pifM7jP-fmI9Z97=iR~~#}a3>|Ixf!!D$!M7U~aD z`H8RW4$gNr)7^wM%T(DAlyVxMr)E9Nhf6?rI!5QJW1yQUV@nP@0Lk4)nO)cAp zQL5%9Hp#zO%s0+OE&(o!$kItV0Za1Es{_rj%)fshe;n~@PQHu{u% zyKKc&4W{quF+(?-}uArWFNPB1Sg(Z&YKvRV-$OsA_pn;WB-%h zhoY`91Ef>fH1CqM6lR(us!Y|=N%{O?T2Mgvm zHwNVMlRQs)$4weD46TKe$@Ifv!yG2oZh!mYA19kTup^raoi})B$=faqxUW~Oz>2pz z7+ifvbDpY{pUp%4t}&0Jwa?Y`JnKi_^g2gxS)P$rOk_Gt7l(v&+QH zY@t{vIM_s~YlACdnk0swtuZy?`{Y{8hr8d(|KI25et0@2mD%)4Ma&_H-e%EFlD3#c zza8Q}I+y;8WCa#7J>c~83DSb@csmeS)`Ms?yWuz}^}02?nVS%&jgeSboOeTMNggme z)$>;KizFEQ9P~i4JID!kgx-#H;HZu^F4m9qXjeOKpJj6~N^E4plB|Y@0;35DPW#>= zvTU?Cy%R@Gb4?7zT8dpmktF1wzfY>&uwM=PG1?4Ymka7+>Sp&%kUV*mQ38a^uxcH+Y&DEG$(doEm)Zpzw!{Qa4^P)HJkiSdLx4_ z>Q2;%uZ{JF81 zm_iDITjZFK$pFs-Yhdhe4PG=HRu1!vZe#c3oUFx!`eIS8d4+(5q7EGXTi=tq_J|W$tyN`$YpUz9}rs`Wb0FzDvx$n3n|yVfo1&bGqk+e5S>o} z{Xmu-KIC$klzD+eLE^(PGq){bqgNSkB}ZSv%Hexz)NT3AX^6QEpVMY_s{ZQ-7v47- zo45A&|Ck)%wkURDU$@)DG@Yl|vlMBi)JXBJqn{{{6{8xY`ksdRJ3U(XdMTQxx->`y zre!v%%IC&LLoEp~d+DnQ>gcNqSVc7kmFVkC3{1N%(<^*6NuCSC2S_`EO!3$S0wm2j z(=hJ4g+_f+jQnPU;!RB`q)72{EpstSN1u(@s{)Nuyg>n-HoZ&LM59W2JrCW3 zLUkP9kuTpb(?4OO1a z8DIvxVLPJC?(pG{L?w*T!mLXUI!D$zv4sKkqG5%$`4kJ0yKHD~hmskf3Jk0klu-EJ zqRBZ>xmEzKnW1Q0N9Vzqx6Ag>9iXn3slw{pr{(11;8Kr$=*`B;&YPTU%f#>haMv`c z5h#kkf1{Noa`SVXc(uLN1XfuTyMZEUlo~~8R(a^90Xk_kA7EPg6@#9=LJj7jaS>2E zdqEamEV`pO1Kd6Ox~-&F9Nv>QW<1#$Kjydl@$1JvksA?mpu}QA&Vkpu-286eUB7IHlECUN}rx1^0Eu}#npb8?u8pI2*%$h$V zJ8wHH6fB#TouB9YB-99!6{&sO$yRP=#EDB4j+(%ynqt8ru7u?hq?h*!PXUpW7B~=) zj;SqXg|LoZJz1-~EG&ec#8c1HEM^lz*Urj>rR*iYN)^-um&kC6N!4pK>J?kd6!|JgTa+({i6a8O zywJg)EZdfz+LWbV^={Q2Ql!MZ#Ow{sq8~}MOxh36fBgi#ION1nTECG-4oD3pO~=F~ z(4BK(`T;2_676}tgf0rMkU^J@WJyR5Bm@eLQMc@aBFkVIC8IP<^UksFvMUt(r zWkp_*9~5a7X9Z7OszX;qZ4*Cn>6P7>^~5EO9Q19KofTw3S=sP=*ztY0?-i;{!GH1pV$JJD@Kgo=c{wQ_&HZ)anF%T@hpn#nQtWPu?4;CZ$Ux)`K@!pWHoF?A zdGvL2jgr3UDSptG=nhG$g|3)>1o9+dmqO3sR0Kki_edUyKwKcH`grb6spdFAb*%d! zMAafk)~GyBjMJ7zqw8!48BF z6blS3+aSdau1nG@Px-zUBuHB1iEc+py9~NcfxIm!pMTkFp=gPRP8Ju9!nLhRJ*Rgz zuO@W0cZ>X@e^Qizz~fd}Ltr*6^Tly1R?%sSBqgCl&pldbvLu)S)YAxZnAkW_a*U1EQEQn zYzONx2cxr~xFJ{7F6)a*pK^jgV(Gmu_OA~0L>c*a@MtJv<=rPA1O zkv8fOS_zVo>*!JMX1|@0%YV4{t6t?_7c?NN1wl9+(?m*mi-XUJYeKPKRYzw}E}_Pa z4cj;nTelTV@e_i`KM!Y>CU3YsABS;$!czB zapI615XlV-$!(z6G&Gt}96sogBq*CZ=-w|}IUDDyg5W$=GYwP=9l%P7WZKD4Mu{nF z$LFA*4k2fFJiXAtzrT3W>uWAzH*zv^Vh}(X$Qis2z=+l_fz^VGN#OD<}w7q%z3qSr@QPyas5c zHAn}#7itg zXtWVqU%NYVYea@6I!DgKGrG@c!(I!pc{u_S%<+ij<`b0Dw!T`JjF2|;A&ng%NESMv zY^0yayJyw&upr@-M?+A9$AGNbEzfO`jMr4e*{Y0=F$e6mJQkdO$l?^I=1 zduK-gaRSanZT0K_X-h`gi9KQq8RcEx4?ym_i`Fa+%A+eHwJ`5Kq&3*4hPvMkq%Nq8 zchGhG3=l`Z#*RpN!L6k0df!zTZA#Xwzxf?WokZYG9ro(xQS2rPvM1`R(7pASt4tY) zV}Q!2+qS?Bf!Dor+{*&;>Azf(KA5|SP74}{tQX|bo%0NzdtW^)NEZ!6Le(nS5a`&a zZNr=Tf=6WAVNP$V^Tt7AZ&hA2FKOc9b2;z7j94gk>J*;!z9ih>ddjmyy4LrpKeIDv zM^Ghr0Oq=Vwm_}HtjWf}#2^rx=*s&v`$ zkhYkM(wo97%sv&Klgi+=6>}GdRL^}Rwcm?j!%HxlOKuzThh07A)UwCsR)!0DoOo$% z!C3B}+f7hv<{UUENkoI|Fsz>_gWdU1Ls29#uFbGUWt9hL+`}UA%T^-$Ix)ir!uZh! zoQ%Ag9rr9sEd5}naW?v>I;NhKaEsqMv6VSuGO6yPSP1{`qSQ5#_K+_k50S^xO|pTo zD`7CRzHWpV7>W-H4Qba6Rau zE#ZGBhL45tf2YtOE%Rg=WD~s$`Pdtj8s0iPA+#Vcho3Fd30DjD1sw3c9dcY~uhD+S-Hvc` zz`Mg9EHrN%;^L_|u?evtig)RsE=f`BJIS*y3Nz2mz{RmcF(6H!Qi`PwH1a9`^{!%_ ztqHb@b^_sO#4=T-^%zLabTtwgj>i zsHj*E*B%r1(Dtgcv_BkdLcz0(Ia!|xzx?~x|J%I$c4UIuiC04wifj#0-;&5!MFMX? zw#g?`sRiLodyW0Ghi%0N>u_SljqH4iWt@06=kGDsKeQk-cw9NiWH5cx8zg6@RxppC z!}E&>C{w-TQ{jso_lBJ<@^%^4=nuJ^Bv*WPd8esb!v!mWOO6TL!O1zprAXeP7jcH zwocfjXP!&q7emi6RDr^5}arbAd$;kJ#p~m_XJJufVYP!Q|F9 zZT#p>-!0|~XbUUPVo?V0S5}C%%nGnMnaXw3??qPo47s4QQ0=x)xx*JJa&FU0LL6mB zM){X=eXgV2!XdjjnHJ~Gwz&kIuQZ>EzEbXqg?Y2w3#(~bgq!@ZO0M$Nl|0P>!6v_5 zicLWNT00vGZ`x((4{M4fdjhbtuz+49s+8Xd$q4U+a|oo zV;i=(f_s>Sc=pk@zhLyMEra`=+Hp!4{n?uBwHg9u*ExFAG6Y6(ohJNv>wus6D#vMIVPQp6%3myM zq94m!rtNz*Rn#Lcfu7m2pz+27j;mFea5Ml$xj!dNP=9@|c8z)e*h><3v!E&Sd008P zrdjk-&tm%8>^}Oo{EEMp-3}Bzv4YmH@kWwWD8gxEl#zDiuFMlFEg)%5 z9J2>LpkY%;A;m&<(q>9s=G7@YD5w@BPu9_Ow5C;w%wEWk(kX0Jwt3eD8R~y4A`60n zLR(e|OY^0I9FVDlv!E3b9gJ3GtI|F#V51I})uaE+kudo%>OJkTlamdZ=(?00p|DQzgc|xJh!c$ms<^?)Xkx5GgiUAtcwycHr4b-8M}M}Dd#`dmyGXR>^&Yy~CyU%7?B&8kagXbfz#)WwB7y4XiKv1 zH%%h9xJ@?B8~r1zCk}USpHS>WiVRTd#bgt-o9e6W&x)4@Rlo5t3}tmu-ENgfqY4|` zpqHaho-V`2H;AFNDw{($Q`O2w#kCooLfqLOIpl%{Dp!@x({D@j>E)hD0z7NL9UI_= zT<|k~+N<0qPG$Cj9w)TDp#&2i$R!UWkRQ2+j-&dbbh2x+JA&~u%0lBGRZH;C(ir}( zC6Id@a=}<#7u^FQtu-{>LUWFQd=qROe7E(T{I^q?GOrFd{70Y~M3-?dZX5o+zu*8s zbK~XE?QdV6{x`7^fGaXC|B}SNGyp6%nff+TY&u2OQR?jQ6;NrL9iGZ;k4hBePwSax zoo`oE3D`rA0!MPL@QA-{E~6`yi)kP zwcbWl{N&_>(Cyv9-ylsb86Ag}+i&1Qs&II#62x4%~x*2~hpK z!*|F9RipHp%;bq;98sbl9jl>YK9&yNwhWbvs)}F#V|AEbgVwCrryPQRDx?zr`A--B zeb?N@R9sLcuZr3*CrOm{$`agN_w~h8*PD-mmr%vjpAtf`pQ}^6Q`!Us%yo+LKveU< z?E~RkJrDYBq3Z=(Nj2LmMQyv0M^-(FW7R+vo ztW>tf6pK!M_wX;e=AMqR%P?CQ5}QWf!aeP=+p<*|rT&Ck(I|;C7lr3iYv9D8TMH8M z#&~V6YL~Y`i5WSch5|o}awJ$rA6P3PH2TF3++a=Wov{ZOvu>q z;&}LQ!^^)!sopoQrhiH184FDTd5}-(3+R{D@Yhd0Ce~y|mxt8z+L_ppVv**A>$cax zX9-9PgnDYsQjd$}ffs!YU~;_Wfp2L0Pj!TOrzjVAIB}-XLSJdQ7tj;a2c#HE1v0KI zC{55pR9@39N1FYIyo_j^na+6PaV)i7snGI(zY-5%`TYu@ebZ#B_WBplYr}U24!K~) zu0@{d7av~DW~%V+O|+r&#BSIc19;5Lkol&4u(jH?;zzddqC2q_IYj`7z3R)kB z2o%``E%pJ3$z1~WN#}YRo!cMOZ#+xZII(jJl3>G}+szaUi<*t#+zOi&yF%``Lw1Iw z%MOb=g*DSZ%6oY2L z^d=72;^ePTp*yy|Lm5%>*yHgD^2EhyB{o5geK<-ID0US^;;>z(YR=Nv(tC*2+c9x66KLKsatE zsT0PzX;2gt={g^H9}sA20ynYvT5G`pNS=3v-!xa#PsfUSXxpcGyI5@VDc%n6Hn)Lt z+Sl~IcXYmPt{rAUbgPwz#aCCMU@KGk$yKIGPcz!-w$}fkZ<6a#asU|g3V7XQ`J@cV z9y)GPrt$#4m%cvpfZ%`3yZVFQ{7n7l4}Sam->9WD>q01Mw6Lesf@H_K)H+)@mz;PZ zWPx+3Mc$^&RJO}NyTx;1IEa-l6(Aqv)(DKwLYt}o0l_(OvS5&bL#fP#pd-*QgtaHI zRw|MVdhAhwmcTCWu82%kEVU3=dGr>tMZVf|&EyWZJt|lx!BHA-IOw???Bptsrn!l> zS35b%SiW%Ddcg@tZ1?5dr9h`jEg;<^$Q-`m=1HAn4Fgl?Zo%r`MP<%n}tHLBGHhGJ^t^&dCE4h zj0p!;UBn42&Kr_Ad}v4Jt490se(`_)nrz@!9O=BT0oh|>CQB$5R6X)3_1f9JOsT@~ zvo+?4%jp=T2wY6MNF8YwcgYJP7n6;=96#jANoDX@rm`*OW?%_WPj*WOrW~SAc_#3x z-FAf}@EnGF+feVg<#*Cg)Vb#QWebVfT#&;6Ixg(lQfitN6+pYUVLHY+9rn;HzPDCB zO^eq%>cC54L!uTX9ybrcaf$XjuXQ9A0(R(&X6obkHv`ex#ZVpiv(P}M(yO~5kxp&U zZfPft9X7btMdzvd8J!)L^)CXWm0%yauz%5wZ9U#DT%t10G)L5*tRdx29H?zJnL8RN z7FtvHQR*b>i@5`lTcze2AbIR@Qr49fm>TlG4#S0de+JEzxBlB5Se8Znkwt-n#J?df)e=# z{=c@EI=WdHE7}{ilUD}Kgo%DjczAjWuZhM>>gH<7=ADb#KM#K4Rd&GFQzY0IXHV~H z`|rPfyK;xrh#bY=ztKt(o!I=qf__+@D2rktC6I;+V>pGh%TnD6WDps_qNgHJt5So+ z#5Md|VVF9u_GxE|AT9#Ufd^f6vV42kju$=sv5onQ?rQsSPPS!Y+LfRGfIPJ=F4@N+ z9_zGmMNrcoj*xW}n?jLfN^PiyMAsnyi@#!T()~= zDv^D0*PIhlJav3VgVQ(|=6Cx&?C0Ex(>OS9F#Mf)#e%PTQy8+W;A$Omp|_-&%Cisv zX_eOoU6Lm7ur>vv^H7T7x+h}1APcDBirVJ@&8~+yp?2b*pEP?ACjIc;lCnoCU*EDsYGK?35^sP zI{GL3ksq>%=%H+T&$Qgob%$B%*F@6 zVJu8uO&h=Vg_m<03+GL^=EQV8c*Qt7`Tw?hK3VU?v(qk<*=Y;KLO^8`r9J{?A~hT( zM-wE_jV$Py(F+XLkG{6Y=Tow6iY{Vd_&v}uHEeH~`l+~(ci45EXhUF^Vv{U!%0+*? z(xEsWD~1u!F9f0uCvd~?&FMcm`IZqXkN7!1AvL2diJTYe!>FDD-WM zsiG5jcj-jGPLFuV9U-@0AH4!fuu4F6@W1Z@opzdPtLOgkSW%rK*S#iiPy}j*@+Nxt zO%{t_#U68O?!AAi{?65T3%|2#%pLQ#fXu{lTHY8F)GUMS7x9N0e`^HPTU#5alFuiR zI1^6CL5jUkkvrII0b12l^XMuSa+Oz0D+BxJe!dReQB4nXPS!3v2}xQU$W9^Fb`QDa z)16^Sz;mJl&a7SDIRr|-v9{r=*C3tyM!o7RxTD*qECf=>)zLl7UR9b33&9@%A8qN} zg(3*Ur^^Q5F0INW0fM3(oL!wq?PPs)R?n(Z92gs2K{~`5P!(cYJaK7M;+^f64Xvpf zEi=f!2Mm(voK}cqMb~Fm`9lGumgxmv;|$OB zyhyc8yj}(@{*zFB{8{gdKg${XiCt^Wq(Y^+ z?9~<4BG*OCPkQqdoe47W*yGp^dRtqGmTUoQ*bGi%1%dS+cE^3@kq;_;-+>i8&R`9VLZ76h74&siHCN8z3)iruaS!FOjK zC7A4XJW9SCB~Ezx@-l~ivW%AtfBa>UEo8|~?6p`ROFp7Xb6v%t1`#^q$nbBFIBZpp zCsGVUZY5qEjt}!M5LJ|ZWG-OGr9#h%QITb0OOh!zfg-CYbsFRhGL^`=pbr<9Mpp~k zWtFn=hlG{sZu1dy{C2qZ)s^Pe4I|T$P8_YVK!ZC#zIgo@NeCU}r%3wfaxW;e#p}(&afE`EX2BR&as1Ff{RYd^$lxDWPyTnKi=qAQ`|p!>Cyx3) zFfjo)D7Kp-I!c|XswJ2)JO)y9dw-Ovys5Z8v(@ucqB#a~9Ov1EQ!rc6EbI|q0nP0) zKAzA!b=j=#flH#lknV?VSk2bZYu$H9D`_R7rv~Rkd4|f4XdbRzOyO2Ftx7R5u)@9{RYxaD zmV08&1s;GTCS9S*cC7(t82JQpJo-SUq?U9?z+b)hN24H?8?r11?rb~q6i;9Nwv4P% zVn7rAY38FJnCBR{Oixa1q%344%6UiR8A`*?R(;FCB__r1J}KwLhEzgEts$y9bSwX& zw2olIS{=O-e)oD6@vGgMg*y}nyc2j1*7CIlSYN`x{-3^nBjICnE^en?AzR?xu8wL} z%-8>Djp@=4KEB}FqHZ#0#RcQ6N@b;cg7(4wgZaEWab^q;mJ|GQF zJYQWlnXgV!>`96oLq?i@W(m73B3A{)-AF{UF9^%)`WcjishWdwQ^%C)Qkc0=hXA(jEpv%tOk0*@6~%uBs+9Hl#>eOO|_P`fU?q z4PT!;hhz(?<_rZc7a4A~V38I9ca&7PnDw+HQ@vi&w4+HrkxO{8hV;Y~dE&cVe&QpoxXrO|d&EQckH0nH%6b zgSCk+rnT(8fZA!042R0S8-jzLwy`UX7D_Kf|n z7sSp*kYymOPa~=ApD8`=kf#35aK489Bb3_i!_bC2>!-yBpzC_04xrkyS5^ zPa9xm^{YVNkYB&%r)-dD~vEwh;qP5Si4z#4Hq?}^K8vtuyiDaG2&Qd@eu7k{fS17A&(C&ki@|MV=&qwTF=bK91!ra?`axv^knpY~o25fx z6(Fz}?-2)udD*;t$?|ZmGM#tF{V}iyVVMIG-~eT>st)=I7LywKgfzvkLX5&Ub&99g zx5W&Qcy^hf7Wg2s3MP-n1zQCOvo$M@3N}kOdKL#RorEDC+}i54(R0}(>p;rc06m+r z#{A)Vzd&}e2Rs(m23>*)amZa-5xqsa8z{;;g-vf&(2oWC z!~>)?B0b~h{RqZ zCzQTe4r^a-cUNnH<`#XspZA3B47m|VGC|QqFn}YsP*EH0bldK> zo$hwGzn$H-`)#+C?bff|cD7$S)0wt|c)xH{P(cl#LIeRtLVIW%Ly`r z$hq0eA};E73M`K!*Hktsl4U5)XuIU^Tc0%kQ~f*pzxiJV^#Z)Mz@$TB#pRPxqX!>1 zs2^$$u_t66Eud`S#eC!-j-KDvR=j8CXu8bYIGvJhvsA646v-4xpkfPwLgwQe^4-&0 z!uQNtIk7i79g6*v-q}i5u%FfEU*iZrIEso94nN|@B)8j6zHN0=%0fO|PIj_OqPTI; ze9Q(F2Pg$dFO^cUZJaA!J@SK5EhLXq9+YP8BZ1Q0Y_gV@FGBWATyzFKu#gwn>K=II z1*Y+~=nU%JUL8;~Wl?@^1?ulC5$Z_~dR$j#LR~LvN}KDU@Fv_Is^s5w0!xB*UL8db zv$e3lGGf5kJQyPteiC3sk2Yat71`p(`+o;)@K8u8pfbfs#aif}pRpMQ{%?ghJj==lSZLGWUtH@HF_F>!Lkcv00%t=WLE&iZ~wA1yop>qUBM@?>p#$b(Uz*EJqGZ>{oxqj@-c(9*&IZjC)$oy)*lD zU#}pG4|)46&f|df&`tE&l(jyY^kvQ}9Zr4EJI7{W15imcGICrIo49*|I{f1V9Wo5A zPLQ(!6<%jGTSD_iM!Mf8%Y}_CY!UF%`|phD9%c+y>(%%lRd;J z?0;5D&mw&Ez$;CDj2G_%sXJs3%cM{HwsL#82?`8|^0ZH9)QM~6BAsI^G$y7*V|QKL z9H3Cl3rq~#rmUH3P_G2ea5mkEZ02qRXt6UtY<`>L_nK5Q5?*&s@6*~R;?+KSlqf;1 z0B~T<%$d!SC$Png{~H^FVExkB%(Ub4H2OQ)tuk>(|FzwI)A&<1j27n@y+P_3jqa<=0&! z50?%rA;Jx%rC@pucWF>=M9q|Z0cI!ucR4ZRh23)U{l2<|zqTf2Z_Q4cPHwSF>AUep z&vF|g@kf-Rmm)n-YcE_nWBt^-vJ}X4T;R3I)=zEVZ}!ZGrZ5y&|E6IlUe-Y6LR;>r8`(!lHDdvg4X(jxG07;5<(wn}!6txFDv*tG0_E?@H| zR$8K^LcJ^qc#x7M7dD42lq}FSkJVBZ;@qXY2G8As-C)!51@*)%_L(ODH#y{gD=q)+ z*Z&&(;lKUncfXDm(+UqlQE}I|)Y|iGyKSW<19LWR00CP9Z;J+^pnh*WTpFAMnSNu) zF~5T8$viz)=k5gLwEqCe)@64U`i+v?;r%|9!W(q^4^I629*FMrl1gC##|V9{sJu7y zB_Hx$jq0OQjK!$_g5e)#wXBW%$9H${v1d?rTfi`|C1cs;ZQeSlZ^{!N=jrn}i=oy8 zE29QvSo%~FT^oSgJFWo1@KfMW8QqITgJBRCoMwj&e!+L%xkn|rOCCk{kJZ`+N z2%R58T0z!Riggr8reY6h3|e#zD&{^8*CXG`6)$~R;Gm{ozRcU4l1WvKA&S_+$OUj= z8qz5IYqggZ7$51+|DLRPV}Oxs1B{K7B8?(zso3q>@*pg2S>Xls+xMe7WXYa-Ow?nx zE=`Whl)i~KmZZuoS0ARCbo%*KAH3gc-+%gsO=N#Mi6CgE7v5S9S3tVT|U!^;DKvs@6=tlFEM!; z7H*t%Whf-UYPU2w5FCRpR7!LvoflZ))yU6*IyX>%g2YNDZBjH#jP(B|N+D-fn;uxt0HgydkVn*dVJFFXlDUrkDX4 zk^+>8Z$qO*H?Ko>!1o%Quj3|)*7|nyy62j7mb+W&T1^+7#o4T0P(Q0we1vSL;p0;| z^fbOTDG1rK{qLN<)dWr|sYoQzPfF>~v2iY_!Gr*+7Pp$j{PzE!o>$nxX#hf?+ z$T4^$)A?nRFDTbm^7}{!@D$|GUH%T?fyFew2suaj$&sak)ik>AdVjKSQ~2wiQOli} zJFyl1XDP?VT3B;YYopX?=3K%&Di%EAg+Y&{6PHfBL@ohK6N)~Z73wV=Ls`6!7Y9MH z$z!^fdvlgcvMJ6%vLo1JcMmHhkL#`5;Prp42I=FZ6>pQy*G!Xanay@kAEkIek!~s$ zYuHS%S!^y(!PN_iZY$|dCYXo^(D6?*nZ`JPb!R?%r>Z{@ou6Y!!m3#J}5zKG7 zcDq)e$mtR;67)^c@Alp8i!CB?)FQ!tZjY>L>LyjW_K9y{B-CMB)N3C}md@$n?tcE; z??^F>f8uMVmG6-~2;V8VYOdwQv4P~_c+9@paqun_MZSirvR|xz|9fA(Pvpi+4$CKl@mX)X%Sw1ti6dTa0$uxO3T*0?6SuS+wknKtBAV z2QriH1i3X6w+ss;M@i$fnwZ|`y;FA515v1Xm8nVa>4;3^G|n&dx;7T1$Ropp^XK1w zL}ok;<5^SxQR4Z)KIVEkBzNN^e~t}qHc*OGilk7n&lWK&+XAXmWs0h0JLzI<@ewAe z##k?r4O;AH)Ro7~XrT_&zh|#f<+huK`)p9KgHix2woVs0b$%I;#ifMpSMQbGnYl7%-K2`hvUywS7EYN9 zQ@LzlF-$n;GGs8rVmKqRb?INV*y}j6*$;4I^TMF|e3s7Qp~tZzIDfi9y(g%R=+B3> zOL8NO^dYSt1n&5Cp!(h}!+XvOkIPVi$gBaM3&EE|5&y`icI22E`R{b($$$7;>D&Kd zbw|cO{ikW9-HqLm#WwNL14_|NkuRy(O!_gwiO%q=;30$Y{b{(spuk7FBwkP(piiE@ z4LG1S%~=ACs!g1I{{245JZ$_p=U)-FVR}654fV_|<({S2PtxO4wE@XIY%47xhB-a5 z&p^fXK;RScpa;Wlokm5<38Q4YOPP(>Cj z*gxlB6lAJ-3l?N+yOl+L+jx4USBeg2CC zRa3j@g-BAC$3fd=P`7%0@`lW?+cJT9-uk7NJ#mcN`cMp1Fh&}ZG|(Ac1A+Hl8VY7C zeQekm?c#;(U3-R=yEN`25D2r6%ij=pi6M$V3Tllvx#z%0Vv%QyJlV4u zghn9Ie8sC2JflbN?fOMJtXYE|*dy0UTgq@;?kFK997U9^^X*zR;piX!?CZPdTAih) z_mK&qDJ0lYmosqclP!-(gjWoMZ1AYVqJJaOVpillpbjyFa zu+RI0RvMNgEfB=Wv5}>XQycsRu!kAcOGDOh8=c~$%0g3n!H1H#=pXArLS?M+|46Hhb>lKeKay8 zv9dSg^?z>=+UI!Ob`@gC=j`J4K}!F+{K3o~ZU;Y6w4d`iP)IdJ-i4t3qZ`VP&Pl;v z9od&TqNi9PV?@7UWDl`I2DQPXe&6ffR-G1sTJmY~YcsM%@v44ri||#~r1Rt~c;91* z37j#9aWS r07_aPhA*Gg|D)US6xLFau#3wwbic(xKme17{KRmH?jiDwjdUaeKprl~>OA$S*Qu z<-1oBCr-6`woBfhw1#B9F?|IUHYnOfDWIA>52S;Imxb9=Tlw1|7j#43@103k1ej}c zwrltKTlCPjOof~=ssbCd_~ESQ!KgxorIt{ClG757L)@GdC#VR>7j2ok$;%~x8ok{? zkc}KOOi#{Ah?;2czrI%1g~6w-q!T$vp@|Wae!pLa1e1jYfxSW~rAg$J$dK}IEd3W& z^UpL{E;!F@OV&Rs&yZSe$?@-f|0-GQ#!)gzfDUO#+e#@iDYBW0MLwX?DMio?8z(Rq zEEPreg2D`zyjcFLKkb<*(swFz$whfNeLq@{mBp7i6?5}k9r3UNgv|ou0u&4xk{J-g z<^R8V=SqUa3KHQx;eV23b_1yX+6Mu9=UNrB4 z?3~W&s2K@DY~W%&aJCI(M#wP!2zBSO=bj%}{gF@qIq;|C#B0_&aLWcs&6L7Kk#iPD zcAx(VMRAb6Z1&9%6n!w#i6VW47gijgKw6$SL1okqdK5-p@xr42B-Ikm16h|`UnXh~ zsMQ$MSHkW=r9Tc)7Kutnu$P0O5XQ(QvOY-`rvSPF3~J;8$%hzt-NY{8TAwH4Jh6U# z^nKy_=!!_})vwj`a0fkh_?)Hnm0UbgCu`+fp0!koJ5IOhEW{0mLdJ6rz>^tqA5Z9S zvQOQ*txaM`)1DxOkyf_UW=XB4f~Vib-J(nsb<_L&D}@GaOL%!uzdYN!d|D=bCum2+ zfk3RA`ciE4N>D)C6BfQ$`0+MRelYWjYmy}_QvHk*?TTc{s~T*Fu}0(H+h6#T8CGjF zza?)C$sf)dxp9T&Q5$PiPAMRlyBAZpIm#YcP0X^WO{y%P0okXX+qDUPyEUo6e}@C$!%KrtYnvrc{SO3gnc5CJ+L*mF3!RYj zk(jV?-YVy0W8;7saqJdmW0M{AZpCQmciq@`W}x3SsM7tRsh>O%#|ui(S{{U2L=D@yCw=$9Ak__$e^Tif7!;j+pv-=vhB-2*Y();(S6C|Ep;>C@PPL_?) zSxYIBD6$3>KJx`_@=v`@vXU84UV?%$b32m3g=JySB+Rw{PtP<;x~q>v9oj^ z>GR&hxlgYpje-VYpE!M5pLoDyBh*q~ip&yQj!n86|FvYj{|W1L=h7q>6vloNX7&LI>|%6`Fu(N zY}VVUSj;Dt(TfCnbm+|7I>$&iib2(iE{HeDyQk#~^hmw4fAT60IEiWU6|ins@>}_q z#;cOZnkjm$6hZ<)7rmB_%!!TbH_U-6W{KULxNRq3z{0QnY}&J2$WNuq$W}MDAJsM} z-a{!MKrf(TGlK3xPizA}pSz!XPTD0;i2>pj-#wtb(an1heNI}(HIO7#L3Ba%%~>WL zcF?|j7VoX&K84t(g;x;03;us|mJ4i^v(pQx!`KSHAA;8GC|GLMK+M@Gz%> zYgWa?HufvR@)*dX!ltqd=KfVy$Z=e}Y82;RV}*-x`)*w?vFFis+x!RvXKoIS1eiCu z7dTy0x}?2tAB-|rJ&H`a?PQa@T@n|X<#}CPCoYgw@Wu=zj>E&U`Z*s6!;i=c9#sG3 z-~V#5b**`K=P!OmHor0POS#R;R6r?UnY>0TltolZz*JgyS9irR|btsLaP|D3^haEx7C*tl^ct_CBw3(l5cc z?w8a^mQ8TkWWSh}kxl!H=Uw%QolU8_DJ`|niMZ{O!;s}L2VU4>(?V|hq;V_MsVXC6 zOveBN_H*bMChVVb{V^*{jQ{cOt7l(#!_g?klz6eSm5+Jz9yx5%WPA6>k#H8%B?fh# z5;TF3KCqgu3ce;znIc7%5$61nYz7-MkFKjqOR=qLt%MKf?WKfC?6iKCGYc-joKJT?E zC=)=AoGeb7JSzg*e^EGf(Bp9Q0WGF~SNn_|gdBj)(4}WgAcMbTbKI6VjTIarnV+PP z+&2asbvEGGM=44uvfEP0dM3DEj_tz8J-Ho}r=VoSLgbecouka=o(bLxM1X zz-+yTS-jo1Vj{ov#s{x^rCB?@E1->o5>v2?*aS-l@}V_L_5>w^E7eY;3K42oVzwO% zf#L-#L8@%fqlFklaHAjZdnCOzbA_w=BoLF1!T6>3|M>}5;hGhU$L&b0`g(|U?fK~A zwu5BH1OildL%hFFDFw(|?4e?jC}w#qE0qy@Cp_HNss2bP3ZHAUtcN zHwJYB+wK*hq<`eEFM^!-0beW}Uh02Utj8Xv6d)q+q8p`9WFEb0RrO;}<6BnfJpA@|{*BbS@&5018%&?26lW-Ml8Rl*LuP0s zq^=^s1B*Tc8k9A>Zb+(Q>cpUa8vY<0#ZQePCrH}F6_n+CykLtGDid)l5G~FV-%9!+ zF;geuVXeG8C?)zat;h4H1^UIjTVw~Q_cTZgB=`y5k@RV4^2cBA65<-{zRbtH28F<4;cWT;?EcL8G(_D1 z@_AQe`fGu0>P_-Y5u0z0#XpAWR>!}5rV(6op4FG{zO7617G})1Ps_5YZ*}8T3j>7? zx-z)6mnPo^x={K$D7)XJuAc$5j7=c1SViu7w(~FgxC7=|d*RRUg?IxcE<6PTAWLh8GYMR%iSh_#48~gB$pX z9Fwj^erv`<=-+SST=8m?XDe^Zk)^f5%b;osE8rz_cS=qO4!NqYW$5=eqPcMS>&J2z z#eDEF8sBDp6o;Je|n}zijrOJMJleAn5 ziki5LwenYRO6YYH*L!2zST8vXB@L*I*h=&blDnSwqae=jl;)X}cTarrq2P|U?+c-( z!7S5oLwP@Hk?tf=Oo}$CEWb53GB5Ip?;E1_@!@spukEWMUn^RsZ8k7WqZDf?l7!8} znaWdp)vT#yoR0f_|D#6T20ZB zckif3n;V;=cpH7G`;_7yMeb0sZBdswo1iM?(^-RGOX0P_LV8KD$gi9965LKS>CVp1 z)!vWZCdL=qau+@UsiE+eaFgx`DbcJ8+Xceo=jJ_(YK!X79S*)Zt3`-+TdtMS52NnU znL)Rso{HO|b^znoi{n1ObFp6l(MQKMef1!pr<2_g|KITdp_h@C0VPvVOAX8&Chr@S_0_{1cXdA7IZ2w`Qiq z3`TZD0v9SWptkxna*>5BZcYw;lY47si4aKv%3_dsrCCxU*$*NRjeLD0G_^wB8=pdd zKWG! zl9e?YpZfj(%(7?XdD%H~;}j1A?@l{?L)q_hJp#mg!{9B)S9Lqk%T>(+Z>d{a9&|uk zt3ggDESDH7PsuUF3|Xy43IUeSF(c-ezkD=4-0C7_tG2HvMH3*gKXe1Dl2QQaz^7Df zqx44jV7NhDN-lElDocQo`twQq6h~wwq-5f*P?Ii4dtzFHEQdzwzN`q#T|Hb7qny$| z$D}h(dm_(}?Dee;HmFw!65vdu^p4a>pHgRe;|XPL|PfvLB~PAZ710d z?IU}XIm+7EZSs8~@u5z_Bm*Rdo@NQ#p^v)sGb`ARGnPt!|F+ev#c2GO$l56Nj#4-W2L<+`wbZjrFZuZMei##5e2*Tk`mdtHp1m|sR4 zHO@m9OY{tH=U$w38lqUChid(!daclko|T)f|B@uItL$^*Kx(HA*0xZJ42o=^Vs}Bc z2Zk7_qO#fhW-kpM3`8yfJknnm7trl=QAj4eJP6Iw-f8{5W8@W&08LIq%&P_}@Q9}T ztoUet`@;Jqbplk{40T?%Qwm`1H&C&~$`09K*ap(4&4j%{)Lw6dh_hXy2k!-3#N4S} zQyYC-`CZZ+h$7qM{oLMYeVO<+RB5NGvL+T#&U-b7`3H^Q`+fdy4l|t7PO*Ah-8X{B zSpTVy*a{QbC*N!$Np8Fd724o8lTvJ^NIIs1O9%+PU@_l*Zb5XaC~l5fc05mu#-SwI ze6V!-y0E=dP>TPK#Gy9D`~Ao#5k zqHqxA+r}>JwV4G+5MlGc7#*-e#Q1-gP?KKw>bwWC9{EGEL9}6dJnuH|h%d@{;AcFK zW$bv(pe|Q$Rt;NcF~FcDazI-cY1L!cMpdZWo%OYhVbU$1u*NrwvyIy&+$Br!Tfv$k!kM^wrD<~jOE$NJz89~u;;LZq zy?rE$9j@Hh;U|?gxY|uAz^l%uVsTe#vnoTSN9l*t5gj+ zT4++>-lNkM2ZJiS@D9u81jVYimQ2E@P;v&hntNwr`_3tKq9|Ki2g$sX0F1ky{qkjR z;a!=uNr9u`=ZdKYb(gF&DsOHbuMD1i09zi5y-xxH??9JAXH*?8L9u^Ey9DpWax#PZ z0wz zn4qH?ed_GWFx(C^0J>^4TU8x0#2ShlLF0=->uh85(u|H2P%jPS<`SOcdM`;5DG(~ z>5%Rvd#Ci6ez4`GMQG?NCc)QuK501mQ1TLE*r4;x>6#KN{*|%{TWah@^CNOAe zaxCb<_H=N=`@QRf76;XXzttnVpxW!T#OqGbGH>$=*dSXMv|MX?+v1O>$xj0bSEtk% zQVPi?r`3&^C*P?s8v&e{9_i0n+po1+qT{iDSx5G;i{0E9I1M&-tCmt!QRE@oJ+K5qsXVN6D=3RKs|=!WR*Ga&vA226lJd8kB{kE4N*YnIpZjR0 zg|H8nrYZuh=OHc7MV=j?UmS7~g0$T+8)xfV>7H=(OLoUB^>@MqGk)cxnt8@iM}L9c zD#hvlNm-Z`Co9vscay>i1e}Q>(@{YwfQNTK6${Fmf;QQaDP8`H;b*q@MPb!csC4;P z(+?A=k0IIxSM-P`QFO%fc;pe!CTVSOW>7Jwo3@;T8>J{e)4|27cx1V{N0C4MsLV3R zM+UX^L#`I@a?0$DYMP$EkhUL2KS^&)N8tH&S0CAv60iyF-8g{_#U(>tgLRZ5nIZ|G z?(w~A;#U45!I%EEWCy*BSH)X8VXqfYzy*d)&5})!KQyRQq6;L>z{e{lax~C+#jvBk z#0(ux>yK|uv>KYu|G4sJvXWh_#S7%f<_kQ#j;zgX} z+BiX_(4gv;IR_EaFsmv6E3csUFlF{p zY{1RO6$BK&+5`HauOLy>rpD*7vHJ^6dnD4sB`7et>a=-b^`(Ecb$1#C2J@>rOy`UH z?(3{x{b2rYtl0TR;o(W-b9Q#ceg9=-h0W$gKc#p`k^7iCewLp*%;^fpJq4hS#k5Sb zyoR?@x($4jHb}=5NQ_gq!mj6n?x4N&0(~2&HaJyPs<{+mP_GNR7FI zeU_W?!P;Q#Y_}ZaTT#jB_eQc9EFsGRR|*w63jGdopAEl6VUpnpDKqy1%f)`JeuKzz zDM_`@3oqa+RZCwE#-3zg=vY5V|71?rB=g^MHKxt_n>Pxs!wv}N9F!zsJwUZ&07~%l z0`mg(JK!JUwi_2Ay_rL*?nGBQ!_j6#DjiAC+OSgsPNxC*9|ENnLL=9L0j2%jMvVj{G)ezw3i< z%+UPAijtu0mR@3F=gPP-etK>2(@81XDbh;C-v2gM_njas<+;#$UIMZ{b>cc$|JO|{ z2G3y|w_@r&sOmZ#+)FZP#E(h0CG=`wG2Kd|GElb9Y3=7Bt3ymWZ1ueFUmo)XRPkUn z7~cC(vw2!FuaJi%CYf-?q$_y`DdRIF#{>q|o{$INlPrZc_!HzXh#TSvmfwf|M>}{Q zM}V0h3Q{?Df{Y=(oKzJ|2iUhlMTR1kmlB;4bC`2jdnEFU@K+*vgdnuT-UU1Aw2T1(-$k?Q~Bx7B>JjjG!R8pNpO;N@^Qv;ELPl4`uGq)Hw#d znj@U^S522wt}z>)Vb-E?KmPI?r|g$3w{6B^Se`a&>$sg!dt}|btKy~bl0VSm%iSp4 zAW9}#;xt|x8Subs3ksr>gELg+@YXj-t9hj{_g$|PHA0lQ+;w4v5=vG0v}}%b$qHP& zdMEkRjkgcZ+blvyDaB!m)M8D8rP2>;&Mig)nqRtUd*TyVB( zYJv}h)ba=6CE7ucBcAP0(bqI8cASwy$-!l1#Kwmel*Xr}z4sFHF`HQ0jUy+9(&J_c zs*eHL8Rjv!YYztAl@$4*g}Ohp09LVk-c?ie$T5(}shP6P2f0RjXJ$`*rN1%axw3h1 zWPq|6!X*dge*f}a`x1yZ1R@NT9B4eUz4wa-cu6X=87d`9c-wpiJgVMG@Y&{L(Q0p& z+>fpd)1!F%DyY8F@8(|SB=Zb2YJgxNRdvR{6^c#GcL3>a0FHP?(@pTTen%#jli)vMedSwK?Q@A@YLKeJkms738F$dw5My3dAGigZzm&neRJ ztg0TUnxS|oj;f?@&<}mgZO&$w52MmS4-_CMqH|R_v`Jm8?2bv3xAME``~0)Mdi+nD z`N2zb_sNW~!D}Ai5yZ2HK?AG-%6Xq%NaWn~9`tD9ps*c|W=VM9*?aOwKm^?e4RO!^ z0t+{nWpDX1!ML`YZj;CRmv4a^jn(Mzqc>2dbtkIjrRx`%BYqFZj^M>E9GCvuZeL7pqpMCx`DG3de;`uZ{gtn z0XD~trRiY> zjO3W7xuMcXKk;54QL6bMC3=mbPFy5x2a$#@bAx3Y?Cj|4K+yr?;^n+bVQnxP-FN}U z!ymW~Rje@dve|PPh&KFeTft}u0NfZU3|Y9#q*l`i`kAonN9)UCE<~nrE7ZsZ3wuUw zay+oCfT-P_s=DWm{6}qa*CFHi)Q$`y&mVW$2gUcloh!E*lMN}s7s>iJ#+dB2F($c` z00*L1$vriQmK{n?&)VE^#{~-uni=* zI)C9hzvB6Xvckx7JMjE*mwjM|5zc&V$n0l9@PngiWd9rEw*sf$kRZ68QULEu4HcWG zToTbEN5%yZ28_-Ngt|D%I^WNLo#lfMKEM=yzF@QGUf~CgQWT5Bn(aJJ7B4+Got&F} z*Ap44){+yxy9K*p!NOLm0?AH#AZkGNsV6o$<^>{U><25LVBs(QbMllI*TSyJB z-}~Xbe(xl12aT;Bd#8*vjvg8Ao&(T<4H%DBmxo16j zPri|hj`2V+)%J7W;U|e-C&kC22^dBB2%U-kUe@LkffzWzN zv5q3im}ASBlPS72GYP~Vkvk+wo*7h6=l!1#GT%x0&eeIDK`ufC!|bpC3&RMmIn4}~ z^MBZp%C`bz{)G*P$f`GHMcQElnG8y?fg-6u$)Q3@#4Sqw$JfL?+!ca+(Oo`J`j$?- zMoMMvG5Yhq$7W&e_Hx12fH6W33)ol}6xYDSe$;Izj33xr4K~GYZk&l_uqP$-D&LB* zR<3?c%*vQE+O{ZtddL@=PWp^iUq1Pc|7P`qd)^Di2o*3JEP-J^e=a-DEBw9AKL6^r zu?|B<)i|YrUmLvX8x8Z2Cl1LowsD&kTTyrb<+iYmqr$64o)xhY`kIYkGE9)>h?^KU z@mXNxS{SgL`k{`d%cdO%W?o8$%d&Ach6UNn{xPWQVb6C)(5fGQA#3t&^SvP47to)Q`dh5L;)RuBdD;xWqq1hn2897SBn-d=iPGXdT;$d;(&^JWAmYHcz9bsyi=76E z;iquqKpB2KXPspQ6!%Se*ZTeH`0L&&W+9xo2mAkB!eV7-XgBbh_H)Yuukn+(`M}qM z)mJBl1rRUeJ==VWBKOTcEk31gt%kU+o_<|z0sMuN!?))Wz_p?^m?+vqTQuj$v=Qc;17uYRrQ&$Rm<>#PI zO*y93F_t-PFXh*%J_{ys(_CmC39t1 zBGj$i32KL3XM-B45Id!tBT__WS-8h^r7%Bm2`33)_()P21{wnqSHgx#0*?Yi&%JVI ze6WWRtCefq+WX&|_+2aFO8lPuD*??DHipT4U3>y8977Vj7bry&MH;Et{J=KZL19ny zV!?|~Y4XqL_1rwJLA_LQbB0NG0SKehA{zMJyn^V*+-y;?=(u#TAWMvdFa^`wWW}Pb zx_k+3dZ6e}p_S&`is6xjoS)-TFzTzL(}c5m$Vt z3&bz6n<*|WpML*?j}lpUUSo_yd$=(kAAi69riT?-1&@C9 zF^OY`EccBGKYm~abNrKmy_%{Q`xLaZfrXks9!6Axcht<(q&|Z z>X9ch6PR34i#C_8B6_@nJPJjAs6vtuvuTc5rhPly(I&qV*1*M6z{ml#ujbBfgSMOA zH3#>*4@miPE5g$)IkL8>i^85DHI!3ibf(#TI29xgXsw!*CtjK|mQR`KkP+HJiQhTY~JK!Ad+Ug(e_qd2nhSW4xJm3ybOyEwh_ z()2h!)h|8j(i7~wl9ZBnBvY&r%~QpRmBcVysgoP8AD}2QBqA=N6gw%h<5`FHUNOph z0uO?`Ho&0Hfy5jT3hwjA&algz4b!b3m-=7gF1+iRCPy_0v#4X9_G-*Z=>u6;_<4Za zPQk*1{7SmS1=RnJgUHrc*c~|?P|P6u!N5>{=Y_tQO((&_=^TX2@EE47VZ7|7=Ib-8HfDZH-Wrnc#u?zFHuj{PQtYS5UMd#Z zGO+UYyl?%CV_%1y8dUbjUiq1l`xEN9@DtBt~ZzCqO)o(p34`V5GF}0xjnM3 zNX*vu`&fb?3|%nnwlI1b={i}3&jpQ>{90#$=Y_3#0nE{SVg;UY())i9+m~E2)GFd@ ziwXr8IXt5Y&J%1WXGpu`CU>j$uJAM1-|LNb()a7gcff8A92wC$_pq7+_sxe#4g4Dvpst*Ax5z@WzVLNrBaK;i`kEvDkosN@TbnsqS-2$1xN zq!V%`SfrN&Z$(qs`Z*=hF9wHD30E-ODneP;26_dxUs4Cok~yijkX2fb9oS!(l@Wbn z@gA@O`?%i@eBR_~#Z>**s{cZkxG|!|7!oZFo@a$fpnNd6ojw?B3~EyV zV|z7?AL0f5-c_?UN9aL<3yex4D9k`w4mJ+=a275mTj@kzhl_DG3R~k8@GuSHt&I8B z>)vU0OuNeyYoVY_VEW=2i^Eq>Tk5}6vD>STEDrCY4^C}`{9F0khi2^d(l<)`eO67( zq>Uk0!yBciAf=x#sGew0UnXtx6P&y$i)XmX*%}1^Eb9%^436?VE2}atRdqX5YF%uO zfA9NO$=V5Iw~bBTN+~iavYCqQ_fFTA&5N63P+tw)q>h^dHE0E#69h|iah08XqMdd1cnb}_CN|<41g>W(6*WqfEKH0OIZuF}Y>r?#pPC?q2eB{|( zBTbVZ&=f;V=Dfwu$+%&KjaS|L$-3_}{Mw3}TeFj*l+ z2Mp>~pHD;0cRf38z7x}JZNwylDcGiopPnXv-hG%aDi2DNAJ?XP=0?;^$roTqjKAdB znZ1zgM{eXMHISj=WZ(6?N5_+E;w)a1darV^C;m0qNe?N!HrR2lnx&5uED6OdUnXon z?+&aLH-Hm_N>m2zfk5~J4BA~>w^hA-L)O6`{o<#C|HG7Hty`Y>0yRr;Ljp3+jnV>6 zjy6@hZej<&Te{8{Wsy7R>-1&_`pqRYcO~ci3m}EFGUi$5)_R`?x{`mF)?@J=?sOp; z#2Io}qF)(P04tqk6QdlICG>f|eUchZn;Zu@?q5FHpx#PO#=!P@mZ+8A4~nrh@XO3e zZp+~7eN2kDs1(vAyd1n?I=l&YJ!>Xi7RE)v&bu%{fhSX<8|V&MvZo#oK=+ova$+Az z_gmp!f{ZYS#>|*Chxv_VC z%Eptbq7<;%UPi?x`>qV^rsJp^^1nQgRccl#Ou82Neoa4pf^%d_8C5LLn$af9;~a|J zN^kj>Yx7nHBLDdQ8SRo5dB69{z@wnrhMJ=$U3xI6OiS<#UIB3ekZvi`Y}VZJ&xWJ> zlIv5~bIYh~?<3A}tXN>nVfWh^;A=ZPO&o4TP_}COdQ$Y7)nJ^o!AT{hI6#q4saRYk zc5$zOnFs#~U8H5I9$*2lq~iqLd|+BlRh32V)ErO)87t;88hy7wGjjgq)j=uTqp~*n z5~%@}EhBvpTsh0a1NA>p6N-!BBF}E|I><9wYD6og?UEhtt8s3f{=G6PasL$@pvDaC4vte|3XAAPk?ML@2$Fj8L^Q5WI*wAFC$t|OB_ z+;grz$2|GNRa~4(_?w?humVEY{=S|pnm{0nIAp6Rg;FF^WHpAx`2v)Gz^=1GzeC)3 z0ftNe9bdj9U;WrCfwScBaNl~`zuIRwXTLMSCX494|1!eBmR;`TxWf2MSogKh-?EyJz~59XARFCy8+@OQZQDU9psQdjZi9Ex+v$GpSx6&6 zRXBE>n0*nHj=<*Z9&VBK{Jk01o6ht@6PfuD8dBGdugRj;LAKh0! zy2&5i?|#+O>{gqWhkui64;40+#f>8%24KN1gR)*^`EQf1QuvVE)Sc6%$!|#yOX_$v zQ&z^LdEI+ES99O{xU!$l)m)PGsgHu6yD=m`(4fX%@_La=tjV)U9SyLajp2%)4V%AN zV4p={0~I%3gc$NGCbdD`ELl2zN$8+Q15othZY>5zmfZuVb*?!&S&jq4(Gi_? z-P5egP47>o%gEL@CW}&SvpDUc6hJ3lK*h!hZq3Y6*NUJ0G=cqql0SZhS7+o}a)B5* zt$e&*MT(#ZXK`qKL?^^`z2Z(%JMC#8>fA!7y$G-3T?WvY+bSsyEO&*_t>d71m7n4q zpbQ=5+dn?I-rEYAvTuF&ELrWwJ_@in4~f4vQHqTee6O*GIfGHBwNJ$Pf()r1x%wcP z9=%%=FSt3YO8iiO?QpA9*sb7^s|4iavv1r!7$y`CXcu%qyOE^4$TCI`eiWu|8poI}~q}>0n zo~x0~zwL7J>fbQ3tE^1Q_zm%wJ?%}3+cv*37>S*Nt%^Qk;mJu7UK z=8KGUb#TYqtz_>M2$tgn2V#)L4Ex-gCFdi@!c!d~NJcjEES_QqlK=N^$*j@1vTnSG z%;3tF5@g1U_uN0*VptLs#j=u#aiRI(mOhfz&-rXhBNQ1}yW;xD9T0EKo_Z4Yg;0Yy znH-GDr0>e~&2l`^cPMfj_Jw(ckpr?cd4?oa1k;T_TIpLe_4%S|C@o0j9EEP|in)c6U|o=! zAytI6RCq_)M7+D9&q>V}T3wzVJI_W_AcGK8VhA1l@orR*8U^U_s7Lc(rRO2L9 z<^|H zQRfdXdsNsrM!q5OG(#I@FRAm#s^BKw>0e$GpVRH(6mbip+sPBowy|s?x8TtcbmQ{j zZ1jHBc+3A>P;M{s$p#i~oYH5I1hue&V;K^1%GYWtV588?1A!&wsj5u+QB;u&xMG+m z!cl0Ly$j}HoPK4dfTJEwB;>xfkT66-$6+n(r| zUNMZ*FRgpxLvgG@@^6pz9wm$2I7rU636j@Qie!o;P_aA32Q+30J&Up)>PaIN{1fkd z*VGriY7!is=U0v4ifgPK=5c#B7jL#_W^mgumVs+w!9ic_hbs~u<7^l2;_dJ$;x1U> zTjmoNdR?}dS3)ls@Hob4CzdNc;a|w1IhFK(r9nV=O zS;2&Qa7~b6pY34dr@HaF!jK!;9$h5-LmpK|c2a-nHY+=~%W_rSyr;ZwX}azPq~Geq zH)q8O#%^t4i6mzMp`{bKVBxwREw6WirlmZIZ(~$_ylH*4G+r}21y_WlVQ+V1 z>%qX@-o*Ly4f#2UPIBpG;FK=Y6>?9II$ozZOWYpWNDs)Me=DR}nWP%gi1?lkT4Uyzl*hmheTqT9BT1;>$6T zS74Ym%X;c7jfTTFJh*sdt-xws-Y=Q>Az3qllf=De${OpE)zfEV{= z{G7JJqxbJsCrCWI(nt3-nk35xNSqZE+n*ha+`ijA84k|g8&#*?g5DM%p+ooTZN7#4!xorT<(1utQoD8hW?esqXWKIL{P?X1{s*a4^5;tTn z9m2$r`^V}dD@=^*3pjlAe^#{HBj2LT5@c&@I1eLF_;zToD=X>V=q;L(=)Ukac^M4_ zC9_JMLnPqflskl^XU4_BF`aRnlK+=%u61qk2=a(0OWk;F*=Do0q*02s6iK3D4dfCO zx1OO-YKv$HDQ0v->A@LMnQ;Bov-IaOSFRbu&49x*J>0`CJjcprQ2)GX{r|I9e0JM) zgh9i3KQ})Rc>xd6cf_mSS~BUnxPZQ{{3KH}Aj8w9kb>y#;uW5`R2sKr%Br_2TrFj9 zom__j^2{8J(LO#qfFz82=Qs9_2E%%uP1f?7E#0;?Q__QrIjbakq$nNhH`8JIX&1hQ z!w=2st9Rcb30=@X-W_jUOqMJUJxm7On0sW4&0>;7Db`SArG*0nx-uhcrd;M!%++u9 zOK=%&@$gt+VI^KUfQv4&gKA6v@dSG-@iJ4I`{rYqm{PiVD>xUUQ+S)y_0a}(A!#Fx zK4)A^b3XipI<%S$|ENpPGULSf_CkTZYr$so&5d(h42}ru5w`MkfGiQi+LGvf{uW7H z{bf?kuZNQMUUE;|0lP&#&>DfX-D9&dIs}~;P6&Q-DP*1v{x#3G4*SxMmyL!S!+@dO z16x||(E}daL2hh|CXa*k)kr;(>xw?@XJ016PY36o;c3^MW;PoozgaE#YIQ>Qb&d#w zI^i{W1ONQnHH;$2o7`(+{UceGvWvdS#VX`M zkA42_Ku6pPq{Yd+IKlZzXN7BK>RZGKDi<~9IS3eAb76Pn3fM3MW?JcqI{O-Rw+&$# zs?M>f*f<%usw~P)SGySMrNN(jyFT-K1*qI3Pp$k#N3M`UkdGs-snHzu|aVv6n}0KZuEcZ zpW$8M6(=a>WUj;%z+&XEav`Q*=z}Eo&-OrJ_=by>#{}pBHt7;$v5_dA+>pnU z?U1>?H1|H881@9}IUD%%z=}B@Le2#CiW)N%I0-tT5re35LRyu+{b%w3ibkZtQ$8qSg}OmoMT(*DFhTM7Cp$rV zqIzBjKMq{+ByO6#mH($tL($}9@y-exq@ey6oZ|P8e4Nbt^FwIn$oRW-D*ct zu&kPQr^UWl=ndJqVJI!iib#{+3q8e8nco$p&(q=$$#ZxlR_o2U zzj;S?;eS@qDTd*obOr9pQ5t%pJg_RjC@Ilac&&=qAJHW} zgVEpE@ZfQ*eYWhiieDH~dvT#R!M{M!Sd%V`ljpmKQ!4I>P8Q_(ADL3&RS|Y=MuM&a z6wIo%PkGf^-0TM%!_TE}haZi(9d2|@e&)q=v4X=hSeSpl7{e8xm<`LXFMr$?V0Bux z2`j6}7B{X=2BEVdoiBxy0{Z%lRBZkHYvML#G3@R(@Sg^@lDNtHLVM+Lf=Xd}aId^l zxM8|UhqP%3n_991?%L1o^KRgm$KZWk!o#9-VMoaBxl0w;Ch02kF>-8Jf#M2az?jU~ z7$(S}R<`~i>u;@i`CZ5FZjzI3jF-=C@N$__T%^c(*nttGO@>|c_LK>u67F{pEZY6f+{aP+u|3y~=p z(=Yhj8>PsySi`~W@}Nf(#{o=^8eTT$$kBtZ#w06vjnDhR-Z}OhRt)4%&@XM$88zwJ zY>>)*I-{MRPF4iR2|g7!MK*AoBC92fV2w!SUW|mihANpH;ns0Yx~nnUNFA7fUa~P{ zMR2J&d3L||<=`%9qi-(ARCP=7E(gBsul{G-+~K6Tu^H1D2dw7CeUqJ^CVesYe^$_Z zC;Up7h4o^6M5#uBvRIZWEibw>`!wlUC*ooXFN)Qz*^92=NjZ%a9JlXAa^}9=cU{J>k zGDTg$<(#9O7 z-ny`6xe*|g6=9Kn&7?~s4eE}_Et-7LH^A$oFp7@D!_ie|=zSvx4lC<7{#W&doByvB zOdlt$c$;*x%kH}GUKd$5eyCgBM=2gqq??LG(Ic#VM2`nqy!C)X(iwSzJf&o&3%9}tOdRTSv|E%Nbi2$54N=+JI-Wj0s6D`>?g(#`o|~2f zKHb80+9DVBPzI`$1@ynW7N|o(%kFcs@=D1JEA-~KVN5b_sHFBX+V z=ZW=eyw-Sa*W&J`es|1L|7`#a(8g}^U+Uizok{1(b7&A@2}xDyM`7Kb*&o+{=h^5Z z|9Rk|@9$stVoZ~IneKqb`d?iuKVEP(u#Q*3s}xp41h+_V8N}OeMs;ZGJ)6Vo=|39O zc#s2c_Z8s=fPv5F$gsaPY-Jx1bB=Jk;#wF~`^S5DGb;^CD~yY3=8JoNwS+J)&>Ja~w-`$>)) zXLAnO;GvXK02B2t)Q+sAF_igC(j&$|v`4ln0$icvFZOWIWoR9_uT9>-kD&%No}O-WTzl%#zCPleb#omIn3ZpkbNa?zQPx> zz3e<3vBJ$b{nrmg3$3_Wx#{{ZNrD^WW~U8qwor-;ifo`_Q99J5>+(O!*~2|Uo&bCK zSkqSR!c$*m31cA>z{s~2A)Dz#avNNpl;XDYuawJyxpu+m|zxldX@-|4H26h99 z?JPA6QMsr5`PgBKO_cWURHx-YyxhSTvpt zP_d0t%-`Y$^5d_kl55|+FduVe9g+Eh#IQ!`2|=STM&Z!R1Ptpr%KdZJ@RBCw3(nA} z!T3zC@#**R`Qx&F`Q50JIVE!pY8(TTrnvT0_*~FraFC%onYY$|g#g3MVK@t8n59_{1xF{$O5l~Trwtpp$lPd z7u_xGqW?d8?*i9UcCL-P#l1-uhTI5d?*T_8z<@Y%F;u{VPG@?Zc6vFzo`2hEd!+qp zdpc*PU)pnKIxQ-A1w;iE)Bq|VhzQ~hm8+nRjsgxUiZ?J39m|Y@BBH|oeUspjm}oX6 zG}^E8D`ekRFzd;?-t}IdN9&gyo*iCCP_lBr`l0t0HAa6cCZVrwd>Y?5ojnQQ*Qyej ziykf;$>jyZF*lcAEqUJ6vP;>|e74Nes=x~d?c)kyDLX%_B^FB`mQCADw#>?>n;8t} zmyt540>AE40QB1!jNhx!%iBfEqV=Lw>0suS`^tQD2rSlb5vzYA7K$s*|d8g z@v;{ZQ2Bv6?qcZeNXQW_Wr0qL#_F6Jkh(Qe1nHst1|9Hnx{TQ@Xi+7_Buv!o5M)Dz z4qBK#*|wkoH+*GOdEX|!#Ozn=4K+a{T4G0tIx`4b9*hK%8)LuqO-q7wmyOBTpgivq z=1kqF5&r*R<1y>gu?$uG-1R0Q0<%M;f+0#r+68cF`CGw+g~*qm-4CqvzTmq zVT{X1X2#_J#qFcWZYplGA=#ZGZ4&9h2f8WD4b7jt0kRVh=te~&y@Z-K@0wtv4|0vA zNejU6=-AcXm;H8=WDfn4eV#4KX5R|mZ9yB{H+bt*g%AydEC{5{3#G<8HKE|UrvW|uMGbVb+SBJG~-7A8SVfga4Q0X{N{iYtQ;pt^4Yx%|a8pre=KaC|Tsf+xabKcv(Yn zt0|I7#UU@rP4E4$BE6-iOSXxr4qhCN8Fr{(fufetuKDCK)z6w8hhFDp564dbvhQo( zGy$V*!o#=85q@UJbqy-fo5AQT#Who;iHb|&HUwee@OnS&i|X~9&zzMUQ=~?hQ=iZ} zWsE*-YCjvdZw33&ywxl+Np#Fbo=V<4aA?0hv3@5noJPJg>tUb?Sw~)5RZh0Ju*rel z^pLiyT@+VJkz!0%Awfj>>`b|4*W~i(-mtbXL;YNfO2_ufJLoKWy;#!*9V@wNAe^6> z54|(#?0JawBuA}}(WJA-Iq19yXbhj%6$RB;uw}*|Uz%Ss*DY8&9?CWyx*KyS5O_`X z({JWwkjA_b=KtNFnP78z+W9f$oD0L|o*8UzP~0_&T&3c03xo_=xioZANjHXor}D9? zT?EAu&CFIW4DaUC*s!3s_;ON z7LYg?qSmT{EhIU->ZQZhQ)v#xnssCI%(K=| z#}8|lj{3axI}_G|fB(`;IjM2puhGBz~CU$cas83)*F@8 z^iJ+9bD2F5l;)>ZgSHus+9W7}*$JMMp8iMBA*dAI$yEohfUXCqFvgQ@Y@ryLN|2zW zS%o=we5`89#n4U~U!S6TNhJ`#LAiJ?um@vc0KrxbI$GJ`OIg^`H^HBCgvOL?lQ+I; zGq`Krtfq?KAo#Kxp1BiQF{$6JE2?5rhW|j=fLkr3{{OmyMRnVik!QiPTj7V@ftAqs zv6RJ&gPUAe0G=5#O(^6Ays@wE$o{J(m&FTGAZUa6LZ>YEtO)H3>5zORUMKIOFMhKI ze4xQuwf0qZ);YTn@XWg#Z9{fT-Wp*5WiI=^gZb?4;gP&=kDjN9l9+K_C3l$J!4e_N* zC=P_p3#qtXaEjVRN9lxs7Ku)&Ro$P^Me=}r_EN-&i8|$7g-%&Vy2PDyMsTOs7CKj) z5Ks*Swgs_mY!*}r?}=*Vc1`P}bqr>u_KFK)>%_%8cS?57Auq@bLdW8_oo~hKHhz%# zgKvMPhY2$8YtH+N(@eBQ2qF|Qf0YedJ~#A^mzm+jU4fY*IX!V??3B8!;~?jItJbk(=7zU|RwH>N}D(abDAEdRf%=w@|vt~|%DQu?v7 zTd=~M683H9>>g&cI?n2&wmZ-j`QGRqidigwg0l6_bn=3dgw?-)_Nx8&g+8Cj%;iP+?rxRkkBgc$$ki!czuG<3j)1Vq+neTMj5(*oMSUsIS z9dp*2TQgGK0m)g?T$!drnd;s&p$^1x@IPaNAL>l@yWxLMGH_;wWmw2H8^JjTypTKg zzyB`XKfz>FT31DBbKyrhZj?y*4{Gf&69bzm~D5A67cY?E6WDj%D2tJTu&4zPKhuOb^*KD}% z+H>j;L3@8_g30mOq2DE^_~oZu*Fr+@ZYbK@Def{wE`T>4+69%tnnYofcclp8vKxUN z>w`+@m+V(!yMCb*3y0g-F41;ig)nlJV2d!64Ug|7sRR?^2UH6K2i&R%o;e$o=bplJ zO0vWE_~LntB;&cYUpglla9iwc9HC8ejjasD>l*yZ6hgsiCw$3}sdTwlz51m3p)3>L zqB69~XOFj5-LG631wn1&i$bZfEE!3>o^Jm}@LjR$wedJ<_T8{NxEomezt4N<!+uXa&FDLrm=H7ey3|W3MBRS?d+aq}^UeEHg zj3<5u&al4bit&`AM=%Ws3T@%AuSFL-&=?r3-9J(s`o?6P!C9P~kPz zBfZ%+R4yF;YU}gfAG6*MM~GAV-zhes?W40p#BB{x;GxJwttjz7}g7seDck>4&)9dcQPv zp5JD#JV6yGZRq8L*PCWni+1`L@7)9MigzV(C9wXavxE2SowL)Y*K@l^C%f4TuYEE4 zfjQN|9j{{d&I1~+^b+G=G5Fp(cG>tW`0KAh3=y+d$kMNuryBm&7o$IhdcWu9|62CP zyda6Bax?d68^vvnuZ%|{laZSMC$s2>taY&f6G2~LfC7(NQl?`GS=S)smlVP|@S8Cp^(ZaGDkP;pf=Q-o!b1yHJ`ISJi!c2scM{P0F{!45!- zWcXn}JFr?6XIZ25f|w5*6MaqErX87i95{z{Y!CfVxMfx?X^v=# zErgWi)z>ulzm~#m^vQwHLQ`lnH07nUgAeFPE{*3?n4`+fz}s{m#J`~$+0kOBcEHA( z{jhRw*a)%u;En%zr|c^xJ96)sYJ`TtamzaJP*u zy_yth2QWus(IB46rw<19OfQb;i8K`ucC*PG7JC&dPk9@)nKjs(+^kCGumf4gE*9T} zs&szNy5m4)1!oqgtfI%p$@$a$eqtFh=3|RocyY23Uq-s~wak3+Lx{B)nh=nGb}5kg zwL>+P7PudwT@yPYub85s^h%tvs}QU4{{&9Z39Wm5_^%AibCoH=9_ zuz2=FZc*Qu4wGCpb9A@cI6rpIvU8^LDWmdoNXO23hxwHyBa#i`A^oKKU=Te9qOxjH zjc+Tv3YzWpqDQb{z)t6Ox?OZwc!|vo&ym*38b#-PvY;cVLOlT6hr#nHOquuLpsQhR zWG_=7C;`^}-2rQ+oC0z5Bwt5QpqD;r@K4)<;}b6{6cToOf=QgKM7rehNnS})wGV5PUCEtY+sy&dyz-#`w& z%eGyMnpdA->2~lj6)ubw8%|2MU=z38*N_v~#9f<+TaEY`r=tG}fUyHDu;e)y3U&+^ zpIj=Jo2-Q7w&af_mEQ)~g)sqjTtjLwvMCM>N+uO&5Xg}4eR&z9r%MA78_g>0qU=zX zN8=VXoo$px5p&0A`0r-_lW)jQZVfj4_uzVe*tauV&q6LQDv{{?^T%q`-xX{z0l|h}kHoxIGjpLyr7%D#fo?ep+$O zGjZI|qebJIV^2d!b>OwjGSD%rhMMAqAs@rR4(N-8)h>)3 z=(Qe#oy`=NO_7aM+#vxpPt)rl0@@B92(+*2SfnjOW<=fO!;>3BFNPYsUbE;Wq)^%n zjEqKIg=du^@%-ei-9B8DSR2cc=XjxHOl!%X8Wbj!9DDOym&hs?MoEboN;Xj(c&{5k zCr51Pw8yq)tV&zCgwlW*4LHI`#qG9hN0Ebmey|Kk{!1P4O>sX!E_b zNgOUnobAx^0?Am;?QxsKQRfW(Spd<$!^czkYn_;~!d(_Zkx$Cy{A|NhhGoh%Wx+fb+gppZ%u!ZH6+ z8+1DQI5!F?u$V?WL1FQf^AGufLa{dcO-p8lC#xo1IAv#pZ2_~e7bWeY6v+9WW6%}R zs_v08iKfwggA$M(97A$Tw?GN^tr09glh)JF&}i^7ek$mIH2zKzvz)iu#xKS^pyfl zWzZ(B4${gsvr?GH1I#n~tBy@TZ)#0s2~#P$AExPFiJPOwIcHN|NWoE=`@8OS6aK$|zDo#qAKJNFVuh`RrFNRT#_JjzW&9!XqO{ zKfPi~zuS71@pqnVqiA=~x~ctc`srEHbzC)#N7tt2!lrq})CN&&qz=ekTA=!IZ_r9< zLP*-n`sp8$E=RL2*lDqz*d#xB!m(D2m&KyKJvmlr8RmV`V!3eO&PFiT7^*Xw1 z6QPIz(pxquQiWJ<@|gmQN;FN1{HfXDO^QbMPN-jNpyNF>=ecec7obw*!D!R*ZDCqT~W|SJpw}&5n%q+6j{dI(1o>uXngjDMGEZ=rKkURKUf>wh zeeumS%RaCt_qr~;ve@YM8cY#lnVaT>SNiyW!Y&sk*&dKM5eqhlwhfX4KRs}EXy%$)s1kN?+{uK-F2{ftht&(OY9f&IT&rA{mwMCLaR5Z=6Ro(!!7NfS2u?ISq z`&^sz?Z69f);vgmV9HBQ%v$$=^gd^0vEV5USz?w@Tp~plP;uLToAui0tsYh;8hExJ|3F@ol z6MndGUBr{yX1KUYanMqVc1)dnRAiX}-NT9sCP09djI+$ucq3}|+6U`5;M zR{3J)B-yRJ1!E+>a!lUtwJ)UKZLhd9;EwUOfAIVa%FqPnVi;Z!0n0; z2JkmV<%MfdBduMuD)3X!3XkPpXX$GCp-iiOmQm(mL@n2^ksh1F%HpW)m(g-ds!5lP zoY~;O%%Y9)Pt?wA^U*wnwtp1g#47qWpCalb|1L2erIs zv+`MdTEjRT_QVh4z4$!BYxNu>JoFvyER!*6{&7MsDR<$J*;zC5@)5;B`u7kOr!#O* zXyQZageZTg`Iu=7t6{GC+?mm&()ta!9S%y7LN|t7AAy8k*b@r1j|EJ2ct^xx_az=i zm1hj2pwMD2y%dO;)%D&5?8+!Sp)u;v$A=V8&|tA`P1H)^7B!B(!Cx~!_Txz@LSRIZ z=~O46xB6Mu!p1}n8%sP)h>f>?>SN4=D=Y>;8@$zUgu)dJP(UA z3^a4qQCak^sJ5{Efll>`Mgpv-B4dOXCr-T>_OB&cZ}fLilYVDlDrpBw+x>w+cCtl% zAn2s5TaYsa=`pHiZU{F@qO~v^gYanH&Y-LG`j`rj>fkhj!uMBw^byHiSJXpUmefG; zh4m}?h7K4BiWw2Fm!Be&*x(EVPkNG$1<|eGsy#;p?v5p+Y+EvB?V-AIBLR& ze#lRGZ9guaGKdo{9sbtc|9ak&$O|E>vsut6LMqsWGp=0B#rFA!37TM;Hl>N>H7(Yyz`5`vY;F!}* z@BZL>zc+1!K3K8jRdU^hH`R;GBvyMV?k+{TU{ei&vI1rai4Un~ENe;ieeoX#43Da6ysk~a*^PpnPrkAG?I3J zmJN1L1DWJS$yO+<>yw{Qq1v#fHYf#xkk4V;kF{)wgv$Ms%mh8<-(PPgD_j_QppiIa zSDi(1P~wt7#r3?ZpM>FU>b$Xf{aW?wHUdLeXX`LDZe%$BXQ$`DdaLgWVo;ed1Qu+ z&nfO76uAKkk63lnPOhUOut`&A!PXSmk^Z~tNl*pv#+ zM3+E@0!PI^0y_uVMM)khU+JB2Iy&D2sFNq&0p{$M@u#EpG0Qy;_}<`_DmE(gB9zA} zVm=EiQ|nY`Nj{S%NCBFvz8RlR(yH(eu$rmhg*XdyG3MAo2XxIT!YTv}X1wmEw{qvW$vL1O9!CVU*RG zu?G@XA9V7`4@>vmv*d-^X7*Wbq_|9qtfk^IxB{8xiysTbsAo_{SA;^CtXrAwzA~(h zw1q<>*vFC$*awT{2r;2|UpD3#9ez5|7`ZblM$yh@Xk3Yrz_P^r( z_klY&17MroY!d ztXluaIrA(ptd4#4-5<3+?-98s#RAa$$rfH_4+KHA73_Q}z+-HIFl{eM=Jr8~0GkLp zpaubnR`C4jz{NgWB~z}raric-lFbO29&vpliWYAc0%oGk|ip=KA%E9B0c_c4*O_A#^IDdY%&BIc< z!DSa58!`=8_uZ@lw+N+pQ-#RT<{0ep#D_gl%oCXF{yl2h*u%3_gO({3e9B>i04BhYa_;Ga)0B zq*isFTl{jnsEcliJtQoX70jwr?+7WAU6J9ID*@8VUG)6G&mqK-Ei9gZqRDxqVPJ-b z-Vwmk&PREnm)ifIZ_lype05bo6B}KwpfRcLBHQRvdZ~90$SLG|7#e=CDnP65qT?s7 zXRU_A?fgLb-^L3X{D&X*6TF}?=Fz>Me*K50mFD>D(C?B{E?jGH%WMg3r?|@$xj@DJ z{VyL}CwX8RQi+i$5Hc)Kx0DB7^vcLe`mUk^@;gTDuyhvPY^|z5mdT{4JA;r-VO40V zu$o>O)h@!Ht01<69n)u^{CZXB?I6@GJtf&J*b5}H<wev}z%xPb?mD~;I^(>XbX zS?oR_>0|fCBZ1FT z=YND3ZNo>9^z-%4`%<;%-!+wz9=05?pvhIQ|3V8$bkSoA-6%Rm=Z-FY zkNR>}0je#io0M>Y0R103676nqOD9xjk-5wMfhg?9^fO~$V39U<34 z(m?mfPG5>2vWBw^J^7410ONSYVJI-Y6=gyM#rNEwNp~ez!6@ z&ktJ>kjb#d_mS^W=Af*`w#=%8!1KXxmNgzNf{v;SJ%ITGbf*9~72 z8_JuB!X3&~AvE9TPtK=rM|LWpV}Fltnjg|X76TSYE*qw?QGc5dp5ogRiT$WhfxAGAF&ivJJ17G7+u+6FU7pp! z&Y$2RmWR(=4|&Y#FL_zDF{#19Kl|5eS;KPIMrEZWojprZm_k2eDQqvi#yZEZ!riQw&&IZ-belu&4B}!b@e%NSLJSszaDJ*5w z;EQI}DG6l3jS6R&^VZWyPNrjvX7_o<#~Aq4LBtAbWdqhjkk7faxx<}z#DFvBo}VpQ zc;mmuSaK-4EJ|$fCD%;uh`6pe><;~ag<|}_b0^-`mBLBiV`JUG=%c${Ha(!YiIO>?;P5Lm!S)Wwby*#)1Lcz;bn|> z^j9bOnDEm4t3C5bh703muNhtnC=Sw#o2j^)U%o8+`&G8s5WK%l7pMmU7KVJTSQ_;S zeMdeJP>u;8(jgrPSPAkDN6EDr0};r!&@TGop3Di?W;oH69T`^anf2k_!wV}^_b*F} zEo-e^cCoQh1a4&cHHf~H`dsud*pcI8djv8l0Dl7NVWt9`oQ?(I_W^K{sj49( zgUwvnWx76Qk*B5=vd4I0sduhyktejK#%K<^>*Xj-9Ebucc9U)1}Et@%>t7;S*lq>C`J<1BEo z#0v_Lfco~}j4ba%bFPz$DR-u~#HKObf)3G+*&BTztEb2j{Iz0=lW<}_%fms9`3TNA zvW=P>Z@I6sEJArf79Jbbl>=@~iXDQB5-1vw-wf-cx6Gn13P4YQK`wfo;mnX~cB=`6m=^DB?W#=_0A5`oTr zm!wfy!t^VzD|$e%U8`F3wN`GEd{nH;aJpcf0f;BY{Glv#-xQNSlJM@tB$73bRG2OM zJ17psy$T>|1#~0_Vd=-FyOZ!+t3tXO!%j0-oloD1gw758f$SuDx{lma){#ZA18#XB zcYx)k`E+F{(rY!+CQg8Xe zyWn2|tubu`pn>vZTf-Zb#uiuXc85eAh(tkW`{-qSd0_)g#%RHV7dFPUNT!dsvqlw3 zq>SxS@Lfdu-EKz^Jn!}+!SjYFfq^t#qkD(CM6gAT+C2uk>NLMXX$O6jzuPr*WwdD-OeL|5&B!`7wnH$pwBH$PH)szuK_7;oG zI_b^8<2m4FtS);%Cx>T;YgNdj)}gMNqS-Lv#`JtT$9oH~gJO_gtE!rk46kebG5xY( z!b7R1l1^c+h8=;0D-TqoaC?sWVlChS!yf()N(HTnNt<$K#`#&W z#_b!wZ#=X&!<|bNXF}}{v<=r94q8RtzwNY~{DAeabmtEznt-LB6?U1de_W;)du{H?*pquO*L*@u<*$Ee3x5-v~|8vV$EgRiW*Jf4)zndk^4u3Ck+@A}m zl(+tT=@)xuFQo8%g7Hufho1;*gS(Ztyb`Ew)SnlPGxkO$j{8)dz#J46MV@1g=d`NK zz}s|DBo5dTi^>Y=>>5FLBvQU^m6ua#|JEH@DjoTr#StI8l`l50582wyyuQ?~8|9{B zu0*R%*5+8;-&T{IFO0QmG_y9f6jw!&gGNDTBQfj(ae)jNnr^wLvuJ|A@}!XK5y%6S zKNS`0u=@tZIF56rpgj~%{iS$LM(_%OX0>-0X;h>zw|vTQe=e<>vqh#ShKI3(w3Jy8 zn9ioN_n;(uz)c6_2G!8*&TlERhPdGpU0Ol^lc%h|p`B+Vn-6Au$poU-uiwZcyIj~8 zI%NinI*O~INF^1wL$ESx@wkUEofA`-Dqt30G7dV6Jq~!UXWN3aKs7lfwAp*TcXL=C zS08*=T_?+>o5OT7T4LKp@oZ~ob69F{O6WaB1-D7Ph0X%fi(b-2UvMvA_fESFU8Bum z9nf-&4URgX>RBMvF^k8660m4Pcnz~;+=#8DlZ@nux9g|Fvz>Mgm?p{RAWekyD|7RM=9U42`DXHov`m)H zHc2{^`9ZZo>(m|4&)yP?&ePe518V26R5}D-b{Z2txAl=f?D~<(@C-+G-gVbyTe9=} ze^CAn{cNqOl}n9o<;tlHa(P;#a(BR7l(?m9 zvGeIZuXfQ+<`UDPERRkJEfOA5x5QRWNC_VLjxDiSfw+CdrvSoJ@IBJmrR-i{@0|NA zGv^SLJ-M_!`^+;SH~*K>(y1ntz5jvkAlb^#M{;45*$`wc`{rd?b8H%0;ZX!>_|vg9 z!W=;-J@-6QOmA>sGqG4uE1KI9`!P_nc7J2=&i#fvbs$)_W@4QHrhA!kYY@u!uJD`N zE}GjLnJAolo#e=F_-2Oy(T9I0dyPHisM-j}L+2Tou=8@*`}@{g5?AtB(p)!5Z;O8u zd3md#Y3R1FSbWg~=?eENmnwj;H=oY%?@%A1ar>U;heEB!Tk>D7nyQbe^_#EK?1vH$ zr0I9g7T!2zBSVUv(>{#b4y{`?L2NQI@0N}KACfeVKrh!2iqj1g2ZaS|sJJ##>yKsEQWl8`36L@LGVsn~z^xj>BAWQ|&$7G2_-mhhWH_JelcPBP3@;bd zb-!-c*Cb4zWr9cZj}vl9Ilnc?b^m4LtQpEbqBvNT4pDIn#I;_%sK5H*RkjrR(hNP& zrA#Zg_sg&cU(3uFCv(Tu$JM|P-WXag)laI9N)=Yk(G-f~#pp5D%qo@uzY{*uK?KKhRD zR(ci52)^Un@3tq}c+@4t<5NI(nHRD-q{3rwOdop&9O&zcwCMVYiE!t#$t}vmWM51T z#Hr`IcgY3P1yk&N3;%3rL@!in%`G?gOXc zh(C19o|tjwz3*D?*KKTc+eIrQyCDve&0JEWphAPRn(q3ybxeiF)_`2Pf$oH@?Pm(T zJi}kF?2u?-UHepBA=wQ)DVrmkV@svQk~LEs=+h8wS;CY>b_+U(^L;e?4Je_e40ia#;~d+2UHJ3)5NTipc`08x3y3(%G9KjY>l`M^{vb+IVxXXQL>EaW;|R2w-C8 zgb(jFelVGKY{RFt2_`+?_d8A&@NcKL6UT~%hv@jUM}%w9-YUzhlt<%hDpzk2WgDov2edif9UkaT{Ka$zcw zGBZf!QrsqrWKnVG_aIluImyY14`l|H5Up3yD;f;;rL*Y4d^9;p_!n$AJxs9N5QBnR z71sAx&HM`%7XD=zAwYX)Wn?(9_g(nk7ZEo<@h2vnXm5%ik<%{hwsf1#yq@CPDRP;L zYmPoLZ8u2(9(QnCL>;mCGleM#!1Ks!z9hT_?pZy4yT>Vde-z}%qYI@ep=o|c=zTH3w&zw9UgCkjmjzt( zXo=lQe;m*penM3|;fm~TSVwH)xD~!23pn-?$##D@wVuKE=nU(cu|V7*IVx%weLC?f zP@N<&cc%5bHN&2|-K$CPEYOV%IeaXP#l6n}6EDAq>fhaa>enWmC2m|Wg-f;1ZnO;mVbd2$xLkgSLXt|Cu^x@n>KnxK-# z%Vm-2EEdF76B;xPx2X5{>R2qtD3sPKw{x(_!Cw-Y42r4oF&K|c zo^TDu)~W{FDnrxRj<75#Dq%Dmj9#N+zBpH0uRIZwKHl7S|ED3&0)U3YAt>a9xG@ng zd(U~7xuJ4>|#g~q=@1QC{U4$>mj#bpN85gnta&B<3 z$6f%WFENCbp{CF`L2-*-7*kCbPdH4j#dbr02bM~CpJzUORkP}1$lhVelS?v$;xK92kpux9Mv{?oc8PD8BIdB5_wdpwYjh`{tPY!&LEif4rw?MZAB=LnYD6r5B zF(~O2w~8Vws5oT0FBBs&bw|X;pyghJn>^2!zo^|G_Me*N0>1V=4 z(ZG-QkZgXqaN(^`r5P?tC=Ljk3XMI07!X7r72G3VSF}s&q2GLEWFvh=)Zshec3rU_ z`rYr58`BICLSs4LfEzlwC}!LN!YlpC6k$pzv>&qvXIxj*%yI^t?1s$XN@DG|6R^|r z&Z{S1GePE|Am?YKk{@JTc$K+e2APu-2kq2HskkoT5=QIS99m8H%Dd>!IhSR{pmvOc z30iTEXQq5JtUzbH?@Ve7n=gbmE!lwp(2azpaA$uco+EiX>BU$S{F{3k`<5Dm;24m(6Tc zd_=B7Mrc?c$#DMM!DsCKI0x?^t>3)1q^`SB(CS{>V9D&pr&`N(|7FAm58PFsKCdgV z@>GVRkQ3~AFtynYVHb2*knCb~A#N|e;oCo2&^El!n_v2!;CUyxU-~VBNpFe0tL$d$ z1lb|`CfyONB1@Q-*hkRQsi*Jwr-YWOiv&rKXu3c!GubG*<+DrKE2~x}PA!VW=AWL( zN7GR5FM(@unBcJaO!9!hkTt;UrTKskI^t#Gs6FJ3Us}q|+R$k&k{_dQGh2bas0-Xb zttt_+bd91)QJ)-{PkvwUl@(tp|8oEL_RRkM>DVrM)oZZEgIM?|d@qML6At>gpE3-; znQ&RQw++Fo-Wv(=k_-E^Hi(yQ``+i8XTDoT7f;(Id(U`eDDD$omz9A6y759DJs)TV zu%)VbCit~MgIAn||KWjti1~R6a7S_}X#NK^BiS^&uxnyt)BG7|WOORf z6S+#?^D@-z-ICv&p;cXDYlWGj8{th%uJ?`@4VF!5jb+8yN2ke|(k?25uDb!ZyRuG6 zJKZWjHK|vwshRx97sbIG@3oH@GK1(D2|lbm`}#`nu4$${&YZS_BvLre3Z1bF1LLUK zhHO8@LCe7|7S}CfMH%PNNjbrSL z!yqyU6$kmhwwKb>SH}PMS54sgb1w5n#69v;!L*94c51`M-}t@1^&t@3+H;8ecl zj+|ipjFS^k^##5xUs*9Qwochb)mxwcYSzD2jxn8Mb+Q80dhyE0TpuWk%Ju0M zRFPKsNpggN(kY;a10do-Cq1v3S>#tNsr5%OAOkHs#(?`gAArtC3!rC_5GWT^c~d&4 ztQ1~!FpI7u@nj1d!#VHavDN=yD=bqeE&?h*V-)7LM-f#-VMQ<+8rXz~7g3H`mF%7i z`+;HikPf-A_N%pGSX(U4i+U1oKc1>=ly9jV9V^K`2JU!$Z zJK)v~gp-}nr>kRE`yx{^IzcszR(>J`pEmg3dO8&v9H2sd>H_ipS78A|KO3oEvgi$} zKF|0VIAyE=tye;`Upm_Y?wOH<2Wuy4n08SC(@Q`=0R1+s{VM}9*Is!=a0RGr4!G@5 zo_uxit*h1hBebv_cC%1a;j+EG+va_E#nLRfY`>}~ z`X6o~RX$zx5vC7#Hn1VQKwL*sV#HBuS%GA;SYtgZ{2V7N%znayRB16NqWy z+Ub^kf-HVv7Oq^geb-uwOQXn2EP*gITpFc?(AMD60i+S+CTJ%kix}*fG!EKrKkL)S z>pYQR&yWter{)h^*IPz~U3O8i5$E0M(+}<1i~TZ1y`DQ~W^hO7T)!KjfO}n00QHsy z{;dHh)UYG^M2J>Z6x70jz73;O76_`9ZIY|7yQvIK@~sa>l?Lbexjg58&^iqMcF0^? z8-&+NLuFOW{lW9z=C5PX8BFv&F)0(GxcO7#L+XT!XWS8*T5wzCnTpR8#* zl<12$W(8zAiqF&9Oc>4$8zweKbj$@_Fd0)ed;Y?awBNe0Lt>--c0=@a`k3F95TqJT zk=FUv_~98HyGMyc!+87;<&EjJu=8k{dKI|XYW(&nGr%~VlkX&FCOXgT7-!#AQKi~j zs`E6tCFNiF<|(q=TF}>p(Ey!iLrS>TQ`|ZVKI^z#K~ZESWBiFsdbTK{z*o2 zZiI4Sbl6~@&W0wTX76qRB%8xG{=7}n$M%Nl*h{kJP-J6(TF4po!o4vN*Y{`_J&5Rl zYl+<<$Y!cVg|FNc9)`^115)8J|7GpVS`|L! zKxFZR&Y;g?8J475xa1pZl{^ac}!k(#b>@aB^0A$S<+Mj46+zf zGeuX7{8hc4$=v$zK0zi>e}Wj*!jLYyBMg!O@}`MSYj-ux^#}muIJ8(9=~QKXd5CHK zx&GG8out5pyIzl&O?)}U?V(5+vSk-A$JAPNrhmP>-!1>8(?D(*AF|%BNrcBYUt01? zz5J8O{cc+#b_&k1B|*Ud%H~h4R@ZxPmX!$b_ak8@u(}p7he@aM`iy?JdqHK&ez#)b z15oQY7<_$3ljxSa1FZ?)t*{L4awmGp9bfegwhq?b+OpE78ujfJr*Q$&; z*P%I}9Vj=Ki}y+^Jn$#<*iXl`Dv?x2^|8>d#^7C8(2of1R_dcZ@;?S00ggnH&YqujW@4VK zNu)vPOr+?*kbaSTK9wm4Vsa>EsPW$xWN3xP!wL^rkRYj3A=9cbwS(shKr{g=zmc}8 z*Ax3)QTI@*0(C@bQDm7iUJQeyqI{ZPG6xq!O`5KUXdEOt+&JVhc@Dv0_xK`_(Ia6B zze1}KA?4{7$;wshsIYdHmkk;_FY@do%l(tf2JdX_m3ly7C7U@VX;c8Kf-*yeT;=iN z15sL48(^(s(w?YG0X4$@DGj5T8sdo#YZK{|bG*<&ef66SouQFU(nx$ zi>>aM*{Ew22S)1(71!v#S=KJ-BATA*<8}~|Sg$L})azdRe75G=%v+L+lKoLTq{&iDUT>dJ25T84u#r9OC|S#Gf2BcmPO>+k zPFMqa^`_I?y-Q{exZ%C|^hHVfcx<3doN!;2B)s{C?z`G=t^IQI?1o>j4$V_nkp<$l z$WU}CU}aRiST8RopGu%)tAK&4h9s4tzaNb3%LxiwRarb?X~htY_>@&x|MuakyZ>60 zA2+=hi*JJF*SsPMYxA&47D%VW`JR2UWGI)JS13l+%^Ls8&~`zYgxzA}NK(RW*4mv41-<{`evPl14v+*oh?ZSRgshN%0OmV>Qx{-=&7v1r%Ve$mK zR45eJ65Fa+7;?-&(9p|O2KUXl=Up!xyyF&hzHU-)|9;KvP4tRUNI?&C-W@c%wuj-j zZv8mtFKhm_%n342_jz?d@TQx+BX1PxKqeeW>w-2xsuIs3j~-N8bt~70!xDmpGFV-X z9QIg?h2`hY+*OTe!)_ZCBfpaPiMC;!$(+RM-_?-$(A+yjdALdFq9ak9gah*b|NiMS#l@2Z}@tC(jz znXk?_(52#yDNHuB;3FAKp%`e!A&!GLR|T)|J1Hv(hrLZTy>Ke-XwXDH^sNwHn6`7K zG0>9edr79*EJ*TBVXEc-a4(ju@M`kHs$XnoLDJW5cJP62AlQUO>p01of@Q*af*a&m zR40uOV9*GE54fR@#>&VCG>+Q}JyuyzqNAtvfg1g!Ub5I5DFW~x0Wx*h862e$S{m>t zR`5QQyKJehZBX_t?EKpd699kx$;|Df)P-H(2D8PqhT@=g@&FZAF3W+yUZ%X1&86pu zA7u_J^`Z`CgYVI(WiJg#&Pq0u8>GRvDDv9OyNWHdbjot-W=M{wPFczpOuh!)DM`M- z6AZe!HIpkP6(0SUM6lZFie zZ?}9eh5GEPTWRB~_Bp6?2pMxWZAwdQlJEY&tzP*0nxPmFvJzhS^id_KIL9=G=TGg4 z?2$YKE$AygspKs4xnx%W{>HJPnEKiY3_o znJy@nbopF>CIq~)A{tvF@}}$z`E&l{YLQO)P=s#O~`mNwA(8vE*0~;}0Vhf?mw$*2$SjQ!E&8n8z^?qISX+^)XPH;(%Z7hAV z5}%e>d}U-#L!Pw7+yhobkSPAm&d8V*Uh0ljz#aPWIr`b%OK3r{wj)Spij~SpkYxQar8=Zi&6^*+XBJ z6)<^FYu3rO%j4m_jX?Y7V(Lkh!5IV4_?f!S_mjT;>>3~!_SbCG0HufP>D8h2GW2`? zV))-q>A}F=l4Fw(f!mbMuJXRCJgxW~`k!#`uR#WuJN`)PhzzhO-Fr@r{Z3odyGhOD z!x7tG83@uGrPs%#k}P-z<4-wLFd{wg!FO_Ba|AK0bKqp#^w@jFHk{r(G&R!Qa%MybZ#GZM@P7nSvJ}PpY}t* z-pO;!eYPQB#d#WgN&RQ$d56Un7N|KHpa_IKg9tRf-5-UOGDQ^fmi%o6`xHmRF(0`3 zC@%&q{^MM`(2}6hWnCQ`%#phUdG33m|9`+uPj68diZxfmT2ux8dBfJGG@oc*NA{wE z*Psr#nV+%v>*n=gmhOkkqQZu&fn2It^mY&IhAxz1my=OHPV2XGYA+cGOA#W0JZ81E zs!t|22{a{w3s9flGd*WYI@|2s<)1z2q-+_ZLC-_W)(NvfS_SX^Op!k|B@{=&F&t=g z?1he1AM=Tm4#13Q2wu<`vwYRA=_6qycVVB_1{*nYh~Su3VrzD#N#%}yIUXAoJ zrOvr2swZdJeiS|VtTWFXDE`_s)nL-<_}d3bjtje`Rc3B!Da8S&!<6#y0+G8yMph+5-v4PPToc@q)ydkAL#!|32@1+ZftFi_j_`Y{?&RYgCp25mhgH zQKDHN-WCcnw~)2RR8_a&n0&wuS<$+B4IW%BJoi-n~yKeTW1dEdng zEn~Y_QGha^?jgq5nPK@s-CGQzWzj91CRLmZBoOH!j-i7}zDHqq!;BK| zcoCFPDMv)jY)k-6%Tze;cfkH<47&BxnS(l!y84Jfu*xI`qYvdVYWDD zgq@lrLXx?uvOF3QoX-4y<&Qa&&*J%@=R8^a!c?E{HuG@uDGtc4a;Uhi^xRwWbV0w{ z&gfqG>FBfc+$?%iY3b7HN=W7)rmx z3S`bQJR?WShyY>p;R}<#{lAu>x96(Q+eSEU$+(qlzuR5Ua{8bog((73Iz9L<9|e#4 zRv12cJLlf!969TPZ8zra1r702O!gz;-HAygYaFRC`{M1OxFU)aP;m!>5+`1ibb4ue zBd>Wb1;OXdRE4T661$(VED0&VH1Ai^o8Q9y>e{)Q9@4MGm&?8C)%hL=K#J2qAh-jv zvDoZU#^8mD2nTOP?VbcnGyKG{qbI}mBv?zJxU9fI>s#-yFGpx(MOc|U}R-~AXbYPs#{{0jn}bya|Ox;PJpS3iw;-xeoYAq6U(K{wR3^wXlw6S+0Q>NMKXN)I#%)iYsoN>XnfeC&i zu1sWFh31@Hdx$LKw?TGcj-st*YgQ)3t))mBYGL$wX-0A!?Bd@%*6WO zeER%Ut?ER`ChlR(5(mo+JZ)Nez{ao;J#8#UUfcK^`7%oZz!xMAchC$R#T2)VB3r4r zEU2Gd!7h1o;A>yJYL3O@RBFYX`0uRw@@Wv~%Azx4s)H7YOQUks*Fl@fcrP;R9R;I< zBh<-^znx@$_yXn$_r*y>;-X1Q#(0>l$@`i!e>G~_bCCU62 z8W;9j!E_8Mo83rpnG{(|#pQTy_Nu1ae6W6LgMT>?{;rJ5oNzf*$F}*@OfHVdqpvF( z)Qv#>jCVQ7QamwXe2_Dmi_N9xpMU%z|pjX#@XaiwVRZ4{jn7Kda8#)}^b zHw1NsH9>(JhW&QQQ0Vc9w2RIbuMt#u=13n2uZS9abEL-<)x>#d;64B5YsV9Ka?}ai zmcvT;kMgVA^B@0Ob!|Kp^4kwBdlr%^Bq z+d)lgj?K(o!RLl5**>o;Lz?iB3=!9+EGZU6)dz2rL7I4h7-?P_6)8-otCeU8&r z!sduubW3b~aISizPc~$dle{y_-$}p*scKQ*pLm*DvDb{kt8baa==N?1>ypkrq{D6bb%On3QwvZ%8))|)W9$} zBldz~@cz;I?X`Lr^Ogw|`+gemGjiO8lYBSKrn!~kS}1ZFMXd1q-AizFxD=J@elQqo z?(>ZOz}dpK&$)WB1c{B9o)m#|*d~@;&8=N#9ObhcxG) zH#j~9$*-D0C=qjfT@YN+=pvSTHLH3FviaiTgp|#|9$zRPJlNxV3+~1Cp?tbhP#xUy zY9?^8bO0k6j*%+d>2riG@v(O$vKKFp*O8IGunThhT$$`QKb~e8ynn9JIyOS|TY;Sb zI12)rymNpXz7v>$aL>FDI59x0T!MwNz&IyzGT)r}eDMQ}`LoVEvJIHm@3I-MnoLjm z)c;vbwpokCy0BC7k(oI=&+*?EHQ$daf>!o7m^6IKehz=rCesFK!CJOhO* z+3L&*`ibj+GiX~>hU!3+<5n?#=y|5CaTuU%qvuEW|EI_@_2aS|X&cF$a=IYkM92|f zgFi}FhV}_k00ULBEdR2XFv+K%rd6#9Et}kd$`XnGIJ6y9e63Ody zY$ct_frfohfjG?%%9v0ox?Pk)u1DYy$hn;A-YU@;`U?l#lDwN>Fo^AJj{waIgC;0c zgjPyWAhawpI~*xO>xkhAoCfLoJdfw)bU11qHQ%u@h5a*#dK zDm~=}+nD49vayNM_~TzTfh=+LcYjQ(UYI@BMKee?P~689`3NZ$kxUU;cJ3+l@^*;E zXw`k7D6H8g-o)KkX83ogu`lf&R1GJGHz_VL_vJwMEkd3`VCHC2e5}fxP(|QbsI;q@ z*iN^yh^Zw2o-48<8qtX^eJfk4JfO}{-&H~z?Mi7vNE7fop(te`l;@QO?oevBdYwX{ z^^HEQ^6bdf{Flbj!k=xVS-)~?X>CxT2~kI0TUAcBxUku(HN(g*iUW4=Vk&Nhbg$ql za~250>g6kiyO=Iv*Q{(+hvXDp>T}j-fjEQ1YkLJL!CS(wGPm4YVs-3hI#6meJ}fIN`i8MVQvSx?Dv$F?kNy(6!wO-(mmzxTK(x; zv!8b(l=kLj*@ss_a;tCJIws5iIH`l3RTh1l-lc4Sz11gzY-zoqT6CK|#^y&?(~Y7+ zK>`DNGeN$GR&{Vj8B}g&GWEgjqDIjzpM%kRJn971bo0ymfmAS=8{Tx={=^*y1lw;h ztSkJCQ1IgWzRx=$1I7#?oZ}6R^cd-Drwvqdj-GEjCc{2R*fEfUZ{Y=rv42bG7;C8q zxYx{}7f>7&xoyUNHLbcTc%MwG+6Xi$8Z0QlO~-(n(dEQ~w+$1Lr_}}-Z!++`Z;q^1 zCQZnn`p~md(C2wb@BqTE&T16pRZVTxB0G2{9uLQFN?f*6?JK>?ZDUP7Ov2)jdNSY! z(lSFcJu4_Ki6ToO1pqn#9k^RX?LGXmYE}P)uc&>9csxy`_~HdYM8dJ@ok9~Lre=Ka zC|ShMt99XK0BDXMlEPd=ajPkkin%ScdGoL~0x3DFW@?ODH+UB78xlC@nKvu=v~LkI z89OV_JL}Yq)r(L2nt<`MhB2o}rVHnVTD*C@eUG$mb=dB}Pd~8rFj#fBcbXm)iV8piwn7Mn!36;w$_L=N zk8DL64!X^;Zu4kFUjlyUzT|%ThXE!Ky|bfk1=-BceRbgqyc#pel~G&?MYdCMeKBbQ z4bnMjRp+^G7TeUWD^M$>jnymj+}p!;O&)MVzB$b~NhyeN<Ugpwtj; z{dC5vsq@{tWUI%Qx*rLy5avu-%e0HG_;iTk#hK#fh~f$J-M4`4!29YWJCs=YuqS#q zNrPG$teO~^H5vNpn!O%51k8qW*a0syL;YSc{*TYUc0>MT)NX*C4K7$|0<1fLl1|(L z-yUs8mj^5yHs$sV#0wT<-}&X&qsN<8pX>kU;)i4@ztzWuEAe0hJ;YVdpg5=mTTR6s ze&e$2gRAVh*>}N!T!NCk%)lPHLb_RiwVAhkkZ&?xTr)cUxp6*uAPF26dobXbdurd@ z_dA&h8_K`G-b_}!Fxc2`hK($W+dz>F$S(LUe)$aP012^1QK#fK{aN6Fs6{jI3|1d( zp_4d#eN5eDQAp{bqVW6z|*=yr2e9E`o?4Afkqg zqTppTAS#MCkcj9ogMi4O!hfwKIwTU!hJ+KRr}Hc9?Ckvp_xohM>s{+v&l=P93x71Z z>hG(d6*GSrcSZGVXeK9swFmP}C=AvCd^S6|AeJV_dR$lK2!VhQ7tu3Mw*MDCM zlgNG@bgh>qaJF#@ytYj5fyn(6h_bAeUKXzAoFk1;A%eQrsOeNccW-DeeI7If&qL8F zkZEOm{Ot@ZYaldweB|X281K9L&aHduX7oDVyH1vp*4KrUhujgzN2wA-m*^s(yrNRA*0pJR7yX#-4Zciw*+U1LEi6Vics4`B z^MNcIzT5vieBIoZya*V_wS~lj#keWg%GSsYKbxbf>2(A{Wcgf;i6ObZs!l0s2Wn>z!ZYhCfl&16jMQwgH$vo__6<2 zNmyT4pDScyu|N-38EjB@R8i`!2J#v9Z)%XH`K5y{57ekCd&4@QjQy1JzR&?lZ&(X< zrVG3yA9kt`--tS-M2?*XF1rUr%=`hFK6eR7@@ zhN^Cf+o7*uxukZcT5+D>vftwg;%z7#iv`39+}_~B$|vrsZjiQ6?G9N%cF_sE5-1ox z;F;~Yl5^c<8UAnl6J_%$GR)YxCyLz^IV^kWyZhd`Vy-&)S~)Eiln9qAPr81rsFrNu zoEK?=8WhLrYBJ>1Av-S`avCB7%#hPDCIfy})5YQGex>4SqT`PrtX80EEO@qJ*l}NC z8P75Nm;PPQlH4CvCWp{ix=koUpr(LvV?5qip10a&?{jqIk=9vh)WybWBamruo*vGDR? z$NjEl`|@vv0@oiHQ6lpC-6W!Y%~0~G2}*8J3@E5vrJ~y)nw9{i=5c@tjIkZ(>Le|a zW(BnJZsIgZ5gtHtgQ6aY)8dXOISlj&1s6OGKh=sH`l(lZ$aP^GAKBX27gUq*74uYS(ljY%AR9o%A3qHcSLwft!P`^iIU!vjSD=}B%xypd zaX;w1846pvn$R2K9xjg6AZ?Ls7uA7uOFL-mzKZfNW{8cnGGjY_DGV(`Y|-yTKQ!mG zIIM=)ZWB1AQp`q*Btk3ha zfzh_wKR8bIN{y!H^t<1_K{kGATpOU}k65~PQA|2Tc2Lp9oE9dRZszEB{@|a8zz%df zqnzr|1vw!#K+?xzL`<<~DLdz@b~x{FZ&Ta@(Wh4#h4D`$3rLw7kMV!>OGlv2He54T zGGLQbbYTC+f*?bU_fxO??(5xP4Z--*0BLZq1ny*yGz|)VaJBL)6!yYuRY_M*Z4^R{ z2WY^xx$K^O#w*Tim5<86qsyJKQffmV*K~@;j1?ntOh>cbtE9Z9E#{el*DC&EA^xuk z+P5%e;i?7u7HWdT)P5_Er>=g zJ@x_+j@vN*XGI=HsIyk>%``9S8Pc zVOKL^*-56D1d6PuqEopVMNUXAL}zz%K`gd1XtSG^PVupO^zB8%U?qZHywT2QED|aCN8REUi8^4j6xw`*cPNqHkV~hzl_^&(P&G+cEx4hq;|=hJo$gBW zII8nN+t=~cZq)&hQ8$F@wA?FD4TS6k!>D>?5VTx1bP1#)lbm~K4YM;mADZr}+Gvyv z!N1{%e=)bvBZIe_dq7s`W7QtoO2CZtLu_#TYP7L4Iajva`Ko!fki)J^7Ak{EXibn_ zi3?TL@kPF%>3>9R&MHW}dIoKN1v^Oag>S7jkMlb$BrL@AQFq)RrJqe>0`Lxm&q1)F zT@I?kz^};@8%R<6K`|Ch#jq2`XNR0Hp4KAic1Ge9#BgH-MO;u1mYsW}m~+y3Sl_<+ zl6%!G)HFnk(GKp7zPDztK4q|OUdk+eue$>rNN#&Q2s7M)p**aE(M$6q#R^*zwaXJg zd!d(G>fJZHm)jyo38PZ)Uiqv1BᏹUA=5U&6wD~L_{5tH+4b0NtuNvO_(9%kN> z{v~VPN&QaZPdmT0_p4jJw(aZx13nJSp*8q<#Z=qn?OUh6Cos?aSV*v}01<)ar~@3N zM0a~?`Fq^z{RhPG+iy)|Jo`Qa%Z9gN1zREq#gO$m4hlQ<;4tSve=6{D zJmC}`+``Q8DN^iN5{6rDTt6!P%S5Wa*)^R0H*_=WJTHeq{(2822T&0VPh%!~TL4}O zgX&;aMEx;zuXE?_2uT*g3woY07 zjd)>=vBe3?-xJH6o%`y>XAWEqV4?9@mhzxiich{vlLGf$n}YI1$e04*go6u# zV?1LWCkwV><=n)%B|ri%A=ibdu7H(lu)PY6XYPg~>%rhgm4w!K>yc+r_|6xgz;inA z6XF?s%c>od<*6C*j?K|CV(jr8T1L;C^czjDduUHfN|cwTm&jFheof*SY|OgC?@*vt znCyX-i6^}e1gf;M5@8i*TYxH6elX&k7|lo~cWpSxx!c#184n=FGvTkg*D@eRiT_Ui z+vkJtnR~o1u44{tVl4QzxNpK;Q%TriKiq@c&Jeun`J29P1xieBn4njdQS$Fp=+sx2 z064H`W1$3~fv<5nqQtNp?q_phf8ESLv!8syVdXyXS)OfxH@Kq*(l@ijyCM`MChcoO z^pkA`L`HK!><$#m5P9m#Wz2P395xaIl@FT3b%0g0gJP%YW#vevgln(Xn2r1@pqyG#5SK>Rt zabt&#sfn>Qh}gmA4}Gq8e{0Mwbw(u4A)T)oqkHw_kv8XZis`3FFBRP-Ju|&ra&&=e zrT#$6UxpfSSoXYzztIm8^2ZFD=rtoZaY;Do$dW=y-LG(`nSv5K~(wb zq9@{d=y9iZ8-$&Bjdz~!uv25iAdv9}xoT#Q8>%UP_NF#SH(%dRDQ?ah<*-vKdlUh-L`9zZKQ>Rf&$~lV6;LQF32P<4vK1ap(K-HCmht%G=7k&0 z1AbqU5U_=~C2p>bYFM92Q+z7u<$`=sHy@h?SN^Yj;WJD4!Te=ye4|ZTa(T;fvd)3? z-Fr-Yo-~TtLXi|II!}I6yjodBazK{9IclqPlfE2SH_Yp7g0LzKXY4_7O*mGL29uUGTH^3sop3PYpq^2hX+5Z%mSCya|0 zd#1{FD7u8720yf0g)-YodBGlA`He#tCs*w7*=^Y0hxD?^faRdcb!t!8E_ugzyIkY*;vqKpB#Y(Y4t>)9c9^XH{L z=0Q0&KFI4AuIrU2a9Si7)>jp|bSQWG7f+Z9kjZ>ngCCO{Cg?@Z;`J-dy$?2sa9m?Z zEchyjh$FsF#IXV>HaqB5FHYh?l_}&Cb^LVCk3-Lk;=H=2cLh(BVZnK}1{@=n!U=-L zq)EQN-m&y`r}}3rw+rr1)3|SsyzSXJf8}(I%TAw%!D>ZG*r4kJ7uCKwkEGQBt3zT0 zIsW?g$}N%_dO!&Jl*g49B=!aY%ZVR>0sBVdSgk6P-;e$2L7CA>O`q}G-;-1a-occZ zc+c4svx_3>RCJnO$zqMDkXM-7BMc{5G$Sy(r`%HfuYb&@o}Ba%cwu~N1R!g1C7 z1D?II)l)LLAM^428oTdm#*7m?n+jGqq5e(d_VH{ZPWpc$T}5`iX3@856YLzI7?27t zprTiW91A!aunJiHo%5F=?ZCh&xJGj4W-GNcbf-HQk?H!GqD_)2@0FeR!fOq$k&y!R zM~Y=DcyS(T#j22Ac^kih4>}5wP5M^jw$2ib8Au~dkN()U$g&J1;a@_3Zyq+Y5N}Ei z83^uiMTUDR(BJD854vW1Zv0`|cdOp+o_^kbCBGJ)gw>DQ!u|F-ZTV4;7JzVALAjvL z+~6ffW3u3b3vEO*nlW))o*!v3F(w~V3xxSLnK z=#J|}X*DmFln5{KR`@kXrFnGDYvcd*w*13~vJ80~sRQoz7HAx_jbw@{s10Goq645o ze-H{vRtPGn3Na*-crk%b?kBMC;WFtmi%iv>YS5-1O-(4mez5D00bsO^dq0ei)&6@fj_NW52mbV0rVTNtfz zfsv5Ed5u%U(X&6VLDUjk>!nH70K=++Ug4*&Fu(xiZO^_1n?uU{)z%faOa$o6;b{#B z`Jpuh^fSNb`#xt#tONVUAS5;-y1bEM5-G9)!D7@8o0?Ft4Ff!6_#zwC=2W4#@{defzSwwwcBs~n#4OWDE69wVkgC}<@| zqUWk{xwCMqmsWN#FiY7kUlo!8QCvK)SN=tg_ryeM*+x^~8*N)0SQ?ELJ|<25%Ju)x zT-?%OWBC^3Di3%*GaD_E>%tTt)iM7DDQduBbKNc=x@3sz#V2@R?`uOe;Z*^L#2efV z&A|(xXLqkG%Xgz-&>L?m;2@zvwb@tabwk|kxkGVIpsMEeNqS}dVYahC)>;aFNVdTRdv~Kq>8HWR{I3bG*-HS|;;dK=3yUMz z=M8qS@c!LTS8|LC&hJn4*OL|O99##E^n<9&2%lvW#UxQA9@S9ue6czjUCES)%Aj&? z=lpo)IM?1$-O*;-N(_y5t6k5X`&re~*WC@#xKs)+g0ry)oDJNQ=n2?y0fs_o6;>e| z)|z1_?EElME~}L@ZnAdRb*hbnv$Gmk&Sv_WllC3f4Y5GbZ;`AP+y{msb_2oME!BB$ zyP`aD11}MzILf>Wg?jZ;v=q(IHl`yufrHmifVVmph)jgH$_gTr4h`mf?RAI99=D2! zQs2VGAAEE}+#$>1_J%=F%q@X)iCp~Njl~I|+>gioLSS0Ioxu6ivy1GHk+DI<{yQ30 z5TO=VzVm~rM&IM^e_a0yS?j=?-CPrIVjIPPz1u`ZpL|}0pU*Y)`ZzC0^Lqf%t*+o9 zC&QUL6Q?l=)|?eKCSqwB%L+sB(wF|+e7nMif#Se{S4b_7@JmxDCYd4$R5VtrA5q3L zLo$^PsBDkvZV#YlqGVIP;DPCa*qcf?6Hv?j7XIi z(m6iabj`dwff|=j=3SU~h)MJNON=1VXC+6=A988}3G6h#zs3lLoVxrW!VX=vG%olv z%`b++vq^6(k17-%_sa7v6lz>nOe++o`Nitr+09!XrE%Hrf7!nsIw+?36)cVsuwJ;V zhlzlPzI95&Y(qIp9w#RxUxc0JPJcd+U5rGPE>PA;A`9S90l-&N>OUYJ^4(4brNB z9PV8ykfD-2O1)bo6}}oDRZL{PEJLK{U1o?1C`<(7COvCWHhn~y0Af6!`mJ&`NbK(C zo)tcn^~$ji3@QZa^bMs?@nYQysCLg-zf}4Dh-7(&C|fWfX>)^SG+ofr>OA)}uYh8; z*p2DC!Fri9+`#6$BVPD$%N*m<KFF^v^S?S* zt=QyuGpLTn?9M)q8h@+{Y!Y7$=-?;%Y;~|*tGRyp^S3P$UjA8Z-eqMaC;#lbKl!z}#InN% z@GOWXw@6Sm?6}~#ASWbsswz)lsOYfaTwDA6t;2-1quK5{c9{5RANA%kqb(`%J|`m` z4s1)}OiJD#QA{629#GL)i&o6tHchS20xhnduI9J#d+7CKhrgEJA*+Z$n(bLoJWdd; z3CLPhEE)ute2sf{NOM#hKWou`?%GJaO075`(?Vy>&qDJ!C%lt+JDh8H8+e2C9{MK# ze87GX=h)<23u{e=RkN^YLr3fUAXI`c&4Nfr5%IrFiUB?E?NoG`c%S=bZaFtov~3!aq?QYMJnKQL;~aq=M#>n0 zCa4qJ0lR8~x?N&s)X1NDX(e|iBrZ56@Qw+N>~3NOUaGiq>-xX1D%bn&pmEtQ2Zq(Q z*~7t_nSH>vHC;Z~J@2tzZ54jTI`}y5t6!G!raAY;B59<5{#tn#oipn|U?Hg1bcQ}A zKw;+I44C)_YSrVMh=j1&qi6SZbl4J-jfp$&_!!rcg5e(>BpI(+2}z~N`cgzO`zVr6 zMQ?{*Rd+;Nz0yS`-UVTUTpj-@FOz%EwIJlA^B(#@_ziKBw2Itw{aCR}*&MZ>i|?Um zSe4|7*8oX!J|V#>kYvt)a1Y%d9zPAou?zUE#P>_RQ!{AV8&z zkvREHKF_$w%-#0Ada{Zg?bd-UMwW@i*i12-D3XL#Aw3Y&Me;4~xNH1(`DD5thFv!b zv7nD?m!)LHT$mkNZ!64hwDU{h>Gq$w8i7&q^=~zj^3JU>2g$bRY{; z-H>&M?hMZf8FrdLg0}VOV4t!X4}f~)?|<^TGv@N$(=c7pUA0f45QWxNm}AyK5n0tjXrtI8J?v-bf*8+W_r+zcP*#i_ zo<x#w;!Sp# z8tv*nv2&^Unq$FOUz84zMtU_*CtK@RDlV2K^B_~JSgtGu##)S^M7V-uIA@-x80}+i zb&lm@wbjSIavWA)U&ocC_^(4A>KR6lFjQPyvI(DYn)r+e$HuGC7mwX8-7}z8_~v3la$DA_$G&=?gi??7;p zEy{tEU7Onh`&EFj#PJ9dU|%PM9ge$p{8(b{qI^k~Aqyn{z)u0D$@a*4#eMhg8MQo6 zP;Y{YA+2zwbT4REtasbzUh2GNLNYZ*o5T)DuQX&fyWksxyPujj7;{+G$te?4a)e?I zQKSsFE0sab+zi(eZkyu9!gL>qepQ2NPP$J52dhanGdF;}=E~=2GR)Mgy7vG)c7y^OGGO zUN9`9pRfTcSXq|If0{BS^q-9j(JzaCHAos9cp>`C1U;=3bAtkHr|5oRiU@V!RJg@y zqxBiO2B}(cfi91_?~a1BpAqm|Q{|7|xaXClOnPg`zc(zKu864Nltk)6)9G^2aWCLH zBuas(0Ks{XTR{)F0lI)^Ks`JTgLi)hjLV|A_!(86@g7|utG`#Ui@|q41HD501E3uu zGUaNZ`@>m9hhD9tc&c$Ul5wHs-`|(Iz;?md=*WBF?8~6Y z4r5C~InCxtVu!U*7J8BCTO0~h1|25c2)$zFUiS3H^PV+`7|-})y_X##R{di0_syv# z78o3ZZumvg$FY#FwZSDLBwJbPZ4iNk?>yS`lF%Gw(KM}0E5y?n_bRnd7R}`8SThwb ze5)oPh0Ja$95!VfboO15r*UO!`o+OzWZM){Zn6O6QA`d+c2m&}P%m1|FOM9Q_j0>| z_SeRb5v+}D+$aT6Einj*kCH@JVT$iJ? z9U3`}K+71hVt~(xacq6a3M-Dgo!;^8&+nLLejIj%v5?C_p8>NtL$c&>*y>CL?)|pe z9ZDo$VxN(TUT;_+*4ChW>3FOVJBfy6@BR}jC_*?Qp}gbk;B;hm=#i}2j#c-n)3<1Bq0_HOoNWlT{AU{Izo^5YZgHf zS1w&HO`ZDE?TF2%$S|N)0J76(ziil?_b>h>+-N+uHn(%5dWms{%HH+!XPCx6+pYNer7P%POScu|@Gva_+y%ROqp-Ttpz-{}p* z@(ye?cPOwPq-d`Il@rt!45^=$JTO8Zg%g(vCaj~(xS~2Wo_y82WeN2Di zKePZpDs}2O+VxrmBy$_Y;`PhG^0N1+y^nrNj6gczl<_fHJsSJbf#H&E0+%fmlR}YX zDq8Ej8`_~hrct{XI$Gw^x=;*r0DWnpY8?mKri!3_YmknomC0d{P zHMP0p%4XGh{U}T5-tL>zq<{>u6s-*M8!?`O1!M+5Yt^#{EMU1J0*rD`{Q^d4pCW2SGnUq1MSr)<@1f8is%aKF7ySY9(m z<0QNQkNP%!EMb~uK2K1* zZ?5m=uxPO$Ja=8VW3&iw8xR+S zq4GgaXkKW7E#k4`2GMggjXbahCNIB+6^N)=zk9DuYz&OZ{`S=-lIXzUd8qgtAu*6n zF*_)-jfy@E4$7Zz%RhQ3YmTaL?IoMJHBbVI8k0vu*F$>opnr1|0wK-s_Pdn^SWL+3 zG$w?+d~$8_E<1$$u=g$L>#oIx7yuQN{oL!oL`&d&>en|LDMF9A29DZQrdK}fbTN3Z zU<+qCaEMwxE4*_-W^`}F4c{KmA*U=lhr4~cCR9}t1`ZF5wPMEJP_-g+)~%36(DAZ$ zVsK2j8EHhmIJOC1o(;kZHrwbIeh#;jXk1X2uZB+Z$@XerfM-ziHcfH{2-fEBVB?>)?)~x#j}y>nce&s* zFTCfW>jvJFfD#T=3x?zUh$j@j{o;z<;$r&1!nITJU0km4c>U~ zvqBS<_M6y>2qT(YAJ(Um@-Iy$2pSZQSk>w%rj{Z#R5XTVQ=wBkq;in+mMYH!e+VcE zTP62{QD+h>2QkNkZ=&LAtRPQ}lt}z74+K>wE#wpAJTx6C)JHzV=xCtj#a94in~g+5hzqhf{wcH5!l8?|%CR+4!X~9z`a` zV;98$NOn-s$iN47z49ZhSO@u&Ha=FIUK3`8qDfIJ+T0qro86k36t1dXk-=Lly%dDW z#1D|fV2kR~m&~bkAo;>o8FBQIQOEz5e@WLdsM%W1dm+^D|nJy0oY6qi)f+<;_uB-r7LQxfNh05(2g=OXxb0mfsxp)OU@1O;k0><5Y+m zZLaia1#F&MfR`TF7*4EUL$&_kV#MpNdyv3rcx`+XC#a@}c=^7#V5E>cu8BU$qC?_G zB#BhhtKDbYg!up`Fm*BC*UFh#bG)26tG5!%Y2TJUj@ z#Bj>_5gFnwEs$v4dgV)Pl_#o2tJ(Din{L6u1qgVAWt>2$So;a}@z=7ki zX(lFW1I4VR$SNxOz^}Lbqv5~1i|w>gfAu`R3OuhK!9G{ii*}qf*Lic;jgAH7wtfQ< zp#+@LG|4H!S;c8(rtXs8Us}e`%__z5B}Pwo!3P)G zh{l0E-4>H&FiiigOkZk*s zr@q%X{iMY6&ca^c^cT?Em-KQQ`0L%yDeY-hp3m4=AoBcXJ07rt2&G7xuQ4wgVv|O8 z;NZT65+h`vXk4&_4Eb_*rr((kl3ZEPEgA_3+Xabm<`<>pj`L5 zvv0GN*`CKDDj4J&Who89TnYRH{;*RVFOHWlD3LWtaW9M=ZjVJwoZdORIoGcjX0v0T zm#nPE%8~_R=0vh6c7N`` zvN{+AJyxAkzg<0PJfuhm&LdhNMecE{h$!_vw)lfjm)sC{$gl@3BnsW2n>)GncFkhQ z2Eb9j@QJu=$p;@MaQ^h{qNb?Z5as#sw*13~vWro6GfpgR(F-84YSzbaQR2^rGrWu& zo}ZkbbdjWuCZyuPZ~(dE5u2cViUB7!i;B+whnwP$Zp)AB^U3Ahr}Hk8n7~>utQ$K< zy6FN^K<+Hu!hGfxBQKh^ncK#{K<9Cjrl0gK6<_u_3!UXwb8du#~UJh}R$SG?|dMWomdxHbgkgSgISH!a=FKwq$;xgdGH$^E9!uoE)t4V#@d8VR#W zbJooP?WJyKC^QO64#F)@i&Wpf<5q})a<9U2qg?u=vq7oHmVgmaudiqN#eL3sgp$O(y}T5xU*uR6(jug>|UGJJ)d#|(FAT=3|;SAF=9 z6CSObHVnB<=v2!oOP<^4Z&^9Zqu*@Dv#dYT7JL3Hj(Te&3$p;kQ>;G z5$;O~#Q<%8A9jKnlEJ1%F(f%UpCWlp{Tfuqt%DruzG#t0q;-%&IQ^92o@ zUfCVRdLLE3peMX)*7YD&Wl+8##Ye+S;kE*|4o_7}s$nS9AG7_UW+NDxK+_g4ESs8* z|2eDGJfX-YTjanyZwu+i7(t96U6c+Xx~E>c`RBQ}e21K30-wltNfLwlz}GGH+!Sb=IraGhF#U+a89w{vGm z0}x7*!?d(!W-fh|du;ZyWy_G*w84GNuREm4u$h4WA`b-a<`&Dk=w`+G5V&CcGc&da zZ_!D3_Qtp;!t%Fv6yxu>^u+snuWT|;DLL%E$U^#P553Q$1bmr1aV`y2?5Xl2O3W`J zGbJW4U!1~4)(bXJwYBs55)iS4?MsHT>09hzJN)Or7MaW5IxIRY2-zb2I+s4G>>(h1 z0Fq`(ej@2P;TSivE7vxBk8uEIQZ|=L>YWAFhN>R=u#tX_4g8 zzb6HZecPjk(;{h{X>+KJ8)45a)VM)qGQh5FE7l92?U%RT`ZbAXmriux z_#BAHjK~J;q!_3?+d@UBagZi8X%v_^{CEutkC$Ah2VUzN>2q^n`-8&D;|KTB$)laDotFoKhCR zup0rri(aBZI2KZD0se5 zjB+sAM2_nmD+h%Vz2kBBb;k?}_9t+f^i5Codte>iLH9!ShHAHe?3`42Qn*fUKuW?O z_~wSvJ=l_c*a_opcer(aD%?UH;N0M>ltM}wBg|`fUGy58Cz9Fpjv56iBTtPBAury{ z3Nw@btef%kZx|hv$-Te&gq&dKpg3^O=eCK1a*<+y0Ck><&H{P?a-Pr{RSJ&uBs?m>}9bYXw<70|dN-wuc;$^#S>N=_dE~*#&efIp(HTEHyNres&e+Ku(5!61<15 z<~Dk3Wh?wX;qMif`<@PGXWeXMxklpA((#K>d&!wq|2VkCTvpCu*Dnh~amxiYoLG{j z*KO&XzteMRhImL;*mc)|2ysT$kpkSa?58spS`ix5iYtS=qQJ&$8 zi-w#G?_7gh>gfZVbib_Mym+1!tSnu;_r~n^%ni$n^@VUl2hV3(EiJ)V$Z6?*yLw(`A^|SxNt#&YJiLN`G?@iw> zNcYn^W2wEw2-GF4rm=rR%Ged2t;fuZrooysV+EFF%#6mW*)n56(cf1t z8e8J26-c=4m6y(j{dj|P1Gi9E5|$By^+>7mZoXQv-DLtz5KAPD6&S4VTawX|@Oz_; zdB5Skx#SZEwlV7_k92CEP|W8P>8GLxBU88^D+YmZaG$p+r~*Q?DL`{d5p7^FsG`aU zSu5WwKH{J64jW+Qu+x6-PEY+w9j$Ru`)u`ArjWzzQ1FAAa(u|7#u<{*th|E#w`}&D+4c=%sPlMk-uS z)1892x$W|!{7Sk|m@T*y)f=W^PI@;*wS%x-8(rqR&%H^ODY_UnG!;3&+gD{I$#SVy@(>E>AT#~_IqjwfW7H)nG^_I2# zd~QAIkQMv21uv~9v1C2DB3wEEA-q<3xqmw>96J=PoCHq!ya!CXA8)5`b|*jtG@WE)&iz&Rr%K16jbV9OjstLauD z$Rpy9jF6Ow%AhI;GAGKL+^-1LignXgPOqj1B=}p^%Pr#K!A|!Kmrtd+a}O&~Dh6C9 zceS#L3=q{Bk~s^VGxVc2O+BJmVMu^I_}-pxX3Yjh;4y!^eb}jz`^AT=41WV;D&P;e z4m9Nr?iC8Kjqp!0+*ioMJzzy>Hm%albXW(eX4>B*Zbs<6cbzOH$qwvs?>E`iXHpEX z$G20_{bZGEMqoZSi{2Y(XeCic_er$;yJW~|V_>Fei>Q`+m+YZCWnG*zoIG#CYrEPd zqahfRd3tU{?03MO-4v5Sk)6=4$a{_O!>2qf zF4^wW?O9K&xm(F{kNu?7dAC~?M++VEnxkR@yFJyQ43ipCt$fPELbcT(+-!HCj0q*+ z$k-1bhLPE!WcsnZ_c=zC{QgvbJz2qS5p&>(a;6EUHc?CxMdGRGVonQ_>KR-0zzfa?;^Xq~&kn;3S&+69eE zo79t}VJGYki9=@7{cq@I z;;>KLfi={`K@LzGKaYcdV;zlu^sv(&w??GGHwp)c!}>R+S%3b_JlbqQ^fp^L$i*mF zOK4V5RuK4*?ukJv#w>i&Eu2^uK`KO zP!`c8T`R?*aG(b0TC9^e>@vg-gui>drhldpgv$7}Rb;0FgAhnKBQ{+7CxN5;x$$4dK@Nym|w)V@Cc-h=8;k9|^l{wr_qE-}ov_m(P zTjC9ot-)>ldJyykin|i;vzI}$%4oglM1OI{PS%SRUZ_ily1qG{`WFY@dReG{+0Ctp z&{uNT&;vphs-t0jZk7@yiCUoWTQk#Mh#7Ul*}!Pj0qk}1>#AGkExEoVSt|<-ye>_@ z#UGj-D~a(wCQcAF2(K=>1be3<-(F_vIcKe~Z}w^5rD{c~Y>(eD|4(5(Tru|wsB@kO zSnYBR3Fo8ID4}Df~}!m1u{}U0G0sUMnw`cL#m~`*FkXrl^nKy#~wDm8jZpYT!6i+3J6+U8Ki2!pK&*l22fjHtb=V1lDWJhMLCXt!^I$Uz z@X~=;`6rVr8{IF^Mt`>d`#xt#EW03}1N$eCT^Qm0Zlst*ifo{wk?MyV6Sbmy#wp%S zarXkOZP3vZ9v8Nn9r4oXw#S?Q688*?bUKr?xwK6{goeH8(0dj3m}-+T*_jUC3&YFDqo(P= zro)1oCUhPErcFfz*5+0%T-GAl98%`L3|FNjvNEK9J`xo!Mx0@mRR*C@-Wh0_vpnkT zO#54wF#^i&Y087)6~X3ucWgE}4h#_s>UdZ$hv6>l{h3QYagCd^DaiiwZ}KfRys!!q zqugijQ=N~#>L)Wg9O*ND`+Jh=zyY)p6Bi-d*BZF~%+VglOU{;5?nKiW3c&Qs57b)mkq zFXj{Tz?s8tTr7mbn!Q#^o1+SaI{rQW|I@g1FW%vt?~~%wL*JjaOOngUnr&|=n9e8* zV3^)$=U422aqreTpPw2H$WM>#c!ykdU;{E>VnFUuOb12Usc1uqN4xwn{Ya)dCb}%H z2)#2m8;VGFP=l_}kDh2;mh*1mMAV{@kPQCvX$)NnN%LV%vr$f9%wOy{8nW`^vNeD{dtdpU1o~mAv z!K>!w!iT5RkGcB;L4`9*dDw3k-3tAuQ{{th@!`;s0Nv%d0M%ViGBGGJ;$M&j4XzIb zu9LNr)}X~_u$$#fm08-JY~ zhHpViAR(ZOPW1WQ>$Ec657DaH5u%xyFRBdMFVloV=5W@nkYOjxA{vqfs{8ISuuE61 z2+RQZK6!Jyyl1xIFWS8D9uQ@^KskNbX?dhB6h5*8n7~y5MbQ2L?)Z}A8aw{dxce<5 z_L zDjG|$bLp*qJp@~6sMbbRgknv2i=-#~()82esvXWd+$%y2y-LCNmk&Fki@kR0069ha zB6FcOT+2nFB^6SZ4TGtb1zhmWH@Ksto)>z6v%@2a*Fkp)k1l{ZrWVO+!F?F!4mXx0 zf&lFX-rni`Z-9>Rp!+>2y2^sE6ke-VWH=vyZ35Pi<8gAZ>ZEgw%NoiGgtAAMx%nB( z>sMp$m%o|!^?x&>OCZ?vPvr8K23=5*H)0*^pqO@ww4zo@w&#<`IIkATiRo8_ssgW5 zzAZoQ0kZivUVj*}!?8;t?k;bLS4e;%o(c}Wp@aK%P!3Sx?z7$t_X7LG*Mu8*I>|tA zYjBI?8DVu9?(7hcYQ-J-F1zDXAk74wqrz^(j_B{3 zH}_$)&N;A_q=i;M*~&J^jwQyY#0u1y^79Rtq=8(1e4!DQ=4De*2~vw`g&Gb);{$T7fjsYn)`> zcAs{6O*p=4in{Z$S}0_yw@41l)&jjeTaf6Z!q%p_ zbR~$z=FT`USFON{R1o6?{?28(E7YEiqsE04J3@sOQYQDRK5LD5hLnf(sigc%Tz{$m3){ z^#k6im)6Gp+~-i{Fm@X@KHQlatWjQ#*>c_sa{b`d>X8oYqgYUn#QrRXaugl^DKD4q zCno~hp+#Lk>6_mU^AYf8XDiS9?}?9Ha!&=}J0@RKoW? zmg=0pCgf0HwZ1b$1>HFx`{q7^Hb4`H9;oRzsE*$5Cavw<}hMtOiZf&iVPG!^(ZI zy&iNs@09}}gh6y}Ra3XSfTB&+!eOVA-U*PxoXkE)Crxrqs?-U=lVMLaQ8fB@m z`V>1*HGa=`nh{aHztb!wTc?nNCR4G8VzMc+i;6~V*50rb9~C}|lCb;kkec5rl1pH2?2OL24&ekH#Yy6aXw=~Lf$O0p z%0X%>dcd_J0^{NL-ErlGg&wlQz_3cEn-sBr2H9OWnWs$UY7_>pj4D>JhtnX%bW4|A zG&8?=Xs_RsA7yoM+~nQczx+{#`G(kGR~HMLUyNsbCR{JC@&sM!Y~^h|zpu)(i>{z| z!cY5JhA)t6JhC3JFWhR^6P8WID1e=O@yLJw=FMm3;~Skhq|||fvyCQpt(IbHC{js9 z_sZj$2B{%hhVoE4**c=5ABy#Kg$60^WYXyi(BvsOT!n#L?7)sYk6w8KH2q4IUtLs9 zV*@eON$-_X5Q9+0Gspb#L;>WTuble0+WfTMJ<$fwUF644(Kc~_CP~ebLHeJJ@ zY9`M1PFP8sqcr@Qd~jou(<4zKq|3_0xrrRjqpX?`1;$_kAo{d9&nmDf(j2a^}((ssCcVd!z4V1*m1d-l4Ym}_6L$uc|cb`EN1 zM)DbwDJFp;>p=k&OpmH6V8FFiAH&Brt=;>iw~f(Z%@kUPAZtdl?KO+9O!~`s7WIzIZ89_>Y~4vPgJ1R zI(&>=er$^aOQWS$9bEgmQ^k=}ux{Qe-fA9J&SCe_L?2M8hd69<&;$QLxyEHbHv?V| z%2!Qmj>@Ig$~JxsO{hzYFam3Zbb<1jny;V}{_uR69g`L90DZUOr0 z`D;HhkGfe1ZXHpgctY>&8gCV9!M8}Nrm8*1@c>`_PqA=?RZrf+z4koK>d8Cq%3To^ z*dA!~W$%9H{sEH5E(_$q>krhajNn*SPz>ZEN~q|k^SV82IXb#Nypg`h+=DL2Yvt>e z@qV52o1=QUF@Y;U9`A^MF*j~XHN8TzY6^Boh$Sohsv)1adP)kZ6gEOO@s=b*Oz~4AdBfVydEx#&OeFa`+W1hWl!|w$D;J5xG(uf`373=qsL+r5QxB(<` z$pky~6a&&{$57?7NuiUJOI4@?kWORQ3$3g*NGro7Cv~do8ZZ_zLT-xF{bK~s5OpTX zJlB9YUb}oBw+j?KF@FfXFS3+6$)FpGgQ{+c3!pmmF|Ffw$nag%|Lk^#rW;H~NEKwp zumG=;R-x)ZkMt5OeYmF|d*aKcPGZqbzCkv!gOUTAp&}C~ z?V=diEA617v6c-Aofg#5r$JyGUt!hT{lE;xqwr(=9=dMHv#PgSVlDTIurF+@1h2BC zVX-h-V*`nW!EAVX^P<=viHso0^wP}41c3^>Tp))j z#YYp=0m_QivLn80TVC2+s1|G2{-d6q>|mjR_fBE$wBcRr%_7stQ4xB*(M2y%+YbfS0Marq@!qEP(B#st# z#b`+kBnuQti{r)da3_-v75C-b&x8$9L+uC#kn*8+R;@HtAl7(4^@565(h-_8O|5_& zBz-CLmgi2-BYFa5x~Q6i?Yi>?t->D90&B`xt(pRxBf_dref34G4ASIf`<#a+8}YMp zb>Io0Lb4jbfp^|JP0*G|F&ikd)_}GaNgnz>sq&+WQg7WvQy>g;XbpZ|J=?aw;>W$w z_vacB@!?19N6DTkM6Og-*m?|H9BV zpfUmj+NZVAP@8+3?(3>TmIbRy@2I|g6EL{gF26P{bz<(C3^t~%h zf~rNWqRboR+13PPEy4=MrAr|J#p$2FblVS>VgkxgCf)JFn^10ci+^>|HbuLvo_zAw zKv;>m5kh$=NK-SrM4SzU(>pys14jF^kyv^vF9W4*@&B?I$!l~3GT zgD~R1oAh(g@7W)rnWk2@Lgu=S{wy?)gB$n*o(FBbFSj8U*iH5;u4ZKmsrR<-`eUfk z5uL>(#v0da1->v>V{!m8b2ix=< zpn9TnT}2XL6NMbVe8KpspEiV$=^FbAh`j1;RuG!>kN4g_VD17ttnsnn)MmQ&%FCch z+@LG^QmWgYdj&RJHeU6-zj)soM0|1e#2v9iM1$XhBhE&nQuk){U&%^#evAXB2SA;A zgddYaG07B3prViYH-H>N?e{Z&36*r%`U1-7k_EM2Q!B2!6oyVPl&k@f!Cp)pfbf&< z-LLzed1~>+x*`XjdR zkZ`{3iBLHls=wCW$gmy?rVEJ8!@OolvusZa|F|ppb$T zKkQP-^6BlNK=T(R%d=dM$PzfsQRn=JW@nLV{u<^OU?Vo$-T+HUy7B4O1W0YzV>fsZwub5M(9M%J|K!(!k)5}G3 zRo%`Hpul|}NcyD84c)djO}!2oA&d#(xgzBq*J9|zSL}J+MTJHOJ(hgIzEIUx$^Bpu znvOuhXq~){pu?KPyAEp44?(uKgT^~-7!9%sc1GaH7P!GNSYc<<{rI8DuRC^jQyN7- zCk0xIX=kd=DW8aINt#r3oSVtT7Z|6#`EE`~FZbx;&lam%ULTUBgK`w6JwWYJ{!tH6<-x_+79qiBTflk2BY`R*C>iL$>7or`r|IP` zD$O+L%b{p+KjN#MrFtaKl51yaT=E3iy1k9o2WSR>X`xmmhgH(YX6_3`<(SH#+fbbg zTX_(aQN)L&%Ap~yw19J0`b2cp9l^Zbd86_JtL)iWaX4aeV|{1|dn3qcCdPBjB-e!ye!;qI^kR^gr-q_x)Rm~=47zvomsR@jVzQL^0MRV~bbYd(lsE-RPR&cs4{J6jOLoQ(wwwwdjhk75}MqinP%A4`~~XCrSK8C1G>x*+98j~Kv*_|6gh3+lgNBuys{2@qCov4 z2&m=x+8bJ9L6zPUc{wP?{ch>~w8O7E{Tn+=V5An~DJm_foh|v%gNBcY(Rn@Kl<_fH?ZD0} zXeEwF`)r|@6pAEM(fNY2Goje5VRoS~RelVL%v4C|Dhb;F>`K^s7zkMQHUcB&g&i_R z+-T=RR>+tXBvf4H7?JV&Q~mX1g##l4G|omKV-v+BQ6!#_iUbxTRCs?6^di?YGo5rJ|?7&Q0i=9EVdj4mV;%V?8Hw(ga zTKeju7(o}Xu~X&9av5|3?S`q#Y)8s|<&;~ArRU*}R}N^S3pQS;_||-*&G>^Rs)iIe zu+2DaVl$3XOa%qeJh2R zIW}9A*aG4?GAta}2oLPCK$JA;!XtA6C7X?n0~;3$bjyb_)S){LeawI#mrHM(-2qxJ zp9Z7kv>#i`XgE6 zz$;RY$y9Hpm{f{vq@u9|8rjpIO0gAsz93!HD_?fobI|P;NGHS6B#IRby4va+Zp~x0 zezIYCv1ULUUBk+ioz%DRYoEOCjM=sF?U2{QIwTcFTd^W+B7yLH-mF2y^NT0&fE^<4 zNUL0m)V7@PaAOnQO$Tn2ejwWr|+uQcu+uPf{w@V^z%M&xC2Vk9z7Q06;Z zr3bm&C;yp9NLhoV&z;r@M3PTq)5ifuZ1nzEzJzRbCbcGCYAMBnhD8BL5|cdsVW@ej zn!1wI@l21J)ElV}UjO79<#X3ji+-8^-7P;n3(euv>6BP(0B^Xr)3sY(C|&bKt@0u? ze8+LNyPT&BV(c&toWAYPzjpHvwx0%0@2%rHZ`)V=Yp2Wz8gZBSwe4y*kPSKFg&RR_qe zh@-K$B5uqWRN%8G$;iLwJjn~$AgSkH5v>X}JhMeM@|kB$@)&2x+!&4F`02Cyg4m)T zOlp}TG6LW6!#U2k;FshhjyYtKO7({aG%DpxFX8GAhUTE_B^np^JAHo~LndyBJ4 z#S61koi_o_hZGA!5C^g6TFH;g19nLUooa!U-_NuO6Ff6jC5lb{we$+o>D3@aBO4#~ zfeb51c9C`br0`vVN9d)_hv|0yGUqDSrOv(4DZF@5wRoAcR#`z01zw5WM((MS1t-WR zAD#LHH0aFborXQSPQ4iV+~FPFlWh(hkAn^~BYfO4iiOg{LMpCNUMpK5YNM-#iNIq_g|Z){90)pSltx}4 zs^%xMy8=O&2fLaj#%`JgxM_!~4oD*b4{}!^rVW?8{+EB`@GmQdoVw*$+M)GNjI1Z! zAPsfQwp?0nsCl}9;0TPT-p&a%lRm2ZMzFagi^DP^Er_giGFpPl*}c&9?lOqkwkq-i z5BpV6^^uj*yRxKNDL|=4UnX5r7TitWbGss14!wD+#pu5v^}2!?R2&FK@^v+RnKZa$ zQtz&q`vLVgt@0o}sQ5Iv-L_3;UI^e&Z=Lbim1p2(GxHfh<9P962sLxkLw;b@~CZzHiHp+s`at@;{X2p4?CwB zm#ewg@>Y?22VSm@nk-lQD0UA;%Bi?>KIq3j^zMcH@WaqF-g=)tw;ujY@7jMuk7sQk)k{N73(jY7IYZ11IYiT^AQ`gdmRfcQF$!h;}ze-;?Cuyi;*MxZSH*jB7y`voXMiU%93K_pwcdrO|A6p3_t~u9`(lz_A1W zTq!4%PSh#2%3*z$GICk0sh3s;*Mp#3E?vjdPPs7+q+3uj(PpHTXMv70#Gf^wjjn&Q z;x9bbxcIm|=)Fiba0|&eFdYh%H%IKOaw!&AE}6KqT7x=Y@zI5zWl;2fTD*;)3#@*S z5xdCtfTk|!>iS)rec5%F-*9n!3O{#>)*p2T(?J=mEC!B&@NGw<$Qdan8!8*YVw7Rt z_{pTXlo2Td?gJ;upc8~fiR&NZe)*fY-~Q>RZ_oRqbP2^SqDcHi@-asFOJQx_Y%=#r z6uXil%c!{Hq%dlwyG9SuGNdx(i>#{(9`$#O5`V__h1G-<`e_PeU6Z?jrQHGI8M_pnGw-oI9V2|WkGGJr_o7m ze)ar2WSuj}?~h#ivnUn{`LtBrDnYX7^f;r#U&SvHBWGpEsdsvJ z0BZ2q#PQ@jMb;pK^Po0(VBX*QGla%sJMlbKMgr2 z{se@MQM9c|-8dInkEn`s*R4dc=8L&8N9Zm5jESrXvSzv;uO2p-M=OWmwxWD3^G9`3 zBRCZQ_)-&D?ZDswpJPP)ES+K@!n%=)yByvXgW8OD#48zS(OnIlTF!;*WhlUi4KowG zuSk)-gtwq#n^rj?vN?{sL~ugLq{dGUz4g!5e0dWH{WZ#WNcTgJvkrRSoE}EA2e?X0 zu`wEu-^%brqa0+7XJ^y6Ri2OtJ~xA7G(A=tjq%*}CFgJa)x0F#VMFj1ip;TxU6t$E znMF}$F-2;yFz9cxI0Cvz#o)OK#WS4)QXtEloe)J)dlJ z;I#!PL?hOgJc@-H^=;TQac@|o5`|0-g{*cj^1mfm3(c0(>0;hSb|++Emos}9t+JDD zai3onRp)n!S2Y_lIrKWvP3fi!;Y<%x8Ex0~h67d}yCa`};tT<+AK-+QN$>vXH+RgX zK)GyN9C&AIK{ymyP$+U=0W{a^46vv>LEB_FFhAUc)1Zf69p2!QFFGY&9-G9|XeMA| zJ2o^~4T>?}_?7iL4x0zdEX1Vr^nQpeHG?4U@VALgO(Elf`ZB3ZDqlMVx)3<6lk`YI zVkRIv02u8^)M=KbhTUL>oSKxa%IjV&!s6g9u{n`=C0~T96ZT+RjP1Mr{MCK*Va(sD zr3X2^xsIC#bNhDdEb|PF1>uo9;)gOUf6+pBf~6wpYWtZiVOA@zhhO4#01%QZOOe>s zhCeYAoS^Z<)Al>@f};|*;pCCQ82)KLjNr(eaN@XYx32k~ zemo+O4qRkxQ3NtCuuXBwd$s>!pqU(rIi@-(t>a*25=5eq)iwke-PBVTCoEbSD{Igbx#w9w76VOJVEe3gq^ zgnOp!8CO-e#V>pG3*5}W=wsO9EGKBZ^wI0zy8gUFBY_%veJQobxhi(6Y@ui)yV^Z@ zW+A)SIf+UpWqxbJKYH!zL}p^RMcVBF0ZU_Yf(A9fd+Rr}F<#dDpU_@!p)<5B8X1;a zL$RwVvWkk^A=)$xI*w!f;6A^Q)m(`+KH zs7|k#Tr>4vgihV$0Ug_nSFb~JHi)T0XABI;><_M*4LVP}Tl^&6?I7R~P2J|Mbw3}R z>~f7D?X;7wgHCTH%r(_E_bS)VWP0Ci?mBh81S$0!L9#>#Pv}64ZeFY3b;Z0Y*Q9AD zBDs-9tcASEhHI<`9NrR6kf(|yA4;D0z6T4xd{sKH5&n}T*~vfdmZQ2k{Ww|cdCsS1 z>i;){M&xiqMbo{%e&4)K(}JdEi_j1p>GVSR1}v{f4)}&wTIjxrfxr$rJ>Z~sS@1;8 z1E&|0!;3ZETe(ev!!lM|zF+q}bCCcpNg4+ZhggslsBpgmd4g^H2G|cDW0udXrTb*3 zu~1Xp?{-yM8=K(W3(TX7|D~JriC0HZo7!*&tJzaxPmwMttcVxs)PK9qEDBplU1yHa z%V$V*$ZM%EKt?Z2TkMYXF+Yvlu{op+r!W0CwO<0I=?3M(eG4H|7iJFBsJ#p+Z1Jf zXZ&_CmnMdJVilgujD@YSGun8wsMA?;qtQ6_^>17uYhD;-0FYQ4K{(k;u}~en8I=K+ zd0dZZlqZE_#=6QC2`AgU>SlM)R|0Q)r%P%e(R4_$Ht0~u4rz}(S%5uJdRcp|m!Z^aD!Sr}!*b!u!#FtZ<@A3g!JF`wYH{7g~={bI?>M%)DcrfMGfz=3hI+XOe; zDHeJ}Zl&Tj`FGR(v$g}=G$VJt3F&*^f@U@dKyB?8ja|cC6Jd1}YD{@R0KWnFF zBe=70A&=KA!*&SvOjC~&Dr1JivmWFIl~up3{HuA?o{OjAxGWiBA^nrb%K)u)t*R!Z z9HMEv6v$j(+XD0zk@TDR%X{1HJr)oUm6U3LT*@TO69^ol6(FX%hVxg|#x}0sS8EWG~UN zqdyH^O}0wMYLz#b1A@i!p$L-`mtBjZmQma+5O+v@?xS%1zzKewAT`PJPd+4nfB!d?GirN>Jo^arUAxyI{vEa6H2qu_-EygINa0D9taZsb2K&%=biG78(am z6k6CpwFo~XYdo%bU3LXhi(-3F4x|Xn<_^qhp)W+~WgXBueFrp1FH)Sgm5{s@lgtrI zR*YzyTjpn#{%wZQoTM{9$tSxU*qod(F(-#8_5ej{AYLoGE34;s(+|FSRr>eqOab&B zNRxdi%H$`yobqlEVk4zO=`wXYspfC-x*G!#6R)-5I(0YwkLwI>oQf0`o^{H4enHHR z*y7kz;`4&tvD;;t5}o=ZWwIbC97q2&y3)5F9GFVq-LkuUNIjF=@*dA>;kBs+AMX5P~A$p9T*(?VM2+ixq}vg}KmZr-lqUB}windYH|gKvPqK5HxgRs^(*u z7A0;hEjo6bQIqMlcwk}PInAi!dX;nk^phGFBh-$3q4w`&u`@*9M<(#rQEUoDlBl>$ zvPOXNnpLiQ_;H4_Uz$97)G>U)}P-8>a8M2wZv`sXcNZ>-5B9!JI}w+6nG@~_9|w1}b3!)vQ-r=l!sHMRcrCiSuZ&X}7x_tux|zVtCr-fjmK z2WQirKBwdxf0X!I;@l0do%#0Cx!dVuifTG9;5a!T+!34e{eAzXQ{#W@1etupxZR2z zx?8X@WF7U{&w9VBduwp+?(t%XBQPE>D9`*fKFDx#nILcX>3imc1&2)%S)d^36=fg` zTq)cVthp4@@7|`^Dr=#Efe@jokltc)Lyt<61&tFa_Q??%>$LLI3tS53ZXAVOo%uBl?Wz?t{yg>smvM=xF)|tAEm3qaXdpM&H~<9c{(fo+M zZ-h3fanHIzg6ErFy(P#F*GDC?I(2buvDb%RydAVbvcT^kh|1yhJ0#H)YcyJ*%WS-8 zjZ2QIgFeYW5~Ee3W)4S)*&A46h1d%Mte-5K8~S4-u+C1qIEh@~wv0J&{N=vM`*D+E zZ&2hq6^Athg;9M(gWc4jwq1|}FUvqg33O9m9H!YuRVjO?SIlf>76`8TYf>2~;SDMh z_r#RD^zajbY6p~<7U3>}MJGKmniiqfzce(PUK-g!=g^v*$R2*>oI7(g2l<)&BR|QK z>dEkkm&N;H<3;I`eRKY@?lny{zZJIl)j+99WpDvqD9WV|1Xn@nzvgcXM7N=2Wyq;T z*cWrdZTYP8Vx(hKLHP;{pURX#`c$*seP0mDt+g^$z766Gxp~fG;$E4@FOP*!CT9ER z)a?tQMy$Q}erqM!?!X)O!zSq2OR>@BSD_AY)Ik2fQM2B2<+uhYFOU zD1b*RbB*LgLRG;fk_xmbt#U_j5x7WsQs8D*xmGI6{dyGD{t*0YQnmVdD z*liT3)WjJ?RH2O=v|(ux-gE7VY2ibPG-qQ4 z7o(6jU;0(k8@tVA=(uDj9XR%9K~3+L$DS!GV4np2ld@!t=u{Dsub&ZR{Lw}cn5;pd94fz<7kyf zrH2F)5=b4(T0Svt&bhst4jaPTu)eP6d2cfJa8`%Upwnt5Yj!;FF}Ks_33?9;M0#Nz zZ_sJy%%jp)rhvcUl?k*8R+@teT1Ot68cwLCrgXOXns`QsfuiAyl75@OMK8JW!6ot~d{++k@|j zv3WPr_YSyg7KLR<+9EJW+@n|msWE&$#iuH&gYJu2F=NnaP>~V~yD~|s*DuC}C^LZB z@{rRaHtCjHRQ^jN{tkchwm)g*mP>MATeZ*xt#>K5iz2tFxI*cAeow&mparlKY9aAc zGTrVdZ_KC(sa4eR`oR+#;Me&jIOCN*>3TOjyTxOPGZs13{{kF-(7P;xWd0Jt03_lw z2}s?k%4ofCK#&%cPM7fv_wNWEZ3F~oDzkxiPhq%<_vY|%bgeQ&P#LXLXTyttzq!*Z zCH5Ng8Te^iJaz;(4JZ7|cmvEr_f+P*xKCOGfBShcMufI_=zwO}$lDyV#pA5{2@C2d zAT?ihe}3s}9x+DPeS6Q23{vL6X0OpCynB#hYbjDi#jSKdq&f*E1Uj<4zd)2M*r?Lf z2~PXssTLYeqQ1;#;n|t-A`J?L?BQ3$?)2R~W0Uvp8CXe!R&%%~-ZkcKOf}T!?;^Ee z;f8`S<^x8tLte~c_13ZJ<@7-hy$tU!d_{vlx?>M|q{+-vi00q%(~j+oi>LkSr(WY` zyrQPhKVvR+_JTNCyG=0nxiH!V(&x5g?$Xx=|3$03!1l=sV*2^lz3>dISbQwMY>=Iu zIpDnngpRb3gW9RU%I-YgVeg2 zE?V`AL*^wb4x2f!P=rz|EcHqz=Xw47qY6w@m$JxHDe*d{I12tlxyw-?48Qm8=7}j; z8=fg^XWRa}D^uWbT3jgAnm6{cMi<5HLGMMf!GT>AAi9i@-OZ&~h~Q*WaY>SHI$M?P z)uJp8UK>bun`An>wmjm5l$ZW<`_=EgZA41Q|GfM%x#3I}n`}OMDYl0qom3oljM1QA zk5>7}ebgxpZGBTGYjy%}xPDGMLD}3KI+jizxCX&CF@>#t>Li@42SA?~AJPZJ`kaLrxfkw*U5Mj7S*s z3|sr4Wxr?oj^#mtM!WL6o%O59HV5{Gp+Dx|oN=@8aoM)2n02J)lV2mZ z+;4?7D6&PX_~5Sq{mTAE#r!3=5<=#W+5I@;li;)nqbbSs&stB4oXJNfZ)`Qi?x)CJ zDh?BI$z*Y)o-T`7DC$w@LWcW0Ak!ZIO@>!AhemO+$2YF>zy~@>USPgQKYwxLWhrcs zPz-m21ar(Ml)Z5IvQ(>F9C=L94a(o?v_5J$s-7=OhMc1<=E=%$ezcd>@;MvK3#XUY zao76~HoFsZ{h{Z|U$LOEhA74L$8s%x&SMOif!%$lJi z(sIUN;Bx8p?hP1=$))wa@Sk=5*l3{8qxIE1WjrWFbVE3J$Z1<-yr=_O{26vW>*RN1 zY!KkH8Z4I0+h|wD4HwH`8O>WMneoj(P{xI6z6b zGQV&8{SESw1MibRF#*9PioHOQbErm906O0fXQfMPq1_F##*Zq_lRhD47eA17gO9Qw zRKIu9jj^Yp0$Qi;BWHz3L_i+hr_2L#MY@Z$DINeD6#Ka0lkL!^KrgIvZBv}*>C|WuX`i~bSsna%3_Gh^i4|-07?_w+9EXk;ywJs!99v~P*kB)AChzjbkIAY1V0I+ zFOXJ>cWRY~{YtIYtG5+ko`jHvGgcyt(+WfF{m%d9Kkqfw$5cmxwt;;(&O)fRIyjBh zv?)$`-;2;RO~HOhm!_3McpudlmV0DNt6W>z2O#r+DXvyVv({$?Yy+zVcJzkF2NWx3 zeGIp80?MTDdwI*fjdtkArzV{xX%1|M_L|tCe2N9CLlzZx9ysOLUwj|%Ep_UX)2_(6 z#3zGK2cINabRi(4Dtry86z`ms&+jD-vKy|z(G(1lMsX`s2J(rE!m7f%1$$&WXZ?S$ z{O0$+j*||zxS>XIe5rj)w$z#@033GbWFgp}CQFwjcn^TTff*f;^f)nj@suH_0{*Jl zL4^jdU=nB`a2vr>aQy*YG9|&=HUXcfzRqJcC{G>K1{b**loM%N6y}LVhsB76RAGzo zGsRA3P_aOCB`}rQA==8OG8=Rp%*$9YW8#b003+r@aKa3=C^)13RpTO55&G^@vV+_H z$$^v6AOkoe#k8MdVGY|w#U%(nkF5A>mAIP%rouHyE@guUbDg<8TtGL>*tyQ+OHTN9 z3LyviaiHFJ1$5X*hhX*=-{E?gd=a`kt2~Bf;r!EtCw!53g95QBq(gilxEf+ar@U>0 zJey$61bDWoF~0YYr1-BJ(er-ZzMJHz1G_MwQ#Jw*I*L6@k<*~R>3b;T2sA(MjVTg$ z2~LN1g|!KF>UA@#!yk}VNFVmb)Y7@KOg>&YCA=VBG^+^&{BF(0yO5o?d*<%gtf)+h z;Yk$e+ab!K+vVK>H`F_!c}~+$b>B*&^5>>9kH$8fNrN*uu2%VYWS!b@t=M>XMu=8< zfE-nDLXjQ7VhWuRN9M-oX~*0QQ>5>LU%g|5R>u7!&g8ZOLu-`@m2i+^A5f%^ircEX zDotcHjY1SGE8#V><+BQXK+WzzNJdC4y;q_UaD>hxEzrT|xtU!PEU48WghSzM1cMsG(~9iUIE>PhMcZ3TUAJ~=>U#mKeV1Z9#AC( zt+Z4ItMmHZ@K49fWnw*jjlmlR6&=vAJwI@{$8mB2-eEEA`ZCW3U5H9nV%5IoScnjn6iBxfGd~kUOlmZJ{XqBi^-$RHl~<3hKavS@LqD z4=8U!-d-OyEKtT|4_Js2oK>T=l zvK3-S8HW>yC%v=LyX&{c#cc0S-R0!g3*%)kH(B}~P;4JX?on}{1|JBiRec()Rdz%b zyXMjtXKNZgfFUnGp{$i*PdTV4Z&2jYNfL-K_QX`o#Pj`bM}nH9smw_gmYO4tei=1? zK6Y)-QbFH#zpehj^xDcI6TV5A|4N}}@AS;6n2y&Nq<>Dj9QR4&Rk@bN9$=D1DUltn z)w~nrs(8S=n_lbFJN@IRRqP(GCD8^Que^ZfKu~`*>@m@APm*B4JC17zYv#C=@v|ZR zjbcGqzK+c1Bmvf%7tSaGQ$OT%W3mSSZmlwxF7|WSb?@w&)LZ{t1?0Wts&oN_B_WY9 zwQ6cBox^XJKOk7Pt@W>h&J8!o7XNLLhxiGs*1y}UOR!D4KM;hqLC71gp_DORN3o0z zlGMmnwnlb4s706-v~NxaWBODrBv(F`CP^Ouo_o^|3G~oIzD#;`?pFFejsA_*P=?+| zP6szLO&$-u@ptW&W(dy749~12hoGLLF9NAgTII#rI^IrlC`RjF9K0lY&d@{9@|0vx0)<0J-UW8pQ#1M4qo6{EXJ|hJ&8$C;cbn1To zaneZ|VH!6E)RJvLeK(sz(=X1ZM;y^%)8GBH!M?X*-nXNTOX8bnW)zYVZcC!$I)>zw zNuZ&QVj+{ipNea6!5rB#kL<{188Dog;exEpso-&eqDni+{n#RwB7Opi@H~9}am3TG6#tsJodo+ArtJ)t}OJ7kX3#$1jZTF$DJq&koEB%uo zzGai(IY$gxX-S@j4(qPHRpB~YX2g)VOZ;1s{K8-e3TH+Ht1>8d3q{hPUIJ8bdYHC| zM)^rnuIv-U`=Fvv0aYzdhPu9fw`!kkfB=RM3Mk|zcDWWz3}a|45V3BO$9LDJ>&zB& zQ2~eDrCE>(z}&_%S3QkoRQOqgE^odB+gncDI%1p#Yv*}@r!_M)wIO){pUJx2Y(Oda36O3Pf%SMUo|4)@LQKYCG`#*Zf-G|<0gs)`J

    !UN@HrHfuTF2Rf$^x#oal>+XSb}$tw1x{NrGa;y+pWA1 zv}G2|Oa(xcfQ*tVEdAID5kAyU!<*9;NA&ax&(9R&MaQW2<1s0ty0QBqI{2__o@d0stmfGt`(Yo!pv`!|9LQNBWY9Y zj?5I7iw$>db}#03%v00%=#>%aZ`~lBG0n8$dOr8Df35H|1ByfXR(&(?2IqKS-|a9h z+6#i?LED%%)lRtc{6}o~cx3OKI0k4z_70up%rNw#E9akr3fH=Mr~#fLM{g*X-XQP% zw@U9Jk3?Q8e>rgT7tC&jZS9?cGB3QdT6|kx`_99+`WM#yx=Y@`YYeR6wMox&``oj* zJ>Y3;h`n=A?S~sX0mj(PZ%$l&$HQoMesPg3B1w}$Q*d;GZx+SCA~y|0+ZTTHH&}^u zNt#@TT^eR0UA>Vf2Op^l)ey9DjK*PPHbVpx1LXfoO1=8X0j0fOzp?-LaXj3yJG9eT zR#SlUzJZdaC+_C4Vv8Z!LjP`A3(WimTAx&oew87~gmQr#v6h4MKONF0UJWp8T@e@4 zkR>tgkhsrluWLOW7p(IEEfw)$ZYgj{#z!~|Gvk1Z-h|r;9nW>c3NFq&;m7}?;@o(M zJDm4LAQp%_48j|@y!(vk13Im)kX=>dg9%zVsVrwYWyMsFv@*DuQb!ihYvn0ZCPXX9 zHl#lXOuL`hcHCpKdBx)M`;tx^_O(#OSm}+)c(~1kz6d6Ib#O564kk^MEL0s1Dh6(s zt05_Kcgc56GceCN%*>4%HJflSs=cEQEnB>Y-~42zo6#@%i|WjuNW2q=?D9<aucDltf4941v zjQyF*H_hb>oi?OmL6ERqk?MEU^}g%|FqDoy?ufY^+$y^sbL;!N-z|UZ<}V-o&`<(h z3%ikh^r>mrfzxy&X%RgJ7I&0c0-+CH6Nb>`*i2A{S5E2`kUK zVTBs1>reYXH7|j7TGUu5YHs5vx^@Y1Bge1>P!@1PfjU8TUPza5!z_|*|1QZo zX@Y110Q1~PTZN$KS~>ni@`}4!fzcEBP(9_(WF5Q6tP^9R$i#2iMloA)RT*1AC-RCH zE&)oOUOsFf-f}-ZzlG?)m4Tp6aabbnsyt3`%N^;FGei#En;Jd1R>8w$@)_OQamT;E z$@GuMWH|9^*Mi9~Xo@w{t+J~jtupa$KlI+BbmogJ0kV`QjNtupDpkVeR&rJa*f;B)kxi0_=L9TM&^n>2O=zY-p>A9rH zgWk>FkW3=Y1AW{&uaqgby&(-{s$;zFJEIbA!duF#NTctffNPCskDUrEr$g`K<8IvSh9-OJ)7&s^Fk>@LK&c%Dv6`Q%8 zQC0rgulM|4_TB`psWaUl_lW0^ycn_(Oj1yg2nMmT7)GRxZRbvBpSd%4?lQmI->hcl zj=!1CbmnH-Y3Fv3io1XcE}#Jvh=3rQP*gTITv`Q0O^YiM!CG1cMMQ=F^N^q<63qz- zH`@R7Gv(xLNzeN|&-=X3^8HGBWv)cgSgeQhm$8q4$O`* zQM>F2UCYg58l^Y6pQ)BG4fGK@JyxeUBs?8n57o}~zU@%x(Ix1aao+2;s2WHS5?{EOjNrtBA95N(2%!_rxq?V9(Iq7gbrJEYshT7#k^B9U3fzb?a~ z$pU_fv{7G6Jcf=>KTSE zn;ZY!xBx$6m(7OX7+||aeFYjgi=a&X29&9vR)fM9^y~Br?u&MOH<5W~_=e831WrYc z0|>Eo2`l8dZ~Xw%)3wgJTgYuED;xbk=7QKWQP}7XRZ>MjyIku^syCaMu&cb=*43`t z&kCy(v))hAFlJ-&$D_SR$On@~%M7}${{=cgBQ4*0N}f!SHI&9k_e5iI_(E_xbO-c% z$Hg92CJHex>`Z%Q-zeH?)$F^O6VI}Zl$SpLAxCINMHX+vWs>N|tL2>*2+&dTbc$>; zsYAu{^Mgxd+AQxL=&DMUC4}Zc>`~S$VsUo~{%-=IXE?4LiXw(hp+) zB@SF+8w{h_gWn1MJndQUT2A05aFJGbZFCK-qdLD1b(B;!IMFrAZSopkGPg@m4dh;0 z)Ym;naQR+CUkI*(6Yw-t32K*Vtqw%4pNuVj*iu+Scaz%bI>q_$o6*k0mILS+iKNH3 z9YD%xm#{<6Pi_MJ%CHCPk@+!M=Emm522lpuk8Bl|r2&_dPGzLin4&2r8fJ7zP*7uB zXBsxh7;OYywT~4tCeFC=e_r~F85u(Hf5wm->{7yRobyYZIMPvxn<#%kk$y^p<&14S zs7>OT$i1=k^YhsAqEy)>zpZ|?q$8?abXSa(DZn2Xn8-Ey)J_JivZ2mKtW(MXKEY(@ zc|7k`6ot)X2065>ZVs)3*zlK8*nhl|e89&W_Q>|lT_rmUD+`dqKnUko@d5MY{j#iKz zYM>werap32wx0`RF|wj52}HZf=a5S8g}o$=KYY7Y94}b~bc?@X1K7vymh1N15C7@; zvFK>J@AA)`8=Y=iL@M}nV~TtWwhHlZQtqjc{8o^1a7;WwvoU%M^b{YxxI$I z0;*6woUG{>b%c~Oh9XmC$xI`SUDNHd0qKXH^$ebnL8qwNWmvnS#mx$T+;rVsH-^zi zBjL2sw!RUoiERR%{n@uE^ThVR=fMdw;NiY&>Ma&4szge@iXzLQa6y$3{87LInbyF$ z$xUQjQ4wqXkVpG$2Ovhf%;o!70b=3{mv+WjX9d_4M7r@bVN52vII>A~Trl8K$$2pQEXfElLUKDP=<2*~ zdc)M+zVB*FUg-*fCJ^2O=yp3H!Xs_GL;gm8BX7Xt&eX%C-akd9Q=FT&cp6@LM3KqO z5#9t*qC{r9NH5L{J}1v~P%_*O^o`XwaR{lm-+6r#*KA;BZT`^_vXq_Y?8bNkweJzK zI2$PWdWs}dnvVpjG91~JlQZZ7(dHRC#p-!)6}&h=EJj97?1hipop^%5>FmJR~%&ri8aeV{DF0QkBu%GvL?*04lt}=RPm(&<@bcQY{ z;f214?je1=_Xa%v{sctW9us?yj0IEneupFHE!a1A@~dW)+~;JyL2BIC5}mUc_Y;&H z$ls1o8l=`v3~%J^iOJ-jE z)ia6VYlB;&E{S$YYj_2s%GpM!5Z*X@ZG?kB8Vf}1GdYd`1g2b8A-SJU`iTFmdqGP{ zLo8~}G*7GJ;~IN8Z(YcN7ca;R>fOPakQKZT*rLwk-<9S-X>ntCGB;0AC@K)AaFT-d zsxp}zAO#$Y+=m^bG#04XaeIePVFwkDKYaHs>)vO#b(n2*GaKo>Ds*PFm7EK~tGzNK zvsFXh3D%xtVGXkMpmo|su-n-|%fNO#*u#VEH#l}aD|k#u+)}*b_vWeT+We%Mz$L9>s9T6F4MUgSD9)lQ{c<=1K_oX7N$8V;~y()COm4+{@F=q9o#ADWm1A%XXM6+7RQX1 z9arr<$W65-oO9cW%m%q!4rB{3i-c6%9T6ZW5!CDT!2c8fQ{sEx-7%klburQ?goU!e zMsaTBcui&wjJX{nHuh6k&1Dl${NP`=Si3qlJc=AuUhq-UEkZXUC1{PjpOXgC8@;fC zDe=n)E)p2j_m~3y-g(tt$H%YbgY77c{rTE=2UertzDBwFjQTTg^K?`7jqfy*Rg(y) zkd4Urrc?4w6pVt#-1~ziiM({dMl2wvj^9MbeBz@w{ zNNo?@$)F-qW@K*UWhm_|7gf#A0TzOG8Qx`BJzJY9>tsNg7*4>+FrROP=(6clqZG3p z88q}1!`Ew%sIajZDOaHnIaO9hR;p_RI-j%3p~m9eNVj!wS((j=?=~;`fweR2w(gA$ z_x2oVQFK9>`92=-<(ksr+S<@u&T3yAejqZ8#^YM#lpnURzWIDCIEOAes2bNPP%j5r z$@J5fNL*uBS)0+vS2>pU&q<%yp~dH0ORKG0v)vXgHhRXJrZ-0Rb1;!#MdyMF$ZoPe zyo#>o7}U9-@7pe05_`};o@CJ%MVn{s5Pr&hDKe-t>7AmM$QAQDr0YOU824=TYoI@Y z@+^b8mp(ne35xKVrr(Zentm)QBgDA~asqOOvE(Gom~O)gIn<0;AHH=Tu8r1K6UhTo z1%m!J6lr*!l47xs_ATVfFCcH(1u2*Ku^R+574Iuk-&yR=BbE zqqlI{H&b%3eCsJq3DXLCi+YI0wTh8s_k`-cs8yUMLAAw3Pn`m_hVw+@30*ruYqAPX zh2@iP#uBv04|RXF=rJ=rJ4vNDj;}>X>FG~j(wbNx#uGcH5wsIQPabX;9PP^bZml(e zyA7W7!B;Pe;;1W-*Nfw4%6CAJr2wk$>r^Px)Juw>w5cY1wSP%STg0cKB{c5uCR>z; zLgVIEPcMi$PLhJYr0+@$YK*R9WIWlkO|Vl>#6M{-vOk&Z2WT$upl= zV&~9+SYPBYaIA0SHH?ls^%(;O6W2-n}=8`Zi=5}!ifZ>Hc?CJbD=+Sd>_|fBA`sg-J-?IJaM`3Vr z--s_iCsIJ0XQSR92OTHzZaf=7@#Tm_#X3r!M3L3N)C=(vtV8IMq;i^O6o~dJu>!{R zR$&xO*@xCqE_CrOR%o4&u2}g-*ziL52iHqU9=n&EA_Nd+(jc>2yzagxNnaP3tY+r`$*ONY|LDPW0NXFinFZgZM@aKxE~Xz z<$>EdCRH1x1E~`aNXvY8NrAFPdc~<^(DotC(p9nf$%*LmW`3V0Fk@#$`la8JHEwKB z3M|mDm6E4XWFw_PGS;G~6%vyQ5k~u921IHWbWEWgph{>As0qh%r4nW(kPo?X{(Ceu z4vhGti(RVv@Ofm<0~9nwQ$2=Pldd+rY2$Lo;dUYB9V0k~M$2Uj766xsa4z~O^-yjvOy1Yp^r#whMK2q!rNuFp&yER=dJ~T;e5U}l~Xw% z?MsEfE34dwhS+hOxC$Yr`?A8x1mCL$FOGTI@d)vVCkgCa7WehgL5pIf%aTgT*HI*i z(qO!#!XGJvQi8Nt@(Nx{NW1)exDFBk?N0SI>=oOsGTb;$pH)^o&H{BF&K zOt3`fqct(8-kMt})*_u8a_w%OF&GiY?-F1sH5OOr3s-yTRN$o;={mYq)xa^x(U$D;j zxb4hiL*lGe+#yYds0}V(z=<{0n>cZ#?H?KYFk%1woOLEUOsrp%V|dn`*EXI}x)N&2 zZi}>NSo+9N?+0!o+hCIftAeqO8E2L(?>#f1<>ELnR^DSuLz2lVS)~Ln!>cj7+z)-v zE^%P(gNsdzF?7IL5E*$NmT)=!&AYFgosu_Cr~Q~5VCR&$aUA8cg;R2hk~dQ1V@lH{ z>6d*WF~vXP$yLr`z7`d2@`cOz+XA5K4QnE^RrO3Ky=<-)PcHYl<=w-{Cw&o#p?w62 zgzA~<00_TC;dO%^-I6^aO}@2c#pFH$b6Tnlp?RFF2-YI`a~gzd@B?I;I5$N=g@XFQ zSo~1p*9=Xya8?s6;!Qd)SOHPXhB3k!7<@89M;;xCvL|mRPwU+`mgt=OrbcZ==k)aQ z*gCP1R}oX_dz{%AUM0NYS1CRi&^u?H&ry;nG^o!6#`6=I11g=WI=n*~&#eSt6iJ(X zyXpLIUKHK(ZiJG!T5%08F}yvhhto_WPu(W*gW!$KZQz5*V@ku~*+ApT0|ZB%ik+u^ zoBbzTIq*I#X%1*ys^IKjV{@Vf-$9KrpA%d3O zvpYzqJUK?B6$ZB$M-?#AA`*A(iWi8J~djqji zodd2hW{oiX4|7bD1exaBWi8N&29g7*Q2Mz=xi55U3|8sq($J=Aq>&X*r&z`<^TRL> zPFvUq1LtGQS_Y7Wn_@hse>?S^a0vV(Cz=u4&U^{dzcYV@bxEz;E|+Z7`ks8{qG&&< z=a!4EN9eukRLAK&RVmXbsdMVwVBb@>bDZtFnG?@qXXMTu_WYBldA@1*THX6(2|Ewi zjl*)=ET+vAO1_pNiP$v<;oB0wHXb%8!-^XtHp2_>abJ1c2_i5FPDF?uw<`ND&%awo zYu(mZ*oePgg=EhW<>d&2y5GBp^wX8h18zR2UAAk1cJTw<%pB~!Qd8k*7V)M3#-F7ctf>8biMyz z@WRvT{BK~_SkK57@!ghj8iV z$8}%ScDd=bs6_!({tD7SLzF5Bj2zU&u zgWf5w6E7Y*h$WgjMS&9glCj}b5A2rZf-)%-XJaj`DKQN(vYiY{0)T)zb3}L zXSP8!T4`rYb;P6@W?=QcAz4bcji%}8#xql$1=va{IRu1uQks*VXXzt6WVd{@)dJBU zq<2D+OB9pn58L2fM-u$6NXkVgM3-h{1sT-oL1jb7K9^n=q|ED3bxO|7&H{OelzB_% z;@wyP#NHuu4v_5qHv7-B4WwUv{a~>5YMD)p#f{-)V{IMJ2W2DB;mMSh#MF^XqJB{W zeL<8dt8-lil+A&%6E8Nech!Nr&Dyv{W>ZoYctT1#CXoaSmWw--{1!!SQX0tOK>=BL zU=?(MSIq~XZ_)!vkswdD)pzM63wqNm4<_hb3)!o^|@#LdqhQ1E|{Xs zCS!~+c9r2_$-Mi{_uh(l))i$ju-_+dCk#@IdMt$HS6svBMZ1e1(Z$)>}i^dHhEWj}|!?_F%RySq^j)_+{{_yA4 zqG@g$^|c{`7B6p?9^fQQ+RChrZ3tO93A>ydhL+MbbeD3sblf0e8VUP>VKs6t{>Ta# z)Mq|7H(56!+UWJef_RgDstWq-}1lBj&j;9o%fHDOY1Xpvy)Bv6PDmI3AenxGYuLQYqdL2ZEYgoH^; z)q|o+PN{6Ka_J;jgWrBbW3|MxMyHcMU;gorUbp5mWfMqo6&RW|L8lJFj9j)41=Ow)vy|YpsJH&lI7G4JJ|}oxnw9kU@_WVV>eJspRa8ZkMf)r2`9T zVmR87gQBzC!kGgeyGQ{ndM*Z4s?!p{I)jU6)ZD1GG&rw-c&t zPqPi}1E2nE!M~W7xI+IwOdyS;<$K)MuOeSqEPF3e^79n=6g17uDjo+or67G*10fN} zI#>Aj`y_J@aJpzz+UxheDAL1j=ot~@K)toL-}_K#X$Tn16H~J|gC02jpw=Gh66q8J z(hoh6>3BzE4=9!93)^KTPM-sCX1nYjy^McdHdNA|DC~w-wMHJEjBSj=w{MqW#bph> zL)jFr1J2y@;e#Hgl;xg~RN0-FZemxid(03UX-`>fKMiuWA$IuNZ}lk55K|xj`96~6 z#!-}73pkZf@*;{9P@0PQCqyXMh$@TN^>Khx!^6Ki#a?A0iHqI@m`V%DruT;)p_{^& zNaB53W%XQx;=EuH{I6AfS(K$RF>)LV-3hf1dsXq|qUg4?E5g}f7w6&QX(P_=xNUrl zl3JLUz2eFoJ9QuOf!#@`NK+q|Yd`ZoJsZS2j4Khs9yrXBS>DeRs7XzNi|3KDJHA$a7H zFcnl48|ZjaCiplkKiG85qdpZ^cBwc3%UIxlGBWeSbr%VevM_OLRJzcad@gG+5( z3!*4P9KSuTxqFq>FM(gqT<4gk$$pL{emw8JeQBOXU$lhyXq5b6EO5$`ExgtfkS_AQd)#x72&1!l4-In=o5C{5XS&>`uO zZJBbAyOMNsd+8=cT-ZT>gSrS5%Nuz&{K}<$qP~a+!4;fALAi84_atBMRnK&b5}jO` zJId2`bl^w1l~Z@IGeo{ygunf7GfaNn@RJ4PrW?Z~$)Z$e@kIFmMeb3W<&&^)xlz#O zRp^@@iKQ_6LowQv8G4REdEZOmC+lY=hL^}LtFA=!0(TTHdT#jL09{yg$WD0PBi%#$ zI-+oI)^t6P@9qz+^4`EX%RMu(%KMhMU0OJ$OlA6xMCPh`cnhVyG9%Gu)zAm$bdy_i zcFw%w*H33h_Iu;KbPCfw4QkWc4A09I#qpb_c9U~VzVL?U@E)C_VjfC@)j`4eE>0$c zSK~3$efvOm-SoRn_sa~$h?t`oWMygCo{YbXO@x^G`M2Nvz&bnSwyu+nv{??{C~F{Q zy&04A4C>t5F@ZSs<;R=Z2@0ow@6&F`3KSF1-um^n0JC9Ityxh^wmvr=5>R}PD2phh z5UG=4weQ5}xt-{;(`Y;mGxi^UC+C|D&pMk>Tv#K> z6?6%jX^2MP!4fUD?wS*UXJ>V>zbjhzrmNb9Gq^nW$_*66jyGMQn}90NpE; z9~nNf@;gZ!CmM^#hkY#Oa1^{gZAn@QRzXoetpn> zu#1!JT_Aoi`-ayE<=OG>t+sQzfq{N(EL^*T)pSFpOaJ{vYh2jqZ8tHzLHfna@dn+e z!?Ki^({ZT@e*BTyZzAZG~N{l~zb?1Zq~$Iv!E=usfT23B0sASG%|cnYTlXqWL!*J0a!7>_z- zxD3gxU_7B`?&{x0nLQDK=Ys8|*o`A{CoFu6I!azmkv){ATa*ehv9+WNsuZx@S&LON zeL&ORKxcZEG7uhnsa~ZmjKZBJCa8R2+LRLng8A410H6|dOVY$pgCOmfrSY@}$ueK0 zBfRgcQw;6EfFIHi4tVH-Hv5}o1okWM`;HY{V}#JCxEk6yMsTr%(9NHH@!51UgqFND zbv4OwV+a9aMnpq)QSzM>$)hx>vc0PPN;DukMN$xC0iZoZ`zXe?s7ZoZoHh^@F==}b zl~O^qAk3aa%*|K;qo?EfP>rR~)-%{_-w#|4&J2YNjj7WRGe(41U(O0K6BjG~QU0v+ z?$!#51W7@;^cr~=jY5u%f@}OXUQ*B!WBZwlTd$)gqhx^a{ZiZVyI zF`1NnD@D>M4Q6c&>ZEUcxu8q&31n)HNPDBJy^5lGqp>Kr#P5WthCT-szez#0)3KuS zR&-7HSe@9X42_5#mLPjaur8X_Qa>aVa zVUq)L->?#~(Ls&IWoQ{xANar@wag)vY-*#@=F(gJvQ=y3<6>dVBgzU!_M>3TXBND- z{E~IT_&EWFO|hCGSQIFWqS}C`ZU`2&lXZUMi3ZcK*@=lUk1^XD%$fF}<5>@!VC5UC zdP7Z`5VXC^1m~DmafWcH5eTykDZ;Axw*txnvw@#w2TUJ5P`hT3W5dFr2exCj%eMQ) zMK;sx1lVuX?YBze8i&?S42+mbM%iRLz;>L^YI>P4?{_KNt^HOuR>X}9xovo~Ko23z zj6A?8mmc&#raB=ii$0)gS7iHi`&D}xqL0k#_8av0g!Fl9hs&vvJ9Ruvj6;C30}x|9 zW%7ns3jWEwv{_Ka>mc>dO|aymg_qkz$sr$boYLemm%i0N-vd8(7pGIwF6$;+loerm zuNL)n?qT2iv0YP5x_Wm=o?n;bwEA-%7U;I|urFt(eEGs z7wDhEA>X6wi9{iK{KueJtnLkPdVJsZ;IPDo?Hf7y$l&q+_tR&6om|N&7i=OaK|6px zFz%IXkr5a#4?=6}^|{4NJReo@(!a6XD_^Lmw|M4ycF`DSsqoJiVv&7H&~mS1p}L?w zz#)bcQNOH6aFle5(jP^A5SBQwWeS{(0+vPi6O3DP+O3OM+%~#mqpD{2H*PLy z|B>n6sj{q>lV5%Sn87BBB-zH9j@bXP42s>j7|WM9a4jnbm-@wt_A?XBGtH7^VTZ_| zhx=B=LddCZ3pYTe71q3I-1e7R#c7ha#~TeePZx2#{95pn;EA~Ct(#?=M)EfDbqCBIIQPD+!&?SMRS2{cw5CZB-tVT(e08P>y{ zQ7wu-aZhw-)bS`SCV1lcho&E!G2ju;KSLLLX)`0SaTT18?VN^?CW%S+5(yakeQLsc z=WJGO@G;1BiaNSatn-1yP(J^L=a8!#irQt>UMLS>Vty%$?)NEw=GXJ2jm!Lrv{77YJ2^{lUb^{*#h zO@DRMcMa;)pRN3UDUlXi?2`mbiI(srSw=_${g1c*Zcrl! zf8o5n%GE!6`?FVppnG1H@~=$`@W?Js|F_Qn`0c;l9^+}E5T$Frg7i@GGc?vp-Z+lCd=CpQx1ZA*CXngUEJ#tSLY?l zlQ@}l8*dfa8QBrh74jgYeeRvP$3WI~bE-2VHD1btBgp`YgbwVe*DhqF#%M z;^1jd-_Egr+HIVE+_o8c>qpBpuMV%3I9YFy8aG}m0a4G0)$<8Tew-plC{2zE`Em60 z2A@XE8W-@l$!o%E1gl<$+WttGTt0yokBu|qLEfWFQmFX1p?0H*{Rxp#r_{F5JrPKVP~qPwy~(wU z!;(32RR#@-=-J|4v6uC`lGU$&$R|G3s za#T5fdQ!yCo1VVf$xo4`Z zSpcREwI@=7P@54&==F4>P+RQ7f%aWN!x=B(NjX!z#S81!-`o_{Z^nkcu^T&EcX*{u-7%KB2sbXfvr!k( zEp}MK75+P z1LGJD=rEsO zmhBJFDfTJK15Nv(KoYHSDhyC+*i{)!*~nCi{O9Qk^brO#Ow?LK<~)P$=aeuj z=i%SND46;Ch@P%zIza5Xafr_W$9U`=gbu!sw&vqU`yZjfp)GdH6}L?l#~eMf=^JJY z-Ff|+za;yg8}Im%#jDUv$>H7j2$%5)Y^?r;3g2jh*9B2Vh>o|MuieEtFNm9)qq;}$ zRVGZW5p)SM=<0x-oQ*S0ef6-qLgL=5^BO|p__rg>$`u)OF>`}%^Epb6l3U)lyfY*7 z`Bynxq4+bK9OYLku^%ooGMS6*|G@bxD2mb|mz{yhg;y$$-zE5fY=if%CcFbavy2|P z?A>_umYBELy&}cG@BAmB8B|%k4VOvcXiaQx?Ak)>{fG#Vj*_QSWD})9N>wBwOy**A z=88<4t*Y=pPO_$(mM|D0)~0awD{E-|w0@3@*V~VV#1R-gy4rPHwoN3p`fE`>!!y@; zvXHEOZmbSy+K+%v79|H?D4o(^+TbD4H?*IC=cC1}0`Au)3eO473EE{_xtpPCqEUR+ z3uag>8!v<9IZK_H&yZDM{Y~sH`1S{Pzp!SabKB&j4Q{!!%EIXn{10+bn!Jto z#k5=g10MCTKJ#W0YtxWw>6PG?plf8!w@4L|L%F7rC@+AV#(`x`a9GxpC$Z zWg}2_U>>1bwQp7_bAiF_yr@#BpV}@eb}gah@aQ@MsNri}`8R&yG=a4_VH0I>V{gTV z@j*6*RM`N02WJNd);;YqT;|r$ExucV zuxGY`xfrx)d*qs!PKW?u8YxM#1QL|xGx~w5F3xK?bxUdtz%NXd;l4WYGH$*Q@tGNk z%w!l}!lMTvhJ@`*_k9bauo1VD*$Lth7*XoxEswnnm0X}k3%Md>&%?Dc$L0{zHIB-|_BAa$M9#>ucI|fX7hU}dvppKtaD2Sr%8lKZS_>QXP}lLgzYen4 z=S2pNIo!N72&@x`eXfm(=Wm&T7USI-Pzt>w!yIF?05~&E4Uqd(-DHn+g=D)=o?;+W zdnxEr(9nifiz&xeak}i*#Z|%D>;BDxK2kdir~5&VN1<=l39&If2RlaK zp?l}b(5Ll(1iKM%+f4JWy0R{RGc?{fIpH)(bK`aMhZdO4r{uuKkv;51WKJIRFw*^# z!7oh-$^_yKEta$_8YY}-8!DnOmrY`#ClNFZ8>OgJk9<^GW7J__YH$^{S)uq5o=0w3 z1JEg-S)q7hD7Aginv^cXVocUj@0b~j=7=-t+U%<2Bl4?*&juBU_9|P&NCxDT;UDeMKWU$RaCy=lTyyN&zs-Bsx~|q{`D#CB zxZV-8Mc+Q`*)F>w-#Y%hq4qsXm*(e39SqyA-e_wDx1DI-{Ppo;*6F{;BdTs}4Q!+Y z2j`kfeOtwOB5ek}FSIOr&6KNCp-s$`7B)>3nErKjK;t2L>_zhM-0|>27jM``a>BL~ z)qgR2Awuzg#*iCsyo^b-Xb_8=D1ShaeoFH}*hU6@iY`e%(Bqy9J*Haz+7Ysu903lX zUb5x2IvTZ5FUwGs4|&6K{7@vbmkdZ#g7ECh;DcOq`{B7avN&B5khG$k!|Uj@8Sx>; z04#Yu8QKr3XoDW$#=NFe;HS0RZQ$GVl3i1BXhfWTR!R_5nO6t2ij!xqoM|FU-7h>V z`Z5~0>!#;1hsh1k{Fzv(j>j?crx)(xq{?#X4(Z=(=+9oP_B}c;M+KWl=cUZA2|pvw zq0jhc(D&#sB-m-Yog{k}@JszRGx}NBw+6Xxws0GUovZ)ahTABHbK*j!cYcr=f;$KQ zy@F)AF$AIiU_|t%n35M#Bp>P_mEC}Wl%Ot2wO7@A@KsB^O1uU=uKVWm0TwGKXF$Qq zpt2gW)V*`6=I79*VR<4P+M!%fyI*_=`n1Ob>O!=Lwj8=rSkG+^aKWP5S&(tI!D0uQ z=0E)VNs$>c(tmutg(N*U5q-eZ2!Lf!@+}nEOldx&4Kca&CjWJ04e02tBg>&J(V)II z=Yz1m=xW8%xdwHKUrl&=(304j+!ZldoDJUNGwC^iCuc)v)cckF_NP<6S{mWmbiDsf zZq2-uY2C1nu2dMJ2RvHg|DS~oc-*AxX58eSMaDb-a%R9|9}myP&ll*+a85DZ6rB-L z=ZtmqsG~8o(*dYFoTzNyb?ir)-_5=7)k+Wc$nKEyI~1OHhq|tXD(iLBvsOlF z1gaM7(P$MP_Dr3uZD$(jy}@6K^j@t%zj9Yx6JASlBQxmSNLLT?qbJRBxF22U;w^R_ z$r}r9U4A-GlpB{V+2D!N#g>L;tFF!I4k-;oGL|}UbL*KJURl)mA!K-Erg- zHc|$11_C5zKlRvmzjuMG8!g7~#`Ps2Z#Kdk-A2hFYqf>aT$c3#<3W$?%V=2g&@0s! zgdl-O?-U!wLrQ&7qD~FyB`TMymmNQ4Rh21wi6@|Jqy#usJC+J zPTL^q{_dN*zHSCd;SA;`sdMA??*$7$G*a@9DRPX`oP6b?=>01|wR_V4ieICkin=Av z5)P0Z^acJ(aTRq|xtiN5@8O)56#DL)bynF==Y@9qRZnl^?TKlLGN{{l89_%#Bc~-Q zk9qHk-``r)bqoGl1Z=+Wr)I%v^({_8R2O}izV|W^?fj;D!GYJhzMijH@#;I-FWnNi zORK0g>MirLgonY4X_VA)b)=qyH$2Qa$hJ$!G5A=a$*zy=&}ADwql8qb|LF4h;j>=J z)GCIW1GB*0VNauYe7%xKqhl8$9-TgJTdWW<@qe<;AGG#Qo|9yu4bKJL+zdJ+OpC$a zY}KA9h_xXdh<@4<$$0xNc3}vu#<-!!^hHbOR{FeXo_eZgy|auIJU4N)qZSj@UP=x% zP`fG3!ZKxr@}&R5=I~Dh;H^NNY!RIsq$5c|w;`T3yp1R2a&|+AwuWvK9Ab{p>9Gd& z4bsT_m{+Gvjl|D~glS?tb{86=L4TfE=b5X*``n;QAlQ4H9CDf_jos+6Yu<4l%HG-7 zPJ0id=NsQ?Cac`oamlm5LpmkjM8R{_q{^<*rBk&gBEWb)OgVQ#?u0-U%AwGmH15S- z3@{wPA>@K+{OGSc`|%$AKF;3DYQ~w6_lE4CHEW97hD&TPqJn%wa_F>>o1nF+Q{3=7Lf5Gtj8pq5dw`73#KH<7)RaE2P1f~GZVM0_ z6-g+{3@!y-J5QS)do8jFm_n-qkWGEOJq$ZZ;Ktp>-x}-sNVlDIY!q&84s8n8o)@G9 z4SH0}KOw@Nj4_UugHP11L3Z%oW4Vl-ZMosu(2#C6#6a?O@cSAogC3wkQ@f^}nN0dQmh ztj1&Xk7FDORFNNXn9To_BL>IM%)+C?Twto8v?K>1<6r9^aMe8 z{j5Vk{dbkq^SylG6)6{7szS&Vy38FKJB}X5 zfHU@VycH2;o!)Uxb99O} zQb-K)BEfR1g=vx)fced|V$#$1y-Pzn{2?!R-+OcDpa<$pHA!m1SNq5FKYeMr%eW%# z1=r}#6zpY#7#8Oii{`v*T{U1sdL5-N^7%#@sPp!On3kgB{*K~*nLUq^#S?JasjS{p z_jUQg{^aLvUGn;zI3YHg3R7iEBO9gVfkVXty9sPyLk=n@>?W)Hcavf-odVN>F7q%R zhoC(eA89Hczk_S`yw_8-^{b^>h+-J%pSH`Y=yZN9bCbI)GCj6NxlP_A0UeI$a|EH1 z&fiHL1XX6F_N(FC_wAz3gd4@3q?_kj`eFO5;Yg&6*mey>%y;!WRv%`~&*!$@sSU=x z93ZQ}u?12mh`b`HthOp@o9et^PmDHhYPxKdtWu(_4{VopN)CIDmvvxk8eHc)SmKA> z^k-PxKHGYhv0iPadmZ+`4M8@ah*<~2;`53ZMzJa%gs_l*aV28%UJ6(z5rNExNMId2G{(FT}SGoz+{@_dO@> zs|^-HWWGq1890-cBA>f4GFDnJqTHwCcPY|4EDG2x*v=VJ zxWsU3tGM4A3zM)UFB7DJLX^m#IJ_nJDb0Gv#;0bv#TaCj|lR+rH3d zP~xocN8{54l_kmCF4_RXfyJD2kTYuYY2ple;4w^Lei5=1eiwMwmm#;PkUDhi69W6z7=dgAHXu@IOEJ+4s%xd-Xui zPsuSihF_{akZe+0wpU$o0jqZbT6O`G!ZxD zsJ;L>N)vrkZRm%hMF?b!>D~%xSFG_dM`SV~CZkjIOAZNv@d<-8m^9ZZ)ex?E_xC245Fk4-l_%IuDfXbfptJ-s@)c8sk8Bz{b4naJ?7JlX~$!~ zW#kT4wrxUCQ;F)URqBGk_!yLfbc+3?E>s`VE{DioJ|D>^5id1@Db1AJ(@<2x$CxL8j8O*7VUQTjbsxW+UBTOpe z{Mi_O1*mx26rW`4JE6!hJT&DmPw| z+bHu-ehvD>7F?0-S#VpLDN7(Z^hV)9&gXMnIr=sR&^}PuxY&iePrs)Argid=4Jh2$ zblAv&4mBvFu2`3#mR`!qK9eTiN&%;_*gt^oFHJI5g0KWz^2V{a^eAY-sp<&|NckTX(2Wy!+nF+ zKnXdLH|3#h6C$UNpv1b7u9#oq2Y@P2?nM4tBYlwUc~M&iYvl}~tE<>7N8DY(^y}q+ z+-V*8ciXI?jadHmH(D0Fl|wcDv}M7c)24KjZaS6vYZmp+=~ok8c?G%F1cAI#>qg175ayhv5=}Ya>^r^Mvq<^*7apo|0HTQ^g z>P}(Ct__2NYY~E3WU`_NE!OZ{QG^9x!MHozR)xfA{{EU#%U=}dLhVbkV8DS5A@;iU>ltneq%~4K`LyF(4pFG1C0TcHeR#fm{d2Uy$#)lnK03r zE-8z~JK_a;y59$VsR9Vtu8B#H9rS2{W*G<)_xr^0yC5{JpLLu5SdlKg6%Fl4Lzh&E zuFFu#R}ZUgoIT^_Hc2orjoTIewcv0WrAGQ0jQwg1`PXNd(bxNiWGUJ9+~l?EEbLh+ zB?kkylhR!AE8v$2F8IZB(}Sv@K?>Q^bqd^5rfifL)VqLO5F4j(k4}M=yxV;sQ2TSk zj}obO4eC!Izcu>7_ndC?95QgD%zIT6RERl>?@ftow`-b4*o8{sM|LFGYhQMv5U;g@% zu5`M2l3DWB)YT;8xtV0DEg-&&k^_ZL9;NwYs08;O-4tHK7}PD2o``y;j@Lk!%J0$b zfciSfRqF%a*&KQde5BU!(y*P(q0kkQ6C{~yV)TjkGzOGIC(^aqn?r%uhpu;xdJIe8 zKYRm5Zajjk^S1uF{xadb}D4Y&26xb8En8Ie*jQ~veG=7~qPD&#xl zq#MUyx-6!Wi6i|Jzc?Q;Dwab=y#m(JKuFW>pS()zCLGM>) z2x|nr^hMEDrA|>RfNFO9&H-s2e_yDX$tN=s)6&pykSp6s>1mxxKWmrt#EaTIrjOJ> z$yes&CsnhsL)gd(#~ItU5y}c-6LSLB(vfC_N&lwZN_M$1!hj@ugg;(O$@fvDiqiCX zH-(#)L)tGSDM7fty&}`*3o#O-3)<|zmeULrwt1ij1tcIreI(xVjx=4?J)^>37c>N< zVFgu;knB<2_thytd3A$NBV7=DU)IAh6)}|2+IYU6-T}-a$bIXCA{SR7$Tl|YJHiS< z6V86G|BAKl%QK~mSZ0B$97?{8A{ms%pfHND1kPk|DkWGQhhe(4(HH?tgr)ZiNt@3{ z5K~T-?Nt^MY=1@?u@*(8WVx5Cr#gqRV;=+@KFcn9$K-$;1LLv#5~Li(X6}y*36lRQF8lf|9vys*(k9ko9SjoD%nl zdZg*lWq_)xJ)H8uTQgveg2h;S)OJxWjlK(h10Egp==Z^lRxha$bVV4&yQHgp4yjB@ ztCcTQ%c1fHS`Y9awsRK?R6R5q4sEc69V=8#*gs>*wP&6Cr(M=7ND*qY=_2V3Ur>Xg zjZxZ^pmykDyEf-ZYZ;ag)~?g-M}=LtbJ|I)P%$y@xH|k8==MQr2hr6 zwq16J+fCO_KOf#C*~wt%Kkn8kj`9wBL%SGOL14nMBchYRD@_Mco?{1%bjpJsDAmk% z=CEgI?L^FzClq_Wh~+t~5Hs-?1vBo8%|46Zrr=GI?8ZJzk%iBaNy(vfJB`xp3tb8o zCqr?#YA;<-qc{}^Txy2Gb6Wfu&p#%s#;~3dlrP&A*L|COt|{5O6D-F8PInL@?7xWR z73?6=nb)%6tED28$vyq=7e)PaA*dk6@#FaWMIWh(fMlT}=9>JLxQ$mRy5e_)>F1;f zSCQ)xAh`;Bzne)Iu-4`-nEIO)HG zNtiV9)6z*T>eFh2dOK%#aKa>HYe{5o_*Hu~^11{%#SWh%pi;JjSvtvi24FPs*$?5z z?7(QZbJ{*u5T-hNSNxe{_E9`SJmN{ha}%N6Zh@;*O1_RFNt6b=Mq0(tFoIQFTYZOw z%TVX=ir=ld=LKC(=mTIglq^uh=C)%#cg6Sb1U>6(;0F3s)XI5>LJx3^^rHHgiUf-= z%d4k5fWE-Bw&C%|Z}&JqeyL0Mu(B2tWEV~yvu1^|L8sIsdLUX0ToT48NVt9iGCRoW z7|-7_!<>EH;+gBIAJ1X1djy8ewM!qm#m(7`fZNtfa=!IW%tG@-vY<9Ll3aeqT0#ab zri+`D{0oX)r!-BH4%*arXbOm=%5KtK5@-*>mX3DW9{N_aPO(&-MYNm9O}d}c40TUX zhma~OleY2n^m&m11l*0h4rz9{etr*#mgpdxyfeC=qivOS0B3f;pbsK0(8_@^`(b{d zEYc2KKDU8-!b;@aK}REN=cU)15m{Te@D5fgSShHH`T5sNLE(Dten?2 zlx0{UGyT-Us6h|>szhe^t3CoR>4xtVxEtGyIwl?i=a2H+5R7|XK`cl4I7;PrG9<&xR+IPBbY=8S;^qyxuu!|HN z4_h6ZB-rH$EsK4mYChCza?nLCThJrZR(o{`5*eecf%J2hG5>#H_ve4s{POod*Szz~ zKm6(U8WAn`AQUxw1tFuOy+f2dwaHa$1%>Iw@-X$K_-j?HaQ-5@ju?xM|2XTKc;+6>jCZBTBbG1hpixP$VF5M0tJ<$bV*~D1CetcL zwV83MVQ>H)jtw`*Og>g@3c=Y5L|c{B0l2in0&i@&ObOcI(+sT4IFEIVM=Zeskbn!$a>N}t{Yu!? zENeO-w~hDPAoz*nUl8Vqiv+`eO)>u4(jM7?Y1@Lj1O|2Wf{UUuB`&bXEqcdxz#SMn zhc2=0n~vg)P)ELeFLf-egKlg?Y_JZZy=WEhQ*{&VIl(!orEZsPR#pfrgyV*shr?i6>yy1t*D^e2 z*R98N5-Y3izD75C@#&o&W|ws5-QT@O;wBN`-5SvjlT69iP-G>g>7duh_o}dQO1pSU zW@NSs@i9&wy~R-32MLQy#^u9v0*PYz#GJ2IxZ^T5C&Th|IEDr*R#VUNO;Nd^-}@ru zZhAz}fsC%lF>wF(Jt5(0O^zch3Y3PdS7C zUO_TP%Pza`lu2qV=Eq`6UPuActtNy1NRbb5j4g_um~hV?RV_=+Fuf)oc1geCxgWrZC7>J)X{3TZz79;7C}I>4|!fBrBo zsUyS;nTA)_RgvsTWWU8Ilv8q$HYlPr*lvs2Tok}I$U(*c{a4KX>Qr4ck^>Y*b;-Vv zU{{{Dg+XcnWC82pWKG9y1G81RjK3w)B+8%zBZDixe%f${Z}JeeYED9uU#LD3}A*S*D%>R# z<%=hh0ZQ{>*bzmexGb=Zmk^pmKkz>UrIQeXjoG4D9t?#_;kP3a!xKVpz(iCvzudFM zH!sSdZkOR6y*MSPK(t$WKelUXo%fQ+M5dfmKsU$xLX$li9Lw0 z0VbF3e*a(p>EEpRupjqx-FVz>aAcPW&M{~TDj_nb-6Xyqfs9Hn471rDBzs04?11eX zxaf%Oxn;(!oqMb+k!+Ok#Pjpy8v?<=;)((_46un5MQKYxl)%5q|4?XgBpgGoS6751 zMtwAPKJXE?chuWBeW&dQKFVb7zHi*ZgqkRSvzPJ4$qA=Pnj1TxA6obg`IH=Z-Lp{` zri%U`7AY*sqVGklkYU*~a$g|51yHq-Zoxht#&k9-b45u|##p0FljFMP+UzBgVy_cZ zuR>_{BlhOtu9Gws1v?D@x?P-Z%FgvorcEWI7Qu+Uvf<<<+^7 z_0av{D!_~lEv}n>SfOQN)xE`gWoEQ&TpRic+2F=#DYrmNE+vOX<4j6psxBGuz?uhB z6~qpod``P;?JND>p!nXxm<}MAdIivaG$3sir-ogX4R{>(G>S9m7S~E(js+-ApAkpg zkR4D2zPV~^uk|^pmA2u}B9tI2lp)L!X>WWjnR~)J2clxGGbkg5#%j|}1X*3>`j7v0 z^lz=!Q;iLa-p1P5trY+4E-gqPlOqd-LBsR$0f&9$)vm@9!rzB4R~T zP+MK6$X9+zUl7&!V6?s~BujWla8Pt@_QzqD!*fLGk;sj--LL+o>ZqJZSGgpX2y_6> zmN#(GC);p7p4FUSO^M>R#q>5vPLSKgRBQ)P5Nw&i+N5?_bwGyje_`^l9|JBLY#X<= zje$`X<`YkSZ|bKr%}%O^P*q8C*hSghcV$lwSh%#iDS0VHc2Sy7rS zWG$u9(_cz}*!sGEAIVb;gktHE*3;)->(sG-^1!e|#*^>m%%e9ZM^{^W zuFsS&V#CcY^=p~;fZGg0JbCIm`Ydxq3Mz67-j1hE2@UE)q3CUoJhbVRxBrGm(AOTh z|Kx(V7sDm40SY@zjA|vE35AFEvOOoYFMRjE=bG)vuTIS?Bqi+Zh#O~lPg>ZKdP)xE zl6x_vlC8pc39f-*e#rwt1)NP1g{6K*`pbxBx@`7|sSsJg;!a;3c0bSi~jVNPnTGWyW7xEKlS=W(Yse<<%HokUECvdgUB^+>%-RrHynGo7PoN82ivdiDDI^Dt}uV{`TQx?>V%KaN^WdT zY$z69=dKUW=YzIJVblrLebFkPK@Ut?rpa|aIigCwwnn*I7#C&?aI#zaB&e}P@{{i6 z%wx8ZeC6ubH3IV_v?Be|Z^@d`3h~_++yrErM!2k7DR~-2Hd30ws4{vT6ddYRH@$E2 zuFmeFulRi=XoZR*(2pZYvTokO1m>2wCwc>?U7AiU90xN81IiX3FlRUu1~%_y8y}^! z4f~#6KJCWF!$$e^E#4JpNdH1)ERV)&aXwVe1Ai z+HV^b+$OI?Yq!H~LuxksA)Gf32d?rNB5GV8-WJgY?c@nFmxKYKDWqtSS~go%&zw*; zE7!>T>Fk&4RUNc(JTf(QzoLTa4HUd>o42t3uxb zwU=Oh(Ii3rS*%dags^J~kX*&Zj?ctl9s@SWcsOk1u)zu$6Q-_|K6u&ebG&h_Ii2ib zxA1jiZ^U5XjU1!o4HP*@X>LqCDXHWbrH$TQkdG{ju4n4#KJoHN2KCnIT>>mS*(gk8 z>I3(x&WJ}IujJ$c^1kGsr3XB=DC-%6I*&O__LF+<4U!jpF0>>P`5FxBdrSepTe8J7 zPhCBIz@uBTJ8-~b+1&In$am5Wu|2V0lFP9fAsHcK04p2lS=l8!K01!kKcCo>?`sCr zztnF%OV+z_sR^_vj&N4DQ*xlY*-B}S#dOhW;^q7=qwhvE^Bbi)#V!u&qF}V7QT!lc z`2Qu+5~-2CLza6ba=UqnP~|p;0k!jyJ2>uk-re!j*qNNVH@{e9O^)NXlaUP$Bs3G5 zk*(ro{DLSbU8@dA2{HxYT@M25VX_kx)|a{bGdoa-C%m8Dwmd-eZg?_1!SO3(D?h$kc;hTI6`WB^4X z43`-Z3>9>O)0xh;-EFtm?RIyzvpXGjTX#F{w41fdmQSPF6 z8x0I9Bi=wFIu43}$e_aieMwXjiR3`SM*HvlDktai1@b+4zxR7D&%>xcC%65fbigMI z)`9bMLx@(f!R?_K>(XvW+Zmnoia4HgoM`Q>jbSBBMp+;1ixgHcnZjHC8P}YQ?E8Ut%sItv@4LfMxAIw@+!b&EA%juKv-oDK7Mz+_(-8AkungR+I@ z1xOCNv{)!MLhS?8F{q?5sG1vqdOAZ+db-dbH^q32|1FtY7Fj6H4QK$@WOCM=am@K- zO!+J942bdz%^N<(b>@-$WGvCXH2a-$lU1pJl0!IlHx=E+zsQLR+8uhxuNrDID~0R9 zQ$hEoT3IPffrbsdnx@2^ORgY0@P;SuTAJz4Y5dvceZ_0gDcQGKGU#+a?4)~#azgu- z=iy--@G@Jgo;UjVukkWt%FRP3FI2|HLOA6v6DLRI6O5pwE1L~6qzzI`Lgt9H$~^95 zY-mPI3!C{L@tEBYEt?l{1^<&5eIJ8!e{(jl@*r(E)}0|;kNT2ulkem#DkG-PS^$iA zWKs?-17NgB>z)t)yeinZzTEikts;{B(yT8KJ{z&?DWl}26e*^nb9^g>RT9*_t`yb= zuH6@VGI)PHC(8vmwZQx&1QjvyBD_Cs zW)=tU#SE!3CGHo`dGCD{VpUB*il<@96(x4;oHUz2cId$Gnb`n#ddUhM zj$2|Lm3(!&dD#6~7uA6?H5Nkg>GU;b8Oh)_tFpwayw317f)3?%)rSv*Qv7qpC4Tj9 zB>Ur*rdpB}5gT!NMwKU~WOlfxDl+UJi)RmM!}4J#BtIF}GNjM`PpZE;%_p1CnFHg; z0s&|$Cy!I(-VSv}ec>CY71Fon?&Vwx(Q!Lu?La($*}EFLAKJRs$lAf*I2LeO+y*3R zxeJ?B8-LKO`s}S{Rq59g-#M$wfCh|Y$w_i3pkjet+~V|z|Y(6f!7845I5 zrCyn#jY4%hkY??d8v1kUpx$l0bcM7+UMwsGB0gvot0Lur)%42H5+8U3>1W%Sd~Rph zV(TGg3!IF^%oxX3BgwMS8EuR7=L+AyF(=x4cDZug(t#qy^;$`{jdzj5Bb`>zA$kv0p7`7_Fuu8+L-2B^!# zw;1H}#(p>??zY~_TJ_0Vag&vxuwo?J-ogqLj$2LKQ2%d!zcM-}Z{;8HCSS0Nv^g%^ z$!ZhY-~md0pCY|f^xbJ`ei*LbE;%jk^N)KC^MELvzRm^h-7c<%IVVAFjAq5Q`LNO7 zHh8fs!#_&#fGzt&Ar@jdNTdAQ25%%lc zvo!V)t6Rl60v*U{q<}I46lZfkhgM^JJ>;~?Wd*Mpl)V$ZE5Zh4i4nu<9EPX0iZ25O z=hg;djNz0b8QwZ3Pzww^pLyzJ%l$j%+~BNpSq%R>5|{^?xMh(`1bN}H5lAZd00!R? zkr~=XW2;$rgjTW6amj+?|K%rLawGU$AM{-yTiAimfj3k8OuU;cN}f)UG%6a4Rrb%n zIXlY*Cu^H;s-jbP*{_b<X0*XQQOx7 zTWM`ab9rrRUd={(EwskpT+jMhZ`XnGP+)?GR7wt{u*p<3&T@VDJ@JJFcU=oTabL7* zYL-uG;OeP&>0;qga-CaAA9CO4^Cdq%q*}61l;d7CIV7>;0fO0_+2y6>WtZo6v%$P- zg-v9}ao1=|ggP0pRqA1m^Om_mWt;%CiIbrS8_N}GxtHL>KU6Td5x|&%PicP-gPEcW8;vhaR-B8J{Bw$&Lw?yFMo-tyx+FxhrGsJ68>Uqv z6R4iPAq5Ey53M3;*3cXcMB%c6Hu?64ubJ09yDuEMaFE165B}z9@6+D(w1Jl3wD*c| zt=oXx4e4;1?k3*`ez8vt*q(H{6=FCQGwR%PM706&oNNx>hC$$4UOlta@uKcw44?JT zN%XQ#X`CEA=La;pMfs9DVfix;hFQSaC?G2!h2H2N8<+vHv0(HL&S%i4c%%3!ecw3$SR~f@1{fV{A2z^wn+|bq%3TcK)TYWC%h~& zH58i@PdlF{d**8uMZqVag0dOpbdLJO^V*rcE*(k(#c`!Dfp^P&z}2AgTQ$$#j?M&- z^1O+j@U>0y&n}u1pubStBnzbI-SqjWE1)!)$^o&bsb$JuZf)T0sG4t=ey=xb<gc0C@7|Ly_;1jNb3}o3#-U(w_^bZoU1|P*N&U#2|xl?p$X5t#Q`k) zL-H>M+dnogPO10nr;^(aoR?f@63iQ<otooN_u#M(?|GKN%|3&9q7;YdkG@Iaz83@y#F!%rjrlY)w zfmj_{jh1+CW7abJCqr>$yavglC9%;3Rv4MG>PT&3sL@Et{-EAL_BgPS`qacHuA=17 zOj1TgqtAONqK>;G__AMa#6zzh=n6RZqkNFP{GYa{FXxue7=W1YPWNJZH75h&z0Im^ zzTNaa8QQVE5tse2jQ1ej8xEY6pkR1Pn8(pevSA@a)3M!MOHk9MftyKVV|_E{vLD8m zZL-jf7eUY4RoMbWtPn)aOURyLP8P-{F6h7+E(3t^Cni` zR>%FaEj*Rk@z-TWus32Kr;t`34Tu*RS|u(lI1BA=l|nV9$#t^59Q6rh(kyi*W0Un} zY=E)H_}F9G;*O0y=^o}iCk~rgu+R+C>A%Hw>2y8cAi+J%lg`cIF6Eq1?h^0io>C+! zCux$66T)JyZ83#bkHN~GQQxJ0G2J{OM%nHAA!u0TiVR?c^I#<#JXWB{k+&$a`?pc)`@xuZvCJUe8 zh1GM<`bKtOAtC2~W$ukQELbeKNjv)k;qd2?;;urCpA*U{Ko@0<%z%ua? zAFaZ`K?oV4;n#_7-mP9!2C@BDXeFo%>%5EGQ~A6p=<2wZwF2vS*ZRlc>9Vt zlesQ6VZR(NQxXd zS^}LVBRZjLDR~t|%BkoE=>5l7wK34yI#Ea(v;UyR)&b|PUUWYt3 zgz3^|Zc?^{6z~#3dguX-<)9NmnYDl!WeFPl)C3@988D-SWvBd7%Jcr35kxKDyqQT# z*d?tUINjK20=YU$UPF;eD!N*-ig(WE8fcHD(@^0HTN^0TZR5w28)2Qo4jOX>x%A-N zA*a4K^fQnsbsc1`lSqXNk{X^5yW}zqtO2(Ko`%7Pw2G?0ZvPTmoy5`4z!s=eLd;WY z7ziQr3`;x@nEOgoydB$@p zTu}LAC;39yJFAAHK2CD{Dus(uIn}aw=%ZZkdxxC&N`F(`2Ql0p*)ce|Pu>e;LU-S= zGv(L>Ow&X_Fy^yOpG)5lG#Z3W{($S#>(DJLC|+1L|#$!eLR`u_9C5670Ka@>F-WQIoiKTkBh zF-VW%h*{9}2Fma2EVH3Ch= zGJzrUXiMBYom(3q=;>fKJhF_SMQvZ6^fTI-U+<|~Pj-!_3gWC#6xV}5l8XML<}E0GK&g0eo;mxbT_HK5 z8V(*`3AyI6RSw)$*Mrda(Mvk$T-pxmhqm-oY={{M30QPXqQBqyzvhfF7H=NKmv-O{ zc(w_iwo&q}6nv|r^CXS*I`5BrWVQ2tYdCIJZS}ekj3+R{ch37u5ZtQsLT-2a*p_Ff zmbJBcb_lzisVeC`Br@8RH2&sR67RsiOP}0yg!ACOI_$0uL>8f?_FM{@Qjz;cq`54hxL@(v4U>o?`GLljP~sxto^Lv z9V0-Z%J?0mhTTruftRBzCU9z?@GEgmcOGz@w{S)hf5WDyYg!J!+Z1_4G~&X$0T#76h~Qnvdm_6U5|H@5D~NPLZb z{La^M@_0r}1top|7+J{Gbyr*imntk%+xAyf|F)-C?At=^oCc4>~b6KQQYp_=-w%9a^E@QD=)}5 zVee!sIT%{xpFq|!H)lhnpLfQ~o|Tjn#0A^ww)%d{xERF~o}zy*>(d!VtSIARtH=&^ zUW)^7qz{=OE1!~s1W6VZedn83#CcKIm}XVU>lGo@WCO_y*UIZ@e6@j8lV+7p`Urw} zIo!u*c8S|L2A%3=RcAy!&^hODH;~QVdj1)Ky|&sDM3>zfmo}YWu<);|+R|WKa8tBQ zP!-rogW^%YEB0X5)91*_P^>fFMi41^)!Y9BuS==cUut zyWAUv$c32~S~W8(QeDa2<-Uz6mG!!Ildrr619wKMvlWAZmw>rUAC|2EXEc-1OjD=L z#IA)+zNr%{GO#?!R^jKl&Z*_GCvwFyey-)deac(_9Hi$)gdR3h(3?f#sAx3DEf8%| zUld-11LUw(W9(sApKg-Bt=50g6ZzQKKWq7e9?cPSSaE<2-~WBKIV#wsdK}ocSO9l1 zT@i9Y-UB2;7;eo_>YykAcE+5(@JU8CV9rW#vq$FypZK8rkN3F7eT$R7(=rk>TFAtK zcPG0{44a0MZ=lF}gM720l(L<3>(!l-3Kx`Zju)Mk;;YG?e=BF-Zo3UjAhB%Pit9tt z&AD`65^G_B5*Yc~^)yD$azx8W3Oz9YbLN<0&}p)B>@*#>mgzA%sLeUi_HT1(GQMrGEG*_JBmkh#HW#YP^T=AiRjxeqA zF1-vmE|I(iX*VkQK4SgZVj~gd%owJlz;nFD7Nt;y-($dJ&Vr0g^XV25Q5fv^vX_kvYlH!>72(LkU zNtiYhN7X8jdMcjtVH>}8A#(eoEIEpvB&j+>=5GXH{aYK#NIE+ZI+_D1U0fZ%M7Eaf;T@yv?TQI2XMc2yV8huqAIh@(JIc&N^G3m-S{z;M_c1W2Ibtik=lHD#6a4N7z21*Rp}0k|)@{*h^1wA6LU$*@b;%D?<4s4x4=L4KI7;p*_&Pl3BuQ5u6E{eaGOXIaKv=N=#&3{z zK=G7jm|U~O=dyTXh}syeLh*A9U@bzo%21qj)H#pBKvtW}I^J2YqfkGgXX;5FNBw}@ zfEJbQ9uDi;SjJ-`XJ7C{ilQ9Y>9J50g@xP$ZdXH*GQL?}AXK+7UCtV&I&f9!5E)rPvLyTM0&Dniu7vIO4Qw|c@; z=u8?p>TnO%Nhfnsd6TtwXT8AL6A~lF{`uD`-+I-!WcmH7;zP34fh$-_O?E%KDLL@C z?4+X8fxH_kRSa#XxCj+^bk9nb92cO4fRZ<#6K=x=P?aI6p#`8lvn{lTDW97dz9d*b z1InVp&cR}4x8x9h)~qMr#LpQqTQ)wlyqb*S-AsAt^Sk%W8GhIV{2bVjSl|O{kzC@c ztH@z54XDnb{?{Cxh4qOLCKRTvIT{$4u)zKHOw~PUkt>er0Csv6^`n2MzhlJ5TZemoN>02q__$$$j|-Ii zJVlxw%Z^mj>eB(M-8&(m-u^~W)PU1dlY*-F8z&p?Yy<>)Z+lxNT7}5 z*g7=c`VMc|`$lm6-4*8^a?yboqyZDSd`Zb~Qsg=njd2UCz%Q9K;65Bai5DG{rO!$r z$H}LP0U$=k8{6C~@6<&ih z5iD)7ei?K%XsSWbWrLuSml*zlM#T}l0AUu;BdHBUk}?DPO9tIRW0}No85MR2Ajb<< zd~bHsL6=FSHn`bDqC#DM>e&~)F&d>KniSAb+U~i`9|#mcK;ki# z*r9-m1?u7J?unvI8ZX8Kt+Rmukrfka4M5Dl(&SNWFAImrsa+EACpZE?93FII5X*#98%9(D*EJt zOJWVPMYMfZy+2Z;*70@%pF*>$nqTC3kp2=p-ZRo-PAka#VpC8(eO7*qUo|a{*Uq`* z^3+>k`^i~(|9B_f%HhX-8!LlC-Fp4g-icJbIB@xbg{qee!R79aG$a6B?n#R{s|6|k z4>?-sDof^^m()g7Kv~Ou?;GMlC*)CWXQ}~iRh)Gm)eDNllSL&^b=6B(FGv;@fUDfB zIxk!v{9zrpmc-95;a=dx&$l(|#tX717UuE6xZyPVgUXR!#-QBEeXUg6^zwjE=)m{F#P2aGq zP@4Hnu)~J=)AoD4_PAIx--r#rmNPrN#C zxu8(AnXBh-jU4@~hB-;FX3B8&PuG~g+>_T{A8JHL(J!!4h_iClNZ=-U?KiJ1+1(K^?8D!dO@_4I{_$3T0=Y8TV%BRq-P)%%&Gc70@Xl7^vr0 z%zO$~d!x^KoDSdl1g!tLXzp((QXJ#J$p{O@F)fl}n0rIV0+7SxNrq#@yWK8wP-c3V zL+I(WJ^7sMcEAc4cKa=!9XPr98CoC=v?(1e}lGEHn z-sRp}MTVeCxN>Toyl-~e%+KWm(jy@5x@Z2Whz?K@uB2~>Q(Z1W=U+OVL}|I$Hg`hW zD_KL1`mCIa4J#?bcVu%)Bl6wz#pw%DpvdwSKGq z$D6;=4IL&U$El~c`R;OG0}GXgsid?0x@5=b2FOH=VWn7tisb_GyxT`0dz3@Vpc5N`CMBZQd1O zwSn;>ydPM@!7>dw4KT_4O}_2GE&_#tI_F*DnAcVWLupUDqFB~1QG*{0O4Ap(-L63R z%&nwRxYF=A77rp_1QHzd&3+`sdQmm1dqLiGwP2|ro5`T_qAF%&(g{SJ?0=VDD#)N$ zl4VrIj6+JO9x{}duXP)8+A6s<$M7etbAL2vfIGk>@^sGd>$Cxj8amuDfkS?nu~k2s zQ^=_eaM+6Tw?Dd3`J(4Gu_*z?$k)25cexcpMNW(4g0&?MHvHOKS-@x0ceHLWRt9Mb z?^|U>l+i7Ea&prVMuC>^)3}NioJ@xzCfaWEmUCYB-dRMI64|c9WFxSkA}glvljYXF9O45|D4u2 z_{?x~>pA!1^V;}l0tdiIcgWPOkZ2r~9dNGUY!6){Yv%xUVZ{t=h@C)P!`Kt>Jdbhg zYvY{^R$o8I^#}g2_fq~u!ZZ%NPqz@J*&eW&+sb?ed#%l`r(Kr_wg+T7ZdJtfa_o_;2EX#QbhsuRTQ0QF80-9XiVjQo+<&1KrYjHTz z1{=h3+V~xGk>^EWdKi|P0zUxk?8!a2aW$uy&Y+9Kn-owsXn29zggCrbQOIe5g6Fyq z)P}n?O#ZjjH>BB|HoigMGy{l-H+=nS6s()pLu_nwX}de%_p(@gQ`jI(n{EW(cXEIJ zOS0X8qd*5umfbu`4gtbUDjL)zx%YuSM5`$BzsR}$YN7{L`eEY~#)b`AOI7nWhP2JW zWAv4C1iK+Zh{0m)p~8Ui<8ZRwoS%t8!(Il+GBiesXZiath?$T4#3)oe~ZjXyQaF>*qwRjiO|-PAcEtS46Ih9=GWq>|q6SI0{t z$PS(rgo>m?PS`h-6?72n(U8+N#en-NLEFNsQHv5`qpq1|YoUUPV&|F?)5#S6#nhcmC&@X}= zW4oe_)&yT?On^9Y@@;&~8sj#8e@nyt=vxPW>}#|~c|$)bC8?wN+YTJZsWdU3dntK7 zMRKTUyQp^F$|uc^_~shN|`!$$H<1&gxtNcKCv;6mY^r$FCG-&uZ~QWrSe{ zty79*5A2Ou8QSdCNZBk`@~2fZKC9f6p5#zal54M z7HeSdROopL(o49@GVH4CVJ)6Ng;p;z&mYF-SFg#J9hnF$9C-P$0E^!6URTWhRfC8Q zBm(iJp^+C)OqS2_cpe=U7Vp=6eBeb-I3kP0YH~%qH{jD5*yw~!xsyG=!}Dy-BtHhk zWd3BefV{HejrHF)=L&Gx@)!%<|Cb}$6&0@*azLjO*zaPXO=XX)IBcg&A!j)!#w*`D zGx)-SX4M&}2HyjcNf;Cc*(C8JUT^qT$vzR56J&EP2{-woZ-P%9ic**V$9itF>H!&g zwL#uFyG6Chbq(p9eUcmzEuVi3_SrQ)g>)f(j4tr#;o_SaTC1==_$pnjG7OD(wlW99 z@OmL9C7^}{^SL$P8t;yKl;^C*Co6DKKWLd9XKrs^l8vy|#5+l#PLq%f{2r5xf zIkr3EocAp+tg!jNFi%zi$m(gg-Hq>l`g`?aqr-aPM=^6rFFS|TfjyHACW)vOQ{+ph zK%_CcL5gYh%&_dRBFG`7h9XzyHU?PK_y!GQl#0|Uvc)@QVmpF%eo=51-QWyeo8&H# zjN?5>JKdoi`6NbFaNne1kSYQx$Qz^^nWX|lRtr;F4N}Z><%p_yRWmzjj2J^rva~V4 zFk-UX)88Uqgnh$piZ2wX(2dbftpY0phMcfg4BOA!=ylLLtW{L4i1o^L zcIf#d<^1!C{@)nku{t(Bl610zhXY%wwI){T5hVwXocm~{3@nc=l2w9Rj5-5a>N)@u zSn^Xx?#t>3@XsI<4rZfvs&;XQlOG-QaJq|Z@f zGyEB)4vZKk&HBks`I6wwFo!iyKT-aC=|nOJ4s4z*WDtrxcX)QuC&>=ZU2(mr8)`4^ zcwL>j*;~gy7*XT%2qdIxB9L!s(o7S(2;n65)Vva#_~Gk=#$JU>-<|qx^1#V)6GpK3 zj987+uQ^zmoFC%@>yPHB^W0sLD?mgK_`sVCp$$3!WDRb+Y zRqo!bS{lA9bOT9nZ{zFvb)jhigS-PkGEnVv09Xc4ZXww}g)ZV8i?}`ym?|XeU5e>8 zvSj)hk)A)utMT2$t+AnJGd6lg_$^}}+6-Q-&_lUj`_uQljOcmq%qwR}vIC8p-88qAH2Dln-#DQHc@&gu4-qRfH!M}w1EcN#?E0?0-&V^s;wcX8yqARKI@^A zU?GKplQD~KB~X8*ybJps``Muhg2iUwFdv8&Sf*U9?oD?$g5`%bJI<4hFO63WT2CXO zwTqI2PTdalim|>2IyfPTkRw6~gcx2j=aE;8*WQ3k8f*R?0Y!_QEn*6Z*!n}KYCn0m$}nPX$TudltAzH}nDcT3{z`Cw)RtcboT!SL}UA zVaRdU1Bt{QZZ%y`ABxcPn^jHHKP)~K^-(dzJ=Z$BpYv=#9?VGO{ zt%!2h{@;^F4(z^eGqDSClza_ER#DN39{0)7ux1&QqrZG2& zBY>c!hKYwrh)z}vZ6JvrB|aUpvd9K0meFC(z))#CDYl0Vb{@mXZh#pv@L#A8mYd6_ zIBfFTf)L6wQb`vJpZyo77?V8uNlqWKnClTUVgEzRc4CyexG9<4{ky(vv>gY2>h~Tw z`GQs10#W@43(-Q!n<;V@n~ij`))1(#ZD)@DV8FHBUybcYQ1VnKZuYtxt}a)Ckg)Nd z9TAzKX|4xnVWJS2ri(#?7W#oPbqMNz0TrOngr#%K1P9+xCj?>B0p8IfIqeM_XFaqf zU8C>2sgDHaK^P`6Vkm5xAVmaE8`^xao>#-1;o`Oi)0-VMP*!W0CIt>@;5vsUV=S%c z0Ik928z~ludu&{6_4B!~wT!sG{Nq6Ei_Spar@SiI!;R1KtlPu`DpCpTm?7_fd@Cn>%84RPCAL^LfAMY|o^Rp!}@N=@+ zfx~JgCZ;BnlJBBODiy7l6+^GTR)PB*sA3_8PQ|;}_JrJbZFCm>fYa|ftaK7DI!9K9 zqC6xTgLVc5(ihpcE{{D>Mi`9=0AZ|0Spj9rt=}wq_v=QW#BBWDPe_#m1LcwlP|i^D zQxrK)MWciXhRHzP3leqM`;0Z(coww)vCTD+tMyIhYnaAJ9e)J`%(T8qUK%EyUJoTJ zcf59qb4U$u7f{h7pAbHa`svuYhLQ-#KT#F9Ro*JR4w%E$t2Q7<@MsRMK-e1%-f~bC zr_&83Z*xQ_{?s{6k7+`P8Ue279F7H^QLnQ@)W#!!`GvV)+_T$h2i_Z55RoeiJ|Wri zW^u3@DOz{A?Uz975Mt%|zQw2?=mbw%iIWMw3;V|{<7DIZPs+{92^@BxWT8M{+x(S4 z;(D_Wgfa`u|^I1aEGaDlz8cnlZFT@f$+p0u%FWLJz*ILqxdRQQs;T@ z_?B@A3XFOEAIMP$UV_?9mY{Q#9Jr8lpn(CZfFL6}?}8ljG}t2#J*Rs-GH9)DwlWoJ zkFw}B9N5y=->H8lH0&n@@8p`U@24DCz`)F!}m zw=z4GDE@~o%CMAas;HB0h*~l~U92at<;6-}0~FdBoaakGMS&OO|uz`ZU&?Es2` z|9^>M)dsTh>OX$+*Odv_z*`M+wiN;UU{{nFp6#c`z5r}48FqoUdbjao`9m%lG>Cjv z(s)iYZLBbxO}^OUGd4DP>*o$~^>$*8xsVqdh&gc4fd$zwRAJiWTL4{Z`mlQXy4OCk za&G#9V$VA69$sE}LSP9O&+YZQ!7Za#&OH=R;5+Dqwf}o~eH^=+pYg%O{zbHk!r*>$ zL*lR-Aqz=DWY*C0J7i4?U5KHEZH@c-kkjsMpjP~7&b>KMyN@!lDKy0V+rV*A7rGSM zU+9f(x%6XC^|{|_n_Oh)hkVvsa@<_Cf{h_@;J}InJ&Oz_s4@xCm6Ns5sGx+DktEW9+8kP zq2y39TR=s}@M3tkrR&@?lzEa|dT?%odk;5GzRpFx%wG=~=6;W2Sr(uliexWBVtV)Y zZcDR*@Z^)`FT-ZJgHGVJLhK&sRuwo5B$}}g&Bi`0d-OkErk?$20^bR~;Ql6gr#UB> z!@4CF_`i;mUf10;W@+|$olvS%L$8Qac{Lp5&K)M+JrLYax}{p*xX>zc-TAikkaD=T z0IwUEb+hRDz*95S(0&XH65s`cb^~syp|*BbZ3cn~PP+NytW4l59RFL*Uo%E+TfTWS zlax4c)V9&YcGOYw8j4g>(HldKh%uq3ZsL>&t_)3~Q-al*p_urC(h>h|Qcdgl$h({r zSScJXM!|ewW@sjjA=xxfXj5$fGEL+hM2hDt%p?$Y^Z-_(^*?+3mo_=BSCWIY7zxQKSfKYqtsTt6xcD?Hdxbq`CHhoW)g7WUj>Q z52zGJU_D!r|5grkn4c%bD%4ZYqL=!d0EUftPN~-!ueENcf!6d=2#S^9&FWa`xezSL zZISHtv8lUaTLbef*v5N?oq_r6p8qti5jsx(PRl^(ZnPw%<2EouG1LgZS3}7+P-Hz7 zjSGtws`!BliW4t7;Imr=>sShXO3}veQVdCcxFEMw(fUf*nv5f{Hc z-gk^FcVLr}W`co@lsthVaa1&Z|A+mT775aaEn4dq<8gdmi=1Ke8vMNS^K*=z6HP(=m_H&Ea3x-3D;C9EH7R&4{N z#8bjH{=@m5-u0r=$TMEub6bU<5*?7xW_Y#FUCKG3Oj8ccEu?q5?Go338+qOJCK+G` zxHX&{Q4HwJ?R~qApGBW0+nx0hnKVA#tUBkDD*#mw<;uAl*G#&6Mk%OBBXts1)Z$r8 z60@nCwf<{7s?*5!q}5-4R+H$sF5&n7@Qc^X3lc2!ls8DRZ3|UcQ13f|mm{i*_|Mjco!NvvZ#&>TOk!%NEpz2K)s4_}kN|9oy%>DKi@h8`qzubcuH2kat=F7y0 zW>q#At-I42=u9wf>8^uLXC!xdkB{46*ZufdFN^gz9Pb7qjSa+boZ_$NpQNjJY2r$L zyU)65ncVW}He2i_3~DnQI9?E&KgSBx)RJW%|NccMY)^&$4|H%0yCkf;jOX;y%i;gn zW7JJQl%AYKfpeG-HqbDiK6}1q1&vp3Bz=7GuZwt46WRk+mKe(`4en$1$WFh}O0t4L z1P9K5ZjNg!(_C$!Iiv9m#Dfu}Hw2z}pBaU&7`e-1Mdz|EOQ$#tmdL^(T zw34G_$Z1J13hNnauk|1+9m}XuNC@RMA-Aloh*%MtGd(Z#NMNpb$f#KdkXsi9vKEPE&h3O=T7%}l{8Q{{bBn{K7|L-&0mMQ`?*^xfZjiP!#|7=o z{(xpxgS0YahkKU)4g1$jJ~JU!qT-q7?0sVTA$8{1A$qga>G(?o<&mW>XQg-OP9`w|&!o`jq5b_1 zeQVAxNEvPs#m~13Zmb5!bCw8um_Eqk(HGz>YQLji?c%m^d% zIA(-$1>5GUdt|8`>(+mpO$7^F*z7&68Hg1wrhGa0FN@zb8j_FBCjWpOera6VRui-H z870?I;Bp z(<=7NXy*($VWz^+27rWN=!0TWlwNkw7t}QiWR32bG`{FzuJGGAokFO!!jh>@2)3*y z!{t$jxq1nzjI>CGe+vZkWxv~AYAlGr_tm-$fcI;?8DoLL=U6AJFIfY>;d>%JOZ#1* z5miNQNk*F>I04i2eL#v_43DZ>9=F|gZD(Mv8 zD&bXeX(U#q-zO=)S0JU?M#CAH`CP%uxi!4aerF@A=^N6EoV1x*MK`@&v0QK+)arY~ zD?^ZC*?Lr1Gfg%`jx}#*^*#Fz{Kmu2h@4;VsasEWu|tjnyTVY2IU=@LOv(3BBp+jo zhTy+}^Jmy*Yz%3egEid;r+3no(C=)xVKrs?pJtk7+T$}9+^#!6ef9XPZ6O15Xe-oM z4F@Z4{N?X|^ZmI-)O_&SynIsV!0zxF6CBh~a;Wk?grd%IKnB)9*9KGx3xrLuN&pr> z?3<1c(YRzuAk;i;U!Y3vgi@_Lytv+_dluRiHtWsk(_t%AJhKjs{&rTMW#4$=6XN7G$bG-czeQC{wQ%WYGGsq<}_Wn_CQPCd(T9 zSTl~T?Ll9DKUkZJpQ0ciw4ztLNLR z7XB8R2Y?pS_OP-BGtbzQXHVD~GOzr@saYq@DZ$u;aUIx-SRe~S%9eOuKl#*6eVi0} z0#M*^5;LJsr_ee~JiULij{f`hk5-zO^s~W)!2c2i_hvhT20ayEh>HZSq5+uAU{D>s5TY?-IIP5 z5U~vmo;S*dgrtwm6Ne5P*|m@g#8_^+^5*Qbkp)1ekpij>*yMx#`nLBRp7&YeH7n$(T6aCyFO8$@{Us2J$GNd~=`C$J?PMZZO(C)5VQxDEE%D`1OM$2u+u-BmLry1@r^Vg8)x+KLSZbRN`ls2P{lMrm zd}zL2m^~fN&5L4A(H& z-KcJo)&^>r25Bap0Ik!5vIEYT?AFt(;bDyS=7#z#C>@1V`o%#LdG~6!~X|e@sZtPOj@%vpd zyEEX5-K?8@k)X6i(lr-|Ho%r4XW=b(6dFeoAOdM`*LtI&EefKSQabn_+W}PB4(J24QpkK2qX?ZxbiS9MiCGT^xQ9>U|G6` z+dP(rR=9M##RjH&_Ill(+YdJD7%ZV@rH!!QtR=VRV2lql3^!lD59?RVR8$48<95+! zyoSWrh0Fa{PHl&(&#S^%=*Py!&t=X_J=pkS)}zH5Gcv|KEbwF5zKr5WQUAX1kEJrB zAC)rwzkW?LqxIT5a0Iy6Wc!^#$#+sDg^FGhjQB!q^^<1Rj^N9F*shBu5`E$KWF-on zYYneiRZm~`EBCJxgWv@3hV(Ae&OnI-sjxfO?zpkT0te%{)iQQu8Gqa9WW-6{SHJv( zEOB6eTWf-e1WF!9ku_BGEq9E&fY(wr^WL1>{vcijdLwMU^5gkPM|;Mq=^yPuJD;;W z{iB$UDQ_qK;Y)KRJqv<$dhsA^Z90YRbRlz6m?O%Bd`r7$A>9oz)8cSFeJo;fr@^WT zFqmXJf8?gavzh?MHOfb-o0?wqUe=s%UJ*B|swLIRWPjMsh<0(2Sggx4i_Ac4GXpdPw;+Lu$a;sp1XFI5f**#Nmr7c4h)NG6Ic{c@;wyE zqoUhohZOnlWmJp&>dZl>3YQhU^6BfQb<#~?U33|}L~xAWHve-bX6kaWK5XUG3YVCv zYvv6y)e>NRnF?|-^rthLRT~5YKArS1PYwTUT8yrGjCqtpkv)U}60@Q|cqL z%vDy|K*Di-DzGpk7b7ht-$s$GRP<0-ik~L*vIs?nw2IBHC1TJ33E4n8Wal~HUN7jC z zs_dJ$qV6qh6Yc|E00VDqCAXEthr|SSd1g8n%Ix@oKK--C^U*$i+I~muoI!_WgY8IL z`GMTHwQ+sWcY$nqX-eOKC3(dDIE#{}Q$Ugcw8>aNfQevaJ;t`dO5lgP30WAO6gt5` z$~lN)AV+>ZzFMq6S(w8O}!x+@$0x}4Wmahg&O zBHzubcu_ZOkq2Cl$y(qOni8 zC~||_mWApnprX76ysF&`i^DT$Nb(RhHWESw5>_*HNp3-fN>M~n=n^PhvWID5kC&&e z$&;fevuwMA&wPC&=*LmDANZam%chc4lMi|WC0|dGb=bA1=VPY)2)7+%3pC6U!Ra@a z2rkI&@o|mu2U_=OjxmTGF25xH$iE!@>ZYwJ|aQ)5$S zi$udujW+M%&%Yxgy>XkRsAE^&oeB8x#eVNOOh6=G`O}8me|gWQN)&I&K|a zo_x{3Esc^+YgX^_z_2M$4EOrDlyfyw`2%>qc9&ur+%f$r;TSIj%EC zj+xk^gOq$fMM`j=jIBB+N!rY-;^HvWA*qOn=Qhd?L##g8Kb^*a9P)k{nh_1x)DGGU`4}b)wRsc+3Hf+PI(({rRNPf8RB^rE~&gS7|2$>(L?|>s$Cfe zct(4~GN7*iU7l}*5jh;;S>Av!E_Pt5{BULa$lB1SB@G2OdzUEc#ci;-2i0A3mWj^lanr z4epM>(Ux+yd#nHVig>#-QqX~vBDw4r7pi{zhO6B(KtxXqs4kY3ipC2$n?uYTccbCQ zigR((zrJoFjXn;XgtpM=lc7wExF_B-ufnA)0_zkqgL6o6cq6E8#fN-8>CMqrZcmK2guWKijLFc>l(GBjy8>gzkOTyI9(#X?dRN7VV^euxndL&29hGal$D7K&e zX}qR@)iha+j}i8O)f#r-9G=g*FYr=%MhyO zZHF-7DNgGgV0wd%^qugp=BO*VwWK-_NJ&K-Ac+XE$Q;pOE^5YO|FNABF&-=!U@^`~ zw(>E{Zpu)=H^2X)2khGTT_oF2tLXARBh|v5y4}4INj7AAo`7%U8A2cm zk3tb1rNF;Qfq0w^ssGK~p72ip0obpUiR*$+cy@W}`OT_0-{ryEXO%HH-s-^g1zN=b zy)R&ZzB#*LoW-u}CvJpkx9)IUQ_61Q9Ja>g%FMUxy^RZ#e!-k8WZO%#Gb%IL8Rb&) z-4w|H?v5G7KB(2%q)_LF)%aj>!fyJKu*N5g-s-R3%B=`X1C|aXMo5fUCOAXG8kQQm zZhq^WL&_v2p2WKkM&R9cTf!y|`f*`;%!eA?@~*sTM9X_O8@G{zFAZ9nP0(_Jl7q@- z9TlA`*eSS#E+LeH3}DHb5iKbk|Tk)B#rdqbY%t|L(qxtcc~OE&UY_yUhT3t zF}!m4wT8j>-UI~ki?itc(v2berR@3UCJrY1n9;@Y-%EVk2&Vj5oV%oYDuMX?h`KZ# zCI6HnC#dL}*OLMVTzkmlpG~S|z)JAvYs~+ghtMnhJPcx|XH_Y_&8pLHbOB8P$_V11E8_s>Zz?3MF#sN?|@UnU~M4r}2p#?h1~UtL0|7 z3_6W`Zkg{rSv$RtXt`$udmy9UIj6#HL;evDLFLp@1&x0L>bC`vMzcxg>nC-0z(#{TiQo&bXT`B$l`>oPr4AT zvYfxseaI;_^pX%lxG8YGIQ(Fk8U;e^=G#~gEKAFd-O%+z-M4<=VT4M{8|ObJo7h3c zfwNCVCQwPIth7I)!NhJ5!GRZlb55msFpo{yj03~OLYS{3Y)1g@n1-B6y=oT3 zg8@O}+Yb8HFzdE~1h6{BaOMGDWTM44&YgGN9CBuFnJC2F55(cI# z$T;4I?IV^w*qe9eh4GCp$inkm4wE&bC5jw)1%V*jh#h<~CEr4kO;ogjzY9Vqs5o~g z?4hf=jzFJ~Z;K?!zc*Zs9zL zda39p1v-pCgsns~q1HL?61UMYuk8zfh$>dcRk&s<&jfx26py%X%@JX>95@#kj6D-L zz-Xqa&qfvnpcB#%1tCNJ8lVl#d#wT*?(CJ`J;=v=|nEFw}S($B~dB}t~mlPL9aD{`0LDuX;%KR`SMMhTj zI$4F=`2aveLDV&-F6su!1bikC^&hV>pWGm~r3pNBjClEU&9po!k3t!~0k=wS2ee=J zFlv-x%=a$jz|%|M`q+ljI=g1-^Vs0knAVYPq2(k}?k5=z3@xa+9O3lr zqvQ~7*@Mukrte2|lTK)xbkIo`aw$aJr06FtlG9|p%SocfJ^=86 zg7&|zm+a^EOF&vtTok#R&YFHFY{&`oMjgOCsgIWdNsA2FBZPy*T+jCJHKgTcD{Uk|()o}}FsQ{Wpb-)OFLRrnr z3d*A|$a|Q*Q1%N_5wNi0t!rKSK%4@9L$qI;=Wm|h!%drsc?@`ThbOMjEli4^E_Cr8 zSpqLVte?CjOXAPI7!bbb463n=hO1sC(D2$&)&cirsiCn@>YYiezmlbfg8Tz)7G$9m?eAqIgf^UAE`9~vq zw6_G0$XN$o*m_KmqNn6-6uCl0YkiA?a6yAjUNyfSWL&#v)rvq?ON0wUgEZOWGz5jf z)kS>}+-WQtP7Q4)&5C}>h?Wz)_c~L{LqEPnQU=_bYJ9oD z?V)&B(*~8hw7zknm3*j3?ojUbObx9KXmV@vUaC4Z10s>hoD4cP;#0DhlfXNu%LUnqg)?}q^HN0F*RC1VM%T0N^?Tr zYLn@F!kdduM;=I9Wzj=(;<16cH@q_Bm~5%X0eY2S$jR_2q8`ve<)G6;cW^DV)Qz|C z#QA_sXUHjP6?k`UC@KzRhGoM9FIqIZ-*3GQu#&Pok z&2E2iHz(73q4F9QnE(3TK%uuwF8h4cO&krg$+uk*D?Jxd6_f>PX`0~cjJieA9dVjK z4@AWb4X*2 z%DM@eZT~;bUYMhxQxZwTF^J zy-qeJB4T*hsy!@eUk(|}@<26eF_uNH3Qd-D@VCLv84EKIG`HNb0v!rMmj|bcl4f-% zm(0c6&I^+{_ds!YILreh*u_j(irK-3&Q!4UYxW@|c7M zHtsc?9B84-pLzNX`&Nou`p#Pk6ibiW?Yw26{P2%Ce|gtvMm{>5`~z~>fz3#(i5dBf zlItk)DM)h)D(MsS?x5s1pB|wU=cBoqZzjB%8Gu)gDe&I`+VC?6I>C``4Sm>b*l5+$ zjWg=RN4#Ua65ec94RH5~yXg%iDX`b8%jFp7n&$wG*B*H#yEdyXa<0#7g^bgmGPwiX zX4Pl%O1CzCkGPuNFNztyVZHBHUe_SjoJGg7=QgquLC?1|R$OHZHOmMZZENP0=^i!B zh^iaExb!Cy>%geWGC@@mB?sSXBNhEeXmVO8t< z&VZ|CmQK+~-*vU`PUq=NvBt*JN9}sV3LCF{@#Trj&SL}); z!{GV9&#MYIT8|&@&rK!8?E0M?7$FTNw&pM;hZIHy6f`IJ-hB*ID-%cTl=b*q+haE-4*rUZ{A+Z}? z*rDat&%EFLwsGYt_IUU{sb{zHIB-;7Z?f{7r{uu%+JJ#PWBE*rL@!fc^VHM#pjDzW zq};Ddf&u)*a2@QmD`ps4a`oWpVpwlrni`AL4fQcuKwmV>1sf;FIB^PnN>S>y*-b|S zAyV&*c7@Iv`T4Yp`q%fmbST@PblD(}-K2<<>zwOfNBY<sdnNFxw1ynv zl+Ue}#mV)OYo5Jx*ScBjRZkFyCZM$Hg|*1#roU-k#mQ#xF8v{;dsVB>e>|?5#Di2_=*qlF9{C z^u?fJ6-qm{@i$UQE^53OFS_R0Pp}&lhs2%|B+iRvwlNy!!h#YG z@YY{)>6={^nLNMOKUTV50%dwDTyOIG3bh+P?J={7_MRtYR$A1x_;bh1Qut(RzR@$2IIV7 z+2)LvM9TW!_kzqxN!U2M4(!obpewzmV!3)OdWfBs9d{iRu`4w@`fd6YbxBAHZl z={FxkWX(TaB<)PsKVcKep>MR&2%FR2!gE)CShJu&SP>Poa902xEi(SEL7E!cAVu^b zs~*y2HGnO;>|=Arn{mKp^c9C48pgca>BnLFBu2Oh?+QO62@af;E-(Q=Dka}Rkz|x* z+T)h&29@OE4n7DI2X#PStya;>;3j!5cJgBzKn+wZ>?0c?TGztV2X2}aD>1G~A9Knk zegmr|XUc;2KlqJ#J+8xgDi$hi_0YKd|JeHyxF*wUeLwO0NL~!t2qYOmkqCk~vKT62 z(P?LT+v##UGxzpRZ|C2cw$k3(b=s!W+Z7iSR5k?_)Bv(X5D~-;Wpz|sMg<2IMO2Uo zj)SA%!l1(coFpm6*CNE9%_b|>I+*30oE zo2-VrSUU@Gl(KNF54q)x!XkQF@9%K=s}Z5ZXdFflg=dVw3MJz%?EUQWY?C3$RQwP& zkQ_E7ADbDHLzEI6+-fSiUC|2lZVj>`26iZN)PZRb=;;iP%Mg&w5Q@MhK=?sVe+X4B4T8>0s&ssnpwpw7F7gl)xD%pTm{V{br7;O)MxjE+;v$b z%wi6bc-b~y7nIRyV!e>MSvcCZ>7&5W@Jr30tIvSRb8oT3QD(c#4PwI44yUXWWH~$k zg~QQYQ2ZSf&0R++Qz?={ML&$}@=j!MuLO!Tosu#L(l@H~VF}8^K8gHx#l}e`E+4Sm z^gee^p7j7L96f6wJKu|YJFdhs&Oel@%i#@&!s4->q3irz374H2` z*e!4)(BlT2szCTL4MY+xa!qyMJz;fVnj$N7jZ3?#K-i^%A=(vBhbRQ`77%@sUUON9p^wpjQEi@S%h)cp19`D7=D-Q3e=*gi-pYbjDiMd!X=3&h)|sI|PE zGak)2Lf6PM1Gb0{3(KiEWvODDJ4hwc9U(g2eR9#SM&2fEjLIR%(XItW&Sm_g!bIjw zlx{lq1Vg+Z^}~Jhd8SiV>5}Bz8qyS%?7WCM5RkF8P9 zj>7jS;XRHOzNw#lZ|`T8g=jB`FJq-_tyu~>W+1R&$OPPSz7NzW6f#fbW2i+7I+;fO z=rN44utU;l@nQS>Te@zL5R-|?_RU>Oiie{|aoBVD$jt0iQOf-k*-J%Z3aFJ?0lVtm z)5|=u>sFJp6Jl)8@j^Y|bWO5(ay>Lxf{q78;Z*?{%1Re4h(K#qhMw7VbiZ8394DW+ z=8_ExRJz+Rwasn72}$8_9NHSJ!Id{rypm~it0!gZ>WMm^(J!VrY3yKP5!7%AM`4y8nC2 z{yA1!(ij;UslLTRLsOa??rTGLfqB*>io5CM{3_9bzyniTKoaHp%rX1U*#?fMCfVVr zF#Q5n;NY&cy}Idi-4|VhV^7d2uZNOy>WarU_f`CMx#f9d)Q+!L{_L~)9rJW=cfXPF zN>kJs5Jb$R?*ohbj;1 z=ng?s)DotaIpdryZWSG%3*2`wy+jMV%Iqoq^4;=EK8{s6rIsmoyD+z1o~}MTf=YU} zaDNJXR!6MAC9L31wfvC0Wx4gSvS+|HHz-MLNXsThrig1`pP;Fukz|jwfUBZ*=WbW~ zxOukyt65HgZG+n3DOOX!-RMg7IqJU9wDozvY~r8DDh@|i3eC1T8!2Tv3Oqz(xppIINgE@JGE%M!O^BBz0{XNTK&-}2A~ zo8S?j)?Grju_LB%SbzfxJ_d!MD=8%~ZTC>oS3MiUssdIhdtoWr&ci~YN|&mD&C2z2 z^ln|C@mnvzg(%+_8sLBgR2yVLA74A6s*zrUW?q@}$DaN2l!!e%B=si-cS;((u7RdVH7F{y>;GD^lu}0hT)+akyNE{+lANxbT2|v@*-#9{+aoD`% znjvB>rCdXiWGXt_6}kA>H^$H%CT7+)k1|*ljhCQl3~$W5Z^NXpm})ysqHV*lGA!d? z8t6)~)Y*7JA|_T;Id;kqdR2)~+i59iKmiT6O;F*UA6g@=p@DF5d`7Xy+R2L|HD^Ng zlQX2BD8^`|Vu7Y7Cc`#7jP9wA{{5BZ5hla2PAAVIyV)%~+}-055K0)d@EoR;(44J? zibkCcBtPvUHG&K#Q(Wm%=(Cb3gzh6(A~r@Ccr;jb2$hKnJ&m1*YctUH_uxrl1 z!}Q7@tzPO#fb;}W&p8+klokh1xnmh`VPj~x>$(#wVWZr+>A=~3=(~+TRJt4=7S&nE=zC>e$}DJjiD_EEENbjTjgOzQ?Q9C3e6O* z0FJ{RUYWP%AD#2MWbv|0`jMCB!rXY-Vqu97&?w>+i$D*zP`EsFd1#*^L*RIyh2@+- zHUSQuZycPRpYqD`wQDUiJDf#`m0Zns-bbn?*qT($`yhteLAv~#qRPmhO~;N0g4vAP zjtaBa9Dc$I6_hCcJ7+By70#l<+AAqohDnwc{=_FqRz3uvpb|xvZ&lOrRU~x{a z;fW8&ZhvDtIE>c9;ixeE0#@J{zxv(1t*$25wfwc$Pm`7GTvraqOrZ97kn6gEQm&_9 z&Z6VzUzYsss^af=L1c29`(fePuyT-Nxg(7WTq0}}rG^xWcQEa<9!}rlb|he|>_-Om*}SD+ zX`R<>IMTn`Iwa&vleJUtcG(wgo&(1CD9{-bBf7FlH`lml14`tV% ztiHC~{nk~_R~@u04&>}&W2M{=dnoHc%ujPEBH6hjFcbTGAUg>q`fv|B)=e5%LDsII zeD+|hAUpn7;my~_n^u~o%YzO<$8~lCK=y>V{66|;zl#3#AKv}>g5OJ)QOYG0i5V+n zX!XM&AaXEhrAeZct0)prMJG@1qObAl{0-p2Hmze6CbRrQZNZP_J7aO)bn~mr228W} zpRI3>BM+UpJI5JjbG(XD##3Y&v|IR~iR6doxowmVI2G6}L&EaYfBpos9MDdWv-*CI zqvkb-%S;C1*f+m*iL8EMT+tFU@NA}(P`$H}iryw(@e=B-ZRaIN7KK-MZW^bQ{B#En460k)OzN)!hM58N%rl4`p8_N~20bH0%pmYMW+AWo*54jknibA^ z5>^aF4!Or*xpJFZ1FulnIjhUi$fqf)S=#5C$Y8m0I-{T54lN4%T+d5eg6l!g%&zu@+|!)ooJdOrQ{gDTV76LtLE3X(lsQ3Z#MN41&pD5aF3#I+Ula_A2ADQWF| zY}ni#v|!n++raloo`ZgBPqO zt4SGsCF}vg#Z*S;Jx_&tIYL zhFr}4fXcbpW&UNlm+;lhVtN>GJft6DSVjDun_Pd;qZ!3eNmIfO&btz%(0 zR==(!>Qzftz|azN4%-wfxqsA@&806(jtaUXpgnx(8D&jGv$788W?P{HN4>maPA}+K zZ}lpT=u>EYt3*d4y2%x&C&eXdm*0ZHHx>Y&TDdw1uC69WBH--W>@GzeeMVUn-YIJn z=pk~r%MV}3k(AQVB~i4Ew>0vM^3iDOc18fMAy&rvg^}aVI=EhHT6HGMv^eQQCuNZR z>^4RmW(Ax#Tegl<$|DpxgnE(48!@)q099*H8}`qo@6k)Zd99Z(@zury246aAai`u`J^-U=-%yKFY5w^3J#;`Vy-Dp)9K0&s)DmZd&0}-v@0Ia zNRdN@gp=M{H?69St_;>B3P3=v3Al2NAf=I@{6I(RCqq||JF_yOpr7@M=UALQMIc+m zg;d7=(lSXr)GTqBl4K=u+(91*s1_BA@xAR&=X^LLg8ab**6)`AypRYa(CI$=D_Jxfx zW^9IxV|K|_k(Ux(rna#jr%9!Q7z0XfVZS4dj_#HvDT+N7F*=|E!YTGV$q#c>{9(9% zf$blvV>S=N{Y`ky`uA`+N0PHa@z^(g?)_^SjSg~2Q60D!WaJW=B;P%Xjj|@y9>syc zeCO*kLD`4BmtYMcwq3U!-?oisucww2E-JIsVjJpXaCqxuMR9Gb@E*CZs26FW;pqGN zZ`FVE(QBt)`5jukrl<|RF+r)&xMz_NC37lWvY<`jLEhE~EKJmg-J-kc46@iaTU_ha z&F=v2iKEsR+cELfhcr8~7YHoIv2OEzH6gIgQhI=m@4?~yjTHrjCmq?5?R?U`4Pq=v zF!*QWhi!z5#bi4E^ycOpu3t87df&NzdIQ-vTtt?`ZfBF(*7g{stfR<5W90-2jca`y zgNyi@KDT6kRd6F@X|B)E;B`E>DMLmhR5Ifx)%c)pdWkzq5M#mUfD>+UaVxyWfA^e| z5V5PCtvROFs!?yNRB=q52o#lFauCZoB<`NtMSmW4Mx+nh2NyT1bxb?a=%m>^-1m+U zZ#F{4%&4$E8419wTbN!^e5z%rmW_qsFwCq3cGvQ|=)V?8HHE@k&W4s?h2k{DN%cbj zi*EX2;f1*x6w*&qB)x(e!WzK{p}=yQMuZ%ezhfki$LyIAV5$7b*-*X}Ny_s+&5CxA zO~bPC1{q40@Aj(ZpYlRgwE?Fd`P~pug>`L*)+#w}W#PtqjjGe40Vizpc3Glh4*Rq# zw5lSnH4&OxejaTc@IziZ(G({VUnS@dURbsuL z<`luuJl@Qs*LrJJhAx)&YbIzD7@lW62Hm$8RM=UQVxN^cZYJpbxP>erDeO`{9JVIg z&489oDZ#vKc%1SvxU%R5V>9(yrd{!={1#y36z?EkbMIyS}{l!!{+6b*{s)6$|{QNe?03S^7J78`p^|6;|82? z7tmiDEV4m0WUrh*#v`+8>4Q{%X&3o{)6wI{*SErNRB$k zbkpg9^@7`?`>vs)Ym)gr5==fq>TF`##5k%!c9HbCRRx=#yT+>mBEt=S z9K{M3&zUQ`!sa=@i_LBAZ{GUa7hU2FnJd^-2(<}O1?4Y2iZVDs-)6shTNp-R+UREK zN;lk^+jCN%^+B_q@MjHX$9t^StMN{Aq_2F@YuUH+j*~NUj;T*d>gYyUQ&lw9?UUz@>FM_NRCl zSmFYa_CoP3=ZBG-RHx~i(1@ap28uH^mXb3pXW%)Xnmuo`ngQ;v**X1hy=)_kB~@HL z=Z1Ts5ZX?tG-V=VqfeAhEcM*!quJtik-@uK)!21ZV)yN2=cbJYE3B}=-H5~^$x6yH za?9D^zm-@l!V=5V(Z>MT>Lx9f*=3i@8Xh)bmG!8je}NrVlj`0Z_pe24GPkE)$jVnx!R>1ug-Bf_pQ|=V0%K+Ns5&MkY;x{CKn^xWr+{ln(Wp ziCsXH#f0KU=ba)waO$|A%bD7x6tHBN}!=zCVgI{mo}>EgPWj{R~4@Vdh_0yRVMER_KKr6HlqfMt+UVa zId))qOWm9MYZF-ZzUw94n-+dyz`)aV;(w(N3+9r`-XudRH?guXK~E{r4R*(C(H>*kaD zW`^xLrMyOwtI&`E7<94I(gPRb{%&DM*kR{|XzTFuS?@B%!g$%i&_vQ1w(yGQCeJNy zX$nkFVqZe7>QTtukY1>tt(;uo-YW%c<0GgZwJi)EJWRIpswaNzwh(P&xv>`QL{JyK zdAgC8Dgd64-mP5!ivz#gO?vz-j2!^u6IgBVPiGb31J%Sam z4LcS1;LA`^n5L+jnjTo_2Zk=kt(3XyiMdFm;9(=~Jh^$vf8zl3AM|o!5Y{iAF++ew1e?c-!R(AJ~ymBP%46a72 z-5yF73kxL2MpFSj@|m&4l8k)xk-s@x+W9xiv}!$advp@omw>*~VAmv(Qm&*(92LD@ zdLyKoG)%cZ6LW6=l%|FF?|l9zIr2|qG;dn}+6_l~8^#M|&&hY3gJ3TrargsfATJQTb6olCSxKJDLd?^6vAI@+_-7WH`HBV1*R! z@g|YlZBe>3f834(Bf=)t2H)*L_sE^8 z2LwAoJgSZzy<`+{VGZ0-{4O@HvIB1DytjTo!DLCI^zUoPq6q}l`vzH(6iS&$k(E@m zj{Y3aAa@AGA8$Iz9OSZ8mJa!{8{>#3AA)8eP zf9mRCLSFOB=iVo&6M&^Z*c#gB7Tun=`}@e=U<3oTbj@C5dZ5jI2#iFS{(k?wBZFcw2B&vX$8^1)eRi zZchfpxu^0zpY^%dRS4ys71hY~f^&{ka2_${4POJ>@H>*HW;A*IcB;vy+LBm>%!B+Is3RDVRX~+o zbMOuj6DeR0lMDt&J;-YbE|c559@`BsgP`-Q!)HLm>=vtqm6})cAAOeM51ieU1DFQw z2A>E$Y@c^kQD#sTO9@}*i)Cv!Wex5qEC^Zim{|qPSt&*!k$+ksLE&b|1U(E)<8O)r z^{nvNX(cWTQ42oBt;#n|anyNR*g`Z=r|B#{zJ`>ITzWCN#_NHO;YEuI8AxnBVFjjfKGlEt&c9Z^pg1gErh^qtlMtHkj}0^z z3uA;fn-E)P!G`B-0+0>k*!z*1x6=i`G_79O=Oj%d*N5A{bJz=BZnlB{oKoJSNDmd= ztjtm#XW+k64oOdm&CeR6(x`$r;=Z-#J4>IGw_R0KPAT;)pR*b|OcVhp0~&Ek&XMZC zZjU>>ZV=zy7;s2hFYBh0=YR0`KbFk@;4ewcWl0w7!L#^v;cLA!1F-GIJL#^?`@{YGI#Q4|Q){qF>;m*)G2;TOur{9{*jxc>m_>&EHu*KaJY` zy&LoJ?`n$5eZ3Ll@qf&I^%DH}9rS*D{jwx6xL&X?!Z_5$z;37uj&;Al%kVe}p-h_) z_N)cWG54OK@VD~y*#)Y6%T~Ct9bFUdbfYZx7el++1_7KTgGFx z4sbV|7}M`{?T037)c)OXf15N1p;^I6<{z9JOnjXq1}~fI)TlEdzp5{QkmARZ=wc;T=3EZx=B^)i9sRl%ZsJU zy)Ztu!~Y`#Rcod5YBd&Om-$@sO9<8AkofRtUrH1|mSx7@4S}VsP~1HAx%NpCLUr_3 zVV`S$Xys%OT`o{pPR6z)*jC__NJXWA7&hiDT(xG6jxw43vI9qM85%#E+~Q2en% z8+U`hf3AF~+A^Urly$@1zu#K+lae0!F8Nit){%+Xp$IWuu?fId z-*(&$>i{nKw~x+RDliVk1c&`0D>{ocbUo~jx&uB7yGo}!>%*2$>V^XT2GKB9$8jIK zab%l7!K%XEK7X-wPzXN#@~kg>J{0Tm7Ynn1$_`43K;B#~ zyrHmr;Lhso*an2BZmHFQ?0myI5Qby^WZbO(qo>JdY5dXdg(RI_(3r!vWRIEaRX{1B zhB$|cUI%JKF~Xgavp_EE4L>EmDAEi1oyzFdF3V>9wTh09xC?q8`Sc?74N;0>Z)A&0 zKJPj_`uJ?e?D&az#et@WEwY|s43F##r$NKL5|>x|RTEaWO=WJA8V<*bFPY7LJ*7NB zkz-W!RZmP6)PPT266;dG zO`+KuP#lJ85QY$51JJiY$%sb9U#u?3;Th@~aJWqR0gVzzdI;H-1#eYp4h0q&o~s4g z@SRzoiB>8qrpH7gxdbadvCNaL;Tj!C%pvpavq1>3;gFU28Gomu!~bm)q_j7LkH{Gg zL#oFNQhG|+Mv==@bUH}FlzKvmOe2Hz?sf%s?9+hl2~D8FDQaL1GJO z5m;IfM-~aI$#VA=C~&w7y(I4uEZ-}Ih!K`3q=FV~I<#fRIJ1!<0Zb3n7DhJ4Cf}UU zed;_<)Ga9#pBABr-?K`&#tvu0qR=YhSmA8k(QOy9EhQy}I>y`$Ef6dE7g)oR1r`f7 z>NjQ8pxu}rSj0c=lMvV|t)4RAgtn^DL487ppBYwg8GZtS8XRU1Wi76+BM80y<@ zIajE5`V2T3mM>U#+UUi?ecIrLNZnQ27Sji)XDb?B;Ip(<8vw&?RQ@ zfOb0Oo%r3ya+6P!Iq8?bA!!`;Y0AuenmkImnS#Jvbk*0e-mO`B*coRp!~ZXvJPXC` zirwZj$xFY`sAgb`X`Kgi>y!NFfzn4O*IwQSpj{UYO{(4Py0YqKbfC(G-=#YlHK} zq=;aT(KQ4ap;Zsk|$qPTWnmks5f$Ay6ARz@Azd zD;OEo0RWXrFvwOWf|2Bj__scmgQ^W zuf4}JSrI3Hr^O_eofYBmo+ro5ilkA>)f7phqDzD&!mLmfZv#gc>#ZM)ao%uW2`yk9 zVi2rPo9(8;`WTM5)A=9&t+3qI3^f%Tp8Z0z+22Sh(=pZ@ofwI-48|Y{>WE-jONJC8 zDGflCIxzJ-gz?aZU??ta;#gt9b`Y4>2S?2KFk@Ob8y!U@9g|IFL!A&0aVicwsr${K zu#HjzHgr^UR%pDU3+6iQwRO`LL(x(;=~L*YgEov&o~8rX-3iJKMeWnD_(d9s$> zGQ;6K(M~gT{#5#nt(KK-Y)Y0m?7Uhj(8Kj;rJDu` zRoP@UA9DbPX!}LT3#Mu)GbjL*;aJC8U+xBEJ*f*2D@cQ z_l@6#NleR-{2yO!BuN~0TOjB$C<>BEDM8969k`$3>ob!STV6@}*3+ZLDC|^{NoPc! z5Yz-L7M4cj^Ezh%k=yXBgO?H$T1J47VQ^;uORNw=ZCw1v9hUnZHtWw9Z%2gWI8?EV zown3(u@Jj0rSWe~0z+gx!C0-ao227co2Qe{8ca`Lb?DKt86KaDO$H=eu(p*Xvh!m( z438o+bEBn{8z{0K8801SJ0(5gupJ6cc5W6KO4Kkj2sB%u2qgLLo!#$*x}Iq>??CB@ zV{$b^=ae--9La1N^yls9zuWLl6F~k_uJLlrIOZFin*53`I^;(OJ%&#E|>N z7Bxupc%tsHNLUg84hq)X+zvV+XqI+LG?g#q(UmXdL(u{1sNyXR4vZ3JK$>a?)CG6)jZ4_Pfu4#WY1_@z>733}s%8QLDdv5$KYO0Xvr^{Z5q; z3-Bb5hk2g6u1!%f>ZHg@^#C}^x#9H!hj}%lMT_qgegrO(5b7<;^$TFpAP3{nUY);$|~95`dTu;Gh#$ z02v>$@~0Jy$(;P==)EIk35U%IXc-M!sn$@+WQrt!XtUQX`cg!)bEB$JWoU1&qYpdl zrQ78vU;f|&$G3zvn@1L?FniC@C#+C0Zrfez#49FLJmh7+L#o(0xEzkZT{OecNlJN~ zB1fobluKw=#0J(8C`D9n5Mm$>IZQhG`b;BFNfSr1oa0E9?`b+8w3*AH!V-nM5p>`g zWkJCswmC>sAg2kuBYYVN;my(k6>OsRy$nj+F|%TXdm+wi>_m%#4#31!J>(w>U)UYd zp-v0c!-9w|&S9VPBCTo!O&(dmkDVuD@ibewzWpsp+9VU8zPYuR*w`2?917d~@d}C@X{eXnbk~ zIR>PZ<8(Fn!3CbHCSLVy4EvXJa+XeeRmP(-W!KkC@Z|T zHqjv_o$TRdyP`A)YBXvv4*)etu#-AWxpxoGDx`$+o`<-@2)bdNz*#L&7DAYU8 z;nfSaxGiz%Blp5Lyb}A(&2+qr{ZRpPV-H|Fdx}SP*Sh67`6o%OWv5WiZl9qw*Pz`| zI;DiabPdP=LHNzIg~k>!x#|R{?;0=!YuAQXh8ZdM{SNEHHO!Q-!@)2EIOvjf91KHd z8UObGc&v9exwLg(s`)!v%Hf?e5b_6MBb8F7P$ZFx&Q~nN`u7}2DTTBWjO49STQ5UPF9c*G-%0T4YT0@ z$trCiF$gtc#%=Q0ac6DEi^;jLb;3vaoOQg^O(=R~$$$GoVZ#c`?KsiGssiorAX8)r zz!^IHfsYE}m7Y7N?)Q5r!AfKuw0`V!Gq$(tp!E>J1uc$@Ilz6J3Szkq@I*}$$r>mY z&gdxdVD)wDKza0+lt0Wdtu3y1kED@399~->T{viMsiTw!DN+l9x`BD>k3Ea{ZIHyf z{If^%;`pt+1P`djk=$0a3ADbA!8v|4^g$3oZHj7yJ#a7C;9cc*2n0(rqzTHVD17Ip ztcO=V=M&dmenu@`tqQ{-RRNooJ@Vc1rl=Z09o_5!-A)8~5U)EMa8}UC&zjQ1>!PuI zesoUkXi;MUqQR$D5n;;>>!=yVOQAOYr;h($CL~?>@3IG^k;6#pHbYW7rEI0hCFn!} zB)2PJxpNzWPXuWVIvSdrAzd_5R*uQ5=w6p9Q9CiTPK4&N!8b#&XJQ}hlF_x%Tr)`1 zbw{k{A)~)pdeS?ar%6*B2r!0XZiZ|LjdL-@WnzTIzFl-na3(}}adaHHQH4n-v}*8} zktNdzvUobWLfuV}41=vl(-aucxd|G0*d-eULBN%|>bX_8Qc(k%3(r~;9SF0*kh6Ad z6>+R!HZD0X^AmRy%-*WqbdIdyFw9ELV3tEEK`?9+6PdR2HQK71sf4(OH97{0*7UUW{>6mkhsCb5g z@|=W}Z&-HwAvv0nLfDr{(qs8(n_naXSBZJgQ#F&LCAVLBd}@xter(CWm#TuhUE5WalZz;O z0nIk_n6Ea^JhKz3_Wk-B$(9_s7sR4iVbnEC_r3fi%Um6}SF+I~VFqkz{~2b*W>nZ| zwrw3|d~Vc4OU(=`ViP(#l>uTGlLX(gSh9kE@O?)}CsQS=Wp49}$GRDd%^#@!Q}ECH zY}qaIob5K=vVB1CN7GvK=G9Y#^l&&7l4906X32PE%y{w$x2sleo5@xGwfoM`8$yPT-)zLBG>vevbRXW}1rSb={QHr~^E zT!)si%O*5M6@?y?;~!rmy$sFOio(m7qt16BXuO?QL!T0#rtiw}i6ue!KK?UGc|B8m z$rrE>>~L}4*xwv`|BJ5I-W0XLw^LFE{r1-St(T_qZh791I2^WKdQWmFaJS2SQb%`6 z*7|jfrGUTfwi6wbXPP^^-ehGvsTrTo_TrmN)b#W>j*w*>Hc`1|4(?h?xrQRiR5a$J zwW@NlesY@Pz8uJKD?_uK2LhH0HCw>=7*zZnmr)!#cP|7m#uff!-D@wIfZ_M6$^~RS zhk>!%3>aG|WgbN~Q_*@F^FL(-wkenZ($Sdl8F11=MrSb}Q-dGBqGO=gu#v%np#uRu zJXBN1yp2|cA$^TbjW_JIsTnO)hHP+#4rr$9E0#PLE7XyThWOab_ojN4(HQevgL48yvprv9;KzdQvP3z}Gd zhV&@Z80D$gO*;~Rj2Rt$mA>z$$&h8pFtVGgMqvicwy+i#gXZaN1@^&0Y4(AD%E_!? z@a)NWdeeMepEZ0UIK>!yWie9uomC5Jfdwp=w^YR*u~(-b*HMK{Q>Av`oy z#riPZ9cRNtFD8A!emm=3^3p;qOGYh0=(PYB3na(*3vuZ{=LQRxVGA^IVS_A_{zz58 zBq*_ktWk1hnOlpfRdiKm7_*k?;bl)TYC;+M6t#%j1jz916ZJ~dCN8WBUgdR8)FD3* zu!Yw7-*MGkfrwxa?}VxwoTa-n&)Zye#s(?2FD=h{(l%_da(Kp_t~ii6--N4KAAHUmPq5DfMB4YEGBoX~`bJjhex5{DpVUrj$|=(7!UPo8pmE$-Dw6Q2B*tlUcW ze!)0WXU$M^gi;=&$N?&PGo4B@0}|Zwc}E28vj+U$j}dN*=#~7=^7><)!*d`YgxLY( zV{2V9e0PFi&<0;TzcBZtI(2r3x-kkf3U_9mld&%V5EUj<*tT;##oQdj` z?GLyJ$<5T+%Vyzk9fd38Sin4)f7S=q%~k4~;YYsc0lyudy^uWWm#+zaq*&}0JE4J& z5w7C*%k#v`CL{)54XFe2ZZVzjn@6`Q?u1tbSGu$+^24fB_hC2n_^nY)`txUFG_ZL7 z=p%o^IxL3ED82nP;s03HJH8-Zvz6+|s(DrPRRxkBfZPwHe;3$Thv?`YUV?H=qm_mk zwi!et$HR!;{JB@|Rm(&p8(467?+s#ggCc_&lyV(KQjG=GMsiH1I4d-VhgvRJ4ALOW zQy2L4xmAg71Z@Z%yQm;co6X=bxV|Hr=Mla1F)!&=%S5A$rYpokPd$mRSh(Sblv zI)A_vi=d3=&a9P+D$&?}Jc_7W?b&E4cX+)P_cprS2+OOg+{;+UwcVr+T9%c=C>}9)f|q(l$b3M zn<*twv^P@Gd1{QaEEX0>y6Bb8tJV3T2Lg>vlo$K!C#Mk1P-EfNIzi!^*MXs`ow{h1t zl0{~G%^pg5iy~dn&2Dyv2L{XzF$1%X$QOBF^LGrxb!;od z>(UgxA*~?;0Ut+b(iE42wZ0i1mm&DKJt&Te18{oJQRtZzfe|mMRaI+Gb zHYwo`3;{aL7PpHKbB2y5#*%7u{IG+Gg`7M)JU%vHj{5l&((x_bxTJn!7I!bE6K^rSISuWNw{ zGmw>&7x6Fq8s_dmKqX&OD0~#+uxQF?=lLPb@R&m9z=)cht-9 z864aAT%&A*ZySttKyZ;c9Ciqb6B=c8^h)OOn|QrbhHtU6WHx_6M!y=d`d~7i4ZazIg1t?pgB@&)szH0Cz}C};;`ipl)L$l-X z0%e6}w}nQHmg_Tj%JqVC%n8*!NEzZ;AMX*rRKCk^fnKmF053e^FKCK7L3GS6vS2Iy z%FeZ90Cf92^|qe zEXHuesY{To^`5*ng=#>{j729eUH`lX8h6Dyk}WDtya9f4CdxLmnF%3qcjdu#$tqL zWSKt4L76R!K~6M=t4|5ei^|2AjA;aqFj3qlSSmxp*?C4U#W3Bq*_rf%f4$8Ct6Aah zPT;ci?zMj{r*)+oTddYX{#iG@lHV-dBSmDRvNCpv>2s@eNo3j;MZ9&>nxb^{GWoUH zi3*xvBpf7NlyjQB( zA+GSxp3){j8biNRs_$vnvhdEZ1Z8Jffdt|u_3|=9;==H3qk|MH4f8W1q>KS%E`_i7 z{9jA(wqVAJ8LM9&c=_?q#;7XMeS)p->gY4_ftL&YR=%{rs(_k2_fMt+C4@mk9RMYQRwc>4nu3$^NgIe5N$ zd6AD^ULq`F8dY8NMN-Rb3{4SdF-45#P+$?A1%~waa_-b~0o$^}c zp2%c{UIV7X6;IeU@r=W2Rf(Wu)vkyMDwG_QWQwbyuVyA}qdGt&O4F+7C2QU4>3ZNk zprBzR6i60I*bz-ff~zN%hwT9iu>Gmvqw$tIqM-~a4x`7)E~%F8g(>WFyAoQ-KM>F$ zD`E};$*xC!5lZFH(OqHeW{<@OYUNQD^KWGY$6U7VkH5O&>OU;$2b|3YTH!@BsSZrm zU=jRouWP(bo`vERC;)^WAaG@@0lMeP`CmxP za8-pI=Egw`eo$B?nNlWDWCazC(xsW$(4i2TI%8k(OlW~Q7TXugd9le-SiUp{=hk7e z{Jr0WicFA*-*Dw;B!RWd=poQPEwZb|@Y}<<<4l8zG0naq-tQOIt*c z$E_9>Lw}cc!>V7)Y#u}7VG|b2y@N480+=#3uIKo#B%EG6&4dppv3d{5<1juxG{eU( zO1Xm~+o@<&(rK11b-gZ#nRQpv6OQg-YzCW@Saf)IK+nhz2%DK(K4<*2>)*03<6!sW!I-<(r@066!$)h#*pcrb4 zbGIwLhwtco3WWy(KLp*MN|zRwbzdrF4m*Qrcay5r69dzQ(1!|tp+UL~H#`p`yS&#- zJit#0uc8e%a>#-TSu6jXXOCjR^W7*%VbKaipI&|rSpTp{@{jjS=$f@iKZCRk=g4q) z=kve}LtT{e6N+?D(J5~2Dy=8fBCC3(ZEm2=4_o5czyZHJx`W0YvBub<6H7!gq!69! z;dRrAVth)g%5vVpw1Kl+BQ2k^#{C9gQzb~|Bk>8HA;>eLTucG8KcHfIE+nKuiOk?J zVcIDsvV2aVI77NR7_(APQ?NStc7O)0-a&bz3gyNOa%FB+AU}p@d1^f*v=6~fe!wYD z-3G_|{NYXPMvi~Vae|E^*M^rcnl%$!Tq@^ow`eo_PeavHd_6ug!q%nde%0ovdo#PI zF2I-z`aC=3T4>#Q2r7FLnOmL<#%!Wtc391|U1#&zgIPBc!|b@JxXp(@`=W2J(A+<1 z1|2gCHN=b;&{KxEY&9R^tfzSGhANh766Z1|FkyuclevHP%-ZKJD}0RmpU8DXWZGHn z&p6v7gIE!PXmi^e)GSSN(Mc1*UC~Xi5+(C-J9cr#j@i{y?mDipZ20N59Y({y@2IPP zI2rh+#N=_v|MAsElElvA;BXLco0-RvNhzUaUOE+xoyBpBv=RhO4F!pXV)S;=$wXCM ztqN}z3U9k;@Bphu9YB5?W;$(0$YZAxehxE^y)XT`Qj#!vLRL);($;n@zli@xFB=i1}IHX6fy2rNF);a+#xHCBEf zcWrNI^rc_?(ByZtfA`zpCe0j<_kV8YN!*~6ofOeO7Dz^k3ncTR5YN6ajehc>KwK|S z$2d*V5z-*r?a>Pr_NlXT>E__RiC{bm#SMawM1?{Kmz|kY9EPg00Ct(NJg`l0hK6zu zXcFNPM}b=+LuG`^5)is=5$%h($E5lesN=|G35Lx;MgbW75ST5NZ&PlS><_qVh?OC? z1uYZOjLYPC?iFIpMYOp!@E~=Wr$+S-<6ozvQ~V%z6(j5tCGzo$Hq=b9_!?MP7Q3Df z`8u)s8jSl^`b(!Ry<98qVV>}0KpgOY2Hsfp%I+Us{qF}qX;thAXcJt8!m9$N<=<=P zZKl%$lXzue*}SySTA$KkKC~f!tsO3mU?S}BTc1A!tBK%leDd0(pIx`?LB!e6nw92A zSrDDSCP|2jmtiJyMU*Di{ZU9eb08251XLiT%|xg6`N^2)&$-P{umi2l!;bnvR-hgK z-q8n_EHw|dx|G24WG%Y@3Wv8rJI&mld`b!XXK6u-dWv-A7#-y?4 zlH!hZo4Z!E>pOYW>9=#|FaK)B+eoN=^3-9UVvj4~x5zn2M%R%hucb1_njm&I$HGe) zwPmqBz`JK2z4>huP+s}4=lkT?3j-8zhX?VcE>OyI6ltWQQ+a6$OxAYu;~4DJS0>W* zLeBOe^o&a8mom-DG;nQ`cn0qDO=vcfPoo-kBhxJ1pwL6pA@pIjs=682YN}jvLy;ef zeLC@_2Xql+q;XuW>K5JR22M3zt^&cH%aTmrVo{>FZpINPj;|8*f$!ZBWTa%{wOZyz z2&lr<(dc%=f=D`miDUETa#Y~5z?yBR)&Vv6&Nm|(#g_RWD=apQZZ%5_g|$A9yxM4p zg5qZ5uDnYz0A%O{4>(0BKCQY7=p;*-7MHPUeL9j4cJK$_sAE>10(Wisy2|V=mW*6B zE(~}7WyA_s7fYZUWFHDY6sFBY(S0DUOdPtsbrfs{jcW_C2R%M!hk5TDQd>GMR#yC5 z&S{D^UYY2K9C|K=)XO!;JV0tfc#UUh4b>oPRAEE7c<=iG^$>IbT3DWXljluYEmQBh z$3TzGadV^uCZfqU*fxjTOgZayRA4_2s&TfrCG<%RMl0qpHDY6n9Ef^9y`_)<% z=Kqk`w$jb9ey9~(*#=fCFFWuuE3l4#|G#o3hnbvNk@LJ9Qo`Zz?MX9C)KJPw3YuYp zvN}4LZ365{*Fhs`F+sUvVkT`&-V8YPMW*tw99c(WVINRXP(2m9+oAt?DPoPk221>~ zDlo}6URLi06czL-H+ml?igm2m*9*tHEm<^YQg`w% zf9>CXL=JG+AYCy-%xOw_iXtBw*EG~y$zn2KHPg{Kp0#ta0I7$ULQtCp)rUUjcSD&v zu3b%0*j@7kQ2Mg~J~|q={#q5(84Va#xhgs*VDU7q>eKHmosTa9gI`$19FkrM-2%Tr zt7=lUKsDUwURNZT|1`!^?imD*3YaWr)kLgPL4{p>=K8ZSG0eP-8BB)x-Oswj&S<@N z_Pyo)CQ#MAvbvn)au`$x%s{o1QkGJrn2JVHE;`T^E_#7MIH6sU$cJFwD)$CiqS#pR zfxeMeb$W7}8c9M{ebf5g^3MNcO$V3L(^Qewj0*^8;rQVnuqPfXS9rPP!p& zC-`%b@H4M|Cp_w(vel6iPs2}}t*{vWZO2_@2a8f&P)DF?z1h)Kx|rzLImsN(RUa~2 zo%T>lh*p+T(U<)?!(xQ@y*nd1)V<(+lmR0OiVJyXNfl44u7)dMc+?pl^`aK9+sb|? z{E5$E9(%F$9F5oTr}5fHswF{X^nGu<(lzreL1w3Y&62G+8>KhP8jwjB`}e~mHDWQ9O6S}_zepoZQ+v{^_RWLHzG#@doA-KKCfCAS%C~m#un!>LP zzVBZird4D8=ztUSrlPkBtD!x}c@HG;7R~DstN`Urtfx*1FYrULTs#9ZYJvKQs#c0y zsk7c?3eAClOgbYnNl_NQ&%e*Li*9g7!d@OcUE-_xNR{DHC_WHySE6&NhLk7Pyeblu z4Knn+``jAb6O@fW5ze=PKW0bB*np}hk31JbtW49mwqM>q{aX{j;^GtMkS-1bY^7Ou zv`3V(k0SS}Xg$4OjY()EeQH*^9a1MUI+vwTO}$;7=Lu=$Mu-M>iDHD+fjU4JOgUC9 z@Aj&ktB3LuOn&3;-1Pc!l07AjSE))2-cIUeUGzq(%xA!h*>U`X}ZDm z;piD|t)i=tUVr4ZTo?mX(_D4C0`-s3K`)Wq#vT5C9OX)AqO}8d`1{A7 zioz`=Myx1#U_ZGnZq-wofNiiy2tCU}jk_F}VWieo1^37|gc?8adO@FSmN=DH7zV}k zF4a@2g0Fh!sbi+j(80 zTk<7L9le^AMzn?QkrvZk5Sosi&`XlM^z_;BJ0b=`4l&CnbkDjLwR}Qj)Q*TQ(V3_W z-*!bdy;0T5ERJl7S{0N#H;w6vXo|Y(*-ftwU*r-u!CnrN9f55$A#Bk5jAvN^oBH|p ze*8Ue6Mpn_LoSmIFN{lGZicq4loI%)c~tZ{QHMOv{b~qUz{ow3$Sta))4zUMQZ(S_1W0k9SZl+NqpJ^YfrkTi5MM~SVW z1KR0I?U5dY6#veMk39|Q ze~=$T*-%yb)Pay@X}oM34;5r_ONcwtQrKi3Cg-Jk0qz9#pllAMTd8W?0j8=^d^SH6 zMUHTnnMrfj;l7*jwVv4~_+=`7SV(qFARn9AfJ2lLvIx~w^l|zDvxG{fR=(Lr?+-XB zNeu3Wwu{{~R zbCny4!xYgO%weDNqH|1QV7~hLOzgRLO#Nv{zZ{=1#+UWdWInz*+I54iwXn8B8FKx( z_TbqMu(B51jpcbIdo3JIMGc4Ps8%X!pf4oQjc$2_#87C0Q;p z!ftXu!roiMXQ7ED#-BBmUGHg*%73xIWMt-5M~0GC4tqcYX4daErMyW2nZppZbHQsQUpQ z@^-z}{O_ygpNVRUf{Z#pF?g3>JG1@u+pqWh1UW;uq&H^uhO~m4oe8ZEZpttdqFG7a>l=N zLG_dbFtx^!`y!V?pzHB1{2ka6d5B?UiJ~C}(OwrUA;`{B{>Qo3{4Is|IGatgBBh6p zMWxGbFZ3yZt%!x6=!n-r0|PyMD6j|q3!PLl{kuWDF)P%OgzQ0cF#>oDdfRb_tl%;3 zjT18$e9<>$z*+@%mx=6^D>*tFSz##Vx=slsFN_)^gJ&115yaWQ*`Jk(WQ`6hnUEQhyVK=TSrpksMPp(ND1*(?KG#@&EWh4) z$Lx5Jj7?`=@br~2 zIkFCbVRSpk7i^69+{FZnKWonZp2Tq&6!~VLSWhX_D6*Q0ZdW{@ixtJ*KncF-e|-k{ z9aCFG*zm7H-Azz3Es@C|OQnDpL(+BJLg#_rNE0;tmaN=L_Hq~+XU(8-gi-=o=KvL* z9uTk2eW}2&Sac}@k2Wh?M2W#izkJjGj7K&o{_YRh00Q=quMSS}=;8H}bXZ!BlaGV2 zKB9!zu2>$lB}%LM+~qobkYDT5CfFO9>ic*ctxBuj?^i?DGVN23O~tEP75=7s%y!;# zpmX3j*oJb*j)jdGu-VFtU>n2c@34H9m6f4drg_AF(XGtt@BM8H-S^f`KtY#x8<2G# z@jn%`D>edd`lnrFYUp~{>yB!!EkEPuoU?+)bH=jgZFbQ3>vfNSFM4rjj1Z_XNxYLm zS3~M(s6;tQFZFA2X|sV?WzWwY$FdO<_Wrfc9gh_z#>Z_LA8RQyZ$-@=B^q&w&kWM& znB1v4x+5f??w$p;MR#U(gd{SXCdUZ>?@P(D84`D++xJW2JpRq(lN|oZ#ZSq>38c+z z>2IWz4HP*^MIZdiWvDMo6mRF%7y^eelH;zJwSfe%`+j+j+gb1GDaXkgkEW;tQ_qXi zLXW#<1@_Bzzz9h4S`xIHbn`z|G)1L?Q_#b^42J%gv^JnJc4@C9kB>T_llqcb=}{Yk)tBR`d)K49h6ibHoU_!G8W}(RU4fbO}b9+@Ydi# z4PQ4M*V(aLOKtUL3?^(00P*Bk#IF#z{NA(*iKRb?Ah$TYGfp(4wZx29KB7n;6&(wD zyGCY6V(>}tlir|f2CXX2GeEg+5Y)@BNNxogyAdFbrBDd10~wwA99ipq-K$<6uSn)^ zR)fPeHwmcSYrXpwcSC^al1s1kz9iX24ngzkWFWod`yw+O)P23r5~w}ULklk*eUEG- z?Vum5gT_1e$R)p&2(4-hyx6WNf=o1yT^3&N+(jq(9^`NHC{qtOZJ%B?IgvRFDq%Q= zR&}2ALZ7%8;bEW3$r_CD;6Na|7J#AQMkuJj%+q;AZ}|40>cG8{G)28&53gWqvmgHP zoTav~T1vr0qPteAdn1{Ta9JN_bR6LiRmq3v5o2x?kvb8lB| znt}-cT*n5SdU)03Q~51#O(#<*PM+Sb%3|Vx_=m6VBBe~qd=Ruw3GbGqg+2&5rdY+V z^cztG*~Z1o!pmZL`o!?qe!vP%TG(F=a7 z`E6s=YU+>M=CAwS8Q>6XbI$_B$4j24VS@-BxNn;vi+>^LB8U-ZiI2!{dESB&iq_yf zhwsuy3=!LBdqhvNf(UhXvES-1x?0W_w~I`kp$uG8@0=s4@Vn=)L3R5o`q<1%o*%mB zf%H35K|)#!$;W!>H5cRy4mjQ5=Ylc?T&`pEg7r#+X-al(7CC}1HDbV+o0SoP z;#n`T0?hcNdB2z)Zn7zF?b(_|$~bIOAY?Eo-S{D;tf5FH6^*hYDdF{?P-E;@FyPdr z0_}-lR5j76aMOfMl?R+o1a;B)+&-wmLU9@to4_syW#Kr&!B=zX#iR$+VDTEHDi-GHz*Af*9D%&mvC@+a72reVG`|xB6Ca)I(rsUo^_8E zipKqKtMsCc{adh=1#VF^_$ikJKr5F3!`IlT8-&yHf14MW7*?M@hXWhG{VM;9?y^9c z?n_9T!y;^~>jnY8etAj+v}M6>Kl-va+)=X)t3Pl0qvSbO@s0tCM`k;o^{uzhJ$BeV zx#nTaOQsdb?^l%z$odInx0#{eLMb5>vYCpmbiqVIPb3;{19`Cw$~>@`zlxXSyEOpl zCbwp3aEpRfA1NL?eK5ap4_1Gps9ybRAEbS!=ihtgIc#o5HX~2XwS7MSp`);y5zdy_ z7U|xd<7R@&k6XwBlEPtUWxE-MvneGgP;bEKUG|g~5$a^&Pb{Oq3EXO9p!Xs{N9F+? zuSim)O+=ohR;8PMXL7x>!w@?NDTp&Tf}wvq}OMo6+X=q<$Oyq|=crPau$GRelb%&f;GRUQL(Z&+2`=m^C7`lC=M6k!}h|u@ca@u;t z8l>OY^~Yk%j!zQ8|K*Q<9R168f4<=N5?bj*C=R2>N*Aa_9wpGK>_|Xs2vl-}9^%C+ z%f+`mE1^^ogSd6{@%fkLC61}Q7Apj>p2v~DVBP-6B13;Zn%<@yMvax;rl`%BPRe`= z{BHV}g%`_DyQlCDLlXVAqBo>VQ48`8t%@e5XkHReAC>~ly;jAAxvBG$hSmd(d^)VX z$qGJC%!Rel?04O3*{0OVrUglz5MNC;aoB0uZ??p2qm)oYq@$v%=tKrtQ&&CpVfBJu z*y3EDai7HcC%~4cD7;xw9=2Hx)H=Z`UaY@Xg+i{Fo~(pU>P^aKDN?~Ph`YmoM?f`` z9>{)iux&O*hJYtfa9EdH>jq>vu6OCQUwvjtwR=Gl+E%(QTn}x4EMuCY)N`i~L~o5^ z`h{ak2r~i9HU!y*MteRL|D#{GS%&r5V1mO7jTOoK&u3g0mQz{u)4igZ<5+XK4 z#<afk9FaMOut8K)cKu%P*r%lgx5V zdpyq0FJYzyp0kvwe?d$NNYM-mtUM94H(CIrC_NHZfZC$k>0>C5Pg86K9urbyQ0Ng2 z%YYL~&w%n{EdQn~W)=kY{y%$P0@qY}?(GpzNL~!t2;^h{MIsD~Ba5M;7H2xsWv0vA z_V)JP+ikwK((l%G=HA}+T-#3DLEIM*)Tn?4R2D%*5fM=dizx1bgNot?647zuASfa# ze9x0aC6PEekZ_~@I=|}4SzhqG|2*&WKFj~ltmBo#O$|bfLK`4?U@V$JQcZM=9i1Vt z3uMn5i^0w~0$WeVkS;In_!sl`=h)1MeB{~0fbgcd4(bukuTU((uRGl&9Zzc zTS9O1u82qgJzMN)&JZO5D^vB1TzY|^&9@kWwx%Vh-|ZsNMJ)H$gWhP75Awk#^Qyta z#r+Xh)+Gv%u@x(y^1@K-6xS8x8OEK|_Gwq7UHn}EcP9cLsD@do-tA|gk)1UGvd{O) zV$mIs&qMIHVygu@x-t~^WEg(K5vyk`ooakas~&`j@Unv)?Q#CkHF@|hDtC{;=0H+g0=^q)t%50iOg$VQ8A`zlJALXjlQ(4g0FHM&&Y z4*3})X`9J+xZ@a~kIUPz_vLdL(-HS_dikzfp76VjUka@E)h_Gi+Sn=U>HB00jY|_c z7W2c6CEVNjm&CX&?t*gdPI~`1%rRnBLLN{dUV?oRdr7=_%4V-Tx=hCI4_z4542oe; zlasA(mhA;PLMM31_9FQOD+XM01a>XQ;DqE+75v|MzwCg~F!|62-6P9Z-kx$}%0q%| zU)V*jRbi=q9eq8ng`Xw)Y|6^83Q%$?p9Z$jXEl)VUX`}+JEk=8PC|QL;fy|5lInuUTE6@pv(*26P^b$f5=B#1Etyp{1ad8cf)chlhNssL5wbzMV-HEkxeRm zxWYs0RTb(K(_)<4VZ+672(cK66I@1rH}EZyb%l$KLXvv%UD?X7mQx!&FDM5dH^!a) zK^nE@>wQ0Hg6OD*NeiuqnpTXEVl>nVmfhl?iywwG70-?0_!FF_!gWLDb3fd)z`7!V zOCZK|-7!ct4m3b%loFh=RaE>*|1|=%C4GuM#fj+LD1@_d@0%$}mw}FayCTKC!KYR_ z^fGL&VB-pLuHz5&{=Y8xKd)JfmvF&`3&;I!Xs{rw_v6T3emmW!GD>e_K+f#WVh*5~ zISLXh=;Kx?3PAOVt(B6O6b#YU6!V!VBU-({m}ce^87 zkGSGyq`cuJhQD}yXm(TH+IG;NeCEP#%3_ON*f9fLBEX zPNXj4abz(M>e!UnLSGYtYF{8X;e~%D^RTC=hE#>l$x-L{Vn4#sSOeYT*s9UAT=U3cIg0V_B?&H0sH96P>tG zRUm>!?OK9o`rWc%1=FfW!T~d`9$Zb$P?sF8&vEj?Mz5Xab;FV40Hkv z^b%H6%2bLZQ}Otn$JibFndg~lXZ?4)#xD0bGwmFW!q$~jN9?_u-iQ5fpX>V@-u>L( zzRPw3OrEmjfHf?*_&KiY-P=a@Jw+GO%P*DY0a;SF_&(HHVihLR?v60F?cd=I53Agt z{)_zdqtRw-F@I6`LDKKW3Gglq3uvDo5Sv*+DVI}Z2^HV{X60M@nMkTuN2h+j9E7zz3_QwUF?4AS5+UG5wYh3FFCp8!jA0{3&-{$rR<@|Jt{t5gaj>xF`$S8 zN)S7|&-ehTX<)rf!<_P~7p6+~kGuQc!|x@2?Z9_i-)i(um1tF!U(bB|@>^Xq8{>3B z!;CX=`Jz_g(ztv`0aStG+r{4*-aUEagl=BGsBDrRWI{9PwXY`%?#AdMnj?0GUk7bJ zP5i`LHE6 ziP$b-mM|LT5*Q~C^@C!TRbvg}F2$!#EBqfgV|LEErFj@QfRm0n8IDmu3;6AIYm7YK zb>_m6TpPGK0OVna8;G`PjGP@%M2DR4NE%ikg6LcxFuJUkluv{1#b371%z>>N7R!x5 z<;9S211NG5T%HbV8#h_sJC$awC*rbp#D;c=iM%dFfPx^#EjuU=(bs&i1%xe-pxTV7 z=#t(0##Yg;$rT_~lm+2`4Rc&FXYFFjPc18xyt>F-SYnDE{8WN!C4zE{hTy zU58k)VlGx`5S9cq%gV+r<$*{QaM7uXeOCuU!w0yvrs~Wri7^X+LH)}?xBZ60%`*TC zHV=mm66eB%8%VbE9^M$vzS)J-MmF}%c{KKQfd1KTW2+*RF#K=yo7ophu?apWUiiWP zO!~K%z0heNXo>&r8dF261yytbf3x%;KP{vdx{_Ok=f-79+X5a3)d>G`aT4^23ELIr z(`xyJGxtdQ+=}4OoxC&h>z;jXN7Y&Gc^-Xkjgea7r0R*`n2;xKtTT|n$%IffTLm9j zw>iHe5m_5e(YgE;{4Bxxkk)suNE68&T06CS@+m)TNo@39!nCU)t2eEVo=a`=XpH-K zPRg5x;4^VK5VbuRx>L4r%6!4CmTOuYKHISIzA&%z&$n@>le%Nb7Y_mAuqT8Bnl z)-SO^&%RHc40~P^3mE2?u^%W!aBf6dGt5Y_Z<-wkD|<)cWJE^Kyzuub>+%&Fbsz0? z4z=Ok>X|20z05wS@L+Wzo0Rw=k3OzS_d#k3C}-A1m@=}M>EEwD7?UoXjW=Ve5c@gv z=)ABxdhzrYaaCyh8=26fg{4I3ZuX9KXfcs})APidf63qWzI!|F<1{_4>*X!&tu3*} zjmsJn8-On3pI4Ug*Sw=uVQtba!6EN2zLG4-r4t3WgOde4k)59TGin9FBhO;{a_3N;%xZ z3DDG%zih$hnWj{FoTjsup_~OVvtZ zb_S=A{h`Kg=8_m=wN1J_9l8Kci=B&mj#wYKf;s`=&xZU;-NQA3=KOMD|G(}gS={Ib zU3X0-I*W>f9h7n#MGB~Rq@|qic~gi!lxFPgPI)@?jy0-~yYjES{L}Qu5KEa17myhD z6sebI(vO7(8cr05I^3~p5#n`f1Kj|gGn~wl7--{TnRtssF*FRw41|yI_}PRH|L8>h znpII|$SD5IZXnxV8OWTlfJ`N&gr3E6D!xh8Bg0)%azynED2U9Xv6#P|W|6)(H6}l( zOVI?PZjCA^n3QPImp9#!>R%DPbn5EBB;|I+$v2E_%cvQHRo#ocAJPp%{IgzBBJKp0 zkxW`2-NiQ;`vOdK@52X^BM~$JQp1MO;w7{HJ;aQlu*^?VN$x9ypjrzA?WU9^6s%V9 z$vmB$E%q%Ct)8ZfFwmEvxcec!=NlV6_fBMQ$n$x)Y?<2A(s}r|R)r-q^Mcc5M?`?D zh&9uz6gn|$I*En|($geaxi^|!6SPi-JpxIN)Sh-MaE6VSrvse%`_#GCE{)6XlWaIC z3E}Cyt#sd%1!Hu=6hRBKNU(4WPBWB4iVC0~=!j@nP@1v^dZJSV`M&3wg@QKTX{dV| zzBzdM8f_g6ZQ5;(=D72m9CGR_UytguP6oLwNNl8cOjRijaszXVNfcecRc8d||M?lS z2NBOd<@__85HV^}RFlja5ieCWiVc+9@`eQkvd6QrW585^;8+*b&WW@KCC-S!f7!2R z8zq1CnRUr}A$cAVF1)$1vAAGkRt}UVE%(+2E}YU6*#f*7$a##l1v@5ZPp%lhzz=sW zMLs4y5OWLR>OkjP7rQ_*z{@lre%Txk`?_tA^qip6zhQP$KRT2C19IS%aZ_6?+|*Ag zWj#esU<;y#sgjjXyB>ySgLLstUdK6^j-41ETiQnm%eNiIp@%rUZYLcXdmv{`mZ z%If$h##W5qJn@*{0zdYe50Y5cl4jZc-~!M>Ob@<8&lRyp6+rV9RuWT7AcsvZ1;UD! zFcx~ts*L5^I$@?HLzNz9+WBO|18}`Tj(c#Z9G4}C(n2c3xNVDuh_~kvXzv+2k@$X2 z!dAW+Rc^svuII?X_7g%t`k=;9uJ_y*dEp z2A%(%wfHC5|3WPubNUC{U*JLN$y~QXM)<+IWiPwd>cP(W2znqlvX1|dbPBFU=c@D7={{BT zW5LbI$-KFuJ4&SZ(WliQ{k zCbtixP1Syw+iMss?i=#gfg8;K`p#S5dfA1`Fi@7~xhn`=m4`I;nEs!w46 zj=>})EHd|mIrJk26Ek?>5^0ftW|Wv>8$f+3k9UOV$wqZ8)aNEgbjs`Hdh*DRZ58fO z-3X}gI3_4oRMC&5m@X`y+3R^l+V8e;0tkfju1NFawo3PfVC4l6zR>JkpRA}h=F!aS z{7)gF*zb1MI}4;~bTr;*%48!y%bJO&y(>J91l^5+t=PP<^RPYW^gKzBGA0AG(!fKS zS_2ey`tb)tbs_g6A+EDw~oFczw22_v?_zRUs}jt`}$f5l2^{UBJG%x zBf9Q6OT*k0RfRv8^k7n>e4}SCELbfhFKn|{Q(Tp>&gc_um8MNh=53Z<^2iH&7`cSm z9Xl^9HD<~5HlWyPQ_ZRuu9=`!9TOc8&1#gJKjny)^Jgsm}`IqIFk$DvgW>Ck6y=Wp znbSPF?n<*)^!E3y1%h2RAGO;8`A?OL5w|8^Su+{y2>ab|bC%Al{WnZl_Pe1b=VpFk z3|2&Bz?Y%l&6L4H!3mTT&@hXs7a-6!2JF2u91Pq0Glp^>Hw?dKDEyJN_^S=oRO~#( zrk^fF3u*Gs7Jx<2(Z#%Tyj-)1LSuy#a);de-5Q|C>4@laX^S+UmnCQw>J>`~m?bHy zHq?RGVYXBcYtX@%o#A>~ADkEV$n%lxJZPFhjY}%b=m}Motk}0^B4nCO<_}{!nNX+R z4aHE!;0R$Y>5D9k%{MokX?vkn8F2}d z8$7nbyUnPZ(=i2lwjU_aOGDmH)JsBBT*vE?VFHUgXe_Lh^AKe5V>o}{grHHy?(Hf5 zW}jO>Bk~H#7z1^d0|~;mQc7rK$f4pZVwVLjnP&RaB$o*FlF~l-qGILSHvSz!k2H5u zdSDev<~7BYsr13{C1*Zo&?KN-QJ-6``nqQqfB%$WzWEotl3`BD3q~H^B~Fv#x|y+< z4O8Ewz7?uIC-_!Kp{x=|D44Z)v^@iWcQtf4XJ+maJ)qz5YCxAtP zu3Ig?8}rclmC!>4ucJU}YaCAC9rcT!wRl-Gu5u~na$WQj8~m(#x>?pSr9`ZmoC(Q+ ztG=0Z!gQ^w$8#}s@@Gl9A}4s~B%w66+>4DM7F5U}zx#;FKR; zg~~rr7~V5|z7nN(I%5~Xerwo)`#~mahybwmbO#xEc&9iSmC^Td#|RW=Cna;-@BWi$ zxOGXnZrqtb@BRSIk{n72%1xP2A{ca8vOeU&Ohd3%l`H*hVh&^_+kh}p!@nX;jfG5+ zaGmTF&_HC3FZXPUyEtyIx(|}{>-`t_9Sofp?rg&{WK<0gG++Akof&aId&7*A@~}TG zB3oT}FLcCW=J!#`JrpUW;!SklCE~OGCCnqK+276%TNTkt>q0Jx``t3f_9^zb=YV!< zYd{?Ylp*86+Ri-v{H0TE*A3jPOu@?A|7zVe_lmeKHk$wRbha7^sxV8fR~Vod61hXP zywZ^TiOoR{0*75Qg*VxaBD+U)%#G}CvI8T{R%K&g_8L;eZS`^8vO#j(!UpLmCGhy} zq2f=38|ZRqknNy%@EXBzKu1Yba2Zf|i=2yr+GWCEuZ<+(pW;cMI zni+czh7Y3GE={`?Y6j3wLvbR>du0H0&;meZlyVmZ`=t0yu)9eV<#=C((D3X$y2IVn z+>#fD{Eu2y8aWDOD@Eejh>up-KV?Z~W39P{ITwxGvq)D|Ai>fW0LxasB=~dq+H+?ea>RTWPJcMbm ze`&JEpxC%)NWT{s|AadDx0IzGW;^s3_Us?XVs3WGg<~@ip&qb}T1P206j@2d=e~RI zeN#>eqJ66av!Z~a(;zO1!OA&P=&y3B!@OX7CXkzki{l#*Qmde{eH@$Lgz=g6zX!tB`v3CGMbr%`T&{?E0an5{+d#+Gi< zglzy_rW(Y zs@%t$E zFdaD*cl&K5Idr4ECbUFc2ERgreSjJT?{_N1YxFRh)L-qkbCj?-!<;p_a+~ynvLxG7HnE7Qj46 zDQhTFNyS4qpL)%NJy1f?uD%Pz0a{g)-)(u4(%9XH>;k=B?Z(D{eIY1EyNuZvazA#_ zYfvYC%WGfAN!UL@CB|lchUhHNpI`&QK43gb6k$HrG;EU}>ct_6EON#ItPRY0*#&;5 zXG26I2u==f(Q_*527{BG=dRZLc51L0M|E$lEGOA6Y_b3d16-C}loIkwMO1u7U?+Xm zw@6VjVQXwjOdHc@6lcn{D1t5LS=?c+_4Y#la1WWj|-<4 zDlI@(Oeu>fvXzSOqLzh4gaSGYW+vK>?ngLVp5QGgF&rydT27g#% zEvIt8;$@do%AFM1PQ@dm@&;eD?6|`3Q`dwTXgyO$SlH<@wZzm!*~MT@zLCTmX&8E# zOX4N+WZs-S$##C_SgmiSQ`%x;ejW2jTEfj;l)pU=TO_rSTsRym} zHs-MOpi{~qUNnvNK*ns(UNnruuN8gxnD-mAr3n4)Yp;>(+`JMO2F82~UZieH*+r3# zCkrJc0Mtp!^tfZdu!cMk#uAz%Q!7+$%yqtwKI?sCs)pI^ho#}hB0h}eVa2#k*g->u z`Nhz+uiuT)D;{}cNZDBAmr4xc>r9QnsH%?Fjv1?SbPyDPGQS-42Hz)D=gi~RD`NK#?^b=Rgun1Q!M)479e;|4qz&UYTOSiCAIN9kwkxh-wnBEb^c=ojOj%>Tfdl!0m{Mw9U$tE%%o%50&LVDxZEtzT2zNdR7+x<2ADljwDfbWdJ1CP8#q z@(7Gj9Mx84G9unI_%>zYCxa z|H!YFzf*CS4;%t8POJGhydiA|D$9J}!(9=$1vupTBYXKp;yU^;tfHM0FN>0S9ln+@ z9||7rUBfuuX94clgT5GT-RWY3bs-OY2cu)TO!S!1(aF3e@>Z|AQ=mQy>sPA5%3q4; zVjzXo0uG^uDJAs^q~t6RS-uI2AMcnCLw&i9`RTStvHfKqBogAf&g9=JCl`)2Tc#WT zap`Yli3=wiz)ug0#czeg6DAxJML;qH>7zGTs56elwFzsD+3Wndpcl43 z?D{09gY^kqxPtb{bI$&o8)!qmpT5Z&8_#S1xbP0$240dS1wlHoPJ9h&_DshWzHMAA51U+MU7hE$v>`U??M#fU z{i-aG(75Bl>LN5u6U6OOV|FVxDR;S>x(!AW3U!q-ftV_)HYxSv@dks3 z!2=n8k@vbmUJ_HtUmlB(FaFwsP-t>$6}BYRGIzGy6LkZU0xZ~J9I1Zl}-q+6%VyC6S#8?;f3X-PQEw()90k<7eRcsmm7e6*^4A=?OhgQttgifUI^8kaTG z$*vgSL%#vU=(>nUGaUddyRl+rQ(pXW_$acCmBBn9UX);?#2jqdC|uJ*l3h61QfLuu z(Nf9`imbydqbU=BiC{=nL(46kz;^Enk6VIO5vG0=4QO!}h}iDQ=;}CN`oDalh60c; zA9eV5a01Ba;LpVV)_Ei@4Jmx2P zY3g&zO|oj<-HF*jZOk3Ph(HM@oimz-4-~sbsEhbRL#)|=tj$wok=@+#NG=?!YqT&o zhbSe~p;jTOo`yjqg4u@0epuaG9Z)&p4#@sgDdsC-gVYAXzl(VZA{{i=nl5Nn+o#_N zXYa;TjK3Px#b-NXA4&JCt7s#2%`WeQlFeStvNGIU?N@h_gm9f0J6qad<5okeKot!8 zX|$>wbp|-lXyt~}WxyFoFJ7>$ui6Eg!ET!UlW!M}F#{=1|9h6qePw`@X#u2EN|{WN zWmG)QbsJPJ)I$UB-icaOGH*j{Zq!H_l!0?#7bXTiFha*?8sGQ*$7&$h9eY#xIe!hm zZAyR8z+;q8&@fkor$V>$(}T02#9@u+E@>HWC%;>~5$Ms%q4w=CxV9U{?TF%&w;RcdF=U&?7ci4jLa}w4iFgxB z(@X&_4ZlKF60>U}^f6|uv!bvQ3`){qM*``55E6aljdZvp^6ebTt9AT~e11euo}=p~ zTJEd6L#_8bTwGe${g)9NJ0uiSLE(yK8Fsm_SRjtYQ43x_DucG%Hb&!-CrbuKnq0aP zI(1R$;vX94M^5FQv+gC~0uUGWX>D|wED#_q{cN0hQ!Z(kCA81m4!h3SWBBYE$iX-L zROI_-p?OVt^W544WGOe_)P*;}pwu*=fF+$$uBFIoD!vR-Hp>OQ^kQ;Zk`R9SpRY(y zLqlVMqz-EFJ3I~rB=c^_u?TKN?1X7H?E%L#|H8otc)@UM&oBFpwZp;1TDb6XV#9AL z_H7lSd?ajO8zG>oVXpel6|MKKobW&ffn24l7elWBpf~S&n`XDq=%`us`0_LvAqjG|gy6y(=j%Tazo`;TZ zk}RBZmav%rX%(&u(S@`bm9{ZSRPT;@-GGb=Q6qGT1DA9w-4&U|L!SCnAgJFLG9Q#w zb}E}lQWU5?hSh}J;H~iJoH!e}&7j~rDGIVi$dgwM1Q0iP#@teYq?&h%fJ8X%v6Vn^ zBrBEFgqEt=&lC+2HKAwy``vP*un0m=vvYZm{qUy(qP>?eS>v^;Ugj04UCGyra_(?r3I+1X3L?4f(HlarH9s8+M{$R-Hh z>LPUfT=h=h+rdbU#IBD@m1Ou=MAtx+w>GLaY6z?32$?4?mdSW8w`jrWA13_nWf%9t zMpa`HQo|JcUh-g(;c5gOB7En zg#{+IQ%VTi=7Tl`6f&7iSK`~?F+wLiTCY;F%&(N}fW42N)(5lAz-nTkFGVE2tyR^_ zuTN#Sd8}4vO*rJqCI!w5*Dz_JsDO@JpANd!i{nnmzWK2eF<-LnvF~Sbxr-ZOe)#c{ zZ%r{H=9iyN-A0OEnN9gA3w%{mN*zV^QSl~hnQ94^Pj7*3RT`+9q{il`bu?CNL;1l4 z@MuBs{yKkyZ%NE8?`2+yx`ZhW@+Z}qh)RT4lalR}YNF85*%gpA7BXg-M`%RxFXRnr zgF?MF2EF9d^lgw@8KxKH7{m<*NG|w#{@$s@>s6j+43&TVz0+iw3tOFh3shuK%5@aH zSMv`1cBF~U*;IkDk-sFiCKMton#mVtXjLmbj*-n?Bb5Bax6F0EpPw1$oZ)6+-i+T_ zXI;JEvVMvULPI?e6Pl_kv(->Oi4ED66D~zS+_w$FX}z8ea-*_|D$8Bxh14wgwPzba zhPI7)Xf!J5=xCUuz&(T;OWO}tztsEu!d9%uK6}b z4|9pIYubi{5xYhn-c6h^LcRZ!JrUL$W0%Eh(HA)ZW0dzFV&=T;F|^HmJ&n>xJp_vp3M3zwh)2RQJU4OnAmX`) zM&JZDM08HdmRqY(bIDD+aN&XtF>S2a&yFgLSs!ei_8&dl_*-SRYCs8t0``Ypq}3`3o!w*6#_3vXr)S!`zZP|8w@?8Gf@fe4~$sGwF7-3~bzTxX1aD};xo zY)@oMWHVH{ZT7k&fKDNBJG|Ei?o%a*P*N6}mw%N8{bt*JT7@g#!ptKClQEZ!3OA+R zU3}oQtdeb}Dpacj9kxCV3k-vzWWbCL}H=C$Xf%g`tydk%>+i zt$a+8UMe1|2Xg5Sf1uQot`01wYeKcE-C-rZEdU9GM~Ts2K|O$jp*KU)jENOz54SFg=%FD(!tglg?MyK&AHSh+D7WnCaBxnGa`b1JnF~ z@+)#Kh;sG5-znc4oX*n*)(Wt4d$Fj`t!#QP)8~d~+NSimC5nn;^TM=&3qt$cwok7R z93$F5{FE7-AX@0z%V+~}WEe@FAnJWztGYO8yP`d=RhZ>_7YNEqepK@70tpoBfQVNv zKV7!iJ11(bcFib`O>3c-JoIu#VsJL0|GfFahwrxq=sVAhGD4eest+&?`oOZMv3o;KO!k( zKo(?RawdyXZlFjy6~7?#U`P?28~E6-B_eO!Tu~XYuY0203-VoHL^P4~@*l`_5kNCp z8?{OKxo061B#pq^9m>0Pd_?2b(3k-Me%OGM}9f5L24Hcp>&PI+;E@xXS*W~ zKZf(I+?GC%Uu+mR96RE|y#hAuNDn#TwcD>zh4f?v5=<`lyUh#Fl2pui7-;}rmn=vp z{2DjyRhliwFKV zulHLodqLD1>h3$A`S(pXi2J2@-1NQ1xPj-+K%hedOf7%C#+1*zr>OA2i)H_`duBK3 zQ`F1rX#BkDtu@oz)kv6%UGc-5ZHrg@^n)WX@N_T-9&lQlsM0@%d@s`M!fXu8UPFpr zncdHE3x`KXDfdxi4;8;EA_p3dc8P(QWratjZ!g2@_;|>6@i$JG8*mh=pLFl!24Xt0 zLWwtiA?}XJ2t4`jDYE!$>?2Qbg>>(<(icfy*y_L(!Q!t0Iek8_hcpVxWUcfg&xg@V z-)5U3_NG-;hW0Xf^cUh~jAra`G3W@m4Q6nLn_T$WZ!5lLoosU1ytIw<5tdJu(Cree zi#s^8Pf`75FUT^qy_5c}4c@7eH9?00w5nQBkvKd0!K88uKjVs1Gw~WU6`|;-R+Zte z^BV39zsLp*YxZ9>@G!4&Gd3$$oE-O$6>7;=?+wOGLKnzYwDF)jr27s1cr1Z92)$yd zG4=nyNx=lb8$XZPE;27ezb_g4C$jvNS%wNMmZ9~OG7aNG@vFvO^4R86qTT}7*gpkn z#nuIP5*C&9R(d^>qNWME4}92W*)r!8bPx8;=Wb>O8^!VGQm+TT8)SyZe{Qc`K{k&e z)fVrygi`LHfbz$z4~&8Q1koA!0zta$2+$E5zH>!7m%7F@#px9l5&8UeBzd|4w7lyi zS0-Kc=>^`TeBO@O9logUy*lu?7+Kg7MVF;FA#1nBzm~rd(khO(@7!OvGgk2Dr;D3a zciG6kI7|BhTR9b@+ULk+O4fyQPmm)h-zd{x^OWs_L!zDHVk zO%$@B9#&P$P$=7}@%-Uo_<8I(>I^3gkNP;H`SrKW24>qt-W^iq!jas|7M8D`Ql6m5 zQ7S%rMpbCF--Q{oo*p#C)zOcqUSST$=_FO7c|1OYoGzY0K1yFRsr zq^bWpCy~nKqqqF2Qs>nw?2=Z2eA)uzJu3nq_+6tnd*#tf7*4ixc#s;@MA|&b1#X-m zMWq}L+Q6D2bu4y*fONVrq*5%(V-iOz`zi8>if>_hNJd~My zl5%RBcgZ9qZrJ61RKA5?ui6A!Z5_V11Sf>qGq!nOlJ1RZhql~$*>#W8V5gemus7|( zjC10Gkb;mRpDz9)^@Y%;xR0+fH4wwehjqDLxFfbO=K9ofudbL~lUl{4(*0qXLD@4} zNG5NWq7UwXn``0b9C{%sgiTu8OoUxo_-av3)I3nbYZb4CSo2Qa$w`-Gcf?yjt*^}^MhTwXE-;-j4Twq;HzyyXZ%#JQc)mb#A#^KmlMJ-`cqjeSCMJkpkph_&qj-OO z-3&rEiMovBaPzBOIE4wkyaV!CrIZp9T-&MmIyUVo%F1fuWsz%sGMs;kIIkZ584pWv$dWfvMxxfpMAB1!b z?AZvtEJc*#)R4_{ubazUKliW`PjFh+M$H(dj<=>sbXjEBAU-URq{irhg$mQ$7besP zI`~U%?_7O3AnYB#+;V7_}0s-1P>#Z0AJ1GF;0MuBS7PfVp2|^q4w=6xcjn` zzv!W=DgjuEKo%JLQ%qbiBiLU5?5l4dCZ63tQb(LHG5W_znTxEes$A9)u~8>wYKBb_ zXqf$~U65pCmqlf{Bfnu+jAm?tC^x1?S}$BHspS{@HbB|bXP!>55QEIFeE@m7sNlgM zqd5K)CxDE46#vN%YuRX*1&9raXfzsn1u#aH@-BWAdp>{*)3aVtN^VXv`XYzi^|CDY zh7swbH_azUY&aeSHb&%xjnTP(`u!L56EEZkfyYQ9x44K4SKMV;_$(_a! z8R*@9y`HCNwu9sct&A}COC}3OQeNCVb@oEY#!Nfy^whDxy=%>G%B64Gg+XD13l;gi zQI$83ZdatZvt5c?uWgbYnR;B6?t?X#_=&BePllum@lY4hD%1w<4C{*NiA)ZKzS`u_ z>-@Q*0tr-Mg2*s?gxvMK8>5A85>yNBce^9*nvTz*gd?!PfSd|+>6$*E&~JZavLMe` z8l2^RgR}xoYME?EsVaN%WU*jz{=gk}ldm?Xx|xkk;e+3NOcJ;m85izO)LIys6iS&y zk)>3822gfmDZ}Xq5I!^Moi+cjFge%-3GOpG>?dCHx_!dhY2{*KTsR(Y!^NHJbCA~( zv7LXPl!)u--xjGiO7cQ}3)OXGanx0xN0Hs4#i6~P$TyV?vxS0`k5=X`xBVkMnrAHgqW2R$<>e>)lKEkGwAj z<&V=a3*TJ9hmu?m6O~~xuUVN7>sGI)iS8kXz7yIC&AKQdjX}G|f@A@U%Nr^iog_J+ z8a$;4`j+t85+;K|N!Wh3dhx0VkTntP3{D~JHQ!Q_1f9OyB}tMZs2;-pf3_a7n0<;p z?x;7g?RB`Aiqzb8{+msJ^?A-_*i&`ItTEJdYuaaX5b^xr-Ec8TCb|mZk}Ft zb8;;-O`&FFGiYYFF^TSTefr&UfX2RNqUKf67kKpFesB1h0r=BXqt1}@SH@%7W8pCs zP)gYGXG6N$C^oncrV(2W(nubS%+Q;t1;0%H8f^QIk;xL z6GeG+C#jrZpl8GXoU(KdWvR4UIh_(7D$JO0`(@8jo0#fx3kJ!heD%eN1^gS{B{5Zs z8$47p!CF42%&9F;gv(sp8p9dqxkY#8`hN4VHAC+!;zZft^TkG5P%IVAo7C@yb$3r} z1hDz0%X*+MsSO&D@MuUI?4ivl7c8N@7}`(a{uJ-s+ZvvEk@?Rf-G|9M7xr?XNq<1X zdKINip-2)H-{aX3Q2^>jC_IAElV;hGpz~ss6CY7?VSM>^&NKIk8s-z6Y{n?jNEAx~l>4>cU;d+B8z?w13M{Qel8#jGT@LN0i{ zLa=jMF5M$bQsz(0jY^Gq2tnEcQ5o#mGoqSEzuRuV+^ExHwpcM&loE^FcshPo)M5_@ zEHZXL$lAxW>+{fn^Xw^Z5Xu(EFP57jboATbxlC4a+t0XgMgh951}sdQC?#m#tf%7l zhc1jt6m9Tjn;<%JeJWB*9SkUwv8nvLaix&No{d}Fo0C8cvV}>94YLL!F=@aSzn4B5 zdUM^|)rVuXW9wp2W4aGoM@(nB z-50n&P;^37_6mMCT|{R>GY0l-o{u@D#=EgIvq*8;+jRFv&o0=|6e+F(|7kvNy-%$G z=`K^{C1X42V#jWkhXNg2OOW;S59c!{=un!2U)gU>fa$Wv#RfCy(n(t<^+LeCUa=|| znF`sA>GPFX7n3%fEmd^T$Yj_ogWwo{e$Yn#ra<;Q_`ZcP&NnbagvfAa->zplL1Z)& zxO{?jaix?9e)o%C#{cev-^}`hlvcVC%7tAK@H_^%B2U%!P{H*wbTxJuBly9U~!XuII(zSL_XNY$$QlkzHN<(S0sXF11$|u*{_#1$O}Xr?ig3W@N3!s z7jhkd!`L0)5?CwHaKVHNTM!$1HBG!T)1I-wv&IZw+G}h4D?B$!P64-5fe4c>_&?{b zH82tOnGs7LWWfA6?QqYncX`%43tWKV!Zi!KEtZxoloEOkH&O9NWeL;!+!91xK-zB> zn`n>j;xC_WfDRCpA?|m>>|85|pyf_5Ojr?EH%Y6?4=NHvJicOlro<#3?nKZG1uD;g z<20m<`tK_xJFM%UT-F`6QPI@M(?B=+$!}E8JQ{1DixlNPwclCw>*O~RsUBcP*sbW4 zWZ`7jj7qmA}=zwaT?nQH;~I%u5$>P?6+gpyyz<>60fN zJ$v>|%~OMI`@q2y#%*d`mQpY;K1 z9$tZHC1^S8q5+R&mSo zyRh?GY=Q30loFPw4OD!umw~<%u`Rj;sIx8s$6IQwR#nJf`#PHm+!wltM%AI{F6Lh3 zY|Pi7a|_gz)x3SNr$gH5O>s^Rm)Q;W=dGVp@xJN@sWYun;<8?djd?#$dOYXx%42&X zGXf9OJ3=bQ779L_yaL*6Fx`U5+76XnC`!--=tUqhUC zi!q$>fs>J=etq}+EV0>^EXla~e@T)Hmz{03unrq2WjaOHg1@9l5g^eTMms-u2Tv$% zVqTa|b)CP#cYz;#?Dcx&oi!G8D^~|LsvJ@;X3xA01t0c|JKTFX;bWAf{^MEJROv2j zMQo6tFZbvs9rSviq)9A_LuIS6$?E7q(g$e(!-s&12BDcY}0`yTx-dI1aVRm@5tFbc=HRGK^{O1^W zSwh1RpZ*g6!Jn;({9dZCybW?cQ?MPo!lBX(`+Hk~gXye4O#QBaO;8|WioUYw?qW%6 zLo}PMZeyxtXQn}57I|hHjy7-*Pmks~1S zWDMJuXLeW74eQ>kSEN+HMyFP`8enwG8ymGS@UxCr4#^%Q4lNKlcpZB3_HF-dFJ5eq z;G8S3B{CN$nb)2-Tk@Ba0vF!$9JW|Z_EJg^728F{SBjBdLaRPYj)fT*_J%y4*WjHZ z(1X0s7B3BR+9MNq@cZ424SP@vf6E(V$mq_fQ2(;gyA9-^bTn?7Gz^kduo)ispH7?+ z1tlfo1;Mwx(YQ2%Y^I}dV=o*AcUA^D>mYR4HXnoKcD{M~S1&!yUpDM{N}weR@PON& z5}rvBmCcqOnVPS@%V*~i5Lg&Vt~RdNaM09$=`ftIF>3nv|MqL^aw9Iw&C5r(R1~&_ zF7<2Roq^J`jKJ;FliYjWI2c+lZ;HDrJxQ)gGA9;A>xGT-?b9C;ROfJ%E%}m(;|eE7 z?}cprdR>`s&j~iKJauoaEGOA6yz zk!O2kGB0o3d^ualKQdLrus0{yD<1ph1R6_NTZN?|d%|<*-QLB1Y_g;i$N{$Vvv?U% zEfL5s1D!chH9*Cb6~*Sj#wL|=Bu2OcVUTs^$thd2I)w9_?8@k#?!d>wCl1TGwFk)3 zS7x)k#R6pMlyWUaRzIOtLJwp)e?f3=RI?0aZIdH5`&Q4$3_`#F8yC8~2AL6_}H&2c}!D+Rj@_l68)&|AKH!(+aYeuG|Dt!4HJu&a9lgj~CtM^tsOa@Xjq+@|!=ou~JAX4hPqZ&~qxy#QAM%^XZAM(S zKB=JP@2%G5^DdkHu~FCG>09ZW%wuapkQo#>0~ozRKazgM(=WSS*5m2jBX__FtD|~% zOS-Jv_FNVcHX89z%@+HtjNRW)L+_PSS=2u$iAF9Pr&=wmxwQ)zPrL}LXPxdbt6P@eJGR)eaH; zT=&+<9->vXz}mDcfYpU;QnIMFoiIJuy;uRM;pF_6XY;~_!0!;L( zh?eP4hK&aqeo$M4#vw8I$*ou-!fuuB3(-OhqlN5M8@NeUo8F%@XXx29a-S>PGc?F| z+I0&#e$8JWWKH<{ibV5mNZ979<_hw|cTQX2S3?@3?~@eLN$;3>Mf%A#<`gV>xuDcO zatj--AsUL1-+c66>>oe)@E3pht#l!!oJWy_XV@HY#2uzXf8ye|`f`6@HZ;t#;P=QW z7Y^eaEG*y^N_mkY=c#xklsGeOH|U#dm?!lD#vY=4X@eXsU~0^ELBg9_72b$E@+^|T zW=e9TY>T{AxLA}!7e*abV7Yw~G(&3`P56HGBkA5?tW1J|wCWu7b^$&NJo8Z20-AHc zG~P)YyFXO(q%~D2dA+(3M-!qqwoT|#p@yuQv}%% z=zB2)xnO1hd>nE3LO^jcilg6&iLdlBBkY%#$Sjij%0&D@({uoR(nd-N`|=DbK2ek@ zNfhmU1H745D6LG5si&{=6G_5!W6CH;t)o$c11r4q=tEfJjpe+!7ke=rUN|TAT^uT+ z>>GXPcg_6cyiwNclgn=E=ECcf4V@k&JJERL$y!wv6SS(;lJaR49z{O6 zA<)^?>xC6dt)MU1#(=rlCddUToIGIc>y;)&Rd}2w(DZ?YOUb-6B?R=suQUB_XGuoX z9l>7uZcK`x!yUDs$|a{n)w~4J4iWH&j_m^$VUsi+7D}Psg;q7+^QI75$y$X-bKOqs z>3QL6dF;}$ne^F!Yf(vo^TJP%41XxEiv&5J7N!iE&5)FMlTtr^3on^xV%=#}<>S5& ze)9qeE{*UD<6aG!B{6%hVepqW#gm3=XdJP2*&G94$q9U;?i{Ht@i2q$FYMVrki~;- zt)rCSg|4LHt75NDN+B!wW&Ab<@sp<77n>hm0flxtdY#XF|4!eE2~J;& z1{w)2a2fcVvyW}VrI7q{y)`cammQVs?iGm*76s&hgfPH8MHi!mx__*<1j!u_y>LX4 zVw`b1;o@v`JU9M#}snFLVgapB zDJ5hzPoQOLmi2YR8d{2~Z!WYV?LgP~euC;ghJ(h$Ay0Kbc0;(K@o+Z9tj z4HsK{2u~c~>t#T24L7kjcn@R?rt2cwX(aZ@gw$p?X;NyKMnzsgR_G0R7qAuSLDja4 zIT+JPZp*>!$v{3A8rd4;RkA)sJKaNi$O&k;E~HHVaYdZdk@f_291&+5X=|x{zShB< zS7g&e3k)xIY7(N}sRlS5HUj77n*0!bb{Vy#F6?SOs8XW$bfb8OrA46-8|wWRA;8P)+l zE(Xbk*~e^z{xHzI$FoU+mIWixn1$LAhI9*4>!y5rI3<9F%Lw^Zp9{Lp+h^^?41F-AYt!d4!m#qN9wwleQVwe z3@uvK39^$p3mx%Y{7M3$B)^LBY_TL!hD<>iio$z#`&9}~ z`SrVLK&mn!yg=F++rfT``aO~;I`Q0 z%=-DgkIZ|jJs)_<$t@S&nlBkWu(aVJrR<@|Jt{ucOBaZZ1#A4TO1hY29tw!Dxk`il z0Kbiw7gixi6oC=x4>?PB0_!$Xu~!JXWBT3dNs0QJ@7*c+$~6HjCJaq{Q!!iJ}&=!HZ24-I`J&gYt{Y36~OWA@PR%;>v0ZN=;4=3qN|7j|eCSs?NOrR=6i z7nUXK=+b~9P|WNmIliWPS8SzQ5ZWJvH8omjvD48;R*|%c?dqq!5UoPfHU?QluKAkk z#I!2heKqki>4&B#3M93p9WwGd;6pQU5`!_s%vXhn=?S!ane?sKj0(W`t7J-Uy|&P^ z`wjhgW9=6@R?V`y&}JVNOQTKnZ0vG=kr+9yH%wqNd{PBQZyvu@ojxZ$k*vJg2b1qc z^-t57_#+d8^3WdKOgq@n*5c`aWQ(>37qr_XTMfQ zsNd)hl2-1r<_YH#+Hqkg*#@gPHt3m(L3Q-n*ApRTUrKIH0_~MT<6eJ>90)ieqKw{e zE1lxrEOt7Q@Y^~RG&$XVzz<+I{p5($?5-+4dApIUcx5^kwplp0nUr!pMbb>H^OwZc z0euS04PnnX_GI?CVN7@ny&yCPmLJUYoEC59=fQrcn~VrRhJuih0LUoCm7;KKGS-*M zL$yJ<`iTxW6njWhTo!nyJ)ZmA=Y^-e*6X=k(Chr(6zhqz*E(1~=j_k6*BZa2XVase z*o{B48_4!o#%`Rjup5<>5}3iusd$WYKw$%YnAQRLYNJuf#2B_fRWIat&I`+zrU=kG zVS$*_J%yA(A94A#V(Hcp2+2VY@f%x3yCzqJvL!Kjbgg24D6}DGt5HE5nV&Ud+2g#H z$lZRc1Ixi7Dv)gE@0y50V8gXO_F~JvSn`0oh6%pCcP4$;x(~`myHO5=nsnkmAaBS5 z&QstaRNRj>OwOTurI$P)gvvLJL@nrme=dhtI^ZeiAKL!94&oZRZcJg7$D?^K`?dx{ zn;2T1@4qODManabqXDlbwm>C_+IXu23#pXYBxM(9HDZ5Jy?ed;h-4r+VApZqbvvNX z3A>}_o{Rl=YlS_REpV|R{ik7yeRl;wry2^E822RTf})N$G1nlTgzV}Wsu~G|m@sS71X3Rv&{_y;FsfcJOz&1i z_bF0ibKNUPpjnK=TnzY%Iz8~fv7yjqNq;6yU?;uo8JbP%E0AT$0Ahv%e#JcKUxFp% z5XqXb-W>$NBbNI$#T6@Zfuu&OI-oq@cZ}2udYM)_IiedDo>n0sU$W5i63{oaP01Oy zgK6{YK13Wp#k9?Vhk1dMO`+J3zy0iGXU3cBgM2WDLs}vs7ZbHh+6XEm*g{Z8x4nlQ z2E*NBaGVhPfx&U$VcyOS7-P2Vxc`q8Iy0qY(6z-Ld-ghC2V2(_fhWCN`PiI?1u_Ri zw@O(kJ9!;*$e%_pojMYJjQx0eYFpr7m|(Hv@pU;L)L9cZKOZ`C;p#~n@~-Hon#3a( zO}Q7n2|A=r1TWb0ffPb*yqZw0s@7w}*n3c$RK`yZH8HPMPIbPa85$Ox4+h(|#p}m{ zD1UP(O+O>@3d!IWN^{+fJcP{#R@QB$lt9p#gPlp4^l~Kz(zL1zGft{=`3t8s##KSP z#gk8G;tr|zg_Q9$P$ma*!B9<-BIpzp#z5tX`jF=u&&H{x(uQb5u=DUVL>xUaFm_mn zq1|$Iec4KDcjTqADK-j5u)RhPJ2J@qL%R@tTYl_YKq7^FV(Bt$r#(2+d2>C;OmLb0 zK^}M7Dcfe`@h_H64KbUMu*^?VN$wa@Yw`WtO({z#vV)3$pfE61^itB{TQAHEJRO|J z>z=$dw(9%mJoaRx>isPEvQ$et-|8(T>0JvnsQB)#kqzn1Udnc^}RPriSH|Lxdmu{{?rRk-!&n66H1CJ!zp0xs;o*btW1 z@GHPi$qNI`X!zeTvMm}bSC@G~1?$r*SeKG8UBlFXRtetaU`#)c#G#`6`H>uZX8t}w zt*6HwC)3`)6-}%uI>n)DU2I6gOXV&=F^7qA_ zk#CLFs`iX8;@6VWAP_(qbMu>P|7GzQph|$U`gKfc5KyKn7mZ8f-SOEq9SXa>r!ihtcp8n^+)h4;c=SOBGkQeLJ=GZlY1_U5FM-@GDC z06u{P>K+uOtq26)nw0w8QFhQf!1J!6?|2Y>GP# zOWP*3meea|b<$_vy&^4|d5t+nx+fdJhQYSl{5aORAwk2LxOK9g$P&7j)+=@feIZ>u zMyvYFa|u%qoX|!yi*eBOLIs4^k^!it#wpR1x;_te!F|wikxo=`qOquHR_{~yO?leVln`KDtqE(^PxPg9l zhgOC6q)qIP1P15)iOoT#OPDV(ihY`dF&@ZrqqIO~rFRe^)qq)Z9Bc**>VU)V-}>&} z*UhHMO`3vZJSSeT|#O1YCF+p*nH4~-U2nk~R~j9pAM$gvGDM}3{&#X#vo zqYA6v@D^hM1BPo_ps3Qs;nK$JQ=j!;N?L(;MXTBpwGTWfqY8?N1K7b!oa1og*wnzj zq6L-K0!pulxv?R)gq4!nKrz**x)ojz$~Q3U*c%F~mGQ$8pj|+;XvU7nKFj#k=Lil5 zm<#)c@S)z6Sr)0SzJ^55otWF6Epmb=nb)W)i7^%xdR7Z?B_GMR&*}X;vhZDgmnHYhSoTZS4KCGB`s7 z+cR!2R9$8FIUiN*4SC>PHD`q_E$F21ZmIE`z;i$P`n>+-218TobwNV9nQeINI2IFY z#IN#*k_}Mg0Tp$~ua9o$UiQ;>fz`{Sd5s~BAw5zxMDCiEr3(f`-9ZWO?BmDI+d$HN zQbaGFY*w~&&$<*t!%sdt#W{OPH}^Q%Lw8CB`9u8EVdrFcA4n}jdLg9Jtb9Z&+39q# zB%gH&!YhwME<(L#^rGwGvF-<@Y4Se!T{ne&(8juE(Me%H8zf8VC&EG2v)|kg@13SE z_%$mVpI<9?j|;?6PYTn0nw5wAxp7EsR#rp6`i`XZS(t0GP9FHo@I8|(x;5@ zxzeZYx@pVd8?t3G(~KC)=>`^@0zo6X(R6}Lb-<|`xKp1jw$74j^y$9p#8OneY8-3d zztMQJ-;|P$2`cP(;WEQ5j~U7oZm0C>oH}pS5nyJz$i2cZpTA$S7rGJJxyY7VAm~6P z*%oQPBzAE|aJ8tw^@tqSFt~k((;wPHuec@7XJLC{pZKURu4k$zu060#!!gxzv~2Cb z&eS2!5&g^#6{LHC>$u+*4*J?zVd_DAJTHkv7Y23X$UuIa-P(7h(Cj zPoF3HXGf%EkaNItzZ0`h%IIb6reD4UyyiKzPY|z&TFNK+Qz8qSp1l8TJ(cv+@T#y6DxE&K>$fIY_tZqN zVmysDLXPn-nZVhqKPhJXmjO8f(a*w3?;C@hI3wpedb;c>MTV)Ud{&z@lf51gv7M@m zgvKMcTK~4dX65>s#jrkIU9wa0cUAGyo&NFc9MI!yR@VBX#nC^#nsbF;3{LNr$lZ{& zY$v5qhKSFcmsgAW=)XT&bl}TrUs;jvyUXu_VjFv>;^wS};!fax)lnQOTcmNEj9@&z z1`M$Jd+GE#h-s+3PjbrVCoJk*)T|t2X+@18Ewbypoh+>THxi3W2MX6$RtE z!*E;ftrk;~cqaao9q;ebjf__!C5xlTMk)$>u&aSws#A*2)IP}pzb0{mX!1se`S;%< zv&4MZNx$*K?=;?XGnl2uUzIN>JDHg!`wg2yJalYnU_K=S*7z(iOG^ir##7h+_4@a- zmu~)c?T^yGoU*JtvgR9CBlrHIW9iPNC%#g>wEf-vKURMwd+D*C75(^LWbeE9?3ynv zKltT=W$|BG`;Drf--@(~>pcGSS%ei+XXeHE_=8&EYUjznMY0W8aqxABA+gM`V#oV2 zsNot@Mw(2?wooLVifWYifEY!;BsCb=S_;71eauQ-jIB*oVQEkfF~x?&a$HzpvKb8V zSN?ZSO~xsL*;90F1{3e4gmVKPu$`mYIAD^T=sozpsNsuNg_oEK)q z{1F3ayyDS@Eq@9#O-J@v7t@2%)L>P2P#mP^V|Yk(3LTSL&mnS>471g;%idLU zRCoZJQ@5@KdJ_I@MI)4lKYMSq9H5El%t7q`=O;zGKIkGiBsgm-|NH{U0QhkqrCW|u;P z?{Po?sX-do7q z4!GF?sm1luyVAApm+2cqOl9AsOO-ppsN4ixoSzx%UaDNji*eL?mcf?F+Gu#;y^J}U z2t>xbZnHyX5D~deEB>h=aj@-gU;Prf^u{C(ppJA*;-HI?bx`D{o;yBS8qGWCRW)Z% z$YEjHtj!B_3&>e$vqjb;Oh%vfZ4{PIJ6qJ+^RZmyGti*?uMQ7 zK+|H>p^aY;LC1_>T)~hPV?#KQp8(kj{9PC#IO>~6=jim9a=c5xb$?l>=R{!n68pvc zmKk7I&tJcnG63zV%hL;F#DNh~?Kl$wS|($F7Dvf8Qe+(!g`0Kk>(zyTV&|<>Oo>_g z+{bSLAU^lbR9wh&S@L(&^tRnPHD(glF)kW@=b}-cXFo5qplBT`WzGj2Tu@E63To(1 zsEj$SyOsp?b``!=a}EliW`@QtT)cvB+Y%QR%)a@rxD8>z=(|q|*7KBKFc^=Puiefi zW#fs++Hny~gOQ0iNy%y`qJge%-adYpta47ZqDOpQ9w%t!S8@hG9JpDzMw#K#C)-0G z3m>3wu}{-T3m3<#@xla_S`p1V1{&C?s9hOe5^+O%%n$D)@~(tvMZL>z2#4uYUis{3 zUIs5yu{Z3X*D11L=?&pS=L$YvQNxu&B+Z_PBo=+=L<{*A1+ zWBRN-BgpKeWMCY(>gltPBR?JdV34)}{%}>L=Kwof_)!>c&$0IRBhojOQ5|MDu%S&c z8QMG!#@Q=PtNdPT@z)I4nBM>XT~f=;5w>IF1Agxq|D=hMfi=2FMcs1Kx|h;dWh-Y@ zi=c(UrJmjD*GJZKI)HozIuW?lbOF1Dbtfo?&SGC=@j6W( zUfitI7ZcaAi|Kgp)v!~m=GOXe7vNJ@_^0XjD!8BW-)n$QE4BQ+D*p=Zd)R%26&Yy4 zY6U5x9{OxJK7`zsP-qHQ`o#D)o-;8~%ux6e^JB>^6QsTD)YY?Xwkb2Y-MTquWcr@C zqMG?GH)vQ%Bt85VVD3BZKLn&%xgq%>B`gqeOc5bD5k3>^Um9}EZ|{_(rzSM#FPkhY z-D6~VrhnRV_YX@AmS@q5PufZ4c+QO-?|)j1?8|vd1}#*lsi<}S>s{6IA!kfjJ|>%? z>m8&%F(Q1?e^}ZSfmAh+T2Jy`$?9@yTGkXX7+MsZz-fc6xt^y26IZIu!MVP;reWuc zN+$z3KX3N^kcW~B!lsDJwA#B)ROxh9u{k&?Vv{>)5c^en0<9K|HLxgMpu*qOFRfVA z4|{ZEeq&-#3|N>D#w@WRM$a$@o_UPjuOA6`?Q1W@SWVI9zaH`s&ujNEGNJ`i>K+=| zLt8}sl4gI9@#>VSU4~sCC(+8+C%Njy`fLaC-gG`w2gGsire)zJDsqK2h*srCo9C7|?B)r&#-mLvL;pE@^1+!N=`ZKlg;xA1*)$Mk@Gt+yHLbOSWmOv~gzU6icn(S1E*wKrz zSs`7JvUtE3CO)0+m{lM^{-l2PW*MfslESX=k$nr5#Ndto&?6BWl&Qe;r$Qi|6_o_2 zx>@0S&6mO{dARGwA%Ub%Ous;;DUCC0^Kn9QgASq_uTRJ>bS9o-bwO>*%u8r*H z0w0{BG6ZaBvn-uH>Wexb@RxMB9R%xo9!B2~hWW4z0fGt)Yd5X`Q#OQQZI6aVnRZyh zo8?2Iez>3-rn8sSC~mRd4CGDwuc3dPZA!*sx5zWY5_3Y9;=OHt4}XaC(JO(4azvtG z?{X`Fy3`tuCIo%w^6+72O~{BtzjUkXh(o$yyDZcD5|nYA6qZVMc{VF?gp+J224f9_ zHs=9I!tQlSQlxmVVWAKH)OW*zdtwc@N^~nEgOPvzvK7kemg{A|jKMudmW%r5lArd- z4StyN+%GCfCbJ!&9Tx{x8F^^Mlx!bG3V~w-v}0m;`9f8Pn-<`JJc0$ho55(rFx{$O zc=ZX4es-DHz}$8k8`kydb}m}U{fk><*`(b4iK_}tVUA0ViyA5mC*Wtjd|IuAu$TW9 z=6Aoh`1q&)SbY6hw1Job7$B4TUsv)rIpz8qB2ne;7v1~BYVXa#a6574UawRqt3zSD zNty^e#v8}l*KYf4iZX5MwA&@gOrvHge|7MK(0V9vF7Ztuy01A3W81LF&fF|mILqc< zx7l?@%gnUikoPZ{w(8j}Cd@S7p=;8lW3tF)r-Iz%OnzBS;W!&7>c=ZCJep&0KUV+IV-rcAK`M>b?fsOjh$8t^Q~_|5f7k(KQ>W?L z&_0sDE)*4o7KZ04ZZEtpjCViaS2lNm-6U>S_R;s`)uc{b%{dr8!0v&t>q&Ns@83_d zhn;u1RR$iJ1@T(9gm1S&ySyzX&Eko9**fa&L8H4)r?ZV($> zV1(Q0Yd30c{nlVm{6Bc>Ez)Jj+0NBQL?wfite+yiRMci!9({`R$ThTf@g)e7>CZbu za_ATiluj7VPp}_KdV-#v6}fNs$Yj?^ut{r3bjtbJ^&I-}94JzDii7s8c<&KO9!rJU zi%O@JzW3=a$14hKC|1n7wX?`5`gGWW*fml^O5F8n~~U;tjxXTSZBL^In++i|KxZL}aKP_oSw*+50L zh%j8ON}P3gPCKpd*|NbVc<7}e#kdKVtwk(e^Hx4oor!dGVIq$1C7tI)fc-d8SJyus3@&xt^YAU zwW3M9%k!A~ZD6InC5+~!3;Q@)?o-ad+#1#~zgUmmz~>z+Ts6B{nI<0=UJ>mzVU4xP zY%zxqV#y|diqW763zBPI_nmX0UlAm%lN330X(;BCZn?$qx`XsBS5symo`0$q&-C** zrslxBp%`c9JT2y1Ss}9x#^av~fAd?C#>{xwvA+S@0b@$HiYOU~YUV-!z(>Q+^)8mS zDX=6fj&;wsQ`!mC>BvumbZVoLeRKlvo*3wEG!V?+srV=$fv36xX(A1cD^s!HGTG}o z>Tt@dQ?l39Myry=;4)f?#-7*)Y$tv;cfP4F@`iBjnekHMq_G~gtZv1p(h-M2$AsV! zhl}ob+)pmNjkSKYtW^P5mz<>s9n-|yrYwBN7$3HQ!{DnUWLVF!~-oXZ79V za*dU(fYOeO{+r)AKDQO}{~ABwj~bCAhF=7xkQ&x~{zhppSJf0zF2Nqh2Xq<;mtrO# zOOvp~BwLXIbY*Hq6yp6H(7jGYtdl|w-M^$MqV-F_ zU?tnIG@2LBE^u0og-so9%aI`y`8;>|#c|Tz&WGeedY@TvS5ekPEct`H$!{(lF!?s zTTH-W(STNZ$h^Jua{ALx{;?z*O5C*2-T*Z%z;Bf-{S+AZp%`5X-zP!YL zc=18jMF16=rM*t7V?gkTG;lG#D4Lio-5&{$VT?G7od$ds`g*xo|52Yjx<#s$sF0eg z93=KCJ@KJ+lr@NSm;mc@R>b^;NyBT!gG=eA5gNNqHkpaJ+}Gz#|N0O*T>hL-#eMtF z*Z-|NaurqRzFL2x)twYle?BPKFR2nGg>@^6LgN-E3b1k0M%yI)%uHt5Cq1_>8Nj?f z8D}R>#X7Y=G~HU6*^}sN&yIVldiY&otH^FT4rF(bMJ5NnczR{qHtBKbf!|9yiFT^2 z`vj)pB_E#kZZVn$`z=JA{Zp5K2!k8!{P0XNIW&Vb8F`zhDcMPi)KF1-=n}U$tyGSnboSyx}v*J`#>%Dfz4zS*pANNDZ5n>tw^T zMjTd)&v`V!vsf&lX98Bs)BLx9huKbdO5<47qE1$pU(CF=#Y1F^?*WixNvF5?#`!l) zGFU$WB$>C~#&;$KD@I7N->~|B#Q6V-vHs<%7@H~cS=u&<@D+kNC zp&3(kLdTL-z$O`+v;=rMrz^LM7=pz`;;OO6H$bNUO_eCKliU|y zSTGdy)Tf8eXRV!);tPRUL8;%m86XiSLsp1VzjDw=P7cPO_e0MUD4L{OAW?ltu*)xz z)1}0x>%@6ZZDfb&sn0IAU4Di)obvp%61;;Vn0L8J5n53KNB{eP_mO`Eq%J#x+F+-* zpIuG&5*Q75A@TV}QHk3_u(%yTqd}IhcUG{TCdTxorxsh<`>CsbtuUbLY}DVkk`g<1 zY3q#ea-5Pu>tO{I^~AFusEHcrC!S9{FVknZu`^7steKH4P4&Lu(Z$_B4vEf4O8F-_ zCuVm_Yr=Ecw@E)6{g-se6CH_oOg{Yc|9&E9W@&|+K<4KpXSa7L@M4{Y^eP^2pBKZy z;q||0%?!%{F#e3#3>M@6I*YF|GEr1S?`J>!$6~(6{(H3?RTD}>uSz5H`6MeK6~n#9&`a95T+6wKGVL(bsr8B}<}6A{B-7 z1UkkIltfqEasAYz63DpsvolzjV~K{|U|YGi9$oF<$BBlb?K=O-E+G@am_^W@$iS0)N4D=*9;S+d zcDo5OqY)Ux%XB_Ecf>*8(u_t+N5woJSu|?>)lk5%Bk{q)tlXLXp}8|h!1)+;$aL

    4khbL$lt4cv4aX$uS4bw&i}2Q6wIme z{Y;S!>R0>0k$@;B@{+Ym9HCjZSl;J0Tz;`!ULT3ct@(?>U_%WJcQrjsB5cRo=)xJL zKoJ9v26JMCpv9)axbg~hX_#@_pFMnly&5nCT{GjEhRa$*V~U*dA3Y06-0Z|)?5cTF zjIjFVjn;jTd`%GX}K(Tog$)Vy}gvWs8qgAe%dz|c&*71-}*b%ms#jP^7SFPiH zz^{tJ6o6ilPUrGAQ2jpz6-hBt-!oM=r+3pSU(TTl=KN(Ll}$f$-k zyr((Ag--<(wQ~5IK{79w#+uG@zd?w5G`O7dKC50qU7vE!XBCyl+w8L0qh58{XQlf| zRfGHn&<#O4N8HN0>}q2KjT<M`4jUFj?bM z6EIz&*k+38s5q^vhBPTNyf#82Jlnr?hMqnbu9xM~m98aIfV7btn#wfF6TGiT7m9Q| z$hqZ)X1gcxI>ePBOCmLWZVifEq};C_hRCJ&$Tb(;Z@559%@0Y5SCRw=O_yXu=0YO2 zkl%qB5705)7}6lz0j*v-Xsv&POtXz&HMMj`Z;WBl(sY9m_eWwah*5UAY1Z*^32jm~ zDK!bN9FtY^+vq-0F|z_TXwTYp+YWS}qtDt8rYJKzvJATaxFu_S-bwRKybRt(p57O` zcqWBE^llL*vrSRibXhEh=#m62!b>4dQOWFKawlrLdotT1TqUUUTkUh0q=%JADx;xK zI5|PytD4%T&fv8Om(RT_eb3NE^eEjX{Oh0JyUX7ZbzsKHX_tUdk~}kmUdP|%u}oAP zRL(y^^s@e0)pRD{sSN_gC~wgw#K)2(4KpvY)qtgRdcB)Yjp5*`sXJmnid`T2*FV4i zOl$Xl1tl}&JaK3lN~0Ji>X`oHfO$syHK%4)B)R5HhD=OC7scMD$So=kRq~7-`*1^y z&gA@(pq&a(PO27mh*1#<1(GyZ1GAyN&d4>zri5og>!r&?XQ51?njQ*1N8oKvjZ9@8 zzS%ZMtK24S64uWdWEJ*sbh=8sm$4cWWl-_h448WeA@X<%2wssfKG_o zVgDC{YzgRx?o&PVD2%En&9VjRk9?skT`pbCH=G1-7>OHDuLE>N>}iKAviZr>?l~#& z=jTM_fz%ZK6_5Ug`ST51v;V89eVQr3dFC{=jvJ?~D5G98-|}&hW^!O%WeXd`7GXMl zR^Gv+fWgZUbqEef8-!`RwJys-S{3-YL!M3ddF+|i0V+McF=f*dRDaG5Ug=-vl>;ql zGk68ORUX-@1P_x7l_7e?$I`Xo-EIXjJ78Rp8w18m8LtG|B47RePUp7NZQe82tEYW+)?6Fw zv|BUwc<$k36sdP}-wyloqxG&r*hcCcc%ygKWFc;(*pn1FPQ_t1YgnL5RqBPoAkb&GIzd!V*SWh z9Hw&A3Xjl@avZ7&&&+~Fa%Sam#l%RqYS_Hs2)!_$7ZAd%Vrp5Ax z@HQHUNdl25PBRg;$8%H`yF)3Bx?kN zPAQ({er3bj?8C~M4Ky zC{U!sDtl}Pu@4xKnl@LB4q^!P3Y6?d6;{{?a56C?=69^1v1C!6I6vZdartN&G^6a; zCjI_*-C@sr-?d)SNnD}}&h$Eq1T%#N!J8AA!QP@#ttdq@U0`OkW z;NbcT#&|zBa72BQ>R~Sb^nz@yEC@p-vAWpGDcy8A@N>H8tI{U*(f__GJ*(ELitLv2 zusD_0g2>_kcDnHOw}XzGYx%z*u&|)~-wT;D$gt|vrMy}?pT9A<&O1YTN^}zb9dtS= zJs{XA$dKN6b~e;@6Jjkk46BOme($qJ@)P_vR$+8Z(x?3TcO;b?m&<{pF_0GD)i>pSx=App)@q*PJTXx_^apU31RL`$$>LX^+#vEkevmTnVh_-(XJ5 zk_Bb2UzO(a@|X_0Hsj;I5RGs{o;BCwx&G?SFYY$SiNg}QEX+I>5?uGX8=WtLDm)-` zwMDc8GZ^=^59oGYn&;KHG7ZlxjDUFz;SCkhOe5Ng_xjehyM>wF33yY zUGs-e%o|VbHhB?0@7NR{R_diwmw6ZZ-Dj@I(qtcsj*|g-DzIsxCO}v&9T4o{5E$aXSm zyp1nsu(kA_z>gthqElZLwz}2vFr1I6@K&ZYwn26%vms#R9hQ+Y zN)(Q|@P`9C%)Oi^WBv}z`?BDYp;*R35flxh&?S}!V*{>!H%t&g=pQ)77~RG8LXabZ z8Ygt9kr;MxO8VCfj&+l5DxWM)+PL6jan$i&uneD{y!ofR@icmHU=U@PtYB*>b~Qy- zQE`|w>U618mis}77on3&KXj|3`^2`hBGyirm2k0kIGf$~_raR!-!M8aH>2zBkPo>< zT^tw`Z6=_&K(Xg2a)yfQkq!Wfv&uJJqN#Gr0eUG`NI=tOgI-ILJnrBl3th2(Q9gFT;74k}-Ab2=o5u#R(~2*#=AqjoM{)nK&)Gz%AA z_t;5r*nXxmddO*QIFM{esn0IvfCRJLr@Sje+M{6zbiz+SSw>KXr`7()gAWs} ztW#(>m*%-6xSv_(T;;mdd6{}SbDGW&=+yP((v$^L@MpjH3~bZ(`fXQt%d=+b)Yt&6 zmn?K|Rh$p*@V+W)p|1wELi(!DZIfu3vo&~ushfDnSb-P(yEtV7+?(FdF?W8r4B)ub zMA(0g3>kkOsWL2LP{Y?OYh_lu9*4TcJ^c7tJAw;93x2p42}-@ACxja=?X1$U155`5 z6*X-UNxW+RuE0`yH58Fsi4yDP%;c_d0?TA9TOPde)4}4O7}vSh|M=!NNi(-~&Vixw znF&;GQEWR!^x*t}P)?J3zdOd2Q9@9Y;S1{Ryrd9K6O+o{A-cj_?9&^gX_WW7w<=LC z4jIdps&3K*ebh9#-~WJD5e!UFki;qWIP0lZYTcJaYxKT-#BiaSk0)wr9O|+-b1L4b zRrb4|0>R#zkUrSb9d^s{M3-oo<_7azk2>1vhFzx z36+{)DvM&FGF?l>VRz2^3Ozl%D{Gb&LoY5|;E-;wRc1k{-*D?M>_wd7c`P*d&Uqq~Z#rbf8TPIhsbG z=zpZF=3`U3Y8xi*to|Q=_Lqqso7KbF>_$!&+i`aa)8GF^f_VVgVdF6tLb_Sfbt_}m+XKi&VO-E*!- z903MyXEchDw1Rs2CjVO1A&+fRE&ou+?wRpnoBXw~11(q7d+(gp%-S0|*5PE7y{OIX z*XRRrvTD>9|L3Kj|8vFKw!2?~IyTgYfOs2jaSc2?g9!ga5QwOz%l+yH4XU70V* zhCbI2JBgn>-=OMP6S7yDI@xac8Z)qrSQ{M{<3Ib|(z4Fyy{}oLN6T~s=8zMXG$J}$nS2i@SInigDx_^f$Pz*l$_!087K+ZfY?EH&>$r7n z8UH(H7{MAy*y?w28HW=jCQbcpZr<07kO)bb_FZz=fg#aq0*P}JdxjznR9rH-z{G<9 z`BK3~Wg9e1)xW-FUJgIQ<><@Vbc*NA>1jbFbI$YfhPBjC8{|st&1nhF9n+VIw$EHY zb;s){K2+|v#2HkN=wznKHP7n;h-9^h4?%C%6wmvlURCCmO(%drIi#K8_w-2j1b5RL z;qJ%33IB#d*P|~V=GQ|8@ndQE^h+TXuxhR5UOHe6utxD<#&C)YthnY2*KGOY&qPLG zy;C;%U9!p<8Y++66mO>3jmYMWYm_(3P?)6@vbGtn3x*rAYq0J!g&YXR6s$pr7r7<+ zu~&j}`=1>{M3dt$7~}2SUT=pDqW%3lKMxid;Ssv=TSv(f2evy|Cb(Tsv1=)kOvUY) zGp}M!Q&c7Jv#Y{4`=@xO1s$AoAIwB;?B1zW;a4U&ti!=>j1hZddyEresp7ZK%y`~? z6x@qs467p;6ss`PYqy=r(3!v%x)MUvj?rptpGB1{tnlw?VwFRpyHp%MFT5S@JFN`yM^~bb8UOVJ6$ueV``R z6Hw^@7~cCiQx%XHMH!_WuGjHl20qUeHpx9n?Q-i`ROnDjZZ?2|P{L z#iqdT%ng0$*35K9uVfZQXF>(kNwH2`3nlV}GwQst%X^v(sYVN-1M3H#1<(g-i4;X& z<3~1dwq}OM;2DhZk;i!9G{cTt({t*z$jh!q=)I>o_eZkK8B}^ldNChREU0X+q2f-3 z^h23Ox+Ghwxh6jaSUMH*8SF}-p2uH<`_{8FcPsks7od3j)j#^g4H}Q0vj0y`&~V%c z|Ge9+9p?4*77Ei71lW!XGQ+#QQQiWSE--R|>PaZ7h&>D;AB62lCTx-5hK-5K1HB;F zc(-fSXROg|xIO5-NH%a|z&S9}dzXpz%cWRwjx(X(4hTkHBRP@V+?#=dQcI`AwlaBv zE0`5brdJ1)jbQ4yO;t0sg+?;kU`V`=!8JnEYn678nl`gzTme5~0W=%QwuhHj4eT{{ zcb}`8B8y==&GU>Wo-3@lPGVP8Tg(I^HSf-^~e)T$t95>rD|0 zRL5JOvVj4-d6n=1BWyZ#o(GD!GzgP;onDv2%M|te?d}IiX=FjnUf&{qK@27o3u2Da zr96ADnbrVkgtd75(N4g5H0&$w=g+=kTw|Ohsy!sznSf;52=S9$6boPYPD4MMt^W68 z`}xOYJxqM`3iT#s1zpA0h>z~-^5&zE|QIOUu-%DWB$S)fAyb#dh})4;-}>|b`<)W%J-i^pywx@r z^Wwn1$qp0vZ=u*UifllxW;KYF8k#a`P|g%Hivaa>dO$cJ-0h9R<>R>9f*zKm! zrf{|dy|KXD(pa`U@BeSkbIr8|x0`&iX%xGGBI~F)J$#QfGcrVvb}=2y1zD?dckHG} zJ*^MQ4Q-EVaA}XqpGd+Gw{O*!-`ms2dO?f_&A$7Vc~ymMg<-UEi} zpz5Qa&YE4PYL*=ct`bZ*Cald&PXrE3eg`*DRHi1s`F-Q|MSDv$K+bXtkvZ@#x5va= z(o-yS2)RndU1M%eyAPYGR_2DfH?~W9*bkb3h`R-CicYUi1@dYe6kUQ#A>Cvhsq^cZ zHRxoxE1zHG+90eGWJRqBTO>NATH>}lte1Io4~~%`+Qa+Ar&W>Z_ULR8o#Wr_ch>LG zaU1!JT_qThS5UpN*nnh_2=CsdsE`=$(5bUjt;~5|H@!Rdu&7Ru0BYS0!ae-6@;kKQ zj%VTAT2uA;*34EPOfdN5!Itk#`L;Ph@&(!M>@~qzKE*=%A&ZK`nZ#WVu3f0(b2sL$ zTVr^Y?~s#0g&jW|J{5YFsak}%`9Bi0C6K}vB@sj=HV|{s$8f!FM)h|21um-yE!dnla5!u@?YjmCN2iXffq6hyr=SixhmDEtNB~JTB7bLdYIaf zAu;|$8Rc!#k_o0gS?do0TAnfQHook>;`=`^7aM$Xuk66nZ$W@?{nW)l_-~W?tXik; zVeXUlQ`flw>kY#72V$$j7rnaPt(GJKdA62bK{~w}g!LdqsZ*a9FAA%rH%N5qZn~Df zr$FBV+oP4m4nZaypRVV(DK5_561+i@H0w0%v~6c@Y-noG>evh^mQBr`J&QV?_pNlM z+r9v39aSy94n-P!r|$LJ6p30oN2P}Z_VX<)%#GCuu<*RSPFqI6Xj|#7-+bqd-x(d0 z-#5M)N)Y ztA}n&q3w_sG`T>#U05946#i)h%HEXGW%MnNeh|iidZMKs>Cgl;OI63i)})K&sF$O6 zubB*$W`{ja2$NnObow|l>E-mG^q?W9i?h%9;0WjiZzMq4K~=fO10#K5z&qPtQ!CT^ zW~;V?hP+mp5ONZn?gIXXSMcv>Wvc0F|HN2yz%`GaMW6hP7t()HnH=z{YS0D4rpkDs zn(hgx^w1cj@KTvySPj+@F3?1L{NlAuZ<`BbaPf5es4uS%Su>?0paws@sQo{3Fj&j5+H zO|ezBolXq7Pg+1gt1JfhPquSho;6P#G4QP6Y zfpuC^OK+9oaM}E;(l)mgReoSvP!_$}KP@Po-mL7U7m_aeKEYo!?r9n4w{JchpC6ll zn_Q1Azrmw8rjBdO-*f-N*BB#+Hl&1HAnUn#%nt0bl$*Q@ITX8%A{hoH?%trpS2Shf zOT2ml+RIJKR%N~fl}g*a5?>i|LY;S1SjKSQLGROUndG1f0)G%VOb$i$_iQ% z-k&z^c&~l!qlyPpj6nJ3jo4EnSabE$l&jct16blYc4i(qovu;MA^oH9;>grdE zgHsvo`A{6JQ*ZKLAWHB)Ash-kC+qS~@a|%v-?$n9)1z27Gly=C)~R=q)@bY+@#s#w zL1sK~L0dRpSXg>~MR(#ED=UMVWo?RNmqyQesRmW@aR=EM4KF1&{%L_$KdE6i7<;bo zPy8=%`%a&D+-@g0Ejf-G%3yzA`G&dF=nJw--f3b;B7$l zFZNsFrnx}+-13xzA^Q}|p@jdT+jdt?@2h%W?X)71rUlC4^idmwt91$n%9(Gs_vv#xpI8K^a-VL~F3LW*NQ6Rr(U3bDr*Kjtrc#P@Cw zgU!jVIPNacQ?7OG@Xu6kJiEnbuk&021GbdyTZpVXkggbE5< zrgr+Af~92QrkmSr+rx5=4i!%J?4>WA=~qARnM5RB4u{ntsBVx4JnWYw(DHT#PaGN^ z&4rbycy#-O{p5y<*dOk!@-=RERf)@M$QB2tI_)=cI}0fmX#BZUTwBzZ;CNBBus!ZBXrYfoW6e)E8%;l4z%_n0ei+ zDQeNIqe|P-4NT_TO1zBpfy_Re3}>?gPI#eGHhdWHbt7Kh&)avC9OV`sao{-G4HI;wiu&C#=&!jEkr8=|?|xYDn|b3&4ND?0zfVG-FHf&FSTe zWOCl_70aK0+gvC7^9Swxft%%-?EIHsnb*TV$)j^%`(mN$9b5NgLr+5Nt_npZ@zLed zikaE;PX0$?%nuqa3<#10!%ST~rD+4}(Zl3R9=dhdkQ_JKY;%Rq2vacHa5g(<89IZ` zhu<7eU9;oH6M@w|GU+*@uDNE$@P;W_P#D!LTj>tc*ehnV&U$ zc5-W$p@X5x;8jDbFrcr2lI>zvrvN*Gz2}hA|IiKp>Mnt%!6lb&kUK2l%<}6#N)p;} zmw-tg8y-CG6_9-dQ^T#y`_=Si@dMtYs7&KThTn{DxRoa6>Gr@5S2>~erSJJ%`XwDsd+TKxk(v^(a=&eSy<*6z&TqgQmaaAK z52g9gfASf77@N+Tl{hl^Ha7juPan)P@B8~>jmX%LtH>a`9XK}AWHQ%>DHcSCYN@zQ zB>$E4z#?%Sua{RFu*h2%v%$Nce>ohgLWYknvQ@6@X5bx*d8_?vhCfxk`ObqM+?w+T zow|;f>bg_W^5xCcojF?NHPR5C3YjmYAS`q#_3DjT?DObeke2g#I9!gj*>hjequ**c zI}R*yg2o8v*Se2SpZ=#^^L*U`s*KPP0oB zpI;oANDdmTlj(VdPLtr9dS`u%F@r6A%u4wLeDYPmtk-X>+l z-^C@HOuhAP%{lYD(sLD;fj-M4)biF+Y%)a>skrs-4fH{Ni*U7fE-znF9lU~tT_D!O z+r5kK$saZQfGpfGKwvxNFlZ_ zX`|7g^nj@uRP1$mB#elQ6R6Dfkwmthv;lhoZ&@FM?Eme?2XG<7tvbp|=~$z#Ogj$| zcsaZ)W|_^~Hd~9QXWASa1C88r{h_}RH3dW8+f6bYcsZ&z!9g*_7E;jKCJyHt{7&`7 zP|e3)hR#2?Qm;I1!Kwx78z;Kq0%4or!j_(w$8V?)Ffw#{V*icPSe%MEl#{bZ_ z&rb*3r#s@6jK**PrKEnTlkI{Ri!;p$NES!2{e_BOeEePWjFH1`dn{y!`hV8*FPG<_ zd)iL20ygrJ!a?JS*%EdouonuRH~HU{tx)Ij_XqEeEtIw@I(*K_idB^!TIG`HV*7Jq zI8FSBwZ!&QZ8{<+3qt+U^NU{`j|FjDrl_SJB^j`@X(91bdibmTv6~&Lwlv6Z%)rkm zz)?er?3`CLoi7gBEp}&QoBU;O*ZnojoWk$07Q_N6zbpp(=INsj1ouM6=3y2|qa3qY z(0?RPgdZ{@TN%tgC&q4?HB8>mQss-5x!PopYw~rP9lQ|c0Pd=kiOI^dt>fKXGl)d zyKLH@b*$JKu?ju@Xok!29CBNFl*_*L)WbK{znn>`b!4KF~rg+UD-1G(w4xF~85|Be_mqHtw05TWcj%Ok#Q?h`)Z zVLN{0WJf0b_Gf=BHQ(tx8H06TYhhvcQy{CO8z9qz|262$J<^gx#!V8h&JQ{=3Jyrh z=vq=sZpL+3}abj4JotFrhVpH(E?fn60y z?u-b_@1)oQisVso=x}62YL)S#f|!0jnvE7=Q~3H>2_Zv)T4feEvpq3oF(_cGRiQGI z;UDb0c#z*9$NnCNLUu@ROh#RmMvq<(dqwldijK!#iG2!*ETdzT=-Nx~wB;2<7!i}{ zpS7M8Ig^h}CZ?KVp)6xB6_*|mKc#6#$MiPU<4+^W&v-?r9uSl(`lRi0XwNRxsTZgQ z1Qj#O124?^OV(>wq-}~Eeou@}eNuV>nx&p2n}bW|Y=uTdIyG{rv*%XR32c+PLULAp zN>pbf{L!&P4A#P#Pjo17iZVfHB<8gSNo%o_YTqgELEXsS0IX3<_DHA%cv~3 zEH`WtP{ubrzl_QTNy`jg0S}MoeFz43o>mFHPEhHxQC`49J6#a7JXWhb6S7A2)EuPU zFlKI9M?JP|ZaHCWQbd)H!pDd){p^UVWYcq|S6E?!yB!n@{kpQLxV>RVm36WdYT8^OVGmuzy%1V?og3z_cyK;`uB2D$EXWrG{q ztdp-^b_Ln*Stnn;5||RZ1C&nhMCI`gNqYE(Mj`Q{-WaVCRZ(?n!YcX^ z6IKmn)0?u^WcHb3!eLWR7H0V~S-0C|@M(vfP~$yIWl-R@-|As`0<4;4%a7abisf}> zv|aeIUtjs!KbPLNp+xOKaHISgYaYFweUKiOmz#cNS6o(AkBcEdZSZ7Td`VP5qku5 z!mGti>fY#hQAKPmeO7H3NLd3JbCY3yY;(N8fERy#;fLmm$rhxBF$5a#vx!;B?2&6Q z(wPyNBq@ym} zd&U|KEJ|GIekf+YYNfkf-uY*M$SSBlV+?y-<^vYH{`Gh ze)dxAE{c>=ai{4FK|X~32O{o)*jF~VsQIE4*v_C+qB*L}5MXv1ts!KzRX3*(In~mq z!)xiZfMu?l9;PxH3?t;Bnq~W4p}yKV2( z+oTH`a2G*A1#|#qkwuVAMP*YEmndLBQA7n9&=?W~RDuit_njG(85x}eGjz1~*Z#ts z?E}yE%=^9HyF8D@F!#aRU${K)bx|FV=fez8Z}5Hb3c(q7Xftf&#LFHRIABHw1@jON z6J=#sPq@XV^yve!vMiH+y!J=u`FhK;==|2BWUURS4Ix`IV#|_9F@l5K!r@vdXud)V*6}+L*{#aJvc&Rw{ zM{6JI-L%%XC%;MV*|8hIcHNc)11PPdnAH@CqQciWpYv;xG;=!VE?f4Gh`HmOJsUft z&8+p$wJVBlnz@Sst#nbnw1M9tg)IwxMU2JDeL-3BxS2V@>jfG~8gzMH@@Wh_#;pdn zl4SaUSVtca1NHQ!kB*ScB8IV=J%9FlHoZLK6>Gh5R$npOjV1)oe)nx-ru7#jPGW+W zJwNQ4EP}c%Z4O;>i@rgYBztco(F>v`-*xZh-InffSvk3am%+Ve{kq#n$YcreN6%UN zgMau3N&j439axne&FOK?=9W3s^Xr`-c%f8O7Uu+sk@e52;vAZ>Bq*)?{+@UXX1~xUp0w_X7i4jn zbLC$bjVCW-!*wAh@-n)7Z!L|zKf9fV6ng@nT5xDR6q$o;>-~1qBk%ggPhWjiZz`fT zfA>97VZ%GwRs$PWM={VcaDoaSAg7(;rz}Nzuy#I#O_l9@9aT=YOQNLGJ5SgsZ$x4D z6cLi-l+!Uxo)8J|mtOMOI=fBSDN~jE0cWiXaO5X@f3#enJ=9X>Cwnia;CStP43(7o zMRRtEqutV7k0_8}1*m%monm5j4Simc(68XWK7iYcLhc|-r%bS98o$38{bt~$~e(xzZ^wT&^|ZS zC-TIphva#}qb?bAwkS~oaX75-SIf_Vr1w6zI1$kIl+DPbGXk>$3pt?qPNyxbQnqUY zu{f|=jy*`Kot$*n7KxUU1+keey2$j&R=H%-F}(3q^II8o##l1@m<$NT#`|VlO+A@X zKt)^BA~~Wc{#!zsBseFUUe4*=|@LQqJVYd@@v_>sH`Yr5d z&F5Cq$)3BMa({j)43EKeuCa4mqyPn7qOkX#OyPT^)rw{2FH3$M{Dm|IiUK^1E)^akE? z%!6CIq;+nf9+F}oEka|HOkWUoKpr^)Y7DR}-#U(pMG#{?5X`=l6=EiBc&BcQM8De> z-V*+n#IUPKvEjW@p26Ncg<_z@DUk{X-pD!F_Zbc1a#>y)s4rRLT*-xU6AgVg7?-3> zdWE1BYG+lwP?C?`(1^haAY30Kpv(rwSZ_?c|AyB;R|4p4NIk!uU+9XJmzDJ3f*r1z zk_=_8+hOnP!ClH5(p|vcuv0SDrDiPizzQ2<`2}Nq=7;~W_wUAW{TIZeHIbgs&^d~T z1!}oYk^u|Zkg^nY$TI0;_7jN-oHDatksu{srTmj)^S-Ds<`g0?Bl;F8+|C z*zA`r7%u$L{2xQ7cB)^D@74S7|Aa9vRAcimv*ECaiLy2wWLLYjX291 zoZIO^r@b6iwkRI7?e5XZqJx3m^hA{^nQb;6ZFR5v-aT@GU~Z=qay!kTCE}BGDRmf_tK$U4a|>L10^9j` z9JIFV@$m>K=6qlWiV=6S+S^br@0R|n@ot!nsj=OfEMj8s z94m-rQbnayw|AnnRmFt~r@L&$;Q_!entra125kQO(tO z$IG_c*9p~J>o_zPz}E;=SnlKO;_M$76QtX0{KSN4Lb`It%)Q(lGwa+S0Pm>nJ;SKU zHGlPjLZDqzHSlXB)`^Iimj(KbIlQ8pXDZUS(F4M zjRmqcB?iLtd65A?_UnT`fq-h-1}F4*vpCgoj2p^iht!!5?b=9**IBxjUNt$NhgVil zZVvsz`z+l{Z??zLfNDm2-jd%;c2o2W6K}z(Vl_p!Yp$mVe_u3vxcW_`+)vU+~ zP#P#)jwsbDp_oF7 zFN2H^AG!IVNF`eUQe$v$IelEZGjtah^q8Tu1AD+!<<1&OIrpFvD_@Y()RGuCf3N%u zeB)&cV*Dv8KxYTic+>Q_@%l~K60&VH+}JMe$vy+z?4}rq0qms0vs?-}8tFmL^NK`= zR0!EuyFK7-;T{m4=Uw!z1P2B`bS3V-OQb!t zs+2zBvmZJN^QShzA_TgFjX(ujPUA!PL^oYJd}obMF2r(Mm=+oGr>KzQR!8NJ_m6Kq z8VkP7Sc!DEJOjf|yUYsTlNMcmo3Yp1ncA;a{*6RV28%w@&csnnEJdQJaEy>=3p!-o zf$cP=(6EB6mOd{$;`0xv3NrWYKkZFf;~KlSZ?k=vul@SFM`=AQ?tIVl6j{M8jcmhF z-4p|KZ=#ru6j@J&w@5(hv_-OCTV*h0cg?kbUd%L0&{(ranVHyTKkzu8SUPm$xG~#> z%|=s9@K@Y|B9k+uP}B<41)E(uW;KVliTmh$ZVKmsI0kCijzNV$N8kqLkhGh|dvSjs zA;4?R+{`((uuScky4^*0e~ttQhUXlf7s;yt-KcK5RTlqBw{u_6Klv_a6gP_6wjkiN zasKeRl6f(a>_pk*9j=}8i`=R>P0rXlxOvL9kW5Y?=LXr$>70Ll0-H+l+ij-lu=JMq znTEymvRKP`>ti4NN>kWfu!3ZcwkNjX^`ypNJ=sq&pdVI1g@1IHyBhL3$aaIh7}+B9 zN%35g1Gf60FN;T|bdkqhZjr}oO4mYmm%9hLOfl@DqX=D1H8_`netR;VL1VRT?8}9| zYI$r>me=^a*C%F-1#+J0#Xd3kI8K;Gj`+LFB8-vqf~12?V5(4b80^hGI)+4o@Fjjq z7O8=AsVU@uFUT5l^0_+d9km?qC<=-LJ&qWnyPD5U3iwjoPH(r*2FlSxiUL^dHjzO( zDPZetwH&XND-tbC>R>@kjRQeXK6-lV{yO7Ujwe?m8@4$ndQnmC1>1mD2-Z144;ZG2 zDkrwsM*ptOn)oN@#p-8F8lF9S@I6kxVZJl_a8dy-ILPS-#km!vD*mSa4CaMum26xl-!f3K}_HjiW9|-_adwal9mCyY|SQLa& zl!>(}tf28!kXiO(^1Rm)lzPK)GW>6wNg+GKVZ%#Jy@BC4LNOH-IZTCP3_qEMxP5-u zvK7=uE;L}&(#Z6uI;1!wE|6BtFA}P91Q_m*1O}aLoKi(!P;5}df+fdg_rJE}D7Thg za>@}t4cVp6Z0GNBkKq-B!7!P0!mKXXGT~P0zDvDxforjc8f192BnmCEK8WLYK|&zT zxqr-&7giv8=Jscl8^(yO@BjQa2VM0bYW(TGr6j?I>nHacz#@xcz=Z6g!qcEu2FnLj ziJSq)2ERt=BaWI4jp-5dqI}Z95ozETdFZ4uS|rIdTh%lmPkv?{un3tsgIfVN=6B>O*(%kXoZ^|EAQ96)I<_(7%K-1u#!zNK;4!`g=cU# za*M+9=wwBiL#g)XD&Vv~q}W5cp}1tb{EB!0qyRVA=Xnl5!uz7O*0G)6L%O7;Aw@wQ zvJ7#A;1sD5SD;0hX z9MuQBoBmatO64gavg~nPNAA1zaaV?}a*Jh-kV+8kZFXsLJ`{YC7jLD)*C(fpH9npk z!YXIj;p6z$yhfl(`WJ6+ zKXPhT=$N$U)V*}$+7mY09vj{io8Uqikk^2#ve7eYb~40Z<7FqM<+NIP2n1C$fn_tG zm4hD})EAT-v{SjCd)T`fR3=-+C5mlxK9_aKE9n8rAa|4Bl0*((N|c-sfGjwWTq}Jl=^YA{j?N%nhWddMyZ;@Y0JA1}_~r@x$2rEw+vQA*!hIFUotLQdqLJ86Sa58Jl6Bl7H;7}(Y^NtNw=&>xzyrX`h9{_zo|){{_Ee7cy=K-8+J>I47Nb&6azR(hK3EUt^nl} z*#dpX?fjB&UJ>U3+j(UevYq2uuob*WuWgH)LsON#^SbH7{BpX(n>aE8~?czMvLeOzQI#(=$6rw7LvvMYsq!riji;BArzUa%Db zR@iE8PB5w!S9uK0J2auB?5wb1Wy3IXcvgNZ^*_yL~IF6<*-l>Yzf|-%jpw!57XO7?e^2RV1t_O%F(Q9HhIQ2cftJ+8QoM24zTCl}zUd z`laWf+_ZDP>X^MoQUnbXLw0?#4NR}|LGNl#u1Bs%7u~hEDh$5uD8W_f)Ae!;F9jkS zjecc3WWa(l0NwQ?qW*cwA|R@$RCd#Gi&Xe*N&rgBvQFNyn!8cH%j`~#oSmnCW;J)V zn}a?1?Z4$r)`Kck^LrIp&d#5)-G3SQE+bn^Vkst?BI~dSU&lCGrGPYIiv+0MApSAn zVdZVkXkaoE+M_*c)pIp}i0vMak88teWfMMb6c5DG!K+baxn7W!IjoqnI^Ys$p#N3% zx4{LVhhD4gMarOi=fN#VSe`A)q;I?e5t&SS^=!2Sx=bCcsK{U@PE1-lYk|f0z4Fy! zz3q_x{k2B2@rAJ+dkkzxGR17CNCFj}4UNV9K>4UTq}0$@1)0Dg89FYNEs~yjTNyAY zt-@|N)gt-AKgB-Q-@;z5iK!eDLQITh$5=LH>})d>fV}n-Zy6BGC^atFF0XZrUG}EF(LIn9nyCGHP^<#O8Tx( zKR1@T7`oagQ;{kfbgc1-_B&3;`mNz5C`vrm_+&s<|2PzPK$=XEp-km=gNjMn%m_h~ zvvt0m>?h`lb=~ya9mf@R6Jxtg#nb+h^rqe?DVlo!$E4PV<8B&*HRL?SG*YC23U8Fh ziAqD_Wk@L;%{dB9G}`?(7Px6>WZu9v4}%61mO8*dAzJhvR|f^=qUmv^&WfO*s$^OZ}V4ObHy?8L#~K*+z7~Atd`>~*^4!F z7W8^#!aiQ5_AH#U!v%{OQUY*DbH5-Kfcz8Rfv27@^J~Bg@-OvFyZ`;)=#jN}PmMeI zjGeb@!@Hu@lSb~82Px(*Mf$;)5mwSjLfR^9=f}%d2r7KE4L=(_d!*H*0W_QXxPwrI za4c{jNY&2Q2|lIMt6aM2HZ2u~THe3tHbjeZzX7?%t77Dhso}KwVP{CB*BQ_W%2^aa z={RK~1cQ$8vVyQ&sxU0aOO>;z{Of8tMwkoWFpIOFt7?<2b84q?S5^e4hV0tuKL6Zr zRMOb!h?;4G+!h)BH9Hj1JXNFf5ucQRcv%kv5$w$AUj}Bj;DZCLF3y#qpzyF9mFpEi!H$dO;j%6P($N@(y{2EX(VhyJq@P z=FF<+l?Xy(dXT|W!{Zaxp1fX)qXO7XlXwDbT>m4)_P&Cv9YRVi= zF-Z%SbOmmowbD`dFj%9@!I93n0SA1o6ZJD6Eswt8PkgW}PIROF@-LV77^B5z_gg0B zf6y_@D@oQn>#%pl%tZbf#aTr%UGJPp*TH_eou3LNZy6Le4yomsf>z6Si90w$Ug>n+ z)aCX_V|t61Ag_|-P3@g!Jrgp{`M1W@YsX~e@KC|an${ch^Vlp}Ofc>YIMzrKd>Z^z zpE)*#Kt^gKPa|m+H*+T@EN3v0=7PfDp$Rx(1&T?l7M~3F(=Rw{;(Pa!J(I~P14~^- zF$XEKp9(+i+fA<~%LUi`o1r>-y^99aAmZ9Myp2Q_mDmF6X-QZRHvGUIPmO7dA9_`r|ptqU|t!>EQPD$uLiuOlJTyE=Fc8 z@`W~cnw~J&PdQ|Dg|c(t}wZL^E&$gEZ&aKMvBph#+^^pKZD zQ2ZEimc#SoF)i_l9iD%2x9-?@6hLga?9PM&$QJH-ULm(Z-XQN%HvIcu#Y4Mu-{10s zz0kC_iMsmgUd17WT6xTI|3t-;amv|2$r2zjZ;b!aB9JWy#RToner_>{$W_eO zM*2#{H)XfQ$LUkd8X1V6lG2daZznG3_eRO1VxdY+UHDcq{RJFW_#CIVG1#`8Cq3%| zHb^7I6Z_n>^>@063`)ay%8o#RjwN@^?6B~ZwNZD;yoomBE_P<+;J+^2VI0%5S>%|A z(hZk+sq)DhsMC!1?Ubpy0*_%hcr{46EEini=F8gs_XMi;xeYqS@G|I?DwNpS+0I0$`WKi<&*##L&4ZlmO(s6|$OIW-tN3o5m^j;g9Bu^!qc;Ib8xx+b^ z{u$?mQ9c`(lmHCCLBo=y&=>CqT||@z*Y#HyNulNsglgI$Lw?75!)#T9Gin;)iizUz zI*gtodw}J=UQ+DxH;h^j52RI?Z`bOKYa|%?DR4dO2d=w?!S_sbntKB~@LifB{eY); zrROBPdz`F%VPZFF20MZ+6tkHkF}l*LY(b+ou-hVu5u`g|O!$%yvV)4db`H&ggk@*jtkE{bbh(U& z1jvT{UlSxi7n$W#JE8kN!3VRO=SYF9jA;&y4=5IQ$nf~=|BJS+o`dncm<0{}N@4u0 zHIF}xdH*eGJbjKWe6{swpH{!4HyQuk_1~T3lNZMG{LH|ZT%s6I+&l*YoBTUK zx{h_aYWZKvY2*g$gX)4dS(X^X%jNW=aNHnw?fhhVkh>=^e#(#?C<^iN7o1V*Hbc9e z;vrZUAyB12+09i&p)a<);fW!;OC(!RsjP8nlO>Dt7u*TLjnx;Fvc178eDR=J`(TEw!qp6zOJ|*&#dN z*evawv)z8Xi?+`SwPLKBjv5zEo;}@TzC*Bake_auM5WnNVCF=l={I{!OWbJM zaUG>InmWDeVbf3a#$w)b&0NwtncOq@mUUAMto~g(iOD^IST36kDQDH?`KtvWuMMfT zBzf5k6^5HICY};d?$^iN%UwGy(tU?ZiUUO63q&g5GbTLfq3BxvLNch3Mc$7IO$rG=Y2zS#XhsM7k z!a=o5Q7hU$HIrlo7IKE{Fd<$Jx|UU(;Zn*jI$e2>-aMt%VN_X57Sn3Rsy_BxjG@44 zT5Y!|_TZf`kC*g#Q|{XTzvO|P%_h|eh%}B+ZQMXHYbmk{Xtu!_*~Kjl0k;FIwwGFo z+te5VM#p9W)##atIAjIXNkfGha^q^K7bGZaqIL=)1F4*35yocAg-uMk-+DLPK>eSu zG7L|Tzkjgqd8g~p%&75c4!!AH%}G!WzBBOdk+%-m-+m>|^P2Nr5<_sq(jw`c(0XgL z494(+H9Qp)@ROCnpw8ud`hl?@Yr9QrD1S(ch?Q#{b9MmpV@!T@VkK$KwqQ7ls zzx@e#<0F3stG92vA%q{LChh#^s$_}~fO4H%I-Sk!kfqTNfIj;cQ|6Eip9*CSez%w` zW#ZIQy33k`+9Pp)RZ>6_XU!$5l-hWiUB`j2NjMR7H!ZA ztC*mH3#o{(z5^qr5COw|P8Fv}iaRXb#wnfdqKoJ}=cN z2yTiovHFubQZ^hNGeNfAyU@eXY8 zn5cD;HD;^}l}DqS#>}+E%Y2@9T7@s)xFT*2P2m)C3pw@ty}@m=iwo-nnGRK)PGPaM zRd_?X$v#7=JA?n7AT+QBgOv@(Fd9sQVYFD)kVuvApJW5$Qb?xKmKVBRvzCyai4BUx1Kgm>zc@iCVBSK z%Q+{NG4|_0TjwHk%%fV_>W->88i<}ZhvK*JN^pk#@!)a}Htl1<@iDS_N&#qz4p^OR z+ME|W()WAx*~;&l)%VMGt@F><#ZMUz>6h*P%ZLfmFRWD@u2aknC}T=QYZqM+A6K;b zH;3Y`7wZ-C!&Lo?>cb|kXpuF_t&HyB;eYo1D=o$`(-$O=Zh{aQ3BsUiv4GRUhC}^hfRW*ea`eJWG zgL5bMia3RbX_E)C5|5^kN@&}{s}rhWYOC|jYQ=l$M~eEdk7r@A-DI<=g(XLk7q~)@ zB?ekiZSNGeX!Y3daLG{sXBYlB?9~GW(a0#747_y{V?Wq}$?DNFy;XX+?KeMv-q}rV zN;*g-T|(zdmQ<6JfX>h*n@B#VKzWC=g2_-+kVDccw<2IH%;H4)MhL38A5YH-gsuo^ z0Nv+Q&+maNah`>Mj5Mz;R!ba>Vg4U^(Ap1HT18kH6Y9usW=p;F``h2;S8pV{Y?vsp z%3#M@L@|3QvWE(+B01jOAwS3I0o6cp{3voIb6SPJK?^aHNHB*7s?0T)#_&_E12 z=0e5u0pCg`2o;w59rjrn(8n!X{bjiTJkJhLemG!X#p#eekTr#954TU>KE2aH`|%37 zLOLCW?5f4>{5qenK==jkESu5psfvK7%VrGOHFKImPsoqwP9usq8{6g5udD#tLaO8 zj`{07mQ>HQEhKL;X#I=~WtUUTA&MNJ!oPg&iul88vcp=5wq$w_)9hBp{akUK)Y99j zlW)}j0QR76O(8pe^r!Ogv%KE1(~n3XZV`4N-w#77Iv3ajt3?S?Pu z6b>=&6n;UBjENp%4UO#CIfCWWas(Z?i-s_qu8TNocBHQ=YW8NRILRRwdNuaCd^l>xi z+zDz9t?^kk8Aruy!y`us5(TR#1K}uil}7TaU+E<^oa2h!bn47Z5J$riR`NPH8@buD zn?qYlf9Pw$9T@haacY7e{f-F%0`!=SJd*_wZna!b< z{6g2}(0f3Ct-F@O>vI_LI!l)r^2M;Zi$Kg zR}59-)6S0&WO}rd9QrGNHSp2aORB|{5WPu;;)D|dBt}XJST(uaFP-}_*~o1Sy#>c> zWV@X5AllPT-;(7`jp1##zvn*?)D*gUGQx6z9H-B^L~}L->w1+FW_1BYrtTiyC>B_u ztpl%-KpETTbM|xInWeX1ciwkiNp{+BFsIVMXDg%_@Q`z<@H+0n`5DY^y4NwzbD!ic zugu|`dzV|B=XKXjyf#uKP330ElLVD?!_;aha;fpj)mDe^nSB>{D4U$mOXEC0ol`9? zQZzb8%(*-#clKWHX1{13t27&01_ooB>Zd(r+RBVlphD3rH-{LD%RE^|X~Rha6G}KK zyhHLd;K5d%=0w?V_DiP+NStRT(2piV00-0%;$_QuS3HO8VrCCWmR5N@aBHW#0u!B8 z<$lYl zSQA$1-{@eQH=)e!W-Y2I%Lhgpa64gR6%@k>( z!ed0MT-0Qv?4jL8hicvh;dyXxQ3dh7BNX-{r^2l8v!78MfOdv%b z$kt!&((4S|%0Og#cE(>yxR!Y@}v@j3tbpI=J(48-(4vY75Q$@+%7Vu@CHVI<~#nVO||25N~ z7|p?%`RlvujpIEwn`bZ)!MQc9jk)R8F)zvugEx0Ls+(?Sg&p%$9rI9BphdD4s1vLV z27_s1g$#oyt$V-<8I%6&mx11l-h?>byK|9jwP6!dVqiitDF)bOQz7~7c>5LfX_I{; z0}8n9{G@<3zjEOnSyDhcL~Jg5CCLXI>$PX=rRBmLx&Y2bc^?VM9~+`KkMR# z6;ccr6RSdn9a5%D|Id13;!zWfod;h|nF|RUH?r$>Jo%6q#S4)4itl+(SZ(aD-wSR|Lt=E7gZe}h0=Vkrhnh*$^joT5c$&)Cv{26w* zb@{5`ch7s7=8ablKt|~BJhgliXbm>;dg-g*X$G?D#!v{lbN_Nh+!VS=Fd$8wdR0~y ze2Z7_+#K3VW6~$frI3R^qudUN*O7eQR_n1Ij0G(tfMd1BRd#6k?vIYOeogPGw!Cp8 zgA}vlh_vA#%UJ`zwwhv~%Cnpb$2vN!*=*1v%=!)7> z=@8BZa^n_BqEkCR3Znc~oGy^_IAISmAL+^+u1K2=x*LJ54rLBFpnfQMUZ93w=71s{ zMM3-A3OScK+a+hjD&1ZRNg2kun;I|ZMgq|I^LPvw_w!HRGNu}RGIDFfv3?VTqLEA< zXd5FdF;q5mJD;DblT&VyR9R8p(!@lVJNG6Yx7umb1}El*@74glagqL6wS(ldGcLC4 zgOSq)#-)N{4pXF*3P1nO6>%m1to;fJ7Cfv{UIgl|HaM3TrmGff=f{L>;ob?pCELQS zAbF6H&U_2{mmto&Jfz1y$#Z$g0mp9Pe&0+*s;HFuQn;0A4Bbq%ztJDo^ugAzpMSL} zbmR#8r=GWvb8jxTjBUnKjxB;Wx6Y*>8B?*asa~*Qi(`VMMXl_m_e*jEpn%FN(s_jw ztV3P?4Is^}kzjH7)xaGyQn`H#)$+X#T_Ev3A&xPNE!`tPSY-O2Imr5iOzN9@3v=w} zmp&)e>@188yD#kq7N(J68Ypr`=e{IzZppM<`^xx$R)-Rg&EDPqeKLp#wF(agW1eS# zV8w6d^hjs$XtKqP?o}Z25w{rC?aCbL_?w)nW|hfRo!llUb-|2Kgy6dCc}Q8`oVJ(S zC{L#=pi(uR?qPPgqUa`iFREw}CZGCYkJca52uUa;t=uNL>E9@ClqXPqZ&ZaLK|4!8 zTEXl*0-*7SXJTo@4J%G;{%(*SR{y#`CxsNTgOv?0T!5kxtYSwg2Fhj1sPH@FiufxH+S`?yWcS7&xQX8BxqEt`qT@Qqj{#cyywPA+omxs8EY zoDZ+Le$*IR9rmZ1u!nXB!~Rq`{G%x}lin}A>D$6&L8WNDun3r%OQ|i~JuX|g|GQ$! z4M8dOIk(%bg&Bt-vo4B8z=Gw+smB?=f38ma7FeU=JXK{5eX?SYN@*lh6O;rU2Zf4j zG?c%T3$vlcpljS+N^GWqB{tX$J+6BnzTWaf<5CwB)gx%i57H+=ggcQ@JE_|Fjq(Og zp?3@jclXeHIWcnv2zClp(F4MCCH5gA*&p_VVT}MwgUtTl#PS7?u*&Rz7n>(c?@C5V z<=Ud8RhTd8k2>ayH*k%iPN4Lu$TX#W{H}&GeAv<(Sj*v>9o4)l> zV-7bS7uWDAX14kba@CA(gjQi52nmd1%`t%D*uMiKz78Wl8jL>n1Et`&(3oS)W-|pQ zSi^MH)|vF}S5PrZ^~sEhWB4(6@#f62!2=U?z{<|Ow8c;ScQ?H`dB5(ZCXzTA`q@T$ zgxM4WG&^bVJ^a=c@yEXRzIjEQND`*yLax7C_+?<(%u7Cvfqk;|%yD_Qe=jt{KkdH8 zzy@mysGBnaisas#D9xR%G>51eJyS2x5|78SJ>qhK?PEiaj zMm1EpMtYU0=3as>pk#U{cv8{Oy|e<#TQg`Co4Oo~84IIc{6$ z#&g^0&OqIW$zF+)v-F_s1nGe^R=?wV!Br+@9NVGseJKpTBBLCQ`CGH}*1vr?S&ybr z&F@uY`DBu8Fb%O36HSqIRCt^x(irL-Se*{33RT=JwOq6KCz$s#Sqx~QTUl8c+x2eO z7jxCK^_Z{|Dfg4~$>gX3CQ2x#kRth1_yF0!bSYQPi}ERR$Wv@rX!vpS&(d|AEt2NY zJDe2x0C@lzIb;IF*1~i;iZq8Nc@Dtx6(QKJII!qynCjCol$thkP})n%c=+lVaacH$4Fo zQ)9)h1z|esP^_@~Qf=1SOylag7bJ6Jq6jcu2{OBB(3ILMjt}UfKZSakVvi0^x3GU! zAH-glP3-c?0!5?FwzW!%>{}%N0Cpd@DgzqZ$IZ_ zK?3=GER!m#;>8Fa@c!`lY`pAKdJ7lB^HEIa+~!ae){GCR=NHkrLG2EyqHM2^1u4)H zo82uDe7@Hhi-8%ak;qgzBuj8 zZtv#MR$%SL75c;GQ2cR6cwcy8lZR%3$hxSsPJ7QO9FMG!}4Me4UJ0gC>5JZHu17W$ee&wb^F`~onm-+ZNgS2 zwuZwpuz2<;tA75=wesP;@u?plCaEvXqEun9C>2l)xWhSAI8t=%RAM2)^4EI9c1}y> z9`QkfjuP1=D2Dn(-VE|pe_eKI(Vzad>~EjBc6*9chlWj0GrI+fL(0 z%f2Hs|7>$!rI|xn^fuFW`?<;$zRExFs&T!63Cet|fPBSrh_qeGVI9NrOD7giH%^6{{5XMfJb=#p2$e{7RLw2b%fpBaF za6S_pTyyhB-*K@ z>-lQ=F^>)zp0TzkH0BpH4kHcXEFIOR}Ba*2#vWF$WE{ zPPr75MUf0DJkc9DqWWY5U;-I?kf`=L@ zltAHs3S5U-fb^>+E;d%$eUBOSBi5~_p?9p;&WBguT%w0dSjB<>(q_Y%qagzT+@hE- zC~}huZ*Wc*tmkg!ToM0$kHe>;)dXrnv!Kkj(;hTi=?3RPZc*3{*EF$)WCWZQq921b zbt+6RA$e0K{lM*l+p-o(9lg&jBS5Xp5fr+%J2lHAo$qmT=}dZ^LoxlC>jPO{AS_(q z2Q9^4m!?b>foqdF2bt-ym`(#)f@*oY(;8=JJh%#kRAqt_v$DWHIx9R0V1r-S!KtM` zlfiF*v@A4JVjIC3_cLUTb2V>iy0Tdw4}VS{T<@ljDKeqMa)6u$ekl$8I8$K>6paAX zxIYe~j+O14wCSFskk(`F&i6b|krg(KIS?5iQ5my|Vm4A_Jr%y!B~REhPn9&aY-X9m z39?Uuw7Cen02=i{E1xX;yXzXhe*IWiPh($2 zuEmwi;Dxb!*^ghl;U~+#c~(oj?Y7>0?nXZNjozRv-c#dFK4WK4Z1-PARvQRa3{uQp ziu7Ys(FSH;$QjtB3_3mVzwNKz65(D3#2pN<0ol`Y13=0| zEyt8unZqe2X#r-m>!q1=v?yr-esKo5Y4pe5svJSLFgjqPrz%sDp}e^8u(zt5yOK9J z?=#SE!}Dt8AurwUiIwg4+U#BDrrn7lH_U!+kvr%!;RqSr8)QE>Hz<;6mKFsi$uyD+ zzDaUbQBW`F4PYtYQ7#6H)$$>`N5AK!vKGW{hyQNX$G*Mv@UI4aLCbuH48Kv9IQrAp z#E(RJ@TKy%w2|l+27I{&tC^Z&k|?r`3eO9?6Qos{#T__WR-kE?pkyR)%CtyeSB2-h z!MdsrI5tby0PE2HiEx0%1`x{|RpSxE?dtx^ScmHc@qkUpiD8CFEpL-uTo~)OhIbv- zS5&-@ntU7LDA4;TtaDpAdBFdH7cAVN`P{vrcz=V`f$UDP^nmZ(DCC14`o%B$|8&nKu@zF#!wjB@i&y6zxYqRh4D^p zyF;2sqldR)ulSCE9l1_18j7@2;Zc+OXH{_qxx2)-{Z~$omu+)T6}3VSLmcUJyzO5v zZ4Skl9I7qt3NBLOy+`=vpp%53PD*=0oGzKp^1`DYx>X2%bpZbC@Oa==ENth);z37q z&e8*p&7oJt2r|@8I_7a68Z35^60aq9gD-I+ornCFpgIcv6#z$h0j!#tbK-$btyKGb zqcf^3VdoT9?5yO~Df_tiTeFna;x*9Uw2M3?-(q<{Jq>Ke*Gz-WwA=T`dFQ>J_bASN z(K@eHJS10aee)6&jW>n%E>v?-_xiMRh3iwM%k&Ey#}%_NFpl3bmPbuvV3ZmmYTeNG z-;F2dZNph96FKiKGwb>1A+wFG_2z1%wwL2cwYKRE_ORA??N5x!3LiQP@WdF_IKd7dZj1N6HdXKRw!D1)_au(pw%Uf> zVQ2szk&;ZK7$9p^Q{jgcYUQBg4p+6jYsTprWe!6#DrPi?W{WhE7+|}*A?Q+8 z%p9^i=p8RB*IocxEZ5_7wpRg^7MLwlHS!&e?KArsAKUL?`xGlPGKu$--+W-~xSH_z z+WE-+S4jiS5$_zR8(%;8`n4Zk54-)%zIQZ>R=<(L+rZTDih}RFa(KF>zn|-sBkh1*Knn9ymQXLmt}MtgH&zZm^t0Q_MPwtfsg&N1k^;jGa;xibnwZOH1H#0ERzA%N z5HHPF9Q`+AfdDo^eH#vlfHykAJ>5((F%;QAg%`NOW(QQzknt3HJaj_9pyT;}rBRQ8 zS(`R+)0h*=W;eCT2J`%gtV-W-gi{SU9H^K;X6}iNaOhv?vn>TWF^}K=M15jqDJJ#r zeyz$_zle>cu;Jo_QUe6#Pz;nmrBmVU{0>Ma8dPR<<Am!Iaf&#Zj)dlnJuWE_p}*{UjlS!2#-+|}$r82vuHvXQh+EA={!^`p zRaP*U&ShJj^pN>kD_KHfZMZ`t-v9)u6a)4yi3&%*QoBgx2zD}Yo{<6N+>^=#`((Oo zMz*L$mMs7$^sus*G(+Fe4oK7`i^_!?LG#1<25^kCZei>8@zI)*4qieDcZ$qKRRo6WZ zX>^NBBenkP*Zsp;Pr{S$%PMDBO@i%4z|u4Df#K zq{0hbQBUHKBFR(j+32^^voz%Tyn(66Q_viGoODU!WpUg}ew$yURsp<@d)4a;=L}_y zPqN1;^1$a3=d}FZ)2e&M4)|F@eSfgOg|Z@BW4nR=*1{@#5(Ngk~{&4#xz zWd`ucrx>Vv&7#7OLf-bnYqASq>|%H-tZ8fqVhmKu9I`7_Zk-oHDxr#5Ek{;_7?&EQ z3fsgIX2koQuV@kv{wwU)Ic$3C~^vj*k{y$a$-9lg`x^%t9_1npvr1+LU%Tjku=I}P+r6Gk7PQOQQo4?unq218eHhZs}-%X!ly8YGiT*b&j8sNzMW5bgbie$GK|G9v4ZTRnLmyAqcJkrM0ssEA7lbYIqvG|Iw|S< zjze}GvY}~dbM}JmLIyHqWSoed_{d3TfuoU+S^cxp@sIC>=oco(+sEU{{ujpCX);(} zYbmChB2`rQU1|P;1bc8#T=tL-k`7%Mxt!b5GvbY&U%6G%nVd{c4QI8>L%Y-REs}cY z1p9|}`$eCUVz0fkQH?xZ)EIh`cg&^#l>-YdaBjTv(C&)(fbT=QLBV1Aq1~5)^K?__ zL%Y6Ncn(8+*n@FbS}#Sux;#3;9*I20XTVq>Y1F0gDJRE;w)zd(-!)V0Dp8WlT_M;r z`=MPX-6f5&S5L2oUCSj-vDbiqn}4#Xn{MP(gR%(b_cptHO1ho*GO9LN1hDSxbM65i zoetTj-24SSjFr+a@E*qdmdEufpYkv(yKlR$O5G2Ihv(?c&W9g$945PMIDm7^z>plE zm=cNmaL88mby$MCTI`J}YV4g8)=VDUy8CzO=7;6gQCOdDH{r@)leC3cwJ zHRJV$@zl21?w5j?sBI}B8$CC{Zty6#gQLQj4rK@F+5Gw#!;N~g0@1Hd0 z=eF6sj|t*%{L*7%7Uq_gPB0<#$SE*qjz``9S(f;`b1>`3zL1cjnNfY_ z*c6h+?+e;FC1r|Q4%$Nqn8ieGkr?|HU>92ae|(oWxaMU~4vF+X0W`^z%%#K&_k2%#mK4<$F{ ze(|zTxXT5)t97Jmmd(;=u!7!8i^J0V7V5p5#`hxAN$F&A-oS}CK{4P}9Hqijc*_M* zJfs*^#VqRMM+k1Xsmek7H4d0eN`XESD6?^1^u1h2q#e!K<#dauigG(WqgC8ZWAJT< zOQXLE5Alfsa*CXtQ6@*S(o$X>+WFvFE`4*!;g|_)d0HX0!D*-#gryY%IRVCUj%FNnBz=s;?2k%~kHn zZtGogL8j!ovqq9A7zj#rJPw)s-SpwGyTRx^L-nWsLGLbzS7*|9teXgZgs6CmkYE1x zzg7S967f_p4?kb;onSY^T02F3Z zc;_Ugjvdm=vI5Z{q{6N}E8#S*H~8ek=FPR%9z6N570<9*>Zk`VQ*(@w@q(BO6F7f8qd*H(Ef)yn5g-I;d7;NSD zQ%o^M3b3(gP?pG#=4_u;>zyo03Ah^^H5qgmLUm_XPOhW}7u*-;ORJz}Cu(xCXi&Ct zvRa;D-|k-wLOJdHd~PLFQ+?`M#mk?HdPH}LMjFkj^0?@mJF}GfoLg_<>g6$i0Sv5@ z<-MKn+$7r1g>1x)^0 zNEF?M5YGgxO$KWW3t$*LXYGTVL47~_gB}>NbeVvB`J9#VZZM!r9-^506#0q@U*}Y# zJfj5JZivKTz!ieKDG(A|E=Z?Wf(CTEe+*G2df$Q*8E^tqj!?S`P7YksrpQ555;V>L zI5~<+Z&hDVoDyThMbf>&O|yEZs^yoxdR=cqf*oBTtPb2frPTpz84lRvZ;TME5?pt_ zDN6z$sD$*%vN*L`ExtORd;iiR=>T0Cpvr9{AaoDahS2<|eGu!gkir4^uEx1X6G=PN z1qWT*-R(-JD!W9MnKXGrOTWY8}}deA>Ot4@Cm>18@1C`@y zZVxm$Hiv%Z*yh(6x@vNlFP1|@b2RST+^`Djsx*sp(K<}qX5#xvzh%T3Ye3yJzL$LB zBsONUH6i;`LgKi)-qv<_Jop7J4|`qGfw>b!B>HC*SzS=k=)IW#s*N7B@}cQhJxU3; z?Rtpcjm$dZt9Nm(zkRciWU=#oZP>`v8u-4Y6a#&6Md`>7 zMYK70NYB$(T+X`eQ(P1@haR#obWNHX%R{NY8?RLJEVt^K{}mWvUQB~Zh*Uo|=NKMuVX_nqa7FNP&kE8n~rw*DK5-@3H~kMrNG zd`+eKcJgbt!w&y=`Qp`+8$%=Cz(%L0&@zWrlOJE%Np-(_5|v3h}-$73vr6vcttH&dlt^=aj%1zPPC|3T>ZUUVO@*<^zqw`VPy`t9SdZa z-@Ghtw{9wWJV+S}0?d8JLovB1T#uByet(vdc6LayUFV)e7;M^mDdtOxd`^X@E0Y9R zu!lT1=#^;`3{4?h7GT#&3Hc1%GgXaGyllx$-vpl=x?8ve;ybRV<}~p&{EPsdx{&H3 z(+A>S9kN#8!QgZlEGeMZ^`Tu4EO-y?);T3QCwo8OMe|h1i-%-HMZ(Ac6ca;2BHT^F z+6nQ2+<-ixwhZe!r&+$VXH+8;KDMRS@%xBJ$K?)~VB z_H!GjK~S;iD;@) zNBGs;YHsKJF?l1;dT*w$n?KX0$FRl~R+~d=!GU}Ce>Ut#?^rRDT(RNalm`Ze;!BG8 zoFX@ z0sqwq&pC!-Hc(_O6~0q>MZAfeBWH#6&O=@)ye59iI2#=e->BI~=0L*i!7O+qD@ahg z7JoGOymv2@`=taV(}|K30!R*tuw)l=PMA7smt%r!El1Viy287IbHpc6vXXbj>v|v% zr9rR1M!M2*rvxljN8}EtgqpyE9IdVicFHWnv{nr7tm1C4 z?iXZ>@K>VsNw5d&6T=%IDCHR=DhloP-{AQ%IK55I_51@)p!bO!1NaBU%TE1EuMco% zRFg(|CbZ;L5Y+*?nqcjH8@cHOqo9cC1Gg(c_NQua{#4Q@juzoxzDjUY2ESz&w>85A^HbvG080NjjdhH0uh39a}%e+@i=ASf8O&1J!c#4LCOVWy12L z$`u9!vmZPN$`#V!-cJaF)M}iy0MtP(5m+9786^hm3 z7O;|kNVzAlPFe4)mKV<&bWsgR`hqSI%%Py9(~@KMOH>;IFnby=OY}~LMvu$NdOzGa zUUS8N7k2Ip*&z?qpwkBCIfUY&S88@R=wbJv>f9g6YIeIh8xA347%Z9FCtk)9`j|CT`Cx3rk{J^8rR&P@5i&fkjl z(2)N9wMMe>g@Fc8%8u}+k|_rCG7_lp!}4P20bjaw=_zK7tbq@u1oq|70UR5YOh+uh zs8O~klRoFZcbZP?T7SEdb8LnAV1c95Abvq0%wGUsg{!DAZ!+yH0 zfVokIWamr&*{$lr^B(BY@eNlhYh1LF7Qin7MO^xigR&~iMe9a++fSZifr}>}w&EF9 zxR`YCr$0YwEC9-;5ypl+4ij>rT@FbOSR2^uTqLZRzday+O8k`Ng4_JBG+1$t zK@x1N(=n1*X%e6D=FOPnXN+K-%cjl6XhAlQU;pwmW462}*B2Z1Fb)_PifoDjmm`e| zM}?%1mK_XhaMATC==g;-GX}^ipmP1FkFWMjaMPVESL|d|cvvLOP*y-@Xzi@nVC28Y z98v7cDb|t=HUmJ9?ZK#@CK^B)A7;B5=ZDxhsy6JVn8+p_QB(-?gk6w~gqVCc6uMN* zpV&~?@D!T`mf@Qx@aK%3U%O{)KWsK-Xu?)(mLc~Mk|;({<#Y@KWlWgMDGC~<>s4iO zQg{f0wgshfbUpSubSYK3NIgPA6)h@sZIQ$}m3t1?Dp`06%$ojDZN#IK7XFKL!?9az z66!wu@b`LuHRW#YWby?&f7OQV$9jWY&qIp2N09+4yuleO&_wH&DJJyDDHnrpeRiNSOIkP6a3zlAOd z33TN9pr?HhSfw#D1IZ*|=lpA~y3Z9p2jF+$5K!eHtMOSUtA{SVA-hURgR?IEgscO) zTq2g+VLlXE`;zI0(%9f^(H`z*0aOP9PnWjmFPdON6gvakq2U9>OcTA+pwdDE3|>g_ zKV;Y7oJ^l1y{;JnXF*%9PiC`OpRbaBbn~B!q3EQxAU)czQdvvyf->}IPJ|$nbLW+R zfX{564YP`wdg`Dn-uQprZ|SYlM_C7NkmEKSv+p!;cbh4ui6UpA&MLT@zQl=|(-_!8 z4>)2M|0%kQtM*(Dk{jj1JFXtNiT0T@CDo@4fGLG1?nuj$Pv;% zW7XsoK^5obw4IXGlWTl3>C8E(KR7&6J-?S7Yk+xRd9=8V>%&vgVRVtzf=8Xs&-(TA zE{Gllo=-P)VkZf<&RzkvGyed~fphR|Kp4%`1Rb&h#7jLb8^87wy{(zITr-!n+OVww z7iL6$u$y8&qevGOen?TkO%F(Q9HdnRoVAiFj|3ke%?Fx%werEYH*&k^gTR^t9yEUc zkr(uswtnD{qGf8WPn@T!*IDhkA>b6jOtRK_ZsZgTJM1fghg?g{dYz=rYLWB?7YlJW zuIi9Il%5gXC4<~ViLL~tScp|8xRo#GX4n@C2U`C*T2@4}ZEj^ju#6zMF zj1A+SXM)!&^KoG%0GJq-6)s*1djI8;@v!~caKP9E+b>3QS|nLw)j5|F^Z!x~tTJ() zWxOw(Gv}OxK7c;1DgzdsOnL*8u0(%mSg+3d=rSKr25X857+CJ~j>=;*{q;*rs%P33 zl4rw93-oD^SX#;{<`6{=fZ!xVQPQ~)3oxda3K~qR%U(&qETlq$Mg&AMo$RX(x1g>H z-lUNhfl^SB5UA=gMA**16RdVRFO2d*+989>SScPE5bYKlghg^4G8NWU>MC!?jhQh* z!T=wm9gG(o>}|qy2w&A*mQwD~rjA9N_WIq+2B_H%n z6|L|@zUfNx`}2xEf*(`Fe=xiPqHu3-enUqHaVjk37plZ}`ooU>+w#Rtq4BZ=pL+h| zCpVGYAZ&9fhXI>IH<9F^US0xO?NdBAQ;{Mbw~fN-wIJFanJWX>jPao9!M_Fkb1_fn z29!aKOoZULd?#lYCwW#Koi1vpyE&b*C20`bM;Zx>%(%G|!Uh@Uj%kVe-v7p{9>&Ca zY%<(7Y*|e317b@nG9gBTJChRt`q#X2PD((wVy9w40im5dGf`oQiDos%{ofRR>8f9E z8h^TPDM@(FI??tUtSVU)1D!y-sBl!dY=Nj6IJuks%IVYcLGWl%v0F#0**|CP94(iqNPWL(j$0wA^tn}*@#|Kn#vIS)`thvQD6A_PA2cr+HW|_^+ru`K~^BZf7 zL;h?mhz;ksp(SXh=D70k*oHB5OHTs)2aM{FnokXAEJL%lY9& zbB#kOHoMI+5gh3ir#j^Zt$|+CHos_2JHLl6kzy}t2^82CF3eq+>U&bTA)wp8N|MO` z)T4n5vXFuaSw75}bR(NEA(h*v?ZYTVnADbi%fB+-2|rncX~VvSiQVzw+^aL2Uw!}t zK}(`0pZ2{byW+VL2e~Vt^;B5PGv8>_LVOoJ)KE!<0i@P%ZF%f-fV?&Q6abjYu<7sa{?t8lN+YnOv? z0KTS;EI8U249-}Npla{`DY`u$qtMji_JI<5MHXPM2Ffb9jC?JE-uBJ_Ynf z&jR-@{^$JV)UxH&AornN1hs4hb=Ld;v-c%%O=Vg7KH)u*iyF$|crkI(o>8a}Ko=i=5Pj@4Nf=d7u7u0~t4zenW%BtX2Dp05>ZXgkt z!UdEQMa`GELLQT$UXN-sG4exXa~TCprU! z>?l|Yt1E=X98s2wR-#c(MzMDCGsy)-lm8U%cSiQTGAK1zkE8$ajf-Tp16z+WBkQq^ zVlyeSnTl-+K?=Ay;V=ljuJyknZ3^iM%TYe03&fix-2x1jX#Cs5a-hN>8EUKBrO2g| z;!)y11^XQ?)Ba>(;XaLA@8D)UZlAl<7d5uT>@Lb8l`l;8rqu{i2Pn3lA~jgS)J@`E z9u%Jde)!CQtA1F0tCb}Ot#dmcJqU&RcrG)bB(h7sD|p1IB2??sCBGcp5rftEC!yG& z&FykME`NY`z-Kv=&eOcCQQYJ&o>`#y1gd^dDwn!}6p#E6ohix@ zLw~FdR2e3RxhLo|bHo*5PmM2J{!^y_1BwQJE?r8taYK;LqxC>FLt#aOJM zQ5;q@NK*OOsZlJ!B8AGR-QHM?(%|>lXQ?bMw90dxWHGZN5|IN8VEbH719x4Q`xR+z z*h-&ul0NyUQ-$}fpl**=ULIWr-6C7ycH3Jn8-eDjNw@Yyg8b9)ItZ`Of> zC}ws@TW5{P%IQXaKk#qt=IskUKPQQ#iOw^_yvKnVf-Vkv%178bkVX6(T~{dU?Ng<-*oIrfYFWUB-Fwsl5JNIAtqV7~}cHCV*f zuFzH1V{81y=v0^eA*k{R?2v^bR%uIj$lK`T9&JxEh?kk)n#xamH^9lJewD|bH-X;3aCKt3*JGxp+6KjLM2JUO!^_Ms;ASU zF9nu*>3FtLH!NL_ytSwb0hKACU14#dn<7+kbK4cz7mMw(g@IbS6;&wp0OI#S6qw4(_;_oVY)(+<*GJ6t)xeBjmh_@4Q>JdsD}Zc7&=o?$F_;- z3(t*E-S8d&p{Bk_ATQG>@w)y4b-EM%q*EsysrQU*hQ+7RIoUHhG@HlbsPB$f+zhy@ ze*5cfWQ8*+Ffs|7D0Tw{|6#Gn-`1hf%U!F6>Fk9Kz=Zx0Fkdu7T%%Q5DI1ainL0$$ zZT}lL`;+AQg`0hljW3x0!YBCgJ8q2YihtE_OldsK@LzOF3ZZhflCGpxt%5r)of2J_ zy6ra%&zvaBAo0w{*nGeX0*OtFUj3omU@+FFg`6epUYOl9h@y|%P3KZ9M0m4MH)#vd z&Q(Tg>0Z!@P>Yb{08>TCqppRV5>nab(OD5#$B)(joszQISQ=FnT&&X@w)1bmGQ3Z; zeApb~Mh`3>I=tAlp^(d7$$|IQX4((Yi9y#PJ5bHsm-U6K8iUnLit8~TimIe*!mdZO zGAJsNCDOM9*%=pOlffMyV|Uv3^>j%7zs@ljk~g~wR+2&o-X$G2G7Wnv7V0E-Qn6QO zcS5Zog-|4mO^97mmsA~pu#^#W#y=)0Sj z$25vmn1{YPdr;Y*e#y4_W z9;XG}J{@CCOxJ+}G3H{rzD*$=F;~2`k;Rhj9w~3%Ueu>Jt7wnOezWpd`0f;Naa@i< z*$|XP%V3Txo{Z98xeUb(4GiwLI^}0b4v@1ieyQ2b4pzuE5Hp5{E#Ccb%^Lw98Z6Ga zA0*Bt_qe$(4h)*rM!WBX>Fna^WCY}9fzI@CV56v#UNfVd#>iZ)`{C#|`V{+^UoP7f zxXgL%dYrJWPLDXn_S0*=Q0sn4{778xy-l)vj#iQgd);zb8h`9LC{)`m zz%zL-XVaI&&EyVQvS1}!#(W|`<*ahqr=&12*%O~VL>`iQ16d4GR24^+cn$EzkLZDV zi+#ZwC8jKPMt|(x9N8he>|Vp>)2I@MXD+yF1rNg8!-k*>>5_c+9Q|{SDm75=7x$+A zZ`DTPGY`?roE?^0YIbH)w$K2>4=QK;4_WEJ3tyqp!nc`XH{!1#R^J1y>j&rt*B48x z3Wcik1ayr+)kU>Iax)xNIeUR^tj2e$m|!MhcWhx+p7{)JOWT3F2mjSn!uaXTlLOl# zGeX4&{t*SQYZcw@X&!^*X6Tw3Sspd6*L|+PRvb9!ULMpP-bX)#^DTmXE;-)0E;&Kx zLQ~lF{%hwgoeNB>b2i(#&1e$n8S{dki2!Q`m`Bg|KMjccvcZtNcPit18HYd)P7|@EGY)IM6ARaqfX3V>S-+`zmb&}Bb0eS1f!=) ztedq86d83PUZ7drGwU#2!GQN&5LoM;;vW}!(B}!=Vy_JJ+>lTaRgSTuM>QhLV?KM+ zu4r>d2fbU;;-(2%CtU%V4sLNzEATdkJIh}=05E3BUDm7ZH>;hh>qwo;d* zI7~K@!{n?)byhMAWL9a>$WDVi!pPPM)np1BeHGqg(T+xeO?jZ~haH)i*udX}emKgP z<^lUeDk!L?%T>Al)$@DgHNJHrn}qNu#|uF!c}me|IZi3{>#vEd2`@UZO*0c-#5`+L z)IR0W*LKo(URQOwUzKM?{N=uEH}8?>j>x;Ns%r5~dBTE~{81-8o2m-m;5C@bx_Rj{ z-oNe&!DG}Ax_Jh@WHo+rzb<%o<9pSw}2Lpqt` ztO)(vBnl(Y`nsJWx*W2VHNJM6{wGs1NgzCVfSr!a`$t``WaTIgE7{T5zpfz_4!qF< z2WMRJp^;*tFk?Rz+va0YyEdGsFDoG_k(AAj2z?- zLv?XuNE3gvSB4;gN%7z8Rl{E|N)uH{)`@Mij2xCCj{Li)YF254!f6@4#4=DCMuDIe2A(|ZT?H=X2^!d4ksutN!?=8NS zz{1%WT*Ft@g~SU{4RWo2t$Sru18mS$xc60c06`;ekyNSt9g*qsR%v2LWn=-fitUoG zfb5sGIO7hWXSg-P<8vUhlIpl2oPUabeOj~uf9vyPS)|ID zv>PoA%@lipBJ~*0Efn5yX$$@|df2TGihvuK6Q12XU14n&{~`TpbfIw63G1qmv3X5c zceqxfx+UEog55}kqQ$}-VUz55C`MklDKClfTh!7x3{3+N^rFlXxUvv#!x}C4DzjnQ z%e~vedS>eMmlCUMf5BYCS+= z?0?qUdGmjunJbpU!p!Y~ymu%fN>T0!dgSUhpn9#uLM!OT+Rj(I9JZfz8qwN#7nQ+cHB5A^B*yH4W&)2_Tx~6d1T{*CGY-VZPEI0)v>2X4ps&oPH)a#0m z+k8tG#0y(|?a|TwxwB?DEI#)!Hoq9+@^vrMHmIkwNe*ln%=BzwM`i;(=hTff9?hM8vZKwta}>2*8K|I6~36?3B_Ae=OtN=0V~~Ljwms|$mD*b=?AYH9MR6N zUe6^vx$USOIA+&ow4fZM*anK!Qn4op6sCjN?JX(N8EF(%uK7$IS>`#w+sa$zVtftj zjWK<+$t#UN9^mvBW5EolfaYUY$p^)3^G{Kpk_{z z7U_^*M~phGM|MMgkGIpsY9w+5*b`^w8AlxP^we;1`%f>JrjodbGl{2yZ=s93 zIIKE2gXx>2CU@OxAnXg`T%H?*RlF>?HV|GWi<^5f`m(r1dWgo68bw9KRiLPW+$100 z=hGMD%Vp<78mtc3JvAYQ6ISEJ+=Hz=pK~nAGtgjm-Y-3vLbks!W(TayxI|?o#RA3F z4k}jT-y+!^TpK(hyBD=}Rv*1*RvMcOHR~7`9(}!a)*1RN@Oh?1K@T_dd^h`5Q|IV% zc@{K1oDysbJu6@0eEUnyKdE|a36(-!_{Y=V(JsoN9=>_++X;3^J6Vkyy|u9tBTwI+ z+B9e8Zw;uqK0oCZa@~Ow*2|0nx(_IJkRk(A?0wl4dFRX62>{J={@0}a^rdfH630Z;;0Rh zlgeF@N0m6PjUJt=@xLH01M;kVkE`=G2WNjVmp2gA5rd=p=!CB(|F{iGYqogfIINmL ze$nuF;EF9R4 z&Ns4485FyoB5N^DjqU9~djf_B_S3_jSVN&J%tNiWV+3imZt}6$Vf>tq`xwEffQD`669-v+(w)O z(1;7w`eJV#WEw)R$}wMY#rwW^APRe7rY7dX6%A+zCZ0dro;8n#3CzdzZ*Bb1hu;`m zy5=_ACP%nAJPsV3={3UBS&BVPk&mg^cEx$GJc$Y=Qqtu&VH1@g)p(3J;n9hikHnb} z7#?+MS0McjHg1j^*dn{jA3spdz(B}js>(f}I$Sss5;qgs-RJ>c7E>3ZX8PQ*uuGqO z#19#D>ZS+6F%^kf(_xQPm*PO6$u1{ta010F+d;}QJp!o&%UpYvNv;-Jw#gvT!ez)C~9f!yz7P<{sJ+qKxNQXo(Odl zee_k{V&NuNRr<^dpy0`cRBEbgs;fqE)m8No(-oe*Fb7U)6u4}SI@R)!t_wNvH45x7 z*boF=1bvWO)t^E6r?6 z7=NN=?>HH%Y0A0N1=j4$9M}UgvomXoQiq&dh-CR4jJ^^FD{Drbs(jXjp?9n*1|T&C z<6#XD9W*bhT2wi&&b!53e|aNY$~z+Hk5*x}9?_RWFD1968t|&D%D_)Jxi-VtgioLR zNxJMkiI>3yb-sG$1G1Le^5($tp9&*`l})i*DWaib3!_`4$U&SYW2T{gSvp8tCz%6bh8t+Z9^L0PlX(5+N>mYDQH_?nS9~1)rO(+V0WqQ4go=4v$Hc z5bGS2$xJ!7d)`j09yQ%%zaRiHv;RiAEnF!!d##$YB|0}^)G2Rv!UE75o7FA22Sn%C z1=9qX%_#uI*i><|9%Fae|Mm9#FE#&tEi&CuHJ&&#KXRw+gzIWhn!FjF1FZRn=|j*W zyPLNuI^UyiR-EwB?9pI6gM{X}bb)6DgCpz?7;Z3mvcs`AP?YUInEkH77pe3+AtSvn zj4zUC6jT0`V((GpE^eMWAWwiX*z2zyg!JSQ(iAetbbGV}-+B2Ua0Q(9ix=h!`=X1P zI!Sf(Wv`kri0Q(=`E)6-37RrD2W#nJ0M<&+ENF+i5nd6DM6o5ZJ2UU`j`BZ|!D#Qr z!qYA(VQbx*$w6Kb#DpuL@9whfk$gaQ7_Mgn(e1_PAxM8V@be_v$UYiJ?vxFQ&$u7t z#R=QOZoZ@yXkR-%yM10MuxhXJUlN)jIU?+%2V@0eH9ylgksX09@fR5{a)lN-f=ufk5E^ z>L!F9m8^fGGIZ3*9=jW(FLV4Gt`k1~UV9#M+FDO9{vUp(HSI4Byh$?C{<4m@Rn!Lc zdPr5$BVQb~J))G?NE-MLe9%mNq*xN#=h5u9!?Rh~E$I33C2@{$-K+=Uw_fg(r$twK zHV607_ypWIZSlstp@NjrO7I-ALe%nx-8713X!TtJ(ZO!XF}|^xnDoqZh1;{O$)x`j zqZ^|({Oo~gDVM{R2b(F|stL=B+!&nY`xw@xgFY?pal+!LVP3j?lmA6=ruY%i)To&= z!EK(_H!&G~b)Wl&<9vVavut^hlX;>nJyk|An0$)Op`cwN z_P|@0#DBXYuUK>+T#$>=rH~J}<9(P$5Iqh&>9b38Gd#~VlO80?oUhL5nVaO?5mUhQ z2dwtI8GZm#0XU|?uDflHz%sTYwnUQ2vmWo0Uo$NucUVJWrZBv2hQ?cUYJQc>z|Pky z1+^c)n;t661Z2`{m8t?){r_yVDwuqSPrv0n+fP5jE@wEIk?HR*D*U>?!Hg7-{4Tt~T0-agkDbd1 zDqVDjDPW)&2F7Lx+arg0qk-*_5b}jGzx}Y?(Mk^6K~4;y!5Lu221WPI9tb=7;kOyu z0GYw>`5z2}|MiK_%Dkr^?!l zO{B%37`tJLUf&=1V~Hs!9CnQ{GtJmVP*0EW_XU^pR835hC%(b*gsCZg89!5&PV@MC zrs9~};_}0U+rKi6?>j6g%tY*y-@K%&=zgccXEkY6l*^AQk*gDWPZnNr-yj(Z?+>^e ziu9WAOqnUL6dYsg!BhYPKZ6rEsFw!6U+ZmHM%sT>^)A^sQHiAkS3-eu!no4!0*cM0 z$aX4rrDp}3NuObg=nkKL`YLq$BfX_YQ5UiYqW0J4^f0GgG|E-%QLh$hjuHtJT4Z^& zTC`~ah)&U}l-VWj9f~-i9c>O*{EXqq@M%~f=tcdgot#0{X*lx>NgtPl)Th zk>eLtE$gAcu7B=IS1>#I9-HL-0lRrS-ScSR1MUQM!Mxx#(W)WIutyP{9E~-?weW@l*+^7KqIWScEgAeKx=|BnfMoe-{fA_I*A zIpKSCtZ3Mvp{fnu#zl%q~-dFUn~0feq^EiDT5C<{e-^sooI9QMHx z2fP^DiYM>vf*SL9S-5E)V~Q7t^@z=2V~}*PSWvx7y!wsog=NZ;U}!Mq)y(RW#|cyX zS9`1_4SoX)peQfxJI@?TrN~dJ+t33fZ0VqF`EQLhknYYe-ciUTq-%FUeEJAfc`ACd> zsvDsly7#Tyi*(&WP2d2RF?b29qE>jQdc$`I9^^L!)F_ZCWH(P8g2%wr2}4Gmbj`J$ zlDwA_q|JV6eu>xRDC^CW4NTD!qx9s+Y7qW<|9c%rHbgDZ?>{WUf`V& zfn#8j?Ys}NX6CJ)-p_e}`9){~-*x&cp>ClXvg(`Fe=I6v0v8_dIG(eNVW^IliIl3=gk zRROZ2fS~}Um&@W}dXTA?Xa%XHPu{~|`A9X5M^gq0%b;LG55Yo^n1{jy%L%oe`cl2K z!D7T{KTwgy6Il!g-WO*YS&TG_Ra0aIiuYY`2g3hK;DmZeZxXJ8-Z1>%l%eCvWLbuj zC*!B;$UIIeJU==$%Ycc54`!|;na-rf=xZ*e*kXzlP_cu+Ej#F56LyZx5=tO0b!70ht+Pr)3n6F`55&g1d1+ps z5@dF=&kR{M2~U2k6CBxPUd&IzL_b!CrDWgnz4iYqH6Z5bKYZgNS?$1^7vPi|7h~N< zv0y?rqguo^=r+!R=y?yx;$H==5jC@T{v8*L#zQ-)q(z`nd`xc@H4F7|lYyuv(W$WO zbL<0HBeYEpy+)6)%NuU!4Jq>bwQ1qBnfm0Dq746}z#cO8iY5fqf%x}j19HrHY*ilQ z=UGZUW0U`MPE%l$PqgeI+@`=`4ah%T{ILZa7YBAj%xqk+rXIb5q_TxpRyzK_`ULdsqpWuV|};mAg_Vx2ePd!=<#dy+!?SLXqp!D>qv#{9(3pw zLy5}$*BXP7p7OHzW66N{9RTcB&$7cN3Hd*rxS&rMvutT_>Z7D2d|9Cfe}}p!DhN{oi!9y>0a%&4M=scB(gVV zvFl+P+i|d$6iMDN6@z@X1F{^5#^Rl5Y>p~5$`U#yEDb~gu*bAF=8XG27kuV0Z@V~I z-Wbpuvp4b%taT$m<9>pq@*ASl;7&EP6(XfOGQ_n@H-qpjS?(J`?0+S0~GKGkl@Am(c-#Jc)d6#m6 zn2O_WDbD=%n_rq~SjMyq!!MCd&ZOEXbWudH&}Nf^^utK<(5^`3V_grnK{Wd6+8=I( zUkJO(D+sKb4U&<_)R{+P(W&7Ol`cU9HA%P%S}bZnIJFkal~Zj+1Yt63HZayEPqfY3 z7X&Rk-hAf=rtE+YTdxGM<8cwV&(!Zw%~MO?^u(aBDsHZpz7I^t9jqbkhw5e8pu4WB zW*03KyY;!(kzPARi;a;rDdZYI!Y=PPA$Pj!`(OWvHJB5(hlA(HdIt{C?lJ;M9>r!; zWGfZhDn;#Pq^}vXar;A-M%?s-z6tjgas&;)qKgzT%OaLVv@3QIPjly(L3MJ6m* z$+z!TX;MfTb3g2Y5&c6sA!XWkPK#Zh_h3HOk>Fa0g#)QCVOLMO{FC63xoKV~=Zw1? ztWP>gR(q(BII=&wfJs=8<8NPh7$zf~BIGyk{Z92?zxdBz{oy}8lhqx=ftgo3!ebI3 znO=SJPk%KvCr{T?J8-4F855JrU(YNSrt{k9lcJ4`*0*QgN^!TKMX+vGQ`9LJjbh!b zW)P~{4ua80G72DLj?x8lHo^HJ(hPUy*m+5^!o=7WLFP?N;`is=H>FmpGP3WVaT2LT zNEV{gPUr3PyBDAVyPZj|e%)T%U8ny<3n+AC%R2oC#K8!C@*%h^U*lIRY=F z%ZJ_USu$qDw+}yS`%QrPZ~wX4@-%+stz)LOjW0;x#0(!oHa!saPWwE=3pRG;(u+aW z47u=o==OPAg0H`}Q+AG^=quNBSAY*RbJ1so0>=r>~yJAlp)`7l}q z7s6CM@^bIS05o*?c9^GORXIvfwTG+_8cYnYV*B!#va6~w;DUQ^WV4GsT^g=XF>!6| zhM1afE(-tq5)5jU>Qg~kbQZLKZ343VQ75z^SL9iI)yf6k;h+kGWwTX2J0eG&(8Vd0 zrSjKBl`*QL%8dct;i*I&HsDnO(jB#+JHLuO;-%k>Lz5UbpsR3KwaGt|?t@~l7AW#U z{nw8a8D1!IHtN(xwkm5PE2D5!Ehu1R0uP54dfZ2yup>=Pu0>XeR(Rxs9J$4Cv=Un; zP_yK%WpE4-oIhe(V(qYj7Bf`|hs9d?A&8#9j=4aX6;UW`2g@-9(fFsnVT&gDsk`id zf|EZr{epYL*G#*P9TpO1I*bR&6**FlKP0*I&G5s~C0>h}WAydca(SshCwtPThn-q{ zScDAyuiE}_821UBkU?b+692!i8kT&I>iIhZOBY=f_saCbscPml+5Xk8MW_dk$Aj*V z#8>{|!#GOK{Lk2^_Lv)sE{RXY*coFSKxMLEj6;AO4*zvo#LkxuvDt4H{mXC2W(SVV z?ls!G7gH>dRp&y%{p@9)sZczdAipQs;aTKT8I>CWPDdVH2I+s0%XNwi&0vaLawB$2 z`{*p-039OWTc|~eUh8t<4!jeBIRCiJ&^n4;Ly=S}b{#}_PS32ALn;Uwjr-(hC9BECkb6G2 zUcLu>)dLVLvJ+q?o^CEMn)nI!IXiSI>5EYYV0?dfQ5LCmU^~)kG}Q+vww{76lh{P( zx>>vB`E<8OvUI6)jccOwHJ1!&K7GQaC2XlPGTirj=Q9uw6kHPzdEquBpYHK{9H{Yc z59=oDMWwvdFI*DmF{%7Y-Rmi@G_Xop!XDXjpB`ZLDwll{jAQqN9#*9Bv2*NRlvZ{O z)L*QJ5EopG859d%tp*Dx^FsZOjZ9<=D0=_M;9;`Ffl&m4g5wOv8j4M&$SP3g7M%6k z5&6FgD14=d8;c%TnwuJzvp8YJyHKF?k#6MW#Ke(9R@xZo_* zG+zv?@x`o4W>7zk=g^V3;I0+eZLTbx6LY^#tG|3R58yN>j+3R5MbuPmMNldz4EBbrnPllfzLtg%2(L(-F7J{T3jYrl2Aq8S{^g%N@1zxdKy`oL zZAmCbwwCc1yPON@rxTr5lU~^l*M9j@&{L{l&Mn*$+!x*$GG!BQsT~;ed(8cAZGE28 zT0$wm?Do@WgVk7{C(9yL&ZOPw=hjTI2Pjfc#TGDGpg?f~YIraU)k{~>HDRD5T@Ser ztO@%FQd!6mQSN=gRn-nkcsa@(@9OziXCouxL0J}tupTWabFFd3mb~7`o#KHg%$wxV z@j_i~YME;+%_l!>jAj9-=F}fsFyWj#K$q z5uowU5Zo8nu&bi2uymGup%aX;aPB9#)4r$78_1$R-JbcO!RDO%LE>C;&w;~htBo2? z6Q;9^r;`yXb`ywq4Uk$#!Tt-#hQ|NOC?zAEt{T=2Ik0 zg!k3XSG*5K++~+K!?OG80bX2ap*Y)PkB2%e4VpzunTyaA+6-zy*QEP`HBbVlAH9lI zdmbZdQK8>O?=HU~NgBx!*0{!n?w@sAP?_ikzWh zQ<)ZeXMmPYp8@%oVYf{S4l6E%-Hk%eDUUv?R5gVZ(^!RzCy~byYth#F1NjqaaZz3I z)>j2uBwZdCy2GQxuZ31!_FfWH8(b+umki}N z?)g^w=?i{02J|Wq(fF67MYYhwZfXYWR-oad?j<0@hYv*Bdo7IxNNfF*ETM+6&~i*{ z{wIC**;dEA<#?`YdB60^u_e!YPJ5W%0Te`WfYe5DC4EM+l{^xU05x`TR4$l;Gt7FC zdch@mRs^0O;3Y7bbU|Pqy>@=VYnxw#4aDpOCdI$twH2>u6wB>L6J;%qpN+`Ljo^f1 zithINfuF$~J!X$36O99}iu;Z1WC_IR=$Z6K082 z=Lx>)qjA9Umugs5x8VGoN?tR$FFPSl@qggcE32NL9?{{GAVs5!<_vGJ(Ea5)J=OvS zZ6DXqv}8Odz)Y(xd*{$s41j6=m-C;HgANRsZX>|7Q!FIEPf)S?65FG)^GptXD_n&t z%&1R>T-d5g@52i<%6=N-g(~dh&ZBn)CxoihqJDZAzt+7m1of$qPZ2Mc_-~RA2CaH! z)T!AoL0aZ{$_vz^!qzQN;Xk*a`WJ80f1E~&g7FpCk{+*z0BudvLJ^wwWXS57>f+SsQ+w#g9r`{OiZ2Wq8k3c*jh! zU4|qDc%b^^9WjUKY;lX{4dsZ_r_qIxhoe7I)Vg1Fn_>fEvHh@%Q(_(s=Iw|7wMDWJ z!x~c2R~}FD9C!_BGFn4+Q|wNPlvA;&P?@c)^3gXDsWK$Ryf&9Y(K6vSUu^5iq|dqo zp_6By^fXlc;?=uux1`5?6MdlKSL=%>K&z6!Y_2wP6^k)}M=Nk^HUnY#F@ha#*5l8nm>UmE&L}g!Vef49FHIMw z7X(FSmMoM7>4W_ZMg}j5Rap@jCMO+6q9M29s0;2_=#H2nvM;#S{S>>|Yt@`w(u-_ zSCCVjzYNkBmqEyEcQEo_;`_02Hd(yvVL@zxAoHu%ho!m~h4iO} zij5D_`eH<`DQcfGFQ|vjQDT&)U$Db3&h=gRzgDSETYPql%K#7a+c-vF=a-|@(s7og zA|^j&)&OYopZe@O*9kvdW6H|xur0A>c$PCoIbzHwwR#;Cs|tmW!>#uip8ds5Fv5~K zp5QjSpE7SGCU9-1@5pA;Obv&_BEpO*=#;FV(9Ys32~5)u@Ki7dr}ikvjisNolK)+JA#E1Os0jsK3wl+a8at? z{(2i(F_DYnz^e#!!;UNa-9)h)D43bpHDM@=882KM^O3JATlqkexL~b6rnqm;Eo0V% zsqp0>$)mAU+$OGt1vAPSF%~@27DLP<=Dnf722+t#E;}0s#*G4<%MX;k$dBDbW-&T<~;=$b_R0eXRmb#I~3d3T5$_pq0q_t-3RM$i+ZGxi9( zy#D47g00WHdUu!n`kYUpfbxQPlj4&(AI~`y^-k)1=yqk^NdQf6y&lzNFRjm3k^X7| z^BCy%)1mTaY}^MaZ?AZ(?Q6fs>t~}*LZy`LEsGThc!C8x2>)7+?28gDWGo9=e?8l0Oghl~qQCEi6^ zA==+39f%qtxbEQAxJO<$PtC7%SD_>;boi3o2#oo6(|8}22PQ0l5bQ&`gOs>$is%YM zT}ae?9Ok78b5H?Z4bfc`#+|`?R>SaZ%IiC({7_uH76%T9?)3@?h879QjFjGdZgKB)l*% z83^XKD%bkAGN^~ z>7%g^1qF19CD@{{;%7zl$lOHd%wH~bwlUbO#@J_wvl>=L4|w~FN#XzlY6gEUT}rlb zLyZG3LO>}sE*Muqv85C##=6^D8roUpMJ~k*=KtzK_K1NMO0k`Himi;wkYwnfaDdmt zE=W;F75`9K~}Q7=niu?sSN2A&^~)$q4@uJ%lyVPnAHEZGJ$M)Pfx zH{bl@KaX1Dxj1m8iW$%4ptwhYaw6%W`<18WSIN4eDiZvdc?Zd1@gS*nM~>xcpj<9@ zEw^vZZ~Xjmg~#}N?Rgyf;{zKvYZYXBC zUA!+a-y_NMCciyq=lAb|3D+okW!ofm{JL3o_`w-}D_pOYFibm))+$wIDrpuQHmbs6}Zb|cI_e(P%KzCMN7+vnY znj}Z}#&pwoXI4a+`;d4y?>O+BKtd?=us|zl;oqBkZed4EDd|yU@vG-|#9%gZy$G7u zz_-EcS)O?PolMbXftq>emb8J%Rj%Sp0L~RvW44p?D@PdC2mIF-(_sJ8b^8wNpqNPj zbV@F`1J78Gyj4)?hy8;pY~Hx(iR8HuyoE+=5LCoGz~8-HF~52ag#Dt;)nIY{q&ocv zlH|ap`?*FoWCO)Qqs(e5wg`$&8bhv0vm$f~FCF)hu*|LUU!waIHM5ex*1ypsL5la~E8D_qT=VGC zP&8wcwppCO$9d_QeCGOO`TKRD1~^^&*3Al1@SGJo9Wnwy;@fv^P*P_ngji#lboZJ!PiuZV_uouY z(hgg$U?$02B!qr*eU7ExWbA|%o#n9xGUm}T!D2S8>+K@BDVsN!#G?ay$7V>$5n#o?6uoEHb=&EC$dp5I4b2*X?LSp#k*w5&nT=iWQ#PrqAb zO@hFIx8Y_I1WD|+m?g}W@I&-6=MO1qtCY1zJbdF0Ng{>3<8P*ZT~+?(PH|>{9q{pF{;WaDlc(7JXz8jQ+e~RP z9oCyMLy4(R1?ft&S9qL(PIuKoNC{n`p;~T=drvu}3~fJ^pZi2^9!g95g5h5n_DA2i zm%8S)mD|y{? z9*u*Mqd}u+4Cs`gkT5#qsOqZ1zq~&PW5%Vj6R#Fi6BsRPOwA5c;t_V z&v_c)_Wgz}XULj~;O4-KQn?Y_wo_~tMYce_o&QQ+g(N$21M}plJ*L+EQ*t%t6cD`V zuN4OBUk{0Lfua*RKMDdTnAWGiP^->N5+zoBrX@Fk(;T>V+l+o%GT0NK-@$q(s8;&$!w=VrPD9kbGU|#f19q%%;=`%) zH#mW4^10$?Vfm@+zui-M!?do_VMA(WDkazY>+7k9-LA+XLn!E@-`(l1kDVM>RQjQl ziEU$b{6b;UL@|h`z7$)W;5gw^Hs#!`#+zpoOH4^t92OpC2vyJmWYW+(h%3tSnC#iD z0jr=g1@<8YVqHIB9SDo05mcE(fHcl64Ne zRROlzaRH57iUm&eEGjmG8CaMtss?uZt+ToXsw{TToD7iX$dDWn-h62%Q~FNU7auM% zTttFUYy#7l6M(##M$sureaW7fnFLVA#W*d2;^f9~0?PD`)}?u-&d1Y*w9lW5p@ld% zx*0_Zh{NlJ?{4zfgpJ5FP-8uH&V_yoEr!bEW_;3jnTLwnd0T`j{WX`&uLA?d3|V%e z|Dy%SI~*_E8~|?0yYJ?SN}-SazT09Uh}j%Ky2L>~ba_8yWZ}ifyDw9Tj^QBv6XzavG^#Va=}=--J#HOi<;`E|)`>x!c}= zjRB+K8U+q43@nGl)u>aK80F&9LdEoip8t4nronP(7lvOVn;h73R2x~2 zB8ttYNDdW?4L^DbRjr^G)Cw!3R!6LgzUo#2shQ1+6T(ye*FrmD^!$-IbhY(R_bZ(nRDDb7Gj zGTxkLQtW1mY@}jQeGAzFpbZbB-B)?F@Ow=R84X000&L0yzXD5jR0YfsX<$x>SFycz z{Qf`t&z{^we0ICt4meGO<9Z@T-@4fP_q7)~@17lb(^Hq=(YY_3lHD&aXGWc{FyI*3 z5qU!SNQ~ucKrATCfKs-~s6>7ao%>p~AG+g}QDvUFbT++MsNW6Y-U#m+b=n@W!lT9) zk1bO7wHTO_`vHu1x5ph@{Hf0#*f{j(m|%l#nVtE!G?L4$-Ozz|(Fcu8WEI5%_i-5& ztCb%jncGo!2@e&>S-Ww@>%IRz{3_* zjL(faWq9={aJ)@zgqb>;|uE0Np_XkhM;T>o<+<#e9!4G8?*i)s(+h#BBjFvs}Y#k)jT=M9mLnBfIm zygvz!=P!A48=b;ixnM64+P1hHqR)lo`7Vwa48Q-{S(l+GyC7&%kQs*_JTy1$)|_`G zKT(_NC37(<4(z6wQCBWdVlt@OpwkZNXDoKrSGtvZUzig6F>b;v$Hlmtrr_|GX)$|W zF_;o3k#aZ5d0|Y+0V7kgi(-M;t`yu?-=>gd&a26FS&}nS6twvs6!&}g$v!2O?w7>_ zvc1YS35v-LyKV7a>)%hGmJE?@K}}dcKVEoIeBCF-zeNfvm829{H9k@-32gyUyCG8U zy-k95+S;gCjvEsrV*1v`SQ`AA?5LjsH;>t4$wcG80XU$`9~ZhWp;&0I$j5GqD%l;E z!dLpd>OhsUn?~BM3vX|uRh#7f0ga-~f*Rj=;jkMdm-o1J0F`J-B=%kPL87oS3OBr} zVUMl6J^(|7;t*LYy%%K{W>F4)s_C4MdvD|E$H#FB{Y*Rbc=LQyNucMd2*Hf-&zSrI z5_w_2V6W_?54f(Cu%+D?(jvJ2npT2Tw7S#Xf>vpZAa6Fx=i-SxkU_O?6{I!h!w!)7 z?2+cN^AnE)hTHn+5vRW3(_we% z3-qn<{en|INg>I?5vLrU{^(xRs*sKtwJ1(_fj&0p(wytBHHIYcZ-vK`@z)EO3hzU7 zn-siYo|>O7zY46|@nj`?FAC3{Aeo?l*BJ7t?iHWX5u=6rykTBF{~j!RWxU&wIQmn9 zM|i!BYl_Mm$c-~t794(tCW{Vyr+xcR0t2w#JhT1)S!BLc)?SO*3E>J3vk$WIg+Kqp z7rLKaBIm%76*E=%iTqMoDt|<#LK&b+_fAQcNL5UC_$15sEBnbGr)=JJ)5T%(Z9Ki7 z|2XrDQRB>t@0#w89JT=4%zkGh!NiVk|C=sf=8Ah?6()C3)@jrUx5apRBlx~LssWV> zDpd~CwPAhIVJOmqR#6RgX;B`%6AX!JJMS8u9RNdCzV@;ICRaE}oV(KJLfHP+!?^BNnU>pAGqZ09bI8C>ZI^+}7CB`dZG*SWL04RlgcemuAZ;7nU({^Q% zpG}v|!1CD>ZF9g0EYtU|zk7fh+e&*pK0!vDxIMsOY2yjhRE%>>)D*jdB1u#%!lC8O zi;LFsHokL-KWH?G4_rk>ti3CN$1k4(_vZ(`ls2-o1V=sS_v{l+Bo?;(Of)x z+CN)1j1DGxmMzb5`_cd8ZtpigG=SvX4-)5+d)&4&4x9s8J$*c+5~j0@r;`zsq$>y< z3hx54zapWQv=*2C)hM!NHAyx|a)O3=m!n&xSixUQ?-N{sYB~HDsdAPD zRnLD&hTL-fJ7O@bdpjzf*Aa6T_QCoq^>kV^$~oY*J&N`)jpE$GYB$uZ>R@w}kHz?| z9yU~XP0i%FB9G~aY#krpr}8V9i_o@|7OnrxrVxC^(QSTe2Far8lt;pfy6U9p1bl`)bE;4KTbu zKjjs2eIlkE2i`a>GXmQKiXEiL02Nyx&jF3YIx-0AylAerc&jk~oUO!phl&@C;<#d1 zo+(lt2|Nqcm{mTRGj#4VmQPjrWJM%HBLlGVL{-lR&l@FvH3~J8$j8PuAOQn*(-rdN zvL2|n(z7IY0()|^3)WJd1&n4ZwUVp|;DzrNfMgo5)UELN^Cjqe!{S7UM&m75V5LzE zl6a45k=A#mXREXfAFQ-r}3|tJs{sd7Xv@px+G}bJX~w_WZk&- zqStO~H_T=>%a*GwJgfF){x40-CmeQnW~LybW#O@fZS*>~^{>S9#~(x1ffUbDlO3Mb z)H+`Et1Ed2c%UP}8<6*ihdpiumCnx+T@q)CvLZTSu;sQ$+ylzd1M&ghPWxz0*a8GL zo`2or%$t=7HuTinIoUsd-l<(M61Zl@1y}u&w!#Cwq$_e&zQ=j!SwhB_x_MLJeUF<< zPS6;4pS=$`L1X$qe_Oiq?<=o>d$xFCMNle4D&mDBA#pQRY7xr+?GfWbgxI}m zSY{1~%;3M(^LGYr4sLquqPSPq2eo;nym}g+hOW&TFAP2dp?d_5u7a~QuJ{>;=zCGv zSAb`6JRS`|cXk=CM@Irs6L#ON*{=dL1D5$`2->Ci?;-EXXD(_Skz>0p)P#6>a*~q~ zqJI4o)ge>DQ-^hF%#fR2<(-0U$!XH>-L7cEJ}ZrKDfp;8a%|a$O)@WoNfNfhR=x}B zDW=T0aL31!Ike{yCwxq+OqcvlITlik{p%W1;lOqXBo@bo6dNhFjw1W1*tP!J$W^Sa zG#opqF~2bCbe`0Po%X!vtKV_smRgk$ZMDPlNBpL!u{6Y}Qzo4m+U?QGgLKwSPiWRY z8S+4`-!?)2cIZ{uJe>!M?LMYUieT%FygCm@;7(nC)yl0AcMw@>XRL!5ClF1a_P5am zVFnN-r}vkU9TUZR95}Fe%m`RD6uXxqyQ$b^FWrk=Ha8{ZXZ{U4=YU-?%1SML7$ z=vR{8I{(%6zZh8b@JIh~DyAdmkf@)2CqL$}0&gk|#7di8;%-4bJq!`QZr4k5*715| zj|3x5mjrjk7v*?dGA_zia*VZ1RZ%yD!U)G766<_5no(!4@Es zk#r!cD7YjN7-S#%^&|ntppt{NOI)R z)$<{{0O}+9)J41E^h~HIOqXwU8<3-E!>mZVVt}WkcEmTkf_DW&la1wovee8PdPSCY>{OftwA?ThjZXuRrf3Mt}Mc%ZKwosz`;lA&{FVf%5V%X{)hICII7G0(0ZD(S z*mZ$PvJe%PPp6GWazjPKKOWmSYUly}Vb}Mjkw;FB>(qiQ(YR2}N{UUU$WkQ5!&M;Z z%auR6_5G>GfnnO71qU0=F6YN^gTp&F@_y=T@I92PlIzG82X=kIc^C(W9TW?k@p)A2 zY|EAq01y#jrr=#2kPmm#vrf0uam8@qX1MfH*qS^1)mMHm;v+67u9xTh(HFLS>E^rXU;gL| zx4*gf3w`fyc>C6uPQ?H%uxvp7tCyA<`l|J!?=#Sg3|d!KYDY2LwF@DGN?K1t?AyYnsPj zcOHr{;7uFxat7J=oTU}c8e#Al#U7?e6BXOZeoWt%mWz7mCr7|@7LvWNT!fi|9=Z|) z+0HQibb3UKU^OX@>WJCof0EuRI!V`)+{l8+{XvVB9Wj_KzoyHT=kf+%%(|JoJdx%4 z)FRBCC%<_~tY%J=KdUrvo?it0!Eik2{z!b;8?)+U)R=Ws8H^a>(1Oz zR!PC?J*2au)1l=% z2ae^MQDD;2NQ9-80CBOngE#8b4w3pkdWc}rClZmTMRl;FPC(XaBfr*?8L$Kc<2(;b z9%+*?=IzDnzgWS0-YbCjDf$<-zz(K|Xz7laIAMu@KAq^CCeIIC>YS+B$d-o`}SpLWsHumF9(Vapk^#(^Vt13AA1 zrLq&CJ{~6=1djL-koMBjT8MGrCbu_IqwHDq;a@SLffXuRy2ce1IvV`!&*eJldAAa8 z&$^wPn{shp%9*G!lkd|cL#fpoV6}mR()nqid4I!@T2y zucJ#!9_f1v&;JRQGA{>yclRHbY}UHeD8oO|ka^AYQM`q&Ep$ z1hwvQLM?qb8cSYO*fovj%1%KP8z95{^jdMnJTfNO8cpA=Nc-aRzH2V#b@+5jPC~?9 zORIJV_tU*JYEtSjldS{^V!(}6Mu6*VaDxWdr%!%7H)td(!koWiSdNB zhQ7&h0oZnmg_Y|B6?+tlAeu=X{M#wdr*FNS9DRnA^G089oz*9;i9X2O@m|YI@n6dC zVfyG#qx%V7JvILf-AqoftAbVsrSJ-0(I~d_w6gs#5AYJ5Z!XMUh}S#(hCR;8+vu_J ztAmb5HZV(_J7RXjXZ6!VQI$T${9Xaxb%K3obx@-7P*jcLrca-A$Cp3(;<-g_UtH?U zb;HFCgcGe!7M|mX$8X&D!CccO6Nl}DG}AVMh2n{PLx2AEmv78167LIMCajWy)zQwp zG5542m)_wSx1f6dX-N)!IZ`zM#SK$of+jiX9H)@`a88zI+TRx4b~9C8a#)C%(NgM^ zYz#RHOA>-aSLBZ+1Z@!2$723u>4_=s9N_>A?z3yf9h`tMeWPFAoKVA(bnRO=D@cJe zIb>wZt0}gMB9&BZys$iaPv{4q$k)H|d;M{H4E6@5%kk75@5A(Ax)s!b?|2u2oM(#w z*)%H{^@22!ni-0^FGCH96#x6O_82IOt6X$nroS000lMV(Kx=YO=zW>(YK`D4DyKjH)p$4$&Wem&D;#+X_8_X~d^$rA-$92eS9!aLrd z-AJ*pTCSmD(YyT96?sQYfw0?sm-wFB05B?^2yP2TE<@l?1eLB7|MHn!JxeQK#R-uz6;i>b_&@MD zNN)J#LN8-^bovyefX3#C8*HWyH)TH`%`)wVHq%C|1;%Zim@h^wJ0%#LIx$lN5lsBw z6#fErU#Z34=hy^H(X08N;Y6`3hwZwNiq~B?)!uqRGK6N7y0ET#kKbxgVA}<9_o{ed zcc3aqs0LcqY~}UWMx9pk_xO#-Dx(^h3W=)5^{)HC0^AyH@!sx%w;?<0W4@XX$;Ioh zrO&Y2kHTWyf3AaV`ya!=wJ-n1)Hik54X+ue`3`C1R|DqnB+~Ij2!hZ>A&(68GO^Nf^}U)?ZBzyVk3`2L$RAEvH=r-^>n)@ zlEKxw7YYx{YN5qKr?-|4rO^X&JW(&(BkqsZN}ydGGB{JS9&!PQ!5~a>%DLUf2_Vxq zZ0X-+nl*kw78^79WIY*bk?_%i9j|urMng{eTyXCeAU6hz`gIGAkb7=f5t(!ci*(z_ zg7IWzp|I8)%k8x?yoSqGnSVc|pt1U?TaY*-amG1v4wA-*OD)~bRKc@LBK1C7femUe zCxN-oU5jjm%Y}Edzi7J6ao82e%*G{)|7bzq%mGUnU&5RXK1+=D0~J~9z&=Z+ks8`;Yw1$@)?;> zoqZdK!thB=QTvo>QEP!+7xe+LKzo_+CQzf=%3Zb)X=Y5!1gC9;yX}<;SD*Loaiy$V zP)&NJ_4IlX=7f7(8vN>m@|er=T#(;_z{mfQ`|Hud&8&=QT7G2;4K8NIft?mJahCd! zf`A8-em5{Iu0UguF7FrY0Ct1r^K@0T$Yjtdxi&>RXx)r)vl=!s2^Ksi^W&T!I2pW< z;s?L^Gl_FxFGOSHg``sKDvB(}OmyDM2~sVTJNJgGp)FJcw%=CP-h!|2$q-8>`N`<1 zJK|(5rse`^ldkD5kEMQr|di zkhRhX2RRfAg$9fEg2@?h1R@BcaxqkNiq3RyopyTPPIq^< z(~ETfYj@gdchhNS+AX4Z2SGp-G=OrEizpXSl>4A~7Z4T2D-zLhm_a~fP~rc+NpMIc znga=qPPf0EoSdBV1<&{7{oe1rJP#F?%Eky=1vNlwq|+ad_ODXb(W&Aa^O{3@c)39x zG@kE(bu!JPkiqjCc`@@_*ix!nzS&yLfRCnR)HDBR1nV4dTMpy?HE4!hZ(R~&eb1AP z+^mZYuR6sBsLG^RScX!mFbs+4iII-XJK$BOtfEy%NS7Gn9Id+kbPUC{(%8$RiuY_& zY=u7bs6Z7`nb-JK(3Q+~FUuQ}Aux;@Ryvz9G>pYx|Jc;~E8{|WF1ueFUWQDR)oWy{ zy^DcQtvPf(=s>rUE`BjM+#4LWPP^fr4$2JH!@!7>XFf1SyxnRiSKi+8S7ROvn_X#4 zFjLggNGpKM^{Xi4w7>3pj8P%6fLeivt#G1e?lsXl(JF^L3h9jT|MRGEK!EwPKHR6U z+GIZ(nA0jV@qbF+c{GN%}yS9-Y#~)(6Pw@L~P3RZjSzUP$}dPUFfW69qoqa+EK*@X}763Yq1b6n#X6 zE&fP-*eyr_9>fbC^`5Gd`6tDz<;il0cIODKKMkYosCo2rvf2Tsd9d9Gb@l5q2V;4J zXR0h9&A=Irqu3aVL_ymGRA_5uG_vRij=6LJt);cm34o0QY7Jm#wka{V%ldCjL%9E3 zpud}b^YZSg31>(Wx6O;~;vIB%hBr*)P%QM_q@!?Dg&>OUmY2=d+rVoeAlFBt*kdHk z1Bt4UJ82bJ0qQzBnH~_i%yGh#DfB9!HEmY(Nic~sV6jB*WAkq|jD|Zb!;T*t&_X9f z-)Wm+Tq(k317pKrF;Q2d0e?0>s28-wAiaOv1@pe+D3&{X+RVem@H@umxb?0&!?>x& zX7?^8x_#1>Il?pZ^MkZS9;G0VT;hhtASxtzf#u3I%$*MNCK`V2<_>6~>rAg9BNY&R z_R5d$80%)(EG$eYVdzu}FnNeI?}grNq>=vI8wjvU-1?nk_)v!E-79Za4$K9?fwj&D zl|{k#A)s0ceKJsKh-HeqWSR6The5k6TCL0#ZUeDpSZN$8M2R!X9dO7*;2(ja0JS2E z#=6C#p!TuRU+HJsYy=HkGR!y}GmbbRXu_S&Vh!#wq+q1hZyOG%pEB@Ct0@*ZJP%M| zy>q&zw}UWgA4!rN5h9;B@)zh-jgy>F>h$9dDKfea+TE09Gp&N1K_zZ6pc9Bazm+7? zH-laYQ0pYwUV{PAa9-Q5Q{l6XO=Ok|3tUf-+x!fF-Q^d+_bq{X=!~zat=V#9a;tGw ziOt#-6Xh;hbgnR$E@d&?3XW-)G)H)Yx78KNTenWr3Sgt98IZ+>V0pZ1=u$j!W1ece z&jTQ6yzr7Lh3@CAnTP+~8DtGd$3hDcSDW(G>p2;n3BUc*6p2!AbdHAoUmPiXZnntv z2Ii=SVk;?9MujCYcjQeVRf}Frq3d$TTcAtZ#$0k(Qs`9X-Nw}Tw$iIyt_LK@7k7#4 zgCT(1#(;dW&MT^B@SH}1cV71T+^t(y?XF5?G?EUW?`d^b;la*%`65+1!DfsOIvM_3 z!E0?YITjLR*R@iyu4=$j3%Tqd?oZLSgCI?hJy(9==G zHi4{lop00lmp+<1`KG7II-i)fH(BR@`7cxSwxwnExj&G28?JgSG_V%w6blhqH5GP! z3Mv`v*uu3*gZ;Brs8UcmzbH6s=9hC&#$1Jo1Wl2yPPl1YIbgFsi>X=wXq*9GB}19*wOzcM2Zf5QG&r%4 zG)@H-Ky3-*SPVbcdEo>Mu7i)^dQQNY=<4;Q$1nBvQX7CIp7P!`u_J9Uy23-Z7#Gu_L zP_>;S?04)JM+FW*qy?KkvDRq7uG7APUdgWl(k!(i((ST%&@NoqA!?@qdDlohufo6C z1;6TmPXXM}u7Ax8L_(8JdsIx=ri^BAc(?(hIq|$0zJ|Ul?halVmD$C1eia1+kGzI(5+AM>1~08Dr|34zPO^}R>3 z=w19vB1ncu2(JbfxaJV+?3o^)py6lC2GQ`F#^JazY5IS>rnf`)UVr14q{fEr&{YFF z)JU-ql{rp@wbI=I6{2b;l`fywC#m$nn&}RoJu+zAs(`%co1`t0R3|(h`U|3&ts$9#97QJ0hRN8C3MX)Jvo))7|Kp%^Xyg5z?_MSw zo|}y~ROt_+G~Y(CDHPd4g*7RL;(oXkmOG#G$xx#8S+ZnFx-yGy57-$L!{5eaxVAta z;gWU-T$U=yDy9h13~r2E@q%QtuCa+FMweDDqKm+v zNpadW<-9z{X}@;|okxvJfu!LKm^;yicUk`(C)+VGSM`^lKkE!REfQ@)I>qViREa;05Xm?uC<&E?=b-qv%X>FL8Yd;>XG=>L{ z=FidxrtQbfqa~ZZUft)dARBoXEJzF=H^$pM%XtvS#ceOVoSd4a951g#H^&?hV`)%4 zalDP~9-Xv)8{7LM&toLQhGQe39y)Aex`ATXQ)C?#77Hp`kux^?!o-#-;n%<2xq!Lu zdf}yiE)+NN>bdpb zp$a`ZLXW*yM$))NzHGRT04hv|8JGf!1<|%#Dhx%d)e4l}!q91gJex`7pA@eHc0H&p z!PblWbdFPt1l48-?bh+~MVKv0kVgn1v8+%PviG&JU06AXJoPzFC2n|Z4fDhiVTNwW zSMMwixVLZmyfI6UnMK#K*O44~j_YuzpjyWCOk z5vy%?JG3i-b)ZqyL8F9QQSe&lejtg~)+Tg?B!ZNAq4#kuMPobMzip^su2Hl~h(Z3- zp0EqT&CsA7R%K_=75?p(Vra7vc)o&>g)G>o`Tw?9tOt+m@84=7>z}d0+MNavNuk&+ z6xmFLVH^bbV6TWf=jOTRx$hS47UISwn(2@tThgFih5zR1?aBrzNeLHjvJt%f?|PEdLI>rm5k7DQJ^hp9JIqe z=zixKA4CP(4UKcpOdhwiR%4rtY+F}ah$)ZnqNUD3(=QNkmHF#yDCYQ zB+_^1lp-AqzVzN_uKJaz>6ax-$YySit_{;DfnPgpbCgZ7z$(0*3d>NY(5HA;y?W=Ymf!ZP z7vzeuts`q@Ie!Oz!SAAPxx;4P#jS$R1S#Gc!8s;_E(~1frlUCRb6m<4%XWAlRvcrk zXo|*$70Y8}XvEBSgFTJ4Ft`{O8^($Wm5l2_JA5j5iB7$u$iPo{t%53X1`W#fypywY z<{X6mP!+tp27CugfgiF%6zQGBq&TgZ8^f=51G*g`=Q^NV=9eFwFUkcMxyn-8o}uY@ z^g1@YSjDVp|F)?gZ%`XolW@U{4ZANUYEN1unEg>*1QHBgbz6fJ9o-fQ@^e|yVmyMv z@B_19G5n@+I4<`7?@x_;k-5OahGQ#Gcs6Xwc_gTUg^Urx$Qgy+(A=((teVjsylTd# z@f05(npX2LF?99#y;6Ul@B8&C7uaHdJfvM#I)BjTEHHvWEt1Y)U^D7?8w#_@!p08| zx^J5L!SqYddS0hP+7E=lhv`jhgS3?{3f7UtbhIbIdxLAA?aE*||*tq~wFo+3_jVY@rNzr6FW_ukjXJzm~< za4PwfTinBjozPVihdZJD6nl>%J(#`0(1BPH{7HeIohDlkx?$*W7kd zyF!jQ%N^geNK~Z|$G;BQ<392Q2^a1a<~ZS%{Y2F*yXji)q3Varbmu$mBi)*S=c@xa zG_Bw2ob) zf(K(NRhnr)-ZG_4Q7~CIe65zs1@n$NeRO0Kgv^5W?;JXVxrP+F(zhRI1@V7b^lwJ@ z8yfa=veMzlvh?p6CkVOhfNXeUY$A1-1QiGvWT|Ydb;lEyAnYuiSC3`W5HF|6yS=(U|Sa5ODgD#Of9Kqwz{M`;Cr>= zivKZkNZBkY*S_B?c_3CRaHNY&`Ab&tphg3l$1RD)F*D))Kg|5|v#$Dv%AUta<21D9 zI#rB#PmcYXmU*ydO|oUaiqj~Z{I7|JR&I-YdcnO7uO}u7?rVH1MfoCR>{8)Di)6KQ zC$A$&b=7sLuqe2~ze&-gxWNlw0CfhH^Hg`mhiR;}*A^n6)NG675E+nJqo7Scmu8{o zahvey3@1o%xY#s$`Xo_5e?h91s750TVf}aeHUDZsl z@|w#7ms@j+M4P;ur1`Q&mrN>k(VsJ^%EB8h zYlv#2b3KhzH`UV<=w6lOl>m z`X{w>zjK9umsG8YbGau?ojf+;0t3A1A^ES@lgRX$oC{eG#j7OBmq08FTh>hgYV5G1=!Coi}c zq9-cYC94KXE+iq=?T3>W99F0Xeafd{wQQDm8>?1S&f5|g0c>4^c4=e-AK&FTL7#dc ztU9GmnrT>TYhk3Ai-X61`o?E03kTC~?MR8EPp{0LV_YO@qPno!quk>b6eX(_%Xl@u z2fUUqSkB)L)g2wPH%T_b6!eiSrYUqgRER~fD_?95JrvX|OO|&{*IijhpQ1awBiZ$C z(GcXyV$OtGcs_=n;@Ib9Xe7(Na+(+0jaj+>`p=Hfx};#1dnG6-fI5PERS$H)Jzt6afoq3f-h_0MpVP^v>2yyYAWsTjgjb4brBcF6wvOA5=aK>iN5* z9dra0=b~EqBeg=OX}jMWRc$fczcO&0LkH48mbi8EYMeiHRPE+nWYU$bg1%|}a?r#@ zr?*Yswy@tZ#c2RC1S&)*RMyQ?m5Oi6`=;dx`+!ESMIH{~?bv+6CAu>z2pTVG=D!~i zZX=zPeD~MC`;C65{kxVOZsfWR_X#aE$UAmZY$ruJsIbMU(7{vV(;z)Nr2|L-`yCsk zi6J+ocjSp719rgmz(s-z~<1iQ-;1T+%E>c@f_q9S_SwG6+ZQ_m(tx=4b|Vxp$*b?4l(==x+^4` z>36;g{JJs1#E>uMc0OLvJI**Vi8x~{aWZApC!Rq^pLKQALZ*hl!to(L$?>7rDIwHy zKq@<$S?XP-1xsd7Ht19J^B(exWoq7PE6QgZ&ID(C7~XERciiw%@LxNtzFzUacBl)= z_JSf!t*-=1x~RPX&Qz{XKI9UhbSqtni7OPhL!C2Kz2t_2_99Gf>+1Vj1sLc-!sL#} z?S>^LmNT|SupXB9JB-Kp@;e@Xl<3zm(H+t6NQ@1yVZfg-tbjX}Vz*Kxi3$s+&Mw+N zyIi5GEWsv4%qumxM6q%Hi8BuStRnrq2j9q|W1y6#U!LogBx!ivN?Sd{$2*!yHypt_ z?>HHhi69wDHf8%MO4@|lFcled0wO{cYtzzfktqCN(PN( z1}L^NRIHss-+m#5ZU&~ag~M6k>}%^sTiixMcg|_Tlw4Ub|+Ys!;4=OZb7|Y3!~g(h4Q9Re)gr~ z9~pPF*=%ac1f}i$S1$+l4H;-(CB~r zg)&9ll;+U9;C|jQ$2+si9W-8_3I^=jrJoA!Dr5Nj>7L-?>74@0t%Qb?Y$jd|7C(bG zhQnFx9XGuA_;}BL)=PThJ?}VTCHJ6RzhlhI7_vO@A5g)7^JfkuEH=*^2I2&XiAT;H z|469!Nd8bX`Ojn>x9U3^-uvVj_$pf{b~FC5VVz)TZU*J>GI-&_{3+caHWn@{=Wm&k zsjL@th_;Ih=sWxYyEKn1<__Ol=H#*IGzUz>?a5;USn0%>tyh0%%&uUwD~t)g1Pz@y zLnq^+>Jlde>h@Ra@s~g~&tzE4ndE>7ZYJaGr+?e#sW%yUgYT4()aS-zfaLqIfbDLI zg$A@7DlFQmi{9md1*|;Fd5k(LWxo>y2ifQ3$MuwY|p<@oXG5gX<$5Vse(nF?b z%7gbv-AMah8(vE^2EaH+u}u_dpu$#6Y7WId(Y)XS*=IiCezDTXsS(cG#nJ2mms+=d z;={rnbb<$z!OTmiJ6$(=b-50D)l0L0I=NYKNraTjYK7Xf$v+RaMaiBQ{8r0X%LnXI z*#`P9q_Z36w}>_b57^ZP9;N$&Q@qPw%yKwPmqB6JRdE-6YLSlIyKLcInI@=<4i{#* z<5&d?mr;io7PxMc^e^a}!VPAYz|Rnlp7LrocJM;e2I}&KIa|l#zu0iIk_rE%hu%Pv z73x_)^CBs6iwXn|U99eHk(>fB4Sfiq7)&ni58^ieoF--X#oU22{BG+VSN!*@kjWwH=CwLHfzm=2jIyTZ9J!*VcP0D7XpsM)7OK0X6hm4OQFb^RH zOM|INwdi?H2%%yfD?Ptn3Bfj5spD?tX}V38>TytV%BkFc16gI?N~65PI`+oA2jYHS zG{256TA0b}4lZ}d)4qa`w`hycM}PY0!~et6YQ^FtYVkqIwHNldSpjaNz?K2fbZ;yQ zEA3!^SMT_VG5L$l+7lCWF7dq7Kuqa^JUdn#6+#BGTfT#?7u}OYGo7wFPGg`^AccO# z?x3nd0>~5z#v=?j@btLJHN=DXD>|zg-Exb}OtoPn$3EaPqWK zEdRM_oRq%&=4#^{5SQvW+bymjYJ(t648=xKWDOPe(STjad}QobU6a>1f5L~>dm!}x zCs-b#E~vKfg>UGib1iS)$|S`$TvOXv6&Yb) zJ>ham@pp~$0ygUhn~3x;z8RDmh6e&w~pQ7bCF4$QRY2b%f%gQPd8OYSBwgLcP{?z>i6}C`k?)T z8|1hRyE>m5VB{jjo~OuJDy)M!GiOQ9x0*xowO8~&eARX6>a{%4enB!`X17M8s18>cMkZI#If(t|~npgAVZ|UK+b2@W}RM{}RE*rqBo?=f>!xqBf{r}oSjoui>HJVD+b5DiR3 zeXt~E&1`Uv3CX0RpcMieCe(^`3pzmUCW}sV=z`FnIWQVNqvHmt;a_UTF(*h(*ts*~ z>>u@zl4r<;AAEej^6#0@0`y9jxq5|>1tEKhM_}VVpCW-b(3ZiHR zS?+)kPpQp#7BZK?o%YknDqo0l&XZ)hZ}X@ZRS7SO(t~>D zA_?D7{wfL;X~W^K=$4)HEOuD#0Pt>+q&h{zz+cYI7vZT)x{`n{6J}-;Aoi`(HZBf1 z?HVpbUbbqcUV3Ams*S1fJvv1r+2h*epDVoP+U}6&@C=5X7rhvHWr7}o%I$mqOdi^C zg0KyjSR@;mz$l7cLy?ux&H|h@Xa-eX-uN@cI!ObhTVLYT7;9&#BuBXU4DhyfjK14OPc+?+9Lg|yr2jUjVTI*EtJ!$gH0EO1i zdD0lx`0N+IjeN~R4~pe~n7WRn*s$kQZZOTeC^nBG+4!@MhD=lijR8U=wE|ve=dIJ) zA^TY3c7wMRWVON<=tL{dO2AE-lH-K&z$i8~uwJ&z8>}}=QEbf2v_P&d zi+M9+E(RWVcpsfH?~6EI#c5TT(0p`r&NualdG}UhGTA?gG#gCJQHq7e+FD@qmuYAy z$%hU>lkb|+ zAW$KjRDyg>XtDyOcUmOaZK!o#b*$fL>j&*3q-&iuf}$V|eZXfYS+WFk%Em)f<;zou0Ro6g)RR$|(aISxvE!$~b^R)R?5g+M8>n-%$gbY}A2V z(j8nnzeTb!a1ZR>{Y$Oy!(<8J=mEl`979!r{ybzipDUP+`8(^NgTj`heZXyBQR_SI=8BuiQaBYdM?Z zUG07eS{U$><-P-UFmhk;0mYGT9+YHyba<@rEO&rYX>`&H)$D0^-6`!b2_C&3a8Lm) z5vgzh7j5>q=y62`hwIqG9$6mcz}u%ipWv}~!BV0Aum@TpG8bKA(pdZ(QoN0wDr@4$ z#8Nx*X(O<4vRG8=64@o=4*%z*F2e*LK$l>ry9zsA66Cl2>SgI9HL$=H&={CKZVY;l zX3Sg+J-T6>4!q``&!45YL5r?suOm4lMgMHqFC&Ky?9zUU-Aj>TDy&pKAWK)4fW}{% zhmL;}$;h=mD_;f`FcHGF&MKt3X{T4qmkPfOQXLH585GTAIkXC{yY|eDRNnMcE3lq4 zQ@avvv2S$g5+hCQ1>XWXXGVUI>QiNjR7bF8Y%fMRb>>23WZN*(*-`!>Mz@(ZGrx}d z#$(O9k7w6y7(6D*|8$)8y7C?kD4=L0Tc$+Kvd}aeJ>4h2H476@9>OZ0n#RQGy13KN zI)7p+vsaQA*gXYBmWpLJ=A9Guu{T|A`#~u>ba$_u)Ga&d)kyCSY!1aORT;Bs`l?Ce zf)2CbU}OobGXF+{nufzjNzA`CZh6zVYS?D?(k4uVt{@o0I+#f;3dT)!(?6o+-z=IC za}v=;Q#$6ioG>xb>2FsuzODB|d?RQ5lpM0*y-=Hh&vBMwPgA6x3fmcU$ss|WtXL{M z0E|J%>8*zFV2XFPJcF(P>W?na7sjX}`Xm?zzCEpEUWB)1+9`4nv{MG{Zu6_WmDAhTgH?4I#B$@TUoG1m7y*~l$^XT$Xn z#ReeCq}c5gN!1mwS2uxo|-Y zeXP$asK5F z*T@=K8?)WHD7a+)NpX&l;}SCVc`+LnWB)2nBh=1*NBOeeemvl%y-O-=*k8G5U=dDI z>@f=BCXkLszRXs^LwSUA28iOs@Sz?mHW;bB&-!a5w_S38XjNN7hhjDj(o^K{%nTuV zD)>K4WoLU~n7-c;MQpItT=yA@VQ*&c$*u>~L7ye&3@}rG_n_C4AU`brMED6rUeI5G z^eM>UWDCczmYEHakt~Oq=Qv?Qe=)ZOiQd*&_B}nI_Wi5ZyX2S+1L~Fmpjs#v^6F>M zPeGTrK~Uu0X^*W=*0>gqZiy;eh@HE-8eQvd zdq$ohGcCo)H(T?p*BhLJ{TRZo|)2*bOTb#v)DLY#X9FG$edz1p5mar;8p=;B7%|DE$K1z`555oP* z=BgD*j{VX+Npq;~@hh>LlNx&g+pUD?BhaMLt$Zty9P9HgD9Mso)P=kk(nUsx zO^cxKaXbz`Gb(KMr>uDxpvP01XZl8xKZ!tR>o5~TU36Z+4)3sBkR&?{$ zd#>Z9DB9$QcscHCnI`FF*R{+^;peVu#dZ3gGxpP~VV{9?WcTSkq+5=^`BQ@PBHcAO zN+Jkv=D6P@c<((D%d3)n^*K(%=pOWAPr>p}{;{tZ)d^0&d)xil_NVRc{d&doG!WSZ zT>C^hFPApZz%A0}Q0bw<7No2fmm{%=g@9`osyLzBtPw49QU8@>);Rpa#oe{x4YG+~ z4F>%&uBY1;Yq!LZZ&WRJyd^*(C6!LFtq0Vu!>Ogh&!teatD$>DYGo9QHNLyfWH9&gHsHg|)#t)zL!l<3PoSV9 z=@a#<*7B8p2cg-RfK=zORck%PuA|6mEVD*W6l;!9ryO{O4%v_A@AJ{W!kkC>=;Coa z;Is_c?k2RLqaw~}$d3_hyhfs*vARp>Xd4D!HN`^H$tEfcGzlDFPgNw^t2{fkNUp+! zHFh_ukdhM{8UfL`?0|y0z5&WQyC)+=nuF%YZGh1rR(NX~%>%z(_G6wNAa>q%5hT)v z^8t`y9tMU4irqkw^}4oJnD2I3jJ+x`G@jb=vz3b;E?l{2<)U?qcD;Aw`{QfaIAFjE zGi`cE(-;_GLp|~Q-zrj##fMF(_GZv4$Szq7ztD9#bIC!a_T21N;-&^Fe^rrBoPUlm zmZ$bi5){cYX#7Z(GzDVfo-1E$a_O3$IxUM{!>GDtr`+%=Ot@6|$G9xDM#CwGd3pRN z`sj@5=R4X@IL(Xg`hOW8oQr!?Z%8I~|K>At(1s1kbpu0khGH8ja*7JW>~kJ)c4+9k zV$4$mg!hmFrow;IBJe_Ugsp;pI{l?;rB3`-)u7ONs7`Cv-2~&PFh8BhToG6Ku7vdCR@d#$m^xaL%>bV? zt)X+AhQ7RmHq)M22JnxNZ^nUnKLJjFpZHO+2#BYsv#_X52m9yNmI zWS``6$Td-e{~o7$J`g{?2z8#md4We|Ww=p_;a!WH&H?! zH~Er77LX7fBOpbLJMpd4!iAixbj^FZPrk}bOgwoID}2fxnIxC$P0Z2neD^ZhV8ct- zZUdvSjbdTiw}3=DUF3G%btA9Cy}^C>aZ_mZTTw6X|M!P)ul>RL*LMD_@#RggW5_>+ z&i2A9Y4RqA4@Ll zKvFW?FGs9aY*FNRZRhC|-9vAnjm;rzW|l+~gg5J|H!??0h`O=p+z2XdKOA z&>1B{jBLK(tm0-ZY}kj* zG_V$%C^ms28>lcWqHB@Fd-h3GO$v>qMWS{EzUEf?#ynJ-?RUH+tmo(bvp9RggxZOU zS6l5p5&;t=Vy)_H7!A_i~; z(j2>?1s=!J(Dw*RmFPtGRLF>(Kjp$p+nuZUSu?fuB1nvVSoV3qGGVQ7wY1e)>xkl} zCNFTGAGZGB@MQuOR!Q^(mqVWFs#iDk3gPKq8nTV}HD5|9{M%tuJhUL0w@QpOqgFaM zgt_MFlz@lxX*d&T1Vy7FrKidAX6#9x7l)gxbTay9` zb}sO`95FY9{F)V_>#%#y90o-5fA*ufauhHh&DU_gXLhY;jIn!Tvo5a*hou%KxJ~{E zuTiw51X>pdWU(O;LiARaEQy_39)v-Y$QkAQA&bNL8F9XW<79Ilf3!DnTd(T+qhEE= zn~on>Z#_rixS0;yEi@(t2Bu>>#X{fFRw}HEzuC9gJ=t@wUxCv`UYji3@75foC%nUN z4!tVwqQ7|ginvvf%`86gla!yRf8Mh&_m#zUbn=2@y1uyZLCyM-c~LCEh_lurxy zyWn@xS0~4o=GZ&8UhvQhnGo7s3qVwK?<|evfKTMq3Q>jVoLMSqV}JH$&-qt|YwTZY z#t58D8uek`+PB`-Bjn7i3lqrs=f*9$XMm?$6nldr*Qv0Lyw6?tki&EC&Hn1BDYQa# z4;l(TI1}0&x`!+Woxjb#^@1WAb^e+|Ye9nz-{VaKb}QztmAA?2MX|xDVm#9vI#m6% z$1BCDC=}1sIW1nC-7m84!; zAu4jM6{GGEB%qO`1+^Jb(g;6LK_$b?t3fNEVHy+Fy-YmB79p*e3-@0XT@)Qt#!YS) zG}51YL*a{{U8ainJT5xu-ljlK#N8pgUGc8_LHlTM@H3PtbQ=ty!dg3w*x;8%JE1yq z0kdY_NLIqsl$s3_OIQk1f5ZsqIjw6G0;a~zp!A@6==|^`8Q>2Ym$eKcX_z~|A$<9P79Pu_0v)B_@K@SPHpYQt6r zglL9M@otI*-#CW~Yx~8Y%K7brrqDC5++NuJR`TM)Up0podhhXq4+?*UMU^GL_|xYv zpZb5de|b}?vnRVlZcFf<)h@kq47*^OsZq2YG82gu+_M}MH&}gK|D{#t#O2O ziJ?8deA;ra7@?|4P)B245^Eij7aRuV;UOAc++;@xk@GN*mpDTkN7}@c`lQgQzPR^V z@7W4P{}tfVbjV{vP(@<36qVleX)ug2waJeNyJdTQK>DbQKEjW1*75sZB&`Ch)Ke?2 z$@@qt(0wfP-sjf_)E*a^G!HFN4JOazT2s`LqJY=QEDm@%2#VIu~i$+(~+{lJch81Do^=p{o85c!MN^TOxmEnE+lMpxz4BhyUFh9$D!5fzo7L&b_N z)EEK(B+QQKv+g}#1R9s&a0w;bFj7nu?O+)WhLQC&s!a?Qh-8cRN}{0}M1@>b>pg3s zR(OxBlRic2r5!XLJ!C6iP^I`}D>qK9^xY@k5*Q84NVyVgwKa_qE?ToVCR~`fdDGPu zoyJAu&k0;il$f``Cce|YRdf(oe6hM`xmTBKgS3_2?+x{Rz+|FU^aEX42d(qEz`L2> zAWe|>2pU~3hs4;I`8JBMk+(sLUe9%TD*qxm;nzV&P#;EEz6#J1s6ptMnEX+n{`MtV zhjFX+bAlQZjosMxs6zxg)jNSeJe8k97lkxPkschG$T9VbpMXzH5}*-#DPROi4R!{p z!skZ%W}ssR^g~%tXahxst#o$)+9~ZD)0no?q(pgv3|oFO zL4`g|*)H2Afd>055Nu7M`*{z(p+YYq2LfnWGb?xnvI;>ZQKbgQg&?(iGYHb$kk|0F z8=_+n`NN-o7W6ncg{sSMX=$*ebBKB-Mik4%Rvfbahv=;jnN~^E}`b&-y9!AdyNHu4gbDb z)?u6pe_V*bc7ab!C@Nv}utBiRvuCcpsdi&vzGoKQMan?T#Cpri$TMU1;<;5|}j!_J7tz(Sm(*d~fJP+@2hHcqYO9SAui%?&Abjq_h=KPbuz zHau>^Ivc2gIq>@W*Rp@m6pH^ zR`?{b>B?JkR8@4Ne-)iH9rj@8-@-l(xHiq6RC6Elldj_ej3CZR;V?xwNPVXr6 zzz99|UKvT_=EB%;TSBD)j0z|=pCY-iTXL@Pt)s((H}SpMSIKqiEgH+0 zV~sp#+yOH3IF>%!`PccG#u;dvZK5@iQ~u~fRDkVsMAMKU-}lDdg=z&Jmnu{jA-far z`H=?Blm@M1OM}ickWq##f?A=wE;cw%f}af9ef;4%Hhnr~e$p(T8=PRaR^Q{+VfYzK zPe1%!T+G3Qh&9tT8Pg1L8i3d=xJ-}|;mQ)pk78RSoiEn%Rc)>j5O3S{Ei0}irsmi@ z6JqLa>z)7Y+HaH^mpj?)ESo4`LU}b5JX0&`oVGi6iF@aydDPK~ZfZr(+@9cGk~pK< zy*cz6*+KUOr+DuV>I<%OItV=ncj)!JJNzzEZRIXV^X7iIrFe8~y@`jBIav?tXMgW+ zbksYorLVo!NY>hLU<62ph6P5FDRvVDv!KoYwF)rxucd`o-r#L@Euzc)RYk#=(yp90 zXonR7Sbi{WHi9cUtO|)?gL6X1gp!MSuSoRh5Zw{|j>J4STN((448u|?#ez&z5)~H7 z$C{Hthifw-vW6`Lkiop|f*^vt^1-PGr>Yfar#@JKrNOzv9@$w?wEB88?q(f}%rp*$ zaB)9uc-LzprgIr83Ut+wx}xuHkg<>9SNIe;{R0$)&9yVbOu)Dt?nytJY|O}DvmS^E zjsB?o5N1F)lMa^d;O@l=%xAgfHn= zl%{u{U22gV^dD^(w=vD3xxysL7r}kO+o$iKXqJ#ul)J{W@0 zC|6$nC=&$1RySCj`Jo+#Hu6aRkaJ=3xUraWQqzGWc zr6wf?W@QJ(0xkD8D(sP{6>e=bz`2R#&wF`AQ(@PG%hqKt;ON$!)pB|cF~jTXqzmWu z24^Ut!dEyJ!|+6KjTYBKmg9+okKsepxPJLR-+#wg!v689BpY6VObFFCF-gpNw?cW1 zI7xCu*c=)gob6T5FXta}kLK?MCRgZO4P0j59Qv6&QPSjJM@KuQDhA}`4uxJ{K#kgd zU|+o~TsN~hv@hV4TbHzl*BrVo@PWKLpx;q<-+H%V*BWsPl-q|38~ty;&?Zluao9&A zsOD!gCB8>ty~`Em2Ve3Zuv@&$d(=zOSn)K3E7Px|1C?n!{daiPuZ_ikZPo{d?eDPt z@FSICt(+cKQ$rnZ_6dtYkg#T_D3i~nz;9e$8Y>pzZm`K1ob*O z{fsR}cMMQ+onnCx<`NZF9&$x|R?-RnR2zGr(KhD*7XviDfBuQ6;|B>&!!T)hx!7p z1>j2fz%iFDpc5Rjpvmzm8#%L(XWFQY>60g+)->{%u1#e!J<3vE{O{kB1RIX&LS6MR zyOTk&5OqnR!b(EYoYG#*bIi1FR~`;Adl=Y}A-AF`lJPo5sWl)`+_K+Nr(9iDoXLI~(?dOr)jbolc6+PW^-q=cn>h18;;b z#_*zv?{0r3g2!7R4Th=C7|J1>3Wi03wjpT zS_VfAt|3!nDKr_4WB=0&Cj@%4vglmmCs4h?32(N$MAx3PWa?buJA8 z(63m4;(PlbSflGG>y~4u2f9NY*4#rG%>caR29C$?x7L}S8*jcQ)Pv&>MU($b*4gly zlVdR5TPSujnzAr$F9xpb+%R(F>KDv3y$Ml9t*#_nZP7W8Pyqu74ucE}KR%Te|F^MLE3y*#yN7m!7PFsF_IAwr0o4+ubi9)R-n z7hc*$)Sm6K()nt|=IMh0u(wH&H%R-PEBvjLP#h&*hWrcdt3?1}GzcfWOql%h&2QW3 z@sjuDe|=2CCqYs2@ZCfV#YRzN4HcF}V{pHn-tJt$Tz72-=8_hxxo<}QCd0SpJbA-A z#^)WkCvUSAgSod-69V;+ko`rqmF#+MT+|Z=$gZT=GK!Q^VOZy+D{=xl-DE|#c(YFt zk}NkV>g~6SyTl2=7^scmp*2E^l~P5%C_Ebh1BJP|aLi^>?w<ov9A3FY zz2G3PfX)lnCF%sjT#!+DR>9;X|eD62>D&*k{+hyD6S@nsxt zG5zdEieTC7LUeH9(H+LLI$T0)Hk^+(LBg}$2UR_nw8?vdlNT&OoyjWYvIR@ByxUl4 zf}FP`Ls=C3KmzoWdAuX@(v^cg;R}`|F?ZjI;b;A8_=2L~LEws6GU(GMX;oN>LvZwv zw=VjX@4q?OI1Kyvl4HZsGZVqqXs0eZojwg!ktiIEo&yNIBF{;`BW{;bZcBwsk|-A2 z0@Ur5^Hj-!*Zd#4Wq7K#dFBV5S2hD<(;<6}WXqH&pysv8!ej|ha4>IXUuxP8?)zcH z&yBaqkHf`=4Ty=YbH6hRLZ71d`gCc7th&9jDxCm7od;6=UG}={1v<19bc$0sugq5? zX^?Jav*mt1CGuEGceP@@=UrL%^lEoiwy*~L{f7YsPzt~L-8*mp6Bz(D zPROv`h5E>mN%qE+<~Cb7V1jFX6}j*DspFaXy3C0xlb7nyF6bh7D#z)8;|>Q%#z3+c zcCD&C5CL1~0ObWrEPcg9OJ3kUKfDqek6l*IQ|aoCFe+kslyfn1k1scm&RE>iaf0%M zoxC~so^|qzQ;V*MBdF^jS+|4EqHcZX3*kAs!e%khsUsa9Sc&14P)^`PDr69AN9L7mS$8N4zihOMr{KY zjCu&bsukf24$D+nuF@etFKLsXoVwQu70qr0+zl8Oa~tptpQ^ef+8=^FWWeMu9JEVf z>ZRSl)z#wvKpe&P1Z!kFrxXF^ie;#TcT?IX$6(!BUySkA_~eQY_@FRsCa;E{<)Bt% z2G;muEyk$PITqwSvQSSr7%e38f0L2(UA+f$Yi{)y19-vW+CEa^hT0)^W{j{sQOhQI#J)q`WSwySQ|BasBILh*_krqv-+TX;|MOe%3W{Au zk#K`4e%ct;_PT%xpr%w>(e5MngK+|sHmxHm&P4C%1~?QYUZG9 zCp#uXGU9}2PmUcerpF%3C$G=ZIJ)T3^5g2Q=SbXhgO&mVv}~u?R0x6dE@dKAums;SINm7D8R&%I9vX20@X#uEH9LT53p@FYK#Z1?>URjOBHop7^}E zLg0ymS@1*C`=5B;jq^Xyn~+t%FS|=l*l=giZ3ClliDEBMOR?`(m{My-!txG=SQrKJQdkt%PA@t0at|hlhbJ@oa}OqB%nnt}WZ z?x0&~EbwoU)Jy|0d)LD+sdU%j#(Ooi6Fjx(jw_Z7Hcn%D^JL(e#?nZM$wg_`Pd@9Z z>ud5plIED|nZ*2c8~yPzau1@zpuxzGW~%5aWj8N&(LHjSf5WjkG+dY))Zv{9fU}Y( zVlXwMLWsc#)_KDTA=C*btiqUTj!Oi_hAX8^Fy@ubt)sC%YOBvS&khLe=D4qA)XIDa z&ozgpcpj!({MXE?nSESw-66)l%{|;N1u_xqJ->7t2TY9$Qr7Qwop7rD_fx-KRJf?~ za@OTqz*0VvB|?uFO1op?vUqy`cr-p6mg+$z1TS^_mT`7_Fg_Q z@Lu*%EJ*+Cf}#sWzuz*Z%C~)X48N2<&7P5_`t5+4$RV9WwF0w}y`uXNGSSh>LFbb) z-ZRzj9vP5f8)79B&%^t5x`9JJ#5U!}P zZVDUbcGNhrFfNRT)&{R(c72bAL%_zmQ5m{&o!)eDvn}fC=x}5A-8=8q|8)0 z$*xAxQ5kFmbfR73uy7mvJI#Lh26tKi4X1~1yP-_gZ)>hT>x4Xam=1ZaaOn$);8`Xz z`E*fmlR~Y80Bn^+GYuq9XWp2V35!q$Juu^h-+1BIY}^j5E!Kyu_8Yk2_9IbPxN%5@ z%ZA2wflo{XU>c-(%vDzu%Z+A=g3%?Zc4?ER(Ai$?0l3kr65tgFeY5CZs4Y#k?~|x1 zNgWhBLEl3_nj0rPCj;h$RD0bAcc6j>iw|;uMR(AymG0J#)uv^~RV&(E z;#{E80WYnEI15JpI-Qe%Mut3&cdolsH#G?e`#0H(#}Q&8sn!;kNe zC9Pz`5g8L{r7pVD^)S>!6}hT;UrLK*EBPHzA9|9MheW#FlJ0Sdlhx5HC+&Xe^A~&E zvY6gEcN}-|@paXtr_J$G9<>=Cj}j$jj6AkW{_oq~&c-1b6S6+L7hVyEL&Oi1i<#!9Ude=V@@8)!%S(GHmxJcKYbXSiXLBnz{Lhhsg?V ztCQ{i%SgJx>a>w!<0ujXIcebki6H5~0Mr!P`rT&0`kBxu(ghX0z2uxI#`mB2*8X1UF#Vuv-kZ3By7H8z^=?Mb=SaDNZTC&;cQV7D<`1i(oVJfUH)W zL60*c;PIPj^sH{=jKAQ==cc}B+?>W`uVlm4!bFeV;(XC^rjSVwt{12kw?rotAWt== zXi5i-%-Cv0v20-clr(WR%tVGcexvc^j_LhE^^(UrTAveSm}tyRkn4yN+L&thPn8Fm z$eGz*he$s!A9#_g-77@7bcbkMEr+c_BYV~NUK8n8lGVvq|C2=7aMmZ+U~Sk+u}Kt3 zq{6P!k!;gE-G8BO1%CZK^aj$PNbp$yVzGGfCbue*NiPM042-HYS(~zM=$Fc6#&uDS zp*yYn%FQ<}KN00;T<~VIeXS;n+Y)%?3bmpZN~3qs*rJM|vm#ej?9_5l<6bjQg(+dA zO-OdIc3I=2ndJX$SjS9W-2tD>*8A;>jhT5>lQE6am3MdwYm~*MH z4CQ^gMsl82D~?WwLPuq@iY!GBN`!>^I7mmeeRJ^OhZ{B=r0&7s(b zc1@lPtCCuAp8>appCCW~pRkh`;?>qzi8v!{q)4k3!1n!x9zQh}I($xikrD&UJd!}n z7G8nyXpdw~Xf{u+I0dpT6{0h2W}t>{4n?9SAm$gZ0{WJFzdpX`&f6CjZlFH6_N}cy zy1nT7JB6~$z)g}2?+U9rD`SF+6(BM+dzUu-i4ILC_-E`{EjAzz&(<&XHXUKK$@QOuiuVIm&+A42oiJ1|@-L@PHkB6ydFz zhP@qsTPn1gV>2e(flH2+pp*GtT(~hoy3P8nCdkmC57e<5awAKHY2xedjWp6_G>2w0 zP4iBPis(a7nRGd*L)^j~vMiV16K90ud_Qq4OMb}7aiw1EdhOL`eOH6+p1ULu#0m0l zS#NMAFiYv0JzM-YE{OL~EAU(${o8HSZ>`~8jyQK5!C|f6{$qa|H*hq@41VaOcRSJq z8?TXQ8xD!&88{MZicO}-CMpcYM5Ca6?>>!PTUx^E7D;|kpB!2~fqV=z{XVInKYCPx zw4v=XGs`$0d$Jr)a0Q4J7YKvDoD)DMq`ua$$~e$&vulfq(0PGtj|3@(hf@#V(NYhm z(GjF~ys3xH^f$G8RD+qb_8U&fpjK2ZF8+G$SYV6d-6{ZrL%FKaJ73fajf3|{gw;!; zN0@f=Y|98YTKUv@-hutkdW8Km$9z$oe+(Zg-vt@4AHiLSs?+|8*C)y*|2Y5H;AV)3 zmB_ZSke)>@obmBE%$sqo%^3QRmFHSFJSUqmvG5Pw4PP&t(Ey>-1bHXqY*o!5EZ8ke z4n!8V@d1qi0?b2@!7bzR<`efngar(_pq}X)N&a&a`8sakuvSnk=xpr6zt7Jh%N9Qj7_@6rbkN9hP{E60k&4YScry<`L)OR4R|`PwHxl=J z<5uyV22VbPVz*FaGs+3~J06*rMHj=vQ6+^QV_Shi5&HdSC-78>PEl+m+3J%x_ISc+35 zS;@h%roXxDB`iWahP*d=LU9-c%sb$J4h|jkK7m>toZuY%3 zYme7hAZQ+z%v^blFB=2qj{H=jGH&d9oGHnM5d%uu!wONiQ0!*>W5ZHC9=bNLJ7x}k zBh4Fm&ynJ>cMghThflrgQ|Py8YO}0!ZX!Dv5IugoV6Gq-!@P33)%02^U3DbjS!X|Q zf_kDmj-B*QhwZ#Gg4#LIF9+-+SY6Z{nhLJyUP(Pa)@iw42VE;n@yZt+lmSI2-43cv z<)SwCz2M}I<%j?DiQtNgVZZ*T4?cz~Orv51)+1{6ga7*M>(xP3!=}*9^K}13o>(M( zO^}BR8|QbrVx@Djtkx|W3R{~)ugJ4q3YkMBi>dV3;aurqT|D8Gj}lX3K#O^|^7p6z zuW?6&&Bo+S^f<(O0s{p|otN`fnSr11PAVUWEh1uW-V-#q%FJmtxMw`xa55beKkl5a z{GC3GvhO`-DY*F=Rh0M|8koa=;C z6F~XlH)f)5*m^ZShn&zi;csh-cKkm*`X)a7*>s|LZi0Br46u5aV!J5v1r_$!3i`Ty zk4yty;MZkPR*kf>c&gGD(B#@L2lb#V8ViHz67qR?B3ri-jEy&*$5_? zpdt|j(P1%E#21`tr|s)9-Itj+|Ms<&_I29VcKTlGbf)b?1jU6#1r*c($|8%n;0DS% zD2^*QQR0FNi|9DGfFgqe|KBeOGbC~~7ZP5^zVS11@45G$V9xjc&hPxr@Av&?AbCNP zf1c~0FaE8?zY;n5ZdoA*xO;`^1Orj~NFfj%K;8L=z88dA)0ye&!LKE8wNE_H3X#RO zQ`%<~$Sf43_-HX93R81im5Dq=_nxx*73@H=5Q>16F9IL%R)hkhCf2dSP_ z@{>@X9m~FP`smC1Nz{(a){I?$&JggI**rZK5akP)2 z=wsuQ;#oRX6WtlAMXi_1oEi?2N})kI&D|$YozVgF9AEs=B(q~hd!yz>E0R_LQ zcx0_z#~=}R=`H)7u`=ANP=d z8z~V!3O~#*4l`&vxR_Z|;A_wnazW%;)dOd_+&tot@1NyY{n>Xu{9x$ovFA@#Z>{qN z#}1xSU%e_`3Pc*+Q~G(=#BG4sji6Q6Dgv4=mv!WPz-?%A2sWAU z#0IlWeK?|>9cJm5H8{;gGB7x_?Dz`@%yeM*G2{L7r% zUrB#`&sz=Be7A0HhZ>ks__h24Tnvw#_+{!Fb{j}GgEZWjFdf@#z}O9v(+ceVJ>SS@Lkc1bfNWvUV(Mm}_qc7)u|I@NNY zLn`fcaW{>@lRHz&rwzK)csILs%MZ^(HZqjPf#-L*b*RrzeG*~+abq<$OhyFT?Jyq0 zyI?mqPHRu}TZ_7_lP}q=vkuBrS@TfzZna$hMd&(nU^uj@;-F!@nLQs202YHVW%gm=s zWM@5hgeLR0a&nb;o+@vbwU6Hl#}*_EpWhrh{KGnDM#2c2(Gwnh?|<@a%&WxB6m$i1 zK%`UKAS)H#q_gQbM&+qfSo>**}z1Fu#8-Vm9`T@FcRnKTM|jspy=M$kr7 zSUu(VQ&y;OUT@8Fen{WHo1-H0mbT9!%}%@%{n)~@?4%SQQ=|jcMR8X--M>;;!^z+@ zNc(*sf=)yBj23<}uN%ZO3T9sN#f{u}s!*g?orFd1@F#E!2LzBIck&8Wq>gk;oBW~M zV;YIu@IyL2f^giF{z(Od09`R<=ui5C?rZRB>uA1M%Zx&nr zdLzDNHogeMrz*Q2kMvrxdJ!fpSz{=lZpNEBVPzHB^1|#rL%qnbyv!Yx0unXzv4EtK z-s{~h*QwI|D_&aq60YYBULakHo(?i-q)uLPB%qkan&?ttg-O z?B?$EelW9}o6Wr>i54sw5gj9q44XE4po`7t*?RA0&nLHC%y=pO^tbPmXeYK$Itv6O zQ;GzNtinPT%vI@@ZpXrom~dW~9d;;(I~w_v z0yyilsi+?D0d5Xx#I?zbxwYaB*qe;uUE_BMo0O3C&#Mhs>bg!)%)Q316)$x&XnKjM zrASb!xuVSU%L$1cQ@PLcN5n>?jO4UFzw?u-rIdtW_Q6fS7(sy; z0&pMCD-pKJ#wEV<^bpyCh^L=&d`pD)i96Qv7EbG7FbGy;(kE4kqO_@5u4}pZ(~NRl zm@2(RImEr>Fy_4x$H$hT9(waf?7%^dtQ?IA0m*m1`k&_2$SXUJOeG&rBJma}$9)j6PKIBO>Z{Z^El`-t9P5`cwq1B>x?UcRV zdR3bW?@SE`F||(V2f)$_B;pXU#m2LASGZpF_Erd7AX!C>Ad}97iY{;jFdFjEt&*-3 zrU&%Ey0eDU8e9))?3Kpf1Y_@mF6Dm4TeE=RLW|LgJUW?V@#<%^&BXuX-3DE{L76Q_ z-RF@>w}W@OHmqB|^`#nbNE49VqqU7v+=NcyR*`o4p;)idt17upn|I7^jGZ=s`^JAK zab7or@8BE0Z;=yDycE7=;iFul6b6c%qoNvwN4?7^s0|SnLSlWg!1{bB-ztq8AJ7g} zB=Gj#n^>8+Qsr3UV*hh)%oq)oQNbD4nY zxuNTp&*?Q-uXavjyC==Yv<;cKG5oaI5VC?j^;_kS-N#bR=)_(SgB~1aTqnspTb0_4lH0-9@g601JEjaiTahukX=n8KVKoI!q$O_t)*0GP zyX;~}9oMO=$R){*;9W$oiVi%&(H;xGG2fK#QYh-66BJ1IRZQXLWL8LmqL?zkMvrMd zfD*`=NcTd&y@zv!gVR%@=pMZrc81E-YoI7-NOxHex9L^I&y}9Fb=>TZ9$P>7Irp(c z&*r^*e`(E*{B*{g6NmK}7?X><4!Pe~4EjQ)G(V5NPahCn5p9(W`nCX>$1$(cInClT zqUH0}kuKh2vQL~iUg>n~LGzr^w$o*1&|Lp(_Khg>H1#OxuQJl|!fZW9Tf`u`D8*fh zd_qMv0_9-4#AHm0!|FK5PS^$$o%Xss+Z}hLGwF)yIKwrAl=T{~R3zcfQfvyzaZlkU zL0n;v5+WE_q-aV_zf4l)ah@xA+iAULL8$hcxLfwnEyE)fMBpo?S4`KbT7!GYpo`IP zfkb>heVydbPL*R}Jx+OgX>9=PcU|E$LI!*$ozJNir_G3`Is*GVwsAXTP2g6+92vA% z8OvKL02S_bP5TPz6;?;U9}J8U8)k`Iz-^+xhdEx?{FJ zBP4B3>^>Rw;ET@J-UL#}F;K18;&w6c1WZ8fvRq{~z12rAKBelDulm=Uk>iC4)0nV8 z2Gbxqq@5KqsPxsjuUoIPpLS-Pc;A{~HEoKDn=>COL6fc%NW1&oTinJMBsju00>vR< zcvh#(Kv~)~FMO)mfoWeDe3fi+;&pbJ#X7r)Qo!VrOGTk4fr}q)866#gm2O5#l$)@_ za5FTY-sPP-S)0ZCNI4)|8{R=Ikl>tI^)+T> zj4-)O*m(P$);VUhEV^E>h7>w6T8>$urJPb6pvYb-3YmJ!U`gCVFn12>6Oeud)x*#u z1P>8-3-RY8_&MmZ6Ow-0;Q{T)H)wM~h%8l(%v3e%6A;{uc8jOV=?7sYp_M*kioRG3 zs2#v#2`r1=M}NW&Jm)jNM|+vU^VXRO=SYSVg9nQChb7(?Qi^l+r(GEjAkQau2U8H7fW;zdjf0aty9`T`>@9^VHR@#@AAft^L9yANLEO;`j8r9 z%^954l1@GvwoKnfDbm;#im)ClhU9U(!;r_WEE0+F(7de;!>g+Uz$!vJuFP@ozKW!U z?Gv3MjSy)*rK*->g{==~xjivPAX-|i=bW*@=pP26Qml)X*z7ksv4@;(0j>3vBAFry zm@HmOnxRC+lr%o*f=UuC;3OM3j&XLLH5xXe;#pld%wxDcrB|V4F&1&+lyyR}+pjdN3`BVcUGn|chvTk7 zqx6CMTHfE6TwU}qQm4}SG)R(p+o%fFfXpFSvJo!NG!Krz%dh{ibCR_*uhYih7{qp$ zhtvji1e=!6-SSj9=9E?Pce%l^PFbPG5bFv-73rjr5X9uwV1kxjg$ZZz4qsi5dgL83 z+{5~0<;P6eSo)32*4ce*CL1SCfni8eMBdvNflj^0uLSm|G3T$H?g0N~Kz7g-Cmt{% zkxTvus77PH*5@7_0oq37iMRq-<&R4`r4`(*AkukzUfitpK_-c#48Kge1!~K$g$z;D zfV>f;H>QQ`^Vg{k2k2F2>0a0#HA;WtXEzI5_?WtwB{FdVUzcx%jKeB&kgN>qkMfs4Yp;MtJug!*C zgFSNGnZO6KLRQMPPMf3{>{D{fx`3Yz_?&oSlR@7pQB)Ox+dw*17Vpz}1`U!yBSBI> zEcAQ%s74Si*dcEBZGnQ?#SOxE-!r1c4w<)x84Nrel`*sG!x5*O3l{#-+wAr9D^9E= zx=G})g}YEfDS&)uI~BE8*|Y%v*BdGNONA%NLlEWO3!FPE{Z@zRH5)h$!bg%WVU~EW zxQlmAlgKe>n&n*!n}qppdx4*2i?ReL?YiZOoMoI$PLt3vP10YOl@mmaheZJkCygjb$4)|`xYexfwzEaz$X#S;=81Uzmne_ zb`8|XOTAFZz@UJ{h+cw{23U}*#l+xBE~*Y<9NJ+3vk65OR?qUpF&LUK;lQgCL=rPr zg!hGiAjwV~11_-`{VYlWrEeKSRXC<#uNI?LpZucnn%b%@R9gt#kr%yQ><7Vuy|x{oYYF(Aw5toEq%U!}k+4jh`% z4av~E=%@i<>5VvVz;oyM{H-2lbMnh(vY4!$M0Q%dwb_&cX8cVkErVJ1i5y&t8mrUx zdjTJcT3g9afWY}B2?js*`C_1>*1r@ zAwY6URm>f_?E`KLeP6PyUX1 zp7~S9pKg$oPVBY3!pFy)aAW2g#sfH={*zBUvQVvFdu8A`^w_a|9 zV%uJIwR#t~O^zIS_{0tOJasmpFiv2zr5ZbwMuHb}o2|fOhSCTwFZJn^l0V7K@Y=98 z-~vf=VtDPffL9)+*hY~oDymFeH@g9Hrg}&tP|l^wS4@hZq}L35{qmQuzLNGTs(R|B z8$>%iv*=ABH|Kr8Jrd9W&iIN+Ro}ewt=`4)lWZPxTZh2%zR?54vLAb&u>;Drkbzsz zpSGOW7b6VQR*T2Mz#X$%xY{O<1dj&c+9_S6Y6g@CbjmM)OwC!aCnv}e*w4Z3XWY+0 z&t0dw4%t-5X_zU>3K?|4(nk!j{NJ?o`zSH3a3K`6p<~ep|6D9Yx;dZBUN)yCB7b%d zto6>2|0O1PTlRbGZAonE@0U?O`@vWk8Ju`A#K6e#sjL&6i>et9gE1Ur&=d*!=xRwB z)kb=RZL$?^7leD1nRK>$`Ak%n#k#gO?~R~J+!N;D841Q~l#`N8BR!UzdR$(A$(sAn zX_H$R*bGf+Es3H=X@VjvWWQImppBCQf;k&qx1x1Jz){RnaCyY z85=SE|Ezbe8WbHlxV)*Fzd1y^jr4Hv?sCvFST$eULSynQ>ihz;R`fjdU34nUqQdME zgW_%YZ|o+;CNMFtUzo0p8Fb8G%9(Thch){DoA}!o&zoHvR!J{evZReu8xSMtmhB<& z0yw=!e6maYm=`HYoOz1t&~>=?Upr2OCvuL-2tMh)X8rjha_txRl`_yhpcP? zRy7Pzn;I)|R0^j_d*y>raQj5x99%COJ8>BlHPoa5W? zyN+Z?Duw418zudqC|pk8QFaQuA#<|M{h+v5a)Q$wh|-_u6gS2vV}%7oKAUeGiVbt; zoVtWHlg*~%=6^K*ovdVMI-Gc>f$FnirX!tFfIGgPirVjmY5IkNED%`I#`@-ZU?`+C z4CIyQK6ji)42p5q-(-i4@j!*?GiLv6cION;He%nIx`t$qy7wLoy*e&6CA==BroY40_5CQ{%6Hv zvT+jGZ!rklDMcjVUC%))vo4xKw zBm=Tskrs(9wdvaU^kK|*B=mFyL--H*&(WcpjD+Q!}In<}q(ZU0w~M&|r9cG0RI zUVGD^F%f<)n|DRr0ySROgUkK;yUa08l8NMIXbV4EjmsS@Os@>WRM^d~n?1FM{7;M9 zmFuFg7ZO8cfZxTD4E*$wZ5;O(`1Qn2iHT+dd|WGMT2Y~c|@w|eWacqeitMcb6WWK zzIs)RcU1)S(OKeZ^$95B!u@q*`P8XCzyE~&=!0$g-bZ>@8QSbM0PNmerWVvbyT_S>jRj)u1xB0tVTITUAD>NjD@_C9#s=;MVakj;ZY93~m~1wP^qy^=1M;Piga-OAvvzpK zsaxFEIwEhC1!x!y8#dd716DV&0uAN*(eZd|vKuxIiW4U{GH~|n@jD5uv#IhY5plCJ z=^jP(5t`(n8YDN|js&3lv(@K=X(&7$Kgpmu5>O!6DJ_KpQEcmka?)j_ zgM1{}AsKL4;q{1a2^nxXO47Y@>5V|{XTJrvop3b-IJ@ENtFga3Y0Y!uw4pc#0v3$B zf|ez%EfQ>1r1->!RC152(_Kw+VJ2dHEherRI9J73Vv#AT6eihGM1grQMmg%t9op#% zvo#rMyE1HJgwF!=lF0qxu~c$^T{Pc`EesIk4|7)PDFxK>)F2Vp4Oi4aMH;pSNvXF% zbI&y$NL}?4JN*i{>FQ(j;b}V6oyh(lwJ!QoHdXU#)erTWRhnbs(y$gj%w=I%mA*^e zCfhAS9_!e^gRU8F^+d0^|Jv2ppL}J1Wb-Rq-Z0g!=NRb?^qM=7YhGLTTH%jl7h!v* z-Ak5<)<6EBTH6OY@A~%dYroeIm zk;8|wZu#bEm7&#Sz$G4J274jiIP~n8hWMGoFvjtJX5VAE$ja_elQj3<3pLvvp=;ze zvg?JhJEtw|PBo=~B4C_q*Dink4}_!fj6bS!PkQgngUMG%Uv^(mnDl+4EvHT{3>1^ zw^evkTEN*b>zIF;d_dOdx=Wd@zV)(QP&h`wjc9Pj0<@uLf7;~sM{9~Mr*&@_NViOD zqn^3Ru9t*M^c+l&yzPlT&l0-~-OyKM%THT}*_OWSc@HbQGV#~hUb+8V{!Fx>i+_Cj zNA5#E4VorS7wlm*@NbGs=|u1p9arcQ)60e><^JSCV=Knk#n}&z2 zxEs-qJg0FiIQ;3qrl(lz<-QcTMI8x!3;&+S&1jPqj(;TEz^#?x$P6ObjM!qB@JZ%TclWO|9Jl6dd?1Uq;fwo(opg<93lfl>& z5er1sq=54Q>;#yaQA-1wAW0K;DY2HlLAKrxD?wre5DeV!)h0Kn2j*?_Fjk)Jm{O`f zAljnLr_WE-p5$B)#z^4pm-cd&23?+gAR@)5V)}WqDdavc&9y+40M&{JK4X1J3WR)& zJ>ot@DSb<}UzQu8-K5&2T06A@l!tQFSm3Br;cqzTg1;hEsW`3T2CFGSa(RRG>cVUq)XQZ}N$iMhn*PDrOXg-{w{E8o2tGX)y>VGfov)_l}!!{Cffv(1&P7=&a0T|d_q==xcx(-tRkH2=wE?&p{s}*2L1oZZ&rMUWO-Yee@EkuajS?j` zjPuxTVdurXUUc9OZ<{SpsrP9a>3Cs+!*LdCu}75R0Y&apQ8yy5ikDKY^23o2WlthH zr`{CzPPs?d@(%k|&DbYe8+1XqesT+0F$qmn3m^H~4+b7n=`}GVmz#rZ5#BfE^n-j_ zHDvvrpStU%rJQ`u`OqW0K6!=$^bY5ND#fIu@H{>RJqV9*?kP7)E(aWuTnnjiTf^H6 z*%{Y`jX^jY;#T^0^@rr9_^{tTkxmsiN$1l-@X0g6Sl;qU8GenCA1RLl#TK^T6lb{& zwV#1@g9h^r9~1n2R|e!NJExY4|F)G*;#z=zlsFy9Mvb{`II{n%r=D+3_~NwoiGkcD zdUEt+lkh4|Sh?zC$iXTS>``7^m^xXjh5VbU`B`{?JLq3Da6U zNLS+iI!ez2Nf6!|PYwOC7{%e=9Q$MYVu1bUGd;n|R!xx19*B-G+bZ8>tBc4%C-#=W zIU2SlevDEap-3&TcyOEiugcThi`_pz(rePG(?7ZynewgMk?US-i8=>CKnyIO^D5w` zx$hO5?!-7yzc^1E4Y8KOnRzqsxgGW1D#4yi;o@9%zZlOjadeGhWdzdw*YZEVL#MhQ z_?UARe5^46%f3*uv?xq4v8A0Alqk`^*L+86hSKqGe(MTZ=fqIjWdWtFlme=qHd9e~ zbc`{K^9e|(;@+iBo%i*w$jbrxP^`AOE1HRcRbY^AMm@|sz`UG21F@u2 z56Frs{BvF24zeOm!VVcK_hI37n;f0z=*i7?;4lN?(5%t1)Ys^R6* zhv!}9tduqZEzN3D402d4{40le)O1ZH?*KE8@A>w%%v-ZHP6{88zn9+ zrn^y|JQL6@Ke#!~dTYXIjS<6UL-K-H-#qRuGW1igIrRF`pJpxE@H(!T(xEC-r-~LV zcW>bD3hyGD{nt;1=!!>iSZ{C(f2ro`oScvf_bf;pE%3?~U7vB(b`i5@ezQ+q96JP_ z>e4O^tV|P?ax40v_3XuFFUyIu)fwixG~Y|$A$QR+l4$Q9I)+?>?Q+OO{#FeH49{tL zU=B=|^gz%r$>)kPRen`m9$M@u2w9HzXaHr|gJaGzHzp=n(`6P|nEg$ZVj~5=e$)xF zBNRDwZ+ouxfL+1s1dCoVp`@L@Bmpvgy{Zj@r4`dFLXVF_=-%{ou>M*d^DRErE35a^ zd9w&AzW+&?tJ$}y|5EioNGv0&qiZT*cwL>{Bx(nL0up=OW zhoEl(jIZ_>pT*E;XtEvMpFJ`z2b~Mnxm&}+X&n)UfktZv44znk+d>;hy-HXjbQm@J zv&Y3*2b!H3qyBogU1YWv+5EKYB+-cjvcT*=EVi$s6q_irk&5b>+ddtGB$(jPNh3G+ zdihd80WkRAf&7LK1$iNufq;0)pH?+l=K!9v5s_j$K!z3`BX(o=Ms@(Xbz1YC*=B&e z_kP=9vi*gbPL5i@=^&-pN0B{1KVvj4gD&~BPK~_G6>i<)vIXcR?V1hQE_hI+GN;<4 z0$ow6JX?JTme811wUe|#I?bSqF(ro1Sy-dSbQq*iOysTc1X4vwX;6{z4m(au?144} zxRHL^4{!37!XLb4hR(S;7blPlPHbBqS%BddrMN+nRxG8>q`Q=z5NO>Q(V|LM9SVO4 zx|zF`g+S7q?S5BP4Og!5D?-=sIys++kxdbc)OJQ3m$%S~LHpev0U_Jsp&RbgCG_ry zoe_%-np`!?%Ij2%v2F2+XyG{dwhpc3ytuNf7WChBtE-++zwbl zVuFaCdMDg$em3OGvdDgRGnW(Z6zMID)lo|E0Y$2*s83~uB7+yIC!QcHeYMHFPH7I5 zx?sfD*w8KSl9u~niGL-x146+EYdCEX`4 z5gu13gkTj*sa#vhy9af1`H)=NLGMxa$ZH^6*a?_BeFkZp{{WUd#lejaonsSB`px7-^Fc{W2GFFIO#ccB%fX(Xrenq zN19wl`HeF@Hadhb^=7~J?jId=_cNQC`d8MKksK%XbZRY($6iVSNx&skRD&=v=qmr# zoLs?;`Bec27Ca#L71iO(0@K`2yQAk35Bo=LoO*hdZ=ZLqe`MrEioi zGQ(!xl8;(Q#YozFPCO4?vVhVlN&ytc^&lP1A2J}uAPug_F?9i#N||(KV6>o;lRrCS z%4yht)MEO05uN9{gS3%KA=2tKkp@|1kV&QmHz@B&^JnW-2c$3)b<5jnq(VxXUog}3 z0N$`uc6zQBJMQzSn$aADb>Vms68A*Cl8umXGdi&sdtvNp)D3H~0aMnuFk|eGuPeUE zG2`ivANCw0%U&3d2ejCRg?`pkie!o;AT6V*7*<;mTI3IZt(rgRvft}dNoCOJ#@zb5 zW^-;cUf6VMhkL^o&fjjm1^psWGNOliNL}ln#nEO#?K4o!K#&HjP#*XJBPT{+s%Gax zP_TS%)*N(iFG*4ZZ^#E-Fex0vfCr3MVEO(A?i;KKSrOt855{I#46_Q>hju{0Q(Z8d zkv|`O^DS$@IPF@RA!h-}Jw9_#5qIfg;uH_fjbc`I4Cye^Tr6WM8<`3+^pjg2cW|8T^%h( zEPJv487quTNEB{LwHCZ!lh@+Ji)aSj4pX9A3xB8_w={Inr8E?0txURwbVOYa=<}%i zx52NqM`9eNQ`!e}6Q*WdlH^Z|l@7X~T=HrU(@zvjJ|=Ec6?&$KDi-d5p78$Ly!An6 zg~w%f^mRrHHtQK{=-ghgJNf#q`SPJ)-r9f*B#oVG>%?hMdo5hsJW2t2p;=Va0d9;> zwrEpG3w@in+HH^GF};rT(VIdx%-R$Z3z?Tt`2Xrwdbj8#WYr~d)_4xxuMeuHuP8S; z#+mtiI7bhZ&--x1(U#lg)7{KKc~5)(zev0jhhp+9hJGWZNTtX+D(X&fb5N`D!MwDP zeg1p=TVX#ERfPN8v8_#h_e+VQZR4BP!e+1?F{n&^vYLV>w$5$7`$gX)GQ~hHNp^9Y zg(*G|fyD@m!(EUOzJHq_F~W0>Zej(A3Ey4)y>D5kzq}w`2}80>C*2%$gc}DX^OgP? z3T;PlqqGEs6oin2(l z6Hh0N7SqWGlmY~nDyXO&bv|7rfQ@BPPAc**bkkybf$5|)EPwXm{h}uU=6n{Uv)ePf z1at*(&S{f3`DjyoAodTdAB^DM3En3IIbwtAI$7uUb}uPgfQ#u&S}$(oXVTFT+89xc z=$LF%$e>H~WVlVI$_iNtcpANd8VxS2VDhZfQA3EGnfZy&rc#9&MedJ#E|LvSj3UT_ z8Ww8JrxZC9$;PxO)1F`skS>D~PO2OOZl%CD^_cFLH^=}$8zr}cqXnJ94qAIsl`*A? zw9D2`O(EBXxwIaL7O)%!b}t)+*b4%RyHYKB2zI75=e z?SkSBgJut>LAcKMj3|SYYU;RIyz==a+?<6yP>1%RsNeUTw0hhM5m;BSjv50jA7qA% zWRh&>D`rFTgq!^qseECU!Iv$}&S^?gB0>)pyV)Bo&_7&yZm%hu&)7G9qDu5qKxzG5565-15ObN5S|g_yVY>?ng{00jMNO zv4$cmsi;ChDdbuXKzi?WNL|ST%3baFxVTmWWD_E+o*9o*?;rYpJQlTTCl1sys8w&C zk}yZRGa_1`O>=7&_mX(Q335pi8!^gPeGG$S6DAz7-}9W``Q>x#erufp=d?S53^{F> zT2e<&i%*M}%TAFFVKtq|>699PUk+&&&eChc3q>g)nv_c?f$7(&R*h3w#l|roGV}QJ zIiqEt%dDmWYT1hK?HP+As1sWa21U?TWuM2j5X^cjpKHpVwf=j2?FS1S*#{st_7c+4 zzV)BW!&vQ+;&WDra^~Q#KJiTC=~VycVGKjd68_0grkQ=tZx{XQw`B7Rvxq6T@HvYq z1=Qu|QBf`Y^WHInYFRaX1H`TLp?$QGiCeJB7c*+FhBp9X1}a4?K22_d(sU}&oi4@z z?LN^B&EgHB{ciQHdw~+AfD;?oMOP^uBTmA?fDE6EERKhbvfU21Md76IND{ zEiVjQ4q3ou2c>}MNj??T&)fCVPQXEtcae9hyqaD`?kOAi8$^i_11=5xd&P($BIqXDgbOY|C~56pFuMK z>$rN|8zK5p&C8#iqb~O|Qf8olc_qD@a|w0@@s@nLOPm^l0uef&bgz!k%`-1Z`s4#H zNb^+%^fo z?1<{1PP!44tMw|NvvX;oyTWrl;&>f2zJi6K>hRs1)r+D9Nsz72g28a~uZX2Bvh{3) zw_(Jl!M5ffe$mrVs%8DWJbJ16Fh7RehjPIj_kJGc9U)0m5B-STBAdo(jcopk*$noG zqvD`OcJ^>*YvZPcW`9Qilh|BR_QLoxO&0#lhm-;+7CxY&ZoPcSw>#{RZ_mqTNP~16 zw{6~rSvu7T#YOIsur7DKCfmJRp3Pkuo~t|_-s_Iy)6s$L)AQ6_%6!RcSq7)$UyHtW zKXSSIZdpBD7Jlv5)!)?llnDFWE2f{CoF-Zylo+7b#C`Y958M8&`PYBh_U-gVr+RLzG1)OI>XkWXI=*{$3Qb5Oh+PF-0c7a^q|Qy8$doo@Gg+wm*#(PCuuqi@@j z(;ocW3TrxTHW+c@&>RE#cAp1IU$xWSVOYVwb4mgTfV9)Q{j{GSUXq-LB|irO2M%hYb2qilHJg&M+3&6(MNXXk_@RZ_ zI7BH9Qe+<$g?Wugc#{asxsS+A>E@8NA~PYuP3avFmB#ega{58o;n|tK&BF3k)u66C7IJeU-WI_m^7x4p&*eH9asqTUYLvmta2FU#m4J$)HZwDF77HjJXjcz zXPk^3B+O=Hgao4r>pVO!TkkA8ZSDcXUhy$NPMiFOrZ}vRd+X)hbadbqNrP+`=c+tk za)R8YZwHT~nScRgfsC;aH;W!vA%pr^O#7Gqxv0|h986fe5vdP7EW9R0UaxL$2Ongo z;Jng*m7%9fK1Vl!#o#>aRu+Kz%(OhKKYLx?uhF)(pTEVfudyc3XA==|;*B2$9e^0Y z3c+nC#L>3Xg`#_|_vub4&>SGU?cSNS)3s?H4Ui36z)ABkDB`Jcnu}~j$7cq^Q5(iq zu=)PMKm3C>yQ@9l_c=jg*o8!#7#&#_&h0u%kwlR-R8$;qCn(782sHwtO@436RUa+l z<^N2^XV94Z>94+JU6=EMm<)!B9i*Zt@-LR;(`X$E1%r@cQU-cZ$bWs$RVP_G4OF9y zi{5dB30s@-+4wp5;Qj;m{IP^lo!HM{2%}!+R4nWWO@8TZ40*IFQ8`TKvwm{2>m^tx zXNii(YZfsNEDK;b7z&o%|NX~1Kd|N;V4#0)r_1La35c7e?enMv5*XAn7{@nj*spEV zcRB3-@j7I+I&|JQin>+4f-?IT{T}_N$-oO!a<$pQzeuDMt0@wXR5o{oz4B(E_NZ&S z1ha%3zZp9cCT+p>NY6RC?dFChbFBg4v=J_bAx64Z`YhH{NMfB(`+GrP_w0ptY^0ZapLlD;F=z`DZ7nQz>;JO%4~G-v(6BLDlM`n5J@%A8DM8zF2n9u5xmGPMH)s3(H z)kcpKmv@&W1>(c1z25kLyxaJH>BfJ!Ew9+Kz9C0lWB1KFZK8WubLvljXP!^qPl}&Q zZaVP>?DC1jXNX=(@sJ{2Kw?BE$y;CT`|{sgUo-wlm17I$fkH(T2qne9jPbXJvdvSv zxuE_yrE4DO8IU&4ZhFJ4WBxi-mnXaD};h>g(<7VVvnqJxb)uNx)ikg5=;^M;N>b(m~ zHBTa&Bj1NEw*RzZ(P91tMX%((d)15#zTUXe*UsA zV>ZfEzPBXC*h7WedrRV9z4PUBngN&iSAT8GJMr0MGmZzW-U({!+MEAsy#eF2zAeL+ z$|3*QsVhRXm{OV!R9eZro6>Tijo1@;H`4Lmhm~ow6%|&GIsB9rDkf~Y@kYjX%m(RC z9e=t(PL7mZ=DbUI^09>*dzn&PpvZYD$|OaKQPCz&8>ds4DXJ9y85LOE=3v)u-fM2Vf-yyr_S^#Oz zI&~w8H%l_kil4 zLsqt6ztfAr;jo z$Cc|!`6ei8IvkKquKFBMYmwPZr`qaMwqQ@-aS7g_Q`O2>X)4_Igco~l_c%*yyFgjA zmuPGK6Cq^~L*+<>*WQUnB6l6KB`PKdJz8Eu$^X{^pc~YwL zF{7_|@F)987Q5BE^Zv_7rG?qsMJa%5r4U{s(k;v4J%%@-cKRu(AdC@Yig30}@yQ1| zl{`9`L{m!zxWm{jJL6TZ-VyqcKBd|`r66SGT^+2`9KGI^$=iWy? z^41n8V?$Eq1G6pzFb7@GY@4X{vnFd%|K;3oj(imivAiC{%`OO0%kaJ%d=Rjjf^>{- zZj5V|B0+)4-VLt1{Z>v+oeV{*!xDP!f+f4nnb46LaLm)<2-%er3kG*;|G5gt^dP@n z0I}*mcjUDNd?HOPPLwcZdgxVobQO@@p5fNQ_2W$yPC5}AWDCv>|T!7$0dnfX$X7r60 zxjy>=cfjb5&?PGqGC{Z`f|8jhugqz`|079t;>oMj!h^}B6k92hNkyT@pD4OF^$hS6 zAQ2BvMVBO~zX1L=F1sp$Ee_p$vXIT`v_pI)w8qQwELHwM(!kfLYP|Zr44Qfn@7g;jW=@S4xREk!Y2si@ zHNSxSmouQ2vxjTYoFFTvMF(aHj(TT{+GJ@{j|vJnUBGTp@zff#yuNBeUW!^Y}OcHZsf+~54wd!~6R z>Um4LoNOH_uIR)I>}reYX%D3U7KCCdDwEETd<}{|NOSKX3*1kabhFGpbT_7utBzC#_73xy|ufQ4KkE3jE=yN%Vw(2y=s0UuT%(< zE@(I^A+{YCFz7ojX?&wHD9;_~N18M3IeIK>=%536Nsrfi3@(8KeJ z=o|C9WyP?GUFdsE<&am7`QL8)cg@^~Bd#!i*CWKVoYy_M8|(emj@Da7(&2cEG4?ZkN#rd)s7jE#3HG&Q7nB-ho6y^!M;csWcdfPiH`71bzR1LkwQsu={;Zv5$aq)CrP3<33#*0AghHSz&sBwiHrxs^*tZ8+6H@ z(ZXNjnZdyxs%gAYr@||p!8H=3ZPhMyMP`O`ik9F(m-m+3e(SPbQZiUV>KU%duuFD; z8Z)H6bYF4ni*8!F0H};8yywa2)4Eb~v!Z z0NdCFGY&?W4=23&<2`R!W8(!0yfI8tyE&^QHQwm~gD#j9(-C^qb0;~XgLxWLB>MA;lxWM z2Ko&wDlk0|yqLarjqPynkYLQ!c()`k3gQIC9cu z;dr!Bit7})LPhOS0_9Gt^1(cC4)90$T*wE0OyjCIRgQuRgD%C~59g=Kk(MWyz7t$1 z$e}l@`XaQ)WSR6a*)dg(M4Q6Ngtq&@z~2{T< zmZwzJpkug(r^Pfjo$9*qh8&p)Tj>2>S&)-g>tC*htmGW}Fn_6_mlTRH^*one8Gg#B zZqg3?1BoKjqnN0veFkN+jlda;zww-BE%5T$3A>e&+@SNHWrmrHNPU3hvh#+WH}nWJ z{KLzt_ECyG6xm5d?RdQb{1Jl&*gI5fxFy02lE&RAIRmiBr2EAk(n=7@ZQ-vBKTnn} zNCP3WPT@V7PW4E#%+H|7RVPCLaOHxniW0h0cwPP}Tsff_3^r(vd8PO)3*2CL?m4@m z>@$4W=inFbzxz9DMKcB|Fuf*0v0~Z{O~#i?>4&mz`7)|)-k@)W;sCdie@;_Mr%mlp z@A7Yzw~(vyL;l<8J1@mWREF;IFXX297&HU$74McE@^5zQ_?KOdt9>5p_iUeU_32(% zz0J=1G74Y)#u{tFhZo!D?T{tAYMX(R5Z1sO3Dz@>zi$I@vIW;qqt5Xy4-R%je$mx} zE=iJK(iVm2RYf7`Gdt98V=is2f442de`dlySl|Jhlw>=RO5fFZOR|WyB`_ zdhM6)KR>0(i5EQ#DNSw47^qN0NR-b7zj)c~3y@b^B)~kTxLN!BAq;m-S?a%4-AKoV zU^Xs#%zcpXq)ilEQ&vKwPSwxro6LS18!h*_qWD8-?b@ z&B{&JenS!_ks^y#!4^u9L6Hqq)Gpt&8F$0-{X6OI!0V74wo!5$y!vF`lkhb6j43g& zR;=V3+;)er_Br9zF>V`jHh!_6zjn5KD|Wbt`8^zAH_myz(wp+1dCoKA9qN1+xGsrG=4X;i1PR_!HfUe+1)t^6a(?b|ORer{+fj@K+Q)4UX zy<+`*Q{K3#y`QJoWWBobm94Lw(=3zk6L*KDxHs`)eD=;sajzzNP4g>j7VZD>N8kKQ z>UaK<8nx%!32@(7rySPc{v77dJ{lvy*?~L3_m&%XD1hSU5<}vg7(?4EFqBFu)=?zM z9H%pF3?}%b%CS^*&;>1EX=umzGjiIFi!lx~%g0&aVxs)3Upe&roLT3!jgW*tJbb33 z6q_irk&4}KalvH$X5lSaqUZ##j@Lzc$XW=UZJbg(Yphnn z#zC?jG()z-#*W8$8#8ELYDsw*YBn4~*T`*TmlGR~(-vT;rW6$vDW{?;+^){apH?k9 zrE27~^G|^u@^*2LTrcfjcm_(5$~mQB)iQ{tLbM;RwHXtsS51bh4OOf!L||Q8B>Q|H z&8(-R1;^+Tp+R#~TEId2mqbo$a6P?~dzqxlv)%8isv!u7(^en1#6DpZ_HY@34fD@2 zBE|SPGh9Y;tg2t>aI;Qxa@sOwhQy`A0n0<$C0LE#AgdJOPN1!U)4SZp zotK{_HtaF%Hub>@8x#KRxBuN`t@z=z*kI7`IHX1vAf0MJcF@(xKlGknRi-KtmWCal z*-2j)7Yb6j6~JVGTZ!W|&)B-w{oG90G`3^Ei5)tAv!p-ji%wFPtHxAEQ$($HIz-ek zFFKKzOIMFie$ntTv(;t{e`Y+6*`eZK!<#YI$!jGRGk+GP*g}yEDyoHVas*dE1#tuYlwkXD@jLLS;7L`#s3;XD7(Ju8-?~EyXI6JxB+&!F& zfk)|`yzL&x{0*9Q1ka)a>%8{Ldc}Padq8j(TM{|J#L6Itya}U0{_wW| zj~(#}1Yf9aQ<)b!niD@iOtMF+^Ks(p(JBiQxSLXdO)R3KGW>dAFQ;~9vk+ImabB2T zpTRjCxeT^vkW@=+;#`U+Sjtlx*1|WDtK11LoAQbLob)c-Zn~pETF1Rd=Y`aGJ19lp z9>{P6HkfT9^&i{cwf0`wMy0S*n9IE<-{!kT zc~zb&#}wI0VUmwS+*Rh0wH-#xeK_LEOG$<>rP++=f3g2L&eUK5#fcr*k1Viqgi_Q}qzacq2NzX{dc&T4DMK*)7!uiFvk>Wv zbw2u_cW#P%r~G+0M2vK*tKui(O4U93EaXWVG(9i(OYzPp;DPj8B-r&mlmCTrs0rB_bx!Qewpc(_PbnHHa+->|`=vynM<3F2Flh841$d63B3 zNk5t1>wQVyF4JoUB0qie!&mFReNJ!*1}}DUV1ksSeP4o>1J- zT>WzA_u>|T#880gNT)hRR|etrYCi5OZ-ag0xPW`g7XE{oSpi7ol}jgzj(T4Z#sz%L z$IsQN@XDuNbv)Kq#1^~{GX|qRWczlo!aEi4o1eyr&9IXF&zJQiX%YdggJDaNOiBUu zz3EicCZAsScE75CgW|l=Ou4uaUB*gbD~(rxT$tM#%KAHH z`)@VgKZ=1E)#tE%h}}Rqt)`lHlCZ@(qv-`%7ct~ItyQFPYt+D(5Ais7UkOT1|Sf*_|nI|<6}5RVLjN| zfYIOof>@8D0L`1$8;4&g4p9arphE6peu3{^Zz#Vo(c=_wvUoKTOn+SI*(0wBU(3^e z?0yc^wa`B|5$Qv%`nd%?Acq(&7?442{2I4R`ntLhw9NWE*0|k=W7|k@Sb=I3R%1k0 zh1G}Ja5FQ$IUi0u;BB@i{fZMSiEgBgLnmG-AGR=lC6oe+PqtH0PsHbiX+GOX6UU(0 z2eME#p^cmls4UO%jSf67Df8dKE#+1P7&Lg_YEE=ur?j1}*yTAR{?;z}PVQqqv0`Fco3%<$c`@n2FuL{)8NvL|QGz;2fnmOOZ3kFMd*0 zzyXOlj}BlM-z0$aq3s?hf3s=oeNeMAF-aeC-4YHw(lLUiynLWu#NF9KK?~n>e~$XW zyk=pihqfcQQF_3?3v^X32phP)k`6dB-P{|t4EA))nMtK#SR@MFnFKYe=4%_JIlg%` zerT(*iIi$G{UF>LgsfZbMy(Co$93#OHR7wz)M)ddF%Q*9+cZw=`7ST?p06?Esy$>{ z201j6A#&c30RgSfVP4n?N^y)LN2sU=Gy8c+ca8gyCcQDhYCB|Wr}!i&TKLJ~hvWmY zwc&b@i_+rKxgYr3T7j$*HGHu`wK8ZmY)M047}R2dOsd0pgLWO+zo2jKr*lyivXaK0 z@fP4+Sr(WjDtN8`D_vhV$&hWHk}#)Hs)yS!N5Xcz4+G$MK!4cB*?h8xvvKF^PZ+dX!JbW8bzfzd3Ucu{>UkQ(u_IRRlaG&Q2!Fx%ZBjB~H96anb_*m6QTkaON=FFi!^6@qu2RX4m5fHCvj<`jMhDPQ38hZ81M= zqZC;bnCsb zs2H>4v+YyLGte{Ar7?nt88aiTz!t9P{I<2VaatTPSeCwsgEJak55S&e46kioJxHyd zhkc6pH;;dPixOLq09ikOqt`kz=-Vb=@ya98OkWq~(rxqXvtoJHfQ=F*&+5`H*Vv(? zrz_$+V@VZv-i1D4NEJW8JtxIYwX>dQ{gZj`x5~>Pm}JmkA>@OZpx&!IM(enVL9M|T zg!T0P1!q8{y3hTXEHPrOXm9ud(L;4I@2tAl(}4gRC9=#G$1Wfmv6Fvn_iGTC{g-!2 zC%;S9IC1Et(85>QOetV-n?^<5gyF3hYwI|v+}+C5$!{Zh7|OwHKM@f~=W@` z@t;j+s1B4zFNFXw2K3u#{roNdShAK3c@Yk@0>j2~6b50~(~di2g%3&+AWr+DlUSk} zST(;Mvb-DkiLe=1%-t{xw+b&v5;>^oadvtO6h-4EawF#)>`&vlP8~1kgu0P3QTz06 z?oMt#je=$y+&bw-n27SQqGWDX2=4Y|(gnT^L1hcNVG~0S;=?flow^V5$nAl$QDA5! z7uK@P4$xwUA^mH=|I$n5DaiLfD;ATDPCNzex0r&qQ;J-QY(?4CEZz<7QpnF~ryq)u zechB*uw(;gsUR0Ja!^Y(gY)E<9g(?o3s74kTXM4NCE+DVwpmgrsOR)bQYYI36AZ&Q zHmERkYKLZ4P?_-GwMT=kYwy`)fje<%j-gV&Q5wg~3rXf>Na6*BBINq$q(9)cag2J0 znDAA~ZIFKAUkqEHt@I-gttp=-QPgN`ycn>XJ|;fGyCjJgU@n-WA~Rz{m(l*%0ZUJJ z@kUqdE7t5UF9=>3cwtOt2jg-V7lHy6$qU)dIqIq{5^Pm8N{czWX2A^QiNtn!AvjNA z`d~54dA*Vr4~*{_OZXb46>g2vTtU0Mo?vw?#)Iu|n(TlX1kTfgKE{?Gz4IR`U-NYI z$DMUaWE;CA7bjjl*H}zmrIcb91!Y!IYatn2r&@edoaB`swpgdSA9#v?n|IW60|%Q1 zTpmbvhYz?^(;ML!TDSu)?Lf6Q;BrO07xoD*t4|0ATnzqrPcmn5re8ZBNguEiG&^<% zj-60rci^0_T=_NY`Xr}aC^1wVmC|QKoiBH&-`VAxrMe=5#nMoV<#Gw{sp8$E@4MT1 za7TV~M>FIj_hjEzR;~+GwPCB+x&(#IROG~4NDRd(sIyt9?)#GVL(&ENLtCep%LZNY z>GYYYa-C1Q7v}Jq(sq$SRa+I13Yjc9P>6*tvChX>p%PffE9UmewOJw9kWb$Yb%X~q z8h{yn7R$fvi!SKx^$Qg#|6-nr-nw;m6DfD%g%PNI4@;mtPATdsQb$GgarcS4==HAg z0d2AYm;O0T1ZR>yx`1<(Aa{GT;7~Ytrm}LM0hgT%(x&3{VbCP`+?d}5+I;&w-`95h zy3=o2;QQKTuXjR;f@yL=x$Y#NF8YA#Du30ScK(3NajC(5$F!s5(bN=3;X9_v;^m6& z%QyKXa<+-{B?Z2h62Cn z2#crw`%ur>v5+1*@%+R>M1CvjX3{8a@wmbPRsr|7 ziz&wg!#Gy#(ZOPf$8f7>fbTwQ>%s8anO30ty-@l1Ix&{-mwN8eH zT$P${RenJhje)Ree~%!6lE+gd4nn0ooo}LSg}QL6=A`RM*Zsw>vK^Cf)wcK}!OuSH$VC#)Bn2?Rjb1dbK7GHe!O)n32b{0gSc) z+wQ2f;elHZb!@QgKeW=g_qC)+peYuWx$O&B$;b7=zJNSIp9@O(Br9&HN+gN0wLzx^ z-2{0JqWPCZCJux;$ix+sN*=UaC4#1RGE=NX##&9MC^5KXN~5?l?2-t|o8e?8T`Ruo zcArVG7s;e8Z{5uEf8mY|uhfQ1SeX;&_4N5^=gKVQ?p_gdVnrOU%@8WlB=~jEpO@3E z5F5$nm&`gTLV<7WUmhX1)C4gW@U{Vz_PgZgYnPQ+247zhTv!R!F6QP-p?xP^RpMQ5 z?149?X^hTyvZ7kNRjf$?P6!0dmWj(oh>np3K%QAX`{BUqlaem3KW<45^@;#wyP22$ z4aZfwDhqmGVT$AAinHA`<;tz{iv%_Ba>dJKT34N~rcG8hVJZKnY{X0mTX0w%O*@~6 zAJy7y2q~26=ryDVN`2yl_aJ|Uq>pDL?L2TZ4Bp6f?~!`5Y+tPX zbEX522rI9{dDEt!w0O4s(zLeyr^~Kg)DZ(SX&M@c^>kQk(*CePiz{J#e zKLC##W2z~1*#sS=X?9F54T0GQcQDw4EB*|+6e|9)#6Y9v;^q#|a#5qWYLaG$r}6P6 z<7(lT$#mD?E#;y-f${P(w-Wiy8IMJVf|mlOQJw*PNq9+oXqJB_Cz|xQw}%voj2CA= zmu_bL-FP-H*>}t0D`Ne-QTN5zddr%X=bH?j_v-?5-2LM|-zXO4gIRd=7XN~_E_Jp|><1o;!T`)7ygNJ#{I%X#BJTR6vnW+0e@H#|yKU&wTGgCRSu4O?R1DE?WsAKpipl|iG`@DJe>W#1EhMuO+ zNY1E7!b!6cA?6d#2*BjMX+eMa*km^TtL`^b$XBn7*;p~Uzu9<9$sbbWKGtdHGP)3B zeRhV|nmLyvG+6rIC0#f3hP!63vSaF2`8LjV5;Lw>mI4iCKP|Jcx{nS4L$F% zBz+QOuL9;l4(k@ZN0knIsXcW1>?&0Y4nD-6ku6H38nT#Z&7BdVCuh{e+Xs z)5@iu6%#jsVkzDVv{2YUU@XGK`x1kDWJ!>L%%txV{G4Y`&gbCemZFOsalrA^znFGlpa93wkIpg8W0 zl~~Q6?b#zsR$Q97jIX)LT^m#wu$pN00-I$yjmMS*^||QivLIX!wN2`QmDFXwT`Ej; zB+3fBn%p$a(1C;}lf=i!a{oRTq>0sXb)m6g$to>_mET~fyzyLlP$p#NPkFb~8axpn zsteTyS20P@o89M<0Zb5yvZ~o}OcX>Cjn-}{d>Z~4@uWkr%(6a$1r8lzTAn@h83u97 zIu4uu#s7UvzGpVxlJjOwh!rO2E}4;m7ztRR6c45=m`0HG=gIVY)3w3-0+)b*MuDsu z#{T*nKt3!qPMbHU)=Gu{sls#>HZrXZN`YR2)s`SIfAvG%vj>!h`q_it7g9?s zyX>8I_OjA9U&XCbU|(z*cb5qHJaWYwy_drpy@oEEm=kuGE`vh;eY|qIK4hPug?BNq zKBSWF; zuyVR#p~Wy~3=b97?q-FS(GNa6B>suX{kib(x<{n`C3Al|&D@_hO5RM7OH||)uPq>g z6-TbY0t~$tBvwt})GKhl(OmJW6xS;NePtt%eIRM(^F%VkMSi8toAl3{(;=W>&q0b%UT zuA4m%5p<-mo&+C{v}cEualk{4a+5k~x53}(b3yNEJH6Bmc=9hoe>jrR4sPa#54~qT z3(GBQs52b54!wx@@7BvqFk2rVc%H0t;`V|fGe~Wt{1^+idyD^#q z-WO;oD3W%I^10paS7)9EksoAu*y2^i+?>A5^=@b1A*nMdMZnJ`hl8LUz%IWQz{q9Sn#hB^a{;v|WtlwQEU zCBZ|QPDzqj55X0vxiHe-L3uN9h~T0NEEHNBM^`gDE294U=UYQ&nK0sh?{Fg7#SSA* z9Be*ghLIXdevl$nRODGfJBlShA>~Hs^XVeM8T(~Zp?BiV)r9xrXQm*l;6Lad}swgmJ`+>sh zjJo=DWH2{=RwubLt&?8HtaC35Iwd?!7n8xKq`}R^(2-=r^fdTK4|=vkoB8=T6OtNV zKlceqV230p&Se#vAt{}br&1&tgVGoPdIb3_RPu`EKad)d-EFcW2=+VMjqU$z%r;CpI|MW-!@B z$-((4pdz(Qvuu|#E~r;p$Q-6CrG?7v;v2G_sijnU$igWJZ>9<>RjHgsk|=%*7pXT_ z%c~=>|1FCDfUJha{T0xVtfAL&v>;w`Nmk~z=FOd|QYy>)95+>%thge(_r`F{$@9ZG z9OPJ-nCBlpj3cw&+_L^<_j@tZjsMrggfH^h6n;V8B||MGl>dXs9Gxs{dQJL{h;SIx)f&ZV zoYGkiC3CX=7uxdTt>5jqGpt^`^M=w}BXm0gOp{K{+%=^n-HDgX`^~12B1#T~ym?gQ zV&IZG!bjn=)m&sC17>dw>_gdIkH=~**48EXAt_=jHfbixuqyD#%rZzxZ=C>Fp_Rb| z){&W}0~j4K-moY*R$eXs+O9gpw&2-fuCTsqM|iWs^Qcj;J*u&+Fm~Ehft9viEV$Y{ z?jSeKqbW!yJq>J$5E{c6ZydKvTBNw+-sh4Ax=XEIIM-a5?dWx|X)K3;4?Cj3`lHz4 zqvU5>k6E@SSgGsBF#AJkwh-k!6J_c2Wf9K!ZA_EfgRpugPS^_7k+Je*g?(K)>whSm zVK)YYKFY2q*^PnI>VJOM=e=($O>@qal)uN5Z3ATkoOn3{;gWt8oF$Z8N0FUWz8hjQzaki8k}tfRa@^>D(= zPVK!SY$wE5i-0*I%>z0Ljw_JJrJRo9BWF~f3tB_W;UI~JgX19mOcQG;^t;!Qo7SN) zKvaKpL_uZ16q8-~?DMvLWcw=1cJ$Hd>$_g|!cAP;7<*=ODQrv1q;vTD!^(no zgIqw7ufr-7ti5pChuYc`9CMtNO&R4Cq^-7W#CF<+j+NHxMlsGNm^MaPaLw)Ui6H!p zvQ~}ab-~3-%|2BpshF7UpR6bstnx!@qY?47Z3oCRBj6ZFm_L;jKt^ZRc>iKJX+kGX z?pR3^V*KPRiRPyWA$pSFcZZwDk9}vyJrZrWy>5-w(jLE{W#IG<;`hLjg=<0Xnr;yh4ykO{I5}b z$W<@K;-(QXDCRHPwqZ8EXC(I6nHQ1z^UoNQPy6wa`-jOwCyviT^0(i*cr7JgO_5kC z5{oF2lE}c@j*;1HVK+J9(cv)i{y~P(HdNSqNju0dWIDwP6{D8U^OT30Q1QcEIa;D) zx6*OqNvF;XLkB2%B}K}p$j^?+vBbk zV!`%E4mXRz9?|)S-PH5r3tmw#dy`P2?7`GJ50nq8ljuUxBl!s13N;5{UED0&KOtFx zNnvCj`^ZoMiwJ_;AkG$54ANj4Z~y2Jw~XtUv>+?23^*q^214wxvgozug)h4fTB>^| zojMUK@LPCYvO=aqwu{plmc?M_`w0TcV$fnO6+u&mqE?bEY*1J73cXqc#iFIE3bNg! zV&a9s{SKjY`Ns4Q2P)q@!~q+uph9iFd)~!TcJUQyRy}1lKUGn3P#4(?zR2t7bJlq; zQRgUIu)!>!(>wVh)9W8UE?Lno1-8RGJ~Iijaqcpc!y`exF1LbnkO z(q>|}{zvr;_U0oO-g!YSuUwGozGqUhB97S-wk;T@KmCMl!TOL`c@@(pdql7D%kvrD zv^N~YS+P}54-X56RLI#c|6uZ3uFqOEnOt{buVtwjoyS8;exD*;RAh&w$3Te+{TN&Q zN@tZwx5|%@)iP||XaSlqD0#%1h&oAiU~`ZrQMO58n7vS%Gn3AolE%sKDD+w!*yj(t z4VVo8Nf9p$uN_z50z6r9b)Ke3)XvKQzbM&H>x!Qi#qV)Xly#EL${wh^u7pnfZ2l=) z)8o+&Z6W1Ac%Q{&(g}W#J+BdPrmz<+-5>&bG4QDByda(nVm2rR*-g?__$wf(N1qGU zP*sXgOg+rmEMGAbe@Ez9nz)mrX(ksu406U%{0>2~U!K5exjb0euTcqif~Hv}kl5HS zCtgcg;R#tVP8+;}bWGpmcVX%^(F(GJ`^3MKUM0Kh{i$AB#qHq@&&aT~1T2ga+eZ)U z`Z51&vBa`wfekR6I8-0ULa*C&tlz?^8xG}M#u`6Q zJsRsjZ3oF9S1Rgu`G`&U5!?~{Cs{oPN}~G*zqFJbO8k?k$Tk^9-YZ;;;~PYvbWzTlB|InBxgaNZLCoXs)+ zHP!@?YyZ3Hf5>ttc8fF3u)C3xLk4s$6`A1IE3IM@-Ino>5e>2<7~RZ9@k)5{8$UZW>I_M8Vr=X-!$vkG2e715kxRzZ ztIwztW%s0AvJ8Hn;Ed~SSDmtxG|ReS#e#duill(gbH%I4d6;oJ=~Htqi~growinuw zH~F1YE@RHjI!7NP*X`Ub*l+yUf#vxrqu-gsUlSDo0EwMM_n)6+T1 zTyd9d-;73amgp!~g8>^&)Qo+KB40hN56A-U=N@-Wbbx+Bx+*@jmfjOG0(0ij@MKr$ z4EV77JGk&sbH2mM%JO370wvYCoB zPTHA}4Y?-S6TEOvY*<%lj|Xa{kdITo-p}pN9 zzA_fKv06?z?`mMjUzXhX!~_!UgPJkq7CRH;#4}KgS=Oi6zM6!|`Tc30>w<|v&tZIc3sT>Q_@>T^j|?NueKV&!_sM<%EeRE
    9hF{WbfLxumOd9f-6+021Nfuvzn&;NMwY1IqwuN}I%R6NJ7m5& ztNe15_$OdfC6d_pxirc;!}LIb3FSQ*Dm~p5ny#u2ZIk7X+v?}EsmB%n);Zr&71U`v z0j%hJCaXGWEWo?Ng*pY?K=rNtfku_|cG@vzQ&w#i3yW#>8ICx?%3_W_zUSumETeT+ zf?_H1W3!|9I|VJU#)u6+Hl-Qz{165c?}X0Kr{}cXc4-Zr>fR;F1y7)x)X+I^;#MiW z%(d9R2L8M8qVoc2Z0OO7$M#K0_R%7D4dHX5H>Juz05}ekESgA2UpQ=1)$7IlD zAd7-^p|@qAy(c-YNE)|S6~~OwLirieY)9;~dq!f56=FwarTnOTIOMm^3wdIN{C11y z1M+n!);}f6IwS`;7nugg7aRrG<3X4XR2eI{r@VobgIV+D7FCDjlz$~%sp_)NLg)p< zZ#$y;nZOr3(;la{&_8N<*=fPbWh++7k) z1{A1mRy-Er>Mk~HabOL}Wi%)kgRwij2_JG2h}yAHsA8rD<1qHkHVz30{V-v+X%7UL zp25nrjQ(P5ZHbM`Yb-lF3<@o40%C*PWVnEaMS*0e8xr5tD~u{ESa)9!wfI=Ka6(Zj z)`qSPbD%@=^tBCbES}!&xSy=X!g-h1sk2(L$D5YdALjh__hj=cv%H21r+#f`xs*JM zBHO4)qb2E*Wzu^Bu-)vjHwb*9R57y6rK|P?EOXPec|`}bfXH47{fJ(`9ED2X0y=SQ z7pKS{Pw)j^hVPvI=zENE2pGlwgJO!)> zkKJ>8GC}&Vob&&i==!3MFfK>-;x}>lAB9!K_O7|2p^jnNj?9$zpkfy2enwU#Hljtd&;zqD;(c zat{37TLDYlYv^-igE%7quSyEY7^p`XRY(u8+J$D>v9En zR`6%k)F=MNBb7kUcxvbIZHerOLYul=zm+Fz($aRf?zEMIzLMHSTTl>jZ+3nqZhR{tq9ML?|oLI&bu7TvKglh_gX20L3IaQ@vI$NI3-UI z9grwnF|*FV^s+4^DPOp5D!2CT$*egXX)s2C*yjos|pddUyyprw~XA9R0*H1pITui{p zl;AD-;YVdxL|M$djT4X3^RO(t#O(pxnW4fw6i)on{7a(i-uQ%lf~0_b5$!zW9L1}W z6?peBhn4MMS$dJy_lGx(tOm<@7bidUeIkC@d11BzHB>n#Uka+w5Kh^nD(7iX;T3zL zD#bmbCO44(9+?VE6Qo#Sc6cykbvG-_j{4#IemU+Yx1{vl4^ESnPVAOISgxO2vWb#! zpkNdtFACB_8pUYtvW5GVTISYtjQ@hh+T9Q>*9lO`HO>t%7RriCO2LQSUqz>}z zn`CXEjHc(cLv1fWaF?=$*F&ezu2LB_>G8Qxb&E13dqT=4!0(0XSXei4Yx-Fql+;4o zj9wD{F3YT1V@Ckz8H8-uvVt9U^H*O$!d=Vt&?}Ogw8E;nfhmOoYb1(@=4*w}rPL-v zF3ko-fmf3NE3B~baU_d-*#XJQs5{_rJo6)kdO;6K1F~yl}F>x2Sb3WHfEAMS}XEb>Ve9oiA!K z)C$r#&1BcC`5CG$UOE1cWk~R!ta$RfjiiNFz%6PCNrJg@&Ys{5Yo9<%Yb-%ecSF?~)hfltcB>^^k|ytXS+* zni6Q5R31df!D5^c+1O<3sL(beOTZ&yV*L@<34JL zv|QQ0BA$_z0L4m=YE_Jf}7*jPATPRB9mL@efJfFoFr?HhtT!9G=jA2=_05nl?8 z2V&*gU=7OPMf0P4I$#-XeJyIYKij6U7{DR4+b>|{L9E8Yc?-vjrxf4zG}(YW7qWK%dH>aaV?&wAMP{Ho#6xFnMGS65Ngg9;J&kE(oDo!5V?PmNJt0)@O z4@f=KQRoDCyg^;%y@q6lc6b+rRg!x3dItCF6u4gm%j7faI4JZjRc#V>&MuUl_TMim z^6U;tmR*(=L-%hdw{!McL8}1DHwBqi<8xLl$smV^j)ws(q8&lP%J@-Laf?2en{15x zqx%=g`Z2)4(*MhuNy(vGH5Fvd!q&X04ZH;N&DPhde6@jPZU?#eGf`F}?4)mvkK)I0 zdw_0#jmHyigE~{Voz_cB{dRGB=v>bF$@c}19Nb|zi~$*Bn1_Fg-GDf)ihWdU+sBs7 z%}#4)tZ*?ypdYkh!BNTpwSdD4Ew@gRH12XxGNh3(#M>-Ohki zFSOfzm0yuyxo75-de3+v2+B5-V*k^04QQAnzxNVoJPjH)0v(4v z7RAE=_JF3waB%pz9>ogSqkk}S?+-_tEX|^&Q>#gz3rzX_EzK%Q9z&7kROCV?ajKD+ zC5_YLQRUe-+%mZ#{;@W^_#r-dc(<{8@lI<|f+`NASZaN;iJ3TYD9(y9=p|7>Se@cB zug$$LFbi7nvzTg50a%j5BU*rG1fLDyMnf??qz>Wk6+e47(2~dFr3%AZ;r2*)v(>8$ z2%B1Xr@U)7X~C=^NuVvQc;g;%tXETYwJl`n4Yk6JGYJZLA7SRy6(F zl@IQ{wd~s^5yur9Rl7Ld;y#zC8KCbOQsvt#O`(ws(2)krd>m{aZ-XT>zuAF%tjyWy zd5Za$Ed#DjyTr8;LPfqFAih&W+1q3HtrPZ0HQnxYLC}u$MmF%w^tm8^Bo-}T{-%QI z@X*A(r70!_Obcj?=gk>rg~AI)k|j#aCcPItdRV8~p``MVYw8%&LRPK+L_-$5GGTdO z0`BKd#Z&S)imb%;RV*876hDw|QWQ?zB`OwSU-QUWmuG`(6M~*yK4M$jzxCnvw@jE= zy!QJ)Ar(%X;%zb;LKP!95<@8u6kMZ*6d*><-<5#j_1i^VZ4J9Hz^IkfnAf zvvzEg8CC=-c6DtKJ`J4w3SdEo09S>YDVjyW(tP(3RVf{HK7Lq*1sYaz!7ngl&) ziy$F=(WH+I4lgJcbpf+KYKnj@;>I!ciewe?s5V1TAKhTh1L*_c-7_ST*Lq%t@Vqik z_>^L*gZ{|Qd1IEC8O-zy2O`hidv)iD*f0}j)@Mq!q{NA5sCqLL9HQh0C{js9w(zQ$ zo6|FBWHP-uy-glXQt6$L`pxH5(gj{y#dy4uUd%r!Dr8RaGu^6~ym44Fy>v{T^t5LS zi5j2HS;QoBmW{!?Q5r2#hDWP8x4lq7EsZ=N8v>B34mkl@cr7HG6BiyY+y>>o`*|0c zVOkzLLc;hSK2=0FpW3Ofp-NhUkvzwbRSVno>Q@25xw69 zy4_Fk>S3AoHGP8YCskZ6KXsUD({{y?xy2e5kTi>S7QbV{QA8=Pom9OtIJ#_xqdH1{ zk|M{b$ZY=(?>6Zt=aezr0^t)$n{*xL3HMLNqlvOF>2;henC2GCuTC>kB_;;f@^thX zk{kN9k__z&SfFZ9-*NAxb0>FrXDX|_^`u?cpzaGv@$U#& z=hNZ+X%g8;)=aIP^|{gTp)LID{IfqAn&omJrpyt>GIOch^w~8_4V>q*j!s-rX+`O# zoz~N*LAMDP_ieIAP)pnib|X=CAKJ=Rfxyj^=>@R7=hQ3e4HYnj0UH?bloMr*;^mSZ zoDLwN>T|g!$zqJb0Hgy?RzZ&&uL8PnyU8(fHmDff@Q z+>fkTnpuOxlpGwC{Z!;S_acaqR6?U-ZfFmXn;w-k0I_7EtVr~PTjrJ~ zDxiy*0{WVyg$HfTYP18rE;_oCTSH$_-i4}zL>X2cV38i~UkcOnt_Q~j;Zu4zJK?Ib z3B4{|yk3_Wpu9XN`qH;Sy@71^xC~Fj?|!o4v-#0foq8A^i(^M;=yMA@N7!lWYW}ut z`8>-+Ih*9C6MN`Z^8E+7i-1%=jw$x&L;&iAi=w;kj{6J?7eSiin35ETbf=qI8( z@@|L_WvPuj)9s5zm+~oJ8zIRR3&f8As|Frd;yFae-MtAwy({Tzm zo_R7{gTe+HgLxg#pBo--*iDl;AOCdPYfl}UKbFlS8`!PZoY=yYnE7YhDR~A(wo;Kf z%Dj*)k*3uvlg^;;ll7CMffgX2qh$=Ws+itE8-wrNM3WNGLF4?94Q<{ihNZ!! z${kU=q))jcBw<#0P;T%JSe(>J5@jfAqUDy(ieoU1aBn)2qTzFZp@D}a4jPoN*AY#J zWVvMjq$Yt;)m~E_m`f@;t%$Th@=c7@ghmFhOS0vVxdr8)bF?odPr%d_+2Xj|iK%gG- zK{f)!I7TB0lR;4N+$mP(Y7|2iOd%#v6}o6okfmc@szxmWA(MUuf9olE0!8Af$i(0+ zUahj-;mb&^cz1XYr-uG)e$%{Omo(1P{lCuN{aaQ8{@j~pbATmU*gSez*IGx*th7H* zuneYHiCLV4Z&C-cE@KHdmS_WsHvNETo@o;SAy3$*a66Yu4)TYyeBC3~rEL9m5AdyX z-pJ+e4(Q*tY=dL7W_IEXp%tl!c#0q`-RW5*9GiM+QPV5hy6|d#>e>wSCOBn66a+z9qw;;6uug)Eg`H%dN@H&X4+=as@E-I*c0Sp&uCK(BQ;dxt;>5ug zD-2{1r{h;Z=oZu=1Jh|uIpl&GV9|_%B3dRf7#;A3-hD31B`qF};xuKuH>{ka_&x5% z>r?0w(&Dj7qA8rU-5;m0L|I%AJ}6mn(x(L$*ax}b#Yc^W3d$`WzzYj!Sgwi(9$+^3 zv2GhJ&kX!oY>uw%d;NymG#lxI$0w1!P8?D>Z#GRHqvRlqREVW2dUdjQZh%&v z5)j9r*K=M_M_*OnH-a;Js}3`qo1(UGd}bYEcow9PI{Kuks#Gpx|4~Omj~5J4sbBrGxW*A>N^|sgv|pz=u_-}lAYl?3RfDQY@Yt)C|etgGr zvT@pAoR#S#FXVn`2Dlu^OaM&C*xNv?fyeP#n>~+_66sbsJY&Nw zy+;Ra{MRlkPk3uX9HRlgp5Gi|H5S$`W;aGo+buRWM&4DiJ3#Yk>GgtV#yz7yYkIr?o_gTB?{^IbhhON(r((V?h)IfwaVs2J=a4mhW)zW6=LWMZCp zJUI#d$YitGcQcNXucXK_DiVu(jKx0K{GB$UjS~I-@)y6!eix76<|{JiRQ_;pl+px< zDJg%CC)>t=VeVfDR6@yh6xj)(2Lj7 zI$4hWi-FV6ybii)`Q+j4vjUFut}_4oS7G^S^>T4gEUD2cjz<^&55Z?9J1aW1WvYNc#wxU;FNU?|MhzDI6sp;$?f||3DL6i zVVyW4TW$uJos@hBMY4gDjD~(uX}0eHP8*XSj&#n^e8Zv)NnU^!gi{vM6x12EOQP8p zjO4vo(_I#_$lpN8QcT(ym5z^(vsl#|Pz50ct&Tt8pB~g*Zq6X=n*Y^eF32x5N zGy+-p4(K0Blxgrtbbx-sF;W5HV9jbuE9{`Rx@zL79(UaM$lmO|X5#=mHbP{esmtyS~il9%kLqe5tDZ75^R;ZMj&ovi>+lXZvjoY%%2Tw{bQwRZx#x z=mi|?<5$dbFbJM|`!;{+p1;$vM_9qadDqaNj{W0LEd`_5_#95WSh6B7jX}jm@g@IT z-mv^Z7k59Y3@l&{xoS|+b%nHzNdqr;7bqxU+0-|^u<0sncrE9|lj_h-(41Judl05s zz|Ev%L0bv5mUz|TJpuPa4a=?)?;iIkej6mG?PN8FZW{-hjEu+Z@%8vcr|Pj~EZ1rC z(Srn;Z+fEEFZiGa$zBi|(Tr?KHjIVMLu`+6WAO}Dc5ZZgS=C(20tGha#CZ)Ru~N=( zh+7Ilx@z&2;1(LIY}*v)NP_zz6366nIwA4%h_k`F%KL%Oaqlv>bw1a(DHi88*hI;Am|1|gK<4r4?4|9I|d$QSygDrc^)+o7@99jyuQIT~L)H%vgmMSU( zGzL8qc{ISFG$cljP3>!fz?QU6k7MFJ%KZ{$$j?*j3Qe_=k|}*ISOA0WHdg;@5@mmz zzvQiZ5svVyh6I5AV8HF6j`AS)u`(&6{@15p6i+b0MHRcef^1>80CwUm@?JA=?V#kq z@sx?$TOCk=fTA@9Y5F=z>+~9W#aIZ$O)2)*i!%UC56E>?rv>f|1RqiXb?cw2^0#VU6H&cTSsA|wC=jTdi&@3}Y-ABN9Llsz_` z)ZoKPyvqfk#fdb+!;t-EIS7^xqz(H+8z)wl#(8J;Iaz0SStfX#Hilv)Ws~aO1zZ|Q z@>8lZx7%I`enq|*2fIkxAW_>c?UL?PCj0fs;sSNj9rAttZLlD^Pe#Ow3iuGr#qkLGW;EGB>6_aezOm5zP>9e3j#S>d zW)kPb(OBpr?H2$^rsSI_vVn@!gLj=G?2=(0kEYJEO>xt+UAo#8Vz8L(%VJX9_e`px z57QddhFi)l7wiuliIHbFCzCm?RSI zSLL1M2RR67$<({ zhkL&D(1~4N=%?*xkuoSb)B>kKwgMzO^z=g1+5IuXeFD_Qp7;K9v}qDrv~+4U>2n$2 zGUv@&5$M(FXFyg_@)(LNry^0`2hE2D%ivO=LB~!N&}_%W7L>4krai8=j*{qKKIZ*n zv8CW*2N&0%4b=smokap9-pS^VSU9Esc-n#&^dS1*GZNcT@q4~+If1Z=zB#e8YDE|I zKDjLVd#mhk_khmvKjG%3%xhk@80_>wda*I3)y5?1oq=AwV}I4|Kv#MW7?1Rdw=@ zG19_%R?uSm4zv7%6D5?zk8!6F90$Fy)42Yq`_9B363dG0R)8NCs+7 znDKf%kg^dXVbWv;Y9@@_5F5wLW-BmaLu~rJ*sMH8C0^2f|G!Lg%ChBgvq;AnveJyZ z=m{n7p~wR&vRtAUYlHRlUKLURbpd}+uXMMop3a++0YRpGav69vp%)4pT_86idNLho z(MTn)&!u&G>+}Xig97F5zZNwIou<1_4(O{V@1Yl$2Q{dAXPouX$u_N~eU~?rpMDLXGiH!lyUL4=Lk> zKz3=fnHhLhF?e5+J}c>FaL*u<@J@z6W| zdqSag#nOt!fwv^+$U{O4>xawaP`?3OABtFcqU;DH)ZyMb3F?AIP0beJW!<4y-Is^! zLXpfln~!Ws_^FxDiqkFLjgQIZYeUzD!3R0*H>i4A%neIMD5v$vKp?bVCSfZjPoaPi zIubbyHDzuXB-Vw-xK+}pgDUByNy&;n7i5V;>5o1aV?{X5_qF1yZb)c25;2J(8cv6M zYpe|C=!V1Gws9u=)cE?jPe_6j*JKr%fh3)h0|R?96}epYm8=qcz78l;sG=WFN%QX_ zhm|SxJ!uylxgts$R|q^Do1jt#+o0ow>wLQ1w*mEKk0e*@aIb!dFmd>ro)t`}Psjc1 zpDkB+Ru)m0!n#Q(Hv=>ax6@leXHW}eBk930u2oEx!r^b*^4n$W+qb;W(fjP)zSDZw zCxd$Ky=Sr>|9K|q2jsvQ(ror6{gRT`QsjhD^(k4gKcY^Ztau!#V|qB%q5GAYZfPPY zAL3#o>s~mAOS?U)ec;=60wZxHbAl;zD}{jbh1um$PI+}^hitXyp^!M{JX1q<%)I_a zncF@0OZ??fsFMBD(zhOdaOk(kBD%hRJ)-0Nil3}{ukzRR>SxEXjwM-v_vZ<=ajugy zpeA)1`kUE<(pFRr(>Z1h!dC=Ut+J2DjWMB0t^Y(r7K|Annc>9IX7FA5IWzH;JdPqO zsmOGgb70}WGdPn*t#$y$v!YfOGWQVwvDGyre zc}027dkBQNHlk;k)(>uH2{`NM8NhsuD&H`7ttA1B(=Ly!Fw?X~e3}IBY%y6lHHF@(%mLNJE23S~4D)ToHBvV5?i(XuaE68vmIFS7%UNNB z68x#5sm}z%A2H$tXM?nI#-y9V~$(K`P2?XH!BCitBebbyD^Bm!5(n5=7| zQSt6^_)ef7$p#76MDFeJBV<8rghxN8gw-ey`QnA67aegi!6WzK?>{F|uM9ju3BDgZ zR#Wm=imad_^Fs7I5N*JiEy~Ymj*%T9D@TOv)?$!sgM`IRBe45t-_6JWv8It*5tn`t z^Zu=ftnb}?XYY^G-uo!=eSO5jw<_n3ka6fYsJ4NE?b4&)2MlrZ!pCdgec9Eg9=(5A zH1F~|mm#EgS=Pea6QUK}m0e)6JPVnFoFZv%DDLeW-%7Xh+Id%JLW5zjaj%A6<9SBX zE5!&tmiji!qWGt2y)=(=V|t+yV*BC-_0cfg)=T5LcpPdxfSzz6lN)?v`fYMX@`QVu zPIccUY!c;$HmIZc^`csc@#9kua!~BCR&q*NFV#5aS&o))(M+nqqb;;W4E(ptwL>AdJTd*DOAx<)^Z)0UWMipHD zWUFO6%1hM{ZKXw}B5=0|73#3;0QF!>p)3@t3Ay;oVlsgy2~qrN-Vk~#!gY%$c&cJ$k14Ure#U>!qi<|jV36n%DDV`4>(dcint zu(4r2oBw2b4*%kK18w0(RUA{sjhzN{NYJ0eIWItJp?#j)f}xDm6YE=(7%TzIfTQ)G zYX>}a!B8crFjeRoiw?2E<&K+uX{GD~5))otIP|t(U>zg_#O_D`>W_acwp>9uEpV*J zUN(xCOX6UB@3=vIXR@kIRyH9;SQAh$`u}U9r?6oMifi-w3g(z1Ei>kS*+RVO2XJg9}&NyopFtVrWLAU)wa!6h%Gp{k)1srhNt{LR$- zHPrl7)cj3UgSt9!LqI1<433fSgu;xX?;KT~6aj^>EQ?8}bb2qn2)qfjN5ANTGvvlizJXI9=E19fmULr z^prO?j5nye#g+8(@Kt_V*QOvyL-MKvp%9Ka<=vn@N$>OQ6O<{>1|5=u4w~wW_sQ3n z$xG=triR`vI!Ri+^6AqOJANJP@Ez#)wRnc%;ci7vtA04$1fEpC^mQcfl>raXd-p45 zDW~L+3)oFXR*F;TD&LFHhLS>Wr+2uW;Atlwa^I+`m7Jb%N!0m9vSNk$BGbY<e!1K95CdkY2=Rw(t^z@g8h`S;0IO z;W;~j$i4tVEDaYr#``~i?-Q+M(e*2`GO|*lT`P$4)v1g98^Gt)0i94ciIv9%75mrF z^KMQr0$tI^Q|4iX_Wc>v-1Lxh+_ht`POs$^N$muLL56LJ7#ZY=_BoYWIBV?`lacxC z^R|6t`#|kK&O5^r(0uG?N_JE7Vu}<}k*!_@!UT8YzesbJC_4pWvKe&x>{5?9=x#$| zwIWeHl*QBtJLyv8lW9-5_3Ek*mc76CU-S`Y)H_6}!ZmM3`{{X4xFun!!gZWVQm<}) z`%c8C?ce$I(t9ayubg{EZAaV;5g`3$p8lKm0S7+Cx+k{c1DU&*2|D$^F8PRTd}Zb* z@MZdeDvOeDqevRePtyEZ#Q>LFV5$1zp7fzKgGQ#z6k)lb*F`U?^gqlyL=O2L@~V_B zp0a!57G)aep0v(&doUJSl?(1pI4R1Yw+Gu1HJ0N%JQP_z&3^k^SARX~9TQd}N_p+1 z%8A1;m(51Mj*_3G$T2GNQdpZp_ts_6VG!87FngD9JDnzKV|EF*D3iur2Isbld(x*t zT?%z-7+PuJHOj7u>Y!Hp=JZYB88r65#|dKr3dro8#wnY)okrtwMv~`;vT408Mbd-b zy)Fkiw`t>ke9%FTaqD7WCuovnF^5P#M+;KH2`ZhkM0uOu&rKPUH1P0H(%;SuvONUw zShqC;YZ?ZzmKHhhaoM1O&IWa%bO*5O-M9D8uppPAi{49;)l^zDdt2#-exWj+T1Vgt# zqAW=aB`{g4y^1^m(gK0%C&~veBTM`=*MsYT-wJz9P$uaZc+ts`9d2=+GwICnizIcD6#BTLg@^P&x{xa0 zE23Or9>+5osx4l15@dSYuPmpv!KZ8>o3;bK1~FV#PrVRe@rk=MFFOfP4u4ZP^1xYMtK>x zEKH|58n}DbikaI1t6K5WfClwOMn9oKoX@!+UpgMEqZ0h$7{{X)F*F3^+6KL0Ud0Nz zl-rZP{q|+gYa;_}d}u8c!Nf67rWY{T{96(OYhW`mmhE%pfOZTNvTA4}%|;tgb>PF26-WY_&4)JXqvL>-R${*b@F8HwsyAUcF-*)ny`2?y zM!h-blgt-)4mh#13)zBx;h7zj9D1iSsmLZlm2ctH+~B)2s<30A98yNfirC;=ueB*k zLt4G=&L|5x%GLDx<8HbNK!l~h*h7u;?0(+U>x>s<^XnbMpk@yYJso#@{u*fUro~&LCz8uf9K3mAhK;W&`BxOVPDL((q!9i;QT8Y-&Euilb^0K;$~S{-h17hq z;sD(%P4g&LZkV#4SFTKwqZ(tkpiQQKV_yJN@lLtOG{}u-@ysqx59fsU`pJit7)Q7#6xc?~Z4&4YxVm;%z*^^g6sc#m8 z+iP6YJkC$KJwAK#olu<=eOkDW-}??U_h(PuPG^vMZ*26>o;>huX%BP_ zZn9o@c0f>CN7q21oz{EdV_s~boY#E=3D5y8)G1DzB107eYHc^%&&1j_` zUPbfUr^hkNB?r0K3ujzUNAdTD;o1wESCDHLCOf?}cd|yK$)0@ujZ8X!YU0>|VKCqr z*m}@D7_#-#Y`@MrhL&Gj`5VjXai?{ctuQ&Xhd?d(2Jy}5hH|Y&G0J{s|39F9vC%=c zec0@ErS;(-z=VuGv#7KDpG=e1^%+%PlOwN8tgqe7xSXftXDRX}6}g_;`@xa7E8Z>s zK}kgHoG(9Ydhb@mwfC-m_u9Lw-d*?4rEgWvxfPK>#eYyZch$Qmet6(#nw{Uj@z$F6 z@6Sn}Q}*HAxlRAJ@)wuhZ=JJZZuDF8Z$+g2^2om)`Q7!1vUjdU+=|eBZ#gu)-2QP% z#Kt+RC@uByL)^dg1I@X2x+1Q>T@tb7+v|UREu!I_tN(m0qV>J?zgj-`>vs~UGyk>` zK6ER!@Vz^8^Zw=FKOOqr{Pk4P2e#TQ`xv4ZPFT+_uo0@6f4TKP3KO=}$9}nwq_V@7 z6Q|QF%&=8J$@3_ZLq+Q7?$F!P9PzTi9`~g_oBTSVlJT7Qlu$=^((~52SCRzIMo2@% zjJr!$&Q20SeHX#g>wI!Ky)M!GV$v;YlT|`pSj;#`TFlmkCJFQC3xT!*!NzgH%j|<0 z79VKirHk+UYT|ekWt_XdtpVqdr6ykGbGc1NUJw z6x7?icPygf-6g+X6KTBa)Z2K~Kaka6TYDi`S!^~E{ z(pgtR4=bwZZh>Z>s*|J$Q-l?QGPe%F)tM=@KBOLaMjT*Ko*yaO0P@Tk#jWR`V*ev0 zy<_E;3CULkAXf6KyEtD9Fdn=vxLB#dENZQEZ2u#a0uRcB%4k`R_ADgZ|BPY!*s=qC&3{>gZL z5%;9>lyb?8y$&jzvcvMlY}D+x!_E{aFg8{jQ_w)oii9&G&b+4wa3UB>R^J8h3@%L%t+%Yx09 zim}W}S!SIi7g{x;^DUs%vrX~PyA1{FT6mgt6VK2}k8>d6o&n`?brPIjAZsK%>!X)~ z3Lf6rplA~44L9p~hMTcf*S()Xn>kl}*Fn|ArCTwm$p=#r4w_7kn(6@~!l~m;GkR+$$d}nS1`d zC37!CINEcw_O@R5+O6H_;N$FGyVFi8H+!nDSeghns~0CObFq>rJiyUIJ?K6+Jx_E0 zb^U|~q|+;k|NjErFGM!Z9S%nEK#$M=?acjRllLjOBlu6U`jrW_LpG+L6RoA>TPTu5 zMb=4rJYvIQ!`f*qWy+)visGTYBMDkPqTH}G5t$;|Q!2g`8!lp_d z)`Nm6&AoqO)N=X}50G+U%y5qkuWBP;0^QHHE6&=8%aDO9#9cgwbE zcG1^WU(7C$_R^{T-K0ug0|`)Vq-;Dk<8ahQ1pzBaIC$(W-SL|LsLa=9zhf(t=(HIo z9FmCWeD-?W3@Z~tRJ2mq0^Vq@c!GkT)?;WNCaj;Ch%;W8nEcH@q%C~e_vr3HffhtP zFG-dMma*|r##7*R&CL`OO>^&5Wzr`jtEOI+^td;x3`_xIq%V^m_d^TTOE>u&LpxPP z%Fc+=c?DjOOIt386_-sn3&fiR=N&!2*asRb%Q5_kgOS7!G?fdZJm0X)JbwRH_95BI z&pB~oe|5jrJhh8rGbyqY)3Skk1@`jII#x=v4yOTlbUPxqPf3=Wz?P^y4~0!2ZiW8K zHMbH-dRzuO#(39b0*Q{oYZ6Dp=}Q!h5m;Y#TTD4nYh&Vk1;OC!fXh_^<>_>+xI5y=oJ8P4eZuS+$9ZMw_>KZk zLmwZvW4F1p4+mNhQfMq#Lb9EBAzW)U<_9RYm?8yK40;NcbZKM_RNd%&w|VdPK0!KU zgOmh#^?}G}(&5Ids6Uh|-S4gQMbX4U|MEyw=tk9$LmJ3Ybi;ibB+%I5Wu_}Qq}d(r zSQ|46EO40_hvI^lZJA{J>Zg-z+2uHB*;4-DhU`x!$|N9$48BSVymrdtJ<|lJN2d?Y zi)@TGM5mK1!RD!WrWYQWb0NB$N%VE>Yai}ijdCc4>&l^RyoSPg)5)9fmK^o8SeGAE z?YKlX@tc2~_g@BfXJ{lgonj$FaR-t_Aw7?tL9&0na9=>VC|=kluNN7jD@gK`v&2Y0 z6>Sz2&d;Bp?WTACG_=gEMsP`jM}NCuzjyT$+SW^Xsn$6hz6j1FS;6O|Sz$*+ZIJAa zCCORf)ne>qm75lYVUko^0}8wzL*je6q(jwb2ltHw%*<$vo&!GignpO!#}TW=!igbt2m6&fS&X2JFTu{RBV;F9hJAKC8)N@&dBe-Th0_5Qz4RVkZQ+oK$EZA0TiWjNul;f8DD}v{h4mMT`iC_UaMQ4cQJLDbS1S zl-;!X7$fvW5gn0@>H`WOz_=z!^qt5pAf8YfAKVPn>-Cu6EZZ5zX=7^~b4C-;r>6#^ zpOG5eMr(_p@~+pT-gTUe!Z0uKS(AOE{OtA#JK$w7CiUKyHoR$By_5$1aV5!nWmYdI ztX40FDfS>mN~oBNQ@6;E(|s`4be;Fu|6w>H)Pi&ci;LjS+jX6u%{>bB1NrH z@rlby5kty3bWcB%A%S8I5@@SGGCAbQ|tdZ(7fVpP!h}R_kQY~35;nE-7_?eQ;qZklH<2H z`t)>diwHMsQ8c-m&IXZnr1h)@8vGntU#L!lV%rhs$W}X@EA8+~^)Zf_vC-jg|Ji7a zZ5a4{)15ZE@jCPMZIKobZOK;Z$^Ms&p1;uw@y96kC`Br%m`@knkQt&|M4f_m(GIUV zCCpU7+t%XI0B)`!x+$`VUc(?oAW#DW(?GqbUeYXS7yWf9^>M6tbI|R%394TD8gOgh z@@fpPqzgd?4Hio?+M}CQ2Le#(qcQx51j8`bBL|<)i##y9O>U&G1~!Ht6VwRCG%JoD z>c{3PbE7`4Db{nc1xU+Q2G^1S7w(&f@t&PH?Eq3BhvYw zv*5yy>K1Gm+nA5m9@&GH(e^!-=PrM&`K4{t&1pSe&Ins7{R!05>?W-Q)1sT?7pE4f zN&=PzL%C*0oXdPNq@94n}co-60}gf@p}(;cK@h86=aPr`M++6e)T)xg?`AggC4;^fH(Ro%A5 z;v7}hdj*yB25NVBNnml{GSDh`2$h{W%|l_i7UPYNeR`v_xDGnH_{_30v85#r^0I zZ~g*|C+>d|d?VL3hnmkE9K);i!|1BfC?}0-DJ5I31E%X>z)#3Am=6Nf^7PvM!Q;ZG30@3 zlByp#JB{Bz^bc3Q6&qY5#BLE%m7?SHz2JIf2`TnL94=KI3fHzn+HNJiBCJ!jo^6D! zhb-{(vZUE`GckQ?HY7PTdfrp*h|GdM`jrP=tC%MF5(P*S<;~d-tFn4!5oG>1lkA`> zraknZM}w##re@p@d3k1&j6eG%G0cKL<$q~+kiDq1i4kg_G4ARL16#PlGO{J!UTsudm;{jDA``Y9tobEQC|q@cR3Js z;+_16v$AC%Y8MBL>ZmPbh^|xXp+c!xlBsz@V?R1|^~}o>WD|7EbTGIMRc0qAI8N zPFW>R5cImED}y=PSh;@7D^Gm)t=2il&?!jsRl1;F)Es)qzgLy&zayYB zaAnv-8h$Dd`xQkr$sa3$$1jy`RHJ0!V`Uq#GT|Sr5W%$=hKivU$qg#g@??htQvHok zA$^Y4?p4MMb!w~hKw*otLs;v*Vot$=JLIwi!j}M~ZR#rdR(U4qm+6`A zw>oB~2`V9*p;8n(=aC{!+7qEuBiC`82r~Fo`Qx+Qp>0MzKJTet?Q;R9_;Hl+_Hj0y*H&0yTnYFH()(2`?qtOqZqrtLnz@__5JqU1C(re(+5^1NvBE;FCKx1T9RvmDG1(rHpfD5}^Sa*FP zN2BPd^}i#~W@DSifUz$h-E&O&JCy}LTM_~)6OW z=Y08MM|(6Tpp5oI?RYLq^wY2i3s8P=AWu(i z+0}xg<8N2~g)DPo4+oahLkvhF#U@ZBj*7VgMQlOx`n@#mjfVBw$E+)WHl42y0nSlpr5Hj1=RF|p!` zH;x1vY14dB=zm(>>rpa4j?pGUL|6~SYL&1Y?+bktnxtCk)+}iV(QbEr5{d-spm69_ zJu}@SN!6@EuNTS;gnOnyIHnn#l}Z}3_gYmLVa|b|AV^e)HHT>Xl}m%SFprelFO;~# zx;5?MR9GanL6E0K^yHmpax=UK@_3N74#;zp;OQ@v_;jwc6}%ger$@og77=cDU?jPQ zob_yhN`GYW|0I0P+~HwtH`Ioh5{DlS16^KFoAjgKy&36df!d$67ym$34L6TDv0Ia6 z1;4Enn?#X~R1C6D>0ckDf^;@1B7(Q@$?%{E4;D$^6 zPnLW_EYK))(VrnJUm0H*2&9HgWm_mVks=9HjH$2|EY^VEmRU=LNI}-1XjWOMwZWO0 zpe(;|&z_lxgSDv(g|-!MP8%HNs9|dmuL}Fr-Tb#9`r_2jz}-a=-c7{+i+{*Cct zneEoJ)1rc7vl-G90v-mG!D1NY-kVilC{uiqrscX{hgXfT404o>l1rpjS?a!2TuSZs zydHjC)j{`u_lE2gS?z!j$jIYkH%p^|An`@q_(~$TlYqydAKDLJq}3C?uVVSu^*Z zu#!%QY*m%h_q~CNW82Jj&2buo?&T1sH?UZUG~iP2nl3|WLP-8ZO9Pej(;-IUn-R3! z`?3G{Ci;$X^z9!`o_F#yGIeisG=0Z{m3NPJ{)C)#VyxV;!paqjy+n~lDrTo98;V6v zkj81fs?vFJ(#w#if^JHqqB<~1)v79SOL98`XlR_412Pb}mveUNa_=M4_sfBS!QBvj zR#p%3*-_di-Xju2c12g8i#`^JbJLAU8+d+{(u&vHdwR+_V6dOm8wN%W(Y1uTq zTF@EQ?((?GU>*o^giY+X()5b#d(GqaHaR=ye$B%~z?}Cx_j6f6q^(N#D^g9%q3K;I zD)cf$-|}h;zsweUB?Hg$Wi~tL6w&$iD--8oR>EZ#SI3EDr+LG9`tSrOguvU%!8lIq_EbI=A~ ztR~q!wZt=Pp%LOTZ6ajE#9G!&=|feDZyb{@%a-T(U1jy|<71IVfiK$$hzq)7YQOW2 z>T5#FswQaL_v^?CCk~ioSh>@iDRvV@;;EP_pKXB!%o@4&Fi69k)jXB8heJ90mH7td z!n_O(dK2R_rW`Mx&rox8F)q!#Ov&VZe>_?KRSPg4f9>lZkm^@vFX*NfjLuW+If|Ud zh4A7jnVQQIYlcEc}+BD5!O#-`FP)a4flP1j+uM_k{ zn2#YMAKD7wc&e)KDuraZ#?S%DP1%uv-K>!wyoditzhDR)1%wxT4DZXlY!M~?Li_cx zkfl0t(i{g_YMT;6ahTTMq*^}fB=F>pP@&BG?c4u=J@~P|H^<$%F@W}Z+1-{bm3FzX zQJo1@Tt##dt!-2rp%lI^w1CN`Ym`_u>Tq)~(g4^q)Pr-wNc}ka8ZT2Z>DNDd{J(!| znPKkE+c1mVb>bOjr4`fE7Zlq~kxnYcyiZf9TprXUzaAMY?(^1e@Yv&90pgjEA4jKq z3^~P^TAAA^QY!@_g(rd>rdLp>JPr$@9=Z08B8S0;?g`^VmWaC|pxD`57?h;aVtY2d z*5eYb?U7f^Kwg*gu1jFYuvUsm3=Im9V06VR-E$J8>9SVX?$#zxxe<~penNu|qc9GL z%_^Z9G2JhL?ZgTpak*rJFVKh1yzRchcfIe@;6#E)TAU<%+`HdO5^Q;A4Rey#9uV}% zk25!X_s!Zj3wUJ^>#qjO)a1_pJ^3eZTd>#kjk{T-*om=s-paeFq1Y;llv6QgZVlvh zN%U<4?jMsHJ<7D2FKl}Knp>VME+p48QL4p?=W~I*D#+~EPOe@S6F0Y-o^MqDpCG+r86BMp~*I2oyj z@wvv0ll;0LfL?(4QwN1Ropx)0Ly9w9Q!WC!LJyB;8GJ-6E^zc9q!fI(Pl3y z?C8p&w;cSXF!9AztWIo1II37N|E-8F19ER%$)GG;t~6Pm5oDyzEI(UV;AIZFJ=;j1 z3w#2+xxE042?2$@Kp6Vlqt>(?~=EOE+{^y zJ5)y1fa1tBBW;exG=&sGR(1}QNI#$tXpoTQIDO8$UKGcC?)|ykC-KZD!`PC~bg%}@0RjhK!mRZGuLZ(u=Zc4w)(j$+H~givfB=ixSw&S(bRLE}1NH{5n(xE0Tb%+exv%euJqtKU3N);e*>Eyv0~ zO`+JW6pTWQxf zF*b{^MAG_2lxyKo{F> zW{&;YI4Q)uCHI0;!kXl*sybP%a6*i>)vt>ucD%~xo|%XSe&184mHW+_=lA|}iviKB zUtK|Vyks>$P*FI8#V_x@ooPR2hUsE|M*fkfJ>~ zrA^+T$Q5@gQ^Jy72ZHpYf;0i*qqAcW*`{tmm@JDogh#|Vy zbEP<&P6ZN{3~*Tcl^u~e%qQ|rS(hL$9Pb2j7P^}(Qy3va_XvdR?}e5DCrce^S3V%; zBqdOjbjUZ$*Q{Z+RZ}n8O5&o-cP@F|5ZO2viTX1%u*8C=N1t{z20VR#&clU;B#K9S-+=arEM7Ok={z+H4y1tH!bcGs?G)kJ*EWF?2ubM?bz7c*B;xn1c}b zI1MKIq5A_`yCG;G^jQj7DqS`m9l(m99Q9N8q6yf39sI}I^G=6+i3T4V_n-WH0jFK^ zfS|xF+mhwyvyYrOqne{I8A;8N-~g686?3)bs2ax9AbITzWhRJo0Q(L>5!YPlVU13+ zp4GpuMb(aSphE!C9+IH}JxmaB=yAUd+Aia>+UE)FAt--#I4&^fhW5F;H-l_*)toj& z&ykl_B)Q5|FxvvPYfi2h0%w>R-9yf( zE9EGB3PKsTyt3&`$P!2m)~WGsGvTo|FJKP@-5R`CfNN?Bx`Bk0|N2>4jGsMlJ)*{n zAG{stcR-5+lI%e!DwA%<=$!brU1WBd6Azy?jT1XC9Me=q;6Yh?XgkD4FrLy5Y4#PubKd;Mp4WhI-fS0wIG-NB z3+itr=e=REGk*V8_959iT(sVaqa*vRY|k!=1%kJoR7@qkT#Q^P`4L|pq2vN)8P$u7 zz@NX2(FqzAUnq)|I>F#QI&~cLfIi}1F{4+NJiaN^b^(IJmW&@9VEN5y5d`=Cc^z$m zO84LUo+3+LnGl=a3Pc+zb^}G$Q8Cq#Ykl`9^CIi$GGPau5B!pyvW+BbVrvY|gKZyN z#uFIB_x+|&aeI*k6h%H~m8AVOvfQdf?=i)8QRE?PpaH$&$2+{Xsg0qg`5%sPx{ep_ z4nL{cHdCAG-=J8}X3_CbD4s!T$bN6s0L>LYpnKh0XD-4DzXJgsV8!EwppvYGNgmju z54ysRo3z)XI^f8hL`@u1Bl}aQYGv5UFzrsZZ93HVt&hx=_W5lV#)Wi~2f#{SOjD?nZp@mL8FQTv9B<10Rn&D`Jc;Z`P2y9_DG>U;bL3-YA$pGJ}g{LGOP zC#u9*eSu;pvwallrDDv=hmXlBag)4{K!{Nvc1^NA>~O$tg4s^xfh`ceZDjJqJJ~x7 zy54#wJLpv|=uP6|m2{-gmt}g+E7ui@}t;Z|)as zt1%`2Hr$BK^&tbm<=qGnKMy1fpxhgTIUIMG77_N`t}LC`s=A{nP-bbc7WNeBek)V6 z&)2NGhuajf!MQTCmtM%midVR2(#5jJk=3Fs!Cg}CI@pI!jg_;h^bJ`rY%6GyV8m&^ zNxz*``UBgzA4e!HPTE1Ykl1MtRqI(S7)B4M0~aqu<*7CZ{S=X;*P%8YbLr4( z07)g(!lU%JsnXf=v;^%2tm5$&=tj{2#m;R&-cSd6^p@@#6u7Q$^LQ8vb80V8hnd2F}g%zt#eb<9ZFQNd-(ci zvRv6iRjyx_=Jq&|zzww@qcl)MyBL>aUW4Jh25Oi8oxHK+P&lzi#gRjiq`D@#CTS5N z$NOU?xbnE)b91OPA1>lS>yi{us!l5F7la7(TEr^ZE&_D)7s|wS@6C!f_ zIt5sJ-2gBfKlg2z;o*afVICj1reANnnj2(sh5qz$>k+c+C5t*9vkDj;q}URQ6ry{C zVryA2_Gun><`hC*3QUq9(YHlf#h|~FrKt@(r>R%8h_JO!m<%3HhgY9>f$|&}fpQ2| zq(t>bB??+Z@l@Y}AlE{JA$OS7)f1F}L6|o8|y-Bm7 zGql$oIozt4N=>b2qQLCnl}j$LS(+@@RC;M}1_ZJU(WpF@D{f`LwT$c`<$^3tnqYr$ zw+boeP$I8fQWx0{Vu^!o7+}tixUPv8-jJmRbj`Uv7d%Sk2HzyrhN$DTSqW`yj>I?) z*zhA^XDhD(bKa%Y_k89IxLS~O{O!uWkY!FB>)B}q@kEMEphz4QGsp_Q=vY9U6qiMd z_IV!s!KYrGN_0~um|K+xUhPDhV{YSx*UA5qJNH|G7FZM-3zm@VS7y(m)(S)iC>D4f z3aFR{1@6=7eD_H7uH_!-vcrCPbODV;pys$C&Qt>dM!FP2b@60_K#R$Q`{$j7{e^u# z5Rb!Mzu4e%NxU%GKf`p#eb`z^n*k9%gE~pI1jM#RUwkkO1bEGp8k(Zc$JBvpxWnL-EJ zdl(kMQ1X$$VxKgE89cdS+*(3RG|3Y@%6&88Dwc1XRoGC2ZBIB>Lg9T)a!4@qKuX}T zxi+=~=1_##c0Pu4ZcF6y)ta4R3wHja==DF6b+3%ofzsR|wYu9Wb{qNxF-ycc&GK1= z{ziJ0+bRnG=%q#)y|of0RAnI9SUcUM$_4%y$AV*|@#77gkvenWwa-8Q?fS96!->-Z zIp8sv=8;V&`s%}s^kAT1(2yW^U6B&jRpIRum+Qk99u7pnh+X-|z471S7V9B#U64-p z@v|Pz>!cG%QyyYHDk-*%B8R9LBw?-&%m=}eMBh&&2LzW{1cV{_rYxVvxPQDSYOytw zh7jDhOn|z|m0=AcBYiCd@5MdKVhKL0Q!k%YJu}^7Pq-nvZF&#M4?C$uYHyty7xU}c zb;3lL!JYskc-i!^t>!EP$QMFZz6!_5h}f|6;@@j$O|q4NeML3`I0Rx(Xs)UZunnE0 zIyI{R(n}#q6Hy=ldVre#hTx$a!q0dCg-X*LS$>0w`d~Sl5)i8SYg2E94&8Lr~ z|9%=g#WL?Sy?Nac(c>B~$+x;pjBu#~7J;tmi2fK_mL zNRsLRtcT1wG3&!JG^jMts>aOKa@b0a1uu7;4T!xkd43w-EmvQ+5`_q|c~y+4mg^u03wtY&Z2r;0sO7Q60^Dg)WYy@I__XEj9;!1L~vD(zR^ zc`r-Q8F8LIAQ)m0tuBh)d-r4)&Iu|Ei%5c=Vv_Z53YO|5C zu7GJ`<6WC1TI3i&HTosuLvLadT-kJ!?8u(*)e5Aj!`j`1DaAfT5oW29@m&Lt-foxx z#S0RAZsvwh;7@-)-ODoZG=8vu5!vR%vHJs7Q%MfR0wqN{6@y9u6-0}X(pFW)+#-o- zORy`V3v#Oednu^7;+8ltW+GHDG?GCt^~daRQ*$$^_E?12$PQ{!J@p(P3o}MsAUOC8 zx83)0<6;=IGUfXOuu=4Ki+NtDIC&tPutDN3mihWFxE-EGo)bcjLQnpsUQ{KpQ zD$QrtBzwZOyCE002iSue)!Ia9j=F2ky*ZH7WO81WmxMYEWUq*{ZFCN-XPQDbd4jKU zBoJfyS)f4CLvqw*Zf*4G;2cdd5TRk(Tds6tNHPfXY4>_0dt}p9Os{*tGDn>Y=`qNa zrfm?R0c)o#0!w`cT#}iMWEB%H>=3R38VQ{`o6Zv_5L{K8V*kzLo`;df-}WJ0Btg4| z!Qxav^Ai+?YgLzppA>3u$R5l=;jJ7YNX^se&m5Vk{o%{}2UPrQWL{`Ja-!pCH@`!P~P zuX5cAITx6FXNW!@vRH^`u$mH`{_1sEyAs`Mb9)?fktDvilQu-B&G^KpRaL%VCvcs7 z*>AtghnwAFsdBoNZk(I(ZaKYt^aaEAg2~p$w)cFr*gJ9IzkF=-G@W*F#gUU)=!IFB zJrP&kE=sTp>zrh>J4SJj1>A!wD?Jk@eabuy(uXajR=(DtI1akPduA;mn6LNK345Jd=LNb`&z%trGGs7 zjXv*e`C4Vai~g(k-tF^V0V}TCs%l}fDAm8$J;(3MbB6-Q?ogf`W@9)2jJk~W8m-+upb!X6Z_%%2c6VJDSl>|kuJ=>GKpMv;u#5oIzy5g?o#Y+ zirk`Nx+Lo%HKC3EOc5J}`IS2Lz90~9#M*70dQ(WPEJ?L50K-5{CMi)@0pq$KRneQ0%DaK z6HDCU=+o1o96LKG*Eb9HyVei4)Z&32t9iw-C(mPu-`w=aw;pWy$bz*iKU_YCbU88B zHd;j#mrZ6DPbU3T%u?Ye!WL0}L|)`>QV4qux99e|bO~V{VXv2 z{oa}lBz?GiV$R1?v8B*|sd&<1b9IPswUuI% zD6$cVxm1`2nyg5MRX?VI;%reCk{?kkSPnkdhxN)M^FEx=RTe+EI2;w69dg6vE7b`< zms#c;QJn-=SEt@9T`z@H!0j_p z0DZtE8=UN7pVE1Q`e74dZunux>^yT0xC1Y|o7}e3R@s)%>fDJD!l8p}u8_m(%l7bN zuG&>!+x@jg81YMU?^Jb0l+J556|k7ZD9eQFXOEgchW8+i1|k3y-k-zwfxnjhzXi4` z*-qFy{LjN4kHMLqG1L&popY=ZL)f_WUs0*~V#s!ZCj|Mz<<2 zcyAU!=!7f_yB+;uj<^-FBC2QV)OB<|EL)Ih@1c8(sABH+FrE5_ESve8oYP7u+ZqgH27n!0e)igWFJ%k^tzkx zFfiw&*JPW0dPQxbV+4|&fJ+&q8`XMYru3mI#W#*gmu1Uy{I0Tk2eES6IY@j$=`S7u zJMQ2&NKVUT&Xn|(xmeJX|HZ$4Ok!V|)vwO#OAt@7>nXAZllFk!>KG&|n39X@C6$nz z+ct5<)?Y%)K4gD+|3n>?r?q<7Zbmz;1H!SroCUS@o2Py;`@V?;2KEL&)$CV*BZ6%? zvTTr6n9%9QzM-DDh{hn4cp0%t7cYPChKI#g{PG(4kR-e^wqmc9tw^I-FdivXOdE(B zK;Ab>_ml)248}yNeS)h{MxV`O%)BN^n*lWmhujD&-mPq02-299`>u~{6E!MUtEy+ViRxWD1%~L_ zKuk+W@x?yy&Q8^lISw~cqlL{-2;s4#;Zysw1pgQB&p0tSIQC~&e|zl*Prg;Z;D)S_ zIi}i3kY2G;xo%3app|sZ1TqfTxp(*=k2E-Z05DQNj=sh{6CH^@l4PqL=(O$#hZ^8I z!5&E}eHv1kFzqo@lOjf%_X!<%%YfN6{*E48JNn}u|6!u2ldpD1xK6jsGMe?PE65Ih zvy2nxOCGkGWp-06BnxDNgqca@8fjhPnR>d2E(-xUl4b69MW(;-VDSHs(go*8EeM-{ z2msC~al$wz8|24y>SujGDfalF&n|S?&Ly_zcA$mtEr0QSGu(E1;jq|9Rg95Kh+ATY z;vhj6yiDBg!ocPPFBkc}Q@(Fz^!M}o#yV{P#IX8@)wW`8PCLDD$g`E3oKC&4I`EDH z(^N6G@q|tmEdwo`jIca$8rdl9RrR}Ug>w8Z|1vkcS|Q95+>+d8^>3v5V|r1V`=OZ_ z%R`}na>02?(d!RLe#B~s68404K*3_3IGK4GWt=)NCr$8x+#%Q##03=zf`lnWl6FO_ zYmuZoBGn&v5e?BDbT?F?=+vlO`KL!v1-H&~`D>XocYEFmU+b~MbH!_GRgcMWdO2v2 z9g6H%Vhf1W{N0zwd^v zp&6%h{Jr?32T(Qdzi~imnaxtC{r2}HiQjDI#BOns)ohkYu}~$FO2vE%4lzQbWdUL% zPTJ$1FW53GKLS%@uLNHY-z|HjNRyUQyJg3G?g=jmyX0{ph0K1p}bK`6$h#Vd?7u+)Uz}c(XX)YMRhdYzjrTQZdL06gvl%%l1a~sxZ|Z7+-*&z1LlflvB%u zz%vyN1*X~b67extWSE+e;kS3lhx!*bqcbk2yzIuLUoAfN$H^9eEL#~|3py-L>!gz% zRsh*RvFj+Znu^gy6$UR6pPg#KKx}ZEW9V%84v#&E;M=Do?&pVy-M^gr&o4Xu*h<*= z>viAZrNuJ!ewQ=qUR6;CabBJPr z>8_ZHL4}8G8o>fPf|wtVY4imQmgKDSHaF|kc$2o7B&nKJ@w3*%!O z&=&fi2G%juHb3s4D%FQ+&r8w;$ilGPe|?yCP&mRd^Vpc7_8j&caOf|r1w8Ig zy04HeuR(g!(0s9MiiMs1G%5xQhE1#iEuuE%8q!8Tl9~0DR}n4pS62s^m@VpLdJ|Eb zRu2PWjnk4Si|CLhUf*WtEl~L755AE+%YqaaspbI5vCjfP<@PB4t90K|; zJ0Y5Bzp*7KblRvc2j$=;2$DTgcB<+{t*VE@M{mc!(?XUAuS*>>H+=c7>^QP-nCPXb$lf8_QoVAMzgcL= z#HrHaaS#^Ex8LkR8r;>Y)xfB7Mx6~41}fmiF_6l(Q_~`9aN8@@W{~Y+RSb|eM`({T zJ@OPc$BRRZ2FTX1HlGcP#xXSV1Mhg=IE5 z`K@nXCmV;OK5|~~i2&2lkU3)~#X>&hb}9x}F|aO$XjiX$t180}s>D3ByJ)>vmp3kR zQ)!@YEARs85OJXwM9)zCVJ)*07Qy3+0@GJ_)KIYFi$9iIxV?z?Z(n@F*8-CKfgkTD zdMDo4E4Ko~K8l5s@?0t=BfLzwSy1WIF8cDw5S;|7^!1_xg1*_8NNjM8cb~9&uAbc! zaYq(9#gaQCfZWjwhwK55jWhQ|==F%}qbJP-U;EMYZ7=)Q@P&DM!qGK_%^KVh zNR?*F{wJJLyL^qYjqT{3=7)%os^|jSB8*qWhvg{QXaUjgGB=cbFOK}g>lCS#X46|u zNvETG-=lxZE}Js??nnKk=l@ylYFj7FQI4AlTM{~Tqi23Vs(*v%6q_zsKgCEd%G9)g z`**?>4i`B-!o>X#v+IpMo58vJar6~l-%00P7H0H?D+Gg^cK$9)$Z{w4P}8kE)Fg`C zNRbU7O%|5p_a$9MzstSgdQrbim%JW?G}X>PWt`{w19zyj?>zoHQhiFvsduxoZW8wm5DR@U52F_Sn)#V)R8b77JD3xFTxxzm^p7g#_niCv8|f(^9;ex z`}-3}9~lyRdnWDOuR0XIi!N5`1lja0CIWQYi;=(`7DH-7(g6s zj|J>7N5s1_3%v2aRpx4ja?(jsiPl zg@s+u?^+;T|FWkZ>BFu`T0}baQ6RH1Kn|-p@6|Bg6tBa*U;Bo6xF^TAjoU;F6M>!7 z{vGDK;T9)G;<_N6>~rEJ(OIjRq>^IGC~}C3sSVsKKv=|z50Xr}S@ngoUV^ecAWorL z{p}mFZGrWoT^@A-vEoV~`os|ZZg%UGuIc*ehUjIA7SR)65xyXLx&Vbvb?OIR{qX-6jsVBHEl|ejh7Dg>U!})*w;$3Zb;Km3oW>hZ=OO-AQD-aY#V7Uddv}J1We2MtLv^5Om#O?R455cky z;8y`;w$Zr|J#aYI#ssh-#>f$%8rsLsN8hQNYMG(h=7-)ODNZ~?m0HbEc@(>gBAHYS zW|NqeOpj+kE|~H#A@Y&oklEd*0_IJ)}u~O=b`P zUu077ZI#2v9^bbk#((IzIqH~~ftj>^V@dZk3tGOD_p5&;+g}40Vdw%hpJH<;l10U= zUU)-xN2Zq*%i@>^Ud75DSq9P+<7HP^Gr`>=aAy*=OnEIiv4hUgsTYEY-(zd!ox zP8ucAjs7=)XSi3fRlZ?%?6k*IvjlbH6nb^Qs5ojMtoq?KD$cvM+^2~T`I-eTzq^q9 z_vEM(8yAQr4hf`uLb3G}IYY%%2cXJv9D`M7z)+(>cBlcD!4d(~LI!Iis18V#mPa^s3Xc5rSnl5@=3cw_&>x|B$ z5;zS8T<**7%Zpyug=x=DO%udU?Gy~SG$=5*-9jfbmGptx#p)hm>{Q;lLjO7N`&W%6)SBw1#{LLG0klRjdf0kKE zn|4!dCq*7mF-T>-p3RkZ3f4!aL}?qvz8uXaQcK zv0w?wc4FYwT1|WhC>DgG3aFUn!utxWx$X-sVCn^zCGB(;Ong}|Nbw|x=^-`5T#?%H zGzKFy%uF^Gvt#`%ugPuF&OiDU+7>JFS;9N<Jjy-}wl<2{dstwaUjX%zOKW^c@ z1zqp%^M4#^q>)YX(>ce&dm{EhY4u-j!AIhUY?1C2;H7f9US8yr!`xrk^X(lAb?V0i ztL$r;cwzQ?hUlBJWcdm4E*jaW_6Nr?9i$aL>Yi||Tk^kr`qO@w3;#0kjWV~*Q`4mJ zf;{>2p@Y$YXc%)e)~4r>$jd5GJAV1o-`G+R^07)z99T@V0_Zcrst=zCK16KnV&a(d z(=U4$sjq`ueJ}XY_?9R=H%{1KH-?6Ndi)yK-q=$2JBuMY_zMpOx#z@&XtkAl{e)tB zDDsGkS?AVG9?)Bv(g3YKEJ?K@`jDWK#szz&a(PfY5J=tgEcMYM3rDUL9n=9AvvP=;f2U5tV9Op!rdW%j z;m0BViI*D``AnX8CyO0qg8v;aOr!L-tPgN`{otFtU1s##?@M^UyJ&ka3 z4lv=^EM~jE?RL!xj{ElWnlYR=ukaV!BtI8fjMlvavckxmxxX2%lYyL zu4`sP88j#g0P|o*5Nt7hE?KKemTwHvHmR;N4FY5|McaI3zILZ)f-qBKV7vVjgt^ns zL3pM?72L)Ci1j=fJ-@M;fEC9giHFU+Pd$+YQ+l-^dHpRP1@(S6&~qJHNj6XR?|f-<|VL zw;*CObPC%>vGASPM8#}({dDSnp?vH1T1Gz6dj`RAQkw#@&pNQ{m{CjcR#Q(u*2glk(w zHzN)1<K|~(e`;aW^aZZoa^E>@=_|- zFHL$KiHGP!sooXEt<2Z^g&TacGic!_(|@*L>UT}MyvZ#mj;1WO3ITUg>;sAz zsh9=@mNzUB>mxS7ZqXu)^5XvXes2u!u6d)3)#gf*J@zVhhsTrZfKJGYScDPW`-)Q} zZN`On^JIquZYeP`+@R2h;ZnXx2}$jsL(mz4;-5uI$b<+VaH$j>r!Tx)rcNV9`b@AE z6S)#1FNpR@FrZvL9RfC2K>W86!oHB|!D4U78IQkIg%|!>Zi}u0MOLN<28ox5jT zC|rR0)5v`T0b=p)aEM9wdc+H_LxC#>5}@3IZt)m8OHK&F>;R$VmS=e5M%Zv=H6a=0 z*|s$ie5N`lUM_*?z>tVv4#k3AWjgTTgWKE~vN`IcW_?(tFgplDD%Z0OkQxNz4ypp> zA+4%R4PF2kWomj=dSP53=K5Td^}Y_J%4+S7=^b9cicgQmEIk{gkq*X+tdaU2%e8zN zwK2$Iq5kx7>k+cci7nJItILGBtSttHj5n3S@b}zk!NdNvaZ3>=WyT z(mDIWas8F_y5n$SH8{hb@IunCJ&aG|XAjeU_Ljx0{6|@I70Krpv2$WK`=pf-Izq8f zcC(*~xlDT8Gc?Pit_V8ZOXm~vnL!?Lj?)HtyR!;w&f$;HLMI$!(Ey9>ZgO?`d&tdpo#j=Pi-X;5$k z?M8*8hTG15{$t?ox93}oO5=~0Ws=g@$R(?9;|YodAK@r$0W(QZV6_LtbRU42?%@Uf z%C>Mr^bzk3?7o0q^xCOc!uJLAyYx(57mYs;g)dX>3+Ny>Wyk3w-gs}-+@;?@iIi$24ZiOK)AXuU@0?r-F!ZlfwOxMhWj4AsLjq7KHP1* zd<5q;|y(39OKDiG6ZPn>=<- zCDg;T(G5VkG~iMyZJVCxyVU<+a2~xn@_s08E+na1JT|?q#lMo7M{b=F>sf6aBue2m zTq>Ea#!2yqPF>(tJrgsg23#=dsuLV-{Gk2b<+Q1GB3IfWOZ0{E2lAN2F^@@eh^c1c znrt`xftf7K{_nuJBL4Yr#Bksqwv89!C$(KWa>;FQL8ZO;2eOKv=i|gv7p%31EWWo= zY!XE_n)PULpAu@t_kq)c;&Q2>I(0(C(rA+=;8JO#@6aI2urJw>jr@S*$WI=*uhbn9 ziEVi>pSX(ssCtKtOVQz&mZ36aHA z`Vv#%RU__kH$*QH?~yEJs+d+)InxWe=(XZo68jCw$fNMwReo?8gCTkDlSghi{i{H? zKUge@RQ}UQ^1z9ws5qpn~wLe#Zj}rG-Ijy(`Gd zu&q;)nN6T`jnoDS9cwOS<;7q>y{R7Fit1zd<1jq95o^VqL(%`p= z+BI5?(p;HQ?~4(e0T)bBO_)*)nxTuQU}F3+pF+sM>i0{S(m(4eQzU_%sSMfD;5LGy zhXXFx-1fQYeCy?vP*89Ka^%xy0MTi{!C-BcbYV=SGkQLi1C=0J*vcWGwD&w8LSj5+^ ztaopO=r<6Qx!r%GabhHbMlO7qlncX*%oVZSE%;^@=UKlpWt&}ravpki|V zuIt_UzuWkAs2yMWUd}>?!I2kK?7=Rt4jpteH+K83M69_I1Mzf?*86q<4PMKF!@Szm$IU zo9D?|et{O}wSWZpbcgIprcmrw3dSrZH>e2W{ut=fV#PVsn;&O-kcz)s76v(D^no`+4Rd&!* zcO=D%2U*;S!@J`0|j@W>PdQ4 z$sp%YPCsEvNTp~Mb#h8JjXCDp6QsbaQd8@BNV7W}ewjun&dQ?WVDEE<`yJqje6cfb z7m7I6V`FH3Ar`r3sR#e}f7QS2b=x_rOgddqFRFo%>+YXCp3^!fF67Mg%U&-YuNS^J zzBkwh0KUDB{eEr$4CAIy89J{7TmO~AO++y!AQBtYV1j-8B%mJeRN-Wwp+Qme2^|Id z-v5No>ZZZt_5PjLnt$-0Z(e@c%an4&Euw6}5mCW{-EKEzdf{XDEwc~IKIfeyC=1CI zl)2T5w#-fly&zobZ-`zu<$^Ge*(uM1Kp#{Si1orxVdL?JfPkXsd3 zDrhs3<@u0+iR=u>;wp(-CW)pTdi&o87uKue<`Lsdwy<7nmFn9^vAq;| zOvRwwK{l<2JxQY?UDifJp?#;SMphlrpn#k=C}V@97U3PmLnyw+!W3lYO$C1)mB^nx z(x8Zwo`%4jQFTv+vYJV6U6~K-aO8$)p)p=_T9O@fGiwVVj&_Oa;OwKCHRH#)H+IwZ%o;iLVA0GY zKl3@@LxDWup(dqD8HBwf&^)UFNdk>`n*DI7VG**8S?V#2=O9CugSq(49Wmc@X4 zoO9?dspDrroHxoqZd(};1I1pT$aydzVYs$w17?sU721zRs20(ts27XVWvz@hR-7Nv zE4c0#AK49Pa)Opb&v$#`( zsK`fKLy+dPO}0*ew8q-4qyls|d)=@5L64t$bt;h(auLoxJJ{#?X!P<;j?!wcIT-e?PsK% zpWSidVu7nxcIPa`o}x${71I^brtAsr^F#6PHrg2a5K1E}HK*upLFFLde4NJibK-2b zDrT1lUN=PVq&p+}T{h1sWV+yIyRylv-=#6E(x>00K~l(E_3L+OW_J1EJ#lQCe2E9R z#z9x;2P*uDeZqa>C9Zn5-{q|QoWT4E=1<1IIqk{FBh4F7yzR`$pU-vm7hxaJW3e_) z>| zqiuLcMuflp;FmGK{l&k2_y-xyx)92VCmfED5Bdj3M4hTUx&S;H5UxbtnNC&3T%>it zB|BCFB&m+5!3oX_$d@BzP@@4?);jf?=)8admjM@KfUIH?6(}M>-Y>Q7bXwe35FBs;4Plp60Z~y2ps0+F z3W$p0ibQlAMg&D}fi8JW)K`EUPZymzDW7*|avg;KlBSj`YiVsyc%Nsb0+|4IL69zyR~Jw+VLz;y z(v?P8;?k^P8${O7-$Mlj#(=Nl^7or!-)BSF1`rT083vC!G0t3q!6?-T5yrLtpWPu^7zj+T@JKFFUFP46+;5X za**8;4N+h*=K7EG7QJgh+1Q?+e?|_BBUh{j^CZPIQ{*@mcT#>69L$tSi7~hQ?(mDE zp*D&Sp;~^s@Zwi4N*m>mg4U7_MTz$!h^iVD+ZB&NII?uYUU#U15_FpcAqURi^0@^x zc8{hV4A{xP<*^)exNj-0gx`bi#X^uod3s!@ihrXiHjxJc%c`#MROQn%s9$;Gt&7q; zx;A=IXwI8WvCkc=A>P-h4c`!-{k%h*hR=0ZFuVWy)$J+^u40e;WG~6&wl;HNKdQ#+ z>rqBA&{$UlF(L4ca)2&fi-ZB~PuSEr- znJp%TO5{w((Bjj)E9PObt=oJrPM2^Oc6DR^7Ff+a;~_-dM4Og!@qXz_ZH^|{6ouRk zq~}0%kxye{!=6e~)-iM}jBI8&OA8b0hd<{;g?>|_w*^MuuiJc@e8de#u4`05W_~b? zvMEMSL2XA|0Z_v(mdA@$gYNUCNjpO~PO73Rrj>Z-MqGp>D<(6mfJ>`~zXsU1R{iLr zG#`AVc3~;)=$_D_24$Ei8R}6EJ?8hFFQ0tJf|7$jJNFr>A1dqa!dPgx!a@thK=Icx zD()}yuR~e+!N@k|lDK;I)z}^dE}Slh?+NOdfoBd!FNxU(l*DI4%INahcl-~MB{9ew zlE+^MM<00bRycpL3dL{uGQoq5FjGTCJ*7WqGh6Is@d;@K2#6gWf6W_G$6GIu+SX z@VZ@|t%{{#@Z4RYo79c+9wl}!FAD9Zmjg@15TF}9U_B3e&w`E}SVJt{#%6EG2(nFp zuw)K!81It`Xb4a7JV-7@9FSp4WdZLj|KzM)Q2KH}mL)sp-NWyq=dPuGpF#cpBkK3- zso!T(b1`Yz3ZyZKCed#2XrTi1&WuB}SyTZ?8|l=DK6>4xMqxA5??9`j@II(lZsqHF z%_c-tfiH_(=Bdhc{M?A*DIJPxj||PenPrk;C|i!!xIH^E&p*Ox++253SwgQ}NG#T? z+)aO!EOTM&m1||a)>BLhMUtVPi@(L|io8K_h}ONa_G_2sw8mD=+8$K@o%tKyxFGFR zEFxXOLl}hT5AR6Ma)InfK8NeQ+#s9sX#KZs)$&}nxRpi!ZixUXFmOugiK?1a02c9h zNT2fRR9tl(h-y>jLuIQzq6^%fbm#yaa^B$jlI^*k3qZ`Xi9Jsp`8AvXGDf`T;2~S) zJ{HQmjZ;7OSxL_F7KoC0dY~g|iEWKN3IE?G<})p(z1!nkqxVUVMVgPE7n_fkMt`PU zq1@rI56U9e@DD`6w_9EBi+dM*R`DGF#pXUZLr(Z39}Sx)a2g!fop8=<{ce1uWx^4= z&&ej+UYWHzh%F8Z`P5QOH3h8LaZ7wNylca*&B!G8c;Jjp?53Ca>Qv~z?tcC1RMcC< zbFHz+l#NF^lpQJz*q}3hoOfTHG6_mPga<;}WH-gDl-=rDbvIoXa1NLvt7d_qg(f8s zlVx{2A1RM(mI6~Yo^#08!a?5ztFMO*sE4(y?f%ujC@c`lnE0#Tl2jM=St_hRkxMaI z6v@C9S3K1Jbw>8^%~i;&V_FrLJxn#HJA92YNKqkWW0$;6jVl|ZE(R4cn8(`bB*AIt z6(L*3V+12VtP!>MCQdeR?7u`G*kG$p=CU3PiwgWT_cZwy`j~rTn0Al4jGq@#7octT zXooefc|Cm(vYLnvhkW7f8U~Kou_LdEfCzqQLc%t^9nM(a#{n zYeoYp95FI#J}xJWP~ZOb_cz%(xGrlySUm3*ApJcl?3a@ZHj)sLAEAPUz;|OD7 z^$c#PdAC{dmegWd6n}lYg{*O5%Ti)xSu!YQ6Gb*+MCCGX0aX;;7FMq*nR#?_3n-@@ z2m!_c)ArSAMUgtyhegrrB{%8SqLP`7DrazyTaAHvXE+EM;xP_C#0?=Q|LxK7_bdoG z`|lMGNDH?mkP8gVomtUX>c?4Pcd(&4=br z>@LMrIzAUK&|Z@l@Y-a!i_bPbH#9UHP(Ew5hJAz;%=KRHY_}~#eHF@=vCFt@PJE0u zo16;kQEUTWI5olu>!RB*yO|6A@+Te&?vnyCdK9ft|JNGO=LH;+X2Ex;c!sbaFWIj; z@)~5tvDaPOP7mz%My;Ib&^VxD(ZM>{oKLWuEa`#ujBR(i%ep-*o#>r(0;3mP<(GKq z{U3U+|xnIKD>LENQGlNrQ#`x0}@Fcyy% zf{0xgT@2JcIZ`88?x96(DCGab3Z9feAl1l0njT)E1ef$Xe06X&d@k~AtmI_?n_H7S zM^gYj@AsnHppJfbB-ZjER~0-OMA54K3q8#Jd{}A z3r~U&;TaSMOre-$iX>8T`_wfdd3;j`I^$_p8j4=Lt;46)vYOd4;?q&}~1!D$)p-HF;5g1U6w(pp)w_Z>)bWJ9UD zAq>wuexdd3K-mq4!EVa4k8${s=fL4vSBCn{f1j(i^>|o1RXkK{{9mJ2W0f#7Y> z>{3oK5FIS0;@Y7&I3I}I8pD9{TG}8y!0&{zoK$6cAZ|6l!mAz>Rdp&X2|5mX5U3C-1VT#2XW`=IwRgygui|KY>CfTz#JmqhRO7_;L)IU;H^-ui%7c%Ii*V7-a#? zBdzFXwH0tyH`@v^BN`mMUBM#_0{}1E7@K`T9R#VZjzB zSWsCWT^qwJc49?p_cl^8)N;mk3pNN4Q4F#j)fBUvB0H(LDrx=H`~W@V73!EQSxc<` zI~DJzyoaLR2Lq}>j0hud>FP>#cF@L<*4R2`?+oC*nX{Vd=3jmd_m_(>w~z>#t2~&T zw)o|dkHR`)T4M*ESx*wfDuZw$%J!;b90fjmP0b*O>G_)uvN!CQx;XaB>;Jyg2~oVt z)K`~5f1-lG3Px)704@F=ST%$4pb#s)^%1B1pu?#%d_Nxvl-5O5s&?XZ$u(Ap%^2P0HBO_ z8agSpiI7`rk{9z5C0O#C?Ug^(5VqPEIoH}`$TFG;BA5d#klmyPv=;l_Q0x@1#)eF6 z(!B!Q*V-DNKKOB!zuCt`e|u*}8c4_F1NUt`7$3|i-0;2%q1s`apAjSG`E%B=TP@jV zsr26-yUMp%9=Bk(g=EP%l5O>sO{JK%6j?*Xtpe@QL1P zmOGd|=7omU{#X6NM>+HY_E$Yr%-m&($on!=Z*7okjvzeAk7ools?cdd9!8C{6Ax-QXao2YLtLt7dEVBk>QRuxBEi>^GCm! zXqiX8UG%eGl1l zE)A_G1wpvcq1-t0)|YguRlsDVT|}}V?t(=p^Kv!mBD`@y%w{zfuxJwnJ&J4c)xN{# zEV3VUwsytN{b8bVkNcypk+>=@98h6#RThel`E|i6*&I>H(O_{?i>euD9d#NcDK{G^ zCrvX$juIzq4EY4-98CPuNB?SDC&$H&bzvLBQdQR^$LgiWie2t!=FDG6HOaB)!2l&q z7#P{3*{3!t4-qV-$bdq)0%@V7#4kxu^#%k?$~?+EPR=Tg#JSESwAv0qmz2O9Q&k(5 z)9F;GN>&jSKf@?IGod813F`Im1|$d3N1%~eD1vAUhJwwdQ%G-t&D{A@yC(EONgqaC z`rWXQH&J3bP9wc7P(@;$m9~|4a#qH8ou#i*TNm)i7kNlld6jr%&`03YOCq7H4@&&@ z02f;qy~V59R2-Z?ak1QGtLj>Qn;CDLZ+td<=fW0k_k6XKnGq+~A?$M#ZAwI^GI- zC3B6Zt>b+j-owAfyQtg)WU`q)=@B`ay?D6Ve`UZm-lfO~$Qi3L5dlvGIBzDUfcE-3U)4*Ptct9x;fEc@MpjQQwgU#xA+c7n>II z6^;Z5yM@XA_0E4AiNu@>yDcmdbEp`HEP4Y?Xy{+Y!y+7AR&-u6q3{x~NmXX6(?LB7 z`B<@^1FIbja-(dw2DN#eSF-@)ZkL^5#gQX{WBsgKkHz15*_-j(g}0SQNxI(y;FAB8 z*BjkHpO$S^=mo8@)j^A5Hu)VS4`aIg@LbuPKF=&zN8vRNwG!4Z^~v6_M*Yd%&f8!& z9z(b%RL$7$x4i6;UJUWJ$&RY@?&kRdzxwlAKt3)sjfX|Jqk)E}kKpagrk7Qj;3*MZIRj%HM~^bEem-A`m09pRFmA> zfOi+^J51Oqf>3=vbeft4y)_u!KSGv5x$a3twf|^DRfdNUN8y0|IqV3bKB+3VB^u)r z;dS9;Gz%@+^=Tc-CF7bsAI4NkvI6_VmyUbvhWmPEJue}`ymyT$kG>n86^MI_#$^Q_ z;C-e(HEZd(45zcK4i7Sr2Tbm^$xithp{3AaMwr9D z#;$kC;dzd_l>OT{gjeXgDU^4s{<}Hc;+A9v=By*7+@fhN?67=d<**niW)DSnQE?|^ z%S9(-6(D)MNCq4m+3G`*_!(Pc<{85B=>nc!aEMOgHJZ2*diWqLh#Fvx^82Khd@eT~ zsD!3OTr1~l3WApSUXac!S9i&Q0W+vn+)g)nCop(bp~MhYHyfADYsChTg&TIl`3YR@ z`W_5+Ozhl#@&Uu#*!!pAw{4|yTsDl~Yz2uYVmBQ#3L&*tM(-1#Pg3U7E=-fB$&JEG zk-7BM8J!yJF@tc36t6jJJ`=r!R|}!hj3xG;fu1x411T z1=NpVj&a9n2g0x(%c+r?1p8k$$KkgB!6Uqqlkpg{;EZgEZEVG5!NC$o*~&jJ{zB0% z{1i&CKfI!>opV=_O*)k6eg?0~pw`$7&`3!NxliVnzx&|3pnS3<9%GDnF2%To(c*AT50*fPZUf_+45H5pm z=Y$;U4?q0S=QoyVXV*`>Xp&<^)B+Lot_qU~@OP^gdwt<+;*9AGFMw-7_peb{3B?M~f<{+F zfTp0mtMul9+1?DYs97ERbH9& z2?-3Kxfm0Etnw<9V3j5uGbgiA6tYc;44Qv-pm7>x4+d`z(`F5Rgj2TJZC22ilePy~ z?9(IW$O@txNA_Czlcf{`4(1jrZc*qO{z7t5+7i3_-E%M#pl(T(=Y2tjz$n`|xi$6^ z{$BBcS+%p)`F$oasE_e`f!?cAahc?4O8L0JS;H@-A5YGw&+@nOkC4*mOH2>{H4JkM z2F)MC|Lo`8$vy@{#B*FXVEWewOQ-l-0HaA*QA0MnFhvvuWCqncl~7C}Me?b*V}6Zc zjp7tC@Uta$&$lm1x6ZkuJOf_Q@ArSr8jclk-9M10OP7Q|jYNN}L zvY}8?z}xJT%r6uf7>$+ZKo0U70MKQpdKElHE5NP4Ad zlFxl=3HI67^Nqo+u`86HDYq-?eG_?g0i%E`&X6)05~rZZ>Y>aJbG~8g#d0ZVaN)2F zi_eQau2@=!)lU#Ese>MQq+rLAToXS#R>Ulp>%98iu%Qf$~(&%+icRE#;YCOq^F z3o2&JKhaLA$B{Oxmw%LE4pZb16^EpJ%OeVWilh4YDEm_z(jJagtz{wS#gL@MrawGo z)&y#kl?NvgZKh|B5*HDPP#|z0sFsaDZvEJ;+CwjCP?Yg26^-s#=$IXSJ4#!rNR*r= zjfzaq-4J$zan1q8_7>H0rZxZ{4J!`lQr2E3`BU*jZ<1=kum;`UVY9ZkQb4_6i+ewqnwl$sV|V-5hSK+XS55A zbQOd0@o;&Ea*O7gRPS*gmKK=o8(~O6$J+vah3)qYQK@A1k=qg%7@ZQHoKs#IU3OXM9Eh1~!9aCn{|2a{ewi zR4iQi^@m@#tb0P1O#1;j_{zlbJFJ#(rzi%f3>&F9WM#j_KMM6{SHjPTv8+E;xk-IS z+$i7Zjw)3TctxUaI*oKN6HP)vxW8EShmGGBpCX%A(uJ7s6@~7C8`7 z(a2~&qk374ru^q~G5<_h`7$rpl>qpLlPLWIF3 zX`bf#^hbizyhi>Or?o4N7y;Y{*6yph8KMmJd#h~q&0eYs1s2V-d>RRRkS0}|>}7xs z0gm5qyI=S7FSG+B&wqfE&a#8#&)<7bXDejIq7C&C=w-G8y^=ZJ0^EIlUP%H})I@?s z{E})P{@im-UVia%;MldUeCUp#8%)7DT7RFPL98+W?MNTY>9^LCjOK+}bZyf%}p z6_^X*3Zz|hzgwHE&$B{|yzlum)=Kuf6^XJKxY(`Cu+Kb>iXLO@>2j%xa$&fz_`ro=oKlrNij-NKCTsxHVvfd^%;8A*@kxGYUw0aMN5VgJgo{7Gd;Sv0>ODi*EZyhATSkd+Nxq z;)IT|b#3G1Q5JN3e`k@NRJbrYnyk>VpJJfnryA5dy$)z_>UiviL7oC-E^X2Tal@r> zExnW!N>D)Ku^SfanUy-s#83C8qxZSXvsHnfY93uR3qrJzEkc%l&=fJLrxtpr$(re2 zg4H&-1lm5i&l}fIIBz+^N#Tm30YmeGpCg75HxzyA4{1NQ%^SW{-U&-hE;yyK&hSK@ z(JM)$T_LX*%xfS;-Ywo(Xyugk$aBWcexy9-Ax=2Oj+FO)ah)G*u`P$*Tw6tQT-dgN zWZ9sZXFJ7|QKXcL>-2zfl{JCgK3HnU+ZUom`3_jSB2`mO2(}0|LPX6#14r1vW~Jg5 zjnqENL5-+Tg!jTD3Pru-sF#_|A8s!ceeAv;%8Bu(LE*udD}r-6!3qK2$9#eH&?B$BJFoVf}jzpp;f!WaC@ zdfSdemt769^bn$Iq5;;v*khnmp)rBR05m;01RGM1&)M`wg*bs}Bo&%Zz~4nJu{uF@ zSf2xjm+gx`O#8z2IENkQW?zD`?=mm@v}9(&*b%y2)Tp=(TAJyB%Xt-1nDs%iha=>G z1j|nF7+wQ2R^+mYmr0c-d18`aliDz+l+FObDmb~|Ss&MRdlG;Ucd&<3NHNe3pF_o6oK_T~XY$os z=)RbWs0)#qk|ml<&x54IBSoWAbp?OMUoCj-wn<&(bvE>|+ij^{RG`v}GJ@B6ZkJ-K zTD+)4dKgMr91feO!#y$pJ$;U&x7h8@_cL-e<1DUktnqhRvS1v67V$wIW-`SjQe+hs z*A@Ihvv_03c1X-=Q7^kwaoo2^ls5|I-1c?Y`I>ESbNW8}YaYUVblnhm(XRSBV!=eY zoBk+S=EBRNTr04yrzH(>%!1h@cP$3(M>}M1CuAMG zo_R6ixk2{+TcTXsY~Qm}i|ZmAs4oVawnU0qMUmy^K6g`s$DGd@?QG7#Fxdf;flEjA zCp$<2KM0&*OU2}}>lhZ=rBWbq+Ycmet+DZQv;x&Se*DB`5t~Cl4KG)>#v1*5Nw&ry%#BzVw2r^dub!XjX%u$wE-QP<%GdH7 z7%q7-*meNq8E5p_15Y03unl&A3^DOhdk@@sVC$fA@^FdELX5?Ay*GV@@~&qAvvL|1 zA*DinQbEuL{^7s`?_|M7kh!}$vrE|`j~CsGx(dqP`BU%l`jqD&9D>)>3i78OBS`|h z_V3Vj%{ifpvds!UJLR}=83{|K@GAc#2zsElA@VCVDpn|y1yGmPKpVn74?pGinNoXN z+{z&7!sw+6Ihz-U8Ib2ch~2zwX$>{nE?->MIIyg%;%Ddr8)d1=D%v2t1!ax7bgj&h zG0^9~+rek;8vnsJjM6sy_zy9?jCucuJ%4@Kd(^Y}b#%*2^WUwpsmeTh0mL}4db$IS zZ&qKQzFmQ2=uVsvY{%9va6EA~Z0~m7J|}Qc1tmZKcA#Z#TipNN4x;DQBIv>sjlpVd zvyEb)im8x_!(xD2Ggf)6kasAn{qyMtV9Ea+n0&A;WIj@P^zrlQ9)2ps#}av1F_0mt z0q*w|jt?kGWwewl*mSy?%hqsg#E3!U0EIF!cX%^ucl zbbsu&O1V?6-J>q!YtgXe2ekYR%H-|Dgd>OBu4l>qP;|?d-ri-`EG$BC&@vt2Wh13@JL-d+iEwL?Bk+hqxo>Vm}7uo}x=##M{N98aOHyC|v+areJ%9PyB zUs*78eb$;O zQ$uoK`-*ZWlq_D5ekL?dUo`~_OnSomJikzMDOXL|E;$|8?6oiCAkj~~ASs-Z<-OJ$ z_u=}W)~R?s$aMJ}gL3`O*jm8lJ@Qzw^0mc&Pe1h_8Mv|~c3=zdXyHA*=9v50*}#JW zM44HdE+z7^^zbt$U5@D`_z61A+1V(Cl&ZYqlfqm1TJ`LK=d>xa{C6t`gk`;9yQ*Mhwd)+~RWTyx>{@lq?C^-|1TigaTSvVlGx(x*H*t2DArmO*EP zYEzX>a=m*wFOgh}Ea3GLEZs+zw`HOmp!JhMZ{t+y=88&ELmP#>eDY6En6V=5soDw^f z59r;mSBQ6N4)|^5>!DsQM>G0qY`cxsAVV@5o4x67|6`kRbXk|1C5MP5)?IRp+THM8 z0!|NdzF?ye9>d7qtr`7pOMJ7K`+P81l76=>5Tr|B^bw}IhJ)^C(aLx$XFAgtw|n@9 zo@Jbd1Wra|?04Sve&-#_{B`hW=RPC#-287BUbDAb8Icx>fkw4sR2=dOK_3~2m1;34 zSQ^+TZUwoH2e5EMK2rz{hju9I0v`Fof)6jK^-Shx(AEAIg1UXKtB>=ShPBhzr>BmG z%SkJfs$BNHx;fWo=u~x+t5rarP)QP)3VL7Y@t^8c^@4Kfii{T_+!=soO#(Bpxb1r9{DnD z4FRe+-$r4dR}DeCvR&FKJfb=PD`hI07u#@z0q8}^0V z49}YlAN}#g&516&Xk%$k#CY+auYlH6Ij>1|Ty#vjPq0-`9Au=g&K$hb8TMP|a-+RS z;qpvJT+5CWZ+&o-+A6F$s0%Nt?c*?iUG>4ZB$&dXLZ;ukG->Yc;sq2g?A>` z@#+QFVA+v0wOqM{KB`KT+?mlU?ttY=ExnkxJ$feyf}A3ce3yxq2nt2{6K{fA6BZxk zY6^mOgx-Ve465tXb*jtC+q~7xHb-Zh?Hvnt<2!K15hL*Q8SG;*#G1o(zmA(%E&jmA z0-Tojcg!Ojx%n|Joa)|bh5rJI$)kX`1^k$j$nEZ_%GBvdiD=>sLncHl>ctvx{2xVa z=FP+Ac!*llccQddhgA_}2-`a~9+Wlb;eR`2_5N$UZ-nEvgAJ6cLVnTL&mOi_jB(io z5{r~gsiN6ir&&mD(|MX?flig9xgj{{eIfD)?_fX`)uFg8uJ*+Kpj{zV)M90ISQXU= zl~x^!YL5%kj_~r;8BQa$HJg?ZjQvyF&_QIDv-K zu#fEG4PSdSaf$^RZW7H-k~@yx=zn4)~NqUe0 z2Tqib4w4svsy|KgHXoyZBma!J6Q~H%MF~vFq+_qoLr(z-2{72d+{472Y>t_=G41q0 zcL!JGa(dw}_&$#86Z){D9%lb1cMq(i~e;};g_d4^?_GROv zXIWt3dAA{z?BoUu7hVaSw1UMUih;0V9ToS{`12lEt%6~bgy{4bs8HA)0n|)T1hd&& z?*Z8xV6&_tI#~kK@1_Ip^#mrL-lyKJdFWjdY1RivT^Vyp0wi|$c>Qwk0`KM29(DQ5 zkH+^X3L=+NtH=St1A-q`2!9t(F@z0#!t#-&4D3eI3*4%qPLBj>zcXH&Y#ZKp*@=fG zl#hB-y)qOLDx%B4VK#`@1m@G-WbY_3-41^Lc8>GlTSjG@)3h_@c*_4Pnr^W#zpIX| zBgI3-Q(QO746&2JOWeH_Q$>*-R9w=VH@|ibWNP*bPxFotV`L()HTLsw6unvXR>{A9 z^7fiHK!7|c1X!)UktM4YI(Lf$#oZMgoVFl zP2ggPjddxZL=v)x5X{GTTbs)V{llRh08`Sb@6`HPCLH7J@QWnfh0AkcE*TU*Euxq$6j+Qn zEX;0=J=t8)2p`ooWrKQx1ubo_pZ*<5 zabdJTAb1d3awrD8SREC&WZZIU={TH6pbF&{F#WU(*OGhDb%KqWRe>8ds|9^f!PFO% z1QjZmphM&yxhVbciZY(MtH|RgIa}gz_%LxaFsz@)4JH@gdi&_hZd9-k9}_b=Re9)P zS(DuCabWJC1td{Zcq#CAu`E9zYZBz}Ooebw@-E=N#Varri&IUVH&)mD!>IiCui~UM z<3=cI2%|Dq@y_Ki(ZHIgV*DS-Y8N&tg;qvo6UA)A#aLXEyqJfjjxDMr0SbNOMHWDy zK66qz@3c7IqeF?kX?LRHMaM~-vrWj;FD_SO^7KyUZgDdvr6E}%0t-B*Zv1WoS;{Ts z=EAF;94j!bqnM8aHAbFp1+e|1EXaY^4U85S)=L93? zBFF0KS9Y&!w2j5OYy^!Zx{Bg)?feFaYM8pGTU9!hsY9clq$-z8!Np6z8!C!HOEu{D z%5pTIG-lR=St!B-JE0gJ%Lb15>0x<;L~i375sO3&-PF0xAoc z{q&Zg!hlu09_1mrSNzfV*4SQBE38y>i90p5^wnw3^IC=)Mh-X{>UO6c`{a+``uX2i z*aUS;bt+^5DyHzysYqNNo3}@$d&6oh?Z$=GUC!Te%klr_-xn?tCsa(x2+S27Ro(XN zRAkAt%b)|XP*f9gPt@+w?lHRNVvOCmpq!6|!JqKY|9;N*@2hM8-4*B{Q5eKVdY)0y zCBNZ~edf6|NMfou7OM_^lf&9}RagYg`DspoWexL)IkJN2hDz(X@Ja==*ayYZN+|}I zfVNO^SkYYVkD5M4;kt-Qb-!DxvNoh(a*OvFaY~>O;_PiQL_=N#I;x41T+NDTgSarL z9U6|-1a{FS{4S^w0R3brRl=*CWJTe_Ti^Um2*Ju!70~iUCdC8iYz0 zovZ0lg6u68)}!pLiTUgEaGh$QXmfCbT6-AU^^7K_ON@~mR-EHqVM-V+CW5x{&++sT zKv_~Ksg)tq+oEaJ{^@?VHTYQ51M{JcNQ-@jK#<{O5LZf5;S&-fuLj?w9XUBT=4%R|Z}sApO_ zVS#&VEYz0_I@UQcC-Q=%CHCf9OMX=J_Vqa@V-Yadr`;!?YIbVYlAPHUQJIccu;k>G z3^mG@W9-8@+%uf!De4W)Z?bHwVp$4JF=c`Y5iCg7Vj`_Yl}pzUJY-Ox7as}%pRoxJ z?oebgrYRtmzYJP)SMnZ8b2R8w6h{CX-X&=y95@j?h&FQWS2Jp})oiqeVpdaR z1@^OoyHcUZqdyAn3hz-OL(8z|hNlC}eajxGa)0W`@8Aa2o4Joyz3ieeIT}-@xD&ht z&+(_!bsUMM)rWiW}30%sUyFH8#o%9N!Lgs zDK5M!Vrd{nUR$gxFwj^1lKETt_oB`LiKsrJL;0!a$yuXqXY5AN$brG>2}(-$Z?;te zE?aNHlJ{?uU6m%p0xyYTYm63YvdsmHCJ|gD9;tCEBj~fQ+wQSOv+?ZhPCLTQV*IlD zwYrzREIJp~o#^DY3p*pXOCKu=C5Euo5be;ZU;zy4Uk zVa`TXCx~Mh6}_HolnL&of%`&QW6$xHdd7=dV;_fnw}Ay z6L9D#ae4eTe7&F?It5!}Q#=on#IVz{t58Xw zPag+qs?%h7NY3nPk3Ra2Z!?`7jYYG^qfMyJnx|nILGeh zxRq@^XB${~mYu+bal#Tl*%(sK$8UR|vMBnB&s{}5zd%_wr;q>eOFjHMd}ptXY=_Kl zT-e^|hKY``p(IhA~kSI5IT z16)9%PRrcIL3`A5``ygS z`qFog$t_dY;ctHH0$Iz=U2D6$U)@364tGKj$AlJT}DCIOpnHL%$ChK;mu$WD99EKm~P z7XOAMxo|aRsTDZ%6a%^GG%7BQrx$!!N@u9i&%FjAn`@DM5UHt>)=xzw7{py-JX11Z zKYu50wV)@e5AqD7;ROwA(1$&g_TR|v`3zB%No7sGT{y)8kpYrxmrXOVuy1cwR75t5 z8|65I;L7@7z-`|B=zYSjX}3V;NB8xB(O4Q5ZiX@r_Cvz1E4kt3PuXiTMiMY~;aQ3$V4OjJqN<1j zGL}FrmaGiG3N7trB_=3f+JY%7SCi}GXs|qQz`RNV5ATE*%Kn7cFvH_0!8=5jWuh9mXcmGD;8=IFc#6W>CX z=Ask}dqI56B>a=A)C;ciFG{ig``m;&S-;zY*iZbg@oGX6d0pZ}LANlGm+tUhS5B}Q z%I*vtb1-aFKWlBLEwH)wy}%=6;W(nVdf#g)W(`GFQ*nDky5JSYgPw^&Zj_m7obzeq znHbHlpZT9}=i7AfyL$2CI9YSo{l;y5H~tIT?qzg|oEIBkIT?-5>E0pQV4rR^E zO+bEa6m}|1_5(vSI*m~V{MSyT`FaLhc0$}@7M{7?X$Ka4n*RPsTo4xy{j#_qkEU(# z?V8@H!4ys&y#N%Su&}>_Zkf4ReSP|NMG`qZDodXyL&^mjPi`88ji7TU7kXLdlAoR@ zb4l_mzHY%t$dYM4AO~GIIMQJ?>Zd3M+Q=HI zxH4szZx^pjGVn81iCMxjNuP<5t7WE{r#3aB5lBaiqgrG8BQt{fVtS=z$`bEQvla%{ zlD%_~xwbVnJ@Dx-;Lc&(3}*?fU1g;MUkr zAzO(ut_vw`;rp$z@lc%Cps14E^3l7m1vQv5pV}c)%WM{4cu;u?8cthmzr+y7IXP3* z!Q(+SwtS>qd=?j;mslu~+dV4BZ}!(q+l6cRpD7;$$E|s5y{y`!U))XF$w#2rjcM!C z6HZNNjjfSYMVqe#K03?2TfFuL_ez~-k4A{CAyDC%L*tiC2?(>8p%tm!+enEEo1r6C zW~iECc2i_06_@k%i=eiU5xiNg-OtbDpCeGp|HhIxim5-%zxIwcFXAY z47w<&-wj(<+r^cT5vsNvgPLChjU6O~l8j)WL8kLzU91<>gzV&R7jGn>iC3V>`Fi6V z{0xu`qKh2Rk2FF=uo;^X!-A8w8MF9I|2Kba!Oxo|2mHvV+*}-CX68Uo{L6r@{b{~)b z&aD0yGflZ}c!97mO`|Ojmx?d=^eJ()Kps_$TN{1Tc%7g-rBfjg7l@3J zkbF$$H;Z+F1w1U*N)nXL$&fr&=mPIW6-thiL(nnT#yDeUa|kR(Y_Fbu0J{ZiJ#_u3 z#A2;7h3hVp#Bl@|nFa+_brh3MkquNFYNBFf6u-|MK+=>30oFTFcRZoe1(Uj%>;ZDx z``%r255I~@rPH>9t3LmORzG#{SuX9K-P*`?H?kF(jQ zU1)ZQJ3-*LI;;?wHCu=2(AWboyWP*BFXRNQfs5Gz_Ma#I8l|*=X?=3Y8M4lW7wFrq zfRjftP(-eWGR)}}bS=TLtfdWN>%G?4^5Dv#xp)vy&qX$&`+jRlld4IzS$!v}+8^(6 z#=JQ2?zweo*$n!Vhd5z_9TY?ORMda}FlWx+*Ge3X!G-8PBL}#Jfn0bpykce3PEt%WMUGQ(1yC5e z+Cv|q)8Gp6GL93FbHeOs&Teh-I>*a3^`f@PmWggCq3uB3CIfFjdBXMpY&bxMBk{q2 zL_xer8y|B*6(6%>Dk%5?TQoioo5^*Wb(1C(f&?0I2Sp-NC?=UAiRMTo>igpK(eIY*We7}F z?p19M7?GCMaEHTQSPb|1!@Z9aEXFAE%CgiJcV%kEUz16m3ookbtzfZ(VxWV28x?nT ze7~C!oCR+|*0f1ZtrP3d)3a(82j25+@0dq6 zzGM-=1TOx$qr;^#T)3U(X}5yKn%72fT_ zm!Ea>tk?Vu$zI_?5wi4Zu@2avu9PekeXiUzbv~A^)Xe+`GRAO2M%RP4-kE5bVZL4T zvtN=;F6^;FS>B+9Ofkh2P$UnE#Dxaoo*=X0l|J+wX{WykPl)bdYiTgCc8`I`LtoxP!)ac%8G)4w~!#`KR9510L&pi*rfUD-*CkZDoBvrkH~i*-ypw zC=-|#h_fN_+@tU^h}I+LcgN&DZ=DKb_fM0|D23Aqt^1D^NOWeTE22;w30>V>&1Rog zRjRUEeutkVfQA>^e1D;+lzb{Kl-zQ^uQ1XnlB!u9lZ~>9C{!&^5|}emJE4JLWJGD) zQTD8PVm&ef0Ny+^{m7dZEIs6B{)iaHk#kmqaExN0cDsR!L&0S%9JmWDL0Bt*l>%7S znyWb;0)+(Yc?(3VJ#K`iM-)nsgejj+lpGdlDSf5zq@3dsWP_Y*mP6Pr~*&uH@nPL(tvWki`7ayabiHWnU zW@e(G0(APE=J>J?gngJ`zu#GB*fFuR;_M;Yb@VIJ1IMCuaX1=l30i$uD0@9yW3S0O zl*NMk(Vsy>%Guec{E`J7quZIy#UpkWJcfMk!fRXTmp%Gsq%$NLlHITGejVJJHd&`4 z$^Am~-Wm0n9W+&!-0(&pXcR4leHLQCu^Dn;oN|!eLJZ*#josAaS0=X1BrDP{{emRC zGHW7GupczXY^Inrima#N3Prcvk(@PE2@;OR$gnww?ydP*ea3)kV`f$tZp!pZxN6z04<9 zkKP8Sucqs|`d_8}_P^}iAMv8ap~zyDM{853n+fIq7D%8{vnbF}go;3%$wxxcH&(xsWQ z8jDF7yem~1A6okMMQH&qf$59c?32x}npLK#)$Ah%UKQOCjU=e}B*(yxJE{gFY{>0F zW@7l8?5LWtb-C(g4|Oz1jmks7iULc~1rziUZ8B`VDyE7lBC#M?uB(1!+qdq@)!(6`mPg&rfoG)Ei6dG6-FAwMFm z4$J22fjqJ`c7^gY<#t8AZz8WQU^K^@!)T9IfYSsBb_+4YL_!IcufHI$Oe9k`ez$=v z9S522!M1rF#e77OBr2|xPLDY5YcvV3o4TRP)mP|h5a-$;yAj?A1s$XL_BqvT+51It z$aS3Fzv~8zl9qpO(yJKq*s3`8dKvGadj_2ms@b3mVf%uWOH4iBTyVl3V`y?#Gg>B1q8Eml@6jbheQB!!Br4Os>u4Qy*!A^=(jp%KWwT7cT- zDnC62HE&HRSuAlMJw(_EgePmzzdaC!m`TPSQvPelPc7hBvU%Yoa*pRU`?eWO>`t)vTYDB6M z*U)AT5)h%If!Y?kb&%EMP{?Wz?P*fR&ls;$VRhv0U_9UNwom=x8l@gOM3bYCmMIko zSLy;D`QDfJVO=FM+@%C|(SKbanh*aaLaBZQ^eQ7gsim-zD?Xn9)rlMU9nd|l`nfHG zz$=oeVc{DXz=KXz|K@6s4$zHi=XWE$afi>sU~B-*r*}s}8(03+YLBkaEP&D#A8naO z8Gyjq&e^a*;$&;)Kd&$LHH%pa`gQeOvVogfa^VQ=4lA>?g<_z#EQ^X;<(Dmu7j2=l zrDb$pU?mU#J{PinO39RaVwC>4I&*_>4&4@Z$JZ!)B&haS<+s7N37C?gWhX3NWB_W^ z3};oRn#X$BPyu`5Otkzke7jLT=f2S=Z)F&-_Q}1@++g=9 zhj1la_bYeqr;9J!7Ph-=?=MRcd#;I)F_Bz@G(UI{$p46r3U*t}!};ZcbOWQUCp&yi zN9N&$I+ZT4RJ@pX-UE+SL?wn5Mcf3=WUB6B<|-Mk?#aJvIO~4H~Sl9E9CVN zSNW%T4fNiL+MToZPF&&LHNglnz0O*FjT~A|LDm?4*l*vo?ZI+c%fiweW(Z4FR(otD z)nR&{D?S@TbTN&73+$5x)#9*O~@Q%S`& z$&sWHI|NaL3|G*(Cce&nlh23;c(t@m!dvb)URypzTkX*V@*8?VV;J6ptXn-$<^V3AiPCu9kXh30-}ZoA)C9{5%n*>e5b)S`paeR za^YYOOA%!Jj6SbR;K?I8V2Z+;C=7u(1^_ni{rS80q2>9{aK_o_gLz-ra%Ztn-ZjaW zMRb{nP}_+vuzac(m9Npc#3t;~ex>%kPbaS}hZqp{Z{HAWBG)z43%V9Xz3j{;DB(h!c*fH@fHlD#+rO{B~>o8j}6W z#N+C$JeCTIfhcYX(&i*n^EXoSH&b`#Ecni9Z}eM`;ki+6F0Zu^>hO?iqKi(BdaS_U zJ+dMot7Ep{Hq)Rmh`Yqq9=#s;v&8$LC@WT{YLFffI2vms#+Z|^*4FsP@7NObaq&l7 z*L;Ei;@}A56Y9QP4H9oP3U~4|$y)GTjkGZsBKx4+tT{z03DU=EQR4~=xX+RNsVRX4 z(gfz-^zO;pEXL@cH-IV#sjdj#m>AseW~k=CiskY3LW8V;j|VIioRnK~DN zs|7V#2Ca5bBd}Y30|137aDWY8$E#H1$|sMm3pnrLK(ve$9|NYu{L>N0Fh7HRzG|-j z?9dDga(;SZW(g^C;rfQ-Rw${X7|=P}L&c^1a~(J{t+6YKo;gD9P4A+YjLYXWD2jb` zUUjmi<62@DzSZ(h`+p|9S^n-9Z|eWO{++XPPR5$4+w>8wv6p;M1*6R8u%bb+D74QL zW+KW!KMLxE{bst_qchfgbpIR!y;r!|r_*%Q5wLMAUM%Qw0BqRtGK8N_?bQ5g-rrXl z#JzvJG$)h4L#VytQ_fUCB=5ZV9Ph3)BN#1HzgxO;BRTiZ%{kfz+5p;5hk%^RKx66w z7@$Z=B;+4~=o-R+nq!)IX64I5$wS{gV7AOuI#8g{C4Z~{dzU#Wm0vKsCkg?&%d=In zfk~Kvam5~CzgwTzZp}k)pyp|SN(#(pG|H}eVB3_)xX^-J9ZDWQ$C&jp_~lS*lBqm&#|SD zbXi!i&_#km9kdT^;8m#;7_-TU4=t3g;>Cwz#S;1~TBMn4Rq0ft$xCZDK%Rb|*+~-)&z0RI`V7f?qx>4Klkg*==uQAg5Hw4aj#zTEYj0uZBd-RPOB}ViWnkFsMfbqWQY-7Gr?wK=3SJQ0e69?B{dd86K%$e}sW`7hk+ z6F9wn*Dc~myt6=H+w1-;ACn9Fpe!t~>4DvJYD5DpXX^qAg7$#)aG&Qk{%!g*;DoJI z7-h~&E?c4G$qol2^5mY;*kFh84}atKyOD5!x-e!~I6w=)%pE7)(k%G9HFkjr_c}F; zOh=Jy=K<&+q$*dA@0B&sP4wJRn0gQa9FSv~fkp#2W5&GUv2=phK#J!)nM;z#k*!u! zdnUyIIY~Md*G!{q4ff2fB!}qU08~TjcgvjAF5IkHGNCu1PJlm`&PWeT6QV+Sj%Ku! zk+{EHc*mpge%Zg4Lu5)^cXs)C+$7;l3q~@OKQ1JdFIfWlgcYCicc1@l%D@mYD zk7yVrUTyt9qO1{(AcF<)7*9`Fg&do|c|zr_Yie;z0eJnv4)^de1>H6I}-9da>6PjTxU6~ zCWfscv3lHPA6dbJdVF&=?P)e7HDCK?8(GU?LjoG4<2v!TQp^?#W+5?8f+Yg_fOeGz zOJR*fD-XO==H`TJFe2LzZtVu|lZs@K4n*gs2_J0HF)f=jz#J=d%>0O)3!VzS3Wq}} zHt1D)B8KDhB~^;U>Rswnq>!#sXGLBXx8HuXERiIK?u zcO2<^#Y{~4%<-N}(o6>UEOnHsjta@;McS^f>C&G8pO>_;F;inI)^?2|`F~2FGhWeDrVoLafr236+8PpgU z$GZguGzTIEmF?bHl3dXqpC_az!Bpb3U%Ky&rhs~3$}EEi$|>JTSScU!{77*)WYw$_ z;dT6H((%8e%)5qaQRKYd6o8{o9NRRqjt8g)3Fbj^Lu&dfZU{4!xu9Za`F9u%-&A2a zW~`j~6G3VlmY(+t{^Lt#2zLJMy&|%Y!;W&B#fo;EVxTXv4%>%vf_q?_ZPKIFs`it5 zA(p=*Bq^u0u;U5Q~AFr*u+$L2y+pl_ughxw)_x`nCk>ETuMC9yshGTUm? zczL0EK8|BLzl%1VJ??i`{*iDf${<`%)`+IyU%S%MOl3pH%F9&(xE;z0W-rh+1Ss6>n+cO=vcOJc71rAu~(tb<~(l7O|)?X`(_5H@8EygLgo z1G{s&pxN^b$p}gkZJ%R0TQ98ieP6jBviCOuW+%ccd`n_lfWvW(=q@>}?g?vizN~<= zKxY3!W#65hx*rm2(kRaE%4`rs)zIZ(qamfFT2fD!#%fhH^h3{;-YKttpvv_rR_FR; z3X14z$!LfH?&wG`-HBa`wUIUS8h5-C?M2+mtLbnS z9?TxG)##3YWM_2V`{*0r_cj}y?>B6}L^jNVzV-3n)}0iiqsVqDQ75_Un-{Q}IUDiW zQF}t2_zAg_a3KM{d(4%$&PJ4k;Lm0WhW@eX^E@ebdgx^0%W-frCJyA}GwddVv&{hi z`2F{OIh6`M4p&9ksL;c93t%Wh%OEJ*)T8wpB-OzH_3v>zt z%JY8f7~n@8P-c>@0JM3?DmtXn?WP|~5BT*!ah0Y<(mlr@G%+|MWfVkUFzvalRgRV< zkN!Y)l_2$jDYBCi4vDOoh5!v-=#=eIw0Ua`Gzvywkop7QjyVx~Sbay?D98&Pb$g(^ z4~8jWZV4;VTfi)b zV!(~xfP%j!#weo<$!CJUqrj!BRi!byXuROUjKsv5VuzQB&%5f3ZM;a@_m0qJb2Iq8 zpfeTJZzxQksbdgTuawjq)1Px^#5mwrI2#$wrK5Zu+hW1!)<2Ydn_Rpl0}$ z#nWgs>_p6q#bSE!^0NF<=AKKm^`3{baTQy;n7Go}7v^WpKVbeje!Gt6OG5D9W}vJ* zE|@+gPPSN>hLRW~+*nx;>hJ%FxBWlNOUjzm^hKn1vJ^Lm-Ho*tDemVKGfa^uR3er; zV6j8Hs+5Ab*_@%+$_U`-%7#75&5$(ys2j=zqy<-zJW$CoUPzbp1Fz2qs#|irpjEZX z(JL{^jUrKduFpB5ITCUt1V`H@YY0e{mr_`JaUS+hNbO}(8PKY(N%#1dQhLxCC{!C@ zgLGO}6_6CW1=LD=!38R%DtMiYes(`OqQ+0Rl%@Pr3$r1CJxmuYZcz6a{o^XdkgSx# zd})Cif4Na*D_fb%>lduyogo9Lp7!2yQKq1f-UPo)tG*${+kVa(lP@flYY%K=vMZze z#;e|bfT`;tlJQQi&qb0PimDTRLYSa_`7QAsT?NB9S?s{UW_qV~FDr0R zAOEvT#W$~HZlP|;B!$D9qn#F;qil*& z%i>GCl(|><8oaVGNUOdDT12o885J)pPT(i84n_?EAvGh5{jp2$z+n|Yxh(2iT zg0SI;;xKegi$6Q8(Oi7pS|Gmt|XgIzv#L7bCUj z)X$`7nMU34ng$)&B7TqBC6IC$M&m){0vxT6`rffQ=2b?WmU@(IpKQOx;b_ZYiRq&yF+TjZl|DEMpI%i{K@ zC+bmbx1!t|T$r_9ht-fnY6tD`Jmk;C65#y=v6U5TAfs-NJu8$g;ti2Q#3&2eC#;>L z(FV0gw#siuHJdsY+ErWMM#7UPOVT9xIPS&spu29=4XIOg0)ueFYdykr$!*_Wdi{@f z(RY8QX(T5VsMVH7uZ+;F{}KMF1B&u6)Q8k$D+_32SYWHNKNzLYAJUEC#nM-B8$HQx zY39N2?mcPUtj%U!<8aNdjn?i4%4kb3V~VdebV-YF0ooKEY1|n*NZV+aZtYC@uj~p8>yzBUHyn0=dh77SS(=e z6w^wPb5x>Hz!So4$s(*9X}9JkR-W5`gRc%rjM<&eLFY?KwY31q@8yyTc0UE?zUe z)@!T(H4nIyM-M9UNsGLjhQ?o0UBolb?SbVUXqYmglcA9QD2&zSZ3;jc@Bi3&Y(J!I zJJFLEpc&6|esI^?vF2>J#)hk`4>(QlGYbnF=mUO71?T-%gA8aoAS+jv8izb*BZ@Sp zm%d8&F3Ow_CT2k|of?NLTt@M2v>uvf)}hGlACv48T}-pL9q)RWFn%T#6?lyK%J zM@arGV2&D}Lav|~phDeEC00s~gl_PDKEFrN#h{w1e%>?B(`v06jrg!LH3K24a9ZD)A4ul^2#ghJf&O=BSxD4t^T%+(ff0tJysf@m}P^-EU-YDouD2=Y~b-w^Ar$l z2M!&-g&l%}tVM-6yAs)u2{zS3U<gdm4o0@7$2C5}b#ub#GLIZ0fMH4@Gvmu%F#FWq7ed5!LGU zo%dh$CCESzM1fenY9#!a-x?2;G}6;sWDlZ}M7`o35KSxe#UAVe^{29w_#S#g=#kJv zp{SaH;XYJan>rR=I%$pr<)y<;`64Tz%oqv##mz69%}vUN@BDxq<*>QAW?^p5Q_Kex zIYlL2VV3)JD;fmD!c^};<#DLcF5=%Kcjc(|s8uz^bbuSP|2uoXuKjlN60K^J} zJeSr&SZDb~^Sj61zWDY<(jgmG7BLu1?ts2X?47IxYBjunM1ec)z6EQQ+uYl{ zikF&pmPgU#t*CeZ^3G zB*KiSyWf6LNs2j)DCp82w}919%mIp2f%p3^W|0oXXx@7l122`KnMAXTHkL@yE%FLq zXgR?yf+D(qfmT%&dmFk2(wJ^pmomlgy7Z6lVtWEUR~5T4NKY5i5cIi9+Euxs59vG! zBqVY}|FSE(l&V>>D_Vnb;{m9_nqs=jh1hWbF`qrc3Oh6M|LM!A)(w-Kb%8&I{LsoJ!Ez|nCnuAgZ8;M18?0m zk12lL9Qc20@4I(k=qWwI>fLiUgZ%3qJ4dZS!CBK`1FZ!hac{~)*3N0?-HUDo_V1$$ zA39Tm++L{IhgN$>bKZTd&^q&*C*v>t((H@eTa+=E+?z#KTV#_zrIqCa=0~g0Yq+Z!x<3a#nf#dl*VdY*ApDRF zni2G=Up)kclLcK1POHyJi+T6vmoDy6JjLW^SnZtAkS^Y6R5~P=c8iv_CtQ?WNEnFC z6g__fA8@5=mz_|t5owt)!%jPKS51Cky*akA2Q>({_|--N@tbmKAwQ4CgYKACIxj5E ztHyI_A-#OjCP9}f!>=Olyt>M_h}lIS=l4T;`h@Iew5z|H4!>Tfy~xQ2Vr9#@`?Y)h zYuA78Yj!)@e|lgU$zjJ9!{J4y+QOg|Q%n&>c2bEZnYmqFSE2Sr7o9HYmUYLNyz2KC zV!vaRq%`)RXNO|BydPw`vX$6>i4~fgp-(}pZjs&b&_j^Mm9Tdh@@QiwN2^oL?s#X` z#?P-B=PuTXjjHkUU%4UG@pjRjjE+Grp{G%)%4BjWtS3en=J(Wl5c=~>lQ z|Bi%QppnLQw_f_L{1E7|n+l41V)n?6EIQz_-S?DG;X-8CHiTjg~PVx zrECwa@}3whYS3i)>+m~4au~sqUbj9P2W7=fOe$?8UJ->jhpcgyYTqf z1J#ND*Szxltm5axq?O&u!(lt~iN(rui(-JZsuMY?R>Wrso=H#Ab&-u>A47nsHn36H zsZ4_Ywga-AUe`mi;*fW$Hc;;YZmjey>6RT450Unm)VNd#6x9aqj=M$H2{r}|%RdTx zLNZAyG%%c`YXw(%wcz-6foxm@uSk8IZ+gn~#7e;}Bb!yHa@Z3`{)9eDHqAR9){$^a z`6#w766NMhe~Ec)7^5|ERNOwh7HTJU+C>=28e=xRVOFQ@z~;GgSJo~uW9;uv{Czn& z!r|oN6${jSKrukj+C(K@@)=d0b8iSX(mNdRyATXzyHL2RgM|loKc^Nx2`z?l7Az0d z^b4SZRE3Rk81BO)G_*YJ<6VCn%3b8;Op7~kEvP#f?AJaNio0zH0Pcg#<4!NEf;hTR zGZd8`ZcJYyj|6%@9RY1~`a{uELdG-w?3Hqds(_~lAa)?a>K5CJ*CeIjGiwUodtYXT zllYnWn*>66^bG?UO*6GNErjgB_h-v zsEd5&T^2JPYn4L~v6@a-d}IZY8CSmft7UO!h-}s?bmRcLRgJ?wPrC)6nkWXOpX#Z^ zRymN*kaUUm&1z*iu;6urUb?BMbJWdL?Ri&T=2HcBNgrVFI^lo9A8gX#o85Er1Fwlq zQJqd%Ka}_72ZD*~rt8709-N;aIKWS1vSN!v@`4IQS#hT4a0Ik$54=$@qfS`vG36Mu zL$I@Rotg?9-M)40lULo=9EGyZ<BbslxRji8UcgIDPLv0OiIuVS4? zlIXstfz*27bpx3WQ^S%g#5CJenM`}1Wo2uqnd&EBwH6rUY!2IoWFSUf29$gGM|mxv zY>jEq4j$x2My24R#r%>&41Stg!4-X2s?DXjPA5A9+HBWxJ|P z+(tu}eu?CXcP~A_N22Iy%2zuw<&(~llTWg;EHmEz=Of!Z%$DU3nv1_9YdE|cg4qAK z-Ov__$)?D9D)C}uw=5~5M2f{PNF4(NGOz?0$>f0o-n5l=-dEj@>0$#7JBD-GE$pCi z`7cL#*41@3ir7qBg}d@%;1k4|z;}<^hG~84rq^WOyLZ|vwR<2|@1DCG$bUPqEYiBV z?hBF*uu&v9pjasf$s?oG05$-$t1$jn=-Uyi7Y!?cLnad#9V&QLCe?NP4j5U?8gT8gEV)E^Cq3 zqcXzH=!?-O=T**I))lkXs|un-=lJ!Z1F@rSrlCJo6w-}h>-c-b8-upZFP_iYB0-T( z^;|IK1?PPCBw5AI3+C>(iWFF^Y?~-%14S~a#1at@d1NaKMeF!zR_h|yf%pYPUW(1? z%G2cPj2SumZ0q#xw8#Fgy5#r1`yDed*8JwkV{&#Dd0_EIuTcy{<}Xo+=rO`q0$qkK zUSl}+?sUpJc%yDQUgf+!;(^#Ms1+FYOqqKc@_hRheOSnQ=Z$OPOhJpHN(^m1C6e>N zuhL6zRprW3{6^gl_^p&{LBIMe%8YFdMjQ6*Cw~6-!)P@}4)p^+!sb@1WlWaE} zJT>NSR>WL2KJ$} zDZuhmdqS_^V8A-T)@T%nL<%|6=g0DV$<;ue^san4WN5IFB!zL|l(2aJldLSRn+4nN zf08^fcfW@E_?S-X3O&xo>}*uyJrEQ|9tIw)G^PeJ(p9mb`Y6u@k4m$Xs#O{2gS0`2 zx74TxVrv#HLn#bo&`k3wi!pe$@{P)CDdG8n2I1Y97Wca`jdNU*C72qJ&DeGUo>>8T zhUSCQZf~2HodxfH2w6A|hwwTrmYA~?bDAQ}RAN=^knH}O;F!0{dyP)GLD&ATzi47rDwOf7l9*NO=z6z%E-;l8@a;~_@nl-u5zZ7z{uo$opK^TPFB&|zm956h!4 z4gFxj18CmQqswNGy7kBP7)4P|xdSXwJn*=}AcUY*Y!B$N*G58jKI_z!Q}8s2$((qK z9q8EMssDTbRAe0zW)tG!t{+cc5OG{RZfZH>C2ekAr!qB<=y@gmn^*Yt)m*#LZ3v$@&zjpm~1?o%*5KV@BrTi(u=ErS5 zX>YU$>%>W-gCMMQ&1DS*lTRB9`!V?dE;^grx-})h?5aFx&ZH9UEONxcBiKVRAicMX zN<=ofyHcHWn9f%>@UUsHURKS!udG-1$}43b&NB$1D?V_WcR3I>!8N)?xH~SHNn>t? zbk7;!ADGvXkRExKZ1qQ;xu%#-SM7&bylKmC>x)mWdDX1Gu-x4Xl{WqHxR)6)|8k8i zC0Xo3{@ne4M)q3ZKA&Q=6xj-eFVL6Nk%pkSNbrU5c#-I1KB7=Z`J!=0&CXOZ9&yD>)o6 z*VIT-nw$Tit1Lvu0Ru&@^u z(+vw&sF=Akt8VUmGb(;to6tZ?zc3ysgkHyC=m^EsP~-rWsNH|TO_x|t?`TN4#=rW+dHF)*MWrcJk)G|i{AJThZf3`B@a9EP45tuHuK`pW- zf`VE2B$h;B#RHBsDnAQ#5a*=l{PX#JzC%vn?lyn%?FQS6eb>tmCbcIHoUFUMB0k&< zpUU2fWTNM=Q8{4&h-!)fp@<49F=bY_ur~0Hctz-{Sq9;#;<2L5`#h{YNa?=Q>kh~n zjkw(i-tbxxogSSLgzU)Y2-3QzF-Z}7=%S!L(k*O}Z&MZ0D@dPj9gs9sg&hDDqCGGU zFzkkAhCA0ZIS~sE%((4GSQ#Fw{cnAL9b-n0*nP=PQqE!IoU=eqJ;l^gSZH} z7WYs6jtg4f(Of1&o(1Z|YSis}YRuCZiz3s>-Izk7Ts=0vI-x4gi4d}|Gq%BD-xXF6 zntAJS`yr_rLW;k9rJZE3Te~>CEh@7BTpq=2qeu>wm_u5FHHYX&v3RiTfZwoJrT}y( z)GZLmUxrPmSpIt)ZN2ekPkq2k|u7P)y$!t>1&YW;~!MR{zPMRqO8dn>rL|jHpM9;MzYd{VS zN#`}Fcn=y8P(@;(&kzj~(WS~emDRFk%TDk{7mSIQYj7uoDdlt!i7i;%0M9f808RZe zOi`onW9lhECkpxN@ca;I;BAyYimeatd2_5e)>N+oOPV9FwGVgW?1v+V4xq7s_sa@v z*Tu$~e?;9675O)$>AZH(3&N`UZm8nu0Ue=I9!Q`9W5q7IIHnSeG8C`6O`lMi1FzPB z7sdV&R$poEF7RIto+*68Y-`@WRlJUraCm=o(!zQiq8Q+{+(#v1nUS9Eb2rnpx5`lu z4cS+U=$b{{z%SROL@Q9BKH!%Xxkr#IGjhi31T~Aasylp>LNLnSBNx;OMVgPXXl+?t zBpMHFCK)poYdlJ+Qs=n24xsvTO$K`mPke$Es%QA8FWAaAgU2n@Et#au0$S1W;jnCq zSx=D+D)9qVu1~$9PJG_)li0IlkX{jyOwu8o1*$%QSpr9F?!-Ud{xf#Je*1@X%AKrU zKXQ_4RDO#a!*i~d#0XVL<-@ha?I z0!jlm26asj#lPLqwXhdXLFY>cVFlfpfAhUf6H1skix3;NRn`0l0*LBU<)U)YQ3#uL z3ok?EeSsQ^9nUZop*ul_8Va&KK~E%ULASD*=%P&vR8{N&KOil1U$!x*J0^`OP?w16 z>9)vv`V8%g#Is|7u|N-ecxjW$mR; z%Ql1}1p$;Sl>k#yDW#3CjOdhQd+259JOi-bJ_J|P^_X$cbWCCT?R%6JNM;s(QQIr?{^(| z7Wepc$D~K}#y~;9*~nTCY;rG+>huJ9Vd)-m9{thlnjJ=trIbYKv-8uy>ut;u{i>$ccn87f^;;eMXnnxnz49AiWbiMEmH*rCvu^-Qz* zY&Tlj4C;h?S)q02ljpzvrL}H|4Z#glKc1-pF9*^gTGiP(DSo}4#`>b__;rE|V1#XP z-zjY(E13>Ozcf3#6WE7G-Nx97o5ME-IWsLX;bqx9C6k(n3Bx(<1iLBWtX$lCPtN|x zT8W5FS0H!y{7~XDKJc@XV!-}vrxMGc1Lcpml^qG27WOIc`vW-)|AM?sS{qp$e%H4) z5UXF>V>%Kp$8^gIMSH}Vq+f1qHsK$4NC})3;2Z@zt6$>4&HwV}L;qkMSDRRF&E24~ ztsr1SKu$R12W7@UO12W|C`&||qv74sJe5-lPFTHb7PHT4+^ODd+thtUGa~z;*^|{i z5IrXsW`O|f_^(|z#TY1ZgG$^lzASDL9u*Yv?=VGyy@6YpB6>ebW4e^3vf8*ayc~6A z_!FR7x)gI28fA3eYedcTQha2PBzrCGNWezU5ryf_`=0wl?j$U|%UkE&Djxu+cz4)o zU_Pr1yX(6lvO#)i((8~Ri7{&u2+eZUf!R1i9Wz%bW>Lw|b=B_QD zCEx$GSDo*+I=*0WKaKs5!|~gFPtrZY0sbQ(jmmVdSp@r&sWih+u!ZB*jYe!-4} z)q)$+j)V-pUReS2UmS3WR}VC!;Lya?lOixIl{}C&lkVmXk^=f38F6b0C{nxLsMrq$ zi#5-l*__fnw(UvYi4|FqW+x^!yT6>2ak%2(j0GTSDdrGGs;NYjgjy}A6E}u~qTRd( z-Y0&a_+6rV)W9Iy&dXNr7VTi_B8!7NA<~x?jRmtuNvp7%9*95e-Y#zpKO0dyr+fZ@ z5^hHUyn_){K$y8kaEq@w621z=tWbdp)iyH)mw3kZRq&FX03`>pVga0qAEpEo;CMB^=f!WGBc>TA?uxB?=ZFTW_#hL+S8Z+#*k>9RrSGk6e z+euS1^|`Pbf!)+_Htgc}>`xyBnK4s3`hx>R_l2pbth0cAImJLdLkX2QpiJlO1kMzL zFzxMq-!%!dmWP>2ws8a20kKCUjLQkWD}MyCGigk|cb95)^lC6a>%E>y5$ctpAE>hY zYh&JDm>${hd(_tnCddk2@X;B_pE&B&&-}}Gznb={FH?QKEy82418Pq={I@rx7ZN@I zH$}%QjjrS!iQDAAap9^3!^%e3r9bq%v#=v!MA7744?%(?(FHl)w=rl)c3P1dziPog z;D)(MP#8i_XA0Jk8&c$0FOdxL^3{!k>NmP&6^hnKXTif6h_G0urVa+<2&`-&b;Gml zzpM-U*=&)xn*<{^+5k{t2CJqXEr1FbJ&iTp(^n>JF~{~H$l{S{`7posdxLIc%h0F4 zc#kA;*!2Y>opF85nG};ok@ta`+jBsvpZ!Tps~n1B!A6{fg~VCkd{&cS&m^-N&`CG5 zG7~d{_i22l5=r5(nXnN_L7|Q=@UcJk)132vNGF4%odj~du;H05dDU?fZ4Zb{Yst8W zrr|K=o2su`7aFixJvba-vr(3CH2mNTWscTBpprxlDp*F2E}ZV@n*A>6Gyq}zuV_%M zoMmk*I9q*U9vBP`}$+yS!LF*`Ve)^uF@6Dg~WyIcZ|M*Fmj$>9!2<88_ zE^DsYt8kO3t4Se;y$TRY9_Ll;rx@Uu-3#eT(!uQW9#CqKI2Oz6flC@UsAC!9QMWUs zECyTtfJYem{nmPEa-kS+InNN42`m|V6=gBFHNoYijz;09Q8!aFpyp!CT^@4zJJX8T ziyMnwh|LeqLli6A%)D1|CCu8Kd_mS28^&j?*9kQsXKbSaQB_=D@P8U)yqI45!vaLk zLx>$C{rl-T%dK4x&aN^xoQYQXQSlxz%2Rd9n%ui(qi&}Ergw&oFWbIpwqstLd<#2U z(JmYwwPv$_r80XqIO{hBJ`Ow=Vfwcn)L@?Y_X}#g?kUoF9SIlx`<0i(`j=OtIC)m= z_dQ`barWX9?I=V%%B2=qt4sDmB^u@MdSR0Z=ShEBI zk&#-}$Dyk{fh06ZR42{|N@H})1N}iT~oOCLmkY@T?WR*he`dWAkeA$ho>2*S8tiP1GS7{D) zY|aY5Og7B|0+8_$g(8Xp1_~Y0ZMH>Tm+hi^6i>;Wg!3dmEUtk=0K-%UT~x6w&&yV#(!c*9mpyTwE~{R69SDtEtVPcQiEer5(r zna6^;Vj~s^?xUDqirfc|J&-)tszwwC-BJAyvTfpxlJv+DZ>V6>2IcsnoECCo zt%}?b*`?AT{f&t+Yt*e*KCDcSTrbe7hm~;m29ROPqjllPZ~kCGIS+W+P(nRdHmn4N zboBr~Kk%sUurfdJW;8x&0G_NIqDflNDo4HD?Y_u^(*K}&l(gf9HZkZi_u^*xPezJMGi^5+KMaE@$Z7wSHlS($A6(%eHUhY{eS!$ zt@lATwi}h;JQ|};=qMOv=J9oH;ne`0Z?E{AwD}Fqeq#fmfj%c~iPY4IYsi5JEvQ|e zl=ix~Qhf@OG5*~-{g~Bca5w$c@K0|&{Oj^ zV9#*2GG9HqK&w6+Qcs^(w?L%^+%=#`5}gWmY4b3?AQrFJ%3Qnl^?pI@hv@#zg1;_N zKwK!HP`-z549k-o1wUxQiD|VTZ1C+`L|cx2EU0S0S{(fgvT3%V@%}8n*MHq>RRJ4b zYv5Hb9E?5rdJ{B_fyQTUxD<>_;A23ih+>J4k{6iMk%ESrCJmK1X55@Rcw-I zkmO0W$$*yw{RV!VQ>HXsHOV-VDIKGhq~jepVwwwuk{o@lej z&N%z_$G?kK#@!;@rg2)?O3$dO3}6+`fslA2ev@>>JY=cS!{)-=oY6&2z8k=b*dZ zHjXB#KH#tq0)k89>S5CG!~P+G!q)VT~;dpiTPW1wN*7(3^fT!+K{k(Mjd7eX+3(LlLBX0RzfX zem}i*Wyn&XmgaAsi+0A?b8#ZPlx`EBR*WMtL5bO3s6lxU@iFeT^WZ3HR~6BZ zWG%uj#dXnssaA!9ItLPC)8<5Ij1B3OpSI0Trs}Id{)*Y`lm+u1l6rP#hr@8XZeey> zC8My6-x^sNG3wS0n|7_LHL`XN22z0Xv7c^p zPXm|?z_LGZ0rR?G2hKfNx#Jnll!r65hM$8o=gAVg<{@-)bL zT?*<_=FyiS1<>xDCFz3Z<59P4I5VKE;%5Y5dK@LIO5R})L??o6Qln%($r^wMPv2my zxo*B7!8aR}Jw^)^d}?8>3W}}!q2Q#B?(;t{*9Vk{8pAau5}a?WV;OZ@L3YsTz?QFR zm2c)P7xmCcn2*|SARLFI;28b#242#4pD$_rhGvMchs8_h%Xyf2>91jZ@m!JQu*htN zep~h0@5wq2mp+wP7?y1mlY`DzVm)+(R>V9MY*=_ewnp@5!CBDZ!s=SBDh1Az^ShL& z19&6kn7VfQw4>NgIO~~n@B!N^(j}lO>>7L~)6QOg5tBXc|rbP4AWKT z8G=4XAFTmKC0?EQ0)r~0a0(tM;h%jAr9nzOmSKe-CXt-i4!{a*FY`vmU2_Hm-6NJS zwpNn-g7_*nbSJTBydvfztwDQ~U;f-)Nyt@B&S zL&(5gfd-*gbwk>>Af2a&x>!VEy9(0Kc!M4ijs4Jrrd8D|Qs$YwT&KA%&IXS&fP-U1 z#3}z|&3MlSJRIIT*$~7;%yi07b|p`OErqons|EFRMiB0=r#Zh2@851*IBUY~9H-;V zg~l-Rnlvx(FIl9B!_F*dS&xh5AE1~j3K)eGP1W`t3E9f)yh}k@l9dsoVa?K$bes20 zsh&=k40#sub7dU~S+FW)E-H*R2(QNtk*g$Kk}@lu*ATF3)^laD*WmpA;A~~RPzxLw zX9VSQaU8Ac{+mbUIn6kj=EQC&j9Wh@KeTg>?YxbZGei9_OxkMA_{mu}*9M1Yzo36{ zvS^=gqjaD5P|Vdi?F;vk!>aWj-7&2YSuvIOb|l(xtLzk{_h>p+t@)ysP|(M%(kXGdUh^3%+CMi zaF+41g=IQJG2qdh7_&^ESPE5d*~+2VW=Yw?O0X(*VLO;iL9hH{`5BTAgveWgi$`M+ z?&a%v+MvpLg|u1v7P+qPMYjThQ*Xo$Q13wBM^EpAl?_^q>gfBPNg~tFJjtjVGCyGl zHKw!5=4$%mdZ1^hRlY`)EJ_mf(pEh|cF(i(%+=&Z49?p9w~W6rJdJn9%Z#Rfxki?f zEOuz(a1shiy2h<@`4j_Gkz1+6r@Y;wj)ZzyPfWKWLzx$F20ULN%vWNRw>=Whgmx*1 zl?}X;i(5QPM0fZ=)D*gXTADu`!pQi^bkT82+Kgj%wzF1aUJnZB{_t3#Rh3ECB=q|p z1R1}Ff+SG`57lw!w@hp;s6H_t$zhXGW??e&C|7o(}ScV zVJ{!g^%HDbYK!b-po`QMpaQ?g@opRkaoO8+_4!!F9Di6Fd z5CX0S?u~0b#%S8AVzGGNC64P$CdFf4OLg^oU2?%0)fmO`ccd2BjlzR~A2x(__w1 zoEJJxxUw5gFT7A!fM?>EsZDNkera}7?=8xhOYU(vz_NPgc*pZoiW#Ip|1j}3Xtban z*T=x$)lGK;XF(o_rRyYJpn-9eplX+~patUmu{Db>$}Y;fptqn}xlOJyh5q|Nv1veg zIINp)hJ-VakoG7t-HW0d!;Oy)f*Y$DqcXt4ToVInKRKe-^hB9t3(t`{o+%-P%Han*nX!TB6Y*MzVuz)tG~Z^#F4)2j05a$ddQkxaAUwQC-j;Y4y% zsFtV_SF5oq;ue8FenX@o%0T1zm&lM31c83e9!)P`<zxkYxzqaz_*f=?b-U-R!$4k&JehgHV!1%WfVv4azbb4;}saN0{T;2afGi6)*1STNA9 zHhR^avQ5$Z!!%vYQyv!Hb;@v&Nkx*NF>f2$f6(7HuZ^8Dv}MCyE$%~}M}Uq9TO%Dt zhOL=%7Gi9T@96peevkfAoEb5j^$Hz1zz#7SjtI9~Af}08fSA3WN?b!qWk`Wq6>AU{ zgA%ie6!nB`hYZ3`v_~|H$;5jDJ zr}_Ue1L;3{{^Jh$fWw~7eG6z@r5MN_T%;1QzzUn7v9M+M4V3FZZ(CC$DiO7~L%jd8 zxPjN|eP!`5|AEJ{Wy=}_$Ncbr8VK%->msw2>pXhtvFCP0YZ`c+%3j_bC8QKo%TAFB z-&T44oO9CU@dLqk<{j}yzV06S&s#-X<4qiBng@anky)`V(7BIpa~+MOXqxA+$hFGB ze2pbBR)!QTuA-};4tUhf;9>A6rCw^STnZQq^i2HlGCZ+!lzvFo#{YG-PGW1EDTi}E3UQ!?0ZPcwRk@Q5h$}a>Lg1tjaf`0Un%;4p0K(Ry3L=*DjjHB?d@!aE2HvD3d z88x1dPiB*9cBtX7t9-!%HO&+QRXYvPd8u43Du(oR3TVkdoaR$SA-y^pX)}vx4FbjJ z5@Y)TYKWvkt{Ztdd&L6)D(HUciPFIyKQ&HEvSJ5fVL8G_P)1b`Rth|6R20rurjxti zL7##B9WY{=K4it#f&B0ZHS%X-n~cNmkJFI!0#eooHlXr3r)z(iG-OTpz$O6AVIAdu9i2or z&#^KxGc^alKWt4pWP`1z!q*_I6W@jgG0g*SbmFHo^6=Rku;;yhcK7rhv3u{F^+oQyhss}~|3(&DzpYp|B32`U*-pB)12xX?_qfnxy+m%_k$a8|IGNyRLEFxzZE{%-fr zenGZz*nk|eFd(HA1C(V&$jpGPA{*Y$1tJ&784p0zDo+CCN>_Q9IEF*roNWG)kVa*R zq+QhwnPiM>UGy)1U4yE52q~zIGM?1WUI~ORE^$a$z{Q3!ag8})1(z9ff?oTt*UWIa z`)}9&L{i!9ayT41(_4Tmhhm^LZv&NxI~0hQCM_@sdt&xUQFjOtlSg+jP{DWG7f+&4 z^YmFCwn(xz5U%=s>Dy;DLNhkrzO?xmd7m9NIJ^ViWr2-cirGw&ja1@ZpF$d&&rmMu zD(O<*e&_0v$4fLia8XYE1Sl6Ihm3WfUCKYZ#H36%kz57Gv3I|f^4FCPMoWP)VokfM zOZhzNanzCURnb_5(l`f_F+))B`_${Hi(OX>gs_8##iK6$F#pbj^3@jRbqU`KvKewFR7yqWHbcyKw>(eU(q0}STf9i_N8k6TCa2J zmHwB-z4Vqvt7kWolW-Ns>y~9g<#x@Y3w~P`Z4ZOU8?~Hz_YaPAGO|O7O;_dfF0e)v zD?f$$)3^2}P9+-4VL!!2G`7KKXH@W8uz_WMwUQbgA*f+8hJ zCbLJ}rL3b5tMKYJ_kCg%Ai*p$=7lf%Kk(D4bkWDXN8Tug4r?3_d}d(y)f=C9%R(Ust<@1|7P0^e5$OC?R6}^xo?V@wQPG(2r zn75SGVJmo732y0ks_TTPjWh*+$wgo?8?=|f<>lA@JN8GnUv-kZ)qZF!Mj&XfK^Cb7 z|2)4P^a_FvPWRO%Mjojt%-qZ0-(*+W%=Ki?yXqXP#cF2Czwqx``^FPHsX4qovY{oB z?UCZQ)5}P)kq(hg`u&~ zu2o*BK1Ciujkm$8mH#>Z4I2yec|I}Q>o2ojwjni1OZJTK#<{)o*HsaRyxiCyP#4(_ zZmupG5rQ?^gHQ>XDnCc2bs1y9b?4fY(gZ~D|1-Bc%4~n~g9gBOJqjQ3vkR!A89<^~#h3#@3lkAlfKfpgSyD9+b;g8Hd z5P>a?d30wadbBzodZ;b@%i>4EHt(f+`miLI*H5P^Gn7j^5-JqU^jcn01U`>OsuPIj zodO%a+CmW8^1?eCKQ*$kis%e&to*Vt{LIPCR2wSZeM#Xyr%4>mYzRhU;Y zWt1RwU}RziY0qu`O$5DW%@N@s{h>swLJr@t%{8W^x|G-hJL-lNGWxI$aat7=$iQfq z#TBrv`uOb(zNVSZxorlzETS$@!(-N>y zUK1aR)6?B^M&0(4GzQjPH4+L8lFf6IA!Y>42Yv3>$d<1@_1v!j=3vCI=}I>w#c=uT zU?xe4xdNSR*FU`98o(*P{|RNSvoq0`=J3YchMLY5X1PyWz^ zsbZ0>#MJmv*m)fd$3Dqr^|BGg0l#(?=EAd;D3zjB)dppFVMDA|-2sjfRCtDi3Bi(0 z4D1~6J1RKuXDY}vbs3y++@{D)Dlr#Qf|_n}V4hahxCleBHS{_rRk;;Ppq2}6fO<@B=tEj#y7YbmwkMav zt+{?{J@HZjt;r(2;^ffoIoLVY7 zoyUp8L2xC(1l1W*M_&)A0rJ)~;E~ZqGAL#pMN(0&5quAm z@`PrL@qhY4EuT$emg$VBEgOJjnJ4T}WWa+{qg;jWP+%Pm;?1}LVFBE3}N zaQsHTCiRUCub+AgMcnEj8-d|K<3{*XW&mP=`T&Dhtw*o858}xOJn|v;wMoz|JQ08) z=vLr^9g=}WPZh+Ip&YPF*`@kWj9ehN&Ci?%{4*XWX((3C35T8ZyuAt#Ae;kQaI2zq zjEP;wR1#!dsrB>PRRbs@20?5j?5h*k3kIaM9_Q5?qBR|4kN7ArCtQQXN5}v;>UJcg zPj+8@0RDt*rI9EJ3l*_-DoHfzhP)`S_G2A@?ltp9&QXj`v{lYJWBT3~WXCr#)LO#}hY5Ij_0okuDYnAJ1cue&ko-C1^_Q7H>x@4L4WI^QmobL&D7iQ*XF?o9sW92%HM0*qIM@P-BIu86TbWsJ2df zan|(MpmrGi#tjhW!XVzkpP{AM%de+v|KFOIrFwzBhtIJ~I>mBDdB8M`S4YTFB`#Qln7_a};vecR{}DE%4@C=`9{ zyD>10ISF3ResNwRz_%Md+H}Ut##(u z&=hevMRD8$gfCR-c8kt1uwCkb3NhFwRrr=`QF;iyMM&6|RfU15sT^w^>r1xmxt(vvRXE_hAZ}-*#$29)CKLo@K znY-Q(BD!F;!phavv#MTO;k9N~sytKB?e#DUxu|u#)w9xkig|qthGc7IH4#1CCO)m$ z9<<$ionW5|We#zx-9*rX?nt;D)lBQ4UPw>hBHMiQU?GcRD(!Q$oDG?mS*5XS_P~af zWup4Nb*al*`iHZAvkjS_R(ZcPJGzJ7sw%=IE+?#r3_uVW#Vv7*m91gR5e7^pMYLK`$7A zo4#cp59txNBf=3kJ%7aQu>$!S|FVLrT(Tk}nZln~(ptJ?Q+$tl$Y+3`6me7l2@BWK z-8h6Ev!Spb3+(T-jh^c(7W!Kg5U}xbIE*11bOrYVOGP7YW$^Dw(J@gol$@3XY?EF0 zZBNL9>^puMB-ec}Bw(oz2A=C_d?L%|-u&E9L>oS7`ZVae&DKO?tF`&DxqY_b12SfA78D8iMK|mxh&yKjG)I24v{`~~IBebfV@DCx8 zmG*|Fh~DnIF{ll~d<9UU(L)!i3w@pD!%p(nr!@7GjOVOVwr$8HB^)ysChVNRq{has za5$M&Vgajd6qAESF!8ADcoxV(_;n3eUa^*=W&lr%OR`PWkzHzujE_i5X-#<4WE!x91gMc zSgaS9Ddr+YKBN*+fZ*Qz3OZe~{F{YzHvbrO{%>5g&M${&su&&Mqoac53@x5_7Pil> z6?7!vrWi%QP#Ca1A)DXrH{#YD{t%#a%>$3Rf_Hk2xYfknd!ts67kVK9buBZ3R`NCm zwkMz#_nn0DutK^)uoIHrNegs=O`hA_w|Un^uJc1(?m~5O2>zbI`4HceG}AXj2Ka0M zYA?247>=nNus|g%Y*Er2vAcCUFlSv{8?C<`es>af@=Iee8e0{c#w11b({~`GmlNIs zNr8)`LRrhOnOPPud-7Y^Kx6U|TyvHkGU>V8z!sI2ZEsRt##Q?H) zQi&Tq2Vytzu|fAKDf3F_^@i$w^CiPXzsT6GpvL@EX~-ZHezx~PkM*wD=}2vL%xknZsA1#P$GA`Nz8 zc6isT3t<-wqPVScC~^bNM(mRNC@jkdO7NcYI*8`usB4iKLB|CnZm1=w!CWIUj@L!r z_U%?w(LIm`D1i*7Q(-YB99RLvnXq7GVP<}lE>C#XtMRafabLgy? z6IT1WsYv|Mf-Rx_#Ml(kj>&C~t}Ma9zEIZg$|#EEnqCKJbU<{5S{8HovB zZM}WwtZ}iiAx5qTQ`sHnlWQd1bF!7Gk!6AV0-)ATgIxvNeS3gP;z2|U*b9&q#Ohki z0BW#8L2ncubRk$A2MQV!XF|hrELK37@t6Bwer9b?*u+^l9NxE43tSu60PFzS&}9pa z2tnJ!u7qOYu_XKFKx@E<@6ml?dMZ_KN>!MF$Y6DRq1MhPooJ^DGD5xwU z;QUXAgR50>X2;eMRa*{6Yuhax+ZU>~$Wwh}K>;wRL6_xr-^1#%m~-l9QUm?Kdo2`6 zX;A#Fh+ai5Lm@8|+KEaenaFyBM16Q~I#djs#N1Fk1sVVfLCvRjPWe3S+(mZ?yP?eT zQyA?*Qk&us>>17XJQR$1?sm|x7p`^*lIKSyB~ zpsAJ}pR+~wAW9Fk2&U4Zcb)A4j(?-}O@_ty@Usu&2XgW;tI6Q*dL15l^OnMFIr3)z z@;4;=3*(7YS=gCEirGPtJgAlP*cxAYBRRJ}S*CIP*!=4%xA;7Z9Vb5ZoPGIVNhMLLX(Wp^3ELOr*$+gQ7SLdFUshjEf zspitF&PxBl@6r5W<&k+Qez;Y^5|v^puyP#mJ2KBe7elG-u%|26ya`Jc8)Qs)*o7xp zA!CLntJ~{Uw;^qdT7-{+8+h%JZFDmT?R*&Ck+46ci?{TN^aDA5T-u(1h4z~z>zHj} z_rvPK+UUoM{UOctQoQH(?P_@4MoibiIab)J;T|C$#VYQ z2@VLGv>$g}aALzFWbx9| z7wIk_>BiEuI&m*#__SVY)oVnR^Pse#iQu+o$isADBhd#G&dY+(N(%_Yb*R=xkGSQ8 zmxxN~R5@}iU<{@UQiUC=HjtPDuUFYjN_j<4-CyKD0%D3FnSdHwC)gszwxOA9p{f50 z)o0yO_)0aO+Grf?`SuM+_MMZpDEg&XSbfKHFFz032MsY;R(oC4rAU^gsF5AKB&ZH5 zB29PHiLpF;?72GeKE-jrI;W=rU(D!KA@pJ(Cm+~`&`)=Kq>`B}kocka*Cdm}7HE%! z1=3Lru>a;#iA`bY5>OdcURVT+OR@;9LRCx;t*MIDgHXj*l}R_COPS8oMW0ruMPpf{ zYq!Vd>2nMYPM-TohO-TaNi52YWPO;-dY4>iF%z39W+O$is6;*el<0WrJZ#;>ses}P zWMsSVd0`Q%k!J)o1iNlOG1>poY>JRy{Q5sM|Min!{q%SL{(|InGSfWC<1RbJZUQ*F zclht`kEXt9HXb$ee!rUR=CJWNWdYg46myUQcQz3dex`n8O^P5_c1l#BF7`%&;QoWT3Y=H)Xi9g7r+*TdEy=GkV-Y-I7TP zhtUJg>*JW?vMFXgMKY*F6qLCPKI;Hz)o2VroN|nx%xr~hF%+deA=g~mdogjkSm0vf z!!A6-3Kui@TSpq+HsfN!yB``zEr)RdgiYgcah76EQ>2+nyiA5XlV8gbBl-Hn*GAlQ zOtt71$qP*fj=*|hGu=pbL1Dpxh|1_sU535Z%eFl($0ZW3i}|cM}4=&60lFbX!LPw!P=k z#UTdq3EsW=%?=hKu)vdrAz|^sA=I(L6V;q^HR?-dJazugsKg>ckpN3ndg-f5$RQo{+$hmx#|*~a_C31*vIC|FOj67W0A5yXQ;6m;T}mJH z>`)ZV?UXey7vmmOFQP@63Fi&MkBAbY|{M=T2^C9~E&2p@Ir%0A-O?At0i%Sya@jU{O$9*{#JQ zpi)%$o+pWxL{80tgfBW@{Z-B$Jnuix`@GNc|CdPcwiXFc%(Y0?E*x_0bxf2Od6anT zpL=7@I_W88gQ`k_2Y`zku0@|aPnhd8emBp?pc@z9@TNC6kKD3QeIzfS5MBj(6&Sa` z<;M75$GCrAxY2@FIg!VY`(94pU;F*~ZFs5t`u`dY%lF#O&~*P;p?^DW;)pj928f!% zET2L@#8sSD)e05$IEmpTk`{W#v(c@e*R1|h)+$)(x=ETBQ7?Vyhsoz=bpr_b)yr~4 zsGQU)Xp(I7EL(PGiB@qElqGw}E;Uxt?NxO;7m2XREKy#m#ER@3N~gvsU>$i0c$@JP zv7Qd6#%gjs-Rn^=D3J{Un|%2~BplSKuZ1>A4vV+=ZA zw$JU-Kc!pR4weY_8wBl4D*1^bj=%QxtKvVm$#tL&I0$p+Eh)PBvC923+r#HJTcVL;YAbd;5IpVOOlIYP_wZ8QdV3`xw z@Y8uckd*w=vyMD8vxh|OQo~uTB6{kn;N9fQrXGDYQ@F)v?bHNOHw`s15ogIGfiV;B#J3e>#%*4WH?l)! zj3^J`D7+IO;2HNIPUiZfYQ%%qnL}aVrvn7pThaldY27Rr05tvuP zRd|cMhJW1ejG$TF&+8IgShB;TPnxCrYlROw(kZ@uWW7`uUK@IP&ZWRCI-PgivBC$* zRk7bXR=F;)I$}pavl>KARhQ`OpjH8LIb4azi)fOZlU9@4eXY{G_I#k|f4=zRinoj? zTJ~}OztJbS`Pz1z`s_Br+(m+EBIrgU5=*pzpGk6R3ef*+%pOW3qbSw>2{qhP?n@3|< zD;cYVaS>Gbl&a#v!@=!T-)1ma?;Q#}G(VksZ_<_wT!;H$rwd%ht$-Oo$+iS{g-`I+?V!dLTV^K0T2Ol8IhVd!R{Y<_d20VVu2T9Yw|UBb7p3$%lSSQq zf*B;}J72Xs8CspXNX&ca_0D^wL-Jj&MO2$dh6;t)QB6G)wz|l}0abzd-l!{}Sr3vr z(26~~E&yk!{LnZ)>fwUmMWP&4+lHKR&Vi64M3ms6xNK{NRwz{J2G$~%R8fz=R&n}m zoES4z_8eX9a(XM>*j^1}fRc4@5JZFVefmYdsr zqh0#hrwa<`qGv3mbjHL>t|6F8fXY4YRYUYjV7%S&bx~TE($B^`Wg~lKMEm$7MuZNA7{_ z$D?a8poG_I71!k1)a3~Tp}!hn4oESZyuN1ltH+Ofi`_fAczhtv+)ZZjM)xj@3cd_v zExDqs&_0+DKB7xxXFRcps%*{~D3q$4)#+U6(lnWMzueyUxL+UV8`@PkQIR#USPk!-?I%4r1!3 zd(N6?x9oOp4P}3$!hf3xW+OqbBO(z4EfPp@Ca7AJn26kITWwjkv9M(H+3vOFarV(4 z!-TzXwsif$XT1dqxufFv1z`|n+rzu;@})i}uBjGa>gn&2~Bgas3jvd@-alogoT8 zEV~M=O~CE$_SG}$mmnAKK~)_8(XDu~_3Vo!XkfJ_Po7{MU9KlMSrg*RA7}i2+9T}z z*X6&`>+Bdi*(TUYA{g+QHWQH;tLlcr;qI`iX>wZdX>Phfe7en}E38C}^3apg zIW+1$ad`cs?y~(Rr`K=40R{12ua%jT@7nFcmxWHZTyz2IU2>@6;FZAoe^qcMa1V%b z7KqbDm&I%8E23eCYwkUsy&zIk;Fd-v1rB&(r!B@>Q~m2fX0gJlL+}t{s};^?Jol>3 zd#@A%L$3N(NP*j4Rc2^P=sBn-#~5n{vE`*^b&PLRaDwkAE-?(oX5(JYerC0Q`;A6BIt3H^q zkxrcob&R8fn|lZ*pP+My$Vas1rxeu^Jv(i#5J}7r2gFiXqn+uy!S&FLgEO@X^esmm zFu{tqV7fk6*z4G+$Rw*kx4|IojYS_e@?$JR%-0)lTcGGULCjx*`{SQ=iuVp^1wSyi zYuWNen&ra$@F8BJyg_goKwBQs6sA>-_#T>HE8ZY%6?8yZoCfKbF{h)s<=;S!OwS1F zk){B{U13;<0Gg88<@+SNfFPtQ@UE~$f~)O(!m7(^YvB1Y+N}s|Zg~FVPnUI@v+h1q zX+IVwUQ~p@sj?gt1A2ghxe*9wCenO1{sFQGC*yzoxbI}g$-TEybqeF;qdxi0Av%33 zU1c))6cG$iZsZY>gW)^94hL%$>EtKEM0u}hgSb1mTCiWd626E7^I{g{NeZFp-f&g5 zpxC2q4#>O+@#+k}Eb<7o(Y;xHRi5ArT8Q(Lf)W;-3SL3=NNwTISTqcezmF^%<9^xf zH?j}I7!Hg5wv7Gu!!S3I5i{w6E!XMT=caBF+^kVaWi7#^5cD=8vPH6jf52aZx#d=f z)`2X5!QOxX-$h=ZbE6_J6la}W(Loh@V*R%AtIVJ^E}WY^!xp;^m*%=X>)L1_5bjgq zGZn+`=huC0~n~8fPxO9iJ(UO#Z>-Dqr20Kpkz`@jBAGTA^$3$QV}-E_L1$cTov0G5~D_wBRWRiY1>V1=xBQB;RUG? z9Vgy+^9udZb2HEEHNnzOf&tfdI}wRuXltE-n?b8WHcec!qo!G!XvE=2U_Njv>}1;I zrxf|TOMzRVz%?echH8;pUB`7y3~+?hBd2Q2XIb;$s?^H=F%PNPZRnmQ)|X8tPL{PH zieL^AbUCsIJ)&JmlsBuZ0y7pPF+wJ+bFaeuawKq#u!?Snd<^pZWCtyu-zvB&?uEXu zjKzaAbl#|wsUv}nZrN0P_7_RZJMO$vynHbIF~4Y<0U&b>11rf60t@P=m{Be(c%lXNa6qaVTzHLw1FX|XVrUnGxPSfR|LKS-um`lgE-6u*0P1PS>Q za_-~l0WiMVk=v$_eTbweAVpF2OiS&ejG59gGM`@^>~5Bb*l0+0UZb83=e z|0!e)YrBj^|+a({500%K}O2bZTT z4$FqTtD#ES7UTZ)gJj{?_3IyEgR{>GDsm%Ue0jPN6@I_2c#qyTRvCdEmrO(T>?ogR zH^D$l=uRRsn&04tlJmfmnII}(c#=t3blW*r+3OzTj-9{#0wBbwm-di#vPzec5UpaH z`$kVq1LQJ0Bn{GJ_tTKO>>(SSteV>~L7|1XX0>!C53}M9JFb_VTDDI7$kqM(%Z>D= z=Vp;yU;?64g4s^cNkk-28UjQ#*xcQy(5sVZRY=3DRSf}_s(kK*tY^Lm&*FkE-=7`KkF7^Gv1vSmwh^fkUqPobVO#Hb%c;sIbOO+gXUQ zU$W(6Klc5$I?8T0-v8yLLrSCFSeMwhmoBj5CGtrVOHe^Dpbt?-M5a+Xx1#}ZZq*Ux zWGycvs9n*l{#?HG#dKbbW2`7+aVpshdB=2Kn@hTHlXy_PRaDFSB5WhSfjsMWCa{`p zR_}A#O;yE z<0E32W8mIm3r8zx_4Sa*lkw%Ojegdh%s7C|)Xa>rXC-{;AESuaS@9j?gtYcIhwjj) zo*PH@rU`5=6U;?|ZXzNpoQ}=8$eaeHs%A!0O4rL!J#B-$OgijK* z9PxcpgS56td2}dnb*QGwC!1WgNbd)uuC+dafNf={bsil7E7g@zVsx)w_VX9!Rp<|Ws#FRP2p>{aFoAxL~L=A z(^NI(KazJ}F~UN$$#+{zIt1{+owE}+pD+Ri0r(6p91sN zV=nHC9Y^I@8kLvNj}LF6_S5SkP&*MdBeaSFXj9hDOAXRQc^`i*Bo8MM>lRuU<6Q6* zTw3w;ykGy}ztlz_r6Y7^GJVjF7tI$;ypEFuQ%BH8!0TX!$Rc@xTQ&XFRwMAL6^T+p zJJq>w-+BA;H&6fj!*_RoYxtir5tU9q5EqFmltW~JxKunuHYjvtA3s0bkd&$hF8v$x zN@Ufby@{!*r0L~A^53Z5MKwalV555YwW^m|e~t>%S>#~I$3TAYkYC}n<@HK3dIF^D z-0<;rF2k7T#*G+GCXmoX{%5F#9BtiHrYdSq7bN(r$}4jaubkF2ni z85LoN0~4oB-j|f)7>xU!a=nq$7}&3|zVW*1ch7naZMsjZph4OysB~!~&k%*)vX^GfqP^`eMsrH2N;vZ`W-OQ!b zG!XJh#?n{{V+fJ1Ws1th0FCb@0f@0S)$}viQFRTmo*1Mz$j5( z6}XMr?Fv7k;iW4i+i91$aWa<%+y>y$%bARy+y=mI*S}@a>c5$bfwO3CcR=ThKIXmK zwaN#pI#6mWSNJdlfO6CIEB>oSkQnFy{>2do{A1fmT|d&6o8-q|RM@aRNH{^ne)Av? z4v6l^j4ns&%wPS6PJG77lXNDIMJB=QBj$6>eMvW+BX14Sg=s2bGm~6R zA5Q71saR{W?(eMiYIeVS>3&(1S!I`p-kI zkF)uW6p-RbwTgDf^>dSG400q(PN*}WtXA@NBpwMSYCffL0BArCza^| zry8WSP%S;=I{J(0o+Th?}0oK$Z1bmt8z0467HE!FZ#Ys&G0KSmNCcLXvGB zAoNu&>F~;?bTG(8R11?$o}Qz@;+0Pn$?nM%vP(XqJ?~@6? zW++jCeEQ5)QVo)o;gqjc+~jXoUUq2}WRqnJ>q9ly#r-6nK;{L6R2I2z;rfM|HBPJe z*<_*fV8{?qKmBc!d>1c2{BVHbUjPJB4)|MDIN_bwW)EgtEH)@NDl^*OW!87=v1Lif_E=(9AR@Iw&{nOKCh*^`hidet15=)wOeh z4pgA~NWAw?ZSs}G9S|~arI*hfS`=qB^+T(}>MLYSXlwYU^rSg>?AEPffkXMLm;Pn{ zkG9TU2`gL5P+UU-1lK7YioRWKMb;&+x@!*5h){p(dA=s%|} zrSH+7zj%kA?u%D6t3Of1@pE6pbGM{LvSVbr&pq0rYLd)2X37a=%z}4uu z{OmIQ9vyGTVZc2mW1mhi5Eo1#BGCt~6}QV#yydiGy;74GimY$CaLjFF0jsViTeJr_ zOi$A&^bX>=WVK|Fo{WIhI7fZ#p|avpcFQrw6!U`Md%v0TkJZMSCQ4&Ai857{?PK95GG}{-Gf+lY1)yIa!inw8W)Jxv~p5SG)BC42m zRrHSMX8Ci_#O4$b46yKL6Oo5OlyxQ1CNExcN7nCNyr5*(Dr&pqcDH7AkxUmbNTJREA#crH zdKjY8{m}M^LFv=tSgJ1~hpO_qtkf#@21i5c9OK!VQ;If^K5|EJy6PmI1kt|b!g_%Q z>qPMb@dJ#%0se9$0q4m+@Vc~$6wyHVutSY^VOS=Fs&}~6ha(a117}U9s)nkGi05aj zQpp?hir^&vY8uq1sh4y*#qk@Y^#c5j@r2z*nl(lDziqz4=CYWwWBjqKy;1cHc$6Cz z{k#|k!Z98tAd5FimC!EqIQP4{{G(xz_3I4D>X~*&2LuUB=Fdl*YQfjRn=sxAPSy_&#DYq&kWL?bLt^0g-<~3O1v)N zpI{5pM0l)KaAX;>lim$qMJ4cV%v%#0&u^cf^aDJHcjA?+{C8PKZfW7zeC178Vat+{ z<+Qr7-%v+I>;t}eIfmVag<0xEvdERbUD6T1!H{zDw)37vU8-jFHRv+v5R@<6vuGVv zPCgU^$_2Nb`+4cS^EPvY&0xbFxn>Wr(JnXSMxF|}Va|ieWn!{l=+iYOrc_5TP|Z<5 zL`DT;t>2y{!?I@eXES2edlt1S%KV~&I~NRimPhOaaN&`I0b13tEI(XFVL$%%S>^IQ zi|YAT#KjA{)M(&(=?+2n{4Gu&(NV!V@*}!v(Vj)tkN$DT*qG+g>PY*GWY3S>jWbuv z+@=rc1a9uM9iyh$1T`511ME~jcKCLQTvL6^#`(-7wPV=&ZDr`dBIciYH50g<3vBBo|D*S(JF2~Bn zxY%#^9?haEdjwLJS|}Vow-8iwRGWNjW&vN!d1bDs*D=QRpmimCpFUugke%U*&30?b?@66-`u zX>5VX$2z>O*tI2v9#73a8wnb7XFPx#t3T=mc)Jmbg;0`TIy!#ixCezy_b+Ww71^vVaN6!zO%6Mp;_aaZ$lYXHSf*&DYZl1K^pKqkuy_|@ zu%UyJW;Il!>(3RsW2Ck}WXO3#XtnFG!vXU1K&>L$?O}-3CrgXK{5AIA(J5;J8K1=o z%tZbdM^>40McFN8SV^q5gq8+q*6=G``shO{%~8h=NsU{()v6yYA0Z1dVfh2CbR{QD zOnFdTci5by<2hNtl$cnZEP{a*O&Sr|sAv%Apay&!RUV+H@Mu)%!u!ODq0p1{Y7%`% z3a!5a5*M`FGT;Ty#K^W>$f6QyQX|uS6uzF-sOn+JlLSsF~^_0lNJkH7m zDE2#*(kxNgbowLy`5AksYwphN@Jf^?`)80lJn*0fi37IyZ1HIdK+&P??lE)r$WG6l zj8Z8WnehN*tO2$5vCm)LH80TUqZaoRt)#QU$BvLIfyI#I#cP_?sbp#}iUA+>Zgi_A*VDZo^@8QXgAv2B9_Tjd z^(c`IiVb%Ts;VJRTuw&2S=Tig4;D-<#^XJ!!hzEqL@azU=gvP?>bMlD`VFKc(X&rT zG_9`tgIgqV(!@(d0|H7segXfc{^}|!0Dl07UjQvR;nm^_9w2k{C2Yr zgC&q)z1fB*I6-1cc<5Ws=45hQ7BF@kc4wiUYmuNWG5oFR5L|H2gm(CT!Fnj6?S~v5 z2G=K>4aO2EjhZ!V2p-c%e{fGJ_BMJO51EteXe~Ey!;ZlLTD+sYjY5LSBk0{kWRG-6 z-Yy@O?RU&jRe21_^;`^iFrKINO?E#|x68LOwS3KCQIq)m44wRexJ8m9)*#_Rg%^?- zRtlMEQDkaj;g=3gRyHWSnD*ORgvvxPvAKMu-58P_hAN z#(w0$y}|LEv`3yja}WshugaDfR1AulVW4r7K>#^YX|`X-jV`dGDOMTbf6%i`eA5in1Aw#*ZKam_`k{ zRX5oYG_d>1a&t~(E^8b+#t{p(b4x_BBuxRy2HsUMFd6VR@oJV{0eYNs>Jrxh=WUYM zK%2p1%rWJFnKAFT)xq4gX{|m+JF@D785`-;XUvXNnAnj$1Or*S93pbLFh6{VcR^kU z>ySiwt010$BcT;eiTa%KM?; zF4qdD3jPDjntC`769-(3nu$y{Ie(@lH?Z51NTFktc|$2nCuY5%me(pcCv5@|odI5# zN3#Fl_KOb-lKrcfd@egI=v+_@S~*1?cOue6TS1OC*?-T{OM!V2P}bv+>|RXaS?Iu| z4A(}vCHrpTrIO2q7b(29dWqrM46;|=PoE`^@bS#B!ydlj_$J?E-}ITF^aR)bX}J&| zo4{NVx9e~Iv*AFK3osr-=}@x&UknG5ruWOcTvEY5xg)FP?f2LW?^QCyaS>5Se$ZMDXRa3O|Iu%P*d5nM$YlR^-8DA^ya4ZV4+cMdVAMAHC zpUdplMP`|uZoYI?+^lYOYL{)5w9Aqu$Azo;MF6MO{3?DAdH?54Kl=Z(6Iv;DD>Mev z3UQ$%ZPr}lTvi_Rhc$E_w_u$edpXBVX1Rj|14Id>MC2(Rg#STxtRmtwpG0}F#}1F} zvvl6~BC5#*3fnQtfi2;Lq&OJMew)?Es+uTI@XZL)dGB>Q=dJ}&@k~{#U@NskSSQoD z+~ya_k3y?!n$J<^{_s3dqCtt&HG%b3Ri3f3d9-9Qv*?j#-H;=#Rc{2&0S}^orqZ6WUMVX8n$NU_E{qV z`Mj;Oo7IV-8|WG;NiRHgnm#sn7w=+VER_xBBZvA@dSl)hVYU;{Xuy$LaqJ>wm}~%o zl^CYbeT8GE{IHE0{w9tA+cYF`o1l(V|MWxi+7-J+3`;f2zPGNzG8(GwmBrLPXhCaS zaKYu$oSu*xaHFq84ARl`F|thxyxa6)I%yX6t7nDfgjLK;S5-JY^#5#;R*~VBuCkG- z(0_*~gcZ(K!15coVWn>FfBoPei{xp0K)XC;S%VrV237aTs6};(B9B(VfMmTohX*ai zbPlgUp~+Z$K5Rgu-|U}tbxx_%+Q59D21SKa8PzU3H z|LKbp=8UB*RG$yrw7@5{!$YGNv67ZYB=Ra0$jR3auP-aKpQowkU2p*cA~)db%YqA= zJy%Jy$u6?k;}#GJclaZhjm=&#=gf@R;+WfO<8n2pZ?FBP{dK=>_-vtZjpTf%E|D&? z<7(`4Caa!f1XD`@w?(8urz)GgOZS2JNGc=}cZVJ3SIbs0nl`3@!CJ}qaL6mt`JwyW z)-%vLs?{r7Az${ue2{e$Aaz$Mg_K(L&YcVNwEmR=ee|B-T%i2=)TKcg7gQ|Qpc)q5 zQz!1GT7aigtHAquB@If9ACCV!Fi|UFhI#W{E~qxSmJ`IM{7>JHeP$Vvboa;7)%4D1 zEOK0Jg05nM0fO9oA~KDM3aX|*_y{Q6a{XY5B>tV@NTWLX&5bYjExEZQ?dO`Je@XoI z{clz+Lm6?Dg81rcq(m7lA+94+fyqB@&U!k{>LFyS0m%dg6W?J660hyTOaEBK51`ut za$Z#~g)UuSZZv3|8i79>i#RoR=anzSYP(5O_%aMFSJ+G(5SBe*Rg}Z`1B@0WDE060 zbk=hdzk`WuRI%MYf&rbzz0g+_TES$fur;}dT)VVcUFx?75++DnnL{Op4$N9Xt#-LY zT@km!qNKzpg*n)M^Vof$wsEi2Xu%~VRJUggSp;eSg ztqO-n<7Ew4JbIQjf3ir{4aBfi)oq!v!wuEQ66Mzdii2~V-p5{L@J3)K zaoYQ+X9Yhx$P87k&?cik#z=7n!Whr6#UXYejFCoq;nUc%ZRQQD&q=WiOMB~$d7I?u ziBDQwHo=6`CClRVheMGS8AabEw<~s%cg;l)yjj`Y>wxQv(XdJ6Wi899P`O%zCxn| z^MP*KALxyCytoD?{!uQ;c7jR53|1sMrimbwh-^W677d(#G>sJcsYpe3ad9;`Cs;fa z?RmxOD;k!u;6C6i53=%MPKISleEqR<^WK2x#IUe*60{16J0HnQ z6h6Bhn;65=vt^#aLWiAmVM!}L9TBpIt2nA^|HOxuOcgSKG)Ne*5s#Nm2Dpyniq0*PPI-A)gEUc&CIh2%cj-!(4#`JLJ0zPHq|WLL!Zmn{q-^GT z>Xx*N?2uFj>}2riouC-pAZ-=o3Ojj~ke8@*!H&sH{WG+R&LI4))>Bx2g`bH3YZWb$ zyDs-4S|sZu9vz2eI*H02jS6h8Tq#64A(N&kZg6^vv6=AcuVBb-gT`22P3fBZ>1XB& z(_GvjJC5V9XipdWXM3aGImTd6kf>ii=v*&I6~U}$kU_^-PKDDNegWt`>-*$%vb9;j~>k?*JYC~+()L< z55zjMkDm{6PzPLU_yunDf?OVVW_SI=G->JlAqV_;@~HeD~DXU)y5d^8IA;%6_3uv$TO1 z^4IyDhr04Mc`KbocEJo|I8w%gPCR-y*-PV@vt)e4U`V^XBqDnHu!Gijm7C#bv`tt; zMvV|RIE=P(qaI>|O?HF#R@4jcm;;52-~8-B=t@MpVgTBBljwBOa^YcdH~7SH{1^rs z_EO0qURrQoM6t)PEKhvaWfw2SH{B^8f}QiX-Imxcw9j5*Rk+jSA$jTs263N|;umQ|o-1 z)hmdPcs(IQl1tDj+zv&eTW8k=93%IRKepQrJoIY$X$Xvq9Du`W)eW zmy_o4e+^PyIOdxSDXK0q4*Iv70j}@_Ecoa9p{PCdd;WvzKNy}?DvgnfM#w|H(GIiDx z49K=t5|Mq*&7PfVY|O>xTCDTMn(<=l3RSJ@k!A-yaNgmwecn!SsjABbo8+2-@+yZ) zc=5ehUDXUWVx_7X!X=&|7&EAybmIaYa!zvA5jM779g(l_dBxc zB~e8Fl2*Ys&x^rYx=}d@%1IM5T2?hvk1?p>!wAg5F9*?pv_Y4Raegs$^hD=nC5bON~nJGlCj^Cb@6HAy7R{g>P^nP!_g9>SSWC z%kqB{=ND`FqbGDFw{NrElFH2d?y#$~5g+B>di^ZD;khw0IVLbqA((9hfGM)rV+)jJ zMG0$#>!hu&Uxq~mH&Yc}&FX{&Nva}IH{DA10Wb7v-WkzkL7elHH6h?~x%saj3zE&V zhj#1cvdB2M%ZohDDrwa4ulU7x|WG?&sWRAyv*ZDMyM)jjkY#jPc z3$ZI&t2j32_=~Uzs_=?c)+{!t3E0|vj2bpB2(Ws@b=2=+zR7M$)+|o?H*-mnC$VA2 zwn1y+V8jtj3_-6aA`b^tLCb446kKYs`_+&SoH_XFo zTqY7bHY6-TI74WphrHvQ1c9@UXYTZcW+M6)RLsRPANgG&sf_) zv`LBYJ%Sk^=-Wi(3ZhN^w{Cia>k$>iRb4c(bT?#aaCg7B#j!<#%u>aarcT@-Jq3%E zis_ny*~#whikOI$MY*CF26BVBZpji{$wR{u1a*<~reF4%vu6GD3OJji+8eOX6R$?L zo6Y_`A^q}|jvZe5OCGpofhCiHLs{;KPy4bmF?t5Uhqvq8ElSnHchZ=Tk?6fe0Z#-sI)g<i`Dp4dw-|P+)}Z8#ZmU*z}yT-+n5#W|dB)rwa=$qW3MXy9g$Ypmz|FtLG0$+Cq+zN4@WT zYo_+P6ano;1_{a$yq%J@Q`17vcp4PCuL%wYtaMxfaa`NKZ_8fq6PDOE?{w5t*^T!Y z-!J>E;&|mdx1+SQS@{58-SStQ!R%Zu3US}iLMNF^@@ zLZHq!i>#I$ayhtUOHjpkepvILKotYG+X(*S#Qdl)!6=|wadS$@htJ?7mj|XCYrxT7RT3QsP=hksnUQ(g`PtSDY~)J+D0R6HYnp5aI+`a zW|xzZoATEG{_?&5W1Ml?zw_2xbjw)lB|8qed}%U0-XxfAg6<$9vC?3C?pLN$-Y*`T!2d#`W}R8y`A-R}6= zjD5153`Tf0U&=Dc5}0bEJvBMd)z~1-BKP}LLS|Z1;@HoNVXCKPs9GfTUdC4IXnq%| z!3RVIU2uWhQS{Bw$erMuN`49jyYUe<6#hb*<8+Y+7Tz>Mtqsx`Ws_P}gA~cfPtxoB z_IVx-XfUhd9R~tOnH#Hann0NyfmObaLw-iI{boP$4-_e*>FKZ&iWQZrY!;X+K9I+Sv{@TqbbFOd} zO75z_ES{cRu)--iXd`dCBG>P<`;l<{A#8j%w&H+v&Z2#u*r2R)(IB16=K=BVtx(l= zR+&5tY=jMGk%>mE5Wq?0BYwb>Ap^^bx^q zBIu1o{pUe@rbg=b)#v}~#qmSTZ$fkUD z>ek3adOhqozr@n(k*UIZ?nJ18De>+gmpiQ`Zb=KovF^)-K&^mlEUf3nCLtikN9Q|B_?QkolYhYFYR+kx5>&%e9DV^UyE1v8Dm}R+k4u z>So58&@37jR3DRFf^}vq659;|XK>i=S220=Po@>VU`}Lbw^IuXHC+P`_{0F&i3T~x zaz)u>uOkp!0zb<8xgy+*Rl+g`xKq!N4N%MUep^WL!nIE0qYg89cTd~lJmv=7px_<}!0CZ@WbZ3pqE2*D$)aCA2g^b1USs63LaLXP=M&8zzdTV3S9b|Nda4 zZORVnlO{`m-TjOwD#?OVnijq}0wXIZtzsyX)5k&ZMx1Nfp zcij47XON~`QRvbij$Z3^-nNCSm4z-EObMQKIUyTy&;i3cu3$zS&I2D=rES%9U{EH2 zlSe~plb_shk|O=p6!T6pF8+!ACc5Yf6I4A`wBEh+syK&2@&ly)NvH3B-#=hsOnGfv#+=O6Rdn^3bm%3@H*&`6m7!PSJYr z!8^)-Gg_7{e|_a;`utR&5gh#m?j)Etg1$yXURAca>%!AQkz*iNh^?O(mDVa6+$sez zOv>*{Hh*6R-eglLXh@uQwdrEP&lc z)|eOcJ}0J$rPTM+S5OghqdU-U1YB0?Was!%{?>DBz8*9tXDvat$qkdVvnAT~ z_tc{l(DA!ZV8C*u5Rou} zN&2O$oNoy3VsY+ax;n6kI_j7WnMF;bv_AydP%&*4=h)zmRLwmBrLwh7lM&*z5K-n6 z-d96!qph^tGai~t8$BnJ35)2_77!k3mwh>>-ld$lA}@C}-)~eSo#ud~UN924Ly^yq z_q#7S3wmG!*UB;lN{L990ieX~+n--ThGDK={Wz&2HDTB!+V;oGJrKMd)ys%xY(}s)K&O z{0%GdE8-(FG`&P*p1Xd5EvAvsYRyc0O^i zjKYm2n-*e=(|q;9wb%c-`X4K8gaq40mKTinH3)Rxx^Sd$IX`3XbbVP>m#UR6@otwR zaFAHL+oQ?XMiyuEn2rxnqn}`#?Ki(Szx*dg>+_R?+h3tCa&u|yI4m=0VyiwQm@Wdi zp+Sr|l}r{!EdnkDsCL7=H1ayHk~TotC7HS{Pg3;FZv}a*9+wrMc~a?8L$%3wxMjjw zgTs?Vr!ngmX%*eUec)E_;@@VL3oBiUseMizi%!#7yh;k?AuF7+$uen!a@7p2Vtc^; zz;a>;%6;2qIc~)i-h1$N9oh(g;kmhBltWevpuegC^lvUt5>Kw24I(Q^@B9zK@yaH!or#KGTTB^~{Ybs6n4@iP($<|BgML`2q!JLJb` zeRly)9eTjv5h%o-xeAoWC(CjehTKA|*m!<;5`OGhng7N&=a}nDKS}sw$BPpd73oX# zF|wO#aLo0Mp8D1H>Zyi+Efymb0P|=D`Ot4Y9ZP9_8wOP?V#;h6M^#;Zd(^w9FQ~0BOP<|$>E>C5q9KEZT|Z|PlOEDj@?xj zGF;?VhQ1|XHb~j&fN~X;V!?II=N(^=C_hPJ^C6UF0B?7@e8@RX)F@7rx69K)JLMIq zHq@x-2#aBK;e8ROK}=RqOGB}DaqtNVHa$UAhC$8CfV2XO4h?mYz=t zgLzy+ykiWw6#+H|>c;p$3lFtym!5PqPE5zXQT;c16}P3a9m5C;@J20-69^`rpks+h ztnJ1jUkT;@pE;jf*h5}id~Tru2G%ty77T%cOsKxPPhOBio#pc2T=5oujReF!pcoe_`0tbF7B-O?stTuj zw9X|KEJ`M+>2iAiIP^DnQ?29)_grCx*K#Kf^0I?%k+$lkERc2xn~l&c;FApU06Ai7 ztcEQG#27~4YYt5u8|)L&SPiRx+v_oZi4i_;9r?jt`p{JRqRA+ILNLb(`Y0C2Y8Bc= z#lbC-tuN|=!7{gk_#sM}>O1*v@{hXZ1z@o?c7&r8X@eWu@kRyO{~Pnb{_l0_Qmqt1 z(ZDWw4}7gVc`JnYWxg>laoUIjUimQOPRK23(yT5OG_SYI6L@L-!+~i|D?+i5FQ3xr zp9&ilzMDWnXk0$4ZAQ&}K)H`1w^6g(D!O-abMHOtoXdvRR1BWANRIPYx|XW$LxdJ9 zIkU+M@IM+A`9wZ3>@e0Jn@>yQ#V>b6a=U{Qey7xVla8+%tm0%iz*O!bg4dZImr}Jg-a9V zm&I7*fK>?bbBh5|v0yvVpkAl5gK8+8z#60|l2-4SE9@2YNe}qwL!y+h(C-b&xHIUSN7a$oRW`FR1%9Kweo&FW5Q0@y`H39INrpl)tf z=Lk#ul9)KRbY7-#SYAgqkm;&gkQ_No*YLHf-WfOM_075>w1r_A4~|Ab%CKnzV6o$9 zZHwmD<`EgYH7hJJ7Oby?G%~g`R#8cTnu4%qMGVT7md( z#34h4B@K9Ml|=slytzt(e~&nTXd`;R$yA4|3tHq*hJIYjegp*vbUk5LEITkh=&_?~ zFKR;`z2AL%d( zPYkUW^zrk<@5t^*d;HTyX;A5q;+rOn5`OezobVwZOXLpuuTV9sj|;a+`pAcGUWQy$ zv$}j?iQnHo+_LnFxJ#A=jR(!@Ei+2pE-ro`{_J}_{>|#k&~%#^ik}lpou~h#Ilb)a zl5txyTf=62pyBG@{5ttHwRvHk-Og1krE=Itn?^NI=g4I0KKM1DE&u}^A8}iL);o)g z3A`~6x~v7o9_>$ux#O+SB&=N zzt1IqkN%ijiph@M?Q%C7R`fNa+Fna2Nxo4@h({nslPfhhdLe&L8z(-a! zT?Y+Ncx|2#8&y`RxZA7oVbHRfZPZnqvH4#YYmf7ds0vE@ZXLav8>;NsQ_3(w)fR%; zOwe&eWL`jFSXWq%7@TOV$<$=3c6gx3G{%&T$G8tK=lA~j22Q|ue3$Jv7SQiZf7U5> z8$iXkT5`!HBW$Bv8Wj&BCrCD-sSMc3p!5tbuhK-CN@Xhfi9#o9733`L57*oRL15(O z)!=DJTDFq=JTWxf7199g6CILIT(BSl`6H|f6HYi7^^G5E)0*huG)cYS|LXm{6Y0OU z<4KC8|N7WBOI}|8TEkCwEM523-6h9BVR{>{BCNz4WM4sVI+`EN?+DX*Xcfp1(wju%e%OeT>V1OxrGZJ2I1P>=y{ zS%Y++=YZfWNL$4)n0xG{Z~4bkX@ME4A>J+jTBxzpch{oiQi3mDfLS9{!HT7hg!e%O zM1J@X4>_f=q9&Hg4yuQ;3H>d|Z<9)1a7hVHo1gC7FR$kJi*HG9%!4Vb67p9-s3*c} zV=4VIJXVB{Gaf*1{1M$v4F%d+)O)=G{BX=fU}{UJFEUMW!ee9Fv2meiEMqi*V`Bk` zlRcW^?fIRADMqZVS`&PP9&xbS^gO8Xj1KK=BAAT?y^e@P>cm`8heWeCXg6;+tkK6E zWEtyyT7JwDL|Fbb>s`qS5icxys6M&Oh=>Qg^dHle+#FszPD5TY!Ov-eIZ4oUL}aU= zJZ!B~KQvj#@za2K^Ny@1WW*s?gkDb%a3;1ZP;g+x0aYw8C!R+2LmCUCe;r+?7~Yo`5T;47Y#aV71jLgpctl07UvipnJhU8iK;x|^59ySW$>Epw{I`0 za9aDV+e@^H&jVJ6o>i{0nB-wR<1wLF^7uZ`!n^KI`PZcrserI!q_9*#>?XInUk8=R z^Ykf2EJR}33{SFJ?<#;Z$*n*7b;S;#$(y@cT2r^obKmED=5WSn5 z1+n8P>8ObX*-tQFQi_Sl7D=-EPT-GE0hWeTG6e*ZfC_w&MuyiG36cXt7qwE$KsW{! zR5ZvLesg+-TgsvsCM`5kUZ+UoV^3`^&_dwtSXQGc7vGfk^Gdu2W>xZ%-79&8Fm+iC zVHQKj!Wn1+Xfe6<;fKX08eYbT*m@7!MQ7eS#iDa*%#Vb+!tIna)ULm6l@|=aixv%GI&%N`nijO!Cp)-O}wd zj?x(Xt5jBa>3~DpfWB>1c>wr7d(F$@Iag^aYbI{YiqU zBj_VUtrYGzC~&_-tRv5_ zT$Li~TYN=aFFPzQaT^Rv3%?@C<`t5qF3sw@^Xmm$MYX(PhfZ;tIFa8@4?ApSig{T) zyz+|pELjLw_3`@_pLc0<8FpxL$<_-)n|zHd-m&p8Sp0jKU&v_Pve4FC zNt}lCk%qNNl zrL~bU)~Gyr%mVy8dDA2u`Dx-W&P+EN62D(pyhm@dV?(mf#E|SJm`sA+NklFWe(1Tu zDcZA4bPxDBQDC--s-Z42E8HJ?UZRQ@R`9Qg`{uQV`Qr*hz|wzKQa~0X?AJ z5=-eBPf!#Jg&l6Xwaf8CP$L<#V#JVhKkq)d*%NC_v5R$L3HOP>junrOPD}uorPHHQ zzhwl{y;oj+mp=U50O^_ukj@g!8G`Jhl~KiPkWvCgFa>`=3L<2V=B#(r0bZ2j+-U%#fWKF)rn?Dgd@*Yffr zZt_nFt_kX0bYzhzP58?XZY??bdiN{EFMsm#?|xY&xy^qF)1OYZ!|RIIDX(f-hI6X_ zHc2MbCqIx^dF*ui2rx42uu`a1T^0AvZ|2m}9 zhJ6a$ibS#0Cbufcz-I}2BE~g^P6nLDg3Hs6#crAV^MNpBEM2E+w8Ar;cAf- zQd^`ciq1J-GUFQ4r@h;!oVN%PPkDwdPXGMdu|?*LTU;!R9WS0(xVX?oIYVRMc6jzf zIZ%CW|86Bgv4A;BgqY>-0z1HG_1a;8+# z*l#uBfPePL(2v$2j|)-&8vJW0x{v~0j$r6p;3-3NP=ibchU@_nx!{E;m5EA+Ry0l) z*&Gw^J`Lkoo!XiV6eD^6r7+vIxNbt=IXr?~qYxF(i{xP$M z9jX>=58bZlhs8X~{J?c>9vEKEQ00JAJmOF$srA|>xfuc`sZyy6>w;1p{9{Ex7>CeG zxUe1|9{pad6VIm+2_@!^i`_O>vA80b6R29CIS5lmD~Va}4X(Br+^3I#TRWzN%{8vK%Y*)~hR&Nx1BKS8p4x*1v!9?#iO5Re?Wh6o{=*MHJgusi7E@PX z2Ll*rb8BFI*}32hnClO5`fpZX#&?Zd9N)0C%?_&MH7c^mDo6-dQ59Z5lV-RVqZ!Ei zolK=WADERaISw@8{55XT?#+@9t?-q#Xpp|nT#tEf;$y~XknGolD^Be>IKzmZj>REY z>6GUtS_5ouVIJmphJOYy)*`gM1ysrJJq{*BMwj-BYW@-+y{wn z(D&WxbwH*W42cbFk!%1D$Ywy9c*c9IVYbFGkMQ4gntLkGlyA%8y=0TUjzu0N-tS|b z)JiB3N13nj6@X9dHCtvQCiWTQxto1roF=MW^{kgQV1g$mG*{FQgws8GM$&fq-?07# ztMbre>LGXWpz_5pMR7uQ38dK5LOVzlAk0u*r*o)|u#8zK0(YNm_Wcx^4`QiCpe3mR z2J=l$Cuu9VOC}>`8AQf(kW3GnJaytb><}4a&UoQ%Pu-uNbzx)-Z1m1Yibf<`M2Ewm z{C;Dewa5JFqhlc|p8g;!U%?3#Q?6cnV}&`(9t)XWE`}oTdzL|7huSmEX}Ux)$X`Aq zaT*93WkIY$2UJ0qmDS`hAIr6u3+u#3_%@ufAHO#XN5b^GXj2U%!>`6|B8gk#5Wzx zNrqV1c^U-mjK1fq5qK_J1$`0aVaNS6>p*fh!8htBYeLtAPRg5n^l#(zMjzcY2|wAt zi(@Sb?AGPD^P}m9%;ys>hU3|rY4$BVEyjEfhEg#voJu0$Ca`zPI=BIwQC?P%8=jh0gE4|312;^*{+z$pc<8V4g|~ysk8Kwi;4MF!i9W78DeS42N@7U&<2Y zmC9IUiMUG;<+MW(;||5S*(6rp7DLkN^MK*`yUBA*tSDn~CYdX$R3`J%eY&Lwpz7qk ztrSu>9OYN=cZrd#25c?vRMBhP@W6X(1K)3T!awhunO;lbGkUhM(Y6V~xEaR_>y8hb z_ng(3%=$Wl0n*C?BJ%2-3g-{1g5#(L#qV(2Aj5aYvtD30h$SACWQn+OL9_apsE^zi z5heUgUK{$5->5EI*7Mz6Z|+-iPMt9Kjx=4gmBD_;FJ7v8Y3mQtmY!2vdtG6*!7+KY z6W`%9IQF}wtZ&`SJqQERtq*O9v)@5&}qBgw=^dkMk2a zVfBR{eqQ~1b9tp=6GM_gFzE!XB_c6vilrRI{&7$r0QEcr^HD#34ZqUmaG)kXw9v&+ z>VYld6;7oA7a>n})UiWisP3`0sW5vl7LCT3cHHbiR^H__8usfM9?RP3AvBf{e4v~5 z2YMs75(4|RiFB?B+P4!-68^H0Ezavmuyi+> z$SZUK-pxtVNVm-9j6%7U!OZ%&cfU8-DKNrgY17t^>GjVIJfNUtRE0@0!E7by1S0aZ zxWxni2Yzr8WsG}0a(NX_^@0*vt!FG%<2~#!=)3~tb!=?*Mvn+L6FB-#+irhHd9^y! z2oBjFH9P1%cHE$P%4E2!2<8w$mm?kBRY8ZKNOou;7QbQeeI>}-?DyI12eLQb8z6+X zYEhqfSXK}HIdx<$Kbb0UyE-S0e_nD&ws+bpKDMgnim%EOdasLhE9BL%!|$hZ4LD+ss~~yL34%f(Fcq6I$_bl)`)h+kn51ui%p&~9G2taDG##p>3{y=|0z})?Z>*r zzP)q-H~V47xB#})QSM0v!5kpyGN=MqZuCU{9nI!xJub-N{3xuZxybZ`^k-V=I+%-Y z2G`3J%(Vkbts1Au*fY|+YaWPgqHG)@+ z7d_8);6gHIo`s8KLm_>e$6Y#l#&TgLEE`fd~e0G_uiaL+{HLDe0vcJw$B_l-c{i))kI_Sh#o`wAwSw2C2#4%)(6NNRX*)C%=` zK+274){&r(#3Mj&xg0sWpLqAt$sWs*Cv_K%zlMQE`X9X2BdH&%kmEDM|Jr1G6nl(Ze@Uu4VSV)K z$mS4u(nbZEP2p`a&~h-+gr@RFr6|b=S`TeBriqdDv^OH3vpTX(irK%t!sQ8qhL^MC zI7XUpZsY@bJ&1)`r0A?X5VOP2Y1Yg>)0Q{vX3b@59NOp9@Afqru75c<;{wTH=P|o* zycgOeCwa`J6jMx*om6~5V2)=M-OtNZp~-6TJp~$(b>YVMSLEBm^2B(girz@@ibMh; zfeaoozOSMW_#9IANcStPnHgKa^Nefk!~8P_x8ro)%gXnDVFF9i_QU{kViF679z#ZH&CIJI8)HtXK8!dgDE4ULq>2&+ z*TwJqY6Eq^yPikunMNRv99)>K+AHhuHosP@7zyg-qt<6XeMdYr`#8}Cc82fg>ftk< zh_#9ven(Y>UZZaN7Uzp@h>^#;(5p+0&9NwMq~qhOQQ3K)su7QWpd6l^3b>(bo@He!{0~-8Z;AxV`&Rurh zm=?)5bL2KsOd3TtQ1RH_SnZ1fS~(EA>;r9z{8`1Y=X2m>KkK1SoVN`W6IVaKpS-Q1 zg$qoec=cAv22#on6fV5(fHuaYJ=D7tbBH1bsQCAUr&L_{ap+rVRKIjp^R9Jn?o2EAliaOJWdu9XrhJ;dUW<13$Y za&{VTNSMmW;9e7s1b-0MrS9ZFoX)Gzx6t=kh@P`P7V?a zWkrO`!oy09Lw1N(ks!F{anK(-=P?1+CfpQW3e`1hB8QZRxKz}y%AH-acq@ZR z%rn4cixFg$h;TMw*y7|>shF}DqC&m-A@eAMls1Yd7*Z{uUBG6tjXlaip^$Xnr<;C42Wj8i@hJlTyE zTL$>omlM8CnGiDKHF6e|s9ZKGO<;;A`5mbgvz{W!RD2;-2inkCaJSK%m)3=?7b!(FUdUvqMw3HO^H{T+u zFO0KMX14HXDJGX9TdDZE@Mh+qZ{NaR)n~7|)LBrfumYAL?5ot%or|_B-(Pq@mQF5l zo`cQ6`L`W96Sl+Ys2Tr&o+o>Ad`tl8{MYK`B%2*TT-c&O0A!M9QbIAiC{jqpSNnE? zSPT$T71Qvem>%{r^6P3we3N|3qM(T`SCJ!00=XdLPsn2$Pqa&mY3#0pd+H@HwI!<{pIn z^y83Gx8Be)?%}W!&Mv5&dQ6YHWiJ?Y8;c%wJEFu0+aP@}c-a+bM>(N9800iNV=r2) zOt#&JhyGmIu+U^#eqX0nC2ItC<(kc2 zZ8R3&WyG}6XTrvdjrH<9P<^Ml3A7G6;NgEHP7Oy!O05c|p||@$<@S_{wdbP&q!&rdTNP8xW(484`VH70HzSsInJ6pPC%=7k0yEiD-C*r9;am>*uE4mL;!XQket($D^>^Y0KXY#DMbR{B`7cd%Cv!D|Syby8-_7fcwua zDZ)O#&vD_E#tIQw%dc*T|I{P@(<51z`dvF_X_X37piu-56d+}q-L>Rew$lQFFQadM5lrr z2fQeT8lx^ogOx%};JF^2^G7VO1${8@eKH^jbIv+raKVX^#?qAQ`}DveeH&L1+PJJWyz!1eDS;7 zbE4vy+TaxqVoxxeb_c)+aAh{6L*7`2QO9o=M0l7W^C!*4|AH!d>laS}CKnF3L3C|W zxNR%NWKm=z72ou4332JXy6-2%jVa;(ZlK>FL^X#|*x)gOpbtY_eU^10Vq0D-- zvK}NL9>gSjJYOqd>uis^9Zs3osRw1Z7<})+Si6lHg#DyV`c9Yn&wqG8FDvlePrpB} z3qni1vU9#!^4!1_CLyfAb1)jTgTGty-S@wF;Z+^~;a6{nhvPE2+r4`san_~29yBl7guOC6K_ zqpT*$b=|)$XXn?yYO*ZRe>|@vbu&q~*%$Z>#XzXL6`1!V=f(rk7^B8U;@waxx>nl4 zYax9>LlS6EfUSZiFYp9iL|+o7LzCb|=a>k zXg${|5(T$Nu|MoHu$C}02GlL<=2nJ{|2XTX^U|~sJffX*Hb?`aU;ZY#;=okS)7rEeH=2t}K;hC^=qSevZbb;60j_y^h;G9zZ^$yt&tV36plrzchJ0$w(v7jz>KfU(V8{)l7 zd*r&Mk7Q{c4ydhWiK3|u+Z0Fbg%v9lQRTTsBi}HgsA~3;pOID1J%%gW1HPuoiZL*;-R5<)dtkEm3340?jNsNm!=;!Hm9G*$g|UMM9n<_5HMUa6c82n41DYIH!~QbX_O+^2OmZo-^}tSP`YamXgn zu|n2M`d@yd^lz&yU&k->zva=T-u3DZdSzH9{ELJSr+Ll1=8)u>so@o>W&jJ8>$h<4 z(}m(A&Z}WDo6l+B$?Se-y{R4ekF#I&6|9#|VT?VZD3`qKf0mVs?(@M_N9pZM)k$#U z4N)nK28~S-ofOik=nUT+w@0vpxgULBa%Of5Pa9YnswovUak9x}4pKa!aHCeSjSrRD zC`W&L(M4dMD&roKWbn$k!}2jVB)`ONqHalv*se|fG;lFtFKqZZeblhBGc)R$CI4yZ zC0lVui&Q##kdx18leR)TFLund(kEn4vFHgqF8;f*Mvj)+ImORWT8`x;O#i8}ypz== zxbB|(7we_Rmzn&Vr44bBoT^XM`Os(g2;i&V|Z0d60iDryfO6XLlvZt46c z`iZ#C1B)GwMoMZdwBg)*sMtZMs$otR&t?Ee7F-e=W$0u`9xBOuY-UlMH za{THv1Y8U6057XcJs_`*zA47I9|bcldFos?aV>Q%KX>O87Uue#K2(vcuYd%lh|Sz8Iva7JlWWlJ>`uU;4T%64%FV!G4@ zpS+k{MTICiq-ok*-%f|l6tJ-6kag%xk)fo1^4*%PU-SwJgKQH{@we*44=2iMouM$FYs{RZc+wtYXajbBa_CfT-{W z(BGjXp0T#qE~l8yG_l7Fo6)C#t#!=Q=(n$%Z*qs+M5;rih+Vma3nS;KnWJ=&VnE_+ z9|oYiB}wyFg_-!rA(oRRKkD1zx!$N0uhUE&Ct{CEtqa;?T}TD+Ha3KJhH3%OFU&EK80m zedzicd72YEmQ7r-%n@xQpPPb72kbz}q){CI@{e!-+{aQP%Viy3E23AA=O+1@ltw^~ z26`HcLleSUcxWFqo5GWPF6N9sJI@!fm>IUyZE?HP-dbN~rf_3lx*M%G5R(b1 zbjv?O*03`nt{Y<@MP?>s3&mtoBm)FWqLPEE0*wE{KF7BXn$o(|7yUkzPbjE8HwcYm zfCY}6hMB1bK3O^YXB-nUes^;41X($gfE?$f!lz9XlTMKo#F2JUu>IZKh# zRQ$%*Zit_VH*iaQhGH(u*Tv-sD!2pl43ZqdW0Fp$KHE>@@fDOYvE5HUiR<&Lol_lF zFFX-@ntp#FGBs@XnHVcezHU|?q$RieK$LE=jz1uSR5_4{H%Zzgr3A0nDl9*LI$tXm zmgXt`>oZPG<12KrIJw)BQjSdztP8I}RtV^Ff*;UEp6w&^c89KFx@09m_n>#Jz_Xt| zGOwN49XbpGug4@Cz3@mobKbvGx;^;3|B-nkJ|`DVXYScd%d~OA=0m1&?VqZ@w#lBR zM;FFPwVCbNO)=oS?0`xfrbcEv0q$U z9o)?j<=^3b%e$YqL%l_m8h2B2bJk-hSlp$mlsOeG(}oIIk!c{n=3BbzKFJK3aBKT2 z5=rq)a?I=*AEKB86se%%uY-6hNcs3T&MFAd&uj76J_{1RS#o1+tB?L8XK;&Uop(ZqRsp6+Z3figTj_WY9=aElk zyKR1^&$y74gX6jpt6%?U^MB7YAxCZay@sq{=is=m>jvtslfCy0ib+k_G8#5YfVZ z=`v0g6t5V&rl0@Z+cx!1T{EUckaeqeea9>D_IRx>Y$&XFtrr###-w>A2-3XrM4AhW z_eSjt+qa;Y-l2v_{YBF6v%#Z7m?+Q%qEwp(>9?^7GLIgP8;)BOsaI@pZvBpp2%2gM zbSkVmAC;c7OzyaB_|;0fW>-L8KzqOko);E3a0X+0qk6er>SH1A2{U5!@=wOwWDcs< zK+*=UI1Z%Rt9U3OevmZsvF)m#PUYMSUMtO>SLbnAc!`5ofx=MO89ugV?T|Ry+Sr2x zGk8pC5vG68x-FR^GE1G%zV=$59dwfK>Y1JDx9%-H z@T2ObDX(VzbN|w}@S60{O8I4>4$8z^>0R73ZW2UrFF~!{&Lv0^qE(cO^^S|BI0`%y z%av98!SVzv@K9I(O%`mKB63;qSjk{jIaH4?+-0^%WT5=U%Y)Z79*63Zv@rN%@gV4u6eYGp{@yc($5`1Hv2J- zf`+Z*R9^G=vDD<5ociXsu9J-~OjvEdnIE%*Vj#}8jfzK`Ql!ci80bc*zrM{KkZY=a z(>ZPQI!@i(@ra8lZJr*B3iPNZ|E{W!E~CmEKzIA}X|dVPOe`-GKc7CV-Dl5#@2!KD zBrsOER#2eXm>22(a_3TGW+W}}CjX*vyU$)>TgY_@c0)L0r(o@`Wc5(34dR63tc^!9@R;ZsaBa?nQ z0r2BP8pJDKw`|X_!Xvrab0a|^4V2D6j(Y=tgm+tg8;nXL-Nw%eF6M6cN#$H574kY} z6nKj#t}g|RcRc>!%U8ad>QsRZyB@N$o-mC2f|!=9nx6+)Zev~$htkocRnofXfwDiZ*=ne%a3vvSe8w{AgHiXWc?s)MZi2B|l2Tl4(fDx1M!TvD72iO+`5x$*7=d%O`B-s4*FMvOB* zZZb}j@~E?_Ct@Q;)_W}nalGU?@Kv+S@}tLHtg`jN?b{oWbstK>X8=hcOC#ShG!*B*+w zNs((HKehO*xCHwD@&FD{R?;Y0!^@I)D)fph26Ff11Bg(9D{b8v>wbu|rtMHIA$2 zG>2r#hrI?DW~*@gr>6@Xg0e|~WA*YWnC#noGtjaf?DJ(RF1!)4()4wOSuq=DNUO+? zJm-tG(^!GmNZ*pT(r4YXm0jvv^44 zinVl_J(1b-YJpG+m{-U>>1k3B04ouWiUth#TJ>BVw@P?Irpe);=utZFmVAhex$Wa# z6XGQUvg*YXD-l~an>C_XEj*JhW(hqd?RV40SWF=oWZ(SDDCFV-Axi%ux{|4<)8usl z7(~ybKbC{GT%)92x|efTo+iH@G%ADVi&tC^LLKhZz!GSddHV9r^0&9R+|Gw=F`DDv zS~vJp6c@Sf_9JQGU-B(|7?({cTk%&4m8~MIPpR?R=><6-tzr)d-tCF3UW|V39&rO~ zgi<-D$<8g21hh>XzK%oclq0c%?~JaGmNo~O)~Mg@Ye^%8 zGf9(~aj&8n$jg;e@q2=v$oIaQ#?|bL`#=g^c42v8Ac}i4N(c6S`8m~UPi#_!qM?0p zbvzvng{^sX-CUfH1VO(Kq;Z$!i8gRcgO)8HTv)SsAY#~a1@A^6lo;KNgq{x&r`$B_ zn!}nqrad{6Y_s{xDW5g#gX|`Uxa#*c;q4-gB`2Y){;p67H@R8kQ|9hGx{C7v+>YjX zJrS@VDw`y^^8R_X{$-rBzyrdSVJNV8PntLPxRR~+z#0{w-oO&JpWf%}7p!0l8UQPFzZ!Ep=$$e&+P6q`A3|%$l#3%78G6RtFi3oqhWryq_Gtr= zm>XH1G6VZ)n{fMV3^^Xz&;CqXu3&}i8BLo!bV?I|Li7HbL3Xj*m%DJ>ro{|y)f58_ zNc*YyTXJKM{3o2f!YsKqFeS26`fy%;NK(k^=-cvUNYZ~4lNngabjNAhII(@or*-AYe`@`fjhiYYPHzR?;C}HZuWgxsoZNYR%lv+y z7M?!({e?Hd{n(+pKo^2J2mTXM6_Xm?#JjS@=(ubV9V97CtIuQDI^CDIF=O%ySu35Z zs0_8<*F;wMG=^LfIv$SC0*{Swebz7zKhF-X-&YmoS_;5gQBFn4gkevRK+BXhgtm}2 z_nO7GL$AlCyJrSg#S{a(=aA%l;PadZ8)v|19vuFhtq;g*23$A7`>i{_5&vz~RW*bh z7Gt@CRsrQR*cLG6hFK$23mGrN!!1+=(AFMzcS=#rRET1QaYXu5gE-;XM?b2L{M(|V zyd^)ROk`FDf9TsHPL1p5C3D*NZNfdV#(2^&-SS+U7+56;oG|~^hg`kjTa7<9ZGW|Q z1ta8w3&&3e%r>+JiqTQz1{IIo#2GOI@_gQHZbDqj?1VT?yL6)v+kdbysZ_8^s`YwT zrNM5pcJG~`kAsauMPqI#)>wq>JdhrROi>U0*bBwauyEHvUxFYD)L7+620*eAG8lak z!;(IJod+mH(1~6cP-%yc#gE*h)a3JOfPCSUBq0tvj*3I8WQ8C-qFJZTReUH-h{Fmh z&AY)UkA$o5agr2L1Y&H^{>og3=w%`3s99T!S=$tI2iyL4R=~q*-JGGiHtXANCY+Uh z{Ofl}f(zqJYlgFQibP9CYV+Dzq zD&GB{Uq_iBA#`6_NcPVpzzsSnLfS|%z-I6+6+iZ77xazn14T=S{|idEExzkHn***3 z)=7%FdubhCFS{PKTXI@hWZMr>MQ2@K{oZvjV5*7*LuzB=ehlELOnklwlQ3uQ-tC@1$U9#+8>C)LI2z9@5 z>Gvelg{=xmG)%(wPKtqTvzCg-c>lQ8N2S?_AQPI-)9%8fR!nIczNR z&JH~+?snH;ptf7mF&9*>tt}i5fXVdh8W7Wt_yw$t!3w`Gg>tE*JQGYpv%h@+H9mp**UvV`snRoZQ(_j0@_SNstA0S06TWJBWPQ3LW5bUWP;@F?Ptq%0E4 zs7}zMK)SrI+zh7_!ipgMON|tf5Y;xzj^yV;>9%MMktEw9ymMN zWJG>(eo;B8aA6~I&di83QA|BWjsWp7Bnj%d`@(em>ptr_+3J(L_2d|D7rl9IlVrmz z6ivKE?t0yX7NnxU%0*r3BT6kex%r}A?@iua>O)GP(S}`hc1Tq~r+RN}AE!J(A9q2$ zl~bdH9Up*&Ly52@1=%?ERHulD5f9mNUY13I&_J-=d{L6el z6NU^+B5sh}7be+UV}`gr6a(FmMIcwst>ok@hCqkP;D0gbil7=KqdPfBV24M69*TQ= z5uQ0AIZX^veRMmy1GP7Ju10k|_P)n`k96b|kme(LNEXH#hEM4jG z$%v_^OF?+-lV|jaP3KWganJN-eTsn`e2A4rnen9Oaq~+ijI3G{b_{wRT((vVl*T6S zfYK;t14Y(S@w>f{?;$b%=F(2}HK3}zpmwI`VST!6K!o+1op^>7BB&(anB{+4JYoA} z={y6SKMTsrcc@Q61;mhNdpOjR{~rYBY=F?Xcj82}OgZ##R(`eZZIhWuUYoj@^ttd- zvi_yXW@3b5hAHwGnZKHqMS=~WN7%;yP<%LSU|y5ICK1GdE<=!S7p;p(C!;bf+ukz& zgi1yBIW= z@vD7dh|I^T z(TBtGdACRjCzEUZwe#p)fo6@zD&83dv~5Gr+n8I8_^#X-<|z(EeeqIgU8xOj2*5u; zb4@rZ%bxcuTM=cm)9auMSsA65g1$QM+m^1(3(_5H#qmk{8Wz6X5NC4AmYx^EMmw}Z zR6`YbVt1mBzmC)^*2d*SCx17S9I{)Itz zRK=0#U6*kZU$`T z)=-yt74o%lo$4Xd4Rv&91^0Yf>1)2lN=G;*(|Y>P{?6Jx6|BC`uA33~^O3R$%TVtN zVrZ-ch6m*7Upmj!d+GRP0oMXPm^UCUivb2ZDUdUVeC*re)F|E*sIwj4Q{3#tL%hFa zAGajI{Cud_b?du`6~@jg;lpSsPUz%h$!kMaMW(cPd@#tI%|&;}@#x`PkikNL1(17<+rcs0=%ToQ5Bk zaQU1I<#@wq9Pu~O8S5=${4QI$ZzXDfz-Lo0y59k)UC3^pKo~BEYCC}j$ z(}R!}><)h-ZXgW-*TQf9V$AKhyFmgZy7tezZ4<;A9kvZ?_f@RW@zT+sc2rne7B)_; z3%e~=?9f>Vx%~-CM00|7yn0w%A?u56c3A8< z9I{556~kjXY&XIt21xf+efhVhRq620yrkr=3$IFR&C&&*P|OfT9#QcZe1~a0h@O=O z-3>b^F85A=P#>fYB)2_Af;I%Vg||t%)Wvi~!O?r>Im4%}I+UBRxAx_~{g zSpL> z>qz74lTWQ9tE0!+M>^G=zo=gN$qx+waO#bI5Yx&F$}x>)WGxwp(E}67ir}?CRsvKe zILd9$WotR(HxH36bs_lN$2ePJw}|pNyMbb_p1vORVk8Tw)rU?!o@YYe;J-*$lO61; zWL$VV-e~4QRZ`wNSY+6$l5K~=Kg=y5VTj|5mC-mj%w$d28VCK{ zrdzUZt_d`%l(h|H`wIgaV2PPjI9^UMr4&R1*Wq zRMBJ9c~3EIts_y!4NoUPrx8B4xCCCE$cQ8zwz1nElD&YK7`{EfSThYtw={r7|fw@lS);1ax8|;&1 zaLmD$XzJlu*(cX^2H!sW<{V3ADVNP9SfRWskLprXaIOh8tNiYePa>`XyJdyXH4jax zAcOP?HBJ5_zSv6bgiY?$&|saSv@(pt53|DTOTVo%#^%| zdf6*^yB8hfin+9uyydikLmV{o+hPdPqY+m%1L0lE;1< zNr5=iJ2$44^v}teUFp0%l)d0`II2to{%Fg8z9}%7nBP~;`fswqg-r}ZUniNEZ4{G@ zsnYmX2$7YIGt?w91)_&B1Kd_|AAQ~D0;7+<7ky6D2tL#{1$4JDA389D_>380JK@h5 z&S9rntw2=#>uJei6Fg*p{c51Vqkln4gJQRLortvqm=N+> z`7uB80lVyv>pJEn*(|+1Ofios@(@#o1M+r$TV%U53u-S7vUS8jKN5E;P+ko*sE56J zVdsr|;zn7$a4QZ((#S&absvyd(&VN9a-;? zbnioV*rOwdXq~$~Y2-t%5M6}W_ zyAyA!3a_3y;UF7+uOTbgfzO4bzpVdDFT_T-Z)7xrk{&1U&1#Xy-t9TlG_KnX{!0vXES z3%ru-pi_ChG>9?w(t9FHMLiK$nPd;(+I%-SOa2kW*BWIXK;tU}irPTt%XnWY*)%IL z*f6&^G%auyXMl?dB|N_!)Y@=dO%e|!NpLM5_Uw_Diq^$8i!=$dG#F2`V|Qg6E~eT! z*>Y_9C}OqV&ftY?dG9+W6ea(z_7OQVliW3%tZNi=g(8=z_@hi|&`N<`R^X}QcL>|0 zNi)aZY+L$~__BQUOzeX0QllkCY9Gw*70+1(W!#|q%xPBUgSyA@sAIfTPMrtV2Yk$I z=B^Z^!qI&9MmirHsAWZ};ke84y}~V`CJwF<_h*gB(ZDy)>H|KPcIj%Kp7S2`E8i3k z$Pxs|O4p@s;|3v=a zWfQ)Fep9!MY-P7{xv<*=rRkG)X}c+=h$1_vcns(i(@#8;<~PjM=;d{DlS6c~Ab*fO zZ!Ds2ZjaByXyoFpRrQks?%LQfw<}AwD?d~o3A)X_V02pijspvT9fs*y__7c2&l=s1 z^S*Tve_Qo3m&jxPgWOwkT*{jKOM>1?qyAX@;|*`*zxw{tw}yn3vWMJ4uya>At3CG! zYW(xU5(Frsz9MeBa`WuNam^u4L@3@fl?|l zEjGdj!rzkgnNxH`x$yi$Dq+(6=Ti*uS>+ga;zrpCV-aN;RYqY38MDX8AYUAMR@tH4 z=HB6rRoY3uSGZ-=a{&Y!AWYZ@X98eiZ0pX`KmU38_qZmAxCOfP7Pq+pDLh4-@7gx=2ah76EQ{)5{k1S<-#7C5Oy>Bl9RnA!0C}SD_O0T|% z6lNsiQA8dHLTKa*t#W zl>Ea^9Jg6edDF?kBUm7A)TO{KreUroAH>p4!$AlRUe>BuH)z`qs0ouY-Jk5R#tKj` zy#!9m7hPU$w^u_zZeSa~iIc;*8&T}hNN=0ns9FVz)z4*hF`FJ6QDJt^^LWDw6;v1R zy}~cLoIthjuosw+z-*8OZi8;1))4SZQK}>z)Nxh_T1d6876NY2Z97EHcxiGKV_+pl zzK^cSLI1r@R_3!2CR14qr=o=wOlF+#=$-!}IWF(ZlGL^vcrRHHX96!2}@9JGv#)o6hOZ&{W@atH-7>oz`b)3OCbrx1N4i zzw`}DPMOb>m$`5>)(X!|I;V|phH~!eSWR!#c4hNitzyh=JjZlIrB$Uc2D(F7$UR4n zNw7j2U0}WZaWLq+VO=U|U zV18HuoSQS3%}V#zAV~zt@l)QTZhK>f zphe<9Y%#r?F6CB6BA3FY=PNtD?&Rvp*P4vJd^|6!?)^nx(RSkX;d z6;|hgishZ^D__3~<-8ZvUFvJTmxayZ1|LkJ@8g#GZw}ZL)Z{%9lm@%0RdLPY>e#cY z0aC>6Qt##NiAoKBVo%w^Ga+Z!YV^#}r+>jZeg7X8m>k|$Zs;4u!{XRE+2VZFg z5~5Mi@B`E|@-hPw1R6ZE9yoh3oroL24?Hg{hQsyY8z6Qw;;Vz|D$OlW3s^7E2Cfbq zb?cYv6+pnqWI*Y{9>ES~Jmym>!X}k|Y4Mz4=)RoZw(hej8h_;s{5Zm|F>L1jlhTBr z(7eB9kXgA?#@C%{Kfb>T;ptKlWCRG3cC@z*Zi?1!9*sJ^x`% z8<3v_X1hk&DhBDm`uXKR+@cNKD=ZXO&;#59{txHjapQluEY*2H7cErh;vLn#ShHEi z-7+7~V?Gh%!^SY-m>X96gW7#)qhwgTM~PQB$i%P$kQKkg{IHD(dB$a|05oGbZQqY9 z`({5Mv~%6qyPCLJmujTHh6T)%>@B5s9 zZd*J*L7kaL*NMu-1)|N|T;W5>A#QH$RdPw$;f;I)J&tj3E#}W5II_nwjW`++})Rx=Yw|JefJ8V1O5ib_^mAFj0I#^jB5X- z-(`xZU)wgWUR^NgEn zy&T+@xo`(W`7ji7M~b#9XZA%>$kp+8iZfN?T_pW{2f4n8i0R|Z(m3IJ$6yRc`9P@n zx}|+#*nUsO&si-#-HfG zI4v1HX)>pYl1U>t=VkvD^{b4@zIc8zc!g|XXJ1@+Lv+y0z7$go__PI7e52%%48ryN zecV1_BYm0_1Ykq!er{f1xu{EBqG9#w8cWX@dc+;gz*xLlH}1y zy0ABE|e7GUsS zt7!3EA@$J|hRzb$xi+B$hNPNE9~r)3$JD%fH5iQ^Z~{6QbZ2|+P>%KX%*+Vhy8Q- zn4bjQ9U$TOBx201NLA#MC2v-q7MvCob1)A%=7#O!$jAZjRK`0|3@(o@^3fPO_HT%X zJ;5DMbbl=CmSBD8nA=UC?LHfM9m2!@nkG&{SP9UNY=pAWZpk%q7Z>tdS@Lthuz8M) zHKk1)y&Q`UTNh-_!a6YUsyK)JA1DW6&T$90DU9Zhw1k5JNxT?;)sH3h;r-*k0RDPY z`&-jo6sPi=rh9BklUaN9<@uHrAuhXpw?g{StZWF`FBnf!mk5kx#Sc85Q!AVKL$z&| z&F^_0@2ne}DYok`y<9M3!%HRuv}#S*F);cr%dZ4m^T{S9jbb)XWGxlH*>A`b69u?6 zDixse#MIwQXFq)#5aH}!VaoB@AtL_7$}a_*5YhVD#u~D7st#Efc95IRP;!7`U|+SD ziq8kbpGRK;gMUWR2#soOKuL8@wF&CdTev7Yf!nM)kM`LuVkB^08(touRcs3Gix_}{ z&V0@mAdCKli`}#aSvoj6_vt48s+hirCJ9OjIvj5{K*^@1VM;vManL%Hc3wPx&yvN% zWfzxfGl+c3%W*=CqBP@+OFjo%8XH0e1)5Y|A5e+wuvlkW<<2d@Kby!Tb<}B6zzRhyYVH3h};mw$p6$*DR#n2{; z=_E}nXGoGY%V`*S+T7ZLpQqjLtaAqg{;}hW?wR67x7znUY^ysJcY*x>fZ_otru|<` zvn{jiz=*8PpkQYumQ@ySvJ{rJqThNAs*k$VZJt9B_kFV6k=?6gUU$$P>8RT!Qo(JL z4g1#%jYpo}bl5ZrpRM6GKSX;kWHkw{>*d{keB{gS6JF}yHU5RHdSL=5JItJuOp3{% zNUBkyM^oXW;{&wHFh+;}JwITN1JHJK+lyCr_y4@tw>!uLi{I^QNh5_b zNt4<0tfH9x6c9JZ>-f7l8N4<=>P2;_S8{I1Q^SvPTj@13^X8=m7RMyd9CceW^QLIh zYt!*A^?MLHT`O<##e45bPbl+f+&#e6^H#o`Jk#-WVf*9$%%V*3irurc;p!K}n7p=+ z_lBh|4I78lg`==mSMIpT8VogiP*Q#qv(*O{))wWlB-Ac(1Sss_;;Z5$;Y9^a*jePCyYXMBEHwze1Y> z@iQ4qc86}*;U}zMYx)mOcIWic@b8e07iOn@-^@zsDCP!5uA-*BUeOzo3SnBESD9}m zm>kRqVH-e}_bz&q-+|~+w*sHLa?KMya8)TTO9q6a!a9#8|GiNk^DfdFJ*PS36i~-} z9FY=>l1n#YZ+mL+pUU8LqE(~?e(XEyR?jbnI+v8l+(-?6R=0fE)45HgR<;o3`>*8a6&x`{& z3zBB|obt`nE!*(C1qUs6U@mK@;TFrg81Ph?Nm~X?Std4oeo0&sXWWZnzTK?I)Kmlu!)F zhZcgm3GApI&}(BSes-x_=|oZzgh$W$u8qa_X}-reJ)B0_3W1JX6yA}J1a+xPIJxrv zdF|mtk_5pnrDNS5^O>Cnem?vDn-AsCch>Q9;_U2PORTu8EwO@&0d9#{#~<>p*=pxSa!@$qxY?iE~2YqI=nlS*cl4~qx9`X&a8{F z3m1-T2(UnvbzDr5ik?yO%MdTi#3P%?p9`m=trXCGEX$tPtAdVcsB?hLc`=PLBhU03 zEFQ8$-r9yBhkO*PLAsMa*<{K8ZiQwY^Srp3&}bFIp0~NTRGJ)4pS(P(oajltu*dOr zK~McdYR9IQk^3ucYh7((ArCR1{{~2;F(}ZY3#k;Q&mlnZHp(G1U~QqT)?r z3eqq~|TBGXw8|kBb3Z2RJ4yK>eBD@yu8DL!2F` zAN~s z?4G;J)-wOwx3*?4F#)3UA6FHT8W%Pum&}lQl46ciB!jjT{>M z2TQRrG;8eaNaB@xXMzu_**7~`P%6T9)_S@+^pFoYAQ;}5&~ZhkZ_OP}ek2||76bIV zigEsFOca%pKB4JtJEXC$f}G!g>{fUQCsj0kVzFm~{F#>ISr<&}3&0BGGyZP%55DvR z6PA8i@vC9d&Mrdd!cOc*X2`omG1n>5O~qGmx|#PEc7$tEeMcg&iwLy^_6U+fis+<} zR?-GsLwIIDKIS&EU`RXwpm&<90jF`7-`$HaIkUN+{I z61Yb2zWgry5_rQ|(zFEd{){E@IUv#sVxJDwwdB};S#|}m*(kYiXvRvvKovyk2FO{u zK=jye@@woOsE0m+eo(G%PoIjZ&N>P1$))14-7pJ^^D<)U>7%qBiW~MQVdXnWpyL%xkER!v%ui^LI>$w=VO%f0m&Tlv z!!cnwZ_~wy%L}MzK;L$>W8GogQ4b}T zc$u85vv)xbU__iCD4ugLDxLI8TNj;XI@R~)B!wVH=7D*6$|gxWB$IN2YnD!2hu3te zQ6{;CoR2Nwc5$xG=~SQdO$ymff8duCq9sk-HG(}+I6`j(s94jx0&{|~r71PUSdw@{ zd}v%q`Mm!|uY14|az|P!xD(A9SkncHm4(IXVYWbGNOt328Z@xuRT^Lqn zX0XyyOfE&XQt{YMpE3J@zh1spo(oP)!=i1G%ZpScLHObpF+`>Fwnb)$c11RafYTz| z!8l^a{p@L(4nUqgs{L260>})5^qsF*3T;7|%A`H{PKqg@h?a`SipYEp+V6zu>#?iV zLksf)_r#*2V$ae)>|MJ4d;Q<;jH(rC8u_hsKIf4NSqka`_QXOvx#UzV>n|X?Co!Gb zFn%Z5eGc2JSl(*k`M76AunD2ZSKEY_A9xp6|vHRg!DYv5Kzav>;1?xTct_mr}A-WEP+Po z=V2@+&ymfrn0(W7c+CzLF;~A+XsN08g6xm2sI4LER25wZl~KrXh}E0jlBD^o!hpf> zVay=L+Ax*}+=vEwE73t26(Y{KMqv$AF@0r8InY(<g~o*{SG& z7LqYH6r0Wycd}O0r>jx43k-3$jtDl;-wHIMl&5kW6 zo&rTy2%GWGn=*pF=-L#2?U5e^rJF8wBfX9^%929cWDj@+oYBcB!kBTP2l?(}1k7$-gYJj=2NJZi?HucO%XTianY* zZPH>7kR=RC^lGK=NN+E4%!=g%06hg3wx41JAWHV&uh+ilMc*rCSNe3Rd!mN@H@GK+ zT?olhRmEUt=!k!px`pI$@D(0Q=WU6pjB(8DKYKn{6JIn{m6KE~=>>$(}+ul&>4-@k2wMDp6y#iWm2nT+fHWn}$JlaW0_F~byj zOvNV%HV5TMP6{BrOWzCJzy&tjTVOG=~*NT%wJG{sA+fUI^Qw=AZB z^hMwat)en?v)@3BrWMA&Bh3sPa~picFch@U5ee8cccvg*T9v> zUwj~}&U0;Cj%0`=2yTTW2oAw7e2_Hq$J|N;En>X_e~~&>UQi#6srG^K24WoJSGIhO zo@Rxdb#COu&JuqU{zjP7Yl)T}{#r^d75O}_oScEx?41F|xb z+rCm){lZfq#A|Vk8J(cP{EucP>J?t^=|m@%|P*>$iKmX>oRifH*CttZ^Rw}ZUAWfMO7DMGq=D`bBAd(+_nk`~0x}*$M6i|4Aa!#w z1HTUz6b+teo`+$W%TYQg!fWFX#Ap?Db4MaDx_nB4>x^j}t>WtJMpbTPlSK1@&wwn= z<22Z#c3N{t(gGRMZL?Dr{Luk`vk5<}R~q{-Ycd>s3!Ek*N{Cf92PujwyYeX zn8Oq~i1J)$6&`qulk>NX*X82oM`H{ZqTuwdNYt{(6J-<4zF%gOI*+pX6;OZP7cneJ ze;IlsL*E1Sq+OAkv#KXztj*o2!n-tBwO&kT&nphi3>@}6M{6oPDu7|n0dR1zoRSGR z_QPQYJKz3y?|kh;4-$|DcpPnOfCIDa? zf+nthem|ES)?KxSeaeL~VTFAvhm-#L4Y47rOMN|74_(#z=mJ${VEOzq&@9n;wQ~)U zbY6w5Y8H~fRMDILdR3!t=R_5;1M>aQV}#dY1_1WY-x~P(YENxo9S_GqhGPr`rh6A~ z(|PAaebREs73^1zx}6n1iou`VQ5S>o=eyit&;3dqr%R0Ac|W(AqL_%m&cf2C~U(mm^+38>|_DKMMMyQ!8%atPm8@tH}ZJIl%$h zIp1@F>m*0AY@_$q#YbetbC1h2xw}+&;+nWK;y!&ywcR@(^3%)4+%|aV6-DFs7$D%g zHxPrX37k#f1%2RoVKJ0S^hpX?T_?a}*MOCk3A5Q>Q|;Rt zf+lD^rwl?f=n{oDhV;IY2YkZL^^@&_l_T(DWsKj~XVh^_K>6Lt!4qU9JD|AmDh3TQ zlYp{`V$vy+Ld9?4488(Qkh~E}-@qbY1+xh`xzti)crB`9MJl;eX;lJ8yLJo<68J{s5|1e%n$odU-kT7Uwx4 z#71a1F5YG%ta5f-JYmT}#ReKK9Okz|rG?@70eNk-R&mW^Z`73~T2&u(NgLBZjtRlq z_~^$kvl$m`26e!-(%|_UXP6ceRpFukCL=B^*9kqWla_`QidjdIHAt`9A;hv=&9E2t zNj}GK`{W6D_6$$H`uV+P<$=!7Ht5#aVJlnF?a zkj~({PKc@8L|Ro+7=vnk#lY_~ab**)fNPK$7Ivw1{Cqgx;=2nqsAJVERIhTT-J@z&Gy5K zEYm$~926HeJXVrJdLVWg_Pin9AGsD-4(pXWfUKiN+!A_7*-z`f)dr`cRiWbuahOP59Z2;N&0n@zh+&^vY`?bmJguM0g%jC8TCyZB_ zRZ$L7%m7918_C9Q$#2Qi-Mb+Lh@p-FZXdQOBGr#pHAHr*nmCQI6rpi7fA%k)C? zp710&7E!YG7|)m9zcadj~tRr4U15Zpq?I<=Rhsw5s1lRr9&g=6{j%9Uk|$^ zb^Kl2_X(&HgB~&biYd%K*}6rh->Y4UY@O}i^#IJOnC$^L_zD4Bdrz9exNP~Pbz3wA zvGYsw|A9LCI}`BUZvRCn`Ot-dx6TZBPblVNiVT8em-L+;`M#xhKwnF%NGBT1Id&@g z=zLzyV!b?-gB@<=QN@sn?qZ<8sZ-I;ukyz#sA}I&X%p{yCxqn#9o{Y=uS*A>=SLox zvn-D~f*d8%edG#cF;V}mNQE~S(MT(g;wZO6$H`3ErAS=W3oySXZ1RW3H{=M=+#<%? zPm_WG?B%_}3_#sXstDv)ifEMRsSNRt>E>p}!ikzq7_ro%` zZ@ghH7t%s!$W3__-79M*O}t_HI5(Xc=2j^?!60QYiC#Cw)r*VhI?>gT&Zy(u7A`tp z#q^;C!}LSnkD}@V))Q#3n_aBzQs0r5aCCg**gf*b@$sL;_4(mdyS&MWlqVnRS~d!V&PW z`NLRTXdNC?tcKJl3;z92mI=-m#9OzbG^gVia<7IoDN4DkI9hI}`k=T#l+L-N+|0e8 z&IwK=1<>qO%o$vmAZX$g!A`GMSQL3lsiRj4PLsifwR7%8p!AAfmLSLnqmmkzAu0~7 z*0P(1TfKgd1JplE^#%po|chMt3V?l-}tijo(D&U?Ijqr--E3&efDo`K4 zDQ@E22-2$F*}!a@hiX4}VtZlNc2Anl%T`Sfa&`mi*$&OLpvDeRnc@5MqD_GM#rZ|$ zq{4*(bO%wyp&k-s%YlzB# z_zK+Gr32}73y|>FlIy2quA`))3K6)DfyG_?i1LccfE;>Xc}ySuAXo zJ3wZzoKhPh#B$V;H~-nQn?D*!0EP?4=2!{9z+ek3rGXaa0-}SyEX54mMjF#6omoiG z@?6=73nvy5wETB{tMB_+lu=j|yJeiAmcKE=BFjWrY#^86((IHmIXHB-W?i=-n2E%xY+u)kQTz zEU?b&g6ufiuPEXib02Uk43CSfC9iWHO#!wvU|*{U-8r>K^w8^2&}vShsDQg+sxE4+ zcxlK|sEge}4#+Ok(e^2H4VuKI9no~AAUL5PyA6D+Mi^58DSGZjQUDR?_ zJ^u()<3@vV91|n2;XRT=oFS72{@a_9<$5pJAh4pO2ImX$FS4`-vclZ!H%|7ouz=cM z*VHzemz8_DEV2tLC_p_uY!CjT2HsxzL#T7PZ<~abAK{xb}|8rx&Qs`siwo2KmBPb7DdolnoGpkgeKEW;(sBcu%~E-rqr`Q2@V zRylu@az6H`2)*Sqzc&KTF!a0>*zIf;9uAnS*UY8aZ5t=d`ea%9-d(m3ZqN0(j{=Oy z`t7co1ZtZLTllL@AYVc-y9sJ15sk4hsp`f4S|F*yl<5;xyjaUUkAYcHMe_>D^q}Vu zHSMS2Fxv?#6N3RE7wVl2 z?=|y#I7$9`dFh|37?5k>r&fL5Pm|=oms%f+r%QSJ6b~V^vFqWr0ORC*wkUE!6(+Eg#x2-&$c5Xz9q zD+}7mO%YVijBz}Dz0FWEQ<2P19fF@>gII@>eEGgwfpOs}`^pFBsnzTj9v9wG!3JuW zEp8LRY$RZQqA&VZ$(yB^MRlG`lPt&~F|Q@YV+Da*t=pXEg5#>YZ{cRk$qUS|>0zzJ z4DajA<2@G>@0v4MS)qQaldlhKl;%k^9bx^7Sj7PeM)8m0EByR5*!B^A-hQ->tbc@G zs144J2j4mQL!-pd-V^py7hHIQ0~t@lWI!vyK)vf5sC*i#-b0}TvQ*Y8-6gN{Li%{f z1j(5eGc%2YF>49_PZhVwC!Q^I0aqWTVjzpr=zws3rR*;CSl$guD|PNz|2$p%kZktE zJB?B_*FZv4nYh)j&b>o)mSaxI3243fk&$ZUx|+anXV`X3G6A<=VM;sz35 zwe=6k@&6WEUg?Vu(D9GAI^rAhMP2#nqu-fpv0tdoj}=|@lVvlrxr-p^`l583x{F-H zs|!8^8Df|M&_}oPbICmtBu6|?u61@%_RkT-3d(@#&wD!e$U2m>_O7^LuEn~n$;66! z-7BaMEe>o416h_BO3Rk}q0MCF|AfcTw)uDNTw5C5`utl^ug411Z%y|yY8T(B-g1dr zKaMIgS(djE41}9*!32sbev5q1^c!>+S^LfnNiJ0XSXMX1V2nMqO!2HEEBFt|6@=5^z-H<#sT8IY=D0KKI)w!T*_b;&9zXuS zjJoNa*$GprJL9P3CcJ)638sgjx`7aa7t2HYK@wDrLpn*UK#3c^YyQ{Zijj9OnLZ$2 zDck1R9Xt@MX_WQ}w8~Ciq41hVl7BP5a#|LN?0MCk)^MnD8C32EMu}C^RsqEx)N45y zbQbtTn)&H-RzOLZQxv*eZ~EW#hdSKFKFwsK^t9hGPv}qx?AsZTp9dMSH!b}iYpLuv#;{UjvSZpuFeL=F zd#Dag8}|Yb7a+|JR&5<{GaPM)cxte*)X6JEEmRh{T(KKU)32lx=mkC0fZXUX2555% z!`tY?JcRrfNIocm+E$h0KlNNGD{`+BI5y*bq$18l>^V-tjfvmUIq^$I;52^yPByjY zjRDSi6X4Vk48-nN5Yf#Z%^t|uj0*9pS@p9GwL4qMjPWh>YR>w=qQEBJmB!A&HFn`bBu%Iy9|IrDEg!VtmHS

    mU#mK_9> zOHf&cypH(LhrUgs{D}`DbWsU())cMsxV$FQ1K8bt11-bkk996(!+G9l4Hvy`%w7vN6(e!6)L_^F-}(M zP+Suoh(yNP^8w3)6oUH%->5^8^C~m8Xel z)KFJ?Wl8Qs_Rk$vJ;(&qN}+%N+9le_L>@Yv2jpubA-I~e#-p6qqP!+*2Um2A@M(Cn z^7^bnHw?yZ0gi~E zlb_N=#k+9KKZGg_i}}|Q4AgDgNJL{IGejzguj+R@Kr;#j`ZjyVP1lrrGZ$lQ+wYSsq?7`oQ@IFw&HOUx z)g3YDhP6UcgFv%hEQIP1kl67&crbViutw<9L_my5H0yo40$XUbT@Sk9f6d%PN^kp) z8LvTY$6lv~J*XQdKx{a)19H&iPE>G!Id`ecx+ko#k#1MFaTf{m#fiMl?uFqUbc1Iz zStDOd*{P4(_aY6wY@6nhJi)%c?1YQ&gYU)3jKbyQH$J>hts4h6)8RsXE5ShUz-FM! z2`UV44?pauDG^=epy5A8sL7U;P@4Ae10d1yKV-?ry+pnkR%@t*PzTHc_@A?<$F{k! z@MYQeif!)2ZZ2Flsp{>(-XEGfykFiPxv+L?#jU+xpp!$k(lpKb(DR~R>TjCASCDuu z&3gsNn029}TX2>Yh;4@8$QxP*@j#C9hvp7%Hf9*t^^#Fm+~XJ|wLbJH)TV-9?;I#~ zcGG8zPrX+n59}xPvJ%LByEy5V3Mi^7^aZIuot6YkRkx%fGzT=~Zdsgn@_1*;LRJtm z7l$K9iq#Rm^3KmLnhOw@RXwZ-i+nJJm4|BD$vEb4P-SpDFzbMNiG?1q&YknuIMbKQ zVx~=dGH>6!n2;LYr(U<@ z18(b7agl4O4pB3dWk~YJz?uPIu}h|#`PIHxMOkD!iTKMG4nSSTHC!lk2h=NB>bwBt zYXWyRw}Zr^X0nOO;~WF$^|jDC?tt4hQ8OQ%3s^C+8OrY^0*i7Nf9E8v(vS?Dr#|TQ z*rNg>M9?kXt{!k(EbNwVf`3U9y*-YvQY3TnoOHZtqp?xy<4T+hnMEk0YLa)Qc;QqESgNM9Fk4bM=F zai~=(_b{&@!l~q9wnNL3h;bgavVV8|+y89@i&*-vk<>$WU~%C+Q@qJ?7c-XWC#XIm z+OTYBm1mR%Ty#1DvCB5qZ742PEMDz-HF(3+$ATid3exB`K$lOJ&aP8wmD?l-1NKYe zn9EE#lTK!nP*ggkgWU49l5n)QY5f~QQAtjwL1_k#!iG>>_A1Cj!L9N>dIf>uGzZ8> z5oTnepwKl6@K*r4?{Y8v+gFb71v->0;E#OJlFP?Le((LkU z;Z*VSr>^t`<0r&Hd!+&6)H4!5*BzEK4MQj5#qm&qq(v{#ZiNjxA$!I___7HzY#P=5 z`=n)G^dgKqy;jb_Vv8sTy1DKAWctxpZ-DrT9d|vT3as8)o_o32X4yDhRIv}1qQLalP)trcZf-U;>U>f`Jr;Y9bo--d+mT7!)x`IvA(Q4$K#0NgAkR&_dr8T$sJ^qSqt8=CDMn ziY)YA1PW1X@Lhhn2O5u1%}yszD@!8qD2L1<7w(&VR)TKxg&1xVBdj3v!IV@6BcnQt zsSDVzBm#Tk1L8Sp-X=yJR+G#UNv7h&mNQm=&F78TA*pO`3*WpT)*Irzuu>`u-4c2t zM5I7O5ISPb`TX_UdSnB9Y#qv>?c`OL3eAJfU3MpBC9vEOC#-i*AweT@5IutWH0x(~ zN7%Y`aojt#VWJJvz;TbXX(U#1Vrz19Y!a06M85@iH285~yvc{nP8{=qy25nRxg?ra+Qauqx+Bn#b6&KKn?v?c znv;ckj+?C<;ckz`)PwL$?nP=F)ydQQ zPKyIiD{<$ZPNs)IF#5z!9uA^s2TrDOsl)gzBmr*b$5Xe%(K(XHJK?@Fe4+k-^|Bgy z?L5oE&1Ov6vQF98#v(>%2gDKPJQFS}e5^1>3{i*-xM8f&=!E`Qe3`aPvc-;0zQ{`O zj5({@=w}WID{mg%e92I27s3!us+0V;c{WN*rItSc=l&CIdGn)2TRQBY%IeL#Zd}T5 z%Eq5Fr(0#Cv%0Wz$_hhKE_sdCM$|&(&RQ_E49NEe%)dr2;msd?aN7j!(c|QO0q)7a zS)Vn-sGt0_N998y zoHS{#w9{i%6l&2sy&MHJ8{uQSO7Q{!BRd$s#P5$!jUqYW6OV6EJ#JvY8gA5DMleeV zDu#$QXjh;AbkwV{@gHFO8?7RU-V6*9NGUb0Y9crxv*vJjtN2= z2}Vay7l>%wkf1DRq4WKh%7CAS>VsWOe;Djykg%qf#Jtz|&=o*MlMiKy>Sb7<2o1E! zG{#LqKp!^igCtu>Sfjh992Re<3{oUd(m^gr;GyFh=x`u!1a=`vw9_Fg2pV+50I+1b z6N0|9$_lO_N5${}tRjo|R(LFUD#r+`d~qezjeRU>CpGOJ!{e%~q0x5O+b%m^-_qJ& z^Z9jlu>XJj)mdLQqHE_APA65xZk2Lj&E>iY7V8M+3_+bFqIYqt$c^0c@I8~d0fTow zZldZ$>qCL({QVv8lzV*o?(f!r`}YHG``%0Pza`T}ZRLF464U7p2kqq)k>wsc$RZM3 zU?oRjkC@XdsP<~14dbFE7pk&FR`aTOIn#7e_jt#=a;7!XtAWI(k#3dOaoXrKJ_G^& z;KbLCy$ihq(*y00*`ou%&w*v@(d=?nu#XW^4L{qrfJ${?fmLb(l01URCa7&hG>VGz z5J9{U(Sr7ROrwnFw30hu))!6E$&n+ZOu6tfg>Hgmx<`}*<&!W;p;v%?^@e{8T(Dza zZF|onoqgM$!oC$N@iTV+nBVO(FX8C2<`u|-9OhHpLoiT1p^%793tUWf0tzmM)_J0* zxr>Y)H~d;#6og!Xi>Or9LO%e1^`V))(ye}7fSmK>XL9sbV+_kp{xQLtIZdIhq~1df ze(pp3g8wboNDvkR)QOL32LEPGG~%+bu|ge!rdkMYnwI3hV72ff1*TZhn#cu6=?Rq> zBMqr3<#Tt2YtqRafj-9ELjzPHyhTx`A>O4!1PTvYfYJWjB}qZRQ)A0_Xs2^V_2~`6Dy_POlbu zf#9I}a1f5UQ&KL-1m$Yc?1%32@j{vRQQqgx=If$X%4$^2{Gq2*K_*yC2sQ*A9GbRA zHSq+bKCV=@M_ZI-7f(kBs|6=V!*BQeeBnpv^&7gS)8Pl6FX#-syEt(N5uW zT{g=2(udo_zh~67Pt6Pc7IprO*}gqA(YBfi<_1AsC89SfE<Xq>D=96#INOW4~uMw?m|hN}3!$ry!z&vl;58>cIXu0E`yr z#QnU7zK^&NYbbwCUlKh9GdTVgi``rO7P)sxc5$&s^PK&29?j6nZQ9^DNGc93m^S`? zy@)K{dUcb?2(p!%ZvKLbV+S(V_1i$@j^Qf(7J`AAE*psG4V*q)F`bcdqw)H3|b``tuOkP08xq4wt2&6#~yqP@9(VR!evJ^<}__CtBqTsWWzv; zbn;UlB+a@n)7J>QD!M8vnpZ*IqJc`Ia=g|@bJJ&$uq0x+*K#jSkxw%}QM}yieqgC?c5wG|X)tQv6Fx0=4|J1_^h{>9fN?O<-a%E_F((sw1~ zMXi)IzoOv>c*xUD%X-%w*E{6U<(M zDkh>aeRmP@Sl%kn^NAgor7oGYNO(N}Q|fMWmW^wOx)4<-s|E6bhR_<`R{z*>mqQi_ z&y%OT{+RHU-2X4`-PP~@aoITVUImqiFVk3Z!r*b(1*0y8QIAGsxZ_m^u0QsG;RPyQ zI&csYU>NI?#KJN{9DjJ)XC(RJuImt{tmKO~N-?Zct6V?+n#Y0#3owv;KwcHx{?& zUP~lKIV1namm8=A7nU%oJyjtaDQ9UHLm4~*2)dVvmoB6p= zoq2OmEqUg{t>C-d`opc@dKj(x>2}Py<0keMBe4@Fvlj#|_}GY$45IHF)LC}QiVNHN zJ~P47b%Fthg^q}>k`Bm$xIF5Cq8k3YNZpj|n~iCdT4jRABWVd)0M!(Z1}D==+AO2gd}Amr(irAFNnBehMA zx`f;kay_6zbb%@NIIGI$wgdfjjl^KF#=iBU3}rLbu5RYv;15iWnb@xG;w^_n&>Bg< zqMM7)jhSepxUvI|FRf|pF1!MR!)|pgygbRgI=jnmqQUVwOuv0mHXieF!5+q`B4dTA z!98?^=y>ERs)X$09*^9`J096K3!o6G<(?Cs6XtmA;T}d~n05oQGv+tZQeiAo$G=U?e zA{2u4@`X1cfi{^=)gKw9 zi%_`!Tht{NmM%c^H*5`RCzuw3Y9gXfi7qHoqsk`?{jQ;&P$|CW>GBD>sAPJz??y$v z67TJi921mslQ?@sTID4wMNmHBuKF%t7j+3(w+#2I$aCUlGj&lXgd3$@79W;Kd$sssclVVDWrZ}tcfNYZ z7oCsf-1ku1i3!=p-53d}{i5xH7M~&@5YR<6hoj#y-@gKUhnd`Siam0?Um?0ho>N?x z>Y^&4XN9OvRw=Cm#Fb9J==&MyN(B*V!ZqI5`V@(NM&vL7T}s|XrUY!1;&oJz|92_- zZwbk`Q3Ub)_R}PD_F5|>u`NCoz6a#hbgvYH>!21Z8ag^XG>do#0YCYqJY~)ZVOsG^R1R0R$a9ZeczjI)1K`U{VqGcxLIz9Kw;jH%5jM93<9!w5I%qxa^c;cJP7v1V| z&nsWQcHjbmfea1Gba4gHZh$ zqzhYJZktGhD+F_ipc;@@5o#4c;sLS*Lt_85>ypp-&HO|B1c;5qs<#cw25>7s4%5@< z9S-Ut6Ch6kZySOrGsJbF23{JJ9mpM|ryGX2h^tgBxrVn#spra5R?jUI8QMV{1w0NS zG-c;u%U$kqQ2cz_eV;r@Ja4fO9tcv%@ZfZDuQUq+f(^ac=q~KQWt)QXidHs5kQKPb z{PDNF8uQwyU#J>URw|&@%TPy2ri%lwaxjw@6dXgq?w}j+PQRteC7bz}&W(|r=pZNz z?}^x_IH$mRmw=4%Ky+0NDdlHEVuY(^wUZd(nhJfNh|b~PsX;u9~0ujl0>H}Oy!JLZ?OTJA;@Xf~q^8d?Y1hlz@hzUbi+ z>jRH)OGJx>dxKz(uTm}a{Ru6~BiwFotRi2i83@3WOhGrDMIHh6v~F%3QzBwDt!!TP zu(o!gVOW1>9a=|&h3Km>?~7DB%n6ZP)|+jG;wZs?DJRzJHg)Lx)n8xo^&9iz-))%p zsTKMu6KO}c^PEiRE6bynUs*>6{ZK0E;jZL7ksIXYXg2OKM3*Wl< zU86eu^P>Ogqw3$7t@HyE<++JqfTE<4h;EeD$~!nK{I`iT+o&GSRSMVu=o~2EVm!%$ zh$}&jpxrnG3{PF|hUOrP@Wyp~_)8SsPdjCObnpjzz*&zP{)|=Hwj6O~Kw6 z-6Z|WE)_FYujVi(asX|Y-gQXSH$#TDPzg}q{ zt=SHLR1T>24YENf-`pv9NbU-0SHDi5wjW~_fyd_JEAbdy@y$P&Q#iWp@}xCEO(Ma> z5!5OoT5rv3l*Ue78ltJ51kPsJG56Qa$17ioy*F)}WP3-qzaOjBY3!DxL8a!Y238{a zakHE0mChgoReSiBAY`)BV3qTSB1~#t>eV63n2<#x9bTqa3(&z=b9$xC!2Jn|C4St^ z{J05cLaLwy1r{@UAd82uG+m73F|gS%#6%$2DiXEP8d|2h=%sE0zn#26vdqo*M*nm7p#Y(JO-U!+Slj&R7+>ihAf%$Im6P$aojk z0$5%?E{RH+))cmgw*vB8Qr-6k7_PK1E5==j(nZ~%mxN$z?9(o91G7*iuX1v>`ZzMBXZ@S8|xD=9SW&lI7zr$O`>)p>)}SNc2fnk-btJ_vizN63|>AL8PgUKJDCfL3$u@Z_{XR7bl>Tp zS30K;GW&rEDQrndDMVRSbLs@uyhL8fyqyk-eX;z4;p#}{-s+$ZxWP(D65m+7hVw;t zTE;^dIWk6mHvQA-`zIkv_RW5zIO+4yrxmgfknlklWw_T(-{3a`aU>Ex#B;E8Q;ZM; z=`gF#wlChk2^tAjoJq6I4lBeR9)0_{w8Gx=Z56r%mH~D_X2p*Tj(?3ABVtB&t^|7BX&%_&fV~& ziqJI970+h>D*eh+?olU08j_>IXhmM_S?-6qnyqr&WxXDsBSM)iYmenNul*$0q3ojK zaPC;++O>3f$Wf@`$!Y)>c8x<;_pp^Nfne4U)Jh^6IcU*W+Dby!B|08Z^M}LHj>@tf zKd+r~hc@`te|0kB`A=sSSJ)wr$B`=+h4ya7#%3K|1HM4Wp-bv*icpek7$^R;6m0Sx~FdMEkD39|= zZ^U1A&iacM5_Rij=YlGNRzW?~3}L4FJk*nay-=}nCT+)#Wg%jI)yBuZw%MxR`t_Ix zjS+QACaZNgZ5CGge6i9VS+Kjf`~qD2}FP; z(?boA;BCalBu8WDb4Gu8XSRHUcAB zExCJcJt)9yc+FH7FNf1Jqf^ijWw=r&Z*#IrJoLuxn)IR0PW;VE=OC2*4dFcF5;ZOL z*{4)Obgw@*TiwuV;?+G&-vQ_w}?m8+afoB>#tbkm*(?^WQ{ zTRysJ1rd$(mAOzaS8*Neyt-+prj>GmO_nO>VO@n5prz@8in-V>gQpV=_3r0kiyT_7O}8LG31@|2p8dnAxpvi;5k0(YK3!=(C)69P(&8M2nft(xnVmU#JU> z9d|ggJaoUL6QV&ePj`zK60^mGwEE=(Db1cq7oyV1n`9Pu@i>gNz@v0v6Yiq=x!trK zBIgxz;6Nz6;wkKJv6DtEfBM*Kt`Rs5KUkVYmANo*Ab@=sa83{mc>RwM(MkU21v<_V zUNUb>6a@U{aWZ(Q^W2csLtOELh?4MwLFIEbSe@#Y{IaN)#MoirE`eTyZV)%aV9`|| zE#a-;Y*Lm=GO5GRy*d(IV44bUIWy!9)7QW{MP-7kGJ}UsD$HY5g<&&V@;= ztYJer^l;YQu&d_9%pk;W*ysJC3K(XO7!n3F&HObUpd>U(6DL3o#XO&-s&2YBtOL*3`m0oEje=GLPL#~c&OranAc1u&p=y0#OEEG6wfR4z8TR0(RrE-RsEEb zNf6Dcy3I-AuJHi+@EyXf3|3ny@;NJ6*nQz^t_d!!rfJ!w9R8nWVQ;d!nvNvON>cV@Kvn!tnfEbbm_^p z4h1h$=}Waj1$e!}jJcl;N(>h56U1{+`Ms~t%t6z%J}(DD+^{S>OPzQa$efPwVb5Z7 z$PVG-CjA@XVFc1&G#CGyTH(S#f+Bdsgyu$qNhYXuM06&1W#pD1^bK5>tfsfB;~`_b zHaHsql*gn%2x|PCy4SE5v3b8Xv}FSg8-{aO6IRd|d*IH~z2?!#R)T)JIUJnG>z06j!lMpi0Fn3+@3lLN)5{)Em}11? zc+d}MUGN>a_n%+-qUU0^l8c20-U59T$pw<>U9&U9dnRpi)*QcThHRVvSB>DDd)DXQ zg1=zQxI6xTo?sLq0lzMvPi=HzZ4_dthH0a@1d~NjTZ!np_kh&ie+v}sYY6Y-){+Ut zfLkI_MwUSRwkqijN%`a*kyVkI+|B-p1h&ydJ%FHz8VavnCk_2<++^D~Wwt@ZR5%Qs z+4n@-da;5EvE6U~fBnJ;m17^>_orH17%Gd$4u{Gkf`PE%`$RORQ?-(P<2C0++qjs3 z6&svIF8AsHxAYSV)35f*9|CUHhe8z|vNc?@Ho|ZNqj&KZ7U!uDE$3B{B@q=Oty*KK zwtLLIRoae0o`-_dq=q;x&3&*N;ifpn_c-S>?j_(P#sM@~d$rEJg3R@-2-0Aaqd$XnAUPgcoy1?`aHovQY$fz=DMP zY&1e|5a)}HJAskFZU}yk{*Az71-mgn{ovRy%r`}BG+-AFw6LkKnbu#{SABuYivwHz^3+)V2ZK4&Ngbm>Vbb$fs~>Kn z8kATF2vSV7P)7q>9amOE4(fIaZ?2>_(WjVhZW(zv*pT~xqe8l>PO*|6dZq@&4|*Ih z!Za_un6px*Rcig$j|alY>yiOEh6nWoLSnhHNpwqwvKEKNEOM>#k>83r8nn9OpM&%{ zmss?Fs4g3~W2LTXD``ka(C3b0xYj9ZXLzG@ zwdaCzk3P?vswzEd7TlxH2AvP8<{eRtO3s2U^5*3Q$(fSIFgB|NYiv%~N^&GDvo0*1 ztgy@;7GIL=632;=J?Y@QI5D^rVrCwJ1Fe7Yv`uu^lr;V=pK4CJ7(ESxZuPPx|6Srj zQ4e^?dLpnhkOztzLw5&(5O@%d9rs1*shA9dId<){4`eUH4z4&9d0%_)`|nROE^7VE zsg;!0k}=1Hz1R>XI&61VNH8FKav{=~x*@^Rhndt`-u#{1I#I4?y>gGdS5ijbnbEH} zsb0^6((PrlkGZ$de^1~JxE0Jx;Ff#T!q4qwUVttNEk4+;3y1~#r(;2H#{$Cmje}={ z`sHu_pXLug`Ss8K`&W`>1ha&oVl2+P)r$@V#Oq|whle9RFkgq>5XIQass$W=>Ubte zxPfVfh=**|;b4sZL|TnCUKyMt5Jh@h)&i`g?Ic?NoPiQc6Ur1}ZXdB%(LJcZxdicYu5pp-F|L+#IqMJ=aY%N)Y6T%pntb zf5AH5d&l1Zy}1OU#3!A*r2n=WXiJ%1}c1H z5Yf1OK_xbsF7mnNac9QP@C8lOJx_ym7;kpdhh~C-1On)~-oc=oB>x2m)TryW%M;!8 z8m}`UVAs?SEidNt<-l`x8g0j=&jLYtLtQb{p29^6Q7J_DKh}Yg3?x=EbY-hzL zNU$B$0qt2qV$9D&-{p~vHY79(d( z6qYK2DJQ4{L!2OB@5HSOP;vSnQ4G2r@gH=|B#Ccp zPQ6_atU2-48b;r(B%%&-DK$f;I`k3f7(@0EaP9jI$TbOqEb{QI9YOlQV$^gl`rd=* zj0BM^h0W(;$CjfJtY6=WI{U79y<|4qU>7#GS*flZLo9qa@9{j%1F)bS2z)MB0Spbj zg8I2dOwYHVGM9wA0tOVFk^FiPw6~YXrKiIn7Gd zzgm5yT0Mk)lm^uBQmI|u29gCVdX!2vO~A5U`pN-fmPL?Qy-BmiY$x2TyVFjs3l~cv zxgPz`QYfRqDR#>^LoFXiL7?j}-|7Z}Ng}AVP;HGYo2^xzaF12&rK>q{-kM$tFYKS# z3`MQtys_39yH93>uYk=2jPOycy0e=AmrVvb{^f4IxzW{S1&0+gDw@wQ4OD~9Tnw+h zA%R1TUPV@weF_YKT`JRhj*e?ik0Lgx#if81^hVH_9csVjt(KVYaII_&(Ry7stJ1G{ zS~*WsC&*Uo)B`h-uxnI>1)Fz2g12a2S61)db@v0u=KrVvTgJ7dY}%ieQ#;rxuPz*J zbi!mwJV-G632HA9ebZ-uNXyK9)cR1!79^XgPV)ZD+cTe0I~g6&Wi<0Y{kZU>Wdf~# zIj@G-GP49~7#Z$rmB>JsMHb0-g&Ym8lj)*1`ETQ{ch6K0&fF$B7=Q#oSasrfBvL}G z<(*SI?qgvqfL^1I+J|@xy%Yl8Z_RhZF1y}X+4|Z%CT?SeB-E*2?ke zb5hmmPf-_NrV_Qeg_~{!S znzImEq3=s%3Prtwd%P@&(86@7K63lqYV~T*qv{<&Wk|UP)pfb2Ma7)!oO`_Wp+K+G zLdSXUp9rk_d~kIbj1u(_EQ*~mcR;ILL0~M<3Sx`?=My;|V@{1>==QLBSrgqAxdgnm zm{yKdC7RQmrZC(yCDWLng_rH{mger&*VxUh^?9XLEQ?|DF@x|%MJy1kOlMF zxG{4y`C_E#!EpJ?S<6%dZU-hVQw_Qy_R*&HsZZN<%_fx&5_W*hCtqx!{?}mYPBOg= z!eX^bOv>q%>bWotHAnEtcsl52@Bv`rBaW~2F9c#k4c^u&x2yFDA>Cluta28Z@(iP6T4cU>pTcF6c-Vat zUw!tEpO~9u-Vh(N70VJhyJQxRhff zTZi{^_VNB@nMcj};n^U=g*7EB%<-M$i)MF{i>d8kxoL=63iYsWNG?Qe_Ft<8B0_bW zU{qnjrdpe`pwQ(z8TRR$Wnawx%LSHd`3r;y>CA(eMHpAKawTWLfBIGw0 z(!Na{`GO^UHJQNqQ-Y}@s53+~=qTI^swztkzN!b zCwwArCH)z_TyY7+$l6F<)Cu86=_26A>wl;%m;&nxQ}4jt^kf<2F6-N!l>EN$;HzLhafZj?Id6aVYCRI&^2 zqJX-6nAaE^>=D1J7<#Xs$#$*HeApo z5zJbGiYKBm0R$;Z_6EU2K3$G}CPQTiYuf~2^+=7!!mqY_+W`>YSwHeXDR(+=2)jwOG-+;2`j`Gs;DrJLYpEy2VS)M_FcD@QMy z(C3*<*U4(hLNeR`lvj#9)u%>0Pd4FX#7|*AfOVYsAD-`J9?kTIn9i)kKBaMreW1oT zR24rwDG_P9kp$Aii}jt8Z~`jRPlpsbRQqKaY@Y+gtf_S0`~kMEcNy5XOz z+RI-T#wzOM*ChS&B}6m7TG6i zPcVrD6^G2yNPU;VX`#`vj85f5?hfg3;Su2|esjiukd3eNwf}raH?S_TE!Y!W*MD)> zEq?M0BPxDZ9#u^hxUgD!(gd!D2&Rmn_JJu?{#cg6T{#C09cAK&KB=5|ahAF@yjb;| zZecpY`lJ^pwSqBKUvSmGPFle0rmJ8X*&DP!B$sSqAR}Gf$4&Fzs@6s2a-i5MSqEm3 z+fc;p1Qjo?la@?s=2uBu`DdtN$H@&FaXlm!Ui;fo_-2Re--u-?=62LKL_4u&bq&4k zGr!zF-|%aw*w_WADAw}Ux!_gP^8?^+t$3)|SYE)Wh)jFXkT6= zY&o@J7(p^dBa;b|m}9}PY?U*=e5*5L+ss?-?z9`$)sbw;q?X?smlisU7E%vfSaXdt z33MDJnCArbl!z`7m8#DRb_XpM?hVpxqLFU^-L}X;lBLEN!d=`gKAHmV31IF+D?3p0 zOED>FpT_S=UG9_NEoV8hnyl;bUE44v6h? z$Sz5;Dj(GA3fOF4;X)ZuL$$(sS)XT&uyO*J_}nW*agmFK*CP{weg|1s8kBiZmk7IS za->O6d;SfHcT7+%+VXlXVi%T5R=9|JBQ7YCU@2PTwbC~O>?JEyOJ(IAD^%OS@V`mb z6q+N=qGi7eeec8V6!X!FYEVh@WeKf|PY@VXQn%vpqw!?7*i3zv^KDml0R^V;92l`gF0{>*C|=#LoRPqo_FfjSsI=>s)vl*=_RiZzUKoo@^$fA>0Tlq=9f4 zK6~^_cF5wrV}yfl&HP?L+3ZDg`UIZ<5o{A|ig9;Z5&E~)oX9+OHT4-}0oa3Q3Pb^zo?fGW#aGhQ93M6!;m z>KRLg=35vx-ar>#ldNob$|U#boN0yLSLPzg+ZvB@UaaD}q<|Rhs8^U%8{sn2*=ODV zmmvpc3XEc<>?RA^-}?v# zGH-T6*-`o9uw)2~29`&CBz&?y~!1|(99+&q-G?THsSYEym#=w4P^miOg?0~{>SePH%GwT-nSYKBxXe(oz z{~BNPWg~EQPT_P?RqVuy3r7}QH-T9l!GL0Ol88PlLElN840ySeIUd_21*&oeGrUTI z_67C$oZ;j7fE(UHygb+6FZaOnT6qy^IM+q>apQSuoLgb3QwH4Dc|MnvLpi8gvRz&_ zJ6^n(Q$!kiU?$jN)h!5`O5+?2E}h;g#e13R7EZHgJ!nqbxMj2PzFnwvFaXT~sXw=f zu-9CCUYs^7+6bv9p?{WB&FmoM!m4qMiT|;SU>*?EXGHW#IVN;qjSbXXk-j*IY7$+K zTpy@aUiGe)4!R|AfwEX-;19ww##;Ug;fkqRX*@?$2bqGIf^HgB;=OaxUkC-(<2i?B z7Rw)T&nT<(RZOvz0_FvwWYjzkTjY&B1_N@)%bb1FphENGuv`)^XYi`{IpkUIUTQ1U z=;#x43Xo9ww(I~nkz&18@~*;2El(9CX{|*8-12Zso;7RK17jTJT?i#fA?1 zy;KW)G(cw&i7#C=MPcoqqvwsJ{-6u%%vS0TwgO|ep%mSoNm?c7$_uc@Hcn<2Z%YW! zJkZ-D-5_IHN#xe9oHa_?v|Ru(olB>j{b|hZU%fu|#f4>$mDrc%!b%F<#SHAFxy^h4 zg0!4BTH3JbMAwm%(_A(FO$4K-MCV{)0DS{%6DO{`M8NXbAVt<32F}!ogym| zuZp_KzwT2d-!8ZWIU2Vm?fh~O0%fxy&PClRKq^v0t38q`vRB$f-gu81eKUHnGWjQ+NHaWaL#N(&m5QWFOsJ+w^YE8rw;p*ACsPmqe{Cj~f=LuCyu8AlR%pV;9 z*fiy)GQyr2WZj3I(AYBZ#0OK1D^AakWy`3oF1+GYnXEX)1Oxf(1yBVUR*PmnQaUz= zm4t)+egkKbu!>v;MXu|;AfQbK-T)L4SnpWRV}Nde!f-Ul=!10A0+I3*`?r%2+j23q zReH}0j5I|aMUJ-P*#R0;2n{=9jf_`xUBXKJ41CR2WvjM>4MlTJvNL?KFhe{lnyr1PdBt=)=$3WVtWkGM z-Z6@mnF~K{rpjGdtTdX4l`{kb`G~dXm_{Z-G&{7Cwct^#3eKC-4#r@E?YD1E5*JOw zC`a1E;vuKD0d}_FV)VI3?+7Z88t&Zk$>U`3&~bqK-h7C}?jW&bu2zX{?s>J)xIfk^ zu{NPri9!lx(rV9K&-3sbcgaJ4fxU~Geb^Z$W?pw_Czh;Z$AWGD3Ys-YT#sW^0|3aPQg~YJzT3k^G;p>md~{p zC2S_qE<~|;fSqmsb5iK;uNh^|v7cZ4jH-EK%)HGe5~_h<>Iv#B=JP{p2A~3hV{)d& zi?LLR* zHxwQrb#tuZt}jT9xmnmtxvhJit~Q4Q0F_1n=fp#XSc>4Rn}$7tM0JHUt9O1W#(eX z%J+)(KFU4deeridz3aykoQQN%#~_I;2xk%UW8CsF+)H$C6peRQ_l;Ykd1d@|s+!%F$c4AfH%x$C zPcUZ*>J%iUaj$ab>woBgx~G3;_S=bXX9uOfeN=i}xnxfE+v#uT2WXYMxK$7`o(B{y zwd8rf8p%=LSViL7_X9DEYOkb=Tt}q_KT_UpiWpyCvaipw3{I3GQnIVs85OLom4TWip=NkeCw9D ziaH0x*&Sg6Zg}M>lOfq7IVU^||HmtJ{Cw`N*@yYJB_(9Na;RmgcTwaXXxBE&(93W> zbl1=p^AYyQdnLPg^-64aTY@8h-p6o0)CZqE^bCAtrP@D42vnxBbaQ0+D1h05N?_S(ZChJ95o=%fkcxu?;D}Dl9OPuX@Qg| z@LU~sQA#j2_C1@i(A>1W@5{dL1*F;x209EJJqZ%oSSO9}nQ$h8OpPHx1zU+i@ zVMJI-h8@a#)!;7M;F3g3+dZ!mc`tYsMmYn8#q6`1y^%JyE<vyRkE= zM=tDoT4xf!yL2qGXe>2AM5FIx-ZiI@ zxgS^$@l{wy{!HWSS%;GCK7!@&*~&U8Z^kvEyz`&lIPsk8(0pZBZFa3WQ-+ zaZg7s8;5+EZL_xVl6jA~P!gKcOrE6@d06CmzoY2oa3#U+W``F#JyGjtr{Ag<%Npgkg)*5rTn8 zi*h2mncpbgJOzoB*G86tx_ey+AexXb&hskfbdar**uWZYF3<(laM2E!Wr`nCIf1SGex80vXv0R& zu_#Me2re5nW!8F*<-d!KX_!cB+O?lHqG-d~;LB8s3!`X{35v1_23EKX^inO>d#Nz+ zABDsbZVRVfzE!F@4tt|*o{iE{$VX@;s~}$Zfv9Ijs`}B)EU%j?$nee=R!+Va%EpOg z@s+>&;z)q<>XAow*ZK-)!9hMI@E6?`<{>#Q8|Y^x{HB>-%Gv3a0)i&Py^vGNX_4n~ ztHABJFLI~XDIl(k5h4?GNkpx$_<3+|cXY*jj08>o-V+Q#jsPpWGJ zF?nelUCX)Qzh&|*?nUliIc%qd>pWZex!{^=_B^BP6u`bKTh%H>v+~k8S^k4=I9|8B zEAqKSgFl-2U7*EOkTp;ey@E`f-2*Y|d*u(w6VxL?H#gJHLML{hiPZuHjcvq_X){)6 z8dK)~!#y7u(NyC7{3lfH8)I2)HNoB`f@vVAdeF(|p;nVId9!;#M3K){=~=3BGTMSN zCg%qnkGvY(MeUk>UA#wPIDhaZjK}DpH%fcM&WA4IeHOG9QWs*!8QP=>_DEKHo>jNa zY*#PhoK$0rhN!#q{`llwEGL{X`Iw+i+UBvvE5oCeTs97ref`kGn`cb!58Ebb;bf~b zxm&zqy$%LEU{?up6ma+gx28yY_1U!k@7^%4XW_CdnUz8tJHjtf8#w17Jf{kD-C9n{ z!QSg^012rY;kkQz*V{|8wCWbSt!q^$-x;DlOt1bI0&LO%mr& z@S$L=0X*ntpcTIk?kXqmiA-nG7QDO=_4X+rrVq67%sO7C6n^LINK)Zk*lKJg>$YBo zf#e1*@yc162cp$9?w$V!?6NgPS);$DM zL{No8K_1}k21_wmQ(8SwM<(+gPK3p$HaMP>DBkT+K;VCyy(=fDi!)TW!{eCS;r)tK zcg+b(tBzwDlnu%(^)s?E*x-$V%SVG7=@z;K*dmh0I}}{D!RB)@^Qt2UqGn8M=E3D( zF~a7_SHJc%>hL(~rpfz0PcY{Q>NF72NeASIgX(2nGOa&QX4K0vR2Y<(MdqsZDb6YC zz4Y9s>4DjBp^~ciTBwE4#&oipvxb|&siBPBwyXEf-8=V;#~BZ;5(h|=EyUgKc6EDL zy)4DEj%=YTAh903L+vDvs8#lPo>MGbxX{4akC&PtFdO40>t)IGIdKVjZej;y;U5LykOy+ULTy7})*}a}FFMm?H#L1=XRA z6yQjDR_xOXF}nD(k9&w;&efb$Yt^_09pP3&m;)YVtA>v1eGm8=t|j?b&N{?Lj-nLR z-k?NPX9U)yNKkB2B`APKE19kk-6C@&=lGd|15>a<0UBtu%BRvUL6UzfxnLx|Q0F4t zc0`-EWQFjtSL41F_bnrkR{ZwRW9p0x1L>X#kggHT6@t1%MC&5=_;ivtBn?qZ=hP^+ z!;vnk>YLZ+p(R8Y)dD1*J3Verek8aNT){cS-|QZ*I6+Y`$V zJHqkG{0i=KNhbF=**0sFGG)3hsx5q-$3{hppp=^&nc-Ird-6W8;v`Kj(R zRQw!63tdzW_ZHpE?;Kw;K0Y+fT^AJ}`c$?gWYzReW@RMYB@c>^23Oh|a_U@sS(2&d zNV0s+3SWdI<@f(;PUH9TCd-9`y{u6C-I)G#`UQpIXEI$7u^C`Q=NL&6e6+j=Ow(>JFV5Srq<*XzpM=aTL3UY*wgfrGj4+>pd`E>Z>a zG$-cO&113FTAnA9s}2N{$+#muu@0sQ11JBVdBWz)z;WT-xs|j`O!sY+=1DY3{vG5$ zDoBliDeb_-E+0(`aFkV$$0o(|3aPsyO&f%WKL1j_7!*3l(scTc-F}RW6a}yDZyh6h z`Q%aaI!i8V)3Z`jDP4SLMgm>*Zui$eKWd1=^9%1C{Ylr?+P-lCga?+$sDf%>MUk26 z+cP0MN8byhkB^gv;~I(Kw`1kTi978M3i=iBPU>zJ5^m$^N=6apKQQIjP{J}md z!-ch-3KMN-H^G3{JD-TofS{`UNvk={{Kv8~&!)-8goU8_rL-=&pY+++3)wh z-}v!`sJrqcNUwY>J2ojN$k1l{I|n|zH}B~ONB`x@ybDnWgRV|bnzd{u=Jh%SG$V(M zosP)Y0E``EmL|OOTcHs$zbhI4Cu)tQXmuA}hw@Eew3%R1QQL|hVkJ(dk*1=6z})Rt zFb%YEW9B5&gF$^>=-9?}2tr0jv{8D^bPT_32V{nKciORa$V^?m@<--1RoDb*xv){t zO0^Ya8aYM{$cH$P5I#xL%NSl<{` zV}0f}wmuZtCEG0;ke^f6$uyU#GUy0w5pZKC5QA!!`vp3AS>$fc$tmkXHHINF)q^v& z>VpAQWYd%jQ*O(ahG=>P_jv0=Tj)Jg`n)ul@15G@%ka(?t*unbYDmh*z-0_9dN4-+UlPR5y9(thV%cQ zy*Ght@=VvpeZ@B<4~A?6lN1yrLRqYAVL&Wu+u3K%ndLv{%zWRR>6wnq;+d^8lhZk! z4RK#UP*4F)K!M1rfV;A(2$d>eQBiOK6VO@|K@q9Sci%~DFrw?F0GgL$s^ ze(vX9u4~!+L+-n2EP8o%&F1N4G*+P-uGl3!>i=x-K;RWO!|{^nD{i~#irF1}!~PBa zLH4ujT0<{{*0G>l4?gu2L3UBxdKqqP=v%iBd|rYZhAMJ@`eAgc`ON0Diz^Frm74w# zC?9}+rpNp6d=%Y4qKTJ|-^;#78_rv0OKf=VL~pU|*o?D$n)-hO|0QRZ(eiw|=x4to z+nw0*95Jyx#gq(0S@Nm)T^?o7OQNT3?BWiRJTDXus{wP9yP#IlDMfkrQSDt|ex)$;2Z*F8GOYD&7BM_Q%44ABf%ZrW1M3FJ>{mkGJT{1t^ zyoka=^?{mR5mG8S=Xv0pN8fB*xU^S#Nu-0qZO44OCoILA($DXl_#o%i=AyG>R%nUriRMbaRu55-W(53_DroEvG2+r>uK8Ljlv!YaCEQrrt8gqR`O67KAO+%l41{pi%Y5+gX&vwr*xnIi2En=wz6{QI9d|7yg^?|XjVO+Iwu zaLC6dBi}*EphNZoGBl+5U6_Lu%^LQ0zZ~~_^wBx_!Ku-$aT(OKQOdu^+342{s>aK_pGdQ3XOMgRF7AGE5kzcv(?~X?VHJF@{*8f!AH4D3 z=~TgYyWi<{y~oMmpmbziR0nGr>#=8B96kc8T{=nHtbSV0z#4GR(yc1_$MZJ&)%seA zvk3sk1U=U8j6uZqPw!RS@xnbThxP^EuC=EaqQlz9PWU=z=t$U{QAw-ZE-mB-0>a}KAd@mo8*p7 z?{>B}3=sX2ImK>Z)<1|DqQ7+IY7pOOQr2c&{{=~6=Jz^rElZJ!W!yo@GAXi^iZ>MF zLg9iT=+i^e*c%~T2(hC(VJpJ#gr#tjSwo_MD1)T1qq05bV{QYQ%pc*9eXr$1`zC@9 zCtep>zy}t|q0Umf=*Vlzlwe#WRq$aE6pZ>NP_A9+)8~3jj%vktm7Nn~z+m&9o&9t*xw<_xVolXa24ChyYI+Z4$N_ z;bQ8=C|1L3#+l}D@Sj(cA}5|{J}{YSj#9Ei6xmP3qkSbtND_EDZ(L!gyk5$0jj|1m4nUtdAZEk!U5Sd`FK%7rRG8V7e6fy$9wLu`UN` zf&nt&7e%`~nz_CFjqKrA+eY}MU;o4*@Y0X4=>kRuh{{^|ZPDkQ^t~Rqm7AeZS|vxJ z>SIBf;Q3-&5c>-ExgsqT6y&3%AU1GRy6yDa6N^iw!GNS!M_el1&e7SN!{7;DmJLWX zFsos%*-)LWfpGA4*H6O^cx$DGs7r#|8VVVW3_0mF6R*aeIEu{ehGE|(=#06E5sIeF zOa1-4V51ZC{$I4m$ZjWgf=-z@L5C>Weu|U>I}qz*DOR%7i1rJ*xZM!>PoRKZgE&@z;P}xfbHID!&Zxoid#d3c!{D(f7V+6$JcK*($#dej$jE=na#G&h&tVIUMW_ z8x|=?0xRMA`%!~Y_apB|De*W8THKH9h6TJ`F@%F{FC0x6IA#y9jH82lU#X{!m>K+@ zUn5CmmN# z#k$AY?J8~x*Ueni+G$;13o_MR+;Z8cbdfLvBy4JZOCwt24$LnPsgvOL`M4Sg?+ubI zv1zcJNoU`hm#;Xhcr3gw9FP z0Cg!aaESM`#odG&JUpKH%GGF1@CMH;_!P`!53|#4pnbL)C*wOX)*fYyAIoNCyqc*0 z`e5}`AET*JB(1F`JDk|FIcj1&N+=mnacGd`b2TSjTIBy2R@nNE`Cjo5LGoQ}5K(3* zQaBr8d!-{T$jE^tQHP-d;)Yb2GrNnM3(+1;uzq5qmFW0Za|*!)w)-3eqi7erFh;}6 zZ(xL%DI2t2Pd@LOC0kj`cq=(6oRSDV1`rZzI-kCyFP~Hko=TJnJhebWuP4Rf%0ymw zw6c%?Sa?iPLo0I?S}6v2E_i+rt)Y=c^MWVjB8r4Z11|toL$A0k3W|NJ*cUwW>1xsm zvfA9Oc z(fS0focnL&loMMYt%>!yK*>I$NHZ0WR;4YjCqxrG2r=cy;*uG+ITbUT#mh+^y%XlC zB){b$4U$B%6Bx(#dTaz6mB2eZf1m%okTYDE#eyHZHntzk-tnO zZ;%W|?I$ZZd9WDoiBKupM4u9z51Sx^KKPo>=R-c4t}3o3pqzD;b;|b=@CxDc)1|B3 zv{J(}ko<$8Et@1@0rbp}^O9Xgz@q$w>hFgdfweZhe=jNd!g$b)CNQj|WYCv?kcvMn zye&0Gb%OJux!_a!hS&YbV~T!KM3>PBHhmcu#-(AdghM)77ohBOHRM$>i<&HG5#t{6 zV;`E0us%uB=7Y#3EYIL9D@g`2Pl-Ha{`GwofD{(glX1!4$@137lacJRBWE$lmagY5t6u9t1u_}j+J~env{l|iC&>=;>m}7eJ&^xIhA_B8j$CFd z#3~O5ilR?+eZ?3B*51GjEm?OyrDR6O#r@O4%Vet)d&H1f7}NZyp=7xf$)TV#5f(YzEuAF!3j>Q{ z6IiGySr$dMVRe^|#(K}{Xaj9!MiBN7r1`A~$L?yRX&<#5N9v z^#J+IYq9Z@nB5j~HK@N*+@L3xGW>!GZHDZLBi!P7vYcXGZ= z!#|PHSIX>Lnb6gY47c;zU7tG}%FRuO)2?bP7y})=@VF`gz;^HrO=!SSblqzi54#L{ zEQ;q{%F8WIf#zc~M~x zn`#T*KAVOtZ%i16`E#dx`x_DTyFDj2l3h-`ys0(8LEYN7xzTN} zVuiQ`O3+bU*50$J-Ec7BkvDw+BfB%@e_H4M=g*C(`EC1dZ*pTASvGa-8exEv^--jk zicjQaK$+vQNX)Ha@cKB(VJGkgqgIGV;6+o@1^$;8>37M?boE;!T zgP*8U=w_yH+T>+kDmfl$^h}4chE4*%1}vkGlbRqi%V4*`wfw?ceP--O&spVOxrzPCrDQo2QGxvm(8>(!BGh<i_}|_ZYAkSI+jdNwJ;=_xmN6~+XmO6Y1kx9z#l?bP z((N~pe79Jn4y*wKrf*B}YYpupM_r4iO; zR%_fK%z)=$9z-$8Owk9tX7NME@aNeMre}7~&N$LPuIlLFs0*)x zKDTtKA^No5L;2Tbyg~sKSJBGd9vzT@s*ZL*u-8hwJTnaAZ`mD0mhrOs7xD)4K;H9s zabn6<3sFbx*4-w$9fk|$L9#-ORSO@GWI+=UPC+^;wAqGY-H8kf17u9(V;J9S85xV- zq0}#*Uz|7^V_|;j6%R{sj%kI!i@ti&U^R|G*OG(X(&q=qT2wrLr(=&;Muq$CPxrYQ zy^i;l7yd}rjOTSYvAxJQ@jA9qvUG}UqT(Uk7P^9i+z(nwq5r0kCP8U%Yuu+lYI7AdUrt>C-JkByF=PT8MYJRCPA24ZHV1k3m#DZlUjME`9g z2_-XUWJT5mCCMB=z^qVxhkFvTWIm8c97V z2Huyd&<#+QpUmp>!WL`vTJJ|~6V*lxktNG`4Rr3L5dt$tx%~k>?h`xTyEe<*ihM!f zVZj*g_P{QWBa2%XzV~op$HJ?^9!WlXAPN~hl35o>#`mu;Oa|&*+#ZDXq7w_>OQe!n ze;&QgvG$NLOh&DtgFO~ev)W_6(GhWf(2!0Jeql1q=S|#}dP)XN&9ziKc06NRD1G`( zRxv0+R6)!P63T^wBLUs)L^1$c5mj`PU?n7*y66MmebB~{r)Z|vgTTNV=u~VG-w|U2 z+h6o8jd=?F&s@a;Z&2w}bhCGODf8e522lFANi!gXf}etUrF5wZvPutu@cDRP3uL`) z4XI23NzX9#!mV-b$x9Cwcy}M3YDChC)gg6|fpuDfak9e%8yhLv28yhu;y18M`J?rX zP;XxyJz0dfX5X$gV{CTIr0g z$N)$bp(hQT2^zZD^-0)+uw8B=E(yF|pjE~?6FqK5T(ZNn!}XmVEn;l``AZg0S;!s) zRv_R;6r}m7Y1IE-6^Z14gj38P;EuRdh2kgYDYAGvpe;(5?vb2*4ON3jT&{av_iFLfMI^}z z1+DVtP_%dOm?2S@M*^>3yxH@i7>i+&WWyehJt4?*lBnrn4`{os3O%YQi8v(OiA?j> z;PHhf#UUr*_p<2G?7ehi*5{qKY$u4yEGJEZ>k*v_YJQEeR{qv# ztPZ{HAttw-I3lyw#4tev%Me8#Q1SI(Xv)1ZIV)oyvX!TNSAgYF)5(J6Qj~wrrEhub zJ4=B-sM%HJw|dry3nJ%+SD$Q$*X0G3K6>ToRu-05VWVOq???ndP2Yy3e`KMcn_We3 zoWGZq7@+g%0ius)`bY%S*$hRM^5KzkBs0sB>U;{Jmp>g09k@ARI?)Hfe6^a>#no{y z%|HR_lM6Sp(In!7o4G4kAJ5Dd_P@TxeE^DgmYnvvC|R=K*Z6NRsGp~BkU0WuQYR_% zIxE!6k(ROAp{C+U#Lw-5J_nd5M(CUJ$Goe{ghup9{_$oj**M-*>Ab-RDAOC8pjK0| z?G(wN;;G0q~AS{u86rVJVSIMtn18IRL*XTTTkjZ_eITIBvvenDPaBOzu!MY zwoHbQl5MDQfRQqLq-E56E9W%#^UkYy5Y!vj^Acij3bWmc0xR7x8QljCM-{CcayRgG z6hYIPT7aI<1~$xJrzKeYhgDJuJaDYAN~*=Pi`y$6h=M9}tT}8KAqg5*8se48z3dKF z3a438!S9hAo{xs;4xJqI1j^3MYWd8*_?Pb5h8XJ%%S%7}1s7OG%s8{$lpnPvl`J*d znF`;t64En`tTdTS9#XRV6uC#mZ&#)U!H z$;v~Y1nd1v*q0aFjXDX%2Iy|0+o`My#EQ>Wc{e*D7W9t>syzaMcr-yAiOYJ09lJd!53sA zaoQ#K_da<$nPWs=aK?8V$f{`sx>v_kHEp3}n<#KAi&48ELK(CRfSn0Vrv?~mA z&xyPO;LdbxsLuRlWANsi-|EnPW^caJ0%Mb3-rQM6V0^piXTKubrx9RS83T-BN(SPL z`BZ$Z?-TEK5y}goK|;!1efH77Vu(gz1ZccdI1Tb_4#+Zk;oFTzpF$<=@{oRgw+PM% zxK@7Zus%8la2bDga1dx1!Nqxt-rlS4-D@6HvKR#zZwhqU7lBv|H{~kMF48aNR|bvv z_mKks?wK%k9gw8WR<4zy;chzb=5`|NO?C=(QEm+=}2$SUB* z{V1JeCkui+cn;_5x`mZKxk4Sm8FqjTG}WuclO`<=gQlv=bUrz;;!>O<1<7CNt>+DFCU^qY0+(QHQaKqY%nqinzZOH zE?!v`m?^6P2l0^ziHXKKYu_6iprY`6r}qJX7Sv0izoxrszB+ zJ4cZwAkCkf$vY038;{+yITu9xg(aTtvg)8M?x5Tu+$Sj}2j}WUS$->LY+)ZKx5FA@ zZijV8SGpOl+b8UfUI%MTkbzmKEL)Tjw2|E!w>>a{QZ8SV?X{78K29aik=f>rLS=ss6wfU^}n z*4mdb*D=DCW%JWjyFJXjlG$k~2`p4C_j9()>`^>m55#7^($BfPKpjxUKQGgIl}+-( z5r@P|1TY#sdt7W80ps}0)WXLz%cMqYvo$s764}D6O2mon&^{A8luyY(>Mom#zr3JH z@b2T7GOv6^xBDM(uPv^Ee^^if1hbo7Po4wh=H#|u8`PjR!Lbsu#T{#lY9OI;Vdf{& zi+qh6GfTm6Os#~~82j*Sli^1(x_mV|BX1%!uucpV3pB7@+)n8u$pAZNc0XOT=&$en zk4nDbE0-6p^&FJq!6A@z?-ShyrU}rn<5x&fd#VO_2L?ee>~`c{Rz=`mR<>)E2cRqP zK%gNhw(-jk7QUDMFS>=uNxW>vJz&W0M2~Oo2oT$ z$@_w=?kzOyB@4>Ebab!u5=eA60Z8>ZJ9shR}!BQb{krQMMLpZFaGBP`U}PV81w1gyH8Qg6}4% z6Y$nbvw%MSSdfms=66t`gcI{qqfoe@(r4cs{QFD@#8uLzI(UN%1#3a9a(KG(zoxVvxMUwm!xz{1pTTwd4~S01q6zYuyu0c?`;kQ0ivU^UPA9CF{~fwi!0 zaeLS`bR%i?9(_uN>wf?3uBYjWxYoGkEZp9}DF|$fdmP(46VKtbxxo|fHJuRZUSvTh zh`kuErX%LDm@kMkY$0+pI^{vK3n$DLF*1q)1wquOlOMXV78Mg5WX88I{H6Hk9!Bf( z&K0tRq&jh->K+q&kweMAtIYx@7BWjM+zMZ<6iGo3#a?mKYk{F(tCD9642oej98K^V z3O%#r(AotPF5t56i?(*uO<#NKo;{JPn9V?2FLwC;(1;P$ZQc`dp4klKyj~deUyPlB zbd;=%BG;(+GxBU#L%9%SqgjBb59OINu*w)~!lf7>t5M6r4nr=W4Qc z`Vb_}@y;Ed*>1`d4jvdH*=`wglqe##hTM=I@NV)Rb4{moh;)!b-t~y>uKBL8;(zRE zi2LjfJQgs%t8W`BO))yby>Y?&@&C!)^m!*RuJz3bDr2n^{T-{)0U$bPlz?K>t1-Re zmKi6Zi8RGkS;0T3xG;0?E0_YWi0GM&$;h@LA9s}OdXSkRQe8;8_dBB@`hCkg!Q_+i z(v(hYZq}I;$3CWHk0>%o#qVKX;cuRP#SNKHa}|ZW?CH=#TENQYbP(*)fjj^N3!qEK z(0hZyfmPux;sf4DLEPrOLac0nDhNZH#qE%02@<`R2c!_FwCDk~Ozg)xs;J^`)VFZ8 zh^s=g!;zrNP!7=*mO?7#J|uhC2YEW~22X4-Q_0bJ#xUU-k~I@-3U+qjg?Vslx5v8Z z21yrvM&2*%WaIRz#F86q6w3jlhMiveKB9iH$}cJWcwi^L8`_0B;F4@Tr7l(t_0USE zEz7YC$&DuxPMCLrG4Y_CcqImO*<)5>rIf6QBD<;h5&vv(p)~9PX&+QRXVVSx5&sfl zxBH0yZu)S{PAHUbkGRe4k!%k+NAIBxXFAE*8CU3vsE)a~-!V4_t^(m+dR64!!13p! z30oA~tlQWrY{I|7xDWVsY3gpt=+b@GuYX62f*OB+&B?GZaXi(=a1kV_lMHcmpHRvLQ^uQ;kG#zFA-Y;0`C z$!AY+%*KwXvOoTFX#t^I1meyp_5j{Bwm%zsKw$2JxyjKuFdIedVPSTogLf}+sT1aF za|=sgE$evDQmXX`DNZG4mr%>PBi00$Nty@>7NXWI$TJv{6F0?ZPg+8gJTh6AL0KRV z;((ZwFuK!_fa`)IIrJxidF(DWc9_8#>=tVE)QDiI-^bO_+rwcaPhohU@PKkTs6i)l zdTb%;wZa2mza}hQctg5p;R8q*G;pWbY7P2wOy_jW=qS#Ul~1~2ge;gN7)LKX9Mb(t+QUQC(}lYGx$BfMv}lqfx%tOaXZ{Jm=B4SsF+l?TKpXcf1dY=r_oe=zjnt(ve}7~OJyc6 z|1L@f)?fz}pBdLEYZoQ(+C}^Qd&T;aIgGE>2Z6$N{X*<^Lx%(jM6i4gxA6{K-Q06* z(Jp+y){gsgzJfV`?7Yv+xBm8PWo`e*h!P%e^LNNaCq~Hw6O{B&GA%_qsrcnT82;(j zx3vE@@Qsh(dc36KrB@BkX(uz1m>E*-jJS;+x9KP`{Til00o17P(<30>@gll6;`KJ}xuFc|l8Y^?~&LQ#vFUzlsB+N3u zYPZMH!0Ori*yYI1HST{lBkeip*?3^w--_WXj_SDb_N)Fzq>akgF$v&0vFlQ6g3Hgu z?r;@U|8k}q8rYv1UBJuzkhE;b#$w@rFVY(&lJZ+`14 z+2q9b2{aDJ1oC%MGBriEQ}G5)zCvChNP*p;@sB=fgTuc<)+WKAQSriV_BA2K`<2by zA@?M|mI3LwkmSsI1F@k zkAYM(B?Gb5(^PyDTPv@PxGL;te*%m-=VvsF9>(ayC+HBj0C`}qxLvfGcN>C^P+P8* z_DXL;{QJwCw;j1fMA|LwK!n3o9eHyH~J zB?Ho%94a2Ix@Hz6f-Bvyy6quBiazv1Z+Kx$3%`D3SSk=EbCgIoj9h!z9)hcu)Tm}y z#-#b#*O;6BODFafn?B2)dyL+3=bimeAE=!68>8`ATv9iaeC)(C)Ebi-h+#_hkRtb~ z_*OZR-(n9#N%U@?9*I__yb!7rSihE(_tNY;AjKNKZrJA%q$ z@b?80?t(=t-st=qUSVLF%vJPBwbIe&Y9WiRl`d@&SCeZ(DEpb%TVaFwvGNv1Z^e|K z-hJ)-SB&8M?fJ~_krU$;SiCOpStvlcL@)K3`(!KJ$2E>*7`nm*-8EP0S|S6;iHf=;EEeQxF-&Ke)>R+aS!_# zlu$gNPjdQZHi_}$nayutFeeLXOl`AeqzymZb>3W$#A%rWEGUeW25t|`XBV;+-MCDi{g_WC6DT;~{1GRgYfJ6bhsFIr^t%D2$VjrT!82 zi=RzIXu^rjkp-cNul-O@uC^`d&Z#4Qb8_p8D#DB4SO9db|L9b~HoPhIk% zf<2E`gM`fvue(Bx`$wcMs9B60@Ktm}O!wTjxH?f0Fd+e>i5-QqY(k1twUE&8vv zxGX;;{a?mwmQ?Y3`P%~z1s$e#iw2_Z!=J;#!<0s_f%6fql^mvSNe20+eEaEJ-)e>I zS}_#<6hq-p0Vue&i7FPp*Rm*|JrzXxVI8I%7R18{6I( ze$@yo7lGm+$(u$_n2g7MN>)yhJyd+3S2MQ?zM9s!hHqXIo{uxsL>-%d!0jYEGa!$> zD)Jec@Jpn)Mekd+VK+K)9$@zF;Cn=}zo2L$AUZbCgEOQYn%Q-7bL% zu|~>T?155G|E$zAkFF8f8@?|Al=bL-$%E{Cx~HeHWUkSiEb1)SKng!+;l2iw!9POD z4pL+v4F2M4;Krnf@9=37?4}F6a|Cy}xLw75EL79Fh%?h0!7VQFyv@lGoC-{UewAUi zj{Ag@96BuAA%ZovaFgpGt3{4IEn5XG@>lddkr5jom?iC>gDP2u+b3+U?2E2vNSh$Irl=#r-!h+< zoHjCSVW#P&uY^7TF`|85-R?b#m9MS<*88J@W$bf6n(oLv?6}{()dM~5-HtnG*^G>{ z(4P89(7}(1(X^DiWS=3cop_-Q>*_HM$W}_0Mv+u1z9Q-vc;67Q%T=^^mI7fXDhqdU z^XS9#FU(N3$*=Gq+cCiM@{zL|8TfX;{Bd@>f)O&N?DCneok+ui6Hhu88W#Fo3tvUX zBP{Kxo(%%5SW=^*Pch^hx5s!{2_buY6r*PpFaP~t%+1B~VSgtMsaP-}pa|?g5>YW% zD=G9Z`?^7~ek5Y8R4duSUK5Y|6Lt(3ea}{o|0Elb7ZnbN`W-uF-n77^NZ5(Zg@qOe zltL>IDh(>CC<3h(R77c|C~-I9vKy6E1v+q0uDI>@MfX%gZwpADKH`ErMhRk8HSIq5)(|Vy6Y(*ijf+&?rb-+_eYD_A7TWA## z8CeI@iD`XoW8`j3EtAEwC(=gi#P$QC#ADpnTuKJSiz+Jq9Df%ZOK*GVe$o{- z09C%wF5L~JF~GGrW1la!OKp`Yivo*9NC~U2@x^{AfX&^~gPKNxjn;e&-h3|z?Tb44PlN}r{> z;Z%|=JN&jM)-yX68iwJu4Iqw$hT++-el6qA<|5usJHuEI>`t7o^VtclMV&+wT`8m`Ud^Tt{jQMBxI&y>AO!M{Iq;0w} zkYX5ZMtE@~K+K+P887@dKibAOVq5f8Y+HN$rH3NJdkp7y$3Wk-uA$5rw`>BiWYG( z3#HnWAM<-m##+EZ|Jkv8fKm@j3 zr0=aV5Lx1+RS0<)IPO2*GhgmFt(#(D z0W=m8b^;YVyqqV30`LzF;m{+?ApC$Zns}1jQ~g+4IsOKZW70WjF+K^XCm+N0wval9gIEY+b2l~$&xD%c%!r0MIZ2PmTV4yst^sGE7ggx z?R}4rfx*GA8kNC!TgJ6rG)xRM91}QvT-9 zLf~aj+Pa8ZgG?+f**?ywZjS4 zTj2#$jW%V)>X14z;xf*RaAqM!YWYtIPx#O%`!B`u?&=!^~+ncM-9lkGdB>m6T_v_1TK3h8I*$*fCC~L2Bkvm zQCXs<^SrNwR!I6HGsqH9XLv(-)@@UGPdp}AhKNp-F$as6(c@6bH_l%>eOS86U0Dy) z&Lw^-2hiR^y~Lh8YJ#?^kIR{XS?e@ezQ1F(I!)-2@cN()hR`9U~;H-4y>thX@m#! zfP-?_9myy62KS0@%up#Zz1wFhu>k|wxUe$z2^ZsaSAhZ5yMk<;?WJ0>rA!nIeF0Jyg!LzcLDH}Cu7Jf1`WG{-{jWhX8`Qbam z{nLzSiqpNTB+HmpvN&;|2Dq8Wx)v9U2gMKz?tiO+o5! zHT{tnRzXhIbix<{j0e-U8!WG<$_D?FZywZrzU;t>rw|LFTND&m(;v#axSd2LPjTJA zc_7>ndx;N=3rX>-=NBt;W_NLM#RTfyxr!9mcK(P9&L#KU9ACP`7$u`d!?wY<`Yy{T z`OkUZoo%RmfQUTB23a$=M{zVzrKt7p=R9@k=d58pb*c531CCL=&9x-3GGCiI>8l zB{61F*+R)SQzV6oHx$$_&33&>(&u66^HQ}SPuLfk;<^;&IItynviu1gGiJcBag=Si zjS(=WcE$fl8)O7Vd2d-F(J*sPow(w$&ICpWDcL@Xlu_}if@8DJ&A%_Y$Qhj1Kc_9O zHS|L00m)8Tx)ene&xZ<;y%}_ z(QAN>7Q$|#^62}K!|s@GG{lPSYm=DaY;-;v_ooeLvV6-MjV<3X_g$TK;<4bT;_6i2 zQ-LiUm={*2Kw-*&^jIV=V4)^bL)Ws7%v$N~c+X?>i)K9Bqc=MECo{A1pEIIonafaF z&~IvnLLk)ss*Je20LTvgo5PiO=&2;zZ5f20&-twM&g3S9oP^+W%S_mg(W{;IgtdsP z*k?k{EZWGD^`h5yUAJ}jN*;UgHLfPU%m!o4koLt;7c!h8S*THX`D6G)YrZq_LFEMz|_*8 z6q5bZ$VVpQSVzf@Q>2=TzaCf~au1YEM}K0PZEcKBQUW zpgMc?thTsLcD6t({g_|tTMD{k_o6B$7;AUnaJ_6qUU;6-a5?Wp@ZMMEH=FAzenFB( zFOr4Smtvyng4zPiBj&Sv6v~wW=g5F$m!U!e%#SFU1(gejg}0>`09+ZQ3BCzEj5i|+ zd5-kJCkXf#f_5S&@4hW-FlUVXT)8|Jcpzc5DSL{E?ujS;hYv^WfA$tW&HCevR z*deiQ=*QkMIopVt(N6y`mp?nf{0@Q%M@*KR~B6pCIfrR!8edt9iMo5@ox_c`n=FfGJH^0 zxGjFQb>MlaoATldY`~jk@QmZvP@S{BmFZ#xRPm!<{e>hj3jsQD21;cDzZ6QAM3Hr) zkz^2atqwXQZ1udvM^D-o0N5}#Cbw7xAd@>LXXnSi5Z-1RA>k6>l1NrMv5nbfGRWza zY!gK`Qt{8$O303ZvE58@rdY?l;&zecNM~aF5wdEe(GZM(yTgt$G8I$Te>wga<}nzj zjrm%Lw`8)4efRQK2IYb7_?6IJ8u{*3^4dVqBit#|`J{thWf@!9!SCR=i>jxqye7vk zSdEWolZC@DG4{dC@bPc|{qCD`quq#WeES&5VK$XGZ}bsX^<$HgWt6OxB1LFr5F@*y z+T5<{tD_;H{8(_zGn)lPy2#rE6)?EsLBdfir)(Di!7-L_S4SsvD)>uVq7MROl36Rn02@R_ znqL*So1F-B4A~r&d_Vbs*DG%FPm<|?!m>BQ&X{4CCVJ)V)BILv0r-h4I^kO zW;}kI)H|_*qBDWPMM?%daxGN+l55ia!ly3Tvctk7ihOzqWSxfGF7S_q=R$o;TU>|Y ziGP!KTU%&ryTP0i4^(H3_z zGS@#%&`3@}14sWHATFx4?s80fq1@y$#Mcj`%9#jAXW7}m--VeeC z!jf6pZbP1Jac3gdMdt!V-F`-RvY&x5h0eHpCTMu^4x=fI8W#P>b1@gEb#W}jS57Nb ziUBsn*{aEv(0fss*i^}@qqm1oW_^rdP^<;Ti-*wuXBmNF%71M7X_q-kw$p;b0`0Yt zk2HZ-!%3Ts`dsL^D3LcV%}?crOfR*p1YV^NGEN%)xA$Z+?kHLdkg--^+y~kDAR~ZG z<-LP9ok=4G(qDCp`1$ey_DL%E5WaR`5U2xi=um z3fD)!HNlTPt7t58OP@X&!(huOOvX}`(I`0Y>Y?H3mGkDz3l_)>5_r|GBv9|Afux>V zpwtAz%PF4>1em`*Rw2Utrb*amgb3#iO&;I6`w!-vf=rTt&YRu@mY=aJ<E)W!@5OJ531DL>z=ppknAMJmp&+Qa|?`)Nf5D(5(Xa=&Umdz#jdG-||N4KJT0y z7vct_nY`t({q$ZC64XVsxUP$Bi`z}Fp4KV7$GQ@DAf$ml=b6k}Gp(Oo7uJtCeu|um z_$2I5NP~Rz3Ws5W5dcPkFy`mzfZ6*eBMV18T)&~z-1Bf+*VKYbv`v&30Hg;oIdp2w z*_j%4YD`;PAAj4T-hj5avop^xeDCl=+`PP?Y~g!@QZ;l-@`Q98*zZWp(NXrasS9=B1JM_=c+hT^``a7h)v7Wf1W zha4Umv3cAbFxF!q$;g~c%?^8eUZ~NWtWEFVOG=zL?h3UgV+=$kB?AVrgH$~7^PGcl z>jh6FZo-P$R?q%9>m*v1J^|S3ja)ZXK(T|>veS6gY?rlyY&mbBG-6ASQ9xUY>NJr3Zh|Vj!=pz9dx^z{o+lBNiMVL$axoZ1S)^Wplu%|gJ#!KDqaV)2awCD^?d@YNtHgCVxVC% z)M#VS$4F4Am9}sVJYq_H)pjcjFDeQ|{;@M^R%JPuYK&A@2LxloT>vV zqVUv+i({)?pZoHejrMc5JL~{6lk$geyYDjZL;Qk7QY>^Wc7a02_P{Od*65o=?Vl+s znO!lvcP1qHxc|5z{p$nCF2zpI+yzbanjo7E0n@Q%2BpcXa$4~p|M=05jq}aZc(ysirk>$kt?GBlsOD(yk)Npd$h^H8IjhLV%AM@b~wu3YUnHcBQZwV zpI+KfbfP>S*h@p6hkPQrHBTk~hyzT7nD0S>sb&zD>T^{KI$uwX0w%&%R*M)3ooZPq z`jD-EE~w^475G=r9`l#`UKmLgSDJZj14izIi2Yv`)b z0%+UF3a_Uhh*yL!t(5iB`LH6YX7vWF47xQ}uSl~(tnyGkjOk!K5`zj&wby_oS@3Dt z(kga#I1=0%1QpND+`@si`>;FE0feuedCvWQ)M*j$O>ui99VA`a;@#q17qmBgsfNZY zZH@&ofRSwzmyPK1d39u(dHVVJJ~Ag>TU*GL>nS=;x#~pOt}Q^{(IZK5om{(N91m;8 z-NvDt&~29emxvX6e`MbC^E@J)*nC*%>_JWC0ro!M`*RX`7hHDlw$sbwV#?z=zoJ{0mm(kn%; zvpf2-LQR*_>cFRs zTwNEG<%ds92)Q+{Gwevfb+2ne!?RWL2a;00eXwOSg2siF{-CuWdEs@;5R@1E_Ehs! zBa?`o6FVvv2rji!D6J14q;nJ+cE4A;w1$2PEFjghAB3Hwv*oAQOEX#Jtm>e0FCD~L zPQr~a<#As)<-}&~pNxKuGG9G3rd&O#y z-aye1gFeSe;b~EM$cb4y{r1dg;i}}>f@?y2cz!tOe!Pzf%R+%B_%smvsOe515z6v| zwt4Bf`OVy9)(L2G+%M^)p)25~xPaB=aVVtH?cDrg&H--}_I$!g<|q@%Fgw+w$}dlG zk)?&c5zJ4eORor3icbQIITo;JD}F4SB<5$vgBv58H09kwj~|$;J2-8I$%0-(tNa`v zO$ZP;_0lK>Qt6fulKsE2B(e?@e$zSO9RPQ`bdF4QXYb`rlexJ;mj_N(uhj8T^^cX`*l`s|B|?VbKs&Ua19L9?gNu^R799lR~Y$Zf4y~m3*zohzI1Ha!D7B2dnv| z{C?n}+wXc%q-+%rgyku)#s$5x3Z)Pg3vZGddKJm1yTVe4^00BYP=Iw7 zBQAF%J4q4!G_qZk@`|zteyc)3G6)QhxS)>pJ-Uciwun(B@@(L_z*5OM&zmso>RWDZ zvNXXbg~tQ&Q`@D(?s1iKsGXwchVo=)FMXr!_Z-G!TGSu+y2}`n3<*LZggdFc0siA&r~J+rVBKkQA=G72d)^?KiBR)WW!&#Wz{p>`_cPcAj819451c z!-fJM%xpNI_m8;!G5#07iT~}N-u~5+KMGe-vgH&>n2a{M7k&+mC8=}GrN*80cZJ~1 z*a_(~S#cB?!arDVkP!qHee}R;-fiiK3o42r%_&-#bg4>F4HFaY;k>W{?0X`w2b$jv z*E$T#6M(TE(=Qyv37ljG2Z#cI9^8|Rir@4 zCe{$6%~aEsJ{6%w@=CXTbA~;Ys6&od0^|I8k2-}0L>f1F7!HApQwyauAIk4@&v}jt z;b7IE^=Zk8BF)5vTCrm-_$(u>;E%_@Y2L;3d_KyFQDvdCsftc!p$aTI${0~m_J$n~ z!cwa8puMl?{be0gXjMdE+0>-DILzT;?Ep{gI?M0~I=R{3oR7`|?VH-Ko)}yiWuDhg zQM>5itaU6X#K*Nmn%|^-$Hs4h!B;)8?~dW!mcQz8Qa;Z6+Wq2R6V}f)PDYD53pS8K zW-D(eF1c$knWm0VvV#=ahsB`Fp?U2jFBK%zMz^r2uGJIOFVude{6X?m($6~#d#z&p zyMUEP=kZGU^>kg-c~{I-4ZA-icm%IZjp=}zD>eO)?1Ja1RsPQtM^XU5$4G{V_c&i2b zg()mNKSX+^yJ@`U7;ine%otN{By|D^BGUaa1+424c{ElZ9~9NjfwPKLk#>X)44@s6 z^Kv*axME3mM4Wl`jMJ`|EL6(qn^BGjss$K)8+OkR{}fU~_8mU!zEP_N*}6yB^(tml zWO`IckU29dlbp5_&qNk@RGZ(p2CUm@o-1OrL*C5|xk=jJ{6nWSjeSIM4ytX2NOEW@ zN#PWF=7$uLLeJBaB=l+(noOq|yQ0R@!)}+p;bWd(w2;P2m)@g$#SekY4=5kgrN|YU z5c|JE0%YBnY3ZNrxZ{joP3J9QZ2i)2h87vU*U0xi>LQiQyjSP_ml2RN8dINlhLWA8 zNIeywLUskJd8ll^>A%?gvH1;dMbp!V#A$Hl}_m=(C=cpCcZ zMJ5jgjtIlY*)ro<_ga`58OD1yygiSq_Wjhn%-d<*5(_2We_6)cBT3;X%e?maDltEI zZvL>Hisu-8qt=Zvqet51Vn(KKYIjb117)-@PduKS1qEfNO+%3FCN?9PlC7u68Y+Is zy(P3fpvPNzTGGMFckPb0<&~ZIcW(tCCjKEdJb@V?0(*Ki=B;KY`hRV0e#E2EhcxL@=7-3s)VT6n+5C8r3esfhar$vSZoiUs_ zaMoAJ(*%d-SGpNA#PYxkz~gw#&?`L%EbPUU!I3c3I5v{%=w5mawG{qy9CjEXK%OYdWblwWqeyXqC=q~iaZ$|Yo*6Bp(lFqu?#Q!+3rJE?dq%QZHs8+egNT(CeE zlYH1v2vi*4gcb#sMquu)MSRqCkW}%pA$~v#`8gzV+2Q3-D>E^uymZ8EzJL)_rfyl= z^cKqqmEV0Z*g%#uTj4lw=7Hpx0BbWPOQA>-6|agd4>9JP&d)E6Fyx&^0xO|{+i`Zq z!f0BHAq%%V>~PuEuLPTC^PG07v5;m&{h%s-cXUGJP1c<-T#I7mb)mqpJ=sHT?MP2{ ztq~(@F{S--+HcL}E}a$^7Ni&SC8|Y%#X!!Q2#yibm^84)LLD}JLajU0ggAn`QD_`jrU4fPctg#q&$40xhIE`CtSY<@H-~vevxg*{H zl5bUO~KGE!iI_AE-a zjRG*m@0+_%xI1E>L>-% zKEn|QFOPihf_aMNb5*BmA?;!mhpM5Qyl;o(f*`v?gBzAd#+q@re7EDyST=5%KRN|?ODZ5BRu|PzuzE}@g-$j%7bq8k=yK*#$QCwRlL{aU_{<#X5E9m3Goj{+H9$qOJ1|pQl z-UIBi#Scl2;v`4ocAV65N*7kq%VQg$Vxu*p4|Gh^!!@kIdB?fw($RZr;3)bf=!V~u zYJv~W)rrbkS6P|7{l3uJrN7#a)`uy0##{M5J2pWikGcI@s&{UfuVlU;wT2cJWcx`M zNtm_U%CnWPIMk?5L*EBF#fRaL|bTA;5Q#h*$WV0u~T-oZ8cF1^IeyC;d zbeurQlwW+4JNz$3kM+=x{C`Xuoj8Ac%ftt5r(|suIgcVwdtz6)pOBsK9+qk&)1_H{ z3A_RYZfheSMi!EO`qM986Rr+VCoPTQOzr^KHHEwUh%-&EqIe8=|)mGijA;<;J*)AzGgw?(p#(@B6l9jK{mr(Z?9s zjj5+@{Q6hs4H8V8S|?sETj=w^EHPHJLrwc`j}DdrO4-vLSbt_|#H~RPOf{Bnci3_2 z+vl!C8$A`Z^e2U6zY}{ZADMV6b(9PQ!K$hFF76}Q7WR>a9|=;Uuwm-J{QKTpLol?xo6PBzmH?aIkb9*Z6cDu1 zN203Gd|(XCQ|N&F201CV3J^d00}q8@BzKst0%Wy|PAEPSjkxF}1`g7SsI}58V5k^X zE=%MU2v2jA75qe0sYgpzAXKguUkL4UJtp7hJL1wR?}C`$T?i3ZK>vBa_;17PHExX2 z=rH)10hTEmrFRp1ghoI~{_$oj+33WE3F=M8^kS(g88ljCKp3BWlBe~p?9 zY5fmWKm7RnDmgC1+v4s^R7pE&lJc|9f*GlGc^cTVHJmWNc+VgM$sw zGczPF-2Loxm~O0XrFzYMW2ap$S@3qxhNsLt2>HQ7LX6xQ=(AE@v3;zVeuvMlw`!E1 zT|Q}Bj10`wAGJ3B+QVpI-nl}SkW?q0g7%o0j2ub^q!d|Hd=py>Qh_LqdsI;pft+Z- z`MKG%8|bT{^&Jb~GDRN%Yg3W%XdpEB&RQW}x>kHLd=l13R$*kkm)*f=VTO@ky>>ps zJj>{`cEbYUQJSBj?CyqC`7tB`Dx!{KX1Ok?nC<{l>t4ExY!2uR8;~mdK`=q(mlS?H zu#?{{y+X2CM-?hXl1$ICpDsP%t?v$l-oQLX72Pc><)2_<1w&FecJo}Ep^~QCYw2w> zc*X;i;h0@u#Rxo85@)RMGjA1T5|D9X55_`c>B!=1LXiONClN-sdFN|+nYcf$)P_n%gDWT$% zIYUAn_X(sQA4yd55~3yQz*7N>I zli&&GAX_V4vffWKf5}sqCF&VBr1jR?)6wQNM9h!csid(?+c_ZNGEzo z;83G?lVFMcae!~N9k9UEI9PSP<(a7Bmw$iwJELv+eakz+Ips6=kbbX|oSmWpVWs?{$9;ke+b(XNw>C20FGE&O zuj1$=yFu#kAqX$46y>6)61=yTg?Eo$e><{v&SPOu*dBI5$VpBUshtC@MeJ;WHu6|x zz3XM@Vc10mfok~UNVV^#;1W(QWGc0hA4fvRS}@-6|JeH$xTexGeUEq!$%i2~f=LEc zB*K6=axqlIqSI+-+i7RF+wN|6x6}Ku-Lc#IKWjUe7VwG+f`S*+018A9@dE0gavf3B z83hLwMZ5qJ)WJbOWKiM%zDaOMqv-0QF zN8#V8K}WRPJ(}oMl1+gz;$`la=4}nwCA$G4&bNbVM5lx;%A%PqO8kWQe-nLGRT$O^ zwr;Z+KYM~2_wA2S?f_su&3JKsZxFSmn}1ng%k0NOpl75zTvJsT?F4FOvrxgY{|l`h z`yTSaf8ZGYWPiwqt4iT5FU{HLW4{TpSfotfto5XT#~baBt^Ai#N`06jA5zg>^Uiq3 zN1UZ^2i=S4QXW)hQF=|9|6MxU>sVO5ve4%~M1TuHGAcj(#>~2?ZB)f;C0{=Rka?ZfXG-`xBC{O{fR z&oucd^7uI)9o!smB}Vso2JhyCY-&Yv(1vgefWCilX9g*pKx(b#=_5*APLUESx|J>! zC8&Feu7`BcDFQ5|PVRW@B1-lB>X7ELb|-1j$zB|yfSzlW5!7qL!KRHbnf2)sn>LVA}<}i61%VEh%)`Fhl#~{Jc=W&WW3b_}eD~Ekh ztw{?pRrPkFaIHVY|58=pPbibfl29Z4Hi*!mB9mE5AqzPCkZHj{cS-ssNC=n_zAi$C z;vU#!-{+C%nIeUpC0zsyGe!?Fp9v%2;*1JIrNJN!Qs>dBIOJa%+y@SfUV|;Q!1}`t z_|8Q7mzXylaRNLyq|b`{`d@7sYk3PO77kiV`3#aft_X}1?g%~XcS?OpfYgpUv#w>M zq#*2YK*_v5!E$D-dfG;S4^Hzu(u=v_1km&46dv5}`|~R+SJ|467X%(G z#>T|xQWy@pXEMlLYwAe>tU)N!gKn3uQwIG?Gg?3~w``7HlcOyhqb4`L8k;RE-uuCC zEMD)2Kk-nI+uQ;<{Po00yj39Q0j2Jx$UQ3hIEYOi5FPVG6B4IBFG&Mtmm1|_e_ah^ zrqGJSh##rWOx*^hc_${_k4W_1;ejPs=1XIO@<0&kBLPs~)@cgK%BX#GLMTXV(kZ@m znnVZdHj&C0VNWqFG9%mVKQy&ep=oK%)g!#ea@RJSY_ z?emHe=ZE(R7W;n~RKZ}>8c8&}!@J!FTw=skf!hVTEKtqqf4Ncs;!_t1l4$0HRt92; zvw=Py`0-Tu^Dra`{vWgRM5$B5e{mkDoWF?-(ldXjB+oQrkhaTuBFtvW1v^L9rMZc%Mah_d|v($SSaqkFSo_Ze07VW zCct8$W~Bc$ndI;|uyfQZU~qs^gOcYSD!SRr7-p6;j1^~S>Z3YkOZ@N%%B)Dfcn4h| z$rfKHheA==y_D&5zX9^kODEh@cBzbXdcd*?n3%%r^B{eZNnc>_t~I`3zeu`o9Eb#Q z+Za7_5=IIIEC0p`;NavLQ*FQc*WXG9BZ8v!*GVFefs=0q96hDpOp%RLG>{tsNX!(d z=AF_?rex}(MT?Lb7psDI3mPTYNiOq|z$kA3($ldao*EHAo(a3rU%&|<;|{FZ(LWmB zn8&^kt8a{=3?Nnl$U)H(^gX(#8|dbcdJS~9oCR&6a&U~dgc|4!2+=s$tgy{5cQmn$ z zNEI|0#H$cIsKfg>WQUECA8XPkrTBj6Uavg#%8ePTJ@HPx2FEGZbOvBR`@Be(CtdCN zvE&rNs;_47v$4DjHNy>bPFQZ3SyffH-lK+Y4Z#agAK0wb+UQmg24vp{>5L|Ye;4KE=mo79s?D< zJvb$@j1)kXPJL94pqf5P)<_y6a;H{8DRK{8K<}5-MJ1|l&dPMxYp$yH17}3POToOX zd4=I=(tejXaVnEbddOZcgIl8dW2Lif7k6W2;}tp9Mm-+w&C{N%ZD}O9Y^8V{2V~)P zuJue%@0xb{71%msxNrTmbm=*7>@R3^(>1v@{k;m%gbh_|xWYET!`u7;OUULh)Cc2& zQtqE8#qPDD>Gzeoaww+B(k6+sJ+iggURV6`Lt#Ntmk6t&jcu8G8!;^>c~T6uBr=%B z%7bWd4G=$=)xKeQNL67hs8yNdil-}t?Q}8S4ti8a{BfQ5WPzz983)FD9+1=Sz3xW> zkGmGmhXQzzvVz~W6XHI4O@VTk8ife5we@Pv2G+(}MY|+h+eETGw#ap5$?i)_}_``hh4$TubYj{Q*f*P#@b8nkPFl8-FG`KBF}1Md0z8@YsGkCPOEZ>9|Va^T6=Ng3@HSX zvFImDTkD$V+a$`O`)Btn2VAhD0%Ah+k-#*0p0q`ELl8$|BCEgx4Y**Monwxi1L{VL zc@Dpd)1~v=qSt6GU^yB6mvv;{1X5%5Emu;PYMwBP5CrrVonx^zk8CtEfZF{ zT0nICE9HM7ONNs#;IWUq%?j};lscIriBz;nxf2DFbVvzM7>>N28KF=I3#u2GA99|# z7zQnzFg(ot&N{`8;Y!`7Fa6R2i(_xz@g*%2NSsyZq?=OTrN|vBx-T+C&?G{--E2Bm zaf&p`;xHL_b5@OVE3?)WQqbY0!HLX$QW%bjJ(C<>j2QJ*8`Rt6rk5D%Z4F6e4$M6* zTQ0t>(zS$bm$j>~Xv#1Hoy$Z>G#bM&6w{QGdeLrzCvtL+}=-$d=}Qi(tVu9wEU-G z@Hhc#7d6oby!$-*JksP{KzEJBg^SYtYdw>MW{wnHdZbZaO1Fw*CN9j@7E02k&=^pw z!0lw7w9g~k69+TJ6&pz#vnMRcwL!9Q-GVf^&iwIK&e2_m2PqFe&;jK(bq&3+L4}@| zUX!YVT}zg>&!a_C8BiuEkmQ8sxxw$Kq49SVN_6J4yxrPGeO9J_-UV6Ec47xQ)|Z-{ z=?1w`xhZm}^?GbBoP1!Xm2~o_cyf47R&sn(+0q*SLD%TjwZ$Zh$7xg`^BNLbE~3db!t!U%!e)WO4}@$|RVZp}pdsuWbi=Wlfp)^iu9fu3$shR{Zm{|EonxQc%8znc z!}z;bAP1~K_E@a+vw$|(%Kz{C$MEmJ=rzUkF7>&uWxcVVEOYNuA&!m;Qha+uI-wWO za_t8oDrS1ONpu#&8(*2QhrBRYOO|RLey{k%4pQT&1dju5_!pPR< zhvV-mr#ER@N87YMdbqHesHcJ9+3&I2A&;mlzX`JdN#Z(xJK6KX#0^hdL9Coomr&#o z6}{B^EZrm79ExSlrNQ@RErkTwbvJNOQ&o+SHZ2V<^8#NPm%J6?qfpU-EXRoqy3*jo z&PY|EuX`B?v@I<-n3C^vPh<+i(_y2Ts_G&2>UhOzqDv0PU}MSDZL{Nn8NmV0l(WF~ zZ1_C=iuozYf1YB2(3?Ad@ebL-V{hk>6)5s4HAsi$P|-cF7^h;l721i?fIhMAoNK1k zsN4{N66+|ajjl09e{`6MI!c;Ek5q9UsJ@;k%@(f)37F5rzX;LAh?hW@SB8_=k{tkm zr_baF-(|Nh!-NpVfA>Ti`H=-KKRUeSb#jTvaDn>wAb=%Wa7?wlF&y5x9jP!QdN#$7yxXI}M8s%s_Fpv2OkYw!j$`iM! zI_S=bT;^IBY?iA6Piw2R1~-gerK--mmcq^$dUgaE43_JZ1|Qa9)crD35NfUks#RQv zri_R3I!zWiO1hM1ybGZ~2jr@tBHu_a?4)mrbxopJah7{mxWT7IlPhi#t?|Vpyyk&m zA=dSn|9lyz8pMDDz{a+gaXCNuaUTEV2DYy)m{=QS0h=*&N*X!LW3XMc0^2D{4a8wp zRJ1t`)34N}s)}J|aYgD7n_u~oY{61}Dr-#qBA&h=|mZj(0#8~g`c5~ZEcu!_rEH4sE~$qFQg zpiCr@Nl<5m>M%rrnyn*V#hj0(#{l7Uzzt2puYWY&cBOkkHfSvCVii*zkPVgeEvlVg zt=9Ajj?+k`(nnWoZbV^faOJEW^rKhuXQisrnEY7l;K zHQ9!jGXkHqF`TA~{o>Jri+$wLy(9Q-{hS|Lj8DsV-+Yra@YwhOWzUe6?lz@vqll4; zey>>-N3{`*0+rI+{}D@BPmwjS!^)g?4XS~vn1R{p^a@g=+{0wajHF-rp=&v0OqP2Z zAQ*jldWieubkw z$qgE#{cZN&&kY(M9E-{v4dEq^llUxzm)K6)Am8P=8?qQjw4Ee=^8J9LZd)hTDJlb3 zM&XguR-<7jkpqr~z28|Un?Ig;$X4T*x23Kus=XNT!-`bX7E7xBS?J{&3&X$xBVRfU&aweR#^r<&YRR1|-x>{%3Xf|p zSa?+GG_|5apXLzU)SPyG_|m>09Y%k4g>@>@%*&a4oVGZmJ>T;^MdDtV)Ix@p`?{7=uc63lv-|pfDcy`L|ClJp1?bdQJE0VR zjG*x4Y}#vKzPw{>j+T8?{kE-kJa5;P9ahtvMyWSYB!!9|~xcT>6)#vq?8Mi^1c3 z_gbqG*d^oDvExZU6^#X{O`_|sEnRT$AGUvQ+xPCg*`}xnNCsA%O!r)ErRt&TdPuA= z%|DC29|3$8ld9>4;MKwm@j35qc`lwi{za7j{ zVU@f_0enw_8c`Z^P=KIc8yx4`7+fE9PYacSz;}3DdUa;AS377;IDh&vRJfk z(3HO<=yY!l$<{sqcK>z>sztP@Dk2`rT6`WsfyK>$GdN{WF+mP~ifS%3Dh%&G0-_zXY9*iPF`A0?8fu9bn2TS0+n!^_oUWze}&GQBvZD ztN?mVlSjOwXVxa|!Y=5+#2R071uQ&bf?j!vo580Oc+WGSaOHY%zst>8>HZz0Rarc@ zFx)T$%MKtFc}i;RB%7ueDLd%15`AQjc&l%}OP0DvvRl~0;F)In?E?|1B+vJZWGz_%am4zy8@eFUBkYcB6TdgEXI#WvLcf-F!hiGjP`MP6$JR{<>tU%w2s ze~6SK!9J!3(2@|+L03m~j7i|;@howH#^amD;6O_b|H8H_j|((-9Ft+`*E>!lbM_Y3 zEuPDQVHKngYq4T)1%-bcXbbF{J9}Ya-)N4zYQY~auO3b10*_(AQn`R2fJ`y~0$d_Y zA|?wz6+m1D!Lr2trf64K>cjyTvlb(!od;ZUfg`X~fW=n{>OMMaUWK+5q$5g#9a6kG z{8ZWt7j~a&_;E*j_Rrri-nC^?;%$_Mg&XObv{<^?w-^eUzC5anIv16yLXt>e2o!Jk zI5NfH*6(u5y$n2w#iW&#lWf6aVql77U4kRCvCI9qG%sxD4==syEK4$K1dMj7S?=fL zWmCnylg`?j9NCFjI53P$a;tyr16Lpu=U7*Q;Ol)He;)&ZRKhV=Ds?H+I0;w)JnR0TG6xPoBba;lK6R&&3GQm>{+JQa=AkPz$L z=-)|D2yAs^yJ{5s?2&)@cG-@R4?p5N*={#eTp#Fx5D+1++ zQhc$6x@F99U>5{T-ouDzb3`|?Ltq#SFz!bmccgk*3`O0~4lE>_hU;eIu@8FC%24D| zYS_kZr=qj9m>B95B{DH?xE(nOJJ$k`7c+|BRd`hj(h_o3zy;)G@g zv`cC5X?Nzv)_(j9=Yoz18g|{pj-Rjm{Jm4py9qfZ+~s*y)g)RGm^BY+k1#?U6J(s4 z=T|EEyZ2xL2s;FZF(K50dvCro*J;2z)H2f`R5Zxbtms<%fDruGfC;qD9!6VxUWdTj^}=WwK}y%0@n1 zfF(t7;#?UpGcDTZkGhft?wNvW4TxAc@4+4AT-f);j^HYebHUf`Y#Nr8EQ`7fOv4*z>vKse;a`So6 z>s6LS7?`6#8v6FzdhH{xyTUKL&WG=yTf^|U!e>AHeLlQCYKLh=T<-4ppAj{LcajcYUQ0^AfMUt>>1UGVzI4tfvsfaHL%8z-?-!; zCmTrZTAE&J+wS$Gcgy2&3QJR0gS^tCOA?zE{Ay=`Z<=Zla(blznu8_gv?(Pi8m-oAv`Fc%6z3YOjWLern{!3rp6j;Xq z%Opl>QBL5)pmObM!BYPkAUQD8J^{_7&!=8ku8*l-48x`Sto3&ajyM5k$Z|3C&>;x1 z8H61$E8dxR!?r(=i`T;A#2!mGAqGv@-xB^++rT++U@^g~}K)2x1`L z92$f|&=hdYhWFVM!ezoOcH!BBI^YU+8!^n5c>IxfYWr;!2A*7Kc)X`!kwaKIVI?Wm z#7C|V-|D$+!dA~@X-4SVv6Gran>DSHgMu5NnUm$0$Xt`$27d$_FAVe=MPl$rTEpDa z;`>CeDN^>zF3sB-a6w{lGq-8%@@!F__1p9kRCtG$sCUYmL_0jH1ImOsf@aAs&l_&Q zMIQ3-E2CeJMgXj#>kixTsG-FUtc1@_g-x{VblU#A@dFY+0YuM+##J^^YA`zMsAw$d zNra6;Y49%jeL=eaT{@05D-zU6o=6;;DOfE`ml?uG|0UZ0ujTZ0+du56uVnwahk1_q z`+y#N?Vl!2vS6d(l?%ToDKAX<{5~s)XHjY>LDC~3^X;I`p(ty7#`_Ay`FHs5qz!KO z$#LjxU-5e8TbV!G^Pejw)JOHqDxSMUq4!8t?eOTJD}`6&A3Cxoes}k(UzuTn$-D2h93eZn!Gy>D>oF^sd`PMHQ)C|%T`u1YLZ5|_y+CX*;L;#RW{d$B zeNq;kW6mS;jir^Ani1yL1!GR>e~Hl#7Gn;O}H;IPIb9UP+mj7@6j85S&x&gfkVwOhFPfq-F~;f_6sS zBs-WaI$d@YRAoP+`<1(1&h#h@*a+Qi58Tfw@CxVfX*dKkjs%NktoWU>FKoBVyxkSC zY=E)0%?MgJEvk)7aX^v~%iBt(V#UElQ0vlb#?py`uhsskcEWr+np89`vQ`!Ed?gw6yFxj0dJIh%BS+F6kvZiA9~v}Pi&Nru_>?{ z7|ttB8x~!hfHCfCN8hZr1qPQuJdeFm7C+?Z8`$pK;Z-AAw*VV{KWJ9HcTZU#)gD^puwzON-Q?RAghh1qQCN0VHm7Y2-4jk&8bjM5{A|33#Db3a z&DVZO5?&Z|Kn?5=bZn*6X%yK&MPuvDUvUXZl(tGywEF#B>JG21LAQl{KufYgx?8Xv zXs+&R4{H|(;U#0UzHr0`da=X$6OKOl|GfY3P1{BrmX^4|D%Mg#E4|OlRCkPJtJwOq z%V%u9#V`N$c7LB=-ZCafoNPCL_xsN<6h z_%)8<8Zrn_JFz`@`I!AI*M8+A+e|H&-3gCD!IE%}6&~`}ott|=iJ$;4bv5XIeKe)e zt1@t5J%i`BspD0fRk}R!!ZGR%w-XzdD2VCDnBZf_fBsMP|5*Htn{(H^Om6bHUU|6{ zP3ISsx`!fNpcyD!Dy-9-1)cg5f$kJRE&4=Zp(IWWdL~tY`M{KjB7k?@jlds{qVnl< zJxK^fa+!?Kvl5`6IZC?2jmk9n3Gqc@3`?g!^)f4{EW$f(DVKUSLG9|6&^~t%}Y?1 z1ozQve7^v-gJO86KB^iTiTVWlWu0`4c#ZcNDe`#c2&zT(QF?8&PrJ|O;rAo(dgpj# zBd|Y1Z4n+j#U61&47Ft4%~Q68Yb>>3d4lU9EvnM_r6$7EewV{=R4Ka5RFWNZjqg{U{Uk-OI`U)Y2Jiwve;4osR4WZL zHfitA$9IvRw^4E^0^jMFd|i^lnD27H^f_RzZ0F7))5mSDcuO3!_M)%&Z%b5nG1&&i z^Vr9@pA`8RX`n^g?y;9?_KFkN^U2+&4A_M8gOW>4MLh@8+9qRxl`7wn2` zg{SDCpi5BX(>=Wl^i*#L)hLU6bSN3Wc*aFZvlkA0aLP(Xx0h*9H_;VLJ!6DLca+|h zbDjj#!!asrb|n5Ms^`7F+6a&+o1<&?s#TzWf!fM>Vsw;rTU;N6pf93B#f&Bzedxc} zR}=9&fBeZGqJQ>3KZ*X+&wl;8|B05-Y8OIL!_;Mt-_T-w_$!NJ^WK^juaGuw!8`uO z7sxUz7rd8J-=jz;6}`_3tO_`+U8+uUbEc6V2slsg54-BOU4T_=CzY2Oy~ncIW)`M> z%oQfjFW$3Gv2NmCta0->=2H{g7h$FGz_X9)^*cPyLT~q3$##MH(>DnUCDp+3W26&&7Y89frD4W?zq9mlD7vv)*_~*a zhFQCxerl*G^IYF@*0%qYrOgwKH}zjzmKr-gYSM}yLNGj2J!b_#@0zx4Jk{#woib@1%g3I zuI+SVFt!{HxESd2z}8ogConbC2x=Q+gKz{WXOue$)waWK{>$eVZ4Jl^0uL5DXeM)O zqLT@RUW_!5sNew*y|t*y<`ha$=KTR}6tyZdVQn(d=O^!ZNoTbf*nWqOdCeIiqxWm= z`rw@|iHcV&-c9k0KP@Laxp_A{4kw?m3i2GL)E`o0Kl1gQmemBWoY|$?_EH4^W4p(v zKK(8rO*k{%zXq&H*-M4mbSbJU<8^usUUgEd*VbvWwOuMB-DK)jY^V2nZIj<8`~CL1 z?R9HmK;1_&AUpkv(Zno>&bTAXcSnXER!74wWY}}jAFsc3*S5)>w-qrg4efj0`siyX z{wwq6i3?E5bX{M-iLv$N|?lZkUM$G5yqjH(f>2)Dz6{ClVgJ5C17;(2W zHg|9{Jc_Wx_icHPc?%*Ij$>q6F@^aWe)9W`wg!jgxkgn|Z1+3@ z7F>tyH(8`qk|abe6J2t6t{*D;p)5+i;KZDM<$B=(?=s;T?=o_ksbG?XAA_7@cX(;= zb`MN+#{}g8BY~05reh&shQZtRkT!y0u|d}}Wo~J3gS|E1YCXusrTM4RpL-u8aiHb0-gmukUDOx; zn?r9zeg3sI3tFRUM4LnZv{!Xfn*gU0-#{I)R8>iEK4hGXqWdJpx6o+}+@r?!XhFKeUO;3bJtCs1T1 z72V-l9~di!$|<$3WA+AhoMM&7*hOfEPK;fXJoJV!I%c;H!{#Li#w}J`AmRR?=Mvd4 zoRSBR<8wfwF{B8~K&hcSEfcbjP|l7WN|=9)58gE`U3#7@0~3(vW~!+(Il$=jnk(`e z&;lon*PNoKBY-RmVx4Gdt()LeFDw$o9Z_P+~{Vzt26q)>dMUOBR5~7-A8m zLw5Dk+J~|R`6>@xt!EMnw?pitL2j6`M4?0G?=`+ln0v~0RjTSe6fevWLZL)KSUaHZ zs!aDeqz3T%I`3j11C&wZi9aV!Sb#={q(Njk9#MbzM*IfboB(h4$}Ca-!Tm2r>@8(X zouq6U$o=5ntV?xdHr5UxPrqr~2HN3+LO>;M5UkNwdFnN*JgT%9>+KR`LU1lyyVVPc zk{je*%4G3l-Jk~fa`A1IL+IiigNH#}JpMUqfUsMZ;nt-;e|n;CGz|(o_GDQa6q-a? z^F9+4$%J$!ZhbG2ae{Fk z_TSIxd&LKE*ijsRlG78)-v{i!SE%=RSX|nlHUOiB&gp3T@7jY|%c zvITx9+cGI(rn!p^7NIWXy$}>KO=M7eu1rz@B-zJYjq-HRG98nB5NGh3RuIR7)pkTq z;2p1g<>mjjW#{0nbHc*6kt7`Ao?=ZU7U(uXjpR;oxAMGvYzY`RRWI4jol{?P1IGB? zSCgLi5D8iZv*&86tL(ApOlMMxalebY89slmTs3qHp<+x2*uzow>@zKq=!W-QHq6^3F?ec zpmIm34!GcWhqZ}Nte`s|Y=pbEdNqj>Nn03bKZI5Ip!NfPn3F((Y9FrcLgM4SjzbRY z!t$w!pQqY#HoPD~6&C6VY%DTG-*yHRgqi0XrX9D@*sp8fY})nJ+H;Zpu&`$|$6Yh~ zo!^GoayIbRJz-&CSifLrL>P>H zIz&k-f01D8v$D8`7@aabWEM97BjLU7pSicFN~ZRCc8KgXn|A&?o^_d>-~Slx6u$hl zoW6hl#;jDT_f4m>#VesO^}S}OP`yGIdENBx z5gO%5!sQbR!-32X`l2_?Ob%HwAx-S;2bT*>td{66XVfW2oM6J=xJa*|cB(Cf?F-^- zun^wf_CzCSs$<(fRhQyx3`9+x_HIW& z+uylD*77)6x5p}6x{Xq&Q)CMjUF?%I3rXC7KNZW(fHI$6xNy_-fe@2-guzYIva?E6 zDp{p!57A}Q?TXLgP*)7C#n;^yI+1XY4v(%u z)Eg+0LPb{r4`AIK^WWI&kVD66G5cRix6pNSwg%jsRisQJ7sjkuh4XZeVBSV}3nwE& zo%z$TG=g2G6XZvb8lpX$q{kfdhUAygO(;q;PZnPPih| z{0hqar;RxZV#}$jopgLeGT9W6t&X1@J6YEWq?!XRG2(VfqOg^QirQ>?X<#ex@Z1n= z^}@VHVfX>>E5OL9*W#E>9<>6U`9n1UdtGG!egGY|lx_7&P~$%Z23a<{E$muYTUe7H zaQ)yv(VT|HPsdx{-0_!vcaBEwfybLB7PW^wajB%4zBvo44-E8O8q22#T<(N41|tI{ z3Y^01<$CtY>)`E!4(LgLx&g4f%iTWNHSDM_I@I~bzu9h_xGZNp#s|wb3z@gEtmK4v ziMn2iK|LU*Y*XYiIiZz-*!Fyiv_QT66~E*Nd@}|)xF67+5fA@%)s!QD@a8ABfi8pFf*hpFd0eX zpyR|4s;$#t!D6~JKfF%>!P@gCYRD|@%~{aql?OsH$X;6&cs&>bGr(=YcrG{7OcJG@4I!>!l7HrOGd8N0yq<^IU7BS(Pf`{x#x+Gdh@yWFsp zPGJdaF`Wh52i-2u0aa(jZly6yhx_0L`A%uB^zWrpCZup$X~tDl2mP0AmpM!0_9~`8 zk|)Rk^6|k~dg?@okk=}9(74v8)5fqI;APd}!Jg1~#c5(H5$=^21Eran@CHe59P*qE z_e_lT*&A+5N1Wp{A^dgz?>9&y-mt7cg_9osgdFE_2-j$}qFkWVbrh+kqT|F%{LFmh zph4CpXrs4ix&#nc>vKn%CfLFRdhtjr7TgHae5&(xuhPlBsR_UR8a#QvxVVZ$}=ELq}vjj2v5gOxv5Sb;~1$bq>jQXu^a42V3ALr+rq&bd#H0r zlp5v6uxu@^QZW#D>43m-4{%GBYYW3t6y4$WrJhC$dAru1CoX#SL%|91<97df+tu+F zWGz`9R7D0{hG{(JaoBFF724KN>eUp92PRsGajo)zx?Vw=Jeg!n9ODJ*@if_ilE>GM zt-Hcf;)pJ$5xSU{z6zB5eeUUW$yBfosjB^CUGROud1nw#JZ19i zz`)>h{*+N1{}ww2hBEW^!D#nxcFAi!WTFBo*g;983BfXm>Y7FD0BfO2(sow4Tm-6-d{HY~`%?)};#SAXihP)HQ)@X^}?P228cd;bmmGXL2}XmZ6L` zM*N|yGN6sA_cqe4A#G5XP)5La$p%%rn4nTg8A)}LFKajK2Yr>{e%kkhak^^XSKf*B zw(L;eIy3GZN#hoK<#CxCh-M6l$LCRMDA&)TqS48S5g(4|o}H(?qcF%b1+8IQ1Bzs= zAc5K>LLJ@Kup*gWbBa_+Glca~_mrn)D2STk+XyV@P-zYFyQtLT#D!>dVBs|XoG$kM zxG?x{s{ntacLAN^YW}+}3WO@>?TD8XZcjYt zt2yC>dbcen-&V1S%SM>To-K>M)8^3AS`!Z}s5F>Zf=eZpf#BqpXfBf$73z`dc7*nl zeF4q%HIUaxRjqa_5n{WTZb@jOH~Pfq=AKgP8UqToxlA@xDk6uhv5!p$1Bhss}<-PrEa9iB`g%^QZ>j6VUI$Nl3L~*O!B$~Y4WC! zb&%76Oinh)p@Z6GAx^1FhHQ${vMWCZB%=(l+qBkY2Vo{1f|9pXI1yN|WVio%n*%{O&Vw<{;2mL&&aIPlpXBn>0=*N_u59qF_9>No7`m z+)JqoD6*4^u2q28m!c2Yr?b9!cfPI_sw4Hjo4hXrCD%oHpX$9>v94Y-Si!H$r}0g_ zW{|TQI8T8A6=**OT#|%cvQ*V_H*?wmIk}#Jn5WO%lkYkXVNZ@dnln>ZeDwpyVwQe; zqNkcH=C3c$lfqWCc4=T3Y}V@m>Lt!4xjU9&x^gNc=kw+ zz2ZOemcR78vu_>N)+#MgxfMGHPjBIA*(R>A5d8pC*Dc#g_fE{iPBmH5-(OOOC_nQQhJ5>Q(tgL zpmQCl=oc-*MvFLc4qfK#G_t*x;7O;#o{1j$HQXQ~-sm@frCe%JcD8{}hdPAHEzF1M@<%4JG@ks{}*=u)7{L8rvb z`+6{PwQ$Ykee*FVi1FH_mlk4{9JwkxToZ#I&AbSyaV)}z_P;ps`N@SUXzVJP`e^1O zpv50>sS3=eFU>nQx5TZUIX4#z?hC^)yz)78tM2eP3mfS~P}DpGT0XGgEm~Btpi5++ zkGWziNVXR9@32YIYoHPm2g1vM6bfn{&d_%hptu>Tcu*=WwN zgZ0y$f4ua(_kb-GqWn_3;&)P6{*p;Uh#6E}=1FhPYNz{DE5KoB zhnsTzOWcmWS}ruqIOGqsIg!2YMvt15XO`B97RmXgJZ};rM z`x;b-15QJC@EYGweQE{C(j`HIzkj*-fReLk;!s#U1$0~y#|c*atrdBzFzekPT43d$ zdA*0!b90Ay9Fyy@dOF%DH8hYkQPEAJ28hHahZp(eOlzN3GOc6s8AX?%$mf7}pZghL zZvoXxX=3p1sY${H`K5V!&0c2ZoF>tIr4g7hu9IYvBs@NUjR4b1E%1!BOA>q+2W5n2 zKwmuO3}VD9$W>WkcuuH+J}b%A=1DuI7{h86TO)EpQ968=pgk-jv{8}}TC2zjZHL~* zzr>0$hlL&*)OxxXLo-XFw3O*{MT6-xQn*Xqa`9*2T=S*PH_5h=_qmGqvDE9+=gc&VT2d31E^5; zFF&$P8*nM6;<3Zal2B+5>653c^_t4**jog`$}8m+KAQpuTu^a)tM;^ZjFZt^Ff+pG zAJJ|9)qDJmZQ6hfGWfg6BP^u|pW#z%D-nlPX zyi~tmm%L3jP9O)YoWUKG8VvL{D*8%9ize2!6B=#qDT_4M1sT%%sKd$v`l!G_W9dj5 zV-!6S9)!Zf-GU4_E}5DinxOtLvcl&SDO6Sl7^Zf3IrWJg(YHSalfeG%4-;$WuQe{X zF?&(uAXFBw+DSg-@rLHS6^5%RbrnS_sc0lS?(?`2)&ezF?dlrU9Scvs9FW4nodtXJvrH<)EA{5&VHxJN3TgxFP}0B$7M7axAI_FUJP0? zu0J_peB8%dZvFDvJ0BjeN-R5{J0X{67J>mpaR*(q7w&(uX%rViURo1x_BF3vZXZV0 zgf_@ql}WAuv|OfpYKx}EoEIwz+czWAzt6*o{4Apf5Lal}cRf3RI#z71waUH;4~l7QZGE>{ADrjgnTBG>*z7iA7-xH&_a|4EKJVEZb1ttzjv2f~$&*WK*7xxgZ~?~So5tXP8!&Fi3f!Of%}%WXw>!{`sO*K^iZ0!MeR!4D z;C9UfkUY;cKn!ebu>UYdT=vr5`7x1MG9y8KR_u_ES8Y}qMe97T`sSNn7fo-c7tgo| z9RN@#L%PFb#J%n%LeS0h+)ZMmKAxJ(^pe9+g2>74EMy~ear#3=vU)Ss7Fwog!V+8{P7xr@;OclUBuM@w(uB z8r=paR@@{)^3pm@4Kz?^YqPz~Z#8@2rP%Xa8k{Fa+FYGJa$gXNcJKAlZIs^*NvCsQ zA2fTNM=`Lzppf0T)+LUl)98b4B z6t6H~(SI+`G8At#Hh`XvrsW~hDUPDZS9I%@SmU9asshX?_{XYPhK}oufgiv zyAhi}YsWZMXJ(o1l^=80?G#^|a&5{Z@1@=YF15hQbteQ#Lv=`!&?G7q^}*LZ>M#8% zU&yHD<0<2g_AO5H!QY7JmH)TzhKt46c!*sB)yesN}|*W6j@0{UnX}# zV#KZ5vT3>l-q`23#_bdtJAwEwac3X0zr22|j)LlAwmouE0uLtW&1GcG;1fupJ?4Us4xlWU#EhFvDCq%Yy=7;ksoQnZ&Htl!+ zHvXl_7Smyz7jl(s=CPkrY-J2~Qfeqs$)=)DsW(U7_Dpp%$P*bIu4@Jwr7oax)l02uwUI= zhmCaz4f5@R;<+YF=`~m}7bnh;ntDC3?ga&an&@2U>cUzDXCaL{v<4w&`ROEh6@Ku? z_dG2t%l9j`ULfmu?1${NS}V3w>I{l(rJ_*-`Ih#80%q78-Y5dec8t{z1biByyRNE+ zFbB{aOj!mkd};o7p$tFGzZB?A56@rh@2vSAoURei```dh`pLgbhCY2JU`~94VuMtDbWKo=iD;Rd(UqgRpt7ynu=Y*y3b$ZNvSAFSfE`?E=r1_Eclx2sJ3zPMv@p=E;ovt zIPvfuk7DkfH}UwnxHuYAd+7&LN293AV^5{pYN>jxsC$;irqg7IlMJp^dd&q=7=`H_ zT~~N9Q2V4ypbfi1C8RyNJoG~Oh~c~im&Tbu|Pby zjO0MoK@vd}l*6k$DX;+1DAb$1+ zC>v$s+0oD#f|pUDgp(B-pYi@1@7prAvrwJ4LV0{Ty%|)n8H{ux^VYlOd~B-fy2?a6 z{a(Mz`-R~+fd=~Xusl!}y(Y`2-!pGKVj>OnPVu`XZtqvny})ce_{YJEh2!RBRMYUp z&D`dNw=t`4{`^R-ZTk>!9h?JJmgjL_6VjiUJC;yH_kj0uAQ6LZQ``#2kv>5g)9#_W z3=JU$8cL#|=>vKDcPq2K3d8Yux#ya>PNvKq9fqHPp2JRX!|xWL&#$fhb2R} z*Q?RZyuPHWZhNAX!>Fp^Mm}%$0BGc+kNnOz0``B?wla>l)`O+iZHEWaoBnmt!3Bdf zru9+4TS?Zwc4a;)P2q8m)C2M?5@jqRma)Y{8-sGVh+zXA=xf&IYtWh=mt7jr+_zO7U zX#5W*eD~bfEI3-a?z=xAWjw|asL2f3CY_+-A56>eP!;gpS z@Zdh_^fCaATfb6g6d6UW+E&OLb*cIUeFCHCV)z;D_JF-!dd(yM4eXF> zt!66b84->9|_rwb}FoQWCLVv65`{ z%y4U`_jz?d-(ww;1oKOF+rnU>yskWVI#(@_u!m zHD9|b+YW7kpctb{h9KQlSsRdJT#?l{L2UPFlg9xWPaDSJ?{Pzhn_$YkX%_Ev$?qqx zCh6QZJUre7mst6-yD4=(MRKWV%p+r7Gl?8}1>M&y?IHgYe&up3wnUCn!5+vIV%zed=0l2na?>KUE{{3b^l1^Z$ZrQ zSJxJktl>29c_911=fD>%n*3Azark$r0RitiYl} zz+r)nBQe5meC}m*Jt3XQCEUhiK(Q=XUCPa&DDV*j5;WV?P^i6WdaE|Y7ZpEni+lj2 zyT`_du}m-*%s3Jg&l;;V{HHOt$z|R;CoJh=G#f_crbw)ZF>PkCyyou>rC&}xyI~l{ zUmbtH_eX05mNmx3-zAPL<#EhxyVZJps>@ck%h1xi=Ub{hEr37+? zZR$j6hF7<@UbDn6A+(Cg2t^&tSf7&~70e}dm+Iss&}y1oO>gx=9|Y@b^_p5yp%1c7 zE`51ZAoQg+$PWkX6*zoFu^Tq*E~pc{42}ibUVn|8AI@dv@m8kW z%Ea8D)XfyR2C`zB9H>ALl}gS6SJy6q5mdh7Rpwd*%m{af8vcUf7=RpN*>B~#} zu8H;vy26vB%RPJL+ol#ME7S<6MCo4DNo^v7I^nuFaTQYy!G0s~i)Di@c&>WA&v|-} zo2l>2NMqa4;QXs*PqQj9;Zj2qXjsxE_(K-ADEEbU6X02}&zhxVVb zGW5->GE5mF>N>ot=_7Lwxb=_0ySFhlKQ@NAK?SN6#7jCTctL}FvJ@{IsHEVJ#{ z)8Fk1`LWnyihf@>@lRy+aHfdIo1Hu>Q?!LrZ$i*U*D5fDQw-UhIDg~R%cL>@bb{om zf)b6cjmg%+n-VNFFBNq_dk4rYKV50rLErfZK!Aq8LBE}S79Rnh>bPSAE~d#GA)ag(7j|Hb z@EzAx%lmCv)_Lm|voNZcP1JkoY88jIl>s<;&?<)IO6<DQ=<7;lLp(NBmODA850*C(iVoB z18cat9MbAq-E_$LR4QnKJ3rB6xc1Rm^D4A(sdM}eJHcXW0*1bJEJ`e+aRSS@9}B)M zms^${$z92BND`0ZEd^E!Nd~3fN|7`(-DI`05IBx+Lx!+ho+)im7b*Ky$?i?$EZqwh zr7v=;SGuPGHqW1E@&(n(*h$X8gw2GG1R^#=IOQcLh)}I3j!v_6T)Av*c)Yu2acRq7 zr)m%z#3gREio)hft%yXV|al2)R27PK}rpEO?#>6PXt@A_>tb|eSb0#$F1>QqJSUh#?%H6 zND`TT7lY5%Y4G=q`oLJIXetX#_f2F5To$|b1=iAQX6RDp7-t<&Lx)F(WUc>3kNxuo zTynsb#n8)c5NSg~DMz|LBQZmdLG#ER5QEj``ayy0X8|7xu<8VL85aAW@is}yq^fRtbt&@&8(!9HKJ|jt$~|Yg z6Z@I%0oRj;V(1y?fK6~bw9Cr^m%5)FSV%U#FmO3&1(#e(okNlBRCF#h45q3oRht5U zqqNTh2{B^C@S;I}Rn!SdA~PfD9=BGlsfGpXRm>VQu&Is{V>h#IyT@Lp*{hy$l9}Qw zRwicuVb=4$+F%dH%32qw7DR}o5)lW z1N~6mqk-Ro=MMWll*J31mAYdANafNb%G3_b2Gtoli$?yaDtQe6e82_oOq8O4g@Miq z%MIHHA8FmhcFA&a8Od88K4)oUNd658s%)dwkQUlPMML-=r4ZADP|pdU&+}w+D6VB@ z&}2;1Ef-#zS57y3b#s@ z4Nd}p)v)Yn7$!+L?kn`x+~<8aRN>R2LfXjz7lgzARnizH!5&Ch&D9va-~@^BKmO)V z6Kr=wT=G0Tj=iz)uA@Lhs%qVWG!$B`3QPs++AdXrL^lv{o<>R3g0OgnZf8Kgyg?2v zKRZ0m(&pGngS;LbN#{B-_F!t*MaRiiOq2gHbK_{zJv^>cVoCSlZm1vHk5{YrK!;a- zRGhjnDi(5sJ*0!qXD&m9YZ}u*a{Z2J>*dQmZwJ*VYeh!-SXjL>M|*Qt?(_!GPQ51E zNykTI2+zw?JeNe`xgAlrW>*Ne%X&jPN5*=>YC}eW99uxK{*IBnap33uUAC+SFNl3% z;XA-Y9*}wgKWVB837AZ-3~qL@LlrEI$a%K4x{&Aj} z86{9gf&shkNmAXoD_8Yc; zriZ=VX(zckC|}Ln{oF_JZ+jw5x-M-}4xOOR6Qf30#>6p3gK0YLf`rp1o_%0$keHWn zwa(WP4apz)!2yynfs|SW81_(V*kI;S(c9=|X1#o!bdB`Oqxz@|(52TYT)`B{D(EKi zNLUhlM3gaYEm<~Ozo3oaxy#Jb8Rc}Zyd-#qYkgFzYK?al)6A4B*GQ{JuQxp0_h8q| z*sjzzuZJ6VwC6a@41e7`*RoCVVHU_pT<32mdtR7u{%I@Zms4sW!Z<`ln+sv}nj5N0 z;Q?<`lIRrK6^3dI*aViXHSdE^4L45=3l^k^6j0!%Jx*6EQB+K?shkR@(OapV+(tv< zS?|&30SRSfg3G5+nfp3P6k;0)oGcCgBq%w&TChdEa?U7Ila6>2Y#xlkm!oNq3UQuvWmF{!pVlg3sAg5c{5G;1LT49xVL;^MuT?aO%%IDrudB@Dqg%bOjNvj< z0>#(nKO

    Mh2(HxU%zryFF@n={u337C0&YtlLU<^SG-A*rPxKMGvM4FU>_iB+cEchFVS_3d)*3 z;8G;>u=%y#@EG%|NWO1+#GxJjxTB9h8rGlz~ih6FS{xAU5ebnW{_l$ z7HD0@s>QV)4`sUJ0D*FCl_ye;rpzr3#uhQ7zb-+IGzp2MEevWO!dgvhp$@OZqRXoA z!&(r6fg?0GsVck-@5lFHA;m+zrjxWV<;*9L#ztz(JZUQEvX;_=XE9;1F`zVHx%jqf z;oH0CUXiY!yeZ)Ag(l<nsptyBi!=YgtE z!wmQ+jKS^gIPodro|oU|ZOP@yl@n^8_x|~l%2KGq{SZ9kIxrpR;pjD8O5k#8katF; zx*6o)MAa%1d>03yc4CXF!K+00|H=CnxTf+ef3J8$^23mj5XiklP#}UJjywz_;zXz2 zZr|NLcDMWJzuk7F+q!*jH?{4wQ+%MHf}o&+5CAc3AghP>8g!4ZkrO1z z8}r5SA#pMAY!|6Il{7Xh4RqU!x@tw_$Ty z4)vEdz5Of8K!DRSH?1t~GlQ~0urS#JCmthJGf|czz_~<&y{&0S@X_rb99 za?&SUGY@-$q(&750-e*4cfj&B4K{){5hf73#)htSG^Q&+I0%^j!!A|8{!3>zh!3fD zM=n#JfzHL9(}5cdlu$HgV?eoWoIV@Z?^X{Ri=EXcJTp7DiDEZWB!!AC1qJW-ZZXa3!I-B8k<)e(H|>%85pf$-+kYfKF~K6< zpKI_w^Dr#O&Y|x^OLO$EdMEHt6vu(N0&AxpdMsOVnrKx??DC*=<}c-b z5OsqRBrHEF_q*?!&crKrL)@iVStL0naM<$LFmN0O{xb-eUO95)xy^KM|1sblOHr2> znGw!AAp(bLq&M`bbPGnzuwF!`($PDi26+wXr1yJhl}Chn{=GL+z55{G3risl?&)J{ z1%V!$6IlLArdZG@o2`{_(whsQwuHlG% z*6qc6AO7TLQWhe(~6*UO{ob53r2YVo@UXSt;#J_a2mW!%GTEz}C7Pcbe z`Kk^Hb9aI5ok@m^;0tXtG@qxuLK*6hW0?VYSU*Fj~+hkf6wH2)+ z#jnlEx8BSU9imteoZCmmR)qF5LoOeSF4Nue%K1?2I`t#wK3Nfi+i~j&UM~NUybqSf zJ4I)O$!`p~bkgS)LoOHRUxxW>YiQ2g?4ZNJTGdwH2ZA#Bc@-Q7{aNd1Jf}haK!ECH zyP^j9c7)#epk`^M><7JLzm$7hYV&h4RZg3vB36|A^HdcIEPRA25UmQid3mrhs-9=Z zv9@(|Y{P`DSF_)7zy0#RAGb{JS;?ZU15(h3suR2>uU)}|d@Z~Att;ZH;4AD=5PCa0 z`!aBXu#CLPtB()8<)&2eK^{xCpG}wosS-Hqa==4(2x2H(!uE#V@`2a}JbAamGiJ0f z52LEWn$>4`ZQ;+4tWw^Y^GuNGzJ>JA7XvN_RPjGi;FAiy*IDe0+PLsnV;w@<#=-I# zK?DL$&@x=liQvcO^yJo05XH} zfMtM1Ng5PKej3suP4v|&+ol7{DV;%fzTBci?lJRtb14=&`fjCSv9O?7{p9Toq7MV9 zm#NpHpqU|kIC!P3Q`kzfJ@Q4>z}#Lf+8NpErd4K5KgOm4>1wPKB8P#qNmgnkehja< ze83ow{maTvj!iXTCWOf`Bq4etIS8AT(&8w&ts{ ze6V({oz7FC(BN9ni5O?&FmSXVV;no~r*X0$ldDfQZIGL=kv;7fzb5HZNV(Y%JGm z$FPghBw}If)`;_p>X3tme#$VH8RxNKmnXbtg7#HYE!73L(|5f#&#Vs~cEPv|%6+6W z6^bXktw0&Kd$z!SAhX>j+Hz!V`fn6rrim#ry|#5kp`Pl;5c%|E|tDO8YKPCWlCGb>N+quBct>7in;En3fS^eUh=hoLF=weSp*DJ)f_ z1+=SRCdrtw$#c2-qc`qIcg@nz#LCmm$Y&9nXY_DPkE+)1nc}2hrw~Oy638CcTg*M# zC%mU&Ph9_`DG(f1eZtFPGUA9K8T6yKdk%Yn@o~Z0AxF>aPA{{9N*I z3fW*bxGN{KFxo>@Y_dnMN%Xou3i~b?C}^;-!PQI~Y4*y4j8KQq@}M$ai=l=+ooSIC zibQ5Cj!OfmT4HZ1T)bdmaElaGeezYf4+AVHb(ttT9`U?r7%Q|umIZ3zYkBW`Wcgr@ z_%?4BI4)Jf1W_9-hp;Yr5C4o92@UvtFpjI6_0;|2$fxexT%UyBlNF1MelUD5-q8>& z0(#y`4{My;g*g%U-4on)@H=Uvm>G(0V#AqJ@^U}?rj1@X?0xCQ`LD@nJ8i6T0Gk1 zRD`w)igRHONbO6~dFyuPHr3%l|~S zPd-34d!WX;Rv9NkQc0Qo$b7B$oe5J{<~&C@@;;Baw)I;+iqwzuE=;~Gy0re+CVc#+ z@h7v$M^5ZAtudQbpHl266zQj8Tg8W>_sCNOxMVM+jUr_6A;W>q3Ozr^wF>mA&MLPA zf$IAqx`j<*b0T^rbsq1#*6`owl?32P!!Cvz0(^<_MjgE+NHg*Rb#(m9jbZ!zz?Z?# zTFGm5Yhxg=PG6U748xP5%mDil#KoYPE1dlN8)09xR=3k^vjTV3(QQABivhK?LeLTz z4$P--%L@4WK{Wz*9)?}UFNzn;qb;s+TZUO!tr9zBv!GuLeBnRe|7rU>OHVFb zdSQ{??bG)c7G=6WhO+968IQ>U{=+$~Zu|W%2M@Zh6AS_yXjfprc(X^5Xj#OH(DMrW zYG89Vs4sJ^M;OHMzT{?5A1(a%TbB7zE+GLYu3)f|P0e-hWBTNKLp5mu{ZTqP4xIgD z?{aTg7G`@4NIw>@Wo~(42y)__B92HJFFq*C+xg)GC5;r zXsRgoFhvex%h0lgC3JRh9BAF?cs)#NbhG-|Y*_Ov%4a8p?DKpGqkmqp+iM-HBFmUl zs(eMq+%ozo=-_Sk*rj+FrB$AAJt{zT8)(NVY=!mTI>AogX0`DgJX)*VJ1=o+oTy3C z;NE2~PANy2yv*E~_SuYtmmX-{;4HeYTQVMkj`OC3K{#(@9Qsp&j5rT`E6uX%sX;kdd;7^5Oi$uoCIZY81bLXjQd0Tju!AFDl z@N;;0%!ExrT)|a^EjvHHQq|Z%+ZzFa6pBrv$Qr0zfk^Zcqy=GMZ>uuZ zS06d<1{Aj3&lirdZQx%x&EXrY&ne3C^2%4eU2H!ARVc5$yiWNla52hfcgTwaYy1Xd zo=Ko`iE!OXF6I1=y!tfOJH7*Wl)9{X6%C{|1{ zvi4Kn=zp_wI%a)vZpKUiM)RS+CKN7Wa1>=sC+7AW?j^S|n2_6V{9does_;ucp0)6A z%O%|!Q?Jy!J)pCK&$(s?KAZ~@o(S|Ks$NnPk`)Xk3EiPkgFto!qkwOf_!jVqF`wpm z%yWbKr)J0DnE!b;`wx~)8!o4xk;AUAg`y`brea24t$<^sMlQ?Q68mdbu`vCI({Y??f%;^#-p9)@NJx|*Q756 zzMuc9FIV>POT3r4_RyKq)8uH#kjoZna>&t;Li*v1oR9{Y&aGXT2Mr+(fP!~JSmpdd zrgj2+2OMS_5b~+hVAqJ}=U|KzLa2xT@UN>yCY$lw@~M9y>zp`sS8Qf8woz;bhW}%0 zgmtrr{P#=FNglZO($}Ufo2FH^d!CRstC7cj)8{W&)PLRW*1|mXLbdG)vM1c(@j2sa zr`_D(;nkE-YRTnx+TtTC?Cj^uXz1wB!N7t8qh9HJy%KAYFd08VfG~|82VB@`#BI9_ zCtOS}E&a2=#wZ8Uyik<`R5hSZ^qA@csiC9KPSJLD3u%u?gQ>9n-zUO6{Q@&T5b(w7BmQ-=xzl@Zb%|JUS7WxBcQL)Gm!sb0# z2dy~0R`1%u_)fN zgKwB1tAaCGz1!A^wy^TxieN12ul9%7X|brR7bZl?J;6Y-AI)9Ji9m%^n>-gwmm?8oSPKs)W9lgm$x)++aeYG*r2=I!OcvIVLy zzk*}-a)WB~*Ju0vZDrw%9a76lD{9wn2ymSptmd|8abEKW^7|tfEt@GejUuT?Da{up^9-$lP}Mj^pjEYscSnwpi%XbLgxF$Q zK3$Q#bS&+=lTxq9whWp&ZN$S0XDx3*gS=y|@n@6zqPkgqn+!-3gPPTAeYYse>9Xj& z1!u*{6O@B9j1Q;bp15ExN8Vse;5XJ<%DlZMOA;%haL7Z2zHzgciPi_S)61f^1?!_z zd^eD6x}W3-E=hA{7x6!m-}gNsAAnTespuZMkgkz-3fHLXCAGZLhyiz8{ciDY3agjY z(1%nvz@LeO#vZGuGMqvkB)Q#qOubUMlu|SH1t^=%d0*(o51~Dy?cQuMXlP{iI5q9hA&F6upbTd1kY^ zSGL9bY^XjmPBi4w!5j{b_r+9{R;l;bxizcP`1O(;kkM!jYv5l1q0D=-Oz9<36^7gD zB`fVKXKF|2*wuiAUMu;9WqOy3(Q#s5#Y!?5<#)P4BL%x)m-%4PIP$GAMZ7uMfl*T{ z54_E|uyVemHV1ZP?RwQ0x|m`^Ii%`&n&%p!_MH!Gy!p0_v5l9q>tWm$x-oZ?|1!@s z$z(G>EGGM%m?!js*|c$jVnHsn+L)zH5uhR#WNF(-ILFtMWzDeYOzS89NH>0DLbVNVTO}W$#HM#tV{*AkLqAFiA*As1M*EE zE;``;kvnQowA1}_5*Z||mIZ0n$WTnQgIO}RfQv0S8qL?>vTZB?({Kl=$&hL zcR6@{K=I5b+0l@jjE<+B-RjU4@wl)u77*BY4kxToOWeMA(oze{X|ZBO_X<6WtuZL| zFzAk+MtKynpyFpbcKk;l4jWN1`pFL5$_W*dg`2+~`nRPSf}YG>FQbgHCY{OnFRe0z zfU@N7$aAWvl1rfUSOv6_iS%A=MAGxA$AQ=}zjWi5cIg&S7^3v-)l1joYwai|VkcbuKv+L1Aa;y$rBk$m8_6DW6 zb}$w4rO9jt#8MAR`s8_mox;r?d-%6yRpQ&Sj2XJfLBFHHD31>mM`NWro?b3*7iP0* zD!iPY&5AZE-EDKd;Mz$qwWCqu0GtuV$J)!6Vu91n+izbXmK_FPR10z9V5^m`gSeJy z_iW@B(9P;(2#(@d7x6BMYvfI`9M=*Mi>UVB8juyRBjRGj7XI1jn_`_?Bj{DxamcNW z!`Pr;?P~VhZhbELq4G(EC9BA3mo!%BP8cmO6=#5paz>dPruW|&J>Y)WbpnR_g#of@ z{9icB@f)m<|7eOKe?RxZaZ6vsX)$4i0x-zWr)!`fJCD}$VGg;*)7*OllvexX@m}zw zS)n;D+2h?PoG_5sgp3h$g5xn^`~*(OnABb78SrnW<%~$Q@!RC`XcCA{TrTs-Y>CrN zu^kj?qheP@91X#U>WA+81NVCLfP$`WmR_(;b(reDzNrgu9MvMNHWeBhoh>fiwbzKNIYQ2-x+62V7jz;c<5gOKbX z^rh4^d8W;;0u2Ejod%&8JV>jI_j(rIUEBZzN(Jn8oq1ih6o_^?a za>$Kl7(7O?H5AcPv8PlSuAR{M0INZ3q_=q=$?wb=lHWGCH&=Mj-V~SywLG0$kGNZ1 z1BIjbIthw>Kpoh9pXB+Gnci2Vm@&S>yUptnH>-mY@3uiv#`uTY=6eqZa)1Sw%Z~flp0rC67JC9+l4E#%#KX$)^qdc9w~-zsfyX+@c9W2?uzX zW}$J9yiFJQc;T@>SaPUeTsS*12(0N9iR++ok_{BQo+9h0*kjXDY_H0QieM8$`k2+oHFBsKEGf#8t#MpudL8G--2q2h*& z&%2ua_P3wz9db9#DwSXQ+6A)KiL=8+X7j{WirqrNeT~IB-e&b4{)ONUUi_53Ok>!? z8MQpDKV3Bi>oMAe=M@dYKK?#Ghd%sfgTigR9h?Uq@y)FJ7o&w?PuG6!yMJ5L3v!{; znf>I>YzTg(GsZX?a^y8d>?XfUbg7%BH@b*dKwotO#^YVj9y(vu00I;sI}x*zb#TI% zBdE-#nGqPU`^%8uC<1!y%CBk* z4A6{FkGVU%1qcJlJS^&KlpO<&#eCHh6dC#Wtz?fr#@KZ&fmaTVMY_XJll>Cdh5UMB z_qIE8bo8A$!!C`oXLAOmNI<}(G4@k8_hqfzyjCyaLB~e(b#fxlV<~s8-Su*q^hEGA`*d+h@9jE z>IzBYW%BBR>gD-#lTyOuCZzk#Y{(Ffk^TO%5wsU#UiKMd+4q_t<|jY;?au`!#LUY0 zPCZ%m+5}_r%z(d%VmDGG#TeB$s<$-)#yh3m;n*ho9NDKF5XAY~$Boo(oJycARM8#m6CU`%hJvwtXpn}^LaSIuqq8uPLvH(M8(a>1 z45Urh;AG7v=PHuYf=ouE?wt*lB+rT0+E5-gA||_!VnOD!go-T}U*p~Nf{rCk$`0TR z=p#=FG}yTp^LzNg<5kFM1z?+rol@_fi3Q?jT}sHx%~ z4Q`a7SUJks*Qic-X)2*z;yT zx*9H#8ijjKx)yvJu7`_v?DmM8u0f4^$5&uuYE(XagKOlZ>z|fhTeO}(?D8zU%r@N) zyJGF-Nc?oKEoifjwI6PC|K%hTu2ee?{DC}k;?@{j%`lWiv1=%@nu;|Rti#gl;oQ2A zv;ey-wrshp=Er>g*qa|eQSYpe{V2<`$$q!zylF|+w!(i#j(el5NvVa07Yir$^UCP+ z&!pOtH_kyxLa)5r;#u$6GLB4-k3>IYVqQ+L{%jJH3{#bna_?W%m@)=_}mfKi3 zZI7CtMGeufI$3*Xa7m=Wr8O!GLv=e=q}tc@uni(Mg4MQ*+3_fDU=8}8yl6kuMDmP1 zok+A($PqJpT|%)tDN;zqVz3SPg>Eq8qXntC28xt z@wThmez5j$D;9GyEdhaxwCo1g2* zcX!vXBrZAb87N8ew_P}Fv z1gnFdkbL^;f^@Gu08q0UucuWFKnZjeKgTOY&@VYB{!o6OPUD~RMmKCXTOux(Cx>JQ zU6Jw!5-&H_r7*6Y?{f&d~@f2{+w)^LJpZdw!0`6s3rwetbRcsKaai=EN?=l;SfI^HYGXC<%Itp=(<2l>f@Eo99bD}8aV-VI;jey#Gz!li%w z)sV}cg-he8UpZnRY`M>ldFoz15Z0f`QC6SMYtn!4OI6|DRvAzZl=>ul=)<-JoDzH_ z-!0L*H%B*mX_a-PETSoLFeb|amlQA8}z&xQ6#|_z6Mi4J3N8? z5;?ws`<(mt;%?hV`o(h{xrLM2oAfUM!9RG_^M$aEhlNZRzb6Q}P<_m>3;c{i?Q>-9 zV|2cH-o!>%Y{v#X>W&11#kDzMWAcw@ivxdPa#SwPzB-9qnL-|$IVyK3wu>USsMv(B z-H`m{7ISzZcCx9F4nd;r-n55`LeFadZNUdaDybTFv-D{2Jvt+#L8w!e)5k#$y2Go` z^A0KGb*oy~-CiB?O6tkmoA?#cZT_1*j`PnbGl4pfN=jKiC3HbRoaiE?OO}hSOYSHx z%DRPkK7K}nur&HOu;7LmJ#@|)xuss-AGLzFg)H;#p+{aBzaO5xD{5zSp=UC$9=>fH zWD4u$FEJsUhIDiTK8~#x<2d3iry+G-FGoA`w@sD>XHGjmSt%~VsYtV*U@m#sr6y!` z#3#~T_m@Cw#38nHBMNa{I+hPD{C@8?oax!lFi&WPRGpNFNpZ-qPa>3+Pqk5eR%4iQR_ow(l*yvM}GUIYnS#GN2gc8PG?@{AU4Jr5@QJ zgNcuN335OU?O`+N=a*yF0_E&*gp@rd&T^*j!I_x3{UjoK^I1|&r^%GSB12oPsE z5kPD>5rQw*xuFtBqYO6jK{OtgYUu5Hj8sHOM~(y4wz#@Wx?#OcrUZ*5&2r zahual4j$e1h6yc!ztk@#Tew*&C!W(P%&=8Ru}~Sgor+C_xXZ4Ex0q)2GeM^NTF=tp z0#z=5CDY|z#Lt1UIqa&MtGYIOzoaj!L*B>e-7@@7(WiNr1sCX4=pK_5S{m#iNEruq zJivN6G{(308`PcuwCqe}r2$kLxh`p%XZ+c$1{EcG84zQT@zOyz&C3eDGiRSXg?(^#df5u4k((IVI1BxP<8$br`!@k>CJa@tEg4;pA4UxDk(<&WB|Kf)B(L7|w zApd~hqkd1*bn}x|vY4bgv2(J=Yy!!pSXgFnrD6w265FiaC_f_1mR#~07Web+kdJux zA?FoOHnSI|mGFDr_XzrV&FbrtE8<&`Tjw>B+Y_SlG^SB^%s`u6m79&5`90n_%lwiR zeKuqLjRpn(p7Sr;nBHjMVH*0iAnOvn^+s7U?5~~iRNA3zQeF@P!8)0j6Jcaa-R9NN z*;8>R)HCM5vI5J}`uMs2x91KrTMfORuvs$}KcmC>e9yT3$}tVrXiw@3+sy7g=Qo@N z%Xt^7Up>Kl>x(8!vwH^b0jYV#h%?vCtWN{Qo~59KH}*tG7h5_1aNu_7uBdX}ed*n~ z1^hc?r2rIxyqCH57^BP1Ir(7GPK4kdWF)4(*{Mk9*Fq+=!L^TnpA-e^ zpamlcY>FDjrsC{4{D1a9a0Wk|U>*N+^GBf(Qu3L_Bbt zNc9)Pz}Sp{u?*ul&asYwQFt1Y|M#c-+m?xJEBWF$Q60=ZS;W9@_B(B10Do{&OQ>lAy8W%Ydr1+u$cpn$t)$Lu zYO17I2(^?`vBn~p?W!(5GF#$9^672}#$w95jVY2~@&A$~OKKp9(#3x!1ct9SR6@ea zNQnQ#QfSE%6pJt7C9&mEXO%gw#xQa|oeHxI7E2b3u*Z>RxoEZMF07{D_;H6}I6MNM zO)56Lvhu{)yt!{t#`b6vb~fweIiz9=X)?R%Cn@$AMQW(ny*{1va?zdHj|nU=+n7(} z4Zx^P2mR10x`fUS&WDCSHzYcq23IYsMHlF0yuMkQ&glbW&^^niNr|U}^`RT&z{W;> z+TAmENi~mU$-Dt+iBwZD|C|K548txv>BQ)6h|Av%#Y1t%(5wh-lx0O1M6C6^OpkRe z95n);4U)|_-RwYabAZztp&$2OqAk_TUK6sf745V5km0~Byvks`8?v)5hJa@fD%G

    1GYNg_di6>6CQ?EMpQe>_gc;y zx@LgR4oQG+S3{S8VT>mCGUZxtJ3J9wvH!I%Yx2hSPejds3a-6v1kJr~efyWB)`7P* zmrY=Bnqr`l^*HveKK%EtSrBE{VEki;R}u6Y?}kF|1=F-Q1P9#?@Rt+q=^4dBV0FV* zaJ6Fin^40CLxuo7M0atX%1f#slDv6 z(F5;9{i@ulE8Os|y&$PoJwb~ZBg`KTKs`UC|0GG$&927I{dSK#VTRxi2P-wUu(B3Z zP<6p&xAV&H|5+YtUL)tQ1uraBZ#RHA)!^=eW(=fvArCFnVm@#))W;V4FA;++!YSVK zy0%AD<-taSVEtEY)s-W`hgAcQ@*-~d*gxS7q&k2c0VYQz3nPKniVW%1!cg zHHQ3knt0x6`YV?u<2fuDmcB5nk8z%Ht4AK)g`DgOHAnEv)AZ1W>IXi@$b4=aDhIYF z875B8YKmD!krh--Y*;7Vt2z*#Chkz95J`(1bn1QV7U_T1WLgKIXN_Q=a|h;qFe zJRR3};HhU>O0Yx@IW{Wk!q8SB)~4b6RjO`PFTb0vn4Au)2eheTPwFGz&7{_UkLIBp ztSk6(Qu9#O24Gq4s%@H46pEodJ7LB;R6MnKm>jSNF4zr=?>~IF-P349{^iswO(fNU z_d*~7FsyMbpJJd3D2s|YJ297cS-uW9!s5nP1f{yfd3I^KgBOkO4(a7=b!Sc)xP z9k2Yx-xsY<4@6$_cz2b^<9$19kxOA{M;H{n1vN+Cf3IuSiTCRjhv+g{mnfS?z9>+J z2FB(w@nS{-D0_iu__&Wg|KMj*qdmzKt-V6xEtM`g@cL6=VqdL8{Ah;n~PnkdbrdpOC@b|hiaih z=L+hmUg-zsikvS=IE_WC6J_tNkUYo}YHxca`Jk_XL^_*g+v$z|aPS(m zs&=SWx?>O*>6jrWa1k1nk=GUp4j{abv|FcqjS`Y)$uDu>9P=R)V^dBsV0}udnDfj$_iE_aLdKm^x<{}B^v>$}pG%sf@zV{r z9L*O0t&zBME>dodO!G-{TNw3#^pXswOZkA{d5}W`x(&nGt)AN>2c424+NBT3)rdsZ z&^5L+Mm9jm6LVxT$D{keZhd~S|DB(_=z-W{|8Pn6F|v_vp-<1K7nOOOlN3pNWOb9X z=_V!}NLB6!uZ&D`sc>22-0lG^T%-Fgt-YH&7OWk4bl0(C!NN*JJ^E1hL-Ssqr($A6IurZgKrmy*_BQZxMfMXl&p^ z)fj;Vr<5CDMh5~ece4W~?ZzMe&AiasVLe)wqH63Du8N2kluEQsOuW?KP$UT2gPYab zMsJ6jg&^zt$!SOFa#^Q8qNxp(oGT@(d`>9}K)woi{Gkvk4+QAM=yj?Uv9B&gAPh%Wc~(i0Vk@UA;k?z+qqCkPC&xtq%C)3?xVLMT4G0^CK06FBa+1wCJ*)?ey4+J+?3ot?llyPF%58-w^Ui!B);3cO^ zJ5`4WCZOkw2V9bcx$+8uHeT98&ljT;ye%r-FWoPLR~nu&rAq_1Rt7!xZ4>4DR!X`w zRo=j%8uSnZM{4~uc(;O~PN9yR4Z1#k1?w-6JUs9~+!;1Q}ee4PE}J zxtqh{IW(#h1Xoo31o>v;gKB1K^Tqm_vJ_?;zgO7|6nusri?LWDjQ>!s-zxX-XJmJ< zKfRV#3h1IF!av6R=Wk_9 zlh$r=#R%Y1Rhqh+PTg^dLr&9`yvI&>A&LStDb6VeR0WhRJ!`HHz%!StXM=&$ z*?+G6Sg9}~=jeO?c#*7rX^;ct8pHNgn&-!9gJn+!2(HI3V97$Fj;! zWlGc**9$yMF&dV#ZTwb=AwWDPZD=C{38(oU(Z#c?*Ubzyf<)pxYcnb41`-F}e4jJ{ zNj1e(P-G7kvzf~LW~NII9WU*7S0)9AZ)MiCV5)Om7qbV1z|N_#k?D@E^;w=)CpuCKc?{ z3H4oihv!j^P0Y>6AoNUYW0Tcw@oz8fG>_tO$rm^|RK|ecJ7`|7TF0sl*dDnf>QK-Q4l<{4s5Dg>3V?x#sSzqk3C}#qa(@ zJygFf_c;qyw+kXZmy~&37VY&n(6eFcr(AXnB%>4Ezm)yY7)>_Uh}~G!CK$lN$?v8r ze-`weUm4e*-?eP#T6-fhxelkL^vJ*$w9tgnvW{n_|A1m$)N5bXJu3y3L@0VA3 zqt+~_lH`Qc26U)U{PO~@2FAJQn-~d)@z26YQUR~UsP36A#HbUnx~t1)WC%<*p4?5+ zvzMz;ogHS&?kB(c*u2@^VNDWC`}{$#Jl};(QD_@|XL2*s6?`h-KBSickpxMGeM=nQa+z$7LyzimP6p|fnl??5IX8#HLW6~E1NrzXgczMeQmBW51K|af zZTGy^PKZ^Quy!yTUdIU&;}jFrZhv3pxz+R>`2S?pHt9;x99jf~a!dUZ1y>`I{Zd8e zAgzFn(>GP?-n^|mz&q@Z$Hquw&niGYwM^I$IV?kYmDl~(-`6H+k@9Lp2k%7SkvBKN zmZya&2Y`<{W;QyUp zt(askwaB7qh&A+up_p~cnwd@KyQRukGL33I&om~BI_Wh*ZJ@Kg)ju{0`$my`KriFe zPs(T9on9j^2p{0Ldt?YOAf^}i@r5>ZX^TQHP>AA$$$?jaSrXaNJLx=NCn)2c1CiC; zF3n2xLiNgnI5=DnI}&-a%0N~CBu`V3exo#Kt6E>8ja~^hz-yt?9XSUvahRpi=MXVR zXpj?H7O|4a2*F(uR#+lwC7wgd>sx-E0BGbnt(xGlwC9|>3&+g`ub%#1tGUhM64-O# zy*Z2NHKYbGage3i>8a!GfQp;b5|65TPY&_k-c^_gwlH&;Sz zXtAJDQX7ESO;%Mxy*VbX4EuG{eX%XzSIrISARxiO;SmwBa+G><2JCZYIi z6w^tOHY%o3uuI+u*}5!xfRFy8wpG$5Y6NZ-ta5@LQ{={YtiI`gjXniLJLyt=>>?eS zQkyk{VQA_eJ7Eg|mKenIpbw-c(oi5Zw?TmpYNL9aYntmh*Gf*%hjuazAM1u=fItZTUOL4s_!y{QcM;_GN~BwfJ9sAat0aYw8-kA)9g`d*UN7N?}5Ix z&A>00ryhhXP-PIv(l;nDg@ad=hNI(~FU|*=cgrq~jV8i6RIr&38~qU$hG9pArK&;6 z|Nixl9!8`z{&MGBlH$Nf0fDMvMN@ec1Bs>0R7{1iQ@M)4YmH+;uUW_aRW)@nu+N-Pci2x(n!Ue z20u5CSM}N_sZpy`$trYMJ9xDL%iLB4VLN{={iS3%^25+=K;%&3(MfL=_mR87^-8^k zLZ7%ks5J;9F|iZVAQ*tl8``Wd!;^gaUFt1MUby-Q z*$yJa=r8vZZF}$`Xe{VZ)w^7c=mZJslqrvcwbdfSLHtlLd9F2lzN58OL*V52`dkDj z-^n=4H-gK_&uKo1bKsR~vk6?1DP}cAR#7ok5hy@E0Ae-RcD=x_G`vGK;9@W}c3gf~ zzFVtjn&p0b9p+|q&VK*$-<*xOIP^yK-^fBs?fVYA|Jh`Ml_ZKuq)0p!V~E5;>=m*T zbCahV0&RQH%So04`xmt) zCZLRBwo|0YprVM3VmhCCxptR7CY%!l-SiG}ecEHEBx(7KK7yy#PTc6GcR;ZP76tWm z>JHTdy2LMmXdB$10b7v`slVG{+78uKexqzzbeejbdxOZvSm9tFp6g2vO`+`m6{9bQSTulzvRBCcl>)ruR*1j!u}dLQqb1 zkm`VgUiTG=(25^7e&gi*OgwMVc;tfU_FT)qD9(iA`z9@QJ4q4*E1B59FU9BRLg6Dm z2!n<8LLXdw&<@{2^qj?>xtc_N?xgO}V&P_2)RnHG`y~S|N98*8N%z~#s3?8y^FSCM zjZr)o)$$(-CK!=)?Y~<8N|wH0yCL9h9v0(Gp%{oTuAyS;C2NB8>rvuug#l}Mag zKsAC`8pg(2GUd}vBL|_ECKyAuKj$m`-@&=fZ3Zi(J{J8P)=Wxou+YvmM`5Z>rnN6J25g{3dOLbMgWef;Lz;< zt({#dxg={*#flqc^)5An1%bIgO^meb3SvD{p6sfI-i;m)>?3ePiu?NG9_HO=4!gV* znV66aih)ALRBU=qR@M5W{+Ka&+!dTB?vxKI6QN`mif*@iEL1c=eJvi|=a(=Fl23=aOPSydncfD- z9^Y__6?Z`%cT7NJn0T&GvL#%eHP)}*Sg^x9V$Txl#&Vv`{PG$4jt4^xeD2hG=U)C* ze&twE7k~XvvHJV``i`-A!|l;@*xoXyxQ)O5(CC?dvTDierUH9ybCFKO3W)qJf-Ua&zV?hzZw`Qc1oqzPnD}BmLofLLpA_JvP6GLu{9bbn(ZJeG7ie5Nj z_lxe)lu9n}YNxbHQ1u~Q3I;PvgA9eK?kAxOWi$TIN83tFT;wc@$)t#mioyMKpCoHaX$1N- z^ToB3p;rynv66i5g!cyntuu|M^QE^%E8XE#mZnqwIXMtsNy8lyyY8bnh2G38j(xy1 zG-}4oUz{^1Vs_ZQ5ev2Q7vJccy^7NPDF5xj?`pHD1>Y)pt8aGJ?27mAytVEZ)$gDC zX7xL_W}SWS@Q-%Rdh~;jzWB!KZ@2!m>aE6i&&+OmOM7PaCaU}UE#E8nw>AH;a`vrR zTYs=~mUiX$vZycKiT&}4AN9_{+xNeX7w!4+`G2T}f!D!>{qOXCuk^j0vp!ly<$iGg zEiL|NdH=@jm9uaD(?}Z1MrL%(pz_?0L$hxED<`A#%7rQarh3sO1yCHTFchrAfGetB zK_!2_6sf^b2OJubCoNQf?Kta(3lnODmpP-)pDrDPrD>!Xv3|Bkb|oi_jQidzo&RJm zv-fnvm;={hvxw<6D7xrwx*S%S18-q+rxJ?%Yp3MWg|A@|_P(iCrzJuo^%3X7F(eh^ zh?tR0{|N44N6g25oxgg~!~I7iyXnXN8Pa5ag<88XB;K==P7VPzd0^e>P{AqtHTqbb z3JxH!c)z_4vV*`<9;hLA;TOr41#N;4U zM`HclSR+I~!`8vXUSMD!mmMa#VO|mDI*ShLyRs-J76Jp+r|qh`S$CEF?!8b$ggwWH z=zGdCFHoA9u{`9uV4mBe@j5kD@1@h-pp$l3ogu*Uy8{oax61bVW4y7!x!>JxI~ePb z!R~yr&+X{Wup`5Q0ijxb0-VfO7M6T%gQ8nv5Me>e62mizyh?_4s80I-x*zlt$Lafb z+<1cU^y((_L^7A$kK@*L5|-Zfa;W~z1r@4=6Z^ssEBnF|BX#Nm{<_z$kR+cRdcGis zze(OVy=ThR=*AD4zaLM1dWgnDI`uK;q3pD*(e1F(ZpH?YH5!96I+OD3yZAzK4h*5ACJ@?3G2qzj1ST>laZK~+0hSDqDRXWJtrzWyJQ&)mI!KzF zJ5*)LI?q%5%f8Lg2LdpSQ!VcV`v16zyPzio_sS3GOOZ>VMjA=S2IxDU{o!@fH<1C= zDUb@uk;RG!+!jOQ@eWcgvMEdMXzYw7cW1PK8oK}7?3S(KA^4O|_+PSOJjplt-ZxN8 z3W6ji7ryc4=xoh`sDhBA%H6V?{)vJVUcI7Pu)+PBEJ1Latb45v9_iDNBD3>jI`r3b ze_)^7Y2SD353D7J(P0zMM^exJcZ3lh>8gLtCp%sm>vF~feWZ>17%WlA$lr^(P zb0@qpC{Ok%Si8b4U%Zglsuu_-;6sUbWl$v_OEU57Cnr#;PY)MxDE{E!BTAEmlqclgr2K9sBfKSUO*c)7+c$Eu&@mT~q20$bN1s zlmpwGD<(GQ48=gV?FlNTlq`$Zsc}uw#)>!iC##MC|86zZE1|r@cHTkfW6XK}fq?7N zwOI3gOOY$@1g7~Gc`W#~sKyKRFG%FuN#D}bBx}zzy(%Q|$<`}zR!TA^Xm@&Iz2<4j zQK-tiz+3m47OBRtWTsmFNL8=Yvz22u5__rd1t-JOw@^_mY!jiVBd3V6RT%tw5gP3q z>mls_LE2{XHldd!P{YzxbYF94avmV$3(aC^tga$OkW);XSn1j(YFF*zUzUBP%HgNV zN9&(hKLZ?HdoBZy^4cSRyy-NW=0XRyI4sSDae^L2s=SlVm+qSy8wE8927>MF9%aC1 zyM)63#u6Buf_9^vY)-c~f4^nGoHme!oe!nIA?{CuscldUxDEixXRH{=J@a^nATW^_ zp66TlbJTv&saqv`A}^2*I|zbU__vM39$Og3=7%`>1dbd3_rEEMH7`hUSpSu!_@s-j z6s=_RJhkOeNUgo(c`@Q*L_6~p2v22At&`HA`<f7cb z2`st;opeqJ))r#`>ITFaAv6Df(xWC*#p$beT({GEjjQ(8#u!<^hp)a$u8t=QO}g~@ zD5jSpx2YJ^ZbQ{JT*tK?uo=PX;+-?L(%WgwdZ1jE7FTpUZxpz}eC*I|7}JJbOtK2p zPNebVT4;-0AKA_jT;*@{=z^h3!*_b3s*?_6k5OD338Nt}zR^R^Q{KR9g7&&|o|i;A zdGSdEJPqe6=fxm4{P+@3>02^-LpkZX#T;XL|nq#Ea zUoQl;+O=M=z!e6DD6oeA7+{^6b9{Cwd|Bk&r3vM`&JJ8<7B@SG2;SUEy5 z2Pv{2=@3xl2l0+Agc(wu=3}fP&`ZB}fc`=`eaH{}+f_a_Q}nHc7#7Wiz=-w=NrpQi zRIBe1%aZ}{FpdHnq{sw8Ivp#17;sdc3Bru?1B->)c8~3?K zNme13USViIX;5^i@_gfi4g&2wT9LUQe}uI4$Q`PSDzT>s?&#)YB;T}k_i&mI$Bkls zwDIW8-x!U~^%<*PBiAkEnjN_AX|c)De4k?aDAG&C+;!O}$q;8S4+U%anVLF54X>I0 zoUZV0_h<}dJC54J5593p_Vm4}@+0Ey9-HJ?BZVViwG%ei4=VeE(|w9038Y6c=!EYM z>)&OI4aW`=e22%nq8oWVbRSfeKRx;hk~VSZ-Y0jQrWZsVR%3NUnh&0>6dk ziW4N4H{9bN-Gjo>f<5o@Vb2izgR*At58g5FmVdhA%z?4T!U?o>cJa^m{maI8%iphl zJJY3y-a5Pb?Y>#f(Z`{u9{bSByn0nvqcQPV=2am)KXaphnb+`{9QNhe_U%X@V%xvK z@@VW32|HDFT4^*->63o*dy;I)JUOm;C8Z{QQZ~f^K}|XplM}KZBw!7i(C2`2@|0+` z>mI-5kl5&^YrGzLrb07o#boSqFqF6;Q=v}1-1UHFF9^ok8!RKiiQXbuo?e?BWyi_g zw7FdfRHntKzUylafU{_UC{@B2Yh^g_j=kVD{Yrcb#yu{ru2*zA;m zdeKtVcIo-R6itR~3%#18X!3nCcp0*Jyd0nAXgs$wvdMFWphH;^)FW8-ZVtaU{41B{ z=tK0-eKnpSp*(4cc+?X&;+c7l!5-}tvCoW!G8D(P=)F~y_sru#Tq+qHH>v?fXZVIK zkz(R0vYd*+)apU!PWcZ19saGtVj%_=#w-(Mf;y|9ZF0xhd|2A(EPl}wO*`Lj1kV#& zLi!`W;MMUryB3A&)N}60>qT|^TG6N|&$-WnRWofcSlkC6#U0#?*f`zebuW6C>sJ4w z(48|*%KH3PzNU>0YZIYbW^Xv^Vd_9WZZkg@H#*E*+~;T0i~W|W(o`J~UIfKBj92Q^ z8P18ppgJ{oZnJv8HQlSv&-N{t9MJO2MQ#+puy`vcw7fF-vl~-q8y(rOkIuA_3I`7E zwV13GCnyG*iw;pSPk2tDprM~2aFBN8u~Rkhop#ZKN{sAv)3-^haEDj5<^V*Y*D*OX z1nclHzP;>`PG@^)4d<}HJBzyX?ZNl7AAd9umBc?k^Pvm}jpuE4{Yn4<>?EHO`I)H= zZYRi{=`~)Doj&>KymFI#^Q0WVqR(N_Q*hXNzIN@pBWVe!; zQARxZE?B;m?0&&EMrTbhc9>!gP^6ZM*~V{`#06y2*_vB^2Sb~cD2oW8V3fpc_b`Mi z^TjRl0T+X`b7^=#IpK=wag@kJ6}(DGH616``NWFvD$_!=i9jQEj5N(Sp{~~V$)4xy z$7)n}s2hXUMzzS11g25~;b&xg-l$0cJ9osr;5BZT8fGBu4XRVy!P&g4 zjf>Okxb8e#b6&5iLfRO5o9>~jz%BZ?T~!S<4+{d}&FAo@O_WI<%FaRH@PhX-`ug+{ zLxI!$aH`WBfqCKf#|*nV@tyxEd3DIr{I+5aS^v`X4D2+qL0c#WiZ3=%FG^+qdaIxtTB;s5r1CVS!4T&$%hh;d2rNov^y(cz1gz3%4nAsI}5!f1Nsa zYK7oV*e-cQz1qO``a=oe&O1eU{uATmLb`#(VEq zXQuFsR^|6c`VNx?mi!k7j;8>(>97R&8j4Av$Vw{as3ygExoVd}i>g$uk|ZDSjsx}38d*yk$B@O6K!bGVUWQYo?y zcwW?v`ZC0PY08u?O*iP;8Y|397ui={=f42;A8|PG@O)M`8jk zjbrO$N#&Eg%#}m9YV>66 z$DrV^!|Nb;ONCQU@pVvUln(k1dBQy57k2blt)732u95wGBW|A2f=z`h%{z^rE?;us zEes3iMk!qZD@RoXYM3p1wPI$z__k=QErk9dw+1G5v<9q0v%}=T&*n`r_es}MMEJ*; z|NL#t@BZ|`ujl+xwuoXDP$brVC}#b;O-~ROZi8GJ>#gUBwMQZ<)K{nGOZBYO_!Zn|@1BWt@42pF8UHodzF}f1kU-C31ysl_GJ)dd^cI6C6 z&&>0y*5EK#ROe@&66XqQpmMHWkt;-6JzH5Qj$j*xJC?UL0^slEx?Sdl+AoRA%2L8z zEjmQEiLj;WZtz`T%fYIIZaRy;=20Qcm{t)4#FNU~v^GA-@Zr9qLwz8iM^PuxVm{E2 zE7Ydb&>+@`*$Mw0xa9cbamh*DM!Q~&0GeTqG;=lu-5y$CY=D<5? z5cnO2@KqGEf+9<)7^Eo#Y2UTprVRdVU*4u}2zo3*gAbfU9V$GaQ&$M9 zyjz1RpmYw4QQ`$%cI0fZX7*Ve5`EHph<)V8A%lfNRgax|6gr=kAqj#jlB=pV8X=hR|Ji#NxTelD zf80|%A$c+6Mj$5z1c@LQD}rGJJF%Ue>2z*8J3F1-|6KOh-5uHekGnhb>+EDYvoo`U zR8$lY6uf{25Fvt!0wSV>TScr|1q+IzwITs8#Ugm2sPKE9B(x-QGzStkI@^9)&N+F{ z8#wRxJkRq!m+upGX7Ut5Q$*8KfHwBg~J{MS*t9=n&vUZmJPp@RmpBPDcc>F=h}M#P03cr0y1js>oiX9o2_>h?#z zx8RvcvYLgCO(7G{IlFyXbkhT#}T9bOzH3Tp>n!mlu*Fw$OTdnV#Mrpgk&u z9kPFuKWH7VW}dd)Jzw&^S|5kS{)PZUz)k;=CA|yNcsKnU7$-^=S6zK}2{Ms4FaO-X z{+(?{RW7EF3#Yp6v)cFMQ%Z;v7YstGd!-?$d(ats~`qK&h9L3NwU zwMN|dZMKZf%k0A`dEdC`*HirK4^a$dG4K%zj=M=mcbYDFKIMj$_|Z4hv+f z*Pl3dSlQ;fK!Dib1( zgqHF9{m+0i^nh=+aPvIOtX`|g@j-sD?2ufbJFE=eB)MxEGvans(9Ofq#{$y?SxkD| z%HXqf)8ftZ5(Vo$P7;T$j$LBq38lh*!RGNz`jnGcaovr;_y6YK!fczxxU3_erTxnw zld8ln0nJMXg=-fB&raAdGa${1y5^e*!Ph%gOQt87FLw`hw zCf@)@OKOz1L9ll5BkxWATVHOA&W9!iqi!9lET$QH85z06K^+Zu_7?#^u^;`M;69@$ z`?tH7TTI`Iz|#uS>%xWbNmk@?4=Ckbirk^%p-{AI>2-5^_+4Hl{7cZESEtUW3n6a> zatqco!+&)NO!W*t=H`Pn9$FQARG7(VQRtOdv#5~P=7!*|;Gmxl9DS|cgZxzG zZTU%(Kj)%Xk2~NnNw5$0ei(zmZN}I*+*2TU zzi2W@RR&QOb1`SwqF6)<#lb zSWo0tC=pFh*1xUjum?W2_e{rmYkT!uw)}%!>}JV&=@sfLbl;pvi(X028rVu9-@Hr3)La zEculUf!)k?`7w7>8`X>a#<^=88=$av0qwz$y>mHooV)Pnzwx$+iyt;@KTkH!ggDW7 zH)#Q-%%jM5l;GHXu8&UfXpcK@YFG&$I=nS9)5>kzbun<(YL&@p&}Tec)EqD40fU& zV)xolvvlJo*!u4nKlia{na)?vze6&)F=)83Z=~ExS?;8iAW(Evd>@k$+o8#Rz45Qs zy{@NElN=vl^%B-WyZBssn`%ApigXBq1v_8eN`}G1+osv{oeO`L@DJO5)Vs7@<3u=_ zWx-@Y_XY;U28}?``4({@5<-?^S zrp~tyh}5SmOM8;|a5g+<7=Her|1j4gM(+PizKZOaN$RcMzj8_meycrWt;2w9z&(rE z?5|TN`L&v}mLDigjskQ=BrJIlf+eBGXlt{?cs+WI@J5SFuzKhW z)@!K)b*wo%$*1jffzL_AyRJj(s)qd5wwZ^^S}81DM$#fpz0p*y!u4LQ*hLXDu`m){ zz&G$QOQqcRcF4XZoyt+Sv5t9LmB<;WJpqy=S5gP1KK0B8?pWfSBSIbYkp4I}-MveQ z?8xAB3pXf{qEt-;8VP9bTz-22+#-Fp#J(W-H-2Cdl*oObOcu`ze z#}bxc;Rot{?r4M6MsAnf4JcR$!5{3*iU*%7kTDMF1vf;e3+{R7B5`w6B3d&&WI&U{ z$W&I}zI)w&MSsV3r~D)h*z>m^E*5n%*A_O@g`xtlF5!{TKBkIaq`tAB!0X0>X6Sx- zm0a^hFH0wbEfiMsPLl#J?Bu)7D_QO&G4bVsOb!fOvf94Mz zF-p{_$#rBqH>1Rbef3AIjFP>SvV;QOIp9#it@~K#`S%@ z27FC7g3$sw^PhTRnnWEr`!(~N*w7BrB8&m%eqq_N4Un@o3Gmqmo(VJwr%T)s#txo| za*`_M&MSzq0P2TTyYooJbC&SfY6Y_QDJ6s?49KDXfK-UlNsBoIxstpnZ5#CC?%}@| zQXlN2|{iHZ}1(i2z>^~rGM3N?e2G}oayss_HgFB&=ZteqUA*NHk4~dUm^iJ3{ zXHgUt2#kv2OjjWgjZS2TNveCyh4#Nj! zzKwGSNZlOZdB^$A$%&a~ zH7BBroTclb$hJxL3CPF6g=IjOdY$)4^h1JNyXJW(h^aXdgtf7Rb|mBQ z6Tj^f7dSrFC+>IF1@@PF0=40~M(47VhreyRpJMT8VDgbE8q)$j)3L8@(!9!0ZJoGO zY~Ippk4ELqnv`iU$u;_4XiW0K*>~H2tk(alEulXbubu0Ho{jz=YxdD3F;O@pbMuE0 z2jf1B*cf*(ZX=K^oR2=D9%Sx#Od$O2d!pP&xhllZfA@5>Z4)7vjW4sb1zG{z`lZnG z7O9=q5?xT9`&n^|yx;fI{5uX~)VF)Z?OkZQCv?z2PHK)y4@~y>w#C+{@Okt$Y363v zx^Se(Xk{{;r<83JX~ht8!t9)=58NM+FOS+Z9q@0CDmAJ?+yh5lWKGy<@Af!6-V(ck zj2*YbIn=tt4EGdA{YHQ>cLczQTlDkdUPU@N&npHVUGwaxy^Er7WNH-B;65N`8j^ZJ zX3*`BRv6WyI7?sm#!uShz*Gij9e(X`TSBfAZd-MGamd9aWf{u}p(ii$R)J zDWt@O9XZFXuyUAE0@d~bDjpp;C6azTBOkEg8p=R8rU;1 zgYP}E1H3-wEWSQ*5&s4x;ND-51qRlap8zGtYLAtIe)m>6#uG{;*bu&A{xJwmb6a)R zrEokPSig$X2b{#xjM2@P18jLWp7ad6aKZu$7e{u;9U$50rL&@tAy2y(?2~k8GoeK; z4D=Z5REIn}6ssZ=MJ3_dQ@pDSyWJ0nv9(pAsAsWuw^XOz=3W}z$7peIA#Enp9%Jv; zn(wWML7y`~wwrnQ)r!E!euI1ySVt6y;hpN*$P~Z*-gg5k0t@LTp7usmD}*CHgp9ns z(G7uSx3y_+-aa>7fFRGCb0`9XT#2I9qFanMdDal=VHyS3eD!o8y^i^Gp>_?g0^f)t z_|sbDMRYo;jU4vuR2+?Q*&ZS6X#CHEzt8-?7O}M^IW3O#P2?_gVM}?P6${K`N;yoC z2UI+E70#vi#BAYJhvu8yeMZ_CqCFAY2ORI8%W467H|9S8kyQ>s3mjxhV_vmrf9Qx? z_aY!*al6195a)*UDv-Ubk<^Ecx}{0XGa<<)-c}LQ&6sX&m7|9S!yqW-u&~+)<<|VB zm_pioM?HNjqD5g8whFZ;6t%$KawPZ|LGt((AvQk7M6P3j&Hu*N6qk`>o`oz2FhQLeEkE=?S#lNF81?FkI zUh1>L9dkTJgC7U)0?>T|O%U!VtK+7odL6HEzc9ZzX;|CAKMU91WGqP9oDp_`Y~qF_ z7nX5AHapH6T0|*f#HWd=c1AY9kXBrPv0MR{wNrA0$jFOGza-c-&q8sNn z&b=(YEKZ9|hE7}#v%zdv!g4+Zps>v7n9HK#Ua4)JB}+NuFuzg?X)T~Xih0K%SGPr& zAUYDOQ(c!<(77Q-$icld|E_p_5GJ~GXzFHl@Tydgqy`UUY1ctV*Gs&qy=%|Zg#atU_mC@7ha4F04*ixIJ!pmxv%kqf>_E*jlV;X;V2@d`-MCcCGuww>tGI zyHGpj+gonin*0R`L{wXu9=j7K3dY}-$)t+8_3zX3|_AEB<%U1oim-UZs2WmFV-05T=Q*E>OCHp+5?O2zlV!Ld8` zaBN$&ZgFMc)YQThU(6@o)NvQap5~-}vJ3d<%4)5TdHoGr+6R{neeSl>FWV_)Hbpj5 z@d)uhUSs}@!OvZE@=V?AEy{XDN#NRpmcoM*! z6cD-I#0|Kz*WdgtOe%G8to?xGEHI+ z%&7?gu74knxAWOgX6h;3!|B<(?xx_X+sSxh0Y$l6-iKs0H}9DXZ*>c;u)Ud5W>O@B zibt1uA9D-*G}wYkyI-KEkqp(s(qIrgg8l*133EFEZdP_vf-JY2kF4^VVOwVHvg(AT zaC#sAvbc)>8RP+Cg_8CpNU(>zQZdF8SITu>TI_sM;DrRZg{FSJ+7?BHcz;BxztaK0 zY1JKN=-2@OPGW6_U(VULY>Oma79%WWkJz?o97b+0)Tt6gdt!$9<_wrZQ5rBFp`%x8 z@W3Q-^GP=YeMX|SNYRNtc*ISH&w;?Y#oMB$M)Sh5Mw7+MIOsT~(3%@w>IOdhfo=06 zE)Emdbr}Nvco&dWibu?S(oEl9bU=Z;{yNpgh)w(} z^SeBN-9v#RchT3>pUXO8kyfineVzyME2IcwgO|mJV-F}QC7(!<4lhxb5ZEX^rONh6 z^U3jXn9yh5xdc1L7{>01e5( z^Y?muD1>?;UXvQncA2i7r88nbRNbSyc}P@+ZR2`{4igV-OZ`kxjS~su5838()OG#c zpVCv|mvZ4f4hz537>Pp~xem#j8AO{aIS>e}LAjC&@g<%%4|4Ga6@&Z(fhK;s5(&0g z!;BH6XdZRD4vBKvUOUunP|w?@avGD+WZ~j$=&anJkAK;sWrF|1u#9Y(C^gQ7O_76E zvSSyegq`pXD*o;pSESHfsL@nekTlZ}dHR)mitXO3W}bd!#BIG_i{e1wE~uM5{YpZ_ zQNPE6{UK*W4ZQp0iu69Y2G+_wMOxJQAZH4ZDFMP!!eWAX*p*V#%U^lTmf3-e0&-oy z5Mkkb7;Ey1_4yd&!(_#gY4w+kEs$Mk85=x}KMSt?-+bE+oi7L`So(PmdA3IGgaU!O z#o0VOZZq|>z)nLK#QnD)~CE$@uZ?Le+Q;G`VBWcs!nDNLy-AO5bbx(SB0H%}?vX>guAEx;jNEKcL7l zDqc_P)UaPxk0^jf*UR8>36dLgf*!;+3d?yH!(hi8ruV1->aq?^!lD%B9*kN&vq-wi z^U$0%GY6r<{+h4ByF=6e#`bt zXqQ~N6q5Q#XXH^e5~S6L)7|T1YlR=v6`|d{HZC1ut(Mm5aI*R~`%lWh@%wr|3!Hv- zW=1>7o=FZ_twRZ=)Kg?9gyc-5V1@7&pHZ*!)aC_bAx*)O@`x^Y1?v`%y6Nf9qP6Qp zrQ%$Pxf+A63_TRl&0LTbsVkt9=?TRp9#U&e=5@61FPidGw(n~=k773*CI~FMt}U?R z&y5%Vvi46%=Z#E>h3XbyO1uK3F-Gj9x0$y>R8MCE^SAa`XqseDQ45=E+%rL;#_mvq zM?TpgsEFy5VGC+3DM}Qr(3}bYGJL04%~M7_s4mWsw&RAke!DQ<7A{ZbNxHB?V%b#p zKyCswnf~v4ikhYGQQb>-N{!&q9rEv-T~CgAR`E|q^@-0&+JK7Uaqy<_I{^iAAApH- z@9TXbr~U7Eo+RsobV1#r#^~(<{dA?mDUh%`T>ccR#f}x8#1M8-_}7ZtKNeY3(K{8h zen(PVxPTUzGRIY9Y@?K05y$cOr8k9Fm?NPz!bj3J%d}MS|E{cTkIrB0x;l;ARDRTG$64f z==-_tL$=lUEJg2n`V5T`xm{2mbwOM+dk}i+?h$lox=FoE=e0xAy`*M#ZqRXHHM|I1 zTh-K^S;Tq)Qr!SVa}JA>1BO@g{PRS_wIV}aGw z5aYn*nc{jlZF$(&W1=_OWkYt~k%ym_TQt#$?|k=TvVq$k+lAu<`>piW4oV5`y=_!H z<~@8IeA^F_pUm977+wZuXYULNMvk;?^2V^ds7~gSs4_{S%*^$0U79x?y#*$Vb*Bj( z%f!)d|6Pb}biid}&@7RFF5%$9Vb5aUN;MJ+qLZbTK2J9aPfODx&jzG;=~ShPv)+wi z*l8e(xeAr&I#t$BQeLb1tLm@Sgr&^Fu?`e#PcGSH7at|ule0PQ3O7Ob<8K`K*1uZh z$+7R<4kF#$G0eFLR z=>$&<30}pw1d2Qmo`|i6)oM5E!8+59@J{921!{Veq@n^zB z&wBc()DY0bO9wco@zw{y@OoK2eCsImK-fzP__ZqJI?|>K`ot~jBIy|dWDe;9gL;(S z8U`i6F-`7p3(nUCZNFTWbiwfRe|j~{0=x3QvP7ci20IttnKxOfu_{Ul74l_NJeIc@ zg@%A0%`UGNep__5>;qArUiUDz+;GE1n z9B&;BT*JHRf6`B<>i2Jt>|V5szWC~CbrFBnqJC1ttMwY4A{DTe44R5~voYcvD)0`7 zKlusYf*iLHw?vXOlK|zwxE*g6rQAS~^;Ge%EI7e`+vQ<^2 z?(uZ0amz2#UXRL71Kq+e}a)gRU zzo@0vw?T=#6q!K=B{)Y*qmje1SgKv;RlamRzZYn1%tq^nP!|G8celB1crkB?Ag=|`BtrtrJ!vlxF@YEMC`otk1e*Ci@Y3c|d?i4oDrQ)nPcNj1pcwNplml0o+h3}NG zX~)f+;7hH0tACHcf|Bs9KR8ZSakE`q7yl4S9j{0?QOb=JNvGnmYQR*SVzn1HDptyY z*be%tNS8S;ZaTs8af8tW6FBKoNHBZjv{NQc+9=* zlvT}Mi@Z%3R(DQZ@L4Nh&k{dt4rhGMNm9J@UdngAsI$AWCq>Ht+rP23eu0mN6diHS>JL-n*skC*Pw)p4-Az%I;8VN35S?dxzvpmK@|5-9 zgp8-mh%hrB0M0lSYII9X8A8z*Bb9WZt>7_iPz! zT{eWs!a<8kjwb}Bu4u);kX|B6@ql9ZEC!9)>%2x#ASgK!MU4U4V&9cfgG|4$MULHJ zJCuF&8Va)`moInT{%2~SaJ-#5hAVbZOpqxtBX{XbbAD+NDu>?ol9O95yvIql3JE@> zltUD`2ZZd*L5Qh=4GIAQ?LEcjdE2~4Aba1;@81wx5rYl8dLWG)D@V4343ONQ9?#66 z%O3ldbSkywA-#SHqF#k|RZz{MTuGr+`yt5;f*vI9qi$9FZsxpC#T@Mx|1w3ne=cl; zN+ArS-S5$>KIh-$uRZ)q-u%n_{ZP@^#ychcM0zl;-@QxR1X0I~m=0)aT@r+wrz=5K zY7=yKi|&&GXef|NR{_J}0CX?Gj_8mjq^I@ayXcXy@`w*rM<66t1gRC?8*;=1pC zh%-Xf3NQgrGJ~#vhPn(1j zL}tvB{c*M}+uU=Nx649}mlc&Oxv^k9GeEa~{fczoQd2uy~abU)fypiQL{*=LW*##?1sQ*P%d34s)ka9_h;`2O5-&IV)ts7 z&7Dg;RG4S$19n-%hQ+IkwYIPt_g^svn47`u3wRWkE6Gu5+f-PZe*>EJO{cQrByXIq zoWA@5b~QD@Ui_u-El}oXkxpYosm~iSunoOALAH;+9El0F#lb6s_XGn&ROpVNgAw-~ zS*-HqgJwPAfDG0-obdrSWL*2jhkI?yMxL~ET{w2gQXF#4*C@Q;vlp6&wa1l4*92n~ z03KnJ#7I!ms}t z3y-lP?QRUVHmnajPj0H6TVp!eXKz}F@ej+v$L!C30&(uTIqpAA{I`;+WWc%ZU&fXJ z_g+Y?cwJO3JtR9D`ab^wb5xciIZG$aJQCVU*N3f|>2zsg4Mh9#ZGE4!&)!{r_wQe^ z7$DW*zgtaqbK4%dumS?YpK*KVBa{-TXezOq2s*jSuzM;fm+(P8lFgtAo9El0N+kE8 z^T`mw9W?Y<6>80g$JN&f?o+au03=1_O8RMS{+uo$-dPdTz7*Q8(ippf z(#A61TwrNfLpca#_M^v2F+JtPF}M;KBm8VjJ6zU$WvMyAtuRuo7(8}|?va_AUIHS_ z4u0on9@wsf-79T-qm$1$>A{!&ZqBmZwr*N3DRD0Bkz;Y-nm0h5vLm5od`w~ir~hg} zt++AFK@xz&^4bX@4l_Dt=*zoPHn~|;&7RMH@duLd!l)XZm8wanl&KV1Yc7P>(*=A` z3fSWAGK7MrU!vgR*#!y5tTLxr%5|YXwS|Ps>JOGhE_&?|Bv&J^ajU!;IF60c)9VQP z;(^<<$X^Ve#sf~OW5&NSZhyATrgm9Ku;xdPWk@e?5MVki_FL#$*hl{nlIxeHN&4i= z5<`puSxmFQ2gbiC*OwI>9|bLACd5h6yqn|Vk-P-rvNosoEW zU~Xf;wBW*KjqQSo%{|j`$_|s*^q=!=nckj+gbQ1gEbMPcmU1?3SgggGu8f%C;L_-1 zxvAz^i|ymFun+07yBW+(n=Z4$E_hgrk?BCk40d>Y=g^ir&*qJEVR*3cMq-kD7K621 zDIPj?qG;F?`Ay8*5a5bjcyHmT8w!%uqFW49v?J$I7Sk#>bChh4-AnZJBTt>GlGF>= zI86gh0xr`69(K4qCf}=g-V4fii%u$Ymh{P!<&ASM1CK%7;!atMxI7Mn@aMcgSlrDN z&MpZ`mSdZ-4;C9#H=DL=oF1T*z(Y_<#bc70sUvP9v<1Uv&yNMS7B=we)hFpT%_jdmuL`xFz#wbe;&YaVu+40)DpT7Rfc1_RxdbMMX zmr26LT9}y6910~)u%Y}eN8OrZfz8Ty{8LCSHxGylM_~+BAlpYN_fVu5^7P-l3>o@B zkmPqv)D2|%g`yOG74HOHJU>UZS#nEyTzFGm64V%)2U+nQK=$4)C=LWZeG?(Z5J`oc zqU?}dTF=WQn}Ux87V(RnlvV8jd^r?XW0MJ^N7vY9%Rgz{x^NJNC4+v984|77g3!cy z;DYBT$CmNdE=IH*6NocZ5<_CeC#ePOjz71 z8R`~cTBMmivfte-9(K}aEy;(M!WvXNAn_5RaALJC{jB9%_5WhK zD|T5o8p{s2P+c5s1hy`H_-f#3)FQ9NK&(+V>Nb|W+65mT4#7Or>DvF==}I4OcQ|xe z*(YNc6cZV+>uKZIMeUfeB|T72hhyBEPXlGTK%-_SQHH z?AvgtQ}ys`>HDNsTmdOet@6guV$mL6uk&-rV1YP_1SeRf6QhI^LS}@2#Vg=1OTA=8 zwJW=XJ&Tjn_aLB?xp>v00%;nr9(tZ;%bFxLaR)uH%(Fn+&HG>Va`U&^;aH3O*sNFN zk%KNQKiaJ1$1zIzK1J%W6<9J9+#>Pru-E+s`ExMG)Rb^qCC7_8)!i5a#AxvHyV%N9 zPj8)jU49qXE9NIN`}kF&cMnBmiCX=y)7tyIR{5}3)qJCDdu&nEYTjumf9PXwF)i|D zy3Mae_U>u_9$}a8p2uL!8ft}Tr{_^Jb~b4&O-js+S4_`1o>+)H`zkjvQ}d&!b?;b! zl=q-{Cb{jxKuWPHxEZ084=FN)gj*QnD-SsZgroJ4KUOQg&lCmii5cdDvm|Vz|A9q! zfMn}IY(0I6mp=zeUSfOIh%hX7ffyiYtuF7ii<`m*`Doa*K-$=1}8~%A7PO6%NTOWXRWuXN*&9^Fm(m*Gy9E`!kS3~{MsFUcc7Ra z|G-WLIebRl_Rv~9b=hhgZUFq-xcoQVEdXr(O8tM6mE3$XE*v@7VI}u6DP;yl(y(~Y zQa)J5>r`S03E6J72jfml{|DFuIY_Z*iyjA#_;y+04O;{G1=(G*nA5|5jgGo5k2PU; zWr<$v<=MQWPy^5)$mbslD1{?*6QeV3yKgC76qF!pmrq)Caat!12pQ98oaUeG;GZo1 zpxHL1-eq^#EHtqYR2Ki&9_mkbLpn4ik>wE`nj3Qc%k9D@S%;>YUL8^}MfPis&z8$l zbDVlg!#Qab>f9Um>*6dL#q-|rEK=pdCQZAQj%cQoM=8>PI>O8xiiAvBbkcUp%0tqL zcCXL-{4OA^N$}TJ#9$Mbe9u9!CAQ8Tg>W<0ZDo)?@kz2bx>L5!OS_R-Au6KbHf(WO zB0*XEb>6xsE;ewIiI3MXv2u-OV+k>l!i)>Fz9iX`K9y>2mh z_ONj~rmi5d=wzkCl-9-`L`>-mCVBIZ{(7O#wn2}}+9ND2cd*3{vI`W$&KKSNNLh{R zZlb8}rE7($%G>g-b98 zaGIbN+JiV+c-VslyVc4h*L|YfIEjK8Z_PjPE1pG`{OW`I$H{x#w!ki|X7a6M$wo?< zPLWh9J{iOalA9-rfWHEJC2UqUDYUr~OqQE&p<*|NCR%7GUMt)D(ivM;r6*GlTsVr% z!ku)@l#r4~57X=7&_CNn1_TGB^@^nMO3Cn&^SnY)f!FYo9w#=yePYqHTnRp%3yD(Q`WSU9coXe3NekeV_`QB*^qHkUr3sHL%QN3Ns@7?973PbiGS`wNg)-#ov0wto7( zCEWJUQ^(arSYx@y0fOwfI=+X$D%yf8vHQ{jvj2rKTuxgduAWjtg5_Z~_e$tc| zi4{Y+bUEn@f##Av!_YR|NbeRsQ0#(GWc9A|Ep2VqRdVU;YzGe$GUb-U=vNDJAUrHd=PeoigO0&4p%~+krYgS5hlpAB0uu@AJEP zBMJ;vJ4r=N^tibpV4@kEbB$fR%=~-OA&q=HI-asj#y0D6hzo^tWNo2)5(U<_3Rx-%2nqw$2^) z)T@_Qk{k1{`4*}l(&&pzns-9BO;zST;`SaXb9YFO4to&6Vt%q;b^vhLHOd63xHJ5Q z?mDq)>F6#{3QZ zZ+k#_c(ZRajcIGxe?B9o1=<*6mt*vtBN07DQM!b8JnQD%3QXe}JhD8D3vRu77V;c* zs$_YW7v71~y2TGVN##X;FQB$ZV-!}$?SKS`O_24wLuT-}4(WP3 zWYc8>aTaP05Lot1G4t;?r`l5WJn0U0VUw4Ic>~!KI%O5)ZjAQhz>fpXyl>OB*yA6( zT?;?{hG{xx*JhG}A0^GRhzvK0x{4IOFd_po#K-NADk)_-MfO3jWqQ5e7O169Bs*bK z{Z7&Mu1J5~p?Oa*;JJ&g6pt{M!Ujl(CP8#y?qz1JVt_99MHbz?q?;6iyXFC8g?C1G zc%G%(<;bVorUCORYBTT7tn{x|&;w9v?O5E|g(!AQX#6J*#aj938Na_Q-0YEUh{;gy zq7AWqUK?WDG`jh<;vUZq%`Qzz@QC1$)DYX^Spa5T&*DN+yQWKc1)9K=LVf_s0VmmaSNV|l?U;rkCcQd8pLXl%pGsyy`@n{)4b|`9Q%*ace-|T0R zH$OWwqn%{GF!BbdX~)T%5=sefy`5BiF&M*W^q=P)2hQ573lE13!#?d2!IH?#0JwYjJOei8Gj#329T9R0oNcUe#XQLPF!7=VT7KVPi7iFu8k} z<=_4B^PZuL#M3~{4`~a#yt0@sVU=u8R28#v7BFVWP7>`-vfir{deL=xZJedOE(fj_ zc>@oF*qC#B@zq-KUNA<|;HtI@wmWH(Uhj?G$PslnCIL3`XGo=DM4`=Lj52I$se`7J zO@dLkv2-dFKXn92vnQ{5xM5>VquBj(a=?JQ*v;JuoHs8OE&0JW6c&|~JNw^%MY6c5 zBo~%46;>*#kWxZ9QSMl=+bwVO2Di%l7N(PxQ4lQH6jd73BF^vu0pwLLGgYjaKWU1| z5tWBtGxd7HeovD}j6F*_1ytEH{iMfy``?ybx2>jj+2l+X`i-pEM__)pV14{78_UDb z+jisNvgL7M7k+^iV6rIX28yhw;_u8Vi%TSW{uODP=ELt?h0NY|O&;A4QxbfVPMb!W z4)ZG5jf?S%HVwxKZ~knRZ7s9Q8Xqjx$s1>t(Z|Tg{$=#%en$nncv(?tON_!6&o=)) z`V-$P(ge}A(B`E(=&fPv!H>32FapXArz{I2v+s4FTKSA1Hcha4a zV_!Sr&!=pKCIPE20XF3$c7(E#Ww{~!bfdtD;IUpQ&JNlgeQT~Wqil|nYL@FE{`JHH$xURcJ zi%c|?L$7fXA=J{Wh#k*vi|oQX8J4!l$Qzar(JI#Sig`n%i#E!NdD+SqIhL4R^G%$k z=T(wsx{NsnZhPSL65O2&1!{`C>1Z!EIp8=HBNI->4xEl}{yEf^{FaNJabd4K3svqg zBmyLeZhE%IC5R3xu&M{m)e@1x1G!N98s?f~Q+?W7GJ*&%xVdVa|~%uy?bokNc39lGL~n*>6}ZguKUP6(kc8rJ^%FD)YI(A!>ea?6E9P_k9Q%0o&y zM3H+`{QXyq0Udq@zPXsVQ+5V)mKNJk6{@i!#Y`s&$&B>QR~t+f$tD_7Q_C_Fsji}e z?o|}XWp>aYVYJ(Lu4SfkZuqR#6p5XZDh7QTfRN0AnLGsZ}cPYy)er3dqAah zaEU|@&EdP5TDsr8!T*vtBjzEE*YS$^i@SvTLr;5bh)VESzPw)8Orzg$kV#d(AayEl z&Qd&#vY=1#d+m0zk6ZB3g-z%Wt<+^5rG#i?H5GpZP+(#$Ojv{~jfgYb%h+#XoBIKv zKm^YpI(kP#hrNLYncv|z*7e(jYa}_mEzjro$w%F4!useUTC4jhawpzi2zlSK%1|w` zKRjMo#zWHGi?3ofi~}sPNk}`pbm3G*VZFyMA3X0G4io2)BmsBfhawCCm_YZQ>Sj=f zW*;A0R^hd+JY#@9JlkiVq97{8nVGiZ7Gx4@KJISk9dgrfM~)A;O@(^fh2v2y)Z1uY z>r{!NVkzjsoTx#?ao%=F=fP}U2S@^r>uNWMILi^h5k*F9&*JInxoKGtyVrZVJilYV7bn-;<=<#@wt4+7o6NwHslUQ! zKcuQ0552k&%EO|X_({rYNx!cl05h`Bft&Zj>m5s{H+_W@Hm3T5bGVN|{5EEvDRk@bne%vmk%_I#BE)FN4+?Fhp=yqYvp;nztmqKoE9W zIy~!+)B1v2uhS#|Vm*PA&Lm}P-?2?;VaXi9ZQemYjJdR_v`9gWIjf1HTnX5DddWU` z;%O2y{}#|HAsJwss$bZpx~Xzz(unD?Oyc31-$5rm)UjzfNp{zL)iOqYpF9;lC>QpV zu<$`)Mx3dr5W6qxXRicr$FxNh$H0R>?E&`q@zZsDE3I80U{NdT)Z{v{-G%+ON30aZ zUP=k^Wjz(&;$9%_l;zU75!xhynRor5AEr%ei_}Se*|Et$6=d)%6`OM?wHSZ=1BM^9 z+l1|LC187`34nH;)?#O=gMKBVz0qf!piAsQzy$V$?Ini;gPUl}@i;$=SnyKrmiHl9 z?ZS8|w8G0~N|{NK3@Uz*f7UY*GHJCK$-5QN0$il;^R;@~64*1HJ|nE>&{g1u42OR8 zQzzqwjQj4JeyFk_Lv!-&8j?Q|H>nG6TI#IOQbs8unPE2-e;KwY4StXP5(W1cj6d$s zBnnQ_1^hiRZE|d4_}K4AD1NR};b@F0Cb~cN%k~5|LSfyU^SrdMoS+K&y$C~OhRHRE}2)sN2jay15)Oj1-3V)GfacoGFkXc4ixs^^|g*aUb3w>by*)8OChWg zmc3KlBOuX%Zgai#{G};&(&4JG^M{DE7Tz;?%@_kjYfO zN-(iukIbmR0nn=o8~mCXSpcO*uZjqBs0Tht@xWGi%gPjI$TBD++AX@PI46T@Blwl< zu#@b@8;eGoWr+%`dnztRKi3*<%K}*P=^uUrmnbJ_fF9~4u= zlRWsXD^QYnH|A52YUR9o`Xcz~?g-WaANy%(qo8h1Ki#1@PJbGjH1lG_r7+wxmqUM` zRWqkSwsBbYaZ>-;%Qn>!+@#u-xx2Rh``G3wZv8xRgPRrO!hV<4R=M4uQ_A}k8N^^R z1U1R^=!)ouKx{N{6TsGM7NGZ=#qz!r54UyRV1QPTw zffJ^9_Nj<0W_Kt^3oPTf>MkKxaukBA_>Y4^+&P(Wid?4IKBu8NZqyCwT`+m0 zUXVEl+u?M|5=7c8s4&433~E-%Oa%bLUPusvomEOi$grz}YL>$xS8Jtl6QoW({-KpeQxv)*ml4I2> z$BbWdVTg%%7e%h+bem7={6=UDikZQqZdhDUB6_^=*;shyqC9)RWZ?AdS*JeYpUYlW zQ5Na**z56WGU7JDrqP9Au+2&$rBTXt6j?*X_blEzcXXi^yKL%I=c6w>A^pLghi}KS z+cS;B$FNJ0=;*J%^SsBVt}-YCN@JQ8$zY!5N=6iV9(L^ALru%OZnau=AON+1)jiX3 z!j6E6-v70K_^(yAH6biC#VBj@=jeIs{c@sKMwRhRxoW#Z_h3IlUX8D#T`@er)vH7-sOm&}3%NHMt%_o@RPnycsF&x@gEy zhEKs{#Y?+GOQ8T3vLu?q_P{3<#B{xzc}N?Dy$W-f9U6*+Cvfa5{sfr0z71cz7Iqvw z@nsw?Y&!FsGfOQn`d-tI_LG{KvbAD#`C->Hg4bzE}~Ah2|Gu?cw(U5!fcrJ&?S#MK-jsF)Y!&MbITI;ui-d zh|c>OJn*wTXdRysTNr}0PbtbPo|TaFg#U*Q-{MEoGT#!BKKwJWQC0-4QTMgwTm0UNTaO3^^%WcbBS?WT~jk@#s$+7SAjj~QzmWLjK$LC<7 z{|B1w+6xP-HJgSLPOxy@9mCC^$lkIoVC0ga=emCxWR8zlR^xin4||pQ9ta$Dy9#bq zOujLW=@(OhJxI#Ciw?gsX*tFnd-g)ytZ=x`KiNAz+xibd&pR_2(z~QVRpG@~4@G1_ zFQJMU@F?^0!nFEu?DvZ`uAQ6ltL1lV*-6J`O%h-(0k6rXFHJTOsf;?70fc=#Y)& zlh!TwyX~lY7zz6A%B97if~v+uK^k|yzpo9 zg`zJm)GURHYs=_AbZC+-hezP@eJIJsd;TyvIW#$RjiJo0pY}s)+(>qGy|L~;k38>j z?IddSYpba$Sh_!6V~#+#%nnV9T<@W$R=##)X@el2U-RbLWslzIdc8wa%)1XgAu2)# z$RPhDncO0ed#=wsoFpKgIqy^lumfTO#|w4+RIJ*zI*E(5;<|A!!cxoB8{HUi(|1sj z>|H1YelhQQMMDv9~>vECaOwuVKZZ= z)qZCarQAr7bSgfDU+jC6xycmL=ZL9nx_4o&1QP@H@Y2Kcqo$!c(F$A~z+!cuvrpK; zGWrj>%m1>P$z(%4Myh5_m`?S8RERq?J^Xf%F{f0!=v(oZ!Zy!?bVWEfIwXNm<2*_I7Y|)BohO5ZO)&245_e^A6aRPAZf8 zn{Qg=%CVnc`ji}qIjOwJA9ec!Oi)jEGZ@&%8?%@$;oX?Kp4txe z$Ra&`LJj-SerVGGckwI4#}&DB2N1Xws*8iWyjuBPLX7DTku^Z`H_UJ4wZlx;WNKA`d@sX|EaoG|h78#QuLZE?% zH3kdp^>m(lIqw+xfTT&z1|N()AKD8Fr<86A!y5O4v4b)8gU@a>Of--?awnR_@z*#Z zm-_yeuQvT<1&vX9jZw{^+%p6eLi^Q|%5J9JFT;O-sF_z`nm}Y)eojboJYZO7{PEHO zpBWaNq%NrX9eL~q%}mEf;!`Q*T8gZu;&+R3mYBSknAK(YAIeF#!y(1`w>tF$POqfv zreQ2hJAP`81rICV@lPSSFU;P@UYTEyu=G8Kh5KE?MVwZ^2zL_Ly;ocaD<;VYQGP z{|@}0eINb6R=c>Y`N5K8^r5OA`m^X%2@$no@W5h-rZy7( zVQ^b*$FC~ejNHUrX77Amutl2uYTwcIq~L{-CXh-!F1NaZQbO_kUMe12)ag_MbMHiS z^DcYfW(_<4T;k==8JwFfJULgVIu>XM&~6Ob5rm{5?ZVyuCWn8DH-6r$cpw>ZOPaSo z&^+$`0=SP4J*z$-2i*JJ9gIJ27-JPF@BtfICU^aM-&+;0Sx~d+-H*D6!G()TpY3jaF z`?IRQO?my`zuZ{bt{Duw6Smj8nb)EDl%LMa=3Vzs6s;wvR8>pb{dzo;m$Yf>p#ey{ zU$tkx=dnPXxy<*fxL24aF^@bW847IIG?E5Ey{woy4!jL*lXmJl4k0Y)nH)g4m7Y+) zpZUvww50kpYmMYvp>HFlOs7aH6@MlyCCpTRd{y49%8KY=v}L@ti@Rv_$2wSu zJaYx@#?Uioa_oTHKltw6Ut6RMQ^<(OXWVp?3!@^{igaU?Qa+-{=Ty9T*L$BlAiAKo z$UOq>3LoUxFrQo^`@ON&I8%+A=P|b-m_ldx-Kt@)KKgQGqVg_Xp+Tw#&~V~|9@y%k zP1Odq?>NPNZ{#b41Q#qS%8R<@YorV5P9|Ai35;Q*ZkXAN!%!urcn^{4IP7cLJzDN+exE^T0|)KdnWRJXxMDK!J)yk(H!gct+hKzZG#Wc7R#6 zXv8g7lEyR25}~~QwEt-+rCBH6;sqO3yAjADq?~p6pU%8jRGuvZYSA7+sn|q}d;#9lT>7GNjp!uNCNCLLVAr`uuOXQuHbql;@K1R+r@RuJ z_ByWnGQR%Dncnj)_<8MWaSAD!Nse3L=P;$LqR0Vo@d1YtvXJ!fO$`|e)n~$nnf;-O zqW2+12`V~|l7_iQNzYp2n6n9ZKyr-v>+A^Ff@?E0a{jKn-Y$M~a7~xO9JP zl3#(Bsf@bKFUfB(CYLUTw3VG|WH`|}z!q^1c&3aUZYHYhg6G`2e~-3JlzFa#+Pkcf zo=qt?QzVm$?-J%jU5zXR=DIu3clkKoN@HvJt4#mwB1mPx&n`z+C>p~YfKYp}#0|;z z&N}sR>|(I}Z}zTpvq+Xbpa0?yBw;3K`SJVVbV`{@k+oEOIyBlgr8ygE49sB8bICMy zbhmy2c0Fb5Tc+iV{V7injCS3<$N4_>Hd~w4Wxc^HR%)TT5K1`qhpt+b6O=8?bH7eH zWsRZOD*ld#w#=iwDKS_}C>$~KVYT0&qK3C0<|v9X zsBU^LJ3t?x_j>R3e%JKRvX0PAKtyvaRQt#q_ek4)!6ZYv#JaE!pkvft7Iz7=m_*Tl zZyIkortTqY$0P4G)M`;3IVC;{Nq*~shGgc+?s*Kx=);k>7ptpvsx?%byfN&NcjxQt z`9aHCW&o52}>cyMdhH0#;=Fp9MYnA56JfQo(F)gqd`yz8`p>Q ze(#;H!fF9$AToarDsrNn`@WC8w3B%6$8L4%Pi_n0vMfN#@AvPFw+N&Mk-t}vE*BO^ z30D5S0ZRE9MLwnCGf9spxYvBkfU64QPzH})#kPo2sG7eT05PXFC@V8_3Eq|??`kuh zI}3;5ZFm!=0E0Mb^=?tz0^+MI@X6E)`{{bwRqwixN+4k9WwAT#DMccTnh50Mmjkgq@b^Pna?QN*7 zfsAVl=*IX9UDjA);VL}nht$fjm&c&_O3xcYM{MfHKT`XS?E!>+GdXq?H$V(sU-_o3 zp5hXza$P?H#Ky<(3v($Y;AtxrzaywRu2WVudo9o8e?&(=l6ZrQsmtq7M7A18b6e$o z;#^4`cr1}r7s(Nw-uHM?b&n`Y91ETaX1nnp#!sGik{9$zA5rfGzPJ*R+1t3?7yZdy z;1)!0G+M^`@JF6lo>U@2|MN6)U)m!?Y!$`?Yvk}toP@|ram?*+P9@y!!a{^4+>QOx zO^wtfz(t6zrCf;?2?_NO2C*_59KJG7PG~oVp1jqmS2q54;<0%aTf8lA^W^BG}=$m8Hmf)dGQiTUnz5qS4wzY0YjuiN1ne|Dq8R?OI* zI~XhM5@yHO7Y5n}09@8aWpVZoGHJZ6%6-V@Yob@L1uh@$5p^2C1v%vOz!mj)VHt0_ zST}I(xB|h!>&-eHI}j#tL(iD;+kd=h>&AOQ>{S-0-#~1yI!)3?L+h82b@NvEuc6R# z(bH*?QMU@kX?}f3iRcXIAWUtQU^sw)z_Kf-ici6UrF?2$$Q{MqLUMRB^obl^j_<=hP&h;=)@loIgFPsJwD2jvg zqR#kfp@YSO5`qq^0_lhX?JL}%9hh6mTu>HB4@SU74g7-e3t;b>iM1nFaM}td2b@m` z&^2@J?)S6ku%Dfo(N3~on1mVVurn@=w1iSZrut5E&CSm6j{=ce`Gg=zP!pCd!*)jr zOR#>wL{b}>>Qe-1Sbg*<-qnRQVd;W;IwQ73(!e_dEkyP7e&1@KuLHpb|4h8r6Fz7z z7pD#uF5hPV*vIo5E>ESB#f42FmP(c)dWHJGYU#cOz3MtaKYhn@#4UlUrP~(wXx1nb zWo?UBf}F?!f@wT!jsY^ms7hWLT^YQO-yE}gc0$BuX02jPR6X6J`IuMCyBwM9l}Ykxx#1U%{xmUe|^R4I@NaX zJdX~|O<+Cz(>12zudqlQQL1cd70vaamW51lRYb3i_)c%WK7uve)T-H8jal&D*^8McJ zKDD81l|#_$w^d%hs7rVoxKKMZp9C8N&P5t!9U3%DYQ^b-PT8=-*L~tem;@Rop2)#h z+0ih8kxZ@p_03W*i&**R%VZhJaN*DrNQ`lINIs>6e4ZR6h)EYDy>_u9SmJSmw<0LTR9PMHA%W=a9?}s{0`JD;6D^J^q_-+phv?~|=;G*d9u%@%e6=+CeSvmBmP$d>DWeXE=lbK4QQu#wwrC5fsjbJ~0lz?2$YxgN!g8n(gY->izQ$1n-I-a$uR1tD>qwlltS?7VLa)_s@Spwz;rg zg8GJWdT9@(ga%$kn8y3Qsg2G6WAG?59jz90XsUQ8=;HY~s=fsoeEsY`daZ14a1sBZ zc#EtM8tIt1W~wFqzL;~hQl2165mW{D`#McqO%(u>7aUK&lU=M#WX9~?_?E^t+4spz za2HllESbqW)#$6s3@Y)^4oI`2FsqZzFZT4uIn__?g2`0BWD@tY!=(TI53kq;qq(?$ zU0C~+TFIL{O1YgP*;M>ia$1%ana8`ztmEAWpDc2&Rt7iIt7qm@?Qh@w{+gL>nl|s- z3p<1FF1P`SOKH4zjnm}H__eePJT|7v`1_r8{MGQ^zWtI#$*f!*(FCn)+3_@i*l}Tb z0*Dx=lGano6pAEM@eP5$D)ib_MN0gzY`kaXzp4NB{o3cMd^U z2qq&TXN-|vKB&02FpB|(-NnhXdKG8<3?4c!;H*>ZkZIAx=#*h5W3Bial(IT-w=(f! z*|Y36a=?5#@pKNm%I&R>|IM!xY}?-ysfGm;4g~bJxbr2aH6&zH?T}(trnL$uw7{&kdkf0=TG>0VoqWyP19Zt@c z%z3}(d7t-LzTczd#_Vpo^tGxWo7z{W4GH&+>;@6{EnITsum72&u$2wr(A%h&n?zY4 zmT6=!Lq;qywm0s&qC%IoEH zU$3oAG5?3QwJLn9h8sH@9F;9)f)4+B<XAg2j+NbWV+iD1`I{M(Y8^$^QVCdh zj}^7YHTPv%if$S@J$4FmnVY^P;xcN23TH>go=;uV<7*~ji`&9=fAK$RzhS{e_|jQF zBF+3lsBT;?YqUblX-f4WMOu-CC4)T!`QJp~D60gHvQ80lUk=g*^BXisx1u}ri^~fZ zE~3^7pl7H}R0-XIL&|-gI^1mYES!2xPzqX$1LPCM335815)4eM{F+xk15F^Q!UB3N zyK?dXfmO2Crs=v*K^E~k5Q_{~HNn?Px|HR9NQ_rE7bsvdpi5|@3h7BwB2$?AKtQBB zN>+w+(lsgrbSB!jZsKil#&myq?X(NAS5F@Jr6aSE`_8k3V>a43Zxv~uYyLOH-dQY2 z!RD=6+6YxBTf%P4u2QCwGmL%XYS%-4)=B)p@vOT!?|3LXxgy45btIk(^2x3j#_Alm zvO4vY3d)uCQwi{vLjzACbS<|5t8OQ3V-E<5Bo27r6t0hK3`>kCk=2ppDcv+S*_+qc z*kFtHsvtHOT$=po zg$y$fhA)e5RlM06ekjC1cSW@+S|g!ZJuXA=<_c;Bg_O&B_05P@BPG6TIhqL=Pq zQiUjCaUkTrY=xJThX3(UIS&Yr@9B&!Zr^5%fa*(({Ezb;DWY{_51FHg7S&ApVzETJ zmR>&-`=QTIPoHa|%R_bjubKjl>}~o?Siy93N03dy++(OisrgF(%z(GCd>RY74`>FR zEoJm*^c{xQqdPfwqqzRZU$Et#b6Xq3L5kNYTFEx~BV#1!IKR;jj;~&g<9>;447*3p zkXmVzELnvvEcOKz08DsL^9pD6s5{);EN~~1ZHxngt7%!*(O*`7egV@U1erKlVD$9 zZp2pYy|}Xxxv~cT)1)@?1W@akKXqttMl=bw(I@F+>XW_)!tN@aWjHt<@_2{d)y0zh z?8&#jS+&^90+hFPr~gP+@LM*yu{|lWg4IS!l|hkpRKiXHuA&9lkAXK+)u+yq)Xf*?-Sf3 zA3+=8gpM(el)dBZ+EIP-?z{}f%bu@(eT8k2BS$S^;n&R}(Zvz#zI;ix6U=#sb~~wH z>gf_elK95#2SUBzgt%2w9%mAr5S!@ks38yhR3&Q?^w0-G(}X9)1}`K#s&HL|%idA3 z_5MaaKlwbJvxnEHxUUuNsr+O?yk&j~S-iTG?0aG6m(y1BOADnsM3F`+p`KnvjIvd; z-h8ytuC0~aBiFQ7zuNiDiZ3mCUH|39Kdk)sqYL)@Ot<^XSXA94YzQg}-WTy89Jf}+ zEn#oJS`-GiCZz7ZpD2It==ztXNfovpD7+qRzfiCwu4ip

    L}Gyn4pUNI39+qf`d*C;!7Qr)(rCj%|! zMw7gNFUB>7pJ7tM4Rk*=9UcldvWMjB zd`XMMpq zkJEErZx&kcGIP`STF5dtUMCk=As~}dt*1yTm7tHWh)5D23efc`^2i?LRuwdZ1y)3W z+C)uYWvqc-H8BO#BO}0Ps7J1uh`+oXs+ZzQc67}gSr@-!JV>@pBF$FMe;=jVO@S^; z!eP3fDV6V#7-n7(7J2SvH$z$68TvRG@;F0(#FYD;@bA#piz@uCGEJmjxdyUvt&+6Z zoQSm8WSLkxCep#VKo;PN9Xbp4&bDT8#oPbak22!1B)A@|h=y!0eZD)=g zuMu)zm-qiP{Ittfzw^ny9yhLX;n4I%u301dw>}o?OfLKNvFC}A-7ZU+k|@?~l;4a3 z0v}+Mtc~oV|L!Z}xPknYzwFrm%E)r)H#KjPjcz>O?6IZddEJ z(K*uGXl(W>6Pyp%bpqjW2D>%bBp+@{-YfW2f>HT{GxtU4kOaJ4ZJ@!=0G1~6KTfr2 zqwSER_HFdNoVm>ni?Pzm>wd!a*t#ZsmOE~2U^pBZ6wkWpUp>zNB&GwTSUm)iix*_N zx|xMA8=7K@)h!LSYcb`i3I|yeTIt{P4xQ6&M3Va)gFv&a$I)2ZqZLBYSej*x7ibN)d1Vvm!H}A z;g>EK*|v~!^j@OnULRQ^-Y}yjOou{sA4>APE9MT6YiuLXo(;%#>w!AHmd4|sfgqU< zbaS=N(5JKc4}9Vy?~$0G@!&TSZY#|Y_sN}4eJuv$2Mt@!ko9hCK+3HQNIs>?rN|cK zDQJAnoLzYLuJR~-W42!XAmoI9z2+#0q#_ZUUhq*2(p2mPdac8}Iv-E4Z? z!O*M9WplDbP#nah2>PPXg!C!5gz42+mDTgIr=+Y6l4P5k*ml<*2b#CS2 zUrL{Mg_pM%_6h39EoDzY8xYu_LJPDw%5{ezgbKYC!tFCYoY^GP8G^eNhTuIj7jC0> z3NT!%>wvXPsra_2ThbTXu0|Q=g^AzV6h08UH~xyIl3qs!K?Zj3yhNy(eP`jZi1Gz} zqArkJG10)1e4Qzi53(ymbobb8KA_u4x5m`@t)S9_6U7xX2Ym*8R(LJMskwr=$P8Cl zPVNR4+gqL>NQF!8F5uJ9;qf$5H&?G-_}7IMpbVWN#s3Jic1Uwjim;K?hnF+ZG`}c* z&>LxQUl8G`p70}ou${Y}46eBEOG7y3vs+g`S$Foz3GSm+}_V{DX|N6-nymUy^j#b!L zja?$njYj%x7kWnH1d%1?{CLovy*2kQ`)fihZqDy^HmxQ3{4$1a9A;^-^42OT)h-HD zgMd%vfqzaYYEq+~xn7g3LIU&LXaG!!pujv!nf2pAzTAHqWC?X2kpb^&IC757U@Id_ z)iD1J02wxlo)toYKSF3pNWpv)eOKn(;ICt-v4F<~7!sc?Ry^;uaac#9SKpX@7+jrB zQIdEYlfrzg*#jNVE;Iuiudkis%yG5rc5_cLV}xv}&kr29^}M43I|s4VU57$}TjGkM zv%Op%u;&^-J5e!Fga?m)CNEUH%>H7sVUlHr(VBj*Ba0@H94phDMyXOLvYJXbCr5rO z9d?gn;}lZoxp-QB^xtH&Vc(;-xo84D@_O{{J2!`T8yC#8V8Z)uO9t6H7I%jG{(f@G z3c!acRTD)Tptv6VObfX|y0m&;sq8>l&y*|(!{vZ#Mln+e3MqT#O@XzMYy9^^v+_d^ zYzl;4ojD&7lq5C5d8U|Q%7og;{jk^pIcVIoUo)H@sEy3?Ne@no9bQPI|B5fkesE=U z*y(&Pv{qC{O2phUEYHTLeXzIoD|mf+_D>k|3~t!Rw4T{0u^?$>_T^uZWPbju8+V_U zT48JprOKkn1}Xt(r}SX(N?X*q!ae>V0*-m2a^{?>lfLTLDOx}It_&fR>UmBw=si*1 z%s7xU=14nxr)T2ZKeY{*xNR2)M{s1jtS|P`tUT$U_v5X0?Jym+g>E=k+!s4Y8feG45zEoIT7>+>hI0jODp@H!URpZLuOZ=B=GUZcHM}twgau zqg4GA>7x?5K>u%#|Dg8?T8D4p7QYtdIcUAqb+c)4mC8ZM`<}Y4s0#HC320j@(eT%T z8r)S@mjmWGS=HtX@2hGM%=nNJsA=-1?VJMjslm&q4iLP?05k<4&;=S>poVRjiDFw1 zJ$e=T;N<93A-YA#gad_+*g$+uu*DDPAU8;n{vy}2)~6)6h}I!VPMZQxYhV)Rv?9xxkZAD(h9e zAdyqIO-GjJePfm*{0VLx?B(E3=qHzC?_LGPpFzQ)*i>Pa-&+5NUr3*e@w>Xf^tmOZ zk)YU|5&G+~U+Ok}c79o4Qm|180(2f_fw?F|=aDbF=0fu}PO~=R3q4_uoVA1b^`Acq zvKXods-r83ek{k?jh&i3Q)d;rqONzJ5iNZN-$^ddM85TmvLo)uGL0Iz@Nnn}f4L zlgL5Q$8%DonTiHYgRoqsSGULXKDn+NsC9ao=IOD7ZsgNf$E*pvAu#LVP*W5BoV?BTT5-P{6~;L~SB(UWKkq+FHt-Acx$z{}wPgrJN$OE9bgQa9~xRPJP5 zpYpn5e{_ksNA%DGC*J}Mmb*>VoI4H_@Ey=`ozHDaw%$FPYui4;r#8xsEeS`n30AL| zz-|-(>r)P09D<56h>`x-9g^F^laaS09{3x9Y`Z_!wH9OaK=Q=R=u2I?{q6akvcIo+ zrZD=hBL9n**ggRk%PpO>)2C9qmaUn#43x*3zWvEBSA6Zxg6^N6(qfVCzK9leJ+v2e z2knc;ghaJ>rFPA1*P@0eQbv0!E=9^bQy*fL-Mco|U4ZOg*K6cPY&+HCsSGTJ-PPrC^(|b;EzE7bz-=|o%S$tC?C6)*1>3f8;We`EyA;3ohxd-z z^0vPq9t{V}d#C7}{4?cAI#;;fb49pO*(hrjrFsr3%H+t5yEx)3-Q+(=n?w(Q^v4u< zC>*P7FqN>yZwD;N6UC<#IYCFsq40jjBu9Q-JoCf2JVBE%bC%O*GiRaS7QK#s7W}x{ zpBhOe`8+_vmKWrOAe*fWZVIJZO_3GD<%N(m0#zBLYB;D#4|e)k+chS30K%^OIQyi= zZ)E(Hwg964`@u&^k{kEMLL2O;sL47?wU#1ls00&~K;4;hLDG#4^{^=RG_rMbQ-qaH zlwY^^vD-1sILFz0JExt;>m`5rcc1rq6ALIRJaUhb<@^F~?mKM~pmP}op$(KOog!(# z%@C&740?Ad5lDN(dg)fdM-nXhK_dS`zf<#^%trOhakUdi6ON@;S~&afYpD!9WFYo{ zc}BguPtYZ5QSR4NOS3`?U`gE06lrRvwJ0}%=!a3+?cb^_mk!dUke%L6jN}S?o5_@R zD_lF_@IZ>Kcl}HhI&y+4O1;sxmELWa${aoZXaREQOERpe)t!zsLHk$T+=m|dVM87r z&^cJ{2hC!Ny2*M!y?ST-In_ihpT{xej*O(;*YcZWJ~8~COTm^&rtIO5_mJEdW|FD3 znq+oSs_hghK|iGc7SvDyx;@6i{&PT+DukNQ3_*<=+XmMAwkoQjX#nE;C;W${5Zjsh z@UB-=g-xLQuwPRv>J_eFbTwWzG3mj>72qXOYz1&O@Wuf!TiY`7*>3RIbK;L;zI*Fs z3uKlok2pjgzA&qoEmn|OOR3gSWF?hwP+b+IGnZGGiA_ty3tc;(j6Du^A!6)HUAi~% z>#BTPPxZM<{BiiB?aW$5iV%1@hSxE=Z2uyGUX4vay~0Fst)L3jR3{V$c8$N~op3v2;Q>hfgeS&h%-Nhk))_$t(?p}3Wq+Pyfouq0xY;O&5yc7vL}&hs|Q|6sX)56kV-hBD2U0E z-WQGh#rB*G_Jn_{D2pTor_LMQlmqn0OfQ`g-KOppmc@dkfwY${R)03@Ojv990~hj{ zIQWG$9K>fCp}XhF+QZk)^}`zMhOE3_P;REH_j1TB2fWIjw1 zo6VZ!7-ulDJu2wtSuO+$Y-k8F&e;nu@a`X^ymp!lkTqVYUH!55sd>6JzNhB($CgDW zip_f2YrIm0x|NF4kq_te3Gj@oey6_Ft}xE&owHMUYTh6V9Y4|>`g2-$YTgd$slG9L zh1c-Q-zmaCMj^dCTCW+`yX+vCxL#kY{Z9i8*D+pTa^EVTyt#k=>sKvc`t7N#@00i3 zI2L=w3gsVCs#b~|qY{t-`J6moP(4qFi{sWvy#}$4JM`*Ox*Vn#Q|PcPDhe_mnNkSw zgtk^}Y=+b*7{$I=Jad=eJcGBwW@lV@W9iB{wn2!roV&wr3!zg^uhHR}3Frf`g~ULc z0Ouh@Jh)mW0`8-*&xP(M_Va`L&IN~# zIa2%NzV8d+sC~kbZxgjDFUxu&^y(5&*#dVc^_92K5B7~iHa$?wFPl*qd~E<5q7?!_U@t=Z8?B#wpS$A*a3pMMZ?Ssk*xwMD{RT-7g$JR)(pJ+{d zy{0{;Dh_)A4%2B&4%Gg4`!7=EP9O5P$hJpsqqoaeXb*~N>Gl{bMY!x|-h$q0VKse^ zV|+E|8`dtY&!=uPDG34Rz1QG{`IF8(ERkk=C6{oZtUL-Si#_9O4UUX zBbAUx)(J5lcTQd`)0tpdT_boHRsqfRiSgirWUxihvk&!Mu_NLS*!izB1r8826hJr# z)l>U5_kFj=3@Ov%;0nXc`@Z0vV5@!?okXq)`q*oNBr?qQ(ymTcRmRpvuJzqM15eMP z6XQpU#L#|aun)uFYde6$HA}u_+BzhKDP{%;s*>bL2jtKmc?CNBbERF7#Xu@M9j-`M z#2bJv8l)_m2sYN<_kGL8C2|gYp0GaS*Ibq9zt7x~fhh6~`XA|Q5j0a9uel)i&e&Fam z!&HK~wgJ}9T`^n2PRHnQh0{&9Das|?>~)0^%Ft_ryPzZ;bJ334gGUeVsTUwH_uEtI46fbL0T0iWphj)+b!wR4#cjSdyE+1*L@z

    0(hrt!Cb zP70x4xd}XwH;+Kv@AIfPtwA%6!L@5f7{^IgxZpLo?z^;Y%{jEhmVnc3tr7>>=Kclf z#F}ZGSF*R*&q!Hpq0|J$c*oTB@?N@LzJaNU>{9j+BzjI^dg!v)YjcM@fb~q+9J72X zenRo+dQdFy@oN?J(?jev$qD}==*PJ+yGq#-hu7fNZs~66wYl5q`taq^Epdf&Tj87V zYHP?^;lz#)E`+!p-_CssKZyTe$EyuCylR_I2HUGJfIy0_K(odd6AYMvI5OorR0f(GpRmCeMf^~V zAY(ElJ#PB0hDNI4s3Go0U9ojO_hlC(og!qoDwe@lC54&-4Rlvj&HU{#aOd3U+DLPr z!y*9pfjVmna>bTNk=j*P;(JqAGykMydrYmsOd~tY5}V7EG02G3=BsN{tQcnUyQpb} z_OjxTPMyZ4&cJCgo#-|3ty{naU|gNk)5Ls=SLGp$IFC2x;`)CL1m z>4Kc#9gUY3dMZQbcVk`-f7E4ncTimU=d z8!1%=Mb=RXJzu>f!)lR!GNkN;K=JwTb+O&L6Y-bZ z@0pdmb)#*3kRuFdZZpLMVr9(w(C(-{Rv%n36Pmg~Jx0EF>MBMz;N1qrKSRnwsculv zA8PH%ZHSSh7?tvMjK+JuE^*i7C$=&Z;3=F$E^qb@a zzeK4UyDLCwFshaA5~Vs%k&mbZ)a%<#T9pR6Hgfw6UAbS07^`=7OgTHhRe2?{Rf(?3 zLyt1S`*Sl`toA7p4|>H8nedsM`L2FZxL>ZP+rhIxI#w&NAj~7wBZ(~KhcNejcSyb!!ZIk;I*P2t zl^dcXgDv-i=7#dJlbFj#ss=u;bsc}-l`AQ6fVs|0br%wRW$_L4S5Eu8|Lny1MwFCErc zqex1s9w6X;&&NGias!ChEc()#uYVq4OEk_WndQcl6$b@*r=S;# zdoW`1N$4_fy}DkxL{&yr#)9M@-6^Pv>=eK!`W6cz#TaaY=o1|J=A{K7r$FUM%azzx zQARZpv%o_d+{p#d!ohVMUfd9Ic(*fxUqr$Zc{a?W!)ZMgn&KZ(dRU4xf$e{p_N?SR`V-lJTp!#f`hSpQw)Z)Y z9-E2S=6=r4)dK0a>r21=JB$1JKW)F5NpA6TU)^}lS!I=YxKF7*qewqglERt?d-bqI z5ABLcRv)56IoU!@E8Xgu7K{Ie#s38f%JqCEuZ@Id&B4(0;D;XD=kAbTxnEj9&9IRv z(3rP!=$b&R5PaxiW_KH$tQ!Emb+ccDx2*-jHwXbEp$RZw(YgkvD+*OMZh7JbCx}UY z4%F7+oC`AiB|S9M9R@E6sSUmfq;g>B8W`-tKxeAevzx}BIOv@s?NZ_j3%~q*(ksJX zYh=yxb{Mlph24y3xDJK z0!?DT!i5V9VG1+MgtAr70ePh$9}-t48g{oSu0V0<;qU=@@$_>cu*20&=k^-eC!J^4 zn%WQkbNdoo5d=OSu^TV0IkX&xW0Kfih*_m`^3wq(*Xdk zWStw^kzH1}*h;BD%zHDHa8Zo%qbK}Zm}Wr}xU?(8w`ONC6@D9}Yo>fcKR$YP{+6)w z;qAZ~T*qY2Yz%iL5-fu;g>g@hTS2Ox zQq@poKb24-&YEHp^+Kb9Ib4f%D`vXvMp7T%svHbzQ*4}P^69pa`{KGJC9+op$sZFm zk7F_kL&6xsOApS0ME4-1yDv!Uz3e#4{#$Ui0iL| zXoJhYUo$5r$^xOh;DSuDxOJO1s=17N> z`(Txe2R~*%W((=lL{~;<&4!tBox6S7y z#PyH@9TvtWIo4V=1USJ*b`5|WC!77J#!X3%ERo%~dWK_(JW?FoszB!kD($~_Nmc=R zUQL3mP=jEf?C$I|!7Zg;-6(Rj9GJ)J%@-a#Ztc*iUnr@b_V>jgfWbQ?HGvLb13jE1 zfUKnC!0Whq;Pou|QF1|YN^u~pTu|fn`15&5xj=`^TQ?(Gm3iL#r(?$2e95h_10i~! zK_5I)735s!h%aooCL-AH5KL?@UPAux!*I*ARMB0YL=4Z_Qt6P@js{6;>ih4<3lRq5MF-I3ka{c~{xT4gr7P!;r)DF28myo|Hbf3kqAXfUGzK zw|jshr!joUV+GR+3h?PrB9j$*+v?aUty2NwM0_|EAl!a3vupj_=RE+kDZGKc;eTUx zqPXUj!}MSGO4|Z@No`n`YNNa`v_rdFS{8el{@Y%uQ-|926}0-y69-n;y7B`z4#rw9 zZR%f12`yG-=BDqpkY#SXs0Pi)Q3Z6FlxjUiQq7Gi$Pn^QVtm!~L0_!uhob+UDap`$ z4HOXKKGDRX0!Ee(FkA@)+k5i@#!J;Zr2=`c94f&e z-6_am-}fv9LD=O?t8y0FyB{uhB7tsm;7uw4`zhB~ykGQ!5D2Iidnj`stn zQWe>b4}Ru@9V^FUteZJ;A2)Ws71$+{DD=Y#KJZ$vkbFU|rxO;f{0GH8Oz^+2qgQZaB6;U8Jf8PRZ99XJ+GbdCpoDCF zcz5Uc zAja}MiiZxa46zuF-|cK#OY+Au9Bv#7X|OUJm6Qr%{@ba9#EA0Ymg|1n0K8O}WCq3* z)1j@EG=jk3`{F@GqPRWsE)W`LLBZT7p|z4B5A5`>mDG3*$m^!nN^(NavDam7{!s4{ zd0%!}(jIvTO8k+5$oXE(_>g3aET=AGhLOdMq#o@SiS1g3Pkn(KM_LcFnlyJ8|IPhYjaj4lYo|Z3?gW z`-;iuGM%DNzjR5aS0DaTy(T}T9`b*0e>^8E^#3l2Kf|O%-u3SGZw^|f-8{QqQ}2Sg zWz3P_36BYJTkwbe{Ag+Tix39EtXpD0Hmhs(%%O84us#uK4a|l26K4hRk@6U7Jbl9c z$L(8;;ibNGC_m#3AB&gziwopUlIF(oF(BR@wUo`HR8X>*O(nF0+TJDhKp5~G_&uc4 zHQiyIzMm?9`BGgh?ttY>CtWFMRdmyP0)LHu$&Z6KT7K90r`+en(%}M?~R`h(;uk@{#rX-|z z`XHO78uBP1*A*>sr&NdOUSeeGnPU(U8@Yq|zK7%pkJUhqfOmz3iE8�`21()W^f? z=oTkD_zfyleiN@hCj9C*3BUc{pa1&JKgyO-s>KvZoH&<-<3D8K+wc9@5k*lq_HuVy zy)*_&1p)s&AZ?Wv%9hU96t4SJ(jVJN7t0LLOfjqqcwBQ|W}%N-5~5qrE)pXx8K^mJ zU>1o>q&t-$Mhbl&5WhniN>^DlhXKzq(vI%-_q|o;2K;WbZ z7&U;rr(ScEJ|Q-EH85AB4%7L9vRG6wshPhr1dH3cGx@c_L#V|uH)#w&%vg*Yy{ zT+-Uk3mcE74Vw)Y;DZ|*W304a9$M=E`&TRzPsnd--Xt5x$_Kfz_qxYwGTKI|K&o~t zm5?4>s5ucn$nKd}8(BoBMb^?c=j>uWQGl>DYOp0oSA_2IN1x=OM-h;u;z1-s+v7g~ z`;ZtRH+H>Cvx*ME#|5w%xg|GzqTjeX(~%}qH+F0}noN=wTpvzvs)4E9g#RCb9#A#2K|S$CjR zshzpcg5EW-%U_n2(MHxFxJR0U@?|NJ*L*LE+oQUa`+Nsnocpm44>og+`J+pL!EHR2 z?N3nnTjrUcoOtOJ$(jT`Fr$}~C6o&CZ3R?$g@{=Sh@=21oXiu27u+sCnc4W zJG9NvH)UjZF4#)%2jPH1AoSMo zD7b74xS#(RbHDq(mLC5UdD&L3*lnFy4jtt|@7vxdy%Pg~&cp~DDm5`jggPULrrq{V z4qhBl8`&4MBmn1@jA7!DhaO1Keoc@V4+_idN$;9K;7Ear`!+=v=qqE=U}Orq!WKhv z(cE;1ehQLEK0dkz#A(g?NhKjhV4N%$IIv)H5anZp?p^A3jefc76Qk!k`# zVGl`q&CEB$iUDTYuVbJ^;cVhBXM$$D02%8#UZ8ooZEyFrUs|TC?Ot_O zk|C?PwTDsx4erNOf)NTcA<*fFYYHBaDMmR{{fy@?`J?s>ee&do1 zSvgSOx?FH$7UZ?)9aH)Q?|TkJUDM)s9Um^6DGo^`IrIpX;h+BWr(DMJwdDe&H+&OM z?a=NN)P~)S!a$<=;tUo`#Bq+#g*Y?T@6|)3pjCNWxCDNbb;!_}FdS{O#<6XMMs66+ zk8SzDHgdxeAUP-R4^4zP2L?$VdSJhJDSd5j8*>VtZZ|u`!bF1;1a89AAXpiPLt(B) zKyC4d9_`F~f>Zt<`YmEsv%18|){oBW_d4*w2}9t(w{u_0Jp^MoN$$I7)m=%*o@Oyo zOMV})hUD;Tn0I621r!UT@=QA^;5r2riG==FO@SYPKFx*KF3A>ASJ@$DN@T(7%YN`K zN_cg{0su=cK9Fo;Oo3GrBYRMj4tf95*Z#Wj?IDlS*LJ>M%rukI*Dk%LS7YkoD5N2( zg8E$0T;c@aqnY7G@Hmg`-I4N@Us({-{GCsNiP4RdOi5O(gm);_ZHjz?qzT#L?D%%I zDJEHUPF@|P>!dG3uKDcz<)Pa`cLWdX8DT>}n=gcf6-e@4Kxd0@%w7_pds~<9Q_fuV zGeE^f-89{!{kX9u_->RgjqG@BtpJQ?O<=ERJER)k8K!OSCp%(p3y1sLvg1LfCMqQy z?JH7tAbm%xs6M=1(HM>%R-3v$99kE`v*UI3;km*k0n4Ke^jSgU^oJgQ)!}y-LbKyb z!v>v-O!RTjJF&g=CsC=s%`tV; zl0o8#8QIjUvBsx~v_nC$c|~_!VVJqX_fD)HCTXlQZzYwQi?Txj5VXGUy>!MAxdY3& zM^?j0_Q}>2dUXHsZg9iuc*S@7MHb8T+L;aSkyZTWIycT=Y_po{vMALCilig`Wr}o= zE|}i{OMhIfb$}$gncF!%7#DT5&|zDx-Z7D4w)|i-UJ&rVPAq+^*|v6qPjiYJgM*{e z0%g`lbZ;R;UP*L+D~rvM=0;a4J4HFPDMp8i1iEx-^}ONK1=9DPlt3xhm0|sf3|6VJi==|VfYfMIc=8uS{UeToZ`9{`aOE#p1e?&6**%CMQn}OP6 zl;2!LsX&<|mrAJc8}NowRw#Mxq^}81feOW;*k13$@y0nF+FQzI=tF>Rh4{|+bjC0< z$#d^Kv@R8MS5CehmllVCc-Jfs9}Uv@f%51UF4>64s+j4BVvQSnG#rXG52Efy71Gu7 zN@ec}bj?92!U1v&N+?EZZ(NK6+d;G+7q(Zr_H)$i*Kc@Q49MXx)c;?ygr5O%V_X1P z!zcriPN~u;k^)rl(C%iAHr06bPXC-iUD$pmJs7hW6ATLbJ6v1~f+z3(-u^!{+ERCL zuwg8p^yZ35D9Vu@oUA$||5SoO*%aoSsyS!}bS<|lM(@8dyNBKxe@?Xxm>>_+dDAXO zyZDjY|K5$7q=z@3d??P`!@WEhBW~xuv-FRa&J0^%F(2PK^ut}`AV2fr#_sA_EA#OI zrFx$t%~V1$vn~2S$gQxnDY)R>3aT`xR4t+``5`Gb(R}9LB-}A&5#+6mK!pm;H8Yl| zbSa>Tl_jt8t(B~WUCV^X334XvoZQH^$PNh~dYCUjX^La>^ye8ve0pxMtii6A$6e zC8`c(WlR?su4YY#yhM5qd|ynMj`Qsq|65`EmT-L@-@Ee8(Qm75rD5GRy~!a1iwoT0 zVozv|5@mYz36deDy)zzli4w)IJTXfejdS3}^H})}_ISQweAnTf=4VEI9d%T|oF|5@Ms^yA#ns#Df~U|YJyC%)poQ!!Ts+NQ5x zl67dSlmoJ2=AN=UU^SaAZ4T(r7Db#VeS$oJfll@8(4N=y(ksNd!nIS^kuK>+WTSi? zSt-33&>evJD7?;v7RxahJsk>;94*@A^@HcVa~s%!puV6ocCW%D*X{AI6Q1zTp+Az; zMxrzE7K&VDu=U}^(^2a;m)#NuB2*yu35Q%b-sUkG@2ER6qV`$E;?oI!XnCp4C*s6|zQ{9-Kby?0j8^YAZN3T}s`aFr2hd`>Z!k z*Bq7ziU2K2s0MuH(=QCn)*aOJl5HU$2o6g2MSxbB+!WZRKpg;+vK!PL&&iuTO>$%> zdgw9u510M6_+jT1hIv=}==x_rP~?E*=eS+n$0I&oyp1(Ijo$b88^*V_I)7lUBcVAr z_IWr$bIn2J45|X42?CF}*;CgTwoANA+y%`m)%3Dpmzy4|f#QJ+tLt3-A>hOEEL$!r zw_OZzuuGjMFsn$rp|h;$inwBPVym^ea3#wMQ zF~e`xsV=EKVH>t}TWoLyVsVk29<0NcJ_o8+pekU~^e&BIrhzuh%vMh%CSX+ThXwZ# zy5a~wSfuK^za|@TM&A8$J6X%m8FAy)&2}qiB!^OMrpP8LAy)`&H%cR0q$!{){n~)C zs?u-B!~AK#rh51j_z#*KdcP(oG&`cwZ(@LDKb{`B60U&H6Zec6^uIONf{+~cXC-7e zKZLlkv+$G#Twa_Y&N?tvMdaGSvp1Kkf*pH>4v;u zCA43G32K#J^TKS94wZDTMBSN#QPz`muCN_6u`riBoK^&N-`r?eO%3{_3fK6i2veB- znp%-wy?*iqFLYiT!`guDXt?-pZE&xj18b>q;OG&c?AUk$ANKC>O8vJeN9t7EIB&>N zr?SNRh6p>Wq3$HlyANv78>MGt$HSOHpXYf^+1<&ueM=neK^TY_^hTkc;ldw$$u18^ zGePS zSEfa-6kb;>5nhe@NRkVjdgd-9z4|ttM6NJvRJqY7fQe|ce0iu|y-rda+@Vbi+2uEK zCU!4574MHg=LFzCx|cII^r~M^ecqSM_2J8-Ka=-*>##!v6T}8O)29_GWQVgucr+JW z6NghbJRUpG(dKby$4=#E=DzhW*Mu)yJl`eDBMy;=9&T&-$QCPKb}gk^Ly?tK0+JOD zhedb9^h12KL_D#vlSjj44vc?aqTkPtPOY4`0Dz?>v+mz(6B zqNVxc_LWglp4rlco6K6U)Z zb?}{@*Yu3V2l?Im{Vim%8~Y%6RzApjN|j2HWGdn8e7z5fDleU}B;rBT@mEbWCiHXY zE~mF)J^K;cInK|%k(1BhH@k#n9=&6$i0HQIJr4E57Fj>(0SrJK582nDx4l4v2|pLS z+|L-R7XuUJle_-I*J3Gt(6HqUS^vUhZpy9fNXi(zIbrk7=@8d)aN^TC7woo75@V6y7RAf|Fbbg`EIql$*k4 z-=&jwgl!HpK{{^Pq#=)`lkP!LT9?SYtq*RMW14Q+q;~BoZKr52^i(X9oM$i|+7MJ3 z3stYvy6F^YlB!iwsDYdyXq^;6a^^u)B{ZnFs#+W@J{UhX#(<0Uu4h=fryu4u2T_;T zch0x1l;$A0#Ap;!;^#&~$BS5pP5sr7D!t0;FqzZid0Kv4hIQvvo<*80`BMK2f^NEg zvaU3G&OaUBJK!(;OFM?3DLyvD?Rlcv)be-P(L;fWhBhdByQ2E5xM z3&3bR^g!aW0?i(OJ+nIu`N&TDnyGn`Aw4tv0^rny94IcHlFS$!575VrULF^kr1P4; zVwsp0RExSugBwpwm#ikHHcADoIY+4koK%W}7X#HGP8&GQ7-3Suy&c-cukZNLitil! z=ECL$<`NOh3G5!FNpvqN133Ln$_C+}2r2$Lw1>jWXj~v4oNDlDg?7OldKZJeKBySp z20n8S%zn6tE@OH~Vgzy(j$DuIg&5GSirXtSxA=?$SK~p6^-N^XCm!&yoXX1_QmPq$ z`P=_kFtuW3$~@A|&n0$WdrDSYF@`;$RD%?`M05;~&xj7uk( zhAUK-O?tp4O^96LqgRrT1$cDP1Z?oL0B)PXJznbN^}NURs%1CX9O--F_r!*oxD7Rt z8Eldm3$Vs9JN4LSM{g4#hq#n8)*6pUiLqIIKjs8q^PE}1+ z&8qgP@iLNXpK2;eZV0QI)k_XeMVna8)1gWe)$G47sej8S$a-9@(-l?8YtQDv4mLsCyDQ%e&X8H@=TXz&NKfY>}O zE=&Ht7$4@Td71Is==R9WcwEO+`T?akkN`nW`D4u;!q1RhmOl}V}AQzVs2xcY<77xc=~gRijHL6syWKFhma zeps3raKpb=g1P~Yz5cxMVLxJd-`b%Icp;YR{ZUPgZKonfLtwdIRgi9dWM>pujbSyA z6zOtdt)vkcs<94sVjsW#Pk&-6*G7hn*VA|3uxi{ki9%wr7b~+b|B58@3*ES}`%-FU zUA9oFEQ)LZMPX){_nkR85PZ2l{obo4+3i>L>Vq?j)tT|>(?DU^&wM-)yxeo_+4&O% zb{;S{d#@unbn)Ly|9r(i+=7pa?(!sJ7;B~EzVjhDWQBsglnVH4%c+EFdXX3ycZ4O< zcJ)ahVA7fBdirx_honzg9oHmlj#)lcw|uIBDVSbF_s-EZOkW#}q;a1_p*n2cG$8iJ zu$wt-R!g^f_5_?zTvw!nFlfGP7vwp2OeqR;W~PBRKAyhDaT;p4&EWUu%>2k!dgcWw z!{boU$%r=5xDeXNREM1*H7Zwab>j7r>#pY`gt$eO?oZQU!) zzTI^Bf*o{DnE7nII$v80(v#hpRXLoi;L!~ML)m(3tP%{x2?_NAW>flgr$##=uNb4Fg{&?LLX** z2224RQfA3PXm5)jvhRMt+u`swWe ziwk3#A9aaj^K)_C*LepPu+aUOIVYCc5pd+!aq+s7?C;c7JEEwg^N|EpBO{?f$_wp zF5U5ikL#ArwwRIM*Jv9^89y`P#{S6>D>HJCf~*kP0|bNYR!wDWr|2-9FEcV91UCEV zPI;OF&DssRHD0Ze*JT4!`rr)gIw;g2qd#a4&o87)q{%AelX~cJ*{@!n7*Qfik>Wit z@?7hg12iv4JEyCRZD87<_$FC}^;*s`Gxo#h=_{mBN1Xzzj40iC+qP#ucEycnA&w^L z7TKfH@T-1z<{1{PL*N8_9O6BA~lQqv-Xr;u;Y;2}fn=l%f zfT`VXNv|SH{&u_iD9M|aD}B36UhY>tPp>{Z|2<80%(a=i4p8~%Q)WkJOn&Pv2WXk> z9te9tFzS_ep@U)llet~lloKzl@V3CC`X9b|lB|Bt;8AP^k8Db{k%Cc3zSX#Voahvs+)6kNFjxC9nNwF>>n4q5n%r)bEzJL8)u`McI5GRA9=M^I& z*xZ5zuux-*OIsih^u*Hd33oQ^#gJ2g!Q&>!H(WhstCjnLpn^k1_qw89v);EdsB!uQ zNd`My4P%h*6lAa|QqbpC-k9C0>{3?GI~0B;a$%qHZd3-lTee$@5?M>4kE!eB_2K72 z7Ot9{13c=6skj6GVzMV~N>xxZ$s=jPVV24y;9O7j+!9w5y|7Y(NyUW=7diwxdm;3s ztMcqUJP`B5sq#J5m;b(2FIaAW8%o1Jj;a8Os8T4LbZB6oc%t_1umchH-pZ-x8Gk1E zgKa&D+lJpc3SErEM5Dvm9o0K0Q}L-JSzF8ufC^9#y?5dXRqPn(;dk7rqiDy?czGaH z;Qv;Comf^a6&|_A$Z~#BGB>U#DYRP8Y@k%>6iK5J*81OKPfM;!2FQj;C|PBPrk4aQ z7M`VtJW7IAGmC}o;S*~(?8Om)$&S%?-YGxo`The(rVuxdpK=7i?hymMQ`pbwHp*{C z4SGWv!2g>$#2!RAZ=CJF)5D6k|NU=^t++97?F@3`Ig4;Cw`vajj8gSeq>oA%X+h3l z*T0q}&yiLI!OBEGw<@SK3<`r_!BPcDI9%#NBTupKfP95#!OVl<$R#@HJ)~@4lBYm0 z1XnUSbWVisqU@IFvY)w&6)Q`iLqC26DGk%%2W;xVYWYqd16-^_y`T34pVLE11Cu(l zg4LOiV=Avkn98&w3CtC?o4E#M`x)%A=*Dop`eKl2HdMfcFN@Y8QL~Z#9J&*$JX7FM z2M_`1@JP3$jOm1q#;hq-o<{SzZHf=&eX?t8uOAd0*n_ZVJE+b-gM02;xb{le^UhEU zwYd;eubg!yQm=++S|KchFG!N3A&Z;A-c>9_6(}r7gUyMWfq3Jx3xa1*Qh=Nip-@)UCm&mYd_1H%(V3*&#vo+wFqx zn1>!1m(2>j4GJ(02-?~?O1yFM*fiMr4dXkF7fxQj^i^Z8BZ6UW?9p%thUGxT$Ql6} zmSM@DL~%Dn+iD{V>3X^ALG#3r+K-|qE_LaSpDnEX^VckXP4h3$-y%)VnP1aswbE&) zRBaSF4wdtyHuAuNK2e)*tteTQL$`wuv2@1zNc3-v?7i0qy%XcRVwydna$b%z)26VR zR~EfeB1OxB)d83V9_}`iEQ!Wy$3dVV=R&H@*sISQmm}@dCNq1(x}r7Rj4Uz|e^?@(sH7i`3_ld`9n0~b&y9jH z9F~`V7$g6>?b6z9k-@Q8M*pPH7g|t@RhdB#J@6mi@AUYe_v0(=LB!*;CvJ-uB3|0{ zPifz=Ev9!{L~xYH*LanPv4RLo(E6D|O`)F-1;8@cPSO4N?to#=$wJK*zfMu1X81$m zoN~VlUPy4KyRIl^nt+zF8&piOaTM0Q*vaSuMA?IvXU`kX#}O~Qyxe*tV6MXAvK;x= zw=a@){KB?w?8@%4a%Q$tsvL@JrV>`YQUf~4Nzep_d~r}4oi44OS1mOKLhxfczy zF7(p}LzCZbQZG7BjKF?XOY5%s?G)rPpFsP?epnH2^BMFps&nW;!K%6V@9i;bd>?uk zf?7q5;kD3_ksjPbUy$U-Rcg9v?A7U0Zkqnk<0BPxbOiUuLhNlPaSl|*4G3<-;}mr8 zvN10$eAg#SZLu-hqdz}L@?IDlQ)gvk$|=C<(V}PYgaI8TN#@Hb#EQe6{theJe<{*h8F^DH(m*G;@Zgln(Z-0wkGg{Ko@vM z&O(?|U^VI{YwvDOv29OxTa&}pa4so`IiNx6q-HXtguY>1>EfzKw+kkdHc@2|@M&-H zs}&T|hZvn<>Jr67tboQh;-iMfMVq{A%FBn-Zfk6-5IBXru=nwAjNJ>m2}w3!>FcIdni&#C7c17(j@9C+o8>;nqO}f z+$X^EqDUl5UIsP5mHgT6`sqo;O9N&5x zBGwsgo4sEZ(FtwY`7nH>o&|RXde(f>N1J(GjV~wOypPYglya_+-+#Zbh z#5P69r=-@6*UKCUMc^ZVwqnp7d+1^IMRkL8!Tbgdx}`~=rVY(}t9|stAG=v_bCHSN z$fhvKs$&rI)iumdVVnn|Cq~^dR6TJ!r|l?O^OrE&9xu0@h&cMWjO@FK@!9Iquq1J@ z>=7E0;Ben$^Oy1!C8%cfm1XeilA8@CQ=~HjS%Z$AA(_jC_|0T;5q{-j=XY#;} zwQxb`bQD`(GNoEYk>yl^nP9SRF63@{$^OuZSi6id0(RhGjH_I~ix(bVt_+)bEZ%~L zkj1M@$-YVCwAEv8p;W+V)kr0LF36K!l$!sY!Cv#d2#d;#(l%*__6*Yt;jLCtO=PdI zEEbPs26Sk%gA1ecq{RY5Y*9>c%)!t_nhx!?x%CSc{v)>VA1JG4Wa!ki|X!_r`@u&^62EAh8~2-oD%mJQw2o4(rN|m8A=UGo|U$>ZwrR)Fk$E0o&xnT8-TN# z49BaBv~`g5KwG^(1`61_C6%!tegpMEE+U?K%COjtil^Me6^D4C!hNmx_O^}bwu|8> zXCL=PHQ`twH@$I520=5rCNg(gA2}@Ro{bVq>nAq~^Ps`7FCME^&?m{4l}ucw!EP*D z?3F8UJJvn9vHXVae^gPHX(%taAd~EH<9IAghogMZT1o{)O#6^rIxjF=T@lft-6b%x zC6N^oiQ-erJ`v1{%(b~!$R^Ff1q&0YBEcX?vVW+m_30E<`k(RdlkJGf49N2C&>oiF z4&N7#!tKROj&zy#=HOf2pM)N!_r-T;3(2~fYlAzqSBQ~04DwvZVp6|2uG#O2ap4P= zM{^wa#uMK$o*mp^8EZOhz5Si7(<}g5^80`_BJL%!Z$!w6LuV~@ioxpey1ha9Ms2}dDcL% zSz0G)cMY+QQ&2P>m^|YeZWR4+$%l;Gf*8qd$!|&OBm$)yqu%UXN(BY0SyTc_9$%kR z!K@D0=-;M3;olOsc}fpmC3q;iHn&5I8a1imb<-XnZHc?)igLpz|5L~F)StYI^Y?H+ z^|9Vvw-u19_`;b#IijfR#_kMV&8p=UvG96*28EuC`Q>%W3twIynE+Z+!EDcfn^iqMVB z?vOXHu-E36FQ}y#F{|0T$}|w((jC$0HA@1PLuEcwsa_*2TQD)h&9gTHJrSo2=QD4@ z%dow4;k&69JuFj9*=N6gha~cg!@6;%SZ_5`rc$b8imaj%u%ayqN*Z++B>TZ=AR$+# z%hU{>Hd=NWjHlheMaRy^eD=~?7D!YE9aoaBNo1)NmB2kpb%!Fisf2oZ6-i-w{7M8h zUag`k<$&yrq)WLB+T-f!rD7CN>?9r9PEq%3mt=pt$}XbrD)s6sfi;1jMeT~-7tx}| zFdcC83R0w*)6$7v{b}rHAd-C~uAAO5B~_TqB+gkRZk4QzON_V}*aHH% zRh7Z^lUhlcV8e_KEneFx!h6;aUx$poJ)kv-@wy_ipQFS~|}d(q+4Bcn&3 z*E{dNQ~o!6E-NNkkf$~MUPl)3Ge+(kTO!caG0JUEqf{vrSxqJ60Pc54kh-l+Q3fk& zfL3;>KZ@YgVJ&L8(Ba^dyx~E*zdLH*de0T% zM&+Gv?pg4sTG(A-In|_W5Dtny__sX^j0^63v-i;j#Y{7?Hp}Z@-zsm{E_)5XqQ>jP z1zUmfeDp6FImQEZ6Kemlam{xrP<1C?W0t?DN;ctn4ml8 z9;uXcMYSm~ayy3_q%Z>dSjzt#o zYnYRydi7z+uJH4$Db$%G{me%`a)LdO7&&($|H{jLSY#X8eL*U;IU>~Q!I&1fBpdVw zS*H@|Wq$+RCs^v=O|SL=;iuc)=mhWaAN0-Q4P`S5E+F?19LW^Ns7l5+~0O z53x)*zuVcgmgKv!=i6X4g;Y{1XdBv2B_w$omDS$nf47SE$6=pPaZCkMB)9>Afr`-G zA?9Q4+EZE#6eNBPL%Z3*r+&5LTc5x2`9C**EA#6$uhqYH;#Wyu-~ClQ16|~0!Uxvt za;FT&_;Bz9GMw^TM&E>=Q~R3*$zS)epyn4B$eSc>tSqDZ#sJ7pD&;t@c2VJjh48tOwJ|HxOZubcx*&i2f zMh~!ywz+Ym@y)6B8gXTHr~gP+jD-<5_Bx8JFtU+SWl&@tm2lOsO4cR9rgp5aEfJS9 ztL9?at)+oEL$D&=RjZb<$B-8^#=eD1_HskxFX=P?$u)@zHT%ZjB}4qwn{to;7|HLK@5=xASH2c=@q>mfXUKYfD_=J@B;{6y zB%e}2@%|QE`6AQX+TcB28N*f5`kM!R zYGJi&2sNr)?y%a2sVCeLmuDX@{rJ0EpLZ^dq)<~JhA*n)a46q0cM@J_GJRU>pIEHOsaa=VA|LUyB5oW%yJuxZ zZcr-716-vNw)k!FtBkFUT;spl^Yf^-=@-0o1Kzj1v*ha{_X^skFA`n~&Y>^LD}8q= zA*hCD<6+?QVh4Q>GY>s-r1%$E)+=N0%af;I_1qOwsL2-BgXn3x6p32kj{URS)qDJr zK3$g+dI0*J4PKXHx@pu3$)U^S`2V$#Vog@*K2cyvUO~)d8o#KC9GQ-{M6JWz*!6LB zK$Q+^hkNBsf%uzK$n}_3&s1SCQ$$}A)C7)Y*v5m*CoPkMd&WTXrBdO&f3aXqEMEVQ zqWLM9yYY_Ap6Nl89v zM{Sy!8(k>q(!7;3tX|j$Fj_(S;F>`hkgglqzYPlBMp4T%!Bt*Myw*n^oBFZ;TV3=? zPzEWa>wv`rM!!wbDZ+6_W~Xi-QNAYVW6POl&s5*$m|+rk4F2QJQ;N))gKU-x4>X2< z=3guM%s)MNyX^R^=Af>qA!SkI1KFUaF$`O!+C-l-dQC5V9SOno8cYP?(09>gR2enqaE|d7I<{*Zw#Bh8 zwDW3S_;`7LZuXD=zL+@jqpwDFs4q)CQe9`tpm%BsbHNLj;{c@E$XY>i$j6dPl7^sR zYm*{1pQE33!3)y`%f&ZY*ZNfBK-D8FV<$kyb=n`!dv3SQF1qcynM?9iHG-WT`Mt&)0iwRbhOTUIrz+6M!_HD1Nw$yEM-I#kZAv zK-D2#Sr=I=yExyjMq%uM*b!0Tk)0GA4dvmJAy~-m0 zmH?wLcW$^rV+t2f$&v}g1vjk*TSwWJ`#2%v4e9B>=L$`e$)d}fJ|L@Kn<&i=v#BSO zVm47E4P(9O%Fh(Vk|rqk)~YZdVx)EGQQ}G%vk2+RD%D=!#g27MhYu|`BQtSkX8o0g zE|$$^PV1zyw2vVZdb27kx(C=uVQBd}6$X7z`8UY7xh4=zvucO{8-ujHa+RJpTAa)AX{-pZAc|aGPG7xNhZ!+4RytF=r`q zii*uw7X@sNT2Hc|4L;Mo&;2~P#%ohzP`whkg^v5`eA;9Faxvz`Wl)Um_nD$oiNv=X z12)SVVMTl(pv&tRv@WeDs1?&3QzI<*%62V=iaIQGLE*B?q&Es;E#b-DI=3{x^FX1C zN~x_vd6y9}#)bDXCTqi(|Q;3||GJZC^+UK`eJ@(J~d3kN;q)IW`g% zPF$J5qC*CJR00Dd75*^l7*04EaBQtY#KeHW>{HkgG5>C0$d^r1&!=Dc>MzOB*JkRu zWj6I(pqTR%IYY%}kPMQ*K$DrcfPX@0zK1NV52nY7)(ZMzh0`raj@An^)LSJDioH^3 zq^FO3%gC+TAYLnIVUGLOdSj&aVaQqG0iXt|2|Wxer8rS`P{kK+E_}hnx}J0kQlkbV z;zUQ@{A}Sef;_AhKs=QlRBzbRs%{GPLL01?Rd9RRonpopl-m4%U4N_p5%<_=Aru zW!+zsV7H<^IJa!-=!K2h2cGL^05~u^q|kGWJj{hWf?&*$q&x zpRUB8nuF2L!dmG=frWHF>a`RDXOFSWak;z^@Kh7LbgIM5g|%2=^C9ey65|s&Vdafu zKb*YdRgc%_&`$_<6Zb{7yC+8HtE=hjZu=njdw{=JtV#3JM6CKY9$p#n|Tr2g&ol_c-AS-%`Lvpst$26{b8so2wBx+yK>trB#C zC>b&zYgHZcYDwDswcbg7Oo;fV(PeDHJkc z_%&^Qb?9Z;Zbh7^HxjS3abY_)3~T|}XeMdwPvL&{a@Uss`K<{yZ||y`Mm`#?fzXL# zt_fy*X8jZcCH#+o&rCL;KwkxEnAUjIfuDjIKux3k2?*826AXzodIF0@1eBvZo2`ds ztzLLnn#`*V&|o2BMdSf6K$;bTZh58*B~(w&)2WIjZORr|98|a^O+7eWbCtOiUMh|g zUzb=lOi#;m2kZdpr&Bu$dlnO6Esqh|UgX?x_ld4Ts5!oU$)RO;) zy&yIk?8bA9-H!SH+uH9;HG!|?t;_F|v{wv#gnt14&Y1|)l*&dElOPXReuT}MtG#RTKv#Ryn~h!c>5zJtE}3Xi{mKR9G+F1{`i}%5M1v z&lVY~UG7$-2nH0m7RCjouX~bL;hp_2}KhSqanNE z6KM3$lwAVC){9Ubiqe*=Cih4aCL^t052WMEdCTQGAH4DmICvRMj^wzjBVWE%FL|^H z29G)LeRKMnI`6k^*LzzEeq@T1UHwVL6Vl|wnZj-h1cpMQ>|`3<_{tG9SHsZsDjbsjGD2ECEYhWh-dcr;tXW9x5OYk|tF{C{m(4 z2y6tl13i6dYO64Z)<+5MkMd=(N5uOJiZ8{A?)ndxT=n<&M|cxxIl&Gc{ms zTu-X^eok{!_ALBpaN7J+VQKOu+y-r|7dQ4f&ci+S9j+TAoiT zop@rqY&Nl-q!=L2_<)M-r%%Zfc*eiel^@a@`9(0DZFJuYMY33%ogUsG>=u+nW(Hmg zKFw?9=|Q7)Fd`?QKe{4vt9$R<55Hur&(8A0y z37#d zs`*pWDLx?TP{83dKTTC2{B>^*&PutijS&uv3zkSuGq+QZn zra$_u3V&{#(ICX_bmh>VR)xhc{n6#KCTc!{5#k1pmkp`?&T|3>C3FAkUCSm!F0mTt zjZ%Vv8E$koP)rI%lBn2yiY~g<<6ua6U<%Pxsk=#zq*}bruW6#{?}m@GwHO+H&qQn= zoPXfMe=bzq@#^>Ja;<^23zq9*-I1~8+kvwS>p)tG3fB7_b>hDcz4y769~r`#bYjh<)$>tPZxE8Ud=;^n4NTDRI6}1UpM1$(6!l5 zgA)Uqy>Z@wwhGUCHwf>4`{_@5 zY!8*%Iv&;xKAHkspKt4Z?6HyR`k8x zq$~45DhL)YICr!PJLUb#e8>W7Wk%JY8bRSK1M2~*BH&>tX|b5a?vg=zDO4=j>mCa~WfqPSi3 zTJLno`GJh{Syi&6o5pJfTu?E@kTh4}XAijKsoRu&?sc*O7vy0daKQ#j4HkZH_bB7t z@oJG>T>$O0>N-zNp=387QuxOTRJtlDC;)V@d?GN|+D!=j5 zDV9BBPMc3==@08sCNbC{2H!@y0h%~8DAA6E!#Okze5FM1ldgm+8HJ%+tUIbhZM?Y8 zy+N4Fb2RNFw}Cga2HZ!J!yW8X)hMk>Z_GZG{gGvK#c5%}5-({LmMK%_weZWBTb`#|4h%h-nV%LS+ z$)3sNoSBdv4rWW*53~uLO(5?vRR$l9-Rkyfi zy0t>H`i^kTgD|biphER7rWCa*OfNo)eB|EX`Ywu2*<2K>JQ|1L#_|!~n_9 zTqH^;gsOc|>Xq5u7I6mHbL(TiVK-!%9yCW+bI9EUnaVGJ{Q_Cz#2yKd8V~bGvM6RV z1@ASsbYXc&FWs{M$uiQ^6}}abh4fh!$Q#Do1(qIkQZR^{6qOZi_hDrn4vyo^`W$1l2IhVQ(0jKkw)3ex>#{8Ihlzf8!#_tvS@{oV#>48w>H z>VF&8O$FiG9U1Mn9W00&^8vQ!-WL`Lr_42h<(&`b?ILB92}mFed&{*H1JnIsD)z3l zf?iKn&TX4J7%;Rw7|@mKcVkK<>`wnP7{4ZCw-D6OSY0S$uS_o*i4e6 zp$191Pl^gH>pky>p^COWEQ$F))rQwS+=2{kn?KYJd#{OIS=1KTWU0>0#jZH9Bg3NQ z{m7@w6+?HJ#zOaG2fdkB8KhMe0mVqXs8DTx2#<+2>+wD2W9)wNC)&`+H%;aw;J3PW z$>!I_!`*LY6?affK1H@su{(oGBjQDvy9C5lQ zOH1>HDc9X_F=Qx>LndyF>96t7EDPzT+j!UAa7mPbE>9Q0rScdTm(VzjlQBR z1*crAy6REn<+#POvm)hg{(>C7$*8yly2O(elR=x8|=h00uCrGLIne;>IV$1{^Y3tj7$Mr?xe)DVMg-p*oeZ?|M{F+!0tr@(Z=ae5u z-h#RUJ&+uqfnL-mRjXUcgeWhMfS5I=V#G}oa3C#s-nuB0sn}Gg$R+!oI85JUHps^* z<^zhHIpT4aXZy5L569y#t? zNaOBeJ&ke+EwT!t**K$D)FH|b-9PIXi4QqO+UIFL^6F7Et1|raeHubF?eppb<3x|c zF8S^hJ@8BB<%TbP6Ux0a{Np@PwP@`eTZ#9rfrhzW1_PKPSoCSJNy=a$U7IDfBivgDDErrc|jnnpdUS051dg*PHW4;c>GD!cmZBcU!{_GV$mEQ6iq6h2w>XN@E zb1jjv5ma2~GPs@_uKuDreOY8OJ?~dc`4d^|#2!wunK8+xm<+^p>~-%e;&uF+(x#ZJ z|8f(U+b+hG^G|R^;#ujD&aV_VhF;-a@GPCG^KrcSFui}S1I}&OY`>b5-$H$J zN%Jo(vxZ#!7ALkjEXmJS;Rn;(4VBobQM)BirTKJoOuVQJNHS2$y>a0>-c4yA>6e$t z8szn`a>|U(2r7`S1Esw4>IV60-cxCVdo3s+Y1M}zF^XyGbDpih*o|zw>-_e+9aC3$ z<*1*Ka!CSsMorUOh3JFZ1D0G7XO(TS9ca(*%ntd_e#h&(jCbApwRIlRy3|JI4v;Pef4-mqrQ-iv=ErkSVQCHE{@^@?@JL%@ERYrBbJ z(kPM&>uuR+H)RR@*Q~mwY?NbB9A4k1 z0v7drH6{_8R2cj`HJ3lNm9MA4yL5q$(AzmQ=Dp;VNSz^c9{hO3VTp>XP;%h3NT z{V-;X8bv#$hl7$si#?O38bl)Q$`3&f;kfV4Y1`)K&^jN?QSY3l**vY2#!t*sCrvF< zX%4-;4ye1arniHx79w#}53mvD(9n=eYf`7_fLeLz|Cb|>lO?mt%eCqvb`xeg^NUGy zPFnVXNQvM#zkDb5H~;?YcmGF9GcJU3UXP5hbbG{!Zc0(wxz(ujJS4lkUy(MW8K{x3 zF3{viZc3lHw?cR7pi4Y-a&*!|mAVKAU~WNnsCfAr55Pb~t?xCSZozn)tVbISI{{;~ zM~~|m`xN(3`_=D$&18~-SIquba@>i-FKuQ<<`TtxNRb98SpgBHz5H8HxsiGcm;R$zFW@=~tPisD2~MNnyrw=q>(1{itEHTFiN#X?gaV4j=#aswX$Eh$MYdFM zC{Tk{QKpIK4iGH#d)x}I0e=_Cwl&?-ba0i+ApfXUCAc|a78*hvR|s2JjG@Qa{U6p}G}33Mh>rV~04d>iL*)OU?+zg4zD(Bn)%sS8L3BoT%ttRw z^{DpSGbKm8%X9nZiv!Y?t7jjK-tKWsQ0wcEnb#fwvj?CopK<&fbpw_qny=JKilv^k zLy_XzIQ5tmD!UsY?{ojny}WAi9av@~sgEP)BFh(S*tdT4N4?!)2ieUF#e8XpWuc7I zx-BfFFB_CgMZ00~vI}A!7*Dw%hI0Og%1l|BM6)OABrJ`egq;G4Q=?2LE|*(m%cIu{ zFh1fCBQR_ntw)5#+%oJLj@V>J#3*Z@NfDB76v<64%c*aC^9EVZEzak>Zaab2xnVh{ zZ4{G3k!&g!OZ_iJEJiI%>{v7~MAa``2aWXc5YsCbEk3}Hiw3V(c}l%As6=LHEa`G> zQf=|v8RTG0*$gR;g)CfQcW6~=6=ed+GmmHI$bgIU_T6Ti4SO=htf5E(6?-_Sj9)C$ z(~FP0>LJ`w5N>OcE!UT4e!VqAY<|bY>~XUiU-<7MQ!JYhopx?vX)CNDs7|+MK`-f( zH9=|R#>l;jkAw33+hZCRb$s#iA_srhEr-<_VAyX}V)^Op`<#G5ZEQa2@g0+;_NW;*%E%&uIcm@5>y1j4J)X?%#b?#X!4R7ABvulCJ== z!>N%-+@y2Mhwp7K`4ot24ILrqwPH)^S=C7by+Y~AcoE*=cr;rF-Vxt>`=iz3iJZXu z#@yqB?Fth{a;E?0J(BLk7^yJBNIu0t1!xWxnkOn!z=Eq|rTJ;qILo&HEBldoj{laL z-W%6Ju>b2G{fXQL!D(ljoV$tJEQyf0MCYBjppJz$c`JR+GoRlI^+H}9oxYnbS-ZigOtjSWH4RIvN23N>W)J!Eh z-|#KVB{CPRIB{T&WkKD>JmGx+NWldP$Voj2)0NTt3~QXj2wI=hbQdtmut`ZW@o~N%fGizvXprPA?Et zR*2hUR!weGuJgX`eO^)#nJL*LUg~y%KJA|{IiBEoJB*a|Z-0cL9^=5U*J$G=ClAGW zJN1DhDbio=C<|O$!wl!* zL>WQlA;x%Xyr?@0EB_6unnk=^0mfx96{xA98-;j_R<*+WhP0J7QV=G%X>dxzD~%`M zN_|A0e|2cKYiUHbYj0%#l*LDV@P8{rXShOZ3^T~>)Jq`A30iMFJag(lKW~E8&+c8w zB8RvI{+)Pf)NBUCQxpTuN5`mG9DZElr-uv#Kjam|l3$Yxk$kNhI)=7*p?qQojj?~w zeD*?^?20hN?J(zzyBaL$#z5Xlh~HqrurZR*=y`->@>h5R8nIRzt2K3kcK4N`8$z{) zFwQYz_^7+Q5?)+Lv9w9m28o9RH=CyBiGbcqP&Vr6({4YyYFU)PlFh#GL-Nx?wzQdS9ddvaAPf zd%LWUv)c(yUu@@fbxIdadGCiNi*tHm=y%D5*T&*>nOU4xin&RVYryCa39L>iemV)Z zCSMILC#c z)KU;EPL|y9dVV`rgLkjq#>(yg?v%d9icBMP_!ZpO!X z%_@0lW_<2Y3=mFT2jdg53i`*6N!ye)3rp#oAg#ycAf4Mfzt6lo6x)D#q(qh@X;3uB zr19>A)kB>Dc9f+n%b9BWy7;*J!H^SRWF9gf0*!r*ur9d4wTGWW7pafZA1RV~O{zwo zcKSJtQ+Q;{PSeNWlbR&W@U(n&9w`KIj9k?naS`u0spiFrp3W+PWLJ-GwO?~gi?T;fNbS=(utFA(WT;*wjH(gmR-Ym-p4;TQF$?JlJ z^LA+)?-)s&Ume;>Uk}BT*e8JHHF49jlxvg+4@*u@<0Es!cZzQC@Xsp2mLRAa$80J7 zIWjLh9M!j~LyrbE%F*?T6O}|Jh_*}j@Zq1;D1W5*4{HHs!z44VTiET-XuSk|g}h?T z(26W-hJbWWCMjm496yN}q{ve$wkhV=H#hu8@}i|w5f8ftje9lpo^R}+vlgKweWGWt z5_<}pV{pM!px(;g&d*w;RaHhnH&EE2K=e8;#&pT<&utZEPhFDocJaFd;4mB`)%>-d zjX%A&FwM_!5U(o$BatTA5Q-dg>B=m>ZW=GZAc0m@8-;!G)%;qBC*spEsM0Fj`!)Ul zx%Cs21lB^`;2vJ9@VYq7Z@Wi6o$WOUy3$X=nqm+KHG;xfJHaQy-xjaK{>Mz-LnZn* z*TsjTZWE{kQQ=h=p@R)S<@FeA=k;H|79;ktWQiEr4(G(-Mi%A>qm{vG-u>P=Q?P5> zQ2Tog0DVP#WFEu|QAzX=to00O)7@UD*K75>#SjsI`sLf=cI7_rQ);~|!#`cQY37L_kUAC|5geHT%3bPwFU>WeUW^O5 z6SmhibN*ud*6L-a!qepVMC>j{g^lznMq(N(T(|!>7VqV=FuTQLf@{)#VX|&>m+EJe z>)e)nP7JvxW^Q^X#Xzll2Nm0+Ok%KZU8_biBA9o#y7z!gI!1t=g_Xe?%z!kEb{9fk zu8__R2L`RnkZ8HI@Q(86!pDjNz*-8?=wK5mlr)=oOGS7d5;WCt44k(jX&kjwR0-MV zJ4(2@MmC`Mgl-bj}l<{C)Cx3QvKa^SlMMY7GYIoz@=zX~VGq%OJSZ zhMriW^Tt9CV__J&PFMhfY)2bHTf^Fv=oxAH2+py4{QBHcs1{0v;IfH`N!e(CUT#6g zc?G)x8pY*$qoN~@wp2eJIomn!Y|W-|jNjoC9$ zN!E);!9MI>f{`=rQPm zQ(nJEld1_SrVh^NQXt{u_0R@B-jCNir0B#}JXtR7#)s;WcmF11dL+-}@w|7k_XJtS z&Es+664tF|9?u4fNufv*6?-(OfPaq|D3XR`vriLa_GHZT)yQ`@{v%eSV*DS)=?O>w z5aKf)R#_*Wj96G@_eY)J9R~`mY6uDG>2q&wov~NkP4ALEl2waOM71eCo86$uc(YNc zr+cDtYO0~{3yy=tF_Pd{373(6_hr%N-n#DA9D~R8v_5daB{5*>R3p_*jzkAkvpIqG zoVKjSj-?41ZJ!f5#=iI`I=>&j>Kb9WKpRmM@R542xCZ179@0G!%WF_P6x0ZdXKp5s zBCA97!pjf{>Lcve-IZ@9{ruj@M{{!|h?wKPm!Vd^P5EeUoaj8smO)Xjw1h^{1ii3B zflmiA(4Yo+Nn}5NoH(&7G|b>K_RzoVEW7hEia$bq?#5TYGS4(M+5`TZcPQBU6ln$SycX;qyNQg3j57B=Wo7hFPPGx;zgKq$41NMc%` zw5=WVbQ9&r!ZAWLo*B?_s6z<83|s8yCmlIB`>Q4#9skv}kI6A7&X}~CA)|?68Yyy) ziY@0|4L%XoKCe@m8U-A(%=(Zlze@TXqZ90yzfFc|kLDQESxFJZliR#=Oq@vfCd8tF zU_Z;RN|>#>%}WMz)!@D?x&m#F4eD&a?NTJ}KM{4B)B)rAR`)~xoQ;)Zf%3v$u|s0?Gu+Hm?}qn& zsWicAQ)=)PvT-umYc_fX6a(ctxlmUD6G-L!C%jV7j)d=PR`|vFNXL0@TCPabDQ=&q zbGs?JDQbn*(qaiJm9+616v%_trrZnpDbDYB^nc;#N5?ZBqd(C;M>&m$^Dab6 zFZ%t`a*6dy#cx;^Qn$XCyeN&A2I0O&IfnnL>0SKJh(h1u{|EdVRzlc5xa9B@b_oA8 z?a71bCiC&Foxl1u$#&vd>5!RSE1?)@>nWgO_kZ!GbXmwg`1_HpLHJ38Zbr7>mY{3F zO%d#8(v_KkbqjS8%R3xq;jDzli)q7Ri^FJhx|AInqdCJL=ldN0FB4oQJ^MF*qIY7r zEHf(zcuXAjt1cY=t@R^r!M$*DDyGC(MLDhF!!DjFXds429_6MqiRtA7CCk(&UAQ?tvO~#)~k+ z-|A5Z^DurG{t`PtwE?OX!!>w+j}reyLws|?YoN!AI|~@vVb=&xPPm(ted+3bu?csH zS-1X^BsnqecAA-~EfkYUkxfQ6$nwDIP|X_Ukr~%RA)|>yz1;%~iV!1%E}Lcd!?1p6 zJGo6JA~A&{VFNxm*%wCVYq=ihoB^oERUaX86dZm@O2^q+$z!3txk6 zgaZmaU8F)mpQAoWQ;{Ga4)2O6bi=B8^w_ZwZHwRba4f-a*o5_B;B0^hn28tnJ8Fv) zMBey$Y2)u-M`1(sbGqrlx1ZNK)JZZSs<#^oBdVbCp*mep8J83rD_I z`n~vXWPI_DSwGmn5DN;S+mYJ#{rwApY)ZC_cT>4Zf{!{POA(~UKK;s$FJyiH%C~oT z6bIlVu{7~Kd>)ol8b5LeJ@m-DUtL{T9MDZL#tboM!v}5o&xWs`UWj+IBb5uR?BJwW zK7tjW?64Zme_rEvtIE;{Ic=p4i;`*$#CDH`--AFA@T?SzQbZke5~D#j(H0pvtiDJA z%07;)FVf}{oRDGjr*OZK6EY@c{P-Vn#Wo24qep{ng80Xgo-Ze?SlR#w`ZV0IqGx^* zY^k@vrN+jI^WrS(3>m>|1qXdvNlxT8wN`bE-$i4x0dw4Vrj*|yA8^5~)rt<`adJbn z%_lJ$kJ>laJ-Q9B5mTc-%#KH|cYQBpJhX65Tt&=63kOuT5DC!(C-wa|4MaY>`MKZ) z!6r!x$BNI1oR&B(t?fAMFMrez&kmIMnEYLJQsNP^Wwc&T=gnl0!)8A4E{XvO#X>4} zNIMrtdi9$+MpG?3LW(^zWqkyx3(_QOndH!}shL3KV{9Hc6;%7|(WlRfG;cc4wZ~0!-}}0>ncqiRg?Ge#?$?3R zqySpY_IvMfb0p>UkXf)9Bts8y(9fr<{Ez-~aZ?!8eGaQpg`tVJPM*x`6iH>w5{4bmdshje}r2G(1JtGqV`;L#_%KIl(t zR~GTIRQW*-4s|Y`51Ms@ZZLbG;pHD5^0X{_8d(hI#GVIBDHL{m z^bw53<03ds0vTnU+ifK_2$abB`8ff&_|+hjXR)Nnvrkqvt9>@kA9$Zug>~6&$}2Kl zJeN+ry1)ubS+g;)a>o`3nF9*;to$r4m$5@g{r2bHv0U!FYQgymP$?^f>gy_DEBWjF zMa#c%W8vwAE8Y8*ZF6se``Ie2px2YMc`N7clN1L$;Xjry6J?No|0{f2gOPG?o z8%>8FB!1asH$MH!SAR*4zBXYJp!FPPQ7%x-d5WB&VvRgQyUA4s7qS?=A9OhlvT~cf zuZAQ>B})>c)_ByZkvU~8w6bbt2SL`<*hH46uHk=5mPX%F9+Ni8%K~>pSuSePHOe!2 zKwZpl4a1UH130VsJ^ZaRkR8oHVq~Oa!5*^}!N}EuT?3nV>->y9(iX33eiGA3gX3fT z&)Rq$2MCVvWNbR{66g(`PW1nw&lD4=?*7;HzmY^IE^R6_gGmO(fL+``#Ts~oD-_U| zB=1*Z2UFoJ&NNiWKcC3cpjqN`nr2LUq&op2Z@gpj=JxDo6}NZ9j0F51rf22L?qD z{m2#c5hQJV6#2*<@522DpwCeV`Q_yy&?4}JPWNwcNKYI`fQ@cJID*HT%l-EAkBz7B z#EBupQg~9vg9dw^p(&$05D3QZ%KO5QV-W>%(dZOT|0E*K?+$dHA!Au}=vr@0t#6z1 zuKc*G!@XeJQDV(l+kVZu=(|u80#CG#)c^MfOt+ zNN(?;VjsDu3NCvebRTfJ5OvU9=Z4Z^$0M2*b&^&f&gkJc2vY=iBi54vm;F&&CA*|$ zl74!?rQP#}Xu##xj57X!%gK-;`1?5hvKzi!FonJX$gS>F)Szvnvu3`;5EVl_R ze9#j$da3a+0YMX}Dg}Jf%K2ywje$8m@T(XYdQp`_C^uJe;txmj7K4NB4$|wdJ z3U^}F2{vLJRv!)0E9)VQQX~RNc;7x>4MtY_h_M`h$?7RNbVhIugy}wsz~19Vxlz^$ z;jx3p%#l|0X<#>b=#C4V-oT`p_V;b^j;;KUYkj4 zpBcgmDFy<;Tao_f0N+^g)bC)xQ|irX2;})e zM;9a>JC!fa*&j>8$8sFF8vYo&Z$C}$z2#v-$~)J|yCl_#t;=q+vENEDP(7Q4v6IqS zMplG}?i)OB`f2oZODGb_>3r^bB>@K*97qG_rI5ZMJq6*|y5L^2!qdT&L57jCan$WT z%9ao0WKG`q`rQFSlYZX{{IIvTrpO4ZjIzF&xYzesmVwt&uQX1p&sH4$I zKM1Q(8^lSIC0(RJs8!)bNcq+3(a1Mmf|ZnmF4%8|UQi!78l0y_P%f8i4J5mI`re!d z#Zmsi%mJ5zDXmimTyp6B^ZEmUy*)LuonK8ihU$DYRl!M2P3RGrH6MgQ@qc~rZvGh` zcr3V8JxU|7Z&G+Z{GvJZvOrw+e-iQ3U1JdPF#Emmw*iu0{wEQe{RdsHkQesOMwE`~ z6xre6bFlvOv(MTolWltD@$4L|IJg1VX&03w+su+BQ_LEQBv7$NN#_n)b37P8`=uG$RChLNR?5c}&IThF=#a zg*VERC4KGIC4XT2qA`;7}`rYG1$SDSOQ(E$m^eb#-!hEPqmYqsanAe}%1^3L);g2yW&kvOOrNPmGr27kRf zVa+NeFap_>bmfhx6xXWYPB5wlk(Ol1P+nuer7C!3I97NS(#M=v25QdD;#HPWLN3#w z6Gwtrf}Q1FeI%EcN8gmz(7C*FUY4>;c4OZ4(0oC)81$SJJHqQ=>|abIjyD8WSf&>< zkmG-y=b!Meg^RrZxfWc#1Qu8P2cRH;s$T|9PToG!V>mXTNR8^}*Fo3F4X>Skg|k{g zt86KdH{Wcf%?uoX&Ov4MIO7=9L1;-aoH zn?RR&qkGf5=9tU8ZaO`@0m_L>BAa8f{48%Sh$xDv3S6pgjv0)|3Fwcmh%~=q!=eIy z58QcDv{}|o7k~ApG*7Ar-Nud6_IX}i@Kjp4uoGCO??W|nGVc=k=xxpQSSz76A?Pwg zo^?FRiAJ@U{nN z@@X_jjdHBC0avwKnjVf__6crQ4wrQ=Z;Y!rkM|hQvgavI>+4A?zEZX1RWBI6t<)=0 z1y7|~RkmM+fteR<#G-dw7V2hHiL<9pw5~D-VAylqU-~Rgz@RE;EbenPnT`6-)%=|- zcVe@#&CG11QA{dDlBw9c^d|s{)T!};4n>M<%B-#7=cFiOn?jb`myqAA%*71tbAG@ zam!D09@K?17^si{)kjT|-?EUsvr5JJbSsU^nFoQ5Jj^v{8eQ=T8f-Ga#gj3l2M6-h z`@K=Q&#pZ$Ho?XWL&JB*g%q=!Z?5|0{utAYqYs~+Ne(;ljB~|o#yL$fCn!<}(SKeo zuMPSg%M=B$=-n^-RL~c8I}G*FQe1JN(J9YSVW8!VDs$>3&jPwu5GO)SEW9#cx>d>-<{qFtz@ra{YJ&0BU#N2jR{dC(e zJltH9K5b|US`c=VWW6>3sWbz~PKwz-huImM&@v1=w&P*!I~yuEH@YB)W4)|_>Kubzwh{cJ2^9%d}KCE z*C+-Y%}ZE)TP(Tb^=K|~F?G3aoCa$Mbu*(?C4?vQz@meMW^@aaK4s7a8EF&X`-~ai z)$?#6kq`2{nCr!|T`VAiwwKgM;EogJhNB>Te{@24mAacW%29!{SaNRKxoMqYdIO7T zA4y_z=rd5@{ZzKhBM($#3Tds{_z_qIb#9ubQLdL|$}@qI&LGeU$}{*|SU1WJ7{#H{ zWcz8=IC2_A=k*gm-SAt*tFBRu4O)3%jG#43=W}&NgCCU7b-AW1%e?mbEq z_&~i6&R`|7EY;2+O%9y{IZ3<%FTmym<97vC(``@`Xgt(DuS0Q-HxTeFpiQO02js&S zrE-32Kmq*SCDtS=vy@NxdthmXjs3PdMUF?fRe&CG5jTW~{PZ2ZWhj_S%+HDQ-7JA) z?7uCfQShvTPM@N|)q0QoKy<${ABwT8Gj1c^xy=Wx29eF5VB3wGfBVN7mPB~32`pGN zv`_^P{}f4237{ao6qq7#vnn{pK$ls|t5YOR1zjAbRIJ$nRXD9+s(2rltx)#baK|0T*LF$KEIJRhH+E2NzB}qp;xO0KeHiUUcX!O)oFaubsce z3l9`a_WG}fHT6Wp!kX%Elclit$dCQJyVNg>-Awl4FPcmLLlQ<~w{zlprvfvpu$f}g zDYBl5Jq2X$4PWQ}9*FgO7C^CTmmN(}zrFpxb4`Zk zoe$^jB4th-x&uDDVXj6k#Q*`-VT=iZ%s8)$zAnC}zz`e6_Klo(C<)L_SAKoyL^7|? z?S!;WPzaHrL6@uiZg*JfetjRW43a}*aL^nY&Jr~Y`h$V7qic>WrrOvqw$@b_{G1Z&M;wU`TNt>k{oW%trI&jIy1!W zrWi;s6(LD$Rp4D-y0S}ID$amTNMwbm735J6l))2P)rP2Nqp@+fkh z-+nhF@Wh5VJy7kW2$BWJFOTQ!fu@loZ5MoveC)W7d?&bgvLz9X)AEV1P|NhbrH9V^ z@>!X)FH7^mfS?-_2Yey*beUI&q7Jlyn*bqyETWRX{l^e(meGUv*(u zB)Cgd!X$NFoaP6)2H|yJioxun=F+qZk&Vw|Y*~v`-=?7DQGx zS`V2IN1|rLv+qClYqre(j9f1{@$fg94gYb9`G6udRO}|+Qc-4L4)DQaB;^{x!pVJ- zl1T7ZXJNb*MX#3np7OXWFM|GMEz~Hq$WYq|`5Un2O#_H=gmUb1q&`Sf>zJ!1g(k!_?+$hl1SrHt6218ywQJ7R%Zw}R{b1-08R(ktxn6z z6VrqyH+RwHO&^fePP|ltqLyJcBa>n_Q6!CuO=ULn4=Wm(EioT}WPDxl$&hlzBYwUI z3a0feIPZN;oa0|1)A`uZ@p#$W=LUqAJdbBLUXe~0# z7!SHsDB?uOo*+(F#$lmc6VFHuitdT#7%YiX1p9~%8cLQ0p7(BMw#ghn)o=sJFf(C4 zJdFAzP9S;1O|8qelu>)7rlMRktFoSA)=^|F6>Cn2r@;anBMqLd-NokZrZ@pqQLDpy^ zFG~_AT_D|JCtXQDbcZ^ahwcyxj7f*Ck^`dj8L5#yk*(3oW@`>j@1*f?QWOLXTft*W zoWH`m$r}pCs4Kvzhj(2Qr_Xrko+5}Bee_nF^te0xZYhEsdOPooNK+^KM2s6+RYtJp zG^6vu=V=PU@s}@_6w=tV|I9ld)_DJ6Ei8vwt_j)f;$Qs6dnWMxzVV$|-}}*= z?!YE*)UD!u0f&RIuLu96$EPOGW-X_bUv+giwu>jHO9^Zv^D+i zdRVL%i?YJcd)FwIGe%2L>-)^RBVtPs)YKTMXNo1*XuD*|61$<^-jkHQ+$VPXF&d)X zbt;}^k>*Gz+li+%mQqZ;fduwbg(jV^Q}sX~=H23%%Nd>9s_6}iJrhl?XgyTy1p@1y z!w!{Ses1OOXPG>mzx=hWitKRW=*@Anb=EG-yhJX2Fh-caJ0K2&vWiF$Bsa{`aNwMg(NMRvO>;W30db(AZuWo=szUyxLd{+by2ISKf z5w#FbtWfmHG&RC{daFm8pHY(~Z$Un2&}Pa`ihv?-zxV1lbqw+~pMu6W?5opM2W?}j zf{h|S7&puVPS*<%0|JIkI9&l$&X=a`d2=W#sf#!zSiBg!8SqRkuiqnyG3wnKmSamA z)>z5P`(NtmLlsyM5&8_`}oSaNL%)H7g6my9p9~w(v(6XUsdq&z#=O%Zb<)%`;sh#za_}5y zJJ??i*zsKPvz`C@W6Q{~(|Ru~QDRJ%=hJEG5BP1W6kY@Uw-VWknHj-}%3Rk9Uk8ut zh<9)GK##c7u?M(WwZt;o_gkik>Fp1CeoRiiHc^|qX7lS+in&aYCMvc^S>fByKQQ~i zY`v^~;Z12~VCh1gq?L3FFidy*u23ehC)#M{nN(2uAZz~`28e|o|cqQI{gSX|a zL1`;}pI`1(;fudRX5fiMI!PbIhB8HXHLzU6b(vgLH>`%DE%WGzd#s0V!Ls*d|6ICY z4dM-U7?#7{dH8$Iq15buUA0cst}leAG`H{43~k_rC z&?o0*eOYsaTo$hgE}*;TcG4--ny=!`TGghRCxY$^G?}2Rx>uY5UVlH(ByI6(kk<%? zJ@Ym{R{~xUX3Nl3>YBTWfhS^N{Th$IAgM8&#Pviu?*@VU`o11>%OJ9!uA z%i@XlFsxpc+xJUq#nU+X7|xs6Srb!q&fkQJ-k-@=l5O0aA17Wp)tG_1oMIp%RbmV- zG^rYdA4`jrwLA^RgFjVXowwCJSFpq5$c&rP95pVe7NgzKel`Ie87+?*?nvs+5Xyq zrq&EJ6%+%#6uYR{e2DnNgl4Gx#2Es`H6Sq56InSQd#zC8y?!>NG+obo!-R!-1EiX) z0be*7x;?QJ49kzPuoM44-od>`X)F-4Dtu)1$l!%`Lk*`-efh;J);HH#F2A{Sa5}N6 zVOgwW5(??j;(?w7T#F`En+i*TQ(Up^eWIsQt6}l-k+ka-wT&M&zv^vw*+B<`E2(u< zC6&umdNjq{`Qkci$+7>`e*I#M=Dc@y7e`Asf= z3-_|59(kpTJy{4w9)#iCpjG984m!5o7Sau{(g5Pt1i`5&6l}jWwbQ!`*hEk$eb8l< zC)NV52+pUGk=%G|Nn}4COiCZw?6JkChCb_%o=&bP8rA(60}ZyGVn@+Ovd!U^au`k< zk9Bg6pPlnX>jzh4JGBxL6^N=`$N#0)CxZEw(nZ?C9mFD z$DiGhw0+^)eM_U^v=a%789)M=2X3Zb?Ts_eGj~5{7qWbcr04>e=LJ$BaMa=uUA<@W6)_0Q^U@H6hS4uho{-jPYyNuqpe{npoxJM zF?S6D#Qh%q9+-WG7#KP2ufg5a$YnA8?pUaKmv^6!Cp*HDc_qAbenRwJp4}nF5lEx@ zrG|sUQDBE{b|C#-RwS_OGI3ga!qQQ)-x~*ZrC_;yyN3p2SQz|*K5Cdi{*x*c6V zdkN-OpqN&M)UB2DWnPY?k|4133caEe3$^;)jl@O`u3A<7Y-nvNoP{M;n5{V+q+75A z#aC9$Tpnc|F?ogeQqiSp-Cjfc@lz7Que&9C8_(iigXc%; z7kg!1AGxob>@Y8OylAlz3@^Qn6I!UG#4ld4BmjO*);=tB!5u)qfNTM1n~(^)F7g;y zChV#;Gy*(*r(@xM@5AcF`4CX)_gLjwP2U!`D|5qhK-(5BF6BjyNGmD~}qZm+;JVC|Y1cmDy`V#5(+9ZLFQodGIN*6%t zSL;_Q)`O3W1XQ=xhJvCz^(n;>V5CB)x060WSBLIXbisUbFa%19V6_D5Zf#I15I46` z-tU2NePjnpp8}}5P40U%%J+L0&e92b_-=u+_s1(58)_u)Kw2xXehG02BkENS3h5qHw* z;rmD_kb&l_ivsWtt?IE`4qX(`OQ7^=rbD1_2jn}1bnB;)JKwOfNU}JfTHNP`g0qK% z%J^+O$km~aBCI099rDV?`Yv1hI>&m3BhI}hhJI9vja@`F9x}mfE9{PxW(KS&If9s5G zq#x?5m(5Cy&XJtv(viU#Pu7BXR9}Sir&)gkClFITOCGMZjLwZ*7di3N!V>dNfAZ<&5($>{E&ftwU&n{3b1qd*daN%_Mi=H|o}hJ_%gL0mzXNh+x# z{&CUp?WUOf-^rwweDH<3FI{FVzSZJRadS)$f1O`9-KOY?EaGPn$C(yD zE(gGmn4%mB{PWqh^DJ5Copnh@SQzHd2SdY_BwwoObFY)_nSTmnZN{<%hYsmz!(#0) zk9Mykk8!hWzx|D5mLaW$|##+GH{ZN&4lP5{%MMNB+CkK zkGU9=u6!hG5EjrkXK&c;?(c1&=d+gg!lmqz}sOerKnEqPrKe4aX$}b>J#MZtHyIjZq1$?Dew$eXG{vgr> zj2|4>nM*20)1-0Y^=qSx7pa?(DKg%Qek-OiTZP64=F>=?gZGX5 z`@sS7BmB>?4vZJ~?2vyycmB^UQRK9%UKYgs|LlDWTvKP7zDGDA`7q>0Am2ygtjhdbJW!Nyx!_ru!!M!yf6h)Bku#=7!ZS{g{yZqZ_E7Z^c7|#OJzxy4xGaye? zH3Oc}->;Ve-Oeuu2_sD^gL>o*Fx-vV?;v+mMo1hO)@&;jb?k*01DYlb9P@}7$4E{& zb@kAXO+6PgOMVaC559uH#yJKex}`VMDJ} z+)uqsJJPwhvkqMJW=1x)4!Vo`gNNKc^FQscm*CSNd3F}PnWvqR%+v&+1O!0;njMoj zCg8IaVl=yXuqs^BIHpvMDLxEZO+f#7FiR`8C-T3k=#BRAia-l=o&Rf{trU11+<>^4XKmtr2YGcPlEWyYi#p=d`H@$gm=V*GV5FJ89$*w6l5S z3Y1a6*xPwO|7QCXYdC2g*ov6pq{XT`UE@0XG}!L52ZUijX1L2MmzN>XNe!2h`%--q z5f&n8&<(-jJfvl_FH9zm68rA)jz?&3SQBn_;7FU9aO;iWL_t(wp0rV3P1ghr!#+Pw zUc|p3MlPj(1(w_AiI85YMsUi`iXUUv2(IuL^Qc`9eUFqxyk=N!V*glqmz;3m%A@N> z>&FF(Jx7riAQy@Ezp|f}UnNNW+7P$$KAt_&WXMQQsalGp^$cy+Lyk&52 zX;?CY%(6N30MJBsF{Qjz-&2xO8r6XD**tPbmLbj49fPI8fucqGpU3_~qqh`9!9uy+vC!)F9vNHYS?^9E%# zGpoflGw%lH&T9A5bIooKFU7WxWFP{3cnhpqa8s4*cIj?t0=W=$I}EwMDi@Z!4l8=W z`s@f9a=SCD4m9%iC=Pq;`Un#wNnxLO6_9R{IJbnUa>bvCgN*o}22@@EF5K*pg8^Ww z^leEeKxZ$W%?Qb%Z;Dp~hs(iPH8U>+b@L7eq$!unn}j1yNrGawK{8)K8w2i0#)j*#V8~hy>PMj9Ssn5q$ ztTbhQ;S&FGVC!dwC#GE%$D}gJ%;yRqHQgY|Rkq7^&jrk;2gbU>%<8$g_lMedi#B!{ zvvgrJ_)=pZwd?UO$%S0guGgmo8fN-o^{RU1-fiB%9#*Qj!d5IqvArC+$weznV>G8d zK3ZUCwY%=2-?>c&OgI8Wj}yB>LqGV1Y=-xYVd3)oL)8-U;dseS2euI9Mw_^u6bt3( z+o(tk2P2*B8n$MpVgH0ZsnALWtz&2*j;WSMKr3s`++?PcL_HfHlBO&pwL-8TEj$#~ z>Y_h)f8wUfiV2z!ARcbTtoX@~f7JZall;ct?=Cbz#qDlGGTF<`$#&ojd8-kK>M0g# zHtVR!PLQ}y;B^JpgVgj@`Tmd|MMn^pc$^XUDDH%{1Yy%*YDh=Wk$_wCTY}D#LATK} zX-d@a94*nrdr-EzAA})B$7xf4F}qQ5S*II;PvO|;pyP43&jaw=@hR&}la&~Ih_*+M z9uJ}?Ty@GSY>*j{B)%66yV zh7~KFhY653FC6meW{SnlbQTC=K}#jCBPiatK)6Ga1#2dnpd@Ce-}>1Ti)c>-Q;%Ah zC!Co8_>QJN`0n!tOjXSN%Sy7#fj8+N8KL9=#qOg>87xUcEbZ=Qck^mtc4M3&uq^4F zcUQ61V})~>UwL?@q#L>l3KiRUz4Y4Gw$Ymd)&dL2!O$$PHgzuVY$!H-SJOlCDA8r% z7MIunt#HGFn1CaYMSuST^0zI@(I@)Ifb~^chB>l19wspN<2_>D)fuPshWgrn4E|-B z0XHiRlMPV zlE5@&H#Bul%1VVZ$Q4QB&AL6#X##&-WMNxuR>C5T3wFg4h<+ZOm6$27OHIiHOcK2lFgtAtaro?M`Dx#JL))j0?4qPoL=sYvQ_po6I?(%j$dInJaYrF z&-=P!7zf@)gNrhTLjXv@sV#< z|LpX4T7P@_``EySUde6v8x@%9dmkFuGF68KMbP7TQo6ynQ#iz=u(kHAxUmjEcEisU z^8j-G&m&cG16YoI@B0_YI&SWW1H%X;%*McI8^r?YCHCvyYIaB=5hgcHGD)8B&}-unIF!6iLni` zAADw%EFQ9^GT4EARx_2syS!6)jfx71cE%xonhIMB`W3yQP)S2q1?kvsqeVM;O>$(v zwRZ{M=(JA2wv0Y%*H<^`+%mPnf@njYO(6%K8VhpP$buZF*anImp(10PwTdCnlcF15 zb+Q%Cy}^C+t3z)Eekf}Xs+?c$T2C+c>s?gt+Dw-LQO9z>Jzk~s8Cf&EKODOy&n>#e zEcZJKG|P7&m%SX~vn?Kdv)a_R1(ov~CwfUf3x7Z~bwLu8AFefE&+zFn=9zAoK~E7p5t5!rIhlczyJT z-esg`63U$>_~B0YL$mZj^KQU+kqU=xR9P7K%zg{ImrBLwdAh1cEOYAex--AbUxWGD zH08vi>QfSkQTYUkj+yz(^EshIK5gnw zdVsgbYk)~t-6OqFHk_tRh)8j>HR3G=^y60wxN2jFa{@G#{o5&@m`ZJODQ$7!4UicT z&imoTUTdL1uSNaQ_uC+u*s5+*r>K^RZhG%@FXY2sIdr$UKfEZQO}$=mRazBP=y?;E za|cK?yQEWi)NL<6MHcUt@SGO%jkStfUb%kTp0PEwbcnD8zHAIF-Ed-$t#4VgWp?1* zkeMxWf@Gabwkl`t%4hclZt?GiC=0SIL**^4Npi3DE%8bKS@`A|s}`gw8>F}Rn$7Bs zu7|vfK?2+^z!*EJ69viGG3;^dpSkjklVLf^|Kj)WkthdVj=-ND<6tLJY&=C)QIUE_ zNmHw+<7J5wcpKdt>6}Sb&m5f?iza#W=1F|zw9-t?{4A%&nx+%S^}V1{ZY+Q{P;3%K z5~xTdW6)Rd_eqmOl9@r0!%N^nL~0xMabOXMZ10Vj{u+qaIa0N9hLe zIW9}PnWH`}!AHp+ua@9W;h@_I;VE$%ILkwBn^j*d{v1jUw!0U&9}GwcIxM{$`dM&% zc$+#(gu>woyiVa(ue*v`dId>O=NX_Ta9R$gx z)0Fs=KJi)SK2daVFEKC&12g9>1H(9Xaayvd)UE&Zt|{w-!=lLykHZGvUV61V1_iq? z)r@Nr7G~Eh7;!qFDu9@NozDYrOgkU+uM&(nwSX&&Eo3dqM8Ou9Y!^GTD_fKE8E1NJ zcZw5GrfO2c-jW*JUB%zO-a^)LO9D9Xo(Lpn#yH366uX5Yo2f{>Uy~rgNX;4WYnJ5s z#d$1QEGblDmCyF`Cj$eflVlMlOdpzr_ZdAaT7nE#BQCjb zH!0v|H5_MShW^taVUoqsdw92(fSn2~{3S5e7*is~YI8F*ePpa-q+hwcW zHCf==b_SP*X2WZq1gZ8irdiA8YY~3Tcq)%SZ5_x0Qr_BYO(f2N@na?u2Pxyg%7Crm zNG+00Yp@({mnbzt*Hn+uMItV=Z>FsCU ze8p4_&y0?oUj8?ql@0Dl&Jj)a^ z4!i9!gXPP>GS#5iF{?rNKrrHG{O)0$1Xe{1MPy2l9CeQr?|J~$lJ%gbh-a~@Tp7@o9plPurPvgTY@#AJ z@s^3uy?_KWu)^y0)on_Uj_&e8sb5r_-Jz-uf)jD_!y#x*t@T1InY4**r!5-Ux|^to z%oC>hATDWU2XpaYx+rK%6^}@Bd4q0CYy}=m z@L=L7J^rnYu)%G?X^m-UGHrf$SRchqJ9|GcxR>~M(Q90*fDj(@%Qb@JX(LWZHC{LU zycoNPQ9pkDv|e%4{Hwe&?_-lty$RDf5s26b1&oZGP~;*Nndfm`FswK@t1CE}DI@EH%DwYE(o_&m$@3T_xFznD4>1R! z2ff^Ti^~>|D|3Ma6Iok&S7sHp!`{dUJ5Sl@?fku}S-bE-i?eW%N`f0>z zU|z+-(y&IC9N=}`CA!Evq0DgE%kLB-TW~C8-j+=SdXG0xoUWRMpLyurksW{2bhpIC zwQ*non%P?+EnPL8B8wH}Ny@!Xd7t!e_RkK*4Mf&yhZZC`qbTBJLyC8;; z4leSDla_uEjMWd(qBIpM*Ix9y;oYym(qwxC2_|UAOyOb{5msD?|INiMa}Ad1-S@gH z$xdf-)ac8;k77YDwFLMqfatGH4SW;d?|!}h~xf@d+Cz&f4i7!s5CfEuFXqR#yGEVPM;m)e3jm@XbDO) zpHSxW@u~hW^WR6#|4Q3_fjR$pvk#d60)M57|I6N7d)t&b++qD5GhE;?0aY`5gx$;z zx?*7uQ1s}FTJFq|f-jpMn+4jgc~VYE_Srq~Ms#m3#ukd7 zVB;nTJ5xty0`JY>&2ZX0>_R^oSTf%LvDZIYP(X^gMS~sKB|c>YzB-BpANc?kiNweo ze6v)!eivNJK>uWm$8nN3N0R`VgbUK`?#D@CsOGviv@}6=XS)nIGm!u>OSRPta28tU zbHKlVU`urZPv26l!N{_HtAzp?Y08r6=Xu-trC|@eF`@`{Tut%^emfSGdK68#gmZKj zlPN-e|K^`G|MSZ~{OZsDC0#+WA5bJ}0%$RPt{MX zBsHR4ww3`QxU6o8fufNVdIpwx{}c&l$|iiygUZ zNFF!y;=r|3&|yC&*|eWx_fljJ6^X0Z=cLSU<+D+N<>7mIM+K?C3$)TX=U?Amx8&{h z)Z52jzrOhGO`skME%)*?_j)?jH(fFWxXA`p08C%@fJ9e95JsVsnHG;_st1C-{0EX# z;!NM|yp7Y>N-_j!X3i;wY(<#K4Lqi{W4!ma0nxXXZFt|Bq8SHXo6Ho==u0)xQ-N&c zF8{(%RKqC^>sNG;Hjfe%`1L{C><>+z zX6LCj{eUf>YSSm}``)}4J5F3^YG&fk^Q=(zv=&9mij8eK{_4L{+VKNTK7W?JZArb1w&EBf+j z%8y(Ucy&Jc6#kjyFQxvUX~}0f;Sa=$2RVH;9d})B_|a+l;z|Rms-UKQHpri;x{|8hD{Rc=jBuEaEgZ@|_fVyRvOi5;v!gBqpK^th7 zKgccYg1+>7^a}Ne6Dn`snV&Av)mOL6PEX(c%!m^j{wDeLz|;_wcLCNV*oWc+F{-P9 zBTntITk|!Cr4TQHKA;{&q6;Jm>mmBq<+c9B6rkMD(pbxcBIWlWZDTn`Edkc(*V6{C z7DJ5_upDte6Z=Q>6qngZ_okD&5A^V>`dP#Foat2yXmkBQUf)q3c?W^fkJdm7^ z3CN=HS*+zxW-zdEa2Di{)9I~VF#)Baz+Rd|FOL|M>7uq+!jMBBg{rs=$P`1^jDyYFrp-|G%N@D zTH4qVr*0^wK_y8pi|ts1SPs7lFS}!ZoW;kPMZdlDEd!vA{`&l9O0geO@Q;nm`KM%N)wFbwF8jz2S)8H*wKE3D8D8ETAXeZN@ngxo zdA$%L`TIM6yCj`}pJM4haO{2DCt>W<_B-Y@362}t2oSz3wuZjhaW??O41Kdc(k_Fbf4i(w)kiRvP)nyvs>N%4yLmvVsfd}I9a1-W zlx7?;U~6JblbL_`*1VnjDgWOaP0_)nw)d-t8+{Ub^Q@6zEE(4zZcqyrBTrhzTHi0c zflQSb&t`{AqVaH4+zby6b-iM6J3jyBw|+$qI$)MQeBXY+r&Am22(<#KjAreT%ha)63yF}Yqj4WzU z6%VUedg(0HwRxTHP@qz$qfBgc(d3EpLhw8(!ZxJ@VaW4c6udyMFz9 zvc-X|4bZ)esbk5fSa4IasmKokTfc+yt=;oS{B}uGWHo|lY7N`XT$!Kd+dY59{N4V* zv<$m2Wc=PJFP5xQk6zQODH3G#Y=C~#sJ31g(0-#AjjL8pKt$hE|RKvoud}ZJf z=;9mCF>zQu6awPlV-}mW6uX8Zaa1IR?OT+ZUgj)FDqfWrDfT}1|L1+o+?zKXYemm_ zSyN!^zgLLYGz30oD9k8cBKae$8PJ2hS7T_kcs}kcHApqI=UW`_RAIpVOvh%zT*~T4GMP z?rKba#Dw1d_YY$K<>v--GCQ+lfV7R*U*y0L889*@S1DFYkxnYINse_jZQggnaC-t7 zn+xLj5GZcQ^2+|zr*5qa+mwT{1gLq?hGJiCF|VD~T$S&K#b$5Br*1ke*EJq^-!V^Y z0co19DFc|H+Us5kwt?WKblCh9it{8*(@~&Y8w&VY1DV2hSs#d}jW{ijNOZ}70-iRa z(f5($(6?N$jij0$R^$L<2qqSRb`}0K$uVgJydeZSO%S3uH2f0#Rq>wY5=8a@6juO_ z7xOc3r`+dDP6imDR?=G>O>!J~O9X{DV^*oX6bppS#aO0```qIW$s#>#6|jd^QIZ^H#|w#r>HGK+H|5EzffU$b;|3ziR3fJ1juV zIKZ)@;{*fv;HPvmv7*Bg8+jgFXT=gka2?YY_j5x;%oo?+C!>pvQ^pChay-Arfipf? zMhnm;irqkwBw%tEt>NptLo|oyVd7`IdzEW%=x*2T&_BGu>%Co@<*+%8)t*grdC8o0rY zbfQbUScB?}20BcP;6uDl=d#EO=5+Ekoe+fSW}-xSzNsLAr1xR-M8M>orfN~<(0k=) zNF%59P_x}AC|5z+_m@C8@YWwg*sOX6OioBQ zljj;s;h*jNMny+Zqe6pwC*5+@!fTRm4bKkAnOi03hR|`L=eb3&s~o_&6D5?gX@dtu3OPU zELZxJ$qWHrCp?ASmFS@!2|Gh;QW>ob^#l@0eu!q2%WQ`VV|IEzGmO7tFPv6Zq&WG! zZH9;XIGtK54IQ@V}_ViHvdO-a6$fy@!gK<#^A1&+1l_#JLpw_Oo!o?i?FKCqq~TjGXyXJe>W|+&=Jb!%RW|4|Xke*nmw~>t5406wQd&=3 z(<`q7nJjdK^F-F#9?K_RCy zsaKODf=2OfRRwi!QHq;=ESL=dW2Idh|8e^FqD)hQDlRqQ4h$GGJX;BDrd!s6o%9ES z?S89B%v>}vh2jMEqI9iqk6QmaNV>g>=|1{HS);I)t`?(8!QSV#xExS53Lp4&lBj2+ zMAhOtpPHG~;!VIFFi7%z;k_V3is#$Zcic*N=RnNs-q*4Jx=p>4F5({+<d~KDDa;b#=x4PTn64mgrASug@i)Ig?dJ z>&JbH{emI`kWm(D6`NkbqD4Jj>4;Nd*d1B6e@2KVO{JB^2%r}XYWKm1?S}$4q$K;C zXtbePplHbhsrfCQAM#4P@B%g`YV^-uosZ>FwfU|!|3Y05^0!>%BlJYmNT z=q?W}V6Dk{Ij~&=*7-3D;ckj8phzBs8$ea1j8{zWaaj{y<$4r4d57E@6v!T3A-UxR z0cie7G8o}{3r`OV>U=l)tr2jvMdoaQ z@p)TVEaMSv^DcjO)+?sUOb%Q2XGStAS2RctlS80I(M!iT_wx(zPMFA>~vK+=$S8waZdNm5^W62mExV3zMl7T?TeSb zp8n5yFSn|zT(9u+!?w(-pIxT7Wy>U}Ex_<_{jv=Za1>5vkaGE@`!{A87Q2C8$XAeU z4jgN$HL@Qi6k9}*d|*ccYO0Ko>YxKE>~{e!NNjAmBL?3ZTIck7?hu|tc2X^^xxyS2 ztaHCFZIWZ97i`I3)1<+cgP7R|7v}SZ+Pg$1Eaa^$(a<#kK3k=FvveCZNAi{`5| ztVA*ZRE+rB*2=Uq(j|8l1;9rEEq;#|$C&VU;7jW>@q@>G3pjlTru;km;eR_bn?RA(ROBcD8mv+%zjc!AQ4R{Crdz-2Jnmy2_nKvx823@Tp5%my zsrl2VMJPuxar~E+B=e~WDb^SvsF-3wta}#~iF>2dveU8(m%R&PMO8DGMP%}i@M`I* znOfNp9|(qdM+xR~E9akw>g6iI05saw@EQg2>}Ft-ta3%=Xe%v>I>*(Hj&@F6f; z_5h$|2z!EMu>;orE%@hk)({yvZd`$wAu`J0?UMDvt#W&qgvU;irLcJHAa=Mo`;)9dQ&>2xi)&`0JCzqfY3bH{_}=Tg5j5*u zN)`+&(9yNNCWqsPZs|mGtCHaOON?_nH`n8vEjhoOVQ@WE@o_a|>r>-;958Y{3Mdxz zvU8}&=G@jnVIU_DK3+4kgT5#6ltW3>F4hOozWQQVm5(o zV^%mvi4KHo6?=ScN|8LRh>v$R)7^G)IdD=fN0%AbV;p4I;*{XiSADzM znd~$Apz|m;hk`XcvY&}(6C|SaerNB@(1ahoc^sUKe?5Er0sGl;Y@37-~a*E~)Crw)}pAuKZOoL|5OXsl{uw35pf<^pXL(M^!;*`z{#Xx z3KIbvr$0Q0N3H5Ua#>nW<2^l4V{zJJL_EM74DJ+e=X1_xk81_2{vwQLg~t6!+)kdI zu`B7H4d&+f;-J?^vopD6WNtbs_7VkpBawUIgI`8UH~uKS;9q7OolYknTSBpHAYPWrtPATS-MoH&t32JeP~1TJ}Rlqhoj3dO@tcNP6$@){I(!(iDF{pFM-#kb1!8tLbkJx(_cIM3TmIJcH6;D1@qwUBeoP{}h+={4CYOpFAeW?3@Ld@Y zR|Qpp4&7DgA>SO(DBL*xFu&Mqn{2FM1A9KX3(_;_Sorkewdbv42~6_^dpa2 z0*OZ+nuHV7%;wik3)UTWe{DwCBv;f4od-EkL51rSE}++6z9elF?hn5ix|(E(3dLvu z5?T9v*~gtEOQ!qbsx|Hqb~(vy>7$mPe)fgYUDLJZRI=Zhv>BP^V-#CYk;A|j1PQpt zfHj~WzY*$%J_ol%$Iya}@oBz|G9W$7~THf zID&lH$LA=3t>8Q>Qp1@vk&RFD%kumpkg7KK= z?0sk+#`&???s9{fIrhEpUnJ|eIkFDSlm%P?V;tFS6bq^8EtrY^LXjr{)fDKI1Keb( zwt8*xjDPN+f0r2hgGZe375OO!NjsG76d0)3CL>tP2|f=^pS{sDWy`z|e`MMX&Lt@0 zz#uWBTzi>bt7h8E_M15YDL&_uxl1@?-ag?rb&*L8_Wm0_C61>%8|BG9p z`L<8KXI3wr>Ra~pOS&tK3oG1L2(lrGR0P*yf8$>#Yg1=|OgtVn)6m95Ux2JZLJ*FV zYs?h^8Seg7Fn>0zc2I;~@vz4N&~rE7!`f4<=37 zLH2`2Ep7&q`7JPZ4#YPhDVoew2kB~N1lQ(qTI6hw!pCCNa!|7JZw|g~z}1htP7%`2 z4OfmEbtFkfty&*UWuvB&2UH|>#irBu=mg$!L5(avpcJ;?IXqMeOjE9OuH5fzx~+dwzbSLGj(1PSaUA;O)eTBF+HaWJ5hZt>{n=lGO*7fXidD%V!;Tv&Ej zyPs4QD=HU4PnNt^ps(%1_jl%}$l)B6mhKW=_RN%QRyl0P*0sK}Ow)Lu!v=TE2=(+U z+JKk>`}bO4`>!`Ywpzn1|31T=mY5n&1KQ|jP7B_YD(9~4tidq3-5WSZHgPjd4(vvk z87+7@6bl7;nb^Iu!8aqMCIB2E9#HL+h9O09k0Q}UgUZK92yj5vD^7ql?Fzq|PLbvF zk*Do)Sc4L8TBYc=Gu5>vxH7vc*$6}94b$K1v*@dCJa)VG0R3w+RB(W!v%weGlAuAx z25m;;eO`|}XW1k_b_ly%v;jMP590mQZwXBm=44Tf!h_Txq?VrP3Z z25f9f3OGwPa>IrLd%PgYGiKkOO|jc4k^%d6VTo6#upV~UZEENU5~eA!oq90%fnP6u zG3-)UHb0M_<=qW*Te+$fPpvSU-mm~0D0ll;x#InH_F}jJWfU;@`nYMf{#xAsepoLy z03|N<@^48zH=sDM5dq$rF+kZ$u_+YUL`AL-$`b+WQJ(0QOS^0naDBExADV98tzVm} z{V#yLS2EKg*elDKyTRAaz%WeV1n^<~_k!%kUEK!Xk$KiOXhvd?UQzB-WAz>?dR`V!<**7dloai zXbXG|BT&{S8|MJ|e~7iB!{@=0pqbyI~I_Do1L1gcS=v5l4xV8T`Md zQL%nDK9S7e|8>NnqYtGi6QKSQH$gdcvRQ7HEcwXTg`+12To}4E`2LVGbKZiTG?sucu+TYJdLBTV#U+ zXBbM17N8vz3#p|veasgDrP~H!95+qbNuzEh0tC4EKoS=Nbs%!O%rhedecghI4*KL0-knEbl_pB9tb4h)mEMj_l~ zQ`upP+@m4~gFB!$`ywxv(rq<)GvnR$cVi^?MsTU>GsUgxO+m^0Bm82zLHIy`!tMI7UXE;oFXkQ4 zs5J@kDeRxXXLj??L$4%8S`+7%Fq*Z#%Yk=sn+$J(T0U8mFqyx1AyR8?4yYERMj4Wj zRMW=*-erELLxY?v{gfvB!$)S+L=0_4r@^^NrFDiRSO)}_;AADl$|Fo zeElDbPQQc61%~S~A85*PZc+@0W1ZX7+sR+6X4J}Rb?RUxyoI@8qyKeNssuSrjL29n?3CGkxijzYXChDluHa@o=%T z>}aA8;bv%JUypj_RRfg%)bpn+|P&xGJ>ARl+ z?WaC@imY}vY~&G?>x0$<@G5|hCLstfq^aPJ21S2Z#X?Mo7cDMSXg*e^c$Rt;DG-C` zkY5mY3X_>=Q4}h|Bk|2v-fdvy)6O{Uf;5zAN(>fhqC~x7=Q8>*z~j8)`XcIc)w>R%v?9CN{IkN zVF;a*-Y!d*WJxhFq&dXT?4GB^mBK3pgjt4j4c;Tpv<*S>O z02&37rLPua%RGR_2`Eqb&bleFrkfy#b#%>aYt{#K2Dd4fyIm7PN9G!j1PII`LeT%k zONj#g1otBowB_?_@drN6jT6AeIHTD96gO=A<(Y54XqqB_N_wHpB+qqb0e8Wx-8CmY zkAxkFtXn)eh%ugPi|}Lo$fUe;{^mblGY$PathLBBT7pt3_CpFj*GQCKDiZ8d-h}nw zrg%wqXl`h&>LS^uf;Ix6@rdz2%|levSvyG;GB9%%&G(pe=jaycj}vO{AZ-8 zW~v!!QZ0RmuazBBZFX7B?3Ir=bu-=UJ>D?9f({6)!T)uLn}Q%paX0ulbn%~@-UNjS z2|$K@5>%4D@NSoNGizP@W_1G#?ue5XGyy8-V<_v^d?>hH&GgYto`_TxD(M7S0nHVDD`=-AGpnA14$Dn1bTfU6j?b!cEe35A&8lbe=bZLfKYN?6rixwZ zp$&~^bLd1DE3Kr(hI3LzkGXf z`HySqWa`dhsFiYqT(G!7(al_bsr2V7sAZI9-OCpjmx|Mrb>T_&kz#vfjyU22&BP;P z*yF@!{qG1xh85_o(rJGoYaBTE22{>t)|4$2yBTu{k=F(5fkwDp26amQ5JQ8udFU|4 zBtKT0q|;S1b}37P48q!5sY&uwxKFq}ruWe?xj86JKRxt2(%CW~2g7>Bx4;q<(nF$rJ}S(13LiXohM*cPIwY09HMN4>Pd zGj=2V#`)onlrc`kqZ4eiGn|k@WuK^dU`jCLut+gO8Pp)vGWELnyao^2W%?3-kl@Rh zHc4jV%h|IS6<8_N6w8KPp3*LMC5NTX`(M+9&7 z6SMTgH#>lxE^zzs9oDfgT>69OOr4HLbHk1s%>h;5aUGFz#Wwc32ey5rDXW9lc<0dP z7Uj?z-5crEqEi81Fm*l;yn8}X(sL36CYEA(bWyPrbRQYkJeIYaLcad2!6*H}u3!J2 zY;jGbD`m6A9J-3t%R{#G}somyA8=?FSm%R120Og zMwY0aVh>ZKj*8q#HwjC9(&^=anj-=E^iA*aT-u)(o;h6dLxl%ihkKH{`$S~`t&DI@;n@?=JX`mtG1kZ{_g&>yQz@<16@ zb6OW~jTTjf9uq9BkX~o9RoCVfE3zk?q_P@GM%HTFH;=)K`AffE_yb}}yX&wU9y3I` zdVPmZArz8?_0YQIrzdo)KYAz#f;>qh|17jwJiH9HVnohA>D6*M~Tt_zW*MbuO35*>oasw;RTI7@Ygx6h?kmB2%7r4(Mt>;jTSx9Fspc1g3Kxe#La zv<+RAiD7!&wF6Gt4&0nDP2E!cN1V(69r11Pn6U(OeMM`9#e}f40jD z>jA(WDGAhiW1a_3#xbjSIwx1((^E}f54jow6bR53crB!fheRFSOtZLAuEEjpA~Lxk z@5=wTEhhX>^&MRva}2<7lBo8QEN=TG2X68?WCXJ^iUr!2BG@OnuTk9%M&9=}^^jYG z;(G7_zwKlhZyWCxX`riv^lR7jmgSz;#>9KP@^FPGz=f?vJou-yf3g0BjofQjBgvfpSiC3Zcl3$T08Xpac zlVIh(rraA3vs8^VzBR=7+LX!6X7++BQo-OWUb5Z<*$flLFqt3{9zMox$=ds)J&Gv? z%cI)9_b=pu6Q=SvS z54g3dtAezl=h!RNHKM3sg>yy_6^w>#r6DI7505CQSkG!+toEr_r>Dy1WYRTKq z2J@lTzNI0{xH&EkY(7BzYK(uJM6n4JS&fd%d0sr55uz74(xARkI`HjGQcKv_sj?73 zV;`B6<83uJ*Yii6+1XCfB!(L%9JtIHxRl3WBAH^>QDiN!_Cxr7$o+B%Iv9Q7(1%(p z12Q_V6QPKVSl3BjV=Tl4Sd>YD!LJgu5?CX&@LuJfMM8j?;62rk{Ys3RkjG@^;oobmDfSmxF!m2kzBuy zpwy69XPe>kad?^tb{{vMt*&vxF7;n8FO*p0o;a|}YQ{a$SJpxd1-T5c&;TjjHN{?) zsy@;LZL^aCo-Zex1B$+UW-<<5IMV&DHMnqGqe#r);(%&^R6}3RJMXmcl9>alW;#QV zCqahD6xnY7UizJPCKDMPa51T#tLBBhrPgT0IIz7iqZOm8Q3AP5Nm@inU^jD*J~V3blU>}?CE$Pbuk z!C}{})3%d(y22+%RHu8_riOY)j|~g(N}4iPx&@S+Z@YC0`{-O=4m|**`FJI6`h96V zogV_Syl(x{VW(W)QG#QnDZ9Mx*`e9Xa@0I(9mbroIeNIE=JMB0e)Lr*2VFIH*f-uI zNs#GQ3(C361<4E&<=~kz{}1P`^}Vv71xj1NMr@9VM+fAF z5q?fzvuVSZ!{&d@bfV?V?G$3cWwkg_fK3~#-OoT6HQ!G=d-7xImZIX(d9&{sPNBZ51vhXg&rxA9n8F)A!6T`^E=k0aG4I|TT{@|2p$jxCFDlD$X+ZG)F6s*7onFN6QLf`Y~7fIGkgj<;u$l-Z5q1YW^{yMi|LC8nMp9p*TRV_FyP z3v5)Bifu8mmd%vKj-viXYo^HXET@@r++y?-rB~>$`gR#P$x+%1DGz|8zFnr%&DE#) zCVRtdIb3akK@LN3g6q`8(#7wp46Dtt$iJ;8CC=j!A2@7JBWW~Rfeun^4Mi$p1^V_o zQLnu7?mNA&qU>4@boHzZ>4or0I=u`O-dkNVB{_5PN;P zxMp>bTwaj^Twj-~q=lCS!Unyt_1;F^TihYO!W6jnF_pY(_n3ezx{6%~)sP@w$^)$q zI%jSc-1m-kxkJl=#nBex?Hqt)daHS09sXg*N?!wr@<;xqoMbq#%L3%EV_cTq6bnt# zc~BWIA5aXr^^+Pe?8+(P7mA1G6-Wy_q8L1ztvV8r>X{_yg`S<0&!;J)MEPMud@ac1 zwy95$99}P-#Jm4|fwVk4is>gep&+Zl|FE|$Rz=fUu)22IN)$Zu`;4A2507yS3N>{9 zWq(sDSuAD$e?VZ*bNj(LIvD{P{bzknP`wV zR99J`rK2JYx{*kIF-dSm$Z>B4b0Z6NqH)?2Bj z*aH;VM@4FdRWl1nH&9nxVh;FsGyTCsZXdG!!aV?sHucBieBQCYy>q+0c5#j zYa4LY=e&Hy)c~M(HK+efV#nJBIk4@>HiF5A6q`(ubyVaIm2Q`;r-@7li5D#>9JDCA zm6d+CK>p<(T}0wNY&5qP%>*X^STv$-ZZf|JF@F{LdK#e7{`{G@$Odk<#epkhfsA0x zPG|?kLXkoml3?`vU`P68dX2~DkO$mAI>euc_mNoFd@7%67H{V#GpG37O8xaJS5!#X z(x2HXMr;EU^P~K*`n5LGvI8^O{oAjVF1{Xkm#q5cCJF-$86nAxPH|ol zCAusvRulzB1>SzGa`8KV)8Dm&f>!=nxsdlSomDye%7S{j=*41}HrbU0ZR!rN>Ki2o z{5r^H)ymiE7WdBHA=SeEHJh!Ry6hiKlFOi+=|>PBuy!e}0@;Gfxwp zf-$pU{?Pm<{KPBfzr*8gKpi%s^@9gfdcNv;!5sQSsF+2K(h^k3*(*Cdqs;XNsO(e+ z^}q^b4f>u4gGX<$9I%f*YtJ*B06XQyikZm~hK0!QgVnppzNcm(I&HKNHBjskiX5aO z@60Im$)Z!4zFEmksS6}cnPXbGWppdS@g!~mxWvW z_XX6_)i%jQa{Ndwn`VxK+TeCh_J!h8`@=j8_T?vaThEa7Q@x=j=ETx z%6ZZxT811mCg4efTXEeya(?6f-3C*Zzen{H92g~LnEyuAuEExaj5y`lXWj{GlmSxGA()#pw*{mGv!M7kDnc)qt=SJnl#Q}0u3N*i zL(cQ67L7RV@!9HSQ>Vu&;F#da3Y<&Q%O4|X&mM8(xJTlu8 zJz*XttY$wG+E@>M$zGC!T6M=QKU(w~&1}fX_ zvUCalMTt2i61^ba;JZUr9aOluUl=Xw6BhXN(ov#bG48Ok=vtYbS<|fn) zPrKSQtNfH?hs-3$cc_we)hfDFIPQq>1k%Ew*b8dfVy(hv|B^sk+Lb3v1y^)DVK|#z z{{F1rubyTwBv*fT;e8Uv&5$^7{6ELYkZh({*v_rjO9SbI4soXHin>NnGxIY=hxh_r8{VV7HZPhT zcDgRe3c4%LRQ0F}LPzg8rs$qu5L)MRoFv*^#dyLqZ^f)y@+c>-l^Xuhu3?D*Cz-;H z9VEelaguL@lQfD2`rr>?zcb@VK%06%eAMSt5dOON@+E05@6+j3uG#(x5i6a~LuJ7! z??z#)bFQj{x6{4B-*!8f$Ib_LfIN0Mn_c7vknfgFeaBRpl8dkEz%dpxnw9B(P4Zsm zI&(r%cDQE+d<=m10vUvWbcWc^&97UO#)80C_+U@p}~8WL2(l9$C_ikT@ovy6%D8 zpvca)Qm}QhEDBe*)P}K;F;E=m(7Z|c*SFoSnzr+BF&_?G^Ja!CBZGI4)Ctm5SD;%B z1NY^?W?sMGINK6BN$bv-`LWPyjCo>m4kd5qZhOICHGXleIhE{pCT&Jv#A6g&Pm#k^ zWQ9+xs68l~ULRKAxECXty(X)Tg=0`_d5^9CF15W@)3Dy~6bs@6Im%Dxr z{r@YSSBGS{te79`oTk+C$|JWCa8!7VUe^zX%74A{Vx2z-Ttbwb^BKmRmok$>)SCRESelrnn3PP*J9Sc$iJNvT&hFs@Yd(dSJL$!0sq-i6!UbLiW=BLO+I zzFyB-NZA%gW31+c5u|^R8;;ISd#BfwNTtYVx-uy?jUuU3{aX=Z=syF4(}H5EKpEjsx`0h0Tz(wf1AS zN!&;~>>g^S(~2b1q@crAyO~Mzm63QhndxUB0}LHIo5?X1R3S8ha&|qW`udp%Of(b~ z=>JTHMbjA`u4c_jYi04{oXpyk_CxitLW8IB;+ai{$SQ70K?e@Q?KD~_Qz&*5MK*w7 z4=j6U#Ta3^%3Hd$N#3IDg)^}r!JY#`$WQ}Y?u`xB%jHdzvO40ta)r`lj? zL5DRRW(du5MV)j#eU?r4Su?HFr$znX*Ihqb_44i4PJCzWDltjPvFfZ!Et4!>)f>VQ+N}&Qt%TD{8Spn9z-!tO)g;-0pkUSP=*ILCjc@ zZU*;8dp%AQC~HeowJ8T>SZRPN8-3E`5UeZ64rvxoS{$Fd6`2q$Ci2iTv8!{fX(4pn z2oW*ULfA*|g;-1e;x48V*lC9%?n|@iM5dN*R6Ov$HZRxfEYqe&`QuLdKsbsSH~VXa zTRi*uwNN-|cmHLg7H1+;&*hm9JOW=c-RnLj`xrA@c9i2me%jH_WK>O$XSK4Uq|CoX zsY!?WYiwcfrPFw2{<*W-{nDWoeNrsS#4y5TwkLM)mDNp(uX=M}D|=H~u{bK?qF<^~ zQw;~1q*2~W7kf1-@SWYi|HhxZtx7Ue{G)Gh!J^{Pzj}Wk zSMGM%!5K?O|J%;TrU>kU$&;X}#VGlnT_tWTpN=p6K$8ReOboZ_aM)PcKmG1k)4uA& zN_z6gH09CR8wJ(kF2JR(jAVl(SBaDs-Lr-!mC}OK=G8vn9Dh_!ke#aeBKBvLArA7u z{lQ5x>@-fY(t)ELTa2O|2^71UBC%8?@+Rfb(IOy1spDnxW9&o9H~tKaIctGk8*}&M z9hwKp$$yRbPtzvhr^FpK(^8D3@%UF?6_25?YW{=ZyTOy}k>KQ802f&HCJ!c~Kk2t` z99wOTOYY`d%wdrC+6Al;|qCM{911S2MRJpM1?iOiVr~NX6|vQw=M}@|A%{ z$jEprhXY5rwi>M*YbkaOMdGN)UU?h{Yy!fn~N}^!quPujZGI z-~V^m5>qSTuwipE-+LroT;sYlU7}^`p{VUF?2RzKIjO3!`q{K>E%-n9-gEK>ezf3! zUNEKD<1&B)2SLq{?_uaO7Ya{Sh|lx3di1Eeg6n|}X;@JvX%u8Zt}TxFl$lh6Vj+-p z`{~IvB)6yjuIP`31>{dnug@i)IdHMqs;OgJs_s+l7Ze#l(mW`>085b+UK;ixufjir zm(S}ILaZgM6xyaVJ&HsZY^2KvD*5W5OjT7V@G$Z=@s>fY``AAf4BmlGCq`THM7M-X z`vkGPLOR_)fm{t;?S4dcg*hnbX6nMT`PT?i98~~=J+@|P*16x8Vn_$~$IIQW3729P z_X5HD_sXMOmoDuTHp%mOHG*t@m$*Odv}ZM)C%PPxAV}bC@xtHt3jd4wh2kSFI&LKF zG|!t;7+M-u?!6QxvX8oGWrdzdr}PDzA>AWAruq#RK$%2o?E2RH$~4}VHo5vg-((H& zbGtWij%?yq0pP%%O_`BrlS8qfe4GjWe=oihO=a?X4#k_VF zE4H_KZSalpZl(vssUezP=Bx~99|)tcqjyxwR>-L&r|LtYD6o&Q%rxAx=8I*DZF=htqec zx|t2C_Mklwcr&HG(Q{I?DARlk#d~~8!&;OVU^UbNSz9tw9aIFoYiY_+njBzN!J*UX zL_x1SPjX9$qUt+TyF<{9A&*WT1ab7KqV*o#v(l6~^evY*wvZF$mvzwMXtg*X(d$+I zj5iF>y0&ob9CFQp*T9uVYv30YJ3x_sD)K~NCjX4LCV{t0l<80Q%#mQmbp%5w8uZf3o+!)Qy6MU8+jn(XQPiB(B(gSM( z;#5%Bphf8z-W8^scQ61vIl(Dam0-lFZc(AQabdLV=~w`#9&-;rZVkbX122H(MrQ3x zf}KOJ-dJ+?*tKig`^Q*)(#Zw*v-h7WC5j z)c+&zP2if$@BDEe@d?Q{6LJwuo&gjS!61$th7phGblPsW-QDhPul;ws+e`Mpoozci z?PjLew1A>`a4DdIawrf%KsgjqZVymL1qYQ8FAzk>!4Xho5aIXvkT{Y+G!G;++Rm%P zljr!tJl}Ud-_Q4Ryx+T&WlNq)=%(=P>AO$LOtCOKY;Se`vClqprJq-17iU2i2$h(z zkR_aY2ztzQ^k6(-2V)#7d6ggw~&4k6_!afdO+fH`Pso~+--Rd^iPFXRX(Ir4exW8$nJwW9vsTG@a?^pLMI63K`ysnjdRS$Ykze7YjIgjnL-x5f=2F z8eor-Mpx3|uMfED(!^Dw0#T|e)pfwN8A2q@s*OM+Y8N|i>1D8)083-o<2hraR>q zW?x@gDZ0R9K}C0}N2?@OSx*;%Zu6c%xEWa2A+k6qE_X}k=+ePwPT)M^R`YU&W#ph= z3lO8)81QTW{i2PzErHDr(5IMPtNe_(x&2Rfi%8BZ7}Q^k-(@Bo@mydfE$QyRle+b)FStD z9>bDWML%hDz9H(NZ~S8ovszMbc~NI7T6u-iHtv%Ezh#T?zG>0H~;764Mq;T1-8&K zgk1^|!XlqIqHE`X=#O`yB0I3g>lkQDoziS~9yz&}E`*}+4j&!fe2j#7rYNQWB@@Gl z{fju+VPs{K=(}$j&B!1(<42^DoxAG5Db6bGfMZ+}Rnhwi@Tbq~x5 z=}yTj>Ar=9J*ruFJ_gHSNYwxmMRL$G5Z{JX;>|0Vs~A5Vtc z(s8{ssJC9I6K#9NBVvwTT6D0$E@wR5GU0pMz4rsV6T)%h3Nw6Nbq0j=Z%BIR z&2UGd9Lb^$cVtV}c-z4$IVzmo*L$BZ&n7!;z{NtE`zhz5_1B!C1|WigQ&2 zt8J@De?V)1ZV$bWn;w{mvb%bEuMiKhd>exw-M}fIs;W`RUr%FJ?ILw(YGay9vy=IfEeE-&pRGl zJWm7hbPneX>5#T4&WldT!I$)3<9Q;eS#I0cZSftkp9hOW+4KUdd2n3ka9*35`P*u< zyd=`kuNCO&(urph1ppMhPHkSV@<7^+FQJ z){Bby8)q~1&(@QXBlDqb`k9q+n|?lI{{wI12Kn86$70ED2Tmo|m~4iNDF#A<`5-d| z4Lb?S7HPlYDwJiW1h)D|Gxq`x$nVo5twQT%7yVnL&CrdeYmpvMU!Fhgblmf}Cj?(J z$uWY>M0z?!8Rl0HooY3l7U_U%lPa68f|{ic+Y=B@0yTC85j)g;ye#66lWC50;Gn&Q z=14t zEn1RI_e-N0gqW`HjgHwFbFPYFmHnie#)^pe;0EYOg8no~57A*Kb_^$0xmS2jrWVh8<*Am=G)Hh_X|1y6@ts8p4qrr|cNix56h7}?ibNiC^{ zmcwDE^P=6{L`5;3?^fX-=62St0a_UCGPoZi>q8D?teUl#mo<}9berJ(@*C=H9#1@yUGljdK@jNW zLX?e{2a+H6NnY@+1qSIMjJ)0Y;dM#0bAd9}qgkKPU4FvroJTzeBc?blC6cDacKS|_ zyoyQ9$oMbQCTH2rh{Lj~UjJqFAI#UMu}Mb6L!1K-#Mo6mGH;vTwDhE-6AXr$aMm0g(G02P}tP$QKdXeWG63fd!f#wJJ} zGl*qqfABAv$0k$V;<&*lOVurV>7hlSoCkd&^>mLc$30OIIiuL4L2h?(kukGk9h)A& zVw@-Kd+M|A=6)-{=(S`hGm}WcD^n-;k;$rANijz#atP=`RL41+`H{RtMbGQ2!4s+Q zuYuGe3in+J?w9t^yE&a2EZaFLELH32)${klCJ8eH&Dv@~k$d8tTFJI0+gzKqo1h;p zSJ@`GsptgV#Y%r@O`BWde1mC(P8M`m(vJ zc@Kl2aNnmm_!jp9#Q0Dwx>eCDDFvOz7>=%vj#Yl)aol%}YyI5gzHM%w%XAR2L+UaC zgz^F#fT-cV;66kLi-Q}S;a4PzhMX=s4>^S`+PVmj?t2^PA=7-)d`5zQdt5L2BP&%R za8Seei6U^%s|9uRHBA-`&m*i86-m4`vMLfMUgrkU$9>$ZA_yA#rv`Sq!X1_BUCgZo z8QhBkhvn)ZqD2p0_C}>f*%9?wNt*y^J3CRMVmFC%Zh^EF^vx@Kxks4chKF*5jH_iVMVEf?5bkv%TWxq9F zT`Y8)B7J1jzhU=CI_I3GmkYU?+1?V)viUwIVuT5}ORI%))=#u)p+wHHU ziHD}7l4>F^cPfG!tiGy_8_D{1)tU7YV>Cwom$#cpEW2oo1KW^%6C097F(90@9l8fx zuNV{8BbB~bIa3Us)@fqkwnBfiL)rzE53Q11VFuaEL(0~vF(Zx&^rD~KUHO_3A-Dhg z>W5@KJA^p!Mp|!zknI!$HS1eYH@pubD_9wWKO6m_2fB$9ChUYZ5h&^nN(8=!b8ulw z))s?BXNL{bRluH~VknlF_APHG@k7b)p3Ak4SL6U%!hJ5 zQimEMJ(9#gjCxHy2CMhdq{+)C;Ux!gnEDABGtybBln90gndKx>h>gTg} zJ@F%Uv*fTw`aJg7sXAfsIQh6}peRCa|5;%dLbq$-Ia$oeMWn}O3_}||+8^KcdX7!)R8V4?_-DLud1d55LNDLKv)4fr4fJ8H$ zki0uaszbWyTih^V1krT{op1)#dpl0}*cZ~oZhRYtJ;y+V*k0}D0o`fc|GNzK;`MIKN8Z8O+d3kR!QXB)!CviTTqk-2cE=Ks?f z!E-uf4v+LWFnHojG=Un`y4!O(r*2F$x3Vt=?GdktB0-u zh8*k#NfDO&RLw#1{AO)EsJSF6Qan$E#Cujiw@)i~$J~5=mtsgBAJVKXn0a|wta1(K zpgdttr}pXC@2&n`%*xf@YgqN0L`9q61_)4Hm*_dwbQcr=4SXB(l^ud)4yx>3{ zKZ{~uAD%{qmbh0-2R%-x!{@k~Y(NpuP4?*qqITq;s{)AN zO}d95uz2OIPu&|tx>m)2%RO1k%1lWI0o|%ZMcL14R-~%%QC+HXyGNp;@LM%2@K*4} z4MSlQKs+~%jv~MZ0>2;FQ_&lP-w{;fr9&0#Q-KMxVyNV`GDO zV}5F$NMhp$Iq>$}LfYvm3x`2D91|3nPS3Fit?7$qz&cR8XcXID|NdNFc(Bpde5Z78 zIw^i-Yz>%}Q62k7DF)b1E2z-D>OAS_iU~uh+prU!D+LbrA*a=TJy5l8Am;#PLg@9l z8BhnSR5AS!=oYH!G@nS1Ym$*Gu%+sh7bxrK<6fmw&^YG}D3F}laAz3sBi73?7GO}8 zp8R@d9g4;=5YIm~X;{6}|K}WYwX#>lv;eF8C=|t0ObkUfQlW;d=PgAoh#=}3Rrl#e zsNkQPvPiSpvJMkw4^6?x5BJ9>zG1X1W%GZxj_h^dUGqtkSw2EBkYGGOh2EC6EAGqF zeXGG5JPg_GS{z)%N%DIPM=?v+`=_dQyCw@RLFe&t-xOsJggq`V!`?5$oo#|vMKh;V zjCY?0;)G`H0eOow9XQ3>6&K{4%dUAP`Qe?Zs-lpGev`KiPp8y6JUrhZjJo=C*h&9n zwZW$PhTnR?Gooky_J64-YaJLpnI`B-qL?idiK9ZFx*tffswy_Xj-j z#!{E_Zs*l;yseyYsD~+$@A0YNALrDoZTAg){(m)_X&Ct)K7TlyU1c>5jvI>DJsA1V z=DVDity~WFx}x6T7!Fn|0Dt!Oxt*FM*9N+onJSDv1C^C%efGc<9kD{|^t$a|KV#0F z?XaM*Kx-3{N!1>Ywp9N67xH*q&EQq0E`X`n(eLLaX<=?o!N0TM3l zbg$-q6jG_)!1P0JP7!c(g$pBn4~Inhu3eNx_i{7lAcQLEHQsrkf>kfioV#fuMEFO3 zo)GmyNbCy1Yk1}Ww@C||UiSms3t9}x#W^QSPYU()P5Em6p{4cm@R>IxJ2huD*_v)% zv-X?@m@1V;^YJ$o(AkjPZ1lX%wy2Ihuts5w)!}n-H0B~ZU|qX)f|*P|kmD}21jtQC z?TnKtW-CP!sL-|Av-B?i3SPPMHt7@ZB<~ALfT+hMmOJEh*KY&)Qzcz3uiv+%0qn7BT<(C{ zoepH4G|`|%ianf_^nm25eHm=<#jt$2Uw$*~^-`aX%j$dTxG9t6Oli1zsWzL?n&bYL z0TJA21M+1NH`YxJxEKX7Ga%KBMVRrykAW!%+7uvOGD3!vpzNam94^%5ig6{v)hx{K z7%z`Ipm-1*%THBh%((%@S$4896UF?O;I~s1-fiCfgMmgjM*e%jt)^HrGRb?TW{A4wK|%15`~OLC>)C?C7cO zoA}Su$&#*@z1aWY;!F*4trrup4v7lP`Nb*^E!G`ed_}dfkF1u* z>HE4Ow#_semX(sP>I+%9f4msPtapF6K4|U6Sm;s6gGbj{|;})MYLWzSn`jEmzzLbW2hE zVW(cuz0=bRhjD?jv#g0oMfXDNS}@8r`bV<$zdOoO9SByn4Xnlm@mi*@69_%A7Ih zUI65gLFVQ2H#27sIUR6Y<2&TEV`d|0V&XrIGCh4vo*syIp^s}rgJK4tqIObC$LjlY2;QXK#De^bLh^GmDNS4_$ zE$yfv`_9`xy5VAkP1)DJd7f-^V1tum0)!Nb*+#*84XqYj^z2bw5@8vB`I2UB51q(u zP-j6Jph=tk{p*qo+HU%$BHq7urVXve3%=(p=lcbt+3qqcqe7K@Bgf6W$iQJ;6bsc4 zJDj6@%c#f82j{m54uUmay-bZ--xzpsa<>vtgy#K#u$BX;97_aBUN;bHmsg7A8O}R<>Kx;2imS z0dSb7kNo@ANSy7yVz;D&ZRU^lx1Keqp1Lute-pb0dh$M$nJY3otQpBRS!A|ROd>@# zQ=zNc~2pT5T9rw+Hct@DWj4+Q7knpEc@6dxU^LpCsjglW5C zpS*Md?2ekD{TTzdHFMilQNkQm6EsAlJNrR{yjZl6Q^Bv`g$b)(*VBV8QT&lVpxk$D zAeK2irm-f!k#m8@=kd=fm3?3pKv^N{sR^0nPir8@3R%-OdHj|#53jr;Ymcp!|c=ZCYi-GRx1wWR`!d3f7y#C_ezea4T?QLkfOmbTCVUGcc**5SBd*A z7YM2yAPL-C+`Qo8;7rY3H$#`OZm+r^2s!bKgGV^ST1lcJ3&fW5fc@AiBk3 z%&?^qJ{_~c$x-x!Do5B-LwDhbY;zut#byO4QD4jgBebKOYui^@1i=v!&NpC9J zNa>OXiVKo;KJ7rxT1gM9FG4o@xU@l*Mn92sd4|n0+|{fN^XpX%d!HulAfp<=YhQ#F z4Sk#hB|g#Z(jO25Yuksn$$lI5uh|F|1>SWQ1Y_zv05E-~Q1 zg>yhkHmdSAhhkvwzMBe75A2Y(NEH%SOsKGft9I{5%2qJSMT)=%k1n2!f zq?of5IZcH&$djd5g#GkqGtiM;0`A*35;p6cw3}B=dX{&2HZAU2SS;=h9&*A@4e}EA zK1nG^A>otFQ0EQ)76-5Nk?nI+Rhxp3a~ecg_uZ_0L=JcZshB$2b*~3xCFIT88Gf_| zk}ITLf%jm~^;%vbly1gwii0sUWS7mntza_&WIXqgf9$UM{>$F@jQY9g5Py)VIp~p~ z9LYnE=tm-~z3!qbIRj9o4Y3>VOUg=7m#Bij!|j$!v-XUjTHOu>-jAU} z_7SwW#Q^Jjwk89bV=Czz-la=>ys|-{YDnH?{l1Cy6tlzDn8k(l@3Z!SAD(*i0%Jsr z%j3SwWGg$gIB>-8fC*ak6a(}N8B{1_2uA$bN>ELD67VqK1B2kjHA&f$dp$Kcff9~`OB|3Ds#r$S0rCxfgd+>ZoS+fYE!L= z5sxqt+DOWVKN|&;^iBg(fP>RY01I zVHAv!0CgmAC=WYT@Ec`)q*Bxdy<=rdu+$fCOQY|;o(2pZ!%l|Q`!+#uFkbDTbH(k7 zEV^2HmeedWCCs=~@g|{d4W^GUyQL{DP_Y;!jz-10l#y!dn&8C>$ z6iJ7^p^)vuo;j(iT<>-XmLMQYDF}oZg8jV;%rbr=iV^qFjWX;ag#4mB*E?2;)d!yw zJ)PodU!}XrO#KVzZxT1KTX9lEhfmEm0!0(OzKZO0V4xf^0ZKl_|Y@y13 zlo{GhF~DxSlM21egwL#n+U`_UlJ@~gyZnhF!u5bR$kPi875Q#u)Tf@+emCZX&rEkN zU(_Y)3`*7Ht%&7p_wBcT^Wuc+8URbrL4$2BUHx4~uQ^H_c5h>0-nYDVmEWrLxS~qn@|dA-j2SSh>v2XBJ;D_hQGkH9PPQ$AV{T5N#ca%tI5H-z^2Sj-m8xvNTFK z5*{0o=~dJq$1%!QHOh*EH@X(B=%Vk*npB3C06UtP38DJSJ&nCE4OM&=Lm|9?{8DrB zNsBj+(+BIoUWo;Y{q^9HK;%tRM_1Dgg58oM-t{5%A-7!Y2n$OSYaLuHJ#D}9?yBSS z%@e^6>x5Y7In58k0=a$iIG%2g>lIF>rZ{-SY#;&pX4gSUiy~L}&W6|0!29?-2%VGL zvIbH7%v9Az*W;QI3RCP8subMEYBuOFt>I+=krVEu_k#O+3eq$m!2fIn8tZ`YrJ=Cm zz&7BpY&XU+1=COZ%Ql)5e>p5{EKqqhx*-|eko=G{X8Mo4rR$j8PohIcK#|T#57dLm zYII1W45V6e#VyiadAtxr$~HlBN3Iy3z>=jdz*M3lin0+n%%^hlU@;%jHlJC6W!jS4 z_u9-G5*-#S7FrD31XsUVvjSDda)hzUY?%19i#96CsD~~Y;xcNxCNn4lim>rq1j(dP zQ4FFhP*GVe&$R_3*Yl@sqA+>>sFT0S3MN$ev50=PF={eD?Jw~p`<02B95acU9HbbK zB->Ag)`5Q#LETgwS@95R(0X_~{h%X_KD0E`rBEKv>n9I_JD?J-5Y!p6xgE26WWC%~ zH-Hs43z{5r#g+c2$*NS=Vu?2 z>|1(DkvL}u_r0s&%lyXtY?Bt$bp?k*K2zT&x3oon*Sexvd&Rehw2^&Fn*w&af$WX+ zmNsw2j+Icbzu=m*TNEbjfhNg{IW=BYb1L}Z!Uj3&cOMYn(&~TQ_Fp$anY@n1Ulk#| zCb{E*g7A8Ov*!(n;x0jIb(lNkBRW4uGq>2x20M33Cvg(4q z!l#Uu&A{c7s6b6sd;t4?`lY*Op9Sd(eE0!qtghwNfrE2TR>MK(8UMNOjBSjSUY)XD zxPIO%hGRHy7|WCaB>gqRHT<`pPSvz?YJHD#@pskEy$am!NK>P0caQMN5uPP+yi`>k z$MB1vJ=jNoEROMy;gm{Gfw~6Kb3YbWOS=^V!u{~iv(ft&{?`Q1`4pvg{rSdEe}7;y zQUDI@FIkWRNb?3|6zN?F@_L_Nl1{e+B0(aHRLc)ZD))j4Xut_af*(Sh* zRUEJ68@P|hime#`{F%j?9*8Nn3F5wTUDC@f@xoTe$QcQX%e-!B@&4o7Lb`m(L4KP0 zs%Y&D+t@!lAm}+(%lb>V+y$3}Ml1GS@oT>$n_jY*4^-uiDx}&$G20Q;BciUo+>0bl zg$=LRncu@(=hLk1m3M@koSUkQd>zl@9e)^d`jc(u_=RtsJqTVnq)l$HyvTLx|MdQ^ zjqo`1V^%bM8>rIxe#}v~~k%v?$c8Np?Tm4ft`?yy{NZzii^1UibhnCY+O^XyM zH*{@4pbHr&td79y-xleB3lfZBAr>^#i>pMaC(`6!F}s4l#pk;JNb7x}B3zga{9}fV zJl+z<*WH$-LUSuNxHdSWZQ297`<{>-PK?0T<3$GO2^-BrFHfcLF2fWx2 zSZ94+JdUmJV1qT z?TVCd4ShYwo8?ky9jI*K1ha!3D^O8Ui@bXO*$Aqrcgr7=Gwk+q4jcx$YXZTm6myv( zm#EN;Ic>ek95Y#N2$c zuyjF(wAMFO)y}!fKcl+o4$mL<(a(RdXrt>f$!!kyX5)&Ks#+sE;WF%mGAc1*kk_Fb zr&11Z&2p>d>t*@eF5sd8W>Dk`h3X)Uq1LdAzUq&(5eA{ILwjO*DJq?zmk!I$44%|G%_>yYv1Yzs_~I?xFiD(P7l%xviALcROPP>o(MuM{=Onl(d<(tNVHRdc#|sGNlH8pD&Rs>hsu zJGM7ijTRF(^UDu*gG`~?W3zeE(_t%&EhIRP0(IeSIs>XZcVd1Xb5i?Qv?>mkCU{-KKf;z?PSk@gj1cd!~n)D{8SR+j8J##)AIV7XK{z zGnWF9ZX~TZzm=OV9#A9;VE;RJzzy~>Ez<42js75mkt>|4T=@h#u}9Ci{fU-_ST)hPyS}a zKY#JoofZG{8Ns$1jfF)9IXVjDiAy=}Gu-*UdnKnhXubCdveWO-oD)Hx1*G|4ukeJ{(eaQp_T1z!KXyD@ojJ2> zZnb&ql*2Al7TTn6YaHWWPe1W^Ft>&i>skQ)Z};gMuVyXs+QsqG#2Cp%F?1|XM`dCx z!fIEv68-$Eo>x6l*y4%DorRd^u^R*?2LStpU)UMFzpSMG%ba>M)nt8&qnK!lY`{cH zaqw|)ZlN+4$>y+H<*yPFwQkd4tM3zwaag^1$4#W|f2(2M%Pu22xH6Gjp$g|+SB~U( znn29Hi~ayfA*y&$RIPdtwr6(h@gEN|?5xFjBiQF0JIwZ_yy0Q47VwIABEafB$|+Bx zm@O2ELsIxuRfT_)bE0CK|3*&IoPMa@Er!CD22cgXPj*{g#-DEMVKn~p_Bvbsq4Uqo zi+3H?>#$I$JGgW|Ur(Q-2bV^A+;_g~5hZM$j~u(jq!^k{^z=z*WSU9UbSd(Krfk}+ zpK!y@$zBs|J}x`EA)9%+&uT9NLTDgSXQT;(Vlr_4rtEc(`p>lL&5yc&st#GbdB=@% zPuqItvbiF)1>xsC(tGYVytnf8%by*Mj!WrB;!?@lVzGWyL<)-xF?iBLZzgBCcRZ3j zlR?DetmmMl`FmGayub11O)Ijw`+(x*3g_11RIc`v) zVAz>iZVb8o=+5~Ra+qBKg5&;|k!BMoHa~g zYEMwV2ph7KrJuNU2r^uTfRm<8&>_gC|Jdcz?Q)uA%t@nr1=*6cxy{-;^4*fW;PoMo zNrSw`3y*fXrSlir=&oObeBu^hVlUWgx}u2=KjG&rucl3r3%e&H73;tY zlm)5SB5}FT4)r157=LWT*3&JbUM|-ChY2e=WlPd|&DzWKF^{A<1tS(FMvS@fz0$qX zTJ^n=4)0V|9yi|Ap4gg{0husHUi@HT!_2+T#`UM}&FVjsH8Tj*wT}u_Bv4E|MPjJX zlg?0Nbkn_27Uq{LY4S)_fx7qnOiiVK+`J9*-huzw^Zl9sm#})D<`3KPv*mB)ctu7I zn-aWrMptDHi&rENX93+QqHP}i62rfnwHG}+aQBRXccz$bSYw$Q)B{JS-+9(y&X;7sG%F*dad+tD}IlJ@y*)O1%pewJke7H zk3+O!>86=m7sLnr6^P@l2gaybn}S34tZ)3+uZ)J{t^DJj1`2^>kiTMwtn#1Pp& z3+p2rUu1gEFMV{=jo@3e&hI!Gc49Yx!@6l8q&3=AkENJR6j_g@6v$`RBE9I@AqRzm ztCD7}eO3D`m=LS*W8qnQp8nv{zwV!BM1+%AQ%bTN*c=@-nd1W#Q$&${RA~OZ^VZv-~CFe6qA$LfAEhL;@Pe<_T zz@ynKEajY{a~7niy6FOXyDGz_n~orTbd$FI>ld^Y{66@U&%Z9gC!4fgz_L2zl*rw> z0NKRiSK^f>?c=Xp&|+0(H^_-$rEr+=f~-=#foWXv^vMak%Ql0P1;|YlW$fT|_*++Q zn{THawyMFx1_}LJId&Lv&h<|G z+tNU();!@=3wk7=5ekX}2PFLg?TRvJ5Q=kK?>#(Hlo7-04?qQi8-gvtrO+gM1X@;( zaSfkW^A2(Df@;bI1{^0?oWg|FukVC|X4CBPfoyl@l&}=P37~yej~kVUa1WE1(9t zIc_aZRpkmh<-4Tz&0<+I^W%Q^?Y#7`TXV$U+PcGB0+NmCaNsmRt%(;?LNOp?T|kB2 zP#5_fcizHFRMdl#03JQ%8LZ$RT~R=%sZ&)6i`Oo~D??6&BWJ#lqq_O!WoI?nT*H+L z{sr0RG~Uy$xaYGAWFnDz)yC_M_1Pm!Kzn-L))wa{AMy^p>`|9!72GJJn1d84rb6}fedicn3}|seTLnMTqY(n>$l71#t-I@ys;Qcf5wlz|#@2IG zL%tvr%}t%vB25wL_Hj{!s011@4D9=O3q<-W=|avW4pJspEr!N4-0gnE4HqUTZM9C= zY>K{|FWbx&t0{8aGVmWRUHpzYgWM~!R9WDbiz3Y~-3*rLZX^}_^&!PruIp@T%VG`4 ztb(>R1KIeVxjw6+ziPB6E6N0&q=sD;gX5x|TsJX`jTCd1BB#+T0^4n(qR>4*NLMfK z3vQzaUEp)F4~Axo4QeFF9yCCHaFL!&SI{$8sv^`$AC!~sjpbw1J$n%fzN-8xyY9BY8*7*UX_j1Jr%a*}hJ zDfH1D|M_3RbXYgbET_l@x3gLlr@b>!9W6E58Bw?B*Cft?!?y({mLr{FAhn!Kg_gJ{ z3o7{gL-vE#=S{CJP>}gT+T#LTQrc=kk$V>i7a_|(nvf=0Jnx)li+_(D*#t*Tp>+^3 zH2|X?x7#5rh)~5}`>Mt~dh4)e!-6JtyJD~K^zt5G-JzxDp_gMnU$;|wo^B;~U7>bASCg4v-Ol*4rx;00rdx`LJlsu4Q(8VsgZlv z@z?MSbY>se$_z~eIHoW%`(z_1u>#Jt?QeVbzU;u+&2130dp+?kmLJh)y6_88j&z`D^a+n2+i?vt38A-q(Ouuz(S&43sTN>x1xD1=Ty zXkTe`!*)SzJ;Yi{gEEKipa&ASWCz;VURaox378QJBiZ(}WqizjKmXd_RzKNZj|MsZ zAJxl&k42Q~S`gH%-RQZ^BbG^*^h@G775r+DHfz@2ckh4({`HE_6eaRw9xzlA*>U7j%>wLzp z3lz2SZyfTf=IQ5SZEcPOqW0l4_pA5K+%IZZTvwKAAbO9oW8EMOlOGgC^-Ft$F=MHx z_j~D1X`%sg5U+zrFM1ZUBM)LRsgs5li{HiO)&W#O5j-LR7o4e2vb$ZIrVDn{7C0uKE|LFj(6(k54x;sa(hhT zgCEm9prv`wr(ID3^epwN)66Q{da{$o%`+p#1{N>8_w1XOZxG|UROFO?hOA?k7j$4F zlV!4sZKaq5io{c)ophxjn%U3q)SM$w-Jqv)g`X4H54S6jtWeiEQuJWk$$#!FutUak zhqCF#;Z5bAm}`zXY?$7Hidd#5SD5%3>VH)5AIvWH>Xn1`WDtJF*3>DSV($Z@A5pS$=eBct# z`$8JIv{;T5*;8#-tge2wEf}mAZF;HCitiY0h1la)(@4X3w!(qu{+@}gxIr;jDRP+# zMRloO`6Xoue+|Eemqj-?w{ueFwz;CQz_@y@xB*nH+ep}~EM?K+4ruPeH3vT*^r)EK zu0U2l)Y$FT^xMSd7D+HCd)_ImMN>!a+o};Q&hO0pHV`;pb zl^R&q8lGp~sBMm3geA*{=Rao%UEec8Ed5d4406wb11plc8lk=tNNwu7NN8uLgcorZ{c<*qem`|2w|$Gm^=vXqD!4S z{cz8NnV&QoX>>cJdxb^7B?+X0@xdj|&A_(3SEvW-$Sh3(okbt^K~O%uU56o?c&>g{ zs{18+u&tRlzqbrQx29**yod5c3f$c?*wG6g#9Gz*)*1!HA zj1hn?kNYl@t?YJC4(ur%FtL7miUC6Z3@WscU&z-r!#-Ju*~VrSq7rM&b!dFBd>k9! zQ8_VJ_zqT4qV^vewRQ#0el*h|eGpvfZ@=8fWO4N(V>Hq0nU5Y0m@`E=?2gKUb_33M zoOtab;C&MGlU>YBc_*F09d>GwBH<7F+SZA_OXG9aPxT~!5zEI}nULw-k{{f7*@bzL zl=z4S*ArcjELqw`Z+N`|WsQ&!yG0r~Z;juu6J9@|ju98j`=<&j)&u0(s=#utCv-bI zfP@Uhhj|&-m?zBX^(57Sot$!$Ril7npj2fK6`G=2^~iJNU(MR>&Ii2VaGOW7*5EW? z>js*XZ625X8daD5(j_I5&HQTm2``_2SJJc?t`DQqlsQ$MRu}tN*uUb^@s_)@+k(SXY^Dka_~0_`dttw zmzQOz2PIE<$u8}R2a65^Gz(UKDoC|^)CeE!vS{z&=Y6F6} zQf`}ISp9K`y|(a^M%VK}GU2N?Z|^EvX2eU=57uOnvhgv%iW$GDe775rN+1)ylsxA2@4TYdmFq(Oi^&ed?2K4`$Jf`8ZV zV_vUZue=WdfjbN9CPla~X~c{%PNPmt7%$H&fB5uXZ@>{B#arrPA6EcOw+&D{Wfo|(bDw& zNVb;jnn9{fzPv>g1H4LkRA?kGlDAJD2ebna9I5eo;?W}A=&I`&VXudz;jA?>TTUF?xf%5x z9D8WPd15y=4qF3oLR`E2<=-G>L6EOm)lbg2>ONB+1v(x>*L{b4jcnKn!y_0hiJX@m zSjaiyJXKri2_V;o__RC}E96ewR`hOQpwX(xf3Mp~_Brqd>5PfFtD+d#F_lrF#UQFS zK-ajnE548*CFur|64=K{P$K6SmakMxI|Lh`02@CeT27KU@MD1qS@48+2E;s&Mj1`z zQEp#w7Tv6kX2SdoW8Mc31)3NGqC;-FC(OB|KJV5h`17yPud#&v`5Qk*|lai=Ix^WNGq(8#x!~cEw}5jWIJ#+ z{GbU$_D~Fz*zLlmmQ>(M&I?4>I!hSC!RF-{aT+XK5km0Fv7b3tj1(AWNgVG%aFhQ? zrV|AO5*4R3bw1nYryHpEZS~)Hqx7`=Y6wqD4 zyx$2BE42GcGj@8coBy%T+53#;`5!0IJYDh4dr2>Q5&%=dPdh`p`HiwfMU$$*zfsml zDn*^Np_(7NnD5E3b+wSwGQyW!?tNd1Y}$u|x}ZJ=9T*q`7MHWcPGGVQFkh|@;~dxr zI_v=RX7gXZ`eLC>2X;v;2xX46P+|mU*y$*@VL>>2SZc6Tcc*(N6xEdhxv>rzj(cQ>eK5H^@&P4) z_Q}t_G3->S-o4sY$C2e>gnXPWGz=axXvaI*g) zD4pT9rzdlT(2;c0-SEueH@fJ;6%#f-)}m${yJFR;i9*aWYQ|ZQra99*<(oz%wS41F zHYs7Z7joci(s>iy9ix~Uid0ge2Sd^qOhERbW@mB<`t?2z-XioaZb&4`@a|9$mC zvfhD_qc=g$c8Y;q`4*%v#YO^9Htc}Hf@0Bti>^_SF4m=qD`rE5x41)5N3eBq3is1& zv0*-k_WH~U8`BO-YdRE0Y;2A9{e&cq7ld4bAzrNVwrpVLQK(xfmF5NF)%9+Da9_5j*nJOu!@I=hH0)-x%@#b* ztc%$|HvVWCJY6aJ9CP}Gmx>|G63GJT5}hXbL1#6ClG?dfLC+P6Dv$;jw?29Z)Tal+ zHmXU5(cD7!5@0{o6+^=jQYa6&WNNl?lY&uRQ`F^ZxBd3S!SWO~Hb#kYILly3jY@p) zWsl8a`};}1vwpfyJuCRfeK#sDdfxTW0UpGC!MWmgMQ^}>OOr>F2NZQe$97J~Y!H!=pKu=FrZRUIUEx$iNp7V-cBR7anml@98pTRFES#q@o@N96Ldbjbzn9dc7vOOA2vI|CDW^+rQJp;v0{LzZ9t zINw;u{VKk8Xx>{dd#+vpxruCXsH$A^hkU>*NayPjnQ*6HP* zU;C~(9fJiP15C+{WK@o5@`Fa4!2@KAe~0`k-){b$#TRAO%eT++ZSily3+!IL!m(YqI9@MAMILXa<6{Ihqc^r-I_(bbfeXju3V~Il@d$f-;TXMD5o)MM$5vx0 zW)nr$Ll>c>K$Jz7ElE{1ex*1#dgjzS#QZ?9&JQ zD|o(Ho6d;?ZO=Ok9+6blstV?+`?346);H zE<)H7(gCa%_U!kZ$B<^R$p(&-6MRpcF`8|a!7}DPArr@q84wOxKb3W#2w8@ zsNHUX_7`BoO0>RU6x&~Bht|Wq`TzE^mkJp8u91@;OQ)e%I!9CGdq`3|lFh@G^#mmr zn@nLe%}Q(-R{`5&z|x59un{i(zgm^iQPH0MaXHCw;4+db6W61VVjw}Y7kYC@7Tw~1 zlnd!QMLXvaDHg4p@rc}(ZS%M(+o38_M9mm-S~nw1*v&ME&WIklJn@d2(X7psWYcSS zP`>7#Cb=y^`jj}|M?ejZ$GenKGwcB+c9S%tBF7#dXO#-I2+P4~XFxr}5w@cyONWdBk^G2m_%P@%P?DX11!iDoUb=Xc0s zm3q!a&klKwGF+D9-X;4)yP3bsSF z{lS^b(v-)8Tm0X@A>9&KDk*f|CB>0#3?geGGTy3TdhWS1&ibNdh>Tb0^wZz}+n>xs zBCm)eVxiXpE1!*;bI51%hf4aUB3C@%((RHg-LD>Y+5?x6fwFoMDoNvshu(Fa1^ z{#|Mw5@CZ4$BhmX3jvg^oN!?ly_P&BBh={%-P;6D+_r0Vx_+0wg)Pz^(kNRa(4pP(7YM?((k^$ONzT)q|Z6KQU`5;sha!J;h=*aX|JJ|YeNgf5;LX-XUwOG zB4q4yf^RP$BZr%2V(sH8CWay#snBF;ollf7jsDmjQ#88Gq=%QGZhXszFKO(bH~tyx zUgr4E+VdY#nOlwkt4}e@xP@ssz%k=s;cU$Y6$$J~A)Csa|=dvWAl`z5(@{ z11?3fLlZNi*yFwrQO^P8ZV&;%xzgU=*c|h6o^2j z13H-FULSTc$naMqbj<3cm;2=+P2%4AMI z()I3bbMMz-v&0ts*=i~BhMrM{gBr&vxA3Jy@@DNCw`P|Z{{&?ZT`)7;`9R3j*1x_I z*q}B_nR8jl{{o+as$3jR59g@1(b0NE$s#81`e zFu`nWQa?FU4?H`!xRoG!2Z~{NbQ`UwQHOh^RS-Dyxe=sSumRH3ITAfR;9_%+jupVh zIJ?hI*xCgB-?}}C=8-vvH76FLaNv=Po3&BGxcPa(pDx-28=DheF}!YW?_Z%V*2;Od z^)$Y4JXVVgMTh5vf8XeT{OcL&>*Qm0C6EqmNV-i7$xVv6L6NIe=+lmKLyMQLiGw0h zNX)+u62@47k2+E9ib^_7d|^qoPnoy_c>ZqjL6;k=>ve~rlpiYZ_ezea4{N%BQUx^_ zVijEqIEkbvS?-ydZbcmLutbMt{po=j+-^ySG}Cxjrlf=DGPvhO-I9KC4}{+jYqH#7 zFP^Aql;v?+MK@*OM~9>=tsBWqLv=qm>hZx1ATu;1j}IAkN)O%=Vp*@wN=!a8GB!T4 z1kPUzI^Q&B&3{Fft2`5wrBX}^MYd6)y+F7GWw46t3r+xUL40tp;sCj`K!-eV23>bu zhg3JD$OC$VCKbp~Hprs*HeZUyPRGOmGIktWp9zoM@dKgJ482$U+V99F2i^vO%+n~| zlN}VZ9SgxjH!t4_n^!1$Lj_Z;fa^xKq++OeMq)rXkM#-hAzgH`EEy_NreL;VmL@Ra zww~K@$78ihP2c-__id92_B!w?Wg*zx3oOO8>MmehdJ>>(1y=KTAWK1Ecth7%?Xqw; zST&Gxv{qe7=SZFe3_D@+pajb7jVl|L#o$v|Q{Lbl0gAgZVmlNoOaNiyt~%qwWD+Nz zv~K;j#As6Fe|fu!#Io~j9Jsk7-^2i=QOpjCY^OqjUXs-ocHJ-RWIZK9|*ez9#+foQd zVwBqwLopjE5=Dg?q|352DMD~jcFc@k@TaNqOw1tVWkU;uh-qNi>kj~bh3dPh{ z#pzef6)7DSEY&7P@Jp#oq`E*tM@-)Wog)U;+Q92LFFGX$9(&}=Pv*ps+d-W`bg`d< z8)T4R*WuQsU9rRYaR6wvfIN@-yt+Vs-fh_FjA}rEL*ahf)_oe21C5FOnA(M7+9mV2J{vFAf%nE1vH%}(S_2H#0Evon)d(->UUdOf9OT&3#F);}(@)l7=;{4a z^>f}%%@>o2m^kq6*FwbPBT-V|0rfUc8AMI?yIq)>syX6aP48P;q!@DQ5Zno@;9v8; zAiK6W3YbAV1x1SffYtzHCp6mqkGf2n5*#5e3{(SzWMhfJH@d0|RI+F5p5@ z;eSpND~Uw&Lc%XPxBaxfd*b_ibDr~@^F7~l1LAI_j1~j;S>n6$?yz0}Onh*z>bhq< zljCul57D>dA(%!+)jI4Yh+#Hpqzg<+Es7bBe8RE|JS^d2xjzEA)+_JI@zdL5}j zI;cE7rd+}3mB4SKFMGub+Xahd$Ro)Fa!G}OSh96e!W)=4?wr*}E`?s1R3|PHFBNX| z?E!j0B7gbBK62NsH?o?nm|O`ptod|O zxg4)g3%pG)6jqbtq?viV-AzN|C~u^n?58hB49KCHWOkzKYDih*bFt+4h$?lO&eupz z!Ec_zrUoIs0l!6_>V^!tGk8CK2mEwwarbgCwa4v0_V*r3MM*2djn_Rdh?4!m-^}qS zh2BZ6(sOehsn%uP7T9C}tl#diGu#Y-)238>if-10nk>`ul%8#5%L~@Q=cri>uaZ*i zp~x;Qeo76Cr&``G(*-G4cc+FtplDVc5t}~%qUzu125_U zovXryInsN6gfv}Zjm}@MY_Q=BP2g$b5x;ceuQ4x9ClJx>6S5R zgXdwol8?usSt(P;Z-FkSdtN`l)A1`H!O-Ee&gZ!DidT!e%gmH!+mES>l^!kXe)a%A$D>HtqTb4@feyQyLt=#a zP}|=|+6D9GH3-g%8pOMOn`NhD+oE>+wx~0FEBQTs2W)AtIZB8<%VIs_-19N_Xx!nq zy-W~merM-=lIp~M_AWEf6jBPPP0OJ)NTSYDL3mI6iRhNJ395|1#y~@;eUiRNZG5PX zISXMKSb%>bY6|U*ESUslhTbKU+E`-~pzD*DxSs@H_ulJ{Tau4*^q)VBBfP@;EH#WW z!?+diYzTSLvGUI=?5Vk!(QXuUi{hEnq|&V|Y#}r#jGNvdEO4u3+C@7g>&5MY9A2XU z8#e6XHHQw09$)jh?AnZWk0wV}o{SDFBLspCw{gM>)yd|%zGAW~E@IU#lIO&BWxttS z*+D7FD6*B(Z1Sjx-X3sk?t-&#T$FUl^Lf3IFz0w6Y8u|SJU1RBu5q5o^3U&Cx#cMkl0DVsGMu4I^eNglu?}T`&3O zmyf!SCa>+ppsb_M+s3#K^osv4)C>qxO`1!9lhaGiC;6*d4nGp8iiG}`#my2 z>qU#IV`Hl<_GI*Ig2a={$81aS-Vc0cnLO7l_7_ED$4e7*`NYi6JxD1)^{|@K95Hl< zIU_J!c70>&~|d3iJQJ|b{BNkesc7$ zg;YBT{r~mxxn?z9vBT{u=~Var+h0G#<>M!{s8ho6S@lp3`Sj;OJ{2N3pUVeavZuBSPKI8e+{VTT2VC&UC3L=j71Qs& zaCSyebHpvf{V~G5Q5}Lue7#?{D8sW$H2|N#nJ%TDd|D4Z_{Js4-> zqKi!FnL6l-bZthXeCzC1MGCtDRGPJeL6t!8I~ds{h!yI@W4lb^WNb!hK)Bq_X^}nt zH{+O}S+d8SmT+T*503Q#MZoGcmDwL;4hLbo#JE>3jD@6v`PkTs3iG=hzsC&~1^OF< zFM4>lB(h$<987x$EPk7n%dlOlj#=fI6BPSOrV@KrHIPPG7F`%nq1rlo(50C%meViu zyaJ{4I?yZYpyOUi4c3m3UG(j* zbkH4vHAv?ylpKM?H5#Bnmu&Z5&}>d&V`ieavQfGzJlDNO9ydLUZc%7+cqy!2@qmFW zR9Nk_#OYczUe#V7fzYP?fi_BjJ;vQA;q>(Fe}Apxe7MOUmj7RE2HEPw{_t_Ll~OgO z*hi5HO0!US(;L0W80xS4zZ(;>4Hixv5jmt3qI7WR)9+w$WQ+Q;q+QSek_>yJP6jo) z*MdO9N#6>;0`N}TNE_6a7YG-F=pirrXQYbT%G`;=z~{~I zc7jp>N$W7I%;o2q<+9VEhQbyUgsGvEqV#N$u;0By&?Vmw-M@_*h>yJU{DFF$5&}KN zTiG^J2LMW8k#Sz+UqzqsSnUB-{$;a_g)-38Cg?uMiZy}k7!q(w9#_gFGszHTP8JtJ zSxgb=V1u(=Kx4^zjIdjTg)ik$!8rn(a1wHpPsFVIulVRC~(UX=kqcddl(xIqhLEQ zI6Tx2m+}w&ZK10P81-MO`H(E;1`H?W=xsCuMlz*HqDVY))j>!|hLRIt6R5ag5Kx_- zN5=)Dd1FgwHv7SFvn%UyAMlF6L3zINt4lI+B;FK}OmC4EYn3?6Mq7-{6vYXQM)46kMVTdyr zJ@jeCLNV}Hi9L})5B>KB+MoaTXvrQ9Ii(= z*_3f5KkttnO|^*=FLA6?o7B_YbXQm_+YeNjWdAD=vO~|dA2QEb;_kp)Rhl#b3b1cU zb*`&sZwR_8si8MAd0xjs8WC9Iu)gh*Z*=U0H5y=X1q+w!e_QbBS1g67xOlftJpEXa zt-2{%$)5JP?hR3nJgvn`{KZTHxUk#Yx>SQM zJ+kf)(EC^sy=d}D`UcpsUfxRYI z#TuDuf;#e@?_MNpCXj7r^T0+*0X6d*C{4Q{n>+;X5f=!zylz3cJSOe3RR{Q+g{fqV zWSa_-`EbQXY3roc&{U>XdV*wn=~Q-plf(U;M)r1xdy;*RakBN!YpNH1=P&O^m_Qr79<#`KF>{V5hGeP;T+T$U}-A&+uvDoRI?+P@$9tivw(Aq+# zj}*|gOcl@QGIh~dDLGp8?vY{J+-6wW8x`0&VSC)G6AIO=2}0|Wg3pq5+#+($`{N+n z%>b8ADRL;HqckVJflKX2ubv?Xc<3YVmh6^v&RWDj5p-2j9A3s;=Pwg(mu!b)EovRT zmA8ydn^Xp>4v(*Nu-P0rtVV^J=iYoT;(w|=OhEbd1u~B$aRZ7Idof$hfRarqU{#$? zX;wsK`s0EK)6Y#ZqnL1}@`BqAKZC+1*q4}@9deN6Yyf)rrsy6oEjD9Bv5NT)UTcpW zBh+)pc{CSnuRq@YV-rTsynok&Bcjmam_QZT1e#y*qb_dz(l< zvqil(XtQ^WFnQLJnFB66yw~{{w>En>h>cszn07bg*3BWu_IVqR9TjEHGH&H7&qWxw z+C>QpeBU z(S7NF%SYblLIzw``h4VFz-t1!gbkX&qX3STsmiv)&jS^wIg1MZOuT$F@jNHaS6hka zVT}<=EutzbEUIp~mU?Y>h??ee2MU^$lHdN}Pujoy^7rrj`CZ8pO0kFnj)vJk&wGdi z&bRm;+8k)IJoRs^sUW%BW+x|x1k@4?$s25^6i}vALTNS&(}8Gb4D#*g89e7Ku`V2o z+fd3L5=qsw;`!yG7N8=;0FEeQ1CwWx)8&!bqyw0|# z-LWTnX1M?GdzQs)PD|LeQl4h4rjDf^S3etzqSJqlh0~AAqqqKW!9r?Cg>7KQh&0yE@wVjB!H7_+Q-c zE4C!)N(ldV@BCWxyI=lx-k&A3!i7*y?15Mz?Jj06xaGoH#UKQ3bVT;5R)cK!fPA;> zlF+rGnabkGeqMRxo+%Kf!~f~oWn@XHwpX4cJWmRyKlI!ZG3bI7^{9rq!>>dH`D*Vw zGs=9he)hO;x9p41U6ys|r7^u~`CWV13o9y2W8Es?JUJ#mxF25iZjzQffSrrdEl7qUCz<5U<_J;B=5g|4q z4vdPa4 z=yt&X|8Uf9*Vu5JubP!0mDwsyaVwCoKr@ z3CuL0DlY0HFyBBe zPlX(33P_so3H)M$+Y*6oXp?!BbGAgZ3rYjaMev*!MU%3cbklehI!@S|$Es&!N8k*L zCR+{;hdAq&Xt*h>^wG=(?j{GQ;%nbNNme?s15{|{0Hsrkk0=-gO=?hC$YS?vf-~|; zw{&G$NT2&>?uDwfNwdI%SE2vPcoRAn zFAX_J23>{$krU5Q8D@w{pcJbpvYgWF^iK9Kax*NiF!=xfq+|p~*A|pG0tL@|rgfCO z@pf;gWroN~YNZLv06V11+*-f^Fe+lN@*H!2@x^@0FVv6htuKd){$er`zp4MxTylE? zNib9ASvX!XK#>QO2DEJdc~_nfp$(XK#eGnMu`Ku~bI31KxkY@Qi4WbYsuGk0Z3x1O zWNdKOksR{tjV$y#OxMUCNpjs=)Q9DZNIZk}$~yjPMZEYVjZf{S#8HYzWPLXXq7xOlI>fBuWE57Q*eQ*D;)4N76LH#auBsRDd~H~YW!j9<)G)+eUn zOy(bS(eJ-$vP@g1F}F$0aE(u#I6i#Q%zbI36c8{z^0+=_kLsZ($~ORU;Q++ou=u1M z`u{;l&(Jc-tDn_gf;#+ex=yUEhCQgP_e_&!`lou|kQGcX=E3($3jHo7~?KY1Bf zO{XKWU`l$@EG%W>SYYDmp|b%cPv0~KN35gdN<>ufXym}1mwHWLV|fR8D6=NI_`4FtF!vI=JHOhKEwehtp7rXvE$Ag)+(+NMe|I4txKRwEQ#aB$#i0K z0+l2~5S~IQ)=(sY(#$&}su1rJoMD<}_vkasN&1Yaz-?Z=xO&!x@C%{0BOPHFtmBUZ zysbaP0cY;*%l5Gh&b=h{uvTJ>3DMc^d!}_rGnpMh^;7lJ3Yq+|wRDe1(H2NNzIH7C z{OlV^pZS?g#UsVhSH>zG?BwUnB( zqS1!cHCQ*1t3qC8$lqR37~JTY2nE-r8X+0NlGXGbwUFUh3y=k~W|PBDS#dz*jrTe884z12Nt=q{!d zz-P;+G}pqLK*M^Ua=V{)b5KbHXkxSI908d3yrxBC@|i4VcCM-p2$S9P0%M~@Y*3G; z1q)4MW^bFmZF(kjTwJigcr)rb+1a(SAC5->+F@?xhU4-_vGf0Gg3AxDG!W7|Tte81 zr<^1+Ud^KMikR`_5v4f^G!!HaU=Q(HpYu@dlO=8eqlhH^b^#v2v$!n6f@M&fgUI^_ z8D8kLfc!073e8*LRE!Wxioob~(^qAMOp$*dbe}h_r7*gTJll3bo~nz+J<#EffU-yV zzThR{nMz&wa^TmdNh_2kqFuaYu9?bgvW8#jX1sf+_alBh9}AOo;mE7Sb1lqLkUzxp zXlk!d-VRLPR%s&t42TIeE0;i}$?=KH1�|{A=KKtN0S0!m|KWP@@)*tRCwA-{P3cGt9M zE!LG(39zuO2sE1?v8SSuZQUg6_dq6swki}k+z&ldCTfq8b&=bcZs?IyB}kdLjIy(J zGa43q*tlU#)Tj>~>hb^0A53`J{YwuSx$eZ1+;TGu^?*|JQRFU)Dr{6-Ws;+-LerQU zx&wab59maukxp>ynWjw?ukyY{3IlfW%6+saqfYX$WDWX>c&w0XlOlWBeKXN*>yJQl@gH$8woACmaw@17^6`_W|$S`lS39C#c9_%WQH?1dDU%GCYWHP z?s!j27ET~pW}}rvDdH)z5{VCnAd69&XGhTDkXC69Z_pa--~ExKIPr8-Zf3ypC~QnM1Dg=i@qP#l$S= zF3=>ypdc=^v0Yw8bfw$YNlgNK%c+hAFhjuu+oMB>#oe3$Gj3c#fi%EmSpK;6Kmy5e zV#5M%>yU^`Ii&!ZvMrS69-Y7f=N968h>wCP1yf*w9KKjq>6YOCx%V26hY~G{dBih4 zK&b2IVPK|4fRq_DLrIgi`yojB!rH>J0t1Jtrn*UNXa26Quyf`J4_kGK)+Uv8hgA8)99>@>) zuI9yumO>@xZpkTcJzP=9l*mtdCU|@Ywf?2PJ#g6msuH|bSE)C!LguJr~!Csdy# zL&!Ow&W8+X0ew6qdFm}v9C=HAfVVm7gN<}N)2zmTj8R^EbJT{QL$Aj(r_}iog%Kx$ zP6nNkcCw9R6~BKZ{6Bjk(87jU>~aU>{Ykz9mMWTDI2BHeDJz59;U+hQw9^f{_Vu#oUv;P_WEi>;qJ;af zbG##a>jd6tC>BnfPq0FB-1K++R`TG^aOY3P={<#67VaBGfnW}zpQ42u2DJ!7N3s^$X{%S0(0Vixc9=I|1lM^<^ z{UG`con^Rr^2}*v5v!vH{Jg;sL#?Y>#ivGW`8rY!&W78lHyDxy>PG zZofO>Uu*4pg49O!&?jYk=uJ$X80ji!73X-V%I>i3@=U0CuaOSO>qreW7ET2cTOrd6 zM>_)Xh~rUa6Q+jP4b!H>k@bG~>UEZe!^-ay={Zn8gb{4xiZ4x8>FfA6_w*~U>0Lg3 zo5PQA`h7a@bWCmZJYcEi^+M%(v&|MOYbiw{MOMRNgnb%Y$N z|H)VVO#zMK!5{A=x(UGO8k!y3Mk!zviYU#gH?Whf9(ov<#G;FY29@eUM#snFMZ#0j zoo*m`kH1#z2nWLCc)VM*bq?g?+sJhf9b@c|(92sjvD^=I`Y`Ld%+1b(z!8kX&=DEB zZP!7uJ_^Gzx}26)^zhd|UQU@H^T^}T2{Pcq>2K!5aSo`2ACec2rxYtGvW(KC&t9y! zFE`K#AmJx*yh(Xql(kYYP5KjeB`dBv+meZ?zzYCZkIynmFscB%V3ay)vx&NK4{ zTn4==-R^p2Dr@P7l4fNdT}`fd;l0RyZdTWk#?X5+PN`eey zxA{G!)~`i<3(9tPOlwvre0}Nbr__~h8~o=r`!=il-@@HY<-N#cEC7R&vG~vxlNb9y z(<5%R_G5+#~9V)#$Z+$*+6i4r6aqR(?B?{z4!gk|E^?Bc4vK3@L95MxUwB5UN3Do zvqkxo0;J(|Ksu7&;nfAB0VtVtob>Wggfc9PI85*5*?0m@Y^>1?MLv^ZGCT}y3YDbA7+*0 zyg$V#(X{VcRseFz{5kL6j994+#Kpai%?LUJy_^p-hvYh8wyKut5_AY+Laxow&&-1I z^o4<+&TNio!7fl?EWSA3&U#v)m>0#9qGaDLDwMtid)CHl~ z280$t7{O4Sp@n}O1?6Fo#tq%WJitMRk~ZFRvrGUvEt$YdvZvOsD*}ZjwR>TRQT%Jy z?3Er`Adp~u!`@{#&lv|Uh!~9sm1(Ek41#SVcO!#3OZDdgLdfvp$<8 zKq(M3h)&Xd(8a2R*T~-t&6iu$JyC^>-me4-EfQVtk;Ss!Ne6fiw<26H_1N{mKkW*j zVIKD5@fGzUre*Kdf4g2r3b_Svo!D2cH(L-_P>P)tDF+=3aF9N{!kz*LsYR`0dgxQ$ zg*;rDooC{vcL|K$&vag!q7xP9tk!^R@;Y%guR>MnwoO$kE@iPNZ{^e$bsM`Ys510S zP_}pxQ^@@NuOA#I>!!D;&qj1ejkntiR3pXCP#l?U4}*h|9P0kW_u9KA?0k^0>@{+A zxR8kxI~_~Rw0HU_#a)W@P@2CDxU8Lc!42g~3+O~X78YLt?JQ%@`0T0J8!CJ1<#6ps z((7T#vOAHepfl+5!GKF#@L&M$>E)*v6Q&rlkppA-LV!WvxZn`1cDo^AOy8L~>CD}ibzYv~fBPNVfMAw4_eP-DeJ|k0eLa6#T&n0p6P2~3`nbIw?qucUqIm< z$)>_o%Z?6C>+)FX(Qr%z!iAuYRP@T}$oF=>^I6bU^0x*1<}N6iyC8=8pG;-1q}I<~ zCgk})Z;nRh`46bskm^N}W#^zXYi>qz-aGo3Y=O=tTlMGZwNq27nT(I12i z{gw%n_V%}0Ne6NNGfa*|&%`qYZdmczqCOame`YElAKd}%gz(JXP5^Wy^G^60$q$#-9L%?#ujB$CD9cOvh&1C>k9u2NJ_O^R%!%LyuQb)Ylk3i2zH z4oE89_60X9k^P`&4VR<-={ZF&cu5Aw4BDmUpf?eX{%E2&Pkw2iTr``>N6&DJl*2s$j zI_Ta=qvlV+^rfDO{2gS8(X2fY0BQ`q5UNFK)EHs6sNVxW0hLx^dGEW5mjKKCOyz21 zSm8nskc?NR0qY7)T-u~dp_!^;UJnG9s;5=~A*X}x2!z;lZ&5X89C_0;JadGqfSu@5UgyVG$_E`8~K>RgFTmlx0B(e3&+9Mvz5LGKVT!H z{3N1sKn{!IWdGdhOG2;nyXj8eE-|doQFU_96y#39rKq)?M^{8wfH!?qQ5uN(i0hM= zy;4A*_5o0~D3d&@{8QMEf&Qo61Df5?av|9t;A0eZMi3Q1dtfd0RVbK_8I5!|TLqd& zZq%OXEL!hUdXJrIko3UMRYaPsuoykW3tgfHtktsY!ieW z63u(ow;DID_B-Q0keYlF(QVNmNTL(3U`ot}Sw|@{D3V5LI(et)68XW%ZvJk+Rs0;@ zA@@w>xsdb0s2>_H-Vc-Ec6lEZ;I~S5Nr9sawP}_7F@b<>gKq9hJbSq8bC#3cpnm=E zAFR*>k@w0czE4)YG;S;Os2Wl#yMa=qA`3`!1c)^#+ESxR3ydYV=}K`9@M_SDDq!x- zh=IlNh0t=4r^UT70TK>7S1{QJD$jq)w|;Z%|5(-tIIXK;h4sJx8wRR*IZ#sKCbYoH zH#K<4gsrnvc=_}#kT$+7i4VOV(kK{kxiRIAcUEAB;7(+Ryj1>KP>VWM*-rB4eN!BY z1FmB~)TBIr1V?@aH$yV%TV+4;F&UDMIboMb`UFy8_6yuhDL|1ekJ3PYK|xvId1x5c zDBtdv?tgO1RnHYrqkl$JK5;oa;1a7iEXxXvXRrq$^4C6>r~7ZD^PwtiyI(%<8d(nY zN`dq6sKLn{Nc)*Keen*ht3f=;Oq^0tszf*QC<%xO_o=qhjMphq_;uZLZmQ50}=jFfLj0Q6@g)J|t(zV9uy zjEXqz%wr{7bC&eELqk;vyTn6X`)%(y?*`swAUUDnJqBai1s`CLtIr*Yk}c{2P^H-{ z*&DPrxF@PhjW^>(37%(}CdCD4471%qRjwm|z|c9!VdjpTt(*AoGwJcB325=skb`9K zr4fCBmCz7RHi1&CqR4V%tHZ1SY%7G#KVpRiV<0#<+^^dP{0w)S!;jUkJ>+7!{C!$m z%!!HLR+hx5!-9GmCp_DEM&f}MiA|XPM}|O|*Ns5T^CvUS%9A1;Xk}j)kBu0e4M4Co z9FFsphQQ0ZA)&tUm-v6JQbp_?wF+V+nW`STEHII1k1K1WxpYQQocBeEon_9348ic5 zwgYDP`yF-kM}CjK?qQlsetm(=BS|k!RAj5!T#`*GK$bO~(yaN~={J|S*U#3^y#4h{ zl63zyWEqp-(>zN*^Zv{tSewO8Y!LUe`kBW4M9*wqg8yPi<&VsU)pm>wu`Om#aLi8W z$%GEeEDV<=j1x!Pt>kR3dlm_SER-eAlK@$$BVx6pRhkmg$&R(vusvMnn31>s)S&R$ zU127>k?o(mj+Ah-8%~S~kd+(agH}ZK*_xP zaO`-R33NpeWZf0iL+8_7^aZGM?EpnHZ2EGZ7`qr6FTqm6SkO7swzB<@3rbcB1R)h%9(iJh=)4$dw)kRX};f7IaXjZwCxH06>5#4oIFrP&%dfh=Nhj?B{huFy#Rq6H>_31IIN}dC_mD7>$h< zvW9R1dHk8mCPBI5qQ#bj!4(~r_d9C;|9Tc?!We;x9T=Bq_rn$9q4ZC4Fa5~vgiyi;KL&U zKZU(VszeALElwd=2vQeR$wvusC|E~Qe#%77!^z<_k80#Od=A%3tw&Ubsn8hzm;4Jq z{nwh5;`jzxf2EG!$;A8=z z%LSkQ-_%T*X;M1!o$p>GYn+&+1Z}v7EQ2;u3Ro*|fX)NaB@re|vMuPL`@KNW0)t*a z%94nN;B=v$-9D|~6DoEpeGBMqe%dMs*+29|=NA8Ur2T*a;y&z;=L)-wuGNQdG9lvv z&dVQJMq8|eDR5T4C4gU;v62aA^$o81t`2_9!~H#Nd-cQJ>DXi3UcJ+j2vXL%{f{Mg z?dklE6H^7OFw|B6zXf{Z)&OA%M4hJgN$%5YCzjIPGj7w}Y^l6Kl_S|3Q0lc&n&5U| zY`Bg9N87Oc*mS_3j)m|SPq&Wcf!M6OZ<^+nEz_9Wq{fNom5XNcN+YE>PLU&&=0;eJ zJZ^fms01o^n?fsn%Og`k?OmHop4hfV*FhI!@S^}ye#wl^ZDDwmteDx#>+sd)s#>Kd ze6k@;n8(jlY~^8|5Dxc-VZVk8qCKGgkMWf9NIkoJcB4n8($M1p?jhTu_8qUzRbf?X zC$E>?FlqIxbwMMH!f}Sr5F~Ng`~nv?ZdixUaE|Q1-3WerG-8EL99=0j17HrN0NQ;9 zr8yq*L|qKKqZWxVsSJ}V0;N@YG2n)Gue{a|7DTwtMkRKbb`QCBNKr|Q`%2fwQ8QjP z3QTeO9Ih9CV<;l^f-*qFQ0 zSGxmb;@drThoEyCE8H**y;^MjQ66c$JP&HpjrI%6OC9U&jsg%4ov*ALnMIdQ2K>PU zmEFJekdf;X$Z|7%iU*XUk0N&|O&YTd0-1^Ya`7fvwO5m@JaSE70gc7}5UFUSKLHzW z&<(&*!R8;GlGSQ0#-Q*(qr6vQigjTOBVLSgFf0l$Wme3ZUlocGjwV?Z|0+|=#DfNk zA^cHA0k^1@tbNF~){pqTl^9*MX^!1fmfk z`+0RikSr7?fUJP=%i&jv^FGHjeiiJzjN!K~I-VcT>~&qO!0W(%bok>=$pGfF!s1<3tu#Q-mOCt-V;av+sMqbg_#a>J{Q!M?mb z{CO*xF4=$!Y@|bvzaqa9QHy$w5QQycT|0a)NjjjlRTI56M8`iZ#rs>-+vePqwF`7Y z9h~WRuZmdbZr^3&s4SGZAH;Q5dgk@kEz~e>5JmmzyPawiZaTuJq>;T&jGMD&xH(EG z4pZbHrHPx+?smm1p6OKIgVvlU=~J>L6Hq>|MP27nuGW*3Y5Ssg&CUv_6g7xfD-!vg zv+e^O=4No`Yc1;BhzgGZmle^4UIQ+lzgjCiMD_{plZ@&qU0#PI5z& ztLvBlY`KiLvNG1Q59lm0%n+F>4OS*F%#1?fc8M{`ciP z3f3QG*E5_(!Fj#WJ@Z%nhh?dV(-s0)DgMCxqd^MHC_Y#K1~PrP4Bb&0hT*=$aK|`7cifqlmR&D8LeTRakneUiav1hc z+30UjpX-sONwa83H&=)|d{fx1lQ#N;JY1`^LsApq*vzA8SVo4B;l|Xy$G9Qn;a^Vg z{j$kn`PsFT>15AvD{?1}&a{|0F-Ir`bPYIw+?Nh%*2ENOv}*iU7r8bHR_N2$h3^P@ z0F;AeR9z5KUp5D!;o0fk;9bJc2TNq=2%7?xNfpuA(Ai3hQ)n_T*?)6TUC@?Ec*PA_ z!SoWyb4Q<$)Oqwt2VHO_*+-U<;>dnp3cEs`uFM8Dp7GzcQ7IGWkFfj6$N)B+yJG#= zNT6pOV6$7>vwv&?)48|HACP8lBpN49Nc5P&OiwAGbL>Ts#uRtSN<`SUexG2U=<3Xe zo&zqQh_A56RM@DbU2sfP5v}9Lho(uFsgR7dN|Dc7G84SsN!MOICfX<1;&woF!AGZf z6wnQsjJuKnmsPzsnyI|b&kOC1JWurO-%8|bQGL{yy+~p2xqcjoW%#&vz83S}mI*E{6$nmjh^(Z;Qc1QrRh1WhF)D|w z7em#OSC@Qm0LXj`j9Zz?3!x3(?SjSMyZ|NEC=igTjGGW6yvHt?@Lt=Si@#s>qn&U3 zPYu0Zd`wi#93}T=bVJ{q4!VXenUn}Rs+AGTf>YSEK&-pH5ZcRo{0W&rRBusNxMzzi z+*k4YX}#a00HpnO!&MF5BS7qk063f#dFCL=GW${q!5 zn6_csJ*en!Q5T6*p(#3EeKX=NL?IW>uJj!ND4sW5_5scFK85@FfBtUu??ooiEKk4u zUu3lthXXg8!61WDq)}u&rNM%eB4K@SpA;5b&5Thm9&_!w@COD8F}79*N_ji5h%qPX zfZ-kH#gAabhk1(q4{@8PO45rzb2H)NZ`#v;Cd;@P73Ymjk$f{WeMBizD6)pqbkK49 zL_V-D46Ow)TnRBA$Z>tdRKR*n+apN{$L#3XfCz`5BarNm7%NZw-Fq|Nu`JPlNn98! z#rvB)Dx$Xse7NA$+>4SfIm&;us5g1UGpEVPw=T^^Hv_lti5^N;{$|04@yz>=FLN|l zI6~x!<6&iJ?REH#&p&+I-()?C2Y)RiUVll_k!nk9a@wz8AUdJ2&UPpG~!> z*NZ=b49NYtw|;t*Wcn9OcZ??Dh#~U?JPzSqM4o$lU4RKPf82T?f#f)`LsDx7f^td$ z6!R^V=Dr+d()LBHWS?v`t52z$J+bDoU2s*RlbmPPzj6|491n`qK>IRAc-p5<(kQ5O zOATJdza!`&RZOGcB%K5#Kje?%vp)BB1Y^`LIB)783Z^;`mmd`6ldnlGCr$d2=-1jr>`jmkAur@9dmUQn~plPHaAQ znPI$;Qoyn&htigm7+No`r>`>2o|V2kR5!h~hy3FC z`&2yy*?y;@&(PWKd!|8N3hu+Z9A{KGLVl=Qg4^~)x!JQe{Dr(@X-J-~Tyb8kTeEt| za=?aZSQLFaG|8tj!cY=j0CEn={@P-?ReDpj5=x~VO9M80)m$N9cC!O_tXn<5mez5>0xBFJE&1=|~xsai3) za?W#6%oTf<_Gje#&j#+MpFBEX$@{fJuiXSSL8}$I@cX^jV{3lCq{?P=eKQ;Y&I{EAc<7KTdtJQf(DL$vj zJxY@w-c6UyZDZ;KAB91;SJXpmm1cSAWZHOf4^WlQFj%3vh3H9*d{cM_7|%@QyfvP8 z<@X`2JrBCdiu1$kg3BXQ*gk0uyG(lZEl9osSIrt-2yU9R^Tmq8&fQ;5^W8s5)=e3=BGj_=&hMsCLQ$I&qHcg z?8H0LIDWc+4@gp{1~&>e(Z;J6GVQR2jq%YIOV3JA(yRR|J%Bm}q??KHV&fH=%0<)9 ziZ;=Y!d}2)ovwd;Ltz>5dC@2~`fnlC0a{f|ZB-nh4}pm(@>|Ko`e%#pNK>co6O_=$ zR3$)i&@l)c?1jzQdT{+!8?GOL$r;h-aJ`AsR7~Z)Cuy`4f^}MiSdn%u5w!(3hIWNq zjEEO!iH{0n12+rHyn5x;AUU50vPTPNchh(_3nIxa>K#$rysF8PnPY*H5n#lDi}EK( z;)ap8WLM@*HBDlR-}6~TvYa^EP-*7pY^4;%6e*-MC~kE^^-$6(UFk6¬=5}=vO z4Hcv%5iRmu)rr@ODO?n#u;|^CdTogSqQv7e99$!yzG4bA;zEr@hzB_$v^gTgJms-G z8EJcMaYD@an`{0U`JM?ey8HDL$PFh(%qlY%XMj>XphzF3>7Sj<`_z3o3&qF1EV{|q z%`t^t%CDgh%d14y!Mi2eOOlMB!*m%`ee1%PgFkF27&mHp;`w5z6YhWpZs_OeS!2Y! zL(l~kUR9z27ktuy%egtY;ykV_=3(P{qyLO)ixd{m7(S($sS(uEOZj-9N|fN=F1REq zR3GxI^vwgD>e(XR5mgQ>W3`th#Wa4)t5D(zt60$~C>~2;?}&OtNqjh+r`iNUfmdZ^ zOpV-l`Tz)zX3;Uio~RD+ukbl9pp!G)3iI&5xZjK>R_(+&bt|#zBEMCUH#}R^<)Tja zSPv);remcsez7FW(2{4%zcm_K@}x0$q__FI0xHCfqJ$~SWw-ggk@tPJOY{KZ1Ma&e zj+_)nW!mgz$A)>%H~d$3S#}nEI#%txK?`gBLovAtL21+n#^v=DuRAl_h|YB+%j;p@ z8mB|H4n3!b+TmjBvs&gq&*?@}1LMRoYAZD`AG?CTb0{btTB+dXpi8A&lMGcu+Xb8H zByv}R5*m-SH(C_hZEjtv6n=(!&;>WZMYY@_DI%F#7S>U1iDP6%mt`i zF=)%SLxx1>+7zK}7Z|UGhk}9=Fi#;)4H`TB9*KOEC5;X2@a=+<7~?nTWph>i9*=Ly z1NH1~SwH9?!;F^U)ePa=J%Lyk{S4z}D*&JJr0jCQI#M@%F4#&;j#8P6?l@WM#Eue( z4|K5C=wDc?W$LC^$=D%>}=EaHk`2 z6QaL3Lun6$9|(WOFW0`qzzG}h+qdthTYnliY*cjqUZpZ&(U&Ll8EBS#9*j-@=lCyBZbT=ezHG=gM*r7xHjvRZv*yumwY1> zOQ!pGiSuaV@%#{sKko}nf)!Vvv|X@IRRhbN0eL@LAkFeKwktdUPs1k}?{8yINbWIq zk=Wo0IooC&>rk@w%-j9@m${fgRQ&mGKOiwqY+#^=(~zkrky5Ot$O_C}^+GrXj7Scz z1D3Q!Li~5kjLDEUZPOqRxqhrRx!IMsc%PTP=-vn>C1aqO&IZA$=v-BkC+f!~5iRy9 zfCiFu57{@iO(bnXhUqsyR;a+w;e?EF`yVZ=u*_pRZJNMJ&az2XK=*rGk++2=y0*fC zqysek3cx=xD4ApIt{0XRxdgM|NBCNtU3B5?Qa!tscW;Ikg+H-fH^@=YyWE$(@{td$ zXHbU_OZR(0Fap&9{T_GZKxPJ-a(N`ufAH%T2|H-$ z(@`TR5~ILHFAt_R_>Otl+MzH2H(!o$-Z=fQs^Y(InyZ%mao=5XoSVbr#H;b^W;5Fb zN&$Ud&fua7S?TlV_d=hRdb&h@(zEAv1D&EniiGwqAXlTe6BnEcTtrYZfeWg#$LB66or?;!DdCy^&7oS91g!g>HkN|2tUTZ_R`Hr4f06(AG&TllAocX1=JH&v zabcdPN`}%=&VmYas^Xyp_5BB3)-!#e&|C|0aF3Vqcoi<o-2SfxCVZXyQQS;&$BBKRHD>vW#p4wV$CCj{gQj5-*$7gjs1=7P3rr*A zfdbBc(BuOxoahH+XEdfRHq)ix4mU$oW=~Wey^!C}OH~xmCqr+CG|75d07qz*=&*dB z0NFcvsty`EeH4jrgdLV=LH+=N(<*I%4tHq!mb2+f{1>UP*!d$@wTHeF*(1BdiwV)W z_DPMYlh1;ZCgErFkm{-T2x_9#$QLULL4y*KX)0tpwy_Ol4dgJuiHC2FuzSFy&&I}R z?ScZj38Iz-?%M;ZC$IKEJ(QlPU9!twS&&HC9nwcCqF)j-=M^+#k!2*BOM#9Pn>i5M z9^$~Sp%e)eSw(4(FNra}tIPr5p5h!@&2R8tHb!MQX7<1~+hlgjnCv8d%llhXOt$HJ zn}73LvSBz1nG-MR_m~ZRF{OY$clne?H}w<0n2>b;Z1+0&g{QEOyl(Rm4j4(q|3J3H zw~xe!_PG~FHp_PLiX)-RA#+x-G(45xEL+K7vqaq96|~LEVM{*SKw*6nauhGXG6i&+zhoHP3a?Ds;c|lYM9_#Ip zM){!2Lf53o2jEU)QUE(9)RCIt{Ab7DNBk~EI|zp18S$rEztQ?23=`0GUfZ1UkBlnI zNVAoIF2-7ni<_RPhY|>*(Gd98gV1LpSqv+l7;qq}r(Otk*w69lvEYV{r|);v5l+}} z-rREh%_@p-ntZ0G{@^fK;>3<|uGvJhj#8|pNFt@dRf!=-g&eR<ywj(c25Z7Wb{Ia|geY@zc-WYW@2A zH@>;*^}BOVsew1)bu#n-Z@{HP@QD9uSXmGzuUgcHVEx<%;g=rwPWKX?J#5#AehE*0 z6T@$D1O2d9|M!J*zZoV7y;rHOCB;rW`5Z9=#Xd?=L6M!%Qi%Orj*;FL^`n3W?^@=1 zSb-a~nDVTlyL~F?4ndx`erCU<$17=4x%?b&F__7Fu<~j4JfhOi>?7rLh7S-=XQK2? zEz{^;3(x7CRiMJHivc}eY2G^&bso3DE8jZ%Sd=}mU>FE{17g_G9l=3Prjc55y5f)j zT1|OzG4{FKP#vrXHgXlMhho$!I9NY*qq32Sb5$sa>df~o@cM~Q}0YZ z{jb&O4ykq(P?d?m*oh6AjpQsyfMM8Wm1no?ke^Y#K#P|le-CUC)N5h`MF-=BD@vZs zIrhTC@Xra!h}`iN(-ai9_WM63)f31CGfUqXA7WH;{pX7)tQ&lqQKIsZ;16qFquYdObsC>FXSjKciF~NP(;C)6^$ke#soq^{Y zybpsbfyba%89eJ;a35&ce^^Cd3>XZ8$$=^NJQTFd^GXp72JI7;0gEsP^crqx@+Zz65Kl>;*g2^U zl--C2?kIa~2ZWDE~+8huK*!g+_Z>{HjSw6JqK=OBgL_|ovMh;>kC3hlX|mi?AAf!#z6 ze(mnut8?#u?J)h3LfaMbj|EG=s!jG^t*?lM89~=(Y=`jn8OGtp=!qHOW|p4Z=jg49>z=Bgn#|I=e|-H- za{2{JMS`svl7;T16jvy68M4sdxFp#W(4xN27AV#Tce)u`7tZeXX;Ihmvw1ai5wAm# zPiOMdq}8z0o`<@lm=AjxTn?1s+NgZqH8KxM&B2P`wKa5{a0M*gcR(dgK3y-~!IVPT zIX?STcdnuNWTAl63-;^FNBuMA3a-Sh#Q=q=euz`|13ya_vj{h#`e7E z++apAA z^xB;_=QM5I@1#5v9$W%lVo976+lU-9pr%lYH55spG#Hda?zwiaYD)wvmSC=WrH2*) zTrg#^Yz*B{z@JTa!sro<9a`~waaog$|6^tT?LU}svHO=EGIHICak1QNf&73{^ikw4 zrNN#&t6;pS-Es;Xr{6S>a|gRM7y3@DBLPf1Q&JX>uqzfgc`Y9+oC>Z zjw(t6p%ej|U1l;nf-;p@nYn=ig}!@aKz^MMx_&fr@XmW~N~bJX&Mp)-$!eMN!Rudf+C?%a^o?6H z^J~49EytX;cE(B*u{L(2s$Edd>vz8$Y$%OshHjQvw}1N8T2N%ewm(qKn!D%GlYUaRSZwu9XQ} zEKBj87A0}sshC(13&2G(${!{8V?*c;8oGebj}_K=q`HDx=!dqZt0#5(aM%AF@mO(0 z%ZNXR>y6gY5>oPC7pI#hv%h_y-$yoaiw!$5W)7LTM7t@)4vLgf8jR6aOAV3QCK>jU z2R#AV22db1RORKW4DM2%YI9Jrv`L17BLy^sicz<=cy2EUVXc@2f^X3Nt}1j7_*90L z&sfEz#T$x*mA-o-Dt)19ie4e*(-0RWA*6ip-$1 zfl{Pmq72IDP_Iu9)b(Wl91obe3Z|#9SU{Jl#1;)mbjQxihuri1!4lbfDzfsU=ME1m z53=*&qu-(4{nv{9KllnN%YH7qHlv-yOzEHffa#po%{Izvz=!UWZkxUb8`vTJ0f#4X> zlwPga&#NX+C^D5%i^Gei6o(hkhg>_PSa^GuU;*z#U!D6MSo1gxK-}Q+^i1=N5g*m5 zZy(-2|25N8Q!)MhrDU@cFS$Q9n>_YXirp01L1~~qN&vt987XX&EKi74PKlm}B490y z6G*!Nh4tFQ$|E2Nyj*rVG*@-q6S!2Dy&z5ORw!&2V0;%YE(hjjn=Fq$CemU#>%AEj z(a5q&;2#gs>e)))PgOf*_Nzk}Egqgm2pky|5{x%n_h*L-_J5K?|M_35CFa54V){r_ zH)utqO}aX>fk2ZXppdV4^+IF|2&kcli&95tLC&*^Ha2?!%{;Heb262OR4u>?U#&>+ zZx(V)ch|#eJ!0dphryjBBC%N^OLgrTV!sEr~Mdw8*=!{itu|g<3 zp|y>&OOkk|(X~wR04RCoqV(B^mAGU$z~AOyHa$CVOp8;wAf|0MJ6=Igks;fex4uR>6Cg zfFMI3c{qEUTdVY-k9NB&_I=6&k*!@hz~Vkf?^?&ea1&DBUq5XAuE}&l3a=te<2gW_bnWXb-pC%-o%73Q(UvP_Rb zmmG2rxV6`2bc@>Qx}YV{h3?3t4(T1bH*#%sZ=|6q>Q&htpz&iP)jT?0TmWI59U$ze z%{T2}=Q=!XtXk2tm&0}RWkCYZ*hv+ij9(=S)b-K=W2x<+mS-!XZAPNiiHYM08zFq+ z^o5<4HA_zG8(XPDIvL$4XcQO=urrl4s-yJH$YlR4agFK`KZ#%IcE#%fQ^4Gtk;K0& z!3qxJ@rMrWh+=yEBO7)b$8HptaWg==Z*J6kn+(v;8pfR>Y1|Bu6EC|V96W?mT0|*e z9iB^R*7;oXju*!>Hv+L5V8G=w`7NITmz-&}ereL1kpnIT)0a+3@L$Yi1$N0RqCXQH zSJp`HL{^2is24Hy^eNd*NDOod(%fw|4puXA&fXSG4I;r`j#;zc31*aV-?6vtuosV8#-?Z1ET&39gO{`KJG-08cd4<+a3 zXzPRZ)3j^FJKZ{G4Z0wkV5j2xOjx%j@avfhvJKjuw~_16)Vx0U+?-E+w2i*E1Xz`R zFnI063vNi%t})1KmrTkDZ}DC%vwd~?)G!QxfTN{*>}wA5=kU)T=}}Ig8K3>DU+OLS z*Dnd($qF+&*&n18yt5|mly6i(@lmE~HAw#$dRQ3iLy91_^1sGs+_-}DM~#+C7cR(f z-bobtWDmR0Nn$qzoRT#PFu|-X^8U!XQC1IGnm$xfu!pO5nncsGjL1?^-N@4BQyb=jjlYLH$p?sSh4*241SGzhjf z%06{u+`{s;a~p2UyB)Xx|Ficla7|^`{Su5!NzYhp52}u{72-7FKz>`ouT!9AX!hg%{wT(A&Q-$ zQ)P3~{5QGX3)nXA8eQeJmp%ybkrcmT{z2i>mNuV%tb@%M&a2raOT$0W>mSr4IQtu6 zvmotn2_)+U%M_?G0mo5_DWJ$9AW;U#bEVG}|1PgpAmVp$))A1juOcV?HOh8rv>+W^ z(pqsfWD=}ol3&U3jdxD)yZh;~rvC~k2A z#OiGf5L>~+COOuluw;2&`txg78e~S`Y)?>JCEH#adsAovsVs^CiiLF8Cw?1?d>Vr^ zq>7v3mqs5K=Sx5FuYuj!OzujO0p~hZgRn^4GbaWl(2|0Wi&OmW(^_sC{fYDw>FsbE zLuS+{=?^CkA@iPWd;a<-Wd&j*P^5qRN+a1kgB&pVhSMl!Cq*@;$V=?qYh*K+b8tZj@grKFIs*($H?q(T&Hj4 zyMOrei@w&0@vEiFrR!dA5tq|>i&x1%_5nIl$e8UX7v0)rWeV8c_0hQKeC%ulM&pYe zYDYh5-(yy&rEVsLwVPA8J)3LqxZoxhB&YSOnLbhqo4;iJBW!1avLD1C2Avx}=^%@`aLGs=^`<#ZjXm97*FxT+0QJn6wG zv}u%R6KNNG6cW!%qfy|q%Qkz$R^i8N6*JzcEl_3|KkeS{Z8Rrici7mP1(tZjfP!kO z3?m{XF6Ur%f~96!dWT;RfpXCX4@{d!{$bST#YSpOlEvD%dhUw@s{%j$$rRWM|0%C`3L>%EFO1ZrB$p8goQQ|8-Is&x7)1WpCIfMvc?!{W*Re{@}*d-fdV6~^KQH5 zEsl3?n6(=S!zy?y=A^)$BM77qaPv9q!}0flkb#+2&XZFGtN8m*QFDr=(!1z{h~idF1Z^|%f@VP!n>$g#FV)(o7n&7^R};&a$Geb-6yee?VT zHoF22yrF1;P41lV_QHDi9=aAHGr3D|$*S}Pi?_qCf3?d`r|K3qEXCi~mXt1C6;Q1| z24-zguulpu6yKCZLtt=+BokT`MHc!dE1Tb{Sb9U;DedyJ19~O}nkN>9EhNTozvMer z--sMBFAr;z$T0_=xnQCiWiVG?k@fG zpHBb$;nK~oH@ubn-QgeWRA;{1vh@9`Z@t&XANX#IIB{Mg_u#BvsJOt;@7+b4WnIE< z(88$$rM=ssoyyX{&D@+<&v8!popwe|oLqWfaU!>J{;i-cI!|wo(3LVgISI_z1{+U6 zVs-2hU~d@CZn&sxm+kZz zQZ&09@*Yyec;Vw3Ii`RQxHZZ_JVb`h*BLKuWFwRt;Rh7X$BwbL6iK;XBDXZ~4kzEQ zK$tDqNu_;tIdZf1xIf?;cx5zl~s27EMxH!lgJpure$tkyxN?h&|#PO7B~l+Bp>otx~bdT>itt^ zCo@CRM&MI!lmppQHLr(8jr82$;{j@9i_*xgToX-tG$sI#Nq?p_9k7DObkE`Jcg)kg z*|^FMJXcvr3U8EO6v9#(3C3_eu7x0wj#=I=!(!Iu18!vyC)&HD4wTiboF}XqljpZ# zjil#yXS2JNqL}|S&xmkX(_ta|p(HRdyoWyYMvFK_@}b}WcLy+gwTgS_8rYvY2n(f) z!g_a|>fo#u&S~^UCR>>@t3#UQQxe$2yC7N@ahh!Aek|KZqm)pgc#yY+#DahQDB?ON zLpdCp7``TS63*^e&sW^?fFc2N~K2597psB25QydFpzRlB}%Z)wF>j=VASPoIDL zlIp0-BVuEKj2^9pP6dfCoV$x^=vuk~l;+Fk;i^2DWQdBW10L}JoJLjM zJEz{v{7TzfmsEBH3d_19Gn-;taG3VS4k~#+eE8M3j8J*>wXgqzRF1XX>A-Q`RulMK zq?kI2oI~0u)FMfEeZaNSA9kL6kQfySq;M64bg8iIiHxb3YPHMrd_bnMo*r;rA0DG9 z2uY(4ad6$O#!`5V9P@_KIY1wk1gqr%w|L&Cu#r^?t#2WHdpksguQ8u}Y11rlfEu9V zp81ByuDNY+SMLel6tQ(?moQ1T)%3yw$w^GkCYd2`oKsUMJ^#!y(5f4M za>P8X!a}A3E-;Xd7@F=1`8Qts!r&f>KIGiz{b z_R?GE$+nl-;a2D#>TAv^=dkXr1^&29CDH?eUQ`I~Kx^b!qnJxwiaZ#ig>+I>`skvu zeld-PtpK1B#FE!3SrAVEBDP-ajEOB)5SjLe^N$tgd)3e8>N&8dVqr(R)V)LU*yBJ5 zSQHG{Hn?AN(Lz#9rt&iB@l+ck?U=)`lPcvG&?SLtgU^zw%=T?_Yv8K;B?H2wnaiL4 z&&FQwgwSJVaSW$pf?&cIEu*K^>zCe>sbF(p3|XjP%M)!9trVn*y68P_`J7vx7=_bK z(F9~VfL5W!^pPn!`{s#D1?FksFG*0=LV>Iyt?m>5Q(TQ4Gw}8*?=FG`OOh;uY=LrU zl<-8DOl7^iRp5API;I2|_#yxHqdU{*85hB;-<`dYq&aZBq{L*wn@53<0X6YahegGH z*@EQxoywKG_0H*>GFCS4GAH>8J-7p?Wsdbhbjfe|NLVAvGA9@a0xjh^CAjU2Uv zmMzQXoF=%zi>WL1lG`pN&lhZ;;OMZ#z*97gKAPZXKkqTi)5#d?EGln_Xt%EsF@MOb z+D!I1Fk&i95K}-gz&?3^irTpNW>6_4HJwx@E7hg`yO}nAr9bM$+!Wr1RLw5Bh)xtG z1#b&G!~vlX@+m}W%b_{u{&oOYOE%JS^N~li5g0fy{TIbtlVmlX9k$8+9z9w%K@A{i zbf303W`&w*t3AK;Q}ap4VOKd8rj=}9g-nta%L~2LrGZJZHa_Saicx44qFxWxv#}p)dAQhu!*Wlyy~7S1e6K(C3XLXZU25w)WYY|?-{j5RMKLK9*-l0E%07n- zp26VHIne<%u+mgxbkmTFpCT`ww=J+JvRzj6rPLsee0xwG8;1JHd;x5S!S-+C*nd5{ zVQ^Son9}VR)_NH6@q>z8SIAa&DR2(FssYM{QK^@EC|?*=lB5nEPItufix?b81`vO=^po9uEu9m zuv(uPIX~T5y+@hlJOC8|^|TsQzqNFgw2hC2C@_P8RLE*hvkMZ_vCn{-C_qNLxmF#s z#11=a`1W^hCVa&Rrvvjj_euFnV{~qqfUlln&Qs(J&;;lw(hl%nkcJTb7^p?N2%0hI z;i4^Sa327c%Y-l;0|p0t-ABO3u&lr{15&n!co*q)5zF*@g$UJlS*g#a2q0fJu$1(Gwid#jw>%$&qFA;XLs=NeYJ&|Hw(rN}zI5Z$)?vaXAH}MZmjn~{qL%-} zJeAgA7t9utW3lQPJl8b(u|gMi$VFYp?I);NroUbUK1+_MS9%D%47>ZRBM7n%v`?Ln zk>~9Fn-yrMrTOf>?qYOJK2Tr&BZ+-!T$3!5>2L?dBvE7w6@_&iMG;lvByI_C%37Lkj_eS5d)}b40n+pOvGu((&DVb|H9{k8&Ts!flGs7Rfs0jvfNqqh zlu0qWDUwD-tt7_-KJhQ`ui~|b7Wl&|Xz}rY_RyoGKd4=HYH2AQD@gOp2&fTX;b4@% zjo(i?p%n42gW?k*_l38Fb~Zh1aPrjb;E}D7VcOq-`w!Ne^JhEk^kRV#yGha=nknvk z4MXo~K7ETYhagk1*{4d3Os2=?te@g-U>0L-Iv937V>P)! z4jhC9evDBw%QlMHN|AV&Sp;dm^}&#wggb+{!xzgt6OlT*B&1TIkq-vt|F3z`qxMhf z`CoN50;A?j<^MxgJ8;|;e1=g&ok%eW6p5pva9KShyCF~;Sf5Z&GG*3orP@FuXJ;Sb z>AP<@vfcuiKWX1{cJs)wSKHIfVeyj8BNjk}%E)_xK_FUy0+pGxI>9X;X!?O}fOQ0O z@&jo#h$jD0ta@lzF|_PeqqT4nDUXp%cGHLhFHlaI*p?#{bC@D|hV-N&*Hwbz1uc*> zkoDsh@jka&|7bzEa!8uZSwZcl*ZFGX+2Y}lt05hd4rzy1Nnko~9^VjTh$@uGPMFR4 z>u!42*Vp+%jNLaI_zSl)54ih7R`E8-+NCx&PFCZ_bhR?Z87o+f+xY*yRkBI*T5pW5mkAF-I^8zOMSpdf=P+#cr*m?FN$Hq>slPg7Ub8SxgctU?6$;yB z7o6(@8#(&V%H^oBIJjN53D(5w{enJ%e^`33M!u|%>~UL`Bzr_w2iH4a6{ZHGSFB!k zz-`&5kS>|2+zX||1@M{WP|UtM*mMum%J*OlufqN3UQGs9q6g`ZL5I#q{%d4+fX zMNfm?=-kQe3m-qO8vpdZ|_o_Y=!7Rf7#g8EgU` z)Qmo-s{uxHap$}D4wLMcW@%hwVlIj(<|su9K>Ei!X2B4zHe%#AX3^hqzryLEk$bd& zj^z}Dbl}=~YU#3Y)U@5J%n{gf?5zH} ztr=viJF{U!R_0{-2Z2t^iw=oeaX)7x7r9o!bwual@09Fj-2&K#Lv3wW>>%-WOID7j zafWI9#nI&?ncWQIxGOu*c^Mr<%cd9zt?Z$q45h#+oV9ae1<2W#Mi+P%cn)wi%o%AZ zUF}@S+f4>!)slK(HrmKEN!G~X6)$$uCn|r zHq96Jh3IRXK2{gYXk1_O%bV!)!GHee5-isy-@I=T~IRj0ax9_KLElS zGyBPHStSqE@E{3&Ef3gifDucL*=4H&k{SCfV7J(My6c zpRh}al@1;BbNta&%fsz>rrl?Zo4Zzo^M#*FYhk&6xD@On+t|fENk~Jy} zeH1I!K=4DKypQ=vk4ORNvShm1)Q_8STJ9GfVQU zz^{lVq?y|cMW09|ja7Gf+;oV)9@Q%sAfMK-6AI617i{rpouaKD76`G2HL(b)FK%@F z&|DG2f|LYW^*Y6X467kGy2SHta}T&c(E&GOUiQN5g&KL`{7aE)9p_Tyh**GnrC_z| zU7#m32uC8-)JQSpfZGX>?kWg7>Q(1;La9y<+Zs`ZWO#;r^fo@a3ppbEM5NTjUO*)R zS*viKxZ~aq!Z^(?N4yKdP=@J1$nwO<(m=z?CdRKaGSU4(SgcjS(g_U9$nZ?!x5d$T za)X`qci3Wc=Yls^QN}gWpxfXDu+XgjWe&VX0yXGS$!~EKvwOp?<`~Oh?=g)uli$DImcn!s@qDb^8%sy#4r3V}4oE+vD0r7_q z`)bK5c8g{Qo{%z30I`)~;wiF;imDa&k#1Tab6GJfJ$P9o@0bLI4wg-w9>$Xqw%*!3 zd1OisG+$nR?lmJSiWdBNE!pqDTggB!FzUrWNijfEe*}XgU(l;(uMKXJq>JJL8@_v7 zo;pjX$_*RfUh(;Y6BqbJ&^lR?Y<0j<7mZw}+7H2#Qo4`?#HKJCkBP9!41s%ZE0@3adGSs(HPx&Aj5*VZ-$n zN|7-fi|gDZS-vMobwQ|X%B%234h}H^@O1oeqF`JVk($e*FyXYL?o50SZ1+ioe z{MJ9LrMsQ0IqAWue~QxI$%<5Y8M!3QanVP7+GWS*;kuz+R!6F5>*c{Nh|A_7jXr)% zlO&1PJx7hq?HC^_BPT_V7uvFwSJawI!LeC*J8(SagvtDuOEK9L$)cjJl73g*Y_8{@i0BWiT7;QM_vtREG~TV$ z$fE@%udL!!a4L8i+@#>oy{dRfB3Vf)czuhC=ivS0q}8)UTo9(E`|KCInJiqaH&^Uh z%0BBUHczm0*eT3H=4XR@DILc_LxQjQ`G6E3Fb|x+%dWF^!_01oY|)z495(IewdF$d zB0&q)b#04wwm(i-P!Mg7sjjvVR*{l(( z@85Av@Onj}I>PAOY}ZQD$h{tWaWr*H7EL{rf+Ti8 zHFvf@9# z+g)Qs_3kyGkWdPgTYI?p9B>f>$t)_&)zF3(h0=bpMqa!iKV+ZZL7!gkqq`p4JEd;%} z%MkcUR2B;HX~WV@oe*|G`B;qnfRN%3p0lChO<4-mEa9DDr<$K_e4Q;WCy6%e4c2pz zHP)?_`EineEP@QwPk_BCEImX&IZYJD>7Wa}AsON(x#wnDxiQu^X4`8laADgI9;dh4 zLMmu7$#o9w{8&h?OXr{iZ!iMj#tgcV4r^tP-|%xT+Rk{GQ8lD8xX6XkpEmZV7fgYJma z#g~{th{N6jh}{WIQC^sRn?D8M88xOXa5U;E`yc-Lmw_sCmC=_38964#W+%lYqpK8^ zFY66E+BbuHy>CQ@X#9@ z^XKh#{$byaBmYf(os61cGG3+luj(&YKWK~py+mMHv6b^1fLN|$}RglELN<+ z1ib8Hj+;F%vcp7{@MiKt<5K1Io7o#l&J0p(GD)AH7>F4ZQc*P^V^iXi$KB%5vk2<9 zU5<#`rMa%#=6wQ&ceQUTKUT4KNxHmJyifA6Y#`u1eUNSw9uZ^aJMPTJE$-mg7{lF_ z;(C6oI6B}6)OZ^L;Glq{e<}^0vf&$U^f^CaB&-G3zO^mWnb+4PRTL-Z^X2Li@W57BDnEP%S)Y=gX{sIz6sn(;Mc5 zkPe$iVu9ijty~!ySezjO#Uw*$BRU|}2lHZqP7FOx17eyafElDXxGC$90=cYS$)&-4 zzqp=zRagKU#Ggp-E-Lry4O3T%YdGhGhCaLKe3yJ1>bi}Oo9CM}v%5A!6Fb~wZ~Lv} zMW@n9qf6#u`q%)lbzyEG)&~K@{9Bz1i+I zD{xHv%W&Zt^CF%XiWt&Dsg6Me3?;z2=zIS%g+jBY135lOOxsV@nPrf)I32Ap-<-3s ztJMZB=Y#o!JY)mtqVLlyNsAZ-5_PKGb1%#Gy7thuZ?}l!0T!7!y9+%|q1-7D*$d&4$*9FcZ|`243q7lk>7J@aw^6o^Gj6J}$2`|3AV`5g~X zf9j5Ney9E^EUU@{;_!+kah!pf%MJg-?bJ5QfAHv`9jS~7FZvF0t0hHRtlEvP+ ze4jR-vLnmzEqNz7&)aBf2AQ+#h-L;UHZe5^DF&p__ZbvHF=&N)NByoSbAH@aCq41n z?FILEumyh;NCXl^+462@y)r#NlTMU&LmfM;G7c>mb}E;2%!Qa&z2vqF3cL<*`@-zA zZnbI*#xoms7#YiBFvi^ExHa6r&8_|JEF)l=Uc2%xNp#>W0nk4fl_jv3VxZnwLq)~B zx*Ce3w$99_k>xLoR)6f-3uIhzoYUTi;YP1ir^xo~avydYA!XP-w`++;ekt+_urk=& zu1pxecEIfDE?MDsnsQ0CAkqjE*N3%9gV6C{9*0?eXz6y+-+=eGNOvx0Y!wlVf!N*J=#642oHbb z?$RWx@Tbx5Y9IA`|H-Z`!DohKD>(5yq`K~M zFQ449d%{Q>&pN{W(a2zX%nC`=?0?N>L`EcK@VDJ0abxA%JFdYbxhCk+Pz=O$cTiFH zJWySu11t)besr3UOq^=54c_qY?B}Iha#b%I;Ytjxf!DgTNqge&=~#~RrJoO>E`h}husvg5UV>0Y*DAlkW|0) zM$g?o6q4f&gf)N|uMd0NX9bAqlD%$;FFG5kfr}O=9!NF1NT>5~p6Mfn-n~EwHsA_t zwG6U_*Ce?-yFjMypd01LBbgL%Tc%%K=X2JFBTL+g5WC^X8mNq63Z7oF2Tq_*c2N0k z?2i_{=*&Er0D;68xnXId^x&8oz0ybIlKXAfH8UDjIp1i3NOPh0B~@oo9{1S74#^{u z5TsLGr2D~>!Dp_@VrI1Qdw9vrW?2@!X2!cZ)wZv{f9mTa&uPVV{5)>Gv{cx^zhbjU z^`z1B6i0?58=;96dZ_&GEo(JjV6(}ib=>G9v9KJ^lg9Hp=pm^Z!LlgD~ZiA($pEF-VDB zL?550Q>FQA_WAT{aZtvWDcJ0@lER~E#ZB2-9tNIE7OoQw@-9j``CCY$vJqB2uyN`- z$gAdp!aCe5<8{;bT`#FF%-R7g+Uxx4l)V9^qD*>Uu%0pwv|Y-jw}ktE2O%M%jF-i| zHAxwv31h>2GJ5u$B^pedvzwGo<}C}78?BA%?9Yoy20LrxxO=Un%*5K{Q_Mk%?5Cot z6)`hV#V(8cg>cObRAgAgi{ofrkUjVD(tarTXaOBu#9o4^c+Qa1C;qv-9RD2u$NZat znabR-I(H1@7}{v$n|(g_dgNa(v4qmK>2q^LrQ9~U$6ZUW=fFywj9-n(&vH@0J!GrUhdY& zy-JK64m+Lj>!M>_Aj>aFb`$96v~yMi?LivQ!WD#6cmoYXxn%3iW|tn)#3ats$je?y zoTt~JIN{geUg(?>xg2Ez4BQJ3q#bh;$ZDuyf~m7de1dLbHhSc8@q$5dz_{|W3FMwN zXp_03hwPqB*;~z1IbRar*+SxHxpXztPmspv7M;3S>(|cYh&~H>jB9x<-6Kv4?&adD zA8$4XYW;?tdbuOV8hJn24F&V%(&OS1mr~(oAIt=-;FNO9$ma0u`A9)-V|!&g+}OlI zJ?Z3T!|Pk-8BNvdcV}-TX$~AhE-^7ac@zU14B1fnObUQ~=XQ7&w|DV%VTRAqfO;wN z3F=f;=k3)DScudFx}nRSnP>1T;XPM&C!Xr&-cq~7yJ{nD+F!?n|Sr$?#) zqdhqyoU5`R|7+MOV;-Oh=)wx5?I0SF=wox^@nof#?Bq zNrlX9z1&KLjqS6=?}PPMZm}oZ-mv_YkF}}euw_0!7fNsV8Szo`#+D+omz}HZz{{&j z6POoLOg=>pQc>mn!=7pM75@jWD23Ok!ixDToO<5Fpi8PDCoQK|dYgNLvyV=3eF!v# zozgn5!=6J@?Sk|2hja~{=mU#?aCfgQN%O0P@IWu{fL{>T1Mg49yjI)GN3ce}1xMDl z*q=7-XZwG=(n_+^7TM7J9Pg5dH2+SO;V3q^Q%6AhcViUVio?KQDJ+bCu$MdDGJ2Doe?FT$v#xmVdPLk?3l zX5_We&B89*3TlpL3fRDCyuNICzhrpHZ(no{eoT`>lGqJy`atkW zY~~UNUhB4-VChMXW*gs-;{Q;3LA*!Y4#bo?RXQh|(=5E?ewMk$z3IK#16%9>hCxWY z^}is#Fgt^r;d)Gz8(hL<&u{Y8sp3S2K8pRy=9Rd79CX;<&=V?&YUCKH!%WZSz@|DeN$TD+UY==$R-D3ikB#PNWkKjw&dolp-gosPh4* zmTu-^x~@SGY+G1Kh`PZ;quj)#L7uK*_f{jva<(GK*#$1aLDEGxxTn!;Lx-KTv_@%2 zpi7b=Q;e3*gQaqXceMf^fZb#~d8(vmloYGrYy!>a$wJP046(EhlY|m03{9VV>-cJO zO>hfh+A(1{f>;*}T3mXx8{cB8x_=d5tcB}X z&)V}EE4WU7L-7Cg&Nt3IeLrKn4hqx!()<#AR+8Oxk6$6oRCmZ((k)3Ook6IqhGZej*Q*|i*Q>V9 zEOg!&5o<$2v5^6{7Bu5bH6z{G>^CdWOshW=dTWspH1C$EDo8H7-9QJ9ww*D7STV&w z8p2W348m1U8$XWR3F(B;KIATyh2O`Nwjw=04XB$cj*EA2iu_9h4VAL_b0Dz|Qw~rT zybB~xHSz=Gq%T~o;9T&(=#P2fjdCbrGUTmce?YsMBUsD1K#~=pB{NyShV^oM9Lznr zJ!#OTM4sZAuY)WsH&899i^ie~{DOv^N?f+i?1a}jjn>(~^<(=lvHI<0+wmmZvwC|S z_vPcCdh1`zc~33SiyGpeNi%`F<+StvlDwJW=Ow{I`>IDD$&Ay?_5{UMvTZD41qTkd z6`D*YSrh|SE}e=h@X8Qv<9@WXjo0U1&cLu%Urtwk%{asS{!+^K$*CFSrU@oKqL_M$oTsANIW_cQzj(4i zhAABFZ)znQfYJ1f@`AX6bBNmx3ZYqa4)=_*WbSQM4tJ~H8rK-buFbIZMYgik|1?25rdvUMkQjpX?GNU6&>ep4rvocdHDcB$`kOTXS)s~t=lx$f3(mag z+slcrU7UB*7iM$c!=px3;Sa7Y)e9`cC4-PM5&)ODY;!G~G9pXXuw@Ovi9Ns$wr^E_ zWubX0^|KVGjvGfN780h5L>+T63kftyMCXKt?LbhJ5)1~Z8|5f|c|0IT)F{Vl4A5Ri zZ9~*YE(j}Ad@hBYH;BH$j-pP_=<;yUY3B|JL`QV;W?{nYl92V`Ni*4-x#x@~YbZU( z20rIwSl_Y?rL|p`OJ8)Of~)o>NwKnrr1{)-trFi{tVVfc*hUV?q>=7NBmc`GPO6}c zGj-W(tdYQaOegk$Wh9K@+)fWXd_!&S+_K48aaoaDhoC3Tr|dxS#_s<&83Go^b+Cv90+u}_-uZ|)o0}}pbuCv`iZ?p1nAICr z!ut&5LyBgd1b@Ae^GMj{(k|T#wK;Kt4W!I7nmvKw1V+$0jF?R_6ZjmqkNnRfm%ln2 z7eF;%D*qp{`UMNIK`F+l6?`JaBv2#{ixdxY%B6$6TDn}?8E{WlM0HAIW*1TYVfiAP zIuh1=;nohmHGSE#jpZq(==ll?4*9AUS?=fI`|R4GOQrkY~Foh_rHP=7mS zMuGICvd#3@A;hO!;_SK9dX{1Xv(#Qn26rA3jec|g6vLI%7Nw|IawdOx6j5s>nmA{f`+2P26S3%V#I66i#5bV#VqT*l2 z95{8F=mcm}VU87MP$8QE%2*xs!nkcOcN3#V;`Upf7`{cn7eh(|t{(;@gk^dTJDs0d z;&RmM{7h_o)C&_Vv&C^d^-AYQ(mS5pyeFUXHfmIDi=0s%Px7#&^QXZxjmWw4>*l|b zb?lJiz!perf}CWEfh?@8P;d-9VtcrknEhhlFnyyGn8bDg-SIwpJLjSFB7HIN-Ye(j z`6}47|@fOZpb|awouv&F1TW(v}rI40FwsXYG0L)9p9@&BbUtK2Tr&BZ+;% zoZ2iC3$cS@k|?r;imG5rTwA@8WQS+xkrJ04&njLur;1ljKL{%g#A1uwutPSZs7k0QnX8VyBom4SJ+fTN@E^+Ns!%Vv zs<889$gKs@lLw~g;3sQlG3GRlY+&KQ8+R569I=GLPzZ|EpXcPrSKTBrZ~{X;#V*ec zZq+bb0olb?S+*>1Y02Cyh^WOIScq2$@IsCV6&VhM*loMdTC|v$9g`FLz$ZKM8~MHeet z#n|#J$(T3nAHDd*oan_Gd7E1USH0i=U`QJj4?`&pZV%hsQ?Z$=Vrz6*o{AQ4_WpFi zNG#+p?Zsr@3lAlDhdNT>A_lnPCCa> zO8hBg>#DPYaj_E%l)+j^7r>M?^}$(Mx+CNOw<}~5V^c7UJ%AH~$k=_^@~UNs^v?dx z&;Hp6k=Q>JKP2a0nq8WECg5qNm}?ZdLPh0q(>b4eRq>Dx_P=k*%9lQrt_rAzpKbid zi@Ro5aP9?Fi!KEsulZl@5IoE!XX(2@v9NXqHb9ByIIn6^>HAazsh3-Y#rCM*M)vm`J%J>>D|tkffoG1}G(SYv^L(Aa}^=mJEL$aKWDsrGwlwKTKINJhNM(W1c>9 z#JAr060^(0PBzk7cv+rBCWRu44|~XluDs~v!hLUk5YOv-^RIij7)OaGhdGtJ_0o0@ zGN0&FtKAQ|w96BHZn~h3ANw$)?#^}6Cpi)$)y5-pzlInw3vONxSx)yfRx5G8^ z9$qXb#j%C30oOD-dhzmO+|Pt< z5L`PoA>>0W0rg+F;zWEc0cJ!{WliLgk#WM$`k4_-O8_xu3!^ z1PRM)$%o+uVU=Vb&?ejs)v5Bhk3a^n93ovGQSYB+x`OtQRG;*D(8-6-a?8knzEqU6Ht&dolk!hR1 zzdPJKRoY?a9Sb?!w>+-}q>EQWHQp8ak$V$!z@uH71GGAJOB{L5kXea}=k#EoTdYtq zJ*Ft~XI@78@zc6#mq-e`$;E+VA;(RuO%BCCq1Ik1>IBqRB)FCOj8ND&@H^7m!CZiv@}?gM|c zz$-C)P3Q_voY#FIvC*mU3>LJ`Q?|HaJK**h+_Cy(WlJdDoz^&W?r}D`j}Dv?Yk{f! z(>aZS81ZN$BS-INQ*EHM{C=bAu-_5ik*l4mODdhJhyEC{0CJb!lC7j3N>d~s3O2j+ z&q8tZPv_hbHv%_dieDeSk=eGmQ)zSPv4z+(2E}+`ESOB@+FH#vV$O79fg;AR1W;pn z98$x7NFKNz_P=F+zRYL;dDeZI&pv1GE3DpE$6XhGvz^*$o?CArDgBK9`G6dtPsHDC zG?FmZ^C1t%wQQ!=+ve}kc*AEMENoubGTxsxjTI~$*9r^E|LafxSS%)~K0S20bD~c& zApDNF!K;shLgb%F?=CthUFWNjqbcm*U*YT*pY+$#`voZwU)QNpW?djp+dLE_FG;Pm zFRV%l8EML!z;uF7U-W2~RVorW+69n$AAWaH3iq)`4P7qj;2#Jn5aTCJ_WiN;b+9zm zSiC+i;Gf^1Xw{CYAfp?b;j?!eImm9!Xh~FhdoiMUL!w3 z%K1sMa()fH>Z_P}c$(Y?aXv$t*s;ZT7j^OvxFjlX%63Cp*n`k$!5Y}rxxMgAL;>lR zW%%^F>Qoq~xwd3btPRu1qXi`{G0c!tga3m$G0bUlaCT#GjIEMuZ4aR51H;<8f1bY5 z!_x?$#$Oy=PLkOH$blEl$4mf}O)}! zVh@jzC++(?X3ifzGUr8d*r==prlZ^8b)r~7HMf(GYuPM%1U>kvbUVh94`?n~j@N zj2H&SN~D;K;#8biM#``LH1}g-bUqF{rJpBjUmE8l)5Q7MPBDoTNuZ*#l_|4^K^D?b z3389u5PmYe*iXMNoFj-QCtkBDgWmX~Yb9p%%a`$cvt5f-pZ%iL+|{r!M;l(j9KmgF z^kPgjM)f*y+;FGk7L@oK*xn1m42eIJT?bnGFJ$w*PUV+t`CWXWifk^$CH*ACXgGw< zOZSjN4!nDF!Ne4lQw&s^pP-_4M(&Kv5WRoeJ66#yJtj#e)!Z1SNYw3FGWR*R8X^3i zB-`qY4_#Mm3Cs41RV2@TC`}iq3)-dF*%|kT(st=K?jY}udyUB8zC7*P1}^|>Y+DR| zjJ{#*n`OBjD{NDovun@1WrWb9uYLU&r1GVidsNZFI%#mnl44NTrh6d>a)^7KgE^7vQvcmdXHbKClVq*np3In}0wfsqbTp9L$0#yE zYIK|HO23Y|NWz1A%((mfq&@c5E<;+-b+S~@@-7SC8LaLNON4k!J-=L`lcz3LH*!+J zVrP&oKs&rMm<=-R1HsX79Cd6NanIb*p8a_B_l%%Qrv~087iN%8Oh)4d#WYbwM@3y1 zUKj3FritJy#niAgsQJDCQ3>^ZIH>doP6@~|uJp%N_0Bs%stMz!L(=2Um84uT?1b8^ zI866=v5MVvYp{BSpkCU+FPnFOw3DL&eX`9y72e$hNuBledno^kUg}}LPA~Z08z2v@ zLz3@l$U#C_p!-_sk3o-vLQpr+F6f?fmZWoX^@6ZFX63#Lywh{r+{PWGDIFOWgT%@> zIPP}RZISlJKQIDJ}NUdY``{zQ?O%3j{-+f5aP6*4DGe8u_r( zUwXl(M)5JoviwUaZIB?tpU_JUoi=l8V6_;HR~o%KAjc(^D&?jJV|t9b@3oG(N4*;5 z@w~F|I9`De-i{<$iA(O%&D=E5Plg14Fl`X|Nt=oJGtYI;|6bTPyavx0qbCIRB%fD*E9X3R8MKYCNz@(I(Nc6 z|AEb(fdg+)T1ewSwN0GFzJNr_XhEeI)ma-r7rukvF!gnb)>;^=O@oaZu9Wtue`EAx z6#xFpE95r2)tBRXXk_*D(Uk#x6w^zQZV>Jbj^p;idJ1{)OKDUhJnV^He51V13)TQw zn3%}j3}PGg(nk^MJZS}NNv1<~_-A3C2MswD&-payG)d(46MP;Q^iILg!vx0}6u9ALNbqWW673puZ(Wv6w z2zsp}pM6YHH5azZ=T~sDLk@G+xdW3wb~)rUG`}7od~vSUdmo+4)p_9x1nZ%8NbUu- z(a1esPOn-p$~&?ac%xi+yB?0<_G-<*4u2!?7Nq?xfn<%v9Ol4Qr^>{sK1wkK6gfmi zT_tHWuJ$k$vxe?>#o?;vY8K{@LmroAwF#>^^^yW_ovI)7yU!?wq${WvaWv@p;@#D< z`>$>G+2Uqs(JD@v^@Z0x5CKA2AS`}8OYSbRu?J*3c#I6JZGKz!e7?NAr^U$#oZQcU z^A{357H}NcE@@1F6HhUlD6$@+LV@=KcB}Sqhn-NT14Wxhy~VcqQmyO78c0~zlWkjB zhQxR8ynUZHVoQ?0S4-A7@M0Y5(?%7nZKIg26p5#zkkR(Gtdf^4Re@=z?U-!}20r_&f+b@tS_GEstAHV>Vq`I0ke}Tngt8*o(ADKeTL+GJ=>VzikT+ zlit`NWjlA)pUm}u*es76c)qcq6f{EGqdxDkH>6&Ah;u>tSd8Kh`b}_P+Xjl{DV>4X z0m)>n?RxUVs{W;OjkAk#)4DRUi`^2*fqhuWF*&0T?1Gy=#=LggoD(V@BL#%Y7>3B z=-b>Cf`YJaKBRbS>E!vHO3dZ{)E#3VSNyY_W2vg(s$lzNbDNJlTPT@7Z?E&OZV&#w zd1S<4?T3XZ$T?w>ELl;{ui|#o+hzlyB-sbjgN8hqQ%m;;U|l&lU=!=Md9l+k%3f zRf+iFp&90kgANOvT_&!_W{TNJk#$rQs&)*xwFZ~CV5kfzodi=ct2{euRv_Zpho;~P zD@05?{muQC&0SZAMT7-s6tiV&6p23S{-B!>;%<_}yqYOiW3l#)hyhtq#7U?OI0Ti7 zD+NI8jzu9FpFuAz{h{wwq582$7mWcFY;lq9C-&oYu>voqhGMKUJK%>EUZ(F^v@ZHj z#-+}+AI5}|2QST1XN$>h(CX>Tis@vCiaJ8p@#?u4Un!fH>T?m&QeYC|-wL`ae7NYU zb4kd4@mWa6s#aW6cF8W%ZEn>PohmNyiX>0k!tdZCN8OMtj`KjeUwnbYEjBz;tys&+ z<8Fki*`tx@=oqeEq_HQca@6-U?7vo^Q{4{lk|jiZChQH$;}#0rpw>l8*V4I?JZ_!W z4M`@A!tgp3)XheA@CRhM+?CFGu#1%Fb2DfK536ZM{tlfgS>6>?Lyw$~{9QU#vv9w7 zbpT$+=g-U2ouB^ozp@Ev#{%*sukjSGq`c@WsMg?2WwtLQL198)ng}-oOI+5167y!C zMkskypLp%WYv}d#%G6su9=Wd+q>H*>^U(gY-2?~S4z|g(OIDU@TK=z7MXwqym6J$$ zjAXtvmgJzWuSb9U;Db3>chGA(Odf9jja~w4Ca`Ln| z?%kLLH+g?9QB_a zc4K>j;wstZz&R|1CYB(JVu0&Eor*Fj@@|9bH8o1}o(JCm*p}0SlVshH1*h(q+fNES z@}zo-m!n?wf$_X%VTHFEb%)Ynxl|AA@AjNktOv@NZm12RVcHcdpiFzaL$cLe*N#o5 zf&=>}=_at+LNS{uvXP3?_-q0xR>+6iOAsnoB?GdOrD>ue>69u(=*P@j2$>GB{m=ez z ziu7+^X(XFxkOL+!SQ^Fbq)4)10ic(fER$&cA!07b^1tK0bM|RqzC7(8Ex_FM)nkqTZ0gx}Yw+B~~GYasYt1DiD%0>|11?6a0`fZbs{dhAsw)o@3yXDbhZ zcp+xdKU9`WhNQ)zg>bLiGaJMU>mU-A&g)y47dgng7gWK?<6*!3AVxJLOJ+)VACfo` zc&(fbW1)CLpnC>|lfCflP_hD@_Q>xf*(VD>T6i3o6SlcFhU-))FtK;Z1IYo88zP7JA+RN_*Hq0DB z36zOyX^pate(Z`lmQy>~uo&xc$H$fpEYCD!gx;oQzB6;%i(Wp~Aw?R5E_aQ*fG!DC z*V3u8&IJ^BCqf-_**pDjo&Dzj0s{fY4j6@dzbU1RQ_!H>-~}?|JZUDJ17_h*zN5FQj8w=6iJS#T^7B#D6$$O@^~5Zk(vTA2Yai>M-aT9GOUb`4pVPo zsvyD&9n&;F7SA-#wzZJdn!>4-mI^b0_b5GBr)qG|R`B zp^)NVW!;qUF&x=XM3{qzD(&keP6lm?IqV~~;J^RQIIQUpym+yJswTG_a-GRv1 zCHH#%kW|Y7360=r!6TC4mKd1=q$d5Y1Maz!a{Ac9^JF8hP`(Lj^QuLLWayLrF|(V6 zB|Lk)-;>9TePHvFV5TMW+Kfd;%)DEosvx;9&8ip{!J}5i#S~LSk)u@9F>Y!^k}PJ{ zHZHRMV-|Bc9XG3B)=kih8IqMic~B+)I>7&&Fx6+DTiLwR1d|ydWK2qdT<3Opfq%Vw zr%L}5$mD=(Rnhqxw$sRwXQ7|yRQO!2|6$Hi+m#jBjL6jMr(lc*b$%R4aZ6t|1+cB$pu@*JQ?_;F5$T$ru#X#=@ zS5T!w>-|8oo)l4=RI!SE^bJT`Tq%iBrUN`MZ?=-$3#yI4pIVol{x|2X5)@JKk!1l7 zBt_I0Urr{K;TP$I2#vgs+iDZ$L0h9j4-gwbgPm;*IoVb>%ZQ<-*RH%v5*-*rkUKpJ zLwhL(^dmG>RK0ViAdb@$p{{qvC<>l5$$BCTo6ebnjocxpjlw*^kW-nW4}KmJ9D<+M z6rFr*X^&}h^e9dkCZjvC!!awEO#igb_ZKfZOfr?FG{#e|F}<>TP*m0|)^W=gq`Iin z=#sewVOoGrqR+5Xbm-;TFU(w~KB(+pTdk9GuhQA`R&wo_3pf$Q9j8ql~UdoL8IDT&Nd2Tm$D zZvy5riYcaGkrRa_#Ql=Z?(w8IEJsjBt6P0L<|3nhKBtMvqRS)Vd38%sy8`@|EV`5K zqKoKzyh{IE?_BTC2;{A-;GOW(^JyC-E`U%L=op$LXXXQ|Ex7}x5m^egE?e9liwnX! zLw5P+2yBs~IRUtjX$j3PT8HW9w6YBMSkceY|HFu&*mZG{q-zG*V3ImFNHGHxc|=9U z@CrgYA$-?CU!);r8Kdi$oIBU+GcfqGvAHh;ilRo{q! z(sddcX}jo$ArC_`Jl6(;o`MIG5o1`nNpeGc0oKH6qD(=VPVG` z40RZ)$j8E+3d2XDRs@uw2c`yZV)_;(E3?5+4!G3;6*2zxZEnW{VCTS2Luxj{>`yHZ z=@>i1j%xwFUzu{m$B4h&;eRTOg%651;8d6G0+-`OpqOscGG%B?Ha$EvK#@>{@)3S=k=4K$U7mu zQXNS0*xb>!6_%f4X(n`w6_%%Ega!}z86i{i#+D+o*MT8ZX#$x-ipi(QK`IKZ+kXFp zAt-01H*hE}*e**K^@ZuY?uUUGaFT4J2j;79_Id1FAk;3QFR?q@BF9#+0V6jdcvyOf6>O#-m@OGL-_&H2n(4UPZxBl!ZI_;~ zU#5L4^Z(xY>Yi^Key;QA1DVjYZ=EIUWUy5r>zSWSDmbS=b2{mV?{B3*&nxnzbE41o zS8M&df(9hG6VVQAGT6{?Zd*`28Pv(6>q%EOJ^1OJ`q&>E&CK`DQ(q$=Ik1_zZ-RhZ z6a$KV%~VvXPZCr#uL=Evlmu=e-I9{Pn&lk=rb86(92Kq0339a`iXvkTs!~ zNvG-yQb2Ek|KABMfH@1>K9b%8DTdszA*X1;Ju&u!$A^3#%TvStrtP)o>DElAIZ&*5vg*K{3ZDQbMr+n?ieJz2=V-3r^_L@CK`5X^+b6Xj3DB9Yw#6Gg*Te^J>qsCCS2{CK6luu!M$8r z4J30H7G{XbLH`cBeJrU|s2kmE8uPVkj99#wwi+>ZW8|;_L)EW>{$-{SJa>NG{8zHh zfhR4k396GR2DVGKQc*W$yOo&OQ|qx}mO9OEzjM26eRvl{c_6NX(eIHE!_=7=mY}s2 zCdciE9cb77t1M3fbFkt9F?Gb+lLL@`?^vYCoP zMZHRYjC6MiQPWz>LHA2-@Oh`)s9298+oEK=-a4AZhs|jT9oC>&AR0u?LR`ZvFPo|L zZV;}9Ert$R4anhL6_y5~u)Vt96?Y~|T(DNUAPfU=gRpC`S6LdE9y|c0jTNvDvdiHFm|CRql8=Fx+QlQ>HKbV{Y1u zUf5`m8#3_W!j1rH6SYA$Mp?KTw=S`UsmpEZlp)w|z*{$}wrIr;86RAi|Ld2Hv&~QM zUQ8t?UYglPXENKIp%^eqRlrjM`$S2yV~ZQ+@0gYQD()dYn6Ht4>W*_xlcY|+pAE;# zTy77It#JPd_rZ~e8Q1pc%#^s~NiV#1MiRYPBTt%Xn5fWVnNR5mB>yWL-DRBiT34Zh%UDe{tP$SM8b8dWz0 z#e(E{hAaE%if>(8y5sAOs?YsKTBX5`c@2F*Tpu_Jof8ITfrty``;=xsDcgUs1=L^GB# za^L_Rr1g!Om=015WU=m}q72Ck-E*pV>PC6HJVr4Xv{#ASNjai(@|0QV(5T~i2SN-A zvF&p7TQDVE$EXL~9xhrR{@CL{2>RFc?t1do=zv>6qfCOW#$i&hcw$j(_18S@$Fgbp z75BEwQX^o#Q~jeuq}YM?0j`;V;~d3&NRiVJy)|aLtNUK-n0q;}LxK=VqpyPq$xiM; z;ecDGv_GiMtBjO+g0H&Z|7Y(_;F?O${c(?Y4#|rl8(?w_DiT4E86lWd#KxI!b8okM zyWIb6e_iHIXJ|X#-nP>=wcDNIE(!<=E~o*NMG#O#M3jAS8CP&naS)Lm7aRmdWK{mo zLxPe>GzStI?HxZOCuezsdB1ty=Y5v%x5=f+WrHtdp55!D#}yByz~dY*uZKd?c=c&5#Wl+FBDgyp zJndEBLoepMb+Zk0o?yuG!BbRNbz~VjBa{zL*j+DqMt+OSO69j@c#mFww7i1*Ypmvk5-iwD0BWO>z*1=9!6H9g><&Ym zP*?4qJhMrF(gKIoSmB3%FtmsRKEf@zk4+0g$BY~kOT--^mY(?C57mD&Phh_ymc~N5 zJKLC|La~}!?X@>GT^h)ZuTPe!~u4jio8z*0f`B6x(0uY5)7{ z8lzSDUcsIulE=-eIPiX{+Qh1SM6qQQDW<~W#Dxo>fopbe2>2rf)7LUOEE-1bNRWtJ z9?%UX!#4w}WCy%-D2Nv>>;!kGldcc!q+>OO5Y}0xLdEMnDwK@WN7t-(MR6~CpDv;xkxmTI zC6f*O$4(mqn_%B`eXc>>NU!du_xtA9w9oTJA2+ATe$ikyyw1(dJ-mLdgfSYJ-+bJA zf-Ii|QH$X_*Nqe#Ly;&dY|W%3rbLkL+ooMTsZVq`xK$d%PoMr-u>HH`pZ#>K{P3TB z(9UN#efW;MMfk(Cw@zBwgnQ8=NfiM*$SP-W`1IcU=RiefzZ^^GGt>!FQZbx?4qqxn zEIR3x{_&H?U|+4DJS*{GVq&cyee`#7!iVE-3I0;HKy7X{9Jb`af;o#SyksHUH6~ra z{S)BoeFP)o7^*GSJS%O4$&&WM7o)Tb!yn^hBOG^0o7VZ2F!NqXhh5Sv zL~wC~o9UGU@!BTA*6)E*K;T_zm7r#NM!;QOmZx5s06lSquDN1h-6N@Vt+tylmA2_ zM_TM0*Tsh}{^7g4?G(Eef7!5BCJFp2lW`esz6bm&p`ysI;j{!a6Xyyc65?Cm{3~9!L zA$_8ao_T{U+1QMY96H&OWY-qiLqTz17T|6#FbiYK66^Yg?#52{f~z$Mo`(lt0U)M(Bv{h<;6CCIJ)2aI>F8u{$V|K!s)c zoCR&7;^3j5HQKwiU{fQB=H{9>ZZv>onLu4nvD+xJg$mmVv7)oiN#eoC zSuYK(B=qUmVl{i66BqVLYx%bq>Xkd>=@3GaT%h-P=1j3^0gDAN!RgW9-12C)yYf@j zwb#rO&#y>niv>FAMtK298-L=F56X|Jy!)hhX1dE2sQfR4NbhRkc!b!ZXgn<6lR4x9 zjVBL{!+|*K_VU?A+us7NiOGQemK9;dV`h!Tgo8R|M$vq5A`x z=w?N=AkQn69;{s{ROX0Q_@sbYzzUWYc}_48a~j%ZG5%~go$uEt)u8}nAvo6EidBKz zows{HUjQ3?Z)Htv&Vkz1W zMEIScnKM}RKO|~#+#k4Kdgm4{s8LokvAp!@khn5*1z;L5&U+2PuKa_ZdN4QGz8_0s zHC6QC;Kkh#aIvAWc{0Jau{uwNv)RGl6nj|n{vs}Tap1~I3!O*Df@@`U(oL#*5g;PM z{bJxv{yG=4uD0TXqhlk49F=`cavgvM+iQu+YyomeAiJd-gnf0 z-|Qyk4W%6rV;jKOeit`KKJ7#E)HO^&oLouWcL@6I0iyiL5JpEcStHd(xhcSy8NRUFu(Nfzgct}5EJH>ke<`0SUBzq|_n16AFVdk%(V zdz@kn_q=yOTdVCM$>JW;BE!sICa>6~O3)vgr?H_OvKk?l_QfXWZHFDpX6BEw&A&8P zx_?DH+3hAKC7NQ_QDhAyD}|4O79-D;fn`c@z{?JP=!>V$DyY187@Ob0CY78p_di}L zV`9N4ImATr3ne$@DfE4^bipClPEi*`RukzApT3~Rpmtdco6IXy+t3!gaAvGRz>qNM3VRh4YO=^XIk zfEqLe#X#7#5jY8&p!mbkthdFto0QR25T5{!Zg|6H`o3HIWL>N#@3s46FRuXDQ<`P9 zM&-s5K^MJ->XApWiF7TAQe}$M=bW5!8HiMBNe-lL&N?^Cs@&5iH%$huKw%s2P{4rG zUS1Lmjc*ne@)A})k5Dw^xWO#RpS@e;M*nwntp8=QiJSlHxc@R>rypK6kw&o){Y^4d zo1gRVm+O@%Fxy0DidzMjrz|7uc`H4yN(P*kyLAQ?v&9QP4@#K@As2&ASG*kA>QLs= z-aaxX6Eh)DRJQZ&=r%ZKTEODd2oUXLz?CtK(uT~ z_~Zqbd+8!wl2lCsFs+()=s_54YVL4H@*jg10bUqzLZ!aV%u*qW!T}@Jf|sBw$cbb`sfH&~l>b1GOqR)^|J$k33YZdRNI^Eb%S&>nou85m(WE|Ft^ zh$&!R_-X_kJm-|JDZVlyW5q z2D1WxX(CNzr?g*Q?1EQU`XBYfeqRXiw8;NbLa%>wweLkTcxNn64>Ie0uGN7!gf7Mz z`r)Qj!otVV3O#SSYzv#`y%V=<8+ywA@Gr(iY+}#PZ;_*~jJMNlvgBQ$*z*)QXHfY& z0}6lG_y)`!S<+tLsM&)#?j#V5Y$Q9}bi1JCuu7IE8eF1Iv7OL1hs3*RS%C6mzi$j| zDE2X3baY@60}88B9PX~`Bl%+ZFP$!MZ%N05>6@JCzQX=}~!{##APaYbF18!c9=P%z| zW?mk{WubE5{gs7c8f@QPCQQO~^|nx`%)}H!JCORU@yQ1$WeRP$%vdqo98hG%XtulM z_Fr62zHeNK{QrCERMPg6wJWYL>CNw@*dB^>fkHPm@TJvMthi|UBYt7XP9Sw}<28{p zN_0#T=_c~&!n5iNQ@Wh{K>)N#a80p8Q7(<;C-G~wc=ixTp>Gy{$}XK!5*X>(MaS{7 zR3~Mb!g$_AL9xp=-{*y5cS74|Wd?X~)+2lnV(H^9XOnL&SY;@PO&8v2wsOZlACT-!YW}Do2kA+(&r!nTY~T3ke3%U zj|DZsT|g|edQv_q6n81sOp0UDg1hOzJ{+W`NoRUtCELo}O06bKE^Z0YD?bzU zYOt;8@S-KFzY_TdgmN%n)#089j`8UqEbiw;$m8YRf^09SE8jM408+o!gj+^E31h&5 zUHd&P+ZV-Ok1jTMiygK`)Pf7#BnS_HRKM&LdrtNRILcoG`Ql zNv^<+pt232y*}wD-tm*|Z5I$kqr}6g@n`LWmhmvcdNVQlcQNg1qt9|O?8A*@KeyA}Xv*x|$yudRYVlZ$^snR5}KB?lnv=X^@0d&2U^Mc&|_$>DfCs zd1kxpoacMB+U-*^fj=HuDe(FxX{S7fzjqNHUz23PPP{>q1+S1;@4dU4R@tMV(@ZSC zg}DNM!_y_a9SVGAj9$(tA@{6r`^*_@!8P)IuQz>q*}NF+6$uDhC=<){NRVR6JX72P zrD7#?mq4#vjQ%o`eIK2lBWkB(iH%74M*hsr&vL|u`Qvtb{nr|nQWyQF;Ac3D zb&GNvk6n$)2X6gDn>=Rs7nZHVNL!o<|KB$^YK`zIpV+`$Ym1`=jXqyq3vDYWi5W67O2&$Ib&z{a@7;`5X?0 zJSCBP`v@m*=Ic;n<~a?C4RQsi4Y6Yp-4mVgd=o@a|9-7i^;`qMeC`I1ys zM96CYBHtr^_XM_fZAXm+bK^3)3AgwZC*wjrn6D0)YBVlB|5N-P+2X)U)j<<;v72I3 zDY6sDz3En8RL&Z#2gkyxItbI3(|w{raai0N>&j$$C0>CJEWrUu;R2-E#p+&c%v$XM zg(-IJ+8!fREKc=UFK~j&#QX2Rzilkl;tpI5W1(8Sl}_Tf@xnukUHSl>GVD?9a7Xp2bM3i zxxUyKK{m`PjTKg=&rkbc(q4(t!pQ#q*08C+Z-1qAN^|vJ-mH z;se^`qux1ioTyO%gyTDm?)Fy%AdjM4Jr){GNX>t)NF{|1?4(>Wv27nyEa(AzgblMR zJu((ns*?ixc~$Ny#`gYN?M?nMk4@7md0CodYKLEqw(zaoFL!_E!k3e&Kc@U^tycfW zo&fY_ud}&8FtWyfx$u~m;l`a*)7N+ZOV6T@f1dVlt&5Dee~5P-~K z!dm>z;Mgvxj1LQzVrNt6|y$)_9yH znYSfEu|$Mk~~#N8f$G30WMBeBpZK5p{O(S@lCnK`0jB zwhGp{K}W1@mz^_( zOZ)Z8e)_KGtso@sEu;IS$LY4$4$f({i}|n#W^9TU!(f~+L(Toh>JR_9>g1LQKZXLA zHIOIL-1n{YfoA$xO}Ap7i>}yZ4_(M+NDN(UYg}w=VxBY8T#@vgp=^37W@6);w9%gQ ze%JdHS?0j@B+0~{Y@pcn6p4h0p>gpDpZ5t}0NPzZ5!4FHPh!Bh&F+&av2vQ9+&^9i zoD9YUk1NZI%=v^Y5b-3^rq_bBLI5K2<-SHRuq8sHr|9IJ_L58mNt@DwnyTkg>!w~<3&q4Px8s%r0)#TZN=Hvn_IyL6G@pt%qQmyQkx4xd=i)3_Sa2Zj3QK&M{s zxpKB1O6RgPX+We|?9%^Q8J(*@Zzfut$y>qbdC^d>u$PAgaTWgcqU&?Rg(t)h$J*TU zf*EpDr2K37;|c!8Rc3!@?lO|jEzst;>nGG*4qq$_C>GdPbE&Xxp$9$l{i>XGyMbL3 z*j6;#gf~4h!B_w{P&KTbtir%1$Poq>UQyYqRE zTt#-UURm#w1I5%>P-$;}bod-`1kCV9?Rsn(Fn>$*w4B)^< zwo$%QjoViQMmc}!bRbebMb<<2X`Qs+wME_uwxkQnV5{Bx-6B1(ucKC3FN$zKOFrTm z2;+4b%u?q*62@P2u*`lc?BM(!2^Q|(;k5=A39?;knVU0CeKB&bUKyn-XTk$2 z`TFTG{5W+gT}t#o$Z}W>h5b+<*Tm}vpsjJKlV*9Q)5kpxGAL-DQ0J!}7{l<)GGT@! zezHy1%dtb}DT-_fo;%0=swNttvto6?Q8M7t(jwvL)zqaPOi_yci#Fi@Fzw@{N%{?uaQr=850K%>Gql6=LW^LQ>29o zYmrAYNIiN&nZ!pTxdX88F8t2Ef2;XMqo~<^HKa!O&HrE*$$In7qTk(ikC)d%WgO^y zDUNyNNw5eei@7MhC@p6qJq+iPXZC68NtN?|pZt&HZ~b?>wwFAXHPJ)w^7Ni($HEdT z{7no`rm8c1@_q7M>P0!hsgQa=b-B;J5G{T>0z?F%AZM54h);=Vzw9or)BO|0h-S+g zXg)PC_Bk2}%m4FZ{aZ7K%+CkYB1z&a1GZulu}lVsusit^h3?neUd2NEUwiqXSUDkl&|q$5wKcy19%O$ z`d1061PAAo1V@C9IN>9G*qqMU05P22&)S=po%w%1`-KrP$G&yjo3uFa&TN^<&g?G5 z-l52CAkX#cq#M2Zq(*%JaA#W-%Y5VIeZU?yScieEE!n^hrmKdYlckU{cmS1i8S>-M z%eWSLrdokp_l{qlW&_mS!NW-V4cdZqs|@aslXZd$(m5K&j`L2Lv(odvZzmlmE?l5P zNu^}B79g0?;oaEjs0*JwSm%jV90N{Um_q1+d@MWU93hX8C#i0|v7R?bKGYPH3qi`0 zZ^FwWk9+8Tr-HBnb}b~++m-#U*kcLxASHpD#XHr<7p;dHl6;cO=&(fn)!30u_=ks{0vytTS%vOv{S?57mDM1{qxHcqa0DWMHM(Q)96K{l6j;+5h9V4TY6G#cnU zs+@Njn3p@<`{g%WFDyVZm_$0#1^w?@<|DoiB`wwnPqFzv%iq{ZL#K7IOR6f_wI9Nf z4;JkNdb5L`=RBLCtUO+YW1tr06>`jZ`5XP%c>oMGx+bVkZ&3~?alfWZ1M$KV9N*(CvSz<;9g1x3_`DEqmVNVo{o(r7mtAzS8ocEXo8{qD2Wf|i*CD~( z-;f$LQ=B0ESQz_y1xn>0za55m5D{^p&lWMckJQOCB)SZU{cLwmbJi=1XBb5P2Anb& ztiPy~8KRn~V`*HDi~^V7Cu1iRid_ykg$FbT9dJ%qusB@UDM}P$?fGbOBjX017sAE# z2Fu_XA*Maymd-i)3r2%g_b=De$N>lToIt^CSlp+QV#_H~LWS*?#w@rhf50Y-FY&V_ zn|S80YqYux+Iaan&nm&SDXpH#;%XoX!6^3uNzMFPEn2fpKFMNy^bx;CyY|g1U*7-b z1+DooWu!(MN45O$==c7(>`Nbg>(fPs+s5c=+Yd&ju;BWN6O5?CKMinMV1$wDy%X`I z;Fa03T`&RNaf+>^NI4bO!)pNA-nH|q*;M+h1V|PGH_g1NsBv#*x}0mZNB!^1`ZVXn z*yM6BpjRFTopn8-du3;#vA0&cG&D;y^aOU;rP0-Fz2s299Z0hno{yKW_OA*=ly0AD zcn^#h3_Vlqf_IG#c=p5H*cz+8-bub<-pXL1V<4Gq;5RGo&RaQqtG{k1+d3m&zH;`) zzyn@m`((I&Q}BK(Kh4n$J*HWC7$wz59 z;HLnF@o1nnor|)6csl_)LRN7fL86(klY=$`Zc>);3=nE58U_eCjcM z0frvlpMT|hu0{kEe)a3;$vSQba^Ni+fiw2-+qHYqg2WJMva(!Z_)Ba{g#-+yHH9 zJ~*vFj+-WG`|ke>`{xSvVal|~f0e3q<~^We#4PxDSqW1pSpjU*2gd2uS0?>VVb102 zu=Ro#*m;w{Wk&T8T{yL*;d{DWRIw6youj5S2~eP<&lNHR)$VEXQiV?Mog-QSq`jzK zeD#e?Vet(6!oiAZx8y|F6#$&<*#yrEzg}kEgZr!~h2uKj#6rXH9{TY-pvwf2A_F5x zhT<|rYatv5-M^W_NEhr!>85*vAPpoeVeUX(A3iad+CufMV_wBG0Cm{vf=?P*Mbha= z2n}wUneT-IM-OTs!Cf}O*%$z20cf_tnPotIC3--{7=6~?eB672ET2S@O`N@r6dOa4 zC`=$@^`s%jc^qmJFkjaqM+pEMeR>@JtyX@}97eRwt(HIN5dx!*`;}~&*}a-?M8%w~ z-#I~6O(H2KsMtiY8z~Y4F)G=;pm;gvdC_WP=SZFgDf+A34Y}QM#)g%^fNz^^CfwRl zUJ+m%TzS;hN?_!XdQqMT>$;(4yN|5)KtkTDJ~6;~{8$znVpp}P5oX1DV2m#^UJ+n? z{+AOxb5dT1^;9je1rM>qtaB^%&+>}%Zh;oK()mzb0_7xgTI71~UQLZh_pGEDkZYD_ z3dbv?!2!8LPQ(cKxBV$@$Q^idVg6gjCFR(=S8kC?Zox6f-L{h^la;ZCVyh`~mI_M} zM)G%qzkrh&_C*hejlMGjDw6)sGuNxFP!zqaK@YZ`{HZh5h<#tH- zdnDw7`<{@;PTQT&i5_|$kX3pjAx^r>F+gQa;GuvM$`kZ?k7(wxQzfs9&L%CSMARuN zc3C+o$8WpeW7!UmRg+S^_s?&oPfW%55$9ET6d)V|Bd4ESVY=>W{Ey7L`yJNqSm@U` zY-nH;oDY3X$i<7zQ_W17Z0oE6D7j2_Jvu(O7WZjZ@{jmIx71)8)oOko z6a^#w{Ss{O#*j*ZyDnUaZTs6)je$+jTh%6s(k|gV=a%Ne)>yFgK3vapGf3%YPlt+( z1}S3OwO^2^NkBO`Jj}g=ViPE`8SAtY#pnDV$gAD4(mzFV&c77&ba6_c`^>vBFk|jE zRg$X053fXy8&e$TY2f;Kj`DtP)8Md8ABm!MkuN*@1L_@Qg0w3{IO63<{9!X2zzG{x zAeGa2wz`=cQWxI+?Vrp;CC|p$95{(*A+Cc_*gENc-;xk0lWPr*VmEo+aIKSuk83Z# zaZ0T~%Ah-8D+E|Oj%B3$s%ptm^IDRZ3KbR#Um&tKIaQrVr-E~_ht8#Ad3x^zLHM|> zZo_8AI#3LIV4MyugW}sc|L^zaGY?B{3JgFM00`Jsnae$Pv|Q%b6|Gn?We>*(~+qNq|^PbDNq=oYmh!* zPBYCAqiBR83tc67q{4l{Ugs?8{Jd02e863`u32$r-WK0(QtfUJg^r=mZuFnNeNDRM zPyeNg$*=u$@*SyhzaqLL`ZbA}M1UoI*w-|PVxgrbfeJgLygRR(-pt&SU7veE_6QpN z_VVrooDqZzR|KGdP%pbi*rR#ubZusYZ;z%pa0AKG*!inz0Lf?|G6XMs{^SG^$2HqO z`Vqa+yjA8SO~o^k4mk#5-^|MxKd*OQl#)rB9^7!iD9` zr|i-x=iX?Lgj3D(oJIFQu{DQI^UmeGQEV(lqN%V`Bs(}o zec1OHv_XI_0Kf>!Te0&GKxRSs~&xamJ&_Cc*D7gu}5b1NU3aW&+ zLxCEZcU*LM+VI<^(t0(K_NzNT4IKOb{>M;nA*M z=h6Zl&^P9l!MEw^{53Z1O0nrzzzoJ?dHJaJXMdh;#Ez3#T|iQ~VaI{pK34Z=5`rJ8Hn6a|ZDy7Vn*wza@}3PY&4wFA(cxLohUXeo_*xJUpigR^zgtC0eGT9w%Lho z?8#Su^!DY+5Oz7ZE|ULrEI4uCwaEfbPA+NycEe@-WL|=Akzl3AIdL9c4dNB$UfZ3w zd&H1SOszIvHT*SFf<2O4A~)1G>>}tO*ERqn@uH|%i8{oqsJkzVmH zS9X(gp1a**18uz7e(vXPxM(FRtQg6*x3EF4XT0pCZOa4pE86DYfbaiB(hZDT<3GED zbMoRTw(ZneI~up0dBxS%ADWBSIBXTQ1%aAWbwX&O`1;)Su;Do77cE{dPxL)b7yo3* z-8W*smit!2TM=(9|CicDiN529lox};Bm!gY+>ZG^PT-)f?0)AL=26^N#Ivvv^Igfu zh+sAZ>}_kC z-w6sMx?a)*p_L@bCRbgNV5tzx_iwvnlG=8=TdP^2On_$j#h9V>WPlU#)l%M2T*TLD z8CmJTv45y;8y5SIr`Qb?SxO4I zF8-dZdCAyoA%!>$D_bZwo+2BtM+A2uyH$OpnMMK57J0gF3%hvmKeP>3K*Fe#&Hh_b@3Zpz77PypKDp1~kb1Iia= zif<^=>5E>?D!mf-$dH^I4GZnFxZsYhCm?&?!Lx6~@yrAFEQ!!)Tt3K~{3jCmk{Opw z6XUX-Vz=TBJPaoj%21&BOm_w3f|77PWjm!u{V_WoIT^B(iDDF@G2HWi21fY)@8AAi z*gL;^JM2&I{LgQH6(*rsCqhwC{~q)2<_5-Lw>cI}2DWrzZ4wfbqFfYwt5J`tMUJeC z;h}{KBIMiT)t;-gz-f!qP+eXLf97FR(HRGGAlUZu@o z?#}8`WT@|xY8q4))wNpW|0&^RFx!=lOoY#-IX%1#^?r!RcPUQL4?QpWo%h(qJ3~%N z@6sDJDfDq@LOv?3vJFF!Js`w!ff@03P7tCVPXF^!^Og@T+i3?5$y(@KDTRd2|2E5O zwU1@{8Duor?RLU1Nti83Sr{!ozo1HxroKM6lP(cF2&&aylaxwRKsaw1mCW9fZ~o$@ znYZN0WVv0VHe&+Th>)<&$>6Dxh2Jz z?^7hJTfn~T<$db^Sbj1Dm${C4+q|QhF2!kO!j$WCdl>`qa1wJ%&@r#TXL-OQpL3ok zpfaVGX_Xp|4eycYi;Qt8^kaEH?~xCZPL{!XS0yNCxGZmDLd)~q-;u6y01`Jd)U~Ct z*u#je@0IVkL^g7(TXEn%=Ufvc?xI*|aos_MRRm%S9qI+I7VDJ{1G9owYEZg2Q*3CN z>yhhr25p_yuFj(HPxxfRPxOH7E__zg; zo_3%85ogINNR0TTeNS?N#Kf+RJtw?}ATjWt2T76xL!!h468k7NnbXa<%IoXY{zTYlrKEx3~QL7sQB^{Z2_|$m&-HDXAt%*-Wu<6p6)#o;&o7c`K%^ zRz(L^0YgHQAotr1lBK|_PzbyV7@V~$5T5z&b3n#3&)W0E-M41{*xY5c;2CCmKMchv z!uDCZeJ<_l6t!*g?%e-%R(+P-haK}1oUFIw?sXd{Hhh1Q(O76(KhTk-4(zffntc6Z zDK?rS>kJ}pgGA@p`iyL$P$q(<2IKpsM?_f#R{L825rNGKtP}Y+7p*dv@p4$76bQqH zWo_0|Y$Qb@s4%P;=vM4=(LJR=8GkrfiHH%W%+4pdA>t?Ry2iik;t17&7Zzmj41d>X zOGNicoA!eCIDPtydE!hT!};UlG{~o+>UDVt_ZfB&+F zOv?nv>2|u6z7%{!Jy?ZU>rn=Id_&}azzJLPFseEj`EP>K#3n(Htc!jut5Mot3WhsM zjDWE-SU90%!sV^v`7e90!%!yO=TbT|Uf#j1aPN0Zr;&VRP~f1GMvn5cORtyj@Qa4KT5VB@_PjYAHBbOA&l`8F*Ej)XV&@ls*l8|A?64k)1!-Zl z6)3#Ei&wZ{z^O>xP4-OFPyfg>pUH(fwR9JxQ0kV%jbA-IOd&^fjCInjex4gTO80Ir zF*hSza{7+D8zUAXta+Mv`5k&0M7ps4qdG7O7$}lwHVM$$w1JLJs`_&pDJBqaJ+fMV zEO@PwZS(~~3OX^s?kOFPuo^dCW_RNUk>-tm4r@j%wDKkK4donI*sR+K>@>IC(Pn(g zLdp(uxG_u>-7+h67F@jPX^^yRq7Qo3couu=I^9$0JyTkg_54E zk&!0DDol*@pqb0nEpnJy-JZfDy-VG8l1x-*cV_pB8>)e@G|gbO8$HNJfV%!$0cUv`27Y}{A#WSMV8s!7&%Ft!b1q;FhG9{&? zbY^`ZZeVqJeh2+fq|wmW3*j?a9^DR9_Apg#ybJU$9&TL*$2PF#)k#z7PWJ()bQ%lQ z4Y4)Dl|90%j4n#GE%X4?8P!4)ah=q#b$;$b!gX?Lz}nLJc4T>P&y=3kpgu(Kot=$N97w_Hj-zgh;?@)2cVGs9FRY( zak15=;(7l}!!^#6@ZuqDatEhLa9lc(uYP}{;$=4zGxu(rZN*dv*s z7xaHO`g@$srT!wIovXFG$#OV0)UV6ria*73{w|0NL)oY_e?newd^3wCZJ z6;|ZGO10ASss!n@7GvE>I*s)vnEopf^~!5JV#Vhnar1w!21nzuK=O^p=KKJ!i1DyM za}dYN8x$i*qgD_!z$T)|!lQuo5P35!rPy-khY-0`FlG zgOnTC$+mq;vn=1IRvRuq2g>I?BwV=NuZOqZyGgL&%MFs8Ma{O>(lhy$O31t+)9aZ{rEJ-WX?#kiu>y?*Hf66e4x3MgU?YoAD=SWuQ zrzP{kr}gl3NP4oxx0`f1SG#91sp?u5OUp2W)bE-k?$cDx-8yRk!X42gsQ=j*GhFdI zgkyXA6P)lnQPJTMZmw+Nu-g_u#W0v`q}UjWL{VYMASJoZt=y~LWjEakURy2b>0+%KS5CK{W}~Ip2220!8~bW|>dB)hIH5 z(9W+D{#d5{Z=>b->-sgXlMV;A9IGY{pXok|y+@I5peB;G&cK@YXz|7OQoo{CrmC}~ zd+7TigIRh!9Zfod2VbVbOi!1saXI28ImdKcQP>rmt$C^}JS#f7^JfpF2UEaPZubY&ns3MIayKCq(`<*1zOPjZeC6aPq90?qMFsc+EH` zdokfg`bp+1M*m|`p`e|VJFx59VB%|3Q!HqpoT9=qg>l}x{R=KEfD-*Iy1=yrnt|JR z{k(@?$P}kfFPaxG-w5rUI%IRLnqNAjXxpN~cF8$7Eb>yjXA^Gf;VPT9i*`5zj>_dt? zpu*07aoE*)R3Bpz73m-cbtBZXeJKgkLJ3rAk%Ktuz4vgSKCJLU5;8y{cY^+a%vn9o@ z>3%J;75swN!udPl*RIv}xt8<)SC{sN;TgR$!oN>btGyc1M-2ax66R*miIAH?ZS!~2 zhBJ3(l{32}(T=-I|E2r(O+iKkCVQuBA~_@NyB#>-ciLnvETPyV6gfnNt>k0)B~!c_ z{G-(>UA#*%zfSsDunwasC6Id3t5IVT`4S3Yjp}o)66Aa3h?exZ9+ehMU(4vKoOeku zz`SyHzTXmDPBDxV<6I|NQmf2WlzVOU&t=w1^~y|Pb}$5RI-pRuRa)-RA|IRT@hA+H z)eGL}%@kAxTU@tzE=ptS(-#oik+rk`&I(oi-=k3J|Lww_V>m{D$HdIpJsrIX8~@t>^S@cU)qD>p%X%Z{IZH>>)4t$E3u87p*HMUx2d|dx|0_P%SxC zvC+3Ps0Rdace&*<%?cy@(rLZ=!uRz`BsxZYZA?Jmw_#9a8&l}mKG`GN=eN>-E2$SD zAxnltmk;HKao%{k&r5eLq?7(@p6gk-HJz?@ZWTmL+2^Oj3dM&WRqi{f&lf?K3T|Bm z5{0+jYxo5K{1P7CG%}n=#FF9Ucp?2czru)@I3aza@4J%k(ndh_e%JdHS;lQ6=eS=L z0CLHYX7BOKk~Yy95+sX>m**<#U8?9c)F^okt+h1IajUGj^aLtaAIyCFmpSH<=4X?l z4vZWN@#OlzG}U04XtcN+ifhuncFr6sCP)4uERPvXXS5Fza3cfJ_!M)E3I|rt=IC#< z?6myuwY`^rZ*)+^(jSG8E(i8_qfKJR;S^2VojkzOVVc4pziLlWGfB{4%+aE#KKQEquqhl~V;rdXfa>b8tP`rsN&-8WN?DaN{@w4ErzvOfy{p{I;9nK1c?|dU%de`N75H1= zJMmNtpo326wUtJ&}VGoof8qFV7L-eynsp z=S|j&EQ+lF_M-RUdi{Is%_Zg~8?OjtEcEyr8gn6;4jhNDO+T(!j3uKwG&#o=Xmav2 zm9ishtRNh(jTU!saGmDS-G3m^XRo=_^NN7Of;*G}O``3C$*L+@322O57L^3nD3MEd zm8wUs8|v^j@FL#Tlmy04hTIhVXLEyLI%C`sGPLrTj$`XrpZz)aYtxNR&$srx`##w| zQgYLQx7mkHVlP<~3(Zt%R9Lw1rg*^TgU>bR#J@8fW05GjHs=>0hb~^|kR-SskmPx$ z(T^Y>ejNN~Je#LEAc>pbuKr60+2;r5L=ZQBz)m7YqX)$hIPAmhVA*Mjy}e7RHNvG; zH9dhGabO?$vI#U!QS1qd9HqkAWmWERgB|1OiS~!a@ixys;nymNo{~4E*d-bKS>b{+q*dApd6E*?mQ}fDibrpAY=E365E|iN z)Ksi6PKJu#M$umhVV3nAeKc{knmwFU|9f>sJMX4Uze2-ziRv5&I%^JmdI z2t9ucdt6OC`P*Na%kI8Zff@_qx@B|^D0@dUL(MDcz|ak>WhE$LL@yJLQ+MXm8MAKw z4aVW=Q}#cyY(nNbZJcF}3@+ZR1D9x7zzoPt%xRG~D^k_Q^rigMg&;w~)VqYctf4T@ zXK%mNbiN%GGoe>CyUo2=hpm~h;IXcmRHN-Bk7bYergXHlyL{i10tgr8?o)$FG!RFd$D^=yx1KSn?RAxRM>7BX;9PYExx)Gbu5Y1 zRMCfnb-PsI!cL$qD+x@c%cQ#1E-U!uup>U>rdN)`CYLKU>QjU~12h|-;DnC}n~L6@ zI+if311~fd!m!BgkBntdbp>8YowPuT)nKW@_1FVpyGGH342-|`(T@G4JZ?( z(@-sp^zOatMh^_+;we2SBUMS;Tza8n#ty}OV>LiSkh3wqCVWjBb4+10DK+mLTuin) zut_N}F)0}o3(U*Auw1G~e$I27s!VpoXPb98Fjk$Nrn@cuOj18BO|3&F#MQ8^t?kh#*(GKPg)To~l;@vkr&x=au^pS&}m6ALB93lA6@IZy%QhtII1u52W;ZL#w z2#2hY6+f-V>-pE*e)i85>^FcaJJ_JtX$bZgFU$_^a;cN9_wH77yC3$&gg_}l&1lRm z+RZUGnWj-7#N>Hqq1rr4!}x#GWhwjHs+W#*q<7Dk#g+nnR(y8Phb ze_a?0Et&&I>@Couoh7SP75-VwR?r$rQDG>iDJVM_ZORd7n6)EdMx0)&Q8ePy_B**W zHKEPVXg}s8{ymnYjg&@q;Ofds6Pt37V!_Dlqr$+HKppP3&=X3Oh##!8#KO5IdMmqM zyjE49>UWFuz;3rHlBd=y+ayt1?3C0?QzcdIg&w+QMQd<`{1$5e0&ALXi}Yqt8o&ty zG_li)gV)ZttxIAxJf5{M&z`X%S}YrxYaf35-I+$^A?SkL?pN-&ivNh$0aYkibq$k#(6iqyhL3xd7Wn`)=huUG3PBb;aRhjO z&#E7SfM}~UNu`UKx)|gnyzUH+XM59PBXrmh7lxrMqho~N?L_t48*9fxTjjt;$Rcf( zAw7uO;g5j)Fb|j$Y>y5rrqvn%Suu>wZ}|q9x|5?)n&6e(7w?ge9T#gJs)UN>ZDlQtwZ+BK2o@#Rel_dP$#6B+ju23G5`;@-XeN3 zh4y6@w+1NBE;z3)asj*GKYS z9eDS$+N2-k0mb%GKnoD|$Oox|y6H>M;eo7Edk0B{cJZz-J+kfKjGk2M)m6NkP~TP( z80lF?uO(OP;|Mh9l`Qh>N4;it7gGKAG~;ld*!djtBU=e;lyjnCk{-I8U_ zD<`G#x)i(V1F}lb$MQ&imEdql8gIMbs!0jHM;1l08B8qyvgnL5i-&Sb=R9{ax9JS# zh^%nRF$hW>g#yeBrtP(QQ6Y^_)r;25J4RN7c1v)aE7KoAwcRzzc~wU6>PZhguS`i0 zToNr6md@-B`AoEC(r^fKL6}7oWUC9qfNR;w`E$tHM^Yp9M0Z5LCNU1|0t035umFD& z#qOX;0%W79Jw6ZRdwldt>>I)^`%K|&cMPq!@sJF0c^Y;eHM)5(H z;nOzkld|dIsNi7J-(cjyEW=?0pLYWFrtZeKj1YPAqn;mIkt3JL5x-i`JDM^YNyJf{4N1hIk{?fEr;x66W_{r8pibbMffwikbwd|gI+WQ} z31GoP@xE3;JG70}2i^fMY31yb(!o^_>+Wil=RBbVuSQuMvQ^s+Bgblb0KEOK_yjmh zFfs%aI|n-_bthp1kJm6_*eb<42Ky?{DDTk~q|E=tU>+3|$8L#6#^12BryfN7$h-&t z*@8<4_K_`g+NY>7N(XIYdbJ@UjEO>YfydeB!H?S7xgRzy<8o-(47K?Nx|y@yIc!Ot z1%^9Rg!^l=yjGhfy-QzGAK>k0pudBz zl00^Lzt(A6JTJJPmo^411y=?63@b=Ezl_f0CDX;?J9HVniw8;r zHkmYQ!NS2d*yi`*Qfy1rxa||0i#|FmfFOf3%&m!|*jS20Q(=2$_n{6Gt$wQdw(lTy zI2Or(e=~i>xD$t44GWu0w)KNpwjm?AGPfftzxY+7D|7#AZ@oh*UKv;BnhA)`Q|vj4 zoCbXhj~IR{txKdYl5Rzs>bywTI-`$Nl6=2>=r2S6B9U%X5B59PC=+~B-~~jDG2if5 zR_cEU+>%!M;morQr3QGc3f!Zo z=8F^QJVhS#!+r$x*7bo2QjoyDrZ9LtdT5BmmdJfT$;ZiLSq(uBrpeN+)&p?Dgr9r+ zzi!?H?Xb4SLR)j?w~;?LMwlatbKcKXgsk&!RvqB?4X#vKnzDsCyb3{upw0H3?Up}i zZqsdf99v)EWO1m&ZfEwq>?x*d=oUnafXT;54=i4L?2$?@6@sC-JHyaKR2^7612Y7A z<=H76LDAwU)#2&4J+~`+cuSTnDdD3w6YAa%)`VlJ6jp_E`tn;di`)=n>8o#hfl*F~ znb7yMUj~|2!*KC^95}^fv33TzHIh_VUTKiZXcAz8vYiH%C0KR3obS+`?jQ&!$PuN& z>yOkC^0qm78Z5Nwr0rDg`vh>KN7a)d>~gU3hetYshaz{*i*}KH4jj1y0gPcTZW+Zw zHv2FYW~kW4ihmulIJdBCgu5jBc&#)R@EhndmjGX#d#hjtlOR1s^y(u%Wpo!wnt3Ro z*k`Y7Z|Lce>?v6MfL!34RQZ0oidpp{6pL#l5J>?7aE#Yv3R~$WmnB$v)jAcunlWQ$ z^q{b8b_TDE5h|Qa%|zXY)n6Y=(XIoBWh^js#LvZ6SG^JoA+Q8IPZK{kQ+UVa6v+gy{6J z-6^3PB)Uv-s}xU z3hBz)gYaGIq#u(AcNhy5dzu6*K&P+-sLLz)Ee!1DO6V@>28apd%FcmWWiQjgtQM9M z{8jcyUJb>`317bcr=s5)*S9qh(F;hY1FvuEOfsO4D7K#>_n|7-HEq_h;98mP3OPlULFt+$%{yrW8hN_ zI5p8dK|oXmXx!=#x*Ez8K(q9_n2`rVNq@hkKHJjn3K$Ig1?=P*()sb%X)G*jW$ zCpG+bsp@nlnyCeTPCP#-&YA?@kW&kDL_Hd8n&|=pYCtdc4WD&>MeygmX=N8{#GmWu zy`PfJFIiST-^Axir&y@cNT$N}E1>%+ppuuO-VEXB80UzP4Ft)^H@W7~n}AaM4zxej zYSWyzKogYV`tpF2>LiE+#m!$eF9%q~ZN70caeBt!TA7Sv>lcXLDE% zT)$xaPF%h;qV$V2{p>G9zDY6gv`EpFqubg;4pdA zt}iTuBkn@4r+LMwg;Gm&NYS5w`dst_41Lez_~AbJgsl3MpWHou2bMqO5rX}WyEzHy zct87}s}rsjzznF4c&UvQr=JyKsN*qkHjJ z%+r?7t|X3&Y+@mmiFGl@HTNXQHU^ASE^YD({`tT=0oK*7aQ_^i`-HWV|J*_BaM7}L z>-zRLv&@qS7TQ6U35#7msF!yy{2dc85f` zSMmoLyOUHPtt7uE>jdViC>6$v4|-zS0ahanK)2FSY!B~%moCEnVjz;sbPG5psjYyd zO`wlo2P@u-%PnthzHrstQ{|F)a^Pjjg2Rn%DVf3=r4G3oN_eP;-6OY$wR_PtSOtj} z4PuAu)ZLoEiAFQBVs*e#2tQd~rA9E24h#t>9UqpFT2HZ&6p5h1uxcDvz-IZN4Zw=c ze*_?`s@b%|guigehKGrZ%{^2LPG7vrV2{D^#~9(GpnE|WZPy{)m+h_U3x0E}zF~tt z#tR0r@#R$IZy^I898G!*lew!TK{Cy~!hzG0mddxIkFRM@8rQEBtB?2vPl02Epz zkl|NjGGwb?@zkYLp{zY#UJAUE(SdsDAqZLzIF$*G)3{?y1>LCwUJ$%)5ahfJMHWGpZB`H_DTC{r^~c2 zown1yB7)+mxPh{0SY(k!Rz*=kP;nhKIH*xv5k$vf20<82M^u9Z3QO z)ObOSG?~*6p|HDNPf7SYjyGG5tuyyo0+tot+}8{IeClb1b>`Ym$~PsjMWItCLNC~J z5oN+2sdh8}1MfT-`c{w`)V`<8n1YIdYk9|IyF_)b0N)WRSI5i`uzct?KaCC{HgDzB zoi7POx|YB7$vm?~%98)2nCyIMRzUAt7}h3=IY5#9(D9Rp8h6^|5$M4bz>-F>bQV^f z3`5l8yXMeczaS@c7&)uZ>7M)x{6Y$!+liX?u%`29)4E9s86 zv}B1Jr=hbfkg<+pk}0y1iZvCcHAqa{DyY)G;dgItv0yx`oYix8aFngCa{WhEMsMQb zinq4=nlbU+#x3W`1~>|th6jJL8|l77KYXf*YY0H zJH(y*OS7uxVc!s)GJVF-93Z0XpLf`^Ly|?izJ;TMD0%wjcQH`dq2xb3fAMRP86`RV zjn_%C8>0loghnifbrh3Hku)kcBls=|7hV(AF>U;EDB(W{CHw{SGID~x{i;Ee1r=!R z^0JtEdUx1St&T0x;bLT%Z_f%KlxN$lN7j0FZd*8FN1?8PU*&gzxhzZ; zRrz((eXsV+SRq}@%cTdn7Z+~lj@}&@-FP`{I7WB-QD4Ol7qxSawtm%YI6@L<{XcSe z5^1-1KhIGNIDvYoROgm_wsPUx&~2d^pB=dk9mVL!a|=GZ_Ok^lJa$xxe-F?{mA$fU z8r&`hxrB0VgE81u6lBm0aCeG(==CH$7{kFykWh9alY%Q5MGW4I|# zoh9m?dI%Wa#^n)~<|8#2gF7ZFcsuky-z92M)G4$Zee@u_u*DD8PQ?)G!SxTGSShRY zgQkso&o;81#f2!))!SND?- zT5oyMUMe_t0SC;sZpcHCZrsnDQqET1f}yDRQ5R%?#`TpA{FqSP%&EKqe+eZ`37%ia>43 z6jAN`Vo@hu&+h_xvLO##&SpnIsVWLkb-=Yqls5gq9dY)piH559k-m@2ft%s=+2;aQ2N9;)0OBA!N38@3{2)M5u6EFM*4}i z-ZRVR3_U2*UWwl7(-cw=0qo9~q$qLHstm1<8GS#$Ocs%KZj8Qii-{_SVt|I3In2~; zQ5FE%)=08zg^|w+8}e8hmMyCDI;za(t)U);=#-mhRH%x?Ly<_-7)xr|VUy;#=^Q!d zIDGq2pUBE|PB_{$^Y!nTF(O|X{55j=rEz__ED&{-VlGkS0u_5*4y!CMT#Kz6=&%r^@Rht3>^h0kBks zAp5#xN+oyDW21keYO^v+ti`@5sq!0h49d%1TcQU&+LR~59;j~7gC4a(y-`Us)BG2= zdEpOSoq*V(+ykv!RixgsqBhPq_YcM)Ij>P6@x1SBw!wbLv(-1Sls=6xU zN#RR_uqF$H?V=dyflxVYv!I!ZcQ&lVu@|;BPEZ;(z)h87CuHprQ4d6fvchyqR0)PQ zkj5DlsO1pr#Iuv(csuCh>D1+1Eb+(2KFJ*HXq)5R3pB%&K<73HbTV?D4F?uS`~S9j zwso)crwiQN*ZTv}qLHikQi>^}NC6dl`*l5t+#57U6<24)2R16Od|s!#I;)jA&)X?p z8kXpvuC5eS&uW()2D<55QlY|P`ZS4N3VqkV3SHKvX_6!KQP_W0^BkkV;%Qr*I#y6v z-Ogzr?N4uX<3|f7n+?sa|GNBlvSJb`v>1VOih;QD1}YX=Y4|>AxB66aZusS?fp(rn zzo$gz@TYQOl}aZ23A$9}UqB_26L_j{L3S>owjW5N>B&+*7eaMUI_v z$KHiY1ZSlQL5Y)wx0X$M>kkH~@4XLdL79P%f2a#=32!5V9&5gQ;d^Ud8}z{ahd{X( z^{g(1^pJI+B&Ac<|4ZhA1ZF_K?3H8=9(68Wa1fn@8CH03j$!525 zcH@;%y@g+1K`~H|RZPWp1)t|_5}D%0nF|J{9Z=!=7nNXi$_806wDM}_T#$DO(@Bm9 z(6(?DZ-B0)PI7B0{rnS<8DE6y@FS!$VjG?1cM??a^x+TZEOG(X#sIJYBnCX2|sYNV<&Vx^c+2!2$$T6az8fQfN{ead_H!RgUPbM&2sk#-K{k0G%!V!|>h~ z&7Cm4?53vZI~)Hc<=ba9=*6Um==md0Zr~Rwd!uqh4UlLMP6;1s z#VwI27i-$#45Q;Lc#SYQPQls+LGGK@Wo|Et_=laMxMWeAa+y)&rAC$uC61;)&KP5d zO}qI-hcLnRrcO9N?ayc0=;4{?JAo%iyc^FvP}MeqUbUWL)=*>>=y}Yq4?iQ^%GC?2 zppq?n`eE+^C}11UFyNDcaA=sFN2Po6z0ccVHM2}?n%?+%YeV6-YaKhUKF$++Rh3*V z%0*~5gg1%*x3o4L!^HT5f*mG4T)6%tYyQ96V!}>g!fTbi z@CZ9ZY@^;?7Hpn6j(l!?Ey-i&f4XmF0|Bvx5fjZ0ih%;_GAj1)w059fVvx0+(*qf9s0CEqL^ z@<7E@Q~Qz*rvRKpuy6dRju{;7ov3r8C$)gscv1!EE?YG~%8Q5}p5N?Aivv=p2PegeS+&AQr z9f3t{h4jafNnU>)^vH}jJ3n5aQ?KH`%Yj<6>Zo@4fX^!50qNQK8T<-Wu`hP>JIw`E z#1w9`udyqJcfL_=>779jUEuPm`_xTQ8T@)WNpKmM!$y&i3Kj6wf%|uOS~BnEJS}z< z=#;JXyhMrxpj#c7uE3@}$QHEnp`8XcyC`(@fF)Y{{xtJikQr>hD?hY`Y-I;FH(nz* zT7a#ZVk#)Ior*=G69mwJooS-nX^RBez{JGqt{EC$*9DjwIBK&aurN|D+czKUCZPcr z-R1*4PVD{o2xt!2-#VFdQrRHcAH60p$;^4O8g{dHH*#^98eWRn8e$#UBvo8AjNe^a=1Iy+9GuWXhK|HPz_DdKPEOzz_T9c zsE>^bA*22_$1hC%>QDWwm(4E;LhLNEArAzEgu>ud>Tbh*d5!O1qkfiZ9p z6b{_UxuZBUJ(!mCqV0D>}^%DrlI zChEfn=)8rE>RM4pKqnO37X_g!;(F9Lh!ShR!nu&JakpaWoKoxH3LBKT@v_NICrW2!G}}E;-q@C17H>hhQr+F(w(p;cNDS@ZN`=_sFE{ETZaxNI0hXvEtxw+A2o zG{r;6bM#vFN-_)k9jHFY+k%)ZS)%M zhY%Mojk-B&U_tVN0q^~ihUgXY59K(KZH-u^I6&?N;5p|d)$|%}jksUjqB%r%2I1$u zzBlLGCTBID$ky?YX;;Th5jg{UPDI?NGno?%k>|3{#9FVM-8PS8XDy5i7*pqRZJpPE zPq!zk-9r%~GhUS1PshR0VgB@r@!4W!SSBvK`ng)`juS7*n#fM$43p?85|S2U1rsKM z7Gp>0?f#Vz%pYeQISL}DyCToQ;!P^>)>vTS#sL*Out3>%An0RSCn=xUR42zK$jl*{IkN@%gADZ2e;GFBw%i4`OOMMoO#Vv~Iphz1Piz&FNRsQe3 zQpe!Wz9_x-fOjQ#nfHN^bNZYbwhK=GoHGjnq-E+GoCl;QGd%$F#PfPR-w+KDdXz7lj86nv{JI_#R@&t^H8s&WqU~ZNMh|+NRowK(t!?y`$U5(Cuk)%x)u6{_ zPvk@6@s`L+`FUPDEMjWKhe8KEii1jL7AXcj@)nkJS1f>~&6FI`fOmZ0pvU7`ho|l0 zc6%j-bZVMHP+y~6j;r3sybSNBe}|3Sq>vQvI*7>*-?^EN6RhG5dL)J9Ej0aZb>2Cm z^C~=}UDl;Iufn~D$yNU8kaCOYiaQMP6U>YlYj#8Iwo1OY{#f@5Yu!XPPLUh0w(O`Z z-ksbGZ3*Cy4ATBX{^tQ+yveKyaHU!VJX|hH@$ni}lzXXnd@pI1n zn*ZfDrkE{P>+JJykrX$cB&#eY$vle5p@@!(-HBSt5V1*>pYpsH)kEK)GZ#Ptp`3eJ zh~iNVvIGeCR0ZGV4SAr5bDSVvfEk7dlKOC`NAsR~+@l`&r|#t39jvV1gm3-jk97;p zFll)sA)nMvBIhke;~2#prpQ4m*0fGRt%VzbP}wyb6;7H$^3|nLjq*l0#?P?xDmMJq zsj=(GbvcsTdZTLR-{hX9SNZnvl7Lilp6r67T_9L^UERm&kTgpZKujJILJ z%mYaqXnNt57_x2i**@hU7ki2$-)<~#?x=_C$t?4XjWLWOt08mW{NLoemt`?#EPbQt z`xRuL8=IVq79f9*Vqi`>OvOI@(p7P0ps6@kKfl)Jmh>pS4r) zZ&bT%y`aXc$hV49vGBeuPSD4}$>*LVOMD(ee3ikQRUr35e>4{-r8a&kug#}{{$O?n z(`LMO8;wNLny?*$Q?Kp$!R?s0Hc#p$e>g+0i-x8R3mz-ZtGh>=B6TUwjjzDyz|IQi zREc-$RqI5#+eQlQq`onvV+dlb>fMYf> zzdP1fF>5KZ8kNUQj!uj67Ua25Wqg0gZgGA1D#ci4!qE+? zV;FL7hRTuOw6QY#l+=uh<6nRM3R(ZspaMv}BT$h`F<@vmgA$f}iJ$;XN!y&YM7s%O zQHw>`1f`2ErI228O8rO-GK3gDL@^Nboj@wejib&{)9Hob7kG5w1DRF$5Q7^MojeH8!{7zsD zGHs7vg;%klCwfz~7QK@@3(JJ-qEK-YDO=b5J`gtYwu+0Pyk)C6nd8(f#Ta1oG?dt* z!|oMJZ>Ap8m~Bf(#MBJ3mz`~KW7u4jDbt!l`XB+Ggn? zGT@Ebs$O}uU$x&6`a-0(UX{xr)n<4{A>G0;Ry^ebqVm)?{5DR%G6T#G{=hrpOa|aH z!14}$w1Bwt3g6QTZEj%sf^Je2R7jTyhHr({F_GGMGQjO8Cw)fYFF6%WPvh&!9=ouy zbKuRkuLN2vi?fO8xv~FZM~571E>S2oRbHYhiv-itIO9X_#F-l)Les)&20m7YWM9ZK zs4_6B=;8t^PS8U)NpdAa9wtQ@Q}Ifxq*QS)O8e2AYM^*w^9)SFYcaGA8Z+Tp0o&;0 z@YSAaAcnTpXB*VAtdN4p?=(F$MCg>|N(MdFhBiZl;xMUGX9Yo>%Qn>PO_gt>YvyZn zrnm8%rFrVEIUN$L(8I99;c2zAi)JN@(BXno83BsFR!a_qXt5ltp1+odA9Tv<=u5L2 zAaPOwE_a5;1T*Q_^F)h|5C(=jY4Cfn$?;VUxF(- zjEmu6y45JWN;^+v6wm#M18H-^&iAn0&FVdNU(b@)_Uq4m!wjF#9qj#ga{Q(7T5egq z1s5p>Qg1C(Y_B|-(?kF5hP=VpiXFw|t0OyM?S%s9TG(F)``7ZeM`}?}pr2&Xk3x2P z=Y>{yX|bD6KS>EZ6OgDl=YK~5XJBRT7Qb^5WkUFH296HMGx+UttkcsD$oI`XLCypW zd0-I9L`uXGe>ih3FN;os)|XIDp$i<4pEOd?ut#oT=mAx;)Yx(i^b4R3S@1pY=f)zP z{*Ksuah~4!ym#i}^#p-+F$9|hnH!TV?C6olhduT?MtZ~#6kC0EK~5{3SqTNbm%|Mj zba#&hT@V^HZTwT>5Bv-0M$RGiWq&*ht;YR&py63LcdJiXaeh(Dw3WMf`-@T|K z0I;mED?vS=Z2s4`Q6J2yQayLl zaB#pJfy2`?!!!0c_`T0Njg|jJy)v=yxb-^9ZPVR$Rz{|x7?YyaR&fQl17?PYAdEFM z2W_ukDl*?a8@c&qh$p+xB1FyMWcpxNAa%ava(RJy@(Bhyv zuWZJ&mBG2eIj!jPJgRJ<2e?^5M`+xCROw2LjRIMp8XhO1$kuH{`y&2i-P@LpU+czE z89Oa{-vf?)tK=T&>ur|6LMTsNACAHM3#8vW)Bn2PkOwM1v-@u>z!?)TK81+wEn6e* z;JzzoUvCwdjmTTolmARsO#%u3k#FB-ib+RP5u32!kRlI)FfE#fZ(adSu(O(UWr6yp zH)=1en$oCPPde4Z3Jha;R1W^;rqkFyS5`M4!^x~nx%*n;a|2%C^UUzblK-Tb>~v!n z_I(T7Hc<=^A@+j{8=OUJ8>c;1>{YL3vUxpHUEu17?1;hF2jty+(1)lGKcLE)-X8@) z4m@%|bzOKVUl8`$O+6x;4By z>NIJRz#UO-qCOndC~}z_vOVhUkz)uhvIV$p3t@lnXfe(T;1hBK+n>@EXEXn}Z#a_J zDGJAI(;eTfTrIAaOih>y>3G5ahQ_di5E?BKMt$8M9t``I>|HF7F3g_00-N)5zZ{Ij?Nh+LyXcx5d#t-H>`*bqK!5E#Dzp zo23tACF0W|zyWVjo)JQqVFE(%btLvI*P)^HF0B4)=n^}O0+Nd%zQn+-+KwCG%8tC<%eozsFa6Vde_7? zio(zjee3xiyLPChJgibdU?*@7H3eoEmc5- zXY-J8tyAK{xjL#--7jt+Ju=sTh%I7lxH(Uawlj7_tUvJPi{4%dOU;nQzZgr=`U!S7 zam^q;=Ycyi#?QHhOAfR1M{?z#5h=|!M04V&`$!Hu8{)>R-~ARgq>^IFC{jYjCX-u{ zcX?%k%U;Jkf!l^fX>ndjl%Agex)G4j!4=VI1(tMZuZgj;v4f8dN_5K0;*Hb#f>+Pf zBDcW66ktIH-mF+u%3Tq?C)nvEZbuMi!*bZz~KJ|gA@_ou3)5h8UIF1YFj{EcO_Hj(g z9P1)RJJo*;{Ik>J#D!2LbxT|gx^r2;bF3zw?(x}bq%Vr+L;RI?cMEH4-UYRh;x#y>SO#KxD(!QuR2}b#909fccl?0 zNjp;jbz5aq@_f%sOXeVnWNeK^93KW7Z{+iLJ@0ATJFa+=`gcRjINKF7q8z}rUF^B?z_so_ zwvS~s?8c!JJF8(+q@*F_vhV;8D?*Xn*5eKJW0(bOm28aIsnRxbGNc)xhBby z03KGxjotl6d-n80-8Ul4_Cod-?G{o#iHHgB&7b`|_RXLEYSACXG~+=i_qBKAq=i-3 zPcaa$t%dj*h!5VJ_o4VOgjR3PYvZ4t@-cX)X^L}X3FoY|1*GE7ku&sJ=~)eQIf3)q z_=Vio@Q?gzrySz8N3P{%^X_p=K&tTOJcGu32k99}16>_G1nXzKqC--|$(*tdgwbjP zQE=WlW`{jatYGu>$I(K=;yU}2NP#%wJ?nUi+XnOPL`bTku>JO{wL#h=F>#<`*a7wG z+kH=S^wI}01Kx29_Nq>aQNj?9T?xumuJf;pEKx14^6Ta9n09Z*Cea1IE_x{s#ea}r zyC(2X*b0{<2AzTqi*a?%2m9!-VN53ERB!qs-)wVUJHP25dDo5egxf6aO9sVkqDTrA zi`{OpTCt6fi`h;Zr4(8?HHrbBwY+@w7C+Py!->c_R>7fBVh0X~Zt0Yh*ny+)4xJ`5 zgClF|o4+Hewk9Mu4xm(904ASeKu$i3ibdxGIY*a;P{D_c`-hOW`2>nIj9M9}236

    #>DmGAOc%iruE^B+E2K z%Hf|antI_U+rHJX;f_!zzW>d? zTbEgY2EarFP<}@nZpur>=G74pzXJctZI3 zn=g9RpD6%x>E8{XtqH62+W}gGweykOn)SWMU&{T(VS$ig zZ^MFlV<-h;uYJ!9;$X(nW!gOdj-VIJkk+_f^@J3E#>g4f3&Xb`6)?q3| z$gh6;d+mS!^fy2I#%1%h{|H@eO^&w1Dh6Fxqg=-gc|XDK zKU<+uX3q3+f0FQ69D11J^7t&jgw-4AzU#tweZTYLP_q~M*LT|Yk!|eKziu3Yg%%Ve z(!aYYW+z3esMw{P>*Nmau`fqD!eW` zcd~Nm4R~Typ1om-;jw)RIJYExvKczJ{_FDJ$qIJRapM#YkX=VWC!J!z@!mkiLc0=P z4)=6OE)YnnqfipLC;Ait4s5>PR-a1F4ZmU$wwNAAGszAcV;gTaSN|maSoJ@bS?Pzy zL+FKkLaK$y;gx>5l9a$UKGthw2HyY5!EaCf?gV}Yy1M=+Yho7&R#bS)uLFd_K#hLp3 z+iT{&XDyQCw(DLyvPUI?%mvzd8iHk@Rcz7@?R23ksQpoQXprq+?y@87yr)gNf!o$C zJlqx%cA9du%P)jgy$anOAzAl<_gP>suZt=QS^;X-<2YTI$I~HHP|kd*9Rm_ORD{PH zzGb9t`BVKx7#$`yq?4LK}QAl@t)^i5D%jd6?XMw`>x~Qx$P`t+z+Gq@y>~I90xm=#&@=M*jZT@#ieSSy(A2 zOo?ASIdzAR8A3n5Ocs%K>>%XEi{^3*2<1==ILDb(Y`=GNc;d|T5YW(%6BN=}BJJ7v zWsz+(67YvS-cx2wIj+_PE}x1T!q_xO7x-B5f%kYYAV-N2rx>1R-6ZSt(PykRPo7R7 zyRk9ZV}XQHiYcN<0Tp|WEDK-e-9dK*YtekPN-(4busH0A3+GCX7An?#uml6_L_Ty{ zFp-$@Y0Mqs1*T3MXk_4o<$^nMSM((tB`B=yg!6r8`!m(;rN7y1&2)T83XARVB9}rt z?kge;&BhC^h;J*h!%CuWlWkCBZ2yQsa~{IEr$yCtv+{~yqyG+Z_VgVgk9npm#y;WL zdZC?wC0n;0^S(y`-z785lThD}q{~RI8@s*@7RIECVql&rrD9tpNLe%K#$p8^n1n)l zAFPq?d*%C;FK7jN!GK|9hp$z5xeEKGDgq1Rnk3o6H_cI46PS7 za-eG&Hmi1OmpTViSk?x__UADV{m2{GHz_vwC=<8eN}Fq4vEjDMZ96p>{d2C**%00& zu8t}e|qCw+<(;^ zB4(RX<&k}oENAD~x^W^A`htzf25q946pF0FwNYGD8uz;Vq2%K^I^|Iavi#kk*(56G z*1fVzbU}E8F8S_wLC84%Y;fcfz}UIz!>|9c%es=3jep|Cj;@`8SX9Wttj;CUE?>?A z`n2pWZ*R~en0Hpb)(^{`Y+j2Jo6F@!7paSbbV|K!X|O3%uz%hyX%Ximsh)d5i2Ykk zMPn_T_k`P^^I@w5z1}mztz0x**s!pTkG%(>PdKd1G49hUxf1k|qLAJ^W%Vq)e)Irj zgC^T^)id|n!OHV@_XJnoebL!(OL<+wL4G^^R;9d^X&`?X{tP5*c>%oynhqj*+Ub^g zTU57S-62~`&T~dTlFxsp!)J_Y%!e;_;ZIgG5%uG!6(!cZ1h;im>@W@Xst>64hm=m! z<^d3HKp)&fx|jQbH|PczLT`k<>ib^jRRx~o5hNKNR+iWpzRcx6Sz%+spWZ2X!&>5x z&0OTh;Ws;ifam->L1wFtS1i~ix$C8UcXAdzq^b)*;iMbBz1-CdwCmcdEE9e#Tq(cm zUpODTZ{Y4B5B!6X-|DFPa4ap%qU)gPmy0TXSi;NZ+~NwX?JALf$9mQCk^sie8mR+X zC3nw-SX{ASUx*eLLYNim_PjZ7og5mwGb{WT_sUjJIpYn=G4S+|;+Z#O+atjXz*PeJ{teLCJc~{UF(!TU)W$pX z(4f!BHRaz=6bmq9Ql=^kwY5bE6`ghi%elN*6FtqhcI7x(dElehE;{-sWIaVl%`bq9Yq)Yw=I5m zRnXW?T`bxciXB$yWFCrw?F-#8OwKC1xm|P<@oU$U9G96*8^z? zV`XSL3>u7;VUDjke6P2&l#iYb$o4bn>f!h>|nC4XA3mK)s|FXm1cYp`qOlbT!7N2{@zGhY!cSbG9Gs|}|B4Q`Ux7Ex+lQ|h+CRy)P6D40|ol@j=Y>?1$z9p7RB@(Nt)aM*kL zj*YUtt6jIBl|`9ICitJSc0}wr{}@8TT-gA(65g5t?)FF*x_QrdJ`OziXI$;NL#!UW z`�(O#eFTq}>a}XtdM3dYB2F!O0=(d24ytjH?=mjnIy(KCqdCTt*0_V$#+WPRGB; z1K%vzJ_YX}pjD3oqnDin&dZn^bI#q6K<2 zA-(ED_|>_q{Tp+()}YzC=(Gv6|WE z`(Tb1$pESHy`G>QHJE;)SL5kO6&i=FQ`uA^=oKnI6@90uF+E9}oszxw(quJtO%&SK(9^jgo9z?8s3 zx<}gWe_Hx+M+kaleO1jgl){l{#Jb_`wRiPs-vfBBo(-~P8z`FmJRXc@UHN;WUU)VU7?NA2u?sY#lS*6gNjXomVAZuBkxqNLP;{m*rEVa zbNkdiq*hsL1co8TJu9MGerKXJ@-eF z#BR0YzWHmY^BW1MG>QQ+y!BLUN4gA;DCPl0`l#3{zkI*5 zQ@TBq1SYkB0|8o0vqG#YLdV&uY6$6^jp^5BX(A81$8L`tlog9kkeumwx>1)T)vHd~ z&4+jgYT%e|g^|tLDa{`J3ImSwgO?4aQivtsk6te`>NIX!vmD`v|S^qaax zByAF@u~;)~qZp{T&81>nlwEW?G^8=6C{pEEx!<7PKM&sLWFYw*pHeiV6^bX5IoT0L zW+BEl3nC7LtY$zM&FDLLXN$TbE4`FRdIRD4S6}FHyS~w_fp9ED82Lujcsf)g3Dj4r?bVbLV}GkV+Rf!H*nD| zR^V{os(jxMJY%evL~c99*jeNpB6}hakxyiY$c`C@NKZ7_tpOkO8X#3$EE@3XmEDr= zi5MTsfC+8vKsz3k0rT(0W<@jMmpO0N%r(z3KYxGTc2fC*wzpt)ZD7iT$EOs46r{FT{Xmu+jz@1vKTvOxjEPyz z5?3~+%4h5vgQ2u<42Q~FH@@f^Nmqk9B3Cn&ejo~(H{+=6hTj$OpzJF6OgZX9yaV&I zL|Js+jA9XbPvBG~@~#G%q(s~0bpeetK!fcjcZmSbSN9VXmnaq$3qYa`(wWXTNlpXE z7{FlbQu_e0vF@3;_w`?NSx-u=ESKA6JMG2kkn6S>XDOX>eQ;d>l9Bd@EYoZ!i+4#5 z1(howC~botid2=9|0e3QhwwtIh+p;fHfp$h+tZq~ALW_IA@ZHM+yd+Wo1Jk9aA}{s+n;AvAJA%jLEISmrZ)l11TA-+d zV%jKjm5NQ|C4x5F8Q~MXa+BYiDnG$ogp#lmOy&X|SBsi!Sh8*+UuV&U-r2mH+^ga? z{ubX7)nj?QPdy8|L={NVT0_+D^I`(h=HiK4)lH2d2tRwFQf7#dqX! zf-Ye)ubzv+QhUypi*WW7@a*rf4N0-T=rdEyIBT7K{whT{WR=ys=Oa(>m=l20&@4i|!^xL8Yo?bF~IY67|Y5A+m$La53QH;>maP z18;l4oVAzLY&&84o{u74#wwz|S8M0vlDi^w8+RL68SO#E?x{K@?s|XP#p(J$ZJl?S zv4hL@nMUe%GH+aHGowR|v#??RKK4=5|ErG2zc(9_q!r2YNf)~b%Z<4ZD=m`bk0}Pq ziaw!Y`(}3pJc1NxCP_ z@iX?N(<%1`eJmdGIP2K~%$fmtBWGU-DwLXvq_tUe0bM4+T+Io0PEA zuVa{9qpb=%F4^?i>M-pF-+mxpndD^;dums4%Y;h=OQ21VcBibHJ}T>ry5V&_U)0a7~#+&?{<~hMulemV;w1AXQSNMsXk=k8r4utHAN~Q4kR0p zmkEnSosnBq=|tNmY~sXIB|#aI0l5}^UED~SUmxBcGvHkkWF+gP!ip>z8p0qq2h9a| zyC}iX8QD)x`qcB#*+t!kdTu`5AaQa$dektpX+@lfv(U2sY|A?FsUlm3;&omb%#e z9DWAupX~6JuKmgC1$N&!Z0~oArDb22%CXSQfnn&JR$`ev0ga{&&(|`D*48 z8SsIXu^eZit>BSlMx0iFkc+XJ={;pKxFPF(QYRZonNTks;Nn@>39kjxow(e_+T~l) zdk_=7MMJMeK4vRScNp+q9lbglyOZ4s%Z%6`vKtbrD6F333?WSR?}AP9VzqPZ?@5;6 zYRrqi7RSo>9WyZGt<4vphURr?7mV{-UKYLGt70C`MK@&NA7@8xh~Do98v{O%z+VQc zr%=1&8@FOyY7WOT@M9ld`*&bv1Sga&d~d+IN12_rUzg>2S?65P>4L7NJ<<%AuA8O( zk_I|ilo)(f+$*aR!CF^5;G^xF(+p02za%%{5V;rC0bMOm5m+tIiNOXX$Y`weIq!nR zv-yFHWmHV>p#3L1mLsf2#eJ8rIa}ZF@;6US9Si4NC7JA|Cinf9ky?wXsf1#HqL`2U zsjw{4yn@C6ZM(cxj)d|p(FGAze(jtCAx3&m6DP?x8*-OAp8;4l4|%MMDuce^*Tox+ z)YeTRyvSZWcPzldYuy#}ns;w}_eGaH)qi6RH;3CL)mCy^nYEs+!}_+_5gihpdOsQR z7?5Z1kC7g6D;T^zGNY19;na^K@5@e!aW&+8j?GB8jA`gc-pI+@?Cf28#{3^xGlbnX z-eQNkygd@DSB%lRYH&(2B#DZCve8FRH*(q`0`@56ZoolB#naL{PJZwsZ^sfGBQJI7 zHY;#Um|HUc!*7}AoTT6FyH8HOU@Xu(7BkCbin&OU^SJ0S$u0GY3+Q!G520cA<2k8v zM9kwk*tx^V5ydr8c7*n1SanP<-=s~6L{M!>L?>Mmfl-si@E=4~ynE#toYhPr-Nt|P z^Uj#ZbAU36G@VpAE{0GcDqhgI@ZEXwf*!w>;6}qWH|AVoj5$YB2&P?*F1GQOe|>w; z>^-xwLu-}@?o$~vzxzHV^ zcqW*gK2k5t2H|$RY=FB`oibggG!|mUO)jMhX`OnT&n1qjONG<2X*JQ8*!1Kol*6(Y~ab3UAPLUTXp$DDvUVRCynei{GEN9#+mp zoC5gVEKL?U2|EtK!LAXpb-+1zVP{aja#TD24KsKy{-F9HX>nul^jLtWono$2z%7sM ziSCdr&QrI`O9WMZ*M-gD8zXkAP@_Fx{l4FJ(hmFu)Fw4HMNQ`1qnE<}+7_V49SBK= z5|@Yc9tb7kJk+P|Q*W8FWy+8ThF?CAotLDLLW%aa>b9y=y_$glB$hh?iw{CH4bUmP zS&AV@{Dz9Ko@kx&TI54mw^#Xf(4Aq0p%uP*DC2JlFXd?utCFFEc%4_Gpe5Y23r|I& z=t<>DVfwS0FgYJ)wxDNspIuC{gIUs+sI;*(5pdt68nM$vpicNWXw%H2sy6<4Qpv63 z)eFx+?*M~lcz=tgMOr63?}O)H&x#gbB)AmP+i2X=@14L|6L?a2pKix2w-Wbh@vlR? zzNl;BLRkFky&HKg((Bx8XBjkS#QQ99WEVUv4-{>9Px}?KkvjaJmp&wi+}KC~QFw%6 zUJJzlA^jAP93X*MO^noF6YUo=76V2KDEc?ZW6;J4N}~oqWf`TFvv~y(9rU2A9jaQA zA`+nYJP0!z+qbpLFNfcaY6IniwV_o&7G5FOVv=*f=YYy2xQDc0&^H+HL4uZ^KINY+ z0OjanL9-O8R4Lp{|GOAYN?346toOb-&&01vpD{GYaT(HaH%blyhMiHWeKpL}S_FrU zGwsGq8#|IY`RWeHa`plr4^dX>w__RET|mlmio zGS$Z{UN5MJ46T7I#<)@^jlZ<%7w(0met*$jY zhTpEcp)h0VN2e#8B^hpvCFrs=VumcH7?8inGil21q3=cQfN}_I8juQol}->eJoqX@ zi}O$v@6ss^oGq&T^EO97L&*9-eA{(ORP@>xynlM4;tJ4|#{vgiL-FapF@ojO8##H4 z6<8+3{QA~6t>?X`1Jf_w?1l+-Gzr(l$s+uD21@k|%Jn{vfR_a##n^QV9p@2w_u=X- zG)$N@$>7}gTIGc)lKWBZBun(ZUzres*+?x#TG(-QI<%TxJ#%%$L2dyJ3&I)&&NO;i zJMfz5*X1~8ir{*^z>4JcHcvH`|0_~ zoW^N&UONQ&AmCyC)?OKK@|DLtA-IZnS{Z^YRuD*q8fj#u!3;jN?6tjWx1HVWLvp*| zl?&6&wkqK*|5YT*jcrw(g{>;57+};EQL*t~1ZEPq4I&G1ywi$0uQ*;ey(FS<&S}M& z@C(A5ng)7{>a^l$bQXP0xZf|1XV6?+Xxel4$)4--iwloN7eI-_J?{&QYmC+y4Hhov zoNTI0x3iB98&PlS&F%ko@kQ^Vt_!RRZJbfTL8$|s5;L9k;qVfN)^jmh1zB2P@)Uyu z?8@oOdMH__Z5B znRF2LtP8-NOLuu%q>&YdB9ROgv~|kE)2hOrl~Z@_0J21r%WRAFzbf{8VXQS7A|F$*1$8E`lcDOwGY78cK(MRajfqMF^;t=l= zY2j7`o-_Sgglk6E#_P-0KG?X0OLnnxOQ@W8)BjiLDi%O`CGz4B&h?fg5uq)Foj6&zQn8pGI7(ld zRY@P<9iel)_sr;@b6b`s>7O&`u~l42ckmN9gC333Drt=D4tf;%ZuT7X$bg2ocfI#d z?}Ic$8q>>nrKMqGSVlOgrcKv@XhOA0QNF;8m7t|7OUQ0F4xpX082zIZ(?pR2RBR!% z&oFh1(<#r93h0h;&F@ICUbZwio7YdOqcC+?EJA4#jIegd5`bDhyJd0?VoLwJ*T&;!$ubza#0r8>&QCLQu9U(m{2CxxM{a{Sh~*ZP5| zGbZee2^>$M#d613uw&oE*odr6cw=SI*R3U!+}4M+(`0==v?uQ1cghA}341L9lhwtd zJ;B(<8e0+;hGItf_!Q)9UEpuYP{fiJ+ygIuD##0M^sg7L2|Oz9l8(s?jAAq# z#>FV^J(f$@;o{@(zn%HE85bA7mpGT)v*p9Ov0Jjvpy`7kOGl*X>NAli9b)~0i?+#f%XhUMvXBK@` zjFMd+N1Y^H^p2o(MmOcQ>5;Wm5bPc>*h^yv)i z2)Hi10Wvs)vR>|r$xyBZza6gLE8%z8O8Pj^yjQfAoQKE_vMMku*~TyDW^+)>v{H6~ zb5i*NCKI_2njcvgj@zxcUguLE2+E3b6QN0KQnD%%YtdDxgt-C*GqL{xIuo_?^VG@V zOMoedwN2ozNLNDhuIk7Rdanx9%NM(*=GtM~VCgVed>V@luy-vx+u*jc$LIe3!vasU z(K_;(hW{Z6wv3h=2mEs_Cb$%eSx1p%D%R9I6^rkT0^Jge9t z#?Oy=c(j4m^Y6`Q zC&)%o7rpkgYv&s@<=h5XgqHs71GXlb4((-O(%h{+_vY>zThF>@J$372 ztkGe80~dY1^yY*gk7d$vV|3U_J7sV#E^Gj{B`TjI{b@LvV049%=#3p2@}XG=I5R~+ z$L^rfU&*3N1ovd;Bs!&0gwUrU#C67mHC8PBSr;K?!i2|Lz7S!ZwYHN8K0{~oQUdcL zSAvtL{Xflyr-#J88)zd-%pmw1Z+QDN;to zqS+_{J*?`elt66ha@lJql&b9zWrb@^K^Ty#fJze8z~|vdBM@ZvOf?=ToSFcn%A4{7 zJS?2l?uDK%`SgltTmuczc?%oWhdnza*`Cdj&WtOL(UMr&6(<$~y!xM&GHVqZJMu3W zdPOsh(S%-^Aw;LdrK9VN@6)g9@E(UA=J3hm_>+|>ao+%hZ}#?X>(&@{I#4w8tEa^a z2E4V^!b=cvLhUJt-G|+j4|!}=9TBZ&8W(N{5oefsYG!oNExvjgbRqHGInDK`aR?(; zuN8cBIa*#4j7&QH&Lr!8mTt>du+wN0%J-mtzL|el+{}N-NesaVYrDLF?tyw{s3**t zoD%4IoQ?n-8_15h%;lTxAiI8UkT}NdhUuM9;^N2S^F=Rz_Tg5Lkw!8jM?9SY}{? z|MCcf=GNTPL@!N{HuJH_wBC3{oM6|it=x^Fpw+0_DKco1!v{R?@;dmLQ`Ye?tiNhO z3Rlm)AS@Q3D3MNy-#cmuvVqgnhVc1&MvFUEaH8sSLi()B>DdHL+;^HKc8c$dfF1yK z%k(|Pj%miGJDs4ihoN} z-zI1ccTsZODbN}%ex7mN?<2}*zUUn>u+7Q-c}7NrNln32whj|gyBzfx79&@pTe3_8 z`Yc63C8|olVnMQ~P5A+IAuJZ$b(sRa@v)C-;GZ#YV?D&mrc7M@v)fJ6%v023=EMr3 zbK@xrl7Azns4|LyO4MysEKWi=tqd&K;+HYyYIv>qF|UEy$Zh9b_JUBEESb|3avhqg z7$pi;LEi#Ubu{+!NaYTIf72wn7Ev8_d`gYtqhM6@ETu+0&pdPN#snDC9Q4flkMredOz>X7jAv= z`sdn&*2I^_0Yo?klU}iW*CMa(-?+h_!cht z$!a*-y?q38P9ob*u4PX+t_{LMXz_g5!uKuq^{I}P_Wqr^@-jigB(0ZkzKdjn|Pf4dj;jYf z^svUt3P#qrK?7e?3t!Vi<7J5TC#0zNg)ECe5o=6W;{l}RxzvP>1wsyCVU#Bw|5a2_pxJLO9eSgNWV`V~ ztKPz)ub>zxax11{@4a>jUiTJFhb$>19*DMgp)O?5BYtwc;G8f&vN`-N9Zzr{(&GA~ z*2%X8^>X_-@smH4*YoxKrC}TW^TD6GqQ1D$K#tQnVuQMj>k=@G1q8!(W0d3k%**WC z5gQ41D)R5ec|SHo=Er+Cf07+C~_7jbr~E<&hxL-17)&n<;+ zDm$T&D2p!SBt)DO#xv*q?**V5;+nvlb9bqYEj3Fyt@1{wbGaN|B)BET-U+CSazUQp6EFj@3@&N>FQcJl=Z5y?pHFiGK~;Jrzt)6}0+ueuW5t}GJZw?I@N zat}lW@b3qa8)r3z;NRM9K4+lEgNY`LvcUEH3a?(-r68RW3BzyOfvPjWHVnGa{j`la zR$!ZWM6-B3&kVNE^lu*}%U&ANM4kmSH&V<7imat#u}Oysp3tZLR^)2tRwU;1?}X|3 zdRe{Daous-kDu*c7BFIa>oJ{Q@Ke>_DKju0`#e5H20d6I%Z;-Rn=OEmOff4dl0?Ox zfLLOyq$+qVq@A#wz|MO=rl;z_#m@(X12=KODREz|(*M*9h}W-r_L56(JS`1a4Dp8) zbDJVJsaSMhvqUKSZYm49BHj=VqVsZ-&U^M`r~%=w2`o{8c&=(4wEHntVWMyhR%M>^ zuZHXLLf3|B4^K4vBrp%q62--0^sAoE;xEQMK@vHD^18(m*|3?cz;<+8hADLw^ zL0yP>Uj%!GuR03N22$mlrH^yR;)8c_lS7sWDpVLW?4q~& zponjc40Vn2)NR}za)O)*=!-&vG3vwLF7+_x2#uAojT~H77X@vZo*+%4kkeYDNEShJT|RXCmS)lD ztz07Q@+|=enObF>AWNi0;9Q5s%}pWcGn_f#mJfT3!*6*rCw{U&>DUPAObAL$u8lGq zn(yu^$tKlq>^nn?un}qfgA~(1kvfy(QZo;mo3u(mrX&$+YA}{lPv4NI`R@`Hi+Z^^ z)AxnmoVA2^grjW=DK(k}?6uQ413X`p08f=SMP&+V=9Pf+YKjz}Qk_v5PvKyJLNX}) zUgDfq>!pQsc^KLsU~k0>%x#*+0J1Tn$i6`_9vMBX*rDjh0avH~(u|_R-}oqyw7W5i z;w==7?orHLihP8DS0RrA`SvBfQA%@qXl+4EZ<6y`==cC+rgH zloitkLlPCkgjt(*)3p}yIR?{w*p}V=*Q4qr>&hFqb*t=@1YtcLj_zVyX5oa{2608G zvr85=OIs!Fa@0a9_R{uC3ZYsLdmsXh86roLewUYkkABo`jg|BEv6=@b&U*dSo7Os^ zFNs02qbk}#qiun-y0Q7C(Tr3Jcd4<|3?<^UcV0=FiJRzFZJx4vR;N>BYo0&8qr%hk zZ#(8QW*$@i=0#_R;fQy8*3-v4w?}H1hGoOdP)~31yX>{N&a2I9?ffkPHv(?>;e=7- z(l&z@gTa2lEiQN22UZ4T!hc*UoBPkDSX0jZgq$RCVRcN7=hwUvPb{Mscmj-QAuw~CJX(KO3R2@|p-pIKcUEp~%+NK}jtPy1otWSLmXZ;Sg z?(^XXdw*)JN#eHth#i%YX0Rk35>w6_iCC@SK;2XK(5F1Rr-JLzAX(1Cr1+vrzb4M2 zEIJ_q*HLhwg@ZN-eca^mH9pNgu1k%NW%S3m1h+NlPxi4hD^)a#<^Y4l%p z%5JZ2F9U>Z64kfDvN_NuyV9%D%cVBp$zy-A;W#dyTrw^jtZ*@L>dLSelXs(ZR;db+!2x~2cq`!?z7uIipf z#GO^a1(ZcLSyiG?s9_VurBqN3t1yS*L4v8g+MDieUqW%BWr{FF3-Qc_5 zoO|xM-}60CmYrWH*{0dx6UD@H2go7d1IBT(LrR2(O4%98C_#x3O5u9wJXUc&7ME~` zp;C7qeR3qO6UBTwBa75Sk@N;ox89{b7_e7;j&oUD&Ry>Ziq+(jIEULVU&*^nZ}BY= z-j+3Kc7+#{C`d8-+ilqy4gO&#RIA);#V98Y>u1_PO|%ND)+6l6d^6c2Yaq%BVN~wD z_j=4VOxUc59C#buf)ooX)Krl)(7w{U7^GlvKhvrqx90V;Xk!(ESoCMBrr-L8(KQ$ayjklSJ zc&ZaSTwrA}rYa*Pmx7IRQl@YA22$d{bJ8a!b5bS6?4v+BAR?awbpe4G)gl>k?F8i# z48|LnTy1WRT!_SORT!M!2jI=7!`d0M5wHQ&QSSswgihARPYfFfD}V$R+`NihP|no_ z6)J!6i?Z;|J~irYPGwLY5N;Y#OL&7@1AP<<5Kf*lF%bZn;c3j%$zqBXfTnl{MZ1~H z-@GJKjs>xtGxM*yWUKf3f3Zb`J3hENO7ndPrsy-*GxMM8{KE!Elf?#`=YKl*zl+=- zZ+6#gJxA6%uxSAyy-}OpSrn5&k*&Dd-3bLs@@SY}a3>pmizq<`2>EJpSLliy%R6D{ zEyhL9+e085stdL;TO@%m>L7Sr7zt|CczZOv+6X@5Y9q(JmIHxg8hQv9RHetzTv?HQ&otdH(q_pH`br_tBT>$OZPY zBi4b%)O47f*$Ff3z_RHtzx&BQR)R{ICsOsGwgqqol6?DI?S+qVhIf4UjMIUgZpWW% ze=61XCGkitB<4oWjRviwVh^1f8NJQSk$L<4Fl7a^Q?|}L>yCK@P<#@^NVWY^oe(`# zBmkHCZ&5`Gx@7Tho`qDW7D<;Z!5QnhVTH=Qq4q%4_>eKNFJqtB0X7FyA8!Ah(N$fw zI&MDcd1(mXYfbof4=DyxqV7@=pLiW|!%69&&(WFMZ1r{bR$-DarV_OVHE!HhP-6 zxvVf08`Op6(dz;`L5DUMwy%{5JQRf-2wM|g;{{oliYiF(#QvZ(vtp?r9}4fH0a~(= zvs92AawGV*%K+H4Rn(Fj!KZ}j^cD38={N&q3u-tJ2&m-V2*#@;)-1;5S@PRp=Mh>YK1t^;5pfj(Vy0nBwLOaOYXXla_87L$gO<+&l9u;+Jx+c|N)Udw;@ zGw!f>*U#w1_O~p*>oG{yj=KRz&-_Eb%4mCLrT;yVI}#MI81t^&6?gX<*Iw- zs{{jKRb&_x_G%yXb*9xZVoOg_bA?aBX({3?1Dmdk8H4!Qh8}+v#N^HM+5v-D0fA zjf&`-B!j`&?r|^hDe5*@>|v`E|7`l9xL1+ni~UdHdt$rq#8_XMiGNSt{iW&~|6&AF z1~v2za>9Z46h1eB$3=>1p#U0F194Yqid35eZP6+1kUtE* z9h?z@(L7uSM{|JDO(I{7hFmqnf3+$UuS%8o&rFsZT4)WQia!m#XD>wWwKAwfv5U@_ zt{rkq@;$B`l4ld#;07+lv)t3sukQH|-{?=be9Uh69Ja{&srjXIUUdF2rf6b#2;(al zQEAq!;xtIpe3Rxy2Cn_eC77#HBu50f>T7d0%!u||7ML5@$1L^FUf9Jyt$;-x)5psa z@AI3)tYjG_#ygV&@wmNDya6i^Q=1b1^ru-y8}#8v9s9}lX;6w~^#1h66tjmSrBuY@ z%(4+KLJnOqKU>`)-x+=cJnCail$#dSW2y)i^wBfgL#jB@E(L5lrTX@R!*D-%J+C@o zC2UwWxqk}LKd6bF7=}CCCesPdH#pcf8n#J|(3q3J#7AZHldML_aaU;{EnKhWq(dGw|F{X#rD@_+OR#0K{P8Cm`g!u%b)*U8+~Od+ zW>6Bx8S=tj>p7nZ4LwHl(EQ@x1|=rpZ!;L-NnvVC8#?KWBj{vhL#dzcz5JgdBLXw| z8?KPJmnJSyWHMQ&QA{dDHc=6qsobw#TZ|aVS0jfRtA8H~ZI< z^e&g)y1n@HH;#SJMpN#1mV-T<#_Q2G_rG#*Y7O5AkT=e4JV4e=gK|Hk-JleT*+`Kj zDxyHp1iHWUb@xhDD|}V7HQpWaZs{TCMrluQr=(Ge>ns#jOnrWdOKjh&ZQj}Vy|DU{ zI1QM0QcG)Byk`| z!~V=dw&49@{um#BlWM^ZyqALh^#?Dy=019>Cnc$nu22+q((|GuO?fMC=6bI7}sU%sF%^v0`^f*9TqDDAK1L;u}M*`LI*BgIi6X6xD*x z_cpKEz+tEC(2I&;rxludIPR5WRhT>UPL~p>7PC~aR5R>!+2_H6VW(9L{@kNl4)@p0 zE~R5$Pxrq|_J*d-P8#`$<~Oczu=kBj^z31$(%0+Z=l+qO$-GqI>5;xq1`IoOhhArf zo#MG2Gl!k-d#@4@Tc=tdnr`*X*;YZ+Hk*$;j^)o970R|XygN}lD$ zFBB62@^VHm0^%tqh9aw}hyfs)>1OOxR?J7e<8E=Ir}jCmpiD=}8myQ;@qEsTk8G#> zjEMN{j=Fe~<-l`Wjma46C}t-`im8Zpx!wm12fri~lf6*|J<55hM<-wJlS5yZ=czBw zYYd8>G3>O-qg}o=aH$}LZ#Zw4Uluhp1M))z+tzqj`5cxl^+@;ck)BqhDRruTVtc4q zgM(3qz|x`3!NJN5Q2+k<_d3iAzOrG{J1~92LfzWmHv}ycM z5~IvSDb&BKmgEX5D7@UpzbMEQU3Jap#xgOy-*V21hwrz6~3+gk+&8FzTbHDUtnE;GQHJBt!E3b8~=-h6%+>d6#@4i);P#kxUB6 zO-4=!<_My}!sB`%TI<{osUEn4%qVgM69hMBrzxv^Qhl|U2)2XUDulEv(Bem({*i<* z{G?<#atf;@1EB2>3->iM6T|OtIwkwmh6|XMQ61jQ$>BCgwMa$Gqobs2A;U`-TH}3z zqhB~aDkqIEYps!KyK(WhKV;t47kvi;ho_sqDc|h6EtO0OeRIrFx4-X86-PIv7#19@oSJmmj|UdU_OCB z4uGlIGg=B2VvKa^Wq}(xm@$eMv<=c8`mD!bus!j`9#o^|3~SIa!8Y$SCRIv}%c!H@ z_~u2jj@>fKfnzWb)*H2q+D0)Tv9kr%*1;EqOGbV)Y0hYL!u`U*89?ZgcatT(V%XW7 zvB@)@$q>f@%j@inetEBAC2zN^&NGfv<84oX;P_1#BoiEbYaiQCS)=^N%Cl~oXIRtTaFnL3RBdCLHJy?K9`?!qIr%>93mE?c~z1;Rhsg;J4R4o3W{ZJxweB2 zO4YncjPBF#)v7C2g2Jl4Y}x4c=5K8`FPGr3-fWr4#Il`YK+SO*6#;qak`~E^kUc7G zU*LVt1J@x}ZFESS=XSSBRf{BxTQOhzDH(J<7}g=j*t}%t;etPHPOX3gP~qh|-eF;sz-)`hGd&dSZ4t7t04eZv0DhbZrYQ z>ri9Sl8nA#Q!ueY4fT_J*W$O0i=W?|NqLVPcxe_tS4`%p(-hM{k>gZE*;|)jjk8L# zTAsqW9@xMKQWPwH&Nz2M{goA(HRSf9M4rw&j)QgHfd@K!X;@ZBE!ht>TrP^^eX(dy zQCOO9gJk)%yWVM_Xx$=T0XZ)xr8=0!GC+f%UUWrR3k97^KzCpVcaZecd*Bn71vY&7 zk~lHk(9>a=p%24zb?Vj`4KP)*$EOW3I2w=U$E@+W(lt52#fYj8wP*iGRKG+t2PXW^80vmLmRHwYhF(T@>NA4UZ%h(~xfrG!WB`^w6>nSFIBC%A&QuQI% zo*%ZBF0$|0~46;}3UyJn|n=S}fwU*2k(>mX{6KS zb)KV`CWu&zk2$F_%I|rUlte9NToxO*TV3#X({@7o|{ zH>_3;iOCmkf+wuTi`_)wuu3%~p*VA<5m|SBC|yCeIk5XvZ36jHih)wxMO4I5b-jO+ z@{~}U#7P3>;!T_i)n!E`4YQN%U_c?1$t@IYWAbLjxF3~O(x1!r`DJm7!fNTx&{#%W z12V?>>QgH1=dwfUYDn$h4*#bq_jn#v+avm8Gc>cp>iCUWp?S)?Kl|(Z=DFGq8}GA} zL=J41bXAQ+2Sla@*)sqA0U(uwoT@y!5c10nx3m})WM{9&f7cdUumy+3j%<8`9XKjZ zE&b?4uY37mL9Ajs{V7SO^Zhztm$^CYI@6@l=D!#7l{1<g_36ertNXuO|3+3j@SK%rGG}e37%-IU z4N_iME@p&lg*9WYyN^_Cz~BrL@gAulcIVj_X4SA-^NPo9!x9_g_GhaGtgtclkFC|~ zziPzB-M9bgzeueEW8<<3Hcn9tWYT_u`X;c7E|=(eTRdWCYh&CSrCU65LC@r{tWsFT zN%xQUy*4L`_sFSH8X1_vK^4CCU`!k+RkhQGycFa-q%IgFka3IJ>6O6kOyRV>ah01+ z=h7v@@&%x=PJi;%d=AKrqQejE+Gs5AV0gSb06z}XX)A?iz@lLfOZ}TdKlWiOM!^zP zCWeFM2`jv@1C{jBq7Ex%(y_v1p?IHa_skwT8-F45eDnud9{n+-6{fTm@6ec2H60ca7DVpB&G7GosSTYCZFlGaZg*%t1XgNQSli)j zP#o!%mrII)lUYgM4$pGSn4KwlI6G2asZ4_nfHY+tuf`kn$D-dxW@jV zmN(rm-3u`sP<1mLcENT4sGW0l!oxCLpX-Dmo7DvQC~E`Qx*(4S&MH&F(&sR1{`jL< z{bmOy&vpOvjeo2hg8j+3I}pa$dhQMRGAb!FfL@T~$E6%mUF-@U>z+A{X?G zR1wO9LP2?Sg7Xg9iP!CJwk8THRtT9u<3;zXjwy|fiRRe*`$;CdZ6pUKJXD#uEW0SC zgd&AhME1fyxen-9P{Eo0#8`niE$)JP3)Q?B zp7sK{5sby4fDH_#?-nM3tV5b|g5Ve#weO7X3UHYoK653&v@j>H)12romq#S zrz|+GkYV&{w6t9@*dG_ zU2$>ccbtt7ssB>-|B`42j>T>>!EO@8BvK>}+zJTXAswPgi5iAO(;-=?gM&mT8^{cf zG4jv8WHnx&-HpxfTDB5ntcRy=&RVp_Tw}~({R|7LVHNZxaiy@Gu2ijo1m|WAGAmcf zS81x@xRkSu*G8_3v&1I_JD}2OPcX7-QhiGmIz=WoopVTSZ_8#iPDUBEkwcp!!?Z(I zIH6*f-Oqc`B{!C&aN~CyGQbFc zg)9Wju8`|)$S<$fxn>sy%O*GI>2ON42!-&vH!Y2$Z`b@Y-(WzUSY z0a@ZD)pWi00qk+$8rhx52Qx?gJ=&5nfLtD+8&D zJ@iLKVSTC$)sTG9wONy`-X~i=1Lbt?)4M{C@GFIr7pJw`xJ+(XrdS@AF_s*T`&~Zx z*Ilv12%=J_jN@db122(5Q(+XkH&RRzMG~>XF-jz0s2%m4Hu?7xZ3j>WhFtIXpCwS_ zrbB+;egX<(z=#!O$9Udemsmk|O4onoQPabnv*VjTi*hTM)-q+nRarq`e zB5cZnTmDgkbSU;QNe|=c7dh?~D-kemKeldQ83AKVep8)poc(XEagK5FcZwp*$MSa_ zc*i%(WZp@pm~|A1ry~03D6*fQ4mr;|WXNZQ4F&^G06EnYWr<99sI0^a*og^$#DpHS zjFq3v|C@)o1Is3M;<#RwB|p56J|pdsCx&A=wPN=TpoTCk*Ylc{7yaUb`ox){Taw+1 z^+Daz7#?O_q|;seRNqs69kL47lb#h+F|@cT-6g{HRGPApbkl8oLqGkzqOf%DV=$Rr z7GIHehr1W%AHg@u`3pJr%5zjth$86u75I($qS$Q&n;kL5p$u+ZVO~#mcrtFcrt~U?G zy(AXKLg4NWxg=f+wPQdbs#%l4*#cFld*~GZY`REN37+c;Nb4%3hvzPzcA8fw%ky3_ z%?75yiV<~GI1T>W!jmsLSsJsr_WSL9y@%F&X?JilI49nCK7(pD?$(U4*(1;8l$FVt z`loe2*=0^Eao8n{1-?s*q}EHD#Z8v$=r!*8pf&DT-K8?9%I9`)nJ4}P{p&rlUf9CH zpV}-{z38F*q?i3sV+}-}@)4ec!^#RGQ-0I^?>m2OT+h5&d~gQ&Y^?Q+<0cJ}RVI~O zhbZPQMQ(vux)h?qoFb^&vBA|4-Mz})$tezPbr}hwMGDexfrLY}V(X{3x->ZNRYQG{CJZXsGWWmjAd;Kne`c{>R6^(arh)WNq@V`U zh?_O&<>EEmDBLB#qDBFsQ(kRuMNsEK?}FOw#oV4ydroo_#r4RPH<2%rmCsnl^}P6Q$yD(d3?=H{7qNvq2q5O2xj zT!Gbl7i?i8@(RP-=p&*I`2e>x@Xnk}9~cm+-F-+}$N8uW zB>{9{$?}eowDkf}7mbmSSOzFZxZj79Rn@#c7kh+RZS=0K{Bl~~%<|WGjM#_chSPTr z*QPmO6{1M*lVpHmMN@rt

    ;Xc|3f@Vzo|obJS#XkXC%wc_So4%H zi`%5kqsyg+bA&#oN884-*a+;SO^o%3env~nU_bm(-TUSlsHG-D@hc6SnUkJK3DN&EDK(KQbC>MfPY@65ntbZX0C6k~3 z=6BlvdH>h{^~Ya|S5V9{ibRfnlVct~qf49Ivi#=9Ff)$3%AJ1Q?@QiBGxLx+wwk0l zFtcU934Ti`2CCz2M`lYhgGftF4xF zdUev(U~le^kYONgU6|-G0158~35~O%cWlp57(tv3Y=}*eYu*VfBv1Y5Zxst)^lYJ{ z;UJsD*LgQe(Obd@OOvu8s73NzWdt7&gSD`D-2Qp}W(A8WVS&Fr;cB!a6<_`ODYDjq z?MS|f9Z98_O%#j*$mx8dJR8(0*()^Nvaqowj_ZqqFFV?nq)^O8 ziX>4HxEE#_j#$i~N>zN@==5>%|yieqDK^I9T_j8EHpXc_wE+?g&>yi!< zDOmQrau0DxSbzhjlBWmJURNxGW9mC6PR%f`UOE>BUm~dvyn3lHnI#G-29S~SIKL-Z z-p0q(3u+SQ(&&GVY+hr;!;ty|B~)GVK(rX>0=XgS^c8SQu?`9P8*vL7+zzNuOYE^* zU>H+&$oRs~cyeGTXjn$Z7{R=$;Zr{Po_T2XCGkKkM2Lsz?1eSz2In;2640oI5Iz)N zC}nhX^7LBKIdO-)m#_CivdYmpyCJZ)ly}LukM03Y>s8b$`%g?01(lt!^hLLyx^b5Y zj7H`6y6Jx=G3<7Y9oS1LFfl<}C}uNGoFhez_WM`)G)dE`zQvf#j9Fu`Oh$+ns(3&Q z)(|_XmPIr6voT-1f8zmbc$jx!r*lnzLGd&rJT#rZ*OH}UZSXlT2+~bJkVr9c6j@6} z^hoc_!Er`K;>fueGE1kMndj@C8ivfOk+;+6KC4^Hvkh}MH&blHgzWEcHIaA+_CtWL zFe=YDonp37WHS{J#p{^6Pob z%NeT-hbdIPII|fp#^{J0R6e`$UpZbzW1@~-T}8G!urUGJ*(hUDOfleMCDhdxH`%;}-=7{aUYvsT_ev;AvVNZza+GF>PZZR(K@ z$rFJvvVQuhFWYMpnhah>fHTHe+u{;Cy!`X7w<4!J#!FWDpUFcfRvkESU^UeQU9l9i zh9WCrsuEOzKqTzwVeO0p``Q%C<2gQhEI(_{J3YQDlV~Fz?!4!Fj6|_pzc?@+GEDHW zj$+~|5(Cy|o=!OAs%?;5b=ex88v?-C;MPFvWP=iWSq_WiH$Er~Gmga$>~!Pr8?Jbm zs}!(_eK_!X#)5tUCfuNPfFCaH}D9dY4?0*Kr{2OLdd| zqZ^EbOPnQ@3<>Sda5=`1WQ0y0G$JYLQl;AZ+@-U=2<>VRCSLV6eE zsyqtDifZshm-L4>Da%=wKxg-6CN(7NK8jUqc0=N@*?qa^ul?9uvf%0Y$br2P3pL+# zv_9ytxKmzFM)vkF&}v&#VI^3sfv=Ta*|eEu_=X10`h&UW>adYB3%=+j*xjv`41?Zi zhT@iIG$-D-hsJtI=#%Jep8T6 ziG+DDp>;t76{8zic34{`+s2?Ac?R&u62lrGp$DpDq^VEQgXHn8Vh=*If0BU5U_!Jt~iva0CC8liajd#8}F|3lt7Fg^Dogw_oc5&9s z!@mwwBAQ$ZMEDnEtFO($-T$#Q9&I#?W~0jT)NBax)bcQnVGSJjYtHZZxvP2J%?o9w zEL83>Bpu$Fn#^sLRe`oj`9hv$!*Hw^g2M)#6nvD8x zirGa$DZ_}9P)7r`94i-(Y!UNX#i0EfH0Zj|d0AkzG*S@5+v2NJp=8O0fF}4^!x^Fv zdFOK@17eC@hYCWQtTfBrg$m$cJUw>Ke{wQf2dZJMJs7_T`V;k&PR$dijnU zvg!!f`xEoAowo^e(tS zx*VSFcMaCZxM0&3agWTyDFt;h_J-P(+iER{%?*a(+{VZq)g3z!=Scp1Z;^50`Qf!w zsbudo(rjYik5WuMMe3-C-HOKW^9!r#?ex`ocg2x{N+A@W;6L=Mrk4xwtV7;UIyh^| z^^y9GdY9X>0Xc}6g%64ih5F7fO!vpU%V8(Hb5e9rQcWKE)rY0{>gXr;PI@-*FY(tc z>X2n9I=vFZa=8!v@KbNo_4HxE0l}HcFjyxyW@C@!ldGc!f#o5qF>~D1sGpu+{L2^J z@Ejw>p@w-92uT?st-{!lI1vi&W5v)NE5uNN%YEN|%ZQlbnVddS?ZAk+ zXo8pqiaAb^qbTzk=XREyk)D!N(;1w&pxd%KsC`%_xdqIoX3fR$LDC{w>$Xx*MK;eo z3#E_;$uKY8caPsCznoe19E_&mZe)R|%(Kj;S%U)9h&k*5b~9kFa*s>DVi@R?JHcsQ z36$)+s$p5b^panivRPBjKNxU{*0*#RB{?_$UlK;{$Zo!uy#qY167~tFKsED%pCxAnh_iD6MTOsQ7I!0odcCcb; zO3*XMWYcKSKi#*ul5Ur`$U9{Z-17i*TNR*f+CcC1k7WwQ`&7yDtFs451)T@T!{$A7 zr4%%r)F7t5*FPCD;%gS-Gqy5iHb>e-U^T`wKfN;N%@^IVX`IndR=DOXhXY_|xVp9dpuH5w8L6I-qXt{VxicpN zw&8Q=T=lTi$c8x3e;+v+9zrT`4Lv9VY$xj=xe3>M5_25$=2!tQ%7-R1_BBwoOx3uA1LP2Hkc zpnKH_zke?L>CefQ7c9#eREtMVTtyTE;(@tT#CrKss0IOWJSw}u$yV?6$2shgQ`>^& zoE2Mn>Bc~L9oR4aFr$9M9F&y>2ix%6*vazsW zWB5BGm+WH3EL*VJ{ywpJ4%!RFa@HkFRdVV&y8bto` zTn>>PfrFA$-YNWZFbOrer+O^)$Pje$V|Z92p;NXo{MO7S&CahKTYN?{>}0sJeA+SQ zJa?=6ifLQjyVT8^VwaVkQIcV&y{|`l$1-}CO`e^ey@~?v9a2lS(}(FqD9L-z^8yDy z89%4oZ`&(r%8jobR%L<=<#sxTx0@xWZ+sM4GA69Avx1jp6pb-4QGb8`2a3tiBOEsz zZiybT+2<&&cWDq-3zFqqR2#Ut+9;bifC}o61Pb9 zsGU+I)Hp*#P6|r|owo=ki|~%~PW-Mp z(!(;Wd7>kHU*yQEu<~u%@3mRznU&#o-0zfR#fME!Mr%=Y@8=(p$d|@i0Kab3 zA}E1kVkxqQia0FBh3!ZpQW;b?%=E#Ug3n{ZFb9%V$TolIIh-#$@?l+=5fa}iFU%l1 z2X(P%+BI@{rKl0QX`6ve&d@L$+{OThO*NHS=%TEq~o@r zVreCPL3m+Ad-RkGOpH3&a?gRVyIyUOEQ6{WwJN-wN?f=u;}K*P@_%^3~s#nBCXkW^B%2I*i2aHA-0!LXC* zwG_}c$PjH5P44?>#lLleZ{F5~*r)zd%Wrw^!ryIp|op&Ot=F(5}qAAXkh0=Yu2 zyLHi>5|9&-YoX#H$R^`raEth(#t4nhkXKX4J_nweJ~hGYF^V}%k%P#ms)uZ(G-bcy zW9RFytZ+pZ)vd5j*~8#ZILkewUq_uF)M{wfq=HaoZ15q$Ro4pY)cnm5lE;FlSjZQ} zE0G_CY%1jXrf}AHABX(R^qB`dhn@Po^e#&UonHNlzA&9^fEz`2Irr0T{7(Mo@M&qv zKKh}ek9Teo%sqB6XKgGd{B2+bbL#WomWG8t_H*)N8KjKe^yR?eeb`kR6^uDVF$XA8 zO+{4DZ6r4|KQu-1se8UU#iL{HBfstRLD3mVha8kV@;gj+s!)#^5U5OHhJc-iOsgo! zoJ-*#i!Y086oGmbp!dl$1KeXEta=XGebg>L2?=(s!c3n|=x9&qQP_L>Tf4jnE@q1j z7bdkfuqW&$j5&Dlm3k%PZUoYIYPOyu>)8RxaSbkk{K-*3%AyzuVQ-}(s`>SFBIyyv ziBL!aB|-A&voo~a(tHjwR8FcJq+N865c9>`f_0(5RXO32L-){`YJKRWAYjEXO&BFs z^gWp-Yd-vCxp_t;n`CRpjcOAMiJe`t1m^_j79n_HKy1J`O}c1UmgM_?A*5?HDkd|; z6L>b{n@(|mFxn1zj$A-)va=lyJPE~`F!df$%ma$tqas#K8+KYb?NTU+#e^(%F`S)- zbnA1x&8mW#P?bv96?jS9I~SFN;y}L2&_b_{47)8e99IP#4BMy9gJ)MxtAK3bW(_R9 z$dH@iZjNAs57dXBAMg67KMi@;sy0&}Y@$99bO{W%3k1=yd^#xECNZ3U>b_O86Qa=R zKqbZwD^){~^-U^uz0kZm$AN;a*wfPnUoB7LuEtQ4MhXK;#UqPMm>5$v(z3Ly>4=AhCL^C)_Y5>bJZP zjK<{rcbCs4x7ZmI2L{PHlj$saDzkJd8Kxp~>3Y$SJcYT=JLG;te(yhWVu04CZE<78pqM|w$I<-lgMmF4HT&JXSTrFuR7on%zgMovu2A&9nbK$<*NH2yYEwG zP@+?nshTxUKBq#wnZ)pb#SWCPa~y}A({^P2?JczthDRg*cReX};6M?CmPgG}wG>lD zk^NMJK76xatGb`=ee;rdfQvlNd`Qk*D#)b=WJoiEl8htT#Fs-i!w?+|z&(t7#i#P+ z(>7E2ic|CkpB2*%k>%4?%P)yHa&F67G*RS|c*V39d79u;VFss-|6oDSjC{_rz%?W_ z<_>JJENEbrg)V2Z zcz5RBmBxDH&1zBt^;o>kD@U#OYSy5gz+DPdh1?}S3XvdGxJdVn_q72m#vQi_;bz=k zC*Qy_ZjQf|lf)PejqCk8=gG#GX1?2FVwUnK2Kbs}wN#Mp z+N?RQJOXjR3{DZJ2dZπ_kn+n<2e=w;z*-ffxQrAiUUx#fD#zTPv-%EC4_FwcNm z_O;xCe>NiJ%>x7fPL94bNa;2~%6W=8N0BDn=~y!}F}%*J1hi<+NSif7u0^ly5;X{u z@;{uIUz%?y|U%6ZS#s-(8+JsGzeRmRROmoO`7hvZ!b=Nt9vA|p**mK z$qC66HEE9isO!hO-#nu^M2g*SNmd1P$eJ_ImcIGc0i{sdt^N!;=oI*k4=n60mbA|kk%5hi*t_C%)!hRq~Cxz-++Ao+VG3A}r}?Bo}-0n6BP zihUr%3M^CK(6%(P^6HGFi-+ZJeoBFa=oa5+Y}1X zWBRJ=lGPC3FAL8UMfsv9qW9_IAEU3$!8O!%x8#t*@IHCotQdEl$6>mQU&U$Q!5z@{ zzdE}xbcd{qZwK5=9y5p}+vA26W~e6q-(Ad?R2J4Em_vXS3_cZJ4@?a~ekV@;cQ~-y z2pI)%b0y@^)&@Ls%5)u&XK|C|*la{BWXLr`6z`cDV&5rXQUhZ9armM~Rs-U=`_8w% zMSc4RZ$~BzWkS=~H6fs7X3HzEo*I zk*9nYJA%X5_3W>|o9}ION**%DR+BUbc1re}I3*<%Q%I5RSTm?|q-s!;vIq1+hFl*8 z-wxKIs!=BpJh2=tQbe&-5LPE{W`L)TRe!Kz5Aa4gYb3~6Z;=d2>Ui1?MU?-Ln|(5- zCI$gBOxOVt>?S~2M03DABVf|yKQ17992hXCO#pL{VxT5M6%~O+nOaE_$x~G-O8_Hl z-43dAxz~s`N`Uo7dmwGUNvaEjyO{oq2Sjpi!FX)5hgF_wowt4%@hL>fUz8L@0MlPtB zkJRzQvO8hhX;inzqda<(zt+&=ISjM$^|Bmo^FguOIXmmI#%1~gbB6iTgarp);#lC3 zC(Cy!T3x{7_(YlHi;4jK{$K_$aQj@=F(#vHHBwBTvj2%?q-5RRUtzx7VKb*VFu}#b z;%L>hCe3D_oo`&#eEh~4&0*kYwn6&aW&SZp5k03U=Qe99#C5aQ_}%7%v^Ymccgc1v zXkpd|bu4Jsq>Fa&Zu2py*emaK#%tGR+sO`D4W%cZs@ai!VD+nafA%f&9iA7;j9A$9 z!F4kFtZU`h=A^cWSZ4@mMIbZrbO_3!eG6m-e#^u2%9 zAXx{1XZ2v8dzPn{G4_b3_Jsbf=|!)$)W*-{#&h%Dyabu8XQVeF+BOLMwxuptT_5l> z=prDP6pQsinW8?rgzl#eZ7(c_>{#wO$yNCl=NeuNZ-}-hgsj7g*_844)IKm7>6H~$ zC|<#0zOQjXQ}pQjWhCRJp)OULEPr-V46K0*sE7lLOXze}nzB0J9H|uc0877(k8J#N zEy#!QY3xv)n)tfZro44?jA` zE;ySMz>Gk%2hL&!Upqa`2pLW1@3mwpJIFZj3J6rsMy;3=DJG60YpIAG+*YAMOb?3Y z06n{d+?#zQ&@O{M&VX5kiE;X{$L;@^*>Un+BPLe;cK>a1{H4LfbrVdqQp|aZoTDOc ze!VK-V<0?q(zuzE4)NDpva=Gbg0hAjrL{W6VbEc0k!%RrqiXhymiE%i7PLVz(hm7C zX0=?~!KnudQ;DFKR0nj@hT;-c^NJ|cB#IMt%6g;+1Fi(&F4BPPq-Q#vOP2_{=xrM8 zf5;V!=s*nE;340qwvL;n?&M_CQKZ=Al6c9(;9Z=3@1ZePzD!u zDA$pS`OAn-u~vdr)$!iy@L?x?km1N~^AihbJ%*UcZ~Fji$`qw95mQD-fgG5Au;ze z95%vg!E?Y=i;+YNWFv0#+AAxf?DoZe*3gZQ2z&6~pY<0R?y*C}SNkrW_cxk|`Zw2A zkZg8qT?h6&fvh&lpzNj?5TGleBC^$Gp2#uV>N8R=uV1m#<&bwA?^f7I&Ac`mVRgqp z_O(Y&SnRO|QrJiM(I`Il$O-8sDAx{gm8q}QFD5~Q=f9qoQ66Eg*yFI%ivFL3-cMlf29~Is17di$_!k2D$+|H3{zK8$ zW=%bft$N9_1-lg0l4xloWSrk1Ss_{65<1uSiu{4-{K7QlE=41K735ZW$x%hGqLZ(i zz9SH8AAsOCX{0YhMQPMIi18Sf8$NY2Q%_$P@1;BWZNfc}Hk_lrPvhqro=)MEDjKBV z5jhF|CCiOX2huS@fZu!z$dbTqt7w!-atU;}@H zPdn<<(f7HHbS-3&;eEYJq#%!u;$<#ySPdRla~c)YT>0L7oH}v449M73Wc~Cmej#1K#Bq9r zsL`@KKo^Sh+7{X@0 zy=@;HwiRY32n~ymMPGV1=^ra?tWdCx=?1u`w!#i}me4@6O^}mhnOJ)556MTS`li;}f_5}ly@BGf_y{uXtH=p#d z^IjbHUq;rN$X-6Am?4VXr6Tr_SP+>UlCOBBN49d_Hc6fMxYseS8eaEYy>gpmK>SeL zD>>s?%2l9d2&ra}SO%NIOg=XqwvqREYf)sQlG{ZaE@@+2GD5I3 zOeSh2>C=+Mz!E(m?}9~_E-YDo z?(+P#h8`1m1L9=)ZI?=Vl`Uf069Le83n^Pb?9V?ac<>`5fPS=Z%iHAqG%{%N=0BsD zE{b$e5xA6Sk?egP%C~M)H*yZBSIfbCmrLrr65T3&ljXRy(B{w;^I;JxOsBKl3q;pn zS@jyGAU|@ttVpLL7bK~BRQI89Umh@y@M?~FtD-Bo5ekjx(L22#Ii=GDFkkdQE&2hu z_ER$G3NEZXN&>QMu;Rt?e#!D2bu)vHVkTj;pW~k)4?+qx&Sm?uXD1z zl(|o5(+@olaC@W&xC2~t#_${~UK+j+_vtdvRe~~49k@Z*Gwy$26+$dE7z-s1lI7L( zX+;CU(mgjrQ9v<@3~_O?#4nbCY>M%O%L=@Qyum#B4rvV1d!>qQg=KKCJO@fIj{OB0 zjn4n0FU#mZVa(${gl(_!F`_K`_cLNhIy;m(FfBuGg0dYH171`<6;UVdkk?2WB-K2u zAJ^g8zh*GR@6{V`wtI2Ubzm~lQ~Vz+;PJ?3rm5*QA4MHl3;fRC)cAw|s? zaC91BZ~iRi-$T{ReC2S!0nhWZv`1i%qEQ+rDwAyk)l5SS(LQ56Y9uB)?4!D{%OSg2 zHC~r>=XIl5shIWWm88&tGxU#`7>a!q1B9+UR7ARe4@3nkAwtz7zaceLoCS3T%tVRu z@AT@B?+@7RjV{wVVWqDYgUc-v+o`JA&7gNy@86`?W_YJ`B)Cc1Js z%Wa1lBvP}*F?zoG^;2Z61EVM31U;z~ zvx$OHh$s)lx;dA9D%_qN8LH-By!kG{$~ifL{jWZkN;$?6WQh(NjJtJRutEp5>Z3DC zi4h&bKH+aj0y~$*ab133`HWu1XHX1K$x^6@)CGq}7V@{kZn=(ZU!jh38$B#t1^eG| zqU?pM`9S*N)CSj3 zpl_s@B#Is}2ab<=xc(Isk zmQHU~>{At!Y_dyslJo%8$Ivtl_AXC)K5|Ni!rP6~PDu~F^wm?oBXWs7(j2;uoc66F zt-jhil1JCcy6Jd#Ef%b#rr;TO5(0`?*SJI|5oj3_il1Lu!r8Ycz_=%xaO%Hl$`bjbCwte zTdV;|zQh1kK)ogv|^7rn&yY#D{eXZub&%Uzu%`7VYt*SRyzqu5u-qh0% z=jJX-Sd^<;zv!Uku(COPl^|i!5PevgD9u<@;FqS{!G(HVfKRBm!`&n9ht%F%p4khF z-7gC6&UomV0+pIBaIVg~LN=>w7w2+^Kn&?LQ|Z>m*Le&Bcxxf3oG;NkZb#;X}+-`RRmRotLZ_Bb|VK| zCh(1LK17%6+MHkXD{Q_)Z%foz#sK>}mQnMKS9kPIW^wPpBq9ro`+e$8AZVblD>CM^ z>)alQjpVq;GUmqmJjltZ;p|hL5GQk2E%?XfAm-0|XPUUC9P4q$Q$QjK9XB76|TH_N|8I=$F)~7O)R?xD1MBdsT=YmPgPMTq5gURkHQk&xQ2rd@@yXFgc|KUN z8X-@-k%iBMF% zL0TwS(x=|R-6%>U$2iG8Nj}R+9OnvegU|CBY8(NU!_9KgCf?k#modhwhKl@I&t3DC zwuQy({Y8UeyJwdwGI*EddY6394H%dW-Q}N4Z+A;y@ZmQ4@>ezo#B1M=0d2E_(# zke!D^mZhj!xPNXvoguyoVo-*Tnl*d;a@6bn4?5TKE)bOC#jk8t0G^*xZBaFA4DB+U zp%5qCti%jH)MCwd?+mTv=O{4V->hlklz9%hCGhYm{4=8byF$~{`xkGc@o#C?6hg*c zSLnb!v^gPQLVL7HxzR0#YvN-7M#?)a~s;R#pN;UUk95xha!7D-XU>b-s zLd+Kh)F6)(_*JR$YqPaTpn5(VP)~5e0!%?c}1r);YFjd?ECOXBZZNGU^w#vb{nppk@BXW-)C zhL=v*A=i7Qi?mlrKIeht1VrZSg`1}bjwOb#Tm}skNC&T`hh*q)VIZzRU|92L2f`}m1BIS<2zI6r9FLs3;N~D1;I0nG zDBEepuq-iLdj}4WI3u5{Lw?Z@%fUTv{+JG*t!|KxDK~|8!t^c=J#%~?05=$_ zq(k+ceu7nbo}VFk{Nk-dz~ko6?YCtFjIoxQN^gJXbMx}WFVwE3g<8vPKwPL4R?``r z_TWdsJ#@RUo?H~HkyKDIVHIvb_w(qG#lU`H283#38B{hIT2RVq=d21C2-~d2HJ^PE zWhHvXvJiGfl4bP7?#Zm>87Hb)oB!o7SuqWY(~oAsZJ?O-6iJ{W?lJnH3aZEd950EB zEARdMt%?$lyX3a(Awh>M<#`lD8~u{3f7hd*vj4H=cRj{8)p29tvgB7Dnole*iFa#Z z8ZzYaRnpqc|5~#cL$;Ol75P%$UVqE}g5A($Ex7IW2UzI%pZebX&O)QPX!?G14ykb9 z2<|x(J9dO(4pHO)6@f|^NurARNumL90;d67#{$tN?#_@Npz`+9Ly&rg8Y9RUh!j-4 zl1JBYS|mut!{@N-`N4o}P>NmU(!^;M?hVfwX}vz^;K;5aW)~Rp7K&XkXQ4P0J=Vy; zJo+9adG3(az}hKUe)L<-izhP?S&NRb=9zJ?Sr~b1F0(RyQ=R^LcvW~O|5dFo2@I)V_^ezp~Sx8QShzcn3 z=v}hjY>}MsI2fiyd9*efRPOmYZpL}D&yn>GJYj*(>?jq9EQ*0zaa*Z~M$Q>99c}y$ z@sOML0_lJhYMlt#aFKyG#oJ&;DfT$P&70K~9LMR9cS9y+gPY#v1lbJ?aa&rDF-K&g zC>gUe8(w6Ek|`hj>yp2`W<-gTNL@~H*rhZ%t{V>y&uC2Tp_o#N?0_t6&?o(W?7azG zQ|Y-s?h#K&-WakG%sB=|Ac7!{EQS%Z(do3)>2iD9-rL^Wd)wdKw$lFU?Q*BnblOfk z#eD@q0TmUZvdF3|qNpGsh~o-~isFhOI1Y;7f~fF+9uky9;^aWWFU9%lE4nT>9QM+23z6e|@OD719iq;z#*)G^%~w;BVqb3Rd|yXw&Fr!L{ncq(uCg zd#5dOYtUu~R|=~5cx|2cQeL6YW_G8mZGksjlnvqR*brmm z#?9F~A=f<2>#)m_g^=z*gDmnbpsW()h*0euyG$D8dgpu|bjy~~?TQ`-3og4!k$a~i zQLf8iQfUx=lO)Q^eDi%zf>*mKyv4OY4&2Nb=o^n~qJptM|Kp_s-GeE(oMp>(719f! z7gZc`YdTiUKt~cO8xq&vZ;mlvk4CkW?DldNalN zbdfuHTphsR#qT7$G5y&s?*jGIU5bO!SF7@D2%RC&Wc?{~+;KBd?5-_Sl|}=V{Kjv7 zPvVElCpd5gRe^~aN~PEx6iKEc9zlRELp4C`k)+-uY}I4}RT@@*pZ91{UV&m7LuXTx z`mS8>)jg>|xprcOUxsQgP%7Eh$e7O%=fawoV0d*Xo1U?Zoq~dH&tDpG^1H@e?&Ri| zW+7W-(u~(du^kj?ry>d#t0rcLt_s7F8L2 z?h-?C`Ja*vfm^*Bwby+f@C#LmBA~lgbU_Q>O=T*b_?!f2cc)8w}=7In#dwOz_97}HYq4*k2mif#w7Wv^g)OV$;Uvk`UG zZA{I^B>r*)oTAyYtbI%A!7oybePJVXiz8MEcUxV4& ziGF1!?0mwq;TbBn{r93Lx7>}e_{o`Z=SbohTeS9@KqZr6p)Mi?I#7V@@$keX)jnQ= zFiE|`TZc`yjq)@7ROjd{0ShFCdJep4D3s_=XJ1!k%XfIA80NBRoBiV_8l);;{f^sa z2CW6q%O%U~$Tq*mE2CHjkVopzLfPT@r=n^1{5V)!6r0#N!XCebC1=;)_u*Azqfx0A~i4MfA6_6HOon7Dw;pXXa-nuqL26Ralxueg76qS8W zc;SF1Cj{Ri9=w_1+bbJ6njt;&C_8X%Fx#~{5iDl6E)t&eZn~nKkS=18Nl&B z5}f`?fDs;T-@TJhG8`Bl&}}`)wJoODB8n6sJmUL_g<}K;p>3?yL$R_92*hBmo^CZz zMr`suFHeQ)ymS}cHs51@N$M+s_e1gjbkBVX=&A;TfDPQ~An}b7FQ5bjZQDS94lT7u zpwl8#Y|CIXckGxL1M?Td#)|oaoUk*l`l}m6LL+wmsyqK7S?R#oNi)IDCW?iI_w`i7 zVsag5^Xs*J-@OLaQq<)4~FXb_-!2E1?Gb=9{1a2^&)B>F;k zy-2r`S3tkGekHJAHbiwH@Nxi(a}#~<58QE!sSqIDIj&um1Z#0F8Z(~#<6*|jcaI#L><=oseIgv{^s{iN()y1&|{sFseBLu52S91NJ9O_u5K zNx$;iVxD5)0*tYjd?<82rbG*bRR6dgLZ&rc_Ox`luF$!uUykZPaGz7A6zxkMZ}sdd z{~Lj5q8jC`AxELfz`nTQfD^mT%gAw_+{n~^F4Nx5+Uk1p_q@ozwqXhVl^=A6e%?SeFk201ID3@-!Aeodb$HNR1OOa3zSp( z;46}-VBMX$2}1ppS_R5n)=54>-m;rP)euU?3}U7j_%g49PQ})#)xMe14iE`L#fv0$ zq8uW1Si^@0RsLBr@L4emjl5K8bPd!m*o1lf#q8LUsBBsYU{O*v7%#JBEmP2|teAR5@-%4C zq+DekRA<-o@|YVMbX2j4ZMkETTxJWhl4qd2_K=af2HS`Fp6m!XFH z!K`FygBDkkNbz#-o2oqKxMsO`o8n^VvpJ2EF+hqe;mq+tXdnDP7g31PXu?arR2jPU5 z@xT6f$6I5aYU=J!uE{!7$LJ(J5&=Yt%Yiu$*z{zdG1ovp@($A?j}aj2!hYv6s6oJc z&>rb|mBB~-u-{;ts)s4(*~W^P&xYakh_EvBkTyBX2`kipte)R%F382jb#dU$j|KUz zL!z_NRPkCq<`dTn*SKDw6TEUI1)-&Z&)qM&_g7Rl%5hV)R%p}k$?GPUGnQUAm<_Kz zJ@|_^WCOm-Ki;V)>%KHwV++K)O`;w;Wp23!x57vZwGA5fA$Nmr30Bx*F-iw%RThWi zt5Fh;#0lk$5b$4Y2n35mt%>mYy^|3k*-w7`F^P0w-z3SzH;JLxD2lA6A`Z`v78FPp z!)ycX-1Xo*)fM4rZd_h|wpLB@%iBloh|`iX?%_Avx0*||jVX#K3$ktLo@b=(Gg~!H zqV;a)6neoJ-rnMvmeF$QpT0ldoY&G} zGd32ODwDkPr+2%2?$XR`WGjK}6kcYEHxu24K+N^F$XAkNX|J-!#h$LG=||522Z#pK zhwb>vG7g4_p^ba6>H1GSj0;K2+@Nb@3pZE8fn#XJCd*D1#X?Xmm5OMgi^H)M8A%DT zhB{Sx17h-B42IE=(Iwe0%C!(!IZ|EE^(!28%MW{{(^qAkGYfgOyy(F4xyZnaf9!`f zj@U9-o0osv4P6UmO6SQ&wD|nCWC7Xaz-~v8$^7r4SXf|ofZG9k4A4V#!;+CCHI_g_ z3q^5wzmOGnTl61bMRf|aVYI1Q+zTLTQs;76qN{N!7hIQY_Qz-JIQt`m?JFuP3*%yx=}jm{S6W@IA8 zyXVA#fJ`aUVZh_ufsH$4c)iur@rM-aK^= z|1uA2#P^aiQ9Uo7dG_NZb&2rkx9Y!<_AXSLL#tK#^yt7mRmIf4plCs#Q+MbsmsN9; z)F}kN;n0(ss;X6Bzg>mj!*{RFkMqtH>Yo1+DZrKqmIU-rp$_{K%Zv$gAN>rjOxgwG`8Y zxKw5M*w2%FiIic1_vPbuJF;wihS)DreQApM=0w~UxUF;e3*r_6|!TlLSnbJ_;&LuD6pFQmk4yQWf*E(DTm0iLM=4kkRC6;`827us=VhBYBC42a$4 zgIhzp<94sy%yw-X>&=1{noi9<_Z-qpYF*FJND`L&2;zdJnq zaKY^G-eGu8aWf*#Z)>KQcjj2&(>eiZkglK}motiNK^fF1mqTRYDG%EI4Ww!M;9E?I zKo>1VL2bjSW3nm%co;{WTjZ_aUD(lp3@<%0U_plWL0k6}r|EFq9mZc;a+KyF{Vz$D z*g~)$gD(rqluJSBJ5L1)R|kQrqKQ6AdX;@nr~T^u$^}QXYG0B zd+&VryXIw&r%d7$FRBYWJWgZTSICMtvczfhEp0O0>(s@M47emdEykt3cWNE|Xl8+= zNO)R&OS@>&vI(cfcU5Knx!%hsv?)3v0+u7pB?(?h>b3k_308AdLp4xz;BI~lKS$UE zXX1cvIU%@*idW%&hpEKV9Q}UiwF&YE^9G7#G`gSWf{Yg3qPG{ zF0lG4EF5@UfgYSe6|yg+;CHYsQ*QAWlP+Mez}op6id6bA$&@zAi`?%A*v4X6KV{b9 z!fr_(3%{ymT%?`OxNn~4aoC`lg`7^Z^!Chk%;ov3zVY&?UVBc9<w#wJsI25S>jG}xK} zJ;I0l+v#FDjn-lA0V)WD>GJ;{9>4zZ{Xazf>^JX6{P}0U{h!}N$Y|DyP}C6c7)O1! z>!GhP4Osp5`eKqYR2tB6{UuUival3VEU13uQW2eyBho|asMqNUyItDr6z2kZd%{>Y zfnUnsrn>CkuBdctqdz5iDv*+&eLSpwb}GGkW+J%Z?a;XMkZu!{0%=Nw2!xW za<9|p{2FDlZ-r}CEn+ADN3yaoGVoFrW^b04k4Sb19NV zMJ$_ePke#S5kBWHpMa&%k7jmGy*(peRUso7Ricz22Ll6V1@TQ#}H2)bnPOoD3z3%^E!YYiP63hE!T)-WxiN^;7zM_frd z!0n$>mTMg+P%JQh#(O(5q?zV6i&TS<{}D&i?GOI)kS z<|XoqVPdX^^)UD8b^J5rn6h4*M(NV2bJ|q52DU}s#ZT4XXZi7sZ`I6OEV%6dXyz^L zZhjg4-&?8w)|vi=Vk|}BX(7jyAe6vf;N1w?rh*D4?MX>CuUxRpwVJMfzv27Q)F%zv zCA^)aHn2~E#8{>DG0vR%HZcN&mKKYL91R6G=M~s^iSWaSU;Q@XcYl2U*9$(BEvMKe z6p0+|4cIIH$g6+;y6@XYBt7GA{|PA_V{X%B6WE`o*b@{vMn!Cau2a4LzBvOwp@88# zaa7v34ZE>Ykf~qsMkl@5zXtZ}kC-jN4cbIGGAiWF%cYb3%IH>zL0*@2%Hq5`X}#)! z`^G@LHyXZ189p6Kytl_?9lzG&IT*nNmqL*2E9X5IM2hPI?@UhA?DL5c>s8^(G zc7(QwhTU9XfAqhC2&+z5M!%-#V9*33ptLO?=*S`mUL2E6z!OWc(G*z&45pgfGvg;> zq1G9GF|e7|eYO4Vq7f;xTJ@n0ag$Z^Kg4tPK4p3ShuE8qe>&k85ne`2Ed9V^ElD2A z$T)D&ufzl=yD2uCA{kT!3ZEsZmqGW?j_`Qqus_y)>mGvS5E80iAcy^RRsPv@%=F^$ zt065@wv%S{f(s#6gbVuHRHFme38E$eEl=dch1pcCB1MxvT_23j&`Qe28-J@%WN~2} ze8IN3vW%h)A8ea$p0ak-B^0_r28V<}j~ zQ&nHprNP00c~xJv@kGCNdN`rrwIkW~qGc2eVN1q6{Lide=3Qv7CJ-DruVkS!ZKrD< z@209zerc|*l&=q7>A%ViI0CgNSQPhNGX0{Y$UQ1pA3Tax3!LyWid74i@$%ps&By=U zxXvWq0b=iS6G)fIm$QXpn<#P(+P6ftqEskcEegwVYgU#EN`ysWncmyH_5N3Z-zFY7 zp;PHx`AzWds^mN5cPHOeHmf@Yi^W%Ex7{0{2o}A;CXwNFwx(0?^65C2&6))MRVl{! z6Fef{NTw@6PvM***Sj>ZMbx0Z@6t(Qfo6#?9axr*YtG9%1)KeQojL`2)eUAV)FUjP zbySlIWpb6!qPp3CgL579fb@kV``lo5fVd&19>2(JaGbiq7Jg*MUtk$5Pu&n-b;k{@ zh5x-_HD$!vbJypmNv{*9f0+Y&I$KS?F3}Xbh9WCbWN5J%`4Sc`T!@lC82o|5FM{D0 zpn>~kYjMqeNu1y5^#68EFA z?%1C9TrA$|oJ1axay2r8mC>7cSV64kW%CM@+vM$|Tg=UcUF8dA$F`R_tqIhfe?R*T z^U4^9-SSwdg83^d!^Jb@eqGGReNLB2GjjlHDjP#mKzwees(kL~E*YF`fNiJ#wWD#u z#klXVuCGXZb6Hy)*x9hqtN;)SZl$65AfBmJ^thG+cv_$Z0W0s8`FG0TVo*UyD_!Bg zcFwuLTl^)yx(wbXep+ya|5kX;&b*1k^dj#=|I6kDmtPW`SSZEBUfgD; z&#!?bsheDD`5OYOWEgpSx~sdG z7TDY)AxE*aWy%5iL1^LB9=C<5ni?8OTOgI~cQ#CiqhXXTQ<_erT{%H=LUh}`nnX#! zPg$#|ga-S=f&-doZaE?CiaI}CtVc-*CbWB%v0>Z+W-UU@tp>-}oYBN>IojzyUvFN9 z_iFsjffpSMTB(QBg`OC?PthFVH>+>XTLBxJ6b-V;ZFPPclqtjjJ(lei%;96gLKalVyC7Cr4X#S0xQl^?%38M=K~9K)C?ZqZu85w5 zqz*upeAl-WD#AC341WyNPp|&X-KNZTgB1EoIbYhwrv^)I?#7+E!(2|h;B@Y)odxqUch7-=y zZ}kITsYv^cdDz54+HI@zqBpRDv7cw9*8PZkJRc0S?xXV5`` z3-SCCHQI-5DqZ)a0;s=3tT)OtWFMj9ie7SovPslJ=dh{tM>YJ5k_=uqNvE4c;1q}A z@E;+4Q9Lt1UZhK(zCu{xSF6CfP1N=N=rfncP{dg>h3itp3FxmbFxH>p2!BpMANTZO z<|T7oIxZ`c<7Ta(8gy`-LIK6@p-2uD(M@uMi^aWD+hj}UtmC7o_qF-;{i5Ct+61q2 znqH?I;q93|9C8++taqTOEx@@%(nzGWq_1 zXR=3jAn>;l#yBGAi*H|DaiL(W6KS{f^ORy~?_)44_U0yi&1Y)OB){fyyItaDNZS7I zUX9!sc#+(f{EozMqr`P!LjsK?gL+`~6bq_`iBv>_N2_#?A1*`jOv>El(AmG1c?eKS z@W^qwOm6Vf=t5<;Y?o_xXa;XRtZd2jSryksW>}MJ4WP`IpJBZb$A^+zx8E|~EIVw< z$HMlviO!VPxz+ih`+(fpDY6!_QmDHw%cf8M*YdYduz9}e{hdM>pEKR>-r}F5H-=$> z+#X@8Mi-}eU@rl?;Q+uC5FBUL%GDO{B<1 zDgt%J8nnm5+Uec%z>jr7<*{an-j<4Ll$q0xDGNfInHcdgB{r56NJhn$z|3++la*)~ z>21WR?mBJ6$D<#3pCF4JIMlBe00z2aM!7tN1K(7nc zXnjr?KErIxP1V{tH$b92gK4L+<{s06k#n|tT@Tj%J#r4VH6lH385AS*IU!q16OA#U zJ}1l&AwfItcROZvfLI;oiTa$1Wp@~i)TK{<=F;H46gJ0cBPnXJ4oapgj`gpGk2_9K z8b|Ve&W!a6l?uOHNh&zWi&BAKRY{zCF%>4lYML14yHl7o)m{ygF^f28NE=;jenN}b_yB;GkltvV)_DmU4|{w7hDr& zH394NViE@O^%%!=J;!~p2Hj$aJ9z)h!6kD{Va>5YxQT)+~ z=DrJ;u$$uss6n!Q@NW5qr}IK2|In;a2ZnX97{vxJBW5_kg8J=S zv@6f(g#7-~N5{w#2X;cXn=A<%C^m*7QB(vL(8n`PqGc17PXHQ0&sJzrs|iff6o-#G zD6BLVwr~`zF%d8C;f4yftVuY{hzcjEW*QE43Gw4&HTbiWlRdxj-Bju=~gxmL@>kJS7es0+KTrPI{ zq0w5Lm>2jxQa2R8i34v@J4`G{6UAPm$Ym-b(f2;kl_KMtu05zleuFvhv1f9UI-Ysx zQsN#R*s0h`mC|SYv92yQ^sGu(4=KJ(aW!w1LiY&j8CNMP1)X#TMB5&SK)Eargbopb zn|Q@Qu+*l|2ZK5)RO)Y(q8>?+aJ!~NSTE04!5d7ifdoLu8}|jN^fD47xB&r^2T~mh zp``^!LA6toX1RLd!bW)~yhAjv8y2TzsXkbT(hl|NMBfznt+T1uZlt*Z&x+-T(=C>P zX&HEv*YEpmtP26Hho0dEZ5fFFMDsdSR{{%^mHZmTbDwrP!K0F2BE0HT$sbi$mkSIm zPnzwoScbt6#hFyB{I})iYmi$sGTVRF7DnVW1&(Xm7e4m=aJq4A`s>HdhsZ8&J0J%xLpo}*YVD_3;3~+cB9dH} zO~S(0Qoq$KR;F~(JwT;{Br6Zx^XLizmJ#6cw97Zwd#7vHq%M9*;C`?Dz&m=Be)^T8 z{W_1;^rASYnz#UZ>;7esdrDR73o^&$T=9SJLHd5 zk+~t@+hFo|^VC*NJcz%?Gg#N8TPxHD*NFDY_=_C#MJ6QXWqPgP+$pO6pf3;>_g+F>w zI(#cUz|=uP@Ls>y`|Y;G*lPhe)Wz4JFN&~F$9mYVi@7FyCcD9)R$jRq)+b%_X$wJ* zbo5ryfsfeAc^N~cLTRWkytU^nzxivf zaDvjfWag6wZ{s4C-S=-rL_dK*%70M6Y!AhPpi3qdapcXz{`-Netp$AKPQ?M=o&0R* zm?;%SvGp*AAkm>0oK-zk=JvbJ-SVrlPE`p|;wRJBWZV4uyZMaw;gS%~oF;|pZb77n z?TvtqX67}Y@$fEjni2;|%hlhq3wD z?4GZk80&GER80%`um?hP%}f=9=yE4rntLZ;OK_>=Cae}%8i$>TqczbuL!2ht=2L2U zKyDT-?B+(w)o3_=euXR`u@fLAH+ZetPO(W8*+NB>ftW>yszy{J>QE(1OC@E>Wz*Xf zk9g=WG`OD&-{jjy9?h%w{M@$gEr|D)7{ymfdFmowD@) zqJsu@xO{oc6H)=U-_f_1c(sisp5GyYsu}v<=#G z3Vpyn-daIw@W7SS_p9FB{_~=F4cgLo{uBw~SG`V10iPpFnLBVx{(F%>>ibdu4U773 zXi={4-z$6QGH}Djx1avU^LO`sJMHcMn=0YWQRw_A*$(9J`haDcWt!N~+(`qspZu>i zKR7qP=%Jk4oX^o!1R|Es6A7H=ejh*9J5I|) z)ijp5wSczZM~kU-e5l90BS;ACPX(7lh1+PZFPs8x19LVSKRMxIT#g^Iu?vSbLm_EQm19|GFy z*oqGNUUY?Di2(Y=43#T2&Y9vpVNg0b>9$gLQk^(yKe#QafnBgWu9I{?`I;_Px-?*o zYn5iZrVkYB%IHVD;*eX@m-5y@gL6S(mRh%y6o&wrzIXogRJxpp7dln-?l$mnt%BPZ z`LnNHA2AgF@AFBQ%{l)ZcB^B7#Scr0K2vomYn1mwp*pckc}o_hg{sFmUK}*1#(77I zx2U6n?*hTLv@zNf?;j%Dzce=H zu!)VyqgXNQ}I*&@z|TDFTI0Ea3Y@l3RIH~-G8J-kJp$6b^9h1!9YNxXxFayVby zV<%3!7TDuj>t@(CmC3jBSA~@+m$I>;m|;BNJN%lq*H7l~5N2jaUjG)uy84IliGMSv zJg~ql(CJ%9XY-H-I-aSKO{YrA+O9>tws4@JMqdOb~y2imA}t6rwe^GYvZ_On6P#Xj=#Sk1#N+bhh@{224KZ4c0w&P$2j_#x+IVe(m zcyejr-wk}mNowSmM7oUQVNsJJ=i6}+fjj|8Lvuo~=L(PP2nP5rj{pxt zSpu7#vusjgrk+nQ-~V#)NF3MKA{KVbxNRxrpCIeJF)f(vxqfQUz++x($KdLbgM5Uf{Q)2{o6=z6|B307>oKAgCh90BM{$)_ef6vCw z?(nclFZ}b8g}<3;#LfpL+H#WZz#MNNT{@`A>=4BkQ=|x^A(+XjRR9kU zP;!A(X(_Z8FAp}XD7rFvuM&NcVrUs`;_uLI;sYy6i0&c?GJ#B~CR2lI3E;29fyqIS zrHj#}Xi&Y$cw=kGBEbRQJ|~Rm8isr=3yQT9U}HcS!C~#YTfWcoN;Fh8!SmX0=KU{o zcNuA`l5-wqD`=e0YOc#V6@C9NXy4`HOfLe(j!vZ-_@$q!;*6@JU1krJ|t`r zL5Winy;g|r#Zb6}d<0m?-@rl%OOhr*wStX9^+zcBdT;GV}{cp|iOdjjnJ*9Ls z4?8plx-&9io72o32RW5y-WBL<#ug4_$B$>QJpU9~$~J&nHYA1cB{J87wi7T+Fabm5#ozvPgu9&WH8SCW+TbRzNiK540Kwsd!eAX2S5LqiqMwhS4I5}2f z!rFsp7Y8CNJPDqSk2QO8Xs)?4O zeOcz8GaHkG!16~9X}*BY&n9zd6{-fT%HW?id3Vc$bP3-MGp}}Z*s1~x^@~?!y~i-lo`F6uC)7?2+76ttaOt z?UF~b!N>L5O0w4nuk3*F=>?3@-7_&`ZhBUbFEW{7or3d{Q|f4Gl6qj|HpM1+9gXkRpsoGdHCgVwn@sM!hsx-{#e!z1tvG$YObz8^*2ad-yn0yJ2Q!Iq*D&b22O2^p%vIgy?x%b&sAXxfrP8a{2 z0tp(LNT(pfzfD!*6)D~?s}E_=f{eB}hFs#;@asb|cz?O6`WrMLQh(Y#uh(hOpdaV7 zpkt@Z3aci`vQ#?NuiZ6P+@P)JUH2)R+73dQ(Y#h=arkmU8hryogGuTd(Sfl0yfc2U z!?I;KjRMLozxivfa6@_HN&Y-PBbcTq|09;9jWH7CN)vz;Q7kw?d$1!uRg)ds0$g7! z-sq&0yldorPRVqZ7>|+zI%aLe-E<5T_vZ*RrRmetf`M&0oj&NdR1q6i&TD3nH=#AC zjaTc8bpA=MWkBCu;CU9<{2S$4Cq>P$@y^SZV0sNUIA5|1rvHAH{KK)%)qfYX3m1!l z*saJtdEz!Nlr|imE#UmQnm%C7?3+F@IVJBq&SuUuPyKX#O*$M zW*Mob3uUMuvrk^-52+>)rCa2CkFM~yE3B-6k0n>io>%0wr*K3iwpdzd!uvT3XSR*NvEm%+nFhb`Csoyez zlf6b%sH}8HihjF72d96y@UbU!aO{iO70Wz~AZeK!bd79rU_TeqjDzxySrogABB@kF zYXB%qk-PKv_!ZDkNonADd4;rq?seKhAEjHBPa#3npuIs7dD(Nf^B)LWHRnk;0AXKv zitP5x?71rmc7W`V;}!eCV7T*dG%0q2fo1bC#Og!+LNR@Ph!HIk=XpEH9tTFtX%n=R zQEUlC4p0#{m~Kcpm(pcn+x)V^_VZ4vmju*@9FpMqyNZ32QeeMSAEFlkqsgSlq|o`8 z5>+=K9l&4d|4^}yzl+B6iM|KsU4SIPQ+|ef94|xmaVcHs)5CAjX7V4=xv)Rl5B@E_ zMUr|4IT>aPV2uz$LoZG?02Vh0h5M1*vEH(JUG+GuAoPx44Wtj#=|^B-GJQ%V$LK9{ zY%$N=ro>8Aa2wPHcXLCfT)4hmN4!(LYcW1%3t2L84?#;ky`DGK=)ykwiDEg~!EIOUz~-gQ#No}SSg44~20mtB=|LTZN8~Yh zDoNh;f!PAxT}36qlbEU8OKwA`6$ziP=O{^CCg=i6+tVKXN_SXl83!d)eNLFsMja5; z3n&rpR-S&-2JP;q^Tgfeys`)xPM8im;#UXX?wV?@wBWF5Y6}t&yC)@i_c|rYp;{YN zIpT#W(*EfD-?Dt3y`?)rDg`}E8&G+*g(NC;+cc*&wa$A~y1i~u@Z5?w4*7JGQvQnI z#m+5Lvc!9px=TwM@D5-#M);Zm94vQ}OLXv%Y<`je$yTpdZhOQOd z`M)a>rq2Hz)ObQqNwgrI zx$4mYiZxaI4pjqW4@v{8_)DSAvzd7|C+E#xr+BY+-(IIZ-k-i%B0TAXBV3t(O@>cz z^(&sI%X?kxeS$3J=Jq)723v1}%k>nyjv{NR zh%P2oRS}*mi3I-Jl&QMSzFTIl&=yVK{kplf{ArKqij@!=(R&Q%;SV&=StlbXvY-6= zV-oqL0R>2r3`!NmP;3-MR#Op&{U3?3EF9x5DH>eAlGOHIs)n2yYalV?gLXdo_e~SN z^1n+UYO(iGPmtMm-RpM=8eD6fJ7}bs%+%cWyDrNHDvr~#dy+NGjVY~fjvi2~LC2tV z#Lk#_Iegx{8Q=ef(O~2VXH6s5zcdD;*Ti6SP;5IzJ{#z1L~f^kXZJERj9i9|^I@-t zLFx1r$;w$tz!=sq4g)lBn>8_l*w750BIORYUm&KR`H8uxe)tk^ za6||F=#!CjHac)n&jOXlb+2Znp1ut*sgh%x^k;82gUCz+fCdxUsFt~fX^~$H*rKl2 z8u$#GnVqhUY?Ar}sgj=4+SyDEA1EgCYV#MJUuk|%WVB85FKj$QR&%pW4jkBnGKWFV zOd`c@q(~e}Iwz^Itrgo_7mG7w?eqcqL1?Gq0lyu(upw;#N_TZrnq2J+h`}?-4IG1? zwC{0#A8|*=s!r79+v33 zm?$=#{!EhSn(De2)ZP0vb6e;-(3IKb8s&P4G^_C<>VHG;Gu&ck_-**XbDRBF*9{x~ zDkej5+A5E`{XzIRb4eJ7joev~e@WDyVDGr0tD#Hwj_#LUj0#3t4q%YO2Iq3m&Hj13 z)pGk`!DIs91y^jCKX0$soUk!IB)*FL-e^bur}pR5$>$Tu>hXiE^D~NlLXk&Q#IBGE z!8W%?Gf|H8DZi6oDi@m0GkDeXNlE&2y>L-*twPVs=AD+EmaTQ`ap~ei4ZQkjNO$Pz zH=9Hcg}Qo0h5yUju*z+%+eThDBym>HE>Jd!wtF>+9?jhD1*2ln6}g>ojFnWQe7oB& zl}^tq2u%lxjy|Va#SNy(d%tUvyktr=?=jJ(1ZRnt2-b(z(RYCnLAP>RlDrlQvu`lB zg7%OU&0b}ae1q3#Z$fZ4a#orA8YoAX^KJy8hO|DonfIyl0og8Ba7KYs3f?@OHvDti zc)#R2%dgr1`<&r4%Z}>_On;lbFv|$PpP!wbL-Hn&GbXD=ImJT$=O9Y?to&At;2I=| z4MIM8UOMm5%sT;%YzB0iwM*`Wlz3%@b@5w8i>Yl8G_6rogQ|k*Rb)U;rIEPy09%k+?1CX1m=8`Ea@=jl&iG}4<_T~M znd8MQHiuo4bBn$^vsO(T6`)0JrTSJmG)$Rkr_n|)6Ku&}F&n@yKO(JkT<8d_aGNhr_x}B7i$yYdUnzUEh z0TBfqQldojQb;;|nIJ!FpHn(0oW#RsAcrJ@7IZ0nO_mOQP|>_Tr*hu4z)HbBemB&P z-giM(o2N9Uqij&JZZFJP4K52eN;myT_m&Ywjqlw~BZU)4oyo^vNwMV=DWxJ(rWbl< z`RTTk^@8&r1q=pd;+YcR2Crk-4m3c7brm97`j_KkTdOnz%q|T6~gVdW1)+a=O zJ#3^ckoE4R9f!<~i^FcrEf^ATIDiYh+hx5stdM`ZCd0n!WN8aiP5PYH3O^+e=G+a! ze3y-;(wdK*>vUT)rY-KVY)FP^QyBkC?`nFi2ljSPg3z;~2W07E`AObI?w^uQ*%SG) zIT@P7NqD}SvakxTcq~4IV@ZkiJaXD1mf_+!vJ5n|WA@*XO5)YlclulW6yxE@7(7xH)$>)3@C$vz_PXk-bYsz1BWgXWQCl)Hlv3TNnhy2 zBu*^|TgT6w2K$-d9{yg~cORY{=Y=(R5BbNG_3t-)Kbq>8S54p28jd&2-%TF-tx{~B zx>f+R;?Pdp375kWfy**#hVWLX#GbD5|287&<8>?FByB^n-#PGxXSqqOP&dUwQ(Y$& zVOVamsZVF?;;1|G;71xBFT%gNzW(X#W(M_>*27Dz=*bj6;$hWaD|By{qIeE;V8(ez zvD>^lJkBUUbSY1|+3UPpGmuY12WP-7Mu&wu=iM@;m}rO#FAaReJ1#iyc2{wGCTy#4 zhz8KqZ4H8Yq|gmvx+W3QRikoXk|tGC14;dDDoi*Woo}F`(51nNbYSd+w~7`-^G*dE zogW8|d3*rA8N6MsB2{zLGo9Y!sjKkop;P9TYdV2gKPp%sjBm2SZ4A~4>GpN}!Q2bt zBAnvD!5a&-YAKp#CK>9(qM^S4>(uJi1_|pf{&ue<^+|R2q+Vr%cC8TWKDK$aOQNNf z;#>*dyVk8Xu#|U3ktA=CW5a{}*2xi?wGo-F*F7x%b4TDs07-~^-V2>FXTnDu$Y^2yYio_aRraJ*! zf|t0qfQ)rI-2*A1Xr2z${Tk)Xuwsp3um-1-i>(;Z^mBaJGFrY7IPYJ^I#*sYT{5M| zrN|u!tv1Q;K+{}lAo?vB$F<+GU||NVMSz9J?RMHS0)}vc$EDqFyfN10xlrvsNu3tF znIwvElT#u@L9SHDf2M=GiD~A&#NgN-^<1hbgQU@7a-Ux3{RsHaR*3H~{Y?yUP@Qh(Hot<`TNJnat999rI>i02^715A~8^dIroBOF=-M# z08xSHz#jfx-)ed%FTv}%>@0PLtYb#+i{C%V)&Rq>hKw2}48IqbWq^uUnfGLd5g7At zWUM8b4h)QACcrpAvCvanIKbx6&*V_#j_Nh&h8o1lb3oBsJ0` zW$Q9G>rVL@IR*BFDynx0)v;`rU9DWqMn-*c^?*QrBwjD8@m5mk_K z!9l-=zz~B3yyd(*L8AcyYiHfWI@n}hUq1*Z%R$9&uZkS&5?IHTs6C%4UMpO(`w_FKiXXv zV=gD*u-=CS=?E-H*{f;eL9#vw3rkM0mw*DOLw?RZdrE;a9kGCyvBnwyjE-60lxZ={ zwT-9N!f@QynWg2mGtBFzU#(+WxdtKu7%7$;w+#blWSM?-d+EE@3?yPhKJsEF>+ zHrbBJhQI4UED;40>%wb9$=-+FNEeQht#~nMR%77B6Qgrr83RKk=co_dp4|;F;-c-l zck)Tb7)$1ZboiiTelf)sQKW#1IIg@5<t%iJz~!>l+oe zWlkx7n+j`-Yy>>3@igFVeBn>q;LS3gre*Rsn=d#PY`7tEX-Ie67iLpfp1ML%Dyi^~ zV)eng#lVVxNHe+>792ls+kBW-jAHro9^(7!xN*kL`r1kUM#GZdp0}8!JMe;2Wnx73 zQ7lm1=1~zyhn&xr0wG>C;1&8Rpo}^m+IC>0!aJs12P{(ELA&|YG~UzWvSiK4;*$*|-+_@+V}hItiakt`LsUdN$aO@5(Aj}`50y`pyP1DrQPH!v zP|i8i=UCV>_jTTFkbKGX$(S0=OBZ%HXYkHL1waGaB03C}_Ji-)qF!(MkEZ+0)EJ3J~u_!Uz>D8jc@ep)tSMA;BTCIGV@E<@kR4MZtF`eV*m zPYWY~?qYGiZw%38Xtuj_fz?2s^s3Nke+-8-kK%?0r-TDAhMqw?o#h0Kad&U-zbQ3Z zm+hj28zg!Hh|Uf6A(JR}3q>|T7&D;QwV1j<5`2zAy#_FgD3=F6hF0!orYoq&CBw(? z#jBWl5s1#02cK~t@mJ5f4~WA&7{7qljXTo2IL(9O2DNu~Gxg>Gao8}M1?1K$nu4&x z2%Ht@xak+`jguCKr_#+#DKvDgVzz)SS`j|x0=_yV<@@69a{==RAgr=-(cCq>N|C3Dt4 zlbn!Y)46~BVBC<=Ju~R!G~?RhB-QLAsb88r&JmL(axcZ^Q)D+4QLBiS8XD%1Y%56} zDNf_xCP1VG37dFE@0<_4HVf;4_i3BTa2sZMsv*9QEqA5#a@Q-8EgmHysN&iUJSg}_ zFDRxLiIF{}%hnDKCzP2uujVgaj~cF5H+<{K=VQJ53W?L8HF#>WG*y-}FE<1^h;-Q5iqdkK;#%Gzb)l!OOBpN9 zrUw4lZyx$TnBh-`)t~FIZsa%_j=Jvia@_sR>lYpMcO@1o649b0LtL#Lq{uSGiHiIE z459Yz&$WaZE8%MCal4)82G@y)KmK5>qaq!2;gi%z;@*tbLFr!eP0Sx zjQ{47f4%W7qt{w6>Dl|F+JWP(Ehc{G1&XbwNG%ls!eO)?OW$Jz&CK(lhe7(_0+5{2 zA-gpaq@#*W%M=~{xWV+eJfhpkN@1UqVTDWv;@GOP zWu*A(Y$W2USD*JtQr9R8B)TiWgwmzl0$MfDDC}CV?hrMpKLg*kLzvCeVaryupp=hJ z>}jB(E#TdDkMhSiw02ds5?SVM>F{4$g1FgvZuTbdf4Wug7?G7Tm3N<%eQEYXS50tO zL$RkQa-51l!rORekKcI_fUjIoHzi4ZKj4%oQ{3dEyG$TT&vwaTLu=?rvF_5`GC`bI zud*EICyHiw$xcn#?Dx@P@ov{GuI=m)oM8kdLgzUjLMjCIHPa_S%)dt+paj}k&-Jbi_Ljvo z`II@$w#k$BeZk4xP;r+Fo#z|P&8&~kHjxqsUdS3v49F>pJx-BoD&pQ-(b5Hx;#4}p z<3`XT0R(3kl+nFTs|j{z^g7judYw|?b%IB)Q=tt0SN1w}%N8Je?DAQUNrU#GvWto4 zHPKjD-A<=D=KS{-4~)vmaP$L$6Q1n8XuZTJH#`+&R($(^7v(;R`jt$s>a)XpCxn17+N0|VEe=O0J*A_0 z<({DNFejhUH34lUHU>no$oqa;bA-Rj?TovQJIr2AEF9Snb0%-Y`Nf?dE)MyI(JCGN z#pTaQr2|`~CKIbvPqCoQa2l+V;;1-AToRb+^YSR3IZG$-GoXBX8GCC+gZ2Usgwuj+ zL?wZpg6z%;x-5=|o z{L4pwi{`?}Za<(5Wy!3^3peBR;|b+w&HK!_bf`LR0v7s12B+iKQS4fZtfC@n6wiG+ z={>(N^e-NS%9;^c4C|-HK4Fq~B;^2GNWPiWOWGbF(FmTeeR2 znAnmX6q`(utyBbdsN!a36R#MgJX)1!+%XXageq4gNF|&sE#c=%`kXN0n<>79fF`{QIdAOS;Scbw#I`;K6W5hR+ZRfoy8G1gji&;$xO6bo)hx`8PLx4l;+2f~0S zrAB#O))4@0%GkbweEWtNaG9u@Ht5tH0sjOp8@Q-<(Wp>YBWi+}?PBpM^(J5IlLDM( zY?!DRSYi5qa(Y86Z??S4xqt2VTg~h0xU4%4Y*Z{%;bqAV`t`T~Svh@yeyqU%K_nBj zquTrH?+QR)&;Hh9@U(Hj%;3lDew_c)9LC%uaoC{01s|i^r51z^&xPlPd=8c&mA(QQ z%QB6w+_RZ!Qyg{fb6PzynLg^=BF|H%0extTs8tzn>-x7LCyK+Qzcda*j%3@DmJP*F zeQ=*wIBhrI-8d{nEbLc6+5ntPL&qtWXzQg&kMWd76%N!i*fQxzzt@otItRzY+JTk0 zWC|+A!O9Yf?4fox7Q+IH-PVzpVB-XoannwH{afY~q86As4ehE6u}BeIy&B1ez*c39 z^Zx@TBr8ApmmkR}9dY{P9d~)&CHpvHtaBrr);yD?XfUq=z3lPKhQN)}j`_97Oh&wpy(-|nz$iiO7eM^I&eji@Nr-Y2V=dNU|r zbwRKT>Y$>fXZ~%)w{9@|{8kBS_@f9A>%oDGhkXGtcyvzSpkn^I)%)AV-R;L22X2#N z4xHAzWwMiPpjc21tD_>CnML9T?bA1NCHv;Y2yS`iN{Ykv&`2#v2 zo(GM~;I(Fmv0xmpC8>9M_B!qNjGTqMvLMz(?!ATW%MIEN=XfUBJDOJ_OcjH~ts(~; znw_#e5R-3HWWYEt$4IAdsxrj;CO2sH`zFKw zYcO=a=$IR{{*_ws*O4T09GK$CLL%p;%D{<^9&ZW_Y#=U?a~8sXRsPvDiYK>2<-s1+ zC-+@&Thz=Po{bkdA>)f?Xo%~>2GbB@+49u1-eCW6197#T`~FOxJ3%sx2tSPY)o&wy z_s92tz2HOHa*AC-k;qZBo}ravzhx1pC(})8dr#%iznm2% zBNiUC^Jz}zV%(MA_3kj&Tel!oeg^}F<6&X4tyTBORzgg!!ja6641Zdjg= z!jq7o$&f9KbuIO7P=CtX>1y{y;{~#~g7(FOc0S+)+HpVq$M<^2dT&UrqQ|vXfwA^| zGg{;rO3Uy`otp~r*H(?LUQyw%hpyB-Rg(xW_c@`X+a$_Y^*5Bjvo0WiE7eP{F$eu1 ziKD?Mfea+1TV5h83t2b4U#BO9G|~7RigCBov7rULRc=KARsIFaZE{0nbdT#9_YB@q za*WP~_rNq!t)fz?4{j3GtCKYCTmt=Ek!8^eT6MvCIC4VP_#gAw(2>-yIc_+~QvF)K zs)a_o1Xc;-{5f8y-yA)2G-w*F!q1@lN9(Y(@n;{7^>wILaafWo8IW1oA_ke2^yzw+ z0zRHjQe(Yy4!=FbMvjp+Gh@|kzpw<>jA)yCxLJ#@mH+OFx$*{=+^_?0i7n`KGzD!4 zEDAp@gHHX(ImO{we%bO!aWc3#$+YgcvW4y>Eg?~C6Ksr+D|cx7+xqbZ7#DkkiyNX| zK+Eo-yj|~q__lf2WK8j2TL_M%YFc2@OQxWddO|QT;^+IGRAZ1I`5by=SwL%xZGc#M zWT#q^A-?gPK`_J^_BdtPtZa*I$oR+@Ho5e#%Vv=WLzN>ru%TE#esCyd>3DY0c+yKn z)Pp*4k4uH9AY_?8(l(`PQdM`{PmpEqr__4qXwnt}Wcz@P6~QZlQ#6_4%LI@kUG9xf zp;Y=FVXFpf*^mVatK9U#jq)OP=^T_fJ*EDXVD`BxbPM$JY;`+9GN;{u=a&iYxYxmk z3-4bkK0zAgxgJ+!NYv2CB>U9BCJd#;PLSLXBwR)wX7r|$>64JDj`Oa2V~M0hs8?HWk{a&)B0Tdu0^9@30rL^JSPzVka6vZfNVZ?s65+G_uE7wHdaqDsUC1)58 zGDr42v<$Eph+YQ8rcq=k6@l&ZP2Ou$4)^V-Nk z4d~I8`t4w|;YFr2M$k?}8#E&7<3-{=r?qbA7G0W~!*3BGe`zh0GT6_AjTkHTuvajq zWvtvDzvzSaj3#S**DpUORTIcflaK!##nw^e3~=LoYT<#eu zKBwq$&XL}lo#0aFxqL!{_8xryHt|bB(&)}ve+2e7i0K_y7J6O@ESPQ3V9o3C2Jiw>-JtEMkQ<7l)LZ`}pm ztKBZ!yxPI#eJ+R=L`yHqIe8@`o+=LCZ7#EFd8&q}NpjqXpX-W~e;Mn=QikQgz$=fX zw9U*?MUpEp{P%a9qB?gIjZ57!|5Q+wNm66Q?sKLMjz&dVX7~u^f=BN6azgO9r56sp zXTDfju>Y-e6njup;n%{TM$5v5_|vctUbyg}bA=xkszviqvPB05ftRAu2bTsSMGAVc zHO|%aZsqAWGldJAgiUsm@)-Vlpn0d!Jk|XWZtpMR(kP8L=SS zT%<&XL(mwB^X?{SU$(xHg z6Q51785BvQB6?g(Lt2AMgpZ*=AP+<$nnW3@hoTP9;5|XAq&12%0m_{ zCC7mwvDx!EkkMlo>xu~%zIo%_zOSvEU~81Tn7NU{<;6fYJ_x=T^`XBJE(3xyUy}95 zf;>&FqQb9M(GEO!2B%AhKG>(E${%a;)(521nC;Y+Gv&~GcHJ*OK-b}%7Z%UgAzfml z93AFb8XDLNJR6y1)0PJGIW@{*vqviY3YkM&Z{WV+T3 z!Yhg%CR2(Pb2>bHuK3I)QB>sqnajCwT{pyKR=BO2*iA4xgH4KiCtf2RB4mjw4NUe; zrq?m;43Z^op4zG@4NRv?19vM+LUgr?^%6aAk@n;i#1+2HLs?g7nscTY(1kQ~_*Qpj z?E#;@nJJlq(rD#8XrvcB2wf-CotS@p9;!8Nlx9lfyjKdXO;6YwEau>O^|f_y;^eK5 z`$vjT%l|H?csp4yIPX!w^f_VQOcNLybdeWg#v26&s+xE)FgT+T+1(WN#~_Zg`E`0t6^xk z&*?J6H*YZAz|FeEH+NDyz5VUa{SNx*utd04aWe>dDm8d1p6Oy*;ISmNf%Lr3>Cu}l zQ?T};i#aq4&tI4Ir%CI1crpo;dn)`7`7aL!&P@4nO{zCS7MZH@d3{d1T!F<51quh^ zOj+W=eia9By=rc3eP$3)E!(Kx?|txJ&PLc(zg6}LSvtmGx5LD9j-%LEibPWpiLiWa zQ#}GwIjAW{j)s1A1|+RpAs$U(;QmRl9u|Y!e&w{y+BK1g@#{-W&G}o@;9y46B%yX)zmh+WePpt7b}d|ZY} zPP}8_B4GQcIlMb<#s)4Z!+VQiJ*C|t@UmP=9JUR`%CfOhay#^V_jxOH3~NcaVp6j)(2?$kj)cq`JkQ8tuGscBt2iL-~8??GLwgqG4c2RO;WiP&4a-~lWognhX}VucR5JfHdY7t7M12cQ#*!5@aMki*_0&p%z% z6ItQXDrgK@0-Aiafp=hI(jt#@iwjQoQ)%fG$SKt@SB0Gd91-*-w9^Vbwkq*t8{Hwi z0ks=X!ZZ9;*Tn6i8fJy`G)dxXeO18nEn6d9BHao?09s$JTknw|XAm$w?-D1VS+_?a z>uwiYR)D@F?uwP#(3MfUVDUPsNb$>2qWD9Wdl7Tp3&~)bm?MFxtX9LchbA&(x(nh8 z98R-8f}1aYns&*un!#Z`7As{IYQIh1NbOPRlAA!NFeKWf-g4rGFQR`-75I2T; z!bHHh66&N)$i^FStTjcl^Mk@U+UaE5O zZFwHkON=kqcvrZj1!U9hvUL+Wgk7YK>6fKQdm^z%o>PO-h-S`g8f`ckYPtrz&S~Zx z*MFQ_<(6#O@n(hU9K}=5lX@8hv#`sf)cZJ}>tqi2{DyyMDKeiG;b zgM>&MQya89;4!G0AD2MNI#m%CvFF&5~KV4sqn7r z6Q@dE0t@$RR4WmBdZFI6GzLvWQ)Qac)b3^O!Dv7r&UEx0b9W6(QHcACmon zYGp31bvYx_LkWB}T>wHNM_kKzcjdjIPoO%cMZOi#1!fgwO5)u#{52E1`0d~yYn69h z)e0P3FIy4N7J>Js`|b8>A*eLc5`1kf>Ed^iC;r2Coa_uH>u4F`6;5EHx_fda zN0`9$+fQZ|lA@O;x^u=1dNmYVNs;|jY$~rpfelK~L*@cRc4C(@iIcT50TF z!kbiGQng<~SQ5WQctftzggx>o41eO37H}8TuT{odfdv}x1iMuUbSwEDX%dv4g#A5M zT%^kHxnSxNK#HTGboQgU!^fLF1(Cp-4S}lXy?pF@T}NC@82V6k?$0FNfejKgHw{Q1 zZKT*#imat#ch31lq*km3&i8C(DK2ac(hT|>>GE3baZ;EY)+iieX?fn~GpyG*nU`U` z!|)#BWL{ng_1S(p*aVH6-@Uz?5a=nOZ(>fbY-Pq`V!f zb-5{pu$%HR-$3dCo2zPNrx!L4u8~%_q|93Ft&4%SLAW0(0O08k`jY&n7)3S9peg_f zkXG|c6u^GRYD?`wE-r;$1Hjbg4R>4i=_c z;-vx7Ei)ipqFC^^&r-42f>km_PtV7mZ`?^`PS!FVzjw_&24>)7_wk zOv^zRVJy%!2!MWMGF_07(^C&McApDZfm5B} z85a$5*Q2@44lxDXfHA~iqxf>yUydEM1ct+AC#-bDJyj|w^~P^}D!+4L7F19oON5b| zd<>@saIVJJ=>*DKzCFRZmxKd~Y*vJEug56micsU_R!+Mt_Kbd1#s5FVMN zrSC)IP&K_2cs%;$75r;r+|<^)*6_z-)C8HSGgSdMt;v z9aadhbP5f#{_R@#UWw`=+YQ0oRCzXik{Hyp@Sv5xE06cM9-1oO$unHc7iG}9r`37c zeeoC$cnvTfF+86?04MN{`_biNU;Ec8jxaKLRtg(urH84IeLhEtjC|P;YQyqy*f^yz znXdb$q)CpsRSL(%4suzOo?61^gJ85??b&^}<9dj1CQqI!2Mm}-c1&WVL*d5_`Bk~EW(mdnc z_PXk1cxe$R7(iUNV$vDWQqgnMNyD9~VVj?)0J4)Us%-rI{9l+5^NanPzeO&-U=@75 zX1Ms2V%sUwM#UbE+z*6kz+{CLoY)PN#6QodFdC?rpxt=lbVXR>d)>>}Ia|Xoo4Aru zbw=!at=|b#Mk*|^!N)GnE|u*K)=q=$WIc_qYp13770jumGerF|xOhjZQxth&D`p;Z z&hx-DwcqlI2(^{$=YguU*?;W+G8#$^0CT{pUKe+kw+a^=77~ zmSU?Ya)63m=vfcRhh@UcOhH%$i0ABOGMN1ERQWt)QEQQdG$d&z6`)6cRooeoEIKf; z5$KZUwa46_b5)F#Aw@K#E}&=Wv3!wx&G#E$Kl{dN591hn{OjoVjiX@eZ2?c`Z9dBT zmNtRogxLa%QLzbJ|1K_8f%-_iR2pdl(GSbEYe?~U(qJ}M2Pw9iA{A6@Vo1$w$lL-| zCDiyEgBnkqaHoR013;-6ctzMGtrYZ0Y9x*HodBcUuPQ5W(NwkKE`&^%1|1GO$m@{e z-X-6^O`IXpggtS>x|Vv`x&U-LT4|(wQSIba!vTinvjg`{teUwQZu|cyi%^am`SkNW zv&oV`f(r~CxXZzca+r~5)>y7^gI7sEAc>6XAg>!Ltss>uD;7dFX@*X*D$GdBr%IRW zCqekW2}1t2W&lqKe35SO7P>Vl8ih~`vC9RNJ4A+l@KpJo&?I1rMb*{c4OXdB@h0a~#v(Wyqa#r2+Wz+5kmCx#TIy<8)AVahzpx>!( zdP{V_)3TUzVd>uOF^d^&?LOu6SQ_V6Hyu?6@lMnojCbpI+7x(d(!8Xv+h8h&g4n2m z+stwd^{zv|&I#O9uyg5~mXrW56i>C4#(#|d-S()Op!RB<)d!6lP#U3DU>FE#YDVAE z4KXdYV`YeeM)_sySox!BMx13fii@w~xJeXZC8^db*z9?0Mv?5mEM(u)(r4)Pyza<0 z*%#C6g?H&4Q0nkWR3@*C*Xz2FIR<3)dqnmB+DWyeAylD{$!F2e4!ug!xrLG)7$W6n zuGe;o1=-~sDz;wO9A4p?9Qk-UvaG2Ur7oFDq~ALnv}uYirisaeWJ5=2jigmj8db*Z zA%`PVW|lGa(mKf3uAh9!try7g(|wi$q1q^{KUPeE?WV?T-FbRqWz>;hF7f!u*DdK` zpIvAi7!y#8H=sIi55?}L$WAIYdv>c!rpWkH2NbVJ0FX%@iK}IK{-@n)1(!h#`6zJg z_3`$ITLpMup~o%R5#K_Owa^Nl5xG@~jSRY&jig$l7Zl8Ch*=!G1AeL%iOjiw84_6Q z3(;9&&(;8wm8pAuZ8^zy;P_vy8AeJdwupk|CKg+yu7<3TV*aQrpbLUVs-2z<0wZ%C z#`^E_A>gD|;6<#1+XnHz&xP7)k6gQD1w8B|O_i^VNmpPWDH8ADIFO%0BF3tj#iN|k z9H9YuYDn<+$gReJ5v)UGuua?B3rl|aqBntOiXO<<@)o{YAF_t`05ms{ms1mVBW7L$ zAZb}lZ{%Hiv*Le4k5^}GnjCyBBw^(h76n&Xp0gKiyiY=r_ zJ{7wxVwJp{T1mP?AAY@slvBp@dfBE|jaQpubih}x_01IJ@sQbke_+*QUCiy^dg!~z z^4}YLRIzk?E_HTR{?ujTG1l}r^x7y}Arl}*!x)V+whY7V{YAjXA`@gX1?yW#q64o= zPzNyp1Zs)}wf2ov>=ik54#0m%O|Oe77A}Q$N?pvN37^T&3$-qIbX0mycvYMnwVSym zJT+D#2i&LJ9uxfRnA~j2rGiPn|KBEn{NNPzE%M1=RTGXITO@bQKy#g9uTkU*_JOBD zW~>d|+^f)e)k(LyLj)STQ;ZGb#wvc)kW$r1&d$mc_dB6Z_x*r`sr^pHkN~~?YW!5> zg?-}GBCiALy;>+7h6+GsVu;3N)09$gy=0#F&2 zc^b8hhIztCqya(SW^Y{p4h}C}4BoS?>z|U8oaOHM%XGw#z-=?~D+~0fLDl zFH}~53>uu~ir2`EgWcerXYe`0Fc*VQ1ot~>*hTJ}qKhN1h37yjQ&ZXcNHh|G@7AS;P%s4jkjrm;r1p#jc^qYFPB%x+?zgy1Z<9DsLs( z?3d>cd-i}dx5aiP*cyBqYy`#N58LtTy8ZH8%QCH(Bwt{qsOxZ0PsA58wIZ3H^4h#k zx(qsSIzqc5Fb0SPSu5mPrj6(Wx+67VcfCuz?t;T~m{&;dLP|4VTp(_WsR$_~W#766 zHC}(e4l=6uWax%3A72`!4JeJO4Z5q`$$K=d?KLd1(1y1`+3B^gey4U>jqiwq$2Nj! zNK0XP!$|SfGVb=g-Ukw$!`NJ&C_lmIQTkh2d>- zEP{oS?Hu|jy^)1Vh{LkRNofM~0+A%o*lTh#VnK@ER;W3-KjVI61`0;{^~qD^X_93T zYQ+MBII-f2%Z~6XAV;jF56LRr&j&6@_v?cAn_78Qtl^`!1w`+#%Rs|l=$1q8o!m?A zNVm~Oj(#dxs*osn`d;<&fyXZfcZQ(n1rn9vY(eJFAr_d5LGh6oQ0+n1p@=!YCKu zqEWZq_^6&PVXhnEbJ@y0WR2e!-qnCAECtjlt_YtvB}YC4YXSZK1YX^iyY?CxqzR|l+ zc!=Q9Jy*S~PiA};9i;}T;aS*+8DBv`j7Az516~fyIj%$7{GaX9JxrMUVa=BFWF0q8 z%5nc?1QH(uVoq5U3$E1`$e=>aAWCS)xv7rJD(SfB@>zStdw_EZHDlWpslmXhWXOHc zUr02dFRwzSGGvWwg4D-|*S+VjPs3jopFi$MUoE=uf5!c5g{LvQzCWrlFx{_6lqo8w zj`*(qx`w{Q_B(wsJ;fEIpDB&o-pTz=Rg){&9!WX1Hn`k9Z^|X0bUPzTQKBQzDme4T zRk0p=dN3koH}>`6L*^^jX{7KN?D^cDRqU%K!{Yykig{!sH^btXmmUQq>$56D5Eyv>eX>M) zIOw=itvuuX(7tjoBZ10bQ#@MNrY8LCwRcTWdHrDbPswoy&Scy)oBWFud!8aqR4mp$ zV~54Ep!n#u@V#FgeCC@#lLWi4PSP$bpM|oj7_7b`RAok&IAbauSW2qtHGXZd+<0A( z8(v!Nflp~*JAqFkaT2(AvQ+t9S5$tMoQ$dk-FbYQ$`ByEkn07#!ebz;s3&z^H+)){ zV^HF5c;UqC9m4L&P3-c3VjNse=Lexp^Ewi^d}UiW9@&5? z>#^v;y#2d>T+alWz%?!7<7AS{%?>$meEN_X6iX=c*pF&icP?gef&lfS;GxD4(u6227Z8NoI$ahDYA)* zUCKucK5{C~+vG9(xikRXJ=n!G^MlV6p*VGxt{hCdN#O9CO%8kQ9h?TKM^YX+$ zeA@Iw6GBAZe;7yVUmAo!lwkluu2Jk2id>{(KU`2Cxy!%H9;FK;eZ0dmC>m?>I10bi ziX@LCKtkI>Qsq17T)!OPzQBmUJ(u&$AtnxD75I80lPWKpWqbn9ACP>%owM42mFy^)7z6wxXa-L$-jN!Rw!fw!H3&{vrx4%+51q=4Zz{1bLF%}nA@ zj@7Ur#&KyI{PyPq^5Q11S4Sc>?!aEO6{+zf{6&8CL7AfSf!l)y9%G3}D)`a$f-ZhR zvF8ShlNF6M@-hJAqajD9;JyGUYyW!>>{`6aM*0bjqmM!eIzVi4!rbOkvQ8UY^3Mmz9(gJC6mW=%GbQw z@3bOdbpTemKa!qs1-fEI>&yqd+mbTbSlJ%VC?6SkZGQTf!;PHG+_;l5^mqMDaQW|@ zht`lRZhIvMwk9=Z;3}rrT@)##Vz;@M&nof%`-0D2Zx*-9a-oq^7qiVBJF0H|FcCVk zkbEV>FU$SnTX>e^mgQddEl5GdT!(6m9doYBJ5A*n_{4qTBk`(lBr<<}I@BmzpV7dW z5q2B^WW%jK|NYTz<|%59-9JheJFt<-G{eC^;LTY?B z1r;$XL6d5fXL0Bc)Yf?(dZf|4gOe?JrQ%@Nx=AKfC=*sxku464iv4D&D5O|$KC~eA zBQNrT3Uy=64vI3OtjkG-fq@NbiB(2M(Jt~NsBc06bpnEKSaxh65imYFAG`4I_I$Vb z#`nJTFOMpp}i=Lhrl$-mkR^t_wc}RRzc& z@hgSL{4WTbm>MQAq;b*;A6;aEbMI>)QzOvLjeGsHcqhpP0pG-s$6j~sPAM>KRE;{0 z%m?i3yAeW6c4fDd<`h}Z%|~(IR1svT2gGwXP;3fClBwAHujwak_BleX%IgK^Uh5Eo z&db!w5Ri1y(kPmc8&)>9T@752VLo^EJ=+%%k~!CejEDuFw33P!td$D76b8)yDT+N# zf!tN>2465FHBlG+l1f9;Ug0C*-&`-zN9dJ@{*Ii3LlH7-^H{BqGW-C$Y ztb4KzQ`LTHf*#==x|rOS&+nF{hYe#D&JqcEb;6#F#%|9Y6o|&f=fFj7%ma;v=rb zbed-^s}Yxop(B`HOLF}Z*tMtn+<5jj$vn9t5|2==X59Ls@6cy1jxe?K;Lxwpyg{QS(g zvm|XWyW_x>3wzD%PCmtIDG-B-#bN|w_php2hV}0#F<0b;r7gvDF+}m8(xqOSZ=kPG zrOrTR^&6qP2o||-ncOSETvDe1@3h~Ze8fg(grKnYxOK4nxb>g^ImZN+cMtz)7dbE< zY5)fM5hp437)6dyvE}ZcOKZH1KU3vPfc9`TlmygxKL#O8H0Y^e#&? z^UmwWh}dKAG*^t1eWW_{+o!=TT}-q1Gr?o;`;phYw)^jxv*5$W{_V1k0A0)pvV3A( za39?)J|R3KDx8L5%VmMBf<5A-S%<&7LVdNOh<7)5Skw8u zkJPr;h6aRfpZ4uqz&EPPfN1h`4=j;vsEU*~7UCbHae>GDc!$1qE2NC^FHB+{b zwupZ|h-ix0?DbjThYupo0!{6$8H?Dqh>!4w+YxvJaMVV(MI0eGJm_HhwMBFixc50< zt)byR(6Wv8vGp$vC-_soCHGEwn#|6n*Uo=HQXDu~419hA>|r*=LQX+V#bQe6QFMpT zWB%>X8m3x#0(uOpCH>&(oFgZhebDN^e~zm4-KSZlHu)k-wd`Bxq1q00sh=~&)<0z1 zUl~43Y=4j8+{q0lK5akx=L8c>ZvLMu{~#;4#r_=FX9gE}0H`)mEEw5!Mn*uS|1(Oy zs8U3qOiPtl1;q!c6-d;sLNSjFIycPD@EI_*9Fa8ON&6oD+12gOmX+WRyHHuJ?Z)ok zJJNDGn_lP!0t%VDt3L#VW!XZ{e9<16J_>3bu# zeW%+nIw$KfZpIX**>YcPWvhx@0j0A}i$V44DsQK}6}FzLJm9m+7eOEv_Ma$Nte2^> zcm%R?e!}%6YbKV{7tWv@<^$HT>@o4xbV~-3XW`+v+d6A>DcL|)p{KK;ggc2W2~(|hso=MM zb(*KFq#K3Dfu%%8b_iPydYGzIxOue;-mab2BWUzk?b0aBppVN6Bxhf5rGYw9xPyEu zRJGFGp^2jEpyi+ebX0tpAMXMg%v5@qJWnUN0bZg}Qdbde-CxyFR>sk!*0_&C(t-zgA1JAQhiU#m31u zDNjUYOz9GQb<`BI_Xpd5hYchPfdzk!Tus)%ks6qgwu?7Gh&i|nQJxio;si|??<(XK))g#w9xA~7;i^=>7mjlW)7L@Q2pd^lQitMe`4JcxS_d z$nf5a%TvO2r`HNh%g)^M8xE3{;|a)}42W8!QS1haq#*h4L*PuP@ofb))BRB>qJ5dU zBsndv4?*>l*4Cw54stCcKYYG1B;;;9)DZR}&R-jmX5>s!W z>%&TheRCEpq%uM8_bCH9gl;jCg#Wi2q&VTnA~uAl_JWXg{QUXBp=3)x#9{4-70==f zuT61H5*K{U)4&0WGI{Z85G*N`F7jwtwY7KaIkFQ2#9 zfps_xvZ;Bc=XbH+a54EIAF9s%nZ!Hr_7^P000(I!#exvdS}Hb+SL~a|bVwULntt6B z^N+>U-_OSE@;K`W#c}km*9|1nIdqcTZieBRndAtKXP&p$1y0bQ+}hq;XxT&Ju-3y$ zSI9t0VMDT~MBEPS#7n*P^ddK81BOUhy>k&=3n>bNigS)q56vU1TzVr>v^37muAt$F z6SFnNUKqh~IpJhnXUnn)&L$I5_qFPOk|pDT8*tz^HHBi6DUygf5=*CUf^6$nzY3me zxwlrC?4ID>BPpC*6|`4u*N=LTkIUvye~@SGd5P1f@3@ZM1MR9TsRMP%j*6b98MYZDW?xSGIeSlm_mY5Q`%=X{p$Q}_(y#p|NpN8IW4jr8w@_>vMK(~e znA*ntbOxR7rl(avn7~8sWGv8DSEgDH+V)V z0OJ*ur%4Ri0W6X2F&J3GtBv$g1E&QN7VqL+hFbcXDVG`KTS8J}J51KF=ElOn*u%td znj6P0yY9^>j{AcNG_MyP_935+C-G*27JU@^fFeCq>_)dXh5EHclI7l5D`e`YXi_vO z@|acb*~l0=33SI$T0B1+}8*hsBYbrnH0PNxQOYvTD2AR<;dvmbBB-7$iWg6ejUC zVF_W4AgQFPcTQv)X}BvnQV0Eu)sjYF)m;rs>I;~n9c16?xU~RTOQTl2=o@Mj>gu8o z&e=g{1oq21pkWd7JE`&$@Fj>3OJsILZIH~fgI%<`>~J!eR!15A7q}VB*$bN&z39$P zy%fk$=|u$08_57sCH#3$c2C$ZYl_+S_Eo6b-U&oLXADF>Yy8FvBb@Mo3+d=XVbBq| z;p6bpYFEoDIV*){DZG{b2Rx1jCPr_iPrKa^Z=D2fLvp>3PO%U)WUq;vC!LyfK(ITw zH*))kdxZ!6mkoB2>nsd5hCQxvn+1o>3eKIoqbAzq*7z@8wVmvB;Jwm0GbjBh#RB(j zEfsq?sEa)EPGoMlrnw(;-tN}LH_Bbt@H2VFBefzSOv4m_6RsDW2kw^Bz`D@vwi@~# zs|CBg@D=3u|LT6sfu}LMz4Dk|ve4})T`g+`PwcwaL#Bow=cbD}9Hf_hb*+V$AsUfU z;)?NSy!Pi@7zy^dVf^#oY@TRYeei4&(Q!R|Yo!v6bT84;XP{ST6Um;vo|oaD20nFa zST64t(Zwv|KlbmSwR6w^Y(x;?Hci|B!)@%5+-)5&n=}40-*Opx795TncqCTV%-j6i zeB=%TnK-PKHA+Kl3S0!F|0)bCb~^X-k3xM-18{AOMFP<9vEmgkR6QAUvhGN27lOoTqe&tXF)9m(RNozRFsU`*U;u&F(-LW@FhHY=x}b zl6{9f;>Vj@+8F%@Dzb1eau&x8_7KQl4X8*+rr1P^tfFGIbk$5y@;u@UWGtXv`^0JK z)FNKG%l!HF>0W!z_t9qa@SHL1@sf2+JP2eCTBeg7w(P)ao*NTFkPJfEyWFtapfqV3 z#9ph&6_;HusxdGH1Ecnz=o#rIG9r;DVjg`dn_P_#=xrUVn+ojL>6w6gD6<96{#wY~{WH-1KTDpiWtZOD?6K z+cF1hYP1Z*??FQ|b;9UbfFb9C8#sm>Zq#nD4vrY7gQ5RvT6d1Td&iHoIdIfMTL7#NY_Sq{uE<;Rjpl#y@v9FdJI~{!qQR@&J zEWiEBKgWGrWP)9$V0{Zo(G5b4jf;x(g&2IY^Cd^ zb@c7vdI7TYs1=|3!1sHg~xrUT)8GL zpch7Fh2cF~=KhRsUY&SXz(RgCn3=`n`+xwqL7=C*qk(0af7c*Szht~FCXZRjJo4xy z^#U9d?=n(z_Db|4*f5O{IRij6=;5%DWgR(#EKSs-nazbSdVmJEO$McT;Aa8Bvu-)k z!lAQLHFKaZG3-y2(}Om{g43uYznK#(UMc%c;hT{rBlE+u?HW@2(s(QlW+v$%#X<>k z1r>{(sb##+g_k5(gg~p4D8iV`X)!8BfOVM;K5Yjuo2?Jq6R1{fWXc0IVJqZNwb{&I zVo3G**AhiKzfXKCT$24CN{S;jVR{KPGNSIp$_W_&#m~b*r5^Y-a4faric6k94x(0Q zL>=%xBZR5Vm>bm6JpbZIkau!#y0;}|GN(A9!3gu%Npij^5YKzuZm(VSzm)wolUcd+)~#Gp!p*EWum{{|W>yYSEXXQWQn78aMSl0B+Q1&s zS4XMxwPX|9Vl5%q|S4D-cwLFCh`H@$i!8bvSB3ydn@dES|#a@Qli$QWNuH-XpE zEXQM9^9myScpyg-h9jl>^+@Uk8$C}(ZI41JoDpNs1{gixcvyLLr1%n_{o6%96O0O; z{J4~8UK$uxn!#un#R4B{J{7w{Uh1-l+^4UKn_^nu!OGmTF}j#NWJ&aT*QS{Zl}B95 zcp1JYNPN&nAPCckH$c@?Wyosp689cglpMjrIuzhba!w3snu%g*Rb<5Q7;G9x12Th+ zJJM@0;tz{obj`&onD)P4hiy=WYeGzicdG!mKd3Q(ZE_Ls?4+yFdxMWE`el3k?~^R| zLia?GhDl&GVX5+YTDr*VG^|MTnqm+)TFD9WIek0yUkU-79AK*Mn-2xyCKDrco|8p( zy)-7~w3&&irq~LK?4x3Ldo2#w?|;NMk-ypbVQ2?k9h3`gN^zipmF3>=RPS2Dzc*t) zRNqz#`<+gabm08K_Pk79rLbOlCU8a6#aEkR-tT*3>G#sU-TKy!XnSH&Jv=Sizog)4J3^Mer8Cg<<`p!Q(I*Cg<6OYNQ5d+_<+px+qK9?U%&+wL;R3 z%sx3ZCbl2acLExq?I?TNQgCxHccnsMa>#xIry%AFE=iV&^|BG}du*Mf;UL7;u}5;> zkJoy&^Grs@DZptFS<21GIB*$6mYI=BrP#F;SwqEk%XC5gG6VU~R#vBI1tj1lBSgjs z4PVWe?YQ{rnXx#EjhOjMOR6U;j7o8Cz`Vgjw{Xwf%?7vd(Hv zM)P7}5l;t=KGm}Q{tNB_9Xe_#G%0Sse%>qFb*#Sswu1J{Rl_#Kjlx}=Ks!$GX3pgB{mwk~}kLrBm!i3g!n$PW5yi)5bnz>gc1$ zGM2}#_k|uKL+ZH=f+N^*QYpMn7cobi$H*RW$IYm~V#q;e=FH#vstFzbf2f#8Hoi3I zC^bXJHj2%mfNT>q0+4$yRlZ+Y8<-*LbzKWJ2r3QaZwxGSM->`=4xJ&Y6#o6HScP0s z7+#C>NtdsZmq`s|irA)x?8)|OP~#32ODi+vtPksk_gB92W6PlZv%Mh>yb@Un&1Xbs zih3kPqJAfhV85b^z79gYSH;Wt$$ag^dRZ?tsqPO{E3<*gr_c2&M0+-S-V}F1>549< zRZ!x;Yib9bL0=H=53HCC0ejbdQ*w|~LAl;%t(!&xwcTtZUFwpn%%F?Cie$wTfK>rT z9TlrHa-diOXLQ(ptgl5CzjTh8XtFl%ZvWkTWU~WTSMD>j9|aT(%S|p7i%mmDW3g)L zR@T6e&?BhlY3UlKlGje}_Uh)9`1U}T*>0wpfs(l-{#{p)PMrzV3D`qc(Cgad~)1`;+WT7d+NGO`|EPD-GSk9%nUC3DHbF?_kf@4 zxe)XjDj{!E3Wc^=;x^{f(A|OUf}7GK1kd{9HH6sKCUFW5b!YAsb-J_B0GtJ&;pl5(-83Gfw zzsw0e)W4T2-}5w>4^w_&hlmi7yEw5Uh zH$h_fLEu5&jhL;}r*G=#V0br!Ud^6|j3M58adxvmsKEz*F=w;lfVh0t*%-B=<<(sV z0r?iPKQP^|n$8S-B(4bA8Qt>gO^+t=dP$ty;h;0jVNjvP5e^5{OLZ}enHEw6pHSoxupYnL1P;X_>biVc#44Zx)bbKTij+kGNkO`p8^SZ*)zGnAJEbw? z0vy*1>ILg3r^=U)$Kyn%(@U*bKRJFpM(OjIGm>=qQMzB=D!}Vdpaf|F zS#;CPErDpxlE`_+_7#ZRr)+-rFtA`gfSt~)V_`5Ww|VM2^DX;m9Tp!}I%8V}mn40z zn}X^E$UufEh}E9OkvqaG{SX@|ObW2s&FdnU*m-Kju5V*$-IyR`7yua)h8b=ExjN~c z_bgKpR>RJNAk2Xy$ySmSMwWXWwEknz7$V7PWmV8-#c9Z9bS00)0b#^q~4j-utgV zs{Zn;Km6v;?~9jF>|%<N{ zMZS$!uLYLa_%a(DH)Rb-%3G`n% z;lEb*O!r1Qmrn8PkS0yrAX+PI^zD+CIAc#D9ydp25#$Ssb5oNZ*+us{<`AixxfCci z@6Tv;*Tn!?A_Ni>SbRNOX{4Yk3F-~T;=^S3A|dh#b@N(6+Cm?Uw4-ep(6Csktex60 z|IfLFlck}Wq&t6US-|PA+hi-1rSVhMk_=H5Y>>B127%}za#NZcR^vN17Q$>MY{P`v zBV%&H2@~UZxqo`cvZcgfF=3_G6}5&*64 zQ_e%|TEv#r--@4w$isfy2!o&*x&R-R)zq#svL<|v2N2OM_TAo zFK8Cy7s6xA+P#TFb{S`o{0`(P^8uJ8ZsTWFx&iBzygI~`K8_$c`7aJgtohBfJgyfZevv1Y0F*G zm}>eoDFN8kha@xgA!W!D46?`Oq`J0E=+Fkp5-7twx9!jQ@-FKJ=#_2a1GEWJ_kZAb zf-G|24YbA#K5Hp<4MkQ{vFk!wLyg&-+DY^0H+XdOfb?B-Sk^(0x!j2D0Qhol#^k^{ z00whEu5JyP{jbGAciRKIM^q?VZ)zE^)6(+%*|r5f&mX})uTpP*@*7Li4Ts$cTcOdo z6RK9Im0k2{=gx_$HSRn7I_OTi2DT;_K*6_s))>#XEi-Lg8|*Wq)<)%KC~jO@S`lDc zJ?dUxTTZeac=f0?TRloBwumA-u=2l>*1K0tP-VFnG1t8=&Q_&K`s7=brQQ@32e1wg?x?tAWx@w5QTd) z5L!{{WgEVpFWMK7?-EbJayw?<3aek2!CrvBi&KXA!kO7$d)oxQkMj22B1gGJa2&S` zo!l@p$vTQXOOZy{t<#0%j3~}+wabR6IJc?^kX{fS2*^>MXZlG+NFSZZ^to>FDw7n1 zS8n?jlMy&+9NsgJcz09;2lzS^WuW@A^doVw>^rJ zH#e3*aKYASXc zkV&DAjV`8Kd}tc>Mzsnq0P$XV)K;i-`Rp5OzFGe5(m9FzI@f+6<*Q-NI%kSdGoW2~ zel#j+pN}=$@oT<*Jb%=Yzsd=}R9VL#BBq;+$J~~@)g+(Wy5zX-9sxa^0Xe9B6bnj9 zB~&b~6%A6T;)Sk!5d5T72ZEY_5&p4%7U^^E^i1$nB}ZZc6SE3WoKTLfQrIoK$Y#2# zT0)j9J0sTls8)Mmbya%wQN<^cZgJAotOyJP@9?}9*$I6o4|%tMMxt6WYE_@Zg_=3= z2A+)^W;lZC&r3v>D-Rd1$bp?2D+_2JZ;!Z(z9;L3&dA&RBOc3Q*85}xtoKP9Lu#w1 zb7vzep58uozx&84xyht#NDjI{);sV%tHjJm^RvXfHV(cpmxyYQ@-!R!3{(Yk0NxuJ^Zb>IZ(~O~>{{ z0k^ehr@r~ni=I?k%5U&U;&(vFOPjnce3g4Hs1s#+-IkAeO3C((vV9@&?C6|8@yd^X zbKqQ%$$ISWD1wH02euxE&8$Zm#ez90qGFFnbca6^u-@~;}c2|3d?eg7y~#?6&=;10!XGn}ob*mV?1 zqGB-*`Lxo-;L5HB-QjwWkx(m&cz1(G&HuD<{y3~kHjX$txBbX{?aH5$o{ck4EBv@I`554;^Ry`+5BQAHN76zW9^ z_)uk%t;}~@B2~M`1@H043wy{iepT>RwiN_E+L)cu7*;(-5~ksxPn?c>VxFg3naJ#M z%koz%P*ny)oe+a57wIQuE31gk zfYw9$z^qc2s-R-ucq&J!WA8|>5PbUrssQ7rR{qCfk_P#1kNA?6zlv)()nUUfby=~h z7d-{OC3qKfn50Rn`2Dg}dAi?0?}agXL5Aqlh|ANf_|~Cib#6!Wg3W*((YtVenA74! zz53buOHz{!I{x-M&15aN#mRx=EW6B1=~jw`zLU*VEY^5Ky0k%>$Dq->I2$tKpi~6i z^dXN7o)OqAsHQ9RMjcQH!^F9D2ow26^>RC7%w|YEWA3=agv%3)g*kt->?C?$%I=aubN3!Wv%oR4peEx46e_Zv=72nj%UG$CS*Dt?y zHpb52FizNTAi`y;hkI|J@IrU2$&Ne-`J0TizBG0u&Mf8KMX{e#ImpBK@D(xc=x{Trr}Y6@yyFsJgeZ_0O4{*H z!o6J!$qmbLH*MyI1G;BS)aYI?;504%dbp)sa@bsw6-(11FAKaB^_jd#rV;l$X(Ric zs!2AzCGatSA>@}JZU;I~+Q_ppc)i?o{a@PBaPx2aV!k%cQpG}R<``_ESisCWDt4!= zFua%Ck)n?7f>myywQ1-NOXXE4u0t_)IUP?Hbj#Ai#^lpC;G5(3Zn(Ca^K+c0z;Qjq z_vPRJ(2~M~i(BKsff*}2L92N?ggM~yR?`=GU9QzZx^Jkeg7%6xdnQb8r%mlj7~R9A z0AhO>BP^9|wHF`sK-_VJ`j?Bzz%lG`+<)qK|B#vdnv99R|8J5ySkICJCk%_tj9m`J zZly>D6`MNONdAi4@csNs;nAoHmxJ^jvGKTzToc#PYsqmXGS(TdAyaSP#2W8LI?vDE zCUCfDus1)Qb;I}Th2C|OO*mQdfzN7^F`j_P?toa{PKpK6lspjj2A)tX1a72zp*1;m zhM{!>qwCn*j%|W*!3QGp6#YThy?P=JLRn%*;2s%NENZ9KFx|X^-rxwH6xL3;B5ad4 z&Ab3HRs4$DVMc~M315CW@nH17l-4IsBr)ao~hfqZyPB zQEUxGDyi5ceqlIj%mXABxh-)=6@pUlepzx9)B;adE#=>l8r#`2Kn7Taeid#l%A?MQ zKbrf*sf9F1zmRPg7X|E;t#UuCOd`;te?7Dgn(ntScg5>M3d1#%_XZo=8CBcG_eqQJ zhWv8ShN-G5e$`~Gb{|nDVRZ0q2KE6b!^b^0eDiZ!|JQF!FnV-nhlDh9+ju+h()Gv; zbayEBHbrhyv3KdM(BZOo($n9ma;$r)6+<1K&mmqH9sixH;(uJ1r%8^v z&i--Zn{Q7v*`%k6I#z^hc7X`N$JgZ*2Kj18?gae(fUqt4z{ItY)ovMnb)*mK72zA~ zwo+0~ou1w!DdS~|fYv{_TYf;O4^Ir);kgv_9*)wLAqOUEU7r0Gjmkf8IG?lmw`x@W zjAsVnv{E?k_wM~`hju%g;8ORs>VJ|Y+?FuMjWB?;+rTYA3dJTpXa4XbzwL@z#stBqTiA+|Q&U=Xz;fPcsws_e# zgUH-2j`VCuAX+z(Dc%2h&yv%biyz~_z_Y>tjq0MXn+h}NA!qS!2iME;-R!hvK6hel z0?Ttpu+NLXU9#tAmVCVqn@_UBz`I7I7irv$IBqnG z95P$1N+}is$GfQ5H8V4&98u~d8>TMeH$uKxh216X;WxwkWzGI7j6xPi9&y%7AbA}I zY35Wp5`?P~nL7d3=?Cr|^g}2uUhNIblu_YURV}Lxyvd|7WxNjhN(gqmF2E-ij1mmF z!^{kGgPd3bW_afllk5Is>Eb%9WwGM4V$2#Uri2}G^4ud<%4!U=VVXD{o(C<2YzI-Cnx5Fe2+9ba16N9%tEwM>{W_fhP{lSL)a?l;~kDX z!9T>W04ss&#OZ!}m?F9j8j!S1)#OAbfxYWmHTj5Ly;BAj7|f$_6zSrP!xe9V<5R-zM5YoJ41|r0X~-{ zJ`%pvyBS)bjs|AX^}=LN;39`E2JN)C;B>#{pcKCfmn$K?QFyNn7RP3g4PrNje`#oh zasnBZJ2n4<7v1(8m+Xc_01&iZ5q7v6ex>-SFhABN?+D*zuegtuncx5mD>y(B@>3+(9lEU+~2nz#UjgHc`_&oF_E zTSgT&D`RT2RgvqLCdr9{>H#gR7dlq0z{}7v5Lf|HB1p}gBy^*=jJ+8VKT;3W&QhB?fPNpX^ZIC3+(9p2)j*%hT z#y1(pZPt;!F)8iri!MoV-3#k7dREmdYqXA#9hGI2W<+v!y=I0FWxDW?0|#)kc?@(PbG zx`=IHHEhR>wPblfnj~JiE2`N)$$14xSYH#jgf8Tlx;!BBilGE+Mbx||>AVZVJ$R&x z=_S>IJ+dm##Guh$K5awTbC%7gFO3jUmXH7Hr$2agBym~?j=5Ng(;|Icg$q(gVwApW zrV0Zrqdnc0AM;@6HbG*r2km_A2lLy?Uvz?oFQ}$hl0>G{E1y>Z zLKAsR>D=;Jy^(u->!o!d!?(xx=&OCv4*@M2*zae-F$!0z<~W^w(Y5DI%_x zTpn%mM1CTp2~#T`McZeFEhl3bGj2JAUCy~#lw;aoht4p0s~=Rv)Q|!Pj?Nx8^FI$z zEC|??Qn5y134^W`P(S#!$ng*rXmY$%*!FSf+ck5Jk_&7%@9}>?r8%uy9=1Fz-UZ50 zs%FkF1>>?gXon9bjUndm+$~Rpo@pJM0D5d2z%9-Cu3CWu=;y!_$is_IjRL}k0iVH) zhQ+m!BZ?Dz#_?Ac2Uwbymt;ZOZU(k36q`no4Mu*Un-RU9d2=)88c9ST831vTEb5at z7c+6ew`O4RERu`#dUo?h0Vp<2DQ7gR@v~=dQr7U>WHpfP+sUg2=7mkl-E$gZ5<_lAte9{OWG(P6q%5ll$s~`&x4fH0 z4WiM888<9hEH?vBZ3T&q_gTl%m2Y3Jf5U{ErxI%~$(qw_hNv&a*Xp4BKrcz+=P1kR zg2)1YJv4Sao_?M@^4{jQC$O6Lfb>QtGHbuSgT5ElG%=Ic8(9yr{j%wLNuy95ePd=L zl*DPAcQZP`RwA!EG*_9)%bT)4@MzSUnR!!;<6RSPqrbYln3v<$=35~wi@wQgia8s2 zlPn*921Gq>^5Vyzjj5NevL&_r>`adu*aia6;`rHj9s0^|fBKVk|625Q$g-?f=1l~q zG)U~ENmYq_wW2t(Rj@i*Cu^ju$Es$|G&dZ8G4yPX;vL)n@pgn|U4fO-p+u2hq7OHg zx38S3%AosrD`(=*5w`FR>tAE@=Q*s=hyRdut3F67#c`wBKYBGcXMt%6y7<4AMvxu{ z-Zif^kO8j?C z#j?5!!uyeGMZavXGt?_h>*K9bB1ez0oDwoB2nzW?uGCd(xIn`#T$>A;o=q=g4;psFbL07c3{ zsa%;3<q5(v-*ApLG}!SR$Z}wVkrxY#c&d9YZSqyh z&63gN@a8Z@HeiIqXrp(lb&M?j)5#B~neg!8M{NhlHV3vSN6hAaFU9Vm$ZnuL<873j zQD%wv`*+caASw|@U6-GQcI;L`rfVjs@byTRNEf-EQRW8c6XaCrcUl~r^7d8nLb$KU z(6$W~D?nn`MOVLbRlIY~b@|8lWJw;GJWMVkBLgZQje-UFaz1Tzod z*a&Lp&+n(dVuH?+<-v!+YjfDTPqM`fI%_C)HAPlXv8Y{-9!RT;Dpg+ORRU_nqdCzw zPl+vvuz9diypa$uyfR@Lo9=YxgNOA@mTfA1_Ti|=!0}G{aADt@4WK){wPMOB}CNTS;dur%> z-ye8d+yt9%yd7`3;R31bY_i%#ODFLYMeQ<881y|Oz>Gt*$TPut{*o}aVp4Hry)&wv zsPJhNUoK*ffOai6PiQ#6u>_K>=T=cN$P3o(P`hxW(!;b0{jg@sd9u!d9T*TiAFw6L zqF4){NsigP1>SUiq ztL$!0hKqN7myifr+7IF@|6JUIXV> z3!&QcL{DRo7lW;8Me2ljDxNavnQeJyZXiX*Y+ zlbh=@*f$y{FqIMy_g!*BllzIYkME~ZG_SGXC@k(Kx2a(VkVe)s8%zI#TC zuwYN%l3*|++uVCxuY+)}{g;ntKV_Td_}Rzoe#N@+7{o_;<;Kp_Cx#~Kc3`t%g{T{M zG??vCX~J-S^2Djg>!x&H*uWGzl-_pBT1YV|o%O`YNOzzKOC&eL@d5iH$|md#ZX@i9 zAM4oJ_RHgcwWKe#!uxqw+$rxPg&rAG?)axk?(_O(9dd98d3%MKBD|IdV#`g;{=h2DkpH5Auz5+~P&&iQ|9jX7Xe|D4y_FvYK0$sssDD z`DQ-uW{TZ}(-3<@UJG#)@N4zJz?wmC@#^MfPA-|!2WpcB*`Y6AuMKU4P3qp7Q*<2_8Sc)tFB9IdIHGXEx!-DYlLRs;^kR%qS17R&0Fd_**q|@}Vg`Pfw;A{@};ilJOX zJ0lj$r0^bwSCN&k<~pnTpvbpIiMwXZH*faTlDzQU!0fXy=H#UK=oSdl)YG8GHw;7# z10I7fM9;m-9Z8&w;JD6H{W`G;OtO#PXd-JI7)-zhGhm0DL9v@DvI)g$x3U^h7rkT7 z15&N*kS6grDVsbl2rtflBwoVT@Sk3)BZts)#dm0*+Do>{X-ZseH>Z@lJf{!EK7Rgglq?elTW|arZ`4=T$_N)cx=pt5qcazJEkr1+(7%2 zwIBIg?uH$OW>k9&%Hh3In2D1 zVKxTmoPZG-P!=PLLq*MyG`-z!m)zdlHtp?qbKA<-Hf?ji_I8@wG))mV+*lM;P+3$! z5K$I2L{=k0#03$Qh^QbV8bgAhNKo;6o|(ZgBZG4WhC8N>zsQ;QEQ9m@=Xsy^S^j_3 zQRpfK-TUxC0ooJ8js=-%li5DDRRO|TP?w?ys2Kk^8W^Ve<&O^s17rC2;uB?`8vUzF zGRowhK5&05~bQCu3v83Nksc zFuV&dY)}P#Km6s3CSx)G&9fb(oL}6`c_V_P)y%Yd7VP(5&b2Cz%*4@hkzV}K^u1vaGXwn~Bx12^y7^B9e}M3R4}0!-P4?37 zEDIu?)_b*5{D^VXdU9CPAwCUsTxsgenI|M&l5){j^{&O6W);AW#?YXXPIrlOygS^X zcea>`gJ2uJz|QpoM%N)82w$_HT$DMhOF_Ju>592lW zJf>reOKn8Z7*BA_sW)PGxBbTC*983b*=NZ$CyvoCH|uEbr?@_f^iWYLWHY37>SXuX zFOFKYN67)dB{-1sYzAjMy=2J6Fw#TsXOMXo&)#zHY4-|rZ~ zzU6Fq#~e{M_)f^tPz*HmsN0JFI@5}1Ba|$xiX&edp~=e_QGLI8y~mO+*=g4|D+I_* z%w;yl{V0gXqfcB%H+t+8=vC?T(EJtCQYEV+;+U1wkP&Rfw4?>weEL*}RKue!+GAuF zeK#UQSi=lody;(U{o&VFPIH9gGa;C9fCb`*nf;Z_x5Or6_2T)>ACNWtHet@2?;uc* zJ0iZ5N^zShl0-%A7hD86sn4WXq>Y6*AT{2^03QR$BW98fLgZ1#fbe9d5qLt$0iYTE zk%RaBzhWuO>$C>Nnqu%vm*wwVWvdibnp&tbY&M+dS}GHM37xM;95cbA%nikDez09$$DypTBXB^0OuX{L@Gm%hWR; zXV{4YI#x31`vmoFx8~N;>AqK`E!xK5o6nw=MoU_>75=H*`}FZpN8!cF#M%xcD-Uqc z@X=42Uwzu6z%krarj+{N`Ff^L5bx7Vrvx;JtA*=?oBRr8gLH=cu4*9s^n_Cijx`~+ z&o6Ir#`LzQ(c&tyEt*Eg61jn4=g^H<*zW1wbKQf`>2Pdhole43BhdJ?ramx%TS1Ila zMcSySG-~}gTe&NGX7rz_5NLL}~03p?LifrgU9NQGwH-om|<%y=kzPBFpF6(&t*44B&OV&Uu zM0|~OhOk47ZC22usktWD?y3KVBR<42|0Q_RmQVSK4{^W+)<5M@x`CY6CLj384`VGm zm94b+7JBZH^s&oB3k6#j)v2#5i#57hdX=yw0*7=5RLeu-yf7FQErG6yUN}=l$NQuJ zt4WzA$p`v;n+%yfysE*iKRjvSIxpo1dY)5vWzqkfIzu2SR*6_rD8QQQnE273a1 zSg5d@>(?8xIU-HHU(}_6)|*sCt6~teePUdp?nDLK5u!sr{bUxyoXfcuRVQoc+JTD4 zKqE|AfrtRhVasM8@sF0oF-QWrIRY7L5ke1Lih_SGr zm)0TqB>~!hiQ|87x$ou^taDxqOswqEce^*zh&m+SXjSYKlrH>8n!YH9KJX6>y+>%T z8IWEVn9PsMY?m0af?qnwc?~PsX=X)}DQ*Kr;;E=wI$Lmqbh`Ee_qKkv4s{kYLXFg7 zv7#{z)T0k=0uO9AdK37jHazOhiXR>)CR`M`WSj&I3to26i6b>xX1Lf)aY+vm2p34_~CQqCmBb=~^ z!$gD45;XR{7R@c`dDz&;JZnw2#5YL9;4AcacYLH7jVMUo6#z6!eX z**hWj*$dtuu|0Fn`%Ub)fuDhK&0Mn7vQ5BBeSJE88NN*!!Y&&3v`7z{%yt29#h_b; z4F6xVAWxPKB8F>#s_%SQn`pwmU1NU;{O4ip=S|dEe)Hh8^yS-BZ+&w-C2>x?2eMKU zmql-yaoab~uP&l8;5|9&7^Oqdu}JKbyPI(0ovb0xFWuajvL-4q`4xTdWq2jm^% zzQv_(v7$8fHSOL7)n0#&XVZOaV2j+M9s1V2$QEt!*YWSAl2Q=vZC2vn=T*I)cvFj3 zui7|kkWDb0OBSqVAILtEwP;s)xjbKT@(iayEG~a+eqyFe{4jIIW4xGv=>yiJ>CAR4s0&H%b8%}cCY^e+5Dw} z3Fys?fJqL;K@nyq6}3);$%Twiqj!98;VF3|BouU?E0Pvg1aFVX4%MOVJY=fjOs%Sb zIVZh5H<4-fx+l=}sjhm08?7vLw~Io|wpp?lDZ`e=ws+xw)t}}c{H|rIn$xaaR+`UH zK(RxdUl_2m8Jwa85UsoMA5r@b(@LxH)^3Mj}u4x2O# zr0n#MnUl=!o7oAay6C(XE$9z-_yRFvifzM&>H9w7K1Uz&!p4+%BA)U0Rn7l-$ohzC zMf1X=Am)mh@wnhDdMT42Y-cYlKJ8T(F_{I>d=&D=6GHN_PsNG)?>~e)i+d&5Pf7)#RC`-IP2a%}!j{4_(hAd`&&Ybx`Co2syAAtkfTri!vDGE&#>?2*@FOkM4@+cF*ikr1rs*ynZsN{)kP;Fxdir zb34AP`61)*;`x6VPi)_bqiR-S`_Yp3`e zF!JNJah^xsGAYOWOvVpIH>O#ZL^>@htQ0xk^tvTX@@W(86eP13rCnah&lq`;jek$x zH{UqDHsWFuzxk%$i%*@=RG+YS{$$xJ=Cru5(iett)Iz3}ThC>LLU&j)n>^3hw*1f~ zP6#_UAZ?gH-hbXcagMRS_BP|? z30sk|--Z8Kes<(5L5uq=bzt}a&WQ`#tf3%v#5JU16d;Dtx_Q>JvL9a%)6<+pMQrrGXk7M3# ze-k^dad#Ubt!2HlJc`I<3b}YqhR8+^BP8>~6R)r?{=xK0^ zv9AM~Hyke#GukBDgrCuFbnw{nf4V)(G7I=PcZw4`tyVIFo0$$#rY!%t5_y+q!0mx> zo2x#edtTC4uS2e;nBGO7QSXm9?$bjb5$<-szG-;~%20Q;V9|6Tl6ljpg}i(fZ^;*I9iACM267!>Vhpg2o$4HP*^MKwqp zq}UshCxaR^je!Zha)GWvyf>^0ypB9cFoL?!d6Yq(4m|-x9;u0wmM@2oDTk4HegUM_a-5_2i*%NkvDTHY+ z6x<5>hiNiFtl-;0YsLc$Cw5(}z@le<&-~5YW-dGQkP=1aGhnwN+5|TKL>@%~ynG^>#JX@(Kean0uUsZRW-XYGk!65H`?L+#I>_H@iyQY{>SPz6<}y*~}< z6SQb+0+QLyvMkL#8pn%)j=DjKF?XaQaXgOs!S~S;Xbp7}HqHyaRLWnfziZjk>9mz4 zR=PBIyKfX#Vo% zU*nm00WL7bat8v&PNbM1Ghx7Z>_PmH@_&+k^8V9a=$xg=@F z-(+(F?|W3xD}xK9H-#I6dtAE&?a~VAReFzqK^Tr!8x&_-*(|z;+wSSWZ>aH8;x`+Q zEV{#shI!3~^KNFAr=y$xGc+B^AOZ`y#`vVJ1+uU^aKk~={e%o9p4x64e+ve6#+!2aYQc=jwZ|tc9 zvK#h@(3JRZVtPp@wBgi*AJ!z0^9@}2Mn9Q+{>|3;^Orh7jT>YjAM;HEW#e8zxiF17$%+zp~Pvw?QB|}uyrMcB} zv%(&Nl3=H2A>9kPmIr?4W~?F6U|bISub8oUPK&lf+{|@@KJ8^T*Q1DO<_-xS2&%nT zO{*h&MGpkA)9UC?{9Clqk}`2(_+|M~Ql-gu?WSY6F3oBdS+`OZ>!zovS52!{G$@YB z6NQ~j?6lp$vb39iZ&nsGdsRr&;qNmZ`mpBd4T?dwQ3@L?|7u0CH%Ot93U!M%ON0H& zE2p(+FEZ88w%(xVaa}3h0?<96iDMe+TOe$BU6~VDPdYuIBv;rU4#6C8O+Z(O@!!`W zZcr6V8|eYJqXct`NGP8Z_AfBq^i}{j-r2-K!w===Rs<)0&?|H%;**5i0c!&c| zTQ?)WTom>}l*we=4gMb$>2TteE81-3>ZQ2PDDnwrsWA7X%XI654O?+I?xqbQg-DrG zJ`lXu1s^+Au9&b3gAUhu9D*0 zrP<`(K=(6Ho_$@J3zE9s(u&|Dx91NE(m3Oj5`^a;SufiF zW_>-G@G~Z`S_jxDx<*rkyLKrfOrYDGqskx$M^hkkVoTIywi+FyxDO~&g<>^O9o4+B zCD(Z|!g0NeO9f)W=S{26L zL|sB)>B5{q)TK*iV*^%u3G#F(GKUZ7o&bhyeUkU=(2cAC_Y+t31( zCo(iSB0pHAFk67+acMq9fjWFh=By4eHcofYF_K()Hyt}po*xGy#vY(Ny2mU*x^2b=!1k1`y zEYKd|WZVaToi_NYCEuhKTE~sSnZYaOqyzx9J@j{y%vqHS;)KPMS4;54zty%6^NZ^z z>xkF4)_J{yKLsC(9nUV@iOq?XUHA%NJps0^Yg~aePi7L_GcHd{CAtqpo5Xpt9x+sp zXK8TpnVg-tZ6G)EhbIN+af|=z^|a@`Q3@V*L+E=Vhua-hH;Wzx_OQY%?*h^8%W`mP zSY7mdbXfLIF9znpT`>4QeH%!f(&^h+26oK-2p~8P2D}X1l%ap0@u#nv+>#X= zU->bqoCfOZBa7EgQ``xPdLNDudCngQ^{z?w~4Z zmccjNE9ef(q78f)$nMt(UT3sqpQ;3R9}v5rsl(@`squ-pD?wK>O%1(b;#Ec2Q;=Uthk z=cn(Gnh>>m>y=-T^-jE9+G7Uw?G%?vkQ*~ zPIjN_&|E{92NPJi+$L`aTA7gu;t$-BcRTKw7jUP%=l<-^y-oOdz}2rNXP?De4sR6Q82^E@dAC690&ZqQK-xI^5e z9zLxraYZrtbh>u7ao66UTJ{9p4*RB>KfFdUfcCxhn0H+W-cS@|0?hUA-7F%xPHa$W z%^*-paR(_!Gm5-IL&(Vpbc_uUNez`=DK zloCPsC7n*DszLM~sTk7KgYMPB3J~DM^Sd+^!P#)W0%CmYCsIgrLeTICAY)E2uamrt z&D76w{Me`65yrxGQw2r2D3NqaYh=B?Y3f{gobaYcv;=s<4g=G1s?zu;pBnv-2Z zjVwP8U2w5!XA|bE0S6o9>2m^+8e7*T()B?``=JZ&vm6HxBngVa%qtENZ=25~VrSxaL;7L$_2;SWOYgk)*-G)7q<2imJl zRiL}PU(}_+9UjoB7e;iD8Wki7BuFd;`axbbs!HsP4cp;^PtKKN&(1heX*cK|ty|U(>_FOxTYj9^7Gp_x$){u1iA!j# z&|`Lpy1i~IN~wyl49{hf9&*gP$)gYQNDV6Fcu0qes>%i1pseOn!~ki0<>JdXB0Ho9 z1sPGEIG!PI#eU(z5_If1DHYA=&;AUN+x(ziYeavez`j9!u1_8%dE9 zdn`cpGNK{&2*p)Wqzv{&hH{z~?WS3KyqmPk$XQvl7WcZje(Slt1gmO#l-Jn_kWt4UdmG*Z{LnqiYw;CJPVT3wUe5|o z_X>!%qyssF4l&fp7Ba`6U<&HDka%jf@^Y~K<>9<11a9Di6?>j*MBvh-%5tEb7h zH2wVG>m-Grr|Z0ClN>TLMY$B0O_7~cRO)PKH&C3HcBvA`N9>AeDgVB5S~{IY*FBpj zYf#qGH-j3)bqy zjX2^%AJ>OH{OOPHa~<_=>tESXs*arZTlj;9E4O}U!px0@@pH%xe)h$Q>lap;6%c<; zas3qOgQh6i#%J0@eQeo$4E17mDd#z;1y<`m1u?P}0a*~hO(0!Zs<|gL#dmMm9>{K| zXwE1rg4cN&`<(RR>u~8Nw-UK-7l{jA>t3VD7Th2iq7>hXU`&rzlHswF1Ohzxj1KYe zP?`u!`Jh(9xTW8vsb*pOPaLV zp`FY%?bfgU`M@h^Ptk|2pihgT+#Gh*eX2NFFBEphdlgHubtOXxxp&o|+h^fPKFOg7 zD3E(q!Gq$xHn}ay28D|LdEFr`TFg}}hqME}Q5<;H2*qr=MT_?}FI)*v-bv1R+@85l zfX#T>q0I}I7=8o8Ba9c61*{CMrI!pY%9I&?%jSjGv`dz>YMZszjQ0OJYDP7?;KT8X z&`kHfweDpT#`gL$w@4K~`{ulsn*abuAhLntPEzDJ71gOa=G_-wCjY}(AjsRcFgD<} z;(_pw!=uUUS$dNo7pg#4aM$LxXwNgf5ZGxDmj~yE54i1rCR4Uoo*$7RZ_#e`8To+h zu-ve+*-JGoT4cNUpErvQ3>evhK4`gXd-1Y7jyXr(d+|yns9wNPzx2NRiYIbp;1%!a zD!%?^;Y)EmcYBivKr2uieR$`$WO0j4kZStz@+?w1T0?^qFO28Sfb${6LE-QR#@ci} zJwQHHVyHPTxOy=xHO4|+OK&W5$5OF}E=a|I4aqry>pbp=?g72)mf5{TPwQv*v6y_v!m(K4`ssYj zh(=73qY*RWt#}<;H(3!uC;!LCgr)q4KR!q@oOlVWFvHD$irY((Jjiw}IArLdTK@dO z1%`U(gA4X7&>dDoeW;|z6*p+<^g1B-jRxA2TmFVC@X}}doVx^OQrGF zLeEq3Xwf)eU~O(D3Le&;Z|}Huc#N`@m~!!ZB>LAT$V`3ksxQ$$W#(s@nO}Z~;(96a z85MO~nx2|R z`=+V)_;2BK$5j>Zgwlm^!P!I~ww~KZuBr@?t0V3o2w#8wmb^<@4<3D*dPwyi{G#}D zG>|A>+N6H{b*#n2F-V!Y?^>Kirz$>_fc`D5!+N!92D80sYNNk3QCOTXcihC>VI7N8 zKHPl&MHBAkzj?NUl=JhXoOnTOH8X4{DegE$YN@Dwss!PXYL9=_>|}Q7bFf21OLjjX zCMX7a`Y_L`Gg4I-aX@6im3rq$-hU4B_>}?uP+)5)uN_hy^a9@<*2k)6x}h@^wtd(Q zf)6s@aheo_;XR08%*h(JpoU!@Zcn?yO5}avNPJ$b8h{c=YtH%@re|`HHpVYt<<#cH zx29M!2l8_5fyk4H{yXYdzmIz3Pp|!E$$!gNQrt3%L_3@*$;vQTLT<+Mh+ zY7z*I?L^6=f3u?wE}T5E)iOozwDWEy>3-uoZ4h29h`h>{Mcz@tW?h~ien7TUenfa$ zk|U~S?h)fi_uL*Dg^iao+dVQwEn34i-?dOxr5IA3nQe|9TH^4*|QS?Uo1cl zf`IiEX%z2>DizLW$HQgp#H*AQF5>|=3<_aiC%W54_oGdUp+78RZGu*FW7RzF#_U)= z?3-1-V6yLB_}}GwEq{h(1qYu+$BE%#r4Z$Gz#w~D3Y#oLSKj*ReS-CzQCD)(!0>3A zY(&MQ>nH1AW=74Ap7up1cG?O}3{$r7peWtDo*;W*575=(>onk=Jx27Zqk=sDJpWGb z2Y%N?y%69l3rq*juM*XO{EECm`M_^U74V?sOSwwN5PLvM{mBHACSK@itKa{AKjIY;uoRH?DaZ$B+ z#d+P0NDyLYUY`KMyDA8~Cix7_xyTIpB~H&I8^8lb#)642IOGY9QA~+(h~LCN`D*J{ z=H9!{{=t$;-D#JmA~S&OqBz*-q)}0q*&^oJfmr(gza{QP&XMZNX&-~YegMagN-g;3OJ3s3qB->P}TWLw^Q&0R@u zerY29tIcvy_b6_FB6p}LJ-sowMm`YI9Nbe>M{Bu%$22312?vSickV9Ye)3pN+#~I}c zVRmSOuvyt4jupWY8Lu25NF$daE1&t$WuIWXXBBj;rm5p(S3PyR-LW|gSI``~jscyU z*}B0+NrEe$PRoW|xa7(dOSUaOZjTdhzo1KW#G02#ad8w`OGRO#GCjyZ)3(#UR#pq8 zerG{_?@foBtH;(78<(`l?sN2j+m6qUJ!7&+*MHsiSF+lP%O!HmW;uo8V2RyGMPUN@ zl;%G4RjzeU01_?ivP23|Y>>eJ;j0cpF_SsxHqN$Vyp%B@W8HxnWwn}`a$oLc$#?T5 zu@_d@aj=8DOj8qn^lK;8gKW8|d3uj4L};QV>GU}l-P=@xA57 zspN&GNyOBqlSf=FgS?Li?VNbCY$dz6)hBvJg7DM~-KQblG&XCWQRXvc^OH2kq1@!- zz_o#C>Ne>;M%NW$>>JMsT+MEUo~(ON9*WimBb=pmmm4I%Ox=|V@Lvpq{d5tg!m;;b8zetwMMoX@8d$}`SXS`|x z_J^(^XT&J?d-HoGFCP9$$%_twfcYHR2MF_LIO3IcK#a2JOi6y}sG2o7B%2chJ|df( z*dZw~b4ao&4th2-sHmENLu@CM&)t~2X#sGPYj*)Tey-nA@8h8JgGu95W zvnC+kr$t-BTxCijru(^SkNl+iVTgmlVjob@csK$cUp$i+P^gtEL71fux6@h^D+1vy z>VsZby@qu69pXEpzHlAp7jr#7VzDtQRSeacoQXTFOw&s)N-s*+IiisFi%BpY*@=U!zUg?&sQQdS7+ zsCD#3=0j2w(5lD)y$gq<#x{&R=7>6YW#QT5Hswmgw11spazgZrLoSo8qve2{cxO{; z=9uiEILHfTQBm!li9zW;LoRo`Yi6bUAY(*#*h3fC*FwA%OZOqP$y9rvCN&DZqE%5q zJ_e~BQwdz3AUde}+e^QirtVhm08tEwA<13@4P(ktw#K@ZX`}c5?(Zv|jG{~+oU^DX z=+ulfbrXYB=IEcK%Z|@{Fw?jTu9;CFaTrJ*pJaZpdHjAyjm+=vyk+Tv@Cl+hv5~Rj zu%HSG=6AAyytR+bgMJ<4?VLm?Jeo?IF!AW>$vXhzC)I1F<>fn4CG zw5^I&lAEj!JBBI*9gxpU5aMt#P(eJd24)keh8H8p{xn0Z_4O$ve$}rm<8D^MTO9@+ z1nkC3jDXL#fZ3!hkEjg5JS?cN>FFcB$)H4wF9#I*c@_Q#SRINP9q?}(hb*z#Kf~$^ z`#*Q9kDQPb)<5@A!duR3%{N^6PW)_>O{#rqLn+xcjezF$h^%S}#X%cZ0SZQDd3Je~ zE^P7wrrmtX(9xbJ8&2t;o6+S}E*yLYIz`bXIHN?Vix0eZy7vVP1)!W7s9~4S#0Jkc zd5b`|)bEN9%4>8$q&-(|c;JLC7WTttwD0bOAja?zUa)cArQm>y^OKukqx##+O(dRQ zMT_%(Aqe#1k3fGq#ciWV3Kf+=mWS%;tDdn;F>N4c7!+(;1iPXTLv}I9G?(b z{%5N%ZXcvZvzs=K*@(TC^tIpjG8vtpou1N6Qb+TjotS9oJu}miM{!WuvWtq!(CAfV z&#il=nyIC)ctY#4XNCrA5A>>dpXDFkf9`cuh*UzTR#na1m{$Tk(*p_|%@0RP zp-N%pEL-Q-#*cgCW#fay$R|zgvGwKj#~lm3EH~g!TMj@^M=$DRi-Y&5b$U8iUE(|F zfm(15oM&sZW83^&yW4T&*5`i|FMLYffn@^MVQ2R%RqvAJUmCAvhnZBj#jZX>wad7d|uBp?<=6aSc zET9WLkGbBO*C1|Ieg+aLJ# zW?_S1K;c-c!ej0&odz3c(Z0)g84&7KPyGv)qzG2n1fT{;xDkrRu_9HM%ocjKK>{B& zq0`if(@VTjo&Jmx3skW>-Laqj7oQC8S@_~shaT~o1?Sy6Ui#UguZv9U%iFqh|4m}~ ztuIcz=K)Ib5xe3o6qihq4OCQR!20R?{ejPENR_Keo!+V_4?8Y2G;6eiRC=BSH!zb2 zZ@?%&ggX?}W)E*%w&$=VG#)RSefj{<$zs|4`iQNIbXgv~o>#nLB)QPflM{%qbV$5| zWNnk^Xz&Ao!wV1|fxxi8hSJo({U>TCJi%R%#v*j$;j&`>-ZNZ78d-$goup0wbRI_Um71{ zomnsG5XIf2$N&|EQUW#Mhd~z>bp!IC=6cRpMG2Y5WgEA{aMjpr9nKoP|w0<)U=@BdM|B_>}x%$U~RSbL#w&O1qmzGFxm!&ek0E z(_M!`OQiM2fl}{m3O3K;ZYj146^GxOcWo||AN07!2JBjVMBX6QWk@Jo1~tZs##CEjZmfQU^eyYZPqx=V9g5iN?P%>Uq0 zdf%f#i5E{~P4BpYV-BSW^Qm|NXG+9!#oM$AIQ>8HsV6Z`3>;uY9ub(?KymRDSw}@7 znfp$v9M*@-S(sk#WS0sr`@FvWf1x1LMo>)HWZU{6exOM4i@f-p2^8Pm^UL3mZB7h| z_su|&PjR^v$p*2ype)Tj5CW+7zQ~k|q9yn+D1QNP_v{kBD$7~Vjm zLvl@U2^s_NqK?o`&{tX}IqzSms->G=Yx$Qr>K%t{%xr{E;{IS@|kqHg?s(9W~EyWec!FobbLJ_2!>{R|37BGDkW4(MvaJH+?`Bgm7QGvXI3p6K}><9Ym`A$eiu ztG_obIWO(4nL$2wVh<(OEY>$jad#g#M&Zu=FUAWHB@bQN zfm0b1i?3fK6%c2u^-2(Ss6STk6`d68@BjjY!VaI;YXb7=4t0LGsp=eh9Z?Z22Z-JB zAq8I`0j9Thc7W`5M`nhbUZ%42T@YnV@=2%LH8-KsJxB8T8D$4toXn;KT$CPwU`>a5 zzkhq6?!w{|Ua=67=}2AqOARLJDRH!+1(5qMlbsUMZtmOUAyir#(7p7j z7tAzuk#nRj^5%wP$xe3SVO7mXah8>qD z{v5+R*>H-N1)GvGf$0hC5WCf<3Iwate42ub!}N+4 zExu~=?5;K1$B1LW}fuq&Q>MLB}P$YoSrUUiUqDHRA!UjMrHUoC$Pz&yjPcWWC3EG1Z7r4y| z*f+^rrCG3NKNPf1*r}RG9E}@0BTdfu;9_<+KkP*Oxqp)-b2Fd*Y$x8ITj6>Bgx&nY z{>8f%ujk5x2Z6w>pXpl^C#&>45>Ph3m_d!~BfeKd21o(eoo*n8->3RmgOMMM6;C97 z%qDI4g}tB}zTZ(lJvaUH9?LEQJ}`0KT18@|jbBe=XOf{rJRQ;rrh_KO5r0^nIvX)z z^?pa42=bc!tirVHNN-8sAPG-d8&ZMUQj|e)P>PdEMWr%TbO+4-p!*eI!#?=kCSaY~ z6*lPhA&7K7{K}bGt39&hd*lhiZu%0FJ{wc##goIXYaeQ<36K|#i039#+y;uoQ&D}Yt)aRMA+iHuC5xWEqkJ7X+;iv@-|NcT(!TjYg$VF5+D0Kxy~<<=)^4rE6rlS0~B|MBE3}9n{U3^ zNM{I})DJ@Lga8S9w=^U4jIXi6u1DNW=Tljl?Ve|p?}M5+k{w=xzIqI&C5BhfIt&|M zXL16u1tH0&R9yhNdK-fi!|%~+C7tXUUvOLKSSrcqJ+~Is%3wSPrRtwa*Fu>2!(BU)$6bKvFkak){G&8f(+Lp2A>Ui3Jd@q3tRQv5#KgL-QPo(CQzNm-wCr` zc|O??+`|xcnmUtQ2{m^7W5~}SHc}CcP5%%%WCz`_G!GLU_CU`Zhz@x*WtQl>|27_t zS0~>5SkZVz?*@}kSo^SfP7E+u>FHvFsD#6Jz>h63HlpIO`y4&+^>4QQHpFC*GJSS! zA_Y$D+nzAQt^lnWi3eiw(r;%}(~Dy2?;l{Kehf6Xh%0 z2nin6U}C_+%L%4J7QRv-Gyy0u(*GiQ+a=Bmwjb$$G9udp&FeSthz7 zujDqnt&p{73$X%eL3!{dw@Ln1+Wz@Z%;#$BQ}})UPTLZ6?l_hGwD;kobATO@#&ran zRZ{U+;ssMGf4Aa7p`Gc!4wTRq4hm^pXscC?Ak`nJe$ObwV)3!tSF)Ly#8tuA{ z7lfz2tNf9-)?`BTq0gm~!%n;gePm|a>M5>{A~jT0k!J;s`mE?iB3ml-N|r9{pg$0F zs^0~%;IM732DMgHr&}TX6#9yg9rG%H`*LKSx^K7xi8PH@#7KI@d%f|xEDd&sAY5%HhccVpaiHljcv%}TdqW#v@Iup+ z{Q3X$;crZ63i$1_&ys6TSwMHW*+#0L;`%7kLq*+X;{`pl>Qv28k+4oUK=MPkhHsvG zR8Z%&#j8BHRe7I11X}hj+MQ%cu{Qd-M!H$Kq(zHWG4WnU1#!VwRQI7$EBU!f&s+2k zx78EUdRn$*$ri6oGqPS-vQ6F2 z>1X8OdAw1X9tyQ$^u)oNlY#%G;J~F$tMGh)<9}) z9Pw%$j4c&kc~v&-N{ctOlep1<=fvBnd^6LQMsZsyvW1E|=oJTBCluadn85w$RqsbzhC=f)MH|@-^@s2{=TI6_V^cs#Q=qb zAtboo2WqbNNOo|0{PPtRf;>t2%*60K*#SSjB8zU6rUXpb&gu#4iXEoq2~V7u)4Si; z@XFI(4V)!eC%EZhWJ1#AN;>FL`As%gf|`q2biJfrqU#WKh#cz`o6QEl$u_&yp)ahj zBcr66rvCMh!80wjOui&(1uH5i>&5HE2qUcgZB?{MyWMeh!M4dX_06EIp$)2Kvr(j= zlP#ZFG&=z*f`^H!D`=!h+wFdlF>+5h4s86e^_W-TFnndaTJ87um4q_VA$GE`@yY=i zoLw3NQ5!Bc#ZsihFNfVyD{EY29K9)jG_CyM@#t2E9>k?p6n)#|s1{GZ{~7__5vT1s z(VID?=P9mRNxRu|FCL0z4+pY0``%x=1DcIB)=HkIRNcKBoEF>_+;055FuTBfKs!p}!r27|vO z%aoOvHtvNmI;d+Cv_XR-zSMn4Q2u0?50W1acumgZ7xRJ9<99o5B>B~6FZr5GPXEu8 zE6ENgHaS&hCZ~wvK=?DCis}#-araCHBAvdjybw?bg|`*b+kc;7!ukls62wx=;<0}UMgbxg?n~jNw13zo>i<47azhMH+v0q;N zgdBBZ&~%uAritRtP~;R9wHg{i_C;Q0PeVt@{YBj~u1MQM5(4T;|Ge!U8KM^L3XcaN zLm)lAhIELJz_o|wexj-;2V{-@Ma+<@MQePOA*`co<$FZ~T*ZtG_gm~?|J`9FsvG16 zxwc@1#|4P`tyDz^t?=j*Y;)B|7@v2H>=SfLKm;KmP4u&R#?euXrS7N>LKmVh2JsC7`>Wr|!x`?N!o z!wk4Bh05O+Ew%*YGTHP};U|#B?hAp0RK#|Fbar~_%vtCY=Z6;qINI3@i}9Wurhnci z5XZy&pr=V( z;1+v|T#wk-5K9a5vKwPck zw)t1ln`KCUzCIu~Y^R`7(Mu!AIU2WW;A%@#uj8^jQviYRMj#f4=MZgcm0R3xz||-- zJPA-6*BAc9eF)T5INH)bmReA;%q2CF?+VeE^#^8<+9cIX} zB5yd$uEXLigENhLcuXoJvtGWC z@bc-NAB~4()`=5TR!C-7dF5!Zso}6ckn5q6;r~W|la2VX3jpK1>EC^GHqA7d4$b=2 zm1O&94v!O;7DJieh@O_c6bGf~IaJi`nZ@2%?qJXy+$JucdCWUbIOqmE8l<(bgu)TVWqek}=m_H$a zasaNr{^Hkm``q74FB#aZQFR@;|Q#sPT8i3WO6i zZoG^-+v6YPg%_$bvGUi%WMYb3GES0J{C3q&?ArpF^avkkGsPuQBoQcsC3+e;q@nbx zNsZ#PM)4_KrL+Lrl`ui>7_Y_J%<{sKwTC!h{Ifs&`{Jj4dmJl|m1AfwojxnmWx>Wc zgK1}tr8L`T?FW|wnk;vD-_G7^hjPNWU6q}PcEoUCPZ8RS{cc7VlM@l%tjc6LW%>jhdor(0n*1}EqFKG791Uv zCkMF$S<-?ub=mwuNCh;~0}x}qBs)NcR5)HCxG!%AP7wByRG$`YDMVu%=|teS%NJ~( zlP4?kT+YP%T#4AmKOJDhl#UNG4scTK!3_PCIGB6O3}II( z?g~ZPsHjulxGewcRklT&;*WLS71Bc00lzBnXi#l(gU21HG%r!%I9?Gg0d_nVuh`&W zJZ{l8xDAA0ZN_Oyj;I=v1em|7rLTMHy^85BZ3b7bC}NtqYYSSm%Y^kztVa>EPl&IZ zw6|DjBWIeyS6U_Y5WvlD`uM{yQs-s@{;dOKoJhgakgX+-Y;jw*?B`bPz zI$=UE`Ivos^1ze6E<0lLU$zv(09vsT2H=tA#64yY1TgXyTYcVsAWRVc?E&mV6Ubq| z&iSkTTC^yXvrknX1_C2gTjZN{ksue!B>6q<9K8ob)bK2Hvc@(WL!DS?snXZbuX-nw#0|iS`I5Sjxqm( z3s1?RujB&P$L7gU9qY8L7+T31Uw~j-GFwHTmf?$dpXCt5(@n~x^PX{-cgM~17fz# z_+l1JM>c@uv3n+Cz&c2Nn)^<{-xswis_G<=c&O$@BCF!?UU_n8pCCFYNs!{(5U@^& z-Pq@#ZMKQ&4yjfgf#R0WL4PuX>0{%by!jaKWZD3rQFkKaecvbD|NO&!l`op?#{4(W zc98O?EXUkxW;ae!+;NK30*iduDq${DYtTIb_Rd6`=^$)tbB6N zPVC28A#AM=ELylX(D-+gwpp9vdcY48ia6S$9TWftY2(@5bhhcpE-;KX5o3+n1D>&- zH&N$K-+gA~)4vIh2O&S;$TNxsUx{c0`+y~I$5j^@qot@^FtOzWV}ATR=X%U%P2@@I zhGG=&bm||AzWIGisynCM~lu| zju$Sd9p2v%TgKhxMEJX(|03#lul;7pf6Hmkg-}lHP*_P4)IvKEq-O#e#7Biq!A;t} zS8hbM|M&d`9s&N6hww zA6kdciK&gZgH5}f)ycj4$=)wbA`qH5Mr4}GDGpR;4?)R3LB*MJ(Ki400;)C89t7!T ztlLjhubS4ax;{U4TDfS|v{XgDAa>g3Id%R<7~KqN<#Gj|h20Tn%cCV&L|1O8l!W9o zXk7OyuLION#Kx;?m^$dfe*4e+B8wwx6cYou@gQpyu;I{zpv>)M&3{<-M_XwuzA-Oa zcwTx?6vw10(w;3}ycei4F|)r@Qz5>#pg}q@AC%4iD%(F%n;$tL7h3HrP|C?6jR)OitMGLT13@!CL~;-(DB>tK(f%RjgRODj>B7Y z3@H=efo9fKUge@L>5^0O4ru#>)ez|v3(?7Rm^l{lF_ z7})I7NVoV*IM{86vlVk-8lDggk9Y__IJ<0oOBZQE&39{lxSy0cF={TDq2>g|eMpgG zRMasjAHkaGIOb}|!;o`dh2#x1er*$UsxsYlpw|hUs`sF&Rc9QDmGro#sgD4E{PF1< zW+TZiSU_Gb9 z;dHtQcwSceY=dfqJSdXG5bXo6Mk(BV%)8fftUA+@_`*tQ+-EuN?tRgK%&ENOAHt!)Cn#Z6%!Qqg<9N| zKACQPY!_mti^UePjUI?LH2^YRPf7ptgU9guFP9vwyCZd|+pkM1S6}8;$hIob7CAon{G@0#W zjs)O_w8g6}SZ55C_Im10xmE;cgeJ3h*;U?c!E0G)9)Ax!&V7Q7v$N?60SLR`<-nfp zimf^gs(LW>Eu0RY4dihZ!4-l*kB_BT?0TM51~jN{8T4rGjW4%r=a+d*hhX?5{|Wly z4qj8`yk#FvE;nCsHDRdsD^>53GoNS&QdZ}&ve6Vzx+mkqk1UfW&& z6<*zRw7nH7Hq6bam+%47s1I<^;J-}&d8}n|!j}XZR_gh$ctS@?I8ylZ%28!pFa7|c zdlhsYxKwHC9(w=y94OxN!~+?;pFN%%8vkvFHc`r|+N#Qu(}3-wxN61>mV`g6KsZ z^!`AHHaMHl4qtSb-|e_pbC&r3Gi^dg{||iXNz9i99T{fm*g$dd6j_G?G@UFeu2q4{ zktf+c>-FY^=b;xqJ!mqFg!x3;3JUYP9RKQ_kN;_=jfv$W{^>rb#*7Xe1iO1?oK`{( zC44D2D zJqWUj$!x0v^~qKVE2KvPbVahV*(mi}&vmMCSeFeE)e1o(FgC68D3+GX`#`q7SbCAE z0k+9(I@Uop!wzg)T8|0GH9pWfw!gO}u)~rtl27o=c`HC6y*+a6c|>wtAw_wqyNYFy zeW~YS+*`D$)KC;uF6tw(e%0Q{HM-RYuj+@I%Vj|*ZKz1tanJw6)Ym_T zBI}l9v|6Rru3CX*hZraqjBp8)DQ*Kr;;E=ura`&i|B&G7OBHnI{D*#3iYhvB`i*(1 zvpXVfk~DejR~^rQ%^&}G-o%c3d42rOYn$c0ueMDw?RYdh54}YmxbXUyId6CfDj-L~ zdp*Uip~xyK%27|l9*@0ej}0zy|5YU{qY#_+sU^($v^q3cWsqi=r1{Mi5UFOz-t|aAD+l zK*bj*e6IiND+Rw%nyk$6fBw!TvVmU|)_J`x0_}t&Y|;*jOQ*;-D(aT{DtjHWHy_U* zWUor&g3D$eRGd(!PCqC@y~ER*2IYYl?~?4G98uZ)hcL1)5?+PXPDYsDK9KO4Zu_o1 zzh(AEmK|Mu92Fbny)kWt)56^Z_Jw%KRO#1C?JCJT6=K%_MZ}-#y&+<80#+C z3mT=r$*KgV|1FVZe`&1BQ8TM@km8CduryJ--LdGoIr7cRFAkS6w`g~}V{voU_pU_7 z0n7M4K^2|uv(tUUSDOs4vm`s+Z@d=AG--#O;vP`;|1DEuj9DuD%3>>Rqa9ZjJPfc@sL>)+d2K46HwEBF|WNb-$I|*YNjM z5RJ=+A@y@cIp~fx(3xZdkT;Po<9T>xROwP~u9<;hzhS{E~+_05k znz8TIYO!eK4!~2O5H(`$FE;%)O(6!qJ#7^@pzK3U^@neozTg_fb|8XN%_@i zOS56+`_>N&b6U)#)1vJd3?Z|*{ zp#ugS7XPUY~di5X13@jWk zV>acd@}JL|Y6@&DUlmkC9*!o6;=DGGY&Q#R#8cclimaxh@au0Vv@P^(6K!+72ZckE zX$c;!G`1{xk1n3X1M4VR`>Wa(%POJAwK4gG4P;4AU zVyMVgx;&sHFjnw6mbN7@S)qr5T4wKpxBb6h(rtaw9Ph(nm$_f`Fl$@qwcVet{@x#~ zQSs%u$BDOXIUw8uG_P1A(I~GMVg(QobV7!6@b3SE@RMHzH-ukG`IqkBnP;BY_Eb(L zpE+^HS=9KUMy!`&?^C3Uifk6In3GBmRK_GQxj`KuP<|UG6I}{P5~S1p0R~yLvO=_r zUhQ!KQiXQW_f(zCZD0*fR-=<*aLpE6^J)njDC@0T65L75O_lNT`S28>P*_P@PU^yKfgBQ|m4ZYC$6|Q;woN#q>$`B2aRjo@AWUB-Np? z0!qbI2I9ZEyUb=`jjK`cAGX>`UQ>}B8gm0}I9s_PV(*LA)O&n_Do$*yIEeb7mRN&S z{8(RJsl&&(6 ze10iL&g-0!It%~2ieisaq|}s5Jy66|tAu1~Pb>{Bazh%dfvBt@v{4!p3dGZAZ;W#Zn8+=aL z&7?tMp#f<`m=doxFzXNKwcfixh#*JUO_GK6+Gg=pI!{~Vc0+EEZ*|@3Ss1pO*)S24 zRQIX6go`AbK!~_5q+faf)S|A+Z_g>5y4I~r(aW^ji!H-eBVqeY@V{UroY%LkYg?oD zG6SgY?fiu#(TP{`2Q84DL9wtixs!_AB-pHo0%e2>Ae1gu^r@<49dz!*0~6D|v%(A9 z3*@onj@QO%_`ft^op!DFX~`$_9j|lRWVNYPio7VWL*>ZnZ&V<*4_=4uh1X0+h393f ztv9WosaPP#Mm93IA>U`FEOC0PCLPKJ@d`M|=pAKB43L*Iwt;P4a2a-B9dnTzE^Xfp zp6qEhBR?wJc7begVl%SW!i?;sSRiuSMn$FvbW1l1H%vPqt=Fc6B~H5{t_ZCaUt#dv zD9f1_#qN6{S$#m74d+Gf4`;@bF1L-sT5#FMGeF)$d?D>ovBiQG~1Qa?H} z(TtKC|JnFA63uTV?!40|Nw>gMBE`bOa03--5MXWTRe6EN7^bJAB=^EQUg-m2!gb+M zl8#r_dbCWvF{f~jBd4ArgEer-3rW@wJMet|$yw3XL`xhDMU{c=UUk7sL3m;ps6%X$ zEuXM%a@piO?-djFs%{5h23n1Bi|^KXd%fZ$$LPW-D<_;*+zN`CkgVS0c2Bj06isbs zpFHEJm(S}(Kl#Z%Idag;=J%6zS{}Eb9QZ(P&Ev*r+3m!(goEMDICqP>LlrCAq1_{| z28zeCepRX-vMCtr1y>7pE6&mfqz&wL*~OqF!3D+Y$(z0VTsHb`R_Og(#73r6v}RHl zUHWp8EStXQxE##UqTWHEANT#Xhu{0IdFr`2r|ffb+=(5OR*T8xQ;I!Lkx!_|41qy} zTR}#7ug5Y!18e~;6EwQ%Ow>93E_DJ-Yr$PYv+iU%eJ>SNo7Wu<$uCd_ zmMHz2;^RN6yMdDz0RF-;>YjSF++;vq$fp8mwEPqhvLcKt35ii1;D37Q9}QRD3|hy0Wwuo)Lm z>Bru;dEsKb?rW)xwY%%IxZrSLFS3g#B?i`rSDAIV_A*6my*5*5WU%x+|5aVXf=f`N z?kFc}m;tmI8v`a`m<}9ti5E7g6E7Y-@%NR)fRZa{V{VCbMo7F`1XIRVMXsCfux3w~ zVd5A92*=-c5IG1f9Nk_xeDoF8?54Os=)OocIkB5^z`{*Qr&!2S*iJ=a)GRTeLfGe0 zI4|EVVPd}v{@2OGfb4OvN1eJ(y@wbva};-7km)>uLAh(>NkM&tg$p}XTS6U3vSj<% z5ARfP4BeH8Ql7JqCJ^{pkOQ>ti zB1u|6It^Uai{KAqo8mU!-I`XMI^+qVX`y;LE!5r)joqLzQVeiB z$8CP1f2G=PofzY^%O8$Z7u-&+Ro1Cd2`+)T3q?VBvuae^UhGgpLWJA8a6O&srK#ZKq9j;Rf0; zTTSEt(6_#Cg4tqeTi(@?#m|_<0t(6@rB~}HHkKl*smOg^81pm9N$Cu(H^3drl#F7= z;rLbCaM?D4WJ5=GZo-X|KmGP68b328{*Y6#mh9x`4>|FcL74?w@+cN4P4`fd388hK z$5g1-f%PGKneDPP;Z0b270?$1M*@~lE|#`Q6Ula20d-qwR28{J1)daC`4mvq@(Uu| zNlR8kMvZu%SEel97qi4b7X$?H^4;zedrxnCmgK{EHjl5_8BOEHWw$kUS!caGCmT&1 z`7mD{^IQA6uI&3`8u8;1tpOO=9RS=N)}G$=#~+yuip=|W<4DbOV^Ho`7?i6N+end% zRAjF3lAuLguf^(ubVa;xh9HeD6{XKapLvUKzT%=kx*-6CBbMlA^(Sf9wZWXxGhmJ`c|eYgL``uGj8|{j4M?%-%OQdSbE~ z%~&JQFPCaC9MCH6Aobd_!3}Jq8~QV*kN|uS9V@~#fMj*EN3GB%Q)Sz1_SnkU<$xPl zPwx1;{PQepLCj|==Z-@pvsRg;jt|l8^IGd$ENq5i+()5UIdX}yW!vz{>#@g?twV2* z@8Z)3w<9y<&4L(Tv)#$+|7kukOd!P;KiFJ~-9wQ~Dzb_B0z{i)1$X?v@U9SH;1z$m z1)6nJ;zS3?aY@(oGMv>_`sqpHK2~LNWk?uJ;+nS?}j~C;_u^kSoy}e%hf$h`?dErGB&j z7pK+unzkZPtvMaaux*+cY{kh_+wlPSQ~R=KgW3tXZ$9&ib<9dpQJ}#US2w#njs%oU z%TnxDA@@tYcFp{zkkyRAb(5@F{82G|jX5Yx1ZkKL)(Yyqi-hsM*OYxORj=-#zwo{n zhG&jK1&{F`-p1K;(1-NJ&%kYVKlz@uG8w0&myubfO!k1VdZuLCG@zgT-_Z23?X?ei zA7gaL>$N+t_5JywD{n~5uE@K26aPZiOaQlIXy|Gy#U|ps7?~+cU>c=SbFh~9npfAH zSt&*ytct(=nvAW0|%WxC-4Xntns%l zFZ=TH#d+N(!clT|`zKfA9YE~7(eJRZlo**0bXQ)zv|zjMa!8KK4!k075*lPngAVCa8vt$oDY6$zT-gGT+pCo?%3cF3U*XtX*Fr zM`zr2{h5lBsNZ|d$GWWOnQ~Y-iieW?Q-$e*O|lAMmqIV^BzMuTiytg2jF?F<0ejVHsgblRj14W;VTYLSu04wzH`if~S#Sd#Je> z(v@8gxv{oF>d!=47wD!i$3R_A?%>9eEo`iF{E4;fX3Wj|K3_%q+AYURfs;YdjDCz&QZCz2uQMhH&cEW^KOivxiaK$9V1P0^Zy7OUU zyPw7f7xo_euj=nP!G+@MZzlYCFyTP)iwM#_fy7$S+b$Z;-_?-{e`_DO z(iwP3uvvJ*w@3uhjY$_@PLppF9|+hCJNjRE@ARnG*0>*3EOpx^e#q#jMypMC(q|Tl zTD7LX0OyMH;q}^FSiNkSd?KQkE}32*o(9h^3yOw(+j!qr?Q%^tu$kiDn*rID_&1+Y zmqKziRtc3(U+P^Sem4A$f0<8qSiSaR>0Yk{2HWOFM5{r@On7{5R(P*zy5R%C!Rf%! zp9oD}5Om&HMs?3j$=_Gun#saqv2BK+=1^G*1N* z-@ym`s_B&z@aoDb$RtuDxXvVrQ3oE|mj_-V#lff}aZ23*E^xxH-#h%xI?r8@PEn`C z2HUCrAMg5x{=2xjk*2|BS;AE+qW}{tHqAH|fP&zQf+|FDvX!2FE@^bfoFcax>4sqZ z!nP4qH`xl4Q5rtp$l_-NkFBScT5m2n?dpnS*Q?FKxy zDYSt;rXHIqhc`m(9H`|&7C(e6Ps#nWbu`au4G2f%4%H?WNwAc z$?D~v(cqA`Ntb~p(kjn-^xd>BY-StvSf}>_9qk!7K-r8nOkQCh4FT@<9|8MVV?|6PN?!=B${A zE%UL~x-f7>C?4y!gw~UF5$SY>=0I4ds$SDKXCdC-#}?D^|WMYlD$=N!JzL9o|6- zHNaEe6SWt}B~-I6gBXCHV&yk55b-zjI0+%J2WSiAGac&S_Cj!MnmYxI_@za}vg z$ZiWKaT~=ZQDhSp+5HbDMhw`F6Q{gfBuWjv!o&tQ!}2pWxJ$TnO0v33Q6bzqy@1(F zV&{*^FA+wE7e3GgbB?HeZ}7s0^BU(rH2m}z)6Do-_U`00MDN6_n<5L~=1?poRb)_+ zv9e=ytRM|awf0OpDom#zy07)E5a||4Fj|`_X?IOvP+|rZWKNSJw`7lA4;?D6HA)`{ zbQJ{ew}m|NUhbV023ubyr5Z;RJM)IJh4cFO#J)JjpBOLiGrzL?4<2U3{Hl?>L)Hyf z^X|NeCqT+EWR1L?Vv{Mdg^IiYc{ugjgwS2C7wD|;(ui#VOM~h?&nXkBOM!J^bz0MB zvO3BCF1;-vR)km2X~)70jtDn)N6D8v#ULCw6PM7u`Ug0 z)$9|LixL=+%KZ7|U(_kf1NA?B^gS#XuJlAsma(u3PX>#}SU-7a434<%!!Ry1b=~_1 zm#lfjzMK>1yq-4UU<<2+Rn{%9x+QibX8$%eN$i$ zdT6tRft_h1b7@u{(k7?V*SzxP=d%eD&rVq?T^Ig<2EEn?K%Cp_o)>P?5HYpvciHNB zl4Qo1$YIEhu6N(c_)YNOnkS(nm+aw}xb4IkJ89v& z6j3aMa}QFHDVoFn$k3$g1k%E4;6AwRjyuVR-0Pt%TUQ*g14uE=C$-80B!)$%?P6#V zKQ{w18+5G@noEJ2YtW-XnkfwLVjKV%RD+F766lc(kRaYrW(X<>wnYl=(c3r&A#3E2 zGCx`3q5JRIX8gQcq%9*^&kcS~Sm39SVhbpekN82J_L8|kcHsu3U--0wC zF2pQCU97BDi4S5Dj}8O8h)d+!Hu-~~=#w4R=d!~$J1pP3MFAdQ6UhXQlTCufk~G>l2BLB! z9Ey>=U0%=d8Vcvl+32iV_NH|<3P&18CsQ~lRnzNnSW_%6L1_id{?Abj!UhC4>b#4* z)8`=pFmAiV3fiQQ<*Hd1Ue6A+O$UPWX7M(!1Yy*Cl>c#vKQKbPSo}bF0l~(`&k0_> zCUje6P2|MKops`kAy^9!@pE=l>@Etz{gHW)JPBH?5q;CU6lXnRW=2D$VzanI6eU;< zyNP|GV-am2asEJfA|hsHR(J~3$)Ba~(#X1SReq9`d+hY+pVKlsT6utV2;K+aH%idk z2x0!sR{r=$)2(T3oi^OhK~#$}4_hZ!fx$DWJG4t{y&+827r1D0I=y^`Zj@>a{9h`+ z!RFupNT1+m;r{EjpC7i4(sG11ah_=wNL-@B z03~jx>{w*c5WuzZde#TIm1T;8`MQ^C)slpHjV`Nw{|bGFCz? zt5i`6Rjzmor4FYo8=2oY^wsejCX4y!D^G@g$4HKO4U_YRss731&ugqRo1E6PaHL$N zK#@+An@$(^jbt^-A{4ph2$7%`+4m9ZJ3aOau6dz9e#`UX%gO(jl`r|tpMUwM$hUv{ z%gDdH{kuQ@Hd0QrE`*|nTTg%Ex4-(DBf(N(o_CVg`n*nCEWJ;zYya!g59;2idadQvdv9!h{pN!I!n&#iCP|y=hqb4; zG%fSV@B{Avmao=*RLkCd{P=(UsakdrcJ+446Q{45s2_N0=e$PKog z7fYJO8z#1j>(nu<9t39)?#6GNe-)_K!>eUQZuf<`Y;iybx$ACZ?gr>OX09gsDoTV)3Kx zrqANZts^cDwBJB*I6uR_=BHnLPOQ)iGgkbMb>7r-Vud)etWd=Yb@8gcUH3-XH`|oG z%zmlSHIZEU!SWw;{Zp+t&u8RQ=F?&GXZnrBYv%ECpO^;+zSP04pA+kG`{#?@ zpY`p=YmlG4NRpw^%T7ub`!eHlHX+2Sj)fg@i?3}a>S)#lNgQ_#t%z{_==5?9sx21i$u!A%#M%Ey zZyf!LSGUN#=$V=L!((KH6Q?kMjQo(@o{bc{fg*7zX4vo77Evub;09zFldF8<=dPG@ zKz2Xkq|`X4T=a?8m{^b}17Q15KY3;>j(FJ;>e`BzZivmmc=f`jqh!@^HDgX3DS@59 zAz(?O*i96PM=F7 zT^PO(M_lCw%$tkPBrh<7<9n4q$t8tOY&9-g0O4bbJx-A-D)LA`GAJQ$ny~~@Xg&+s z7l2Wmd+zTvi}SvHMZPrX^cz>?M_jW)fdN_FCcQJI%X{PG3Sjs~t`&ozd-nEeIYF6_ zA%z=Q&Ej)Vh<$KYR%o&KR!|MdAXWxuK?(L5$v#N!$LF%bZ_cg|W%}iac6l6v&FG~; zCY)MIYNS+Obgf*y+TMGpsP}+E)z5;R(q68YYfH*a5Dq@EOP4#FPaL1 zN)@3MBT<@nIGzTO#vfF?@H8&FhyHUt$g61p=lf7?wW`ZRd zd-MPtYyr@ip=QVz^}YB5xwWv6)7l9Rsh~jy_;_H^Fod?zM>H{PvU**33!UVD>pO9x zb3ZJYhvlY_>8Xwr>FFqPP1Oc7Rh*!P-XmNVba0+yAz{ndbGRQvM#}z&yo|>EGvBF)JInmeON)i4!{E+V&l>Snae{EAcQgaf)ob_pAMi7s*wY;Gz!QfJ8gki= z;Mx_n{~Ok*aoUN8BjvWbU047W`F*2n*PS!Is8=%@Vl491_$mq1Zx-6i|`NMXhwPIEFn* zAS^fkGo_IkAg0yR9m*sHUii%Y+`LR#l}{}2Jn8VlR`I!cC4uRhyzumyYuw^Qx(a!m zs8-xUcS4zyp5ErAOH%YQ*fLYLNb(tMFUl3e-Q9I!d&1%FqFiGSX=858sRl{uE)Z$RfN(CZk|DP8ax7 z#-r1^wH!H&E4@!j(}c(9ccYoJuwMDo;5pW**K6;S`^+xcJ7a_j97ul1u;b7_r7K>@ zpt8pQ=-_y>-B`9VsFL)%@an*cGgG!%SedmHyM`jsRAhq!!zj8+)!C`}svRsQ=Z_(3 zo(!HX_<7QL?2c-F-or9xL~L3Y_$k@w#4`zKZ4Yrf(5X%6TN$e7b7$@n?m`~Gc;EgPGSu;^yzqEcIvq2q-vz_*P_am7N*p1I@OTVu z8T!HL01% zS1)~y=@1V+uG1oAY5$ui|K;KWBu%--6iw@%T`&*#hwr+#h_l10d@d;JwR-xI{DPo^ zJ{y7Byzi|MRry2()`j98@yQ7N^bWCo4biAzW{EqyV(s&zzo`o{1Lwx~Z|x%){D9-c z%iwAY;1p17K1K4V$YSX^*!#g;11NT{1tZg}w!~8IzHY`!HSE^%B3S>HZ5# zw>;}Qb0!Xbv>Zkz_f=46hwP<8)38+W99t}1>wD!VWeczfu~c<2aHssl>#<|BB{%|v z*cAy}o$*2lrP_O|?O89`xJxe~-M}>bu`t)KlP&S;Av?)E)oD84?VZHw4O0yA{le&3 z_b0C>E9TTl<2}3o4E6(HwF=pqE+eRw3Hvbxef>?IscD9&$DyejbGSrOCeC zvvqagXJ2C;g!Q@`WDn)ZzNSV`{uT>t_>bDvv^(hKj8xm`+qxkkJEm9uEXDK zimG2*UqDiv*yjZnm?6exKgEI*oQ>5Tm?(|Ovar}Oaeozr?a-|C1uF9fMI&$<0vTq8 zzzAviZPLzgurG4(eCZZ{U7pIQ)-CmD3o#LD)_P(&Sz%ydV84qglxOPNo=D*Xgr^z< z?s4UPhnI^p{_=^2kDhfWxCJUwAJF$zM&`P(m`)=N!5#91P!x;FpLa{NA@~!Y^&V+- z9wbg~rOv>b$$ksquO6HcAmpn%w9{Q)2pRWb)X^{TmGYT!oH+Q;LAz>VSjE)b?6B_e z0htffkkNIjdOiDHkd!rn*-JMGI+S(cN4>#ULC3_%REBSn+zDw6#;FR;RQH^91LYtN zg4+nt^(8Z8CqP+r^LE0&UY=>5h(7qxSV(p`@kCT*F%cc2Sjf}fH#iY(^r(TP>`W*! z#9-P=QL(f<5T!M0mB`SvS#c{Q9W*;yL7Zivz9&;VP~bXW}O$WQp0nxFm0&b<@O(7ExSjejH2P7IxN3+ND)=QhL99iR7~U5d_e-`EPQs5x0Bk!4c)YQ#Vk*G1yu!GqJ3e(yK_j zSB3P+wSLDc%n5^C}f8;m#4Wo8uY=jnXi|06KeUi~4x29WvP8LTT zub}n@&e6hk-+2m*Hi^L&-&&O z5@Rf)t}#$47TT&=Ic2A2hv)UVdEWKf%Mm+Cvai9dch+Wse)>vH38@fX3>>uuu{r8o5 z`#9u=@&A19x-p(j9*>7DhGQD~P8^Qq$mz%JKO=*Id>yKcnDlHj zCH*dl5R(X^qs@cG_}YSuK^N8HK?k4A$?(|y_q8?L#9VwiYr*FWbeH9~#SK48k+%?C zky|fhJZ8WWC|%R$@FG%+A$3+@9J?TVb<*Hk_r(kEl6WT$tM0Q{*QQYHc8VlZk^Qe> z%`ozV=FzRH{Ge{e#LC|+#%m|(5>Xqye1d^Z)ad1XE{QYqCN>F>t>%_qljpxzRpVNMLjXAvFo~m9O z8+m!wQmriJQS=vq&s zxnorD}?S@o}Is@DfR2 z%BMF}b7tn>Qa;?Xc#IPt{7!Dz7^=x2+?cc#>90)W0%bh9F5gR7%h2wJ|Ls#czP-jc=4#w7c2y&+^pMY)JStGk*LXq3@31wkl zu-Dy-rDcLN;Wp0|6AI`@!ZM&i?wWpmu6<^y(I99LNmieB0o1XSxX(uuDsx^VN;pDg zxY9);Yh6@mJ&@uw2H!O$B9#WVz8EJ$>B2bI2B5*dB*2W6D%TbT=14o(zzmG#2r*(8 z9DG$LDeSdV*8ECNOJBgjm;^~=q7J%gYU_(TJl|MG6t&~hTU|=?YWT!=zI&OhcjDD> zuElD22gT|svXzRgQ!fvE6k>{f=}hEFSl*CEXUY8HSRzSME<~XM%pJsZQq)H7 z6&3&qL4)FOn65)y;uR}!NW^M{L*j^u++s}m4T;m{7agzie8W27l_Rh8Fr7_0lwCp; zNW!$hWc7xLZBm`7(l=S1B;S4tlQ0tb@fq7%(dHF)}xlRj&3~$9x%DqXzPqyR^#{UO*oIM)#X}&$_@F9kV_8Fyrt|a*SU+vJ*#Iu3LCt^%Q%KB4??{?iWFHP#g{Uh{%I-*{?#l zR8Xp^gjDTRO>{)=+(T3AwO8d?bee2s=!W1_x>>yHTkoxXeeJ6&CtR2Bm>M-&Ya2tuZfQzTtbv7qVOtzT5*vJqY?*AkS>p}F$CK-b4r~VN&eos4 z^{j6qViC^g((Z697u3T_X)hF_>l)oQ&O{U?Fg4<3Ff(BSipfVeGFx3AKt(Cko>m9O z``SK+How%N4M#rkLwm9RF)s@={(5}ihtGP}W2|gZP|MWq)2@d0xh(UG39S?uLW|rk zfN3z2c3_>&p42JS(_7WO?!`1R6+0LhgG0(6AcIdj_LCPtsM5Ss=Vq9vC>NRLAW400 zrl_M9Q`7;9gM$)3_rMtW{0&jQ~O{<=JlC29XlvW7qv?Z_ujVbpv z%4&K&>7`p#dO96kqD8OvkX`eR0pQ|XchPxtFU(I>ByXS%@PujFl(Dk)p7sVGf0S{| zJ9Y+LOy>Ep31-}ATi(@?#r$yN#M!bC(;LEuu#RG5DYBZ1G`QX%d5SB3r` z9)4c>ASm`jFQC?Yg?219O)UQ(x9NT2&&u*KhoAAX4Lv7S%H4`&jIr7hb2XQ>SQ<%K+GuTIf3Pc#_qz z0?gBo_pMX6(eJODToeY$`U3)@Dp zcwe^hcU9W%Q~LKML0GFaDbEc6#-z0Ve-qG9w5)FVw0bm9ybE$#(KkUJ9AC*$dwP+rWi`0;4KJh?hQi|3gy~_s@atxL*5LbBi3I zLj((L1Ip7T4f})V7wsJkq)$4r?f43@Q;e>e_P+-xadrw57@)3J;XG4H*Qn~f z?}h{228z7kr7p#m(CzMrrzfjXCc7N!jUBoC85L6O50J+_Qu+V!kJcu|X)~KRj0DQ7 zZ}HWGyIBmOR}?Ef?b$4>48)wKEXqVo-Ub1V^NL$9Z*(&Ntz)iRt0s%eqD=0r&EDhV zx5n#W6hqJJd-yiyw<7!IpSAA4GcSBz+fzB2d^UkZjUPIX_fqVAigZzt4_?0_UmEoJ zKvm6q8C2}|-@E*;`3sg(7wIZ`8J?ED4kv;t)xe4VwZS{PsrTZi#m-Mwmj{)J+TGHG z1wOkYGBjrcQh;N)gD#Zzkh-u^IyU&8GUa8XDt3Maq{n>WUF5zyqHMC>yHj{TRLXF<%j*1z-d ztLLSqYqXO$@rwc$TcbAczedC6*inB!JfFu*45Y-V4Xw_BI3Q zpU;dtN0Nr){Bq)T)gcRTWl}6~P^M6k9ZI8mpCEop{FG#Mhw26?dpTCH$W4c4$jB5; zy+IFbc9hJG6+wu%R*8XIP|rj;8Dt^VC3&Fl6&t)Ybe+sm3oJVTXmm*7c$ynP#~*$D zFV^`k&y+jEQAlqt8o=!@Br1m8fPC)*0cK`FR1Y$3B0#(!6ortnv_X-`MuW1AgtOpxVnWT3`SeoS&yveA9N{JQq^ ztnDDfUE~(;zFqbibCDh8w8CqwcwvKA&!cP1jg3QJ+r2_zo^PagrN1X}{PrQ87&X}z zGm(K}K}R)-icET~Wx;zNnEu6N$OI11}%*#XCc zGK6R8brBB75ijrz@pwlEji*2DlivP6){5MGg1}B(*vg^bU8-0q!YVb8@Sq3z(RFu} zsOHlOm$DY%7B=ot`z1_tI4sNq#2Xz?A6UDz_(sFwfB*Ype@kF$mAxLt!2kgCPt&I? zQ|jRV#tI0JN6l75Jbqw|4)R~GzirLY^JRC#iI*@OOg#I%@JFAsAWl@cV9|V}W?wWP z^_dIjm_#_nFwNF9lC}Z>=ZY~J`+=|F_BIj@t$Fx7GFoS-Zws?08A9u|sUXeWr)p#N zhV_6R?rmW2zaOB7?Sus8oU(^pd^vYUn&yVGMN#CpPPQxvs^hgouVFcKlh7c8`Xi~4 zi50~A#<2UPsb0zICV_r_hw=u@KD)w}3HJNt&PY~Qh>9QuYSeaz4~&Mm9Zx+RE#$ai z)cE@E8%LAU;>1oXM@mcVH<4OyEo+c(r?cGqUd-@o9ka4Ci*d9CEEW%p(aB3w|8jJi zc^$Lt-N|c+ez>@q6T6~C7ORyUip`=(1{GQ3+B6?`nIVcD$Z3`<61pS zhrCUjCuPluos>5z+A~$IYlg&1Xt!qesu`uC^qC10&wF-Ft)5jATH~6mjtRX)(q|e4 zoxw|OlSq1NC=;QY#2aEV%g<$^V$lVtm}M(=~dxOr#kT6VV_ypLexocnX)RUMNsANz{$K~1YR zi|%o+1+Btb;V#-Bz>=Y3>TY!(s2bj!70qm6GUTza9b^!kh(I;6C;cU>n>}iU(aJbc zv`3Gk#~sAKX>4I+@H^UF6PS&nWy+PFA4^d#RZq9m*fyPRe5y|T2ytekDf)_LPXQt? zoKeUA;n%;kE)sc8LM$A$F1T{eBaKp>sg(8R?0%PcU!(e#=l@2kuFW8N%Ani(_7A;( zqj}b+DBR176RmXJJZaO6C89gZG))X^+P*F|OI*^Yv$)Dh%8fCzZQe8rqD#d#^z`@%CgwI?ux_xBJZX<}+UP|_Ih+IzT0JUl~x(+(Ie zUc=+O;fV3Rx^2E@|2M1ur}@O-#Qtxwh5wsNv3n?zNkt;1V3{B$blC*ttJOnt(=qz2 zXT3IQ(#=^Fq*(=2V(j24DBfrYPXyr~2@JN+pOC6nU@?<=D!{f=L4n5)pkMUEp z)m7g#!{hholHMXm`Q;lp@wQQu1(-geSV-1AMMWMU`yn%F>C8`IKcGyIs;Ore_PA$E z@0tz@Fk$-y==FBU6GHcRb_z>Ks~70^F3fk^CD;LT>$>o)@Lu5}D3Cs;+Us?kd`51N z6Ck=VblNAz@Lggz|Xy zzM>R%nCjK(J{|tLObN37^B7Ir-|ttpv=J0v*@q);HvHSqo_QidJ^g^LQKq}?^Xef< znm7^q-bKOdNmh8Wx+u7TO?lbCUSv1AXSo+Jxi62^*xNYbPk!u<7~Wy}oxgll~(h{0`fXt8vyZ@u=N zGu~T56Bta-PoW~FGF#n&e3r$+(x!lIxgL17@~^cw&OJVuz~1>1I+H#}_XcGRts zd|tZtmf7wwtNp)C&J4Fz?Znol&BD4|q1a0l`IL&h4jggm0DFGZq-_MT0tI=!7eL>?Ov6FHIuNzXwPm?guW^_b-72l zSz*$!Y1OQlqxZ*peeBgHDDuY92#kyMBesByd$xP)mp=b#oo02`V$domwu~ahRAeX9 z>(NTbuzRMRei0VB$UeF;JciZ7IszQZcGp7nejnY2;7Yj;L$P_`OC$rtO(^}+3>(e3 zThr@ddbowDCh^Ly@O~E<#i=?}KJA@o=256snh2Q;p`fjd>*z(2HXv_GoZhM##R}Td zBV)@b4A+KxWQN;hc3RPZZ})%u1M6yOKJ%CpXWMYpQMUl3OXg-#nCOZ4ZGa7`5@b62vvXSbo$K0AU>jNl3r&Fp~b=3lNP(KdSTh*ZJq`< zU2$-&5=%rc1#X%VKR1PL2*zf*SXnH&5%!6qQ;y|<7&Ea4`G!p|-P%+?^?dYW@toL` zAMCy_wDyagHsgxJ(;ZCxPZ+3793Ug?^@tN8TOm>~v|t@8sNIFPsK!R=Yye#RaXWfU z+R_6*+;%TH@$IjhP0NA;aVsgC0QpNpnPF-v_B2IKQju#s*SHmdMq#IL`?O^B#g{kE zUltncUE|tDM+Ghs0XfrI=}sY%N*)Wd?+CajX*50@#(Iwx)<;4^yD02KzTLETl_gV&)r(>4)qA~+j6qa2&>9Gq zq5hywjY4tY_cSQZljYw1F6q!&sW?_p8Msvu<)*7sc7|Xn6ZTyuFfuu+gR@4lW()k$^(jE@dq&_ovt7v7h=Ci=g}D+C0XL%GPP4xBi#^u#P4%CQYaz|j3InOgl>&@ zV{jCO|JlV54mWnAh0kz(+T$8Gd_MZ_`Ki_k;!eAC;z;?f^{(}<49pGHEpc5uDV<&? z>=lBiUZ-xRS38(Ab(pcS9RkBV?7&Oh5Ew3W*yYlC@p&w@#`v2_ml4#(D+9BjtgY95 zAP=}sZ3I`g-F2_mejkU~hcAz>?Qp?qIyepWDB(9Mf2oRF!L`f$!;gk!#i#< zksP5|U=ce&MIKWvo|GF<LO`k|veX*Hry3XMUBe#yco^oJynQ@sh}U3t;;pL*5U&7dYcAk-$a-9pyG^kL2+H z^Fzw8ezxVs$32hkfEy{p8J7KT|KP+dGkRWa%2-1(oftjGEYNe9VnGFaKdyw^=tm*< zLhdMWn@NZ3xLQc&yr;RWDRO%UeT{c8iLT$JR-ETEKtE>)jUK2-`LICUJ28r?rm+;| z5@`Xkk`=!FE+$^pT%U!p!S^R8L&C#C)D$TXxC5Fc(R|YAI7ia%;IWf6;s>6ikET@q z(hQ#OUUBUtm!6xTORojgZd2?{id?56_4LQ;X0b^kJ1?xv2jYi=dfCeajc$42AjBu! zsy?MQxau|uuE;;Q#-!141j&So>9(L9><0g~kWS|0>p2VF{|MGzAA>bRZut^Y#MY_X zr+ncJshbhHCDb)08OkjWxgMWeq;40(UTUpp35lNALsB*Inl#svX*xW?+#U#2GfCnv zRL$aS*q6_OqR%5x^qDEcG!5*qjfS0D_ibJkkc4CUTk%O1>9mOb~Qz!s7MqV*cOo%dI6+jbA+X1 zQY~ls&Iit(4v1kM8iPyJ&(aF5*VRtDSmIbi?+lHJz%9;X^-j4_5GzP`)7|pfIBoF@ zm_VLKuqlKmBnH zJs-4MGs$pJuwYVsjq6T%w|~qGjB%yUbhzB^@Gr@hH$VK7j=jq3%{#Br-Qjij+XC}+ z@`vNy$H-DAc73;7Ob{C=HjW}ORAfSEV{kINi!N0yVE{_CQ`P2eS(5eU>lC!7KYnclHX|APok3xq8soT8D1J;Xr8C}cN z9)jBA`QDilxP#;(CFF<(y-gja?>zJ@m+9h#22nFfRJVmdlazrb1K(2~@DOD0b$b6c z`h#M+*Ih5`gN1sE26nzQA1;;*w5b?a<+lyUR>RYUV*CQ2HZpx_5evKUz)G*fFDwtp zgOVn6We5LDr*X0OeCVi*d#P>vUw>}~-JxH4D9NqoW+4@A0nrB(+e49ikd6vG0H)MC zRK86Ik{CcM!moHBe0k09Mlhqo}H!P_E1y_ z)jQYcR{Nt5uX9?go|stXW<3!;C+VLY(`F0Un0ye9)YYi+q?rG?Te5>a=Djw!ZS0p= zHcp#kK5ym*u6)nB)Y54eS6s!DP-`2*MoCg>jE2MpcT8%7dfK$HA7opL{_W)Q@aTr0 zcl6D?Kl(iDEACuB{nWGbYQ#OHUYn)J396>Id8Yg06;OvJ1s?ULO^Tro!CpM9@`)8;0%*SrxUreCI`vs8_yPl(_xAPa0#NqO)oVdUe_AiIz zRb@~tFw*a&BD=|6I!#z682qf$)@#qrTM=4KZwIZLDj$=c-_6-&f^*tlVViQNd@nGj zl?nP>HcT`f6@Vt7<3hr=dNV@@d&u#q$DO+aZu^tDcKm^7-E(OsIr9$&C2HQi^7Sk7 zKk3;=GQ3i!?Q=P&?V#}@ZX6cF_TS+JeXt?e@yZ(EWmze~miy%e5j!-A>LZTSw;7$G z;2ze|-r6xUsB}xs>lp*OdDu~3(3g#8gygmzZs1uln%00pX zKOhC}U=7o5$-XgR_2GvQv*C6iG6uWmg%Ib>mR{x>Qv8o*guHc%`X>408S_f+TEO`l z#a^XIBNf>!ZjxS87EP_5k1OXp!02*ZQZuz+UcDA~3gUhBvi!Llg7x&N=>_vZOdTSl zTfFPFWg-Z^yx0L!!e=G>={}boFILa$bE%x1F6ncrcdMSI_g}`=tLnA)r)EmFcz4RN z?_yz|Dvhq0x)^r1+x$EHt4QJ0xK}#G$WYTmiiL%fQ|Y)@(&zQLB+g8Ep~Jt(tsr2J zusXb6TSY$ePNlnMoeC;)gMWeGAbV1-+taP@h<*}7T$&Lc4kDlPllv!;v2Ohzt?RR$ zc6rTFIF8xBXcDpZUWbJ9CsdaNC~W~7PFPIPLYLB*>rn4~H@w+=1Xbm@*iN7M(7j81 z*3(?Hf=i=XC02i~^<5Ua1c>x1frcg>)ys9>j3D;Ooa30fKW1bOel#) z1`&S4J(AY^Afiv)tg_w{a$57kvCFep(7>K{JshUnMfZAKg{?bG2Op5~)61SdC5+@9 zb3bDmQPz(o4I`A>NZ&sAjh5i@l-4Tg`N8nir;sb`B((BZbA`uzr`rnunf^(q{1z3F_?h zyC69aEBW-tnK*Z1RVU`{6p9y1?tnTEZk@J<)GE`# ziekU`ge7q82D_)h^6B?2fZMY2qw%xNuzR;iTSl^;8`zz&SfdqEYym~`fdgN(6}AD= z>4KNQfyP0P^KEpyqDWE~PK$Xl8>A-JY4=YrpQJmiTpa>>*y_$u0|+4RpWa5lUo8AW zfT^$8Zqc02Ay1}T+Cd)5QMm;@^7rcm<)CM|Z&D{xyKRo{XG6cp7ERDy=#>V`VeW5d|KgHtVo;7C}lxVyw8zN#O4-^=-tU+uH5_0=ri=XJ*) zB>;Oo56v$RY$8kD4$Z&jl_M;ib4ie*SpkI*Xe1+u)8+^US0At8N>~;WzU*@jC02#mrMev4s>VfV~n~vsf=X{>o+9 zX#xgA`l;8N*@wU_&hqXG?@)FLF)Vx>wllK?2Ln22tgkGZ+oia|90_O^8=w&Up*Lu` zKzT0mxS1H?Fryfm%lCO<%Nk`hBv~AP<&HlTOsQh$M?Y0?a+pCJ=F*7p_854>9t`s> z{4aHeeHtP%BWPyg508-*{1D{CzDtS)f;Lj@28zT%tt9Yq9Z zL2+=JnGrI9X{B?72El%xPR5?>5q9577#%8xJk32*{d@0&)&6FiQvKTc0+QmyrBUS; zR$@QJ?xjdJiX>x&6Y8sDo?K2)rsOmCbMuhl{IEZAgh0;RVSf|@Gu0WPC%ax0B`FR* z5^!4n&@*A8={ho-XUeb`L@zrWHdvQr$A!(nfY@$BhMh>BTpTjI_BiB*$S}K2;}eU& zU1iNK?zCG89DL!(QIsjcN)eb;Ap5t6oB}p|dyls*V`CF0Z0X3(ZFpf~+`8|ronXBP zdZuQf91HWZonoPwd@=-rN+GP8A@6mEkQqj7Pfva**X*8KBc3?1XwD6qVJ0BzzYd)I)w8~r zeM__<_^cH2VvK@n`a)Qv+Y-_%Yz?adCUl3J4vSY0CpM$P`auVu^Fqb=d)rTntm_;3 zglL_3h0IY6iM2}TxNdjvWK!wnGfrsQ2KW|7WgY$^@B(B*-}plz$Ci+}|k+dgTO2dP_lB^9=m@$2x5u6cFM(aT^Rh6j)Q zaMNj^Q&YQq?+b$1z&w?%f;@Cjhg`4gY_`;SOY3==q;Z|US-s1;Mum?_a^kRWzJ-7E zm_Zul;%bzS{5t4zkY2CXR@3cXmu6LkbuwEfJ%s4Y!@xYjIuVM@)oX#2UwY1aqx&Ps zQAqOtpEhZlaIpY8aHRdR%}{YTIis;ePkHvAK6`K7-`VWt`Kh#2)jbCjJ_j`C&_3+^ zKc=H0K$)Ayq*(|3pQ%+fM`*uU4Km1@!zv&wGI=ZqNU3;$=M|@nJUVY|ePCWTVaV_} z`jC6FiE?ZD-m}gskA&d(zaB+iIMzCUHo1(+1f?5Dsk5QL!_M#CRtVVnYoFR3Hw1=R z7?0l_`J*G>Fh^MKfAfvENjbkwLMPrOf|8{nC3$Bl_6$WnrXp)Si^ZV&g%sh*zS+_( zOviN0&_#;Jfr=B1(ILlkDQqL|_UoSA6w;?E6{XJ{2;L<`Y@9oAGRP~{J-bMKk?n-c z7a%G}{~9UDkT5OT7Yh2}Xro6BZ3K2|T}MQ#X1S=9#&6G*6$fKJfK*;pd@&@

    74JgZKKTjvg1C1lPQ~YqIpl=cuJ9+H$)AI#gWOz%@mtPkyI*n|Cb&t z?EZ@JXF8MncFNn&z*)g$MeVdg-}TJqnd9~|IE|rgI5JFdoUS|GgFkrmjjb^zrzCL2 z+MVR!Xb}}Bc3Ur*jekAGo}kE4D)t6{xhL}V(7P=3ggByWaRBnY_Xi&gg%TvyIz=J9 z#XH3#pROiAGr%BVgA8JZ0FZ0E%`**A{Tv?Z-r*ItGM&2w2vHus< z+@08zSdsLvgWU41$OeDRAs2b=g*GrnAP@9VdO}d`UnAceypO*wrq_LEG`^lNqrw#) zV;yd;=UIoxm22PG|34;6!{jp}^05!{vDMI%^}Sn6yMch*yPzuM z_eE5zWGD6eA(sYu2iYO%;2jOoM?V%dD`AW9SztN6hQx`gpk>}V?>q7`pF{Hsp<7mi za7dt?joYil>jN^P>ZO;1+r|4O4YFr}?WDtRpZvJusPGzZ^^|P^tH^n>Oq4JMI?Mp} zp1&WA*~=^RyiOiRCxhJ?`Au>nn$^#b!fr>KJ+^rt7p$IA#A_#U)9N5+d6}0mMXT%} zjbY8|7N!V3<$bVW`&{!D47sEVy2M+g4Z?&eb?`mA=Jhc3%G0dV(xR-RZIq}I^`Q4w zLq2ubZa5HF~rlLZ=N%Js@J!6$E0V(b7zH3ppmR4{0f zu2Hv0`-75Z8v1Bqx2%0L2E9Pv00ppmf}h|N8ik?kJ8jt5Alw~^sn|Lm2wX*Ky4^I!ARo$HwbLCldWa^$D<70{{<2JTfPa>h z@-)3b{485I=w1byzd-XufMo{lS-4DePo@QtfU_9(%857;byzhVn9S>u>FB5KAd>|w zWoS5B$V=w+$+t#ui4{&XZfrx%%fjLVE6SF5n2c4fa8nyeer=2uXv~fu4Lcs zX-%8YTKZG3NfgkE_jMbJ5-^)^nyhrcBis~K%Igltyf#i?8tku^&J21sxYr7BqgL4r zeIgCJ>S&ZJf4*b4cex5##Gs9r&fp~nJ+-QQRWqCHf#;o9HNd&=lSEkqoj`T*LAHc0 z4ZsO>*3Qz3@-SjFN*Nep-#iIMJ+kEgSu$OQl9ep$y-=Wf8 zF=&k5nS)!C1=S%%Jd~KNqpz?7QcZ&^mfBZ^Wr7OEEm_+9b&6cR(|TVX*Jm95rpZLT zbFA-Q$tfrH=x>`DovRdknS$&Kq_?oSXrr(}ep7Nk)S%%RKQkjl1B?ja0e+na)`4SB z4Jc3$B=8!AopfIezJE9at3bDk*79~60swnK^12+{ClIB;WXgWBRgkoT1UJnV z>D}-m-T*OT^ESQO3oq3N&Nt$6ofk4wv+0`>ybLPXa+*9R5>cbxU@IFrA!<@UQ^GWV zlev1XfA6%2vLejiK+09UbSG&PcBw0(b3F>BDf1snE2+!DxgI)5kLB?C$&g$JlTLm_ zUSzJvf!UeR8@bg@E4m`?p-)6z=G_SCCAE-CDVos|nCpIM1}7j`8HX1)+W|2zUepd- zI3b4mc~}0nFcV@D(|b$FZYQ4LPMb||dWr?Lp955EAA^Lzy+J9Yi{DI=BKqf)t1kJL zzJVT1DG;1BwKE_;LZ9=>=eP4yef5#3a}A2Zd!hA1j{i1SEtsi&es|{7L1X}%V6PA8 z1v@tEV(fO?P1kumCX2P?u?PIyLoUXa448pyD9wZE;r`_CrGeD^H%_v8a7wmc8a zKyZs1sC)sQNQ3x)FTxhNZ%#DZhp zOp# z@rdIQm@rOo+eP8Omsv0S;O50C&ZaFd!sXK~oD9>XKe*-}vn>32r82%&iorTb0{A`4 z1^shs=&elej2=22a=>W;N$kWeRjy8R=@|P;3@OGO1X6^qSRhd*3(!kJ0X# zxt?2nOQSc>tPgIdA2990tm(SxL-Nz4!>d;U247ht(9N2Nn?7kQcion|Hr`{b-}TYb zpH9nX{o%uegjd~7ZgAZ@y>7OqZgzQ8Vob3J=7}@nJQdQ4micKeK{Oc!;z0IYNAC!R zl0?1RWhg5(q-BK$eU#LwD(Jcj8fDUyH#F^VxvDaB-VInSU{NgW7 zHszh&Cw$3=ug%gs!HnWINUC_%m!t+VvoZrpb62MyIyOd+Z{|@88eb0 zk<}^emhTi@=MB4bg7k4oKnIhn8lH<{oP%EXg|I?9tE`dcc);=R&&c*U8+l;b8i+?@ zeTJrm!Hrm61hq#%p{)wLE7%Ak8!gNQaYjh7M^#u_0QQ~MDsekZe>Ww@C?w`Qn%SL_ z!@1{6bJ;Ob_TxS6+8GE9FwcLME^Emr7L$7S51KHDvqbt^QQH~)>C^}d$9d0(8YoS1EG zr2}t+?15l8!MJdT=Hv%A>ahWCkrKI{srafk45bvmGq9541gwvayO|%o z>h)huO03Jc?}}w#&erIRS(si*%8+${I&6ObIcY+8z^=?Mb=TVMLgubYeBF>KV5^>@EB{^ z53Qso2sa}J)^WsLICCI+XC34gc$p8_mXOHMmuSxF-yf3J@@Ivs?W z8sxYL)wvhUFz6ret!R}W_%-r19l%%3 z@_Fh7ryp3zG@0CvtO~0Y9FSsgW^4!@2k>mRpkqPkueQ&+C^z{(+0%aU zYm&~*|8e4VcDWgp@+o#ZMY5^b8aCJCBqWWHTaz4iD;%r+&?mz3t^$BVJjn`1-CL9k zyeX-ekCpKlNH(6~So1WF*>x;~&*7XWXHu@Zns8G0R?T~4o@aUSuJ!Zy^NQ=zWPV?`ZBY7b@XTL1 z5Mt&Z#PMB79j&yKLFNJ^C!WD_&5YPaicO(Nl96LxVn}pru1E)&cQNIyxkfsreX=C) zu_|A3yz9>o=YWdmr%cH2fBKr+S6}rEH1Y&c)}+j5pI_ztJQY6u7~jKqZyDsKi-0&V z=-xAtabFI@4HL+OjI`lHtj{x}#11F_ho498f1A9Uch@JpNp3qa-)@x|+hu@aA5o;2 zid{x6EnA2L%$zq$#F}@j013AkO|E~ta2NmmdAp!qQmZ^QXCIBslO$;HimTxDOb-_2 zVmPEt+{GLXuA-Y4Wl9r3Hpif5f(<7s6wf4sKz8Ytr?~dYnim<4#?45e!K3n2IT6@o zF`dEd=;$Ls1+K7C#OSw<-o-y7M&W?_P)`SBmDNxaUH}1cJP+FBf#jS<@0RV6N*+n7 zL$*NK18OvH;p-y}ACH?V6no@r=C#vzy>#??*-8kPFYSdXE=SpB zdcWRtd_CFm+JtMM>TU#wubg5_DYBc2y({cvHn|&rHmSEyIT*YAovgoG|L+%}5!Yq# zX11%gK#$^W)2bqMQFv678ei9d_}IsaeUb~b@yri}1^jx@HMpoo>hUwlb;%|NR7G(! zZ&qDoyW71V2xh!zqo{vcz*E_7ZzwU~V<}pR7()05q{pXMg{@OyRCR2@&oTc&9B$~C zKVvK>a(if6Wlf zzKP6r(RS(hRykb3q4r8;-P8<1KSpkfCh{s!%wLt4F{WFP#}A(z{c z8(`gi(x8g0Em&$=RG=wo2K7q_R>wVXUu&jhpnL>&RsTbr!N*N zG^^{CIf8tT2h7%hWZn*mPFfUrjm_ca^>T)xx$)ujjtg?#x6bn)Puj_eqrXGbT58a0#4KPtw>G=@Z42$O{OwjND(S7nCtjXTEKk2_pHal@( zX1|#Q(NQea1>{n(O$xn22O?*K?gk>t2Ejo;(6ySC$ZUm@o+f3A01wbA_lE9@I8Po3 zG|0?LXF8Z_vI^F=9iRa+=4x1*M#<)}AMN3{=@Z6w5Vx7B`-^E`ax)?2Pnt`=BMDB7 zlmauPY@yh6ifn+$h@yvXkz%bE_%5xWSFAyLQahcex&WOJHRn`MCGGTD&xtb*9HB9` zaUR22+@R4J{)eBv>e^?>ZOW#L0^7n51r3Dvg1AIKZx3HL>#6$*k6iyg{xVS~__ytz zi89?Ry`V|mBui&jgywjl!gP*n7t<-c`-a}DkDtuD2W|%*cTtU<@9`C0Q>(l};QFAX z_aOAF&|V1Rh@a66jQO@PU}dyZIN^udKjo>1rS1opgrXA%RjkODpieg_YE?3(US zq*7i1((i84K{93xhieKVkPTnUYk_GBwD$rMp$gTplay%`#|Eg;AYk8ZZ1=x@|5Z0J zdlueeP8gU7yZIT)UGBF@ruR)o@BUD_PB0jHn_$xpEOJ=wIxK!1ofCp=gkE`0XtVmJ zWNW~Xd@m2T-XVpM?%Cvf9J)Af22sxsnZeLX>MXyKs+9Ebay>GXIic~S*KX_Xu_40T zs+hgA6EL`$nB)`x_H|1^S1u;TiI+cCGVP(uIgOX1QuYvv~ z@}jsi@(@q6HKGliObkLmqrwjP5y>Wg2auytY!}PQjzA6$Tv1~=5sHj!u&&5nF2ZMj zls)b=JjtJ3&`z&7J;}~X9#a3q-7iI$03-jSCX4Lh7P@m{FXpTn?yD&l9)!y0g7}#R zFGj0u341E3COe|bK{h;_egY)9J9Dr|PB$xKMz_k)V_5h!BoUOP2Y_gYojQQ8FzAVm zr-0eufxAH99v_?(ualAGQ60=l*G_3QIqbUmElfc_F|`x zxytqdcjlz4k)k#15;wQez+M4i7isMb{2a8-px**zN#pRK9R?W_h{lQ>>$cN%?whMD z`FK`@S9-p2L-MCv%oTOoKN=*(jQ!)VCmlAjz}|D`6~$1s+l9rf6Z0R}`#mw=iet&F z^E!=PC!WmvvdPk%{LgD2k>gWHyV=9rM6rz&IY-4JtGSP#>yLCcto|#32l=3Kn->!5 zYvfl$bZ)z#OU)u!CrG$`MvH#F1R11c_H63D?a9xdnNi>>#{uJ6+3* zn+yH!gWI?}q+DV3qKUU{3s*Q=e~d?AQ#voKDx=EQS=t;ft1Ks;+N_w)<)W0CYi5G? zes9rw{;-SjUo-R`L!YNu^eh~Ek>NZyVV0At+5lImj`{e;bGfG#&%K-TlP;3X*Tz3S zXl6Q#DON|3JSz5sMO(gLT;?Gs7#!XCRbf?pZ0m)dOE!v;LkB~s^*8KU9kI?$lR#F@ z?c<}fgx=2%Rh6)sWbzybo^cw)F@xdrPV`~^Q9IiNmPKtv>qs#-usE?LsW$`5A&Naf zk$t$Bs-m%1-R{Ws0V{crE|KChW$T(>*F`aL(+6&@Z_6;XEVj7Q6!a$HTJ#iVAi?~MD<4vxhxm;2W7Y}IVHbl z=ylK{Es{)VnGz>t*f65@I>dLU=;yCv-}xrqfzWzhA+7hX<<--5K)~F^-#oKj*eTch z9}#0WdL4b9PH{cz`9R>f$;4umIDr6+fyL<#du!dVKjEMF|L5pAvTxJoF=HkoIpVrZU@dn<3Wletm15i`@+K;lS4V!U?ddnNzRGIH{w zM6(k+rdH!45c;PzOIyXYejpGpt5Ed%RQoi_G?2#GFU}X0M{N*iNtZ6g1v19}C%`#z zRtQ*kOFGiNUuId1_?pZ$JIpLdI>m0F$a*Rk`A*1@LOrZ@VLQKETrUU3Aj}1o@lLu< z07^za^0p&p#bRPD+ zKNMtmo2Tldb9lufC`5b9dA6gN>hH-9(I%CYThx;1Am@)!n(64JRff zfS)nK@kpiEWQwe%Vz<);hOXuX{5yu<(ArwHRWw1MF@u0j2r@fkV%{9s-}9v0w9fmc{EuV*Ti$zUJ&2iEz>rgKK(6`XD+8-KNUm~}O z=C-T|lAqn6y@A&ocG=8ZCTjFb5p>cZ5Y!)35`YCLNQWuFW}vuILZRvEdC7uuu_jN| z!8DT+@sW^KVE!uRp9AfM174S4`Gk86ya!b4SQT+J1sJ+qMfgDyn^q z-9yV@5#K{^HS`j7TGfvF$-Aa}!Gyf{jo4(w4N3G<=!6ShZjl3*mIbiolta?E|YF9z9vC|s8h z^^E%@`#_}bF>`fZuKQYY0%{ox=Pf?%l}!VTVfQs&{H$(qT&O;Bm7p5>FV%pM8B&O= z=*!}6d4kuiaINxsV2`X@^HPI zCA_p1wdL*O@P7OV;P7D_4&~-j>wZn1Si00!e9;O;I$y8UD%ZK`!5zpB>ZJ!IwSEqa ze_H*cZM*eWPj}c3ZnxfPC4ssXe-gjyg|W-o#plR5Nsc6A#(?DDf~RkN^7aLKSyZ$7 zUU-$boX%l8>4~kGw+*u;EOlXZdsoqJ^TVoEp4l>8Q&l0>wKWX>GheSuWk+XJ7QjhB~uzW1= zy`3He(Jy{HU>OVMf*&Ui_F0J>=c^j%UJwVW_J0(PRJnWMol@L(&Nr7I=iA3O5Z=~$ zCb8W#wg&s$|Hia7wj4dX!^2JpIe+s&uEey^k^l9JO=SHPkP02Sp2()yZ4}8sBI1x7 z({U&sd&oc!LGjsax-+s$m?Ajnbz^oyNV7U2NwvRL*&6mBN`oD)+U+LEVGUd6i+FWVURX|0DgF$I#H0mODB6W9WgD4gqIG98>*zjykz|<&wTLPdAMgv{Q?uzafaGdoyh@X@oi6sk z3alPF9~!UKNE6)hRHa@W%){yBejBH@y5+c5dFDiPFz+rM{^`T0X3$@}BE7=4ieEah zjTx3kaE|Ne%MknqReREs5aYDDZ7bv#Z0mK`E5)@%Y!D{H?gE%j>JLhq37x91@v!yY za#68p9CfKLyM?3e<2u}uT`njK^IKIqDF;9an&jYCs%-u-{8Wf3)Q5I;}mi!oWQY zq55Rr22WI}GYIh3)5ir>VU03u>!~>rvWw1%NM!KM00G^FX)9$|L}qMaQWXXjM3*8R zYHG$e_6`_y(m&uoMayMJH6Mjur(pH)e$Isjtw$HE3r_~L57wjixyux=t&O^FP za41Mq?o+?uUU+L*i*$eR%1}*<^oXjLbjo{VX#o?^yKk(BavYY7SAIJyZOvCq6VT+o zpMFG6IPqc${MZqW?L~^cK#}uQY@uf{ADJt8bRPrbUJ!xAfj(wP-tM-|r%1I7^e_YZ zK%S+TeBHD}*>Sd%ha||5pB|30J$_=FTKHr;s7D z;k!?3pB90Xc%Jq1Qs1)Y#}KoiUz!QH{F(C}yd5!)n$2yfXU2w~u{g(^nv5J0n$alHgXY zviqG6eR=>@Sw5RQ9`aYYN9&qe6O`Z1Tv2+M;v@{)j6JwaPMifojz#V%rlw zvZIgw3xmvQWS!UJRGq(f$HxS@rXN=$XJEzC!e>>XTBz`k(IP;ACtsljHiE2<@|J}MT%>8psLTvpHVp<3K9lA zQv~TuIv?I(WX9x{w=f@z*D{xU?E07M$UoBAenapBdO5jCBg1zuuR+*B`eN3S0@rSJmXC4U zKA*dxmwj;CkV}L7itpCwi){7S2)LYvi4!6|b9Fd<##j!vj)+leSE%3q@6Ys>GmX=R zYpl#2SS!Czeq7Nb?%@?FJLFC3+uyqSgGXP!`fn8rFRI(=8%!OYIIAi+d&X_qZBTj6 zQ)NmY^Up%%N;aJt0~}x258UUbbs1$Ao&-IvOxo{*bO?n>fIa)knv!|*1;a* zahxzR*`+AsUCZWtFP2w3?;?$qn|UMo6bns{vZ>e{LFJ zm;=P_UfIsb2ch?TKwOc&Gv}U9vpUnY^UWT*GpICh$mQyS0}gN8IA;4YMoidxnsw|1 zEc*6O#?y(yi4kO_6UFei5I?}hQzk$W+Mbve_GVz`^xJdKlO8B3x#m{n(*U8P99|n! z587cyH9JS!9K$${1w6I_U_8f74J*mD6c=;a>Bfp&SRIWr+bGSV^J@?ug$hW>9;GuI zNCIySw2p9K8OZ2&p37K|eg(&DwQeXz@mVMR_Hg4rSZWr&ILSCM1ICInA=V=r#k;XO zzXDpRr4khW!FnMOu?)Bodzy63TJyCWFd(T>Pv~=!;cyfiY?NW85i(!iCuN}dkXbx&(=WJUfB;^)&cbg=Bq#WZ&SSF-Sr7?lG{_r zDzl2_0g8P@kzOMs>9DF)REj<#YP%JILwE*ML^p(XO49-~Rl(cYG-z&ydKp!8-;5(( z#UA&Cpu&;~I_N$0AOnmyqa@sj6SSb(kBUyvY*Y2rYc~i@)<$b^Lr%nMuf(YXM1w!@ z&Eep^^EOV^bWR^2#{Kn@m7d7F+bO!vg9xm@VOOySlJM@uG=!xHjMv2z;7j0}cA&lJ z=(rkS+M>7vjTY3J3t% zcqT}Ogr1hSKX{!R9*bj&dvtUcl(8R(StnTUt5sf{UlLiNNQ%gb0HRdBD&I#(AD8Pr zF9#31Y?0nI7?W=L`N1e6esO+H&{)ldeQm=iHe-zD;x$3j<)43#;+fF&>-YQX$%@y8 zo0V(!P;8{w6pAEa0qz6I$Kt;({mXjZv%qdSwjEAmu1nfUyyAgmtSB8Th@jCW`T4JY zqxs(-{qo1Z`)|oAid{jGIQyY=jNi*A;m^L7UC+5>j+}U5Wu-^@8H4)6Ds@)i(Rp`f z_sDwbN@%RLiQgQZ=+zGk^_!Bjq*K};t8oYoZ}ef}X!=IKhy%7xAO6wxzpduf=@I3A zSH!(^{g(}0+Df1;V7m|{R`cl;W`b>o?QviS4E@V4`Og;vKU`p%N}9gEGM`j(n@XH` zsdd?GDtVt`Pg3L<6}yF38H_Fx(rb+EKszB`r&%7B&>ehmr2Os-E{V)jAsu*v&4%qGZQ~Hx=`c=y^yH0)R&SnfvN7(D>eI=A(Rd0@ z?0#G{vm|vCdz>P*RP1qhipvA;%RA^|{xVg`!YVqRJuj+-hD_M&Go9&DuZybkh-Zgf z&PjWO_#4l_-;aG$X0?RjtA{f-1>)}^;bwRIeOb|KIP&U(PQ&j4ddT$tq2WrI0p$UB zTsnUZFPPCU{p7+dlcD+3yB$YJp%e2g z!5bMtemY38P?541wG=yrS|uie2BPcfyMmKmX{3bK3%cAdLl3v}iWQ(sRv)c*OPRki z5H&cngId^nUP6dI4Exyr06 zwJ&{rMHIyiQNab3MOg$9*%VQBao>QTj-n`=XbeU`ks!B#d?i*-D;x%3Im8ELd%v;R?WVO8@lireYW zf=Z3hEEi_Zuqud5<_=CRht-o~S@TLtQua^1jX=q3&y67IqbW0Xye&FB0uF@~1Ng{= zTnBIrp!FDcGFa*evuMvaCP{`A;AqL^nfZLYj0_+VBDkFA1qE<|^LeFoG!Lo7)9F6v zUS}PKg(C|PcFM3p_A%M!T z7lK4JjH%KgEb3(&0`_|8R!o2j>_8xouE!Rop_+7yvSiBjnft*BW#g#bF(ATv(?7~c z?1&g;AsY90|78l(@EE(@M6(3ExRl-sPW7-}24l z<%6@cf_KY5%cGTkPGyfRfygK`YV}K;AVL*P>A0#kE66yWR=Vhl=tZJT3DnmCYcEpgVT^??*C`h^<5ohD6`^B9 zk#!LC+3KAVQ_qpzu1FBGJ8V}SOEGIH5)Im|`1))^j1_V@B+?hfAQKE~>vK2inlB@) zGk<1}!f6$HWp?VjzNYcsFNs89i3G2g-3i(VwWKjCf6yUXl0EU#yj#9o0?v~y0ZkBF zbxe8~0&&uP8bSsXRnjJ5aZoA^N^+Yb#t{ogSV>H$OF53OqhpSo}tVv`vx5?*by| z4U_h|UYge@J@a9^r(SykV#cF+=gB9!)1j9`^;$gc7WYZ7xjpQWFuyT0nkzG;#{Ox6ifGzUD-Ps&&7 zB1Fi3Q0x;vchI3mu2)}Iw!`-13)S7}khBN|X^w>$B293!l7XQqK2QYIH=V}V^`W#! zRDJI_Uh>Hn=kxA+w#uxwi9vvw32Czy5o;o!6QsxOSdjR`7d;6nL3LBL*X0Xu-A=FD zuP#Ur(Or{f(MP<}=|Qja^mXZK?Q*$QXZ~5@g&Q)SHJbJAb3(=|M}Bf=swvOmlSxQ+ zY%u{(^RP{FI>i7HM>5t4H5&GfxzT7VX(Z`Bn8a2i?{&f4#I?alU^eJrpvkRMLt;sy zXN!8d@~X7ldl~Qo*-(Muh8|P#Vm-8&j>L|h527-iUUW6}m=x`2`H;1wTji7#fCe+0 z&a&y#eROWt^gn9-{0`Wse-!`DD_(z$k(zn|pA;VUt5AWjc)0L*#0%dL(nFer$4M)} zN_XWDLU53u$CL~6+@k$D<<;^9t)x%2K(9@w&(h0+(ja@Og9-QTgYxx*)^rPxI_+j5 z$5dY6`KY5=?;<;LCjRDkji$ymyH#6R%w?D6bue|>^zn!156<8A);TSvK4(MS!z4gQ zDTH2kll{}W306;xQ-%j+h`PyDFr+}VwWv&-ZS$Sji-(C(qzT0=`tF}8qaCE*rT+{W za2TbT9jhWx4`rAt5=}8HDH276p{-umIz;;q7F3Kau%TYT{z97r687GQ6PuiNyRHrHQeysfojOy3H^D8O9~=h}DJk>~ zw=WeBfK;p<>h>nm59og79a^_7BrPD9cf>KBzBJD!NHYgXqpd*0uh{}m@0H*EBW)BQ z-QRORNh0jnkO~!Oh6#&x6tjjRtEezNvt3p%E1bN7H|T(Z1BnryJ4>Kpa?G4&Tg>EB z$7B~6qb(N;Kl;{xPBem|{9PL;~%^Lxf6p8^uZ~_(9<=^WZ&1(|wmB#Q2 z`N@)a&ki7!%#>X6&gFMaFPAM6-4DU^jV59HtTs(5{eUDv>{1uMXToQrkS0I%fpfoU zCUk}igrQ^ZpTw;PJ9O4JoeP|Agh|lCPn$@I9m|gT5s)}dF()W;lnUG6gj8_hP#?yS z8=4J)AS)mu1dIJ(um@)Coh8-szR+IhE=US&8M>dMIV$OdB1J=WXOL$jjo%<`2QNa6 zu$9I@r6HdVrjB8KXjo-@vQR6?@U6*;s37UYh z-GVSlOR@6=n4W%t8*rdCJ9b79JLAeSnk)Zg#M3X2em|6SaEoiRV}Z101kKE%R~TTN ze*hM@x35YU``&?{9gvZ_c4DX77Uyblw09?+sv3ZLn^BIj8u0wW{i6^)eL|C z3%^QGA{Y$n;&;zVo>uHp0e?`V(r)=?W*npb)s#fRZa3a6vVQy zzXQn|x-WnqXl-yc4Was(qDq$@n-dfoWd?T%#aY-`)Pb9X`s>s~ZR-O*T;*8SjnY9OM^~mu85n25u$AJ8JK8NH5YiBJ&Rlu)SD)zI~^zYG--}y z)AZT~Y274LtJZny6*k#8S&W}yOPNK;Fs~CkeiDAX_TVJrYSuc>|0>z?(yV4hBXp@< z6a&TiQmL>=@1;SDczv$%!pBE-#xf@v9+BRMy^bjsO}r6wlJxj@I)f?Lr_VWBg4b>d z(lj?`U7B}Blb~LJLoBcrmblBIVK~A=^I*Zb)vEU|ZSD^4K6zpuGAvj-lev)p(?9ZH=psIWZZz!ekGv{-2(tS$#;i8f$ zkRd&#UYRLMn(z_QTtaap2%w8~hf2dR1`cmc_oQV|Fp(LLPdQIHTL<=aXOx}QD|FcQ zuGfsfDf0QtGO}wtIX+^34pPj1itNSG2my?@__5cTLa$!Rt|ak@nOS9abdjb;0s|I_%fy-7)2m3h6p*rNfLl!j4U-jQRSC zeP2pUD@t=giXA&DS?W)pB)bD2<%pr>&sd)dW2ai_%OpB#i z`!N>i4EIfP+)00_!ZgG+?)#ilMLS%+dZJ|Cz=j{2W*xF5_mwy$xYzlm!UG285@z?r ze2uP|+0CG?VNF%gJeZsP_L(=F)mm{=810rwdxDgF!!*m|C0Suul2r^D%J&po0{WE^ zkcS+yCE(ngsyE{JC2HGA1s2$u3vQO~HoN?(?}q__#{G@hF?0vX8BflP*x!^>ObJB} zP+@x}ZYDh>dt&GG1r_g|3|*kZ4NHrwL0%e!+1lRJtq+HmgIIHT0H@xVNdWrW~v=jUN%Lt3P7M;;Zb_#VzNS{=9lH|_?e^HO+o{2MwuW%b3n5@5Gt^4b3-P%I3-px-{zK}UM9MwJm|HX*hUsv0rX=- zH%srY9sSd_m$CF}xTwzWUT}gs$77AEO_-;IlJ&@yfgDa>5OCw#GQDGzN6lgSN9k?5 zOYAc5f85^=d(kqn)2 z@$z_i#&lGgK02*Va;)Ih%M5H832F<1oh+O$u4$>3ehlY+KdSoEUNg&5?Qn?}+zpQQ zZkKl}hacBLj1Pw6TnBF}gfG&2J&w|NKSi^bU|S%LB|BUUM>||@1c9*n6+nrG3Fbkp!@VME(8{7idq{~EQ zWG`t^*3xU-BYo0+N=BCIF#qK@SJ}0^-SCO!_VU~9ZdyP>cA91j*ln~EOZn**zbdTlAcPt~DGq%ZKgd6nLaeGOOfXTQe| zvWVBkFLAo<)@`%m;bR8!aP(R6m>ZC93*Va=VqC>GrYn-jK5na+{rb2_-H27}2*n(t zNI4aT5eR4H7oo66k8Joc%Fd$@8{F@4n0L%;pKte2iwnF?x(P~x6iBZ-Co%mX9lK{0 zPi_--&DCo+kyha+3l~vo{0KP1v8sg^=t?1!t{2AfvjljASI^D4KzGeeRHtcj=5^9} zZe!Y5jTv@N5tm~@VMW5}BGZx(T+DcOyxTrE0!WXFMIfJYv8I!R2gN{5PuOVVNws4x zT?r_`gvUV#j2DM8jTm+Z_cypy(^td?LWUfT^|B;Mk8^QgoHEuu$uZXb0VFRCI^h2| zt%X~jYv=<>4(K=vL%ubWR{{S^WDGx ztJElx6o3DE9a%G;>>e=%i4+4_hMOV%LpkUX2YVWP=4VCsLcsF8l?mWixyG=ou%77 z_c=X2GAvk;<1dgqbi4o3@m*l$Y4^{EX#B|W_Y^2U61|p=Kg6$=FA^>ILKQAr%^Xw1 zu3~ze5^pc@SRK$6jDy|=qu;>vCGhS<3+{vFzDQ*FD0p**`s{NWtBJsZ_u(na)QDS_ zz6wMAq9TuUOgD;}?hoWk$@cM}A`IXB=26TZieyn?&0b0FHM6>3Yn_6LLTAWU$E~gv zQ#0V9H9|;0@zgBdV%|ZoGtxfgXJQPWzBhS0tP00pzYUu)b1$!nuJ)>c%vj0_Fs!cu-4MICx?qx33?%y_ZEo^L*F2KJy3DJAdd9L80B=j(rc?@8n6RY zdd16Dyt-u4HrFI4y*dsSKVbJ2oe8LQE`;LlQ7{4$hhr-!0zX7KpP;UseqC89$P{6} z7;=&%Xv!cmrH7=~F|{lBQ}x~Hb=oVc?vOi*bK1qc7G*TAc4{TrNn;qPHKf%r0Ws!a&_^|_>$HDcOqGP5(^d(#yDuN#@tuqB=e<$pw!~uv#Lx}C zcM(k1aghA_w`S$v?kWC^hktKY=E5>_PHXkY`sfLaacOuoj}}{OW=ha z+Z0&>FHR`hgnQ|R<4@sjFCClePQFC5HNQ?50Y|q@U9_asL$~^L-Gb*BXqsw(`5<)$adKvQZwt!E< zc5}%pc+8V*tle#PgOg-?r6w--g373#d~kN$ITFuJJ+WgI8B(%_?XWT_CY>TFR9J*4 z1K37%y)Nl=j38B0O&UyBw8EqP9rgHn*>Pc;4I}o49*pJB*EY}f;i#NmUHg6W;(r`F zcY;x--1zq^eTL-oIv{!QP1HwC(ypK;6UYXf{fVs#iz4(libnR(d-x71Zvt0!%GFllv3m% z6;|SeR1EF%5?Cf6qtST}A51G?1*!U8M<`OtR6Di;nwo?L*QB07nHSA#7a!ucf{kQA zvELEhjS-?uu%fifqo6cet`i#m(tTpxaf}#}KeZR!Z_zwMIasVT+u>vlwTBa~Mpv&g&!J6`p~hz>u<=<#EO4GQbFjjBP%Cu!?u03uOxeJ>M9bK*iG+na{GM~7bjPbl zXFmM>kVfb}L1*-0(niXGmlI2jV|MIk&Sk*bxd2_Qjc*@(|BqY8nrropIWq?*qpm4d zaS5@1xi1AP4vaHO4$Y3jzmR@zlEaR5mMtSB$7+gML6K!t7)p&eWv*KR@Q)1qV}u4w z$20N3K4mt&GtD9Fho;FWT;zrwuOvH0;31A;)=^{)6^4c93&bssNXby_5ii&EIM?{y za)l~30FdXtsyq`HkA#FxVerJ@Z+(ARtErr@+puqzH@}eo1-OYZxUo=P?QJVF))SA; z(NJ58aC|fpbVgyURvG8wLxb-iIk^zWw8BpZ$ zVrRE$vV!*rvVyPAUMz@ksp2PkMnkAqC2;Jv(~rLxzzrA*s&>$nP?wFPR^H_wr$pARYDI@2L%LO#phkiQ zFwB=tNO04w;zJAu7@+FZAE6Z@QJn3 zKwr?MjFmv;?dizhaa*NJ(Ukg~@v8MY0vzRW6P_iE$38(8LS*b8{EQJ^5{OLv?#3%$ zGeV^PottT-fSW7WjyF8DBamK2F%=XkrNUzQbx<+_`LD7Bx@f;Rr*s*T`nS=oAPB)j zT&J#s{M#yU`b6_8rWVe@!V{QTl<1T^4Jbd7r-9p8xmR(AZjc_Hl%NJ$csv0YQDBto zji64_M4u(CkjWjV#Fr@Cw{;4T9F|Ne_9*r+aDR;Dow4w37)^FQdF>g9f7)$b#I=)? zjWGK5u3!F!Y_(&3_`nEYWKj%M1xTa9uy6^Mg2sgIa1pxycLr>ko9}grSv0YipTM*V zH_h3_tMOYhsolQb zNF8h3QDv&ca51ykv3Zt77mN2yrnC6hCf<{`1R+g~>3N5;7RZ71+9KtE0%IgvcuBI9 z=}5?z70|^;M>O^ti|BaPqE(yUPTMhK!difN%p zBNbL7Pk=!A9dBIwx-QZ?l`di0p{!%Cb2czvF8bAG$a+Fc&Jc6%13D=nThK$0jtT?! zp|(9xNJBzvsb90d9u~4hpj^uqY$ov#G?^5zT&ateWQs2EqP=yo?j?eiF3|9bRL@@} zxIPp54t;LS*W=&aOmx{0e-8sIlKJ>W%4rC{LoZVT1nno%6}(maOv&+ySuwmqd}ih+h41Aj&Z?3j@DQRm>1zG0X}th08jT(Aqgcuw zL<(@NF%T)}!bMAbi@gmLpE_KgHUKGu4)V&TT=VGS>uQA6ifht4zPj~(hovP>$z+`X z0~hjxvz_;P)QUHI>##(^D%ZyWcYFA!#Mk+lWDUg6Rx?740X;0Bg3DIG)&BDXUyg+Z z-i}uu78dxOE|+;J8q{y2oQC$t2Lb^^iIO2US@cO(@bc%zO@L)7XX>a*gcY8;$q?MZ zVKRV?PUoCH*kwwD_d-R!vC!SMe&=N9PL~|7X1DJ7CqwI%vHUdPR;c4$gILM}5FvQ= zIZeG`=C}aBMsq#tuT1&3F@OWBUQcp zB#Ca{>@MjS;+4viWQCSx%Qw~;7}nonkPI8c7KbkhNcOyY;nlBJZWyxEM~Jq$4Q+xq zPj3S{0<_-ex?OS76}YWYbx7Ofz(Jq}GVi7C#X(1TSKV`F&tU+7bqdisF=6vTDT^60z?O_D^D8-cOzMe1zvdcS73LVi2sM(*;u1)RfWZs~Nc z0t~ydMsmvqt-t$JhxoPPF^dz67t6{#a8foSO-u1(QVc16C%x%?kv|Db(PkD$?@dicye*XB7c%%j#1>WVUcg8k*&(e8(?_! zEy^#x7A>im+OOKn%LTSsJrgfGEJ<(&c|M_1(4xeWt<`>4{B&r)TIKriM%k}p#q*{(A51Z@ob3b+6=p7|mOB$wjSQx2og-K(dPUD?X zXXP0exlNkgQl!RuG>J`@q$iDXc5KyWS$K?F{_vn0IaWHt8m?H_Ko3O;Y`?WPg^t-6 zHtlHRb9M~hy&-*{G76b~m;N(kz`<@I3?=1$Deu2=bgZns=UT0%re~(CrWsve>s~=(k>*?S2Uu z`Z3lxK^^6t0%;;x%(YxLpvb1v+*9dB2)s*BZ*VSvQIY%9aP_KpktCXTOwp)Bnx1|^ zJp9BEyiU3}G#f~mqy0MN1=B6leau1GGuOeVHy9N~&$`HtvO0OH+LTk7%Vx-qZ7wX- z(2e5j%A^1^!2uUNkok9#tzNxK-DQ&QbDC_K1e$WYyjPhY+$3BYlse=q&fp=!BE7@s zZsHYrrO=TOqZAAN4c*#FC4mEqwIPu+N+7++`YW2%;PfnfaJ;|fZq_@`_95D_y29c^ zjGa^J7sIQjLFKLh*3hBDMEVp2sI-9b1%kFAR=d^u1fuLupBn(!$FcSuc6lOiIeNg^ zxIz{E(|fgKB{v(c9j{#w*F7vRZ41S0qF{c)2E}`&B}}rUa`s6eU8z!)gAWvgzdv_L z1a%j$-a)QpxNg~+V#(Zb!;Lj!PR#q3n-YKJp*EsJd-7++B-xGw`^!emcP_<1vePcy z&c`Wh2P~QdD+nqgamtIN#~F9#SDf~O-CozFOojNjBta@vZK#~y zDQ}scpsrW$^+KL?-5!1tuUwGs(?r*UGGcXja6|#C-eKO{n%|0DdW@>`ezM@V9~d!n zZsw(N@4WPe{EKt9m=NepaEk1J9G|j<(sG;hx=}E$1+t3 z`~ioOU$lna3XPkU5HYiVDN1L$v2XSYovdLK5?*qSGeQ zyJX5!{%-b99?KnbpXOLT!;-$7-gEoiNc`%zKhGydP~)ANN5Key$)l7vT8TTLM%|`?w#`tT1H} z2YrLtm0P+JIOU^lQ5BP4J?x-@#G!SYa>gb8n)d?Q}Ad;*payv|=F%9QMqcJY&^wbN;i=*e$WoK^kbLj|I* zm*egFcyD=|X~L@Ax|&&1P!X!7{B3RxB-R~@Oz}^t3*2t`UT|w6SZsA*dgjyyx6|q~ zUZ*tK(6U37O}DC+X)bw}$y#W=7TH0%_=!AQf|`?LFjW{WT@uXb#|^~eZWemKWz;@G zi-c{Y;-%3(u8z#c#$LW*LF>Q)^Wv-%A zyw0~ku@h?bHS+a9?zEFG6<6?T>GZ(fsgQ0K46cE|QuoWC!>kg(DkWUwp3LtQW}p zb2kKJY@E5)9~*%yQ2=HN9Mf}lQC6*Kx%Q&hi2)lOKR-BAG~iImkDpk|&xXpP&zvgL zM`Vq2bEe-KpTl(w_URwZkT>b}N887e5O2r+B$kBuAp$JSN`%D21Wkvs4k|zVBLHL$ zexB#v=poK>)0VC@<;P_qmQDT!8b8}rJ7vj|XuonWW1{Un-|LKcn{T>RiWofnQs(?N z>w%&?HD@p@w&CwVxs#51JMH}*|3?3E#X*mM#9{!G_-sH}d8W-A zjvF9O6@MJZGwx#kaJ>5%S!~D7oy4f89TY<=2 ztc3$)m^~Dh!Je;vODD#3CFE=Ukds z;@ugvnkk&wD!T-f&BwiM;JVeiH z4J;2q67vN0F}f6V*~3qtXZ_j{n7!1a+-zpLcigQTh|zmN%+Hb+_lkk2Bxz^RAPxoWC&d z8^5=e6dD56Sg>woc#jyalHYDvu>76!`5E)u-dp$9sjt_)_t}RFGoalXYWW8V)T#O5U;pe|^*`+T=M!%= zzOmwsh&LD73Q=4E^l&kJ^xH-#Vh2!v^Urvu%or~eSXlTkklW9gBiO<`;5E@HnggD# zilLO(TfWsp)E9NY0*Pf$>t#7JY=wxC6KVlIMh;-3Gj@FZ@( zU!W_22V2kEs~J$VgNxU2op(RdTB#GB9-)wEiqV-@lDq1YQcgGSjd)yEW~z#OtLvpA;Putq~XU zk|mdz^L|m@hrBaDI;A-GIInZxLE2r4#NE#^+s7P(?l93|@!WCm$W5peg`NA8sg}cK zTW!YM7E;Uq%F{d(Pk z8>V4LTttZdW}%RT5rB%JnAH?nf#p4VUE-8wv!ez1Vj$No_AZ#(N=I2+onW1w$2ZIb z(Blrz>F5U^u6-gS`Gwj$v*aBg)a-`+&E23QWI&oZ zj;Yh?4C!k3bcCutqX9e2!2i6uD~R zlqs5hs?|)Z>7>XP5BX2Kl&cw}9X=dXrLP&21Df6Wvk{Iw0>; z^bpB*y)4{X#hRib@UI>%b$#$ps zrg3^9#wj~{`KI)?T=voS`$n;`^gg7+HgJ6DHu4sU+UV8HKAEu$+2>xZvU*vrTdkjN zx%|9erliDaFR2d5fbv4fBd9|lZt}cAYUx5?KPhqg3@(<<-Ws6WH|LBrQ?$xShjHMo zKuwtFb#BfDST0+YM+t&ykK)X0_^HO0S=9MJBoMN|*!O&}$6ap!VPe>SnyS$(ads(m z1k?+!_<$}|JXH^?K z?mF$3)(F?V{;2@-@m5Vtp%Xp(cnMZFpjz#LbAP5!?ZN3~x8LoH=iCDae2lB#Cx2}z zCOgNIBO~p8B6(EU<#!>i{zebMd4_BH&haySqlvItL*(AH=e*;1_(_o1?rH-GNuAqi>`)#SQ+mfAtyhln41*24(Q(^s6482B1bn~89zGDliJAiorX8gtMA<-_fXAKmW?3r#Qz-a%%lE4E zOT{TDu81S$LXVxQ4bz$f7d&i^ryHhS0K(=t$gk&U}@`W11r8-}mPIM$2nWxdX} zu_->Do-u&p@gO!hVh6=2#;|d*o!{R&mh#c|yMTkCdbmW2r~1#g%eJFV_# z`tL?l*`YDy0tDr3pi1WJL8H6ow##F;Q@iq*f^EIA#gxTSg#70Bf7JcyXTSUTfBjav zlwuZBB%Ec@vB4K$8FEkj-`AN+5iW)yI|c`f%sC3CnIvhnAP?$GpxvQ^MlI;7GTJYn zZU9$lc1Ww|A%2qM)`_cUR(f4gcFyY#Do})r`h0GZlzFjA8-vDjnZq)}R=S*XXwmDY z!F*4;PVLxE$P%)LItCVqcxW%w9aH3skI%+PjBpX8abtPoI=>R|e4!gG*4<`VEFO*|9nTA>hM;`BEtc>RlyLVVg;9$SQ&1XPtI4 zmH5-BZxnwWO_u1>?2~p1+7#=8xBcYMI~U(x@#ela8h%m^Wk=Asi`m&W77S*vJp5x; z6Lw6DqK!}={_Veh?PtWv^0>|%vU@y`1P-5$5{fxMkwPl0Tv#?CQ&i4B6|%M+p+nqa)cBtpqN~WWK&_?6aWbMhO~Bxx_5tbuemZ6&D?WO(0jE1A~pRtFojL~u_YBc1XA zhnh*91domoyjCf^%v(J7nDs+l%>&sZ3CMb2ZLm2Oe(Y4wH`lVe%xukP~74jj^yu7O^3#}QKkk^;a{jEM#HpoGS{Cjqzrb6kA5 z!QhG6ve^Z87>uSO{oF6g$CN0Qi_gY>bu3V`hpXO?XjNejw5*-o3JjJ)bwzZ)^sFRGv;*oiEEAoDSfzCOEL}Hy z!=xtq(~#>kpR2Hdhi~>=Kezp=0TGAC(HmMtCub5@s zV_{ovQ(?Deb;0|Kc&2dxEDrQocTAxurJI$Rk~_lH6Cnbt)V+p>VLF2j_)`zTTsa}> z^ah`tkhlrDUOHu7g$5JY(|tDaH#^$cUTG0fjF7Pq17>yO1Qh#K{nc;$l*}>;laEWZ z6(oz>j@OQrj}s$gVll-OQDh(5{kDkm=vErjhYo5gy-tcwirU2>4<%({9scNm%$03k zx&@dJq^oglfh)sVyG$%p5I%Lb-)gj@cXkLj^xNj!p%q2?Dmalst zE>|G^m<~fD4Kxp9)``dR7Ce@pPac^C^`|uI?Vh*Ll+K-t?rO&d6PCh?@w|JAGtw_r zRf4i<_xv$RwN=)|@AB_;UNtu-q;~3RZv;ZZ#G48%$DK2yz@u7zX3_?rh(d00yuVMe zof$JAo-!w6!pc*|9qX-+-;QkxHj0{mDcqGr^6faJq-KQhI7Bh!6e*#?Vg!rdx++Zv zV;CC964aO)zn7Ow?~8ECMW zNRBBgAZPVR;1b?3#a8up(8j_=OPz7BOh|~!;2k0D5d5%RcF^ncOv8t?kG=;E5Q9sP}bdX|zm3c1}w#L0cjFk!sm0RUH z42G#!J~~a+AY%Yl9msP-`~EUP|ExiWGIhaJ{MSzp&TSuRp?BOVT0bFN6gw|lJUI7) z;DlW79PV<<@v?uVcMoL64mu1B{ckNvV|Db56*k8HoX|5)@Y`SJjD=6gj@K&|KB44k zc|0rzaFlK#U(P)4TSUcCMbt@NkxPrBPMiFLIBH=GrT^u=&~sXALFJh<%N0w{9MDP) z?*4u;EilR!pUvMrMwZ&KYymI!Fxj$!V%Ae62G!ek{|Z^DUyNYzjjPhzP(!s-&>%^n zcLd*%cghQBB zz?Qe23d8#0poxeY{TB(!Am5^s-YvZ#Xi@Y5r}A!TlT#t@ny^)<*VaMpX9H14Hl5}k zE;=Va?A@VE_o1o~@Vr#q2T{UnnJzv$r7e+LW5LKHusrL)0+etQK6h$!Q%!Z27s|WL zqM0C1OCp`bt(fN-L)z)6=T1!XTVu>(n2mm-UwKcUiP z&;d&Tu5!(vh_=`sS=qE(t_BXiL5Fn79)6rMZ}##DY3|Xy>&|$NvDda0E0zG_(OP3| z6bu`WU6iE!IGX;xQFZxc?yq`DjUB5lUyM*)8Y!lMBK1_*F6lw9&FTu-1#zTz$-D>t z;LSqfWnj>abnBa&14O#&Y%v<}Ta;UXJww<3Y93I;4LUsJvKtZ{rC!^awUbH$SIkD{ z0d$V(q9vGBm_B)l91f9g2;5(Ig2tL&mtJS&D{6y=IOw3mF?EvQD4pYG=moim(Db85 za?1sW@A$@MqDz^V3M8$mnnxdy=y_3A4ptjhJ!edKdIT)X4~9W##34JLMhkX16d(Sj zhY@3c*i*HJ?09K*QWYZ*nNKlLtZ_FLR>!-qxTok+9t^4CH`5U=_kv5s=a~D7&xAW^ z!+G)KX1YyyL*4{?E5j8m8k@>rOwxG;(nQbQPO+K-X|j7HvzZqSTr10g87rNx7PCTS z#QbnMu@c6(e56@z|81N8UDE(dmYBsF*BHS75JQzrNt8rN7f;sZiz7sdl6LS{Ck5y* zgsq7Vcgc{pgVGS?ltmf-xBjDi;_tHT**-Ch)h}>*f9+RWta$U6e-AJU7R7(-wvjz{ zym*}+Au7r!256g$sIUxPlAs1^j+`3OZtG>|C*?qynQG8;$=rIM z2m}plgCEcZQ@en*8-gY1kJ|k&@@yr7xT5HZRmnz3`uLyby)zaHZ~N86*eJYHfe#)x z$qAZr*y%PZGazK{tamREYpZ%)Vgwn|7y+{P9Q5du4muzoeXq*_;E9I>6Vz@JY1n4F zT<1rSF5!yVnUVvZ7~_C<)`?05cfdu3eDg>Y-$tiuQr(6|#veJ5-@Am#4#e>>X0?=` z0pnW;w=tvX33!=auncbh?KN06-KhO`e;{8ISxLLXH=-6=RH zZv!V}<~OcNmw~S-iHY*g3f>h`%I|}CtSZ4Nez7=DQ4M#hg;o5>86~svZmXi3uHxU* zoEGLmjNV$w5?{P~PjiOc4PGj+5ExIJHtT@qX`f)(y|=D7)_k?9@j-ONka*@O)^D)b|C6}6MmgfKc;zm#-;Q_D=SN6}V-$0kB9&BF9=HZ1-z4v72_`TXz)G{+u~M)@+QN$8L{Kk=e$Rq)j*0}fsO5iY5Y ziJs9A+qr&1PQZQFO|pFc$=BP2cWCrpUz4unh5H_(Yh9NLHbH#df@(#FAY4>DXP?u8 z9;kSNJM3QN(pNhKhF)Vpm0930T{p}=w;VL=iti|vUTV&JEe)nM$nADJ%~Dl76=+rt z0jqq1y5Ikv|4}G(QB6ntMF~!XfK!@E)l}1{#9)CwJ(_F+YFxYctC@;<(Y%igcE~*6 zE=ak;Z0Ta3>obR(52?WB(xNP-!$s#ww%Drqj#dSZB}AH5Kc8K%jsSy2%4ad5kwfz&<{DpyWn_(WJ+S))4@dvd+4m^Ra4k(G(B38tOS>5 z4D(jwp(Cbgv0U(B$5uO*>|RtkDyLtBsHrRB2R!sdHwjn!WCib}uS=JQoR~GJteklo zD&&CjqrL7^C@G$_TRPxA2F2tVGw}E}cZ>j79(T9d5r13!@BW`rz5U*4>TUArXf+$` z7zrICRNiKaxkiyIRM>85yQ)EA_*tht2Z5C9$Y$m)gikg|YM?R>R>N6Gdi+~u8T=&g z1_@?=qf(;RUY}k1ZoB7z!yWO7H`DlC!Fp}coC1&kUU)pTLE5G;T!AV~kUY>WZc)_x zl|VRo(VV|7JP?Za1{_l6-IK?N(!K6Va6687_RKfjxhvb_TMs1MhN~Nfu9i5}Gphub z=5+->yt5~C;r5|htgy5QOOIsF*bmsi$i8}IMz#IuYetmSzjHH<6xi`5s&)jTswf7+ z7)l{(0ahHW)sadYih4ku4pG0q?zX&|e^HDTIndXryC!Y+*Q>XMq(ML)#94u9Pgkez z0TSgsio%)55(AvusWe35FeME1#mW$91;lcC$APa*hjhIk?crY$>kQ6I>@sv)5vS~M zW`zXXlsvp{YsftO9In6FAv4MbYTUoQ|IZ(&j4=A(?6`9zo|{e2jxEHH1~<&6mq{_{ z6iK1N(&^39+candM&2nb6H^MiZ@|VjH`Iat7cqz42BcwxxKBfK10Ml4J^?G;phLY3 z>&QH;_sMmNp4HwO(0bNqSaScD|1#wpri2f6yJ=!!qCh6GRyvj5qU@Xwp>ROP*#g89 z$9XB!byvg(Lb?YO%;LdhG6(k9%b z%9e~9p9LDFir6u92g$MH#p%olIZ;kAB@{UTC93EWKO`b6k?NMrjd4s0NCax|8hO1e z(d~4g0UHVGTY~j|x^mBKaZbplIh}56oFPuUlr9QHzLR=cIM6U-D5ws-N_k8^z239k zze}0mp68P(!G`;t^@^o_+2TwIP(+Xf&1%L9Y(-;&&{I?yOQG|yA14UC^5dI7*#6bp zX=TYuC*eQTt@YY`ake;rf?j(UbZY$9BVD3gt}KVb1bXdq<>$%*MU^U=S25L=!syY= zaVkV)gwVszHvFwCSt@$bt-qLlkLt!SWn-95g33sOdc9whe}o9bNp&}Pt6ZCaB0Aj% zFILM7p?IQh^K^rjf)#5qy9r&?7?55tbEDAQmPyY8@;9&+Do_OgkonbEh!dv3&eruEue_a#x8o4o^EPZ9T zQKE-p?op(Z3ab{c_XDC@NBo@%o)%rX@F<<&zHIWIiTz-%>JuCehSCw8bhKo{q-3a@ zymxAXI>|AQr>lXI^hk=lEFb~49J)?vT);jRvc`d0kmy+>$3(?G*HRkC$&}ojh4F0o zLDvF%h~aJ-X$9v`JM2atex7d6#MGe^9piZBgQ9>tfxTNYa#LK-7Glzs;>yyn}D}Qo2TE55W++Us*C|PfVf>F1#cHe7|D* zkA7~1Uf=C(8EF{J2Vuu*Oy3A`bDLsrQRD^{*6CK}RwDvT=R8u>$`T{yLH5rw*P zISnO%TZXb!F!mD~q9&X)6oLgg3J>#F>SfE7Ta*dvXun9+#z|mMJ^{INtzZdCp?mr0 zNY4*GN#f@WeNJ28HZMc#NL8U`DA6I6E}Db!6}W{T{8}>6C6gBQf?f0?#|1@nzF4Pg zgnHB%2NcaqcS;!o{KMV7E?2U78W5c7z~ER-S5857Jw=l#S|U$yF|+{{{{?(IzA20? zcQ*nqYw(BrND?*KemjDR}6@qL< zt!t7r+_y@l*KU+-lJ$EeN$dOv#T#cF^J;+=6W7ZoaiiZu@2G7xq%q9>Sf}@?{W&eO zl)NZ^vMId;m+iV83zvN(gv%pltlO`iQzS79dii^$O~ML6jcd4Qz+tm@HN>N51s@@` z&Lw_#oD$UC;v}Y=9++{$O|RXqxhn0IBR?p%EAh(;wk5R204$pZVGOuWmrXRTL-ns+ z_?X1LV2V%P2tcGz43xJ>pu#YQ9Bap60M6x!D_mQ_yRE}ST`bF$6cDGp@0AyVd}^KY zWP(yYoTtbJ{H zgI!FF5;612>Uq05%(UtNSC4QtlSpUsF^0CC{-14bx`W^k?v%&Nj!caflmwm-=&rdz zhKTH-IzPAqd^`JQuklHwTLY`4D8!)>K#Ov1a4CI5egLAMs=<)bq`t&-fq$n05=Anl z`HD2CFrYJ#V`71$0mXjD&mdS67#<*ZF^Vh_>GqPv6I=tLN?K`lxtFqd$-LrA7 z&M+d_azyQ(SEO(uPFdsuzv4YphMM(E*`m_zn~j_i2PSvAVRKww_K*@mJDn*Rvg$NS zGes}pzIButwqMziB5s>(`~AyE;s~{8EyYArWF-|=BS!xT+}a!b>7I-Ks_C_>lR4sW*cJKW1Sa^x`O)!UowTzvYc2OQ;(o9R&YHU$jS%% zlW%+Qjc{@Bbci64FAZFFjDSlV#jK;q8sxWs04{(QgDSot$L)COX&W)q7b)gEMLxxuM!pbaDoaqe z@G8BBvM7^ykdF&`2dEo!&PY#aHao6V7Rn(JVTo^vpcA~UjY{N2lt$Rn{OKP~ZlrO;t4KCPX*2sY(QW=sI`S-H>PS z(8C??lnR?DDCQV;Z22K=ix2Tj3(WuL;!@YblGg(PLE=7HxK`^ z)d8?q%zz5mvFm@1B8%ritk_8+3|*) z#U0fo%yWzJ$y0)_NgGR2U+<;ws_yyU6IRRnJusM}lfE*4D8A-m=sm?n?K$|_B-}7* z9l^#Yfi={cAbN6&#{jD*2eaap|NIZ$hhqt|vSY=BCCmyP-}SOMr*v5~PuE4dlpU}s zYE`s?OT?hnW1e@Nx>npn`aL!XqWz+I2O+u@!s{UJkGEn1ly89sS4gI)id-T6z@*gr zv;qj0&(#Gd9K+>eE{xM9WF!kMkI(&^WoOT`y#MzV^3;FumV+Y_?F zbxz{mex-GCQJEshxa{y7L9kIKqpcs%U2cghl$TIgB&{GO0ui&8r z7G0tcq07hXu=B40Q38>9K{3pE5KgV?O5_4ZWd;ATRfK*K5MffVBz?dc4LRYC{FcPw5w-% zQby6z@6vyU47@a{QCmj{ooI?#Ns%Zj49Np@;l8C1jgcuTS1e(yO2cUOteFEJvj?+c z!?~Xq{ms;?WVex3EPf&k$l2k7l=Q7&t%d2%P!D+L2geHvremVjbCwk5%(P)0u;MhD zF%&03G43ZP9shZ@QBWi*f0{}5zBJy(Pe%xf!xU3VkuoZ*i=W3^IcbyX{-h*nk970& z9x%4W1zeS0nHxP*uT7-;zSiz}#Bn>l0vytJ+)ANF&vDTf*9{Z*&#jn>Nd~2U)pAf@ zL^nVYKNu43drgv}EDpHhcUZB9pTKVd7UE@0pR#vuo==;$iZn1GfUm ze=HSo%QeQa{{INyhdq8y3xNIZE?d^T{U?o4qMQi(-}NNVj^iDwN63guihVk#_h zd^4nrRJb>|UvNL@wRHRr{z0#5MLPXO;NtoDz(9WtYMWf~OM=X8Y>}YO@VM=n!8<}K zhR%v0Fsj07O8_#}BavFHdygzFAUPnO9VJdBRLUb z1$kj@NntHjh!eW06=K=(uU0NwlW-$%uT&qBqe-O=B@8fU>4E0waBO|WJW$yBTRb^_sq+mNXCo+z{^`A1vXUDp>{xO@JohkG z-z^lgiGo=OJ4tp2Mu_6&!~Y>2(RH5`&COZ56PiPsIMz7Am8&NXIz&raWvzj^ycM&bk9g+vV6C+S zV1J&E?r_>9*zXJGn|lJ0w%IU+rV_rpb_Rqys0W^?h zc+^blB!|7LAy=_z&MGEdvWMR!yysjjE>+t)PRVk$)W(%iA}3XT$5fEKBm@gA^3YIR zqE6iou)wk>Iz6Ox09!bUMqnfZ3RFWGbYNR;QJ$I9Cw(pg(E@pBV`LaaVr1l!+K-J3 zP160NQ(}r^9kvg#k7N%C&DRCCBf`?7cl4?2H7Owl}fJGsiH|Wqrm%}y_#)*gj6V&_Z zWsqR3hayJ_YIOGxI@HK}T~@hbAnbm6H57fotCe2MTrs6zmqL(nIW8cc*TAcGYz@Tb zrG8ZbRRIPV=;D;8#Q4FGO9F^MPP%egjU2ym8HAXmdzCod5_}=X?`_E0hc9wlp6*la zSSoB(){0Lk{x8Mj(b{Q~gDLL}WSSBMy(9vPg+eG&5DAgp37UA%4ycQ6Ul*tXA$$0#Kqzu^R=;3v$mhb2Ax~k4 z_1PIL1&7Jj+Uj6oSL!R4_rGH5J>()?>~{yimO$3**6VUDpd92zrf8QJIz-E!)0S`8 z+lq(gfW@TCb2&Zn&cA+cYIU*OcqbM+%1LI+Tr{>C!}05&2sTFNr_)%nLeBu%E@t`I ze6e^uU*=5qV=3^spRJGBCBrDOk>fr-QtWIRcFjdv*zxki5|CXj-sjsmw^=;s@%RWi zIj$-XdSPw8Qu+d<0Ns!_NXiNRERa@Xs!vwR>6^-K9*=?FSZT!~_ zSI%i0Nmg(UB;4bIh6`1lwpTjnaZYOuU~)s;W9&Wk$O>?b{C4yErY3*86)!AC{6^)a zd7yX7j!H`W*1^2DLDf?m4FS+Cu!i2=f~Wp6C}-~Z3vm%ufd-ue5AZ%7^t zxd&uv@<>MK*0l1!2>izxdf3@ z@#6M?(NTc`Me#%eUN{QqFsShV{gR+063q(<8=d`3Kk}XrG|xA`=lA@M@3+hD%s$qp zfb)D?8_3SrIawI$=JDxKmVIMRiw-NDV6j0l^YgtHxq|SgBx6Ro8(;x8?ltrB#hLVR z21pYy9>)y2;B(2!nuYKGEhD6KW`d+!s`Jj6vFUYo=!InEk%f2}F5MCDw!eZFXUJHw z4R%J%1Wu5d*t2?lkHEC^nYsA~b!6Ejkk1^xnb<%v>nV~*MJMpiLn@$Kx^4PpC<;nd z4!SILzb-_r)}5lmUmf?33EMw^dw;^u9*@)K@4PA4f8TxfyOxr%PHRi7h`=6G6!O}c zd~p?QUxC;OdbPAF4AK<aj8@9rN)1Y>Em=i1}eE2W7-k5A!T(14sg};*ck-}0xCkn<)k?;Cf@v z7qppz@+;6qB?_=T6lhB^<5LWo0<2*}eWda146+q3CL>}8B-l2Db-cv?=|47GVuQ3C;k`v~p8EbD z?$Vc}TcIL)>G#gV)5x;m_{rxw>*HlJtlNqaqG{9{-e3LBtERQZMWWtIGDliloH%B7 z$P71oC$NRS`Cly||6En2t+Xh`ve@@W}sY$xGbVPydtD{AzMYSjml70gro)JiaPjFr!8ww zoE8^WTIz3yVm`+}zqD9%N8A8RqvIJgd17B|g2a=@$L)v{BqsdxFMsm8X(rD_y*j>< zY;|JS_Mn+-yNhCSDF~cI-xRD5SuCoV22EWLJuw`z)f+FNFVZTjnbxCND}r0CvU*vE zTc51WBN;@pl9dS|tG#YQ(m0O7|Jn_@Rp%BY@R~f!r)4={OKlTgM%d(7T(UckxZ$Pn zM~{w7G2unOAoLPRbz-~#`N^`sc=L6J-rt043rO zfz_(EdUmR@Ts((|9RBoSMNQy&@pAc8d;+S~NAJT5w!&7|Nqj0_KwRPK5CZyk0i1)Gb}w?of7IfXmB) z;Gmzm{3^(_{uFl=EhgFAI+UIFD?pBztyX&}25?qHMRx()c@wa^K{Li1m!!>_0-#|& z01;kD6Ofpx=N6;|obsuQiV@ra3YDsbgVIiNe&*%r11`%n^}=Ny11_6D%4`Ms%NgD& zpEKgz1r@@C5S`m!5_mV<)5G$@%9#hDi@fX^>>EAKEIb`N9Wrxv_<5`QeV-5$a^jP_ z3(2mLg8$AN>>uyD=X*+ygv7L zP#sgN!l&AmSc(PlJhfJJ$#1Wk?F>!e^~ri7QF*q;qfhou8(ksS%MPe_g@ezN>sJ%l zPiHKshGe7mm8DFan;ke3?23{nz!@`6tgolv>izK#zh=VEXJ7xu&q)n8%j3kzx@?B5 zMv8&j?2i#yxe~xr}g$w9cO^k5$my2);BXw2dPi0QBLz4e~B|VibsG2^u^zA+CKIT4NRQh#%_F_(*lX6U$_Uw{35(v13V+cH z_OY;O)%-3cZ0hJfF*XCelMNlFn044j)(Map)5w$WupRW%Gin<;hQ7&mbv~Z{V}@Pf zh7M}t#am>^19s7!Aj@Bxbtuzp9ok4SNfcR&>kzaIf)1ytfgTkZkcPhrSQHEr_883@ zPh-OgN5kgX(TCEN1^JfYeJ?YQy?I`S8WYZJ8eQO(E6D|_hVg{L(3IK) ziJ^zb@#km1`=1WWB5N*?aN_Oqb~9s0d*fqQ7;K@ zBi(|t;{A?}5k5VyHo*cGoTuM(?3Et|h;O`aT0_qLFgAkpP9p2f>Q$CbWEM>%15|WB zy*a2z(hpn^UE*xs{W&$lI=WW6#WgNyKXp{FOS(F08Z(j=9FM&<1 z@KjI%zd=<35#t8>6Tx9ns?zcDc&UDS;O|Z3oeA8)M|2=vej5ETi1Zc(BnCCOHp^F% zvk+HKqc<2LbuUN#it~Pl1rI*o`NeW&GU#kxs%U*kEq#&K6N&6$A7=B8Gw1kg`C7je z;WGDYWL%doTMZOC#f~*e=A-}nU$+h}G#RMx9Ql8Rq|AxajOWZ&qK_%&BZ?fQqM;hk zH8!Z0T$AfST%0w=-}E$ktH>RdB-uX?Yw3;z<nC4u)2ei?6&}01N+6`tsyscV@ReM6 z3j-$xq3}HfNqt6#D=q+evUB2eWIppYo^jeUO^h7uso|O6^t+Gl*O4WY$TqWCTTd~G z6j@D0f8@L0H_30eWC#5)V2jryK|eoXev13u;1j~peSkLp{~7Ty8$bLJ-*x0gP9MJW zE3aka`4UqLpL%DKi6I-4uP9cl99{{zJk&^Up@dAFKZ{ybrqWsk^9 zvSYH+J#xsPu&`XvrR)`Viz~#YidN06ifXZYR@+9K`1}=U2)B0n zdG$xHO?}aOM5<_*O+|vFcN!ErwPS`*l`bvh>uI(k1oDV{WY%wa*n@ZRF8ABcoQz}^0O>)c}Wb5m4cY$uypA`f32!cr57eELHS3xIjAOhS;RWB ze;)7zlB4tn|78(Nf)~GX<&|omM(9sz(QJ~PQs+&L_0bDIaWu8$XkcYDE2Bb-%`YWf znf|Tsn4Hx&4)y$m9D8X3{MXEU?sF7#h9XTMJq5fxnXmyW@QM@F%t%%iLu3fb7dq+H zOdsi%r^t?zZPW24aOc89+wnFggG(Ye&*)TRD?M^;p9<=Sdfz0!p-0-`tlxbwuzZ?v zlZBM@g@Letf(@8KffaL9`Qh<$wv=g@m+oPZ>BJ;e^E^yWL9$9#Nub!F-FtDsZ8>h) za9?Jl%^N+qp1*Qgd%_lkO-ijl_U;h4b{&~QZoD+$iZcsf_EXGf6zQd+PlhdW+Y5*> zW{-f~i9Q%u#IKx=H8Puc`vcLf-}q`>Ad4GuynSDp3j6g7A$KE7B9oPU?)3_->aFS- z%fjAbxAEhnYJ$!&Jy1cuc}4;Mn!Fbl!8Rd|)-6r)%cNQ3uWN@KRTW*YK>Y}4A%U*(P~SkS^@7Plb4$htq~F*R1^6?_cz0gM$l~ zP<0D$eBJmoH3%ChNWYBQg|UZJc&k+U?y180uG6 z^J)Tb)5|CI{8QW{t?E(eR)^I{jM@CyVoOHx3@4X~3ftHEz%p#(wAC?I;z?MYUL24} z*h5pd`@wSLg7PxU1r_3&pmC2K*oul#OzfCm;e?6_vv#!cEEx-55=Yt!zhMEtHxl)! zD)`X0B#9BNpWG$J73N@I?3DAoI{J}MUSxb!7RXs?Ra;!Qc(!Pc1fh0fH+^owL+K-E zbxHRqcUvaVE7Cn6yCvDfa}Z~UtpKy;!Qk~VA<4?i^L~-Pz)~;KX$^@L1w^c&+5u!h zJ(1|O7@gP?s!ajeGBp_yP?T4njly`2P-qWX}2<0cD&5Z zu3SIax9}fKuJMOC2d1>s{i_TPMdbjY3c!%M^1YWoITCdN@-g!Ex4reO@w=<>lPD+=HZi0gunh=yL zGPJ?#`8mm?o}(7Pw_0uwQ6{LhNK^-oovmRC9u0 zpWk+MYrRExXS~!9qmBgftd>%cdoYm|Hr57H#fHbPFtSjG&jzh zi4bo)H~)(zl+_?C@kL#kelR6@G8WU`Es%^TiNxJ@r<%p)0gT@q24b2#Dvz?!hS-9c zXFM|sm_K7Q``!Bb9~S+Or^%@NbM@9UWW5uwSD-C3tQ9+*Vj$7Fm5RRVjifY}A&wCv z>UW1`BV4F36u220d$VU|g|SB4#iG$OwiRF)59M07%y97Zk7C|UYm5Enc%+fInYT}; z?%eEaLdhfMcs$XL)b{4QNqSOdhMatg0hPTSRCKd&h2}iZ_-~U2H5ls^9U9%l?6*Ju z*L82F|1j_E&2MjeyA$)1a6Oso_-@L#nl-2(k>h_7LUmeIE8aWivDH(~n;O=zZ4D#B zJ&IF4ae`w}EV@~kQ0RmUsbA|ONG~#UbI@fJJ!zJcICiKELB`^zWp+W?)O8n3WWXqoN%~_kO>>R^w!o5; zMK6ME4{q$*>880`1Fz57twZmZ0MOqtpZr#oxcuMAnSt zoj7sp^e!_9q)`kQiOoh?d8DMm8ktO3C>rEV&}E=imAl;n^{Fe$9IzNwK{+$FMrDQF ze5Kq|KTcgs+;CvPgkvV&VpQu(#DCX`3&x_x>b(1VVnvO$(5+K#1 zNEEd5YJxH#WV>N{F;t;cDfUS>%q-@WgH_3y(dy5(38COXs1GmkUM$K{vK{RCwQe?{m05wg#ni&@$Nsz2;Q5AV>RrRh3Fxt`ZU^Y2SPh;aYlKdKk%j*>-kY-4eKu|^>zM5&3bw{0q$^^_!#EA6tQP4uss;1gF&Lz@Y|y(pb>+7*=*afWM1NK5@d=|7%B+WptO*0M6_3 zS_x9?!oaKHZF!Ww|y<%G6{NF8qi)?b@1+LW0E8amd!0fP{ipEmTn}W;` z7VA0F`Fbdd(nFcteu;isI|Sxp=4aFO!1cqPC4KHlB!o9`$%-TFT8|39WaS;N-7~Z* z&g?%n&&F6M-{xP$@j-6u(YJrjUi_K~Dnm?FP8(FYYzCFb+*4`vY0~YD>PDzm(4*`M z$l)(l_R4m7B4$AK0OjTPO&|2X?0qY|Ot{#PaE4pLq`BbwEE45PnpBNqHrqFwz8l#k zhkgc=GIg#fQIzSy_9`;f^$@PvtK1@QS86@_J=&owvz|6Sl@*pnR}waz+=oKTHet2r zRs6Luc26XXgl;(p42~649Q_-cp~VSQ6C*FZ`8&&IIj02`B$tK-X4X;68j7q!;@=X; z9#+nU)uhP#0<_3tsY9+P$I|AUi+74HQ0ex;(yavNXSKPWdNt$cY!F zD`uehgktI`a)OH9L7ye4pr-SV<*kCvlpcyH^x=km7nWJ?49t_|O7i`x)!j5Yv0!{2 z@RQV-hCWJ+@6;}=4Zxc6G>ICc zpVC8Y2t8r!rb-5N^CCVhec5ypREe$-VS1Z;irYq1akMBSKF$eu+)&k+`1;phbSXDP zl%75a)mHfn7l|6=JEaGqW-nK?RIc^Vh2b{13q)g)xu}A_(j$#7pY~1*@L7$6AsQV< zI8FDc9_ECR2@__09x&TvdG^oz({i$N5&_zUVOIYj#el@`9xD3&Z2dG{*j1=M)6@5) zopeP=qHC-8Z#}{lbD^3F$|q-Ls#}<1HUAjZ!6Wq2h}x!X%fA%mYCL~c4fUB`-h-@aB8{RQh5WcM(xUzo{%5!R3y6Q z`d;+D8j9?#4mU}o#kx5%?SIzzrLWJLWim0ZU&&cTa!1NdIdOWj&deSjpqRZB*-b^K z1Yjz%B4m{h$hU#J(xOQTxcm6ZO=)i=UdecT<+68)_;Xq8e8@2H>L7b{2nr`q^*nnf zlwnDi%9}Ll5S2Eb>*+2i2CfK%${dXyfiqf&Jh6SlkL(8(>*nx8-?z6|`Y>E9j1#9k zt%%oxV2TnP*gENb(5ijxKIe#ZVOU;;{!g-U8#$+dXK)LRO$$b9XglI$M<>ZPk7ueg~UQpKUoD;-JM{13v^Zogu0C#=ag-9lZ+`xt(Z6}~ z?YI6Yr5P7OIk9VEC9l}$e#?Eis1))7dq|bA5h5{HNcPNZI+0*gQ@1zBLkSQ0lm%2l z%Oi^E;0`)_X0r0FTG&o!PyICfiukHLY2GTIw~hpLhT`*&e6rLys@*{~79c%=l>OTs zt-B<`3ICobGc7mbyW-!HL?;f`<(n-aI*Qp!krXQWG1+sE(BNX=?i?Xd!9S|5qSu9_ zzz!T@XsGlz=#rs65P%VuaoEF*wvxe-CdDwebvTSrygSkVlke*kCWxd>{mt)4@(Z@* zDKUdhCdB|tbQ*+!fUxX?tI7g{Y|<`ixqDpD+Q=QyO1nh4SzQmTzc)i$G#e#%+%C;( zSM<3bbX^jhsooiN%4dsT`SfZ>7ix~)lsFbX*0Gbc@VCK?2`BCk?w=zYCqblbc=Ros zVlpV;*o;o$<$mLmbV=~eZ(Wii2lH7l;X2Va$!1laJ4jdob5Bp?y02W4Zr~pmv@7;g zC;0oR4g6Jtf_Yi$Z4gqg5Zm2!n|*4d|L`5sss8wHv;N_a@WplBmA+T^y{(oM9$Z`< zCoXfbLin)*C>B;Tx-cY+>QUy4(Ze+sayk~skC8L%z(#Nndt-bK{rZ-Cu^YFDw*EDZbOvFORYf_!`IE{=K`E-_FPS7>) z^69m6k26icTe*Y2DQFd6do70-LtP^|{`umvh<<(#zgeCs$`{uGrPz}Bbq>h2*!Xb| zcmGB@w)-5sB(~&IUP{{+OZ)MlYMZ1+wnjPhGg*0iYOl{CQN5x<+$CN?j{5GWcE}Rv z^@4cc23R=pxo$e?A1+D%epMMm-BYv)SIk@T3YJ@814_#L#iF99jj!&7{r4`nE+r0a zSEFR7FM#1oVb5>Dq~1UMn`N7c(-yB-X(?GgsYsdRzDF7t6gLUimJ)FR|AIJKSst=) z`aa*(uoeyGma$KI`J`B;V^-WGyH2gu=@`YFTOG@uH@F$Sdjm@YEa`jm%vK~V#iUYX z6EbvMlA_LUI#~yTtVID^0+I4W>(><0qPZ-sAv@?o&t|WF_f!#H%M*;xG>l>gYR^yG z`dHlN!D&;*x!;ei7>ibk6Fah2WY-RY{Od{eLuspQr3bq!yxO->oGW6hecP4Sq167c%vU3YSCP0Wq{VZ1MBc1Y@@C$B0g>ar=JwIsNs9S-nrF}-fp=!G* zV9*66Y-(w&f4dTLB&bKBbB&`=IT*7)59l`W5gzj6Rxd~Z=`K5PDsVI!PjLObBTJle zJ?Dh$3E%$xX^EwbpVRuUR)qP|B$eWPXcDz|b$USvOi8>yG09-0;y;s#yt%7qXuoMi{J zGx+kL%YNxi<*mS)pkm2Nk0M2pH)IJbLe@?x53dSZ<2RsK8)dB8&z0yEsq>e)YgNY- z-2~&fh7tv102dh-3;40Gt^ip8?n%YbfNB`NY(IP&H_yDn2~-mbPQ0(TELw8frN~Md zlBonXCZa)J5rToSo39{v#&It*G$FQOV(97d`^gCt6PMq2-}$0r0{eWhK?AG$@3@w! zcKd>bIO_U;n0O32PMDbRrN7m$;G29CmjIW=B-V+2lXNrRB$;B?QDhAjZ6wY>N(<}Ccv?V<<^gY~s7FyR%<>x$fV<(c z*kg@6U5bn@X#tMHkKVSHw`&mZ?sI*r4%78og80t2hmJ3Y#>2xx8VpP$4>ovhmd=-#FmL0C5kLHwA7=xd#fx4I$|&SqtA`tA2U#`0eCFe z;qSWJa3Cna1ijif*6k-5PP~NGm@Q$uDFzsq^Qq`;x?7qY)+Q`wDu9#etaulb!er8@ zsMZc5Ah?jhMjIIrR72+_k@!X-zuO%_kPx!hfBP$mf&#vdhqV1E!i&;2koN=;YWpX) z>;(jS8!m1ZtL`TU6=PwnbKYVuYm9Y&y{i1%Jw=PAOS;7~lb=s_)0c(EdHwF|{9{B_ ziVXK^K{dZcQ{=tfpoq0dR7Y1SQUbcZSInzYoOW=1-x#cmy~B(`MZrKX63Vs8HiYd7#1-n8EN1?=w7la7&)v3| zPnV<4tlO4vy3BdY$F%(X^wflAlH$b654743>%GsVm~4tB|L9OT(&1cG0f=)72 z-y0XSO|nV$$iIuOo|Ov%v1xRzeNb*oF&C;yZYvWPhVRx z%TZqPD1c$^R*XL|Mk)Zl)3J4VhzTt5$=!uy7dJQAiF*l;n?b9bVh&JbuaPbD9C#_n z5mggd8B{20QZ)gT9?+M=mqj$XmPJ65*^EI@`80H(HFzSigMrBb9pRmU`Qfdy94W{b zZS^+r5z;3;%R}1uYudurU*5OR2w%r3A`}9a}3UP=!$7*P=-ftP=Fo z^+8KSduQpP!R&^zQ{5MtAD$vw;;`V&$aBmAM$=DvOsNwtq0iV#002&(r>e#W?mp zt?ElYWT{W}lAxr_%yE6JT(ydyFU|xx#|rUXs3hMS$YSKJ3dDR*O52r(_$7f$!`SB*&8ksCLCHS z`tyt4%W>Jeg1$r72+jZvdbMY}@~CgRM~h}PlkatIL6fFQUKP~>1WMQE92V5es{|eV zBwOWD$M|uNV$R3-8*O{t`XV!epGy7l_Nk2%O?%;`%Y%=AECHthoH*tNEC|EGENduc z6-DBy=qIF0@5BV>N=tp397vM=j487TKhJpIQMbave?~`{5V0{^p(CX)O?0Hm3=u~u z<`4za!qF%Hmxh=>2Fw)LqXswIapNLl4@tYO&};G%HH)k(n55RKl1RNG!RNCmb}gLNhpgqXdzE;wSIPE!v;-{@fe{Pu zfizMA5B=k^pt7K`#7k@flG#GT>15P6`&rxf62>BC=EPedD`I9tT|g!s)+>MrRHVjY z|NFcoKgDZ41|NLp-7s3)w?(1 z+ybblt(-gP^5K9BvUVl$Y9S0?LGN>KQ7j8#@sZ+yL~zw{kE;{(5#-x$SFRD92)^XE zS8ZV1M%nMenJ^|SKCwZA{-_t93F&40UFgb;3SKs#gOhJD(!--c6sIlAgxZ`P9+n+q zFH{89O4rzD3e;ohQL;eU*dpTy{qt``BuKjCb^(^{a2d&FjC^f0@UVQxI$TDIrszfQe&}a{P~PB= zONq{jAyi=op+brQGL>8^dRO>CwO(c*GQgFo&%i}n?uP3}HoaK13FwGH>IFNQ*N0@n z_80fZ2B9EhLG3#rRS31WowQzd>y0%~!?YK4;O!;#U^YAUf{fXqb~xb#nTh}Wj%#6< z2{MX5vs+2Q3q~V(!VE^06a%6B{Z#a+*Dp!eMzv`Aq3HTL*}=<;tPG~XV6 z8kEy67L?9A<+D$<%54X|+k1UT)7%1bC3F$^C%N>-s3Spb!U07wuW??zaEC`-U_J0N z-2>OS)NKbZ3#umXlS|UYz;*V)J;jBPBZ}V0V~Q`RAzIy2pEZ_J11ApQS}8R!22#+o zJ>~^<_gFEx%F}qU2ukPEpqAe*Sg_R?aIr+Tj%knQ>OU+?vLuwW!o;{p)E4mJ`+sSV zSmqAR4M~23E=H|lXmbGdw{iV6xl|e2<5RME3~q+qY1x^+`IY=$OXA{}B#Uf?(ioi? z^kvv%QF=OmW}GM!)GW)V?FO!uK6e)B_l*%-PzsMK;teParEuvUMLgqX+`)S%F4n`+ z(X|=7qj>PY)_v_4mdRr-IC0`2oE6%@R@n`8shsVS-=BlciP`iyK6_1`!#e=%UMvRp zwJId}GI+b#OlItjcf4^NB|=6q=iFY&^p3bJHuh3}C0S{vi2CYP zCVXg)zgb4My);RlN;8-jQ4EMM?4+V|<|WLlcfU&y{Y2)=I+FOMN74_QG#}=^ad%;} zCRu5`g1sCS^rfIRN@HtbQ^*=+T0kxR-+z1m!0RoV!SFPHy|8@xmcX^IC-W~!?QCX7 z0e#EG$8>JrFT$LEzjtz`2{Rvj*j`34$W?=Y~Vs267 zI#7H;La3S9%QvQK^fWet_V7{hDRoMPM<2klPbOR_v0Em(9%=x0TnM+4;&?^3+Zoda;tH zM(Q8zPGym5?0<)k&@f@P;KzJs$LH)%&nBU6O(G9SSd>Wp^%keDs?u{(AQzwxvII(f<7<$Y$fIjCb>r0RJ zOsMh|lTq=1xBM-#iJMVz;$^1P%&6?37zp`qr=pYmx}bhxt2bM!*v^=Q>{56+s)3*# z^m?Cb-ivryes|rkDpG*fhs8udB{>##&HEe!RLd(pppl5*F}H`m+Ho|6V+GN3urTV| z|N3C<`6(u-=of@uBB_%g8#jF4v6Et;d^K}O+d6ehaquDa#kr9Cr&kFElrf?$AU@nC zO#kM2U?W)NbsPMO`5A;sr=gBpQL@&uIkkm|HsGp zF89YYvn-yATB3t>Wnsza<@8l5rlEsl1Ox3n^f<2!$##WL1Kg+ zB@{j=YYB`0AlWw7WWImYRWEB&-L+c^Y>bE=T5+EIv=gW~A$p?bS3mw&OP_{IZorAl zx2@DO#7^2rZU)DQmaB{S%O>HLJcrju4hCw~8ztM-am47mpw!I55(>t7*G z@=FbC((D9`R<)j4GjUQVL7~`q$jlBU9TvC_O@&qH6HOdVB3IwoM^k_ z%}z0!U;L@U$1*qVv?LlxqBx+&;2;jhhI_lK^<=~a0q7eu8~@L z`J`K4O83PE=s3hyaJ`}flp!2#fJTiS^QpG`nRV>Q)Boj?+~n*Y`-gwLNY**AvkQW* z!*WI2DF*2Iw_vNPR&CZ)RSOcLF(RF) zU7@3~y-%y^m2KdoGP+I*gub$FP};qvxIQO~f1SMN_%Eg5+ccVQrrp1WBMfN$n>oTn zIIp=rb1ETkEN#tB9E`Kl)@=p8^1(crx^b4xFO-mgxZFQ zC(n=9QQXcoD$91mmjn}58WD3Pr}*uvp5S_*%BT>hh}u9LMlY)*=&;yBsF2gl*#-;4 z#)i|lV|{q~TVIzXsCGjIQ{y1?FEwC03GYFe01~%&JsptcAZ{%vP;$TA; za)x>jlPE|)zi3W1^dIk^$1WCidnfRoDhL_t$N7x+IUoRG@Qjh{d&{~*^vmqzjwUIggT6H&_3Kfxu1Umy^qmtu)()BVx#c+zz zdTc?+g@mAkv11ELMhK-keDS&aFBN;AucP6_K8F<@jYmGQK6y~`jH0IXiX{O+V7{73 z4_y2zik4z44_mKTuD&46pTb`BzwOhbsP=U*0u0~#QNdswEQWQ*6z@7EkjlH@#;f*JmQqjnyWGG$6PCkrRt@L;w*8`M+%cJnW@vsSxr_1)K9>*;S{B_oi z!w6y8iKTCUdVVYdUQTQ2SYtT5wDf$CY95%K?8$isIigk=qwBuP? zX}ZvDh-xc~t|a4XNsKdQ+tfQ`cAq(@b$jyb*Uqea(KRDid#)8+A^rSXdOw}x-=SGL zzu60~y|?d;jc;!L+UD1ne7*8p=ic}ODD8MA0aqV3JR=3)(eM^Uf~1eEX7CE0fyzN> zzicUQjo$_7B2TTV*7bqCVPqRb%ZcNcvjq|Kuq8s>l}352$a6}|j4NqNZiIUREgqfPAI!Fi@f z_X0P3L`b&ATT%-@T@U2Mj)xTjVNA0cvTe!A26?|nMF=<;u+hKw~bN1Qb{56*hWi(`R{2s^rORkz0`)TkfW(u~#aOzwy;~ z#%Jnno$p~@#<QF`@c*hjlMHHg1FP>tfMBP|Osdb-3zX zrAU*Y1<|T9)VeUe?Bi+d9cHU{i=Fy!ZJ9`pcEgq-?eYvaY&67wBkPM*4TiOW7!xt_ z=^3M{;~ZBt&9hAivI{1-4RRu7=Xd6qBI)myYpO{ew_a{1F6}sG7BDTN7-&2$rJ^w~ z(&o{?&!Ss2DCTx>?jDjaZjwLp#-AmTsDZQts3X!q_R(-9upYQ+3LvCX1r^jvi$qJ(Zdcde>}f6?%S-Mku`;4{o?dEOy6kXV1Z zv-+bmmSn$9>xx*R?@i&AzY#M(QE=Jt?_Kv~Wxb3=8u5BrhuhYutgvxTyK$Lo>oa5j zJKPqV|9P)a{GxARuEuztaf&NWN$m9bG7QDLP&5+v)yNUhJL7~ZkKeCMVOBGE&NhnE`N_&! zI@L2P47DN~r>y{;ry|~pxf$vMKunS5UmwOEQ?3$>X$`b1lAeOvNY=={-(1j@oINYK(NHP2S5dnKGsi@|pj)#{4NibM(>QZnq`~g1Vnlh7CA>PtQCH(O<32Bu^hy$y zm%}guR_vWdqnflmZTQo(HU^YDJ(wM@a6*am?mB<@lcHb0=<&B=Z_EG}1}=+O@A-gq z1J~9z{vuIqz&f%l;<6O46wK`lbu3KHk;MtL=C3;Rj2mdb`<4DA%W`L@-7H%vZ62bP zX;CFBOMSE|T^L(JN(hpN_dvC3qHC*|#e3Dh?aFKNqpq+s8V{gw0?3dD;ZSH82j>Kk zi7`JPc>mWXf8@h8akI&_No3i?;SEZC6myRv-Bk2We@M366kVKa>Dx zKjvKtO%!w~+vp^}Y#N@-RF}}@A#41W@|FbaBs--C1Mkw8q%d}qye0@w%Y=s0TnXg% zlo@aj%MS0uPpxW;Ykqi_5}vILx+1h=ow1-=tySSKX^>Zh zB*}Xtv1VYC>}DuFgMarS{x3&0!$ULez@MCE*m=F1TcwXuEpr1-Yn!a}ez(esykSpM zBHpV$J*7m+w$V2|(KL;R0W!BGHiFgsrExgp1lEaP@`{)vGnpgtUGZ;8BDamU^Ll9n zBtEAibzt&Q})~d4T6TFL2 zxFuf|byIL4AYVKV__7gD7Ix0@i2C!5_7j%wE|>Bp=QW7LO5sA2s*6S^XfOX$adE&V z9*Z0NayO*??BQ!+9YaS*Z>cZf$eG}V4M+YZ=40<%_2~mkcz`jBJ1HMZm_YiEcQ4xN6X_gZfn%U?|pllB|+9pVt}mheD!|~ zNrkT{Yo=*c&0h5~X!BVt+QyHM0_PfCd95ne=c2TY&J`_{YdsF}^>iFH9(%CS;py?T zJM7@qjUmdhcri}AL|O4-bi6!B1mNr&ElU!Aqa@a6J2T+2eRl1vll&&oJI@4(W^Z7tGj$_f((9)p<*{vz190LsR`6ea($7AJX5{O(@@5{gO|X=<;Iu+ucJTX zU-fQ~YgNVp4$Nk8Da@qLLqGyG*7F7V;&QhX*)mu-2VFjcGOU>3lDYT{x=gqh8ei0^ zaz)mcsbSOf)Zu9G#Chd<>bE}s!r=|S+Uak?)9(rnuOaE&@Z`iBr)o1il~4>YgYE)( zb)R}+#)5}}<&!oC)`4_f+@uXNd-(nH%O&-~15wNTS~N&|@L@my@WP$`AGByrseAY( zJ}G_={;?|{EwCD-TTq-mQ4kL@?YHHre#p zyy|O0%_HV`Jkh>1s3|i;O+LjyAKDHo`uhBf(h6~>xDcv)>lOE*n&Awu8_MkJh1aAd zbc5>rRGoCU6rZXh8T9q}O>&T_J8}>V%E-UcSANVw%3O1BMe0 zY<~DJKW$~kBKPLR%a#?nH)HVl^pwI^*qRv^#Crnw1v_ZYpFIh-AmZ60IOguroF#pJ zCKHl3_~TNdbK)&`g_%((q!`%nDPYWx9?h1hHnCN8(rVe$M<%{jV1+oPMws;%% zh61Mz^aQ{DKLlEqH8^eH-%7nja@dmjb*ehLA|&5yk?W4IgZ?)KM?JInx%}OV9RUL_ z=!0OF*?`Nzz}P9}ZtEu((P;rFe&TrRF-nx!8x&8C#|b4=-j3DpeA8rDCieXD7CADB zTs8atHd9O^MNUDbl(H-Wo%??G7+^NTFJ-*kuzffkqL-ca{Ag~uTfcHIzjjt4q-Od; zuY_jP$V%Y<(Wo50F}Ly(o^Ix&=TNg0Rv_Aa@4PA_kHA#1K?M`RhB5q^cZx> zP`4{{CCSQEQ5^iXQ{>9DJyAuHZ2Fe)hJeM>Y57@tyS3FBSt_gS=IyljNuiJy{gR z&iK|P$RuFE|Dij|P;xO|c9;R%abbsF#$kvb9sLK!vVq`fRt+cK;aF)Cz$_3_BeV&( zsRtvDs;lT$uRFrip39*u8~?-7nrjf>{aBUab(~btOWm&vu_n1Dut7fHVyvZUQB|u4 zTq@+qaee@_)dyTq$E}mb3Y$`&Ew1_~dxFSzggm|AJ#~cx0-pD%^%b)1Ka|y9ESnM+ z222ckSWgU&Scl)v5#6zEmz;}CM&J8J{z zX77;g@+^|*MEj`*VVU@%GQq3Fs|(&AP_}4vyg1Lyuxj2G*Danyqiy1!BwYbrKI?+d z1SW#Fj_TWEvqW5w^pvf6_JuLw>S+DJKYp<^HjoL*Ksgx`kqUcpqzDIG&Mnw5AMXvg#LK&-P)-vtwy(b}GiEy$%+XqP z^N+W@=)!`nNL0bk3NujEw#rVD-75P> zYRx8Ulz>2Vn2lra6YGE&fpv1?yBB6p_=71Rr_5A}$fqxuT+eE=zJ@`Hc}S7Zsc2(6 zWxq!!jR3*2Ca7;h4zc58+w@Fz1q288%CmVHGqkFTkP2}%ZymMe^>a**5~L_$)rk|W zpBL+c8zn5l#Im|{{xKPwtgIu&!Q03>h$*r<0XFUDzESRW^9`*kFS4IsA=ZblW>ACu zQ}Id&5aRDkqnlKT0vvsR6a$RPDEI*amiVW9EWRD8mn{}$(;LEyC4lRgh;~)I zEB+<9k_s`F@(sFRe+3>H|K0@NYDO1^9k=+mX4CP?REYmH$gw&11#EBV7>xVvS4{|9 zxL?>ws$Vb!LWSqB8t?{+IZ2V@RCJYW)toE>#W5Q? zk@%Q#Q;+|iCWUUTviK=lRnwGe!Nd8nuP*h^SfEuEx^=4AYrMTsXTx=yHU>*)wieoO z!pRs~w{%}Wx?qxJ7Ydhfu=57GNtzkj9_tliUcyjy^_Wi8C~xU_$H;QZY($IYYYso3 zl-Rv+ipkFC7ld9SsZN~J28yX+Hf1NpfQD%%72P05dxGMmIM;1J1~3?i3kvd3VIx6| zTY{usk?X(4Z#ABy?lKBHc~LWA-MV5YIQpwy?uXQ>>dt82k? zD4dz=#%g(c0)ekKS(y;h4n7AqCLRgOqhq0ZvWw1@R7DlRI`612>lmlRVGb*&DRJHy zeBk@d4{fCX*f9OUj7EC-jBHvTp08+CV}J{MNZ*yb4bw}2m9j(gRsxXeU*p|zyW_TD z`a>v3gAOvoSxj)QG#9?VJ4N`Sqo6Wox*q5F{yJ~$a8m%XfDE;%&0UiI<@R zW)?4xVt}JLi;Bjj2F2-)1YHS9R(|?gf&`t_v*K;;2f(%^@Q`>Fk^*JwZn7NE3`OC% zkI$VEC(_f{)W2Ai=hH~LZf9W_JIK8o2xkzy*kjmEaf zL;QMqhsH>ZbYLNvl}O`p$-?_{x@RC*c1PU{zsIjrv}le58Sm|1c!Jae1`h;m3P@IF zO)n147oX>Cd=>e8la=M+UC?`NJh$trYES5l9w<-0$q71C`Oj2uSTeJ6*-AUF+Xfq| z;hU_D6q7`ewXn&8XpNr6;9LV#20wH+a6n+I(H++^)ow$vm0fZ2^bDC_;esRcHywL6 zdzEzRmrX_`cK!E%Oe#kT+&Ql&MlP7m|0#<3m?9rh(YN^R3SDr=-21#9{_-hh5f?(x zn3M#T2A&X>xDN)NCIc>ac zXV1LlSplUsJA-o;99HW_*_*%xH_uw6o^fwf0JCm?Qh&3lbcP90_kSv1Mz)WHC?~ck z;O`6z+Z0nw0Y&nt==K>&a~p)n_JCzp?U2!I=hX!DDtZ)JkA4sKFzHd$E7k|F?Mkf* zyB_Nm*>r*=6#_db>l-h>LAI+q_?M&@{Lz80Z5#(jw7Vg)7idO>jAy)L9W-bDeC6tk zP7r;c6!1S5H~5|>Sn$xI$p&AiCa@d);BGoEtkBJ{sS4OPe|+?vjnlvp3D1~2``)sS zgc0I+RN;3H{>)Mr%W2zEtSDV2NQ}E^EVIKs^*fj@tO+a^bkPpZaX&qCHlyO{R~>qV z6DlU`{xEg!EYm{s`jwnjB-e@eLv?10$^nXjIP`8Ry8LSiulFuA{tJC8e!2eHf$a)( zQ<^lJsp@~yy}segtzU1}lm%2y$CP#jL=yWHZ9;67!L$AI9*`Te%RRSEzaVbX9EPeq zXsD6byI&9=b#KyK`AV}!t2)k{vu87`&ER>4%kkuv{qV#IJk*=vpJXUZ;7ObMo8OV- zm!`kE#0+Yg6tkUzgk$u_K>v*rCTE#czq@qG{QPiaWWb{;AT_%tF9B(j7VvUog1Z#= zBlE>|idy>5Et;Xdw%wf+^O+w#s?5i+_gUDx*@u)SoNQbhbe3#*X>hXJ3@2F>lTHyG zlswS&iou9&{8K>adMhwvMyH~kX_fW6qb9*N_a5b9*F|1Pc(=x<1Q3-K);zD=twel8 zhEh7U5=S#uk3opQ-q0`~ucZ0P_E$_m@qf4cEwYIlP@H%}RB8s49TWpGm+e6FvG5wn zq7w=Gw`(L(uq&MX>otN@GRexE@2=tF)e0zX!hX#5Z$e77hmH}!4~*?Lsj$QfSg9oa z9(v(fuU+8|u`Z*xDUMj5p1|dUKmTC8zyy_#Ik>ikQe)7SZA90y}^EVxPX5CS#|Lp-nn}Biu zpZt!K#oTO&6GKL42AOpfvxXw8sOYo+AevSq_cNA5-sdIx#mi4ZnrK{sVw`T<;NnbBwF$qE5S<^wqB%(~h5`_;{?Wuc^%3c~`g1|f*~1jN3&If%VK zTMvu~jpAbpcKp9@Hh#PMiL-zL^20rKI+ab^A#Mx2431Gd?wD zOUZ7Efkvc!Dtg^lE&*p)M&#jX%hl(C5(RA}W@-yGDYk(`$!_URIzB3ecZa?x+b1>Z zrd^iisEc{~>2k^PNycY3`0J#LA$Yqoa<}w8X@^?Ew{T)8)>z&fz1)=!g36)%pg|58=7ofQZ`CfGubmyyR}|;BeXcMhu;d! zG+Nb(U|`CfdXm3QeOQp~Xy0S>Q$M;Xwe>AdhL6(Kwm$e`(aqu3whgj%_b`J8T&X=u zyITdDrpgBV*ffs4AL`Hi;{;101*bI~5b7G1<4vZRbre}cMH@KTzYlGRD44^3uD~*T zEZ40Ff$HzIQ_39^r?55?Hq7+nHNn~#j=0XrQcU>tuQk87EUe~Id*sBPh!uIrLUEk` zwt3Jq08FO&beA|rbQi(-;&;?QGmBeGx*&L z(DK}@Sxx>o77LqZ*>RI;b!1Kkgref_UI;R6Ws19s7L#lz&hs8ITUho|3^=JpRP>PW zD;6|E3ow*4=TSOA2P!EFkmfB%-58iBV6p!gm7U9`#(QSd*XKa9^Dza;ghIVyn@2_9 z4!VGUO@5xYZ91sSK#6|3M+Rwy9@GbXwsUM1(w0+Y6LKEggXbMsjz0QxPRN-UqR#)} zYbNBhe)D=3+0AVe?7Z810`!{09GJrt16=GCR5VuBH+Z%QyX2XWUD+6gEz1vsy5(D^ z^(zng6uNcGQ)Jkk+!@*+ucaHIpmEbYK7Bt9O!Lfg(F_ zSqO*%YgMhXPDP@tBQf*zv~fVu(=R*l%BD~MaN$J{yY7@8fAz||v*Ipbxxu)}Mz56c zyvRh}63tfSKG*D-?aCf#8bltQhanF@^S%}65Bfb;daTvpo!tKmyZ+|Q=s*4Z-QWKv zT1qo6grY{sH%)A;?oGFp*mT(0d#wpor{NuSTF1Nq`V^7N}c`l2K6YqAc)YR_`>!f#v^^p@E zdDKU|dPQk4yFR3B;peh$_wijS=<92Rv8Y; zR?rCsN!^pK*}lk$F*aK>8o&$(&ohok31Q=F)&Vn;W5WK|Km2Vh9p+A)<+0LXK0rFd z^`N+{Ro|sQ@$MI$7h#1wW)Ro-WkE+prf6)q;5vC6aKv@!F+J~>{Kg_no@1wVS*>sy zcZK%L&ht6F+ZRh-@Y;n)3wvzON|HE!g;j-i7rGTV8JrX*INd;&r_C7l(s@ z!@QfBiZM;94KH)DAJh-CZ-3RYeZpx`QfjscJ#L8DFO3W8ckh*5lV6_&!rPN}k#;EU z?~Z5(3jB2F2ygajhZgaE-zo^6m%Hbyn>6L_2|SF$CM)wIJ7!gbYMDdfGFqr0ZfuzG zQ9r^BD!0x*(k|du#I2s(UwzsFWaq_H#l}Ujs|C@43G|(r|-|%98@o?A=S`zUI>K=I?*T=J=%dz@(+w?Ze* z_!#TN+xY%o>A7@E{bi?JkE|#v7klrW-><}Vr+MC5!71RGP6_yc-Tb3n3|`s#vlh+A zswKfy^yZ*mpRK_c{o8~OzI;jgx2wvM@NIl7cTDxef3>P@>hrVKPp`fbq#IujEP!yt3iV&7{k~E7~es=}{B-&=cRPpz+xX8foxbl{M4w zjRLPss8{ME$QYVU4~7E6GJkm(UO5a~t86-ZW`P%uWrSLfVy)-^5RY;Mo~?ixX$2b1 zwUNLv(lt(2XTpxgx*3+7epYDmjEVR=u8XF`y~=Ky(k85&nMoh=IRxAfY&w)6rpQ?A z-8RwyIVpHjtkCPR&Eplp&~qZog0&5R46m4X9u=0qCDi5qEttX|_bal~02 zhWHQv`%TNV^V2=ZPQ0!Q)Kb?NNedbVH#A)r37rFV?FB?)xIcL zh;+2$w{md=vQ^;O`-XKOHwE2TXeqM5Whd>#HBDBuBD8+RGm{|3zZwkKfq*6T(1G&J98vVt z(q(V-42}2A(pPCOljGAm`^r@@rBR2W;Zy zN~#SK%qVIH>!vXmnXJ_D(&#gB!Iy63cZD#QoE82plq$jlB9mP4$1$7E< z2N)=~(CN^Hji1wMoGP1biM5$^&~P(AUmku;0IE#bnt{Rv z?Mf_FyBgX8Eg4$%TCd%{Kzjkrp1~bZBI!_1!^X*fYEx>w;IO%$VBN78VUIH5gIaBZ z<*MSW;X7b!9qw}`QOsJ3Bv8>tby?&UM}h34u6h{?9CM7B+iOB>g@wI-pRMC?g2lu? zdHr~+i)nk3_xal&k{BnBaA?h7kw`JCDYAl!#)$tN*8xSp_$2=j5Dj4K_qe+aI6byN z!q#cD=X0DOF=1_1tiPrDpB0fkJ&m&TkJX=zo%}t}DwpTOCQ0_sOQX|*S)+l^6ZB4N zmF=Cnf(IlKXy;pHy2$lWdK&UKFlWiix3qpuA$fu{i7xCdwt#;@jF~fxP38%Dlxz+# zSd~vJpN6-Mt^6gCca&o(@UUgFtk`s0E{}TtxJ{PRreIbGyIj6wGAwT#>iG#dHi=v_ z`~IGzm@^bnk%>=j6rek{g?Ul$gyEb{({?4NgvJoMD6u!*K3 zWP?XNT^h(-ldl0J7NbjG!TEs4ygF-nz4>d5su$x3`q3ySwF${)8YAJDz-kHBsC z^tDqyAB!RK8P-K_@MvL51F`QHXiPAjdwmYlN#IC(<;M46*ixS)KU++TgTXa20BwNW zO9HNqi~jh7FP0j{5Lxo9RPgm53$E;AJw_ho-wJkp_tz%J=EF5{v&l6l-Z3qkID8}BM=|#((ru)7 z(}SZ_7J=5q)QH+DyC6>T8+5^#3s$HlE31UHU_kWrRqx&i9f*r|1@2LRYp&zP1`Ya` zL=Fb-3O`R^jRBeMc5u;oweU`=syLWMHQO#Z?f{aN%QSF7PZzoY#YR-S>W*u&Qm+74 z41K$y4Fi^S8O3fvdjqx%6N*GM*yeU&I#oLkgI-Y?Nlg9R?4OWXs^ z&^?O6z-1AseksCs?PqtNAr0AlbT-K^K%>M+cIT6X=(W4SnCR9LR38!HfU@RCf z4>>@$GHB&8X(M~wfq=d9R|(ck0fhGR2P-Nyx;S@&-*OjFP#en z-577gzWsA#tFVyCrp)MWN=UYM&dhH-z=jS+i=OZ{@10R=CRRvCPCHWuT8U)d8`Q2_Ze+cSUg5 zQFZFrQt$1``w>QEOK2&S+~~|V;)%WF2vh~*Y8D}<*Tl1ZZdZI(k7*Mnx;IO781&pQ zHvTnxJ^B&HoDH_Pc5`#9(!S{Zvn`_3u~=0Hx|?Vua{LFXIpI2gFsw}Ey52_@GH#^b{o{O3J)xaYVH&LEi!>eH3# zhZ8JDNE07hMK+EmyRE+7trVM2kz7iu++T^!X_SZ3EOk5z1C+QT7Uo}#+sQz*{{`P_O@E8C>*hS*NZ z80VmKaIo;g?BI`Z%28gJ9d+RJAH9atitoa%HAgG{{=fwAV9(Afg94jF?pwp>)+^HK zqv~Wo^nZ07K1Cks^x89z`wo}@A_t>Zc*$Ifc)rs<%qOe%uM$K_enb&<# z-abg$l=)I*+^!Ma4DEHR6Ws+12tpFT=+Wzj$M5`#XFk>)dN~vv zT(Cj9_BV4Vy;|?Lbo?prLo+HpPR{I#z^zWu4#}6MOLe8n7SS!x#>$rhUwq^#%y!Tp z&Z}2o4#J!Q(W~K4NR=NF|8>3?qt^8bOpM@f+UbjV+b6;VtZqfUla;D=P#?eL2_>+= z*a>ATDhw{8UyUt$=%?TE{BL`qH>k~W;z7Sp3;rv--x+qFy_pp8ix9i64@Zt$&3n}p zTS<{!l=gwLQ?*2LPO#Ct4btr!0@LYDbOpVTBnWEB7FZN@Ouhijng{4yW{>DDT?Dll zDBsy3Y*SuV#fTe1ZoaBlukyYbYQ7HdNe~=V7RWC7mrceiH)$@);@M;BdQg8Yf>j&d z;gB_fGa%#$v-Wdk@E;Wx2p#_BI~T~RFIZNq$OHt|Hc(n4y*k7Qq5)X*IB(uO zjN+inT(8K|V7?R6?2YPj1)OSCRZgg7uFil>q?_0cL?B2nGH;$k#(j`68a5OLd8Q){ zaeG5kQd^h5W&z2Z%Xx7mpP!v^;kDL5EBNf9SZL!fqqI2vGy`3DwoJEk=5eTSglLPm z0m#DhrAx!iY&AI$fx)zS6ZG*T?G+{~vo&>eI^^ol3CjF+2&_Wi6e!i`bHk{iUTr>t zxpln;^8Qm(#~yc_)ZegSWiwGZ0#f)4VfR;Ve9t2$JR z&eDviOU!Su9W|6Meh#+{8DvE@YEJmqFGwwRB~!HKGD&!8>`I}PUC~o)8b#Jo+8g5@ z%L>Cj4#xSX6`7IH&=dtO5O0{==a#K07A&6u=*pdH2#NDuAnx%Tp(qYNIGGK_i@`Ab zD@qg3zvX2C$j{D`_es)A1CZ@j0Li3S=&wlwfXun!vzTBS+dP-(a)1V{jgAplg3_;U zQBM zYnYR^jQD=Zu5jq48 z&>vPZuwF*d`5192f#zdeSU1aK#Ciz0RZYqX&x7U!InpWWI+YJXQw3Niktjq4lSCl~ zgt4s{0(^LdL*_u|VCPv&#P`UwO=DO3S+G;s_rskegCBNWI8|h@!cGyzZly>*rOgT6 z9K2b!+aJokp=%R+>e`f#$W7sX@2k;Ie9s9=W>yLfl0v4*Hz)WUDVbU1(;`~zTTNe* zZE$b!E+4;o+@}*)dG#uaIok4)hWT601IReyt%4^WbuQH z3r}{nR*)&9SdentMrpAHN@r$5YY`!pv#HswLs1)84^y4qFCIcKl@mHAUk@vd*x{cG zZA&`TJTbSyWz#rUBK-*_w^6dXVM2?S4m(j=Oig?a4iGy6HvRLKJ#eu1nz$K$l`k&8 zb76bJQGT~A!XU)|HAU|QHQ8+u1tCq*rIPIe{8=w~KSQ=SYK`XyYXznB1z9!ueS&k0 zh3zN|1qHUxaK=7&aK-InpZ*n#SK(qDY6v#vGT=@ z>U$nK>=3)>G1s|KIj%9Y2M=8LI`cR;9?r9w@wWLvKJ`;BY)Lr!OL`{O%9aW*$nHTP zrZM{LyBB3AqhrKnlb1vn%c{c{kxjH=WQ<4ufZGEaXG|6Szn|;p;g9azvy~%v*=8c} zItPSd(=*v&H8lz8wif1~V8g)!m4avXa$Qw z;g(K6@tudUIy_*KSOV&Vqk_w!W@eF{zWCVG=p$_2ti8f{D?n-hH3mS-9uB9BnTJn2 zA5zC$Q?#nEEO+SDa6>ROPf=pW*3-e}x5dlQjQYlJ$9Da5k)E}P8fMmsnxfPFvc;u= z9psogL5fT?P0`3qb4Prd&S9>T1MZI?QBXxLhMb)Ie*L%(ve_#^_Hq;h+hJPRXhO42z@(o0)*G`1E*P3PX7kLrA(b-&DecXPk=$ z>p|j$4eMK-y5~9N|90BikWwp~pG~pQijxk>dx%FD_@+-@IUYK^lnVoN8IV@lE3BD_ z6s)=QUH8TC){`RY=o_!M%5)Xsn=~u^KJ%Ry$6)8WDctSM7tZSi+nR7EzG8kR-DM*T zH$VKZtD`JV%c7;h`$(S~ftKb0zQIb0jiX2`r9D7*k-Ids+w`-qj_jBA`B>V6h|ez` z!Ht&$5u?Wc(X%HNXSVm_#U9xy%edTwAgYJBig+TgdjPCRjqR`pQLn;x)rFllH*g*8bVKKdg_0 z8}>h_`RdPYl*cR*v)HSLNkW#J!<>3a&T@%|zg0zD+m(Srk8<{vT*Su})VldCS((OzYX;eQX zi^6(5jdCO1t32h6b9U9F2cbIP0D;S~iUE1Chp3q0M63;fB!AfbtT=n#aB8$&*lKaq zXm_d}lFG>Kf<=%Esuu0>PW3x3K0xPB1o|;$v2urhs^1pD$C9S#R%m~1RP6AtjKuxn zSsMGsWk?vY@6p4$%KuxRc)n?wzaBH0?-9dja?WZ>I8L!gC~}a}K6;~CeneO;D)Ct; zsMhTBjuT=VZl3=I|D?#|nTC)h!V{WKm<;m#a~OkQpZro-Q*;JIFQ8o){+BnaH5`K; zX;9UJ9Ndl3@qPH+CAr9M+Mq4G8zlFj#oiWnJXWzL_<)Nh6od3~;1@N}-S0+@M-W zVe?=~pfNn2H9$u));eLYV8>9=H$-fJxFq@2Z~s^Ke?R%nPyg^gvLzI|kRmaILUCxn zVQ6>p!tAJnzZ58JYcyRJ8XQ%ZsJYeS73Y0eX`t5!R=zs_)fUlJW&w$bG8?68dNWfP z@d=YdXN=LI=0LMNk6kmd&kZe^u3ogs4@o*Opo?)_Y#BKtNS7k79kPZYynwDgMfGuT zH7L7$>Z7X>W-`sH+^JpDGAAVnY9{WVcp54u@j1FyL7f6CJN23>k}gdU80n;lJbxgN z&Ye2&*#=3Iz+6_||9g-}0m*oEZJ`P9_jN=zsymqn9tKEy=0$(%QzJ-_9+j9$_DqXb z=UdD{f&7AlJXxY&(XXg|dz@vaGtLUVNYcJwE42!%X}N%6w@@S-(wKBcSfa3D!Zq&- zDv{KAW~kN&T@-;ct=a-qpScW*Up)UBEV8`^M6E-dQ0*`s!$goBzhP*XaNDog>GN*} zSYH-7!*?3%Cq@!0o%-%WBfnGD{zRkkd8vn=0#ACvEYFmk2e zP7f^fG)VHLIbp3n#g2s7FfnCgU519A!S4F1pt97K`h|lt1zQOh3C??qUvUOiV6*OM`WG==?Qe5G&Cndkz*DV{@2kN?j-^< zUA7J%)yZT|Lg)N5-#)hm;-%u7$jgn7Sn1c{GwcuDvG4xrPz?4y`5fc^o(J)vT{oc< zc(|y=$6|aMf4XxnS=rMhg52-8Nni)4di+zsEbm*z4OWUFLSC=5Ys!H~IR$hP)ex7sf@FD-4 zhY#b_m0x~TC$M1UcZa$Tl7)l0H7<;ZOe;jJrr1P^#8cY8t@g&uNetDgy2`Yh1<|)N z)if%r4BLM3eBXADJ^ju5E$H4M~}r<)|~# z=Z5TVcsS3yTV4=Y2&)eh85~|>ZVSUMo!s$@6iAa{po;~~z{HJ|Z79s(@C38nL-BM{ z;oKVT>mPVO_fQNn^SEwp+kZ;-YJP07CSvjGcgg85m^A?*rvdXz8^yL#J;rJy*SfJ^BNrps8GTc zm)hp0lva3Jn-UefcLsN<=8JonWgfb8e>|yIV*vnuLAe@T3yrA>T;}(9*3pN@t_pxA zIX$q;G_xtHW9n+tM`LQ0YYWAjyXBwy<%>_y5ERG9%Vz7nZAGTYX_;gc41U0X{yJRh@N!lsozx7xr2 zL1$=>=Sq(vP<|U}zsGKBA?E<)IZs`kzW(t<%RKdjJrYawE?hml$7)X6MzK&pzJ=13 z_;f>tE1iBsgKUJL722uWC09aGTCh!YQgH_&00#Pggdqg#%;Z-+QdBjBmMKmbH;HPr+AqW>igR7Jg_AWaoOpHg9;=j zxWuO^dO!W>4KpMvh3hp7gHQUd58CQoEbVb*aLt}^u^$-rJk4>}@B@bU`@-)`wOEqB z{Izut*)o{N>%z0hek<$pF~#nnNC~A)^Xq{4V2TQ3CI?k15K&wMwql3!F1;@Z+o3lz zH-pP2;^M9mL`PLkzV|^svRZDOV5Z~M6$EaBJpJA2y85wuW&L7)Tf7=*%q^MXkI!*O zhccMQk{JpboV{)WNs;fB+_k`*bQ7U)#G^^FFy;-v}G|u%=ptQ(znGps{*?Pq#pL%bsf=CC}%gd@B&vUT$%?ZQA zcb^;PDfUl9)%6z90&%CZDM*K_9bD>OCOKi7y+DQ*6|}ZTwTEAo#Yd-%SxlA_`MoIy62<-slM8-M-BmA&Sg0m01m{MuM7U1gPG_jwC0VoX%R3|PgjY`5;(nSgj5s^z-t70y zn|gJoX0H(Iv8w5AIn)_LXgMJS7w#!4ykX#3y}FiE2RR;qc0+cPM1+fr5=8;M{d2KzPkRRK&E(AF~$Vcs}YnOV+wDa3JI|064i63lY3bO4|xW zS4SpP2jok8yjs-;NuB#urVWyKcOc<@;58j)f$H3mETYJx4P|FQtp*Z(Eh3~utOgMq zhiA`ihI3d*v3ZswcX=UY)JJiizxuufDN^6xyh7?+7%8`{kaCG)p*QX6qUyh> zZ+a(#g*Shh_t(rhI;084WH%P7BWCr_aj}O5F;25&_X$o^@`Bhau`Mg7*k%tng7ju7 zqGZ3%d{&Qdim3F^Yo7Sr6Mp7%h;+!3gdLs^HT9hTyw6YA|NcL}Wkfc3y?@tTOJp<6 z|86UK?y`$#4mEaU?~WbQO;D~rA8ZVE;O-c6k)){Zf|5;=I0?cpx941*tyk|;0&yzD zdaq9#5fs~x-KX|oB*4oHyRUrI5Wdb<1&7a?$aTRFT>}H%p3M}SK>;0xwneo2TNh;y zJUdjC9$WnPK)-FC|6L`fsdq^0geQnmeppor-f}l+DJ6vLfG(BwfhEBYJm-&qbn#Fy z1?X}*tTNob+5Iz%6SH{Zf=T2ezi^8S7x+H0LdhM9y+x7hlopFeJ5`{hlnyB)la^zW zm%;BSG(0TzDE8SJZpuJf{2{E7t5=*Og-o{`lMVUOm0sKYlBE?hd)+WmnXS1UfN2La z%X2%p&IMuXC*y;$wF2U`DXI=1P|Z-J)Au~%SzTJxK{a+sWRhLu=hiD4)vgUo~n~s=@%o3WAtj{v?snTUiFNV zCc)tYrhkYX`!m!1YwLY2ChENtubd>QFO7-%*vdrZQ!Fs5Wl>t}{WdB$3-7V3$J7dM z2B(Fsow7i(lmuR(LFfX>=RvLl@||m?CNVgPE@t`>HZAIyx{Ad6QQR8@xeDl5Y)h49tWkb}G<*{0ey4g`dI zz0b~qDlA10Q%E^&aG$65(?rr-1r>|NJiT`gZ-zVz^E4gxEzjvpV zx!6Ln*%aA?gwpMywLv#!F~J~`6y43N z!X8zsC_!+4Lg~bPK?xzL^25%hmEZ{fw|I*bKs6-$6$sUr)Z_aUH z5LH=$sF-45d0s$iYbO->bp#ugYu|V*+cf95P0^X2z;5++9Q!2nL1Uvc+IXtAJzUhmMN ztl^-^3R<}1KzRMBId;rHSL^-(Xq4PG_Ux=;x;AK;ck0-5=@QQzrgi#~=|$kFEK;mg zG=Ypkk7pNIq`+%B=+v?83h3H*FIM6;=;W3$%c8ETYK14HPlOfHyFOcJyt4eYIAL{A zSA=7yI0quffw+0)mAxscn;DC>aSL!;Ko*aN=H!7fjueVrMUjEr)FK1^}LlmWQQd9In**;~+>M*>%@Gzi26n`=Y0=cfG#r zb#qsEx)fGEXx>Z)Z?|S^_#XdMX!gc-MfKo{^lPWju^mru-yCbf8#zt-a_SW!RIzgK4lQL#*|NmhRON(e3$h$QH z??A8eV|U~9%AE;Z65@JMrANB7UX%;u+)&dF;bB=0jlS`1h54U> zoiQj{3%{`*)@OQkr3XGEP1P|umv&kCSze%f_ras{`lmnB{`x1sdjAhHnsppE?O zgXe62U;WR)xXBL=?3~#JiJtUeJ;-FA&{PUDJa2>=A(zw$z57Qc zx@&@bX{AS2^xQT1HgN3C#X_4aKKG}uopLa6GgC41X&K9q2j#^# z4hc>#e(bQX*xUB)`fpnR^6`)S-y=s{*vMSB8uDg}ZKB9YN_&KC4M$ZJb6K(um$bM- z#>UzdRoP@zJVECLiTzSk`zIy{Q3b`sBaZ$F^!f{1;F3(TN{EG@w?sFDx~oh~RD);i zIP=ea3%a_#=XZoG7!AEc1CtG_D0U@9 z;-C&i+yeCS3t#JnPWwh*-F|nYyfnNGRwj-x=Jx!{o?c_;yMKDa2yO9t_pWQXFHLsO zv1L2spb5iI7=6QsN*xwCw$KliTbM0Sp7sxbg`KEyWR}-1T=7DM>z2^ok2$d4Hle_0 zLG8j-5FDLz89w)Y&jr<}Hl%1>B)t^gq8jzr)skJR7ugU8HkyA{xQ@faXXHrz{YI z#cKuMrNFo|VZC_?!zmNFQ)We1zwq*ar%p4|Txh z+@P5Mzo~!vqPIOJge)Dsjn46_XZ7kBajEaF@ykYEQ=E`&@a}c%owD||Whw(Kg=54C zQBc_*nJK;EQ{m+(Gw|HOu@4T`#^AXRbkIp|I1CnD>+|^Vhhoc|^FisDKan_o(KQ#g zDEU^i(FTfLkAJN8G^y~65ogW1>bYlTp5UnDQeZt>r`QOq_Cs%^g&hli3~TsJ>_Jf{ z^N@@PkZ|1jg2_QxdEsNuUpw)G2^M&0;$y4G#+L>jyRG1{m0|%exs-NaP_?|8zNSDw zWwWeRa5m_)pfsZVHKTlkd%j;&bQkGWZi0&KHJ)pwd4ero`Ap)Rq{!IF2eNYcJzj7_O8c+6IlV}LqK_T)oiFeH*G5Hc(B}WD zdh`5O+ulwOa4;r@4TaCKF|;A)x{KdDbhn53^u=W?F1#Y2Y|7136NCPkJ~k_ovwN{-pf?z293|qs95Q-ojEE&PC8ucJP{@cQ&hJC zlg2KJx+62Gj+~cl3*QVBmzVs@CigltyzDqMIo8NQO(rY#|OFcUj$2}Y9ZfKA0_Iwag9#uIhMU~^%?FseXz&38a zLa)ZBIM^J0ZVrZxq|cw^*o}3an?CuX6RBSO>P1E^Iu;Ozk6fzCUMxAuXW>L*XYNdqZJ+#0lG@?us?UK+|G)Mso1z~m z0S`Wx!V1Y+DJVNeHnRulh6#IRO@a=e?IX8B=DSAb97>0uxcO!&z%!8-P7{>%9zH z8U>f7*x;Z{SU7t7%p}#SfP^(^bX%#d@uaz1CJ0LKULtj=M zmtBPx`1r6ih@-VBH_7*m?_@SiE_AM6!hi5TpFx}o1Aacpt{v~qwQc!%NoE<21|ckD zY>GynjXFiO5aTI}6h$7b>U3!W^oHrxshax82Kr`bf^-{w3d+w~Rf_~G{q!1?T64Oi zJ5*5F7yyUw`iMUGzbH#k)_>uO0`l=-d14oyR!&)kUJp_%1W#(QS>Ie~_z471Ez+%f zG#zBIu+;~fcw)o`$x1(rrKPAA3rl@6nX^G{kQz>a8b)2KpcZK5P^GUvq;f*5>Jg1e z<`z>zdG(kRuRw-x&EZRVrlta@BH{ll^9IquoyiILjn(IvsFLt3MnqS2zv7aJ#}PtKW~4MCMB z{^tZcL4?NX_T}^8}SU4L7w$HElWkR^^V%TLpT8`!I>G0hk2eBwD!M9U|?4|kQHf5qa ziYV?2Dufxqs7#AmGNLTt_Kvn`im)9LC+>4k9)pC@M*YW!?@siy_$GyYKio+&T(}tn z8oUPRK@?H!R*K|P+M0<^e6bm$S-!$c*F$H`+5uH&SjTysP+NKjt4pT zlG3S<_9`uqNFV>}-;tD;1`?oEG{EL$Q!IEQ>6G@Mtc%>GO}r*?9vNYlKCpAZ?vGSH zXpI(#Q$b`?mnvup-5Q=K+^*ajeh2=488HAG@)ea+(PJM@2DdN{hKAMSdcN^w@oWoD zzP0ZMMP!c)&o*bRM*b+p9-_#8N*gn$n{1zXQMLv2>k1;PmAM{Cktvhw)F+rr{<-uO zh`_J(`^*<3X}xX2<>(>oM|(+!D2_TGctS7|?z=MDn0g7AbWK#1{k!k$O$J zv|M^4v?Q>QSx=gxvD>&#aR6$nOO<<>>e<`q^uTU;)ufXO{Jna0GP7&OkU=yYfD8^B z>!Vhq&SQtyOhnz!+4y!OJ%XjvC@m&#^o2A|bVO6Bh?K9>2$!(@aQ4|}_ljTc6$O!sKp{o?S+A#8B zp!Y=-3F61XO>rK;1QoPR(ize2(e9yFUxj|sJlSEPEE`{aEGR6!U zb$oL|)|h(n9uJ_|Tski^(ff4^k}9VDX(=giVZ#J=Xuy=Tn__`yeFvq@6YPwj0y)1Gx!7M-7)xEoREr{_g1!p*9#euy#Xe;7ud2{pVj!F0eHzU#Jcr z4l~n2V#e32uTLveUY&7TlEWkht(D@j`NAFsk3JC|pyyN7q%k_p?}2i!Z!ai{>(yO! zhAfLNB*n@k;mPP;w*{1eu2;|3lmWxa!0R$(c+7m&KeL%R_xo9$w&>wwQd~ID&w;Bn zzYaPjcsVJSwvzRl#^^^sdi0~Eb1r^M|KqARV}6cw|H!y=NYoLzFR08tF0v{5$ak}5 z$3O^cI|GINVxuTdh|3C@SKaSNRh)((NU_v=J=eXtIt*9~vc zt3Q)>cy5*zF-H3RI-1{P;^5JE?wIizDqhZn>#j*s!-}NCshe_Plf+dwrC92lt0{#D zagAW1S9@r^qELWz2qKh9xi|NeUAXxokum(4P9 zbbp|9G#a@C0Xn!3B1~=V)u5OX=jdcvwN~mE_9U$02 zf?QB{$ipv{PSmxLrik=_EKRej25K4d#rYod9ETL%Sa_~YbKLyH&*I(wv-p^8QG?4O zC(mjcO{3U#6pTW@aK&l*-k7SOk7c(*u_Oi(PKJvC7Pkucd)o>S%)N-hX zX9G_tF1)KG+a%5Mm5N1*Dv}ds(%q~U9iXqySRS%XlBdptzEsQ!HOpHkKJnM9@p7ln zNA_T9P%Gn1K*6(@EqX7?b}7Rr_~yE6crNt@4D(=yC!Jp6hYI3mI@^3Hs=TZK)s`*+ zWEoX8k&r9|9&E`~)w!UxQ*>DxLrA@%$ou{TT=dMFH!lHp7kiui8r1c^tvD#FSLi{> zYL^C+m2in;RzE{V&Y*BIA9Mt$CdO_)W@}%b4f?w7Km>3#lv;pzTQUW*Ct^WQWlhv|%H zif)fQAYG)2RVB|@GJS`%DLMtR)ae0z{*DWfJb%1+Bj@>#bJ$tsVM+EsUp~26UI5~# zI0s>UE~uE{ns}t3@O13#1jW-EMrdp1kA7MaU@;m8-dt5dvRru8TWe(o$|x4*m2H&v z>T6lEGK5Xh>9UhEdps}7;@NciD9H6BK#|Et?@!pJzQv$9x7&NAUoVt3B?x-mP|Ks& z%^*3+V&UyNd3R`@U_ZeMyH>&WsPup~B~~-VN2WzN94U4J;W-|K^-0kDuZGn(w0hTQMfQgRZD7w@8^!+HfJTXTa)cqG6NEddiCjW z&_^@`Zer^dSi8HO$q*Q6KqQBfrEzILj{!Wa~!?@KY!Gg3oVvq>iX{ zpQ!7u2444B;$}Rh$f9Nt zqz~i$76xZ&T2+UYF+k68R+&k3F;U5yYd-N)bSN%_Y%aNwMuK+v$_Ytc2EP;)-n~a% zHqopfgk?L6r>qV!gyj3>(D5wl2w|`WzE13g%be6A!b{GP+^H8L6WtrtjUaS_e-}Qm zLj)DzxwJ7HYlDgf&8)6d<+5tM(XV>f+qR{-Y_OlB(G1%X5d)^waf%94gCjhnT5OwL zhSKqN@P$w0XDB_Fy!GUNE%s^KANns`jDScY!5{qQrbeMgKj1Jg;zl5#7O7+ zEmGiz4B^g30WMK?`YweJ8ngG+z<7dUKzKCIyIXEVR@hpBrkGVqFP5*+Yl%)m~+@{Oqi)j7ejtRdtnVgg7bm7s85HuRbg<3Et#q`aW~% zHC}ZzvZo|SkHQp+7xcNU52^;?jyvHy)LE3-AQR){FKPu*0@K-wVxRSIDM*QNs^Ep{;^|3sp z*!vW@hce*^kOhQl1(oTN=M;E>{vQS3?Vgh=KU4QZ?5?4AjN`2_hw1S>`qaiTN8XUZQ=8U zy8`Ayy|JoZ@x*@(sDHN8IF0Kn==qQZsTM92=SyJ%ry<2r8~6qMOrD5jyKEzfE;|`= z^qAy}8`UXeka6l7lp%LUB$0gau94%X8#ou)H68u~DvkKrN!j7IaUI-kL- zws;xo_Rx&5_^`{WR#mB>3la|K@?P>eEjS_0^(c)f9owo}>QTq^f*f>-;DVrz*%DA1 zVU+I<)N7KM9tL^wnxc=Y6NRlR{2jwyklF_$?u94UnSoFp$~G?;QTo!kzZq?TQMB;` z9hpBGI^qY0;*uyffg&p??G4Y9GvhpR=~Lr$84;@^``mW=I=ysh@E2g;s~`McXCCGC z>RoqzEH(Vd$2NrcY~I_2JspmCXR{nhKX6Haro$*to@of_RVGF52;32vH@SJlGL;qx z+J|B^6VAl!v(M&6@rrG~N&moNEHWM*7)@@!G$EllE9bbEVjoeYo6}UsOSBBt%UMnYB~4 zjxE&~(THLzOen7Z18S>5ef zm5@1GIkm&5*DY0E=3VK5#b2poj}xPErQcn;W@3VL0UOU|grS3;ORs?7J^ts$U~7ZR`pB7Q=`OQ$>S*F7?ajH@Ip3x}t3zw<$o)2{0e6#wGpzrSGt$a~it(#S4;)0PWYXf;{k{4m9W3hZ7= zyG5QU-A=dC{XbJw%Yw3I6-#>{s?$oB0%`bZP{&FMs}q&d9l}CbOr`lP@M~43l5#;c z$qCO3-xDw&xS=a%mie##1~QG9YrNKs+u?n7)?-=foc4gG=-4?IWiZ!~k92$gUsE)0 zbx9t}F8TCdzhh3q&{_BFwL@IXs23d{>c=xL=tJ#4SGixqgiF3DsyJb?pWbh;Fl*L& zP2Q}%!Vi*!pE9@Q8IyhwsjOK$mAO;(el@~|38|uCudMAFnjzrBuICQ-qO!}YZ;Z1T z8{@3dizMx(F*X%e#-@N`w@@UT(%yU{-~Z4XdUZBQ_c$rqCvQ~bzKTDaqO-{@&mPZo zkEGWd6_)}_f>FS6lMJs!fk#Z5_PXtpe>yQwmLP0~^r!QXGX%5@G$6wT1?T;zvmV}< zVnK_WR8vl}UmCQ0Vuh9+6bmW@+bL}qK~3o+Q1zD%q5|pkLz3xpcRI)lsPiYT5F~^Y zg~bXB=ypj)*i}_GFz2<1D(ID}cG%9-bdd_U9=lmuMA^Qr)2rzn9w{Ec2kUcBczSY7 zaEa8>2pT4Co??g-0Yf{n;LN*Yhr`k7y1^&VtREPC{G6k+6cVkM-s+4<))aZ49NVjG z1oB!gH^vV&BEupL&DnAtM^0qnT;zE3+}gTpT4wIXuekL;I|KDhSJ zRY!c6TNwTO_cBX9r(OX9;YnwKCo?U*fw^Q{KL(q5!Lok~nw4ns88 zgEq9jU+^u9XY%GJoj)Rn`FSR;3x0CVYKA#Wv8O50NNG1SX<|qXs2e0uDv+(Y3@(Sc ztl_Tv7AT0Vn0b}ygn~RQlGb%9AA};Yd=@n1UY`cRBeazH;y$+w;io=$;(o-Z#8mKC z7}3LEH(e)_IjL&KhOj<2G_)Y!2Bkm-(OrJ83%Ul>2Iaat>N3kgsSw(5FJaENZ<==>VGxn?Kq+%yFibO+v9|@-} z{O9JcRextB(P2L{;6ryJ6l@>O3mT(-dNJ__3JWw2fAgISWYuU=WHkbtC>AOgHc;9F zGoB%gXAvbv45I)Ji5( zQyH!^D{ejb1}bua*PEic428VtWL8elp+8fvXqO~G)a4w+Y#)51&#k|!!K9zFFt~$8 zZ?|U*1Yj_HpgTm0ed-kj@&frM;!nf~SIliUs)qvEAsWl4X49Un=OB#hRNvdbYOqZK zJzIs~x>k~K@H*6sHu>R#rdfV!T&YTj`Zs8Qpk^eZPPssQSXGHd)acySDc6r!ZNd;x zz%_6VJGT;1@-m|iwS(c#2hh)-FNT2H=iI^%&_4(h{a$8qUz8txrID=sf;At3 zxZ?mPC7oh7P-H!&t(|;clA^jhHkYoQ0RG20$vMeQZ&Nkg=Mp40z4OKW|J(4DGjDag z<6wjgHb_H5&0tS*>@jZC{N)b^mI^Eg`Q4$egJj`LAx(x!wNY1ELd zp}QrzW>vnw0e>V2HhbN3v=MpXAli+W7e2|c7yR%`C-QAwRhOM_INVa~q1JVK)`yr| z-Af~O`Wod>x4e8zx}|Xz$?OdDGBLWQ&a5ICI5u{V_|r*c;O$P^=C%-nI7*`qe)h&m zTdU!+_~5VuG2*IM_XM@mdxDH6^{U&UMtQSbFW3<{0uzPJINFH{o4cI9%MTUrRhNDy zvdkrOPOtrhEFVpv)@#57o=UN6DUwWSw?lg#EL>h~jNVL)qKcU(qxX}0u#agk2p+h% zPHz(>`|a0^;855(2A1#A`R~5}H}Dz**ImNIQk%ZzWdX;}&Xe~^(qOKr3r9b;TOm1< zVu8OS4GYr@5{#MYP?q!L%x&Sh^d1kR2-K2V=r|8E4XzITRitr%6fpKOBiUH@1ZQKL zAE2MkH_lGxa0B|Y|1J%Ruwdl-*5uee^J4 zHe?hH0My`{{RVDmKeM`d%vNYGH>2?L=j$fqEtsc`3bQxDwwLq|j z3E^Q5D$iHW0|PvtJiN=fF)*0Zdi>gNzCO`nU7oN4ZHbLWF6WX z4#9W?Qj^*&Pvi_J@c_)=wgHb04;zEt%?~hp{zz2sSO9ZK{H%sUhLC%3t#BNA34Mw@vI7 zWjdW2wNmAb58Gx;h6WOwhd5#PB>~BW$34%K;k#^@g`+?o{>#?ivrwm06P; zB!`u{e?UnJpQ>pmfbuzyz<-oA5))XbB8%+wJ;L^duJCyH`u3Um zQ*^C93j*egjT6kf&GHKWR+SE^Vi&4*h3iUuu$)q_#siCeFUWel;@K96z+VQsS$wAM znj%kt{C+L;PTyA5<={A9s3AexzzhK{db<8yudbbpe-_pb^|`J6QU`=kPJ#GtzWAJ= zau)o8jXs^K6xBL$D_o3%dKOMVvxt$=^l!w!!DabLczyn^>(l)1)Qr}Dt{&!lzGXfO zNETG@H%5Q!E&CVOO6q?$I; z_3BCwye6GK8{|Ce#m7`xd*!A*r`fFUtoX)zuUb}D{=cbwpR9A?<@ipkW#1Nxg(}=l zNP^m~=usUa9kK=DR0S-o(QGx#YbWSxrKe?R<;-PqMX!Q%;}c(u-J~i^#g3Si>~H>Y zVg>my2j3ZMa1)^TS-c~;fB08hw!xRg;o%@nLY9ww_)T)j)5PoQzHdA*@xTOkK}rP0eAh1L?&BWjFNTm-yTT z4Yh=j9sURCzE`uR#)#KUa*BIlKL`f#D=dc`G3d6>;5E5WOGn>tAI@sph0~E7tLe%R z1d(37KfF_w9+jecs66-DDN-iOWn#pu$6O-~bdTee)NBW7;65K94ZP3U+jW=x9(~a> ziO2_PS%dxr=&UlI#by2vsvifTwW4FuCeT^D-Hu9{k5GR&1GNJ}4dYClbf6 zXx4R?@G$=jtk>E=vFp*kX$!=g*lWOkTs0}*f2AKZe3^7iq{o2zq9h=Zn3*rS$uX}s zP*HHoWHSIn{hNHiWSVaJZ*=m{;0+1y+PbDL8@A_gHlLDS8v`JmW~!XUdRZ*8pF6TP zgHJ1|lSkJ2@%)U&A@8$$$6JiXw+ep#E3#oU*=6;$6;f;-MRF)@7G2|;9BKZqnR9cU z@7dsb^|5hvk}}ynU!!j|czY?3FJ4Zt?JJ+D^r#b=uc=ekOvD^vA~`K+fEr7ueT=-n zJ{u>{`dvD1kJqTUuJiWw(6hDwTyE(O$_5O24StgoAOM~};cgD*my0CW& z;?4uQ306{U97SSLkr-oPXJ=Ih!IGv!wiMX1M$}OBf9UogqW{{F{Q0#XUz0>x5aGXY zMFIJEu)1m&9(O3R8!+w%DRv)4YANkv;bI|*tOAo3@=IW;3D^kDr#>jpf;turbT_%J zK!yqAO>1K3V)cDbR5F8%{^muD&MY~Mee}qkhP-KJ?NlQ@7l|zJ!6>DeFUAx1Ju`_# z>Kf|~Frl;S^`WE99>CbRP5gF;h%#<_v8CehPRatMCtgpElU_I1rOJV3;Q_Eopx6}@ zSw?B^PG8IFP+3lwHtwDV^#}idYcT*Q+=D+@RrP0cN`Gs+*nPS4Fv)&q6;dV~{h{KQ_UK)ni zYu*+&b@a=R(kQ3rc?W>5Q$U z*o9GY!U`od6btG9-5_QQ`TC~ldP%MzPk_P|x%7k3D}rieqoPey4f+2TQMIxjdV)$L zx~64L>SPu&y>9m+dfnPId#824S|oGQrLeZ>OR_G~tHeBC&)83x+^L%*??q(M)yfXw zqFy{@kw35vD|ZJLGfPLGljHzBt|QPj3=G-Yo~Qc`utK~rL><>({TEwK`)yW3l|iu( zcSxnQMrA48ro@Em9(t)5d93v6B|;-;7*z^&b&3Pfg_R2FF>!BR@mW2lO_b=~ENPbacml8S zb|yoB#GQ}m((syzi-Y?7TOiVk7Z~WP%mUIW>T}y5Iyp0!juD$wCQ&|Aw>_#-*!4Qr zwhu1hKxMGUmsXG0W*wPD2vs!x_Suc?2FhMuB(t5CmbQ|bS|`5>M#?XqROPNK=!Lg zsXHpT9J)Q~h9VI-CR=FyHX9~X`QQZma#Z_|`u?AD{_SO?aBB-c6%HD=!WOT&bLV!@ zN67+lZ6Gp*&YjyL%8*`^Wm0*Xj_Ih=^Raur)a(?EKyAU0Y$?oNFS3QhJr;tNenTk*_Nt7U{nYb0aAUvB(U-Iu{ zbQsx1VUAW+<%FZk8`GSv(K1vJnf##PBIP;vZ~4zV-?WY1exWFcIU>4dVuM~Kqyet@ z+@Jn%Wb?$Izi>N-{qt)_a)aA;{G_MrKW$l=UJ~2E!Pk_nLAjn3RlTAOayxl~c(#pB zgw^?FRUY!3=X)nhyNM~Cv`1v1n?p(??vBOIdPl67oFL( zx+pS=Zi!k|w*s#b=L3XKA>31ZINdXO<(2=LFs*iiB@UyBkF6pbM?+|1VD4=z#X@Ci zE~RajBPj|BWEtt5GYuiDL+TV5n21%yLJO~fE|~Q=pmGwPSvM}5E|+eZ+7WPw>E^fbRzQHu!aEZY=Gk-9=An{HFYM(PYqqOkJS9Mj_Xv1jU7 z^jZ%2pMk_chWj;eUr|$|Le;B8dcaoI5t2EnRdq$VI7qK<@v3J~0S^1$hmHARgZnAW zt+oaOET3b%z&+~E-`(aBVzD`~DIG;*s|&||k60PQN{WTT-*QTOOkFKU!g5`*A2w&? z(x3RAotomYUAe-mRfSS2U7E9#T7lBZB)3xJ(FST%M&&(^xk&1@(Wi|D*-4{U)ff}m z0>(a2hNVdWO&(dA-I{bMhFWg=R?N&5S9)BhOUCKu3wH&;a(j3Xau$97Onl(-%(eXR zvuyXtX~U^oa9!(4I7<4_VpTvPI*K z)eplgRz>+|-A1x~uu3@>_Gph;S(++}1tP%;O8c4eqU`QCgJh4WO{q&}QdBLXB25py zRknu7lva9V&B~N!(4Z*uZRD_Mif#`mVLBr&s;Y=ylkBdGjG$gAn<#^^3doGOkv84b!F!iSFWhwa6Q)kO4y9Y^4V+t&Q(+^+A?76f&D&+iCX zz^^yVh5OPWtv#TeZWYC@q(~g4y{y7!LP#M=N@74swe%S(w#jD@_A*tdA3Kupw=dD0-f?|(S!Z5kyD%8<(W|qek{7p%0-p$z zA`?R93#;V)_jjuB>Pxa>rTLB;;oaZ^u&TtPF~KF4`*LXmlcULG@F_z>qn(I)fysI9 z=^-KO1FWK9Qp@WrW=@1p%PrVoEYf` zGQ6OjF7_##2pLYitCV1MII^gigt+$-Y*fWC#L!^zyfJeSWS;jchIEb`rT=7GlgS}c zedMe7<*t=p3$F~RCLqoA>JuS;eJlLhq4m)5uk88%9=l)Y^B?2zv)oR=Aki4tHQ~o+ zt$Np%U5SI1D3>k}x6z-B-{gl0eL+Pa=GyclzkINl$y#*s=L5GJ74%vfycqMXbUfsWodRn z5z{*P%}{VRL)x(PXedyy=bquh*q+}plxujw_?4?)Kb;(ESq)|SWvwCGTsU+K%b5X- zUIWEKV(ViH2>-Dqq*i1IJQq|0RiJuC&*XY6i9RJ+I{u=nYg)%-T!&zVsGfNgiuu75 zRTg~>A}Q(c-%5{r@ZUus641jW2;-vi{b~a801}0>@a8 zaqXAq|3+f@h2LD)@q&_yftXoOv0w*RQ`%Nl^2|72D4V6@*>oB@og38+5`<7~V1h6~ z00&w;T0ETSsn`w<-`bUD?swLKzP~4W+vbt^1kzj>8ys2ZwSxKLDyVKULUq%}rcfGa z4w(qubj@<)aKq|@&yltsHkhat2&D6z$R`rPUj=s`p9 zP0{Cu%}5x6)2mVL$RQINU#s)E`FYNUBQUk+qgS`tMo(PU7vTu&)J$9-R4iBy%;i{O zvO93`xbu>BFNX{p&sS{MDD!-#BQEEjTLuaBjryAJ(gU^(JT8j~4o)330teMKa@|4o zqH$(Mlz#*u?3yB5w0 zSWBtHcEH6#>}oc*f#uTuR8FXKcx=ovu_^7R>r^gMRD+j`NkIdBW{xuq)i99s6kohB zV&BCvfx6LBi)G0at+`AR2D2mY0H}4v<;5Lc!Xu$mPmb!6HSyGt)Jk!)+&oIPZ7XiRs%Dl5GR) ze4Gy#1_ehLzB8g?WYls4{%+`z8oRGBnmdny!ip$-kY-A5MtOg;Bu~ zBC~|UHbW0N22M>jqsKD_+)yla-Rz~Sh^&!cWlqw4{_Q4)P%K8%A$lIsRg)edl7MeR zhdJRGal0fzh@moFtaq2{?5sARJ-|q88zg{Q1zpqk3Dc#?evZJKo%ng`j@q}?56>b0 z{Fnc_oj2Cvfi%B<`U8^e!X8M8l{=b6v0ym#P|qq!fsUL!{|BBQyB~T37q>fS-uJEa z*ec8PY*juMT?;=7J)y18y%q1bFgQ_|Ngm58J<0`Jz1J|#SdLf?#!v^|>LE_p;pdV3 z{O$G6Ffs9&d|Y^q!$By8OgR^1y~_L3Y9`hz3I#X~Ef8Owp<5u15%+lIxECv76);kIdFb;q7 zoeN~uOOxR(vVzejiiIT329%!dkfnrS$r@6ZZ}e#qEe-;fK%nr#@-;k$E%wj|YGVD` z4XTkyx^Dw$m3^uzK|c|M)OnjJQoxHK`!1G@?e3GR6@OrF{d!d${d;?69xmWkFt9 z@xi=35(_wD(=Ppj#19rnb>R$6ffZmjQfw+k)>7I#U%4pjb=wL5MW?l#dGgw6f-Rio zuhmSfnq-g|Bv(BvJq(ktdKx6B1tT^2eDN^zf9qxlY~$>Fyi*rO4o3uj&zBK18G>~G zRQVRKZgN>z7?CKfrt3s|nY%PHS|5z;MTS}^`xwDMEN|!xk5#~KGerGj!z*9(LXa(9 z=LA^kgFeGOk3!1)|087-2EJ;$MtR^3BeczHQW^C_ZSUS`mPzG<%IF$W_|mwghpncY zJroOV6+0!*cp|ZK5Z8w6VqPT$|E{!6%AS(RN zsVcgul85!6ic~w({#xF8AHMs&bI(2ZobNMPm0RDsyNeWzWSDw_NQOkAvD0sIo$}FFb{FStvoxr&9%}o`6}V2UIJS z&*kOeICc$!Iv`GmGc;S(74u4_l>@Ar*+Hikg>9u3tVM`5)&j2Ki39G-xkssKkwL$- zjG*x`DGt1r!x3+5)?k)Mf!x3x-tTsw>}N}*1(;ma32Dz&biB)i8Wui~;qGADpXvGN z+vJBPqjGNcrAg%C6!OIEJH1V@Hz{(1ibh(o?GT5+((eNKU|1cUB4~hj*?mULKo<{tAZ=R55fnXj>_9LDr%r1KoQYMJu=zBh$WaYQb_p%XtnESgPA;rUS zKh@j_&;PW{hUH^~9C$9xktbUt?svTwv1DeqM>iwieJ zM1{lFS8yNzC(3Q&&CaD^UCchAE(gZOBc3Pb?wrt~JX8N!i2&RAr&i*I8v!E?cV~b0 z$H`7652pC(FFqhK4!pdRVdh$;P;4SaRv|t1r-|=bGF%@<#hi&Z308(gtn5ZxUd9U& zlmFiz(^Q42j)lkjSH0Am_@WqKkmlcU9D?D4+>$X9^(rWh9#Zlw+sM zpj%z((&ySB-a~eSPhR9-ATN}FxKpPKR#oVy>U280VaQtB!R7zj5*)+Z;sumR%l>Q1 zFDzBy9Ck$E$R=u3Zg;~BBBYKk@J$M?oo`YqPZF$&hzm^&MG14P(Zv~2DX@;mPtaI1 zCWs90OFM1zgUGklSKjnD*^-9g$0%UCdD;zCr3+SK=wjy7&{fi8^%3FSkVhd6sx31bVg3zF z=6Bvqhxs<%oX9>3c^0C>`MT}Q&zQi(Wqc+izPKUKE9ORqE?4l;1dV5B{ zfd0-3#Zo;(SPdyAaLv$^X@nvMsj>o1Uq~yYK^udydS;vOZqTMsz(;{PD}vUI(@{9FoXbW6?%KcB|JW2BP~o2huZ7 z70V(sG@aVV;o#Gpf|&g?L8&xmPKx{SFs#+V*kZA8!z_Cf+>9MLHUr6s4RIqUHgWC` zEW>CHYgjk}V>^Ubq(i-2D9J(-(V?n-Bhd}*1=@`U@q?g`A#dpvx&2lva|;O7GU?** z?V)|HHL5t)?k>#OQDMWeWdvMsqhf@q@#J5x`G(;8Cg0`Kzmz{CO%6N{>M`?E+9|e` zB3G&C_uuP}OciE@>Z*c|2z!N>=JyNR1&}|0YdE<6cUq|{l;Dv<*#SQ=F1T40jLz^y z>2BYhv+gPL1CIy`B_F(p83Z+|TY@9P9{Gw;yw1Q=YGgmr)j+B&9)tN4!f}tk4y_ja zz=sshieoEXu154T7s4>7wn}<~t`wF^4Nt@M4WvQ<3}*=T_!}yR3S}*dR$_ee(9=c` zlN%4rxSbpRXDpG%4`!Kjzp+tef?4MDU;UQo`K7x!a2ywW9m_(xOOL^eG@$3DlT? z40*5TP1LP^NS=xdW%=|flxOO8uA=kl{c{H)+v&XcyjZt1a_8JOI>CK+V5#)5fBVb~ zO;1FA2DwemJHNOPXqj2b!DMAD+%)8d;JAP}XgCG2_lGvZ`>^gc zk8q=%4{^WdksJr&sQzTzzpqqVGEQ?4I~Pcj;hXBBs{;~5?W$~91(imQ`!q$p`09aw zzO>-k_f~#U_wkS4I~Ua)HFSlZDfQp%-yeBFuy@{u;Ix1S@qV8H(HTfPv=@-9L(8x) z&~npGsNsbc#~r@k-r080Qh&!$kxe)hb+ClIoiPed!VrQIUI%3j%6lGbop(V|=zwa! zGgf}-4r;sT&EVi^id{S5(I6ITWjH0W%%dZODSWNM%bor2}6lH4% z=jhhR;)1Z=uoBp-^XODQ2vcv1ZnBZ8h#)N)I!^L=S8#uaXAM6p-eRVHb)$y-x#^zUW8( z%MLpvanvApDOV~BC3>a?l+xqa(l89u8oqd=p6z+}6?3~4-HKDBcBW|rLdDxmI&+L- zY&^QhDnN`da`1S>m@ef8#0c*9qGhbvO=MvD)3NXrf(0OJ>!ZLB2s?~OjrYn3C}tNnDK&y zla>=w@LTNF9~~GVcj^y{-_x01+$;8hKkwc6SnikZ>S_LCP2KeJO2j~ z@4%T$d1hU_kz(}}SxZG9ckWPZiO37p)sx+tz4KzcAV#**Jr{&y^}s|PKc_>1eu=&F z8;5n-t2^7?uI8{`X4odLPkr)NkLhYGYc^i?KOER1aMW>h!0hCNYRR-5?c=vrf{sqE zwo_adxdU=!_bHys+ZjwJyWw#`+NsL)uLbIhs^ClWC&X&-gpC11!+O_lo7}MZ>GpXQ zGfdVa?)P2^By$R>H2X+*P%O}&6;RP9G+CnO-o4CvR}ihE&(Oy}bTA}9GG`{ZE}6E+cN?jrv!LQ(4cjhgj@qL-3YlGdeCy~HQ(7a_)yL@gDfa1; zv2=pj*KDV}RPIjrTn*Ds_tpPyiI~roQQ?61X8-*G71RatP>~>7FErn4g47pGf4Ljf z;g>bLK)4}zi3DqIE(-Q_9% zmLy#q>{eUpvohoV`1j(Ao^6^q-z@nq+1>edLL2ay@v^Yi8xmg4j{9O>hQx8B3r;yj zkA7tudT!54o<(l+%Y}B_e;HXZd3g5gQ;K~|kzSKD0xera5TfwMF@3%-HW$o%XDr7UR_%wce{&UhjqJOmVQcN3}t=L)anS z>+%Go<_o@)`Q?Fsh@a9FRT;d0#(tkop*7$O)_cW;$4@y!$0^T6@ft8E5?XvrHj}&Z zq-8GlYqF%oky75u?D9(?8Jc@a$meGshF3uJ7Ueym)Vd7Pd*x~%UAIzCK9%^jI zT|gxq)YzRgrrQ?E`rVMGRL`7|ZWEpswrehgU2(xAT-2u=s%mV~#M@cq_R}u43Ls`< z&!@NAb>sSv|KUFaOykk6uCiE?@4!L4V`d}FK8giimNF`Or~*DkgrvqvAlHKV)Tpsx zNP%pl+k|Jl^$`Z4uX{=yJIXwYBppx-Uor2N@(4_tbjt!0gf*&Vf%p76=ng-~gq@n< zv1ZzCAHxE4i(-veWf-q9V$?q4j|DSeSniDaEIudsx%UqwZw*>zNpd73f)bakMBYpK z;d?Y{_?-tAAPd4a9pV>;;*I90&CZ4Db_TQFGc*mV2cqL)aT6hVFt)zn^H#@pEw8?C zT&M64zt+t9ysr^lVwX&Nr0Qj^`gVDJsOoV&?U@wZFKkeDh&xq>{Hka4Go51F83!xi z!UwZfG-tC*_+fTi_tt;?-{r0z3qPoBQLJ`@Jf;kdF=+t@s0qc6v9qm1kj;Q$bwe9} zSbRZf@`AeWv(4w}9G5{fOPNIn%^4Z(uus4l+( z;U9D6XUpmp*Jh)Ee&AX?1LR$`@9UZuZq{st5AYLJtMqYrbJThFYw}!r8Tp?^h_l1O z5oipj`<-6U$_&hkbF-Iix4v5}y}e=kHt7iSMtF6LyQtZo*pn>VrFD-T?lcbuK~UK3nQ_FwpS5|bVIefiWskpzCx zR|k&#gY_FW8f~W7H2gfH_mI-iMyMX{Ri(|;H7b$0Kc7yU*{L;7kc}K(Sa$-PlM|+I z$^+#_b7yCC4>kYoH!>P5$*cK9P#qW^98}kzg-I44_em1znKD5O+oC{j+KFEDv|j3I zsJkZUXGxdaJ5x!@t&fWOfzAl<;@GfC#jz^=Gk3l z2*{yWFccY7^t}))*U6`!yW&rvS!-vy!rI^f{v33|fO;0YDKrN%vYwE5iGkfBdq$fW zlkEAjPVnw6eqb2pFxz>K!|fy;4eTh6G{MAo>FOe~-+||~=glB-f?^@;ewd0*rk4Hv zvxQ5(aOJyu7xa;%WRLIquiskm=&QHh$X}(1ZV+CSRt1--Izuj!O!^7A9l0i`z@tmKk*fIWfp<1iiEP5uTy1}3fk$)H zEoFh=W3tI>_!B`UNZ32R4D7MPW;ldK-sX+rVK;EYW~B7Ir%lWM(Z>X(7woB(B*THB zbjS=!r4$QF$lIytcIKk{1<&?5w@E(`6x^IUpjs2LP2MgztILgk&Rjf%S=H37-$qw;g}8U?KZ zE1@25$GnHLx6`enDp;Fo462(w;B-Co0z@u9RK!DxRo(1($p9qpJbnv`H{DKoZt=>{ zY@2z?bD8IWQ=R95upW;L&2{$UpiZb5=_LswV5xz^9OE5#L9{qS6E~|=T1T()y~3gS zW$l}@;}>r2uQ1|u{5s5G%h*r##(m9Fht6SnQ8)zhkQPfn_1wI@Dipd7t@BU1jN)ugUYFX5qn4 ztG^5K4x~!>T;8Cpqt|=o((QtC%GQu2U_Oh5C+EHprMn`RZ5?pj>X}PHmjjxn7X+mX zT1km}Jh>`Efz@o+4srSHt4zCUy)aLkE$k7+0Cmf4pod#EHIrUGr&-YyRUo)7Zc=n8 zie2mICDSTlnPNCE?toK)pwcx~jX8PeJS$zwWY~T~z*5Q3)vF-$4jFV|aYNYURVb@9 z=mYnNHv6ylIx1{YJ`|3WUFQH8MYz6%(}npz*=)Hc`ErSY11CjsB+4DsVrDJmxg@JH zLx6%|`YHjgJWhCs$7-MIrJk|(0q!9>e;4~F%kU3J+yp)uK@;WHs>51THD?`_SzDl0R#o`*n{n=W*Py>VK?1+-*q$#=)MINq?ki6Q`g2!9>}n`)4sdBkm0zNc>B}qeX~sdi<3;dkL12K z{>u@w$>Sc1g&5)v&l&c*#c>6bo%_ zGkL&&m}5J_rj5Ub(4Q1mS?1&NS%Pxl#RiV--J8M^mp+mu*y8?#Abz^&GN!{TMuNeH zih0k-QQf-l3^^nCpW^}3jwy5A@i1AOrl0IxOwvYLif~+uN%oo9l>&+d&W3HU zf(>$YZEpL6S`;glNdh1WJ3%K2@`OpUi;86RQ<%)Gb$c#vVLC$^f%5byxLCKK^t(@i z0q0wSTgoO)L$IBzLdJ)P@$~uV_i_)V(?#EUY)MGS$I3V`ia2-`3uV9x1=Mj*RK*lB zuK$gsC#;(Qjoz*HIV(B%DAuyNo}*wrA8PB6I0bbC@A=^ZA*{{_%EO?(v_ZAh4+|<< z6n(&-Vu+NrD0<+tgfF*6fq7M(A-4qR?Ly{(fy^rtWD6UEE4^(!z{6V_-*-7Q*4zGd zbAOj3Onw~qQ``2)`Hz+pe2NwHWW->^=~1#YNAYbS?>lTt`&B&2rXZ*%XG1WQpP93_I|Yn}h6T)s#Y6 zx3eAy%x(#a1u=6Pm6>!$*v+{y61(Sa4TFzq(^_=De5ZYn%(4GxYSurO?ga>s=f-Y%!c$I&@ zBpDK;Kz^rK)aG_`c7`Ta((j7jf3CJCq7zJ(p1~0XSy}o>1q6uO!2yYh)HAzh_Pg6~ zB5Unyx86+HraN8}qq6PY);3E9M2F3l;9xu414|Qi^ioluVz)9DsCv3kRw(Iq#!1*3 zNNp%ykSxRsTbPgq?DpxPb;j^gk$-7erAvVzON2S<9WY}sY8AnQclmv!sR!2_rSLSl zbTyEQReI+!m!T92S>-X)3T6?B>ZL(z+^W>M+8ZADENrGa;7~vZByfS;IgromuyxQM zwH|%ekEyn}k{9|Wea)-ncOnz|0@MES1X<2+;nH#ae3E14y{)I%breaVq7A$-hXMcw zIBkO;dIMolmSDZhfO{udBDv{05rf*01>pswp_?Y)=a**%zL1)Lv2xS(pOZv>z;NJc z5@e+hONZV_NppHJv6EeV>VlJ)O9(oP~Hk#OU^0pLoF1pK9#{^&lAJG>!@04ZkoWr^^9C>oMDzs5h;SVew%syqQ>mJ{Iz!No5Jv+Rk ztm@_AH%!nrufb%p@At$De_*m18FwWw$T@yC!*LfsfK_$4&FG-mHi}%MqRYeDNX+zd z`f$*OMU^hAB6ezD{OsBf(nvRie#?Dyr$|XyoNv8WXT0J*Fi`Z67FJiUeJ)1<8(pfp zi+%t(=ee>63^J9a2pVW3{i&g;Oo%dOsCNZ!BU{`d?RS2?HjU^~=cP~k;JrT=2y*H4 zY3GUV(^ek~Qq?tdUg!}aw!f0u#MFv_erC5cTX;OIS%s(ZS&1xO(xQlq%vI}_2v##K ziUwtkXoT!gt59hRpJntmUc$O_Nb%O$CY&v7El41R{BY*LW77#UoE@N8NR!x&oOij} zL*imZ5j`-c21lZxc2xzn9Vo;zH0^>bF0Bmc+dwDcm{m+Cg(d^8FzY)*bH;mRL<+gc zELya5O5Bt~0SQxg`}7*L4EjOGU=P{tqi0e=9{@ATCiQYryX~@PYz>~5P%y>=%J0|a z{O@vXa<$LsTH#rrTAvmLN>-puaEJKuTLl7?Ksw`H>r)Ek&|AEY1snrQFj2!FZ+wig zzxm#FwW~kNGFj&3wM+*dsyI?&(Am)$hX~y^w=%&Ej~2xQ$(5LQfpr~ke$Pbgar-)H zPjuHA(|EICO~6I6ehL`n;SNC_#X`s|i;6D&`)dY9s>=bjnr)#8o@rz&uv0x!9TJz) z8%a`R8MF57n^4MgigbjXBx~Jk=HGNZ42p)kJ=O(xgf-gR`Lyock2DNfWqbVGsX~l{%k7ige#|WJ2QHb7N#o4cbpB1=%MGN*hJ+R~gJJV!2 z9{*6eoNOH_OV@$By~fPU?V{Kn6i{(T8w0yYFNJ~K!GL9@jSG|3d3K#`SHH0Fcur&mah0Zj;GR2^~ww# zKx7qEMlcnoLv{lWH&lMNURGeK1?{kLSq|OjB1y5Z29$n|%Hy>g742S~+UBSdVV?6c z&j&#Enn|Cc%T%W!zb=PfO8Q-6Bll^}&+HH*lhD4|j|C0l2l5!GSYIQ&BK<#P{O(uL za-%%jNS^e8f35wVrNheM>203A(?y4wjO9dEI?7Rw4CZXtoh)WgEu&DlRx%LoTaM~l$o znawamMCzMDNueY!G%s{Qc@c)bdh1^O(A5+7lh>M3)v%ZiXS6!@Mhc~A4!xXkgF`zP%DI;+`qn zF2BaKLA<1jE|6#XC9B)OJAN*Y3Hp$U3EHB7;-e-3VlTnHmx0pP(Cyxby|Ix(*)EUn zIjzCvsx1l~@@y7JU|rhqRA;Fj-s}l1ZF`H7C@bwH0^3fP5==97DkB4&# z!Oj785PZ3if!Pt3$m-C2$_MZ^(9Ntaj%`*U@YaME_%;i=L7f$|zICV&o1r=CvSn%| z5X=|Cd`-9A?F^0QaCbb#lJ-R$H=N$kf+7F9(r$vLfSq>p7!=zJEk4-JL*{rtgm_6ULoGq z@4CgqHlw8-5cDc=Sam)oEbc8UUFJ*~GPt3X5K0pooaet0g@JW1)2e^JO)ug%05~kk z>Z6b5&I~hI6{+)rZDgkd`#|+(cA|!2fidC$6@7h1nXu9YHS;oI$Y`Sz+%b5&Kd4{W zp};|AC9}sL?10BLNb6rh-C$y*z)0QBU49?8hlH-1JOZ z)Jf4cVFRotWynI;Xrezgh2n-CKo^-G=-Qzk*&^IPfGb&5XX&Xl8?+F zR7J6%LAf88FQuhYRL?@syj}H)@cImce|}XypgKzMU`1GwKk(8)xCLv?jHxm@j2Uj3 z0Snm%`HMkd+U_Ih(qLFzhdN6Kw*;0VldGnnkB*gX{4MMZD*LG8W#z$)QZpDudw;%2Bn)idh_ z``v0(g}#@W#fuZ%FNlxLSpmA}OJ!~1SP2RX00B+6Gag1>nnKx52$Y#Hb3iDsC{kljNc6p#Y>gYyy-)t&avQOnzXaa<6BK zXuDhWj5Xm6u&$Y**(clLp}Q=o1q#FzQLeV$D+@w?4XSb}y32akdYDf&1AS_|%lH>O zZ4V%$<>7{gb!WH%)Oq%{emq=24m@~qZ~?7$JElcZ^2NmzCTU)gcDW5Ym4r1z%BDTD zWk&8Os{u3e?RGo*?OE&_D$|gZIsI3^CHg5aNgY0&$fa0#!I=~k4foK87hZ$!vms#Z z+ZY9X0MXD5vl^ri6>HzFpw?0qRI=w|X|`--WRv?ESsv(GbtqDaJ*bLbejir9V9O!? z<)&@)G|MK%0zc(3=tS5C-$US}&A_w+#6U^9O1}OF!Mf#}d76s0FPW(vM zDDG6HI9IydoO_xs5G1lmG9%+*h9=8-m*7}H9!!T)$e>e|KX9#dDf^j&K=+@>q8HN* zao)u%LO;|dh9VP7lB`o@^nepw^6B$lgHEkXwky)TU`|(gI7$WPPro0AUd}!yKhUVT zlB!;8eBe-`QB6SSn|X8|@O;EUnlQ3?7D=uOHakBH(Jd1l)DA@)jgQ~vA3F;cI=w2x zQq?zN6tCzysgsw;u9}p`_oADTmmrg+42{7}K3xYV`Oh53s zKN7Tt0CiAcN}w^jM~8L!C+Ile0Vn;8EU3}ocW{acvz?jRWXRrpo1E{3X5anu{L2nN#;-SBF>3*4pj0Rk8$fsi=adFV8 zgFZ?Qgz1o1>X1v1a-F*!$HehL#oCVNe;qee9{hdRe9QTi!)6b2%#cv|u~Cd5Xi?mK zBYxJP6SA4%>bs2saeQ95wGiObrY)}JhQJ68{N%?w1_C6e5$X5kQ~yK~9C+cN&}{VC zOtEP=0f{~=yC6^1cKL6IrRh}>x^CyfS$jQ=K^CBSIpZBGDJ6HLMyAXO1B$iq_-v@K zMd!b|4PNk=B>iqlrmG1a74LlQ3|Z~K@PI7gVatk}D0U+Sy%60f-l|4xp{vmzb3g2M zM5NO*7s9T%fLD?R(&q$m${#e_q`(SK>7}m5q-fQ@w2dg&X7r>a)@0zUcq0e|9m1nUv^sa~pus(c3lO;kx`$v{Aim#*HuO?XYcPx~+&>hPO2SEK_@`JlLWRHOqLYFh7-uc*X2g`S8eO&&1i z4my1tbV9h9T{RaouIz=iahpRsflHp%vdA(;lOq!{RgFY#z2Wfc$RVvRA@U+HUE~Va zx$9Cyk5r{nY07aW+f0~;G82Xbh+}{TZJz1R0|MfR0U^+ z8uJH0b{=SlPm$YiWq=6HF7`9-W1DozxZl0`c=YMsc)Nt(SM9J_zTz2^ZqGIOFqzDc z3&|b__HfUdIakLh_Ao`Nu`n)?J*Wj?H;mrmMusNc5BVf@S)tuxz%o{hl!vE-uiFJm zNV-;ep(G3V9oodV_slfTY&i!&A)~6&hUCeg31AclY0~E@kl?wS? zCP`E!>=hypCW5W9(w_N2?Sa#Aw{x;RaRzhg3NEq{dp?~>arJu>T)O}%X1r({#}#v$EYj)TYf` zR0jsB$qiNsq}zbhyHQyz*rzz^jf7LTfShXR#tK=tNAI-E@mCdCbflHA88Rf?KP}oF z?{2DB{Vy$xwfWc*2ZjwtX?L6W-h!qtWPYR4rF`MjZ{pT5#a6``dJTjMHZiwnC&_Mj zD`;1hxG$A#3azmTp7v~Kd+6Y-zxKIw?z_+c`1e9TUV^*u1QUbnvev0ll`J)zo{@BQaU6GnWOt}Y_` z9T+3$%`kF;VnML}Fa&aydM1h73d_(ece^5OmEKT7cs~*9o;&C^akh4FPNTSnt|$HO z-C|vzqBlH^1s=noo6Mk7e&7lhOfA67Zln4PP|8T+DQpc(sADASrglknx!OwaZkX9! z32qP~C*nzQPeeS-_fY2sC+Xu|X&M`pxQ88!Gh;>4DC)5zEE!FTjwyQQdnPFTwCrd7 zq;U$lZ`Nb26nm8-EmZU^-Re(&g}#m-%6(X(c|qjobdtKhhyvIeK$zx!nq#uikFJ4r1+$JufJWgty4 zd;n~uSO{31qN4YQHw7Crq;w@=rwArrUH3dsFac|?q8EhRbtoT@s;;7AB&dM9Jov6? zU9j%JbeN5^EsB2E#fwo^LJy3t7=~Im^@>ZCaEnK;aOG6p&_3t>8LCH`hLx;Z_k7cn5{xnAY9EP9F{(j3_E+$O9uRH$-67Rr6 zV4fLDHd3sfB5SE=OmoBk^XUd>-G_>pNYu#D;gR*Bd;AwSDq%fydYqC+Zw-G#W7y`~ z*<*Obm~P=^2Paot$^E_B1dZas@9!m94h#)oks0Pe@1)ogiWE}O9pXkYCf;SrbQqs* zB^hpaLqHmv!4nyp$FLZ6hCU>26PI~Z%+r;KP++1>d=YB2VYfVdujgq`pmW7)?0Ql= zAIwCoB+m6VL=Q(F2>JI8U&xpZe#3Y8KYG+VctK{;xpnE2KJSyRSV^oRo4zJ*it7Bq zHTk)yoxWM}hptHhO;JZZb7`P|*hWwda+&A3sOG4(q(Pb=UMD+ab3lyHD`UgR2o2fz z5HE~SSASd=D>j*xg%>s)A*=X#F%DeAwB5`IrBmz%ilkD}I7@0#tW?5U#G!!Yp&g1g zx&;;wAaB=@WP|m&ZH9cCajX~c>{8xP;@SRKFIR0$ul}M59Lv^y`}?HYfybOHX5D?7 zVxc_pBo#gQttKeikA=WYb5u4-_X7*zpA^|4-tKl^b5*iRQ9&&eB#0`gHW!$UzmXz( z8nQI7!Zq793%H`2+FEBdrUClXKrHZ zYbq>j0(5!lHZ~M0WPbJ7ArX>(AKG zVr^HB169^OVocY4ZGP z?5I32a=%#J(Iek(x1+p9m&v+c53sXL{)&@KyN~3)HqPu3Gw)^(#qOfW4kQr9RIZ!C zbM%%Ob>3IJfuc3M9ZFAd86ZQ`8I%=jtZf|huk^+^-Ldda)nmA#(mO_iqY%N@%HV=M|Z|E33A|2F2^Lum>5$iOA;Z;kYT(* za=e(x6F^eFO$dpcS+YV&ziUjStuesab$r3W*^rG-{J*;7|FVpt@Nq>P7#ti6-NxuG z&W3Zf1-|P=amu6eO7BZ80?6R@i0F08fMK=&1NLWM%!=Ueowt5i~Y2!tg@G%8X2wJI3+61@8e3cg;I_Y2DetL8$GJ}VRx1c1vK(kzkX z4j%+-%1pYV54bKNNFsuSnyJ9qtWAX+1JsM^cg>V_hV7Qa%xFQsYp)oq-m2(h0Y!4q z=^7P~R*jIanyZZU{3sygH$rOul#}yo6Ux3&a?Fd|abT3in~?VoYI`o7*=hJyGKc%qMxbvD)!X1iMP46uC15Bmpg6clh`aM}+tcLvb9=t&`R1Zh4`H15DJ4NC#3LZp`J%r1NM) zncOaq2d=%UJFe@2edrEQ&BQ5>y8^cnG>c>*)l+}=HwE7)pydU=N&mU7|Id~Ie}^rj zPyxvd#{!CBBNk|ZApfsLu{>(fsa)0X zdK#zJ#>tW$-JtQ|WE3ki&Qttw@=rb0`+j0FQqS*|D#%rSp*shzCwOi)aNeWXyA-(v zp*!I@RflKc+e1H%N*1_xOSHiSG8^jX7jyF7PI)`ezu@gfCp3%tY51!UjL!~(ye%jy zMHwjz%E4eFS{4Bf8E#>Wden@)o)bqM5b7rr+Gz?n+C5@R6 zK#Aj->+QF;xxv-42LbydaSN}!L2q+IbN3R^#|O_*>=4gWmS|2-iFs;C4B@bbiG#5s zdq$BwZ6+)tfPhMaxJ#N8xkd&9lDb^jA#P`SM6JqN&jF`oHcLL>6eB6n4mj;6DiIRy^r5E?iijkv;&ttrZW_uHeDS0>E<2R_d0fP^$bD`S7d9o?A!?}I}-;dUYJx5ol#&Ej7x)b z9fTvRd`^pXuy^Ou*QR(urCTIRVk8{cka5TX^ zGkhz=Zx@}G-Sfc7%Rb>fk4(QzS&ZZf$@k9|?h~YGyP(LjRCO%;48+uI-D&)LhvU;6 z{bln$%ddAF*3o_MR-K2p$*yP+V~ywp)Fgl}r8=$$a`JaRoG&5~rX z&qeEn+1jHb5QPMrSngReea*DW88Lr;%87zW-&xvcOL9Df;$X|tCc#CXJDJ4dpRDoz8Z4e>#YdIM zH^G7i#%rID{XPS*2ziZJ8k82$AU+&aE9!QM52y+#aYu2xrl?#0cJCh&znJyq&EGy3 zRj96l`)b`G&;r+X_;p5I^;j*ygduEm5$Y==rV4$odX$HM_NW><%XzJ&LVh%`obFVW zdDK98WtVhtY>L`;#fS++V}S}UvqaTTlNC6b%u?~wUwlAfUK_In)N8|-BvL3gks_<8 zXsnaEG+(z_e8)9LvUo4(+*S)Jy~kxP#?npJr529Uza7&rCQJX(pXDY79!n-&>8(SWs=Mw( zX?x?b%hvvT#`JAk+xR%HOK@^V>_76;K?dL%1rRqeWK~M9}oi!3B1I#UuX1t$6Pv-e>QF{NB661})F^ z{^ZzqOm>4=?e{fu+JWsxmzmwTMzNPEauHY!P1MVe0jpV}64Fkf5Yiuo0@iyE&dJc6 z0pS3Y4jy#E^LdcH)~w0Upi(rdM1w#m6ha~E7!)_v3b2%_K+>ei)joryCQN!Olt8*$ zk__8zhZxI6X^F54_@y!QhF?b}-2`P+w*<9`?8SZOu&C3g~|}%-C$})gKN&=zn#2Nut7pFzEy7Z%B#*BdpX6 zVObOlmA2_rbT88sb_kLkvOFGo9v2=Jrm)RXp9r&Mu>K>!%k%bux>42zQr8~tA0&D9dRToe>C#&NG zxK#<{^`&853`UL6tU=w}M!z1?=ZeJUc*70)i5u2uo|Y|>EzXy;D2{m<;&yG~XKz1L zlt*>K12Z%=K6`wS2dUrffEH+-B?>c{XlYHdBhUku&A-YVl7b$Z7v!M==g92Ly*9>J1u1MR& zD0#gdB)e8f?t-4d`|lxfemzMM0jt@xhr$nFzr!yjuwq_PaQn=D_YAELwNiG&X7-Hh zp?`tXmjy||>)g)*u{aRxchBiotaHDn{%^y5h`n^r7Ow(9E+lk5qw$u{o(98X58X-Q zK#Zx~S)ovv2UI1>T|s)LSa|Xk(%XMiz>H$+cp0%tKl;``;>i-mTsve^jIVn&UiW_1hokbNJrQ@9jIRgZ&d9-9H7Itj_x%v>zhd z`2}tqH-1fyo7t%S6uXBayI?uv%?@vh|n*K^>S8EI?@`cR}z=w=^hA)+5)4!U?(}qE=B6(I#$X`bfNs zO%uSa{sOETrTF5U`}%+B4y(6ZLUPzlDUP+RAxRt(q@A@y7Bxp>=CF<0_9U|Ey zl^Hp3>=$xehD8F)C>ADQMO1Va1e_`sl+v00IOBRa=l1L;Btx?|=rPj;JP?_5bJSDM zM`1m5gY>vh1}OvQcaMDS+m)`s6HTLDJmxS|veh#l)1}~t+*PEix2pG#EkWTG{Wmro zSs0R6JYy`Uxp7qfP4*`NCd};WDvKrg{01Zko@{{x*f2EhquAXPDWjs>#Hitg=}cR- zIZ%UwVqjo$;w72%7LTji6cJV^9iShELp1~BNg4d=8juwPy={UFj3|DiXR7C3fMgh5 zvu2H3m0Az#8G5G7W4BKSy|_h@tnTr!oz=1lriMXmEchAep8t*tTV+YI=dj~ZrWqF2 zP;3%KRs$RT-!;yt4%jcR3Qq9UVQ-_*SoPeAcDmv9TGz`X^@Dwnaa)cN%8w?Cr@kTm zy~&znJ*k^Q?m6(JI>C(FXn?Cku+*$SBbl66nZzFY|>!7IiKFAgoq3HF)wE8r^Zu_$|lW=kVheUrchQHhJx8U z)rQ~8H9ZDQo%f>87w@QE|wj(*{4t*%hgCe9L7%xh<` zIr!x|yc?C};WY~5FNIE39i6CdfVp`yYy7R!-H@(427W_yp;4+ZR`QWIN)@g#fc`b% zH~P!>r?kFj0^p_ZEt^9gP9bZ}`Z;bgyJRvMprTuur;w_+VoF-Tv4FII_$g0Sb&-#t zBC!tyBFj__$`w=cjLlxk;TRY9sqy^If*VnP=is?;g7G%h%*>;x^JF$`w9T zET03sXCci|HHraMfnZ12%BeX(!gAE}I7#*_cCDjNfUkn>w==yV{jSC*?p2-kY>q;y zwn`UctGI%>KQa!Y%54y5!7K6qGWu{pf#73UX=qa3^J^45A$xrBX~tHJ?LzeKi`Rg| zv{qCLdTtwJpZ)7H{qUZ3Q%|wYo|VGaqqiN{limB*qaqUw1JnNT1X=#tcvU%OBk6jI zT}P1=SQ7yh-w1gIf#{xNT(qqHetUztfd$p=?4-DNF1tXR=A)W2caTGHbK`Lz3yAexFDWJXe?O&6!g3{ z_;?IZVCGKZwWyzsKDY)8I^5?}VQIIl!-fE^@wFeAM z3)>!fX!^ORME0cfx?pU}hAdX3;I&h4$w2WqJcJPF|zYxQfgIjs%n*>O>;*Br20mdMM&s@ zyJ{5u@*@8=ko`JFFxnP)BLQ+WcyXWm((V(fmUf8GQ02gx6&zNpP}Tu4=tGdwjLYhk zE|_GsF1RfOoo6H)0b;%U!2Id=!(xI8!K1-7j>VX}kI93vk^(s_diB#e^J`(Q*kRyC z8c^+b#*^ualaeCIE{_8Fhaxa{5F;KVR$l5Yn}g@2Cyebt*gKCZE$gWqb}fseLaHNd z1=RYZhtkf(fiA~Q;Ad?RZ+1T8UPHeCa`Gm@4!W0Q%gWu41IvT0dh}KAn|F`DszDo_ z=QmEB`om9uY#E)A6Vb1J@^196{_Piw{vfAWCqg-}3E>DOA!!H>8Bl1hff;g!c1<|u z3-5Qv%*{#{6m@CVU?M3{lFO4ojB;XN!y7ac;phVJxIvTrUk7T(GnIB==x|J>Q)LAj zEP$!>#4&IRe#oP zAv<0hR~LeN!xC+(DfSRWDvX(*pZFgRIPcy~3Vh=t@5?{&Uo$O_He~prBVw$?g4wlv zZ^VY#sif3(k9@QwauSDkK!s7i|JUM=9qz(9m)H-kdJ_Ct!)Vdw-iuDYF3ARP*J!wsvQ_3~783W*EVHIuCh z-R7AMDwLtfq@U3}@)e<*A_tvNaRRfkP&_%orBVc0R6)7~mmI3)2kFXsZGsEp1^wM2 z%wQw=@Ep8Cm-J3M6C-&H$)H7a866jzRkOwuW_CU5C6uLe1Zil&37kvg+B6WxT zdxZspJo-gQ8KCKX{@{4ITV8c!OmKQtqc%8yP#IWZS;Wgx{aOU+h-ouBwbe7WhZV|@ zm%CH7RiU4%TND0_UNtvay=rcT=78U6PaR%aD?mX}-DX9<+Y(8-qEOcFh8b_C#TP*o zunMLl8JYyqd(Q(_`ewWKg)fmTk?io+-4f&nZgzeaqC4q(Hq3ZgKHUqszV8{kv`_}C z$v%RyQVZosenh_F{!MMD&=GbBmy*{vCxPRMrZ6(+LxlvhZFPlo0_uC5+a z_gL|HTYU*fSMvJW9rt7H(WY{^N> zQ_+|XcAOkjo$=l*J~yv3EJkt}m`XmVqLaUH(D^)>04xmeA?tB5{4V=%fAc@@FR>I~ za@dGJhn!Nb){sZ3Xam8~Rgz+;pasI`Ch&}l{9){b)x$#BUQaBg1mYQwy9^S{4-o?( zDS)BUJb(ITO^$Z2=RsJP)U`n3IG(XJMyvb(wTL!RiwH)2K|6VG2yFWatl$Vnb&?0>q9%_T9$V83v=5H~9dI+j5@y@!~gK zOsINacm59~-hqRBd1eUONU?f~tfit`6sQOb7{V-;z4Pkjg_6ZZ{-t3+do^8GPM*u} zhdp;4OWcjmD;|C=PZ$ip)vnvzFc>Lb7nSq}m6dcvSAAR9lXELws>JPcPR?y-D!nsh zs8^GtEtJ^KGF$m>tbO}dHfOWTc)fkcwd3<#e_lS>WHI7a1RaC4z}MtoMu4hu*rIeY z#U@Z>B^7-_12w$jCe8c0cILqJaw(FYO*pCnhE@UFW^lE>A=f`0h=0R`2;W~spumAVuhaB;$CD}CMxiS>2Q;vT$)YRf`8GXIPX=?=+;cz?Q_WY zkS_+#Zn*3I=e_sWxtE7!3w7JwE(?Z)iHmKP_-Nh%R(1Z%Uc&Gr<_+3tkJq<9IbHnM zfA=s!rRgVo7n3vxh6*J54>Jb^6bt^xHlUTC24!Y6Y70Cg4dQN(yCJnOgSjB>RNZl1 z3Dojbpd=i6Y;7T3peEY|!w52VK)%83*5GZzClR_-VG86<=W2_?b!d0eSeWi6 z0cU+jSZ-)eShfsV?$RX?txIIv#Ou6vLl!rpOb7CSzWe-JeD}c;bcVLpr!)+BIkD|J zsHIO3?S|}zyKb<&EyOv^xNtWvU=24#BW#ZZ7;BGl1M9nf^M9i>8JrKl{*9|-?Q0Xu z-)Ux^wo)tv7B?eT6@ucTv^*?J)Tk_!;Yu^Av-Q7|p=opZYeMA8sRjlZ zF!z)xq7oF|ljIj{}l;E#=!BteA35Y0!&rQ3Nc z+XUqw-Oi9CIUy;x!@J9BSPZ{u0=66H>}(gCY{$Y28;+1w4(zOKH?s!m6uW^UsZ?}6 z&;w&-vVnLC#RjfM9GcTO?WkOLHDW+@HKGH!y>Yo`oC5=&-mtp6O@joVw#RcV_aHJ- zbR=!cx$n4|z)|te*UpgD{00^Wj^Y=XL1`1kZls{kqA?VYeA7r_@D0@zF$1mc@C zxcP#C&GPWO?ih`;_fTT$XuepQn6&*#lm41>(AmmVCM@HOC)Z#}BSq9L2BX~QUQP0x zby<+Ko}peoXCrH%lk5y>RA%~N9?~{>JMi^Gz&z6ruKw%@I*)MICZ=cRVObk(Xsb$? z333uR#v7yAgHAol40U2?iYRX(C;@+@DTMwrHmz%e^w@T0m*A@4N^pY^)0GCDkg6_4 z&_Lg04vPkz@)qK=lLYxLYu%p9TNH=gj=<4TOb7q&8No3e$006r!0S_Y+|fPwlfdNh z_$>}>gE;(_7R73}LK!X@={}A6XUeLjnAkK54!LkJzT?)TdwkE~dw1PxzqmJthNi$j ztsq7Gh9(Ct`Z#GeG##Ya{S?_lMV|v%(9#7rm@-iNYmUkZ+dVs5c7dd6m$6Mz_rIL~ z&c1(o`i)Zy8h>&w3W|f&OBX%?{-2w(j|say%KaN4k-OXLp|B_-Nrq>dLI$O$$WhUC z-wyBO;D;b6d0N|`L^877GrPsM8`rEyk5vYcFGx7AzxqjmzvYbCVY8e#=D_EJI~2yO zN3;?K34s=n%`z6X?}%(+b=P6M)ib;NQh*h_+p8-ChErpDr0qI*>x|l}H9mclI#%7Od?X+A1r7%73cpJ(d#9#P zyE!Lc(F=DR^zWKoFB@>$q*>SKLOUh)g})Jl@!z|lRZjvuxZSxT{BgSJ!A$HS7< zIZ%6qC99}NifM9P%w5Qoyfh!UIk1fJi03x9Q=~L3RlV4hMmO^!NCKkwvxlq;hLkxZ z*X<+gy>Vx;U6!xMVuN0GdTnQWhFan`qA44ke*bx2yKdHW32vIGzYh6T&(J-0F9|!S zO*8O0b@*+kQGIla$0^{OM-Qk2_L|wQFqMvm4|QY6bPa>u201IrG;u70wS~=(-?X8g!eqDRSgD=eAE^ zH4#rN*&H)m4Zhp>$aOrd9F7~2Bpj?9dsN-@vRS(@V2$O65PGik-s#e*H7eMSJ1>Fj zMROhJ0}QS<>~VbXKi-WGHu)qg^*uXD3BSml0|VofnFmuzu?Hx!kBUYm$2CAwiD}#T ze>?MpWK9J%bY+2{6KH>sWo*!?%6}`1y2c>01!|BwywKp?$RP?`>ZMB(4FG!wV(cDd zSGCdQ^olu|KrB@mZ2So->>eM7A;&4oVYZqK`DGxtMQlL@?*EI)J`4FmmII~ue zJ>v>c`5Wk7aGvu_(B}$ayedf5?xM>*vot8LG<4-zAFv3p@L2;zM*Xf!{q_fyhv#W; zztu(W@xKIxmnGAV2z6M8v}Iu7}pr`O|ePr@Ru*F~DG{ix2X(Fqo(HrgVGyjT1!I*i zz;<)#wh)Zyz+-Z?xPcEks7(^^B~|8+C_Z+^>{?#v)*@_a*ldh6UT7Na z;zzwd7XH?Ro!>RS8%RERZLqV-Orr8R#XhCT<00%Eo_Pg=ytOk6G|OgTTzAmvj&?63 zIskDhcui1T9lV1W%vpotrb}8tV=x#X)H-StKXR#}vu5lMLdhh94XlvgH3STo1+}+;h)8=lhjb(Q%woK1#UX1k&hI{(Z8SE@Yq6ok4*m$SK4vv}RFS z_*P(X)z>pX*0Qb~0vC;yF6%<-0uaSFiNi`${^&RD8-Hzp-{GIQiOEA|@N?iz&=@25 zJ*Q*?6nToZ*n_UOg@*7iR%VK-!jzo_CI*kPYf{@{&c5{qQr4 zvp}89!p0q;UBTB`UGz5JrcjNW8l_JWi|F~Im~Elt!uv2Wc1S?@77AK&Sp(!a!5m7= zaaK2o8t1k_8)$(kVOh$@et07UosgFtCu@!3sDCm0K49y9Xqo`pVKrYK52xAA;EDhM zm4E!aj4%Y&Xb<=cg0+>NBDf@NbUz2nkW+LzJ0{{16!NGY7JbrhovxTUI&uDc%b!Wi z43c9sh(iKh9;rQOevQV%a&#NrX7dk!^E3H3zi%)ZY1H7i$%VI!Wag2P zeY-`;fEZYdUQrd@M$%>{v#KH09tl7@fn*4x$~bdq6@zZQ;#iM!{1t;s64}_UUD_9h zg)_KnD{_J4K#)#V4!Ty+hn8&OZ1#pAGc>}@3+!0uCgDM<08jUVo3lo&m8{UWz0bns z8|3&?*|4*!U*~nwD+<}WJ+U=tf$<9X$)c;IFAV%7+`-=nQtgk0(Y&k)!=9GI$Y`-P z^$!yO(Y%oiblLw~)9BGVk``tr)>IR)J|v2l?cW1?8zlnpn9-rWFO1@?a_c9Z5N4_i zs0pY77ot0F zfTUA0$g)VGBK1Yt=wM)C=Sk9}K$BAEU(GMj3HhON!9({8y8Wsg8mqi>W!rr0VP+U1 zVH6ZLMgewj-lqJ6U|H^#uxx89>07C8q;Tu5r)o?0BdJCvvx+iq%=c!CG4s2q~OfnsibyK~l*t5)M zm-8-n4KTz%Sg6|xhq{FP@Ily}d?QtW4Tg^ql5Ft(=DPgJ)s!Dg43kXitpEHCNn*Au zO&47zsgXbi4t>(6B+%9-hY?Ojry%@kZh)% zCqg#SEeO}Jso`Qf2`X{7X>nQ|J9rkrDxtcD_$ zROHZtYb4n(9kyIbBsZ3xp-n$)Rh!xO_*eNu3%XtU>D$64Rnb>kpl0`q>JHS>t`5{+ z&0*`W+|pgiVc!LQSL~@y*ud5Io~_jPzirMVDP}2L5l`JrSp!8Tpu&$+2G;$BEAN;( zwPw7*U5eg_lmHEs7-s^7R4ToX*C)9u(kOt9BU1?#Vo*s~Ai3l|^=D=AAI0pUTHNiF z8!}s0IBe+BU;Vc(|75T*e(A0KM9a*=IPlI*zmbKxPsutc(ndwra8E&|k)AmfbDR!( zE$onFd2W`k0nWHCX_6Fq383g3`#c=p6okK>xtEaH8#9i< znd&Js@7B5Rk$+iXfLf`~1qtbN;0>X*MtO42C|Msxo=}lkys~}XDj2Y9B;Qqif)&qG z*GMv5Ax{@%gF(C`2X=(j==l!2kBYIW(`VPg9J9~ABoM_#2Sjadm~olML!Q1iI!~|> z@*&kI?O5s3D83EJoVye+gjsZ@YnpOR#piK3>)0ly^_mdryuzb zhpj*kq%^@XHu3`lMvzTfFaDN854s^iPP23n;2FnCQKTq#vNkNL^Vg@;bUo9G#(ZKUyQ8pG|x-F{Cb1|dE0X&Y2@HLW8|8Q zjNC~|c8nr5RHS|^6M!WPGb9a_d0EQJ*)5){MR5UDP#$xWKqjX$%TuS})hl@#oFq+$ z*v|#!3n9+^8zk+l?epsW)fed1fkp0pvkyx4$bcslp<6v?&`nq8mlgu)n-WN5cU32{ z5~1p%p6n#`7uz2%4@_-qT95_6<-pJs6PRU^U z?4=@mrLnSSq!e}=+7(F8w$nKUsqDMxWKlG4&@Blv!qaD;<8?wb2n%_2vF9vaJ!GnD z6d7*KkO{66uWc5!2E@RIRaD?EXE0E;kcitw=lHa_oe04H17CjnoE)gwi4Ec%_oP{;PPW@n$R;9C5;52>)w?+(7(_R5^Yr7*gC5!8sIb53y5l zX8mGAfcX|H0}Zig!;K0~3 z!#-gNq6dPt9x4sBFNm@v>U^Am!R=_{e&gF~C1s${NX5+htvT%d0 z>9aEh_Id3XK5m`;we*oqrJec5J;ZO_O{DuV3vQS3kG+vU;rXDUa2#OIB}e$nGn{j z_B;Za6ZYQcW;omvjgr-_Gc#_>Up)TV+n!W|5tdq5yW(OXmA*=PJew5RbT!fOPve3C zQaD{u-W_5qZ4lrcwz*S>Ca__|7IHad)+~G z;OWFnFmNp=S#%m#!?-|J=favcNm>Vr)n_5>635cjPGeXQeb_eoIwWjeaK5p0#HmKq z=#JXi8ilTO6Z1Ov(rW`w_%{V@TXI;D;t}KCFWnZ>6110Y;YWvE;%BgP=<|vep}G~~ zcsEJT{3Ea|Dwdukqc>?3o7ijp@_dV>*&($)y6|HvT}J2m#<*{E=?_<97fI4&*4lt9 z`m{&>yfc#OIVS>QWg||wyxOB&N!^CsS-e^r+CjGht>}3-EVJ9~iJxD3*po)%D_mOn zNRLwUXpHzS=Y{Fq#bnFfffr3?griQdx`K~{wsBkejX~M$Vox>3avPjGM43vgf>xh| zLu6HZIjC=b(XDZnDQv9qotE@5FXrNZn-(pZsrIRuAGHKROM`AN-1Txhm>KoLp4*_K ze~@JlbF{Vnewu#=->NmyhIyxhTnuZXKdP_(jzl}~9u?@?jwxE&LCKORvW1G=DqYFb zZz`%QU}q_l9%NqvngzhGeu=NQ1U6mE7~{?|%#KbC#`oIt6C=z{e}bhjs=tfQ$mg_VG1l0QuUApWgeL0U*D=lKca5%z**YY6OtWlnhjl&+C&T z;+>oM>gs?!B8|Lc{)zA|`W#=4#6)*J59s*C^LUB!-rzx3o%T-7tTn*l+AVq**5Y!& zN4=LG-8bCs(nFv{nsW&&jRm}o@BI1rK3xD;tSIJ)>s7bIfw*){XK$Mi>)EY-y6rO7 z&VUi8R`SrTl)s7x>;0x8tBtU>o~ zE><3C6iq>ukOHAz qrEUkgB#rEi5BI|FXI7wO^a9F55=@Z8(@aSb@^aT?PMw|-S z!1jr2l~Gbq+^p*i)_-3cjXh+87$rZ4#tT3j4BQNuC4s!t>TTY)L-d4F=+8o$MJeB5 zIeeInvBPGE1?>CPhg0Fhci{Dt89saz*KFe=Ddbwtpj!thOzBj&K{V4JA&=S$I62}A zW;YLtsUfr66iuzA%V*>pU#x`oq>vF`%q*y)Tj-jI`(mx?0BMJvf)U?N@jhWbdsBo~ zl|E-680om$NL1i)PRz_0-$BSeZY90!n3=$39{dcVLpPT9EnCHD4{HyL@wEwzm_lZ} zxo&FU82|3OPj6nGN^Oh-8ivJ#VWjP6WllqL+od2lX6~drGR7sqg+k ztHQ&6{tMw9&(S7-`M7Xd9BUGLxT1&ln~uY75U>brG#~hg{4xZ;f>CRpw|14ph>lJ`Tb>AR9e-E zfF01P(q*rq?HUfY$=0=rr#@TOp=*V0+PI_rG&*DNCj#)EHJn7&@Vqr{!;;I8;+rAp z6Ld;u;90RK@$Gx}*B{`DW-pKlsuaaiBiOxOgd=|#^Z2KM(nLAUc0x=Y*v zIad{mGM^4wmkUv5sEMnlPCI6TYv-W2 zwH(+WhM`6}OZl8evZBZ040c>lLQt=Vl`YnUkKSSoOxTlk8<@wyI1>&PE^Jw9nga8V z!U`_g{?D_q{rw6*y&c^@htU$Dr)jVSQ?%~-xBIPre;wt96*r>x=y;@(n79! z7ri%V-EuU>I2gslxIfJfE8`1{49c{>zwUDe zhZby`kLQ!5y^@PwdF(Q97i|Ba89m-tui8|eoJIFCvi^=6cer=2@bnK2C=q!5Y8q*n zK^_~SlalIov=(XOxm#y#dKFIdz`BYj-n)4<4 ziK|w1CgR}YG*-_19&xe=x1MhK4gfFT>A+le0kK0SYJ?BA>B8XJ^rn93|-Cs__HnNnPh>6R0?O5YmWCEaiyP8IK$q*V3so z50Dxzq|0fPsLq}wy~M|)_7NvwAoJfE0UGQ&-D6>II3}Fy1p0z7)7*-au(hBwY7dzK z#C&T;(3v*!-Mj%)4j3jYEeB4EF+*juH7E<}_t6093Qn#`I=J<8In)wV^HE7ht(R3{ zzUnbQA`@L*3VmO%w34y%6a#GRRkidXCJW-K*$AWi>wO%^FY*_CIWd^bwy6EnX_YDU zn8QvxX2`@ae7c_ZP^x~qaIIG~l_>9m!CD9ZSL5a(vgM-CGzbS@2R)p1@)X5th`t_G z*7+af9pl+LBsYbDH#=z7xM1GAjI$LzEpzDe=C|EDDdk7;5+d$KMDcVK-{_Ry@_p{&?4%M&C?yFo)%eFhTwY|TVC z{#ZIHeE$2f>}OvWb=B-vRYIz3lWkWzL6&(kHYo#-L=r(SdWI%`&Ks%KU;+3q8K8;Nybr0g?7%5F-QOp)zWWW|?m3%B@0v)1@l@#`bn zxILU`R+(^{_aW~g&Qp3EW;QdRvQ0_x?zTo?|T%TGG+LA zs}{aySbsWbeZWhdPPT9`sT-2r^l3|tT%`BIxY#I-{)m&>UII^yL4)>Xq7k>kb@QOv z(D*mKZ+j+FT#!bzau%*tqIl5+Sk*U+d&A2YAN1;;(n@nu&)rf4OsexK++-dBH_ zU;p{^FvGO+v&)P2lj3(KFmlmo5~`(SAoO??XvDm_U7q>u_S*m}S`2dR_b64KCH)I~ zXq2h?(_LW1SUcnKtSG2W9i$I0dGh|A_iBCC&A?3kvIVQ9IsSvLo1KE9_hhM4&wdw%v&t!ZS;Oh`#r&64ig3lsH;FQcfA{hvAhr4~GQ;|F4=Pt2aX z!`6V-oex?{3^PZGQ`&jha&p+;m1G&s1=}fEB1IBlE`Thb&R~!YmmsC&Q1BpIM-X~W zk?esn(Mx~5X`x-QYXM|mx`}o|tL|%S;`dNdgV56UAYtn5eh>4!&mX325O&wzNn2}#XIu8`V@`(UO6-hyQk9G^k}Rx)yFr}+%5^r+ zhQK!J`agc-`HxdsvpO#DK}K%u4&KW(t1hVC7p`SpS==mYmm;l-J@ZRnJ$;KH@v6Dq zj&gzo%S`2K9X931j3DhzAp6}mCmZtChow#KFT75KfB0GW#_$^fb)1h%|Gwqxmp=II zhjx4SEm}T}dOGmaZX?Z1oSoVuPuOERm6)-?S&(woCN&{E ztv<%?5Ut|HzXur-T7}($hZenXW5&T=L-CGy8lSBwN;9pNVv_IUz*C8t+Nx5zGN@U! zcHWvmb&jAAgxeC8SX`AX%4e%loV10XD2J$dlJq8P9f)Tjv6sFUX5;dPWm$9^RM=qj z9NXs!^5A-^pmKIIye`(CkS2kB<2PaSla-BA-+w31H}yXpcCsa@J%x{id@gNGuF*$QZ%9A z7NKv{j7;X|BvCmut7_@no}1jp&7u#_IjO9H(gbzR{6u-P=;AB6&q+R*N!@1IC{L>E zDc!`(#{50I?y_mL%weY;Ghw{yIZuN#+%Qf3kVxIaKN8U+Oq4$k)+q2)nqac8@gzsV zGF;$mKglncDQj2G9oNNJ5L;9^Oau5NPY+yZtc7$$ol z!D##}g@w^gcKEQ%=lm1XQU@lmaNwwknZgLY1WH$MPgoaS!Pf+&$*Kbq*_-7WMI39h z{HXtaqQ<~{9jOUeBW|=;uCt~1cr^{#0UYoZ^W#$5e`lCl1fqY6Al(j}8WLwDcN#Ta zHbjv@D4GHmHgrmLMYyOe+5~LSHLwl0Pk7WHPmVb4b<-#?c#=5mA&Y@e$DbQAJ3LfDPy5-RE99$Of>?QZ6NU4peaNBJ7S{DaMz|UB`?F9xOlgzDlkE-Fj|6}8n`JOk-`)4~Uy@zpX&5-L2Xn;8 z%gv``pq`Qql90=iMZHq>UC(0A3t$7FGCY-TbdTqB&`A6m#lx#m>YK|t1@AlxsEjyu z(7nKsb`O{Wk>FLMs9v-yWW*^?aESdG-!=%D94L&;i%k$>-pKs>KgMkgGQcG!si%nS ze`l6UXN`bVNyz|Nhp5Oz*LYr>~`R%>|*taq65mQ0(m{_q>om0hkMPlX6Zd}qLC#MW4$})H}Lf_ z;>@MJ;n4xs!kHD2Wdt*8g?_$&?|*(^x)JTLQ5rKlQunmHD-Zk~&RiBPRaJ+c!^j z(&zY1K^VF#@xMw^AkA%)3yPpu`R|b>u`3n&$w{MVQgqSDZVjS)E@#P=#W|p<&`KV< zb~Qj##i1xBW(Bd>IoV-nFBuSCqcayAHbPv>0 zR&s93?c{Ih#()7b?1~LW$e4cQmrk3f8`d;y)(4&-BTkGuaA2onx6ztrGbP(Zkr*m+ z6>tAM*fgvKKjUJ+=hDlzpai$X6j*{EOZu^Po9#(X*G=G53Qc5@jSF-6K6A!tFB?SxY!_5z=GG)LR47LlXaUgL&cBtT^@X~Sl~M#%7KNd zm^rJ)T!Oy2ItQhS9}hxVgRbfeq)bs5IOu90;{tp8}11@+L$lKe_stY&mn{@c#mjCJW)UIFUpcZB)0zFxUBJV|=d zt3jvH)U_}gc3H8(kaU0OQ=uBerCP;F=Mg8HJ0g>Z#8e0U&G(y!#5keS>FuFw6W&F8 zlI)jC@AkydXE#`q-SUi(0+0Qn>SVtsaJrD(0Gi%|UWJl)4xDVG4|-M6t^9jZbyL8u zkZyW#Ch{5Ca;}*?c)*~%dHknz)jW7kT>NmAX^_-m7c*%_3)U@^Y%@hRVH(O+?+*GI zsbY81U@E%k)$_0O2grH*Re`@8918*Pvc0_y%mZK?>oG07_BYo*FvL7wu+x7`svI~s z>824VE>JSqs;#FYF&cJMxj*!w`$jjtgq2s{!a=E?H>GYixFKsfTCcTUd4jVfea?|t zYrWJr1Gdl83$-4B8!KG~!czit=3|T73!%=3#f(SrU^(5iy2o>)c*T8Ima=T_LtW1~ zbk6*Pa~s87z~&k4qQ)L?$app&LbaX}sx^}djt`E~w{;+CKW_M~j{*k5tNV}`HlsI{K3 z->EiW^r|{;1TQg$)F^Omcb|pz8`wQ2Rkm^0 z24vCeA%$BVD_;%D*XQ`n(pq|e584{=b+IApG}azx9c7Eh2KP8l9OUO>FA$4v7HwX9 zLZU7OJq9hQ48s49IIZI4(AlB6p=09DlYD*gAE@LZK!^(zC_;_=RyulDqYD|*KkkKTCXSJ=M^=g62Gb- zU39Qrx-O(aqRtNKfX&VlzfP|rpL&qDt`T+2RwF90QV3Zz@y1FQYzsT%(o+QD2|Gj& zgKLy0+_RL4%1yAd)eE&w3~jjm0L9Eie|r0e2TXlsCV+Bao)$Bn^&Mdk-33f7z3hrv z34)4QkJt^8aw>jtR!9lhu(dNz`WMg-1KUE@&0sjh%pSZ^Yli4By4RK;GR0s2-4eq* z=ev62K604ZJmImj@9j%$F6i~;5LYddzTV4(0Iu^l|n8Y31aHmr0I$<%bt?>gf{Juu~4JdXD}`LpICv!}EOaEgE);3Ow#L z#Hm<(kyp#g_wArd{fC{}_$OHSvrLj9EcGwsl=|OWlgikAqu+9{ZtJnc7^z zh2P~i;-oLxD-FkFM>RTtRYAFPn|)F0;5tdOr8qs)Au%@3lYitB-N|f79F{--TKdUv z2NNakP=W|j=_R=Mu9rPpr;V_Nj9{U^jO0q^g;)GSs z2faE(X(8%j_61;Xu%aV8wHLtnb+BeQFe~uuxu#KJhn<7W#CdxqPlJ;{rWmvEdZiU% zxc!akcgd`C5ckDVjg`TeV)B*VFR|jdzQMpuV#wdnoAoX+vw4M0OEb6z)3W^9_G4tj zJ7Ze*8kv@4O17OMiBRq?$4JRZpH4}0ATvz} zZvc5w2*b96K2Z)BkW>MXS}Ico>Nr6f3uh(d;OU8YVWrY_%b{ZGQM}=^6#!!%D&u74 zP5&34LJzS4K4-rDom*s!1H-4t2tIo#St><#QIT2&s#A>(gVacBSzWq{vu?W0tuq*R zc8){(OFXBG?v>Pnst-cM4z9`MKw8R2S?P0zB~QhfOWPqlU%DV&00oTxwW7G7yTM4*lcS>R=CLy?TdkeS(%6Bwtj#QqG5^2Vv&AJz zdfFK_r_o!D=H>C83-P}ddg`97GdsEa=yF*WjWZf1F-q=Z<{3z{^h%*t$63J157Sc4EDV9zVfumLnZ$@&Yf#_Vk9jLHj+-)VZtovHHI)o{ zeevVKGmsf!pvJG<7XE(QqHbv$H`%XpS*4(qj%Kxooe*TvsPa+b*&deVsZ|~4RB~cv z>t~&UoRT7T`SK^rK2D%Mjwi$ZPw0#6qrN$!jllo5Q+5Fe!FDRQE-(D*|7^3iM(ZSi zWMiYD$Cr5^?O)&Zzow-~uY<;cv*_puDXVk zI44z*>X+(wfYfkHCA&dnW5j6_uw$U}s25DYoqctQXO?G%f41@}8I;2JVTyHHNQSVT zH7r>Joah?Gr><8Pw}E7LxAT6FQdqoTR0KjGIY7pOUjoFm8hH_9fvA&QFx9bYT=f?Q zLuAIpTjBILBWwPx9>-^JT)TctRljj68BY$pOfr-4)CgsbkNuvJ2MaWcJ#xsY%nqpw zKqYYu_Jfp}wAO!*ye{AreTxNr>bTfWl12v{2(6(Fr{c){&<=ViIEr`K^ORqc5KH{^ z4O&1|WYD#hr5<$Ccb4W{<(3JmW|gzCUb|hY)Rc#6aV3l~&`f$-#_a)_4Ub2-M=Dtfs4eNyaf|M9P~@Nru30Ib`D zk%4jC*YLgXMfEK(Ol8Ykb2pJZ2M+4i8^QGmC4|GHv2 z>o9?c36$g&qT{4=Zj#IK)id0dNw;~=aXq2VNPKD=@9w`3W%tm?MJojYf>cCc6 zJ-1AFNOelo58NC)>KX_n%e@D#m%Mij4~zB(KECv>}98a&f-(y4U#9E zo66(L7Vm>z_ZDbX#h%wiL!55UlMn(-7Omsp^$y8_ao25(hS_2mSvX&fHe>WljMI1* z{o8Fy1B}i@{_j>&!VE?ZY+M?QU{puRswq-IMP^Wyuq||G)&`ck$FBsum}XIhvWH~x zHp_wLrG0MJ+#bIs1%?uuMM;78pfIe_9b^@O76#?d3PORJ85d05VWmq-m*~~8)cVk1 zHmyF!PG?_Gz7QI8$5H+j6+4rpN0xmYMLk_MHTz{N5b3})4md15_0!)xc|sXLG~_mP zfebr2Y(g=lFpS-uj-zB7DH2UZra9jyt9ZFW^-=!|!fMY!soILZs-+LovbCD!y;-j% zBd>Pa;+fjwsiamq@Z@ABwX%(i^<+`JqY>-)H(6R0lrM8O@DBK5}keTk6$c*wn4Jm?-;v7q~i~&<`?j>tR*`{S3YKMD`W8- zGSM(v`T92^j(lZ~VGddIk@qH&>cCNyN~1ZXkdi_Eel8WcmV=U-NOB28xR^BA7*si1 zqsWLTRqmw+*&12h+!p>r_d0)2R0XLY4O9zaSQc-(L_<|7#0qof=g?K`RXmOIfbszp zHrWIo7Gq%4%N*yYm4{!~9ivHR`i?D2|L$#1rbKlgr0WM6nEj#6BH%h)1JN?nJ=G}S zgFbmhjdcEyY>RS0sUQw@M_4GaauLrsqd5hTjMJO-PB8;X%$i3{@H6_fvIrN>q#pP z9AwQhnu)ekvP6m`P?4xHrPfQL4v@`DPSHojC*9OXl~5{nRGCBfva5k5$8Mu<*|O%< zBx8$_9=}iX#>DT}e_U&tBl(UP4l@~*n6+?@pU=8UYJrgHHn-iihTbe+1DnaYzO9~l z^R%iH0S6bi0cplf)*(AEp6N%7spYU&8q5Dmg4*Bq^3qaRsz&ir1(hxtZpEy-o~01t z!om&^WQR?BK34Gb^Pvn#qi9ons>tIN&9jTiU_9htImUmzt$WtIu^1;IX4>C&c|P(r zENFgHcziR-7_W@Of#atTbstm4QA)|6AZR}oS7#p!YGE4@`QlOaJX7i8cf%1ohd0#UnTr$IJ)Ip;mZiYxgYM>fTcBb;d zLfx}&&aXCz+YNqn?Gv(7npzGKdB#n0kxV&@6k73pq`?!D|-U@}+ZDWn3$U0m3k86%Q< z*Rzm)j-LrxtEqykAy-3CjTy69tAb$r#=1sjYM>blAnV3X7=dQ``F|?=%DcFjdC1us zbcqjxJ_-+gT*R(1cY|TD5Gb}YBW8Lw50nj?rXOEsh@5@*#NQQ>iW%gFQ84NpB|A-# zlOPkf=%C9UsLoqEWAtdWT`9%EN>*lXzt zH^3?E1rn-;1YoVDROJP53P z{Be&M(^wyIKjo;A1qCZ{y2QshwH}4;n5@b!ns*`~Uss=(rEH+F-G~#0uhJ=FnRLJ4ELLihbWn^vYSm9&TThGV|-CsiP3mM1_V zU*M9w6LL!DIA7wQku=Nph@NoN=tplC-#lxXBrIA{eN%yGl_X;a_w`B%+^jy2A=;` z)EL_Bi7Kue-MZ*}3je+9*()sx{9`FpXLLxq=%bPXx}LV$$WJ^F%&;+Wf7WX8k8A!a z-?W^^VOLLP3UqLlU9b$fFVa|p>^RO|`cUBUkaDU>xk|p1^s=vrkML_b``vQ`?t50w zY2Y4-(7Im{H-&0dZQN9OJ!xP!h32w8x$F6dCRNR{-!BT@lkDS@cEE5DnE}IeqM6|H6i~uwZlaDwzW=?c zq`Nur++-%*ttu!DWXqDI!%kVs4xNjG^76w@TW5ChYnR5cRy2z`gX35^bdk@n-7=mS zgJ}u?VREZ#Vd_11!wmGp+TGX4RtI*4i;d=#3`zzBO}jxy6LQE!#5EMAqfZ?bAxl`ccU#Vqx9)=D1ob=3Js17q!qHkURRweHSjw>U^%h=;|JO<*$x zw7kR(V<=c~12eQ_1}{2c>eV`|=VHcBZ4_hT+l{6B*(;FqNh2?r-z)7AwoA(+jw%n3 ztwDI^A;%^t7=J*RVPpJ0Y;l$uHX@q8x$plNEXjX${?|Qn&Ven-r$(0KCMCO0k*ic> zgQyg=-{a?(NTKN0vw=G->4I?LMc|ISv9zDBRi(0oR3AbtVFQU)}vps!O~E$%wdO z))m!B(jhG154ZxgilO&P7wn~m)f)VWTf;@=mp$%V+|!lEIdMVQ=L_klmS@T^gV#%I zGwB0Mp~whcRDE*SY*W+oj_g{SF;x2eTqH)x4}TU>z=sMU+=Uu(!V>mW&-0;`F30G0 z59?-8zH#svBI%8NS@U9||JMe8YcM_1EGdtC=D?;W)<~jxgp$3W$mdjKqgc<$)JZ3W ztYNhvyD@Gol`BpX{JBn^1#96_ekrKiV$tDlPpBV69c0;vlm2rjy_ZI1WsL%F$NVm& zA;cCs?|HO$P=XMQ-q`E9?a}_gbF;*GBjQP7=&P*MIkl zQ#7VzL{qH9Ck_zi2-u2RQfp^ zw+QUxNl#+T#~mYUpJZf5ruY2Gzs9uE=pC^mW@PXBm)#cr8RN*SWSP<{UdbMJq?Hk= z!ds94+RH9st>PUZBY%~+=vhP^w`)N+eQ(hh8%}+Hb7R?5N);VA=xe4_vCe-JyMT|S z^J;W#kWfL-#B|rQFKmOTm{sCe6`@W8$6^%^PC<`L$Q}$n2LZT{JdMWCj|Vy*yy<)VvY;Oq$7+_0afs0YQU}4Oc zFoFd&ulj?Esj#{^aFK`^R<}(q?GRBh6vSg93#Q5?BLSpFxy21>9x*Ww3!Lp{Z7hY- zXl!H5XxsEKFYZgyw|&#_sDClL2AoR#IpTzbesQd-pg2|+76q}@c%ng43d-B3B{30s zb4Q#oyL%wGHCRW_G)fm5@7&C{9Uz#F|0oV#{4&HR}UU02__N zR!>6G!PscmW=gh+A~AZWL#|#k%eGG^g zlAVnq8V7bmD~u3PK*?a8x)-(z!8oMM-RE&p+0IgL3|S)vZi`U$A@*ndX3;I*5vN?% z0NEjW7`!%M#CMbXTKA#g$9`C7G!)!MuY&B@YO>NfQGQp9+J-gC6Ye$wlqG<9X)eqh z*a9<*05h#%?(aTf8vx_v=d_BfWtL6rz~S`_BX}iIvMm(ZOhumlPAvq+{&ZJbw+!nT zp9Ge>^|D8V$sQNT9e#{GvlDRiEk}_puwuT?eK(t%O`{!V!UP9^oa7X6dG|{5cn}ak z#jqr4b--aEs((Rj>7du3TRaC(XcYEx@GO668p&t|?UH9nU)v{WNm0T!%XNRHy=6Coh2Xw_M*n@9FJFt}lLW3>x8| zC6_^0qh8Ve>2VeX;iPvG6#GhgG1@BI+fj z;RPNi$sNy?yz|TBSh2D$?>g@707$PTM}40MBO8xBEC!6MMF6pr4>A^L9?m`sXB&)$ zGB&1~?0#pA#!(}qv7eGbMOTi#p68+0S+dsapk$9s$4ZBZ()+=MjpuZUhb8OeM)ZZ$kbQ~%h?wKW2t1=7#tyYr zj#;y2OleQXfhQR=doufkNBuR*F1j3CkU`hGpbrV#`p8j$GP_9KqhAWybKYSH4yz2j zC7l#BU0g8FU`(6@%0ncJnK5zT5ZN&!V{(9!l~AORicDr*BRQ~K83-@sm%Bp!Fl^aq z6!#WwgF=dUPKL0U9+u$ssK68`+gRh9oCeXS=M&ar~dV`Wu0HHU8dti(j|lxLs5*i zr-O-~zJ;SaaYt5e!OW0cKe76-=^E#)65TPglEHlB!sY5t8kqkC7fIG~b(Rtr3Z zna9v42Nd;hI}7a|dOukOb-}lVO{(MH)WSaQ71bcSCZLwJK18dk@*fFK@k)^n&(*5* zEiu%JEn9Uhqgf;Tb0OC(ukW42@JtC8llVSMZ!wRHahw!tp-Ptd*M=p}j~`r0Ax9iI zjYMm-h&e;aPEq7I6}e&gpzEKn%eusa(p$nEGUOh}RHpi&DM^|+;)MG%O^Pmh*l7?b zk!qEz=U;~{zYO-;fL5uNRm>W3+8>(78wSmlBYx;{3_BfBzHo1&J0pZ_pJPxt`N2R;*}Jv@5cdTA7Z0zr^pHA}Z`Nay_tQ;X$v| znM0g7&Qxmkn3+CPmt&&O6B#)3*BP0q9`If!XSlN|j_X(;ks8jvA=zHvyJlON2@g!{Pq=h6&QI{o~{eWy2bkK((k!{53 zI5`U!k?ll{yFCL0IQDVt=di0*at}zl5(RLgd~s>`NuM}Q(R`4v)XEAV5e^CA&qz|B zT1Sm|O_E-4M;@INp1K@T6)=wYBVz zmc5oHH`(DMv)9sL9n*%5oxlE#!7=@>#-Gh6kC-J&I&c>41|w?O7nJOCiu6;F>q6=R zM*OeR$JyQV1>s2Wh*KLEi#HCC21yOAEB#8MhQ87YS=UYOhi6y1WFuv48y6BP@p`tb z(&g~%Bw%^1qO0GFb<-%yJd?Rdh*3kA3p!?_s8bjH;p63>dUpBUSh`IxD9x6Am`bk; zft(18@8N>4h2O>nS}th^aIe>JlR`$Eus9xnGyK|Wg1=iLsrBhyQb!ttQt50Of;ml! zt3)eWNqxMMDj{8dnZgdDQ5>FK;TO;86(&h>lN9?e6vPVf7X#v2b_czmlMS0-4$CV$ z|Hy_dQz>P4-2D&&tGO|z^<^o$o)aODH{uj0XcX5APjDW(NAc7s=Yc_FJ6l-GtSiRh z>|j#Ie8mSl7Jt=LY0iv9+qOCPT<;3>|8D~MOOUe!hWcalUavS7?trk>qlKobJ@NuOegN=kY-JUtD{)JJzS{C-u!5!k@XDg40W&bKo&z5&+$`Tc``PTn_9|+|0c8fIi6LOE7G2_oL_fp)bLxXF$H6T@NaVeC-kQKlmkf6KVs-VLr@)bhCllGAB8=OK`*?d~IrCq*HcO8w zFN-Q&a5afSCOr|Auw3m6ubkb%?UbNp9%O4|d4gsjdBXewE5Fjrp7)!qwY8qKz8^-f zuH!Bsi;h12^&$goQl5cjfw_m zjdIuAa!?XODwW>36_AXw$}K~PN!himcIk7vP;zL&fN0O`EQnm(4mu7Zgyq7AQlt=U zmo`fWT{Vh9X)P<4g=B#mh1UPb`BbOD_W@ya-`N>j(X^#E=H{wd;}8Ngoz z9^7x5%FU#B(SheCGX&VTr6nxm75~SLbXLgk_lk1adA_T8?UHTmG@yJ+ksJ4f1tJxG zE0tQ+Y2FS|4_)+?Ca6==s>&Ca`IavGcolU^2vaY+H(aY~0gpS5b(Q>4efY~-6(05T zUkLAbj<&e_!N-4zV_7gMTM8i)D>KE96Qj_)9nyTIjws8NH_X@i(H&Rx{L){Msv|Yx037-^n;re(I6-}p_O1TkW)1B z#aqpMjSX#_Va*wf2*mI)C{_48rz<70!J&qUw!*PAe$h-07a}2;(^O5%^ zlFDqSz;PiD6-;CCT}a6wOq>hJ##}5`Dpz7vb+hOqtsZpigbI@sUKzh#p>8Gm8d}^O z!yprGkU@0{e#{eeJ9lss7hNU!%QgckD%6oEu-Xaj?1c7>IV3FyzOh9!{>1!yC%Pm5 z$(>5moT_(Zacw3w?2vA`Q@5yre^Ic@zg?QJ{257eZWGyH&?kEMmVsiTUaWupYaze< zttlv8&%t!u$RkuCj5QW7vjUMLTUY3aHT5n0E;scdQ6&cz1Ex}(IH~7v+0et=KdIg) zf0Ox%aOkJhFHJL@UI&Ha&W+~snM$NBUV6%wf^Xi|QMN$Rjr_i_Pn6BBr|oSO-acGh z_~}-*+Te4XjhMqBJr2AOjyFoFTs2)bLXj6#WP+e#R%OI4abZ9kH-Xd6>XIfzfJiMr zeo2f%EAMkX3pI%u!gkiJh)Z;%^HmSvPUp38pKxmB*b><}G;TEk>u2D`76$ci5=>ic zl@0)HP!)Xu!Yh~-nTB*hL1`g)ZZz2uuSL($s@k~OAsXcY@?Zh#IkdCt0zMO7q6gr6 zb9rmTBf;?;RD3}HJwN=SxX8Uhe8sujIUiCYu-r^D#(Gw zB7TEb^|@&6lB%Gb`FDKF=Jw641%{YnmX1%+V!#=DA~yc?rZn(zqA`?UO_B6%kH+ky zQrCD^qKf|qO-SvFVdgh-hz&ovJe3)bU4HM3!5&Zg%j4(}YnfWxqw2tu| zSv;E|7>eTUk0=wKg37hNS$D;U*xS7Kh#K$oLE+BpKgo z(@!5Nmj8poU{+LTe^Nox9oVc?8<~}2N>)ITeN^N%euD(|heFD9E@Wm%H{HPQp}Y9m zJgsV26305=`^@{kaIf2EiUDD}bp4D>VCO(*HiMktMFDwXOhg)-8FuP}ZR93ZIc!Q? zQAOFLLWB_t%$X7!;DMPv`DOSI=l{C_Gt-BDG?#QR3*S2M;(N8x!s{s|>!HYJROE>O ztb#?Rsh)^j)^+|hSeOhub&30dSGPf21MEUgV$^o5bIWksFn?HnKkSe%%wKdfAKO6g zlqyxxO0JT7(pdKdPTKrig1mW8!b^c|LbIUOXCH5s*QbHIol~)G0eP%AJ#4}cQ)%#4utyww#?Ql)2TFt@*+g=(K)L!*G$Ug6-?>QA{ zYVJv~@*rRLDfmRt{w*0vd)rLkQ*O%%3)Q{!E|(MdP_`GgWm`Rld+%L$CW?& zS;0aBaz3h5)sp;o203SpkW)d)$|-UXHdI+HE;aP}8I9umVl3l_r6B}hYPpr%=o!1_ zMu%t=`#nk_yK0krjp(-UDOBp=^ntX;az zyF}U<+D~tmwuNjU*IBLnTGoBBRy7d(K=P1t(jBj_MB||ik73`4R*i=aA0Oi7FDkMmMLc5kx-DC z;J{s79!0yfMskY8MXD1(6jZC)K~^eLSZfav4b!%hu9sG>7nhX7`ee)@L3yV^NYowmB z<#~~G19LsAfz&&RT?tv72{ZTmSa*$K*`H_8!?dg`tF~hFx;pN}Em--J9i~cc4(k!u z7=g8fl7RquJ{8#un+m`?G=={uf-fgbak=bB+1`$omqw zrmlQ{uXsapFk~Z`+!6{Br9rGLh7qyY&eD0Wv%Xo{H~r6h^JbpRyy?ue(>IgpOgjrM zxQnRZ0vb?RWEVj!Dw~S91B!~`27+iUt$}fE^qJAeVUF5$YC`9iY5b5z;}A zd#V-;t8FlG;P~6v6+28e_n#m9p(!vUQT=WLspl4$aozn|@`;(%xj`{kDRP;LL8usb z6?IL#^K&b`5(GJ-E>E4I;2cFSvXvP$miR+}S@EGeDm0y++Z&NAOAx%@47C<`Qn$-1 zYsPAi`j8|++T^s!dNt+(KnGi2y|2 zn{D7TvqTR6U=5Fp9wZAxOb}a<)V-VJyYLRzNi)!tQ_Oyf?4x2Tgdh3#kqZ7MNxsJe zZJesxCsA<~sv*1O$+8UkfI2atdP<>p!gG!CwSv$1c#IX5-SnY3-GRB%Jk>tmrsz)j zPAOg)RGg8XBx?h*=#PaTfhh3}NcZZ6XkkHk7QIRtKP6>~)#0`FXrFS<;fQ$3_t|=x z-5goT%%Qfge|58X7(XdDenr-BGf^%)pXHjNZ5zcTQ)Ckr(?M4VLGvEDFqVn#Ds`+{>?CVZUu*0-;b z_1y5`!n>in&G4~~m5NCa7#G#6l@DOQpgeq|XSOO!^EtDTzg3+dj?Qu|{W((} zGWwU^W(P=M5^@j{M(L)0`jSiDCXoEFdfP>^;Y$M&VAL5=(YKRgpsHaT z6>~DOTey$E+`Y;3n)HNX(5*JI8>AG{rk+q#(EIpJo;^}X9o6>p_d-fWDt$rNF3k39 zlK-4UB90!lqT1Crg}tbMia-&eo6@!JDGKKQj$FL%s0{=Va66-hhcfZnSa ztQ`G;-B0=8VbNbKSJEtt(-R8J&)HUUUtM%kl;%a*=La|SPKiBEkX6d^+3J3ct)A$cF#2FMt;8cMt&9f$oC6= zo8svBhMsLgxM|HhF3>i8$rmm6dn~nXTsBw>vDp#s#CD1SyOu)5-1DuISJ8c;5Vr8F zpi9K}m~ua0UX^W|+9|H|8KMUjIg)B#g)l(?>ywBxnzNet$UOdd*f}nvZW??i&l|>J z=VaHYIYqIuXI(lpUXbRoFswmtT%i{EAO}efl%mzs_ht{d<%pJhfU3z&pH5(TgvD*D zq(ZzatVelob}C)&2Q=|k?#|T|gd2O<|Dt6N8FUf3El&!*1&ACw>!V#LdIYCY55{C_ zIH7352it%0yq{^Aqg}J2l5Cqq4w${r`4p2wkt`}^351-0U=UvB4n^7AEIL;d&s!qb zEelHw=p&!VPlAFt@WnLCFaj~;_Wqz-is}}W4uUe9I1!i!b*SBn@zPXDrMN)6ZR&2P zyzCAcmN5;0^_#g3i_5B_T#Tvzt>uD;Wergpxg&ZHzeu@V^7yDRdMy?Ez4DirzkYu{ zrn@F7kpQz>(4kl#asRc|uT;UB0=?gjo{e*yJQdgv6zc=*sI#1a;<_u#w6ue72Ad4a z?eE^*L$X~sh+k`F@yaNslp@7w@nVB-fQSZ!%tBT~g1;&$^G{N?DDOv9z-~{Ls4J+1 zpBQk}BYXN;poVFR-b)hswe%fGKIsaAtVvlqZM+J(nxK(=M0?dY%`2n zdmV{$`$27GLUI(Mp&xFb{@af-JuGRcT-N<%;m69N)53re2xdO5UR@l4JAKFwl%&*a z^B|n!6fU4q28;tRM(M@zmpB1qV#b}+f@j_5MeV6$kZ^#RS04*c3#&A^BnGi8$mq`} z2d3+=G$>bu*_Dvursxu+25$8|N}6S965U<}s$rqt)@P(hQA;ntTk*y=vS7%qkbhNt zHyE$2@@P};@d4-T3Q1C;cvhDZ_Mg3rr;HJH98hG%o{WBKEZE|Nq6xqH&Y>Yo7NRdn z;Dv>ss25Ty|8Y}!02D4z@}$jwV7gvir5JEZ-tFiEX(NP2@5BjL7ft?6#8{$nt{Z10 zEYY}l$pzmIRjFbJQxURSl|>`-L<^}3>Ua^$7?Z+qz2`{ZD6u;r8>D5LW|awJF=glC4_(jqoQLU1QZdCD)| z0SXc}j71Vf$gki2z3z`c``yp~@Ehq;idjsN*wLMcafs7$!pMY{%-FwLPAe{JJ6I-) zX4wi~U8Aa8)hw%~`@}*Esvd7OIk8Hw(`7Pt}i4!I!ELi*I8#6Q8}B%`?u@>ms2> zO}RE8?gjjeLDfS+C$xHP3;&p+pT9n&pWh{@rQ=>G5IYlS_tg>)f_m!n?0=c*<#5d`l(0c@_)wR+U z;sgPntfk|5U7iKNoZO+Uoa*dp#WXs$kNB`vaNw|x(iJDGG4XbZxJzWR8t)cQ{xez4 z&1$&r);58Bm=PIoTPbD>{;@H8d<%6B_bHD`GppnFlcF5y9~`n!gIPiv+e^r;#9gSL65Xqb_N1* z>h|z8Of#@6Es~$}S|J~DYaw^N8srV$2|%V(26{utih$Z`u~8+{`awe&D@LJu80v@v z76&6M!g3NDm8&t;A8+)khw8z5wyG1^kCtK7aR2=TG_0K9HaY#};Xm9hvy+~wx*e8` zrQUzNZcukPFYkbqg6&>;f;!R#S`NLuc3AX$LEo3v3+siuBaiTp@Gpv&g*6DXc}aqY z$};Iu|BKM>Bsn4|QPe>eUWa0btSLG{vRV}{!Mj?NZOYXEunNsyi6bX3hvLZ!F6=$U zj;8}3sNWw;`hW}1TrBAW&9Xy&C=RLX4O}P4dA?3-kaulYERj1B2{w(Ct#GlaGY2** zi!A>@W16gHr2bWJB7RZ1}sgxgKU;I>~XeO; zJba{EB&z0B&s#Ys%^~WLwnBl^eB%H}uv-pm9aEMW<<$J} z?b;&xkh8Pe>>v8458qf?a56m9EZ>xLUz6c!pBH+aq)f76pIkT!1GL#AqAvx_1b zIE&R8jS09Wqupuv}+njC^5AUDw}S8hiv6A98*lu`_o1|jXK0ZlX;wnFc-8nP6i476x{e%o z*R6}lq^^juMHVOmrn&2?WbJ;bbU%OP9Q+A?7T&g2{>w2UZ{!5!&|`Mw9pL@@Yc}ev zp~^iLSzzoWN2Dtj@6%rJEmrDcMKwXIl$l{eZpOiE6{19ajxB=oqrNUSUbN?#+t~G( znz27~D-(h1zS?a+{K}jET#;mq;T8D+%R25;_DJC%U7Mk;pizBT*DcKVA973AR)kdW z69_Pa(LK@>QKjQFAJ{Qx575VB$AQ=P^_MJ@FuCkCxGwhDa{K$0$->1mE$znyi)WA!!II>N1QQV>U;Jpu==i`{egfleEZ+F_K4e?*im}2s4nqR#5E=lCJ z{B~hz>@o9?(IIc|52J@c)zy zxnZWwUBQMBU56sUySMxeT8{|m-;B1>>H%P{U02}W8 z@MmxCvjhpt)7|A&Hn(JY(fn(W43Nq1iD-&8N*k1a69o;gzI-6PliKO-!=M+-Kq7NmvUie6CU z(=Dx%?U|d+Yn*eIUKP|yI~~jxh_l&pSls9AkAly){bMZo-7dVP$dcckz^kXr#CN4N zJOeLKlJcDQrl94bb>y0VSI}E;8B-YgJ<}uwQ=k?N8upPKkL465VHQDwd#!V!V>3=3 zS--~z&cur9uW!U#mWMo0L&4^BBQT6r^T1cF_iXg-q}vsE zU563an%Az*Z&uWV}gzE3W?aEihMv&Cx%#k5hRg^Gy}Spjr9 z6(N;AgKoqBV&pwG_#`*!88N7BPD_uHV-ZBXs; zz3zKh3mM*UHAiCXvmWk7<&5E8@plt^Q@4N8WSzCI9oaBjJs7mgt5k(4oFmWUF{;d% zpBu%Zjb-9_JO$Z%WXivR+`+TB)3~6)K9_g&hiA zO;8?aD10hzmVNXhgwK!>6~bo5ZCu~OvMkY+!`X`KWO+lGI%S%iHE z$8v?;T3DN97H!D>^*`6XU!%MyGzo;@l^+=o=AC*d9nF-Nnaho#X<@)(3yL+4Lot^V1qZ7wp%jk9muEj5uO9Sfirk=@+F97fW9no|f*MO(FYTcv@;Q zL&gb;fl`qpRLr%=1ouXDqxw+>QepsZ@AN1O#P#mk)0(Ga^3k1HNcA*)ezo^)IhGCW zni1z+91*YjEV5MyQmJ%%ILuv`V}pf9NXdZarb)FznW6oH#>PGT*x&~qK>F|Bqtu-h z@8kpNH{>D2iUyQ*z}s%GA#xf#hJo{RBv@_1ZQ4@a+Ppw$0@KSEHytJ`xfvxFc6=cL zb3{gFGR177NFo-&^(fK#!A<51-fg5aRIgqV^{Mn@4P@Ij@-Sm7U3+KN@^NGv+zsRDG0oeU zCROBf0q#a7?hbnRn;Zip$0EppEjuw|e%piJXX?LYLeK}<`|prrF6;o`GQ&g@#hj7AW3;pUE<$P=W6m~=z<{G0_OksLT+3u)GT(=#@SV`zDcJ< zT@BGQ*m8w31yF@Zi6^%qPE#4wBEBi_q&Eb1hW02?$-A9y59|^Yh^~4R1KWSA@HQ`x zm!a)|-jb9ktGLbQ8f-lnzq`YumS^liukMp2aVv{-0$5nBkvSYj|HNs&o$%a>D+euw z5nZ-kjYY8$b8HbpIU-bD&k-ff$`LunAdLQ`pSWrZB%Ziw98Pe8#6+Lw=s=mt+K9Wv zZ<7QUwl)Q3)+UW&woxP*W}sPR^hcnzyjhkJm^I^+I(f5FHTiO z>nivS5O-U{d?rr-(Y+7n41J}Sfp(W<)w~=@jtJv=SSXFDFnTqn&*<8KoD8bwBeFz5 zL1jqyci^1G1?86J=E={)7#;g>c9dWK@nN0ic9P3FA}kbR8~G(O@5}bhM6#dHNT>3y z63L|W>Nc{Ge@(PqvM;zuvJqr<`y#VI%5h8dac7sxn+D7dG>m4yfC^XC8)KAHmR#1N9(P(uwjk}>qzjE zgKUc(NP5zon4Pf!CN?x-N76qsyJVJBV=ik~SZKv?v!`F4Co7IzB*2>U-iRf>m*;KK zHouI5uv?g&UPkhjg`(XM(?6wdlATsp$-3!9-@E`VuFXL<=QKflP`7d!B<015l7XBI zD-v7j)9QF&b&d}ym=ZVU?Zc77^QMzaeTkO|F@ndi&OTE*Upkv+#+`_ z9P`aKvqdQsvzdZ1i>U;qGK}dqMVCppOIAebecF|~=%hJ2eLsslB;2B{mv@sk-YRl8 z)IkQthVkTxB^!ElOcQ4RjB~G;(D9JJ{imd261igbSkF+*Ns1h!VvroN9W1yGb)fn} zTSA*|0pPIPeEo$%)meDpCcM_yCxk}>;i_#3n93d1l%YAr5;uE zkS0yi3B08k7u8*pUXx-OsqTVzUPO-M!>O>JmjwI3Mv|}w`Mu|l2!MJrmCl}y@zw;% z8R!SUtX$9=!F?x~<4BmH4$_2Bb8d|(or92d1Vz$u2t5lH3;Qk$A{LNb-}^>x&9 zCcsSpUHJmC*@Xd9Y6h5H6axXH9aK!dN3QQlkV$O-fA+llJ$GI2i|u|%%3ZKes8^p> zm-E4-HOMjIkFG4Nv->^2pc{lc0?IuwsR226aCz;#^t!P^#int$qY)>lOnm<2WQAqg z%QF?1VM%Vuq7Q1j$udEO_)w6poyKhdV^O@iK|T<=O_L)*>1PbTIK_@w8#k_ivDSw( z&and~b^g8JXMK%?dEYqtS#%E2P36&-8RP!*MVOA-6(#<9HI^G?YmsXV_7vf)UY!zE z9D!*H*gh+;jY-#L(U>h4_d=ES7PL4c9STi}LK;MDg)JNjBCa?Y)q`>%VAyeDWsNjp z&GGLfKfNZ_g;zo>HL)imdwFNQKMbyyr%BVpj_~h6Wf>kP3ECXGQnqQ}Y=w$VojI@( zJ5&;8Ov$&**=0$MHRz_RjCA}Mw8Aj^)qtz8G}7GzX;i&-KWyY3C*4x0+|X-p`k+9& zUVT?pHaAPOKky^pau1mPks=WFQ->J=QNkfD>@IZMN01!ls`M(n@VXX*r-79J!{3zu z!ez{Z%%T?=VrBNd;I@v>sITzo7wsA~_VKayDuaj$@%Tf`uYMQv_8))t>ji&!r1@&s zcRTQ=Ls1vswv7C`>fcIOqP%f0Yz@I#NVkY|NJ~9SzA_L7AL3xXD*MWNJ<2OeOw#UA zE{(jcC~%Jxbxqb!(W~#ce-r}#uYq&Yt0Lq|=;KREAd44JyLHNu7n8#W!W=4#HK#0BM^C!EE<7`_te7s(dl1hYfx9wdBC?;E9s`wM4hVQ=o`W5K6c^5yVqse_ z7M#_)8>IBEz^V&(zIM+Tf6h6Z%cGqL3F}?61I1_Bq2vSY!D4>BR`RQ%U0eGrz(yH#jlk@YPB z%^fUH1cnQ&qp26`Hft(0I)m~q3`MIjL9)ynC+i;m1)}TsS_V7TSWk>iO-u*C(AR#GQ*6!&C4Ce=FfeIMh=(NV~30ury+9PP3F?O-Y zDqVQb@T%FKVI9SsqR4S7rkaOa%9~Y%a}T&*Q)Gvihivp55Ny%tNdbSeW=Nb15>@@m zYA83%GBEZH4l-;d6E^sGR`yC-83%_BlbFY9U+x zXt}k>Wkuw=?+PR|j7)NPR8b%Q-ITBQEI9l9JM(qG^8tLQNy;K-JwMT1ci*#D(dJVG zb!NCGP2VdfGVfjhI3CS!Wj zg(<+~BkqalR9^@g_Z6{iAi#;jc>`{(XLb-+SprS`FArJ0r7YQ(#OcU0gOQ$MQYf++ za)==jxLt;|&A|CIS$Bk76?b~+qxJ+J0pV6CNJHM;HB7zxNd^gibgdq(<3i|TJDeW* z9^-*rMDgI|bx&cz#!4K|m*5 zbKbjPE9gg9kdcQ@!@X#B?SDRRvLWBe{qJ9st=#H=TzF->-^>)|Q4ACWWrDAYoP>(q znx(T-0+NNgK5~Q{6jso?#kveSE(j|Wp+51Z4-zV@_KpkMttkXoMjxu7{knFEmED7@*?}%zPmGulIdSaB2K`>04xr3p=a(%m9#0F~D=TlZv^n z#Pa2nU$@4jo*GafuKs!^zfjT_vDY7{!9X*9sUQ>T;fuA4qmIFDZvmYdkgBPcInPvZ zk93T*6wY*MIn9vNoOANylTBE;{XbX#Mpn2mRb*Dn1$^-tG*< za7KsOY-9T11eCC&!%DDZ!oLpu_MBys371um3;VGw=_eQ$siynH27#&*+AZ#@l>6j5 zT;4kaC8G|otq>Zu6DM3;zVM?ZmZFDGW*)ilf|o_waGPIc;3e@=PfUz1wPyS4o^4Bf@?sR+GWE#1la#ZMEKTH%xa3Ppki9RJ_G{ULw@hQm#*DTD(4i7 z^C;)6O&>o|Ye%+?zO_4@dyakdtxYF2e>_+IvI!Bh-#gbz%DII_T-c^Go8jj)#hjo> zEfrG-)#%FvP0lQWSEJ0UD&I)g7Pfe`9W zX&bMP=urFvqie9ifRrxrkaTx%_9iBiUc-!MbvkDZ*g2+LZ)3MW^L~8qxqmfTo1nOv zual!+8fychnj>bb3lsygLiJQk4ey9x(+kHmpGG7GEaWXXr5;e}1Mp;1^kL?(s$YrW zSL9SprCZ5`S5MCWQ{nujp6Nj5wU9TU>LtZKKq(i{4Wc7MQ>pa z2A)t1sFH>Fvz9h~0s2b|-4AIF3R}qy>6IWzq5IyS8bF9>spld3xBwf~_-D!V0Y?P= zN-iddGvsVFH`dy5z+Fq;_f9AJ{5F4r}0DaRgL4 z00EnBad4WXL0%zF5OmU*Pp?-O)8!$+1J@^ORO8PpJjl^6lipR{=0V2L3~c!Yja+Yz zeyMGJ+SZJ$**OP~0Jj&`WmlpbVyo&c(@(x6AvcyRnshBvlcr5JuykOeEb@3@#g}s; zv!lY!R&bBni4!i0Cwli=#!^^97VWfAlokqhj=Hylb@)GcjQ9j>d6Z6gQcvyjOJXOc zd;VaLrJZnD<^h&B|1hjE434PL{Y+8cn)x|QgY}9HfsUTFo(_v7ijZHw{d?UXfA+hd z|KT@}C~VB8tp5Wif*V}b-=6;65R=QXBB^^f$+xxx?80$0puHWjvMr|=$T!+Y#r*ZA zvd$onTSn(Idr7MhbIp3imxFsE(tw^Rf%?-2x1>A#cF3+r*2q4debBF8xi9K=_{Yjk zFD)n6BpLLz=}pmjbd_vS(dF4qX90OdsZ2jpuWt3Kk+llfle2R-O78Q!f=@$O4deO_ z!JjQA&=Oxy`}|ZO+t7d=Q?I@uT5hRh>9SUZMPL)tN%ZOsA+4eJzEU7AjyMTBH{e@r znO3X--`VrloI%C9NQaEC4UZa|aDsu2eLA8UI}EHWaVG@-;=g<>p~1y5b>aFO79e%d z6~Z-4A35y>0J{B^r653ej82(yD!ABRSHl~cjaiNDKO9FSaRSmfE_Z)(bk*yY5-coA zCs;)u&ogS^8e-b15JfCip_keOd4SEb74mqV4s{FAc`a0&d2xK)S8k8iWTYMO5TVVaanEaB7J-_^z)L8RNT-Ze1H1kW&Q_NY4)S*N+>^RNv=h>a z>gfh~sVc`iSq6DXcrj6%sBI2u4nZ|_T-)XX$lq#^Be!r|5ahpxw?lbX){IW&C0RSY zi0tGqB3pvs-W*W{RDPizP-<{$@Ny4bqV|{y31z#5-NJrVze?9Tt9u%1Q?>xBjG^GxiD6A0B8%!L(aUI8rUvuie4LVYfd*1=I{0X ztM0R}b_OmE{+llI<<3C8x?MN`TS5Jv*rHWKa%mOcFnljb?*Y0g>TdV~@?51xPf=YN!&Aavqe-@ZoHyKq({2tSQj9_^qQ zC>`F4)fXwE)gI*_6=F()M0$!Gs4R~aRe%Sa7?9(UM;V204`}nlF;{XG>@eq1V;W{i*$qwE??LvEuwr{A>1_i(fZ^WPX{jgH*dPNUocKq=90(ES?nG6qVntoNYE^3BLaNKN9%_`gY05ns)Y~+%!R>xq-M4z9 z;o0ulqpS)!pzfrHd%$nD1A=y9YLuDC+9mD?g}OrJeyjQWd~`-ZZ{#+>ox>qFv{Boqg7`KF8r=@ZwC!c0yGmGo zbqn)>136Zd&(9Bc?25plVB(G$);rk2q~ZPOXKIr%iaz31yV^l;j1r*7p zVj6i>!fVn83M`?n3R)JnFz{2lkdzBzMJV%=COM|KKkbmPC!#{UHlSoWD%EAtpN2O@ zAJCkU_VBk)#jBsIjtJ5u@w{6e-4Fw8^=jbTVM|z_|L_He;A6Nm2mWTqhm{G8I_~}R zjzAM;YF}AjM$%mvGl$GDQ$jIC6e*x$Y7P6@y(9yQfqDsM>KE}=&Z*{QOEa|f5ZTM| z(4Cl)Memajxgp{IhRIhwdXxueB)a&NbfeWwbS!BmEwC$jr+Nj3jvhxgxTxub9 z+{3Tnm-rvm#6vAnhm(qvSw@+Y`C;k98K?ha-TOQFCWGP@=(dQ&S&QhnZbS-dmPan7 zk|<_9Mb=U=pU6)WNV1XlM1V91s`qS*%nUPR-l7F@w3T?;$Z#0_r*-Af>+Dv<_s?$s zV(FZ)_yCInYw6qa{Xl6ID@szvic06kOV+=%O}iH=fSXiZ5t-UMvvz3{W*E<*ZZXyc z>*9GJR{)Mf54@81r!5kdD2}MQAT3IFUsw+uqws#(@z49Cx(*zM;j!6v(9vk)#y$*I zd;r(A#WS0}by{h%C7ad-T_PJ@*f}XNvm2Qd1BN7xifK|Gnt3t6__Go867G8*Q!MqY zrO!tlO074iyNq0vY=eY^5^?p5laH zD&tgOPnZcPiofW#kwR|aUKgH@PMg81l43w7wG1PBD9LNw8Z)@CsMBj)Kx6r`VM{@i zCu>vQ;^)EES`U8>18P9RTn)@#?Xr&0203Z~ul7DAE)cAlqN@__@oA9fNa7{9b+F6> z*WD%l4Z?PM32$vcCyncIhs<&I<83svIdyzpaieDgsg#}bc_;w=T*I#`T0CD1I)@9V zvJOIg_DFIn8U;FydzXI#eWQW@pjzD5{O4tBl z@A*HJ{{GlM7tiFq=acwNe#W0DWtlxan|D2^#ShA%;L1v-jaLf^GRvO-c=mm0x0`oc zj;^X+eb%qn5Fu#tO9sKq4n>{H;UIZ5h+{;{qkbHH@aEmz9p5pnhh90{^AmEyg%{4Z z%~m*y4#QM+POdPzC~UJNccydLR|7;VlJ`ndHDOpg{oZp~!@ctBp||+Jw2A=6>>W5soF&`uvxT& ziYZhizy|d)(QWx%r7jo3E;*t@eo2#u+&VmTnD20lkHH#UaYO>7ier#wIl)%N5lA6q zNNUDx1}CnRI0MVVqKp+lZ1<^3o^<$}*Fwn6q@6 zAeq;tTElp=?zm=jvP7L5Z2>+^V>^i`lK zMM1?18iH`sF9w|TLc@U!*+XuP@}AH%ULLPict#q}YgDgc@KY9D7MGl$Dj40`pTeQkx@A+L%Z9fLu<%gIUSreH_nLx5|#n6 zb(~plciL%A6VHU-M_=D=8Ny|WeYVr>pk;yb#h5yjD9I6lDo~5EkEGJDJo-Bhz zl}2?N4kIKw-zbyQd_GG1`P*l-&a zmu2wbB{XTj}z9cQf|9WAZ(I-|&kWCfXs)SgGXoM&d~%EJfEgEi5)TSDF^qEyXY=ky-1qz?cR-D5K^=CHo87$e`V3-OGV8Jn zZI&X~JwBM+A3HmLO0gCztFwT@baSAw!11a_L3oL{Rk)wNAE8$xH(^@X#qa~#m5MfA zr#wZJN*4-HJ)%un6O>2qWloYU5UuOy_eS6ySY?t8w7V71=WrntqQF6Fo6dlc6%)hx z6gLPp%-bCEtc$B+-dj6(UOkG-Q|n=6jrIdGU6u(d#D{{$ZIT)uDcgZDe8>3w<^+t1 zubrK=`gPOP_Jh;ZtK^&uPi79kwAp8qr&qHp;gil@E3mZZVTMS3ZcCt=j=;jT8W-2iRGRYcn@{K~KAE29 zx_>_~;EhE0BhAIaxpk5rk~L%acVqOuSFTI{ep6Wt=~I`Y&qpUIA1Hd1dUb|qr`JcG zm5?KG`;`x0IrPTzR|efuzxBlrO6T8}*V*{ey;By!BmUhk(EFyJM~7S^OTIDhzpCQa61Wjg#Sjwn^N z%bL8?1v>N=4`~|YCrN&IHu7GeQ&=3)DlF5ki9`o9!d>er$b%OU2h{Hm;<~Ou222(^B?VK?8Ct6{dL`l>ftXMLoF3sU3Ru%QD?P5ODI%Y8+opEmd1sATV0nmQ9#Z6UDyB!dhRKGE?`p`{O{Kq}Pm1>dnQuL) zLF+0sJD7t(5W^+6__0xaBu8T82ErK51DL3)WNkc5x@lCS4a841ih4z(x)_3bSu^$r zCc#Rsji%Ab+0=^$MGafgI4L&&gh7{uf zTeJy+X4&O=7_m0#a@-89rPFv=9`jJRDhvVyS@c6s{L{-liYFHa54mBe5gN?eJ{A8$ zzocPIhU+?&k_$w}9p+8+dvmAzi`D-DMt%%pm zfE#?&{jEkG5|=*aY-tB#Q(XG$%77p@UQ)X}J4`-V|^ow(I@=4K`#_6dudl#xHrh*~|sF)9ft6_&SvM_i)lz+D^$kN5 z1pv87GN{nujv>~FG|M)I?bVF)Di|KOjl*oiz&N%EC--sUuPPVZx2!L6*#HSk#n2YX zzUTVqUQxCO>aIYs*#Pi#LM_SZkSx0FE90RYVK#m?!UE<7v)i2hfg3FNufN%9$v(%$ zXK~#{J7M9fTgTT!T6f`;<)X_>4+LWF%+eJCM|Y;SlIYG)%Y^heU?s#X?<6ILfb{Ch zpiX&yxUQeSN}D*fiKpA`ec#V`t6^z-J>-z&IO>DOs*_sCBX(VJr^Fh|`XQ-8bBN z-n5)wxM0SGoe!4um1_S1RXTA08urHxo7$)aR}Dn%=Is~=pLy60kvLI5_!%iac)IOlM1=T@6n5+%02M&gIO?Zsjti|=Id@q z^PsdOUShl%zecZ4QI&ht&KzoripnyRW^#!>{F7i%_4vX>(5|^AYY&!-= zOg45MJ^}wxPqZ?N|J~1oD(#vTm1LU>=k*^j!%05HK;B9g6|-A&d`jL6kfy@x5?~&E z72OnVJZEH`OOPz{KBhj&Jm4QwH2$pV-{PtF_3FKnyoeTMs^qA<@ze);^$u+V&v|Tf z4nV-DBWdqNPJS*`_l-$I(I&jKhdrN64!AI0E}7xwB*j4XUJVt4!nDZTgfDnmKxd#1 z%UBJGD0l$rO~n!S+>?}dG?iL$~Hq&3dL-uU=(7ETGFt@QJ*Ad z{R+rOVo3a4POeM)$n}t$Q;bE7Su>m+VKo~ouIMnk*@+)Gp<}`a?|nVNk~oUX`pAXX z)GRbps4Lv(i8;h50RU`b1v4_>RQilk)q2540^CM!QvM(9wAx4qHCX5@p!yB9dQFK zr>@7;m=kCxoXR^wf=$lI?eE^*L$a;0(77(QNv)ZKQbsYQ6e*@+kY+nxa#egsk^6Ei z1<7}FcS_rMy+EOGiq{g_6n#&2cXnUoN5D@PE840-QuohDvg{;@=Vf^`MHg%8<(>4A z`4#jbVQyqo^bH^5(KYGE!Zu~w{Kst_XA!yK$;|p$`!FU9jl7v1Pg6e~TxPi|#Kl!{ zVN|i~A;o_4y7V76m3w$4K5JD+_ydA#(~s~A1$W3UfQ{pzF!J%UAKMHV_AZ=qhJwy!+WMl=kkx@W|R9ayaHZtR@t*;BC~KJ8Kh!z zRmr@12$Nt4rCD}OTI1U+Yf`U**zJ}ethTug%%!*t-Y44yG*4-~O5PU8gT>Z05g9aI zT^I2My~Oj55W*wEV*kuARPwC|>G#Z{cl);iqd)$>MEZbqL$Ain@n%`Jw86Voh+>|& z#n>$C4P5gAmaBn88qjkr6XgQ=6ZY5+@!%vSepeC@cpKR?vUx`devff!zC7wEsaGrx z+$Xy;t6f+Vf%QUo6$8YWHx(;tguE#{v)cRdUw#w59qWu-w!DWE3@7{|?#2nr%|&KWO4b-jw@B()E$;Pmyn$}g1luT#cZHR0u__R+r!((ZzF4%1VK}D zd1Qy;DwNA*Jf9PgMOO%y^TuZedGx<+`?G#@%eWk{KWl4KiObSvrqt?ZJ?k0TXgN^z z0VHOXMdh<3q4$Ey*Jba&2`;^3@Lc66JodVeJKyBWoB$oRQ$g zkSOy>Xcb*AJfMm5E%%7$T_9UDtAX6m2^h2qU{9G~@Uo+!H{gl)3uM1u|h7OTewAJW7Ie;VEtovv+Dw(wG;TF658 z9{vU5xEa(@crZc6VX)w2aHu@RocWe>5SIXr3%kC0Gpr?0%o>WUq+(!=4mxU&L`yJ> zC37rN5OxooO#rgHcl&P)J+;2fvi8GegEK7E9Ghi#4cwFQyj;-^1_+MD`J@eKoxqI_ z>7aGS^x7OzquO!y;As!srr~}(e>|-_hh67nFD3+@Ui&w>X|_4>t#4l=>s@%Z*==S4 zc2Gze9oUwjJe8@R|BfnL+c5Wv9@%PS8`NZcT?6d9h?Gh|hVFrJ@Umr<^ox*zI zc8GLrCl96fWjzt)9{a@|aJ5yqE}|#G*w2pVw)3-iMdE8dt9VynU%}BxvJ*EW4T$mB z!ARnUn{R)A@8zi`+$?$5cQr}n7IJmr6qRx_loe7;9!0XLm=@*duq$2o)#mSQdf}KR z?mPPLwvcMsI>@}QFj$Um_ zXGuy-4OsaN*rGuhcqH*&=UX1Wa*i&QF850e=<-C0Z|9aT+Y1V(%sC@a#10B8yVDbf zK7FGv&;*p)SJs!2bZ*m+3kTp2nZc@rV!$^ppki*#Y4IBhN0tNRTI!Uq^EU7RelBbG zTPIl(yd$6>9IHHr+!6!Q>CzebOsQfAlQknpRPHzAhU^N-!fdTRVA=D#HAavb!nviY zI#ptT!|D0Q1~M~OX9PsXemEmeA((E0%{zrh){>pvVB^BGQ?(gviYW%dxcMN!=$#;F zr~8QRU{JEGQH>K1s>m8OWRdC-3!tGm=(bNH2s2{683@rLcPcWpWAql=;>K~QYNJ;@ zq&&fHa~8-@JqUGt%7Uw+rZ^xh!N-J}#-EqIMYeE5jSCwa7Qvt_QXoS6fgu)T*jSDd z5~#$932MmVk7>GC+?+*M$uN|sYgO+C?*2gq8Ye?0B*}!V14pl&9$SDhV&!p14nCT- zXsFIINBfydePqeo#yr_9dVk>l7j@godJx8JHYmd<0i!_+5alB#DneGPZu;Dx-2|-3 z_^y3NIn8TgS>+?+1qEn0sa0vn|XYTh*gR-Likr_0qt|kVY73;OhkQtjrH&59R zhTfpILZRtXl1N>!HMPyHJmMx8LmhFEL zt2zveqx|q{|LxiXutCrYg&W}IE_Y9a%(i~-dqU7`!8yw)?J!cnS%*^?3dYzsfpJ9B zV@tv1$PS!wL(_k^%&)QJA9UGdrb@F}r-)(-D3VLXpvptNa38-rvK@l3^-!LVaobNq zi?!$F^+HT*LRn{A+Te%-I8c@c_2W||m2|tLF1sYI^jSWsF?y+>fbId~o4g@*U!eB~>4_OD6ZHwj# z)Lj5kDb`%1LIi&N2LEZJJ#tEbV<(UCV8-!W2(1df1sAW#;XWf2Yf$zE#8lRg${SPp?j!B z8&s@|G<-xEiFiot!xOAt4oC1=V9^mD9|LaW5j*(y{O7ix`kH{&J}>k-Npa!qg)%cx zsz;6_UgG%8!4X3+R%DFm%Atqs$gmc%EB&wiZ;T~{!G-q$SW*~J zUpGnFDm?0c39LcdR1ij{5A)l=5a{rdUVU(OF|QT&7g5x3g|dwxLpNMReQ=D^WYA~a z??==t+K3L-k#~AkXp)(oUPjSRRy$(n5RVmPBMgw8e}VPeBEOfDVIxUpyI)OEQ*@Oq zR|LE`@?vdMG;Z5In6(y^ygC#+WSdpRk&6TmW+ey~2d7Cs_{~=8H=j>zoUd2kn)6UP z2t~qW^r4`$^twolvTl{7_%3sbEW1OVwDpHIrn8)0XV+a%B){E}VyO$_vT->UrH?A` zG|&qv3s0I{9D(TuK)$}*!_n`&EpV|Bt+sUK&^DaVI^l@_zkcIc7v_eryLMtC7J#%E+t_TzPVsvU@Qr8E}ZJbk|1}} zr!#m5)G#NwH;W++w~2wxXpHV>XtA<+)h{YQajIC+D!eP`f!y5G$y?pGdK-4ETa(5cvru@TOP^+Si6o(aXVzgK_NwD5`B@clPPCAZkV z>rS5Js@XEBj$%$xK|ZgREb&MVs#|M@i;8Zmuzc!4afN0j(??+J1h+hMBR5Iz`QGPu%8&mjWkK`5WWB!f zs|8c`dM^%+6NypLr% z-IEEIF1%P~Nt!c8gWHst3}ei9K`|m-GOz8G1VPF7Vxvr1{0(xw8aKZnhadjI9Br&1 zSuwef+)^GABu!pUA@@Q*oj$Kxi&rzW^`Q5V?~hmTUq|^#V?dL2(_sg4tlh$nrq^c{ z{N-6sSm@!`(mk`%q zJ!P|VT%llgvlBnCL%~XXj#}CC>bsT%| zcpd2yoFr=@LAE?(-*m@~A=g2&{*pdA>#$+tHb^eZQhIuKzwZQ-ORU|w@6Tj_+y1l* zPgp5tF7XLNAGd@fv@f6}S0d$+NBp%7ifPJmi)E z#X^oVV55zg4JdeIxl9M6b?2~a?8Z&-)1!g^T#S;KzWtH5RF3ZIdHGpcyRbnxKIpI* zPFuiYam%=Te9&HS*OCF|;$kjhns@ zsBKi|NKjWbR#YLbQY>L0r6rzMs={3&Oxp)vZ4{)b#$L zKynxm-55x2AYUwzKU^#WS=rD4i3E`Urr!NBzuR+swruTsw!;s=iN877dDq=bnpZ{J0u1VfQe31~`vLHF>t8|8c&)n~O|*4bs*6x=vVBNam5qVNpf;pUii-P&n#11o~ex?pc5NW^^`<0>nXC9im8$<6lFt@ z81fX71H0&jB9ynQoRK5hqS>MucRFR;vBG`4$8rZdR=V$9<5_O|Z~~U=uAW$`S}%kh zn{mr`r}O}%BJ`0R%n9uoX_7KCJR6em9{?Nbn#k<%MtOY5X8z?MN3V%k$Iy1*7^fxH zjvRa;??B_L&w9EfrX^ufd)D)y|D@KT7(*+lPu`g|Zln8XP;4CLNA1V!fZaY=iMUMs z`|pB(V;OOAS?`1;hH_Zj7_wUfHA0F_kl5BITknN@JjjA@g|v{nUJddNNVZ4OuxzMR z#!}@5d2vLnC>_d}kC2LxA_Z()2JD%!PS61L%FaTPjd)>qRh1%L>)6wr&P}A@@P|DgZX1n@u55J^whiG;&*84=T!P-(t59p8fW3bayU8T z!?FV>eddObd4IoG^)1r^`h)EKcgQgpUO?Y6^Fx~`<~&8tf+Cp2m?C?~ubnQFwELxN zvu0$_{ZX+|dNl@GbxrCd2}nsb$di=Iy>Es#$dQqdc#Geoe4yx2rb^Pn@FLD&a2bifGPJlD$NuqtEJJKjVjeh#ntG6E zAy(8WE(UEol&!LPNzEBr>_*7OE4E>g6SOA$ZT`xKGfdEW|AV%JWEZzwT-S{{k)vin z{E%WG@p2EAEMj0LGwiGwMjlgb^=iFWzgH@KN{w^ao*5f`t3U^MZwq~g@O8?GGZABRq=dG%fi0%Lb+qP*!g!|3UQ>S&-n{Tpp-@vHN% zy<*98%0duobRA&NCD%XKAN>Yhl^}6yss!0VZ~I&f(6#w23T%*9$p#f2z}Hu*stW3O zu>*7}u^F^R9Uw3mPLzTg^Fy}nnQ?B#>1}r1m+_}>{^`QMn9R(?o?m`UY9t>g=zR-5b zlE~vr5}nC1dp)*L46Fz?P%$6TO_2wM*QMFK4PQMcxjZjBx{?$?47ggbNYv~OB_m<= zin9(({Cqm5I6?gB-8rnqvSUAa?G2O7x%_7F7o?G!YvaQE*xhE>Xrq`Gid>^&j0Hj~ zql;mude9An`-5(qwI{`cZry_Ayg|2WRkr|tt|x<={cL zRla!d#_5~k%xx*Q!4&a9x6gd>uFrh2hs`QHb5NKn9dx@-g2t6{&&k!lq~} zG{TDgTKYgFem*0u59y|B_{jl-ZbQ;zk%MlT{A0pFx3!X9xG#-K)IQaya)KEbtNE0+ zPs3x=3qP4*skO+(1i7wp1y;I|GxQ^6MkB=wR+6+Esz=Ve*iNq)SCcb5CN^PWxbZmt z{$<@Kvn_>bp4?1wVFSY^RujA)wi@&1-&DT;0P;hEot7TOJ_I&`g1rMLU10~szAL}_xyNu! zMt9*4WVs86`!dY{u$f|#D6*c48CI}rX0rLINZC^ftF#je(7uDxrv?9?y)S`lGQIPk zSA0YA#E^?X@(!p#gaL8n3IpPx)9$pLnVoif{M&82-P!I+ckA@nZkx__X4=66ZxIw! zKm#a;ARq!BsGQ<`fr2t39vq^hI1VTxD*k`JB#b2H)x41KiFS9kBzcbq%=3MJ&+qvi z-|u|+PFaid92*8no;yia!eGccJa;@>UBC_l3z?-?ddh$Ns(F_g7iYtPmmHR6xp>|! zKg=M&8bf~(SUsg;YL2v6SR1f~E*2hB*9Mf+8Nga(IAb^x4`HR{!s;oxyffiBQ!`z3 z>T7|;!Wyrl(5>R?Db3+YV)TgdSaW!~UkdLES*l17Tnn__+QboSLzrWGwH4G}Ir>TW zC+0OS4x49Usb{$)yB!t}0ihmQa!5{ywwrDZN(3s6GaiMcgN&s?(a(zYGi|d;v}zzu zR%P5oVfSQnT)a?T2n$v~tPDyTu$}=!$2xU`B3;_=vf8IbUNu4IVY}zuswuHv`P;hT zY<3kpR&A#eC#8)=^VP^)s@T% zqQ!Hi;gyO`c@;_b(;io+x?c{31j9~GP!G`<38CN`TQ?4k1_Y+CurQOOIK*y8EP1k; zw+FY2jF6dqZry&e%z+^TWVgdoRjCxSjv`4^%vwlcHhE$gEGI+H5Vph(f?Zf%^m-N{{sbufO;q}=uly&O=D0TL>CT8d+1_l4c#ui z{uUm)7g;OnQ9>o7dKo$BWjJG~AK1*V@HV_FlHjX79$4$XY)en+uw!-ZG-%AwTn9NTnr09tRGuUNAw*Ns2i}k;7C>syhmA zSIkP64k~-Yo1>d~Js#Vj5NelionTQ|u|KkpHHi+$GrSjtHAg4$Z>mb^9s!KJpLe_6I`+C4I-(7ReZ2m9_vVl%A;*)R*p{p|=q$PTgx>-;toBcMv0GER`i+^SO? zxNIQX1ZwLjCW#`6R7^8ec2}xxro)DD76#WnR51)Qvwou~en4BdeWDBG$OA zbWf+(xB)LqRfMfWDi&vn3qHO!CKgAs@xdd%NpUk;jvv-+JV(~NG?rt#iRIW#F;HHx zaj0a!McxycJ9RU^AUt8}X^CEc7eo6Pufs}S{!Yd{hxmQirsrAZb zopd$b0jy7bz>3^Ay@dhg_!g!&yc{%bN+t8f=^#C&tri`EV)E*M#UV*PAf0#FqY^kg z9tT%UhUjikdvLQF1%~E%bq9kA{SkGC8aa{EL9eV-a)GC9Q-7hxl?~;7VXSi54OI>p zGPGd1Qy;plaKo!~a-F(Dm>AxxiVHX$RIS|Wi_!tHVr^o0z7$(;PfJn)p4BBeQj~pd zmqz#IX>=Kn{OHH^x6C(7T-GWF_F`DJR(fIhPTByWt^OUVIL}O?9dLc_?hDL- z)u>>dKl_}xcK3%ZR&qjhh29R^^Ex#6tYot3xB%$d@Tr$0*W)A4Y#(jCWKGbrZ=w)8 zbnKCR8j!WbjP4kib?d)d(@BcvMi2k#8s4++iS+<$w{i&MY0_@%Cg^N z!P!AEB^234#pLn63TRU|K%;7`xFu+nsyX_&Bw6+F)kV=IfqKSnIF|1B=f8Y(W>zy~ zj5qTixLov2_4{-g)Wa3K;`K?qjUGMppFjR+_v~~~fSD)0;k8P&g@#))-ogV_0r9+5 zHqv#l+E|zZ`dOcb!E)qpIRTmq|MR~tG}rrnx>Uu1<1hf7VG+Bp#mVQnXM_OZAYCuP zrqe=xEJ&yBAl0JN;7c?|CjfKhIbIJvbiPNh6jb(l$r9gFio=n1Uv@?cosomWx;xpS zGkE0x{n68CUjF6OD@`QTfz8Wq6G-G!477e_Q8D{x718Pb_m$mXC@O+>d++wDmDP*d zL_LC(fIEfZbtefPK zT*VFy3yUPRX3u6ts1Ym@=UJOc@py95WKOCnrh+1SkWDMe=OWz_j&jH-dAVikEkCR* zO`6mSq0Sxt>!&um9M$aCY<8`8*(FB>%}W0rs-o9YnB}VYAiOi%t(SiwVBge|z%B@& zK9ubMzvqt24yI&cu@4e*VY4Yd@ae)#?-c)%z-R3WID9pun<>kS0Eb`CdMUS=a@eHN zhgHA-r_t=T9e5wavfIWU>uSFu35NU|VUcQt#%z>x)-!Hk=Sw0Rt+Khi;YZ0SXSop<@4qOzw;}QNK(1jLymZq>``W?eNDjf%|&}lMIi} zZFf)!2Z9EJG7$)b(=GA=myI5!pmSo(a&^%O0?df-(Nv070%y^h39Z6Lb)$M6Z=Sf; zzd`ZP4YurX`Bw48=d_{N<6*J<2t9R_^+CZCUO)Wq3$M+J;ae?oyeCfwj`wzt6;a&~ zu&WHZue7@xWB%P*JI&_z+50mmn>g-G&7s17Hm+W}8{$W#=>^M7_L?lUofOkXkxNue zyx<;LA;_9}He}FgQFN^gU(=;!)9|$+sA&eirm3$h@YV2hnf}r*?vxh!zrT#qV6f3^tXBakCY0SrS$okjXzK{WAPs-ua|9+=y{v^=}y2g)a$Crf*yPKVC+hJYtS5B{~P* z%kMJ*a1q5o%6khHlQFFggn11=8>4qd7P=kd)hn+?1Kl@h|IrxDN%h+YoB6daWwL|B zaAAXbAPQ9)TY<6rIPXeuJnt-lUg}?Hx6ihd&cuk3JZx1O8CIWt7dIp?j93+^FaqZ2 zd;fTmtmc;bbl~V=u?f63Q4COYY=9~*GT$dt`sY}2MbLH626cnFOS77K{C2VbpwnvV z*gJ;TT6nDxHvAZm@AcPd4!Z2(+uJm9$WaZ05WIah9&B({bKvY>HA5hY`P&*&V##fG z+yn}#H-SPe#Xw>6J}PED=@l)Dz^d?VQC}#a8EF7`3FFDN&?F!~+v<;7_6lK-#|7RE z!Ab`A=s>(_%#}AN)~M!*v8*|PXmQt^FGY4WEe>SR6mElEH;`)LaKOoV!|fVRpYX1C zh8}C6mA@7`VqXhq)E3sq6h*^7jT|%V)@PMqnzz3ZL{rm0OC&kmER6$)edo6c}9;epnctqQzs52=yEX!xD(k!cU- z8i7;lby7*XUK-%UjT@fex<@g0DDo8*gA~3=K5P8CgL@&hrNx9rl22@4olK|x!X*Vr zd^fmjkCC%M>2!MFNB!{mVq}7IwOcp%@aO2$^lGTz>34|_z-zV0BhoaZC=@w+`v5cr zREbB^jB3C0Gn-~yQC;@!D0(ET195K< z4|&)=vYmGhUiVKb^{5Sis#NIj)aHv(0S+u*`}b;Qec{qHqeXspA`~9pb~ap=7~ae* zp15G*pws!8)pRqjX+|y$7c@+K1f8SlbZ2M*^()&WZ@4iV`TOPlcU9)e4u|D{V9CN5 z%5S!+OM^P)MdS+!EMWt#do+cikDc<|`sMg;Ui9(F+D5@Hc@lq{vI9D#*2`}MCz5Un zQhE+5tAQIjovwf?v0DEnQCB1|xgbmHmaWP2#bI^oM3M{Lz*Vq|yd%NjaRKm!j^0_a z6DWpTmJz*YU*Rl76khpB#^!%DC*5RWt3<^BgA$fDUYahgl&tcpl(f?~0_&9>sup=H zoIWf&U}st|ldpu^6qwv<*AJYgz;WHuqrQLXHQ#$Vta)MCO>KeoCPT2+`)WkhghFyj z)~rs8!X5=^@4f&{y7ewK;1#dqU6REKdim(=eAW)^@>qR5dJFL!jB!KnlU3q5!;bni zH}v|3mc1Tg^nR3o(QYK$xOqPgyg>)Gz+vS*RTKjOp>isw4iwm$qtkp=@sq@f(tVQ# zo#r!nnndO>y>F6E9XGy9Swts$FB-q@wfO=RW~*0J8{4URNC&tlapS9jXLaqw4bJTm zi^dN+#f`7^FYqex(t#_!XneX~8;y5$gKk0!?8WL8r);FKF;cueLEcmEMvXNbAg(NP z!gM2o{!kHJLkcYI-W_;*b<_kW`zQvaX?IdFg`v0wPY`HVyFQfZ)1+lk`lT1TXyjFj zx+R!)2TAO;uJv@>Yx|)Js94wq#Q4doqp#nYzHHi)V~?HCm`Lpe2+|knr@gJ$h_BJ*F*= zuw%oL$GhL}+EsHAK!^2eSY-XMudc#d-@cGe-&bbSJ&JC69)$*dSNK)JhUm`4^ z@IO1j!F=kt!^3<8`+VjGkK#!05OaZJ79HXFew~w#k=n^9TG2`4&Jt>&3TT5u1qyU- zgb)NO1~<6H2IfO>1QjWA>1uhU=r-K4G6Y=E+^Mzx*C6m#NuxFfgeg#YEm>6+fIHc4 z&^_I*fnRUfo+zDSHc%vm zis@D5`5F`}P`7Q+>85Jkn@h+=>1q&IA9QLXv9L>8Myly8^fI!Vpr$oP6E(u`o9mo| z85`km;&vM+%SQEg?l^8U3rihYlF#YM+oA@Q`+RT9QE3!CRegOiM)x1ePXgcX z22s-tT+Py@`V(!itfji2Bwq!d5bfeWcG}KsVcI>`Oy~ker)HjhsokmE#9W4GPP<2g z0yjsPD%};eLa=J04w?_w3b#S5=dn|f^0ZsMs4#RcHuPe?^Q8Z2|AT@G!3Z)_BLI|@ zOQXqU+|UG9a{|=3fe)^H>T5JPzu#83ifrZ+C8;qnP^A? z2Eh|VK`x!9>X}lnNQ1z2gQ$~E6VAojc0AajG91Pb4xXwICJ;<@p)-bf?NTN<=Y-S- zl!kH;nY4ONZH6(cf1`C5o-X#?XCA0=STn>Dl=;fD%ws-5(XIR6Zk&aIyrrr-{yHg2 zALaUI^2@v$K;r1KtWA^zMK#%Ue{g(2F29m)lVm8alji6Fe!Soi-4WIc%UV@HtJ|Pc zHh%;3jAnos(rwjhIJ!qt${%!U_s{UL*%D_nKp1U|b<_R)t2h}QYHq@7T`oqWQ~vh5 zr^)h{#^~gk7@ahVSx>>7#q=ug2B-3n5OmNf(Y?$o-4FM6y^t(xl_dG#rcqBT+XYE) zn_=a7v&I!m&l}HH=ba|xXPWcYza*O^Hb(o%o2pcKhI&3pr!UDGqqn?!N!A1g=9WA| z79TKXfG|#=)wrr5^fi2k(~{7o4e`H!78PR54{fp>*NYQ zXSTLR&YcG}0KZcV_wMv;R6`-zfq-O?DDPF>m&ZkwN8&wpE-e_%H$0-R&dQrOIg5^) z?LG6xXWqd^o0I92wU!h)aJ25Yi501&n7tI)jU0T*s)dSJ@e0ANsO!^nh0t^ftgWX& z3ARIdm~NnJ19tJsUAjHXsX|$%Gzl`zb-Y;dZF-lcIXZ)v|UcJn=C=U12ED0pLPs zXkK{i1eb^Gp$8<@`fMAHrGr%ue;<}ew29&ZvVmwIL13s@HxO_X^4CCsCzrlHU3-<^ zq!+n8>(w5TFYSWz;FS#g!cK6iN(u0h-3iyCgBAB$O2A{M4&EVp%{Nl~D_-sK)Av=N zKekPI^7ZY!gU*kgkTlL=3xnAaIlJx6E9S`tmY6;+WEcj^A=+fs03SIUwD_gm3T~jQ zl?xfymWGG{D!)^>%SwKGxV$iS{& z?KlIN%MF*ZFP}H3Jb$6M+gOOr=S4R{cG@BG)%xk+F#lYpZ`9Ya8@pW}Zu;#EqkrRa z_i!@V_0nw0n@rs4Llgszbv2m$j_1WoA4bJ_;*!}S-=)y^(q(&KjuTf$Oc}xAneX!u z5}=8Z=wn=s&>jZ|39En_6elJiUs9Sln@;u10(Yl0JWB(bHE=hQU}p%5g;+P5EZ7(g zqnwz4?Mw{{ECDSI&j>jQ-7Q%6Gjb=J1I8@7(a%0QVywNre&P;uLSl!_H?dF*U-rQ1 ztdm`d(qc3Z)(i~eVdrgN0j)V##s(^wG(q0;7GIn)3a_WYJ4w zU$RW>%UX(ALy-h32AT11xMBMh#8p82+x-)tlCj zCWGX_jw@RR3Csxa>a{>Y&B=)mp}^GGo_CUzsrv=7wu-x0`J7vii?Id9nrr_XWggmf z*kyzz=De9-?0*N^ESIWod4Bo`sur$Ac4_v{!gpEUUH{%kpRJX5`x%a2jd&7-Y3JnA89uHvCetXtOfBW6YU4-XAfWy>V^c9{n{D@C;T3M)1Cf;{3-_tlOP+0aa zSZt81uO-(Ss2H%yph9?6RVygfFCwFH6h|-*T#zw>;YWN6CuEE}@Wv0DCmDT?_qYDX zugL~(K8NF`Gk^$rctLLg#Xv1f4i$sFC6|KQ-Pbdv;pL&bppgNYV+{NW;KVMS(68To zpgVF>h83ERow|e#iVhX3pI{ClU3$&qoX;1MDi1sHTfx+G0oPdCoy;G8xbTlgxJczc zk0iIplX#P;dF(jm5k&^5m=pi{=QK*EK1L2mG0b#C+(W~mO;qk%SbNwUjT%Vx&WC^+ zVD6DARKp`@xaNL&Cg_7-qrb} zcs57p@$zQY3Gnt_84?5KkW#O#nJJLd=?cc9cu$_@w12T%lRHQq%kNAs1@6UURko%e zTz_757>E^+PN8CQb9CqA0|6-k=jas8A=rM_JNNRD?EaD9a-{JOyR27VKk={l?>H#~ z8Fh6DcquEQnu4kYxl>=j(7A=ph~su<$p7o%5@#a>54}KyPMvZ%|Rk)d8W5LGAoRB&Y$V&3TY%{hk z^eT5LcggD}Z*X5Z4Sz3o!i>~1J1U;mRBc&pt~2Jao3A1hgk(_6MvA23N?J}I^4sE? z3(6v(hXdeEy*hBXH%%=wvye&CVLBd5o9B$LkR7mV;? zZma6g3x9pXe1HCulmfHt=JfbF)XH9RaS+!xRR0pCHVoRG&(EpjYu`M=$~9(mLM z@JMPdv;0D3qglk1GngcC3(xQqxTT?u7TFCq`E*MTk}_CQ?vbU?t+dAP9!Yp}fp|;6fHa1QdPk z9_diw)(=drncgT|o2=>sh8X0E(Z+_MJOfe%L$@uc*FsV8qHe)&NTp4xUTLf{(1N;dQ9e)M;>8VHlRs!Ekwyr3S%WsOQ70a7b=o47;INeyqhx=1N77XrfU*tIpfJvy%e~<$GZSFXdPjR;py}OzYf&{c)tT&>U*SAvM*p! z2-3zrbUDwjo$?VE#El`|ICaE$98=Z4=w9%EYc8mFVKfmVgXp)Qe7jB^4?I>_nw#gl z*Y_c`<)f`R=oK%h@@@^{GIyLF;NcgKWYUM-~JJd`J(} zV#cBc2DW??6e>d))iiDK#@Bhs}I7T z#rpG|G;obGy()|t9w9wosxYgXrs|Y`9+;-?RW=7`4N9%eq)37t-RaU?FZ>?}zsQ%s zo~%UWXK2ZtitYeDm!J_KZ*#Tu0NL_i z_+%_A=V0jZdvc?l-*K7)$BnA{9{zs^%)!AWc;~n|H8zaTX^d`;Zi2M?3c&!q90GW| z)n9@W406*R8N32sb2PSUL!HGK)&V;z#{cq<>$gudT8_?{!IwyyrJhI!h5#@=4@<3Y zr5Lb#*;LGZri72PpYOXs+{<6((o1@!AJKt{x{-GRACckuAhce$cR$M7T zJ%#4zo$p_QGni(IbL)Vr4@@!H&kl0L#E|_PdbOMNW4C)j^{*<0MznmaJ^M$pbz`tDhdM`M?SM+=Y_JhdNbIb`ePv360 zBXbfK2Sys9u`XAcL3Ua)Bn}*VYcv7CL5kT=k!s8=W_##-u0-`u#!4&Hsp$5*64egJ zwvh_K&WL>Rmjt44hP~6h$paAR#li*LLtTwXp0I>U@=2!?m~KBjhrGr2fVKFON_q)N zmkuhIP|2#D5w}GUtr0Gvuwx0?XRUOA{wS7;gWffrPn>7B3+n43XRj7NpeW!?mm1zGpWP8! z1A8yT1UxH%Vqz=F)MzmsSx#rcrl}p)r(XUJWoINrZmx$kK=ui%)7$8E42ouGuZ7hK z29-6uRQF2|*84)yMGA!{fHioxOSAH(3rG|8@-ga=EI0*G2CQk2?rnjO_^zIr{K?{?f^0b!-a(`OBkJw^u! z=#?Q@U8g;&NpW7Tg0AaKayjJW#I=GnA*W;@>(^+-3C;+xVoY0H!ft;8GWW!pSB5)m zp&d)f`iJS#448YQWZp347z2`Syz|Nnyi!RnuWvg0<7$g9_USRKhQQOe+Vu!K1T3T# zsndSikIZRbo;Dc{?9{N3yYvPRxa=TD#QniNE{TFunl-?+U{GVhhMHF44zFsukB9QL zX+9OgRC%d#@uaLNH^Pd?K$ya6>H9eL^X9E$Q>9I~(wm8ETkCpJwL*HCAeIcNSV92+Lxe9bc zcU>-umx}T8Aka)5j;vO$aBq&DC*CdV(xm$3@{w&ZmzO$qH#B^w1k{UeK!y2y@fz^9){e}1?7r$&RZ_OIXiCOP}kL|X2c7>uhF18df0Dke^l6VeUTyjS_LQQEcgQ#|Y? zD+nLpV+}4+lOc(a7FBMNfwQcDmp7%ww_dRc(g~IRJ5)$EmaGB|(g(a9%G)&Fr`6Ye z-*!fiIh%&nNr^R`4}zDfZUo;Au2+^Y4M7#q3BJnbrhmm zn$^fcj(-8_F5w@T1C_)2fv*9h__ffSkX*wdEbr|nCtTSmjYa~&r$ITYBNh>z^G1?? z*pOFzCZdpRcVMshlnHcdD5jDkd#M=E+R`gl7STE2-83nxArXBeuwGQh?^1%mmaH>QnqZ5jji_XcTN$ZVN=IoB}$@=eT4ggEc*M{3M?;k5p(%sTP%ZJoMZvJ?Le) z!$1cRA2c-PHN}_VwxK|?3o0zmGq;26hRK3+LPgHq{i9cn81nsX#T>HUfibkx1VdXW z2Kv`FQ8Ag`gRrAIMc-E44POv=Y1(D?9kMIIjnT262 zR#B@tr>4Wgg@q+^&OC7qlkanfo|6nTdAU54w9u*NR0J(iW%El^b56s8GZrSnI9=R; zF`_vh(M{ZdG5M{@SuZ*;40&U$b!t~>&$=CnfX)!nN_WT;A6KWkZ*&`wjfHB#*OSEw z7+>FI|2Ld~@k;jDPj{HBm^iG(U{Nedr}Oz;u${${f^1DC)W<>PQ#pL-l@)}ShIgoX zL|Uwa!83cI>~fyql8?4cWZ+fA8(ZW9F4%$%P^#0kPQ*ZI4xIjUfUKXh%$*Pr+X;Hxc4L`MA!STF)d2cvIeu}-cJwWccWWViu@O*V%N z&@Tt=+A9&eJ2sm4K)xg%Fbj`(HDEQ>57Ph@(rP+Rd`g@vtl{C$F7@w`1Fz_?T+){mbCV+14VrHa3S6Er&bJ6^>9kj15z5wV^uU7^!fbk@M=tP;!b*YF z-xWgCip-_g`}Ghlp2Mq#4rkNlGxWS6L$_la%La9CWEvFIos+}`JPym`SIm46oHa#T z0h_j7J`%;k($t_pR)AzxgF@$1GF89!wGxAt#bMb(*9s6&8oV(gtr$p#0M&=hw|k@*LJ-Vky7t1Xf>d zsbs#mPohOOJDe%tg8}S-_k>Q!31Axs_v4_mJ*~~~ua)y`55L)t@7XPnh1l;a@|@*L z^F-!L;?A(7IFm!lWQ`G9T+3&a`RVaEICAPdb@R& z{RpJZ5WpEtV*Hw|n4vkRfQ8)hoi~B} zyvYq?^nu-sFfRUS=8$mM<%DJWn|Vz$bQ%!d%%%5fddLdlBhTw&UcHMmY0dPH(W7nx zEcWTQw2FMT5w8c7bZ7)k(Y_;-{082G{=+<{G!?N+Wc2$wSsI%^!8u%C=)azan zF_s3!Dj*GcmL^ke-w)+~{jl)2#_i5qg$E{*FJ7AM&Jq({g#n7WN0B>J4EFcr(z%ls zDzw{_34+2<14UxH$BL+K=o-jVVK31B)+c@+Es*_t-8#;-fL7qVGB`ay9(*eGGoHw~_NqzSO{PWhm%FMkzg^e0-20#2n{W30vxY0s8_H1iNO( zM$JXG`H%GfSK=|KY*5@%EL1!WP7oM+ym9Etpe{f_ud0(CeFB4%Pj^pPAt(S|A?If<2wX55?^yS3 z!yUF?+iZk`(N6nQ_j7WEsPFy!_Gf=zWJ80m-+&FZ;irSB7juKsQ!{S21EEzv|8%_3 zq(pcAK}+UY>Q8at-E_K%nMtIWc#15iVleH9%L+OTcf#*LKlR+X*v~w7^s58Mxv+ZX zIgUHp%h`>{Nw+`zRcyq>A4(_uFInNh%S^t>{BNL`6x_P$U0!rwL8=}+#K%q@VRIH} zR)!oQcU&?+FaS%jTO~<8+j+>FY5SGK^5j?zg&|+U@-Q|($_@ofTkBnWy7ML*;h|Z% zw2EwWV0i2`fkz?50A6ybm{v)R*A^O+IyVGsj3n&^6dnbk8gcH7t1d{%tkcYM&j`_x zPUX;1N48QSC5O)nAJ4M3dnGK8D2Uu_N*{eYbCGH(8>7d7T>vQGzW9XQ~^ z(mI>QBr{zkbMkG~lXs2LD0*}im<|uocqZQ$e;$&e#2_Lk&vz_r&4_SgAE=m&%8r}) zpZ{#Pxf{YI>EpoLVipI*pkZ#*sBZT-#@tm_2IY^LY{{^Bu?{W6ZWyDtJsp`7M-#$z zU`()ta64&Kx7V|}7~0addz5*BmMk=ctz=RasS2b99t#69G{e?G!?@ZE-82TTe%~%{ zF!xYd9Kqv~WL1X89nYPSnbH-cLwQPCI5m&gBPb*t3fB*cMQH#{jXp*&=wp)$6c|*etFk*iqS*u`xnI`vWZ)e)p7r2Kz}*Bps9pnz&R

    0%x%?JPOm*r-EIijGuRkj!wrGqLU%v=^Z)%9+d^qR zF0&Jd?l{V-559Iq_UVj|XYBRM@$U9XWjj=>qCr?QA|tjAKN=(N z(fz809(RKFh8|NdnbZ=oBDzw2TY^EHOs32`Q~r?{LVn(L8f5XMvRh~zvrg0@No5De z{8UuK)Jax(m3iYBt&sf0QBSISK@lXyy-k!D+c0HB*tHvgqi}#mg3FA^Unp#)W}G&L z!yz7not4@Zd*sCt-BE@q4In%Q624yLUfUweK{W3KxjFN(%HW0>p~otnx;P485q8~s z+iW=QvWwY4KXyS7=ru<{b+Jx3;DWVK3%oE2gd)J;?G^?;RwaVt)UIe%9Sp(zLLWHn zH)r;_=Sa5(m&=NzL*IcYS}~IdQuaBnY5q%vXEfSY)njG4>No_M^m1e5am}{KThl+B z4th2R!BxV;JxmURb)2=4i#EIUj)Ww!SjA~p0IX4NiEN#2BDKf|3B?%^=S-lfX? z%GzlQ0Yyf-SkMVNc&nmyARv+zdo(1ONtJGQ82`{=;AF}25 z@mckp*tzDY%tWRqs7BF+vhC;f~g%-%cJfK zcLt0DA9Mr&GFg$4fP&9Gt~VOyKJOh-T11y+pGa2BSWSSB1CixL!k(F1=mkuUtlFrCLH54V!NPQ0Z$ zYqh2vqu3)9IS48r>{`&ITkU%Pr5!N`WlJIs#2lU7Ju^E74$ht zG5ORnPetL{`bv72QL@+}_6lbs?uXtg z8`M3}ge#BUMBk?ukv{kAfabV5(Y5f8L@Q#C(hYQJ@Xg>+0FN(}__-e=2NORi308%c zJny@6R3k_47Tpsgs3KyU3oMMy)BGJ0686GkitX>B1V-h*=_ zLaS+zgqXcfULd$hkY80N+85msR4CX+T1h_L7h6duu?e9WzP&)K+2VQ*+C@GJDIZ?E zd1TYTd6&Jou`(l^qmA7OZrsfHM(gTdSyrKm{r?(3j841?Ewa*TeL%6D6uAo`NK=l4 zJWw7M&h z6OoLwAz(wm4MDc_v2u@~NSXu^ir7y38f>4N*CyZ-XBu%H3|>MW|oG1J90=3JKUHZJS8|=^1RcEo0+DVOnHL@1kfP0 zjay|s^ng7l!QQa#ifvxy&_sLQx|jszyxq!8Y=U^1xLpAqZ8t}(l^&zy)(JL5tb(SLhDa=~8Y7sD91dKTWo!_^%fd`}^R<5u_{UNO zF{_VvgD%zu$zg&~;j-ZCCKX0(lQ2f$uzM=)Mv>hkj?IN3-)yu5 zEcJ@$gRH@=SY9FB;hn^4QMMY%K-gRx1&r^pz%nwzN3Zg_D_bQf74(3X{l1`lAR_hq zXt#JkJPK@~)5_R??Ld3;Gz<1zq?$b>f4JO&6ZILFsG~RcV?sGq;9Ba#Ew`Jm5&IXTqJL^ZGh+LoxBt0 z6*wpvs3_4TNEK!V<X5d)%snQx6V2ut)@XZ zH?Y<(-=}rf7W$t2_N)_h6%=Bc%f|7{*{^Q_mQ3p`l+Rr<3B`R^zJBN%hJUJlBk`Lb zG{r#`O-Ls>L3(D^K!~O@np6FCgp43Iw}k=Iho~A@4Ely zJYnH<3nJcquk`@g!f&tTy#F#%YX$7R6uXNeWmJ3@_^~KloelgNt+`vuU%Oyir7f2v zh~Y<{yY{f`v}aQc9NMM4M`M4kZufqb_NsT6>cg)l#A>tY<)mDKKkG%>>)}UD@;QI~ zSehp@l8cgy5|H7?U+;YkjeAm`%~*dH!UoUT^$i$-u^AkHq`x0aG`FV?8k0ZxCX#G^ z&~f5Tl)(x*r4(CCL8UmXcxCM8bXG*IdjTwQ^A@|`6cq(lO6IlDjnF+VT~!7>61LEH z!jmEnhIFZtnRdnXaO~qSuUT!nt}(9l=a0YJ6n9FU2U{^bR2Xn!th>6ybdC{XMhXnh z`?P=j<2{K5F)OmJ{*ol|LyQv}o2^!e*+{XO6j?ji<4L>OV+V8H8w>f)&q;r2z(wa@ zHa)|A-%nBmD2jv)nis$IPK#U1KRBW9<@$)Jw%SN92yKr;JQ%B9FzbMgn~KEPML~;E zVN{47cW299tHJs8cF+mKuVY@b?RM+5d~+xSACg*3rPw5jtfb;mV6SAZ*+sy9)hO*d z?h}mjW6vZzXda(AEI)Vp;rb6NXIU1KUwk}!8`9N{Z~K;@@j0+k>~z z2j=4MN1m4;WPV)J2DYFFcC72SDQ;>0Mo|Hj zE$orvtqI~{pnfKVmU$b(PC|r!AkZ}WBTw__cNF^REuw+IJp#OV)VQ+;LnH7F9@TGd{0p@4U`^_|M+= zaT|=0XwUwMEO+A7rod{oSx2$y6j?o3d<~7c@-!IP(?YjmX(u_Lfmlbo=V4)0NRN1Y z`&HX#FPev4GAAS~S^s^^c5FNiAtrN#U4p;%HgkoWfy9Iy*$`37W zT~V#dtC6X~YoOzq2K=aOx>$aLEs-?GeX7FW1$2hATYv>f8@-V-d6cOZUn3P%e#9}x z;I<)nZ%74|s}914SQZ*$p8_h z1DwzT+RU1(nmuL=H@ly{-4kv1h+}ghKRdEN#n50I`EXiTaKtjO+!zBF)yl=}fXmST zM!n(iy@4MpZv6i1Qr5D%xPRVro~)fjAc-*~z@VqtJc{H}@p+m$1xT@D zk>@?vI^}&CBsY3q$%z0#TWI>*=RV+);!`d`#ho<&k6b~5&Zj8`#G{PBWN8F+QCGi{6SXG#9r&LEE#MbZHxgU;#odm3_5DYJuRrhY zTQjC+OPipr0MO88;)WIyDh+gwu!Tk#=I!ctzGQ9ECM$%m9ULjO0=IxUYoK|TtaS1fb6>6v&jp1YhIf)N9Tet z=EJ$;fgT>1`C{=H2heb1=BppjfBC&-QCYqsIfisNae(YYD>h?4#r9F;Ar=2fls45U zf+`2B#TW=aOBrMR6Sn>}=DBtPqc@jl>^#-uSI`>iS#)J8TdA$6my znrd)M@XZ1y5mtT;xM0JHK7o$Ku}=Fwg6>)MOcULz+V6^=a4;CRORY@1EXS>ZUM+!g z`j=bC;)rzByu{e`0XQ&z*yad>TNzu%UY75bwJR=$terLwu4Fo8=@B_F%0sA_HOfr^ zy{_;+_X6OZpt5wm0*kKC2wK4mJ|ws&12fpjV2x2Pw74)Au$lay5n=5M)H&!CPhIE< z|F+|of3mete0T#+9EIbsTAZu-Q_@l}Cen-9A+4z+?wjC;*btgxl*UgX(ej&v|M1M1s(8G6l zYjCj~=^@%m!A?J&dbbbOh7GvnY6fQMH2VWvRjmPioJ08Ew0w zkVz1CD%N;DR5)x2<3Ib>=EG65&%JI9Ly?}KpFBA{dT=MLSLTr2lSrf0l+;oz6j@aR zMOvfRv_*D7n7ba5^@*_^DlO^&R)-V@)GN^HcF<+ssFbpnSs-3YRz>dz89%*Sg7_p^ z5ZkKWJ!eO>wi=A|xu|N@!Qgtue&}S>L4U&BhxG-WR^+K_afEF86G$@|rK_%^LplWW@>F)DTBbC2i_qL7B2sQKu+} z{<1sOIkI$ND_I_rFD>^x138t2Y^~qc$N}J-mI}7Y7qb06r-K`Uj|C^k_WRuPZe^Mj z&2cD{dj=H1&iPk8B}g&yxv;u(^nkG%Zp_~B0*v#n^@*N03PT4&GvB*iMhczSi#}>) z5-TWnH$}>+_+79%A-Ue*wtsF4EK*%Sv&SGD`m&~&3pNO7r*lC|EOi<gc zmQH;`P#9JW7 zoFM=6+uv*d&)dKG`Jet*wv=KQQ6#};x=nA#=o9T5h8HrZb8d!5U$G%lELXnd$T{CH=zYxO`2y-{8AcSEz1{Esf0) z=Spipm^VjwU{-Tnm+F87`()&SXv6!Jbi zG#q;2n0?6m;;O%X@aleB<s?pGLW4lJ<=)yN{7Lc7z?Bq6B8HvLb3!QB$srbIwG{NDB!x4RS-kcgui`)Dy z9_gxmu(!V;Ul4b43iz&gcEANgFc&3xksx2P((R|Bb@*@B#+MYQcSY_?3$<*=Qdb&SQdygJ`Q>W0@i3Z=CfK*gN*!|E2%)aQZfUVGB z>xi(EF7aAI5<(BmDw+m82lu+4(UgW_+pS6nE+3I!jnLkAubO&~6b2^4UB!a6)3od4 z{ei~Fc3D1jtiXdyLXHG$k3yid7(A*J|Id63ZujUN^l7hk@@z;^)G4aLo4(KX0&{S9 z<2E9saT}_UoEa8<|4|V9@$+7t2;y_F5be^mGtgnJ4b%ziy$;KUpN3(d!I&MiYm&$8 z)9fFJ-{R!)%eTJq8;g(g`nDs!ePGPo0S!6?U9gTQ#;9C_X?2H z*cyy?Xfr(O=tecJW%pe3pmji>dkRP{Z;Q-%Y5uCHO8Q9DeoZo?cS99IJzc}p%t>Zo z{CY*61~V#1)F_r$N{#|;3mdA}FwkwOl1`X~_h!@Ay^Eyl1F(#;JIbhB14?yZ_MqV) zaslgHlcm=r*A+N!L1bZM8IFt0_0yzg&U~kpB;*CY3EzI}TIBPtEVVwQPLv?tAN`?6 zu4c_YUx5nQy;J*CmjgSa@o1qY(`#es<-jk_G{)VdPmssTp=Z}Ky8@3wVeI~BycS1l zic42z`QzBRnv2lA6kpYe_6RnH?w#5kXFk&{&k$Uh(W*qL#BH9bF@5eRUEil_6E(&y zcq8l08*w*(dGL*m{uij_RiCXeJZG3vQ(zq$NIkY)PmQSz19@>lv%aQ z=8a^z;(3p)HRJ_5>IaXH93M?gcY^X?k;xkHfzt?hG1A{hLtlw5~|{!?FK2R*?d!`d~;yyLh- zK463|d6|icKm6OVL(h9D?FZMPE%q5{HCf3l6dxotppAOq^-9qLCG_j_Zu|4fNW#4X{e zM2n37A}O-&-h1qVn^Tl%0~N?@x>Q;h)&{)p1aXDBL;~gRD%f30v!&3tNL?p~tu20& zQSYzXeU+fbPiv0xZz3n7p~^^4>ogF?FB0c#un(vCyN@^YdEOHZxcv383l4HXR2HgD z@Xej_*yV_C$+WW2$1didr%ms)TgB@`E=qRI$fkEiLZ4qh!xT{9@1C_XTC0mK3&od5 zeBplFZg2L?r>n?|v4eG93;+4Mujp-8D5s?nbF49l5pBm3C! z9jKZ8*!#we&4S{n3Q&00c7#I$Nrf4wi;Q7TdDL}p48s(H-Sbebf=4YS7e}e)9k@uS>N} zbLWRxa8lN>V(WtKvhBp2yaUezxQX5WCw;OeOjC&#Zc`X#Y)m8 zsDzFZTfmjSPdCcTy!Cz={7IRdvo>1fSiy)tDU+V?1Em2E4}r$Z}Fj1P{6oz-j( z+p264Rm-b2y(H0dK~S}3f3N|1dTbM31D=6l?k0NWYwJT&1P?tA21EV7?4I{AS|=C> zS7l^S;GDrRJHfp)g}%IO;#(H;Qu|*QKP87>81vF%WnLO7ww@xVsrYjswA&naWyUto z3hKHyLMc&LK@|lq^iPe+kI*wcGjDia7Iz7n6t_d`M8}v~rMc5h4#?I$5@rZ;HHU>Y zOkqg7w_)ma@5ab{afu|AbSf6b9#bQ!JX=~G+!(nt`eW}jK`$wY?4%oF>Y!I&onpOD zCQ#9_OZKDK4qu(Rjm*n{7Hk!ClVY=AE0h%OlJdpq$N*$N?&1$HXQ zOfe*dTbX$`1ecjg2{u~H@IA)NYmB=ycjGrw{w?XXzVGLM_x?|ZNz&T;k9PDd*!y?b zUdR2lAI2kf-i5&Lc&*kp%j~pUDUQ5xHhofEDL?^uAX{CMpyq!)wEg7!eYc(A5xln# ztRonBl(+B#>%^bjobX^YidN2xcEX`(h1|*lL25_?l>0Wv?FaGIe%T6B_fG7*rl)(u zlSA%1N?3K+xwQ{Tmepn05gc={HY6wA)?V*Y!!IjMVDZ+o@_y-kZr z$u$4*7OS`4R6KJQ$Nc15SJM~U!oq1?3yz6iOI`SvuJ0dz^V;j}bMMSu{njV1b$@l= zTDqW-@eRp?kd0vS7OX%3S*!?#5p|-r(`jcpv!9uB=FInXIwNOhoUQFSna;m+rh~gw zKvY0Q4X8k55d~C=61K|X%2H4iH=SSoaoo<|3lu(a+y8L^~$+{O6aXVlrtnigq+o478Qj zAfeX*ADy^3_>dE1;p-%4dBsq|G3ZbrK01Ai_~vYETFBr4T_4Z{CW%jwVtSox1(cN3 z(moU zqX40f&I!y9+Zfyl8M|88MqqnJys&WIZf<^9mRx_96$UJvK1%{j&%KAu1hbkx`*rvK z|VLzQX4Hg^F?1qM%wNyKB z2JLX~az4sypQCy#N_jOd5ILG3%a*%h9ab`3<%dMR+kLA1l7&Fn*&weM;8wWUIfls; z>Un;zN!lSesH2zGk~{FN(DDkzP}bGz8YQ_k?I07!JIhhE$PW5rJ=Yz&;Pun5chW7g z)6){++ti^b@fvhk=eo}IbFNORqt`+cwxv$;M1BkOUlYr|3+BHC@Al^IHBAkf$+on} z`p632CjMDz74%{0#BscA$giyDemb{FSt|h*WdFTPvTz$1qC4D9No;V#><;OAw}P3h z?1*Lm9@D$V_PLvmV*S_VKlf3-z4jZ3irUiQZmL7UX3K29|1xuO8V{JA!bEPnG;vxb zf8=q4@~(JW$W8Iv<$1d}apsFS)WBkDbE)AZ#1M2E_xShPgI{{n!A{#$=4` z%kbUwKkZ&>; zRA-~|vJA!M&vNwrhMfZJN8Ay#jDHXopVRP40}NCTB-dxD6*s?eXyIQzTl7KE8y`YB zGQPOB_=C?Ee+VqUPD3x(EdJo)q7QpvdCcRK12t8asGGhDSxlt7x&xF~x$e8fhZXg# zdkV$_r{}tBBX3}dszvWs{QO0)xW($pPC5fzaZL5>oSqjnTowMQGB3!cR$`P-+v1rX z9oasbJ8Zx1JUMJ+OtL8^gCgm`C`@DTdYT*V%1?r{ z8WzN%i;NV&#q=dn1=I>0l-?PN@b|k!Nn$zaA<1-$VyL&kH9aKBuM1wOBx5=@|q{CGbPzv3z#NJ6$7Y&XPlmXV?66y->cf zc1iOR)2hs;YwPTIXKtqU@)&P92wtGs$@7aNXmnJ#Gj0~+E_o<-6v>YTEl*Ul2l->9 zrq?>{^4>SUS)9oq>VS%zUoFLc0q9$57>Z7*6|3dYTGb=V6ymrzruAaUR)@L8Kz{a( z=3(ysIsL2Y2CJlqi7FvG?3gCzh>^M6OEJ(hqM;&?&MrZANwg^#^afi91aW}E8CRot zZ)AH(hDJM;4iq6lg3SjyM=158L5F?Z6iz;rbq~pwBVNErCHbL3MbUXNlJ}$u-GU+s5XM=7SwNCZ;_N z|G-AS-zHhOm=uO5aJAsL`JEGI_%!=|)F5w=ul@gHIkD`sH2U}_)FNV*;SqeqQQz#)o(R>S>NM6KYKdntX8TP1}2-<$_d*J)_c@Wf_}b(~Ir zWF}3caqrzMLs7?Wy2*cYK!c=4xEc2J*ecNKoX2GNu9Vc#Y0xlGBN=KJNOrqE^QLD8 z>|r(sEV1g!Vi}^pT0Ne1WKF2ELNv8nBMkR40B7X~)1pc8WTG{ipM4aQO_2;L0%ff( zz&4?b69Y{dsuag-WJt!jMYcXA^jIX`TT8wOQW z-QgXI;^0#8N|)=-J(8HI2~(^(Cqtj3l_tiTbE5yd*reU;*So8A{bEOyh9w>%_$ z4y+(&zeW@l;z#%i^)`uwzd_4=%EQYcAt(W)hxNVx^VFPkIhJqOi zmU92zS)cPUv=7@BjQ1lUZW9A6yP5pQvnSfKdKNBBmK&?j{j&yKKKFe(*s!QXC3Np6 zdnc0G9q07Uy!KdUmp|3ih-mYj1>E^APw&t|~hr@zW(o2@}ug|Ixo|0!k z#<81MFfVgPyKD&#@z`$%XUW!p4YQ$cL6PW^C)a_n^hR>;HQ-tISV4M#3V!!%yI)(a zEc38}RU8AX%vi#ar^gRB^RPO1{(E_E8X&aShtow$UoZ$=GJ;So#hjwZaVnzNe-|u= z=NBBO&j{KCRdmUm$goGUqB-?+jBk^2vv5hG*OFWAwJv46O@6z)o0Pdw5t`sN;INxJ z;7|v@cX$stB*Fjfa|RstIt@6aE@)C_a4Jc@a=_sbw6mh5+$j~b6COkE8sdSA*m ze|;^vG!(cy4SB^g1atwog}y05eQjx>?`i%RAZ8m#jf58-vxk^@NWK3c{;%UHwY1+o zpSe=Y9$5_2@7e)Te5~fiOfd*LV984zxMQ7?2IxjerHh;+=VuDBsv;e_Bsw8)Jm`S> z5&DP5fDFs$Y&;mU{9V}JSscZ?W4dv&+x{anTcK4RyO;roU9VTo9B}w-{z^ub3NXkI z!y5jjB~#SymRH`h|NNeQg569%{k*Ntn70BqhFr{BApmY+ObxMauciJ6j=1Ym<}7Pk;hV1leje!vsd8?xy=^bIJJ-p#Fb zKFG;e7qyFNBrTgHxx zVT9NQ9p-J#+vnuJrwq>SW7o&0$$$d}JDvV3;x~Vc_|t#C`@1Fo^+Y1uHh-Jtqmu1L zj%_T(tfR4nZdveEQD`%krG@}E6|ZLXT1{LvPgJ;%PT`@I1We_Oxy zb#m>cDKJ=VB#PEYF(A6wtydObPTi9}knWQ$qgDhVef;5tNJx26a9ygC-k0XePAfW` zHK)E z=&O?s`nJM~)gsubz$@YEik;Aui1%EiOT+Pw4kb2!G(z(S^4c$> zw9~gzd*8UUaECiy?comkB1w>b^!xB+hosIi9`#Zg$>^Qi$)5sQWqL?^}#843h+}naketsA>2R#HqJPlnS zf#|d}Utlx^BmdbpJGOA1ZT9}=24-cDUfKF%S&C`M_a#XsnF$c@o2Qfl21iwov_+kEngKF&4q}HN1aDS z!2S?D$F?ENk_@(ISTnPoQyy?pa5(%^XvGI2-33~w3!E}|e2;6fzY1BH(S2Il=AOl` zliZ!&4h{BckjC8Urn8>j)tInjj3lF1<3r3>PrtgceuIwzJH5YLPH+K z?4w9F711f_m!&FNNCroRs^LhHFd+TR@thc1om}g<5L$EZm#GRxH)VIkH|IpVZJtr> z-|Ad1$e#k``FLkCeNKQftYB#_a8NlFd!2ARwcN^v#fmWaY7VV%#A-zXp`xCXrl4W7 zBH3{)#tg*b$zAa!S>wF3^argZ^0g23uO9{EL(Z}+9GPkwlyOwvTP@#gee?8NA7LN8 z{uce|MkNR=!E^oGAyOMrAk2f{I;CK7LE)*yNE3dCAJ4yi)$&g{&7{hWCa`(8bS#0!lW+WlfI|N!?4WVs$UhQIgD7?z<~I`->E_jkpA~e@ z>QH=j)C8VpI=u|YF|hZ%oo<-B!~goM1JXEsQYHo^tN=pQ%>38)Ow|tU z79eJn4UwZDo6@f*`k(bK+k(FE$#BONrkIOfMiZl!PiWsXTX;+iIG!cVJX%($=!gBU zOf&eZx&^Z@lWle!Cjn=3MAR&cV)jrZ4Ok7xWl=U{RSrW|1?u3KWm8k+ZGv2GS5T7Q zzJ<|DlHbQG=GOd#lKIg=x*LNd9tWOl4HAAK3lXrS$PZx zwn7g(WAcXgpT9ZPfR^Uh&wfDS+3jrXI6#tTgsoJHfjp2Jn-#iTy6N@v(&;X`aq2Di zN1~J5B0;j>jcNP+GdLJd)j2-~YPV!MIjD>i8;tSWU9XmTo+29CuTx_Mm94-uJTCqI zfdSK0v)!&aW|GHyLy-q=h$IEcIhw`Y?&~B=QJh={dz<^fGz)q__oev%9mfP&?Lr>6?5JT$L$(#Y$)v<_nQG0YvuIyW_058NzEkzX1JqSF)^zR1~fRJs8a>RSYD z5;%#RR+++UoMe#OY==tB4X{l{Rt0qo&@$_{1^&>Z@W<&@N0u%v z=C;dT4mI*uE@r#UdC_$tN&-F;ZJ!!5Q!PK@pUXut5LLG*A>@!pO!&6>)`)xze@`v> zZMluHWHjsD!)oE0^htB#pIi+K*RL*;C1eA;h0BgZa*!(;;S}wp7$6ziMnxR*Kvu#A z53L8ZEw<73T?ZXXovT1<=*j7gyw5xbq-V$~kK2NEvLfe|uc_iC6QD3-IigIQAnSo= z)RFo_I7224kLE#+%y_tq>5sIkD(0d+*|4GQ;+f@`D#)~wK`RLM)O*o zS51!P)YEmrgM!tQ8c-@1}qxIfpp2a}0xB;A^Bu+)Z1}JCfr2E7rUc~}^dRr$g7G(OW^MX`(1so(D9XY=hm?SQVj!PkBT~F&9OEQJn zI)Z`&YPq#SKUPN$ON5xaeT}2TsgDdc==@Js%^`Q~csst)sO)m(BxczpGC)Q2b7R4Z zw31TbRj(xH7u*X!Ai{hw7VociLA8@*Q?=8Zl&I&s)~$r6zoQ`-k_6$$e98mv`jm0* zhHVj-g4JkJ?ls(nAO;GXh;_1+IWhVQ`i^9eb$2EY^5;nkuFUMXFvSi_H!l?OB^ zv-s;~91xy`=C*zF61c7MWWPr5v)&C7+@_T}XU@=wb)k4~BT#?mOkL(UJTkVwoe#*R zZ_Li~-xG2K+UPoF>&MiIGeC^CjgAZ|@xr?Y9PnN&^gJKX;rw!V(bV?7e{3*~3b8R@ zcDxjti6LbuE{itLESQO%1iMAKbT#biSMzkzl|aRq#Me)9lHYPlE5v?-?68$kADrnM z<%Ww`62G3t{aLLf6OOHAqMkMp<1|`N4``(`X_uG~k-=z5|M6x6S#QTiYp;~rC@+xdVQ62Kox01lY2_}~_=+Gj&8&(XJnhCO*eazo4A*J^>-Bu$^&AVPS=g_ckQ857sN2y;C{M?G==v&FDJe937_)08QeC` zdf_=vz1vpBm8sp~1>p_KH9!6GHyQ68d!tba1dL;X;V5pkwNJ5vAvNX14?hqX3{}*& zE59Kz>~?r|Tu_{4WY%_2OcF&lQxT~OEX^6BZbqTVT+XL+5Bp&{4GMCgD;pz(C%Mb` zEs|8lcE>WA?ba?MTf!bqMvvR<#l2IZZ=1@ty(Dgr89}#Ien)}PTP8P2hvVize@(EoxTi#qisRh=#;_9^Xu@yu9s`Pv%C!74BrZ`<34xj{hTJ{ z$=7=pw2M9my2egPjmKr*eN%x8&>vDGFk-9gXF+GgXT)pf6b0`0-|bpU@)eMQa{&eu zdd=h#g0I(3u7{YyImJ$|9&Q1*)fs6WSpt$hu*PbT*kLg6mr;(UG!ZXU^X@?-^P-`c zREq4RB36)HG)T^Z)&NidWD5JpDOb?OtPNH}-A0S7N7@;Rs(vbz9>@>VKt*7bWG}Zf zRMiY7sW><#1l0&^=|_zd59}tH#ogwYt-4*yLMI!RD5dTL6f8q9K3I*0~eH(ZvFMpBlc# zsf<&>UBwLlH5i0Srn_A5{;P8>xjQ=~!$`S5t%&DLo!cyyX(ph4BEVO4sPK_nw zS?iv&zQ6E!(TfZ-I7xoVP^*3rQd7;rk3=!dXP|h5b*Ra7hOa&XR12vqbZ%L?rDOk? z*-SXj`qm<3LOMuZ9 zwRc|NRVI!8l>gPMtpRnxkw8$~A&qma3W)b^7N3%*IQBzau1rw|6X<~QfMi%D)#pF^Xhl;3!hHahL zz=4FkihnDn&-g*vCoJGuIjLLmf#Cjo%1g5<@^*Oz)Jt@Ey>@$5JLSJlB{R&RY?DOL{aXqlT~3+D+F&ah-?KE zSrTiDTY(USVvztFBpH3#wb1>p#PokL36p5!I@+W5e{ksmGkVYkinUsHJQ zH)V^k+bVZ6{11rPVViJ}#4ri6Egt8CRzd=C&;gy6)!sP*wY<-z4*1ARF^<>f{t#;6 z3tsz_qrx%?eYNECmKnx-;}=2&l^^>1UXIA!D;-sC zQKssFFI4);jBYMc3upj!*ae#Be#=is>#U~t!pbmMVw+eEXr-IY+oty>Em`xT*KFx% zEGO1FVO16?RKxOA=K0v86`qyiye#IpzB)2Fm<0L6IjYHMGtdC3g-5^~(0pNOV;z z?c$vvTR}WUe|5mO&Z|M$Akn|-At{_jWgL{7@1cjEZkJ`l5{FUXvviWPO%OCD$e3X0 zxwss|4XhwTefROPcmBDEk1!MiMZb_ZD5kYeNq$AnJ+fP~kp>B!kXT`z1m&QhsY=i5 z*aUJJNYRK)Nx1b^%aOh@ms9CD=zv3AbFZeWMA)Hq2bw%Jfhgb`#~XBLlIM^cR%=eT z2H1?SHwIvh_K06r)8=;!7H3k=e|||SM$_l91gEln*>mhrMfYQ*r zu}20+(a@-nhhFqrr$L9EUaig;e#XwMzUlq4L5IHSopd3+j34cU5kh?8fKTJR<><7tZ_@Qg8RamhRcMsaXnncjAlyl*f!YolW4kxq7@VLNVeS!c9ye@rp` z6uFQ6o)4i=yq&=gt*R%b;HxHI71VfqRT_?WG$?h_GG4!Px?`?;JJYDdPSA`0etGeM zH*UOresSrdjX(O}?vjg(SN=!An;-W4Ky~o#M$ih=57eL>eh3FLe(-}QkF0_xDvAD0 z*gS{tw1-|bS^uQIlBnJDG1R*2-)sF_x&Iln=^9{ParoQ5GYm_XgHTaG((HIo{<)Die2`**Vrd^00hU@?=(2+U!6p7q*g1W;YVzNfmI-Ju7}=tG8VYmZR{kZ`YG`FU=+?!w4|jC}s-<^Ak}8nx+k* zcVIzD^6R3tplIAmSJ4GPL$PW}lUt*t0fIKku;}!2Ck{Op<3_*K*c>mzZ}5L+6mF0h z&~f}b-@QaOzBK6AZ-kCr6a&pp+o_10H!q8pQ*E*%LtLZf%w2J(-(I&u>LGWhFjsgf zwAESTze4tqn=6dtW6Z9T&U0S|{oE6P6_y||cqPMAX}jO)->LYzhXEwNsClK4B)v2s zfr`5kwP2YPqoGJD6_NF=%g~;m8n$GEq8+#yD?OKmX*j*J%1IeeN+$ZAnN~_LWuO(_ zA?e=O5rtRFI5%gb_HC)SR8d9m<6$PicBmN}ScW}W+wIx@rJ-qe=&+~ixNtYwZ^xeM zX`|^cr5G(m4pR~LL_J}Zz`~g+{9BLnC&_e9RxZ+I`EaDIXaco7&=Ybg5O*loOiq9& zS3(BkOz@Sk;b)ta@$=S9KJ|asPCgL6V~Co2zyZg$+wgy%dsB#;0ee3#+i{j`vSTv^ z9VsImo!t}zU2;2Mr3>HfdM2cCR;%+ePDNO!tc-Iis9KsV9FSg_8ZSNlI#gqZH_n0} zUFerU^>I=j$HOBuYPR!ihK=-3FBb1z_ zmse6afs~1>CnwWQOp`KKw%6?;x6%`vD%El%UAqUAGaAJMX*b;6 z>J-f@q#8XcJy%a|QZ{f>INhRHA*L{Ku^W)CIGAL7YMT7yL zB}W`Le{IRyfLbqo%Vv`@Pgvx)q)ws}U!=R72c!wye%URFCa~5qj+X}Iz}o37069iv zJ?#})bjETZvP0G{YZ8C{&qaoHO|&%tO4X&&Q)3uZpDGqKGD+fFvmZlaFUqAFYIK^# zg`P-T!vYbO&(YI&vPaX?W7y)%4~rK3<)5qJ@|tH<_$6VFEJ4;6mgJob5x`o>eUeOX znX1ZzDhO1JQAL5}xlc5&E#s$ygRQ^shRc0Yy+9YLy7lTok5aL2`YMmT)6#>mp6j4g zl^%kpH6Z3~RY0+ajd^QhP5A8k@h`hgs~XId1AIK-&@I7S4zlMSo2!)<2miyWpOJNQO|78R++9TZJC!#@*=C7Ku-|24HPYxo^Ns978pS}8IvIGS0wQ_%KavYIxE)uh6|usYfO^lOITtv6E|6N3X#&;q zD)$=#h=wTQfO;${0I#d%)zCm-YdnqFKzudLHaRkno$oeG`JTH0Ck^i%T1vLs@g}*z z2oo6;1LR}7QE2Y)I}MOz?gnz{!|&9>(HiI?L9w|en4{d9y9CF=kxgVqz0S=m#~Ai&3%zR54Q)`J!hBu#DyJ z#dMo2laEZ!n`F!QSA2EQ?vdo2Kl47h8@Apl9akcPit3-fs%0<}UFM^e-*Jwa3jNJJ z(rqFCn;rP;dw+cY4-q1oaUc{m3RGU{Fa62QK!aKN_;2k+WDmRSfE^cTR~i|i&nN~6 ze)FgZy|`|gVjCY=d!xBIf_U#1**aIyNd?v*?9HzS#uwEUsAxzAKKlXbhHyNoIyCRJ zXh`cF>4;7ceFGDoT<5Af8gev5g`zJ=1$&fKddBiHAx&i^Lp3g&G^&$p9Yf`Rz1L;B z0XmA9s1mZnj#s23Mqt=WF(AXNf#AMmM{VJYR+_$SC(olpDM`!`B% z(+$eo?_3t0Q}#l=12ll3_-@dJ0Ik!ppbT!Kq`Z&lPvZ0EI%8v)jt@TW#$ zw<@_+ouBS&z{{Wa9bHd$+c93sjPR0AF;FYDmx@r&&kMpzhRt8k4^lsb`oacm zYfO--l7+{IMxRbP8Oc5}h1eN#k*F>XX{Q!LUBiG>b#i)%JMv5A z0z>`bP^+1Yp_v*tYDU_Qr{;d_Cs?88m3Myq!ylOXGHg1C?0Dy7rY@pg+R3>sNeej? zzEikYRKV4`Jbk7CS_p7wrEen0^xgXEZAl&Mo3KawfT+Q?H00v}*4Sb5{3my(g?XA5 z;XPg6ZO4ExQ)<@(#A+G7RrIabw#gs7UL~v~`lC$Wbb2QUw0$}ScxSy_b*3mul8=&X_I^k-$=LDD^?tm4iN0#Z`Zu2mT@pqC12#g=J@hkHn7-dH^ zDdN(;PG0z4Qsmqw>!O#FEA%~ZahEfZLA8=r=WgCw;0iB?Lf;+kU!70(O%mgk zCgnq-4XzdROg(JVpp4yy^S}3jw9_;u$Hq6Ye`SQw#^ z{;lEIWH2kf{L7_Z1)K82za&^O!zQmAB4y5|a@`L=R3CoB9w(jKVOvu%`ixlyiqW67 z<=G7CM<+}*C0-IxKzwC{%d(wfwxZ(=^c<+r*UVroKSXt>Ty#QCC~WJ;tRhmf>dAtT zH({xj0%*^jLCfGV!dyIeMC;tpz5f%zOoNZQs3jwsWWHdz!D=HA9HtmpRt``R2Spj& zt@1)j8@e&5)p@^L<5b+YHiJxhY;c!Lm4767vzO+s^^dTpRx-rY97QLPa zQMzL1eyE^cg8Q5!{&kM~A%JqwDFJSIAYL`UI5=H#eO8}%P2gRZ^Fd3jh9K4u8G)A3 z4y^+asB|!EE}TRAOI!`$`PD_Tglu3J*|+18lzm39+DS1`s<4fUXqOejz9&B{K~^g* z^E}|00!8=jWS{JWk1AeRIImu?n~O!c(1VoWyPR6j9Xg1F1FNCL+^mijJ?6%<*7482YonHa>w%~vHd7eVj6oLCE#9%h^98yk^H7||X zfVQ&{F3Dz!iKoa0Dq>jhTmuwCz$VlzKKwdZ2JQ5GNHRxB^24y$-sTo!#5}Mv79*as z^`XndfgV$#2)iv_GoxV9!09EJ!%N|0bB>Z@Bvx21s&&RZA5b6&9!p|59g1q2wI=U) zK6+NuZOa-HlYfMH&tjB?fO;o;`uu=lv(cHmoM`MgoLXUI2ns0X07Y`Ch%%3jWNSdL z47+sh`Wz)a+)jEMAEch4I3QP6NqS@;OLP>fO3-4ID(aju0hk{aE7XZ~q56%@SYT~skLx8X)iD4ekzBd~Rk?ze-kYQyA?sA7KN07^1^V&*@RRJ;D ztc?{Uqm5E(R(78@Wet2u991*qhX@oyc@b6|CDZ2w#laO}+d%6onLf{~3fS+j zTH~~m*FySc#|B`nfz6?W(?1qCSrazq##&x>;P#t_^=GdSr;C)bb6f2=6o1KRwW_6< zQxrK)MU*;M(K{7`bJTL|6;0r3<;S7cM1^H2{cow|SdP{N>M-Ebmis?)L;lRG+#T+N z4yX|WLN?A7VS9n)33~#g-SFvrX&g^Sm(m-;p;iUMZ;@eJxcXsZrmE#Ty{^E@R~!t1 zy1;nvWMP(nsb9YIs0i=CPC|GF-m8;lAG8|_q`q3XEI%>_9dqYjpPctgt^raGeh$mY zDm!**b{l~tfnqjNWIYwJgKBVx}!|k83RF4EgMrb#DXb^&+fnoRHPysfn}z5l=lc z5hvIo;!uGH zF8m!YL{eutu(J|$C{GKc!r^J|4H@2q#B})d&`ll6Nd9J*$Gl3?&3nvS!Re-ZNv90o z{2w^RsCc#jZkDenW4v=T__BlBe}8h#!;~S}3=#1`(N1AAvlo8G%^P`)^@9nr8jo7% z0?t|9U(Wg;nAtshWC3B5vI?dvZeF45V_t)D{afWr4lQg{9uy6bQUd(MZ>(RWU1W8m zVZG@Y?L!{>Id;=yw=wbc?|e-vH<&5qiFb?0PCGVJB}Qf{mtuf2Ig5(e2MrsY9^3un zI73Zh+&Zh^j+rx71fCP%Y*)Rtiw2&afXMl2P&0c<-aizk!_;6T|CGE*JSb9i zOFC)2te#pPHKmUf2SW#2pbLna#PQAwRt0Dmbka@oV;ohkQ?8S0t51glDOELe8K+%#4BCm* z3apxHmgP(P#7jG6=fueOZo|0W*nstwtr-t^p7reS@BI9vlVMHzSas&V$lA#S8q7!R zYPL{J0!220I3}p(UgdWBVT5XksRY$jF@2+T!T_0GQFQ{oqfh=htUl8xPuc!}-IvvF zUEF_HaNaXjcz8+j1ZK1!Pz|_Q3{A*L!=hhvRKrrhHZ8Os4U?9md$gx*ckaWI*_%vv z%ujcQ*>T02ne9E0Wz1_9_q$*%6ErQ8$k#fgh4a+%R@XYNDw2OBTm#9XCT7q9xq2}D z*zeLH-xGSozr<@#sQ%W+vgNJ~l2h_*ZbNY5d@PJA<21|KL1oM8=&>9fU#%gdA6XYE z=53$yXGN3WHmo$2zrOG#Im#}pWXHXRtwsxG1I5%)R)9bx+r|G2U7pUds3;Vep z%A~2+XC5W3&auK;!*gFfb;b9<5S1x(tU_~QlMQ*$`AdCSS-qo+}7Uk0V@->zID=0NYfVu ztPX+N)?)ew1WNHfkdrHhSk=|xm#IV-idx#m(7V+OjP>_PzCcyvdqn7I*wI{=IuiZN8gKCG)GUYu}uy}MVfifVU!w4`*kn< z@6F@?bM+?TW(g`|P;nf#P8bTv}2tc4}bca!T$V9)~|m@ zw%c*!|FDr^%%&I+zDuVfDm_<9kVp>awcn);=w?etNeP(d9ZOe=oNfQtLyeEU68ZpWy&VuXr%im9QoLhb%?<~>sBKeX0oT;lwld_veR(-V`+2TrR z+)KB)C&Etmim26j59pZJC}KH<3$GEpj$Cq0%B!LKxEi|K1xM^O+))g7G$~Q-r&fZZ zod?3dV2?@*K{dLaVR4o%_}GE!;lFNfGwp0O(-L|Pqz#JcL(*MOV@k7f228_ce=?g&>C@zB&PV$`*gWjX!j@pX(UzXEzmgn+W)A;Fe#^ zF&Lg-eKB`0$+cs{19oOa@~VttiYao0ipX7z6v++BW&8u~>$#Uj0}dT7JG`-2Q^Rco z3gu<|0${~Wk=N5Xf=GT9Ns;FWv`%%CJ#hWhKn;uZN}?Xo+sWzJBcsw>thG>c)s zjaa+_51k3YhX9RtT~2^!w2#-$iT7@W_PIzt=+|!(oC&E8yy;OAXlNZv`dpqyztyOWXx@hre{Ou8CN^A z{XQ8_RfHWID>GFQnCnI=%rq$VT;ukT#@u!c(*SnChJhS&Ghit!%sp?b^IwS9DNOyF zmn7$8#v8)sNc4M>=@vyRv@9O>L-K7jFx`^Fej7*^-3i6N6B+EW6dOZpjjez%%DCVB zy<*ynPWA!*uB=&n#6Ooi)D;Vk3seK@pzIr2ESDAvb<+>JHw35gce_s5jI)|DHH?w=p-s4fm`dJO5^tW`1e|1n~i+x z<)n<4LTA#MzGvOngn%xUJY8|sC6(Upo#a=&=suYcn6QNiS_Ye7!emfZ5TV{b((W-$1d%F=IZBZ-+&ibacS`zYSmO`+-oPuU`gBUW=jm^)3djIC-QwU9 zcU7IF!1pL`D^aCE2xV13epni^o)VD9!_In5V1@!+5Y-_MwW8mp29_MufL|3*;a@U$ zJM_~h)64ieF^U=C8*Qg{xPv4_e%P=q#GnKIH&)n<1xZ8ejHN$J5Mt>|R!Dk<4$t7f z=o<3r-0P&=KRYZc97`{?PPueBZ`F*OLD`Ul1R~`Wmm9P37K1o51gk@Cy5T;1t<%_N z-{RSTN_q0TKfSN|?|1+B-hcgGw3=d8P$csC7_jsW#&{nqluSC@_W8GdXs|aaRR4F# z>B;0vqp$TPifN`u6BTg|GAZeF98u-L0kmJ}XoI9Bc35EN7BbRmo%&s`gp|VUl=*z= zqPjUJ1ELBlthWUU1^;Yuf^4-iF9=TI1Icu9P_8=)yl<1|1*uwT904yR@>WV}9Z?zx z?^czHH8i$$Z2np-PjyX_!O_r~z;d5+0wj&n1g3=Sp>@(viH^P}YIE0x#xOAq3|8r= z=aza76m0%&e+D9+HO1`a(QXq8UEf^tE7Q%A-LgKI+5T+vsN^3MR7ewr39{wlTv^n7 zJZ^Pvf=0&Pu*blDKej(uVFlaw;y3Q0*sSA3-8kE|9JFqD4oy6~y{pdOn z2xy(|NqZpCR3j|lWfy%h7K-S6>niH zK6~_{_r)fJU=4r1=a#8gXJh>X%5-joU3y|y@#6v#Yjn<9KmYRq1XwHJJ(6-f@;YT|MecLJkk|M z;p#ck=h7}ik<4!TytE-?=g<@VF01FHEAphm%B4AxWj`DXVWwp*vxH|Bj`(?%pS#OX);Pm4Bi>ji4>Oks(8zpN9A3C0?jcowFgd(9F^ zJa-Tq+`tMF_M1@?{bgU>i|&!+FU0P}n?W@m)zal8D<~x>J?MUTv41z6@3qCFn%)uA z0b#i&%AX%>5Xn#^35}r^KN<+&eQN=xP4L_K!u}XIm(~| zhH!SxOM!^bQE74T&;_KT&^l$&g$&XUVZYU&LrO>!xhRMUQK24Ef-Ft3!yP$5(xFcU zH(J;$hkqS%O()Y2X(X6V32CJ_xE6ckI~{Wzbin(OrZr7b3*J+-TQ(QyYWhf>wGDq03mdd{S`OPhZ8iTeP36185C8U@S5gIQnkR*$>$XB%8-k6GIOATiZ3ZB5t0x zJ>xjT3OAG9S{wa6(~Pzmm52{g=?~XQm+`wv^ptX_(U0V7JnQKjF8bH8^P{I|JZqgZ z+|Ny%E(qH>7{|xT>QPp6V83y$m_+F+3&GL#_EsoK8zOG16>JLbBk3WZb5AQCLC9XM zC?V@y!R}-3#@gS)XH3gW0S@~OROn7+wfY+zp1|aPY#{0EiZSii*(Z?b9#I@|h+^_7 zvY(2;{8hL6ey8s6&EYCcN1}7IH+0YelVTcZn?xt0jI)8HDKdPqCRc@bH$zrz=!wlx zZ<(sV=F1Y&5qdyWLfYJczXKL6>@45!)Sr?-)K|_NaO(>#@OlVT%vVC=czNM9!X{uZS4!h( zhPxFx0z7)~^?0v*_tFI#?h0}#9PcZ3E*VRGj#Z&!7FTS61MOjEW4woUzx`d)?zg8w zWxv*xm}%A1Q)^~IYO;!s7VP6!1Ps+ipo?1vZQIydfWm#v;&$0e$$qCHlC%8KjtNL6 zj1eR3C!OUJ<}otkj3(DKA^AdK#7x>T4a&OfK$l)m&zpt}tnwCFoMSU=;<3jGNFK|5 zE()+f1D7zY%Ye=MGJc&TQyA^02T*!Qy#QRaOrefO+O&#*mSAAuOpqaIUNW5?qLx3e zX~gnr8Y6P7Xh%l>Tg@ZqaLkgd7hUuOV=Y)&lM;dw7alkUq-~!;X{fQ;$ina$6KbB)+zUn&r3}#hlm> zVGg;`-}5ja@)_5mfm#BhhV=o{S6k=Vi6)(jAyM|#kc7{1RBi8$A4=tB74}G z4?EsXR~ng*&nN~~o;)gIBd8jE+$Kw>`ypvrN@sFEh4Qx~zbmxPxkFmR*Ep4N9(kmO z{|zUId+UI(-2e33iLxIZevG*)<+Jrk;L%z+t=NcDG#j8!p#GeN)xm ztbf>M3jC2FCey>yba1^1_?lMvvm;*fDg>(A#5vFP2j(Zs)`T z2I9S&#Z~TSp#T$b;}OHe3U|SITlyeRO&%L)JoT^*PgJIa{^~{F*0egW@#vaTPhSHz zsV3zL-X`B1!9mCzXNQ%zSNc!rG{a0;JkvIu)zOA#wXLCMc%6~*hC};Xe^^ac+i@)e z>`O)j>k}zv6Gh^H3pj)Bo-1dKL{fw=ueDk3$k7B(9Bbm6=N*~YH}2QctJX@k3= z*q~Oj*J;2Z1vsSgX?(2)gM!o>xEs-i4>+j7kLa1Lb=v=GlX9=qDd$uL6#CEZcR9~L z4P3(YLqDq>@pzBmG{2HyhvaEK?0kf^&P~c(_bmRZ$+gbgLZAsCY+K0c$=4()Ay<`4 z5@d(wC3*o}l{80i$vsQF(R*D`E7W1FBB&MxHX{GIaij#v`eGrNAimWnZCt>c2X*ho{DtUo%E{7`X|-$%i-C+ zcy;yU2IZ}VxxX3smzYJB|GMwn*MHIZW}`Bh);%9YUrpPykE{*0XTO`}SATH!_Epo} zv)z_LnW2GS6Vfm117ST>99JO;{6w!;hL78lIktRetam3XgEDD@f0l!mHGz*@GB`TWxmL@u zI_K!DF5XpcBb~>&#KAi+2RHh}LB}I*vZ2Y>Gba=(hIy54ePLBXt#_H$(<;S9C(}=#^E;D9g^cZ^21YOoU zjtR0{cT}QiVsZqkcwyl@RZeIxu)G7i2j+wI?fuwzhU7tzJp76YAVahqh{0{f4>gAA zH!&7gpZTEyS*<{?G@tPJcW;9&$sYXKA(CRp8z!Xs+sc3=x|Yrmp^jQM9QOcR>Zhyr&A%bpSLmm&r#lfF*RsEuvkT&EJuJu^9TxOR!CQN-( zw12&StxKgNmYkr(#H0D0)3WGXx>UL<5aF~Mgc+Cl^>Z5;wfvrRLpb(s55Vto|6`7) z$&J~2LXP-a4>;oliV1G4S8QM&D5KesGRNwRrl!Vj*C8`z1%+NahGpuJZo2`j1WF$D z5o|r|z^+6M-M|_2kDQ;XD483}1A({QAu-`-Nw5qOEBaABHqvP=Wj%lxEg$1Yi_vga zd%+4VulOV?R>%#eMtS1hBC_+P*)*3J8Kzu{0S?V9D&k7m3Gioipiz(NeVP38Twrj= zqT2P|hn%XV`GQ_Ss*j$%qXCG~0;`~KJ7?-LMO3e2d*u&~YSn;o@?q<-#a zup3p2Hx`mqc6P&#VNq^mHx5utE=6*vh*qd>L#|CMtVWsfdO@uO>)6)tj`6#AP09l9 z0Re80v`&L_c5U>L_;??JJfwTiH=MMKKWkmpn@=)OsujDB-)~|@N z7&fMkuE}e&^FPK5fBf+me=}_ZVp9-j$8mkL4MJ(|13w@9+9se&sFPrOm|DJ^0uvoY*{Ni-m>b=>6F zIr}5rh@oFx!KwD&q9~ovaI6p4PN+J1nypOIkG9jN+ zOesaQRK%t!7oCtcEDeZ(+hvW?M(Iw{Ia{@y-|f5?N)WWu2Oai6xk9rzLv)_o74Bzj5bo!3ziC$Q472}u#ib5$IETvjCrF$@!pr6yJp&8nx8oz7GvO~58Mo@m15vNK~_%g zPS2-1q)0NCzhXWro>xInS{HqmgFEpRb8=`6rz3QSf2}A-P$Q}F zP|I(JYQ5sQ>%GT^C>tPY7+GUN(C|a-kfbU8@bBYcjI!f+jTy!$-SmopCgnEoZo0^? z#7jfxPf^RWxtnJ||G?bqv()lK(s-{XPC7(#b}HIAwUQoLE8PK{meKQDW<%GbbDsN` zvdv%roF0(Ia_+e6yB^Px$G#QZJMK8b_)s-Gi%))A4`xq>w93H=ufJEC?wHs(Id+^N zFteAclj!R&PYQ-a`jFWai)UiNV@?68F-_owfbEW%!e;S465|F$E$uQDrYaL;JH7UX zs`NCen2^#0VjKNB*G$M;wui2BEq2!U{gZ&^_n*Wjz36QK`WnoVxvFyil7L=5#_g;; zdzO7aEinw1jbz=sSgk*ke*c3v{7l;b%+$|keG40MaKnm*_hN9q^e7iKln;9&^}jaw z^y`Phd&02eG-+xP$fe>doil0$#Vht>!;aU6uo~eRQWUyQc3jb+hz%|T>AP%hbh!2B z&yvX*UhSXp+qyFYqp_MC`(4{N-Qr~l4B@({ZC8FnVqTgoQWCN`(7g5g+mC;w`A-92{?xq3gS4?TD0ZBGTxO&z&_gkIDAGwq zU@O2??hf}lNl9QDjZE%Wysvm)fI7TnI?esevQrZ{}JRAyTq0aSgFq-Hl zWIXXWnf}5ZOFePiGEO@bYN_S9()=mr-$H}^B^yDJ>N5EJe<+XG2J#A!+Sk%gxnbfD|%}OWC zo1@XP3P1Vz-C}oxiE4Q7&{DFMor$tvZ;n7U`H19Y2E_o0&~7RMJM3#bE{WHHaX_nOdn z)9WG-)I9L$;Wo)HKp!qvm0J;bY+z+($r^FmdN`Z9^w1}!A^~=rv;<fGm2aDO45Kw zqlLnbB^>bx8E)6uXi~zra4KwwKHb2 zjPX#+d6M732X?1W%oN=gwLlF_lv|oeg-P!{^zg-6N9c3Y(=F>axq}WRf*gp~;3oH` z8*YhF%qGvjjgEHGLNFV*y0wnAj#dS}MTj2NEDb2T3*v9OKPuc}3XXP>oaD>?V17XhtZsi5Ca2R-}h+o2Csd z4~P}&Z%T2@wfG^4jd=Dw6k0SS|h-UG5*DhDpW?`9+ zR<3nI{n!l9wpPnGgeB97yhpN|b25cH+*;fL0lMjZBu3dbTgyC>sagnnOvT(JaX;ki zkMRrq6J(obRQsnZb^>_QVGp262bqFS8c)K*sG(B}-845}-6h6*@bsWVtMij{^a--l z3;&89NRDAxu~xdDvyKm17w|U$!ya7e=RW53gyn|;PG z>8HqjDxy=8OmA?@6sFU01m*@2fUQ;SvYiTTC@zc3`MrL9kRVm{yY{;#(de#C2qut^!r}o}qjsbsvKGs`JR!jyVqmcodO%xMHkr*nX z*13Rt!=+eIDqYFo2cy!ng(JJe|9Ub8cZiicY`@;@-+D`Dx)@MV_|~`U$-2?1Htg8# z$}ocPHj3Fof%6-I#5QUNYQki?Ep^JDZ11h< zc|T(;^H5Hb_?|cAJYutpv*Xl-8HS`3AOv0wsqvepEsKIuA5(HMoUut;Ha1HbOSaNF;OLC zhaF?%h!HmSQVcMQYp95|0Vg=^(oWG%VI8Ltx+=S5{nL-~8i1>&#si4~s)dh$&a(l? zKJ(m>w-If~Wm$vUwy6h2Jz?21^1|~%_Rxu5ZGvmOVw;ILvBQh;Vq!DX0I*|)7iv=7 zvMy6wV&*Fw3DtAmq)4Hvb8DBcoZl&d&Y?pd@!qOrzr9ZPq*!`@{^8)nX(O?n0_%M# zpFI+*DX`yNPTRKEy~dM`uw&1}Og7?>=nyMB&)v`84;;YRp4pxi0WHC;(6_CxwV8m`$5Kcc zn-H5J>B+#4f4Ly*9Ro-v_5A0Tq{5E3G;K!H-bgX^6sbY#&E@0@{r}ke5`ZStYi&RA z3(1Qi8-XN45G25WII_XfZB$vNje%X13p zbpPd?L2!DM2u_n-t`$HIvP<}Yy@b~#$#ia0?fGtV)I!Z$7l5{G&>`v7=70)nfX)s{ zV6~~bUV~ewesNNDm~0}Mf(mMfs4wg`{8u5Ypw7QVelp?%87)m~p^N z`26~#FK$`F3t`9RtqER;Mc{P%JNn#GJPyyiN(LQnLS4#LQbA`x3h)U*8Ljn;l^3#m zoK8S#dWy$YE{Q$a#%M8#k z8xo(rYr*sGUqN#4Mks2_llDkkUC`Z{6j2?Nqr5g})vQ79-ToVAb+FStv1Q6v^k-y) z0Ixn^YuFl9Bi$pu3#}z@aq7I6aZV`L$cnAc{HtFOi-iocg*AILGa=)-?;n5tx?Xuj zRdBmWogFK$iw4T8g_0eo$T4i)g~hS?yo?aw#=9?qt2URuDMEhp`}fso!V53 z-n=OM2o36s!a)b!PdL2@rzcfaPB`+U_$#^~?8>}0)zJ4eaQ&m9@9Az&3Pu*;!#osC ziU^dE)S=q{&@|Jy*5w7sYA{h5Q_DR~r*b#DC2~@^rStBFb+9u88r4mY6Jl%_hgY!O z+zN70m_yfwcdM51&$G6;E%nf-w$AL~W6Cg+<(ImD>8G0m3pHfdeQYaTU|bq|?+o1tMd zar!X&*T0MY{h!|X&B6~KD=q8(Qp{)jx2d~5pLgQDJ7LLl)(7Phbp>Q&7m<3;XI~FB z{Nl}lD)SYx{0T--owA8aZU3#_jJX!E_GNPIg)w)R8>E2Wr(}02(udx{0Xl1boe~)v zvC&s5EAf?MaxHyFc#I(VG^A$cv$0{qy{MQ-EDXSYV9KA29f6s2yzpp=9w(5s^>}5~`|3Z#WBR32z)eDsD97Qx# zwC>*z76G)f?xx{ih{``%x<`&lwYW9TrZw`Fuuy4=B0-!0vnkY_biN>u)d<5b2#-U< zlXSuGTg#n+7qNU65s)0%FH-Le+2qp=1P_?HUrqP4H88PpWa7 zbVzhPaC6AgkP>zaq*ubaW`wN|SJ#RLC1qjxuBli%&;YF~^LhJ~)!Ydh@0cS)%(PLN z4tuJgVJ1WDR^6TaTy$Ng2blANyJyJ;JKm%1Hb8zZB?I!COe*@Y*HUpUAT#UR_Z9%$ zazYNWk1HPnAq7y+Zg(kRU3SM3y_$JL4pp)qzXI2*$ROWQe7JHN+e#DX4j*bU}& zF8$DXF{!4@_%>Ba<1MZEFdFY^+g;;${PegnFve#5*|CB#vHYd($id#vZkDvVUUBLb zYMjpT>%qO$!AY5!schg_EEkw9iM=krX(`g_w=iRhUhycm@BK4*GWD}+)KUpp4A(tn*{=M*50g%tUSo{+K-dZH?@4t9G_`=JgBrE zsL3K`HmV&*CLw)q6nYXVSv*BnQ_)zERx=$-*zXdkyx-(=RebwdmOwmNZgUXyEE>i= z`Hgf(kN)vXwLp)FOm0dCiD!lhJ60=D&NvDaX_O2^%tokRVqFy`&099NC@e$xxqF)8 z+MF(^jClyri_4IERS;IjTFmL1n+HVZZQ>Tkg89!#&6o#}(dMb~QrhM`w0O}H<8%gl zjZP2~2?|B*TG(|Igndx9VC8FT7o7Oc#i&1a&0QH@B1!UIwXo*(!SAPeH*i)$sag@c zgLQ+8ZA*36d)@K6Z6UyXp1dyaS71F#f})!7+}FI^}z-jvsqhZ;!-oT|ARq zv}0%4BLfTLD@q2zr)yMnM#yGxXJI?^HLhJmBLT~^jx)eclWDn~>|TP+1$03{C@O*2 zQ@j9KpW|6|URaM-Gjq`4+MGrj#{*@n|0x=h>GG8E0-Pxg2*vPnn(~0Jjy?}=-gHbC zRP)z*uW>ybs*$V$!p7ksl(dka4O&kl!_t*o({BiGfUCBfKk`|vz-)%qb2y>8Te-&R zured0O^m0&J{CIU&>}tt8KKB1Q%yf$6^>J3X%)ncheBr8#=wpl;`(}A{_DT=h@1N8 zXFf#x!dSpd3zM`UQxsCLCznzK5ulE_(^j$|RZWC!c(pbw|8Q)=w@8FG(M|qIWM%G5T(i+sJ7AgxcY(EM%$C z$h4B}!Y&rhh_S0W7HZ?|E$?$2Js8K5(SRiT-J+oD~<3!$|{$Fo^3Y@28NHMi)s8a-^$ z>Kn~v%~M8-je0QjMHa1BYJNEaP&^q2&Qr{DTNAkVC37o2+O)`zDS6uFMoUef#aio;EAncu^QQc+X zs0zVN_vO4Ic7@=o*aiZ~JcvzX9a|M$!;}5^rS>4HUXpA`3^+ql>{ybN8%UBIO17OM z84yT%9lJgw0r>jqCG>rH0|!&|diYrH{iJ*a`Fwlm23{fLHI(t!02rFtXZQoa8Eu2* z9=%YLU;5FrSpS%jbfK=z{N9IddZ3(p<@5(6X&NatSdT1929CBgD!N0u*MG5W;h;ld z*uqjtxzCk(%}@-qO!eu^_@E+5ziTENlZh?~YutMwt*DmPs0Jk$g?H)2k@p?MFi5yj)4~Pv&aW02|Qv)cNg?Qt9 zKrdN%Og02f4MyhJE-VIC`gNR5q@C`XJLu3Sx#j=RwcQF7tdj@#Q#FspvWJqI{mUx3=c3lQ|$U zZzk)GG7Xa7x4Zs5e8fn}m1K}T>x#Ip?mExp-k`mOaWHd0xL84 z)AGGmLlEKuZK)cx;PzIFNMWnXXp&YjoMFIUPcm-Guy^K+fLRTtdx z$-`b5ya8vVxcQ8ITa@DaQ6AML*-3qL*sD$TWz_j^)% zNKch3Pq}Fcv`hnujyN;mrw3BwYwIdVmK|%SdIQjuQ?e3@6jIR}Sc`Z$^fH2+dAmZ@ z{Xj(oeN*P|0+;`QO9H1?oGR)k*fawAK|o9P6;vuPSv7Fp)7^ejEQ1v;bzbdG$x?K? zLVwHZp#92S%=m~VTqP?)WWvv4GQc!M#?drW+TVRyYMdy+BtO@V{kbO6DzcSY=qdtC zR^XyJOa>*u0s`%JWL?l*Co#NDF(?`7j)Kjcz?)k@V|w=xG&sQ)O)-(S&E~TVRZMWg zY;4UZH*Oo3(7qrz*=it+)>5)H6j@0{V-GImoIz_w-6gG@l{M?1V$_(02xg0EmC@>D z5B-IDg(LaL8*OB*9V?t&25X#7$-s`?L_y3uSsMHD`k*3qJ0Bz89nuo%QDvH<)DOs5 z%lMct0Tq>PiY`T$^Z^~uS%}rD78w@iy?V3Oe2j@<-rSaqVy0U{Uu$$R&eF8ou#btn z%RCnH9@f&$;B>2UIz~1r@u(IubWf2EDR#%kdpd@fM#wmL2H?O2WDXEZmfiek83Ba) z>zr5Q#>97Hv+wM9H*BKW0IG~6pkVEmBtVKnvC9`8sa*95?iqe37n5AR@OUVM?3z2l z?d(`N8W*jev2u(bS95Db37o5}f{=Yco|qV>8)~ZtHq0^fY$T3p3>_)z*lip~9*Y+{ zHc3symg7P)LT-zixyX8e+nYL8mvpVXjIU0MEDemAi?z*pQ2IRJ45@j|j@yOUPV}x{ z+{;*#2;KKY*MpCSB#CXQnJmPL`2vLDNYhyP_X_)IF3!aVq+>wI27 zq1U4hICDOO>f|AZ5}@e8B(}1!djCAur~G`LEft6XwwTI=Fi7CXIL+tDHtSQNX`Z}^LtpCg2P_5+`SRol#s+Ix`%blq&@(R)|28|dcgIJ zJGM60s4{2n1aqW(oaAcjtp(1+94C4U)w;jNnAuF@r!4lc%ds1!+Zk+bt8k zQJCG1T{~8T+1>NGB1&n9&l`}xw3j&O{W-YHac7HJ0qS&#zG2s|F7=@i8D&@!c zSj)Uyu*0R&uV&^g=PSW?f_rGZmZQvMXVWP_zSAw~f<(3f=O)QgadXh+;6{33s#~Re zAzoj2du}6**((ce@qJ(d`!Oz&kqZWZwa!Vv!6nmz^&a<3@2nJ3I9{%R9lJk1HSoe$ zQ!;?bK1?3L4l-?ujjRg)Gzh@$bHpr?EG7Eflbv!!RUGhp=t?lO!*u@`p;MQ8RYaiA zK3xzGjnsyz4|P?qd)Z$>=}%c$vUETg7oy%PYVbrCNiLMDRkM*m5jz1`0XJ5G$p~B? z|Fj$`rlsSwx30{aqX*CjRjN8tFpV5FSd&Uh1~kfhsOUohJ>iXlP7*VtZC)#M;Dnfn z`)U5!03gJIYo|OKrvuZdqpf?3*8K+S7aPR#(t%GTygQX zZ|LE3@6B)iiqzY2#P5;;u3IS?WJG>SMW0x3QFt^Q<;{KZwK@G^E5oroXV=S_E`!j2 zrk@;8meBQ_e&@Tuz>fXIchGwyug&=)xYWOze^S=Wk9FG`c)_>CZ;QW1bxRs28<6f3 z4*91?ZV=S4lGt^k&&bG_YX1@%?`l+etXgibFo%7OT>b7v;hM;f-~n+WdqBD~ymfd4 za+6O&9$6dhk42Ed0rT$wGqf+vr*;}AUo$a+>{#8I$RS7lboZ5#fObBpHH~s`Zi~lN zT3s2iEJQ29!q5cHY1VKaCCFNCPw*potbD(4$RS;jF0ggi5ev55HulkE{%dWet6ui} zsBY_NvYuJJv;F!W3FN+wO4Z&@$)KuiD;0g&y_L=YPM?+HuHhV%G)0Nz1c~9*yfi{# ziDma)(DMO}qclamPcjf)B1PpSWC_D>`6NXT!%ty8{Kw>txL@i~GIx=7HfgtG4fVia z-Fqn67ZmBHqA$5uLm#zfZak|j3~P_`dHKAYkem?U07pjlTsn!B1PY~*ZV?wioba-H zJSU!YPMYai4v-hdD_kaI(6q+)y=abZldg;5+ z>FOA{HYZJS>l>ACRlRxZt-PO|SzXZ}I?hB3=Eaz17 z>jO}nAVbwN6(>xfF_|Tv>@$CTd-p-(up<-wV#kuiL_qR$`r&Y=yiUlx-loiD<+0+x z3%n=dwmgP+AUvDa6?ts-P6rFA+`Wpk#jO%5yNAytgEsBRh-=6MCkAq4vH&piIZ?$M z{qO5F&We@s5v0eC?cUV}{7#Q3*#JfEQPFqA@vKzVz&Emot1~c9a>xNU!+5k7x(cgz zPrpct!?(D`2ILFk+;ZqRnHt%blA+Ii6EN;%&D!a|Qk6~DfInx*K{o)BFNYtv6yBvs z4=UuJBFlo%u8rZf&&sCPa{Aa;=3SWwHtw8EGUQOrPJ?3CMLcznctBd?4pcvhU>|4G zY2IoqH0l9o4Vd4_(nNNecPck7d~+o5jk#8HyQ5G0=QtLGsWVswtg6r+dK*lWE*O;L zIAT%-GWcdIan1%($LeFJPuD$qK48GbZkrS_vV*7O4t;#{e-}qH?rwUdXGFE&8Apm2 zs1ZT_#Am9tbi%4^!Sh^Sm>%#PE9fOUO28Algl% zgRVMA1Fgs3!`*PX7jP3w>;2OOT4^K6!OrGLE{K-yW%u3W2oDF?IEBM3pct4YPfl!z z6K37Wr+BVmXcLbuuqOobxYuYzYdF*+Z0e_+9kch zj)??{Zq<6fW3ry$G^P3rNs(JNX9~A!F*Y@9y1h%;RH%F6T6OuZp5BOY4X8mhOnN=66Zr zgZiBhI3{rRgl7s$q$?xui)(?~;(Fj=cGb)})+*H|=ri3&_ehrn9)+j>a#{Kp%z5hM z7Q#Gm`1gx1@09EZXZb)rkPce3=J1neBNV=~D)j_!_S!fPUI=maUE#Xhqu45q{}m0}|J!0<6t0g|6slfUgPD z9i;;|e+g^VJp8#UbcK_8j~p#UjVm^>94|s^**k!4tK>Joe_#ElcYgP)5C2=ZjFK&( zNQ{9VG4I5pUt`R}jMm0}b4FtK-)V@_8@Jy*^pj#zX~)LxSp(ztQ%ZJ(B8REy3P>Bs z2*K9nxLc`m!fkSj3m_6&KcJXd?Q%%Dm<+I6>D4Z3JV{d^J7GL4*9m1zE9`lO9FU%; znR`Qc#1S0CvC6A~Z7u~Y#Ix43pXe-Q-TauOPy?RcAqV^%6jo292qaM++92Bw z;i@)PE`0(%O*f{Nd)m8(6w=-Bo8Wy2lLu^=y)5v6a!Fu{sF_g}o$c{93Y=ykO*hCi z-a=ddyx6$<=>-X4nkeGh;ezBa+g;*0u(5;W-)!D?mqxk~Ct2w!_YL@#tC|D)bYie8v3na>QwQ4J-lq(}u7-6d@as`A82rc(EtbTO+t zcvCqItSRi6F8fc z*7@?AwQ9`3Vb+vZ8Dtt9Q=KwSzpe*I?D`-4l+?U1+Na$BHYX?vDFpxYoIyW+)Q>^xE4~f zG^)k2cKKCtCpS$|5VkYClZ$hl6c_tLe?D+>N2B#SEz*GHh0cS+9|Dk}} z+*a3XQ1YksJOtxwhf6&l$q%tkdY|^56hfI{?WtPGznK`*nK>1pojqw`tY$>*_^$v-GOxFMsW- zBs=Wbd3V?V0Q)G}ZVGziL|+QOC`@7}OOHw}2~+*D=trD3RfA`X;G}BM0sSg?9>W`i zqQ!+LRozhk-A@|G20;Z~A27g<=UftIvN4l)(BT$5iVlS4UKfNSQ$#H30XeZ$0OOy;s_n2c8%{`-xE|NT2XbY9za$b)>r%xbY?NwdNrD|mpC z-J{5DD!NV4%1`rd7O68`9)NwDO~(dw2A4sOZ8aD2e>=Gyo|(XklkSI^m?so*vL2`) zQCItK11Iv6KM-k!UN-6$aakDVZ8iqUl5`EqPNkzpl{b=a*4hO?f8zq-UbOBCsnC#m!t)v67RYB{>rO*fD>f9av4@k+J z?Q;(Jr(yaRD9Wm!UQR(6PPlf4mUIgalUlkU?6_l9ke1tkQ4e7!cO5AROH$@|;}sw( zC*8_~Z(JpdSaqyh&i%pACURcGyhW@>@@{i{1=hgWXeDai@k9ulGNqvITbzk)Y&+gv znXs{|oG`t(;N=oQVpz62R@&wX)9ETtYg$DUzun;x<}JOMlUZ??Y2#`7&)@yCaZRP& zrrnsRlB@~XC*0x|Hyg^6X!RxcMZ8k~%it5Ppd0uzvbOo1XRZc?=vGOlGkiaHciY*U@9J-<&3L#ZV{>6+k=;PJwsb1^Uz^QU- zaU6hNOJ6vjj5sHXfgFy{+;Vi2m!%nn1mjy86BAr6kXe zrBtJVEnGp#_E4mZir&ty{Pu=D&(*!mxe!mCR;y(FKFBb)UjtqGpV;^YZp zMjcH~nT`9DX`GCs;V9J`LtisaDYaYU$3(7clMwiifU6TqU;3SIICuGEgX83~8;k*W`|hUOwO@^B2IoxeZ)E>K}z@J$RqPr9Uwqwhqi3%zfOkWS*Bf2PwAXxS6nq7=q z4l|c|)DzF9<8l$6w)<%R_3;cutxkHsFpB zl~fF`LIEu!i=Dcan4Av1u6gfm_*yRwDSeXIfU;0@-&}XaIJvG!cepS-Te;P}nSF-8 z*89FZ$-CXD3^@06x6|pZ?q7xBW=`EflBM{G*b%v(poGE?iJ3d{JB@sHD{x*T4I35? zZ*|At37^u9MJpVC=X6M}2Zh=vxf6^x@R!04V`X8N-CM-j@R|77ArD=rTf5zFUD>(V z-?$a6-3q0rG?yK2EPFDdFcf<`0A8N8(m~l`mPC3sLdf7#m`S9dMdyzhrxM#OB1|L) zW6yIfR2=9C@ZqaD2b!HjRPsFxX6C!5imh=Rl&zF~Zh zjPsbyFPVmk_pfGrEZ2Kretu%gNs>AZnpKRB6X#PhXvLTXiT!hje208H$xVJ`RF5Cz zLQ3~Y@chcWhr(rV;+1qdpWPoE5A242tET(dne0JiAAJ4M^+402b!{}fCF8cR>dgnI_KYUzQYNOxI>}h{EEB>AD zj|3IE-@igWWfthRW61)n)T44E&Qr286gfpjqqilU{v28I4)`AzCPh>S-406?6#zHh z@)_zQq*s_KRv(#n8cN}yR5?qT2ij-zAcTQKMakwK0>c{s(G*Ji8@W5<3v6C^zO zkW#*zT$`<~<77i*@Kg9+FAJ#;P*<~uFGMs0!gZ;2RZchj2j{9|0=1sl3qcE^&^{?9 z=OwdG$gBA~=@KxnbMV)VaG|@G?lY2Jsd_a59d7v8ZC%c8aTnn8S@tP%_ zBvaW1Y^|9yk8pd#Z7v%}tr#QJjC#)2hwM`)zZ0ZaLMxMdi^(oK4oMy}(2iA<44CWp zQqjjf(uL(h-QR7hld5FtX;v?!+NF4OD=+wN@@e<)C0AW{v3Cog92d?p(@v*9Q=H2o zt1elJS26PrdTp?dDr@O}5!(BG!2{0NzOc#(6a647!yU8itc4mzh?rP-SS_w?T8@lo zyjNZM$@%9!EScoJ*K@reH11VEat*y5RMBzeLAE-R-6P)SjS1u!Z*1TkoB{F1)zWG{ z=7=NlV}ady(@N=ujXF{)6`;RWBm2E9%zhbw(Oh{Xr{is|BSOS zf0l`6=PRYPbT{;@j^V9Roe*3QT$CR0Pp2OWCn5l-<42hb#G`6_Kgc+;r zk+Cn`;0gm|7(8d|6K2S$kct0j%oSm;zD8n#7XmXZ&j-{g;zQPYWBo?~aFm~vUvXX; zxl#1U7xRJ(B~>$bKFhrf18{)6ur(x}%+ClM)CcJ`SIx9RFj&|2%5KsuUd@URY7aUZ zj!ly_sua;KvQ_|18s}f78@`~Tf= zpV`lQMgnrTA`v5`N#Ujh3r{SH3r`j^WXXpChCM zl8CTWBpFJZlf2cL&VABsI^C}qlI}7>)C0n}kaRlN`5D?LW;`7Dx zN46|KU7AfQm<0>%cQZy#8*FfnP%_|;I7mhBef!QEgOcOI4)L|wn534bIO2E=w(7^o zGUrb29k_!zaEN`>b%=AUu*NSt3Q^!8^hp5FZj? z@1*h{J&5}2$N#gGQz&V5-SETMclWZ(8PVrj{$3d@wB;YhUthL+Og88H;^Ial&pjzg;aDpy_0uaa!PcXz9&vov^y1ss$*YX8mQjwe|FZ8!!qv! zio;&QzLHg7i5h?o~}D9zYdWujc{Sjd<#V;_DW`<7_QLwPm( zq3{B4%`6lp<&u8L9%=G#kXu2kd+(XdZAr>eDODR_4-}TaW*xL4G`D4&c3bP8oj@FL^qevq&_l~lLo@*&fcb>yOr$k0XSMRnraz#^#wRUB+r&c1yb+_3hR2N1Hs#8nuXY(1quOmU ziiyJ0GWq@)8(v<_*(2AAAo_>w$^V=Zck>`&ykNFJX9kId^L;-t_T!J;71^;tW5Tz; zn3pbq-T_b;m`fja9CE1P7xK|WL8gpn7%A2*w|P)O31EA8jQ#}EP)QLT-eF0afgNw7 zO{5tB`GK@@?(k+hcg9jNv=ES-U^j4bz!8@~)LI&wW3>UTlT(S)v)Uz@4;#Jm7`?VO zJWf(AeN&`2TKKp4zb6THY_t>_7%dr;4E(OCRP-9B96w|VMejUFiv7x6$cYmdo~GOz zxrVdWL!A-QAKVd~F932I0S*Ll$$-q$AOhsQ6?+5?e^}x*`x9Hw*EXRS?uu+YE$Ya|HhA7Z9!M(q4 z^~am)euDRsrJp{7o}H6Klnn)lZt8Dd{oNU39W&PSc>Xj=%v`OskzfQzSBr%`r}KF? z9g)8{gN3_~G(}=q-?I!EGlPK5SN`z_m=-D@_stG@Q!hY%e=_xLa?p--OoxF;_?(h8 zQ{*#jw^_&9?7a}$*>M)8DUQu=7dDG?ofgXyNRI!Y!;;7{$kc9Atzc~zRsex5b~xJQ zln7#C(4m{3F6bj|s$LH8zv=H4h?dPg#&1)_vz}aA6L#pOHq}013@`ifxUBhEP=|)k zu3)A69E9@prf}UH>3*C1(?}t^*ss@_sSU#nL3*7unIDg_Vcvmh1dXGXsEm&mzdKWp zs-+)ztRd;l)RG;0{HqKwSVYO7m@$uvE)3i4Tg%-nXp}7BWGIe7N?QSY!^;}gZpmfv z!7S%IAX-U|cNcvr5YqopQ8w1YU=pK)?&>na-ImFq;-Y9x9B8FG>5>w=uxsJ&c zA;0`eD=afk?RIhq@9wTr6Bb>daXC4Jo^^% zm(Re^5B7mb^Dg99b9*HD?5nPIoa4$=pj+3ljw^8#_V0@i8k&1RnZW6jK-X69O713E zs?T1KDLKk(b1nv~l2tk3Q;^aiYT)F~iX~^~V+D1hAd|f$aPN%0zLkDBx|QxHIm%3S zcW9y@MUWDCY!-5};B)wQKS;B)^K)kvh8FTy$>M`@MGr_##E{v1L+0U&fdrapsF`T| zN%pBMPd&braVu-cRy($R4j3S27bS!2>Rc+i>~)MZ*N0aHt(DgfC!tgYb| z#uX8tDm0?bs8Z772fd{|cC&BH{(SZYh#zInY^3q2U8IX7kvgxw$iB$quh>fXSd0KO zSCr{6W>(FQQ~WO(=P{T_0MHSh<=3Q3JfB#J1pk?D3 zy?^`Np@xR<{vTsv9VU4Uc5LpNpyBC|-eNV2ugptR)XW?J5|uAJ9tv*-e#S*w?ZJgr zP7P5pueH*N;Tjc=#J)nw(zCKfZe{Fyuha+R(w8MUg5@DvQ6KaLsrJ7U{A660^bn_) zzV`~W3Z0XusB&8CePy1t#FF`6gNZbZ{=Q}CfBf{U_Zt0`dPQ_J`jhpf)Q%Mq?72rp zr|T(M4Mi%c=oJCGLu=`D|2UVv$kE40&vZgiMz<(%8(vFR1Y8G&wUHoSYPs*(OkzbXm}mw}#}~vCYzCATRb)GT`wlr=mM% z9;WfXHr1!Utbe0?0ak-x{IHqZ#p>a2@!uCvL9GbL^}j8upf1hK6jV@m#F>Krur}33 z>f?;z|7R^&_TziMIH`i{1zwjhU62t{NauLh(igc{wFrjsE2%Op9A(Jhw52kba(Z6ZE^)RMW7+R`OW5>!NFKnOcAteIZwo4|txLAkTA629y; zOqzn0L@%!*m=}JK4Gl%0RmckH%!(ZrF9$b5?hKkphn;i0FS}a{QU(aH2EGg?w9<)V zT6LSRUflels9`PHZpY%L&OqFhQ8J)<-$g}(b)nqtyThf)zs|KwiW!w@ifePqC5OER zrE#)&!9pw$IXnL}|1c?HAL8_gtDIIu#)T(_FT|wXBKFyUyila+KNNu8-89AdfDJEy zN$0V)INr5-5xq5VFf4PdwEiR0Fq1P8vH$adadexBW5#}6PQ*kkyx;kdXqRhIBs5~y zhQ`UTGgYyd`T(R|ofEBgQD@W3NI_JUXAa#N4C!_`aQe3@&jIIr0R|DUD7q%#E)QEh zBm+Hex}axVVm z(*aM@1+{*Kd?ePzvN`j`menw5Qu!I&7=hS?Kfj*#S%FM1i<}?aJxey&u`JqcAd7M- z8F;KSq2moKaqP?MgUY?sBgA2)exE|nDVHvE&SW;PfQI8a9wvHZ8 zvWFD8kJTNxA?c9bm!!H~9IpJT^IyY0O&Cb0 z<@XcV%QTRB|2!TNXJjj@;Y-{~XK-*SEB*3#J9!sawaP=TNVTNi=z;NvA&1PFo#3m* z=40K#dAu&^Qt?o*mfI3E#Mj7LxX?o|3h`1a%VL=CUAh&)T596#oRTbVt@E-t>~O6cgFBWY`x#b zpmvXT54&wr$Y`PMS5L(KaBHq{ot%ksw2Y8^-T~!$zhyz#X7zwly?%NRU!zDsX8|jj$LI`ppQ}a5~Ey?q} zDQTix=yJ&pzZQ8uWbLFwIVKVt*37&Uj454~OF{GCZD7fob7C%vO^#y(Z_2MobgmI@E95*%eT#lJ9lFy;7OvyBZkP`YCbL>TmJu+f zewbVGzbwgb)i*CeN^oJ+W$DeRJCayx==J3kY^sGV&c{8Ddq68RA<_V(L-s_u&piol zpQ{D6v_{qAw=>{USP!s0#K|995fJ~I+VhJvqW$K;0ZL%_cCW+xafWoMAP#nX= z`(oz%onV;Ru$W*TYu_H+_Um7|j#xEyTTheqFHCS82opyo=xnEC85G${MQ;(^cV8mc zgpK@c6K@eMk>BFn2jPHvEzh@4G9bOh>4j`i;2Pn8t3lLAr%ag4yWLg#C34zSWeaXa zVZ>6)z2J-eW!w2J>^jj&NS9dtT3giCzjm);xE zq-cW}M6wh+2=4^SyV~$8vp2ll+>MAd$^}wu|L#VK^_Hu)>hXHCyZNUdBuW# zjq_aa-79z)^+6KV8+G>eKQ7!3tj`IK7zcxFK=xMm+s=C5-1W!MQ>+g-{?ES8{9lr6Et{BQ(y)7`k>Rc z^e3BGH8U|};%VDm4m(fYFh<35RkLei?e|0+7h;rK$0ww&=3~$agWm@@h2ZKY(X6xjd`8;02x(je&;Vs!OB?Z6UO zGjGV@bm#$awZ88%xGy<@XKKnfuBOrLy&Z591jbiER6I)vaaE zxnb=^30e!s?XBma;z>Z-4C; z5+*jZ71qsOoGE9A+^22;sGV=r^p?skcFPA8Jy!z>q63$>Ucr71ae83 zH9woiq`XAZ!^dZiktWjOlcw0t)`A}wS>w+KY;nY-zg?mA{w?AmhYhSnyjn=Y+Uws- zBP%eZ+92U?cOY^DR45AhePP3tp80}oWi!2x?ow3Kha~HR>U^pc!;$wa8b1J1D)vOQ zhED%~_#Mp=+w8l?nLhA=j!G99#VRQ6fzH@=BvtywJnZcu+wc99JIt9gX z(TBWpMOB_>=RXwIhHF?$0*6Ac%&my1d*=&xER^hEX9#-e0cqDAi{o=x**?oG$I~5Vt4u%hA!si3Bl+%E9;Y@ z5p>uQ6G%%~R#hwJESEW`u(@$O&M)fYmB=!7x!J7^V!{`QhNTvp>J78*>Uv#4Nl{R; z^pblu7_Li0x}*Z<7*Ky(l@cjPEdv50jE{It3u zDIvT4(-a$775=cL#hl#LPAh;;UyY>uX^Oazb0REr-4c>0SPr4vTDqBw)IPw^0oeeS zF6E?`YRuXhT^Q%}Gr7cQsq9xif8z0Wh_R3E1<^bvJbEeN+4`P)>X^uih-`=frVAQ4 zbzT?>vJ|aM5QJmRMmN)mEgT>6KIwhl>Ax}7x5nw_+|zE`Avrdlg+-6KbndL0d1(-b zN~a4%JL$o>Lk{Tlz_rhFhpeS%0RXcXeYC`~a6!#^oRLVIy6NyIYGdLl6O2vy0<0WH zrGxzYak7k%J@UBl%}~WPz&=lsr9f-}9*3HMa<7Vr9vb~xZ3?U)#In6?mun#{tZW)n zpP_Kst=tQ#c)Fd?Cza3BrT&(F zZ4M`l1BOgAkNt*1AwV#C&yzsO;wZ9;ioPyy<=4zgn*o0mb*yyyQ$>>ZvyeAqFl7$1 z$2{~5E_^-Wi$8nl4O;E|;EQB4vwS4`HM2;C0VH-(GT@BKhB%U692(N;bejU(2UIG` zy|i=zwV1cm`5Lzejxatj;F3%CvDGOq&2*P^?W}Gnv}~sLho&i3P^qG&zFG6P&RplW z!e&V<=Agw`)ES)H2sv}&&wukhJzC1Vj!Q_l9s62i4FW{>DA{d_+=9wmSBSfap^mnK z*2tE!+MRS|>}6re(sEsFrGU-~UCl~UY;ok|bY%oy*lRbB{Rg3QdM>Ak^4rE4RAukyp*%6{@cG-zLL<1Sb99 z?>FFj6KZ+Y_5N5otjh&DD|gW@yN z`T$*3Mh<<-^|mM0a-=Diir0#7haF_&uKpP)tjP%+8%dc_(=sx=sAZ0EKX%fKr@yFA zeMnZ=v3SZcP){2v8Bq1FqoQl+>w#5H*Mm1Hn?Xncvv+7-Xt7@nXSK&hdD--S=m2`o zVh7Kt7d-JH6KG64oApL8gNECy-`F70LnD30Z~s7&$Lnuz#|AF6vmX`a$);rBYe}b~ zv2mJuy);jX9o2?OUPf{Z@&!7E7bJoibih9&mHajH)V;#w5Jfsr;60eUA5e9NiH~1mb(t8^ThB9z@h{u z7H&T8nPoRl=1V!|8)SnEroWWW6+oKK{BgQaFF_>#c%zN1wPPb>mw{+Yr)0p?york5 zC~B76qAx-lt&88jC`6vjdQ!)^Crni4`_{3RhiFtlcG+96tZ1X^;L-BTn()j|1;jP zS>k!j8+Zm6Ouyvk%CboN?tGPT_>W2W#*XEOi74i=uU{0#P?x0@z?`B{oeex6fSkFP ze6^mpJ?mJQ^@8UbRWthxKPGTw^a&W9z*!8>{pF6N&+SqumLPX>R}q|GCk-rjssvzj zs))GhcE&wku!>Z;K@%Es*IGu5S+Zu#cK2x)tn(eoS`)d}IOWkqu3!R*XLXA+9Iui= zsH+*2^hvPZ=7{f^XX|?!{&$=CI<51L7|im5NR^$nuk16XjI>*8)C8C1P9DTemW4D+ zFwr^Ts(K214>wj9o5HO53+0KZ3%%5IWN~)y-{>M7)pg?-Jwko z>ruDzx3hEUyX2;9tGjK=49&#x)8v3noMOyyd?;|gN}=E0s*b%=NivxQjO|zn)fgDD zC6ug?f_nStPVSYtgObGX15t~>gzY8G;$pr=ln}8=x!iGxpGMY7fl+6{Md9BsO9vvi zx*sDc0@xozzboWUPMY}%mT=q-EB(n5k^;Rqaklk>Jx+hSD73nbW)+0l!1FrmB7e^F z&IX9B8M?Rt0ii-i~Zl?k&7YOoFW~K{HLIVw5%p31@Nqa)A{`DI7 zbun0o5y?L8Jl4L*Y$5D6WzAvntEc9T#O z+#!cyULHsfurZ#EF=sx$ji0pb`TDXm-+NiFh8zUSy(IgE+4USWP*A%m8St+ZL4)6E zDbOnK3+I)vyQ!sPdAEJKqy_8@*eY*z?N%Yp?7<&pt9S}7S+JAO(;A9i-8y+0cnqf}VQT-?Uo|OiYks#@J3mE4jCvu@xG6gc->TGv2nT zA&k*sY9cFNzf<<+zZtg}#aZZIR7u*{lu(*QL z!Nw;xdym}3^nA?8wi0F-Lt;WLV;yXm);{BG=&0EqIoIWS@$&N%Q%;gpX5z(;{Wtpz z#7jOU%cWp16@AL}o(qs@s5+#3Bep4b3&A-o**Ei8c)Cw7)Zk!HpqJapy)13vW<$%? zGX4ekV=|2#y6}bca%<^+V0(gQ(UK&km0&d9qFV;7@n*EnwZymobjH{wX0pk)W3ZU8 zmQz`aRI#!l-&D~l(N@o{xn=Ya7{ddR`(}2{?c`o^U&O;kt^XL5q%0U+{o;tDYHS6XW$}x^R;m&qzN_{hW2u%eZ>b1Qnl_zV3>|NwAN|U}c01cQ3^fnzFDq zMJpe+8x_Lq(oH@?!S(*}tSZjkF!hyrJ=s;^$5Z5L2QQ7Q!q=>zjr>+cuj@9g#$S=i}0SO%JFTccZ$RS&a*OH}e zicW5e7)fbXM0Q~PI#AH)8d$_QSxB{D3F;mTQ4Vv6V`E8loNu?Lulx9OA;J0q=1}x$R$+^=JA)pY@q7U z0s{xO9WEnf1R03VxZyKxK#W%vTE#(<CL@K(7y~Qm@iQVV!va&)4 z=n9vsq(gdk{?@?bkdn7anZU8im)-1@F?;=OYrTy5eNCG)qTD}!o7N-a?%UqS$YMLT za-o&wC={%tWNRt1hKf!SR7A8Xx`4mwoT%QvilfG6F-4N&uRPnu$G2`K0LCnktuL4c zz&QTfsjq+BAUAHk@Pa526O9@^Qw&H~yQ~a9u6!sQrburV-*ns~*(Sq)ArNFBHP~m0 zo&NhnLB$}c?sZpGFyIwPr#}}AI70^Tuwy18BrZhV=Gvve3A=+^#P$Apykem7u_hGF z1epl~)#$PnkteU;bKQ8y%mg-e?8!B;zpjB+-dK zKzMQZR?jQ*Qk=0uElE%|PoqlmzAWyOoRH@!Q{9#XUXj*wQn}WIhVg1z1Qp}Sta$!j zXU&@DUF~!OpG%x9ONslIHia7PP!t0-w)L2dy9j13^SIC2@**>}6Ugzt7@@aCobNOx zlfBHgzjiD?P8wJ)jg+i`B6U=B1yl`n0S6%d-KIJTow!kSR0SP>B_^;z{)n?eS{7C= z+(fzqA8~T%MZ=`Y)pQrFQJrS>(vLWYojatxoaInabO9Rbrg-djUHrqks2)G;Dx^{E zaH)tW_r3i_d(_8=qAm!#B})hvAFYDoqiBHMW9QNhlXUWUK5FM5* zIHPCRtU9sfdCzajU}-~_b9(7r!c0M$;*uNCllt!qsGzn)ltAOKE=ddz+CWIwxM{gn zGxL3``GXGaLD|q6LZiAQyeRFZwcLE)_{db}HrGS}mMU!Wxd-hcPLs9XbzVho+q`@D zP}~r4R1DPY)E37T)+%IR2njPJ1*4TWIhc{6qP#!f`0f8)hHk0TS-nhmK^mxS0eaqn!cD3!O{E0uS|wyqdU zfMpUDR=9CIXnK}$rNCG!A3Jv3o2Umst2lt6pe5EqpTZW_ zNpQgw9vRTgbm{j;KZNAeqsQ4|Q)}n{aES{~2i2n1LUB=-6kRpgMD3<@DS+(eFzXZ4 z2+dMslzG6V+3}`yQ3TZUZu03?jx@|kmM(X~fI^GdM&ZOsV(YO)gafPN3MYK0^MluX zMnXeB7cV2*#tRzSvFTE4;B+aaWJMH^H%IGuuNU+7$hSCZ!Q@ac0uOF2kc1tB9T4 znNnCOiF0&F?@LleN5Rk6Bq`(LCvIa^h2r?h*kjg1&uCd=bUX=QnMTie_STP5IzBM& zk7&1Mjfswj_5Q#l226w4=&*(ldnNc;&0p)?23zLItn4IS`jbb@g~gL6p3M;>SWNW{ zWe@yFFK7hv-%TOS%z_1Wthc^25KEUR8OW)#RCJpnfzt(c%A?>r!Mg0X;m|ZT?(2sB zzPKMBa=`YnYo$HXe$ZtbSQmw{9PIzA8x9GVn5+r0s&e_>ZwjbpSA)^i%soOD1!+`hOLq=) z_Ui67(XFnv^eR#rvBxQmolU2B>~JXv)2Pa!SV4ESfn7$gBMnfBy8}2LPFs&=*4mJo z9Dv3?Wg1d1Id!UxsjBSO{xQL!g(+2_cG3me!r@e^{;)qj1>bwppGgv(ko>=3Q5u&A$#2S+OkR|y4vG*l#O{M4h9`S_a!;pbm`@1GNIp+&=zURE(`@PHa6wfpv64+`7yHN?TO zKfK29)Bby#(1@6oo38zu#E*lm?a*7jg<{evvYv`DBx(%0lK3bOxGWCtqIbKsD^tfO zaT;m8Flyek91lFPz|nn=gZF~{jUBwwv9IhNT&L}S{4{Cv56m-AY-(tn*rHfSq@~VS z?V(mIT(}U|qEbjYs1^6zs_A7RMZ%TxO>*qD#wHV7l~BM{2La`(Ne3t4=WL~h-wd32 zg%K)<396MSsApL8@LPMYVC<1U#FPwuYv%{yYulKwyy&5tixLd>FNac@gF*U60$gLt zWt{+q5f;nYaI>wRXdA{J$tCPe+&53(nr$vAz{Zj|@z%(K06{vZQFvFf_O(hXnMvc; zDYtvn2QCiohVsiwY9l$~i=UxjTy!C*nA<(2o<{lKjX{n4Mt&WnHB0H0k*g(G4|;;W zIm6z@yv}dlQbs|yOf!DOt}=ZjUt&K+Y9R5 zpukOM-$j@YlPa{EM4GV3;Ew;^bDn$86NmDXRX-(2zf4mz3K6Jzy^s+hBC7L>n) zsgj_Uv*`_o%{xrjh^F4o(>-)JR(54fT7UZ|FZyzm&cT9*6u%^1)?^KTDX)!NO<$Ko zu>-$j~@`AGTc~}p)>K~Pv^eq)+2k)a-YV@MJkymcu8P9|$RjT}kv^JVUUdWgx?{|BiMT3|I>%Q;Rap zKz2U#r{8_|h?V4FkL7V5m&oP`N66YA)lJzA@*B5i?&fTV1nC`KOEjieODF7Cto>2H z%i16P4gdK&p1JGCE8lVW!)x2!bx@_*kQrk4usUtincb>=Ld)#G}RX^y@mZ1XJ= z;u&~shoneYEE_;ZU;El4pXTU%4!%ni z9i{8(MA4aOtDi1gpAu7>Wc{Zh55hCw`ers=dC^O~2TNE|Gsp@X zcA0hS24RH_Dzxpp3&$9-q1;yS7xLJJ)c{WH;%+iQQar`1qR4V8<^l-6)&}KIsTP8L z`L2WN_@6sf&qvR5hj7RVzn)XE*V~AP=3nkvKsJse6(%<_pJH+;vW<$td^r}cZlkIc zsuZYzucQ}2A$@%0rl%u|sY!4*HTjVT%vE9%RlW=d{&G7Fgq;dR2~^FZ$w-TfB84dd8>K9*310 zV0%UisL`#;W(8S?0rcsC*+KTN=Of0xZ?qzROI-0PxxsGBS=12%tB3VRPz~MZk)*G`*3q~viVM(9!WKrn>OhcM*)6M=wL-7Q z9lApb9gYhY=s=OC$!A$KluqX>3IuK3JMN_s2_f1kn?SBc)uJ#oreTpysvI5U0|7OY z?}*Ewtw7Zk*{*C4ZB}gI+yV4niA-{BR;-|~m~M~H&hQQz8F_1n;g`yQLiJ!^lQf;% zEXAtv6_lYDJ4d<2yF~$Ic(P4l@Ou@KLzC8e-U@|3EEi}~tGWNl3aHOCv-bUR`oY+5 z@r}@%cW(X1WYsvb-DCkvr6Zg(m9rq!+~TD!89} zG|y@jwhOVh{|0w;#C=(sdz$;D$R07u3||aso1V+<5!oUFC-Fo1;-!@rJm`2c;D? z{BC&>kYTG8TZ396&P1m(jX>anbthOH+Y!8#0VU|^pH2tfC_w`h;#d0S30fd9hjd|Q z!*0XBk6$SgR=Hi}pAE|h*c)}(tro`mYzDFgHAkO@5nH0Om0EdsXrDXo0`|0$*98>ZOSOISf;%(g2&zX&iw+_P@|F7tC~P^FI(r28#;C2#r>vNWps)$82r z+><~`XSgb#)9bEQ+~OVa)J;fHZVam7WCVgRy~zl+7!HfG#|k#ju_YFN#D*_e!G?M_ zNPXPAtc8sqkH@D-zl+mmNy%oDg(rbxR#RjJ6_W~G z$ldbVpygpe$T1~*io+*!_z8Ia6c2x$qmQz(8)L&hU-mtP5fITQ-aA0D*u`a>IND!h z0-s`vDWaf-DMm+M_wMt!7FJ7l&Q!IMylIAN+|0?FU6;Yivm>x5e85FVUzl@Y&Vb*5 zA4E?sN=gZw2~<;b5-9HNdsWZod?2XLy@M_g)p3qbiTBE&TNrx-@%adQ&T=&PwVfgN zFAp`<{zlCFe%GNyvW*>PoOrE)w)i2wZYjlp@ZU}<1`8-LXqOF!2g;#xX-#MnC<__L z6mCoRD}flu(DaikzYncJ11`w-eTsC@+q~=MEC8Q|Amwg42f~06?a&+4K`%fuGwe%D zWv4*@+_|(HYV2Ij^X=l@KMXda=Eiq!m5_XPsBvOTbJPSil@zmwBBj6-B)b=S1DeG$ z{1W{z)|w+o^ivHUT<&%)T&=t)>2ufd+hwX7oYIIoP98YT+4K!>Xc>clwff|kT9G#m zR7>yx_|riAqk+yc%$ga#D5UI|FS8?d^tR`@V72muA3t9Ck8y*mlO6vt$I@dx^2){#%C$D+aRjDW@NO?G(?5iD?_Z zUr&}f@kTk@WR}w?W*tS6sF*@-7u_Ui20k!dKzVpev{v5dp6{PW_W(anM`+i~QJ!w= zrW&7YJJ02FhGXT@Qu}U{P8dn5(TNut3#rDAP~g!9fy^!52D)WbHB_N@VJNWB`z*DR zei;goV7Lx=qG0o!qpYTv20_8{fB*7N&;q}!T5&yeAXKHH@vC%BiMP|H8b!D6 z@yzQVER-1xDiNx%q5%K%S9_>-_&)-IY)Iv*L8@{QZ(qPDKL6G~`=k5(vwHTO_tE{~ zPri8HJPGZzl`$5w<5q)=6j z4s_X})~-J3GuJFus2#I)$9rvajIK=cyGwFNuY}{NQ2G}Zr#u=7eGJS#`nOQT!WOSKAWsiJ z*8e;cXHS3W8Tj>PA0w82)-dJ_Ngqe{noL3g#Xw1QHWhR8TUYdg02;my*p0I$gBL2> zJ-$1l&%Ic5MU(?NQVl+h{6oAt-)bQUd3?S2msM{s|HX-)HoS#rchGkg7egALf`IL< zd-{*Icw}UE+2XJx{({x>?YuQ2|9SkcZ_A8e$(Z=xzbC0qyoV|^L2(Yn0G)0Ik}1># z^$3zZiVe)Uh6r7y_?A3xnpR#QP%BdW?zpQmK~k|vTF5c5?50gF0(R3xKf4A`h6U9K z06gq?_BzN4z+(gU*6YkUTAbEi$k3o zR&b$yl6Y*d`8McjStI9l)`^8Z+eJym#8p7S(;M0vs#+0tKhmiGn+F6fjk3kC10UT8 z9V@ih%W_yTj%BnAeAL^(H7+2T~DP8_ZU0-qrXqEw1mOOZq>202O9-iAO3 zE-FpZW+=)|U}}KsdUV?baC)r9g%xvY&r>Yp;`=8$Z;UnKV#)I0!(`y4X*%0tf|x{# zSwoSPR1B6@ZkljMxg|0exB&IdCJywYEzhDgSh0Myvg6;(2_~J^<*+~*xj*vM zgwu49aDU{USJ0^_@ZADMw440qC;DY^fJ)35Do8w=g0@ z1jKO2lzxur0t~VZj$3s`3>F6f1&3q>mNCEnbJKS`j4RHsE|d8r*@?HlAWb#I>&T)Q zHAOa2F}>uZM~ycC1)UAlaju#K^#u(+>0VmJSxGLP8>XrW+5}r*jmO^L0T+~9-3O&A zv5x0pwiP1|g2t0!*>X-A0?AqC&mb9#t*^B9*Ti@$!;h@k;gN@HR(}9 zuK8W_%Od)|Z0tA0k9&M-NTaltr#eOs2er-WfHYK}`>N=YV5lvDOyPY}N|%BeidYw! zI$pI_P|dr+DdglqV@Nv<2jT?BTsvs{#!zepj)m3f$;6NJ4Ljh3b$(}|xj3HFu0$4u z@H(V@a?JRwAlKy>@6DwvXR8&JQ1n(l{wf(113E!Ib62HHjzOD$ix0%qaYu|P=2W-_ z1N>O1gtbqoBCZehWBVxUVEc1s&wikJ&JVHtnjL6D<6ZwGFj|`rO2_|&tYMc0a^5gA zDKN23n<-`^{;@F{KGK?MW%-IEPMUimRNvkJr5GLms(gb)E7!_)dM*|uK)eON8@0_5 zTQIT2gw3b7u1}+1beoc+%<(?wl{PC?evs7hu-xH+4+w)5%}eBBxrz!&K(Jdrk3Qmx zMfIZb1zDehERp|BN%7{CjAWoUBqzOMc@#zFGASDz`_EAhi#tn;Vxun zQ*ypL^Dn+eoAQ`Bv685r*n>G>VkL?w1_Vd8Q!yE$M&K|2QYTqv&=t`sk_6>)+dVFE z8YOLRP!$QJPf&;-P%NtB^}C$$>2qHZ-ASt4HU=FPoeRtot>EBQwUN1W25hTp>6MOR zyB#@ThMVq@A7cd=%75AqkD7O;T4(^>Ah{K~E}$c{8{7mvv4w_gmGsGXd7Xt~_c%d; zV3*ehzI`9cXWtE0Gw|%8?D{RM8F1dLNbXPGT=DNlTcZ8mx4%s;IHR7@7XI!9lsEL?5ar`eup#%f31#FAkuoCfYuTCP&KJEB9GoYamR#( zhRxQ(>Ja?hbdN>RpG;CUNee`{p@(jgAuctm$KV;umYu$O@+h%-48~mCtX^U+K=nf9 z5G|->HAxM*q6In147z`2s{Ga~Rh|i)L_gG2Z3P*jWxOR)sKjlOw#qT5G-P6*+)4Yt zu>~{s|0|vz(K2QZtO|Q*UjF}*G(1~q7TB-Ez0gwLQeF}d&*8_Y=Ggs3ox+|si_wuXPABib=~{!OtrF@bCFRQN+9=n5$)ZDQ+UwsKzp zhKv8V?5f;FU-_MRgPGF?>@2j5?FsIYcL-yH3xMJc`vhC%mxY>9b-)LoU#m>R;K`%- zYI!voCY?BT{uiksp+70S6B8e}nqu}-PzoGV;++Gss9FWuihkL} zIq=A*_-P^FkF7qb!EJI?wpd5bdLElq&EF%b;yn__31S7Gxu2PhSG2he26&sKdGvMq zz6UzJ>Cq>cZuxqpnFmLB`GI`C=qm(LF`xSxmM;9Om`@ptw??*N(av_PdVRo$S|91Y*N=Z+i`tEyS zM&ly;t7;3`#V*+I#AfHDiP@>47>MgtQZZf9bP2|8kqiOzNk|fHFd29>A*29y;cDPJ zz3Nw?)RQUT32Ma`!k!3?p6q|ycOc}e26xv$8m$5Djc(Zu>Ao;ko$q5GY@^dR(KSg2 z6`NHWH_gaX-Hx$4BaaF{-aRff%7`HU#jAIay-pmCIctK6dWtzrk%K5&_Sj!TVkgEz zEz}uiWu*D%=ID}ua{oW)QyO4y`7?xJ;2dNVJjdyx@j4CZ;+2VG1&zYT5EM4Odb`JC zkkQ#7X;602dGA~mw?t<|Y*p%HiGFLzvYGuZD}uDLH2-BY&yo>XDo26Cu#VZjI9Xnl zhVgW$Z<3%No2yzqy-+!ERjvi?DkS{EX2du_zPLLgMTsG}*r;t#FjX3|14)ySvn5+O z5UADjDaS^Q00}JS$_CJ|7~S?$ErTXMEOLr@*|vpBXrxfyr#t{g{>ID_P9uL`a>xBAOVM&+I5TRq(K>9iDW_kI46FmDu6Zn5zF}f_0NAY$vorC zCj92aRV)_r(}%uuRs6T>^7C`=fz)cH_=xX0UP8!9-d;ryjb6ttSZ;FZwXd}+7Yiz> zE8$w1;oJRlv6UVai(cEP!0(pmuj-b?2RR05)U7#))@A z7WU6g(o{L@pOv_KF7t)-pl(tx!3{LBrC?vkF4?J9?Hzf2=47$M#xsYq>zD6xSL#&%N7D&&tTHN&|brMc~>}D-MaZ z@+|i+6laCCvJC+@fOo4mqEvj{yO!>wk*mIhbA^KheWelAA$_2m&;uMsg`Df5N9n<< zFd9)!Hz-TRCqN??iWMg}La#862i-e3?YcwJd+m6xaUdax`@2Wrg!rWJRI#A-oF02KQ zWHGPH9a&)RMfA(|x`OJFWRc{m{2{^PD9V=|c5C`7jwA75K2fajVt%m$KU>C2V%~S& zf5*5e{VjjrO;Yc~@z^$#)v|?R&QRnuaH&A4Tetiwzgl>A%Fy@C(YJydWo^;PRN}m% zA2t7^{HGX&UB-Lt-$fq>4xysCxRXhhBUeT(2$*9$7la_8bPh**e6#<=wucCo!;>bXhgzj2~j7><-kfWwc@dVJO6OdGIrb>tYBne zg|zlC8`N2b(Qy9GpM`5OpC4v%;`pqEFiVPP9h5UdBXra-Mg7$_c1l7#i99GRe>zVk~N8iHG*3B8I z<44n^>+hIVcN(6!QMtWCe#G*f9_E2`-X~4*qx7-n<@-+S%v$IxXp%ODU?R9xo)48w z1_oKk)huzp$wyLKWq}C$o<|Y6)?(JMPz(W;gWs~l?A&#AC(Uy>Y~pOr3wn^A9%}fW zkn&(!;QOVg7#qFQ7UWmlGB4Us2&Do&heNJP?;!Ew3SWBi4oBnMoXM3e$7wYvomwaMRN+ zOLB#VF|AyMjg)=dDuBGIOPm_k$IXU~bCp7c&GCI)WL$s`hEf$Qertk;rYn|O62NqAxvmPGHkHeWo#pldf zp4hBcP8^-HzzNmo-s!g8qh|8pD>giLx|PX##eGpF(^{erym3_=M_ren(PyA@xt-og zo-6&2E7i*MfCKyj;4E$Aw+S13ETdqAD+UY3!C~Ps;-9xTK&~p0W4=vx+WnA){X10I zhi)X$?5UM!1EteO0_wV2&_Jw)XZ?Dp11EL3yK23?818Dv9kXmNhTFfVr+v3DK z^1*~PB!gY`wG;1!s!Z0eT@(Ygk@-|iuDr%uwM&NF2|Bt`q?@$IM=Otws+Xx+ft#l~ z1lxwzc;bPNxjA$VfiC50DC9RXK==`}Iji#3W1Uv|Lro^;hKRZIq*ORKP-u*wZ9X|baWc~?}Zu#mI#j`MPU;cpoV5=BQdl#54dRf zo&3{uCJ&hq)qzQb4y8``PS*(?NS>x7RUazPIDreDmTMr*(NORPB6tUEZm8WR0saKp}+M|!lJ`YR}h$GkNL)=5$wpp3p8)i3q zYCRWG)ii#$S6mFGmY+j8PHgaj$g~M5vzE<#2#q5*XB>hCkuM@nl4Jf~NRRnX`NJSu<%hh@Lfw1HK!W_ovm7#?@_oo-C8>abj<%*~A+%00dR1 zRI!2%VFIrj*z1fDpiExn?Bs}t^ufRa!DFCqZjxRCiSbN`ZnoeSt_JA5@!V<}NQ2tL zj|OR{*w`CwH8gBrT5JLiEBi9$TJzOe=1D21U92pmnb1VkLcS72dQ$)hC~O9DWwpR9 zRn2WytZ=h&f^E9{?A#I?f5N8ASm9#q&GtF-UojdO|KC>4CmWpD!0a(GFxx2x7itfd)pr0hg`Hc*Y>oihM13T$D#Gjp%ee=DIU{z~z>_H2h#t>GT}> z2Jbd>34p?#{kPaprf*~>Ka#|B&f!YaU^pKI1pYl92yT#{n zhzhxPVuAY>%F`9OVY*RHR|`>q>==;f>TJHZ+8@S(BZd@k}R^DqcZHQ zFfHU)q;bFg!gy!6QSB6*ve$N@0+5J zgd>mJMaeOK*XS~9u(_jb?#M{bW%oooZA&5(rv14&8|6#ls94~xY*9clCW2W<*DFu~ z!XSzG|G|N^4%#CHD*Nm6$IRbj&a2_Hj)?`PjOAgiQe-)RJl`YU;RwB)#~QZ|_?+*p zLP-atp4C^`sr}-^vXy&$x}>T~*wQ=>y%!2$6s`Plgmz{!7nsqvcw>>9Aqs16lvoD= zE7k_~v;52}SXsBRfiZvdHQ)YPD2Ue4M``VpK2Oy}iB1L)@0wZmzNOFn?;8Bln)^O@ z?r6TUyYEiBs@yOAHrL(gz*c_so2SWYCyuG+o4BN#C}sl%@Q682dfgXu>VPn^gD#r8 zFqZnvyCdLkM4vpD$>%h>*YPr8GupMeFMhm*EU`yS&@X?>A8qwh* zQdW>0Cq@Tk@P-sPl~YU!MRrj!$CQ`kz*P*@APcT}XV4dBmquviJ+n)KkrDLN1WfAg za@`=>3j3N1vvWb1^FG0s$Vi5470gW+;X%ytfOqSm9<2M>}`oi*8j)f{#n|l42(neNb2n+F}OrL#^VL z=VPB7s7_kg=(;GU*QhV0q;eEmdi#1;O z;8jRb2?YzaaEFWWg0wJ=2R?{;;!bFQNbx&I@;EyBV}05^BcO}kq-dwHR11Yk)rth8 zL}>?I%Rfrr;9y4IFfxWYlP6S#ubMUBg5o))vJPQ4Jnqd3OxBICzQcwYv+x){eRw2O z$7;r$x3YN+k-p_+T%az#dhP>~;=~IStVcr@sBDVKqKKM`IjO)dp>m(&%9Wlu;vD7e z0Daj&6Ks2Q5<_pCe*4qsQRVirph$Si>qaQ10&56j)WL?wcKnV6ozK6P6@JJ5QSpPr z(~L0r@Ne1!WV;i?-+;@R*So>tDU={xY1N#EP=2(Ll=`ahZ8sPD(hN+dYQ_9} zvTk|SPwfUM+oAKcsly}B2ZeQKu!7E*pY9%bic`aBLo61`7avS-;-j=(Rmk4Sdwml) zYdovmyCYil)y#IAZtHHo^|eUWj%Tm)Siyp-JG1t_c_8H_iN;!p2i*?69CA^zbZU8^ zs!e{72b-QzhYn7)HCDx5kYF=TNA|Ik{>CNd_qz@yl5NBFI6HBlX`RWkQ%W(rDYBD_ zf!2&I)SbD1{t)QC-r57NXQDG*ZxC!G?U(g&@4923M!RgQ1m87ESG;}ts}-Ux-fED1 z!TjEqzyzND`o9~m|NEKfQ_7UdN9oJoSa|B~!*7(m@s})n)udR1jp5f8OGnut!1B44 zP0ROxQPL_kqUQKJ-?~E9I{)uX-Y3^m>-cYR^K3tfS5zydM@A&kjt$?sRmG=3@G9ZR&8>2Wo zk8C$rJ!P}jIPdli1YCwfY751rQ)E3Ab4hqf*yNeS>yXBgKJKHhKQ)^{i_^%*_V9@c2G4_KZ%%J=5{{a5BnEf&O2FnZGGaV@MCA|kt@Vg=BH zhnvqPNSdkrN~WdCzX*S%x9CuJft$`1BbQ2g?w=1Qu)ec)+zveJ?_>3xI&Uq{Z~rIm zEpvuiHU`9b|I46*d8h%|NHMTCTSvuUD<~QeXum)qF&(W7z+xq+yI=XU#5XT}eZa-= zztMAHSwj3-sI=>M^%?3v%rJT&A5=xxk-`_uZ#`~;fCCg$Ns&ENOsrsMWSO{&_n7y9 zNe5B)^f&Ryfd3J)i@TPraLu3(L8p6`-+j^&-ORyr)QbD0++PE(N>2px8EE)}*ObyL zT({F*ycNKLoGPyh!9YnXRJQ>k)znp_V}{*O#1bTyyw)ej?S&M}cJN=Kf3-<&gwOiq zpz|b+T>+#M`zB>3z{;f7a32bs&xh{MkNB;&d7-2@?IdeK92t8-)k$k=E&Zh^>Nl4kOK~B7Lwm?}5 zJOvXPq>a*AUN6{~N`TEFf)oS!oD%O{Q5P94=e*o*mcyEpTAItBlcx zxIgT1QC|t8k+jl3M4Sc@H(bB5gB}UcOw3Fc9SLt4i5$ptr-IGgKUqEo-#mA8+fT5(vJA5m9$S)Fb4f7h ztdx%b3t2Oqv*N^S3UKZWLC>^OPv+7V$UM2&fp08B<|n`Qf5%+4f#?repi zR8&bX5BQvqarnCtmx4BVCG(Jp7Q*zwhk&2^UIcIL8K`xT2^@hUvWF($(r#F!X} zr$0CZI;;>wz5io%h{WiR2=54gN0MHe-EWbJr;E>!?o#UwRyhiaR zzb+bM3!i$P4OD69>)v-~;u&hmUTEpfp_c}oQa%#vsYY7mt34jfT{_RMPqIz-`{{?+ z?u$RG2iSQ7+d;SIEHFN96igTkzlkF>ea499+;UMRtH-i+#Bx* zEREcNGC(*E#!GsQ#tM$|YP{#Caju6S7HGncaWBi~WAEw5 z+|$AW(I<*@ujMn7fPyxI&caK}BiqB=4dZ z)zZn4nLwv}QPSrg8YA&*_iCKHb*o?=!} zWH}Y1Ra_Qsm>nxK^#9teiHhZ;XU&adIiy`qXZ6%OuPY-BShv%Bm3pCE84D{{X2d#B zRnFw;V{$50CxFy}2!KJ$`HX&*JBu`+3&SJAd@) zPZz%59KFb^-2bre<_TZ^IgQdhP4Rqj=V9m!hgj$ z{5wvNo*9^{{PIdX(kR=_Sv9@EAGCU*qNZl@CO@ZLdMu-DI6tXCJM(wub#qSZK3gd0 z%cC3g1p6(~i`KKL@hV*tRGNJ4@J7En04|+0%WDd&6R3BXORG@ zY70atR;E|tk)}&(1B!sU7A0W|MC&6l81pFTQPgfXU`&T4DRyRic$v660;4D`j7ldz zN}y9ssj841nzY#~e!55;>|vST^04JVW0}tBqqFDKDe+fPbm;o)-F#IIxh}}p z$99?(MxwGV={ir8KFe0Fc|+9#(dmm4Y)Kh#$yP3jN^;F~TNH*0on6v>0UG{2_$%}S zf)41?ARyX>GgbM5Lg-(KXYlt8tEUY}eR{RC_Yr`6_Q{Za<`fHT3OAh?dlm>8u%HI} z>r?y?1LeLDfHQEh0|R@iJb~N@-w(VjMD1||K62uUS(yWm`MjAG(FD=?oJ*+NVp0}e(ml9TJhg))xET0 z9DAR_%4Kj~tK0j|H~;#gbIKL?SGs44*6TAdZPCleHAyx6o^M+ln&^q zECj-;1lUI(mNfEE0D81D40yj16ns6l#We7pJTbdy1pVG)r4J(1On;>Rr zZ_rx>XlpzNJ0qGEpsA=IcBj6K6%j}>mMVWFMS)%LB`|vcy}2^I&^!7Xt)xZq$UR?t z9GJe1-)1Wh$&!iVt)tDru`^K2hqDZv6`TXF2N}JPl8)jylIO%*>BA;|MFqt`$5}BI zvx|$xnn!#Scun8D#ar!`O}9fF<-o)$w`87{bHulQ=5~()ZY)(Oc{H!z<-*rvqdw-g zDDFpAx&5@B6*f#KBY(_crjjE{Vqz*kz#H(IYd^G;}B9UnA#I* z;N57Hl?tzkv3wm#PP-y2q$;HBK!ec?rM?3$7;0_fV%||7Uxns*JW&fw-hDvjzATj0a(PP!WQ^sXeqw zn&F)Zk`t_!)DZz|2*#{?YZ*x9->?H#Sk$`-9~gm^dH?7*a@&c4wZD zRL)K0W^p^9d~(1A%dd8EFAE{O4)pEa@bjx?iP61$WWGv^xZ#Dd4_|0swj)2b6 zdJYDKi#e-iA%A=qIiLh|@dwieIyxgDBLMqgwDJd%>X6j&`X5GTUaQFRs19j{WPy$@ z5TO7xR*@Rw|2lp92FPrHlnJ1;iC^WG$!iT!l?pF$Pyk2cyw#qp062EjyAS6sGy?F# zzbu_e?m965*O~xu$yjF5Skh0$WJh!ev1x8OuYXE*#LBQlXvAr8tE02&tK#P9mT$H| z5BHhqmgri3N8rB5I!<-qKxi^3fU6a8;~PYK#TS7^=9>5+IZam!uftAHH>sCwr+0C8 zNU|f^WzEqyztspMB>J6+KFqldGB_D@9;YojJEE28cgc>}D#w3MlJ@A$%JrPXoISyd zgVO_g;GCA|tjX&++dW#LhCYsT2-l6jL5|YP#|{3yE5_ZBc5_>zTgkq#bjdd6u&kZ5 zMc?8b@vI6_leEdFK(@RSHZ6D^ZXmCCpQ7>F9Kjm6Ds6IyFp2w;RDP*k!UIO)sypu{ z1WGc7CM(nw1Etp+s2GEEmY!HVk3JBPD#sit+9|Y!D&)q(*!uzuIH``f>Li)2O(3b~ zNLNk2lG(bfnlFR){(K^*A<^6eW3!iYVt`oi*D!){QF2*`1zy*@7YTB}^6Z^#$gd;M zJPIyio*Q9e(0e>fHiadTC1H6`v4aa75S(BF%rJQpg4+lPG#V;H-3u= zvpP;bljhqk!;HO#|ApWjU8}c&NRjX0kyU}V+w+v}wCsGt2%X7s2t zCjR&DNvacj)TJiQNe;z8YfT0fgT6?oYq9rbw+`t_X@Q_Yv=o>FHcn{=|D=VpJYZ)y zdMbEStvE)Cz4Pcpuvx>>!Y$s8r~hR!^QP;9y{6d0IIIlFnEY7{YI6;9HU`9rqtg}? z&$VC2eH6x_ySXKt3dv4!CX*;i4~Qez=n_tYkA~QHp|LPcPY$!d(UWUN&Wo-9c@CD3MbvfqzPa~g zSdg*YwKTAb>6SsyolhryS=bU?98eut;ys$>p4Oq`iFsf%!6Q1C6*{OBt6d*{&A8+o z`HxGVlSAW(#>BEWQ%n;@PEj$nlGC0I^fr%L<#{0%(HTPLUwkD&R4?m^h-Y$Li(zt5 z6rz<>%--gq2|X`F@$kAxA|&_K0WsJN<8? z&Q4*(j>vwOGvDif8FDhGAkh`TyjLWJYSr^GxYopxD<;vPF~}Q3C|T__al#c=_Ii; ztLY38(pER|xB6k~vxOOONsefwG39>6f0JVH{Z@z(BeNVP;LqusVX!G!Hwjbp_4C|j80g~I zM8$l`LfjI4JQBO(9?}~oG`t1{@9r(pofBH2LvPE3EN+9c5mwwj&(DN+XJ&{TV2g&# zg5j-A7v@nWe$}~!k(kEW)3WCK!C#eRR_gF!x5UYF~x=ptMCqtHrVJ6s4 zhgjic%=+Bx?~SC1*NLOQ7MgftgEhjtl75$B@A|;2;^atd*KUcj8#7Hscw0aDmW(=oYGqUQMyK0Bn0zS55>@_^6T_{k5X=Yd~y+oSs zb&8;Ogl!>smg>6q4neMHc^Ep(SN-qKXoIvS^4?yQ)J;)U`#ut**L!fvWg)UUAC|>N zL3Krwv>N!t(9PcCQ^;LEb--o22k?%CK@$n80Q_yI+ovL|Q-lT+_}L*awiR$gOqm_u zeu+PN|C}H2jewhXZvDq(l@t5G+f5*pPBH5#l7gjGt3CEBlX$Q$Y2~{kyS%!*n|<1* z-QwkhEZ7--QBv%^029|4z%?;sM!viU4tU3Ux*hO03^ydJMQdzBl=6Ui&BN35TTZ+* zS*UDC^Dh@`W#{H@<5o>h$(#@RRwm9xP8T8I?Dd!LoGV zmyG$&`6bC^Ti9oJ%j#vf13Kl2@|!afHHnL@+&Jd{&%S( zU-xd3q6#?BiH8&j3aLD3pTQ`thL82%m;}cPI|l;isi|j$q^GWS+yR%o*`H4|`lsLC z@gKh-o1NG{-Dl#T7E(+;MRKVaY$C!wC6%GX8>!_7Trgo;6O=(en0X**Sy%^zX*Y$n z0%;n`9;5aQrV3!gkqcW3>~n$^bxsyHZJNVS!3qFROvunTjsXEXQ*!#GsOKj}xTO8# z?KjEU;q)n;*kiqG0--jFxlWO5R7@-BgH3Hdh&oldotsk>{^7!2_X^0=;E~mCTKRzq zJH=I#uLj+ZO!kW7HA)Xc`Vx=jPi*$VR-#&Zsh}|89Oo#V%xh2@&PErd9a8t@{zvIV zzZ0aE?v*r0_kH)ke?FY|_#0=U|F*D#i@$-thgW2{6;A!s7soeTb8b%I)aK~8Hx9mg z=bMY(XnE`4o7i-JEe}Z z+4Ymi1|NMti4NLAFh7XF6D{))T-!YN+$jSty`*w>wz4v^PSQc2RMbjhqq3(Mc-I?z zPKQ8B5m&Vw@*}MTcVdleF1$b=>hxZRgYRE#yDTrgXQ@XY|k?+gXN36JNLX| zv@rg^t(s3Z47Zdyalm(viPhOoF;FJH6?zSze`JYN4PjH%3Gd@5&+E!vNK!*T2lS85w0ed@B1^qu?j5D9KkVPDqDV8X0Ceq zbm-QJC;p&`^>`xDiOQyz-HITpUm(DU%nn6qh)PEn2cRB8rf9$g_gJa&ZGqcBe|M)x zp06swFTqc(Xo+Y6ChEcP40qzoM{;!!+EgFH&y)Wi_WTLD7)M$Q7b;?YB{P*=&8?v zcpS#$T7^yEeqg(Sssw7(lOyu|t0I?#6-+D)Q(f^b73~zOJ^DNxyZVd>C{|3kt*5X9 ziiYw2!;4-kWyk>F>eK3`+Q6v{Gzd!ex+QRcGKfxK?) zvHS>G;bvb8Vm5Ij!HL;WcKyf-Cu8f@{-tP!aS{5UD!Pso4!84h;vn8}lZEL3#Z*#c z4;53d$mHo{UA!{!2Ida)nAh*3lid>IYb{;nhOf(g46i!bJ^k^+6PEhm@vHnacYMvJ z??&M3;o!UasUBb(h=4Z;9R<2(0PSF8YCN+W+e1szOu zr!PwMsZb=U!o7UW-ql2Ub`ItWa{uPANqjRiY4+puO1V8 z^_4nb>_Y*WFKMrPEg1PDu)OFdHH8d-yIde>=iiVz3JA6%zzB=sCve0^Sy>fo?t;T> z#~SgnWO?vmkfd5HYl8R{$#Zu`$cRfcPZmd$rJ8IIOIXl)M;+X?D z=u1|J7<=HWZnQbekkh&z7I=e@P#Bkr6h94rT|h@@H^>mGG-Ru!Pp+X`K}FE9!Ir1y z(Hdk7E*4K+?7#!;ka6mlCl*8-ZG|puVmjH+&I55`eYo=`mgNM+)Klazww9o{6I9bF zYbCW%qL~lwY(r3=96H_lJd$|%;v`_0uBP|HSK9#k*T8{8ZbmeHs5<*6vcicym0S~iY@nD_imat#O1yhXCg&h z0E9I1_XPLI3wa0m*Tf~j<9u1%Il5&EY$n^bX|wGrR)&Up&BcFyh!GkqQ@eJPoiB}{ zIbi~%DvE)M&kAs1pmz=BW-rV+rL2>5x^5)U9~-jAZ2|sYlN8l+Z%b4BZcC9-Fj#!F5DsBAE;J#(2qQKhIc@Ep_)F*>6Pd&g=?{E61jQ}mBA2w1|{2iS@x7A z9N=oC@hlq+3oAlRb{55WwryA&GiGI9RR4<}J;|Umf-{0uDSlc7{(l`O-fOr0VzMmm ztW_q3#d!8Qm6b^ubF%!8oB!QtQncUu_P5E!m&T+#Ffl2&DCP!5bRfSSl0&!n?wi;r z|JVnpyn8suc^`Z3oZ0~ztyqbP#4zUqb6xL=Yl1c~nPE9fjIYFzL~dzB9(}|Wt25&n z{H|WyByI9E7#vj}w8(X8lXws;o+ z11$cKs+e<;$yVZJ28r!_P9`ti<1!Sr--Cq1G1r~p2B}2R(x7`{jox~pr*s37VFiK$ z!7cG~E0lw$qIrW*BxDpoSU2 zF^4QKU4b4wvn9r5N_a>3JCgK*C9a`>Y6v8pGx1Ms!mJC;gLq!9r3bp z8oq|aPK5WoTt6*D_wPnjhv@jD)I-e@BKl>)F`zIV)iOkO{e0yYFM50qn>`F=+Nx8^ zN8)m~ayN*XAc1F{?_;RRteTYWu(-YHeKvezg@WM*aL_T!C>Sn7`1XJO{-@?%hto!G zE%*_c9CSsIBBhmNx~huZirrKiev`CW5zmxH06yldf}-uwTSC}~Mu2f}5DtbJmK`+C z&$`;mF|I4WKh{%E7CUjUEz4vLSw}HR6p07XHYGNH>04m)rZp?hPEZ*Hu<+<80zy9* zRvV0AX3mj^SfOG}^hf(Ep5Gwk#J;G72B9`!{jLve4Q+{jz?4S!Pp;tJkZPcxb@0<; zUUT$Ds{5M*KYH+;xHqLwbRx7UZRs9C8ZIWVV#mQK82GKMAab~qD#U{kWqwErW+7YMMcp)J}G{x z9EibQmv>3iff_0W8lbZ#gC-j9^dyx&b^_GR8-q@{tqsH2NiMBgCDU@+Weq;5@&ss# z2930MCRL767p)vO&KUAI=}`l6IFMe_@He~eQ+CLghr!PP6|WGv5_fade(8MNY8zrF zE8O5O?j_5^mWSam1p+PnSt`_Y&WAqjSi#F7^*H9&|M`)R)QG*~?|kbDS^Lt&?sl6j zbz3PWgCd)$7^JW(3h#kH93T!k$+Ae9YX<^1fz#=Y#f;tD9Ayim7lSMb_yjs0TZ!7^ z78`jz8_kH#rpw0Pv-YxBb>380BS?;ZrS_j>$#6SoCl0J_HGx73#UxWCfr`O&G1vpG z923qMxb225@_k?+)PcG9KexvK56kOTRKk0aK1NvltYORtG@wwDA-V&V+geEdsgQpx%e`BUyDAO0UZz_AnpS=^ z2-U3%pk8`$faBn0vke(^hpaUUEYI^e#{9yZs+&#N-ibYB3&h^3VP)dxS=&7-XP5aO zoV#$(yer~%SuVFbq9yt>aH9Lc&Hz~{P%zwexKBJ}uGr&a$i)tPn-O@e<}IT!nOn(k zCw1e_bGj6H!4rxXBCO1f?C3~u>Ce_RWx3CU);!4Boe6=$np{M_1k57xw z5~*s9r{T&B`kJ`NYY|5innd7ylqKIKOXBJ1ewnI6s=uNeUQ(b~!0M%g{bUS2rMA6V zJAamK{xGpI=Uw+d>*;&oXT((Dz`J`$rW0?hs!hPYn__lSq=1Skgw016Pse}E)6i*A z{ZqGiX1YP!nJjOT#-ke76YGG9} zv=+rqD5NsLxUCSBOglkNPr|)DNDY+Abo4!O$*h&*Q3EW4J`Y>>I!-%WYPHp4Gkf}T zM+hy}zh~_QvLD~*7)jfZ6Zn-y5+s0tsoWGMQ;gSBpC3=29kBID}uDL zEbe`NieHZ~MWj~hCS(QI!Z%n>=Qwl>&X$cR8NA$KuNyZ_Dm16seYz^ac~?>g?LYbEBU%Q-FrGH`SgknxqU%BEXg$#`^psYIWQJvq&lJQ-^1%COgyKo9Jc-i? z;nqyxULFNd3{D0+OEGwr<6l{UW9&!Y*m6u@v>qRnj{gf;1MJGHs@qjpz); z>;dKA13?FZFb`u;w^FroZUy(tYJ&Dce-OrH2V8c^PQ7}16eG0k-~qP7@yO8s0y}st zs44xad2N%07C!?M_7(Br399XMNx%lE8$qR2aQ*n5v!7B8d~?EO2&?}{AV zDQC}%`L`@Rwuiay72aPdMj1;b?mU};iUL=&~fuSXD)Mp#7_e< zu5NMFXzor}O!nyco_e@gwhqH>UTiyK&jsE(krQ%92Q7SADt?^9YU}oq9sZcKBoR^#yf^$?LO6P!{4FvSocs5D9 zq=$HF<$y~y$1oy}We}G*>$^jMB@d;LgRKhDz!PDtg=h~=;Aj*lgPpcE?yr4Y$IwOy z_WS^f3gg&8(1{tjf!1rthI%c!;FQXMXKRE4P>{Xj+h-ycD!HqA;wADHmOp z?+?^MSX0k!PqZMXeOCtBDxM8KS0pXbT6w>0uj?UFHl>sQ$x!WKd;QH%jbq)-9enJJ zuTPC){{!s22B&3A`^CwhjR`ldB|_J^+sN*hW-Wm-t08MiEyYw(WFHmNBu$my;S~vV zvJPqa_(H0XI_r6NQfAmnc?vhhZ}XHCa09k+!G zTOb0Xl{b*KS^CP`^T8!E_Doy25KD2fJZ0sC9-+p~pt)!W0_RNxJsQQrg$r%BPc}UX zW)H?QKHB_4!8CX(p=(SL7klYHz><$*jIg0*2Znyf1aDW--Z2MkJwxH0Jz4!Gsp zpfN(JhaBbY0HoytQe)Ci;}!NWP1Zu?8LP)|UyTGM`fDwln|odpmYM6purV!89Jehq zfmbfYY@U(@&YsaOzYYe_j3uS?R=ONYK_dWT(54 zVqp8V4pdMf)n}x{XyT(;z}?L)&jN5tP|^~4buLO)d)U4sWqtpx#)00QU_HFe7g)x@ zFe#w17p6DQHSace+6}LT{_$(x*nOr}JPy4VifI+h5kgT3ojs>c*(FX5>*MxuZwILK z)YecmwJN$K7#EY#*vGH~$rH!Z=E=AALRKId^XcWw@0zpty(Er@1%AI0ZylWwqU9V6 zD)&E1x6L{lkRi$pQyJuz)ldqVqih74i`|@2WV~jJC2O%T((~w7e)WCxY>kC9M-Mr{ zY>L8)(J%iuiYEKZ|59u2<(FrV-aA&$rSm#+rf^sOLRXKgZdjf=+x7*bX6*<-0~Hh1gY{qZU)3hVCgYzTH2N6Kr^gXpf{7MudK(F z3b!$<1cguyxnM!HR43V|)W=qOLF?@zFv+F{x5=SoAa5FSoALVkXwdPTHvM{lHGm-~qTuN(-*oYz7M{hQhL?`xIp)+`h*P1~wn<=u9ibi*$oIY+k)Rr6$>IIpca@i7Tl^gQe;Ze0>sa|)Zxz4jn8=&fP8)>Y%6z-!w>;A& zyA|B)p^4|#6tKFs}Qm3P=y|IEnO{4 z^vezVKg;qZ!dF!Vg-$+P<9kJiM&sS zFZHH6A|L>R6KH_&1-x>2xkP$Z}9mR1Zj~#BD*UJW7 z+M%9n1;s#3Q85)$Ca!>;II?wC2c|Pe>HTvT{_}Sl{{Fe2it;(RbiWJmdm0Wb4z7ZO zr|IO#CkGGBT^xL1-rzyFbP)yDCe2$MeD}%G>bVR50S7O^0aRljDT?fhoT1nnA$D#x z{9^t8uHv&@UQPU#ptNbw1SnIL6CI6pi(rPDIfHSxWNzNXbHcl>UEZn_!qz7t_jVdd zgwZJ;DHxtLXCng2vj!gFcXy8b;Vb6B5vMH_un@H|BwvyvHVLvpy!gPxCP>u~@?4lju=Yg5?IHk@JbH(qCJ0e5g7$}J%*BZ2 z-)I<~F}%%Rv2(g7CHyGY-?$+Ce%GNyvTYoxGqL2Q6a(cPJFyrvgD#B7=WGhjoQZnU zN9lT~VG}*o&I@y{vUf^0@hTTrR^1ONIneN2kM z5!Av$aG_A)KZgV*k*m3oFe&?ge1x03CEV}L`ObHi_np_N%A)rK#0l!dx00@)JZB&^ zb801@G0T}=r)x9X1$$@oI&}r@(jHT!`W~YXOtSsg)E0vSV>IkR;JT34;5hDN|K6V@ zM=cr2IjFp`aNMXFu^2Q{K{yyyM^a^U8j1NOB0RyYWG8JkANYmt#|s~hn|%L8@uLqc$@ciHm>u^oBal`Y6rfF^*mV?1q#_nW&jgU-7$u5xCA9*B z_|k+-fO?mB#Vl0E&}&c&$F{4Sm&1=I#$SFj8y@{-M|X;^Y09a7drc|HaNv0SA+zBw zrdUV?@1P*AZU^qO<7Wq}>+-UYblHZ9mEt6qy>lAL22v^3YafP220UR7%%&j7mxTpPx(rp<*U6rrQvjOmJ5^=SG*%zJP8KtNu_#@*WPY9dF;$JS zNBG3QKKwMv7DNVAiK0B01$PH^x-N3=l&>9s)#JvL4cfZ!_ONPMSxB;L3`vk(5j?-{ zeE9Gd5H`dgm%SPm025}+`_~C3L^Zr|@k5gM+8_$@3WGf942p&Rc0Cob-s4lZbF}ei zUHFy-*X0W~{Nr```S2uW4e1np24pbp%sN?>us*yNGJfeNNR-8(0fWaZrxgro@ir?S6R*WPIxv zQHY<8{X|95676BFUK`6I?^>OvMH3^*0%qn~VIHWm6bSIX^#V*v8RaYogfDe?I?%L0 zS3UA9taMwh>~qSdH%*A0!RZrO4C1g6#AQ}qx!>XgaJ=N6=>GEAm@Z2>=GSB~zsC%9 z+b9<3khh?lj%t?JvIwd@=wf34t1~P|Q08{gPgmwrpe~*~ucZtj5Wv?#ucHrm^N1?Z|4Wz#!0&9$Gl@8$ zm;;b(&})#}DUZr=$pI!N{5L#zgvVG67YjVzVpykg8x-w(7YnACCab6qJy(%bem;!@ z7f%4&*C3x}C&dC^LJk^~a**iQ=ZpRE@y-pv9-1|!eM;k;QjIQ(HL5Tck*iV&A7)Pn z!EUcgU)a=Zn;`SqrrZ~-+a7#EwFvTDnc8ytq*uSFwo8m1pfy}9Svj(rL(w(Yv^aE9tsG2tJ&lL7KJ$xe(UC zFngT}0}LL$PMH6T^|q2EbvW>{us*zBS_8q5Yk?`2v~ms`6XYO@%avSVbT`yHoiimB z(jC&G(0$64&>V%{?x>iK!Oz*+aUOilXl!$PEkkuk{kQH9|K4k=61e*AxOd27C+H>} zY|vLyKz>A`fzX^Wgk(%jr`NheSl`;kqK*GzD;~Oy-_g3;@q6eFTPL1W-@4Y4)yZLP z1P6mtyR-|qNYNyquW6u-RQU~xR#m+7|2GmQIP=46{MpO1M^d@zz@7;QZD6f31)_DE z)lK3!(iVc!@VX2wuAcjxmQ22=>XNVTFBxAXy6z3SL337#R(nFaQ3k3C-bbAcidNgF zz|HKC6=smjhcLU3tzX>z-3P&z)aeduKR7tucM2=%RQeIglvD#J&sO1qnJvu9`G3gK z^!mpIp7yGSn!vcg{n`tPGoFuyRUQYu^Fn(?RUV^YT8203wjSdb@5jrSP=6=Ye@kdu z<~aE}Ehf?YmN^c*5(diAL9vr0id{pI)l>w|MvEl%;a8-oQmje>2k92e_G+43s; zzvg8t#wz|=_itY@*@}j5-^wDz4qP#H&g|P?MX^wYUQR{ChGqLgVUlW#LbuU7(lcFn z23&=^K}A|rJ5FMn1(m8=CDMS@Y0hE%RIf!QCiFCp_~$`)01^%+F&6@pJrk6h1G+*n zRfjs3t-H4raHT}w#wApfAT0INWKE^w<5(}W|<;aCSFE??uN`$vQ|elW^-wA!FD z!6-<&ug#QGWCprSiiK5CDiv}5@2<-~X=ch?7E{eoxVz1*N|f!gk?EY;PDf8j z7M8iRsMJHU?cRiGW=EB|@F({M?yA^geSi?uvh(Pnoi+}c=U$$HFfW0Pp zVh+|YEY-ZN-J^u=|4OhlNBsB9-YSH;gFJf4%qQ~2l1Pc}oOG)&N1Fm|ubH&2o1AvL z=VE7gz&H&PA+H?ak)H62x5me5OqL}4936?sW7u71r!T;o$c5&^_Wf)?vQQ} z7`&3_^iBg~LRZQ4P?fSHWCsWt>b-YJ%UtTh8<`T&T0S3s2`YJ4doL0^pj$vYvp)QW zJdZx-e$Fpm)J@L88RdVcE0}*>ZU+br6EhZ*)JPHX;`KlNsO6iM3m%7E6mu+LjE$=; zOrm!qRBNMWk?q^UW@uZOz=@Zu=?-^sw&|X?HZCrXK4#n4bqs=%*XS4+N1E8O>e>dw1>#VJ8oEe z+@G_*`?poPw#oa`$(N>IoVz1rxu%*<@y(#m$ZMcm(YSxPrXJQc>&d{YZEnYajs1WA zFbal+AJRw9up~E}{_|Xu<5BmMs7zA&C2QWeXlCk;Q|wWS97fSwY)jQOiCgDq(rY2$ zdhR=SX6_Q4gZ@(>+Bb4HXVINAYm}A1wYymzL*d~{a72&EkBBNo+0z$G(qz!)fqtza z1ewD?=~5UQhQ2VakNcdum{`y}>~UKb40p%-wnFx?9NtKmKBY;Fj1m_nH^Z@EqHB1F z89K6T)9(LFOTG*j#iid>~4KI(N^H3vJm8ra2>%X4)f_d=N{i;01UC9iDhsc@Wm*%RDf^w@wzza^TDkxUt4U4L-*LF~*nN#%KsfZOnV$ zIREP}dm-(^nMLw=a#2uC_dplwN6Et53^HF<&+8EmbD|mY!CJM`T2xqdGaH`73l-F# za(=C_%rQD_G>;=|h((h~{k$*}qHd^22H|?Y-AvN>g}K0dl{6mzH@V=yf#1eg13&Ex z9cyti`0H-kwjnBBycI8;jD2HYqnQwMLMOvI#U;W#kW7vXY-W-@`@C+* zDqYK5>%)<_(5Q=;#MFoH6h3^jboLE-HhoimXcVFX!$ir8!)oI_yih^~KU&*3#^lLp zx0d{cJar-)%^ZeUid{jGWmH6*%%CuYV6Ah3V8tx^Fh#!lS-kqOzxp2bID^+?cii~* zk3K(tU?g;Oj{BE!(a}NYXb|YD%2K^6pw9?@SuyYrzG0v*OKiV1KI=S3@D%BU04g zU{OvlX!Tf7WFJu>eouPvQ-8kA^C4PwIxlPDxDnm=-)(4;m;m!((fGfRRSvvP&NV~w zCW?jjv}7s*36^`@kT`k2rhqx2JwB^Uqbv8@(oZCp#GqI@g;G!p$-YC%(`o@#x3aVi<>8ZQ(X63mILDpj~=R(5pHu#Ar#4YKd=J@Nyt%L)InKq)L-r zm)9u`z;1!<8+e&r)-odmfe8+_#{w^pWXzs5|5)&C6J*AA{rWR<#DO8xYzCS06bp?( zXQ>DS5M#$Ur+|sb(d_Xk4T%?}OoWR14dG}7z~i_TTnlB8EzC6^5Q8zwmvuNNF`Wb? zmm%J&i=2st2xY=1w^|{F*MP7MJeE}25Q2q|x)!DsqPJbb^eLZCt`$P&FbU)>F!_@U zHIrCJiH*$2f13j0@8xu-e2plV=z3Mzz`v3s*sM+v=1O=FGT8=K&#ld?wyZ@UKU~d= z`tiS6hE={)WpNx_{yBm)5ttPag0E5Z_M4m~qV==y15OO|VJNjl2Jc*nT?`J_K=A_% z*IjJCpBpd}ei)H%8CK!r)i`h=0!MUcr|_y2t0q(F9G^vkl?={U1^|sw_rci9S|n`^ ziG<}+YY0?T#s#K@t_=g@g?Zuz#UU@%WdPY%C-;`gRm z>h1g^p5(Rz&r&gC2Rk-B6#IxG_o;{_WD6Z9Y7$q1YL$^jKa*~y5dbJx>=K)+oIRwJV4G@H#vZ%TZ9O zJPS&ukRI+KsfTKRoYTs@A;;a`f(rfXITqtxn z z0w)fXjbt~sWg4m~VRlAH=zAvc`KC9*BXwZ#Jv4KX+9=jQktQnQiZmlcujzI*l9g^1 z-U+CI%1YgGW+T8SiK&5vdHJ+}3~*A@SD(?DC)poj}B(C$>_hj7iT4gj?w?H)g^Q9hZN53Ch5Y5^2N@d%TYG&$(%f|>hO)g_H$jH z=XH!;BT3U93x^7;!1J>sB{yAroeZjKuRO0MGumMrMRuZmI-IM+kJ0%13q>?EMODj+ z!t0bRZ`@U8Pd`1ON`%Ss#{Tn~=r0tBv$2`~tgyrDx;!WR7^qjYDPe^vxDle)rh`$# z_pb+bnqH>_QHNKeca-uNALW42h2oH?8**pTKmALw(gdy(fB&6pWX({e#10%{2C2b8 z5rZuhn@W*QRKzOip>0rL9#eNlenyTfF=TDem96pZ65e)=h4oyj3`q^~6lj=Z?uVj4 z4woev^4Se@=P;o#%y+Z?GH$Che|pgmzU(2SjJfyZEvi!LqF;e5%Ne)sEB1t?%f4VY zXkT`$4lKd>@M=0=UKV(qTw*d6$+8OHbS*MG*d>@aOn8{NKGqMq^=ER!V+j9S zpopJn7fe`;(y$<6aSwaEr^}=)zCfPXQ ziGP;x7Af=TCPkV?!KHbB{%Eye7pN8w?7=?lB>`RET5V(PtRcpJSdapnGsLZIx@)|* z^e+J>|D^5RJB1`?DF4KPV?Whq{z)mt?x9E#6;bB0f9VUa5p%7z4*-8|FRmZjDgd4sjpsszWN*?UI9@SERWvFSM%I7*@{##&^pCkle5u z!rdA3g+B1lGfigdSD((xCj|~14mxXQk`7WVWL@`D5g69l4Urk7@89F0*Q5nw>}91} ze`i0&rMo}|5(P>6g^NpspG$%AKQzZ@L8Rn}f9$jc7&B^6L{Ho8r|Swu&KlIw+A2(9 zR)lZV0Mp)f37*}~JcW$NQ)aPSoyMTp?RUgKdfG@=U3SEf1uFQx;Hv20o}u(scE$KpNP^S)R&3|36~!+YAC zS8)%+5PpK=#!-I~a{SPjUBbFZdPT5Uk}BKow<81-NL4D7UtHzVPFMK10u@ngFeWhb zL5(Pn#`GlSL6ex>f-7&KLTEL8P0$1@vJ&AXaDwtf%3SbOw=2FfX0>?K>u|zl@By1a z+Q!RpUXUL1__zL(eN1l42Ro})ldTTy`T|GvpxLH~Vj)tK55yz=tE@;##l$)&KTm~~ zkq#YRG;dhBqeG*I8?aK^?{`K%kOFN5y%Sh}Z&vSz^m!BA;hZD5pun{GQPKj?Ex}tar9C@%MM85#?kkyZk1bV!*YnKg4uRmGUp^YA_M9k?S3B|h#W~1 zeVr*6ZdOD1O0mOl6LelK2}q_^&EG0-5j0Ed!lQoi`1_xK{r$B7;hoIJ-O$bNAcm zFJZxH$QmDKOLka3w%37S$w6|vUB2J16*5{|-#+q~N+qIRQz`^{-!-F9La;j(HEj0;CIz z=omrdBv_=++a7>tAPs0Ly*&UmFt&kpF@)4QpYmuD-<=per#vLtvr5<wwq4MZmG#39M^ZlQz{Z5De3-sK9>YY?rjDH!3koW_bh1Wt&b0(|5VC1( zBiG@Cl;BqGjiu#SK8_K=G*I(e4pJ$oL|Q!As7Hzy>aI$oM9uT?zya}QHSR(I^t51X z!s%45bw91CR8$Ds-C*_`X!>g-yFuj~W$89KKMsOVfYt0+sDUrzMO||IjjW;8CaP=z zu4gE-a>FVxj(>b3awP1r4s4D%*ke}%cVncamdyh1dFf27`}ia?Z!qao}}22mg8G z%))tFz>UEL53ZY@isD4j=!7DbNlf;1p8 zZHTwDYS)^im2Z8+lx}$R&2RmZ9CF}v129t$N>ZJpSSZ*!h3xoAj9zh*SsRE|G5t1hQ2*K+~t&<$m9*D6~?m*Hpj~y%`wf^qPK}43rQ{rSlX?%xaH5 zd3ERmd8;xZOow`7dpvfzw89m58z}igWtSSLHihgL_KzYnevjK(C<}`dWB?i~NS!8= z?qD~oQ|U4a37rYj2B`n#=?adGk6Gf9vHU#VqTByYA9@Xm{8k^n<Z)w$ixNTMAw|;ZR&19E;b(Jh23Hi3~A%YP(UI(rLaQ7k0OcHZI)87S`E9r1XY9pL(GqpKB{iH^f^a^62 zcgyhUotU84*t%}_g+F@ukN1UJ*>nfDc^_&y3pC%v<&P2n_WOuG{OO;6v*3^Nr4+k_ zB9X%my4dP($(-~s&e>#Vs=reC35jyxb@moB%aKU22^5K=B1#~E1f2fbr=n~@8`vF>2oL*g3D6IRLeE_2JEe82_%&4cav%FEmGRl$>)d3+!--oRmpl zgBJXH!A{aR@r>U&zionZURCm)b51ZPlp8dUNfC`K4|^xK(g#JGCS++*hao0>IfERb zNx|1-MRZHQKg4tRq_hbPUpOPTixzk85Zp^LUh>#NJFA*Wa#1tq#vyM=JzI< zfb`w%|N0xUiC+lCfx~S3%wU&Cu{jjU!tj5d;It;qIa9k13c-;jMz86ho7{|oXLuCD zCJ48Sew*B*{g=*6Q531Rx|O*^25j`kj;S=~a;Og9O``0Vd>kpbMg}47t8ah1cd2Ex z-eKn|j>!B#m!z&iVPp!>Aq{7fsmS~PN9K*A&&!OAeQ(SD>;h^pW$`Tqm>Vd@y!ftm&N8PctGfiIR(kys7Hi(jz=c+JMk^D@pX;Ftw7@hv$=R z!C#QBIY-HLv3|xezen!7Wi_f6SSNJ}&rbYIo~he19oT6j+;E$i5PJDzfeAk!9DjI>EEx(v4xI5zH^a|bijAj8 zEXqiALDeiOB(yNaQ+wQj7o>$L3fw*_OyL%zW-Tf#F0=bfZdClXeUXAPq2j6A(=(*k z$ziihz#2ceE+CF#S5hQ~ir5c|DOlM9TFm`LL$>;k^Bk5}4?fQu+URud!5<=UGA5^N z?+?``K-B$m&qA_cC^O-}(@cpONOC9^qVij*h;+JQJpQi>#}cud;gQsztG;$VJc-!^ zi`zzW5$g2!v$3G59_?|}W%uOzaNth+8nl!tj!Ra1M+ugCuMNaHemi38)%Y2hD~9@Q zBd`o{8*cpc?92MOlL;$%U;O4{63H(%;=q#)R2&T|L5io?ScW~5h{E zH41DY16{4MihNGG__eZIH5@~IZHKXs&lqxlcZah54m{&Xraj3Z0T3f&)2Y$ruNhXjkU zr4CqA;i1vc$5$SR)$Rrwd&@snBVWNyVLgcr)-6DNV2Y<~0{7R%M*Zo~sjpjBN;+)hpQD7c zL9ySjpubDP;HyK`oPKW2hJJnSO1fOS+Iw?YR&bxwcF7jDg;^S0qso;v^vej?KPxb7 zd&A3m@IRRse8+n2VH!t5ZRWr;83(l)@YYJXL%DxH|^3xB4V0Ssa#CLCi zu1mEMY%erB&sskIsIW@dg3Z##o%Wl%OrtzpNbv^=zboj^{^4U+(-cRuRc0C275YY;gX`*+Nc(LN1pm6^BZ(p95_>F_}^545WPw!n9m`Ou{e(HF1 z-0g2?o(&aWbUeGz&KqYA9$IagV&>xvJ8ETg;*yRhMi+n#e?tUy}R4gR|8@{XCtP1_7Izk#p{68vsyle2k2 zVr)a_y$VZBqg1o6?lYO7RW5OY_0)PnsULP<8tcgn{DP}C`2($N(zVfh@wpAZNAIvP z+V?*8ZT;KoN5?g+-a9rp+8+yI;~;j}0^z-KNH;triv&;H*G)U-cMe+6^_muDmtT%H zWx}JdPMcMrybNMiSsBYI=Vdpv(XoKEudW$y@HD&F}j-6(Amx6JJGDG|oI-U7zzELZM$gWZ zuDdO3miPof9JuU&qrwSzlEf`cJy6xg245vL&ZB2r%m(n)P%%4w)Lvx%?#)M*sh1av z>m7L5Inpr&f(Eu(jXA|k?W+D+=7Bukxk1|F3c;HqDQ0R=?*SzlfCLhGuW=K7Z2N~v zUyX&qmSWJhol(JUHhl8W22D0iFi+W&%ZdI=Ho*YH&7j!%4vO7Qk!@5&f%<}U<+Q7z zD?PG-ie{I0XFy$OUhpoL`fy~fsT5b!#(nxZg#qi{DhfbawEECEZ-e3qRNiCOx(VVU z31JP2GeoaWvlIFLm|+9v8FuW09B#n)UX-sRCcqRrrJW|r#*s|3F<3{ji4;koBCwJP zm9-(h&>9pcXn=;eZiNn6A)b*DZKg)iYV&*8_Ia>F-Ie=c4^fZgxCZ-u(Pt%=-CPbE z665IH$^@~6BqmLAQM8V25^r<92))4hq=jVCrvu7D8krtfRO?vlyGXE-88zhc!|muL z1bE?=`eYOJFaKgP95WXgW{|7=^0$tgOdyZV>|Z;@-lj+^6#-Of+5$-Gcd2f><^}8S z!NzKjB5B^VYz1VQvZtTYq)fQXE}DQl%UJxG6%4gbSVz((*f6zKi|IV%xXzVShz9M( zGf_JpPXT^gD7L~Ak#T+c3btL@rP}5T`d9shS8=jTI@uFzQuG>B23+IzM7|jMi27B^ zb0yjI=hBr77VTp7XkKWf1V6FvjPn`i77%QVl=S(18m3zW#omcP(g0_J=!hlk`J(gL z%w_8h3)7k-_-Ug@Rq&wWfbMy?kHS7;!ap5P)4ZS1bZgXIh!ol5^ z2C|pku4}xD)o0$Q5!Y*q)%{*G?9B@;2bJROiXGBXouK$Y!(#O8`cm=Uo$VtjY;|CB z!co{-qbia{Ppk0K-3eIj5huIvl@$z(DHz1>b1I&^MS=1w6$B%4NdY%OvKNKt!Pr2p ztK9*^2PA_HhFuW?H@}vf_e0CVR)=*|IZ9P)mCZo|+Pp|9ea9UMl0ZDSRdv<_e(Bwo z&V*kz!8@c27cR_|z>mM#@#f{Qj?Mbc1cfC!I0m6LBzaB-wC5nhH^>gV-V%d=!1>vW zC6N+chw~oyeD@xZnYitGe)cGQ+f3gY-d{Rk(W+gEtvpbcQpq9H;-sBz&0)r4FOJd^5R;atH7_V2K zpHQRF9r2Ho;T1PPORj~BuLqAh|8UIHr1Lqmd$*F-kvK*rsvmb~YDjIaay!W@j~sj>62k7gHvb<~|2irt7AQlvEHuqXMU&5{Nb68F36ATOk!Z8d`bQLdpCb2+t_zmJ zw#n06vK0xkjh5j*5O=EH;I=@A4)Ri2?5(BmwgwkoRCH9x)oQwJHx8| zKli-jxZ-i#^VUEA@lT?$Cgepe3phghoW4{kc(U2dKUz((t0=OZil`BnOR8w3h^VI6DvBJW zBGv$nSA94H_d#?jPSByw6_&Zwh-;M4BL=jR4gGCp%f{{UNz+`DwR;{C?Uu!iYYkuX zHeS^>vl_H5qa~*_)w11!0=hN4eg1Vh?mg(yDXP^JtGeX}d>3i)E`#s!DC`gvY#t2D z65b*)#-f9jgngrpM@pR6T?{Fdqtj4`0D8dQUK zh+;wS;Q$r!*8!&f7A8;Fsw$m*lB7?`^u00R^Uy?Zy(ZJ+_RJn8K^7HoYp<`sn-zs@1}gVG9PH)Z=7-OENa$B5&C;##IO%qr?EDJwK@*Fs_bHWT52Pn3b zB72Ynqn*ASaLOZH_K+@-?U|)Z^hWBsc+tn@bTgydGwX^V6GGk_W_5g}S%s7-`{?_j zs6dAHWP5Ndn-!815<}s?T+%K@UwD~2>L%;(hw&vI=yPfciYINhX+#?tRED0Vc%ov+ z)48ECgqJ(!?)!iFg=GQNYZ9>GD7QNA3{u|WX#Y$aXR3Dkf$Ih!LqzhO5l<*Xj>pS( zT8W{T-NG)%{?6UU)w0@wqfEe<1Y9*I$}@}Tuw~;R1ZZ2~+nO!h>=)&^EI5ngj$%h3 z&sp6rPjBd(^7>vnZXC4zU*GzB%SKi{2E&0PIvj1U58p6&B{9!nP3yIFnsyqAQ!fM> z6ni`{x0Gl{y2%&eWTn|K&M_}uZnr%%|5s9G8Cl_j4aY@2SIlLA*8%Fk8EA|ygZ0Yj z&mu6M)*!-e!~T*RfAyc}FDxgL*8~w9^C7k+VjrRoRZm+4)$~^RF5p!ybJ_SqeZS(V zVKmC11`UZd(0JyJ4!X|Xo15vLuLYPEE^Y7LDI_@#yl|;DTey@`>>i2~Q4tk_GN5)z zm1WAKr$q(mK9@cWZBRfVPNFxe!4H(gVLE!}1mrO;oZd~c>9gQ@p7Yv4Zc8D|9w*a# z?~o$Fem1?GE9J&r}_rg-n@OeK?Z2HIYTs zF2AyGVaLZk*ZS~k+3A4YGb+hepES0G?w0PLqk;Nuv1H9#8^f@{C`))uTo%|VI_hcz zyJ8K1USdEv4)MaukSFs35VfTE;ad|-(^1147e6G4uTA8>&}{O`pje2E=&6V!{@9mz z*5jI@K721o!5kD|7;UfSb4`)o9={R@x9T~5c>0t=0iJOaiml4RZcIs^Z3Ea0 zIcCpx4i{9UufNmkM8ak@i5Ih^}HSvfo zAu?Pv3~5_zhzx!v=3@DG{~|V-nE4mi9VRQr0gJ`ptXc}iuA@lefZ7gJwPr)O0qISQ zgthf7mf35Hq>&Qi?^M~SrTBS|feoMY%WlPM3>-K5^kL-O#g<(}4(rl#^aU-Jbc*vQ zx6=!eWfVme4ho_@u0iWMhDRRy9u}p_(t;OOD4N6-Q?`Yq(sz9qL-YB4&?&Mzig}}V zz;wI}dNw%5550d4`FzXYR%3JJQY&bV4u9sx0*gi*N6AGCVH1n?sZz{ z-=_Wy+?6B#HKSK?%NI;Gz=>72n6lMCEE}60)`md&$e?$L zVUN7gVx!`Z4;zBuCAZ>*kFg<3R)1irx#6%jg+oQ7NZLjgOHyUvXJG2?rtsFAO`6*p zP;gT0n$|wrMfuj_;w6JJY69hDX9*90N(Fe zpY49SJi0V&hqTw}s{9-nqlZAdWKbB~Ks>Js)^=6UUY1F3(j*IYwSqmf7qg9IH)&P% zIdw5W+X7kUOfA0qLBT5bI6<}|UAS9)R68u6*?RDqO;h|{3wljmel&ZINwHnpx!b53e}M5xNPTbu8eBWb5IvAJ|Ic2jnOGGGdh_R3j#)|$hjUZiWV7l-V;I> zNpzX?j``94IE|%3a}UsN7~Jj!=`~o(mJpf+c>n`{66Pms{YA|^=!dBwt+4-ZjF)jApMqkt9QvLRtQj)=Ma&cf(95S0+iYXRc zs~uDXGGrsoR~%^zdH81fl=|=@kIQq{%_x~s0i=O>^O_;oa3`P@C~@xz%BEzyq)3*` zsShukvOR3SU&WM?fLixPWs7Qwpk#(YQK@JK%8q-2xH;Ry@I0I2%1YS0(o?lHoc`){ z>WwKT*hGEkxr(GZZ~z79RR#r6c2aB}MRKT!ZFHA#z2t_xNnGrDJ**TuZ%PCSlBI5& z)v2K9vyvQvzMaL2>ppEEC4%x0V1od5Iq-^&p&_fkY#_D!-iBidbw&RZ}hL08Gj}*sq|y|ieRbhZogE( z&PM-r(63iTcKf8tjN8ZP^1#*JUwAZ$*8tb?QBp+jnGLySuQb-?X5#tyjtnf%?_=vN zeqj0Ie^>qINXRoBcvj+IRC&Z~6D)`f$d$J#uRvB4pGNl#{+Tt;Cbj0#a^Jur&L4|Fkm46#VY9e%+R(#eOs=! z+#6WGs_7~yqD>2~RGovYF;G0!D$`sR1*d^L3|gOZI#q_WP(Z;1S!Y-jw=nG?5EZI} z7%yf3pw2Q)5bs>*8&7cMT&K}(VGUj>isMj5kuPk=I!wHT=Cc@#QSXDVbu|QY4;=xB%QpdJVRnm$_h5sc8}YoF z7cj=irZ5*RCARs@HV&Ld!(VW2YIK~(=?43<_29S8~KH39T+sggfeK-*-o)g8J>xyO_=y^P+XO^yXnwT zO`f9HlzBZ3>UM?nHmpfbYfu|1CO8V31d&<^%Uik_osp@%9CF}EjFFr#pZ^D8$DL7MoX!Z{2zDdy@O)d*7VQHKV zv~P*t!zwPZV%$cI6+7jPzvdBrWSR*pPBLu?$>fI>2VV6aHp9wpiiLi#om51RaJRe@ z#=1rvDcK(enm^Fflj>`L*#2too3lDWQEs>UN4hWOwfim!__+T!)X?;I`yj_Cp3@{g zE9()bv6WDdeN2-k+AiPd9Vgf-&sOC4Tw&8}(OdWoQ6tCAb8}-$2>JcjfBweJgqvSo zB@0NxP?0zXMh7g22hBU_6bo+aMk=CEeUirPPBjpnH-hHivT-GfdobOsaapH0GiB3+ zD$xqE+_^2}o~Rg<2!TZ1xmz((y2odO{nCwLA>pNWw9P?oujO>|`095|7%7_YsDJr^mqjv}>4)&IO`v0tIPR=Hii-w%1-78CSO_G^&|U$1#cI_R|V3-g5a z9=I4f;9D`JL2(wP8RA9vT$jO$yH1HzU3yI-X@u&@2SE>jt+>D1vJT1sPic%4JS|Ko zIRLd!eNG?sIsp$$d0-7)G37x}D!oy%5o-T*SEVadTitTx{YS7n2^a7f4z|wS9v0>X z8JuBkS_R6knC}0YZbH^aAGaJJ+r~kq>fkT*UW(mKkwPk>%%#({9(1zT`}XS_l!jdd zTI$-Vpl^@|N!wJYpU2X%i-Kl#CFJs>{r815D0*CV*w6*;LK1UU*rKVD*G|2wL~<_} zp8Svuf<>@!k0v1NtM^#WBryh6@nq{B=_~&}+duPHzNJ+saDUH-2z)uN3t80@`^}N$ z6&=_papV{dgUi#zBTs!3cM*eAUSipeo9aOl>Xxw1;_5%+x{d=HwMbd_q2 zsE5(vk6am=$qf1qxKDjfJB6pDNRw>TEovk=0^NGwPO@rFI|$_s%XC@`reRD%&n_7b z#(GIagJqcupU|!YJIx$LF`L!becNL7j-Xs>CT$WEx4E1DB8S}eIHKMOQ2 zNB|WVcvg79uaeXS^-0fZyWQ)<&QyQ>?g&cL(r8&2OQf5D6T&esHb8}WZljxlV@|ku7!4Lz)0~bSFZNFV# zaeDKNWio&dE*yA9;z$$V%H*2QT@RR44rswm0a`dooo3VQ0;sHmvZz{Rk5JFrQgZoK zgU16J{qx1E?q{pxvpzUcZ%L=(uw|4SJTE&!(_|0cGXAU&&zKv_mIe0E8w1j2Zk3~$ z*irJC-^PIU(G*(dH{7-w%vaxs*Sd>x4*6fXNEd~l z^3v5qMWkNS1|paCK*+Pv8yQrd-Nxn{E7^oUSo*3CJMnOMpGQ<}bgxqn=&TsLdYvKz zS_F;$%bByvwBUTwGRnRuk6K5YkLAm6$ZO6S^G{Cyq;@u$km|2genO%gxX2CKiwEU> z5-B!;B5_p2&Iz^R=xH1Lb?Am31nST}rwZu~X@g>vYg78i#cBia@=>(;QC_eZyHE4+ z3rpKCo4F`dZ&nwD9Gr|wx7m}Tlu(l2O)|9QfoJ^A`Hcdvg82|yjSE|5dd)w=kgxdR zVs+xN-?Aoi;rjUDWwP!|mO6&S(jXI(O|eion@&ZXP?ZDe{eGV%bB<|pg(v)vhHUct zXtiL2|7y^D+vtrPp*7;Me>Lq7#Y}+LoDijeS^;9!saw zzN7`*=<$wlP9!Fz6gs7y2Ej${lZVJcIdCKuT+u-oT1T;o6iJ{Wkf3@EiE&QX9`uNl zwL<=}0J_$*w71?Y7h-#cUSn5?yzuSjfteSswdao2|N5(fKoc^SCw1;3`Qyk*vqxG+ zvHK`eViZ}+hSWQx$B`Z+o33@=?pNu%HW1Rg8ztrRUe|in1+p|yhpl(U`5xpOq5LpS zk}uDgi|d5`3m1v5dnbg}(2yK9E)iN_iJ)67Sthx|K;9mE$q%?>PtVlWvxVbFJW4zs z{EOe#m+Tqopz!+sIc^X#tKfsBmVFs~!mAE!Sva~l_RH2QV#Csa)jW-<5$iQMBvP_O z(yZP!u|~XIa?Pf+)mDJadVE`PGaH`rOWBhj{mx`qZp~gjjofnJwahXzDfTZY_8~>O zsE9OC>FfjI9^p9%Xc+6DG0g(SE8V1Yc3ViDrY$6!j+31bfwKBV0ft)}6bmxs3v~CV z9uyp-Yw2tk_c&QlBPtX17wKLh zMIartV**}mybV@3gh?>}c zck-IX9M{;tbMC)>J>F!k+W!6O|B~gejf2b~O4uZBWR{8a8Z1D;`s;PTb$d_{C0Gq) z$!DM;Bw4s;cDl?yU>+KDydXaGE$p)QlOL|%=5NAAVSB-1l0A+bF>{eiD7Kg)1xU7+ z0V@n$lCVsJmc7TlL)>qev2tvH+a9VX;{?b(P^(O0&{eJ=r%8T@?zXg6xlxnqo2jjV z`VL+Byf%V`k_I|`N|X4m66gbz3yl?Mo#KTy8W&!VoYSxjK46;>;AK@D*Eyb4{tpYs znUE81_)td{Ij|3uY6j^9ijAYlN-82=v@fI}3>311VfnUSn>%jWb zTiE3wKd|2XdGM~0%rp*Mr^GSSC6&`xAn>k{(y+F2pAIB9J;nf4a}<0!Vq3 zk0PY3L((&|W*5{9b<@-Znnz#u{0~SGhL+C8VUC=7$eS8M`#jEYNho8xKw0!y5zpd87hrfz>zeADFrm&~Kj@wDh>94mbbEab4_MU6S zloYY?C`cN@%KGR#hrd2>evaVKyllUA-#=}fpFy8sH?Vr|-OLq1f@r@_g6J%C_!Liu zWv*>UmJiCF8I>V7Z2P&UXMR-i8Y2S*~0*F2$uCG$U(*dtBR;_n7417`R8g^ENmuPEZ!MgN|XAjXOk| zf~rCrqpVZKnEf>*m$aWecgdd;$Nk+51xm0ge`U*zXO zkH}*MGVm-|&>nP-KH|R}7V%30I_Z1k8v~2R-+J>Ng9UI|&K8}8wi~e_hmqIL!DG(L zE;#OL`i~noO!}b-cruUQk0CV{SW`93rx519&Gu(1TrU$kc|C3{B#+ z;2wEwur6Lu1ML`V1GDK|Nz!;7^7?G|%jGdS*-2U2Z= zGn&t+8ItB-m!l9-efWdWRQUi8cM%;UD07LFpzHDP{q% z(^s#-uzy}?j}X6m>8|^xVosoic}yP{ciOY%#`GYD`FL&i7`bl5>x=8S%h#)}OMk94 zOU#BU7;%*=` z?Qv_R5BdG|Q)nE?l&ligxZa#~OjRz;o}7y<@feZ3t3=(9B&JrGCx`@{-#Q>W1^Uze zZX}FtCITNTv^S>uLfmLpq~tMV#O?+i^2?E|a{2Q}yzpUu7p!y4xPWfO{)FNiGbj^? zp1M6fLwcQfeLWmEmPt05fhdk*S5hQ~ia=X<#{K+k9dbt@vCBw?)ru#~af%gy;JDJh zyQ`*G{oZo!aoE~T4!x`va8NHvk>9wzZcq`dQZ`V#{A$J6H-@ceZOVN}m4KWESEYxbu%*9-&_EaZo`SBS z1kbV%B>J(pVHv^54|`Otxsab-DXHE6Jy(-mDgDMf=g3NacEy2x+8i^xvXNrfQ!r+b zfj2TN_mf6pw58B@B~$KqRay;^8h};e?9&0s&WIv=Bk1{P^2XBhOYORKn_cjI%hq^@ zwIUqN?ids6Uu|GD#VQwUR$VVB1)?xmha`HVgy*d{M{lHIxcCjWUB}n#=DZBY*l%nt z+&b1|IHHyX9Dz<=em5Wv3#83vhGR9wuA<0tDxyL0g(_EuD_+P~8WYq1Co36i@MAWx zqx2&3XwFZ6U_u15()S&5b{uIpdz9BH_6kKVQ4vip*JmUPw@xjfpKrA=2g9<2w+K{C ziw?}JHXYH=Du5PYXL?oSVYJ9_?!U*Gz>)OYJZN~71Q-E)zkO?e`0_vAFu&G0kx zE`RgSm0#_3IuEX4DxDC9s!&a0(p~to8nQ0C zP4(>#t(NsV-H@jUI%r5GdLJdJQxAJw7OkIP)lgeAG#rbKVehtPba{O<9e1hlqq^L& ze>5RXrg$%mbPUzb=D-n>IJ4Yp5g&e3|GT z-9Jwmyt*OeoI&?11Z8SV>6AI$iX`TcAC~(eCncVc7MvEGsV%4V)b)8WE`3g@65gxY z>zttlzXtggZ$O9T^}s{2Ue!SXuGq`G^xjWBZo5Kxjr;Lg*%0SzP;Bwqr$W)9KBq=z zyDib{f=Eie$H!)@us^q7nUBcyI5ZyC~6K7@8xw0ER2Co(2hh*hKRJ6&aQxr%; zhMx(uU`3(G6rCfu0CgUHPMz>JTi9k5_WS?(|Kj^%%@>XP`!d1#t;=!~1XX|iS{+$E zj^vvexm1dUK-dO!G!4F3#nkU;A{Se;dN+w>aaalk_l#-5Qh2nvG91^V`fHn4);SJqJrOX*B#wa0eP03L45k19QQ8+3Yfup#}tZPN0CGXdVRlwS-+Qk z5-OF?XrIWN=sWH}go$=%KbWXo(k@L=S~ZD|3`NGP`C{A_?`h3F7L7f8)lxOwVf|nZ zh43X`H!52}br0}+oK^HX#k=*IP_ONOhZdZ;T%ZH;5I4`3{9ajIW zM`@Z}PW=6Mu8}nkTo$v-%o=Q=Sf~%*L`6ge>*|T_Q+0!4s}NhtAf!?;ONR^}*^{&B zd!g;hextBhazAvjBoZSo+0(Cu9hg}=p-OJ&bvR#;40aW)`F0QY99|$9+jJqr-%>r% zVQmP99%7-o+ZD*@Km{{W@-Xz~%vNn9S>%lEbBjeSGg}13(jrxzGB3DgCbH*_*0jMJ zAHyARp4aih2le@1qPG5RIUw8I^cs}P>-V%#F7KMWK+sPwgsMh~-hED6g(nn|bJmMT z3rReKeD-Di7>u@0=0@M2V47hX-njT7NqlYM{h&EBi031NV$&(22cZaeSg6vtc5MQw zUgVh4tqnXPkDm50OxGRM7<7?zLrG-4$Gy%fl41+#%aMzPR)d5DTg*LJEwj$^w|ya=0z%Kf&mNy41D z9l*%Y=XAw?-L!q$C+@nm;PSu}#V+BBiBX=L-P2rjv23{*!!UYH%gnNnZr3U~=7KPk zk`SgVbJ^oj8CC$rw@6@!|M6yx(wj^%#95%R^9mH}Vh|=l^hDko1Z|w5YlmI2;C|F=8Wh+P*5`x)lsz7L4U}0oiPw&A z6+``)7I{cVjh=`9DDZfcBRG}UC^&9Vp?rzE=UkJesryM(CMg{%rt7$$pIkJvKF28* z)Nu|Es4A{>X@Q{X8P8%ZiLdyGP93XfifzH}$8VPkbOzjIQo!zO#5UT-NYw7|YgzBd{VFs^W#qTn0L|doFoY-lz&c z^8)S5ac0w_yg)lv_%&bxG|fr#FRnXGR=hTgB#sb%S7@Skr|KYNAd`XYNUzyBr;xd; z{P>=$uE3?h1zO6CEoe9q*~i+wINH3Rgfq;Gx3J3|;p06{mekyQ@;i=e5<%W~@CrJK zV%Jb)H5GB#qZir;w|n+FJp?U@T4k+gC+YRy5dt0ymUoX9AWWlaJud7r{G2!S`C;Ip zkuaz_u#3yVpn5}{O7{STLlP4`?W&8;_%cUb6q*aoJ2>-zdY@mPpVt=QlU1#<|KE@853xp*fmTk2T@uwU(UO_b?AZii)!E|V^u zjp=~8(8C@n9u1!MG%D6xkRcW>cEHELEuZ91dGlS9+p+g&KJSy0{M-%)o>^|0%`6QR z3tEZisfdGopzDxzix#Uo1slwH7AvZHUFSFgJypG;4%?PF+ka zTdTZ6dfXxdnuCg+Q!q1GCorlAV+yi0sDbHm)obDeXUJ^`JoV5Sa}Pq{LVn0%$u>C8 z$USTYEJuW{ApmFHt^o{j|4#hX?LC$Y9zIscf#)ZVIUN8H`I$HU#NC6crZtsG691=AEiRgq)kD>7YHvi0{@S_F9B=vOxN}kzmU8b zvJp(af`UYl#fo64ppDwjbb2~_JLjJ>=k!0FMP|-8Go91vWIQvS1r^*uR6qq~Q5Hc& zHc?a%6cx1!C}_G+Z)V#=Xsy^S?=4Gi8pe_YOJlsU|hNa zHIDQSNxw@L@EYz8NS0#?D_!v@XsN%JuA$rH`H%s?(&)V46`or90Jjl5Aq|tQtdsR} z%YD^NQjMaUb48af*cY4+$+0r>$d08__OwMhB-#mh1^%0*|6IZWiM-<3NHV@9yj<2F zc$%D>v0m9Zdtu0$cl6pec3pyAK17N6$zg{Cze{Gd07F2ngprHDo5}zs# ziBsyYpgzjnMWKsA%Od7OUlW=U{4XEa{js3~cS6uXVz_xPeXrW^RE2Z>zVh1dy zBL|N054#|2=VK1x*&o+VS4*`L?AaK2hi0gOI24#5$`?PLVrfmZ9zZYhZ)~|{9(L?j zvF*Q{^b6Vmw|npV9U}|atzu5Ramq9T=W2>sNs$#)G!{weS>&KYHXgj86R)>QddObo zFdHD4Ju4w(ex4n8w)x*rEVjk9ao#37V!^dJGxL&o`^*kGNRXgg5)*Qe6F2Ite2wQx zdePKHqdv))z4N__nKiREQJP;=%>08!h0YDqH+0jPV9wSE&H#7l9+1oE4ya-dLFM<# z;FyqR`6c-dCWmhGTBNuvKIxG|$9rZ*XzW$-YbAPESQz_2>c3TgRGLPFoi_PwA;x6OJzm|i zTxAw_Po$QgLtpfNKsG8*DBHMY5{;^do==C%~PEJ3t_xJwaD6nk>Jl z)Y5r$v9G#~?Ds&*!3BZ>r8*r}AM{#bgAF7C_O&^-;lO0gF@~5;N9;@t?eW`Rn>GYE z?NVf+#6N5Nw(u=<6Ti&&xFTP?K#-?|ywCy=7yz|-Kd52pk>26P1(g8`GF>4$)2vA3 z;Uh@qu^YBV4qb?>^K{Kfocd^b_t=Xf4N$N5xo`A(&(WFXn8C!#odGc%gP}9)EJI!Aa51m~x4vIkA&cY2>7Ar5LE5 z%%P$SxjjIyp9RbYaT8(DJ( zW>5|N*!v)7VMvKi9-~LzZEu;8L3Y7LvVL3s74P z+R>b#7*NTsr=l}~Z@=-aOX6By7PpVSmHigsz%$7#!oB()Q_b-sv8T#|_cju@O+_O_I^sp#hB@p!(nb z##mF_Gdqj{_RJ&E zqvL|Kkl_`SLVYE}Op%_|uooIvv7Z&oH64K-E3jc?56m7o05q1dVWy#JWWzt2KT;T$ zB-OEx_L3}f{fbT;cmrjJLCaJL#S~FwD+J!?r0-vZ5WhyXBT9@5C~^nL7~uFzl8s8S^>-MZ=ZL5=*qQ5A0Kikspy4ymkr z^>DQIV$=KoTp7eHvYJ;R+r$`pE0L35^>*8nJkvJKov!WASEA1MIsy<+1*oIhCTb4pXDD78Pa}Aw1~-MBtE?7o*5)#UIh~BW8x3neY0G z2?np^-iOjfWYZ{8W3WI^lYh7Q~Wj_y+sw6bJRU=h1uY9QbN{p}P}NRfOhdVgS!05!}2@!is; zQXOAd9(_}|TwWXu()HNub316D!YxOzPk2%ML>3cL?Agl&#hvj8dMMm=hi}NmX-fo0 zNTR&R&k?Mqb^^W`Lt)#gkolLlo9aQZ=^=7rZ^nXpMBcYAiLT)5QMIsz8OHU;>Vi0(F|=*6**)DAb$x@#U`(?3DQzTyoiTA1#iUVW-GDe^fuP!}!)xViWa+`8Ctd#tQgAFG7i2lyN*d^_9)R0|dS}A|l%d!T zc2hmHyMNHKywlGBkiz~SSCC96j^$SyO?ffJ0LOa)6}>`!SC%ZlF>Yz(orr3XX|Chm zqKhN1OXiQM@;M1Ds~S}sSreKNR5){KWQ<^!k4Ckfdxnz`1Z6F*D<K+(UYv-r9q(yYLdlGtWn;NgfdE~z6a*)9G?C4(Hm)@7=(c31f<9Me$QQ|9p z49b1wAd}N)>VX4wy5D;5_%XfQcm~ejkYd?Ao@$4Jqh$FG*b2d_giWL;>GjOqzpATz zx>6SKw+IJ$kHn~?0LUFS$iPQXU zOeqe#GvOGV)2Noh8GO|ZjfHr&QFdEg?&~gv)c!IsC+d@4g;vSf~E-S?6A|glG_eLTJVG4S|^8xgrO_cwk3{Sc*Y$F;?Cn!e|_FD;%cv zv~MvX=O?N0R$YeZbU5lq=*j4^mF@h+VSVx%x)QYLA5SUa^t%+ywj0K*gpSoqki8(4 zre6QyzZsyDeqHdGw6KGY6DNLpji95Yn0AU>qM|Dz??@_N-@?fdU6UR1J;v*lRdY9b z9V4Bd+qsK)u~Lm{5w8ywVshv2PsL++Z@2+w2$+$U#W*j!+5oLLP>yzvE~bpJ)-E>n5c z=iM78HbfS*U~y%7K=gIG&5DykHKv312VRwS!VGQ>-0EA%nTtjyL)T2wD7`XFyY?aB z;05z%uS=Fao39RSZF@2!dF;*blk{pJG0hCaYjg5Bd#CpKHuC3yhzPpQ)w*|rVYQOV$U5-~PmCPKhT*^& zbS|xK;vEMR2oTd{h+;a?MP zxf>Rt%5T1Rk}P-Ph+n?ZRHsqQIto75=p?9Y(x{3eH+z%|d%2~YrNUlt?+wSJ2pD~O1eu*Rb(zjxD11qyZ{nyf>L#EtdPP?F3VEe+F>qOoy zNCafiXE^QrIpu;m_vF_mm(Z!45}*t0bH4)%3e z=f?U&NH&kIlN|@r4fV3o8Njig?3O3NW@!|KpM#IAig-xhqE~R|*u}80M9)({;>9mm z0mbaitk5%}{KMGrpAR^=8yAcxm%cVz?Z-xz=qAN{Ns()i-WC)G&e8qQs1DPQr{=zy z_-1Zk#+wHv^@=$eZx;B29uc<&%B%8$J?=1#lurl8#(5{csZsTDcL3=VwpCvZUrnC4 zEEP5KpSbK}*1E@up15R@5)t0{*gF-}rq6S(Olu?Sm8COtxqV!X>J(EAdVDL#JqCH3 z`(r=%?iJV2J4JE2dmE*hL0FunQRQxl3(JRW{jp|@cCl&K7aLjh@RQdIL?%l!2-52&*FU{qah_3(`2HE0(w|p^lIeuV#fdJf3NlRCuFp zvC9|!`F{$YvGrp%KdS9#*_i?2tSd&{mTbv)4EE{T)Y_Zmh`9~869@ddjI7XEiaAY@ zQ&ecN)4!-wSq|v`0K*y`J6)} zZj}C!XO}-)0qxVs_H8m+_>!!HLJGLOCzynr<7SZ{T96+xahAyzJN!mW83~09Z z^4p2URdNV(_rmWk`-|o$-tyd;i?2BC8n^H-uf6-K7meaKpf_R_0P+}f4p6(xT_Z~q z`|T58>2Gl<8PuflFQBI!aE4@?1P9t_w&LGV3Gpqve0TytNbuM-}k78ZL?|HguT$j z*feFVCd_#q)5+ua9uF}f>FW2cmy-Nf%&|Uf1p7*gsh~(16^-un6=0S@W|#!dX~>pW zfty?}n?FXQdiZ9s5>@spqssjc(|?<{eAY#Amuw;F4$wm$X`lzL?HqLNU=Z%p^;iBp zZ_mtur)&fG016!d^nsym`fVBLt9?G0Z%d}aiSwlvG8NGC0Kv&dNfrTzdexY6-#i)^ zp9ShXsFg@`#mt3$uwdncv4M=0L)zuG^B-^S_B0rk4{A4_A#0pCQvpJLgQ~bTQw->n zY^0*K;KySndaEQ~tm6X5DD%E)WkBT%YM*>HmIHT9X_Bt;sP@Bbm7YLctCPZUGzi$t zyU_}0m)W&Eh-Fh@#t|Q}_J{ZHxf)P%_>G#slUQ~!XD3c{ZZg7D62+{dNCN6FRnuvL zwG)xo_`lUem0P zljUiOeE$S+wEsGztV^@1xfOCF`H( zot^n{JQEYLotw$Y_rC@0m>5d<%e+#2rfIBpvdzZEdLS6tl9wLXiWXMZg!;)FM@u;d zYx4Ue_v*<)cGkp+osldfYqEx75-Ac7#iAgyhU738ghk;P{{=5(Ag*UKCfLrMwHRVM zuC4bUzSk!eDINx%IW_CN@$m1gAnMN@zLCj)@q?`R;T*NFkNYUE%5EgZ?jj3kfI6UFMx zL{cEaeGaZ3@l0k|0x(N;0O>~_XQg*qSWDFDsA^%Sylhg8pk90s+EV)5o1}#tJ+pJV z;tXkt($YEL$#%+*F?k^Pi4vZ9^j>8TIS)C_+^_|smxO`Fxe(d^u>1yVLbuW-^iDwJ zeo_;JyQL1VJ5%uqeCHavQKrT$=NYmR7Pk!gtT4-c_k^9G0*XVOj=Cs085#?SPvGF~ zC_|b@C;4TD>Hj(DiW1HtSInElfErIlR1RGkh4&tS5Ga-iXmo$%FCBqGV@wY1h#d<5 zb^Emo?-)?{aZBp^W@If+QA{I6j#JT41OZJ$dLFr>W3-avA-xgDLsDGz8svTM z1%hUUE)%plaKXf6IaFo)?}4iB_{cPny~5I|yTN@PYFtqgcwIDf=VGR+S%I=yr$e<8 zsQH}?+^N!Z4;Z9ORLkj_a+QO#pi{=exPL+@ni0KjMSbMMaO~ZMHrW~7^fI7)Y3$q@ow_77DW57Ya71 zK562w;4Pf?nD>Wg=aS{8=oD@~*sD0^3T(~JfUZVzSR=nGbTuiAXq1(R`h}TL(VFHr z@Gw?TV5Nmdb=w`gBv$ZBM4ME7bgg&3pGI{=uujrLANkH@9qBoSlK(ny?`#YtLmnkU zqe>aO!}HvXN8--!V=(#xbRO%sk>_Vxy#NpAHv|oxwcnCWKBG4U8lV-N@sCv`*IcH- ziGwMfDxXxKO@YjuIFXm)pD9lA>!vH|`3A9Ui}`qKKY643>5eY2)$DEIs1)(Q%*4uaxTIMGt_-e#!3y4kJPKCCv%%7s1wN5@H&zi)<*K7 z20AXN-#?e`nW|f$VlThxe3qQ9?GDJ`*1Z+BG5pBMw$Ckx{5V)-A09mq?u)Ry5{v1Tw_!57KzU4gd-a``oqs zE?F6{wcnVchQ3nRUdDM}gmv%S40Qs%vK zv5sPrDYBZ1&T>Z~F8nywc;1%+V;?Lpy2WK4GElVB_{lE!m}kc&!Srioa?JAEZ}PO` z4q1Kso!1>bme&10ru|+P+N#dYI5%S(cN=_T5&Al@F7y+S9zY9~c?$xxK3Y0SxtUuS zaf~cdsWa$iUAI(1XqaB9_Bp9kcE7^t+ktB8L84qvEV8X2%q{S6zg&~b?i z8p=xX{A{yCkEO-I==}5V!u|{^^f<5S|57k}g{fglp@`e3(ZBj#^v8eHsS*y5C$j&L z?E3_54KCS}qV|{(qg87s>mW@pnd^iB78;W$<@Y z+E3JEz7zY+8AglYDvC*<$Z{$g-P#XLXrzuhfMR5-3&oEnhBtr89JNrMXH6vuifR?dgfx?wo5--mBzBv$j`yiwLamxhfp z736T*b*0b!7YaN%(*Sjiwg@~?Y$9| zQJtW4i~F%YkWtt17GrH+HQh2>y*R8J+H|v(I_ITLj)h$rbkUe}1%zav1R3}A=-nib zhCuLg58y9$2#%~ri@60;f6fuGx!ipAyFW5mkzbVjvX3-5v5$Jw$S$-|%teZ{QqfqF z)5mQJt@6<;uiT=GL1iJ6R~1?n+8fcxPZ@tvzD1H4bw`q}$PvT|HL8lpZJv4IDdQXY zc(t2zN3tX+n=YcO-1d5|BmH9Jdh8BJ8Ji`_;TF+7q!fxTu7DKk{vc#vJ5H+kT2TRz zBczN!!|75zAp3(7U8_TPL>7BBctc9l4Mv+$BU=ma*T}2lWpU$Owfq`7W$eKC3DdQF zYhY#%)L6!og)3ITjMYLzjlIv`V@kL5HL*w*$eFYd8`ABYFjiO~Xc|)|?sq{(x2yC` zV7!JPP##^&Ss*y!cY4?=3Rz>x3LMzumP>C6e{9+}_ccL>g^oH6rzL9f*fOHN5Sq?B z5<@tD3$Wv4gkTUoTdXAwL4EQLF9;r1L*_aME>+W7elA4Xv&LfqA$rvPE_dDM3)-aW z6>gsa?I<*_*K%4VCwG61aO|i{gBO8sbkfHfGxO4EWQ`pQre*Nym%gbLH~eu_ z@?2YdVCUWbS@MCeO}Y;);iZs(ge0WfNuN~_eY9RzEXBRh-7b@D#pgb7i`BRocwm@* zdxtB&|EdRN+jZriX<-fCShf4afp^_y*HZiiN(;&9(`g+oXn_sc~X_z{Yz}gszcdj#K0)*2j0#*%2sTn69Yi-*E?gWQHB| z-6_Ga9dZdKh&sG5QBm$&<m%BaQ!SpyXY>*T0<)98vtU5=^R>MbvWEPv}xM@^^CMRUK2(6-UQ3cz4UsN~+<# z*oSl+Y;v9&CbqUmBjh)k5=A&|R=@)F1MXk2Z(E%q%4Bks-NG6u&dh=62re<{%1qAC zR|>WtC*IFus)sWCke^}Y){LBY;l1}vf$>U}_F4c-KBtLa25hYtgbz9DI_Sl_G37!i zFt+J8!Tm17lrkEaMHXNgP8awEKJcfWvcqo}7NOR6ujP_*Cl2tQG+I{bD5jPo)l@VJ zYovv=leMBWzr~}n{U7_|f^i3W?CSYTMqQn{(?5}SSD6dVYDbwy=Bo5^J|-W880lPms72O@M&GVB>{Dz?I+&|#!W|jURE$oWw%Ip=hTMWnBNGHhw=jIGaaqm)PjlT~~ zpN9mAQI1{3*cc)cGz}FnRz3IWsW(bZsTbKqbDh}dtuVsLQ+01_K0#X8W)7Co<DGHoaUg+lK~p52bk0oVW)hpu7S zKe|?6Di+2DDo$)tEXar*49etWlXiYF^eir*I^>lz@5*ZUOFc3pG=3m|$vw-tDvb$w z1XT1|$t7{a%u^!A8j*o-^bim-@Tfz7e{f*yIa?6o#Kyz|LQV=dbFo#UFe1e*UD4r% z>Kqum+Yfc+^U@S2#&koS_!TbHIzS(XcJdnUWj4gsSj^oJ5aLkVG4Q8vI+XVJm+L7z zZ|?!9oDOzAvMB~Aiqf&9;d6+@X1X?oUz05(d9SN64Y<(@%H^PJC1z}CNa6Sk<5#^_ z8Um4!R!QP({VqAmjb80=lg9B%?n42C-Uk_&Hpl$7?9qDMTDLBG9O|kted-^a9exnJ*X~T(Q8@z=md6@bqNU z$bT^vwP53>I`QITLGB`3S>c_|*%1g_HkjW*(i_kzhFaT`^o;;ma5j6Ca5}s&t$1$6 zu*Kil!(&*`G2&X(Kj!fbfst8f)*c|soH)L=#mL)8rI@u8Nur|b#i(qdO9$gr<33g& zVO3*w^=7V~Tk4SOFt+`$o$le8Y?j3aBmZ37bHKFQ$7$0w7TS<H_IBrnrCHu_~U z#jN0B@~k5dEyG1#yjK0sRpZ#{RYLaw%~ubvNzouk=e98I{Buz7*6P2~Z`XK78WVFf zY&BxcJ?y}9hyTR5+Cs?d#D1v-Lf!>}Oi?%e#HG*up-v^DG_XuK+-T5Es5Ovq;LKYb z`t*OxK6}+0&g;dfuZEI(u+zYo`sL$)VVG`nAKhK9jjd4i74TaEkQWL-UC>X;MIZPD+7B+UJRxcaorZkFy3 zs+F`v)rs}tkd~;F>Gw#PSHBCENh6C7B=`bXhs8q)>_zdK&?>hy!J1HY+{87Z+4TM( zz0yD)eUGRULNndgPr$e?b~Iz?9!ws-71GM-lgV%w3R-RmyAX;St;4$B=T14kXDo1y zS4N%*$G;B=h#rs|QY?YoLE;&F-(>lDg3Qm@6^leJ`p#*TI)DwCO?Z8VT)V&e#j&=} z{0#{F{r0+*WHUPiI&nA_N|^?gnUqmXF-5jf(Vg-J?~9ULzXZ+|cWhtn5g$;d1A`Qb zt=;m<*5~X7O)RUZh$*bS^JwgSoQeA@$EW(T(MDZEO-Jv0i_|WlB=WDdS!XY!nQ;eg(E>g zM#vI3H4y#++f9{E*>s0|fad<4Y(G?UV>sd*tFNr{ZcZN@|IL@CH7HJ-CA3huST8OX z6bmqD*YASbzF4W0FQ|rkXOw3^f+c;&4!%0BsDZBF;XyB3=)&UT2&6tYCOhah+|OM+ z!LZug`%t=wY;xiaPL0v}QA#n}DN;y958$8`=3YHeW`Okt&_;~m7d(Zr5R6c)<5WTe z&#H(@3`&4tuR2imqhQi@+0IG0O|mnVh5#B1PRn!iZqo@?pc(Pj zrpx&r2GIQc0+~Zrv2#(JIF5@Gpb_S2x86Fa%%#LK=lXpL%AO3Sub789v^d2K? z#fCBQlhoKrm_ct9+>xCIy0)S*^Ib9WiZ%DcSD3&W9rgu*#nFE7{u|GUCOEBQVxa_ASS>D8LEfGQ z)5YbLDp<0xgkbu)T4+cJ_pOH&ZLNP@Yu;4@^-{%(k0^KG= zd9WScN0pspyFijKlc(h;P3xhTfv{ks^szr49VWdKYB|lS7L`T?j0Nz_9%T;QCf9<9 z4qo5K>Gr5{I~4wF#oqLe;U<~@G5XK#s%uR)+zhv$*X5ehPV!5C-bl|~%XjGTQI5OUX`|cE= zgeh7VDs`(P4@KvCUd4`nEV<@tCLonJh2# zU;U=u4SLLbCdFDtqZDh2JSQwh@EA5sUolytTByO)s8C)t?FaMfW)AaB>4Zeef4#mFbJi;=h8k;HGYXCkjg78l$nujSzK zwRqeds8u2HVOOQeZh4Y^mnJDzuT<0ZbfE+bzT5fTLady@hhWE@k*B0 z$o2=~sY1?H!9~sr9z=KG{Sb@~xorfG_4QrxplXMMVFy9ItXlZQMa$ne{)tOP$UPuF zE%2`f62pF%61NgJ@W<8#^pGYf4D0h~@Ybgx^fGPOkYJKDcQlZ!G+nye27}D9u`;tv zj(l?MjaY93>^^K5(Lz$$889bKKkYIyUZ{?+ zimM>fs?%M^!mCxJkXs~FYZNC)9(Zm!%6$JOK4{_4YV1;}293fTx<)c@NFtnrXUpts zyFknA$hMvAol!6Sh5;-yEBRfd)`@M@C8OzYqL>pDIYvdd^P8b>wK57b(#SrkXE4Pc z=xpV+$sjP3<&otP3;$tIXMZ3XgRPRSk^_PRU}+$Qh7o2Q<0O!oKER^72jsX1zyhRQ z$Lp_h*TK!JqmNUt0~LA&x)P$Cmu=n^SFpg%ht!7GQyC zs8&ht==RZI6qd@tq@a0-o&lHb-lI~X; zd|dH}cQ8VYQp$ZEctPDpdVyYB_j1kXZg6(5$oKPBdpr{B>AclfrJ&c@Bt1{^f-}Ru zP*zCUliWU?o`DzE!^g_?M}OPE@I(GRxkUEwwp4{X@nU14Dm;%~0h_8u{_e=QptA6M zW!bdbZkU%sFSt{_!S&&kN7H+}s@(FuvqbC1W4IdI2MPphBT-rpxf&2AH>R}u-*&?e zm@?S|$Zpjr%fjsd6O(zg7ivs~wfS)J?{@4mU5j3+6orKq2e;K3BCY%|)I(tP1k9@} z7FeG!Pch8vjrA*z&9ljJ0L$L2uR@g?FyOh)UR_DDzcy>pextRhoMK8SQba{xf33I&}*p|>LdFVi-JWh+t0LAQ=Hx^~l_r3v&mt1mzlrxB{Sa%tRXBby~q z!05&)i>UC%d)4bC*QY>*XK2?;1H^3R&R*8!xd02BKf=ttTR(kg$=5OAi54UeSbnhV z>I8QatRc+`tmB2%N16cY-^jg;#e7Xtv@hw37N(wmBVah}%#e^^X6EdD%Cc$c`Jm#y zpTV>g_W!toWSUD}Ik7iVZDc@-DF&)N3#jM^?oFYk0bSk&qC6UUP(V;qSCVWH1WyoY zArgWew>C;O?>^|F;XNesS zERSRFE0#@%xgft>{Xtrw!JGtV{9_f#{n~gckdz!0W2>MTAYm`2q7f=ugSWB5t%_F~ z*g-E4gmh)HN)NIOR8YZk5|+-fc*9UU9-1*zkPPkyUFsi z1e1$P6*~gg1@9*{LG`lvV{Su>8dOJ!dM0$!%YvUk&JeGDp^W8J3#;7lbsyw`zEB>7 zW5?MY8|U9{^KCj1D>FmU@5eTpZit@WCp&S;jfHJbjimcKm&8aow$ZBy+8<*CEz%6% zV?b;Q^roKId0C=##a&pX&I&6<`-AobZi9_Zxvb86+r&bNMpeT= zt|;WLo!G`LlQ@8|-$p1sH6*4`u=Ut7ls?Yhw{)Cgb^7ksU;LVE7)5p)ne{@70e$3L z2!@LG23+EA4a8t&B5Wm*-$J9plw3A}mQx_CNLQ3_TA3DUk2IG?xl4^IM$jzY$ZO+v z({ZCV`fc=FGz!GDrM+^8eVc8a0b4(5OV3-L0W&^>^L|S^Wq4Gri%Q;C-nj(5BzUYo0dfuqY49_)1{O-PA<(iHuB0ng3#DK8LQ z;XSH(?0IfKPA8lBc&qZIqL^mjjq{DM=&tpU50?3p~Fh}*2X z`>m9BkN-=HDqT?z>C0ox0-LxfSYXN2n!Nnj2GAY#yczzgb4F~J)k3hPK+xyDC$NOu z|3=gJvvTm1F%zl(SxBL~x>c}=NtQ3>RSWkAAw6e;=pJ1h*cg^a@0KlvZIpfsRsdW@ z59yxBN8)aJ-aPv(G0*15ZZLQ@q+MQFc45qf4@SQI`+toP8Vu0Vw9CID@$3wc6Fb9O zjSSF6ibP7h@e&A*UL+dDA_b zq)m`Cy6zL_eck6eZ^=Ylh<3U!nYeE3GX}QafK6&(Qj=EP@k%+Gba{66TIXG56w5ufY?72MI6*(`0 z0?TWxkTdd#=Fi;e2ITlJT)vg;Vz)Fov2T3Z2xs*a1Hju)MK_0@4y}-FVzzkfQuNVn z<148z#x7wxWnZ}O@CS}@&z3Ri%02!yH2S@_B=-ZazqZI7^dW({{MzJnMVG9TxuTkb zz7cGyp~k#6u!*lxUFED5B}IX83aSc^5!{2$OKc5~X9~lPLSAeOXFIo9kpo449bR{) zY7`f2%AVR4c&5gTCD;bv{vW6R?{E9O&hro8XuW9Jza4r zvT?LM*QOUwD@)wGcrZI&C0_ZHUv*|16rlghU9kBD{xv~c!qTSahUbSTa2kboBy~QO zREa1{^hwRE#Xqlm|KY4Z?C_tHGWMD`64dV;w`OduFYq-&b}50=pzQNI3Q?DKk0vgn zW;cB&Vvd~w!V(3h3&o&gJ3;`4vTQxfY^^Cd1;H)DqT>?avVg>~^Ix2JYrWaXq9jwy zYKpAXZ>{w`vo~IYo@yijO;;cXBeH%D7Xj?kDYqU)PaheU7fU9b%KEJV7uTXzP9)bx zk;O(t9bZw*J&N>F(MZ3I<&o;g{wcfyQI#7KWMt4`UVLu$F05MUzt8vD~i)WM_|$NmVBD zQrsUyP`EyF(bU9mUM0<{tnp=Z(ImXSm{&s=Mzk(nMxO3=bswEQqgHvHmlDzE2uIikv)hGB@9MP1`Bi)Id1zzl>O5 zmYp-FjhiW|<8F;i;N-b!R4Y826nX{hWz&-UdgxZzQLPE>ao4C&DI0w#jN#pU!?CNA z?z_z%>WF$`mw)8wL@0w)hWjgebfs*85#>E=6F|g z_$|Bk+&?$b)8I-ae;QpLBPtPm`xPP zprU&tkf))Ezsavv(nj)pq37daSGRW#9 zq{Iu!3Uxc1X2t2S7{45-fO|ZpgtNp>)wM7Gce9@Fmk()|J6Ktfk?|gWJ*F{|uZgE> zA)he-H47QzK`Y2gidjLC zrBrm4TZ#NAldKQ7n9LW=>l%$7Y^#{3%wpW5gd?7f=5>b`a%}B0(M$ zFJXm2tK^iVR}7v8Ha8;c3(_6l<}DptWN%rv7tO8}co`-uVEk8F#!K|u-}@WiU@c~y zS$lviW9OYZaRg+Gk!?w(n6(s1qN0&`Ig@iD@MK`p_=M?KrG*juNiFaA*p#s?(vuP( zcptmY8(%qe@i3VlR-q5$XB~O;xrb`6E$wAa3=RwJWyxVrWO-iQ|IO}bgtIlCF+#b@F71w-5zn_t_(B&`p6Hje(el(7%o zuT8GA>te>vk(HyG0|^C2auD% z$>iuem|w!AWA9ngjb|l<9P-{T8uaX$@YpvE*5bo!C)3DobA35Z>|$t)EXYxcfz56m z6|LWcBGm!bj_KP&)z#c~Al9zs#4}h0t3Nq+xAKwaw(xdZN3@i!tPWj8fYpf`&s6iX zlvUyCHsKX{JB?LoI=AFHFN3b~K}iv)udX8*qB^o+0)*l^!DB6pTo6+1rj8MG3m1}2 z@K)8K)m&6R>!ycbudIa3U@vqiuU~`?v)3$J9V%+pha3Z#et+a%Jz40)OHY;&aMn;v zB1Ph1V>?w#--Un&ihDrQR<^Q9afRE}5{U~b(xp6sPHB|CN6 zTh83kh@HD{dG{22hK_}@lPL0shtz4q{x-|~Qik<6`9sPlp!dB(Ucp(?!but0{+awIo60UaaF8*8`ig3wl2r#|K`zD16F>1YT7nZ^0i6cG#H_%mSRAXXb%;AFlaql z5|kxMl^hYQ^o!-hgcOL8dm94v5!o~5B)R8J$_bl;obq=hHG*=wl1dr7f|un!XU@Pi zP>2AY`9-`2C5qgCp)3?a?e&;UH~qNrt|xuF>1X$@8GVeYax3?JB5&f~a<#=a*@`X` zuo)k+#?-S1exg|PftvwCe^H$ zp=tv*`6arx3YW<`IY^#u%j`Y-ysUb~ZnCW!(k?eyHYMiv$UYrC-_gVoZD7Y8e6M zOv*PJ?jAH1YhbhYabja*LE0e$qH?GPq9;H><-}XAC0?isqDHqCi68N2C1w(*p0n3 zTw^sJEt#JopMNoG!|&EF5*Ze)EPhHGNfS;SWHp1{rt$3x)zjvRPH)L41(A8YpNkUm+nBiZ*Pt8cpVeo@bM zMO-#5i+rW(mn;-NVhrW)P>!Hy@;NWe4$pG$m7gcg3IsKTZ$*!M)CnSe5&%<$NVB8P zj93$yJlb)aY8?cJglN5eurGv7Mzf6Q(XO+-P1h->O%GXEc@F)@+MmYGI`Z~PYW26W z-udD!_4)6#&))cc$E;0M?fZ@Yeq-i_fBAgYsqe#@y+D#JpSqV@5Rr`eecLOZhm9& zdmCq6cqjgQH)fvuY5c7DKfOL{-ktBYyjAqm^B?5TQs0<)>b*_W`484ob>F!A=HAk@7BVDoCBZ0TWt2y)-X$!rgT0H8} z^zN}xstMOn8f4BkUXEZ+vivYCosI6byv-iRIdQHKJp}Ct;L!7E3XS(jm6XYrj6xmI zgFy+Q&7K>^t`;tLU9B8=53||Ln{HV5{as9^BH(U)S>hK#ieNR%IjW%Ft#MYuy zhXV~D3eNb)Dv~=2T0I7TCo3ohXxfUY=*RK}9(N{a`CmYBLJw~Vr(B#ecIVguX6k}j zR|#x8_`4&!<&U{n$raVTSv5aPn)RnWKfgQ^_Mc$n2QHkP75mnoc6|Hd%>T)zG^&({ zVSuo#WJ~qt^zt60j`ggVs!9@B;!xp#k@$>gvEvjIOV#P1>pG zi9A8nZPJs%eZe_UYEenAV6w{GU$z7*>p9n`$JGVv9nPz zcLw&*m~(}{4QM2Cl%-zniWNcn#4P@`8MK~Zu?)(^YWSsInVdon-jBw#n?`9sO!Ydg zZb8=6E19MuB+o~mo%hQF6{Uks>{F?c&p`yB)2oEWT_A|jgkRz}1a0RQ_*DxlVQ*+Z z+h}BvEFM^GK`oBpm<|IWLfDUBB1!1`m zB&Emhkna~IgwE$34?$`B%VFs2AmanJP(6}>AwretWH}nARoryHRKEJM*G8|h2(6@@ zkGgs}^eQeiP$kQIq`L$52kr|>00k{5tLTaBQf4c)l4|b3h<1W^E}=5RmV&311I<@B zTxQtmvBj_9`y7lwbF;|(DtKf1H(}>CS&g_s-p`OuEIu;Uw_5&In~C>0uW6t0Uj3C(1{0*xexfGxM**erU~9jMViG8_ z9Ejwi`h*9#>Con=mw)e-X@cxjaLMx1w+a(1$F#@otT5rc9;MHm{WD&5M%83_i5IX4 zVK_TmnGd24>@2%I(4ZT`8mQXs%~%Y?3aTS-egA`WTXaL57weXEL$(GMh3ohou)ch+ zLRUeGxu^l>qfCZ7n0Q8=r`xz7`)V~GTfH5+7~%w5gD7)O1`BWu>DG?e?LT_Mu#&XC zdo7ogJ8{7Tc%y@ARO=|FmLkgOGbJ$e2c_*91%j#lJ%nBF)c zL!?n<(Dh{C>304Jc)Im%{nIuBXQ;q12bRI-hJ=Pe53w>e)Q_*E-8Wr)*dz~}I81DT z{9BjqPL-5-Zvr?hyMLn#@upSq~AX_#FFIhji^d!pQ ziMZdbPbNFf)kQh+?6(+sCx-n%2skU*qudsbEi+fR^94_2t0JK` zr`)wyc2&ANOpU6=xL4N(sK1oPOkWV7^=Y4S1d@tsbZy(|_=$zoE!Ra;lYkj5U4b61 zmM*0FUG&9PsAarw+6JIIiignpT2X;$iJQ7tmKJ73uhO8IFn-AfFM}Nrt6v+xoz+K{ z9TfM|KUnT#K$0?k={~a2iGA8VMyT0FF$EOKgCbzbVqW_cEeM}2h3?tRu#1sex-xQ) z^1S~JY>UIJbd$6|Q0Q%FyDwTRWQ3Ok6jMWyDk}P{JaPI?RUcii*aG50 z1Funa1jO*@+wSfB9$KS%c2_n3e&76+{?U1w3GPzk<&2J|$;~oIR!Nhu%v4 zDMk=GZBO9&N#K!$FPG;14cGt2gQ_95Q~jI$IKto%RNIHWwpLr=tfx|v>NU6XyWH{*F%yv;k7D7JQEXA z1>)tIynFN$mppC}odHpk9PSFHN^x)k0_wIn-D6)+23-U`&oOdP)F^H7?vz~)OL)!R zKpHA;p2L!zEebo_?EKYFN0{oyJFRiCAZUM#fflmwsV@k3N8XSo%~oSIF7h*^N`d&H zkGpId{@fNm7sK_P^1G9lO+$i3d*k8xX&fR#o*&50XILR*#9zLg`i{zAS+rr}Qpp~3 z9*Gn0sLvW%n`0CM4KW9)=$0r$mPf;BiCQW>4Sac*{o}cM1*0~3pGTbM<0wp5FjIIjRm$0dEH`VN2Ic}@Hnl4iRC`nV=V)7z+3-BW(KLM_+!qri~N(w-z*Q zI=s52MSd{?U4$ib{03K`GC^e8;*g@T2$LaAH|wbw0zO#3$50-y zjE~&kjQ!KAo=MCa-_B1CYXScF60c2?`(%fIvV2Ysoiesue9LPePb<6QzKy$>w=uAW zt`-7?A{6+~x+F%T#d}c4hLnnZ?)9=7*)DOFTeWbdUxEnfTyd-x=03SMxo1Ki-7)1J zlwGdpXnkTrcKFXR8qp?f4FNJQvK~g)h60s=XDma;oI4!*U-?;n1Ka5SA6Jk}cGZs^=2cS^w*PV(FA(JI*~nP-zV7~%(P)f^9TY+D{UO~YdABcw;1uIQ7m;vpG0QrBT&G}fqOgmg?f3}+U4Ya$A#C2{+>7eWuycvR)~ z#7m>n`aB>{ysEhgB0QI?yf%3`=PXmpYm{nKEBzX!8zkkTf{7Z{p1`|w1KA)+Wweq+ z9-gxU60HK3(VAn;fh~~IKWiCS^mh*R$_*>g+EqbkNy;c-MjdRzb14Q=zL{XcRgJaijFHe-C{}(#YQswUe_= zd@m)e=aT8{^o9UkBYf>d=VF!fX>@l(}ybOpAGez?S|2s`eR#oad1|TmPwZ>iYj%oar z4I-R4Fk&Gx+ZEnM@}P!EeSwBHcj+3>bj4TRb-+#q<#4EBaWlNYO}*UXpb}}u_6Jt+ zhGi`pihX&>)Y|9rACA0jO6u*jtBM7-@hZ1_z#5PyxDJGsao#AhsOt`{kZod6XFz@3 z`vF-ZI7V6}*QWvml0&Y@hhxr%GUb$gCiR7*uv$_^-Ykvpm}S_{PMvqEom4q-p+T#W z+i{#?j#A_>6@7><5p~KcsS;6^X#H!(N6o6OvpI*K-cZPWt9TMJ3YeJ(2^J z7+nk$VvposfDGuukW5k0%=_e$_}0`6zarsKu{R_4=FAc3Yv6-P=ryMi*A`p8F|8+Mifwqk~yb@DA+kdHpQbann7~tam@ZLGHmR%aydH-c(hY{-YCS&NjeX8rAmm@0I?r z$#lJO+I^3OwIxZE58|ym$1UTik4Bby0W+7ouKH{r2!;>4YhnFF+dg@0Mqp>zuI9Y< z&6xx1&B1>yCVQQDw{y7Bpt;QNw17;+Iz-hy&O7!Vx}Lq(i8(`h^1-aEVdZ+F{nFWGMGbld4Roo-LOk3qb^ z10@9I5JZqe5tUOEj~N66&4_p)5gi8y0Tof<|NA9jB$14HNnoS1+s_Wk`#$dj=K1FL z{GQ+O{RVUj;)VC3s{mbH@Z3`7u4+9$XMU{;%CDiqfAtUW>i(d1L8kyE*YMqLdcAw! zD>(C5&pKmfzOcPAngYm-#aIjW|RyAaav z1aC~Q-Xh*h&iJM3-I?}ksO$}F2(_xLHn(fB~0vn zHxeRTUUP%f7jNeL=uNYAI{x#EACubAO0!)!WZQ0Gv6?9s`V#A@xP>T~)@&?BL&i{) z7+%;88aAmuE$#-t0Yx?NA~Y+z==Vj7@~Av2g(>n`m@D3|%7mCMR4}UOa z;c8he-#F@?M;D#zfyLJSAWMzm-!#ZHt!4KH)X&DN9Z(9lFR)pep-P5dfRC-`-}AW6 zOZEA{_|%Zw4OtGOxkv`QbfA;2w>M`w-cWlQlJ)`4Hq-_m{Es@)3@iEnXi`am3vYTK zSU{+fVvkUyjEXz|ddGMAzGnKbUR_J>lMS!`q3^Y;{|tIE|GxPDJ^V)7fHU)6r~ zGB7@4lrJTApVvB|H*5}Hr$`H~qyPSH%4>S{d8UWn3tefM-+T8X;QrY=H$CJJ`(Jza zBHSwA4ScQbt#|+Dj^H#~N4G{D4gqeUsB{NqXvPsO_QAr&DBEZmVH%1ij~a5I8d zC->|ldtDeor!5dvPO%VBC-t<^1rJD_)#jTQS;X9g&C7JHQMh&9jX6b3ys$FhL=2AVXRmuMtfR|=YWSZhEBSi$ zWktKsotQfMis)(t4jHp+WU*r7fFFzb$NGjFe*U_&!E>>L2tzd&|F#r`RsGSW6w5|{+#^Ii>gp4y4pq&N897T54+2U%q!Wq+yI}y1hEvwpΠ;V6+`)#pSPDvL*_1+4zkO z>aSIk+XdY|$0CvSUav-;j5NA9#Nd?(6s9NS#bZA_+xdxG4YMB@Ru7KzTW-L(%ipwh zirE#p@tL>f$i9HLJ}R)w|(X|@LU z5F|v`WKUlvo46gvek7R>Gvkb);ptyH`N8b4VQhT+RuRc@VK=qTVztXC_5ekSL1<_S zGTS#RbH#Y%9iTJjSF5%`Cl@k>#m8!m0*dj_(orf-^|>&=9CFP^f-3{E=_A3(3d{hY z1P5xAS2HK*T7I)HHW*|XB{#~1ON7~=tyu_1WoPb4oYC;~3HQmB;iYuKPfm7a(#kcS z+n#sA4Ge!IicOfoR^UrYfx=m~e12%ALPfFj+JLIkR%tjIl#*@qSrKoVnG8>OU0R#$+XJ-%MPGfaxQ3YU=V$v|2$ zay3dR7W&8vskj^l%SqPpFr=BO(z6)W1Z_DKZEckvcZV`p16>7lTUcx#4`givyktd# z{AzfET$37;6`4J=oR8meFaU}=qDhkVMOTG(&dxGb$AZqB(lnvdwg3kf=!xAQ@yhl` z&g~AT1#(@t@M7rx|9r!YoV_!cJEY2m4a_AANH8!vcUNvNhk^+)=XQjQ82eO;5bkVzbP;42JMsMYv3QU=r6ZB5iTQ?T`POttz*&{JP zvhwJgmnH8cDat)kfMDTwO-lvH5}6z(3-}Bhp>WeU>L`xNVb-$C z;2TlF>rmv+JU>6(FFz{Z51*}>u_Th?x4_w09|3!-m*a?aYVD%J66?b?ml{nME~Bw= z6h9W4XVi;o3%?VcN43Kq1{71g`!UIiBiozuoVJlP&@&BFa$ z#c6GDZ6EMOQIBTjYQ=V?2HHnvVl`l!Xw{tCGL#gEoShSR2GDbt@wjUa>UD%joJTBXY##d9d12JazWJ|VOM-Gfuv*5xzMzaMpj#DhKSymxId*xhk zwSk5mz3(Hc9w>3C481H#WSW&4sI^=o#A^lfOjk^tZ+OTC9m9TUJ5aT3npQGLqcPSP z40siV!2jmm5#TtC>!8XebSf#3h1>M(klQWbsOtoeAAcXvV=|hGaCF|s)@jcPNLH?g z-EYT$G}}Nr9C)&pt*MvOs)?hhIk4zi5Hgdi9Vj%g?0!xP4z)j|MSy2(57ChvJdn7h(*MzK@tI_d6d=sj7P7Cq3UiHtBK!#HXGhgSysWJzC zb0t~nz0zBQ!UVd=y-_%?>37~RO8X)`XQQ`?k=|v6eb3PS`{ZW)rA>ePHze7G&siXg zJ|d8lNwJWiPNU***^(6~F<;26nS0HrZ;qZZnxFz+o5<90j|Jtq;#=NG3TIWB{9B^kOMYp43@_JM{L`qQ8JU026{?jt#QKT@Z`c0MmA7Y1IK8+8te|Fvnxc^ zzIoGI*vI^4^|f!E{pXK=(4y8VZ-0H=S3Y^IMzAMlAU28DLmFhj`#hs|MzfzjdVDcu zS7-j{8=Cz98Pmuzk8_z3W_ywzn6a7UyRbcZ-@=|$P%H?L9E2T8c!I8yFAK``Kh2bR zwR`kQuPNF+I^^eJ^^3oBSrQMM$6xNsThvE+Xx)pM&5`?ozjXSkhb|l10j1J-2u@egYd|_;zxa}{#0T}lO}*8{(ZdZ-m2;7F>fE$qg{D{p z>Wy>ah0X!TXMLH+2%OKF(kb_B1Lt6Te#9$gEAvjdx|-y1^O{{Z?hSF{kv`55iY=qa z0o1gq3_S=P3AZC2OIoGqy%;?llb}|GU@eQxrxeh4NDfdhFJTwDOV9ieJ;2tBeY*{0`?i67C$LTv) zgUs;xO+oECvde`xDbRHt;q(?$EEt%*RGcZxeb{}=7Ymu=?%>l_gzJ@983W8y(FIc* z1*odm?$yWxXEzBN0t|Hfr#inKiW@%X1L2&Xp{fY45FwLhgF7b8A>Uct8XYI% zz$xoLJc(@sGnxzh|1$o4MX(t)b+2zMBN;A?ni>n#lu&FTMe?b*g0FY~TRzb z_%gaN1;)B?)J)Se1>`Xw1bt^+lHL*JFvY<1o#J^+iiD`!LK_5|r`7UuX0!&iiOT0+ z6V-&=rkng~C5ydl`5%YeAXi@B@tSin18d|g*bFbXvwm`d4YmG)T4$|=#pUGW!h4ks zl`Jfz7;@`X%u@W};yEjc~^Ux~WJo2M5R0n(w5L(6V#R}TU zxZ5aehyLsh1bo^?S(j`T9G%chSph^-nZ}fMI_XeumEMY2H*fWnLPm2e6i0UoG9ic1 z0l9^qm`;^0G7}=Zd+60MH6b^@-7&3Gb%EF8sZ}~Inq&`%KFf)FStM@(Q|J+Txq%R~#Kbr$bp9 zR12{d@NAP6IgAdpYf!`pkJmfgAHoT!+W)a{iF>l*miK#J^(sw{B9m^9fFwjp@Ev-w zFjtJf3LZxWMaM0atK!7A1Cum6;S(RTkK6py1B=KuZU<`@##X6?N1Q{kkR9Je#U)4L zff(7no9K-q4TfJ#!5GN*6pKn#_0!^~Wy7Ibvo>&P7}CmQ)90qand&0>NOeG_DGGic z49-LPJ=s>TaiRxz3W^-GETcVMCRi2s`SM~C#RjP%3O6E5G=cOPdb$%_L}gqF`yese zqK)shuG2L$tShNpwtvk=)zogFFhpi4432#?t=t1z1iI*R;Rb1j>MF@$CfIAWZ>60U z8M7N6wzrT`LMJc1ed^X!Z}YbDPt~dC$tD*Loj{~yM8Itq#p)=MO2swGp!#2xCf*HE z*xrZ%Pbf~ua4J?+K4kjAuLMrs1u$T-oNLxWO<~-xLBk`aQ9c;yNo zK^cjrOmYic+J3Tubcyu7%VcQ~=y!@W8Z&yVz%=r$gTOI9&Ivt} zOA05GGZu?fO|c+5cNk;m9g0gqAA4$2VpC#~-fJg&C)5y|Jzdl9nWOBU1B}LW%Zyy{ z9{%-NxFo!DkRr(mIyUVXWK_;gdlU+G?-<-U6f-H4m3j|$kBP`G+ zid|2UwN#ve-UG!5c!ojmRMX6uGDR4Nb&9p%f)jh;%4JSxjO7H^NxFz5^Q`N@Uy#_G zjSBIK@Byz}VJ_&7w1roS@^|5Ig6Pk_tEL z{%NSB9`U{0PqEN{T0q5NNEWu15w<@AA% z-1tX-ZRf}N>|xx#xh^Y{*Z<{`gEPz<&AlJWR*;=8ywOxy0J?}`K?f_3in}-4;0G1a z?nO)@xe)~$O|rsNUxE~@@v&v|2IVW%MPZ*Jx!WD1Y#O{#qOZUHUvx>UE&YTAhHex>d+YI zyOQGQ+p-O$a6ul`D7(ZwfXd49sJ$MWyw|eJLyH|eVcE3|FlY=2v3F{R#&JT(q@XPo zYd&An0xk23!mtZOQ|{3SOt{eHBQR0NfHi@Z=mOsRqB1Cls+h4YG&d+mk`1Dti|{y) z5r6!-H!6=B2ZQtJnvVfjPh74~XL0_EG>=((0?f8#**nt`NE)}ypz9v`Nx6mPDWF(T zWXqxAdgXT`kauhl`odBJc6n(qQgmQmy#zNWDBKU&$AfkW9dI@&X6M=+pK9gTsB8&gIHs= z>|8i*B8#$_j2YRqPK;5u4n&wF|1~2B>p89HUnc@b-i5VuR?ubkObow2dS!*dUl#6mf8A#}PB5Og;O&r`Pk| zQwRampj_Z$x;Gx5S{T) z#N;r?3965v*nB5Ei;|qB;yQXLD36qiO;>QKA|a*wCJVBKcw zvR=qOiv!9|icO=)4impz1&vX%<-V6BMsmtDx<52ulnNbOC?UT`xFNdqYds4dypHt- zCSpsh!ykVn*t6`Tme)_(hsCfFpalXP06_?-p>S>`l67o**ls}S$7bX&tYhh@1Q0z85gvX)mW*VL8 z(<95IGb7VI2k1>xP6J``GN#Q#GqNbQm7NIVVnTzRV0vO@#u-Gz)9=b@YYQ~PBX8(? z2Z)XvJX|>J0o79@LiGD67Gn0fR9ulyzo%aPaY!CNkI7JBsVwrIVzDf?eHjGD#OVRO zymGcO;8xgK=*UO|Vs^c{GXnJVWOZ~NAM!sCIeRZ~ovb~o(nqUo4KE5Cmo17jWS)TM z(^tdG?98A4zUDXoYOU_zvd)MNy@pIxPEg4VgE6_6Mz8T+7Y=G>{J0UV! z5#H&Xf<^m2dv2?3-<-}E#cj8_du7g9YYJ_b#fA-TZ%q0oE3l^z3IvSRlE?svhM(m=o~@Q(Z0o!Q0RKkq1|DSV3HV5nK{uHHj%p=fn~@yFPgP z*V(S=oHmG)DS2t$ssLuP*{4+--qDc7+WgQb6HMtj|uR}L0hX)!8%H#fJ z)aGebf@XD#x`%H2`bAIx&zS#6aCugTJbwD-X)Q8H*2t>(3BnIZ7dR~k0}lG{4Y(-E z3fmKNMFN`N(3{>M+Qmy(RS3FhoYmQ2u@4EOdm$EAornUQijys&ejm8E`T38$0cNoN%tND7Yd<3j}Jz0oXKg-KagJT)CB zkW5N{!%+LYOE2%H&(eFnkMnlDx}W!f@|urcySLkjNUAz7x}ENu zlO`T?>zmWW_kjC}JH62^9c6Mm=`{(rn{-Ehq1?Bl{gI2-NBD<>C=il(fZ zoUG{cyAidKYzJ=H6n0tUve-`~YuTeb2)Kpy`N0+@$1BVDb@a-~MN^7Jt0p^pn|gYE z$B3z?Cvx(YgTc*bX|r2<@2$YoWXTKT*6J+W+Km*ujv@(ETmuO376liDXxb!to+eq* z3;jn%(X@#S%$ZivUQjrB&7V14P-yg5t(mx7)~{`_Smqjvt)$2iXw4zrK$p}-7e_xP z*lm^r2^=kijOJ>U*CMvfI8L-)175|^9g32mZ4f^>Avwl(f+?{V79+QqXTN^tsy}l8r~PNr z{^-p={(K#nkw>`MH;rD;K8jsBOOql?k+pebaaEdo<}%nZz&-@oF_DvpoId$-dw97A#$-{h8K&w6GPXjW68|ALoX)wu74{+F+1A52!ii|$WJ_M! zS3B(nF?%URZn_W1$`{5<$+XzMwoq&mMK(}z7)%#XNCw!3hDAMn_Z2(<2@*xeM#q~dDC=Oz0? zF3fx+flSN*?>$vgfS!T)8kDfr|AWr`25F~rB+w(QS{wP1XUjb7zkzPhJR0A`3pIE! zzB2W$Y*px9$o4qF{}^qBxUB!^)?~C<9lyrOj!gRcufKEB5veg34&vL88bjgBE*j;b zA=ez78Ug}d1D=UYF&LLd_uRmCL7KQiv@Q?_l*F9%(PV+TSCNl_HtAztfkv-n#T6fu zaCC7rdN`XSvy3#g4`~BNZIm5MP3XN>wn5_{NKAX}B3Fq9A2X$?n;;sTGBrB{2Wge) zEgK2M-D>e3~e4iWERoE_ju(CmIj)g3V{7kw^YNU0- zD6n>dn|(}X!!DELxP{?712>cO{o2b}zcFuMzis^4O!6@|*T#iQ6xLX%A3mbk2NbzS z#i8enlo_U;V@)TG+AUh;p!}daew($-{t(n(L8VNjjJw6(6kY~QF^S;?q#M#onv0=Z zUcM^2u0Y+x?3ssDkGz4rI1@75P$45oWKeTZ7m0QS%kQeBoq}TT0=cn(278^a_@w$g z>32p?3VT|Uc!j|Q@_tf2ALcq1eOQ&qWP4=OSy4Grre84RcKZv1K{!;$=)jg93I~j1^nG!l>y86`BT^Pu_(5IB=86568WIwgz_}o1a@DmG+pb7>Gmo2 zSU<0nS?7D6e>~*CJd^Zbrlg1H)mJ2W^u1SlW?WaVoNUs%Z1+m_9#EvPyXg|2P2t5M z2SV1v!Kf*GG5?C>m^cM0DszHryn5$fjaUp7lzmW`vQbnK*yKM3kH;?4F}fSW;VNg# z!7+{KveZ7oU;q8g@0&rTy(N51T1Jxwc423#*Ww#spxAbbT&Chqlf6;(sz(ykUu~7X z=LH;yjk1K;0iesR2!9Vqdki#22Y_;}Q8wHWpADP}C@GXCHnRF-XKb?KelS*9mGa_a zd!wP`>VB|6ddnN@e&N9ZukNW?Q2eJs>DC-cxnF{}f$pA)Y*!}1;H!#$g3T;#2_zlHnb^31P^q~@JNbVu}GWIeZ2 zmkawY`4&5ej$%_Ol0wCyAB_r_TIFu9ijZ=TT`|Qf4f6XWD-7~38H&YJhMs)2I2xlV z$K4H5RGXQIQeitGGHg7|Ke;bC*C$lJ@vpEh1LsmdnJ3>fgztMLS$CN74! zUxO^$15HPV0x5u2#dglts|$Elz;2qu3@9!#TIDA30grNzOFUqC=+w6hdoVV^}a?^zqmn$tIxeqAz9z}XFlA98{imX%AgqUNw?|B_l zVSuDnit2I4Li1=0)LvKA@Y8^aqfdSgc)XE)<3rN#xr(I3)`Xyv2SiIfQ+*aQJErv$ zO=(CET@%tML%oj*+8`eQ4OJtt@;1icSAt{VLSOVc4`Vo`iAH@pyrhf7;4!AnjW>{r zIZ@mL8X^^7a?GPKG@CuMCZvMyA(a7b;di22rI6n8UM192@UXzkRND18;%QzphGm@J0NA*NJ=R~18c>+Aw`D`e;4d>?3~r+orYFUHM!#FBb-ijUN7^(|1BB1?$5xXuDpR>6Hy~ z1i8XyWvL9!2fp4IJg#7{S}so5uzGep-+p6(e2u_tL|!|;D~ zb?4f1rQvO(Kdn(~v0yF>Hr*7ac4?P_8iKIW`YJ{d1>aR^V zuKWyYy>o?&7EM4B&)yZc^h3wrf^ae$FUfu}&0t+A@Lc&FHaec~1>Xx^F|8#WkpXfo zbeE`!zB_-^i|SZLzfNyW7b z?lDP$m;UL%f+dZ4{N5L}1GOu#pw> zI>-GJ`Q7fwY1!a@L!$FQCuEoaHIa{0ViR;G9t+aP+s5Gh|6UpS_B^vS@w|H?nUqc@ zEf&_Kj$&&mQvIaZ7}M(*Uw{I?c0od{UanPkDK^9)r@@dL_Ag*H@-?4TOfgj6uXk@1 zZ;CQ1s_)P$so zFN!XT_IdaD75hK()J<`uh;S@RFuLg(_3D^ci<7Z&-IKtb#s6|B*o>sQ*Eg1t3>S80 zYb+pMLa`7b$wy(!UWEa2to6V%r&ZpU8))4WV>F~wwP{M12#B<&whQ(~?PW8;o*-bM zS*VM~#vtqkG0^Q0N5RX0-C-+(-xd?ivTllV4?a%B^O&(a+Jl_%JSqBj-}~ikGi=Ic z{&6MQ8l>;L7PoRL7!I1lVb{B0OKz8?Dhrt_Cb!idhD+_*7}T z09(f3a=)jkf2|rNZoxy>m~@?Tfw_9WT_Y8>iL+ zIYOs=Juq6LJ%C67Y&oU;23eZ8SD;m%nz47TUY)6WNaL$qabo!XpgMYC3fV7Nh_t*; zvJoSe)Jf)I#Ei~)X?tgCKK4-X-1{)TiIJBix=6ixu`px4Hn5$fsM-k@R&Nn6;ivQJ zz$LyZFOlOFy&8z>6ulB1|CGFfelqT`>gFpY@)I!z`Iead;wz#LS-l#Yu#**kxhyH= zHHjKzA52YAISn#SLdvlJ;_fkx65EOd>Z%nBi1;VVxh`4gNoY%GC^3T1m*R);+2u09fSeF%9$HuF@W2lcu4m~ zB=g$?SIaa=k&+a6EG%hO#!TeoPF5s`-3)1-TLg`y9pJBw7dGr!{!cfK(O!%bA~=2f zC#~G5f6dPf7S-C-l_b@L!E)FFEPE*yqz|)AvLD6KC|`G!mnPOlHpnYPrrjoit@PP8 z^B&o$8j4B_+AQwlB?aE+wX#>e^~x$~ys!d#Eepf zI`4NhbaQ}SeaAB`EM>}WW`NX-Tc(!C@jB83U5dHKJesNFpOapn+aSsUpZ8-$c9dS7 zK@!84u=#YiPY$yX2RF)Y$g*g?I$c~(Zw_dplbqIR=&*6JHdszFRt14+n--bjQ|wAZkAj0klm-)K8oDIq`9eg3?pdRLu!nYVDY0?wG%D@&u*&EyPw7JR-s$WRSy zKPiFW@2#)~`3B(6gj#Qt1X2U!71l;-Hb4?QSy2mS31!AH>v~lc?*Zx7OjQGY2!wLR zYcE7=;*g1Qk~tb}8o}9+Us;!8y6kD&MsZ~UuSEWszgSf#zRI2ult}tK_wpX|SA}kx z*6Xw_X0*?4)NA%G^r#Py<3&#PgSz(*F?0FmoonW{e>_1}aC3ZI_?(?#;rMK(*i95! z553aj9OZVQrU5pQR_XGWmxaO0T@VTevN-G*!A)m8?Su8o+Y2M3=VJZvIKIz~{Eyg~ zW?=m3&mBj|Zf?ME;SA$(3t$|i*!>hKf(&D5CXHPmP%Q`&;K;PRhh81?$rlsFI{FG* z%sU0NfRIvbW{}NkD^nJNmrblrON2VH2D?&`+({EJ%!?l2B`XSeI%c`=GNx0G{zxwG zlnNQ3oChvW(6RP>pMLK&Fzr$XADnK6&bRjb{Fh`0H|V(VVd{_tbn+-Rhay?%JEM2} zKz0@MxLQcFgIL4a};W-K@@!&QQG1Yf+>2 zdlHr3NT-!94~o1#CCxqp+V0SPh_iU>niA&s+r#Fw`G` z!wb-t0#B&(Y=!wVcN7D|R>Nuu9e&HnX?ZC$^L;Og88Py|ztK$AaWgP3?6mB)z*ic@ z?x4svDh~028ueJFr$JJ^PNEv_7U|J06MG^LH|tKArNNQca;p;Y9QsjSWu@ z!{wP1e5GyGb!*Klmp$#IS*-b|vd_cTA_F3AaCpT|)n;YxM69ay$7|mzTR%1a@3~ox zZ~o~QbG*z}<7aJTA=xlmg^LUOtpyf1NvBv4hS(1D2*9qC>;GU51ibQ?Rt46aS4js7^%0wNEO9`!*Ymg=drTrzPpaoTY%}ayr!nBeu6Z3Dbo3{6f3+@rdb;=7g-8w z7V>=QRNO6ZWNv`y7(fIMDxTH+-I4vC+hzK= zJez?IyYv~7O!j={pFM^PQl2`reD>51-OZINuST1Z@=v9EbflOYQe4+Ml12-p9HrPQ zij-4vy^0R-JJSQQne{V^6j@+PPLj((H==e*J0t^MEwc0S{qk-|<9x{WODn&y7UCu= zCvOi-Wb&9BvK1h;r&V^7oS+VXOD#ADLHyA}xwMraOW2~3Uqdy?oj7d`O zJd+r@YO>>-uR|z$3Ra(9a0o&8u#f!xKUp&ca@mPoIO=ADLlFB&4|<>X?SXdFM5e*3 z0i@)+=~9(uomW1@+paNZd~rxdB>aTdGOe3_^KPW678}~Qg{?9*de%{-_uxJRg za-FmW8_zRThl0@|)f6(VY&&Qi98nFqVT7!6;u?MHOnF{R1E$gCetEntFKbMfuxOyo!&s4R zdKv?4ID_N<#;B;^3awAwj89*TI<3POf8m|+{W)e=_Gf43?j?nzC2w5VVFh{T5l5S9 zimjl?VJhxJQ6oK|ND6rD+x%nwKPFOtuA@=fNv}4&{cEyfN5~z1q2eROW?sGIgnanM z-yq`$)h5Gl|DLQcJrGYND{^@^=k$3#8M_#+eL#+T7f4HaOJh#}lU1MRad$@nbmezF^KG)TrYl=2J0U^d-qskd`XB4U9ep`kGJQoO{7*U)JOqA4LY{$G$gT(JB{eLpRF?-8P7dg@bO-Vb9O@$9=!I zrm}R|LM9vJlx?D7@g*^q1tb|23KN+!@rMvb*&f#S$|bR0y&KBfE(@CI4KW=G!?dih zi=kUyo)8OUm&vi*cqSx@YkPYdEj{t)z`W2fvw^zt?OR18$A#->>nvQHGKxJwkzy(? z`IV2R4aRDf7&`Hj-Fa;EdZBb2_FX{zBgSI}ysku|+1ROSp-&UN{{$xsx%7sP zKN?2-22Q}3boSpO&N`w+=)y-N8(M@JDnt+Jsb2m@hQvT}h5J=EnB*y#R78k%yKBF& zjD6Sie-mP8xS1H!Ml)i;$L-~VS<`>@yw}~E$dHZIveHd4C_XpQJ)y(PW7jGV51zoE z-0X$!aE{OOzL8*mR|SnMJZ7KroV1@n5y}DIQ~U(pgt6%lud!XgFuRw-&rHNmZooLV zZ7}-tMY?kj$RzK29uet;I`$ENGcR3L%dcjd)jj-Gj82#im60v#m6Q9UJLV<_teTu6 z-0zQCLJ*4u9VuPF!LWr3`=jf`yXj?$9ZYh7R#_3!L-guhl20Vi?GczsZ=QBWRV>LR z$T;aVnhP#qvpmw+dgmM{F4*=;iP(JZO&4swN)lc8+zV}@BSQOHirr3;t*9b}1nBv| zMGXC*D3yLBw2QvZOZ7on(-fu(Vj>$Mzy|*7EpJ`)rWmI!>Sr&(SeE>=XLQb8ZuaH7 zx!?Pib(^Tmo?vV=eqNE*hb-)&&rSP?*2iS?mXOPmX7zz@wSggNQTIZ)wq0;WTpoUp zZYNcwm~P@VPs~a8g+K`PtF*XgYf_vSB->+yLI+)0W0_cqFsQD1b67A!HME* z2%iq{v}~*NsOqc~xUnA6ZKCq|=sCoP9hq}DWG~Sm`Dm*&Q4H=ZG8Sa}w|n9KldM4U zN$_%!i?ld8(Olu!FMT{KXLgrUOIbJ|!Io3;#j&}|`lbw5d?FH4%}QeFk`G%2I=qz!`Yk$dRDIeK+AtySSO$g_W4k!^JO zoF(AKF4D6*rR5%IQ;H{?O^XvMoW1JLnva_?*&eq&?7z*MkXZVI7}D*+n^2-fjY0fN z>|=@yP;sj!*U!2ny(8-O*$`7BxX2jf+5QcJi=oxZ9=a4<*gs!Wlr6X`KTYnv@{#9i z1ym||_wu%dHV8`O{gP(y4sX-r4IrS`95EzHR%G!Eer=K+(@fWK6cTWiK*`haQ&4c= ze#SSOPGWlH&EckTa1CZi;Fl=%>ZQyE@v<<}^+e_feZZqrz2%ENG(O)h$e5oo|MILO zlFZ2Fxo3D+pv(P+Y)E!wx~b0J^jr4Lub&k!yvKBVq>KBc8BwilkEBYP7@jk~R@KWp z?uh|lXBnm^o0=0|;orr{Fjd|N=(Vobwo&N3CnQhWDm^rN5mIj=htVP|ea4cE35kb~ z{OZ~J*^hi?B3^U)>|M8>=qD@7zG5BX=YlF1PXE{-;xQ8NV6JBgDHffan@ukf?({n( zpKxH?YgQ%#%$C37&M)exf4_&>Lqv8+&P`Pgs@6f8&#OwoxKdgAn zR8@>$LH?6Z-13#H;-Xc#EyCJdxNKn7Mj&nI^~;iY>YAb${8zntxo@>^2i-fhgYJ)* zpa|^9rM7pWM?NzVuQ|Cqlp?Wx)AJs$PmS)AUiKYy)6qv{)v7_aJZU{U=(f=VDxI0M z;A~zZQv~gQML`f$@hq0t(dQkf$g*o8?M29lV}@PxIy#6OLcTSy&EL8m%tjv-YCl|u zd`WyPs6w2QR|a&^S9~yxRV77@>pl>H?55kjj7hfxz9?D@h3JFwgYJEFZ}g^ly+C93 z?lqr1LTzBfcroRTVNFIX559Et(;xkLPHW=22bHkgJ@r2{dmk;cE=(dHPA2y)w(Xk~ zdz~WJPGJgRZQ1}K~3`%PzysP!2`Y+Q`1FaXWx)pJ4l6L+h1|?0pGMKIGI+t z%EM+wqvKe!f*$9$PQaS$`h$Tob=+HKtSR0L{yI7H!g#t}7GIgm6nl{(A0q2$zvq5z z;!q~WmV0z4c1-K{>=g8n4&|osqskpUiUWZuAA z1+lH5vWoh6TBXisTPV(s*EH=OcjZm<3gAnT6Xvo^^KCLH}G{xkrOwx7Ue~gsHc&`<34(y)ZvD zq?UJH)WTmHtXDt$(q&048NPcaBwmQo8oW1-hjGWiGWtS0<7&KL=J~If&C&g@zWGy9 zGg{B0>zY$?#lq+`Q7p7ZzfZ*_iOc4-O1Fg$xs~$b;k&&jBpJ#VpbicSVfKUoX@2({ z4N7{Wxc88oX(I>W%vX>m-82Tw#1!MEjy;CR{*=g6Gbp}by1hiWLXFj&xb5e{?u{nu zSm+)Jn4I|7M&A{X?o3u-v*Fg@9`e3Ot2|Ho6%F(%>e1^QVdfZgxa_{%0A<_x`|gKl zzUoNbzYAL+8+HGWeR_Pb@_#pdhkopHhkt&?Cv<{v0sWM{mlmFF?)J7(Q@;DTo@ZiEKE^8#RARkZXi7dvh2I^?UCn0 z(mnSGQR-)VyDV z5_8r^ug(>9Bdp}yRn*1uLDqDy~&}E#e+Q?icJ%(p+MY z1JytuMU9YJUB+E;><+mZ(8+ zD5gKU7Dy6$fhPtiF$3D*s8t(T8@UL%tkUSLNT-mLxhk@<=huJ9`b6cjSqK|vBx~I% z;6Ke3cN)d(jwn$`s^?2~7FZrB_c$40g)?WI_d{m3@8Dw${&Ai-V1tLA6YSmISOV`Dg+ zj&qcW;oVc=Wi6?Y8KjTAHStp*l721XBD<(V(J$?dE{bW8YO+8ku~AkLo+Rsw9v7SA z0ER657spX!IfHEs9r?E}_s2et$G%<9uwMst`y{9N~A9B-y7S(FmUS3y} zUEx{FPu1>s4#L9j-@|pFZCH%93B`te9Q0!|RK9uHy_a0%=JC3)(HXD+*2ffklOorF z)g3GhGAxwL81U9$D?Yfc@{Ael$w~5oasYzBO3))Ih&`n^qiT?W5ToFN>f-B(%%;$u7`=MGJi!~U-6#8lAWkUG z;T8I>ntUqeKIv3nk@V2{q1fMBY3%N87ZmWC{AwjVK3(zwUTVk&@VFX0`y*CP?u;sx zwv&R`Rg*!b0LWqM=#;5DRTyo&3gnGV-c2C|vGpn_cM%WDOWw@l^+q>?E?Xur#%`Xr z(f>HP4)ZTu|7yFyV4gUOPE%q=BRMiJdUZ?_J@Qwih9sK({;>kk2{y^r6h<`uIk9iI zoox#7bB@xtC9TqjydI)C7Me#7XN$)D+xyh_#?DXksprP~0wMKm?0K9$qW){~zx}Rxr&{%!BX`LM zlgTX$PoRxrFHqz>6}LuRtlsK-L()H08`vD)P8y`m;hSSSRkp6{_($d>D@=DyPhbP~ zcHzOW6VP;Kc_)YIpV|w)MJcb}vy#~rgmdTd>*y2opj!@eKc)&?%1$un4T9q&pU#48 zKy!F4f1}rNvRbiDv4?&TQAgjK16}-rvUw+H{HyG>UiK}FEn_=|>vpanC!6ECZe_IM zd+$B(OjlS_h|Mi+qAssyCRf-9=ksI*uszE=6zQZZ!d|G^x$@5%W;QSf@A!JI+|aEVt2?1p1-4XFy9Aeh(> zk748GTo~AzjT1a3J^tFS{%D#ealiO=+<*Q4$G=?o-;xy+yObjF6Y-aR=09ck%R2+D z%k*6JZ-uu%I2ANoLrC(?~Cs9G~J#Ew3o&- za@tRuNdD)_MNOYCVQQp@Kz*7{m0o$$U8B+L5nf`-=Rb_d^;ks>|L_0o)udS^zm2nQ z{kG9zUCmU8*7MWoho;oquuc{33x)}gYT5lwv-zDEpBtR)zw4e8ZmCzFbR>u5y8kj_ zBZqWDb|`ZB%bUVAS0o4Kbtq7pLaW>m-C%6))EhY?PLuQi9I~E&*P}Oj$gPOY2|66~ zNV1l_9Rb=lxbsw!2KheTT}dafljG1$!Ceu(kQT>AnOxzg!sm3UIGI;AZ%+sw((#_= z7QX{_rV4sb$j6@fBD`c|1vS0JnepUR?$IbH29=CP&`?1Gr;E&@`#o_m3aVF0JDBPj zAA4e%p@D`jYs{BlgV-egA@DEq#5UEXVZiidoC=*djms)ff`o-Z~_t$P_k&?-z z$-I_mPsH3m)Gu`#-E>R(VL_Vg`%|2}+y*h`%k0=GG0V)Jl=6N7=Z=AQ%r=L`W7eyur zb#Qp=5-oGQ*Til+#?eV|)Vgu9W3C$k-hNPD;AMu<&)UdBvVog}?YjRm;FKMCR7YGcaR6OPKIUEST2D)dpagD{DYV)q8h=6qP3w|LXjAL$FtIB zB0QlnLETXt+wuI?`mg;c#Eg-dX@B29vRt^}uhwGK4^V6|MfM?K!a!^e9Ai3|Ws!-@ zE&g%u3qS8wBr=0;1#eb-^TJOC-A;dd=u0i%=v@@H;?2av2{V zxbxi||1#)y;pc;Hx4!rBm(G6)2iATw{f*t!d#}I8wVU!8AN!b~FuDbN#&6A7a5=eZx?SO%*H4y z_-y!$^3Viax;^y$C`Y<3UD#{3(REoN>Y_7Mt7CRYBImT`whUt?171mirdDkY`dY)8 zDa;wHtJRP3)q_LGAe2fEnGmSY35GgbpwK z>+9A?$=BTS<{4R2&dHeMi6G58Hq-Brke^rxTuF@;tjeF3;M`yR4`o z)%-%}le|DL6&j)%p=PpOaNt`Pp<)sXf%nZnOYikg7oQRDow<>8$|0C+e6F3oB1yNH zsacI|H>T01F)eDxF=`*=tjx%>-#!yk&l)i2+&+=O45HW0Z#hcVa08JGyC%CWK$JqU zTPTu5#dV6X`Bty2X3j>fnXA!Cuk-4H8{|uaHJEb1pntBoL1uua1IJhWS)ac7K38zq zIjzIPZmpZ%WL^F8f|NJdsFkTR(#dO;h0JmH#&B#4(Ikr7y;8j^L>-C>I^FZo46Sm@ z%O^svMGQ??MHN?gOjIWOy+hYNnQpcn-`ex@Uy>azd>)2I{1J^Kc@&#Nkt{0ioOc0` zl^Fz`@{1%X5b4bD``;~q?jmd%Li)aJ$N}DCFeHn;K^dwX=6w(}xWmsEwt3Xi+XJ)d zL#jvKd&jX2jxzwPHx7p_!RQlk@>eGHcX-@?-oqzY+ndN>NF5cHfOxc5(M|Va9aOIP zfG@u3P+-zo3)$dek*40g!99LXBGUzgQw6b$;^!2^Ci1bK$3WjFCGH8{S4p-%hUGgW z5*Z{ut?}xYI#J8ZVc=K-X>=5f;VL)a%%1;oxyWp1R&T%Z3$m7*kK)3yS)e2uVUG_SimajH(mk;XX?;v$s1~xpSXVJLAtj(st(rYBdFr7FxL_M5XTMcmD>Fml)Hl9) ziENw z)y%Dkf!I`^LZ6*dWW?P?UH<+b7E|MGlkNH3EjMrTyK(;PZGY~g@FKYG#yoS&cdUXwHj;aYw46l`7?bC*s*6{K#AQ{|Fs&_%c>*XG|7bd*cGviJe)|a&l z`uvLhtLTff&ZzD{r2Bw;Cp(Ua88cXnm>ECBydYpXwfkznBk4UC&gR%i?;Tdvgdib@ z<|OGc3MXmC5xutYXAiDzKRteT>Q}V?)68Lq^v@kf$nMGHxWyK8kYb_1rihBu&C22F zph74$>{3u_Sh-)XVhR7=D+Tgukd7aDPpibc_sM#wOu{7m$a{6-EMQf?`HN2$WPYXY zN1b2Axo*fXc-2JXQ%+qQdp`3~H=a>0I&%eH5L+_&Ki~JWmXzfZa&_T&JTSG4aAh}9 z?0Sl@Q8Ka*q{IfrM=z+4LjNYY}>76P-zGvc(0wWJOBw zJ-VC+5x8VUj;fC2F*js7sL6d8l`-EyH_M8GHwEv6=q<8_K*a7L5W+)2>lxvFU<1-c zf+`HQy>`*-=k-YkXCwh3y%y>}?~@#rE?{Ti2DXC%si6#leRw&{&cKe@*>r>KB)J{j z=-&hcXhUuVyi^}!8{~jzhvMk;h1vAEX`mYA>9QrWoJ{2;LFU|_yky4SvXx=S$j}Q@ zyp(ETk=9Xc0!3C+aX096I+M5lWmD%G2;&}icVbM|esf`$HL`y?r;Yskci#82K0$NY zqg?mcXDbkJUtT9J_8pu(KqD#07N$6j@u6td}Lc{525%-BrwUD%O%-(q{HpjaT-JxIm<^_t>OC>^?#UmtQA zB<0qSDso}|VbA8d+F6?eZu_2~o4hZP6lqONF@41>NIiK3Rji;Q65d4D zMD7GOhZ}Q{p`}4yEHahFWCm^UCwi`IQvd0S%`p zN@^sB!4vLsKJq0)lV?mJGm&UT=x{)+T}X4*AWn##wDaLC*(+wV@lLtAn&draIpb3n zHtq<;mQmyY71t#jP+TP0GZV#&g$F=k<_6I~_ptz)l4VGLpy~0c^nvhqiG^GvXLQ$~xw#=wSlBsH^+lXmYue>f4yxlHQCp{>y2p{mQ3hkWj zge{5#q-=UAPv3AHShgYc>-0JMtUF0vwpGJMZ|EUrUEtDL3-d*vzGzlgi0+f?>K1h! z-7))2=$cu%LB3J+NI+A8ggI_|E?^Uw1)2TM@44C|`Lz1_kt zlsvsvz?JdE2{)8%`{A^YqlZyax<^NfT^J>e7AQGNu~if)r{aoy4D^m^8$}<@KIz^r zItFekmOx`kb%lxJ#Kb#tOxy#_7N!b(JO*nj=R&btP($R9Te2cOAe$*voKht+dfz+I zAPjPW1=cw+Qr_+-H)jpGEfJn{2N+^+JD#;G!V`RvR@+gRc;s!FO37#YS=7oqyc*W*uGRQxnod|NY&?)F1N}JdPOP>D5@HA1^!{b1pJI5EVI> zhYk;mr{4XOdDyt_Q;vYdArw60VNM{SzRJJKKJQ8A6zNF_#joe1zUJeIyAirb?B+=d z)TDcNC|9^&mGnzd10<2JRpRj`jW&RQ5Ox7o(3*;noh<6SqHEJD-yk%am?UWTG3Kz> z^G)QImqE3&Ncy1y($5gor9b&%p-;U;bCKiTR|Ue+H@P5 z)PV7Om3Jnjx^?bF%xzgU)D^xcN$2^}%thvHuGqg`PCB`@e7JBK%yNtJi2D@VN0B>J z+wixB0h^lUSetiYyJt@0{C!B5=X(&)4>VCwcWx!mnujXY$utPox2U6$Mm z+a3uDe@IuE8h{UGgJP^^MPvROYaV8Z@W@7-0ELvbCL%@=jKq@!cK? zjc*yM zjyPrGjH74lsJ3r6zPB}Pwi!odGyk}f>~Ud7{iFpl4pS`Dy6=akr0)aK!h&mxBIth5 ztDAz(P1`(e9ngkd^GVAWj~OFmZw z{Zh<%Hh<@`B%ZpaD1$k4isfWkP%gBx-6NNQ4RJmH@vQUY7*i!5whFdCPkZk;84vdU z+GmdA=6n6?*dJ2T}D*bc}Vt?Vo5<#AckN z3pQURi7yOJ@+@$orP%Eh*-FKokQfwopa}|#-|Y_m3O`!W`LJxGc&T0K#SZF^@g@^&_XPX`5a4q;G>yxyph=_O2hu@sG2l=ON7>4`W8W80)C%R8 za)pWFLZ1?UjMYutro#a-Xg01W z6gcafQ)i?E)6`2E!nJ`b=9GJEpSl5J$t8?lsd++UI^J;M2$CoJi}k8H`MqtBto_lJ zr&#;Bw2-(kd~Aphm^vCwfv{frHK5?WFkjR3YCiDFVcwvdJ)S^yEoQd^}v zWyz`(|)GEytCU|ICrO6)I(u<;tz&oY) zT_!8%JLcVt8DLMKX8V-`p!K-$F}{J$YrX1`1g*h@ zL&F|e;P-u;QSj9CwqcPlXYNI7!gZGogV>-@$A0N9`fpvL<)K4C<}zKZu)3|CS3S^*%rB~q;TpMXhA1}(ET0w7>-f}vNslDK^y&;(I zJhi-zU9b&~(atyu;f9me6`C)IgJOf89ef7J&?PI3EsWJneMnPC9#8K(;l{<@HUh`c zFqWs&{@xgC?QL+u#PjEe4C^00`T4&92UsdQY|2;`)Zu4*rvKZ|g3aldy4N?BkqmC> z7Z>h?huVMnN= z#$Xi+(tPOE?~8K9mtuC2#PGwa^?V)wLfFO7OF^}~DtfsvXGSY<6=kVzzOpj1e7=*> zWe04YxeJ*FIE)mF36}ol)##~a*tE_*{|-rV;o5*A3m{}rEb#bfskma3i%}F}fNGHp z6$U$o+?p9w*ei>v1p?$`Ub#^WQFBXyFvzB>nW2z2xlUX*{#}$~UwDgka{3!iKFFkZ z{wtccrV(>lr_~0(*-GDyL5r!&l4f@B8;;rQp;ckME{{qmb{0xsIPP*u0P5#BbqY&KfjA zD&SR6A2~6r!x~2;t&6oQ6F?Do+H4jy)ckKAST@b%$lU*rTYn;Jo!F7dF>_>gQF1{1 zRx0K!jIZ>& zbi%(l-`kP>E+50cvhbyOK~>yJ|9+Qd_;;RZ@;@uY^9HqE)-T-{urjRI>-eHL-=&U$ z96NN4YGfvaB5My*35NuKZ^Fwvh5yz^YT4n%iK8o@nBnCvC2yukBNdYz{20>Zq1Iya zYlp*zl-zqG?na;_a|44#NF8z$x8xIPy9_UXD($4#sfU^`yU(;m)Jm|{A}#Vh1C=1| z#^`Q5wPDfShyuvn#=Ps3%3Y9c4=FE+T)b4c3HZB|gF8$IqOeS)*ne4uPvzp> zer3R7zN~FBgs&%6@Js>s2yz2UWL(I=2;-zdHW z0UFJHNiLH~_i>)Rer@J65$Lccd8IfcgVHVl_@a><=RH z8{{!K+&Lb_K?dG~$S<@6;~b|EkBtXVBRX-wg?hi;_bu68*tpP6JZo9uhRGG-|I3Do zP?2UH`F?UmdKrFQhDC#VMTc9dqS#-nScYN1Bv}U!4jm1Y@nB?dMsWZxhIc9blO)b^ zmBa=kPMob^Wlfc(G(g5AWL@9$>65LV-o+{Q?+(PTTE%7+@H7m{caPyi41a?YKW2xG ziT^xXduaRy)A%ap(qh`u3VxapvY3hln0}BZgoNvAw-uz7)5RO~O>*c&w@&&rt@n1c zVK5B_I}D&@uyqtC*3P<9a`hkHa34FH$%*F>E7?p))t~0G3(C@0c|73NNFQ^dARTE8 zp^VsblF;yC%xs3mizg=Mz&b2c_cy(1DR}sbOenyRJ1n&FLLTwv;!H?>OBEhgoFz{r zwGvE)!@6}OEes#q5?WPHao8 zBv$KXWiwNSNz;IcAdg=wX_sK0T&fTY1@V92qy%b$1Ykuqo_|PsJy7WReI0!d2%}JI zbEv8Ddm*!KL84ou0N2#@A=z$8)AaMpqKY8sp@RrOsb$>zlFDevr|tOY(%B z(YZ_$a9fnQuU95;e~(cY4W?Sw0T*5ubW-fEsSq1x4sK%>NxOgpz6R7KI{}x!&s6GV zppY|E0}mmZ`_R!vl0K`3cMJ*_OXzxLcjQ&y^3Yqp_0n4AItWhhiA-P)3ibnk_!fcA z%D$B2fHpBi<}8c(v1K8u)9yrCDLUQkT?}l7SRn=@hv_ir?|wqC=%Po^BUr9puGT7w z1gPDZqJ4I2R@ZXj)zORVGV@2Tir0+cly&LF# zo=HHfg#6wUo@?0H|IgdnVmLl;J$Xml*GHZC7t;(if4N~Uxy5dVa$-m4GqV}0osxe< zkv3#V$rGljwnoE}+DHY4#bnva8$cX}w{UyrI&w69#@VL)j8h}|>?=z*s~R9TuS&X^ zWGQRtn~Tx6(NsxuX#A%OU+=aB2sGOT#`w&#Ol92ct>Jm@z+QKkJoH4Syeztwr_qH2 zLDbA*w+5gg%Vkc50OJN^-{541?Lmeb7H+6u3uK!c<+WZL&S;p-&KFHE+^g#u(Iare*3nl zLES#Lid-Q_-Mcu4nF@Lx>F}xH9a_9A6gf6~ChGp0z2PxW*g#{<&g|IW-OGE9|EtLq zJ$rOeOqyNEGqca|BTD{|BKN776R(|k4YNv+b^`Weaf>&LUJWi|cH}hye0J}G&v}hb z;~G7{e3gMNlztqdRUnV4Mn@+wH^L2KJTZEp_kc;)NM>=xZ#!4h^O|8M-ZVZFJKq4E z8*l5Ov?CT$o^nMA;FaVFv8a8c92npBhv53@v~sOX0}`BbkSp^%!6N4(0TQhw$=U|9 zV6fy2*HuWm1iv>%_$xu6vwMD0SSzsZSgwNDjMT(Fqhma6G0LpG4(DCN)C7HF-C`5q z>VB{)iGL#-G0l9_!gP-lfor%aA?X~Q;8Z9ma7q&g z>!pvoA;Wl$AcF(JfGSZtT^p#W4K!9oLx52)D-P`Q8gR*>waVJSE?|wW6klJM9cd6h z^jhx;+vB=$EpsH~>@0nt5&O7iD<^BbbaZ*>agh=Krc6Xy=78YjsTgvO=~Kaefi`9%g zrS{7A^S@j>>2362Djw8S)dOodY@z1(XZ}%t`SvtS=`ZHO+=#d{gd;KC7{jmDT zIs&$HRnPB%U|+kR)>epeKywL3CRH z(GI7quskLGbmWE~m>_fWM@65JIwyuqrx|4Ql>9D5nyHu@ejm#^Wf<83#ev`=-{gSQ zzW0fiUY6br;N%4dj)2C9*nm``{gT`3)G%|oQ#@}Q6?TLwxftqV8ypoLTqOlQSNPnELB=LP{`9e1tbqlm*H*uyo5kdHMn*$9Ji z!tD#DCmVn87|phwo3)lQ0qD8s^DCs^h3qkV^9huEJw@WF7<^NQgAMN zZ@_enUulfnw!FzPx^T#CRv%pFjnjYgZ~wd3G7#gmX**Vm6i!8#h3n}xk;%b1oE79L z5Eb>3Q>2nx$>6oe^d3bf{gl}0lWh~2Yys_y&$>-L+NTjK&{AdLS$|`htmCxUu#$fB zIpNvoN0W(!7vp9NBwjS0!~<54nDWPaDyilC;k1yjGO_heub6(w9k}9E9?B1sZdwj*tqTrVzyT z%DLrGynP_5U2-kNfot;NX2KTx@?y#{9SncoQOC2x&UDTE#&6n1AYzaRkGWh7P`nMi|!?>=dYaqNmL%c zc{ac6iG+cVQDuZ0^l|4XwlEPe1s^KXFWE#UBS1~>CdL4s#eR|n)?rkf@BK8{z%bsjP3|bD$2OWIgTs<#i z_7EW+Q!sk{_l>0Tq+RtYl7xToHAOmaA@vxBK;a?B{oKMm5QzxBI0ABl+f88f9+sxuh3F+7;Q2MlOA&^x%+oUvC;cC$McW@yaQF>Dhn zyECP3#f|Jx6T%L+7p@>V?EDxf-qAT@hR0)+yqF?|R7_da7Sb&~%xU2jEm_V#%N*o9 zC5fbhMxDg#N7jWym^j?!u?`8|uN11EvjO>x!z5lzP z-4&Q-oVB~}{5#pmZpLxqmGgeH87GC3CsSkx71Qfc3(U_@)ZfavPGj{@1G9aRrUJ@n zw=WuEg)UWY1Kx-Y?g=2tI60*6~xmRy-5gKbWX0xOJ|3E`bU5u-T* zmCt7(d$wKDC+kve3jhzeKzbmm7l?F`oCE7OG@DhmydwVFAwCS^@$eNtTY zr*rS~OeW-w`W>gqI(8<+iJjJcW+o(=lJB6%HYx^5dgHwwa<@)9OUKK*RC?*t=u^Bp z_0FF>cnU$EzI|Qoc(lCu#Iu9Vi^m*(Q1zetKeqHsN7@hPUGsw`((tfP zA|>BQk#$r|6VTh;5p{%V8iK2*t(P5>z>2rz^}j$;l#K)L@PW4c0oLusow&Hawd68$ zRv(&NGaLRYN?u739Tn5U>w+r43OZ$GZqOakV;Rbe>{7Hrs1>V_F^yg?Te+xP1*OLd zq_Rs0KN577UgL`G*UZo=E(8|&V@~{;(9NnkMJ=#*Lk$O~Y?iV0DsGZ2gDU#h)VJ?T zkgB~#(BYmgyyM*m!IU*KCTvw~gv==J=oeqIgaF%f>yQ~OM*HUfGVfdF8aOSCtYi;7 z|9l0(qr00c}q2L+Ad6R&Av`SGMn zR-{Uj)r35WD07F%D8x(qI9mDhh(cz-1(H-2ZU=h7`jFEcgBV$x9!apUyxPqmK2N$m z(j&9M?nGrr-jCADu+n|(AUoRuvIz%YJf|4%vi-qhv}a{Ure6AWWb+~uYJyg--%pOe zGV7ZgW*9q1$stOA3O0%SI!TeLg{OV{HiX}A%hp6bm|M&om2Tpk64mf7k_YqBfi-iJ zZ;?Q&*dI~^Dv+POeqG(5UK3f%yWw3rXJ1sYbT8=xJ;*xd3Q6R))7^oUOb02BJ|={$ zsTox>&yxgUnR}t^u5z!}W-fY7iQE$Kf$qsNNSgZ^{{D&CExQ446mX7#tuMTt`61^T z*94_GyS{sltac@tW@E6Ol5eF*;t(|sP*O;A;aJD3L3SK0yeJDlKI^#u?#a>Tc=6Y; zc%v_#nT!K=A7!VF)vw|gFR)BtvXc38MA|_=^lDHvdf~3DR)HK|Mwa!B(YqH-_H=JE zE+zvE!+ypqf{T4CTLhMZuWa^koH#IRMIKfUs-9iEo$?}8aUd4-VNr9=oNWPp9#`or ze$Pa)egiIrOryMt+_bl@_%Qfb9NLU>(|lp^xWi8O{M3z_h`62;FLA7h>t!jke7eaa z5oA3^fj7(C0vDvv=a#$Xtgw@=s`xlvq1;G~B^cKB8f3iA!} zGsf~07^DXS5;(y@LAxRg7T47w3BsekXht>2HU|{kZiZ-0hv1p0g?{D%4>pSm7Gx@s zF112j<8~;zX~rfGppVkgh{!e?6L%ZMgA`A-l16d)G_4Q(GU#yc@!mKW8CKf`Rt+>; zjE&-&Oz4s|km}E!f6LG0vltdf+$OtSNr{=yvY(PepgfC;S&A$WqSYQ(Ii21O>U;cD z;jYjuelAl@N<%Xs>eZ__LwC%`n+-bQz`V6B;Gp-?M@v5b4)EEY^{H3w=NowUK}x7= z@x#S+5j!0uA7h8tZ0$`==sK$j%Xzm})8RwEU%E__n}nv!d7 zE5vt0*Fwk^jS6fDzbXQ2QoQttQ+vc1hZ9tO6hOJ<b!9`cLT|D;!w>wGjn)?k{_c;F%`3qZkIijRlDT^uSFg| zNBP(t2QG*1^TY^^R*?oA21o**$4?N}N@BzEK+!WTGKa?O10-?6yva6qY|$?J5M&^b zVHiolF)*AY>yqw<=P_f!>XgF%{o5A~?0>$sCsDgTiF_ie-kp2IF_m&qD-5gR8q zLck9+EGoEzl5eBPW+P+REy1<`O%qcblmnc08l==c6#^vis`AhT2HhV!!o=Dzy|~T_ zM=zdo{K3wYU;RJJb+*%P8d(bfkd1Pqdc=v>vdWCmAK*tQqiKK4x$ zT?d()N}$B-ciF->Nc&w17uE>+U9#Qce5=VC-)y(k;E7un+fiiQ+Zg4}1n^^glog7o zJxl+I_p0nY?Uf7H2Qzh<7{2OR(|PkI{?$Z?k&!&gqH(2lURvKm2CXea5E{>FF*8P~Gu<-os&>s&ouw;(0t!MtK=BB^({ z@{En}9XXfSAF*aVvVt#_^2vX{ZYeL#Cd%Tx@gtIJW&n3n@|_fnSxld-jh9Xz3fV{F zQ5NS=2rg-iX9q;B9+wq8py066Z>v35!)7Bi@`NRpMt0z!EAP^qJilq0S3=|F{v)Y& zVrN!wHr_WVIfU72s2Gf&X%)ENogmEfZ}!2+NRsRj6U*}^wP`vJW#QGjt+`oTRA7aid2;h=*K9! zq#p?e#XI{v`aFsiMIbvXGLRbOni-8CYp=oIrr!m7t*1*vm-8XAm&cEFw-h`YzYa`pa~i*WRo?>!)J+Va)gBH&d2`d<2B*oO;%N>I@(YYJ^p z!#0q|lX-PwgSdoB3j0+487Ey>O)A9wQlx#lLe64B)}jh!H7Vd$D6b0EM(4PQt81L{1xiu~IqI!aEgs zFXE_s1=J!psC6OhKa!o$odsH)%Ymqo{hY03WRYX8T|{n1#(+~u7bkS~};1v5q#1C>db#OEZ0I`{J#>{=O zzlj;yqw4p6XtxwwaN3#XsF`7Sp#qTYsX|SeJS7^W9i)k{>5Ci2%a$!m ziZsApAP#;pq)upV%DZG3fm=Ix0q!rWl_Z2$E>4@~h_aFK*#|Q$z{$wLo9Go*GtLyR z?zz9REP;1gs8}g#M;$wquSyWEoZAJC$VK_hLHW5X-}4E21LRo{+QjpFi@6 zvzMDW);O`p$5=hu=RPZZQ2Fg&jdK`?(6;pE;*^l3k7?8d zo|MUAX!O|NV(6L4_|6I!Q`di6{#DB$+{n^qC(gjN5^5_AT^(8Lcc0dj&w3J(a_W8I&LDlHKN2a?>EmN~=hu zey3aV@%+-6eYVKrjRSp_PR=;aEuc2a>GS#zPWqVE%60#GbQ#&@#DQMOi5r%=luOC8 zDUw0OBr-KL5DErtj7;{%uiF;t!mC7AL;GZpBsp_7xF>QGgl*D~JgzSsl+DEy>ICkP zyx}=%COmpjaLe6+^Vh^if0SXf(e1Kmo@fst7vALz(yclFO6j_Of3U+X)e4c4HkfAAi%c zz|zUJl7E&%KbVLAG)+(h`WXXx&W}J`D^r;>n5R6+uB0u(8AG;r%-3Vj2G(EoXrE4} zEljK}{>2BDag^6}gc<{tNN`s)S=qwV#&X*e-7LiI8`z74 z^9J=!@$;Z|i1+6~R8|*W23h>5Ziw!S17{=l#$^cp_}!iex9n@$Z2uI?8Riv%ij}!% zB{$ydqX^^II*8W;)ufh%hx#3bGGivM1P*G z&mXJ#<7<|5q)r=VvqCVs-t&xd1s|ArqH?1SGwpQ0OQsTX3h2zml}cQUYzrtNn470n z06AQEzYB6^V;QSM?SXLs*bckVD%d)B<8Q}q9`spx^B*i%G_Qyav9kV2<3I8#@H{u~ z>1&@K)v0xYr=(T=U$_Nmym?mm^Z%+-|Getk>2G%Zz2S|YUsqoPJ}9kX-LzEDDb>-H z{<)6YWgZ{096#%{G+*kmsp!M0_L$8Q71 z|0FBxG4=J|{&W9qQ#{1wC6#;G!7S&;aWT_dt2hv)Srgd~pYLp1v&U`ujHA-M@>Ke{ zuV%mBekdf(<0pl+$~3u58nngxjpEMeUCKl*w$mI7>7>_qC33e1U0(=E89od4L;g(>Gj#h0JM>|f*k5J?=71QC9 zt<-XEE`AbSEP5D`X0APE68n81&tYM2gFb2)$lre zw7^#x$1j%lidq-k2I+)OZpZvBS33h_}*T|&O%Fzw$AG4ldg1f}rjYweH1s@5LJ$5_{938l)4Ch*g2e>CDq%vb-*gpl~P2@6TP6MMJorw%t1 z&nbByMV=04!))Qz@(gs>qLXfk+&a)+)Lag0r*malG}4Er3j3vXifd3g1RP(TK85Zz z5={Z;mY_nshI`V@NTvgEe0VHL*5`?{Q44Pa5bE}dyHt7p>zO2(j>ZjXoOTiz4U(`R zj_MSpf_N_s^q=P7*BVKl5St-=%b*ujD`^3$x+>`bPCF>w)k=WV+eZhf567Wt1uV{x zix*3c&ibGB$EqYv0)yN@&t@igm$@f?r5Ff$2VC|+Nt=$7Lm%`2$&nVw>ioBHAj;Oc zXF~0)Tcc5CrK#WJr~b>Ai;V@>*&S|qvsdt2{lGK9Y&%0#!y9|v(4;FL%xe@s_s#l0 zHWn~o|9;ZFahAz4t$NpQ14(_!+?q188LEJi1J8dh6_X~fi%JI!TnUL|HgHk2P^;MI z`Ef{tx`l_Lh98IY&xh>;*)b3&doEkysU0j;+!}n%L$83E4p3W6Fjc|zQl%TgW!<=MfE z=q?!&DqZF16(@m{;FH%M4MyUP-hxIKvRD4e>qmW0Dm9q!>fqOH`|Guwz#Z^ThjnTF z6FJI^@4O$fb&?CarKSldb;d7#O_E$mkr_&|DEVFrSXg4}e9z9U_eR#z91xPMr}sze zL0J7F(FWv^dioSdt@pd6&#axff?pWDDyZJ4-{sl#<ucTb;Kbn zoH*~&bMfsD=LVW^((_}7;K++67mI?ep4{c(*}a!8;?~ z&%gMUGoc3YQ~M`fjvMH-$wCzPkZld+Y1JO3cDhKl?j~MUBh#G*|VY!xbF+z7qX4BDZn6#_Zkp=1fjTM zPU`}L^r=(_s>}Q6{bFqMXnLxmbwMX>ppS;dN7hIk>`UROjt?Yu#lxtbu>uM8hlfA? zv*oJCX)TMDbxhA|hMCAK*ctsy1T~1b@zW?ql3F9vkG-wKsBeAPMHavq_LSofSpj2e zYyRSQWF|jl$L7!v$@W*qPboI@Q?e;JR3oHOF-5+2gm-1J{BovjW(%(qgQy;Hvol1? z7j6f#)2-q9MHvuI!B)mAK;Glt;r=KH(_)fE-9iW392T&0us323)hpUR&V9=Sl&5ch z<6p@sC*A|NV>a~HDEU>2T*7^y3c7}uz;yT&`)m!bix?2(tM*6h1m~jbBD9c4*Copf z+D{kK$>QFKD&HDGX>j>09nf~2k#AiG<)Zjw;6n9_b$?>FJjvpo#sQJQEa|qaS#>)Co z^#|_#bfyVZ%8hHwNyk;>Sz_n*8h;Z-rXSG5xtK-6uB|- zc-T3`gSp)*l(Xx1DVA0Zip8M7Nw#oJBmh)w;N>zHFhD{E$YlsGo9QTvGhqzb0*G<6 zv5ukMR&vuai@|B#7%LeIZLr*dEgckIGm1H&LLW-_qHqV$0wM|jr1~=KU@(sH9Oom} zVK7?6ux?dSrzHo=D-zeY!q`$Cy3zYMq&pS+1Al*kcP+1MP#vU5h4kxrd`xjVNT-Hi zYHtFlxjM>qAb16ivtyzDvKMs>ra!TmiP*`G8l6MA`Bq6 zNSsr$5Lv)9gHoSB1(qw)Z1&UzB+2lh0T<*%$e~dt7G@@6pW8ia=tmOl21SJIW;|?x z$rrk$7B}q$CF?L5ZL(3g40HyV{MP(|9~>oVFPYz}GxJ*yQF6$X%cEirK*e(Yl6$g) zOP)x97%*t3xK^<umV)k}tF4t)&L5hr`_Vc={Ve@Be)7x!s>{70K!l<_;WUUp32{$1ScU7?xEK~*{A zAI+JuO7ef0cmWhvn3%d}P0T(^u0S^HTjx!cuqN2C&-1b96Twd4%zqx#CT$OHQ1ABM z%zf^g3{q0XL7TpIdr7mXO`6T=QcWJHY_l&feAE|X?dSuaQ}0c6GdZMZU$6KxS>?ph zc}R~MmO!$Nl5eI+0u^KA;v1CWHY#7I`4rFXiH3CxT{h|Ufi@4cqes@Ff0jn$x6+UQ z)&z_HYWuHNa>d|gFU?i-bKnf1^6ZsSmSDy*^$=-=(8rtP%^g*!u5Lp zYNWtMJ!afpu4MW=4N=A{&emxf>^p}(5sYaqSVLG8l@8As=yMpg)k4ou3o@9@VuR;n zGrV4y68jujplQS@J9vo}3funL1g|3RXFnxpofuvQGkDcg@;Zvt0=&Lc_^*4vx{_P& z2HNwR06GI8)$*r%va5qj;F2XD2vLILQ{N|1opiIPWFcnTH>gj-^?qqW_;OW7WEPE^ zIu+u_(;f@(+GSuEZBTd8^{|+(laudi^}`{ ztUE440hUky^F0#FE-k@{Llrq~seZvJP^}`9O>us=>X2lv{6*30pi&W(VXbg-A zwf|=gJ|BAP(@FTPg{78Dp_d90R+dMNV$?z&aKSAZbU!r>;0`y64+kd_8ywEh-<(~J zY_kX8v)bjS+1`nDSOk1+<&`MY^dWFtl0gnRv9AiOD8nX%3Q7)Tq9>@BGokr(tpsUG zV)=a@=cEY?E;uTg@@ZP1ETxVvSkySkcCs>93yYC z=b0o!Rc>5x>!Ds<7g8h~)Zk8on(|X2gQs26GOmG#+I04qZ)}DOyH$?uxUkaFI$TBz z%l^;ohHt*?*++x>!rpp-2g2fH$hLq4p^<&GS=2%r#gkB>Xg(Hfpke-`W6#(@iPlR?JaU7s@GB;Rz!m|)22Ms}<5(CC^&W9Ee1uoxwA*}EJdi}P zD)Lsi7N(fvo~M-`h&32?dlYdie0$`PkjZvVZ64^M5tyX!tn9|rg#oTVu*?c@T3A@g z04P#Hk>4PLD27^Vq$`6pRixGni~BZ{b(5hoI(%qt#>FJNrmS!=W!KV(?^+i1vDxi& z;`o;pVss=~!YqFr6RvR`x>LzKjED!h+WVsAD!gXkJ?BY0@LbIO$@v?X)043a*FJuTO*L#t&)M_AI`%FR66~Pv;ZM(1wq2v zgWg@7-iYTx#-G$3_?#Qd*9qcT1>KEE9jg9d< z5P9&-!4Q+(nUngzn@RR7v)HLJvr=3ujhws7? z-!(J(UHUvy1Iyf3NQy%n!|aTwaRA4P%^43otUvSh@1O0rtU_nw)j07Mj+L7AJbo-> z$D&UIG5ogMqu=Q z?0uD;Q}WX-Y47@)rko#NopPNdvzv09cs+F7Y|6=_8z^&QUS#zJ;UGiysU67H#9?EpC>!(}`XckcNAg(kPDQZ*sp#*yv!ee`!{puxX5~JnP8ItSrdXfj5)e zzFhvxe*Rvd#Oy7~!uYQ5+FHHW+gN2)BlLxPiv%AJjoZ(YS-b0Zt zDhA!$T;-rJ0R3!?kWvD+<6*y zE0_7`wexN#L*o2Wywcn=d@2_oq_YEtHE1F!Uggtr=oGKDoit(4=7S$(gd97;m2ZxITwlu40Mqy8wv;vQKl|Xa|nM^ z>@O1%c`KQfB12R>wIvwjjaulgh@~qTdDw$X1Y727@H7;i2<#GptzADF0BMHaA_Um)`QN;H7~E zHx65|sLsKAmJ_f$jxjiM+g=tVF#AQYBU#I9n)6Ip>IFt+UC@)jPkpQGx>qm(=|G_=70;xRe{H0q{Pk@UEpd;LzBe@P=!HYb%3rR z()d7WWUg|vcd*9e?Q+7Ixo@{+=0BS-jT74yD;aqPU{*)^ zbmL-OFGCi%JYgfOw(OC$&2ms|g$2uQC)`};Yq&noG~K*grmiIUPVA^$Fq=S1DR~J+ zj#4pwp1Sa42F!p;C#Z2tVDv$2AtHwgsrd|&b{XiEb3PR4!h6YOD0CihxghP5;liYa zhsiGqOa`R<7fr*W=2JkUpGud_#LdV;rrNDYg{f2nF4#+hVzXaal-*2*t$;9}O1|K1 z!cZ`Fz{*>h@-MH~J+=(yv9T&ny!B)y^tV&KR}9_?of-yy2;#)?DR!fdtS$H%^+5+; zOm+XP@+&5T^5X~Bc9RoMY)~4^49W#cewHFtR7?)NQ~WTZjXvxHZbO%9n-WDOF`pX? zkBd}|;^v@7zM4G#oya8FB}KA#GE)yi%zgeyte{n*q`L;~25MNIrXPt;hLkPV9Hg`8 z*V&BXl975oei__`>JnM322!cESpz;QV#w!q^H zjEPOR#)Bkw=EXfa;kQ#vU{Yor`vZCILY#K507^fG1za{#@^uthL&X>?5%c&U%{?gQ zGOpnq&#N|j#s)l?wV$jLR(P1&rJuSf%!G&g-+6eL+My3EH z>eTT%;8Q#(LWe>2BV=p1Hei=8d`MAz?vPhDy;0sv@`N>#c1Rx3(MS9b3N$C8+99-= z7OBw&^r)I(?YZrZZ8bU(7dQ=Vu=*I6w-Wcwz!chGC$tq~yR!dXkD+FWKoixNyZUSxR8DrTgdaqj!bY z@Q#pTpy#=)xaSkkZKvPwTa@clJ1DdWyGbu@BTEkoIv@O|CP{Wz)=R+IRn&nl^RbXl z8Xe#^$dYM?Dp;(59qRvn)CTwaQF_@S_crCYjh6$!Jq+jLFQsM&ZsyOz!g7gSwAX6jB$}#W~_=trUyTK8pdcc+#<_ ztOme&b43)F6ZToMQodBg3Rc)lk(x4(zpP+UdsM4vqZ)V>T0GZE68-`->NWz{v1lEo z_b&kOn}t7`NJh8wq8)NDha0`^lzb~i5`iL{<-9;ed2so5mOXWBWYFF@Ygo^OuYHYlsT9*gxJ~Q{$E)dgyf|1mn8nANKK%i;V_myxouc zxvx#hZ=FaHniFsKSSdo&lUo9e(rFTePawSiOnfR-4+@x9{Ei8=iY%^Hu0b9kynH6K ziu8H5)3?I$tU+;0kV>Nf2<{i8(zs86^e-o;X{y}};`>Z0jm)Zfd;@K_&_4{Tjldac zK}NkYZk&vGu;u-Azk1n)46iM$;Z-QxCB3rEOg=xCX<$m03}7mp$#7{o>4{-pS#S-pOrsr75CU&gSb>s#TkcrvM)?V#SUB?fr2qFSf7qY zp@yQ|zdPgfm#acxZ$Jf&>2mG<2~5S}2XiwQH>h>I()kCxvcw6(o1D)`8dEa3QPAms zgwv+}5G;?e^(yW)uWWgVvPzl8MXQ9VR^tH9m?MY{CX6ndehT|W19kk(mQj{hWK+>f zuxKUOODAw40o*{Z;HL^9uzU}qcN;hdfw;R}lHi@GOr6wl5i4Y{AK-C6;ra~`_hl!U z!$ddD{*Z%F-Mrd`n)R|;?<&!Ni!q=74EJN{gL#v@r_5@w$1z!qceGheGE)~HQ|y~a zCV~^YyjC(1ngmVY@}gS{yUi$^`j=P%!Qk40pYdW~BF`QCRza;LO}f)=9$TS0Ht;_5 zJ^&skN)rya;9|EY*ir7l7rtv}9b{HUV#XK<|2zLl_Pk_)w-aW8w|q*T zOOb3U1{W?mx{7~L(5mkFYBOjn{O>(k9F#?u@($8v?iZQdx$hRK_CihBZ*(Dj{022% zNs`^3dxAN}z_EWC{|W3-9fZ`{EdGY^tW>@bJjR>0K*-t>#dO_C14Xz*rIXv>chm!BYFg=7m4G<`}CMaK!z%7EW>Es&y*rQ8&PO6vnc|7(w zL3gWa#I5cr(b*2G_#N@5vwhu-FN<)iav6kPq8@*SB89y>ONen_^nE7fq~ z{T-+P8n&sOP04}YGL4Ffed`w3miO<;-bV@49_AD;HtdKfaUm9z8KP3c8r0RwlOdO6 z@uB^2zr^E-N4hW}9M5Zb+h^qkYzuJcx^jpg!2Gt|z`!Wdh3gkL^z*MXwA{ zYR!Ohnvz#gq>PGLJ^${kc&~dAARomXj$F>YAU(u%&@Diaa1NrWNwTwaTSQw_0#oK* zGVkQPm2L^bI`LtjRm`S^UMbIQTKgJ8lwwu;av;=5F~^)M<3%n zla>1zpM%&OZJ4=hR_@$9VI^c&9<@U`^h8iJ98ng>69(9#-i|_V6N(yNtACeldu32` z*bGIPlsugxS}Nu&eJJ{&__W(okc5g2+r)|GcZ;()w?%d8O61V z)ET61!FJgD>y21R?n&-DUQ2VlFqtq8#^~U)=Yv>RF&_2Neiz=(4?$@JcAv~eNhozt4LFAAr+*E-a>Mf2WHeQm`IMn zSh$-Q=p7fJ%gm1bt>xm`Y2ztY7Rsex8(bQnl$LmAdhDYw&<5$#=u^DH;Ka~w)lTts z`IfLwqG#5Vhd_$Fg1RSL!`&gQpplqVCqUBMWMvm8TlrvKo>z{;vY93bfbneTjfKg< z`>d=Ub-L|wL7>UF^!!+|n(TeaVlCjw4l^N#DR}`!@~N0Q1{0?<$?gD*x!}5{KIDm{ zOO?hip|*K6%c{BCJc{Q&hG0rHDd6-ix~!<=_6lGfb@`{s!(@8Z+gJ zrA@%D(l0BsD;0uiI42CG!FkjsY_Mu%9Z0HY^6y&;c|(Y6*f{K_MQUKOyhfr`>=b7Or-eFn&1v?BvwY!Z&pPtN`U@W= z;NrYti8ucFn_Dg8acnk*o%b)Z7FNtsenxNi>k{ex+LSH4SpLcA_K@PBI`yp|rhWJR zTc5p=MiqVQx_Wc;?I7*=Yctp)k0r_-=aE?a;;W{uxbx*A=RXr#z|Hk(ltCyzdcAu) zogiFV+hqD>&$kt0nb zN1eEk5Omy!!J?XypQcC!71IcgQ!0I1)G5iNi?nKs$i~WIGMXbR9Vm+iSR7iD#eUXQlOk(6w8E~1nSulU&R?fx) z5G!5T^)f4fOifB!cgd1(@)hxSt*}!9M?35k?H474W9@7Wgj|jYw*}~#PIA?^075lS z=%&!P&PiS*_AQJdOWx?%|T=KJ&^`v$KTa@ zn&zaR-6BiL=2ynCfvC%{87!TWL#otnW6Dd4PYche&#T$Y(FP>*^s<|?Y63Jmi9!6# z3->1Q{{l`LlL;X^&GJPUv9V_v$EZU#Zj6lDoBb|^PvPvdOrFg)d$~I)c@jmoP%-5T z_C((E>7L)`fiLhBsR%`Sn~U-WUjtC8t>EAAJ2ZVl6FA$yip@J^+xJA;v$E&To5}Lq zN4ig6cDmXXOIQ9?4+E10=@ICq&HP={wTfhFwMQKk_WxJpao)-u=LaA7h_Xe+A=8UQn)EQtEq-YQ~f-?&gxZ=Gc`y!qVq(@!G ze135v2p;h)D@ar4{!yi#NUeqwuc)omYT!~iQ+d=E_?&t@a)I(Z3F-{<_!`tM!+W^8 zwl7jf43XIdRyZ;P3=lc=8!NbCynk4WgTJ_|f~J2o-;y(;SqtVu>+UF!|iHjN^zM ztXP?wsmh=I@v!9%fYa_lTS*m8dTWqj3{}XARG3VIwd+Tvd*y>>WLv;Sd8g#EqR77m zVrqvLI^10~4%ax~!ET(#*_s_ZwuXGI*wV#ivtV{!4-K;XhcAO(coh8Lpvr z&{!veW(8TUVNS|a>Se`&eO{9UA{=9uKpH13*gxia=uXg=%TXz!i)f_ASQQ2Vx<>IQ zg7vcAh=kAqmwlesfb8uq$(>vlCpO4nKfdFCqUE~}zia8J+%fyKU8Cec$ajf~xnRtG zZ&2?dbz+0KgxUq1GY3Tb=wnb%lNpsRETNjCs{OlE+c{mF>ykUO3ORadx46SSU3kYk zH}t+w*{mAgiRe1ufdfCn?;NS&)J5eo@7?qHA6TX=`JHYFNF^`%-H9dF)!m%^KtY){ zH+-(8%dKx0&J9#)PR*oz-BbY;RKjH!e(q( z8GPs6&|k6PK+cy-n*r4p%G_}#g1trGW>GHeGXftA3bf)D0WOUOPfozTKq1p8*PIFj ziD$T*D>8`phvbXX$rkQ_%U$x&VKLkZW1bC=S+te?;pg`(o)NB4TpbcS>^7u>mMt1>l@tDA|bJ0K@qSlLsev~dpyNz=3@UoM}NYW(I5Ru z$6mEQ`lBT4P0f*$T#jky$0gWh1&MPdP^>p>%Or`CZ=uL0D#o}#SPPP+urxrXU$nl7 zZjF%nku2_YoD}xxugjL#Z}cKYf8J47*}ZR@GfgLRvOUVJ8mETA`}N$XOLB zn|b0jlng6m)-9-ZvlllzJFwXVZFXJSrS)(A@eDQ1WGn9f$E`n+wXP(`%vS88INy(X8|ATK9}DXumoLg!-CTSXQhu66PyMvOg1^gmGVQiaFLh|yVVBos zTPLjSh4U_0{^70P{KgU+PU~4%0WFHl8g)wb%&E|vIj3P2jv_2r)VNWuS>^G7w=!&( zd=gR13bYeSJ+lq0L#}Dyfkq2WQf4QtS%$ljy-TIlPZQ3G-(^c|AdB{E1z%()rMP zmJ*|Odm=Xr6PS2+V>*kDlNNa(>Xy6avZBwp^O)gT$lUX}xfnDZ3VqHpCiM__lFriCg|1u z>(OOomlG%39y6Oub169l?K7yDk6wGs#RV>S-S>RjzN%Hk15r;Ne+9qK19OznZ_&%j zW~M^=0R~Y9ToQzLXEuu0duk4cT@Ys~j}T<3b>w20otYY%WM6zAvMP4}&dN+p?eAOd zV(Ica?G$FkiACE018<;PAg=&fhc!3AwCLzboU=@$Xd@=VI$*G%>! zZtLIufRwwETV@~AtCai_Mb2YJA+jN~@UoO`paYfV^U$|sVGd}8YLHzk#m8h^vy>@5 z$a_!%u`AqVGhJ>Jmjr2QC4HV*$}SZuw%!aVc(#-#HV)j2Pw)5-#u_T~`>Y1e zd4rwz+kWwuWeASV!rFQNGGZlQXP_}OKPY*LJcPKjJo)Abuw)Cc%^^2AXSY@_qi30H z%&K?&Hjq>&_EE~rj9&pI&!HXz$94(?<~E>74J5Y1R-Wu7E85|O}T*^?+U(c$PuYyAYuy!#xQ6O z0R=1LM)~ZCtg;m2c3MxxioAB2J5G06NPq;&YPvxV$#2ym37EYGUW;P}-}=be3=8XB zIOGm1SWNBRR{2ZI^reyefX*93vX-Q@dj5J@7b&5xF(;Wq2q10>=roaFXcgN?yKf<^ zdM6bzHp2x>i9>K z%pu8CZdTRMC!#c6H0tm!TNXRNR#HxGi*U_#S&=K+7?7hxtyG6t7k20w-76Ull@q;Y z9X-GJ)|4oT$y2@Xt#3DzEw4-<_mG)iyO)wffMpLAQy@*`9*7zYre)KsNlV0mC|J)w zrkC@x;1us_lEoFl>!z2OydVs0(wfLVh@@VzbHR`0uucFzmR;HP{3o;5+_KCl za9ZcoN~%M-;Gh6wD=oYvzr%_yRjrp^kuoc3n&zPwR&zPT99SO&+eg}ZAC9XTa+WqM(@ zy?aNz-7oHw-Me?%G?RB%-QOcNfkNtDf**SeW(Uh!b|$ zjorHOE0*)gD-tlZGW+z*dnBn6wbQK}9eq!x>6uq0$`js`f*dZVQKE^RHo4Oe3lup5 zELQtGUifbzmh%Z4TsU!tnw4pBF!9nrpXM}*pZjJ7Xe#~NqV7wo-5j~KYvuLZ3JNQa zJM8kmZh~|)cn-bljzreWLwu>^Ev5p}qS)WvXeYctO zvJa#o+q zFRmq8cHs{v-nuO{b8Qb&^8FOqN5vq!qh7YtFT+c#sP{b}!uZ*0_2ICcehm<5#ULzf z<#{v*?G-o3KMpy~ft<5mQnE0Wj^*d8`Zz_ZBw32r{*V+etWwv@KXXvUs2$*CD`Hr+ z`(9mL@XJm&fT?B4Ob5w@J=`+)8gX5CKDUC-^7=T$m~J*$B;F`4F%DF^t06XtQ_?tjA-xU1#xg*fwP~9?t6(_Ws@6_aaRbkMw^u zDWt%Oqa>Hjrjl|>4s%Nh6>}aO79)pnlT33a6b}pBwk&KRr-RQaG;N{+w*;^<30`@k zQ{kFDfoDSVA>Kby-7db$tPI=aR|IVL&$%n6)h#eci$!${j=_@V zF<6oe(aE3&buLp)9?Le$Gfh`ggPR~#yF}J6E%#0cwX1QN7<4`d5gyqSNUWekdGtmG zz3hpKKnbca(j93^m7OFuEKdX_T%iY}@`RJ0$ZRt#>^bH(^u-DmQ`aU>Pqs`{WaEvz zd?}3Y-IHBi@}}UsO6{7swm$^E?+~R|7J}pxF#0dmE53 zey{BR`1$u<_JX}nr7gU3(v^@$md0&RCoqqK^Mw0F7fJf80?#5fOeoyczy@_4Qz7n` z-T|7UOwc?)rKZ%tc-TnBV!e`u=Ny`3!fN!*ClmWlCl+X6Wk#sHx1;{f(lL3dHhZi% zD~;l~*{i}ZY6B#}S2@R+?e1E|-hgr_9X9Fh;LdOY(;I=eP#WVR(L+Q(mv5~5MfxnP z1l7wvqno2wNd|Wbz~AB2iE0PSk{iXlLhTu>n-H8V;bljFS%=fNzmxFJ_f0;n(D#>9 zNX;t~%70|$0pFqIP_KBCifQ4sMHP#>R2W6fQewHeX1#k0Z!ND&1%Z>Z5SuwBKJi-G z+(UphO|^dlq*TALrTYxl9%f8&7${ zcBWN^E2|vn|C#s-1SWe)-eBkvmSi_br_T*P1MC?*`&rwP58rB7k|e{Z4iKYzZQ#Va zFUwUvB}KqzilZz&saAf?zPa=w@OsNX{s}#$; z6*em(=iR#?Rx(=7Faz_iD4z|g38Ga&cPIZZ!sKTYCMFdP#vgVt@%2kzEL0e`mQ^ zwn8G)$=$;olpYG%wqW)AMD8wSviKr-&b<#jMbEj(9u4XXO0BYa?v4PwU*^6epoM(o zhsU|fWbX`44G)jw>>J{=ctzuQ$)CS+i@xl6%4&*t-WmGM|8V=~Uz%1tUp;Wfk9_1h znv3GJl>#K*jB2Zol0T(LHx+|qD(@X|*KFhTl6C4+!V=(GF;T7{rnLbZBQb-mDYVZM zgZRfqn)9UF;FpqF*4 zbm6Ehs6ldJ9cMq^$mz09s%M}AHD}Hqw`|D5&89o&H@|_ZgIdz+QB6{mi3^KVl}x!$ zA1r9H=(x}{xBCQjIdVmsc0re_UD7A(cX=wsAJp%%a0+p_*i`Nj^fpQugXGepivCMVe~x1IN6MLslJZeOJ2=O}W9 zib;yh=RS%kQm&PKEL-DS5Bjy6qu2ZG^u6IVAn0M19&k5Em)`Iz^jRlv^KDQ!F$01Q zpB7J?=qf-VcaP!KNR> z+!xseQtj6zX+iaX!b$%?QH=o!&ZL%wkXFziI$cFHV%YXcHxlo zkDh=3p=E7|)5cz`RC^c~MIe>w59%uzg2__m1|bjNMkReo4B)|%0&1ufo;Pwbi5Vhwvz?frhBaTcH z#+cpNt-+D2%VsJ}?uq*1PfJOfU-|! zpl=Bv!X8m80dg824Kk9TfJs^;q@J~M9($As4KtHuI1CV7PXhe{qoPR$=cq3R%{zV8 zORPt@$M<)#d)d|}X*=%64E`akKD5r8cldiv(AAkH%qTamEhi~XoE3G_49o{8IcN*! zP%&2*f3Wx(y*n(9spf-Y4AaDJ&o-hH7`vV$m9F`jQJERAXDpU)qJiAFLQQ%_Z(UM$}E90&Cs=- zl5eF*A{CS8zaFCgvHT~hn~cHls-QvL>|?xg<+WULMCd5ISQ$9mVPvHP2VK51ePNoV zuliDDF<9|mF`(ZlJ|>RkXOIH#46-dCedZO>%26d(H7LqL9ChQeGQlyCz zmX^3NoSO*9BFZNCrEUIIkm4D$P(* zM9G26?Es<#mu0x>O#4GV|dYFKR|w8}oiAdS>`;bM!3#48iVQ0JpC<`QBEcL;_{Z7b_!&8u_q>rj-}AiB z`&_=y$eHZA64?(O_T9@{H(wV2pL73eNjGby=~6m#F<39f%pi2FU;yqYsp0oU6vMmu z(7jxgF3ix>Op-g>__}cVx+nUEad_VP*FB22(|IFiKl|qPcPCQ(;l%a|^y?YXICCq- zCR1c16@5gFV7~2LFW={r6M~u~p5G4IWv_2n=w3H%u^axetHyWF0~ z8Be1yCpJ>7Gz#l*+Xh??V9#Qvt+rWJ8GHb)ENuXLRxj;#g$4=q24B240NuB;HElqa zpC>s5Memqs{X`l^azcRAxLpDH0?7)a#SEX2r|~S!Hl~2)M6t-uU$G)5q{e5+1%lZ#wW-Y68BZiYT+}0R z+T4Zcx3@}Xo1oYCp6nyC(}~@=)nze6si168m_S2}rO}vG4~IsrgP6iL zENQ@t&TyJw)W2bz0J~U+_K3yl)S5|zwEd~G?z=Gx)@|`}BXul9P(U~5|(L2fRJuWcf z3WbS`4?bSk7HZNaBG<*4WG}bfw)0wVoKQDM>3oYsN5xH&0bV>m$#(1NjMt}7dU@mKGA9%V!AyxY!|e_ZmJ zWfcII+94-iY^)SB6wx`p#fszP3bRgrD9HQ_8oUUuOR(dnPF1Oj5#pFHy=$QmzunLz zxn6+VBkWJ-NQGk~D8}&@~{@ zy2S_QO1CmPO&p8F3isqmzB<)$pZokx)33T6RJZt*Ei8gM0Y^&QD3Ni>eZ1nd`8n%L z$|$OiDth;kbe_q4{X<1m4JjPW?eD~T5b z{ZPnpU2!GgnA%v+S|mU3c{Bk3HqK!;tc?>=qpKFUYBOeDGSE7tI+p3b_o_GbILw?C z)$*$4oy?jrY&wTr$JnnqlLz%U?Fy{_84NR7R>_JwI)SODV<)40b$oDmi7f4K+d3RZ z(J)hT-r9bCkx9cezPl=$lsU1xt;tNAoTOObZTSSYyi6Cc>{jzD_y@e>&{w+N6Vp*P z^LP5{r7NKAix#`fea0(Us2A+^D%G^BFrb(a_=ryPYX`eLU3^`ljSt@7`=z(`ir-0r zwpepVl;(fkt4;~sysyrSo-l9>RaTq(N-%b-YMTV!`DrUy+J{fc~b&| zFf*CMFF8h4S7J|c0 ztVQL8Q2?M5!ayq~!ho|M`V#?>o&1n1K8~K?c15SGUV6*TFb(#buMEap=qQfENB}mJ z$KLk62BMdK_p*gB2Sl^*+2naR%n&ZXXOM_BJj;VDaa zyjTt5#I6@BSpMs0;u?Mi?;3ba)<^cZ7l0QfHlSTzzc535YVmamX8Pr5GI)>_r7noD zo9|?Fs64yM8OzU}bnJn3EXUQU{yNoUg|1o?bc_tSjAoBGFN+JbOdX->Hc;$(imaug z%OrOd@#0k>rRumSNHkJR7PFJ39Bn{^C9$NKsbac-AO*)(iZ=MBkl}kDF8j+SPrT_k zwP6D+%vU_~GlxOMIxI%f3RL_T*ZM6PNx1|Hop;qFR&3W~MX?f^W##(h()WRF$;e;2 z%yBreDrnmPWc6`}-4L8mrTv|0J^5wvzYdTFZiz}x>{a{1OxLth>@|wCP|=tMu}8i; z6utPhZy9`Odb$_*xf*0SbOJ$tm=>dk^(;EXb_3@vWU&H2a@+h;*cjj9VNp?qTQaXg zrM*h@iaL_Z;B?^oN;e$f1P8qLtHC(WW9q<_5W*=y~_D@gKbEISczqlW(gsMp*S)ZeSjBkhkXbOphu} zXIPE&VmRK>sg6l^O1F~bUNOS$>gCk$bgFpf3X|bpEpYg1HC~i(5cs%;azlyhojofV zla~4Y$-Wb0b{@Eu}1?+Lp#O{2ah3$5x4A# z6*D}I_>A!!T3@I(oj5>S)`7g}&2?fWW2M@o?+vUkZ1;KU*FzWbA4Qy}`=R?csFUFY zm=@`N^{OCko5!J`-E-P_&vxL+P)oLMdKy#;Vc=P(g7)!rVgz#J#YArMgur9IuuYIl zV-~`Wh#X`2A!?zNfQI0)GYV`I+^Ywn??-I8VaZzQwEjISJe2#td4KVxx2`N6{+X=E zcr)S6r_e_cGom&qw#=DagM|YMY=f1ZMp;)Rqq+0cpZ=-iRnJIS``+Deq<*LQN1NW! zX8ud{_dfsI>Lq)B5Vz#&*H{0|op0>_#)G#ud~?lrtKPf$_mc+@4&cB=#7?4r`81rs zF?I8O*-yeu_RB9mTdBXH>!BS|chLra|8V;CE;^gu?bf4c6X@mZ zh@O5ZX;dYEGq;UySL8w2C2WYWx#9DzP#)7yYMAPPjyI~M)zU`U(E!{}$GS~)FFj5W zS=AXMZcGSAoIo_C<^T4PpPNAR(n&=($gXMXgFAtK z#V4DFAEokCDg3D9RSFXLolJRHl1AGCwsDK9O{GnvAM=oPOs8p6J)lp`9CBF!orQH8 z=nT_J8i0MQQV=WDW4}J?vf&I3^PY&Suu>X(;^fl$nQImIUo%#^R0g*R`lY>4 zW?9M83-p3?x0DbpgEC(Y9SF}6NVVzE49;kOt9OMb)fI5q5z&Cas4JcanIk z^cZtOjb~kku_g}tW;Q(}h0z;m@s9Z%lp3R+7e&?bYKU{SD^~fz27mbTdc#`@ex4Tj zeQ?U7lFhRYN(&7jNLG|aX&Ys|^BegGrI@5~bABDYiyj^Zg)bp%G?nTea3Br;`;y@| zsg|4mC3$}M`3T7tDfWWJ2<#=&@2>68I4yk;H;wo9>x;iO-vq!6->l7~h?~Z9V&nW% zGmTeCu|Pe42r3Um1MXjV5BV8?wucS*A?Q`Zk_lfJW-rln&>D+HUy!p1PCN(YaLVx0|EP6_MH7 z&*jAa8jQ=QTOXTo>dpnrV3(CB5z=s@`P<=Lwr~fmCcd}alYA$=^M~2 z4yU)*dCiSAZhwDFJgim(n_Ee|6R&H9W{N^bvD+xJ6|He2H81j%8Ip5wNy97)<0ejv z9A_17Mg}|Vurf)pPFs;X(im|JR?TPSgxo3H{wA8|VFJi6u9EjjBDZzNiL#j~2%> zb>dv6le}L_5AxrKP&aVMU?s|#C8p~u=bjV8^-M_-53l2CK|~#$5>U(AyXZhfzi5+p zD|ED}<#p2;?s~zRX}cvYY##5LUo8*MW2Y5)EqBGc{iGE*YHmoN_kd!ba#L_~RGR>2 zsSIB4jc#e|<$4ry13BBNwm@p+A~fD(uK3~vDaw0q=qHv{PF4zs64)Y93p9~|u4=;s z>)1gT6Hz+A^$}<2d-|!WO_1a}d2cyqDRBKd$2p#p6r9(Pz5N%}jQ?47O5x(0iVr)9 zfvzuOR+|7zKR3A5@^isVfyUB`Dy6m-B2-w4{m>1)FL{iy=QL*p!WCJZp*|T#3j7j( zdAl^qvS*{y+9_6=NuDJoN=$WWj=HDV@@Bkv9qHv=B$=~ULB-R4SbuIv3h5m2nvgu^ zMrf8_m+UhowluG%?Jg7 zUCZksrRqBRNJK49FTbZw^)DCH@|vSM$X(Atmz9xsgG!iU<}fM*1fY`}$0L%4ybup?Qk zFo9=FV!$P`d9;jBCw9&5GXqo(#R3sXIz$q_e_f&vP2%;FJkJuQUI00L{A-dJVWlWr zSR^_mxgS(1IuUUK05a$@d>60chzsnk1$KkU=rS8Wct)SkURSMy=iOiKe$8@a`ihiU zT3N@Ed~0S`sP2aCW!Am6Y6d0_1A&>T_VO4r6o$M*ZGr}w&bK5m8=^^<7JsogOQI)Q zWE#o>zHjXv_X`V=KOTBhV?O$ZxAv+re7=3QozXRZG>y~3 z*hK92m77Fhr*Gc+3p`4)rSp0|-&y#>0m}*(r?rEv)U32fd*$0gw3U$J-4~(rh@qA( zr?i#oosf=mz`OP>1U1~kl!VNGx%l0hUnGCMINEsohz76sDz40~g_MXm781CQX!O8e z1zm44Jla(y>U=kCN9YbUet^An$yWo>enTaj-#4k8#Z{3Zq?KDRz_hU1g*se7&=vtB*_;WG_@;M2yst{lj~IHMW^JM9@yEYW`o-%uqTiZ z3^#yxgSLe{S2Sqej*|qOo1n&l1V9D1<~W>zZZxBhhQ#COh}Nb5kymR3~=&9x}s2 zKE*=UZUX%jXC$vF%U8gDxAKJSZO>Q-#wgqTsq*)J5n)+v%Vo)N;!Tm2+T1b^ ztH|!BaKcE&Jw#kZC#m6475zeJMLUo#o#cAR} zIWz$XG_sZff0<_zFCU_lUEuA#@D{jS?s`^D*TIGqp1_84S43At6>jPFBC|4xmK!>T zwTqRXb_9!cE%WW~{#EO1!cF1Oy9Y=*H#dtD8?jJ#J0e24mtw)H&8MOty02AiT{!5X zd*fr$t-L3w7F>`W3mBrYqtXfwT#PCO1JVtmqw0t5=fQEEtcVXThP-|quT*^i94D9~ zbs?%b3iQz%EsAc}3$i-m5CvopFxe49&!)=fX}`2=E9|tpVJnS-lNI&y*5G3h!Q)X7az4dn{U2^E3J!Isr z6PvPY%?j5ZQ*1v)9vYjJw<>mOuuTnCcjeLr!c-=m4{^U#=B7N&Z(~p#3%%lv#6&S+ zr{q<$+7*SoEpxPez|M&zL|yb{rtytdatkfZUa+b>M$TfJk&EOMROO6J7d@TX=g*%uUhk0Q6JXl!wWfq^m=wpqF=X@)RC zixli)j3D&LQl(mBcc#SfRm@d49MeW)_&PECkbAQn&p!%IqYsKRW*VYdcSL&WDJXT* z8wTwS+{N1k6KWd-?En~z3FM1gq(>DM5ny|#(G@c>Aim|z(jcUY9q?%ItTLnxulBy{ zZTKhDl2nhx@RM&|C3oD{2oJ~&9hIy2+uUNQ^zcm)hI;lc#U*JM9ZQxA4Ub^Z48yz0 z)6EuC4%*R*DP_OQBXw?P#rzeP4!{@PJ5FpNS#k3rhdFXzVcjFBFwiiciff^@c{;x^ zJa17g|4!I0$RIBa@AtR^0Vm}0GZWXyugI>)lbs?SF};n{-m=|GE0# zWbHHpE|U>vX$r-H?X-!C&eELpdIYal(vnst&u{O7)P+aqu3|QTAU1k#@XenC4P$}Z zWpch~<1fnT>$djYIN-YVuX_}u#ChL7$=jb%ubZ&(gMwfFnr!E`UODe?Mu3!h1jGv| zHjg4XRP+tOWxCe;4=bc)?iu0+)gO)L#!BmR-t~gv8%Ex>eUe7V2*vv98-iWZRFWZn z%5RK%@T31oqH4a`_4WmO6;WFtloRSlUso(dC>`nrtPK*XvZP*rL#S z*Gviati@WHfygADk=Qkt?smr749Rfq5gGPHsbyImB zQ~A)ndz zuO&Js-mRCLNx6t(3n;RSif)nOt_Rxwot^On=sIy%t2H=zy-R0kdmxp!%22a}Jz*j1 z4vB@3BO214$~8TtOLZ~$Je|%59#^EJsTY;NZVKsz?WH@O3DYJBn5&6V&z?Z_S%w0*rpq{Oe>JHznf48|*SOcok4Abg0e7W-OT0`H6R_sEgJE<&g0YvKi$D z7J02>1_K~b4vRgvd!@o{T!Yd>fY=Ja_0S#mv@OyvL?0_^=f!#1Z$YvY{e-b)Iia1- ztMAlqw_I|p=xWTa2V(a+s}x$6a+QyT6Ao{nc#zGG986y$5i6QJuEE-ds|+ zej>zDPOL$!5KEz{iXJU(Ojxl3sv}ipp|=7Lcw?c%Q$A2JUS@m2#k`l-0{OnFG6+}}${{-uJJOjZSQ{>@{0dc2X>;zwK0XjBty4 zT*xkueCQ+j*Gl@1dtpd(6cRvOliYF7;N5W_;FUxaM+~}T!@VkSo&t-I;fmwc;W))U zVX?k!0#53Wdeo|CUiJ87dPv#gN;;X+VpF!miw9 zFrGc;@B>c3n4+y|$bQus>HA0;6xR*Ykd=w`GciZu45a%*kg*x~qE`rUz3G~ycH1an zZs3@VY)d2VxPfECO7~yC>V;~?eeQZ*kqFtu@`bl}&>;sS!C2MWtH=l3;4a13P-(rV zp00PTm#v<&DfpbGm{}pbya-xLc8Ths?6S>s-GVGq?_R^lxC3T#CeEm^qg%}&A+Ep~ z!?Lhz3NI(%OkKKcM}g&X!zCxviH%w-$ukc`aS##465o8@athjjHZo|xAhp9IXqNL~ zuVNCHCT{x2s2L{b@A4;qD7UO8`-%Vqs2N5Cc#*|hCih~wSS7pw24`}?Z&T2+$)x<3%=J8(%(U`sTA2l zMWgpsFHeaabUDZ?opEu7UJlkPux*PF5|A@89N&raO%A#Rwf~c&B*wAG3=7JkKdmn^ zbKK^??ry?I**Cv?o_y@Yn_P(Oj=;h;irq@VHy*v3k7?N8bI0Ckl|c7!&ifNy4jsoX zcikec4913g*h2OwsSa57!Q=pegFYBGG#m7m&!jvF2Y>v#uV1muHgej;Y%6I{xwK9Lxqx}hs>lS_O3_U^p6P~uR)xH$ z-beh90vRjWi(j8ad2^BWi!kmTF0G7x(i zJ57LPO0V)u6<~aK5;Y6A)npPD5VMaU*XOwIAwm9%20Km{sEb0oUDwS8j-8K%z=^HumhFHB zi`W`kdtjlSeiGS6YmWug(OVascRx?B2!hkg9yi^Vw)x)mxL_|=D>p4SY@?3((@db_ zJ>j&_OlkjpR~Bo6jr-%i%Vdia!)Cu3Y;q|UJYE@8bdqn{>`%NiCB-xrPHdO%rSAl_ z394r|169&3*Zce&IzITCUyt&s@;?7Tc%oOmY%kpq4pBV*DyGKgPSDA4$7+!YAZ5HD zu<1EYNTH5W_30Dwk2!HMvlagsCI;0?dU;jko*Y;uTcjQG1XrYsH_|gX7BnwN&~_BP z@TfyCyyp`C`Am}r(ZsE-B0Hu*1Kp8shTRkkbo{wgbgF+EJzN`}O@G3BI{)E(Z5sqT z20XgpYduF7LbpK7{X4FXhr%aZ223?kcZ2HXXxeteb|Rh+X@khZ=b&yqwtvyqmANCn zaZazR^M)zD-0{PNcyk{BmmaiH(d2!B}u#l}%&9r#YA8DyVNsR$E!(!_<- z-Z^dmKcv3_$ZWx?Be3|gIXPi1K5M({ z5%uaB+E!k=xM_BxS1B}Vh4#L5FK?eu4iqBZ(NqQ8(WHxEPgEH^IOhs8=$GiF#m2r^ zC10Y1JjL9=zIi>fkwl|SQ16;XuZ*l^Vu0hRUVfi{jEVKkl(aKT@$p^IAht`iJhEMN zRiZuR1&z*tgP~r2AgGQm2)U&=%EzJ}%#Jkn_pO$0VX&txwl~tLazgZUdcYOP!^U5s zRD=bquvNoHj05myo8bAs+i6SOf4bwZB}+|;s?_JKoOC#`qFQZ6ned2WA5i2zl-fXX z*6z?jITU0KDDuU<@^jwB%Ciy_3b3r}eAm35DlO!tdR-4T)NY%LyJLPjQwWT; zM>M-bho8|7cno;t2I9XAMq;7-VShNl4mVx&dB~f=PsV?b(_)E}-kb89()!<7aua?< z^qv*gLOtvV@`c69eLfjvBOeJqK+4Zj>9T;313NF~3!ly}X43pIgQ}93ZZ}r$u?(DX zt)yd3Xs3I>q`WRzw$|f<5hso{T4}^r7+%W6`K}DwBd?LR3HI;`=)tgRX*G~T-Q^$m zgkbi9#*iweH#|`|=+Z}bD?Sqqy4)4?Dn8|<1Pr=B2NF?-_ioPtx*V>b6F(Ji5jW7s zNy?9PsvF|VOiI8-Xr_d@Qp0C0it7qoF?E^}W?B}JIc<#6N?BMTuh|b2F2>ELInD&Oxy)N1#eFuq=J7#gP? z*9?z7qpFs!R2&X%jym}r&{j+#kfh|xuS1Wqhzd$<3!|&o7>h%O?VRB7yXVTjJ66naX6~T3^nN# z3ysNBsp$25+|TMfn$(Hn27z`x5a#B9{^@`XdhNUn9%f%v(^{l>$HVThEi4wEInbjx zET8cKVfbms{j{!zif+$VFEl~K{o#pZa_}poaxR#Gr;cK4DN=(fXX9JfCCyQdQoXc{ zs^MpdOX)3QT)TEfonCavXIE5X)ID`4x%q>q-#WD9!M7V1Ux+%*J|cB=DcKlV8l_X6 zQSH>^`N0fP9q%mrP1!ee{}B7m`nP_A+`8z=I}Of1OwQRG`-AI8SgE@t<}4MJAFcGk~KNb=8U?IX=j>=EoU!$v2?>M7EO*x0-H zx}-C7BiTG-6|-G>R#K%LqPyw6%r@UPa>%{LyJX?e*OC@wO7cmkELVCO8Uy0!YPv9d z^Nicxsowip9Niq1Cp=1afdj5*_BG)~&%5)xXWX8Ljfd~Jmw6S2-9>iK5rGHswF|h#$ z)NuijSFGvx91~d%29Ft%EKkNlmURU+nw=EylX1y1mE38~Xe-I#S?&jU>8_Q`Q2}`T zA@ILjk^x21H@xzgjigo4M5>?_#h9WT>+2lPJuc7~&-~+kh8r}BSH0hBS%YGw)B_7W zx@1ExWuW@=6xmSgab7g!vdiNNgH7qOG>84G>GQ!*9%%1osbhQ-T!aAc*BCR|_ZBB1 zIImmdpXDXTExCQ1R&iJ%9Wv$x>40gl4>$;`77V$dn{HB7g$Y|Wfz`e^v}ra@V4dn7 ze5(3Y$Al3FLoS%7rOT;;0}?13j$h%`^JlFE6^!)<7I6M);9C<_Pa#+Fdi;^@0kwtl6rkP>b9yNDp@16^E24uK*u=ZeRr;iV;p~_PF*cibN~?^)r)vF|rMLF?04wGe|G* zOQs7F79KMN!v64f#f_QoU-hnVyFp)%32-k!aH6=I`~6P& zt6qd~1~#6Q-cH4~mnlvEI!P?Y` zzfzojaa+HAc(*vsJJG@1zI~^yMBW?xvp-tqJy}UF$`=Z{4jCa?8l6Tjts6ZoF|eMLPIVqW-1nj9D?pkTcvmBW4q<`ZWa9MfHgos4Hh0W?I}@aD;@O`b3n^T zW&HejqOh>?{8t1mX*+)Ywk6N)S0wS$3M+20@`1Dub|{TeSH4>V!J`XNx96=4dMZnY z5Mx4Mqi>9`9l}8w^i60QeS>}|k0U*FlW1d*qX0evU*p62hzT5Y_~(C!{a>$o-CGtE zDz!)x76N@~E{%!|nV}{@rZ0%ZCLzD@l;baOQ)s?t{!M3Tl5)`-PMpJH#el^vP4aX@ zkMI~_4VIiF@vwVcOi(Ry(%uG1!^jgxfnef20>%v(qiVg>tyY&xXarC!X0HwZtKc=?MDy3 zXY!JLn0NR#Il;{(@5JlMEi?aFGsQx_+j%OwK)5Y(x7#NFY*Pk2% zEABwF&_EncN5PxbOFl=mSt=8>qE7#FImzG#EhmolRGC4mm|`Ihw}6VSqp|%9iWSR) z62t51Jjv3fOH2GeRbhALO3?)tEtq|?*M#Jd0`<%a@ ziw`~wzM$>Y4j|G7E2dZRK3X^LL(wxG>v}?xO=t4JT@(nlWx0XL3Uullo7rW0X46%W zUkNR~uR-jzaqFp9?!r(;G-EAf^X7~JhW>(D`{+Bv_jYuvBncNbos(0U$m(l zj3W-3$!sV$3e73s_|xoveAP{)K5|{s2TqjcsM~XwGhadz?-Z|M|K_Lx-f7ZJ8dP~q z%Is>ANs@fWuQj;5$vBoHI=D_h?(v*VB+6%Qv&gbm+DegeJn0P03^}SU0D_E4pmez( zZd?)!uQk+>Vpo>}$q_Y1|GNcoUeJ7yE(*NwuI-gq2Om+@_-OU>%4d{ArO&3|w)uP2 z2Nrcgx4f3ntl)Ol@G63v8h9R(nvue8V0woe#P-@~8{ z;9T3mQF3N4AGZ^3LgXYl=QSXbx*z{`B3@u8wz{o&fwMGN8jX<-epJ0JdnqR*(X;vNm=Vck+aqLPpKDisTo8yEIGjq~$)S4e8 zFPLXS$sa1BYDl3IqXcSNN1&vfVj)m^fQm*!pBUi^;UmalxwYU4WLiE5?+Zs{{3VBX z6-bDdQ~2{i_ys5~E97mN(;QXdhQqNt+A-;8K85sNw!-ZZuRIJNzo~2kV);HWjg6DG z36QZU8?IG@NuFS*F3=NzkF^!jF0kSTpV$5__?FxRpS0Ki?e`>k8YwlCQZ~gxmO~m9 zeJ=XCWCev5W|w~x~fByCRf0EFw3!$8t z7RXAmYm01UP!mL@djQrQZ{Cws6QpX0p`f$Pf@ytc2e9eBh;GL_B;~|iZ96-0C(tI;LzI-SBd-aBE*}_bZ_U|Nqvl-4hq7E!P z#>1y>MP2x(yKl!&j00lau(zgzMqIZJ>X+ZK4*UCa1H);P#_9X9?`g8ai5D!eVn!@j z8!2`JMb^WD6?tERnOx{a*+pYRjdn$zWGK|J2Fs}5x@`a$^-+gjv<`sLa_a_8Hvh^J z6fde5=iSf|D?nKxOn9w}-bQU-QonfFpL3--ng&_E7#)I3CmSoa0R>)vhr;5;$*lv$ zJxB2i%bW+Nja*vEVu13YkO6mOFW0H^AQg1L9rHrdf+n8~%4T=miz0IH73&BXMO92a zT^Y6IN2c{7of`NaIXjKqGgH&oC>GdsnyKgtw|dbkS%L73s)0f8wocW~91Sq^Ju-Bo zYLI0S4Cv(vlNGpbw<@6i3xan#APmK^T_T`P!3*)?cxD4cZq~}`;TDpp=&O8Ue5n$ z8f5kI`@V?;LR-XodA&Q{!4U_)RS8=u%jfVJ9I@?(c-NqW=P&1PmDbC)s}H?X$-!+f z$y(87=y5_9_4|7+td9v@PubIJiEgy2H77QN%gqp1M6u9Qco!ADNwFLVnSoD))|U7u zkn(`?04Q6-K(h9fywVVkUdylN>FIdydunaP%%}5Pq`*gS44-9bDuWZjotLe_)QDIL z!q9?vsDy<)h_P2nn}swy(0(8qp3y!(t%Ioi>+gz*2{0ut>1W8AuS}K=bjujA`QAdY zNfb%M_8F_(Zwnyrp`Vm3Jh(6+uspDNrZ!JfF*7-$7d$IiCTnL{j44D;m>F{>``*0l z^B-saXCWg-*d>bRchJexTcm4*cNI63%Y~)kmBBr-9%Xv?rikR}lZgwPH(Y4w*jMx~ znN7NDUhAu#?_vxiYs(~e70J8`Rl6dM-s}5NzAfaY;wglYGI@pJ_41YeLoTt@qzRpq zCYVH_TVFj!+hDI;a<$CT1hKNJrz=y)$)IkxSl5KegHh&ZjZu}NN8sfC&BJ#d|K!uR z4=n!T8zpbmfA8rM{Ssh;OjaCN_`uuvT%9-vO48HlJimic4v`gCxfpjI=6}EBl^WM= zPI7VH*h>2SsqejRGC_V&@XKG5?M}QSI&5ZC6jCgd9OO{ZNNc&=5Zp{9rPJ$yo3uB4 zyEbs&51WZR(b#;>1{lucQ-%>q)TkS6vIXiQFJep zl`Q~z*&L3ze=2Z#%&V?JFvsLHxRXqrexu-79Xc!)j%10f70s_u`o&0KiFuOi670 zr#Bav7MwK2KjoACPQ2iJX13rQqu8SqsiLAAWI9!aqzIgDee=L6ve}~^f&nS2US1rl z?TP5nTp{b-@`Z6BvD9kMrw}~K@biV~le_EX0*wjDn&ZQqpfQyk z9ZZijsTjYN9~Y2APOM@snyHu*6nl&!N1+gj2Yo>_x%3_{W2;8&csKw#)omUPJneo@ z*mU`(cs~f&%Wuxdg2z2x`NGl#`9f&FSf&`1=R$YTj9ExGh(V-mO&RoC%+laBs9}Q# zQ^9*XVm{1ksDw%;tUwrYF>;LU_+dXeuO7dq;Q}Qq+GMQD90BxvI43AgSrm9>dcd%U zwySsr$(=?(KqFk62Pt+RMT$XU^9S6E7Z=fKYRp}&Wc0M&|4Ys1QZyi|7Hki}m`aXh zS5)Sk^`ul&FKbnN3B_bcSvu%)pZ+2g{th@4ndz~|qZCr9Z%Dd)4n^)Kh4T&s=Flfp zcy;{3c>e1@KeD5Mo=<9rBTfQx-i`IUZ~jw+#DpCAU*B#d8=P1P?KXpS8pUp>ND38= zRA@u$H5leOCEvLuCUVHdkbBW8x~16Y@dVQ7aD~!7^~+jxRNWxMX0Ve$*1*X(J3@!p z0h=$Gf)jM7^u619lxISSi@(bXvU(cHG!sNJ#crg?1}ZvUd{2GE59`jgd74zOTsoVs z{6;p_y0|nvjc#V`s~BGLOLF=8Upkdt2XOoPoi>N1aMfF^rH{pFQ|GLBO|XW(MVcqk zLT61yxoV$}7Oe|I?fc|{mBT7$S1i74GHzJ?aM9oW=X8^*==?ud|C_9J;>9G_OoydV zEP!_t75$9aC&{;6)kfd-Ob)sQ2CWt(Gy~7nzlLae{a95C}{@~)acjB65E851`Rvy!o zfPqD;SJV;o&4Riy=1?LXRT`ZWGAZV&@xA7XjbUxz=%4@iHVC?C7XGl(NJv`nipE`zg*KCb>F>*-Y==)H}Tu$$9?jc9{!h$b9lWl z`iAnlk=$0`!8+tKXTMl zTwyh0L&Wir8xYH0r^~p zez9x~rul0z{L`q?0V{aL%=Nr@woZOsQXcRTaE56w0^NtP#hRf$U}aFccw0y_Ii*<@ z3A$&W4-))9PjSOw2pY`pls%xeI%xvGRnbqX2!7le&$~hOu)5(#8JpO{XQ{OKBtFrh zF=kd`M&*EeJhR>Fi5O?L2{@Anm*?oR1vs2+o+*ES^(Wu4q;r3<}@W$5@9-Q`DLu%dAPK-8oyykVT51 z+PARQqY?;)kmCz^>3~7Emf)=m3s1B5PMjW7Ha4RG-}fSL#Zpx>>>(FRS ztI?_Gzic8U+>9J2RweajYu8bV1%ktJq{`gMc0%*j0S{~u-l#e#Th1Jgs1#`v!}G;G zvQ&smAf+94kZFXbQ+gnUNb+q}oYJJYZt$(;W5e;C($D9YYK)ITD_~i>0v*EIj`^LC z4}+J_k$M@VZZh}@D`X!--CZ}tk8E0p#sUK;$UR?XUV1dvXWDXa?*8sTZxe{#J2&M5 zN&U)z=#UwR@+lU!RasPYzqEkK*5I1dBE>8iZI5h|=b*d;$O^T3`k@4UI)g4(-Oi{6 zT|V|Wrn#j!$~Sf`G8jKCQru_lk)M6T4ydpqp$UV<%DfZ3{4O&k&ff%+ ze<7`9L`ZWV#R4T%5fzPgtg+c0(Bl&1t{iePHWuy;M-_s6mV4cR2a~@e3bxuk5nCd( zr9k;}O0y>tC@=*b@_gYf9)_NQVmEM0L;@d254mL1NES9>Jz*}Q2?ODo{xF|+qG#L| zs{ftyADORudOxlbD}-1o2E3G6GalJ=ycaka#Q&r41z;R3oC=K z`<->ysj%ibm&Vk|$%BMZq8xzXAP9^;Ek|Gk|II&J*1dm49K88vFxpPBDd=&G#yo5= zOM;DAbFH&4%!bYa5!)iQ*c>`pQRDN}8_m%+k3y<&(yE9!fXTSpj4pe-UoHRIk91xp zz0vYU(;rBZ6YGrtIs6mD4gK-fKodMPC7YQG z3v=m0?_Bz-NEvsJepx z|4NXJK<5V7-wLl@6AUYrrY)b4b)&IjIB^tlmzf4krPwVLNur`pf8)9&2C5EM2HpRb zkzBQe=?vXSx|JtrBQMJy{t3EHV&`q{Sh|Lrh8t@p``)t-m=7vnYqQLC`-<4DRx;y| z_&-^Z>YqkyvHYWll!HTL5;5{jvNnO`+33kT=nA}b?ss5?IT|^GmY4x(Fehx*DO;8zW8I3r0Bye*!7@U2ho03ukw9a_;k5dE;gCzs>$n29hDKkma#K^6=%8E-mF88@7^28fI z6^Sf;Yp+zYi4_kT?$=wSpm%Twf7df#xYQUD#>6H}J2GTCwMeguAo&23C+*MJHi}fn zsY*tf$-bAYtCBS7uikytn|L6vu})Pf>Qg+NUoU?YQ5tUW(6gI;F4BA5w8sJp=>oc# z&h$7a1(#YM!4sV-!@XD02pu#J%Cv3tMd?K;dVY5ExovHs=JmsEcNariE>R1T?oqFpA5cV`jt(LBjIwUImg zDHhhDTq-(S^O$}Vaa0f^9B}U@F+mxccBW!hu3Ha%Q`s0=P4~!7x>n4J2{|3w9Cgp5 z9je)Dm@}#)a1F+g3O6Al;z}!!>D*{M}-xA!d2Whj<>E{@^3{azhS~lS>S)J zAq7sn89Hf(jUyBb2}t`98<^6pQ^hky^7EdZjJ9gJPKnuO?cg=hB25PFh}U^;fK(L7 zEl7str? z7Z7!f;l9xsO-K_)$|wLD?Q^(ZvyPNemZ2%Xdh*5BmzdBKw(PSuQsKntfery9(v{9o zEc7O;10Fcf3ts0%7Zx_rHPU)%tpqnViBKRNOWMQA!`kRJL7wsw)P${Y-6v}Hjh&`b zJs(>Q4AyOee$g%Wb7GtateaUKf;69fup@4cIs=`{N_{?4u7XxVd*n6JHUU1_7`!>M zfF2A3Ms1H;?;77iD%Jl!oEK6bP3sLG?=t2HJefg^E09d%IpK-g`|o)juX@C}McOLK z&|pBa4OXj4QE&K#*$I3T3IC^}YFL=M=mwdgCq)DBgQd}*s@8jcD%w5k>>GAZZ?lxz z_+er>lYP%Q!DQ;Jx96VrHL0D#p?43E^jFNtsWekXdnp#EH1dHtgK3wSONzzm!rh+N z#MOdRinR-KrKib#@7$TqQH>1x^LmtB@`HZ0eDwa|KD;@qT2P`qOLAu(@p~e{y%ZAm z#yM`Mx^sPzyNn}PhhuX`LYCe9l&Nv}; z%7v~Y4=j5YI&HF&mHvc!Y{Skg4L`d?n?W}6(Tkr2jDb}PCS7~k%A_(`Sa|k1oIo-4 z%(|@~S_XQo#7^|!*EV%Az(nLUaTipA8f(c1BqfcHtoOi^)pME_Y4LM)svJ$JvQ~mrfhlYZe30IwYP$BIw1SV}MdQAj z%ge$Y@0R%IbZ+xI!VT|lPW@TpQj={`>T_03I=JOKIq`mG_0*A;%p;0@K#}`Yw2{Oj zS+Ux4hZ0vUEZe{mL9}ox1>N*ALv~z`au3jJAEqBbCO^=ypX2o?!TGYpP&AZiI5Cdc zrs$cSutcXClpk{aTxRUk3NzfN0o|4Kov@98sZ87AEQy|+;~6HXrjKj#H0RX~t^*3J zPlN=U8Qt%H z0YxEiK%q_Kb%D3Bl1>v3%HaelJ<{lZon)Na4C=F6-+e4IA@9`pzJHBubYh?HUNftB zC&i{wU@1qhimdfsL9%%mP~7Uh5jq_1llL(LygE8dGU%cwWrqG91H8S9wtAly^~1Hx zLFtnI$;iqbpV)2IGUFb|Nj*_#ZxkG`1j&opFV4I5BUXU3dYaxNL0AxRYr#j;E{S^R z6YpG?{MSvza_XVHO^^~$FUY176pc|IyPbV^pHHu%!fo3U{CUC7EX<=xaeTmWY#|Z; zJFNrf|2_!JQJOSK{0 zE9(y!{D}!GKR!c!i+tw9tImBh>EEQ-8x*;U8Qax@F7@Gv8YYMSi2u|t-*2sF(VTAS z>VW2`lU|P^vgw_4q4a=T1(>|I=MC_5Y=v9i>=wV~s2mMi)ETpyh#oSijJ)tSphr1A z-6TID&(^GrY$8UY4cYSvk| zjUk65iTo$9{#EeTdvAF*(`oWlmduM0-jHnf>Xan#GkHaHj_=FiHc{L>0}!`^mqN?B z_8H9;`OS6T+{u_w_WP54C&)@}8p(;RYAA;tp^-LGYyw5%kR!Z7b}XPZ80Zz+6;Hgg z=@r5lVUll$fXkNZrBbkMfUOv^Z8rOj`uDzR>n%HjTr~SSpEcmP=do6i%jiR&`X7qi zrm72vU;()hzKdS<#-qr!o_&g9`JNzr5!)e|V7qHkXcfGo6>bm0a_O_ATy!WX)1zK` z%MFNVVLU#cFRqu>0m)Lm0LnYO4@b5|Ci&I@L1Z<(Q+jXiXLIYquP|HYRJirdJ`|M7 zlqh#fyQMFGS;zQlY?T4G*VH(AjE{ZkEpG3z(|Uuxb7~s(wn@S4p3U4R)lR&OT{FXY zJ;k1($SErNOL@L{y<0r9l_^tWwKWogX%DJGXmylfwDMXA5?dVkcHO=gg2$L$Q!2a|9db-7qA!bU^q7 zRLvGb6JzMO8HNr~?0$-rP|+nyJ;W3?1*12{5cpg@CnW$^rvi08 z5EW{>$#qFy)Io8nft3W=Ew%BaGqfr6hz24Kv=)gmk!Y%o#y-|(CCmAT{UMvUHl#KL z$HAtIDGEg*L;Iu$bE#_^8z;~2P6#p29_9p{sY}lMLxyF)#ji-(6?k+;EL&SCHkl$D zsp!>nR{1sge!@GdHXb1fH{hojI8x)8s##Z|!nTX%XsA5<^2c+)Ne7MR1IGEx-+${T z!G6Pg?LD;{NTw5Ol^Qc)mr^YFhj&xacX>x#S9_nGQQ@|EM%N6T>M6g8sbcE+%~98c z>!e*Xk`+DtbY;7|2&xlu)lKRFdeG&9cP#j6HqRJz+2k4<&@NB&*9&%s)<7_{YSw^9 zHQgDO#THGVg`zcaaKgcJtzmg&ZTfLv=cG`mtJ0cgOFy6$cU!(NMV$)OL@5 zin=5S%Hs29<3WepQsdQaU_l_!lz{`T1I&TJSv~&R8s^vxB7d#*fA_NX)QFRy4>ep#nu0 zSrRnl(hZf;Sf!L00aV=)Z5|lOtbz*3GGJELX)1XgbgbJLONpI?X(?P=&#elZlQ5mv zTUL^E^4pe_@L!Q-%L;*fI+QuZh9LXhF?qdQn=kBky`ik6V?*MZ(r~Pdnq+?<7l2zu zqXTX%Ovvh!npQ2*ebMho0yht_6NfyD%+|PcirqnxR4TgThZrUAl$6rt3lF&MllQ|> zleCssDL5qQl%&JG2fUT9chT`|nPg)`VfeP!Yk8Am-*AD*uwrunB3_ts!JKXH&M-kl zzi9q-vTYhEGy84}C>E0Tv%!7{suvt5=e+9$S2T}Fy0Be-O>tCGFSs0ZoIVhdOk&w4 zei_x~*5&^|dY>+$Z_D>Uby^xQ2r zX0q=gr=`YuU%NLHfph+7LW?3>AtYaLQz*{6h>|$7M6Mx15KOOn0kyX87l_63D4d3~!_J#fSaD9_#X|QZ~ri1luXt-ZTBc6PgbZV7?s(_tT$}hh7iB5%s^2M03b5?>X z3t5UROZaMS${HKy?`FgKGmptY;IXL$M+aX^Ux!&s$Nt}Pq zyHb=yZn;;A>;XmFkY%AXUVb`Z;8{mj>uaeW{m6u(bZX#xBItXB6LeLiUSqIoR;vf8VB3>RR&P=xA<-xrb z#x?VW+BSi)8nFp>HW(ex4QvNaKIBF^Am7PqwK3v&28Uvg>@i%_wn)H+G2;b~4O35O z$YXZNM$5DtE-qpxhKQAfoGeJZ8I*S^v^|UJ1=kfv)K>y}NK$y#qBc-Qx8`b-eB;<- zK8gH|ZuJ7kbK% zqLMx!PV${BC_I;qt)O`R#N-_12mJ5ZmdOxK>u9l(-e6b|uv7(!#M|h*p4h+Dzy)mR z=5yEcs;CRDOrFqf751>TFgcMiYT^# zBD3VGj&KvOm$9X35 zab@~A1k=v@-RlLLrgw=P5zC$syq;st7Q}EGvnk*kmpq3(E;T!v$R;P&EX8IJ$fQ^> zHg;gdXvh!fQfR#(hi(rWf`YgqKTNgip<`yOn|n%goMh1_RHvZncn$1XRzP^>qp*tv zM>|rX7{xG7xpR`?cm{J)E0k(ZU+Bweqn+3)wL%+h36 zUiN1lBr>!r2HdYe4nmtCjqZX10(@B8Da&K3HOD*;YaEgpK4Azs$O5*$Lf!0meAF_` zY$ZPXK62uIB)Y;>xb?d)XFBBApz7$NT7Ih6wy?6`dwd6Ze9x(IHQ8ZSbm$6L_cA{Q81lfC(cdUBxR%?pJ2he9R0{2Pqbu zGsRT&C*GMJhvjQ!Sg!yJ6P7L{`3|`xD_W$hy$$JYrQyB6y_f8^M{Z248*<6?JTF4# zvE{<5z|!!G!FuT@-k(5!ZEe1|+ckC$mRKGXFGVkZre}HB6(-irep++fID2jbk31ST zt}GvXPnt7jB2gYEHfgLxc~Hw_0ey_2QHox!2Y5jFTDzi)u2aTDLKS~J(&h;x$?nXdtK5UrdQOF zLS{!qPKXv+T~kP1K!*ED;T|ZOtE1x?Y_z0p0X|KgY8P1L@z7xDzWe_q?<^-QPyN%f zAO1UQ(l_pp`!16$PHgK!-TjFA=v<10#(o)8^ngbrP~SDG5IT99^jSNB*-oc==$7h% z9^OLfPT#VHr9j@A8>meTe+(Egy72PFSX!AUImOQ(t1fclT4q98Z{i-*#%lXJpcw~nZ6dh?8ZlK+prFM(?!%htDq3dzNgjbKtu5F~;u zwg`q1wSzsq%yiF8&%D{U|1ACF&6}RJ=e(*G%EbhA)zIaXo@7f=zG(@kyO>K8>su$x#ymH&iAGHXT#MB{(w_2+&O$# zhW~}Y{ZOFY_9i5E7MO}~n_7cHwYhXP(IiVt!tB1`|LaDLJg;Vn!TEC&Q(%PBx@yk2)B)U%LHL-1NBVu-w&0a4VFhKFA(m18ZQ} zc(54mA**0nuhzF>WiuwczPWn)^Pb4kN6P1RNwWxqNVv!{4|xQbV6f5m-mH@Fzw3G0 z;an~$f|YWL{KSHO?!zwu^#clcEEeR^hauO+j<8`16r(xBe)YIdA$lHdZuqp}OIu7M zD-LT#%*0AEL9HAmJv)Ti^V?)sW@L$Ws4#CI2zRICxwg9QP*p;f+)8ed(2j60dBa(v z!{l!J9kMbb)GI+A$GHao)ydy!F^L_m!oz`cs39e8#Of!7VpdaR6&0nUvBVjS$4Bmm zR4Cs#=%l0d1mGxIJbo}48feS0Vn1Ar*0Xu6h<>!X_uB?s$UpLbo1Ahc9Y&A(3dLNc z$ayNNFT9IrJv(K?KbzE5!u#Zg`ZvG9!kV+8^~$rM1%AlGUPV_-xyvhpnwSDECI$Sn zimns24-rjnda2fDvHR^{$WfeC3*zCAcvx7B*PGP(j_X8c)Q3sYTeXq-keAfYEfaUq zr@aq?WJLea&t^3i4I*h=kK!=7Go?qNr?703?}Izu|9igQVV^enVG>Q{`(0!*-Tw3} z3|KyB#)jYOqtQlc#5wbWHj4Ro+_0RsO?g*gK-f=DO*li+9C+mn`sX98OaaB@Q6!s+ zDhGt_=B^4F%FAwrt#8QJM!~Ut$|7N<>#jK|5P8821Rd}e0^23hPvh1jRB&|AL&`Sj z*5F;3vhXZCj(`A*k2XLti*K{MDJv9DjADK<&B*|k!q0#GA&DMsnsVTEFzi2%08k>u z#8YGiPE+0CNS~U>X`o}=Zt*ee-#$2a#>iL#lV@~cm&@j1lFQ4iw?;z1fqfb?5(3wO zTJ@ed(WOP$@76Ko%;2PoRK4R58B6d26QNx&@w86&Ms9!Ilu*tL6Ppf9Hzs;OC1w+> zFSd)~MXjL9PzGEzAQR^LUPWZetK<1#EqtVU_u^^Kz{c0izk;Lq8;<){-Ol~_)=7pb zB#14`S|bRW?iRpo|e(AQL$JaUqhD*JfRUhlO? z9&o$JCE?zP@e81`0_`{&1a_c3HfM3#L<2TrmIWUogHG%^aM-LpsOlJrq*WBNk|N8g zsK;7Z|8sK)+9?b0V`Fn>S$C_e&z&;Oc4V_9&^V^8H+4uNA5JXKwL8oA1IdSJ=%3XO5rp zdES*kR!zC@HRy+he>J3CwdHG{K+JA1td$&y+!3%LAZbA^ce@C)>abnYoStyx+twEh z9waq<{aL*Gl;ke21_}hLr`FJYl4fD8D&w z-YRQs*kIf(>4FtDCjKyLw(@z`bI6<0C((MsPK>e41dxvI+#hRr!9MAd6-TGcsb> zSfj?Q9U0N74L-6$&4lAGGnCA*#L1ZQ+dq&L2VVY`7%f_IC)zLpMb7Z(N3J20C?q6st$?xS_N43RQ`zH}YgvvE%;!#7sD?K+qG8+A&u>x4@bw zCNx2jM@I`ZJ3TSO{;KC9L1It_y(RR$adS{C35OWXE_TN67u}l&azxBeE>)G}NrTFz_W>Ok+Omt2!yn2(C8rxf)O8FVuPcT{q_xurgkG=Rl9Fp1Xi z>uCKADROM4RidXG!@1-NbfLJ2dNk|aEZ6`zu1uSp8>(*!g*|QZ?eliHqKGfL#78FQ z(dZ6880t#XB<+0}v!P&{t4&@#XYUk?Tx2^z>lyHSHsDx6YvKn%Rb8e!&`(ZX4xBh} z*$8f5XhTE6C>;wF=^awQPCTeCl+#c&irKR|sPe%?>q=K0oYE?BRaFw+BH8Gv(Jz`> z`TcbE{908zGzI;sWrEvs&57U+8dq9c<%&7D$%IMSL-nY7Q8mbD0SmS^yi9}@XVv5c z(V)`WR)PMokss^PP8YapYGj|xN|^-3s+P}~;|;TC15Dq-Jj{Oc(O2r9_ptBAsrv(4 zhvdGSl!?6EvIHRNSmnFRwbHxZo)jIUKYU9eF#6s0J8K>SqfBDdUmv|6U`je_hTAiG zcB2$)yZeE_^Mo>AyjT#8wm(@^J~tZ#RninKUKW{XoBipQj=+%dw%NBnyCYU3;J8No zOaK1x)?|Z!)AI7U_et_|mc3SNGzn!>3}h&2si+*44&pCJ<8~jWqm)USweAfd3e`a0 zmuWg(5<(JSYe}oB2;3NyeE;{>cQZRC|a8<*C!Fpuew@A1F> zyjN=EhU(~(s)yoi)euc_ZfF)44=Y_i{NS&<6#3!=9#C9VdUwIX2$}Hhi~;>PjRgtT zjIgEmnn%JY^UH+(+$E`|l^PDa+ia#Pjs5 z<5<0eW5fa%mJX>++OfmMH+p}%*EDy@VR2z58)>uqPR@1D7D>KXqp$7T8#XSM!F0GR zL4@f&<8jOm5zE4>#9;<^L+BE@ne1@jaLP#|*P@1EDk-v;idqxW%3tH(F{@jiLu-m; z9h1`}{Sb%K`j&(v%`^&)FBYHxT%4;WiL3R+5-&|0!Oe&a(SWQ#l*G*gtz1o+NBm^1 z@3JX<1VbXYTY;CM5h#uqh;}Kou!o^p5_DeJCdUfYlu6d&&JqY2c^j4S&kT^PJtDw)P$$_oU4#18{gX8P?o}ZzutSKY zp7@IH+q|<>E zDfF`ACD4`PuAAO7yI1&tYdep~5~Q%O0yaaDd8CXIgQFHM|5dIva;1*DMM2ESm11Il ziX4=bgA+tO-WcCU|2ajzIiQAafIOHS*aq$broaMMB#nGD3;%CB)I4RF#{#IQbZ4_m z`TN)Zz`||7rRvP^B|@#T3dlFF$?tL+q^AOU$$+dCHdGq;9m0%&TFJOP%i%ws%`+Xo zXFPtgdr2L(7W2a&fBvS_F!vn&+Sf0W)eb!O>@b>pHc<>D0dJt9K4%7;PDhjj6IYjX zum81Ki+KHlPI>^!gBtjox!I(L-@rXW5}-I}3kO@}(OS<#frrK;selc7A5S&*$1gkd z9$0a`-_&PiQ+VpYYbP__><|M(i)8nd6&wxf58}FK{DU!o1Iqwpf54c2L(u1ccJVUU z4c+`%S4jE`qpDGEWM#He4A6+?P*F(CaoMj?p}|-`?iC`bL5#E!7QDFTL5d~J)&`|Y zx7=xiP9?HdzK-6;QRm67K z{XjDs*TukegRzV&;(YZC_`XF?o!R;e;FAv-8$8$}9${(dOO^%&U}wtw0s;YQU-m1gq~8X{No`sID3TXr$< zsOlKA7@}aPjC=`Fr}a0Kgw^^WDI1m?mdKz;Q8~0D-AeBdsN_8v6Js%8=9*dtuqUjC z$q944kNtcycaKb9fXw?Plm0?hI>U1v8U5TqF{#LN5w(oB(X%A1oz|dUaEl}liok|= zH;12EwyKtMGT?Lowz11I9PO`U|6E7w#rEgey|@m$nDKvS(i^7y(`@`y2M&dpkzL2U zc)dnOn>^7K|I;czaXT{?QsLX=r@fK7Pk*^29Js2do%SB$JkJAi(qbqF!vE~$iVT62 zJ>DOi5Zj~tagX&SM-EtMM;4R2^mTc&x{rCN=;rp)OGt@l&HN(nK3Vd-OdzM}@L%ly zQD}yL_vC{%3{5`%RY(2A_PLJQi5*U|vLlX5N7TRPpVxk1m}V~gxa0w8awc6y(@Y!1 zT%*Wk++=Hq+7^7zb_aApHcy8Xef)i@ZNRRINv47K3kEOff_A`ao^)ySOiaCcdBuvkgQi99X6nWV1NKfi##DN*_rFQ5dtUd< zfDI8%(&SQ~Jo+w2(;&Iwib8DT=IO#;$nv3D`b?~D$I5y*u1hHM_WSLR4Ax`qKi+(e zoEf=yNxG%DZ&IlNg zRfe?78f44mccceZm%KK3eHNC;gTM}a35bHpd(G%MSNO#3WOh)d2~Pl$qjCY5DVwmOlnr^Gr}>odzCasif65x$n1z3tL&+H)Qpy@ zuwjS)zuqz+DKK`%_sL;)4v*uya-_`&cV{W)V~R8&mp*Qr_qm-EH$c)tQh2@}Ql;mH z*7|7jIhnjxxD+QiDaL|lynkdppMMVd%AltaL zke~4g4A4^bHf}y=39n#s2Iyt(p0SA2=~^9F3tE{a>NvrG(|xbiHNVXZz-JE&_LMEl0~eP3MO-tIvZ5d@otcspZ_@7Re5UpQfADy@uYj+*ZjJ z8o#TSv&Fp-c=1jsk@Ewl*q3)i^vipNXIzfL=Gc*G(SmC%Gj7HaSuAihVZv{^SG*Zu zKvq!3zmiC<15bW1*NvF`c2Nv)*6pC89&ok_Ab=&Qf*OMa&Pg$rA)WSab^lyZ32x6- zaUtC;yEm(qk6{NS**#2prqy!p(Fb{_A%m}3{V6{$=oo49sE0G9eyNacv7c;#sK-vv z2c(9V7i8Vk_QY`54hm0nZPf{ObKM`m|COJ48-SyVUtUc%Ixukd83AV-#lWUY9u-xu z%;aqqA0t^)D`sEh9Gl-bU#pB36o&V63&YQWpL8^$gDwm&;Uv!6{L+9^SKzVv15PEP zZ6JSzB19e1CXZ|t?v^KnU>`Py$rod3HD12x$+jQZ;pM;idvgAF<=7yyM!UkED5VE#O8`(DgEc3Jsct#X~$F|RzjR@Fu8zp08& z+mKCaRmr3Ioaw{9rPtm3xQy>taXpVLR$IJj;y zFgvXi#ZIr}-H)iDTlqEg3UZF@1O~GzP%gtS&Ve}jX3`5^14Lzw(iX{oST=S~I+yl7 z!|LUC+;z>p>92j)(*VJyU+#L3q`okY^=_j%FP~z77-=&VRU>=EL#MinKKNxl)8f9! zdyqc79Fs$$|u!4A+xIsz~Qq*G))6?Nv# zE8=MCy1bQi@k+%zWl6j`x;|o&suZ$FZn$MKO>?e_v$ze?IvS52@oYS_<_3iQkelnt zw&z)SP}IY0jq{fb0BQN^tz1(2!Z<3Yjo@^UVrnT;MMd40r*QkhTZfuAegQWIB6w;3 zSLsU80Jz(gq8d6@u*c=z%u|9q*JgE^|6+*i>95^2{LD7|tl;Zg54Y5x$Fl=Yz0hVe zw*hMDIz96^y)&Z)wH*9uH8k0S?y+Tg+9zuvSl{(d;}ipozOnV+e@!+xu$|d!1dKw8 z$)`vz6?ORaD`H((9ep!wHCay*=7DPp#q>S1PYX*x+oW&SE%0W~gBtoW=MuVD{;}|~ zC=PbBZU;BW?vpaVI(p%f@7<1Ue*NfA?tE|Uw-#DG)T0lH^#B;XUu(~@vpREnf0p#W zOObD5_$ux(MhAP?{!&4urMT7j* z9_(|GmD!m1PQ|O2o_BBdp~tqc8(bi8N6jlpH-mrk#YcGuNs?D(7*fw*0tIqwE*5ml z(m^C(eBfdcK#V;I`yj%kS5^R-Fy-jnxu)V1Y?j9kyq|4Gs-nX`OO)#Meu1A(v6#0> z{fPG`{h1W`$MjlmK7@y3nHw|oVYrmg27VW>F#G{GOSDsTRJuffiTLThk8jQm%MYvY zU!-bQ9}%pVbkTKhppI<$+)}^4e$cSsvbbHA%k2(tRcq>Ij-t7z_|)^(4E<9(w(=q?8#J-&_xCpa(XbSXTb?kNbU5%*=p&{5q4D}v->beu8(t~$;d}eY z5ppOn!{;i*aR;5?>=099E8QmV<^r`^6DLKEBxo&?OehNO@$d1+-LIi9%vN=R*k5Wu zVQ3jp@)d}H=SZ0$!X|@GsH%fkx*_cpBZ3DM2cURdgWFrlzPAGli+hBalbmo29}0Jru%e<{aQ@P=2F~#4}h;)EB zX^Skp-8^K9q(A!pRD*endEaX#$#`J`e!w<8B8_W1#XvwgA9g0cenngv(#6AU^~5%p*$Pjgr3YDfq zx?GwPd;{_}aQaD8A7hrw)8#PO^EBPEQ;G}R{nO41&v7()^wqHA${hiAb$D34D|^|iXN?+W3Vb2@}CuuvKPjxKtlQmt8$29_EV%9Xl;UfB<0j)L7VJc zz*?^HQImS`?fNgD{NCMHo7HF3Los50hHu3b+)}9FS4`>Vru$}yGDCODJ`HGqC4XvQ z1OK%5@(8VOQ%E5%*7-zmvZ_W_PcL=GG^0vpEw`M?4r`Mg3ao=fq(weQ-K~Y?SW)+c z#W6j>9%x2){Qa{9-+kU?a}L0wdnoAFJu8{J3e*hpd>Z(g2H6__R$zAp!Ibrq+p|C< z49Xeon89H5SjS*xjPJMC5i2V*@ijVVTCl;&+fnX>J5PQK?#oNK@q3NO5pl&A5r$DejP*d!=%C}p#OS*oV6KT!u1_~$$4R8xO>!J~S$@C>CuI~rVMm)kp4!RPC}&5^!6ReUU_=G7M{`pcM$YEfU;w)qEnX_DtY_K1!ps9)!CRIJjU+ zcZG#;io8kLHdG#U2RIqhU9Fv3z8ET7W8{mDt^8mfBcpjXE8h9zZ=QDwHl)QW_1PLw z1oZF)f<|c}=YLGv#WGM>Il^POofRl1s%lzJni2y$?2^cgaA_)28*wn`K4>X@^70|h z2iHB*A%Jo)XgAQL<OjSgV zs+H`PZ;&o^>yY-#PfqTd-9djAp6t6fa@AW&ynTL0WnEBFn@z8h-+FuSyKBF!i>i%` z{}FB_Eux^B;qB(vPb~a%;dh%O&oh0J0jITJt9b3to2YYdultV&kxh{wy*=Rc_lIw3 zl^wJCB^NoBu4@)Ne5=i+6j(x&B1&LEglq*+2?=Zk{n8|EA}{Id*!to(W50=aq>^TJ zk*s`j@~i(H_vLyZ{;4PDf#5*vo2saz^FRl=z$1@rlB9e)?cX5Zd`K4=Bg5C<_!K(n znT^UG{qlTBzlffl1O8GT2mg-iUtUjLi+nL6z|?CLTqgr*Lr6zQ#biwmY&YrPLjxU0;bs}s z<$*bBQZ|IZB|MQwAC&D6I7Xt`x!AS`$*9N~I%y1prysGJ7ANdK{L`Ov43PT6k=}Z; zZ4qcHNSrAQfh z&v<+`kCZ=$&^t{1Xfy6=X6Pkhd?>t1-Etj$fx%+PPEh^R;1+YIOSk-yB34m2o=UTp zK5hNmHXhGe{~hL^_GrsnhvlUGVcUrx&M+8^1=sRdk^**4jN=lJk$NLjvX^3@Os1A;fb6Ey40sz4xt-tfYg-Pssr4=eh^K4aB%^gs~Wo8CU21K z;9LsPGJB`oQD|BrC*x8`1Gi}MBA0g9*TM>_t%6vfEXjem&!O;yK$}=6cJF}sd1vGm zTf8Yp;v_$L-nS9jW)J$+)0HpPfi}~w;2yVa+)c2|N%u{rKL3R-veGq$>iD-KU+tU4 zmNCYf=gV@ut@$ce_p-8iRLcBoOM?uiLhRgf+0Op60&i5@mGEqf@Yj3UKU)WTTj zKDPqDot|ffOPz117oMbbGe3DbkIoCa9MR>uZbk{WJ8&SdSzRW(>3NJ?U@pV*xFYDr zj0S0mPqR8-IS`n_T>-17OTJqLcVE6WvsqmXM48*?o%X&2(W7Env*J*o&f6v?ML+H= zfi}{9jPjWkmY&-9r;Dz15Ce3Ioia~AF_6Q~pdh|7(q&yoG07B3qM}xEzu*r+T@F0N zSp@YtN+Uk@Y#K!_^r_Fp+KX()+S>18b3ZGiGBLaVa*oP?jH6NiTtkYVvjt9r5nA?B zOf^L+^b4F6keS7W&QQwc2vY8$d``^GOqJBc@9&wQNTdT6xp0EMdC|1M# zDK;n?rLB2jEyx$1HeDsNsb6qlr`OE-`69^+%A4|8*oPmB`rR@_)hMATybg2_D;3rb z|0qxX>8C8k%hP+Y{nb~O2K{=X!L-CI3qAxKmgXV!TrKj*MkAxMiegq$WH}XuESvZJ zf%yYXN|T~e@jnnv!Ev+%5w;)iXZLtla=hHsPB`o)lbJVPAIZ_&Cv|k}^fe*fa?K@S z9lbnc+f-{wf1myQOj}qo&O_6I>5Xp*bLADxMgN+oE zMv-+$2CP+L!v9e6oIaO913bOJ`-`N$Es}n>K_|VbC<_=r@6}?=JoY2(j23P3SSeUdFT=6s#z>pq(1Hq8COF`eUespNSpAnKErMd zxO$ARr+l&&d_O!jb=q$Y*5%gRRny5WXR^$weC%_I>7__FMlvx-pCa!i_en+I(on4u zw@`Jo&R>Ic8D*r@XPs!1s(HaGt|6e0bzqH=*y$Tujp(1}j69SS8i>7mV#9&=O$A5JN% z=WGbooR?n;DI>RLXpr9#WlK|oZqgNjFGk{k$PeA` zrU4iVQ7<YO|JoPb;dGKLodTPg3UNrCZ zjIto6T;EbJvW9ZgaV@1e%FXSv+q@ho2^f&=aXIW0BTeRFjI9F*g}2Kz$TE%_i{&Dn zH)wQ|8{Fu59}@`ULDEd;PvL3r{?I4tlj5G(J6a^~m^9ATRqh3^tBu&yM(O^Y(D+v7P zBtt8+771=AWQcMEc9!?H4k=a|ny1`e_8s2{(`*(tUWx--6*EaWxNoEv(8MS{E|(yl z3zAM({4~ICbP*^RFpF0+e&qCQq4AW_wCSZkJp7MSro6ol8?ZCO#EYw?3(Ok-1KdjQ zm5@-`0<82UZaNU9d@n_=zxdu_($CEoYpy|5KSkas%~Ev%yOpL96vC3i%axx@hMY@` z*)%J6DGCJJ!mM8{KQT&d5nyJ;{6xRjo?zvmOjs~M`I3ua(y9AW%|A(uGq{H%t#vZR zBvB+mA1kk(dJoc28)PNIp}56uk2d)(*(OjDu-|k&`tz{(>5qQ9-OjRl`i?teuXr^k z*BW2MfiqXk_#)f64ZQoLUfBg}-Coi%r%d!n@ler8Hb5>^7fB5`=FzO)s)BSHNPxKWS&|26pT?|1U*rer7NrZQuIFH%QA1;mCE*~yoOOyy2a z3$s|T7bY>yan7|_koHgNWtpNl&;iH%OH53FjDLZSUM$E2nLv$R#t8$*CE-vT;1SQP z@Tevl+^5vZP?A6&W~>2{t2{bWbRzN!{7S2QJLxi68}onzF-B~-5?DWh4nvYy$&k61_P3n#VNG zP!j~oda%Cbx63p)?`k$o7g8=IeFg^r2;e*5mXOcS;p)*&&I zf{k^jKpr-%N`VmxTTTrYkB}vZu(&fDHe`i}34J|(dRJkXLw06;WVDQ0mG|j$s*eS(KS};<0wDQ3LMR4c?@)>kSuqIE^ z4cp1hBra5w?3=EP4{ea4N=XI~_dvY8`lZa!TslU2*iEDJ?jr|@zWK+V&>W>B+T?AV z{ef*hkkk!aJCL=zVp0a3MXrXP^jbT_){`Rdq7T!#p<0zz85i2D#BzbP?hSNiXo0In zpR!UVT;;1-#;c)M`PN7pT=4(d^af@BylR3Zy}dKb=jMP1-4p(a*EM-+;5qnJV?s5F zqLrXL0U{}qT)JmkEeA+jg~fBX(1*N^>(wn@khI`8{}TUiGQ;GRG3B>^ASuqkT03$x zI)`FlcRB-vLpa63hrE(rDh_UvcK9z<7Y8@1FLL5su&nc6e~zYBGAF%?gYUd_5Q6A< ze(NhGK3Q&s;Tf(EUrG%q4z`!6*S~dE82m3@TYG$9WeOZO`u&|VU;oBD15W&we6*G9 zVdvyH?q5dE8o|AuVjx1ZAL9Bh10Z{kJJZc-OvxQ^%Jl3}L<65#EB_zY<@+Ge(xgn} z-IMJB9f>YpYG99I=t7mzg%+kUBA<)OT(~=kdgyoEy699U4JxZ2@zUJz_B0Okhqs;d zdf;~2<8(kG&+r9TxZB6R$8E(MVd@NxdOX&*;WB*{2e*nHN!oY(U(HQvJ_jW$+GXW)+vE@6QWkfp zFvd{*h9xhC+`2~oCZL#yf+anr@9>vu;5K25bj(&Pjg|Vt{g%*3!zA~OL*Ls$ zDjYZ(e8I>v9j6$mlRQjCmCs$zG`MUQw@4s4bB|lIEQUd9@WkmWrh^&j^^c#fIWnhG zxrOc)KN4$|r|F~!D6R{*ZqOwK?F!Iy%b}LM-Q>|VEz=w7<-*#a@CY?^Fe7*;2LyDq zzQsYvGp09FAXfnVWBu+|Ls8KrHLx*6qyGi)%UB}AFk0gR5H`pd{=HZkn8A{kIe#|5 z>D$*&5z_s__)JMg#i2_kGSL&sBPuG+c{l7s{r$Q;nz{$IPB()u1w-nVtczDFe&p8( zNnG9Yk`y~U`?=loDyH0-)vV6sWQI1Yfr3}nN$=obHWk(&;Cb8-#hQdtpGqLCx<(p8 zHghqJ2hS&f)^auHT@sAvR(j~qbWZM|^ErK_N4P&A6SmEg!m%(0K0y^} zi{vBediMdRQ}ihsHKFc?C%Lz)`#@walhf~hRMi3f!J-dj2s1gGJ!|JTtCu+^h=$8l zo7Kym@o&O%)!O+*!WEDlR3p*BOH#{m*!AG=&maBl#|9J%1Z%!U&N;AAyJuw7+9{@u zBG;%WWFE?(v8rikmdjVAO$M}h@0p^VR~(F?L{#+DZ1YyuTAEf@s@Xv&WQ>Awmx6uP%)VT3Wu46L@ z_-e6vkk+(iH3*Jd`taQ)f1Wg8cxad2aekXTaB^Jc#Z)8Menc^G6p5vxHgg+fIQewH zF|Ovp2tCh$sZrbUJYWUW3EOW?_xgjuVfka@yFuhrc5br+@29RXa+@Dg4AeXHQciq!E_&7so`x%s0S`- z$T70q4|ZUX-DcztZGi}VX1o|jVYOa|6n$(Biij13Etd|_-&};P*e+5-<7P>dbbnxm z2tOg?1n{k3@r3?TN%(4EC$rIWv7niQ`HUDxz^!2X!W8-HH$M4!v+|NF+;TQVC&5GT zw?Ljd29}Nubd1|A*dRI{)H`#~Dc^DZ<<83Vcb|7rF#UjMO&tuV^w!`guJ;Er*m)ge z46I%|Ooi9YXqoHuLZBGN9V$s0bXqfMKQD{G;@6;G z4)seF9#_QGfhEG5WPD}|3rAsiG{((`VK)j6>lAeSOZ9uxz#*$pk;DFF#7vk`2iy=^ zq5 z*$JFV-r=xiQz~6?k9~uSE}}3zfzwIX1TKSMVjtuVfRN=pR0R2Akv}Z3n+H6yIdqNW zB&nT`6X96iqcOY{Bi|6qK99jbz5ROGeAALUHgIy_%~mr7dTm1@U%2G%pcj7j&G-en zFK_(OnFR|WKphgp!-!yI$ek&N$!c^g58u^aO>Cc&%3ArrjreC zF=>^t(~nAEtG!!ZF&o|73sW0>3#oBK1r913WY`G=qjt*<8OK5=ua+3hhVZWN_axDQ z%|?-t*~p|A$ZSueqV@%BBPH}I$jwQgiT`HG2b>y(15TUN15SNRC;1d=yelITq zU|%yOG-2Kq{HUFaoJRgpl@`ixH+mkG--P|XM3)%Os-Y|holL7-B*=CfaKfAj zJSw7deJfpuj}ka{Knf#|&YF4&c8_ciq_Lo87(F9?+88rn@K`yY4>}&+nQuT%(+^^D zNVzjPXEZoRCNAyr1aWq$x}&n$;{Q;Xz`tYALe0O})9mQY4&_`pAm4#iFB-q3|t!y5Q| zBbLras?7%ewy=d*qR7b9ws_59jK|dULbPp<-L>}Z_a+go)BUtjL zRdz2Jbix*>_p9}d4@G9^p}i&)zv-Y4Oeyi)AE>`{M~cBsOnX78VEs8Q>_j2ycsnd> zhXg`Jd2|Kn(ITFb!Vzm)Wd;Z?ZE(qT!C&IC-&v9wni&cuFrQ1RLJn|o>&o&4>v-VH z1aTI@?FGTt_iOh2&eRDq<7-s$JA}9p)si+@_0;%?n(@e4JPQ_OJpHQa$qND zyOC|oq8Kej(jTjA7_<*`vK6F3);?RS{K)-~3e6%0@-)aIyqSA+8mz8L!g29Ewx}8_ ztc(e~V6a$y$ru{50`J6E#bI%O9G<3@C(I=s4vdkHj5Z-3QA|HY?jr%kMNX44o-wF4 z;M_GJ%cUPFZq3vx+rkE9E%RXp+FGNBvt&n8z|&sLD+zbltUXrvo4}Lq=$dB0U++()OUb6u z)PWq=#*!K%yH!jvV8{xosH2cg`mgKqQ&1bztlk@zDQ;B4_JC)ptWO;4(c{)jVx(<8 zr$ja6*342U)>*_W3_S`h*NMu=2H)=Rra2Y-t8^?g;IvBqi9D8B#LMF5(zi&{oc~+9 zQ54f~Yq}}5k{Jf0%8=EhFuZ|(N|XaB-uUy_v=n(Z@SF|EK2gZ?6_}cRM`RfB!gwS`Jb2mL6*NT$vD7eFk&kr zm0}=Qbq)GTyA(KMVANSt3=AjTHrX1cNY zY57}iou=3j6W(us`EJy2fBx%*|0SjwCqg+cwuu=8Z0F{>w9dlv^*lNu=MQC|i2X?yF`KKjarb*xqTOaV$>~t*3jh!B+ zLP8t8j>=ZaRtajU6ms;0IOEB`X3=;*4)WP)bf4_jymc5ww`t=2-~Yci-ZFR^fw42b zPY$#5G#q%Ifg+|6p2k^<`IsUNR8(JhiK1g>J5ZP04L$*jwt32u;n;)Q30eqz;Z<|i!feG7E6IU;=Q*sWcA&% z0`7#=5U)y8CVPj)hZ%1AWr|ul!(YeWD&7~+ChveGyIz9(>ssYmue&^!_TB0+vKSXu ze-pb7SmA=o*%AFtfWc1*%J^3j$#o_NjVAbA6ay4mJE*7((XE-QSm-^rm87W@fGn<1n&e=WC7R!-$BmI#0*=sQ=$1}30Z{#vgdNuirw z1EromDAspr%Q3Wgd`7&uWu&Yv&;Y**wGD&q&oB4hN2IyhXomHP9@%e8M`FujaoEv@{sb1!@5&@J+`$(vL; zus>87eq%;-aG9TR?|0vLe;xIwN8f8wuTrFkp7b~~_l$bvx3#G83*(&IAOHRfo7Cq> z1OJTr@n4a~sS-Nu9@015@1&c?V4kb^oS8^VHS7w3u?#H<*afrNk_GokgLM3rR?uWw zWHFy=+w61Eyv1yMu=WYoEQj@K%75xZsO6j#@8_Ko!bue9X?UmR%SN zz|6yEl)w!oDqTGH*=ELZa}3P|bW^yMAZ6MnUj{YP2Cnqw(pL8|jr3SX;qg6|;pB1q z@%m&QC*R`T*O^kmJUR6^Fec2Q;_Mj z(&MV<6Y$v$hj$4H~g4$thMT(Xni4_@!~ znHgd%i^koGbl(~#Rh7%VMQ%zr4CxOpbB_OB#W&)dvAx|a3Tx0C8x*^x^`2N7$oDnG zUJ3_)u#04}i@iARdqAp;VlO)=W*bEcup9^Az zDreoTU#ScTI~w)RHKh23F-i?ac-&7h)fB1F*P&ouC7Pk?Nm@_fD$~)(NtUm~Evg~H zhha{!n^-mgg?1(3G42nBq^6)QMMrn?v1TO=WbCq4ND6R?q)88{G_CyM_U%AH)9<#? z6RQbwRLE7fOmG|azAQtQ*fGu)b+Mr#zt2HGxy>gP)SZyCI_rbaUM`+qu^`XXzcTr$NAJ(5XFaV?ZA{;f{0X zaiJz#)gsv&HndT&Votj2pwmLz0+1ENp6y@Esg0;}rh&`NZB%gCt{?OwZ2m}1%K0!vhW_6Cr@bPm=?7S3t znxrW#pOeX33Om%;AX;#rB$CrK%B0wQ$J&PQu&IGEZ-M^!uJ7lJ!E<$c(0hTbb0%d* zen1|@fGe9tMRiE~Axu#{=Zs61C_1eE@Z5vaw<@G3Tqi3P zYOt31&MdqOFLt`%C)>V_#xH?Y58t>en;w~e_|H|W_i{|`f16gSI;_uPri2!g2Ab8o zB9aA3;jMHNC>Uge%2teD6{PH9oUTW5L!IeT#2IiZ7vBaNGYpR;NlPZ#6SfAdjSamv z1?s4su|hB9A)J!! z^Hnq(5IogQwJ|POS(AzTj|$)RG&m3HFg_$Qu!1EMr1P;)h!2QEN+de z6>6Fnle<7vmd}Z2()^JQv00rpwV0#j*3)(3?D?&(J5=R!6GZV$bZ`lMT(HffM1fYO z+2hRErAI9oQ>*Y}LH}cIVje%mnZJuR#fq87Uo5Kd-zq2`D#}~yhq?YIfVa9?olR+# zSNv|xNSC9;R4wmnc)hY&T}U^Gi-fIo&uo2LjCoguWQgo()Sth{KZ+~sxN~{+yU{sv zgI!sd6nLJjbta`ouWByEKs8S$6;;FkWL9xNr{{itv$|eAlLdZh}lr=_RZ3i;_ z@AS{&_JnVSvgr%Kec=~}NEu^Qsj8*3+hn~nyQbm#DZ_7T#Aoyg(gNWbytiLo3re6} z^bIH-_}g{)At036J)_TUpGzX|JWw?j$!-U?hhgV=J}b!bdB;K69e+iP{xqvQ z(48*l+}C)u0Us-UumcmdciOE{R+3td#v{qiCq9&~fFrgqE9WJf0F(TTjl z*HQ^+F0S;@`lc#0siY*lU`QZWQy@x{F*75>75PN+E0Xh4Shhk6%-9PP; zutkDU110kekgh;FT;u_%2g&;}3L98zP%JT0=G<6Izhf-z*cmD9z_r@<3`VNN=cJ6Z zJMg>}Yh-rrQ%nyttrLv~1z7UXEL5q7oNJuk@ci=mJ$EFfX;R;$#v*W`2R=)54n`9;UI zIqZ=4@Ydj+Py_O0f7NUx+Z`BrCyXpqHN{j=q@0S13p~atcPpnh%mwZew=7YU`r2DL z)PY|#y>WX%!gtTWx_vu0)_L31lBpeDH`M1OdtA2An0&g_IRi3!Ye0jimA(t52U%0G zULgUBw$_6D%aw>l(@+cf0=LVpSk}YU-;e3zizeH~AG5}GnXzeOF;!38!wNyv>#zUu zyQcNm4m-!0p>)vGziXA7e9PzVntM`QAZP-e72AyR6OZ?)CoICmQ+u$_MOK)Y__ZdF zO{S3?GXax_5E9Gd;4Vv#7z|EK(MY$EYu(ea2z9S61WUxcd?B{&dT`rYv|*X74$0Vy78qyv^Jcd5KW- ziD$P=CtLT*5SjOy5ZpTlAtp@bg#1mqk9;mEb43;5ak4dI0rN26?D^9?nEPJ)@Up1{ zAe*cj2S$z=!9aa>M7rRg5D7$=q@Pz;j(kZf@ii%a_(z(pim#)c|@lxci{9~`9@AWbF zj1>16(aQgcXc2hVMm_^8~};GL8lAug}fA7<>az|?89c- z+|Oq`PYW-be`80>V{p~pp)))|C;TT)jj1@v0*Gvb^ z7OkTV2MAC|00w++up| z_jD^b221gWBfa%x$qVDEW*Hg0H58Lbk$5U<6Ll}LITA&%E)pbBJIAS~`zF`Zr#;4J z>U#_!i>JYnw~Wi@uPj@9ajpRsZuja_$nF;g6=#f4QAaUQC0k2HwMdr)_kxo>#6yN! z>kDtba`~0az#B8_W%bGeZkxRC)e=#&x<*n%-{u|mxFXI7xF$!1)(jfY=W{-nb;!|5 zZg7bW8oD`wbHlBA>Y%Wkish{2?Q*>jTZh;fJ8o7V=UkfwH45rfRU*{UE(vY~jv^}n zXzAQ(&cS~A`LTnpdS2vjOamqkyN6>YID!%+iv=F>T_yu<&KYweO4t4qUN0C<8 zes<%;cBWw0``BBvf(P}(syW{=m0Wy5wr|V`H0tPlWt+U;qm!>G4DaXa=;(R({ag8v zXcDGDM)z*nDaB6Dlj4s(x@M0TnAl>;rUV>Mf!ciLjh=5B{MSFW|M3Pn?o2*0dPgr( z%z28OrJ@pf+rm!6hUp`}?D+*AX_95IqlLD6Xp5|dPKwYf(g*P)fsueFK zg|B!ClGAo6`rQ+G*zSa~MlzJ1Hr&H;W!mH%dU?bNQUE$jnuH;?zml+8A8eU7CFa=$ zaID8=*e9^^=TnAc=n|`U+;JDpAOEH3hwcV=y?co)BuVU+M~?fKk?lrrNfyOGEkQaJ zmExK$+9lh>pp!F{M!A81X6|A^G^p?mIw3zzzZ+J&9fsO9tmI83hr&0DcW}qS4vh`! z&jPCT*PUK#8Zg}hbXfnz%&y*T-kK0hS%nIsot~FmHOR5BQ+`c#ge1x<{5mBU=6BGQ z-kPJbiyX|cwa7GC|70z5a7`_l4ZinTU9hum`O+W!;(0GMOZM#}nBN>HI0hAFmEK5k zyJ5;!FAds?7D>K%91OsSQDI>VM%-opkF0<(LHp+4QcW$#3*y?Eu``&Ngg#h{q+6CA zI_QLyQovIP9Le3XLhcZS;hIU8TsOP#1Ur*QUzk5WuwntHVQXW5EWtPYqwt4^P3g5A zc4{$0m|aCDa4KDq)Tc?=K^Ft5wN|N<*MSv@3BEq7js~x{i|!Df6xYp&XEH;9AsfjB z>t<|{Kg?+3=Fg)@k5$rhvaKpMrYqewC`u7i$~5T(JOXb(LA zeL5(tYtfe9;C9PVk^?hD$H!E|A8yH09r^xQ_pB_xo84>aup+-<-h~II8^GZtQtc)= z4qVQ4zzDTv6jMyW$~J1H>w{UfoE#|2%2%xAc1f?xkaP7iD0yg=$H=u1pd+FW@~T2s zDvBmAa%qznatCC^ob9|V?)3p%B>CL6+%2k8id!@5U@CerD+X#tHgnq{0$1YG;i|KF zDLq?2j|8{TXUnl_QA<{Urv4!Q%Jyg1VK{Co6){tXar%{1MH~Ho5@(I5hWtr?ghJ=J z^dbm?p!3m2H_h1?aBF6%ED;FY$Hxwi0W%hjxwVbVW9IfZ4}SGcgI)RknY3?{1MFgH z4vd^@Mpo!!iUD8v1Qm7W%`4({qB7D-x_C8Y3%xU9x%?E+=T%QF6V>w2RW0D=aS=UD z>MmZQ$AHrfw@hZhX*c)2*MQS%(gFW=%8$aCVtH!d8TD0h78m;}1?}j5X|zE98FBZB zmj+q=`J5VlmpFlQmNdBCm~$(v0Tk7)ic9@cnKVhMM+ptBeD8%HBZtP2e=r8T8JQ*v z4_L#m>eDwl&pR=yzH~I76YZ4_)ff#jsQRkrq(a6?gKUkzCQp(iYLpgo5@+YC7WPT1 zr=E0O9lA%nFh}(XU|@NK=48O3z*J?nTNd}aaM0-i*-KXk?h(fX*U4<1*=aivupZ~9 z-eBJ7jFRp>ab<4ApsB2k8TFR#@bbA@WeKb#nv{9;Suf2cVJjbKM+N!f4%*gF_o;s( zT(tb@KdUF(UT5|69XD(CJ1GO%K8E?}5p!%g(Y`SA6L1`ipk*nd7;tsAP*Lkdkjz9! z2UpC-bhcWbORi07OqX2gn#S2aFIKTf+~koYI-_n@m&me!*r$rV5_nNqyH{)8-sO} zu)o$KX_FuJDGt_DsJdXs<@z{ADOf(#;lJ7LKUoo?;GB zWIu!nWjj=>$+CH;Kx_A`utruY+v1-m(SeMEPF5q!bZG^uja25LIFD`+wga0^D)20= zoqCR=my&N*>+dfRbxuC-*+FOWGL`$qt6ynXEfL(Mi|7^e%D`~8^3N#B#U1|r^aXDB zEc_N!8;K512a=}>*%s>--c~`$7+6mnS%WloQ2Nq$e#ciBp!CyI6V8w{c0P>b8dL&B zvLk$$0*ZmDHXEwipcu9!9AnX%W(f5o4Q`Gq-%m$lwX2?o2c}t^Uoip*Y25>|JuXcU zw#fy0xgpYA;Oa?}^hxZOzil-Pze~p}E2NuV5<;kZ;V7W{& zEfi@+ez{EEB0(&#m5*D_dYX(XT2nUZ3i-tC%-jqSDvV%i_exh->t%qtQ3puywn(CZ zWDfJ7AkiOH_JF1iC?{h$mt2uiqfL(NLRejy$Z4RT9DXmtG<_UKpbjV?=7%OEKLP=|rQ0wM5tH`|cW)uh!A1 zS=1#3DqODS0IaHW>E+Us1lcH&!}}t^9oGzy5*Ug#Vc02`t|lzaj>nC&xr|cR^l%NKAB4j`~kZZJ>iw!TIGr0 zr9hjDSsl0=+^oDLd^Bq(e8({OErg_H7RDB>3|Ri4=ogQk_bj~oKU(+Z(f`=-OQc!p zpjYyeeRl+b`fuQ0Yt$YI-r6G0?I`RhI>@JL=@ znK5BTM&`f_JGx)#e{1{paIZV4T=)SXp3DJ8~Djan^2Sy=X1u5*mSX zeVcqmRBZ76Z(jmVj#z0mbW-+{OemJQr&=x6!y=R?+VILj-Nk8zYKJRH3&Fej3S$RQ zKVyg=s|UXMN0SNpQ^PN2ky|5ikvgv@MOK>aj2}_l1B&!fG4<+d`gX_~aU7Q}FOH6# zm${%bj7JWRjL0aJ)V!5K^*gL%_b_L8Dgl*kx%ymQ8jBZt+y6Zg1!%<}|o83sJ)X-z`h5=MqVr$998g3=Ti2#^2y5fNu|sd6LvI zAX9^HQpD{KPGmsw25!R9u=l`tu8nS!-~%U#UI(}1*!`f#mf>M|nWwCBJ3a#a{fdId*Td2Q4F>teN1kcWjr7P7`<7*X)UE;fzKL`VV|} zOA9+~n#S$^fyK6<%}M$t$+NJxwia$jKY=@%p6NWi(tJKjEI z*^(dxL{7Y~uQW4Gdnpds$a8^wTyh!eww6!J_t~Xr^1h;Y#8!D9QQs%E9=!|tNpti& z9dl2;`VL5wyX7&Mn>?ecm6Sqg8Ra=aR)n-EF%|%BI6-c!l9?B^Z5~@4!P06U$OyoK zjkod{Ei%g$>IH#}mGvyeJu@udbDR9Ga#z4Mx%p9J^j&3ID00`QvUipBl6=J`&0jVK zT$q*To&j}$=zx_d8s|H7Yc?LF*f{B9x^3UrayMDl)L1D~X(jc_M(rIi&h1dD#Fu$s z8B({aLqE2W|0df$%#ojU;HS(SAS^oZ_odv_#tT&R#W4|3r>&Og40fM z&}w-J7aXkSTsalGRp`Z0E!u;yqZy#@`ZR`h)0;h$M7Zcwgl*C_cveb)+FsU9=PN28 z)LaXJ`?L3wH&iB=9sAbXm&jTt-bU>;1I~7e z10%VGim4W7&nks_PQFUKQ`9}}l&VqFs9hfgJMQdRNYe=IIdI3(`56#uMtAm{tbyk9 zDE!aSc5}?rKH9vE^SUr2c_L{CZCOq8f_O4kih}ltc8DrKV{f10PVfoImcTkX?t6UU z>yV^SFD8!de;HwBZAZ}v4?FPGzdj1`{F=#lEM52AACf93HXawvjK?X8`-B43u9#9t z`ZPy>%r%K0aCen$Pzctp81U%uIOS37yL8I3Db;kYsE_oD%3p5L76pHcd9U13e>6r7C)+VH<|SoJND^deD01eV%*iH1Em;tFqusC0FN}fNfiRq#Dh>Gd8Brn|2 z3;!3Ng-xI9u&?A74?6gTps$tlZYIip_=~HS3P@I@gW{)L^y!OelkHP{IcQr4F!)j#@v`5rUm+BTXNbcN2pMVOh zUGwYJ=>hfX8sBQ5OvA=2kjsP4r&&xoL?bsXHYJCf?ERwe3Xv=@XX&qklAovAMJO=zkV~wY57bfNg9sR>PtxG8msyPd(m_9P7ZjL_t&dxVVUaxSgdb9kSAAY^$^*Z`<`UKe& zR4-|gb*hoeH&%K?{s{7h@t#*hnxk=tvkR282Gx2^C37?6NW{&MYx57&h4OZF8P(?2 zu0HI$O@7$7hAE4*ucI|KTVi2ajECk39_P{4M-tapP00D_jnir5fH34ZZ$UG(SPw_e zF^a3B$U!OwMXU3}u&}qqplZTn`!h-obSPVtdS+iBe`$VMB(_=eMFE{Ye6Je2v3abx z%+!4Zt=1Vcwzw|S@`X$bkUO7z-Jnac+ASkA#qSgpIMguhtl@!5QG<3!mX4Y(SP@#o zEO&z<(On)F80=I*G5{1Vz$!n%NV6lDMnu??lR~i-@Wzoe%dm+tzuYs5$_UncLEjv3}u_Ohf13LAJ-?!gNvp^o(m_~615!CiEg zuAW{wpT`HFj~3F3JOcD^V4@s5@7p9xgAN5|L5bB_vMCUXXb$@(!d4NfC5nP`6!v z#0l~W8%nb-DF)T6-B!EhcpHkHcZ%O9;2>fpG7^^D^u#(+Xmg-?6se&#%oeYN z*;&#gKcvVn28xP1Zo)DT6A3FJxcaAvul7cmV6`Dzok8}G)WhJs!2{B02A;zd2h`Km zRLs?vyA>OiJ#^t~1$25~Aw-C?8Ds~p)U?TxJ<25ovO$+5pVOkI**iry$S#kM0;^_J zYRWuPqx70YaUHO>rhMzNqTscw?7-}eO!~}Sbc1YDP!41!2c-SvGP_1xT-B;I^kxVB}B-EOL zri9`^c(wpZ^|6K}kvRwH;XX(j8_hwcu1#`&K{d@^pVy{|l~&5Ikhk5t5L}!*YK7`E z@V`?g@+d5pI=`SBylgD(K<_4<-YvbWKC4W%dri9CAR=V`ZMfGuL`J9qo_zPM zx5|HMf=#~LyjRE-Vau8mJ4%CQAnT^ME($adV>)S6ZNPPgZw(n#@Am)vE9l#t7a2NX z_C_3q6gPTA{1F;U&@0?f^s8pt1z9JZ{Q7`9)>++r$=K|T{3E;RGOhw-z(DU9J!lj> zM5(abs!qsj@mNTKkSt*uRkcCjP_dr`*F%k8ao{@bkPEg*t@1de!-X|R!Q;zpH+*k= zg?3S|xyG*30-*wukfJJLuYpA{3pD49i)J$Eo0}_#5bT0e@Vl6sJi7t(i4#8dpJuS7 z*8BbX|CaE|HnLWX4a|m~JuIKtqdqz3J@eTz`-mKMU>hVH_3qZko&V0d<(AFhFUWSy zN>96y)yUY_jw7^ocXqD7@p_#cy9CfN${Ax_gRm(Phy)`f zguTU>?zZmN{BqGRcUTgQJdF`2jt^R)6G;{A4a|gDzxmR3C10v6hW+>{V&KHiVYBFd z&pVzE=M>1&13r`8nKMy~iC}b$X|9DoWE~yHuh0FHrL4hog$^qk496nC`Gm?BbS!cd z71R6O`8}dfK$0O-cVq5E!Qsh7*$j&(4@}Tc>#+D;;vglvnTq-EtL*<=Cp+migx?%+2RhkuP=(QwE0okO3S+E^z@J(=iQ@;kVP*I8geQ7dK-|R zVZSsgp<|*mmvliO7fFCo(rF2kWVkg{diGN{6R6(j&;AFA zpF(z;&BjKGOQpzK?B>`RoJF65zEW(M|9ie0wr?adnG3LoekZ*%xSXl<5d6Ui{e_3! zFX%HJ_MD?%TDS5eguI-0ukgtSZ~e+fXk1`Q{L|%G3clWThhpg~OM*>g?!BUZ_gm`j zX_(#{!J`x<)<`=C(TT0txAx;e}KeDDE6Z&QdW* zMU|-oTAz?~dNsF4yf?5p8voU6j?gWT!>I~NWUA@Sepq(bCb>hOQFXD&5j#WE<&{8m z*Bt#(U_NlxVG+U-|2%)7FJkxUy6L@;Tt=U@9Edr(=&q0fU}M+-Pw1hkcGJwE5Giv?)OcUCxq%XI5IXa+3c^`4li5W)Y0#; z)mF86{iBN|Cxg>2MOIiAaPxug)As3mXYl8MNor7?9D!N5Yx9R(8dOknH!s^a5fXvM zf)^AaNM&(bgd68d!*0dqd9?L&9e&?<=6?7%_I@pBftvj5hsw7t7bvGiij^fNOV>v9 zenlWVv)-c~7{2*j>1GvwQujy^2QMsBuJ-=CEyojp3(H3xdMpSRlL|NeXNZ?+fohr4 z^e#zp;!>3&vn3~!;-EE3kGx|o%I!}VXcfD~t5?)Tzq5hbq)P-h;bF)-pcn}P)+v5hRm;Supbn&vISd6P zs7~Y1mBlo{0u0HQo`Yk-VHjNyLa2tXwJnM_XHztp5nn5XN+JD^xIs?8pX`kbjn9Hsg7 zz>^*7b=sY-ovH#EkB_g@LLL!XvOdvZ)lrHcl;h^;vgn&Wcjpd3ig78(c->ZE5STyj zktc>CCw%|&3-*u|l3qiRRFZqKFTXVs-KV8KCxXi*OJ3$Nh44VVTk%mKb{q7n?zzE( z2XqNu7y#HdFOck6Jy1V>O>|b0qOL2E7a+``7h)NNC=Dvc3?h!x`H2s#(Lpz1r?jaY1 z8?Z$i*k57ddqMXPwI?ALnZcC73tS0%)g?0rkl?kKt&$bUklUfn?T`o;%t2T-dle-! zp%R9CCgUqXN(J_7aML!AcF;oI1-&eMgKF`tb21=r#UYRsc4-iHAanqFE-Ml^VM1|iYX6T1{+SjCM~p?q`AS= zR7^*5?~MUHkb9|@AbWd8XhJATuMN2vyNTLio(A0U!K(TAXosW|lr^q@C+T~59*-)u zc(q>D%=NN;;d};@PG3?UgYLCdwwVKkbkNATCW?&$BLsa(;p-yO+>BiT`+!Rv(bB0# z?+7^Su|^lo2o85q)+WKZItlXv#Jr=eeE(~6^JEgN{1mimIq#w~^X88}_@)VB#a@qo zM(Ui{;p#BMeiOw(BSHfegB&uul?Ae!;=`WTqEAbgu!AmF!wRMkx-1qSWCmSMh_V%f zE^7laxIvfoGmeDD`wzOrarN$3W(>NVj?59&tMJaMsk`KZE+=7}1{!`Ux|l-vxhy)) zqg07MKZWjv0(m0AQ5vKjP=9V1MY7Ws->o?MARMm=zck|#jJjC-*uO(P=(5`NuqfRx zU7+yy1VY{lxJN!M2=Y{hc$0X#30b;?6;)*G6!MYT1m;s5P?BU*F+2SRqDmCTf7>Kz zuj*y@l?|E-ZI3uF9P*!0o4sm$){+Jdw1`C2_}qSJ9}w1oPZUuY0ngEkAUcPOQ8n2b zSfUusqKy2Xv=fv_+gUuhA9UptxNLe! zR2I-tW6Tl@H3r;I(eczO_SnqhTwy?#LH|3I-2)9-f4|E1it<4Y6PX$YS@iXq_Z-o@U>|5b3nXK^*2lOl zwamR+NpV-W)k~JZ=XTY*okb>Az2>lQQ84eQRp=Rik8L03GX`+bB|+a)=dH*1>aHyZ zElYKrHnm};yv$gMkw+Z_0j`_D_3DMF@w7}e@D&3~#BS*&=5ffqkWn{f_5+G-nEhe_ zvELhyGLkUNdjIAV^1L_VX3?2+kNT$g?pzETV&h>~L>-iNW@u`G=@VbQgD<;ei3|j? z_lP<`=5){{ojI%F{z^Zz&MfxF}%j; zobS|wM~Z~7AmHTI@4f>xFp8M_%Hy(4K5^3PicVc8Bw~R)TvEhj0a5G$dQ0G4*E%_z z@k{VboAm&$?S#IEE^v_-N1Y<&k=LSe{5y0UDVN|oeEehM_=~ig*rjB%u8yvasL-DE z9Fp}xokyk)Xez`-@aiB*)E0Q{^Q@yc`Wy$M!xA-K?}xK_p=-pSXu2Y7fUo@^`7}ta zpBWvWuU0<(v*oh&g7jHhS>*P*b?T6N1Z)rlE{(P8#cJdd&NuJ_v?y0hvtdF<|7dN! zLkKi%9n&87h5WEFxixJ0I_wV19+qfla)L>Ksd(xo6dB0Di<}_qmhUHsE9N!@)ZTm1C^48BB zRFI$Gyj!AQ`TGo7(g+J#Tb%bV1D(O)dA4;Fw}v7KR1D@X``O%Ac+77li!RC_viGE5 z&~Y|lWGvYRZm2!*Y?xx_r%=O?w98C zo0&C{9YBMJ4ZC_x9odTt$k->iC)hwAfs$Yt-mjXq%v}gf zMxQx50%`O&6L5!hAbl)}{jtUbn&{&{D<|8WIMrBX29QFEgI53DR7|Dp441>yOAbWb z5!ca}a%hgmqeaYPpBlPf)*OAs`#!LkB|)D`29W11@+}HX^eXnsq*p|wg|=$pItX~< z6CVXuEI>(aq}Q`AGK`aD7kmhqcUy0@j*pQ#h@1CZpJvG&@ie7~^RB+YwK&|N{emGP zcHV%;6(}9cruzsiQKpWE9367Hrb$h-c&#d+@Mlg1Jz#U?pZdWexe>#U@^Z<)->Sm=VCIGkdIF8I{zmlZ%Uf?{77BH2cc1AXDrDJQ3a(y9m(qg}EA zb~9I~E|2U~w{y#;?DGfmR=jnGs4Nn1!%K-#$?|l0ap0f}UO}?ydT6t&aLaJ5^tdB7 zzq%{zqd@aB{qEiL4n;fjgHEB&f2ZSSDSp=B!v+x*i&*nC~_h*?=c9tCu&QT$>LnkV-d~Khk}ex+|xCBrg`Ow zGC~D@W&*!NQa1S&0Z>CxM}6xLnsmK7!7f zg)8UZgxaVT5N0_WHi4OGM34Zy3Ip6Y?-qpD$-POHX?~`4DR1b9`$>kdG@}zcHI-&g z&0dOwo~&Fd#>kJxBdcb97{7vD0e(xWp*rWT`!-mjAS``Q-QbzPC6k3!K?Ojdw%?~x zdz`HBUx>U!xHIRC+jAs2%7z_zGgnL&UV}`gQ>HX3-VeJPP@pE2i(AjkSU8K+;s^(lUF(ss56m>nJCd)WaocEX7HFej;HZ@ywejEh`X zO0q_Z+&S@{2t+T2wg2v;IA|N+L&Ypptz}X90#uOU)u1|?SvI9Hx{O^GdM3It8jJ3( zMc;lc?q>bN` z+x?1t?au~m2Xq)0CiBT1_mOoEr?&drUms6BixVTtN|w8>%b|1W zws%p)4v!(WcD7?|hm~m*0vanrIOO#0ukR~)-q{1v8(S` z=yduaoi4|Y-d%J)X(#RML%LVA4R|B@W9kbGc1%~z?^WLmNoDKR>jF?iXP>&8-ZMW> ziF8QNzJ!FZsczZIM4>@7%jvRmJO*@G?uV?5JigPHl&<@q_lQcfCQf?Tw;r0`SEvq( z7dEI;)F|zz*W^k&9OQFa|IlroXzN2bMA9v>cYMR9xv~)xHjQP^n*?EE(&PoSzu2^&nYHPAN60c^J76cSQOGpg z>8z)?bre}c#e6Kf7gDe0ljR>OP^uBhH4J5OEy^7t{;+F@vQaZJzKL&p%H#zcr!egQWO52~!OJ8T#B2JgEB$f_d)65TnO%>W~X|Cx0SCby(p0os&8>k3z{=RE33d(QES? zJ-0v*Ejc{LpTDh2ffo2Opn*u252~S+z8wgLilMQ28|ex`oDO&lc);mABMy;!D#b4z z7;B8UO%!8w6fz0rg3N5JG+kcen#%T&avfCoBIC`YA0sE^f!7kibIPL<!P<%t5ykc(?WX(DV#)7qh z?0nuj7Dm`aPg?b`A;PkJPRJ(Oi6d}UitXAV@mHoeIIBP6Go}E9OpYiHiJPPMM-;NX z;iXKys#?6=v(>YU&R1OZJ_2T>NPQQ2h>-QFFXB?z6|Z*n9lC&SlXQeN(AD&6&kWGE zE)Q}9>I5TfBuipvbm4HNx0mU(Pi#tR&t8)6#5U!)nN6vnxC0a^rDE>CmgQX+Uc;oz zKPM2G_q;rN(B+X&h1(@hsLZ_;)e9t3JyFeIcGhXzqxcS(>06R^c75o+z>Q2iu}olbXzU3A+iZ&j@K*K4-Cycd=vbe_?rwg*D&f(T2L*a0De zY|iAgKhQ0-2|s<`_d8DFoESeDX82i4aY+@nRo;^dhg5ueu*#D|9P@MSI%3oMIDo@wsI`ML2#d)cfR0cNAf45#$9A2XMv+)S| z&c0+iYo{CMZwRFl!T{_Bf>BP&MQ zLpgDI6ErOkb8a_L90VEHQ8Bo8&IrwQ)0U2qO&a1$W9x=wUQV{DbT)vgjaZEn!ou{{3}qvpX_xk0_)yd zfo<8SuQyQ`fawwj13S#GjYB*G22ZaSlm4oBOKzD)enIv(Rw(Q5ywpkJxtqaHehycq z#`C%U`0qrkQVYeCVd1e65#zh9;Dg7x64S!+deerF$;uZd*1yYa$=OVC8z_=O#q>cd z*%v{6Tyn5usA@hCOy#>-OocV z7g#JYz>*dyy)Kru(Ri3H=eLW_M}F*c+8y8j-=jYGD_&MUU=ET>Q5Jpnm8;$-=JA^~Hh6ZZIE7*fb7`XxetAW8n-|MOY z;Muog3jg*26X93ZzBBb)>`<)0Ei^hhbif z9cFvfA*{n;gsdYqeRAJ6%a*jK4Tlp4d#yB{wMmXJhe19T*?f=-sFKEls#JDSfqlIuB##oeCO!DV1fos6?TIlV=_ti?ZmymGwp(JL?*pLZcHJ|&0=Z~D6Wqpy;Mvhxenh# zl#y;xuJTBo%2(6N$W91Pw1uo8d=6-D9w)$7fO6xDq`7ij+c6H30|vhal6qL=h^^(z zqL;x8ArC2Ul>Ho3f9To)z3Od}58a#9y91YjAR})SCcda>7V%fTk>LsBAeGuQ4WGt# zD3Kle?6keU7!>K7(=ok834xJFu2JH*;56vf z9w#^E>NRC=RJc8Qz1mSeu9=}2lYKH9!7&Gd>=UKF^UL2@>N^T?cAU64!ipNCk&QWn zyYG(d3QI!yR?xbEz-+66M{XCRDAte*S~83!qn%5q(}NB1Uo?At0eu1ap81~qa?fmf zjTn_wm%g$D0)h1)4GH9${cP<_DEJP|rn4ZZoNchpS-NH*2Z?tx^htu7ZV=4E#0Gd= z4Jyo!p^?r6ovnKMKo|%f3&_lm35O%ClLF4!g4`jmH?IvGvI^vpyRr zE`=h=AUp}PSgPWy<`>H#74VQLkO6vv04#A*Oah=1rPV0DEuWFldxXuk(9y?!zaX14 z>Bf;QHC`rUw9ILGm!vo`GJv#W*t(QSaX|d8r(#Y7qa(9L(SHmG>^VA}M5JrF~4A-`)i^;*>e~x&&nOnxntG)*F@P z_c4u3WL4rL^l8s~>XVAzs9G@yQ`GYQ*;cG?tplW}| z=kA8`LO78;)!bh3{+2^uV>;~?j z_`V9`Uno8!K^6y;IKtRkj(*m;S^0?KK$3MI6*CZ(B3h?Cq{ybXM|aS%QPnhd zvF+9rYf%gEI9V3mpxO*A_G>`C*BGy_hcfLB8cU)Vj)1 zVCM8mW))YiDyCw({XirPBTt(hi!*kEj&*aB^td?7QWni=J=SV7qaCerB1|i<;q%7gTOW>V+IbRD~a08N#{R83=?V} zfY=oq#@|awNdXnPH{k3~Os2iB_^qx(s{dw&QDuC<0N5*tT86Wtu$P4J%kH zD2FjfuRZyN-+7ktvJcWM93NW`p8-(orf7r}jjGM~{huil(jI#}J_)o%#g;3%j~Q#ZgbO!1#(9?v@@6mL76&T|F7LaCTFCmt1yma zJF(k)$jtF5rMMD`6hgt|)V8S|$`&Pmp6dm{fJPvJ*-2Ls9?Ru`00y`>>H-=S`Ci4a z!N$Y^avbV4iA*9>9F7FfdQGRUY6g$JCwLR~qMV$zUd-=@YKLXwB`N`Ru{O=6os8fU zDB1L$w%jBLIg{KzSz2l7wLBe|apH7EZa`gQ^jhe?H_d@ta;8aftMI{)lVeh0D z?hb5G9+<;#^f@2gOg@*DD3ZNOd~*qZ6)Fcn~@ukQEJQ`mN77nerBggp6*+VFGy^J0BMMsh8kmdoPAu|JkR?>@|duna{n zP;zJK(m>9LzppOwg^oH&Hk~2AukN7lDPW%jLKSPoKvk9=oIR^YRW4a0&K9ln`kzvOq8YCRS0cOWOP#!;YEk_Ur<@|qty!!hlABJ5O@J({+g(1G{GV^#Y zQ``lLoI~XkW9k6gFjLu+iqqOHuBV~j5)A2p#~LxV5v8)9e_#S@n0YQeAYKi2>ZjC&EF><3qzO_dqJ&c2>X=c>M3#p5EfRdJQJBb z4H=%X<0w}8QN$^EF|8+!+Ff+EvO;`_Sth#>)I+v-p`}ZwySe+&e$qq_x^z!FrP@V5 z2-zFj1OD~}$Z%ejm5MF~?hpPj_=u)CI?WFYKJ_{bRo6=vy>y!X%zfDuy=K$Q`vk8f zG8x?ZX;{*U_vv67Qmo*MytF2||?xW=21Z`L{_2W|>g+{s-;lWS0}~y$+k< z$0Qb19LP+LO92Xeq}# zk+u`%9IJJQJoDZ39`#Vhb=CXsJSaHFK5)G0i28_Tz`bHRv@o@^2gMa);j?b#^S3t` zth|lj(}I?~$%i%#O;wqobL?AhUm|Or7&@TVH7r}To#H_JZwnQJ;<3ol*-D_4Myb~< zowZujK?6r;_N>j!IfBUQRF_3|(HFHe5E7-b4o_?CLQAp1 z9n&iVw{v#0DWgoYU^I;~p1m%wPiy;=Wg^mPqs3N|jxEZJ@O2SEGXAPWgZ5#QG^Q!wSY3bUYiVi zX8t;elA??)Uq{~x9Gs`u9EN>Rg$l0WgqbcF-Uk0I5XOLmKexmKXr=WvPY3Ez;O0UdS{GzgLju zne<)Pmw)hkljZqp{-J5)Gbi5h#+#At4N%;DirhtJcck0H1`@23!v+$tAzPt#x0UTQ z)o$xG3-^&;HkCakYBwaLbf4kfJLwL( zmAw|Smvqv`Q-%jIUV{mYO*7j(jEclNA)edH_7gr)-a{K{1hLTp^KcC+BZ*b7`o&<| zn-tf2`=TYGgOHuH6Bl#tFkALgDQ+!Al2EF+oy~w)9TtM^S8wNF#Ytqc=xQ)+$QdGN zWgET6F5gjqi6*yG!W{@P~QX&?j_PV5d@>3>>7)(`{RGuV`T(oQg7hbjAPV4^|^ z;c8{QE83evb$Mil80qdBJ(J}{U{tU)vBbBR#q}|c42aPDH2d5MZv)~Phj2#Q4w`5A zIraz8-`+el%`z2op1g86RO3`=1i#>HNg4r~Y|oP`#z-Q6!0w8|}v17ok+=Omz9IvvZoG zku|kiQb%v}xvFYX^edW_m&L{21)@ViHR5f;_A_IgrZHe-A_%6Q(attP^$xZ89{y=duqmPn`kf<^vMz#MT3}K89^H)=*pmMOIQVWzbgG zLzjQe&|gyU+C6n;;J^e+ni(g^rb%|(Tv{DjkjbEwf9-y0p$QVjz9-eB(}^Lm)GTjv zpW^OPl48?V=0R#XBBMuVylHnE4vIS8VEHtj0i==Q#WfE(kVt)yDnpehgAto#5LDeQ+@ zBQB#efr|{ZI&$ThSj?sq<;TQBE=X2ZEW5y*BC%2!{l?sS^`58}u5S+JXbQq_d*hKI zHH57HcigLH=%FLB+ckrMp-xUaISmKPV3vD!(L1>=RW|c6)Nc>DB+3iI>(nbidE^C2 z+H81h`_|{ZMix=G=e4VBIoP4*=!dV?h&v=H^VfK86K&I7m<`&`Gj7aHVcI0O*mK%0 zHaVhaUaRcF>}|RU*yz|~o`$VI#+O#Dirp>bAmFlESx`?oJ}Um(<{l__pQ6V57+lez-0^9!&w*x}%} z3F%LHsg7S;V*k_i-+WEm;>XwJ8AE;o&F25eKdTOntO_Chmt8Vfuaa-c1*9^D~xbX)G@ZAFQw#Nn81zDwNwHwm< z^ed}ooFXecGQ#h;IRHLm1&ld{>~9*E9nSmCw@rZg>c_o5CdZsOj&j`$0_Q2Ni6V_u zOsuq$*`^zsmlr+&in?ijiC*Oay=;kMrzlsnh(S^&+`E=2N<}+G?ejLgawKrj<(RHj zoHBpeyvFDhD*YSzKYIN2$8VmA&S4IdQqBEo>)pHOU4v`esnuUieEm$c@m#&+K*Rvs zDQSX0p7BhHqMGhkR%r9WcY$sMo~x#lBe1!BhiFsK<^a^^8+>I61at8@0kE$5%NR6^Nx)JJM2h|x&9)tY;pNFUTA>&+* z|6~7avh5J{jg`lPvegB~rpdQuA}sD1H(o5pv)7>@yio5~FVC}#$vG`vtVG%N`*fo7 zi@Z8D%t_at2$bx0)U-cqnrwo^vj%X`H9?S={5Q@wKg6^^ZS%`qPYPZbcli^u<*bt8 zzyf_p#q0=8h`7Y0KrD`L^vKekoYNuclV#3CYB01tr)BX}qem*xXQZ+N9=#&qN`=@0 zoKw{XeGEq@WzDi!X}WJYXhSE5FOtSn*t^pSPY<|f$hW){D{bJ?Ak0^(?G=^yHiF}0 zA8^_PH|q}0NLTCvzz?Z4KYrN+pl|Q~urM8N?Cwp`hIg zsKPDshQfV!pmnR%qO-H#2mj?OwauCnn*BcbZ>6?g{dxF)AN&_OpK)X?$4iG)_+y`? zK6zn!P3`pju!TL|RiG&i08GRD+~a4pUiX()e?Fs5^_&ILGZE%0*h-JBh z(?(&e6fkrEl{Q9VfLFfD#8I(s>P0tHd6-a(f|FtEJdbGnO}pA#uHLFV1!sedJ_r@-eryx?dH#ce`h#1y?Y;GP8A&km4yZ_#d{2Haz# zio+AUu1(`{SnE}f{rgFCbOw`8ikNCT7qmW>N-MNI;(hWX zfoD{C;d$ZL<`25yS>y2@vGEM5s)8!3bdUH5UF492qOCZ3)_UTI8eQu+>fd`@Y$+Wh zBu?VI5g=kkEbpdg)l@yCprJvrL3`7)hcu{`sJhuxL2XcHl}-1G;-mxa*ann^T1dF1 z!KyoyrGH45=~nw>)8`ie2VIu;sh|ONx8kF~NA6vY*lgPcOAY|jqJR61zh(9A3xXIc zMaXs2FHXmvdDs|dflOYF`1JH!lF!8agfCOJZipsmb^7vc>Rzk+{qs&px?5_DW9;#2 zknkwggX(98f{FsR1P;0Ms?&h6ZI$#fu&Pwk{EomTQm@_z|F3XELNC>?Hbh67R<#)m)Y*w=2W7u}!xBur8fT|@Fl+VD8Bi+aoq z0_7A}Mv?u%#sZ;qkd4zNaK+)knZbVO-mFfM=V(AD>a+~J5M<84@GMjp6nbBD<15^t z;vcFN?kP~^E?qvT&Su_!@4t{Xl;1|yiVeJy$Ug}c0@^C3%OjnBNb|YU$07VqdaE}y z!HeRf#)~LnATcw3qaBP3Mb-^r`nrI%mIY2v%kes~1+h}}^g!JvIZpD!HV0(Vi=@Tj z8KFS6P9xoZ8?2q{-J!s{k;x>BgEugnAUJp?GQ$=6j#7cV7&)Qxq2_x#lLb4ZdODtp zr|iFE8m4pna2XdWtUhcVE^ZM&`iGZH*2e$$74MLZ!j>W@wnO{P?9eWXgBpYFRLrKC zLm?H@tCZP+dZ6$D!*eLSU9&9umLl1sD(FMt_P8CgE3{F&K5Dg_iBGH@DAqe9ood`N zqo{bCv{dE5`kHZq$SMfNdtpxIvUpDuRQ|bUYZF;F(!%7tUKdEj4^KAiptuYQWQ}5K zUpFRp+Q7B0(6+f97Oxj!X@4asN*~ZQG0oBIXB5b?LgRxDtIH_7qj+|q{A6I0L!xcw zvouC5m=EHBBkKlbN#L|UkB1t}i4BSsYOuAzbp)etN5yeaJ`JUJ86c~GC$eoO)1)uiK@&S19BXIlc$ILEC08z{oktKH@(aR{%?bV8;_yve9eqO6zbdIcd;!?K~vn6LI#R2n5 z1{Jdm^bi&WG=_Dmw=0)`mJ_PKUtpHb!gJ94%B)o14rqXdC(ZA1$Q9Xfx&)dA_xUDp zDEzX1RHBQ17Wb9nVTi{aTtV7k4=~mI;b7)7let)7sFR}KJQt={XR&F z&u9BS^7#O10ry9y%r6MzFU(xc^@p^Eoaa!UJ=Lu&B6TW=Rc6mR>D4C5qRS#NlIFl@ znsF{|f{7K+(++5`I)osYOxn_>%ClT%o({`8aqz^-(gVeqFW;Zm9KC}{WD>m&l70Tm zrxg2CzT6eOor|9`=(2nYUf&dGJO+3D)C)CgPH6I=ei_w%hg`5TvqN16jkLx;z~GyY z$E;bgc6K?f4GrG*k0yUOXzA=9ki#QId7U_a)NbY@pQX4@DN+yhmO#jY>BVZg9*Xcg z==&sBZsci5D94-uRyQj@i(9y#h2{pA44kwh*Cxl3-`i zkPH52t0kOp_*h*B%`aGA6Tc_3h!?eTEXgIOMTa0OlO9j$JmeiO|yOpv+cDXpX(EAP@N3x+o z`Q6fQAgSx0o1^cuNuur$+I{&Oy;#(P_Fl8mmJNF?iEJ)K0v|R*GB#s&;9kwnH5&t#}FPk$c4P zOo`$aTNv?JTnu_s88fPBBwOA`@Jg0X`@Ab&_f+lkE(GHq8PREl zTQ-2e+;@0t1jn4PZbT0K@u7aIX-T>M>x+LQE1Y;q$u?V3Hc=eJ`qoh~rOJcTF@e?w zVf0E!p|yFefQ}y&A>;4PtYIo;yXivc51W{E#4v%vaNw9ZA8AzUI2a+DGWpf^D4nIi zyO71iiQ^CFrJ0){ zeG3{5C_$CPcCoqgouQ{8r^}~>wra0>-<{VJVHlvAP7laovS`cks13p)d91$Q z8Cn#O?Q0JkBoGkh3y!s;5r8N3lpy0W>20-pj%5?7)56C}!)c4Mi@C;P*@<40D{a%H zil8$YLlO1rN*RhJ076Q9k!>4dt*3ykyoJk^FY#>h`UJFpFrDAQ){(=$iBNH}MCGWo zj~(D56bUxnF9XDY%1wUCHp%){lIX;KN}idYqNlhtifp7}(9#?VYSlJJm(RPv zoOX=^nV&>x;izF^gAa-~$Oqh8VU0WiT}@^{r>0w z@ms|*id#&PSZkI7&f7HXxNi~!5Nfr{ilP5oDv4Rp^@R%(?u3xG_&zBHk_4ELgAVg$_|o~eMDdj4>+?fd_E+Onxs$Zpt)(|=Z4 zP&Z5Z=C-rBuarghs`H?N8bu2S+}k9D-er-Esxl~ME|}lWo+5{ocj>&qMwPMU@S5ln zb4ZEe?+%8P?GUnbWdz`K47axZChd}C#MNncHddmm=SVv6YVU`dfh#nMYP_2zeWa`n z`g67Vs;rGZFY0lBpt>>l9Jw7(#MDkuhQ?MF$YHZRhS?j#Bi7fF5u$3;_a04BTPpN8 zEj+Af?qJ~koGe*9zyc#|#q@IjTxm9~*CmVee!D*+puMtmIp3u;R!bbE2{mpUA)*|BCF?la64U<&@{gY+?(3it^!nLS=7-T7X1vNUyS zZQ#oJT?XyUeb5y7#A4b7mtR^td%QBBJm-i1GJQO(UrxM#vQl-uOW7Mf5Zngp3+?nS z*Cs&Xroc2!iu*SCdjG?|o0UhU84&M1A5=&u`4vu^D2GWfJ{(}9MjGE0MYuWe&IA7szdozuFVT_e6Ap09iYmrK_jUYCDtz4j}Cw>2rT_P5>^RJTf6a zz2_XduQRP$$724zjubg@=BM5)^HWQ4Anj64#jJU$Ko%#x;1ATr8KDgyp zN*}(ucsAJnSzBBmh9rxi`$)RfXJLaXm3`oLgYIJYO7}`ZfpBM7PFN9`ij2_1fw=Wr zh>yXFCCP?x|3YK`dK)YRgX%1Kr@AZ>8ip-aVBe;D zHr=>Il3B^oFRA{%A+6BsA<=16jg+w;5%fSF*1WZe~lAW?Iau? z8>8D%`d9M!g=sv8)bp@1>;#HiNs;AL%z#Hn7#1!dY5Iq* z6VvbY^yH0>o~NIgh%;-`yk7pE2@e_6z_-ZB7iOn)%M1^fD6WMf%~T8~lN+FFGLc*d zT0p&~73iF>Dg{gO_X0OMzmY3|f~W4Uv;p->6O$MOa!{h$nS3jIH@s5Y%2sOkyP+y` zJ6lHgO7P}x(r!pkcYg(#k)g49r>=_ds3HUU^M)I;U>lDYbLDHburRLEo|g3ka}=;l zH8QCzx=<67$zwLnE{wg0yh78*Hz3}6*M!daQ!=?-lDk-BNv*mZW&vkmB#3_^BD1XUHZ$u$k&h1D|rexD9CUKw)1&VlS%$hMrNTg^QV zxfjCkfEv(BEhs@F%M@QBt`rZs7$vfxILqtC+#@fgvRDsV8d)EjI(3ZxwS!S+uu|5s zWc^BG`R2DRrM#^uW+Q(&*2mryuhW7riie-T5`iD^*uQ<&f3s&^vGwD8<{%EZX#Gn& z!Y;*mn{a-htG+~=EKJ|`{f?73VHU=D|1zN39_}BnrMM)Dtfpdahg<}a%=MxBpzeWx zD7xusEN&R%w|!j5vJu$N0y*)+H|FrX>Ng?bb2swmabRbCx?O(IDWbG zCVMETO1W^N_e3^MGg$VqOxD<+5oE$9{rlj#w(*qVIU!0Cw(G^lVaOsyI7gFO5%7gUd%IUK)@-tY6svt9)b4hH8hYU>SSLMKo& zNeJanf2~>5BgyyWdHkg+MQyOrMaoDUCUN`|Vr`F3eKS(k{Q`sG&4mBtuHh3aesez{yUGsE0-nBTq zc6x%BO^_TN9M8IAtbl&jnD)449ndfRtyTNHrARx;kG(Ge zYbwps_KNqCd>FD3%#Gki09h>ALJL~a>TP=8re=D&yZ)Z(8kxVxS-N}XPgi$WFJOt{ zhNz%|8bDcO2XUh$Y%aKz3KVLp5Elf&QY-~U8WsL?l2DRJBo`7UR!#Yo+&!4@z2`gM zS>E@6?`@K`q$@njy+HaPxGZYhJltR2L-&U2)yGNSLflPVTu(PII?js~ZWkZt4Z4&^ zeyXhX?%`*J9p~K#&bvAwyM2!o^Qr`(Lln8q9q&BkHb_2U&x?zB-QHz(Adwm`U|we5 z;M{~UV;wM~EZV5~cL)FGRo9R~PUAMgTJ|nG2qe^i3HL!khd{4JM&torA|LDWF3rkS z4Z7S3ekiS=Yd}9FPSr%8oP-{=aiV0xC>ZMj+)$Djd|zmpfc%_PH&`hlZ17h_588Zid1}n_Jue1f!d6kZ?2rUVJ_` zNK{9>HC14O#InoVkCAoU@-|K!9@%FGpfpOcog&FpjKQs}p7$tpr4V@GvQ>G09Z+#l z#@PCJWO8Sb^;xhAP2wfxHH zLvDK|?Lel|N3KaqmfmD*`I%5_)F8;`tzxu}R>K`RKHP?Tt}G>F4r7R3i@mEE!d77t^a z4a!Jjpz^ebB;Vu%HchEX;$I$dFkh@yb+g^@tX5H*pvP-H@SfstgcVK$ z9veK@d*Jz1OsiW>R2j3yb^ju$bgJY3ReDyjmR%1maNVq?Gt~H-oe$wxL*BzO0_wukPa-AB?2sXJITVID< zj4GR*phTq$!6bMCy~(4|Bav}1;psS#@|i`j4<<%`%HD`?f5$Qe=d>=em7rP%a}8>u zx7SBiWi?UCloo{oLUcC;~QVz)L_!oq^ZnX^leQVLS zdE0pJ%Lm+dd&NbT05#tQaf9n+5z;MXFpp%HeRj=QqXH%#NglIWy(uW(d&jIcL3L0z zq-jsPZVkEYW9Ncf4kPs0TWD>=o2pG3AF2A9}#w!WvPlhv*|u^A3WU-=q`eEAg&0}$?c%~ zi3Y3rF*wr<_4b)wcUVXkYLx{6#UP6a&&ecP__!i8t_C&xJgzXf7{pJ49;79PyL0LL zu2poJ{{tF1cZXbz;sCgY)YJ&-4V7CtVaA7Z(9pH#dtmMTusm+d`(ph)IFfz2BhUn` z{oQ4tJIc+oapLOB6K0@0Oeui0u8fMQ_}(>;eX0m3f*O_rs7e`h>5^nf23@w01SZ|T zLvWm6{X&E6ncwoR@Q1$3!3Ex-GXCt6Y$33vb6b9HmSHR1e}MPUSFgsMhrYRPiOeZ^ zs%rJuYnR>t>3aLjk`BO>Esz=WnxuQ~{~cjM&f6a@Dk5c0oIJQ-hJ{*6QALrXP>Cx+ z8`BmFK5f1ylJ_?&2iOmSQ-E{@4&8Mx7WGDg{#3}bIl$xMb6BNwi{oDdt@b0q8T>3{ zWsxO$#)WBpHilJ&Z16ry9?mJByDhSm(QIepghzt=33|tm;)CM+*LvRc!Yi7?vA%AN>ktabvlJ!!zwh!tZAr2yrCJj{YO zE3qNrVmKsI9`hjVQi@Rl-!ufkW zO5~U`DIqofy--{YM}ddx@KK3Fn;rakuVza{$Ft1fDV?X>M6&5RUb4f6M(;vbJR zE1xdyhuWxI2B?{t?n!O5V2^Bs7l1bK}O`R9zqles?jrzHXzdM77LjQmH&J zw*%Jca3Yc43Dm7su(0lBS71kcqb!{sfaq~9J>XFnfGiI7l~&^fvoR*7iT46L$6y@q z>uY^Y!0CHSwwmmjMrzEy@%@wnnsy4Qm^J}+mFDnPN9f!>UUD9kkXJ{P!_uQaG%--G z9zK#NZjb0x^)bml$3YhNg7;NvHpD%eLRNtT@PsK49_A&u;uSi#es|;67s!Njda0e) zeT==f`+_B@p;b0OD2T5HB2N@&05wyv_J*dyS|)8qyRwsh#{r}N9Nm^%T;4*=AQ5-=7gs0Z(>gEWi4~s~Ie~gZ8{~&U^eQQ+5tIea@#^LK#IeG9dPnpQ4_NfT zLaLj7LLSL`S!6NPxiyE@@e4^D^O01qe(IgbB=H}(pPo@T-}oHIyboM}Z)Nsz2>$)P z_t47{F;1I`w1RE7BN<83mXK_{MyJC zvdgo2UICr&S0%cr80{E@Qv474;dRY2ec*bem!*R zH?B$ka+B4oFEc&B*AP1!FFzou0T9|rYTgbOYsbRwfkgwED%`hqxzZl{}NDQv25NpzdX zW)DplP_{z_Yv_QqSyl(cFl%P#&%FxOeD^}jRhr>L4btWi)cS0Ps=6VU%}`O3?$;`tFdO(GuqkoqlWa zgLmhf%#({)b(my3alZ1HS(xA;rPxoA5-J8GDIX|NLO4Aji*1GKtQ&!A*?Y3XtR^?; zstPFyd)W>eC3RqFn62uOq(A|E26GH}4cASF5Ce^;6Wj`^Mp=vWuH=RU8GUnup1LES zVIgH-sB;GN3zOJq2kZ)LPJo{JpTE#;wy|!jIb_>nBxTBlg4t{zO)5!+DdUiQk3xgu zAy~={w75tkG)xQxBF|WIlEPxFCtGD;7(ypETNUqF1Z6RX1N#^CMP{isM>b&pdy+8Q z19hP7Gg5XOq2Zyj%Sdq=A?GzY@5TITJAXKGochC4WVI8U9H{#pky+YCDYjB1iHf-* zx*{^hmWEtVT)@b(jEeVo#$G_Q88x;e*5*grb1dOil-Pup_xTO@F zIG6*h&Ldz{K`9PVq@0S;gLQe*`zk@3y+M%bZ1u89IF-EtByn|GKjtqrQ3@CQP#S0Hw@~hOA ze=`}GcQ+-xM!L8e8Rwl>$!fEedOxLjM3EloD+M|TR5HQph8wCz&`3)lWssb#3+N+N z(0?}MQXE+?Ukj`BM5gbxjyZK6{Zd#3=P?<4OqiqRW#ok%a=AhFsDRtm=sIJ(@lbR# zgKk@ns@&@|5WjXoxkOiZ>HMZ}D3=gppK21*Kx>Zi+JPa`u!-x-st|Op`xd6EdU*)Y z0go!+wujon6M-18%v2?V`e0w=;QSNww91=a4b$}m)SU+0^`ux-9@!b1$l!dN0gr4G zwDHPCr{o`oU+~in7xqjHdM~wC<2-3Z;h*oR_gbo9IBg)&ioQjM9IMg$BX7B{U7Y+z zDnY?LgVsc{Pfl1WNfVUJXakwQ{m@lms3;!_P2}g&Kp?8nV0}#@#1S6+O)98Q0=y}E#28Pvbukf37j`{9~q zz`ek&<}j~);@~pFgxz*_94@)Azyz0NH}W=+{AuKr+4uZDr2u{6160f= zUXNrG@0z4pee|2xq1El8I*z)@_LHlT*COMD8A=`RnV&A?lwgxM)eogdYUobryE?Y? zLkPwkih2~ePFgI=W3uVxnY^b!E@8a!nkaYf$H;@G!Twhetp)?i*fA)f<^-wHuA-3pp(>14}haQOt+ZB z=CGPxIGDU_z?^aJ`LB+ zGity>w-Q|t)kJTOLLad?JTc^;Up+m1-9zb3FZ+!va>t{Bd^M<1{SJYgzJF zACUQ*@KXQAmI{)^&5Lo~e;KJWgZ%+YQAUwss8jb#lWuvvL4e%%dUYwdx5@LWm;#tH zi`?$dTFGyK7KfFpu81D`!NNh8R(8;(Q8eh16$#-XXw=ibHt5nx23<;LVxyj3-7)7x z;9bE3_V9v-k;y)Ge+edxe$%vSH`tpVH0|%d_1-KKY&w>NUL!l5IILV@2804i0lZw< zR7|-%lZPvz8X6NOnjO){`7P2|A&A|sWiP&=RThW$LA@{}Jj%SVLHFVtsNjbZ3$S*^ zco2%Npd-tnIhXS=*kSxRf2J)raI$_=Qor-3|EDyeMSb=kD@i6dBjdzjAdpHP;p~)A zieicsP%+KQWN}3_=00O1j6LPh&!QM|>HgZCU@T$pP2@)$@=$cF3g>^ehiT5GjSVyLo}3wgbW(GRIz9{>wH?DGV+|!cIIf zSrN&2y108*){AJoaHqVxP5feA|sFfNQ( zKEwh0|G4YMcPz!cUhLg<;%tKzfiDd54tQJ-tASC5`KQ#aQ zO~Rh_3CNxOhdsYGO*RQ@6QfD@G_u|-SpSSt3{d1T^hneD!hl3dStDNsws{|2!owiC zu|fsvA_|1Kn_)`pZ1wMku6dN$Yn;_Zp9sW3(t50p(6q3nAikMRpO7}ls%X7xFyacc zeLgH}lbK{N+8aDELv(3IHeF4+;l2WBZZz&0Whcuc?}Zw9C5K$j!Yggm?0do#1C6lY zUTD$$b&5{|_NW?}LAmA%1ZFeBwu=pQem=#bi{6iTr{oQ;r2%_AJ_JTl4HoKUgqcEg zt)lm%vQ-^|PB{kpG4_wm;ICj^-78h(wd}iPZk9!WpOcBpN-ZNcil+u;&5UE#vt1BA z*vr$+GnRTiaL03dJ%-N?uxX0PvSBt_uIF*bNw8Y@5kEM#a*bt0j+Mff)#^ND2YuaZ zr*P-uM%gXiMY=q)!Bwl$RDhj0x2)cOvc5~>SHs~Oouu#5^m}o6y_|O~@l8sx+7cMA zls97q#VduaqGTVf2%6qxt)r zKmXf_EQOrd0k*Of!lbjY;tx3`@c%9Tkm>fi47si<5wQM3VsR)G(UdOkVCzYFWV5n) zthDo}-#XWWT+s8`)WY>c#&P|(+F~BN4##o7RP$%D%83JVd(46+$&_LW>T|fMnk8sGKn5DC(x)F@_0R;3SGT_e-ni-SgLi zv)C6GSws}Y{CVg;{e=eQ_xE{Y{%B9S5`1hx4cb8=B*1; zE|N4SHZmapF(P1}PbqRKl10T}bS;gk0S>DqzOjtA0WyM`>t0oKA_Fr~5EN;`m_wCi zyVpv8E1Dj)zSg| zT#ysT^&#~$0$LfAVmC$7sF)PChF(YRd)=3xCW&GsirDB{?9)WY25SjYZ0ptSIvaG1FOY<3lQM&i2T&7-%h*r&-OnRVS~8)1PNpOsCB@~Wpm2Jwai-ugVvl_F_4+Y7hbv8wka z<7_i*MmU}?d7Rx&S%=Ln{{@t#EDD!d#fiaVMMP!g^sVzQiPEH-cq>V^p`QH8TUbr5 zSI2{V-KxbI|0|+mz}>=V_D7K>m-KHB7i5s~&y7pkXy$JmqZB~?sH0+Pq#69nP|({T z%XPcMKk0c(ovQAZm8-YSgG$Hnc6qn#9IwfB(B%|;F}jyzFx|2$!Mf-{m(z=MppSU^ z&6EGU{(E?K!}Mq0$o}dekkO|Ijn}P5k|<`=7cTKQ@`E+*T;mS!l54@A{o;NQ2dW@_ev=Rgt+pln4{2SH7Lgp zt7P#8MPf+4r>1u9GjHS!OI4Kc3`K{=-kDfox2WBqGtx(}@bG9r5z*v|OL!@44xKE% z1mq4`LB}Qd7)Q8=cKWU-{M}kUz`B(mB?#)g8Nc6tt?(-onUb71_GV>DYLVhX$3Pd= zDA2jJifZS!v$4W-adXHepfXDY`lpFo2urlQWStkOqw(Kw+m~tg?`q5ka^wy9b zB&6o`xF1Fx$UVR$|(C(LTLd4Ynck!pemYfAn+ed7L zg&?1IBNXKr|58O~%t;WUoJHu?dZCN1Tyo!(Dwv32}B zRlYFG7i&uLec}VlLCvUk$w9HBuw)BdtgS>HIe&11%ao^)m%gPo8IX>!*=gkHD@LdL zk(p6BODTYH;RLL|T_IHuWn-AIZ&rfn=T*s|9L3UdVHb0}nk2F}ATi`q$&C=5TNUq9 zNs3#pw3}W5|8>miCu?QL!#)rdG9W$QD8oJgr~y!+?sUEo+j5ZL60Te+#Ots_uuGN) z#I9I#fPPGg{QPUsKzJYc!0*#lf>O71y3~g=@#TaO_EPwbJju;wSx4CXDIZq23zw7}HI%j?OM~Tc# z{mn7?Czfr*Tw?OhyI3PunvNU66x7q(#Ta${01Oaz@*oj6a4y~PKy;2<*0w97Rgn4B`0#~#S+z2Co|Em_>60D>+Dy2UnOYovvEVcRj_68LJqI@4r8{Qg6? zoa}I719HgBfb63b5E0u$#q5GOs<-B+pDu>^ z+*&a>kM5|F)d9|x~0_>^*`f5)71x$%UiDx^_{Wnm4jDbo$t9B>6LHP3Bf z2Z0+4ln|3ax)=X+g*Ui%g`f1;J;%8JiKs(b74k3=Dc_7oxQ_9}2JVGd*!lsEW6Wvb zoHthg$*dqx%O(t`T{>B5tJo{vr{1M33i?>_e9N-K+<5+fJIswJ=LB9VUbQ>IGHUO% z7_pj8K;r%m+EAE;^b}amTpGmH8gVodqg`&BK^yH(#~$N^i7CnJ-u~4WD{8N2db~== zaaHTACqx&5vL;W+?S(n21jDfg_ruc31(R(!GR$|51&+nj|LcQTA4?W(r?nPVc(E~P zf-$uewoG0by>szuca6^NAlV3|3P^YZETOtNc-|lkdC0q(f5BVR2HZBuJ_&A`&5>1f zk6@>pJ=JE3V-V?UV zxA8J0di749guq&U0`rWY6P_1d=^qExw5NqTU_t)D8`t1`Mf3r`mDD?TmtB=~%5!+V zk$QFOGHm0zMC!k-SL4q<(NoEFui^8$rHSL<=HZB`(JYiL$2emzefNw3KNF^YRdRe2 z*~<-6P8^l1HN#Xnr6{FH5fzicuZUhLL<)S(K93ak7%!c^GZQB;BSnwtps@r`a~k6R zLoWTKB3i3NRV!0WICEZQL`E3QZ|xAlGqwfR$m4_BfgjthX3CzZ;S7zJ+z_ev+xf*p z4bt>I{nD8kABi+b+GJq4z--W2X#GU)eVDmKsI*{^&?-_Ci?Zo*c{g1_ulI=eKFID+ zXrS3gRy}XX1)B_ShSoy@^SS(kT|qO!u%4Njc?kx%pZdnd!^eJMvNccd6-&ugZnnmG z14iVjnXS1;Deh3Di;6*IhCjwZqR>ErvKc5y(**fqq|G04sS7ZQDBKX;VtYVxAzyfl zS4C@mHil`+BNIb1+){n3LK>iRFkgag%31U#_Hsaut8PvU`_!{Tu2mWz2T@4KMT$E^ zpI^|+E9KpXcH}3aDAt!7w1cM$H{OK5#*mAVWV<}Pb}qKqr?BqqBa2n^|)N;aXA3Pl`n$To@UAne%y9mK;*xyt@)iL%Z}5ox~wqj7)zt} z@+zX!{fAtzaw^*(E{#3Nb>3HkR=hYIFMJPnJiuY7?f9u*t1K6eIYonGM>_S}tQuQEr0z?ck8=gy}OTC{-0BK=6qL2f` zY_JbwL&w%Dl73r~Dl~!SH|5iRM>aTdN)l-QN5EhgrAWn+;25Ln&j>T-zuMWgY=&$3 z+zz_TKi&VR${;L+Vp1p=)XuJ$B-=8M0k?I`CUhM)o1*@4;&jV2q|;6|R`R0-qizvjFcg^3}&hX&{d8$W+l@FW=-<8W~T#&||}ocC}7#?1-OWDvN(@aWE2XQIl&PYd3I z%8YF2Cs?=0c%LK7^f)2N++yI~IPo#|ZPpR=k2A~EmWtwB3h$h_m*}LKmve+tfF9L; zXn7`S?iknovtC{K#P7k!x#nvsyC}q4;<}aH4#Io*%i#2iL zgn^YR87$<}s`A~MmF>zy(dR)6wR%y?^l|y66C5#H5ahTS#`-V+o88Pm{JUvod^>v6 zd~*A96HZxUrY7-MO3_D=UMi+Up2T18aY6;ekv+0B>2ScbAmFY$QaxzW{L|@7ay<-H ze4)Q_FK{>?fA5{u z9<`pY`wzxskA8K!?-Y6Ob7PM(%{<<%lp={D8>yIcpw#*BE1P&mlMp8?V+=hdy^H?` zd>xxm;mD;aeC$Lz0s768 z=dig4ueZt1vux6V{u?Ob<>D}uM-!Y-* zc0}!`Naz39fqeDevJc(%rC|WzYrBJRt`qiGk=1 zRYS12aOq9<~OA%X54V&W%?rKM~j=eN4C@#oFLG1VPq5 zjKlHtoP(P!We=S;)on${a61zxjOQB#Kp-N5-ES%E<)ttw^+r}iXQ{4+cg)Eo?aEJJ z%rzUlTSJf##6FYGOTOuGkN8WTX^&H!e4i;l`}5!Z-jW{mmC8?A;d0$CuJkVu?uTlH zM3_MuTrozL=a=r2$XxSF3(+d`{PO&A#A`sIsAJBFz;mPnW}OCE4M@#ri1q41`jP59 z$ikje*1++b^ka{MQ98ao?8FNrJZ>0`m|tFas2vVihf#AwB4b%x#U-ua#14;@Qmp&Z z#*iG<11Nbp6b%wWlD!@WNsR!xLNeWJ1lRp85S`m6suRrN*%$0GnbLS5k~X-=ntWi2KSMG;szuDQdr{?x+040 zK)gLXWL)tu!hVdlIsDiO|FmvMM#))H_YQ5p={~#`|Jt{j$a-!buM@Au^UOTnos?n+ z1!ES2;$)cA$x@wPra9@_Aic*}QNZoKE1}g>7 zC~KkD>WC|L4)#i(Npx<-^6nW!@cQ0_pn_KU-m=8MLq-LpJjE36N}-2zo{S#1h1u zM{Km&-^#Ixj32iVb6Q5^MOjPheqw^swm*ISE9BDW#zA>#_D0;I6gMey9Yb=Bf;^@} z&?tBurB$Ag?vZvwwxEc#6HS`{QwF$cJbgDHDWIKX)1|z2VEJ9mUnOgV_DOsIo`Eh! zb`AfO5`CI_`r{?7k~)t^3k?qfGT=qsz_^HqFDmd*m8eF(N`YZJ;{)=T;z*r$XK7nrus5c4=c}IdDxNGhT&XMBCjIbdW{QD)2lOx|3E3eUr z!`9!Z|L(25MVo-u_amRPWTg|QEHliYvxQP@qR0j+=3Xe6wn#94OqQ<(YC2UtbQyhv zsSA9`cv$~NCUnNeVV=-ujOS)Emx}cV8x3x%nwh9HlKYn{4({3c%EED&`vdfPN~es^Yd$#={J-`H$p!C0NCOM-C z)HpF1j+=qu5Tz)mNGTOl9oPy!sU{_?0{~DDVTf+gL*IPAO!xjstRC2-I1zZ3^er@$ zr5Gvdw8}c)&d@{#$$&7#}`Qj87Z%?P6c|&T)`>G^?!asJkOXG)!V*z8` z<|M89X}l#X^NZCe&btyLR`gZxccq<#D`D6DUc0>>$us61S$s4=3&q*1WcA_=eCR}^ z2VIigPeP-6k>oPJUH(+?NLK0JAsBSI;9W0nARG7t^zf0xOp)g{-k{4x?{@j1%b}=# z(N^z4mpj4r;!N^XGU)OFn-DnYlChv(TuvL0^zi$H34sscs(Z6zh4?JJ`jPy+>ygE) z_%$I%l~9ojzd@|Sk+RTAo!v^n!u5lqd3d& zgr6l}rqepUR(MmfPXzUA^y>I&+o?f$F0)%CxUaOpQ#S{>pAtNKB6bD5 z`_3QNPApGpeNsA^E&BS{HcH&ds`p$DdG@TwN+r=s<2zPMT03(-4VN`FN9-JuDs z!yi)6aL;+wQ%Rakvz3h#=CRpyETx;1#7|Y>sUbfNb`0R*I5>PyRu^zorB!WI*t8xW z8(tG|J*?h3HcUCeYl`2IKM5@3y06qi>LnDrgYp|Bk9jZXBUn6XQ1Y_?Dsyvbtgh87 z?CzhubYwVy<)zQD_etwu8M^+#z~>PdiQ*o*CNP74-R~i+w9n3IdR_CJwY^4l!t;ND zfgrOEj5q4$-?VH5veH{~TvhIMmYfj37rbU#7QHW^g1$F1VOo6P2}rl?CAZXDNDu!B zl;o0UbE2JEBDX(VscDodjCVC`5EU9_kIjyq&{IJ}+{u*}?F zUhAVU+SoBh{ zGR3z+*{RNzToCBo&@evqtz}NpjnLq}mv`R_wWK%mb#9NNt}w=Jz2Vwk`6}i>)b4-+ z`hZ;LHoOOk1i|}Jb%LsZPT+DrAJ7zld|^ie;%8lyl?DvD9FQOP#x0$j)?>f}w~$VC z_!BFm54mZTM%q<0)2*V+d22oUSUj$CYmh(S;}$-7m+OXBp;4CLc{8+8mcn9N zz|I+(domPb(^6euc{f{??N!a_lqJB^(J1Q{wUIs20ayt2$}e~~i1zXl_-CSOeSs(r z^_1G#qkf-C%DmbHSZ0g2{qd5tH}ae(PSIHR4myEp($m6qzMG?Pz11Bvv02)7yK8ke zwsvAdrrA2KR^M{@^{XDXxhU%tRFbEx1}#S&`*fVlP80fu;bt2Kf1)2TuKTSI{wTiK z)WY2hZd>*lap`?_;+(gYt|SAcs`4|%(Ax_&Xl&B`d+42@q@^jD(I!A6lTGJAF+{c< z`&nN$nl^yR(o($aadtab@SfWrgH3Zz*Z1!1CwWfn`qZ1vFBO#H5Jk$V7>uU0Nb@9y zA#Vd>p8Kw4UXZiCLXhUPicaC_kPXTxw3kXS?1HCKNf$vT)IL#`ud$XPm+6n(%5Q^4 zN5j1%FIcexX*zQK?ZeVB4m4Pz%1>VPio2r_WkvQT)E_?NeL1`p zoKcOD_H<&d!#GC6R#c4R(G$9b8!FCjjlDe6WKsP7L${pl7|n}y-o+pRqcWn+av!CD z5}7?zOa*;kv5t8v(X8Q@d#wpu6NVL1ShEZ9fmC1A%hUR#dT$jUaw`yCWfG{~ zS+^*cem~sMC0`!i>Cgf@TjBGf6 z9o$W)gcSl0OCi6P1rBbnRY1zUn%~YKgK0a{3ajtFh5ZCcJG$ux`Jvz(AE>VLTsu}V z%FDj+vCmBeDq}x={8w0qN=G))7O15yT@OI*v2sEWf%tU)A1(3Gy zqGARXC-V|R@*ou5@7}jC#VuEAVC2YSZifLa7~KXW{-Eg!>H#-H^~#(uAmu;2-~bRo z#m=%RpUim#M%=&&BxV!Qh=({}Gv>Hly}$pG{Ff%XG4vAMUirRf7W98aDS9aK>GJ|7y?WcM z+uk|6Q-Z?_Fc^P-)@AyX;3U}(1^nwtEt5kKSs9v}DEQwpZSG;|} z=FnJSnsCju%S-Pr-T39iA0)q(Fs)g=m#3R|gM9!gft`!-smRCPAgiYb7C(~=&=0+b zk7Ie}elL82UX9mXmtY_tkK(7UTfAY$DS4JJ5{B%W(FN7J_{=l%E6iqafKHO;kV_#a z#67-=q$@mDs9lsrzqfdJd)uNkq2=sm3+~KpilcTcFv$sbl*c!oDLhQLd%Kk^C!3uZ zccAY(Vg|~j6k3Yxq+&9dYruzb(K}N-;Qmo~hqUyK-xfm!?*;E&K34*p!%y;#^MKqK zpwP|M1!k$**@oyxvXhhHu8a>O76a;lU2YgzJ~Qm+mc_YTyj>?=GFf5F#bP-GF z3cJB`=PX_LZg6?h#n->q;Hp(($Ysc7d2Dd0=PJK6!Sda6IzW66m(N2kc+Fv89>iuf z6spp>t!I;c_IiM}J-d_5R&~lNqCp)t+spXjx4$)hAaRBUx5>z2z-`~6-N(>pglwRJ;RZ&Okc^ z=}aL>*v)<<-TD0?7yRRRT-XW(3lrjSC|I{8q0P5EEDPzJw&%=Bh2MF}hl(nXHo=wI zo1>0PYUnGopYh)dj`u!4>(AZ(jvsXM(X$x>=C?U~&pHG~SyNN8sWVnuszFOC<`E%mBdHg_nmBy?{LBe zR~Yw~#tC6+ucGl`s9d;CSO~3y32ydMRpju|{($dXI$hi^EssnLi4D$| zw9M!Rk=I+i-Cl4)tLhX(@g%%DAPdVroO52XSJKW5xpc|ihqO|wsM`ZQU!-gW+}qWJ z*%SSC0R3l=s6GGMhn6Wsr=5tbqzj?YQ4l9=kXMPiMWBlK{p(==kihq&Z#00=VLwS^ zR`GYuxGC@D;gOqOdzozEGsz})`}`I*ZN?r|hd?*au7myAw`&2%D?}b?**3WqOYwkx7}p zqR=4Ptp@^yB-a)Z?4oytRnsj}JfKxJ1MfFn&nAYHko&;#4VR?TA1IH|?aF50(YVEn zT@3R|A!THgO7?+GyV)I^@$piB>iJ1~Va2+Q`Qt(VKhHNI$3?6ORHaD`(9MjSsRwrcmBOT%xKP3v--`J8TSH0%wuT&- zt5vp0%Sk~%A&{&diarmlhbyoJ2{#YVUdNmWEQC_O$-nd}DR|mZ!!?iISbJQH83X===F%UlDG$*i6^(>C4`(k;}*pZ*+pG8-S zjs}$R5OOPp$cWfINo^w|hRT+HBs72{Z<&;B?b6^j-mo9!@3NA_Pa}KHUg#7`v4tX= zsF>Qhsew=vEro83EY%@zkdnffMhA_#7|iR9HD}oO-^1ahwC@Lx@icBPrPIzpvD^RX zXPHKF+R%)Zlo9f_)WGVfMp~gXN?Dc5p~sn5P(snv{jc8W>C+r2FfD zCG1kTeWU#rLuf04Mp&BXn-b25@k8!>8`65v zRaU<2S9_@b&}V)<{M10L&&IH4{4R1MG>#7xbD&hGRSq9(f}*;~m}zVP!gEtFyk+xn zj{idIc3|H(5AL$8`+pH4P8{&J(hIOQdZjQ{2rDbCvLK)xBC#oM@!pyox+1zfGKVAs zANM=R=8mkEnWRJ7&YqtIM7gt|z5?V44TrBrCAn4xrb(MYAhb=XX@*2W4*7GMe>(l4 zm!>kXLyEm;JE6NsRwe6@UYSuVHFVtfyH$xw-O3}ef)mo$2O&rdi3L>>X@~F&J{wx` z2Ry2HorXG3{2kx`<~NP9(Jcy$ymW@2MDmtFd1-jM7_$+EEAY%A?|flVxJIu`5@vhE z2WG>{v7LPd8GqLA)4gd)#No8dAuH4x9iUVM4hXU$}6&jgL;imd(JpS}Of0uTE9y_sC;$W&f>ZQDG^In6l{7#n;U*6bS)%&W;0nK4RfyOk^ zC*aQW@T9At+=|<6;d=EeXpBZdAeA%cxbjtROsc1IXMY-wb6{1-2JZ&fR*`Or1|8-P zeDrfMJEp+|0d8oO)uLh%#&;ac$}k}-H3`vN>yFYW7@1o1@Wk&d%dI{qH8fU=vzq7( zK2C7uh5$8GnC|v6c2`3}$s59g(fOWfQVnLtQGp$ru90S_QZT?n54=3GogMJ3o8!nL z@^MXn!=AVU0rm^_?aS__39aI5-)bW3KQ}>~JhR35PD-(Zf>DUML2`pqSTrzLW~J$p zW)>Yxptk}-BO8EZ2%J)F@oY_A;rGyZR<9Qu+89WCFuokI>dvAfe`Hr~f6 z1*q6mQ88E>buQ{0B&5@rO^OUj68{KU8&$0s;0=Ex0hNqnv9CIF_2#*2|~SW_dYr}1=eWJUBL z??h&|Vd*q6qrw>&&rQbITN3~exA|y+&y>TKi>lAbd}L(_hFpf4W(Mo=Ko-?r?F#Ts1LXL{-C+^-z0&X zcjCnH!5p*MY6qoAp~x00#^{2iu#LdITo+Is*atJqkV~F0gI@uCC&zf*^t<(w+VdZY|mu6c_Z+U(=>s9fFtL^vYv0_~~M}3vX^y zf(lpQU|=nC)9YbG(fn0*Dg_@Oc0L2qXd4qi%=24?DSvDgnC6~km$x4y>zsJ)v(Icc zNuw0oDUwXZ^w4EXi|KUWeMlcL{gzymihIWUZ7fzGWbiWXf%mY%x&2M$$dFQH`rD>oP=O}Up{niwi zNcyFB1CD`C8U)cQ16PN&(ed+Zq>vO&Ve^IPvNQ+|^FZO$H;rz1y9kYPUSi8Ud+1r~e^FCZ$87HK*dmULvk0U?L;I2^tgmAddHj z{$aC6sA)#Z^vT*rieIq?>JQB3ohnLklp;r{n0P)i&fuV*Q`TVB%Okf`-+MBo>c|3X zaTBzDKN4LA-kOggbfPg{o$QmPx)jnOO<8D&+jq;>%G#hQO@kMi_rejWxV5Q=TBr)VE+eB|;l7bq^O)O109FVxUn73d4VN`>xJW{XT zIV*h*oM_D`b4-RtwQ+4V+3m#3X0Rb6>bHw11uTAYshAe& zdZ6e*vP#Vz{??Ed>HBn_=y^9BS{n?u8RUnM221Bw34f=s#$rkhbgDvG(*REgo^Zan z1$k-j z$#}pp!l+no+6@rgY|Fl%Exu~0J?^y7u_B7zBiY03kTwYV!Ad-podrt)VUS8y52*`C z7h`ecz0ia5`!pDkWTu@30%i!WJPUmks;TweX`p@4dKe{x)9Ia9{rxavz_r$wdt7A9 zA+#4{#{3|T7jS~il=YwXziHVf&n3*_#M6g_hx3b%|b$ zWor-EdvrWWQ`yh1kNmc5{;Dml`SbmekF(o0rx(_FmzDq8@@Ok-Ld^Effgh1=+#Fjc z&H(^--H2sV4y6Fol0n5lkk&WF5Squb1Z+&Z!WuLc(`n!k*)`+%Yt{2(h0rtIC~E@E z@WhZSEXHv6K=neG3=VZ#UA~law4)XbHj;e~n zQ31X@(pZ!54sshFRcS%)6&VjJKsv24q@6`sh^?O2e1I6!xE?PJ(yaEm0`c8)(P40f z@qA*)CUBas`5abhYNWlswR4ZEP?=9t7m&ypV!}}4oeNREcP3;^CxoHrc5~FNiQwv& z|5WpMy2;pd{j&8>WGy#ik(r4-=7Zw2Pe83*Rp%}L^?DpnZ@Q=63^(ml}c zRmBuYjs|=h{?r|{MkX}c+(v{OHn=@|B3IXxww$t553wSm2dPR4+?()1K%yq*8mNnBx9P>_?RzvrZy&gp#DXj5$ zJNt=fwde}7L6Pm%8geGeA=tMS-=j_yri0_b?^yR($M^R<-;z!wyX?dct(9VI46GOe z{pgmovrp)aVb$|mA@+|m3$`0(sq7ha8*6Ad_KA)CIJ;fpf9uR9ZJQ#FnUDD0E9a4G zqZOw(@9$5ZnVFPNDaBoi+@@kMFQZl7;0*<8uFY#=(*1K>DSF4AFm}Oq-$UXIqi~VBVw9As3`T#uJ~4*2Bf;As5gXvJnIm9}1d0 z*MzMJJ1y^#9St}UpjCE4nn`m?nZa)pG`M0rNG^~qfRzFphJ*`u12)evGy!NmR!6jh z@bEQ|t$)UzkQ+ZElWc)ibU8RbNK=HF+vi_Rxw4i+t=@%>ojN9dw(47wQ4l zupxIBxGVX>XQBANBb=4d2ZZZjkNy<uD4i2;6$y zQ;Ft~_kc$d|4bkzms_MS8hHDE-)2v=6%=MSO~MX0P}J?eb=(KKhpDRYP5e zUVWZ-Ra_yt!|rlz@=O$?nB2BmY3_%4{hmp$4?;qJ=Zwt`E+zczo3wQtp3m2x-RjT{ zPUGObF{_GAzwNOUW^h`oVMVMAYXq7-H;earK+|X`ugATNx7vM`tZ{Nf?MAfO3Rj0> zbCid3g6otw2ipE*i3%%UKBiSc{Fn^NHE1#Nn0`{l>mjHu`M);XG!UG=eCJ(KE@@a% z7iL-#%YU!gO-e>v5<72xfk0c}h#G-vN>NFX3M%FlT@{kTwh7+jcLv`H)++BylSwv| z2PE+`!tx99*I*O`BNWJ)b7@vNbbwB-+2a5LPJ=kn zi>-rTl*P-GZOr#GU;PdZV4Ve5xnGK1zUg$me>*m3bu3=*)(wn~mzk?x>ml^kFWpD) znFB;zxpb;;q8NMDcPY;cPXyioBece?ovkA^f$bpB;20cnz>76&F;QrYdOs(;OwB6z z%ima%s5q_LYK0bMzgNFIkQXo5@_M^`m7iAGJ@@YGdUdm`O^_izC2#X+d5A7f7x*#SGxF*Pa*#&?htcBq)g?=PEenuNYv@X!BBc1{ktPub*Y&w()enkG6bW)HLD4!2QpS~0SpZ~1-atDZ;328Yw7k8cK32&Lbr#7`lX^Ta)m%@MB&yArs4{*}NU(jYjZ{8X9i2Bruq42;|5o+kq` z_!|`t!!XA(g6CGmv}H%wT0DvyihWl9YMQUfw)DLvTTS+Gt9f+ZKoO}iGm-l#1yE!Z zQZW_Zz9xC(UJ1I*rPGs`##t93xmU?xuRsc$wWOF{%MQ9^E!jbKGyCWVlJ@J2}a zB`>gD#f?5Ol|R1t0>f3&~M^pltSR1P%=cKy9Li(ug$+eAr?4Tfbh`@~W?j zP*JQ~HeirEjT7Qlj_US8trFY5su|rJ2Mp)SzIoe+`ejeC|2a<9WNNqP&5Sop*5oNK z^DR>Kxv?f!%&f_IN^zDVr>Ge0NXig)2sZGy&AZH;4@mI5DEkz0z>xo6kSV?#Fc4hJ zT$ByE=oF0tU5Ih7S&et?_iB>X(1{`YnL}>%bQ!zl^;_;oqmJ`7@eWBYss~+mMaMJy zq`i?3rG@SnWld6@TjBgnF>a;%o`TChRwp7&(y#+z%l9eyv&WHufc zA2h>zH}(FT2~h8DN_dTQaRZbS$M#m63Cr|TiboXbp<>$E%RUKi0HE0izI0Hm=`t+9 zE8TK~POEOr)tpwPL+}PT(_wv{N$O{J(|h1ZnX6{`r)0oTtAGEsW{>#DOwvtnj$El8 za6daA`iH9L!C%K=iPpuGyIvtFY)0To@5kg>WJ{=4d5YGYk;e%)cp{Q^g$zNN%Ygez zbvxT2f56umH|s^)rfbeFDHd%D*+({b;+aD}*ny(iA;xoud^Ye8C^aqYMxVMsO>A%; zQ#B_ANIu%6XXKCM+d}fiYhL+P_xNDUuzJxww_Zc;RLT9p=&kddrBV$2? z0oO)A+{4qWj{}RtdU7Weq!Zoam?A%%S+HmFgmjPe1k*pGE+CJ+Gz;@gdUZXGKZ|F? zxpqS`yEWtno53#)xG1|1^G*A#PL<9rQQYr-(EAJ!;C|wF2_y?s*c$1e99fm|`eHiA zr;xXOz7E{#1_9oOzXLE!fc$F{T|1Wp8f^xWjjN)UUp5+KMn7g9NTb*s>fh7O{oqw+ zgUVK62SAlzqoP~Z3q53p7Bg%oEAxU?$#XFWO`^Dm&V(j`Dp6e^R{vvjE|fPQA2KRW zL4#6tV5#stLC$U9SJ=3qS!tqAZUg5AAIkwBqmPc^r~{z$l25b!6es9Roz}YMpk;j~ z7oG(tp2)0}nI2Wu1)w4s00nzL5Mr1Jhsr&z3R|mE*a~Ga5*uQ7ELQ$b4q5K_;AA^2 z#(SlLn*Cc#j-MBoL{1C?D}+Qx1IlifyNX|9tI@mgEK&1IZo9fNOP~2uJX)~ z*k>xS{wW^ZHyvkuAS>O^9XEhkS{T#51Hkrw+t=KkW?DL_JAR`faZbE+N;g|NZKf276j=}C z$!skUC%1+ak#?9}^vYIIMfB}3M~vcF17OoAkM$4-oabalrrymi{&$556Yfv?J|^3_ z<&m7&j2tvWP%fo_GT=-qW)ttJZy(G^U`N7HvA0LECQPr+rPHAF?ux2czDkzPoK)%6 zPec{s59y0+(&-*TvbYgX?iTAvbv zM=q*wkUR!r6e?VH;?m(Jp6!t^fPa*%6PUJJ_yyh@uM(kg7-ox zi{Z{X`bwZysr5M+l_sd5*ZAXh3Aw0lkd=Ae^cr=58pVBJ|CXvTbvAl^S;0O z(5)veMbMl!Ghjv5tRSFJ+9J)9oZ%0-B#N6ubbJ)jL{ZHa>3!Eq<$<}VTLs$@4q5U% z$Hd4juDBg}uOqiP;daW~P5RsJCUaErwQn_%^-k=}Kn?PUcBP$^Vh07I5R>Nrfc_++ zhi?65Gc1j+k~FDaje8BS_a3jGk6FR7#zvg`J z7bbK(y;m$HSErGuW`l5#Qrw|P7Zp=IUF$VKXDzA4d~HxNPq(Bfyo$Gn$@T(0IOM&D z_*S-Rw-*-bCyKkE-WH}O*Zqv9O;F26zGdB4O*v$o1v);?}qG7RlpuaKiR10 zk)=UIeRW8&=m1pqBb^g0Rp@)(n%>BYXh;i39u~LJc*-DVSI_iI;S>AkKJz~&@1}da z$|DEdFF=4l-aFN|Q>E#2(-?X)L(;|l(!Fk)_3$fog6N$tB%S_9R27oUYgS&GRXWpL z7KG!tj$)b4A}@K zcK}5q2;#_MsE7m3bUJNky1dToYx~~o>%6u@`>%anrthV;(`C>RcM%j^5LslCRTfcE z!lDR{jtYnxMO2W04#NzBB7+M5bCRee63vB#7pHIfDct3Ef5DvZJ-_oi%lDjfgRCXx zDRLZ4y#1l)+@LNBC=a2E5<1Y!>pWw48d;r=>8Aq^X$xGz8xnrxm@CK?tl^v?h%r^PUnY#(#6W5t^za1BU2%f!3fk`psu!7@+jZiVx$lOABT0teT(-@X2Fm|&}e01sN%TE7Tyx&?KiIr~Y zx+E+Hn!3ZmV2UNk*H2FY9f4RG@m0^7#Lb%J89I9NYvMw;aE5=XA3h*7ELZDxUH>K7 zG>zmLy=uEDCXpiBsmOHh5$`tA9B`ZWsqgT@^|B5qDmp64dHJTaL*6ewD7xeP5USUX zk`uDQz!)2Q*Pi_AEPd%7o3&3KwdoNn#cjXto8tdCpJW=m{F1CFdyQb8NHNG`U zd*)7{?^PvAqWz-6N<(%+gd&^PN?>1?Jr9U7=RI((q>)c^(iRxD7`G{~M@@czy6BZ3 z8I(tIx8NZ;Z^yQ7zmd4L6w^+TYgA-454{E1bZQWCHA4@*Vh&iF{J4<*kP!#m`)Gh5 zA_2&D$2D4#s>f9J-b&@%dP>Qh+W=WBEIz-FV@92_ z(vDN!uDBQOV8B`a7r*)oiLzsyB^%)^mSQ$hWIYwxCjUZ)exG%3Lv<4Po(}n+P)wQv z8THLu1c}k(C+nCMB&M$T^(WG|4Umv;@OhJ*wPQ&17(wD1#Xzy;B`WekSdYhk*`OE` z9;#*0^Xh_+^YrZ}Dun~`hk@6HyIoVAn!`{1>@skSbcX9kV+pVR&UN7>2A_rc6)?dn zh1=(9!n=8OuA9ATJafJAdd{2{?*{1}M%7N^jXQH{IC$f>|7qBtCo6VHGJS>}Fpq24 zA%@c+Q#1Vbxk6 z(_ZFw>k2f}oqVr%iTvvPjZ7E0NgoW;gpY`~GyB5&LvHgAdJoGq;dkgVW!EbuORjRd zWL1z?p3eEq{Zsj7?ixX(v`21{ur2xG&5Rj0PRz__t;_6iqOKpj__{$8MU?ZqNX?ff z_2Qb5-fEHu}X4YwNiIsq9Wl8hhg0I#!Kv4I^`iZb z`=&acVm2{R3%4(f_qqX+-!4jZtaLffIS#q&i9Q+7@0z{rSqWwQKwp9WmZXz!ed{XO zV#mgDkM8_N!^gIpzeZ^0_9Cqcf^{dabL@$>qWIryQkk3ubH-4zTzeQ z@!Dyp=qg?%9mBc))zx2}1Q1v%cH>3Dwj#&b+1V}3D?b|e$Ff@zXKZ)UgvQR>=8^Ks z=;J5$ON!2Z!TV!bF4bVW+;Q)9>X1bV)`r*c>PUgO(KTbqO2Jj}E%z=FQu05sV9rL2 zwqH)TZ}S*-sBynQ9`ZDhz=*5cp|E(b*jL5_*zqCI7?00A8wX(5WVpUGkw}vr+bw1~ z%wi!7=2onn{;B+;5J~!0PHz`p5MC3t%}28*Tc+n;%n4m-A3RvKl^+@YA;6#0yb1in4GoGw=!aF69Q&tOMT@XszJ7kr>u27miP%Fw@Y_s1@KdQ>ukr4M)&&oSL{ZBx zL~Q14hDuD_Hmk18FIj{P0Cl9*smwEJ#!a!>6^Z>GjsIY|S~pyU+o9XMtKxRDLoyin zVb}?U@xVVXzw|23tJ>|A4tzIx^hQWwPIK;;XF^sNR{3HNb-W4<=75+KLS1E$@R%US z3eW>{`nj7JC<1^#;c>{EEs&nO+Dv$W9&)T~pedJrCHR`Dq5LJ;XPYsrZw6;6OTlZB zAJRELS9CLYgG-|1Bf)lEOChxr_9Bc4xDi+nsOX$O2!T%=cRe)4GRH65x88XC_#kCG zhb_)7)Lgu1nj&Ghj$kvn8)uabQcPin?iaq-g}0se$?k;jD z<|NI?qDx`Ryiat-BOhucE^KGuoBY;%?1+Jtq-C$3eGRf{TptD| z$PfCLa9fyisz@2fbja#F3tXGSQ{MUL8<|vpg!=W`Z^gWN`rj56cB+ebxb*s)$N%N#8%;kf`;X%xsg6+mMg8$e z#D34_@Frn8cfaTP@bNg}ESBU%Fl6zEjAP*H@BQf|Qx;&mHEzxD`XU)xHVx@_&72Nk zl1QOb{DBoM29m6@980-{GcGStr3SSMHwEc;=Ge46Sy9ZXbF`_%Q5fTu`Zxs1>Jjsr zV+@X(DZl^9n~lMS)kx?Vkxq)(WjNSzRS6_Cj9RCvDFzIh!&GG9jO!tl0j0rM{DjSjFnoMdL!ZWuR0mj%KNGd|4r6TgBD+--^U#k zlR%LzROH7({I+wvuW(j7-&u(2dzJk5%Rm<;y(X@tuW&GH$mYs?^cQ8B1oW@c=;v&G z%t`|GyMNelB(U&BCk?3*77AM=n3tleg_Nd39{wzc_JH_%4|3Z$AYJPh?j7z{Q2Z7G ziw$md=(TX(i4J3h+o`$EZ)Lye%_g@q{kpRH1bK%%*0rB|dTAE7KV*Z;8KoK;MaH^f z7enl1QU**`c)uQeBmE)C3ia$R=xGRUdd`ZlA!npq!n+q<#k-pD<9^0tt*I1QV`ofD z!f-yWYd(fwD}`n`AX?ZG!Ue!DB;4=n)~ zWu1^rw2!-4G$4*9yM5c_2YvAq56;(0>OAWpgqP)??ym{Y6KGz)6Vdba6sr6i-4W4m zX1%)V7iy=w->7-_K!inbuuy`}dYM_@vqa{sBxt{;%)$drZ@=ih1WVksba`kkI9(IC zYQ_1*g~3_k9C3~0{Nff#_e*Q%T7)x8WNz6Q%MokYeAXJp3b9j`|J%virob@M;U2vO zNU$C6rAiJf@o(3&8x*@G$Qz|-U?xv$f*C8XmbJcHOXgwq{o8N2Fl*-5fBNr+#mFbE zZGdRLH14(mqc!y=#b_zgPDSpAoCD-jRNZiEkkxqZ_skLNoBdrPgU;Y}EAm|vlof>b z&dC}2emkPRIiQT>mwJ|YfUrJrXGa{HV+ zvnbF`QZsuHYy)&!V66G{ykT);ogE*4)rLom(jWpi@iUE#^BIE{5 z_3M{ufMx%xsMoC(d;-U3pPW-m`ei+^yzQUUi|of5u+96Vh~VhCtGsc6!;**w{tKhK06oKyomkZ)jNxEkLd>fGEvHCUBDL^^0Ifuyud_;;?4m-Z;a*2A z|I&HPzfSy*8rtBwcy=y#m!)i&dBa&z(%_v!fE!;;OMjT4RuZg31 zu{@yT!u?<>Jk=;g9gCDqmx2E4GL+nmWz<^{KhIt!pLKT(5V2YyryM-`jnfhX+Jv`- zza_ELNTHEXQz>RQMG~pV9aOrrY>#g_wjc6l`L8}T@#+E z#JCiO<9GVq<{kITo1Cs+HS zRMb2b%07Y-`gSPHi=Nj5HcPJORkz)t*!ZwEs1|;Xr46hLp79_W)P!~ah%sg}kLWjQ zj$JnGR%fQoTe-`KLlscXqgL;M?kH)q+xeeKk8vLc-VGdasFr}|nnb&vBKyT%{AQ>P z7$hSO>*vIUrp&nwNpm9(`d$ukkSLDqy?Dh&A{nO?jP&&wU*d!>nU}&?Ihw>?-F7)& ze-QDa@AeEY_FFBI4KAu~UQ)pISup`%@$yx>XBTlXDx;ODuFpbK zH%U_Dq&h6!03IOh*SrjgM9at*9IWb%3%Nq@f%6Np>2`Ui6DF2|C6XY&?zYbnZ(u*w zL!LM1jX2bK-gms=wqIPnSjU>bQM6HnDMkC-cF$HB{vbAL${J8N!u_*OSiAtNpg#5J z&i;2^^zPo3!t+p3jkVfcvimw~>ZUYB)D+PD(y+tjCDHCz#lsE_{KKKQd3qAZ)y_!t zh!kprfdgRnRtisq4m)JP5mG$DPfi^~JSypv%8@{1Y((!98)Q6JSk{Ejb3cs509MeU zg5H?==Y(-gTWdGSF)%23Fc5K#U}a^4JBu!n|ZAn||PY#$y{4?XMK< zo}Cb?!tO;^IBB!{x!vMI;W=(Q?7m{>4U$&j;m}NIdlbVdgjN`F+-zvB8YRevfYlX( zmQW_WH)vS)h5KFD!uPoixalT3On(Zy{A8CQmxJ6=Yi!J|iJ{S&hXq{59LfqqQyyfB z_D#g)W4~JfV#ei@>)jzMcScusqF}E}E_Z#{Df0J9IQbALMTfXCyaaih{5GUu*ZZW1 zQiBo&!wz`9HYmppk2i9oeHy?kj>pG2J%NdWI--@}h}+JYf(}limDAR&2(vNaGTw`k z-=hCO49cg(A@wv_!_HS|$H2)nQdQe2CY~a3RAf6p+Pzm+?Q;-lJUwrnb^j$*aSO#O|Z!+X9#SuAbK*+_EPDGNIetDi7Z7KbV3AO)+|6Et&`W5AAQmWrPmzmWnsANRIZx!V&ruBf70%tL7J^*d5-_ z)h^J4AExg>g~Tdehx{P7XdYf0cF5-*^2ehTA8SbbY2jD_15ZzCg)6Mo4fWaJjjzAx z97ni`iHEwXy+J^0ro_aXY`SVuYrw9sWJQ1MsR=y!ud|g8x)7^gReG|&35c@D=^Y%DHM}Lk)2fJ-)_h&Ban%_n$89PeY7A|{;@AKvkYHo^4 zY3xmc?9h1J2)9&5r=p$9Qg#LmLA$TWtwnNq$qlz8 zMajGvcYRX|r238{8Oq84Od!B4%kzs5$S~cYYH|p3EQie_tuZ;vjd|Ft+Ip{Gu3@!t z5GfCmOgmm}jv1{sWfW6Fk$qHTzZ^-%y683Ak4MJLMQZt|ARmyoW4ic)e;7E?GscrQg&<G)mkhY^{*C%wvcryXbHE5UdnpDYczg7?>60Gz=>jJ?uqie}<{orL z7GcibMA~i+S&klBG(B*<;g@= zJ~FRP#!A)`Pr2t{%0a`%Ph!XGlNrXEM(HVmo(u~U@Kq(grNRS?&xrpwB7_z~$jShD z{&W>*zSiSy&>z3sSG$>{Pa`!(uW2d8K$3F-6^VQtSPj3yX&|I4xXuw?)&9UC=U(VJ zj;Z-Y3o;g{XF-+LCZ;sJiWET8`U@VVa#fz-3K??2nhWe!RP|CLw~cgR?R*>5gVlkz z9PI;Zzr7ZI-J`U)*~905hb%_0dt>c3y>@lPx0fz4pr-kUt1?MByEw5O`&BL)f&U|l zsiVj-Dl(OK)NkNr?X2tK1Bj->iT_^uE?vv36>nlL(mlbN@NJOc^u_#Jf>peM1$}d({By-f ztCAJeBpf8=i)+Qbk}W=3e&UQXq|&!cbc!4oO=O9);w$m=yTalXFdh2NwDKED1C~xk z{(UPcv11FS!3Z5kDW-}d61B?0W&O+nyHX@&Bs_hLL%e3`Oz+E&d-)tI7_jOiJ zHeDGI=b;W>zo=(|j(;1%!c@|%&e>_ovj19~WSZq>w|*Bhd28=BEH0Sc2gxSeMK`5d zNh~B`pAJ3~oX*V;8R8!GP37cIF6%EFWUvzL1P8N12IZKLT>2e@dO7s3-ajU%>{z|r zGEy%s6r-WYd357G*#RxXlasP5ocr7&UW=qdzA1E9n5t0)B@D#x=8JWua^t(8@ zmVOua;+KH$vCcD&KryR^!2ylZe9kV=K}Y=tCH+8Lo+~&5)d1>Qmf3P_(Dn3sGtMg( zAdeNgrl_y{{DT+W-~0Zs1Fl{joTGlo;(&HYvZ87_6fU;&Ct`3|@MS+W#9~xD^`vbt zutUY4*8IB5(V%nczf%2ovWlH~Y{xt8Jw_TTo?_xC5<^8E^#yIY^K@V0Oti(( zGP~rtVrGdSi!7Drey9ZlKBEI>{$o{$edxdC)obE@F?v#VNYc5DilVt~fzhs)x$R`@ z^jO!r;4)4zUA6T7TRl+3+D8URJHdX(DPhrXEdK;L6Yuof%R>zGX=`D&$as}Q$B=wZ1bo2Yliaw)LlH%6txV(JME-Ab*P_oC-P#d8znaW5Yg zskE~O$phCOSX}z~XtS#H1b9_$Lvp0CYmuaikfG^-dpm!XRQ0}|lMeakdP$Bqjv8@5 zYF^}+N^;C{OqjlwvaXU~4FmJ*%Y>x<)Whs{K))#ZLhSQs2lVQea+1N$)U{({=BSZQ z*-tT0kx)oQo`!<2{lEZ*T1C|&iJl9YbC}Q&?UxC}UtKib&lOz^KBB~aw1a!nYm;jnZzn_;k$NSO zb9u>0`Y1U8`=}xAZC<6YSaue8H5#PIF^+el=Vs6?A-8#bq{dSd9`6Ock9~K{OL9zd z#pm0&sUm!;#_Ozm9VwM<6V@=`oiN0c#~Lv)o}3dr`|Dbb;$zr{ph?&u1u|f03nfwmdlaro!JID z&qc;_2mEUKwL`NE3QDWIJEF%qgd`eaoFU^5R4lGl9F4*XSB00MZsQ(#~5!*kgX{ z6Gp(8vrhaXE9;ngUo_iissS&n*7zR>62dVO_Uw3H4WzB36v}3b*+`LfRAde(l~)N1 zL!~e-q(1~Xb!~-%XG_cilstO^8{IXJlB0k9QPo6zFm`O>nDN29+aRl0{BDmtF7%Xh zJvS!c67)RTyY!m4#yu`{zc_}|>$YDuD83V-m1Q~Zu$ewF5wr!sF)=<&Ca}@zhur7~ zrs*v&Q~`;Zl$TSaAoz^(fmp|FiKarHuRhbES%DqPG2?quH3%E|+ep-7PHSFqEPpG> z<9CLox~*yJfIk$eII~^mrl5t=wd)!=&h4uAM_7-@xaeA;*c8D zDg+Bte;Ol-65I-IfB?`s-VK@7C5EHrU*;a20}hrL5wgRpXX$Tc(~$DwKLCp@#*VPU z4Kb$bm&b5!1l9vHAz>v!LlLn+eF3&7x|?0s_Oy3M@Uumc;1w!u3Bcqydaod;rbZ8eS!R-o2KGMn;qXf{b0zsH!@&n4>~C z)RF=ZIMln@>VcU(#h+~$(Hf!~Lz-hlqO)HWf9{LUp+Ezs(!4UZkwSK6jUBsg zpp)Pzv!;?_z#n*sicEyn>Y%hV1S3(EpsdoI56{uk%Nt!Q`DmQ<(9pmwV@W2!xKh~Q zhynC3!l_szp-b6H9c~Po7{8gYN_?KrqoUnYfK)&l_^drek^NLWV+b$K{wG ziV0g2b_(02LAL z6-(v(5T={*)vy32c1iy$Kv_L10jW8y%4ch&q|q4 zdE&U__2qzt;EkML@C~cY(u>=Vk@a?L^6oWSKN2ZsJ4NED$a_8+^FX6RUQ~#52p2JqG{-}+m`B3xR5Kt4DST_F?v|m&h6)J zf<9RKhOV|&L*v2}Vr+t#dEVo+CTYDX0je2JQK$`tuD9yYdop-6hnyP&)U)({t!&6@ z-QfdYZ;$+vBzs;5B#J;KDVT+mvqKoySNpv0#R~UXy{`7_1R)>1{N6VWW~pEF{2!3I zX{5vG1-w8p=P1%ZMWU+{>^^VE{kl*0()T&YcwgmrpHnD2;I5Ue0*=ur0n!-Y^>%(y zP_}2g3&I4`%i18*xqNZ5qLW+igATKUiyNR3D)x9+I~|lKNn__7;OB_5vFHoQ_45Re zE&_3>yw5Ee%uC4b579zUX2drNX!pUBH{=`@RvL2DH;rI(4=jLZ^PZ0jSf=ZOrROXF z_ay<=OxI-rP6lATue$KxWUU?Vce0ESzk^~DD6)l$L?4&Ny)_`8LI_vV!*fe~u0X)~ z&a8YYpTZE&2&I7%3kb`C4*GcO+XhQ9Ul>HI9O}x!}FJ0%`EwAQ&9J-e-Q{MMS{{Ow;cIlEW z&xBAVf~vz4e#&k`V5X8kB&m^%I9%miSiH{Xv%pd%q^t;HMH}7q_9%4Em_9~W|hU~i*#ex_Td+X^5&OIDomSQQc_XjgoqXz^a@ zbLQm^#gKEV2*M`!!yaO<40mFImfmSd{}5+%iQ?DEtqF7nlcCvuMam4#l1Jw9I7 z@*nc{(rf*+z&;H<4y3}Rkz&z$62rMOClxwD7myC3FL%T1 zNQhW5^DG&sNG<(}$1i$IKVf|1Sm_-q{z0R|R}4UD`Npj*vVR(BGLlCv#Q@oFB^9|X ztV`KPZ+04-F-UfKU8I`>u5hB~JqYXZ$O_I0j)NjVP55!2 zdM~M@@vJ!+@js>`#7x*UyJv@#`al zGx{gavN3*XJzMpUPexCkVka%TCHvhn=lj1fEk5{?_>;|)CFF^Z_#zp7uAr0K59Nyq z@_u=)Knqq6)@Gj&ALrRv|2)Bv?uZ4zO`f#v>Ap`pQ%y5@%_PSjdi9U{fB1=W6{nQj zAiXK=a4V)EMLN8Om&WarWr0oFOCROkl-9hVd1unTTNYodPEpwQ%X~8Xmqc$2oo~GN zJ;U;oN)5e5&P*ep8ZAFpDW-)Y8Y&X&fMELol|cITskfa}b)GR$(y_q>D8Tfow7E`d zr$*@zxQ%sHA#Ff8|M*^u9D_T!$te{Ub9Gc<8d3;pZ1*`DqKgJ$hzPxMkad|L*U=Nk zh}J;u`%%A53~)8aFRlzY4w>1|C7?sz901FgLWR%5FHBm)j5r{^2ZuB!h^jWYM9^B@Z{9b6JY3~9y3zHr1tIV`uXy)t!4u^I=noUh~ zDsPwj-dWeZwX@a;wl9Q6k^B-jG}U6sMabsVLWc8N-pNo{hRp=PfShVr%RUXS5p!*hOS{D zL{gVWmhFgx7FNNEh3&w+iHSJ20EJmOvC$c3C$Yf|^Jw{&7yt5%M|beuhI~OU-sSa6>`$gYv+;;w1_Xd7UKn|yt zd)EC6>8%A>OanOY4*Fgb*G8PPd4@0>02%2T8-Ql?l)O04?WUREW)iG#2yO__1A36N zht3SjafiY&{^8J#+-RQ@p_BSD;@@xS_c{9CIvL02U&t{^#O&A3ORwl;xCZUv;N!58 zM6=T#cD&+%pLdk@NT8T46xmEg-Y5IGsSv2d%I93CWd*^pJUszKuAqB*zO~6Tp9e)t z;cfG5Jd;0k;kJ4I^`e&#Wh!&LyJUMLwIn8>%)6dmGwqIZEGhC`J8jann5AmWL<+V) zWv9lDx4%1OzCl|oZOhq6a@lDM`_+Twgpsy5Ofd&3vY(3F$oo)UP8EoYMRlH|kKcRv z1)~u&106bd!*aHg~@>k^JTwc8e0*P;q>tugvIi87SI-23I@wnijn5y|jfBm9d=ob6vf9eZU8fD`Qpus*G)%~kLB2qubq!B7vy;^4ep|s z+je!E0H&U+A|^5*@~X!Vrx?~IW%|KCk%ta;8_*=XjMU2}idj#QHB{uFWDih9wFZ>R zK#f!_nv{~Gr@wOxB>eP4lW>KV)|k3yrf92+0S`ZGCGV2BFO99T&j>wf6a!U{y8u0; z$#>hLUio3!hk~7)O0O=LHPbrev783Fnx8wjD=<@;;uXU=uvjbE=!$7}XMv~ro~LcK zg&r0bAnIxJV-gVc%h*G?QUgX#zV)rEWD7ft*zxWcI%bdB{qCWd6pAE)@62zVYrm|> z_oGFL;Ti=pLyb6Kof`Do!@we*4$&D)YLv-}Zr;`aoqN`H`QJt&9v4UeHb#@!;N<0& zUu-h1e>PKQ+#*5N666(8RR9gTKD;{_FPP~++oJE%^r1;OXQ#sLcKyEm*A1IZy(4Bg zO!Vmlm@$jfy_=S?9GWTxnkL;z(T2OB7VUA z4!_B@S&;?Ju%JUQy;;ZjkG<%z*Ep<6$a0CbnOb>0By?BOg)^>)sIc{K7meh+DfA^_ z4X28n_DA+pq{4*HMq;$l-N`KvD_i=0JZ-Ct^NQ(%V)zGngs!c7E>!S#8Jp z)=*D0%JA4mF4POog(MB>mXTPJqsKn-LANOMY&lnBs1R;%gZc8&2r|64QHnc zuQtX1DaZf}*`HOr$v!&<#%Uv9R8b6YSC&(e6`W&1r$STa^zok@HHT-vQ~ug{X#8+P zuAUt`uYB=EP9NPt28125Z62lc780*K0ddI``k^>Q)HmY}w-36!o>T&dOs2R1; zq2_|*f~1CnkxVTguQmtV4Y}=q+Da{@Ss=(#?JzOu7@lT@AS&t?uX;{|XV8w1J@udB7>6BkRc%uOJLRG0e7H-$@47?Zjt1O zfngvCs>=HLn4jGbeDQ_BIpSJ!_+?ytMjVhy6`zR_X(ig=Vh(tO*Ep@>=}RIf;F_}m z+(yH4+|VogcSm;X)i<+k!<4ik*IqudHmI5v9SW55q+9180*iz^Zc&Z@K~_3t ziu>KS{>_vf>gilHJB|gKVT>C1<~8W_yK?$%-WC$e+a_rcCUeg-75w~=O5qI}%ae|h z8eR{5Qnt^E;k@Ta%_7K*defgXoz-rehs+0Qd7`%sMvHud&zt1zG}2>a8(gE9D-^k; zr%Tp#;KoqLhy#+_Uk4r*pf~p1=?A)1qNDDDJXF4xe`yA8Fbp&;I1(>GrEx%uT*rnU z@6{qlI_vd|y66F-VW8!*6S5y<({<38yI;NuI>HaR;JyM|k--K9-C~$}53JbGx8VrX z4ObBz|I5xfUBI$@z`bfwDg03}BJ543rFCrkG!j!j-&WF{M^16nEEnUf3Ez%gO9+!#)cr>e+1TaMe36gn$tk{1EXaq%2MTKP6B zT-dKuyw~~Y4^163W;`2eCQZ~KgS1k&G|?SNb9nmq8h?2+;(YjCdgttu^pop*zOTRD zM@LV?PLvI@b@C>>D|+{#q6P*7@)h z;LO0f;~O$fc$T0iG>-QGG#l!zBfildS2)|}Hizd4PVtWN@D6stYmq0&@64&;G{|oA z@ID)CS`Ms<=@;G8EWpPKSoUi%{Pril#Iy|YOXB%6Q%#9moPFF@A+CDBq=|KHw}Kiu z9r8h^s@W>MgM`d6ykQGgLl$^>=1|rkf97n~yLrhu>Aq={@o9V5ewR~7d>ZZ2dmKyL z8`L0+o3Y=sRCq#`BGPS~vZ09;wy1!H6G+*g=#xV44Z@$XWHVQVgR2%L1;H+hjq)0Y zBv?b?vDzEQhp~9XyuO?IwbbT`q$}7le9WXPG(uTmqR(v_i^TMFWSHrMXU!b7Q-uPL z+G)K(zpTy|`I4Shq<3kS%7>sl1v5QM@n1&tg*4JVt9=blf#Q)PUpw5AQRF4>Q zKZh#kbsjcW31fWc7Rq^yCvAJlyp|Yi8@B&|~08gtHw~ zWUH{9kMwPs$_Dy?d+yvq*bk=$p>ML8gDZKipf7Zz6P#w#n7^=bdX8JdbiCb#7boDo zeD?Qcp&TCT=%)^zJ^i@vvXg_|W-EL(u;tgU82}Q!^`Cx7s_ZyQ)M|v>vlIhv{ExsB z^xf@m4?C2lL!&v*Oi(GwIKCMbpp_Q5een))g}x_`kH zKX^FI}N?(jJKN+Z96(ep-(I1D-;mkb8B1?oy^S|q&- zGsRF>5$~1bIw(0ruMr$k6be=Ar$b^{fX025G#QGxFbx2EX&-XL6tbdxj}JI~z>cv1iQ;nL9dMce{1b zkT&N&>=5tOBJ9p)>X*e*67-yz{F~O z{sQ{zF|}z6r3Pc_)ItJe4*hvEiS6WFne*==N+Z>E@y6j?__>P&3x%nL;ut-^S(beDtU@{dha zb{64hq8~Dj=RU77OdnFhu2?Yy7=iX=@VhmGFFe2Rf2oh&L6n`#@ybiSIm2?Am>r`t5RiupEn(Sm;z=nlj zeS)G%I${OZDZf%S-!LuFd7-woX3BPw96#|uX9c=@(f-8$+e~jgOR|>0YjO+ zFlXQ7m_ovC)r1-Gr@Iuxw)OH7)k)bE4i;;m{Qoy7lPnSc|AO5)#k}~(Ff^ueuXJyJ z(_mQG&SPJl{(svJ|3BM6|Nw|aBymaepMcfWK@=9krp5=bJumT9XDt!}~gOUg0 za_VCtUfs!@QMQqt%3;|VFp;FByAnq@c~ZRt+r*m9*MkN!UWqT3`b60Gb}HONYZ zWxki^?FwoOsN(GOV<8>OzprPHv`os+p2HS**=d`9-SVfjMFwq?BL7J)DPyN?>^LX` z4e3WEmK>*;qZFy4BGGl8E5h2r1bKxYCO5;LxI_LxS}-GAiiPhaKK z00ShpzSHvix$CAQamq1XVXzh&nf8TtDRV{D{CYatuQWJUgdy!Jppq?Fv~K!Qzm2^8 ziYTYGRHkwlGz{w#Cj@oQFMtjm6N9Q{s2i6)nGhJm*ZNd2Zni5>=J3P`HJ6$g@SF>(_;QS>$p>GgaT$WW=F1sMhvd+VRT5 z;xBca16j$!e%<2l|K$g#`G#d|>Ba5G$a*`@c-(8WFeOsVc8bJPk(I&$d6pNJvEQe= zWNiV-3bbx2g;kPz(S9fxXpuw-KBv0_Emr{J#Ke=w*rE2xbGANWg<8t}%;xA9owh5D zY~fw-D3$AYaBAp5q3WP$x7!eRom_=oND&*8hzm;{a{6V$)@T?tKPzBN{Z|p;%r$5Y z2a)n1$z*5z+VNKVn30YtqZsJnw2zAH4;*q%o88V&Rt$vP=0UG7?`GG2?k65Mgw=FZ z$PVsK-$Cah--ipgiH>ucU`LH^y@8jJ=>x}2_n)#0y*PXA|fA817HQ?sZJ1$bv{iVUpx~Zc#P4_8gkRo@f z$aL-z@6PbwR75OaNwv~#!kvmt;M^(#PRP?shqw>Ea*&(D831O;d`TvEr}Ct7C%L(> zmCoU0(W@60`2t#`iJV7gI~5%ZpZy%ptO?)kcEr1l9Gw4~zb*eu$HKbct*-Z^Rh%A| zPB6DGNY>6Dn7?l^CIu%eYL5hfx z!~1CbJxKnt{O?#!bdcLepAoGK&El?JSRn3n#y@(*yE%LdxwYUIRzu%2fH|I9H(Y$r zn7VP;_{CM?3lqrsE-+iTm* zD@zR;Y2os7?L=+I_AhjL9_5%hO))1aQcp!9Cn@H>w)6GO=Sa;A+@#Rl8Jj&rYi4|n z>wIHG`2TFW2DW-13-hUEjnV{pqema<_Z#+8U3c5t$p)&i_j!zUim?%-u6s^fn7K+FB^M19F1e;WW! zr>883*4QJtSwE=$$FgxwhNe+HKOc0~gl8xVWlgTVOB23+_MQK;ah~8~Ua9}U0)tw4 zPaR%E^6l8h0h4BwjZ;A}6MfJ6>gExLBg%r{cDmbjg9}oxV~XI=bQMOD z3IsL0hkUmBbUFPUMKysx&qJ^@Pi&D#dP18Z#TANAS_L146W90D(-i!b;W$8(CP0 z2{uz&Wh^^P%q=(+=VnkT%|AV`jO?^yl>&`WN2!z?ipio#I;xbyd0Hqj-0C{ys@wi_ z(b()mOXKdRLyiG!%yUtd3V|O`q@`OVrBDXbtT+pSGJF;XZFRloWhG@Xh7>Kt%9tA* zxH!g)*15zAD^ouItj`p*s*hE*a$zp6a!?>p8(Oul4FZckz}u2 zXo;K7y+`Wlb`mu+*-OoP;8zORr~&WTEfOV2^4dI8KMtQ5l9z&uu9jdwXFQK~SM&6D zo27WJpD#iFXfB-NrW7;U_1z+#jMqogaa=p8ruU20PFKYbc{s;95&T62W9|A+ucmt; zJ!6I7lZEhn9(}7%mZSqe9`ZpibydfBm9!MgUwJEuJmemYGKM<5;@2HFZw4dcF(#o ztD5fT7XDwc+hYK}bDiuqt%&}Tpu$Z3^lf>QusIk5lX>DMVS}_exR>7R(jRgJIOU_3 z{&vhI;qljH^#jXwi`75+nG+W6?mcfx1NB1HpqOFQERWa>p0;E~S8$!onqoC5=TY$x;5}uLhSmQxH-J-yVro@|9#BS9ArA3-#y6P6}sQC zolK_oV1*fLfWg8!{v{*yk1xIh1&fnZd+9x1<%>}>bib5BYnaItui@LaOv)CD#)`wt zYlg9m&sE2!f8k`%7VoPr{5M(4PFvVHKeCu zs8%1HwdEVk@*95qRqGCH_&u|FRqeM>r{v|%_e@EQ*=$wpScjOQH%{fG2I&jgMjTQ( zz)(74(|pKN08jp8Dh>lOEk(Btfjs8a;d|FM%r@wcRquIhBq`GfMB7K{q`KRn1ri#RF-KAz! z!xz$BikO!$+M5No>3&YkOU>b>^bkMhrSsvoT;s+BkfnGX^I;oaWQA91+o_#DocG8^ z{Yys$**lHY87Y!O6ax&NB~;`!@1!{?0mHHy&he1-0foY0htvL<^qmEGe%e0;t~5ik z;c5Rt=UqV5xpDeVQs8=K>97ptwQ+jo4>r6p?2zrbd-i6p{lZ?4v(k4{d?260yHk!$ zJur-)thi>Sw%7Wv*HQ_Y=#a5bttc^e60%zwV-JoeF-Hd&z-Ug?uoIA5pL?w!PNMSGX8gy=vz|!Lh6JR~^ zI(e#NDW2?gX`*vQ333A=exp=Bqg5;gtHS3os3wjhR`9W3`|m$WLcTxCpm~&=)>V<+ z>~_a?>`Oahq(BQO26{DTQ;~UoXGNF&FAFljpZ2jL$+5{b-+dGG%)CO{`2*yJ{ENW5 zvVQrnY+ym(98Gu)CqGEdoaJ8k)&^ruQ3e4geo#(uEjKUpq#Q~xCXS>rXP7YT$DDQI z7vJ@coNn5%o6Szhe)o6;;4#W_PokKe2=2(wq@}VRIzIwyh#>E+Mba<}Gv4y)eKVj) z&8rzw^mCMnGg>5x|JT2A%KU6`^jijX@|T>$x5x=Q&Q|I)61aw9&Qqj`itM5*`I{Jh zXNLqiln=UGUQz&A9lHIOp88rntBD>YxgxCG(5F0N{UN5jA>;K?Uo6+aWC2_hF|?D( zO(H2??fkfq{t!Hl;bGhG9^r_%7rJ`jmT$xX=Q_sifa@KKc7=LYHvPakO3>%F>?Rb4 zVL@b~1bUa#1&~e2l2&`dy5#Y)X~}`*I)>E}Iz^>iC!Yv6zx}Sk1T6benkSoLGANQp zMIzQxgVxCh3Dy|_0OP%S=uYkmfeNvOb+5WeLAQI8aFF1o76~>~9nsPF_t2HxO%vvF zd8RNYhO1{zJ)xVw``D+#lv3MnH`!)LvI~Oq#ON==>}yOp+Tdae6^Wuwqwhtz-U29p_n9!?4%+S zf{<9~PKbI|g1lyqTCqnmIHOdS$XB)Vqy3Kit_CS~FD&Aml=T5#)Vv(ARd(>t|L(^l z^XJbx@e%BlO!{lFNv0ewPvEYpIL*ZyQvHe1p%MuVHqg(8=z z$W$=@Pe3=hTk|_O-Ml*2=xIlM)l2RMCM(Ke8&NIW$gG~G2``b|f!c2=(x~iMUFTI7~rmp3uja~VzfHP`?v(FuXY7B(#U{&H?S*Q?G!z2SJ0{; zwNs4fdJxWF8+1#4h}SGpl{8;2R^_Gztus>ifnqF{OFu? zm%D+x!YotyEQDsFxyOBKk<85tZH!bB`?Ugse)d07X9L2@zxHMm*)R=)j-wH_i(+ zsgTa{Y6K>_WI$_!q|-?g-ZV2a=tBXLWVZ7&KyB_1J?B;$GQ^GH)$r1|S&qeSYNrYq zy^R|O%>$BW+>%)v;o}ALQ_9YISk}$N3eZ&Nil(oaCSuzyZp+-QuMDboJ^f@=| z{;t|RJ4J+nPwX;-+7J9H=%(B2@}WylU_KYPGX5{=`VWym{jYZ-|Mb(}{Qkcp#WdqU zC~A!LaOy9X?|aQO|HWL=5sS#-#sX) zP`nH5`JP?UK_~*PmJN_f2vh5?H+jZz@;M`cr9w^kQQuTJ{w%O!@d2kUWz}p=xL$yx zM}i?Ns+XyqvOU|RR@3~}kk4rKZ!HnAlaJk|IsEnaU5ib9x@_W(_G_IHGag{I6LN}S zB`R*{vgsjb+jiPD2RKWBZSGlHUSFTDyGV^4JLIkz>9|IU0k`cbDl+%$ z*TkR7x4+Wsc7g_r7VO$yH$5?XoBZ+;Jgeb!2UaUj(PNJq3RA;kWjeNrwPyDe5|Mjx~4AK9!UeYpW6V$?v?cDCzBQR;)|TTpnB1GbfP6F8Li&NbvzC@ zu~UDVRf-M^JLv6ST*pn~!;jRunv8sf^>)sV#2 z32$-5%+CT4oOv5Jm01_2Iw?&P!P{Km+8?5JN);4{ zGnSq#ya OQFt6#`e1u#=Uet z#+3T!>7cm%M%Ew?Z?t!6E5*c8WD^y+I=p~e$&ck_(%U@1>m;wHi+on_(Cajw-P6Rh zg@tNrVjnV&r_Dq7+|K?6Q|cePU9HSm^hcCci+X5tKPMrEU?dR>F#yJh(PK70F^_>U z0t!>sKkVB&-LO>M{C}-~BkRT@FtlS|Ubd08*hw*f>aA2HW`;w;40Z~@E{oW_2HPR6 zbyDTXGl5)NE6IlB3-zqY(;b%MMQ6X-94C)Hz&u{Q`n?oUpaC0cUK!g+A-hT?`&}c+ z$402Bq!=h&Is{FJl~{(b(nkyFdbzI2%6Rbg=?UB+Z3DZ5#(QalH!&4X8Yb5Xi0IBv#R$PMWr%HP3^j4<~CDtqG^_W%g#}Zl= z?T;S8VKzwqmG6knzr)h+ zlg*sg7;FAj9YeL-ug?C@yE1P-gI+1=DPBpk+36Mg{g;vBMtbES#Q?cYF?3i9DPP@f^=lL#2=SziRzvme-^m|a``RaIY zVtGAK;iRRzd0PYU$uU6W<|U**`?Y~b-?`#NoBw`-YRLia-%yYAs+XO6Rq+`Mh?6xLEMPYGb6 z4QHO7&`MX#BV&y9mP&0c=DHZr^0QX*E{U54-8DxGC5>W|DYA=-Jm7v7^7spaRp`OT zldS=&2I1C#L*f?4)Sz5JucOYEl;|x`6}o7%XFlynx6(gz^fcprNk`9Ni(^)zu-}b` z_RZ$+ni8qntsBM+HEJ=Zh0$fQu2A-oJn<3Vrhwc2w;{V19l^F!6=PtF4P?i7 z+-8^9LH3<*^lMEWbWgi&?HAe5b!4=vOQ4u76xmEgW&^LXA?_X*QVH!vRK1QVqDtXr zFSKsQvJ-435SW++R&4oaJT|gQ3C8=J_%Zc1s+Bjkxz-w>KDx8VGPS->iylf z^OXk6<7DLDw~~^vwncUv@@X*A9Y-mqiXs)@j^lSM1jZMV7kb{O#x;x1Uy{bH4Z7}+ z#U|&OE=UbW@0&h@A%|1%8^vpfPBqo^Af$$O$x1^t;T8M>r!sN&%wDH!^S5!UmlSe` z9Xf@D+&Z#HnM`WArEY~>P554y6SAG)4&FoGravV2tU>cQK46W)*qDbC$C-Ixo&WdO zGld2at=o0|mt@nI#w(L&1h3r`lSq;6k64?#T@TJ&&r#`0?_lo=MBteDfZb^lXPhCb zO71}ka1cRoe3DYe#)F)R*o_?6&cYY9l_ z=7&_w%TcJhc?Y@R0}+>r@zOPKFke;*Ryp6|!$DlgeISd$UaX}+>Tp+zp##-HN$sp1 zlJ3ALL60yS%8I+_H1RF>(r|YDqMs#=@!(@jW%Bsqvp&u8d*(IE?LP;;X4)g*OHxT- zNlQknk~3N^5B*5X(86}siH|S&`^=YL^xpLOb2A0yR0UliDyN3z`P?ifdfL<1R!?gV zKOdeTU$^v{c$e2&zcSZys>i)Z*}`DWe9n!EZ8566D;QS z$rvUumE{**Gsj@P4E$KSn(VP-S8cVCJyJq3uy4wzB6pHB(Yml=dFp zGM*tg>wxANA27~a>_F2|!+q0~#DUGiWXIkZGh7szbT!o0LPQaqG8iAq6EwT`18ZHf zqSk5m^jtyp^bv=<(v94H**U+-N~mIwrE!9Toz;>vEm2+ zv&>X`n0f8WzIHnzdtU#My)00fXilf}d2l^b-GoxHqP|33P5JPoMpW7Oj zH8#taVzYu)V~VXo8y#Qyi=3jb7*Voo8t*Qtc3|W5sY$;#P|Qh+9HW4(q@MngDG4|g z@L=}U**et~&s4gQmlDuMKv^YXuj^?r0eGyEJ}kQu)ErR~TJKyefzW_kXT&95d9Y4} z_ccevOwWXIIA4$uh}UHEdS~67xkQlWid<;KvE)yV!_-W1hAge}it3krSpjd+&PKJltz3F<0xs&gOau2h(@-Y3omj|T;$o}1~ zydbOge>i)r!`AxlvkWKJ^z`WNVaJK)$1cgV(aiLI+vhlmc3?9Dqsfr`xpfq?mLh8m zg(NL9Oi$M)1+^=&F1FIG1cHuMz3ZcT(v|_hR=Kv8|5SDWjNqvo|Jl~)SaS-q7s^nu z(3)X3114K0`n0-eVSwPFK7fJpE?R`yd)gDgE?+ex6M~GVLA9 zq+NN#Gj@6xN=0E=J*Meq(}mu)N125vdfwn-IXo=9k{w0wZ;g?77|lq_tLJ}561mN3 z9oS7*Y+`gWCl86Uk=ygN-6`B zdFVm7rf86tgG{d7$0q9mWFZo)yezo)y|0BEVdA@F^-i*v8%!LytmK>tOpZ_t#IkEK zw>!tTL$3pe1tj}k?Sg_vZSSicQ?Nn=O;3yL+@um1vj&_{J6T&9m=zT3kqQL`xLiIe z><$}ns;9B&pi#aO#O3s&CmP^ngxoAZUW|WKcqW}ea9Ne2?4CBPz%9oY%iBf)IP0PG zJo98YZzYn(eW%yyU(J&-9oF^Bl6aX0jAVbi21}tl*~YNNqIkhw-W_GTqR4ZbER!zg z*?wSo=EJtUr=Qu3%}=tAIwQ6D9g^9jWry9^%FEuDN{Jqqsl&2Keq4Mv3-H z7oz!DSU>aVYYVk_Q9@{r?|NX8z9FykZdab6_b9i>OG0aSm!YI6UT_1}&ktnVrf6%` zH+Y!-sZ+&FU+TNCMsUV`;fHN>kG|m>@_mpxy0B7Q>%Xwy$?!E$^He@JPO^?9xcy_Q z7TdWyDOIv$#sx979o8|WGhmNLzf%vWsGM_O=ArEdSK%d+E5bXf{EHO}^~H6W8iyq% zd9|>7m3a_q_-xsMhc1@D)S!kNq}#1=Z*u?qUv#QOpVMC2I*4UIq%Y3fs>ZVY+u+w} zk;MznkCC~u+)6zee{61%Mrx4VN=1Et!33GnB8l&ce@)^Yc!`~FVq?-M28dXbsmN== zb&7+u;qT^%M)7CLCHzWp7YHrySK}vo+7ihzP0WOB{(kkc2^+&~lA=4(gEl>81suj? z_00~4kp`_#dp>=~Tp@C}331>>5{qhNpE4g*`1Y$X|2tLEBaHRfJ6XHj{i^u7Qj343 z1ub`9?RHpG1xunv`FTMlT}8BbQyIYp)o!m`_gIhOfIL24gp=Sq^a^VG`w(oP zo37!ZAYd2$B*bKju3>xD!0m}0vT9@4PZ;q4JKO@?lQx)-P7XUyW*LdnoRQ_DAiM=x zWA*IM*^qzN%oruN&(1xM`7*ecLe1_`EOthGwrmHjzye!SHa*Ucg%M28xOY8@BcIPM z5P8FS_kW_PYx{xK27p)F9l$&z5Cr41}#a>HZMCqprcZLzbbrISciJ^QbWb z1Sc%$$Bk$Ap{>rcV_^i_G2Xwl>sQ{!neAid_)4OCY1Y~YOyxY7G=z3IqHkl z(52#b=Oc>69@_Pi-7eSXJcRBY8;fE^j}LZ=T|YS*lpiEE{kzeiZ20TjUm@olc&U8P z#6(=Dm}?Zdf*yaQ_im9P=|YqUtLXK)2`#dR{BE+==S#0v@xCxDw+2R@**Uy4K{lN% zJ4Y(TSZA$MHOf)ez7wkOafQ-G?^mCaW(DE7F8Z*4o}Oz6a$$mdfe1QRG~k4VZO3Gd zn1oj&fRZ+d6U0I7zM-*Q{4U`6G5n%?3WO=T2vdTkj&!7?#iBSCQK4u>T!8k$*@ZdP)mg zsmz5H_OpQVjJ_M;l#O9F!Z%oVXv%NBC1FMjwJ}?fMs_=}g=#XfP<0e@kRsJoWF1)& zkobOsUg0gpr<>lxPYF9p;w1U%%_Mhv7C*-u%VSTeEy4-0Bk4Ze-PpI1e@h@Pxh&31OGTL$9oC*SI|cj7x*cx#_)@7%`P3K zvZRfnG1LxOhY90doGcLajfIaA%}GPJSReWJxa{0xr5c)`HC~{sdSe8sfw&c)TrQM!OdUfX+BaJjZMVypBy0RBawzU zuv-{7LWU&l6;TX?7j{yS;0z8xhi?;zW@mx(7e=Hq#T_G)VK$A!4UA1#WqV7o+C`_T z@vfS>dD_72*1)@=gOrKMyaq+3xFztB6kLCJM>AuGxwf^)7;02-+k(#KbP2ny5uIz_ zATVNP#^!GwA&_bhrb7;enh zZ-;8;t?HYzN)_>-AT&k~n8TAbzJE4*c+YHz{bd}%Y38_|^xvhukz^hT;4&j|+q2`rw+HrOo9a`oNU_ow*Q&%J98o8;rR;=ykudK8@?IT2<-2W0EJH9 zZh5jS*Y5R*Bld?SC>ilaJ0E06NlDW#@2?sSM%~XZeMSyVAZ;e^WfR3TQsmU2!?aP3 zrN&TbA?#4BA|3R7>GIdt^6_mQaEcKm%Z`dt>2i9BACL&rQ31J9^pN%djkJiZR#%3DS-7TuGYC<{-R$zzN-{b?P`tHr)wDAgDJQ$NNlub0)s; zP4Zg*TuA}B$l%#?lU6|*2{aHbs^zLaV@XdLCnux);MTS5JvG7{#c@}Z_k)+l`5IyM zn}S1Y$#!n58wZ}F)R?>-B@|OgkzG_|n-DbwRxS8~M2Xk%tLZIj^pY6%nj;>#uJt+Q znJjFFa)BcLGS|EGS}6Ok^F9weRn@{`&(oR)dB0$-Pn&SNTb!WD{erk(anAc%P&VDb zxBU=jBk*%Wwgp%B@9(cM*W2SVvvu6}flY}oT9kveYpGz(aGsMZInB=ogQ)|-3fp)4 zb9%s1R6M5{o1bn-T(Qtx`p983<5)DA&d$vbMsFH~^U%$XDUx;GaS#Wr;*F(nnxStj z?P!~O!=3{TPGjGAp_n>5ngS9BHWn-eB$*o2GeXyUsw9pF^Pa<@P=4gMWg6&MD@w#L z`Ngnd6R^n$acT5`tN>oSK7P3bHQ2VO9!aA_O~B7^Lz?CS)ggweT8tuM;>(9SMcWuG zIXTU5S8d{ztNYwg?50O~oH_CeM#GaPK5(g;s-JEpgf_}C@^k~{+?9b?=?0V`H9VxZ z(IPumG7m*~np7vmhF=>qy%jK4XVhcY-P#Q0AEQXDvezsDr6*J6B*a%N`H?L#nI zJUjba|7D!ijDP%tX{2KUSzbDv_mDe^fL*`#XZ#Y{gY(_#`qCXI|LEiyC!sOg>U z_XSi7b^uuw7HI6Gi|Bawx?s$H#?Pv$v2IYNqu4zc9}$)b=~P9$I_JT2H6YuH_u*_= zr&=cHP-qMIt?;PN(IpeFl8P{}G5_%FQLIw5^R~MUIJGGIK<9WsT`>jqguwaAhu=Z3 z@Q=nI4+;=vYC2%}>7p^7^MIUm)9&+2Qo@8DT!{Fcx1ozV=nlVX`l=GU7#)i}zydU6 zd~&D`8)CIM85GC02YKrB@)<_lee`kr0kXq^ojP?U)8oAqQ$~?uD)P2G(WjhVEP^6y zegP!(_wd_P#ghPmgEUq*fQ2;ymUBtUB2UzHOj2(31XUlywXsaMyp8UbV+xM8etw^8 zk8sr-NbLf?X+!1qR`))bN11?f$ct;urj0Re%}*Hd9&VV{z4lkhXne(v`%#LXanhJKD7EJIr0 zgqv~Sc=rP*a}qO$wJt1#Wmp2Rl!vnsTv&C1F6XHQSHEh1%Q57k8#gV&#PHkfeuNt) z>bFI%G|zc)SWK{Fu^1NN#dB6pOmX?pWT%S<^&|M#SOUZ%U<{fByTjq>7jOc`_$!x$ zzEg~Yj3#zv71{dIgbVhY3`x5v2BO~CRAlMQCCYNDPP9u}BPf&hgeJPCgY?DIqfHT~ zy;6d%M>OjNFKPsduOg9EF%R^$)R4Z*>ry83j=y^)H6+)8=dMRg zV6u;5_E4mBFzT*%@vVYbL9V1t*iE+gDcyYls!e{|MmUD2J?J`mn2<|C9MLrlh73l zfjykI?APo9?GZFe(IQN#g6Yc%RkBB;dnLUrFkieSFjsQS zYQW!88QAC^%j}wZ>Qy`scNlb$639uFA>M#DU^7suQk+_*wKJIePzwJEa_tNd&Xj)+mO}2wJP2m5L^tA0}olxfDhmXn4{`| zmJG>S{y%(%;PSVBe`lAuTFpxW2o`OdLYGdB!Hz)s-8Ij9A!poABiUuuoDWTM8;3k3 zdu>``Qw(E*y?(*S?FRiQXNfsWg2OuPSr`%4D;J9jKzl}~g4{Ipnq+A9`N1R=XCrug zS8$E@19F_CPfeY(*KemfQF4x8aavqp9Pf@tvJ8|X65J|%l6gI$<#VsO_wcLqB;8Mj z{E?rWHAgK+!L#QF)_nJ~?%_1DjO!2m!G8z`J^1H09we(?FlLP%Cex2(irGk!L@F{@ zk`RhS7)xDmLasJ2j6u3B&P+jlT$L+9NfFG%wF4~JyM`NJ*xRx78Fs*YUiD_;Y$G_* zeKIzXdbeP(DdT zy?>|9JdDg`n&QA!iY!6rHerr36B4Np!+IDsL)wIo-l(Q4Tnxno)x49Cg`En4!t?4g z?)h|<50oBwHqtxbicjtD_O%vNql3)z?>^mouz!-#&UDP1eVL?i^T;`H5_P$WRoY20 zFkWR+kw>JxtfJdVZYF1&P>{ASCdx<_qX2Le%1@gcZvnB^mXtRP2LIUM#WtC?i z6o+KfSRjO@L3SI%tVPRHbMyS3x8{md`yb>58qu;csk@Nunm~@5Jmw0D*+-E*ROCmC z(wuLAI#?`oBRB{7vCt1?P!Yq@hE%#FEMJV%0xYZV;ipfH@js=>rVSLfs2BF}BBUL? zrYKbZMZcBlUJ$NRRfpkqO}s-;XOTT)A2`Xh&EV?5ymu_%8SP4S^bh`N7np2W^R$QD z!^!qL@H?qn`eP>}e)7Kj<;Nt-foC;36JW||)Awa88BHjbOe3bjuCe7-ALNL3nVc5FdVi4KE zB(VJ^JihPLq4(w*=ccark0g;j6UbSUDPTRt0I_`y6{!m^4lwGhAUy$ETwH{-$kL!} zdx`gX1rXEc8A%N8&AP5ceFVt-cqJbq`Vd0d?Ug$5;=DT$Q!k&}tk4SpWzWc-{CRMr zTnF=nHD1~cf=ALqHP-eX)L_}CF-VWAoQhzm@GIb_@i6(>%J7TjT`<%_bJ~8^lV&}~ z={m|0xU(dgn@`1Y|7GCS9h&8oOEJJ|nn6W2D3$<|M!jE?6lKLz zCHkU|WkC&!78x>yEYkA=krI$>IN-l52nyGenPR_qqFt@rBVUxH#5glX2Xv~epc?N* zb{W=efq7Qoj^e}&@`I1eDUlsEa>&Bf+@M$w1zf-Ybr6W(u1>o;ZLfMSFuQ20=9GZm zN1vcbt;j&YmABLrSwEV@H3xRNu_Uf_(kDRa`J~&TMT=7D6RHwLLg?6-5z{`k2oR>X zjLF`|Kjpt~&QQQ*-Rr;+VHQ4vUEo9e=o%yx#056ERzvhDlio0?R9dQ7O$r1@=z9By zeN8&nB48L+Stj?{d*{~r*v{V>XBcT;)K4I{9e9SZ#)OoxpJE-A%j*e04x&=kVcm+Iylt|v9-A!2ioNZP zS^Ml*$&T;+wc2QyW~BZ-fn+(bVFK6SkOHvX6a#gJg;eAYxAM8AzGvqzI=J9dX{REK z-xJy#vBNEncaE(1#+CVTJj`PQwp5?(Zq>iQy-d+|x6fX0f-_AK*I)q4pMOnx2Y_@I z_)_q$UhyMo&4M`Ip9Y(=DUK`#i3JwIte;1<%MOx}T>KG#lKjy;N1r9R-4Mo)7D2fK zu0-p|;0o2IkJX}oyM@o*qPA>#5vOPGxQj_%%+YjWv?j$)X(!2Y2eu}eCe~yl#UxTB zfr>PgVcc*|mhYfXIX47Cr5&z?vn2)m_*b-;TWxhvz$(O82w9`Cu<9$Ux``WPMHR1> zj3y%Jz)NM8h+q#%gBdDv04@vaQKC~G!&lJ(+Md~cV=6X=y=yUutb&5|H@(}M<@ELYA7uT3x073#n0YZnAx;anW9|pe%S}sW3Ex* zcRVh5W@>O5(?Mg|Fy7sxIPLXFc8a7q7xT`^&dE+fFx9Z|+9Z4Ii`B+WK-Ql@R!vI? zyWrUkxuU&_IEi6V6fMFu&n_AlT2Z1X5hhomfc=Q0wZ*`)U@~LvB_qL)-8$8@ER6Cr zLaynjyFVnG9e9ia60adM-W-a_qR4hs);$5Xq-vth6H3b!Rj4Rzdyb;0OxF|0Y89bm`B2$S6LM+!eU zVxF+zu#rQSOp1*>L(L(iaig>y27&bnl1a)EclZs+2dd85Gb;JG4Lk^$ni{Yu$bMW#)(2*3 z&H))elHabN8s+)Hl|C!1Mce3*WQk3&4qTj&G_HBzAJp%SR!5ns6p_zgnuLW|6L;YN z#XO?OmsI2pQs$dIqdIWE>X_oO^aAgsSc|+E@q%*FDbJN)E-scYAhv)SnKB%216N%J zUE+Gmb-*b@)2>XP+9pKqK}1r6vrcsL0jG6A zv0-=`zxukeR)E8Ai|pb&41}Ff?bpOX)rC$~<-b?j;*9^O-Cp}OS{SKiD^(8bFMM~^ z?r)hBNO6fxIPeUdg>I^Y#t^tpRU(cSA*q!1fct&<8X!<A8L|C_VwVIqynRv;de=!3ia)iIMQ#qA*c zk?y*0ntm|rS7h@lA-EA0k_rCg!Ap*l!xC+}D+rQGDg&?2+{@FBI4QR5L7QAP8Z?YJ z#+kHk{SK#5Zu~2U0#>;gvGbAk-2adj+(tPEwlOgA9x@Z(L@`MeSw}^pyeu;38l6Ni zd(zXow8+krDt{=++u{*B)6Ry(v~#%u$@E@(?!Di7tx#@+#<6d_`zcw+4H}N?^n?(^ zP-twU7|?^*V#vrs@{XvG7FnMwuHI22B$aMv?z%v>aV%3!HwbmA;u-0F>25bdih(M3 zw^yHQn#5K%g}pnuf#ivOVQ$6PPn@zfe(178dR_? zhu^N;!^3OJ=@>64O`^+u_rNta!;X~@d4d)54I3e39Q%Ou%k~L>9!>I%16vrDBH?k4(D`l+pfHiLsD-Nywomb7Q9Yj%Dvnm^w-^t^ZY zyM^SyOEdIbG#PqMP|Q(^)KQT**sKpc4DRH6A=#6c%C)ybK68zk4xXDqGG_fO<$&)h z9x93IRF`-Uc^KN#S1lQ06$aw=0jH{|=O*DL&6@rYBool0ZUCx}>Qrl(1_fS*lb)l( z?l65#!eZ}Yb(}|u2NGds)6pV?qE#~>E^sopHkR*m2igDq_{Tps&jlVnNI9@IVafP4 za7J8~_PN3e2puor3k-VX{s>r3$JC5qk67CG+nz_4v8*DT(V(0nWNg+2Q zRv8zpPyOTn?y=#2k&2&+NhxhVppqv`yLsU_Z@c%F@bX;85z?zRcUSla-KmV(Q2BZ60s4p?~iABb0&(g>lA3H}$z25t~?;MH@f z352pJ2GaS`fVvx)jUhbhljdCNty6W$`;@RS(y6fa5)WPwCjeC<9j%9XxCHmL^4(uH_8lbSZI{Gue`Q-;cm^?V4+4=Kdk}ceT;=r~Bn2m2Bx;qxA_t1|8&%bcm>GHGI9(;!&TmvxY)*mj0~+9SHEAG%gnGyRvu8{iwM-jKs^i5 zWo-h%YL713$|~rQe~(2+Eq$04HD<$$*k=GE1dtuqA}{0A`O?|RCq|2Nbb@&3u zfd>&5veXKug4w_?l^XKJ>*3y0@Btmi6}W`BC3W_&;r51F8wh9ZxJ>N#@<5 z5A%y`F;&HW;tG>NxUhddv-d4N$!-b5|KxGhe177vX)`P{4oKMyNmABB0nu{z?b3E$ zKd^|n_bSWhCIYLe7DvS<)z+!7Qr;=r#uzSHH79#Ulj@Avcp)mz^v>#@)}tsO6#~1M z6`MfA;tsd`LUw43Fi4L7v(vbWf3DDwq3*0rhN7+9Fj~|ZT<6?H$2+%(u~Z^v z`qBX8hDWd60pA>{eRw-e`qfHEnY3Y#o#Nk<|9sI)3t|MA>ze`DN04bk17$~&o}_T4 zGR?Wf8z$Y|o~z|hS^e}syDe1=J!EBH!@y#X9)8*;C-X5*N78>6Zd?ufE?K>k?0vy< z$j_Oqd5=)cVT#m3SsO2omrY+KI=5T1A>ZvG$>(j5SJSIV2VDj;lOBEtT@c;^*=m@| zZLlAh{+&(ZjTLU^)gLygiaq0an8?-xtTC7td_%queD_)e;B|gfc&?-eMLQE9F{VM- z1>8YsX!N^TLACy<&X*8omG55Pm)66Jed=Oz<|$%0ZL!1a(Ju~$n!EP7uC{bNOYF1@DeCI)e^?I~Jpe zNCJ}(n&WFI4e4|_Lt}RAR>-pJpc~~SkR6NuB14|6-LYdKP)uxvVaN917}_HLv(m@o zjU&<0WUM9(*t1s!``i|P3d@&~W~F9g}}r(5zUhyD}CYMb3g zkFbJpTsPa{?Gt|^HX>qW%9RgD>`UXR-)REutrU|?k&RSjm4OiEW296*Eh(Yby_P)# zd6PFxjS|Jr-9c~lIT^4ste#wzH%zS&)YwyabqH21fw`Hv7;?A$HV*}_P9M#L!*P)h z%Dh7(IZt@Sjx#CYT`rZpljO2k3lmWkUhPn9d<}CW?Y!I>){PdUV%VMbJ+R=vV_*BO z(P|u@ANa523^%Lcz^J%mf{HeZxlEBuROGJUctIDWzaA$meYK79OnNV~Z{B%D-&dAc|4 z@yO-R&v?FEPg#um8+e@WKFEXDii^hciXy{}(+=l&a|uMd1vdm0=o zs8y9Fxb}^mzDVg_4&|~BTtR!@<$Q1}u!f~hYztaNR?4-Qkbt?5KmiXmf}mvu#34z_ z>#r|)qjy2^*Ean})xVvZl+J%3+cssusRxP>>mhxz%_WM`skZa;LJxVN9l;sXF!s;^wDSp}9VCnH zRNQ*y)+;NfuYfw;R7r_AY2pZe9IGK`6WjE(5pzU-vpYS^z8=8e&2ff%lQUAt0SN^k{22ePPxLcTe7Z zE{QXO^|{tqZGX!hoiAD}WbRVdge56KCm0r4a0rW$PpZmxTcwown%ZN4BZ~C;XEQvcM?6h!bf;Qhx?P8qGLC zzZ}nx&3elU)YujtEUmGl=JRjO=2K zks#*N;B-lhCf)U_5-ANf$uj-M1__qA8=uY#V2h>iY<-v=3nRE2sc21CG-F(!yFTo_ zNH)GS>+>>`^?5eMK(7+8!aDVtO`Ug?>8TmBeG&p>_WBrLyo``9G~#()GXY|K!Y*&0qy+KW{A+!&32 z&w)oF7XMx*ANW**w>WECUGU@KY;BrzWAKJa>5{P;Z=O7Si-H?^7(BgYOtzZW{_*l} zjrL>3%DB0t%Yi92R-4fKJf@gFihMyu;=&bk>>I?Lvb8=jez8HxF#G7BG3yNr|ElJ| zo8QE1l%or-1Y}$fDPrAf)x9)|Ca&SgR|%#?A=6!xs*An{L@(>-_Jj^N-BSSnc;FEr zWys<;gZOy6vY7WNxbfEdeCf4cy^DY}SQC6%58se4(dX7XYqJKVE$EbR?Ni=yTs>ji zNbHa++6u17s;MbH+9R~XlKXHPzs5<<-Z|}N#NW?a$wHFAZTxazt5jezex*|kNDQX{ zVV)mwOp(jdJ?d^K9O?p|mmV0S&WO8cJrAO$GBDRC*6kcY*0KggL3p&N(EHhLU0WFn zt41zcefcxm**PKRv+o%(^6it<+vF2&7;#`9%v}?VT%(vP6ltX*S4>Fp>83Bvi-LH3 zTF_~)qvW2RY@yh*1(-Xph`Y(H*#pw^vUNch=dF4r-v6{$zhIrq9p%QEJLw{Nsk|Yu zIpW&vdb&8QQ*(;cLsC|Uq8>B^layb;XLCfiG>3QIr)ug$eu+2mZ>lz{J7!f(O;Rp* z>EpM;y*a$Q{%xw2f@=!2pBMSH)1c@Y$R3J4nR-tU@49U9fXW+|ls`j04@jQ$uYkBqj??~bVQf$&a859FE4ILG^UYVmj zJilT}_ zXSX`~vz*2S$2IE*ZkBz|JOj&N7vwCpu5H4D^OsP$3z{Qt1vChoCg%hkC+Snu0xm&J z>WcIsx#1ZvxBJTFTVfC{KwlcfPoKfg83%@H+%M9B?O;kG)+?#9g2>{t#A)BEhb#|;>( zzD1vzZX9TwB$_=Wa{@VNGSHM!42(PlAU*7r=={5t%FmR)$F1gw5^=2i8Bl-U=ea|;O?!-6_Yg;YNw?H0!i7fMEI7aMAXzn1ae)I*c6XS7JeguPQY4Xz zygRXI?g3#mIUvNs`pqhYYO-Gjy~QWl?=HPXmCeUF?J-kMwW+~ zr$#s`kO@n&18TUP=6#_aP_~1IcQ5D!WYepaCrJ-q`-ywQlp}PtAbUoOtVFiNyJ5Ih1^ zrv~OjjS3z~1>v@bkcx#rz@P`e1b7QtWZNRZ6+NrQdj!~WgV%G7rIi;9M;bf4MhHEO z`$_sQzG*JIr4iuy%Yl>r(!Dd;TsG80pZF%`kq(C)EZv>oa0PW@Q4SKl*xXJ zsR~dlFNI1S8uxbP zy$}qp=kc4p9&r@GwK1}uHBt?S&ohp2L)QE5!dY({k@f4d$=@aiN18c1aH{%M6SRFo zF%1+sX{g-l0va+*v(dIeqB*EeXsT(XNrXuEp`Z@Mnpy37!ZJJrSAh*e-C4Z;AunH? zp~0vY6o2RNPN6$! znx1LX(56m>$$|swiYe_%)Un`V;l>bYocu)WNXzL~_tMQf5!_9{sF*f|mI2=hTJ#f-|W$&;fAImBpt;BU8MZ5-r6j?#2cCGDE z+@FmMi93K+v3Lds-fzwt%ghUo-OZuHMt3Q@Nf~LG^-h+@zrAG~yuNt*t3M^R4m>S_ z@Z}KB#4{9gnj$AqgK!%RA6q=Crlv~Lg7h+`S3nD{hks0_JqFCJnGgxbb@x7)=sfn) zV$2LL)@kyFFu(G>Jut%0h<%;e&AE8@7EDWn>awRDH9_xR`TQ9}5+<8D; z;iA7?Rp`>GF~~A~2nsyPOo-HhtJI~DcNTqg% zMRd=}2&%j|^T>3n9UF zyWNg+g2ecW$ig>*jF3?LS-X`KI4~qmnn0q8Vh&KGoQh2JX^|EC#go;_Jy0~&peO;M zHz-vq(WC`cP2CZUg}nn#TU3~Lb#Vq*i*_Z_%7B8B_!oZZY+2yCXPV0xHnBiAzZ-hk09N-TfmPp`DlUOc^FjJ7y|V*H!moyolDJbG20QG!f*aS$q7 z*YHy5Wg*@2c)@1>1|e3ooc1~rHn!-n5E*DL#t0LJui=CYN*(rxfGNhgh$ePr71`>* zbCLZfbCF#XlS`3oDiW)m+Lb`js=2F-g(<|XfE47rGy2>xXFQ2tp+b7A?ywR?4?mXC zE1eiNyP#|iZOsZ7ZLPWzTrH1h+h^s>0WW5=iy^yqKoTdsjAwF}wYeMd@|~Kk=gE2p z#!Hb2Uba&V#3r{=kp>qku4XW8RKEs}lQhcF?|KV-_=!FPPPwle)PDB^H~1;dBWbfL zNx9n1uHxy&w>OC*Unmlr#=^U~|-a|2v`B_9oRs;B4WP=!!=mRc3;smZNv;G?vIn(9gYf$wRgAjDdQ0H*zNHt z((g2aPv3!aG`E-xBykk8nj$Nx$d&J)JEU5MGH5&8j;Q|^>;MAf`Q6S9Mz*?|eID}8 zPxAiYY6L|2Tko79s~vb|3IbO{W)UeAvx$Ny9;ugLf*M*_y?mxl@F?}(Cev$n)45Cl1DY1`L4G^0L7~OSf3#@*q_MXV7J}lL9cYh}oIo+|J@@b3FfT-OSWvK3NcPd~ ziiB4yTvAk-q8-xJP_BAKna$f1a8LGlCg>aK&u)~|dmr>hBFti5ijU#SyRKE9cE^o{ zFd1_L!U-nSKbCrb@S=xs&xkhzi(I9;OrkyJvCdDYDpTCylr?6nx78v@*y_tTw8IG! zGFjF(+00*>i zxpex{pLxFM#4n9@z131 z$P!5+xhme^Qto|0gLDY(%Ey8@s2Cb>+Thkex2up-IeP|BZU82b+!X&{HWD`9`)bFO z7Fi$AT6QW9f`|<&IL1k|owFO|m0~RU$CFyTs!_gEx?f$X?h^sPFk09I{cV1|1#z&S+6oz%CFy)$VXx99}1+N!rWsJi{1RW|N-LfSpd=OaIHa&E>PWxYQij zuX4_0{CXm+b>NN5(jrNXV3&LU)DGdMx#s&wb=N68PS!$Q37(6lt|>v!0*OZc-mfU1 z4=i^;sksT*88*$05uN9C(}S()RNFKgCOsA;198k9mv+Tw%?c>oZ-bO4}(rI^uqky1gP#Q9vaoV&VXYVbFoA{#V78qDmkw|91sTTN{6GAKLy{^rQ5^qd( z#DL%^I?jqm%b~BWXx;i-xJ{Dh<*eQ-GftAlcg4RZ@e@eC$=jJmF(8$c3~9^$eX~kJ z%O<8nupVg>9*{2O6z!EP;gO9cjBOIYGj47oZ>g zG5kyO@g&p4YHXyKM2aL(k!ipVSpxiD*>tOTUsyKiAH+LrkCGx!WTV~fg&9F(iV#bX zF=!=h*YD47i4!s$w>UHTeXp|DjN{3VZk|aY`?xt>9QR*FnoR(HjAH63atO2Rkrr&1 z3uqG|Zx~KXl3Z2_6Mar&)mIG<**hNjM+X$TfKU?9sZIi&0Fc&#-K$XJlcdBsj83&h zR_LzY`bz*-$lI z$F`)1zRxobc?q(!zqObZY0g@dv|8h`#-&LGG~S04$lttV#szUUUGEQky^V5YY>5(eK@l^s@6rYQG+w5s zwn_fwd^;Fd+rY)Nt8KA&;q%Y$yyzTrdHim=cUJn;PQ_`jT*>X(U3C5YKG!H9LH6#D zJs>CD?9iXjc=jB+`Wda*{uH;-vh3uaw9dvsss3N8|3;QNaN5B(lVK*2ViG74M@1$n z&x5wi->(7tQnKI2=>!xS+m!{1Q?C~B>YV@o+SdRvZm?ML!FN6}4-h$QXqhGCQwX5w zRsz4V{D@-tdwFkOlQ#py;W@JGjhJs-nUC}+1^jAygKN_qJX-i>{5yE8osMB}|J3X8 zZ@0o#_B1LSOvvy#h3zfJlNLE)g_@W;CCz+nc}b=sENrK^KD!R+%AbEtSv~*0q7BBP z484YtwrzGixiOXugwSamVDzMm$718i+aGto=*2miFlkJs^JKe7JLwQM3Zqpho z0k(nkYm*Agaj*ioQQpJ9Icv9|xhuw8?TkextQvXoR58pQPZYw)5#5^CHLpDE5=;eAx*Lyd z52|pbXj z(ox}QpxJ>M+_}6@eOq7-o9KgFY(?`+#0OlefVal*$G*1B*rnDW#MWpSsXa~z8MnRo zuZPAPU1>{~2OK5?PEf8rbl$g?V%AV(B^8P5hj_tVSERbzuhDLT#Jw?j(B}V0OCG%W zJ!7%Y?ZG=N`R|OZ%fViQ5x>*p&XQzq5kJQrHOXERVCPayHbpY1$Suxm1lj65*X6EH zx0)huFKqrr!8@P6+4i-w5lPCW@+i?7K^Je`Yc;_rkN8+x8CU_yXAO$o0ry;Y@e9IR z7(2cVmSEm|BK36Ne&|10y>+2^x)GP9jsvfLS+b^3Q)&%A#;+9enYFdkTE%5#%j7Zw z;ol-p{jsa$mK*Y)pg$j-Fa;is!m;q@TL}^SVZ+$+rTUv~|7o-*?f?4Td!&V%J#k=& zd}(4&Zc)q)igaK_X$jPi8}#9Uw=sM2QV@a1BvG{-GpnHr?!!O&U@Qf?<(>=5zo4v}q^$DaD@CuHPG5TQA@-}Gju{FvcB=D$$FX8c zj&uxsa#rGCc*oiJ2>WF0Tk{WT#YTIyGUdt#B-Vk0$2(0d%vOp?rpQKQQ|h1x)Lp>X zxWrG30?}8=KFuR{NG<7e>!9z6s|Zpl?G&X>%kzp7ZJ0C$K!E*Kcn>x^HTOJH)D(l~IE5{qEw;n4jWB;1M?CCjzl(kM}rG~XkJ;NeV7r=ooBb!9EA zt(WpzU19`@;NaDk1GDU1C8o#W;9}?5jWw_u99>4lgN+WVKlt<4UUYf`9P_It>O8fH zK0DRgbg~X=XtHM{dz>2MXfSk%rQK|!-m|zFPS#^Ql_H&G?oV-8j~k1lrvr%O2YJUo zG-x}G%>ZK7Cl(>X?4GgM<%Wppc~j%&8Rr{K-(Q+Z${l#Vao%LUag<^}MC>5WH!z`~ zM;SAHeIV2uBlUPTohrEm-uR67F_{rGc2eoh5$(Kg&nDFwu@cEdtygvE#joo*4921gC+J}vka!!O~6rFZ}AfBh9BmVAFzv5;)y zww88a-`s8!#L7nl6gGoo3#_nPp!F`>e-&bc(9b@Zvx^jQvo;Rw zWdY%yAzW}Z6jMo&{RYC77TI<_cOC$cLTv-^&?ZgHqwvpl{t}QH zLj{k!g2TKO?s#mpHp(98(xYya5WDqC+wwxDMTJib($E-T7tT{PMlVpLfs4$Pv;OR0Qswy4QPyYpE2kHN=lr_-D*p zEW$Q)s=<~!q2vt{gvf;`)$c9#G^y75RAf{n^JM=o~G|Q9_k^oG?xZ)7M&MPh6Xd~ zT*+zDAn&5LlZ{XmzbjZ<01fDAnkuH`O0d~3I*&pLDHQZTiSHi1LBvCgrOP`-7&}0h zVyYjmt+0Lo0-4#80)FqT$8x=I36d&wKAXKVG=}>7N;*n}^oHfC8lZEkq!akvp530i zg;*%v0JZWkW47=mVp}`e?f6^&zVB~7M)c(ke1A7db6|%btb2wmgbFDJ;?KEMWCuN@ zpRi~MiM9INffE5WSq^KE77zo@$+FF_KtQ@jaYMc@2!NLjk;rxK6@pwzwQT*Qdyv}z zV%xc*Ce>}|PD+DGEze}KE??n)K?Dv@LyhZjO0oTLfTtBwV#VuE3YZQyPrzyKaMFX!Hg8DmxZMG zV1{w_46Hp&m{bwGYR(uW8e4@C7(#3eiDAug!pOL@Q9sa_ds|o>7@3-WMXF@g983qp z?63Q?d-w-JP`7#@q*lFObsl`D4f0$`)zqY*Yo2MmC=r@)lutq;8$6#u;v{Ljn~K|X zJt+#>LFe%k?80VYZP%C#@WUSL=nrt}8poY1SN!1rOf|1oblAo9L6fnkjADSHzkrHt z;uX!$r&HBY3k-qA&lc|cW`#@M^diZ+?-eOdlP(&lmV_-#nFM5B!kT$qbTMzSD8sc) zSpDs)H%HOXIeTg?g@*dCQVTD0m=at`j&=d@vBHa_zdpO?4}Fop3p09S^RZloEpOc zEhfEb?;Uou6!Zn2c+n+3zJxTmL{YtDzw2G*OX+RT3_8QLDdO(Ltur%R`<*KM3&QcU zD!4(M;d(Zr8Zu@K+o^QEedEp8tsEd>??KQ6W$cg`!FNa1&){`VG%k8>{8#H=$x3cs z8V63agu0?3tLn`Z0|V-MDzXsMw=Xh1P#c1jfx8vk7;U#_p-Tq%fFIIHf~^tSZs9KP zF_W3H0!b$QZNHtr@V@@|MW=E~mh~w+gCV6BB^EBgPBoOORtB~S&$$Cjswy_9Mb;g5 z&OKgmo~NyqoR@AUo56z@6|&2tL<1bFum(D*+T>?#!jmO^W33%A8Wl^v!m4XH*^zOT zTYfpyyySzWilSNdK(+?TW3uQfk~Z-}_^%3BE0QJ}{=m!GqeuZa)_@cGPVw(CA|ki8 zBdmJ59B$&)5e{3@`pJT0rQbG=G|JU}?~v0D>|X0K89Odh%q5CkK#wezm9>hsn6h0v z1w8Iz?U#^o(g`{G%}kv-P6Axav=+_*qL;hqG}-*E>PC4d`eVbfI7zquC>F~0I~`D! zdiT!io`(2*HbkD$OT$mu8c)N&!MY>tc=B3Vf5E)q-eDauEG6{FI;2x==g0A|R_K^=kRebEx_22G%+OphQkk-i`^wlHS18%%R}62J{OR3W z(~RKg{gHeb+2+9WoNANqE~XfWE9Oy=>nFClVBlZdqdX9Tac6z`dI!|Jrv=^OXTkL5 zD7hEDnP8X4g~todul`vqPr3~-MQ61rA zQO1e)PPuIEtpl(6kl8>6#XwG#j*49J=4ELC(8rzz`dDzJ)2m1v?~Z3Kzrv+Kn9b{* zwU?LW+z+CJm;537S=b=1np(rlq#K-%ldFNxngfpNUpsg+*OzjzHwN#w*C(gma9nr! ztA9M?H=2RRasOq6W#C!tQS6CPX{a1|?X3KBx=Wc5)*P|f{}8{PH$Zm=>og_vVNEoa z>4V%bZZAX+Z&xW4r1%>Lp5m^eXp%kB^2dSa9*0c~-X4m9Qu!jN2l;m88^`CPinGq= z(zLuOx86MVmE~_W%5|#8vx|5cntgsa3~3G1l8Pw<>QBP7eRRgW8c3zX=`0$*4%szN z6k#n1uJMlXAMnNFP+W32bXRa2U8}ATXl-X2x#45b0(EdOj(P}sN3)vNH*UKC_!4g1lXwGJrvI8PE?yYZm8%@b$=J-mYn?Mej zyq)}`uhS>XHq=l!$I(Kir zR;T%@5ioT>zw{Y7^n#5&Z6@e$qL@aCoC4Py6saaDF9uW_sJ^f^wNc(BPKET!Hen{c zVG?k5=s?oj!21L!hu~thqBjH+cUyn}%HX-hl0{r7FT^@Uo$BKh5hUD2i`sb&ijLri z!G&trO%Ww2OT%Jb*&3D=R5i5~*u*ly*|(jaBFd(9nq1M9AZ#<)56Qv6ZR1hu-J(R^ zcMk8%=-+{7zage)OK zeXbmO`T|mi)Cwv8Ny=zZ4gEmeMUP!UtD{1I z-JZPWRN@PN+WQtEWy_VWI{fHWLE+sW#cf=JgH$sYKD@e2K zUPvm)eqfG&N$6eIZt&bb03D0RkZJyT=t)>eX99>i729MRCB3|8L3>CKh$iM}%7W81 zH)q+_hHyoR>HNZE&!!fG9VH`-I@H^5eElacx(*^TwY0gd1mTyG&?IGwPn4*5#$o@Z zytOd7wQZkU&{3Ac!h)u3dI>jJd~`m|%RDpNVGRmP7IrOU^0zCi74_uoeC@Wcp~> zvRLz2u)__74%_E1il%-&hWM})83xl})6RcpW89GO-rT@GbFngqMFxxfSk8jp1*aCQ zdOiA$MU6kq`Od1>KV5k8hhvC@C*5HYB%WM7b{m`^F<#eI@R~UtG8eKH2X@V}kSDkE zFhwq!N#w8f(Lza^>gFuSrNfOTl}>Y74q`Jp%@?EvQd9<<(wrNE$7J$q5t>XU9?^ZF<|@r zljn4l<*;~8LpDDB@Qb&<{-RSPBc%n_2qgMwvA?m*V@yT~^RBR{kIioygMChOkZ~(M z`8Q$2pszh>ax&RJ0aD(F8t3B_bA%#?!8ivm7i@9vSHyVic`qs?O@O5`tL9zG4rowV^C;<-cPbu|3RP1?lf2BgN^>J*TfkXJUo8wP^lT3~5KO*l+X?0 zYO;A+H4q`kO^)$0Z1+3eV2YT*?G|xSm|@#`i)A@bID;3ho8`@0eI+;Wu18v)7m_0dF{(5u64Yt1x(gQ42Jn2m-W|#fD`9C2Ttn2N zzu_*BXye-1rU)|cywM=aye(VqOJ5uRrn_;z`<(ycUPOByc7qoPC=0u;HkD@Lr~KzL$vY>A2C zZGCM7fe!51?a=79rU%a&wNJVSe}zWIw{`#ba;A z=4UvWa>upA@BVh`G7C#~+Jsx?wlJ$mF~81xwOe}#hUl?a6IaODb8m{usTKa`$vs(9 z1XlARkIfPAZngv#!+O`w@16B;p3QM#)*vUCIPOCDr!^~%n@>ep#_$KOy|b{2;gNrw z4@em|@p?kb=Qi@rLb2{6f8?skbjt(}cE*etFF1|tjJ=RnYUklw-z%I<$rH&u5H*n{4`kOVQ*3r$nE-uG#cHzbMgI z4wxJrgCWRuKAk~;(F}1=a-o0XKNhr=n zs(9OdkvC+ipvXpH}_c3bE6NZLl*I(5~8 zp0*T37BpqkOUk-;T?#Rpi$OuT;c4)WJ6R&9O-zW1Y#esJ7E2!tY$Qk?2OL~bYK2m8 zs3MmyZc^QKfm-2;DQS{5{A{{KJm7@XfYCsl)J9`Gd5=6zFAvocq384)@z|{}6$4ljn;5oxTqqa?!BIW7C+OnZIg5@fx#MaE{<;>9?j;vPEA}plS zQ_P%oPM0(U#)JOIL+IQ%L5bKNF@D>0!Dt`&@cY@jV1!h3$9-qjWjj*ME3;Ul!RW-s z*ZAC|5~z7g^hp3k8k8Z>srnTOVb|v%4sslI%cZJT;t|y21+xg{lmrULnJnErk1ORfy@`z*22P43^6=?Y9u0O)7(P>oKx97*gLGhIiP6Rz4kniHfQ3u&Et9wn_R*YaO`r`sanM~ z-q$=Eg(cpQ^3Gqy(|)w*;}{a-_18ZBqklO$C0S;J^@*kZz@e}8gNmi?j`lik7Nd3w zNf|JPF1~y9+vCV1rx$9jO_EJUn>7@(k|N8f$SVI{Q5x{7>89sH#FAEQ ze+@fC{&q9;FXj{^Tm}%wU6c_P!jem*NrhUXsQ8JBk~-DGM^hg8#XB#&=GiduW>6y! z-6tK2rSe6K7NN4Jq3i|)Dsk=JuEbE~sY#Ed6)t#@U7+@S9E~__H9cs17>(Q2yqkQ( zI4%7+XWvb7#DS-!*G&vXGsT>x$QddUWoS#MwFWi__l3o~*9Aw3_DcJdy<%;vcwZRE z)q<G*ZVa=%72X4uHXSDygp8{O><5I*uRg>zA&a_>h#>9>hjt~NOSG3qzqlnI4fjdDY#Qj9+mTHsH>q=>5^fm;o( zBxFM8HmKPIET+(5-{Sx(cB3=WfRpuwbb}MHzJ-yYgTl;@;`NPrR!*K(ggr$afgZy*xKDTns ze*U_!L*Q*Wp;{8KO4df2A~y1Byt)NFip@+3eKz8@yj+t>|Fh#G?|Z*^V9x32uqz%G zvPsNeMm-aqs&*PCIK)if?61X*LSUaPP@H_3sHJ%Yo-IhfL-& zyD0{IZiQ6jCvT&Q353pTCU$;Iddgc@bo*aBRl-6$w%18Vyc?h)~eV9I0i6rfW`AKB%{}qUy=IoxVeE3M2*iBnR0K_+r10u zY#=i`C+(JZ$s+}KWj*p*aRYzT)H-hN%oEV)F`hgU*LpQe``Pg5KKUhWJs8tNw)IO! zF^<(c?7Xkw=^xau_b?eM6>om~ELrQshDxp(h*KzL3kC4dpn;kI`AERN=$GQI&3G+E z-aQjf=FqKt%z@Bu@@$|F@lu}C$!`=WuzCw?*jk;*z86@*cEaY>pM5Mhp<{DG;6<{@ zl|W;OVWdOZ6a%reG)l9By7-UKQZYNELAHle>3*1-&c!2RGBxy&E%L6GK-mQp&Yh(P zWKjW`D)f|AM-)LXd#&+@gT(aE17y7(Ph>P4*mnucW@EP=PRn7R^^M|xkeYyU{5#*h zOg6Z}7GgM1c2G%Z~%sAJwkaXo!t5%yW9LEow6RjUOiM?lu9T2Eg5%w+k$P>f{sVC`8UjN&O2*qlCs(F zI&sw4iUGS(zD&^QRmG2G4iPODH>3Ci^598Yto3W9F^LUNl+LXG%9A)r6{mk^S()Rs zu6Qdfi%-(bx}>Y6m})kZZw4LKpay6Bf^nv|&Q>un-hzOgiu=~vp}SR&4N%R88%gm? zqe&XfG)XPR9HOA1n5Grlf@O${=-u;fyxOF$r#H>$RN-+N?_k(D_2yX{Ub`ZWq}qUy z{%GjwseR#fKoY!fVg0<-@||>D$kr*XK@6dKxVr&jKG-0sOIW4sCbzs3BFZ>L zbS1Ah+#qX^K7vx%$KPcKJ!L!OhVR0jcWkajM)T=+?@UXvOl*9%yvupbDQKBA-00X$ zF^LpOpfspE9?8$mf9Kr%g3vgTL1GBSXkZa%&AfMpm{(v|?vPX?D~rHx{jE&mfYa7R zj?3HsFZe%A;^np7N2ZcbN9xh$#Fj_Q#No*3rMi!mSrhFIN#;-E~Z1w1sbQHp6r;BI@RI0(Ew*wgLO%nVPx0$LDC zYNu;=33f2+z1kI7bRzJYBFKzWX_2h-!epI18w!wyn0>Ht@Y2k6%)Nkag5L|Jg24Bd zL$8g{UL<`U#+>~wdNXI405ju^(||Y0=Zp`wgVR>?>w)K`-?Kp0HMuKj2ZMap4H7;4 zZk3_Od0Lp7mqS1F*y))h*Qsit1|V6Y-RXIj{&#lZd!Ep--*xNYJO1IfPDxDQ6W$j7 z7m0UbA8mn|2uh;v5M30aZT7Jtp-A$YYSck(132xNCzuI;Et5vG;n5i7t*U)T!+(uG+ z6#5m91LvB>0<%Do#hs4*cwc-u=~pJv@$mBk3AyaVqT`{N==hvsZc^krrRiUg$Q=^* zI?oGAog!DfO#)<$$o6%Zt5cwDqgM@tX)A+M>AEQ~f)enq#7}9OGNg<8J+YCO4q6|y z1gCRXN3=>xytmFAoU1)4yhLg^opY~qn&pjCuR}R{A$Pm?uz8c@XM7s?*qayc(cU7t z{wu@wDDmgU;BF86*(UAtNDI6)l)3x)T%7`M(8h_diJVS>eM`>*3n~77#`IiNtiUIk zGrLxqO0Rz=50?dm?5A`_p2-Z8tOU`7PjeIab4-9+d@Xk!$z!(xi;>>bWBw^~viXgAk@ z{h?-tm{EUNx8#-ADHC`ec|JM~DPyeA=fvRIW(J-(idjpM7)q1q*RIHf=739=ZUH;E zOMX$Zbe#Emh8D{*4EU6C{S;Fiyu_KlH!tYLXFZ!@F9t%!w@;hfm0N?i1e* zIu>z``AmtH%(!fKELO=`SDBsI-}cX$?uy5lIPuGaBz+|Bj1vb7!J|72Cq)zkyXak% zW;L&Y+cvizNTsggERiGJnNQWRTxN6B7h#e%K8 z%l`X3GH0*!t5S4_fe*&#u+LTTC){q2Wt>{&T28&4)2@wxvI$2c8=FthT42*2AFbmk zx_*Xqk_j`{|D)x9$!d0nsuTPAa?H@Qm0}=1wh_I+X}k{DiP>Q#BDuq_Ahb-mB1jv>J3(W* zKJ15~Bxe9pnJA{TVIK z@m_XVDce@wj!SCeRMyqV-!Er%S*j7IO-Hbz7*gqzihj=;uXZS`zzXpJ`DH=1Yz3pO z;gl;6hMgv*5Z)PAF<=`G9$OAJE}|WVS;xT$PWg#9B(>aEO(w>>mFgOj|ALvN$IT3h zgA@a-7kep98n1@e=oKkw;O~jZ5AT?_%`GOdg1Q`ffw!E~z~WP<_-CD8svx92?~T)B0#`)|fPlmn!b^-IS`RsLQU9^k-NZPV9NDG&5ZaDFzUeOKJ3hjj|}uZP1~uVsX==`hcOh zGSVVfL|{*#{Q>Pza9kgNyynIbvlbO=DxC)$8Iiw+@p+`F z@n{h9tgT{y$go3<%e5chV@!zgc+mR^+3duKDKkS%4#i|rB%RWHdy#jGd%HYWJP71VjubLu zLW=zX@?=I znMSJwO2XzEB-+oV2Ln@qrz20aOpqlwB0Cj?#t4!{<_Y>e3^Wca32&A_-;e~)DnFz{ zNRne)V@TW1BuVlj@3X<}Q#(~d8Qq}6#voB_LimW%!j_Yxe+5_}d_r#Rf1e990japF zD2n7bF_4a!0jZo~fR~|&(iC#5!PRpwtXZnAau0@#@vkO;d09WOI-4G2cj6kLQkXZ+U;@R>v*#J?~bjdOTXMzARq0;0z?h9*ZdR zACPU4?4MW7j|{0<7#F-67$&>qt&&JyI|CT=>ILQpXajPBtoMUb0eTIIN@GTewgo0$F!m(z5{YKGIEp3Sa+Tn>;SqI4HsM1uPS}7?$*B zOWTNz4eiAFF;+~x2j9FRMn14wQUjdZE7iA|bJDXC`wA%AM2Jmd%(kiQIx0GA*?(I) z_1Iv-d96NceOxil(Iz)$-J=Jv%(p-YL6Gzrt^7nkY>ye`c29<_=yVFGyDpHO^?GMK z82xPm$g{?fUBNKYWaWQ8x!#go@Y$ROCx(F)3c+Rx($ya3wsQ&>G&9JQUaRaSx9EIM z6BL7#gr|k%3E&74u9E`}@_Xd%p}S}GaATQ9X$^P)Pb$zTE|cAYko>n6+nm@)-fw0k=Ti(2aAZ@O-Jv;joCr2TjZ!2b zJHZr82Bau5zK`7)N`>Hz>GzE1-4SB&2zN^Ne`dOX<)z*H-M&ea-YJ*VO|J@G=d&OB zr8-38j2#+|@x@q&#t2cR35B1&`R%V3!_R&1+q~3y&}v+?XZjwxIXoK-%^F^V)FB}G z@;BW!2z~h+j<~9bdGl+QEiGRXtHerYP6Lg-IftlTkk1VLyAY7XTstJ)*zTP?`AYCz zX%Q%$IL??nF>IF0Hh2uvHIDV|KKa!jEUGK8%&D`Ix`yl(sK|#3W}ru#uq6Bt>7kVVSPG^2l-5Tagmkz@c=was2c@|qj3)N*Zh0gha3||{`gVvF|vFlZ=4f{4Aae!u#sZo zDH2O*4s#8XChf2RjAQtTR zxHat~UZwkv>8qvMx{$h%>*Hr!n%;F5z!(SV@boKK0b}CzDs`XbF4hXc>yTkz$7|>0 z(A^P0_mKoVCC!plQIA8rBs~4Yh8@}Z?1$YlE{E)%z0-1@|12bFs-;3?_E&{ zl!9Hf0qxXyUY`eIq|!Z6)G-@L{MzN1N{}atlCJfHI+TDy8Q|&~xivX4nVqVW!gyYfIA`|8;C_!Z5ndXX z<+pXohie3fuC-}$R`P7(>uL4p15Dbb{1hb^w)rmY|Ls$2l$Tke^bt zdF$1A3*#0Xk$mi%0|pj2rca3wWnuX72-K>x7alqN=lRCypJjCd|Evll%FizJrxq0IghWekPgm`=N*seq7Tka^{))xH)Ee~ zyYD0aK94v^J*`$9@v{FjAN%7TdY|q3sBtn@Z?p5hjFP?+S(D!am>D@i$U!B-?Sklf)A4tk62~_|)$AEm<;N65V5k?4vzw&_7Au z6;UtB3{Mj$&RZE)2K>UyAjkbSof8j7r6+jI6{5^-ladm;6lN$RGUKUXtd- z;HWYK#~z9S36V!>9=$dA!*&j~+0ONcqCdALb?%Rv-q^n6(c*JzogyXV8kj^m^dnwB zH;0osb0>XFwrsLqok2R(c;^BCFT_3agJBs2KjF11Wis<629C7{|NQgpATxrrfK$)^ z7WO09LFUlsKjV0rAoJsz?H9;Kb{fZdZ7X2Z7;dj*Pz<>Bwo@AHP^njU$dfrm-Wxpo zxn*LV;&Md4{0!+0*C{gT(ukwbvKv#P_RK2v?+ve~n|`T(JC2HUFl`gA^e8dK9waR9 zIP^a$dP}YIY^AaD?&*jXO@txlR5~>%HE7TU8FsFER{;gZaC?)1uP&)xeHEgDaiSto4R5#Cz>Kr>Icai04!2B{ zMR)QK$SNiexHL$sdAEUrr$-jY>47V$^nlB4dX-1}5YhGy(r?E-6HD}rtd*XgwHt!0 zqbK!O{H>PVnVr^=Yo!x&v}-Tfu3Yc8(iQo)OMQ`X^n*&SUJd>i;NVVG+?Gf2>ggzw z#Z3Sge}-sPKo7UnH-~d*hE4%3Cx^!1h?9q_q(Un9o8z{0KDd`eL74zHWmDBIZhSF>6#J9q3 z0m=L_-WBLo=@5I^x~#ArE7s?5#1-pU$$RIo7p9t2nqlGGDBz@?m9%fp@`VDj+0gbmO@`HUdPy@qp)t`}FjKOVo+^O&+p zjsGU_&Z!4no}NE6qe*>pZVd;|+ZS&3z?lhwc7o+QZj69WH&(2;_ z8(>+d_N;;8yp#`+)H_@`rBTdwiX>B-W286;Dt{Ve%LCdWm#W_*M*^t0KpJHDVW>R5 zTnbqZjZ&oR@1&2bwz`%2jzi0bQ2_x&gI!mBgvqSiD*y7E4;Fs4Ja}#1cfvnd8oyYt z?xsKHM*BeJ*tC}5hhngGVqX7Z(ckJ9-Cndbfm*6ppPl~n>itCn|McXN!(_^6@$=Ze z8DZ(|a_VNr*%o(`Mf;xi{9i~6J3p@ziw&!LC4ZEzmO_L-}pWFRQ6Ewbi#jS^2Vh0T;_U!eWLE|%u zxj~T*N^?@U6ACCh=6xP!bnU-)K$+r~>fa|xoMAliex-X1De}HHy+C3-@qtc}&)qz0 zmmr(&P$zlS_(FycLFEBJ@xxa_ut1; zzrPW>L)8t8-Zc@C0%Z5bE0~Sk=8ac7<_#BX?BMoT{y6C8Fd*Iv$p27jg4^uWzbBCF zmj-TiW^mg}G0;C`52Z6*OC{IR)JHGS0^k$~k5Iz_sw zju%hXb1!kq{4oStMdQ7C^+_S3EyXXMRMAb+W3nn5XInpi4=0(iJ3FEMfHQ2>?Qqz< zO7$B{hL@LQvt)%2CMTesF;?7vKDPy0%UFIM{<_=Y0Y7gF+kx@Cne2N#=TC1YzGjjr z54q{TAXQH6t-E9L8CNJlq~rOi4l_nZqv&Ic@Qx=PasLg z%Ro0N&H^L$pbOH8A%9{9w_<)J^s4&eYjL6sPi^PiMroV~`LbaySrw4r9`HvF(=Ph- z)O2w5K~F_s3qbl?%n*MESs9QW_X}aO1ZUFXlt^s3Z^kjYn8TiVVYC5kwpxcZeynew zMsny~ob+$MvJ4Bbku*+hEL#Z}6mnaD*mSQ?Jg0%iepRqHPm=2uE$&d1kG1)ufCB3^ zdU}KHg2wS1@29`jgbDn&^&u~M7sXco))_f-bczeYPFZx|pg*Rc9}#uWY=`#Y`4YTV z>0UMm$?Bq^2S=Wubk05gdU+BzQqZZ&qK)^im#>%i5tJQwrkD6SMhmgQfEDpEeCh0V z%u13_-})rz<~t_URN{MDMmk>_V>jBY?aV!jxkHiLlx7pR(&LN>dwJ*-s|343!FW|9 zPfif6fDTuc9_gYQrc^cn4XGcz+A(J*9p%>o`FvN!y_^PNj`FT_9j~7@@E`F$j!5TX zawr57WC;;EMdQ@#^Z+!S(W{R~6nmegF@@`>?{2R)@6AlDxPiZuzQF0G;{`q8*XVvo z-FkfVIJmHBMgzYU8VlFc#op(nIt8{Q)~hQZc#*1Tkk-?W=3F2+k4{k$v0T!?*Q@bN z9H&KC5Q3ucY6_P0U=^kHz zk<`jJ$net^N#ERR`Y5jw%&?;>y=0|pHaA(4=B8JpG1f&JM`zNJK?mm>WVlglQZI=N zs`D9ei3~~+;lD>XOU4wqax9e0(8211=qG=wwk()?Nz5TDb$=y7{lZA_4>n7>$sLvU z63DAA>1tr*%o8L57d-kBfv>1Dv}_ts6^vIp1$%g~TIEp>{`6aW?pg9xIc)})6$Yw9 z^Wu0VLWn%Y`YjL2Ri%WS@xfvyB)S`4(-&63=5-$RijD3@R+i5M!PYBPzccBdz~4`q zLOPsyQ@+B?mg%LK9*W$eG#yNV@J`tFnc5DfW%hu}r(p%c0T`pi|<%I5>{2u-maMu`7qrDZjf|n|~HuD_bT_B7URLtMP#Lw;{D|qd4+-D=tp-%C^ zfXn|Ty;eH2-%Gn@N+r|hQ5~R-6LpI(aN+{5ihE>7edBoD+>ax0V+NbDqslnZ#^CCJ z0U5@~?}UFcqrndny!OsdR}F@pQ8h@lt^6!i<@7A*krPL*hwi8EghS!@G%bD`8a+to z*cc2&<9VX}0a!|oGqkuP1J?`KHgAN@`-C4(s(X!?fLH92c8aWGXZSer<~_^I@Yzf; zi4;iy+UNzvK?5$0{07+(s9+dMCF5qk-~zHi#&}|5 z;rCzhvovAYSh!9c0<~iB>=JzO^(#>M6z$rxaBE1spn$v5wMl(WofvO$FtXXja=9X3)l$4;)D9=RZW#?_D>Zu$%@7Dl1xq*=1UGud;Q z`s8%&^4S-KRl)fo552%9(Q^*No_XV>7I(**$~MAeM7IgMg=aVV-r=P`w`_UkwEG_` zy|@~|chc@%M0e9?d>Z(>#WCvSfOI-iPz8w#<#Q5Ahes9EJ&mVmV71!p7HmwTS?z)- z@6}OnTgI%N78_Qg&?&YwjxoNieEfn(!MJ`a9ELE0wJo$*^bM-!Xf-}fkml_by!>$HJ zoMz_4TP5K=TsVpJRoQeGI91Qm!1T;1Q|8WW_V1!^kqew&lE5+A85p6~-U+XquJ>61 z-4KioDiTHO813bGF@j{D8eeTckU%tf_fkf^yi(>FwxJ!xO0@^x|fz4B!&04n)N16wQ+RlB#hb zmmd{GzI=7xobkuRN4|^_&n}Pu@g%I_@%Tmu|6~P^37OKsnTt#s>U&3ix`!NOXUsV9Mj6!C zu;SX26a!=tM=8yI)d5u({r+}uV|wO=P;|JW3fUvrBe)sVAWIdk@fviw5ZdC71x!iu znuT$oytJ{vX?_j3=hH%R=<`$eOwj@YnM;A2XDw~wykAAHmv_l^szam;T(<~Cyn@Um z*c=10%wZ#biL}cbrFac{G#S6w0QS;zG}VZ3VGTZ8G?V3+5x@D%PxAC_D}zmNTAkFr zhwOGGkR>^69V#hiKSj!69RiT2rAMAJ^|orIhfZ;L+73SholjwB?t?-sWX_wld$LZk z!>@1wW>faa@m#deamY@=!b5F4r#8@d`8pINUJvc}&@;Dr+99{Ydfo_$jNy_8|aKep73fPE^5r(mV@d$T2?v!;HFDfqoz|ACC>c3X~Az9(X zTd5spA}^6*5-1XfJ#v!d{hn7tkX`sXCm|eq^||FjB~P zSLOVfFS>b)+tmsv80w-?_f^qY3V{8#Qim!C_QClrlACjngjdmf%*w<@e0=sg$wER5 z*$wvk-sjE33LX=8oqf&IvUdraIFR#dN@Atc3GS3vg%kioE0*K8%M<;8krm)G-r@z; z(KZM@Z^7+_0EZRHnC6ZAo}5ztn##@PYMn4}e#tk{jb5Lp4cz<|fIpO&f5Ofo6*1 zq={mvMlW#RN_)ZwWWeMAZ7P5{FA0b@Z;Ly@qAn!W0X^cL@PmQa_(G?Mp|a?T`6WO~ zpFSOAm#ic_QIcUCZYr2jnlR6k!6i@ml$UWrAeEAx?&W zmf+yD9CydR@wUC-mfK2)dE1{z3RXJ8c{dY(-t^<2hMVNnkIQ$ZkrH;k8Rz}WNTZp7 ze3)XYDN;#kHjo%j1HX>D(=$IDB~J9z4Fu~^dO2%I34J&GoV1RcIKL{`SX*+Ca~!G# zZiB$kgG1%%h*tjN`})0BfoB;D?sH~uSDjXEAlY=bs#tbak_O!XtCTyKNa&e%Jft=a<1BTN0F?rV* zh-#&ex*5hZBaZEIk9PUk-p5MzP5hTb|9sDKSIx#gc48yR${yT6ZwzjQBH3n1u2`E9 za)Rs*HAX8jhkNW&fZcN03?g>Zj(PM?Dk9(bmPzvjM$h?2a+ICsabj4sn`xf&6!S4f z8YoSQUpIX~c^Vm^!G3O?YoN7gqISwE=A*~H-y>3>Q(!bwTg@-_-pqkwe=v{@z;|~w z6oTJXe6W^<-O^;qP2R@f6C`~`vA2Pa3NTC`RCdt^7DX-?bissfOzPJu4$a#Z(kSis zJV6XY_I3_k!)b!-@1ZVo=orKc1G4L!d`>b18FGL-AP)?67U;AYSQgr5bRTWU+mAmw zsZ~D-b^P+8=<&ne@IO*rLof7Fa+PJRcuo zHjP0TX!iEbVO3MGk)qDP@01yfJMg|imwaBUef@3reBQP{&F7C~r45|-t(6>C%cLfG zbo@Kty-YT+lSfW$2}8T(VbQT26q8DkZIlN0#hd5gCU5`ZDt?ptvbaGSMPai#Nc0v} z344a#jZ>jb>ea5xrnHO81O>u1u4AfUeLk9uBSwdHTW25rIPR;f`z#q*tx%b^^6!Lw zvM@D3rzn`6F{O@I<69DbjWh(`3_G##D2biZ9as{MWGapG>%_UL55<5}sqQIq7{#c8GzF=0)=`B_&g}*9X11 zM@gBoNnuQWx#^V$hVMAxgatrMLeLJYZ^IO8_y2U2ugUy)#GF`7bnMIz=QXUzK{NBC zfMRx0&~XWpehuJwM{2TdZYLvje#Nub$l5uZXB|^w+kpmIkG#}Zn=>s*j+bgh7sSod zM*g5nlLBvqpmy@)E8z6U@#wVIrYM>mf+6;R8Lm7XhoK|?*76Bs0*uFl-cQJ8c7SnW ziBe_;m>i0M(vx&bv)1zn#CU3zl{}#SPUB(dIve)3t#hHjEc#iEe0bILN@A)Ds8!}n z>*s27IejEkRl`Z*#w>88oEalhY!DKo8**l8(cITfNQvI~!=I5uPMp@#Vz%~YDCQ(Z zK7u0lSB}%UT;yjurO(CC+;YjuO4P z+A||0AK048B|x6|N|$iUVBiaK-$)^sfK6p4asLG zdz@HV9XG?@L5hL?IeRHhG_RA6WinMcvlF;FzYn&1L(x$ye<;H@ftv&DoOzWdki{ zygWK|SuDLHz-I9kqx!h!h9uaM{DqA%?8NY~LKV|VSMg(+^TCE8->uP^SQ6eWDe~54 z(HDf#ft5_l@VDM`-rwjcE)Gr)N{f3QQHmk);xKoc+cqy423h5aez?B_Rx|Ndk|NMzq%1pR zD@o=LIjnQ&QgDyr6{pRX@Ln39ZW*V1Nq}c1stKEMNY85E_d)i1wJdseCoFzi$SoQo zoKb=;0o??1I}Qm8Mybw5^ZK)tN~8ITQQcFvMYnB^NvSQqmb;GRjg+$GysjtUi5~9e z-A^&)6e*=NNg>_d``jM!TKV0)7*aW73)F}-OOSD@6uM1fhhw~wu3WkJuZ@c!ZHvC+ zQ|aE#+Zq@tI1o`y_et`@`=ME7x$+7O#B<%=wR}t{&jxN-j8x(DE#ggfWO`)@k!MUG zixZ=Q!$>z-w~}64v$DgI<$;Z4apK*W6;6m|2~;RS&!wR{nVUX370|8{R30tjrOlFU z({@dTA6O)_)GoHnNXubyjx}m5?swG3w^g5)QYOLj$n()@GT>qzNFybn6Q|9k*=8nk z;wWY^bmZ(t)#<6-Y0Mu&`2_l%;w= zcW}EP7zVC8+({jytAf!oiDe39r(ba>al=R}X9F^x=#Y`_chtF%ta)CuOrpX?q$(#_ zuH>-UJ61+9z|K@?Hr4p<4@ENKz%v7v1uJRkZKD2B|QE4Vk3MmG{ce#`X?RdRwD6{Bv)g|Z^(aqbVtaQIDxWG#n?NK6mZ4TW_9Aov_ zzA#orh%L{V^%zzQp@i6J1iZ{P~maJvx*mdGKoVot*7#pSUfH%yv$DjAtduCa%2w z`nSGXds|G^9!k(nBOo?p8`&P(DI_oh@9C-@`SzefJFfeoG!X$ zT-d;#St1)2+3}T*&-_-9M9Rur(KYY-KUm5LD~_X$!AHc=v$h5fDc{x7a*1O(gGcz& zhEBa4&)sR=hD*KOm%Z}X%AEbTRb-bF+Y!L%H7t4l0LAR1NExL`6J#z-7gbPY;@w_% zBfI*#c{;+ooqPSbKX|pJydb zs2_hIefLEtO)ew#^uBMd3@ecBQR>x6i?4{^Ym*;_&SXWhdiwTHR)&qQQNaQjtQU7I zH?RZ7ggXi4mZk?QMdh>tlUUizeCpjJtMZG54v6U2#s3|d-}ogfX$ye9|y#iF&I`+)>S8_&Z4!CFsDT5FSoW5md0ndVmIsqKBmFdbPH zF?}0|EayQ=l96%+r;G8*3aNBIt$HY~o!J7elue!uw4S{VvjtGkY9#h&*dovE=EV79 zsqb-OKw0ViY@mw+PAZTT_n@K_;{DoP%7V~lph3X@a>e6pptHaQ+oc`T_=6S736{F& zSyD+Sc8^-oFa}x=bm-{jNSXx6f|#6G>U-P4tz1T030oI_q~J^`HR}1Lw6ddj$98rpo-_+ zf!d7AxK;K&<8ut?uN&s4i}oq5+cWLJa+%COvjNk10t%>mIi4oT@#_|{gd{kzZTu^TF+B&;rQjEV&kuQl7^pR-sNtIr`b@u1p zt$w}kIrJ{^I0zHAt8O$Z&5`4XPpm`;RoHv&pT1fm1T&Sp>2&f?TqLRt?xzP_?m~^| zfJ-ZhoOW51%Q*`4ZJVa125g#|JZ_O%)_F~i0M1BrCV#v1l($KQH2rGtJ7lX9ix9|D zALe4sr5LbAGAK=P&~Ztndpg;`!%~?%!8=#ItLQim_E5%cX}zeMjtt6!E|td2!ZI0@ zHCzJ68npb8?3i9nCxn;y>bY0=j*}^)6D3B0;z$F9l_;4QvfG88X~IhHFQhBU4t7Z& zPHc?;<>N496;ljot9(kc1{|(A986f(tFfY?RMscX<*ai3$m@XY;exKI2b9rMV_fU$ zd!&_L4`Y)hC9+knkwGbbH(rf#)vK?1Z}+}Gw`=M#RUXtp4agEA>gh=8t{oU2pLv(Z z2r(1Zo|$&cCoX=|_4}4K%Cj3@Cx(UvEQSe|?G%$tkrg`rkrK3gFg*+kEI;biQQ>d$}y1&6gyRF&&}UE(N+T<0GA&}xXAS>oP z;w3|_eyOZXbd0VS=SyH)q&`NU_0kE_?Y;SB1DhvXn-LG}iz8OBp(guO|IOcIi~O;$ zZav9x;tg?)nRQV@F~H!to6_WRFoS5l-+dpQ;ugvEN9JIi0vp#};-t?24y0rT2b9X{ zd@gYkc!Ms;ulBeHUYzLhHRzz>v418eHNp`zRUU1Fz%N}+d!gn_zho2_S)pd4>TBO# zA~vB$_V=$hk@c>y5W`KsREmLI#;uel-9sNyDsJH4@<#5|C|>WYxtwe|ir2$M2bJ+E zF57kK<1t@t_fQKC9C%}{NzA1CWo{w`PQ2Fve$Qc^ z)GCTOK#_fvrdjgoD~GuT$z}h~eX4@7MPDtc;obMC<6)CN?9cZ(H14b6oC;bquanLK zTIox3N`N=mK*s@9qfxe8_BQ|{#yt|p*)};AY#SgbZxJTR5nefxL_fR$JfEaMcqA~_ zPOey%0J1I~o?fvFh*$w=!taUa+aZs)J^!QKP72xSDJRZyJ!J;BLlkq6A{CS-l}?s? zB&bz3N_C1P`7J(x1>0jh^g>2YOohf4VB7fT<|A{Owpns=I`*A3$mYl+=`-a(7;^KZ(y2j1^q$7PsK(JaZ0s;GUr)P%#(df_fBN3aR*hw+IH$F0tn_|s z4;%FVOnII34mA?QdcLl?*C8n-o$6j>?{$3c1GjadpF5GA2e8sc)K9v8%Tt=P(d^W} zCy;D*+Q@mGN2Jb78||f-5{m4hG@HJ81$u`@yGAje$yrmPZnwu zIf>9D5;HD`cmlD?!`QILAUR4L(_UpHWU+?wNRv3=oOLMI?LWWWGLyw=Ef*`fDQSL1 z5l!=q|Nh_SEDbFmJ0Lh<#jQ+f9S9?2EKv1RL?J#VHS~)!6V8!jCw9jGyTvefOdiDm z{c|Rz!IrbllH!1>2yHxbK#{3B06v)g^w56xuvbZV4X4~Yl}=Puy2mpI!`kIo9ER;| zZ!6m(23@X8b2F?%UeLTU2+ug$nN^z^7neGXD-&-gSzlaX{$lwZF` zyR1=ROsg#wH_d~<5|o5&CClZHcundJZ?1mrocjKpp76a9>7v`>bjZkUQt$Zsve*Cm z;GcVbaQ(Z6MRsjvJ9dN&7fB;7_NY!;7eBxIj`Dn%2|K_3c;0SORxeW#2ra$zz!%i(yOmQ4$m#`D$Wtv zArL(^KHadtyJ(kYDfY#S2|J@gi1|&{@iLNaGv&m;@`FtnTAkFrhwNsjuAJD(w2d~?k-%Nzv%({i zYLCnQpD;agY!imb-c>Y??xO2yq-x1hr3}%`W~#8%q?;ed>4H?@Ztt~Hd#N+Kqs<&N zqyDtA*rhqpgrDNBq9~H%N{*O4_;QLVrAQIfR+#z+;MTNB0hNA9@&l?^W5_EmTZrS63hbIfk1{hCn8ixZ`z?e{YgE{^!la&$}J?9_EQ77IhwVRnw z=PBl6iZoCfWbNBAy=!4ya7%EBtdGQr(wI%$>u$-A@v^}aQ+Kb5dw_rR4D557IMq<=W}V2yiuAD$yXPdKFN|SRVw{xZV|f-EH=Oz zzOf$7F&jLx0v7eBua$gg*+|(+2jXH5#CRp?9zF7d({kLqigi zcfdh*P5QAiWB!0kzNC=TARBO5slF^v4Ztk7Jm0mjn~e)j0OH|Qu2_+RF115}>-}O} z2V7RUu9mNt{iKXgOX*;#*#mnt~T0FU{_I z)x_a(i2D@NOOYN*a|@EZZ_`~;NTur$KN4rr?eZ&pt%1b2wab_BuJ}Py;tG11K)Xw@ zgTbA4H7{p2c5BEJ?GG60NBHzaRq#VEn8vst%z=J%Rl*of37tTU^AEaga_@nJ(!vEe z&&K)P-b1-r*gLgcQa4?v*yo{_?Qw@%nRUWKL7WKp^n2VpVFzD1U8mYF)9&<4l3#HD z47z=1s?tIr?NqMk-sZJKUJ%^V1X+g3K4nZ7ec8X;8{;As^mcFT)z%OBfjY$s4)lP= zSJY{HDaSD-aZ^m#^Z(DvC1i^$G)f=7Xgete+5zr>R(14BvW&un2xM&P_P!dDPZDP& zayN?#xzVJVsi)6QuM_Jz-QIF00rNme! z(>U#B+)CmMwlqD(Kkac`by}Pxk8)c!dBxNkWt9+#gBtipJrY&fe%Ga2BxmfNzH1#{ zHo^9(1#EpHJ6~Z1Tk1!j>AWmmHEh^Xofr4ew{^IJd_tCq=~sCIFbsO+u}qS@o<1Zj z5Z3T|!}cvepWcw+Q4$^(+$hD4xeabPf_3sVFf1G|8d)OhF}CcCpJ%80Y2n{~{GumZ zwJpLd-%k1#_n{p3Xnh_V{H{u(JPopHeq8Ws-ffQx7;#a!#PQ4g_yMzA?#H(|{3kmB z!qk2{W1`74U9l?Y2pM#7+ASQ}ZYC1eQ_MPwtfn-Wds##G3AG1R-DDN7N?7Zt)Rr$_ zJX;X)Gz{)GRqPN$0IaA77%VVm^`#VddLu>3PWUOA6|c=TrD!jc=~MA#~NW zq)CsDaEeGlj<4~0KGnp5t`P7Qm&k4&b_Bt_Z|O`Ng~fBZssj38s7}$yH$WhwlV1|v z70@ov6Xeie2-nK7WeEOiI2y)FT0JAdah*@S^A65xJbOo2#gAn&A!{CI-spA3&v?Gm z-S~5uV6+9NkF}HIe?G0alGza-b zZUe8BLVF1G92Vqr(n4_7T*W}(8ImVRgP9T0__(!?+oRQ{~H(sE8KfBRcE7q!0m*Ncn)x`Vp^ z?bz2k=0yth>O4_)NH<+AF+jEV?cbFDeB)bt7quw6NtN&%GvJc_t>!nep6=b`A2j_^ zTl*d39A|#`=Sueq-!FW!sC!?(|LrqByedZazGJc~;V!xe_G(bX8?bp!!dxJU7N1t0 zRTj&#L-1-M>F}r$j)L)(6{5zV+1SFYcj5PTPd6b-6}$Qn**=nDb7GkSJQKrY%5I8* z)+ITh*j~RPUOnk^`XCrj_2GBL_abWf$P>O`r*$&^f*OioD-ryLW5o+3DB4){7nx zThZ`+xj+^Mq_-LP@bp+9|Tii8C!Ax;0FKY^In*iX>1Pj2B{7 zFK*$r$5r(K*gP)<#Dq=(l$#8u1UpJBjJP0d)_TMV9CXsUtQaBFe8Nw~OO!7<@9!>U zzGVL^S}d^5p%1zn_*&zR1*_%0AeB6A1~*CGAiEaaAk%h7oIq(3_*T%Tr%jkz8Mc9reCi3UJ)e6T% z6_;OhY9pkv!u*VzUb&E74t;?0CA|T$vvi88;Jxl+eec-<#b&&+op_An+2Qq@ul?Yj zWu%G?;GNiSY$dLhB*)lxzh@PY&ujBUSj95l&>31`TktbRD2(d%-*$BW^H`j?PK*dE zPFw>W&jUT#q_{^)pke%>c~GmM12u@ucP+w#P|Wd6^vfAXnAn01`&kyY!=&Sn-fw-i zIK?r7WC{zd^SVv1SP<{lO8?);3>!gV`)b;0r2ju`J2}U+!C8DQcOA)dCCAKm2Ky-n zI&79wnm&)l;7rvulFI?fn*<@ci;^Zq(~wTbRwBqvs%J2-2cif~FF&-? zLuY4}I`9}R70ssmSr>?V*`rAS$@Q+Y0UCtV#_IsMUG z9a9W6tagLt^I*glw$Gc-Uf0=KrwK*xk+Ik!P8|HPVv8)Bya!V5jw!D}{4Ocv2sfIk z5Z&^If(}I-=RDx-(&DR&;@B0cjdAH%A!&@?#pd?*g%AJwqHk!8y%sTX$v|u4M3`{W z4s~*_3BoBQV1SZ zL+g`7zc%TT7U6(w$>Td6qoHiZmUZK4j3^rU_O^z<|NN&Ww47KJ{6li~1>g4iYM0?J)it*UIvZqgp9erYK*>7D zzU<1#2zqkXx|r%MF8{gZuGwjejUaPiSWIUF#jK~uI!e>Xj}oMcz)OOG1B?)0p8<5C zJ#V?5JRiFq*a!f-eV?rpSqH#K(d4&J*L7L8?R!ZauU2{yBW>398JM7!<$ZI?N)PR6 zuRK8?!G0#)keWNrK(Y}-f=UOnH%jTAP*VEA?8QQEhdaapP z8(J6onY@}#kypX>6#4$pd`_gG-(y2Sj$=q;YzKjL#bBRhH=l)-D52cbUcK;&2{IM4 z|F(+ka$?AQWCoc76a#$EWyWfw$gl=!N=T$&MMx1BS;MrE3o7Q{Q=-~QrHkCONa9=~ z#8O54KUHKZR_ynzRb~m+agnmOMAizm9yfVCz+eD@?>s??Y`X-i7fXcYs`FDz>JuyUOmNGdUSa8hdr1_LmBMMHYwQDhs&u+?`aymNH;IB= zrd$s41q~e(d7l2K$@bClnWqC27`qJyPwVU3a};)VUn zv$^B&zy0!?Z<#dE#GZfuj2v-d4b*0~?&m1xEJe;xnj-I3ewDC|uBNp$l5_KG19gfM zgJKpI-V9O%%XYNsBfLSL4qZLkLC#5!A^AYJ5>wp**^iJDc zy<(j^)(_$hn8G&bl0}{Ur|pZE|8PBZ^Nnpkz5d3v*RoW({++TjKFtzz2)>g;x5OOG#s!u6CB553h1bGjUj@1Cos#o(GxQW~c zz{Zylj#2n-akXT#DB3kgw2oI9+%5y(U~v7==;N|WoFjqg?ZH&slUUOl-p|FVf>@@X zt7G))BY|0}R(`c)K-R%M2jh;bZoGQW^MTC)s__<6wr=Jt`@boiVM5o6cc-o+sqE0@ zym_Fc(hOaN6q8SpTuRdkX}||m$O3|u?I#tp#1f~&g2vmdLf_uCJW zoldN(pyl8&zs)|1DWga+V8C75_u3WlDqg9t_855iv)x+fRzUb#bGQf#1r<=-@44;2$N}WVW1Z;HQhL_y?!; zJU3QopM|v%Ps1gNeZKfPv&!>`FdS%~qQd`a}IYSlm(&e{6Sm9FJny*gKLC)^-A7`Sdy7Ze8JH?(u=>0@;F z*gMFK;`t9>3#-S^dA!fP%<6r1-q-P8_a7bnYGuqUsphZrY!Mpxha}jr4Yx9V?p4Ag z988ZX799*o0|Ka3u6PbNKZmBodg7QESD=y##$3E(K-OF!JL}~@O0=T;k{4#!l@0 zAc3=Grb$Wl{zAHv?0CU4tg6j4YB9wCg=#*fDfaI3z!>#;VJqm%0a=&4$or_0KY+4=o9JtuWW3&m8!=u@)I}XnW-So}G|Ri(Yfgp}pe) zZgg~9q(ZE(Bz$NnaG*eckhDn-c+zGMVbqV^Y!>@}u>ROb%2RS$-qP~#O#fHQvLi9! z{oz-?*8Jg@zgzMbG0nIT%6ZK$Yc*2szq<0;jvuuIH*h;;?cij=NHn@`f!8Z2(eJ2A zr#SsR5Gg~Ql-9e~@vy5wlh;m84JXAfcVQjxfd39j7JY;h&x7EVY%B0BWU2HFQ2Ih6 ziab%y>|0@L$M5yi}NPmTd)-#Ycyd%rOuX}5dWEOPawLDGO3l5SB< z7ezj$G|Q+qc_B!YZ{p(ZXTSZK?3{EDY|ztq z`{$ih?C?70)8SDh>ZhBT-Co<~=lbZ?ShjynUas5*tYWfGksR=M99Z>@2L2U34zA^2m)_y-AE9%GLm~HEOg545YkWP4Y->cvnl7v6DwmoW+-F z2Ac$miKEC`N|Oov9YEx9T7`WjO-Xy(CddAN?0pGbQ|Y<4M>rvQF=Qi{oB34OA{(i=YVpx9j36!ZyboP&z{rs&$##Sk zK?6#NP^%ODk_7ksOXWN0rMm8zGk!~lrEf9JwUYq(T=ctgh4eD0Z4!C*ItCzL)F?a_ z4J={CbXk^MzXoa?y9utXF^S>EbVy?ft%RQCC3B9ex?+qn{R+Voc}4JM&@y+UU6roz zvZp;58~`LK zoZj~dv^C2fcu!)mn}uYGSw)c*R9usCQTPprKx3EY9llnoi{V$*@Hm`#@?4*qE3W7E zs%bNJYLwckhG`Bay8h?xyPq$LXRJ&-L$FnxUPK!1PSpoA#ZY5;BbmS=W#SV zW=9`H|IIY*nqv*fi)JW~=bxo61XHM6>BT;`#0|=-i2YjV1RgJGhlG==h!!#S;jWog zBy|C=a^G`&1>Wlh@$KJ0@I~1&^m~|9MrsB*{h4^SpT1Yzmh^(LV%merC3yy`}_el>~6S>=6 zS!`@G;{5cpjo_Z+?Aj(@`yU@TmT7R&Hazx}u@rSY67CW6g7v;3|v#tZf4)62hfjIF$=LwIaSutZR>sj@yTy zfeOFU7-Q4|F9R^-Ln5ZNG7t?>eUC4$YKB2jIFIK3wxz2j!HeUu>yn$h&-*z>>)8U} zYNX9-BWu*hNY7}qdWmPdZ@sik)k;6}cp@wFzo}gBv(R^`=P*3yzJDRN=iPUg`#;&A z_dxDx-a1J|V?SsMu&q#jdTPX3lE&k(3eb1;3#;T&3=msqQgNLmL)J;mDzoYdRPd!n zRYIaphIXfF@8sK}V}2iW$Na5udeLmd^cJY+GidUJ#j`V*G|BEjlWwUfn=T|~(yYEO zIIBG69^yB|A~zKGmmJn@cj1{EidAIJRYw|ZE=rWgN@$xMv{EQ$4MmcuxK?P#vVUv_ zK`sh?s)dXceQ8aKj##5B+YdT5zfOb5snJ|_QASem-Ef->x%r*jB_!9A)Y?7Ia*EkQ zky0wIQBeJp_x`&{ft?(&X7jPnn%GaLRgCKawPO-&4L_;v0tS&i!05PS;uHDBS#7!= zb$)C$T^77+Vmnl>Eb&`1@pkx?;AUWp$niGm(qV*V@|p71)J(dN*$UOA4}>LxLoQ4> zclx2f_|gt^#AQ#N+ndW<{qt5@%F5X`?0oy^4|kCYPjb-?J0DRD@Hif$;!ecwnQ<&I zNrahr4^^E|Dz_#kHyAyXkUvRhk|X4vSJ%`{+RAb0fW$5>i9%bg4ieb3!J*F-RS{q+ zHiE^t3i=c_YVOnR%2qm`P7>tFu#Tr49GDHtn@a2!W6)S5Z>Oc(g*d>S$Yc?CJ)v3? z2uB9HD0MNMTm~^taN@5VpR9OrZEPQ$@_+RkvW1(&mB+5O{dR9f0mT60WH#`qKpuVF zx7)wcHhaUwg`x{OGKkV{T8DRCBJLq$F3FLd&T`0?Lf>*!|VZGkb1<19;(6j82Ck*$@K z`!-Im7iSAQrrrP2M(GjBFs{;U&)dm)c5-%HcNr(?M!o-=ruP*#AtJpm{XI$k!U&Nf zJ0W7E7-(FXM#ZH@b;(ak3xyjM9dyRmp`jFrjPrsHx`lLzj9v}&QxT-7$(H&cYr#fE zRtR1_Ea-9)P*^y)+b9(&QcdI&#}D z9!j%@W#(ksDd+!0Qob%JHdcqo75+lT?DYrE6-&_*R9r#1HD^-Cq&P z2|)eih|3Y#7hz^gH3rkEAb3&eZSTaI(tM1X; zefKJw6sU2A4?Bb7E+4i}cA$JKNBgUJHnmb3cv3~$xTzKXuGz_wk^SBJk0|B= zMeb8^*>o~FO?Q*@kaoxfI7Q!4))9jSQWuYg6#!FKi};4uTCW?zt^RnC^(4^dTe*jO zd{#rp-#Zr5MV}_vGvh(P11a*Pw?N8c4zyv<($>?b=mc+lM{FA~rniXmNW1hC zwB8HHD2Pd*I@P)0qeSzkU5O0ZmI&mU80^)(K=;wOJJ8$VL&>~!8b8{lNucgJl!l;H zQE8*`@2ZFuxQR>h$*GC4;@?S3_R~>r-?ZR9 zE@A9{IK~fO6h}P97fY~nNfD}*N8g}C z1~&){>MI6y6T4VC5mY2{B(G*+eQuDfSFHB7{*CNYCS9qtLb?n3G2JA!s>E=UZsnMb z0cYp~!Hv2Nf=6SF@-q2Z=;372og4G1x=eK>AW2XILAx%Jt-ayXDXf_09=*YB5(az+ z2lbF&brbs}3}8g?*O7UvVC6!}rpTW9K0QgEc?|ITd2HPPfo#7hbP~m^q(}l4how5b zp+?#5@Ostdi2E__0q3hg;zXqRl|#7Y0(SIXm3-x(BNYw{hfH2@m9$;DK~@o10Oaxa zMd=fE0AZ|AP#IRLyA$%U=DI(y`3jmNTfC12mI&{8<8gx~K~gtsOXz=R3wO$HOh<~U zD*DFsC-P>JGYQ%OLaOrOV7$Jx>`<{)(2^t`!x5eWHb_TTTp~q{8lPEut7mT;D;(RYt51SS{S3 z%Mo6cH|ny5tA$39`R%Kqa7?<@0cHM&ao?NQCCfl1hj}BC_aU|-1N2&9WX$Y$7e);S_UYwmtw~F0_=m9Xr z5#$4m;GRR~6Tk2oO50o>FNs*%=Ar`+QyqGIuSHcuz69jhCJKw7P7u8`!@3f}8Fd~P z?wNnKQ8ynNw6*IC)9C#`Q7^vdUnOXz6Fg12 z>yfqe?eI-urvfs8?k2;(M|e2`uW+sSa2^V*_EFynwpkt2ac+9B@t}wL=QiMc>#A2L zxx{1O^wCfyRxqVNV)Le!(!rJgx-;V3K#Er_XC zRY}c)%aEaGb=Do2{mkR=>>@fn^fbKyO!sQ~X7mEVmN91V=;8I3r2C`I#02!RIkKI? zC7$u&`=B>iK};#o=OtJ~ktv`a&OT6Wk6o^Ar!!;L_&$IPt<^zSAXvRffb@FpFy>ma zf-Lv4j+6v+`?(XEg7GLO`5v>KGpkD{$4iV6W$b}W7s9O%cAP%hu*eN=r&rF2aI6gB zZHS4b3Z!zvCS8*P`Y&hCDNy8x=FpIXhsVy3Zl?=oqxZL+J`uK23ud}yKzQ-8V4j1z$1 zGMvk<9=a@4WJnBj~z7& z?d0nNin&jbJ5(ID|6b&IB z`uH2xvzb7@Q6f7yzQ@<3yF~U*ZdIg(l+S)-Y0l6>V}F=xX!;Ed9OIACw%O zPlu+3Y@E|Gvv5wqY!KI>HHt>vsjpm>FY>%7&6nKw$&xJgTom354LcWk76f-m7klD8 zOW_W5nD+fn?DF(JJphFv(1zkxryS)D64Xz-^x+GxBHst!PH4n7d%XM^1j zYPOyy>$t&=$8xIJZh5zzVjyW}E4CVHr5nVhf#=4Qs`Zzom!#1_=EOYBx{+^WoWnFhKlt%6$( z1Sgp>BKhK!5z#hFRO&T*JK4p}66LY=anjBbt)>{rAKVYD9^m}J^xIVRl8NggZilZ9 zNRmAm2UP}3!s2Hlo1j6nCZ--~fUxHxqy|HFZ#vxw_1W8nn1f}|T#>Jffa_6N<%bqU zXOOibY|D5=Rq1;^RKGB!bzI5p9-sM^|1Lqg&wg!7q#ip4A6hQ3XD=xzi`dUr^;jcaS^R<5ocWk=B(%T|E6#AK| ztHlLT4PI;CtQYT~^Cs7et)6`7vjBT7;<;VYxvk(<-s1M4M;JVgg zH(- zd%4%Q=W1Yc!t%(O-xm14?jdJ1ByWq>fpg@Z7qWr5e5#$BqE`dLxp7>56(>N9n3COl z^-DGx()16vvPl__Wyone8FGYTYA8~LtY&6lqiXc)6vhL2M$XrE(AU4xB2I*=fz@NK zi5nHIifUCRgu(BPIS(Obq{ZIqUn3}=f$g-*W4A#cjw*V+D2q;w!ic&-(=D#^flR2c zwa^KE*CUOB9^X!3iGQyDBVw&eI7u4EE>`wN)+=s!4;g%~hEbp7;PyFGSaE_8wR1|{ z7Dpb-FNjZ#h4pf!?@=wboUHV|p*{o)TZ}F1*NwXDT@>9FQ|XNY3M(6-G36ylg{WhS zLmq{+DllYo(yP}4J#x4Gau7zYS5(5Gf0jpfe_=fHXYAz6QHnW4k%Lqms+~?% zd$hIRQ-^<>7&ARTnbN6x7QOb#vtw=ZiH%aqh(Zg~{TGjTi2B*MZ76H>`WZCS!tsS2>q#^dx)iUd|m91HB)&QE~B-8`F0L7brHw z)B<1jEz+Ve>3U^##JYDS*rdBfs_AvIyV7gnb8DewYabNQbP;1zqL-1Wr43|*?1<2v zHUh_IZ2!)=>jA^KoNNOA`eXx&em~K1ZNo*3@Yq;kSqWhz*kY6D^^F>&iZZLwTyZCc z2Zo`O$R7A+Hy-_#{+ov#eXcCkG_4k=HnN|iuI+pwQP+*Xvj5Ppevf(GJ?Mf z`JV*`w+*EMGLK~hO9SML@huRGOqUs>;w3#kxpHVv>I1oFW;Hexs2qniKf@|7UJdZ) zdT;pPJz_*e=GBo(o1t;yAHQ{lto0lygORrx4N%(N-!-z0V$vzHg^IKIK@WuH~ zDw_WDtakkZK{i?MTTkDWu2dJRQ-!eL{a8~Phzs2{F(wA{;N6Q1E>L(53d@#Tk-;wO zLVq0B^t!X=qot4@(n0SDXoT$cMop%6U-*{LXR0-^?w##{rp!qM40NkokFg_Q0GTkN z@Iv<3zc02F%v%OEvS0heuP_=3L7`!c4_98=`)BW_lqh9>_@HulY?**`ch4Vjwr^02PNG^>q3pO-j(3Dc992 zz%#ZOy<)hSEse?4-kPdM@2RyG%b?BDV#6P>Wzy*eKV&G*lV#4zqL)SK^P`ZVcuBxY zpG=@OUg-5m)$0MyRb_m59vHsJL7eBEK;gMw59^O8mM$~pM3Ud4{9B_r1T^*tMg zzJL7`A)Va3_WX4Kku`P&#fwHV@gvDoV1AopDa)t^L4PWfIVS1Yu zXfl}!ztU)Ay6gsKosEi9+CyO`UCBRu{F4V?z4Q+~U%RR7j!g+NXzE1U<#nVr`h@xj zh$W+-Ku|W}p4dF?(X?kC2SwS!>(XZ)hM4nB+7P?dZ;i~lk9jZc zG~QPZSvA|`2f(X&O}<;uAbw1Ge2%M<1O*f8qfNTjaYx}_&@+>_qzzfCcV#)U9ehEY zbu2sxnOe(I%oK1xrb)3rYImr9bK!S1sHIy&kuT2SBuHLNf0!!s;&|d9OtrO0(;VdjT z4?M&h*kSRH&2xV=#%9j6&WgNBHghv)cq~iG?TqDoih*+VEGq73%mGXKW0PW~?^@p{ z@}1#%l3Z;U=_2cYG20Z+JRS;RuDo0Yxn z61fw3C-UGMt9+4cXHURo>0^O@jcjo^F0hbC9*2PFBMg_S#;(J_eunfrxW>YamwZ{S ztk(nS&a6KdzZEQZgjphe+GK!#Ll52kekwE}h-uMyyS4=*ri6Z*CA93qxjAU5d3 zpM#5CCJo>u963LG-UY{kGA;%$kC$C6b#coA7Ao4OwrTr5O}bQdHoX88T{>Mwo9V{s zTSIS6EmI|v^FzsM`g}f|+TaQ%rv`PwMc=Ici^sb*H8%3;k0+4U*UYoC$ZqZZkYYM0 za+iwB(zZg&iQ15O$-_u2S_iK&_RY(aeNZ`0|Gs|7#17@|(ET9>?L$?ORG%}cL6awI z(CDuQ8#FhgQD5DRhUEUcAR=x?-}l9xUXKs-iGJwOS)#oeeg<;!R!onVpa3b87EZr4 zwKGPasjZ4AlUC6er0rmJ>k~y)zIT)tXMq!`Eiy}syU>7f9R3e$=&>c>;_UUnABs+< zL!n86+W{q$44SeDz?*nekm$E!x^-lIo}>+GQVO8^`eW#uo)UBu991i(A0&rDZwl(j zb~$fX+w7vMCVu``9g}8ws}NbzVDdu?kgsXF9?R3#j4lf-jmhh$KV0y3Iyr<>!o|t4 zTzMJ$^beq^M*QTxH#az@_3{=HEQz|+-?@VB*R<N2T2l#jU-He1SNUdwoJX7uSYEWWkMvBlN z*d2;Tk85fat>e1M5xH4|WT^+q32k=Nc3~@2ST_oG_!=~5Eo0$}LDLXXIj%`@G;ptC zUM-ErxG@TEIT~h+TCT<|eFeQt(oQ3TJYKuQw<>h6dW+JaF@qMHXXPwz)IgcH8aazo z)g{bhI#E<+8Q7q~27&WlL)~T!+Uh&W2Eg|}FjtTrp5&O_ntC6_?50Qw6<15=jJ+FO z?H^AStCqqeNSRRuZJhH1ietBiUkR=wM?^hp3!}cW(mO%7T-~Ey`bM_!7U`gOL*478 zuny89-ZAFmQSZf5TOd%7;j@227L7yIk)tx(i7WI_*|-@rrM|-r`G4s&aR$drM|0gt zPT=6LWA53+$tT{n$(2W6`}!})Ay0D6POhA$7>IO#M8#n#qFIemyGjd#agO&V~d^YJ2yScdx$N`fl^P-PDzLo8C3Q+x^2^?_PPg1^(~;;U`qfyL*0o z3C`bv{hROJf@9TjCf#p0QMuo%oYV5nC39B4oiS&@+joAn;~V$?sp_9U`Nn78zWUa! z*;!xx?O7d)?sZ-XVP;h+UF=`2GI|vOSWUwF+?xe(2Gw8;6u5^g+&F=Xzh0ELCfr%+ zm_+tsc!R&cpRhF6+VExvT{_D8cOx)nBWw2Q8TH~UNyXTlVGVg9u$%@lR}>EMdUg=M zp{(BQsEK&H8e>tB=$gj0Ir1|!DRSldGyXRw@Abc^vSj#Jsbs7rmBXbWod<$ViVbmo zoxYtN2m>wEr##48;8-2UTX5{LQ<0yS?cp-!gtj(hyD(LKClZ~*kLiQsA4WnZDKsa6 zB83VFit4fDNjkJbY1drP;Mr#$xP{HFtjxM@1dh`(V?c%H^vrX&<42un(L!XPTSFRtQE;}1w&^!%p6Zd+Yrf({9 zq2-Y^IGGCUjkUDi>Vj#b!R?RCxzKTqBm;p^8;!vRXoX~H2!Q^V9J(R|S}sLl{iPoN zvIIaI&>y3~*m@fco+0e0)ax-9_vfan8~uz>-G7W=)wjX72i!Mkns4V=&`Rm2)Xvl6BiuA4O{}_JjMx2aN-CEl_bGEe{2=nDJ%Y(eYwGl(hHhne1ULsRo zb!rF@oDwisG!z4e$2ej!aF8dy$G45%P09j65n-ZE61f?Tg%o=N3ZsGj4g%It%8nGR za2|D9dPuUUS-nkrM`>2W73S!|iI^_d>yak`1or289U4epm`7Kl=7qtBa3K1vTfUA4 z4j1XeW3!mW94e)a>N1T{o+>hvLhqe|-OxyF?(&dE-3CF0I9+ZET^>?R&g!Znw)w5~ zTg9yMf^L`Uxdkyu*?EDa(p8ki^UR~0xlbGAoK>bPLG?1p#1&s7zv}lN zj{OFBYc;bpLclB?t95LRG*%9~ug?0^otR^;)8mrs*$tY3Y}rIg{iCV2)%Tw&bv2}b z$EBbr?9AH=ih)GrJycuKIv3&HQhC}V*=D2i8s!` zh;$|d#1Hx(3R}g*hvO0C`ZZ`OBen-NdpBq*y{l>KmPz-i7*}6y$|_lz*Iu8bh*Y10 z{=Ke?lsW;Qm%aHfoO20wU=U95p}zk9`KeMHe3oy%_6}LaWBBCT!DlPQq)}u86=(J@ zraR~kQGx(N>RZDXK=-REX?*w*(dN->MEP_vXa;}+ohLL1hc7U|86U0!z$?b#hL5oi zRGw-ZJ~pfgyFk`+bMEoDYNpH%P1zI!aVDd+2{(ol9!yK8)5G-@VQGq1IurQBF_rzq zg#6GWGQE{{FlTe)6;sy#G-p`#7c~6`a`EDB;VzEe%+-k-stOBrJgMTA2w7!|-LP0>+R4CO4 zz6tj!8V*ny=*1s26kc^BC#Z}_o#1iwpKVI%SH-{XAq_mXZtvQunP!T)LXjpa?!may zKr}7P>U+#qe~T~fG8A{)IDHp0MbS+Pi=|TNw`+P9tdP26YGtK?3xShQe?oOhV7+&R zq8N(cp!%gU0=g}!7freVdM0O5N)Wo};z#K*pEy;WPWK40g)Dv+M)!JLAl*LtE5Wx# z(1%I!AV%M*Iu%f=DxE$5ud8GbZ304y7Vx;#OsEy;H!5yNRz)NRT%2Xlbdf6QiKrB1 zDO8NUL@f;wV*}xi(=j%zal_ajkN@(KWA!)}KOc`RWR|*eG(phN#f%Cpv;@*e3{y6f zM$L9#_lcx|r^{&&4Sc)Xj{fD zx=D9l0G`dLdU0jMT2c^W(rpeZ^edE`RSlZ#=-g;)3Cf3(jfxJs@n@Felo@5gNB(Jk zJat9hrpgv}!Y%8^?Do4j>xuk*{hhe>5R>jQP}d|0&XIUZf92;U9d33>pUAKI_uW!5 zdw#`RNgV4Zot_O&FY167bDB9$uQY!>@0*{IWsa$wUl3WtlH}O{AxF&P!U9UvatYp; zB+=vJriOs80F~45e#JPhK3)6auYcyaSb8zVm&b8PmIV|P&PsADG_q|p(6P~aTW85M zWtuWC$gJ6^Iyq%H6^_&Mj}pt(ll+qL*u~qNoS&}~*(~5mo4$9PEakBUoM~qPucw%G z6iLQXhjXOEKTnn&by^A-z=B+Z#u%kv6mA5w)TFsA9j<;kmMM7sKrm3;=(-#}hRZuA1j=H*`y`+}thJ91Xv5H679$tosgPO6x+D^#Ba-q!|YpK~9n z&70M^kfmu|>#K_J4id)ybl-!pbLN1u3PE2%n15MpwXUs904? z54k6c4DG=?Kh;C~9Mx>Lh*azyE=x^4QKOwbOH16a$WqbSe&m zFSayETx$*>Y&jxUQM4T)lv1T-f0=aQvq0^u~T>y5~@q3mjq>y8Hwx}CSA4oXxL66vT~xI z{zGwxn|oyp3Kdp?GS~@s>OOW<3}iriet75|$23STb}ElOAuLIgRIwA!{a?{SiUB&o94Mv%`!61vM&QfW z4Q%cA)HRx7)j`Ex;nT_I38ak$JOvK+63E4PzTbCUwy<7$-zQ$Q%quT=nOCpWed%P* zJm>)w{qsXZdYtSxb9+&FOT!{xmwRcn4KGbo&i{#|yk-q?inf7^pb(lgCTi zrKK^L8w90=<5&6Ch_YafTSeDSu8-1RlVh_BYrJlS$_yC>Z4lJeuNhMb@m@FEC4b)M{q`~#rwoob{pTGIrqiK*c?+(t_@ zcp%y2qbpD|z!<+4PkgAb!9{vs`g@Yh4KDok43Z){xELu043#usVU=w6HBQcuJ)Qh~ z>$d2a-*HuLa0^heZ1=60mSrgyME*DHm2Q?Y)_EBWL4KX}k?10MYw}{?MF&inEg}8D^D{4}dlFL#(`sA7vHFURB zpC4Kj-2$2T7D8U+zD!jY` z?a*KK95y_mKFe%y#D>ERx^MTEDEz!*TgWd6Hdq=)&c(pbNn+5J(_8&} zB$p&hMYSLWa_DAxmLy5AVse{Ae-=nI*2LZmu8(M^o1!x#Dy9`lZwGub)}Z;+3wjMC zd-eEc2=~m`Geh4In=6P9KkcXA8`?o<_-`H85nCI!L57#DA9pFHo9ItVJIMW*QWYjv z3{A{%#Lx?}7a;g$OkuM=>;O5y)s>MS?k9CUOf}tUy$9nS*ql07qzWgv) z0d&r9rSJ0CZ1A`MyaTRdJevE(k4(v=aG zg%NVcV@FwZ1Y`REJyQLna%q!|n619LniQG(ifJZ|f_4HJ)tnsgsYkH&NW^V2Tj`}6;{gdAoL z2kaR0$+UW(og_!G5@?~I*!3rWNc&;`yROMGo^@}WdD;`x!I~1Shm-0viCuCf&a9 z2F;4V!vQ^hC&D%Xr&Q_e-y1Z0RNEMn?zqM%Iyt&yT8ntoJI_37zV*yw%^cqLEs6MJ zW|w0uoVQYqB}jjs)T{K^(OTaiz8s-Xr#s|3ytWJV$aRoTXGiI~rnHU&5d`w?O(uH-~o!+N0HrBT)XP#v;$C5k9|V-K}#Y_owQM~iCL*$9kdU!Ww2rB zMP|Y1I_9FVIdXr<_U{}NRBH5FqH^WU5HWA@MH!&4rQ@mjO~1~Rw-R7AY*1o+6X`nO zXu_8vGdzdx5Ell%8G5_rIwFQ>O?dVJz%>Mux(p$~1Ude@x!PiFUepjCyA zVkB6_Q+Vx>+2}>=_2@s)W^X+j3x&@U2=@+x76a-CmWB#x8|JRuti zc#{{-XWQweU#6^*L8Zvrp4*Ki9@I~PZsD0t)X?sX%XuZ;{ ztM`Fs(>bIJYZ0vFXR;$j{QZAdJNBdHvMl4VPl%;^EjGqM=U|JtX!ONd&=oloe9Xg+ z6qpvLz%mfZ+5-Zn+@zW&@_NP;J% zNcNY_8!0B0B5SF*Q$n+*j@djD7Yvr(Q%F{k9&Q{3E;(SUkYOInu&Q)!PrL6~xc`&= zX%FB_<8Nxg?|nOy{?!JJkc?(n32;lh;qSkUblD;NCdIT;q=kxW5n~Gg^a8JmsRwT} z@IIBsq^fU=)9%W=)5GV?c{$pN~nE53M9%&0Hm3GpEOAHN-lzwdY8K z7)U*U`m&1NKV?7U`{9L4BlIb<@)>D@EMWP_ruWkiV%GUx4eIsC($>@aqj8DV;D_~f z^-2tBn3y6U5Pzn!<`@Hg9{#9K^(VG(fghGH*~VD!$kd`Is3W#f15J4HWTk#PRcx;z zw>PH0pmN#vfUn=>KYuUibuWwSeap%=wHtiFonOB^8zS1ejPYh@R&tKi&b|Wse=3Bg zBRgkbSO3|hOApGDT=#zrYvKcQ8sXdx-GO(wGEoiqC|M5m>Y+5RU0NhZ3QCNR-c()?d@Oapsms@S z=O#!*dnTN6j0kX9EAx2C#1cQK7gy8RN1&S^&(smowxGP=4HNU}<*`WC42^~1Otv;d zcyG)pPQUQ80y1a?O@N4BZTp%wdfxfv8&dho@tZjz?>P=d*Sn}aq;wuJ5+ zy)CF+`dPpc?(MNzCBY!E#%8Useh=3Z-gdhtl3uKN3lUs>?o#s4?=7gb-boMY`xz8Z)j zcFUWjSy5)y-ta_`8S)!4#_v&8ikf8YW88obPB37T3T~h=T|+;V6GrL zxUIB#tc#A>L2Mtz0I5g`^s`ZK^;;ur^fT#@iVPdb7ATO8tT6%`N<0ZGrSH;-@c9f# zitf?X3haEAEn5~<5Y^yyjGPZW?bkl}9=#j-#NLj~7G9BKYVpyq92yf@FmZTQaJOsI z{agbyD*|09gpz8W$Y1y1U|)WwzMFjP3#oc^X=n#o7vTnR@v0~9+$_Ip6xW~UBu+-= zH=pVVv5Aw~x7L=EOdji?Lv{k9jACF#R7Ax!DHaDOOCSfcQ@9Ee5HLk8Rh=NIBZiqJ z;qlWjEowJ&8fx86OScQb{|sROi2dZy`4Z$c?Di?2(MBJc-3hsRm>*iKKB3(mSOgxt z-AoBELt<@gWM+5gN^k(&AFlbtQTf`aT zFGFcA6x{(g05+F~Tj1R0sOAW0_d4!VFZ~W*JkY5+A5`YO*xSkx@2p?XRp{z3nqIN( z6c9O~YsAIb6MG#OQ(RWrJdQxIEWPSTVRWT;P2f}Lj)5>pQ(8Cy(9w&vES9h^eOk?jpt1Y3Nw+ujs{HWmGcp&# z%W34i!q~8$b{%>+A%~hST6#aghMb?C8gZ7SePR4{P-@ao{Nzy#kU?isaRq>!YQT|6 z*D(5syqaDjY4q!uVDy5vG=5Y=xHiFxY+NKa zDX?a3%jgePRnqH_9XMQ;$snfOb@Td?kvK_{k@fd)<~f=tUyxM}i!qZHvXErZP;i7z zq3;8&)v>@dVE51;h6K7)w=rt6OqlZ^Vi~|qhcAx#YK7dUWK@6qN+VhQg;6qvc1k9l zVzy9Z6BTzXFnPvqd7)q5XQS?cVsq3D-C13#`hlWe+#ohG7a60aPfZnFCU2bXMs?C} zR-H%4=T=L=Ec^57a2UNq}Y-M z>q3X;&3`^mPEYjnTZUzeo2l{USO4;#uX|*0#i(>(+(6EjI+8J7U#c=|@c%slpZJ`e z0i9_Igzb>XRwYaW>XsuUPXdRKZY)n$B}5;+l{y9vRtZy}z2>s0Wl^w+L*>bq`7IyS z?b2c)r(yHC=6K153jxIqHt)<;{J zZBz_@ajDa|aN(RgGZZIWPzS!UY|&5*5FR_>Squ;h{}>jXVjg6s7E4d{9m;lk9dI8s zF-;20ha9feuVp@+29iatI0FIKTvHMA)jtoVPJ_p`DNCJ3xlf{~htBouRy+h^r7O{e zK)qC~{X~jAcuM?{VmTi2S+XRHgAZtWz)8|B-{LncRtz_YSoM!f50X&lC3Ht&~x@F#rMR-*nWQ(Q=mWvDJoiR72V_umx`!s9R4j?}}BB}{A zXim!OBhCh$4|4H~`RXVB60^_llxru0)BDC>kvyp_Zd8m7I6v5%ZzQF!85>H29m)?= z45WutQgNHY>c!^3r@?2Hcj?~m)PwZEt+k{>grB#=H-#MwJmy~@+%U1x;v};)bPZa? z992N7KzO~l(i=xOO|OZ?p09-yyC4CwOm!sSnMb$J0>Lwndewc=GmitIOU69&Kxay& zcOfwo9B=6OcOB$ii+^^=k5B&K>w$gxh^bpn3_OkpL!XF#8m5S1fG1-IvV4^VR)j4D ziX})E$`372pq*+lTCw`PNnuuFcAEaWe`~b0S4|@jf|Z1P1a9{{VC7mU>4Cnk1r+|l zp1@dIICLnyD6?Dwju(d=@)34$43I!DGEwr@S;xp87ym1d4H%X zqWpcdHEICq(yF-ahlHiTjL~3WZa+Bwas;w717T%@(wfS+Tv;s~4rJh@az=+6kTB4Y z-~GwY1i!JNBkZ?tyg_d8Sl=wRW0mZpm`;k^175$kt^s2sEOf#9aYT+Gm^!jN7%D<7 zItK}CQ`HBwh0s!Qd8~!6v_Xjp1NXdI#FwE-^h0?YePw!rBpnD(DrXf&>yu;;#Zi9(ZzlV>rJ?ZH$c6Hseru=`6jMuq zzbP&&D#NFTzVzcoZ`aI*VnR!|U!&Je(gMXgu&O#3ma5(YsgE~_zFFN73w^XKi>G^D zMp5~UMDJBmYhpTMuoh_dsH#ww6%xjx@nGFAlmNhviW`4Gq#-L2!L0P=c)_X6;##a zuWuMiZ%-Zvp&|FapMu&%F_3hzj*7eO2Q{M*tgyK9P%L1jV7g4-sKKh`91D46n=0As ze^Db{+4OdV&3IX!+Odli@_30|XE)oG6tkZqd#O0&FRrDRd6v-SUP&{TM&KgPi-XgEN(H>RA7JebKm*@LGD)q-Lfh z+@v#U9+8isVVYT0=(h;s(W}5Wb3L#<5M$G>!VE__@@mt=MS_G~jtt=8rDAZ7&HgRKa2#w7t#UXs-cYj#NVK41a_pw$_7{hRD_+DA~Ro}tP}E#+`3{N z)RcVw=DO(l4=@b3DIt7)K@X1>=pB=9xh#Ho>~=e32l>xQd}`;scW6#ybcHm2Oj?jh zvqaJ&-Y%>cFB4Ts4~5lEuVWgcKPIh;V^Az;(p{6ch+Ll{331t;|U90f7S0$^5u;FAH3rqCZ_&>(FB#3O7Yn!wRg zp(zVY9pyS5=Q7|}F*JYx9NeLbt}T-tjTbKRhR4np7E`B{#tOnF#g$--NweDTg5Vy# z3y7dP)kV?Za61O|%5C&K_;0b#Eir!1+a6jg>)UZ1B76q_QDWf;cBRu}^+#!pF$(#0PRj3xe=P3xX!CARn$@SJ zo5GIM4bWrtrvLe%PoxL^-AC5XSGZ*S;J1q%Q@Z(TYXOB|eKM;(adymInla~mN z$KLomuUSqx__x`!(G`v7@6Y+uIv{i*N0`BXH4tE$eHPdq?pHE+f#P4 zUPCce6gdF1PPfUfj47s139kkB-!=i4L8D(ay@K5HuNOD!kbm^7ZU>D=k_7ks^MsY& zc)Xgf7spEu#Gch1CvB?rW72#eYsoeqtcf93ynf85G8`GxwNYUe#FV-un`Ka_=(qeG zaPx&Z=6-;iCOV>?Wq942z$|8pqOr#wc234h!{el80`&8~DD3d5ftF+iUu*EI61Pk5 zC^3-$iA1p&QM}z7|97|1GH8@=pLkApa6-w5pa1e3KbmNhG#)bTUXsOQNpsjv((I;~ z5{m4k;*g^jDVQ*!>49Q(&~Atf_Ifl!8}D?wjb6oIln}F37D{sEkZW*HkuHN(u<9|a zbPzH1iW}Z|y4M4jLC;k2l4q)PS+9pBS<3ZIJqC@N7sLuDSiJILZn*i-@LjWGIszBh z3XiQAmPCc}8D+sM=a^>CKRx@3yiJuY>;!hVE~rSt_I5kx=(`~i*nLal0j7@|kOmmQ zO^0&7v}vAWaxHIH$t>xzSy~KJ-I`ja%A_xat%3D02CuF&g~QG?>U8$Ln`V=JSWW_i zx>!(I?HDI~t@4*xf|`egtqLTjMg!DBgT7$qJaoqA2H#U16x@tnuC%!1Tf{4X;rz3J z8^VPI==CktA->gN&^a|XI7!e1Of9XhZP)UA-nhw$7qkofh-up{E90aHH?vYDv`y+7JRjNF_i zJjM#FmiswPc2Nv)1Lje2ADr>Wz6fUZs#xr`a6hcszgK-|vc6SS8Idd8E-asMg}~Z} z*5f)kNr1W2NrHIq1j_zT4t*$eKa?No4~16Imm}`SSYkf&Tv(Lo1egYrDQw4F2CJ1H z%$)FD8)W|2_QxCKBOb48KCuJAC5pK~k#kfWw)bh&pii|^SO%+_R=PCiy8nY|C+Ezc zKOYOWvb0Qi#1ZbbM}3ogVVP};brr4 zwU}aEMYpQjm8CH`lNvRhLO6`!;cPW*Sd;Uuqt()wJe4yEe*~QBh)ouuiaX0B!9ptD z8xA4pI{dTM4UnOOzo8X6e&o=}UQNt2oMDt*m2ZvBj&f^8Q~!UUZJWdM zAK-5H9pd!-`J1zRcQq2J-K{7Iy-qLlsf5Js7J8ZFAgT23B305BaVhYqcaw+E zc+8;bQP;S{tv}EVI=#AP(Vl*1zVVx%zV5v_tjzd%5~NrGCIs})8Z>u8mij;w zFs#?OF{MWIp=hD6ezWAb3hDF@hCx}_mNCajJ$Usiy{qUX0dn^kG%dccj?UEX55cN> z7!14IwQG!`?5Iu`QfzS0jBI#tIPL-OKiZ^2Zl@i-$3t7ByX6JK z#8^zBtyew`ZdD-5hh-k{y$@&`A+^Pl;fUcdg}`MnALPc}P@P zDmo}~fdkj|AhKd=s8HerBI<^D>2Gy?$|`DX8aeQVQC6_D>bKyoqnP6qIZDN~jynn+ zf0ChNP>!IRw1d(@iam@@BRv>4NNkPXJJY0tvV`!Oz+yob{Zt77070u_y|6tX16D%& zwH1Qyn2$9{vZGTE2C3pBnH^6mN-qjyRc`IYMg#c|BL;<67%LzXn=TZ`sCW7X*0!aq+k}} zPK|jo;TnBa-l)s?#x2P?Eu z1NYJp9%ZO2Z#=qW>PM@AizvFv$_UcCb(wB`|hhxhMo?sS8}c0 z81jrbzWMC?oF<6V%g*1$=YPHJImwZ)n}rq|Sv{K+x$+aDUJon?&eYylH)`~)P+#2t z6!wQQtCNWtUTGu>T^ey)lobK)+Q;;+NqRI@Q`PtBc4c9-dxAhN0SqT7ICn88!0=bN z&sqJuD#rpwE?K+$jk>axJ3b1{Af0M`DezrlPWQpEee$XZtg6>viB0yh^hRq^SgY$X z#{u63Ot^8s9Xv>+I|tVxT=g&{$yq*^7BM>47f4LH%vtM&dqX?uWw4?Jik9%4$*56s zC6Hhuz6w==>GbM|6}pn}TLJ4L&=KFFsiqqOZcV-ruw)3{&R72{2l>VKb*WH<9FhVS)q@6X@sC+fOl&{#yoY-?NHk*sghzdYANAp#7e; z1Iej?dU1!SAm+Uq(JBTj)Z!&iMSa^15D3N46n5@4>1t&i^q$y4KeOto$moCFzZizi z32hKB461fvW27Oy125WJgT}n$jS+LFP8~{2fWKOnu*3u~_YV7u^m<&5FaZBT&ZPKh zIU#1?D@68pJ?8OY+lU0onIN~hdmU$iGddgxaMS*TX~oTIn`qgvChP)P&tqRqnVqo6 zrWokJX{6$=`4`0W2utDrj@XrIeU;QoPFbfpuF>C#1PlOC%lfF@p}iheU{;&e)sXbk z7_ok46Lg|!r}MlEfN$T8|8buBe!{?wKZ|3w!^mG<4IGLmhQHdCuy|sym#-1?U+8Uq zYyCbX=K>c>keS7#!-apM-y>2gJuIwOKA47Q44Uo2M`UiUpfaMzr<~dwes%JCVL4Uk zcin%k+u$`6Xd%kn1+Vmf9+%@t2*umb5evmro}?7KY1U$+)KT;9L}L9BiV^)KsCc#d z;gAD&6$89-q*3EaqOX_C%V4JbCBwPw6em-Ao;1 zNmWtI0SY8c9Ok2AVqRwqa>eb5&6S>~8zS`a;YP2VNl$#z=@ji+aQ&uemn-Xi^CjDu zjPWgEgQh`QAV3Sotj1fgimferYbf4o&_Kj8*q~Vz^|8k2l>~lAyza9AglU?h+eKAw zgMnK={WZ|0OCTo-Hh=6`{Q6=p1CNysOF`@f!N=04;EJyEX%%BbqP@ay;Nw{DbJE9c zN`!SjoSb{s``vYdUDQlnl$tF#nFXx^^C2fcy8tfk5@6kDA>>3NT!%o z6j?#VVbUz3CM5{#%?^%VD!HbP}AIW>g(79mEYMxycF1d*2p0 zniyQt0eEbdvN+E$7fm0R$m{f;beyfracc-uaqO|hBuLUP|yjg#`7W=k~=u#iTzX=_A9 zKqdL<)c;+I0w6ib4(fjjemQBnO`)W#f0jpf^H`xka%DfO;V8vGS;j%EGJP+XO8U-w zSAMo<_9AL=aI)l*G&6d!^-!zhq4c*LuBra=BI>HNCNN#LWa74QmfLcEY=0=9dT*)c zak?oYSF?%iq_3(|)unWTz!20J(Ly>&LU2vsps~kYXgn7m&TKizD5rP1;b_+1LdQ7L zWw5YPpobaty+L!?dw1ZxHNu>+t-c3G%}Wv|i3=$FGw5gO^gpiOgPUwZr)t{`0w2j<)L;WO2h{>0%{L z0%Q#pPR#1-%-n0dFdE1;!{W57r}DG1l)`>H4n8kqz>m*G55kD7wSSXQwIj zm!XoH5BR>B1ueGqZ>b0|<^S1|*);76WZLIpY!)$(?I%tA`rti-fLB)iAD zPyK%Fuz5Y~=gApF*xxZcdz=t4V%6_^zpAl`7G2%XD@X>9MGJ5(^;;GfQw;PB%ctUM zL@BcJ87Z<(xjvoF6Bf*lpQ+y#R6gUfe1Tt+qQ^&nEh=Mtj$luCm)fA&6xO9S3od&@ z=_d+~4w5Kb5^_Cq(}Zrt!STHwkl9RDMd|lPBj;O+tMi(j!Hknq!=+&2h8fSt$0H>+ z%>1cz)PIshZs8gpXNTn3A#4l9Y{DNK2L;4XYhE-u(GMGN>Y(eT>mu%v z+~8DUm9$55QJ5xPJbZ-vT*2g0#Jps%cgFAjWGF>@JoeJF^eAklbHU~~G_Do#%sLBmNC6iLsCm?>a)5=_f zL?=+mi+lh5P=!nGHAMe zAZ2NnxOcAE4@i2fKnt7Ym zO+UkLzCCT1K#zC*nlsRxv~-4g89UGhSR##RPy6+QZ`r_iYg)}+@*$7G*Jfv~T%ee9 z6gfl1#Y?t?R)krbuV>R18U$qFM;dn|(Kj&$%`#Dy^iY@{Q|2!+3q~K;)IhrXN?$C9 z#=Mkv=)GX;#NHuKh4#u=zX7~&fTRng+XuN(ka%Gq4U7y)5nH^r_+fDvBs(09IiSt) zHfWI7>oO1{r@~#E0C)3R#65K8tQu`;bWPw6Uu=Jt8r}^nd*>WH{~xgBl=$O>t&!bp ze|q9|Uw*FxTs%|l^V%&hjVYS|{XE*d_0G+qQxDYXi=0A))1$fWLQaSn@ljW(_H|El z!X**Pjtt@QSUqyJT$ZK;=~JivfAH@+1&U#^ar;Ax6DUT0Dg5mlo_#)<)<5aV0&eST z9>*Bd?G)A;ibsY1}E01=kSq{!83Dzl;0*fBgJ6bN?b=N->Kl z5V#xR%g=n$T2IffE25`BfDs$9Jetq?zX&G zwMKMJ-W7XaR7)@O>Z9S3BMA zrA~0iydX+uAcmpJ@aJ&>&Q66v=0CRzc_X6WE^|nVywGYNq%)Ai2*y5OK^GkoJ3aAv9DGEuXL&3W%fly2fm->P?I!+dknY~4+U^a{kqh3L zsy0O(+>;9uKzKESY!L(PB?tyx^XVaj{yQ{!R9nP2CXSb8i(QF$KH0tb8-K{A^8?n* z%X<9BfaO}39AzR(;(AU>4p1TWU*CMTdk+9%T%M(=7lB67FJ-G1gge|#U!t0 z7JcK>w{$o;{_|^}kXkFYMy+}?-$=0y6gfwQ?xHV?!hP04Efk8kAgLo%X4detXmyeU z<|MmhWxlJ04};W5;D%eZIDTi~UP&k96Oj{CjURf0Ku~!Rj+08?cGA#7mN3oiZD*w2 zKN#5SkL-~t?k!9EUDxdoFEiUpdzeNzbn6E-kiG-}h=@ zkTqTXfS&R<+90}D-(iBn2< zHu{fj_B;=lhiuNz?h-uz$lsqmq8)SOcYs}_@LHvv{&3o1pqOwuULDt|ve4JXBztH4 z%Q9`+@&EoTb*2tycYb;OuOxEDSd|o3`#m8!dT824v9QhDM1|%#CP^EWar|;RUs@j6 z;j3wELzZKd5|8&QoOw?P*a%v7V$REqL~C{Nfc z$C|rgn z>M==I)$()fS_L_DA9)zm7qr9swsT`(HNAg6kRAs%PgTfCtOTRz5B!O{9K2N0=q?X$*r;cQY+3X4(VZ~n_{~t@-Y<}scdr3 zAV(ZB95(n4xdYYpWs)b3@oV&}qkDpDU83FB%!3`f%LTd#Q{(PQ?nC?ol2$=G-5Zo< zw+dPj@);Z=YH4vWlzvN790nXJ?c#*TK{Krwp6`^OcTw*aril;04Ye-kU50FK%MUEA zamS-r_qzGo<5aIyuVI@?yUc(A`_ps|z1jCPfcJAy35S>gHaqAljkGQPSQHq|Vs>7A zLUN2BHZM<7Y1bu+VJ_INmTJqpo=<(54#Ux2jRlSkFl`)$70den$wvpdJzXWMNe;Iy zw-tl&m>wASQ!J>4lmJok+-uB`XS~y1H;^f-@yG@C^i6))f2r2W^*$yM!d4o)ud!F4 zRCJ9|AMv;@%At1)yA_+E9V%JTE?w_+Pok*=!LTCwL|FfUxw9hqxdoU&H=iS9W=t6% zW8xp*?*0B(=Z$Qlzjbd9$+hA|tWFOLl70i|LLCu0t(RCL4nJV8OuSl-1Ye~P85fB7d+4-)i{@aY%m_~09=IDYhW zyrT!yK%%Gj{_?);G$hIYbo6z&^P<_3rDjHh*}CP7TbQx-PIgPzocMCpPoVB6mX>c0 zxGqu`I<2zpg2L7T`{RL4ax=uGxqXXG#t^q*Cw=$tBCdA5?4g%B$dEUM|vG&SlUAt4*;D zGS+=9^GM#M$b`gJyBu5)jK!c#@MHO7=rPSTN(Sr^Iu^smWN?9YX3R-7!%5?~7{e!-X4hAC@;x0^ z(jM2%Aor~pD^Yq)+d~xlIYs)R76#Uk42NbmY|*`Co9)tE)9rhq#vPBaSKBZD$pwsx7s{8(0x1@-d1~O6=%F`e@gtx03A;^;p_3Vj2 zN2P7hE?OLGuSOn4E%NetTjt(&-Ytw_JLRcfpe-PP9$v?KL9AFiLMx48X<>_Q3Q9u? zMOaLaPpFG|*PvIbmA)l;phyeAEx0y_ge6~})A;u#D)!0O$_7acqmB{d-;$z?aLl4( z?Xuw^2tV703BMqX+yAfo@!OORfP)T$=OB^G3Bgvp72mFB+F~g7LyAODp}29@Bouag zBO(qiu$YM&*P57Mzs7yuQWt0b_AeEaAyl+tKo}uZ#3G+W#pXF@NDd?}H7r+XHn1YA z!n0d?5a=U5H)Ck!#S>=&G@g0NFCNZpclWZtZecqsu9PX(vlTlj z7Fuq%Q=#3?^@5AQL;|AUZ(}#zUvESG@BLRwT@Jc@6x_vMev9n6) z9{wP|!fTbv@-~GNNTv`j5lcGXlITDp|N9$_B=#jsClu&GA(dib9oa^Owtz{%@=V~5 z0mFb?9OsFKBczjVr}M;>cAfMSpGeOJ@GR8rP!Mx>VXK4XQ2>?Ni4-Bf`t9%3fB5Nd z-ueB1KW(OcHsxa+nav~^AIP{D@w<2WfBPdHLN@>P%{Ry;EB1LG>Y<^7VsBIA78Q#1 zznLmb*JuLrn`U+ae*-3}?=Dog`p1#pAnaYbGz%Te;#aEtPLoI|w8P!fnx&u%juod_ zs?Rka?l7lWya2_!mq9X5-6}{6fKKu<7rgVZRQnT(s&af&+@k{;Wt)SFnKZj?9$|Js zAat6f&q4aEG8g;}@K`j2ckav50r>@`t(1Gw*Ev1w4WT>{wBFPmMOAgzO;Tl@O@0N zz+BWug+c?dAc}`c@O_Rg&Mm%Si;m4oCoR5E%$%sG5<~-0Q3h|XM0-))6WpoFQk6r= zGqxUA(WP{yUp*)_HtL=p^eq##Ku_luhq&?&TE(;%2SNFHe{! zL{2Q!uCL;qCQyr>rK*QvQ^h!N6?EG?aReOVP_4WklHd*>b~zw@5-K{Cxpew#GOG|l zfcii|f!o+20;>Sv^Y|O5FXq7SgVz{x(tJTGvy2FJ6$bVNY1pd{F37@C`j33i3eMT4 z!rprfi!#Q$G~+2#;bO*LY0^;t^Y&?vg*prLUS&uP$>-)VTd{3ArDvNeC>D5C%Bavz z8tnu`O|)%KTCM^N`*bO;Km~(_Di7GwpxHZ>uJBQ#z(qef3p##U+A~}jF=M6WE-!Lw z5Tuihkp6C^quuN1KCkGx!!`%$htO+*WCml~4wGg^?!(lno8-Jo7629yEv0=&9 z7sRF*De%fv0UO>SRY4#K|6!L?4!zBzQ&a7bsKA=?skJN|a52#-p3rj+A8$P8sxvF9 z=*YulhZTF!hx80Z0mTAabPg3-0-@on%hdSw#)YI6YNNmd!IGGM{}zAbIqXmj+x+GI zzr^y8Fr`^~+!p-UPP#&{6;2G>w8I0XM2&sosp2Y8tb2BVxy-?+IpKhSagD>MAHB@eA;1wgQu85owop+dh`JLyINk z9Z(HrMCrit24jK{F$)aF8-x=?rb&}tU$RtZQasmuxSJf{=BZe5Z10ku0XjjkP@{K* z3O)2sHzj|;cwTM5t$A4iYlN+W=D>3M8zf2EMTZFo=U*ma!WsrDvqbwmtNgY~2R(87 zbizB8?zhLQLpJ+)`13f~XLo|hgsO{FuXaI-a)tKcdclgOkYus;-XyOqULoCYw@;ER zK28@%>gakwmnfDuuyo=8ognbWwNs-mj)$#ruWf$ok37R36|1#2M!HULSy|7oreob3 zL2e02d~uP=6Ji^84%FtIos!tzv)9aFuAexMB4=LPGeW^hP)UKFy##oCsE1oAzvxL85 zIOQC0GwQ2vx$20Kzt1>sew?%C;bb)ata!_8MAbe`+#moyHo-juL@vUFurW@ZX8~7s zKA5TQ>o{S6j3a6)#g&6x19HjOM}`JCrFDZA2ul`$e`%92!A5e$4+VGE8(v+ zg{-kZqg*BY*t03*0lQPWO8A&=3aJy9@YX@`-EHwka+x$Pu3lF8O1E>l*fJR6jF_i} z=;^_PEzEciey6*-%3p_=bob27q>vk8tk?lLqlcI(iiOzR0buYDqy0z~g9Smd2HfQ$ znFq?1AX61;)j+eoOC=8_B(c^7lY{CZd31?`yRj(G-EQ?{xR2OUp06j(0vxD)VVcIo${&(Im6e6(N{jU*ef#Fv&eT(gy zrNCphbxAE>TkM@Mg}%wtS8duXKfQg*4mdd})9$3X9evr^8JYthO1H3=0&B^D{SN0s zr*P6|dlqss?SjiprQMLClHU}P%hUwN^D{xQ{E4q+mW}5o*ffAV`+`rE>%j>iRPzsB zU;WY(wZ3svQtWuqDSQU*bRxYnm#v#IByS3d7gx-VVIN8(6jk2C{wWI@<}`(*F#`^l zy)QVfpHadVI5r4&N|6FTVn$O)p;IrfSDxX3Z1_zfcy~STiYUqH9FUWh!EgB6ZN=Xn z$*a5vgq=Xt3aWrij$0Nz{FR0|U7~Z+w8ewq(=>+k&}Uw|DZ#sLNy2^Gpz@^QC?f5N8X6DLqiPxrGe`ijosZT|Y*Y_iviv&k3q zT-aKQt)WO26`B~(P8W%z#m&-GF>Hgo7pZr9=Sw>Sb0zABatOilQf6yHvx$lh-fD7F zlE%lxGxj$^;ZmdQhz}MUR6wX76U!(|i=9B-pd_j8MqXzX!mNoMwrNx3dY&US%Tkos z12=4=rT#4q?4&2rP;DBFCTv3{j5tm(n)Z5ERN2ek_82~+kcN&hd2+zI8D0K8i#E?W zC#$1lnBu@8;cC0gnN1<>G^pWZ*=EjMkxJKjA&6qT{PBry3wDc4 z5-StN!iyf{gcYjn_y55Q(wUrp*}prDlyci+TXD7-SfWQY)E=eSYKl}s`+%YyI;B&g z;yc%)*)v(u;{jBw^h1}9fCe|v`wTcC807EgorgjO6uemPQ}5Rla?IztWV8JmuLm<@ z-D}(~2~aQ;?@CdQJQ?G6+rN%RrKXQn$AGB*;i7WSHfUlxJTr4yfn#6r0G;dkBG2YU zGdsZ+gBP2t7ysfH-0}71uYGK&?)Fl3q~{gUJGDUU35jW6` z4LnPnefY%L$*}mzYPUs~^cI+>n6n{$RHF(J=A z%4vqI*LD1#&9A1v>@nj0pjzH_Ai2y|Z6zzxpl^B$b0oOZ@2s?%pcZMXAX{~J;aS%w z(DYGD4ytYiniZhWA!Fpk1du!Os6~HrLvGNFH!m3ayId+AtT;1i#1r1^P){R01F~iI z0{s-UH}YcLcYCLU>;#_Ap^tc1g1`XwIBHdwvS@6EZ3D5eo03%TXb0$#B-Mfl%5pfH z0JcUiO%}t~Tfg{J>8D$XA`^RxNr4rI^-k+8k(CsCkRto3&|*j2;NV^c8>!U;4sFU7 zXlzbYL^}4tN(ar>jj;5^^HSWC1X-&4vhzGG-Q5mn=pNF9?;0I@JknfIoM!{0#?G;L zhbnh;XG)x*F*+mQk>mt@d-kPSsgR(Cip|=9+R1BkO%f5K(V=%dS*UQbHq(4I?D*r$ z-ZY{uMo=S1ZklaF;7V*khR!n76ga1PBIyjg1bhx^ZL6ZTYFCYd0+z47oG?Tti-8wB z&j}*abHDHPjhDUB|F-jr7QbZRRk*vb8#-6}NiPt4@3ZRxLd{3?S>;a0n|_vCJI0s| zF31?;af_YegbZp0)9{W;xA>e2{o5w8hg*Qxid~g@y`}01#U7?e1r^#L!`={J?>Hr| zqDw?wiqfDm7Yxd^OY8kM`bI(2Rui0CEgZ1N|ASVj;^Go`A~7J!E=3t-mqSB88>~#4 zk9aK=Hq}v%pe7g3E8_J9A*07ZFXV72;_sMK%v*_)bPx}8!OE&hWS_W(W%QP$7k`)p zAn$H^|J#O)t6cnD>)q-RBV4TOnT?*YJmj0&_m9_rcMA7RovKW*OP8eiyiz)ymmTnF zU=+VfbP=*nwtz}}}3wT|Y-c50_P5Q`~1PV-ev?=l@1%|Hf7X&Cq9OJ}* z8YXOBl2?TQH#!*1YYHh1s$ovkhxw(vEFmfqWARH<$PlSpe0JWc;9SQ&OqQ@yeuBQ> z7^7jF&s2rYi*@?Q_a5l|WXP=XFxULQU@o}B;{^kn>GD_p`0EEQI(Yo}{IrWC zi5onu*qH%U+)?nzqgW7m%%noY`C2OQ8aH*3Qyd>PNwR1RtDGb~5|GMWKReBJ1GDmo zcZ;Y)(IZ1*GVo;L_~~<61(~Wcmm}T{vMK?d&a<18Yhru`xr61Y)fhQ2KKLZA40OD< z>{T6D7QO#bE2;d#fTdXvEN3YeIDqP?(A~o1rRqygkqRJ4^aBECpA_H=#}eTb$!n_nE_Q)cGomQ!$5c;jVAQXdbwD?55v;&-D4Dvjb zx4RtM5X{k&sjm>aLR6|Y%4^k5T9u~8I>q(YOIp0n)<&T@b~w3fZ$fqAo^ zrp|&Xh%j|5bKW;}27LRANP(CWAGiyqgSrH?c5+k za==dLFB}8~`Tn4YH!Sn&xxM9Q?|amdS>Nu+V4Py8*1K?TyxR3WL$G{7!YxKHui=TR z2F|8>LAP@XuR>5MegK8w8N3?M9O-f?2~LtCQ&c_gp8p=E<|>*IDRS}LD~UX$K(Z| zzxE{ypm0LVG{5v$9=)qGFTZQ}nJ@W-n^$Ya`8rjGTmT;~2VS>3 zK(7~K4LBCmRNCQ_dp!mm+Y~5~HEe^Tkvp6Z^AFE?;#d#}mB7jz-zSa(_Gx1LUEw}q zOS9-rGq2liWy-x1m{qn%NVSjpgZERG?wMC_n*plT81<=fyKZ+Dq~KNyb5tlo;FRT5 z?EUBW@waC=6^il}j?nv}{sGo$;eSIxAcljq4kR|*WjMEptX6F6yRW4!m`vu$is5G@ z^Ms~NL)DA6Yqc;}vNDVA^{4<&?}9*epI3!XD&2yUUFL#u7&#ga zQ~qkp^01{~317VlKh?LKA=M3AhD60Nh~q&N2Sa&S`ghLzg7+n9Hr=S4tPM27(l}~! zF#2$ED`ukL7oSch#cjQt1Y;>~l-A1d9&kXB`-dJiOuzK-oGKAA#3X9kxOaP_hDDKl zxA*W@HhCT8WzmNT3iDOE7P=R@rxVM#PETiX;y`&ighdXd$!E(Z6VJ0^pcskg)wpf* zDPcB9)mo;1v^f@@`>`j^6hw^uq-D=$?^u|x)OkK3r+->O(q|ATMH-dPE1_5*)80*m zenh8>w=YQ`$pO!f5)~uI-SR$&-lQm#16rWz6uVIdc+akti?6^Qw?zC%9>pJ^HQ$xJPQk&)@06jnwC zvA|DGSg~Ha98lvW`En8Qo>FbI{ZO$VGuDX#KvO3}YMJ*f31K7XGY#2B9=70V@lok- zoODZ0^McFokpwG_!`ox|-}Zg91z}MJaQ^gjXZ8kGb~Mc6uMkpS9Dt>&4|U*AsWiAi~IEF@WH94#pd)hdL)|USGkFV>@O8BPl{E!^BVzY8x&#auM*fSJ4 zNrh%WD|en}l?Ww6lBAa<)xfFS7;^k;8jX=d%OAmr=-b&?sEkzkVWAtF_W?&oR=AcOE^v>WV3;&YNULiI757OCGaRx6R zy1~nr9Jl>MP)Dz#26#Y41fq<07v?YDNheRl`~`EL8}39@bmlr@gt8*Qf8O%#d4aB8v;V-HCFrA~wpDg4m(2kv!CQKl^3>o%nh z0GIb~{B4>43a9sPy;04k@4k7((5bcB7R>9_3`Mmg`-XJWcRkr{k17;|ZYIQ_d zKwYisqRWD}&8bmoxD#{e8YZ4p`0RCyuv>{^;E6_Mlk%kPO03XD#~G_OlVmB-rM<&p z5TeoVMFA-w*Scl{T?)h#;* ze(E42_pI2FiPTFkeonFd6nR92URaVQZWSbect9^G8mViTD921_n~w}mQFh7Bd$vL@ zCsC2ei&X5QZ+LXf&t>wL$3kx-YEJa>B4+~G&~dl5vU=Mty2pNvLxlk7b?G}m*DxeM zU>nUgD)WThwwN7)>M-OM>fo&srl^o(s1Z`*&_sZJ*$s}_uo)`_gn*(_gDi`#62(Jv zQ>m!T1;wV}B&w^oG4=9R=t{s-*CkCn^&^KU*|1H$e2o|Wf)r+-H0`) z`5lyW%68F@B&krPl>;*6Me-I=4Af`UlIY-eX?gHY$S0i_74iz9_!DF=Xdt{2p7hwm z9N}Z5Wwmpv>%{Dmfzhx8tPFm|ggzRF)jM}KealdA*lOJ)BND<{s#byxOF1;AkUTSmv~G42-d#npU#y5B`Q_4pxf`BZUhEfv6apqQnIT zOOXd0ZoyLUp#!=h>JqyWJBS8eT~@#ww$Uyb<@73HgdHqHn&K^>nyG|ZEZ8BUi!*G4 zJ^L4z6b9<5yh;?ORCo#q7V&H%=oRNd4bJ4v2nYN;zkV1$Gigw9!q4;>y{^%Q+b%9X zwiPd0MmA{`ivUaC%I0B6rhVZ$`@FZ;P0-n$B&z? z-3-#M=Kx%$*hY#pV8|XyLPV(W0M$XbnK);g<93;p@GdU741QvkDqq^*(W1bjx5G>& z1YU7>0e&O4jp6Plnk|*LE$R)zFQ~c&>Q4^LOBCW_kVE&73KjI1Y5EVZC<;Xlkn zie4q$tkN*)Vs+90R43)aeg|ng@Zn0kF8K{9_?mw4OugTJ1=p z52x?6^~QMUZ%lKs)gi0yE7gA^Ysd0~t$5qMQ}1h#K(X-@iJ?MKVhcpJu(|E(GU%(# zp;rsFoo-d4BR;0X&tyD|ccbYT9&Z%$Uo(zj@qhjF_bY$ zE>7^GPWK@ce6QJ47^u%K8kX_8|~I}r-VPw}xWTKkWMkYpNK##`B5H189;&Iv8k zs2^3{Fw81h?Jmejx+{&j=?TOxOA19Oms3aQz1s8b?cR+c#jm2TdNIU805CQ^rUS#+ z2$ng|2^dt3Tix3)J1~$y;gRnEG^y3g26(U~9#*CWWbryx;Adn9<_7MW$~I|J*B7nD zunYvo9+eX?rhU^ZW5GY@T+zpGzV!|{V#Nu8TY7Hi1&TdSkuy|iq}N91TI`y2z_r44 zOVFO6-EO<+R#Hmi=K&FZZez~7Cs0wXXPVs%*~(t3@+u>oWYA(pL-A#v7?VST{uqq| zr6b-6xq7#7Cz}gW6phOL5WrPe@zMe!rbxGB0+1PRrp<0Dt9^>GkrJ4`o8F&{Z3QS>3rpmCm})wYm!?3loEMgQmwD$0?UkHSUYFzpJc{^vo*j#aY)YK- zJa>5G)f~H~kTX!$P^TNk{B}`0C6h(~$C)+tc23rj3OiG~@;f?bIpW=lN961mrj`Gm zo@;%bVy{xGtJPn<WdM>n6w<)kYU)?2( zlx<@nyweFWpL)ng+?2G_NnXeR5aSo?(BPm>5SK5hv;$SCrC5aC20dG-T#MB*M|`kZ zU){v)@Wx`-aF-hrwN@Y@i-x)`IMpt=;sjmXz4DKIdgX_G;nyZAdVriFKB$RA2dwCo zZrPoN;qE z%!RM$x_~1-YaF`#fwUBkpgj_(pJ3;!38z}h-0Qz2)84t>w@7~g3w<*F&R;6mIA^H% z>xR2+F0Qc^+Y=-Ea9lTV3yxg9K--7aFv#PZXO~YQZ&AB0`-HrHm|FO9qx@fgd(*)% z*=)5=u93uY6t582st$T#FC^4QMr|8c0=40(nD83FB{g|PR-Z;+0=d^a6w&ha-k=~_hZF{ za7H9;S{)9|yYFABI`2|1*yiz=G|oQaQxmwEm$|H1-Wal#>Urz(J87@q_}aw~kZP27 zL!hLQ8L+RTKVFc;;YY<#qF?Spa3V$D3Y2qiUk6o zBr5cVaD%j*YMHaz{kYrcyRI4qze>u>>Rx z$9&KtYko@?FFF46YoCzXvFhBc7znL;Kxm}c1`2vWLT^YO$?iLciFf%XeRcFPc4`bR z%KmEHSF^qG+6g88dw|yd%!9IqB3?B_JF4hh=#{RcFSyo9j?D^piTi3($N+DjqzBlG zw)s4CDUcL6LOYSmpi7c{Q^+-@V{r>8X0M{&|EsoK_anB-0g{lCYI?soTysyX`)4k< z=?sS)I+81O>qNlwoT)MCgdse}|7_fySzw7uQn`kF;ZfL zw4g|Pf$AL>U&EtT=)yX@7+WGRLS^9<-S9=3`mQa#qCfX{aaqo-_v@C?x8LKJI&(37 z@LxSh`wX&1ueJ0M#r9C-Qz}%eDTQhLIDQB3DqHLwA;dsKAzyvTuN|mpD_t{}wJp$; z=x5B;j1GL<==2 z{>RO1m7tOj3PJnafC)F7jw3_z23eG2mux_Bhh%yjJ z>2u)pHJ>dI-_r>Qqs-MthbIzEL+S}hgAAymJRmMkxZo7mbQnIJEj2 z`C>GcT5rla?CjCE4ISi{Dw=P^4c??!CCsANkt={`$i?9Ipc+2riLrYP$iUZno`a5+ zcF=r9S@gq;)t3U%<*_s~6}O7K(GH8D-81)R<7<}N2hrtUF3waWZPb7Q4ON1>8d=Py zc)8U~uSsjJxq)E9c$uezzco#_SgA7h|A`E8Q^m60Wtc#YYLwR-O|cs&vYra9|MteW zR_4B%MWudw{i}z+^S{An2=ihB9!!`$^WRdC(7fN!tL5UBS@GuDh;zQjxtlH^hrBaU zLqL8-geuaZ|dLUN%l*&39i*M z2W1pnN|9nJv{ZDBL0w-hDNm{xBu7xT3tjKZz*DF{p8_=>P0A7=#mt?*+b4tfI9Ofn z)43pyk3B}o%DVX_j2c_ufyLdSgt_gkEz?D&R8->xr#qiAABIea$7qkk=<#?n&gq(Q zpW{fj<2-jAaPo(LSVq#UxZ=M`4s<9y%i zil~9c%Ek!Whe18`b-xEe>u1!9GI%Ei$L%9#AlGI39As2z4_rUHDI}lQ5VVl*YgV!tA05M%y zczRJQ+IF=tOn6>-PXVzYZ6#$Eok`l1*G1TWl~<{g)=UbZo12V-3+YFUnNXa=JDDT zw;*@ePx6Ja?rn&%NgtpO^qbUOcE7?vU*}aep$3z*fBQ7*Pa9 z#}y@NF-n{#Y*Y?u>NN9&Xdh0JXkTP@!hg)?!M<>sxPjw^!7VTR8>=%P_Mi7% zC0n?e4=Y}B_UW0A9Eyd#ayk_nChQWG24N-wGgoR2ODYXPR$wlL97CI;3xx9uL6H%- z6tF-Jcqu{Q6XO1K3zOrPCp_Yl;BHama6^jW3S%~4JU5V_pIwHOg{)`(QCS=2PqX zM1pr=-IXtl>fswE|3qc7%{neh)Jf{trUC{Lu=dyU9dj}`N%$mj#zb#TW8=M*>PL-$MzSO zRQV->?rWa788q#tj&hQ-g)qm|n=l3^bWC@BY0F(O4wLc^KiKu;+7ORe&T&o(gvz0Tf{X? zF8q^CV{LUEoo;{7qa)1 zj02~={I9-am@77tWF45-t!VTg7RC8w)2meF%d6;BOtE`o$P<@d-W}Qhk*w+7vhnqP ztTYL+*0#E$U)=XLBu9VVU$x>@#Ry?~W?;E}{#*#0)1O0~^!gbC_N@*PGqwaj2BNPt zCZECT={~5b%yG*CA_7!WXw-aO13Gs#%zEKr5+gnqbkP>lh;FF8&?r0NgJ-+wT{ND* zKx1lfKzL1<&)YKBTq}$j*A}gr8OJ1WxnvydJN_kP(-(2Sqpc6n?07oq4&F7G`2XAQ zjE0PHNF07QEAPuyxwu=HycDS4fNvzYJP?If|97BblJ_xs4yWq4e8Gnw8WtpR$pBk% zuEt2IQYYO-U!eO5paNs^F@EjzV^XsW{8VH-PE>q6YiMbz_`02j>9|>XWnQwfS$fh| zjZ}3xG_O6Ld?ADj`?7v%4kUYy>2%w zmi~J%>9*onYm8pwNZ53CkRk(AXp;0KXgi>K7fxfiPZk}uwAEje_-}XZC)h)BNQEk5 z!!~USRBXQOKMa>+m{wYQdn@p>s8OR%dz*GFtrA^)P!+w~yAX0a<;p_QZpk6<+m7u} zSzj+}VhcqGk_Op&#~9{}GRX#)s-e!_^FWen;hDVktI?B9X{t1VTGFQDrWu-jv{;HN|>Ox-Xt7 z-)8fDM@=8exXUocmdSc;{;xm(*#^TZNh8IGZ3^65px$4DS2p#O)x-L2xJj8E<^z7$MA;<=T+F*wF>bW2g1Ei%W{ciq3cvB#w z7q~F;!E`W~b|)vuSZ~(8^}h<9%L-X*M1UB{0BQMlGkEI*)(6zv_VS|Hspr?xfW{Q8 znuad(-o*{85B@f=S)_AqB9m|Z7m4N;=(k?`NOtQjk2@$fi6UF5P#vWskUQS+Ojd?1 z1$wR|rp)CAI6qi`-RjW5gHzdx!_MfqOjRwDDQFtQ*qIkPV~ke$^WXnu!AO~b@1NF_ z-Q4hD#UZdo1c{xpR9F|+E{)??ONZn!e&>aIV3pO5QpDGe0ZXu=L$uzd zkF1rMD_m!ml9tin8UqItJ9T2?Di661J*l&Q{Vqx5h92u(N(uA=jdH%TC>Ds}QmN2p z>7`l4Zt5pXi|sUx?%Lv(OS9Uc;yOMk*Yn9zNDvY&`qd|JLE;&#@5;$ek|?PS)vd$NMSv5k-1Xi0+KC4Ei%t zRGIW0XcTK>APS@1Xjigv{Ime=X|2jOmf5OzvPdb3Rko->g_4cfE{1wl2fgZHQ{Uwe zZ0YiPd4ydAHNeXgeom)~>%36Qdy`l9qAl!d*@q75>;P<4%b_a+_xM(VW>q-9QQ1Y8 z({)}N`iN#Akmw2oO|)~~7re6rvgyOnIH3JS&C+&xvs9xIe%rZ^bi!u)Mqn?r2^?d( z{QK=2{BYsS6GEZ`R?S@w7oh4>uaj+Z&>L-?%yq0q@0@>od#+*T$!dLKBUz?uX<|UC zxYQe$m>Q5r5=Nzx^sE zO!&li$YxjIUb*&YKTmrE^$ddIU_6)O+smsnTUDmn5n(2C#L0oC-buQ)crt8+R_s6< zVI$n*bp|?D*V^9&i3bpHmR*()DAWn=IdmL96*{7C`rURqBF+=97w&Y`1e*Iv6%WJ7 zo%A)I0S9D`S!qe8jmg2vOc)wtnD{;C>kV0Mxnv)#*cURwiHqgE$ey8oTD6%c%oA3L zZqpjAa0~i5W4wM2uo&ZM%bi-&nY-GMQQrtjJ%&DTz1O6JuYm^hYGElPg#S0(R4Vp3tn2`A9u7_}2RL$R>2K2C+APl(0+_Z;`n z-T;INkdiE zyaV*u0`UB)*KRED@|qA>#+|WI7bgR%QNx)1@WEepzwu2SOztkK`IMaG1{3SGr=(L4 zCRZr-5=B0uLQ&8M3b&n(@^(3ZRq^fgZ;*v8YEwQCJft@-*}t@Hkvi5Pi;i{3rs3e2 z&njW2J^pML+;P*K2orAf!V`UVmt2Nzj`{2hyy=<2tA}7uv3LGlHSY3T+*;hATEeMK z*&xFNoVr|cPw|<&(iJHjtGxG#)%aU;d{2{J-Y!|IpkLYxqSEJK|9l?Y>-eBl`eXYQ z5A0UvQ!7kBC|=0?H@BC)UG0WvJZyt=AfuiZ06UvhaTd@&#o0nsM3msJ zt<=*}XPUJrX8d={b-u?R&3somEl|@1>qP5_&hqTBNjpo{Td}{BrDsdFP;3H4;!)Zt zX=$-TtRhj7?4BeoQC#M)7T)l@r^r#=Sd0ZJDO05KGBSr8kTLSel>Fp`3~Jujvi7^_ zkTJ-fjwC7E)+8$quT<#asE}eIVY!P6Me(^dXnc($`2pbieoluiO;MJL(t)NI2^Zkh z!=QSpy4x1j=8$Y5M_k0~C*}0xV2$F?yn4C1S%FS*qGIG06BZ7xbI$=@C20`4$Xe)YE4@&k`RO@gNRN29vEu*X9? z7$%l+a4ff*eASG3Uj6g4^5T~}{vDjY+1C5Td_N*mW>_X{wHW~;g}=>gfn%hijSgoE z7T;Z%qC|S_d#~)2cF#Tu1jtz4d72y(qtIR#&{sygubG#q*e$_x+B3U_+kEN;O(EH; zV?KDcmv`T*0Wgo_9P+MQW@dw8K17d?P_9=wA(}e&=}rH+I@4n#R_!NQ+}0{9-aj4H zv!MGZ7WUmmRA?NO_x36(7GZ$5Y{@x^CNf#VT!x}y%x3qH!*jr}wt?_T!`J2iWWGgpIV#SCk)3YSIC>FL;JApWA=E`0Lls53IL>gDD6}&H~j!^dU)&*cV z1r#)aOI3{oy6Xb4>;Oe$2OOf=M8!6?2=t$`19PGOz;ZA$0V>aXRL{;>3_AZ7_1m>G zbWjOte@{(TS@G66Rd4d+DK>^8A3_)oR)C`}X!en=vKq9q>tUy3xw$kx879rL@y9K8 zg_9#P?Y^CBvz-nTAE+<=o)~A7$ zb8?uzg}Z|Xmextmiz0)kNV;xhPECS_?h8Ni%vAj31PzK_GMHhwx#8k#SaB-I$X2KT zr10Lm%|!ShtxuZdfADjoMhy|SIL|Ir)A>P91Qn97*@OICrW9fWsq~;^v;AJE2EOQ$ zF5b+mCe@_Yr72|Bt2ZUs6&b@^BIV1~-QPB+rQ!m!XUmI`BQxUfqmX}oVyNM5wYy>? zs@xcY$y6cd!mv%LXf^0yspFmYx@nnAfyTd0*`(A^__r!QR(|Ae$ttC1Gc%D1fA+NH zjyTzsX*>pZ0P)(7mEv)8Rx@r)TNjEY*C zX4etYqFB!#l5YlDUv{PDO24DQ{O_}2tk7E8iI9ZYD zHm}(1Fp(c>11!RW{?)2Fx>_FYn;+D%c&kg5=!g#pANy3xqnxIA7EF2n+V7P4-|=`O za(e&PYlHiZV-;<=m>=snN$MbHVgXnS3X- zH8(1w*(NAuPNfHUCCrA!__JVQbzWDT7- zZY+xS|c_NFulw>N4mBIx&T@yCX)OG}FFG!6NA zadd6JUUV&&Vj?Zg8g-g zx(|}RRep!O&q*p>FW4XDsqe^g8RU{$6>`R+{9=>D%E;83{U;}^O#2Dh@WVMeth}tc z4rnK^KUqEKxmE^^%Zei&P_mj0`Ij8uC-#TuT=$FPtFa^gIKL7!yfY+i8d}R{>4y&O zg6pF6Ij91XqCBBI4ctqiX5en^oBDs%%mMW2EI>NcM=@lA<2#Vk3^&Wi!TX*jdp%ncB=b5ByQ z=>YMBm;Pf?WyQh6+TD<)#cKCHMtD~$ z{W^jB;sHNK+^e`LI68Z$I}|HQJM3Bo-L`pOzX|@w$$z@?jl0XWdSl48Q4Rd2?NTi1 z?e?_Tjz>4yrr=~uQ(QTOQhV$iHzi)CSfYPK465(H?FB#l}-4h6=^~5Z0nq_^g9=eB|fJ zq4&$WZ4=ywmRs7G=phI{n2PS_9<=a@|EF5vHW}_XD~5#;?zm+49!LVV%Rx;U$g_bh zXRq51W+U)vYv_qL`L+7%T6ija!6T|BENffhQmF_Frel(&unsg@g~kaw?WNtxAAq6 zfAgyJ#FBn+gQK`f@Eg`$z&0i3GJg zeHtD=Z6jkIkgZ_k1UqU59OC#$l7j5FN4y(k4@J?emVAG&2dq{J;#|#|6XVzIlmrcr zE3sp-Lve(UE<5JPk$13PIv_IBKAStPMnP=sL$hG(s`Lkc=IP9fjhD@863(sh!g>>& zko6gDcoHdgBSm7V(9Lri*aJR$m>lNtoL-W~HYhuw5aJ9(=3>5m(`;OSrcRE@xG;4< zGu`KI^Ro}$)#2iI4L|cGpS)zt&IUai(kB%A8AbZ2(1E21yfwa8pfS0d)<#94*BpYZ z*cOeXODW7gZ*>kS6drW}?h<=E-3rVk&@YW{=pN<6CwPR^-%nw5UN1J{|4@lj&04Yw}6p8#+_edGUesPK=EC;U{|3 zE7h3qyarlo50!VEUqVHYUrL>-7wX{mUS&uP$+zMhz$raLQ$exi6e&YPv+WgV+v1nd zC_+(bhgrcGajLl1ZpX~trEP9;kUK=pnR>tLe!y;kfx1=!-kzd5|W{D!~dL=(i=&g z?NwmZPPNOVcX;m&4u9n%w*lTh=;LT11H3x$n|k>9LDlq-O%rooJjibf!KN}i2F`7| zA$Z89#-%afhGZWZT% zJNX|!|Hb!QbRhZh`DqtPk`+6>pdB+Rtd~czITXo+!WaM2z&vrcEp~{d(oL*p+w`%0 zm4oI0t5Vd!Y+#F#nQ$K{c>$I;>7}>!Mb=Le)dUgfUEZ0GviWonf0sNbG{lZO+qc zz}l#xGOvJgXuLmFd{|Tn)K_MoHBM7wVAMx{H1{lVni}gh&Wrc`NF~+zD53|V-;%g5 z%#Nr~&!b7BSm=OGqC$7k7eu8qQwnr-{4h3c&#H0_b@U z*bm!apcYdBJ%Sh@-T=aEDA2}#V;>sY#ze&cuMe~xp4g9!gjdZ1nu&m7wAmZ$*c<@i zX8s&om+g7kcS6OU2YgxuF@AfPGP?_OKWPdX^xVupp!`(XVPDA0XQCW8+i!CzusaJh zZ`@m%8mJ?Vg*7fHH4Oum~ z?1Zeh%+wg0WV7@$d7ijca48@RwC=H~#DX9hXF^TD&p1z8?$TF(u`qk4ZUwsY%jzI;VOf`OpD~o-`dtdBS0vT&5p%E8FQzRSgpj zy@*qZ5)-iTY^|{z6rLNoGESQRI~_J6B4d`2E-Us7^K(%iae%5d*y3o)pV<1 zm9S*iBmYN&-XOFASlv`b?*g&i7XNzN{6MI+1(BahNudZrZgEgM+6ZT&sTNbjPp~wv?6#u)iSOf_UAkKx^{!W;y2QbUGK@Yc=cbI$vR9v9Q9f;m) zDy#&cxEZgbo88AczS}9?_s_B^e3OLohH!1?A6O#j*<=BmO$&xl7M8$ z=x<9B#crWU0u@@zi(wAT)BYRlWCrY!*+11g+5yj{D37~c5gnNK$dYQ~PeG(R;D&L- zr!DvSn?L9qu+f1d|Fd8HnS{*%4Y|>8HI8DVDYAhIMP-oEkZ!1cx$gHMDAg+yR*KwP>GF)Q zTM7RgLAfa=)lB}DX~XEicWH2QuwnJ4gz$g;&d)-B{nKBq_`QT?Z3tz(wv-qtHBE*6 z6gKKZC0>>=ikAvS4yYlc5#LIsFOeQ+EINhyQ`CZZI!mLzkrp_d00u1ZS3EzY@nuzh zcJQcSQ$ClK%8GFV#hRmnJDVwX6Gh^v&?c}v&C=txi8HY|sWWhgH`dFvDT;$nGN;)R z%WC(IF*Tr9Kz{Yx->Lub)8D-F`~Q}#qu4bR2^-@p{`|9+JhLkK?e7@U)ID!Dtaw#2 zLS5G^#Z9#u#rtqkfk2OJwO677vrtR1H593$ zLP0u6P!oJbwAU?;Tv6Nz9#ABR(|NJfy4UV}vut?`Q{%Rcl)9zT`O?OaJl<+5*Aogw znd4-|n%BF3ymII311nap9JxFHr|VzUoQ_-G^`ApOxbXJMm2t~gRLqW=S;Gu0eI#!T zIq{Z;{e2bnNPaOS%duLuiH}216jYOh%xL(GeLO)>8Hdam!8~f+@gJ=Da&;0QrE0S@ zSE8=8>$fjog59>Lxr&izDE3ToZ}ih9q(I0jC;`@}Q_=Rk_^$~$t!{y zo!v>|%ILV0y-@RB*$l&!l+{vB8A&oVOIPs^dZ`oSt%8b0Ss?y;)mNQp2gDpoAUr}2 za?CSzLc(C8MuGE7<~rM)f58MF;ABOn2b_8DO~X=NtHp+q;@lQ;&!JwhpLa`C$;R`a zNLC9E(uD-S`H(`wTA6E1`kY?gBfBJr)zCVEl``O*UkuVZnJc(rhy z!!G)WL=%|?f`?T06(*b0$KVi&)L*^!~df8>qutK5Jzd_dS@W3CMKd>o86Sa-!w=v;! zi{`20c{#IDd9#wQwsc>})0b@egg>(jPan460Vj(w{qe^5Ay?fRq>7F_Om=Wf3RrPo z`Hlx1A=taL^g4ZAX zb1?PxND?W0{tw7;EB0C1^bl}~Vn3otJr#OOvPE3P%aD}K!=C3Ane=LM*mcE_&5FLK zN86QfCe^EO@daL_<9U$G%%N|~FYpe#4%yT%?vV}IoSC)5wJBtPSL$5hd)cFj_W*jO zW0y8CyPe9HG=+c?#bO+yUN-!d&X8R+j`K*~t9T4)9#pA_bnH{C0a^L8%2TQf^aJ~* zkXw?6i!RX3iaPqCpvL3i;waNgKqn0qJ>0zbgQ>{jWQeGNt!qA&=zt~v`x}iUmYX56 z;y7@D9t=|{7E}keQK1-m(FXs4YZJ<~fYKS0g_s6JvNBB?sS^s#u^+No8b7NtaKqw> zIk%?HnD)L2;fdR?n9$t~|M^*BqO4(O)hI?9bjM*6{rw^ao*CP8m3i{ z!OIUqIZmLqg68XLx|f$eXDa4jGdWyAFrgWpz(dB3%8u1tpBQ$sS#8wCNN-vxG!Y^F zZt~J~_G)!$P@WJ94jfQXqR0NC>=ZP}7B0W7C}q^@)k5Ge3e1HX7p%>+I4rorVPqAV zzzlPG#5f$r2=7m;zx1ShmTrw{U*><4BwMlb0;FT3lE=F#7V0FksL<8IeX2VQ)asVl zl_0S~iVBm+SXPXCU^Uhq#IY#qTIqUS5(k{iZGmOJ>f6o_7h#(&kUG18o}}<5bQ9nJ zmI2O0(K3P>_{j+crryj6Ez^JP|1+MUM`E=G#fXoAHRf2SRQsBiNmU*0lSQ}EC!KQW zI6m^frzmS*Q-3DM()=76tI?;xn80*R2qC6U^Ca%$gpg@}o4wc5&?7Oj$fGuUw=JgD zAX*i+6f4dmY_SXt8X|bbd??42G{NSe2^f>}EcC5>{3rD*U3aQ(N?Pf9`xS!IuMKDlX@jELHfC)AcAPazo=CbycPz``JekQibs{Hn zA14b)o$4?C?bmgF?eyMXd_s;{v0r;z&#%2mu}}_mo(j$5tz}YpopM|mOT9Z5LvNhC zQ=TT?BJK>S7j!#ouW7HW3#nyV_-W#cY-312(SB+T+~$+#UIl5zWcNFgG|AnC*XEy~ z_b~a=(L~YDPS)OOd;s~EK#raw1ydm`%(iyVNr(S@+Q#)>#*JXYdtsS+6)caJ! zvGz``e)42+HHJ|!eu^WwxZ!GhV64>81ie%dmoh!9?V{K$3Ivft!-NM&iF0IdI`0zc z_is~VdUlE079A%kK;@7{cg$}QA@ya0ZQG(w!CL-hrU%F@AHc#^=(Nf<_4O*cZP8_h zDC(I(+)bF`wO%XZ<@Sze0{TP8bOse-GE&ja!p-OaAc(FFWN_ zvtqzL4fZuTbei~(cjdAgx*xpSn-ZvXV)_=Y<)45KuS7+)ppLGVwhP(?NBBM76^oAf z^vch>^z!OpxjE*O#^f`2AC}to$PRjC+qU|L&)67r%_oZeoLM&m?7bI|EMn#6VP;3H4;;GO* zOa^~o(U48Da>J4gXb`NJ-OF3U?4N&(`Ecn8`D!TCv)EcQ<~N@Mj>df6QYYX1anQRW z9Wo-5Z~YgE=7tO_&NT1VL&gq@O`^yaDm2aZ4#`v@?TCg+H{9o(szH{h`1?kPIAZGg zoaCGYPo5^5#VB?&H8|mbj}gmaDM;vzd) zH-zk&m!eE38~Hcg_p3mUBy4Fd7?p$c!ys)l5#lUGiRCUpmI--~JTNlYz-S)NcLHW) z#GDxasb>%-_Q5!&j$Us%LFw=_=rDMW4A~gJc?{piijf7CO`}|s7>fOnA`w(*9SvMQ zlD>uN?Ex`Np=h_o$|lF3Difgc|Ji#JxF+*Ef80lWLh{9si(rxg1rlL`=x~@+#EIJJ zw%c|)?eyI4PPf~Rbhq}}X`4(GK&mDNa?|iV*^2ylzw{y~ z#kLy#5ftS^F@S7gi71w}c&wyhpAM!CurOMRr+19ahV*FopDcMZoYoPs!jXZ57}%i! z92twKF9bS83_FkilqF*kZxOFZQo%0*!o#sCJiw8&dAtYbhp`#%af~K`@pa$$G}Mwe z5&TYfDKwv-PbLM?PxI+>2sLRzQ(FDl$E^P@(5m(WSGC{kvyI@%mj* z_N;qm6GTDc-rnSmXEBS@N*{7ZL9-LCMRX@*c=Fjd-YS+g9UpFYurUGDpaB~0r@g`EEfg$ZnPBsW&wCD&&xTvoI59ZV%)qgZlEqPE zHT3)N&hyu>F9{1IbwIwg-n|N#+v^0Yq^V;Z4#V-XaqR6qv@d<0bz_p8R`lykn{D}n2 zuw*vU$Hfu5n+vnCvk^aIVJ^&mugmklDYvv8OgOw=J*%>W8u;omDub8DzXqM}Rjyxx zW~Oe&7gLKSI~DX_LMCq{+d%yyimWFWuV zKt-0iuV?S{+~jvbSma$X#eA<3%8hn?2HlV;n?XQui)(ht%lWECD4E$QHu6E1($)B~tKwP%VOPkvuw6o)i!b zid}<8dq}?o8v=}H?m_M36YpBiHV?dq89Hr5)^PTQ=`lt~{FCbjsbzA{N>;2voUY1- zvO;L9!34cW90!AwNOhG96&TAxGs#DW|1o+#U}QspdCCz7!My^ zBG}@k?PKepA+cE$2Q{_YF3J6nM94N?2*yXCzk=2O75pA)1^+4y&96w2s~4rvnV}3Y z=rFKZL8qY$K)mNoXWaKs+_oZ|Du`STTjW3Q{)I0dDZLE^xF|vq(8+Wy%;LscJLlZdetWd{Vp{@)tnBuewRFDva;N*JZwF1W|)7f@&5L=o^V!7 zjE5(TXsR3w_}B`Wq1Rc5&oJH@_0#LOFTCol@VlVd{S)Ms>k!?V*}~r~+#%er%$F2F z$2+c)jJVJ#lAyig*34ZH=*|@kD1u zPYC(8G1J^E%HbU)JHwhqt2{EH%%xSDFuMYDxsY-rk-a=(EHGpmYi3XyEkJC)|Js&@ z@6R_)NsrlS?~@8Ao|4X+O-Uyx8B{kNrXrhVXE`P8IM&1It@P0CMCtOt^f?8BK7d0j zT?7=BXigGC>$#W{%2i|k*q1?-9?|kDf&?370gF^k8q6&1o_A5$Cdp)X!8L=qqU7(l zLsRRDd6$GZZGED>I=_BylLohY_>YB`y$0{tJU@119Yf3@V#P^$$zj`p@$5=w*5>Dz zuBU{W044dGb_>aWZ6bl-01e3^R8q1sij-23<>8-mjt2I-toMU>AGhCSr`tx+Md2}! zdeMLVk}w(=Q)7cKho|@>ziyKzg+2yLHzZ+LGNFe~1FvYCaKC?wvLblt1X!0r`v}JR zj8~+}p7?JChL6I~oBZoudha3t@|BQ9cIt?B0N$Q~W)xeOW$kGbObF8GKh%;%uUK|6 z#SAC$lq{AatEk9Ky63HEP8U1wtCyfxWT$_xdWrYN{|A{#8*`JR7{QU=YTeR2Ih@Y7 z0vQBuh8zCT!_Kcj2D8+UYonkks;u<=S zYNlS`Baw3!y~~%eI-?`eV2%@8abUcTkr5ilp|UqwCa9Q{N;vVt+e$iXh4_p?_BxB+ z!K(^f7777C#pZ|-Q50_sm)fQwwGl(i0APo!e|mF0+mdPHH9-ZKn;}!vQw|P2jnhXu zu+CM^TJ8p7=m--%tZ`85xrhK`A$F|+J+}ZSF624HEJL~~hnG4L_;RyZjSB3ecN~Ol zgo?T4c(!K;Pz?WweY1+Eu}zbYOMpu>S>nW#&o;BcPo!jPDYAx&?B=hAM35flk`C7# zK`g}l^;{6HpQ(d6rb#wNQkhR*uk9oK^p-K%U^cM?|MlJ*-!S3g&bae43Fs;Dx93B*K;-Q1A)KGMBF-eJ!MB~y& zr_ZTY9eukhEODatCzR?n=)h_@I3k`b)N z|JR@I{&qAolM@&5S(%w?-@PQ<=)G$4Y4Ke;#rG)bg@W<)Ip+hD<$K(=v)20+NE7^W z)los2inSzDRVZ0Cc_Rn}92RcXB>Apk<9;pManlkE^k{&^04IzuJ@?zqTuWkw=Sw7< zcol7hKA|GG9tyv=v7!1>q8;416$Crzvvjy7O4H~Qo*C>G(Cfx|&7p(;bW}!wf~WU7 za>KfL392057h;;B(&XvuN!}|K!v)E+A+Br%C4->eUMh0WJIH!^();1`O@2u%tlf>` zRq#)GUleV4v(>8rBpjMFMUpL|cDi0%N1yaQ3%X`2-x%PZg`$~0X*HyvD!6eRs1|nD zbCGxsuS%h>i!<4koQtBp!VbEIa~<@6chFx3;=Y~mH39&=WO}j>tgHj*ci+5NY1t?3 zv;ki$ozH20g)A)BPN&Pkq||eZC9B8@VGZXbFKf(rpP!DCtq^#6!x(K?hrn=g-6;13 zfBd=0uKdd-*DiAYwTbKYnOU6MlWwHc&oY67)bSlZ`-#mW#@~DU$HOlYl)92WGgV~@stptVt zr%^$cjdHI^UcpMg-*%7lEbL6chJsSzEhsH-=U-8{6ku`#|!d7XT z4>DoKP0Ixe!$m+@*vLwhW()7VxBZ*tZfVr5xtpb(Q+0|&>5aE`h84~_Bf_5R262vH zk9Vo6*}ok!lql6y?t?6bxV7A^LEI%>Eo&AXSERZc^9cKuI#q^hi&vH8CR97D3uq7N zl5Uk(NnthxIq&33%FPg-Psu=W zAe)N3LK5U9>|%Zt^-;onZT=(>IFc7m%PxDFX~BzX%85Y=Ro9N>ah@^Vsd##Wu0l;0hgC zL(p2baCdud0zQX+$zInOPsAMTiLOcpT)MgS?0pc!(Z)jpvO(N0i4SM$e!OT{?FAhZ ztX}j{jLs=E+&@c9&`Fu{+dq&*Cw5JW%)pRA$)GGEg^D~5><-3aDIf+%uN5mlK}5p+ zOE&tExo{dY6GiH6>|FI@Prbx=Zga#^OlP0}SG(H@vk@4@$eTUK-aCwD%W*Yx|4TpB zgp;24#Y@T7;mWO?cCBF`4kwtIxd934| z1*o8uw6V>tS+tTIfpyO^RnrXUK&qL#0`!MJ3pz1*z~y3itjBfnMV0nqc(3TXIL*tR zw!<{?qX3fGWHfd_^7!mdi)C?xm8t+-S>Q@>VWZ*{|4VkKdyMCTEEt&0qwkNK4K*)d5oXS zy64_Wmx$VcwPTIE9x8Vasyj)wZxtO)RnUDPfO2hSn+i(mruC?srA<)(tCM4XvRQhU zze-x=9?y>BV2w?s0+*$g8?x^xsO|W1lK*c^KFcR-md_y9Uz=so((yw>MZJ{l zK1DjI$nB7LUM))vg*YnG9TtJ$n)XQep4rz~Cnl${F@$tq(j^^m`S_#zlC7dT(&AC> zhAGKrQ6H3xw@HpGdSGz`zRguHJPh=T(rV60{;A-_ob{ZIvLx2QKx~&Pls%&RCF?o) zWVPpj%O@WlkyJo8LuM$>&!$CJYG^x+*{42t3<=}jBPhda2XP);n?qHns?txZ56c2U zxHwilh-Os!RmhRyBIAknZ4J1VO=bhTrv$I8j>QRWlW{EuVh0a$*Yo ztI?Xy&+AL_s-ofkhxC|}&kTk#=n$%Eq*V{s3Stz^&@{Oz=ysffsbz`uR^EETH zCJ+lDs)3JKr)u&{=ioKSxo)h!8q7ImO-p0zp#i2v0d}q;SSQlM;-SM8E7x~|JZY0` zgurBgun`)IVFC2ggVquD!=DM}SoTv|X+-P+rps%yDg&#aRsN7?66?#LPGC|k6|5!8 z!Rn%tcn{eoEaNZp?tu=YDtZY>>Q(ToBzuLa^oHp-Bq;$SB%OQdpJ%#YG;iP5<&d z*5Cec?ITWGyt!oiZIxv;jFmDHBaebMMULz{&7#VQIxi@SYn5Vy;t0Q5BYE+Ldu&F< zh(BRuH!?zn^TyY!fA24vXqt|$|M!KzlNGOxi=1UPQEj4R;D4>7B9WgeQJVAS5m8ix zp=bfC7z}ysChvP{Q3W)>TVVW@v+?>9c|^Oy~T z(|Q@x*H7GM*&<@4x8gEzGIdFd)F9R1`@}y%urAmj7oubBR)V*&TiFW&?_+}r?{BN;DOwAJHdV*Fdz&u47M_*a;qkh#4Ump1;ky{kg+= zk~rQ3oW;w64njXCs5cKyv8|zGF%(&W1w*9*y${r!Ozrbrh&+OaJjS*M*z|YV^pc-{ z2M#}C{Ur|*E_L3oqWs_wI{sin#Q!w>avHfkfvhx3x<0044=K_^MWVju31Nafj(kqE zwIH2aOS+*3xlI!1szrjy{RS3u^aPM18|fsE15=k4>8r8PxSj3_>2udsaiBUC(xn3~ z*cfwDwFlVI^CfW{!?jrWUpRTUv__BtjmI&v&5B)0?SAF?zzuI+;a>yN?S$F;Rj66j z=y?vfa8|ytcyhT9{>SC;v&u}@a-RfN3O(Szbf(iLUKzp9d0mA6`Q_qw1t$2V zao1lVu@j(Lai~GkQL@bx*+@m+@{DtB=0nOHNtgCWw#u}~mXqebZ)&nZ*LVB^{HmX)Xj8 zNj52pypSRFrs^ih8`b(;b8iSg3|n!$Zb7mrWhzJ?P41l$B^;xggOOl&44}&lyVR;h z{(raGsnG`*av<2$jQKuC@mIX$E!pzoU-AftoPJG=MtjEhT)j-QO5-o~EF_zlEvcQ? zT_eCPH>7Jdo035YZyS)!0hexha0_%Z7tuwu_NG^tWK-Y@v4LJNmIKvsv7A)b1^2}Z za7F!qw7XTXj+0A#Epl`~_p_lj`T;>2LVGTdhGNtf$PZ4BR(=kBoZYr}?bdA{XTpkV zTghL^6Bnn=IuP)LhK&AdO16?B%V6}u5!asa1*H%uW^5(@Pu6KR1!?UzYD^x)%&7I; zUi!##HhDcKg*>ggu$#o{XZ%Q>fKYa2Yr!f}3X2yL+t&)qZ@2j03j!LmVv z2_l?$HnGZqkSJ|ZT=c&MvAA`?C@Hgi>}i&9WStcfZxogVTe7EEp&dzb)zA2Fqq@NB zHoy1>e_S#F_m@sc1c{CI$^9-ZBx*{wSG(AF0Vs0^q|v#e;vax=DfhtqOTy$yUF;@J z1{8*#C)MGdG+wWhw~*dRMbNUhEHuWgVQP~mft4?*h9K$S(|6GQARu;_?i6QxWdpI_ zWA7H|JcnvWZlTvva^tNY2-}Qcb!^LMSed#R9cKESW@O-Rd(YJHg{Q9BbmS;oTLF^2;Vn=OzS?eLD|{z}E&S(bFhPFqM|g=ZP_1UdyKmY{2* zk(CLZQ`gu~stUZ=K39Tus(khZ$mt>}Z7OtD;L&w{r)g+}h~_iG*hb`m(;+f6tadz8bG~tb#_fOTkJxwdzA5?ESOV%-qhB&bSDKJ|PZ=+Y{-E<1HWvw~<({t<+ zNKc1@2{NAEGzJ?B{%u~eWr32@&Nfy`jJ7bdX^=k~crMI2Kk zON$KnqI$427+d_JQV=0N1#YT?DsE8BFQ_6jGx4K?*&239~W~d|LY)n94 zed@74Dws_BTHDVl2Z3F;U#H;`~k_L z4*uxE*BUk3>A~Yi$?4h06Faz*2`2F2PM zg;WIs3)iAaJy3m^%eoVSX+=l-4#R=SCaey3ql1qz!s_^oC*M4oy(+2TU8H*>)B@~ZJPr5IAtek;YLwcr5Jg`$lQ_%ODBvOQfz|9kcq^> zsCfE#M)071R7bA9>Xq|`x&s5dFs58lMq`ynxlgk+hi8ja!P6IKmkpayVV6&{{UMA{ zG4A5PHv=pewpLbzm%Tt&B5+w~k-8$d4Mh1i^Ezg3km*#p?0VNax@YD=P6Z!38MX;` zff{Xs91D-H(hoewVrXn0j3HlB*tGqvj_^uGgWmT3tHZ5S@KRNC+Ntytn#X}b` zIg|`K@6)Nszc1n~Cs<(IF4mU$cZib2w=+% zA1x(5a9!|X_Zz^>H{jBw=<~>;kEo0)VOXnF?uKl)P@6Kbo!?1XStB=`BZAEHlc@D0 zHUZ@Mi>zD5?k8QjuX^qeHFWx2(#Tp)5m3@(3$-YZjS7UwdkCCrD2k10*~og9#4?U` z7rWTEduk>!*SAXhJn^(nann6M&RS&_|b9>!^_Fh`% z28*l=RXt!KB}k`80bJifJkaFC z7Pc2elPo6|P&sJku$EA=Vu}<{k*Qwwl2U3tJ8pKXB#zfCv)pggv}sOjP|y7V=@xAW zuI4m&M)5kK=lQS%IVSpDj7P3dj^?c-n@E~qMFg^kLai;Fuad+}vM;6>Bj=fwvD~r- zf)O2OgdD2;;_Vrhu@oWU{Pq{WjQs7-f4$&OgYJvd=6kFJe)8EB%B?c=m!g6;%gdnR z`XIZ*E811>gYD;aw6UDzFdfUfAbjZEq}lIU=(SpQ&NW?iEUbciiK~azO?g;?+`gDG z8_>}K$m~&e-LVd!$mFVnKQx({4+?(OM;cz6Vzs+wM(GMAyGW5{DiWny7Nm0W=nX&< zxIm})YtX$pi~g156u(}4l6PNR$g1IJx3TLbJu?!ecpBxk zkie+gtc!%j0ZU?CVLi>yUUFveTc=aQ?i(QD* zqdb;%&^HQ{#B{jS_N^sI0ERK`>p|ypm*5#?^(v@tc1K;xY6VueMy+gY0|WkD$xq4B&+AP^LK>r2){O~Jzzb%lzQ~` zCQ>u2QB(5$d-L_*Yx|dr^SkDrhJKzZQNN@BRED#B(*)V@RXa>qNeY!sn$7Y~x|>@7 zgzy#YELP83#^+_wMd3{vSHDXXZ_Si4ZUwF9KH-c4r7y$L=urH!r;q$RGZb^beabz^!RND&NaRWGGvdni%_6mO6y%}{kop>{6IH;cZfZWk|%1s)%y0nBFI7G@i` z0qf|~t{tGC-XJ!p{2=Khc8aVJ*M?qXMe#lhs0`dIY?eND$Jp>1c`c_28bLq~kDca+ ze1HoV2h~aJRh>IJq?ubs8HT2HAJEum!>VdqrFieT-=`MAG{D^4dtgUjRzA0vq zkgd;NV>tA^ym;=?oH6ZrED1aaK6fJv@4qIk;a~&dW_h~*`RT|jm>HHG)}%S2(y6x7gHJ06XbAxh!s(r) zhm=g$i_TB40H+kN4OLi;+Hmr1yO^kNot&L+ZK7K5tv>m56=~FL zVATmu3d(~U6l2+(W*Px!9cX6L$Z!5r_`YQ;n3ayJbk!d3R-meb2=^_{x?rPt+nAbM zCU1VE-;+_?$LP&FZ#1@eTT+$8X~nvaN;6Rm5>Z(u`}oW29N`0sM<}*3MrCLMPk(KR#3g9*!Q8N zg}}K8KaR8{hjm&{#0uRjR$doD#dHgQ zxBoTGvI*x&t3h214M~E5@jlsAACB8tX3*~v%j$s8e7)ozr`+uf6sqVHYvjdzyAHR- zyB^i(TRh92yNqlM^+#z=>ANNyv-hX|?~|je}!smKj88`Dh5Ame!&^sQO{dsX`J zeaSJP1a9G<7Gt9G%;XIsR881UM{`On{Q)=in&4R6})v{7~&&(9MR(RC2myM5! zq&EfDbMutR%3??*CQ5N{u{eqs!`Tp=6p&75v6?jX;%l>>U3YDk!$|>04wRRS^e9jN zy(W5zWrLZO&a4cT9y(5!ytygRAo|+^GFng}U8d?J4Pq@;M`N0$Q@k%AhnEADixm!? z1fKrcBYCk;Z+74YquIcDBU5kYe)#dLUb@}kRnN_WYJ*12+3#Kwp4K#JqGj3awwVvY z4~1pA9)@Y;0a@?9mSnOIb6;lg?S>epxV75{zvMxTa7+DC_21dQHO(_OX0LgZ++a4( zIPpqpnVD48LrT^|kuEB70n$T81)U{Wor}4l)8agN6tA9(3nlJZk_Da#s!37-9&x0!OUHlc&JoZU%tZb?0mQ87(`(0K^_bXEZpgpR9zhqjgp&S7z z26YO2bcr+>WULmTSmAC`FWJE>^wLK0I^3WK;Tjv4ZLL!5Mu?wu7dj!M++ban3VfH^ z!Kb&-7YP0;%c%->P2khRsI=MVj>!g85574g307`e{5VW(ws134*w=spNEx)#>QsGH z-0>?*Io#JIPh>?rcg6cVzdqp7Adco4+Wif+E-=<5qFNv1JS_@qds*G=OTPR}Pct%N zOdmdq>%HcmejI8tOOn57w~+kT#w;B-bAKx-Ss6u2smKNx5uGcn(=N5!OnR1X^96WqAWFRCeC(D}GXE>W z_5Lrzv0V^ptF^usr1SmrdHHo$Bp*?Q`$7x*pQM4-3YL;yE^?&z=DV=8&! zn;Yh&O@lnleU3iF%~;~vP&z@ljt9BNSXs#lU?q49Jky+{SUW-yPnjscQv!j1Wt8uuIxn-@r1 z?s)&fPb`UDn5>kXc$IEN0wT@5OPWC=+0T8q3cAv7dsr2X(?$#bAgf(m&emcYJsoO$ zv*>D8oado1?RioWyv!3>_;Lg{-U9EZL7c8ikXP`d97QlSLL3c2mSZ5bY~EK3EJJ>; zR8zGT65UkRQqB{9AYk-N6YTWNo7AYu|8d=q*Zk{g4VLFNX$}J=&>Hz-*5yf?$bzw& zkQjr&F=PJRQ>{-zrISy5?^Tyg#&VfxUSU9<9KPt-f7>AL^QZuhup&|ftUJex~bUs^D>w5xi>hX1B~G?CNm4velnUal6sGOS=L!Gd4LmpD^{wq3cZfVTRiIM z`;snc3-`0HT@q%odP!fz77)c-17Vad=_M{+fr5Kqc(qI8!*kfJ(j#AeK+eh^u^eq| zMue&{K-S;lH+_>f`K~?Rmmek3uT7W^)Pja&sn=4nH57@VBI{kZ37-;mV`)RXc!d~w z>KqM$=dF>Apm_d92OmjqO02Z(F>~5Au9dDa!w_fiV%SE8W~kJJDF)&c^{!(xjSL+k z8^?JJ7AQl1D~40b_~v()=>trrqHg}$Qj+e(@z@%(FMly51M2uZD)I{FCl(cM~ z6wi6WsiNbR+v#EmpkcIQAKeUj+x`AqMaAJIf<{f6Uki7)f8Eoo@hHOU?pI<3 zg{&0%h~%*3H_6m*nlyt?zSv{EnHyj z5v8$HB6{6Vk{*c3#>jTDGND$kOLEdXL5}yOK&L(4i+8P-UErtry&R7A#F#bg4RidZ z0Az--n~$PbPBNLM<{4){BneJznxNc!h+mRU$zTT4L019GC<@)Ai49)s*7o)j|8loh zX)ZgKbsgT@@BwummrwyJ3fazl5@bqKRz=IF>II0qM~N`a0H zmg3k-fXrky%A;h3Kk&a(ga2M`Ohz=AqqjgK`q-D=r2pdw<1AxvOu*vAD`+dxIla%p zz$VT4Z(V@ilTWWoKfN!(wQ?y?9AcQZg?rMwo}26614MOwW2b7w93C$nS-V^iAAEL- zWsAJix;0js-cthF=^lC!uSD5N=Sxr^6RboL1m&?C9!pQgmZ62oP}^k;ZF~r$NoM>H z&RmzOO}@wvN_M1@A}97m8q9o=1C$K3lgg>cW)XOsQx-$jBmV0pPkwY`F0w|R;%@_W z?#n`LYUn-{5=q{83olqPv6rAGQjPMl=YUH^@TRG4ss`~L@e&T|=48=voD@J=ks7bq zr@}k3AvA|~RI@6;RW)iqVEwus*5*u5Wa+NFXx~P#SiOu9ipEXQyH3-XP^1r?l0^1T zAm_{m>nJ5VOp$|-ru9q@YvH%4swCB{q1%nH?7*c*E!pjl$Ezg$F5T|&oE-KtR+Xet z6ZKxzw+r4`{5`y8v;4AO0l1pktUHsCZ=6dgg8aPTzN{-btNzT7!w3Cg3?d z3zI`^FkW3O=@lkTIzj5_PIA&4KYipLf};k_G@w=o) zXd&fJY?GSJV1AsE9id1aGU-0yX2R+=f6^17wm`7SubW#vHI8%Stz!BCyGY$SwTs=W zt`pRP4m2K%c6}gOp!dn5YdH%uR1at7@r~97t0yby-M)GJc6uRdylIWrr-jD4yycuD z;ZH*DhqTknsd#maXM^|@oys{%dZ79e31aYE3yoL76(h+?I~eFA#ycaRkK44!LvCrE zoYwiVVm4!0@!_d&T|>*g0BUd#-H=tc-k^o>9lCFl6}fiq8ov)f^u_f!i_Y+e2+TbNF9QLI2+zB3Ib5| zdBkzFJHz5Q7z->F=zWfoG(X^p=OWX)wuhA<#}nwLBLMNng~tJ7*q*d#O;(-}COkVul(VqPW=LR*=WLB>K_l3Wc6;8a|0-fU_ z>8MSV3|y~uRAi?(Nwkfv2fmJvkjzR))<~-Yj1}GkkY7AQI=JohM{Ofn1C8+YG92v@ zehTy37&$TH20p&i_!|>8K3TJT2Dv_5vw#yDilt^v*}atPK1DjINbGFG7_t_Z(d9mu zg0D{6DoPUNva#?c3y3OuSla!d@0sO>y=;qkU$A$RD(-SOSbp75-cWW56GL+Zd%RbN zZ;k*D6ZGcqJUE%$d~Lj) z#(oIDcc~?xSlH~YyNgqN_0RknC6|Gs2>t7{Vpv`CN!?%o% z6kO@WLAyzpblaOyTGQuoA@Jy% z+Wk-?e?wG7Vn@$VM+JFC#8K@`O z=_nZ}i*2MLbN=}f@LbihPx6kaQeAV^1(MzV*e8djh8Kd1B(*RL){w%`T($B17!AGc zLp~Vzal`;W|4i#s)G*3ol;6W|M*h+SkR@B9{m7+P%wu|D2AaE+>^21ieUZKH4Z-MC zT;jHb_No&nYPZu35EEQG@50;yuN>ZqNe9&nuZT~=BHgIjoli~%9-RZO&vIU~D2fL> z5lymYQGwS9;egBO*?D0ZG?3lkdj3hsVi5C-S2xMDO{~+id-<)>J)&(Mz>ZVy_T`)+ zwQ13Yv|&%YZ@hJuhSu&ytZQy(X>93Ef@ik_y?Kc{_+jnRbE00KK4~vdP&P}!iF+%} zZ@>jJCqcmomtcEm(9SUofjhWeDigMd9I{K@8qC0meO!h`|^yg(HZ33w@vo{5l zERP~PAXvw_uU^Ely1H19KBqh{NL|L|u#N%O2LUjQ-s|_k`EPq{uy| z{MpxpnJ2QND~QgC7gIo+He@lCN6CQob~_arjT&GOve7A`$yNY!9gT@}Ef(Tn@dakB zQs~XVrk0`F3~5QskYZNSz>14`s}}mISNBZ4L|X}TkC@bQOdjI3K=OEXd&Wsi@?mG4 zFTv6c14NsFmJstY~e=Htq@qz@H6NofthsY z%v@@0W|k4)<2kG3hy#o)-MBa3|4rV6!AVH}p_VLiVuO-mhL3nk7E6&;hH7T0WyU2k z2H-NGxdn@k#|8!FBWM#Q%x@W!Jw})q&;I^pp4^0q+<~9$A*oJ`i3&4J?4o2lDUyR^ zoonQ4JvO{~SXl)AQ3A&Y#AQ?3fmpbY-K|E}lIGyTAcr8zCddryB|SUn2n;cT40R%8QY|q-rqCtz zI9cYzkjXHE%mzx9K#_PVGBy|#s-bYkz*e&^;8HMX>?BGLLwi&vEYV{j9_AQF!`kRE zLWZ@6IN&%VWQlCS>xm2?#G3c7qn4LYS-mfwy zgRx9Sn*_?SXNz*Ne};k8KAm3Vc8czcK=z8w z@_O!ize3g?H=W{&bnq%97C=cT%sHQBl?w2hQ&2VWe`>1g+qskdU#7|D(0A_0Nh`Cu z2Pba(jy9`%=%!@%C~}92T4A#=IoHp7(yO=<(@Hetbm4^I| z7Gp+BJok9F(e>ggu$uw*V7Cpr_SVfabdKYSyO$jI(CX=ZGqgqg^DM|d-3c*PiK0^H z4qk!ui0W}*vC#OT%Xy{r-XQz{T)8Jod(>CGw8qUgx?B-Y$`z*F+uU7BZ4>lRAdT=R zn^97sg1#(D5v-8ns;-E%skGH|^7t6kf_{rGX$KwS8Beg7e&wtp^#jiurPEeCuzx&0 z)siXrc@jq_uJp3P9vlM|Tt;^-3>5%Z;K1v31g~hm<#uP8KKtTn$7CXm}Ogb~Uoj$<6z~4@1cwP?G zDRS8x-qhwzTI34XK!&O@xH-7oP3Lh|*(TX4GKf}T)o_A5j)SF5RWvTR+9ajGCjL-T z?$fBq5#&G$+J4@&O+XpyfW2(nPN=aClwtB%*tU8_ z2NbDM=Jp!sgTVpx1Tvt&)>zu2#pwlqIcU^1KE0Q;g(Zo4MY+^fub7B-=_-$Q`jbUG zZPpb0wG7MihoY2#cy&FEwKGoZ9xxkXr(KV|U-m&4+cc%Q1h_<#CBs!`Ik5{3+Lc45 zxkO5~mLhAY$YM#Y_kQJ)$rb#aVMW4Rbsxld8iLDxqIrla?HKzzw%aKO7^4?F(mIIy z+5`Uls?SzClzBi}54v!eR>b;o)X5ld(L;+|+016qz8Us&QC{$lpPjND1!GGkY8?f` z`7v|j6(59|EC%0o?l!W^iJjmRX4avKl7Z0EJ|Gd6l+CW957TQrGwAh`>m_@FP}TrE zUs1gDp6ndC@m7&?qiD&5BF#qcRS`=kbC8IKpf+oYptddY?OKX=#|mik;YN zv0|6*h!f<%65xIkO1H>Ju1hXPCKAPvrqnPkiog=wL%GE@bJ7S9hs_CqR+uS|HJ4 z5KpJT`|b-&+x1TJEzhe8Az*F$IJ41H4q0-%Qw*i;!9_LvZz?fVz@BCbB zML6UN4ttg%gl+DyISh0b7_@132vBbZ|N1YMMK`aB-Lg`mgATR#yQGQIGDWekPJzry_h31OdWO-w2NK|PL$>q`gD!KrDmFNu zKrfTq!e?y9dJ*y&Cr(_br?VBt^U&G^T~n9jB@k-ra5c*KHj1%Dvn!%YT?7x!fki*Y zyBAgjXVH3}&!(N@Cra-~g-_Y@1a`lJY} zNCvx0-ToFVkoc!%rQususN#dqYtl4}cX`Ee2A^0lw;y#v*jpCZXqhSfeC>)Lw6L?jl=I&Q_ zLrcdCG5aFi*l=!CkhkIaBfaGOw+hlNWmogfUh-2VQ*b%qwF0Y1v*>~lT)|}5MN{BM zC23JC1^PXlG7{v#NDO;dy|P)UkfT~5-3)6UN4{6X-#yZoJJOeC>wS#uiSzmvZAX6a zDc59A=AGSefUJDQvJ>0Q0G&k1Hc%viip0L{cy%tweV6;_C5y!d2CjUMBoFK$zu>mv zD_PUZIjz#_aEAlvy>OpE{jfxu@5;h2=`J06N4N(ZJ3YitA~!PpB1Sr z6IidwBHT)WyK&W|Q|wUYim;>3!25?Q^F~f1r$kgD8WH2QIS!cYF~JOz=Z^N;d%vq} z9*vBU6I&E3GD0z&7B^tVp7n(DVV?V?>8+AfVfvgd^=DGhg@1Mqh4kv^O@Z}X%ZnDN zGGSee_d9GpVgi=oR$ik8g|+(|zH)f}tFG0Yq56zP zC*6{+kzWLl`4H)1*9r64_ojDE)hUqK&tU^S^hJ+k(1+gdsBP;$$1n+Ts`%p*1A!(# zq_Djpnq)Cs^g6Hm4e}&I7riBvte7GNROFYkE{K42xFSDRy`+NQ#V+>k2H-%1?c;tI zoPRKe7R}q@rL84PLBRy-&I`OsXG7X*+5uG;J8l~4f*lsgjw?RaobFNxpDlCwHeWgmoA8d9=|;T&o8Ok8Pm|;%=bU%BAq1T}+bwpWsW%k6sALZ*+vxCXn;(}R!i`O0T zRmoM!Ddj08D5ZL)vKkeq_yaCDiB)myfSIqIkF2ltl6xEs48yIl0^{>art3vmJBp0= zNxT)1X)K!lfaJ4FfN-J_3RBD721ERq$3(hXXp3nU0o?@hJFfB=LCnYzka}imo^3I; zIm{p>e{ai2ViTl}eEVN6lC{icF(=NH?=l16R!Rn0%w#IEG7wc>QfL$x>gKmVw!F|w z3)Ek9EC}&l4H*b2cPo%&sy2t#OR8X!YldjeVc|jGPq&L9df{k}5K=FEraeyPO#PFO z<K6k*(ZtSLH zP@|YnMIQRvC1D{*@O9FwCU5lK6xbuJ;2#c0zSynu4R1DSE=fVzIlO{@mUWVs<<{>~ zC8^8lQp2QBEcq<~%NcOUu@aHX@%?P>@v zpIW8Zt=pr{coThwouVC4V1@&#n$aGs76$NBb{L2ElD zgA~(NDl%J$|Fl>?en7P-LVIKqc8=o$x>wZVfxV#ZG}gbxak_+w&=XqVg++a6sf^`) zKnQ^h6r)Co)zKU!e5^iqv?r{?=erS~c#ktdMYXNuujGk~^^1p@<(wEqprtURLTWW7 zTS<{+R3z4amO`IY?weV@Y4!=?jbM0Qij)yN(;l~czq{ebQ%!j2d0)JgY-JX9abi4x z7X1)B6jCy%?#iVit3=Cy$Ho|~(dJIY9yOM;)zarzyDp~ZRG*>so;B{Sy%); z?a(5CGt}TcIRe~Dp|PhI8Uym#$N1O|v))6`{m9Oi=B1z+GUL2>(;iTH_Ke`w_y6#J ztF{4&iJ}AaKCRL^ui~ju5n4<;Y=&l$XaDRx_Rrq-a9d1}G2H!*IyU7y5vS&u%t_-< z7H5!BX6D3+Ya7m*nUl{c*&&J?pdt?j_L9XBi+Gp07rpeNHdT`bmAwX!=oBl$Ghqdu z?Apvq5yY@Jf`wVcJ0A2z+9Yq%EcV|lIH~Gzt)uU8TevCodD5$Hol!P31MWcg_ik9+ zOgw^nHG#O*!Y!M*+!YzuYsqf^WfNLw9OL!ymQNTB_?Ten#a3pRv!ejo)>RAt+p*jaMvRq~;yCJYApXBm6GuIK=d`AU+ZEF1jyWZr0!h$7yv?vI!sQUieHp0H8pV~Y z1oq>=(TuZALore?*!1vGTx8uz^`e)Zuq=OY+OUh2f(ES6#~S`T$tll^;r)_gez~Ga z^H8)TqJ?|N_Z+{CTPUrf;{?k*69j3HvThHl7M1(l5hFyBU5oimnhX`Ty#b%BH#P%R zv5$r0Gaq;)$aM;&DYh?gMv4;q8i?oNWE~~LEM=%ImkxA>m?o<;?NCqn@PNCaIGw@SK2 z7sI!ai{VDX##=$PFd6j_V{_Fu)mBfu8hc4{=}!6p-7l#ZKk?To@VZ)dv7}eHP1yS4 z?xyDl#M%)%0?33ZYoEmII5Pk#8n6F!G@_JFym+!AN@;Ag#mZi6dcn$tK6k9QN+WAI z-Rduhea*~E$If;bz4TFbz2Hb@de)y!mgR%GAI>HBUK=-fjaj$tqVckRiaeqsPkh5@ zhsx&Ohjz+~{;8afDYZU2MeXD=?`qa*afaNb9V50 z$bxeIMZp4{0zb7$Q#iYxzecchCiaOU7eKSLj;{6ZfoT;un*ki)Iv^}@`%a@$>GhuM?hm&JS%4>r`AW<4J7v@tk&{fT*A|LP{LT$NDk>k4x z9Ov_aog{{{61=6dfJ(nCwJu_80dEqpj0u_`m3U70x=G~p^n9&@3d&B&+b)bj2)uD$Nc{$@l z-l=|jG$CRquFSI%A~ptxQHmWoK|vJ?6?7i~O-!S>E94Atm^Lbmm*fa~-dZ8O&Z=hh zc~k^flAHWkPAtpNO?jSF1YhN9>xa49ZDY4xlXe-35S`xdu&(BbME|e?yzomhT z_rc|Wiu*0GQx*Z~dQD)Ce6iU02*=>dHV9eR z6!YEZ$KMV|tV8Hsj^Hy(?Gz@>_fDKtvZAVT6ym^LlL}{H{5Fjp#cSYIjdd{2IHIim}4SfX!_+N?ec_WSC;y^wK(#8fuhoXmQiVa%0D)&Du0p zwjkrF)nGinBfsI?#%230k>RxdhLy1{QYT7}5@75F{>ysFX93zGWj*K@B0qS&Y8K^f(M3zj`4P6Y0=wLyB3_UPN4 zfeRgiD<=4PYBMZ0904Hf_=%kMAOH3PlZPts{@plI@5I5=J7(VKWlDB|B3}aY)4#NE z`{?vJ)hc8o?U$f?2}@kW2Pov=!nave<1^sW7m>h*xkYeRs4esF5S?&EJJTk4L~;Zt zTnnLW62vWl9}~J2p_Lj9({a&1A$%!T-_l`bzRHBp9!Sfr-DANKyrptkTJ@G zD4Moe8q3L>w8amL7yygM6iJhg`m~Fo8&ZB%e4bT42Y(UmR@qhH7*C$GdiqP6V(kGZ z3v}7M`yv=xhtn|TXxygr50yL<#6B#V_*W9+#1P9dgIF>p+k{SaWVgCUePG^&sg0W2 zxgY&KlYK#g{Hb+t6#1K*G^^#G1%W!?w5o|2-UqxLVq2ac5a#fB{uz!q!U!JJk#C3n z$db_V`Ps^e{a7pPo5oN(c9wyhSc4dMwdHQuMx;}e0(o;1xx%Ro+$%)-5AA+cjj};} zly?;P795Tq=D0Ch4M$*Q=rgV3=J1($UY5QtlZCMp`#M%UV-Vb8*Frm*wo6?)JBQcj zZeXNS=)+da&}qI6W&aMxiP_jBQiS~a_y42)^Ur_xi$DEFxRjE8MggbwaQ_>6pQE>n z9yD&YBpQ27kdbFL6{S+LEfh(jBKLSBXFBluHjDZLqbBPVDfA3x_ZPQYp_YyDuL(q~w9#j?>ga0dNsXJWmtOVi zo_So+z`|ux9H(8JLN_W31CA*YrCpMf-Umf@p?teDyirl^c8PmX)WAY#w^h1H6~|fZ z-@t-(7EqQY$Wa0g-PV1ooniJ~I*klBjyF6v|5Ja>*W|Z8ksVz@bgxY%Qkj`)$)jXD zD6$>7rH-gh2sg_+L7R9tSsdC34IF8HEl`|xl$@S@8tRNzxt*Du9I-=_AUQqzuM7U# zC_YDU5tZhb#VQZp>$lf;b7&R4L%3gwAkrxooSy9f!($)7*x&F3A7ApG$^@7rk^fjn z3Y~bNQ*Q==8cJ44kuqpA{Fh6@&w_fsc}a*Bj_GsOvs)$D8B_oo71-;M?Ulk zsiplrU}<#k~QPhWI}_3^J49upJOgo;vz#83AIv_`r?6|1kk#`HI-tq@CF` zR&fcaSLE5sJ~?E;bfFS8FVO;@2` zdx5`vrk!|onGK3WwBBewFdex{&7jzs9FB)%b@*s zT2stErK}`*&=tQYJToYXa}?;}+o3xrD(E`N2dS?f+x%P2ieL&$V%1@eRp?AtYf*A}C0N32rQgir6?# zm#@prX;1t0YtQMlb2=URKbFS-oH_E&B%xw-^u{ePre^3iL!{jHl-EpO*v5)VpoU(*hG1SCa8*CkPq z>zN_X3$K==T=|g8#i{rYpD7gVl-+_Z)i!anM>$B8WX)`F?}Iv;kNmAg#F*1D`lS7U zGkR=W&aeW`q*ed-OuZ!?{Ffvl#Xug96}`Cv6txCcx>X^u(YbWGv=`jn6zNHE3x98b zy2TrlQ^0}=`O9^xbm(D(dM#s3mnbE?UZI76N|T=h7eQO0Wea?s_3YFQ5;EPiis=`J zT_@>Iyoyzttzw0g41_#$shE$JUKhdizIq+Q8BA{t@1>kvDnHOi)-=1{>5tlVs1}_Ix{#CZs?&k z+!euR=>gQd`NV76tZpUxPwn!LYv@N{iR%5KE(^D8#%y_Z zbPm1OwTiDE7RG@`u&@G?o$;Y9URdkd=ar}Ip;x-!=GO+l-wMkpR%E0|wREHJ`9Lg5 z24+r9E{(xt4Xx!4y07wn5{A_rQ0KG9=SCz(6hT9Sj9N8q1*es6(zv%-es#jP{O7qp znBXSOl?uqGPHdbunq?b?DB0%}`HYG|*&5?iqIg<6AJOt8>|q$9r5ob|;QcJ&*3t+{s2+0`Oj&}CmBE|z_rLCw zS(2fRt{!#bU?KxS8q)n{$m65a+>zX+2P!PE3_~xuFr$|ANwY*q;|qm5yrRff6%y1p z()**d5f39XfnBJVhlv#e&Bzi z$J>OBAJ*--NH(+crkwXL13A)>-qbEimPL^rka`CNzzw_=|rsELMmb zUR7*`ixtm*cU{cOubO<1+Y9PGB`4VV9!^~L*KOu|T%u$bC~}U9X%p8>&G6m9StYnX zHN!`p>6_`R_Xp7uT7A<$8!``>!I#A9TCp)w43V6wkXAWRuM+hsl06gRI^c@W7!YUp z)=WkERV>PD48F=;3Ej7-PJ37J6eQt{d`7U&#YXFK7?t1+&0&vih&)Bn_0)5q=eoDjB#F=orpaq#uszZ73x zXu?*e^ydX+-xQdDkq+T8N>)#i8Y-q8lsET^o&@P74`y9nyhBk-Ydyv~*QzQH-e!``N3CUUGSHEe&Kr9=jkEp@nLgi=1BI(cS00;^jPU3GAov z-a5a%QCT6mHX8tkO~5!G@OYP5Ef>zai}-Cwl)vTP`GpF;Fl_YGf^+C%H!ZhBQu4K( z^l}~^Uj&|AR2OabYEm@PAH2W$m4xT8=NSePOCS$F=J21)K%Nmg_wSNn8#Cj#?~s%) z&H7bgW-4cH%#wS;<1i;m-f^kGhCs-^tw3uQwXBxyhBkZKL2@|WjqcU>ED zhr4D8yB9E zHGqVLPW8m=nm`X-ScCS8(wfaCOVctoLd-_A&jGV(`QFA)X8ddESD?3qerTK*CN(qU zD}oQsQ6B*hb-AG49|?}0$4=>sX=MSKFCKahuTKBv&yALmI0nM?Qf{ZTGawOs(&o^s zq17PsI~1nYMLmebgx>#aR5MN%^H*?;FuwCLv>ck#@EyL%6?ysMmLp^xJ6FVc|1xOW z9=U5vqhwnsl1#%S~Zj%p!v%8^{bBINptQ=#S~x>FLtvW;j^b48s!h9L$B` zEz>s6~!)TX=3a_OBKkPx0~J*QSE=0My-;^q#r1r`$tzUS^3FX zd)tI(-r6^q;C-xCN{YXyeU)YWZS>Ns-O;dIQ?XaaT5CG{>ix(0*wuzf)HJ0q*pjuj8uP6#A3c%`O8(iTDV~fdrkrKk1VBT+2oO&7-0+|wCKY?S>{ctj#I`_W4>U>rHEb=zD$ru-|)?(_Xbo%mPpRb zQ}2&Ltu{R*0g&qGrc@K~`BF#s?AidJiDJX5=l*MHM35yPu+uI}4BWqe?~s-l!YSKk zo}`}wtxO(wp9s5>Ak!oL$8}&){A9)hzo)#qz%FS$@09E^w@ILJtK+n))_tq#JLgqT z!Y;`=!XAe?&fm3R_R&ySHIoy$iP?ONv5BMpIyE-T7L(${X-5W=QYc7|RvQa&jnsN8 zy)rxvS-+&n{Gdl(SWvr1tj_jbAA(i3AdLmL?LZuCkRw3-$@~BJv|H zDzQX(GteA5Ilt*b_TWiX0QJ;2ekgh(}Sipmn@e zJP>wy@fl?vjJ_)57)z~eTaabp*o=3`3S_%p`_nHhwYEn4#!ieM2K}wwo|TLC27UPc z$4jq?x+QtszDS)4MOz#HJs!5s_^RLIbCG80`ya$}{ycn3tLkF|lZ4||Eu0%0pulQs z+1|{&a_uO2=bq(8n~me^#PDI*mX~nzImYV!Enn41YG&#qkV)iIXsk=)w3A-m8o%Y@ zp7|GiF8N?YN2f|6*FQLAv4sT7`^A6QkZqYMa$3)qLA!h#bi@|DR0fqX zpwov*AVUn-sA{l`cPQAw9*E^9&c;|+KI_P1W@9min>%@3lixpVNeMWy0bxi9^e%3e z)QeWD*6?!Z>!SCxul+%%!ZTN;#dNZ;gj*$X-D;qrNCXOx4EaFV%`mLzzZ%-iHMp7bjZx{*j$RQgrpWrKws_FNhpemy)qUc> z4q39OGc4 zn|8weKL|WUmb2RlJ8=vaTEa$j+;5^}2^85t#Tb>9u8ThL+9<k)Jq>&2MYNejjkiHAIe!nKaqVPoCO)0WnuHl^!qyf_$UO|fPSh#og6WFz^SzyTSYw4Aot8PPz zb5ZIH80IlAWXM#B1IyvoFDAN_LSVtyD}FuZz1(&<(AKxyp3RGWJO~%IcR?_LfJn|s&0k)GI3n_wf znNBGt2%wg)0t}i?4mV=E_Ao0wYB}M$i6ffcF;}_67cawRu2Z^PfZdOV_v`ho@r~z< z+j8re?}#;Dq-PtqbyM?BEHHuV&2tNjN!eI#lM_4KO=b|Pqh!!la+r!)_wt|5E@_RZ zf>{5b@~QZL`_r|b?_aWl+NN9~sHD!%?3Ap;leMp`pl*8Msn5THrwXY*0gn|tRST>% zjr5&>!l>hPt)v`Mk?nlcpIoWR10I`Zj|bvSU+t9OJ=HS{qoCqSbT+h)+s&_@sd2M$ zH|IE{JR1q)U1heWO}|;x$2DQdCD>&-Sv7_1GMknZO16n22~P=P7cQiYbLPWzUih>2m6!M5lWE({_Hmup+7o2n^$*clg%? zXULH^x)w}GxkN8MzhsbGwB&5?R$0-KW4v;B;;curmlk;G`lMAJX~E0MeFzA|3UcTK zXl64Chowld^(`$p8?FGyr}CPphfex>_p3L*eesnBa^V{rzje!NyX#r+V$w;l#uFK) z@d@K7nTk3wt1p?=-wJNk z;V2k@fs_;}8k)R7Xs7qPC^51x4|>K6xB6WJ4VuT`Zd?sje=h14G$`VQ82Ly2np+@2 zWF!{ET3lqheB!stuP6eVP9i084^sZXK{G2EL&4|2~! zS!IDx3st$Tz^IT1EHFKE5oP>`Dv>+HXRR}W&vRaer_GxAVP(0T*D?J(v*l!z30_+? zk}R^Ho#k@gzYN^#k(R50l7SRzEfsTF)+fclQ741|dOR>X4f;+H1u77r%2U3e2=q-+ z>8S#(=|ZHi5_zbSMT?^2n*7UBOxoz1~*iWDiAz@F~#UB zO|QDccKCPrLyQsb0J0nZ{>Yok27(<5T3Erm=Cbl$o-w~{@iV*#GX8wVd$GC13O|!~ zF8gtYrHH%JmT57_vX}W@bpx5ndkfUpgF5AHVkGo{m&ohf;oHKm2*{-m%vi?(>Zy<9 zr{!6$MsWkX1TLRFF%yNzvq!c4Dm#dj+*|n_o64VQo{F&S{EAY#y@y}&&$I_;W|T6_)}18GSrXLcYgHz zNT*q&V~m+}#3@$jp#JxYJsFl7+idE%oVaF(L6N(~YgJ?^csM(xpDt{b7uXuyo#+&f z51)ySY}<3aTFE6#;uxo0T^JZ`j9ee6B#Hk}i5mK@1;(I^z1&T!KjrZ!+PGd=J)oU# zWMwBNUvTq&&63~HX<@;@Wq4=K1&;^rYn1~~eDz4$KPzrwd}x|-#hmKUZb&L==q8BX z?XfOeXRkQ`O;qyh-@mK=zi<5R7k~VXXf-8UL6O)I4#bF;c6v{1hGB&b>Xn~oEmNBI zyeDHm*i6dV#nzlSE6{AVL$0S}H593$V$uV9>CS*BK}o!dfIRMPQYp$7{^?poDz_ov zPx;(!q*Kxmkk2WT+*Mqgdy1n`fJQp6Pr7Q#>M3=+e2zw(5w25ZaVoqU6g~7k*E(Kb zWF^%m%ZJCS-5?K5Dybb*B~|U7EiC7j2~)YbN*z>D#R~QX*lfP6HAf?WXVpt-A z`eZGyu|t66(~di4Hab(|$`4rXc-c6qPK+9c{qYshG0F-*@6#+^1*t472!rCWJG_5t zV3jJ?3e=@;-AY6KXSRQi(9ouTC1lVulin^LbnoM@gFN9KUYaUhusO8A&c(z=5U>Lh zgB5WAp2Gh7)HY&TkIG%LPLnlG9D~g@TVb|PvSf-RQ878n+w<4DsaJ-_`}KnAXPbBz zP+UH6KMX0#gRk#U#Pb{#^kA3|HgnA|ghNg<8;vmvD<;3C%bji6>&Va~h(hD)o8*pX zxv(m@P*EG)7g+};yw^9u9lmoZ()iF4ve~;W^f{VtHgkaAHivxRV+_h}4xF|sb4ABX z^E`%AH|aP2KeBNODKz_b?x1986xl+>+$K4T4~iN=6KabPX^HbWACZE9{T_|5tK>V`FND~+;$Xf@$tL>!U`YGyBXNg<+|oYUsf=+w|CZ{H+rQPrk75;<6R&a zazRez)gnEFQ60?-W@$ZGK-Th6haR(n?Bv9%vyUy8l`n~Z%Dg%P;hb?>x;eBnx+J0EB&OR<6{(kB1s>7=lL((_y#C*{uT-DZKh<26xoP+Rf*F;$fHec$m^GJ+GHpWo#@);@jNYnaYe9to*lY?j2@ib zk`i|H@9)Y?sBnMWf0=9l0~`CykfNbv!0D4s#pEhcqIu1XUK+*4cAFHq-`Xat zi+ZbO+D6Yo&l&>h8|9Eoi&qg)Za4Geu}K5A!O${PAZ!!Y(E}bf2SG>A&hQzl(SvKq zsQ(R*V*j7a==t#v^M7Lt>w@z-+=PL3A!l(92(Kmhor2~stct%b(r_B-gmE?K>_013 zvMxB_jmH_X{}d;AmdSkdzwWh~>~!KqrPj<;lv6TL&Mcy0uzfEdCiY^)VQePnw#-VA z)=IPe%j>V0IR>YOR=|uJLX0>Y<-kZ#nx}VEJ)Q@OE~y;I#3vx!(@DYU!ws&d@pW za>C%euA+usSwpp|ik6&L<%sk|Bku8975$iG@f_%9FhI#l2s1qGxYNuK_Um80Fx<4# zNdBVUK}x2O(`E*}hLTlLkkSUWq;_SuXooMfcqef`jo7rPoyI&@5;qg}gxDCj<>i#` z?_7FyNsrglY3fARQol`$z{C^^cFK~tyOw6s?cf2LPGSXDu}=jo3`vp8RgfH81{6>g?6osz| zZzAQviBOJ&>hd>ywRDE^a8QOEudjs0toYDN9xWg)-zLUbtz)c_9l&iXj>mkS8Qh$o zytme}MPl?i(vdf9LKXqp2qk@tFZF(^k zGXSjkt*Tf-nePo?sXedKsRl*ZKVC%2Bm-gnbe0HT>r^S>cSL29B+z+ZGbJv#Cuony z+9|Pu3aDu8iE0$~(t1f;aF!@mkT3%n`&Df-?uhJfNG6C7%T>~z;4qBEjF2%_NXoyl z<9!zsYKlJl_1{RW6Qd@>3^j?AY$HY1Q8Bp9MOG+agh~Mlo&Pb8QtKikcAxO^2V@0_ zNvUsc&9cn-zEEooLwON6%zEG(ge68BnhTsvT6%Z5Awg3hG?Fy+NuS5c^3UmFXF&e> zn8RN%qvieAXRosq>vr0Lcm`qaOdzd54vZ=}N{L}r(<6zV?(o&~FZmQlV^!gYE&0&^yV5_FoM7IHUoK9Xu?TmEmLiqlFnB zW2`Te{#y27_N3vB&aMN0CQn>g4dBEsZMxZNw2_jnqsSU6CjZAB-`uxUza)X$NqzcD z`}-Z%&FHx3v2GZ<-_`%ymLI%g!o#Y~-~TbGnL@6beUMF*>>~<@C1cw8>!Q;HW!%15 z7!}v4kiEX1v_{m2WXL;#7;=lSj?>O>C$Tf0N_KcxyJ6z&A-u|^k!0uzcbnWuH)LeJ z$0zZrb{d~r1yZK1s-7k9Km4~7udI0e{RW7G9TWCJS=$W}KAi`psB5OI3R%4%p401_ zCRjTqL!Rgz7c`OS9Cs>)zpJ+VoWMk}`W`xOC_?<>|9m>rgs7G8P2WH=*>&PLaXz`) z?2A!C$%-hFkNITuRE%WCot*j*Sc;|&dN%o~ulQ&~Px@rYZ-gT!dnT=usgYt5J1|=1 z>QB6o{TP!-=rC8i;oT|H`*hB%d$c;%0c^PT0{#TB{me&xbt>x}%b-4+>Le##lNiGN zIN?}Cpx!>cfmANKB~sT+Yw*Sb`^rTVF=@~Ic*nVDH7dp#$KID$UwV6u)MQ7tB!yfe zsb3m90_^=GlpOLX8Sq+W8TZZLz|LwDpW_)Tgxg5plnwxQw+1wKKUuQx`z7DJ`txM!yvpI`Wuh>#6$axy z&v{Vz!8DUS3Hn|2TV$&fN7nY6S%}?~3}Prdf#_EHm^MV$gd4a8P|sQo)y=w`S(Uuv+?tNCOHzdwddM z6}kRq{L=QN2~<0Ofb>8mwV^;)4>a{qT{m-(n<3MBWXSh=4@hf) ziFifygUD+CJw8qLuZ6V^Hly6nQ76Wp{E44s1sh7K%_y^s+Oyg2I0(}*yV{azCx%f=yk_4f7(esUZR;@4FSe}&N0rAK}&r% zz2)V#-o1v`>pjmXkIwHUjxbwx0ocTG#Qfx6{?}SdViKoaei%qdnz)^E&_{943$F{n zbdFl9*eSafrKc;Sa^dj&(iG`E-~)bMg}U%rvz%?_*|GB#R_0{Vj}Cvz@iMJKC%^mZ zHL}TxS0N}X84>x|Ny$LybUPKZ{_EF8MiGy_z6pMNMI&WAbSm8L9+;~auOqc|K~S7G zihB%vU+Zy^;|LpIF_|nuVp?k~264b4E09nhiNubvw{+xZprMe*guI&zXe@AWCj2~|W5!NRmf*yVmKWUKfI6veLgJnFwXNT<3f zUBkOdbe^j{6M$*s!t}e+GRY&-NE^qupH;E~J)@l4XPmGbj%tq${=1dhOpSDHNd7Ft zf{(3F$aboZ^STh4qr}jJgTqmy{$Lg#=%{BMd4iQSciw!LPxH;PmrcINg7?q0lWKN3 z8|VGYNSm4Ga+;Ezq(~!#*-Y9@ecZhP>U-pq;C9a)-W8E;V!dR`S5XwOV)|iuv*bQ+ zt2j3}iPJ3Un^nju5`hH}OSB|n)nHSyzd+f&GYPa3t&5}-O5m0h9OI8G{kA@!&k0-YxCgeY}_*y)} zyo_^c0^gq*uW#}RGvO&aFejDlVTUIt-YP>)%?M9eOUa;v;~*9DnXJd}Gtp;rEL{(mlZH*b&x9qpTo4885^L zkJh6ARS3ItG z)Bw}XwE#nPENWb58L~U-EY4on6zLsryb2rYr=aiqR8kev$K41KxlG};fDFG3zpIeF zu%AaV`ZKm^>PHV}w=?XP)H^@?`zKcR9UIGivfcZnxlnal>Rmp2b?6Z;e&`t>G&1Pk zNpupFwHlJ&RXFm4H~Xa6&Vt#)4m@Br3(mXC46pZ@`nqW$IzRXFByx$}Lgd7b%_Fmg z=r$z-70M1O220w~IPIRz3gZIv)a7zWH-GqbW?(#L_;sGI9NJz;&)l<+}! zzmAF$dD}vo`N<&KYv5NlxODs$el_=&=vat`-UaoS*b1K| zssIWHbt}hsaD~(>`hY2-5@Z8Y-S>*%0-nf_h~7$DZl9sQ%EFUJ)mb?=bX=-1BYb=g`Zc zFc%%8wco+6`uLCs^d{0GPYc#?Z_nQ=`BXmS@dWB03hBg<(#2P1H*ssB$36>s@yjIp zf{Y68U)(o_4Y0$cS&Nuwm;*cf7NdvIU(n>MOpBI2d`22M{H0m6Kv;f+19Xa#9jC}K ztk8N4S*V;@dptTouNGTewV|;BbwA0ObsX5TaSMbe{#r!+EoYWt!Kf9tcwwJWal`=d zSf)s`A~#13xu90A8dG2Oi}pb9uTNUFK)nYjiVU6l?S>9olu4~pf|<%tzLg2W^6M7j ze~TmDQs)~A!?1=Y&F2jJfRY{1^bDYlyfhjy&pPw1fA_p3H=#*&>WwOr%??dYY=1x` zY6P0fC>b!m6jCvbP)m_XZ;<6H(|nHjr3F^Il}>&9RbA9^ztrh^{u8MC8(MJ0uPS(} zxY=ufw5rkrF)vZP@b-L+H8(^YcUle75MftEp`r$*#NoD5( zo$6#rL&SE)KN2{_Kd4)>ldj_>aP2eMqrc5IPygs4?Q(`)#OHhDqgP*a>Euk|Hot1O zT;Z|E40(x93tuOP0vw3XY65^-5sCO$dX)kZ05Z5c3JjxWjx9Py4d9>?tk5yJ;lIKw ze`7Kwe@j?9kKCF;71m(i@OXo&3hnxarYZrz6Mf&gVN*jEZ%;< zQzUz4BBz<&7?mD95zo=M zWym*qehzIUI(fbbw9)kuIkUjCoOL@KHPRtQ-|LkTxG}skAd$Me;4Xwh)Hw5rf?Q8E zM%%PN5OF9fR)Ff281X0&?tdI>vOQfFS|oKdXxj_2r9?feq%08~>vlVvrOD>tZ*`^x_863d^I`Gvs!)9ahYj)wr-?EPLMa$YYN2ckFd5~gR!*9vZWRYVpD%IsCd zH|~sC4Ub{xY1|R*bHh)`t?UTXdg9@_WEUxAx1Kn0Nal>0*{G#tkXJYaL1Dp=S7zYm z&alkOr%1I2`6_L2*8*<5{&k2i` zXz30q-iG#M15`-ZWmd4dVI!k~Q40jzuw7bw*Qfmo0L%1EFfV5Gs*%uWs*bZ-LGp>jrdM20el?uV3Tf$-t{ zTqv_iklF)z6T;Oq42kKI&Cp`T)!Xm>_D8nVkvj3thoO!Xds09mO|({(<-2!oe^h2* zm-|VcX7-K9ry-g}M@97tR7`0PRJR9e=yrZDy%I=4F)s-sT6kE(Z6GDw+ln-}YrW?& zLj+2l#xX{y^5{g8$V&^pPp1XvK*UPRP2`nGiX#wk%gHi9%2Y$fGmVoV%M$J2+!ysh zHnYO__WVR%FBuAHhawvd*hUOSHPLV91DUXvJ{Pdb6N-ZJIUAyHhN|(`|4foE)rJ;v zUuMaiJ3W(Ov}4OM$E zNBNuqf|Yz2C?pkjr+TjdNoKJ-|NSCw2J3Hqm(>W2)?W zirM;%K?prro4aO{rSgr_Iz9|4H^?KLCSEHriZP`~2c$`ynvhOTBYj-4ea=DAHhJEh z8gfDWII>BYA=k_}HT$gp2H+iet7<__$gz+?_lto#)fUb}NjV2^cMx(WikC0G=5~e2 zKL=xl7wX5~jDA~avMSm9)SD#UiLDCo^Nz488I%k-!naZ}gPu?^Gp9h59H=g&%Ya1> zQV3{iG$0whHMp0kyXVv^&`C}XG#0+t!&&*~=gs(HLI7bL&3+eH0c4U#%A@-t6F?*% zeyx=xI59wq%>a@~$$(8_8x><9%!5)X4Gkm`xa#be?1b(vH5L!&(LLab^poYD$j1t) zDOjA={=bcs;MaeA;}0>v`0X1pfBwbq-u-Qih?cn!iW+0dne=Z9f1!BSgpjpy@ry{$ z6tdndY59bb4N~M0>aPwE1E;*;Derg`)}Cb=s4qVbD_7`LEkGNiQ#I1(`2_*%Ik2k; zyeW+pWXP9I8wBg#=BxMLrPT7D^6CPy&i^ys1xPjTlXTL5x<3^p%_Ke3)5K-8PNn5{ zLf<6heSo?ihkpvJA!*`l_ouTCir(s7yu<$-Xuj6-x)vT*X2HAY$j0zBQ=ak;iq6s{ zbJkA5b*7bf0wO4ox`b<)^hfe@-tGM5;Ym?CRT`(fO85pW2wH9!_vW%p&U#_$Z^WAR2T55GIb??-b@D0=y3{svO;f~ByaC2B-K z=MW`3K#{#v4Au~LExhLuyP(YX40h6*jxnHPTwQ9#dVgrd)I!D_5r8)s*thUMc&$JP zZMz;>P=Bh%hL=a=_T0g#Pcyzpgy6Fg8pAVYe~1i^Y}-#}d-?k7o||9v z$R8$_mJ9Mlg^D(DfAGea)G)<+XZ1jsXVAS!Y~%~*lcWO-jIZrB-OrwsiGbnRBiiRG zGZ@B->m)v2_Z7?DOsCyMF?3sAg3jFhpg3NKq+7BRdMBX^7ZTX_NVa<|f3^Dn(QuAS zj)NWOmOx}!87t_eTOt|){!|^YX->BA&a~C;*zR;6Nbm=_`$V_A;)T6njFqx zMwvYGv^+By@|~V7+&l<9;aMX?K!I?JFcYM*Yn8`f*+b5N2^*S;plNu0vF7JQvBe5a zRL}1V3w=#%RqHSIzfHC|@%9Rqp%Lp+J|)Ye$Sz1U00>}8AwL-_I2;Al64N)4%wRoV zr%DLyTG&gcPT%Lf;^iIy1omF)RM?D>E4(jF2YRJlB8@md=!z`MeZoMnVxnxqiWTEe z^cHql(Jzen@GB;)w0-M#9@)#zJ9grT-dQtb9i?P-6w#s>YbJe9hFz5h=Ny;!MIu{h zxi9iRZsk-ejA?U2GdPMrb%GQWHi2$j3pG*5O`6oY0I^FPXCX6VQ)0E4r)QW!?K{Y<*WiB(#5?l8UQkh zLQ#T87F2K#!rmFmqNv%c~4zXvJ?~OXX#JC!NJbbvmU#+ z6_NpQneX0!BFceqVhbSiHN|lVVg}GIRs0i$$&$>@{4j~+Ik6=DGp?grJ@Jr1Gu; zPpZ4P$Q>FF^wn#_3xpsl7OdVT`&@)i*$pp@s21^wJN++aaZ$Nk68KagHkvhzp}0a|#7be)-%~Zfej|m$;~6Zkw#w`+>MvbPW{P z?{f8iosv_$>=`S0$c&K*HDf)}WMB$7tSBZu^kDQ3-`c1h_N9Ec20EixqOs4|6G{1+ zS6{be^K{x!F9SbkinL0;Hvn9eSV4c}kRmo(-6gH(VRqoTTM28Y*`8Bq%~wY)g9N-w+?!rQF6rUt7p40o@eItHz=Aa!JK*ljNe{n3 z&M(4!GTi@Ef>GMb-1=FD1MGUjYKM+6B=Ss=Vx!YULa^2&Wb8G>%Cn;(FwPlPhGz2a z#Xnf{o@uk3^{8#$v3~tD^iics_eUgM2l)>=La@S=(P@%VN7LE%$K~8;0hWt3) z61v@|Lo(#D+w)k+&Cs(n>ro)Y za?t1Q4S}t&%1(`AQyg)b*-rg%@A@~sXR=A}cE8(2J{qgE%83(=ADadBu23?lkhnm_ z;7Yb_R<}|wNsnIbo*{1{_j%R+?fi#seUa*uQP@0P?GMel?j@ipXUse9_rs80J*f>b zqz^BMuZwC!(xdVD8@`xsG}Qn8tugq{v}*74XboK`Xm-`f)1y13*v`AlQ+)6~j9weFJwR_>$H@lkP)%ca#56DV(8Ez-`t#_IMFPV}hQ6!#Ghyh*m__3W(QU!P@yMgQwwtI5tU4J@DvctlA}IVCHh zND&o-q@W0k^`2N7+zy#1^q5<`(l~>3&f+>CEHzg1;Z^jM(SBvfZ$kPKi2)kv%&E9t z8+6YXJqast!yNh5&~~?WM{sDd0S&Viu_s)NV@+EysGdmdJDdZIL_0y`5ZeBjP%uc&fSF7Qncy5l}c z>(N0sMCc@j9;LgABV0W!F}UM~mZfy5Uc7nw-UZce*`Bq$oyrY7Ll;xLFe&meov<257{V#t=c70D~{(qgsr;mfVUaiV;dHm0nZBBlXiXot9h2SoD3zCDd9zt zaZoCWV#R~fG?`9?N?ko3&EYM4J>3`GsyeSq6CV@y&_=;*W5@RexQr)bzH0 z@XVi(O;f@=tj$+gO^Nebn|*(rH#FVk(&!h5T_@>I9Neiib5aT^85CLNQZdWLu>tj> zJbE{0i!fJsmdE?phC$$NuRr-GJU!xI|>7d>|H@?k-} z$lJ}yQ?BJW4#+ma*ntsbb};^l%lc{mTKgj7{>PAQfYnTm&A3QaQ^eU3l^5?RQmVSWSJ8ikxVlql0?bkDYBl5sUY{~mPBjix996rI==^z zr=#xAz3Pa4%DA&&)hv%Yiv4e3XDU9p@s~@#GGXHPZM*$Q#}u;6EI8js$?j34hl)W) zZVZ4zWa%+Mrd0fY1<^xSm@&lFCawT!wNB}th&V{-;J-4rGPfcsR?sOeksN^BObZmB z=*7_g1L1NcnZlU*K;+T+tzPvZhZU=)gS3XEauG-)aW=}f2-~1~GQ~|JuI3q)Li1el zUWmUFE%dAoxoDxX634^$93Ha$fCq%oexydf)NJVTD^N>$C3_ zzUc5sbp#I$CW zyJVdvYo@^dVq{=+3nfdYND>v3#X-8><1Fr_+}6=BmHF`d_P@_oQ;0>v}1uij@Z^k~k*)bH{=QN4Eypj{hx2 zzr)?s0e?#ggV99^&Kvk446+92L5Ow}N%ZcSUnWW7HG))jqIWAuaP>qrg7@1_YH5_m zjuljRtl${$=yO=6!R%AF1_+ieve^I*I$*YNfBCQf^tR0Wu&MZVUN;+r3`VYs#gq(u zSq&9~RSAQhH^OT|4vG>2Kl96wUl(aO(BarCN(oO6TuYiIH$#s=A(J7jZ>-*|3c3f| ztjF}Hq?%`_L?~1giHD`K9pc|k1WC^WS!J==RbD3d6ucG?{-Lm@W`KUDh{hSzcO=*mTH;w>`u zJy&h$MnefN<^>%+?dDU)0tDs{I{1hg1Y;z$C%yd3{qNi2qB?Q097D$$h6yu;>r`n# zGgj)>t#rJ(7<C$taruF2_tK=<`fRs(iplCLISdv6T zV+lSu4LIfME2PD{hwh6^hM32u*|n1Mi!RdY67EO9bZU&+UJLrf&*3J-idnaG8Vtt{ zp7crVtj0=jZ&VKP-_e07)pk)3`yehh%lP4RL4v<5DC@gnF zMWZ}fRqz*QbgEi@xi6mEK5KwH5gjEr$so6p#xt!dybqr{uR?{c)$Y}9INr$fb-Xih z-eyov6m(-Eal}bmL3!)d)#bl8nVN%dczTgLPHbx8%+gSgDcJx;9#Sz#U)fJq0@X4W zE)_wiM@K{&v~skBR)ZXyF~gL_*(!b#)C0TX7LU}Kcctof`cN=PvGGnpe^DuxBoRGwCl!gUDoI z;y#I`BI)Q@j1xOJ47#Izl2XO0kYkGOh#d+IeIlsZ9|$zDfWOCUPzqWphZSc;r$|Cz z2i#}qWYhZ(kbGqtfGd|)_xjfO_Rw#y@jl>{5m-JO%KAa$7OAY!{l?pIN4D3ao&UB5 zs3$)6HfjlCC0>n58Dqj$!BJ+))2~mb)0E<1mx2kr0EAEw#zkT5Ky-Ra`dHWmZRhhw= z!M3h~4nx<3@os!@`PcXtY>D?dv4LWU_YD&}8(MW+VJU)kHrM+g*#rbOdb&_v2vU^A zP@)d%l0_glnXMdJkO2nha8QnNgEu@ZzQ+d&=qy*1D#(zd7NxD&7;6gcf|s!dw((qo z;^EIW)vWBlmop`OW;J+!Zsbm5!Kym?qmHVB=gfJkGozjS=FZ(qFB(Zpm@y zw6Q-1K08zwXcOm)(%sZ8-if>(uQuQ>Lv~a(`n;uzbf1gxFGC9dYEc4a=1rD(lI8Qc%g<-O7JLF5BtSF16}P~m~SX`h5$m|i-~D0H!Zkv6m)B#iO; zc6o(|x=(UIGC)p)AO-&W7qdjGe^NGCW`c|RJPM zTK*o7JG}L7dCClB9dEn$#lUW_r!GBS{ScbXkS9f+^WE$Dz`aNM)J3Pt=QNP5+>?P@ zUOKMOK_0Q*epICvpCapN84*`?WW=AvYCDVh?1S4 zfMhVH#j8)6&uIkeo7C`q3gFBSO^+^@>itqM|w7$%Z-g5aw+EJ85n_(*kU=ae?H`vK(iBBE|A*-i7!AP4;+HZs5)K` zoy8jjF?;O6kF4y{_W-)*mCK(yVbDg;$;G&4IT z%^v6d)=SIi<3N&EJ?$_*39=QL^f_N-PIIVtV+*dW!RJKpVaD~F{r{(Uz6mPtRjcYq zkrTVTC(U3{MaiHUc0Uz^a+G=WYLd-Go%FXixb@H(^I`?1K!~WHet~pIQ@EMJ$L^P; zIm%AyGC_`eX0VQPeo-FXI-|q)!Gbf>L2X^U%M%FpfMit*QpFcJ=Y0DjF{#xiMq0&W z?gqCFZdz_u_*yr6u~FLsm9eayt!aT_XfZ=&jA+c{#Xdh?{i2h|AOG%kQ7qK~!do>V zjbydzkVhxKoRb=~Roo)4m8_awM2bikRR2P^o9nX38{*5Jy02Ed4bD9`>z+7~ml~8M z$`sz4u2bC*?Fu(u-5J=$EfU8HT17|vc1~PXYj{e`f7%Z?=HuFW#0ogndujjc_m(t5 zP8-H!AO=EXafI6`2cr^nBk(AAO{G!0AY_M9YCw(%S|+Y&Usgd0FkUdYQ~e~@H$x5q z7)W13_7aEK?^o)kHl9h5-kp1HZmAo4J>qd@?@Ixi$w}1gUs{qNIxT1z2oFmYT?@}I z+Al%cO;D1(5P=6t`LTT=k%Wj%@+BH3Q4g zD(|7udSF@rEBf#zOa$Oo?)kvQDu9@~B96qxKgVVTkVzqmrCydr4KGw(3`6Nca$q~Z zXU#(&rcLi9lZ!3aIcx3y`2CNR5}&D7CTyc;_xmhsv3(@I8{V-UuYT;}`Q^^qeCWIvI&v-bS8@0r8a?-81^8tMNSA7i5NpH3^c7)&2+% z$UFAXNI8a7FaJEq72u(5^UpXnb^r$B*tT3?2F6%hqQ7*CX8+N&qZW#P8b$6+A@OF_ zma&s%Pbe}-#q{!v!41?)kXNi96cEz|SAKFpe4qDmSh*sJJ@McKVjir@$rWuIp&{L^^0&{%# zGD(5(D5+Z<&oPcz5z!4)CH?cVAa8)b&13qO*C7ztx(O7jNVE%v34LAT$uCFArSr!B zo-B$!U`bImdd+jeJrwuqitNth2UaYlNwpgBt zJj-mY_$T(xv(ye{km$_e4swynr&&_rv0RV^+=s24Hu0*L)<*Sl_Xa!w2FeTkyg+pKIpWh*?h>Q?!J49w2`CE;Oqm1;sPOh zKUnettyK-KDN{}5<=r5ylS3cm+~Xv`rtDB~5B>gqdIiA)NXE`z9<2AzSMC$(U?Y!3 zHMz=W`csef5dG;6`^e)Af$??lGMf~45K*>WCQz_ffKPij+_>=p>a%PbxK_^OYe-f4QBX6seQp z`G$Z?9vSjt(n%6$mjhvuQ4DUM1T>Gk!)|fVhim8Gr_X!!NRy)Q0d&ALbU&nWkx8K4 zv%&W6Df3yjAIi+fwe_gtCh!dYsmVrbv%)kNSaB^?67d9pRQnO|1IH0(@)_tph9DQEk7UmF<}H z!7sarB|-d`q}h#u=N{<#plYlyY!HN#vN*-uo(N4qfk3A!6d?2cR+0b>xH?s?vYlTm zekw_k9+zl6cGJrQ&F~XUAm!B#>%^5-P*gZ# zxjanCpk-wr6_X{aizpNB@NW{HA$tDSMa|+a;44Z9Y!Y7KWXu@k>r|~A3^slyY3F0@ zL_5@%<9&F~04Wg0hn6XexErF3lmR%-6JDD8eQA@j*(;u-ackrZ(XC$bp(ptE!;^J( z$?8p!?FPv-Lf2RRq58TBJxeP2-K6eIgP!YV=xL#3(8PU;ib>_Bk!qj=Z43S4Rf=>I z*(Q4cwGgS(+l3%E0DRG#$YuJd-yoe8at+ShcSnirp5Q_E6@s?V3b2t z47MVyg<8H0IST8jQ|44D)lXhEmi%p-mO8!KJzJ>8G$@jsg8>3r?M%T2@b)3(a#K_w zXz?;K=U_^!JK{kkRsdqMJwIscqF9Mq4W&3uA%ieHUO*WyP|%e_6%X>Qw{rR;HBd`B zQCsD36tM@8;fV(`7cQ}k&Nyv74MV)8!*@_xwEzheV+}2l8@Q&P%DzYhilN9Rb^+dV zbUtXY(N}<=1QS?&-0Ydda<_evD0mAoZSNBc&$wQIa7(*q3m5OOw}Wj*BusG5$9{~J z?>u=%$zqu$NA&3Yk`sG9420G5tq5wZPYclA=EGN|E+Q?(wMRp$Dvka}%`y zV}I)7on?WR@rJVNb!N0^OA3})?zCShWB^;|2!HsW$|zlizJoHlo;T=qOt@msIXab_ z&wajVFixRWn6T??TQ?XpCeCcC{ETbzRa}BymXlRub)z`3N3zQd6DgEz6GalJm>#cE zS(~_PA%qYM1W@tYPS-0+q3ZzkKrwj#e9>VQE{y)CbrWi*yO?qD@ca9IZ`ni3AP2IJ zv_OM0mO~m!+|W70)^_aKNC)k`T%`_X0{5s`IrF?ecNU-cbN$?zKk+dV($4#}d*%2W z-?vodV-uir-oK16X!z-AeQ+N*s>p8!rQxWw1bR!H`|kH3ZkGvMd7#cDI>v)isunMl zV#021bX4=;Cd@%2*ZAofANxIqN-Rfx30Q;Pa?5Kcvtl{SA^o{3OSFu3Je;@uf(a9c z1CTcy<>+h0djA4JN0?5Y2W@D`7KL0NNR&8ykC7?n2-he#UEtR2UdO{EbThOJr%ZJe zBG^{uSOS8woJzYLGaJG=D}VKgrAy+p8(#+Z0~-V~0{3|A4d?@E0ChJ^G>9`62=tIB z8gy@Us37g=DPmz4MnCMpGt3wmYcZU>;n&%g{NW75+ZxE;CkH|wAow5}@b{iKVf*lB zX2l0TeCu=l$?Ai5Uc0=haQ)Awnl>-D{`=}bNZbpydC{0{Ubay(uydP{C+eSylUR$A zO*Zi=A`KlR>ESm+yF?pRhB~;Kg^pV`EZ}Gbkj<73i^m;y$PDDy8ok4POxS3fckw-v z?8HS%P@Og+u#rQ_vMG{5#nkc}X(U`u4!loaRO}3_T(nVkOxQas2af6ju5fg!Iu0i7 zo_bx9wtC%}mNu;t>K%a)ntx2}aA>fF#4tV_h6U5W%#avkQK25r{O*}R6HJPRe!QP# zIWbJMW-uwGWW^LIpkh9Py+=uOo2(7?-8xl9U?0iiv`PDem#Ln*v`N={o&ia5WFxw~ zIL^CYiu>{t3#;XK6_=o*;J{qHxQE`q=#1!=s0FqtWxkb*o`jV-MwoG;kQs&!{&WO< z*g@vog=ruBzyulT`rz-7CU)sP=Uq=h>v*Icx=zWiP~;L4&u^R64Outb&!A1xhOUh6 z6(M&~nWSow8oLg5a5lLPxs?0rVRO7a9PxCP9(0EkT^CnNr-wKBRr_bi??<-K6_MDq zqys`N1Lv->i>V|!XK`hKhCVHC0P{8EQba$9)P_P-pR9Jrz*0MuRCP*+T(IA<1>*jv z!%Q4kQ}i@q*Jo36#V)zg^E29%LVx&bzfxVqE2aY;MRqkgP~Iyc>U=*;5$aH zW25Tx7+TNc|0WZ81v4}Dyt zf9r6kWi)841MBGL9DRlrSSO2%?sfmGap~%VQHb$~GER!NZ%7dvoh6Ko?w)g16dPT+ z2zzofg@uZW04O~WR*LROPm$sXjL_ZixFmzL!NHXH@ApC0+) z23zWDoj82YP+yB`GwM6u1%gk(!2rF_6zONOJs#*==pxnYU&=5rtfGGdY~S3B89=iL zZD^1y%oA$qEnkJY6mFLq*91Rs&;Nn;_4Kcq95Yq*n-7*eT%vzD>*aQSCY=mS z7Cq9$kVg6lP$7TVA^qT?Bm>+!{61%ecd6p0p!Q~lvWE1~Z@tyxwLW}6tmor*dDWB~ zKH0*!sKbiYQ}AvmTy)9pK_dM$eBAU9TJpy95QXNH6#}EKb@2EOZ;9 zd!#ASnuXReIoUoj%wKTED^`5-S*;MxTkn%8TEEn?AD&Gr(Ru$eV8tBivnNxsB#Oji zSf|{#oqx-#Q@YwyD?cje@Ti3^%SFyj>E*@G-2<`~-Btk7uD1k!gYyqd$t(sHDzw$e zjDZpijZg`i#mk?S3FL2J=+Dwf%#5^5H$-UNhcB+@Yz`iBx#6pY?Cd>qgll|QjW?py zb3Mt2+jpvlT#Tu_TG(0QH7q?f7CKmGI>z}#to`g6t;CoY#(7gKF0-;M&g%w&Gy*hH4#mmZFhI1C0q3~QFGnli}Ekm0il+`H1E zpiZ!qm|t2wmd`o*m=!E0XZ>{de}tN>NnA>ADJgbh zYjVoWnp9IVkS#hu#bElUPyno8ne<2UV(ZvcslivO4&byW@P+t<4g@Me|wXpq=J$##fXJHom9(>iPk@;dQOnL#iunX3E7pECb_!`J)CQ^_fEOo+=sp&DeK5J!SI>2Qyc}15%V=MjEoq-K`*bqmf8BE=Mwc0 z*Qt^tEh_jKgYLL!T$o<*JQs}N$+2#hhwpxVKiT;pJsEEX+L9b{V!OeR9EyASy67Jr z((?w1;F~iW1Z&)yxuB*6q+ok|9>IpX$?uwYH=PGbg$#%)rjZBYi;7fkB{k??>Dlbn zNb4nc!nzz;jcFaS##mbnryPg0@^}9_S8iHkRHxpkBH3S>HKxXFjVYsKdni&!#dJy< z0#5PT-Li$pC75s8;HMXF0lx(6G2=tBxhI0K73hgyHy<~$fPM)b#;r*Q; z4~JJZG~S8Dc&*;2WR6&ZY{QHROAdg}(i%@C$Ap=8PV^rqE5@qYc48|6#knI&u{Kk( zM2c(#WcZfRK#~%9SAiqxNdl$H-&NP4ABc71F7;+ zJNId4=en~y+y8cFxBRB#?#%AY{4$;Cc4rW8cmYKL1vCWZB8V3RMO1DU#cEZss3_jJ z1XN0^pq8S-|9O(olE~27_vxq2xxD91InOuG`@GNP`!Gu-pC~s-ZcA278h*BN z5@tdcMPWAwLttZAk*F!AhrR3MIv z!U`klJ?U@BhLK>z-{d}U=MPsdMv8$Frwl365LwwT}tBSqgMgJp;d={ibHY-ZVpxX|iiN-97dvrWm7Obx)CAv?e zkld?Y)e@GeNethvfZF$c5KIKFY~dhX4{@Wk@cps}b4sAkL$BIo?txX}e?(&$bANh` zphSC6JK%;#S;5=j=$f>c9&m%i3tuG89&jsF)%qoi9?J*ZDy0vhclJeTwJHnLSCTAO z_mM9$a<5h*BPIl~1t;81iYz)0bRaw2oC+Q`P_bWUg&Vs{8lf_2;)dwNzrJII+LfQK zh$Q!U47H6mJ#>~&WR^}O1IV|Yq4+?89R9gF#72&;%NxrVyS&@gD9;wi+{p{iqpJvx zL9X>e$whjf$MKLN8W|S$sygY?>Bpw#>I#C8P$5ZxdE!e0PI(>Co|Pj9SgRNl|DgV3 zc9v|991&6zP!n-o-XnZpXi`}EDkaT~pKKIhmeBja8`e!@H>G~Bo8i@Tk|+uG_=v)r zatv$XeWRdA0388f6ZcU-gAliC=ycwpMW)quLASI4!W1Tj5t<4#G4<~8q#sD&SF6j^ zt#q<%d8p;@%%Ur%z94;c-iZ7C(GV*P_jeb^ll+kkDjuiG9JkSeWfW6Nkzy(qbAiEQ z*sQ3UZ19CTC8NwJyWn>L)Y_G4^?{J?YtZ$oup<)2CwtZXUY0gZl{7GX6bqV}LfG>! zg&8of6ni>hCjii=FfI)1klG`5zl=dGb}hxKCPu#DY^65xw{XA8uGH2@7p@AoB1iRS z{Wh|fo1MjDBmImGrm86BCj%r0d%v5v>Wr~bh6v~?4o`(#<+IT>qLWYw*S<-CkZOU z)`lD1#||-PW6QcWatD*$6J9gsX50G&c$;Lyvj5))`}qf+N2NxIk$Egi6K3mbMKyw4 z$sXz2MR~$&s%_eXOsB5Rw=u%;kv;TMpmi=Xdj@)=wrbB$Edx^DbkZATl%$iZf)z7* z=l}O<`u6OxLc>MC7;~utyV4oKlA`V&{!Y3*aR_f=#KLHTk*%t^B~m>W%%bjimH}+{ zGjissBZzS3_c=NjCqzuBFL+!1b?egf#_jW&JfO-KRCpNXW=+SVdKI+T*4=?VA$x^4 zU>2^UBP zH}iwXiwm?n8fJbJQVeLJ94Z$5#9iLNF4PJvyVwU1E87}@HV8wa6&^^Jhqa#ZlB#L_ zUV2O`zw3?(f=$qpJqw~6YvyLrC4$`%E>ufnf)Rjf6bNzXk^ft}W%pPt5gz+_SuBxU zVB&>#Y4W0gq5yQv>hZwh^>w8l8BrRJqGH5{U3f7$HQ=0mF&j(G*cL_C9Aw+d2NJMO zT3->pJiKs9QFs=3H1nO^|Mt0mv7_(PWj`mD|3h|VJ%W3RzxwyzKZ#4R??A-T=%-py z0Azwa3o_;}eW^YyS63C11niW?aH9Zgcd8XPbmGcvLJrpemI03VZ1N&TjAOFm7y@_Yuxm05HnSoN&Wm z+7`e`rYzjLrU-5E!h?%>4gQ#0sGd)FC zpuuQgimo|qhw!dge+^987IJL7+_9rU?kK^?=`l{oo$!rX#$eCz&O)Sql9bKg7=SEs zX8vQWJ8e;8r`OM48_vID_G0^rHQsteSZdJHrd6xf z`4}{}XSPG6w^`AyE{3EfBsIlk7d`TFKxadrY?4EE3X%k?BwgzBbh=9l0PN7_ICzII zbas=t;FJ@1C$u!)dD}j#fF&H?NuQTuk$OhVX{O7&kDP&G3+&TfCCU>*8aG{NKJa)-(6BKV7&M%m`^C$KVqggsF=jv|A6D0bh zAuGd8u*1RPLgbLDl&%l_C}RD*GofqdUW25w&;5Y#0Ev!~1EElY!w20gnP9E3`NT{y zAX8`}iQy$t#_)W|=IEG-WX7h5MCL?5lBgo$6y$d}0XZmR=c8xr$T9zay%-S>rv2SL z_Fj7~e;I#QU&7+!$8*o5-xF)LUULs4#UFX8J#n5 z-sPVr(ANF%!ygBpChZP9t*=7I_%lNp}lY3aIJui6xh?su$s2Bnd!Yo$$!QcVwCq`2js zIt>TmP8?!Qk|ot4OMUbWbk#Kd@^C0?%%P8k3~ICJd`Mk%4wxK8h^^*$=9%L#c*aO} zo8ry$I>y6lfV``}_y>}}EpLR!sLHcJ)fS3LrN~Ag=ul%7J!Fm4dqN1?+*xl|biKIG z6OCI)PM&rxEWtzX96UTjBwya*2AVIQbM=w1V7~wKp$*F$%9UjMNZDun4P29I8vyL5 z7;x4XLE;eURVT@?Nf~(NnLf{M^<@QQgj8y`1MltTm*b%%8kWSVQZI1h4PkYo%*}l#F19OwQyoZ_ zjiyPam?VmRh}ng(X&c>dtO5Eoyi%)VTVkO?>O{qqxpr&{d}UE=C0vW=Tb z%43Vf1uO)=mT$){oqy)zVj1PkQQCOHN@QuRBbw5oN^KJ7Bk4>j$YyY4B0p9voc zGoZvmpCL$*mCrjib%Ury-3L{7ar5<$ymv?Bd&Eof>AZ-7h;Dj`E1YT$q3anoi1pSN zXm!?s=uy(2?W-hst3y~SBFt$`4NBzNZxyEm0wKi};eg7V+FmwgjR2Ak3MI>x1~1s( ze&mZ}6U!lsp;@6{=G8q1nfj4rqKGQ?D0bD(#Rx3s#Q*+ptT+yW!bV`@al#E0WeW~{?+xpEb7AI{3FPt%v)+7W zv)14(W$F%k#?x9lkARYg}Kf1Wd!HIJXw%{_1;o&&+LxK(H-C+OuOBz|ZWCoIaPAI|QX-Vw8E(&PEv{+KM^A-q6m1)L^F=N#oaJtB*}@l^&4inc3A&YEA>87OeqQ-M~8hcrA~gQ6ntz*Ws=DB!ZRXU*Pz9_kNcB5JYItmYajk(&HBv3s8+1p4%<-K;w+qTk&|LiPl-j9!Y!9?YSRul$sBOOb zM(=j-OThQJMNqEQr-xh#jg!=ik%})%p)U}eAk~Ti6|&df^{o^n$?zKtnvGtML1Q=5YeTej8Ichhl)7zyL+1vg-l3hx(KlHBdq?S@SY-dIA>A+4Z?dLt8QRj_#mx z&@E1KJphdz%Wxd#YIHaXhGFK6<&a&DPPxNw=sc`mq8f|x;BmNsh0J27mAl%Yf&TP5 zu#quIDvQpN^^SkLGe_~_6cBKjWa81PX@6U&FgNRMA^9^Z!%_kZB^|<*fuAVb)wr2* zwbFP>E__AB#$9uk*>e(bNkQPTu3%v>z<^@0Z)t!YlbF!n7<4}3 z2BJvYsMuwp>xJ>;A$>#M7<2eLSAmWCLQGOfQ%tSjEoG~Cv*4&WOWvUA4qmISCKX{l z(1WT@c3GOBUK#0n-7y+004`i0ecLJ7O^yp|zNT=qqNM2aU;j6Wn*@g4aKGpVibKs}JO4hta#Dfdx*Qrg`IgUX4tor_j=Ozow5KEbqOL4R*DWL4>GY7kr~#l`yh<*9 zeUJ2LWU(M!SG>@29eS*U6w^2t_C0U~avVa&a1&#sV;A6rUFZ4l{Z+iky0Xkj|ItaZ zlE-e?92FOoqh9YGuOr z1LHyqBAQ~>iBbZy=xx6BVSS$WAXL+@t^xYUBw5F|ZpeG(rwR7hPLl2RDECTaAQcuH zwQE**Kbg}MQ>`eU_lQ{TU}6FWbCA)*pgB!ikPK?>R!FJCx$db`IX*~cwGo5x%1vo=dsHN{j=-%PmH3Yk$App(%yMS*$sJpSQ}j^ zE2i;U!X$&{GjbpZ+cez+gKDLyUEmaz=3EF_u;dINeX=BHnmw(~SIQT|LZ0&>SivaB zGbAoFHGDwTtu7X>^fYLCeC`W-#DClc?L+%LD?Cp}XHD-BJ4M0{NP{y79GKG?L)aH9 zRnwM%uX+jDMsiMekqib6eO?mu(A9Lkc<;P4{{~H(e?^#4-ppjn?y5_{QJ3a_z`N$P znuXoLK-f>8d-aBVy+_K^brjBc?L%{LhQ$|4iED#B?64TY5ktLkY8j8B+Hc-wtQL#sqlZ_>=1Js$%~GFFG2lMUreaHd*ZY;q%Vn1ZkLC&?yU_FlZjl+2@>voAphM_7%gAmvv5O+~D@B<*mzZ*@QW*V<8ge z@s@~%NVr*X7;5@b)!2*6plMKcLWatj&}~rkZi3Rmq3bSJ9m6ol3F5<^bM@gp$@*pX zCAfCr0eiWJF=c`-qrihj?2Pi$U*`O+LD`iz(A z_$`NF;rNtJ8OaG26W6rfuk^QS9Mgh1H%R6SKDX_Wnr{FqOD*m7U+w8uX)^_jjg9olhr|6 z-P4INe61g&yOQN>N>=5eDOA95vs($Z5w}-Qv6}9r!$|VnH*UIB1HPaz%_@SU> zj!Ds?d_*sme4@;s(G{5oEG&a=JH5_BzIpzX9oZ^7RVQ3j%J{ey{ZDovt+!k*lwqSf7uFMuRr=iJ>ev&UR;39=5r6mh_Kikg5Iw`vFEvS) z8D6hySML^fM*zAY?-p|pkg2Ir)98%mgpP^HI?7=bahi4PgQtb3W$3IrC~<2*V%dz78fXr&hJ+>~!ooP!$T&d6d}LU^Zc!0ph2G$Gs#?w;KViElWg zK~p}jG^kp9Jfz;UMijTGD|DqZf({PDVuVG6{nFTB!AZGLlf<5j>~}(3lnaj?_$-NW zCDN;ccuAH_pBIj;E1MPjUe<3>bj?W;b-Hpb8d+T>B6;7wd<|aY&Yc zHtH|jm}BcogJ#gJO9VvymLmU$f|S5!MWWqM8v1b=rIM_=#E2(zwSt5bXSk=;CAxlY?`zAGr^Fh$C!*psALx&Ecp+5PGh zBt4`?loa-Hr?I9if;*6l9OamVr(3mo{}%%uX6IyMJ8p!{ndTinc@3!dmbQ=TLXP`)E<- zk1x^3#H%Lt{_4aJ5+?PKb^udmAf91gn0CSP@OHsvn4v(hYSM}ConC0E18IQytdm~= zgQpkVKh83~`HHpgq`#s57b|)qmzrjhYdqFIk8JSJO)+;U(n-ah1w0ru1FA#rtzr}X zKn{#v;M9ZQKxsg)8f{jr4}(rJxS4{!G247=i6we4=+-DmRqs`80Nj}9UD`bnsCo|1 zGLr;!D_1KY1}DC($IgeaGtl5ChB=_xUJ6TK6-9uES)2ez$ABDW$s zp+h1xC`x1w`d!z-{JLCSK~Sq#gAh`B(x*f6MN~HF16p27%^O6rP{AQtM81`$?5SXe z(-CB+yc|EZRdbxQ)r6gzt|RvJK)l_|vk?BErHgVy3gVI|i~Z{?Jz@K9pL6C2B%U*c zOKx+5#KeN4*f*zHRnYR^`L82c+!in%Z=Ilsaab(%LAM5FNt7isun-ut@$pLFFGOYJ68t%W zuji=^R{*Si#kc=CO+T_w3j4$o5K$KWe=lPdG8e(1-S0H z1Q7PZ91TG1=WxY&!#7I)I>o9|Uf=a^zb0FGY}LRPYM7EJqL>1T6goarg zy$4iv^euS_eJe6s_Ye%(4Z>ZD?e0Bfe>m{DsFGz{HD@KQ;!fSl$a4CQsvv0lOPiSn zrYWX#&h6Prf}?^;*HNjD1^~k~hAV`Rb8=dNsGa}%R?au9a9LO`?j$uY%nEeF23!pk z1BLddaRut8b9E+qRd_f2tH%Pyc6E`WLZiq0i0Y6HqT0p1QP-o7M%K#PHS0X|CU~G7 zI?$YE^p(PO9_u`Eb(M6D;=;TW5lt$6rPQQAdJwbWjIQ9wTy)o15jY%;&pqY<*Nw-T=c=rj z%9`>Y|4CALtb$5xP?AeA5b@5UV$)%Ra#r_Pe$daPFox?3bo;#T`1Y#yk_wUjQ|ZH~ z9{F|w(olV@x;@*V*&#d5Rw=j2bkt2s!YaKcU$#efQ4o`5>Jrn=^VSGTO z6(|4t(cImnn49k5uVa&(vq4o2#Z*${2o+l(Fp6-aoC>5aXF}7!(Y;GsD$5E>kRX;) z0(X1w_Rb2!4K?nCvE1pZ^s4lr>~RRjNShV?p6S~2KnG(<6T~t2w57w%Rgx0eO=APc zLLe}!RpA+Si+E}`%|V`D4jlvH*j3Ay(~tV-pO>F?CsqvYcQc+LtGHo^$2+K88w_ox zm=ua6Q?a?e+ZBieaGTto4dyVEPy$}!B>kQzRahHW6_Pow?d3TkW7Kly{$SkaHSPA9 z<-B#bu;hwYN%u%o)!2D8UUEYkFG-z}AJ7zYR9CIIDp(yd;I=P%pD&K7cdwdzmp(SV zUH!SDG7N(Ux!|!&Rrji@re%UH(xb#tO))8f>tEXN(q$c1<5o)Bzg*()rQn8u9*=u2TZ zh>9;C7^fpnsF?7RE{~u7ZLKxOWS8X7b`wjP(BE=bs?U{FNgKQkn#6Ft@F|$X&eR^y z6T5>ra=P|--v8H|hySI^zU-Q%y7{>3{M3ACvI3;b@sbh}FS#Ezo^NpIFLCHi4hOAI%LW1GqqFJFT`>=7s2Qzviw z-pz&Mie`skY&A2+5yj7N!X7-+$M-i49pX^Y6)tNu=Fo_ji$EKB7oH6z>PE?2I-p~QJifE%OQ#0U} z#B_eWAZEbriRhN_xMx$$34%lh4kiKo^y>NDvyU#k1$_qh3Tp(nlua>xo)2du^JE>pYSK0N3c>B!36r?V)gg&IZ=2xQ zeNJ-av<&f=TekL%*S6X>1K^UH%iq6@u(S@iM%RnB`t%Bc^nVi8Men3DHJ=2ooHXFJ zVp0ORrpgtTLCYFEOD0DnH%lzXH+|AITix5{uAFoN%5FBl94|7d4u+B@rDL7hrNV7;Ow3i7GuoDe26C+VY8Zpu&6>n5M1Gv}MLs(kAeC&@nTY0~ap zujud?t$jQWKMYPkh@ruUX1)36u`FLaHhEZ0=Wm#@{CH-AMEx%psuY|`wqS%tSzOa?_ZLx6hz{?H1K)&8|X z2F;P6$MTY>xX8|kQcS;tEH=!#8+1$dstin3f8f5`C2g%n5};%F7`9rDG>HqYb+sP; zgT2n+tv8KD3t%h;IOleF6${#YuPo4S^+}vr0;!JN%pGUETnDA>jIZJ}2Pczh;;Ez` z%Is6OpWYzz*#2Ni&^D=7g&dG?5Foc$sydlT6V~YN`kn;6aFVVJLuRRC&~mj^-1NPA zNZqa0od$mH;m?$+TBLo_0%4la$dpqBnrvCS3V1W2IzJ;go*WhL7o>;Oi+6gR)H)A& z4kBup;&|pEC)7M^Fgv1V1smpm)l+GElo<+itnSk6((IP+hBR*HL~5V)Mn}Boi1(6LI-Y7%SArMmR^&NoBy`8W zR^H>=t-2{s^H?8NBWm!Dlf+3X!fyLDXiA(x^kT0o90AQUR~hq7`&q|+{@L=R#NSv! zvv~J$e{zS%xoZhF#q|9Y^BG0%Q?U(76z)+FE?TGt%^FficPsnV`pq*dL*9dw5lm7; zc99lohuY||G`s;CN#**QhhUC1tVFM-lM)rQPt;X`9>v)X!)yl4mk%^3F?$~;YSaLc z!%INefQwnB@QYJqbih2NtAB^HK`oLj9p=Iwq+iZr$x^`FZtQl2c@BqlkG*$CK&euy zdgr{fY3rC=Uzj?_oXv1pr@u=2y&ij)MCH?v`+|e85M@Qki&05GA@x_f8TKhFESVvf zNgudywko38aU&Fe*M!35U1_gc71!?Rf_tQz$BOH^&H7bOF=r@pii$l2TcUT{)bC)U)y z5xr8AxWc1D*d1&D(oIV$?NunJuA{dHeiq#b*9s-wPIUk|6*io~!Ttaz*i6_F+Wv`s zI~$fhFj%hKJEz}suj<0Qel?T=qaD((-XlF647LP)L|ZI$DYxmXg0Md=+Cr`Z&ooj{^aNZux-XT@s@x34-D+VfrG!(x93EOvUc^9CS+w zG>R(2J`8LMxa4z3WG#We4o$_YU6+T+4a83`GyDMYrr-U4VfIAwyj^}+sNoyDOQ&yz zLZLl$tN7sSH_dDf2cy%;8F8BD1!aXt`(k7nGiVM*ACPBzX3LP9EGeujv|fCI7$xnh z>*{0jdhvbXE*A!o#)uYIf#n%FR3rHRsQ9ht1}9kUU)_$fKanRdjQ=mw##&CKm^Bnx zW%2*z(`Ws&h5CVzIR7Q%#EE4mZ%5E$_sF=Me(iH(p{ErO<=_0?d9rrW2(aPrN(p2M z!)v26DP{`=?>e?ueKhi_V5R3-_rX9gmn!LxX4J}C>39iNi0PY^RiqFO^~wGf5zaME zpCL6zFkrjNJYzB^-eNbJG&+xI|7u0Yif!@J$c>TE!Q+^~6B~4VN--Z(eYm-B(ZUBOU5vBB*se zuiXed938@KzDI&Sm7AYP3d5QqgTe%)d@!k^NC^JKt0aHM4qYM9qwBC&m^}@>cv#q6 zHSJoMzTG!Ic!l@FsP%!%1yEak3!caMH+nsmKk(>{sv_NV0)YldaJh9<5- zQ>NaktkfP*>yzf!d*#y)p;9_Y076za&}Fi1zT+CjlZg~ZiDW`ch&TWc8H+5=wy|_nI1TN{!z@zfr-bOi+ z(;W?73Vh`sNVX_?=uC@K|1C?qA*k zgOdaoN$o=Ys>L@KHpSq{L+NArP2ZtM;;464EKYJbn8giRPEaTg0+JK7sBbQopQfz~ z-NPRTo+k0!f}lJ$kc>7<;6{pBPmy(0>_OFb2Dj%afrS!`LDdK@hM5%CaSR(nYGpK19^EtF{by7BT&u#!QvY)yIl$vM64b;Dnc_k#rN=@N8Q_LV3B>Oz3__2!B{#fU zciemQyK!#F7($5SpQFPCM`EdocYGGq+E+kw(J}muv_i|r;i&t9nB)C#W1FQySzdT6 z=@FFz&Hp*s6H&+9dTG7*u5XsCQaX<61%%82N27(rsGh5zx%OkNRm054`fD=Dn?!1D zUg<%K0a~VgRBT2t^5Q_0C~%K-xd6&ck$=?O1$>wMh!(l_L5;Vow~8;!1Cm?pd8lvu z{>_D*bhdENt&ns}k7;Y+H=9KlDF*(zUb-o)VLG<-EmAakZzS+Uc|@PWiTf1}Z``4u z>!@Du2oG^G8To7RZGGv1W-N6k{Qb)aOPvXFGGmJgAXGMr%IB3w*UG!;zZL;*cWL!V z>9P$dq|Cnu961ghbIjlx70gF~rL!-6Gs|;^eUC@px>#7cHtvIc>|=9R@&~Hx@{h&2 zbbDlyV4H6p{XhDeUpl^fbFLY~ttP4)_tV_QmDcttJJT^u4 z+pL>86a!>H1}fIbB!!jF+Zf(QnibjJJAIpC_A`}X>->y@UY|axC*OK3Rjb(Y^F-%yNmo)!0;dONn-T8wX z@(Pa&zuNQtgNt`Bda|(mm22Ow)b>l7Vphy(SLe{5kv-DfaU+%EjE_-EoZExHpE2Rd zSH0-SWbRUhTgbNh zPTf+$DcyUwfX?8l3#saAdJRbu-1RLKR(RBl(}K1r8ZLLjjSb(fm$oYmntj?_XqvbUA}|kBS4pe5U0oB64L>s!ry$OAQ(hyeiB1+B73cb5 z7`oIq!xbrU&H>19(K_dt9Uvow2&vA^4_+FJ%bv%^35&~q>#LX=*A(+)<|Ps*+2q$9 zlpEbj9{YXb(>uM~Gh3MGer3Tar>0*&Pp+Jh^t{QOc)y^~z4)uHP-Psv0QYb{H;@QQz-Mk-FL&j_0ZF%3y)o9EDeTo zb;W{9f+U$ia~ZlXK&%EiOxx907U2FD*;y*XI%$2yj|>{T0jy`_eq@1s(Cvv|Zs21$ z&k9ZOfZCA!DW66>P@R>dkJ*VH;xDDnNtv+X=*#(>b;ORM+q3@ss<+jR@{^ivm&hg_ zM+S>+93wj@2C@aVQL(M!emZACjSicDBuCZL{hr%=v0I%v&Vb$O+SR4g`+?uq+^;TP zvJ(`|DoGcVrnm$IHbNUUB5Y=H{s}uGMle(+XqL=OkXuor`s=HWWIZ=$8;>>4ZW~Nx zQ4CQ0rc<%m(^|#Nik|3t>Hc{JO(Qt#&4o<8>IS7{lRfB$L$C=<7M%cDkFJN383d!j zlk?H@^o;ZVPy)-5-7L;O)!cg7^mVHqdE@r^Omdi;9^vtZ zr^!Z-oT8XI3IgDQM=2q6i>e!x&J^mlDh z_mOVcORt%K05(X~;{Cp*sy=y!;sePk9rk~%q|bYw*RCa9bIxkdYqNFjf@PsN$KHiQ z>6Ckg_zhX}AIsZ(@eb6)`dybp>j=TS?<|e|?Py^)N-(jSH&5LdCCTbP_5H}L(+!$D=4dd| zSS%BKrrkQdn}+kqj=7IP5=hF>tZo(Kl^l7BuF<~+^5{_>^}xQeDJ;q9!4;>y@KL^4 zw#PWx4E+76<{o_QYf39_PJREi>trK0CB$Q6W}glGw^IxdA8w^$u{%}2*GGz8xk-$f zb9r=MRPF53|2Sx4YXB{;~F0NdAl@*)3@Jfs_cCkG}+SCV?PamJ$e?q;i>QYG*{hCo~P3yEQyz zntp#Ub~4)#LFLURZcEYG^~GAB$-d_8~ncKcidE_-AiS7;b)437;IlSq*@R4gtg$RM}e z^S0Ppn~~yF{xL^|%U1w}EcxqJk<}KRk^ZBTWaUUD%RII?b8G~kMlqWxvVn>{DaUja z#2MJaE|Zv?&I6;q}t4nc1?NKK(S>r|8qq zr}ubf(f5V;OscwbPMoA$+BT;SuJ;f~l7@_Flep9Ao;5DGdCILZ{DPB_#EzTW6{;l? zt3vr*$>cwgbv%ye6xt|{trU}vfHrd==jaalK~)yAFnvaPe47<i9 zB_mjt<>EccFtKrl!mwvudB_fh5dt|AfBEzO`o(-JJc5?3-9-*fA{T9>aFSwxk<{BeBowpRw-Rsfgq22oAGHIG5Jci-vrYzWJx^#$@IQL6eztU=SQ z+~ZRVO!;@EN8zkMxIeUn)XMd{B>U&Cp4=e5rR&WoFz8=2kYqZC`8LmV!Z8!nCSc6Uyr30kjK_6OD$lAqBzeA3ugD4S_so^-k?x&$OVbpS7`QAnD{zNbE9||Ef&;ovFU+q5CLq!!#B=0r zY1Q0;kvY1uc}9WBD@kCPJ5G`%Yl`WZo31;z@SVTp3Hybd6et&9IjShx#&Ln+`%jwH zV^KLg-bb;hoE+Wl*-Pr(cgS-f253@LktG(N-0=T$5-pe4J;GQ{$Qa>yS6!Nt8+_Q_ zwZhx9gJ!=JhChg^7R_{b{CRDvoPb*N6^7X3d}*+OPG(%hFDIFIc9STQg?0=r}A!)P2vm zXfJ4P=iJoZ%{PB`!^;Yyf4xTDBFWrxO?bQ&g6>qqXqvJq2F#jFDz+aiHpr0&wkM45pm->fKurYp2gmQ5bZ@p@6z>S#<#7!RE@DnxMu$EZe!#s8wk zo+y&b3dCcKbl6~|iDE8L4lgKL8>K;Aevhu-3Mk(5Vo z*DZ_0c{}EA@r#q(WBN#*E?c-V#t0piu z74@uJTrg(*!H1mQZEzPT*E{SSU=0npU zy+Lq5+o<_Oi8r!n=;#8xZG+m2#N3;q6KowEZGiGsGrw^qaLHxSdqu6nT<~#o67_Rl zzf+O$obTdto7=J#HTnK>dxBA@wHc=G?U*@J)p3!l)tPE|-MDWZFVZ60H{jsAw|!_F z&bYmM-a7oR|8d5EeYG-6neR1et2iEfkll2HxFn!alS@~T99SqD*mC_gLPnSnD+&Am#+#A=|-;AZ4416L=>!EGjB}5u119 z>M%VLi;T^yU%%JQaBSa+Su1Jb2efd}pt~lmSkSE4Asjt{ESn5QzRn#nBhTaPtL%u; zF7#>^TX)KfE^U6FteHf1+L#;}6tkHkDO7A~xQTurd>74%4>g&Z0XIwzSw90mPq>>D zD@Dchv5;e=I%K?FHoPQ`@7vk0m32P$_dQZ}2yab~m)s+w?VYr|U0zt+u3xa+uJu~s z1HC9)#U}cJyp!JG-YnJQsxxMn87m4#y1@ytBTe9yw7^NS zD+{XWYE`8aFSV=F!uJC+O25#+47xQ6Qq}u`{Q}b@k@`MPg6*3I-4HKGIlN0-1kIjs zk6tLzH>*zqa|-4!nP_Ct|4@_R(@O94>85YVjUM`Tb)2N?rFdv96$fqH&uFeeMKOM( zz6JP;Q^VV#WC%}@tQ*sK2i-mZgyPrXMsG<9lV3o%Z@7f0|6 zf#?~Z&cW6Clf^6S7aJCxo=+c+fQA@7~H`m8&>$bc0Wj`#{H9Nho3e1tTEV|KLoVkjNAe{iRt{*yT-MYH07c6n&C0umfFur#JkyC^lu;w9>(VjeP zes6)+ik_i9a9onvcoKUOD4J%SwCbwWL6G?qkx2So~Wp%|mWUCfG? zPbrod!>>sb!@+xt!9uL8#RTkjb;Z;^a>h%4Q;uFVbLdck!NX>C9gUAcx%Gy1tn2EKJh@7vjo`;niZHOy(FJbpPs5l z=WC6iUeV$4P?`csFy;}KTfORSzS+9L$Q~a^Q$u)MVS3e=i;(8u=b5cbWCm1t%W17$ zoD2m}F;7W#R2R6dSLatA`tC%l##z29^f(!G<2C_rqn^NRJv=zHo?_NfWHl9A;gJ$} z(SL;>xSeMdKs{)`r*qa6uEKHvKU}AD%J^TZCPw;O5n);|=LX5-alohCM#j4+29h3f zkvDd~uaQ|cqu1xL_W^k}RPz=H&X9b;dAbu4oo~NZ4-U0zx?5eXJ58Ub`v|svgXZYa zw#Oh$GLL-p^)zAW%#$uGNq0d0&oVolBI}VR;q>#TD&G{nGRKOR?7*Bfvgd`dcR#ej zRwcz8p~xXBHdkk&E!*T8x=(8Ed>zkR^uIq9JwWRN_3eUVA@2)K3iJY@DXbs5yF9#= z&i3q8mqc}mPI%|(s(|-ge^l2)K5*|=16xa`fBST71!T}XBm*j_Wu4J6w_++b<93Dx zGe)?vL2)Dmak}=&w(b^tK3U#cE-VbO$jNnoYP;s3HjCaTxaE6)wmEIBldcpeLdire z=~eY9`qc%~n}a~)?dp_y7&mSqx18Vm%l_SuR_ylkIQyKFmYJCN)@*@2pDeV)9+pv` zOEKFil10U~tFB8rJPU+KPJLclGTERx61_5VC%uszoQFB8O)*$A(=SXCo%AcAE5a%~ zHbQIiPSqol4Qv6fHU&m24ky`$tWxA84*r%*UDD|8eb2fYmH0e3-2pw|urW`Wum!pySmc#t)xW$?sj^;aKP<`yiA)R>5HlnohI*A(rxMi zx8$Idz{h^~bUWm59G3546jg?u(xp$wxjJL|JS{U-itl@#0?Q{3c=L^-^YjzZ5pC79 z>X6IKY32${d`Hk1RqqKzJPUvO<+tS0HPIP@>}i; z$$8FoR*c)-;0F0Mf1Z$O-=X8_7yysGcPxE4FrH;0m4r~|-N+5@`bNQpc^w`|r*Y5^ zT`C4mp=3#;pj}W0$(9X1{p!QP$g2Y6V@@sovXc&;jvhH9OyOBSWT{pyXTY|#WTlBeWPf%a};GH2dZi3>ie>CV)_;;mN1g*}z`74shV_SKb4YapWOa?_Z zQ?VEoMBiNzbR^sr(M>nSRKC_L#GWiCMA^`lV|BC(T(xRyJkTuY~0TQ{9RvYcz)XC;x%LnJ3m6s6x zNllFc{lU!&z=kEm3QzNCC?Uc8hOY z+P!l-B6rA-1ohB)!jod72+y-UP2vp24(O0~O0{f0-rX$QJGXDnR_&2RdBR>Gdg_Sm z)H%hD8bKms2EYh&I^zc|W~tHVOGHs9SI zDLRw*3D_d{BVEj@@qDPTTNEzC2~5lhD%9`(;V*~m!;Vk;bNL(TAS|KCquyJ{p!}G; zUW_z?w`YDv44U1fom5J@rRRZWqbcUB{D!=T-Yi(^(c$@6Y@~}kvG200R+?cFA02Rp zP2i$KcHnH!O8&Z^RiC_ZZo&nUF$ppNhU@V{iUAU*94hv$H2>o02KuBp6><+-#XTaV zsy`JP1&B;h}gJ|JhfrAhI4AKYc55U)|p6^dL!!?smy;oiW! z-hA39Gr^}hgZtQYqkF3QsHQXkB1af*>VzabqX?fsQkV1d>cx5_>wti>PlMPne?L?Y z6#MEi85r}tkw<=?N2ks_7fy~R@sj&dFg=bz(!6~7qJIg{C}aEpixP%f0$f=Dv<-LR z+i&@vr0e9F;WY{@*#qwjq=Gn*X2lU-UntM99y|}23mf*mGZq$H9^;FJ1$TugL-B!R zK!tUf8$>H3GXpck2F(^hxz-roHzx_|Pc{(!4zky02T2M1#N3VRL+PMfp(Mcr^i^DF z+Z@a#$C`RW;Qkqznok1tmD=r0gVF#*b;y|t9+`S&wP?_77ri>VimdXsw5`u`$MN7A zgbs~g48aWaPTkcR>1L|bCI*{>=K*V@0cu$sRes|Rt>QTb)SmsgDYpVveb+=`wAF_f~DXrs5?><>}z-eeN)aaq--loI7CZjaNK1 z)+H_G^qWUX_DIcFc`SV(`5%@*SWGc{D6)%+EuU3B3)%Q9Jnm1=_b3u%(WZzKq6&y3 z^(Z05VbTs^y%-sUx@TvG#z`6@)<7h*j=sQrN~Z#ib|KJ7;hp=_w+MUae);|B2F*c_ zY?pZQ97e{oRv+hr!aN)MhHBj#|M31-y%Y+HXeqA%q*ShjF{_@}wlFnl{FEIEFTV)C;2I!x5P_bFEt_ADW_n}LcljE3)jh%U3NU(-Km~ZbSx6yre)5``ix&Ept|e9}9^KEtHo;;jZ|UPnmMi zZHuC7PAA>RT$An$vOJI+h3$mmB!{30BJ>%6&(z6|Z%Q~xYD|c+Y7tIf=-^;Zh?($j z+r54N-l|=Ge1`fK`RIjl)!ehuMQs#wlOorKdR4#It2TPxnRD0woDvJ~i-acdU0W;e z^Jz@*t)}b6_2O19JtA+NcRTY)V8$5KRzqULkZYz%Q|vo5+MvNfdSu5u4Jj^<1bd|T z9f#z*B-@#kKy!Y~I&aL3$P8}@%L=XxtC*T0FsaUwI-75_&a*?_E_fJ)g>`YbJzpp4 zF(-&V3(n5K5h2l5byk&0D?1Qb`S_Z&)V{{$F&$ArZr~1i|_y#(u*Q z8IS#2xX-{%?;K3I+yA;%?<{`5_aDhA9$PiHY&6Lgin&CQMk;o<_rd6+P;XuCnXVgt z49)C4P72z5`#}k$c{a{%in%1{p;N=_#T8*aqM|4WWd{}pWkdE{d|-nHTbdmWUJ99= zgArLl{qy_ePiCG7=#A>9>lR{eOjFDUk}be)829ZP@*>DSt}rJvm(ME=+VHdM@=ldy zOq%}z?<)%)%NrJU2RFs+gtVL_!9}uXp?<~Us|y`BoE!nybCdv=+YY0Q6SyevohyHC zpDM$Wk%A>34gR-i=wrAwe5P=dWA+U1ifYd+K*CyKese zWo)GICK^3A)>^j%s=J_y5bF8#V^8D?($|v`=n~#TAA@Fh(66{I1SqXE=)`wiSBw&Y zjzE1R9Wu(VbNI23>eh$buSi@l!eier%lZ}1oK@VCTmG9uIy@U=>Rx^4$#*{g1(Y(S zQ&~T2Gjr>n6r50%s%rhvJ5wn=8I8x4QXJnDlRa&}piF&4i?#mMbe{r0cX}PtK$U`M zzo1RFk;F+VrFp_a;nCn)KfGHMbrLFXP7s{s1i3#Qs(+=MU*0RKR;}}Cgmv)_6z#m* zrZz?Fk{{7FxYx}p47?KJI8X0r!8H=#9J=Npy#DyftmMC~@&f1*pgtN2dEu~IN1lHS z4w`#X+>Jau%hG!wiQI~S7O$dBVU`|n1|5bmVZV<&Rv4>1oA+LbIo=2?)4?q)j=>n`P!oeO=rPxV$U(%6Fl^Srb4qus&0P1j z>c5cX-1h1`&ba|Li(!7l6pBfvND{IwB2#OTM~U>L>XzVr$zWu*@SGo{j;YT1jX}U8 zyjuq@ISLC0rgG|#n0L4RSYd^Q^q%y$WCM@4QV?|R*1ZD`+O!j%uRFf*#2#@0n{mqsiR0O6^kiW&5Dk>S@cJ~NHbLum8wng}G z0ehgW{h@%C8Q{iCRbz{!l)$0fElANy3B<)lj|0wVl4OaD5xUn@Of3&Gr(0R>Laikf z%2jyaw`9>DYEl9X8ce#v?5aUGOMVOxNI>e=fZKq>IWcyU9u@R#aQM6vyDxO`bdPx5 zx9x)rEU~{vL94iU;T}3m8wVbVe2D1Xid^YApvs3llioS~o>)tNT-YgY73T!CtM7>n zx+7pmUJE-8)kJ%RM)UZ4qB{ClqbCHqV0(jUbTZSkJ~U#zjFR=KzU5pvK(@0sJ@wR&F?R%4kkto$z(-4|GA+CFwu+)~+Wz zxGhOMHfC#VmZTDjf$EjrRBTm9Q9wND5Z;g*MV+u)!5t2;fq$rJ6=TPhp$!bKQP(^U zKtAxGTaPkBf%~IA=uw|VS4{0!zYo^0IUV?r`jEOuo*JG%qfp{}Q^+BVJ#Trz`%XdD zmy_HH_;BV|y}2iDcCj9MxzgZuOuNjhn|?B(Ieg3X#opK`xBg84Z5mn59Sy&NdnwbnmPM1w_~5MU5y)5Y(l=n%b;=YPj}RkuoH~4f4c0F7$Iqv zzoy==Pv%D0vr4g$H;O(aL#M}Bml{*}lnioYq)H+lnca=n$HoUK%-F84do}(d7O5)$sqaqA6i@(mc}5 z4l$YpUY1hJVGJUHUv5EP+dQ2TRGMV8yo&|y~Io^jl+EVw* zutrUz#&QFCEqT?0vpO%_@&{P6hnqC1Y9mt>(xToytq~d@SO6dAkw9VMD4yxD;O_s7 z2i7jlJ~UBi&?E%ir>jXq5aw}O?yrFQd34=jV}L=`dDCBk(ai!;TfVgjeQ{#ec_Rdx^z=w?lYOl&Sq9;B0Xzq^b+# z=;(Kq5@C}aUjcJ4aOYZknxUuF1&@W#LMBuohoswfHBz3MyT~>|e~BzQ4}w7%K8KJw zra=Sr9>gT>RE-xD9EQM z6xm3{ey%E%t@AL38+7OA^-J}e1rJn>8iVf9D?ql~D?~eEr&pzRyf6Tw;|2pK%laA9 zvBO{lvvk6y*Xa+QuQzyXX0Yguayr31NpO31r|x6%UFl(Sv5(~(3xewCjjuH=ye{t! z{(pdZ21DjQlrMkXJ}AW7YDAU*P_}owW>xqqsFB|2^;mA%S*{F(Qe{Dxg{&rF-NVKTq3R+?2@Bf@1!(oyf^&oRrAKj;-aH zKd`PZhyF3>4RV^t>&tDM^`)6&ATja+6XiniGoMDCA=|v~dm4OhtB|szO_wYxQ`HJ9h1-0uNpGu8sc+AQ zf)QC(uqA7yQG*NIeqXHNH)wiD)wCU64r;A2pvwkoj-NONLr&!8OYG?BeVql15C}q5rcP5LWMt4X!c24!#g9kX^j#i)8<<*ZH#GHxchq-Vos3$|DU}tfotka z_x4E7A$c)mBQZ$-ptV>9MT!dF^CZENNNf%yOmwFGDQ6F!_n+r|-e>v$ z>22h+Fou7aE)%x9cYL$+>ltq>p4X&qgY*3^+IhyS@cty~>fGf2d@l(O{r={Rt6@f{ z+WuvUX2Yt#2NSe(_a`C=-(ck_km@8o6Q{N#DN>DneQZiE|lqf8hBo=DGKeN zi$akLR zg0qUs2$Za-hB;v!gXi!#_BiA5OM)t3%7m-bTIEvi0d8H817Ts?@LQWBY#GRrUBk`^ z-Vy$8uVn$1(}KlHWmhUlMK;ae>VXlMw0R5n%=?0s1-se%Wmxmq3>;5sZ(W|7$T$!a zLr07~K8D`yxNUa$_{pQ!Bi}Rm#F-C3-~6@{FTvx@yxe|D@sJ|DRCEVz6bJ^HVPJp1 zKC_2l>5EnwH!YrXU~+kUyRXdSN+R72;=N2JWNIRrJ|^6A5_kl1Kao_Rs)8^&#Z9PJZi0$_P( zSuDwSVgS~gx$L_s#SV&;QPEx8Ym!>8k6ybSkU2TlW57REepH#sMSeTn$K%%wnvKS@ zRX#@L9K6PO=MK>#)no2%C{!NY-yFEd^`qA|xo(m^hW^*PcGenYZ@R2abM|dNVrRO3 zKIcteOU3=+LJC>epK3jwyZ80hbo>GX5=j zE8QaQ@?140ai9@G%5l@6>B&_IlT{*yiiHM?30mlE84H(bLKE{pf zo*(=p7ZXB?KmXNVNenxLIB}R*Yle^{N|8X3l~lBm7Dx~E5JmiZvSz66EaZ;`0-hR7 zTOje&O=ECi$<*yP+)a?E{OVh0$f}nJ$gvJ7FW5*a)>H87N5{Ri&FwJ+G0gi1c##biS(7p{(+VE4FZd0(}SL-i7+XGQP=*yjL;d&m`W7S2j?nJ;c0iq7)kt zrqRb_#$HTACuS3qr5c!Jq>$<6WKHgbW|XzkKB=Zb-0rp^s4-;0*qFEx47h`b+W2@@ z_HBIjkKXODq)aU}8-*-N0W6y-?4Gb;o~uap)h>?96cWFhQev5nrTb_Q)H1iLxd`6-^+Jd&}G8opg2JPNAV895)&z88lXRqE;TZ z-s_=~^CFW2B$d!Le%MtHQjAT?)Cmsx4vZA{Pv#!jaKv}uPg1WV3KL-5ANE`z>z=dT z$DL-d$fp!J6v+a88z|f^fWCg@Cd4J{4$mdrMSKkVCNdhV?Z#zpF-W8tPhgnD)Gan> zDjmDX9TgrN0R+=Ct=qw`?TP%va*e~ri*@2lTGue^fb^NZKoM1QcIPn_i6Eh2Qky1dr zbu$D-xUt+e*?mbvF8gRnNhzYks-#?>6N*S;K4S{50b2FUQ!{h*hO>6TH?X2c0KrFATgOBQ#_ol}9O{ zpD`0iZD2rwc3r2gis%jBJ3BF?YI0G;ZvSph8)U2>kR+(ox(F-lR$hkiVoSUT>Xbq)UO_6x;f1@;IObg_0bd09KHCqJQM;=$9H>KcTOmu_zjFf+W zuVqJ?(^fiJ=|0O*?ec1r)&*V^TogcRS5M!PVuS~Gp^v#y(&v%M)H3L#I%HKl_$r6~ zwB2GD{e_|TI&zzp#Tfr!#fG;m!SRBucC0|jNci6+su5#LQn|b_BQ*oZAfOGv2d=Tg$Y_Glh0DFQW&KIh_6>;GvQ}e5>5DX3c4x6f0F| z#+k(^c9#~qGf<;b_0K{^u;ht5d@I6{dE3CfaY+DLy=YR_l#b0CmSZfsqc0U2X?7LeqV#MK8>P-mC z9)audUahiMynULUcgPV-WcvVQ6iBdgt95`36Gx#w`Qz7j`+)FYzART|)UeX2k#lf+U=8o4EL#L!8JzeC zojO~z&mRGpL1zRH_@4{RbM1F&oAqE8ZdJKeaF))po0Id!s5&da&c6CrGRXv)?-c#w zS7g(0bCVMTW3L%7iYWzT3-hSxb%^(&&Y>VB8F zmzqPHL$m$y*g?awDz~_ozSu)L-jm865Qw;|_}Fo=nbCsc3wYUg%evKB>>2TiVk|%Cp0ktl8ND@96US}yEW^Nli#XXM!K0OjEv)Flq;dvajE?dl z5w6YAi_t2zp6BQ)w-u5@Tt_UjM+*;=?_ytUz|*Y5W0*;4{C(ksgA+^^=Ei?t`hcup zHz_%Ba6jM7_H3XO@Ksn#MXzHP@l)lcvLx_X(gSjUU8*_|$fQK4g9?Sn=-uXfI_SJ3 zTb-X8aaKrr>Tbs!@Tb&umc1HIYfh|mUu*?(gB%sCk+B1{0+usXJ{k54ldyRAHVrb1 z=h^F0cIF~>QTr-OHUu`2eJ6HRtZ*mvkba;9Lk8_NWH~H?8+ecXD_|R064p*A^lwsU zlXXlko#wqYY+oQA&E)Q%byl4Ylw|9e8v3-b3wlZ&;DF)A)oX;dO*j}+ywG0IsiujkAWzuS1OmLw{lP>;l8L~O!OJ!LjilKQh;w||kbgOZ zaOU{3r`Z4ef4+Iu)lz1GjoabGps*sqfF<|{iF~?;teYCk)f^C%0IkUh$Qu6xOpI-~ z7<+6KN2V`*-kbh%RS)UQ!bV|5cw$H`Hw!2ROXxhmPgF}lRDP?+PX8lxAGn(v6xilD zW+uV*0XMpm_W#Nc|3p@E3H9MOUi*fnyz&cTHlR^wNa1rDrC3XmBr3X7vI2r%u-x6^ zvsY$pVKg;pVfRiMQ#`g3?`&W+{D3>=Xprjm-j_{LFaKXx%_r+$n5fq-Ghl3`6uA^g zE=Q-zvnL##@G;1+Yu0;1t&$`cf~XGy5}7RbZh5zC65U5iqYnn{p}c>z*F1rmjCgaCa46(P5(dSuoI_x+sp>PnNoaA zkp^gopbI&rk#)QQ{|s6Oo)(6`ue5VybK|-lkInStMN46R|v~O}0(Cn4bj~_q)V$j|*B9=eUD6KDn*j zbQ>y+VV(~^cM10hebuu-{{CX>=v+24+bE%8hMOO~{-nU4I`~sFrQhY;l*6Qg)8*Nue(Z@E zz&o>W`*Xwg)j-^i^}^%z-nhL8wle}&a|);G)S1jtuu+-ZWT;!Xpjb2!Ywfg3q`uDf zC<#kY>~ecR*kH@ba61NaRz8fy;U5=xf7?=L%xS|XR+Pi`D|h(n>5Xn0RDn0tvw@fb z3?0x8MH6`m!WJ)$y@uslMa5oB{Zfy%gx&Y5#%dQNDSdnT}r z7Kax_I1W~$G!>(Vij^l=N6^pQ9(~g?Ey0BL0^=gz_upt*EqWCsm3 z_=iFdh5oa4Y5ev3_Y{^*qE2f;th9X^tH^-N-uIGVZv+UfgFr$tXb5GIHBeo)DKuHQ z(Yq?lp~4xKUs!h2i{;&p+h=7!#!2@?R7p+2kssD>K1bHRFu{>Bv*1WJr2vn1GZj65 zQPjyP^Fc71uT}1o?(|(8byTTSUy<*gUNOB*R_LD|kk6?SVrr-~@`~tUkk)fgaBp}? zSUP{P!{t%gVQO%0f}f6qli|r-)!YnAZdj)^Ay(L4(NyH9Zq9(_EUgmT{2C=qK%EuK z-z@=70?hhmh9WZo5b{Bb=zvSA9E+YeO-z`sr}Y7jc3Ve0@(vz*cGl#>zn5LHTWsGGHC0NTEnF6}>HNpGvEY<)=f7gJ$i7OK!-1 z)1*w|YVv91bS>oH;f>{pv$7Y112G0a+1ZP)Y~B8jr3b>sUN~|4j1_;SkyqiH3#0=v z;HT>8)Cro#x!t_f2@4i9N|U&H$h*af7EkH~>JzPUJAc7~1%>=h`FSy((ki><2?`B% z+c*qptWY(?Q61cL6u`Flx8*(Wtu$CjWT5&@=C(nOwM85^b;&;9 zI>t->Hff#Zq{7B@IC1vQ$`pj^S(vdeW;QT-pCfdc{Fn+ec!xOW1UDpiW;KO4%q|Q& zYAlS!u=hLa;K`e-elp(VZY)_ATu%mEoR%{mLKj2bjn$MQo+2wy7we8|s=Ul|2UMIP zGkoFL)vlTiiw$FJcKulGu(B27{ww@vb#5j^ysv5g16e*C4}cR_x#yXoWId%wrN|oO zQA4st1B+T+Aaa6hm3W{PsB6GW9i*QhL#to|Tn1eQhpt*JVQ3vLWj~qkYgyxLr8xK| z$(fTYJVP(zmNUivdO91p#(GJ*s$Nzz=Mj?~K8EiZ|9!T`Z~N9guUPCyij-t$T&;FdcskcCfTFf@ybl*BWgP^%uD)Xb5B_9wI(~5d zG$GwDOb#yDOxk?$cty;3@)(FZgO?~a`gPIARHsC!0da4Z4w|rPc;%j}_`8MGf$54R z3Y}~#oj+~4v_RTJG6K40WkPMyuE?S1Pb#l_>eL9o82%CZ8t2Ze4mCbGM`euPw=pSx zI`s+Qn8tRmJ_%5)fiy}r)C88$meE7J1vSb_CdPr%0iMOr$A49hDkX9|berCrRS5B$HPRC9Lk`~a zOBS=Sg)wvN?zne~CnSw#ne4<)v6W@=F`#Np5mrW4xnZm=ANqlHieC3pP9}p6@fpE} zF(s#AdW9XXOK6#CSD>G!oSxsJ4@ZS`fmIWkMDlaPNZPKcc7^$W_APE+r6?9d|J@?1* zyTBTDk)+C(2gPzb;Y;|CzB%Ivec1P$ShL#gILY?Nr(+zRBV#uhLrn3I7w3pgR)gWZ zNsgx0X^CMb7f0YaH=Ar{=i)f;zl@wRi@4NKiYkiig&wRBJ^g^By5)mHP6wSw8|w!& z$5clG48?+Jbw?!c=*;yiscGA}thR zxk?^=fI!Lp89|}2&tn6VFcA|43qSGfmQ{pxK;Q9U*M*}rVI%$~4Ea(Fzd5?^N9AMR zy_Th}R%%3ZRUNW@{#(SixU1Ycp(7=Y)9!v5+^}=ZZhx)cV=tXLRo)=TbkCf;bk=#n zfyr8Bn)tMXrKpVm_z}I_(Y-gD?q+3XoHrCR{_pySmc{aHQaw&=W~`Lf=c$^Bp6=pp z^G%(w5dM2Kx6l199ZT9dm65Ou#Rjf=QH}pm9<&y7v>?3$2S;fjz7$X+0{IyDINo8| z-xHZ8A=!s-G?CRW5#UcAQpAx#DS&!v1BjbZ-SUeBmyzA_Y>r;mCjTI2);X`;suuCR z@PU^LlqoV4XFW}~(kZfh`jSY;ar9d?y!Mg2+Ake1J6yoZemL*9Y~0pEE~Bv@PP_)T zVn2*+6PP;emh0&S#!MoJvaOumN_UY1lNYS=s8`(r?XD`Qvv+g{g$alw0J$9?{N<0R zR9C$D&n54<3}FQp43p)i1j{= zlI9S-IKvNDS_eVsSWoAuQspgjRPf30%Y(ZIT-qmSemQbH4Zq;A*HKop>-dVpJ)M(H z=v(rhcRa~px7K!E_nyGgdWg4HN-2sdQb0xD^-5;u8~#G$lq!$qUluLmH}Ddpl7)%P z84T>IZ-^@BlfgA45fTll^1EKy{@H+pJF+5izsr75VX1>~{^znz`BLsjd`FosyWndy zhRMpk%{O1k{k;h>@?1HeeD=a1Cc(_Q4N!_l6!{#Qz!r1kysytZLT572^&?Ic>eM~- z6;YEq^Cu{cQRa!7)+v*_<;CGy^cMOMr(D)0X^`B5gm{g(UslSw#^i(^;#7p+l_voq z4lWh&iEX|y!E5gA;^fea4G3NB7ca}9Hz}_HC$dg`Rn`}^m|SBv2kO-Eva7On!n2A} z+3x8%bBg@Zc}ri;0DZeY-*i`e#ya7r@)N>l?~H)`JiPZ1!B1VO=z0woD0V@hz0BI+ zbKv*NI(ah5QX9-^bK_QYz^X04C)>I^0^4K#MrgUmw663Ku|KD}OiYbK@;9NXwRBB%_R zCs>U?i;LeX3MsOz1ZNXCc4Av%rEc6{Bw}Wv_cv{cS(% zcxbtXe?bsoWyO!piOZwX+*9Satj~nbqfJwmhiYyxnhd&@X#&*+)5bj-4&r{l3}zXX zVdox=J6lu(E-Td!-RlCeF>R%~JQ&1Gm7gfVCCVWwekTM83RLk(<*bkoxMaC^L00r3 zy>h~aphMt5FOeJuEiw2+gRGcGCw(a-k2d~7)ifF`9Y~NR;%`$;ALJ|#UBtgFN#b6U z_mbEEz0Y2+_UTF7NAg4bUJ?`h$PhKC055Cs#?^kGdnF5(1ZOg*NrNO;rBi4YGfunW z|9yGf#@9@~&D%H6Y$SV~ID(`z^Hz>giX#*`L`7%N&7{i{8w~!|Eyt3MUU7GLHOx-H z$CJXrSP%rN(ldX;3o$4K%sHW{bE z>1m`tM_|m@Im*gkx?i!(d$BQPP8<-jlD3Cnb9V3y|{Z$2T&>ZhF1tG-EkYG z%63A@QJ;H(f1f+rm64!>{lJX?PwYQr1b3{f`j5n!gG(vDoHSD6#OCINnYpQ^6#FQ$ z2U?s=3uk0*F-Cv0I4dHMyG7F=L$iaGWyrOj5V0}p_RI=s(r$T z?DxMT#_fCI_uS#AXKX;KyoheVX~*k9#DL2+QovNXH48Q>SHhazK9Cs!B)-hhi~s;n z-@?ws#Qas&YsqZ-G!L#5d&ySVPSHySCjUZyAwMs4z$Knj0zGp$KiOySpMLGukJtzl z>kn|y8S6mlJo!7p8z!r=Z3=Uj)HtzKxol=t8Ysm{iX5Y&k^Z=U7MhFI(63NKH}IOh zKb3dMJHf$x;CnY{-Sj4PI!W`s_F7xiS#_^i&#RbF8mv>7D;7ad!aiAs@Bp+u_Y4$%I{%+*wLTdBEtPEx>!cT8^5GyX zt5fAWJXd>dW{@TolI$xW^9@~kuI;lcyr95`*CxL!>X2Zt2Py7!%2gf%E=ZKBRn~{^ zgIqfX4h_-e3`1-gX*coln!z*a63r@)N;n5)wW{MJpYG;t3mb4jO4c(W=~FbfxJYSr zCgg4~EYpj{8%U*4?~~$(wQ_y#ASvOsG^EN67j4jOBh)I30xRLyz+V6x2zUB7hb=fM z&EjB_)~RnAl*K{2Ub9i&H6n9gkfjFDU^rML26i@2L}mR=&!$ut-Ox+U$3(8B7=v6%+L0F4w0BN<&b_gHs&G<(yh2boMs zd1qNH$!BLuoOp>XRn-|I^#6*kQDhpl9xf;`APZ;sq+l^O-Z3} zn|Bg7{#%&o+~oR+H)^OEQWzDpA4qynQJlSP~n`d1+jTifFhzU6>QoFX3ZS0Wa#IEsiGvriJioF!sO+}xfGr5_esqzo6 z%gg58fD8h-JV<(n`S9fMMf_CxKp-5-n%v4ep-i8MasN(wCn5!f8wYtT0L33SRAXlw1nY zU=>XVjl1A{C zWp3iU(0eHq)-v(Z^r(9QOhhS69W5mwdWL%-z?-)3hqMq#nLyk@I&A&j z591Oz?;7)apT6f}$vgjoEN--Bcu%4f2^3jLMPnu2Evd26=8-$*c*Y>0I~?e2LH2O> zIr`X&@WPc-O!H39+mfYZixZbW)R+x&Ii-MDe=!xUCoS>@AQTz=sZ(c@qa^9I$D;h# zo7DNQ7ruUXE_PfRk712fdenB0)nEXdLRQN&0*=ss{BU92T%CF#EW=;VtD3wi@X|LM z=B9G40288p|FZ=K7SqS@yY~IHSdzrPBTahAl56kj9Fg;UtO2if_Q^%*g(Uh?n1Hd&O{ zx7B--PhOY>@p7|1-9Ae3fFk#(=v-Ahn2M|ZdY_%DTu2es27=TpCK~I#>%F0QNme1K z56`97z4kc-TxtWiKzkEAy$K2cH8BQFo{dVwPBEv8 zPMy-AJk7c4ufZK;*WWFJK8189ANnTpeUrH8ROkC@YC&7D-(?y90O@zp`}9JM*Kty* zN)l@FeX%DiIlO_s=kByc4!JYKvMsB~pDW8`rB=K?=(JaxqBlGvYNZSbB9?&1j=ua3 zS+8g-y&(vz#Ww_1PRvrJ_!)(&$HqpnGf#%8%+7{r_{Ui{PbbvAInO&0e(G17{`2x& zty1e(J}n>kr&Ihu;W3v!9JVGj^{ZNCwrHRK9+4(9G*^}D*=^t%DUqF#p5|R-YNu2= zyrf|Sm>BeNh968j9kq^$VHR=YemUhM#&WInTv-w;E26`1UJ?E6x*W;Iu~4{>vxL!( z4K19-FxmtYi+jdm*E&qN-OD}^m^_;m8?U}g64-e*P8_`}GFw?~rWENESw}@@b5i9o z!5jUSheGsjZ9rYnMacawmtLP46I=+q5$j$(8CJz@14X}*i5KRK*~;1`ykzcuC^@(+-Xq!_ zd{2CJawQf2wTJT-!RmY9t-RNb+h^5PZb^|3Ia$I=YQLyWc3+Yd*(PgJAOElJACTI6DD$Y2xK_S}*9kocPmaD?ZKC(wwd7KN>+;QEn+~KQ1{)dY2EVA1Z>cmGdeC%KB zNCuq!r?9`n`usDToAU4N(ycF<;Gx#Pry+~j*&HWMKWCVMDVb6vQe+i?30kK6Wd`Dq zGNwgw-Lr>apVOFN#(XqwLy-ABW3d<9ylSoGJmj=FW-BvGCWFLhI^ZC?H>H+2N;lH` z0^_`kIqRlYx#1Dqs)6$;&~r&}b4Er$PxvC{kyNW}mLBK!aq7c&k|jWThH^JKsvW>r z-=@4B(8wGP>z3ow@EOVB1^!sPmP=n2Wdzg*-GL|NLemHCX_ZHWU9h5B!t99H9GLA< z0^KM5@_6nDU}hUkY@9%>UdA3+tPBnHzwRfSEZ5C!az{?={92Kl)6>R8G)_&xhj!Wj zwio=setk{p#716rVC>W*5bbN_#6%sH4Y=U{`83vYBR>)TKLV`SK8mA3mF=Ix{>n!& z9GSOiLekHfd4+6rVt=L5Y>FwO6p$Fm#jN2<($75_u~Sy=nL(dWw#g6s9`@8K(P-#- z=aB4BR16vO6UuXeXQYdR@6ww+V}MZ+s}pqU4%uzVec2`NYn;6*$3b8;;}1rJ{lMCI z2Ro~B_};9Emfc29J5^cfEK2h(^iKnt*2kjb3Y@a)gHl7W#`?18Q(l_)PU(d?k3>i4 zgY!BAnjkT`OgKT03qrn&DB&7Z$3qD@xm zo*xL^^uY}>tkriUVn&3NL0`mSIDz6mHZBJB>-g;RmTTpwW42Dbf3Dt+eX6>^UiUSAxyTQbPa}C=0WgKi*Q4x7i@G>;57HjX&PZcd5weXsFdllO7>JGcEv+e;+YtPbQJrMOFx+f?-X z3)hC!i;}}HK`UG}T_DoLaq9vP2ucJ^P;$12ujw)9G!=)p$rC2VM8R6JE27WiYG@IE z9kdgp+(Z7fZjOoev758Q7Y8z*9+KA0u)S1I>f-KF=zZ3DHNdi}m#hoXsk^va)QMxFhBzWQ0!P>_%V)N33=8<32_nl^ zBu0|X=d5L9m6;iROey*(@_>p?m8VWPV&Kx<=%zu1pM5GcRM<3v@;keQd)#%hog$5a zHVC8xQ{{;vdsUtECTcUpXrPxqgYNUdCS|0~tcH^BT5c|#Eox(0=_MY8;6xvoTnmM7 z;I5!0vkl1M?!@0U$S;W2d_IEnSncj6|-1+-6BACFd0hfgSY$xEK4$-7A988 zFZTtOagd+&=JYnkKpzIu#{6zspA?+93IeTbdK#%XTEuZvOM{>Axi}IV#>s%Ki#=AJ zHY+D=eCyW*7kx|1u~v}9IXOX>5X?f3TGok*I-!CVMzwT zl(toj2`|6>sOKnI?8E_KprRXsfVGq&i6RM9bc>+eCi!n@cFesp zcZ{6cq0XeO<2>}9vDjv1MX0t0ag`+zgB1pYEAm~kDsC5j~kQX-z7t}E22km#%HsSPJIRVDmMf*@R0i^!>>eeCSXuG6h-?S_2nDUvg3=d zp2NO;EJN~8{zcsW5dH42qkr?qcYZbh52B@%VlhQx9G8H!5t;XAz2T)c84-Qxq;#@x zxYd&r=L*i78J6Re;wVMxsc4i!&Ig7l6hqeN#m!;pwQ6$dr6iWB^67Km4ka=gP$MOO z$Gt+~K9y0K%h)5kgt_FVx$cSWNzf-LT+rZ}Eh_QbDDMeKmk#Dy*BpmfrDSmUC)Ew(P9nD_OW zI%q>F6c(tGC+_gYH8fI)U~+4~t96)UR8QhdO}^|8 zYd4=GYlrh?oj4^_X69pLQ;JNAY^I_={u(Zj%b8g2aY2jX9QVHLFr6E^n4c{wBwEgC z=qTx&S><*ixDZ+y6q3WfkG%4ko*9nEjZG+d+9|cC-5B;+)={$bNAqGV+XAdq z-|Ltz&r(t^Zjkf>!7Ij}GKBrI?r@#Dl+){elD8uA5G)JYX^c(Y(rtEx?wZ zD17Wy!!uOdL#=!lT?*s@^}tDxOYau;xHcyp7S9qTCz6&+_#!qVn zo`V?ZC0j(@moEv7tI>g-?un3mI z%iT2PObmaI53YDy>A@CQWW=~1-YeF$LfgmzrYAykLsGzCPESuiAVvHhvc&%?Z+TG8 zoH|hhPjgR}59JkCd3ZGte8Axr1D!XXG`0-j-}s2kP<%j(c!~Q>I8)8Z49yJ1ZLRXW zpp4OA2}cH0ldM#4RAPP-X?rzz=DQs%^wU}%{g5(%!Q6x9^jM@1i+ zyM;E;XF{zG4&knO#zH-;0J=9NRi5V20QqdKvP-g7YM?(~o6ccg##=P>;zzY>5uR&@+vuZ711b=%Yo8=lPj1 zR6OwGT_n?qF$9nqf}!n{0uq*mRPSf&f?&%{HmEX9XFmv-9lvf38Ae4ERD_4f=t({6yf4GOBYWGOPU>$9$1yGl~l!kjzu6A-S`Dz8I04{DS6O-f|&)^y8sK?gh`nw!LJlyoZF98M$j z#ZcHh-h&>+7q>cei{X!_UlRamg?ytNxgYfO!i5XFc-wqaCk(h?Q)Q-WV^|-ZGpAOi z+2z&ek;FAdcc0(4qp^ezV3-qZIC+f_l@c^bD-~jqxIb@r6E`5OD=XJ|U!Mond zUppan!Z5QTYhVrn)UH3D#<=iTlLDT1w-}3KkeM1v?E(*Z)dlI)eeOqPB@_3%UT126 zYzD;zu7~xzWOBFr_q+5G{JV>LiSsG1T8Xp)NN$B!wvk+VJ=vfz=B13?yQLhA5hXIU zzDKYtPf+D{?6s>Cmj@-;rJ&d$sCasiSs!6#uBhLB_Wp+FeKJEEfc^UVum*626NOhj z3;a7}H)bJ+;zbBdbxp~j3;bI>8pn_TjU_r*kMc-wV26&2CwfIn(`2ST{?2}q#cncl z-hdW4U^bbRQ3~i+D59bpcy~OtoEZKAutc}GhouFovyy(7HEze5`w{r_u&j~J2uK8O zfQoQrThwxn(nmxOyMA-0i z(UYgyZ{vI4KY!QKqF8BX>)?IL(~J8!`{yqFTOX(2rJh?Z`(WYUH+#NU6j4K`Fi@gT z-xKR1Dyhr-mEv|s%Muvg!H8a~{obC{C$anDI&D2xd{9>VKNpb|<~NNxN*Xi|wNUMw z;&*pUuxtS)+n{Z6(-`cA6*(w+ckF!b?nWfKPsJ{{NoFA-@EtC zYje-54}R_UMka|}ZV!mvXA|3iG5icN^r*eDV%_E?r_~;`lqPW6e3TU#f_(Zj!76z! zT^?+Z>Z?)JOHX;12j75H?0^gKIMb-WmO*C(+kL#PjuT5DSbdJYPgn=SFj>@b<7Z#L zI+{7gi7ky4vc!|Yo$|jGQy@O9Q*TjpLUO&BDxQ0N>_`Bf47iO5dUE|(ZTztG=(RUY zE17Lmn7gEg-AcxZ!*Z9+Rx%Bg;v_|mQPJJ<1VsTPkg*H(R`ABqHl{&lkfg@mBMgM$ ziB{h2nU_NF6sDxH6SfdC%D9ckV)+;}gCJW#lc)mnmd3kwd2LV~SKu|hPIx6}g)Vam-b;?U1U=3# zU_dL+`22gkGtvX3!Ik~HAABhe9)rP;Vf~ZU#NfPL9+9CnH!UmsoYtMOQpJZ#jToIb z@xKXkIARI92%e5={rhbdAbM z11`q&l4grSr`+YWm1&e1@=TZ>!i(2O#>k+YxO*m~cK98pYPiQ{5@Kw@$xRy@7LwSxRLD4P8>b}ys3qRDL zC~c~|c<#dY6QGzCa0dXu6A8lNsb#{o0i83u<#@x|kbF8%b$~#{UP!zbW;)sTzCF`q zjNDv3<@h}FbJ(6_Wv<3$MDG8drwK7lKi{>0YU53Ay!hns>M8v@j0p#507rujj3gY^9O@C0~4;#PIQMM@}N6 z8WOhenZ@<2hQxUnwuLnhZdwuqv*84F;^mnYI$_k@=@oDBIsV#4uQT**V4_D6TPQ&e zisfo{%I?q)XRZKe9*P)3z;CMe-3_((waiG?Pb|=6xO9Y2XZta%uRl}qZ(b8j_T9X41G@rZBm_)5;^NE0^P?;lMiyI&$_&0gveN^yuHwN!MCcp0|? z3ipjvV@*o*0JlOBb%MfJYK{C|m!RB8PoGrYb#G+Og4bB(6T^pvj8<6`SS&W`+UyM| zuA$ZCm zc=mx24**LJXXmxoJu8;PaZN~?w&D9n$x(BDrYV;<-y0qTIFr;oeE8M;L^|r*$&@@ZSqc#%0ZW9G`5W8 zAleQjv~nyTK8nke3O5D@n4odw^);0w$BAu5o!MaTpcD|#E1{x~l1KBd&TSEQgdKJr zaM?d8enKsS0sXkC$Hay>bDxJk?1P2-=H4J{$a2PbfhIZpA$<}eGQ~`TqZty#c%%%&eQhBe$Zisobjv5-<&&|N(d+RS*%n-7_v_YG+d<_*l1eB{|FTfSPy{{MSk_0 z-)a8%&aZ#|hyQuPh-ZNg`$5CTC*{2#`;KK8#c8dHmB5J3t(Lw^UlBE_b#FDyy&^*5 z<_*d`RlB=RT_)Sj*$CysdPgS}&EB_-qiq~~vpXHT!^&PzvD*vY9!LG1+aMBS{L_g*h3Xzx>`qnnRQ4!uev6YuW>%$h-i2G%jvRrtL z-16Av)j=<8HPqO|P-~?NFA@y;oQcebETkQ0Bg6IZ@!aPA@LZuDv_bRf)9`nhteokdGI%V0;#pX%6fr&a zuiqbbyjUP)aC#W_Ircut3L(_L-#+_)pLgP^Tp&S1hP#|Odqh`2R;EUEgvP$*MCOj? z4%s2ERb(v$+iuOw<}~n5^Yh(m8T`B7<%XyNdfRu%RuSXTQ=-%S!J~LDhWKOnwG963 zcflv3c*U;Z9(kVM#h@-y%MAa-2HC}+QZDu_8t*zSDdu(xR>AeBM0MN)lkd>R4YIr36i%KB<2rRgU(9hMK;Iw=9r5!Ch}og< zjqfh-rcHs}o*(!fC$UZ(*ab4hAyND_lwvhS;;Cp{46hAoovEo0->0ewH;Iel()f$f zdh&W4yJ0IR9Q#{5{d88K81JSJUt(FS@^rhJ^JX}Sm5P)mWh0Z!ZDZo50@Gb7=aK;P z*68ONSmBY2PA}_}XwC?-Ir(({v=qMtMNDwDz`l|Frw0d)*oqrV%Q5W3?0C>RZt}jG zUG}`weQyff6I4mvm*mW8lU)IgwauPOxu;dv0Txwm*9E=q%Vu4eQ$pwYWdzoW$IO(l zg@PSN`Mc1+mbCw0% z^y=~)GZ+jV3tJ#D^!l;dW(A4yi{ARt+uty`8socv`6;PqS6<-6)os_!0Cbj8oT12R zD!PxeQ`BW}gk$*oBDBi7h^ml&D3*FW{T%6cS?_frxZkBTGR?c+<+Kpb?B`{G!+KnB z-@V^umEWg983B5pPQBW%eRdbfBJY&GuQ@WeGhlJ>-!%2FcS1cS(h5PjFf`C2`(dUc z58T>JMQ0#BujjRXms+oFld9a(e9B>|c2t?k%@y92Z}dxKvPJpg0{@GOOjd<;wxG~* zej4#U3kbh{|FsXoO}P5u?xIXm@tl!{H=3d6Af>3GNEH>0DdxDTbyG0Uqg6hVY@3J~ z!xXINpo=F-&r465=RbJrKP098cy^(VeR#>27_)D!@EXn$7 zSHEpS(al-4pOKH8cp=nbhM4n|qM0HeQ_+vVeGyVc*cM$n_dZaIod%uZeix)(0J{K{ zwXM7~?>)X{0p|tricLV@gY3`SMEjU{ZXY0JQ{c79i}{sQlWeYjL z-oFlYb2y#t5#0@r;a3X=RSZUNI);7U!yROYjp3f)m}Bg)(G~E=-#_o`s8(QYzc1^)WpVFwSe2#!I+Q%_|DkMU(2 z-uU`I8$39h!pc{Bp7c8)IBjQuTV}5`Qh|KSF?A~jCY!z6R zJv~@W}WbnE8hZM`bJ(kbfE~On@;NX%imLECVoTuKG_4vdXRvAoUU`__^24SDHBOYSG9olPM8F(fhk zgp{g>7}u;A2|2{6<~$OW%kFWmOy4FwCb=?Qr!EP)A{yNLIHDbn?20%?+GL+kzsYZ< z6GQN>Gpb%`F(*f=Q#UbRoNZwGSSv(<-QE>VrB!j~n#(m|A8% zSv7C07LTTvWBdCb&D&&uFRSEun2Rpca<@NM4!6r}j#&>U(lLB^mLg&RhIbt?XRZ@yw6se%1 zPlDcP7u3{m46UIz2i62F3r$dDhQ{){B0A{>3y#3Kzi*5}o%)Nw2Gpe74duBVs*O>} z;rKT)Z`U#nuF#t5{|M+6t3zsnF3i~s>|Z^kT7Y0|lqM+31Iq(DC3b_PjYzY3HX0?s zp4`47ec$0g6LzL$e3(q~oOqQ3Y#l?qmR*zrI#9PGRTPH3dfkz1&_t`gKM30d`(=Av zH3$QUjtFA;yCYMeCf`t2ooP)aonokg!^QJu|LYN&LjE0BOc6u67`iwRE4eD2 z!fr^g)#%s?g(pKf+6{I_r!zBb{BKRb`Ca3GO(UN=F>qFn9}1jDl;U%W^dNAGMEe3$ z<>>);>CK+qoUF+QW^W2bdfWFGHcHNe6q=sS$ct>d5PKm4zF#=oDC@B=Sa=J(g2b+vE)xb5U_o?O}!^|To z#(a@S1_|tp#F>bC{IGmltHLMYA1HV59B|2++{$|~xCh6r9PsU#VuD}K+mfYZ%S!~R z0EU={a!LVJjm1=So(d`c@ExsuxeD0R^FY$hFmBjnyu%kMEwVr=3fZ@SVlGgVqtc5H z5Tgj|Bk2O2a*_Ik&w@&L&mFP`*=gVb)MRri!rKg2+c(f^&sk!$=3wak>`a5xDvhKs z+gtg(*VtupDJ^ZXBG}X6*Qb-`meS$?XBpy`+1LMU{D!02p_X1*iFKd-o!yb|aE#g%9VOPDlT5ui0 zmiBe0hFGLgLu<&h?74aR%B%lv$%xE`<<^O_Ojd}UTY0Ud22{jr=w;j*pk8nDECZ@d zojOmI#lbUcCtL_>QmeKG%!B)$D@V*E2LO9cUE1?TmwpIhu9y}z>W$l_5_E; zv+ng)d0I+BSW#X$Cc6ky?FWESI{n{O4IuIR5ku-Z*X@^BdGgOTZP@zd1CL(!L#d`B@*L3eIB>G&ImrAac@R7=ayFn zeNAP&X78jpufab^mH-NV?l*Z$m`iTCkhE!(wDK@pgip>@HPC(TSAgcFDx`o(P+SAS zkNioX_op05A^%8FGQ3%Q=Fw*XV%T%6jN!N|?*uNe^k|>1w|xHm)~oUp!MG$l$kB9j z;=EeKOQJL=l6l<|bJxCg=rev<~D08 zx%!;7xc8e)O?N0oCq+J?qVLd`KsQFGUaF4sn%^QnC2Hlxfm&>@bWfB{jdKZ>4;W7F zpLLPh5O6PCr~Ydmr&QQJyGYa~KQwo7aQWOk&L!Es@T;P01S@i5_)FazeJ?@;t|J&3 z`;)lWM|FYtE1e<_YKh&b_m)JaVE5C@rs?dyk~BvnEcw&Mr5@dk&?5LUApS63UN|O%oHQ zgZG1#YfVb63Wzg^1}Bx~5WP5-U&tSDL3z()pp|S0NoF*=<+!n1zTC~xwH6~syF(Jv zPe8wbZdF;9!LbSUJ26bGRP&`X9dr_R1IeJbxj```R$b=P*JYp%ja7b#m=ntMi79?a z>$Ajr8B}u`D*n*vNSElvTcPaoGotq}G{_$LS&m3gH1$k1UBF~=Q~chy>$Efq&^%=} zpNzO2HZE!ZD*V04w#akkeDawS$7mAFN_Ph+#UqM*PDM9LQssMPDl12$ypj=h1Cq7`aQj-~40Io28Pt8Ao$%Z0m@7UrSM`cQQ@sN<3W#Y*>N%~0e&#+e^B17L^ms?CpJH2}^VD|;ob_`8p4EFk8r0Pp*cw@aUQVE{edA~sA0 zSfu!E;_7*6+%4et76UJ8q3>l+sO|LxS6R3-a8EF-*$wgJ6_QgV3B=GfRcnm*17IU*ZRX_c$}KKH^?S!4}3y{%w1cTZZ$Xi#-{-D?%Wnz*S; z;CE;d54h-nL2o;AQIHe1@hMYjABb8)c0~JTIN!8W|G+Yz$;QWVV$fNs(7Q^PMKp&l z^6rzDxvucu@^YDN2ni2oN$g%KZj&z3Qk#WhY#?_i}|=5j`Q*jTgPBChk%^vtM?#9FI1dE1mLeh%?%tPeLE99e&g z{m)EY_~5*+$&Y!gIKF~to!Ir=Z|3@zPzo?5Td|7>+p#f@2hpt` z4bpv_BMma7`2b48YWkYIedcyz5QfIUn$F9xXvWL1n4(G9r2H)Gk-H=M)Yw4a8NPzu zF~G`{jE`ywnPtg``hxf%R+v(2=#@lI?*uni4;5DTW@SRtnN~Gc6VTA%wh2K)Zy2NP z-{dY{XW3u$be)~^=1hr|W~i-nxhICup$eVeCfe{HiSxEYqaq{=X2tM}!z-A?kWT~a z!*9;G8m3jHdEbcWP;C>nyXU{L-O*JFFs?RX;b}ACa70+QCIiaXVn;);?Zhd^EoKu_ z3Z+Po)OwfAPdzV0daP1q{ARm2RP#w|)W!kx_;_yCB zJLjxH1H1;rfs2_m51pc9VjMAsH%j=Q35uzXVC0F&10~Q3&%4}XVtkSYi|r1OCS@jb z9v+FjR>f3uIR3=w2-c3@o_UvRxD-o;5t=RH72bz@iYe@(#`GmBOtYNUMrehewo6}n z#0)cEy3_Ds;QP}}pqh8BAf6NsSH|hY7UifJboNpTXo}uJMPt-zQ&7AFsur;1ABT5g z6Y#j;A_r2Om61p#q6IRuThc%-r+Yd$V}kLnDoEDC8;lr zMFQTtAxu}flmce1j6oiT7FZ443x^P?e}}9x(qK{QNKFuy!3?;xflLINC!^umuj+!{ zdXya=1G<`#Bj`&YFv`;f-#)U&a_!7!DdfcF$I41OpFZrWsf^UgTEw+eG#i!W!Jh{1 z3S1guFh8;U@__0H4Q^kTmXcm_n$(N;()Yq)V#8h~$W4K28uld}1`5p#dOL%*#u0It zkpsqT8X6@$p1!@sH6Yz`?d-H+S}Utz9J7k>V^d%TLE_8Jfr(r^#L{ng`rWWUV%sRQ zf9`1Sus*R2Q%F2+&O2}a$a1oIdhP7QnHwup3N|9Oh%w)$F$xN4m8(4JRqgUR?i!Cr za0>hrtgWmItm0~P%voTq>~rt+(WFb#p)jr%8b=%nDci7PW(~gB87)Abaldu!OwO1s z`{(Ks`9xmj-X>cSb&lzxD`6{VPMNTezAUVuyST-W`t9Q8`?kqy6t!MW>fRr2|7qO3 zkAHMl-Ao*b8`IE@3_i~ocRQWhOuhMS%L(QM@kp#pM_AUMD#wNH+8<$XwvfNvjrA(l zE^pr^OxWd<*ga(LPlDgJYzT2$r^8B%2i8twS8E}^Nx5+0LabNGvEFa4DPZ~Pw zfql1RVkWmh_1A@esR+jvE^;eWgf}VkVf}KzfmXq$f&XImu-sxdnk=riK9>yVhy1+h z|LypVY0e(V5w7%3fpLL&P(QK{nB}<@;khQi4gSbl3WwC%QcA%{yw*)*92XjnvR*v zq`9G4?tS!%3B4p;yz=EG-kYVFZnq^jXE<#aKtCt&Tb619Y$h%z_Ky#lgU=@D?#lZob9TL-&;gO5eO=yzzMR^dxR0u*DQp zgKl-E5R|ByVrrjiTNqZnZF5WF+IrDXCMR5EL^gxYh|N>8`KEy|L82PyYYe`x3aO(sOT*ctY}G z$VMes3hc!Igs#0`*nVWlaq7aVBY_E-sgRm|G#~1 zrlvgP=Dh6JKk@1c(5c(!9-Z41RTH$2e+uHcgNI7vt7Buy%Pk4O9svQo}zy=At>pl1!y7@Au)>yr3@H?|%c z`lFGFK1nrkxZsj`hGoJ52w)qM6EY?@UQTTMwrQn_`|Z*Di%AcMG zmWI>tNS6@IwA`9gtSFI`xefZC^)8X15&ux5)8O4_MLBMjyhOocP~7a4w=18Fb~`l3 zt;j%eSO#>g3@0V^o@mDt zobcuETS^Ex?efJU3V;d3bK(<KAi77>GuI7k(uec~; zBWPM>(Xgx}79@!qWwLl6EDocaEsyHB8F*r+*A{Yr`#^OyQ)uBO7MFZ?ZU- zzqfoYx#z^S-y6(=II)wNC6mbz6@5ybAc*_=HK|dF4MoahMCaudbgky9WYdiE@_rH{ zLb|CcX}xgMjB5T-dL@KqbecX0rfd(s>VMTMJ-8_fi#*#5N@!RUo+!BLKcs9EZie8Z zPK~2v(VHbPq8L#Qy%xC0Dqt_QS&~DiOIEpE1ST>hJv%Q?4=!`N7m*>o<((d!t+_R) zK(R5zXsPPmD+T)_ulg5zZ45yhm8r>gEl{Kf7fJdi$44wgoSp|`1*islO$D$vd3LC{?+(tyJ)A4=5T@`bwsxKQ_JdMr?puumBJYD zOEys5#q&NQ_uONGAA=W%f;ms=JBs~Nm{@s(L0=Cm0GEnxiL0Sm(GIK8=Uaxc8mrGA zb^PygLdm2*W&ZA+7hQs_NDN%ifrY_Z;HXl!sXtMlB&UNNKL1u`$|guyxzEw3xItp$ z`o#|{A@Py~jdf-oa{|SzqsVG1`jGUt(pdFbp{V2m^#^Dl{lCDEP2C5M28*@vAD{W} zj(@BkgWZAovS!)g*9&O$ro$LO` z(%E2HrSUmSlOQcA6cYrj3)6$o@6d;d?cZ(7cgc3IBhPbk^PJcAF88Q?b)jihY5ZYq z7CGd^@so>YPya_0Q%jK#spvB1o(xDNgX)DkRlRKal=al9UkrYy?%OTjTrs7BzA__c zew?To;uy^sp?&=tQ0(YbjUoAh6A{r zM0C}EH}%IAQ)+_hpecWe?+K5lDA)wS=T*p3pi^nzd@Nr$@7nD;)+fXkkXuFv_p>74botSybmb`2ztw?=$asm>^BJiS^-(AcIC=f zvcr3zT%lKSKlC@5)Nc|5_!Hr_k5ohsxW|Z^qLA=kryBHL`3fM5H|Sjwxxo|lmD-h+ zkvcW%7&I})yJ}srAF>(Bw=j1$Jhzvh?%GK%Olt&jj3@7m6>xwD$6{`bfP>pA5AOI^ zHxr8fqP_435;q0(*G4Y;TPOzj*Edkn=Ri9(Q&XV06Id^=1`e(vmk(dTpU4%DLeAYY z&N@zO&Dy_^%Xez+en*|-_8dE{9Qn|f8MpB`w$8f@v#EPNAU$4a0iX>QIZ-jPKj3ja z49!hqL_O3JkHY|r0c;%bJ%-W8eoJ=ye|cIqAUSRLou%1m*uP59CM-3Ii2UWgLLYus zbYET-SVYGGrDe1j zbnnnymloSxKV!p;Jns^2NP6CA_Iv;DRd4@rlF9GT>@58g8Fb+UE+>wfY&FBiI*M6M zk(EYlXg>{oLdNV^1Hw8{&&LDntqy&|?SGbi!?H7;rKSA37tomSvI!P*Br6k0yDv2R z@SA0rB3+zMVDVY}!dMIo<(UM|w{&ok$loAWl!+%kWE z*t)PGMKvjv<%ivjikG~TN$2{<^WG^BX#%PhC@Z1!_;^nRT@;ufG34^(e&bPE;8kel zDPxSaoLQPox(l+&M+2@%u`g)6_QwwJ9))+qiE)CS6TE>QNAv9v(=xOwxo1Dw!!3y7 z#LLi0vt_7^VnAxWl!{&|I;c*Q0JUWc`H-I==n?cRO!FB=x)zYSLdk&a@Z4~NHY_O8 zBr7+N9>E>xmVHD+VM>D>yWg(Bsh)+#Q_adAa#W+;K<zFng_fAId z$1ag<$DnPTP=niF-?{ux<7p;xUi(T|nu$OKG$;$&5v;tXFsSVWofp1p@3SJCM`H6 zuxvVBWt{6Gaj9mVYFThQTzg^0{^>A=utR~m=BRnDQ?CoF(4YiWvF}<>dumG92@%iM zEsL2Q>q|~+)})(nQ3ox>aX&n{VU9 zYtsQU-)0xZKwsMqD*7|JOAg@{WcAQ1hF$9DiXar-)1p#!va-?h5|F!BC}J6{F}jP5 z)2*b*6Be#KQMWKpf{PY*&yUwPW}U}zK*b)uG3(#8V_EeNC=*l$JqFK^Ar~u*4^9k- zZDxRor?NxLE;RL=i zwjLhyIOdV8jOXR^kXf`{sdh-`mXVlortq>faB!Tnb(~-SE0B#PIroGDZroH_g}p z3hcj63(!N~PE2r#|1OWC0pmpXQw-W2=1)f+-vK+!-~Fqg+TVnn-|nkkPj*frm1dJs zL^1m*vIod$=?b8=(4J6~o0@6{y_b9EhZ`DcT1j@;PLD0J+o8+I4j_a=;_SiD`=MoS zNMMaSS1sC@kAPqlD3bMbCKN;?VhsW|tYnQz>RMMUtrK8(zjrgglU@A8^;=ZrdPUv;5VKA;rF# znk-E{o#5Im*4B=>ERHLHI3t|9>K7F`vHd}BJ*x-e~d$Xf5mo&%6{{}+m$nWaZT4=fvKNVK;-je5V@>agmtiARY|yJ>i+$35*_2VjTE zAHP)l4a-VOmfjl7lw^=%U&G>#t333Hi@Ac?FS0DQoSUMyuEBP^Fsp2I=xj>;;ykx1MeQtyBk~md%Yfe*C``qnb1&TU) z>%TU=@u$<@+75(&ZNhvRD0h<$WZ1vVtw#{Y8}=U*Bu1jTcL_A3rUQ`ushdP9H*NTo<^?BT#m! z((k2PR`b6ksUeo81Y^bJfq7bF)997v&{w6<*Su7;TNEqT?(%4XVB@iQ+&#`QK*)3p zM}&|4gE-zPzj?mVlDdb5A;&0Sn9f_q>zgwuN4CiQ{CM6@492S}DahW^EPUZrn;=q2$kU<|z?#`te6Z@ns4Qyy3o)C`R`>0vjO zWngQrOH2GOgy+ljbXyn@p9Fs9j>7IkE`?NTULmM= zO%ODP6b7%+42L2KUPE{nZ|D47^opn~NrpfEt9tohF6lZtU+|JJGT+q!?gXD1?^9h$9+s+;s~IgKvAbyaE+;ZH8(YZ1~O+tyi_mq1Om`;`1a)syH`o ztGr#^MjsAH4hE9;M!*&p&72JyQ0@m2@-m-FKd_H7?Y#(;zd`kB&@gBd_WKL~E8U=C-L?OwyN4l%q#J{BY`cmb?VJQyMpdZ_sz2pHa>US z#s-|v9nvmuvIFOfL7I2mP4?x`KYjfiS>wd^CD#lPTPbD>1>d+q$o~BNz2R6T*Q+QK z43O68#-@fedVs%r9{${8cy|IIV45$ENU+N++cF3{60F1wCjHSRW%_SS==kgUxYtOh z6Qg5=8P#P!#oVVz4;78bFa(ltC5aQ20DUutl$MHerP`wK4tlvDPlRMHkSN{cQBOmq z)Mnli(FNH}X0I4}Y!uoCPpsy~VE+w8z5EdUu@|HY8a$W%EOXX!?*{=GjnXa+(EC)A z8ww~0rkmx%&<0#+%MtN&(G7#*v!H1y-l@^Q%uW?vMj^@s)i^P7?7Q<%!8qIfqxt(t^U@ieJOUZI4 zHefr=mbYYz*+7x?RP;6FkYAo}%B&PkN0jlX*RLZg&vy+kW!B;8^M(k(!`nAbGy8I5%Hw zJgZX|@M>M-LoRu4oTf#pP5g*9+Bm>~I}WftGCQoy&Dauy1C0%|Hbalyl05I7F`-i2 zy7q6DG>J|d&S4=w1nmualY1l3f&&h)!A(yen14nJX}2soo^*yAmn3w8V7J|_-01m0 z4$6l9EzDygQpzH+qA&sr}!^hvzr`rV*mAm*+d_w z7$_UBqM|EEhi2QCJ4s_yz2dB7#oY72WSC8#m3+o53tkma1o?mt%?hBf+s7{pf$gsA z%78K-6t4zHK{zNV0s+q?O}6Tc>jucH;aF|2cajxzFF^EftE$lJ5hUkRypC~;qn3&iJ<3%M6R@!|IUK=ZWdz5Z;slOK z-ye?7k(lgEhHz6WiJwC9%^qhR#Q>+z7ApECzgKw1GZosJ7U%kX7T6cC?$yQD!}kYf zE=crTd}w}?;*futYH=I?JcQEX+!p81wXUDM;W)?e$+yRG9-cmlPwscrZ*KG8v?=Sl z$UUoGeG*8KXi7=e3)WHqeR?C3NhK5mn(6zfXe{VF5{M<8u(>@NWT14wfUdFIdf25V zC?BZH46N{LJQk-yw0FB#o6xu$LJkOwAnR#;fEII^>hOq&Y=NT5@#sqs7|wr?lx zSV8_2?dAt|!#q34mB0Is<(4}h7X8g~@p)#KM}7pZ{b`at(zvL`w|m|kgesXGnt?a2 zO1tdV2WH#9o5PoF+kkeuiv7#BlEQJ`^zO>yFZ|pR881nIp9SF9a9>IHiy`QCC1g`r zNuVL%hK>3aiadWky#x3{#^tDT1Q-W^PiQ6il+*78ShmYLt*6S;8rvX8+3-{z${8d^ z051?Urx|E4pAf(to2N||eC#!(K#5I^ogq<_7Nf5Pu)x4R7}urjumr*8pj>q}0p2oN zi`l~ldCshR!gHS4^f6+`Z89e;n6DyMUvY#Y>*bt`-DFXY{yQ^GUP|nHGuDzcZee}r zwU$8rKGJ{LM={`@=TgyIRXq^5EtXzhSSr~f%L>S$Zzy{C88e%rP;da*+jZ)~$a>kJ zpbh$dA%P8|i6Dq5yXn?N7fZXnI;H*7@XCY1$6*V#PtoWg3uFg;v0`LAZ(MTg-uI&K ziji^Z!2ANQD{fF)mqlL`FAvI<>NMKROs^7&f(pIvN|y={2@|%uaGmb)Od*$hIhm75 zox9)opBG(XGw&JBPT{7k1I+{D-D+kAJes*P(sar_s?IQVp?>f^zUiO5+{z)LJ8%FZEzyR#8YGq zl1Chz*+nBcL_AR1Y5U1(kFDdY2pQ8C4>#F{iLs78_WL+tVv@_L|C28@VM6iuFE^6) zPK=2?W|&B$7$A~Lp`w!@j+!UB$zKA>&!5WIF@vE6JZ%B5nK4i-V5^Z%m84oKg3@m6 zF*4-^9lylbhLBNS>DZ^<#ydG7WHRIWhchx0Ld2hm-z5p$TvaDd`0X{rQ#!?Lr${P> z+LGrKVGc(D99Q)IuvJiMu*x$_3IzqX=4ji!mIff9CORys)2@Z+G=*ND28^{-#KtKz zUN$yBar%fcT^#$@Im;~TN=6s(J24PgYE3dV2fXWDVLyv?CML6ig!H#PF(_iHLrUg1 z_!Yb26$iYLqcBPJ{`>FW7Hka3p#d#L5mi18%kqp34qB0qi z{NW!SAn7lSF)24QCi^J{R;xTJ`l|Gf5UY;*==JZI`HZt`7#FL>=;ooP3QQ*+ewXCOPj{d>QQc`v%-dP30x zg{)YmhP>R+Jy9N#1?=n1va=XO-6?J7!LO5Kr?fWQ7<$drID&e8)XdmE=Z;@XM*eZ^ z>xz|x(&T`nZUvU50!|y9XQ^RGqhl8A4s7!5<@dqbamMpQ{;_%6!iHVY8QCprfX&W_ zeBjANza!TDmN0p$v?Y+Fyyzj-ni(}SV0YqC$HPX#7IBLh zi`VgZzi7V*+n1|B>U~024Q}9Y6ck2}%LyElOS4_x`IgC3{awfJZjw_|$R}nlUJJ!s zqR0hgmxYSKkd;%GGWEjEQTIbjcm*KdQ|5-hkw74B%6Sm~sTbDpjzeE~AAOv6Giv3O zJm0<$yg5%VZCBm}US&+o0S|{Oe#-JGhoKnt z5S8&ty&{{RMX!L7QEwSX7;?c8hg?vn**KE%-hBS%S9R(mfd%|?A)f`?9QEKfizC;8 zu^o?|FY9MG&7$+B$@aYZV_(atiqkGmEOC-HgJO77)HXo%-!Jm|T$j5djYd9@ZL}+| z`QMtoJY;*w64!>1WPYo-hurhYQ?2u`8GzQO-=?!;!mdptkMAOG*!||r*$*sR0$Dom zlK6Y+Ly}J5Sx*PXoIKyt{C&X(Wld36z1o96^_qyUeB>8n^VdJ}rU^La_Vqh$h2xpv zhwYZQa9Y#B0$4qrMW=_gD{CSS(Jg+}QmvuFsE$TQ7Pr5PC&rI8fzD>Y;?gz&Kz;HN zoB%sX@|(TPKbC-Rn-Dwxioz>>kx+ZUeY@9nukBt%5l|Bgy*XzB&js#Pob#MW^XIHc zy1r|s#&R$GlGHG=?40$qiA`fC|8_70(NN&9%%@p)FQPm|TO?j8N|R`BE3?C@f|BI- zBCrN}LN-a90P@t6eqsuIyeP!xalniNRU%MBE*ZNlWR7*W@p zEczr_59#D0`kpKY;`JFIq`9Bh6onOQJEifwHetIWMwHIrwLN}UMcp%y&b6P!hg=SB z399Drj%=K_kJ#Pfj14tU;q&ys-uPMlo3H%r`Rv(F%y7cOp4}L-X+aJhM;#67RrF}K zPD>S5(^|XFstvPY)5O~_j=e8o_e@SM+4Y8{G`Q1xr!1P~=}?M>9W={5t4JmE+tlzY z!3N;jC7=PF@82h`6l8#wGnA_FV|hA$V^oD~NRi=QD$((a0uKZZx#;*gZg+$r&&yAh zosV)5a%=*}(`Vn>fxTg29FH9wR>Jhu#=RSd$J1Wn#AO96?G-yc3V1iXFpGP6p*Bg? zCTswj+?JpLk59jr1<9Ysf|>{iA;$WAbHS1I5gc=N-M*8`08*oxgvL;eShNw^b_VYA zhx6S2y{SGXQ=(b7>KNJ1t?tN)w>L-3EaD!D$)iXP6@BvGJAQQK&A2yo>ihDe0XLOZ zf>_=ax7&hR&)ra>cAOqif|sJ&E8Y~bA5u^czuftQA(wkJUY9DwqhfKpEB;;{H0+<| zu!dF}@$%H-jXHS#)Q{J_(! ze>%Ac={ybLRdh3Ss=`_IVk}OC(vn;$#zeH$*I!0Q$Qf_18PH8z(!I2tmZi67I9 z+ISPEHE>eI|9&&&WmIA|EGQ_}@ji0G8V>C31lcklHi5*Jf$V$@J4ia^_YY1s*^Sth z!8OqN%6_qx-H;Q9N}x7=gq2xOF>5KZii&QKqeG3+)m~+RSF!H|_U@U@sts5%yMCfx zy!!S2UQ0d_F5ajUdmb!IC52u^zHPz`@>p6d?GVd#WYKeUFt68Etf(0%Y-yD4X_BStm>++(t-%k75`KV*l*MX2t#>^ZZj;B-A z%j|83Y{1PZr^D(AI|FJ+bmkh*WJ!MeQQvX0>;>bo0ViaHz1T=G2^3j}&01Iuj|4PY zjDQpoD1G1W)}cw0IN0N`GCwxYy_I_%d5V3>@VOfQwO|t}I=^|lkmSBJ`xs#V9f6la z6mx(gMO1X5*Zt6X(6z(Py(59kXOxTM7o@sZ&`nV}vLw}Zuj3HpLGIpq#enZQ;HE&C zDZHaMLZ@EJd^WW~m=;g~(c5Navhuh(9VD=_Rj1VzhCCqbjY7+O@Gg5=6`N6I)r?r2 zv?I*QKDcoDOD60r-}vnxl4DN1r@3N=hO-oNiXtCT(aWjfFRi5d$bCf{jr7mmUOIj) zS;NEM8Jc!pHq3wDoO=EVjZPCIIu~+#dLzF)Fl$~t$&aWd6+tV*jz_Lta5w1B*$#xQYZ$JKO3p& zRM|(OObvF)8MdRH;R%9%a%jG`Gkm4!Hi%5)&t<_W4D`;uKen%w>%VslQ^55j#&kWW zDRAEX!uN(h{HbNd2}`L)p_iVH5!FDhwL+$SK(N*Aig?&%w`d20A`=JZKa@euMh=bA zArsgGH3oo8NO9Op0+32c#uqJX2b?xh%u*+?fgF@#>lVgKnq_CmdbgAIFaEH_ir9ix zTgI{XHJq?I`8!|R@}}jUhs$2sd95K~*$o+6#+Qm>L`ka8+_hID6I>f1{E9>a6SE4A zm~dM#gucayTPEc2n95(RH2JNY6N4_1O-}5$7Mr<0*%Si>(COg*1bjv>5#>n??5$X| zin>SG_L)x=xD~kNQy0P)Bex#Xm~4mT5Q6b2MOg!4C#YOH9O^I*0mBs9I`vjPO$UO2 zoB5c$I)3JOI>nvXe6UCpToEU!c6#K{r7~>_q;L`h4f1jjayy{d!TkSd$`f2xeBFD> zcp|V)To1q!fz8sC)0>%WTBmAV(C?GsW01#bA<4?M3$_I!k!P~9*tIouNTE{?yWH^F zE$ZYAcvR3U$lgU|KDbk>cWqsuO^m3gJ9&8$y$o-|xA(!@MT-XPTo5xZ<6~}-o=mx2 ze&7G&A9lU?eKwf9!;YGc(0Cqhvh%5D|4p=uVe?GeW&Ug*gxfMx`%gc5$IaxA{6%}= z50JRvG=LLlbh6F7mn{^NOpy%`b%V;+_53~sv>Jz{2RFEu&TkXO%40-pff#G8AlEx9 ztZT-EV##J8O|ZK<>wd$Q7hR-d7OoC zXtS~oJdSFgiJNjR^F5JM&h5Ex4*$`{=hLJ)v3X#jNxLH6ADF}Io3n-o`fBn4QHuyU z2b!Wjd2N5-DYfyv_ytA$XJGo{=tv?Q6&wu4G3~M9KF)v5bwq zB!Li?NE(oxfpcOYI~(x09##QjTN6TcG-}#x0?Vko9Dm~DpU8e@xw&Q873Puk-pM2` zs08?v*E4nVdK87cyYgKAvVb_AW1loDUxQ8G=S13U)8EMc@L9`rIc>3f^yGhOu=F@a zQ*S!)nv!ega&D!VEfhc|`r0?HNn@bds>;ySu~n6$DWa1!hgE~%avlOItU*B!xh|~< zs+Uh3qHTq;X#uhFFgWHEJCuFJKmK1!kAurP;>4j37GI=MeNJo=FlkjH6;_O>ny*ur zD7tAab_40uX>?tHwkoiTZjgU0=<(8agf_^J2JU&aYCh0V+>6jFfL3X)!$8FmLnD0= z!?9x!G-=Wgeq8m_7oAQ~Povv#Mf?cLKzin4;{pmx^vHAQRGGG&R~1xF$McQ@qh1?r zJUKxz(y`3BqvmjxQkO9$cfG3iu+Qo?Ch;}n?iRd9b zhpE%r)!D#yXiSzYZu2-8{4o5ySRa0vr)~3CC2xYsu7>IzE%rX4!^ckxH%Yc*djOMYl zBYcg=unYEBw9%OJ{RcNl`d@S=j3=_{ywE2sk=2TzT)$*~xfp*xl=ZkDlHOM4%!-%b zYn{4Xd0bi}aF{P*%?fcbf!6MK)VZ(iP<~(;5qZ(7g+3ynm)wCvZ%@oEErhJgg#4Tnoa&DzZwsN|_#ZDDq->GQUi% zZ6)Yucgg#SL0&IUGSRe(&WfY#0`amRnD#N>Jx#quKAu88Guy}9pqT3vxq`ioNF<_D zCx-z$e=6_HqN|Hy7Ay;1@m9=&>s~PnkmmLc^|ChMQFXB|Zd5;&SMu7G(3hP3R0R+vSMe zG;n&%o%bnyKh*oJFPc#H(_81ZlEd7J1f2I@Mw-kXmJ<{Moms|5YoLPwPTz|c=$O~f$38K5Rhg~rJjO_ru{9@mGei*gGYwFA@pug6x%#Ro_X<$5& z+nOlXh%-8#GpqJ``|y|z*lka} zB4q){vtdQ`Il7g&E2sA3!-IP@v?{I!#eT zn-JU6YJ5v&dSOe@=2sI0X{vari%EgYz#L`hTBIuJ82ZxbCH~ITX(< z3%)+BEaWOQi#A1Fmwpmj7?|$b1`b~ueMx-8|FLujFEcFBKLa*;wSqFidvQUT3-$Eh zmr;io{qe}6-`l~6IR-%Ew^Bc&>yQ7~1R#-U z9%-2yhHcUDe?+Isp(jzN!G6}%X=i}x1tn+O=)(co{0U5G{_O z#Z*X#BEfaoB|(re6BIP){az?D*{wgQkfvW4dW~%5W~!XnPda3V%iR>Siy~Q6H11D}=rZW} zz&Mh&G&s}ekU9$%q_e_H;-q;+zDUed7CI}W|zz6+~-GNAc z4-_kCpinC=S-Diy6NzRD>`@g^>Le@C>NKj+n*oWzjG4nw5wVmExW_EeqI)&qk>I%m zOpLZocr2((e2f5qa%i+%+KrLNygo0oV8f)?2NG7ziw&R7ehDXYGwEH|%%hh4v|Jo- zC-#h3Xm&wKLAY^hJP%#yr1@*T>)o+$zfJgfLD!5u{4KCFmD78|Gc=igaiUtXo%fkE zEucU27BmjDK@G++pwb?2PmH|feQQpc+n#y1=A;E24PT?wsRxv)jP-?yeXgGSj=ezk z+z;Y-!JGeicAwa^O0C*@{nuoj6Z=EE&DNdm6q8Dk%~Ujc94iE|a@a2=E3ZbzB7ya^ z)t+56suX`3(7@Bv`hZ~<5Mlu<#O8Ex{NA2_IrkZV{+RZ-f}>ek_TbkuP56j?Z^l}Z z=EV3YGsDL|ih)XqTq+t1aypE<=^Zr2teezZ7u@jbptq~;hwcCuq(xl7&l3%}!?N<^ zB`#b=5zwT_tYxl9uSgMNupWu9qzl_z?QS4oYPloE%A#Wrh**rl2{Dt8y?5{D;|Vr9 zaoq%KusK=TCQJz|4AyQ{!;dnzo&bDxoLrq3C(82LAp?dm&kEQ^X`h|G;f2=eujRCE zL1C~?T^^F)+AL0^AE9USEC&-uVfh)hyDk@C3QfB*tP8>b*|gRuD>dx$*DNpmlx~pa z?35?X&!;|ke+k7Iopynq6*O9&vOv~zU;gdW*>Zu&{LD)E&T+Eh1tT8JG&4z?C}tx? z5~yfUSQWM_3xiL{AaDp2gw3+M;x)>0aRH6LODC|7+%N^UPycgg%&r%(Tb}Eme09UO zO~8oz?a}+>v=gsOx6MFsg<>vKWQI}q?U$j}0Nj+NBw5WstH<_PK z-5c6UZiGGxZH8b<9lh?$DgWLP)kh2}QRX+~@)D@!-}+<5ThH$4Q0$OB@Vppa6?pbn z1;2bh=1Upw{g92tP+&e?q}VPqe$;Kj8i){=D!0jZg~!6X@%%m>1`^|WPyU!Yfth~o zyk=!JhyZ^wXT8c6yJ9<*ERDlBUfX~=b}U&bFsAmLEB}MYgs=CCrv8bnbz*!0$^MAY z@ivM{!QEW+UHW#YPTlHL=F=wZp?kk}O}Z@jE(nb!M(*S9qH_Y5zIsnqFVv}fNc+UK zVsV9s4Kr((OSr+K>0`d9hY244eucb461fHUoY=|PX9lJWih(A{tyDC+CtCD7MPjq87IK;m@=Oh^Q&V$%TSE^A41iS{YwT~l2reOZYy<$U(yIpx^#$#zd_o&F|Z`{T+906nFsN=tk9Wb?x!woOG##Ess87d(_ z|CqPQZ3tu&KMPtCb&ie=(^h%(%k)6oe9UXuC5PT0cGC0MOl{2EPTu~oRq{3u?Wds~ zntOH^^Dv(-j*u8>B+Li0^DA~pSlJd&I`xw!u9opQr%h9{#NH0kRYB+I#K>j5oAbJq z#bVUo%=M3#4EpQTZ5}la#=)rRvmF?tZguDhPQakP`=7!2mM!})N&0z@*|L*HG21AT zLPej5EcAkUPWPs$0_90@v9#8`l#eyR_-Zv-<$5&W*vvseE%~4vq?uc0*YG=p4s&5g z0>O4XJ+m<*?|0Pk+h?zn@ibgJae)F$!}T8jBd%T09D^zy>AV8S19i@=s`BPWHj92^U%A@tzC((31*0fT?zvdAiZ}$0sG^|F>hkbzwGC{8jBH`2r~B7zEh6> zi)l~%#)rK>A}6fn1D!ZZ(rLDsT&9?d6lsKfAkSFcjm@|F6lY$;mXmm1Q3ST&chKd+ zbx`rGhr&6%LJy4r<$)+h@+?dI0!a>{_qbZ8 z-bEi*p9J!EjK?4+4$|k?W07+Rw4Nbk%txU-+3zeF&YX64#KKut6BN(80Abjzs{S|L z|KQ!NV1|ZV{__6c-@PWhHG9bA&rn_R?f?y3ym0!@@Bii9rJ}3STi#bC-Cjd3e|>)` z^)8mR+S}0HeRLmnF5O0OOf4Fbl)PJT77s^)CuLj1@*G?^S7#2h+BC5(~1V9}vNU?&O)RGieymfUm+lqUyJEwdwv+cMs1x1f8{fcY zq-;nJmD4%AOuuzZBlx<>%4}tVgK4bB0V@Z9>DiEvHA@mjUnq_GcZB0-b&w)opT?NB69;r7tbD|nPUWk78h-Jlm|=;2Ho4-&e$68@ zKy*U?XD>#g|tpjR(1>5f$cG!LbBf?=y=Z7 z9D`=;0XlPGrRK4xp}{p%Qy#L?Gli6eq(lFGtN5l;i@Xrk1W#WK%yupF0dkS@kT&6M z!DG)ZWhozvV2%K_lJU)BWcBoHsBftCjTe-KpAcwM1F+oc30+eMjmuu1Y1BS=cAI9=_0)lif?G+?{w+KK;Q@o)GIc*Z*t!eQDGn>hfSHZ zah^RwG^3`^aAKU;vOcIC&K6Dkx1%zXbrOFjewQRTv2}t?)`;aQonp3AB$bLz2}rR zWzt97?yy5n28H_b52NmlCp+cD28E^Kdl%g)Z&zYC?`m-UYdfY_`sUDm;02u_%U{g` zbAiiFIb9OCW!3Y>(R=A^RTkZ@L@vA<*$!Y68gO6FPjkzdsZ%!y`{p!7A>9lz$LQ1#r7sSGvx#vCz-As+V4EJ83(#NZjoFA@- z?$ckj+*~>B#++rlR4TEX{vT8z+4JkXNTrDgq)w>?b;OeK+z5*xXVbF z9M25;ce_>yxBKMyoc0;;Y>G-!LdqLzAR41eK;Q9v6bip)yWWH9nv9u6iuj-&dcC9+ z$QoP3mA<_T>}F)_2OlG*MnYgb$K2ra4)2zNHrbuN@2XCcrA};j(#`D728vlvk+oFx zlRylHU>0YI*79+CgXwS%Dd+Kg<0J016%-?GnUF(HpqLy|FL?V!XZGkI3CdcT@n<9K zVqmfHDxiDl)eu@KR-E;Y5tWPUg@+@NxMlOJcC!l4nN3^q@pPGa&Up5En-e}Les%jd zqD@BQLCBvKq|J$qM$F`qd!8POxl55dRCK2_U0N)~*}oiAKwn|TLBFGOla)Pm(U-1C zR|v33*m%#b@FRgEKeaKcSehD;#a~MLVB~V~`KVe_Z@S0&Uu3)^6PT{D_{KXvM@Eh_ z=$@j_o(H3$$_;di#%+|G?6ax$0c7Q0T*;jhDE0a`6 z+_6NIC#mCIQ|*6s|EoIn%~>&SaTGSi#qzGWb@H%E2Cp^NVmaQNjS*9$L15=IzmmT8 z0cAqTpvT}DGUUPuW=@O}s2v{>yoslnH57?M?+2NAwMR62!rS0*jmN}$JvfK90K(35 zZPRc+ytgC6Qb5CL_eU&}7|R9A1;~)8MJjQF4p*7m@}MlpO`@tp8r=kia*luGpMBl7 z%=Ocyb;$0gJ-u;Kc^|;_qD`$b|Y0w)~lA!`Y56)Y8PEg zD*Tpfnj(rKu!|jQ-?q7K^E{wPQpHP*qxMLlb;&v*FtiuMpN(jrBX)?3N`nWyO{n?l*-7U~Dz^==6XWTi8J_YeCWj)K zRCJrLd&XIy*!!sRU?tB|9nNXDJuI?yZEkA?=e=uvH@~`6 zkjgtrZq32ePM#!^DNLHd#i;6Zy5#=FB9R!^M10IcP zRIJmfkIgF*=W6N#Dg~&tbvkTYn4xAX^2Uc7>uW5Rvut={H$PT-Q6|q{ zlzZk6CX*zQ{UnlfJ8>X5-YgItGnpBr$N&|cC(_f9HHzn5_bR8?s}8soM}7>&b@u=~ zyMR^5=x?CD7>R1hg~S8(zCbIlgTkyXXN?i)hHDg?A`C9mp?t(c4_xl`+8$ z@*$U4kK=x*c)E)|I=5d|=2jXupK<29yuw_ds+mOSIENn+wFz-E;xnt0-3`E9>PIu;m|iY2@XK+JM9eKjBDV3*7k0 zjNLO1FeyR9E_g>gZ@nr;gan-!8ElZ}^FAW?+*^3{P|BCYZ<%(J|EvGk5M+ZZHweD_ zPyhDopO_#RkkJZNCfrs<=iOwGKC@M^lVbD~X``Zbs>YBbP)$0dxGmTiQtUV2sjUdu zD=zS=pqGo{Lk>XC?rzUUPb3`w)LWYdf(1Ye3nCe)j*%xRfQE1&EAJ=k7%W`JlkhUt zW4E-9S4$EERg#NRZJq?R!q#}yXmZ@N#$Twu5$Aa$REr(Nc|ezw8@`KYeCsfT9uGq# z8YtN^HF=U8`m{RT71m!`Yb;(xeIQuGeFwd3fnX=cE$`L6;Y&ra?ze=lCxx zENRi477#2HXzg={{nO|?e!KGgBGli(e!3bsY7^>(hvqj${mppTJ5dxPFwGpF{|NyG zeh4RUP~J&DY_Uu{aA~b`Vm3XNGzI3NfaJG7v^5knQ9T5x(l$=*64yi=(WC+;W0}wY zPg5X{ex(29FL+PFKOglSC(E4JYtArpLN-!N0!7wQ(MFx`Zdc=`DtRg@b@fFw%W@~s zPT+Tm|Fh9v7%!f{pPcN)q)&iwEnD?^zMwZD(Pn5sgt zS&a6@^UegG3j`TR@ogvuKL$+#6T5`iYFi9@S+)sw#QBpE-~UF4$y%&R?%7ZFaNG7e zaWMF#nH?#kn8Oq)rJ|33{6r3KC8W;Md3zP-fPt!pUoPJK>c08cq!o%Bw`GEk*>@m@ zmnM1i#pSbd{a1#SEo_R~EWr*)kXQ<;7iNbgk(=&a%EU;$%uqN1S03Z-3%=r(>t7q* zq`;(FRiJSkdl@L3j1F5Q#M*f~f$46=-@o~$2|tSt2|Gw7H~cuUcY4hXKMfRfh9W1a z=tEFJheb4d!nN3lovFDsXV?WJD2DQ>6OmnXDKLz$pP^G9QtuV_yK9lmEK}1gI~3jz z^4Ian8p`xNe-U`IwI*1@9D=~I_H8R09KkS4e z9pZT%ibn+E)1OMQbO)=}^F+s`U2?7_&F+}8hBMV1M@T7 z?*jiAl9{!VoLMlZFq!tU23)>Qn`W8IfOfi;-CePgA(+Ih2q;#V>`dA#|NUE%{L%)AZ+iRC)MGco|i4*UOSPFxB6~^c|EcqM_Wr(=F;H5#i;6xrb0;*#mWxl4){uB!xv)pt=vkq-9F*hM5uD2_r{j4? z=N3U4yl+lx$oy9nZc`(Vj3uNhKi1vip!Bs)2gfjR}@Pn^spAC zNy^1r!|D}!x+vm?7m~Wx1tHgJJFhAzOH)oS_r5AM@&(^_L-L1Jkt?9+rpu(3Gwf2Y zC=xe;zmzS#K<{_UHi#Xg05XgmFyN5`$w!@fNO91$oZqO9a~pPPqkBU$HI=*zA$r+; z#a=@-J6Hb~5Ho%dex5mc_8wPoUcWE6YZ=6IS{IllV29h7+n#x%Edf2G0jjUcL)Lna zf92r%P1-ogV;FYaw{WsFlYbj~;?wa2@toMwfK1y64|og3BvWJqb|hhGvR;Ng2GE(5 zBsnx63B$F(1wIRf(N{-h0I}5sLW-NU7&ks_9w0R2-LL%Evv%eeTF4}I_+tY!~fzbmwSyo6F z{nt{7p;cn3tKO$1vem$Frl<7*eRB@Yuk+4%MW@d2$&?Jav_pl9PL2GFxlo}J=Xz+q zPJPaOxpyfL&-X#i-x3l24LY^aojytSgzMD(P;=i7jgh$3+6f)Gu>WdT9&~R~BuTbH z_-5GUhL>K}Mc31D)D17BR%68@H|#uHIqY|29ZsB#&?KGzn%`MgDvTCzbK;;dOZCE5 zRS$i6cFydB!Mo^{Q!c>oxG60pSNfMFz%B!|V9d_II!TtahfJKc;RqP(v+IDf?0{MO ze{y#Jm&ptT{bu@fa`UC}g=5WXKKm%9mm=M`LM4MX#yaMRrd+%pTedJT+)p5QdrrKQ zf57{g;H>u%O+0V$0dL%u=SdC+7}hVHdUs$xFc{u2 z95u@h(-0L#(i*+6L$OY^EVy2F(+$NDpmGypzi5`yXw03X!NY1!(C2{Gf9&sUkndGB zV(}ch$ViwJFG-VNkC{%b^SBt;4ChcnVJ8%F?iPWV$E*THwg&&KGPew}L2y8^gV82} zWP7tRPh>(pM@ZVwbg_FtY<=`*ilZ*cZ0nwg|oc)3mTn z9{g87qJL=?wq<4u+kJ}Zp~zh-x`aG(ZG20+(es)t<=gN7MY}9?)70$}y-Xkek!GLk zBiCl;{M0zA1s0}W<*m{=(JF6vEL3zQJ`C15wOW)X_WDnvmrOsanu#J)9M)N z{r|0(FJ1=J=4o_d6hxmNLZL>B61R5if-zCrDev*qsqu9SP^814^quM2B5|^^B`k5; zb)Opy^uwh1mHS-f#Zi|RK5`ulO9*aJYIAvoUgbVFzmPn2F}9`;c;sP=8|u7=?O{XfPG3EduWG zE^%)JvwuKpop|dHJN*%Ho(mN7F-7XB=rkIAXnaJwp}zwlzC8a zQ!Tg`vEI$lSGQDj#}%IPItUmiD+|2NNHb<$4>NS1oN?O_)kIspH6d4iY+VP7FPq6dKY~dDAD(XMPvP5M>x2h##fz^Z#=!y%G6g;bf% zUopi1;nZF#`i9rNxhUV(3lwNOrR_ZJNoMN;OhaEcL{FfU8{f_4eaOdb^>OGy!@e@@ zadjo^LCWd75OXmQTnvEnWG{amqrD9TRIB9ILbL@yDQ{=FN&C)@jbFVDmvNdM=XFqL z%MaID3IV-P0U{Q;pGIDetV}RCcasFfe!Od7licaqC@q3Y^`6L7-m#hYL$m0$3u0#= zS!|mS>y(cxY9UZr1|i1|poz5Ac-VjxYnO~2)z(I{-#KTya{R&+n!kqc=8sZpVO~(StbP%$NT4&ENiG z(cNEWmQ3q>>-bw$KhUW!iaQkPVQ0OYqI#joD?X$#q>f%brFY?3?{rC)C{1!+xMGS< zT|(?mXKr;S$AFOuVS@VQz{~}cOmLjXD=uIB= zzzy-4dr?FVL}avE?5G81b$;1S+A{BK_d4=C`?|8K?nKXvUOADW*&k3YKH&YaCscVC zfXEXD?HueuSe-3frhj4x-^~9!_UX6s^iTfF*Ckb!9kio;Qs?!l2}_@Bo~RKjrk+$_ zVfEB)NPa`_z;&-B@KzeVEMP)GV-s3T8(Z`FbnGj3wES*&&MHe6!)eQ`STu5wa=e;< z(|yS0JabVzmhnG^%9*gD8vD~4PWMCa#U zgfia-zarl@VS)05X5-WjA?|6Bno659t4(+~e0k8Oh@x<4M8M7is3eCNt*4`xqcK6p zh#`jph#fku`X7DC^+~+!3l;xMVx2e;3mqmSyv`(wNu)?T624(SWZd$a_HTnW8_fyLvQ^Y5FeeTo2}jN)+~z=XDfY;nm3QinSk+0x!gxqbWqj1*6`D@#(y~-zVkb4sG8EHd?aDevFRLYuMn3drS(oyXY}f^f4tpn%b&TV5KYyO= zbOk3XGMTE3nq}Fo@3i1x>A&xhr^>K+DGll$wJTJ6#TUqKpo0F;GZhN4wOJrT-5~Fv zi|O^ReGG&jw}GxOXMUcfU74lH^>6cN;A4?fre@H)i#FsM8|0W(PFBXs``z;-2fQ!H z`enTlK(~+lgxI-MO_G&{9{%L1+K$i$d3snUFKud`D9;V(4Rh$6S#T&STev42i@-TK zNgOf7X<4$omsc0+?!1C!nUXp6a1=+Vz zeaL9ypJ~cxfsSVfl!F;rNOL{=VG+P;YgC%=CrCePlp+~+L866I5d7Hbkuz(^1s~x7 zWvcLDcmb4}tqa3qg`UVEmjm894N`ejlRY!KXfzo{XCqnp@U82M-XC%)a4T>dazT+3 zl&K#wbldAxDc&f}frG5^h?g|TvF{$pN=4W{a+aS4)@RrSmoF%?d?gkbIV{U3ZJaO5 z$JYU`&1as|Gw-~QtpDHM9M8y)I(;(BiuW%Vva=iF{_)~v)i&yE_S%Eb@JP0*yQ zqswL{&dKvPUQq`n3e|#2UruY0IT~z#H5NSEKKj_NWXJQ@FQxU#O_u4zKY!yY+2F)G za?ryVu_NC>F+ex7jfzfy*0o%}W`FciO8nDQJ3*$Pk6iUi;NP6r89o4##p%!q(g)2Q zd*B`9pV{fr8Gd=;1VCaefS8C>F*_ivgl(u_XAWF^%>{EHl&}b7mxt85#)!JU1VBi z{-x&o`^nKM+u_>%1aCaD1z9Z*K z8gOId*`f%faJoq96&db_{V=720X5WV$e~fo1~)udAP@%<4(9h{YRW^9>9W^1PgL#& zIw6YIka|T`V3`lf+~kRniDArBXyZfsxd!_4-W>G>+jxTgp|#>0IBz7q>=&y)u;e0g z+UOn&yU_8lbk~cj1K#~0AJ8VeE$dxyj_w5PH$_zl>&2<6t;+TMazT1nHjUh^S@aU7 zP`POqbdV^ZPk&Cv%%x<2|Iqt>cvIAMX+`8s$VA~4?XPtJbwi~NMMX5~HpVNFf&15&R3)N9D4l$W9zn0o<&2^su#+{vaLjrS+d zf{ieLdKo9YQ;MHm{MLV)tkkXf>t7|etQ9~w?_)t$n#G+SQcNF3da39pb)rYPDqV1B z{$S|+P@NjHOZOpUb4A<+B@k(JS}-`U<#YjW^*k*$(r43^Gg<%&J&N?;`_emay#LqR zf=v-{O_Jttco$u;s8_gp-mnV>r0QLF0)rJc@yCcdy)LQ@{3QJ(Vb%}?jE3Zwpb!V@ zDej8bC>x++>m&bR7nI9Eo1N@dCeT*$@@1(qZ6}1LGc`34J??0w*D{cehmT(!RT#XP z+4E|(RGXw)$8>vM;$eJC+XD^0IS|N3dyWbTu}qT!SBhpC)^)&sQSG#Sj_i26ppQGH zGGQ-&_=g8b`b)D9D>qx{_fyOsisS*!JxthudpyJ{?}q{@S|DcGb?OGlXzC4^JUH_> zUC7(ZKkI#39WQ~*gAaCdx532s%UVO1ita_AzqMX<-z`Z6ybtOuO_TjaFVT|N!?-=Q zNOnBI{)*^ z-^eO%C~@M=J7}JcKuHS4fDgElicV6s2}__yGYtqqk|~{fV?b9Z$XvF$wYh1HY&Gr5 zb&*>gMfYiC5;$5AD)i%E(TM~F+!W;>T+fJ|>E6@47`dC8#9 zmP41y@p!m#rKQkO>|B0__`u+Uz$9e_3Gh%Hsc=C6zhgtR(IPIQ`rH|mkjHR*|-=o57(#ELW z-?*~qn)G~BfntlSj=sUGP^_2qO|TXd2V{(OlE-u(CuC4xzrF51U-W|D?3rCNjoI#c z;4{;n2YLofedmcT%((67mgZ4YWHT&A-Ram9oM16oaqw-aWyhk^8Vr{1L=+sT^i2tZ zt+5Qq8nD{IfeQuu?Ad|{_PZVT4mU*n`jtQbbCAj1DC{a&N^+do-Ka5hH%ci6s&@;h zXlzo{mbo>^5gI5$a2CX;4|wZk+9uxl`Bi>b#1KLjXls4r1|TR!Dz%s-1l|^iHSZ5Z`jAS$#jy(#TrY%Y(;TD+6PVC zkn^n{|NRS=Md2*f)H|YjLQ_bQ?~~u}J$fhh^}@G*pRB|nQZ|Sfb%d{yUlgbKT?n}% zE?cNmxhgVG(n_$T@c#EZ z-`oW;+53uKV3qfU}r(1x<)%b$Gh#1nWi0YvWDOq_e=%5rUIk8`QagQ`P49j&b zOEc|bF<3|8fC1Jq9PtJ_2CVFNC;#ryuObB|T+B-O&T+EhrHPwmnxSYD#cZTV0u_y1 zaJ7Om0TwY-(AqTr`-(HtE8-+orlwsv9G>Ry(1mf-j5{lgqW)IeZz&`?y5`A=mlzhg z&}(mAljbkFq3m9CA5zQ}pm1M7FBioI>r_Q_kDx?yGwPjs{v}}B*75Trc1P9{JiWoQ zB?w<3A#+h+y&TWQ^SZr&($Kq_57n8>Nm(J!{Om4|TVxmS9s8h(14^Fb*FJrF6cWdB zg5B8svi-X^-!&PQgFp9>liS=3ixcOcSDEFXA5ly{MebA4SYfGC9}R-8rR=bCV(e&J z7NAq_`@xAfyS}pd`|DmG{QiMOgI`_oM&~apUXOe0!=C~XW-c#1%#@o(zr7-8Wf*R* zu&(FWJe|5CC_4~x&7HiC-~o?vP(B86IOS(xy0oS9jjJJa@j{k4R^EU+ErU8!o6z`9 zZ$wNmikR*Oen#y+{uMDMlGpN46wYv}UAfh7_q4N`b_f;M%Ljm>b&KqFXd@&Tag{Z7 zYPJbc@aYL*{XM!jV&nYYh#`fx6FrT zEJ)%f3iPtm>K>p({Wv1S{bRDk`@O&Xz9~xY(@&n9XcJxtZjLC1!iQ}7NMIs$>w8&8rSE$m{N6qW3yq1Ljn@BV?@i#E z%F?uPuXr!X!H|t$atjb7K(Sb|7+TPZR&`G=HQoJ9Z_{7T^h__G%yifE@~`em*;{qv zzOtyGf)YSBK|m1%Q9*FS(o(?&MR5a(pca;bB83Y7_a#9|B$!JQCR$VV`^nwT3FJKY zJ?Fg3^NhznjoQY%(V`XGT7?T1Ly; zEFO^GqptC``vbq)-%0waiRqu3N&@EJB@AWZ@c;9HnVD9hC#KxdD=!`7L1&VE?NCOqqwxAxu(ksEmx18yT+yuD`#jxc3gcSJXfNQTO zvMRKn&gJF%tei3sQ7l?D<#RTXDF7y!V!B(lKcLM0n0Obj9yamK>U<`3+Aeu-)T*F% zd6D~4A*RH{16#RPnIUZQESs6js~2Fh@v)#&$H58v+ zCp-;A@^KtQfpPLgavN*vS#}6_7)YLAU!sN2W~(K*qDV;$%6-`*75h?9^AClBQn1 zF#!D}n7a)^wYzvoKk=#j)+|)R&I|4ci)V0|dk;7x2Am#~^ROct3V5UIq|kFCY?W5K z)iFa^n1!C3z1jp=ft4;UrgMR70$YM!POPvu5`8OpORId~UFrjZIZqW!+|uc076b>Y zLov9Ktk^Yoj{2po@PnCkz z2q+dcFikX~1$Ay8lM3GFf}1n-cUDb7UF4Nhc1jKeo+b6NZP9&^dcl+4NL*7xd4a4u zusYE8D6t9~&zHUyj5fDN%v+t*#4k??H0-H@)Bm=fdfpj_g@4=;jwjemHzTR``@ZeB1yN6MB%#B#m>L z)pbyfQXW<0-Y$R2KPb7y<^^Vo8w5*yYltm@Wg|%G!NmsjI|=g4wn5=0^t}sC zH#3WcNIR=p&R@d1S2_V1(uknfdWwyw$XY4}Hd?AOx=CU9jeu`y(}!7gNnVYT;hW9 z=LAB;QgGOOW%3dTExvoZI?^zWe6M0>1}Wt>jX3a%x4~!{IYO~D6eJwSvVW*d7qn8-=`Pvvh-SI8KwTZ92cBIo&8WcL(u#^~#>p;$=v-j82@qf*cEt^!8ZZ?Y+Hpna9gWfWbLnhz^V5qCjmTc>)RK;oaZ@jF-PkwtIpTFdTO~z<$sFFRa0~ zxIXT^L^e+WV#|>`gglA`q-0VtaY2XOHo2BbYUr!r0(~&M6r5vdz}KixKx}WTZx^GJ zt_o_LT+7G6a-;g;*Q)=eWB$`0^!%U2?=-37=dKr9Yb3@VJuXNwG2UZ$Y~@`}NO9a5 z|99{C&Hc3jDen{=^CGt$I3r?}QTA{@#Xh3Q11e@fo<|SJyGR##F!jMyt@5TzI(REU zp#a%k$~`ljksv*rZt&Bj{9ttyY=u_}?7rHMc5RAAKEN`+n=W`h)DX$j zkXb!PRqI!!tYxl7K|-J)ZklGjcv%Qu!o?6SQ1$_FgeDa@Uon~ZwkXvtM}@RkTA-U~ zB^A*cY>4@T_-tupDbTb(pjQTTyOuJVKIx$u*v%d9Wj@WyOKvxv2c7!ep16aT=O{^q z1Ef1DjYQfa&{wFJbBLy)a!;L;AcnkUeat?9#k+5G-wC<5*-WtA&-vQb?gnCiPF^gi7_w@Gb3J8FduMU!f!r8M816I!Ti?9~}j2DEI>lV^}} z2Sy9zYK`#69jDkM6se(NhLSsx5)UZvWLHG+S|=-Z2QCqA^IIa^8wkPbba9^)S81{C ziTvGxngrieh-&B31>zfWD2~8V@)D0WL7%rqFaMxV8ND@2lPXIEKD|{gOGCQB$&9X@ zp*r+7fu4sNB^|KX4eGRUPD8+Aal;OD)gQUPD>+uX;t*Q+SYj45NBh-hV#9D^{D=%++c#zr~m4=BzX#f znBa&vJ%?hqQzRWZ#?byR7HUwB1t)@gf+}x#!I7UhFLcUC26 z8QcuU3l-C>(fND%-(}So{POO{{RF2M*>N}K-}|@U&ao!f&vE}U0@R!%;GRyg+bEJk z#dHgbAi9J!ii4mtghg=ZluA}Sl9$2tp}1nZqImw&0L>+~XJ~%<+k{ARj}PeM#;?T< zkL~-m*Q;L3(jx7MIPQ$@0xWyf)X0|yV>RVajJZKp<${U8K))V+MuN^t+s+VXzi93Q zZni6%-r!`}CTGsR{NJxSV_NL&UUD};N9RlO=zdqwzHNt1PpnWUM=`JoLc_Bmv=|o8 zT1?CVCs<7SXy9quOvCzPL2LdxQt*nA*PJk#J@!*9BypBeF$*t3Vjk8u74qBYZC=Mg ziy@Cb7kWPg;vM{2UyM;StFMUmN9DlArWTd8Q3^S(F{|U^a5gBT1v?k$3Cd7U`3@gH9BB8K+gXqe*O>m<9HTbQCgyF$C}Ggad+-14 zhUw(P#eO($5QLbSmR9l~kp$lMH=5O@qN|cqp3Uk^UWclWl*k{Es?Zyd)KDra^}795 zv-&706&)6oL^Z2B$;~-e>9gW^sNK&c>3;g71~wO@!V*B#b*pl#Qh!wZW>)lF&l6Ca zVnZl$gW8Mpknxpq;%by;zb^mEg=q%6((=}Y4@lB$lQXl&$O>gsEacs3shD)SSH9+r zs(FWluqZor_T#YCio-#BJyC#UK)%npS)Bu_yB8S!b-jM}L;AE5pRgeyMi~pZG+vmD z-ka^Ney8X=Tfc08Ns%9OpVT<;^mEl{;Oi;&G(}D##lvRs>Z8j~pK@Cl`A~9!iHCHl zJ&HXcnC67*);0vg7j_GjD(YQt zLdC#Ic{QCeH)c7zKhtTZn3- zGbYVLvm`+0eK-)M?a()qqpGJL0ffLulLWXajXXtGsKMc!6&@GW$($mG1A(#)?xN%Q z{IpUCH1z}6FjMaky^2~Q#E!dYPQ{bP!)j+B8GjguGfqI7bm!M+-}tt{HtqY7|BuP3 z(Kg=>yi2-eWUpE%wwWSLApXS`i!d&E!=vBrdbm~zrI4V9)8~3jihSl+N{dm-^X?zd zL7YKo5eceWqzS&Ye)^K!^X`~U2nGzIOW4^^ltUkcG-NO2(MyEr1H?`qOdW(Qjlrq6 zT?d`+e=UXVnwCQDyP{{PYe=XgT?{%-!MmodkIeT^2Z_fDUYRI`6o9@{7j5yz%+81# z#im&XJM)P9Y38G<9~;osH1pCVa*^8t$$@bPSvn)OX}2i$CPi-G0_hWFl5f7S3(BA{ zWkw&%#=>zugP-OQIpmQK8}f4b9d}GoZ3EZxP_a>tD$S!54I$_l_C{_D9l9avavO9) z-t=~Z?psFWMwtE3a2C1b@!G;5t>y^SFC{B-=^~eGh;PCLR1L=yh>Nr35OVK=TfM++ zIGdt``*#<~r#fdW99;l?;i*30TZI5wTm_umrGN$@A;#6!*(^N1`9 zx7bRJHZ~)!*#SNydLRGc>ebmiQ?7l7)$K4N0+0?#H7Nd@$E&76WU@7^F!H)bJcFzD zL8r5-V^A)a&kV^<;!}f8n6opagNaI%s3K{rK^}b=tQz{!yR^ww^Yhmz9-~PguB-o( zX@1{p;xRIl4T#0xTHg%k;%SLs4$|qzz|Yni);p`)uUU;OB(V8cUH|gkAI6KfIiGV* zAqyXW|IUKc|5Np?OY;|EflL>T2{>sUEef3tN?%4Fa;va~6U^w{T7C83Ezy^nBAew%AiNf88SL&jdEu|L@Y{T$ zoxLw`dQ%;D@i=||@foJ&xehzmm?^l;Bpdk4-%JV8>0qQXqUEmlUhDn19P+CpB2|J{IX(VeiN=?bfb zrAMV|{q#isf%(OBx*a%XOuqW(r;*X(9J4fDv4&rWJ0sZ+>4 zqj{u&VnG=;n~K>OdQ7rwTI}4?$QEh4;)=@!;NWN(;^Wxse;#%O$qLM3(^N>0dZT8_ zDoUSCeN%B5DD2LYQ+}EQz|3(RID|&u06zN*#(%(VOmJLG{u}>CkJ*ybm^kkI60VcC zIT^rG`1vpYOk!U#r=Hdb*@+YzN0Bx9LW5_Sc`Z_$Y3#dWF$Wfl@N2HI9j}|m@04A^ ze=@ZnTy`xEoV;ts4E8AESoHC(2nQAHYF;frhhBJ&ZlVtdwb2Wq96Y=*T+{ESFB;cm zL#;51DWlJK(5XjW7o@pJI%s5iL^2~gq0Ax}nOHv4Q+uGPR}p1?H^kwrf5ASkO)h=&xk(E1|qW zUxw7Egcb#2v@$t#1_N$Vq$|J(Qh_*;k5Q`}Ah}=Z{yvfvpx5+#s17DvlvUXup|NJu zW(!=H;^f)66@W2sUPf`0Oip>?$M-SJLQmOKtBKZu!`=stW}IS*-AR!hR7`9L3J_)p zGgTmggWB_Y+1bE2_ORbYJMHCd02nJ^VZJFRz)*LRm&sps7widT20w+-d80BV1hXrl z9J|~TtEDhu0WvdF-4fn8Pa0*1gVsk@)19+XDMcsffHLeZsE6#A*tbGz+$l0>J;V{WwM6vEclkevb=Mo=SSouw@{A*C+&3_*`Z4mdx0X2RLm0UhN5Ojq3(>=O+lFi zIe!la>1BBe#CtsM(Yu*kUaIsIdB`h{*bv^VJ{O)JkxeHkkSSQJyfv$i*7FiohpzJ+ zzMjjwMcM>~&~F7V9(pP9sEn=_bOS{OK5>;)11Jv!qNwx1fDYAD&rGP8)bkGO1Wy6D zi@(~+6h^L`QtjLE7A})kP2n)>*}!H!o6}x?U^zrLUfXlV8vnlIZVQPS|9(5HigHvr z9_7&$ioD={g$A`RK>Hm8=d#>Nm>VAZ1CkY(e|!mKqakOegFYO1D5wxi;7Xy=7VpGD zrUTGrVvO59P6Jz@#c~W-{$$bibrJ3cQ}n&svE z(T6<_xlLH}N_&3(a)*oqPfKPTGRwcy?$ZVNoo5$Zll*gCpT^1!cAojGHf{{uhJ$AHX~cz*9J zJhLZV%qU}S8EEuti&2`}?Na)7>g_LFqY*GHt2XpRM8{Uq$KL7?Rk$)k7ifqtf5UMi}gd*lQ19{DDI z9-SB6!y5>S^@2%`PH{d*Z*bEj%-yO;7k7!y(cR!d#?pQi#BHOmLsc*4c*6Ww$om-9 zKh?f*QOSxr`YI&R6h>Z^9P{dtCxZUQ!h2p_VGC=S_|P6+na?T;%QK;fb9S5H;R0YN zUmw{Ziw)6v_sHYITglTfq$S4(F=(o3S5$e|&U09{D>vOXr8O0m4s4jrR8;Egbb(IB zqiROF&xX+3uE?0S%k7?98&>AR)xD}Bw`B}w*I_OdGN0`wH4{NN&c#GM-vrssGCk{f z-RbXSn9g2i%>Lrgh5ICDJu?&oPD`fh=(A98rjwraj14)==JE!h+^j30aY8w|Ien+k z{$l$dm|rW5;$3yzx9q)Dah`sL9Z~O(Wh=;b2i_3@6ZD8wtUVOFiz0x{J|5G+uvP*+2B>`r9BO9lP6Hsx)-!=zE!u56LwQ+Gn@<$DRN3%t)r z3WWCYoEV1IDxBCJ9?yE@gpo?nex22`qcurJRBYUBZDFpX?%$K%!C_^vxK%sqRYGERfq z8R(M}m*2f4GgF;bu5C2l(IdHXuV4 zxB3uCdu>8S2aIr3M6qB=@~D^x(&Ijx{2xf~vxD&O^JBfBTejaSP;k39JJC65&WhRk zeNFr9LH}m;1*TJef;lcPr8khm$l;duLL4V_nJij{PHluH({5I`edXI3NU zxpi}(kO<;C4YH-d7e%!UrnV$2K6ZxcLcObF4|`c;c;a*EMbUQlh8*?W2AwipyXknb zKBW{i*FgaeI>AZoG@R%OeDxx2sSFZ@Rw@(u`+UGFQX08cp_8^q^Cc(wgHG!sk4-!E zc6X3g+35j}83A(K?D0saPtlq~LAksn*ok9B|T z;{u&X!LJ`LzTt6NStBiBEfQbL=NkNiS1bT&@@|5LioS>+0SEsH8hFQ ztP`OecMoVTHF>Sq7SF1nQV^8H^1>Q=g{xM%%)eBP*X{x(Z&lDSF^X${0<~)$5n9z> zb9uXf>a109OCyQjYL9)}r!u)iX^`pXbroI@5UG$i75@Gmph(;VU zHqk-4+oeOL*#`^_n2%Q=F8Ab6CvR%Y$8*jN2IKjJ@wG7zgHfDBlX*X_{jXO&krR_H z(>&^dl6Bpz9YLBT-*ch@F=kYqBImqOvZ7tFZ%%~@(lb#8C?ASW8i4*I66k$5Tw#gNnKMj)6$T711Ya;*=ls~>l^4P-f>%2IT5=+Y*So_dn~Wz+q*!!BCP z?6B`nj}@-t_tRzaKFMQpIjEb|3)b( zTcoI_bDZqrWkA|y0T0>G>;(-=5ny(vvB`M9*a;{5M*Z&UC+BY z1gIMxXiu=309Q)rGSLj_CE;FbXmd3oC6m@b$2xF~-P$~KzVVaqX1!tX4g3GPY9ZO; zz)^#8qeaIKiiI%3b}HsT;2B8@P>aF-2=>hL7YnO-Z8XFkrNDcr&~!>NB&($U?H-flUhJyo}h#WKt|- zdvB#;&i>#USOq<+^Pw44>;q|=pn|uCq^L3^CuN!92huY$@!vAwj=MY)w9Q4Oo~3pa ze0)BzWCjF%7b01zMgZ zR3{YIYn5wAwcn%fr@jlGJFF>ckmZKQ3af!R^N8?>5Mw-=Av*>2X$`U?;02DI-2e$z zsnULM5#_mJ@{A^5*x-IR5a%_m?;h|eLD6iWu`dR@Mb5T^nw7-+UOQ*LA zky8r2fZKg5qucy3IIslLUD89k=qi^dFiGVNWm9XF+nuov3F8HM^a&=9z8lc{md^Wh z$l3)7JOr3sCM#QjV4UMTW;0IKXVM@4?QYMj&Nj6-a9>DX@DtbbUZs#3mF|)`Lvv1) z5)>p9`G8F}cnq>-(lsj}Ggd1M@Lr z`SMJ>@p5pADu>$q?zQ<3=C7E&)i+h3siDsV-u?Qve;CTrto2C=f+AO`28oYbq-)(Y z1;VFccCi>nL&_N;`f0+b4||+&LLJd#;cCE~z~7c*5Ns7X*Dyfm)}W~dIqpc59x zuJqF?F|8L=wqV;m=!9X(l)ya@&jN)&2u(ui(1Jx@Tu23_rhV1Juq~bqrH>;lnotk%Z&h7=6iqa2m{e!b_&cO_o zZGm<4U-U^Umzfo=snfQrDnb&Yi`?zFF&d{qdA9nrhcUrEIT;kneNz5WQ_qT-MEeYJ zPuNjlN*7gMYmmst2*mubt88mjYPIZr>Nu zis7)=NDTGMX*e9$CmwRWZ~q4do09RUZVI{Uz&2&wqZxD@m7nt&dDr z^vGLDDJh3^6)1U423Z@ti8~(T9%}}&&OG|4yvpqY2;b-hbM*Nzdf6Pz(gXpVa?gIZ zM1Fx7;sT(GgUU9TfzipdfkH_hy^}5hw)a&o$kh=mOc$3zrWJBc57C(wM(zSq7i2Aq z1sO>cq$m}2lIqYKKn8OL+&|EO?j);7YQTLcfL$M1AZ~%*D9JY?91?Eu(O9U4jh%Z! zt^pB3CH~G`VR`gz9W=#6W`AYQ{i+rdj^5c*q$p!vd z@$;i(#RihXKJe5k&x#*Fisl8Tn=bX-#LE<~p1aAb1FqDGyXkGw`Mk5@BJj&#`7} zZ*sCt)IxPfyT~v(uHJh6KS|tZ#J>)lHx8_eBNpLl6q`zs%~VWsc$uVe&JMTA=raF( z3l?o(a8=SF&*k+*Hmh%gJ8%GW;NqF{q>e6LaBaR#rANmaC+?UTt2f)-Gmn`Mesf}* zDJ3VDWIP8(ju`^aHbE`_fi!P=DL@7Dv{2{>De&qA`=ctS9phgKt>>-rIpVpU^aB+^ zj;h2XSy3K+MzTAk6L#_4vIo2p@h6^{pnR)Uq5`FDtj!B!%^ganv(F1%+T_BmX`2i9 z27?lu@|_c81-BTL0|O@8$kc42*o_oPq+&9Hs+k(Wv7l=~kD}THSSDBH@=)HRSngR2 z9Eo?tb^?I8(Xt#E=6bNtb>SP6=*b3PEL$0J3|O-rHU$|t?ncfk2^71IBC9bR*+C=0 z5HRv!Qu+i@y1Y0ZmW=U>t?Ylr{5&#RYE03$SN(310THUKia(I2PMkV$T$ckRLyW{v z9L27o$Vw{a1_YhjA!!Q33HM#>%HA->V80waV>Gtot*wu;{eEk}!~fUt^I-C+13PNg z7&&U5Q0(Uv=>>&FSX?%%hbVMG|Ez>Q7kcN76xE+L@H*c~eK&zg;KeidJWAn_E?K`~ zG4sT|U-6l!H?oc{nV-O1S=b86-#28I~<5A4uGmy4lE+0U*IHR({_6K$G{-PO^5O<>l&oA{po@#R7H zK7shWGs}Zk%q~$JCrLBv=o1jG zcsZa<*B%pd&dHikznjy&hc&>@^>Obdve|**x7WzPu4^=yt$+T{oSRp!8N+YqxBrl+yT~r;~0UkHZq!|(!4!k1Das!oPp&b`2)*;1y zwuci^CP`;ZOL^7nXSGoOh>X3uAqXIeVmrL^KtdE)EsJNs-x6l0WEgXU$ORha`mo1! zPSBYAi#zudO@mSn>y9%MExRhLa^1vhmv5JTqFfHU-1_JvZ{b<80)4Dit~EjJ@=RBx zf?qdnK(U^GR;5*DQXd{#pnu$s<^#u<3r>uu8GE4Pd7qqcLj654M`>z8UK78HnS%XZ zkcZ0(U*}n=x*2xP{jP^rnIG6E2iXBQMj`1>6%}@?XnQ_NBaXNLX+$%-A4X?;r-=>g zkW9g*Rucc(cu)$BRxMhJg?-W%DyH`v*Cc6#92Fm$5_l9cXNn-X z{KMS~Frd?>ygFyitOvZCGwp0+^kd8gAj6i%?&yFIcw6ur57Us(YhpXh1c-X%%RQHP z-sdmWmoFyy#GJ|J^&&^>+(Dc=Wa39gD&CU%7O# z^N9tyyglSRX_lsWm-^wh*7|{c$q0;h++WhI)wj7VP}u*tknU=jg)0Byy>nzOw^_)6 zopbp{v(Q$G-9o{b#b6Mp0CH1c*L*_0c-p4$Jz^*`X^|$rp;aD_XqLnEdU-$4l;{HN zoe}nIthr<9SyQ_n{C=sp$Fxq`OzH7*{&GH6R%CdT&Fqcp@T`n(4F=(#dPpHk^3`D8 z5~={9Qv&L6GQu&py2=HMZdSPrC0g6{edYfBKDW&F{DL{XtB#u){fn~se+xAjk=4oF zyGhY#*#!=4Doz=hiYkiTPmu~L=4#Zz**MkoK(b0l1Xj;$+CbA0E2(et77MXzpoX6p ztZAds(+@Du2w$zx!O4(OIT_VOBRyibBssjA*A6MC?TX!j=SeZpRJDMXMHijV^tqOM zwkzUcor1AvkkGQhYhVr7pTpkt$okj7JlNk$R_9n~X7ZTcSr}SrS1g}W2VU)C%!(-k zPRpkx%}EZ}5?$tz!RpWJJe~&afP2Tu@+o)b6}qmNk{)1#?{@5OVZ7h`*u7eR*Zem> zO1#c-BlA5SpLCkm#X79h?Yxmmc~ zeVeaB1=&5(^`ZpvfYS!Didivxz^Tu*X!?NDfv7IgYQcaLaG!Y<_^)>FnVTMf$DhjY zd0k}tqr8#ki?G8SUH}q;tb~ zT0*>4&KC!s%gi}np#Bx>Vf)Bww+_0MNrvzrzyUST8yQ@lVE{?1EFD^=9ReafyIael z(H;k!tU-yzZQCCA-Z0H?dZieG%w#d8GVP$1l`8#MSjcPk)X^wnjuE0}7KDxFwy+a3 z>d%J3^5K8hWI_(iTN9b9@+&h8Yu#lZc&#Jp4jewNGMa>TQ7jmTd@2SvU#T+9#(=J{ z9>sb0kLR3`oCm(BgMw-r&sx6KIUmy@t6Y#a>f<@((kwT9WeAjZ)H`QMu;PBu2`&b= zNGp9bXH-ulYu)UC9Lv!$+H5j>&GvvXkB(8+QQAjV0osvtQIjF=-vh}lcA zu-)H9#l%C>NNHrdA~qz;?HrxV1Htk(K_{fX9-Q4L-R|8Jxu`)_2&V-;S3M3xMG~GR z_?EjYs&&T_qQ^X~a^>_>o{R8d@DX@x9(b=|@S*D-)pR_Q!M`=j7F!>~P}>GK!}zfs z#S|ypOuFOqs}D@M-d_{{|2`v>6;Ny*MY5@wYUL_sv-l|QV8CM4Q~n0wVqu?Ti>f3l z1Jt%QGk8=6id&ezbl&Szkd3RM*N}T2^{^z)oBjYYx69;5J>wZ$IQ7li4C|m}wp*)j zoSIV8WLn_iuzMpj#T@Z`ums^bsvhv64ABx|9xe#Ys}2XA5}tz1i>8&_am!%JJvX`4 z1COp9W}H#RVH_hqN-ws*#L1XUzIgK5KbR(6n#pZ!R_gto}f)tVCWQJ_E_fU+|DH97z^iQq|#XP{++3|S7-q2e`tD*gd*=`vKW<^^vSR|^vP z*CZM6WG3B97l`)<VE(YEZw+3D=t!ZYOqH)QEu4S$Arc1ZPCiZR|=Iq%*$lQS~ zjGK9L_T=s3e=w~mF;lp)CwK*re#J#?ovT$I0M(&9JMxx}^+#LqC6CqIZuiW8$)jvs z9M_thc=vb7P6lgI`1vpYOk!V~pfxaTj;M1@q}Vu$tf6A=xT6sO4&Qbq2AfbKbfP?T z#$dDn*>>dQHU3>jo_%-V7r!-4$bC&VOJ>q{E5ChBvRt)tw#7*A`Tk{IuVScReN$?JO)!eu!o6IF62f*amBAkpipP2Mx z>(dQy{AD#`Fde`Bxc3BEIt2XO7 zZV3iI@f@gCZG(ISeFg}LEF>x`Do93ZZ=rp*?ni7y#FAG6CcTY%*G}cS=OnPS9 z^qpSi^6f184K+L5E;HGx{Q)SzsP_;~%r0kK&>065xHeIV-0G$p>`Lb^u6#sRk7idK z7zPm68L^#Bp;(AXY($>JUGhZ!aVQnjoFYXL`T(-#kf<2Mi7+EU9|M?}vDc3YM+9&k z$ZJWZ#ZJO(DfXDt|7fme#Pv{36~K;CNrh)&y!*&)aE-p0oMv zGd#i^f%L-kGu)4@f2vpf>QjFMTzColzN&d_g-9urK8ED&na#fLP8YW;o+_H6^JsLjv)?`8s%-fGqW}itC zXX#I0bsSI%S<{JjY>ehnYarHc0+hB8(NjA1llUDmJb1 z65FFVL-JsKGQnoU@`Z=ZhTFo9xS`@-cKvX->5Rf98tlLx6*KeC#(<3hS|!dvs2g!z zk;iCOdgiI}eDH<_tJ^0ON?eSKg$S_8V^{v=7ra+p(YamTJI&ZeCkgd zHZ;B8^*u$FaNE#0unoyD+R$vE*aV8K!yqAMUfLbWx&?4Ou9gBiK_4F|&5t3rk=>%WR5;j%( z0CHFxWWcucMl<_9I#*EMr^*GfR2r#gT+wWs7B>?`w;G|D9g_Dggp8@ddEwmF@L@iv zDKK53a)FEkJ4Vc`XYYnx2}O^)=Afz@=BFyC_{yQHMHt?ha7$t-Dr`OKFYE8Zn!o(y zi#5BHcwBMOlrSwU7MqtwqUv6Ta|2|-PdFwlLagh0Z0jg_s^bRNttyCq6cl&Re0Wx#pWgaJF6KPyH zVZ%n_Zb>)4pZuL`Qw%#9weABAS^SD+k)#{#SJqQ(JVn-G76}Ga*70@+=h0f#66Z&9 z+_y~Rh!21J7LEMy!xQ+^yw|Kx^ZR&b119ReRP$G|?6uk5ZZ|?p62-2kNIVsTp}8S7 zcl`$UilmtrE9{ViZwy1c6O0;*05ddYOdwj|qwYWX*0e|iEWTHZ=k?X#gbcj(p4G7^1uPUC{rn19NY-;RE)KjH*=2;UObS@b zK?N%2=A1-c_PjVqM7tnb8loksP<)lo?BX2}X3xt|_L03nV4NggHn${dopZA~j?{(M zi`F??hYQOFZ(QiHtZS>DI{1^Rd8Q>dFJ=}wFcNkcAt9AwH&Y~uirK~MBW(gy&`1xc z2eQy+bvsjs#D4xy{m**tCCm87g33f~f{7S=!vJxEfK{W)sj+zj62^SzuyT0Cdg>Mb%GB0qX2B9LHVfa&=Ps+)R(MrYyy2VZ^X+kkM$C8g8rlxKl>FRQS)9q(zkcf2L+3T1>#gGCZF^x{x<>T8G6h^MaulXl^6`q9;b}cWaSi+WAuJ+ zrr0EktfyjX`8l*saA?K_CO+s8(;3mM-VMayM|t33GfxmW0p#CQBXR_=-Jm zGyzBEzwB3PN{ZV3(fqF)kTJREKRzYLxb1@+*q?vH2;fZ=dyXP!^~|Cz(!+sOe9bDL z@56MqUGuBF`(1PCM50LzxKA$@=F%%jfjE(06ajqqhk|l>xHZSB^!xM@($9PHr2_HB zP^8Cyf3rBBIWL79u6eGda^PSPT?WP^9bM&;He)wPT%Mm+3fx~gbY-+=t3oG5m(EFf zhiAVlv~6=~bJ1Yr`3ms$55=7@FZ&9tye(Jco%m_XvUls^AOH5b%z&$t|M<|$>WuKy%5GS#RE6Sm zA2xaRM3#uFLi_2h;r8D+jdR#78#^nW;${`&`pR#bdi~8f(vj(}#G_4s6))fqg>~ou zh*K248VkRAE55|>HRbfxJFeOOo!_)rYodgX`4Jlmbs_8;z&2!k27YZU3EHKKQo(?6G<@Z{Pl}(_m4cVXYCI{+HNYQ$`zLr@k}aHL-*5IG9Krg$;%{9B+crRsxy-K(7pb3kjD5#^5O8K znY=h~-ot`Zb(2?37t`s1@!}1n9||^#NISuu+*ZY{nObGN^rlOdjX9Pq2hOnJF#B^X z9B{s89yoaif2RDB0XQvRzm-eMUNPXnj%LKd^(e(cReCiQ(?c@YGD$9PG1EVH-;7qp z=ddx^9k^n4vSKl{Vs@iC2i$cZ3r~@cZmJvRH>sP|-3ojxU!hedEA}nGM{*61)XcxB z-m(CniG`*EPT0YK(=K>szzLsB52&NFB=JlVZ~vT5f*)e-G-X2=8RLsHaEu$6hf#w6 z_fuc>8o;)&q6n-sFXVO8Mee%;ulSviH^|PoYzXh7E0`X#M2bR)8XatPwaUBzOr}0O z-`b%z{(?t88y6f#@5-iE&BI|7|IehL!(UzVQ-h_My;wJsTyfy-@M9yZ(nYbiDbhj3 z9DVm1Og!iysS;zL@I+L-?5RJpdTQV5C7WjsIAyR)y_S0pIIU1;ito-y4~S=~c~2z$ zym)@`v>x7?neA+|8kC%-wX>fJ?kg&0qK;UHp*3Eu0U!Sz{#utVdSM>j7gig!j?M8{ z*sRVX@w^@=N{2Eb@d71lk&{R1J`)OWL1c(u)16TEt={M(KQE8FOkhAuV`&u_mm#klLSbRZp z&$-~ud`XqJre3s-x6JEl*j?G0_u?1q{#yNfO$ymA*~|`Q~`N&CPg@ zID&Ox;n{N%1DNE0`*I^maNvzmkr6!7DHawfDO5}^=+?#x3nTk^TS3_gw-)*S_vF{3 z(*4)-(O1{Vv@1&H|3<6C+j+G4y$8qd?q!dUXUt#Lr!DStGdHCzQGb5VfRNH@Pkut` z92g-wBZOR_*hY#pKvIq0r>;rj%IH4V%IM1^$v2&@rfdCn2Ied8fICIg;GRpzu|q_E zJ&~2sTICw`o)CNHQ@bQ5fA(#2KYgH9MXfWiRDMwrcXbb*Uz z`I~r%t=gG-8A|5j1MLhlJZY8q;GmNZnj`O|A?T14T^(8%)F*A^vQ-)nvc_XfhTB+* zI!?%%^nds7-S?{Rhp;^4j;AJHa+0rU2JfwQf+Pn`M#K_8m^7J?Q*MBG=Zw1QR|Zpa zD>~s#a?62D&C1CmegB_RY%fK6sF-y66J?9^v@*lFpU$3Fs{%c@995b}M_7YwB}ff+ zL7sn_ho*zB7Q}&`_JDk!b2j+QhO`xK@-VS7kM5ieIToKVtx!@_!7GeRopvoWkKPrc zSq;G>y>&tPGR;=sZaRbicaH)2`lw}Y$Pt(6)~?(Snr4~|ekEj4wF&T~OHmt@!B6we z3de-0q1*U7Ch|8ki-p-k1eJIZW=5CT*$uMGL7U$MX~fb<)IdBG)FtYI>ZEg_7e%OJ z_m@tR!LJV8uc{UtA{u=MuZb>Vf6_M^P1P^CR7X2-POljePt^8^7k2|7A0yGN=y0GO zv07!mFyq({do`x6H2}R6BKmi%iwvY^~~mKh{gG2tDkL=&q(? zg*)9F-D8EQRCt~gi!`0Gd`3sFa;}Un0PahCq(yqhORH>j&m@|R$c<5&ByqknN!%mZ z;Z^`K(mdXM*Twv6zA#yO_d$Ysl?#4QkGv2xqBW;UF1`LOoS*ZdYD%+Jxd*0JoUApF z^~c*7+6K(cNTvuWN)@~dKm(woO9FC0S+rfTJJ?~tnsh}PV5(R5N+ph)QKiF_@s$w6 z!eRg{7eI|Vu6rk&%v7)54B7G3k9|#}E)5higI_&LC8*A4jc7t(;lht$uz}hb2Pd1F6ZM@ zvBWly?srXxvm1)LGQ6!Ze89)5_39m)>y(ImXfXj2%R!jp^5xbj7kk%zw&*&K2g?) zGd{*Y&kv}<(V%Jb)B^9&X7ks<{ac^pxqW6S= zdw3E2H^H+80MsH_PjFc#zBgZAz4Y3v9#h&W+|G6~{car)7{HKqmuOH@NjhYW?(2a@ zP8XnAAKA$ih-(?FA4JEVReJ}^t&N3aZnVL$y!Y`x$?b)|wXU?#G-KC{EG-T|<4==7 z$4k=y2}k#w_mU)DI^E1x@XAC>z{hkf2tz0+UEJ+@kJa39$9kR#cp1kt5)<(3UwykE$O=;5PFM~N>V=T&+eOz4 z_D7|KT=hqP=vq?Ce6X2n_|ET;NNR&~uIe7adwZD@x=pZgY6%oUpY^;djh~z9s#SKA zcE+Y}#rW`VWSE|vjNk6O*z)&^Mu|0|mkzudF(Z13l7?F4(tv#-JLsZ_Qc08SGw`?S z=)~w=@PRB5VzImCCa(!D0`>Z_ph6ns&;2|ti$NnK_BiA9MDjjvfsT?(^X#K5WDB^w zIM~)->HDXz?VDj(ye|8|YaK}+tp>?)?Ichaj*LU@qS!);|LRF?VZXwA;R;t0%kDe)-tDQ6p^HE3us>IP!o5cJ7$+wGzOaZJTI*WE5QornzM zW)mP8W56b7+P~HNeX;!SrQlzUrEVxjXMBv&mo0DobN`PHh8iZI)ydtvNzrSQ&U?yeYO11GSin?JG1oon z0_x>U15nTOV9<*2Dwk%@rf5i%i5FvTz)F4%XvP5bOL)8DFAFe==J+xlU&vz-= zrKz&eD;2{rqaU=x4QMLCLTE6j1YduwG@SV2p`k&u6e*UDu&tgmkCe|AhAi?jOgXxF zVb{o3Zc~oqh9XF%(Uh~3V!`6%fNov*MNzX$${QJw<((Sv(FLYDa1C2EuvuUsw8 zsFRAVpkvg}+d;<#rXfEro;U|~sF)!;L$*OophaP!F1n3(pScb>hV~hCqp$jf6N_xi z3$5&b;k~|nZ@lW6+J8n}DXdWkcH>W!rSf(qkQ#4SHMsZi;@C^PZr5s2y>q7OC|^?@ zxSiDj|70;_kv-tS{ugy3Ad9;|e9yTnASe9PyuCi<@@|lFXl6C}s&j#{40-{7vNPOfrk3)2CW?$BPz`Qa?gl^ps^Sj<9_EqZVA z=z5tBdM_SI$Hqbfj5#Gw)#Rw^pZ#K(>iRxr#g#{;%&1j$0&`hs#11B&!Drv!DRJ23 zSMxK~C~KNYb-Wc`rokgF_Q-)}FEcT)9^OHyJ<1O6cg+Yt>xtx38Q}#!Yrz-Ou6RVU z_-)}i^eN8_=Ukw!-y%x)&GJtTo)Bwgk$EzisqAl@CQVvY#B6jmEO09S;k|QYtpgX5 zdC3clBvWv;gbSlJ5oskke?p-6ZqUzIxTJP_mF6M#eiP*AeshCEe@7bj$} zVcky1L}fFVDk`bH zlAg$mUP$^gQOMv7jtN2rS8z~|J$=1Q!NCPx4jd^m<9xv-O21omXgVlmV#z8lBQ(hO zn;^TR$eaEdaBxFu87mOuarB*4AV9!-uPkcrg%XPo#@;m z0?+DY(gE?}9z`3yg?B9SfT}8@K)jg0A1D|Gow^hkrMYmoFw*w%K+~b*3Wed-hG{>x zJ~uz*oC$w=#Ko}q_><=R@5w3$UVP*lEk3qTY%)bQP%*fRfUOa*n5ID-84LhQ;-PPS z&t{!+{+pW;Q4}D$EYhDi1&* zF^Xs+Mu4ydBfzLvggKH}RjTO#5<$E-VF{C*U^8qwY@SQ-B7V%nMzO*t*xO(|RB@{h zku+|;SqBEs0V7*eM6m@FaI(jgdtwSwk}r5Ouq63u*h8q`Err6qhq73gRg{iKP7X|5 z!EB{=#T8L;M3x(7Fu_w$%AxoAAA!=~7HL|*i9kE8c*lpNXOMz}7g1n~D}4jg{T3MT zGJDbA+ep=GgO?T~yqu=klN70=Vo-xhzavIP7$i)%H4A2_4ChXQdtyY1M(dm6iPG)| z0zug&2h?QFLe6kCjYnEvRF^@>wbA9CS}8mm7rxCe!?`rF-18K{#|p%7!&ie;2>O!v z3xT=LnrxC0ULCs9J)2HY3|&|v%%gK7@+0;z;|MTYkEfB?GJnh!V+S6l?T9eoN#r~~ zi|igPAI^cT%o!up)KDxW$nB?ME`sMY8`iU6P84x$vZ6+H%)Kb$ic2DYySfiXT^E5Y zWBTFFCb1z$`Te|9X}zoxo=@W8y;}Y*(7(%sm2kES1nnYAh1rYups@1m0!R5?+@o|GQQpU>liM*SvigZD-WBXZ1gQv4YI^X7~9V^E$%Q^vElCrCva3ksF@nmM@uj5LtpCyRBcx z0N<1P%ilMpM}DPt%Vq?paE03F-tCv+4Vj8rRF$$ve%5m^ES^6g*K|kaGl%^yiuTHN zQ4@y*3lT89${b_NYi_^dPXjnLWzKU5r-Q z=GMT|U?yxEuxq1)Wq}Z?Hf&r5$q^#MP%-n=q^V)ZB9xM|+6 zdg6}GeEV1#8^FbO7#YzMC%8=df%*^sHP>Ke{FkoXN%p-qJDKxFHt7V#9;3)%DrT{; z(EF*Rg&O|ZtUgK}k*0S(T(Mx|yRGw^)fe8m@sHQOHsF-}?`OW2_Kg9jm0wMH_ZnOW zZL6WCShbouceh)D0+K016BH;=_18s9C}_!mw9%Xcfk*#oQ4=&Mnv)s$VaYqq>Rr74 zuV4A@qCegjoMP+fmWTsT1=|vlIZj8O4S;9HW<9$+0#246URxc1^5vP>4(wMklZm~K zS1f{^&{VBQqY`LMyTExiluCy#e*2gf^WV9}V{g6-yWBBvbpHCQ1qUJx2IhMeJ2ObB z121D5j7-K6imjob`ZT8RYu6-=>R2fJyG~z|pj2HJuZ`dm1?kt}y1c{#>_hk+NeNu3 zqgPIO;8z~JYD%oIRMf2QimHPO)izQ?XTY=DeDfk>g?&MHfH?^7=hE$p;pf5DOgSso zGK=S~nvyOq6+I%>qIs)m}G#QMEsK+zLk=8s9W1Ec7&5sJ=G zEL4e|pkf+isZ1P;vb#rkxZs5pgk(rZN%H_oVklMyCYG@Ek&l7WpaH1Zo7p=dTf;M5 z_xm9`yH?pts=)|Vhe92Q3X>KJgvd^dyk5zQ2HBda=`^Ntg8%@ePvRZ?e>s0SAFr;xzy)QtySoB%cXPx6)9O(!ddT_&5En6QQ{%>${7 zSGj<+uP)+3V2TR3Kvm-i){N&HAKAuIRE%sh5yzZRG3lh@?fa%okuUmV95?~a3?HPP z46v6;@U|C>y{(PXkO8unDk}K)6^Shb*AZ0Tykq2I521?F)EvW$hNb) znJ(E`&nlOzs@tM?MyovTb3s}fxrFrdPAeNA9@s5OR`fy)I-cp2)zS7u#}Q<#Ybs@iDX&$gW3JAnoZKP0FpK!E!Xt9=KDSH!;I{5 z&$#f6@Z_oET$gekGDpQncwc6?Y3Z`Vu1w7oBWDFR^E6vjXT0)MAT<@8>UwG3UZSaT zS>|8puT`FTt6qMeU**#7a~vq(5_~`R?2vDlYI;!Dl@erugOBhjLk{& zPWLlLyo?#d6 zHq&>0Tu3nO#17}?XP>~pHC6rwjQn=j(FBs^z%JNYBRrH+ER=;6Q85|9cKLy*SYb7I zN;CKeXJ;t8L??rf`y2&z)oM{AIQE;=-SptM8|Pn>tl%HwBkgRBbU+Rp8v60uAKrAo zDy$c5nxSu!$18@^gh!-QRPJI6Prb=#jte29m<^L|tv)r6kWtnr)V6Q_r1!rKsQG_& z-=9zJPaz3Lx`K-*vjY@)OvUUBU&|bveIT%ibkVUumzBrdlimTd+&}xa_kcW)(ctaF zWCs*&J^CgT+xVR#gHA^YHmHHAA0N*5Yw}3p6`Lza(TCaA9e-*ssAcU&vza# z&Rqyo7F$3>38d*uY_j6o~8~i4mjZv{$2wwG)BgCe0Jo=z9y^{#8G!CUz)aO zrK!vdm$i@s@3_qfz_dsYs=7&=;HXR5)Iw4&g`~SDpb-`;oZ!_l7#j=CnAK*9YX|?9 zx$n28u6{=aFfr@azcs6n!m11`Icz0&z_e5c?iCkOm)veT7g7^E+7=Eq%u!qi%ev<| z%~+HFW%8^)I~!)Kx-ZrIl`P}7qjq3x0iOL4uKpy7T~CpCDyAr+MY`6lK%B_$qL=fL z8Y~;A!Spes35Jlx_F{q|G-=W|=l^VQo&gq(KU|hWDjj(Fd%vX!g_pPzQ_o{(f-eK_>2tcSku z^Q%Rr;tCb^WDTS&H!-8EgkN^a3M<3O#88*t3po5$1D1Z%l=@wA#DTHYYJ{b~Q*1p& zPQz*)R>JE^mnbhd$#;{>F2Kb)_dZw5(ar%xWCw3@xgu%iRe5VFBW?!g(D8f_@Al3X zYmo3Mo?jzvX1B2Yiu2Q!2+z`*lhe@qQsu3a_bXBbkeu>HcJOW&ELVj}*Y%Oe*b)~G zEgT( z!(;=hmaPmq2D>9}9XRX)o}?LJFo9y%QDik>5D58~2!VBBy<4?!vI1%PtuZbfM#lJ2 zOYmddVq(tv|EsM3RZnt)^ndX&`3+eiuTQE~A}PZ4A+_WB(1{;vOU8M^SGNEqZm;>r zGr(PCuoHWnGES3~4m|th7`^$MDK?2B>#3MeM1@R0h%U4VYWWxVH#|0ZHLEMd7lGP% zlh?^v=Rp2vV$oqCGH@=Km{;HMDV&f&eI;k}x7`iM_+D+=1+vkBkpUHEBQ`Tx6brtP zG%98({~oMFFEC3W+XMqV&2QhIn!v;S?F@bixOJP<_=Hx8S^byh^@=vlSjtb5U7tD; zG94UHGIG^mcf@c*$*<%8tKS;_KnD&Znc)w-4Y(8tvbWrlVsN3nkHq^i)oI?zld3 zK2!&V%UCwH)#d$GDB(y}Tn1sfy`&xjj7U6)G+QEh;0o3hP-bnPc zC7_#J5$%t9cHxRB$#>AHL54)8mr0`%`^kec5FL;}X_faQm&tICY!y7mfx3|s*Og}ul=xH!CKR!FDiKMWJ=!T zu+d^Onm_0}TO!O5Yc2zkS2L5Z)=Ad!?D@W%{cbHAWwXugcEkL9GD>RGr0ON-KZ-WY zF|H3!B$EmU_Lejm%^GzS3lihtU%aZh^9>^JIq%PIgVp z!zHEfo?uO-1T=XRkom;N(!mDVa{gV}QL-|;U5N?Vm_VzE6&`wHHBe>b1t;>MF13+K zkj3+>dF={pp;_s&jNc$D5bLBhK^<>Baj&K~PVEBX%5kn_EJKpfG{W&@d|=~-q^|u-M6pOY(?xnpzWHDIGS&s2x;iauOfo0MghiAlpu?jvgn+4ab zTcl@IMqeXeI*?S(83hG=wby%Zzkec;V|zkX_({rH-s#G1*K#Fa~4@CivGA8ig) zw4;Yt%j^x*3DTl+!?hDkZOg&vjCYL3)$adi?@QpCO3(E@;yEN=4A}@KM^J$Pf;h4m zuHZ~`I^Emb_IA6S+x|PZt@QTJ+%9u_b8TnZTiiE5903J1fU?M z>x!ME<3mSn`N(Y=Wy_1kVVm33@;#qGgQa_F#S5&2-i?%?T467_D{oa|eA+-G(Fih0 zJI=~ld15v_eK5RME_c{&ZcpE7tFqoseAm^o_1J0MQ!CBHb@E)ghb))tr(#WOgX)|Y zxShGPZbAP}l5#`Dd*h2)TLC-XnB@f6iQnrkqNkYRBF(Q{e3v9Taa^R>EH08ovB0&Z zqoP}tkD!L=Q!%O;ltgq!mddqAg5M0m_Uddkiguv`iah1eWQq-1cjcY(R9Kv1 zf)3OCt#Gi<4NuIaQ~cWDWMn#DO|6DdYzy?KUV#0{knwdCYHWjsrTG}LJ?y}aeW!J3 z3=@K%_}f*feoE6+vuQ)nC6esKQ_WtpsV0YF!6(U}qK^i~13s{JT^lpEctMviolXPP zZ0Fx+F`kaNH;nEqPC%LP z5F0JSU#1_~nxY+u!(fVe3yHqS zBP-8=f42avbSmRr?*`nPc1~_bBQzE`4Y59^kE4Uma93G}(>p&{8}_14Nf>Nj<)NPu z&)2E$(Y3C?Yawa{O}`4dR@Nry2106`s+M;N$f{d}dGPt5x}2$pcB`!(IbpjNtXFn` zuma})c7z>r96Qcfp)uUB+GC$}XbdwYQC=&rds`O3a*3=uafxG|nGH&#*v%A-8Sq-4 zRO{3+AemerVxS*L<0UsewK#3%N#ps=vKY~j_AE|qd1Dwn?pPYbmvOjM{tIRkWtxYc zxIZ}y3?XET*<+2T*mV?HO+^D`CKTbcL*_l5-U~c=b~M5qeqkF@sa)=6zjHagmCpN8 zslR!1O=))lG|ePJLa zb7Dd6rsa!%i(U7a-YIS`#CPAV`Q~p;xO;u)k!j>(C&t}sGu-u2>_dulQ_;pD1_s%> zkQ%=$P}^2IeJwaO11{U$b7vVe!1PvVBGtPjFR3cUWmF3f(gzlVD zf_(W%Lz|{54#uFLPTim!a5+WlXJEM!-WG`sw5NGXMEBTgWrzH-SbLSJ0}iDb;1(wd{Ki6rY@s4d!q07po9xNOp}9MVPNfMUV`!#ON9=n%|=_%+{jjEd!mnu z6%$RWgqdq+YSFpvpzq38lWWp^$=b*o3FaQb`M{K;F&U}F%)wTljbLzma182M*@ zuC-ya!+uV_$Arm(|BSMX&p0hktVCR@K(I1BFwPyF2CTNJ71r?cU=PD)sx=hOATFo-+{Mtn7pqphu%^x){aaycFn|_?v5I-jbVKkdxo~_Eqx1a9JcL zc0hpzdq@^(8^uDa*cK{!hx=vbU;x(W0WJ1qsDQ%qe4JIX)p}U8UWCT-Y;`T2AxWAH z8EWd(Pak|`8?|^F=sh{y=imsO&nKKwV3gA(*3HVUpX+!3&;*iS?fG>dspICSI&s(f zr)KcEL9tgU(o98XQI8f}U$Bg6Rid|IRCU4NMQ`YhP!Kv+B6C4JbD7yLJ>-wf>qouy zGukz$#kb_Up)Jg~4Pisg4PFL#H)|EGOuyom{G7TLnm>w&9$47%Zr!|3|Dl_Aifn@_ z3jF+0zztFwwcTAWz%3x5ly~#+-0RXF(yyp>tLEcxQxk;u(+iR%-H_DmSH!wPWnXYP zaCq1P-O-}&a~O{1gAF+J{*ON> z)L@Qiqrf0*1`!}~k51$roVO1$9w;%_EK8RRxNMf)32kt%QkR6|F(}lO7x6N{QQqh6 zxV9i}E8$@V+J&oBwe2^alEa?3h9r*(L()zm??6!sB*M3=cY%P`Dqlp3_< z9Uk2?wtAIqCc*0E4m-dJ6camF`j^i%84wqVx|C!)u>twe%z*5pSSZ>pqM{R`_DHHh z)Cc9|vIhy9ikYhns&e0+vsY2Y&z;qzdBQIRvyXAL;~<`#=DS(8fsf^xdlkLXeiwuA z2_KK`lQhUrt5T<=`PQlM$gU|%T+>kpjAZZk2lJ04Cy89YIINqJng103c;8TPkg zSAip^CL;(&jVU{0$nticmw#CqZn7?7*9F_jZYQ=bXU(ij6~*c)auCWd7Ut5qvo`oA z`riHemajl}av^;v@T?Sc#Q1sY)F}!4q_EC~11`r&3c(Y5L*s+A_@R&&8(G8y)>`C3 z$YbiDZfd{e#ehTSytU0G#jZxiiPl44i$~12hB2Ey$)7d?_rvuN*44B=yOEds?^U&mwc?= zFW2V3naS&b6$OlvPL0*s7s3a}sM=4j0_jo6_9npqTyapNxm`wu>m#?}d`_d{ypfIb zlUslN_tpJ=C%9Ia5nK?tj{t!eETbFc8Nu5Y76dnG(E8|9e=7)X5yDRX2crFgO0va6 z2ZXr06os!oB3Z9yN>HIHQ#~*@2X0|+SLbusjTQ8t-^k7%;?iAKSjO>Q5--L|gb%Yy zIPok6v@8*6)0CZlHIj6CRY0qW z=iYa$Cwx;=@^GQaiJ80jd<)S#u@lp5HZ`81*pn1F21}$_y#(0(T;oIHnR`=^pr?t| z8t6nxSyX3agKDr!w}q~8@0Ar&*zOoB(5ZKXX|b5@U_g>`m!jUIGU$MGwV;qPfK9=? z(rp=c;@w()dG+BEvywbCSE~JdN$d|+eA7Ka3mL7FA_CJNz!NHix zqX`+05+N4;%X2q51BTPwIblLa%_o+1ElwMUu~Mm0%G)bWA;yA(e9@E8Jy33^-9cj= z-5J&90EbN^!w-$^xETI=NA2Z=i-~7{mn5-dS8!TfSfPDLl;o*WWm&Vd_|BKngR(OYR&zsfX5;ok;OuWi!Q%1a{R8C14L1=JYPp~xG` z&lJb$^68B}w|!c{5kMZ{^69yPdYBI!{>s`u8s-MtY)Boj!RaM+-kezA`d$rC>3G7rL<}Xb9M7dXS*ejl?9#17GZWwm}n62D$L*ays`6d_E zf7!YMa=?jQ*b8PZ>~V@cLXjFOdQ0G1R?o-g#10w@<*Vpw@7=I?y~vdMH)%FWig?wL z8~C6-7<@)p%4-0@?vE5@QPuqJ2tEH8{Rlu&C9C$n%$$W-TbJUTR|W4J(*`NToii^4 z;=OyAB8b-J&Z^)61I+X$4W64Xsb=0uA@4wy9o#FMZq$syOmYKJxjtkr}a{-P{8f+ydt;{@ja}l>>*f*U{qv)5U(7`)5kE#HazbyoMJmLzHq5S zcW?s6#M4{m|BJ85#uN_x^L~=?(v-VYm|3LV6uXlm`Be0E`lj+ey_A#-A1InMABeK) z@~~BtN|}@1nC>cHHm}$*1i7{Y%(|U%1k%<9@M6>^$efJ+pN*vWrGX4+qYf#stD)FRij-5) zC11KGT}s_l7J`6OlO{Ewht^BB^N@_;o@Xx8Gbct=&7=loLD6P4b0-wrscU(6pmypE zNt@O|?<1)JZ8t=}a^RHDFtW7Q8r2e+VSe-RWqDP#HY zqdw%)Sb2x+0ixmIH8yCqp35(`q<-NN!*yPlo>(D@IW4~Kxl_=^+v{CMKj9%&Lxs3Q z*%fi!6A2pb(|Q5^?4i4OcobDKZz{@z42mKiHjOx%l#El~kU1d6288Ebefr;Ke_LR( zHt%UK{E@8YW^J4|=>^Hpq1I+I#U@eY1F$xObeqUJg$^@HBsa@y^X{@rlPNqI(duxRVI(w<@Wa5 zZ(|D)N?|7{>*T$%0;XAZ4m{h-;;e9OgQ`_^TAa$H)5}PoTUBHNKa<{q4P@oQ`)+ZO z*v-5mXuv<4Zl_z7TU-s1k7%tymg2X_*hIew?vBDR24RaZn=XsOP#>fLT2;BTR(oWJ z?F>)gZ)37Ov85R`x{bj<5P+x%+{W7GS#yDjxdj>ZmQ85lRQpWJ%m+^;eg0`-)*Q{FqzW zFQ?y`_`L|z1oeZ`f(%kJT*HwQ7gg4oOq_J}QA)72x7frAzc6(>$1_ZU#CjtzOU}F9pEfGh(g(w#d!UKM ztTaFhf`RxA>GVy{JRtpFBHA-8d0KM_5|oe4AlU~+Rsi-C5~KKO9YrU8k~#P9Yeq6E zu0n+qkThe_0`e&gxZL4y3^9IkfN>dV#B7JmNO!USvBgWbh7Djj>f3V?NUL`j zq=`!B--d`>19L^(D6gXLD{K62h9>Zh&1*V!QW((r%yI-kY)63=Y|#~QrcaxuJ$RF3k0%}j8cg*e|DHLWe#c;HN%bp<1ifbZoL7@( zLqsjz0J)~qVh|6HN)T<4;n99YqbAnt3^@(Sh7$z$G{{dd#!m^ZHwfa=_^s`5?Jao~ z5Pm>Ei*%Q)(W^yx2BtIpjDAJBszFxl+3NkozmP8C?QlP?&_D~9VfbXv8N z4H52r!6wAq`p%tVlE*C!=*02jV`jiFquBivDWRgz)7T-`NB6r_^IH`qbQ#qxi{+36xx=wX?*mh?uHQ5{S|6r3Aa3%Z}RiVAU!@G`RqTjzAD ztgz+&iyAby-r5he(g%LDe?i8V3ct1=?qI)n!9K6R=f=oxb7i3H_172^c7A`N=QvsB z#B&}9m<;igHd1T?MdGOFt)9T|t-0~tYf>x{TS0C~uSzd?42XLe`;DKi58TFaw)J*) zJCBp8qsspyGfrwkh2n2-G?4W#jYWY(_Ye%FQ!H>4r%=)L;#xTf-2jQhMW&ZDX<|fu zZpRhIJ;x(UBCv0^Lfq!27gYGQxjB?9+-%Hk#M9@70X~>r@2HLTv#)C`Tkl_zP;iDB zLO!6_^%RMvqOpl!uoDld3sF~5i{Y66Q-2$Ru>JABD!1$dcu7EErE37=eEA|Ry3)uI zuYDGg`M>nQGhlr8AO3RuMRzjr_1CF8A#!^}woP`Q4^&b)VcXsFq+<)YK6^+UKG`}c zaJZ#iPUYleOt?}1U;hzeax&H=b?zoRhsyvs@#5r^*+ikISm@#|rJ_Fyz8uuA=l}x4 zTH!U}?#QeDhj|YpYlMe+x5*AD+U+I9%tqJML9iH^-Kxm<$1|Sr@mRmhvnz`DJKT-? zjOW(EoeCq0_=OQ|^Y-%ap2p+$a&c|N)KE`j#Lsp?)JV#EdnQ$WX6b@_G3n^UIcF=5 zhJh|pB!E^L2ssb>8?E6xXLh<(1+5m?YpiV^8e2hO^9}8EqIE;__JqwHekP;R^E3HM zvdxLtvsGpWq?lrhC{jp8ulU~XUwpJ6g-ZX?&4nu_4c%HfDM|Twe#f*tD1_PU)gD~y zqf?~@{IyWl7=RQ}Hc1*%`!8A7%KWJF6b zc+gJBFx|#FK!%CzQU6l<$9XS0!Sdzl8Pd4079o%)DB?r*`5&XP)4fU4$BXADDHq}4 zMP+meC_Wg^)C*dKp9Q3aA)!KJ)Jg}l^N#^1_63H;-MHZ-l{@D6*M~HrB@(X~uM_8o&KsK-(WS*b9q=P$=@*OSDzOV|0e=E#Lsk z7ungsSd7=x*Bh!q$-n>Z58D5F>$ku7<2%xo6uXQfF(X=;;qPSMy~`JMZ?n{5d`T9- zR#X~s`3uD?#=hteA??^KyMn4R?dmQb%D`Te-kTc_J#G8ckAk&Zro_z#<`y)Zq}6i|7@>G3js<8XDJ$s9;yt`UY6Z+;tqy9LocH~fZ`ui1W3WV4;AI~O z`R|hFN11Gk+k@jtr1S+_4KlE8g&4fw`qgT1QyMir(16sW$x$1~Ue6fOGEswOujk%K>?^^hB>c3;GnaQo zfPsE&R*IdHCH>-=S@h5|#|9kxAE`Y`oEF5Nl;BP)Vyz{ z*i?#aLVZI_e`(Cm}z+y`8WJq>hCU>y+ap=dcaos7+t zjD$_@nE%|2;D?cJJ*Hy`zgqeF8z!*q^keRmDkok~184gXC$^Si&rswf)X@Q%DOxsI z49(vu_}dcE8vk~=LD51=r0*?0`}%rjJ+ED}6&lcR|L&Q4{g(wBkG{Wn|h-3 z_#V^M*u^;9@lQ<0ZzxLxxAR&Ee#0^P>D@Pe3GZ$@sKy8qGsM`iVKZu4AZkf$P6-_O`0y{R!@U)b5xQNsX8mwHx;Fld-U${D_P(Q zCY#YP#uL|a_O2GgR6z5qHK@r zVgC#H7+d-D$M|GMaXF_q(|KK%_>);5Stbx)s8GL^%B2+WHH_SY`@KM)CM-*hLcdRx zyT?D&Z)^4JC_p&2ORPiq!F%$eu`uI0FWyO+nZYZd*c}wfrlR+eE@*h}P+pRObRu>d z83SW|~*T_r<>>2`|mE2iir3Oph59yOkoTRCE*A zmRuSIW=g`_X>B!MKV$8@cJMYZWVc$dRJ0XnO&gbNzPWa~>~OAeXYlARO_YnogPq3m@u-tTRAyJT^7})}rzHnu1l^Au`MJjA!y3Y-r;>%5^XAsF{82geEH>GwwFCLpE-_S^W=_EQ|M@c7C#2sE_SlS!x)<$xy5hN_(t9 zb`%_0`x}MlJyjcr`FXdo`*BXQ(!^i;p4Gi>GBJ;NnLi^HFIWxSWivx^mSUkr=r|R< zcHuQ?3~(ejX;Nn0XOIIq*+PL?H4L&99+CtwGYwFz>dt`5H` zRdzyD$ZPUKwG^GIGo`;#m;u5@kd0_7Mm}ivbvAEU2U*VU#9~W&Fe^-t z##S3(TdHQVz{YFKq<54kzqy$y3(TF>=l00)dNM=*!(5*A(CZxi+4{2{CV=a_y;q%7 zTceE{taWm%{;HEB@mi7+Y1b;}ZGkCUn;LMRUIw&DH|ezbdch54F9bacWrM*@a1o%a zj7d`F2DGwe62lB*x}d^g7~%q#VP8Q9;f51joYzB4zx>8eEo)Mo)?2YsEK?)72sx;G zLj4SFo$#LP7WjY#`s44b*jQ=cAViLETW(2g_N#VK`P!Jxn*-Lv2qaF3p-z7Pm1CAwOk9F4PMig@Qr{F0DB7Te4%GwFLy`4? zIqG)dZN+veCIrUa?%0Z<5e$Sw4xN4Y{%Ola4;NH8@2p9zEQhio=Yx5n^Rm^V^Rg1i z4xV5PwDvUgjU882%d+X~Qmm(F@>-*WsEG00<0LKYKyZ~>dn5?dA3-EWFW5(r1NqKW zWIRMZ<%*DYIwm+@ z`Y-_3wkbh%AzVhu_UC5(Oxs7=`aK+Rgwymhq3*U@`dFmlop}1OA`PF%yDG|*lu;|x z&Fs+64Vp~ySeh2rrdhP8NpmL{KQ)2EEH3XJOHn_TgBcjzz}gNeb4xM0Z?pX>(Wb6V zmI7v0l)w7@x`jy)**ZqAn3VNuS44)WK0xpDmw3rvR!&;9D4ovsO!I9e$%A818oWo6 z3`z;9vQv=GD`u96vfMgpvp{}-`5nmzBe z?mJ{D7|+EQc4FXIk(b8|;;9)OkOpkyRnW0>*3N`fAlCEO%B%PepM@;O$YxkrT<7S` zZ~y4z9Lq5N7u$cFcp+pZ=g^?OK=zTnP`s1`iNthzg9i2O@SBGeAj zo>W|h926uE=t@xY*0y>s4Zs=1Ap2~=fJ-YZo%N7USSl)&YY)GU#9zHrN))NWWcOOS zkQW=7G#RE8l*P;lYnyk$qmb99JTVW}(Euz7ZgkKwg4JVZ`R4$Z;r%2ld#!=M*8SGO zGF*ac_W85_JjDc{=2tGhOA=oiPo>xlZdnuyF*+R;ZKSFN&f1<=l9a^(8$+Oz>}jD3 z9>bCb1l|BJ$=#<-h(y;0LnV%`s~iKUY{kc@&ih@&s$=s^FiBVbJfH021{3FX)ya7? zm>i+l8j4g>(RTyR37aGCP0ycMAj(ti3U7|sLAOMl6V^c$W*&2t6!9{EE8`fwFL3iL zr2f*WcFangzTLeKM9rEsh?P_ZkG0L}lOIv!s55!EwJRLedw0;ABo!e~_g!L9&g>F< zmRx3H=cIwYOD7rYyf+3MSusnd-J?R!vx`n8{>KhWUVWzx%~@f}Pa#_Z%cz|mhi5(e zR7RD|?gNb=V-O`BLb3Xg13m}5_bLv1b&OS}``P0g@zW^Ae8l_N>ma8IiF(K9D?u-M zRuQL>b3uJdo$7#FlcG$uOMFY-s!Ww#ASDq=%J`5*;Snf~+&`1)aKph2l%#A` zmeKd!a5;ph+?$dLHILe6SuUL_>jnC$v08-TMspO1`J#0g?#5%>#|beLK6|_N+wP{h z>6cf?BC=r;0Up^QQHo58)lnpkiblbL1lMNq-~-1tiaT?)iBT0HN4=NJOMr9uWB&`J zX#N!`L>vq=@+J5AaZ~a^2g5N>-|Qicb|}nl;(!fKL*cy9hY77cy)Sx?Sv)at7SO`622;fBD4ZKcIqs>_vvc+|aM-opDWmr9=UjfS>Y)9SZTNbrq zW=fD2o8B)>D^Y3-c_-#oig!UI4watm6;`RafL4b5+0&Bn)(TT> zG5ZL#kUtfp>j(BFOT8Pk(jk>AJOt&d*#E9A4_X<9AMz!+ur?hhHo?)1gWVqX2sh$1 z4$d3dT6pcZKePq^LgOTBnI4-@>)I%ym+yN~ejZGSySyom!2z9g_v zyzY1ChbEuoS9^ZlN9vq7F88UKvvh-Efkv>IiY}Uxz^{{U4Cn~$Qf5PZZvE?jS^VeJ zDHn*gR=hE!+HW_rN>mW8?E)#)fzWuq7SCV*`uoOnvgy69Ny!`~fXy;Roxp#%;L~A9enw*xt;|bf+5Lsrha33ylj6Q* zX?mP?Ep5dnKAdr5ezOeR7*KL&G9dQ*QSfQ6OvyU;KDTk2ABPNr&9E49&3Js|1dEBk z(J%Sqi>~F+71{@SFL{32INxQ#RY8dmU8@i`vNfbC$Uq-gl|&$sy3wW8ei~jIQpCe_ z0cH&k1@3`hnjQVvVFuV{oD5?ehPj>Hc5%bW_UJzsS$2SYF{kLnfm;xp8nVVOq*!20 z$f2SiDzW0}Gd7l=8>kPD~OF2ahO{+76>7#?1mYO8(D|Y9`_v!TudHK;b;H;K8fL$0dU^f03?Nm0x5xF<0!Ir zh~6bGqfyl%(RUmIK^6mon>K-EnCl$9%jZ8-RhGHK7s`TI$q=THozhlCrrQ;<{?(h} z4g6LJ(Vid|eQ$W)nY){5p$nubez&A`URbMF%DcqoGFg5(t`FuM^8ZxoFpKeYFl_?E zI8QwIyIHp!KRucyv@Bk6T0hlFp~}vwNrNqn8G;TPx!)QmucngSmugl-v?@Ocw?j3- z)8V!4r|lrH?SA$;%Q^^#sXds`lTz<)N#OKC*$*o;O^H9eCSA#|=BC8P|;PgennMK`z!Ty^0XCR*&Zvr2K)^Q19YC=;UD!l z%K9n(sr*XgCuJ;Fs@VVJSz2%BN%9qYS> z`8u0-S%=LqhK6cfx^{r`o~nEBd)z?k_1F2=5}4J2Eg=WncZ(y0q!#cZHkprQ`> zkXH+!3k+|HrI_pG%jI!wwmO!FT!x&w!_3CgkvJNQ!QzCY3G~OWoQyP?q)oYs46@&e zO;UrINjgfgA5x?W<;ddLDq$CIy>FsV4qX*_lsr^aNH#`vM{JgXK8)Lj$UCzSNE_K> zq?$>WREQ6&)9E_bN`8~(GQrZz4Us9#*}(e1N`8g7U9*oQf|SRzE0!^x^p%DG-2q)O z8-U&A5kU>Dt-h7~D!M}4Fc>f%+nnb>AejM{>jozvQI$Cn+gTHk+#dCOL^g3-OgnK@ zW}g}Caw!%(>P(^pK`Wq{r+L z-=bIpMOvungfCr_X7UaIOTkVMSjXbkrQ|TL2ma0G9gwEene=_SMwmrkp4}*KbL)3G z;L|p*-(|xT{5MaH|JIP=uzr^!{(WFU=y$QW!9ycDr?>6o);veSWbt|TQ0Kd%>ei7f=nRwoa<2%+dvhIpJ|Hw##$nD)H&gyx5 zW6PA*FcZ`if7Wg#MK2in7>Jz>2?$qGY&iuwBGIcSC29)ucN4>Pb4)wiz9- z#)s%+IUwnnPG4kD94{vLvGPdJ3jR$6*3jT|wcdBtq;yFJ+oZy4%I98Yx}f&=x-^e@ zwjVhQ%jfp-kZ7y}WX({;b-nM1C>{w(X2%==ceDGCcEdW{U;nSlB4R>Nu}j7ovdW45 zqii#@Y@*miifo{wn`OlT?OsQ{Q8YZ$y~$Uny6;x$gS_4i9vdXc&2692we{(9K!L5d zw99$cQ83I3Y+|_Lt^E-uWc;ACAcK@RF*54RknthKR#8L`QsDCwd0OmqT`EeI12Jg= z=_R10cvp`5-Y(*8_0)FA`;;HMHAXed(k0qXd8+J*{A0IN;azzHzf=}0Fvu=UdmDMu zOuQV}q=dx|`@PP2>DdJS9ey^b(Pp?-xC5J}K~_MoiA>1 zk8$SOLGKr2s`Ub#dzSmhg{9N!!9aisdOQ#0+nAcbt5dsS?sB}98YQMii4@x}doQ&A z4WS7o3omZ^kgVl~66Y;8Bp`w|#KTFY*i95kq@q{wQ3|a?o-aWWG-w-u7$Zc34D=^% zO9Kou_5$-V_7;q1_L3twY#-e*p84vV%U-c8%===_!HFSYrC9GAuT_;HS}My8#8z7D zXv|X|a7&bwL|_>38EN-nvU?V+wiVgxe96H8a9N>5IZNHDjOWKQA3_{(^Sq4#C1QvG zHp|v1)0D_Qg@+P-b!r3@a7V!oC`9jby9NrJqMn z#@&j*Z7d!sfg~9^;5rqS{~^=-6=1r;y1~b?V?ocG(>Nk&xXEdh`$mSE^{K1#4>`&g zy)X|uWy;l6w6+|m#@ZDLt^+Q}B6ctUdk-8xx0pYAn;~F+nZviNLtr=`I4JAYC6*kj zUo<^VTqk3NR~7XYv3X>h?>1O1#WN)lsATlK*V++-V;fE!@y#RO(>hLmJO4W?1SVhQ zJ?(`*lGRSUF3K_UJ~mTq5=A~hWxRNZuVi~JXRZizswbg)pqX)h)LEDw@}$~*q)J#I z${E*1vu!XiO$sBAoL!GvhruxBVnRnueAPT%KrgI$0_@7_P-hB$qzwii;QA zIj&O)7hu?R;FdLxUFxTIC)fHW?{BcY$`=ILC!Mb6Xh~~ zy*Pg+s>fu7UGz-{&ah5;>x?E%m9W@zr6SR{Pke=^n|;`;dK^~1#^L0Gra@!#92D5= zaP)!Vf41~9UJ?#5D;@|o20=$GYJVrq*%$Z`FsEW`-MHEoE|{>_NZWoN-L=0yddp-p zzJ1NLi(GbMKcmmg3VcGbw<*#_MPrdm+q}onKAYm#uEywUHG>3#C=G>M^>i;3N7eWx z&4~foDeQ}0Go?k?AViseYnW;)1dm0IrXYewVd4*o-b*Z z<%`}fi)vD|(9N==-bX#PF>`Sa)4^GI1yU3iQa$7z{5l=5;zij{<9)C^0+LCxc458* zM6BpI)+hiP>rqIh2PE)w=rvIt^ePD0tqQY-PR>SUlxy1&E*lnAoTjLWyzbxsbSzVp z6EA|SOi=}aKoJDU(PG8bU}rsYdLZ{@HZV9NU{Tt2twujX2WZjH7$4g-3N#Hn`u5&= zt!V3?rkc!=|8Mn+$YyRXj1$M__L~{69TW?B#%)wI)MU~vyvJ^zEzqf2LK8>}FB^oy zk&|J-Wxp3v*FiyFo&2dN-YIe;q#ydj-(Rn+0r~*wqy=Bun$57w=!^m?Lu`m$euZIg za@o!Mq9wy27jMRiy@u@N%FLD}i3vWZwA&rLzsPZrK*!%Rvp{ee_`{ zxx>J0QdmLcHuW`W9NQv%dQ_*np}fyD%FhL2N(i0f)2h@d8Op7JZ43T%bMUqqp50zY zd53v;vvf&GU_-cFjM3NtGz3dSJ`RfrjyqqU{@p2-T}sYcf7%stpX|7q;aJ+vmRs z_Bn$aY=d+7U$-owbXs&+DVOY)^#+##HE!0c;2ywwb}!u=xN&kk^uT6|VqtEn^VsXR z1-3Hy#Y~Gk%FySDbgEOLQ=)<3+}UMRGH<6xS44y6gKyQpUbIO6byO8GUeh)&fwv{_ zhR-rl8P(`j=dlw~1cybt>_m^r6+uJX3~L|l0wGQaqTB|4z3Vm8%IIge>(j`=;fipb zc>U94w*EOuvBxOTjEKgSFw$Nb<0nvGa7bMfh#Wvk%9Oz3V7(wsg(4?-nik7(la$B{ zZ=j8YcX>)!8(}k@QCu9BN?GbWAm2q*q&SaR^}YqWR?9>djB3%$%8 zN@JFabm~6dXvk$80I=bJ#~F0i0p{|nmZHBe#ri^7QSde>OG$+OCd2%7f>u?$WUpsk zq(S(E-|uowh@?`+pZ64-W!VAi!kRr`LGHdr{xA|tTJrpw{XgW@0xoaQZe@Vm>Op9d z5>KoTSpto81@s;|o@sLf_$4Xtd8W7;gq1#PS~a|d>S@d$J^mJ z{^PfqM}U!dE=A#-;jeJ{EeUgAV>Km z#(0_a7j~FDu=BdYuf6r~W6NsXFV?d-as0(f&FwYiXYwjVCyi3{S4FLgrLy%(RDgZ- z+HQ3<%pHx;NpfAV4176L)T$A>aU389bD(I#qO?C)!Mvs?ghR6|sxnlUn2O<+*-YuE@ zCldRDt&;!*Lr}4WVpDLYjK1K}J>wJ$Rno9(K{n|)HY03}vP)?!t~=_jp8-j4WsTn> zo6DSf# zMd#9;k#|F{ge34ULd8_3+d)5^T9W4UxjDohjQEUg9On`5;fRBrjKxH1asOr7go>W; z`<^08xw)%O9KeOf(;*o8fMVBEBo?du8&o=VQW&xyf?63+P8CM9LdF01Bgi&BR0pqD zhJ&E@kv2DP6DXdrr`8bN3+9soU&fH3-$k)d?6!l7Zc^!h|8}|CZ6Q|GVof71R-aPs zUzL6$H2lCQWl<-;$giL)32s#dRX`175(PTXv_1p_qmTc=c$R}8W20qxwl4X?r{M;5 z?E9=+7T@=N^Wfi2fC+kM>QvI^#BniRmT_c-W2 zP@L&zUlHsUz-E!89^`#~J$8_N0d;oSVI5?{__`AW(F4(QO~Cu*`FT6Z9w!FgIWwzN zO|h_EIz&aUQ6BcX7-&%R4=#?bh;J+U+?LJH^vPxJ^Y=Z-PH>9 z{1{P#RHs;Bmuf19U})@iY6J@BVK79A1G;^`W5UpZH~oJ`PQ75W9ds28neCb>7E;%Z zAbZ2RBJK#@#`Zxq|7n=_GFSysBD;8NCY$(RP}waf zJTJT$f_=!qnM+qQ6~Mgt^xKS6X$SNrAkPC%rZ!Hc4nf?AcFQ(6TDMw1?wWkZ(wDO0 zo?zKMM%A&HT6;7w9(pwJZ<2Dav|X4@Z-{sV%mmuIp=jGK=+IFt>F&O5pcitdJ-k;89T6VoTZRwSjKDOi1 zcm~T=2IWZyWgele-XUOo9yDwN>+|kn|Kps%I^lbZ{r+s(UH?L5sI7F|KP3D7-+BDg z{O{j?{hc!;Cb)=qM0Vc$Za^LL;R1{oL+U`96Kvz$V2?qv4K7b7410W|ZTGMam)X05 z>R)uKLi8a9d(JSLhp9w^B4xg|Re6NpN#6-h3#$;{9KQly7!Xb+%ZS|BZ76?_DuV#DCgi6g;V%@#=ODK?fO zYp7^6JGnHpy5`bI{nHK|jN>@fJR3nVk9Hr$LTzFZ6@4i1!n9A67$JoY9dbl+TD2_ro@cFa z75^HPja(1=*zK_T%CuAlBjxc->lD4m1+vBi&#aXn1A^lVmqaUoqk4;a9)*3sN(QTsOl_J;g$> z{WKN5YEq8$KE1&<)Tex>@>+)=#vb_3OLK>K_jcf`FUp!)Zl7X!X- z0@cK>Uw=Z5IB_QFrWx8BDYl*>=cwq6h>ekzo*9!NB;2J;QufTg2g}VpK&)#$;5gN7h}FOjj=5=MEif z@sc|EeF5&rKy)io>H{a}&f!IiJgXF>SL|-FoE`Fu$$R8iTb4NZW zz)#p2`bY1-uX^yxpaW8rH3v{N%lh1UNQJm00x_pWzU37`*u!|o19DUDErV$%2=t&Z zu@-Ykh@LLRUcWvMEyRz3sktNqDTizajm`GUKIpRf>uf&pvvUbQw{dBO3E_O-3h^hi z&(A*Kjn%Tbf|9^mps~*P{bY8T^p3Isq{}jZz9|dZp9><}=G8;hfCHZ87yfgjdbnS> zi~V6bD?IWtB{aI~P~N zjt}0Jfw$w=*?cSOrj3eTcfc|bV~?xHd)vN~e9shuTmAd;2jmR5jFA&BYwnl@ z?5rO%$R4geOxhscmoGv`uFp-|>A6mB5XO3(hDc18JZ=gy zjq3%c#haNjsIjWz=gvw8D(4c&hM!Sw4j6FR;a*Qe5Yc!;Qqz4Vs|e9v5g(KExfOe+ z)2G0yJP;I!Vm&fRt*lE~qKFZt(~oHF9?wI8UA!(OCK?7@GJRTvXj{r5D|XLS+oX!; z_sa*h4DPua^S9X`MSab`6~wD+g(@>DfZBN!kKn6$~Ha68F5 zu7A0H(pgUv${K#L|80`u#2KbiGu-4+EClAZQ_)7K9H%Qn?uK?i?x!XQS5d&I$xHUH z^tlo;Sn!Pk0R~~aA|p(v!sXNv?^2o8SlXGSTq?>3QHv3V&-w$j>BtyI-=^;v%}Jar z(u6Fthhny^wbNU;qRsiUH=N|E6bl{W{ugiu^Foklej)RIPe zhXg)Gr?9;RUcJ!+yOrwX=Ve{eE=5`xI7n3< zJAv46yR?OL8_c}!~eFhUwz?n!JH)8~=Dg1-|N-bcTp7uty({(}vtl;`x@Q=;F z9U~68867y=R!mC%@pqTJOvw7#xe1LV)rqap0W+-RQ*16pvZ!e6!$#stL<&q!J+ByM zVA2{hg}4($COjeNk_!pMGay24Y&tW#)5how9$qcr&~|3)!+mz%Scm`Fo$ULGlNp-0 zru4>bOEYBU{mpdSCynKA(Q)XD|64>G3#blw!^z8`IB_?F2Gy`OjRHaX61f^**g87O=WEn#e;*hY%f zQ_)FbC4u{4wT;db!aN>mpRXyg2?5>Q7||hM-n zM80Yg&Utud9&;2pOKOE>^alRwNyan#71v>=+c9`NpSPUR@gOV^c}!Bs+cdLDgNE>$ zvUf^}qKaNi;+cD%`8>U#R+#Gzck{fY1Z}Z7t3@Ape9EmhV5}z2%BwK!Zrk6F2@`IH zEvb{3pp_|1zCq%hc!n!91B;GgfrxlB6^)UeKKFL$>ctAG*vUFoEHBSCroHLT6Ta)7;r6qp|QtUM$JyHXk{0?D@Ev zi8)BIu(sbzMeE3V-wlyn5gS8tu{N(FWMfE+u!xtz;L?4-rB=K#WV^ImSnp90lIWYk z90qkRovJtDn4~;FyPR!O=))f_SkFtRI|8(LZoVi+R5=x?-NX|40YZfLCKkf zi#(p`&@5_I-t`&~>!Eleo>|nSSux3=K(_FeVO`3V(ArZTv@+~d&1bSxB#Bob>LE)+ zm?BN$aWw}cjTd()ShEGNf21$sc#M<#MbT$gyS(Tb(EVO0Slz}u@2TB2Wd)Fi&u#9YRt>(w_`{-`JY6!i77T|*4*>mn0I?%Qv z%j)bi^k#PatX?xCPG)jR#P`L&BMDCIlI${bNirxF2trbg)mZ3=?jwm%Ofy)7RYI50 z*aMj_0zz2$;gR*~MQ}qfjbwCL$7Qw|-JsiXbgkTqn>jgg=zlI*hTWXD=F>_%5F@S2 z{R{n(rtuMdUwI{@o~{WSxg5^Q!yDsn+b}W4FB!$<+%Qp?bLyA>U@{xcU%#D0_P#Wc z6KKgE!enunVyh`qfsvCtz&OyPK~c7g5Q3}rt`K*K_eXZix*}@DWz-7hsCR>=>}!X9 zk^5HVml`!Y=vH!H+z?(yrIHlVs%X%3e{1#kyB8!;i$DE;x4yi8L8E4q1cNL0`RRW9 z1rLGJA|a}TPEl`vtx0od>NZ)PKZaVU#p}Pa`-SG_j)1R;ySR&@4L?xQ7RASr~iEp*!bp=IO&@1$rO^Pg0)XFLl=oFr|nL zdLV`3)91coCYA?RIij8V^P#{GqlSsuUF^T1_~m)VQfteKeARtgKebi)*C)~@k1pOe z*`CN2V6f^TO>A%IjZi2L;`O^^(xno7_O)DEb;k6c;{064^li3Z%*p0E@2cR+sgzrj zO_R~B-(2}CS@Y6NMo>33WHL&jSa7^Hf^?svThcM7UHxuJ#K-)-QV72%DRby0z`KBf zd^}RgzenG-Hf;#HM3T8-!-?~Qd(E(sL$TW_k^usK zQ7P)BqEuj$(B7JxI_19GQqd-Mux7F+Aa3S>%i7855M|v!^fNF!el6tYlvG(Er2bB; zQm0^#iGx>09D!oyde|8oX1CLPwpAW#0!pTDRx;T&35ckMzI_!G3rd;?u$UT&+Y2L* zt-VNb_7$vc9dKD0R;z%5xtSY7_Ik&NmIs#ybf{lneRgGVgvQ)CwGnd|`8I${G{Ev?1$(s#>*@oFe9MuM= zS;%=kzPl&Iu9h)+E>SlpwlrC0??E!fZlp*86^-dzO!t8`f!1nl56}x%vpYR@dgMuS z)ftnY*Kal|v{=806TYqA#_mUc*7p6*IVQ{Uu3l433SXKj2#N}ZxF_WlTSk%nPz49f zdMJbRv0GzQhR;#ASV5~QOO30Y!NRu|&=0|Y3^bj@3OeZb6M+g^n}YHG4$>8|OOOF& zbNL{kkvj_yYhxzAeS_R_@1WoAjBHomg#M=o(gfFLalQnVp0J_-&E=S5Ih^2QW&)qH zF-AbW8Tg~WT6(uu+`|3AJ<8pJZqOTnM|)CnnOWg=ReBgyTt0eLTSKaV!Rsh=YNd}O zF+azLI>tw3`{kVG9Ot#nx!SB1FM7%BDiAp{kWG@qyxVimc{ORe72s9zN<~eYvvjxH zg(>B;9x84^vC_r>JwKf`*sUn-Ibh+4T(AMpxuKm7b3!h)n)&3WWq{3TWA|1l41U^{^LL4wl`Cd0cf6(J z4in>SVJ9?Ly$vUrOzeMu`LY*XW4Hryz%kTKAn@H2)Zn>t?p}|3;$^{`7j^{LVeMq~ zIS+rpMpO)cu><#VLIt%t|BjE`YWDNUwC-1>8C`|#fxK; zl<|_du)$vEXO|bv&w8~CUW7BMql0qA11@K$t&w*rFby>37T3{(ewg>bV%Jz+2kW4( zC~AD$vUb{OJr$6p9O*dORV$x()ymogO`4mYJtQ;m zu{7Cjue8(ifNMoawNgKB!sVQ?FqQ%4aIJML3`4?5UEW-9+ftA7B?+%tkriF_rnVrk zP_|T*C+(nD3$(@-dF0~8+SyE?ciA5aEG&Ef5eS};p`}iaL6eLyWHhXxTWI5j2IxgB zpPK|6F4%qV09S}Jye!O#agzgXY&^9&@82F+dM{sO)OFshrZrb<2R-0l2%W@DEHvBu zw~$h(dFv)i<}YJlNCKtk=Fax+^JQ917-u$PVlLd?h3R=k*B z<@hbe?QVx7;>H*yyah~);<6aL*mu!(~V?^uS>)hYI#pH`niwWtL90v+pEyA!q zgabpc)@P7meANkAU;Apyiyn2o?WLdE4+-6y;(nLTDSDrq;`D&OoMku5vICIGHZSPG^Frk90y2R5=7RB&t z=rxkYEivrG(+uQ+hb)i_D0T-0)Z5YVz`+69;Gh=0KkBSBclvTUu+=6h>*TnJDPf)J zh@uu)43PQ;NjosJiy2F8_uPZ8bkAK09mPm>fbC(9!q}*>_0%TW8&szI)t&kIZ~aYZ z`F+vh^<=vf&p%KzKVG+zJE`cC-?;&TeVIHgqOVd^GZo@SHp6W_4}U`D?;8zp zv??f**Zx|b=1%Zh&^kK~N=WCTuStuUkAhktw4(Ff>akA(qO>Gt`fh%EXfLS{Ci`bd z?T>~rZ@sPmNI5g#WrMcc2X?VQs+|9J>+fxma-WFAVlA#G|&eFl9b2; znjBUv(c*_DukMI_fd_&YFJ5e*cYtVOMpzt+VhBK)kB&q^;I8l%I>x;{G@daMIKr8B z9vM3g3k;5Z-&0P*;=Gpo|H*!K^d}}Oqq`$|LK>ae%5<3l*+8)^6uCx4XYlXSOGF9$ zo#EO9{w8LL=&n4^uh}0X@tBc925wmS76<1^Ze-h}$fY`Q4w$<&+Q)u}q>b3~geBfclUx#Uj(h}VpSTUBN(Kc| z>KOHR^fS77IqFJ+*CA6+Oz>g;L3I~To26do-aF;a+}#Ww>Q`*}QZBt0ro_cfkU3Z? zTJClmguQ=e^q=KNRDVYfBEUl^<$wld7w-`|ieaB;P z0GrrtGTbPgz7`@JTBOX!&25_P(C<6QdVCjdUoPwc#DERkM~=u^J#(XAEVE&Cv?3^G zz83F!-QNJ^(z!I=6OKXB$0wbRW$`bplvxm2Y6X?!EcJ$0Y9t`riKTA)$gZ$#HICy2 zj}!AyBPLyzSRaQ6fKB^aNR0%~+7R(;XBgn&Ai-^biL)kemUx%_ z!m?)Sg^CMXp&sn~)+b-u`IAfE>0GdbMjyHJTS%LHQ;eLInd-~4OMbljhxM=P1*zZc zS`fD|_MhxLFArxDa36OxTf38e_ghEvAKL%?$Wo=zioBuTry#OL*tX!mixxD5my1`B zOrAdEfX@N%Ps8;-SLv;RmxCG{z8D_*TsS@Lp%*&%v-PJvO#H-ob22;1zq%mOG;3|j zRb-I;+*a()`#&SV=00RmaFk*{q(~LEJtqvdJ>QT6?K8DHD1T;yCh0q;7xul`s6jIJ zwR299Y*Ds*2{b&H2A^We+)6Z;2`UNPljTX{fr$gCqm{d+bn%{@30*CwGj*!1kQ%S{ z-XePlZJJfWW1rund+uin?1~H4cx$}a*_))%Qb z&tkKeSc(ukExvSS6IKGn#!+M~6}^;PlV-s3@Xp+gAte!Yzyc7*+Sxoj=b4)h+X4yO z?quJyI6-2(lRZ;%U491ntc!4My&jbcGX_Bs`v8jwX7Gu?h|^Ug`S zl&9S9)7xR1+4G%i(o34eZ{hrr$xDmO^C)JH@G{sK(c@XY;EkmD7SJ{Py`H7Y)2eoL z>Xc%pVesrrnyzo)HBFifrib(^T4=m--ON4o<)BUiL}IFHrfnWXsDYHZm?`C9qGNBJy(Er}4~Y>yP;@Ers`l_(&s#v) z0<__c$5{7{3`|ZyqrQ{)kG{V*0qqZUznn!rnMBr_1%w_`>}M3|p`tOSf|dSDsau1} z;#%C;wcr-Xp;wZ>>Sy%16)7-61o1yS6$3o!5Z=M~O{%O`Y*?ethP8Tt`%<1lFt#d9~14j`|^0(ShA5q2FiRv3hOdO;3`_?Wd* z2jfZW+>faB;aO^ftWJ)FdR@?4kE%!@WD{QPeJF4zSt>fl^vWQN^i<(yIcN>zL`KBLOMulDyhuliMXjyQvXOpJD zD{UT9BX)qmk6uy~l_9ETvgy*<23c>!KJOdKK4`egR3GN&j#s%dcWXA*p||b&({GkM z8cU_1^QL@=l}f?8p;tnTMPN{ejETL$JRg!?v~v@jvF!nmc$Uqm81XLlKPWADxX98y z;$k+O7!_8WDO`=AF#1w(R4_@b9h9Dc*2^SVCmnEWQWONf?N#dKl(h5F#|3r6NscKSe6FK9D@vt5~1q7PTuEU z0ZFw1mj(v2%orxZT9GUWGkxS!>zglWR#wAT_R3m(lP157zZ{AuZv~N{D#epr+)nwt2-aFcr^c3rTY?B*7Eblx6za@NdiucBBz zMGjKYIA|@rPI3=?AOpR)PZ}3?IjA&fu~AL(5wlB-JQA2KU7RQ>i$Z?4J5%>EOW5_( zism=7gT9!~cX!}+UYBCIypCRsNx6)$R0b;?uqOh`kg$~PB)Hm8&JeJKUlP$Q`{e(# z_a$&mo$2~L!U@TTAsfM*l;A=HL97rA74=3towc)XcV_PGz4M=GnZ?f5naMbJI&%jX z6c;uHEuaCFMGzMj7gSbJtX5fyisFg{v=$d|p{VeGUlJ+_oSFkkZuH*vrzB_jg882F ze((1#&*L$2+I7TP8S$XwAT;8~*?jQzzkchAml;B(UwWgFtmXzG7mhjRSwJYAQoc{Y zEJUGhVOeM~9Sf+qB7K6i(9Etm9oqX|hs4pMrtq}DL&9=lyrfg^ya)kDbl5z*W4$nN z`S+Jv$9{&VMO}EkVu=S~Y!9P+u-qKx!;X7^r^tEJ{d8Jvn)s)eJM9(wHob6K*H5h5 zAY3*a$kNzggzQtk5AcEmH{yU+c?$Co>UA7UayR1S*#rQ4m!*yPn^+!X2fzrq*_VFz z&wVY{Y~U}5#|u;jL&Aqilro+otEni^R|~_MF(`5@yY8O4#Jx85+IyF9kcSFV#zHz3QX$@K@JkM?o1S^UBZ;{`t71EpkUU{U}WZ>_l8--*VC12&urj~R)n zldR(-F1y`fiFo9H<+8M3K9Zod&rSlB8%Q2vVI7^%?;b1Wg$nuYmZSe>UQAUpUAPe9zQrohK`GlPa+Qig!A*lUP6l*^Sm2RIap8FIU21v9{PpRYtDzrEUL(i?0@@U2tE4$%z2Lwr`U1^E9v)i5 zuX{R#QIQHVkPV?_lc3RTyXfsS?~33D=&Sq317|w2*WjzSK^L%%=?pi8@P{LqBhuJI z@MAkOY#rl-v~f}2*zo(YNSeEFjE+Up9J_o6_Se`wm8n_DnA48fYk1eIVoIXmEN`H} zFe##(zVqr%@udKRx-@)yQ10{!ZDZpmEUGo>c_q%4@Inv{WChOLMMT;rWg?#&9 zP`dxaSK;n=W2mY(V!yG1ilf@*U;knZ^;gr|KN{2@Q8?U@U^^8N+8;LdaiAp%`A<9s zVe@L`f0Je0VB^9zXNv`Fk||{pMdGO_BqBh&lLfU~`a<7UVVXQmo}#7KVNSm3EW`e!q#Mm1uQZspo% zKF#^yM8P_b_52+^nKQEHWlCY?@!Bl8JOdcX$3KYZ8Ec1%jT31*LO%0dY#e!PmrZ{4 z*mCO)waadrShmQ;GxpAD)@5t4xFmOSt9OjM^^xW7O}al0_czD()qT$0+-Cc82D0-h zPL3=!(YIP1WwssnqFMKM5)jT#+S( z^-EjnRw1^v#4gANrGj4m3LYj34AA|LS%us5Nm4So4>skOnZsIxHksee&yZwlu zo0WH&DxWrH@HaB3O`p>$>}2ZH$4H0bBYx4GM*1ob>e1;=<|dOn{h&vYZ=5E5F8)?n ze_kgqz9O;R;V5PY;y?7)+pQho;dKly9LHmEb8*9NN=u=LwK*S& zHFcQ!(j&`7j37pUh3`=55bIGo$Dn@1-z~uXdE&HH(;$O(T(Lz55x?uc$y(S)V@Auw zD7y#1Y3Gc%ePJ_v`bT?&9X3~}${cI@X)b|b7q((7WYysD2Hu5Y&V+X79#LqTNTbT{s{N&HO`RCwnL*XzlEx zqMFo=;!gV1E6b+52QlG%$#&_%pcYwIq(R#Wik8@mgy;UAr9BnY7?iDT2C^HhtxXvp zOVto);}0qistts;+a_qON?|&}8ssZfx!SJCQ_zZ40fKJ!g^Vpou-Wsl3kFPsu&*j3 z6;P?)yso$2F|lk+t}SoYGYGba*)(3Qk+4n1c*0;iG z3Z6BX9nW)`0@t<0VQ=`o#UHdOfo@Akj0@WokQ*LiQ&K49T8bn9jo>TBNf@U;KC^_* z9BA*@F%63Gw|QUSVT|8dj)WOpwgc9VXtZG{Zh(Dt#fFDtX+Cpd@06waY$v}UqB`gn z{q}P2j|BDdt=`SL)AC9|Jh?H$$q2HY!peqex9cTme_ZNiy=P%@_Mu|YD^s!yXnG9l z)T#Yp`cHxkYLsbRBB})L=DUI=q6Gh{sr_y%L^q*9+DTur?GNYRvToP!k99Ck52x#f z5)5rUN#8ZECYe<4*U71o;%~0|`$Md4=n8X*QnpZ}8RBmPERf0mCK5&DV$7nJzELrI zqekCDO6S}YCkna*vCxmd%ujEe)I%~gDa=|(0q4+2<(V%+zZ97x7wbNp1ZBXh7=1Sb z)S%wgK@IZVB7-);|30(`AVFikB-tOC1p3`Dc81OWxzpn%DVhz_mI#nqpdN~k2R;T_ ztbr~DgMMlx+-n7QI9>vw8Q4nWds`kU?oY$uQ!%EWtWUw<3HGN!b3t2f-800}hLfe; ztJ)LPq()-z!F0cqtjyrQ!}QaZZ`tw==e%S8mPgn`xUL7g{0mdw^fB9bl;90QMSX#T0zZjszFbcL;>i=caj~n(IbC6vxBRI`U7O z7~*Ytmc}|~wxDE$r*5nt`^@`a5R?ce?P&}$qhxxF#2D`>j=xyR0lVyjp1?17N$dy=waWdVwCtF;gNtt&qTHD7vz1> zbIL?BaK5?ytzVOkF1$fLU;%{!N(n`~IaCxn+(x=i^`Sa#0_sC{(n$8y=hir{S;PbaT&V({=R&H2b<_3@$SIwlhmwz`U2Utd*V|M~&`I~XVUK*F zFI0}Wp=L%az1q9a4O0!c%piqo160wLO?n@aJ!96T9{#0sfx^h{#o*E5V~^zsyZIPt zKQ`-Mr9T{tI0o-6h)~PczV7SOiMIp1o(^WmS6^A#{?@yLvAKt}?&OvWJF;sm z>frh)?&dcM~f;HPr9!u-hvDDhYI1NZp-SI5o zUFYTStAWO_GI)h%UEm5$>6{gs3lJkLQCtx>(x`leJ+t*RT%N5p!bmW(;=#`>*F#wMBBc!=+ce@7DC zp-XS*Gnz>$q2w@)it42g(>ajEE1Pmewl=g5bYPJAXe! z=KMT^x;!Y&IaL0fX9A|fAz!`KpqysHb)#C7{b_TN4ky zAjQ@!G=(j)#FxRqiLcRHrMpogw2i?sow7-ehm4htu~o3o`ZkBZ_m=i)tZO5=fWmd{ zAz`Vc%+lPPu?nP;+QSa=4v6>7=u=z>{Iql_z9pqKY?1Mh>esYTRxp7*6 zTsRsFUFt)&yYZBAHAP}U^(xRf2`Xtq5>)Y$dy@?6{}DLY0tlyJz~X~CRXa}!%z&7l z`rTt>xeG@$vMi=}J*8Ynkwhx0o9Xp#Q+LpD%7fbKpmK47zrH+Zi&9@UC51UKu?zHt zoI2^YJo|$I4U4&O@EcAK`b!Oszfb*=85j@0{FR@PYHo!{F1&bLv4Bw{rL3dKhsb+U zOA2@wNO@?J8XG;2sjFo9lE+@qM2s)711!^Xv8F*@z)R8edfg@%CWE$N{L#ikC3gMN zkS-b-mN1BNo1n&h59pp@Z2SaC@c)2xNuimn*E@sG2#JTDk7R$0toOTpKni^unJhI7eEzKyVO>qx;DXvmgARqU(PbnGO~ZKqd%e%myg34?)G6 zzTRUi@FF<9w7`}{Xb=S4rgf}WFFurfPHwg}$G`rKOJwax<@heVP1$W>QMOP@AWYv# zMP+Lr(K#Wxp7*thf|J4?pDWUR0SC3^!px9n9hTpt%Ek(he(}S|%;4*@HuLXF*Yn#H zoqkSOu104e#yZi@zMp-Ox->WT&^$9pntrq_o0M`}kX+bxJ!=7<4=CkPiX5h*P~{03 z-;o*ql*eJ!MQI+i8OMrF5WPwJH`)KD>I8{VRq=4mIIKE9{rq&`qs*6}s3Xcz!IE?& z0?1N)PLTX*`gT5OEozZ?{fyu4$$A_J%@7U?6}DBp>vR$1s{7qa;qIF%yuUc2ir2F2W{HNq8tk8PP|UH!;qXY9hJh^1?Au_(){M4H013Q-q)ClKK- zrtT|x!kz>;!Pr>MfQ_iIy36rznc`@T^GV4K)N zU-OBT5JS{$u8LQgP$?hdta<-P??vWH|t(HP10O=lf2(zG0CTt zK%tNYwe>PwGh%$&d<*FW|1SFG7cNWxd5u9m({k}yUZ0{__vi;V=HL5b$NU!NA+U1| zzKa%>i_68wBAayAzIk{4X`SQj1*hpBzBoK%5=VcU-6t8L;E{SQ^s3Ri&&y?Vcq|QQ zM*2v^g{d`CeRWWV#K_c=WdB0nd{Mt!rOY@9d(w78nHe6}x6+uIPUFP{XG0fh*DTIk zRnF5gYLsyPByKlwLdiJKzx<2HIyUdJC}D}V*Yo2gSRD;9jRM{v7n0-51N(H?1PS)r z9d?ErB<5$Y>}1TV%!Z`kb7cJsv%x8`Sd(%nCFJ!oO+uzEvL~u+D0MlmZ6lS6Wl-*X zNZUnrkT&4<$(@`pst(p)B{#kGi$%Hg2OgKD`U~VRKT%LjC#Y5k9BXZCg4OVK$l>nE z_m@psXT5uIS=+&~A34l7s8O;YMs-OFdSpH16e&^fofGZn*eM+{BvkY+DA)w&~mw@#1?~J9@&xJR;EUkWn(uGK}u|$+7%^VNa zMAv+C>1_HCumcEdj?OkCHmoCD=scwby*F@zx0HBsZnUf}rC< zjLO=0b*A)50FKwnev!v`yUxa=<;Z z<+FB>g8@*CkO5LZH9l#wd-S?F^`VPJD?#N4)l(hrR*xPh+~=Ol9l!n~f4p^2-(@#T zEMfXeVUaJ`0rGa+tbKE`q&H?3K+ne!-m#$b%2R%y&ZraG_mO|r*UMq5pEZUpM|DI5Ftd#(`$He(CgcVpKD25Ef6yddH-Qdb&}Z z8mvcExE5L5G(EOuGy{3T(aG^Z;8#6Kf7Bxvm=TsQNZ|Lo70`H{K^-m1kX(`ChHIm` zm%mYcfG(lC$Xz&-6pSwFQ-OrdklE1exY+ZKt@4U<)#yK%pS~oEBN0 zRDV#rBm5wgL1bz3q@T|34AghYQE$=UAEzk~ZIB~BUKe8~&aWIQhDYXW!k@})Drt0#-y!Nw1@`$?2Co+7KMsNItnf8&+}}s+2;yQkbUV#NA|vN?d-C6i70x5tq+F+kea54 zob=YW`7Q~}*5(C#5|lK-0H!~EZj(=vWSP8SV4ixtuV`R<0n0=*B={UWF(hh)_P=ADfZc=a0*eu=&;L%IQt#vgx#0^JhODoPtA_UH!gd0lMCGT%`P1G?X~cq zKB1I1DAGo1?`xp6UNLKoQE%nmZ(z{4AgbTYFJ@ylAkQ9rrlZQb@S*kKaqQ5b-KfFMLZyt#C z%R(OTjP#~RWPgfNo`eSI4}@Kj{ccTaY>Qqu-k@IZ*)zVIK{_5R0oP-{bf^55JX*BJ zGn=lI>DP+N1%i-|CHgI_Kv(eVaBgakk96V`q5e(b-q4ZU0*N zjTeFZ!T`kx7^TDm+ZmhG*Lejz z+?gZGY3_9Nupj%t=_bG`2Kn~RDG$vEOcbOAW&v5@V$m&moHBY=gFI8S2v;opjkik^ z1x53*3@dl~0qwwVhMrZt#Pkozr8k3G0e(cO^d_?i8yT^=mj z$0gR3Subf^I!avkUq)EipSD3YO-yh%b1Sd`_EW25g}gpD^FBLhFHmr@dgEqO zfsA!m((oOW3!4y@PNif?y<#gAdAABPJ(4Ba;2_m2ilK4}szIUdVV5G)BObd?gdYq0 z+)jxek}}VGV(ViWff<2zgUj&g9UFj#k7utV>;Rg0Yw0_GG#eTwn-P)QE^KJxEPUsF zO8J-~pHfk~-njGsPRw6Q-Fo}dSFSA>e7E$6tv|d!A46lUWY^qQ;V$THtpRsAkN4l3 zg76TiJIEe``dq5|VC1P8mt=RvJwVN|nZK5x-pqPG{S|46_dTz_550O3Dh}yC1bXXM z;ox;$f`W)kS`5+kDXud6WHpdoFAF^Z^?to{ulR|oN!KOVHTS~A)4FEe1K#7goqh&& zm+!}Z>HeoD-10p)|FLwl_bF`-T^YGFWHI#^WJ;C`a%jA+%Xhy=L+CQUHj+(WWul?! zyKFuXd@3pa(@){LQ)&!ew#6Qlf$R?q08}uq^dT<1~6K zxI?mJX9EJJ)HX~wZtmGe=7foH*{61Gu9nZ>Mj}Xx42P#kk-1TH{d|c_8p$&~o=aw;1;;^fAS8 zZDp|O-44&U@l=DbS+(3fBd}SQ$2&@jl#l59VMj@RK)D!)HmL81?bg&QmX3WtWh06{ z+g989;`0z<%emDDCNFy4>+V**nen5HZ^wRl0~Pn}4_?n(kh-ArwY=|jzP@Ppcd}or z`N^8EEXrHp%$0`IT-!29cDjarff>O!9VefXu;G8#qJxc*$3YPu%2LXOaY~H1q{*^? zRm_>-G0nA&5V8|6Y#*j|%YO!0gJQUI>cW)^EdEAEcn9=cV>xJ>aYtA@Z=FxO zb19%)d+OP($B|;yTR%2?B8HoyC*<^V)->B~;cplzWh+H4Q&FFK-T)1`s>z$ZDuu;C zm2?S&#CGvN6r-HsRdRh+m8^jNRDNS-z1Ke3PBCVB3~KC9*yvFwZjzs$_gIRr68O8J zkuh7lb3)hD4gc@S{J*?+71~5<>ANA1cv~ZjwL7P8eHlL`!M}aBL3=O&MV;??Jz*Zq zOocv>BeE^B1M~xNfhK*zy@>Kqpgf$8zmGv($XwO-GD(6a-Nm2peQo2HmVUWoe#(Ll zUD~&=zScGW!86&aY&0>pIwD-I`&@g&j_*lhlLO1w)jCrv`D|2K~e|( zt_f$z7Ts39{!BO)0z0{aiUTS>cNXn^k{uP7zWCK&S?kihAbDmM6}u{4HVsYFvPUFS z@PL#>p5Ucw^Ce3KOT6z1v4jTswoz;7HdrU8xh@=lGhiz04G^~RuK!R};A>u-ntrx- zAxU*7`z-8z9;E~&$*ok>@o!z3e_8sEYfRz%`#=e}i*I5LJi@C~w0qTyYg88@&v8#y zFEj?8)|Gj#59kQFI;BaMDKZ_|;*+YyjD`cha;{&prBP=4ipGAN{plQGXW_bmhu4ne zmRWnZTujOH=l7ZvrZVVm#z^Z?*}FwnMfZrERGMi4JR7lXagQ_Jyc+k%RnL20(LG^{ zZ@V%}d(SHg_P2+G1wgvd%a0cnD_VtzglBn~f^FW%puZXaZUEip?z#K;4m4v=r_IJm zx3FVR?{?fPZU(Jp)|)@FR!!z&H(WSA4=dvkK92V(C5Y9orJ~BAp#y4`#v6Q*cMN#c zq2?cHlZ(a3UA9DY&BrNiX-@&b1xQcta{Q~c|9P$5I`hv38ZNw~u+#})@)y_;>ghy} z7EqEYjRr}Jt%E}q`wj8|UG_x8VPc1v5ndLxIFyfE3DMhlZs zLn*5$QcguZjx6+jKwp+N=|1?zRq(U1R48?7C0Q!i74l%py@>Z^D_~7EHPoPxjigd5%A>qo%TDo0VOgA$t-1mjnNn+ibb!DOWu131v58tgN`0lZ! zFcRN&3Gm%z9{H#0zs`uwCTC^(m0yxL7cPU`Zec_=QOY!mY@ni2m_lEW zlfs60q=XsJWY3lCf;RaIvPu^V)q{H#Ta<$gKhCN`8pVu{buw+bpOY6m&ZjL+^9wUR ze%G>1MXqsMl3du9EVd}B?WUA>DAI`Xo1(sP^AX)(Y17T=v%&`hGx4=M+b@G5Y?O^VG?T;)rBL zE{$bQ{ca8N7Fn-%Tx4A239jTPo`J!cVgZBwy`W&o{ilw?DJAlsjD$_%7 zTOREhOCj33BTDCN4=Rx12O{59D>OxAgn%ePaK|f&f)?hn5B8}o5!KOHDt;hvSbOCt zF#62ZYqS>(1#f`Iqy}pT*=0*0SzKP&`#@nV@EPQQZd9^=qqv+Or$i+gXnlPW=H#7_ zjb)Gnb2oG%9R8Y{HTv8A;=j%|TcZr-Kl90+=ParCk;Mvjgi=D!WCaz45>(APtiI3Z zFXI{btt5J4gL52mP1gvn!V|Aif_tU!aP+;nPrBr+oT<4sB{mbpw zPwVb$QBZHOM>oIM>+ArPZwG{{dq}g+bX%J21JM%yi=oXrr1`~nhlF>%%C!Ast+;fY z6&V}+*s-MdEFxy&9{<$hfV9|^}rF|^D>9ZPU6BGi_sc~`vf_a*3nRYwRs7=GM8<#`>uVlk4 zvmyEUN3(a3!WYJnfJXR`BtR9VET_mpD(c`jFH6(?lO$CtFc8X~uoL2Q;!4t?EBWeW zX;0X7>2Z+zy`zW@sZpU6z(-%cEG<+FJ&M=f`|M~XcrSUpE>)Fkr{akUud7lSWe?$+ z=&#_Nwe)om>ZNkszUnIEs=S36D7;1Ut?7Jk{g+ny3nhZfj7Nvyp zCIc1qvEK<*;=fe+Uu8a3q%fC-b-}y%@w^jc^AzCU)NRpTmTuCX(#D78%5t<{xhDKr8q-{XZlN6Z&9_}IGA#rg;E?p!!seS-v;l9^Wh_&FA zcm>dM2c7ZT#r!wwcH%+28K``qhOSPsjoupYq3E`xPSOan5%#bZgOfX2=!|GD1_!mn zPj1k0KYsQX)(Hcb4Y#nQ0Nzdq4cCMGwLr_XC!iXd*Ki|I^Y zqB||W4TW;o<$1h={H}T3iaJF8CQn_?84+F933^*qW49ylq^nwZM zP-JOMssg*GJF&?s6Js^1+HE6SjE_~c6tYJrz<)#Po{ zu}!3fNnui_NBb;7W_nCf16$zqY}!O7m_%Z!EG=q*?Bm~?Ij{*w(!oWjNDo==QLb8C zQ8c30^0@~`hcr$odP(`~xbgpDc6{Hw(U?vSxRYiJ|MEDc1gfqYDhjiC@sgV%^};RM zTpqH^7}RG;ySPqe@b2~QbE}xL)f;12eQx>uGEaP!1>F`;RF73{yiTAk%!eG``@GVi zGS4Srg*?5%d-<$0e#ML)`Iu0(SD&MaQ(_&#W^zSy%5OdB@#fMG$TE4^%h_}Rzl(nR zh~jO78gCx$a^Uzm8{KliY0g}?fURHNyzxaO+CXarrYaQWicFm8>7Jvs8l!ur2k){4 zzk}C0^Cvrg&%Amu>NjTla$`=yG;+g*H`XgGiW)wpln*G<4V3?0-5>~jV70^Hw*fPAx*Dq%4y(#0=_|+v9U8g*AFmwWL-g2dmLd62)FEn8cS3PNs~QH# z)a26jvUbmU`5h1Nd{dYKsV!(R!kUIm&2?Ur`ivhE01cenJSj`k@0KS*ZYZptz#4=m zpn5e88!zbs692pa>`ubO+mGGlMtlOIndMsD$A z7cRe9YJt@Il(LH=x2dQc&C-yi3!>*{NsUB*5t?heBJ+5sf*OPLn`Di27j2|>DEr;k zlAeez!KE4ca`i*fNaxV{UHpsUuE<>iSc}&MR`61Tvq(Bqr)}fuTfO&(B$5J+{zg!m zTz@4Xj<1gnLA8k#M!!d4kgc2Y@hbzj=V+>fFA&JA!b!rck^%&GA$PtPDoa~IOhcv0Pv{+)cT~y?&Ujf|Hmw27wjqr5EQgxz> z;;UqBV87eakS?f@ubRByt4@pFo(q#fzw&x{pY_I#3vgXH0m`z?`^fj0?69WH)2Jxo z?N`_EjpDQ7hfJ-)Fr`Jd8mh65Nb2SL6!DS*{)TBzTR6EIF=s89I`?*;vGlgMuK!0^ zdRuTk+wIw*)k7Z?)bl~x!U)vy|Abx?-H$kY;}cg1+8;R5EF6EvKK3;3J^?#;9{ z!}oKQ0%zgz-Z*i)pFVf`Mvq3H4EnfYv8XtrMFxjV`|}L?Lp2`A^k@hN`B0+_3;mn{ z2Tm{;bZMLb2X;5-dve29yv#05>6hMUB&%N-mnP35>X1$;-=|=HqKx!`aNyJ}favi} z!TN}SLiOChHs!@o1GHaaha5(VQ1t;cR&d9S6F>tWac|P2jInfEyRgH{!ZL(< zX4qt1=85|#%y1qE%on9Nbulr|t!Gx$f z!=Nf0cO6^#AgGtd!@B+p6F~X~cDTipKk_}TJTGv@VuIhfZ5S9_N#N(uiE`j;_66m? zr@vP|!@Qs@xSF?yF6!>A5OhK+`~fK3Ch+%4 z?S?JR;2DCWkzU#hPwe2?>J~BYMeHazDZDY`xU4ATq_7Z}{+9(51Z4Q{1kH$p!lZy@ z0crlJ+E#iSJ@6_xNd$G`bfeD-_lKl-@*TfAIkw8)p_BP5eBzW@-ut}=?=YydAcE1P z>(=$T-Ge{(b*FXJq)LGI^tt8v6!E(G>2u?GV+EAGar7Cs#t0~O9R2>AH(q_sY=j>3 zH@`{BT{uX5!NTC2qLe2na*T?~pfl-X>NciJ@I=rkyACso|DT@Itys;AQRVQDfxgK4 zDLchRVVc)X!BL-`iqZ*{bdRjhty$NsJ|m76HEABJ66E(}36f^rUGr#91Wy9+E|mYu zkd%v0E1{Cye6Bd+!VEmu&208J9n0ZA5;y6pg+}3N-6&DyFz}2KMeN`ishm%;jf*wLNHO_&`e~R#;_0LoBXvPIRQ?q*cy5nFBD+s4uqhC9>hpu+hdN%DBm#S9YekP(gr9*u0eHn2&W z!rYrwJXyb$Z>012=f4A4vKr4#v7Y*+GrH$y1YVbBde(b%(#yUIC%22*c%LZl z_*Ik2U_B0P(r_Pb@zSvA=75_aPv_^n9QFyLJMe3G0v6eA|J|{i^1(A;buC4$1Hh%Zkw>kVhx9^^~ z>ccqauYW$RF#91H6Myr2lHv|3sYCmJvMD7vCK*)J={Y6`uSJ$SahW`qPMzAJUGH5_ zH~Sdo7KLEh>?Yr*Lg82>dl01KF+q=Rp`(#wBTj~|C=NynC!DzMo}i0TTz%eEXX43q zpz<-&#StmYIWe}->yx#`5lNx-^6S3IVBJPr_B%=lae?e8KVnq3af0kiLH@I9two`S z=YL$d=7>co`n+=Sgmz#(ObxCV#`DgQLZ%EdKF1V`CjhC!_}rk>mksLDIVAzL^xqa8 zTyRO+p;$^fL6)jgR!JX>Gzv#LcO*DQmF8tS^IwaO%^$qWKE{y?o`&1O5s!=tKh`&~ z6&#qZBHFK$CbKQ*`*=ezeW>U&c6iLH!gC?~x9Holxco>D}CS=nckYet4>vw}? z6$n?ccri_$CO4>&73>=Q(A!B)*J!|FJ<*PM|7L0OnIN-+GCkv;NhHUGWA`-{lfIWy zf-AL~ifR=$$&8RFEbuxlZv^w8$8h^jezmZizsWa`w_SBF!l16@_0x}iPLd5X%7gV- zioKY+#^`HEt8n+v3*XH7cYT>>!k02%|HEdg<-6PHwU9A9EkCP(JyCH)d}x<^d$@g! z2+ZMVpfIBKFdwuREZFVH2r(Y&#&_K`&pU9oOAiKZ;br?mPZ`!YA{U{418=bixfxP4 z%Xk+(@$gN-y3jLz5E+b%G-$7i+nMU1PN>FD3(TTB6c>UwOf&d{%&+NfD~)3g90Z34 zd)Dj*8+!xCUa*1Ri_@Co-=4npdCwN^QQUz@NrHc?aQ~O@(7QqE`5%|1nDH^FI|J{@ z9`QFzQnk&x7-+-qo>!;B#NP@*MqqvbT(&95pg#1CCu|kQORo9$ zdKL3x+^bYwf(G>^=^pY#U=-%iF{)D)z7A#Dg= z8KXMOYmqgB%9`nYiq|Q>jKHHFyMr=34@^9(_-HbA5!uBuIRZkXfsFM%+#nQ~QuzyO znxW@vlf^;4c2xl3e_D;CU-OKx{qfj_@hVEuIpyD(|xlZt6E8(?>Q!)-^TpC7YggJDkjE# zH7^$kLQIm<4f5nU*maAd<++oWyB{Sn?t45_JeQAMeqi{#j~Y!Di`MXQ$9m+QFRokq zyhrwqL==VG^ugZP0Rj3ZHMXkeOR9LNb(09R-2WHZ8-NMBX)dMA zqR3_{3U>`arw&n6EGtT3I>JyY(e#Atw`-v5Z~9}eJF`|ZCrLLymFQb!U6B|$UEvuw z;if>pPE|k^P!3!WZ2gKyemWiZJ#%|bUA6>flKekjk(!aB_~+|QB*BH3uN@W;&Y+Y) z3ZF_vEf#I`cn_8OTEEd`R_g2KCjjYc4!zC0UU-&Nfe?0Y$a=3zg0D-U!ry^_eG0kJ zK=!EGb!8qpVuQsM;{-XC9ppNn!uS33ZYJX7j6b$a( zlP!}k@xBneHUb&;>V;RC^Ya`JlhI%VOr7JvVIG+qMw(iGaBwUeqvvn4oWk5!bcR7s zbd9u)Y3Fwa8q~WcG&r6Se)z=OhM?j1Ir_+V1Al({dEXdykv7O4?wgydUG05Di>@Zh zads=xHPC&sboxOpqyokIi=;u0+m#B^-~&18kHro0EbYmVPP$-zw=hmwE(G~K5Y{UT z?|0h~-YTq_f1O}*aX_yx0(McjrGm_%wu0#QLP)kFZ0K+cu?o3Cj)`wfcVqH;@b|%N z_gY}|+5wt?t== z!lVUa*{UA5Zzcghc=Lj6jP}j`&Z@T;X4=J7=38r1(j3EgA~~hj=Ufz zJTDy60F}V;-8O44B#%B67w}4WCiNs-*^*~3p9So2%JY0Ye>)=CznNzgwh|b&7eM;BhWSZnaGJrSu^jr1|PJT!$t&R-0t;2#Y6paeRKp<}@f$}6q*cDx$| zb2b_v*k`Ige-w64NbsrV3qrGvS#Wm4VY2Flu`$~$Y)l%Z+(40JDk^K<`heJgf{3GJ zqsKXM9xoRH_VPTdl z^UULALZ^VK84Wy@mGU#8NHI82vKbLGO&=pjRv=X`DB1`r94aI+WNw~6OEE%%@ufeu zEM4|zv-NnZ_6PIHT^EkLBv>R47LQZ*QRET$90Awl(V}wkLy{C$saP}NT39XJI(rlE zth7hk6naV8DXyjaB9DMlBa+PJ@miPzbc?dgvqsRYM*pHi@rYz-bD$Ux?^+5ehV=n^ zpswzc^n?mzP5ipZ1t^nR%ENI_s;~JP!{3)Z3Fw_$PU9y)Egm0>fg8k7vlqUaY1XAF zvjZl=7@(I+}luc@3koE9>STTIcgIA1`@VDw_4$W!>*!J@G_%`d1)H#iJ zmx&HJsjl^X4C7=&HUSl$?nk7|gkp5~-TD-{yn9otAT;`5K+La#8E?km^{;qN#1@V8k|Q=o8p0vRIALSlpMJ8ZOm4=;@vnd55?SjG<@`f)2wNy6 zq>VRHQKqIy{Vhn$VS8CKbSN9tScTi~79%k6vqQRXrTlVei|mvi_IvcZoe9T^Hivtb z9OvS*htVMuFl10}bK$b-gQzQCkDFr#iP!yODP-SBY%eannLcfS=vqn%T`W~pRFbk@ z(WWQ{ehGuRPqE*lSyw@h@s28Ru`3rJi!2vk1F5#8h(1LoFJIEeWP0?->gBbN*Di-; zr5>`*1~m$|bONy%njk#iD$L|z9^0V)NOFzoj_CCQL2OAT1eh`fN43UqEGw_)pA&EO zZsR5UkL}4AE$lzTKl_UzAm;v(9ro$jZ=JPXik_?eHp}vanbnwSslfvjLBM1Kv<@Ur z1GVI^+s1$%M*EO$U~KcvWhOXcgUirkx|SUq4T?kY*69V7f+ZB^xGcRht51;Nj}jd% zvWw!eK1`$gSK0J6k8bEO-NpVjkFZN|-T0RGt6>+c`${1hS|WHN3ky99|0mS8 zY=eY7CYjq9%*6$F$08r%!k#XRd`KJb9McKgwdH~~upJM0dALaUsXQWu5>lu->S4UDIht0_?R zjH)Cxs!wMu6|J26?%uEGe>ro(ADs$>!RfLE5QEn`^XIpJ@!P~;Ga!mO3zv{wZs{Kv zUOtXk0Ae4dETKpt078}nAvFwx>YKv)!Dv)Z?TW1TfZFi;o*DFwnKdL|RPT`wmDlK> zq)e?>C-E;ST7?br0^Xrn$y)st-*PC=E{-Sd9xS3fB2tm~JP|82f_9e(TjwVqS~;uG#ElQoUa%r?{CN7Y5HK7G~!XrEH-< zFF7hkfSe0PdL5I;paE(POC(s2g~bVbRYqARG<^UoCaBcpNfBPA0CS7%L0}x%6Z{`b z9#)YCD8#zvTS@2e4ZzwQOMO5JeWPbJ(wPFx6qwE*4R2Ol^Tlyc^#rp(YdkYGCMhi( zrHwJQLzPNz!tb)l7nx#Bf}e|d=VYc&vvHMl1i(f#T%R4Z&4PJO`-y}x18hT5@Hw)c z8^Bz6U4vz8h@+E3DYsH2lZx6y;y~^y&!NWcwXbInIgLTx_Wh-Dvy^XMTigQOmH_93>&9VWaTu~%q)^JW6iGl9!|n*w zBhdFs<1|Y|;JYUB>giS*vpMDBL}wPi!8F>2ivbtFev@v)NStu-($Bv*^{{odt;^a9 zw#r&X#mjNt|3E1a?3!+oduKl$qPW02`ZGP}KsElNF#Znp8To!Qar z2M36J<|;Y^Fx((=_P~q|mDzN>S@+UulE%#~bzwhbzlEX6r<9OO&Z442$x3 zy`srJ$MV*X8d?s<*LnS~Hu;#r^8Kn!XUV!31}udZVA)D3LB?Pc6?K{Eo0>nlLvfwo ztSgEvgbI~H-v>;G&UCa8pi#*GWO|1#fAVTxha!Vc25G&IB)QNq;%M{3X^k3cfNWT0 z#&*eXR;~W}&&?=_*|cN|xy%hEE*wgEVu6y|loDbqH<0Ngo_9%lUy(m84)(8Aye2g; zGSh23*LWHg8zP|-tPoTT(k7fD=jUYvRw`nqLu|KRaaRzdx+KN$$6e;J*ZFz+>+(Dv z@cOEYd@=H3&>jo`Uf=4e7sY#36@ofdj&?J@jaMBE^dWSP|3a)9N7{gWa($*IPgFxr zsB6f%;2QF==b}Z6nm~RnQZT=0z;c9 zWg0~`P*D)W6({o%IR+5!ESX%+$EvjsaA!Jb&{v8VBntWz9W(@T`90*Mx8vd>77za@ zRw9f09rcErlXFJo_owF_A1$)0 zq+=b)9Wr!Ix#MH#-HvG}13r{zvz_}~sH^llsnrNNPZuxQ1PthKJV zfW(E}8WzbeWDh&0u9D?T_NsR9J5e1HE3u26Adfcqr4RnJ5fp=$JMHJu^saY&&7Mco z&-N}PsoZ>B7q%QAN967^gb~Pw|S%X%*LRHUqIy< zRIB+An7Frjx6ghoz4aQte?a;aSIJ_JEP8v$RdG(pzBwN{wD!qHtPI@~J$>y!ZXCK9)(>LUky7DSTTE}2Gyf4ikE?sKt3Pu?W7Nzuf;7=CJ!B*eEvh9 zPC9J?j&WM|(B~HYcy6a(4!=a*6Va^0MDjt8BHv5Wa_=6m9RFN8cj9PwPdpgsv@Pas+%$OYMmOihA!gLjK88n~P@=x)XcHQ3$IZjHY5GP%JU_1i{^ z4fd(C6KdE$V3cP{X0Lo}EN+Yo*CDXDF?rfu{7fKZ>45EXE{%5RZ`h^mP<}<0$*~a> zEO$HZa9hQc{nka7!)p#)*kG`fR%Xxxn`LMm#Fz=HF=DSe5^Z+L*ZCOKNXP=*QcXa1 z3o8WFp=4|G0uBpyhS$qBlak4C%Bp$syp-{J42|7ekOqa@4L+^HPEaaM0Rxf}($3g7 zYi%Pw*q4%_XY2?U_U-i1PQz|%s(&uKA{Tamvn>{(4U{sOB1u$K#W(AL4zpQTFFquO z(n-})QGDn=Ud)8el54OSJ!UF=?Z23O-c)kH%;ybb_j6xAw&GXMJHzGrDUV-$7yZgV z;;SZIv+gwf!2^dT=6W6_Ns>bon{|80L((YfT!zFMqEKV{t}Y*m|%yBd!0j`v+xTH)#|d{!ycNE2LZEBGczO*JcdNoox^r zd72%0^pok4&C|{Mn!ml*c8F|yVfHmgEtZu1loIH}i?H^oMFvgmC`-iS zzYfB#*P#M?VS^l#Kbr^kFOVs0gKZ3MTy{<`j%b6T+6Uxdz+-4TEt%Tuvv8{>U2-qt zs`#3`&n-t&9o*-ZI~f`L?UppM4LU;&&of@z2~Yc%+_+_3py09vPb?(|O<^61HG*XS zc8_zay|TP{M-=hAOlUi9)b>EYBZGcODrH7tgI|MR8MFsIBsW7^lzIHJDSNbzw;`jP zcRS54`!L*Ao!*ky4E|;VQ_%mTy(H6x)5YZ$CSy0H1iJQo;0z2e^(vSa&l~*Nq>K6H z%5SH>ep;8pU z-Ow?na0q@z2^kA}!9MY$Kf?_&|B^ko>Uqx~8)vSbw#;vZW{t8ss6l>$Y~Wp%cJuS0 zU)`vHO)&P{RFJ_#_TATh&NOky$L9=Y$5Wi}G0x}TZ#N3eCS`i+caM?f+l-JA`ZCY1|}Y#Pu`7s>u( zSxAT(7}vjbvxwxm@LE%AG1aA%axX=Ssi+i99dr~w0Q#&>y6E4|@)BUr()GO)^Kz@4+e z(+NrmeauIws8--ckDdrTpo(r_1>HG)>&xj=3w=+KL_sCpCpf}87PN6f14L)*W$h4A zi4Uz3Y~~}yp+Sx3i$abA^J%NFV#@unY$&#@CY8Z>+@L-m(4jp_4h9_wDik7hySTx*ET2M$;cU{?JZ7zAQ=bSk=p4 z8E{GZ35bX{>w1(+yki5p=(7;veG*Wos^V2@PXJ}h9pCq57lNM%_VLpKv*-fZL!kLN zNsKbnh?UwWfId_8T`fk2<(1G?U>ZG-cW+jv#3)-C0IIQq2Xw8p*X!wTgku29uCP{W zv-NW@@P0et1f}g7dBulvW0}t3) zl}Jz!2DPIXPt-3FbqRM&wiBDiGiPDQ1-1fZ$SqEJ$qp2CZSG&kBE#vzn<^F=P9u$c z@5ta@ry2+eV@GVXUpljz*(bMoiLleO*$NmZ)ZSYBi!S~0sM0zX!_riNJu*m;gHj67 zA_#G$FrCafSaO?r8-=Imn21I2WQN3vY_T?fv7ar*7XN;Sypi~SZ}3b>`x?__rRIVGmS9Ug!a zhk@ehz?|UWy0t>zaZmldH4%@?=7?CRb=FVWIPu;0RfV9_{72IdJTh3HkOCT;$LV@u zTHsxJ33)&lL1n@*dbxMar1ewQPk9^YS7z^YXbu0U5Ikt?o&wJ9uQ(x?`pOTh|HIlR zZ6<34Nbdy;V@L4mU+ zJS#K{_yf24R?X{o``aQcOE~ZfQ0$vX-}DW_;>kB(j-94|B6tFo)duY$VK;-M9L>DL zf*OJsV@ZPkoVbv8eqO4!`uhp|ez*5v98mj2Y8T{3J`}tuG6OWE;mwC?>{?9Y?I3Mr zo2DDY8%pO~n6S#Hn;!=%M*VJ;K&rh0VuBfxaxq?GQ0w29-3%)i_R1<1`2j{cA1I}; zsPpfO4$$`^lDy0Pu|hnB*-o#FESh~r78875>atDCoQ%;+e=7Ji->npxwPmI|DonaP(?SLJj#0ff-s`s@l>y?K~MCJKgliC!CbM{CR z=M+l%#nGZ;lE(~ecp)se$rm}iqD80WjdTw=K@$8y$I$m4$n#zWF<%sd#*`sme*-k5 z?1?1n$sZH4tOv8>`vd=!&YoZnw6wf(_Fa7t-(GzhaS296X|mKz*u4D&SF8DLulERT=)<=S+R z(-PpiQHWxFXp(hP$P2PIv9yq6YaapqBbKh6g90}cY|ovZ?tfZXCqZt?KCncGCgy05 z1ji|%+!ExzbEl^;>Hd|FSXxU8BAVoD$t6j;21Mau475f^f|rJj0pILkvj_6oM<4we z^~TV<&wD1ZeRda}%m0L_2RgWVaTi?}em$gFcVk8y6XSkA?11=U=yLZyw;1@YK$>>qPO8UXK#+etXWh+9CkbZ6= zcdlzSfg^cnCNY6huA#_E$Rv6ln3_7ZKMVw@@bB)xqh#AIX_2}L0{DlKvp>BfT}~;NQY6}PL*&$iuj%*X1`3ar*S==maymSX|m*g_}(wA>j7NWTCn8) z52;Uhtqbf7GpLt)Ulb#Ui9U<|5Eha)-bK$(WifM&vR#T(ucVB3p84Pi1pCi8%ZThi z7$JKzZhPf7j?OhZ7y(OHZzua**ugks;b0u2ltA-WO+__83O7#K>kTdFkhw|WH%v{{ zCJL}$!k{h}U!wI(Ay4BwOC2q66xX#w5KqzN}p1xD+K zvwHcDUq$87v9yLumD~er9429 zeN+^ZyDit$E4uk5lMQMN^p;HSW)1|R^uk4PCostZMLG!CG1q+8%)JW6t332m5CvK+ zN)+gu)Y)_eiJoOttRhDGzO*|6Gtx;BkATvrSqT-l6|)Blo{oALWyRqgG@3X2c+3t# z+X(34dgz^vkxl=*q9VBym&x_!YPM8mnPB6DxEUPG2G=i`^?I8S!FspjPSjTX#_xZZ z7+nq>qdX@rh-jZ}P~Vu*sxF(J<}X zMf#4g26?;$4^+XDb!b)#Qzbx>z!>J{@&R5YC;@1N4e}>5KAKT0+46D~Y=(^CD<)RbJ+emF z&Nve}Hlqfi<2aBE8~M=R|1*ers-^*?!c%wziaH zxo|A5+QK4~P)eXL-bqF6@X4glh_PVtqIAvNBHugo3QdFXVrV`8HoXGGzmU^;Pk%&1cqJF)mQP6+VzURzy2MNDjw6ft9|Nc?#Y3t6kVr0W&egZlqJn_bK>Zqso+P;khsINMQBIlcj0cfu-FL`Y7q%AnEr8lVDcdM=m5Tak!J5}rEZFrEed2;LNsF>%Mrv?_PnyUi zCwOy~L48Y}E$Vm6gOY=6ZGmh%-8XBEXQl?JrSp8xL6={i2>p)bf;Mne^v&Rn^v>;_ zTO5(X91JQ1R-9I0hde2=J3_z4Cv#GPCQ)#esb=&g9@(%*N>ru>)kf$~zt#(sI6V`J z6>V^P3N!dC8Pso=ykVSQ^OL+Ey$wA~x8x~Ig8!$gem6`fVyQ!e{90HoeU-s_h&&cG zx@SHDpSd9pdoki8#pw|k_fEqveXZNmxi~E@?ESKEVGJ;##7i3FNTH7mPS6qPwB>5( zjM)UBp*J|`b!)|U|1fEAF^XGRK{mOt@i}N=e0ESu;Hk_7!G#6?pS>@EYbrhW_K4?@ zycn_(Opbse0Sq!Di=m=UoauhsPTSkwuf4ao%WW&2uXdgGW^8AATf}`~Q*i-hQ5M;h z#SLX20hduh5m8(*5gi8y!39y_d!8gJi9~Y(;YRz-{FIZk1<(6G&-=X3^8a6we0WvS zHTNz=TTgqRo^|7mPFb?_DB1Fp59;1o{C2Bm8}F|CVn74kH@lKXov&8S3b^K@tBNwX zCV9e6`QF)mip!D&&kTbMvP1kCn-K64>){Aca55p(hNk=DEX!zL6BA;kz}BcXXLNV5 z{e>%lvNEW~Evzny6HrbsDGqbakVhbpeOGVw09^|Eo z9|aZ7Iml}l+sIQZ%>~*UYp;Aw)<|e-$^edx-^^;Ek zX+kxBHUALZMEAR_=I2R_TNQ3sL`@;M?FhVZXYU)rde~|TIGx80G1p4;e&e248U6cO zve1dGOq!XMNu<~~imU?lN;P5y>eU+P4Bm=JWXRCd`as7F`wPdyHY@VNTR7w(Crpg} zk0_USEv?A&6-rKAZ(_x=)CHw8nd%+km45S@WPKj7k=s40rlbLTFcusjMSmLo5Qk*gVJJFnk~J$&ilFfV)54kRBoM={ zg3;Dy(`BIh(dvCWB0CJFD)Ypc0EF|JAYhgph@}Y4%1l5{`J^UVnLp7c0 z{sS%#G>0yTmVMV~GIXD^A60b7p`B; z+W{8D5VS|C;U>>mwt3|1+Ft&AZDTIC;XWgMighb;;Zfm;hmOU0NuvipzB^0 zQ*bW1< z+%8T1rC`QCm%d~gbZr=_pNw--()av|D$G@xjt40K1s7%b8;){N*gX^} zrJ|aFIAo7^+IS3MX+P%ap@$(4s>{xkBt^R0!m#TQCr2^T%8*LcKxkEDyk8sx1+4kP ziwy3_6Sqk-=jebg1h2=LzLZ|+nMPMk)>ed=HjaC?^Xh^cM@&SRD|Vi}k5;zK5Mcj@ z*91F{PHq2gkO?~_oy9REXB+|Yqd|+Gy%Y=Lpv6?w>38DaUJ5cdopda_Y*IJ9Y}^BO zJugFi?2S6<@^Pukd(+!JHU?fN%f>CCmQl;c!3xwfhcTYhs<|vlqd!p|f+FYy{*k$t zBzL5Zg7ZOZB5sBrimX;2r7@#BvIc1ckz+qhhPaXwa-4U;Lwzm!3*-rMT+)xhxDT;l zb6%@Svdmz^et( zunyzzxJYM8O~?@47X6+iIC07cs!;|RyL5^Lri)}M3JZ*qnF8M;;gX;%8j+ByejHlN zXi?lDjZQ(0_o|s7T>FsTE$;&5&?*rI-xfF=8yw(ZHUa+c!C?riF!n3wrvCX=ryuTh zzc;;2x{^6UYf*LMa(Jq;9oqfQ@%rV!+UIZ}m<=cgRG3}o=nvM7#89YsC+65VOQJug zP3l>r?la_yjS0Q=VY^k3(7i6pWzbOcxkI`a%jmDo?1Yrx9rvp=N&;56B}~hJ{?L5k zRu&!cK99}*csyZR0$B`9Oa?9D9Pr6IXf0N=bO7&P8Z3twhsAkwq(ZVf+#K*@W>sGJ z!Roi!et;W(?nIvR7MUiV6)BhBBeC29Do*UIL#5~s?tSr z1JGSbRXz^A8>(#y(Mz8&M|Mk+!f?JCC8NY0O|~OqIA{L+oxi%h>UFb5W}HpmYZvX7 zbkRl3j-b2pY+m`Ky{aNFWScE#dSPauDyYW6_r-Wi z#RM+pCQe*wU`1b{*lU+B61ZVsPo5BI-8M>ZhITlpl=_7uX`6+4;Vm3;P_(7^-z_OF zoYs%ELRx_Xt34sial1A%)-T6bTkpGac7`}D@E#qnxII2usJ*Gk49n)_@tXOgWE{Ar zcOBAnF%({`TZf?*Iope**Z;W+1(1dd>!)=y8VGh1XF=^OGLM)3Ami;+<;wAC(9}^a z7%ddx2-;>iOzy}~L++q`Yku}|f{$qdqn{IciKK8_z&P=2Qf9V*$){MbJ6TlJa^XS{ zF>0w0`?1M0%?&k0Y`~J4kr*vFH9h$6 z;P05s#UDHVc#VAW+DtB=nxW%7#h#@|3l)_^V}V0@7-;VxYy47BchV`^q^N3Mz}ql& zz@=PpNuu*lqql)J@zO9IWOqVOd+X%s;!995f7uHJQ%Ee3{$W8)Zop0N5>nUB)P)N3~Tm&umPhV9Q=5)rSW+_Q1a@Tmw8n$`11JURW)cs?3AHNH>ik(@MHnaRd~HH<41`KA1K@aSQJd9efaGl$#T`!sJsL zfiTV!uJgmv%qCeoa78B10%FKMj|BcU309ai0Nd;Zw^V79tW(w#0rkQ z5FN1jnxJpm<>t$lb^1=bpxSEY^Q@!TM2f^wQ9Hu#MIgAbfF&`aN3ls&5)1$pf)A7y zJZe6$$v)YP*XQh+V|Q@F>leRp$S2fflrsD>*OMKujpO`@8Hg$=7Kq~ZQBisti6q;k zr9|5kQ8p`8SuL*=VR6fDl{Sv4@x{$61fs&AK^TGBP^FNA#_UmzZ<9caJ22jdXF%1y z_#nT=SBuoydBUxJ3k?5ZtrRweU=8sqk2-Y&^C@)JL&N-d<8k|dYBSvIH-UEHa{FKN z=32VOPHR`JthCpAG}1aH{@vwT9+Uy2t<6{M2shS7rh$pcl|T%qNd{t*qqfxMJl=MV z9`}p5nU~)G(wy=%nV0X^Y&k>LzBc9sq&o%`S8k(N$SH22qN)WA_^vBb)%)fBK|8{= zmAocd@!V5T)Pz3R)zE>^oLPElix2o!P&AGu;~7)h1=@C6r}&3`rorYxf($%*!e~Eh? z1O@vc@B|BXjqpn^;$&8+pRe|)wv?@O+DXTX5al}eyntPAJbvTJPh)@J3FxoSoKxn3 z(i>Qpzf00QvsKdoS;jM}df%RiGyLt+PZb4IWBeZTn>ESbsr%6f8@~F%I;!(uSAS{M zJEt|fC3^42Z`AvKa9MD2@?M@!4x8)T;jg{S239Y4%)cy1nAI6rO`0`#{`G_SuM~gv zgN}dsp!5gvZ!h_m1OKX@`vL6j`_98}!`8O1d~omWGheg~io?g%6MS*_`BGFtHk4D2 z;1BbDb8M{15G`I7Tt^06Ua1fROhAJtrFe>6O_3E~oS?fSMwlxxxHY&~a#%j`#Y18X zN?v>i#~ihel5hSdV3wN+5g%$>{!ErT@zj)UHu@VWHi;r@s3_D(&v5TiESa{Nr>CoF z?2R-M+jr9Q_eu{=9&HJ2TVU|i3^-=QhrFwG7z|+r#-?QcTwqDlWt9yJiV#%=(#0&U zDj;B?r(=Zs--O`TWnjcgaBUMgzPV=5zxcIy{>~fdEe!hLW6ONN^VuTj{r!oRu zuF(XtP1P++^@|PL;~g8BGJC+~BZ%+n6pivCraZ7}O0qy(Ehqt1xtwY3f=*eVB8$!j zL8y7l`9~rhJwH2-|he6q=jrn7LbcJ2_S+t8|9%+v)5)W{z={!6R1yr zvS)6p@;)gQ-HFi4i@ml=@hS(xVtBY1VJr;3opoFcG1pLk`$64bm>R=x4tFG`hGlIAgJ8tBHQD7pmOhEoE{v@>OLe?XQL9@n zb1$$n5ZGs#&t^l%59mJC1vStQwiI{Q&dc@2tUFfaC!LF(J86Q9{(EOB*kv1)yB)VhxI6`>06Jcwco_j0dEe9Je`(ecyDV#3&@C(}R zSH1E8(!JBLyS-7+>z=AC5*8sx2JUE*6)+AS*XJI#%>(`1m5$xP$$pGkp^N*z-R1<_rklJ1vxDk|JM1el>O z3X2WX4yan74$4@Uin_$rK@USK-0niHDhyZ=7Wav_K$;`{9CBUo8Img}G^? ztfRsK7#xRoc&Bp$hV!m6S8VHOf7@hiW-K_>PAXm-XQj=|W*(#1BNQ+sMja*beis9( z>9eATL?3zB9~@k0L`+g#)1>%8sVLJf&@?t<@B%YUq<(=y0kd_cMo&7H9 z`~gY7OON6_vqE;9Bnxowz>EyAq6NJA@XtVx>VPU$9W5-Gc0}F{&d+7ZcKVLIMS4J0 zG3S`_e9(rebwL+I8Q^x?J4JGMEe07s4x4tq9ru12@mI?(=;uSc&ig(<#lm10H<@BL zP-Gny)d|tOy{gNi75qlt2Jty)CW+>+2uks*^17iamN(EnFyFRIqerbR8it1(1neA7 zJ1(A+xkJAQWtdQ@N2Q39h2Xj^Jyl$dXBDpFOqxQNt%JnoJHC9pHiKq)}hF|#R!)PZz`a2VgCG^;JYjr9wlwk zTz0+hfD5vOH89W_>4@k*XWKKR7sAo7*l_x|)wh+FfN@%DVg&*VA%VCMdYzLL-TdWD zx-dgs!=%xLZrG}dDMjppzAj0Zv2aTHj^f)GJ_KwSPdiTGWMQbSqU8Qpou5<>w6zAkLdXU{7+#wv)YH(Lf&G)C&emi* zD(o2AVO+uq6=Q?C1^=*-Xn$M8=2;oy6_E`QcO`?jjTNkCm~Q&=lwv47y(8^%%Mf3l z(4>5F7S^;TMqW}h&~f|*_j=**-)= z8g28PLw$jr`#24e^R6Z@FZhr6FPNe$ZC}5UO?GjMt~fEZ&q=d&@RW1@4(RG{Vt*9Qq6utqbRUhcMq8up7ZqHn{X zcf-z?;OS-GYEz%FWDj#%PbS9<7@H^-e$%y76pF8>Dhnq-zM;ya8qx?k(sh2?ipdFl znz55>0JoEiV zQT(K8@+ae`R!x#3n_dX@2iWbQqnl@LQK9H*0y#5Ec||+$F=7~a(TzCagX(V$41d+} zk-#qu&lA_Gszf_P>-@5iq_$1kP0H2UW~EMvd(&7QbS@_aUxTF8Xhq##3w(@i{P5FC zmB~e!miCWCl0A+bG&A9QC>BZ)i>RnVPiVpls)zRO<-#u4#UNdNj$Jt6^1GKn+2={~ z)d%UqGX%(#Xw*o}3e(Ba_f6#M7_j}O%aetR0((QRK$7GhX=TwXuJJFG6nmA;8d-yH z$M_7Iv&}YN^g41r#_j!e+Njmf(?5>>wFx*uznM6ZTyx^Wxy5F5GWRI9haz3Tq|fW5 z6C)Zyqyg0x`sF!{_DEzWy-r*z3$7WZuIGvG(f|U zF%^Jf4|vgTIaIz$@s4{Ztpr)MPLSkFQY5+M0I5p>lcjzLnlV{I^k>_ZNxWtT_hK{s zXRbR$cinUfyz+C=Dn(q#btZ|5d*}1HHFFa9C`5zbghy*9L%V$?|2kX-dTqQO=v(Tf zcgh=qnK>(D+oVdtbtW<5(>D*p`&Jdy8L-?9qJd8<@ka2CFnc}h{$P4{IK2^LmeuSi zvdj`#3C3TLtO=+l#=nh=@ay4gNR4;NtS03%=fbh_k{0%X+Zb3}?7)wl#=v=3+~1$n zx0q+LN-lveF=Pq16}J;NA3?3n&#dE;=wSV1L;ue zDr~^dRrUuNcOn}QIOiLEvgj(31EeuJ@W(5^v~^Yu(&R_Izq9aLwk^lYK1BZqz(c$yi)qOXGKUm_%JA}cgkyrNL$0tqOn{xIyet= zSs*8iA-+6uqYOP71IJXESP$_&=r@Q=;vEI5Kf`qg)SF%Vb4z(|B&oi9udOrr}Tv=>C}g2nDP1W(-{+YO7&*351X$Z(SDryDP<=u5J7qx0S150?F-rMPfS!EtDpODR}O4UmC$^J&( zS?H{-6x@5kO&SQ;KKML16EU|BI!NU2#Gk)&uBR7S&nO% z1FGXt$MNKbaC+|d+P^U@J2fJhvpTie55NEV*+aSI3Z4s!oOk(Yt$er(mS(M*wNqL| z+nHtTBSqt^+hDRXCCTb^<$Aw)4RpiIbV(9l=g^LS^8w=s4)cp0_$hSCfp5AznUx;= z`XdrO)Jn~X*XgNdGgSh`#!_S@71bt9n7VQ*GW7zNrgE3Bp4N|YD~)N;Y#De5j<3~Q zTc3$EvKhUvI+f}p_dfSW?r8V5*g<_C)E5oq!A4!9qR7-qz;wCCJApy!7)LlDLp*m| zkYH~9hq#FYc3DTlP*vlPRf~>V>SsOQa^l1W!-`5)jqhV`RKc2$?!=O*n}W0$;#**> zB5!onON|Zx3u>hL-V@!b(+pRLjDe1nq)TbP3SDtcM!IM}ni|Ab+Pv zibs6N9bT)ZnfXj`ThS=1=5JCr$_^=yI%JjeWF&`$izhd6z)$PA7{anplh2ENW|&-& zU!0m*NQ#|!fqcTu?WmzxSTP@ z_xr=v&5ZZn!?t@Ychh2&d6@QQSQr}0%skyUJYaFc(AaV9(f|1q6NY|rVDneWxp5G; zADqeglwvz5qNAevJZe>_8`1#*upU9TJG6E9ZSvN(3ovMkBi2b@0}7xeL3bqvp?Bj< zkg6tSk0cWnwEa50{+$Gao6Nrt^45QnV@Ne;|}dP|eq;?tFC z6J)R(V5NGLRL$(9VOeTC5-kL;q)ob?z91TK!QZA?+331Wg8P81ah;zI0;p#`ZTXJ|AN3>bNm|I=D7%SR#Pim^y7u#tok&-mBih&?emyct{0}3w`b= zA5j&Q0wu&9avj?_`fAg*!qpN^j{eLEu48`9>{w@6K5V5*7tQ@1Zzy8%&~~{(LWRzs zv1b}6fzUhXG!;4%Y2tK{x*Mit=kj=L9RlNP&*jd;JD1xKIIYE)cD^>mQhSR_D9(uk zx>h8_o-|T6L&0dDdmMvZ2kBuLc10F5jLewkzFf%>ep+C5c-gFU*GdoV3Hk(`Lm!kk zy7ut0#mnF=$Wm|d>Q_`tp->qhK-j@+IxU zDJJJ=@%uijN!lwG>#H#Hd$<+a2|je6doq&-eNfTD)Bf1e)J3;@XfXh& zy%5sKOI05T7;wR}Fagyot(wyAu|fux!nh;VX=oDyYYMAI5Is*+@ zsfJ+XjH8sg{473&ayXX`vNgPwh98$qWvOb}D}ZPHCajevKo|_ zfT69MUi9|ob2BA+m{1TnZPH}f$lI48UTyPTwizJvYewS}H&870x;JOC2^8ws6;))5 z6N3W8{sw`fkYe*Fl0!u`2}(8Zx3PtuIjT+SCgq>-GYBt*?3=RBHzn*Eba7xumhp3= zXp=gbRD|@=AaUybnE&iDW49J+^Ed`fw&BIBNMT5|{L2WEnT4IB}&!mKjnu zQ0zL2B%*M_HE1Vmlr_+gW_S9mci$z!Ua?A14YQPA!_+X>9AsIn4GjlmSi6No_OE#N zz)4F+Pc930Cq{-9V$v<1`KrracjZUPjp>WX6{Zv<53VTA&FPnKQm=fwbdzpW!;RGQj{J=pGOC6Y`|~ESXF-5CjTVG&EQ74p|CYq-RgZiqST;4ukDeafHa}@ zJOM%QKy1#B79!6}l46lgMDhqZqzGn_ZV=&?(j@C_N>IRdTBl294aYN z)yI_SuGiU>5M;R^s^HJB;eYHs@(Y;}K)&In7-{%V*ty-+GFE0KB!yBk*u8@*!~vIQ z|Bt${wv{b)q@UHx@AsP6k?|j|*l(Hg;bKRec->+p8Fc(>mn1vpUIjg+JYF@uM7T?S z(0jMI*8i;N+?;db2GE_&c-`(NggI_CP_uP#?iuFE=aWzn)T~LIaG%8R zkxtk)<>{A0%KVHGF(c-;9Fb2GzPs_^aC(lMcTOc%dXAc8OZdss!@}OsxS%KbW2D+i zXJR#1fg-O_w2gO96f@I!9FoJ>MTsQ9hJ@igQKo;4uzzYDG{r(ySCgz*-YCO1+v5^E z2M)@eEPO3Y$o znG~BrkyMa=gdG28suR%Tn&7`ZAVt~lvOH)JQ^)IfSshVL-*xYISwE#Cc<=0fm&}Nc z@CR?DD(_4#^{po*fg2qk9kF-dhcl?QT+31xQZtF?7I1aqWim9( z4odu_QS4@lY(fT!Bn8x)0_RYY`ySv8H!uc&BFbeFd415ZUo~angw?YfLpnhk~YIbkBV1j51wqn!*I`;mR|8U8@*bjcG4SH$e04ZZ`iy!ajSGx9-!Gj)W)Z2cFkyf z;PiGn?@ay+KN9h8CbRL&;$QWVCMV7Se{N<7u2AeninO6HQeimu8yTDk^aZfa3`txN z#Nb<03mRx->C-6|@-O)Ttr&JUXVYb~KB}bmOw$@@Iu-HmK*k8Y0k{%2ND<&kkW2Q> zf)ha?PEyF`Ow02GuLGIWaC82AO#3Y-3x!xd*BH_xc?_*Bc)Ct;8#rL*qt9~;iWRiz z<)9yfZcn2u5qDvM6nZf!if(^=#m>Xd=U*-x!EeRu7_r^Yy-(c0|IKpR!?H1!%gV`l z-yZO722V{-TV@d$5Ih8WAS}|nB}0Ov=STOs$2#&mYyr^7f9uQdcy!6q=^iGt^NR~) z9!ccptvK;wdZ!sAGAI@>l0rotl3$Xb5XuhuQT7~ffB0=t6Bbs2{&_8_0Y34)!A0sa z_vJx%q!)b)Jy*_pF!8FFqczWs<75j)o--vcAL3!3apH&LzUpSe$cNgNKa=H7JT+yT zjrvB4O`^yeT=-xy2uvE?vi*VU{5}Dujb_h8F)(3@uF@?&D+y*HM$Yq3t)| zG^31V{=@wXzcE4Mqxj`-lIu#{ZndO3sF99|+#_J>Bk5-)AN??ZBO5{eou4sMV`E&su=bm+W$_KSX|`WRSg@%O~v z(eV3i^gb8dq&()+DB2}?P3mZ)Dp!x!n4qT*o0LorI5G6jnpvfz6nlswbyU>$FvyDn zcjBUHXGO(c>Eq`YP{?GKO(&R&YGZ^OfrJWa7tf0R0a_K(KAukbQIDeE zrIVLYI#Tajs?jNKDsnwbm_u~=q*0vaxPjx@;5mAW6F4aS>IdKXs>#&+{&e#9$j98A zWG4>kT`@CErzp0OBFCsGoXHX=?D8!Yoo6!08h*a2Kx%NHfYLlXmVLs)YGet~DIS8Y zX$WkVfTlv^UXTofxfA`WY{q$x^O-kiA9J&}JxokturV{t&Tl^1^GQH*!w zgk-8by5(mo)9my_e}zj_&xu1gR%B$d)M@mRpgI-SJ?Ip*U$3MG0xBlAYEt|T$vZ`v z0ddT3rkF-fqdExlwrX}l)x=GHz2^=(8?=A=X0K-Tf^0g~uiIPu7q_ zOuZcUj4<4W53?ue8(FwTeSgW=NE2{29w%NwHJiPl zGbf(6N3xynnUf`~B_&}0{ z%j&SeGIG#(0+DAwhXWRF7Kr`%YMEta1Q!eB#CZiPbrgF84)V{668Sf06frTtGo$@P znLMEpxW`lP$D&~=mPHb=XnAw}{6bs#ETdAnJpc1b$qH{q&=S-p}wfG$L?w5BeO1ySOo&v$_ z;{^9M&)n{@gO|wP4BhbUFs8^lN8Wd5ERE?dQXcL(^+8|?tSp#g4>R8ZKY}}+d(6>HgayGZY)M| zE9ou)zN^}9S*l-b*nUV5)T;KYD?_SCvJ@FukoZ4=z>=#Mgh={4+66{3v<@0M(15EY zjm~kenX^MQoC#n!qxHVHd&ku!TKvC?zIWr08PxGTC7xx zO`*s}D(WU^D%SX;N(vTBXg3CS6HGCzp53X?BBc;kKwwKksuF2_YWyo4+!p=h?T0KJ z`StJrsQvGs{qE;~{*7cQ#V(>qv=!i*b`Jx;wq82p@>l*-X#$Am=+DYYh7(UoRc1q8 zOtCvCl21ip#>vPBWMEEQA-fb3KO0;Uogz1&S9*R@$D|$~6IC?sS}QyQwK$7q@iI6) zU#O?E=L@v-6#z=F%loK?@9Q==W4_da!I&Buo zO4IADA77FrQVX)E1T-BRfUBeDrL>1$q!KztxDI!3`I! z@!weZ&qcnG7qFUN@4L`9JM4+TcUs`$fJzY-XWn>2H`+GFW*Ch!j{H|`*8bMyk=~dU zKb73zwq$YM#TZ#;R)6(?V((F;hl)z|+abd0JFL~!E}6QL6h!p#dgOa%;W|AtpabL_ zVHs|u4oGG`Iwi)&wro(NNQ`($BV8aUXSI0VnT)4pL0;hSgak2`zZCgFh7^iv)GcH; zl=s0|m%VxvkLJenkhBgyM9)s6TS;sfCZTia&%;aRM$a~=cOOtK0sYZbWrAmYc#*J3 zmwTpt&p`y_AK?l5>xdC_@ zrm^#d%RDRTUiWHwEbFudST~PvSY_F|edgs0YoJKseA{TOQB_G^$3a(QT3g0ey<46OEK~ zNJKXppt1{6qu>|K@Q%FrbrVR&_WbHIQpXJ>PMnawY6g`F+#%N6i)Qo{OOHb$<2v)buP&?_0kZ=iMbB>|oRwGvFD zpqfq-Zx1^Vpo41OynsW>Z1s)lSp3wgi4HD;+CH53;7s@%*JXP7-heLP_cmUS>FsNp z3*PuG8KhW@Kd*=5x218`h_e${gY1@#n?qY9t9w6m_5L4k6Rv)IV$5lhJQUZr^H%nd zeP-y%qgc>T$fTk+h^y&FL8|h)Os7E2biJTz3RZiSM%?7Da%&f$ez6v-zFK@r6+qsc zFcC>op&FRhRzhJ-qW~{RRX&h*yF02BZ1|9{I;z7uV14%bOYpUnaVD^6^zUoQLT*OL ziD3j2)1c6QBE`m0WEBk@nn@0hk^=DmnKF>=}U@M1?@TIl7^ z|LGh&oX5%eT29a3d2Q_fJ@Q{)e%oXzW-K_>PAZ%@FW6>gDUMO>5sEZWQOTg3fUDtU zVX11P0vwb_sI%0Kbgm>-xl3H?t4$B9@i{N*b3g1_Xe^UJ<&ZW-4gavKw#cnR4dQLx zvSev5xft0g%ko5^mx=Y0bs+MVJZq(<0`wrnRU+gB!kP#?gtyP3kzPo9ov92t72Zyt z0$CjBfeY@HE_dVZpKt*BQ^eb|!4>J8fIj9gT@@ev$b=`MaP4==87IaQv_%f`FFGhz zN0BR3)Kp8xySo_(1xi^pz&8OSSaj;G6iI2I7A=h<9<*h>LR%Tn(vzjBe$64>bi1gIDfK<&RZVB}c7u^vJ^QSvcS5PWfPX)7 zwDLThhheB^=74>ihQWD57|k+0I@0PV_Y!D4OW>niDNJOYV5<`NF$74dpx{rt zlb6m^^K+&l>FTmD?Gbq;-KSWosfFYL6#X@Tc}evHfl}j54McZ%O05`;*UCt3mKv>K ztFlMFSB^1etR=+0+A{D6*NKY*<3fsnf-hCAEep@0Q5+7WFiE7*7>H|hO_^%;*y!5levQ`@%27^~lm zx>Q!KW3xPdhG@q-jkJRwW+5LdHU&!sfblON-t7PJL<)*18ZPIW5K$2dYNN}l{ z!N{Umupntv6cTXG#|(mTlF_b|>p)o%%PL9&6To3Q!{9`NDv(Wr3nB9tx_I8N17UkqMsAoEv6anAh!LAuhoiATV@QbgxZbe)@VoYDI@JX3WYfIsag4V*z_>1 zTG)kFWQL}ZH@AvrD03FFG_MHP&D#KH-^B$nOc*`WrZ*06n3`V`4bJL%@{o!ugBC)ISl8Ytq`4g5Z6k#85^ zo@%laMqa2+f@TH1gS_{_NZ(f^RConMntzozb6T=m>Gd4)! z3Q7izug!oNbj@gN3f^0L{aYr>abrf!=j5;xr(`?)d%Pb=D?>`dJ0L%3BkXKOi{%Vy_h;L%X5G9DHC}Z*G?Fea1NzF#(u=}UO@-Ta zNxG!SBU+drc_C!8%JE}sJ{~rYwE1NY|HR39P${C049lgl(*{$lEQ1fxl_5Ijf?JYe zZD4j_zsuUd^*;SByL|^n-TweSQQO#e9F79ZFW`oX`ag}GXj#4gn#5hL)cU7~^?;BA z@;CKLdm{cf!mWl8dpUe82X=HmC-{6h?^S18F_5Yovom?1%2^3M#AweMHjeTCLS6=b zGfpOA?3aG}cHyf&arE)Fs5(RYq}tV<31o2?lC|ljRnl9~$Uf>RyO+QDrd7x2yzo!I zg&QhP`5&nenieDPmyZ7nS?$CN-8{49#b%1#glnNF9k>x&q0gd3b(MzNc%9<7>!)(4 z=z!`&!!BgHDByL|m8xXmHW(!nVlzCRu1%kgphGvf!GpKx1OHduXe22zB8&LDe4+L< zNpVti1ax8tT#%g9=D5Si4T!C%7`bn>?M|FfG3J+FU3l2XWHj`1LNAdNCpH>oW=12Q zVnHD&i;7ySfId-BdoK*%z$_F#kY~|(;sa_!2hMrXd8iJKo2rvvt;DhFKf|BNZ^ zf^~jz{7%ts-yN{GOVmPIht;Y+TpU@z%KRgD_&mLW6OP9GtbzTi!i1v@i9u({dTw!3 zCr;JuGDBT9#RA)4I!5rjWmR6-Mu@^t*i6$)j(L0{I!|(@t@g}O-v+L&Pu)+?IziIJ zTS4Luw~=ydkM{|Zs!kg3U{}X}SUp96LpJ9ATU4ZF=#)z^#EAp!R#+&PkE<2rM`o%o zDISJb5iHwJRc@2$n5&`#t}ACc7zWRcq0P8>?n=k*px*m!nI)rv(}qs1P%Y?b^f|P} zUb}o3AX9&jV&xmVNv#Aqateu#=;;NcIovi6HO?819bL1#bDCMke*NF?%Kl`USrl0c zA^Ch9i8X7l9H7`o6nQ{JwF?d@(}A6;$P3slz~mw?5$N{G^n%6ACrmN3PjYWYqPRZ1 zRj#M2>6Q?^G)tWVEeA=8D+-WfGRfQmMMjdMRaxw{9om+%RTXa4^2O{uQtD->KHAB{ zKDT9IDYM(*x;BL#67y#ul>%4ol9WLrHCl-M3hQ|2Z=@>E6X;hkoP_xn=)KGhI8PdY zT43eO96CEZH@pNI9PUj=w?aquhl2tKy;9R4UmoxowAdMcyK4BJ?#Ntb86=FMd#^}^ z|A=b&KKRO=v)kA_@$1oP>Aaj8Tj%lL{pS*6xc8=hGPZxVNx_wFX6}%7gW^b*(Y_&p zK@SfLRgkp+c)j*cTFIcarK7`qoD7t4`}Mp( zOK8@GP|gc*VnusCX*#a6pK`W!gd#&l$*gR;R~pOiR_#_@aLW#J^q>zOEp7l9e5GTz zt0|gNINx(&#!KES z9nF*h$KkeT66 zmWwzIit`4z?ri_vUo5!=Unw`&3f~|uvsb%S@G%g8vD|IE6p)$IA|+0e;-LD1PYwS# z2$5+k+%AarN4C>Q^LEwi&WysTT8uZGC#&5$X;f&e-~l&NHQf{1=Z^P_6UT-nDM}$g zu$y!UN?;0to9Gk?(@GSlXCbfKNeCVTBhMYdNzc`8*>n}bT*V&wR+yZ)d)31Pzvatd z^|LL2AAF8=@c%UL&Hw9Sa%&16{Q4sj?Zm!us+psbK(VnDSxH494r)~u;Pd8)ck+@H zNFg|yF}|ly+Lqyedh#f4SRei&Jl(NN5Bz!GTP8^S_{PZ;vY(r$>%?iKRx&n`EDQ>!C89eBSc#j>h%B_rq5+-Bi!RWl=c05?_PsxrexV_U(+kxsaam!N6`RA4LXRORg zXbr%vKv_&wmo5?_&++IW)*MVW;n)0f`?nC&nhYjAo+J1jdB8+eJYpr=+B_IELiBHcm;MnNzZtVxicjn2IX#YVt4j zMVeTH8*%`Gy{J{M?R7_@So9sbAf-Lvf&!a3TjJSR_fjgq#_OA{%@dc(b7*4%bim~k z=1OR^uxetGq9PLWq&vwLrZC(tPs5sZunjr3x*FELh5PR{MCI))hkZidd|R!uzbGsQA~r(&b`DxVwEo0O&A z)udUofmt80QHtBAHF2XNzG@c|xDBdZXIpn;{^O57y)xB=5f`y~FUjI&cbqo{Nj^41 z&u)q>p~z00v3jA>CJhJ@S~cljpS;-|Iv~QHv@`sIsjV8lWQVB4t6!ckS?zfb-bq*$wND z(S{=eWHm3B5GYgwBV^>?*c+}9)kIU>+u znA9jQVqgQ}`%oAa4G9DtK#+SFpWAR5&$b6l_IwZ5YwjjV7ya2}PQ=pxj3ix7Y)<0L zYD%NWvX3d!M@1FPy{b4p_pUt6_b6F5Ztbi){#|qtvwU2u2Dj6LE_mOI?2^CV9DYZ- zp4TP+Q?n+;Z~ZqPynE-%nR70N_X+x4a=#b4j>{@~%;SH5`K zyN^D^^t&Ye=;$|e^7Xuams|55efja+5bZSg9c+Rj_*uA9-S*~!~Ya0{gxj*`|dcg$_p z^tzYK+~s?p;GU-M?flMZ4aX(akoLkJV>OiPW43R|Z{oJo`phS0P6zzLgrOx{Vy2Kw z;|LgsLF9j*Q|xCHxj{vJ*zXc2?xuH4=mDk0-iS_Fv!_n6kblXq!fkhCu^6WXteIUq zz7v=$Z_yV*Zb!f@lBCF?n_Txw3&PUqvrylpmqQW=3zu+pyot?~%>M{@SPa(~PXmS) zMINl4dw50gJsZAoR-~tI$#SOE`kxhDoVi@6Qy409vE8UrrkC#aKBC6TaktGAa>)F@ z;j5qS)jj2_SxMV@e->`p|ry3Vtk=otL=L6ez72GMLp zIemCK($)uYDEGx_GZX))51}o|RGqdN#tO^pa%hw2g;uA3+?BU#&d{|2RAx(O3YcwV zgE)bY%@O%@|BO`SeQ369pqGqW3UlgZSfCvxH>Tf}cPh@p-g^d3p}${M6vGre64vJ_ zQl4y#1d^ZJYb1YiLNfKkzm#v9Xfja&zpI!>HafAV1tp$?Z0B~01-YB8R8*jr$XZYa&jn=jqKjL1k>!7ecDvF+t@KFXP9g(utvR-V7?oDfTEu4pC89 z`-|##&B~MD!eYTM)}yCG#CZ`2mu>Rz069-=z^kO6h>52vcgwq=c;c1}g8^lrtEUG} z0b~>yaA}k3J*vioAgAP}0`=~4rsaVw%4XNio@sP8t*854P$8{oTG2Gjd2ENEax9x3 z)(_WVS?XRW8L-nbjfCxa0ovfuQ;cxh;0CCN1MUYc*{oiZAOJ8r53*G86uX)tE3i5} zhb|=@SeBDEez7do!{OcBLl2)V_!;_2$L{^kzxmJoy2-?h?fKPbr0%r|KVLON#A%8> zNs$v&)IY8&41P{7(cw`) z=A)c|_;7>oqc9V$79ede`uw5la54B0FqUsR#_iVXIlOGIkEzJ-+4rBumb3r(JDZVev+$_cS^!SerOu zmoJji+?cMtqUdu^qcx#!hT5CeeB>EqD5p`cg7U`i`nyezCTWVJWCCmo>}#7PTCmV#pf zpDWxpyEck)AiD}#*XiBe*wucRqSgLyu=6Le6>A_9dw#%FAp+TyeuHt+D^f21tp%^ z#qKu*(8jdd^NQ%I3=`jpk=0WC8~PwB4au!_ei#tE3e8dp{2XZW?35WU+Mu>MnKCSjTfwvNl8y@n0yLm|Usa3I*=b!ezqiGN|l9m89ja zWwWmFP8-y?GN-K#(dIIXg1hPKOnFeU6i?`v?{mEv*)OkQay@&cpNNu~YC19Usyudr zA$!pz!;<*)u;>}sThSxm3d7(q*{Vrq>Vld*w=>3bbxaTMn6ewX zBMQQfjBKKFBu&xt2F3n8oTfbL`_xZAvCN2bNisNbMYEOsd6}w^cbmQ>Y1VA{Y9p-L zup%DB;Mny6wUTqfOm&B-431(4POIi9iHEFK8m$jM>{`uRG_ystaoWei4R73$;<><5 z3?*wLHn*7%`S9Uoe!r3WwCnhuf49_*bXv#RiW+2=x>NC(SL2(cs8tzREE`=xrVFC~ z%P9v9r}f8g>p)w-gJX_xvLj=zFZ*ey#xw=#!zLw@1LMe9GfRJzVh>TIj*7yxGPV>{ zOh#=2BjQ**Rd_>+5*eG8V2!(Z#*&u&%Ewl(42%cPt z=i~rmkwJe^i4prmP`=pX9UGP&)=6W>e}x-XKs1UHCzM9q^vr|T77H3W=;dzqkF~9X zX5I6A;Ssw>l+&O&FMV0r!M|r)DsXU_m7MoqMy#mA9VW$I2x?Rr)fT%IyVa`nAmI%S z6Qkc@ZY#L$MyZ__p7?>A+_LM_X{R15eVhkWscPtvN#v#aVJl59iDiqzio#aNjuS2R z07IiVeIyJTf}293XIzZzQ`7sa8V5`%J(VD0}a$9d;E=j98kprstereyKF_t3k{u( z2e|--lL;JC6`I^*$yM>1SQ@J=7Lc?&N(zAq2R3lR%v9f)ZvRxsWo*pOwGBXKmyF7X zuuuM!@H)0OTxHq?2UNh%9{yO?NmnxQeuu~^<|MsDeLWZ}gt1P(lfLMM`(j9LK$*uK z=`~myU1h4}^#sY+ySx%3TSV3JYDjA>8JFjK(F@Lr=#Jdtc|f&v+#;}N-Ev0t&lAXTjy6AfoxV-c|EF#z`?T>6(cavoNmDh z70$cHob=wUa!YIBwCf}*W()&96>fDAO@iLgE1+0tU|vg7^g#O@rj|w-vbF-&e%sFp zr%zAj2H3(Q-xgc43Ovut=e*XASmO?mtmjqmSGnn+E-yEr2KcqG@3tD^_77>4KfsdS z@)_ru+x$w=r~$`z$QV6`r(^2i4eOAJSev@dvft!+csQ@wBUZXfvecDyF>{7_BC7s4 zlv|)G?=)Y-N#e9^z{{}h4t*XI^v2xAKdv2xJ(-b)g;*%j2 zld;NIFNYqaX67L;gI6(mnWtWAn1AkuE}VcL_6H_SPvrIS4CkbWHOi8uOZi36Aka?7 z0B0KZzMUn#?m2X(LNCQ8^dv>Kpn-05#hTC-f>odeybV*eNZ)pt942?%@3_}Yt(kg` zoFl7+8%2o`r@UHxbil1?H z<4}&+X_G#D@4J?%WiGK;C-z*dB%-5(*Sv$QBCVS3(p>LN>KMNUaWa8+ezxiVr+uE_ zYVO4BFTd(N_$Yr-Mi)S(9B&selzbju8i6EN`{yp$_HAvhIU;OdEm;e#9oz2@RZq=M;uFbqcHmU!<;2+uG=Nivx7nJ#@`S(b6 zgn#PaCjc5*zX#q|>1`f4^a=$B*u5Mo)@+PD`)!96CqqH~IY;@UKbU~^$EIIQBcF{U zE6pmUA5rWBirk~34oU8jE>u-hYNLgnqEaYNf-qKq&a0S)vin@7fj&29i>g|FAfQ{4 z6oyr=-K1Qt#X@Bx>+=&fa8%B~Q#znVc^Q8T2pK75=a9=D+++i)l2uTUwdBE6&>8sf|7=-GMVN*-#?HiqO1QP>&R6-OdF=?0jrs)HUv z)a@Rr6oK?oIupl$W(xjls91@eXjO2eg`|XOQ^Kx<^eAFME;&iD&JW~EffEaARz`P} zZAS*DQFh)%#)IUDnUmejB?&fGtSpTMs%; zk7gGQqjb~0L9iJY_TAp@t2p6;nt37o`VURB*0~>*J|xXUl_fZ@*(2R%E|QL7uTbP7 z71b^48 zSta%K{P`u!MFuTdj<23R2~{wq5m=6g%em`@ikP#8)}}0VZorM{4KUH-nl3lIgz1GU zEEH<6UU_aa^lYC6xt(bRoYtY&(KBCfX@0Ea`HThG^XDUv0uEe^ut(k+q;2=u7mQD+ z%awy?Y329iu$#-H=5X6#pKAS453ve%-l%)p4p&$u%=f zDHOYr0x(5&GFN0+y_*ya9zBNpbJZtR>--MM)5QJqx}Y4oB}6anjy%p?M#X+iaRE6R z3j4o_+m-WHee^(XLPy%9-~NFlIq_P()NH)7C>A2TX{b8cCM_kPr9R#uHi@YuP>jPS zR2{62j8uq7;sI=8k~+TjAjANZcuoFVBt~4#U+32}-jVi>>&Tl;6Wj)q)7jk2+wVF% z;=@fi5xLIYMv9ymC!pRmi0QDJVk;=JpNhgIcDta$?V_X~7TMPvm2=a4XxNHiA`Haa_g2JZ8haDU%>e&=_7=X}4r zWP2qr=7HpGFS!XsBn6=mi1FJHQYPyam&zUpJLw!|m#j3nody;=l;HsV@5^ri%N-ez z#dEs@QAT-Ja5l7a-d!ukyP^c=1Ma-_abSt^fyXYcO^`J%2$|sN`3@_rSh$!Kgs86^ zDu^1-K;*z4vW0;NOXZ=w4#=TV2OAXTgqa|Y3o8t$Cs*uo9tz$w$Oaysv znr(B}y0q}?_|=?Kq-TB$)IWF9D7!LY(6By7)L=&Jhz7H!PhZOT@2N&`bo}hq; zAAChJ?3p|i{)!~9GaL>)zu9Gi8ZE^@vRx__S>@OMb|?6}=V+vW!BuWGmX#%Wt2a;2 zhT{(w0F2^!7!J_})l2gf1Ah5}PTKw$VhbLlH)8g(+kUrf{>6>&8zJ(W_TRLT(+=#s zer(d|S11N5`Y*zQ7VzpcMI3OyB0MMRoZqg>p_i%hIKb!*JhyEljw5%x;i@RXLnun= z;%(*h^Das|jrW~V_ASZ`+pXN>z1eexXSWy&S9Z|pqR#2qvf+wT!iI%8+?5M< z(5HkKJj=vYbgv}EwV!7=+puuoB0SfyFlG9_Ma7HoK0Oo|8O~dcv@rtaxfjcF%UWnz zwl<@LeyH1@pBggF2(RY18-GiZUbE2&R3}G_PFWNKz)vj#)dl-n9&! z&;R44O)f^5{7HTBcVsoYfy#mBKRG4<+DtLY6xl#UVgRO`)X9caq0$s_K?nU-1`S31 zaRcj1@&$HGuX?d-*@5Q8`|Wk)y%y0n^C|!fiS=hy&&1DM*SV=N`N$CJviGYt>R)Zm zqa5`{J0G?-n}mF;-yS)SLp&aknO6J z>3aU=uv&VPpihpuSG%UIASu%er#1v`16{3h@g7;ytfq)TNakJce;R%^MW9Szm&c8u zLJ}{E4=DGkrI+Fj*vwN1`R!B0Okd%*6qXp_Kz#u2E{9K?wL;cA?ry)jD zU?^-*AUh$H24N)kq^i~zlMCyl@et07@{bFR^3U~EcSv_lGu+n<%L^^sQsHr6sNW#0 zgypC)!);tx8pVj1A7YOp!yT5f^c$J@NB%}Eb$tC!3CUyU@;Y!(x7Gw@6%@0FBBhWT z4$uPLJLxhnq_V8$98jit-{c;b^>H@@!~pwC0x5RB!qF)3I+)JGPhh(h-=5zpFQf6A z_10+CrYLxcJdMM|Y`WP#f z1%`w;-CEpMg`a~I>16p{S+cwqWP>nAzm9te9yDeeK`H9o)$V5rQtzDN)$l-JfOFmr z7VyIFPY0SV_tx3Ae5_`F_V}K%Sz|Wbwfydm;x0Sx>!`^n_x@}d8ZAfXa>%bxGT^k_ zHwqNi&xfUpa=E)eK{8#m+Hb%qD`c5Vx^E98$LZl+Xo|?C*LkP7mWjK8+Ah;$opS+i z8^jZlWjBG>3Zq$$bGC0f9m_llLC2_uo8*nrKU=;L<9h|n1}huB0yo_k|J*z-4!YfQ2vBILDonI zoz}>;&ZS<(9t8r_d+(5HgCMflD#senWO=DaDbjEv2Y^O_g)v>yE-4aBV}(tTECEt{ zP~!mTm@2|FiZsCn&|Fq$(+*oS)Bo-#yG2IqW$;sOl6VJRv??_5hBXwEMv=`_dYFjytT3%vdjGWt=8| z)cVgKm@hjy?6_rNiOC=ddXOI_c;tD;f4gMOjI-R@g{PQ?V6+EK5mD1qy_*y(1eN>* z(0W`AlIoR04B26Noh^NnQG09*Fd0J2sEyh1AS;+q%Nv%c%?TbH7A6)b8w`vrL)^Ko z@@Tg%x>9)94^w&@6b%#3*iM@k8WeT1ER}{+u2dHUwFh2sPL>~;vE4gu zTE5Gmlfkda^C=2%20sN-XC5rr1j(6(CNNxgj$6Y+S)x|XVZY6s3Mlz4^LYSeNaG0I zjqE?G@nbgSdD-sB_Q$%-3O|$k|L|74w-G;ugWuglw4;TI9oPY`G{MwvirGm4*Igt= zcQLTrEU5_IuEKJAs1tbLTtjdO>ZGsBflOX;T@1CedXOU4D3I?C3#aO(E%c+1I%kba z-2zwLbsm^yNd3+Avd<#7jbJoZv{<~)GK}QiKEMB*5i-qRzMVtLMuUvw2AP1XY-FBH z4aHPbq>_r<=)Bsm6~g-u#TlYUl6cPW@1}^$%(7{Qvu%qtig?aNQY_aoP;c;#`u30F zAsqY^sG1EIpdM1^5?sVI?6=BKeW7}BC2eTzyico4%jjHce)#-YPiP5u_t*)UjU8ws z7`Atlt0fb#W;$ORe_HT2w6RqZHV___b z7j@B(W><5L)AjTvrdLwTIp$msOM#emhqL0L+UaJ=S%LyHpZT$b6+R{>mHeXabsu@q z*XSjCJdZ);bHfk`8q%_0j0J;m>O1Zz*bMO$T=qfsg%aNcqTb_~J{5t4^#S(9%(x7R z#c^Z&Cs+YxQtw{hp9UFSSlJ)d+encE&uz|_IHiXu28N;vD)Omk4X>Zq!jBddhxW{` z2)ja1HZ)xX6J-B0ke|~4CI)j1@Ve5F7Jjzs6S|+5>WaF*O%Z7zXH_O_;iHPZMsa0! zJKTXB9++geHuTm^oy%icHOR>B3Ql%Ufi^)Xe1>htrg0(YQ#ZzBWBl-e7FZ#OntdR@ z+q~9|P2#u%Cw5w>j6393AUMUOg?C7gaDnxsgmgEuwG(hDu0_^Mkriei1dcU|Eu_{j!?jLY#XC>8tMIlbkVsxHMeeCq`R1@FnU1Hv z32??(@hOrq3)zzDWVv*_2y}kBWqZPE;N&4c^vTx&-=q457(ey2;t>~9NP#;7rfJrh zK|6ra1ev4mSOM3Z_xnQ4r5j%>y$TBgjl~`<{Dz_aL0ABN49daU#;#cLR7YH5RmYFG zexf!lTOp~gjc=ZN_aY=5IIL?Sm3Qq>G4`*^{(O2{or*34;4_q*C^_^4Zt#gliYz7OWjP}tSG?+X%(;&?Ulug zI;GXTo$eK`rJf04w!)cJ*ctWBw{Xx#;IfRJZ~Zvg%X}5dVJA)&R*Vdb)9Fx%U&)VS zaQ!F;RwOZ}sT%|oVHCjSp&C$6$I{p)2F9JlkUDyQFeEIxk6=?!d^$8$2pYo@O8`Pkb zfgp!&izxBv1wQ%bcVwyXo+$qs9^TU@PxtK&ye+d2DA;4*g@MM_-9OYGD~m#9|LGt8 zWX>Stu!E2VP8dvhMSk6rl6pR5@b2-%sSbu>)D4OjezLqV7`cW@IoUK$k+@x?kiviV zfyKx^9V0|WHer`-RuGx=w!1voT;t?LAJ>7iT`VY>0$ zgb^%{)$`kB*Sz+Ka_N)fPCR+}leZ2MFN`L5;1vyz^R#71jOMIfRDNR%F*+k9PTDhM z)o3YC4s18FO&pX>6q7`eL<8k+KE%$NC74ZlncL->rNTK24nc>Ns9<9Mk!55@T4g{C zXc>8%-M1~H#l7ag&c5#D9hE@2kMrCT4?XmJy8mjILJDj(4U;zfB0kJy{hCaYD7)CQ4{_WptG?$rYGtYHkqXP6|BW67tDJFp;agY?J zEQS&>9BguFU|LpTfmwfGF{HLk5O7Aoz)B2_xMN~AKPJC_?{)V(@;F*vpIepR#;J#S zcl^^pT5HJnKmr$B{X?r#POA7f<+$A3D&OY*gg}-)hEZ*ngKxRBtn7HB;TZYLa+|g| z!^(I}mVG#Q^xuts$AV@0dE_d)R7eM21^mRscyv+>EDN?%k?673($_fgq$Bu-$2Jet zXjGR1HPF4l&0Jk@WAFjUs@1umuG-GvYM|xT!x41?Y4tjw!pU)#sz{~_hSJw-oOacr zS@|xjsRF@vW|L^`!qSjF?xkr7yc-^nunS55YK-63OKS+`N*$hkNmwsU;B|va+;&c% zs}^n;bV9xf{6Ir&xHKeHk~J41!Ij)?-kCxB;j5wS3bGGj94E>}jpeXfUOqV1xpOfb z+@gqrFYuz-8#_8jB6Zx`F97=MVMors?sBnNi{cnvKp%aEq)lJX?F)|&N`xLf;rtk= z!Hy9;3bh{GyW-_On7Pmtv2#wLq|2knV+&`SZ>(Q8X$tM*zu03hJgTuD7*FSq{Zn6k zCo9uL?OGGI$b3q~#;* zg{-+;x<;mRYgDS^yqEi9fpj#l7M7#aeOGv16GK7*XyO6`1IBEIXn3<#+dR-=(kM=m zoxw;g+d^YbQ!0>cbaTtRHcpKeY;j)$^0e^LxMq#@j@@zy6e1&Zk~b>rApxUCf${~g z)X)md8{pdbw?P#`JAK#>Xa6sqzKs=R<91<&8n6u2Ib9Dt%!|y}*dhm>eOo9=!%{SD zkRhO_Zjg2dA`48qC@VNwuH{^Cz7aG*qm!{uSOCLvFWTWuWp>Y7<~lVFyI5#JdB)hu zEkKEPRfq~{?f#Vk$?_&<7nqPf*H0jQ1qCBs`r_L7(X!}nd~6V*g<=al~$gVm= z8;7QzfA&5_qTiUINoz7RZJ?O-6p5uGn-npwxB2RfnOUMH1@Ja?N;AkYuXtXc>%^K1 z=s{~CWVy5LbEf2vfB%1n3#JQ_5sK9R)67p_n0{$%PEtW?o-~A)l*vd zSf7VfZ2SC=D~{9W-QziTxfN8BYgTZEs3GD)1Rm2U)`q7FTFBGD-9c3_v1n9Y^~?~Z zPP;u%8`KuDY6>!9YuxSxqF1*}fR!29s&;M=0(d4?gA)@u2t3Y#kA)+aw zQ1X=NqSrH8aSut9ck*?L994m554q|IFUGtM$ds#-_WPg~XR|yCvKtB|cn;6l3ecvg zvjMzK8?@1$WqAM3ceC2(8NHPyH}hgiz5{1U9yc+C`zdBG1(E=fE&N!Q1YTT3pKA*D zkjEgbB&AKSbVKg#R?;JEm93hxNfhOuFHYqgkgfO5^UtQYPD_j08D)8zy7Ktp z9rLOWhn)vmsPO2Uij`d{lJm+=8uV3bJ`JZxpJRB z{zjQAjVe`gLV$@6X+cd8jVTolF_{8mDb~C~tmL4oi{QfMDjxFG<=AlMY+LNuZW_f3 zJCm>c>{_h3=(q)K@9vO_up44rEla1Dll5V+1ml{?eG;lJnpO|dE-aoIVzsVaIods)<2ueN4)3`J8+nG zpNYehPcgX^$)X~62H*0?g6ujq#_fwa&CE@C6Q`5@023QJg(bclfd?v)R4mDN`2h25 zAh$+UL3i_;CAEGa5St8od>ZdPo(4r5JK4tZLt+g5ZN)8?tzzu$$CmoI%yFr#>k=DZ{fLWUHbDxil(K1H`bY0`UQ5;bQzK z^?q24+7HFjV5czMrCy5Xj&L(&uyY(?mQt0v1F7+dKk*h-|Bc#5o}BCk%)p`&@7 zfn5tQU+FHtmTrVuxSh<$%bz_ci;|!J>No2D`Oz=`<9GihUP&>_DH3Jczb|dUJ||gW zg7S>{*|yJCZf~*2RhI#$I_FIBiaF)V0jFKA zDUz*0O%XlwyJqVGuFTUZ27U5WsnfKIUJ@lh5zu2K6*P%D=XY~2LQMgxls83Gf$_P{ z-xVA!xU=A>A2J#i%g&MY3op61BXgZ3#vfbHo`E5tyKdz^__5c;-Qrp(S!!LNQ5;vC zVshzoV+sOV2^>>n#CGo`0I_t+vhf)>M*cxDJoudY(Ls{&nnCK238acC2I^l6AS34M z*TsLiCI1uVX14G#2#mqrPP&f&z(d1bH+{3`5qdyYuh4J{Lw5UixjuAlkAUk-7C)9f zzocIA{Gwuy?V!5X?_+aoWGl>A>ie-Cw~U$hyaTSi?wlN$lr$t4k?B^5BT7^*1I1fInIY>Jr+Y+H;SX}6E{1z(AprBH5zW_gEi8c zAe?Z$!eeFM8McCl?Lv8mzriv*M)PsqGC$NVFq)Ykf3$EXDQ1`V;J{Ybffm@I$52LUoC^Pe^!Z_J+R(C%2ee6)6KeTD}X)9uCHYjcU`lTm>?=z)yz)dlQ14=Y4KHik{yy5nF%w z{pXZEMqK>5sAfIMWQPj}9(+JwWCSisDP}iCc2beqAUdv5R4PkFg(0OJ?U2#Q5o27f zJLG`Rso9|GD$)jN6t%t`(UjD42fuE-s7a`8;$zG;4)l6cCK= zncFSLf&kn1kbhVle8}ASa9DF;!JWvR@pN&$qINMZ5~-sF?ZL;Kaj6D51y{`3;y1xI z!%BkzP%w=w?JeTev=L0SYAs9zD1YDr2RGB`%f9Qer*Vpig@R7Hi?S3=lGYP1}$mfPNs!4Nb#x`l1;ZqtP||!;(ZnLYKV~G)$0W3 z-3_hS!M0e7{Sk&_lucVBf6}A}Q{F!Hx>p}+T}!zTf(I&WjY0#p%I&i467^wa2Shs! z2lbGd{nT?JT~C(XZTo(;&>|~SH#z;=Fe}331NIbuM!nW8!;#WUPaW|(nw!k2anoZHtb&xq-3j9dw zQegdrd@yOM?_ZV))O~c;qH0wY5CJ@0ydIL^4f{8RZF2kfja5DIqV=_VkFe?q$4#=H zU;5q7-x&Rn)obFzNaqx?Zt}>Nv!7!6DDs4g>10nVF zst{L8mkI8BKAN2msU2lrk7jEWk3#M$k&n7tj_eRR7o2D!BiEsYA7Bt5aRD)<1r9}=1umU$H@AVXRz-YV?!}oeX z>TxkLZVbg@3~b9OZXgbVskRfpzQYpYb~mlVA8>n8@l;nMHY)z%E9c3&DIkw9@}Z_t z%w`I{e&ikhi^4~)t06;djWlJ}O1E{8Sync?naShyhHUn|hUK87Qy9SJYPIUIKeE%OX@Ut~l_!*^q{NreMvRyUf!_5{ToqsS2|GDn4}je6j_)y~@LT%ZJM zIACAd?+aN317gR6J z7j0*5$#5T*{S=9AmELe3>~!u8C_R7E1_Cr zh;KRFCu7!CAu@mE+LPCRd{AMH0IP3c2bI4_j^>yPP&w>0+k&K&ftNO8*1-kZpmU(` zJ;XnG4~Rf;1Ep4*6p*n*f5HJ>-+aMD=)_=WjR7B4_wjfxWQC8(oe_$<@sL0{utRHs z1PUEis8!7uK!xFb;W~LXUE;6C&`r(+ga=GJ)@lfu-e&i`2ai9=F;A*yQ_AePUKoLd z*b%dw-4wHvBKcHg5w}wKgv2pX{uSbyxoJU5FH+}!cIsz!Kl|usOB*6S`sV5P_AEvo zml&zWEm^*_!y{&1%sh=^X;TD}zbDI+gK9Qbtz#0E2?`*nU4C9ls!(;58iiJ+L z8ZxhNT`V*|=EFKp;FDKZFgq|XEL1RKjnAIY^FA4(YR_uVi$v#geO`-;mboT6PbY?A z9?C}^dao8AJwMg87XTC=u*c;Isin6}M^U_1S-#I+PRjJVBAWx@c~6ZG63^SQ^_Jy< zh_~Xu=ldDwLM5HW(InS_=R!wKEXH1nDWga+6?uv&=4@W_;a}H-SGsZPRTm8KVXZRq z*k-Gm7$_LVLgN*@B3S}Yy}~b>#yg^YYw7JvKK$EyE}YMX1X(0wxjny@UQK;4T!mcX zds=~F-`R9A>5$^;LV@y?<;Sn=yz!yrW!B`CAN|q`ETd$!+KqzpkN@(z*C+QtYO6jt z&%K0`$*tqZbD{*foE|$Qv%d7P51q6M6R&Q@=2x)71U2{enzG*+t%*qTFQMeYXk~+r zyQV?nO%^bsCNupM>7ybK`mGGg10I!wbg!h*yUAmLW1+yX7&`dhm0%20_Lg>}At8U>&51;AH8v;-KFUv408Y8mC}- ziNDSTnV|}UV!{WV4*EQhbcLeAGsJm7KQCWoAY0G?b;3cv2I*lx+#gijRYHYY5R!Pq zVu)dWh6^jn@_P4y#o2U4m`0IF24wgr+*z=jQ|GKvtOkt`oS&hJfFXM`KHwusw>V)= z<^l*Hm6KM^VLyxyy@qQT$}1$s{B?3 z9Ty+v;Sv#sbg|xJ(5XtIpNT?69h^t5x4=?8m8p@V1E`q$TF5Qdn#2XtQA%U7~=RK%5X|}6BZOvrC})gQW=2p z{Xr-EXUEBsX&<<4iTYN*funxGl-H}$YsnG0kI zSVw{yc6yomyf9xdf%ayot69Kk+|Xdtzz4lk!_7-m9o9#&P;8p??dwolTI}&RtR6kc zPxoFKf%J0S-1Xi~5%r9g>F2dATp5w6EDGB-?M`40y^_}(xM~WX=^;ho)tuE+nj+59 zEhK6No-rIdPv_6LV*4sZ7C11oB&NG;`q-Ll*_jjPZ~F1fBQk8(3LJO=$pXKefxNU{ ziqfyRP>rmoHgQ$0=^T@5tU{2<9TT&2d*0KO*S-GmGSe+LOhpVK3w5TldeK^-GBCt0 zCd@@O^q5sWJ!BRpriX_=g_W(C6f2H-WbS%c5co0>N7PF}n;i-*x`AM(24)L}?E|WK z&i_6Cp)O$c{2kZSmShEeFU&ZDaeaI|ne1hkR_nl#wF@Tmomz^ip-43qc@h=@fCLwW zdBq7~JKd{!)rvd*?ZMaNSNu>51jM0X3om-?dQjN0pqis|(aTCW$LT6ri4uY-q=G7) zSLfF%(8 zQ^X2E1B6C4goi6Koe$)ZP#(izt9IF?f+L86iIl0ktY)bUI zL{c0$7z?Q^BbNPgCx&W zY{1Fzd#@thXO~dtlDH6m+ZQKe$-Fh9#Gx3Bi52?Ps2|TMg(R7T#D@wm(ukL42c(p=v+F)x_}PX2eK&t8qa=SDUC;) zB9i4@ynaYr)01}5E?K`PP^!)7_iUldebz2aoK_9{r^u=qT{O0(b#4}xiSc;D!bkMR zuyZ0^K+or!B9`Wp0e&851C)N86ZP|&B8ul{&c|k(A`U1=-ZbF!(DN9#VIjDn@MRiB z>t~dSS9ossYyiE?4Maa3e_!xC;EGouWoQ@O@OBmH4LSAshK0%8wPXbo@0r0heBmJo zjl&|0Y9?%P5JAJPqlu#q6Y-m4|H*tvV-vb_;8~Z20kG6FmRBc>9}?MvAK?SHcO%e>JDj<%petqqH^i_wb8!(eMt)9|@3Y2zIN?E~Sjjsf zuxIGo`u-o)ld>BK*!f#Wr5Ao}Uee~ME*Q+VM$W)CP|SLY#3Hj$F$_TUQp4I}+?=>M z8bxAAp6wMsUfpNbf#TKe*!l`qpqQlp^B+%|`??Mb3JczA{+F?Epedq;lz`;IFG`oc*x<_2@6Atc!eQt3)6k=J(gI9lTi$Zt$|_I z9xI$s-=7sQ^Sef0_wo-)KOqfMNSDbw+Db7uDRPa9d?e}gJ?7O7DfL;4inwdS<2g+c z`CjW_rRJK)I!+%KbM*~pAZ>cq@YU(`E?Vc3>f1w?aB2j*gpZ|-ymVC=ke=gR!+Xi{ zB-djE_j-hQhenYth~?rmWvj9a#GhCE0QBXhpqTl+>*5xEr*AB9744u$enzFTPtX*R z4YG&H^5Ir-rD&T+pWuq%m0%ciZ;sNzYz~=q-V3|zdedR2-rpF^`<=iDv)`6Z{R4@0 z;3#3f3I4ZI%od#aM`BG|JFTY;sKQ$O0zrvKHeDrAV<|_Ge=kg8Fz*|si(U!17kYYZ z2Uz2FY#+n=E>^gml=shB)#gOa4m(;|pk7Yq>J)pqo8}x>)(JNT^pe2^8QfJtIV52& zRtMp+`{HO`2iXAy)mW*NOXtGEs9H?YIVkG?T8$k}K@YCl)?Hxl#4 zL}j*{7@Q=ENu)^pFclfvmc5dinUw*to=8YG!6gI3-D)j}hOeKf-@U2%#Jm*EVNHjH zva>$=k!Pyp6xq4(_WWXx7XA%kLx9euKkzgM&zw~43{D`85R^TmDC9n%5A!C%Xh8p2 z+vU&wor&pmR(6A0|Mw3sm&IU*>(o2*folLQL?<2#*8`4 zom=l6!+9+0l&_{%Qxkv`)*zW6q%6tqXq{#RN%Px{za>fRK;pngq{IXySrn5&5e??b zlnV#s*E}}3EQi=UNOsfb-E$zCQxkO0W4-XO63^DsH4uxzt$0zM07{%;dFQ$?*URoe z85>OOKiQf6eqFln@Z9B{4HYEoje*Hw6PT1yOff}vQIS<&ziP<%J_@SdO%W;F?vP_% z7o_*;bgxse_M;WF%C5<4xCNnS_$T;R#l@mN`pA-Hf}wRG{xa&8{4m^i!>4ShFgC^Y ztZP%mW)dH6xYE`*8Y^s`JHEy(8xKt%yuoo3;}>zGAHHieCM!06Ap})A77`NbA=?o^Q0ETb2eT&_Fc~rSzS`(vT+4 z7%AQXVK@@c>l9*4ceTsd-V$q}{(_~j{=k@C0(MZJ{^7Pfb0&9(jiFeem~WN00ux;v zlM{4=dsg*Kd{dq(HPG~AQaS|?E};_wlGu^IQ#=GvVZk9g^sSMBnmo?UC z;d01eU_4Jlb-edfQsH(@kgdWu*AVW;pE-2bg7xplUZDmCmc=KLE-LTk%}5eDc;<_OU$Z7w%#A7CY?H@VCD5c_)PtKoO^YbdY2? zusd_e1lh$D1A1KrROAkNGp9m{0s1^~oOiBjC|mU|@FZWE7Ukb80kSVypO@NDTZl!7 z%@VC>(5X&VDjb+|be0-3pI~VRyp@}BgA9P7=-yTc9swO&fq6ueHrZqc=ANvtEPLG- zeO3i&R0s{IGV$w^tI;1t$$N2WVCi{eT)@A zCa*k@vS5bMsC=#9CqE}!*#%r2IC#6?!~_*m3^*}4ROE1GHD*u(-Twh)F%TCT%*ars zG}beg&F+M`&hELLQ0QFE(FWDgYSel|D(5U9JO!NqL}9bUZmHN~0hjrBYHMI&d-~;3 zr)C=AqKaE{h-{lez-JlZ#O$OPNDIov_-6|D6mPdmE2yCI6xZji_Jc>r>6V@I%%)32 zF%sI$B!sQ@!NP)H~cNAzp)lL%1Au9<)Ji+zWKK3{QN}4guiGfRjN< z(uXeAk^+qzR-;>|mmD=gkHsB!u;To3FZH@hZPvrWPm5x=5_6fhd+Q}eNZ`t0CEXsp zO_?uv2o?rs#0OLfur(mt=M2iO0*J!FSwjut{|#6F@yp-*B2r8lH z77xc*j(%r1Iq1L^=(33gI!!SrC~}O7ER*$$Tj&=4PDoZsm7EalL$YSSTrv$Il;>EP>YII7Mgp?FKWs-@XXng_bgcjg^>*kkT(N@VR({47w5$5USj+ZBv%(Y-+R3LXx8y)$7AN~8M71NDh2xp7Q?U@>R4i*2mV zxbS4j0Uis0%OyZ*gxur!=2N%BVer7F6GuHu(? z9QCW@bVzr)Y?(bSQ(%|AG4`3+<7TyMxo<{!*Bv*FK48VWjXyO4=~vA=JjpEwc9NHw z&@pyX%ma#aQjsX&84W4f2ZmC)>m&^U{qktH6;rM%_xP`zlEW<*ADofva?2w_R6(`4 zJd$LHbS`zm+ps`(&hy5sGMI43^HQhf&^5g6-s#@ig5z{En55=_691zi7l5uCf0l;c z7h)1;xp>3pTV)m0eSTJOhG@VkF7OznidRs}=T^84I3<1lJg#1ywFCn|I7>z^Pb--|d-anO_PQzhieut1R7j zv=Ea$Uk6K;Y?R|KyejNohns)==lkX}X zetWkPVov(tI0mH4XX=7maFEof)~N;qqXbug^c!8U6a22wL8tsVaiO(zGPh2-)q9)w zYR*McC##cYxwcMA;nfSf<*W_xM4{{j7`@_vWt9Cgy5TpB(T=%((sP+?a$q}FZZcBm zQVfW0WFW~@CO27LPRhIz!5YOu61lotc7eIfQy*3~55c;J^{ z3OY(=71GDZeh65z?&?=O@sX#;g^H04*?8MBDn=QKCoSL9G3|9HuS0M7n)s|ZgWL#0 zFRewf9ag)Vm|a3dN3tBL$&RSj$TO!NpZXE3go7+7gqoe#hMn=hIX5u`FSI*R!?@4_ zLynC>@@aFd&@%aNPgA};b$F8Yv#Wn4Yu=bdty~i;w1r|ILa~vGM4^E~3ae@LfUM?# zB7%HDC_xRx1BU!7yu_Y{=F`41Y_LO;+0Ay`6)pQ>8fC;rzgz!VGT`K}SFe%cjX$zPSOhQ#}P>2RL8owp%G#M*>nPcG@ULCL0Ja8#gG+*u&oq!3U#56h5Ow)q_^j3 z6vJ)B^Nw;>PAQh%1;MVq@NGbH6eZZXuo2oiD$1CZPUmuV`<$8ngp>&kpS^W<1@MH| zkR!^&Gx4)f&-7kksj8D+Bkz)5oW6>eFtgeoYpQp9#p0{4@IOJ zMHa*va=lvI%4H9Lg0d)lO<1Q8zOSMK2qqt^uBp%#|F2@x62{?seep-1{~|GteI-uX zGh~$m&$qHo#=cDylSGk3h#SpbHXHwMh(PV?MEMS%0bmT#%Io-;8`d-X-UOD5b9F{79I6Q_?cZqRGLW&w| z6LoCRuPl&+@3;*<(YhvTCwtfd&xbxWlp|McNW z*%c1*wdK+eARla~;zb5>5C{3q0UO<~d#{nF!P>b2B|Sz?0R03cuCs*5@K9h~Kab<8 z{vY=*e%uK#c=TjUQs-pFpcv6;#h($0>tahaF%r(iGKD z+NT|wQ=B5}z0YxP_%ub_m2ZNTfgElgSq{|qP(vv@?>>A#YHF3qo_XqA@aXdvfGjN( zkJ}9ER^F<#?y2UHR%V%@~)q~BR zsFi;pTxQ*~x4u1FUCHX%JFdUSuiJ3S$!JpwpZ@&MB#NCU>%eQ<8WRhYKrwL?SxZG8 zRCSRn!YdHh8=BV`3bOwffNdSJZ3R~r?c}?8%>HaaB1?ri2p~_6(-(suJ657H7SDN< zORX|?qugijL;rF`{M#U-m5}{Wy^R!2A!kfpuR|1bkRlaSu8|DxJ_t_XOreE;LE0#*q7N=OOCE*P!qF1ZL0@Ej z0_`$>&pixp7beL50Vx&=mSe~IU@`Cidja*u*9{tEc{MFkw=@nMvjN|&Q z-%8DE{A`h%ZGpJC&aaZ6z{{e`VS-k|IUK4BC=RRR*Kmt?x5*Ap8FzPtSbaq~dU#ai19zj9`BwF|MzYa?<0nAxJt7XDNii^0-$q4l=IS6H z04q$?O~B~bCO-^$1kr+})m}jED?R4C6pB=CcvSMURHYyw*aL-5m2{r?BwUoky=|A? ze*W~w(eKY2vBMTC7i!Xh*vU)G6+Pb&%VI(QbL;F*`Ax`*yBbi(t?{h(KH!|>-AVUK z_R8~SH~T!AjVt`I%*u$~kW0dAkb{7h2(Qy98r*Yj~X|Nbojf=C-)BxTrJTE5h#Z&M2xR`LC7;9eeML87f0=%h*5nRd?1Gf2Ytq$%GA5 z9CunrETp7#x@kdDxSrnfMV!5t3!#pKdr5eT)bMIRSq;zF!||}(4=j*jxe+^@ImOf+ z9?xLoz(&NvVAG&T;2|SOvqZxwSMH~;3e~-m?vTridL@HSeQr-B>U7bLu#{PGFqzpH zfHI%Wk}m#EFU%=umh85(wharMm``==fR5))N`4ykvH4*0Vn(h5PlGKCDOmh*(of@^ z7*a<^337qPOr5R5a4a$)CVA(JF#VvD&UaDYlj2a*=Xwj4q)?gSkVhvZ#M)h?a~yyG zfIRO&TW?wRY5RIlI=L8UIe$`L{2f{Cz_T16z8m3bY^IoGifo`FYsB%K`0!ocEu;%( zE=A!jZU$Ki$gS2Ww)pJ}KTqG4P2>o|3I?|B>zA}=83v=IG)|`f^snd4gC?ORO+sKAH^gk9Z=s+GK?p>W>BASvMXEUI+G2EKu+^!z3li8)>>Rt>l5n16etz zc!K>^Ltk5kAX~wLMLYM`Ivq6EB5_zdVnHOOT$~WrDmw*2@y+rlq?TUV>APDM3%i&_ zzZA$4{F<3=c?XH%8PD&hm$_W_KSt|ZV|m8}#_QT6P74cROv>@kb-ByW_io{9{jUerdE-Y=IR-{hTOU?=gW+!noiQ!Bv zP;KTo+hYha>B+G|$D|QE>{y0QaZOD49HT4iBvS1q+3c1D9T-5sX*9y|EvFbjR1p=q zmQ*U%grnN4j#n5`5PDDYM7DB4rmB+vQ2504oF^{ZXd$9<)bA36_IJtfniyV|`_+K8 z5nJ9$b=JS#6wyWR2`v`_4KR28j12@sb^7TSiAGgzyM#z13dV}47wnRaQPnxO_kg*d z`&yY73%)Nd8mB=%O$#5RAp`Sb=B-;;L{83*5)8~6fHc}x`J)9WbW^U}Jh#cUZ1#qL zrih0jr9f_n$Iq$iAaiD5-d;(~%%bq!E*ix{UKgphOGL#c$gnXgusO@N<}1JX#ee!3 zZOyNXYSxp?(OB#p*rHUMSd>zV0Y1c?R3s8gMNy~?yvlh{*eBoSu7wiKlQ1Av!4eM^ zF_$ZGu5?p=c~MdLSyeX|SrGADDcuD&Xshz%yGWz@5R`#y=^GyDq5)ad5Rflj79KO4wSBLT{+hGK!syM|{HSFtj1qC7G*Mr@Z|=cZ82VQSVZMFF ztX$B4$fa+RTsj9B3GxL{V^=RN5Fb!J6Khmkshd!mtv1jHV3vNqppCoDqcr4zDxL!; zvQ^tW+UZQ?9sfERO2?@41^I$ioU`C8?Dn>0;U??ud11^SdBAqP#HxE7H#Twh%RgND zbE94H`>(g&B5k9MMh+bBk1?S*?4g)$iabC)>Nuu7@Q`qq5VEbK1c%;A2Vv?;`j%XM zK(R}h4MaU#$P;P4=#~t?jJ)#Q(iRZ+>JCw3T^p_)HPTO{`2TA6YFeUXTuh2- zmOpa`j@2%Z!dNpE$`7?c?SYskb6Bb66mo!pE?JH|c5$I;;Rg1fX%B#+Ks|iyo-}vv zr_JDfy5J=52PDcrCpga?UO<@9(T{}irLbP$nd|Ts`M}(|Yub&F2c(gsm)8{I7)jc2G+87Q<0~IRea2S+TvHoZzb9E zO?lhGjgV0}=(kN#$UEv69}vyU;P!^x^Vmh_c&96?Ay=vpScH@1c*j}giYX25D;KPh zYiF*U(hD3e&s)%VHv|;&Z0uT%6H+5Aj7@#mSfkG0+ zBvK?EW>v7tldfoTVYml}D9^iC&m2=a5?Fk?2u`=TQAKtVrJWN_+g%#@GChgf{ z?jBq4IP0aDm534$dit@rL2;SeNk3SSDK4hFxn*9M+&ZKMQ3+mVHod}cmAm?uY$p)C zfybD@%T{&EYv+~t_qnQ(T(FbI1E>q|w`GER(l|zq>LMtmb<0B^jO{0ev~j!T)v6Nz zXaU@`G2p0JqbN`of;LjJ{Fqlf4^*&lGnrcvhUBJ3X5?H*et|QFk3!3eLkY|RQ& z)&O>BNMhCM_%CJ!Sn9~}!m8gGaj|+$d>H9`%|bZqOhP#Q6w^nMC&1tWO7@ijsHLO_ z>iz^?%i{H1b%zv%AUo+Cdh4{fd1@SwaH!PNsL9qJ_&Cr&vtSrR2b~%?!-A9b5bWue zr3DSeF7tf!^c8{x4Xb!nKt+jqk%#=QLo#540;cWVa%`JLR}xAn^{FKpIb=Z7p{{dRd4FF zvqX&pbUFmtE4evA_8yq6Hfh7I_VDlVTdkM2@K4Hp^US4?*?`1>*F!A`C$EY$l*B^D zC}@wAKsYC5dJVURyYj6aZsnqKQQN|Xh?=i%{At-caZAp8U;j1zyZUe6`x*|72YnCC zD)UI?>s5Php|pq@K#u$Te>;vAFMH6h?&6o-ZNGD@W>S+{dJf)I8O_Sc$iHnQCG5D5%$qLf*=uE}y-{XFscc+P4H zxg<_b+u*A%b{UUJ0mQr-a<4QUmM$0<9(M_Jhs7AwXOjUJ7ZrY7E`>_{Pl{ z$`g3BAWy6wUJ{6soX>;qPho)*ghHS-9SA6VxAq;zF!k+l^M`G=>e}2crY_WRwjyrIt z$oBk`dGgC^l_q5&#YGQ6)fm^?Ae5)x?!DbxFNqdl{*xM|eI7XXd#>;-3`TAjU1+Ha zc&2K7v7RYf(6&(B$Gx@?mCg&nmoP+MG2)s#ckjGZIQCSA=~KBhPKs#M^(mw18x>{7?I)?f@k6k(LwQ`coY15BzejlP+gSdD$B*|<+@pC z-;S~~Gnr}6vOZ)?DF)JH%BaXSOOOQU z>W|ldFYaAb13yK2=GV!V`!+?SaI~z-ceHOqe$Pq+hx(Qv0HcbSHh_{p$Ay zoHl*4`TZU5f=2l_PAwjA+CgplTDI!1c|M!Id1~>U1t5$ES-Kj+~v20kbpFW#pE(iQtjYyymF(SfNKruTgV3&%-nQ0dYnI%vCupIPdY9V>1%5POr zk~hqXi?~;Xm!V1B%XAD|dme=VmXFEh)gb$IAKOX`Eg9EjSo*04V}lO*fj5*E){NZ49-6A1m27n+2)0|_OXvP(pxw- z_GdlfPIfP@!^Yd*i~ITU&z40&>BQ2Ip^y*;lw0MxnHW^I4O&M4yzUMO?hw@C}L=*M7PkP1VxFZ6Ftb;flj>#cf$NuZ6#RA+`w= zHt?w4>9d@>U?neS+xTC;?Mv)|{2#Z>PX`K~eg&@qAFG8;8qMIqw$ILidswLTUMcA%sT~NgVPDCBQO15Gy&JJJ0bj{n_ zzyG1P5ru_=-`zvB4r~D{O`N^m6tj~e`BY>UEJ3FUKp;K@EgGcw(&xWAlp3rJs+XP; zu9r4L(KWz}+fGBKOD?@$dd7e4!WMqE3h&CM)x{ni()~eCV9gr_OQIk89mk9`IV0Zk zactViaE;>|jD9OC!!_waT2bQ?BV-nQ_)!bdIWT0JO(1iIVn86Y76;80e%W+^&oeMC zP`;1+07(C9NUZzYHV?=S2SOUG+J_|xTu;lTZ^`uhHW#RcTJKWBQ?G*5aRUWLvK)DV z)wt|X$$tQx!K)Z>0 zqcN=3*$7Xs91=%Ijpd{4@YHwU{qLAN<_;SaN;kn%BE`g0WE~ZWRaQe8GtClRz(bxo z*DIHP7}6=Vy$!Ma)D35>#KfzbvH5vcn3(+4cjuIw%b33*3F2iY2+E-tD2~#iMS2$A zIlV9Z;{`d~1Ub^GVR+)4^mu5LAj?%38qG_m`^5(rTwZj8Ev4%^$h@Qui{`Yi4m$)LGQm+X#q6R;0Tqb>mO5FfuvNY_Xh`#- zPSWPGEi@;nSJL7(=oHU+;87>56dJfI3>jAEy|PtxbQ5Qr5~6_%w+7+F2h+m!^sc!N zU0daS?hp9mOzIx36K#Z<(V8CDF;oN9X1}Q zyrSuGon{A>%A#A}H;2j_G8M7_tdD_-Mgw&Ax>+5;r$}PZRv;U@GHtq$wTXxQKcj!6cXA)RnsQu57$wrSRW|jX@GU(*6 z%dwy;JJQ`?1PCPul>-iO$kj<0p@4gQ{SFwxu*^pgt?JUs@Q=Mk|JJ;k(8^raIU&^%L=(rl6wDh@gQb`3(gwFe$_MgGT9E_L8l@h-!m*atB($mn&-L< zt<1P=nO!d}>!*M|&B)Oronp39WD6CE zl^e*)TFC7JI>N(#$gP7fDF!t@z(mSqxi$!C-LNdN#cdyy|4c}B_SibYwhw)@)>w6f z<0i6w?eS7?^ZGY70T##o%ZP=_xI~EM>s+cOiM(h*p7pjzgjGIN=)55eMU>T2c*GrlukOcGp*?SYXrp|PK+#{Zlycn_(%sB-&A~b-N z#ZXZjwaeVu_dENYRp#F5%-p$i=a=csbf$y2ZwN|MKm(|NAgG8NDytx_5 zaP#C7D_ht~vG*v_LPZ&b)tVi23^^voZl`3whyIu7YH1aH-Ww@eH|cuZaumRAt?F^x z9dSo^I51n-NtT7>YwL6!p{>#$uM2EL=uK#i_z1{gmX1SVg2&J;RUz2zw`^Rku64mS zs^QDc^Ao?c=X-mkOn@_XHVx#^Mv?x>6cUA zc>31CdFE#?eyRMcCw_ea$|caXeiTs@g1>1oxdjoy4>ileDgvuH%W#~5DqEBddCUPf zZp-@m>mR+Xu;5MlQ2O5_ZZN#LFy$TOoBHk4QYjYVfJs!;POqziwQMTX;R2UgtzegA z8{5PzeR)F&7UX8j>X~c_SgGss4X_$xvI;3R+r$p`cm|t|;X!1uksN!D8zN`ldiT(u zED(_?ejG{K27`zTuUfHIh2_y>*=H2#MiHlG@W3h||6FTQ$bqdkx@5I#^wo^&$E4k- z*KM6giO5WB*G!-D#Z5QT-$RBH8B#C{6$?&vs-QT~e3QAprbc;9+91xP%TJER$!P#}{JRhUFOmj!@ zAas}Hx&TOF;r3oP1ac}g^EK0^6&9GUZjm$OJpB+d9BadXkFPP*{JU;3(Xj2$6P$3z z>YsM_ZD`QWK>qy&tST>3{{ED}0>M9=?5rY-T{w1>W@V8!P;4AU)>2Wm>U1auH`yZ$ z)S!4+sQ_W*AdiNRIk49V@fp(*_i`JdL3(<}PI`4ytL;|IW%uwLd&Gr1!Vd{gOeqp) z2DCtH)4~REjWRx>Bce{1u6;s7D%`V1x!3!>oCu8hfwl0fc6OHAJdgGw#M&Ei1O|p4 zgBL<5ukFeJTb&C&TQL{*dpP7e>J?~G%*)Vy?!1knH57!z#_wJ9%1G=2ISR^+aiFS*l>ay6Z8JMb+i*HFp7&ntdE3@s z)xnRR_j~wW=k;}U-ESK^>;L`C*8kvG=I@I(gq)Xd7w#e0Bc?}QmxI3Nd{NkT)W3)J zqJr|3(`#3X`82hq42?xjV{x4 zk|HS<@`9|OSziK6A5M~Xc@wBl?}02A1Wrzq+;Ak!Gcae=CD4JNq(nE$*;p9{O3Y@% zu5bMCZ{kLY_tt$6MHY88DDhiWWZCDYnqh~Pec4E{8z>S7VL1%t+i(^gYmNU=vKa)^rB zEGW>R4E@Hi0~-AeW{pRzAT>;%uFdjKoqQ|+bFfg3{d=2Ac0~leJx}y8x6v?d69>N zIj}9y4C;(Hs!8@UYMUoM6IBR$XU2&>R+S2|XZn3<4#xOHYC?LYaTxXonxZ02k`uT* zEb~87aE%;BgO8y2Ma+A@w!r1T>v#H+Mi(|Wi>wkp9TeL}kya{djYn#jxjxb`!z6}< zHY6z^h2HAZLL+$w ze8vcbLq?_;@+x|?Fl&KvC^vk+y8$|{VVEP@be{rQ^W-DiI8n#6YaWTjlx4~cLdw4b zB#~^_?$GMXG_5oqBqJ4sphi=sFC96x%P_h4So3Lvq2u?e<@^*Z5=tU0bI9Tg@gE`#yd{5 zVf-}@tO!nlwaWlm>HnJuu3ZQk+*mmEtpB{KGq!@D1NTiXygYHpex9aV#LL4nfLkv` zwq;_GUyYznciS^o@bOGcBq#WxsBx|G{FL*+cXv^7QBlU+ov}`L+nEZEee-A+6b2ZC zzNek{liTC@(jxaD+u)1KrUp3zEt&Kh5ZpsDMo9Yu-=SNLu6*eW?Hq ze}yV()Z}o6fd%IQpx?;cSQyOvovSHYU@MlyXJK(+^THvm^v_1-$h?oirA-w)A#J2Y z_0%_8cq??{e9TTgf=+zQH`g7W(7n@%955VGF|K1Q=R zVb-yEcc2LpT~Fk#YhsQIGqlA7j19P|pl*TSzyWVTD|MXc-n<*~8}bgh9wpDAe+kS~ z3$R)?)0vZY&tE4z3awE3qL56wdGZ4#+NTn&IUzlq?Kf|T3|NQkI>8%IyzI@GaUp;7 z7-IpHcKe~fkY{eZ25{jt0Z0t@+ikC<*i{r+PDQm5)IZ4w0#y`#9jfh&_p9ge!1K|= zV=xEaO8qu`;q%LGU3it@D7#J6Tu|Mb96Mo0_-CJLfsE`ixebj;mGo8Vj__=M^Bt8? zj9o4&oiZwm&GUgU&`daRzBz~DwgO*@RhQZFIlHV2%fa2;5LW1!2GpnP{2Ig!9y=9j zvQpnHzXS3tVWK+66FbxLnDwvS_AD313R4_DWwjuYy8GQHKWd)8_sih^MwSEH z_XYNd_B^n5w9{E%uZG2ExAcJ`drBTqQl$wWDV8W>yvpR&V!XFbcT?U@7yRS~{5$$v zN4|Gqe!1|Hv`Bh^ZSu*3ofH1MiNH^%feN;V0{7E;liE}aG3R5iuO}1q$qcz_EMftwXy`lx9Ly@eNCF|SE4{_vR#3l zWMPk63|seF-77tAM&C29ZGSZ-V1;0#=aspaq^C)_C|`3>Q>UwfHcA6?Ao8a7(s4WH z<^VUu7R|D8P+c9+Fha(%fp%dyhH9Ycf-?h$9?=?aiIq@s{I9*N;G0H?p`{cZ-m+G}NGwyfBvo6+Y8^CNPEnGkX9W|Ba_ z`k}NL`j_s=Q=vcUk|x8wgzi#wn69sSY01Q`-l-mXbLhE2jP7;cNu#>Kt45e%>f|?f z@%Or6&K!U58t)qAVFqc%k<2+=yVYlDV1D?g5$X=68}gzJKrenq-3@}-@yuc9^v{;; z_5L(B*7w(^(9m2(@Abw%SC99t9N$G6#QCa|?)q#=CXLN6OQyt$fZxPn@Ea79!#~Lj zeq+wQx8_U#T6MqX;85y#J?aLaT-^Uv;jdi)8j&R$-orl9tEXP{g{*r{nhZDV@8VwF z>cHIgA1ygQ8;BLfcnRB-_Jc9XkJv9yn1!rHdtoK(Bt5DI=tJtBlOJ(WZ1hW*1-zTu zyE8VuT%~caA8r>A>@o<$n{ckDcv%uEXVaJ}+sa>;^};zyl6$6QPkGE3y$*)e>25Il z1ed}Ot9zyQ6-}cmS8JXRJ5j+kc}_de4;9;%Z~M-RPSLtedqcjK-7kM8LK3voq(WAw zJ5F!Y;$CT&{L0+ImwMc`%d<%;`^>*Z*GBHnXd`)atMvA)UGg&|Q&1+V5LVJvP-lbZ zkm7fl=Xx&(8xhld+liT>ZHW_*!w)mN|NhR8Z6&i^){=0DU?ZD+k#wCO%GP3qRpPYX zP^{HMClxh}mP7XXvl;{^yz&7<^1{Y|r=53#7dFP;xc7RT%CcQbne@9q5Q7VEm-4N4 z+36I!jUp*jR42q`Kx9HuqD=~jAyC69+b*mY?^mrRXXRz0bCOJtI$frqbH+^?^p&KK zfMc~=mtARo9eP99p} zq>*B8Qsf$Jt3mDBz{J0_4-3Kf1!qC0K(quI^WsEZAP9*4Q~Rf_59xLL(6fug1sRxI z%)Wp)sIR*J(q3V*qyuPNo2I2t(-(xd6V!XV@16T{Cy+qvcgR<`CuoztVfevmHpVN# z?;K>Fun6l|(Cz8HZif`*lZ!&u_-450sZO(XzQ#HGg7x@em2O`!GGnb|l0m#_OIW-t zd14_{ieeQ`f@}@YMj$gEbnp2c(mpSNHEfjiN8Io}4+%?NC>!(j=;R7pu^bMSnHpu9 z*ApSO{p5$Y0Oeo@W9k8lC3nIzCw&xNqfFNp(Zzu)=G=lAN&)3toC1BGpq}kep+?&C zUg(hjRnNaTlKFW4=)=E**L=8c?r%&>%eyc7<_Oc>P(hAV9$lU_(7Jg_)uHOufDTHw z#2L=)P@u6Ju0svtglqiZ`gY1oyJ*WI)%k6|(&hr zHIWB3`ecH)iwl#J&DEe&XD28Ieb$*L`GF!nw*H3^7EpZW(5_UH&#&FUb&VjYv4YTX ziUm4>LMkc_WN;8bXj$$+rOi@bBnUV^8Oea!efEGubAXhH4NR#ZZ_-jA0>b-hlt?6y z4wUuSbg*06!5o`uRHF`UCT-N_2+ex8pFN!tu+;aOaP0(C#l=xF=_8XIQ7vpI=JgHW$3T7MCg&vpa)IbAsN>xM~e)Pw8 z|F=$eN0}wSPC3krAMmfF-@T-PP!7rUZuD8WhkoW;p>PhKh7Y4p4b6~n81BW6TWezf z#kx$OVlzpAkzTCY?3i=Zy*P0H=oDhvOp;xguz6-Q4tGk%eaGKoK>o0&d>z@&&w#kF zzfx*tK=LUT8d7qpC}cfM@PmG9<{IqPk)3xf`ydplsr4Y1BtXZcUQwsMCWWoh0nL@5 zUbl*g6%#S;0_9YAM<$)@7ad#$N?aIs>GrBsXF#hil3hF8ZMF+FpBjy!J~$RjJ6Hdv z$X4TmPg{ZOa@`OPMGSNMWP|uf_|m{$|NQXMDSC|L<%V|=B&jU)sF{{OrOUHmwt;B^ z)~q7BK#d!zCbb? zGyeXEHMUzSJ|@eBJ!Ou~)~fk8#j0EL5|q*}QM$ig$^r{~58$GvuY)>HW{(h_dezWdI5tpRnqM-hqI z%6S>WYpPn^7VRcMm-ivyg1p7x`Az?Z=P&5$bhnt>vu=?@=>Px6JMZOC|I85XfkF!d zgB14ZG_o`=LvhN9OMT;BI}hvCh`}`i{0zAe4HGt>K4yGn&)gSXO!69-jzv@>NO2V_ zmiTQ1x{wi1*Qf7?&j&l^n9pzyyO-C}L$!Xt@a2E4DqS-uwdnH`ip-eD_nRJ1Tg5AXWUG&28yhQB$239*(pfy>mnCJ zniXgjfL}@4Zw z*KX~?aTt)6>j$J*id{{S6;xE2rUn=^o9QfQR4bc2Nt>dQDfcb)YcP=wriK*ne!ZsM^t`qhS1OE-jckv$VirxmJeLU4gZH6r~f)m6d1;8O}r5Sj|+@Lwo9 zZ0=Wfp7+WZ!+q6VX6R(C}tNmh)5 zTzMHwq4h%9FQda`9&|cZl>dcQt0}SBlM$)J)}|qbhtTmw50f@CgyKW#pWRRvKy|0 zJnPI;yx=dZ8}a?@hf3+;r+`_E0-mSVmDDFfr>Kk zP|T{QnP63*qy*g*pg)P|7MiO`@Yor9@c|&RoZyG?QS-CMd`CF%U0ZOxP?gji5ZEK^ zf-=}A^p=TTo|mL8Dq!bf8|Wm(Ny#Oz54?B4%Ck(DuWMJ0;#$KG0$f0H{CRFOx4$4F zWQN6Z{IO71N^)Fy`*O<4796A40*V}@qRc!?Q0b#RHVLYTG=<64z zzg~3w?(;RzTz3aXI5f}RmtsFfmApc9EA&V}ndUN7P=Bmy1#$UYkK4dWX-*a-_^oB} z`XkRzd=EqSJKo`ZybRN(*fGx|yM-Gi2Rg6++SY!!EK>@Ht%wUM2f4e#$;k{XAmqp^ zt>BXY$$>CckRO4?sH5qF;ym#{!N{hS=hNIM7-VG``}3w7+BYmKOYStmLsH_xj>&ba zRiv6?&rswP6@{dXDfIj5T1f62so43OUXBsLMg_Z#4GiweYu;M%H9ZooCjigRo-aW; zb1QvE2}P*tXQ6q52g&$-hsNIYi@XI$n5 zi7|SnJl(d7$7LbG(U-Gw_M>U%WH$nDPyg&wtu8qtIqbGBX+g)llK(tCFP%C)ZyR;$ z%`Nlmbc+QC{G%n;<$vrE|8FXb&-J+BaVnb}(FTRgl^}TE4s5g0lh%_s?-H`q7law< zMRPupEd`Y{+#un7=I5Hl4?Q2Llg_W7{CCG>IM;>pK-v!R1ykJmZjBcPXG@3oUI zkOB88ENc!le2&H0kkH|EQSo})cmLgjoYV6{zDdsWLyilFc0sDKA95Nf_69|+Qc>tJ z0i`hr$rRBCrtF6O70SfwQ7*5G6ndP8s!62}Uv3eX2X|juhkm|73w-D$xpt}sX4?*QI zlWudzJ~Vy2SCLViH zAZ}J1^2i4ccLy2{?zhToD`uIdp+j;DoyGjiRxI z9$A7^QA;>zzY zlpsLR4EDBz8MW`9^{cSwRa!qWI!C;|jjsD;YTtc)mn{*G%kHH(sCAIH3%vpqGSkOM zFqj7cn;qIDuYA>g(R~p*6QzPZq)nXd(d&k5euEer)^jK1Yr*A^8^TcF9^-#bGZprr z__H;C;3sEcniq=41pR)T##WvM@f-SzAC~^~!dJ88$sXXgW(Mu`XaSxCYz;y&`f6ZE z!00iu#r2JLcD^HVHBu;9fy;ro;)as2C~TkYs^hXYhGRLz*oT?jFZ###n6xmX){%2+ z|1{V!`}Pm!z&p8Te~>)EmEy5I1nh$8>%sB!|N~lt~OP$Wy0F&I|Wy#?0xB&_ZfAa zt|G7tX!K(RzO@~*nPhVI`pHYvR72!4CF!RrwCr4f9WjLtb+}clU`+8hToHrGP(Rpr9wi5n;4eeID? z<$3Pao&Y)0F50i^k{5cki&hBm)6K}c)BB!_mEDJDt3WEcT-XNPDGz}pE?ROOy=0_t|1JoKHi8zyDQkdlV}B-*L81%fQ{wi|3k{ zn!=F10?5=4P2aD+51Sr67A2(68DaXB%JMMms&?Y$xc}?N&wQa_`@FM`xuJ3MuS_&+ zu{)kmI(x|A7{u_l&suQa9?m40cuLkkq;3~)kqk{8EQZPGK zP&%bfm*rc}UILxo4q@8VY#FMu>;e0Ot%Mn}flC zTa*5^!txkDHkp&GY4!f5L|>9JNqNm6|!5;9+ zoN$8QEVyh7;qk}sv9`sLeBx*>9IEFirQGh7CusD6!ho%$5x7ti{4nb~x|YQZNBfZN z1n53z$RpX$zVNRFiG4oJ;D(}TVa)hO={;?sG-mw5QL%0|Q(+e*Y@Qj7Lyx~ttp1i| zcalnVe}$aoSB>Vv@sba%_A=Kgww@w&R8*EP3al^neJKCzQ?2gUw`0CoyZ}jmu~xQJ zl&LXaErnLW47zP%AMG!a{ug6;ChaIRgHs4ei^F?Wrq6^_=E#lZ(7v;8-*2_#z5 zDX!HeF5LL~y*D4dY54j_^Kk5)UInxF`(OJ;^}Ory7S&>cRK$%w37#i_GCxyO8=fY^ zzbjgD;rll}y~9ql+6Az_Wz-=*?Stpe3t!&)qQ~rGNF#G>_CcsrMCNhmRFAm97?^C~ z5$97FYk%UD36;WvzVVW@y1&u)`f#3 zIaUV6NU_Nj*$l0vGkg8Bg-y)6Q1J|^$QQ|N-8$+bIi-C#q_6G?=#1 z$AuqJ^56zQa*j%stm&uX87>h@|cQ`oF?#xte9nYQTQX41$qwu643oF;H3^yAQ8w|d1n z*xvV5v**)=A+Y05hoEc=jR!>`r^!nHwQQ>?2B1gc$b-Hm+6wO^5LekGSR<@ao|v)( zmbPM1hOkk(78bgl;bo#CV2djXIVa2s$H$?d4j(b9d);E(nHo?ax z=0QLWl-D2CY0l&VPz8h8pbx_CD(2t)WE12C3hVhvCwC9F^*3e_(B0681 zPGg<_uX%yhA&!lGgLFRJ@=M!jt;<>_jDW_}e*n^SCPyP{fKH$F(MvkU2vX~M3 zrrGf}=XoRYvq|hp#Zg<{*%vAg$iXgK6!J)X-rKASr+)wq%^6|mrPws{fvDZ*ZF7+= zE^MO*9Q*L{=~wxA)9kS8+;}N@`ujDu@!EmdaN!L#M{v1Y_#mM6XLa9-rPAjcm|E{W z!gyJgszKcUdNA*b0-N9rDDrsyE1I^OVI`yiT3R$RD zL1m{l@!AQX*itKK@Tw6u`j{nk@xg4U-+zL=X5C(R-zHhMY=s91=K|dZ^st*Bh2fz~ zh~D6RH3(E}N3`TQ0Yjf+ihHGc{uO0b&YSKQ3{}2b@*Y{j4?`}Th2Ca`p#+MJr${Un zbzoBQ_&q*G^~R8%*Np0`;!L{WRjin-1=gq$P4V+z_P{&rg~h-poqOhYRVjyTVd1jD z6OM^SatQps=R|lks>M;rPHTH<)XvEPXs63KUQ6%Z%K;;-KHA*gjsFm@YJn7w}lsIloG6Qca&f8 z+^?|RpmE4j@Z3M=(;K)!GohvMpA#+1N%P#$8zk9<^MycW+>h5}7sUeabvhN*Np8qv zfb}hJ-p8;d&Qa{0oj7N);65~JEFG7pI!Bg`OB0>}61c{BH{?&{D}R*g^T_wY>`Oqh zl?0^+3!lDz_TP(M@A&3B3zH;^=QzpE@JGg{5V3stbTS?Kfg2fvm=o%)@BI6_whpYz zt}+}hDKbQ2zUMYmCaMQM>q~-7FIQ=v`KJ5TE1(R;>!GMblo|9C2q(?uZ}o~RbL$mJ ziY`SC^=R@LsPbIlesUI)s5b$1UC?l9;Pw0lEz%{;ox?kZEU?yovu@mD##L) zh|0s(297E)`1FSkK7~#?%MTf!?B9E7vSlIBu3cG7wtjA+W=E}-om`5Ata2t56(6xk zQsZ4DE++cFe*oN{YlOR{7_je#fayij?vo#}#`~12gK4H~)#s%Up9f#~rCzrhWvSq- z=DY^!n7Zj5b4#^L1AG0k63HpB<70OQab8WQoB^=RBX=g))*^jQ5B;lg_wF0QuuH{{i- z#R0$Fwcz>>OXini7)eYh^!DoUqz#2JPf7M>taM_W5=SPB)0nbL# z$5U*}ls+f0;V57-gzcBt&NBDrL{snVnUBpmNONANy9Y729u+qHT!ey^$Ma%G&Clfx zj{y_uoD&OPpK{mMX&o4%cj3wb4$+@f*zx8^7#L7<5u*>0P9yJltvU`UEi&m7?k%HC zNx)sM5m;OV`TAX&97Vn2 z$SWYaV+g)I3KC$}(_l9$tRHjw5jQIGs|)68EL)kNlz+yP3>S`xm0Ka;AjRfWWG@wk z@$zQXcK_P&V<=t_c1hD6oaWgeZuF@Qzr}Qs9@Rz>iiM$Q*puMq$(eNGv^z@U1nd=T zpo^#8QI;s0CxabPG>5m+O;hg9SnR$lvT*7F&^*YbKk#z!Ud3*Z^m8Ew|L9oUaD!x! zz!|k5Df_J#of)=OS*qz^^eHl|L#&0lE}k{|`xBT%yAa{n$+l}WUWgc*xb>^=F&0y@ zDL(Kr+33Qi1VnTCC4Vz0b~{B%!hUAE9)F2PHLyHYpXM$G&^|XVnh1Epa1(d;Ug+p?4R)0*t171KGv++Ut zwYe6c)P8?SIw^2rpa6?wKcJkX*b@{vPDS1M!u5aq%fYXlp7-bD^RLdk#oV3#7)Syz z*9$FM!Xi2zV(yjmQhiQ9O~7?|`;015slm+LN3R|X`paHTtgKsD0bQpDe z3oB_H?=NTm!!+cd*HL>l9|5_e)yX*1$vWL$;Z;G>gmbU6v=*}6#0NseN#D%c| z=lg-g&X2sXM0saj+cT2wunX^mIJU#MPJy zoFL3sHK|vO(za+AV6qbqBfIU~&s#()i>XPO^t(S01HVYU3&R6^-+qQEonoO5G6kGr zQ~VuBFd!BW44@Z7>Om6WggdmfH!+E95gjX998je`E;NE{6K+nO?Vwz#Mu*NQJyB^j2sC zZBG0+7(^q1&8PS=ec%s`p`LiQ`t^Ob(YDV?FqR_@_s=4_5w_HIx>P|t=@90KI+!}$ zv6<;~t|)`PxF9ib-RKdYcT!Qbl3D-O7>E0 zE=97bs6vmvtqw+6&FNl9?BX^VQ@BVAQL9d;iwTw?fy;rcDd>q{R2GGsjB4!n(w6|& zd4k_s_NX>i@EpCyVa@^{5LlZ7tD|B5cJ9Y`xw2!u0waXB0a}+Whvf*yUi3aW3t1}O zUJ@2-()4`$Bx%RSzR5>aqa9kY9~Im)=!`48P%-Al&wf~GOVGxl$cYN~`4RU;C%mBm z18Eh_C-5uzAav==`t6!(_w7)9(d@KN+kvmk{zY@S?YZS_B!2jmU=G_D~4sPecHp|?OuJqXT*LgZ+u&= zINS##1wAi(Q?-w-J$=#p{k97Y&_PiI{{Sdt2UKk|=6i02)dkj!9eRy&MgmGI{cV>1oNoDvQ54b@deQO6N|*Jk2ng%++Y)c2 z*bNkkqoRtT;0&8XbD%ji*)Ng0JMS-#e{g@^tuN&HY*TayTSi%0^m73W>f4(S{=~Me z++|G$M+^DJsk`a6;C6cBH);VQwYqzBUF1J+$S>&DOS)9;rZ&AS?|p(h!gI0`axXMq zHaZrB2SECkm5~|+yPoG|IjBb~#^>6)r+mV!E*zbQq)k6N^C?S7y#hH&O1-hD4QuK3 z7-Zi8oyIj#b}%~7`D`BT2GM7ajLN51UmidEWy^x%Cet1w=`OsW9JiY20~EWDB71<2 zhFli3`Q${jfssh~dL7t~)1*1vsQyI$$@Fe|qYyVLDw8~Ap47S^u&k0sIHb1rM#K8Si zuFwbMASnrjN{v1cbTieIG%>*0=;hoBw}ErW0~rG!cH)`Bpbs-_)tUy@)4Qhn^k#CVuEKoN<+Vv01b7%YV03YvPbO`dg1% ztFoKH?AVDZv7&5k_pACe?cEt@f-|B1A{{g*-ZkHoLAOAQ)4f*;y)nDCnI!m?$u7x_ z>OHcxGW|8_kw{GN{NtVX*2;R_%rfNN!qwV@sTM20M~KbeDfD$n+}@q>b|h*v&}yPq{d8?@RhmREg7Cv^C0XSqmK>Q3Um| zLt(+bIpg#SMtFgHwz=nQklJ?FE&7`H&!m@LbbhVH!MWh&*6BKg`==}&cad2mJnp;A z?++*}v20wO&V0(KJ{a7|EE`v=i)B+}Rf;P5L(K(Umf!^00qlLZXH^EI$WooBT0Bx1 za|3Z?S9rmgTJzQ~XV|u#zfda_M{_w=+k@bFGi?CI>1!T(^lkCX!%^0$g}JmFC?lJ9 z-cNG_<$LlYy|z5|9Mtg1bE?!alHx$9CMXFsLW}lCZ$6xdC8g%t!Gpe^D2hUmw!4L{ zQzy^KmSszvQ{OY-Ki}@xZndBc7{=izxL^AVwam=%d7x8T0xPG!^405P{a{)LuDeV^ zQ(*t7+BS*>HmxmG)H1)_(t5?ANvi}8JxrAJ3zzzKN_R|Ks7C_ndd0^xmip?kat+0g znwZK6D4H?MFo@bFIgPTU4}!x(v>D!gV>sg{H=>n4{otAH{^xVDGv?SC?ew}Tu&h1z zAu@}bVGzJ_##?{Rn0A5XIb%5Ea=`cUk4IYU#->bFDmm!F0ohtB8*zeSAwpY1MSZA= zpAaL-q;VSz3tyJDN?ocj_m_hdJ1TOehIM-tvdck8$0VEMkrUA^G!?pbsLuJG(i%-q z6cfBf->g7Zq$I%#$+18<#K+J1u3=Fq16V73cY3YH)%8P74^&X61hj+98-5-^h{Y~w zSgnfu4u%UGZaXsA%6K&8jr}irMsQ&Xy^6$u(r~)pzFC(@o)C8|3!%+oeZYsBvcPoh z!oG_O4a`QbYWi7d$ta|chY673&b!^u@dHSD%76dTw!EIBiv7YDZpe4fYlJq1qw|ou zEJsmJ7Q2^`au9<(3`9cJ%1ZiZWVtXYAchv$5K(6aAl<4 zywxkyw1Rt#(&|<63X#KI9!DS_zXm)&ZP%@fwk=x_V)18I8rt`gTo?9dPFwjig%k^t zU5BWs263k#!Oy5Y3RP_Sc%jiRNu12IPfML#NhfOd1?wBch=debkE)C4%~znXNplYP zv~J>c=ol3sgQ(Fj(RY{ps3u;RISKYo4Pp#m7ra;FI%l9F-cy45sQQ|uXvoWf~Z9%Wvq&OvF}q3O+vUGy3e zHri!L8pOQ|Z$QAmEVxwE=F#R+B{(2Y(X91KAQwVP1r@RmVHSKdLFyOQxSR44@$zxE zn1>Ma?`D?0gypS0Zk2(^KVRq(E88QyJ0lP5!ZB^Te7n~Q2(VtzL7!)6o&bO6E#}JH zP1EB;L1byT;ss8cIjf1Z>&t6%6R(+b-Bs-En?L-k?JD*{#dJB=subDHu)CtgUNy=( zP;rZqfbiB7*col}E12CtV}=Z)Hn23cN(&?fl4kiDVU@TDT0G2hV0#5^rma~mRQ}%# zgWG`CX}rbu4uRtsMVe=lpZVWtvg&mC(zf6# zah>j*tT3cTohvLMqeuRTv&9PsBcEQoN8rZ6ARBFJ{A=lddC_}Lw})(E&(l4seS*6r zRbEBk3b+p2xpvXr;jrS^{hkNE(N0th{;(6z@l{{UH{rf zVQfgZe4P#*scpJ$I+ZS9Hxb!Rv5v1D)ue z3|qE(1?0}Zaq=s7=IP@jfJ#U6NRjN_s>~7ScS*LfKqLqEWXqZrdnDQZm;5XzQ(&xY z*>*4Fnrn_c3EIbepvMn8eU@mLKY1XE+XT%nI2u0^Y9<#>Q*%%=6^WuHtx%+GU|Q(L zw|3C$7>vXm4$cfZB-}4u!*2T4(qC4-b>v$hSe&k1!>-jeh8n+EH&6fgyI1G6gfIF+ zwyfBvo9T5+r@?SRJ8Q4oX>xvYXT%O@vFC>tJ0{R^fU@J4@wnyoE3MxVT8xoffLjb% z%8&WYg$s9qs?Q)C^4kifFkD-0f)AUxRX*6W6=7XH82tp|?y7zrM)ne$e= zaltu)b1wWqv}XH@&Qc4q#$Sw{qi+cmf!RJ?hW zt;o~BNSh0>;Ejtp> z#5|#Q(nW$yx|2Zt13E!9-m#)sL5bo+_ubMY#omdA@i5R(Md8Dl&5>iM4_w3f?cC3B zW9atH_uF6eO*7V~ba_@NvZ2hsa58X6r3!#h$~#*E-C$60yPYb8;A|ya#OfF)Sr6NH z&mRf4BRJ=n8wrD0ld-GbT9@_}i#zge){(oUik~~;!gl7il{->Lu@@+Eo{Cy3S}NM3 ziW311Er#@;Li|vrV?(#hH%I%>0q75NfVRa+`?m|t63_3yE zs>zQin+zv*Xv;J;%A?w~tiD8XOWLS3Fi39(+wvM^Ic&2G%pTSLX=Nbe0JlC?J%;)! zBq&S=Mzp;_t+cs~1y{ospuokv0gP!*Hr!@RrtrBYj6V?jOwdSw4h`VLDd>DL2T(7tzukqC*+r}Q%*;n4KQKp7rF=o29 znvQd?m+B9I2#A5HSER^Npm-47WPOb|e@cqXM{b* zaN}h#H|MhU?_+J(A`T;-AnciTQxQGIz^o@9GAHP~$XZ>w?2@2fj)`_Ei9x5&{L*8(Jy zapWYqFrm#IYU|W`JitKxjxRl)4~MwKASX7|Rvz^25T;E;s8pECEVLZnR8e=AvhW@~ol9W&K*@gGf90aDB!e|LN9k@F- z1z~V!+~tXfMsT@k!DR3^-3d-ec*=)%}PXki`i`bS4g`!3&?uM|EyMRuX zY?+XuO%$OdS*;Rjb208~K36lboizG5eg$(7YK8*9u&!xNzw;N{1ft72DjX@n{Hc3q zWlJ`(rfQW&6@gR#ukl-fcKWtO*=}+8tU#{onsbHQ z^1pmT9u1_oU_X2|`koBT3EL;wHL3KaXQEwnrbnIbO0Z$-$*KDU>tE@Vn$J1jN(?sX zJn+=F+6*>~!_VbU>w5W{mi1=EAC7%Y&Wt1Xtqk=wioHUSOH@=FQyh3$Ua!D%cOW#u zO#=vz9v3!)iggpy8s6(>6dZu&t7{=`%x-C7p#BQm!Qj$QQ8<-** zqg}T{_d+ZjNu-Qw95Y?ATiWYZ&$LH2EA(9oqgR`zQ`+NpD{v*#<5s8bnYrET5ELPO z?2+Wvrbr-IF;e4hRIeA6$spT@Bd${}QP!%nB&pK%0h_%J%(cgD7{xq8rTh# z3b<*4rRtETosJI9mIDsq29$Q_b;F`4?CkjLMm_o}#e#<*?U5}(x;m`UGza?{>BqoU zh&j!Qh}b}^-n%jvvYIH>TO4R$ZclH3m|&~4ixkZ6bwdiYRgnf^>SPlS8&+DGV|?Z$ zXf{+%p`;FubY*TUjahUYt${9u5Ob`c1-5s+Zun z5-c{;n0>}_EVvVrkc&h&{2Ij2O&!rfH-j=w0e~EAT)K9v&n;%TR4wko&7M%59Me?`aSH#kAMHl`-g1lFF6<`%{~Zjxbzpp(bLrL((DdLDdSOkho8L$ zd*9|~PmIndUN*yZzghm0zx~R#d2L`-uM5X$I9k+>NJ~Yz-qqqQ8sIIQw+%=;Qr&k8 z(&=aNoQOt2iBkm2h5(+uC>||a(D$cuqj+l5j~i^uyuOew9 zxS~(!tJC+*Y5`(8pqmq{WmCgaps+W=uK}3mniwD>%M3!*#3%Gwg8#-n`LocEf&XJm zSiHmQFrL~>rM~geylUZPX$h1w#LDs`ACovTiZHPU z9G_VVJ7?AbxA6i8btix9fq$*EQ3>56LguhIk~-BOTG+(g@yv?^Db_~FTnEC;~lxkCv4RPxN%Vm^LyZp;OeG>#m$`j%!>Eb!y( zprUf+4}5ZGEwuco)fGt}liRxArcuilWPK;`jpq61zj@#nYe6uq;ETV#psS>x&CGZ$ z?zId}(rXopDs|Fp*`NZC^tCBaC*Y)S@^IerXWwG{^xPLVtQ+`!Yh5a7^jNNvB1@MqeHpTx*vooJ5F6M|il^$a zGzD@%syKH&rV?UEw-D0^y>3_o6(_nR&>z+2hBwexggq)}g!e{{WD~B3h1Q`S;)Ud~ zS;jy8;Aa-VeD#KV2f6CPfPoxgzrg#46nmE69W!X~zm41mMpl{zo>^cZ;*0z+#5f|im!Hw#@mweNfSd3Sa(Pb?!NwWgy28xZN$XY5Y z&m+&nRKj7Z(u|dL(>vytf?{jtq$SFcCvLcAW$^R%V`%W_opmE|2KqG$j?hpsZ2SVY)P9tBH%2!I;qz!!Oz_5oI+>NHG&H? zmaJqui;2T#Muq_pn-QFI$PXZz-~Z&-FS_en{-s+c5xh)#jmLGL)q)sF(<{pZJ4vlB zM$$zZJkH9Hmbplj8dj$(^jOJ!$RsH)DC=}*NhLJJrAyWTuVIpVovzz^x#XeOqX;M2 zmSF(K+O)XDChpfNzssM;_GjGoVC(Y^?S1dnUZ$1ipf3`*WXm$?)G%0}u=Wj`q<3j@ z6u|lNiXPS5ObXBLPC`tdffzn=`bKchPi|AQtak1yTa>tL0*Ip_K6UCA_g$Ja;C$2L z7CDvOMt=a^(A)g>%z@%NEC>d^4D?8l$p;8sug_QOG^SO=_=Z(-K--33oJm+F&#K6FR)PJ*#E4y z%@J~>;<6=J`){T%IHN96oQH~N=YAhKUXnd;p5sxcp62!DUAO*Z{hK%BFZwF;zBEU$ zZdQj7n@sLRwrCH?Gbgpu`yuevuDL9@DSt4h*a!J+^B}dnf7<356@eW>qdIHGmQf&N zw+Fs9E-D^66}mV$FZfc_8G*lT>6*)~G8{E^`E+53i3Gu<(QH|7Ei^TkWpO%*_Z^s`!n|nD(M5? zTo-lh3rSz9{$-u6A~0IAi!PE@Mcz@C04-pR_-2^-uGF_Gzf`L<`c*5;$B({#=&Lus zyzXaCaJ(D?gJI3|kPq=PLDWt!?{yIt(0u36u2hoG&jh(}U2TmO#Ew%eV6Kpgsu1SU z%VzGR@3497vYF@U_(vekgL`h3tS8iZUq{%C*y|sA~7^P}l$(oN3y-GpeQgW}CWJ zQEl-fX=lXW-np~j<9T{q#!k5FDY67)%$pcA+j`UVdv zO^mf`yhr}^=KXfvioy`kN2(cQADxOEOgk+m5iC69yctxj1D#cjm#&CZqHjG`lFgW;oa#? zf^v}kE}>VEX8HhV3h$qy$8dBXGj6Y2=PS*Vu{W}jd7z9FwJM>w0a-Ab)QGoROdVJy z{Z*1M`ystWy>9kRdY#Y#7#jwFZ0*n>$0LVO@edO(+xB`4>?3qxPnDzp11I?^sRefY z5^aTd@|>KAZXrxG@~$)i3v#q%4~SDE)rkY<*3Mbr2?0At9rk-uk-5x8IfAbDHYB{9g6 zZ}`!_R-vpJ@-5-8*UhAB1>1$DI!auXMnC0gH@G=K=&aMfTA50ozr48cCWm8r=?y=q zFiw9cUpzlUxJQ;HI0S)vb2-9o;RSlTFef55WToV~eDqcqo-V4hCVb$y-1cqX@pl&3 zMnyg+we=h^lsaHEE(&Q77fW|b|M>ynCRJ3ce!$&BG4O3V>WqB`!%@G@{&7@q9-Csj z;T`CYy0F3E*w*gx>V#^xF0zfiFFN6k+mBjhw`UQ3TUSNbgy7X4w^K}^sLGrH=%%Hf+m{a|Z7q1*_T80OZ!A5fIIeq~7&($~c zY%3aFwlImK)Di(=RO5c5Un%rxtsMO)&n`&xe_)h8^Mgd~d-1D@CD^ggE%gjp&M%be z!ahm5Rm^4+#U@ZB9@rN=miyikt@CTApU!EWcviMD@|g&-`b@Jb!|wnokhT5)y!s}j z(!Yz^@S-k@ z>CTyHo=FqVlHK9Er8{Y(-)`w`_lKf~bjh@nP{vlFs8%}3&e&y2`dp4de>%su@G>RT z#k+r;YwLat+{CzWP(8=WU>GSjnIfCbxfg7~+8w@9Y3jSFSF}OZ5O(W;0~?X${|;tm zOjou{SUWuhT0@=0y=FDb!vl@g5KcMZ22Jz#J9qxkveu+Nt{g`mxNtql8Y|}e9*TWJ zkuEA~&+Jdwl>s~E_PA9ETcyipHUJk-3Xt}t_@ztqCE}yn3}F@BAzUY|2;3w~732ua zMQ4l2etH{yUjsU)bcHN!Dwf~(x^0}e@g<}Bp6{`sLXY>=d#C94%*hTgsxOhvY@x?8 zbBcs!V}C zT>{FBq=U(pA)kGNxK&z0pHVNCl&C;$BE>IOu*|DVu@wj>;TIT~J4)bAmG(lN%N-@s z-Q$i63EoQ-&66{M9s8p9uZIV=K~0rIPjZ88unp4Dzkm0lEu}f1(3=Z8I~-)^DRhfC zT5>^Y)OL|NU9@D?j0*aypiq=a?^aa=9?`ZyeQk=cL3~tI?i(9&Oq?$}^IELnjA3;;3kujdya6^ z%MO@yo9S^725d$h`5`~E^N;^sf85=In95g6-Xlx+?Xg_gW!h$iuLO#Xr${WiOeXxG z*OUpZkn2g3qC~ZX&5ywK%<(W8dP406(a^&<{A%-?bF`nwQpj{VSAJBmM$`=@4&8H5 z&V8dYPqqqHiHg81VRvw1(1|HaLV^8*ItwkO`d*&p?-KA)ND z{t7wE&vv-5%krU>?YK^{^%SWCT4O=lzuk~G1Cjsx>H^tiL4z3MG)DFP;A(M^v@CGN zoQ=MZ6t%i%KS`k8DgQ$8g6Lme(4~gm)-4~G?APVl91ttW*HnXY(J7xcpJwrXRToHS z?xeE>t4J&a0jdC%8^&k&;nGz`pVFq#?a=apY;;@P_e<|-v;Crj)9KxGisrKs%6+d2 z;vq!ROhLV&CZV~3R&j;l(7%D!-Jj`4*4DofZ1RtmZa@${?}TZQs@q5 zryy3)<6RdNEm;p6E|iYIK1Jszng+~(y|5VYpp#Fu=I-?!Z?QXf{_EO5$VwOXa)DO2 zUr}`;#X?eJ0~K`(Qbk5J$|oC`bb8~&OQbmfD`Ou`gY4ria^DMC^bOZN3cR|%F7V}XM^q1vi z7eANQb*q8NNh>ROlwyIy=>QcKPgS5M@Ay$#&3xPVGdGwM~5fQU6~ttj}qGyAm+Qf*$SmLw_O9+y+^rTsUL`Nu_?y z%36wDMUmymmT?XE58iGG-wGR#xR)F-gTr~fcFg*4h8^)U+zX<|x3c!wGNW@4kmEPY zsK$WkhdXSFHHp@C|pvK}0llwkc~&pK6?=ZBskhoNc!5%6=er8525iMQbkjmyCS zg54h&_|g}8tO#rkE%azpw1%bxL`$;C8FizNxjQF?{=g@bPLaWi5uW02>KVZcMIkM~ zGtlK3=ea4mz0auwxl@2-%CFQ9zCLuSomdlWAd= z?TzZM`mJOSzs<1==X=grZG4I;7KEh=sHolZ8ks%wK33J~){_+iBRQkIOLht_Nn^&Z z^E(7g#!pE1^b%krJq>jV5GWKG{kD4F*7ZHrJv{@;S9Z`J$}%8r(@6>=56I(Yr_jye z8^&X?RG9`xHIHRfcTCvmxl`a6wApKgc=*!RP#EUXd%nNd+p>?oP-U1yFp`?93MAXu z9oi-asZlW8;h5><^S14rdOkxsZlnyh8&BRn=F2bo)&}X@3^NWX zAzXSfq*;Nw*+q1R>W;FFZllW}Z#l41}}Y;yh4N*H!Gg`L;oT?Q4`V&Oio6xlEAf*$#abAM*6Xr69T^c0LE#9 z91b_{->zOLvOp#%@mp17nG4qz00n5jaQ#M#-9V8zD(dMsZpfGE7D;ZsmMzN&YZt8$ zT$i``L`#aKTfK{<&ZvbQavJQP>LG`5_(fi}g!*|<;g#Q8fUzLAVj{Wc!oXNz)gaPM zv5zV82w1I^NWEoL8<^biIMFsix=)W=&sV#d^`i5z9YN8jSkMp6mfZ{owPD|>dfVh9D$}^tkc~vn?5$Iq)cF6r zdELUNuV&J5BE9*kle3mfwqdFflaxt{Zt$R>ghrd?-{4gv)So1eJX>KO-|Y#3K@bG6 zAAFoP#wizgfp5&W_qNGxtrDN+5ZBFFfadt1Hf&*{rd^dkCx6bu^YjBKgKMU-Tc}>q z9@#=SD}W?$bhlP~VB~N^HN=2NAJj7wEmo;{Zs-k?%+JGd;gC#$6)<*DED%4XQ&9^I zVf$4(0TGu;3cY;N3WH})>ZI*CAG8c^U51$Q5nNhQ{DVhA3tHR)++xVmaljPV z|NCd4*!2`yM@79aJty1iXHv6{u#79;LQo|r;48_aH+vueC#`&LS_Y375!8sRr zS#j6>u6->+-~FPCC)WyaQRt>~$vIiHq)Jon+XG?Swh5PgTIqWWu7f7(PI`-bsR(;N z@&Xz{LDq$A(ToO=?FMguQ{WuDT8+vNyx)qL@xz%G+wsc}X62GRek+X&yCCPROkgR+ z!shxYddn^Jc{)|JOMX;Sra^+ZjIdpb-O}?R;Cy0H0tVwg@U9V`r@IK2<=|(`PMcfq zfqgPbkQ9(YFAhjjoRlC7;Z2{#0oa+*tGO+#BWuHY-7@LKX)WNtq6>^tLXIJn8W!UE zqi^ufR_Gbd!Ldh*tsxc+tu(anCAs`C2V^A1x_3zH3xb3u4L{Pkdpx^~HpmLFm0^(&*?t(b#C?G0|H;{l1!wiBV zg9`t1lBgsS%?Ak^?Z5M@e%BYw`{camJ?A;k0iIVg)FSC-8gAIcpxFaIea5eTOH#R+8^?9nNQnu!vM2_MoztnPD%s~>?E>S1 zw&l8Sr)$T&{0MY|G8ZN=xHn3lebhUfe~WCCHK}^&au>W0ywR{afqiP^P3L^vX2o!e9*OY0aj*p3pNC=2;M8kU=~ns)s3J@VV|$!Gyb*KexA=5!Y;Qs zeM=p8P0#UZ_BAI#Ic#2fhlxShKru-aNqE!*ppE46Fw>JK%7Y}p4Z1sMYv`n-)!IeH zE|7fIpxog4UvDhTGB*|uiwX-?Y>iu;`cNbk8kC63rKyTyul>-ooo$sF_|}t*3~Q~} zBob)#XyFEn+{VQ_&6!PHBJ_@H2BB1Tv?IHoV%AY)Eftj~LLv+F3v+uQ#Y9k7Se|HE zq%Lfd;e&H!9fY1`1MG9})^9U*{>JEle)wZIchb$x0d?S*$tsf!<9&)5q{v+m$bfjq zcCR#EBDooM81jj%;X>klebAzm$PS_MISaUK$buz8G zlECHLz|P8(V7X&gNFwintY6+n9>^f}lfuuT?}ng&&_*|q46+V=&)O^5=ic2P-uz{V z5q*z3kvZ&2Wudp&KCeMnhkMFKGuMDwJfPv(ab$%t`_`S=9L=iF@b(5Q-HGV)= zee^c@Fu4uF>7#dCZ{XIwExQAZMx}nq`cjfPR*zK2{ryRei8r&KVu~rU7nPm5fN$BX zOcLZtFlY;A9z~w^%9}iLL64^%@XQW|w1w2`m*YV-?1&TU5A#_&HBo#&}6m?AKp4!p2FWK18w{IC6W`1R@70 zW*$BY>!v&jbpAxqzzK|ydN>dfxra1gMIKP-pJR#c` z)U4EnC4ijrV}(lNA>JQ^5&;l|1KJ%64FbGUr@#AdOn`1?6*w&&iWt(#kB(fWX%}L7 z5tyz7hDGw;evp%bE+pe6@Y)VUmaNhEAjJtpQ&Ni8eCZ#J#^-0nKOZ8E+_qs3yqbM( zVyfCG<|;*6si;3;A2W*)HwER(Po&f_rweFWivvwDq?SE+A;lDLrso9sRmF^@Ym-4jf9hK$O9VG1B55 zcN4z<52e*Qz3^tu%x3qRzzzkvys3&2)k@$_)e^i`1fEMtIQBBfKJ#pHI1W%L zh^Op@u8D%nHWjl^L-g0@gLa!YZRG+N2d2JQXy=M9Co+InFTN&40frXMPU^@{`o7iu zm4+WYT=dX?7in}Z@~Rc=^h*}ljrjl6L~zE(QzO{t!j6@H_#WsQ&EY6E;>(gYi`E$Sl3&y{@Ck#|D0KQ{W+fQMx z9cMwJ4flJA!RJGx{i?AT5SwT;E0cZ7!&c7+@zBi#6LXXKxj+BI9<`|3>4P0>Sj!*w z3-wid@5dWHWQ?Z5_5Q#mvgrle!W=L$6*&|GC5JnxsFP$*SgQ<3FdKl8i}5%NlH^Hn z4S82bqgW<{G?3@lDr*wMAxhbngzK+nLyr5js!5f~kB=O0Xl*RSm|Wuo>rn`Qd}uP7 z?M^Ji`rLO{hrQ^8Ic?;=pc`J;>N|@*?ZX!GGk7K8sL@h6e==>Fe!8qhfbHhix{)j+ zU<|`Hb=iaIQ{OV$k7GZ({3)q*;M7mMiLGd+m_~}6eng{ebT@2i`@_*(um&Ywt#WJR zaV6^apad^e*2DI9!#`F9ZjFQrk9KfI4hO&@c_sLEc)7~~(jW&NT{2#79OFf5LiA?bmH5mpE-UApw9J0u___?%YRq8^@C>3b!(L!pcKXg2%+%s$45 z`hxH$gh1Fw2GT{R|z-mh0_d|NjYgrvOd}5)zle z`Phxk7j2E4T#Lg27uHRvjh^CUZl=us;lIYd=#5+(-I0lF_;$1H&>c0yd5F$jSgqD7 zI~6qoeY&vG13VRXv=};&_rW+vvl9O_2uSIfbNQ;hVYyOhrTKaRRX~-yG`iypsiDZD z!}|F_R|8GQ(6atxWv(=m0j@R&&a+ghCx(K8^y+SMXFr=8*=RU;WP3z#7$6CpL zc>$kI;W;dkjvD~q`SA96C*yio@Wn6QC($oWwoz-cIwny}0!3C+QP`I^SCliSLxJsA zk*8^E!i$xb#2O1`4XGAX^SVP?G89jo1 zjyc&4>VcQj$GNu_PD7RmnqVMzDu-?omrOUTg@ybp(7p?4r~Q0v z5`fBy*w3QDec3FXEG}@wA3Sxuv0?fZ=#5y(YY?!-yjGy&;9)KFrOBDoB;Lip%-iA1 z;^LaA#$N{YrY`Vc=|chzmgr+P0@h4sP_``KbGls6F1+i~AC6+2_}^a79(o(K-7D6; zAsBc!_#G*h!>C&da@_iksMWh2DhUM2II{Awz?y!<+o~cFx5(vDxRe}K2eSH~-trT&&Edy39JKTn}$;Bn89+MjgKonX9?q^ftYyxKUL9!o6 z4|y9`sed~&~X3?ynS`L~gfodot7t7{8Vu>mDsyE3muSu+mCAHSso^VG9FsS^_DjFa66W<|NIRWTmk{ z-^9|U6u%B7dv3vANP_Ci4Vsl5N=)rz=Cec5%WU_uJA9tZ9ycI8dC%TQy`51ab0_5K zh9wTX^|s(@R0ZbF#(Wi9q1wrGL2^@1OXkpO0y1Ha(!y3u!DPPTBg)nGihhvPHLei&FtNxLE^}rQ}(|PWT=tluX@! zqiLtV5hcGZtX)TTIWS79O;A!oF_3WFLq)|2&=1K}-|~qZxd33sB(xxPoTab!YX+5HTIAW zV(Zy|)|Z}UG+=zbd19y1Lz5(R-I&uNuKC9^*&$i9$aY&2Yp2LsOjtXft)AnCiQk@b-uI#xxZiahcG@s~Sl$EGVLN7Syk2Re>x9=#V)+;)J{w_^ey?@j zdH$@Zxd=7Xq_;e7&T!KHj` z=&pRP7_>cV$&eeWzThM6LIb@H3xKuCErEHWVR!_3F-maZTWfpDViQ|m}}D)?R>_~RNmV8m$~MZ*ABZ4 zw@_uRlNY$IbWQQ=qigxS^l{LMZvm0Q+w?j2K5%+lH2vYoW4bQ9>eVlYBFeyYNv3PD zyq@06zx-Oc%XZ#vJ8#T5AjRAg*cB_BkV4%UzTC0MxORQB?w^WC8Mn2|fe8$kOu{>- zDCQ%I979%IyAV0~ae-Kid(OKSbCSM#cTuY1Bj-I~BTkL( z=RNMs#1eok8u3=X@P0_<>?DEVy$j;FK)ixni2EU{fCN!N{PT2GyPm~e9VceqO69At%7{@*zCq$d1JssvQ^3(e%2i6F(;LqQA^P03pvs`dJ zyhPK_U+32@yyw#=XjHDy9Om!0v!Qux@SXzQ>)zNw#q$AR0+cJeO^f(tNiVQGb^z)_R?Xv zIu;ZreC^fJ%2c5dF^ey3I!abMaJm}0&W}njrBTc#iljheS7}0MjA+R9s(+Wd(if$+ zG1!JYeXTOuXDgJ7eHPL~e->h&nH_TqxJ~z%L)hh_Wz#W6d}xZW@kZq=BRC#1Csz@z z1B0W?1RQ%Q2E++>Q&EG!sT_pqzbfs5?BiYVIif`a;oCg-&5xV5&hHEFJuVOVE2a%Q z#Z7Av_C;jUcjZscd&IlxvIPfSR!mC|$X1s?x#B~6k*Bx;W-K#d7rby&My2*QJ4KmO zq@H$49C($nK%2rnan_s+!JWCa(kv+X?147opD0fH#)Ec867<1p(O?m3253U4mh^Tj_%R{vUuR>jy*QEgzmFcZ`zka}K-ii9H3|xbL(6~|W9^|%ly~<_jBC#h z;l?(SIE{eIH){RYQp{G0Y^I{xg;jKeuthV(KOpU+aT}Q)(4x8g+8u~}qKR+O;8DBq z9mo^S+b<~#u=^C~|H0#!ho?R>dtBf)4-Q+C99aCp*6Buo-2AUAe<7OkHxq&v|-*((GFE)KistEU)An9>wDyIKZ_}D8AiQM|;4;VDTM#8)vw-KjWlIGc= z#QWHu2%bn2eI%$+VS32^*>g^yGWBik0;WfUEQ;)` zOu?@og|an6iY_5Aiy%RosyIpNB$!Dl<0VCWD(?jfLbqT@@u~dQ+{sy7@Sni(V}10m zw)@}QzN-$qz^r>q85?Ruhy0K1c2fA#1k*k?0eL0Gz~;G>iW>IS2{RY&@#s*joL1zP z3{SlKdsh|0(|Mojcj zUDzw@o~KpTe`U@0h8M?BOaC}|MsL)#STpUT9-p*B%l2WdqMFQGyDTCH%_3ZO*46k_8W>2jwZg zsC@%1-=QBbd)j9QzbZ5&9!b-fh`{A+z%->Oy9y( zM-=b}WvH!Q?vm$UAud%XN$?4+at-9^nw3q!lt+M}N+9ley27>uXqC;pO!eV_4h51M z+X7c9F!xgt{D_eF?g5tq+jX6q7nh`yV+j7^EJO@V4aTD|$ zrkFz%*^j)V0n%&KzQ~i*s8*wLO17fX3&gcSc(j48e@!oRkO^40xMUq4=&lvPht)_z z!vFfb%4Xt|Y+67%G!cRQ-B-Kl61qH?&6A)2WtKYEzn2D@tGYitWwC7*CR>9CAY>E# za01Vi%pXTZnJ108xG@e)akG$OUQZr)XVRxgq9i*MG?nS>u-veHJT`jXGEp|YcKYrG zH@tGCIqIHCY;wgBN;ZLsHG`=C_+Hx^M$58iHt#m6dTF-J(9n6*D%L*(jT|s-FnD#x08>9jtj?QuQ)>L>9x$7E7oAh?3q_~ZTPEssNX_F zy?bpg=0=^cs*-GXU`OVNi2>O|F?kfpfi#eBf}27250qwepo!4tfIfnI;|5{7Tk3Sy zK+M0-+bc$L)1YeugrT9V^*)Uos&PbC4 zB)#I&N9%xRy=wYrbh+zQMK=Ey!61xX;4#+BafQrSW7y-GWyoyre6Q<8m)p1rg?C$K zkA$Qu3`KbN$$2PlpR`FIHDOlG|ET9D>(KK2j}>VNPOACon?{rIn{#R3B}W~2iD@%2 z4`(T+fg&G6{*tU07?pWa6$j+Ku*l$l`vQ^(?0}JHpjBcojy%a}=fhBOI^xu(s0cm? zO$|~NY$+%xU|V_gSDVBqz?Z^SQEc_>y-;R~(g}GI)D`Ru>v11(stYWjtAzD5hVfGT z$^~_jG+6=zmGS8?LbEc5-o# zXTFJc*os06+u~%;{j<;DKnh}b%gg|y_kRJ@v>ZP-x zAial1Gqo>ZGY_N1*q7)E=~L%GRX{>uja#cCPlOsRIc$R zS&gb3)R6JNiO#4MAN=?x-HL%H_y{LEGd29rac<^|(9?wd7cX9mt)&{>0UN-91J~oA zcdWY>Sa7+#L9)#=;kELaYx#L19Y?=oVl%@LFcbS>9NxAJn4RAic$mu|za(LP3!=%W zzJ>Z3f7#=5R+7uh#$pbMPEk)L#iW={nzeW_earr5<$u_7#$0g!B|(J+dHzQCLV1!v zYm~P!D9v>!F!lz0!jtA;-`{%ynR7-|J)o}KIX|5~>&0%E-XliE6%>)f=Hw8t1K|U3 zLoy)5cR&v2NQqJixNXXr5$pc$C~=&LPNB8HvC<8h=&$qm{r^t*$wKqciUq;YCf9vI zxNk@jWx271{0v?RNc-i9CSh5B{M#B2toty{-?IO?Wve|_1jAvY_bWer<+j}DqNdOI z)o)4aGy**_Mr{zXC}t-`(y6F}5v6Wr(mfuJ{%+RXeDlgz)-J(>TB>3Pc_8f*bjjC; z$G_RVXzk(?UupbV6)0(8$YP^w^L%?l$2y#h@s+bPR5;Nc#uYbTm2~=@J%KiK`K1^J=9Xuq#x25cT-dcUZuP$x7q?>S zuTQK!o(@9?Z z)-M^Cm)5V{$|n0AI0DjSvhLMV40LCxprQ)pd%_IyjWup8wj%9wPm^^EHUu{U9kN4- zO`me;Ugu&j5XqVj)U{N_09`U`LvWG+*S>uLYZM>4w#Zim_e1kZXy?!ak|=|uN;WL7 z1kW|jdx+odf$GTm;L1VyPNr3ccIVw>w+g>{8YR0X;+k5A?=hA|t1gZIi@*urQ~rHX zORRaF7#G{*xa+E=a<}+tMdH2U7DcCK<+OL|0&@eK=oSqe9GZ;>h5Q}-CP3YC-f5q6 zlSX;tjHU6zgVUQ$-zfb9N0&IsVJ`Wt>zttex(ik z?vS`?V_jV_4V+gx>ucAf@83|ITXa|6E`0RZE`FtNFQ|^E1?>FlcCTt)cSsg}Lal{V zt&Oo|9d4}gK5dI9PPm~O=ltNTd6Nw;4ygkt^enXB=;jv&4azb?*(~*TueFh@rr(5a z>n9?5pnXa_57K(B9SW4S&pQ6p_OBeWv8AaqaI2R263 zV;oEvT8~Gv?ZLzUZC_&^JL3Wt$KBo&3#0+;6@r<&dwfV9ToE_A-<6-19;cCy)<-7? zr^$xs%!Soz+(cJF^i)cli!?~*Mq4c!Pz z1xIBQaw#U8BD)}}!e1_V82S;^l48WBlOORBX7I2$hCc-FNrW%u( z*ZdBtp*tp;CTKFIJRdwP?@$cOZ_X>5SIyJPk_4ZDwxo@$(8R#=7)7HFMnhoy#|3wG zY8O5-G~bWSBBd{lp}AmUXg;EtV-z_`MP;eC0@0`vqG_;Kngv{>4}CEbJ3uPL$q||A zBhDSrIA_FZ;Ps49q-7%ovQ>6bvB3jkz4%tMGJ(OicC9jy9SsAD<2t`u;37ajmsyE| zRK+#_LQmBEx=%L-kBGNMmb+~8tOT}!o}i5th8M+E^0+y>fc!LGXxSc1qhN)H#tW;+cx{4cZ%P5Kg(zRQ*H7&KkEmK z^B$+IiR11p+WLR=4HqL^{=}aD1BrLwy*4;0qr8^Q6q8Dk^;8trRbZ#c7R@Q>D|jeq z%PSr72f`Eb#|n@shYS7P@R0;{9L< zABFl^WF}z8_eq5h&b+k{vI9zum90`>lZ>PMerS-=?y^>}3AV(Cc{*YLLfG~w z2BC>@1>LhSO;{KB=xX}R5&8?+Me)Z#Tk25aJJ1f)o`uhasI~UQ%n6-WpgtkqEB?+a(JT-aTZpk;a7iQbX2A(F2cZPKDrVHc_9wySnvuv{~-V2u=Wnj$BusJrrh zMJ>NeUc|r58{%u39*|3K5T0{Kf#&sY4T2WUjgWf!oimd1;6~+Ei29ZXmoDf~`e?))>oys1+RL>a3%fwG>%JMS=Lap4gdAqhuh9m0#nSA?}}@`5V^EvH3fban5Z? z3i?{(><}X&cKBs(Bzw788OIF`kW(i3sh}7T>iG}_1}jNET?*m_5DluQlXwj@o@$la z$bCVniZsui>eZTkP;d;jJxGc`hPUC?j+tt?1Z-nlv}Ui1!*5fh-BSr?uh-axnTRzkuP zpMlo#J)j$vGY3^vA3&4Mck#)7McKSmMJ0ILY=e8RbDpGG*-I|KX`J^}5b0$rf~x|% zUQgwp7p6`hc4~$U8U78@tP!cxe{MJ2nShY{N$|SJ1LpE^4$H=~pli3xvj{p}9#^%2 z;C^ao9Y2mKmGpX*gPFl5Q=qo;t?P>td6T)iahmN3S`IF+b6V6WcKJICh;dQd=alg= zS?R#rdC=D$<>qXnm=ua6Q&C+I6GjIItw%j{cuD1#D~-yBy{Hs-$*o{Ko++0P-Q?Mq z0vS`jB@}Iae*DjY$uSngu%aF&PBs2kl^(WRt#riRA@uX!V z(*56?5BzQA3UX+Sbkrk|;Fl*sr{fM@CvPAXoUozL?=aN8j z-~fyT<^E>P|D4mjw@sGDTn6v(GWFiUh>}^S$z|ReUb)LIw{jQV-19yKvT{M*t9Qav zT?*yfRP}VT=Dkrb{iWC$)sjbFtmeh?*7;SEZGNkOHgj>#qn8iL`=k^1ZLCEPM-Owt z5c==0h&{uNKFIeDmX{=S|(fv3C?mC4A4QJ25@LboHp^U zNi)@Ze2RHo^+(4uU*m?(czzg%cPv9E_?zYXU-YO+2CtpIrpVz9%#C%A7OfTRqE5Wg zAnXx;9#-y~A5lZ91 zjS<*XMZf$**Axk6r?o&*XjWpeT5M=B|J@i-E{006*-(J}TSoPS&m+@Z1DJ^yGERaH^P*z0A!Bou6JI?~r6}R0YS~?otd8z(1#=wCasMHzT^#d!(JN+o)2XO=Dz!acjB!}Ptb6ncy!1zwF*eo~`%6Fkw9tHq z(#-&wcamx3*Br-_Ckt-QsIsEBJ2il2~eA zz)}9PH{upQT=d_l5RFAFIL z4d$$~BW6!c5ShXIzWyC^{RW3MEEW_QHcY?bf&mphab=BLow@}Aa`-ni8CxdGgOq0v zy%agQhAk9ss@myCXKYFM!3C$K$ah)VG;<(4QPSm}98vC)>T?77Yd&y=Se(~Y|ANpi zd6M&xD-1MIBJAKf*AviS@$y#;cTSfQ-c83=6Mx0O@0z29u6(>WG!Y^NVKSZnk z%pJ=3+(Bq0XHM_D3}?fTsnab*qU?c7GlTV<)3H&^{$2X=zpce?pQJ2co!=MU?ZOi8 z4CMyr1}J{3ncpBjPB+qrg4POj-lbH(;=P8zEpxW0@5na2dVW@n2uU9&nM?dq=}BG! zvo-VpM24Unf@>e$E-VlJ(D%bxkoVO8jZb&U+k#s(13XAPfWh3Y$=75?o?DPXP1wTO z3q@mN!}_ZST{@dzbb2rDytaF73@-`8jsc?>n~zqpNp`Ba%(eA1HSs4oInz_WR<~LB zx21~LC~T0I7`VQeo@fT$A<(l*RbZ&7FXE1{$zz}D2F-p%ioF`$^CT#9lQk(F(zpO> z)LJre#5`w&Uw!r%&)&iS9stEb;z zdiVR+-s=4Njel;@r27xbQhe*_KQ6uUjR~(C&zS-J?eW0FnsF!kOt1FSO7kRx!)~N3 zqz$kiL^^OtSwluzPx8jq2EE5ra1vrfL#}L*3z~>kJ|L!f3X1Hp6PNv$obk-M-Z)Hj zTA%Yiw^e54l+WKb_hKCuBNl=%>lbU4g`V-EOH98a#V-f;)NF$|5lF~*@su*nbFD|C zyVfruRI5BdHUw+^ws>ozA8O-bJCx!xQ_ zFBD_b0%;C>GGK+ARLn@frj+C(myW!;DI!r%7Xk9R$f=WlHe9k z)u_@XmEr*qi}}O{O55{F5sal8bE#U@DUbhGkP=XNO0A932nPgU#-XaF0;_CtwuM(7cBjSAy_D+LwekuNp+ zruaceaD$X)1-}afA<8`RNtHa7dDI6SYbdMaNBmEy>jO|PyhH+yp9QB|KhhQtj zP0+XPkgC-mlcqVKSAS48G@UJWDRyBEI!VpS0lG-l5Q_J+6;=|9?G2@|4cN0lWVmM; zN@FCfru_JWe&=5q;dN`lx>w1qmnJm0(j=|=1;q?dq@RkywXejZP!6^8Ly;+bJbl!g zs~@`7XN0bTm>{lg8?d)dt1MMfC%;Lvsr=l9D}v9%&MQsU0gBQiPFWyOk6sl!@Vaj1 zpsXKkAhz{N4eC_1gLrVW;E)XYf@z+oVS(%lX>`XUT!=A>gFkXektRF>B3ZDbtC3s) z8R2uGx|#4zQAdlmsR|Z#^7oORfX?|^e%>wa&^j-=yP9Goh!GuQc*U^?nfLNKVb_rHtk2qn5V#KKrbY$7nswUMj zLAfB_W&AFUy(i%LMq%tHIGGB^CGWC(R{WiLxRA?^-*NwC#DaD}qdP9Bd6KG#Vi2@v zGr~7SK(YqgU2=4}lJSo(O#De*PM7 zWV>S}9X2IxbZ-~-x%UL^gxZ*4`G?Nf^B(2r?}Pu6swkOtb{4FH?wC$T$Jr2cPWCAe z%X1lV+U=&3q04SqD^cz-78)mlmY5|19T1_othD+ zJiiX*L2uk6!Y_useH%Rzm>VAHvwDF3$bRB_Zb9nYXeeWT7Tm1>*S_t}bjT&^3|-hxkb#tH>HQZ2InUYh7^fmJwPQ`TVJS#;3>e z);K`G@h6P$am$uz<+f~}7rm)q9`wg}x6wUQJ>=E}8<2iQllX*u|E%#l$>aNw;VmoY zf867b_eIO|KgLFU>WRa*-#5=}I&93@LI*Fj;;7xZeP+izX!(ZZnqsd!QO=Cd!wUI% zq9oCvjICDn>sKRvMcs1VWF#f5hKTtJG2WTC`aNzCdGE*KD}3XA*(ty&hQz%zu^JGP z7*%wYN-^sxvJRv?g(b7td{u!AGvIoeiRpS}MMJ-xJkhg!1FI*&>IbbfG>9Ur$F#+R z(LWK-|A7$~8~^g=8{|AUTsUyX`;G}NIw|G`MXpm(Wsu!Qoo6k7r}~;SfoW4*llIO7 zg;)M@x?b9%$z0eSbRD`S+@oe5PZTghE}<8&RWonP7J;hoE=(EH+m<(zE|GlQ4_d@+2Mub8AfQ) z^e8Su2?hS#>v|1pEhtl;rk4lx(I?~;`uA$#H`U4eysn2n2e%x6W^S@3c>ZiSh6$5g zf{Z}B`K|7KB$pe|9N6KlHvw8H#T=wa32F(q3!_E(J|~HRhYb8tOc|giwZZRY(c{H? zrQiaw8G`%Z8iR;Ht^QgjT*HjU0O_TdV(bz9By^mXo=|UbUZXhVzOI&O`SbrUVrSkm-D~8E1Lv{so1meGVj!{AMMd>1$|Y6mlk`?U zENAHTY7j;TZS~U$9}1cvSH4WWU)rqsYf1PyO_jQzuY)vsITX(rZgkEcb}}4whvc{n zJN1X-l`|5Z@Bq2U>-Bml=$wCyJQVbLq3%f+^iOCPwv*@?jmlQykMdzZ_b3bj}~s4%r7m+ziuL zh~nfdQG;4)z+4E_Vf|?fqMxWfj@_Nu2IpPU1FF~{wosBFDf7ql#r;VYz=8R&9u{Lw zp8c+Ig2mJ~|3h-v92QSUW*oTS!2(pUPHM%x<8n+(=K3QcU7rwWr`H79pY>%vxmH8P z{6#yTTZYP#_TFo=jjPkZzqzjDXaEt&)Vy?_(Id5E7TtAuhZgw4?zJN;1g7|Ol+pC2FNp4jfwpZW4&z|AY+7Vs2_ikmGz1dH2%LXe}`=5<_;R`HQ#lb}JsHLJM+@I-$aFPArSz-cXd-K- z5omHd>NDO-F`Fs)^`nr*l);mWO*c%nUSi%=%Bc184zbJrJ5r8 zed~BDrcD|s^b^Yg8TJIiQ)6>NhT|IIqpsO5=6VuOyQ>ZySFxZmp-&@rdtfy;q!4w` z2DuVa!jlVK4)}cvUUt7=8Nb`+eC1aH<5Dtb%Xf~G6%OpGW}2*V8!2W3MUtqfeO@|X z6Re&NA$^F5_0U&>uL*BUuPU|z6IeGX)oUgbXuT0LdCl%8mO(IvJ36&!<4>b}jKI(> z47ok(-(RE`*QZ4B05J`}lDunx)y zupBbOxe@AXGu2r3$e!Z0g&?7G`J9VlWTn{2oLG+*OS>2Q0EZh|l)_lC>X8{q|7$YI zot_ zyJQPDQ{%WslpHd_L@vc-Q)Cwvwb8dic}HgWvsu#})F8h{Iw8?#I6AjrMQD6r-m5Jd zO#U4p_sB+H5CN67Xto442z3Fg$)@14898Km7Wf(II#J!uzk8N+{oYKu|x9c?`AL*sfsh+4Z^0GJ0y)z6o?5Qpef96(bP-& z=pu0m{ZP=XX?r96&E~h7es<`M{3Ta@koGUH@DdcwbNxJ?ILo6WJDZ zgZn<_K1dE1fs$Q|=F;NSC0(FC)2!Ugbcrf`GraO7)w~Axdb*7)b4K~SUZ3s0S|yHQ zb7V~rP>&a?31DZ;SAD-;UT5yTI4lFjf4nZM8`|Qs%JqeYRD}~fPtD;S*-tQ zSf?^Vt7l;wuYqu^Oh!gG>KkUNt9b?d7H>Pt2TyGF$DBBuT(oRX#){q1Lq0_>I#$|+ zao!i?<-Rw(n!~D~8Q@0O{nESY(OZiF71-v z2fiUlxsMv}Z^n*u`X)QB&-mLkr}bi^7qe>X_5UOZ4!l9iH}OlhQ%o8~Hc?SK$a;aH z9K=wGm@!?iebySB1xp#WtL=2~N@I>9XmJahoFIsC$jGsa4Ku<3e;I z>Vpkrp93#M4JHdw4aHPaq>PGMG*0e>f3Z-?qxK!*U_3z1;uF zPuRncR-7bT)Hbk8wng_7us^nN<2>Mm?y0+czTa=okhq&o201MHho!PzW|0n3n1MV(!eP48rMGg0js={Rh*WM$}`aGm^9)r!zV{EN(f zaFyZ$w-9t}fnA9OtEhgtQj`y?iVao}lh5D$WIQ-ALdJY3o1SyR$&{=qm);f{akBWr zrlVxF1LFj!bE9yQMlqWxl0rpQ1||YJI+WcWsL8>0gPbLxp;d;hjz^rZh~415*c%a} zrjaW+M!jm+Lr&nB`qPm6soyh#L*o7GDWqW<`OIVvu2akvid>?ikkf-2+U-Kr&W>Zc zrB!sA`!2UM&uhUqy=uJEXKR&lOs`iFzeBN2b{d#HmljrfjnFAji&e<4qWeRJTr0i$ z=;Vl8<*;9urYlpg$GX(}{xHuKoOVpaq{L z+U3>-G1vR_X_6#J^F9C`s^RxoIcJsfB)Q<1;@eK&@GADgpI#AsX`xj+DXud)YE>Bb zp*0ZTG?R`Stn0dd^^4by5G$SYzbi?;1KXdEOt621Vu0RofQl+wa!nco9kW|BC}G$G zRr9;U_R9;xhn?#9#Wen1MaPNi_;JiGX$LPcGFg%?DN$7lviUb)r+*w2#&I7sKw{jm zu^-YPY6Qn4@cDB8@-UR%!1wSSs}O4qdM28jHix!VKhv6rxWW2Iiy6wiMBs&@H&`ee zNMNFa4lTMS-Mi?9q7IZWAAW7clAAx-@#AcM*^EMYwaj)nux3s;fW@leZ1#|4u>7L< zN3R7M&C0%>;uwK+=4S#~ioBl596Y z(>jV-OOaJnR2OWZp{)e=I;jLH@q^yY&lDWB#jk7xx2@zW(S=6wda$MTz{=fre2tX@P z4sG((&?KT&c0LS5t+8p|r>bPnc71Um+sxYWWkar8Mfd_%Q@0COlU#7L*G9(rWRw2ze*O;NBC%-!*Xgu?7=r&V z6N0`z@+o^lApIbN(XgRf7L5Y7YhB6(I3Ny%1L0#j8Y6H3n>*omT=OKUiW73^@u|c! z_UG2jrMagt8RwSg@`d^;i%$OZ-xnK!H*e|Lc2fS*(D7PLjMT>zbAlrEu#+dZ!WsnU zylVwRd@a*1yc#j2fSMEk7EPni9`8!uyCLN+!}8>a&!sW?Cy;$(`i_B1-l@r7vgco7 z|Mm05Db%%Zt^dl3Z{#ky`R4t_x4yn+3A>&uejC}M>2$9DD!~nThVxC}j2u_Bg+vE^ ztZd}%a>Mpcn0y& zT)%Nn-S|G(M{ij(@sFOJ%)c$$!yQM?_n46>AXvzk#*d(&q6X?t7R*uZ3yTIsiKea zyVS5j8y0k_tDwZCMbir;j+^clR0t8zaN`LN*#DZ ze!;{$`G{hUQRFDZxV>USQJcA4h@N!1L`=;1&|#&unE-Gs+mwgj94&D>^Xr59>pO=6nLIAvDDVggew`+V#@FTUU|XOXkuDlzwmpK;=m?m zp9w57DP{)+?rqd{XgHdxz%-_TJ*ZXQbnj3=V{|dZ1rjBV?%0S%-*I4X*dTZ$I_O5) z=-w*BGD*AR#%h>6WiZAw`PRR~2_{oNz4fmT%oXZhD3{lQoL-kYm4BO#As`12>ddkK za2@goSUtVgv)JpRxOskw=A`%K*Gj@~k%_Dz*3Sw)$q_N!2A{ywmJxGh-hY1WMK>at zEk~vH96l(uV{#p%jw7lbdgtQkNH$GYA^r^7M(*%sap$6iX2XZoIvT?&Y@cw?ENB;L zPk?ZQgsCZ+z?47lRm+eV!v&d=`->f3=2-$RTb>s$e^BVzDBMjWIsLeIhcbumRM-NX`&}|Fi#Gzpl^C%mLgTr zqTI&6=>=V9QxzXM!@)%|=*oT^^dM5JEOO~m$B1HrAhnRl+a<+Q{dOkYrg*Ue2BxOt z8E3Xc1vfL2?&kh$bN1Z}<%?KgqtcGIzQZqxCi{=j9C zD+8eTX?}lT^PF6DyD%M!qL31}RdqpCqB^J=;FSWQ5YJXKI$18Sc5WXrJY`4la083E zX|W|dOvkbemM{KX_sicHH@_eL*v*}EzcjhRRVKOi`xG-sk-Jn>F@J@tR(YSdSDdOi z`sOv@be0F7kzf&IizaT`b+5Z}oqSWoifLNqetE5TPe@H5-j5OWlQbR*2o1W{fUtWn zT_d<4?s4xAKlB==!?BpMfZs>ol|v647gU?dlbqq7SM5?~sxhyRf*`=5@gEScA>AQb zWh~Po#|T~pJ){^AY!eG1JsyTXn>C-lcJ{RoXKfAb)XNYV&dzy*5D~NiFh~()y4txRz@xbs z4@4flYtO%5d4J~H3ZsGAlpJ`GY@9}b=QfJZpG`5)9WjH7Is}!lyZ8m+_0E^%?5ED> z=ySAAm>8_1&(5h{2yP{ALfYvj@h(2PqK)o1-LWMduv9))R>@Y1x)t_~gpKDDe*9Il z<40_HaTsd{>#$X`Cn5`u|FaP&-~E_+lboG4#_#XIWsSE@P@rcF)&qR5HE#M&d)Rmno@a-=-^XW9 zs~<@9Pq)gVMHt`#j#eHNS70kXc9qXP6|Rn0q1)(LE7j{nAbYDseATO#q$<*Q>!>u@ zM_4D5r7njGD^v{7SI^`?QtFVZk4DBBmYm~LDA?Jo?6Is@KB2F~9a{vpy%)=6xR#}OB6a1MP zG>L&7uCN(x?#K}dnt|L4*_LT>Ad_;5rPSfLI` zYC*Sk!|ziMawQdXJ#SFfA999fb0K4HsN}1LdcS+}Jz=NhpUiD^*E3<`c_s9vpp#%8 zaqKL*B-}9eh*L8UxPAe79%(W(qAdQ7Kj)g6n22fO^6r#Btyuq>)CfKKU*BjZ>&9CB z9N1spW3u|CQ_MDsY@wp|NZVa9y!LqX(YM}ify`99aF4W=*`jV2_W{p%7pbTBNDqnE zMz#e*yiwK+dJ}fs-jA{vTul&6=BT&rcW8O#60P2J%+H9Bf|2hZBpD8z=B_Y7LlMOQ zJo2chwSo(RXrDbZw94zkLqT7>)*ycR$`eV~gAvy~cSs6mZ1j0r*!8;bfV4k69~#Qq z1ux@($5UT-dw|3;JjU>qsekz6l>!$dWd6jS{{xA4V90PpR!ok)eg> zIR~UEk}SF|Fki}|@^7w`P4hfYHZipVHrcJhFEKd5vm5rx_Ki7B$X~Uiv3W*u<*l|I ze=@H&aafk7h02l|)ltz=5pIXrT#+F_V&{3bFjrPBHWuEs=Vfk-O~&_rDc~9Xkl%hZ zaGWfEY4+7X-yG$&Y@nDViX>p`8)Rl8PxDqTA*SsmL!x{<;-D;13y5Ru99#-r8AayvqdKL;$87xMgi~Yr@%s*DE=X#Q7 zo14A(*@7KQ+>Q3)#NWStm8_ozZTd!k(RNY{G@aT;MH#3|26@$k-Y5vBbE}-*MBjF0 zV?21=96mUb53*NxbB?S52B8l zN6ffH@?X4I8fsiM3Uzc9_$BFdw5TAw1lmv}FeTyGUaei&PWl85!iW9`MlnYzQbk2IdbA6XrIIH(AFPAc zBfaz3JkgL_JShuc6N7Jf^z$o858y$c-6;zwo0qBXP^5VuQnM|}KH*6b+pJ7rP)M;+ znB?3l2Igh~ecq>N)(Y1|Xpdy5D~XE9l*B$5lDAvO`&<^a5l8 z8{C1#MF5zsv(cl7rqS z1WPeImosbi{5!G>f~Dg@$yifmbL5QmL&kB98*)yZQvCF9ZRdl$g%KxY9emw?O z$mIOL*gso|p>d#KVz2mwx~7^Jvc6D}Gz-P85TjR@zkv4o-V#E3 ze{>6yJki6DOm&vJ0Tvtfu6wjgJca*F3NEY$!zh2lTA)qjIc}@T!kgD!%_Hz!wmgoz znIjfr^U3`3nQ21wUz7P-KNcmM(mV_0Idq95S-Xoyl%LWzHy`X~HqB+Q<&pse81+ILKY6l+` zpBj}7^6FXJoVR%-!BR6RHUvI)&NKeJsj|_1PG*D(`gVegIS2BEaz89E9kFl7O1nl(dq3#W0ou(@uSnZ7!4G zuu*vnk_ab&Xa+_jVizx4RKefwrSs9Mv3VXgh$#0h6GPz+{~9QS9pg_P6gZj=E9THP zk61>+Sd5}Uad6Nyql>E1{hlSu9N0yL48f?}Tr$NZQe+JkrRUY_|G?B7I1*Uxj2!sM z1A}quti^}G=O0? z18o4*N>TQ2o%hzz*w94pb{d;s6)R4t6GBgta|`UdJdF=2dc@cg7*CD$*UnF6^NmPJ zSNtfC9B^QdbAp8QakH!(N z(E`76Q(M__^zd;-=`jEU|{>nP)#zZ^EP5>E?42p%npPFn=7dS0R z)Z1(SgYq%rNVhQL8rjOt;y7?^SE&i=@+k)P$5~VqN|A2&iVG}(rg)gwZH4slh|?Ok zI(0Vx7C|hv%8~>Zpc(XDPmnj2Y?gJ0up7g*N-R&!nS(*McHtVg`}R!TOcYz30K?(o z^}?W46LDZ17%dhYm@#R zQ+M;!6paHQPmN)Z3&rQ|ubpl*CpZ7=%3sK;X(Y$wd$@&SAa1gOimD527p^DqAWyQv z<1C2|x;3|BUbcD%UnfkNkJ0Kxem@9`*mrd|{RlXHsHU&l^^BWEcUUEZt{siR}aP@<8L+{&8zx>kQR;~F5H_2t^TtzmWMA*OHB-yAHCr=WvD??Kt zEZ@mrCi)Cyjvt2H4HFxRa^^tG8fBj=u>Sp4&d8bbz|~>fB#CCvJEJnLU31d^noP1CcuQ4lvZx-U7-$+@ zgggUezoH^gi>6y}%mb)xnjOBIc-Zy#Q^lsg+Zwv*Rp|CAt_eIXUF#9+(Z<6oF@hcZ zM|V+|>wtefZMfGfZ`O2a9vz?4*w~JZ&t^R~UZW0dYsa|RtapCa^^fz6!1?Lf`Flt) zw>^{tM{UlSz^s~Lpo0Af6?HJ8mDwZJMy~a&T38aECTY=h&fhCO7?H~_@!iM!+dbVeGpEm)>1j{zx$S9d z`mnd{nf9jB&h&IZ@eK${R8RuQOCCM|0a2bR;$u{BP*Hq=AU*~MLD7i{cdaC964{y! z2{(GqufGm^e|zr_-0Q#Bx4!lGFMb}Bsh$+!4E6w@{Saqg@a_G{vB}msEKX}2(PX8d zUoMaA4ljnje41Wv{3OVZz7tUsa!3+0X*Wa};dId+=$yAZq9p{6YeM3@F*w~6e%7@a zXas?5kgfML-e>n9gWUvgUrqCziqT;LjpOR2w*=NOc{WGEi4_VBbkd9C^79%|vaLLDh|o;#@>F4C`3 zNzk^^%d{f3{GxmjmEy#vOohc7lTXNi@*;~+R~uN7baDWN6gv<;;v{f1$2}{S5HS1{^tc#KrQBTj z?tbey^|LC)i9L5T#p-%#Dz{CvPf^Xu1%Ge0vNXKj*Ws#SIDyy#li{9o^cX!%_Grho zO*g5JPd@FeqP9A*`haF+X6Y=u2^mBhO9*urq`Y47x-+Ykf`T4Hjn_Z>xvOTcaD((~ z+#hMB#~>Z5pS{Z$*bJIkr${5|;{|@QpXQ)%G0SkbPF@Nf>of&|9A&Ev(jU@EJ$uz` z{O`7KNUqJBp3aLiRM*-8Shm2#%;vQ5aP%Oehski-Bz51v`C8}6F-dW7Vm(8XRfi0r zE2SB(rM_6(vzmwLl{)#?-q2g711SC0219+KXD%5pOEzP!9=%weI|EZJJ$EuYUL<<1 zulny5Kf(+u z%?oE8{S6H(uKfK^rj_SEn}0c-`jVL|#feKfR#{MuJ|tvc6Vw1yGK&w%GdM?}zkNFS zNVwiT#T#13oSdVFRBdC!d1UUyLUu!&yKdn5M(W}FRI@lifWyXHv)K0{v3*s!cTz+X z)WDzg>F|K{3gg>o2_xBUQTToLdd`Yz8z&cmHEk%2t`y!hl!k2d&WgAtE)ledwbDlD zlvM-$vhK}-U}t#etcIX^|E=PF;P1y0rGDwcE^=WJQABL@I7uRWkzgE_)Es4&q7i!U zrIS7UlVp2{CWBKs2aB$ekPe;KXssjtORQMu;j}K`GwPJzS%;`-B0=~9Z$d{?WT3vu zvHffezU~oy++#p#?DY^A;7K{==p9#d3h;+(4}^E>1flDY}dU^;bHsCqq~z4Je*h2-}voR z)yrNBaMQB@x$sLT?}fFd4V*~@3l930k40v#HH2&ftMyZkKW7Bi@dvew|1g^Vh)(R# zO|#H^YY15!L9Haz1M_Pabj{0SW&1ahU(9QH^X$80u<#8db{p_B%u|jZI(=x%ZELTa z)5d#fJZ!j~Ndg{X!&YXcbf3Ql+jQ7Ha`R$g*dS!wz`S@OJKw=fN$CGDxVr^sMd7gFU@+*QInwn;8`Lo(%$OR1*BdJcR>hNsLJxUc7v5{c%rsoP?G;a(S z9cJLLqx4@eRm%R<`9CvlS2}<2-S1Lu%v>r?YosfT2(D(+$`KOmolTtS6o z+;fu2nW4=#jl|$CTFmW_|CahS6WRj)P`!}a=)|F;atr)yC1gNnyM<6|-F8Tf+n!_c zB6b(q1sz#yS=Bz80cuc!O~6)4H%V8D6FGHI7;u5VViqJgLM~(xfrZoP=*4KXBQ?*c zgikBF(O+O>?~FSYk@LZSFA}(9K{us$?`NlzZP30Ek#Yq10<)FZXF3us!^_o13=RLV z14sX(-BsB{fM)NY?E!8l*H+0A9U{# zpH#G}u*7w<=N%s?!4EtNg|N}QulTE_4n&HLkf2#69ES;dwc`7WW#uosuMKz6djhr6 zo1R&02&63LHG4OEJD8$tcv;yd|KXo?%P7x-hjY}kg4C#u1baIl!05jKcy@>-pT z$vXYz1!^Ic^vdjoc39Z5nS>0ScAE%wyA-nMxyJvtsItF*3ida+U+HkWBZ}kvRS%5- zj`EFNxgQDpCf#A<#-@+!z*Xk4_!uI}1g*zUJMQPK=}Z3YW%tZt)|w&X1XEAYzSKN~ z3al{@sSz$!U!Znhd1%9!+G6yfmUsblBC<2;nF+ z&j88bBP#ezHjmh~@l(k}F;`n#oEDB(1G576mAY{n|m4qyXT@d}r z2-9@>>piiUY{jVfg3a9<%!O^_L~ovO*2`#E&^o*mE**ISm^0u z=EpB1WKboxgHSK!U>A05bKV6pKV#~krk!e$pMV_20>M!#b1LShT=IjqE}NuXq9)f) zVC^|X?Nki9=-qcJny6l~M|{E?R5~zy62Zgd5e-pI1?tPK=vc3)~bFvh4&_NT@IJv@4 zCfp?%U|k26g%~9koYs&TK#7|WbwU7P9%|6#^yG^?{MjeGD8!%9ylnS6s)}1f)=>^H zpY4Mgd!WQ}4n~+E_Rbe{T657cS%sW8)sv!OuEWHIe#ve3JhFqE>{mxF5uF55M@zHux(4Jrgit+?!iy5TgJC5o7y{chjv zzt1u09{0P4v{VH%?~D`U=#+(~Y9wU9-%aiqtZaBk8Rh#3GG&+lE)xsZ5v<@x4R2z1h{<+v!6Kn)tZn%l_< zglsiItsvBfrb4*?xlbMQsP+2)5<3E!*`N$YC)p7`O%I6Qbyi1nOn{i4^20;avI)Zg z!if{0fJ51AltQgtp`u9iq|#H%t_H$TT$r)^4(}Q163hJGX2W-A zZo4u61tWZd{yQAzqjO$cF7$SMfHm6}4NX>xVvDRYm$do%ux-(#q*5xUW)3vT-?JR~>#C#)gAsCzYcEB>s7B81Z;V&9$dDgMH@p5nd9mC$0NkfeC@_5|u_A@rT6`7sd_QQpf48sp~X#1cXiG$fzB*+Q)u&Q5jWV9quRxTF&@*vX_2>geDGPVZz0 z1?oL&LXHKuOEX|uTQ&P~c)K(Y+_TGO4Y}kfyT~S2;GHEGwTV(i_eEn-Ud9MDPu9Cp zLc)f}=~43&pGxXwS02cr&?c&#dN=eoM{_x>gWCcoPQUa%3}Jy-7ugI{691z-V7+7F zTpyS4QR<(*V^SX1XVrg29e&A_N4JIYI7`S*6X2Hzs$=wx9-#JwAAsa&18-MDP>~^2 zc!Jb)k@<2A5d-@Ub%SppLP1X%w6!xfz%(*@Wh)I?QY|d4w3%C|FimrneF})1I;ETC zdiG}d6|VwKJkdWRJ$Rwwp~u^w_9fPS%Y?0aZ@=>kYM&F^ESD{Cbdr$4uI+O|{S|*80z{dk ztbqVArWW)7ljNse^N`%8FS5=(Z;IwfWci$RA$s>^{PV0P$xa9|C-QPUu-eq{SOaTl zRx}SUfEOP_*2k3(g@&v<^w$>%Qo$L2#Vd=bmLquDsWhPG!Q_?4-I*~WT}IV$hg|w1 zcey6dsR_Z6{J94>49i!?5!Po%o%b~UaHI2ECofT4AvZ;7 zz83Yl1HoULXz8@#G3;S#{FmDN^37h%SR68X`OfQhslK{g`m$5nqK-kfeKck;>Z>sj za819%JqZMXwG@VjaTASg7<6(B4P!x}DSUSz>_C{F%${4XbmSh}W)Kg**fk{4c#O*P#HX#iuwr)YgW-$*ObLXkN z>CiEAo}T^i8yYl0lGvZsfLj@?72Ikk#J`LFFLqDSWA_Cs~g-kn|Rk2 zV0yStp7>Vk4~#SLY-c#%FT&8vf}|CXko zy&H%H;y4a=elZ1!*~d6)AUtmZMj&xsH)wa-=?7358?HfP8W~^?^XGxne%Ibu(q6s5 zgpY^p%zvk9UKxCxx4_47LI&o{AwpdP39v{EZ1m;n=Dv5*IgeOT|hVO>wQvL140jaM~S!E1L(SneI?mdMd5 z(bs~GK}^KJnbQSV{I>J=v-{ZyKn?y0%^^VUV&wlHQ4o_kpv*KUqeEt38+xy3N|*^y zlD})xs2xu1o;zj%ky=7lMNpN5y3zw2|DM=*u!qc6qN`QMDpRz``y-MeM1&bO8DUr1 z)jsL4tRgEIR(oku18>7=7rEJWGkAcFUfw>THcTrm_S)r()zF%{pnQKYn}Jm$HaMNE z3d#bUmGZR{N0Sa~E**zp!u%JE@R-qBIzHw#E5kx3ttDh@2r3TD8fEIFH4Cu+Z#r4+bC~KA8oQ>B*>3KsSZo8u)BDHj z=-9Ep1w@)au^~^AL6yH`{+1RC(H|saP^4H#sE<$1WJd$%pH9A!RjKG9*Z9@@9#XW2 z*6>sOheDxn!`ROO(_eJ*cG&1)s{Z3!ut(Wyj|Syc={`;etY(Jbw3Y>0+0?ym^s>PIDQI9dV~$*Xd36bB4CX*jU|%^f)m?4 zG^CfPvF?EIoO49f%soJ9rP%hRV#+F(=21Wv1-JkmX0(wo)lbtO5gql2 z>LIc7OEho!EXOT@qXgDdB-r^#(S0Pqy5pJCQh?C(FiVQ4Au&H8H!PEl$qN{FOciVX z@$uhvPsSDpcKMtKJY0LnaWXa+%?RXe{Y*JiCVrY$DJ_~ zg45P}#;mRUlgMQH@^AD1NF`37fZ@U{@jio)LI0IhLXAD)zz}2=@E(PtNtvVE1RhCt(nJLsFh**@uVga#AA8_8R*B~2Q1DG5B!TI5ha z>~NaLX3-D#umi{F)!fe~y!rhXgTi?;E+`sMXv5M4*JrMg_K_Fn8|adY!6w4ucL#s` z!wAtvPz>{=V~6NL!K$5d(bHscd{CcuhFbTMd2Nd=%%dzq1}I4*)L$wivd>~y&e4=}4|?{HP24?^IS|_GBb#9Nb=$pmD)uPT z$&qys%R{g)nNHp$SrVu@=-MSp2O3oiUnsK|{^Gkw4i$vVi{~Hd9rRGy_wesxYqA8V zb&k=nIGm+6`Sp?Iky)%XjdqP`P_1=9PJj-L%PBo5{Ld&zo z#1&~B7(1=1eUYpEGNh01w+E8Ug!kld)aZILlYI}qlliuvwNvfc<;jU7G&CMS49Xy% zpWc0)=OsT)i@Xa`C<}PKt_dLtko=}oLJAuuxOGV!+kMPjT8uGZ{;&g&={3yKlO5!X zIlr9O<3rJKe#NpbxEV7F5(HhMu5iqeEM}cyp<$BA?h!jm3(yGBc3{v<;Gki8V0@P) zj{`gnQ(T&nDM$!8?0P!_yGynMg=}fW20`ki zRbaR3vakq}Pc{fJnJt~XM`d}O^oMnZ7j75r$_z}gr4)|_~aq4BN7vA6RvH30o0nshRTmm7AHj~p#yytv@! z3WQVC@Ba9on*aOrKm6h^zZWheWJ?Gt+FEBg4yrUW(IR`s_Uj?9vSWC+#PPC6p~XO@qK`^uYp|^Nq#`XCQ|{YDm<+3V;FIBl zDb|>n_NnF{)jq~+tKg9xZ0J(8&hraui|Br6dng3{pue}K7g}exg*+6VoBL4MP8C6Y z&_>bC&^)pxWXGgVxz2C-lw?1kM#o>J_q+j&>CwbXDw1UeD>(3`74+aPDHuN-u4 z2IpnVm%eFjxj3zvOJh=<_Pqk7Jj-Ffg96JXb16;qL=E<_!eX*(;VtrO;pN^(fHclh zCT6d}?uU%!ZTG)m_zCG3q+#sZaUM6)_^u{})cB3Me^Rj%pd-3@%1JUIOCqQQLR|!b zwj1uJ=4f&R?UV+Oko~Ga&?$W+JvGPtQyBgav+JnsZ{M!J_Jxx%d;3nCiS(DhhUfj+ z1Pf`7lt+E##7X$^7D@O+gzOhz38L@uNK*K=FAeaIT4=5EA$iAFxX!zLpdUPf*wwtYKCwP8)%QgSU-1U{dk@bmzERB1$i zL_MeQJGtzB7~ABRJy-9Zr>LWH!}G(d0%Me|s>(UL0{;2&miZTj-4f$nEwi!PV+`*Q zStqHJWV?58Ti?3`fryQwCUL9kDjY$*Hub4yU;#7zK86^4p-MKtp|&*Oq;;1(nr3t- z*qy-IrIX)-i~&QhLG-+9XC4IhYcNPXdOhk{OXN5Fu<;HyZPMJ}r-ECpc6N6te$q(j z*Gpo@%L9@o*F`lUO|GavzTg~?K(Z#UmTzZ6#hx$R{DqD7nAu0~wCmwfs&K9~A+ysO z(KO`Cb)rGZoro`C6G(BrMmT+@0r|ICK{)G8;on6_H1K&gam13@Jd>Si(UYx=gx zH5*tlyhiX?BN;cab+^OXqLuFC?c`;6=8(vu;doS-M8y;wj3hjYn;C&)+^bX)I>eyF;@(uw5YvrIpr%>~-DnjUDXI4IJNHf&cBX^s|p! zQmxV9w2KRkyp#RnWqDMGe5V3@Zsn3KGDE35Bq_)TpoB0b=sdq{25f%ddi(UT5;#-c z;lPAB3r)jFcZ~mR%i2F%=P$e>x`QUO0ae9Weuu~IU|4&4_(qyj^ddtklQBnktPo@n z8xt5=$l1XspWXcN8`e}kPOCO($amKGmAmQW)ovAG8f@3C1!gboe_I|rApQ87cu$zw z^(+dnt?{d!iVY$hjsPZ7H+OhASm?~s7RJv!Uo>sL$ts=o@fRIb^>F$9PV6gbv#?Z; z5war$)kvr@PIQTTB4`JDt4HnBWcN&g-n|J3H?R@nr&pwjoI9c>?gOfWdj|+|jW;H+ zj#2|NtEOax4N6)Tt-$!8>oYOCHxKfy;#KR}nc&dd3^At>qH&}cGsoNWk4J_^`UOhtS=8{c2~X!f zTsn)o^OE6dt%a{RcDxK2Tm}iX)~##i&AC0k3BdieO?Z4pCDFvM<5d!uW|Yq8lwdme zLEnR3z*jh{l(koJT=>YNmrN4G2IrB@uGePO@LN@%L)3CHuQ;*=q9o{}gZ6yN{)l|u zMP*Y+57`A?pA-Dep8JBmYUVmydq<;HL%u zUZZG)wk3^F@C_`6>e^^9LK*In?mRF5(HaV#(!$^bmGs(3_&2s+a@m?H{qgk`rnpQg%ha57O#IVA%i^bGD5BME16y>+zY;v9&$0I zcaPygl5jsenVmi{nula?hW?o*zQA^A5&JN7WY9`8-TS3EU>oS&6Dhs4G~$2|6aS%! z1Ew-nD&xEl`j$&HhvXTYcv%au;A*9MHzXvqi!hl8iEc%=9=C_EmgBTTE_dH2@} z6MCknf0jh$IxOfCxl=sold^ncdP2C ztX~RoX`nOAQ9dAdcx3sOs*0iAL67h5s2t^a3ZH9LwTQLSr`L8zCHvtxSmlEAEW6!y2rtfToZ@`;G}R7;)E1%c!`Cc$V)K8ZG~|aN(pM>t^zt3%w8&% zR?C1ekBx03PE!}&(_C1XVDF{mxq@N@jpxo}-;0c(K}5QGC0mn#FxhH5@j6p(vCeEI zWH|&RL94Soidk2@+C;4^&8a!tq0vOXP%}VbjV&;ONUg*o!M9|kz+wg^-!&Vzcz_PZz;oQD(dOF+ol}mlClggPu#fjq)G`aDZS^4*m z%gr(?+ok8`uAf}y8TZD=pY@DgjspsT2`rup&GAQX&-wa2>l7*4OeCE=%||b-V6U03 zl~&H%3ArNJJ2ip27LFY{E`;2UNQXiYhkaVlz6~aC)p4KJFn<@5FUxtA@adtSy&*Mi zV>ToOou$?@+r~Jt&e&zKyyOxxaLQ#6>TM9K{N#!h>FNyOB69uYUBcAB#Z(oyT9Cl% z^~z%n%#7yMvr+>G$TY8d*3Hm5QJpf!aNjQB)!B{(KQp{rtS--j)Q%&Iy3hW}Iv>Vq zcQG^>E{5bNq%cKYHssR6T01|^EsyMHW7;@oPdQe!*3w`gYU@WGdd|o|9e?uM+`5-N zNBg|AAKVt$5Jn5=FxZnC&Eh#-J||~kN2=UuFdP^^%|L-$%5R;UM;iEUH5vy4h5 zG=ZZRLPiJ~7jWlS0i#)CoH*D=v&JA-Lo>HYa#`35IRp)?p3o)*-U%#++Tg2FO`Y_x>(`<__jFk@q<`DFPQ_ARj2u>oGRwK_d6ttWe1~N+ zM*eho;1IRMiJkG87Usn|LY7ES@koJsea60^?I9Jyov)PyeF?3r&|6wUjvZInQ;FCp z=cf;h(GfHCkrpz3o3-V^#Q1h%6P3nJIqPCyJaQ`WM{CRDSpq^Q7C%kJn2}if%LP^MJUwbrol-$hT{1Q5 z7`IM=8E?7lYqKB}V<6!k8$j5OgD1wwaALFj7`-}aOgka6&JA$dO$$xZ0`^~smt}-u zUICIJU*JQj$B{@}S00dMKJ(CPLoVH9 z4L4o}?QWNHs)AbNSJ@p<%v>PAT$DWDJo0u#j~IH$h_8e;LhT<45y+f+T^;aM+ea|= z5r+9J%mm}K*(6*4_4i*`hkBn~k(}2XNZ0nL#0yLEk+;2_4?G&IEDzK~t*l&9Q_NoG zu9L?_u8cgP?3|Uv>0}ujzS~Wgn8W0m;}PNcnOW>VK#%8^#D<7#G0k4Q7nU= zX;2)}sw(up6M3AS2SnS*ld(7QEN~j8PAY-4lG~i3h@uEbyH?G_%oI0fk2?17r%PV5 z-s`+n3WR1ylcOwD>{piZvB~Kk@=k<_Rl7~p$m)>akdJZVE0a*2W5sPXJH6WY&+of! z@-o>XEx#yVM5VkkaheJXlPaH(K3p{p{ccb z(k|CVXc&?p=n>J zbow6jg=Bh+@7T8(;<>_MDF5fqWzUQBTBPl_cN(XgknzcWZ zt7e>&l`6^=tt#X$%~oEL=6jZ~dc0#~8UB_}+f~Of1eJwOu|BaUsA|6Y(Q024bRNl$ zuB3EM?3=H$0K;}dR!C4=A%l5B2bmGps#-pwS)9s>mz8^WO3NbKJ+XYVDWq~vtEyR? zt*rBi^G=Mo15TF~pbU#+x2od(mITLnXRx|M^Ld@p0^ik;iH7`Sb`|_nVT~RFPv^4d z5xrg+#(P4%x0WQWy(doVd!updAQ=DC#{HjJeFU zK*>I)KtVa#u1RR5>Y!nIu+-NqOd3tU2PfVK(e!)3zVb+p-$kYidS?|Xwud8Eg(KFL z7YWlQw7h5%2VDQZZe_~?O~z{Dg0+=Ywi6qx`z(yrU4#sp>ux91rzSPQhE^-R<8w@| z!T;dNPR~0&LoVrLD-|E6!BA11WN$!*Ta1AX8*7#EDV<#Bq4%hEtDPF}p=k)h=W40X z{aSr>^3$FNIIBHYM%r7i*aQUY?aA=xc7+4I+8D-U9j{*~|6sBSHeGW=FHxJE7&es_ zuqh;DV3Fhy>MeddqE1g9h}4uRVuKeUwJ281=;VuVp{WVcN_)inp_y)hpnwP55+_4W zhNSwZ`XfbqudI}>+2!BM?IDjy&chZbC1|Ccxg|`pwF_G8iWB2x%+R8g{Nr!dZZ@Yi zS7}^o39S2}DM3Y?4v)mwk|Od0(giEU==@kIJ{@9YEE(Hng#j`gAbA~IuvjpIA<>kwur7sJ1BIewW_vHigEq42Fk#%&cMO3 zgc1WFIb`Qc2c^s*$v3TQ^qf{{&{W+Gyx!$=Gk8-N8WzZ41GV4vtb?3tPHvc;NKfeA z-Z0}f3woGm9XZ5IV=Vqt&o8Ylk5@!v(AX(R)7QZ*eFM91AEg=uO=8p;mh|?+&IyR9vfCzEYPQ}uo21RKEGBQB`uUnGnYUISylu9~)XjCs<7P2g)KjLC! z)vV@82ZYcKv)U(*+(>nCyCwGC#F)Qg29A*7gl{pGT~Cx z`Xm!v+Fn2N5tZ!38{iTP0Av#~sL0U~>d*Pdg~`AiJ1{dvxk=C{8Fa}Ci}S8_OHm!> zqz9%$1ZPuNpI57D2`kwzA4+{wl*L|Kz-O{I;GyFSD@H1K2gurDN=90R_(RvXSMg0S zS-I)*Z>V@Dh6#jz&2}tlge;YyHW2D`K`lQOa#h#(C2}&Sc5tJK8uE%1)|b!a$$r`4 z7~<6!nE+7i33I$;bVN}nU($4oHZ|17Gt>uW}Yz9;EJ-CmjTRX z$a7WfbrM)ytGIdWG(}y|@~~vTU9MUwmX>#io|{`ouJLOTF9&OOjUSFB`?ZO%0lg!Z zlP3Zn0eqfVFHa_MzyUpcWUW6Z#3og834 zNC+vF)QavyUgHk-MXwZAmrn8xDn9GDJUp|!l#tUAedM zOM7JSF`vOP$5-xe7r*TL9i*MeR-z_AMuBw`FSzwft7cWZVJSE?CKckw23ZuZd-nnN zRj+Hacqb2=j1EfjU@6)W)#hMwijm3QpgkNTbknT#s$-Z4&bZ$UxJs;f#h+byoLFhl zu!le5bSvw~fdxAiEvf-h%U(7M%R#m5mC_}_(Y(t6E1@Gsw{qDmR4COUD_oZcR{LPP zQm7^4KH^k|V8hZns$03|+t;XjlAVfWvraq49XkpbIdmnWhf&@qZ>18Z73h)2qvO<| zi__MZ!TMot%q9@B)daPIP#OjG?!dP=5TOO~D=0tJO7FYp z%VTA&lEpkQNfa0O;3mNqVQF4?dk+woz<}NZ%39&R0>LhSIL%fT@cJVv=RgL6-%f>g zCO!uHsDam^d^sd<4ps*CkS(liWx4kiiKZKtyzJ>qz2iJ_ihF?=ksjej>&3`>qFqLO zFcO4nE(X?FXRSG{rK!?s5fD;uH2M)Ecd&oB`6 zh=0dO9%Q6s#xF{q`L%VW@GBDVp`nn*Zcg_jE_f$OtEcXWs`tMkEdoz#XK;JS+CZ$( z*yLB@)hX?iFD#F&5+?E%0>!K1nq!OkW&G$RbRpqdQMRdx*URbY(^GKoMTKjx68 z2SY=YPASqMUf^#J&+;%9{cErQ#?VmMx%+dgSto{L83K79T)7AzMXI%ZV!a^V7K* zeeJ}L|Nh}a>oixIY*H-D*Lyr>L27{orkq1yfJ$7rhh8~0X%&>V{pVs z0M5IRoZEFqY3-JOshq7esY%$xI75)bJ0DapuahjDjtOb|{Lg|XS9Yv;Xr!HqZNnY|SX(O?M-+KJ`J?G!edKN5A6p~E zY26?+s2HGdiCH=!=0A$xJfd7SAvbJa(8_>1<$Cr}XnWnsYJ^5uonTwX%TfcgA(gRR ziah{ze%txIlTV1-LP{K5p*0FrJdvOy4(L%ajPsT7(=}#TGd42ui#W0Km4@GOoiv75 zM=qms$wW%Cffd8cfs+Ek4oM&N!X&gD_JQ$;GGSf zv8%Y3gr`(H3=H&H><&sVtfPvkV#)o`+iR^^v;KW%=(c{h2Qu+wlR|NGXpkXOc(&d$IZ$Opga$T zw_<-TDtZdgnG@H<(YzYifUlca6^QpIgy`7`oW-0>_I5JYFL{161h;S;^PlR-I`V|z z46B2?d;+Rqw5sqgXe4pA3{Jy> zXx^YpGAqps@r}p3B+X5z=utq_46UO+d`4vl|5{lv(FD3{zrFBZ z)XG-|ba@u8;1ogzmi{{A4aL&Jbn^PlE>VjAZDBoU0FtlQ%}L^%fKAgXRsY-*-n$)N z>0u@bM%XZW&e4ZI{NUmAmz{45qnQsqaf77uIHTy6Y!M-C(piy4C*S9hc|Y(y9#I|_ZktrEz)kfQ>hs8kpq;^ekzl~X z^UtdOCog#92aGfX%_1@K-%kG|Bghbc`Io6dKQ*m2Jl?t=P-lkY9d_P$26fkBMd>DF zodnfRs1sP#ZV9ZDuu0a*V_2s}g}g1Yw5dALVxHc8rDqO63doq+M5{dt1ofE8n%U3JQMSqq`@i#4-t>$xyigjUm3I2Z%Q}5eL-C;o=j|YYN75Y_ zpZsH3*i^yLyCr!}O-QxReNpYyRg$fW0YGYtXOaN6>`lNosRiHR8b6)9bYdTgGiYvw zwguFYOGMx>yyAtwMkjyzC|&r?9_XV1rLDumLXaMPi(39Lel%4>PV6P7sTyi>%@Se~ zCYCS{DXM+Yf3uiZ9j=4I`8@D3nl#eRT;N2wFvym;i z7WQp9%q8==A$RUO);I7M& z=>gb$*I=WK3{1AFnUUTHyqJs}!`BNtJ6ufAEBg9(pHk7ot%Odzq1Rcggo%VKo}g9{ z>OM|0a2ltu8U=g8Glb`-EcLE)ui{@LUx{@&Jt>*r@R^5Bq^hI`C>HduK3oJXkHvg52>hK(nji#uvv_`LXeDl zpE?%YLl*F^&PL)}4chHV5y?t4neGZZ$Rz3{R-ZsD6dsG7g@e#&?c2#eL2VZog_#*( zf1R7Vjh#2WMV`pP-v~p+tTq{xTF(FY()?Rf*02rC9>6byQ?2vtqp;U|Da>9V$cLp> z?_LwqCfWlbwioRcZNc7170nBtdoFhUtKXeE&4e%)fwF?iVTLd#Hq7^1Agqj#L40cm zGz9UyiO>Oh7Yv$qk(G1Sc^1ln^0Wxnx=!gXf4!%s*A*@Ds-Uz;ow7}&cTWQ@%-W!B zK*606h6x&H!Y{aWkeKzXX%-vJ^ToV;;StV@f$&^uJbTfexOsN=k&n!oRQ}KLCfvj> z4{m_ql+#9As5A>CtRZA^1ho>{e9W(1&^0fQmF?e1em+n4ou+vX*{8qoCEFn57tY~` z8|c@i;gWD}Zv2=}nh-JYzTZ)5@dPTv;>E8eWNQd2j!>ihsRb6|TK;x^7rCEZ%D?Dk zC?#-2+QIYRj1BMJpkALpi~X-MdiTz|%-{LRg;hQ#P$=V9)>3Isyq(%>f!1w=4D!bF z2=&FtPWg7lAy$$5SydhPh&WSlT~g_O&=);BT@vUfHR}rq`Do`~68`gw^y-WzajWWc zK5`9S5^jfRXhO(Ieil1Vk;&Ho5P9nyq+@IaGQF{2e-Pu0jDSqUsbW90b|5p^L^<*9 zh{oN#6M~nR2SDFAfV9^jIcp7ha-kB_Gm9yyHuE{ z6X#u6fwa&(d*X?+C2xA1TZI@LWTpHw!U91Xww9)v8xv-<(#zr5bDKks1@}(ItY7fU7VvsqtKE)IO7rQQRqmZWv5384+L&nu zn=oKzkc<%oo@0|tk$60^u3KZGFr3)_peetr5VWX}7V5H~LQuoU<2n+Lb@EL2erY0g zhBanzF^Hc5X2uFGhToGG_5*i`X(y9D<^TMN(lT?DIWaOyEtZlTLIxE6>4bVet6Tgq zYR8<%KU-8a{Cm_@16t~56a?V4M!?CI*Xn&+RgI89*TjVsJcuFUh@(-%04HV|1`ZvJ zn)UA{{E%hB$)64n9HN#u@s1bBkIiUq*AcQrf{KU69KgnOMY1*o){y9l$YcBTe~uRG zGt#kD&d)sP-~mSa+VRUi*c3Cvgo^p?`EgXiE2BLQSzu@nA%oWJyP$0@b%x&!eNIn@ zrv$ACyBFEd&Yapu*38>ZrYkyO&zr|;RW$@%o?6V#aBU{*fp+lGcP_yApv%zrF3%fu zIsQ(yTiKhJgh!M*<A!ZX$-M^2kn zM^i*}#j7Q}j?94Uxfr5MkVCGZ?y&pW>Ew1vHo4rriQflav&+JEX@#POo5<;xE_d$| zC32ED>!<74b`~Pt7;!m{}ix(Lq&DpxP`x&SQk^2thSs zy7L{MQr{ftUSK5V(Yt5*WDA-hgIs`{b-j$bcFN6zIhO9|)H?ogxg7#DYbw!%bD)3#+DvR9dwU4S(<||8p%wdxs zR3Cu@AL-Z=1##ZHgjZ+d6+2_aT4p21j`d9TJ*US`{oOC;zw8B&XTVy07}YcRQNRiB zdRCqCC~)!T^ZMC!N*ui?s8c@TAd&Xx@;c?!8HM4ksv5;ASv5R$GqfHkUy3I?78WB% zg=6=~^r#pvtswA=j^oz;+E>IQN8`y|!pWN4M_vJ()I+Ia4DT!N4YG~ig;1ub1s2r8 z$jz>4V1?Cj9!_^Ge#~_NLoAs;>d@mJL^Oq52=n@g3?Sv+I(bHzcA`epFO8QWvs!WBJ;-g;)F`&du(UOg+^a-XfdzC! zSuW0UDK=8{bqyFPCj_PBbtrh=B}{O%ae`tS8AzB3d?a@>0tun} zx2e7Vy_O=ZbA((b=tdRG`sy1{#{o_9bn<(VO=6wjvRV18gH$3_>Z_Ng2ma4els~Pg z_GNec*Kr%k6Wp`>QwvHV7yh2OC_>M^BC&f?vrV6>jXJ}ox$JopBb_ns_0)g_>t3o( zyRgu7N5xKuSU!PO5;@9EJfI}nr@-9ZTy*H6!>@Lt&P#JkoXA1@B3{-_YLJQUBzd3G z7%MOEQJp-UY?Xo4QNSC8y+J2R8{sm%mGN|xZu$|AGlI)_|8I}q#x~irE`ctKsijVA z&t_TJvsyy7mY~)cOL)3S?1XD*0Fo@om!ITAix&x&R@mKd_9DU9j*Az~WZ&z|a3Oo+ zQs&DZ0?Oyz0OJ4CzBPVb+!dZzLZgEVgjnc#hIQ@r)nPjMilFSdF}y=$oumU=x#Rf$ zd0#+&!EN^($5TMt4wI+40*~wqif3nI1QTM?PsHDqn$*XJq@c6Z`d22*3|%SAQWbIu z8EBsjLVbqH6l8~2u%m+yfVb81?4v)U%Ve8p)$}t8tRF<~Q{#QDs{5gj0{X}s;AC;A zuy<(voX z$~h|`sNE(S2#lWwRJpy-b~%An#A=e%@cY@jeDC-h`yeiA7K7OW_N)$RltF8tU#$UX z5{zf)sIwn(Y(azNdh`4wbVvYF^N&M4!3^jm%T+) NUs;O0FOG?Gt8mQB#f>mvsukd$B-q|rj(8EITpd+1j3qHtZvF5dyF)GLSF1x*B6 zpxIp?nLH=X`;cTHB2U>J-bcno^+@5%5a+WtPwHpi1x~W{q5|cTU}Rzaw9Cla>)FRg@(S9-1BZEDI zcWe5D+OP{$rt*y9mbgvYN7YUpt+_F_XV$_^igSEdfs`SR=S42PtnBP6jth+E?DDtKO!C08a?km zwRAdalkNP`efu^GhdDrOo_bU_G6D#(GbGi;ngqjX-TgG|9L?NP-=))1f?$&rCpL8S z_P^ub&%WR`Mo2IdmyMt>dtz)38G&Lv+i%$c4_{f_nD!+ud2vOPYls3z!}-NauX z(8R50?W1mn9*Vpfn&$?tt9LF5ACF%MX|RKpXTwXgQ4$vRo(Q&Sscp^h#+Y(6+$&@F!yo?T8PmW34O9oOb1*sTs?F0+iIi zD(;dfWQa-{+HPKi{EH@Yv2d_e+J3cT-2CN z;zt3ks)OMToIC#Ytbzb+`jGG624fh4zAVgS_js>oAD?lClj(jRctl5!nJ2=vFL<6z zJL&^wdNFzNy)x@GN}6oNZQ#L+A+AVE!G)IrEiewvTS6R~caZwr@Usyb%%unGqwY@{ zlzicZPK!EG2lsNoQL0O_A#zYsFEU8lCdIG7!lA%31V9w z|7^uWv8xFojo+yICl%|&h7Rx`nC+mG30V?BB|J`W!MKo#99t{J*pa5#4gWV<6=lm? zG`q)kL7P1^28WC^+Ias>8)L0`-)IQa>ZFHVJGhu=hHgyEM>|6`Nt#1;dTDaWYWAV| zP4l9ou&FqZAe9(;UP1?`kdDyW*_*=_?cdmI?Qda{+~T}$0*cOU#D-`&7m7Tj)ovIX+A-&%AU&{) z3+6e*2XX|rrO~{~IlBT9SkMqelpkl70}Jyw18htIP!_WZ9OxmkBH+iJ1e zrxLOa1eHvvGgzh2A|>&gT4@Fg8>y^jZGk1@Yi}eys&+H>f~xhZrN-oa8viBC-s-de zdPj{g`b?eoEBn@=_DR-G{AY{Uop)uVqus`o?33;l%F~ler4NPqLP*8$b;T~jV+dUn z^lU`1qku{~FFk_meh?nxX4<%XqB;E+YK0STTyiZ`*+xR9C8)K8`Y2U9dmHwHZx+{q z4%13E$+c`lU1%k2Qn9(%GL_bGiB-?ObsJ$|S{a^w-ccj;Fc_vpecT1_1OC=EKCg(q zK~oXb5VU;4k{K%|l=5RaTIupB(ZS0&hj|I$zP$&ODa$7;2C585@}B@Xeo#@CMBNTL zI(L(ILr_A$2dtO&a5mr&8ai_P=DE;o2><$QOhRk{pi_>YAWZRinW8-xfGR zcG*@{EaW{bpHSwbgA(NYple<$CLlj?en7flnR}18N?xkE6}ZEr6z;suN%zMbw>|^+ zYz6z#j5E|x{(AR1N+<7DcD`QbjdxTCT2;t5*+@QOtq4k2)R8^X7ID55Q)f3TK3T(B z^MiWGDNp3U;qkZe@RBft!&GP4@(Da$TSjrj@bjVn1lqSAsGMLzm8$C_4He_W>rlD{ zf|3YX0zs_?j$-c~_cjr1W%f+j@2$apryV4oGnF30mpaVtHp+jPCmlOPFaP08+4sNQ zHuYt6Ck+!!c6O_CR;;b)L}Ttx0s%EhVTq#D2v*&y}N% z@G#d4&z;X+7a750eAb7*>E@eM#>!2Xe?!GPvC1g4P#I~2ER~=(5Nh;(T;PMr+QMpv z!U@Q!Cp)-(Wa9M9FijEr@N7Ie7jOplGG|x`L1X2Be>L#IJNMzeH%$0gy6y)*rfQuS z9~Ug}ae|P2PEdylb-B0R19>iTyb2Y&1nD72Rhusya;Xi{0ww4!!H`R<{0KBFK0SHo zYwJCl$b-IViul*cWybrUe4-9|zC#v_AWo)rTMA@c@vjZJVEbgeLGNyy6k4s0~v~9@Oyn^A;{F5afn6aUa3zwT+GQobX;Zv4(#r zvWvS53RgEwYf)uR#p@kZ9|y1F_=ahxR66-ux0=b@rd|mh3T;*GU}MMAZ5|^wiyRH6 z=OX4g7a4(xxEvMnm(f(cJF&|jvXRXEGzo-kH9@T))OpY!xdnsEd1NkG3r$)_jDY6| z;l-D1fX|C3aljqC{_oUV=MX#_>2cmj6-AR*pbgv3-pAeKR{><}70^nsl1Krbj2_lG zR)%1MpvAM(b8K;7gZv+_Hx7_8{gm|TSMeT=3 zwni)6=zThT(B;t#{j5Qk{?O9kogRZOOE}4{O_KCL!!5VH>*rnu!{bE2cD|n7Bq{ga zr|4wu3t9xk)*;=b-S;vp&6OeCK8`c3=@yZzL6lQ=Kk?cqG5Ita`zPTvyWv7SS z=UwfS``z~Mwy3HWe0=)D6ym~LP%Fy55OPX2hO##b##RK3NQ|R+h#mpMs0`v;ac`wr z2YZ}0iIFA>b;GM)lI(Zey(I8+xn39_mal9aLpoS9QP~0sv)jk)kP#%teVTemVqJ^# ziewYeloXxxsfI2rw?r9JpB^DG;D%|7ISs4^7`h|9tz*E*mOHQs^#+)@Rw5Ca!+ zAJwYDCSg0M3elj;jmZy%>49JPuJt%U4L~R46Vz=eQo{Bf11#gS8>SsH^q5)>84Gyd zXf>BDs2ko`dBNx?@fj5~)N4%{`HJ9(rpH)AP!4&0<`roNbwiS@JWJgV?Ub%?-Kn@b zt6TX=oD6$ZZidrjpsv0;t5sDZSix(P;!T~iZaX|;G-8;H5(Y1N{697PbuW8CPPX#; zOii}Z5b&uK-jpT?5~H$SJLA_WP4-*yKLt5P2rycO#{c`^FXU>I-}PR|-z8MXD^tlA zJ>J|^)aiS+IXtaw?XV7X@#|2kB%H$u`+!t!}1 zfEc0&h#{`?23_#K9i(vtpJ^5s3iWJ^O%#S{rG-GxJ#2p6h;`8mRxDwzxw)3&YD{`%W zftmkEgIvIa><3&<)^g$}?e^U_9tmEVLo3Yd5Am@-}4bk4Y|!TgT&tfttAb=d2e`**Pu(H z=fUu5w*xHw8a2F`^)wn;~rbQ zSrvby9=SN}-%0`Hs99P{JRw^}P|FGRsX5!j+eAne1%*QgUB_%Wv4qKsfoggGSe-Ef z>bOg<|9a1K6COVKw6luZ`pS4}4qBjRHz6w{s1j(HDqa$dbYZQk+Mv$Y3)onI)Fb#wbAMeOy{u`BZq zb{#0hZrHb7?62cZ(1~3h+yMDu&Z|n& zEWogaki`+yN zd{CcuhFbTMRhkuBpd^cs0m*S1p{@@v2Rn$cmvp`p8E4xIeJv%~6!;4wnWUg}TzC4C+)$*V2z%%-ld>GB` zyoz`3AJ21rO=ia<+0m7h&WX(q;KVocd~7FVP>#BlP-leYkr0mI?S{I)1CVjBS5eG9 zNp`X{)owRf+u6wDVQkNXy|Td!iE5K+`k_%_n-uc`9^azbE;;FmlXL?39f5C#NPes& zEPpxx(TtEhewk|OCog+;ZB@{+u*E!JIB1s^@De!P(3fk>RtXCM+5`!UC&uQ45hTXF zx%GqW5R(pBsqNcIZF6EBa@0bHR1-3w6$%9L zfjUX7%#=tM=Z00tD=6#)XAc!)6Mife$cxYH05L}R8UJnFhQgU9{Crfcs;7!x8E@+m z3$#@cGU(x24!-(#(bKw9nWwlsDN9%Y*=Kc}eE%d!QR*SufBV;7qWmYz+#7uJL^{6% z96fOAq(`0Q*Kv}kZl82+ZX##P!X?yVzgE?{kYs3IdHy?pJv{FkrH7Q;7S-4A{Egqj zz$sM@*lE{iE)T1oYY(W68|MJ9*#Tcw}5OsF= z3RbmSsdtLl?lG0Kj_)ng5BrF-cZdk@ztbD*uo>LPN=XT!zfDp=9 z^T>MG;MvpTKN{=|yI9!;IiSuMdCuN1oIUPkch6#suL5FxpP-s)hW@nuQTCMGTB7PHWDBOb+b6hqsA{$ww=Fc%2IF5Wv>|C*ODF3<~EUYh^i3u zM^w(~lrHsNgcgVFeCxnmk8dw!`8%lQ`fKBUp?-nEGAmM<%6_ajy|l+niWw#95U2 z3=*;j1a+TK_lZ+QH#zG<`hZ!b?N8^M_mxDiVO5sowwpvG*l#O{M4Fa|TaHUJTg?CXp~A5e7tM3l-5g z&P?aF)An|6@BMY}mEK$X_4dx~P3LQ;)8f7(D(HZ+$tIhkD2NIIBA|`}4l0VcU?MKV z3<4^m!uLE$P!e+_2NG_yU;C@%Y%lb@|2*&WKFj}K#~F0V7wk}A96C##AGX;)OMQP< ziwa`;p0%#%mE7QNSM~7MfVxXD)br(uDx`YJK^R?I9f0-NP+-T4<*f7s{yE>Z9_Ztw zsF5QKMKCW;ZS-wX^~f5O*Q8lMT(~+?yV5h3v&nxwV7HM*cWBVXC<$NXjqM_N0{k67 z^s`>k0>tjAbU$~^Z2Y${%&>j89e`H-7zdqX>)CYg=D!zi0^qMdn_WouI57Yl%vP-` zN&y{1hj7)pL~^D>Cl{xWpa)yQ1rbX+b8eNogVqD{n-=NcZg4C3#SyKNl>+dLjfgv} z>X92lJ9-Gqlus!J*IF-wOyfpgt7QF~+ISJJc3mXbZ#TIDiKGhPzf6#C5-!$!I>j); zL3Lr+*oa#kiJ`IBHdaRLwK;}o%7Z_VtngdocW;`*J)zOUWGoP{e&;7aHfwNISamZ=!J$7ZMuJupr`nW0jB zzwuqml1Vn3b0-E2Q-P(tju^X=wdgmsO7a8_HzC%h*%qu=d(P44S`HpxIu^A#Cq@K= z+FX^7PJUBb>$=KKi;eKfUhVwlK)q2MR^_8z>d^@SngQq-yEW@c$Q};V_vZBMg93bTDvAEmfQk!~vbMC4vCP*ZpB z0?OKgh;~{h$30An>T2kD$=)E89$Y>FO7ThWv=tMQJvV~pLNUDxN}DQ#XB8c?Zo0^% z8(vBG%a#UZ!@n(>GQdvKq&Qvyw;MDsk@XG>W)iht8(z`E~yH zLK0W6+<47HIq+58Sj3-EcGKAuj=IYJsZ^JZm0eC$TkTpAnDZ1IR#x& zcr9OilH8xw&c74X$X`AoCP+VJ#e`fRokFloxKBGphB7Jy4!L!E9n)tLs>5S<(55m>7ooPjHnd+ziu^UA`+ORDu(g4+Wnos?Lxq z$qmiIC4$RhT+&)J@!Ty_FRL>oEt(XvWonj459hW_Jx!mTbZb^4Q2e9Y1C8FiEmQ5c zLOq=?_+SSp4Ij>KpMQU=ev9Sy+i8P$3>#{2L1kF>dus-i{o;1IdAyrLjMHJu9NTT; zSPaB$F044Xue)rUEU>KcWgx1>3fi3Mjjmln44`8j?f6H$4DaF9c=Jx%Dtu?=!2ej* zHn3S$oOn0HPzh1)*6(pOw3hTvDfUT;+8gw&g#a-Q879aWVy#r?_%ZjeT|!tGh_T&T zELq_>mUx{LZ)+Ihbr_s!r%{O+Ib>2)jr^j>3&Qgp*xpEW!1$CMlonJRQ778&nL^Tn zfQ`oQ=J;%PY_Wru&g$4E-^QmeefD2JF>P`D)34npElwP-yKlCgxk)MX6ltfTu?i}S z-r}7}UjmxZEcG^lo`z1M^$Kiu?v!7UXb}(VK#^t-cb^Z|8?0BP@oqqqQgKATM>ZV?Arp9@-(%}nx@jb!d!oEGt1Em3Xq-W{ zO@cgL+8co@t#&JCu>iUg=zHQLB%fXlff{Tr$fmXUw-{2~xT(_9SY`pizG(?^<0$F) z3pNTabF4YJqXCa4n8ti;1y8Kpm)Dj~zHnly32Mte@Loq!UzvV_ax-A=rW6Gf$)%$2 zLq7o&kN@o+Y+@eGdnk_)bf}Af)oNYjT}9`tKW$X>M0C#5g%A26X;o3c3SjEq4J`vj z0ZVwg@THtwj>Teil48=;fJzi~Y+Wi3sg0g_l<-R}Ev;6hLKZ7ZlQxB9J z%ltCu9);-q-#+}a)>tXiElz+wO7v(j7XW#vW==ksiMHUo5!Rm@8$9d;1#8yEwmE0G zi=CLnT<6iO9XrUS>fb#oz_<80-w~BiHF|Q7Njn;2i%&v7*Yr+-W7sdhs#dqHcbn&(feo#n=ks zZF)Uyzm}$5f=y$OBH75x?Gc>9Z8!S93PBM>uGN-;yzDsIDDDHcn69JUGym6IovZfpl}p09}9YI ze)gj^c?&HW7uY0&oH*Riz|U|+zL#?-vW9z}&W{{$Ipba)+~u~#y(BVS)Sz4ynCTbG z!7DAA`%tF4l(Ww-5qPGOI9>GFIp^sWJe}2GtIe-_-Xsk)oSjL~KBb$|AvdR3Lg7|zqiaT@932t5EGCJO(TqFU*xT_V7SWax4(SLnkU zlwlljA%-pCkVTXBt(Gbj-u(n;7*6u$v7MemQHarY8g{i)= zoGd{srx;33S`^y+(6mrvNGJ=^4RLdx@Nj#KH>Ic!hc`m?i)1UOcUl!#8Q|aGr-|xi zkVe*Vssj_ItrC2yYWB<$WO%enPP;yof&sz`nMEBT_5-Mt5bxQH8E%0M$SS6wW; z3|pp?1l4-Cg`S;L9D$byUGhYQip$Woc$1SSh;h^210j7&{FGT!%W(1Zp^>XBOQ%AN32#4=Z$1c{9L<-9*WU? z&-Wx*GJ$}=&#)-zCQ7k^BI~GVOy0LjFvAQ?kwAP5-|dy2fHvA5c6;O(KKu?_K{5Q9 z@%YRN6waGs?)um7=l$LUiUozoyvc2LfjcK&tX7%z;`dXE2NdZwmd|xUa1J%*wI@mX z)KV^%Q5XpiVF41PLSZjD%K-hGUMGQf_q?)D+~nO&s-TP)>+NELV>ypu|6Lfiit3;% z`0Ew%zPN+eq9D@pz+(aBZaR5U04{G0N>ta=Zt=j~93CdZ7H`nyp=5bfZ$xb17t);R zur%tY^pOg#X^!@3cJHEfxa_V{Kbm)dg)7dsunpN2P zx<4R06ajc-YQIOhTl&;CaGu(P2P1REgLYL=G?_#DpuqmNm6auZZP_<&?t9sT#<(F$ zQ5Jh^v6=ykL9^6Q1iA;PV8*|j(PmWGf8t*_A}dt9zSzC_UWCb#e7`h5jTEz6#+=t{ zBTZ&@=P0FsO;|Zl%KYrh`G1O`R=B75mVqC>Hd`i3e2g&eSTFM;6={~O z08Sf5X6R9fB{v~loF^D`xkjMa&QSi9qB_deah5AD1G(;gSr^^zaR53Smk6pM<1pxg zPax}RMKFY(r|aanMO+tk7MOW4C*`P0%~50E879U$-BzDxwK~0ax$-xzFT07wj*}Hj1fInXMXf2*b^4AteYzjg#K%{&QOAfe3a(o;bn;H|S zMvs{hpM2-nAAA{Ra(soZ^D{^hJIB|F<2R?x9N$Vx0c+VoDms^w#%t%}a&*-z*$ew& z^blJy#44MCUn<$l2%22+qWpkeVKrohriK`rB8?>98+fOnOTU|YmwUyxRIS|{+#aHr zRFhn&*x8{m#!@g^)2u3v=yFYvA6D)2(Zlj)2QxJ$2t8ZRp8GT+Mn<@Sy8c8$Oc2`Z zl6H!$bmAyXmKi9vQi^zrY=nSyROZ~>+&&T`sNwD=wU7@f4QS`5zTMA#{1u)2s;qQ| zy;qqRzf=};{o;Y_e3jK2M14aQ^v)C$GSur=A0f%?jEfWRst=jrsE|^C0Cf&FpX~K| zL^hE_0hx18&1(?d3@@fPkuFIwaHaNeYu)?EHIl|VC&>?9;jvHrgx9Ru>V8JB9r9mS z#YJA-w2tmj_tHx!Y`s569ECX+=z7jt7*5z3RYSu~U-8pzFFV25ty$WG!o$)5$q`X!oP`gRGFf}y)n@F zV2gkjwS*Jov7+N70hEynB8oUzD&8tzip;65+sgf}IH|{vx_>EGEQ>gR%@J z1UJI1lC>UMJ&gr+M`kZV9X6wkghMyER=)hUW5~+G4m_p(ABoeBdR#3NgU|Pp%mc;x04Rr1sY!|%lR1gm36G&$7(*Dca83! z_O;h7S)82Kj4<#r>FG=m6fy+VfmhOZ1+>0z4u%xW!I=kVqK;>&2s{l4v8_3eJ75JJw;(ZFKKJt; z4=0%Ry&C-oTC#Wo0qV4271$doMJz?uQqeJjHn;sg+Eh`NSX=tW8jgYOHRp|G((&*B z&EAczui5OOaX4Up%~pE4ocFs|c6&#mWzCV(E-(yLMC~Nz4P?fC;#W!U@j}_MT7pM- zKF@EpC!WCq0Q6dKDR;fO8pIpw=(CDbL5@eWt?(W07&+8Dn7$gU;7duq@?W>TOqQeh zS0#%`q7z5VO3h4KE~Ut(NCp+Hr?HFOkf_BJjS@Dx&|Bf)y_G~+SclA^}T*amT8Qi=+1Fycsy z(wePu?A)*~e#Q=)rwFnueptcgHJ7EyS1h@}oEA0=>|SVKOnqy~Q@gwyr2X#Lj|$a6 z)2ld0t(+?AG^p3y@vH^zYx}~0aqigTgK;RkUSNff*C#HR^cx$sf{N+#V59ZcA|JsW zPn1F!r1XI0lPQvv|2&;cKQDOA1gJZrz`2BT)m`Kf~3o^*d$KFBp&kT zk2}6?0{3{hvKHT&uiXE>t7YV!P0rtW_iCWPF?_SWjZ%C1QTO`{KD^kBvhB(|c%u|(XLj>cGtykXydt9p_)&ZDX=;f8-Rg41)7Sl%e6PXiiCNBF4uP2=!s|@4+xe( z7;;4&hXB*a%xhrL(YGhNnKrwB(VqJwS@n|9d%=Ep7(_my6bTgBL`8SNb{6VGFxHMS zRg}<1ZiTC%m2d{uk~U$yuN~@)ZJ#-I(6D{nvEIoHjeqXkCispC8ms{E-Tj(!rJJuhxvHk>~usaI{D*H7209?s1J z4$0)Tcxh%c~HAh>x>v%a+(*iGxwsH;(+2P{_#D8+S(Tm?ljVDS0)x_l#W8Xx2Ia7%+~eG;eMlP(oh^3VIGbL)YYcAIwvO#a`l z%gg4Y?pUv+k-saTo>SnCq!X$1#>fF)edME%HsSDxS58+%8^EJeKw%3MD8=ppH%9uV}_$sY4}dejG;B-emrqg}Y)Z{-BD zL8li)9He_APLf?yS~NLyjq(WD^}=Y*{|W?$5c~ot_RW3xHJ%A*(-Xg2PgXc_KsVD2 zI$J2kW{SjNK-ajC8Cn97Vo(o}+QZ>PZ8PJ&5FyC``_W z`=k38$yO(JKK7fLiX2J-8=Q118Vx(tE*hoO2LocFpmS>w{8FP!veaLT)B~vS-YQ9R z-ODi+A!w0uV2Pk3q)Ccm?DnoL7(Q*b&%IUebofwq{mjaOycYP*_hgoxJFkcZVd(QQ z7PBCO))^YOIjcB2d9Gg{IptyO4;*w^8M-nQ$gf2x=IzmNnB%MI;xyj?P_-Vjot;aES4;M>P4CPHp z(oWhyB)becOvC+E_1?ImHR$yvo33vD}n z$FrLpRu8%u#o_A$<9%~zsKH33Kl7^KHmk`z$@b z8ioqX92#S9d%X7fXs>%A!)%JGQ+`KY>zXU>cZbYV0gZs$7?CR4B}5!95#)+1xY%PF z#{(l3(E(Y6^5B2OCCt-(&%N)NY>s@b|NErDn$2YwH72VGN@@rSb zA6=KXXtsMM2c-q&0{Kz=oOMbR5vbrk7VnzW1L+_1VG`tZutLQM%7bqyH&4tCN)9Rt zz-P{YcteY(FXURtNySBO9sL(=(YveNk^=@@jzx64f3!L>>8(AkEt0?8BL}7oxGbC5 z?X}bUc6g3tEu;#t0AjW0eGr1^ku@lD-uN=CB6ywh+24YHiB4hsBO$*~11#8&#cia? zb)@~ZqpcB>l^UsSgxMOc{ppj}EZO&+b}!8!3xE+2>_Bam#Jvgm>i^CCw2c55?W!{B z=a>OtE!dy2c2(b4mMAASNDNC+G7qDkK6ubZWLdWmWBJ7+JUhVu z;W5kRu~)>MVd#NNg+|pSg2Fe!bGhQDgIzb|m4a^ARCzGMyJUOWHcYS`b?o;);R!Ch z>{P^u)P-S4`e|T1!GiZd@60hGOS=qd=&7rkkJ3t<}uteb^lXU9J279+4)ab6F&`sVMyYI1ZRy!TJPBGpbD zr@LzA)SjUf4HWqlilr1KbGFYj{`-89R7>P4L7bsbYNIr5p0!dbTyJ)%a{|hPH<5yf zTF5OP@Y&+sAwsIZf{0_@%au=*SU5FQCADk*hcT3~P6}(O+W9@=N@$v2Znz06r`Xg) zna`e`Ks5|sRzI5mQX7s~fr<+3%xL#9f$E9k zp;)e)zA^g~rK9NH>;M=Wrl0*S>;QB17fZFOWX>GL?~x zV-?jTJp)2QSn#mLI~^9ZGgH98@NtpLQnvtO`roR2o+x9&w}xvkDthF3q7vmcdH1w9 z-aR+`ACk-MnshY+Yh)lLQ9$zxAL)?#^JG5oQZ zwHRTwv;*j@nx&CXz8Hwipu5#|Z}Y#I09NQ0{wBHV#1?753`|{=0)jcWspyK}w9q!; z`cPs!ry@1lNeIm z=v^MRX14JP_-F8UrG<8YfDu-EVfA00$jHB)LpQ0`K)E}}dxAcMHcx~)pNQn=vG%_Q zFEao>hc#w+S=-A82)|q4ZGxPBZpam~&50pbW(K)@O0kn7SyXi9++NAP8Mi!$bq+p%A)2yf9KweirkN?E4+ZO{lEI=(6H;J0WpxHo2BdKXu~F5$tqKKyH^|MTZD0m`N>m|!0u2$in^V2 zL$kp?S)aID+5m~qRQiedrdKw1tENG{!X58yC&jXMKHdjp+kTfg%cpnnS4XD7Q;S{k zdXsC>+*JCJ{rMoq3^J2t85uV0?wA#@(BIW$Wa_?B`8TqRU4McTyHq>OOj10h*hrCB zDjMtZj77t^rVYAO&0RD5SYW;w#5O+*bd-%@niQL1!8Cw_&hPDc|1RGIi?_~iJxbQF zgM|}Y6Ihsr_0lC#imeohr=mA`C;94VD2nkk(tqD{-eAzQkXXoq+!>y2^asq4dVmT;rp7*~%&2Qs0TbT&Qo?_*H zyf(G-y;e(U4+c$)PT?8Pm>|7)K&B6C(Oi+Q^K1;>3TjhEXadT3g73AN^2%yH2xprc7Fl4ULO?gUOh-%-+7&M0B%hB5(+b}Wu@kjmSmnjGT zu`q#UnR@!rM6HpeE>8d-v4U#c%)`^EvJDd>mAXA0&Tb!;|Id|wdD%BA3GyM@(gLcP zS1O73jrSc7v1}`oVB0iXdC*F-#w{LmPm2};?=e8>s)hvVxvIeat3B(YLo*|+}kL9;+MxQcWdN3+ML*i zDzlxH@xZ~TDnHwG))FU9TVKn-x4-|LE8=QW#Vey5L}#UEBv3#-yFr-7D+)-3U3EW{ zr1WwRs&9tXa6fq~*(+PU7fP&4XF#X3|5<6J!*{;tm&D$?()+D;__q#gj%8Xvg zSxt(n)+a-p61p^KnKVYQj@L(8G{c{X<>a_!(D%PD{0xGc^|{iQA1{!{TRcq zXMB_$WLn>HiTlUe-5N;R@6XgvS~06|2#hvHqY|& zk2>@WD~s{kZ2t{$ex^m{54&qNkc?L*A6I3zBo$K%5S=QdqIXQ$EUF}RqHg-Oca1nk za40gvr&-5W$mFSx~fzvsodkHx*=F@hG&Y0@Cl^Y=$yjck7XmG^w92<=q z(<`?7i+wgAyRWg+60etL*35jQa(5fgCmaV(zCDdnvNpf~4;fNsu3z9m^|* zz8XVEa7MtAH;sWxBT}?w;ydzwsP(E+Z{_rn3UD#A!*aurvke=DV?`Gv*`d{e+F0I3 zVtm5BNt`j^#_~J*Y}&{C=@r3^^y8oOTb2(wt*^pR5cG_7G7V~=4+mmK5v!qf4&`29 zK#;9?Wq90Sr`h4vb3VsE%4AObmafew2iUnQ&YPGd=gh2PJ*5E3xN0goF|ZO?555Eu zA*^7=$`aHn+fVj^+X2R{S%t;XP(DvOq)ndXQ@Ut_dL|yW1DQMOnE_#k8&FdQbORUi zHr2(s$pLjy1)%%dBi6~0Fe-<>kFu~gr8;?v>Qqp97#6ajPDF~T%BLgjrnJ^|EK9Jh zz_YZ8HXe@}nBV@vZ+VuUFN4dtX~7k7!Tjs;3InTfi~yB$b2)wPi6l*)9MH(W6nI$J zN1CDL^Of9aOyhxOCzc()WABW}1Ai`L9)(L9!?bMcv4ZVk6N z%-&9;-Cp;oU##^9S$)r)Hxxc0+)HPvJ;f%-@5C8m2BoTUw%*!8^zu4_ zoj=+%P5}q=#VM*BS}(!wiJH)w(6ou;v=bi{%&jBSu4lEM{c_LOO$$`u@)5$xn;FBq^$sQ|<=V&Da6bx!`rv^|GZ=H#F`1Dqa%^ zXtroD2$$=3h2Idkn~Qhu_p759?t6dx4;Qv*;0#&$hMo`0nCbpMqWrQuZS0z9?E7w;+$?x&Yg`P83MRPAo-cn!)e zT%F8TYdR|U*tjBDz0F3DFcp4h~=FTCI)7zb@Bp9CUk-6`ArVbAev7a3(S~5=-{KXKYb*7*|$p9y@u3f zP)B;u<*>S&q>6B{iTAbjJ8ieuZ0khZ?HAfO5VLWx(&hL1ue-~xzwBw|Yas{ypC~VJ z%6S#S0T(f$JcI4k2Rr zPj;vs{-k3+vqCM!Z%Y52#6z~GDo#z9F+*&Lt-I7!l{=FqXC6FguFKtoszyCIY8JEo?qIF;&So|)4(2Ll__ z&9HNlK5{rvdo!#ks5^4d1#d0}ev1a>I?pElG0|mlC*6#CdNs0m-xE9-KgpvKRH`<6 zUY8BJ?B=!!o1}e`o`_nXevfsYdb&E`oabfn2^swE>+nyW2>%5-DV+-0#_~kBW(~R+ zN#JpJKJ0HXp8Qk@HJ|vW-IMjmktXvu?4bA5N;>=lE9kwR6!AOqvez{jL%YUA6)?LS zQ(yh|Zx%^j9m#WIb5w7(0X|46fRb__s&FFH zz-L|qvW=1OJvXQf0eSyX?vs#vAz&wo2i?L&5PNzp4MU?7}DkB(6~?mU7=C+xkOpO-HODOI{9Tk zXopobtJ>*nBo{oSv1ENloVKwcWyIk}eFHnB%zpS^KOc)?l@kYo85FDHAs!nmYEz!) zwFTdeKtCgcYy~&EG5AqP9B+wWPb7FM{PL-%rZh^gO}PSWO{w%2D5W|UcscS6bYLUr z)7^=8h3%Nhc%)V>US3{t}jDDKEk%*XYFHVGtfU0qsN}0dN8G>_&~~GLLniHjjyn)l{yY1k^KsFA%dUQ{+GCVX2|y{bvtD#1MTfXld8k)df|GuK1U zjK@!AYho=&@cV%8e0QR0mvr+#+CC<$*(DX7H&#h<%=T1?lmZaHnTj?_wsz3kC4#s& z*MQ&qAr>2>uuRSy%cLJ-FTqju;wU)(^gj;96&oa(8wkhZDfD0N-nQi3aayavKnT>% zKMyMkviNn=_4FCfY|w{6mcE!bR{tc$>pa~K%EJeuvP8*KJB1fC>lH`H1rF3M$X84# zce^6rCA`MjKCMiCdQz7HpLP@wMvnj!9vlP&Rl6=U2au{y)F??ZQ8b zX@v`+oOl&ssO&!HnFY%4CCYQ2#xy62fw%Kfm2=!>R&AGoaS&7)hh@faeLzbA;0fU1lA01%JJwkF@ib2SQvzu_?5C z?xT=zx8WU6xo+k8>*1yvSKk{;-4?BQ`kOY{olspX3be(#Q$8rbzYZOJw)i5V4l0YjP!6rx8j{S5tv4< zpyOw+2uH^YwDf?#LoX}j76p{M-I)EQvM9i=OPD?3VK;2RK+HC>xa%+1U-p_6y$o9m zOZ{qiRny8<+IZhix{NM#)7HxtPs$UW^2`KgeJnD^opQ5ky#lv(6~P^Jo7;ZqhF->L zgFa`w!oXtc#)JxsvFv%sj0!7m_G>?n{#jax$;71lW^N(7UNRH&shNqXpcD}OKR`ua z<>#sz1JgyllIvc2WVAXB%0~XMdyEEVpSXFp(YTay3q7}R zpYWPBHQ!se;O_jhnnZQx+|x5FNwX&5pPLsP{n{NkX9u*62~>=$(%9j}462Yb*S%hL zV!}h;x+2b1ri*m4TAv1`PO(+AEa*~1UBnmu?NKeV3&JM4j$Z4glh;KofAc>Y<+ra@6Zr(n3eiKv>qmH9^xbhQ zooDmP4d&Y3@0MXa=2vd*3+l8X+3X+Yt#dJXD+LdK`w@v@x1Dz4C8+ z72OV6MK`$HP>U3sg>>=75{_w4R3u4}@nwB6x%IyU6_VBG0zlwQJo2f?PrkI6U3v#{-h5Oiu z)(!L=DXJK^RaE|hws+UPTl!9>x}TdQIm!h_7CSfGvNm~)ZmNCbyUZr<^Uwb@*D@&g zQbC3x^wtBhwv8^4uP9!ka3RO7KNMZ%@aEzf>#LJ?mb?1+gjTZJtkb) z$J8*R=Uh(Apryfa+WZfLap?$Z3+N^-pl*ir2ra&!`2cBLWS{CY-*za9J1wr_Y!&VG zGTwuPJsY|DSs5O9$F3>Onzny^xZuiy#INrAe&Rcc@7(!??$;&u+5J90b@qVA^P`Ua z3^P2e`BVR1e@bJcTWeI=0d0gJEmGd#A%0V}#{vuJ&B5&G1tw~{M1;g4}5Y(J=S&mX5?fk1#_iz&l##FX&H_gil zI7#1{m7+@X>Q&|j-t;P;S~vTo0!b1pIYpkR$*z-?QMsz~ymV0+mFJf(N~70>9}aKw zt|!;Mn!_6HPL4PVNHaMu_QA4g6jqR?Dt`CJ+ut#5h*tgK@I7+smDv#8G20NeQHqNc zIZs9B34j1qxK?f~(9-Uf6^0GSu!%KSZFE=CLRZV7!xb%OA~Einpmb`;twGMz;}DF2 znypMV5;$y(D3F{atNjw>H^s%sWrzeeIrP=gJzi&m@dET#UkyzF7P|`g?U@ekdA=ChQ`Ogq=FB`t1%N7VZqiCK^y94p|PC84S<#kQILygiR z@xFRVny8tRAVK)nDp zFSzur+pf=*$*^f-blydrsoKZLlZEx2hd}B*L7t-S1Zw4Cx|qhjxScoSFL?8Ki47`7 z7|O1f7yLEnpFB-w=KEF2=gDSvX2yxjWcHYunG8yiMv-JHdaqX>sh3?5Z;Y&dZ@IEd ze$IOh8T6}Bw|Vc7bb*3eH;oR}J@-9cJ0t~AECGsX0YzTvYMmX+UYMqk4N8WsDon%K z?ffrGwNouCFWJ}*Cti#g>NHU_z_=sCY*GPLKp}BQtE5f%|CunK!fM-!E#Ll;$#{J4 z6!jkY%!%23?wXl{>y+XuMcSxnW7C`V2KS0MjdumwhxGJ4aWioDq^J!P{*{seSrg|h zIIo+82g5KAP#gyR1?Ywx3OFqukmb+=vJ_RnhdvBx7oi5Ph9L8DvRA*nGoYQmE6@7W592dcoI`4i z8RK@WI^de{y7Zbj-QfD*)mBN7&svW>QFXv$Ps9)U)yQpi=&c8DrKmpnrXg=D?U%#0 z3x5}8Gw^XlDZMZe=z`YEyqfzyY+kwgx?GTzYzGjn*7 zQULqeF(8>lRf6|5v6&RxyJ@4PZ@ zU34XVP2LE4v?;1foHep^bqz1aZPf&3vu0^dERP+GCd*%E^YK4_<6Dx+CN%l|zI-A1 zWCAHMn}(f~0xDc~P|vW53FWc)XZs;7I0Gi zF3v5H4Y(K^>&oE`_C?gps-*iQT4;4tJ8+9L#&>M$Zdsa~?+Z#|2v71*<@8t}*JUhF5`imh zkmEMsVtyW<$MaPjYgcQyU9|D)PLF55Db?RLlQTx>Kfl1#IxcdnTl~?2t8jTIX;k12ZkcOtP_;CkU+6_icZz&nlC;S(CC^cz!ZEbKqpmH z<#T+VZ4++O+kgJ1vEbzS(d>7I6;57%{Daec%SDTg_wB?1B8IiMA$n<0@7J#xvT8M< zI%Th<+^tg>%SoHs<-LnW?wWXzQ{Um4sP3nGl>;s%bMRlxua&!%L>7ALWL5UBKOR2) z&I+s@ZD%_ziRquO?s4K+5Cd6#kL+GZ0#t$NICYS8Iz1EWzQA>j%npU!I*Rn=(03tP zgpA`3wI^eN!0=}s`I*_a)W0^hWvo!yg)0zOgP z=fwMJjnCx(S^aILL9@enR~C)xN)>bxM4zE1EZ+C3*J6-5OycaG^hBIR@Au3G`qt5w zpNz$HjO?Zu?idxkm~msJs^|6f7pvr!#mLVa8YhnXFq9k{^#KyN<*ICGVn9sfjW;EL zarSJ1$*AUej1MuxgnNGb*W69!rR=NkpCM~snMh5pnK9Z%DL$cKexeV*eMS7&>++*W z(#or&dm}y?7oO*>6a1l zzN_VHpcZ{^&!m8P+4ef9>~&9`9-lHq`{_St3>jBq2fart8P?Q^f`RFkvP>cDmycRpp1 zA4sK(16nj)kn}_abI>84UJg0~CkaIH;ZurgouY_?#J=~5-7JXJiMA0#j^u`E-Y@SqTb0r&1^BhwsOX2v8eS!&9nXj9X=4pmD@YW_`}RbX z%}G&x;s2!&2>}Z@S%M?N^@_C6R!Lb{pTx1~Fiy>=Fk*F9?R#?GmGAwFW#oj7HF08$ zFvL_ae1a7dy)vCVaoS)=hDQzA$1UP^h84MGct9XGJ2b}0u zWpIW*=e?OFI22$!e@Buf==a>Gs`FnlA=Mv7g-#Re4C5l8a)^P?zYHUa(TuM`o?VhoT9olYoDw>GHq(DyCIko7#DbB_E;=~ zmFakTYRq_fI+A@qu)+)VVa&1BV__9{;;=1)m*@AGQ7}}&P0N~gV>*HaYJk0qCTnj(C2u45g7C2?R$dOe^hx?X)_T;!o%kI9pjDBVZV#94A(Klxlo!FzdR@Bt(%oEjt#B> zE%FlhzO;KZakJ?b&8nzt!g|$x)f&y0U%_f@)X<3u$`fOSQ6(K8G3v%a8~-n|eZ|AR zX*T|b(cQ+*OK{%G%bOpbw?vB5G6FC_xSQ82StecLu_@wG;1=&1uH~_jd9&Gb`1uso z7Vj(?DF+wY!AKcYLn z`d1(YhK$jWOpf*%)VLV0R?ugJCuBOsBS)v2M~|G*88us`SG(X)d|>ycAePe zWe`~~GUsgZXi;dd_#ydKygG}{6Rn@*=#|mRWZ8-rD~~$#l%K5gYnCc~PP@u5$Op!X zE(tFg5;#x?xea<2Gu2pc?bxNm>a5r{%T^zC=t*X)U@g%0vy2N9<)$sn){TJ|$QCCK z&h9hYlw?zi42q;t(M3=-y_mW#Z-=U+7ENM6r~gM8+^uh7<*czZY5Sx@>R6z40hOjn zr^)GmSxg=GyCB@>lT001&??u<9E%>?fHKTue)iFMfI=Ov?l@stj>1q;XiT}@4QiC# znf!?YJID6C;n-Kq@SbdW^9+wV^fbG#sMGFRgga|>mgSF5yPhxEe;{w6GwCHHPn0X~q#fAVXa+Z1VEfEQ;3u=k9d^JDw%aOq zdReO3u<=2iI5@_jbCWQ&O!@A_iQZ$n9)H z*~82nbkC0Li@5UdM||bqO)g38x4!TtIwy8XmYB(t_EHL9{P+TRyiL-eNOiDPv_t*z z74Zd+F&#&q!c$#jx8g;2wu=_5|R#!(|n_G?puM z$`q({y5d`^hRm~0eqITQY3wqLiAqu7zNa`MHn?k25=xdLO9OT+UL)OJTKv%EdRVp3 z2RK7cd6XzCB^~^3X_BOTZebYQoF{7Fny5*TM&VP=wGbdr;8$>;{kcYlG{Ze_+DTzwP-=KWTDe{CsJKpKFxjGDTXcXkd}B=LYvY=d!qLPK@Ay z*uXu6fjCStW3aA@?t{8qZAVBfCpKyqRJ?Ta(!B8QzL0A`Tdeb`^+~16<_x+daj<_i z85+`XoH!m{*UQRh7f005Sde*4ls>hS-pvJVWEjx+3=$%tVeBJl-!Krk!S|uEoqs$i z7kD_44lB*I!b5w<)7Xe@LoU%F7_-4hjRZq>7+aRdrQA#y`-}G6AIT~w##pu)#y+7G z2^84`OQULCpicfcu*gk&M~Q8X$zC|gMQ-4LHw2cu8=30uVOK_a_onOYpkd>ncDRid zG+ujfVCF252^#7AE!Rk_wQeIPP7D;7!6k)KY@^60NC4dBS`oZ{5;mZAf>CHuVAn{B zdN(%>C_DPO#Ss`_$e~f>y(zd^HBP``2O2aKc0hy;LokD;$K%7VS<)#w?QWfcAQ6p4 zh6g&;+o>Ko6l~{mlB7w}W>qIWuBA?fiLe<0HcXp6Z()XjwE%FkqU$R^Gue%w9oqgL zx#+}8N}rjHxJ@ZKD58U^6h(?hr$;LNsj6ML-w*2|a_Dwo3v1D2s`qo6_$xghft%8z z$>r>twAj^n4cX&bG|ikflbb`ZyvaBQlDhQEKM%P{+W9&Kp6}pe{Vt9KJTKELx$(Z` zky!$~tAgL_zLpcq`8=e;1G*b!OI&-WWO%H1eGIxRux*zdhK>7iPJ_Bprl(W=*9W#} zwu%nKpE93O#?#&XgsC6-R?@MfAVW?Lx5W{Sj7(LJ(K*|yLIsooK zR2wT=$*U0RRQ2LxycmHEdtJLun{ALWEGc&Vy(10w^IshPH_N4jO?ttJfx)mMHmfil zg~5;~A@@QmfT#td$6AckU@6328XL({RAn@H#^YsMSfIn`b&NEvcDnwviW{3OlTQp; zn@>1p0i85zfE0P2os^9*h#vI^LMHV5Anr@0wMfvGy=H|Vl`()t;8{=%|;<1g^`?DVn?!!Vn0Yr77o z^+7l#$rL78|GSdpJMnhtxY-JFh*BJ&$bKq%H8inykUsK|)F|`Ax?mf0cjBHo zn9*s`G>D4ndgT`HOxPxEX-vGUN?%d@u>x&QGQFknb9hC2&Vx^z4ta7g*P4 zmn^~l+%YSZy}rBs-5-sG9o30lpAs{R_LP~t1A@QR0R=txfuE1Q zIB``LLv=9Ho*q@Mjx2YJ=jpXQe ziz3rl&5}GdVWJiZJc^a~T@&QU6SX_S$+$f8b}nS`0C!!dBfO z&=~q&!_QHpWY`#v_^5q%ifod}u(ZxR{{e}gK=zt_BQq%lG>xWE(I-ivAyRO_6{$g+ zRR%(olLnp?J=9EBPl0k-tcJ!&8PJ)mpA-`XXW}@Gbn?_?(u9c)mG2$(`!#+Q9r0=n zV=|j`D{)=tUEuE*-g^JlbWjrHSPz8~A&?A;<26Y;rCBucg52>enhs4%iB$PX}g3^Iy6vms>h0P8%slW_VaJcGR_E#h3qB8{s>iC@`~u_kgQiH*3wCho|3} z{7};E*yx(|>9q|()}M9cS^8q;Ov}Wx)0z>6jChjdxBz8_uqQT4ecS!)9LNE2J{3Z1 zSi~_7=6GPGSqXB!tH@##IozNN%JrUBKNg$xiSk4Vko(C}qg1eQBx6v%-@Snq^_vZI zWg8UCjR(__4ZqIjTUZ&B*HYdg@qs2kc5mmNB_zj*6VJeyJgoh!lu|(Kbq^I?9^4_r zW}g~PjNsM`9mn!GLB7p5F|ZN1&X(}fqqeK^{4%(DI-7fobkm7nyCUx4?(@MGplt$t z@=(AYE{=kQEiIZn?#Z{ZiG6oDc0`WNsIcWmR>+|eo_t63mI*oGAAZ(O$|pcC*YF8C zMJYhbrH+cm<-eMrA9;sY>j@EVg9EUEdrC6kf=A7s8%1e~UeSQd*-22%0!Jw`AB+0K z2V4>Z@5}M{(Y!`}DG0kfj_UMIQ6)~V;XLu%N#B$8$YI*V59k}hWUuV7+%O&Ig8W#- zd6>2{a~k5dr&UrV1?1ei)jVyXTPg_1uMb@pwLPFw z2wkvgp@nXl5TsijX*k4#jL_q*Sea0wO!db~hzi3KcJ;m6Q*4JU6zpL69|y^3|mAJD8(jD9B9R>qToJm@rx~2$Y+4)3c|h<)t7PEKV^cv* zB1OH^;}QqujPP`I;0lj6;eOI619f}ZW`B%wADL~)5ZD!4Hp1rVQt_e>?11s&yS?Vj z|ICt(+iCaS3a4%#~(HFj- z4jIL*oW%l+&a787tFA9-o1dx96J*n!(ucA|4T1e~)UMIV`=+dO)2WQXn@;%!N#B$_ zK_AE;mB@A|9GEa*k1R7|VlN;)H;zwURp2xeR_^~)x`OOrhZQF-SEw|@%3ewVf_ep1 z^wH=m;%3d#*{kJ?1!<}c+$+#q+oFjPoZ|FFELUt5LAD^`Fh3`vh}RKP4GlBfRm;6w zG*vzY5yhf)K_zp#Bhw(Wut$C0aDNhKpDf!qM{-bY*OkXOPh$j)2|>@z`It|9aPyno zvGgiBF=!Zi6|=c3!cVAmbo|srb=~|X0L%&Xas?ipSn%Ous>v}0D~uDx0vL>AIO4+l znJU^+3EF8L6NnTJ+m=67eb&j3%uWX)%q(>(y-txQdhC9aQ|wtAkfqM@TE=N}%O<;o z+uUw)ssoy&2j561+oI~B8Xt=%mPu1QR#B^{CMniXCV*lqwn8_$Zi~7J8m+OsEOjH@ zuEIu;(>1j>~Hk{~+tItX*JJ2<60fg`w;ctKV=} zg9}iSWk{Ruoy})i|eDf=Pom@pf-)Jwx)r2IU6dWzrVSdPN*> zg(gXJ2eL(V5v$zvlEi6goGWr@*mdjRRtj=wnA*ZXpYy8t6uBC9+C2;EB6_D3`}C1i zza&Uu7@v=GEqpWGBUXgJ?SwknGfQ2>>7v^`MhhtR2w^fc4ChA)9##g2^4TGHY)NJK zd@G0($N3rP4P!+{*3~v4u=~i%ryzr=Nt3gkf6pyL4yyFndbrNBiC^q)%=AR)oGO7<)I99c4&`$ zp;67ENs7gMcY$(zy#i<)u6p$V@$dG4_P1gjocS07fJSMOYMh63YZGkpD2_T0mOnA@1oZ24Kp<@;51G6MWJs*t$UhjF zBhiO_9$sT;_PPTy1rNx6zlDnh_u=t{*frP;#NDS!Ea%8XXfjqE@_Q^srZD69u_DxE ztq5MHL^gR3d#dqL=dpxfS$p^~nJdzZIdxeLmT`$BH9*OwT>Y%qRN ztzEEf4xZh`{B^cgALq0g?y9@}buT-ODhBED1Px#iwf;iu>jeR<$ z0Qr_}RCKH4u)3S*Wu4L@K@r$M=&sk(dVd2=v!qq>sj4Yh+bAm*A{W)5OOv!!g4_4; zfEGJMJ+*IRM$AY9eDA%QotA2iY+M&7-j6eAJH`kyCKm&9N1f;a5M!PvW#XFYjk0CZ zI(p@VY%z9E@0y-5dDR5SS@G+D%ez>VG5NX|_Og}H)q$C6$H|u)=W3KlLCB0W*<-r? zKi#@x|5zkt=)mp;84K+hA#RCqH;>F&C_3T(LRN3P`tM}D6BmW& zn?X03QY2AiD;1q8YJ?~W>Nm8guoChSIVVY&*d$F+^$RWuS|CY~N>}sa-dxlp+Xy=H z<75=r+XTn16hp{m2i)Rsy5#-JgpbEx?2?enPK=MoX88DmQrw})O)C04s5gFeU0!Vv zZ_J_Y&Hqwxo(3Loarck5irV=fUzdM;Pu5MJB)4WE5m|wx$@}aa6et-8Up}GxM_s&{?x{y>#}7j|d-{_V;Hb9_ z9gy+wSyhzqZnCr`>kCzb`x8X{UoIiWe>_sDK6#0YMZ6 zZ>U_o-$7APyb%fDI5-H33?h8blZ25(qB$gCqx*G!m2(f|{h#N3-skfFFZDXY$qh&9 z&s#xhuQ#h|y!(lk205kB2H}0kC#;IB@NZ-wg6CkF@^Mb74tHhQscBLB&v*XY%iagq zChH>@uzfIN&)l-9jr?8y!!A#FDLge2c*G0(A~8##&XBc|n%6$q9aI=TBvo%F7-~de zl0||0LQ+CM52_DN;+>3EmqIZn^fGa70Lqrh^mpy^RJW1~!S~CSdG7>gzu(B|d za^J6xH)pKdV={Am#R_N$nQ!qd;VwiqkR)` z*ab=BniQEqRj)66UCaM_tFjofJkMHdoBrUhZnk+uzIu{%G~|;8NysC? zrJ!?O9eO4NYh?!9mq*o*4qBZ6vPVZ~}IA(`LiEL%!?(SO>R4JjLt#<#ct7M(k3k4dE^?}Ep!ns7sV%4jD&7Q10= z1Z&Uy93S{Bo6N1twe_^o3ibU+eu^YGu@%ZSu|lgUCY2&9skrJ;Od3^1m&xwCsaJ3? z>xnxPuc>7<&(!!v(T2M$17M8Q=j6`z@BPM>l0E00bfH#!bcFINoxZz1m2-;(X|63o z&?MFuc}A5|C&*6kib!NID-J+b@KUOXz7FDChkPaqTH~-MHh{$XTigC`?Y8%gRwFR4 zwU0E9CD-W0MJ|0NFzBQh9YwV0&tS#Ma8P{Ae)>VAx}AFw*3h@2bLjY(;UN51&!?uN zf#43SsoG@ce3}%ngZdjc0Xm0J{X7 zLT)%M5FVvLE-Om{uGMhR1Gwc`7it`UdS(i;!+~dzX4yV%`rv0v%vV&eNV?L(!mtct zbj_*+=GJTZoK}v8n@g_;ii-qL2~3xyPp{_aBD#T^qFlObMuTuQsfgU*1yMU;KUDcV z;10?b@z#n5XFNOd&wSsCejmJ4n%|D zJ4-a)xu8kncw`s`kTpJR9CvGP_ZQCii87^fd{mBTKjK3chf zl?64yhF$CxvbFXbwSJ_neWHD?v+Q(?u_&6*ppJPlm|*hGMw@>-CfD-o#oa=#F6RwnY$?QK#( zgFuM@)&7QEniS`SNP_clHtyTq#8{88=R2?V!PV2JY<_E7uz8J))7laX#XCA0Rj$;S zk|=~YuAY?RsH;xgLEi^uBBY$s!eJQv5u&40O8WuG3~`eIXeD~~Kf4

    zY}}~UfBRhb>mC4U6&wT^f~F8GZ0(x%f%bo4m+K%RMlSL>5HGwCG5Q--zkb`c)rQwz z*=?6iVy9!MMtlTV2sZjRs&UQG)7>CfyIr++ic?pFffHi?{i}@%t_$|>ch#9MNE(Q3 zApH#ig`KdwS) zhmnFy#>OcxtRr^f)xe%TKjI>E{i5g?w-j;pqzOTG)k4OjQolOzH?o)!Gl34- zy#Xu83)ASL1LLOV{BV*59o;`xE+U)xp~H>kC2FkDv6EuAQ=|lTvXY~wJ?si$vg8Bb zZJ^T0B}FzUTPN+E1aFGu8KN`VHs4jgi@f_hP$L%3=L8?6E2ibs=#ik~u`Di2SQ*&o zu_xf3V16hby(cih*v>{1U(gJ|kM&z8gXeVanBNVjjnj=yik&vj24LmQ3~mv(D@*;r zoUHI_Rh#R$A4V_ti}_tcg}CgH?SE6Aq7_|M$fDpXr7C1n&_N|9nJF<0~m zl4egdpcwwIS7ud!fYUO~B5T6S+*(&fI~vsY^3mfgJ~sVZCAu_Q?fcvA^v)_fGs&+de2hM#PPa zKkT$~wfdLQJ;3Ih8r3U29GfmG4k(M;B`*-xOLxm$fr*z+r=x&*=^!pR`O(quPP27A z?5x?iE5xFIh#%>+4a&;|7wB}+vUq*i^@&G9VHsa5*cS$AEl6nO%0?(v>*#YG@cVPY zQ+#nq`#;3jiVQF3m#2UCC(BABQ~W57bc`XXqX*VnC5>jEP^6bitWhi$wOLC2b?T?8 z_dT;13`#eu(?$9TrF22CPJK(B>&11@SEP@kmooSPeZ6Lq>xl_|{10mY%-sSQvupuU ziktG3gg%c3WgC5#&J}esDJ*sr+*0m`wE)j$F$cra$pX#CVO2ElC3dSiXs%ItFkA=A zz#|%yDaZX*0y_6<&c&y&sGiv#-K&5mgo^O>3gqcan$e&vk|jZl^)2NcBzeR*D6o{N zu7%)WfOEim7PG=Pht_-R=~L<}&^v`+Sq=@mAT2Ni2o6A_>)!H$p!@DSpDzAUi>>$j zx@CK@w+?+;Q+-eiz1(NS*B_IGasa6>TCx#1!V9UR3%(>_YGv zQXaM^?2+(zoKAgKdR(=|r&Z7?>=o)*SD?xP%)Hu!xd21FFhkw0_`{sfmiN14%tg}U zZAf?!`!J}|t3X~dwN~){j4hu}|NIJ z<$WFq$pTHQx{zHR1ba6k~Ykwbg;v)UD(d)+xHjC5ygIn@6j zS0v47ROi#xun(^lBummsiRL`L!>a+Tj8T!VtqjB|*H2iW+%9YeJBI1Q-C>Q8j6R@6 zyXFd~*fJ}_oBp9qts{5w@{LEwluVTSTg*-4&-ctB>&Ac>$w2R+kYb^Ue+!k^poCp4 zkbiR3(o!i#XG|aJ#e1Qe9WDGW^dI7d;Vh2aM2z#%}W3hUmP>z<|2|ql$ zPhogd;<}9GcLmhG(6{CAM8~u>5NyUn z)lwip5TyBToBULj5!XiNI+_1~5Bja0krzHW^{4P}-%hz@TdDtADFfQ6$Iqwdv3LDD zg-MZBq5{t}*^_|Aw|0}e{v&Gg%roWK)H`)n9U9EO;j8`X@2`Bec0lOEw$nF-?NVLv zsmssVm^16cytU~FPIPq4mboMyBGcO zYoG2O4%dbIx?n&%J=0NCk4&XS9~0|^6=i2czm^KjKzydFuu!?d1FY9)dSC<*i?!vrxHy}5)I;d z)pdDCB$pO^NYu)-gH(aD9+Bh!5CwJL)6#eZ!2>u0M81i84j`d}Ov zi+R!|`93b=^?W(h96|QWM{v#AfBsG5uWa|hd^S7oyNx4uw)34#`sXhTRx55!Hzc&E zlO)$eI)Sgi5K-{?dCL2wUYX@pIb2_Y-&8pQjea+S-^jzedcwqMqimt!w&e+SCVM48 z$*ya1lb=fil6e>E;8t<%P`*64_TpVJT+5_7Pnb zRW!L)ln%QcBMg+q992VosA@07(wyys^8i`@ievAUCyZWx`PS^CM-`UMiu9iJ_axno z3l*TnV*o{CF2zD;{d!bTLh-Ba*ivb25LUsVq?RuD@}xF#HgFn2Km+zwm7*$XV?ye` z-U9L4ZGPn=v-AvQ(qBGZ&N<7=vQS;8&iqVifynW1eDf+<>BbP*VFi)R6bn1=4OAj# z5}QCweO)-0r#+x;r`M5kECM(}Ulrin&RCOv1>}mK`gM3gWrl2Ti$btei8|aTjXOhN0dY7AW8#-E*A|1y(z(S}-fa>FLMbox+`wuL9|iL1u3ea`d=ec$rV|1KAJaFML~W z0h#LWUuq<2ZVWPzcpR`;l!*^Q>p3J+mgF-BJ;}6>prI9UeY&UAH z>_$1of=jcFN-PxYj7%1_h|dQlOKMa#G^!id1*JlQX1P}bIJ|d6b-vfd7yLic?$uU$ zol=)TJ=J+>jjD4}u>jb~VUa;6*?jk5!BJRfx?-!MHhJ%#fY zWek@`Sx}?hviHBp6AzMYH9091yNn`>sYDZl7q0uDHGu2%h#L^oqjKc2JN=nDFd(nT z?!GlV%acbhwax4C3D~>uzsz2G65L?3ubRa)D6!T^r@kAy$lLh_%raB|5A-hyP@-6aK!A9R)uNu14FI{9vm|ZJc z?(^qD>ONge{kdLQp0G@@OI4#<7_rIQC|;}Ri8dtclzpmNujrhStGy?6%?vpDm|wD- zzw#G8lnZz*Ebf~x{TG(IYMbukv#_}Dzsz1Hc)n=1PQ5;~LR>0cNzww=hvw0GL9Y6s zyoNq6oi8e3wvY<~*I;>Yobf{V;6u3N=!Mbg_g{C7=~kvN1gXhe#4W;d=}}0oU{wf* zoJ;0#2Zml7qEdXC6lb7SYlNnq*NmGCn@cA6g+Y1Qv{8nWO8J;DnY zRNYquy?z#4G`)4xB#KG0g=)?iiL=sj!MKdot{RyK2vu@^|%zDyU}=H$2bmS z7drI=;ZYS5lCRe6hImG&)(}=Qt3cQptJfM5w)*Ug-xYF%J}~Plc-?q(gw~r; zj$0NBQk;woyTjRs%=|CXa15?C0GT)QKcryMkJ1?BvEq@nybP%?^oQ>^D5AUhlWeH!TV?d|WfLMDigJkKvWM zf_m>-2)Sm&J_6atULjZMwQ1rNX^HT1xSmdufHL59O$}Wy-VD9tRnl&zUNt|IyBf4d zrl*f<^R!0-jszq{R*yARp}-%|$)J7$R;FZ)<=UdFq-nuz^nTzaSnrb>wR>i@KX*%6 zD8R;z%LH>lxq`v6LBlXb&uo$t-n@V>`?wmUbbHhfe)!J)KUlE#$A(`-kdNIMYs;({ zGaggyrxfX?5>b=BbyEH$%oE@7#VoI-;wX=-jMURc#c{tX`hD+A*><3P!*p_^ve&y& znJYuR72w5~)a%>lflaRnS8RMXCkp|D-ESEK?<%oyE?0C@I6t(}gbIDcV&6{TO<@)N zNKh)}_I(+nbMTd!!dy+gv`evWT#l?b?t%8Iwq({r@{yoa$aPFOr_7;u1ox2p%4+{4 zstWSUe}VgBpx5V$bbnl`;HWB#Nv61F(nC_%ecCQzrS?$_-kU30Hm;K0Cc7*+L~=EF zdplh+>&+m{82$L-Ka8GZfnRyhX%%U6WB4t!g5N`m?V`v%D)Az;=CrBy#TycqGJCuZ z%Ilz*vMA`@SVP=F`N!fUNu_9m-!1vAn3be8s1>L&4oqJ*y)mJM-W`&yUJLKfB|xGG z>ZnQrx)*ig`g0xMQ8jv{=*u>~AV!jV)JRkcS6%eM%frAjJ z(!pqkgl4i|^MR-wHuX9+z8Vs0#5?`Xe_M-mcgW3z#bfY$&(fK}hJ+sJ4P}=~Pajg{ zg`9rNkZ?FWC$?Uk!mb*7F#He@HylzOl2=)U7~RD1nQi9)^T*5ZQ3t;BNs6xpdH+9!BH&Th1NiZ{X0+A2UV z(Vu$^oT(fA`#g++JtiYiCB$acK97z5xg%RA&IcJzo%&Y{_Km6ge*Ptk{n$21aF5iu z@yd18$|BTLEU-Bp$D$~7mCl6sGD*`Kz&|m@w85IOBz&pwx%isU8~%E_SdtN2Md$ce zk4*xOUC=QVr!gD~4IzOO*2Cp#n@OPn%jOJ%b08>>#m6&cs8 zM7=Vo-HW`*a6J%tY*H)}Y!bFB%LJuw9}Mduse*dXD*7O)3Fo~xVU=#`TLr`nGT*?(mKu`u$xpBdA{n9gV(siFq?YXW6IHYoAXO+p03N5v2<5hBh<1 z$uS^5=#^ZSWQZPvbBjg17>(>9rSuJYo#!p(*{F-qb7Gv@9+neY8}Wun4CCD&`)u+~6+HFW%dCK_?#b`T@l&nng6M`nzH1j1NOD3uRJq~M zIw?j@yL`Gl_8RaZU6J0Ns?(as)zg`RbYZdLgyaJWTEhqQ5qiF4JA5N1TZ)3N`d50D zO7kUW>5RA*AWbh6oFHkzTYO4{YbFmcb-chg94qF~z0VQtd8cp5XRC?2mANBY6~z^Y zCC%tm?Sk-tiScK^R6Tzjii-rX!%?cF%N(yJ<}5~!N-*kk61AXT9E0&4F-yt0Mv^)3wQ^)|SK+}mna9D~V= zKYOb|T=1c!sBWUzvTVJ5Va*3*8NX%Ajj3FqifBOcY(2%Up-85=&7eVC9+XdK#O?~p zVpdaKUp5E6&5Z@vSAy}z5ws=zz~KTE*cp=BvTUu&0*;&szx@Nr8cYhyjV(pF6$J7r z7SxP#Kvj|67r!xVfvB0(Xb<}?fo)CR%qQ}Vfz_c;J-Wa4&}WN}`HFW930?obXV(8U zBs2my=c=flX!CK`LW3U~tW1WTd34TU`#YsM@?%l9^}lZ0dT)nt2s7mn=g$o0ObycO z)Qb|%$WR?`#4FYvo`2^|tHYz%pZ`I!R-@xK|1%X~u^sRIrL}@=<+rf7aagX-%Bs9i zvAZd<6RZkkEseB6ZB&@ZtBlMMazJ}93_E5!C)EaVH-+0_KGCvLj&C15S z9F$%EQ~2*m={tU*UYULhb-XGqy6;N(YgQhhH*NuZ&;4LyZpDdmBD->BW~=O*r_9K z1Szw8Nx%0Y(kj|2E0_i~RhUb5kH`=`soTkTT|h-c(7x`Fif5=yFDv= zvRNHCarPdv4q6BDdKINiYzD;ctisedcFh;)+$Il zL_t=@_7Pnt!LJ0AZ z)Tx6WP3ltUvCo$*VUPf5J(RJa!ZmV2VLc2AVzm?DE-QF#p|}i+ zyh|lqgf8%=_-x-Eze_NkdBM#gsL%^lvT%0elqLSD!U{PSAmmOjizsqYoA)!*z`JF| zb>dgnf19BFs~2B(ak>s)qhC3FQL#2US=OWdNL~`X5=uXoDC;Ae;x|lR!fs|z$^Im1 ziM$oRAX(Nwb0d=#Tr5b2g6sw*_JnlOSHXfjav?w17|nzAoH_9wueFN$X2Ng&Xq&fr zHa_CUE0UdDjw$epl$KYeo!+q3{eKW|walr5IB_BZ92?3$PPQCv^p;utm#yLrEhO2E zi)IU~mJB_`!QN*h(B6tpf|ljAkUOkFb8XHp`N1%#41;B)X#Pz$L-D?(#OM5!VgbtZ z!WLd}&(m26l@~mF2f;ARiFm<)PWsqB7)IEnQ`f61fBWLjST{Brb~e7<~;`9G4siN0d7A4h+FxCV-xrV_A;bvJX9xk7eJTg6@U!l_5SuvKeG$PMvX5F^wr_cLftzuGGPNOgNw zPw;YRXiOCjcqB#~5#IBMBDA1x?G;iE^y@`HeT&+Kz+B6rZj`P@(ZL*+=g_U{57kwo zKE<|xRM{s{N5Y}OHr2D1`A}pG(2Jgg)F|4;x-;SHHCITI2)Q1%1$55StIH;rO>SZ` z`6Yn)Lui-*;(7QifbE~*WpyU1ist{t+XAbfT_p=hnj6Ea*a}u#DNauTKVm{vOtz%o zYpbks_7$d$y$GFs``9X>fj;ZG9n@a1129oCz$QnYP?pm<(Tky*BY8^sl<`=cMuidk zQ|q)(cwvNU{YA;|$0BU##(QHs((XOtQW@A6&3SHZ#640qb)W2LOp_|l_fUA3vPn@b zJpEqQwzsUc-HM$tx{UdiA-$ z6v2SU@u{s)baGtP6#tPpo!u_z^lxGNox*WAD!jeO7C8ZROy$(`wtA_2qAPBkO17h< z+6cY&4+%0po0{BnOnn;bgYJ91Jsx()YRVi1h1E^t@J9WMz~-5jrRTMlyd;wE#!Js} ztEJ}v#evxTZYse@?+Po50ZtB3dngC;ghO+5jcUEx+!}~7l<906bBtd4%DTwifw~W% z_H12bD`;D`im|=C%KKuNZpVzGm_?F2xegdD!!s25vJ9pMx@B&v)<#0b3D!>E^Thkd zta@shQb*zNBC{~;bEklX*P1jTX#1bP{v``|=2wW@Nv#`$=eiYm8Yu2GMNU!)>CqSI zI`IwPYWgBw2@Y`%oh>nnlZ02o*J`un10Jnn^Ly+lE!XarH^x6=c0ot!#W1Ks3Ohgd zBE6c_G6iJ7j41wVTcX%|;c>lTd0KDhu>!{B8(90kgFVT$^nSHeRAESS=)S$T+T;g@1^<5lXQ z6#@$>4mRrsDxpV;97K9`I{Q!sJ8}cPDY#veqp4*KbS;=J)N?kgSRM4dKn6YP!Yb&D zxD?<9N@r8a_OM2Es~A;EkWL{7I^;(K9S3H0?C5gfOGa_Yeqrl5+a4dc%~sp#+A-2K z^hTy4%sh#UBt2eFAj}W3-EKCSNtTuQwz1Ih4$2`en%fZ`JJ0DJVFcHl;$=)G+)G*0 z@okGSVONKKjhx|UOx$?W)M;f*u2b9=+aG5$H>jDupBP=Pm;?Ij0PQ}IPrn1i~ zg}oM%8PN*w%KggY@od_Zvap*{Sma3JG*pM!9k`Vo4LtoSo~u{giP#pjS^--nIy26E zlox!Hu2J2TW`#Dz7Yk~j#=mOn(5FYq&>fjUt)zzDEl-r3^Qu?f7uL{yiae$;WPMP( zYMbJ8*x9fvGq-AzWqn>5igJ}>nC6McXBRI1XB+}BeduEFEDOTks*0~A1#avDov^}8 zCB*?b`Ccla=!@6of4j-P^8hx(TghWN5|Y;oE=iFOZ++0dxPHJ*s!weoY_#M5TFBK) z|FG`0t(w8v{cL6syl<1=h+g&DdF4e_KfBpyOGGZ6tGPW3Z)u8eQ*~?e1e?fV!LHcF z47SO?Pm*PhP5JYmze@n|{I7KAQ(gv&qKf1*Y>T)*BYSc?<=wf_S#)l6jlvkP*YBSA zz88=p-t_I6jny^IfQ3VIG<1=p7#g~Myk5VRv2*3;EJmgItG9E>?vYwk-PqGTYh`E- zQ(P@Ysv&hrcf+>3L75nlFW>9es5Wu>A}L~*GDC4!iGFW7``)x|;OTU;d9Y{qNB693`Oxu0|Dj{v*mEJ=tPslArX4&x>9lQiFrO96jdSOAYKc#& zlVxdI%$M(Bu6!Q(TTL9kO%P(lSYy6uyxCd!L4)rEkN)~8qO#avi@oXnv2q#NHi^_& z{ZPv&4$_AORKnt!onE^=m--E;milcMTu^j+U14s-=uSXN*)Z)ESFT;^r&~NTmCJ>| zMwYBZddE+%>4P|1uI3O>BKLd2$K|S9P%@cA@6g~%kqXD@?15>{3eY}%3jCh7)o;uC z&C?;Ba%A7`k5Jg+zP`doSKj!}H!Pre{k`rVkQ4kCDL1w!9ah-COmP<}(nuv>*L<@g zQIaAm(df#jUy8lD-HU%lzy6CL z#kL#rrK$n7)ART^?$qQ{ld(NkuU_hRmhP2iMnCjMa_16J`}m054}16y&;GD`#^cbw zh5YhwW~{d@xUo~$a$Z?4zDMfCy+B@}S3eXK#teAuh{1pF0uJk)^m#FIUN4R)^F1Dx zMb|4!LROFp7ZztfeQ*vx_TfLHr?; zMcAi`!noh=Bh8BCk~^rdlSN-4b#v3%k{Q6bum|J;4ul#Yh71|>RFEJ%tx40?({*80 z!hBFKfQ4dL^rqkjpK2gzzXMFWYo_R$)LlST2}@5aB+2`IoZD76Du@q<`|x8Y0^UBv zi*@HpZMR2mn~t)x;VBMiV|4FQ(V|EPjq<$r}|pp=Ur6u9?k}-H;c6 zG;y}9jm?!W^TInvYi{@2ErjS_a# znkmZ8jJnY!W?GjYVA zM=}_lCN(k^K8bn|RTq}x-QW#eV_T=UiZ65B>|Vb!e!y>dKy#Sj;Q@_qqvB3fqNG!r z$Mg|>P>~e7`paaf>Qx=NG`u`UcQi0rn9e4XjuJ0l+d`2HDq*c?gCH9uu0MqClZDhxwsrmk)k*r0aQP(jC#U^yfdbnyua%QmhEz^V znq14Qm^3cn9R+sBmWFl^#Qd-u^MeD^yetr@`104zlGXf@18%$w0R!eRcQ}*cHc~KW z3E0VCYDBD1Y;zDSY?2kJp|3rCv)M_qAuL1a(jqYp9|I0SK@3wmyJ9&r2GfHJ`D(qgY_Z*Z}MOILk>7x)2*+pLt zPn^9UlsFqf#%JsF98Ed1Y7WW}=S$ATj0apC$HGhf zX7hpQ8`Db#cjnzyW_Ux(8Cy-?fxh;;P^9M4X&KdgSf*xlPuj7&c+H3VuGDq1PR6!v z)@_Z6okm!!Rcclocol+TX#^?w55z#?2KTh%#r5$iWE@88r+19c0Y9|9nbqrKn;hg* z1?RqSx0g1|2+g8Tl4|I<2SxK3&}T)3hia(4N@or92JdqFzg8XBIlX}>nVLP zR_vGqJZ(LcQ{UOg%BwQ}*w`U=)u75Id);`i+h~Q0qZIcZ1%`caT#eM`LcBO zM#wHnu>c#ND`Im@Yl$JWjoTl;e3F3%@dr4@s|@stNsa0TXdPQ2+3Qyhs;PMYnrQ4l zyeiGnl=`Ma2}?C?2*qXRSojA4P4VECdta63&!`skXtxBRjyRMuz<6Us)R;i?96XH` zQT9PI!s0?be0$;5?^%F!`G@6?NF%?Hnj4#=E-TozQCtf}t|2{0H(NQ|L<>@*ST4+^ zQ^|k|d(Ju`>731AVZTW)Je_S+*U%_*)vc|Vt5=sxub5iXpn|qh-KIv}@Iem)t=Awf zF$29wQVu<|(5|h;!YIi$u3LbL6PRH}8Z`86nicJ;R8L)XBp`W4k`K~M>`~o=zo0F! z-m@NLD!SPl@*~DqDhYul;22{KZs>v{G(4 zk@fo1x4Wo69JSnLzOGbuHejo)g6dIxEH&MZqm{}Af=e`a@{`+5RjCHZVv0EvD>L*b}dsCvq_UG{FAQX^=)*vZ^~1Ok^iR<8UbZPsN!eW z{JrLU!&W8QZMW$_nK#UFN~O3p6j^2QqfuP#tn`FRhwe}sEOKL9wuT-jpI65q!DmG0 zT+0g*6CbPJx#(lDLEov}a*?e2%q(@KRsh;gaS-^~0v;BFzR=})HFTSADV+iZ>y44k zpt9H*Rjyelxv0{s3(3I{-D>8M02CRM8AvTDs)Q&6>f#tBOuu88dX7DC4uqZ=W$kaj z{V3Fel4Gy0tsp!2p~QU??4-^LCA%pOU{eI51F*|FNp8>5tNWrZN$-X}VHbHXos|F0 z_V4ca$({MNU(0&q$!j0Hwro-jP`suJ%{(8~^nwg(fr06rwcyEDE5CT{^#z!KNM|>h zNFD0Jx&?K?kDz3DyI@aUXEAp*4hF|Gf4s5u%&fcS#LsK~Ve-pfDu+=GAmjxN-)2R(stMF7F!}*x z3fRcpqaF0K4f`m@(9OU+}!)*qAV zNdze2hFSE}6nBy$$A?-j+M^njZ7fEaOr3xjt->9rY1`QV6%R(m9iK`~7eVO=pd!?( zj{~bV_FtH#;8#t~g5@AfhQC$}h#|LEpBHXk(_~8`jC6Bka%3?n!#p5{rpi|_Gdl-apM{YJ6zFCKr+*;*f(tz13eS}{Y%HX+qRAlA56KgIfcN>@WuGk4uWK8eZe|HOnmFrb-~#FcqdSICvdy> zT_uv`6)-3d+9gkqE)gKZM>4bBdpxSPMh6jIvpuSd&z1j5Ya8h20}(eiJ9g3nS@g2V zcdBXKfoWa7f3Ak#y7?b|zxYu-Ynz)KgoMpqu78{U#^uJBopixeLxt6HNg`dctZ2&R zd3Pc*qmx8cLS$x2lz{&&KjPw0iq%XxIN27Mtp_f_VX>X}Defpm-lG!UxyjZ*4=QMia}D%^*psANkSoYgWP3Kn_s;4HUOuTo z*-D>MXM3gu?G>f7-KtVXuWpLR>LrW-B~KY(>%u<{U!q(wJ9Rb$haAg zU&H&rzA+i8%r10O;M+kKEENoXZ!gJqV=Ms^%P=gJP#iSs=VSfTc6ll6WKmOMdvI;w z8gBWtU6QQetpPV=10GqyEi+IDqE(C(J&o~8{Vw|LAx*0D@dF-%;Y}(~mk?eE9`L9V z4G3$4a-ebGsAfTj+!bD{Qy{||5HG%gA7p+u@!-S>7RYF}@B1Tp;z8WDf(f*!he2iy z#jT>qaw=habbVwljXPP?vvI{-zxb&;5HWLP5Eop-4-W?)%o_arQo{NWl+qFX5jMe>Ow}cc@}P_xgRmET+SF&l`ymu^4#=I}pR?N&>um8N^S*ee_gUm{ zSuWY-ns!)cSQ{6BYr>khe%i0I*pHL3(}kpK5=osn{AapoA~#5p$51)!qXh zXT!EcY}EjRFjEz9jQ(hLWk5gZ$hT{bgx{Tdd)8eqNXo?@gS1?=05_$}JqM-u=(l?t zW&rVN?3BX}|5pF*uM!8n&?QqBziO%fbD2HC;|uMPjP6 zc3v_#&AP>88~DD}k?piDQ++A|0{dn@DAXn?WsZ7Qi3;FngyjpF(P#Y3nk5B-cY#>S zMeE|&q3sH@1k17AoptlBjLn}#p_`PTB}`qoS;-N5W09dU%_ z2TV&8ir~~Ju+s?%0*_2 zW)O57c@+uHNDQJ{W`L~{>NSZH*n87wH1$wLw~1RGdRTCe#-Pl7pn0DeMGZJf)wxB zlXdIZg%WI#)V<3rl;q3s65R#hx2dMzL_&DH0XruR)0QZq)4xI4?|mcM)RN1+*&uG? zszf~j<$>7E+r+(D@3~#RCu~WCE-kW1xb&6c2v`A|R5jtvA#d)3ZpcdJlDYAc zXpZ@CI0~O0f9XF@eEtoKQ~aCr8Q&)FeP&urpv88WyL^t~>M3%HO6c;v5xqyYFAVk6 zmyzQ61F99I@RhCJ7roEpM{bOpMMxz4R@H&?DQ_so&}eC zEjjJEmTV4dn2#GrY&v-y*Xh4dQm<+y_vk*d#V1>|M3~8T1>*>1z5}ZDP*HJr+RE6c zs>3rbC$>ZB)Q$>Fm+KOVXv`N5T$vgx!|36}q5sQ21W_>ymsvkk%;Lm~C{Q z_B}uNj>WS(Viu&my1{>e?vdc8YQe2J<+25X9>`UpGoMe4Xk$N+c9BEV&7)llfqp8! zKZY@{Qw|Qm@5eAP=GszQgXFde3_F&x;-Ab4qTSFnU$4pzuNGg9sUth-ouLbb8~v&x zQ5Eza-8`O6+5EuaD!cgXxV*qYz4LbWcmBQ_oZ4T%_*%xdR!}(_%&eDBX;W7Q>{Zm$ zD=5PZ1C9S5RG;MZLA_dg{_L~Qe5_~qVQH+)G>-qZ-Pz2)QkLcFiT9J!z^mf6A#UiL z9ll&8Q`~Bbte_GSBN~)Q6*@$RH*T(!^<#eV6kA_EUhjCBiwXbvIX1W)7)gN;Lg+o9Yjn$UebV z*rb0F^%48-=Ut1p7sJY~PvKI8t=e{u`K`rr{I20=Gss77?C7qtO2q9-5FD^O$Rw8Jb@cLbN3$hS*EiGHFaS|_84$h=?-EB zczPz}7oJ|)?~OvVz+4qr>U&z=EkMQykov0&%LbZ}ZuX|K&nqQpw=_|*U4WzGt&c$z z5KDd9!B5sLk{G=|h&AiQVIRXMs)bBlSSOt`CTlNsQ^?B{^ft0*{ zUH+GwY=xFP*fOw5!UsyX9MamQCu6ssn`!JU`&^(0?td(7Zdlj{O7@@R*~j zQ5eOzlRo8l6MLjotTgB@pw~Mhat~wG3^-ym* z0}>Tj&-2u0Ixr{6F}-PiQ|Bzm@cY&mU(E|L6UtxrH~xKkED>H0igk)!DsJcbJN_ay_tE{DQs zZHIHG`PtEU*$(R3hwp^gM&|hJrrbEiX(x`CBRLKW5HQFDrcW-Q3aCxtSx|nHC^2|# zfXYP!-KfqN;dz$>=;_q)gVNLcTzwRg>K73>6no)Lt|a=AO7`5C&Y{xEt`t$+E{g1= z681A6c`x(67_<#|Wt!r*Xj23yxebZ|;iBlneogUx+oXzkW;7-54z8R$$4YIB49@q7qik zN|cn&$emRY`1GhTzFPW_+=~Czug=Fe#iz3+T&w!H`nD3U$`7em)~!MBzPpzfV5E=ivN@ z&N=53gnx`gG7e!L>|T$j9;lkD)WUf_y={xI<|^n<{Ic zwRze-Ix7@MxjpL`?QEkpdX(9q&&ibUU&7Cp^nI-*^(z)za`fj{J|c(RcoW=eWgQwRu7M(_se}yX zcyzV&JeL`CT5*xyK`#>?0)dF~_y;j7BrWWEkd4Wb;gY}aSfzLoX<^W?6wp^lG1ST##Yx(Gsv_ww#a2NsC`0BkN0{Sb z?W%2x8lWr(stFo-%(r@dtgcb@1V0gW$d51^Le4`U%%SOPy!j&2F<99ciqX8~2|He} zq80=U?0eZeW3dx)yTGI$aKpDNOyEbs!WeV!C3D z3cdPhOp}W5ob#LtN9zTiMFsCK;&}rvkW6gY{p-Siu-KmPr8ED99GyhktbXDbDDE6Z z>am0|i~c|j1ac6`*Q=L#UzK)+?xums6ca62V~Ev(n>Z93dn_|~n>m^-zIZhfWLyH3 z0aPIv^k`6m2dXhE2j7rm$Q0=`mV~4N0cwqKuS$1_tcqVDX#{HIVpW!`!?#3=7eM_{ zgECvRFuKE}N2SAa_k~OSfNQ%AMqMP)L(u?ejypQ>ypb<(2nr zh0xvB(Xk_Xeo%wj#mEVBalWIla&VR%zwAXgb$le}Uda!qy0`v#;#U?s^XHTmuaXXa z`z7}UI$35_@!dyp4=B<_B{ZsY=tJa=Pd=D6T{~SZUc*7U8E9Yn$SH4VUrd@+=3mQP zQ1p2{iqX}38t87p5yiPc(^^;WiFYN<0!dfVov9^)hoP4k5FyBi9{!|RmEje0P0jmR zw0YQk+4AU4`j()YKKc2iS=jN5)#c^dy?*sd-KhvvBIx&8HOG80;<=n&B*6eV+T%_d zYh*_L$INc?K;9`e%vdg}QPrrh#ydS4`EJ$+nfa2+=@piU!Ha-8WCzFb)r9!yjmwtX z28sDpM!9iWjGgc?693)>JwoRv@w1Ysh@sZQbBDZDgIk! zZE@#;-!oO%qv+P^)msAEX6MjLpzPnAG{F&;OK#27s}D%q>2;Hnna22XzY2Me0`D(n z&G$RWHaZ_b&arr1ejg`?TWnWZXXzqhm_&|P`AGXH4qARnQHXr++|zPXg)|mMV@>0^ z7@!SH^7+ev$2i@XOtD&%+AX13q!Zq)5rrF4}jS8$tb)wBUQ<7laK zc%o@E{_&`A_WLalG{F`$_5N77jBILlr%wQpvIWTmnt zazKT;2)E)F3`roQX)iK&BF;f`+S$4F;+Cj`A#Jm76XX`o(G+O2!fF-ef$z$e`x&6u z4Has01!w6ly2u3v$Vt!`)$~8_Ci|;t#hs~XwvvF?t>&SD;!acKB$crAl@kz*#h>93 zZ4SNkl}#WB+aTT+1hjG}W|Ku5W46RXG#_G$*!8@eIZdD+dqv_nirmnJYw>!fShT5y*Qfk0G;OGM%H)2|MCY$a=jn32MsGF(>R z_kC!x-WLb}p& z4&6?XV#$0n2N7njss*Sxx>Ry|RNCymKt{-Qhb`ZhY-Hw99tSjLkgIt*N9 z&D_q=oe8WHL_mNr^Pc(@$W)<<{H^5D37gi^7`c{ zY#gw9PKDF|lr9;xZM?T5e{&a7Q9I~`65aL4n^Wp!`7+c1UM|@dpfj%$i$aY7-KsNw zt>TifRjM{-yC&CD*QL}4t%+U}+akTi>awG&Bi92rM!vkk2l7@^m-y?lwd-Qve3Rb- z^P&gf#AofC0FC7j=CuI1Zw&vu@biDxT0G(J?Aw)1%H6mE0}3vN)n~m&ao`45Q3<`^B8-zv$ zxe5n7s>NOMd%y~kEVI6Asm~K&0?A{Jk`-}B$w{IQI;AcEbCf-Ov!`ve*FkYcinL;mtCPMxeHG~=)q*t~WV|%p z;F0TkUo{3Gt57xM6?#cEeO;aw3)J?T7-K*>d)oiBzh0d#GAN)YL#e|GGBaV3UJaBY z-e;8W0koRgeS%t*?r31LFrBRw?~T1DN|xopE*>v2jlU*3Gui+vAdHKV_B8G0;LX8$ z^({dY*+JT)l9|q^e#M1Aqv}oE^>sj9G;RYf_+Jl!=x8+TGcHMyp$|3)pv2wH-WQE% zuw4YqVHkDba0Hm{jQM$m+yXV#-@e#LQry^Jfz;eEgOx>bz^=Cmb0sE0+!m6~8o_bd zC-_*ZLxQX(fPsjNGtzUh0G&V}}FRIE!E)GP41_sNm)OxCDqW2*wLg*5nK z3H+bmZcrx6hQbGsyOVW+_7A4ITg6Kt3JHW}9m+fi6n4^2q8>z@g&}un(pe-(vJ^X2 z(?fS(g@jBxo2O||mTMa!d6Ns30DWWy!S4p*Hc79xcuIq~T#KSE22DexZe8g0NOO_X zS?P8GRy5U-nsAgD0-cx90n!1(GR8~n1M*v|d%bN-51&3!*MNa{$wjT5_!)Z83f{BbB*daiL);ev_|PHUUo{|3#B|@tQwRx;$uM8ID8BP-Ahcn?P@_gY_7Ul zla3X-LRoO7?yTvHwhOZ8J5zD%h2so*UcNcpj%4BI`*xME05^t3wvEvtZy9LoVOOWPxsLYjP`WCxT4rfIY2bH1CVW^)fz zH`&9&ZBuIl7fO1WL%vHw@`DP!FMFqZXNPys>!33s;&ari0nA&g7!wS-L`ehpF);Fc zOxo1jxFV@B05s|8ZgID`E?~_Rz4}pLkyO{fb;R77(-8v{l=%KI=@wu|z)KkE8)+Lq z@9OG5*_QLTZR?-C0-dluVY>5ki(;xo_o1DA5yAS4y?&TFb-I;#;g8>8GxWldoO^+N zGc*D$CVWNad1|i3Gt`z2iX@S_k zc@@HRR=*dfJ?}Wk{~GRx)e3pxYN&V8>#5Wl$=i`lMpL6sUs2u%d|%vI@u z*c_lbXk!=2;2&~aRiMBz@N0C%vNf_DnsP1n!e`0GhCYXpAD|M_xDA2|>LbD0&-c?C1VwV&_iyb1W{rxk z7`|Nlh0HHDD))ro@wxB*9xH2;xK#3pv}=mwMNn0Ck?Eozs`|8NWQ(AiuX7d@c4#uD zG;?`^qk)M~U0Wr(CpymTk>l};{ALFx+%l(dS%<#j^ai_;3o@J_*eJ@SkKk#X)kg zkV>eUngb2lm*+i+DrJT=x6;|e@-}s|0xMiP*-Hu(!x{8IQT0BeD}`=?qvR5B*;IyS zD0HZ-R3$q;Nk@tcCJAzo3C&`9#D-eF#q1ilT&CCYn8qOS> z$&`$z!Uj(uI~lI_hqDjl1o2y5e_up_DT2ijrM~xJ&vYLu{~F`>y-^R{4(H>0>6`3w z;STzNDkTUMFy+_fE#aHu@*%I)6n~7~r6`gbM9>d0!$@y~&zGrg%{e1m3Njx1eZn35PlGK{AfLdw`&vtT3K3@m%Ai{di)k)N zkc_U*8P$V_n;gQ#Xa^s|ee9PO%(Qh{?Zhu?1E2V<_BKUEL8Q;L%luSaZQLKK zL(kLnI~v2o8O5#kt@a4<3-?{V*Ur6A#0>>j9`#-(@ABIYyYGJ)*<=t0LCj!buy;Zff2GpZR&fT4^`l@VCQb1SH0K>R7jW?$_vryc6vkgH&7#+ zdqbVQmV639#{oaIdRV`HI*^kOzIN8@JAYpyam|XWQY5^_B%+bNFRJ%^pgJu-HnWvR zMgc6c%`zpTTn2?_XOJH_p1t2yXPT3K^fg;Ob~{4mWxhLtbLqv*9pP$G8*2dOu_Eci z*h|s|`kJgkW)wdW9wn%q)6HHN_jzT>@a#Qe1jXpvvv#YV2wnWN`90y`KRs7HTl2WQ z)*JT?QtrIRmf4oy@o`Pv_rDB!aEEV%KP}8V|HgIs-);gw#0Ej?+?8<^)V_#4@;t%2 zvcrOmDL2?_P_K71@WzxpL8DKb4+v?;YzjXta52bDkQx=zT?=y4=&$^A_Mx#T8N2cB zCEE%jYbh>;B1u#NkS#{)HQNP7a6O=^%VQC zTWee1?Y4PgI|bW|6p4XIPmZCDN->&*Y7mdC3fdl>^6sc&K_0V*T$Emo z#50Y|CT1tIKQ76i$TdP7t~iSNtbyokFJ^r`FuG0&P^6i+>odG zY!xi^>D8tLEpd6P$S9`WY(PeyEvKF3XGVUR`?p)RDPTTk#Eo57JDF@`M#DbkHa3|7 z)i5aCP9yJo7RErX&p>B;>QKgEUsxAi7j`~42V2pQ|7K}8xINt3=yGj#bUJIgJVWum zq(NC7ndJA_r%4U>BSTJhv{P-)s4-@{R5^LIdV9`_f42?IyRB)llQX~=S-u3B=Zz{O z4c&#%HndA!C@++|xX*B;U%I0cJ<`3dJjctJO!$NJ)_=<^q4}-i4J{;j5-?N_k1^;e zE|Ve~p`Jw49WqI;lD3CLpk*{uX%9a499;P`ZoGl0P>yMEf>i;H!g|=lDc7KBHJkrVj)|ogq_|M zb9aZlH~*TvU6m{7);7f_GZ)C3Z#B(F2`@bEl|GhVc>V1^Co_K>x~MV!LsM0NQ&)#) z5jGn5j)zmi3%=A(HmamC{bz_ulw?aue#jT}C3YD-v$QZMROP`e?qxzn{Lc;RA#_6u)5c-e7*&0JX84rK}`?Dl46 zTr2SIK&2U;G^%vl1tT(Zxo8?3#)Zj9xaf} zD|>N^;%-vp29W5fE!UyU z^~?#ycI&NP-E0osFE~c8C0)V_x?*lqe6gU0#-jHUNoMS2RT22I*zAt%PHTY^3dgOA ztO|$Jj^YfnI0EjPi&yr_H~3$hlNF4a<8#3IgV%kc>Y@+MFQT);YUp;79GUNP2|hUz zo*sS97eCjluLteZ?2>oI_IRbrUSQGq0d2TyyiWTkGiAURHlW%M1YEj&S)K-! zv&*%H8A+m4VR77t%B`A?*`TWg_oDXTIuUj`zhKITJ!PJUV+9M(uXFfvUSOHf{@Z&2 z{uZzp=S5v7nQjc03M;VeqPU$D$$^q+k{OK)Q(ahLPchw1A=%6BE{`Z@8O;#QLsXo2@$hHGVFd=9rA+6#FOoLH~zjV=*@Jc5IoGH zs{)b)xR1i>$MLf*pH94E2>SH$aXU!f{B0#;aY26je(!Oz*o|F~tyV6`I*LoB$QqE< z5n`eRXmC&vwgbp$OBoZ3_6|+HWW1-^v3YipKk*#Y@Us;DXZDv!EvQ(TdE*ykjT@t4 zmlZ0uP#lO?ze^>Y&}{Nt&F)j`uytm+a7);m8OjXk(lZ5Bu$2buovWq!lD?>xs5?`) zDaHjV4&%eThM4~SoN0C1NBr>d)4|{6z3i@Ndh~92;~aCS{++ik3=7s{9DV?*7BNtN zjBfjCvdaqvhNh7>I)-j^@hdNMP=`8W+g|oHrB#e+f9UiDO4$~6iPvo?)!#CGi85(i zs+g^((?L{N-!u;Ic%fp#?>@KdzrJo+Q*O_${g}KziL_gp?n@MRfg zE=MN$HXsbHrY}m{u>cu5Fs_6*f;ccnKf8nl+CP@V%g|{IkXE2Xd|0xujls)qfOHJ1 zPIa*-=|d1W1t}PkCaf1U#^f>ef^NZ40k3*?2d8kD(I3m9<8#kz3f;F&VdHyyKC~^Q z{*06;*(tkz&#%+#z_j=L48ZV+$N299({g|c+bq;&779)j8O4{TV5IG)Z;qxK+~XCZ zK61+2)vnH{;qEwcM|EXF>u>(^ZHo=M{Lf2glO8wrYu8$l0WO)yEu2UOsDzG~wPdeq zH?W9efFx0J5X@hxs&a0IqFmL+J%R#*8}fR=CF#f7Q{*Nn;F+&!W}&zmt^n#k)y`1h z;T4~yvxPWTI*V7Hja(?iikd1>zF-M+g-HW(ta@>$^ejp_K*4^E_;SRR@T-wmn2y;; z!qa411-s?#V)GrzvOX^y+dOiq>I|^`;mEhu+%=AtMKYc`s z=-eQ@uPMHl45$(%jq=CxqA8c>b;94gPms*q1IE&npilgk`dx!c-cnWhC}7+48L*YyCtM@BzhG0HMhWhJG}1Y1B!c}B1bV7(5`A>_k~qTR+Faq<77o1byYM5?w_o<;fPvl=+z$Q?dOV9H^}v`7iRkq~Yq$-fo>{km)-M-&uJdK5 z0xKhoqdBb0-48zA4}A~dD%DZCIPf&#Sv|$M3+l% z@!iN&gc;dmCM`ytt9@RZDRdkaHaEKXEiY6|4EV=y-JWCdRI=Eg@c6R@=A_S z9JC@Hq7o9Je*fc`$6o6qvtWy#6_f((7cFteX6kC-EkyaC@Tn!PSDPinBoRjaTIoE{ zAnNye?6p2fmnGAy@y=wX92l~X&1_U=`VD&IPVb!FtSBc}rMk+Hdx{=yC)_*eu`Y6{ zANCvpA-`F)?!dIAaP?TA=Qxh6*A<7}jR9-?a3pOg5qVp1^s}pE0ZDUX92HyPXe-6( zDUwMgBny)n6ZKy?2u)#YNLOrC0BUnyjns3{6NC{bMA(w(WWOp=lR7hQ&;wLD1IIG^ zj!n=gG2+^Obd8a)lnOQbzK$RTX=an0z2bQk_K@-qG)`*XU_j{N4 zWYGoM<&#PTd&n~3cJBiEY~-D(NQ}N>l3uM3s#n}q9?*0HS^pUA5MN|Cv;dhd8Urlg zR$fMiI(guWT`zlsKo;FWHh3+SS9z61bVMNGMh98R^>HWXbqlVG$2iTrC&OV}@E&<= zH}J!SP*ALU#p0lN$Ta&%4nN93H*Ru%&&r+ILvdvkDW($gm~3IEv|HOM-aM;a{DC@S zN`p_ecnuKrY@U@j>xS<_NzG(jkJ{J^kw!(HdLX=(IYaU#2Ig>d8(Z%M<(rUEsF{pX z&U^h(s~MH+hX~i5jY5Y}_9C-0Hc<H>u899B;vmF?w1CIq1fY?IkNDoTRwp6gfgA zw2JctJ&HA4BQ%U0n!eq8M@W*eLX!lAm+9ce(xzFIY^h`Ie+fVlgsrl~4X7NpfQ&1B#%-dZIT|+$Kz9C9H$G z?rf0$((P4kq~`s&))s({PmRMZW#>@Yk&wF<6>27UuSc%=Pp z+AnNlFjdZD5V^8ln@V4w}tps7BtNKj7Sh*=z0Fzyvv4x?y%!1L+1 zENA{J^<`&w+AcT}e%c>2#fsP>|E^dl=RFa-cW%9BwRBCaZn&ll+D+m?FUsi-yr>1$Jbx3*1HG^Jax zL8cE%Vc&fP6bfeaYS(ay5m;K+2ZEG$LsQr*|G#GTvtfdN{u4=(?c(CLePwo*i)`Wk zf4wemkYZU;2b44Ih|U#2t~e||tk0{;8KQ2ye#j2n4PK*;{bF8rgW4DQgWuWuq|Zt* zyRmCw$4^a<9t=4~8^yU$5WEj~+P5hNLt067SOyS}ZqszKSg`izY8v!u1s903qa5}o z(P798Ej9KAa`JimkZCyZyLwy7T06|JcYF#U)But!A4^xW&*z4w#d$o%CHL zD#PhofXJc(CbiLx1sFztE1ZQOi`ms+Zy+Y({ zH-Ps{49&>1uTGr{`!g{@v4{Jzf35mj>Q|;(cDfzvQ{9gj7t-i_TOd-jZ3p|(r7g4r!Er#E;#5th>n;hrM-`S~4CW()a3dKFM? z-1E&BREg@v!$;UB=!@!SGsX~CwVDikGd7^uTESM@Kom6GmY`*0U&4m>c#zDTV?%T zoBSKV)L@kgHb0u7l^)%ouA$2ziwv@ZPz=?{bZeVAr)$zU4>+U7$cQ)c15WDiO1@;f zUH)ur^oX5Zf46oq`&Op^9^VG~_jqiElU|(>dLQbpH+b!an81qgt|$ane_Ws3g^Sd% z`5OhAhTZMD!|8)>owBX8=7SG6c5UpCwqQMrK4|&0;wewSY1!{t%N&@d4>A!jG%M}_ zB^b~$VC6PCKI!aokjKLIjBzN(9R);id2$^#EF*hn_W1=_)}DgF@9iboZoKwXTdh4M z6bH%+`BXxKk1-}omL1)w?t})hS_T;5aFdp!G16^pT|i=7mFQYXraCcBciH=jC{y$% z5;bE-b*mVezp{nv1*@i30kLx~T^Ms!q*o(4T;br35(KXEQ!F6hwdzdxl_5W6nguf7 z*!A;YlFe=mnFCgkDWEvu0?WlM&_nW`z^GUxIqjb<(G?4>LbpabRLnl2pU915=bW?j zUEr*IU%1G7O>|~7aP4DW@E-k$c@Xs|N|(tR71u)U`k4%if!^gBW*29`#T5pGA6y!> z7nZ;5F{|NRjvF2$b)GSJEX<2c3>m-`#;2CCq^=#~q(Tll9PZ zJRTk_A3%M&7}-OC6VGqfJu%vvj%UOAZtSJnNdcjwa!ZgHcbV&E8$nYYs}@a;X@|0B zJnQStQ|0*VJ8^0q7|Ommd3E4tKmWbO17UMm3HjKKJ&-k4%n5@O_n0D|Pzg{~FW8{1 zLRKG8q{K>9BfV635GZAI8KGG;Rt+?&&q}d>^qRa+aU}dfl&+1b3s06M$EQ!%s}C{I z;}upFP_FG4=%G%wjXgBIja@hegthaSMs>FUFHUD6UnFh;9!tHtIr6a_%PP&cAA=sU z`(8!x<%1p#%6{*sL!&9ew$^5a*#PMle!Z0UYa0*=K_^VZy^n+hzn? z5ZL=;&E_%QtTwS(;$=(22Sajb9MhTI&W@X?PnW0X z95`Tm9K6K7i5el!J3+9la=WdXj2+RI%4tcHYnggQIZQ&Dta4g9d;hzJ@ArNE@zvJiO=~dt>RwbhtsPozL)i-q}PkTw(w`J_ZLPv84fXYhH0h zY~^{l;epua4de9Bcp-M;=@mCaY}Lx#w)n}8jM^hXi4gmEniZzCr<(-=ZSX-3YLcO! z9om+!60DtQ@aYzGDs_t_UE=Z49^wIzr|XK-;qdH$yZ|ym;D6z(l*Mv9@qTie40!Mx z07~PD=)V(w@#}=&{LfE+x!}L$%P4L!MG{>fhJN-x-Rcb4Y-MMXDQ-1IR!|8Bx>8gP z9x~F%q0iy^#n0-`8hZ0$_zYb+zOVD<{rCrgLxA8;t|WSXeT442NRmn`Kon6NBn)>_ z37J5_W#AgP3p39})H2BaX5dysRd9)@U4u=^WxzbRgU$=VYiek`s)hXoc<45T88{T- z?PPmqUm|^86>v>6yE)iJfO;`to|>7Fk6Z&JkDvTds<8m(M8aR!kuo>-J?gC>P)Bix zD1b#my{bR32I^`%qq=5yYwxLgf?L=P(PhA#lFn}ZQoYGtUOFjJ(h3?>7l{$XxR*_W z!j)*e53ehnQX7^YjrVoY`vedDo0tX2%&J#^;?o(}>U~53e1&v9Dc4+;*K^e{>?P)u z^v=|=^a~D6=Gb8|B3OooAN4D*RfH=1;*_$Ny%y>k(;&FYcFn`>Zx(%Bp3d&|y9bU+ zQ~Y7UJ>Szo-QG>{yCm;2*-Qa*N_XO6oiE#{coNb{FBBc8pM>0!BE#rmKV%SXieC(D z7TLZ%em%3Zf#xwY2(Q^i?^fwGM*pGfkAxfOgJFdPyM>?l8E4(2OTaNj?onsWlyj#E ze(%pb@X?*%H%)HqDXw1BR``kqKSR_BZtK;)X9es}$q>rEn)t?BYhVBU&8TCvc{Mwt zSnGXCy$v!&;Fc&_h+#%L+e@;*Nr8qZx|itH%e)%n8-IFo{`vT(c+@bvDm@U345WC+ zah3U=M3B*|iEVOnGn5knJv!vtxr-m5C%o0U_&+ST`AzdqKXTKJJs;p+9#(_dO>y@q z(n%#`(YlN&#j1K(t!_cDyQ!+-)|}13I$VRwwWb8N>4R(t@L!VZvOPP<4$VGARe&x{ zTTl0>i~**PJD804lUyI@Z$g^8n%*2-FE0X1+9h5=!4^H8eSc1s8n5V}PfuAvt)TAH zx>DJu={myBrnuW zsQcj${?)en%1)8fR{EjW{)lYhLl9csFg-=kIpZ4H5xrr0qU7<6JJQxTJiAd*4y|Ml zV>e9SF#UpbfB1mM*=b#}0gwIR*<7{sme&fOK2?Wry{c{waHI&^UBk}fjG3|XlOEG) zUNhsqJ6};^MaV3Rp?a$-zLpfYv7tI)wZK(U9O!oLH8-FPg@-XFoF|8#>U1_=(kkAg zx<{95n-yoJCsev4;RUo=GmS*Xo5=e9b@?Ld zCR;rJ0kF4k7Ywj{^6n{Rb9%{w9J&GaXnj#nLRLz)1sIq^)1SbeY@zUw;+lL(T!XqY zbfK$TY)6k1XQ09U7`$*oE&N`tM>Ei{G z;p@qB#&pISg7&I%aun>+{L!*4(* zUHyIS-r|=%E^|m)Bu$nbS5-sH|0DWi0vhw}^cpTllL@rQbz#{|y;q}uRRE}ez$JS1 zMY=O)&|~X#48q{8P3ji;hLH5hmI}QbO+B=ybjIk_c)9a=l^sCBaP#rJV<%z2JPa>G zGO;w~i^^z=AENqS-4;^(nMq-uvhrFEQ5@tnE2xA`;m{u&S19UY_sujaj=X+#{x*73 zc#5Dxb`<2QmI=>jupn&lj8a*fs)E`h-#yb1*23Or`r}gBdgYn0_V_KncsB~pWc>UF z++88-rwYL&a(|=e_ybCqFFxt;N{v`KY6DY(uE_^I zx|G$R4yH64hFwIo56f5j+qiPbacpf=%`C_D7;s4Ro?Hux_uukDG=K3xu0!!abwrm&OZF; zm#)jZ1!WAwf*Z?R4mF zoo~CH_Me?@+iel?EQks!r~ws-AR-8&sGJmm@dP-iQ9L369Yqlk8C3Y+cM=^EiRKLn zU$kFmcgTBsVDen={oK!exUNB;FCu2k+7;o`L@?7R>&k3jIP z!UO?i08DEdCKqv==A!v$7tphsdgGx3lWWf-whhT>Z7ES1zla=5m8ctgJvORDcXN^c z%*4|NS)^v^UeXIxGSCI5I2*8n^m0oB`=+Bxje$m2+1&e$&}#Ng;Th-w5bE}Lf)BGPj^A@VW(`HMAI3W3=NJ0 z1YddNZr3VS`IbyR&)G%VVXQ+_8p*InCA2qU`3(B3W&Wu_1ClD>dvGul=#+UF?e!mV zj^z{>Z)f+KyKTu<%T}*q>sYakdVgdt#GWFYcJKwNX1i>Q`rP@j2N2I4z(p5X0m6Ob z8h^hqJ9v`W+r9Dcm;Q$=XXov@?|&K5S$MnYlpJ#J>!>*FoK0g;9riwCrm%(qDvMhb zNuW)EP1u^UIj6+BDT&hYgS$NtA7Q>+vCNG6lDs=@!vrrgHvX(R`$w{j9X8z9S;@7) z#zsn>Mv=8t+@a7VQLU0rUo3W!;M%upW?K-j61RiDQUwGdSTE<`!ZzwR{KAv$pz*@t zoOVMk^_>sGpO}sB-`Yd+UKwcATR@|nl0)RTgo?u?b_ciD$hm5u_d+iUY>_XL%7|i4 ztGqBsk3E$-IyI=pNcXGxLULZ1#w_FQ561xSu*YGa0g0xcU`M~n<;_;LGud=eu<6pU zN1kev|9Lp~*kRVAB> zJ0aXDKB&a91xz8W52)oHj6EaW8LLyBlkAMuDwcW<3kzcK(^YYq*8@^dANt>y=B@wP zHr1h7!_k?+?VK~xE@`ea`FX8pXKd4NJ(}~5SvgMBaOj1lZ=1pLh`aUYr1F(9G#4$P zb&8Th?@a^xC#8W2f;^1nHz;Z(B{9fn52mG#to7Of8@9FYToIQB0*A5nhiM9KA*m5 zsu@tKq~%p)%PRw@gBF0=Maec{RMXz!v&o*vhjhp42skTmBkRHahdlscD%4bf&KULR(L;t-;5V#WypJ^ zksV&#IEB<@ftM?k94yj#DsGS80Q5by@y|hzU%|ZVOfi)H7RyQlPeUvYT78vC^0GNA zqi*sR`esNz6%0%+kri-z=p4blN%x=Gr~RqBTm-1u|EWKRwGhKTPeAgBAckVEKVDpAsLA5q6_KV zAS6{=6Pu?hnVlYboRs*NsP98%cphi-7A)?*|A@TF?pqV zB)#UBAUHCmV>awpGNfsf3h0~4QsG6?&K#Y0cOK@m*3G{=Z`fnqq%No-&ZH}*R!r4_ zzFQ}h=(!9hEFdwe*WfC2*j{ad?+NGoe$O-7a^`^Th%%WBy%i28f@z;F7fsc|~Ah!@3ygJfOiIc%v&Y<<}^hWP;Fjq zUM{mOESuV%u`g_IM!E9Ax?SJnS3Pc(k@ zw(a#ORYmS=f+y7dB6)&&+mg66{3uk5jPm!pu_v;{VhLGI$yZTiITcri`iwpr6BVLq zr{cob-uBCA&+1|OY1dt`4PQ0y^U2<3M3le#-f6Ou-A2}Zu}(nkW<+RY6D8kB!TiKw zFa*~KO2=o^y$NqHS~bcg`TO!WZQ!qtM$X;c~>3WeR$`H zpBWu3zbacm(%GTIjf1VA`#Ay~d6XPV5VnJsqt|8e`e{QD+suwB^~>dEPK8)U34QlX zz1nmM-4Fw0V6I1AgXrs3@dnY|H&2k$u#p+$riuvG|g+Dfw24XsI|<9)qlh7)yIr`oLC^TE{FDv_)za8#oIEx%3i} zFTxxNMwhTJ9n0QK?w?EDIn#IJ==bZoZ`S;Gy&CR*;QHI)*7-gt1anX7Wg8j1-E;-7 zTvSC0jTsoz5$JG&NXYhC%jdQ;ccgnodbI&aRzW`6W235st`{}&hdgrVmZ)r62eQ+7 z9K5rhZlKqQTxBa91@cpC;UYxac1A|=MX9VKkKeZ@?P2pKxbYHXgW{*kCkOaZE1|nS zJ{;;-$dF72Ihr(Xw=7F^MSS6{3vVr%g?!HS(C>FSxN7zdxHE;fb()TIU)IaG3QbSI z;4rG54B(=lk?-6Zc*nf_?3&KGN2=X8Yj@dVU2LM{&<}r%ii;N<=6B3qHCHbyjcMjz zhL)cVqBedeRDtZ{VuE!mH_6k) z(4D$fbPjp~R=jPP*Y~6I?{&_@j9(oYC@t`B>{4xh+;~2rr>DpDB0(^r>i7%6&RjTaA^q`68^P?WD6N zfkGp2d1KfRHM4g3R4P%X&2{iFPo1MEab)J%Ji`hlug&@Ib-S$jINTN`WfqInQ-+Pj zJk;hxo()Zp^gsxfq*ZdSN20P#3z&;*z^Ks+0(#w^sHbM`O~!J80w|qql>pskt7NI* zmJ9t9N1h(PijVP`IeAt z1bf!A2{tk}d+H!rSKiw3-#_*mztRh*HYla+;S^}RV`pQLswXiWN4 z2CemZBrfIFa1!U9CCjH6_@8l?@lQ{E5b12c#$GIquqC5^bTYndV@X;u_fu<^<)y-n z4L9bHe-MoxNn1v(aHOsbhhP#XwFzunm_NMD6)X+W=Hg4-e-Z$Tp*IX{3#k8bAi9m;$3N{01e8$Wo&!PFO>>c99h}j8Q5vJ8;{}IQJ*1j$0(QzKK6)7r z&E!~SX!N)G*|UyJp6wYLCBW*w30!eR=B}}{!n<)9gUwd>4sIn~BCPQG@NB?w|M;nQ zqlY|BhUd`RnGZjFqIy&3og1_O{^t;6*}};Yh_n42j@%zTe`oqwu;IR_w?&hHxzclzSZtQp#S=f(llsuCHQHMAzzgi4xb69?rFRG3Oj>&oPg4JBjAf5eIjY_K;_;x|+*?p7CTfDwAL4Vttmey_Uh7F=wV=D#H z$(g9OhCOgoolI^-A17K>5>Vq9K&1hT?~-CS&DB&x5AiT$`w~b8H-4@W2A#k~fP3$~ zq}pjTIab(kc>bSa!fHS#9kMeiyAuA|WX;dywrH`zxP%f4hlE%fdxbfo(#^>Owv$ES zS)y{NRZ8LRlr#vA2=wYiX^-FTxyPYu3GX{hYIs`J@%cME&jItfR`JL`87kurGcCS( zAb6MZHRLxzF3gjYQ+S2S%}VDM8jKw~PeJ5Vd%_Ak)W3f7*ZUZ=)4E|z=y|f9T@kh$ z?{!NpmZMxs4t>U1RGf*~)X3+R#vmukPQf-gYO4a7Y0Qe*Riq@klLm!Eh~f=bj`r9BdAAsq-@ySUJi1j5!SW9ESSW#N-91#I`HS{JWuCAcwozD86i@3PcL7?aG8V z@HTuGuyQ3|q*Y~2D)Y~rlrPvkIcbJV(a#d_BUU1(u0_whJFovB`}JvNq$GYEu!>~5 zap6gY#oX_v2qi~z(lb2!!$?2O))=wJOC-;9{~ z=QbQ6E8G|{Ah9u`?<<3nZ=gsj71zYy$6YZg(8iIo*g(MmwzwEniR)vcRn^~Za&%+RQS%H?##a;u9$X6h%9S* zb(41uZ*x$I@?vO9wDS$a^Cp6Y(Rki4PQS(q7gQM4|Kp#TjmG-_`NlisEIWt9jVqP< zEKE!XCBIIQt5jS&r%PDMHOMx^ZWp6A!?IAFvQJV@-R8E-&hc8*XMp4HbD-AKDp2Ta z@x&cI4N9$|F0dtXuV~3cyk1Rz?!6Z(Q`&sC_!h)0kKH%pn2%n44Fr2@IPJ_1y5NoM zw|ZlC&w(+x?MwmpVyISuY7k4Kn)!FXk}S#>T?j4a_JV(Zb7txEZvGekwLB>Ok1A1? zDbpC!(^&FfN$=xeHB|%MOdk$yCO2rj;{_;Z0m~6fmoq*-AIV!ZQ6J1S1MTAqbqy() z2)#BV7lA{R90Fuz(EKZ2rasE;H1^M|C7&~mq8=VbEgD6ih3VBBiGjaaSwC+HN3Y&G z?Q`hAdNj8yV12+%|6FbZa1Lj`bzGj{lS$W#Ps1|NqTWunPTRo2ds-rgg`kWYdzvnx z*OKzsZK^ZQ(`9n@_x}8E)BdaM%z@h~bo}UNC;n(%|HdX6;Kp{U%;GJ0THucjLRXlB z%5u;x+!C}ss)o0fo8_0!y~wHX%JQpG=+!M8yjBuZ&EF0L$GO}lsJlcC<4;Ml=xAs) zUF7!yTVX$nM!Rz?h#ze{XPs;JU)=DP+0wN9G%<&ikCt6_V?dm<0K})1ypAGAsJLrP zJk-uz7B!2m@&~v`ISe%F=Od>tvme5^;gC{T4*81+Dpm1DbfXdb|~Y9=2pq$ z8SS$5Trf2Nnf2T@uhVp)u-LDTSHY`{*Z|G01#mu@b0?&qYz)ib9Aipl)u4@90*b5s z#-u|g-N|n?BDRrxHMrQXjgPXcOG9c^mjceKwy8d>h*-^S{->+pX(f<*^mX{9m;4(z z7em*^oFXZ_#NcjOh1b0|yJ5fv(ihVs?E~rQ0q>LiV}9L!2L1)QoVqe&m+#G)tAWoQ z@75~55XTF$!gt2zarE4KJbX68e=nrZZqBUbU7mV!>N9X1H$1)oyw-4CG5l$dHF4NW zRqSnpdmrmRfZ17JCIrZkUiO)=s^%vvZWBaGyxxe>C zquDk`ze$d5{I1y?-JDf(mwd|39Jz7$xYJ_MJ5R~aQsgr#En&Mbqt>3|(pmRJ)>$;pJ3SYs3K>%*W(}6O=;j3zpZ48p9UqPP+{A2 zeM!{eP!>dfTO>V5XGBvp9IcMhJDa|UcMo4BfhJL}+aIX9H&1wr8_Gk

    8 z`3!n1fiBWpKnFke!Ot+6IrEeVU8J{w+V82C1c&a<`zGBhFW3wT^p2Ja)b)Ed91h)) zY@QOKceGTXztsspfc_fwWl*LB>PQ9HCUzU}tpaf)cuIsiQUO+L?Ac&|!p}}&P?iLG zK=TW~y>)QYpU*j-5}^k)zYspA$5(*LYFIJoodlXei-@dto7aLv_gaOgL}&&rB6^hl zKZPS_a~B4^mp~T?>fB*8)^nCx@RSH$AgE_~6b@Tf-S=Zqwgfsu_qxYgM%Xzy7t2#3 zbcXJAKd&~>ghO{;FM&Zh63B*T@M{D*Yr=&sJxCoN}$>_^SC`N_AHzXpQ@xps5Z?!sSxfofaJh(6U8#4Eid8&eHUm zR>yYzp+$fqsuu+jWpG9g_zvf)GsNMoC8qy51taCmb6TFb&rw@00mFc>REZg{**wc zXkz`>>iV|J89p|A7qpC4nB%a*mEo^KSw)`WjD(&?8#TF)6YU zKKPRBnKCF(0&S(WJ!ej9Xae-KK2M3zR$AMmF=w3?v?zF$HG_UjAX|FEKD;>ycSYH+ z<|z@fr6;W6-rxEFHH&v+&>sm@PTehs?^?KR90SZ_B^iO^-b*X=Yt)E8QG z=Y1}N3M9}d>JmpT*o^B&9J6^!gho-9xZbNic-Gw+snM(QIx3Vvy{OLcJ3Hq)9J;fk zcuItNQJrDdxb&USqH8{S3@Va9N2$)x%zo-Cfb4yFN`#J5oxwDv8MY{>_80~gOCVcn zQL2^=?zlN!ho?lymRb~S{t{muwtKZDgGwaO9GcZK9$t40w8(M|Pl?bRn$?Q)y@=Zw zdp~kyP^kpkK;`JR=I?R2d+oI%9skfP_*@p@ zh_8P+Pl-^+KlB>;;oe;Va-U|#pj{H^fND*JfAhcin3zAEr$p!geUz&GrcPM*?sIiB zgB&DK4{Fhjm|a-OAMVdnBGiLg^!QE>oDus~{}_WDB~S(p#Oz(#>ci>1QIn@cD1!!K zzf+I@3y02WV-SO!Bv9#XYSFViT;crHhNncR^fqm9UTKdDRBojtGibL2ilAFib?*mB z(4yc}o)V!5x&;+>ahU-vdNd%9L3<>S@-eY({A8Rm_&kuOL`eCV@Cu$}3sC$i%|4aa z(OwA@LFMSr7c6l!;p@p$A{0U8=;G2P8UUqL@5vx%2~>k>s@EpB2!ca*u?A0xPz|c7 zK7X2oy`Ypu6Bx8l0tM3~NLp(6Z-7D;^OOh$(j}`FtXKvelt2rqlOK~A=L9X9{f?(ZXd!j- zk`f`4Z%#&5Usp7N7Daie^sT&(TqRI5s+`z1 zdX5E+{ik?Jgql(1v}kvbb}UI z^-g2ZaS0Sn&!cH;j#~q?vJX#*P&7S{+H_P|3(%h9`3yQCfeeFb?Uu?|>;-wA;3*L@ z40bYdo9J>IprYzc`&C}BCnZn|f4X~j)oKAPTBpKOBGkg)$te2Zq?G_|T-=vI9ug>n zDv(5nq#tDGWL#fs_@IHf8T>#0=Q>6WUvIz>|c4&os~e3sT3C2dPpH0x@JW@B|?v> z6xR2mjS5_JTBaQtbWQ?wct@SUyGE-3`d%d^LLJ^Y8HKq2O#|p>kTHYKOQ7v^2aYo^ z!4^f_;wcf@PIurg)tl7;C{xp%K^G*@IO+uUR6B(mCBho=ln9NZPTyULD1 z7bQ?QE%4rG`|Tnex{B32B|_n}z&mu#X_@PQYDd?e8H@ANhE2H|4?@_xuuA~gNG zlTr2hBX2>A&UFlDkgo(1ZM34$fFZaiO?!q3FBt#)EJ=ed(s(nNdg_FqJ-sH z8{Cl*y_cs%=r9!}LM>Db0E&NC%%IB>NI^sRYxN@Ip+%?O^OOiFXb2zNq-$M(@(i>F zR9>)GB+y>E)qAxt#!1_!hCC%gd+ApH{FB*ofVv(V$RIxnw2xMl4Z3tJ5L#6II8TYt zK3Y+BE9(v3f^Pqu!Jw-W$RwXW{^j!e7(i~ncuIs!^652+r^|3w>;7bG2Kh^%VRUEM z;c!w9ptDnWN`!{donc;1$TT=~{{C(Zx+a0X($K^1RkH%a_ zCy+q_66h{1rM_>o_8UMS>++Nc-KC|}Wz{2R0kn2$B7*`Y(0W?iGcM^nE>Ia`$x|Y< zp4Rr}G&>PjPP&B(86&?2?ddJMWLfh_l{D^{D=))q}8+Y2H%MTXhQ{0iBKS|p6#=?Auc?&pKZyYTN3CS4dJ!z3r+(x$&9B& z=o<~;`>QOd4lPm%a%51j1hS#iTVurs7l2+=Nr{jRo!!&4%(ghp-mXVrIw7U`yE zGAL96rG%*~W_P;O2%w@WDG^Eub256k$#D%p(fukIbVmYZ711mBS_EP3I=DYiiBMLN zlhKscj%NXyenfjv^w zS*^KOqVcA>&oYQXQ4**+J<4AjEe-?7av4vFP<48g57ymQ0xh~6lgyxK3ABaob)Vvk zk^wq)kEcXv3*GBN)8;P)sJc!bgJLAm2^t@TE?RB~&?H@+5}^|`K6=yiV-i3`J2VGZ zUPt#N&{vu<8MHe9hb?b+@{|aDr5Tg0H%hTZV?Xv_(0vJH6+)kG{C(#kv}oBUo)RIe z5PHJibq|CV&7L@cL9r4@iz-z52RfYxXyzoI5+NcI{*dLb!N~*2{iqly5e&5hED*U)Zi%*noc{{8XO&mZw~*q z+Lu9b5=fVxM}8i&69KxrhNnbGm!3x@U;E+n$n8NagW@I7o!?aHH7T3~(Bg+YB|>+8 z)4CCd=e6O`8FtBGP=W;NLQ|Z6z5C7sNJo#SM5qf*aca-mvJjw%Ju3Q@*U=*hv{bpa zN7D}{GEVO0DG^#qYkO8|=Y9l8BU^_-k0nqV^@0ZakH?{9y&Rqrp)~3R?b!4jTa;{~ zV9*l@w2vkP+gMmVg+r$@mZwB$A592)j@=vrEy_DPpFxQds6Cww3;KBc&!X0IJS9Ty z>14PuZacmbR-@dOK~E)+Hx)ES->{nwElR7B5+QFYXxzT>{wcKR;ye!qB}t%ebTUj> z`ZOA#ZS#3bgu2nm5cwe*ucNVdLKyT+0;ztW3BhKaCjitpjHg6Ml~(;Vx4tkJTC`0k zjX}v0$cs9VesK$?0yI~br$oq$I*_d{^cVt=&W?NrrAVLz>N7l;RQC%&8$Im z*JCHNXy7oO5~1F7*3HRx#|c4CuPF?AA%T=jsf|){v;f-W%~K+zTuQxTTLDg=-Bz(; zP?`jqLWgc(h7T?kIIGH2A~b~#-Gn#);Ky{PFLPngO9`|;hCZfa{v02CgO~G^2c~O^m!JY3|3$i}NQzGO=U1G0wOCq5~t|f~ZlqrG!(go{p zR)Nc+=9lu62>qoCR(+TsKKN8;*)b?f0(sC3ew0shRXB97tE5E8gJ$p#3`og_7WE7E zV$eGY6ie$yEWUpJ3sB43JS9T0v~FbU?2>B$Ez}BU(0d7_K|R%&oY1)dooc~TBBVh* zRp%AQ@SQM6Hl#BsTLL|%&WU>SWBmczZp~97^q4v)*Xk@A4J}&xqL@KB5@-_beRvx6 zcPK!6(|AgRCehx9cl8I`0n|ue%dql-{UCvqUymKRWQ_wrAF8B8Ncr{HhzUA-06KqS zAcHkQP;muX$7(1kkQUJS9R}R3+ZgplvIFdPcc1=!*nOrN!ah|G45D zpJp^qiBKvn4*%|)hD&a4v<+m?R|ynL`z13vFS!gYI;O)@A{0yeCHok6!;iXFb4+B= zHwhF>eTE(hhp=9kQ6(ip!PI9^jaZC_&huj~gK{NMF3s^Rc{UO23|~I+lnCY09N&`H zdvNx@rHRJS%IoO61Tvw!_q5h!-Qm!+8^co~WI}gutB}J5-K#u7~Rwhuq^S5sJM- zV_5G`-T+N&cb-ANCD3%5Z8_EV8xAdX+VhkMO{dwGtUJ4o0d#tMG=u&~pf7pqishvT zD&Wvr|I1S%^d*m83ENM91VFPhG8yz&0vXUtQ4*t{Cjitvlcz+;fL@AXSyHDLK#vz! zFeqOFWerzX%vzg{_qrAqJS9R|!)f#E<-8vNML*FVR(Zh|NTB6ss6{4fI5%-4k*7px z`5Ah5*8FE!_de6tfI)>4Xd_ilj?|7ZfkWrnkEcXvBUMh0b#A^JTBLE(j6p>bXjLe^ zkFjN>CqU&@QX;e}l-`-yuQ5(xpU&URpkfIWLKD8@b1Sguw7h_)L@0zNeBX?GvJP4l zFyk15N+gi-MQZ(@oLmkqvNz=^5mLTLEhL~Hjwbd51Tm;o0x3VqeKn-v4S=Qx@{|ZE zKgoT3j$J!wQFg5q29-&mL>gN9o$|udJE1mDiBKX9Eg$K%{Rf~^tMeFCE`e;RO1%4B zP6V{*-Wr|~AzP{vubX;286f`@O-1E(R3U-FX*b96oDt6e+L6jrA{0)$IjX((#yW#l z-<}NGsjv5+3;A4WG!a?*8y0T|_v0xMa;4EkWS?p{4?FYV1P0kjpf9wf%H3S=1hlBv zA)XSUFSMlU`jIX7VI5=A-<1rq*N1hCp+k!elQJeGz;Vtm=l}UXo1f=KW1LayGvTj_ zt>7u~ugY_yP0Cw*@XU#v@67(HU6ST2-(~Tq?W0-%`7Yoo5mLU(!ZM`KcxX|tXkP|7 zNT4S40qe6(&2UR^^BA5Ip(gYJ>w~Yl<7%boP7fI5D1naBHa9bm9e&WFz|K4+LPsO% ziy=mfu`YUMUk-zuB#<^O0X-WR{Su&!`*}))v}p-wNwpD&;Lr{Is4~3rqT4NjGN^F$ z_Cm@CfV4jGln7-|;i%2TMh5{Z8?DQrJrZbXD!o~{T}3%SPfd7AgqEh#GsRW6EkLz= z6b#xcfi&srR{LFsn_1pgNr{jqUESd&?;`-(UwZ+AoFz~`jR=etn>?UJbL;Sw2<6j= z;JTq-C_pP0ZfDRw2{g7wO@$(DAKpf+7x9z`jjcgD1dZ421gJ2=gF*WxkQZIB*NqEZ zp+#AdJS9S2bit;3+2PurQSCz+bU+5tGo{W=x03+1?7&n10nszXd;Y6m&?19fX$*3a zK&$CwFzL_)-|<`5fu}@hHJuF0zK_-isKc9l1|5_@*J$W*EAb-!Mrv<)N`$V_&_iW< zFGGM%4{B;udBGl%Kx1j1E^=TU?5Hmr%u^yXmgeb}rdlTfH1Kd=2DwV0AN252SkA^J zpsGiBN`!vU!)w~dMGFBc`aXq0ZW8FzFS-Rin1l1(nLl_+gg*UpGHU;$?@xe6kGEow zy9D|{b3n=KOhVwS)1AOmBJ_jifZiVRwgG5|j|+niOCaTf$xCTBGXa|7%TpqxTrgQa zR|D_f+tscz=!gW`MYKG$EKXwi%X`jlS`e2!+yw;K<+K2SJOrj5cSGhXndVW&fD& z>nB2sCYtb+2>qe5|Mwl1xLwruf*pfSNgxfnU|+mFkOt7d7kNsAH0Xjg^;X5xd!)KI zgFGeBHG0uj6aQ5u0Cld(QzCSYUbIzD^?ypr3s;9T=(GenOe5=Hi$_ZV8oh?6MCdS$ ztlPNe;|AEUhv^LRl0eE2v@JFoyAYtmaXckL$`7>tyK~4|I2mSlEMbth1lmKt(G_i7 ztbTOq#8V=)hkhdqlQuZt9ko--xblKMBY||OLiJ&)lL55oxE)W4kS(Iw>5*#OCU2UDG%5(9c$MaXLw44%&4T?Ze9m-fEt#%G3bH>a-z{hiS|=F zXi3DOJWSArLl@rkJA-^B(D`aL6+_L};%X(=W;`WA=d0~D zdggZ+58aPd8Y3&Oqe~LVkf7eL^-e*H!dCN?2pJM|e9TlQXi@z~dJMWOflg7)t=w#V zIzVr$q(tZx)!bI^8jbZFo33LRbVUO7rA~gLU7aF;9(Us@5$a2w{6w$D_-#A;gO&{P zlR%y{Du2EHA#S^#dx)n*$dg9peVS)tkzA$NkwI4_Q0HsvijSQx*h7ncR!NCa=WFx^ zmAB?Nv`m?Qojm)>RJvF!nz43ER28FWno?WUfpnJKj6#`ySZfU@3ckE*<2uS=l4G_ihF^RONqx`*$1N`&^( z#QMgp_Bi5eKg@tZHzd$t8u5ATTDtM1f!a~$q(R@F zr2wT?Nr_N9>YQ|+71t1;u+q&83X(xI-~G~IR7-$d%XsQPAe!%9boBCDfM(7-#-Ljg zXg-Z$8>$=C2WZ$Vo)V$?G=}}6w((zpHr)ept};tl1AkY^XfhZXmmMGiI63Y%J&)s0tNuQ;n}p9J)15gCijqLSRAA3N*ZLqFI%7MY5+PqIu;*Mn zkGG)nIVvWV*HN?tsvV`SsGUJS9T4qiD8e|IlDKbUR1tGAKp@J*B?)xd&HxN`y3NJ+Oyj(HnpY8Z2PY zeF?OaYC$VCwehA}rpZ$xw3BK<-w=*bK0%;3#-KO}>>SN`xNLV!ZbmOZz~J-qf7RpvMv@JDj%F z2Y%TEEefo~QzDcdPG?<%XgnFx&8-;pL;}^IO8yj^dYu4zXu(q=RD&w{XZF`P4K3Ps z$Av+O5~zTdb2uCB#C=|)!gxxA3TQb;?@?y>_7%%kR~htF0%=jH)BBbd*6YS=^OOi_ zQK@reybIpzj5a1PC`kfME1=o`Po;RlcDLav5t>%uWMt>pAKx&UpZu9Y&m_=(x&^(+ zc)1yVqxUI1B|`h@7IfxuEWR~jL$6w6E3czu3G|Y7MFoEH$KCF;dh?VBy`)`HPfwM< zgOh=_iZduh0!^p8w{4*&&MenC!c!tNo$lUSl3u3+RHN9KL8%gGFm1o;-rxlm*psTH zL})N=ztXt8KOUeT)6E(5Tmo&S*Ss`o@?bXnMv*gkN`$u3YhK3H(Z&&Zuz3c8&$8)h(V9A%E(r{*FyL3N8A#nKy&dB+y{$ZuL%x#NWvEAD$AS!PMPqH+eWh zbGC#t=%oajO~e0!-V^aUGTq8kA~c(Z|4!E?--8y_PEBXfD+yGS?sd+)dSk!tTa}au z)uek}O8>>UU^1$I34>lspa80zm^UvkgWt$?08fcf098(|{M`}&Epj@fHLmi4O_x9; zY09A1z_WP(Ep+865gJKT2JYsNN0wxZ96d z99q^gHD%CS3FJk$Agyb?aJhTunLH&zUUUnx%<6(w;sasU49bu|$7!N!&Y(;QREG+KQ728u12nTGPl-?+DhzhoxxEmeMB6|HWl11& znr)eJJg+@Kaoc%Hgv@ESB|CCZdw{Id6B+bQ0$rzD(BW_G;sM(6hNnd6I^BZKSUo8R zNZa5$gWgLZP1WP*UPsvyXbV-hYVQ1>TG07p zJS9R~sJivK_5u@J$5^7rpd4JsSfk%g1HUJ^_}tG{>-vBH`9Ie&hPcz4@xof;;)e)z zo)Z765O;b_;)&G}(EQj{V;J;7(tJO9N;>zn!8N@#*6@@F`O#By)Qj+r0Hr^)WY9+m zq)K(sIv>MT;0L`L$5SGtN_EljqdoHh8rRW@L7ybh7@FeTratolK&G8|N`%JH6z7PT ze{uS3p51u{eU?BiXgTzHl@KiZce3Xx5o$rpq0be+tpzRm_%fP7UnJ1CLRty4CBzI` z6#a^)MCeT2u-9-Vdv5t;QT^vwbm0VukKt4ltshCqB5r;0Nt$4QzDc_ z!^1;~<=f!U9i3;upzjjMf>!&_uDjZ;t-w^p^?;+xi>mA5}>@MTNw0H0{K&=w{OP()J2n;@stSpQ>9le?l#V)-q?7I zLBAxBH9b@E25aJCywf&3B|_HpO!1j~=_9mg+0!5fcI%WN$($ zD|@fVYN<4om0cQUDJ8QZTS$q}5K4&>O49FrcOH-L>GeCG<8QCWvvb|oIOBb;i>E}W zCOuPb4qLbwT9mM>kU>QfXz*X!0Nba_34l)S<|z>x{Fm+wPqjh-a($~cuHriSDS@ui z#L=djAzA=gzvC$px=IsAQD^+{)nRi77%=FU1X@O2ra#?wb_ZzCK%NqzWz=O#k8FAg zTGZ%-8H0XHpjWisJ=V+ZFhI>t@{|a@qV;Z-Z4D0qWTCl?LB$g2;4vD)&$YyE?UmX* zB|-;}*&FM8Xf_id@Ab|MDv>}5q4aTN-5)qZWxIi=L?|JYUTo6$;|+kiqy{qRj|8%% zg>{R+tv|q_`&B6=LbkN9?&x<2cXKrFcaK4(66g`#RI5y#upFQ!hCC%gkLaeVyLI0M zXwmfp9~tyl0AxNGK#H&2PsD!SJV`*PF@Ano7P##dZNWfEvURc4O;5EK9{ z`c^3=Li4FIbM=ky8vyd3q|2ax5=goC;boH*SY`2ZGEa$+a__@-zdIWNx^+dtpmGTm zK{W{$sWnDGi=4xFN`xY)Cc*PiO+3AgHRmzNZs33KY*K!~O5fDk7oZxoc}j$oU$8pa zzdDw4gf6#b&`t?tL9-uaeZ4*dbZiAriI4@&ezcm^;}aaZhKWZRv`Ye6Q+I20r)9wa zslMbX5wfQ4){b!(0s-1*7{#F763C5KwY;>vHvqJ=KTnB}8?9=2w$Akf$n#hVgZ44hc5R!+RE4Lr#DG^GGp>mFCpx!~XG<2-T+LGwbQ< z+Wq7=_Z404e`$Ih!Meht*N z2dLDGr$p%3c`E1lppQ#4&Q(4t^P&ddwSu4kc$~liO}j_RN=hse1ACWb_aG~(0&QDh#q|7+j`*2|~7rmk2$FZ&%Fx?SZwB|?9yIT%}Qo;weqkU5qN@{m9QG=^<* zd)xqkg3NhJgaT*`yQnB+DL~h5?PQRr1giF&UbNLYVi7=@w|Pp0sy(+iZnV4AT!8-R zoMe!f1R70mBI*+L|6aPD(uSu*Xf(ZvXjs~498E0Pd5b{@Bv3Iu%DWrf9}O)UzKf?s zsF)t*P3Jw`1TAv<@RmW|5-5!(U>CkK!<*`qES?gfG@5|j7{9R$pwgkG4Dyjc59qrO zQ?xcug%)KQ^OOiZpzl7+cw+bmAl;y5lPWIQgA&LtQcY1UKmImA|0<J;yfI> z{7(rCIwpZ8(d6w)!);h95dWE{L}(IC-Znpf74HnUjPn_ETmmUyDyU+pstSkh#4w%` zA>~U2=ML1v?N^;oYfi4Xj!sA*ZB#U_EkL~&@RSJ6rwg{5r*AKS zqT&`Y=(G%?o~pIgkkbG;-{q>jplj{y8FWSh^;eE2+TX)I!~XU>B|`ma zH1Q|F>;ym?_xdr&PXZ-V3Fwk#aRprmO$01On2dk z-Gksax_F7FM5sEI>HeIq+XA44TALYkP68FvS@-_*YTS!wR*$DdsF=>WO`8n+0W@aa zK?a?dKpC`tG<0Ey5NJ{N^*kj)8MJ=nbKD0%c=S6ZoIw{P&~|$8<;ER$11Re?Pl?cW zdhlI-poLSNZH-sTJ4JWeIeb2Fvc#!?C;N^_!U~iZE20Qu9Z}5}| zU8V_kXEQe(A6;m@j6qi<&`zo~?c(3ECA4U>4o`{DPO3G1c)Otrv?zazGlRk;P+Qto z?-lVj8K5g$c}j%Z(zf~m!SNRX%1I1lP`CsNpedTMBmT?)=-Er25}^Q^qA_S@{Q;o- zKKB@ORRW!)uV1Z=QNeldmwkCkgig}euO8jLgA;?3Jw7rhLITyH4&)H6YB*Td@#HBH zszV*f_dOmigci;FT5VdzbrdOqmeOxzb~_yRr;N_$DG^#qztN7;OL!gmj@M<-H3?)* zm6-z)CgR@kRTFqhgsiDDbAan^yw_!%S1>3_0{zRPp6Y{#I2ROhfu}_1Uly&yZX9L~ zht93)JO*8tK#eri6+4^*y1`kuq8d+$P$LZo<4!YwH2`SxY+DB1kU&GILgU=~Mdbi> znZr{eG=wTNI=-HR^9*}#9A(f=31mYZNW*K>#{e|%CQpfw4Rs(tRd4hfTIAFuib1y| z&>K3v{j1c#S6VG;%2OirhE8w08$;>=RK7liLANE)SDMgxZ*Q>_TJ&lIPl?c1n$Vd0 zGzTX%LSOx2&>ac%hZd^zpA3lYo)V$yw0>loTLU}!Yd1V%&^-xMk2?AJcYXAsMT(6)B|`P6lOKJ5WDL|X4tkZt zp!);ieT*twod&gZ9{(7Q^Pmj=pZ~LtF(im?_1a5^!e5nFDJA|@Awl-W_Zp4E{R_ht z8Z#;`x(AZxucDo{2g*<3yr||-o)V!|w9|H#!#9MsAL+oLhZ1Pv4>g6|8eQxJPCv?1 zBDC-aeH^*_0h||Y@Mk!K9!a2!)FK!22cB@~)Ju6vgf3ExT8*0Q1HX~^3`+(*mO!uR z%&F4Pzyu(dnLH&zuj$NL(zOfTM)qFY$)G0^Xg~DJC-u&xdbYp zvu<=!2u{H6@5ECgR6u84o1X^QlNssMY-Ys;`$7V(pf#ZBThnu)MNOS~N`zL>8j!uY za~(KzMw$H>lqi9;Xy_4|x*Zp7n||Oa5z?ZeN7Bp>DF7`un#Q1)66gcn88n|*SVN2Y z58^2i`apLE|Av=w<6iT_>lld{&4pmX94D6QO+LCF$m z$8{C{+AM=9oOxRQ<8mm15|Dv!XmFEkUl;4BQo#eU@BuSPl=E|J@?Na zxYrt>*xLyVN|8W2=&W0H)%GQ{$o~#ciO>!@>zcmtz#^h1t@9c5S^_PlY5yMWT3iJv zuTn~cmeRC;Q2!j9D~R2uIjiD2dLx09tHYm7Y+nF$+m@$9NVz&(&!@H?96JA0JqD#p zAmwKNgZ=J50BGM^o)RJDX8+6?cb@|Eb&x58-b$dgG=yJjy6Gc8*9P;H2(6_dyqQt? zL4bS@EMm|*31m`5U9nZW+DL%bdh?VBnN)EwK6}Y^1VHV6?`6^Fx=(GVJ_o33 zF;9t5H9G5b-#5ga0@tVdF(^#}Eu#nDlObQ$0OT``r$lHOJ@|ryEU?14L3j*<(k0MF zI=$C;`r;({>Z?2@LL2GyzSA<^4Gvw$CK(LMkU%lCXxr8>vj?>3XQh+~#n7Vd-P&z% zu_@|z3p6f z;5Se{OgzY-&k`t(p0MZ5^DF_1n8Z^e6h}{3i>EsSphbN{!Won;f$qGf3)VFR*MQVR zc}j%t(B6k#9p>O!S4Zn5gK{L08{NG_hJOx(7Nu88iIAJ}?mcEImVm0*6f)?G1WKjd z&@=5_@lmd`jHg5>m3Biv3De#Ihwf~G)|`s#C|3eaqoo-m$E@_c35lo&++!L2rxde0e=U8~gH<2pQk7H`YkKf(4Tu5129Ns|0FKr+2IFiPr)8 zQ7I)t?dkOXb3-)}TI5q;!=QW#G>Rr*N1xoc0wC)`o)V!^Gy&U2>$Da?HAgx#=$izp zQCnT{)ci2+6v(QS5}_Kk9gN?9Nt*~z=&?WseV0HvWon8!whwWC_m1N{B|yh?eTfk!TckG3MEitEWMqhB)>O6 zC6!Vllo(4}do~nqheKy`t(tkob@W35#nAHErO};n?s{w#Pl-?rEuUSns5S_ow3fOI zDw065 zz_#VcLnO zL})IZb=e*%uK_x{yBCAXBv79mnt^N_=m#zG+rv{L)F+23%92iC=Oia{B7^=(pvH9Q zX4blo^9(5;cuIsC)1g~!u@U*-?{yZf@%6I+s?Ohk|b&YFWV9-tp^ox2y#tCD$0#s5d zB|^Wb7qq+XgEr8j^2LuBv`Ye+P)}9=NMC1wHZJ8U5i+5k>WtXWUjW+lEQdk6C6MYh z8nz_%=nK%)=R74ss@JFrz326J02%kxm{)Nf?U6ubG|$jt=EOMw>Gk3%5h|m3hPiD< z7X#G9y90yvN}zL8$M`k0$zXs!R7#1^IjUpS>eT2EKvj!|Gss>7ZKJn<{yp0cU*?co zDJ4SN=q;cDhsM?dNO!s=gB&DKi~DMd9e*F;B1n0qlnAxBZ*M&Ryq*(4H=}kk$Wa0f zCg^zKuq-%q&ewTLga#Aza(}BP0I4@T$si{QRE^%~717Re89>h~r9`M2z0qq^Uo$Kf zXus|jgPbMMOIk~faG!~C^0t~d9i8sdY<|TB>neeQ zsmh{yBmGA38>PAOln4b=mBrtY890Vr{jnc|+$4}=shZ;Yl$x&q8vco=M98s}UVSih zDK5PYHlD_y{SxRdmAl_|U5UcJv-R24c0MUx}U%xF9|e?9uw9P{#aLk_5n|c&?tIL=(YHF z3R-lxbv}a*NTA;I@q#I8R`cP|Me6XB2=%6q7p!d8EeD|M+cYgIt|M;=WJEOyk5bb7+=$hPO^Ov}oi7o)V!sG)sJD zs1M#5s@Jw>&|wLrPIIcYwNBuComw595+QY(Q(Zc3?|V3OtycIk=!gUwL3i&CHmh+A z`@K?1ghtTa`>8>@j{pTcj$zPI3ABMmd}F-=H^HH^dBRg7w1Gx^kD58-bo2*<3hhEb4Wn7&?{nthD@3yP7Ba|R0YO|o*8^{=e=DU#D404Y z+4bG_1N1(?j6s1C=m$N@!ykCqLW=?ec}j$S(4+iXlnXXqBWo+&DEj#Mh|#O{HwlQvNx_(uLjn}fA4aS z{Z;2B&0j?|z22##YXS7RD^H2gDyr$7bhZLl>S_&8p&}yC{J&X=J@z|M&lUrlj+f2xZd9+9|L`0zewWbQyF>0?m6u zmE@Y1#&GC#hVzsN&3i%}_4vE^>fO?_3I<)4K&$CCGCi^>t~@pm7GKV^D|$+D^0adz}-oRA57Go)V$$G#h_mofYgzI!OT17D^OaeJl z-)#83IIKx9Y0Xn20$o3s?}vr3WLHW&`f$6+KZ4$E#R#CQz<1vGwEe$En_!@ zz@gLq_=`bTCD1DBn>FlggwK8TPdp_;tEg`_?(73gfE2?USyx=J5fZ5IotnZi`~=?X z>_+gE2o=7gkFC7X!OF~00lgR$DS^spa?k0==;3haS_blz2$j*~UgynE^r1x>s*@OW zO#;oOp3Lv6+PHjHR4FAwbEzkzo$iqh(7pw$7!)OeMp7|eYnvBN(4yIvJS9RSsTl9o zFXOHN`8;)F&~*tE=TAfU-0CL*T9Lq0A{6IuZ(R4Y7VgM!(YwH)8xrUmbueE39(xZU z3w@pvp=;E^u==`b47AAA=@El&N}wxLcwDn;8IG)@oOw!wu2A7|td0L@fG+0cFzA*9 zYDkZX+j@0zz5CEto)V#k^q5eqy)YG^ds8$PR$NE7CD5sGDk8GHjH_BVr}C5toeHNa z3!PdxJaoF&fkAg9&=q=2E*wQ5z zSMV>i;3*L@p_NCS5gl9sGGA}Wpco0Hd{b-JXWhe~MLjq0ln5!`)H*hPH(p15U+iR1 ztOT;9La%Ep{S^Q;OXMjLvZX?=H=`bTXvveijcc zI`NyQM5rg73~CXVPD6_hOe~MlPvt<;(Oo8nH}-q(3bgp_ZkuHDcBmu4b&o)V!T8n&qKi^t)iX+JLpJ(fVqw?%cf>w_E0+8Oec2r1td)g^ogPLg|hhA`-f z1nNZ7EP0Xc7va!Zd-0SAb)sn&RX-1$t~C9Yz@VoRD3|UGo6MWwT58MhJS9T8bZ79} zbZa&IM%~8dGbljJkf%f_nWif(!9p(KgYn$`XK<#4KV^q6F$i>vW+xRk7s8D3+&0s28o% z`OF!D6B@B?>>2b@0_~tq{?Lex+n_}kb$LpJc2FnZcB0F0Xpx7LAA^!4Q2JH+zJ`Yb z{ziM8c}j%Rui6{?4(L%EpiVh43`&+jUFl?)JFxT*wCGQzln8aDlfk#$r}Y3O56WQB zD+y#mMXgOI{@V`FxxqXoLMBwydPOw^>wz2kmNO_t0&S!f{Gu6Ei~#ylDJ4Q1X$61b zb2l?+QMIbAmQ-A@uO(37ZTc9ccOZ7RvMZ%TsPMMEakWQVtpS>|z=%O_B+ytI!+LJ> z#6#E7lBYyyERA84mYl#v+a-~+7?di39BI1p(1_D3;m}RF##18XNYj;5ewbjFxTy7J z2ECO)9V65fR_%X|1L&0wPl-^+2zrUr<%vDu(0$u=kU{SxP&iG|SZIyIDvMXUc}j%B zX^JLnQD_E0dhf#-^j-o*QRn2R?w(oDqMm6yB|=fuIdOW^$rK>XK`$AUCV@<7#FyUG zH5{PogLz7XOlZVcJ^gW0fL0tXWKg;UvZEID9#=XEpe09mN`&mFMZGU3IRZ4STx)5? zb(A52-qB8h8r`FDPwNPkdjCs_&^y{GFnrvUp#UwNZ@?m%5@)HcyF=H+8puJqp2L%hA@$7?dS}HqsOJ zPm`9%p+$RicuIse(i3*)l!G{$NO5px&_@ZR2%&j~`k~(e3UlNs5mJQM8<)LZ{u_Ry z!5M)J`XqsJX@cEmr{f2J#%A)A2<6fQ`{3Oja6;p~;vR!OOQ14(OgJpuG#8-4p*$r* zW%QUB@b3n0zj8SKkwMuK=o3K(o2R=1H1h;ciO?s4idQwqNBPRqYBm+uQH})KKt0vI zw&qx~ZSj|hpO@gp{vpSsAa5BR=ih z3I^p$Am!B1birQw^cUYqP^}eDiBLnj zU_EElj|9kZoh^gDN}zq^YKo0NqB=l}W~}Ea5!zR7ZybLye-%KZUmRsnz646A_sH$r zV3Y(<^F*E!p>%qWTt(<5=i;hh()VzV@*PA51tYs+F9m9i^jU8 zFzCAk@~7$Ond?vBl0ozRJS9T@G##Be#U1-~Yx93Gs6YapqAqcoW-MMu$=`TNgicYH zcuf2UJ803nX^oauT(E@_Xev#0ew;j66Mm!U={zMuQ)#NR-2dkWfOcH##h@P&$cgH& zM*KQo6`&bWJS9R-RDV_LWnCQcb#FO|K}8a%BS9TCHE`HcQYj@u9SM4N#$+zEsOIKX z4Eia7MpNhHX1j4XyH!*vB|@XAb23-;8t&%k^4g6-za&sT^%?F=8d3^}uI?M25}|zR zGt4q>h~uO0{Vy=+w*D3(qzfsZ890rw0pr3TXel#&{3N1?c#Zw~mlP=h6x-&8X%ABdO zyy80gBY_mOvYhTb<{dycX7Q8=DQIQ6m&#|X8yOVYfkCAbXfiz}CPz-k;lI{3o)V$S z^q4Sm*1HNVN^L%ZL4PICXJ zP!k7$Hm&0+5sIVF1DS<{JO^mrvz-k3CxMh-glX8J({F&*KIbVBQhpJ}FE|Qssv`_e zGN@bvEu{VXj}F?a0;JcSr$lHW?cYDHZH!M?eUDoVvK#c@o0SY`SJZ;7);Q^;<;hbb zWI(&3;`O3S;m~;(yk*c%2^5>8rYO$5^8;FBRmf8!6q{sk{H2veHGry4E@jXz38Y+` zNY{V*9-udsQX-^Wn;76T2ww?X5YlW##Ra=t0^Ou(wB-kK)c}eOW{;SfYjMH1!%LnLA?2GrM!uSj%R0qwxExptYRzr&p+8t+wGQ5pty^>~@_p zmO~w*NBklNxebCk#(l?58(ny7{0y2uBZ2?t|EyzN7^kL~AGz!mG{5^Zo)Z76g>m-A zhc;wm1?=Lk_Uy0PFKNE=Q(60S_ha$H=x#hELds8NIYwo_gcdn_`Z35|0xhBUBzXM0 ziv^Qgym(54me6|=9zGj-0iaD^V;JNifx6QV8a_X}8yvcE`8*{;-RTECG+>`MKzAo( zFvwE^CD1bT=53qe0P>v3QzDc=%g`RTt<(XUbD^9;UJ}TgdNLEtm-Yr|z(t-CA#du* z)Sei37a*HDtyWcB-3KJlMS2|fToQwuD+kx*DG|CzkK>ZRWd{HXTyDf5Zwcf?BZ7&C z?_!nm+Z8+|LQXUy_|SbvcYth@W--V|0__Y_Q#|gPgAcDO$vh=OJHu#oxZ7V`OFh_o zGlLFFpib1HYSRtRz;Cpr4^N3uCu-6C#S8GpsD9)igAPfccC=AKH=~&wKtC&`M5rBY zlxU$j0xQZ6{0(Q&VF_eOBZB2=B{~2dF5@W?vZN6~i-*;*%XDJSO9ma0K<>1KD*b)u z5da-D=P40#r!7=dA8)~jSEK8N3_2=-*3e)|?|dPSaXwc{iO?DvOtl!b3Ga15t?R9> zxQ=`!P<6U{?=IhE04;LR;VBWSPIvE-?e?~T)7yTt0fUZ7pa#F_%?}X|v07=}7M>EJ z2ES-WM&d)oUk+0=2wF-$3cwyAwcKYj{e8THdlZ_Uil^ zFIe9M1%u8?AS=2BHLNzy2cV_TcuItyX1!B2!|j^&-mcWs+1BTOS+Ey7GJ@c z-r%z-3_2%)hEw0W?HP@&(4s8?JS9TIsqa0~p(eh!=Un+O2A!8c%1?5CiYr_X&}x<6 z|4WIG@{`$!C_jXwG|ia1qq~IRb62cRgQIxJErrL2gpMnFLCde=YpKq?c4JV81gf~#na)xH zX#X3Y5+UWi&Th*%Y*FOE3k(XCKp9jCGb!i{mKVAi@stQp!q6a7!)Rf{HR~I@@|= z#A>C;c^c~~uA^`Xw2w+|y0@=|uSq;QpQl7R{(j73C0wpcPK zQUV>Lo9fcs;kd;4d@E0h&@sBHj><5{@sVEgP6l0*K%1$0wtZTdCA8>YrIZM5rs`Qm zUzI@kjgotuWKfg@QY9$;x=|uPQ9XG|gj5MqGp(Bl&`FP547x6XI?!io8@U*b1?aRV zPl-?m`b=$)F*A|?n*04NgKkKmeCp(@1XnKu$f|&+L@1v+`Gcnoumvb$Vkv`eN}$)& zIoW;w*K#YrGV5q$}6ZN?{no>t*05mFRUDfR5yy#U%|na`lR5@^(8HAS=YA>II) zTJe+!je1O#Fi8h+gL9);%?%aTQM?3FeoSXX%##-Yy{nWGA?3$(_SId7+ZZP{)o0K> z38Y?AU6EAkhY!95&3H>p^Y$}hIczbi`wjKd72^kD}B7xMY7nD2f6_&dXp2$-oq)xq{_&df} z{4g=NoIy_|&@{R;sI54w0l!h}b37$N)9B8iF~AjH+v}#8vaA>SF?s zml{urkU5=otA`)NjWABuMhto;fs}iQp6Lz64bC$c@{|ZE_Yj#~d@>Dwqx*MeG3dDj zT1zcjU#%SLuOgy(N`%%@i;Dae0|D}AznMWVBv4}-O)Sd^ZV89(QU{(Ap~f_tup1w? z2B5b44l*cF0tHd$MDfE0hnDqRcuItVsB_}wUUU_p#UH~N^il#TzbT+~Ru4OnOF!|H z2r0iQ(56*WTpoQf;w6KUB+zaeEFbzh@FyI)XCrw^gm%+l`Oc938qgx&(}fI5mO#&F z(_c5g_c-EPb%v)z=oxMLd!RUj^9;4C)Z0{X9ler3jp&U8^{3tshZbd4N{LV-dLzM) z-miW@i)<|o7?dJ`?5G2|=IZ1t0By74DG{=x4&P)TC-&gdLw~u(-h4PpUyX+MN69VlnC9XDVhT9G_0%Ny~3G6 zsS@ZDeWrFx$ZJP{#;@cl5&A@*sr7s@NDF?WV-Euv^i~2bpb?*IV%w(xt$f5&BD8=; ze1Ej{8v$h7`5uGbNuWE_-I}i359htD^>|8z?of9tHt<{?K-2eqWYBvFw3mLP@frv5 z7S!2=r$lHk{YLR^gRuh{@TJ=3it8v%0!7o%a>={F_HgJ9=kk;YMbpr7?}vUl&>{m9 zT?VB~AmuAW?yvcUBfc7@JS9TPSBNyR8-)cbi%%&Slp%qB{8m#uZ}9^QRE$pZlnDL! zZEu|8ox2!Xbmq@o24zYhCt5OaPqyg;hwfr2Pl=EdEgAUxoeKu&(JWgAeULy8>FynW zE)9D@akF_!gdWn}`=9k;Retb33AlsHv z4EiX6uF~o4vvNy&ICS>fJS9R`>GbY&wfjMUG`6KM=#vC;q$!#mi?=ldsH!bbiI5{r z(R?-Nf{%%Wx4#(lSpvKy?}e#cWHG@E{-KGtI$p+!vwHQG{f!DdUKskGWM zK6R@(K;J8+L})6lwv=xl-VLA|$9gd+M*_{K7H!L$=><^0ah?*P`P3qli8G1;N-3Yn zpf3_=Grd$W!ROR^fbOeI`d> zxx62P{zxDry4BxG*p~%RoQmQ9QX*tTxB8$4KX*WjZqA>^BBc_jjK25edT8XV zP$y7##zJUOD_x!vp@A{??N&Nx39ja+taUJcHK>cZ^ck`d8B zGia9tnngF&u?@@>0L`t-QzA5rZmOSqzrrdD)fJ`;+AV>)(-(iXc2?L2)_KFMESpp@~U~0y? znjhiNrBvf75lW`PRNHO)?gO;e%7{VxB+zRb!ms~b3)g^_Tl16%y`~|&QMc&J0NLN3 z#UK|6G=IVK2Nv}!v~iO`lqQ9LC=4`}a0__yIT0J_<98H2nfP%WA(XydW2IUG8#W;`WAwP>!u@2^QefNF1c zW{{5r+CVc@I(8X*0LrM85}^$=L-nw+CcewU>2)B34oV)(y>4`~|4bAf6H-S32vgOiy_MwCB)A1|60_FQ~iazAqoo zx><*LN`zieck5}bq`Cn0E~;i%aUC6zK$B@yUiIMWM$n?#KY2=oCex^Vz1F~109_rY z%b=qYD58L-{a^fe1JIuFJS9RA1#}B~aSPYGV*(Wn@|8dX=utj&&CRv|sRr|u2o0b| z`K}QUenE>|)aNnim;~~qzIS1BD;&dGYw(l^c~ajy`|fm{3LCb_mO;lQ&<9%8>ae)^ z2WU~(#XKcKA81wU_ZEMg3!3}rD1%N&pdPdv+VIpDe3j*k$2=uMJ!m)dnz{q9VDfLb zCU7`UcxH#(?(iEa4x}*Xlmxm!r2?Jz@4N<(o;Od4 z&;=?L2z(enAE0l)e=+E^1o};zXYagko&nJHVxAJA-?Vvl`$Ly;0JWXbXlKO*dqx7; zP~AxO;azyv)tt#wB4k5#BYMw=S^?yHqZfnxBv2ClMrEH^TVeD$%D*#PQabr+`1Ug84@00!s@z5fp*E}Uc2dVG9Cj1iK8El4JU{Ih0@`#{s za&=F{zV|2vPl=F61Wlv;w!sJA+{2F;6eNM-=+F)6JmD;~sNWHu5}`OcbPq?_%!Whf zUYf(8UI<$X}ikq0zK4GqlY!>@y_H)!0>W9i5XvJ*lT!13GqavyGEJQvMVBJ`e4hOofU2sm^t+K*t+1qoDx zF4%=lOh&+O^rKQrglfZoo6 z$o}(A23?Xs_2>!Pr)WkVw8%D_r$ne8Jz)p-`Hr*1O^2Uk&}9jexgBpk5<* zN`yA$&|=fg88}NkD)<(ILL^YD8#JeScoI%W>!0H(5o&dVUQwR<8mG|?sJ~-Ss07N( zqIb{MS%{yhU7*2JB9xa!-yf{e@jd)T?iQsCx*~yE(4Ar1>x=lN*0~FKN`zX_o#Dyv zzPJn>d!yOziVHSO0)3`=hTCm!-G<-D=O#~y&}W)w@XmjZYZF}?88Rqb0yQmFQyhL= zfU{f28}pP1H7&I_HjLSdgbh36xEx0&}94%!d|zsgx3-Y$_G#KRFiJVfK=Y|ln5ozW8&tl*zIsKWEy!fC{hAt(tO>}@&!KtdNzosL@1Ny z>og5_;)Cy!ZwP~~Ng!qML#)nKe9_jXV>~57%HoGUZn|5bMJLJ<7!)Oel-q%Ib4Ov{ zd&NJV5+UVwU`yXmeF2&_J)c3>C6MyFcu{Y|d%&UVID@A|Ncmm7_e=ZX`cY80=AMe{ z=!OJpok6#tI*Is(NuR4cB|@!f-*NHYNo%1+gInn{=%xfp38kT>dzbUjqIRu$N`z8E z>FZa?iC8^*Wv?lNZb_g$bb5zBEWm}T)Al?iLVM`+ewuL#JCOc)ix_lU0_D;AQEF}z z6KK(%uRJ9}d9;4yP-XauPAcqQ`KrdFYbYAD-&^q;-5ETfxkk16jk_7FTfWM`dezGH z>kO8!w6b2-&C+Jox~?@B!@Zc* z>JUIqHmMAXmq5L!h%fQNWov+JmhqGb^`aua*#&>=17!8EghBTtkRA00SO4gZTN#WV z@stSJQGf9L?wvURweHl^zT*13FM%{@ZSeZe_kPf#DxG;sgfwVvaP2WaEFpQar!RvZ zNT5MfkKej~r;z}i*~?QRG>Gc)^Fal1N7IwGlQN>pa&<^6vwRV z#{g7#mZwDM!3ipp?{?!nKn?4dFzAH@x*A6H_|?Mk30uD|Pl?dgFneQX>*lzDX2kl1 z3`&$hN2pVy+41TZXp!y)o)V!W)Tz<9-)}tBKN=?NVbIIL@E*rivqlUW>f&by%@2Cb z|MP#=KPs0-ohRF^fxl|@8=eyXD&^8>qr_&909rrr4Ew8+B+c(n7v035T1x;LX~a_^ z)SoW8i57d80OWNrnnB4DXdtb$>^FFE6d$J1x~$r(+yfQea1fqrAVMsnx6SoD-jo|hR@_F5h|tWnTWRQan5_) zotBOjSNCfPltkA;>(i!Pphe50c}j$m=vr8E!GABb=zhC_40Hgj0-d6*S8XG&;>^}2L!J_$Q?&I;FD&>s zwCK5a7=zwRpvCmu?-9Gv7h06$!&4%(n4bHOt<+)w`da*gL1_|b53Ru-m>q%JdY+f? zlnCviHQ1!3Kk=>MZqo`FlrDi@P>WoReQ-oRX*y4d&Ro7WY&Wa{uJ1kneV0L>CD3D<2XZoO^8i|Oy_ly&=rPR$S*;n3 z!$XIOSq#dSK;w$k6#d@J#7Z>tNjxP&RfRh^#1(3Dp zHpKxnYTz~oeU(5KRBxg`Ykv-$b*+tfN`x$^-lX|K?Gk|695}+Dd!A9i!S;7WKe+w3Z_28=lH%saOgb!c}j$Wsn4LE zG3YZu!&MvZtGHkbWe_cZSnvF#4bTWRp85}n7C@R7`1}J%e|}E}{g6O@w07imDA@&| zl1eEN@}sq*F&>`-07{IWz@Q=tG=i3|yZC3}e2!`?Pl?b7<;Nv#C+-F)RC^_Zeo7$a zZ`5SMB;1I0sufR(kn%UWqVMJpkohiG2K|yie`zC{UfQzA5-2E3jVhpdAZRUP+`LB$eCm&QkGt488#Sbn9H z2x(-i?Pz<%mTE!g;mQOBe&!AEXq+BQ89lm8YKueeMln5!;$*(?ow-j2` z`2H{k{gpsw)c4km-qZr1vPvltGNZotobZL%XV7W8fI(#v$dm32QC)lDBGso#DG~Cd zJHzj4ooYafBDdQy=$`~Krjy~`YBRihd+gvT5i+Ke!MEc`T$-^>JIH@Y;Gr$k7f?sfNbHehu| zm~SeBc1j@Sr|QDfbj|@3a*U@$NcpKcm5Ex0(4xfR5(e#(KxK4#-*o+H1klM6o)V!l zI=ydP>Xix5`>{=3D=ygG5@;8F$~!plG?u)bGUF)`+C`u8-nKax%X>qE`Z8#b1ZqmN zTm6^%t3r!{gLz7Xn$qmn?D30Lp+!eEr!r`-1RD2{zLuG}-V~tCwRuW}#yzAj)te{c zgvJ}2wG6VCKv$?3HlbiOUPtei@stQ%p<-CuhW&G(MZO6h404b_Bk889>2@23|4*Os zln9Nan`*x;(Wd}v-t97j93@Zy4VE8oXwnQ0-LFb15elHea$;T_2D^ygz9gPbMMD5?p{PH1ojpy*FLB|@X9CdlON1T1;m zHMEvn#dWk#0#&6&+dh-_VZUy&$kedW* zMoTkCc9dR%7F}4*QzF!imS!eg8>0=-h4?)T+Ao3D(1b?g_H{M^6n2lNL}(37XiS-T z3r7=fozF1HT>?F&H&a$0TZH#IFFl?Tp{Mj_N-K4vY-mxF-O&v4kU&r9^bSh?go{m0 z_VAPlJ)zS(Nw=y2K+d1i800B|`qG5P^G-`}_;2-@r$nePO=!ejSQ815$IV}5!y&Y%V)<68Ul3WXv_T-7wiEE^n`YST)A8aZ$XQEc}j$y&@PbV z>Wkk3wEx!t26;=MNV+quD|*ue&bsv9JS9SrbZ7W*#iBhx@62X0$VUQwt)i}&mOi8t zK!M|UN`$^vaiFPAH!OzT5wwv(2PIH_8u4Y$9`+NU^}#$PLiK6HH)`ku8)(ty20jcr zB!LXhQ^B%n>S};o8uF9~8JwpJHbQ4LKx&)97<5f$XWAV{)n*-c(Z$@stSJQ#nWA^S+wUqVB)D zGU%8DDyQdB`|!&+9o@c|r$nfno<}|2?ZO=fYiEsR&~XVA_gGEwYyFJwaOe)t<|z@1 zdu(s~JuZGdwCLmYr3^YDfojvra#QUgtpF;y!BZksn^u-1^>grLD(7^Z7<5tssnE=c zT8|l6M;y|Ir$k7FW=_^kJ3Ad(G{!lAL8l~;?gd&aANcPZw5atyo)RJ53$$N(&t_3rcg27;3KN~*rFhw5}_$njHjnr1$(Og>TMY0FM-<8 zuqD6Y+gbpv)8Hu)YDdGCR!0-D5+>MU2!qZ_Ad_Ee3eOYUIs&wR0Z)mL$uD}3qoL^( zICS%(%o!9Qft-HRirDz)?EzYGou@>|>9@VHuG$eSxiM|DjX{AD$c!H42F5FP0o1H9 zPl=EjJ<7GWeXj#8I<)QxgMuVbZ7MJPY19NK2CdfflnB+P@tX@CvJApx$B+vtz z>ipyTWd}enkMWcUJ)o&hH*?MEaOi^mu3*q*31mbQg9|#J#)8SsWjrN9Ml><_#n>HN z)PII6gF+-wW13T4b#BG~dmhc?DG_Q+bE-EdcEdHkWw*{TC{zOZ&~7TnCd+YRP;r~5 zM97DBQ`Ii;Yy*eRqQyf7U6DX5l&{TxJp4HPM#Eb2lnAY$GToOy_Tu`H#ky<;g-M{- zbW@FgHtsP%dh2;ggkIB4b=^QMynAcDQ1hy|j>08S`bTQf;;DN9`cWw*Lg^oA>8<7w zZ#ZWjKkFEGdaEhk^=se;f7QjaJSF~BJH4qKy6qJFg4}hD1?;bilr;Y=O-c+s zZ~6qFm>N7KLT71GBK~&BO=!_K8#@MFlR!R?)D*L?O|1%0!ZMx`A)iO|k(Re5?*a0A zdYnN~5~%l6`o4{2Z=C7fk-$?T)cdKuaardwd>mKRyUC#I66gp``w#N9_J$U{sFV_+ zBQ))w{l^1OQB9{*2HlWA%IALEp0GOr{i&1^A?0(w&KQqhaOgC0OBi%h0zINbw>Np> zcYt13N{P@TI&|9}X&wcrfl1Q?6<7By36w~+rmu!);vKlKQc8ppsn)dXl-#)h)d}j$ zpxY8CgVunmul#@;{!%KXL@0ySfNBO$xC2nl8dDi`M*`KM8o76ek9fkNdsQhVLUpJ{ z&ST}{5P-Djtz}TO1nR4*uGnNgxB);}l~N+qSJlDTXG$U7f\FepX>ZKl3iWSuIZ z0BwlkDG}OCeX~DHo8YRJY2C{Vij_b==<~Lt%YAXznvNDviO>)Fylr~)`C>S9CQF|( zC{6piyeqDwcnOq4PsxPAYq6uQ+m)w8D2JYs5y27t0ea%vnL+m?(DMYUC|fxn zU*w>+pQl9Vc>-;7>((3B>5OtmG3dSovZ0Bi8Tu3I!lA33$5SF?LlZ~CyDU_L7VR0i zkU>K&_fB-iiYsd?xgets8#?^iBKyV z!iPFs!2YvdtuqXIB!P@+KIhQImm2`m)8r`;GN$>Qg9lx(sP*pBXa+r&Kx^rRD7t#% z_5*alhNnbmExi!s@ygyK;n3B)o5r9g5~!H=9KWsE1ZNTdRZ59aG3`0_H+RR!gkJl9 z40)!~|l_W-n_Z~%jzNuWHMyVluz35SQ1fAEwD<G3 zt#*f2LW=^Y@stR~)57}tMyXhnFd=j!gI-9WxX)^elt!_c06ARYDG`eMY;Sx?ulYl0 z(Z8BL3`&$h=SygZ;JMUl0KKTiQzCS}guaP)u_gXSX)D4Q^il$Cr4{@>B`!EAQFj$j ziO^PB!S^_rfHes!&t5PnNdlcpR8yR<d`2wBkFwT9|>Pk`RI*F9Kq9ler3yJ?1M=aVEX;J@v`QzEpRW~ib9 ztgv(P_IpeqF<&}ApI=tBWdiI6?@>jLxQ0-;6iCXHp#YY8-#exm_zreS4f zukS9nT>Vo<6C zN}_6|E$1)c$a+~_o)V!Xs#ena_u(J3C~I*5gWgJ@^R$9*G~mcvXi@YMo)V$+w1U6& zgquD<&+go1&^rlK^jA%xVx}_{pl8uMB|=4iY13cny($3xX_dvG_Yz1!{kj$*WuE{l zZOv06q@aG?-E)q50IhYbdZ^+$N|QhZw3eDNq^KQ0HcmVxLIt#zx+X5A4?wQpbQqK_ zfx;r_TLe`bh5%&wou@=7EW+M+MDm%&05vun!k`QZG?qG$YW1t(3{{J9JS9S7sRKEC zdG9iS8U~p&C{qGWrS)!?G0%@ei~d$hiO^J<>D6r=r~;6Ct!)hYAc34|<#E#6KRN&{ z*W@V?a;BBXY5Vl10pzvx2!paD&~mz|{=K8(0nln2o)V$ubWblcmjjI zNT5&D-O_s910NH^j_{NSeWLEx{470Od3;j5f$IV)}h7v0A<$WDG?e+7i?D9S6ruiuqvBD-z1Pb z4a8ax*@pYPuB_%M5pt)2m|yzNd^mLUfw&_T*U@(gWI*@2yWK4DEmOOn@stP|(7jHv z$p;skI(2W)paKc>ids}ObzLAFx}iOIN`zifi_%8b)PWZLJur+xg%aouEo8i_mNW{W zL~ouFp*OUUp>uKKW`JrISup5_1X8C%H?HR-9F@PWloBCzI&>x(gF6EBb(|f8iX>1b zRsEUiYWIN_#f|4F5z3^hzg8zsHUp?<=y3-9lt95$$M{kaHWr{7S9nT^9Yli^e3&=i3DU-6U(Inv26x46d%fL8ZvdbHw#Es;Rk zH2goN{i_Nv`ZDN`1j?k@t)an_`vdf$Qc8p}X?Dy0QrA%c zIpj`ZP^ko3T}stT;l9lPTAs&KBDA`cUZ?ZrSt>wklh!imuLN3iooco(Ea?SMQl*p# zt+`I~49{|Z19aq)2ZPEaP$WGjVhXd<0NQk!r$i`{9uv89zTte`TFuK0`X_-rX}x>F z)b2O~Ij1&HiI69)ci-u38U-yXTlthhuHxf)8x@$H- zpDU$ANcl#B{3LCBOx$tk%%EKoD37YphkqaR4j@lQo)V!vszRSQWa-gkBSLy2^&O096mz z!=Swq=rr9_7d!c2zwS+?ln9-sn`)E&dwT*@R{acv>?Ke~IvGa(+<_0iU#dJMLLKR3 z*j(#lD}a7jMKj1j0)^5vn#+)FcE2IckfQ`zN-vrA8j?K} z&bn`vQX;gJUNSw`^3G|1Ubp|pASVfAMfbWH^DPnqO6tH&sZ*DDOoyuLp+zbacuIuSsY^UykxMW@lg@8skgEhzzErS$ zrtMjPI$q!@5mLTXP<6ve>~77iDS&6_pu8ghx+>7ByHM z#-RNYD4Is)#x9n@(4u#hQX&*hqjDRUXeWSFpTA&`y981ec?4!>$Dk@{mByzS8?-`sCjOsC74<5}{^aX|<)LI!-!m-e324#dYKOu=ucZZzDZ6gEX?lS0*1iDCD5~9wTe+H<*U7ix5i?k)-@|zym_a4(ai$RAa zP*o-6hPl$52z{;H#r<{Q#&I4dIf{dN}wgQ41GT%2m5BN-td$NEum%Tsz2Or0o1p@D}w?g(0Lk% z7HFNs3TO2JJS9TsX&icdlw$xu7rf6gC{O}vQ<-j}VWc6P3|oA7N`$njOn3If;JE+| z{qcZ7K@#XG{YE#JZ@>k?4n;gALQm;8npU_5Hyllvk zrBh|5nU6zfXwjf+YNsl$qjM5y{!7{}+OZ9u-nFB6N`&UWq|eUS+^zwTW#je?Ixm3+ z(GcFZk;4}Fjruj=DG?e(L-;hEA{^svTl0T(-FH~e-}?t}LXr_OvLhiQJ5-diB7{On zW|5Jd5t8;GA|#*2h5jWh0Z-l_!h z<3JsiFn67-fa|X|ounxi>VTtK4h@yMhFY}b({2Ji=Ro~o*G$BQlXz3u^n99Pq5iOI zMrvHY)d<=$--JN^9B350y?v+O$IsD2IhtajQSkQ8zZZ`u$@g7%B~Snd>IG9_HoMaH zqrZ`Z15L3|FPI8@XzGjSuE$kJ5h#!Y&4mZHp~ z4OeFN6p?W2d|-n)Pzy{cf0dH@l0=?uwXJFcY@Z>^#I>(kqnqr|dFzv7TQr8Bx$Y5PMfg(82NElh$f8C9{ zTl3e`6bp@nk@Y*f4u1rdJgX*9BnPU84X{5Ib9bVfA;*uVSg0O0z#7~!@j#HLNMHBP z=O~H;i9%V_l!~~%2=eJeQ!FG3Wl>8untnvk_kA)1isnFD;9aMywGtN|hwrB;7TN;u zx&k}TsR()+wUR(F9B4C)_{NU#+ksl-A5Bv%v>8TxR|e|hdC{Sx^avEofzHC;NYZ&w z9|VTXb&=d>B z!Wz(xjcPflMNLLV1bWSZ+TjaZ|3MK?(G;DfDHdvnFYL&X9e8rD|EJpodc%R5pwMe_ z;|09FP~RoRLQPQU_3>4G0&3CLMZpA0pXa(nm#e~)I%(-aFm1;|}ee+Gh*?`07v zjRUnppW#FKPd)TE%66tH7HWk)L&?&gc$co|-+BVQ$AOw*`7B`Qz9a+<*-ld|)C|jK9eWSWLYMA+qiBkSjA8Ej?(*=p2$CPAPM{19lmqh&s}7FEm6?M_(-aHkz&yiP(|fq5>U`dS zK<_z_2RuhMAt&(L`{D(fVj&NBj=Gn6;9<+Z0xJS#av(Xl>#97Sm!M1cwUDM*NDl5g z&0+q_(4~8+>P4U|4pbThWx8tpao_vFGMZwc(jeFfbG+^_f+jslB2YF5DuEWgI^((< zwdhBe6bqF=i=;og;Ja=?w=V?B;Xt$D!{?r4nSh`z-D!%2X2XYXZ@d$(tG}-+ZfT|u?h>0UaVAhc z2kHjzy5Z3+QxSApiKbYn8@%hz+s5FssHl741S;S_t*{+9@wCJy1lc;%6brS&cHrsL z%skP}aQsg$feJZLDohMsp7vS-K^xj=iiJ{PVlcqq0lswmw>A@~hyzK$>hP@w%}fOC z(55LCl7Q9W?K7p$pcX|$iFui!5or)m2p>qgS z!hz24dQe4$GkyExYbZpmGjW1p~1!5(u+gqAI-#L&G+zjhw zXTCwu*Dfg*GJ>1o_Oxwbs70dtR0&kUfhNEd&Fc1mHUw36NwLrbn4(D-S&Ju~KF01Q z&<_q|1*`u`x}V(;6!(gzSjY-i|IbKUMWGh?OPLU;k^^;vq2-yHe*p+`ok&wG)D4D~ zZWBXr;lv&bR{~XWpd6@H8qsZ1B7)Xg(i98jK(*3E$rV7B2e} zL5pf=iiNJg@PFCJt;GoHP%9u%4F{SGFIDL=iQf=ZzM7_3XfnK1Pp=w!2tkRytpuv& zKz{|N7_C>?gCLJ*G{r)H;S{5vDMEM}&AXq3cjp6J$APq9%l34~Uj7I=-k+veNDH=X z*Bfc#Mfv!Has;a9KrwLXM1MHkLXd+2O|ei6T)O(6?RcT;C|U7 zEkL*T+V$xKYUV(jU_wLI@rWXV3^izqg*L&2#?vL;aOv9FplSmB;y|;L;DHSiDn(Fs zFio-0>?Anm#e78(YLV)QzCNAL(QghU77u5k>m_&~X!b~&Vj;13nD;L5!z~K9B151S z4rCh$KkjBc_A!Fqn9>vr*#^QG_RUSaG4p%LN&>ZVAZPf=jPM4Vn+Qt#Oj9i63_qD+ zH~7?dbm{z*^a%8a0}Y3|`k@zZ&OnffGEK42aHy-lsO*Plx5`}12-L=b1aEH{z1Mh_ z*!KZVv5?^HeZgu>5Nc6giyMLdav(P-Gbl9wgx?Gytu)0#Zct`Wv;W6KRL3}6D~>?_ za2;c8K%t`i*e|#?UVA6~$NyQ!CXU?w0-+fVCnqvP~a$@g#soyevs7J3Ym631?9y+BauOl<<0a-hC2&9c#DI35xB&Y~$6>I>5>qdqyE zMo@{J5rM99AVK}rj-B@FP>a&+X^Mpe^;d3lj~+yjV*PCbnQ@>7SXg&Uyx5AMNewi` zLJhF6ex&%QF@pZB4ks#Jy`C|D(s1qafF*^doQ z1B+3MRzIgH7Se^;4{omf&@Gz1wM@ z*t5>($ch6kfEKlkc%p%zOZ#bxg%&`I9&gF$gIc7VI)gyg9LP3CSo!ptW%%|Ultxo5 zWE%s&>hfJ2wj{5bfG{r&&Fjt@$Ao~`zXs4wCfowTY2VA<3 z8}{QJ8MCcuiiJAh(s>t1^h412GAjbvaiHuvVdWnkArH{ssJ=^zg|h45_yQ?qJcM_W z_acxz2TE-dRz5ma2%k}QL4l@ND76jNyM07MP>UAaOCr!U4x|IqEQ9h-;o+gEGflCO z4otJ$zBC3`S+up36X-ez+5%HyCTZ$;(YD(knqr|XFclV|@2iViG-;cVU*~h=z<~q@ zZ7s0z#RoF{?2=+3!9iOo{%6J_NGwc}KsPv0G`zk2B^~kbu%k2l@jKY~$0CKy=rw979tq^amc;su9OUQH!=2>k!D11D%83W<%;Z+-I0_k)~Ma z9P~DGkIuq71t#QQAkZxiR1GU84b^HLs6_({Xo`iZVZ|iQH@hCSNKM|EK({$i81xx# ziWgTRNL+!YSSSqo3@7TN@mxWvb2x#VIFLQ8r5Xu0;%(P)_i2iS>|rgn|HgA`P>ZyL z@(6T?1C0-bw|78~HK;{vd(adMjSq#j)YYE`BFI~>nLu|rkOK4>-sius#Zzf<`e{} zz83cHe2yM-p!!BZZO`hS2pau{rdX)H(LyE5eGu-PES)4xAU6&q4&AMgF&%LTS~8iY zSV$bYTZ_u2>`;qN+o}@CodbEm43(X@44#zOZ%0!sKpq@O2?bFWSvy{$T`} z`@0gzlLJXWPt|$9A+BRI44^3%l7OD-;frTt5acEnO&~80^a{SP)q6JMADB2XkfvDZ z6?|ds1|Ks+kj$Y10zKtGyI>7SapRZ+=+aF*Oj9hh3)X->jnc=Z)DbDI1oGxU0no2| zsc3Z;L5Z(viiHB8U)Rs-JbrsiO_m7gd|-VzkS`QW?kugv3VXo`h=p6btQ!D)eS^KU`7PUaCo;XB_AWjQEme{In3HS58wb z^aMtH<~~Mo=w`@OK1m=y4s=*>mc^-q_^zv3N>eO!7>ekZysfz8CQCGcK>i#^7&`d{d6O=qOZT!bO|g(Lbn?&Nx`+F9 z8V2bE3gAF>@HaY=I20EV$seLA7OI24(Rgp=CFs%#y{#rtAP3Tg8OYk~-gt^8t4oT7 zbYTYaOLv>C2&$XZH?Z?LdclF#!N$xfAu&JDrE8f?Q!KO&HfA3Bw+najdt8$tP!I>2 z9u9ABO=md-8D6I;7MdOovs=Ia?ME$o*SwNI!5l~vp!WImaNpbK7frE{CO`qM5la!Y zYLgy;LO9SzSjcD_csviaXs9MlvCv0Y$XIE1Xbyt@_?rMFMJ!+~DWGLoNpfC<3C{PKzBZ_}*#BCr=v5=rZCGvtjE>KB08AqUS4m1Mx5a}h} z?}aX%_bHlUp%Ji$=y{^rMAV}E>=FXKT5j&2iFb1-f1oKA5**d?;e3!hg39N& z6DWcM=|I7xS@sz()S^&1nqnaxD46Vce@p~H4{b)j=zL%!Ign4^-pcD-i=_~B&X%TF z$fvKR%A$Eg@&1%`U*riC#esC2;J~`yP0tWC_A5=XkZu!<%1^1_La&yk+60Q`K(R2P zu}5238$l&1G{r)(FrlIS%^ugtjdV96Pz(nWloy7|U$sS0YnK!Y3Car#{bha8-)L9+ zZ34w|AUW9QmAdeMjy#&)K~pRw2m8GGdQIqtAUmyK0=?oui{OV2g(qeYM=jFbOj9hh zsGFsVlw9y71Z4+h5h#uW1;e{es>A{p<7EcZ6bl8zyDopqBs{eA8PY(Ycn;JC59~;@ ziFk{g=TMqrp*DD6z1IalMlBkBQY@(RIZEI_f)l;o_a8JHUAh^kXo`gdCwdJx$vcFg z;kh#il*oZhp!mV?y0_JgKasQ9mi%JI~c`=~`~s$K+o&4Dstd{m||3a{}ETSikXlmX+TgV#0RBB=Y5 zBm%wRK!Sovef9L2s6`*Tq*zE$FgdL>5x>2odVL{KDhHYi@468y!k!@Lu`o@s&|G-e zO%FZ(1GVVj5uxDD=O~Q>wZrVz*S)?&5VY|qO|eir%x=ZG>b^oyUWO!r-g2O*cf!i! zlP``yP{(_kVxg#aP{(-jAHKcs%~c}MI}W7t8s6T=-rxxh%Xu`#LOQSE2oaNalTeGs zI_VH7odc;siwu2KexVi}xIlG{r)8 z@KP1(__PhRXs)g^f!=eVFHl6}J<|$L(TMJ%DHi$yMMPN+qn!|>7ZpyROb(<7Wd`2e z_u?_EN;FNekRp^B*qlCf1VJZ;<`F211MU9}gxbs zCQvp9s)4t+iGBw@ZgkFRnqr|Eczd^tC~rZ~o?P*e&IdM!19?FS=moKp1?Xm&l}A%7 z0ZF-oD zcj@jtNmDE&3(wK6p7(G8|Dt!op`FiBAqP@~nUmhhzwkCj({!3*Aw`%u(Wvyqi}Fn~ zr3qBTfdq$h$_B*=qf3`Ii>6pea5(1%rJ#%GW{9^{B~UR35}XtG_Wa;P1l_TtDHal( z6IlK2fgXY+N_P{egaf_n(_7j7j^;@OHFQa_(7QgCDi8J7Iifno_ev%N`n(jK$2jl( z@5L{soRdTUD=RhnkN>ldaW5=nY(MJy9R06Wucj&Xf3+7DGH%DuT!x^#KCa|{Rmy38 zU+9}T%NDIg&~9IvVxhj!H%s2}umnLf`$Q9{i~|jZLo-TJ^j;!pgeXn1&|o+;W6$I6 zxbFJpp#lPxbD&4?Gfb{_qO}OJJ4{n7^ay^2Y4QM-cc?|D(^?7ig#*opZ_4>Cb!iCF zc}r6)G#|bxC!}8BLr~*qNQ8Ah-CsG-bePZ49yZekL4mR~#X{3zK1Z!;W&~={GkZA# zed9p8;Ki7g@_s6Uu3w`m7TN_bM(2w!M<7V-yC#9YbD+u1u(L@?GZsPLyQElXa z!<5VC5u~7Ul0X$4=m@+He5G1ThQ!I1@UL)g^=ei>(>cI^H{op{FQ2da7JQDZK zq#n`~3u!{}gTm&&xLQf)uRnn*InXEgO5W2>!Ry^>|7ePZKEYS=*_%6feL;L*I)SP< z&?P8Pu^VTFtCd>1q*&+@6sSB&(RzHMD46$@pXFZ`9Z&#X{E5qAP2! z4M0$mWZ&@4=ctAQ&4KmqLaQ}+!0R!drdVhWtapnportHDGp%I^RLg;Kg5YoDm~<7j zDA|UlSSTk5s{Q{=b3rXST)&b)bsXq+7*zYqFB*uTB@HyiLbtR`~84=bmUsEeZ%RBTxee3WLR_1Im(~2)Y|eQ!Eq)i%kPN zigFQjbEG?gesZ7{iSVl~l})&2TXz&qvCxV{3za#w-|&K9&x>&cYUDsZFtVPkb+;dC z(XLB0#X>$XvK}L&frp3sg(U=P;y{9vhz#2n<7t-pMKr}if|H0=PPVy?E?vl?b^W8Po=G>tv7HWpU zl>eqrc$-^lt2}{zb0B-Ty=~7c;911_KQzTc_HcVwc9sN{RvyWE5jWpjHm_4Gv^*95@hHW^Rb3DHi$$2QoTJIUa<1m zDm?Qgf_#g!2=tc&wZo`Q@tvFwYEfJXO|eirjM{8&onDEcjq(iy`p1F(z!s`MY5I7E z>YxHmvCtpbLbX(G(lZ2kIg3SfK1b~wXd%qG?fSg;D{7I|eVSsSg)rw<>32H@L1%u> zAW#PflKv;Ge6aE2Bm^lm(i96x|FclB&X&NXYyNB031p)3-a4#{X!sf4wP0etX#Zzc6S8DPopUo zN~^a}*_eIg34$Kl2t{^2M;07t4UG6IhP=Q_oL6jViiOs|h|fRckST)t)JhV_k^`l{ zXTo;nO;gmO9(6RuLMiZ>==aDUPy2hVQ6i8P2g-#{d5mrzo)ldoxo1toa9)avRP$^WxI2b&|t6HTKXo`hOp%NzS-rdEhMM0*` z1iHq7X2P3cbwV|+zp}ncQ!F$S-V7s+UgAaDxDxTG&Ik572P%Z6nSei~_2|+$ex@lF zDuksOE8nqaQHzEwokJi84rB){I^v)qilCkMMSeg(YEjXYR07@RKuOSnJolqD4?%u1G{r(m(1Gk`w|WVJlCM<|$cY0jh!Iv^ zl+j}Yf^J@?DHd7~W1+HT>xb(In%5v4-T55d;XpDlDsL2d>xUqTpESinGB7INU%hQT zf(+M76X-4ndJLEDinz)g1leiO6bn6uOLyj0BtAg3^_ePx?r|VhxV<+-nqug!A5F24 zD%{>3Xp-bK)O1&Nr>2lf#MYG{C; zZhV(^13@)iQY_Ta0LwZqRXGUqx+O=T#~f(eFJa}BHE!()x^$bSSZLcX3zf2hKX~{b z)S^isHx5(_os+HIFAYMFb1O};P%U&$E?*OGLoEu~dXhly9LNzi&;C5r3-5{=yp5(< z$PqTr%9KyTd+};s-5}5t4m2F9mCOf=jYKV~kE1CT8V=P;GEZmYa*iou0tn>6f!g5q z?yt8BZ%I%dOH(Y=2Di7=_a#%&r3*GqCy*xx8WRhr+-%7|jap=Pm8Mu|Oe_qR-2yiv zXje%!fxI|S6if{67`fLOK}w%#iiM(JVsKEM=12t9EA)-+e2$)SpqX&#%qN}4v&1oq zG{r(Q;nMjkhu|ZR>|A6BHelE7E*zi>bBKSzMxC@NL!CUz8pv%#<1)EMf5<>g>5v&Lh>+%HDC2V z<%PA8W(0c1foh>k?BO>6uOB5v(G&~SLYMfTc^+#tQ6#Er6@rRyh{bh2M=v?h z3gO<$ccPu~ida8Knqr|9!j>v=9?#w&sJeazfg(6i8*H9E(wu^)(Ox&u6brS%=Go*~ z%e4?Rd%Ze=A~}$@pa7{cO#}Un`fAV=3u(jOD0Rd)H3WU~G9XYC2g-$m0xX20bP*Kw zl%`lH7Y+(ow%2DLf{c4x6DXPkO@`^{rnIa_2-+n=Q!F$YrlaEuBQy|HrsqYV7!K45 zUE<^FyRRb1Tc4&_s1>@zTV&ThLD0LHBm%{9pf&Ko4*QcCfFPe(nqr|f@W7@Q9>Q~~ z=f-~_&?^o!99}^INq5I1X!!)1Vxi&i3R+WRdI`0t`I1n4=W`UtfgZs;!;CSL>=2Z5 znWk9i5zI3z`*Qt%mJCWH2^7zPGGP7aO7VrQ2>R0{#X=dde&qBxG#%A3N~wMBrhy1S+ zIn57(N%D=}7w#cQ^*K$kP!LR#cU*XZM+E!(851ap19gWPDxq#Ixb8ZuKTWYvcbK8- zqdp#Q&-rlBnLx=LXcwIR9bNDj?>_S|peYvG1*d;s>+ryJBLykp1WMsRvQQTFd+ywU z=+Z^KrYRPZg|aB&cYg2C-{|nfJOaJuK;kfJtFuhofuN<*G{r*VFluYhjgLoA?3HE$ zz2QKDW2xjuhvL1%^`-%Mw?VHI&m73n@Xj zR&(&-KM2ZFn?s;94s;)i@NbP;f+t}8SJM;=-G?IlhXxCOA;`;1gFtUN&;_{beqLUM zx92!Kr70G=0C(L7wU6@>g{PE=BqRFEc6 zHV4v#MO%re75J_*Qlu#s(u74@gUZIW2+F#zN}wDLGzYp&H9>lr==OGVp(z%c16`($ z)O_53UeU6fKp!~JH#qRw!Yt4qLBm^ViiN(xfzQ7N>{diA(%WJ}pj-}=1s}eILe+Q# zt=&pfER+QwzLuwBKOrbO+Lb_g9B2mo(!t<0Q*l{TNeoS~&j>8Pq`^Sa- zQHz2#Xo`i7!wgmN&#Vvxg#^kGsE`9)gT)tHQ+W!Gtnh00*M@=a6S3Uujoe^e8w zj05@B2`m4~J&F&0TwF<0EaY2fp(5+J4sYCRQ0trA`5cvVppEc_eZC_s2DPYkHBGV5 zM)<-W-%*TjhLb)r1p2~(?!&xjqoX4(m{j+rDHgg9^P+}QqZXn|H&SF3fxdE}AQ(-o z9{e3|RvOfYrdTKlMiVZs<8V#F7JWSeed9oa-ye+nrId{>ozwxEVj;or51w73gI}td ziDm@)&VkAWqY2x~xVx2N@=E-L6_R!*Pz47v{|G-|z1R$wyU(3OQ!Hfu z5$b_c$Ki8(MqG{~&<_rj30>lvzN2t$kB|vXu}~&-iQm*1S)-fbOlb*$DmhRZyi^rN zTr@|2qXT6$#X@QDQcZI9I)|X6%i0N4#eufNXyT9d&hrSGznrF6XgiE1My5L8lAAql zqf??ZZjiCAjl}fh(L85=wTIX zBrh8%j3ARpnqr}cRTe5Ewmra0GfktN2vpC3R={1?{kkzO=NLMcrdVhN+;t~H^QWN} z>6!);sDT58yn+K6w!7l3@oTTr6bpsCg01nU%QX=+?`sx;esUlUn66BlP<9txx)I-K ziiI>_y7I1~IqrMMEo&f9BM0gY-y_G73IzzdzMQ65s5g9%Za$E~)k^a`#9ntkM@<~4 z8fH!gEWX(jLH#{xiiN6S=H!RRUU_usl!Rmn)Xae%zym9>FQODdVm)Yzg&x2IEAgxn z&jmf)txlj{97yn!8PP{Pf)I3e4^6R<;3qSNO5CeLExH$FK%n0oC<$O z-z+X0+`Jk=SzS^rG#!4kc=v&!_-`~lJBdJlI8YH3Kg4Cdu|m-AE-4l&g5rm3j{Y{N zMMI{3Ay69!>L=7&SwG^*KLoXRNwH8rAxjmJ>w(`8^wmV@P3Lp;mjlJXscVfFa^4~+ z_zF$2Pz;>9c43x0o<`}s;r1Zw9% zSD}b#>6};HQHy#nr70G=3PnWmNt+L#OZVJWhd>=1Xco)`-9K^|zw0hPq$w7f1#>|m zBYWdv%h=`%1TsZCmQ%TEEi|V_iSV+HzrOLY93kpz+zHV|R&=n5k34QO) z?&m)t=(8qGv5+V9y-Rx@G(=E*LO6j;IgphDOx~8y#l?8Ri8RGRRt`|DBzMgLK}nPH z2y~SLJ%o~*(Jv-lLM_UhOj9iM5K3q_Xb^lR ze*DcCkDztxG{r)L;4|UB*!>EE=6Glj$dUt1Y7$n~Il5gRL4UfWSZGocJg}Beg%I?- z{V0K~I8ZoDXxvGRFGEnTZifG#Vxe%D(2&=Cgm3T5y7nYu&4FZ~b5i%u8n5x~+(lC? zBmN0VJSYi0~wSh zoYwgq*>j-FFh0r|Wrly_COn&_Sm-j0j~QFWxUXOpc~l zNbm!E0VZ>c(4~vNp-Q0Z97u4ETz>Y^2z2SL-=rxP5}YGfvh&6m1W7gQCXfRM8V)yu z-Z)WQ9NzCIO|j5$xEa=u9orW{OE;Pj=mrOJg`Yqb89D&3{!4D6DHd{tpFm9>lc9>B z#lEfty2*i*pyb9V*ESzry5Y}giiMP*}dYStWb+7ZB(c2f77+qo(}kWCYFBqbU};1%D&6lppiarSnT@CD3gSGzC6< zFVZY=fy$Lcnqr|T@ZoDNJC6@t-#tO%ZRZ2)#DTs+cPna2*d)}V`BF5+Lf@dfCH7Sf z@AJB9CP$z<94HDN*yY{t?nBUibDCnID0pCp6g%Qs;;BD033QhOmBhmrR^vSGss8Pf zVxf|F_$lFK0hZ`*G(qJgf$ni2!#{AI%qiD&)S^+UG{r)Oe=JmTW~{M8P=NOh0y%Ra z-%w#?7s;bj5M=2?Q!L~g3g^kZ8RLwg^}PcKbe{tWe(A4V?yNV0wusOa3kiPdPp)B# z0)lqxr4z`71DV38-0p23{J_fU(-aGt!l-;q+httT8kJN{pa&eN9iTC((j!re?j+L` z3$+7ur(z18(6E*2`>yjja^*k?@W7f?Z#{t^{fRWiLJ9D|P7o`?#o>mgG6Z_afxKWb zG0mk0PtnL&QV2#fxy9e&%-}j{{ z7P5miKIc|*ysiGg0CxhpbD(!nh5p=lv@NiAGCz zq84TTrYRQ6g*Bj>%d7B+;N})>0(oT5wp>yyDdpmEHoBs zE~=N#xq9vlUqbNd2z$$%= z{)r&{WSU~3LYRQPo^k*$2r7-YCQuLuGKXpZZJP(YM9`KAG{r*ZFzqi=ZH^BJ3cch- zpkNMU2P5m;GmCNmxw{EXv5*~%tW%!5??adFV^I=;LO9S(c#aNhzcoZH3M-~57P<+~ z(d5O2BN4Q0(H8=Rav;IRy`g^Ife4x*Pg5);*toaKJbMO$4mu0H?|hEJIFK#W)xYc3 zZzY1X?$ZI22rBH7Vxia2tzD8DjQh`ZT1o_Z z$$|V~=y9=63a*5CxtXR|$RCCt*MCdl@6o1bIs}T~Kv6L5A2ByK6SZipA5F1P6ioYP z1*PD=*>4eJ0!4D5m#|4WP<$JP-u9s>7J3Prl!tuWj0;pecR3R%iUTe9C#O%h+a z{B8U!s}I)^}U97wRA&UUVq zCAxGYRcMNZ1pDcJoj>4>pq&pj2o%qO(x5o}$AdSO2vT@NQ!JDQ#o>xAhIrA|O~{Zy z2^`25`rcm-eZo6!Z}y-m7V?F@_cHB;c;#`$K6?Tsav&2JOgS8yavim3@_w3PArlx( z4fxqj3SGLBNxlS1;y@Pg96b)|Fhh_|GEK3N1w2PN+gfrEbZBxafs#2;33N`Tt=NRu z7q(8JDHbY$&PmF?Ls1AayHY`*6b|$T4mP=W?ZYnAA}v#zVxc#1u!-2=@>~Sj6bol{ zK1Z)PkS#n%BZg)+B1pG{rdY@po}esiJUZn-f*BJZNkdIeI`pI=;{KRVxc2# zF#F-R!W}^scU1|L%7FwM$v;L!;;FExdo;yDf{o<1FE4i>=+2+r1WMyT6QE1H<8KSz z6?L$UrdVhKbcxNi*Y!Y9#by%%z2!hhVK#o%p7cNH(jC}JQ!I28X5%L=aEd`t%S%@R zz2iX3V47uluTgj^EINXwSZEncv+QkEh(*xK5zz!n=Rhtnizu^hBfje-M$!}uxxg&q z3Xxs-3u|IjK%fi`BnrJCr>8gZ-UqF-G{r)q&jy^IH9&kI`#^qvDvNPt??ux?T4 zZ&X=CQ!F$g0gjl+FR4S9u6c<>cIN||$${3w*7!pAVi5!tDA5!Pt%a@eamKR-A!v@P z9D%YpP&Ukqe%H{#TTOdCq$w84hI!Gce{Aj`$m5SDfwDQ!Hh2Z?&pIxOT6C_BrdVhj zyn^O#8hQdj-dj%+D2D@mg)i(!H7mSg@=KehSm-N!VWrZ;mLkYJ>;{28a3H~t7x?%F z;4!RSI8CvT;KvJwRn5aI_+=6S1j^+=g5S2&6x)Yq(zm1v5EB4Os_;x7}tzA)KUhCulos21jec0ap_ zJNfG$(i98T!d%dNS7}@-@TqMjfeJW~GQ8_nwN1vmqF()_DHc+Ocin@5tMR3i(AFbR zAqUEbg$yBsDEx&Typ5(GeJIM8~Sfm}4-XC3+*iH6e@3$2G4 z$c-nL<6e-@2zLS%bD%3wc%1w|ZzzH;j-)9Tx&npA|1{fOP>Tkgiz8482hxHM->~x= zDiHLgONxcG;KSE{P!!(h^(3!^K%Y6#Zg>UBCurk2)gvEiiiLK=D`@4+wYZde#N2iQ zm2#l*u!pF(%mq9b)MFk^vCw$fL*#P0u^3&t1-C|j=zL(yIFKT&YB_(1e~d2Oq}w#b zLW+W&0+&xNLeS_Ic> zoCx%d0~Nqb{$b-7d~T0Tf0|;U0+`8vzbGvNwdm}DU;=&TKsTXF>=N7upY^xrAWgB* zP3RKOzVj0Iz11_a2vos=en-Ltd-Gjc^f#LSo~BsncO-1GcvEVRT2wo`fj~bvP(K*9 z926DC(8oD6#X|jH*pl!~-3dW!9mR4xpQB0+^Z`cY`=>s@hjUK2MN=&F0Y>HVPL8J$ zw7qEtfvPxAAAtPUzV$+vZe}x0u}~j?#^{9Ng(|zP>IAChK!)(ZiXQBTPmR#hrYROO zga`Ig?Qnd|%hX^40@ZM!Ah^AI?AUSxwdhd@O|eiA+}=?ShTs*k5hJY$RLg<#U~R&? zWY84^HFim{P#&yJ=q^5v>*}q}dJ(9O1C_yNLenfCS7z=wM^h|R2A_#PH@@O6a+5wL z5vZO6Errp<+acK=s6`t-(G&|Uh0#QLt(!Kw>#7vL5U7CzX~YRD54?W77lINO(-aG7 z#KEWB{R=K#GjSHm>wJ!Wa-bW~IoVYH<067K-=`@Sx&fV&!II^JQHz$fNfM}$1D$~m z^s4O|eifIB?@oVNxZ6Ue3=WP%8%#tnuAYs`N)UgN+`c!f0Z5sSzHw{H&rW7BYmC zPg5)u0c8eP^*`a;JEZ>}0$FgNC4y0Tkr+OG#bE$VvCtAhcS~*KE7T$dIYs~NwJVDKngQwbw|+9{#pdO#)06423@A1`y~52daU|J@>>+eCfuGrzsYyfyuq$qh^jrm+s}obOPPvK$Y+s zNs=h9LYMCPC7NQPN_dSNjTcTw(1}mg1ajm+@^B1Dn9csf2wIm2$pYA#$%kII+|jk=`bQN)jd21L7JQN2y}-7eSjhS ztQN7&s6`VsX^MqDz!1K4lW;!-nFg5==q?Acfx*;+vhNf7k z7gWL|K4`+*MKjY&2y~wVt%ln>ii;-Vj`WrMO0$Ko2?43@GQg6gm=5S4ybU z6bsFOa*hYtA||LsxzCIU^oRomKqbtl;8Y{jqF6tgVxa)2gekvbiEB;oi#ZYKF$X#f z#SgEhp9n#a=|Gxdp~F!8aB9l=d8kFH$ASsu#(@MsMtS?3OgVz?9;Yc568spY+7ETS ze0DNDi$Lxi=oc*7W>y5^j{4RNnqr|}uxK09*t7?==| zHT7P(7D1;oXo`g%!5UCnlJi3Z9hv`yK+ibPoR{#cmjjD&i!|hDiiPI9gy~APwM7WB zxi3`Q`5gIipcm0lYw8qZiduBkg{D~OMYM&=-w*pPAn43LNdi6RKrI=<%8@TOenimn zcA8?LmJAD(84FKtLD0kPN(A!fK%3!Rw^P>i>ckQ%H3&6W3m>AInaBU+*3C-JcU|h^pB=k=sisC zjc-!FjqbYFyXO!nj03sB>TqV>_$LVR*+Ww-_vJjryZS*XSXIFrzsY)7qL{CJKGw6k0O(N2^7hJj`i-Xyj5B;1+~a3nWk9i zSZ_;}5#K&5q8448kV>E^4kQW_8ik9NTtkqd6iu;^C`@QbEgX&aOI|dqAW$?1YDp7T zp6jNJO9iyeX^Mqf(kxUgdo@i(E!tNu{JHZvis3-U@Ek2tDJ@1VTK$ElSjZTjqfhg$ z;8KCZE2Rk(%Yg(VK7%7Sy%4l<6-}{_V8l1#YmYUkMYFtA3G|8sdBJmZYH7oG1dVt~ zQ!L~K&(ZBP|Jexg@3n_OaU3WEDx52{uRTCev@lJvPzF>uw?AJ|il7(+69UC^pir2i znXqOZuB-PtL{lsj3R5(PBXdd+WSZ$ppac#S2lZDkwoBtVRsSrSVxc&wzd9coRga)s z3!@2?$bp7JpJ8!RGwwj@EutwF8VY@et!3>a5oF|8K%gWJbPwiKWmD25(cehz7EQ6x zJ(yEHR{Q-af)@X5B~UU4+6Om-=lTgw7;2;`7TPDc8SZVxM_H~~FHze0z@~7ZNLWiv zZ|XS%LE|-OiiIL!Ew$;e;~3PUgMM-Zdd-2>LZ4x7j)xV z!KeJ|qC&iFLYJeE4h&MerJ5xC2eGkOh4B9<_`5B4}=7 zHG$GO&>bia-*vM4chsU8O*F+qcc3_YQ~1(}2pXu_x2*Fy%HTk1Fe-m+@?{o+B(!LX zh16hFzAQQ<7(rp7G6Z_hfv!P|lwEG)D)hK8nqr}A(4vp$yWwS>)#9rNl*xgX!a~*Q zPgN4AMJtEU6bmhdg{mh7vbej|aZHauSsW+{-VAy*Z`LCy<2X&RP!hZus_qUmMVGEU z-Hbrl9HzHRU$Aiw5SqZYMvNwJV`8|>2k-1Ztl5rriL%H=>Ra9Ffo z*{~4^dQe1DETjU5MSG7=!Bu~&m$VZoj{`-+bacy_06dc~qeN3I6b;kS?#<09s6}3n zMwfR!upc?lODMT{)z&K)wW#7TO|j5RD7m>cW&|#n+}%xqK%Y2}6fC{H+_?HZf|R<` z6bnhg(%YHQQTSb_v|F1%`5Z_SNrkS3c2+K!qHr3}&~MtX?37pkis7Vxcma-ReK+H?GhS zvj`?o5eM1^Q=OCcufBqyZ(UL>v<;>@iyR%#p}S6@Jc~fZ9Ox@dN1K~F6(DHb7n)+B zuP_}w;ic4L1X(R4Xo`scB-EnrXky5#wuJ~P?2=-kN*GPZW<=r%jnn772=tW$Nkeh? z9r0GlI^}q$w6UR1LSc;Rg&&x*_zn^Es;EKshj>@jb;KZ)2?Ol47A8 zn9$ISm>7;)6kH`qpdTEl1J?NbAGYHz@!e{gVxbOLuHLG1XDDYD-XU!(B2n11ghdd9jU^~Yu!}ws9Yn6rdX&03iz$22H!z-jHd?} z6Q~;3G5T%GS4`FnyoCN&@e=eO|7RWJ0T@ihD!#@aFQXAO#s04jz+mdfnm0EP^xD{& z{I6;_%^v`z0(IY4-bc6Si;FbHLIa>wpj((RuI=&r98REG4%8FY=`@n+@Orm>DNV6Z zPgtkhxytDvYLV3PJOb5mAo(O=<@1Z=exMfhSwT}QB%cJQMfKXS7ePKg%>=6FKxG4((e4fT3u(Z_QEx3-Jnio~(U3q*9Owv)Lp2{flSVhgX=$2bp(8L3{hj?k2W^#JvL{e8 z4}urt!V(>PAcNm!n)(k2UW~sz{%%Hh-RUA<0{!AZS701EaZp1K)S?~5G{r(!U>xe_ zX@a*RdMl+8=r;$dffwV_s-|2l>^;_@Ba65H{wu>t~SsV3*Ce7{OH(W)*rfa+OWO`XNBmR?)W(6NVAK{_HvunXEP75;EF=Y^wgqpc zmmsKS&>jN)AV8LZE*fXb*Imo_XH=hAv%m zmlO-_fiBbeZey>b7L_Kr5~!U633u?ry0P;@+zqL0V&?3Dm)X zB%xdTXlghfOpPB)Q!FG2-P(qcDR>3{?1cgXnJoYBolQ3IQZ?vz^B=l&2aRcpg>2xZ zny~&Q?x-hxZY9ta4rC5%K=Tcv@G^8@DNV7EIjjLGnry%|a)(qUDmounQx4Psos*LW zLvbhY#4?&2xV(}<%f7HWgHclwTiTc|~Lk^uy=FS9W#-o)?{$OH(ZL7~XZ(Yv<~=NulaBc87nL(sp2W- zPVTgA!U%d(MN=%4569_9-0r>-K}T2W5y+kc&4mu6!Atb2r@ z{!h#Zbd3Ya!=>wKu8-$MzjsNokUU(v%H;+s2ny`wPN3@?NbswdV$YslN0;uCFio+L z;8!nO3)`k6==q^I0y%J?UNGnOpg$J1cF{mw-e|l2danF;dhae%TSA4C(#rO)x+v=+#7j(Y_F2# z=*rFq){z61!a{~utSsJXD{e(oEK~{$8HY_OaOWhtRGvV$IM5MzGt}6r;(cBPWi-V? zN8ruyVxtfKp+i??Z35lqK!Wo|RJ?R^(4{k5N>eN(IBz7tSm-$n|MS-0$MXz7M4Sk8hXbi4L+RSw|M~H+@uD=vLTbqtDjF}!6470^ z`e-nL?s6a-=x&8Rv{-^#G~JMw5JozKyI4&(&m zqe+`K;?KmNE-4mrg7HzJ{^1S;DgT&3AQuj_6MCwS!2$lLMYAhuiiLJUPgSaI9G<)G zw_2S*4>*uuvB}a=9q%Ff+9kz8g2g5S>7$aUMJt{d5XhAS3Hl7zUyJ@hEjs5xQ!FIt zGqgFm1R=o7g~vx;Q@;?% zjROfo@k2nL+fC@wy_`lYN=A*ySYdfLp&gaOT13iQF?z92F@DgXLJx#ID zGg$Bbwj>mv$9Sh&l0Z*5&~hl46s_5qgj%$}hNf6(ITTE0Z>qy*!Yo>;L?90iGz*r` zR;zD4grI(_Xo`hq!SdPiliBOgr91Xihd`bjNFRnRD)N)f5wy~qrdUWHhAko^jPPl? z(jvwL^5Q`Ep%*lDYKaemMElSb3*CoaP=EKdtEfd|4mcC&DF;f0!E*Boi!ub&cS*5O zDh!rOe8#Uv(B{|S1oGxU<*-Y4S5yKnB2sxnQ!G>tyL6ZNiN_$Qa7G@1d^pf4_`=rc z+v5XNKg-e-3!Q>5Y@Yk|Oa#5X(M%v;4m1>2wffjU#j{&qZ_*SC4TV)LH-*i3LgRCj zcunU6`-}sr!Q`#TRXx0Z6x&QwETjgLx9VeR@b^f1>l_04aiASg9PYnnJDxcis7+HW zv;&I6ubxO;gf5+btOkLeb0FPNm?cgxU4ia8`&Tr@Lb{^cMun@UqGR1#pJ@;vwJH`JoB4)z2J;6Tc-#<#Dd&KW^sH)x85lwpmpR%c5i zf{y(5B~Ty-QiDsUv#k}Mv9PLzrdUV~E}iuDJ9t@VidHItUT~l&_#XXoo2H0b)Y>J* zLQ(KNQXl#|8nvh>pn^a_9B2c~octRZIvzoxfi%TJ8(`)nK4AcYW-i!6pfC=!I}w(!7k|)0EfQKt zQ!KPQ(L&{*%U1josG2uS2o%nNGN7kwaI6V0VNbkCQ!JDLJ=OWHqBBs7hSs1UmJ>`b%`{^LT8HLgq%YEC^_oEgybyO25kpu09wFw8OrNu*>yT9hO5s52P~0--N^g8vbiZLV#X{;(++x`+UK&G3^$7Hu z1I5BDvElLl_hRz2zodcbE3)`%2+7{v24|UB30)5~>zhTkVtlIl4f;MQ;6bt=^MOz0Q`=RL4#YKqy?0o2QIgoe^ zoPTv7RtP~aBWa3-#A7T}mO5-vN09LdSpwy8psUc^OwN|Tvxx2^X^MreLT~fwMYDwn zx_?TYKp#0!Fl_y>+V6@-Z3d@liiLt<>qpUCpheJs z4*Zsd_tQQ>?Ia`Mz$gm-aK!qGg1!j70T{Xtlvqyf?6bq@qOz(29j@1Y< z)%Zf7A`Wy4F5N?e*Outg8Ev2`7CHr&uF8KCE-$q46Kd>yj*2-@0gT%I{_#47Aj{`8 z#Xpz}AB^+oVJV)V0cP=4l)Bu`dp@Hxmowlnyk1n0mDJ24Z=0K;Q z_@PbrVF-dmPtz0&or2;A)zI5`@32C#4uMKJkg$ia@;X0HEd(i)&=d;^d%!fyt^fJe z%h^iC1S;b|!O#hex2?vnASGp*VxeH@1b)mb#8b*rkDLiq&VdBK9;+}_L=Cm*SCdSWI5$G!iYJoSy z?hkc^2wJ2=Q!La1Z-&KB?Wds@)kHND=o<%`1?$~m(_RN6C@Y$#SZEfkcjq0uH3dP% zqlPqfKCs_8P+xeC6jZys80X(COA|r%sA!I}$BpIR1GBe7EXdr}) zC}k9dCW@lcAVjH1C`u(sp^}u9N~NNR-}UZ19>3G|y*|g^UXN$ze&6Ga>)dw(6sXkB zU*(FR*{^ttLJd%$^7q;@FVvzL#B z?{^9hhoVFGVQvb8$^=kb_~N|v;oPodDAX3dIKTWxzjOqxi}=Ezasl)h22&$D zJ@iG;lp8!np~oPKw<4Mp@f3xc)xv9gdz$0UiQ2S# z45|@8me7HGFylO){a7)brzm6z9Y`gAe_TWq5|qoJS^?DMk&L#K?Bhk~(Afv`6otAx zf>&97eTr91W=hvH=%)b6h0%o3=;^Bvv|NU#D3lAMi3Mh1Ip|~vS*BRqc)|V>Kr1@J zd*pIY$RTL=a-O2lijGb?YsOpRySMKh4F=T-AP<=K_kC;Oh@g-to}!QkO#9DHpR*pd zsFlKE2Gt9ows7~hE-V~|puSyrib8GS?!D#xVEhIZIcsYM{T4tm&KZi&_-+!i_P4;Kwb!0d5zc(vM!)*MyNiDh| zC}lcNQAiqQ<9)(9?nEuR5u(DNLjq`XG;D-%`HM#rheLUaLYt!qlcilRlMJDBLYYs-tO-7^L-x# zt?$iK6q1LxyJrQqZHHRqbufWJjsj>4lyh{e+kl_2)^8|vv@q#@jfFz{?2b_&oqTgtY2TxH*QYx_JQ98b0=ls!R&~Xt2Wl{dykK@kC zxCWm34+zSlbWbhE#Shvmmow;u0MdpDjY%!^@rxfjuHq>QX~TrZ>HW)Yqu+(~gQqB@4F!`U+LhrQH<$e7>l&{k zHv!ZWcIgh^YgdP$^Z=fsP*2#U`@-`denF7`A5{jO7C_sfEGppHr33_RZ{R5kZHKa` zj`qV9(CKZlN}oYz1ke$fbW)IchFdgcHBV9K2uwQZ1ch}((7xEs3_2@-_7%cF%&962 zwP;ryPf=)JA-wva-h3&7j`calAa?;|3@5|UXKJ_udA2W4QOFohhVEr|cO%HkA&5a9 z0>}eiA)=W+avN%q-VvUnkO#a%#8>Kz1%h(FBr?cT03}N{EA78&j-WeVd5S{GuvuxK z%Y^|58aK6=L0$sL4EhXZBV+MnqLmI$QOFGX4AJAhKSxl>h355**O9jX`UW2il@A^= z9JMIim!~N74L%rpv3krk1YN4@%OD>C6agz@`-6NYA;`R*rzjKwD`NL!;+`Vti{TUo zofAL`0A&<>!DHC_i+PGd3ILtFXN{K(cHUXVpz{J~5R?~Q+y8M0YEe`aPf=(Ploxu9 zI^G(!XhP3@47wnI>L0?0?`$7!1kG0BDGJp;gzDL6Bk_{K{{21-@)ba}(1FaTx52}f z{Reo8LbcF=98~;k9cocsVKjsM1kfh9jz+W$azQQn`hlk?vw+lX(m2wE~Thd~zw zkUCV)4z;;53PA&Pd5S{nP(3@xR{?Ju*?PH#L6-!OHayCE1f0OrXfv+x6os_mQ9knK z(L~guT~Z3a8!uRY0rU-qmcy^Ez!&VCW;{iqZ!omfu${6QL3+k&3L6Pvtk!8=&}Hklo$3&>(v!Op>aG#AxU{*GsPkJ#WnAHSTX2| z0MdYU*vGr1@q%w;PoAQX2CTyxt6syS@)~Ow1_cVBiND~*HHXbTQHy@>;3*1C`~@4z z?inmWztNl5p$rNVK)YaKaE0tiDFkJ{;VBC3f{DQd%}+K%q}acITf5F_bEb{ipT@H; zc&28L)-5`99Wio;#nzD$Ys&apd5OKj;q}uk`=9|DK;n>L0FU zyZy!$^aslJ;wkzECiPdH>QL^4FAF~hC-w(k6ZH=x#gvKQV>we1K`QRu`C=;1y% z6NXyk`Sl8eLIjXD+`6o`#^MzK+Y+9lkT%@9PUMu8A?Wa|#|#P;K=a`GTYdSeD>`(B zvw4a_^WgfM)T|jU;&TZ6z@RVz^b@8?ZThD=AV@!mrzrFjrbw0LMm<6;`X}A2q46XN z7eF&%=F+HmD<1BAm*FW2&4iiD?|QxeBB*L@9|m0$K#!nP^W&ZLGSs4n>v)PnkDya? zx%a^=1igDOnL*bDP%3n#Cuo1iOLi#g-Saz!7tsJuyvLQA1Ld@N_g4b-BLwVnPpUfp*D&_pY|K$BIw6Do}$o1ICJt(-^8n7`j192C`tf5fr0Xhr&YKXSnV-SQRoQ_lnw0i4xtwH z95A0j(E>;fR+U3{EGR}Ts%ny=kQ%HiPyBxLBZ4|RSup6X01AfbnO({`c?eo|l&2^Z z4AV1*jCWNbXw>%;42lsz2XD$~w2I}Q~wxAYm$r`|*cmd=D6SsNi6^+r!F#Q!zQOF4K=g+G5t{`a2 zi>^|`FXv3{@I(Lw!`ji1-)VU0p_s{26bgp5qf6ud=RN3`)zlgEQ~;U5b<{RvVGe51 z{xLj7Av3s+ZVcarr_jcoG+cq@#zzc~zjlMA`T>$AoPt_t_6W{C9R`L{ubfBmDvZBp=bm%reZquys zI(jaE@?n!x_vQUBq83>t@)U*gVUyEY#c4ed)PB?u24x5!>i`+;;uBx+q+nT-6osq< z9ChMV+$^4=&~CV3wYPn@M=eSX zy2PL?0TctB{NLM8JwZ@>Fi%k^20Hn5le^%Fqb)KI81za2J%jIZ)V*tY4?*i&@f3xg z!FM^5zrMk91)t5{F(_LAIl>q=IIQ$Hf*LmS6onjN3~M*CZ(DTemOlH#pd10@3~xOQ zy*_j@f(%o6ibBrt*26?2%d-g57~Zpa;|2R#07?3FTE||bB1l@5rzj-p*O|R_UW6b^ zmkA7dBY@IiXqj+#nI?jso#ZJBrNPj0M0Z;}KI&JwltFI=&{9~j^v-&O=X07>@f3xY z!iwd+%^Sa?78w}sVoF!iR4A;`S{XucP+@Xap>~^&D~mU$8Tl@Dzncz_MHa!om1eNqM2e8B{2MIz!?A zq_?GL-If*_N&JVl{<@H`rsG7OiF zx@|IJP>}$70R6g<+K3ehvNhu=3O#^+okLs+-e-9<&51$90%$y}v@A58j$fO-C7q`z zG#*x3`ZtV}L5HrNMj(Se3Ltw}$cUY1CyiQE+ayIHdsxW$H?;(hk0zdZ%%D#KXfRBY z>$dg6yAl=7@)U&z!zB4O`@thoi`vzGV9;j)R0MND+wFCfP>W{&-VC$X^|ODB?1Z`W{*gDiJ^h&}ZWts@8&3=lyAZ{oasgy}OGf*!b*vvc>pJ-I z6oqVW!G_M;9TL!?``a>`LEiH<4grL!AlFtrMWGN_5VTd1?}u6>zuB2VH3G;VdaAL? z!MhPu*d#?Ef9R=>A8XnQL8|E?45}4CyJ17;x9yq=2pagDrzo@=Hgwi{4S$QEH^Y+| z^iu%+f(ebaU+3b9!4OrRqR=mx(8!Pb8iJszlb;#%O8`Yfzbzni0Qoiw*x`GO9nEiUH~nJ?f9eS zO3g&j)1^E`q2;h0|D64_Ur|&RD1np^U z!k|9_s93(0c5q-j+^?J0hNmc0EbpZAxzo)I1g$VT$e;!Rqz`?DhQgL>(V^4Y%u^K7 zhdx97&O^9vE8^J&2K^O4KVYG1-wyd>2=Y$lDGL37g{lbGKs-ZIEe_rHsXOeFjMH*S+b(3X=tMIjSd<4e?udj zD}FKPkN}zs-7T93_wWRJNF`5EXfAZOHa&l6j82A_1>IUTUa*G+Pywtg>yNl?j#^Z* zkf$hA04vJ@D`Kn=^!%negX{&68O#M~h339NP|Pi!qL3NP1?hcGz?Ep{x)?IZK>&S% zCG7UAv~UNqs4GuV=o2hqt4!K42es(9&2|PI5kN~aVExGU0Df1{x!pWPp(PoPI+>Z4 z_){AP-?%c!Q2-r+O-{4^C?%sqm-&{bD0B!mIbBxC7>8PPN%J~`oCMHIxO*GDJaZjE z`zP`gglotkwaw$nd^B zgH8${FDQn6&@8h7L2B_lMIkRJhF$aWJl?kTs<%IbP6?n6P|u*YSsstdANS!Y3Uz>b z2G^i;+<~0w^ngLG0;oT%m0wjeTZLM5?kG=Ds6VWgzv*;77ah7^U*0juO#nTGThPJh zM$QOI`N~ridJ4B7Jp~141hv&`V9;p+BnOX)7f$CL5mePAMIkwOOw2J`y%$08S9`W; zykO4=pagi~jl9b3#Rxhc!c!DVfEV7Jd1rtNkhEKAGU%)TS}j>wo;(T{PAInKDGIHI zg(}^4cFL%ZQO|f8gWPpd9b>P$1zJ5ip4)|*|J02C=l`r@d<7TXLi4NmuL|AFQ}nNT z1sC0%nrK{8-~7og_E&ibntvJQb9Q*#evgjx$0jKXU55D_SKp#js6~5J+!^F4fCj=T zdeqYte>=r-C{IynAe^EDmfdqfkkrvz4Du2{t6>rV{ZgC zuXw>AZvj+w2NrEj%>xiLwUVbORCNbFm$B6pucb~~R>2@20Td0;%bC48BS?8UPf;iu zAo&qj@r^Mdwtd^ii|(8NS_>~v-S*}l9^-h$@f3yD!pl=Vd{ma9L+992l|knPkgTL+ zeZ!sws70%lc#1-@P;yf>Xf1A$!y!EeT@XO;<7KqhOv!17pe=`aibC(>;XRJ$6;7fS zP5WfVAYTDw0n`2$zMZo{P@m5{MIj5A_Wz>sDj7k4<{o8`p8y&IC&R^2miT1Ina5KU z8UrW8wLy7!ec|iXKn7hDKp{|3)w%UduIrY8_x-5V?#KPO!?AnOP1O)C`bU+z{y~8>@(hDqPvWzC{zO{L!R#QX$UI5 z>&>8G0i+DGADe6uMHj^p7}kWLDvM3IShDvCT6%J$o&sb zQOFzyyl)y5@Orn&DoX}k7eJET9QFGbxRKp&OFm%%BJXH0Ozo z_SOhZ{FOiZhdf20IZxo_PtzT7Ifr|{PzK!)KwIHD+NC)hzXQXvKTlC;D_lnwq5d|g zMOAjm47w?RI>K~iVR8+wW2`*HQxxh5)0K{?N9H5w$fwT?x+Q?DVEN3*z;-G+8SFmu z6osr{`Rw|$PI#s_P)DX+<8>4%fFxhB*qXk#0z=byib9gFSZvdtg5M?6GhiTtZVRBj zFogeoBNCT$6g5dvXfF)mbBy2kp+mPvY8r#?2%rbBUHj_c-S{hihRt}2LJwfO_T;)n zY6xnx&V)fx0w@F?6Wit#97cyOw@HdZA@G=pa6N@@K`WjdWKgsK3WNp0vupZ{LXg%| zo}y47EC?TS`Xsi-Et+zdrzj+u zue;T@ogIQMDyT3hNdWbNodTabw<|;~is-^q6zT;#1*}TzaOb4hT$e#l1kl6ZP)gl( z$aw_CSnw2u9{z?;r}pYF3AJc$+6D$a6+o5H-8y-#;~fO`N#`jFRYG^`cYpJ81lg-P zFeq67b%wLKK=1P{Im3eS>IiyofjDGCjO z2j7Mt)7?>v7JPfapl1T88YXY2UH1+}Q2$b%qEIzV-i8}`twPWp{Q?H13ZN6Pma1)_ zJ`+I~=kgSVPQY4f_3FVT2nxI2z@Rh%bOo*>gS<<4i=1}^Pf_R!Tt~etcH@4XT?eHO zjTdaX0GbZNmUI1E&O&Lt1<&491U1{L%Agzp_YGSBd!qdy*6wov6M z3Q6V}lplQ9k6PsKa+E=D1&{$O+IsDuh6_~op5!SC8Ni~g%F>z%2gVi42IUDLcUT=xU97YNL0NNnibC$NI{b9W5)ag( zpI3_*lrMk=!9vwiVAwO;QxnhkKpzprv?EYv71g3@Q{r4lo@(GB#JdJ+w@TQQ|=5VUz2Pf_SeG< z0&G_Dmn+A;Ams#}qL2b?R#JG?eI_hm&{qKz4U*A|{2r8W$%AisKH2J5D_Scse@f1zmJf5P^A>`>I-v06aQR& zh+6b9h^Hvj7v_TgMjRf6AZ@uX3@R5un_z#6*}}XA1kG&6Qxw_+`%~t`f5C_DlC_LN z<8}0109nB;XtTN;o)|p1gQqBD1-GEytG@h0E%JLakULoM=p%Tp9e zf>IEvZ0#TfUDle$ph^L>0bVjaeW;Buf~=b~;D}XGZ1DRfFf$M^7d-D{9ET98@Qr$XxXc|R%p~$d>ffu#{ctwuHYB=N#1+8 z2R~D)o22MpRp1Acd(E;=qcbObu?G9A>IBW#hG~C=$#buv=6fySDGF)Bw13^XDVGrR z=)NI?>IKkGc==55nU9qSa*5|D3jKta&%6ulh3l_e6|5QbTL5Lqw9j{xcq!^7KF>!T3lY{OF&>JP)iXxqo*(V^@2_Bw+a1W+EV z;2UK(vqR9gCMgQ#!3zHTd`mn@encakL4O61?pK)oh&I+h(28+9MIqg<@Xd#}vZGOp zVq8iY^iKedl4_;>=5h$$Y3p>7rzkW^%1Os`u!$0aYCpH_+IZ2~&HC?3n7!37cfI&= zAJn3%FFZw|z12_x+V3YWB5JOq!k|L}s1kacQmxO#Bj`nw6oo3Gx7qRXqH(B2w*zz; zbXWkTKt-A6uQi4U3b@Qu6iR`LvH^1DX$Wd1Ysw&d0b~Llj6LZy@RahWCMgP;KnElA z!2fLaSF&(mkb?jk2T#cj?oPP3`L9WeLgU~m>3ZcHp02$5%%4F=1dtU}ILEHZQqkQG!o&rUbM^En4bBrwQP09}LG_=qLP{1CKiBu`Q38qCHoxNbED{YI6?3K--h zfU;qxcf(AdLGJl{+6CM2GHEcb=k98q5`R4e`d44hbSqBUAGPRJlN5y{?^c@o+$$9w zx?lN44Du2{k`??Ddix>~l>UyVC?r|I@6vo!CW5LbHt*he9eE3&OqeThZrHd4L9ZwA z6ooQjuHdrge7y2_(X%gud<0M;ltp>D=HSkWwHHrOC=tq{4#+y@p%(enYBA`X0Gb4| z@qNPg;qKPPpFBmONiZ9K_~VyK1dU$4ib3ZEP#>6$cTp&HK!>i&3Z9}+ADE5bJt+r2 zkJ=^JGU$Q;T6i1Y)Y>c@*8_iPlA_SU+m1Rr`VPT;@1Dv&4DuB~-C-bhpulf3YEkb& zJVl}IFc1rDYj1-N-H$`j4Du5|6Jd3DO^*Eu1WDQR6on?j>hK7&)wdAj`7WD57X^?N zjQDRqw;DXAtDn zoTn&c1UJ>edH?hE=8`Q77<5?xDFn%A+ior>Lr{h}Pf@nK_ zkN6h&gfb{d0DS@Ij=m=D1qGkuDGGf7Xl(en)u=@)YmylhEP!;N_#t?#!2;Bxg|$3I zAsr}wFmLG64MDFLePPg50n`iDfWp(#Vi6Q=z*7|J1#3X(K$3T%_noc17`3RTNs2;}ccJ@rHNpe2 z*bUPd6efT^0JQq7kv@W)OnHhz9{{R->wX6PM#}Lf3ec zwW#nQPf;ib)_}Z6%HS%CZXeY*otn#_ z8v@7>_7MH(_2eaj6m@utLWZ!1Xu(i*T+~{7p^iZ}1<-mZJXZ9XdKp0j{CJ8&>!I-2 zuiMcoph@DS)bBOG4G6 z?s%|V+ayJyYS@y{ZgNjt=wFrzn&M?}XW2qf&%g6x>R- zSL1aQD}bC*;H-N&7H>(gY0XmT?983^irYeF1a{7D2{L z^v2WCmxu8bg)YG&NL!--RRmphbYM`t0J4T*%TDKDd{f=x#8VWqhGC29s~p_Pcg*)^ z&;tRqrdccPGvgL-Lx*nhJD#G@nr2QqavL3SFKDP{0)r9+&|etw`8PX{_dfh>lA_RG z81YTdE^tAIPTH@4K@SB`=p%R@%{w~_wdh}y6oo<`!RxOFZQhEYJJNp{^hf~pyeFd_ zf4DcEV81NGQxxiX55DJB-#Q;b;igKx8!y<$0!SLRB{<> zE^+X*r&6anN3vs-@-|DKOpRN5p(Ay1gy(ipY}zp3@y?+OMz6F`UIQNG}W^dSW8{KHce zIt-8USHC*FMZZzYweo!$ucK4}q!%Zn9dpqm7eOUWQWVmQgWVk0y|fV2?x`w+(gaX5 zxHDw!(#10;Uz?;T)C}$n4F_-IWu3zV^%;~dfL6mzb^6z66V#&9$~;A()o@c?n0gv7 z>$slS%%JB2Xga)}qCBZHp2RYN%i@I;*DGC(>v}CjLJOo82_GQp30dyP2u*cQH0ukhs z#8VWy4P)372U}c2&_q;YxEJ5+dYs}Z3VnsnN!;73xQ=m1nJt4}3m_Tzjl#O;??Q)8ww$LZBm=*Z-_LFr zP>V`*eHipc0QH07@b(e)w-A&(i>D~m4~oNeN9krEXm4;dgWd`tQ|L1!1)Rmtqu{GN zMIlq@GgN+bJdNrYtD5C7D0dcmALH14Ws_r7BMZ=9CEt$!=l`r@%zq3o+NvClN7hf8 zr08Fj{}?`6^u9CxK^?Z%u)iu#(ER2wYFjf%6IYV2-^NoEY7V2eXFVs!qZV~a@7%BP z>dqHHIj|rYUbYXvK_$0Iib6TCAgH!C5EnlfjZ|aMI{}mo-P(o9Z%(2!XX+@PqEIq) zYx6qx(MBx_KDmHF1p+7yHXLo&={E{NHm7)sLTRw!Xn)H{F9bPMSTg9n02%`;CiQ!s z;MotGN}i(77+5h0*>fI`ar&-wVNjs}^7EI`9{2Z-0cuhERXjx@KY!T8nYa*FD`h`c#1;F@Z3-8uCfIkx-H5n3@Q>pE#bK@HQ~n*)S@MWc#1+T z;ko~5_{PNuYB>6ZLB#^-8ce{xd>h*qL50V7ibB_50ygbJbT<|s4hlK|=nRTj}367b5SsUA;Js3TNa1l{*FLx*l~ zz%&MZ7C=EzBR5Fj0*~5io1`cd1T}J<^`9?7kWTZp4EiE~VxUE9^qb*H^06&=ib65a zqW4GDA0gP?-Qa2G9LkwRAlDpXFm_cRThgT@f3y3VX>*7|IJ7Q4LYa6pzi`G2TBD> zhYv_azftq^JVl`#C>3aV>7F@)a(){!s6qhEghTf_upOQx5BS4V6q*T#?u6-)W(az< z(wae)0_Y~p^sav!_YAe@?kb+5&`p@>J^1hFYy|aq=*FNb0hAXHlM){%;oTgcnxrU{ z7w@Rk>Ap066Va!>5e)hvfIQ(AG$iLJ-rs${A5T%p6K+9A?%gg$ExKx-&Y)@mbO%bQ z-|5BlM2F7Rfu|^R2TG}*|GSLmbJTK68B`;HOyL&PCrJ*^Q2l9=qL3-vf?A9Wc#K-q zb*${b#_On70G+9q(Jn};l0hx{*Ca)uGxhM%BIh0W7PQY(g+V_BP#YMvxoz^pUm08L z#Zwe&1EaQ>*kMahi}ZfzGU%58(t_fL)tZBG&G!6io}!Qz6hHL3=+FW|9haFhs7?U+ z!4ARjWhyG@&=odGQOFN=2zK4!iYH+8BODl1FMvG%$Y_5Wlu&@6t~Yp!LY{x%GfXLe z@eaYc9RnEjTL8I3m-v_VLKg%ncj74uxkHy&YvIya=+Hgfk-(ro0!Z>b+=Mm9@J-ck zCr?pG@;zMBv|qR~v;0K?gBk>oK1>{KzORhC#8H_%MIn8dIFg=g+!wW|eAHhC{S`pI zP{-JEl8Y2NbkU=EibB3n$5^JL&=Nr-U6hm?FW7$qNU=mld&-v2xcH&PNuHvRVhQwu zV#eUr;VoY^8DuwGuvrP96t@(7_nuqAQxw_@(C^%~dr^zt&s@%+Ljvd;{6?coku}DMWFVFMyuI z#L=s&VS4CfNRi_y3O$F3qy8-mU!g-6upyE`4g%-|EOEZlsle|WIc~~R6gmM*oGaVL zocuHG*gv9 zM+MM$fJPN;?u$+a_m@0Hq45AUQ)**?plieR8FWkl`B%$m_h{c-9zpi1JVhb@YItq$ zbVs~2GtOx!UnHp;K_yz42Sx8?`8?_$Y%;2%rR5KJy%Y6yF&V zKJpZW5@7jEvvm(VUw3>?5QCfrQ09FZ?fI66%+R5;(c>u!W!{Hkyv{YlQH#p2Co;%I z0NsW@!#=$Q8xWKj!BZ5v4Sk09-DcrC!-tN=3_2-*T0r6P$vx2@5tP)4rzq3{3XeUl z58OvBYO}lf;Ku9dlmIG$b-I+TNAR3#Ws?+zN?@Jt@9CR(1^;|bUk14fpo4Jt&M1F^ zx0+hK<|zstguC~Gz718VMcveDqvy#ZYbXEXK!#u zB_2U@LV1cpTVXE9%Qy>f*{+k$VUULax(_u8kcncuOcZhWCF6={4h$>G}Xl)`a8RR;3MbQ0I3mD`hfZoEW zeCg(xcL+M?%2O123!`$^t^0=}NVCk6LFWX}ec02ganlpe!>4yzGEY%x1QZcnf2`nvS~O#X zGJ`G&pb5}txc0W&G}I!Mkvv7A3D9R4;@jdjf+|i;W01c9`ko293q#7Z5tQr7Qxy81 z2^%vFqk<5$xyFP+0Rl(^-aVTk_Z0s|TD3eyAq{xKp3JrS#9}G=7oPt{PIo_8+R|Jp&yw|kTWlukJ=$<~{DGC|Ddrb}37*0jd ziUIc+6exgB!!%mhjGE5~Iy;c3D0CX8(Jt&z#HaT!+gt_(2_RjV7%cR##g#CB_wf{k zbYWs}`psMFs70S&*D)wq0Hr`D|GjL2C2CRV8=j(23Uu=GE}1MuQ2jW?p^X>pRRJUi zwLP8o87m4nYB)JVhaOsDxSTqWld( zB|i)q6e@tUpzt_P&K55j+^Oa%3TZ*%@q3w>76?+AYt5iA0dz=GYkF?kX4Imw^LUCv zhv4pgX-aJNQ7ffiJim|CczfpEf0fV9i&|x@qzr0oP zU^ya|rzmt74xNYR_@4;!>-?8NcLmVmUr-8Sun5;Pm?-cRg%K6yOoAF zUa&C&$N+jlZ5(P_qu)q+4^L6Z0D3_lt!7FgC^lDY0`2A#R(wE8zy^StGR%neUo{LLXtO39&b4|47KR%B^w6a6F}im93JyL zJ0C&O{yasYa3~HpbJ^Mn)iM73?arY4vr!%6(Fw|0B^$fok#)4R$N&H5|Eyz_Y}_mD z?_-btDl-|LqJNcSib89lqdr{LZv=vt zrDrnefdC4422)|X*2yABn5 zBY1fIh*H77QFAwW)y9kNp#Tbi4n_;Z&V>l7Xp*8(0CX^pp9=FwEgD;;%AiLANDJyl z=GN}}il7NUc#1+=P&YE_TPr+qwAE0bL5~Gc5lpjOGFph&QfDsaDGC+AG|QPCp?6V> za&K;CP@({mbku+MIHrtR6n=}RC?x5qOCOS(fuQB>k1;4o0QtkbXrcV$lL(q5&r=lg zhk4OD|5UtuX0SPkK~DsbD|9dp1g4Ef(8?`5MIl${U|23T^Fu9a`y`P;PX&+)bTDrH z4#C^pMnB~#3aLN`h3-O)d%8V%tN)~2Ac2`g^t7SGo8(_I4Q;cSZ+#3pezA&1g2T;pH5zkpp|2Iib6+Vn&t248r)|{@?5~6R|2RbOu(9aZ8(6S z5#BsSp^h*CJA2182h^fhHI@v@7C_VCH;R~EITArdwLC?k>F^s#%kIPFg+EugFepa= zt%MdW>~#ijyDm54DGIHG7MUiv<5jKV=uif|7C>@vXVBc;>Jn;Ec2@52(sV7 zQxqBr556jsKA+IZP?9Y(s_{C?6+rz`WVF>HL-A~UeGX4is9y@a5?1z{9)hN6C^INe z0I9%^oB1PM7oZk(AIDP^Qh^;e8!SHeLeN6@X$;C2K*<2LtUQ^3paC8{MWJMX2K{)B zcb_FynlR{{0P=_KR;L8n5(K$c@f3ypp}SR5d=Q`B@{8;kR3LyP--uK*IEJCY20TR} z$u}b3zq)o29XjJ^Uk1GwK!#9dp;fh_3u@7>yF5i9L#VQdQ7#BU(7>+u7*r^LY~h0a z(`h!|6*aUQPf^GgE?CvwXFniFV{a~lJ_w-pu;WH~oC$7`iY-r3s6FhsF@82=HG*>U z>KIfcfHGicc}~4`GCFjL`8-9T3>aG8cN^6lLH?5!M>k%u#RBN>ODObG?SStL+q8I! zLVsUE&GwX2e-RYnt-+v=0;o0ABt+VE#M?$Ze0YjNt)V7iz+lhw2-2=MWY8x8lm$b} zaQBdl=+L$L%~KT0f}!OGZ}W!;I=j@GL7xRsFq9W|iaUX4<8LhEDGCKcd10-pGJeaA zR*V~iz6c;aI2pQRMJz!r>KMyY6w-r}VR^6eQ>aDr6(bn*RR9ISb=0r63irKry7Lr; zg5WwDxv=Fr1WmF@XHba%>I^qk%NVyWs71YZ^Av?T!%a0{z~;pWO3f-|&^H0}99ndx zzdD|CJO7HODD)g!ln_1z7vqJGl~rrJj!Fg44w&4l`7vV|YSCeJo}$nWnA{ulK4Tzi zk&~+mgUSR@CUo-uR4L#(#`A7GMWIaST{_R3U)W0J?c+ zGhSafa*d}bqz2IP-_dv*V?g@=22~0m6SxJv@2G*Fu+H*4MIjTo1@#;e}j1ju^PWR%y}D6QAi#3wB9RziGL%R)B*v0u&TTh;%(7zjwI^8?3ut&d9zRW}h)e4}@lTZ)*#qTL~Mv(7S7FuSbWjr6NyJ$Qi28hxI5PjSijs&PWE;3m_F3 z!@g09!^IDOnxrVC0%KU44OK-5x{;a5px*+>6`;8t=Hi{yhhOp(gSKeq8YY7jt@PoSnHUtfrzCvH4N zA;~9DyKi>FRe#x~sto!ofMTEnIqT*`{DPq1GM=JP40Ir$tr--JTJ&e8K7;-VAlO)_ z{WjgB9JMG_m!~L{R0Ye>@BYdn$R&6)gY4$~_hzLX(DxoY=Rq)nmR#j23hjWtcldUH zd>t9LI>w+w0!Z?zmOc|A@H~TBYo4Nz7A z6Hig-3M@mL77f0Ipc~1F46+wMvtX7uV#K@F2=YqdDGJSkS>jGHHQx{vF{qe94g%;e z)S9-K+V?tw4h-fg3jKvz(^1YIQxWvwU~~1x>*$C8x(#3LsoG|WpRmq$JVl|~@YSB) z6+^EeXhL3J2002KJ*dCBcD&>pYSH+7o}!Q*)L&I~QO`$^{DdhCauPtkaQEKVJajmM z#%l5ug?!=ey}R40_Xt|yv5G-Q1yB~;g2o0&e?pL#Cr?o*3vNMSHkUjRl-^*=pko4P zDiloacaW(;(4D_LMWLxsFd1>(XFG!W8v8KlxByCrX|zE7SNLS8Y?7i-I!vR@)@oCY zpoG|H2AvQbgNvt*E)0NMaECwcvPoJP=a zDXafaQD_6qoVY6-!ndFwMlLLJS^((*w0K=49+g+G6opPhFUUz{X@3M| z@BP9ccLC%UFQffn{L$G6O19-G3c1C@3sf5naVhocLYZ-m*O7++vW1(fl>1IRF*xM| zPf^GgZmQ3Fck6;$G-rx3gFFRL1XO4o_ML}|Te`!1ZsVT^DGEhEg@$t4CHxz`@SDaW zUIM5clK8?#@M?qEI_1*V*fA8G~9htHoLdc?%$AxHIfh&uovNaV>d@LdtMw z$hCBOgrL2f>=@)Dfb?MLZOX-~dl9tEjHf812TO174N^NINGaKuLFWWe6m*G)8Z1#m zkW>m!Q78(!#ASy4o+C(W@I3~d7eMMz&rs|43olf49l}!-Qipnm0nY6T5#-^R%b*Ja z=o>8f#xGcgm$275@f3x=!GdqRTj5d!9VxD3kgotLgP~=M+_QKc_TopLqEHzOEw7DR zi@RGnvlYiTUJ-r*sQE(~?av0Y`k>z^ZVpdTsQE+q46&{(ey?eVFbxJ>6hH%EcB|pv zFZ@P=uT4@E8VIvngN(eHTp)t{XY&+= zY+)hez$`QTRsp%I=?uCefKp*oSYEMrTLisnlA=&5Yzn&;&>dGRHMA^aP@n*EfyeQZ z?kn+#;Egm-QOE@z$0ss+T+(=P3%Q!4Q7i6$?#t z)~$Z1!k}OQBzc+ew2`mz=E_Zvc#1-jmkGbK--K7S4)vYIpsNDN4%QcPml$iJ7R~F& zQxvj;^@S6wme-;decNNopb!D1T?kJd-D{9K0O9V`Pl{p z1@2hRpqm1y7(V1NZFh%Rs72d%@)U)N;X@vJ7gKQ^qtt5~2Hg@s4)B%4vDKx-mnB>RdmB*#!nGCukfLg&Dyt4JR`5nI%gL`w zQH$IbRx&6`0R4vRDD~?9OoiDk;wcLKhU>^;OBC(|*58(&*mxa93m~6IaIbq^hu2cG z@9-3bd>+C2!usoY!%i9%e1l=>^ zDGFsni@XL@Uq*-S)5}#1dL)2c17S--L5uka3d-Us3b_Wt*US&8`6H->+CBz77C_Bl z2>*TM9lTSZu1ShQ&0q-c-?PI$1Pwmx!=OX~bQUI#zBH>LVplf71=7Co88 zQxv)drPMFFT{w$cWF1(;pr-=p8%!LXuC;7|pu`}aqR=;(I5L?Rhv(ehw@{eec)=zM zAjww}POZ(ph@gm;JVhbNR}z-Uw%>?al)pxeK`8>r5!QebtzNf7P=PT|QOFV2fb8wQ z;FDop%mM~I6F?hb4d{gWK0GgK8p~4@+6ZevR;}G}O~NWgD+Z+sAj!Q>US2C1wP<8_ zo}!TCURQ2&K?5DSySrT&lqP_p;4xub+Z(TUU)aM_6pDh!gwnBJctYdSt561|3!u-i z3|*ew$`G~4CYz@y^cj|+v+}M^KrPBtOJUG+0b~q|wxj#y;g{UTkKri_8N;Hj&g;I5 z5mbBj3xhHQkYlQhw!(}cTzH)1&Qla}OoeIxmc_U-v$RS^tMNK|A%Hr-f?(nIJOgy- z(thw1g*w23;HL|Yv8Y9F7ArF-QvjL4=Glv9y5cH})FnJcAv4%KTQ@repA1)SO=HkY z0aOO3x9jt_xX)k}$x{?6gVWo?y5P9;Ty%Io@Taap}Y0(I}yNRbLB)J7mI1<|(L35t@GALUBeMy3Wn3X)f1@%tl zDGGf_g38RoA-fQ?O8Fjxas<$~8hEiu%Zs?mV#FYxqR_V*fWq&YAV|qRmqD)uP)~Rs zRR)d1!~fO}JVl|N@H~1mZ8@F`8dhA#pf>{O0W8g|w)eqvK|hM*^SN(7!zF@<{ zd5S`-;az{eeU4ZosE^EI2IUDL6_|kx*)$!m4v%leQxsBx8OWEx&G9mHjj1(*@&%9$ z3@tUa%Q4hr6Higd28Nbej_dGL=h0+02E7wN`f&G-b~as#F4(gvJVhaWxO=xeGb{-m zy6wXw7*rsDt^lNxR)SxXXspUp6uJV?r7gRR5ai;P&Y<@K=wbpag1p(Y8nwvsG*40J zVuGVig}q5Og0d@08B{2MN@3E;=aTLN1l_LWDGHUsq|-gO8hlf=T`H^HcpZHZK%e5^ zrGmPB@!p3`%Xo@HpWdI3Tngc6hNNCvOg!Tf5!8R#Uj~&3pj%MrHK^f78G;&` zq$qR?3cXJEX;p)uvA30`HeRsb1kgZu55}Pz3$zf_;|@5OHK~R^|kqoL3K&kL3H&iZ4N6@e{JVl{Yc$D9sJRiTf#=0t#L6rh% z6hJGU?#J~E`agJzLZbjWcP{$_YEj(WN(NO4pd6T@@jY|@BWh8?Jf5OZ4ouNJ51o$R zzM^|mUZ?Ro`XPWk;5X_ws1NRYf4#+16!L)IXx0zw!>C1T6h<(pS^!C=(H>5_8h~0f zstZq1NHUFPf5!40f{b?RGpI%YS-_U<%+;UAAZYR~o}!QiY}w9G&%|e)M5Qmxuu_2wL90AA|k~AU!zi<|&)uqSkftJVhZrIO{g{ zEx{9m=H^ov)F6OzVNSI}d;wleoMFLJ6v~A;Rhy#uA5n|klU6b4uK<#RJw%IY{P263 zwm#u03dzA9B8RNhGYFbCU>}422_Sbk85BmJpN@Vb^?^J^A$K?#;!h~gM9_gFJ`A$c z`|r(4m!U#qfwV22_nzv=Qxv)k6&gMM_~31fB}LH;IwXKj!^x1nrS(?SqKCyiMWNGh zGQ=nLZ;e_cH9dzxhXs)2BZ;r=j-EqMeUlW0Bp*q=k@cLSoAA|3}W1jL9g@S7wb@H#jz%#wZ{j3<|B!DUadNpj*5OnC&`tuZp zDgc^Q(!Ub5sPBFk1|1bZuc6ROy>`PB1kF3ZQxtj)g~DYW*nnU@zCSME}o*0EiBFKl9&I1p_jf4auYziVK#nV%Z<;`p*xzzQxw__ zv+=fVlkf%GV)#7jl(ol<{&*U=bj0lQ?TTsxG zTL=m~%TxaWML{Ku)kj?f?X0O|&{+Yr6Snla$r|AIjZCTKDGKd`Exijy81zO^k)h(u z#tYV60G)ve*a5X0qfm?T7xNT_&cFohE%Oz4<QK$~? z41=>a;fk`6b)^hCFMyUni}r1*c0<2Wn|hw2&=P1-Y)+6iYLT+BtZw6VbU^^Q!1}_H ztiw|g)Y*imDC7d`3pac3c0-Uuk_v--1yBG$*VSV1`>(>E@Dzmt09v5m8;?Wnhs|P; zp8$Fbr?|j54KMMQgt2ZsMWJ|r^qj47 z(4mX`?!cf+0!Z@q!m9X39no*}qJpO=Bzb$`>4u;91**C80vO~kfOKKhHg(Cr4hTx0 z&r=lAg;Cq#$+PgtdfNZdb>DG4fA9am%|ZwzTiNgIy+X?t$}ThOO(+W4g*HWIg+fUx zQHq9CDim#%gi;ypP|;A|>(#maey8hmy^p^=Zuidf^?E*^XI$sJ9!D|AT>uS*P0GVo zOxHwEuO~c3p`oxzIc;bmUNITlBbPxQ0!RiPsaD!2qmD0?J5o<$kUjoD5MA_pc7^K;9k(HH+|)~(;dQxqzO3D{p6Iv=>3~u6G&qNB~KW)0uE#)fLnt7dM`wkmNX>!yR{7 zA;`EUmO+66NE0SB{(c&W7j0E*d5S`sFrhI!I~l)^Hs}>FC`bV1L!sA&Uv`&Kix#Zq zDGKF7p_i7KTQq7>U}&p_P504b0pyecrx!j@=!PJtFrK22QwA(#NJrj9(AQoA81zH{ zNjfJ(y&_H_=uK~)qL8F>aw~D@8U&diQ)N)F0CIt8wBOsZ)*wjdI8RZ?1*XyZ6zJ6; zDB|sU289S9#d>%$lqwHLkbOE&QAn{Ke&WZk1%3p59C?gEp#mtQ7WN(g{hwd!yE}@f zD3nnPJ=G;A@bX!?tqX&~1W*+eKWws(-Go~7&W@)jR0YKkU*;`IMVD@UMKFV&3LwcL zPD_p_;#uOgKX{5ll0%$s>#E@cz27g%WKg&OIt(wq-Ge{jlkrlP@)U&*!;8;bFA_yx^$-=@)U)Zz(PjW}}&;38+Pf_V5&iCcqf>y8fMRs6`#$Xfr5E z0Ck1aIo=PukH<&(%~BNV3a4|l?Qk3yOuidx#GvN_NEO~k=Kp@jBfhv{JVhZ@cpqId zT+te}$ll(bLD2%}R4CK~8_8s&OSi;I3KhV7UGS*#Yy`E?if2%)0BQ|OoO2v<@g|GhW+@7_h9%C61wMGad$Rj?2E7nK zD_TgarKXmtq87FG;3*2NXko39n>|q(wdhM5X|1OF=%oNMhuN)wwM#}KD6uV1QOF!- zw>nks=z$=q-9s4kN&qc@FGIQO&}|5+ZCk+-4~X1;eOZdrff#YEh(}IfGsc zAX}(o)Q>x-f*{8;JVhZ}sAH_Z{{9n!1{b+9=#2o{3Z4AINrPe#RMjj+p{>x#Pg*b$ z_ZhY<3};Y+0IG&dcWhO=RtOrih^Hu24VUip;zImoSaK(uL5Tt=6gv6sl@)OD!w65F zqEIMw@^|Q5!?mV;q-zTj zjd_Yf4KmgmhCk0dK`jb+~N43x9_U#4oI>teH;OayEOFUro}y4+7#~gP`5uoZ?%5?WC_@0Ph8N%N&o+3Um;D)@qR?u1@ukeL z8;M%03(Mn%s24xALN$`d}I8a*}L5lZzib9hl>qnlNc!IsQ&3p#E z7eK}}P-|*6;1z;~N%0hgjBB8d@t8fXXXv0*3kpp8T3&A4S~9Pw=bUW5hOp5 zrzkW8>gtbf>53-?3(cb#^hp4fL5l``O~z9+F(-M7LS@jRb~8NiDMl5aau}2&fE?k> zpTO25#-bKQ2wRtCMg4=XW2vzA7IE>BTN8LsoR+eNs%uq;$oyXmGY6hO8xi>T*aidXRS!+459wlIs>vVLI}`aoAH zj%HAi0GbWEp;v0Y7=v0gS&648G#hq9=UomNh#-TbS_~={K!aeC{MN3369}4cjHf6x z2qwv&yiQt-ApKMW29*e)`A{Pla$zh!G-GrcPf=(-)X3%cvcWx>^6|C|DiuH(Fcr4F z?Nn{lqWlRwMWGCs3iBPP=z?0*{pwu?l?k9J@EIA}#Nw}aODCS9&=mNL?0_aV@(~_qsbp0WeLnphvMlEvLQNW-I0kks!eoDBl?sWw1+sRWD z+8F>7uojEgB1k5ol}^)r^g{rxg-bVX+dp^$_D8c6h1SBQ+p1@WCxp+|E$=%3qkJosto!ifbPLKw0=?HCj^-~@Dzpa!8mlM?IFBq zTUW52L6ri?5T?Q=EwRJn(6lc+MIl3&3M;j1i?<@~S3ky}DgpEy#yIojYs*oKj%)A~ zg`UG0r^kdNQRvbgzvaT9Y5^1tFE6Don)tLR{o6c6p=fw{nGVX^fuQYwgBkQ&00l#- zz^r=HEYzYM4Ln7mU?>%sml-3Epw;U$8T3a0nLvyD8t>!1!^#_Yib5vPBDvzSg9z#y zQpKPe0Td2Z=%=?WU^9M7OS0TcxLQtzKNYlEQ0raVQVAlR20Gu`YCf~vE$8B{NT^k2cV z?$q{G2-=X%QxwvF1@%{zFYqI1*>ocY{S`pNVRG-EO))(@4IJD#1IRg+Rdzq&wqz>cIJqDBTs@Bv>9|koFAT8)_wKwXD=L%%2c#1+= z(A}EvSsIHzqvX}`3^LdH@10Hl@ELXPDrbONbbAd?QOF-YqiKJ~wMEdw@b3&dDS&!M z!r}H?{@~*aE-M!l2UvNDmfmA6;3Ni=ZpVd5S`MuxLAZ$vr&A8U0q1K^6jNOC#*&u+hQi zF}6(SDGF_AgcH3ymf;qu58KQjO98a4o3z?gkG8nbOZFe0qR_T()*9h0TqDqDRA^z& zAS(eB0aIZHs-Mi!rAxNtDGEivRM`0*Z}Bus#{yRdSqmUTm@DX>x(44zS?gd+XJ~Cf(Apc>46+qKg|K3xad%QY zf~L9h6om?5#YCsK%6SAu{jOnA9lm zhCL&I?4Y9T=&S&G10BdCX2)?QO!;G;qR<=Y zK+a9y{1-u66<0FIUI0lZjsn-~;zF;jN<2j&$;8pqBlia*sPTjegB%3VefWB>^Bf+H zz6|o_JVl}V@b#V(@eQ9n>-gD`LFWWed)UVKdTbc}@dD#Ko}y5D*v6Q&uTNjpB4>?< z3_34>++gnd@t^m2O4&=3rzqqGbJssccD#z9Rc?t4auh&@um;p${v58dSbUSGC}aq0 zK(XB}<8R&57NrcjAb^Ixg-P;GS;x_5RjnV z4001dGhpUqT&?a#1TAjIQxuv3Gbbxs$CjcNE!c8}K{o|Zf2f2xF>E=$Vb!E~t$+M9_huX$-n8fPP0vseK7I z$90Ux!+459zawDy|HuvZy}eHUV34~2DuS7l!%IG@p%xuF#Zwe2f|-+E@@?_b%B-OuP`3oyAiWN`&!|>4D?h(WPrKbu@$S2%rYoPpH(gAlF zbWZ^Nfu8E<;dqvKQWu`0&_}po zAOG5opWe263K--gfFj}4h%?(l7NZtzGvp}>MZ&2Os<(6XP>XJ)wOZ43ANdNP>}PO> zbGD`vf-bz}DGFsjgCBLBq~9Mw%BlkxbYB3igdH~_-A3by!Mt1sA@Tk0K`!NPR6hJ>=_;2h#5tr$H{g~vs~&i_L@nAB=fWU=0W>U8N-ei@ zZ(O=IES{$*G%OK%s@hW?AZW$V5C#PZpbB`_y;BL2LYHpTFrK1N1w8A1MwRVB(9)Bc z40n# zh!1hvc!Q@Xv=(0F8WY~&d2iJgIt+Rufc&5=s-d*e9<}IDOP->TACyHUEF5wieMTv3 zjTjUxfG)v8)%M9fCL$=YU!?NXeH$p#mri=7Orv&lrYU)OP?+Q78-Mf=>9D;2OCP$Kx3kCV(Eoi?7Oi zb6*5So!}`7J%kru5&0Oi93dwkAR3j}>@mZDHTOtAYZmE1?r$XU{Q zP4`i_09p@=AQNKhrXr}7Do;^pJuHG;?Ph?tjU2x;gh9^)P%C^Qvjw`At*>5QOt z-Q5`UTmUJ;3}ng?+cXR*@DzmSO2>Wt7TBE0FthQ-!42D zjTaMtHA_)Qx(|yUi&Xq5C*s{vqgh6ovNU~&bZA`K&f`)tW6on*91}2B@%tUpJb6Z_tP`nm8kFk91 z=~-QahnAz}mv`X*`9JFzhd-85o9L;v3H?>BGCW29s^O2VG!`Ejyc$9M^&hgo>b0Qx zn$ps0x2IjagrJt2d5S`s($*Snb-MZ^C?ztHL2m?*8x*J{U8=l=pc_#+QmmQu59`Lh8*_xtb^h4xiL@vyaZB5KjF6Ef?XZn{JPqyv4k7lWno zfebQcJVhZL=$i#9Ca5Cl+h=74B?+KZC|z4Iz&#YVD37NolnSM5eo8)glKhkEd8 zpj|NZkTL1~5kYskZzP|s5oYEx#V@jCUZF@lmexictD0KJ5Aj(x#v{vfFP zW}c$ZODN~~bi?!$g3=3K@{gpmYJ` z109Su5nJ$x;LbpvqL2@CFeV&ok&U2Gs|E&T2q4KpTfQ5++M-K$&YGtvBspm7__VdA z2rA9#yT0j$%@jbT@MMrTY``zCm|UKsP$@ha{wa^a%h0=2RT%V601bk7%9|Hk@Vw}B zHJ+l-Ab6+ze6BqTwJ6I)mqA$qC=Sl;(K=Ul3|+c=*LaFTad2+W`-6c~5p=iSltJ$W zkRH@u{gYtb89~l}d5S`MP=EDGp%@Q%yXaqGP__U{iIr08wPJ`ig8np1Q79!APUlEV z?SWeKIWdqy9|X{47dURzToF(E$0YF-g)Y0m`)FKOe5%0NF=-6?D1a1UQsQv+&T!PC zU1NEQLW(dcp=0DxhFaw1_=7>81W*QSfc1Yc0Iv>*Uf?MTWxxj5?mEBY5VYu*?1rZM zC`SNo1n5wbdLnAk!b+Z^&_;lI)fC{esBG=g49XQiUtwgu+^G)NY+u&lDGGgsk#)-$ zE!%0YYy6>7L7d4QxuYeLa$`wRD8rl%KKIuo9?4;0!Z@HjdOOY z<5|S4Y@VW!w#P;Uz5Z0ipmG5usWlz1aO@@ei~@3aib9fF(^Gxd z;5tSFHTg|VH*AFf+5%s1pKA?!5j07irzo@qzTSDc8}I~dp6hr9{SZL?V1_EPQo$KP zg>F1Wp?)w!Wu+=+aG~n3YXt|@fX4~<(J%cI*P_GoIDDxhOH~Sxq<|zvGN`W1q zd7e{Hiz0^jGN?)b-GSl1jHcNc^cme8%2O1&1H=F059ZuJP^m>cgQ^9P8yr|ylH++8 zLGhM6MIkphuC1Np{T4uyPJY13j`%F}sINRlAxS5Hocb$8)S`{*()vyJ z(H{Y%0DXqR_cts;Et;soQxsBwK7;GO+bt0EUoMnp%ri*<0-pNs71M}H#4YC0L8+q-0b<0 zV07tn*63}e zR{*KQJj3{qncq;0`u5@}3aP_9!z9}=lMs|@8qS~w0VD_2O5^@tNW@^3~HM)FRCco}$n?7`9A5s@WPrFUQw1$b9jCZ&vCBRex`k zW%0yd&;*{MP%o(ZJCI?w9zipldv0#JVNVJmyKp%1_<7x1)S@9SJVhb9a5!S3i`fYT z9c?j%L8k=JDwy}am3{&5=2+O0rzo@v=DpSPHjYA2n%+tVofbe#;fB>dITBB>-&)I4 z6j};5?8ubaxcEW&i3x)&1W***u=kh8K1Y|XTQE;iC<<=aLGoUBRIbqT0)s3C&|SEX zh7ReD_l7q#OHt@9+(+{tjT?e4oyn1h46+hHQm~pcEqjM%Mf&Z98XcG z65g;+|F%DdAkFhKTbk}8TLH8Z9;%6F-r`44cSoM0&`Nly#`;XgrEBZTl^J9wfaGDl zyTx}WchsVh6+A^Dd06j`KF||yzfxQ>pFw8?kQIz!7q;4sdqMI`d5S_-FoxZx^mHb= zbVa_~8FW?v<-^na`nDQevpvL*rzn&UPw&B#-sz()qRJ zDGJ5^wbB?d?kZmIUbo4eK@I|F5cGmn4ox|Spt<@yMWI2^3)-;P{T*u2oVX|kofAN{ z{&0q~+*Ex`zYs1TW=D>aAIluh{1Z|qD!k~)+NOJUn$?whg5j0^QPfx3YAsg7uF?{A6e8$2!rzm6(=Q15V^F19wFQj)d=!O8&`y-`RI!Y@HL08-H6ovHuKqXAEEB?9XSNgUL zauq;#VBY(&oK+Npu59Kh3f+Nu?-6ElcsIwS$h!=36F`1&!-g->sYFn#D4wE_AKb9T zKmMm4*r?wN2Hg}u^{|`6d9v>?J%!p{H;ko!)d1*T@;)b79b30pt&R@%nn3pF%BK<;_zR@`t^6tItP`MlJGf z7s4Pf0rU=fLAHaYI3nnBd!C}uJLm;H{c3R?K`r-WGU%QF@__ru@8Yaj1Qj+*QOE=C zqqL1qgAf$^x{5*G0w^9@)VQa-2tilh@DzpOp+)tYKk#RCYPkHrO*gEM02%```Adcw z8zRVf1W!?D49w(jU*ohHwP=>(cn0|jpo|bW1+@DT2L#Q$z*7{;2!Z!eakn$5j&XLi zHiPajMss<`z2tpP+wL&Q;hhTATG9 zL7TT3v523b`FgOe;PX5G&8Yb+w(}H)^k7>-Y2iV<**`JfotxDa=0ja6hZ#4 zd5S`RpwKJ2XEpwcz8vGrpoaqJJygP&>LdmrD0D1OQRqEX!i+R?_y=9O_m1%l@)tn9 z(39D|zugK1J-EPA6!L|h%%Fjqc<$P+{5yjJ1W*-}7p{r6N=^afs-6H{{3zzQtg&l4PTD5|wD5MLQ?yI5IYXq%-ID|og0w@BGvTS*`zz9L- z{dtN)5pa}cY0I!o1TE}6n?XSW$j=9=lxGjb>vXz(c#1-PK2{pvcGl!0==-tF40VHyC|CfUhb7LI?*`(*RBf{qh0eng=L_#r+^y|%Ih;Wu0;oImHoXj0^w6a% zYnGx=cj#??uz!b76)>vIW>BaA(uX3Ve=Q%GB1ol*rzoTkMMSl~d)J`PDB|B*289Wr z&zA5yKAngcGGcb{6oo!pT4{vJSKtDbjD(&$n{L>r0%&5Gl$vr|cLUTSk3^oL(8Ms9 zX4!Ka?}olJVhV%81&}=~%`|?K#`VBQNAeVf>|tr<%e9Nk(50JTvyws21kgK}&$;ki z8}A+NYRgj;dI$44qwXIzMNoXH34>*)Z)tGeb88LCQKjMWNX+?XOh489%*m2Rvj@lmIG%N6@81?eH-mmXCOfLPhWh zIvb*oUwl6DNep@}fGhxNH`)b%8IE-4DGFHtG$+^_uRK=oDP>T!0E&Yld>~W7bXWIWg{S>r7V#8?%%M!z{f(D9YEkN3O9s6XKpUZ=>{H7J zcxmRq|(w_W8A5SYY_Bo${_|N3m_Bdd%v_^u@XV2r}7kq zOrY<*Cd+F%g7(^9VNi+y+9V~dmU-rs4T7dS@DznMNm*-bo@0djbvM2QGALC5J*kD- zp0^ic5VZR%Pf_Set(8XR+FB3PBKf`!F~l1<9D9;D!HBOT}Q6on+4l)Y_M;7M}b z`#%`;RsijTat^PF&+*XmupduRXdjewguUH#8MP>H(vP^Dou#m*HV& zo}y3>7)=DaFTf){8H3Rb$`C+d@U64n+PD>cMup8%6bgfH-SEo0_yidqUhou!GT`YQwlp{$UApvsyBPFN0NK^UgofOw)d))K&r=k#tGCkVwto^H z@y#~3Wl)v?%7zZ4drIdI2pV>hrzn&S9mqv}NA*K3y8HPqgWe0EPEg^j+%Xo{$Q{Y! zDGGIh3TOLE*{KL>nDc@`*#gK6PPtKy$ib__MRR$ILS}Hv%{xW?90aNP6fo$609pr~ z{F%c?;Z4dte0hpO>!6cAv%D0a+f&-5^`55t=%WDg2gvC~1MUT7cI7Dw`2(cf=(q^A z=(W)R27MAh*P)!le|#P8KwjU=Qxv)moaQ6;Y-ID^jQG4 zfrX4i1O5F`i?&|oDGIfLg$&z0XU-z1=N}gaEwXx7^ODJ8#djrUj43(wn6ylnmt@9_!9Iqf0kDY&?U$2_QSDR$8!kP9TE1KjkS3*+I3^9lu}I z2)f!=he6*3kT=vM?3{H%4nao!c#1;aP?O-fXO|O#Y!4VQs89e6sfMkldb#!p(m%*k z6dF=(r4jH(rxrnq$@UB?5QFf1ztY;(kw-x5m0N|Wx_gKF!^byFN2B&&=TmW zK0EA+*HS+X<0%R)fu3sVPU8-!MG3a?3@Q;oZD2I5?!;7|Vw|I&|H2`TH7=p*Je_GFGP=x@JEbEx> zb;CW?cWrozLXu@2+mk*2pccvhyO}{h1W+9mdNp*D#ua6S%~BMqgF>(8!(){Z^!$Z6 zgMJF2Juqxp_p%NT|6N}46ovM{u%%n58a^P%ey|&ZehHxNFiTw4+aHh0Zw%on3U!BB z;(#l!UZNJon};)~QUKkCDvNhJ6=$PwUD!#UqR?%qvdFxnYk;8eFWC&L5r1kf|6gqhs9uo^*S%~BM429+=>V?ECyXp*l9gK7oPUwLV@ z_nywUyEXSdPf_TvytPLCx}$Ok%ItiBL3IMi0vEESxW|;7C?pYWw7?P{fVG1U3iK@h45uao0N1ML30kcGsr>! zX@LWGR5ALZ87S;Wpj~(i^dRs>>1N zZp%{?GKJNDsbGip2wGp%*SP71wGlwyVadR(^a4H~cwRA2QRq7?8SE`N*8xG>7pgGG zRsh|F?$)37;rMt8^+h~Ip}WxCD*O2PCxY&K>oUkr0A*%Me!Sq~BXsFpeRzsOnVD7^ zoz=2K5meXp5QEMLps7#bJ}N)41VQh+@f3xoK7nstx8cJP6kvRXL1zV!7c3?gUYnqg zpl}nOqL3FXCT{McjEh9 z{)|*l4tJT2pzTKz*k9!+xNI|^f<~*$bzFR8c$B9oGy^JVf|PgT9#B+fF@r7$AX!+x zKKRgWC~8soJD#GDEG%E2yX38kTGU}`hy6{T&qV=b00%~%b2)~0auha8QOE!ejEp_B zD+ocOoc>|ZB>|)ckDr}wUGc=Noik5SNDUr8yL(>88)yA$<}&EA0Gb1<%B?(y;93Tc zTArfN99UI8w=(=DYSH^m+Zc330QG|B+}FM87U(ky*5@e-^@8Wz?jwiyBWP&&X$D;t zK(@tFYPZ`j9)h5L&v=SLw#86rS?~)_Mfd1)n?X(jC>DCDRr6d55Y)9VPf;iqdZ`9F z8}Y{3lgA<%_=a?^-hinH}dT3blbJ_xU%qxGZsY(I*DE2p}I9n9SI#gC{fe zig}7cJ}z)Hh0Ptj@-M&OFN3ZLpxaRDE_G7}pS<$9S&Blpq13%b;WmCnEWX$0K+_F- zT>$lfO_|%%zpg=-Zj3iiQK$!O%G?$H_B8s|smn}f&Nn6f%Qj{-q*(XQCFJesPdNZUQI`x>`Z0 z2bIyKTl$ixD3k_Wt=ZR#r($TxWd_|8Kyh#%DeT>V``tT-@)U*Q;67?Nbz}vCIypXK z&@BOU(gSvXsz1f$z4gsf6guevg_e!o@Ci%Yg5C@XI#gYF6-W0(}2I4^erg52|YibBRP zDY*NdycdEL7T7SzO8`mgA1D5fxQ!s`g*-(eN&RE+>1;e#chlFCLH7iZq{Kk|en&j? z|9GFLC?qK{m@>`sA!<=f+gJv93!rts;itSikHQOySEP7~LhF7*cWa^_?%;cE$!Cy{ z0J;L7QCxp>ybI*iR-U5J75I$u{akSIQE_BT)292#R{%YLo@#mUIXs1yAH`D?dH_Au zOLK31MVD^>03`<97eEbAe54k=8Sk0cI*_L*)BwdtkG`bi%ILkPXEDf60JVb`-{Qi) zcnx-n1y50^9lZGNXU5@Pkn^W?40<4dI>L%&!Pi~5&u}D%rzq4BRxIn+_~YvJ9dnN| z=%D~w93-W-|8&tUbRW%{$5Rwq90cp+YK8aEXJqB&%piXOB>5S%b>IFDMv&7zo}!TC zXVCV{I)P^&C$$P@P=Ejmuar{z)$aNe1dVIWQxpoXw9-g)n1WYYuIpzo=#c>G0Tndr z1)Cou$aFJLQK$!0(0sYsF&tewlNXf?3KT%aF;Z%kifI_qddX81Duzv&GSOp<5#&2W z?oiVW8zg`_K#QLI{)&f|mxl5bg*rfsc5Qftw=(Rt9mk-@0w@%guQj75EJl}ZtsPHM zC=`~jt2Msh)2t5{E@jXY0n`hImTlUPQb*9@BA%j9FBn=H9cY(_E}gTcA%lVikOEAu z_jxw25kY%q^Av>?U~=6xQq~Yb!@bWkC`15-!#qP(@HHO<{cV<_P&mvpbWC_=jUby2 zJ`4&KKnXCKa2-DePdM$7;VB9wz-Xdyz;!$o9k4r&L16;O5}piG<%$hai`p6T6oo9| z$;n1M!+1S{76q6wC`tg$h99x~HO>{U!5(?UQxuvFKVo^e$J6l$x+dewpyvXp82WV! zR~X@+Lpjlrrzlhm{W|Sy(Q^kUMgUpErCZW@Wi*1eyyhtiS;M9Kn3C8EwW$5@8V1D*pd5G~l^QyPAn1Fu6oqo& zee`7ClMx6CvgvW8>4tqFfL6h%yy)d{JP^yVHVN^bsu()c^Y=2E7zO+o6`> z!DVS&<+Qnwrzo@?Y8ejM&WS|OHq8|bdL@8-VCF=(Z_Y+^>Gsd&DGK?(%t^16rnu^M zoVziD;sj6-%n}=n|BTDzhkEc7g@Ry~xb>J0MyN%BwdWZWFMv)p!m9H6b-hrFRO)$( zLZ=$xGy1w~1A@-$K48#m0ptwxb#~@Goe^ZD$5Rw?hWWY<-eY7DWFL~ipf>_Y8Q!p$ zvNzTu=-*JDqL4DYVU0$`;bF^+UL_1l5I~Y2J!tc6b58^f>djLWlKkkw`lUv; zb~xH}A0-MPNfFKNh3a^Qx zw3tc+wZ6_%6#5FUi6dsi%TbFAYfdvLRRDE_;s5r5MYw{-td^%J)DedN&+QK5HCTm} zw;7ZsfEGY)UF-RCEm4b#nx!bT0BY;9l`OZR7TNemGU%-Ux(Q_M`a`xpf)sA@ z6oqu4&(PUzl{C6^-mOd-^j-j!z~+qo)8^s#(dpJaMWGVdoUuH#U^;@fZ@bK(YyngQ zn{K<>_S=A3bYnYDQK$ws-Ci6Rf=A_FVjeN*g8=e?*F=8HyLkvwdBIZ@@_^Sw+bbMP@9jPk2l>%IRfY{91B-H_MbQC(j_-bQRpok3zxBMQa1#hQ60sg zTmke5_V68u4abK-tWo1B3O#~7d}H25;9c@kE{hoSSpdy}7HM3!#+7IzukjRx=0J;% z4>ZX{EeiUxlRp)0%$Lck6s`DJ{x^T zg?cD}*2St4DXL8@5OZRn_5`(@Aph$S}?KYe| z5J8!$JVl{Mc=44f$l&|P=+-O-6$+pr=&A0$7djF_3vcrjg@T}`+GAG3In<)emg^Z* zB!Hg4B8am_ktc#YTk#Zyp1>l=&eLAY5VU{&Q3e$YpjWUF?Q7Luya=*r15Z)t6>LP4 zUKEXYC7usAR$8VF<9PX|tQMJ?JJ!c!C)2xHh*o#pYOd_}il29*jRWjHW0V}^w` zfKi9Vx*_(}$q z3!pDBF&Mc^Yd3--Uh@=%zT{eI$d=s6MbP^(a%N38Y=r=72e0xSea0?D(A}{-MWJ@^ zDlbcLZiApW+i?u~A%H%>vQGTW1}6kvvEwNUeSl>hCzbbj`=jZXr40HhfF!5F4=ZhZ z1VI{Kd5S`kQ{f-qm@yjFF&1kYGU(S5RL2-p(`#0otab=)z83$_|5?X41jeD7k6Poo z>qCoqivCqYU>sU2n}mOz%<dUHTyoPtfTt+r0?#XzWkc}2uvpHA zK~)0iEkJ!PN5`S-yg{C)DD)N}#W-8sH|uW{$DnEf)EV|1dmnWSMNs#>JVl|-u;i0ptpIx1Yl| zyuMK1EJYz#xVr;iyt;@W=`({FR3m^k!Vup1WNrYubcxMU6xs+w_(j@|eGruXMT0@L z0>}!Uz+Ls;eL+yyZ#+dID|iBTSmt*LL3#7^8B`~LcEe{>>e3?5u zfu|@G=Vqm0_kH0u1U=dFltB#wsHPrj? zECkSQ7#A;;KK+Hws2&Sr2twEB{!anzRW?FZfQ%NqR?_E zxzTG_hWE2(={{hPl>kbGV-FO@ZSR4g5IvrvP%0dIFt67DNAwxJu1{c)wE$9m)K+cz zs$+OIzUVJcQAqWXrG||6czmkI&dntZvJpTZp%bXRvOAs^UB88=DD)9Jfwqma@C5A4 z*bb+f?ju_PWCpv|7M7P>LYHpO3!b8o8SGk1Os>Y0dshc5GssQ=U4xOe>@?452--1( zrzms{M%L?=CH6(1QI+{T2AvT=bx@f(Hp>d1@cQZ`Pf@52Dl^mXR3{(kh3caEq*1bcJ+%BG?&_^irDl$*QTNu~I|7DP)0CIwr zN1Gn9xbLm`nx`n_1S^l7<^C$5OXokb@9Cx+_JRP4gXv2BR)d03i)=^n6oukoy3)A) z{0Rt}Z8M!g7X{D;$-zQ~o9`lMuq{tfXal^;huiq>M$nMrH4M5WfNnsGb{RIbL{L?; z6oqa;i=O^F9#4h+m}|)p!?x*~u! zL3hh%=`=h;HPMTwD6|Q>TWfoieMeBM`bP}9DuBY;ORJT(bLxjK-QQ*@3Wc?|)`-+I z!^L>gdZ`R@5LLH;(q^NBOvV6u< z6jFpb#@zT?{25g#$yzksM=k zqx%kN=G{P$%4wdWP%#|cckrK?_|(Jl+(it!E`YpYo$kY@ZV3n)n8#BT@`iOfg;gz% zpl@BarU8R)2p~z#w)T;v9|#JZ%~KST)NJ>BpJIz3C082;xeB0DFoqrBG-o}6Dw?Gz zbPC3>KhFHdM~+?lMZAH+P8lIxi8tAEZ^Gx4~TGVevEQ4+epp!5a zmbCHT5(J%I$x{?M2~%NTliI9B(D$Hx2Hg@sk`u?Pr^^K)$oDZ%QAl#)_)x>*2n3aO zX=T}TAKex}=ioC^zb=DoO=ouFDGHr~&nTtsSv*+&*F=dy?gA(jX15AuYw&FRi2Xc8 zp-`CJI^K5H4%DIn>9ZK*A%M!^hAnz;;D#<;`wX6| zHvXS-+{y1Zlcy*Y8f>LuUFtRlK~*=7GRRW^DME`}-{^HiElP9cDGDhy^_E)`%pflTv>dv`GU;0^5meSJMWN-; zC7vNS$P=}wi+=`#?g^kM7+P+3&&InsMh5T{g`!|+>FI1G?ORnraYBY^I~ght~*WoHEaG2$r--Gd2@3%)1ti0^RH zI0pF&pz*ML7BcId0)ieT^Av@~!}6K)G8;TtUOiEpLH7kv_h@)}zX|V+Ak|4cMWOD| zFro2X(+yp^xbub#@)JOkvZ(e;ui+=dO-G)hkfbbX?VnS4e~NU?Sq422K*fPDSdJZM zfLheMmZvCG9B8GXvs0rHwa9U;4}%^GpnK2@vdGiM%h21_@f3ybK`-dVgJ8TD@A1<( z2Kfu1R45gQEMH`VT68&_rzn&Pr2@9EjdW0p!h3&bP=EkZfP%@KhBrn(v*e_+>bRRtuKu2Na(RNJ-ya>|!08dfqD6Bl5AM(5pg3cumW>BC2Qh;yW zm(Q!mqf6(V!c!DdfN$Ld*^Btg;5$N-K|um&f~3kq(-WV(;yjY4C^P}8Ec|C4e~Mbv zXr<4f#{$R>hL)eMZ2pW|lxodW6taV%Wze2d{7{uHGGova0kjnQ4DaW6{|7-C%~BLv zD(N%Ksc4N_)Je;gLBRsZCPhlEu4Mmv3^hwp$R-8$v`#tu072zmPZ<;v@4=<{SqM7myn;c`1W*Xn)xTYR zM-@SPTzHB?Ay8MZJ1z&W;9oB{W>ACx%5E*KX6_Sr7(v@Ac#1;Vt*tdKD-_407U}3X zGAL32t${_5!i*`p2s*Hgrzo@r7D1wqw8pRUk%12w6eWO4Vcy$FXFGn@bqL}q3YEgV zx7sI@C#Xf~-4hw~Tma?3ee_lOi#KXfm;z5xCc-ZF!19 zO3=x-J3YE1YSER?+Zgmx09E~gmF0;w_(u;mCfHp&SYnt9zGX#0N@DznMLwD=^NVS=$ zMP`+e42l;(UNA*-@>VCjn7Fu#rzqqFQ#5bKnBfj&tK~TidM$unL5pVk$4){mDsPse z&?{(B)}sG8=B0PwUk1GqKn5`4+oIvJ3_+#MQWP?P5#N+a3-Rn$nQULXrW-av08N8= z?->6s3lS9Dg{LSq4d%U7p6|g&!`bbd&Y(mAB$+wM30bF%AX8(WqL5_fq@=_q7JV5O zyj#PdBmpGZvVFWL!~{W;vUrL@k}camt3TkO<&v4E3`!P2yWux@cRf9Uf8b@xES{p! zZukw}{765%dG_J$%M3~pK$0K(eKX-S9>aRL^Av?7Kla;8_Qz~=>1^5tGALC5?SmKJ zr}z8tG}g!@{CgWOD zpOgv)y%j*IFW~9@d-!z(ok-;=3Z=e)gKKQEx}p~SoFsdu={`yqK-B>K@|}afrI z07=F;`G<$$xogXe25E!>7dSpw*46%?lCw(&(RQhdf!6uMdkE00SR z6%q8d?_CDH7eGs)C-bGsYd(VR^y4WCErp)U2AS@q2%2*!mOAx94);akM`XGQB;O=&}>xzq7kLL0eg&N@Q*1x=YGJ<4OTb*sX={^de zIygseP*qzO)S}jEJVl{8I7cqi=NaBEy6u`0gFXqMMKH~xeMDt0f|g(BDGDuuX_hrp z74d9*W7RAMCYm1hxXHc#H(uez~Hns{cpG7R^ zDGKSsePm_WO&)zlcl?er=(7NlY%B20k4Qt8?*0RwqL5@;!J+txb_i1Loq38vKcIBYV8e2}u>Qq3m_hjh=mhLbojLR)t|+T9;VBB8fPJaeZT5Ua zEpkiBU{HYo`VG}e(F=EJqD!|gnWrf98>*F5+D==Ap!x|_4EiE~w!q8F!)t421P!0W zQxw_)FRz)`x?3V>+9f&rrW^LF0Fs8{hilQV$`Q2WGEY%R8j2shKRDt88Ir5VG3c8B znhbN-MN@8#K+v1tJVl|&Fn1j$72u6pq_j$#LEiBM6ou4bnk9Ys z7|1iR4#y2U_NJ2jlC^` zo|o|yg;ZcZN7v-GJ%UP>Yci-p0L_9uk4FZy#3O>R6+A_uS+M8v?8h2w1daC7XV4D; zWcCv_Su_@nMJ*cffTt*A_7nDM>2$I}P@0T6gMJF2fCBi1?X_O`hK=dSQxpn-p9V7B zX#E30VTP^@`XzueVbrEI;|G2poipMo3T48mZTXRLXj8(V1_3k)O4oWh_)SBQ zmMu?FXcCmJ8IM1RM-x->I-GC1j~WHgfJ)edzij|s?{1aPQxqCd3CASz1<2(kP6hLxs+p3+d&%$4CrB6IX zA-T7f8WZA<;+<-(WgmsoR|L1zU}G>rIKy?d|;K^e6?MWJXI@y(uNdl*3n*9J1kUI2N( z)}EM{9e6=-?K+;KkOyq-Q7Ll6@1wq9sSI)uK$3~U0|{II=QZ(^rzj+u7;OEgpc1tx zZSW5UofAN}?@Ou4XY{m2mo8ukPf_UheJhQ*a$~&^)ZJS4LeqV8UI6*SrCWJJ<|u+% z+3*yF{Nd6$CVAB(XnD~n2002KKX}&dUUR+}LBoo9ib8(ytZQlZeg=Z}&DCPi1p#yk zelz^sP&>TxxO^T@QRouGI=?}9YJmP z@Dzo*!d%b^g)O)eW@%C^gRTgmPq5?0`MeuGqilXMPf_R-?6?Wd*nJVT$bDo1gRTmo zN_YefD*ko`wdmp~o}y4CJc1tVnRyUF#TQy#Y`Tw}1dut*Gqf7pI~GBy7kP?8<}lB2 zpyiPg1ld$7G00f}^@7k+2`u0D@pz8wY zGyDihH_Nk!5LDSLMWN5|BOqsOTj7PO4~JYBbVC58!5emwW+8sVh8^Z93Z=mtc5AT{ zE_Z*F70e)40VFw*VD!zb)u=@svw4a_k`oE;{K?8emu~0OOa{3Lpk6Q?9UP&IhyOFC z@f3x6!F06jnS1z#Exl63pqm1y8itk&uZJX|7QMR4QxvL(p{0MGJ8qFjrQD^a8}^m} zGJ=wu_&Li<5EM|wQxr0SlAACGOT6G4q&uELw*`=-%AzDM;5mY>>hTnXBvlr@pN9@c zpHX$VHiO&+&~;cVuk)0`^9*}Yw;Zl&= zzdS{uYIxRNwHk_Rds?j>!XR$}Q~@{ayXw{;v&Og;;kdFY8 zykXN0w#GFlCgD6qA;}xo*xw(|*H!h|%phL@WC;}--5kG7L@j#Nm!~LX2^AUx_MZ1d zpV6=b<_x+ofZSnJUY@cCp8{IlEJY!A7?rDPjNONzJ{hhI@)JO&@IKo3+42`^QDd_d zg-qdnq`c!EF4vJ88_u8y0!UJ8nwf59grKj@QWTQZnmQzXxQbfz>})oJ9tt3JxMBZn z3~)oxRePSIkUHG39d%{#PJyk3H4O3>K-$m?a>$GRgdp`Io}!R8^nwzyEwoXK)Ry+V z(saWH2%x7h9esDl%{B-cqRmqjdJ5CgIjKcX2vYH$%%DdCs2fb)n#(HTExPf@5F zOx`9f8-#beuWYxHL4g7&9i9vxJ1630ox$ySibCn|WKh^wz6`Z!=r$7u1qq-M7+U_) zZHJFb>b{+)C{zMN%QH$L9tg^fab(bA0i+2hdlbp6S3{rCAaCeyEzrN}k066VJVhaI=x$Y)4E>9sdhWh@)pcZ8*&S%h50ptaz+$=aUHxog=N<2j&FF55UqSpaD zNuF=Mok8IOXbViUyziKmhM)&0d5S_?V4CH$e&kxzqWRet40f`)zI zDGIfKcgpL--Rcl@cDg%*A_R~r982|VfIFVt+n~Zz6f%Wlsg_@N#6QDy(J6{Skpk#D z+}%oc8xNos9dPC;3VnyWdr)2$o+R(^H-|w{0w@KB9_PlZ$Ren$S&Bj_F!abRx_1^` zy1g4481!5KrNb=Z@Ml_b2%5c#rzn&TvxqvimqsEe{ApjOrW-a|06m3A{S@82Y6SU* z^Av@i!lOP?uWJN?f|XPl6eEDv!m8HOp6gN&D6|$vV;4mOls6w)S?Z!JVhZpxQ{~PKIfuK zXQ6V1K`#YR5|n`EcI<)QlFMiC6orzY1T??pvr!1T;}poCR|3cdwz*lDb;D2ZBhEZU zAs5)@=63qbDFkU(r!pu`09nJRZD;xE@Dzn)pj6=biq`EBG&Vxkx#>Q7Er43XzSLI_zn?--+en_GP;1ziY8+>X z*SiPw8_l3M0>~4db(M#AH6o~@S&BlQ@T^nXaApT;(H&DQ1|U zQWPqLxq^=0{@ID3NfE7Fn(m`C0b~gy>)**+pQ08Gi{vQ^S;EMAN%ngM1eFgQz@WDR zs5?xOZ*bVs7eOxv@f3x+!z6i7ua(ab6l|`_pmYIb0}s`&axF~|d zI`FuPLD>Q*EL=)Wp-a-jzQ2Qo4Dlh4jVxhNC0;*wORfSrV=6;nxg&fGiA8LCV224bMqeu^$Vj%~A z_)NSx-H4#te@_Wi#DTKmUAJmc#&875HPaLeWy8C!CAsABs@Bv zqY@4@B|u!mEJ5W0x^(UnX^MrW1i%>fj`$#S=>qIW6R4B}$wIk%Wvht^f}9;_iiKpM z++Dj^18c{}5EHoLG&labAlSC~FSky$IUmVB>-rhO` z3LYZp=3<&+As={qPnmTSPboKg55C>;z*cgg8892)Kj{iS6DIKiO|j4nn2mq-W_cWH zQD2E!1ghdd%5eVG*i&N-P>Y&6rC3NA&cC`P8SaIkv3u4NsG0*+!RoNwAJc9K8o!sO zSf~nChvOvY<4gA<H zEc!?dUNOm4qA3;%E`Y@*MavfmQo8C-pgIn;9UfRs`+<03aF#tyvCwvSV7*PoEJ0BA z&-VnX=RmR0qV;3P`=U!%@QbEcC>C1Oq%-L!g0yrh3G|x-{eXf=&o&FZjdAx%nqr|J zP%!y>oy;->nLqD&r{g*L!+|2<_U`Frj$35wPg5)u3AcA$OJg>IBBdt~sDT4LfC=`7 zmS5Y^rR&mD1Sdil-wb9e3klT1ff}KddVJ&$JkPLj6-}{FBa~7noAtxvqsT{X z1Zw3#A}vr+X1UAxb;rPBe1Zw9%hVVC9+-vwD)S{J!G{r)O@Heu2v&9v)Xz*Ko0-3M=@6AfW zpEY@8GG!rxCdJbf3kiSLBuVvOErP-nP7%n01IYrkBB{0wK^}@U#X_C4%RW!vy@~~4NWc(=n9ECiuC6E;d`VQ4fPOTw$ zG10}Jrda4ZR4bLs{>9VLihZQ-c05Pc9LN@Cw*)cjdFaxOmZm8dvW3~L+kS`VpiAev ze=>n=I8X$9CJef&ZAH-b12n}#5%8IK{&Wh(K33&|6p$(+XZ8k6JXejHX!VEv$$s9NB=65ShL> zo}7{(%C11rz)|O20=_Eadyg zT5#a=M*NwWCL(&T<2kbDK#eeLY0UFUMlI6oLQ^c%2*Z}#JbM?^q5uO00y%J?3|KN) zDZdK0$axn{u}}sq8JKA7cS2Ct`1u5Kfav% zxpJU`Fc3?>`3ldRWH!+h3mt@k*tk<+atJcqBIDKZz`AiDDR_?Bt>VH_i`2K$6bnhg zbL5dUvkgJV-b^FV4Gwho9sI=YohMfjv^$2TSm^FMICku_!Z`#*j$Tcmn;htJ9F*&r z+jd3J%`r5^LYJXPEVG-VC4$agG$zn34&)63G38_ny!5ugil$h|8wO(YJ}$$zw_@Q{ z0=aXbqp&D%P@RYiCVLjq6bl`NMfu4CD^yV(<4vt+1oBvo>KOkW-7>Fvo8o%Z{2My- zfBw%p#%ZqN8eKkZ5<%Cwc_mG;|Eg)O)`I-=#<&PS{&^DlSKa0`e-jKns%13JBgoI6 zrdVhb3_VV66~jgN&wG^<=ne-u3$uuiJ^v{o$h$X9vCvtVMGS38nT{^q5+jNG9S@x+ z2NIq)GH^_YJAxJ*q$w5>o;Omozdv4vzL_|hKzBKiI=ruvA7O4-3~=9+O0%WEF=k~0w1&7@toVSt2+tg#ep1Q#U$lYGyWVe zwx=l;a)cF=bK-5dgV9oMPN4f7$P2zHr-%9AQh_T!X^Mrs;G2@$KLOXsZCK_>Aa4#N zeASz4n)gGOZtQZJVjK_4%HF*HY&! zrzsY?0?TLb!XNcPko!Y_0{LG?*Gz^377~F8*yiTRc$!7sxROB6Igs!;ohq$HRRqZ%rYRN@9;fqZ>;zmR zx95G&2OZCmKL-;2kVlH-pi~4MPopUo68?}!&xT2V(4{*%VFH0(aG-j)>%3iSDiO3; ziKbYn9`3pg7OuEpa-g*~fnIW;xo|T~kWk)_pqfr87Mcq;L*Yp;TXeD$YJ-+|YL(t)DeF8;rpsCQwKdq+t z4M7`oXo`iVLMPuTzOfQPYv!CH&?^ok4Lw!K@TR*6Ql3jwEF=v*)tn=#ixKq3)s;Yz z97y>4>=u7U<2kqcZZyS0!rx~P5=_KL9)p0jko6=l~dX^Mru!oiQ} z(n`Bfi%J)LBG79NBn1a;&3h4c0$sY`#Wcl2QgG0gr@lP?!nS+Y5-6Gjy@s1X{&|Qf zf^r_v6brqEn_>U6UAS&!sJQf_j_2qN2TF&@y+vkfiU?AXpeYtghsnL7(u=rSsc6q+ z0>yA3KUf2*aWr0nppd;Z#X^3t22|Wuig(<+PFzl)SPrxs=0!h-r0heGe-cfx&~BI) z{jtgqcZts_9UxE~2Ra4C595N@%t6q~i8RGRr=a-ZXU^K4=&qBxe3?LRIglX~Oe&2m zz@v$;ol-1h2nCZ-Z7&QE)b0Bt0>yKnVi+Gyu+Nl0ElTW^VxeLfAElhxn2DhCn(+im z;6UAB;wVq?jt7GF3uuaky2HfLjNE6^2wL{2gh1~&kR)ufC3EJ3IZzoapY_d_az-uE7o{l{Dud;-P`l}Q2>QE2fj~(ds1(LWBi1jd zM^MsEnqr|+7$3RV*uFuK!|VA3O6EW@!Yv6=t?dZf6-`qt6a#l%+mmg07SUz!RsyAP zpizI}Q|_LOC&~B7(i97g`fDxNyyDb4)S@|O&k`t=1AT+hgtX>GJkK!v98Iy%HyBNn z47!dtIG_7+i$L!=knk6*T7Lx`LznJQE={qJ@E5Gw)t}+Fw?H$DKxrK4c{u#E{gsor zUpGlWQ!Mm6+*nW^}xd%O@|`rU#Aoc9fx}0Q_pXnLC`<#X9UXPKy!bKYwRm=`+o)eOK6IP z=Ki)8EI-l*zf`RsClM%{1C4>3VX%Wi3u@7}Cp5)EW8h{us2PM8L5B4zCr}OtngkPr zJH;|u5%jN9iiIY@#Ne!;QBTmNJ9ALtNyl^ag#(>{;s53*ees9y%psa$p%XCt*V-Rl zhM*tsM-wQQ19<^7u=k!Y)S~J%nqna@fW`{)oInK}=pRgI zjGnRq_f(ClX^MsZ!Gy+!h%nrNjMere&{qzm1mmOYUQzfnk+_7WSV#%RM}8BpzDJjC z&Z|fQ6>^~Smf{-4Cwr_ymrgR0rda5_rM2ME=exTQ6gn)AKt&v=4BmC~AFRHHpk{fR zVxcm4*FC=V`96Z0&NUIJm;*h9H-n_tNZcj#J;!9|XgllIUQ3klD^dTG*z>qc(8^CwUl2da#R zgCBo-Dx(%fCDIfNRmQ_FSsWQX5Va_HbQ*!mInZq=U0Yt|YKWk#V`z$nZbRwXunMIp z1ZiBVBv1th>IHl8?hYGi21ig}mT7TCt)H4_p3p+e@G-4%8E7ATMnXIEkQYahhVGo-hMB$oc6l z)FQV7Rs^c%K#H%S5+mK0L*S#ep!!Ss(XL75vZO6l|eo5+6hNTq84fXqA3pYJU0k=|zPI+XIRyH{fs)`kQh0b4uknrY zr70Flg6Am7y2mQiqB{~B3Dm%W!r^zNH_o4wj#^|SNmDEo4!*Fw6IZ~j-ZSCXo`gYUMzdFfnNTRTdW>>x9r03t7U% zpw+e+c)m_{&q_NE#Ld9iLsiji6vvnqnboSP&f4)`%eu z*LVV1av(pbWBhkwJ08L-xzQ90`9U4yxuh=Us6`oVB?P*_ftp{!9+Mk(2T_X>+i8l0 znqR`!_;X5U5%hbf=<|+;?ji^3108kWV7WaAYBQiI7U}~X^#GB%_{BK)l>&jRI8Ya; zjZdEKX@ww}NSb1yE>Ih&NE)pxYxkJl1$nqr}FSb3CnmBGu< zJNGmY$c_W~!F-OstYitgbo2Mp6bt#md``hlS2xt6>O>j;jtBNC2g-x*zGnAxc%80K zGEK2i9(?!1;>HFc=;nm!1hVHqi=bPpeIUaLwdlMOO|j4-=+^o#n~FySiH@rYpBtbglq z4MBlh?Fn>^11*J8fi?0C{ZWhTw$T&|Ern8nYrBTvdf-6`&j{qifrbad>ag*j1_X_J zM^h{`JP>NO&->zY5~fa0CXh1+3J--#7xfW8M{ij9|$eja)LU(KB(5x&36?ICnP$+b_K5U+bhw$I*@(AR?fv&<3UfEeO!6^8H|XHPzjZtotoO$55lflffJXqN%Z>;3 z4hPbLU7Ya)y1z!3uDnx9OwdcAR~(f_?(1CzBI)`7oY>V zQfy>2YLUK_E`jcHAmQwX^~@D`+JAB{nqnd0?1x45TD(rT#N-%(?r|U!xa*{Jj3%N> zH|q#Zv5*Peb zJ_pi;q2=Z?YjaVHjONo63+ckpa`?94nW#njchd;u&4J2cvr^2$Z2Z7ZzDH9mR1TY! zx;=GmMbM@8N&-FLKmxcKGTmHUjgeqyjpgqlX-5ZY;cl zj=JEw`n9`hiiPIJ!rv(9>LCOjO`JfWM;u5E&a1!r&=4OGbUuluSV#@dtDm=O;s2XO zoU(*KJ{-swzOcJrwNFBquK!e;Vj*An!rnb~UmUe)kmFtgJ?22&;Lwb=!@cmEp`=ra zg}T9^8LJ~>a7{u)l@)=Wa3DSSA&L!f6INF9cjrfFO8CS}oQG{r*dFtprs+yU>w zH}76VAYTq73zyC(C)60-47+>K6bs41rOO&BR)bnpwy&K)ejLaTzDK7DD)E$Zh9OO{ zkRN=HN_rP7A?RDmu)vN7_BjXY32*Om$Kpk(MIos)#X>#d?foPC&~pTxoidj|{v2pM z)G;2ADjkR*y{Rs)Z+^b zP>b5Wog&an4)m+LgvJ@Gd&LNPS3*-P^sBp#V71$o)d+H&<3^wW4rJ0Uu92asjqB=n z&7~<8GHHjyqF?yqVmwRtAOZz)pcAkLQ%yVrS-Yc-K*1cS4f@`r5{fq?=*>==VxczZdzY+{ zz{L;q-%1B{JVzlMNE>eNAqj(bB4|uJO|g(R+}?sMW{1(G+o_~XpimAJ3=%va!E zkj6xsVxeG|&=8dF_<*4Ewkrq}#(|E(E2ueRDc*Oy?g~w@&@p%gjeY0&6G0872M83- zf!bhoxcz|-?tAz7K~pT$2CKt%>4Du4v|ynvfg(820cerK=EHx`-)P_>nqr{?(4yPP zavKmN;ps!5R~#s=UR zngjKQs=w7HnKlSooDCElIdP}6S~xza|DXzK;{9k{i>`8uRPXuO0kf6fVIHdwG2->4SwlPpg0aB zJg?qh(z|cy(sc`6Bt2S6HYj zn_~PCK|`iVhITwh=^W?~%&Gd%HN)E&hfJp_7CHoTs-xRvCLrkdjWGoJz=2AjTIo~V z+Fs~yRCtr7Sf~W5m1e}g=!2l`f3*nokpl@&)9pIomHD(O; zdW4|RLo~%g8!O-y)bC0YYSF;tJOX8LpxEzlYQ#*3?`lEN`fIXb9S>{{2NE9jpJq^l&-~GGqA3;< z9`*k@*G&UK8){||=nDtZg@vlJl^^k&LB5uzSV$KZst$S;c1O@K9bE$Da-ePSDZkj` z|NYo#b*B^yZG%sFlStb}1l9W=BTya(Qi9vNJ!U!X*Xh2bDHc+K+xzV1KPd>>F6&I7 zd=7NaABtKJ&Z$B-!z4MHVxfEfQ1$nu08i1>o_Rr_0uCfxl$THm!|%HCvoys*!bSPd zZO8GFL3vIZfxdE}k+3{!7G8_1(E`5E6bp@n<(Y5K!qGgtO|NT8=8Nh zOD8jvrdUX<3V!ft#hVxeJ#dl=?|6=iIFLEK87@sWDMXO1GflCOIlLJ%AGOG!I!5uo z69`ni2A#*aB6qj?zB^O&(7!5PMCpJ3^MBSc_PhbpEZe94ME@$&E;Pmdt9ssmKd7EP zzH`>@UP7R6oaVoVg^U9dmUv{XyoaV(=rt^4)NB`PL@lz4+e@Gl4s;pjMeo_V;|0N8 zZ)u8!F2lU2)ZH#u5%f~gnn0x-NO+QGqEZe1cv+03DHal*4;h|`L41s=dpepFe6hyeJMK{B*PAL|uf}TwBhK2Xg zrF*4bM4&Pb^b>lUTW?FqAjokZO|j5V=xq)jqalZ&-#6L`RL+5f2W<&LrsEZphMP3S zLc)W#1;L+xhDCHd-4z^Y89d!%HhjZ7GNKx2iiMWJ(;ao&22WQy>&zw4PY$Hk z1gG3gFU0F~Nh@iJh18m0ruVNbelae3ypce^IM4tX9=7hUAB6r!^+*qQ)bo;p2vp61c0%9G!6J4&YSHUtnqr}y&^KFW|4#*7x_RR>2voy? z!r>L<6018GL4zjH6bpsJD@dgIJYHYed8v*-wHzoDrj%W)j|@l9RZ@jR6aj`Oi>N$`f{EcpxXM9D_ zQEi%HAwT#V{TzC_H~Jerd9Z>&zd6umSWA5zHyD49&OM|l7TOGJsegP2+(OU=Q6mEV z;XuXEW%~5JRv)!USB$1us2I9TYjc;2AjoHjErA+1&}8U8kD4lIN08l4nqr~J(0_K< z{(t*Y$Hn^)sF4H3L7(Akl#U;QL=$L=h2o&kU}EfpH(At=O(4);4rB$pIgV&l;>yhU zaWut3R;_x@>gN`Gpyqu=~1A^l4MI&zFo8i8;B7vGYP&;gm|DoCJjG)6yXo`i}VQakp zuwixRuKRdjgFr1DXbzk=QZVpQ2!ft^(-aHMf%8Vj+&9^TphSsn1Zw3#gJ6lX-L7`VsvK2m5aPD}TVj(3c z^xBvifPbLvs7*G3EI3dZe2-4|>92xX^yU&xu}~R&kF<0A?xVj^U_}FgEIE)3>=a0D z`Gt#!ZvUhy7P5hz0&e3+*COc1;sH?|59|dFBo7_P_6K|Lbft+lO|g(XbRazgofz#)pi3O6AAFD2Hub$Y8qU>yGE>3J0o!!sCH=uH!AeIw~~9LRC*5v5@<9=n_wT8HQS<6K6pn2M&}AtHWKAPv8R?#=oU0 z7RrUy;gwxYMj&X^=(_}RCcV!Gsu}~5mOI4Tov@3!>nMV=m8VC9c zRTlb5iy9G>Z9!8k^cAWs2D%;EjG!+0c?5FeK*A%+&zzs_ilCZKDHakQQEs$y?G*&= zpVv$vXAU$SF5SWCp$ibSdp=FE&~&(TE&VdQ5tMRQHoD`1b>TpLptk3c-%4DCo^y|; zSf~%w_7sZt#+{SHT~!Hmodey6xofkZ>+wKLU6iI+=swI{zpPYfLM`gLLzh6V97qcK zb?Vt;aJ5oNrxXiGLBB4vU%4oPYU7R($c+P?h0U|OHOAu8bhF>m6bqe&&9lMkKk(eO z*>Gn9-QYkeP90H1P^`kr`B^^IB;f!sOJc6ea# ztdHt}TI4W?rdVh@Jg`Gd9uLHq&O_==$8+Stf!v_a;5TC*E`E4*o2FRE4f+fsYqBmN zXnOku0^R072jGZ_$E_CKQHy$tDE%+RLI>c8iKPo<3=tHzbqR^w;XvJ>lfS0c82>!b zxotGXLfxQ~-~Rh~AcC}l_7cdG0~L0W(AcHEw;r{qe=tq4P+=DvfrwrPF5owou_n-6 z4s;D}@1YOEHX&%;0GeWKHdn zc#q2q^|E6Ko#qV4-T;tF@L03J;?x7McbNRVuC}_-|y=*OWj$94Pg#xJJ-s zS6t39uOCgZQ0iZJ*UcAMi!NQ^5jO%o=0N*lXqm9nZYXL|z)_lFq5Uwl6v-6KL{P+s zAObz%KnYL++P6x!7(tOAX^MpspafKNXo`h|>qnEb(~l#_ zDsTmXo^zl)_)KiOScn&V4TETkh4SDtQFTEYe~(Q18xhE#18ssg!{}Aoc!K@33{A1n zCU`T5G#cSKRhh%K1bV@NdO|PgibS0rx^z{YQY_RHdO;q(fAG!l`hyRFUUHyMFas&( zoI4yrk3Z5B3w?qa$aiMz@J{N;DG3A$;6UO~t+dGa5?(8Jm`YPDBo5U|R+9|`=w>K$ zDJ4)K2a142kfLt)Y*CAHuG16?MZhA+xfR|!5ERiU8rSh01#zHN(64K_H2DsK?);@G z7Fq@Ux~;j#rXpzBYDEGCbD-0|p%eOk2W18m`_7t& zF5QD2=Li(Wfs){Xee17^tCh@m(i96N!2?@7UUmh75~JM-6wZO>!OC)RwO{~hQSBR= zVxf7kvb@?$6HiBX85>TZ2oAIo$_rb3n$93dW*kki&`KyTJo|Kj7iv-P`9o?_%XFOQmtV~lZ^cJS0t z*J+A{N}`|uX>-9(z z97CW44zwJ)wfiR63`WqPB${HO<ikU;M^kf<-zBxGOgf}khUX^MqJ zeW5@_!XK}t);SvxD3JrLhZd#jyuh!KFc+F)q4m(Bwsn&WQH%U)EeMpvfrJ~5wxntP zLM^&oM^h{$+;Akm_~U#8ZP{>_K*=2FTo_C#kN^G>LCX3x#X{%8tOXSw(=-vJ5D`V7 z6dnY9vw)(bxQlBQT_DD0Y< z^Cj8|wMgeeGlAZ7AS+naI!K&6)l^_=ch5V3B=y+h$IMAke7)-SV z;lb4PGMZwcP4U(OpIQER!0V%}N}zNO)1OZ0(#?2GQ!KOz%5;1BWRxJtqw8@3edIuQ>fkt?iG_GUuuhbwSm;h2 z^kf_p@#^s89nJ*$#DQkQU`p$>Io_nKx|60@XeJD%wokcx5x3~g3j$?uplIQg@`w7~ z=+eE6p(z%MhACy$zwdA{-axr@0)6H{zJJ6udVfy99gL5iQY_^A2P(>Rj^IVxy~nBu zl*xhm)x&2(J`I;rFFj6EEYzNTYB$)p(z%s zj)p_IcUb>Ieg zCR+hlI6riyDHdvi>B^qF?&4#6kGAY3&=(FQ9RcsURHqfFMT=T#iiM;jps4lXm;iL? z_H3{wP%a0Og3Xmz<+tMs=Vkgd#X?fBx$>@|2p)&_3-=~a9tWBTZ*SYhl2cHN6eDPg zg(kw=JJsmK7IwwlcgyZvVfx2K6^|m5Om(Sh(HA#D7;Nv zL-kP-9`M>7rYRN*Z-ZmTqYU>U=)uQ!0)6E`{oqr6EqvBIbm@XW(G&~ygHQPvm+Oxa zbbf|>V#fnp$brtl(BtNnT^kW}c_vM<&>0wd7=#|h-=n8)a|u+$fwsc~8|kpT1wlh@ z(i97AhX;1y%R+p-n^WsX0u^(hu`pM#YDhgES#N8jDHa+Fa|KOBqqWhcySCMoK;Jmf z9~g&vYnLdZ7VX+bQ!Mlc#-Zox=MP3u;~O^um2jXos5RYveENC>{f?n27HWf9)6g5o zHy}uJOfZ2;IZ!>!6=<9piI--Ejio6Ts)xCP@k=Xkslb#=83g*yfkfeExDeWW7PUy> zGEK3NDBKJWH;>0}@1gZ|1p2{&Vtn97^ctn`RG7?fnqr|CA6TbbQSk^}I-L!DlRBQG zG7i-IT3o~G)`N?n^s08K;4t)*9(=6Qw(i97oz+8dhLLDUpDIGE*&`%EJ1dC0xwMMK$ zmrmK3rdY@c7Mo-Zt>z#|^ph=tesLfxxV=|(HTr;{(HS(wLRN5l?>Xj&C--D$Jtj~k z2eO3G#H*q8_;|O!ol-1h38RS_g%S7*`{iZ=fvPx=@HbFulfP@C7A4%GDHam`28w0c z0=!?cwzZT%)f^}tE}d55PP~v2(?(M)ln$3pV|*q48wKf!C3ieWH5^C>rv0CM3CC-x z=Qhw33+ceL|6m)vZ|G)-f38TNS`KuuRa_&~)@mENbngB%#X<*LVYAYJ-FQ{&x0EJ< z>NwD^68LqbgY)on6x)lYSm;*?{Gy^&6s`xh-nWfF^&BV=)))F16d=xf|5sxob1$Z~`@OpheINYFgoo`*qeXG{r)Tpcka-*&7#nMO9}L zsF4FHLf^aBa9vY$=^ob56bmUr-`jBQIDF}>bQ=lumjkVaNhkFJEj%%}L64?bXf;ea ziO4G8N%DOG15!F3*nb@81#C%3H@W48TBI6CQ!MlXwj_wywW^`J?r$#@0yS}pQ0Asen(R*qy^`Vq&d0aiNWL%4g_lDK+doo_-1OkGlE)7P=B+El7Vj9iPBIR43YI_zrry)qH z7frEHeE=K@JL@t&k1_DjLIPdjK#Sng%_{5HjG$;^nqr|vaOqS_gYXKzQi=hAE^;6- zcm=HpoVyZ1eN$+(iDypsjH4+Qa)r0IZL0?U z!uGSeOCW0w^bA&qgWMYNroW<2DHeJLtHZt3PAa3jZccF&fowRC@Fb%2xjR(Rr5pK; zrdUXL649}M;${S;EyyR(B@UDd1^hd#rk_HP*g~3Op;Rc~-#WkB6a=~5Zzj-X4kQLO za-#;V#GQN-Z<=BuF{qKts`bEk-MOxE?>in?TMjfCJ`;Yo3YVi6Efb|F7Mcv738QHF zuBb(CcBm5Q3I{p~Evk2K#KjN3J86oAPC|>s>wB3aC^<@(Kz1A`8dk)%Z5UXDTJ-!i zO|eiktcck}&&FTal_QQ5=qd+lf}7!+&RKlvv=nHHg__`IP%F;Er2-$$I}^yB18spz zm+d_dpT6R4PE#zj1umWBx<7bX$20c@fgCtcI=mVB`?Q}xm(D7WrdTK)-VD!bC2@IS z~9H@sVl)I1oD~%x6Yc$0| zJw$B;4O99oM0ee;KT>HO&yf=c(iWDkwV!;8pmhy2#X{P!R(@grWqc@#*k&aHIdh;u zm|&lOU?=W-S9eOWP#{dO-(U0PFlv!{)Di-@a3C`nl`GFyTY_5D?KMrYkQt21hqrFS zkp8fJ1iH?FmO{zR(vWFc2%05NQ!KO;N^V+=9^)b+rBOy32t?;bu@4|Axm$SI*EB z3yH$bAXv8#Pj$Y}$so`@4)hfkp#NXx0- z-s1f!4@IXC$eROwht7$7>ri~Eo{1PuvCwzuoQ&K56~Dcecjyr40SB4{Q=PpMRvkhu z>baAqSZESVbxyeKy$bz}tU`?l^pFENwTWv?pHgFmAm=ceVj-tCC_EnFv=l*#1FjJ0 z5eMo6Q#6O&)?^^4tW%1G`oI*8noiy#T*qkom_R=3P#xoz$-3$iz4h=nrPY%D&;MD+ zD0NF*!&vSlu48<8fu`7hmDDYm7u|Q+6E)whID!1D9&?)C3;NIUU03O$=I{DOQ!LaA z`p>3guHnU|!SlZp=m`h%gQ3Ui4I*|3`q3%HLVhsxxO)AdG-}a_TVfwO9=fL-$h!(A zj#T`lzz)HwVmok6@4j}LVxbDyA-E>!s4i;J z=slVQ^5sBvu#0m*{iuhiMT7U!6bseCF3yunuHoI#W~ti<Q5e|6NNgX7Ey^qd2Q!9qsz<7m9ZxnMj^u}~N+WUL>%KLJ5!E?07>>OS` zyQrN*pqCs-0P{K977oF02FoQh#XYFTNQD-)kgL00+`(7S~9z@-aXy>gGjL zETq#6zm9a}BOV^wi46SI@xTUhAU&vKta0n^g&?CYG{r)CP{(+E#|7M?`0XkL3gSRx z0GbkWvm8NhchD3IjR8pDcL`T3-H%>FpkNLpJPb`e+uaC3c5i5kg@lKp%_#A-MYp$^ zya|CqIFK!rMGbAMR6@|I;WWiUwon!&HdSgog5pgb2o%bJ`oT-pH)9w+x98Srnqr}T z@KRk?J=q39G8w)E3gbX+P>i=o8T)sXVT{?B$u>^|bK&RlYn>%6!evX#u(G&}v zg1hd1RWv@3p(b=8fucB&2~4x>b4s{_S`-~dQ!Hcx(=2saw_l(ZISt-Lpw}Eo8Qu(+ zyQkp;8Scr_6bmWCo1yf#y(EI%&R7yCngi{Gs=xbo9(d)^{wz(g&`zlO`*6?!cWeI? z+$GQ(4io|vWyUemF6h$bex)fE3W18UOLej{QH!=LdQG4h4%7ghlO4B9RwGDxF-@^h z19VP?d~?UM@uH9O2^7nL3}LQdX_R^&1eJ73v5+Cm6|8je8G>4*)vbjc@N;x{qul3?2lg!o%9oVTSnuuDj-XkaXo`jMC2a&n1{w;e zMe&iU1d8WCL9p_uFm5F-ch8BUDHaNXmB${J^6^fAGOkOdF#H(5o3u%gl#=r@arv^W; zM=dh+N+(b<2hxJiMC{tZcTtPh+@~oP(t^*#?9A8u5wupUnm{QWXa%eRjR<=>9zoN) z(G&}e)u^G-A7X_qzaer*%BvF)FRnWT=I!GV6k)}Ct7GCZ2-TTfFg^b5B3bbEj7 zDuQ&Ei)3{?N1r*+ZaCg8qiT*gx^#*wXo`h)!|`sWQqp*T${cTb0%dX_bC}%A7|{ps zeHi$FrdY@vCikxHjYvf;QW8@qP!60jvf>FHZM5Gz%!Cr~~I zs)bjOysh031f|ZVDHf`QSCI8gNBo6NxzRVfVkk z55Ck#mu^X$GJ(Ewpt{%a9Nm$^m6`q9X^MsFURw(mJaW^)&_*2s6>=bFc&UDz{1dMZ z+ijvL7IKD{YT(R5e*_ss84;+613AV+S=8t2cy)N;Ynoyq$9ULD-Q$8Sf_f=jAy6>~ z8UoKz<^?aj+r6|?iiL*2bF}x)XS@S+r`clyed9nM;ZyGQZ_{RU*KImaQ!MlmKIOkY zkIX|YipWnOPzeWG2NN2D3zo&iqcGQVuj9-gO=|59$!K za28Fm(0q8;IaK9%pcc7a7t85*j=poCI9Pe??*1+rLHk^3iiP4}<8714$KRvti)o65dO(?O zaPD`!OE=~IHUgD%p#E^z85O2KMlEvnrYRQc4|iRivD{YlH?k8mBTxkgItoYd8M>~) zqlrD;Xo`i7!V!D}PM*il(e<4k1p3K=w!rL`mf4MKs6}TDXo`il!0cA_+5r5nTaXY= zpkEv)3tmA3hqopmXz)9lVxcT}1*JdC*^4gSlTkSYs^maD;CnPRc@|y~a~@4oEYt(O zN4c3B@M2=|#YO^EaiG2MJ^ElaHXF6*nH5d3&|df+?e9^9mvwxK27KvwV5>Qh7#ugM z_QDX4k8T#z6bp&Laieo%ilfk_yP>5*pc)Re9fp=G&S~Q1(VYuviiNhr(6ZObKe-6< za$iHBS`HKdXV2EihU59V=N>f0LIH60te!~KBLo?=9wty72hxI3`E;c~JfWf7MpG=L z1*7uimSwLI-ns zXc+XpHO)+L;qiOHSOWd!KrZk-+H3Q48EVn(1vJG%F7Q2?Un{!~{f$1{SxBIN97tFR zGphbVKLiDP(i95`D`7_26}2PiTdM(qnmAAa+}`svFX28zWE)MfPyyWDQ(o(&BgjqH zl0eNI=pzg*-Pe>?pcd`YqbU~p2t!M|o&tO`gaq6rPzwk0f+?EopVGt-bRm$YSjY>e zXdbU?!`m1o2D~OvD+js?pK|~D`*?47d#4l&U4>8i8K<%qbm>}8UB z)~H24OlgXR8sH1tK2h}yf>vcU6R4d72~Poao0GBuK_jwhiiL!yfM!lPse+&?Rk^&5 z2iAQ3e{WW5g#$NCPL|jsD1J6gu}~`6_RPAL|$ zg4O@&cdp~Pp!$+30@-k&*|1O*S#O1xM?aL(6bsFUg{sg#67r};?-%yU?|6_(KAvLV10@1o=0H^dy$oC24MBGv(i96-0kpPc_c7F> z(r!x$WXpjZ0do3~h4-gKiqjMeIRf-IC)5emF)rJ)k3d(}qdLaB&3oqh&%S#E{j2m$ z=>Pnmb&ST|uxNW<+nHa#hofS{U8 znqr|#&}G_hV=0YV)I8?_fv$2O+ekQ@RATdC1oc*@DHgJgv=#)lW#GxZP4{C7WY2*f zg}@Pf68o1T$jFDVrvnB95_%LtS|Kca~L1|IKGXhSf~xw7fj`{ z9->P(f3rwI$3y4Hfu_TvZOozxxLZ493r(@mbXc@iN=lbPP_IaN0$t-kbKzMSTj8RC zE}c#kO|j5icovTLedmTCYXx-zIdP!8Sg0ErzE2K8I}~Y(h4Nyp1t)e{;l9~1%S{Aw z=0K&euMfW5u18Q)*UtoU<3P*c<8@q89)4H>`e}|5uh2nYcmiuA(p0CXd^&< zr%uyD(2-$N2y}}B89>#afxZj=(W0yJG{r&&Q1$m|N7V)dJ+jgvkUIxbe<`k!QyX{y zwaC+&rdUY*rL|ypj^RWEi4+(S$b$ow!Q8d=@NSn8^t)4vh00*=dc5%^TnUq*euY4{ zInY@sJRab+UKc?{^Jt2N&O+gFyYr}Ds6{>QJtoi{4s-&h!q(LdC_zw3rxXjFfT^&+ zA(~|fy4vL(fjl`-I=mU~4phL?EE~Gg6bq%po8gpd5$>F1ZTn83yBtXPE6js*F5pqupBoZ55Va^JL9DRjIl9Mz1TgfFdA$7zx^({UXo`gdF!We#s)9GbR!idK7OI24k-^Y@Z%~WWFKH6!J_i~H`+&+azNjFm=Vh8=p>ePe zXtatv{=z;i+D0I64s@nbTw|q;?tBE<7t<6AooR$OgUd|(H`=z?j6e@K(Ay{|;Lq%{ z5kXV6X^Ms3M#1j0%RBL4YN(e7fgW<8Eq}x{s#>;$BB=L$nqr|Xf8fvzb9HU>H?nRE zC(t7f)B?A+i}TG-2-0h(DHdvh+k5x5zk3n1Xj2Y>d^nKsfS@0R&v2ii_hy=6A>jc* z`{s$DXo`id!n~+$Zha-XbkP${2;|FwgeRz+2(g`mAg4(*#X`ap zR3^SN-GZP|R~!iB$AMH~G%+J$VgQ1EcS^C4DvTzqyL59zP|QzX0zKzI$}pPvH>lrz z1l{;WQ!Jzmqlpn=`K1VQT$(~4e-0!EJ=J};>hW2Bo0icO3&}xGwW9D3E>H=4SV5o{ z9LT5wmd_rS8=@9jJ)$WVGODl^Jo46dK`pZD-o3cvIeN)~o*6F)VUKmi=+$_@DN)d?0NsIpUvg|6Iy=V+)m9+j)*EhJDN2Wo+N(e_DW zHX%qdpQc!-1?EM~hWO!OOAobO1PbCnlJNEp{dD3Fg8Iy%DHf81x3}V$F^AFLsOKF^ z0tIuR2-qpmeDK&(RTCt z2L$Oj(i98zX@y^~N|eJ3>v}cL1bWSZdVdnv_^Fcr1wkWfX^MqrfZa}>*gc=q(3QgIVH@-aGIE`$~zXSV#?KiBl)vxQAL)dUYRx;yKU+=rg$PSQv*|lx$B^ zEHnZ749bZv?-BH)-kLxO94Hq?d?^v<&LAlNH%+lnE{ymzZ3^QNq^bLWK<_xvDp(NA ziFx9Ip#FL^#X_rKLGab~Ts$#Y5FAUOL=F@Wv&0E*DMJwy8bVVn6c4k+xA!>VcU^q{ zVge;`AbA+^DQcvKA;?pPrdUWGMtnm;wekKGi^C$N9nVoR2U38Z>Xhz#@dImVLQ^cH z06o=!=UV>gZ#@yA2`rhm~<+3yoEO_DTUA!3yp1QKoR0d74&|Vl#{Mqsx7fc#Xo^i-o>yhhMYWtw84 zo7L8WeAQ2QzRv%W4uLW`kUf0(68q>DAn5*Onqnb)`0#y}-JOXpT|t=Cyg$hQ#4JnPy^I#f9(By0lIXrMvf#+KWa4m2jZp@WB4bcCJM& z5_?HgEHoS**nk76c=i8H&w*tf4{Rw13V^$=K(&2-5!POP~r4lp-Rbv1Xb6S_BRKMN=%4B4Q(WGG`iI9^Jnpg+MsmXdSSaodbWU~~ zKF%|97~`oT*s&oc2Rxp zlq9_S%*dSn&;MD+*zy<_GBUo;L;tE_7Bt2Ft6CnzGPG*?A#~@o%klZjhRkUEYu`nBk0lc1+QxP z*;*2)mIKXzJ&##2Q}E8F%U5WMg=WB>$CHLCt5J(&zTYKK9S4$u&#_X2iv()X-%cqO zl7Y|hJ3Y6*2pTJRO`v)X6a!0~TGMLrCgp}sDHe)>B~G<(a=1#l-`#uy{pLXKuocnx z+ZhYgqWVrL7IKHJh!s=BdZ8B8H8&IJ4+lzs7RAn=q=X=iR+?g=1Za_u{_jErE!`nk z(eZRQaG-JU8rjir*f9jn+DTI^G!9-PFXx`XvmXWTRSDF{frj3JbL95z!KX!iO`|Cm z8hQsxZlrZ&#)**s+7Az0+Wogy1A+doSCf+z?Kg0;^09-5m3JRDK7LI3pz*HVMw1N= z96t7LU@@K)?5F#`9A_uTw`cQ=M26z6S8qF|*neQ_w~3n4CA08+YDA>W&yGv}k8>p| zLc}$0_msg!6IY^WiiIjdtOZNvCtpB++a2=L3Dm@a4#3y7>)kIh=x@7fI8Cw80r zjWfwYP|oGm1Zw6$GvH;cu5x5Mg7R%?iiKvt%Q$Y_?0*QFQD;n`77kPc{q8Of4@MzK zzMiI7s0RAoBZhR}grI#Z?FrP%fd;`M)xf%|cphlxDw<-UL9j^W>K%)J#lrUOGXk}7 zppY|APPP4h5o%F%JWa7s$Qf8wzHNvnoSdd76R4d7Ik>`aOTH6PM3AitO|g)JD|EGV z4lP8NZi7QPfy{OP`#Dmt5!cwR?1Afur#R9S3#r$@;ZOC?@m$^6T8UpB&yfWO%7@uU zSBveqMf>V#iiPrF_VM=757MYbrut(DWXXZ%2f=sj)iONiy?7%{vC#Y=Yr&nPI(Q3X zw{R^2UEo0f;DNooNe_?qDm$fE=pQ_=wV@loq86PQZa|=m97qm+GyKAhzX#CGuyzDZ zv5;IeEEilK;f|p8a~1@$;y}jm!CbrKNil*(ou?@lGKLRkw|RYW@2M`&lR(xS=mE^9 zo*pTQKbTqhG{r&>U_SMoU-wqjqTLIl2xP;7R>2@lIcg?yN5@6 znE`nOy2OE$Z^G^Evm+i|y7z%J#X`zAtp#o_4r5V^OoulU=rRW?@P^G91LYnf$an-z zu~30GEME`O!apat!(6tqit|b2erubJ58}r8dM9K+&n3ZE?w7Ex&*T0K-(X|HdB=}{Dm#(lwzUnk6^T? zwxkDwf}S5E&{Ym}0_Ibr^*r&-VDC>;EOY|qQ&;aU?1`XV(#{03=Rk4r3VIRx^Dw$} zTl&%z3&p`J=uuQk2!e7>ydaPR2l^NehXgFUBa5JtlQhLbAH(6K{JMqpG zFO5#QCRNq(9650yO_*?U82u~+L5-bKETjn&PE)%JaCb|-Y65|rIZzL%a*`-l#V4-} ztEQ>{qwBuoa{T^3fSZ}wd&?eK*^;EpWM@;zEW5}qDw!cGDp^HKQW6a$6=^F}B&*O8 zN$dB%JCDcj^!lF9@weCG*}3lfx~?N8w)`-gq+(|34Egtk$i;q2nS2>`8YYR#Z? z63B=4g+CkI6rT*UoAHzg`Ov=bvLP!?0cyAPI)lzjpq^A>(07n0FS1`yy0x71VH{Je=2i^TT!&4%pn2J9AdV3~Zy0HZvb1QGy zOA;u6p57DFAH~3DH24=!iBJGNy>A%Sw*jc{0(AymmOz%lN}8{H2je2uw<;+SvJAG? zTKwb!c1~(OSi~Sl38Y5fx`^LVpP@xxtE5CojlOk*X7-&4EjrY}oIy?!s2$y~nspbv z0%%D`o)V#Ubi-a;WcCrD_B+or$XNpIqQVx>st^kdK! z2{eaJvtAUSmk7|VB%TtXIdq!!#yay?0rY)9B7Y8 zQGoKQq(n$pF)Fuuo(hojw3s_pG} zV!cU9m6Qn8p@-`7YRe8li++_aW002wIz!_l!!e(FLyMYLTmJtk5jsQTqqlvwUWyPqBg+&9-IYMrwDNzs z%|91t(cptTB|_G;^8fs`#TIChR^}fD`AVQ%n$Wn|SjP;YmY;Y^gmP&@quMVkT$>1= z*s`GVhP@|&?4p!3$E2Ra!WNH7JS9SQw2gSOh8F%zbb}je4Dypeir*qA&WXUP+l(qH z5mNjX!9|U8xPGf$dp3joB~TXC)}`rf{t1^(ONpmMD2rS)8g3n0Ba5IDMOQ1Ab zNt+V6$`+u!Dk%|4qm{JTJN$0}lofn}K>-qIE4|A1oc6`RvR?>KiO^Phl}~Y>_8Xvu z?K~OuKmy&T*Tk^>^>OLDe|w%1q5JfjaLLl^2G9WW7YquNK+EXMaB7qnmg+d~;wceY zMqdVx!?{ZU(n!r>&_f9{R56pE{-Q5@Mk;AMB|<}~lmGkbR4i0&SwR zpEI^)M*wtw98ZbRCOZ4MM;)UY(4u2bJsA`vfi!8v*CfpnYwNZ+^OOi_(ui+J=ZBU6 zwfU#bpkN6!l3wLC{O@7aZL1QV5}}dwD&Op$jT7wJOII=|L<04v&uC%qz=m+?S}x-$ z5$aE$QLFX6a2LqIKnn(iN+3--^|0%O7AFB(_mHPVNRv)IJl8mLJ6t-8o>v(3SOQJC zNy|EMI`}>^?!{9gH07qXmV<7WNdP5V2Q%o21X3L8nXRjiCmDp;@RSHC4)s(yWPTQ) zwLelB^i%?^4x)2~A7_-or5pK^r$lIV5Eb$1WX}i4Zej_8!X%Jl?@{yi{#aYLViHe@ zkYexAgE5~20a|mbZDHjN`%D6P(GAdfN<1Y(0aU$SP)i-3-p5z#Gw8VlYDT>vRo{!309w6^~K@k#Y0lklI_iJVdP}^rbB|;16eROiitsVfa=y8TYkrK!}K}mDm_Te4?jp)f! zB4nOGr|_B$G6txLjSqvOB+yXmsm5OK5DidTm6Qk#rJkzg-R)TTzxrbggI-9W40@=# zZ3}u1(9let5}^!wsE)pQdMC8VLnDVlFD1}ynsh4jIr9jhUE_F4gl5yEQ}@D)0|4@J zZ1}tKK8lt=y=kgbW5wtv0G)Q?DG};TQ=OYX>*6l?%`&S)$vfGMT@5}C`JMe`J<%ySbbP^fch-qDG?g-$6D)|Ph1*Yx*maR7!)gk zo>4F8{Oyo)04Y7>DG_=`y`a5C`@ci|W8HSv42si-a~wn7>8q!$v6%yZm2o%zpZ~M| z@q;IAob{Yk6aK2U-FZsVcbJye)Z3gnm-< zRV|JgK#LZnJz-FS1ahLAZdWs_tI(qU={zMuPKut)*)zCNY51571|>?MWSZPF`PIw< zpbhFgB|^zGxmVIS$`M*r-?4&0NfIcFo>#W{iFE;L>BLhalts_0&VQ8717uvi)1S(_ z`?UmGN;lo05wBGN(yYN#BD9oly5it_O#y1RT%AG563BukU}Jly;=b^820SG~7Bm6- zwogo7fSiIBF(^d>DNbJ5srSd8E+7dTACe&4f%(WKgOEQh!S4E1V7N1)tFaJ)ROF^{3Wa zBfh+S3N6z2C}2>U1Tv!^->Z{ccQ-(zJb6ll%;?AW?uYmK2vCE%O^Yh;qjU+>g+|t@ z+17^v%BqqQp)NGCZc=mBX@E@D4P?-J2^3CGh9kB!u_j^sdY%%YaC$O)y#Ei^7b?Q1 zGw6c^nnX8ju}3jZ?&(DEln70t8}?ADZ+mEwQ4eDVWk{eTD)f5mWc~{--JG60B|=G5 z=yi43VEi&Xv^m0{j}qu;w34Q>hRH~PE+6J85jq-8=QtWTr?!iiM0{$qBfHQEBRP27Qu1&*_F;n5t_I(B~>C5qeHHtmcDSSgoXT`!j<+ zOQ4Q4DKY4A@(+L-yYrL?b)-p&U2zL=D(p<-T7N6=qc0LDl^&|$uj*X^Xj>DW5}{Oj zsLs9}jW1&(9Jxa5~1_-WQeoh zkEPUCX9hCphXk^s+4$*slhXm(KZ~bC$ckp;TYl<mIdtq8heVT zM5qZBKU_Jz8s~H3f1F^@FA1bga|J^S)o?W7@{^}TNSo#gN+&vwgBEp~>B*o%31mbw zy+2=dT?#E4F^i`}$cScoXFZ6;O-@O^FBtS&0?mw|BgblO#7&ug_jpQ#W=2?R9cz29 zGqlK6Ih#R$B#;iB>fxu<9rv^DZp2d}q(i5AJV=}436R;Qy8kNgqaq39PW`%_7uVwr zJ8d&hiI6+>>%P7FhXpEWFM2ZQuLL?q>)q$iyu_s$-^y>%E= zEP*!A_Qxly({X6os}E0!&<5K6s8TS)44|&(Rx;?H1e)ilq&dVtI1avbEza|l2+eb( zCC&}K{{Uq4*MdPM5@-r_AP1xy9sy`fF;9um6zV|Muyy+m(4%=*7*r~O&QssJR-V!b@xJ>@7*rvF8c=cg*ZIG90c3xGr$nd$6^FZN zr!;~Vo%r0gr1FM6F#o?dE6t~#>Pz>bD*?Lsg{MSlKJ`?Kn`T}ID0b>d23bfT{U@|E z(<-JJK(D9qlnCiRp}i03vG}dqc2}Q4mJ+D&o|0x!YU~<-Cj0V~2o>J5)@pd`yBoA< zWg}At9h5**X}+$}tQYuYuxQLvA~coe>)J)l9R<+TEoT^XNCIV1@xvVLi+D6e&#gQq zLRnP&@X^_71wb3(d>CXUf%;Ij(l(_KJbqp$o~J~p4^=BE{WisUh7PK+46>F$idxe? z-&F$PGs>xw5+Oyc>GL1Iu+OmmL=J;&B+zE6R%)?39Y=hNPV$rpZKi6awbq&w;L>gR z)3CJiJ~}LcT)47+DE0-*1V z*BNwN0v(`Vz$h$>#V&Dim6Qk_pkKf!sND>o47r1!GU$W^8l-?CK4H0gA2ps5p+N+l zcyekDT)JYr3u48$TUBBkRdLB|;l$il*k`1-JurlG{KAIY=NU>U*c@ zExQGf_AQk>jURsa;N#8V=qLGyJ_f*WfCwA#p+ zL6;>^Fg?BNy_D~8nsX0J-VMiF`D1p3b-n+|rLmd8_g!7aLdDFc2 zGLOW60PR+}#ULjMq}TyExkv@yM-#j8ln5zyfWH4W5I4d&Sv+TuvjjRukDzxil5vXW zy(LeH&@p-h%?=x60xjzGNFXoT^w)0K;;!%+HU7#|BIHGz{<2GbDgZhbv;SqZobVC9q(jrK1cf;-Q8EHiFln5o#BFJCc+PH-MyZ2=V-IPH2 zvGjvuZ_2QAExZp;iBNtl9j`cf*(_+$6RSW5xk{h{8i>vBICB@Y=&ChOiBJIz#CmV{ zISx?QtP}>hNuVk8!`HPsnXUyWr%FnMri5E-eGfCnPQI3I5rb|?Aa9!8I&OLaZ`j_` zc}j%5X?AOrN2gBEBGWr9tEpa39^d4)1ZquBhToUE;qvHeAD$AS*7Ri1J{$M|pp(sp zGRR#5^`?SJi)uNz9e7o9o)V$nR4|$Mp+3Wuy%W2@XSBwIr$p!p z^%?AkwaEmiXXIuEc}k$`G^g6oc^YGVFm zV9*^2^gB{X)97yyen#Cs^OOkvj-*4T7ua3|sPTks2Kh*!Q#2}Xa5!-hv?#kuN`y|) zs9agwF$AD#*XvfVypQfmpx}5ad%NS;7@#3HcuItV<7wt3?AA0xYE5Gtt(r+m09`BKDG_Q-W1I$!dQJx@Wp)~a9!Q`l z8spezzk2}C<2gJfLQyov`IM%PdrVpel`tq!0*!O0cZzB;-U~%lQX(|YoyrTF>{Wvn z<#cLSqw?;4D1o}Xq?!bs%lJ!-0iAhDgu2lFg}8bfYXWp(=ST)Ul0en}QW3tlR~B5l zmF7GpLe>A$d8o~FaJ_qZoIZnsB+yE#yDkZGJ_yi|c%BlWl~i{f(`<2DXwmlprVI*} zKm}Bw(xB@^oMx#xh^IuTfC^NKn>X?TXvv8)3<{AzuW3@^S4sg6J=9L}lnA}1NeRQS zRT}_mmgB>qPzkh}2D}#A zo2Nu*FTEuV-v&MfXyTn520f8L?lcay^Kj}0&=?<{5+QdQhiZhce+-a$W96Ea_t8@c z^qL+)d+!_J@2$;g!c!vjnjS%S=Go(p^(7nnF(^y|snM#|<;DH6+JB@mPl=Ekt!j-a zNeqJ)&3QS6LC++RE)5T-pK6OGpu?hhN`!Q2cz7Vu5l7als%se(E`bJ8$xWv;cPs(@ zS0yDv1F7U@?cApLWO#3D&7kKJ$cj3FKbom6fiJ_eBRnNSR@4dHSog9IT)H+n*BKNc zfr6=8DeSjPK0sfqq(mr~s+C54iop><^=VHT6e)qK)2n>aGridWb<*W25voqF@~-pF z&x00?zL~+GC<*j7hblAkR7U{R(v_z~=xq+I4!;P!4baz`)oNAVM=vCh;=GacMLm@O z%C5yzBBVHP#7*hwK7ewUcVf^>3DlCNl-FFj^bDX_1D+D0mNcc@a!qPJKoueC42qUO z$7xbxk6T7DK&hcTB|^t(QX;!{PaIhf>b96cuO!endZ@m#+gJmje^pW<^o<^>&vHx} zLW`DJn=>dz0wqzOVRrw49|6*};VBVHqCUgZa&@fP{+)fEL9r5O6&3IgSyF+UD`UU& zlnAY&0{+xp6Q@9n+{XJeC{6<1u3ks;=UG)epL?GcPl?d&>NZ-Ap;I;k)Xgc8LGcpk z4wXecXnGf+-&Il~bcf2KVwE4=18C#l0tO{Wpkz8*si;;bykX}R^OOiB)7eTrH;tqM&T7R4a?gRAe=|Bc0NuWrYp<1-5nlXIq zqQiJfgd%B%YJx)uF060rqQ{`u5~vpaR9$Yz)7Aiu@5)mmRO<^BOx8Gqt6IDF7&9nY z0y)y;UTIbfEV-Gsm#0L?ktX++)tHY1v52H23`&teld10=*ZCw)DZhTrQzA5(`reD1 zS%ku8lsNbngWgD>5PHKVU0aW}Jw8KtN`yk_4V!J2?E=t+bI%#{RsyY{$-P-`NBYBO z6nvhiL}&#~?)?liu?MJU*=Gj5lRzgQ(5T$gbtgc-s-#5dyFnfCh6Va?M#vaOq5D7cuCQ1lpUgq`BNUE)yV~ zIXopod-Lgsue%<{%FMp^TGpw&VLwZtXS9%U^F%dAfGVn_MCchUWc0mg_6Axsy46qy zeUU(cw7$^V>D^#}TDImX5elUB1>L+Ug8|yMb2fv%N}zFyRjmq-P=FSj^OOjUqgAaw zJ^tYKD~)%X8T3s8t))vhEPM=BIM+|*DG^#rmoER%23&^zH}WKdvLw(^Dk7TP$_($L zv{5`ILPx2HC{?eK4qUn|4xS9kmOw42u3qa(dn^?=afzoys0G#42Rj@=D7EMXgT6~3 z#bMELD$^gsXXN{rr$k6`SoGAr?zjfDWo9;men_BcuIs)Xd&aX^CDb^Ue&f2gK{L$n_ILU*zX{Y zCWf}-DG_>e%UWyosSRVHMdSDAFeq07+0)nC`RT3w(4rxGc}j%r>Fd3z#}fQ{pZlFXUn*$S7Tw`uc~2yLaW z_uGVRkpL;*yuzRY33P?lfa;$yZUdJtuS!aUuFx9Lj+Lg^Qyo!1gh9U~&`yFv+m+)y z!>|TCB|V7TeQTKLB$ei*=;4wX@}yh06MUgr$lJkZCabS{O|%mZ=RfC z&_4;}@Q{8&?dw@wZ3%wLQzGQ>&|1qmISps>%e&uYP>BQ@N?UtAMSi&jElTgfQzA5! zw)W(1D8*UgoCC28DwRNrVT()e+p7TzvEV5YQVd&MYF+LOm+pOP4ui@hkm8rk%+lWD zu*D;dr$k8c%Vv?o^YEapFC&%fSKdeE5@-iO`(B+?ffgl>;wcf@K~UWBpSWr{*{L6c zDkRWj+Cx;Nd~+c{t(! zN;Cs$(0b${fP7c2WsrpgvY{u#k%8TCPW6-#Pl=F?q64|JH||eKduGibO9?cD7C}xd ztb-F8p5Z(tLPKZ~Wa_3xhR~v)J#R4RpajaKfmp}9LN)k|n)T)>5z3^2m~M0)9xAx+ z=u-wAl0d_#&oK1WcYGfiALA(z8b*DFE@hjs&(JGU%`bx<~Wg{RZgcL-l17o)V#ZH1FLspynKa z_L?nbkgWuILzA})XZc&gXS8?+Pl?bQn!HVF5;htj(g7&jYBK$UI%8FWkn+0rcW zVGR#FTqv$eN`!1_mU!Z&ntE{Q&VNp1&~XVAM3?T(4nR70nl`MwVNXgREjkCIyZNs{@EP^< z;VBW)qH{0~b(@S8&Pgo?G008=714{Y_JbW`0g7nFQzBGEFTOyHyN{tozc%VI=#&Ji zL8~nr&HL{I$Zr!*iBJt%ZPDtuY#~5yQN|29ErGt#oa&?3k8c2E`hurK=o`(c>Sv9< z1W@AOqYOGDfd*0s($=&aRyw5);VBUsNF7MCz1iIWYJC0{gX|^H85&I-^`ET|E&5R< zB|>LtG@-S_0IO&m|2${VSqbzqg%%V4-oq&xlOmoHp_eJvS~Ito=tGN^>VIL-ISG_X zQ=KEG>}w4z8aVZ5Z{#C}*Hw$|bwg;f{*4@}&bwSen@w9@UXg7Q}G{1gZo)V$) zw1WTl$a4IOcG@|fK^G;E3w4?N)1TnoU0fw4LN3&0I(Kg={s4i~>*Wk`kU&#uhAPRw z%VKDeRWeVB&{UeCvYr`(OEY6e?_cAKpN$=*yJ?~7n|~{q(n%goa#pQY{!oJ&#XWOIZB{(>I5d-Ie;DYm)SfeLg~~A zjL(U{x$A}#-!RBY0u82>$4fs;m&2uNIEkl3XfUljruGZLBdqIOFJh3h1o}W#f8{OX zT>sHmlp~t$l zLmA{EfxbPVk@Y!?fdI`}$5SHojZS{Zt5f$lJQ-plXEW%k1gb{$z~9oF%>d|56i73;{n<} z?InY5N}!hX^mf-+im&5cx;!O9E$Qi9&(R*Ab+(?_404q~(`Z#I{?kd^2ejCWr$lHP zt!mv*+;;|A^siyP#+CPxn*?L$ctWlV>`U; z1W^3YRSa^MK&|Q0X@vE_0k88go)V$fbm5t?o5gJVWx?Ng6`1B6HXv&~_66hx)iJ!wBK{I+CDkPCPVvQQ z|FkM85n4$#3I9d~;K7e|LSh+oUjn(%t9;amNGvbRtdbHT7kZV~?5&NTky+PV1_ekU zAA0evGip@}F5My(o)RG+dhy-;)E5`l$5<;jt-OyONT7B!)0;6kd?r9^ZFowA+R;p} z`ip}J@EP6D?#G}&2~4hJ0jDs07NT*TkaqQ4;{_ zs>V|yl&k2ePEy9fa>tX^npNIMk0nq#mE3ro(7g#zQI(VkrBlgG_bZD7;L=U~*@;0< zB#=AJGj!I^!!5l!IXopo?ljMEz4RU4M~8IBGU%xU>Pze0>oU8ohZY%4=P42DOY7YW zwH-!6i!wYHGbl^~S<>>^mBm*_0~F=OQzB$Z%V(F{&)f>ouV%X#^h^RN>Vd!eSz%B0 zO>>?SAw@lK_=lKB09kFjz@Tsmw4GYC;Zh`)MHy}9DG}OEEwY;M0B26_CHXVxxdhUn z*TnJz-Jigvvwh7|BBVjDiM(O^W1vMdh9xm5LIQcxroWB<^l@Ua>u{bDAy3-$_p`z3 zjR1{5^@~A~5~zq4GD5=g;-N)tPV?`GUqb@uBKCeGr3QNT5*aKo&plXaUe#J)RPwQ0hQxu9|^w*mzGp2ECL( z{pmC6V(L{sN8+;92|%R#ouC3j07sBOSi*wTp_f` zIf17{sFW^Umm>>sXjwEMfR(gj{K8d9LX=e3jezjAu}y1k$Fd&dAH5AK)`u zf0w63NSmfQeM;NnEU|MV0|q5YAa$B0zHR4;Bfhnbc}j%TX_nZvtN#x8jN(o9G3d1f z(xF%R@%zusphf;0c}j$I=v5xHukU4mb|qY9P_hIvp-%p^_vgm}v^J5aM973X`E^(1 zKLIFY=tBmjNT2~U;`_5!br?XGhw+pM4WJQUlhvJZ^X#DuZy5AO0-4i_*rE>8@MOHX z7kNsA%xOhz_#Hbe6$mOWV$fR&$LSi(~x6iJ;E^^qrL!)K%vFqA>55-5tk-V+|xz%Rp(Dk%|)qOW%+ z*I{`0Oyf3l7?dV~TG0E*++sG4CNis}M5qP5k5p&tIY5g>n{8oGx&$(yzW2aRGdIAu zuJaC_5+MWXdoQlvCJ~_J2`3r!UIN+BrHh=feH1`}i997jHgxG;7Z@%A=;c5!27Qn~ zsq{XI>XujlkozE>5}{OjAH8gQ^A|w-&%9(%h6HL)vs+_Ljd7k~jy+F_P;HuDuxaNdl=-ImfRtDart8x$%?;sZu$I;rNYe@EO&wGm$}` zCD3*n{{OnsD;1!ODk%}#PQ(8z5rg&v)Xr!XgT6>0Pdc2_Z0|MPcbr!xB|@HbIA@2$ zCOG0Ve{RX3uM$X~UK7LHAHk{4#SuIuLi+TY`2J*lZD`TLzAg;Y(H^Q!P-zawhdOKr^@A zU{J9HN})xNEpr~f1W0EaPl-?pErJ|zJ9Z7~7?V?;GUy-HG48Bkr#@HvK}-0nk~8^# z{?9tbb~g0$ngu&Mz+ZLb6HkeMRXZD+{pi~lj~VYg=_C8AN+iwCrj2{ocIx90zMx7< zgtBSl-rWHc2SbbUZdGerdDE3jpb)z09*!`>CvePdo)V!Dy6O6P2gU;wq}-W7WfEvS zO|#?%UwICf?qMUI5~1-l&C-6#LVWK#nv7*oxda+SrE8507vf^mt&Kb-LW8JuE#9>? zF05aOUCf{g3FJy;QDuEPe}WcWj^im2a;37U_PzQXflK#u@Gb@&Soq&Nn-nGM%9d^+ z0F6}RDG^eXtS@U>7w1Le&Rt-Tg#>Cs(=2WC9c=-+f1amAs0mH8ba;AjAGD}PwfhXR zlt8zAs4{a=b36v5MRlGMq1!&zTH^y7Rsi&LO%j6+N+1n-c`bhFbQoF`xR$3xNP}Kp zXS4^72B>q?F9scwKriTdwLEH|5w*QHZTXMcAsYLDsNaT38X@Ovp}U) zQvgb;k`f^m>YKH7-`EGB6_$e-WG#Wl{GlWG=GeahXwX5P5}`4FsBUD|LKlD*z1L%q zjRZ2L&6Te!u4DgM>jO`TkTGqpJg-_87uI)cn=t6G1ac2i(u}xKhNS|*Iy@yp?jhD% zo(p~&L5rNO9%YcN1lmf2Da|!ouo%z!8c&JPRvJvLQl5ZQ$_;DXX3!A{^nlj8C%o0K z0iRJ)m6QlQp!M#{@nuEOqPnXh7<5zu-KQaZ%(TI00D4~~B|`UU2;c98c``st!oD!* zm;~xZPwy^0kNN{N@flBvP&aydCk7?p0YR5kYPYYvkB&ObR4b)kViM35+PUm zjI<6Z;U1HBhq^K7gan#Ja|L~lX5#joL@S;Wp;xMC7B*?>WI5~v;(deup)^*@VEQ+P^*>QSNBuIfW^3ze_e zJ_enVK(FXj>RX0hU!X-c-aI8jujo|jdB?8ygG;xkkt2gnOQ2N6(aKLXahhdTW1bSB zRNBR?P1hdNTCW z`i!&j#;O|j3gVEP;eU64R)>FYgvx#e(Z(KtPx5}~#9_4cUe*#RzHcke9>x+sDA(v!hr z)|yO!#@yj45$a1%h8cUJ@C236rgjW+kU)_%U76i#BJPGh(~PG?D3Yct)e2J?BoQzA5# zz6=eb@8HT~lInK`IZB{VnvGw*>r;Pdk;ed@5}{CjS9PUk5R@~0MU{j$aVH0p#P(QzCSU z$~he7e8vvs^125ZbX5X9r~TbZm-}HaXka~_5~1g`zgw;15pJ>=WaPr2YZ6G8sw`40 z*5SNpt<^jwLb_CC(X`9TRdDG>KMrBgbqVB0Pw#56?Xl0$^9fIhkRLt0FKsc!p=GBY z=?uCdfo$lc)|C%>;j6qxPo5GX8#<}AZ7nxkOC4rY%AlJP=o|Ix=BOtugD*pu!#pKI z->6^LAZ6EkxO9v2+I6bDVO=GVt~X8YeK&dmkXk-ZiIA>09TwfW5KBPqW{+Z!n*{Qt zxuEP0t#LwQ(;S`>Ay1kM@_8` zE}dOtGX~w3K#G$eR(IZwGbihs@RSHCPJWnlKLJk__`JcMLGBW$F^ypjdwsy3s=qN$ ziBMx2!|FO@-hviozPigG4+%7uCI-)(`uGtp-QyUZ5}~m)F<8%ez#D+lhsQC71tNUxOBnF5xK=N~BBIA)-DmHfb*J&!9UJNO4A)ddd0y0O=d> zln5!#DAV0wnE{tB?8#IH`ADGSG^c92;0IPXJ3Zwo5jswDsy*{=;oYn76rjda zBJ_`*40n9gMgjEd=tlB0tlXaG?7=3NYWD1pAyl0l8~DV+fF*uqmH^qrOr?5|xI15n?%3k-TB zfo{{QylnNZXn_7!Nr}*HdX=B+Y+wn{wE_1T6eNMhQkU4}-g=z8wHwG&A~cq|#5-o} zwE<|r$s`5^OQ8OA>FoA)#qO4l9Z!i+f4X$7IwIW|I%mG-!tjgPuyDLi#d{bUuXJuR8AJDG@59FGG5mYcGHtQjRhxOahtGXkv4cEuMp6 z{)VSS$dpDCL-Y>gCW|(sZ!_qb1d5>vjWO<@aAi5SN=k%cXhOqFDGhr;8ZHqG3YS0w z0+lqkozp)6pHcs-JS9Q{0d<=N`x9x zmss~*k7$6B$2c-5S_0Lfr}tXh_i+IEsq>Tw)uE^NgW_X{0or%@A%k8?AVXRavoY~~ z4A418o)RHLS`nL=*bQHN!KH5)6eEG6XqNcjb$xvC9Vz1}5sIQ&;^XTgahBN9@Gpa6 zCD0G*ZdrQRw}Tcrt>7sU`a#_-y)7}g7q4eTt8SGyY@7sAoS8Xey73NxDypPJNO5N7 zB>QoAp!e?H!x$7VfjZG=G)G@+D?s{vcuIsi(P!j$Rvo*a1ftB|>+o%wXTZcMIWLHzj8agAygsC912h;~Rr3%k6V{N`x*^UA^+M!Pq(JHQkOu zNfKz;ds@PtUIQm@Yw7Wn2rYY0Z`jBmFQG-xy}cOpS^|xsV|&+BJg*6#k@p>*5}`44 zY;W|<>sD7MyWXfkC2K&!&*b+5dS-bkPl8kMKyCv1ZjX+7sD5h|fk`K`hGv4H)-N!bz#ih4BM|nzw z6y@$d);?Ik|N6%&2Bk_MKe}P(>$|RmOLzAtPl=Ep-LUGJ{hmV|W02lK2Bj^8^B801 z9?`gOe+gGiYAxXZ`9JFzwQtht96fhlg}*AbN=p2zv~SXSw{l0U>0R*9h5c3OlIFjm zIk)*Cd(Qzh_7P8s&>Na_YjY^j1wK%p&Y=u?FM*cRw7=)FD>x#!(uJo)XgN*$H_Yp< z0nmb2FNr})3>TNEIxPXt5n2VzrlqrEW z(qO8@!3Gy?{T+Bpgf`M(syN&V4+z@-cRqtYNuYT8IkO(i9kCKdub8JqD4u@KtXAvQ z3*gfAnQq3Q&l2cP!#bL+i$~z4tE0zLB6O#rjn-nfnri^+=V{NNFB0e>O_Don$H%~D z)WM6VMCc(+l8;e6ic2#!8~QTns{~S~h4oppFDwTrw@ONc)M;UT{_{vjXwlRSaSZw< zffPSEx8!Y4Y>}EVPl=G?C+FVn+U*O_x~NYkZSIxf1BhBl^YpqX&!uDyxzbp)ZfDwX|nmtXr@o7h3c@`Xht#B~WYnGW1Ovng=a%eZ^BE)SA8w;c?%w_@Q0j>b)xO zqXG%EiSDE9CR_2+{0$Q}ux-)}*Nub%ZG*gZ?)IF|Wq{-ErPApRXV6~>WKY#fS2ay~0p!1ir$oq} zs+AJ@o*M!!`tmA?LB$ei4{b@vtlbl*{a?lKlnCviEeYoR7BvIt&yYd}{gXiZs7u`V z`hII@QL-9OiO@dk5^D@MUICEl>1MqvZ`cwEWJjOTQoEmH0Ge`!r$oq(KBGgKCD>=^ z@Ou!0N+r-u8p1#Rk^K{(k}4??x=BO$4Fe8it*Ob}84N0uK#ENk`+wwMQR|6$JS9Sk zO%`7}RL4h9vj-*&DwjYr=$OPNyKQGfi}I_aL}&&blNdcyw-7!f{|?6(R3U*B2dEaW z-d!CapN>2wLW%=aC&p_W1*l@*Z3Z1!)b~G!mg!VPWc_K@0D%7P=P40Nry?SiCUvnU z;q?0m23bg;r8JsYk>2e*KpQ^rln5=Q(ZtpXhpI!1PLKP_AWI4Kg(mkpcTbN6$X=7D zMCc1m?ls=8js^S!9c%ZgypIk_pr2G6?iRAK4M4q|cuItRQgL|BrwrU(xT$7$1|5<> zH6POP1x~Xf0h&;Yr$ngcL+Ug5Kfyb)n5%?yiy0Cnua_jLS|HdRj)~# zc=(L2KQmyEwFFA1lAEAGU!DOJAI?)Eluji#pJxP52Ix<}{S2~^K*@o0yxa4pc*DNx z&r>3l97w-7-)i1zfRKK!Ar@hwK-7CGbehYYfn zK)FRqnz4y(@IIRJo~J}8w}|?6XXg4ti)M~{%b+6?=rz4z6W)wfhZc3$fv2 zCXGq}df@n%K}RJ}5l!y>I{gPvT|42#QzBGElY3!7udycKS#hhrl{f4$31m+l$TfbQ zpFoS;{_&Iu*;5Cy`VI{owj5eAj6ugGP%^#vHcTFfBffP@c}j$m>BXluXc7)BLm$jx z&(SPD`K*v|?fqy}S-U)6{uNgf7sE$=WVC*k?$+@|{6v zB+w8l=ZJIveHWmQE<7bdL#Ui1e|+3@Xi-4@`u!^JBYO#C?yjWi6Ifgupo0x~N`%ba zY58o0{y~6TR`+JmSqU_pHd%B|Kh*`G+%-HULc?j3g-5d#9F<>vIgvrg2gF=%NG)q9;Rsr!Fg?MT_%!N`!*w z$uKprogY94=Y}%KK?0f4Zuf$gt9}DCZyryHkQwcEk8*Ox-{v^rm(HL|66i5K8Md`7 z$6`EPf1VPd$Mj@~?q#?hT4dIuj6s(r&`5gmZQ0rGHMD3}OP&&;k@Vu5u&X0}Mib52 z_piKR9VJi~s#faRI00W1OLp*-2z8-qrB+j#;+*Qagi#D~l0YH!J}PT>5QqQk5_w95 zLg;;z;r#SBT)M0w3mD`qffNT3N6!kuwTU-sJS9SkgNT#ozIg`F%+qEJx*~xT@1r)c zWg74q^*F;*BBXd9)#wr15TI*$_6%~7KqmBMcv;Z99zdJ(c}j##=*uuOO?x#!w$ptX zbX5W=idt_i*1)o;^?E!dLW-i+9xjPk+cV@&9D}Y&pmFrAQ(X~PDtd0b1>}jzMk`=sRut>)g%uEVO8+Gf#=o zciQx)n_rI4y6iF=2HlcCieFUJ9)GO`K(XaKB|?f{R5YJFY9h2KXW2~#-IhQG^qQF3 zOE(yxh~+#bLI(7jX#057MS$jogfYln0@={Y^3A{=I1n2a%2Og_Lo3T}scnM*>fI%i zK^_w5BTdn?a~rG)Eo#}7r$p!@P0`F>^P@RHx`(O{sJxFnB~TE}ZY@!Lj2+1FRy-v_ zK{UJNl;VE^pld%mGssH<9rC4u$>eQ2p+z@-@{|Z2@}==nqFn+&59eqw$Xf!{3!uZ$ zo*8@u=*(Q65}|ql)>^X?IT0GV&U$e_CtNRMvVM{kX;L5t>@@{|ba z(G6Sv!FdNjdy?-n$X5b|(5hwI0mT~tT9v|6A{0WamUhQ4E&^z?`fCQ=lR$1XOMK6; z5AM>PI+mwI$c<)+6YJam0jT_PA%pxRP%J&Y-_0D1Wl?V(c}j$0>FHg!b+_IC8UAZF zu=0lWmq0J6@VMk}A$B0wmGG1Zy`;k9FDbPQ0b0LcFoW((AV(T3FI8HM2S8~o%@?K~wyt!a~m`s?oo0F@@( zX3#?kR6y^erjN4l(!EdQDG@55_fdJhB{-8maYQ789!a2tC@Q5s>t+gcj*_6vAQgmmbkx;|fTCDbu$ z1!*xTbP?1sX83$jzjp7_JZOHm=lnna=VH_2lQhj@u`CBZ&^{47CH_^9PttLeOUyF> zsxi=z{Z)@8&2L7p;{%bG+yLr2h^IuT8NH4N*sh-e(1^498T3Q~DO24@q|xlw0JT5I zQzE2Hbt4)&fjBSvs=$#!PbJW9s#5;8rRy4iuKwaF5!y{v$`kYxanbhEv_}jIlR!;r zeIe$rBe{+2<{B#`2FY64FT$8qRv51tYs#qZQSs_AS6 zP|Zfg3<{S(`{?c-@o@D}Xi;jFlnCvkyLU!2krF6` zMr|ry`&t6z-HWG0D1=6BYQdJj;4`XzXe)!FB#;iRJZ9NH#I@80Ry-v_I<)ebdus=l zMJ@PX$DkJyNU??Lno&>ujJjv=ln5!dP+e>P8fQPs)x8_`Pl-@Zdej$}8sZkJhZWx$^hyG4q1SN} z%{I6ObgJ5q|4)g~7J414zW#=Tsl=7_hg9B2F%swlL5+ib)!@=auHq>X`asaShviq{ z(j`3W&7fEbw1Gwh_ioO_iZb_bo)V!AG$L5>%Iqyb^}0=BP@DuZrBZ=seOxXRY^kl|C1|>)!MWI(& zQQ-d!ratnN2q_A^+GdUY1}*9`;VOd?C6F&ILmT%p#R2c%Dk%~2rDf<<`pdEaO1cut zpd<+tP7|^%30Sw(_Fn+X`IpY1*Al4vFPfpc=ll+!hb25ELfwDS z-r;jKaocssf-(jrOCWpdZQ3$EoykTy;|iBn<9h$(#E~6 z7Pw+E*q5jN1NuuhY>VMx0Nrannn7y=l#fZfVL;_lnB+P zFN6K3HMrT|ddOJ@rAnYuDvK&?8jTBrE7W*Ogi5I_YEA#bB52XwW4;VZlR%#f=%m&o z@AIKWcE@>2ggzHoYYiUXv<#rx*>Ma?mq5p&#YP2$1R2 zJO;g&K$~ed$GrMF9?&9#X*?xDn`t*kzK$_gSsZk4G_>+Q`XGT6KmPS?mfLuMrhD*| z2q}L2tI6B3W1&UMl~fs&A%WWEDQTKa(((pqU>%+kp>}!nTSA>X;)=%yIt!rjPkBm&!l?xGZ^)=o z04?lzm_eT;&hP2ZmC+5WGo!6CK=x}i7?dr66z7`O);@_78pdmR zN`w^Wnw~s6Yb`)i!C#YdqM*>&Qh0{ggl(ylGXdv;tq{r;qZK2yO7T)*5w0YYbev z@p<5O8ZW~ z`*7+3?rAMaJw$MUG$P*VWxO9b8QX;g47BWUh&Tj?K zXP2)G`YVAlsoedO&SqR+7;}xML@1NW-M5Bq!becY`br}z@1tS~R6#q}HXR>$3|dsP z0Z)lg1?^nhwtsSKXp#B4?hN`TfyUBk;>`IA7XdO_&r>2amPQjpHD+Ni$ShimK_wEX zGgZPA#V^4z?7~+(B|@F462>!P0B-GBGQ^NUr4r~qRnLC3d5ULQj8Nk#5xP&+vkULX zeTPezdTKv|$|TStT3CM^*m7N2q zYc@}bkWB^ky$cO-S!a~)BL-DSpuco##O$JGK?qIfDG~Zhr$!XcPx}Ea8t?v=K?fH9 z_hzNzv}EwvVHs}BZ0NyLB6OUV4E%f&5UN%Fmq8X1=wI_XnwQ==U4Ry)R7r`@zvebt zr!9u!)7xNq>yec=tfd6US{P4>kTcaJ{7Fst0MOhXa~X6<0==izmJ5*!5&>G;lcz-JJ*~E=3^VBlQ0bwq z46>3yRy5UlA?xoifT~;bln7bTROirk=eq-Bn`Or!YYC)FPj9~$j~WA1n9Wlnq)Sil z58D^wA-A@=-VCylKm%#s+k0yTmKQFX&Ql^ZkmkMfH$C18Et=;Z&7i{)s0US942=DX zb3xi3JS9RssLG;NTkpvL#n<}5AX^FKMvIAg-F9?@OBY|8r$oq&7894)sW}4Vy{i7G z%KPYu1nNt@pxC;f@IdcfMm!}#eHDu!xqYyXac@X(1|5|^^M5L7HV@6|4=vISn$pwT z*yc$eXi<^{Pl-@ddV0J4*@H*@xBPIBK_?{8KYH;QZZgD5n35_f5&B0jzGWdT`a+BT zj=RdBlM=|5Iwv=qm7jnX#c1-B2-#BS#AsBeF+dNULK$Qyfo9Sp$iU-zG(g9kc}j$4 z(j&;|(}zO<4Jt`z&?yO|IIF}gy+01b>X!192r14gnYQF|e}D?~%NTT80*(7YGbft; zvCQDle4Y}aaX)Ap?bYN40B!VZKf3aUJtKkU(-cjwdabeCeZD_WiO_tSqN)3A#dCmO zwHVDHdkIuZeFl%vEbKElx8x}iDy2TdgBU~Hgg$)Z0tTIxKo<*@H0Qcai-a$O(k7k~ zp^JsqTCruN=b=S2qRbd{P692;rbE|fe#L_Vy1n2j5n7Z@`y~(O&jzS@|FaA_FM$-Z zTU$5n#nqPTsyrn^irFof;Q@^S8g|N;K^G*DC%s`u8;!tK%f_d9N`yS=4eMEb3$DZJ z=f^SVq6A8y*Tj$Vr_S&hO(@_g5lW!fM2q9rK>~%{p`SH*ABCI#@~Wgn zDC~~4mhq33rU2ayXf&qsKDs1lupL_D{*tFes0BTO)+N0i3YX5P*I@=ZOQ7ocw1;TF>Oz3F z_vR@Ps-91)mUlbv1<1noCWEd>paAMX7T!%B2hiptJS9Q_)PZc&V`B$^I(-Rakc$Mm zN{^t`4q9UXYW0<;MCd9#f?njF!`dzP37HJKDuH^?681TVz#afuYV(u`^`Ir}K=5& z#MrMm?O)QY3;V0COPc?Z9*lvWZpWedvCVl(gkI8v(P(nDUC^QnJ2V({Ljs+rGk!bQ z_QT zgIpz01WmxcdvXiQ3lFRFln6!81gwUyBaU$nIbURun*>r+E0sj{oD416a)qZvNKviy z`R-HPxYu4efI+t;(5OpFnuEQ@)CH(vBc2kWQJ1W>0++112VVw_ov#^mTLLA!((beD zUC#mPZ_ZO9l+h{GxF@6cqSckz3@{~Yo)S}3WNSwQFafzoy zNR3+LG;PFbfP%_rFvv>+*@x5b0z?jogBH0~@RSJIhg)lP9&!Yaggvo*BZIsp&~~c7 zS}{n)7NCU&JS9Tgss75Y$;K(rq9K9D7<5MhrO|u;_1)=*0jl|sr$i`?-uo_UKX63g z+{T?jJ`$)gy;B_C&BA%nO>KEfgc{R3Wo_JiENXpa9?77)5~vGZy74vY;X0kqE}jyh zE_CVC=1;wbmoDQggM1~BasZXCnLJ2>OV{EfPl=Fn0Cg~$eO?Jr&?F^|%KPY^1WI^F zm6`V24*}#hnWsc3;UO)}Sgf84P|w@l8RRE{G^r-RuiyMsfby%PL`air64qPJHwWlu z-3bixmp~cR39NO`wFn@SdORgU8Po~Xsb1O;pm%Ex8FXI)9iT~xX_~cs018{jQzCSL zCM60Rs^N%WLd1Rs1xTP?G~nGo?nxj(9V2;4gnH3{w^PTLSm7Mo--$sFB+xkOGelc| z#SKUPsyrn^wKLicifGu)A%~K*YlumxA;r#hBK+W^s zGU%ZM`b6`hsqbR&$xvD)B|@KQUUZgHUmbwd^okkuNCGLkwOy6dvBJ6L44x7pMYnd% zqa0kcjrV9huJVQrl0YUjJba!}zXHAtS3P-3giL67n4taWC$wmJ{oxD>mOxFZ(CdKz zWGtob-+-q?s3{eCExKD40Z_rJxeN-CKriSuQSWZ`+t8w~Mm!}#FX%Oq*wP8-+{z-h zGAL96snXZm{r5LNfMO$gN`zGD>%IMT1db;1dY@v@V+o|F%ycN;)KlYTHLOae`z*M#pTKU{2boXk@qG=*LhhuhZF189&(ea*`I z=$QoCKwpNs@hx!LzlJAIiO>f6GWe)@8v&$Mw-1BDB~S%@M*9u7<0QFCJ)RPw3i^zO zPHj>PpqgtYG3dDj@}#-zjU^g5FZ!lRN`yRV?z-Eli{SvJJT+oagalGN8G4&_D}XP9 zTNqD?kmAWudU8-dfa-TW#GptC)Q1XGlpA~EJWHJ$g+z8ARa9WAybh21QGt0rZB|@*Iipqx&~_ zN`wZ`8#b`x>eB}-MN&-Ek;s5mXPleE;rFD5qgdWoHU)OiNEkH?zqZt$< zfi~0R-lI`Fa4O7e1y704W}4i4JpVU7RO<&XU{I_CT24=fDO<}PK#S6rJsG^8 z{U2TT9oOUc{sCNQke!(opE9yVDC4anl#ftkWbcs?LQ#@RN|O{Sg``k~Qc)Ttlu#*T zl$lCXqJG!A^LYGD*Z2Aye|tTio%?-{Gp=*rV{1{1Hubh-&@%z_6(-5o8>QhzTg^T^ zMWL@SNj`Fb&V2;!-*`G%qR?nm6tq4p(0qcz4W&GPt>B1qq7+FLI6qkcdK4m_6N1-tp-m~NV>mUrE@WU zmE~~nG6uaAKq+twx>n+EjG(_wQWQ#oTTs%AS$Hqr=*o6Fjn`4O01AiBNyl3=@s@=8 zCMgPqL+2#W*g6g!I&+g@49XEe4$x=#sr?(@y|*slDGE72pJ8suUc5hL?;QgM3CVOYy{4-qs&i>D}b2#Oy(mgV7r*mb{;40D;e}j0Lelp-+jmIFa+%_;wcKrLMK1#Xaw#BO^_(*HD0iv1yDLvp<9+tGDOhy zsXRrYbf`k_oAP-MYLT-4CDqnVK``n*cI}-5hnRZDt^7|015EkSXluIJo4k z18PyT8!im`E`aQy@143+Aq7E&O;QxHgTD9Z0b>FXbgEMrgNg;vLKq)?kI}`&4`-El zib4xve6&&at0RJbTcY-+pFLi2#}l z4?eYCEYk!Hi73IQ}2exoqO zKBLgdF#QovQD`vyMsJtC!s|!IgZ&v)DS(Qhlb`c$#3lqCRp%)R6+soV6%3-*@)lFoY{ zO^?G{dya(i6osVo-jfbFy+F{Gwwes86+o*~VS}?n^CRdtTG5WDD6~4&K}Rn|9$&EM zb{RA1w*cA?yP|w{b;N5ETX*vmh4#a)sAk@2ctS(-#ZCs*380a1))jXMRYomp_mZb5 zG!o9bth2xNqC=NA%7a0F1kmw9Imzuok+@47KANW}bR6DZ*wE#2F@mPJUSrT-0rUxa zK`|vc+fa+z9Oo$teS%()W1#X41TFvaghBNJ=n?E`m3Lg}h9JAIJVl{Lu%}hCw+rs% z`_C$7P=f%Hfti!n+P#wzbkKmOC?o?jCrZl`^ihi%F0|HfypC*_{`Y33Z!qc9@@(O3 z1Z7_2DGGgqNvAEgJ7N%Yy5(R79TGs&y86GK!FXbDi!4u3NLp8aHT@f2ZK+wM&mcPi z)E#=N6Tf}NEqbz=rzq4Nda5y|EASMJeBxRL9Tq??Fov}|Jg6%=bf24~DC7cT*u=zK{CE?CVz(Ag&kSf&I?KSFxx0=4T4`R>}0W=s6-Qw`{$*4si zj_?$P2E(B{+kciRs$F=J$rOsC#7@XAm@FF;7vbH|d%d4j(#|Q#?f>>1O|8b!U7B&aJ({pyL8)aT>fu|J(6%=+J%m%~KRw zoCa@;@~yyAVP*@n8002^vf;Vk>x$}p1nHUb6os% zojD%Q=OmrrDGDjU%46(nCmRIS|G3DYlL9Ccwu{ajdw3XXkxnU3Q799(i#EvbPDW7P ztP}?M2q1rW-RLUG|9n^~(SWBY#gB>p8&nwNCxBMM z80TiG!d~kUS;bsgvBY;Lj|9OUc*gOOUjNmB>jfVd7v4Q>nK`q+sYQ>4DuI1k#HT=A6SHYn3ME5<%7m9o@Ro#rPew5)SO6KrEy&Ees5OFGJ>@A18N)3o>EH!CJY1u0#Gvy6 zC=@y;MbodoLePXEJVl{U=$srHVS_JNbtiKMT@XMu@Z!f7fA#S?UFV}bMWGsa@uStD z{X@`i^y9k=gDwi71~}{7UDp_)7Tql7DGD{fSvTX?EZkF#G7Mu-hyXH#H-~pC_-lZm zwAnmGAwzg`_#CZacTkHKMWiw4k^oADDP{TX(=Q7&~<&vQxvj-SMWWzJ%?|qV}|K6C|m$pz!!xwpbI6qkU`DCH>%9fQfezN6ztqZZBl*PlVx1W*`siS^yf`k~**bPi8Z zC=9y9KC)-r{206Z81X)$^6oszB-1R1pskn$Jbe;!;ZVMn==mov^v@u0c(tMtxkS+9r;xbQZp%$4$ zU1LzR05X8n+x?cGKZ3U2*%u^J40~f5(v_E(!b+mdSgAxT$1dPg;?fi`kk28kw6on#SR4!YnY=>H8;a0$) z`vR!@eR%W3pv`?zi>ABt6otCqchITHZjp|lBc*>Clq7(jNjEF`#;-@v;xeA1&@)(A zFBjyl^955>zm;{&u9lp=saV5;-F zdNqDw-KapGqEHA-b+&C4n}%9s&|(pTQU%a@C>5}IVRQ+#NTnrDQD{As3bcKxrGp@q zWtI$jAb^6)AbUGrX31bbA(q9tog#(D#lH zJe!SLG-n4-QRp4?z0Y15h&RFwL*t3 zcyN28#_Q;@0Fr}Ee|jrD0ukh^&QlbUgH3;i6AJJ=!xfid40?q|a3hBXvZ}YNwOA$2IXcL1n1kiWr z1*J^P{exQ6eJ)Q?=sWa+HXU=GkD%r^oEVfTfX2XRqUGO>gAtU~Bt@YyFq)VZ8g?E* z_DUfPdM1EMa^U^IWxjX?e{W}=qEJZ=yftFB{96S1?n_}%mH^sv7haS2y8S!UqJ#T+ zib7lN!kntU8c z^LUCvd!UoA8*7`5pyEg?2IUGMCFq>I9(wfzf?izbDGDh;=VXKCwIv8z(bbPZuLRJ* zWI4&7)eEj7Xk0g*qR_x(2c1@msy7i7Wf{$&JOT6$KAXO{*I(RIE!oRc6nY1rO@9#V zh2O{c@l`g1@&%Ce6(XlPo-{!%x}C>U6q3F|IC&yI_pm03TJs^o}y4EsK5H8 zScKQg-$a@-=$!yM0ApCc;|+KwKj%76QRo1SVSQx7@SJLDM^^^D7eLaDFmdw^Z9~6N zZYQ3ikaQzVw^(!hMgq;9VGQ~pg5be-q2mU91Zi9F)PF$m;8Q>S5g)p~&(aw5Q2@EY z?A9dZIy~uA+ayIHH<;bJ+EW8>NfqZSPv%2O25gTD8bts3{x zS(oc7H^1>ZDiT2Bp|~agz(QQWA9j9S~*<1wt~zq$f7LfZ^WQ_=NU?dFJan87Z{;Zpy@$Gy zj=dM)i!LxDjzQlAkn~%t7QqAY8z!%3@)U)n-&*ZE#~M#c7>vnhP_Y1#eitAxvFtc% z(c-Z@MIq^T0lH{jH9&{1n{O?Heh8q=aI2R*>V$8Mo&9)MjvL?l3P}HmJuy)FS6|WTxCHd-Wn%`F3808_SmInXHy1(2!gz{85#J3=zVJo2K^L35_qQ6=`KEr zph!iYqL2ihDbbO`E})ZPz!nb%l?xz6=ma`GF~&=rqqg!Cg%qI^*nOtb4g{$^y~dyl z0dx^Y)-6BeB%l_J%-|^sU4)VK^ymt_*mQElQwCKEpfRvIT)N8P7lK?y@)U)}!0K>! zi5*^fv~ekBP?Z3ZZc?tC5Q3qdt~^B{=_chlZN}npX!N(%CXLrowE&WSJyynD8P8BX z{LWJpl72l_wfC$c=r@{eG?+m(0;mLv!>ecSz;EYRGnc0*R074}eUG1Ljv(DgeFps! zKx^Ri-Yoas5&cH-iG%*P0trOi_mGKYl>_oL%)pcd8p6)@

    H+$OC#JoYvqLMLcJyp5384VY5dlPE_w6jdef(&kVYY z6VG-p%j^bUmf`8aTxBu972)c zCg))tsgGsvr~QI;VyDlF5*>jfHGvO=x>1T)1lioX*-IIulsdDY(FvclQIC!ab6D%T zhLU*6)^pLlAm!nSf@>t_Hg0lZi-&F4;$F2idviW)5iU=CexitW^e8#Gpv1`K=-0~^ zFCQE&_Yd~28pNV8fr+h{Q2YP!y~v z;heM4W9j7fkot^#+v3x zK^B@n3f(?W!GT6>Dl925n(TPjtN&q2K~^XF*GVgiD;BXvymxWMVE685)>Ld>fS}|r zCXs5y$n{0cbK7?`sqFs!FUY40iAD?{89aoKcl>g1XLT07-|)9?-(nwUOeXqS6@Tk? zZ1z{wrK|UBy1kF>!I5{N>GNeQ$Hoe#wa|eUr z$qS9ZouH@n;ENoAc_-^%Pt@Dli*%}dG-k|jw2j^p|9YAUXwj?cx5a3&&D*iX$z$Bs zA^WE%T0!9Fq>*3^jT6RLso=sNlBl=*?2_OcEzLOr635TExL$xY^IwxViOeJcE)D^O z%X|*bqFgC-$hv$A8x$$$V26BDe)Kz%@^%Lt`0|B$0f@W>tzuaT37@Scp|7Z|IlHK5 zJ&dja(@Lm@Ivpa=RFY^q0klweYZi*Tj9=SOw54u|>@4U#=+>b%z~-*C8-wT}ASW zB{yG<@;Bl#W0rdn+v=FVoT{hWDtUJftw69i0?-&`2`ZSI7!B@3?As^COlEl69Mhc!%`e4Cdzp>Z>-pZEu1vkrFfK6 zB&VnZT78MQHS!#e5*XziZq1P_?dlJR8ZUxSQ-15;A;rB?q?kj@#Nn~a26~~#0}3C& zj2>8o7OPFHO522q`t%g{Y#1728&Jw4yb&EYVWKNe#ar+g-}0PlAc+(TWBvlpT8HkaW8L9&3JyZ?G3NkST z5{i~t1g|;P{YRl_tA~vwCUiu_0x~b`8=SCiLV+9le}!0ur1+v5+R}Yh}eSEJNHsBMU=C8+VJ#a-5Cl~lpH(Q`#}hg zn+GdV98YFq0g{*ii786b7d!PRY8Cl4jwwq1dT?DZ!5^p6f^Ma_Jy@Pqi;&bzi~49{&IXn&b#p9jJ`j7V1D9@#=_6&r(Ujc;;R!d{wjPTX+Kw- z(f5bu2@D;1qFN8BQy7=T_tzJYuEZIAe+Zt$$Th6Z=*uVU-1*>MkiAzm&glD#)eNr4 z==($P47M)4ETb-A!>uUrb6#MqfOr=`$m%kt!HdXY}Q~ zs!!zlOVS-6Wl?wktKgcCIPh)sCb_^1T=+Z#~KbW-I)o&vJ+(pQ$ zs63+|5F0rLVPhL0b9LPJ!i>KAAh=v~0RS+zkqhp0x&`O(EzikNp3(ORi}rdc zC)?7@QYX&nchofs1)!AxiBsH7`Hhj$@79up14^uTQAR&fCZ`eAA%PD{Ls`%689%0L zAvGkj-!MEny=!~kKRP}6`rGdYNB9Q(Xx#5}@Z%d{&SgKTUqqtMdJ`q+SEGS(%(CXP zfJ*mLp_iBNn}Be~@l$I3f|=U1xwvCB zuDUI$Jz+(G5JCb30>k?i2e6s&Fa(&)1cpZflSS-d7OXV{m`n(HEFKBNn;`^y_WwBV zv-jEm|DXTwVurOOx0Up-Zl=$kNSYaIr zDN}>i+jR1O&#mnm0-0`J(PG1TvsVgSsORc%XdS%J#Pg6cTHm-inVr8-{KtDo;fqYY z-&Gn4zg&a={MJ&oC{4D%XzP_h|IU1c%KqaD{6|mE{7!Efy-mFK^dMhG705HB3*<^W zPk=M0^tqt z?pH2N_O>70xp=s^y`!QIa<_!3I4Z991l%ah%Xann1EK#syxCe$pujlf1rz;=Y&sEr z^@i;IDlzF>AfsnOGFHGRo+-T;RZL3}_|1lv!Ey2gd-{KN=q%`&|oQsJroNg?JzIoazxbcrFvn$X9 z5ZQw_uy{U)XoI|gnzt8~U(T&A;WlR)w8zI@_G{`;&pbR7zunO8DzxxuUUF8h_rML+ z%{~A(9%D9E0s703m$tc9g=M>l2kCsH*{nhN6?{PH;m?p;u$X!%!b&r9L9}W10kF56 z4_^i1@cdlm-&S!rX+0|avR+oI=QiQ1qoY>S9o4Vtjw;u5i`i?v*Q@FN zKJq5FXYZnn3-{+cRb8Hk$#pD;KQkh#N-*4(hE;ZtRiA$yPaC)DX>4Y=*2viOG#=Mq z8nCjX6_e*Fx;!JxPa-p;>2tC#KbluI|(lU6NkhW*PJLsQOfpSiK#a#T>&FYYmg{{^tIBW|F41KD4 zQ6l4tKkJKUQoDRnTsxQgvpG4xowfFAX#S5FG^qdzEJAL8qMyhye5!34l0Mw z{G+^2eEj403r_V&PP-i1;<^wOp#`L*5vnYSzsA5G^DW~dMd#GfFqWwPyFY{r`!47e^$KVjNDiU@Hfew(lZ zyqq?Yg6Hycj3Z@#?Be;Js;fo8bzw(gg=`Mn<%{|Izw*0D`7Us68awy2$JdM(cCKZA z05~tvQ$gQ?8kX3S^g9gFne$zf{@}b6AAG!;jq7-OfFp@bk%}V_9xtt`M77|uVPL%| zU(;O}>9j1Hil42pStxzVV7^3|N{-4+<9pnfs0C2Y%P&?)`^!lftd|-_t?X+y^Cddd zj2jmv?w7h%JppVVGgT1YyPR4qKd#?Vv>t$j&MxUeac9e}#C)8YkQ(DnXA}Qg02$7x zcjuo_C!P<8`v#}0oOs&9^GiVDzcBhAeU-u#t*TTy$y=^(Wmn7GSI?wQ-8$Mg8PwV^ zl|N}WPN(}Mg|Ez`xaR`4*ShuUdKzv+F;wj@yNac{QNz!kM;8eoct5EYi3i2~gL~^O z68Z2IRtea%aznbdX+DUxOtcK{gt`&G*%#cb&Y7`W#;&Qj8Y@NEsS;`>acNutW00GO zxk`2_+clM(Qa98hp%x|a%{}p+%JIj1>mF9g?aSx0_e{5T_O^Gm)(YW#XE9&^X zDO>sAc!dOTm|7X}*REj^dT~0vv;`g0hj~MP);_(n4W0gG`}F_P6Ys8Vr?bWj^n)AB z{~A_bR7?fOXD72;@s~aEne`{|!OIiyvqyKHe|~x}-8(-W+;Z0at9XO1&hv+9bQw?C zHPF7D@kFkdqDfu9G&$Irz_eQg+Yj@LqI&IfXgbS}hlMfMH{OD8$6o|4*8!|JlP29?S;U z&(oOCQ`N-b7fV!c--aH&n8=XuQbu2hArqDjLU{ zh2utE3~H`oRA2}+(Xcu*Y_KG&x*_*-TdeFuU9ny>GdoJLnjtWm#p~9GvlJ2*_UoAvGZq-%cjBD>a=wf?o zFX!WP9z5V*z>3OXP7v?2_&?F~6a6qfIs35#iJJlVMVY*UM;NoS;u17#-Zld%jXS~$aHN;gz3Rw+5y!*J4Go8!(C z=(He!H7x<6YAu^aqq|(5i z_&8c=MLuoiN%@`oi;wK@9YSd#|DQkw2kxIcmUc>b;`giQwO;J$430mxv%8Cx#@U~7 zjgA#z`5!FwOB{5Z)Y!OL2sodZzOpsD0`~-Fwe`k6CGf{CkvUi)!|jxDq6ImvK%i(m z9Y0L`R3GC?BJ*>D1af@?^4r*@XULEHxTE?yPLd&`x$=SmL*S2HVv}-$K{X9F_i0*% zGPxCB(yx#%Ic+gb${%?Z1-p78pi=Xt3n@I64d$%*+*R~d7nA%#DL|b5tLL)6$}VU? zuFVKY;x)+d!+hCJgwX2(W9h~iHRqadT$~h#*qtJQai7_LSs!San@IHb$xwU7Ll23c zMlZt+k2;uxp|QJn0>4rk1MAQV{6;=C9ZiTtuyCPDdV?MD;wyN%aO*qP?lsPBv6Mz;@|Z#;$CG%!ZK2s7yX`Ps%e3`m7TY0kkNghk1k|GyC2zYf2HcYwFHB(asdt{FxzLTv{t1Q;^e-DLUWe!WjXaSq8_LS}Zlh4{J|CUQ)&{MJn8!O~mYS+X#Wq>CX9I)ypD z2D@|me}UeeZ%rsK51FTH>_??ftNiMexM%P#Zhn;wXooeTxILfUy#5XbcC)Wq$h9y< zjga}`;f0J*zO^^G2&VyqV+&2D=7HWV?azyk^jW@^o5sN^R~JL%`h14%b!y*44PhL{ z4BsaxED)0jdwflWW(Oe;D$0gO!}D@8__|oIx9}X2E-2?8OXMc73OM>FmFY zX|Np6y7Z~S#ZKgv;eTDM(JQT*taYzZqW8?|j-KXvC%EhYxp_wVR(SH{rA`qvn6?b> z+^ir*B!+?eowC4L%bbdgUF)~AYn$r*qIq!-LK+otMPA#} zub}sbkltAA_v&h>-pl~~RCZl@rkp$e@~bm3P#9eX<~}F(5%d*Mf5hT)B_3*X0b`*{ ztr>8U65`X!E$~A^M^Gpm+RufB&P}=M`H^UXRnDOtRbQHfq$ilL*vq8O&u2i1Y6@6c z$Zes{*UX!7oNJMg%dnge*q0hsF}L*iI5@O3afVgFy#cLACtImS-9h?X?dypiP{oL4F&dnT7bEOJD-8KlZW9*utiAiW61zf?YNc9>^>ua?tnC%< zbwX7%r`PC)LX(vb@d-X`!*{z#4ij1PL9ZNx1m;17=@l^^ELFj@e}|BOQ)oIAY76<7 z16dm}azcS>V&i$ZI9>485%E9wu{*0hKo!<|gRK5HR#x@Gqkuao>BPiw z4HvzYl&9g_c&d~miocJxE_faxZ|=x?ty;WQ1lp?TtXr$ul&HK%B$@_cEd&*(n=#*0 z+z0YMp!Y|i0-2ZbI9K!3QC#%W4CRX12`R$F#)Wn7+zDO76E_-i?*u-qaea}JpRrr{ zx}JEFnYKyi>)@iDuj@w&T0SJCR3$7Ln{Aj*EpdlqVGDmi-SZ0M{v5v|;@y*jRT;$p zfnIw{kIpH+;nob_7`qg?F?R8ahQOP>$3Sls*Pp&F;|1M5=%46`??mrOBqnCvi-X*S zOZ+-)7UNDK_mWx3y7~~RD1uzGM@U!Pd{aI2Yo*6axv+t+E|^h(h_9fvR90mQzg%|; zC8M@Z?84f52y24Us>}*bE30b`%9BNrxbkMN2Ea2f7*&}PEM3)oEwASki&ynu&KzRMIDGY26LLFQ?{n?X9VXM3%Ip_i685U z*CT_{s#=e$ik?kNx>~Yzl``_-x>}-jwNYDFtiBz+z}ZS_#x$uPWZUJa^R%nj^w!2*XT?h z`Hv-&_?Lb0Y3PKrlZagFXEUV_k;QyGj2UVe*J=DmMw#GGW_LGPPa;4y_vXV32RnN& zZcSd5knuM1Lrqd5i2iKN=7~*<{yTNWaEPlLr|KS=j zH|+KF%8XTb%UZy^_@CsFQ;C+eebvxYotIGVQpfUlh&K=3sE_3Xqgrl8K&~&oU%t!~ z_zJ&k+E_Eu250OJwkq=yKz=Kp>Z&ZReOIo3QjxbiSCPfkF*<1hCr!mW7nKEe`fB}- z=i&&j@^_)>>e4h5UbPq7Ga+j0`Sq$uw{sQMiEreNoizs_LgGDA(8Ru(%4o>ah703ax45Cwt)MCpxU>5DS%b5+?DmChfEkDrXSi+ADXzA*XVh2hrK9 z&sCdGRC4iQNdmW|%X`^8@}o`xuX$;*dl?=;+I$Vmz8Zhb5bu>gH}UU-M$gsiFyM;Q9&ci0--7q|R!)-l zOp~-WzQ{G0tpYC0dV{|(sYb}bi6;T-vz^(MAu`JXG z6mGVy#e!H!TjL( z(@(tb5tJOEa{BY-F{GsrLypAaQkVtTgbNycc_Kv_2PAdTl0m;9?M(|r~{9__$FXk+rLZ8nGqVLn0}_6uanip z!}IHJfp^TSh{oCmq!JFEtGTy|aM6^%y0<{xcz)~r?&R=rK=}c(7r;Gut_XV2C6xVY zzUMCsPIK8)Xk}fL=hiBxZ=$SW4L8tGWoRj;i|i`eyDC__NY$$-Qa9IjCNivIJu>THcvCpj zj!x>Yi(8#JIT^Gwr?kfZyO7nD!{D%8m8&`NjXiO_A&ljQmZwk^+kn5u1oWKsHPTt${9=J8qYG zli@maQnW4JYz<=fEH5Uyo+CC&nrk?|1A2<7$&lA=1axhq7+=)_7ia`;G4zast~U^YMoMN`ti`sHutfi`%RL%F7tiNdHCRwV{t^hv<{t zM|WNtB4ZD|Nh^V|8Hq22&9*fdaYF;;JRvRe9bw4Y-*KuNAq2v-% zG0Z=F_VPu%EBFmk+k?%Ub8QPbhfno*YJa(uABM?#Yc3AebcV$EN=#yJcDvildn{WP z&+TmIv& zsCUI2OK}!9?*wtO{cvydP5Bwg^x|{V?d_et7Y-M7KW29Gj5R*t0zL6J#5)kt@@mo6 z1ZYD!MGX_rSmmyzMzXmoavsOB%QOC(Vt5&Q>-GI)B10=m+qGJS}0kw7IBY_%;`Yv;7yR zd->bKsexGbRK$VUG{J;h4zVSF%mJb_I0{5t-6#!SNG+=X=UKD!gzAx0T3nxU694?I9ck zq5xbb0mw;2osp2x8lNczzLg^aKDY`Qi-Rj1@I5Kx;6s}as5lt9PUqC(WCrPNmXros zR4Wx-h)rR6fkPaTP4iP?Q@FxGq_jVZNL%|TBK--wh;-pbZwmjb#^Q>?Vl8{309_#g z`nZaa_ltW68%BiOI}ZiIgT>v+mFdA&+?~P!vu?d+^@!jVuIb0{y2`=pXG*+2y!oKQ zYv@)UQo-q}IORnmR~4|U9KcEkKmlv39|f#GVHdEJi$bm6DVY8@4cJu$*m?#<5xc5S z5uQ-+d_decI9&^#b;~<}b%P4AYcyo7jS+G-x<*syHLA8cSQLwJZ z+pYs>(g=UILioL6W3W~k;ibm~F;2zRjU$5D@53>>!NKgq60_@@s|vHB8~n|QC53YqbpccQjBp6i_SK5ryam8O=#_ zGT(H-eJhO1e+bV&VTaMDXkd26DS#VX-j2)2IOBXoBJ=d-x7Vq(b&_pxO2M?EQHQvWNH9LELgkuR z8H_ohQZ4jUn2bkoK>_xtkbo$_E^$I>eZaysR|i#tDFxW_7Q=WnwYLA&&u0Xji=xiB z!~(3eJ`Px0_c&nvNxOip7EyoSz^bMOv9PM`iDPxC6|Cx!kpmk_2P8}-mjq#$b{1FR zsxfqJx9Kd1q_rwZ`9fGIH)Uk?i8YAOZAMm$N`yvMX@4B3w)SzL`jd8nS|hUlxq((q zCyEu7JhS3>9g%pgQE2kUMi3iS1_@iqJVE@bGuTqm)h53nlGcx|c@s~?ttu5=eG(2L za+}fBq8*{2mDa~WYwI2dtv_iOv{YpAPi)NSs%b~8=&J3Bqm?Eg)y;(U$Ou*`9gr}U z+!Vx{B5D2TS|~SVboGfkh|q0DSBsj20#(`{2db@o9H{=JU7)%}UdH~- zjINrU#EPz_S#i8BBY=tqzgD8MDKiOMNnkzZg%Y;$N3N!g*QLcJIo>81x4 zr`tP|*%Z3w_710c5EfKy^_e_~z)c}b?OZMH5(-&qdmOU1=5fgSlXf9X#FdkNZ2iwh zF0A=W;bJgiBpj(L+=r-7Eh67Wpia(_An20-m4p?zuka5z#|$=Jxw zfzgo>97h#F0#hyftK*x6AQ_^#+^cr`H5E^N>JB1q zQ?OF;)FLaPV3n4~!D{Or2dh767pz3AnfK%AZyE7a6P7C2636HoiP8JjiOpCn4Hoee zo3IB34^PR%ToRg+sDgNql|r`2SrR~bK$BgYE5nhL3ak%F46kjjC=B1l!*I^;Bm=2W z*g@oN6Pp%A3B{%~J&sLV@i;d9NxRtewx<+W1@Io!iK=EORk0Bc&UH;QdO#s@S&Rm! zDj+c>&zCwzmBbXphdK+K-%X&?=EfkB-YY>j+B~H|_jU?&dT+^R4pKqW=jI@ywh2s& zjf4VIS{( z@BkJpK}H4*iHeW+OQqjeOJfl0f7Bz9US zk!J7T)~JLD$@@;&CPQ&KIz?5HOKFHA&ceM4xl3Yka5odVr6UEgOXad7NjpIn_RgXu zbh4rFw23c>r0mwyrF5;QMC=5`b_7*o!#QG6kCYGKP_-?9L)D+O3sowI^*5Z*&L(3g z%tfL!)PNebDvns1fW+)TjgF5Li70)Lz?C!<1P+}6_n#2JwTUo@q}Fzs3bO2GRnivG z?dZs-+xHk?*u5AJwx&EYGI}OC$zf#kVGv2#B!_^GjV%Xv381P-ATTrFQfmwY(iGQs#9||Xs4>{u z+Q(t*Puhho6=3{;3r{1dfUDU^A$|jCQ5>;jT}on71<8SGATzLUCIhWxq#zLJ46Gj` z0P1;% zZ(Bab`qW+{^h(-g6pA8$lq!GxQ1z4LO zgGgE*utslTlAWth(?LXT3Ro(pvbaejX?;AV+PcR9>rdJRtd|N?aYq<^G$YMT3ULTM z#AtisSf%+%Ob#J2GJ;id%2eVji7JREbp}(a#AH)t5J_v5nEaN)BtZ2!JBZLtfl5VI z7Ex&=?T-W1);T}J>F6)jL6(g}5ql9eYj8C@lf z1@SA(i(AZ%$#mEz(jb!7kFG{?&#L>E;Ayzk(>R}M%Q}YXeo`UK4sF0e7=n672%1ZU4tJxZ8 zZKcU#EKynlftd$5++_hltstW7m_=?<_D6y0XdeY?VA3v7;}OeEl!XypVM3BN!HTY? zSy8->Ns?1}g`!^w(jc!JM@{L%FehOfAv?7gP%VTAq8dE5sWwQcs!$CMbS)*T27|Cn z$Tm%giP$=3k(@x(NU6!uJPuiZ(k^6OKvJzHCxuZBhPg7{R54@X;9REi&Uq@tEik)UF`2@! zCee9Rh)G13>e!~oAl}aIQdP=JO#B5gJIZjWO(=_!1fs@(Y3m#Zra$Rsz%;s4$C{E9 z>QW8Z5yvF`x*M}gH85frapVF(!ch`Y5GU^}SQ1^TW19+tNXiX$kjPw+5w>Zhq_&(a0utyn21r}$I3WE=yMS~PRcc#SG>QtNOLeRXNTDv(pp|f3 z($phn4{C5s1ewHsyPPXk7m~CS#E&|IC(-OSwuvu@r2HN>^D&F?tu|b!lUVg>IEds; zu}TF`7WW7g9K)(@0vxOUq+P6%J2Dr@;**-qFxK3oV6$7)tT<+A3KF;3Eh+#4U`iuo zXq8M9#IHKT?vLXAGh>?zgGl;WGGcCSKBO*rSPfkAute%O;g-~=<{+ZC30@W-2{ar7 zuWbVyy#AzJ@VfY<)52H#s^x90sYo^LilcXZ;YoF~!-L}f!M)hc4jk{PeC$GQe@XaC zP72~#oxvC0WW+9r7~8BEL{jQV<}nHF`#0}Z&|YFf8`;m2Vb|y9AX2x9Tox$_lp8~? zt$rN2{-j;xQbES2=l`h@b~P!fy4i5h(w8hfssMdjtPfVJ0NU`G>R6`mx_nfqNk5F^ z6=ZC9g!+WU>w}y3E4+>?;kDv9D#+YXm5JEYF0A`w+aFz}dhG%@*HIHqUB~Jy32c4a${$m2aERGTt zZ=I~+_@IBHw|RFAzj!+<6>CAG9U+(4;x4mDO3L^+a&6V)$n_`fBDY$W{CfsyHA$(O z$#A&RETxAOlvl;_V8nv5?JHprQ`rU?V~+|==?4gm+Ef{2HF;FV(R(*HR2*Go<1i3I z>#~2Beaa3ZbJL8Nimxo55{Mefh~X#91P4c{KWP`G@j&C^Sp1m*Rn1eXYE>MpBSj5b z*i^_|7psHiYRF6;U`fnMz6#=1PDDDkxpdFs`3pOH)5Gc3%P(Ktn!YmIdg|__(bf}p zudS|cJq&-$HM(_ie|x%n&-oWG%`A@Z0urbNg;|?1gGkCP`z|y9x9*4}BdgEcL1b=< zSt_!!=t|4d{CH%wm5*cApR|ivDz5mu*grJjs_9BRxN1A$a2=E7FOn8f8{S&3a5b4k zs${PqzSJ2~iHg&*&7VOetyfj@8V2okg-EMOtolqIMCztkrQ$1#z68pR#aCPXI9B~h zyI8FeU;n}YtmZI9%TAtMamdowE!8PId4q>>NgYX-w@nepw9As-f_PSE$faVeO|U^E zZ4f$(7Nuk@HlWiZuzO{(o0R$S7;7sZ$E`nU7q@QKO9nY%##qg6;>B2PP#mx;;psa-6K0<|oSkE7OBJ&sy`(k^O=U2LE^&a4k>3KUWwM$CzWm8L%N>%+*v2v8{v zkkM6=T@XLADmX};2uf=uKX+qu4N9|h3Q}aP$)?*Nl2*xs1G-HaTYcIOB6CxeQu|jH z*J)XrA4jRJd>p0zq+OK8Vru}ZW_?%_o{(f1wJHwQ^+fqDIy3^;s47UzN}3DeRSDo| z*KrWsQ;#|>+e8~gQhvMKsTNyVLDnO%Z)Gu@K-5^h$yPj$SAWtjUWu3zcyQHdms{2h zr(nArVoV&S8};|B+;%xcrZ9dfbgo2LO`Szl;^r2R(LD4+(8<5KC7`@?)L{^o2~SJ4 zMDZEBY>7yES6Fl=WqUN7I+{n}8JM&S&uV2QKHgkXbcuI_=@MnnjzRP^kwhKUfSgQh zHcfRYY_bSTAZjFZI!ec}=})>DHh)u}(yVC064zKbESG8^heYye#i2MQ?isv`;81uj zMIutNR1iP!43Qrp5NY#d5J^qkQ1l|`;xD(%E1P2l&v!E5S?p6uO!`b6MB1j9B(|R` zEPB$4G&~+JZMEZ=^d}v~B(VuA`#gWw0H&rV@y&!obF^@u!qP~b8r;LhQe7re{5e+= zp^}q=coMQhOi>xKwxWXQYF)8eF^Hu6u2%B~^!B=1LFywAL0QBk5H%J-Z4=;V^(XD3 zwc6(MlLk;VF)7&98Z|2pTH`@DZVPL4v@{`_#IEG1Ac(MHR*OIuL|5yI&5}VR{VW+Y zqg}0}+7oU~ebNphdecx#rN1nu5@PH?1FzB5x}r%+!LHW0U2*i%OeJo| zYg~dFh$i8CRG3Nx|6#=@%ODt_wyS(@M8E)bHm*Kx2a&o>sSl*taL}#*HGF>S{O;uNa6osik^|JhB()JH@ha&mh(C3VEcU@e(Sw7U_l%>fa>$%2$m;6WsAGuEiQCmm$Q8yih1AsugQ>+Dv3=M>RGdIY z27{?yRmpEaH1S9s>rz++iJALX?r0c|iHNVOEEk zCM3B`(Xx|gR}`|VO&^V`Q?&BLtASj}aY3TMQHZgtENWAOyx3?qR`O7h_Gk z;*cHHI@-yjq&Yx1N;a!LlgO1U7{s$WLoQJtUbT5RJW0v=aG=2{SywAa%gl`1jIb;U z6o?v&u(s-P)cTWlQA zIt{DVPPNVrYO6Ng29dN%9vr}E!ul}C+Yw5s{VR*>1fs@JYAYW{sXu8KrIlmrN%Pjj zRZVz8l3~=Ucx0uSPW;Rk9U6gaEUP$4Vph^z5F~UKSBXa;S8bvVA}PPCRlQ2|)a1(a z;OYH?i{J#aiM_TEXUsCdE1PtEQS-PHHKGP@i<=nNk{QY1etkZbhWN(hEuSs z6)`3bQ~JRY&z;N2-{IOC5~=Z zS#&04dwi>EYaWNEKWP`9l!N+}V56%Q<|4~V!LC-!nm9gJ_2t3b7FKMK8fXk8++_Ha z%$2JRP@=1K)#lD1k`nO_nx@LOxoB6bvP2DXFvxiXu*o7UfvB<2X)7Jara$Rs*fhFY zS2bZN*wq>^77oibk~S28jfOF|e>EUa7{MeWB})bIbE^c|;+d)-x>{Flo(v)>zgg8$ zT*~dI&(uMrZJN$f+fNof36vVcq^))wlm4V#Ojg-``dzK7nw}KwY7N>GhvquFt2OxH zFvh4-ZnLh|Rmn*~JgKv!_DAr(bC8jd)IKTW;zOGcsCM$@Ks)(paWaGN&35+pTD(yS zggycRl*LO@K7c2;wgqso`jd9SS}nQ#Z@Q^0KT07b520u;G{NxR7P z!Yl=4^O0qki`cKY-*3oOr^zZlB9CAgYFOFK^J6XVE5fJ z=-#(^S_R!xyr7HhO6!sfLEs|*TNX)4*&m0kt$iG}{-m41_K$VIUDG6`$`-{Do1CTe zh>Fh}Vr{TeAwFvdB*7|KDpwn=rk( zYb>4;h#J|*chx;hG6Md z;RHFV5E8tStb+JfXIV0FX0v9~We`cNQW=#pvk!O7944ofy*RzHm@%5jmG-wKWP`D(Lmy#t^SD-Sv5MS}ZK@0+X??&VZK@&R>|q%d`8yDcjoF+nMc5c6UBBJ=j|F89Ru` zRNQBPnT)9{ni7Z_i>bEmalrbMb^+^#P4duF{Jw!zO;ciFRofHC>WV%>tqV*&OqvKy z#?_-jR-#HwHf09!`C28Wd?76Koieie+#N({1{lfJf{Lsx!qS?wKOR|a?c+f8CmjW9 zG}MYeH_)nCOR+kWXI31q^n-CVibmep2#DizCpd;{lFNel)scvS4nR{JgGC7Qqk2W*dUTp8D;&*Yc(gDl`FEeAgimohr)0(tD4pdwF zI8gmbyFiTw6o21pMpsR5VntWebU0q=>zC?AS95Fxu~B7^ustedC#o83Q*96gR4DQW zI+&8Rt*-feA4Fht5uh?(tl6B3tt^rgh#K3u+M359>rdK+Y?YnsNvj&HNlxKvFk(#{ zsWiEXUD`#)MNmq_Rk8tY{iu+dh~g5+XrA>U{$3&d4X8HZo)yIHh!#Oa)pZt|Nm(C_ zs*dhahz2I@LNpdg0`28SH5eu&Pi?|EFltm3sq3c}*m*Acbp(&`$=)KazkE2|f+F*a zJ9{q-`gitJy^D&d>qmvj^illy24rLamP&*llt(+Oo6G8GXCZjB6MrR#N`XPtj=)O9 z({&byNm(8TtF3n&tp223uvUqu^WwL3Z3*TgaazG%I7XL99>o3X#AYm(28;NKP1sAq zr!_<@ohP9wSu03{vNEz3k6EcFlGknC3?eB7O?IRC$mjZ3qXmPkM*y2F!V-uY$#osY zG?TAuu2QI(95WaW&JlS)L?V~jX^zPh4!a~eB{>E0`OeV!-9+HD z=`o0;_sZZn+B~I#VO~;tL`vyGwQ4NtY@J%}i>ogoMh- z!CF43h5GBWBz>`qit#r~F)It(F8MLsPD(Qz1%{M4JudRbf~G!%|l znvcZfH;`k8@klNdBv>UQ1%W_kftKh}UAH+gh@@<=vfEivI~1E4GRpe&97ODky>Qg7FEBe%BO@T#jLwmY_CM3&GAzl(;R1~Qj*d0qz!KiS^ajjVGy??z!C+g4Hh>EM2*B$NB1~j{YkrkrDBTDe=z!JU@n!g zs_ltmbxE=IY6DCUQzM85AWy*iA?=0~pN=zG#IZCyvN`AYc31SkcK4%9}Z;N}@ z4Hi)ebQ=Szt$iG*{-j-?)`+ZsuFJzPA&J&loyjvDj@OaKYmK6kHxPo@I+2=$t)#Ia ze$`pAiHfc^kp_{pesnc@3r%3wJql2j zVq#^&+MYOC=@;Ob?0qkz)gvSDrF1~TRFYc|PwEV&RCKlJHHf73BdFCZDtwvysY_pJI0wg0M$m=gJ~EfvBctUqZN zvP4`7biWzZ;0;Z33U#TD7ZD@jNTtb5%q~@QTm-n({7HZw6;czG{@RQh1fUd;uR>}Z zFn6F*Ozv5IVhE{dTux&#Ho`M&!AW18#B4;xhE0u5Nlv-i=p?#SH*9(gA}PB|RcTi9c`c4q?E;NJhA1_o+H+4Qwn)NPcNg+(k&r43xyxFmL-ShMQy_Jnx&@Nnmay)E!m6*fcNszIX;2^EohN@go-a8LxI z(di2kr;>nzEb*OXwb1s4dkc$y#BNw!PS>zXV%BHmAWCkE*?&V}*5V=In3aA&VAi$* zfmwgzE@rE3KtE~1RdbO-ZL3k+63|^D_l6{{nN6)x>B1PN5iFT0h;wy@;2)(BY?EUU zQK=#?yQ`JdvA=y4B!qpgP7To+@D?P5Ert>fVd)42gl#hr5cVhTLfAbdCBY+J?HdtS z6O=;zt8oJo7;bR060?CdE?ylFOQs6qUY&vXhbR!+WEn(MHUlD{$T~Y&62Cre2a%eA zQ8IZa@oSNlaQsRiAn-ICZ5t4{^(XG) zme}tV{i{Be|5GM-HGe78(Hb!?0pDc=U2%DR(vd!dWwyyi#HJHy$z(xXtTPtVKc-M@ zlW7o9+3X}!Up1Rr$++uNdJu`5;+KrO7Lf_Zue3jbUt9YGe*KBN_^lRqf6s(2O=d!E zN`Q7{fyr1xKaQZDoFxHx8mW@df;dxW@%9HOq}rSsL{#>fCY#(8v``V0={>H`?LlO2 zid8b!T8t(ftI`7qtlIh~urdQ8Y&5|5``W)SVXN6sv7(e` zGXl6Y2dYL<${QSkZd4sKekCymaj&D1Dcjr}MAZ86)@WDcE&?=qeU16bw5cWQ?j{tm$_H@7Iu^hY8<@C@*l2+9 zVQa=)n3QB2v*N93TpYMd?OfqhH%^;E)A*J286+Td7H_E<@{ K}4-kLk=`PRl69G zsv(249buP>wo5$v6bM@qkx>$xqkRIq{={AE5}_y1XJ^)sVLlSJp@!IHv>_H{JT4hG zIkAh$nCFCF>ma9FHPT9!4C4NsA(g5`E!jLBMAQnIaX{ZG&uc+mCRVv&9m|ZVeeDvD zCk4Vrv1;p|z^XrS7pt*28vv|XLxxF7z=qRf)U-sD9dYkm3$6#H=Y+8upRNyBm8{b- zR`O#IICO?wXu!i)d7Ghwh{|q`U2fVOOGa3qB!tM@v>HW5SROlSNm`%Cj&0o&sP!lA zqLv6bfeu;d!c+YOOPU=OYLP`ON`Q5|!0dxXKAmR!%EYMT!XS>+**+F-guyXtvvCkn z*$wfi(NyB%Vd85vso4!w+4jiP%0@eh$~rHT8o$!-$TLFCfy*SvfE{qZc~$tmwe6-B5SJ% z<&m90*l6-=>zzQTKXDhK)xszw_Zp#^>=bI64Oo)^=SmF#k$7e{+6JVgaVhC7hy!(& z;!<6+OE$p<5tZFF>*y%yw$x|%Aktb39(70@hOi*Fr95gAs5OdATk`}m{fWECB=%{X zuGu9`Z3=bG28~JplqNMXyJmxfBFK+UpT?;qvmox&8BVFL*(IA!gNRBwJG-NHrBX*N zePrWvdl1RPP_&M-$gG>(l=4VTpyDWIZ7UF%^(XFPwpuhbyJnX(sVUSo8?_$+-8C}n zCZi~`Yc?uf7~?d8CAkH0uFeolb^cqA|cI;xe(2@$z_qH^IeX56R0`#jyQFzeA;JV{R|2~9g-}f5 z137qV)Q%!tr_NAIbo=}?- zpj{$C%OpXe10!&yG(cv_BguEU+DN4sPaq@%scgnmP`OECtxxm%1_a^zOJ$m-O{C;uQabBDvYie!}Kv7SKK*zOj7#9TQE_ zJRgYpG=3#H263;W5pQjJ4kBv(cx&`Mau)#_y*^6_k+~^)$#~17M~$Qh5Sg;Ae*(S! z#9j1~!N)CGS+`V-G(CzHZ?#nk*wSy?)lHK1&^1CZjaNyTK^&?xyi(^sBb&5?h+034 znmvkE7C@38D9SNN+)*8htRwSiQy^?K-r5!*5bICeMQqb}t7%iLcxxJ$0PgBSjf0*% zpqm3A@QhE{Pm@*RPUBb7XAn?yhF`LVY!i48Q7hDt0}Wg0x@Jc{iwKdJ8{BK{Zh7=6 z5H^ZkTl)ld{fWESC8BMh3D2q_Yx-2UhK$&hfGm9@6uX9u42(dT(g2NA$&x{wsk5Ce zS&6cFI*6zhGUI@9lh%-Zx)36B)6AHRvpk;ENO}N~8Qc0Nus48gWN`4IDT2>fn5ypb*njP5;9Yj=i*X&B0U9%u-N7*sC zpXITmK-g%6wRKOR)}OeGT6fDzQ~S)W*^y>Pg}P=DixOa^c~Q)cT4Y!Rs6=#~Cowuc zDvT(qYZinw&lwPmvb$yjs!iB63&I`&kdAp&C}n;;v^vViK^mC23sTBW{c5qqe6M2x@N~V=>~x*)%C;fs8xDL2*_kv800((2<4HT zK-gFqb@Wak)StME&}v~6l6%!PJJw{UP}gk0Vgxv8niI1HHZWvzTt)~^ou#-`*X-CP z*dVxMcg;F#O}Q-vNjpL&wJja31s>ccacf##W-EFFew>8LtV2$qY?n6 zNlnbI+2EiEM5EKE(_BerL7cy{axc|2JGSXGh^SQjWOvQ3rt6voshOCiBxjVF$cHM& zJW>+~8^x?`1p>4F#9ho*%XiJLS(uN6Yp82BYCi(HV+3KOu*5mb0<)JkDl?5>Np3-4 z(HVlNuGz6ouR%m*H^Y(|OSxtB$v%ka449HrcEHe$B4gvF1622ot1T3Fs!j5Q@paAmyb|D|s#m6gopK z)ipb|nKp>1?5ROKIsRMyD4tT?Jke)1PYGg*0uqGTYus%ZoNVju>lNp&5kwO zDbzI^F)smM`no7)*KA~XBtegB;KM}M>{v2h5EttV#T4TSgk+rbuo+LLo?Gae1!x@I^IgY z3xd=ZsS*~8S5S{VZ1ZX!9RFPs* zahAz!B%~!H?ebA!KT-845R#D-*Q!T(jg@H-3lcN);x=RLGLHhKd;mwRV*wnofr-0_ zjRqK>2W586!lYzTIaZYNjEe(znI=Qks4#hhBP&Bx9W;I=IR*&`o$YC=H33(1d5!ke4-mD%3T*TK%orW#dLD zcJ&wYobdaVz}}?KXDhUvA_}ltXV@|)?}%0dW>3(h_dvY4>18FgcJ*K9!T2`R`R>`?$Ij|!#CPi$dr z@PJ-Dj zw8}1;jZ7!MKIaFKwQ2H8mW6p_Cs1lMjM{o95b9678A5@Ksk&yDHQ6cDH5)Jz0Z#g2 zCuY}dV8|L!rEw|gEr z!bXv4Yo0)+KXDhC#I}^vH4BrHr#7Lk*`QGgfYPKUX4hC-rsWERBvJ4>YESwZY00O1uL z%Skx`zGZdHfI~Pi@n#S-ltkgX>DFHNPHCV%i}qLu(1&ASb)H6&~9O zgpK0XwgG`#f8s7~iTz&DBQaQBB~M^#gg%Y zK%=wROEI2SY|afLDx2|S>bZrkS&+6P{8EMF6&~XWgpK0Y);@t>f8s8Fsg0fwU6b(y z^O4XEF`iJH5}>WlctS@(ut`qWKb)nJD)}ym`*(&^suy6zX51j6vWZMVeJ72zKJf>U zxoMe-jI}(*6DT)|Ra^fAR{e>)Sd9i2AFw9lX@zDy5vC;|n}qR142|G8stOvplJkPN z*3m$&&ALHErOM3pQLTF~t(`Au$D8IMSDx(u?0F79m z34}=96tQHiZpqKh~OEvn!eh6)Q@4#wCDD-vw2p!sHE( zpgF1zI^Ifh4B}obo&(wOi_Da5dJZCL{g`SrK&C;j&k{mpZi-$q-ty>CE7Ai9^xFC- z(Cbg!MK2kA++b|oQn8}xQLK2YtxCXlt$DoF;~)qnr(AVsD=9OGLv@B%GTz#x9Yoan z@z&~XOoLdTJ%mWz6tQHy<#9K|9V#Qn2xCC%%`c&O` zYYvW}IjRmi-b(rm;$EHYZ^;_6P2fR9tx!V_bWNq}nqBc(M2N&qu}el<9(@XgjqYx3 z?GxDbC+=d`4XD)F?MdtA#T8AT3fGVkn-Y+%aU;}8^4yl$d%K2p_>xFnjnJh!Ln@US zgOCimRE5kqVAYg0WRSO`^ByYBuJU+N$^-B?>*ycHYGC3~ti~eENnzHIVN&wYC!8Lm zro|z}| zpFpjxdjhro#G|OCc6-8nYZm4s0UPR?MJ!5yb&1MER2rs|@`Ct&XE3F@W>;;}4I(PL8CEGOHOUy{WFnMX z(?Z}%Ba}yW0%4;Fwe?OQ)StMEP#1*MMs+^xJycz@tD5W->Y5E$lK|(K3TO+=uGzql zH84-(Qqo%xXSW=v#e7##U9+n;!G%@fG+<}DbzI^v=sqRn$*PXnhg$$Kr}ji8mE%Xf;fL?@ssMB zUA5^nh^W?*LgAF%Lc1Pop`|hFb9`V)6CTP@!;yJlB4 zsVUSo8#NgL-Q zgbcf6U5UqX0v$&oY@30Aus?AZ!qpODvuk!$6P!X_vvK1R7+xjwax!^lw$#SO3kPZ% z#FFEJxL0Q&rdSiJHp>PPmCc$6C^Gqm)~EX*Qn!j<9?uEX8^y0}0Rq4N#9jPW3&CdB z?5gHDg$NU*T?yzKOeZEgf*dJ~!m7&yjatcbL7b{H)KXout2WaH5tW@hv)gM?<8ZdS zKIsRMyD4tT?Jke)1PYGg*0uqGTYus%Zi)RKr)zdqvz;y(6KYcewDd($O!^W!a2Ska2quxb_E&-N1VS=+6YO})rbY#=z=*XVZ%0_A zV(l7_@uWNe=T12K$FUlixQo>qvDRcf!KCEbY>4qhSPX~kS`x++F*E|&s48gWMi@_> z#aN1~xMs6%kQiGZkh;H;Ul;n)p1AA$h3WYhFYNJ=5Qz0B?jklCVCXKM zHO+&H6{S1_62PVJf~rxc@&-rH990J$ZzVbAYU7uTw>CWo5w(82HToV+gp%hlWW4oR zLWs;wt5RgV<f$ zi1jD#A~qUeBJoz!rdaXTG%f+$WdvDqY5fdW=iGJUv?(-=UrC=q+^a=~qqUb5R1JB} zCh#DlR;VEdx~9@~&4RSdT$#I6)FC6U@#s?^Y;<>PYoEZbKXDhkZa}3~b%t!8S zLtV4$SSKn0S>r}1cEK3)oc7t^ z&92!s&5jCn%_0^hz)C-qcfXk5I=?$PJRFS0(qM6S|NQBPo|_)b7N>jX_qV49i@TF6 za9Ds0i-4G#J&jSxg+Uy@GmKJQvuidR2N9LsH5*WELJBen%LHlDEi4IA9u*3Nje^ux zJ^@mH;>|#6cFnG77F4Kf7BeY<(se2apC_45%(rPVp~4uZVJayvi0^ln1XEqJYc}Zy z5!G5zCk(Vxe}1LTt6xVC#q`}gk&7MvYTNYwWi#bf~3sY*%X=7wsf6G zZBn+!x22BeabyN2?jn=emU6mgVN$Y=6zZA{8WjiVI!$U~cFhI{MIaiTK8;gJWuGy&l2PYPPkPP)yf#Cn9QbE1kAzQ(1OHA0lz$+69^j(!L|hm{Q47bhF`O5c3tzFLWBv@ zTm*D!rW2F+Kn|W9wUXz8K%q0#QeCs_Hq!be;VqOBi%L~lD*T@hENaNEcXjAkSSTbG^7wZhg6ys^# z=G-8nvKddN4O-}$1!<2$+~qNzK-egLZS52I^(XG)w_4t8cFnHSj3?Bl1Zb->p3qSc z+>+DvAxk6muL9#~-Dcb%NM$pgg8EJxYklGmB6F*;mdAJkz}}?KXDhUUYRNy zg{*f1)@jBQVOj#RY{nDuc?7G`*--^D8o845g1A;^F_z*guG_2|MAZ7oHToX8n`H91 z7UVq&QI^Mg0%4=bwe?RR*PpnHTr%vqu(rB*)-~%XL{s!orL9VUmS#RNDU8I>2%sq~ z(0G;f7X$>I;gzgM*#sO!)LLaGzqc_BVtpnMB6X{=mPdhFmp(w$lWhwSi1jD#A~qg; zJlL9V&8}-6RIDiF8J7U=dgay|1f^xl7H$iO&%J7Og zTRn8@c)M{_=u%X?1tA%SxXF)v^5xp>U*uCTKw@TwOi9f^Gubt}!J|znAHWgoSO7LYd$Gnbk?&u@AWi1W8`!3V z);@tYClKNjHe7>}FV{Tg@G|w$9`hk?QjSIgbKDd1NOLHX254 zy%PxaC+;GY3M7767?OMS*6fBRJB7Mt1J)$Kxk3>axNpq{hD?r2NpC@%y~T)WVS#*W zcEcvvAfmFnW*xPrq&9)19U+t4mhz}gAZ!$ww&n?B`V)7NS!G*Vw9$&+HQ}kLO`)#Y zpiv2cu95&{cFhI{MIcISMa$WeX|5!*AkN>}R`kE`iFcs6xJ{=)M7>K^ao@6eyQ<>e zmfyQ`X@7qguH8uaRYl<{7{_*HW8`9PU(rir=Nb@J|B8>H4rg0SrBB$G%Jv4w7LI@a&wXUgObmMqNCqn2FK&Bg zfA`AX{zZ7y<y2$rNK@1e-E#JVR!$z$?n$SYN-?SU)BrjX=^k%p6P4IlVss0 zapK&;<-PNh*>upq^B6YXf)K|aJ)E5z^iTD&Kc77~pZ)ndFjrWkW3AC!<$@(PPn?(= z96R0ry^H1~=deM=E65zj*OF)8p`nYHFD_mo9PyvN|1ooBxR<{q zUXOf#aBVrzFo=V#(GZsmQ>3CEic8{Fciiormj>5@Kkh!d^U@HL3K3=;AMpeJh=-S- z8_b=({M=Bk8{w}nzM|yDh7lj|^$1@#@r#bF$d@3Me#aUTlv=u={;SQBM(C= zj!$LRSJX56FAwJ6U&F8%!?6FuVutBUNg<~GXlE|(4(8xrzWgvT|1Uu<^*L-ju`OQX zXmBw1@bK3y(3;epI0vBugnCq7fg(2-nS z{8jUl!dzSMRr9^2$TAX?6R07Vxzx{z_m$t3t8aDbtomoJY5MW9uUW$n!??)`M*o~m*i46l0S!= z?_8gXFB>>5BZr57}+1@Tzp+b~l{y)kxc+h@Uyyk_2$@b24 zZ+5tN_QK@SRD$&unc|$-oQD*r4=Jy4KsmQ?Qrxt)e|dK4@@y-wBOVUs#qrbdW8!ng zt@@kUpSG?XZk>|fH9mrhg+wy3FK*Iber#uVcW^xW_t2fgFeHF7*uu?1?Cfn%U)h>n zf%*1N;f{Mryj>5tyMK7NcxF1;+sgh9*5!QOMHN-f?l!5l8#k{9t>wJG)4iUpI)YCt zzM&^>j}4jSHu=~?M;$|^wA68qR?`h){#-sxrXq1cya|;n-z(87xSg+p@92p)qC=hA z*_{rKXa62X^)!q~tfz`w%Zb}z13l+#qUVV5`r>PP;?teE=>g#>R_WgZIo@@f*XDb_UVq_~s#CmOe8%7{ zd7T3NRwlfS1|*>u&~V_YTZyDr{9(Q>!QU?)Ot&wepY81L!2`0}JLji|LvljW40AQa z$0YtSLgI`2+k+F2?e6a%oY~(VQWJ~FU0vIVxJm=!>wDtOq``D0C=eK{3uLSdR49eg z>RU0f4kmjqOcxJllY<%Vn#XF`){+g;oQO-dJ>8v6aLIaYN%mWQ_)wt9bvC%>*-Ujq zj%I03jC~?kL6+y9+u434Pc*7~3LcvJS+ejJ%ssNiXtC%)xF%oX7yN~*H5>eI#9C_P zZ869V8M9Lu4y1L}wT-#^-|vgJVVZvQ%HHJS&iS*mo!R*dgBz{?9J*oIc13&@gvAp7 z-8}x`u9<0AtEIXUt`)tf2mOK&Ly&P-mh>xzh4TX0DM2N9@Sm2yEnvJh~{ zv*|ECQA$+}OdN7Qkdfhc34=dQamf24;fE zG}ywMyv^wiLYzJUU0AngomkQ$y#;%Z1GEM-V=^`#vAK8|uxB!VQ=vtoWuFrt@^Zqg z`jA0!@r_8~civlbd795(qu|4f*K%d(*88`?H9p z=9F5*c`wxI2^6z`_bZnMbMUVQ^ZOfO*;&)vLBN!Lo-MN+Y?*t5+%gO{Z_c+2;#2bf z?9zbv7k)$!NRZX3E_z_MO);!BlZZWfz)QUD#p(3Y*7nXt*bLv9ms3@C1A!wyP7~h} zp9V-<)VkalADFL~2e>~2R*y(-coIae*qEvo=q{*^j$_|}A6Fh66EMr>-r;!P=Fa30 z&K3&)F8||kK>FZ%<3FO>f!Sy&UMH^Kp3Ej&&+i{xoXiIEgX2#>@xDihm48ma<>hV! zJTsY{QL9e==fQmOcN}Skx0aQ|m!~@~T!6~s0h}dgM3@;Q=!ci~52sttrY>oFX(e8Ud>-#(!V^(ep$d?)J^eyv0=@sBKRO*ytWVaFTo5JGjCWTWSbj= zW$RIM6<^u&a!qSMXRVK$W_XJwb$0lp{UN4V@m0QChQ*P~KSQ9KErBwW#Kceni=G?-_1Lp1GO{Z@hTz^7FxKH(*|I7-T zE-<};V09_NgPWq8Ke?o!Fnq1$QL@m;_o`1YWP+euL|%^c!+Hjkn(f6A;e_I2W}!j+ zIE~ueUA@#hIpyy$w*wI-Dh|bXwvd~i=w*~l|=O%j>NGY}3Xenc!B7T7( z7pNzbi>0M}4#TxtC_laA=3$oc^{kMDRKSBm?~ILHdIwBM9XCY{lPh3Ybro^8Vc3DE zfNkVGT56vjni;-W5wbr$p5;j=X~`6{o6(hr`og zhDXt-#jiwE>nBbsZK^2e@QuDg3Ib=51;&7Ar%rjMXa7}}IqDt0_}_ct6y-Gj%)sMi zN_eX&>sh$Xa{8g^u;PTZVdAt}D+SKx7RnUG7mGW?Gb$x4FkmG?cgTd}Lnq!h}S^YQI>m(EL`m+tbmkV-GK;p zI6K(co^Ed)?!sAGK&N=JIJbc7`_+)X#mDBvMp&mIDGI^PW^z5At#UV?b^ZqP(a^`T z6_$cvzPl5pzvLe#pGMuDsqkkUK2j9K6Z086!x<%g6p@~_#P7LyV|hQoT=+vZYNRs z?|R}%M7zaH=Sxe0+dn370J8^mjbdY7C|_+PDr##OdOk7uUGim4{fUX&ON< zLcMMj2<>>#rh;a+W3__60LE|=L%QOn5AmPqizdJ}1PnqO3-V1OvoU6IHjq}s5~uY} z5a7k!CQF7fw=?pbp~>{z$OxO^db99?2Cu5WJs~#2UZxr@9oli}!Ob?#UH}d-z^VXz` z&~`IS3sLPUaj^Y-G9Bb6B*o?;k#%#e4!@v5B4d~+7+%=NO>ss3P+z?Z#KOF8R+T3@$BD?7YBU`3gIfXQ8c)|P_@RK6VDEwaav=T z%fNF;Wv|Hh3$k+D`GA$ybaBpn*Y1=BzhmvK76%`I4~2gj0S6Z*kljMU9+tPyAT3 z>Y5T?<(o}5$?9vMjA4_qoBB}DKU3tIA8wvJ|cJeZkH)fuX)1nUHwvSPW>1;lh(X}l@RMGL~jtSwR{XcA81yPI#y zOrU@tDCSZUEf$Gk$|y~3Qs>pmYlBMsy=FOd?&L*C&j0v$#vr&({c%_qHc9kSu5lF> zj)qCyM6&BGx=gC1?yH-fJlewPq^>@(xM4D<1PZPKZ6tGQ#G(k<#L4VW>7#;r$=pvi z%Z9Wt@~pAR7A%D6Mp=;(FKB3L*}9!vk`Jr`rm3x;}s+~f< zgbh=rN7+qTsU)jgWC|y__B~v!y9lIp@oT&Os3)FDl9kv22X#}@G7A?j8aGg{9u;Ut6$)$2Jsk7?U-LQxfTUuRlqKJP{HeB?Q z3X#?o&=sDt_q3adwb$@YNrTr1;GUF zp?-Ad`RAtx&{i61oX1aBy#g?~l6|x%-q(tt`lJ;XwHmwccl%A%jsQgW@D6Y!e(dsA!s@cd+Va!{5Shwu5q+TXWNP2^2gp(X9uL%b#J2m0b{ZE+C5 z0*|V=f2$=$4ihiJRTelPGA7p-v;bwS{^y3#hWg>X6^~=i&h{@oUY^mVcRY189S!gT zQlR9^NJhUbukuQ66Y$|-@{fX{tN7Y)wBT!SY+Q7Zv}287H<5~=Me2q-x!U)#`xBCf zA#Z1s+>bJbNpkz;^(&h$$WXIf0>r;<0pK~up;dQ%A$9BvCRQdHg`l|7^e6H0I(L1@ zO#%|COJ6nVuIU{sBM`Nh?MV-GTsOKY|F^yLkz8F$P`#^wYp4qJ)^F~KCu;k5)-@Su zySGkx@PmaIQ%C%#I{hqCM}}+>V$S;&G3T?zdj@Bk5p#@T`J^0J7avPt1VzYscf|}a z6FNY~sowt)R=y0Sa|^Ai8uzrG_>XBcY!h%2*=+d#AwB;=uK~Tx!H%wyV1%1i-HEx1 z^~8T^LNwBI4iIhp1FvK4LdcpkMm9-0-zdSRI~!y3{i(SmA<`n&)UQqi9VXcl(x7bO z3RaM5S~VZO;`gn~%AsCY{cZ<8mhnh5t0K`isgaSAD0`7tr8Q|Cm@$ zNgmJ}kN>5^cZzW0kzQJLdLq*hKbXeXy^?VCltq(WvSW|(Mnp;7NDdw8?3Jl5a1CmT z72cq^q;M#yrOyk)Ek#_?0jZ*C$RaH9tlE{ZF(nly3d2bV>0H9Kc)>jbI+O=i~G_K^u&XW#Fb~=zQA{~1O~5D z>kmV{aQ{Rd^=4D#-f~urvh^kP6sQH#MaTM=VEiL~Te#_Txe&p#L5y_OO!vb&u4Zbg z8HwuGdg9*fYmC-#PfM^O7Qn2d0lkt8-^BujT;8JiRnQUN(62u^MZajG9(I|<({0cL z$$Jpgv>7YLry2m%%aCg;C?iMIaUJdbdd~Kh8&Mw-;P0-VN%`;3sYd|WA2MEr z&KDxNW4(v56qrvSDg(yweA0gT5PJ+!d{v+C;D|QB!R4cf$aHr0X48X<)9szfY`VCe ze+IJEt#N&amY!g}is(EHPN?W)@xqw(+xeS{;y*9|k4>j=dzIj@odeP9u$neU%EhCx z{G|A;o_HjN?C3OQJZn7BQ(+VY1Sr>(_zaaY#NVzz)zHHv_4gLvz*tQKQ_hfL3IbX$ zUm3$YP>$W;yn{lVwe_>xd!rmyTCiZ^iEVxAI-aS7VKgq$0+%P6&9*@RqG)nJi=ozv zx;iPJk%aF62a~^EWDS)S-a|Sm;C--$Q&n4|hTnmy#+m!H)`Q;4+$LeE(wFz+sS+(! z8PasJ*icc~Myak_nNU+Jsl&jO&tOfJ4sd&(w~^puGlB7fNO zWOjtqaJt6kso|H3WgyD%ru{I84wK>@)d>kD@r{flBwE4}Ddxtiq2h||Ric<;@2%q6 zk~w|dOeJ5;i`+90Ycf5!MZS)DVsjo|N9_;nr_mPBOh(kqR9r!hOr~N87WD3vnTj!1 z!HB8jRqjW0u&#F_FGh5>My$vug)@fZnAfw2yEG(9)}aB!3p`N;oa<_LG~+gnss2^GE|@}qgIQ!+tdx{ zSt?uB@ULz_cLg~z0o@QR2$y(q___`ZGjS5)g z<}yoHD?Kbh=^+i0l;_7m(xmvHv5`$UfyF>)P=lnQM~>%?hO zkSqW^S5;ZHg1b4Ysc$Tt14FRU<7fkBeN8;#4ko${n|x9mNpM$VMEepyLr zBsP2}P7@8JwaTpU2Q@23Nd=Bd8_BDZ`R3ZImCmbq2bcP|0?2$`DKcYG*U1x z^|iW*z5W1IZj9%=f4AQGmKy%?gDFs>59rm`rgY-h*#BX@5{fI#;M!8E8VgmON~k+N zti@5qYk&MU#%rL|DzU#eJ>1&ed2z~aZmdMH{hDOrSuNqlg&wFsK)1v-FHLqYPX{+_ ziW5Cx0UbYf?9~SaEOd>`q%>AuG!$*Vub}?;OLM%B6&>MxH}Y>A%E-~1*=CFCb)~*U)uxp5gbGvAG>giyW7$3M(L1SlGL@@J^o;@D*A^XH#P1Gb zemB7i1DBLP+I)q9WIO2^(S5=qGS&Z%_%}UGa zd-Y4p`V%D!5mSuh(c670;*a{?B|oF`lAk)BpIdw@FTbCG(nP8gmS%}zX)TSQbLtkK z&lFD#9%J3&6Ihh&$!4F=ZcB{c-g8oXU%d@aS?MbDi#CwESL7J2udfr=Z%<~Et>^a- zE>0lr9vpxAiT6FSKs?KkH^bK12@!qao`Az;He0n;yZ1b^-TO}&_won)z@KWCZWMnz zd?-A%$qdro+0Opn;{N5?rOUI$rzTgX2Ty}t7WMT|Im347sf&sL^I7k9puN3ktH5ps z7hmeH0^@QOa7SONMdb2Tq|w;PomzcLX?e5Sy!aKyqZll^WebC*%oTZrtH<&1fd-xd zBWMN2cJTA!=NK;&)|iilVWQfdJ(Dax z%6O=mr8`&9^a%y$q2{WhR!S+-M1NON@%xFZPD*<3NZhnl?ESV5v)aw!;CAt*(>D&T zKk+H~e?8|v>?8w4wU= zrTETNd9a~KK(XAG z-(R`FU_MBcXyx;l0j;hmp2dG?K|{vrMsk11wLg3zNycm#H3;Q|IrAqNuj^J_e$=wB z&q$K1yEr#V6R!+)%Me{LuT+jLe#@_?E|>MxbUM%#M(8Y!zPVf5d}>w;7u2!I7m5!I zKA&}L!XDO~O*Ull@Kk(-`0i9C!OZ^o!BuX3naSgHlq^B#76$v`yHXVfZ83Eo-@JC2 ztR$#J3@%&kmIT$-YgIAt(*hAFu8bf0*@w7 zDplrxIaQrgO-g0kX$lhRyP&LLnnJ1iWYEv`x=9rC=#*RhbjtK6_h2v2PORN5%TAZN zYtU(J+Ce$`(*3U{j}PELGQ)0ixcltk2M=c6;qG$$oKmSG?IyR-wJroFwqmZegqz%4 z+#v8EFD4MjEhGPI%1y2bJ@6gXkD6G=MLL`=!_@H8G0(j_gHpBlzEt~)94tk8mjChb z?6$M()E^tMED9hd)>^sFDy+tf0m%~Cj5VSympYOCrc{ZnEv8On>$m?VOJJ2~!S$ux z1XhimGJ&-}s=-_@f&KMVxta#RnDhGZ7Jtd-%pqG z=}%G}htZtx*HRV#wLCg6YSBg0oU|*UgDwi$k)i@)Tw&@$SMign5+fyNpi-D7F;XL> zOpNqLOUAWQr@b>llq)zbR7BNGmgrxa9PCW?&QB|a`igH(Hrst7lJ0zHDsz&7%*B1UMHxB* z_7BeNZx36s{QqC<1COVDZgS+5M>BE-xlOJhH*Q|97`we*mo93pB9y;=MPGa%xjqgi zdoN5E4`-8unbwI*EIBUvr;jc8tL?F?r6*5|gQ-?3-^bE)CnXAk}%Iv$LCpfJAWm%GT`4rD^6^ z9s0WXkH^IqCime-uk1}O?wmh6+nJreFu2kB&tXeTXA7wxq?Rp6+q_JOqoqU7$^5sV zd{E0}xz(P_ZuS5C*UKu+q zEf7vDJ~Aho)DXjPeLmZ=ZtuKQZeAbVd8xIvCVlWMy{yaShA@zF81YMO>|niWe3MMa z?iJ5g#KrRaZ9_JCpefbp92kpVY@@1pZfAEofK+;SI!y1LQlssmHoIBv9MFkoJC}B$ z5hP36R8=`nf&5m+rCgOakk+MFp-b`}deme;q8IFC=LP9aGCR0D%~m^LQcQNU#jXj6 z!YgX-OzE@B;!3Ov*8NydKR-Q~wtH~!BRw%mkeyK?Yra4# zhVEC)iT(D34xYtfOoF!UJ#61P&EAfGwJ-i!d$u?{KiQqg3kQdvmp->GSEu0JqQ@TM zEt~(UFTS8X6#yHrxR#`KHMbCtrFx^0Q3X-)Z(7wR^f(navRAnFWccpdM}KE$`476=i=BQkAUZP;~8W$2iqx9BAU z*n};HNpBa}qUUA&tQPH@S0I(%oF3Td8Y2F()ukh6{&eK(A8apP2h&HB%2l6l_&d#R zf|8pjBv~*0+$6oX>53UmuZ!%~rh3H)i3%Lzk9uupb?Ty;CjCuZgQ`hID?#EP^~AFY zuyJuZ*~^`W{R%;)E7AUIv%Wm3H$zH;>SnP-_(!n)g{k7hy~nP z9}#~rmpGVPRJGPztc7JNNomW)LqSjL$hmR6+tf6u$2PfS#gDo^EiJ^g=tv8udt(m^ zi0_?ibgg7q1tlFV1o~p+ISk! z^s*H6Fa3E7ut{i8`rFEQojWVO`1J)esih!5=xGM%WNoY=X#=06k#&R}!XU~y#K%+b z9=(HuCW)D*k7Y+}4_&RA0f!W>?ugLhkKvl%5YS167bzBbAa28qHorbX2=xK&hgY& zHdU_{O_L|b9&D_+~qx7wM7b*Qez4#;Jac_e)xqy5# z{pfXTxLOfpl8{${)%>H3_toQ5S2U*R3x)A-ff{R+%y=H1jFj~4{f297?!4X}B7yoc zl2J1U)VPwUa#}~?S`>h&1VVr-$OweIOzFE5JYqgwEZ^j8s0fWjt%!K|{a8Q6)D6B= zDn-_~o1BL#;6Gu9Wju=CH}}P7WnNdQwZfCFM3YHzN}#&j5Gf-SrYf~)Tbw5GpL*gw z)pcW{px7~>&(tOe!lmZv>t5CT!jIo(ca+u;C7-DOs;PSRwZJSl)1S{4F=%4v!DK3jf>y|2zc$m^~%bi(|*dcU4NfLgIPZLk(*nE)jp{ zTL<2dXe1T+9o~amuM@YPz6Bn6&ENm+!5?mbe;yJ)TZssm%Q1tKyTeQShtsWbVZ$f) zUYZ`vo|zt8ygZv^x3djy@&0PK!g71#+l!argEA?g?fgTPmbLIC)X0PD&dCYJpMR!J zRC-SpI!0JHPn}XPy#>nBnUUd%?9AGEC)T-4rTP6k@TOW-;FbYECTx>gwimnc|K;w@ zqb<9tdf~gzxw)xYRh5?rs%XNsN#&+U5zVcHx+P5lxt`QHRXOH1 zm2*xNDgMfPG)fQzqx1o#6+sYqI+Q2k1D*|^-FuE<^C_YX5Ec9sd3K=i&9(Mk-Px;O zbDt_d$EbL3X{|liT(etq{??ppedtqC{hywFIL20)H!4n{ygpK14pXzxAN=4ZG#I*$ z*$Dc}MfgS3%Zugf8fx)a^61qbQwzZ?$}s}gKjNamHe+HC*!>w6d6PP9Ej92ugviz? zapBx7IBO|47lh~$c#FLjY5jE`F%!nIfp+6h4gZ7npK#;eIKXQjI#_Tt|0G>s+xomCt3Lf?&lx7_Gr<` zkbGT?f(&IM3qT?&5`lCccC(SG!$5SLvX@2+^8Q8gy;(A-opEpfMs?&H^u$X9-HUeAZ;W-Vcr3Yka_@Mw zbKfTPQkuw>pufdE!?U9u{x`+)qsW2b*4F6UME!q*zuXzVQ2)0yphn#6<{{lII7Ah) z3s+RF^R(`wy2+sZoFmFrzVqeewP6_E(hj9~rr{SC$!&P10a&OT2Nc(bsd)UbW17u{ z3s5Vo9==SC1dQv+GZImiC`1QVqDs6p%L(J938avZX6c`($6#EfGBwS3py9Wj<**d; zYde_{bt9XSeF7Lk&RcsuOL6)$-M~4ksCO-&P;n~lqdBGc5v3gC0XQ69db*S z4%Pw%-cO7f%Z!TIqiPiADa2z&zCu|?>wB@nprs82im9rY=WU9ovJM5MWjGx=tR`5q z$iGIFHd&a!KeyP@ca6xD zvjY^VKaR~6f1ze%k?ciAO4i0=XKJm^!m^w@y+w*ltdI}_H2aWBC(SG^rMv zYgr@@~)+p4CnbO7;AX^UMD5C3hjZ>(KslhTNthOH8k^h5W!$BZ; z)bi~?!Pds^ndx@_fh;CSJ(ODu<_7th_+nY}1k$fSjkOpHV*rk3SvC5HY#`Rd@u5*; zgNA!gjuoz4GrRgkpt;OU^G*8LMD|NFI_Ul*za> zcF8+B4#~1`942kBaXv0vb5$UlvsVLWgxFpHm}L%wc$$CvE!0Hr%5K zhEjn^CqZ5Y5kw~9Z3he4+#&-PKvz*qFktyx8_CTcxq9^~1~z5+UE4iuWp| zKg{SelVT?;9(5KSC00n(BQ_JAV|nfzblV!|bN=noth%mZRtf)tPg`GjPMbiKI=Apbo$!Ck`s7DY0KUDdR+<>6;j#^=c2fnv* zF~w8EW02TLLu^u{43bxXrJ{p6@p+>pM(`lWlcxlAg2L>ePOJp1jt=VNKce1V!BGXN0O^Eymzzv5^0~>Pc=>1`BJ*YvKcSLr6OF z8Z6c|h)vY|jD99}NNzh6h(u1tbKn#$rET~>xat}J}*vHd{BA|Ld8-JOUe62Sj{z5%)49gH*ND`EA z5NLzHxrhE^@m`QSzvk7_u8^JOh#EElP(duaHKOWYWF9Ft2P65;HYt8BW7S?Gi(LVw z$I_!#N1lK+Iata871V;Q>Z^@I5Ig6lZ!vf)CB|WtqmsAK{4Z9RshT)sD0VsSh;-cugeg;6+LobM z7j5Pbf&}4k7GH*7mcwjQ-KW;gqG$j9Dbc`uTR|myA)9JP*Hz5tJMRY2i>^*$Ra_=c znta0B=hs<(+NK%B>TU;#kt*3fd$Lb8YpieGm|c9(>(YH+Q+7d9hp?HlxsPj2NmH>h zWIY96{lwGI*yDr5%BIR6;I%5gYXP$THd8D=j-wl|TH7yrzH}a3(r`Mo|Ai+yiXaPE zS@-V2irjy_L!L+PH+INt79jA{M0efTd-d((Zb2aBWy5M1DHY^eY0? z2zQXUNYNUY`pXq=cC3SM{2Lv+jYJ0s2ER-ii^T>btS`SGHe6 z;{k(;R2KwHu9h#zh#M7YHoBK!fVrQM&=}i;A#$FgLa8vdGU!5LC_ADoVN=LtzEirp zS?%+4>{Fda&hTk<&M>UFmm_;!k%?A;MC@Xad&D&_P;bkJMJf4w2l z;yBklvud)IWU}XS;-Xl1Dwlhrk~kT|2IaXN;dMbWlT1__7-2@^(ngX=e6fmLF}a!& zu5wlAQYcnVHbm@eoky*+352JW#(ro9g*w~c-tNQO?|xL478(npY(T`ED60WGd~=Qr z9FW0*;HyzW!$L}DR7&2KA3!h_YDAkT{G+iZo?YbPKI`kek>>wIvd$n5JkE@NLUeL+ z|9ILcc05gP%I~rfu4a`Xmx)4@(_wtczM9!hxr*m@Wm26AMW1Qr+{J}18v;ZWE>__o zs!tKuejQKd;vsoOUL?)LymHIz$Ky~$>g^Rjnz1QPBU&jA3Sa7@lRD=$9hrfhc%jis zVE3xLz)mMl26mo~g!b2WxdELZ$mlNH2fil*D#Y# zXT?Q=JmW!2^>i32cS0^VE)oQy#zkV)+_*?CkL ze!q+6X4!f7t~g1L-8ZTyMTk&M4Kji3v>isJqxL^*U9@S7ZBGl<-~+g2w6(RuaQs}D zo66e0oN7kO7JHLdbrY&pX2VUl-CLypEs=eJ@3n5i$v6a0%=PUq!^p)Eoy)i>rgF9& zSd#ZF`bHrJ@AlZmo&NUd^u}}qvy^)Wv|oKA(}F+3Oc+{t1VI-XWvNLfyL8VlkfmR$ zs^Rwdr$fHt9rYW0jO#FKbQ`)zU*sn1^7X$z)J2Om>s|M*tbg#0Hw$K`u4{K7g~sBL z<|v3R`TIsn@xrWLXXr(1FuG7*xsHu4Alm^fmsFp@WpV=lt%IfujofQa4M?Q2>Q&oO ztQi2jF9XZZtRNpuspFdk_ta=>*juFkMT>ek&*HPYW)it>(QxJz6jrCBb6dj+n>EdI zhXJ6g*WhIrZ{FRRcrUn*;c317Yh+IWbGjB67LFcMquPSzgt7Fx9{faMi2W^Ue&WU7 z%zY*rSb15-IU&aZr}nObc&0!(eA>qS=lkQ~;MnNQnc;ZYZgTji3-X*~H|Zi&R$A?O z>^Owq7dvE#O{%eGVynTnbE3Y%Mu~!wTbun-L7t7x?aAr>R)1WFy5;S6C$FzpP5AIQ zlj&MGs+lh9g&f*uZ6uRp(E~sXRgb@0VkDa9;|z>&O8{j==NG&(pxFLs2IJ0FIW^n)LyiwqgKK%<)10=J-KL25Z|;T^q`i*cld^a zJTqCow}<^5131T#pNs1B`IjkgS~!ms+C0FVy96-jS>*QKdLA%G`D{2S2kfF}*t_GR zBflsBInRiU0O=A2Fi!jn5Hs_+dFn2=%4)mV73A|c9M%CiNN472qZV}b8x-(V$J&>W z&tEPw;AK*hS)3uWBsNe@tKy_pn_RN>w>dx~xn~2=rdgw3Xjg@_JzHQt1a-eS0Gl=i zfqbls;vzL*O?ryhC@^RyBqMDyOIQV_Y4c-$t2+)-(66UF z9WigW2G68JO{G-SHa(&Jftj=l&Ln+a6ByI#Nw_@1YZgg*eO+=)tJu$iWiX=l3MhjS z#!!$921OM>1|#sK%1VX93t5p8)TX?UFXq%LrA!FC1qa16V}#-|h!Ord2lU?9tpvlM zt5pW1qd_K&q`TA+c!xglNG}u1E%5@q0)D~MZwzT>-`(*Hd?t|A!IWu{M0xXcf=Ew2I2W4rFxRw(L0fr zQq`@Q`bxp zwQlZDrjjdOp&~!)9hdAOm)E+R1YZ|KFl^q>Rk0%i{U?? z0clCfZx2!m4gvD&FK0+WUhADRZQq=LUwo&uIze;_Ap-UUr&+=zFbpc;Hyz{~X%>#@ zMk7K_W)KBOq==3*T_x@qhl0?Pd zR(zFBMDiFE+;KEa_Jf<$+6pHAVbXS{)G`mJF2EPb%wbTOe5-HSrVs*MHOT)ieB;gJ z?p3L+{o$xTfspGsL=3%sYx0-$ZPXl<ZB`~r11$%!ij84ohd1lfwGsZLy>3dM;l>!mtzbqV&E zLAON~-z`iR`B0iyQ$H(Hf@mV+eY<3kJ%uEr$Xt5Qrcj9F*_0L2J)62n)kHHO1mFpV znM5MA!%H!Pr}{hhR*sGS#1e^n_)q0`1SbvXKk>t?-LfN2M(Bv28>tSRhO%`I|I#}2 zSwb!whprH2RBBt*H8o6BPa`??pe~Z@&@)mGvN}|yDRC6jR*yAnm0x#+n&qS(ILCSH zk`es6pBDVR&_CsAW3)>vUu{r zn`MyVrCkZgOVX3s6$F4PEU!$YryhPhU4(jgFLSuA6$NA(@E z4~)DFBR`g=Enu=MPpf529N6?IghV2SL?=o{CF$W&VjLm}F-kOM93s9{6+cppzrMsJ zibDhuU<}gpB*K!&C?wp30}1k3Y6wUJq%=kS#a5wzq3qaidPw%*jcNta`Us5xPCCU; ztAx;i=$;5f3u9K2b63Br^4n4)p$NgO&B6D2UHDHKj%}z>Z+^I>A-q$!7vZ?*9I7l~ zqM~1S@62KstdqBIV1Pf9A~-`QO=r<7g86*zC57|iyp5(lfD`|T?7eR|JhwR*Z9`Z0 z>lU=ZNj&nQlcuw5J4#?L_L>^l%LQENMt{)qP+EOUeE#@&2+-)U+xqH|PP_H}hqM&R zEHS++YpN0qRZ1pw_WPzD+`>PAD}$M+HJ_m4{mTP*BF_(5U)noKeH-=lo)6?Ek^}mxYS7NP8 zq>$pI1H6ZD5I>(8Cf&Asw%3LKlnC69q5Itd0|H#a_oCr=x3(k`2Ak+lHmC5yW0wyI zaP?7*vMC`WR01jW8kJDr1AXAVJ*mQIs;!E5#z5F=D>G-|?;RAlZe?M()Hm&iU-Sqx zDLwJ}*iFx6WmIq)-3Sa15ODLd*ujC$L*yizv{m)MAZ&*a|_wtA%6YCg+K z6#0$We&Tx+;Yl--%wipS%RQ2mWEM*p?`e=^w&Wk28D+u3R8WYZN+hiEQp;k!^whFk zg1J*A`>qwRZTuo%>5$o$Do?MQ{G>=eJMm4HeqV>w1XhJ5FEFf6OC105DbdB}d&x6< zFn9j+>d^!rqi>cjAJ)C+AbC&Pg{RJRRs2=njO!1`8|NuE9{(czAaxrlH=*WBXz>oZ zL16jhV6FdW;zGmnQrS} z9F9-yj<;82dng7nE*tD}Ilv|QcH8j0z0jX?>yN?7OFHBlsQ@?8PEd!gug$>h=Wj>? z<1~3BrE)G!wK^`YvDa!#>CQR5G zPhuN3XX)e8_y`Dj%tqdVbw*InsSK%B6Sa5UES4X~U#(xYwjW-tFZMcniq#HuN>E?D zlXqcpS|<=VT|aDBzvk=|?x^o}7Z&K~h|^N(Gel7kY7EwF^;7R)sD`ANOr-dBRZ;Z< zEE5*Zt+7IP$!Fp6SyWLBYB!G~4U4i}W1p7DF9(dwiaV@E(FC9C{<};Z$xLX4k=CCD zx!L#Tfbn>dw8$rE2C*9AO0dyV#9)V(oH9r~Ir8a?=~mcwi?Th&0shZ`seN%LZ@A6k zGtZBGk0Y9?jx^1&P^u+7KTU%@wIu|ipQf{t$xjDNYYyVwETJ-_a7PMSeB+8ZNXWG2 zSW=Cq!WLBE`2Sf(`I{8SZi-9qVRF@GbphO*&;-t;cP;tm)dRhKd%nT{rM%AK$mwiXwa5LcqVMjJTE#Qqbftp90W=8N0{Wd14it|njc~IWuXL}-Rvll z1XW=-chRDHtrt}MV*A&Ir0StMsfx!0nP!CdrQXTA$>-;nAF0vzit@_cZ#g;)Vj36`(njVkjkDWOWhslgH}(bJ^wAA6Js?25 z`ju1TJvq65JT03lD|hPGI@Eh^W#VKqxDB=;-g6nbMi&zt8b7{~gxO23T)l#IK_?%_ zyyvRA2f`!Xr6TjViiq)I%t|EO03#<8--C6vVA@gzXza)1bbP{I@naU+qS8cUuGNNi z*=L}Hy5ohi6g?tOj%TIqsouCy_VrkytWKs3Wz(MlqZPIVX-50fZU`&B&F-N{A*{YA zgs@rp=Rd&;_9R)n*9Xo;)?_V^f<8gAQ8a`$r$>KyL7yPCrJzr&j~DdGC8^Vq+>8zS zu=4+s9vK(4D55yUc0sgJM6)r}AH@oc1f7OtAx>ZSI4gC;{qA#b#EFB%8ls!%~Vj>ER?RW|&>R7B4uod7g|5 zjaldAyX28JPG);|P)~?o-yM{#oP1m7!x?j#926a*_FQz$sjD;mR`Xa6%{{QTPe5~v zVxBC}tcXK!$;&(B`RM!@_jk??S0>Z`cq(}^i!E6soz;cqrnOzTRD5ca7lYx}w9k2= z9gfAs8yGWB0U|G5B;NPD(H@4$MemA}R6NR7#jYiD)*8`FKISR-cAL;qyU36q?vT6C z?ZYjjt*w=6=3f1bplJr?DeZdA)S{8(7rG=TA7>6?C@bps^R4-%vdt4TyrZ9Ev}_mS zf?w3T$kniIV%>mSONmAY*4FfWCCy7Yy- zyGwF%h-P&?v$eZB9*i#3m)B#X3$582eCJt9*=2W9|GSMFY$F?=6Y1!0#;HYIsn>5? zW%*&vKEUx6Jjw62kySi3+8Xv2>3>Un@#GwBPitvpcM2d%)6uytcni|GzAM@5hk?8) zpgC9c1=8F6>gX{wmbW)vgn9w~kxa+uht%2ulA^ywja`lsnJp?4Cj96!toh!OIDhhU z1z}0p%mnnt{pb6TDn2$kb7nXmwp%#(P(k_#(V2^6i5I{EHoUw`cH1M`Sc|hE3EDpI zvi;NH@|y;deXv7*puMx0obGS+dE>$IdHAtyZJm})jYt=chMCR(yF*^k9t!}(Sam09 zoy{%yW6oZ$XXHXykWaVDO?c*1b!Ly~eNsLnx%9U&-92}_E=jg2zI8##-(-J@26ZhO zEr^UbJePqr&i9}rv~si=;0h~lhdM`Y}^7{lst;h zYVn*|1~P!{!klPmF7ofKnhu8inUkpB-=4oVhL39B6bB6ZTIXvM@;7xiC7C} z3MI8Qs}B`CjSwhi4IW^Xpg9==yY6=gndAbYgZ{Ey!;#qCkTrKQ+ABEc^`j zUc*dbi5sF(*cmRB732&z45tQx3RVQ%Y(BC^zJXI+m+}+@Gn>!m)KXN}IxFO@I1gax%Y# zeS2&a4srrB2_K^+wz`=yhdZ^Xl-`sfk1=KvtmG4Utp}Q6jRG9Q6-Lg*cJwl!++v56 zk=J48W6{83FewrD7#ct^+^~g##BeR+%b|!07u%a zuW+qtMF;|vM${%n_@7*BP$JRG?TSU6@A(BQ~B zzRr0Nlw!R?<%$&S@Wn>;iu_0}eO%fUTsOYpdw{*{_4 zA^#&KE2JJbBga_Uwjm?2PNvM0QLFkKvvyDkgM z#_Afz2)oDXzPTVbhbxZ0F$2KOetlS6{x%+?f74o7H$YnamaDIY=>!kLpZ)OXDE!gm z_MeOjyiDXZqKAaju!w|z&}JYczVZhynTM!rw!VA*EIgBd`ghmD{HTug9RM` zsYSktF8M%|Su<}!e7U#x6yKnK5VZSt8Oy zFQRKpv3y-aKORdSz1m~?A-G0aN3DOvMS*R`#2~QyGc58Zb=X>J;B^R*tx( zT;OU~5O|Kg7HR!;9x)Tf;el$##i{moPe8rEgY_TsNI{O}QwIlFx0CXJ_9%Q>b3m?b zkT(`FKxBi2O{U|~U^v*EY(X>TJE4V=38C5t5BTll??&k|A znAH?LcJ0xk(INS|7#689m_Ps$QIrUz^Dt+lnLA7!1_BA>*bI69BKh7d8Pv|Ww|}EL z@(p_8C4%n7AaKH%7lXhFTM^#5v~a|<5yd^jv!k8O?f&GxUgu!3{3vo@xV1GpHyKU% zUoOzU+!?-b6B;Hde|Z(zGu#=-|0lZ$bf~!AN}*$_|D_0&t0(u4M?3fFD9G z!By%bZ|aa+vUIQ(DDZw_(pY9x%-&U_IM12FUNFbh`d+LsXlVn1W2!3Vd7I*?tV2O* z8BT|es|nA({A*NclZ6TVliReODYg!h(WS+1oXv;jq!db|!3f`bRz|i)wl?yG7#Z2a z5gA$Vt`V7Xc7P)F$Kko+FVu`IlD)`C$=X;l=nRk@M_r#*&_cvs<~#vLNH>CUd{%{!a185w1k=H`q3M= zENQ_nbbyGBS{17%tZQXkb2uX{b?HT)X#8lvu)+=OFT~eA8k>s0kY)~jv~5F*Eof$m>MiYlC}Cp;Qt`la1aO{wS0R}u(h#! zX1d*fAd3l759QW^xj}v=zF5{gf%Gd-V=czQ7=WW$R*n84d#vEY@u5*;gNA!gjuoz4 zGrRgkpt;OU1(Vg%US5%Y8ipypy3J405(;Niu^$p{jaV^ZHW zQ|7`7F)Yav`TK(0NF^r>QOKmFHdU3=0`C=#ex@XK(FlsC+K05?=KD#!Si4qFV&ueOV_WqMEE#;9R4xSAH(dW|9hZq?A0% zz%)xiUYyA*(F!-)Vsh!tVmch1WJ0(=|4l(|%G6;VtYnc7+-zTa|GXgUk(*`Df;iND zlle(2$QQ>^$Wu%}`jXcct7-jcE=PJ9z+)&0+;aCrl;{?4%Wt&!9OC`X@ZE z@%B!%SP0ULA)+mXkasd=F8{b>rKJqs(IMZNELpq>bH1%9V>Q@)RI^@X!R%GS4Jyp* zNGEdsLu*^J_2(Oh95SVll+WFp>nu=vs~GH?O>)&`8UN3LGIior-( zPM7{COtk3m;Mgq+`f7mz(gfR><4pQwTO5f$Z>yDJF#e(^uft{W*Ae ze_Q|JaC`#Z9IeO~{!|VRMW>y((AJmG`d8O|Gwp3kLMTM@MofNM4Tb(|d?=(7CPN`> zeZ>2uJt623M6??8h_A3Sh&D+2Ki@pW{a9Mi^VuRg=n(|K4^?_2Hy$Xx!&aExf$yzc z58)HSuw}AtIf53>NDe#3pKf zM!%91{;+TYQ#1%4;9B@gFqT!JztDGe3zR3Z<=ZAtKv}T4686Me#N86?@5T3QDv+6m(elnv2-VLh98O_n^Zz z-OLVMzcTaqSZ?=Fb=>uxy6dQLE|Q;(ze_3~syHkDLOpN{%Me{h5|nTZXoJ7fzPn0A zEQ>>};OrA>@?-H{kUPKT)e@^r4VZ#hc56h{zsNLFYzjv5oo!P5TE?oqMi#pQ$^+9J zvMXUw7pOBZ1+{i+d4Rx2wVh@6spen`YO)%~m5wQ>SCj{qDn}~E{AXi1RPtx>CA$01 z562e+OLVE5V?teXU6P(spo0xYv|(aS{Jvr6_EPhlx>)747)k&MRwvkW_#!ijacqIv z+lm%7dz+G>)q4%;DPICzX=3%&#vzEEGqblCtd$z!;8jc#mq6>kSXri8;t-+O<+vZx zRU;6jOleU?h!-~k?4njHur;ZoA{6VP&HTw1ARNZJI1?i|LKMkQ8Ja~OshdR`dhN!t z(^gN32Kd`lJG!o7K48I+Y3EFS`mcK^3yg=C{`B`$ct3D_QuH` zG4_J?tt+|auxRxrey;HJ`0#y9*#u1-!d^tX9Q!EjoT2 zM=xHrwqNvZ={&d;io;>;gxqh<#db}_NS!KRWZkyM9gK%k*CC(uO$7|2&X}6D0fsV^ zLuW^c$XG6DX%)49FFa3v!`KKp;3~y#&=mJ%y|DF-#}rsta!jaRqEDVN&Z2Pxbia!X z+cYg%pott8lUV?flRGc#$woO= z%2j!+-W^f|n7iRe7_UVrF4I#;ylKNM&dR%=P0&U!!z9osNPW-Y_j?p28}9d@jb6qf z&Qgc(mp|31|Th# z_iCW9Y2+8P{ExDK5uc#3v&?W?3SmYm7YuN<5x>^$vEG?w>=gfz6waJ6s zzK&(>LB;rW(bQAcU^mpEa5+||shIw|z7R1w*$(L(WR!=qAmeEH{Je}r?Td|ih8Q5e zuR~6t5GstTjIMCF2?_`3;0*h(*n}D979V5%MyEMC?#XpFFmt;y1qIthJ`<`n)u7;T zC${_oG0|; z^+UdXcZM(Nnpp9YC%-bT!74M#jEtFB$0(cW@uZ9wprD*-p&8Q}VI>ChGBaj^dCVA! zEzfigd_v>3*#WX|buWE9buHkN3i3%DiwtVP14u4BPaxY(uFr$V*Na)IlI`3S9xpZ1ZW}v0&2(VAQCq(fgEE2>}Vox0C;w1Uu ztcaNjcon&@ACE)XvbR_KXdG`7pfMOssV$q32Jh80GlUZlELsWSUYQlb>9olZF1^uU zUg#$1F)Duxp_}+1JDi-*Oc*bEC+~owWC3zu` zpbs+y5{qW~S-B2-p(c=0I@Z@40{v1};#LwNQ%cwZmXW>{lD4*#oBcFkH&;2^l#c3Wq=eRxqpzKq80?cG6d&n;WKyW{oU zfoGsy%n_pUyln}0uv(lq`Dfn~h+y{AwjI|fwSxuZE57cE!8?Zmo@d!e zyzx}-{;JKYm&xXYnwv>)k#trUmYbGtVXhoO-}PX)HSP0VH+MJZ^S0UX-|1p`Rd&|B zD^AiQ_q-QOksnM0gG(MeXjg2Y^)hSOwvm7Ejk649{w<@etrdo^zwD*>p>BL`+U)b` zrhAL@zn*R=!ECIpzh}f_@N)80=$x|=9cj2JW?;4~Tn-@@!+pEt?TciCrZHn)6POR< z*u|ax_UQD+bOghVdk3^%OQg#HGh0}J5JXp4bS37E%#5T$<2byIh&=&KXq>0fLoBhfE7qmFOm)cD z5(*GzwK_xZQG?Ni`bKqZbOBfU-91)g`9D7;w_h$g)=o3GH zS=f{}tNQ+g(z{t=PmQ*Qy+!(8tnh|2EILnXCXlNX4Q@_BU4A+`w*_~^&h=f%J%0)N z)o8iGn|F66;_mtwo{ihTM)nlFh2_PCg`>w*3_qRZZoJF&AX^}QWk;zU8&%RBlq^`^*NFQI_Y`%_nQfgG5AHmLTxRjDh?H0rWhs_TfOR>yBC2e z9JLPhCZ8_Q-L_}}5DoGVJtVmqZ$bg4K)D+&IsD-Eu)otHj^=X-Z%a6b6w*8xnmYxC z<_>bAcUu;QM)_(ue&$bnoz;R)KwxK%#f5s5WCLy{`YDE)@mvtQ%WbmRo@6Y@U!ZVT z$I7@Htc4y|`TrXp|6So}pBTiF+EUBXO?~G)u(nThuPTaG1xdLKyIM3Cf&!@wk`dyTS!WjNB4A+uz~;a=K2 zJjlOAKf7wWE{Uw**66a7)UPJI;4p8t#<{qg;Ko?-_$~ux!Yb5}3?rIAmlmbTfs-i6 zD;9}+8cYzoG$6|?t_2}5uP`kL0Sm>mAkb4`Sr8&0hGVJI3++ai9H-)th7S4bj54FN zrVac`GaQ&>0Az8<rzhfwnsTOS!foB%U)m^zEUQ}0b zDrj1b;5}*ukoLYsj=GMNycb2r8@b)+*hKF&p%3Ux2W3fl(ELwwk_@JwBA9dqk%Ayf z9F2mgG+tq zQd)Mz&Tw-*QKdQx&!H(|1I73Dkaq}BD;k;r=s@?Bx|m%(`V z96Tyo(V1@P;r^8JI#;%ouXS^OG8Ix1K<)bq@;y1!@&a|E8PZeACx*}qdygR(`a2{r zmu_%{h4cK5Y;PGB_kQ-jB(YD@BLP(D!5N1(6fLP7*`QHbj4uH@%;(A){%pSEr65NW zO}HeSSX8U)7*PXE1v91GmKDy_Ba8}%p9zu`D|h2Fk3`AhO9YHkWy zO&Nqo+l+-eO134(%(BpRlpvJZ9v-8AhhP1Zrj#Aj$J<$_ls(i57HkVEK(P6IJM(5n zJJaEKdpH>Nr^A&2TZKBPBX~04u+;}?3&y1;GYb8$==i?{@)v$Bw{1q;=Bh9zR4gn! zEu%Pj*lXJl3Vk7;FUXoloxOYHzAJZ^D@Nfre7?3EVi*1mlb0s%(=@3gy*C}SJNB7R z!-lDXsB{3w{j;+6$IJ|YKDRzxshh5-TTP&aUN*yRP#yBThjQd_JOb~7R`>13*O zGGJUY;d}B+PhMW?xIL&O0QT zrQ>o5Nff$_#@?1_O%N#MQx$9_L7-;W$ZgsSZN(G3he`#1&gCoA1*&X{!&b;Kp5Y~N z*a}e?4qI6+(P67Au-62UnVb(KIWhICvh^;Ou=w5#8H7$D$S5q}$h@{VFw#qNWD0>e zj!aoF*^#MhP~nkwC*F}kf2IV>cdEa0Z{^46Pb?+4hyPT5duS?v{tGYO*)28VPy*zH zt5~A*rr~OxgTJKCdzOyN#d#}q8Fd-ac`KhvcHXmVNF|zgK0y|Cs(b>9K3eLhW^Iz! zMX$A`6SWh96qVaNNXc7fJw2aLGu$Ypmn1wQVU7ap-~SfE3TvhKOlkWt9PUox)@bPLvIU561|e# zeri5bu7D7V_{w>PIDiT>qXCRa6^JBW3}@9eJQTTO_l16QD^=zv{I!4AK0)m(o14ij zZ_8&F5k}b*TTF^Z2+X`rG>Gi{8zD#ze`o-btx}R7x_Nb!GjhJkylABY!F4&%9tb^< zG)C^%>p%z$g3sBj5_;q;!&sJ*E(E?+R>!Cah+)U8k}G|xQh;w0vl?^wRfQZmTFL_s zp4EtrJjiR30uC_Ih608I9Ug^bNJDz)G|BKod)FNk6OpulL(pNAWyF9(e5lHPB#4o| z!lel~1Px#Vk~1H|eu!H^IFH+HyfLiFFu;`) ze6mUh4Sep2%xD^U=TtLNAm(04KbX$w-_mc(JeVr zL?ym%a+$*_SSD-JuzPWA{@v4&$vM=DSUsC_Dbcz(X`^8eD?kPC2(ZjKrW_+~7duUi z+vUG!X~S^as)sM+VQXKX@ZGZaHTp z_*Hl6@x5w~;JtTfK-gL{eg2z`dc2BGTV&rvZ&HpIg)Uk)hFina)810`zf*E-vnnxP zJrrjgqlX{Pi-vC7J=^QTf69Eu3)|o)pS;W8Cf{|^FB*<_Yx6N-=!X7ea|(|uc6pop z;n=oJ|vJ(97_Vz9H1$>ub)#GNwq2Qb=nVGY-Q#kyxn&KmqjoupXiXAv&{;` z)0l=OC(xzpiX#DjO%+N_UlQd8)`u~#@aD(gW~map@Nby>YyLiuRQt?kSBW9EF?&w~ zZV0cLmnas?(A(>glti&u!FUIQM6nCTF8O51KqNi{5k!gXQ&xIdESH=fmMge{fi^b$ zeTOs}lRUNVl9M3v?6gHy5vx3s75&vGwY(v^-`q-$_nr~b{l;K_^bJxm3+sBLJkWYi z(sicJTh+A_`5Ie)n%z80sR@rE`v(1KfAh@lc)Jggb#L*+GoN$wvSTvALGp#9Yq{gi zb(6sINGN<%eN24~**ArDn3Hq6li?v`rGnG z8pw{foo#2{@D$*<6Vse#<7aell4O=t!L}`2j1UzONk06~7RAda*%*XMGXjiT67v1SQ(0V}05xHB) zzY~egI(@+A`sH|HDy^%hJiFwx9deK7_S85klYJ6&8iO-i<ga3y=(R|yIB?0JZ*w?KSBS6_7KW2w@(+oM<4L3V;R6iDElV_*TI_JDl^f;24 ziC5XC!vX!$f#sKJT&MP1;C|VjnWnRPL*s`I2^c>d#JO2IWdsoz%6JDQ2aF$%r8Fvr zK=JppZ1T48jolI#+^n9QkibK+LCS8|`K83yw>W6SA_^3yWsWzSKfvb>IK;{;w!n7p zY}6o*D@SS68drSbo--+26M+f~igYNnHi9*52GdZc70V#lqT^M(&GpHfkpfc^Wl11_AHBe34K2C&$Jb z9bxEY*#n!`pvur64ch{7hwPP1Da2Z=N&eE1C%vK0lcKRerVfDxzJapIXXnp{N+z5Y zRg)KxyL!(HshR}SrS6t~kycVV1)E>GAl&?Ur5;pLDf1B&>M-F8_LyKr+>D;z1@d-!2Lml%9Gzi zN?(dSiLe?a-5{0)R+T8(PqCs=!ZAYXk0M4p%%(L0YFyURa2hQ{mqyqdHA}yvJKsOT z_6Z<6Je1C;(w>~$Kb|&Is5iuD8|F2(;Hfgfq>9DZkH_gmfxY6#9Gpa@fw(NIZL73J zPiPF*IY9QU?xiS;YkQqN1zN~TQ6KWuc#b)(_S?umlh+}ItU8r4WQ}=nt7XQ~s?vfo zqgiP+R282_3{~|tb3@fXLJI37RXkgYa#1xI37qgw5NwnRspbw_=uVvytnf}y+D>>U zmPZTk2$E~o*k~Is&2k&^uE~+>;u-e~{qd;3b9!j-?Dpz$+>FHsv#ysGLPswe@(9Gn zx8^|e&hot?VR$!re(%oAgrS8tW^L3|9(lM+HfZSNcx@PcRviFC2wfqRJL~q^}O@--9g#Lp|^B4 zwUy~!9*0C;(IL;l)|k4=BJIW%7y4W0hrI)9`vk7GDCWoj)kZj;h8E<9I^-^FZovD5 zv%{6iv_GCoj$E}_!TvWV)QC-bi=?x2HrtZ%jwi)A#lcXup>2m`nU14`N7sTC+2Mb}ivas(Hqa9FXA%n>ysB9r8SE30vT=oML0$^+wHru)733VWpf(cgD5XP@XsnlSSyR0Ub$*gF)1k<3wU8$~HQE~X zAYk4amgs`X+1Z-1(h%+xz<;Krb6ap1hlpRY&WjW!Z}j)QioPFejipR2V~-wFYp2%E zbp$g1=;tvVpNCQK7byOrzeO!vhW41PD3c@fD6(ohUE zMrY0p$HR6@03R&K5420MEcDyU77fsWGkAPim+Z7gv$39Lg8;OA*k$WS!$mi3lo`yA zU^Rri;)zAvC#U;cecsBid@cg3^@>_;l#H!97mkJ*%irjb=eH(_fSapYh_%gQdV zD+;7u8%qATO;*C2qpFd4c$bm#SZeWZW4e3pcwJI#QEnr63-KF*`dIW7!XOTbZB-Y_ z_Erb+;D#`u3l#ZUo1zLhhUzV8%O_UO*@cDyPy7gI3i@l$EjI3X?Fk(PQ?+--EU74k zT7$(0F!`%(S_}r&nU9d4Yt1_wv{DNdL*R_PJX!I7PkiE%lPhD%= zfKAT50Bjg)-JXJrhAzS3Xuvbtyvc6 zEDr?w>`wCB-m{U=XUg{*$TJ&2lf1Y~ZpStItyquiU8MB_E*7rIikq%X;4x#5^X+A zQBi2*8wEyUR&v3BGHY7Kg*mCBx>KuVY=;A0blLpR+RA}1b#Z;ShAKly`x~VT-#B3; ztY%5AkYw{YYrDnKQDl%`&FE*+bh1qZS%}^lK&Hif%mR5^H>XLXnE^F&tPPKn(|yKF zY=!re&bSF3yVMd}kY8WK!173a)SzLrNHNr_3p0NuYA{9-VyHA$=rGh+o0(unShqRh z>gjdkMwF-PtZ-iC&;jbvBG!kiOB?4eZOu@m(C;G8>pdqkg|3@U+3}HnFKlww68V^a zG*crlIvDSsgNK$Y`r%YKoNHuS9$t5!*gW*~@^DyAov(Uxe=?OE^Mhi3wII)wAC3E# zz5@=?Fzn$_^tv%yfbK!J%#h$SX3LD=Cu@$7^`Lc+c7ryjXtkS@UFo?}JTidDaj1n-!m9A$o zN1ChhDRIh`0Gxw)+4=hxsH6O)fQ9jSurNvH0ulhya>9-1WSo4dZQK;{c}FopS&}E! zDIkm$2to}V6f)5;NRhb9G0$miE&-#vhb#5C$~-Pe{fm$o^0OTbPbE7@X_0l}Lpn<@1dJ%<%ov?0wqS|KW25IT6hmRr+gZZX4E$ zX1W0T9y{7~NoW*p^BGSG+h%jJrjzoHVbGwa81J5LMryo@4sM6582-uQf3n z<|Ebl-MVvOIG)}y9B-eW_G!Db-ZhQCDxVx8@5<6E=U_cfru&HB7h}rITM!@a?L7r` zMK)mCJN4LoJlv9@bC)xvaGo|jXlPYTPnDh-P43y^t%9S`tIlO#bL+Bq%9=+OQLNC}rc|om4fAE8w(B%hgef`Tt_(gP2 zDX6&*zX`Cdn79Lmeuh83JPlh)4X_U3v6OF>?prW{?t39#mgxKEo8rv9&LC#CIC{@+ z?Wv`Iu>KQn?cZFGo3n|7!>d~~`N=y9B!-jsh9UwTY!I%=bUYdi2b+^E=q5>f%$>fE zlWN8~2GdrMcf`~_!f2C>d;<6Ck6nAT=m|)^{*C&7m{1O;7tamZYkYf-upkg^gjW^< z$vo_4Vgz;XStPgS$e(t=z5N^2d2i6;F7vf60xT28xd^aK*n)6j#y?$dfI)8j8H3mQW!d z>SW$_WJLkxTA619jgkbV1XtQf+@NV9A*kj+Uf&_l$Wgsol)%}EnOk|mFMG(0;VQXF znk^=TGLV+%BIQ8K2~-I;QRmLu5>HTFAChn3R4@!ouu&20BL5TvBw--||Kc_+qltZk zOmt!UCCbi&GEE8{3Fd_4=jb$w2O6ewMwVMEkPz&;YO07<3@8L`&&`l)Gu`e#ki!9~ec@Xbydv+8E_RttknS1dAQ}52 z(1&IjHM)cBk$SI=j)5ASGTdr%1rE3l8S`++m6>DQ8(9!hLR^+L)gli^`ybAMIZIN7^p{ily&|=?Iyobsz7kr07JS4U9oJHVBV=qsB7CG-sN-n9G%_SA;2qs~32u!)m*`w$kn`ofnzszQ>4yVT;GI{bZ0K~!RDiui>dH^h7VTVtmsw&K!3<%1 z^_U$*C7A+Q$;ua&@D~;=h!kgGRB5=}eQN^(+9Ox5Ud14wEK$qgpDqaz5=S2D#JXES zUM(tM4MpK9i{06FMLT_u_!K?KDK0(~4jUOcRUWy%LTmRy4d}4RJy&|94Qwcm2n@8f(s<$=~ir~6x z#=I0s2zqEHh>4CC^k}!Pv^YaW6_fuG9rWmQ$)G2_{t+ME2pWvG5GlM7A7KX&YlifH zy?Ka>R9blRnIbm45#NU&s-#6~e8z8mzR4rZ+Q8RVF7mxbL@oqgWJJakVuvCTkh%b@ z6dSsUj~gX$_FlhWyw`htQT=v>I;4cqO%RwHx`~B=&9R}I{I{%oIQOV}RxW}KpN+mc z3JT~CHEv=#q%Ki8>{zXzje^4WCUl1UXmorkHVH>WH<@!%j8A1`Y$^yaYD0G60eb02 zhTDQ5?%&g+QLz+bZmdAae_gfg7RNlJ1C|6Tc~x{EZiqT(o?%5g0 z4}uosle3hkSHO_Sf@_l@pd-jE33|yaSR#cW6f>u?P=CEfemUl?KJ(Fnjw-%vV1G=| zDqou@ec=dT)jOguS2N^`UsYIDthr#%IV*P5hCk8*3Mwo-4Lfpn`J@Gc#)H)y2whzt z4?5?_C(4JalXhK1YNy{NKNEdRluWI+ApSx_Dji!FF(0nf$@%t5UUb=jDncV4_cg-8QF=Z1p69~DR(y)yu7&4$@WyoR1kPIQP z+wF*jO+`CER8?Hp0^!GT*x*%b`$emi&Vx%DEXDzWWi3+Np-==>z{t864pxBv$qsog zy~J2?sA1F@v#qx37uq3}h>YcemR1#mX*+H+8^%UR>{dB*gQm20Ov{gNz%}-gD_5_e z0EDHm`Vp;}VVp(dwui)2PxG$8!*NxPTs$N%^Toq~1jw3-K|bzmOO2suBFbf6zu}Y@ z>KQRkR7`_Lu_TnMey=WgKp;=-M%RU%rij5q(@17w0<7fqF3gWXVPrR7+ireX>) z@P`NF6Il_kA}9>pQQZO%Ia)qHJv>y1GkOyU5WV3*pt?cxXm`2WaQ+f^s}L%TtBkI2 zxXF$uN6Zu=xpVV#PXyo^gq-|Fr!iOU@S!>rn7Lh>kw!LoN19+Em%t)MFgad6jFymwlhyt(k4UI{M1%n@`x$|qswSi zT1Xy=53)nbb(<UtiWwB;Y=3*Z4-c*TQF&MB z%9nAg7xatZ5@Ru7MwyW@6GRzhlOJ7?9$}KVX2%Om02^T?2J>i8iDnEXvozo3pP>6m z9E%K+zyr$Yhdd`I_m8K|vPDcnjO1UwDoZ1ie;7x@xRQJMvRiL8XXr|eI>SN&?;|rV z+Cqm7u^4A5ywFY1V|0%#gl^)4*`b@hQf_APBUz!3q=nc0 zGA>#sRn7^41aZbQb{cM}vkudBZDt50=)(+w#G;uYkX(lq4Sl9dGE9q+$S-9jZY2>i zcEA>}jP$J_&nRGO1YvoS3cD-6Fwsik$4-k(6Mv ze3|LK_DfV2&!nQ@(*U{Xb#EFB%4fQeRT2O-OM0`-T705pcFkH&;2<1lE{X}= zwvU8-8I9Z9yMx}ITefz0$LqU;(mlB&<_J-F-nN9>Yx43;3G_!^r}xDepqWLEDuc&$ z+B|BlXj_OhIDt#7$ya)+{Oc~y4Sy(>=ABlirfNs%8+1A|K*JLuMB*lx2rYKIK+55B&o!OXv9w6(Ru!1Y{; zEkd;1)am(3H+iTVUvDzoYP#v(BK>cPYyCa%WD`z>-@&?c#>VF{q51|jp+zR1@{hUzj{W61%HE?Ev!HYqAM)A67xoOsczju@yfI#w&)0% zrTy}6#|f|2VgBYev|GN&&D5nfM=%rcs?Dm~+UA7Dev{rJ>8vi07k9Bj%6iYeE9)OT zbG?GKL02*C0}`j-x-2@pCHY!H0m7_SXXrg@FuG9RsE&;;;A%gT2dZV{^5FglJ6Jl< zNTX&HaYS;eUbGc`;s-DbP4^{R$@>#Z?`DZTHQE~X7U_Sn!W+)8=sc~NK(0~&h%1PC zPC?0ZIy$!n7PfPJSMZm-8ZB3N^X|?>++82TvvK>^$eyCNu)MgiaP*jZu`aB(AF{hh z>ReAgqBz0+7PUFia&KlmlM191D`Q*08FJJdd1pb~V~`v!Y~%j({qb;cY;@+#a6B}p zd9|db!#geZ{G++JJ`$u)ISB=-&6WtqL-_21(l8I?z!W2!W&)UB6tST4z}%}$g28D zKlX8e?3UPT!V&vrOGZ1N#uqZE_Es-CS~-bwE{@4rZ5Q10GnxSO2;1j3?k?COMByOV zXU9@fett>T>e!G!#n)>FN@!D1&mem`V*~}r5f|(zOpeKKCN#$27Xgc3uB8-Arjfg({@xBec}Fef$p|N3xH^lf9N530qR0A$?FSpJXNx{hy5J`jt1+3^|^$%C7eSF zX&wyCodQF12RYHZEek`Vd^H?D^CunRo?q^`%E%82?98#aP>-%)z|BNI#V|9T%UO51 zO*Y#v@dfz{6z=L+8Jy$tG$ji<^$m(kssrjv$K$^%(h*=%f>}%&3!c^4)%{RmtReI!C~HPjdMu`fl7g>t$2Kw0W)D0 z>PWh{CcdT3Q?N@~L0++#j#`lH(yHvUxE6%Kyu!2~1S}NKfGELEV2A`MwwAcLlAzY&AkShov&m-&>MS|xRqpxiCk0DrgQ*?Tu>ayd_>J|Z#wju zMRIjlZipAv6`Ts1RwH;1j!$tgVlF2^-ispRjofZ@Y$9JL)53*5U$ey4nDDfb-fN4z z4J8m!5Ge?v#L*~-N<-1L7MzJc7fyZzCDVq6LV-Cn2wZL;P!JMF41I!Nm7{IH7e?x*3{!z)2K*hOMJeA!! zvX~Rvr87+A0Yza+nT9C?@s3KibEFn^IBx-zNaYNJv|!ozS>S_(LFD%e&v-AnbyZp` zO3S4J)?I2wz=YEJDSCaihu8GK!Ojy|(?J z&=>Og!ZU3)V2|8)8>R}P z(g7Uz&&t|AGZ(~JpMXBMK3u6q)@!sms6&s{^Fz~kYzDUWN_%?ZU3+tkF%JdzfYFymdBeJ!~< z?~r7c4jGcTYO{_M@NBq3(n1o2uC_xG@GB|)HO*N~1mE$Iqv#S^@TN(Fz;(1)q0I*c+M_gwtq_Ibu$AQ!9k#jxdre?T*;p(T`9P8rQ@<*cdsxEadoyGZ zI)xyku&B*;y++w7K_tzQDFos;GG)PJN2ab(^_;Ybi9ZUU{h1Oh->Lr2y_FxMKe3eH z9{yAL?Z(QD;q>Xh@WKTR|4!MzI26eNubd=0ZyK)FIrvNJyl3f<&O5?Ex0Fw)`D4^& zvi_jt#mV_-34q{*h4g&F!^w&( zslX3NVEXDEbbUvUc7tQ1r^Bc+O0+E;N%Tr``>FX%xdMbDS#uj96!DewTHV}FsX^h;j-*3Pm_?*2ep-0X#jAg0uxZOxY z*zupp-us5bbDM+FHgx|4^e`wb;1F~ejXGk$AwE=-;6Z-wh?)9eLoDbfpLXqrEh5@de!*9`AGLBOzeq7(X z^lO%K2ZinB>%xD^U=TtLNAm(04KbX$H@P#v;f2vHIZ{L=zHV}v!zx%NYtyj% z7}bH=W)8I?R?p^KO0+Ic+GyA#&fXc#(2>bGrW_+~7duUi+vUG!X~W2vL~xS4JgF=t zzJ7c>1cdU~ZEbakrQO;>2dtc9scGq5S^p4kn-UV5O-{rp>zn*TE?xF;cV{@++#20C z41Pj5iyv{8r+yTRI`4DqP=5@@E$55`zv@mszE|xLy!Q?b2wSUqZ7r@E%|<Oio0C}TlIWp0;}|{sa9%WY+wR$37yeV`GhSF`Kl$Wc z{xZER@gaePdM!z)uYsoMzJ5+&B-N(G z*J(d&v6Y#F@OIw`To%Ese4;~c&NeF$Ph%REoIsbVD~<%jsB9W|L0D9JB{__Fg*QL` zHcM5^KL0iU>UE^rXYM`ZaET$dF?&yY2qMlnFHtO(p`vmmB~dI^Fy6t``m`mt;J$mM z;YtNjBKwq;9u~_br-$VVH89Y|3=zv|S2XhX9nxq_a-{B(lOXc!w5$BwI^-LW$|G6P zUwu+ib2da9oLkB9-ZMhl;1~>%zEiq-So@oU@+cS=Kx>Tk;%jT`GvA-J86c6U~) zQKBoi^)C*`Cw9l%E3&x~L%fzvX6c)u&%osqXji1ZMBh{!-jx^R#RdcLNTzqo*-RY76Uy$tc-NKj1C;jvH z5;n&3(T1&=d!jVH0Rrwn7O_$C2-5jLCB#t8d>o(uc-7keJ>A}7ud}CE?L4@o7H7JS z6jc#2=CHv6lius}0h|3V$6v5m1YN5VIC<$xk z!m(cCN3TUv#equ(eVfB^x~3Q{IiqZ;z`x`l5*5djM)AW37>rwEce^ zaAig3oEM57M=~?m55a(v;#E&2X#t+9*8WlsJ z`1@HldE5BLZfQnv)1HSz0uRLoDZ5?g$SdFA%b*R5XjPb&Io@pk0G~VH5G$|P0^7Z_ zQG+vfgTkqEGW{U(Ao&ruo+B4nN}==^w_%Vc|xQA@By-KbuUe`TuZNy zEI*FVA2E#>#mbH5f58 z{*-@mY^>1{hF+FEuz3xt956XHmhc&e9Q#~viTtG@Px_QPPm0C@nK}d(_y)?xvOt`j z)YG$~6!HRcSMPZtrI6sSs(Ybdti~?+EyLT9Qhij>sKV-)w6uOljbNU<6)~?Wsg&V* zeo7K6)hGC9G*_*r`oyOZQ+@iH@@|?X?1P|cod@WR#LQF5)mW#0w!gjIhk3yL zfmC_F0{$LS`cmvkgw-hN2C*!#szk|tiWQ9#juBFS6fxRi)}j$mago`}AuY<~6qBvU?;6X9HEJA$z^;6F z;5GJRh}T$E=Rj`6i%?t!SCKDTFe&v;db!!S*TrY@E~NVdQ5)tQ7)<`BY_j%34)DMA=TVr3*D(xf)(BgO4|wV#PVq2om`Q<{xf4t zy@N=}A0nmlB{?z%YEL*(>g0lCqi{kbWW)0YwaFkKK?;HdnIegV90Z9EAO=Bl%^DkR zkxvVYL*6wxvSB>qexW}e^>#UHuuqXRu6Z{1`VAYuMMNm3goBPy=gEgPsS!*X14=5n8Eqe z6!g0}+8K;a4<|X*s@c_R$$|r$;%Y>la_EZlBrn(Gpn$8rO6U~QiO;1^L){#_hB;!PjHsWj zx8;e-7Ef=v*%f1cL9tzcmGJ8`i{u5^(tGUUPJerJdSg18p1!wtK>M|9$6VqEc@RV5 z)0&<6vulZnsOM-R1!Ti6Y8~>@4tXB7gst}<7;T^5mc5)VA2Vpn#*p`R@q_7_54|hv zAIf%hxoqzsFNOHJr1tGyl93iPtLd4o-QDqEbfLbR9vfX~%fg`h%38iIw^RDv7A~lb z#5zjpP$U}b}f{8kHj!c(KIVQ-QCw?r3A&d%19)&GKs26?9dAT%AF z+ZsYY>g&6Lx#W%hzE{!rLvLfLqsP?CtG$UKWd705V>&($qu?*l)PH}ATDT1DFUEMrY0p$HR6@03R&K5420MEcDx(e>ZTljz{wy z^0F@3X^m!MJ_;l%UVe z6;)DY{~Fm-^cI#E7Zw(dh8fG>=#b~PCW(NXs~Ul{&0~BA%y}ycr+(A<$8E9_-W<^f zdiRv!T}H}dsl~gE>F&AXbxE;Bx%IMC1{@H-A*hc8&e??uX$;`B3!_3?(X{6l8~42SgpPu#+B;*GRFpi|ar0i&LB86i z#b8jK`3U*B*1WSZP6XT)w-PQWx(62cg{)>v{DNdL*R_PJX!I7PkiE%lPhD%=fMmKZ z=UxCd47F}g!9_!ujK@ZyhZ&yQA#W?lv+}iKd)VJ;5t;TcS?}XG*AyZ>sI-AVrQJ!M z+j}+=Doy!*I3n%EU2;3-`3N-HiPV6WZo_~?%Xm6wJ|M2f-R1clhdtF_ke4H$f{s4J zK49;xnt*(wBCHe5MlG#e~`I*6Uwuf#{MM|3<0VoSJ(6u1GFTU76<351J7B62|N`;z%9s48jxQt zSPoP_#}Qxi(HV!S5NA(*6Zvx3r~qU{U~(=8Et1uJfH^p+JzWe>2I}3!$**QT^>0Qn z3dRdn3^@y1QAY=I%U`-8uSL$TYPc@j{oDj$n8AAy7#N6p1`SLXYKX${T&UJC8q4*u zZgaxbQ|rc!X#IO+v%?m#K3rYeIDc-4i{+t6q2EQG*LzN83SBpyl1=n`VFR<4$jAJn znHqV~!Fcx^%vV}rsD`S*A#ga?$h17X?mn@3_^IXLu$($y_2&L$>Ng^Sr{q@)@=W>B zxNqq@;1CVN9u7sXD+u)Og+TWp-!wPpA<*sXtvKy4!np1SugGMsq)b8_(|Z~HN`rY88ybDv<v9M+79_%DsM69 z?*?hL5~*;ce7-S&8KPejm@XY&WVw!!&lTiEG-FihubsJVSSy<8a$(e#gko4G!O}LL z@szM_HYaueb(ow16Wqc7}PPBi` zrpA)Dq$rAoQJ^S%Ozd}fHaHn>omt`fbp$~j>a|=T15p&Uf)bZjlD> zfNSYxO@ui?!cS~Ws(B1~)ao@bq=wDDE#Qa!@aHJKq=x5y7#UK%j#TG&>&}JYczVZh zynTM!r|r&q*EIgBd~%4qD@(7OgY~!-bB$qLB=3tcW#%o25BK(-;?~1E_1Jwp+>)Vl zmoueUb{#T3Ul>b751KuPIz{3B344mKxS&`px|m^*zRC)M0cFW6BQ z?4)w#BlDv5c*&Mv9h^RRFPJ&Y>4MLRCJD6LLM_M6Nm;{Tahq|PK ze0Y(3ADT%32I@uu!`mUM96#uY=567!)5fBgt8suiHd%%+s+NT3vPw)*&#NdwLiolU zy%R+hgv(N%h8fQ?<<{eTnf-7lFLZ9?NpiW55b!LA-w4fFV)CI*=50q-6i}{}c{b1} zN$Rub&|9FhmPcOSAB3dBHDx$c*7CrPk@igiyYbs5&=ujap8iO1OzS zch;78g6jH^d<&<7VPJ%=1gjVMrx+lKW8WYXU0S0?*?CZ=iSK;-{8oZF$;z+RnA1iw z6(PTRxXQ2E@8yeJ)Vs8_8lLlJXD8CB95yTdLd~z@*ocS+YG&FR>~-c<>zpb}p3_T! z$g6UGu^UOGtyg5M((qQ3WG~OsMQYe9NSD0L9c(0Cm&oF32ikFYAXVfSW13So>;j|3 z=+tbmDU^eWz&DslX2+X5+U=jC>uwD(6OJThYW|in7EnC$!m<$MYItZ|!O}Kauc~CNzTq_+8Fv0Cknk%~=GFH1_i3 zXOZ*rBtRIQJ;kiOEz}WA;=l<^xy;$4=pBnCwmHPWsfmq61a4m_$k9B_X~wc;cDQx!H-zB|lM+ z?~2=?m@=VxOF!O(VrN2&l#BG8SEOv{bjehJoif@co>E2sgj4pDomi0&lo=yITVf#V zSjtT1{%ew9!PS6S@NwksmJrip<^j*1F6Oj$P8)jZM~2&iAnxDOqfxOGV{WWK$bVh6?6y@Bq@ZX~ zDtT3OAZ~~{XP#k2Is&l;?g)Y)Ljvsrt`8t0#+PO(Pp^O>kpN}z@S2N^`UsYIDthr#%IV*P5hCk8* z3Mwo-b(RF!?VQ#HX5Xm_tB?hutLx)I=N$P&`B213|4j5PQ8KmOg7^#dtP#vXYzauv zv1#M2IJvZgC$EWidEC=A?~GVwWvKs0@>-**{zc-MBa1>PQ*D_MsHKN$vP9B!# zd|g=p)i+ zhvSQZwKW9guym?%7SzOM>Gs0=88MH2-!OD@i5W{>paiKY`%D1uRj1bs07a4!;mE?W zuoWe(W?>;s9jT@^_CS1`nT18*t4qt%n)E==yhtgg>fw-{$kn*x(RCsWF-$2@rAKSh zq^ILujH}$1p3E05o+ACRnK}8ggThr86VjyT5koiSr`6q*4LxZinQ1GC!~z>_Dji!F zF(2-nTO=>KPO%nco+%_B_cgct`F1^H98&8H&XUw+Rs$Xb_R3b7quPO%9cHCw*jE#`kt#alD zO=<0z+#!7!N&{+U7-!MA?IAJM)4VJ2a9HI?zn4G0QA*uQu3Wu>LN;X6Px3NfJRC@X ztf?5}8JsJh6FD?9jj{;gly>f^4Dats+S?Z);wm;$SEJyfk@~IBVP-um|D9@Z> z!2!pW_#xhOQGy+od@I@n-g)SFaT+>yiiQ9O{UMfarQ6b6P)FMon?zs?h0(K z4FkrIcV-!TMRXdW4Gs$_zyeWgWm;5^bDq$Ma}JPwt9xmjb1l>g3p!pGhYXMjeqA(G zdJRrcot5SRPZ|ETL~S+|Q*y(vW1q;1fE7Vu;Ew7RfXLDE`RU=ILY&c?K!E7wPoTO% z^JsUu+j)Wtp~ASz=n99M?09m-Od*mv%BbK|kHW}rbeeP3Zm2VXncI~qD1b5}pNZ6~ z7l=V;8{}EN+x>&iyyvgdO}|r3%Js@8Z3diaky$NROCH$+zPMl&AlsRzDQT0TYJO@f zFL^{2fzf3&DlH_B#0RSY6_KS}pt~t?D*0&Etyt`8od>6lIMBGxWrftm{DeA%mh(J*~ADe-?xum#@lr%j6%%(J-#$UcT(s8$v=?B2dbN zD}hSiM`m2Kg$^5HK@1jFksziQaWO-(Oj9HB!&wnC6Y$ClvmcK`DX6zs{Afg!82pC3 zL{%Dt!Ib*c!Uu3?hH&D+L@OcOE3-m4oi-W5rN4B`3*7`gM)%l4=q5gx9lGf&!R4zq z)M=sHN3ud6Nei$0Wn8pOs+cH1ws`q!5d{H2Aw(Tj$>df4M=p1o^VvL;|fXZdcC zf;vS`_U=eZL0P`cbYGidWE0jec07}cf=>hFqSw7?FesntLRLut)GX=EHf!+wFo_kT0WgdwX}#+jGm-?(TSfcTl<~cf=eaD$gCPnA>>}=#RWk?~5@& zGm9Km29N8sZ6wfK_1eK=@)h4?sbIl?S(s%nkq+PCkz zxIA|KJ6$ZV%FeoX#YuYPo`E$f@`Gt$aLHo_-TGwD*Fdt-_l^_-!j_TT4CV& z%U+5f>c-cb%(j|ty0=LGTjE-O&pX+KQ*n*8lWr1*6prZ3z)d~^vu#NwZ(sC`{tMpg zv5PzX?a}Ft=?F#z_YP>kdPao>e}kDVtUw5&D=fMa^G0^5mLk!1pdpIb4x2@wU;gbl z;ng}!<=j5My>jugH&ZX}Vuh6To_kl;KX~SP1#5$@V%P`duvneBi6{A5LIJ|8R%hrv zYB0J`->8m_F5qfEk_W0~WNsXG9W(i02TKPUY1E7&jz~_`i?*UqGl{O?94mQ$Lh0Qs zv8P5`!`>qOFIITN85W(VH515HiVjyk1qIRR=-k$DLI)mm@eKiZyc#W6c=PVgMBH5; z!?SVw*T|j%##1dWEF3+iV)*H3@2fVePMez(dheg~7D;Dyfo6A+)VZE~L~(-sEoyV3 z<=)JCCKX62R>roN6MJVt++&a&E^Ooe^ZoH~aBOtu%y2w3r+MwjwuNrYGmLi^;9$4jCX*YOH(Mil?m{r)RuRqLAbkRsW_S-;d1d$?5)9e_Thj z<%zpfm)9#Fd}x}SQW477YMwPxaoq9WG+II?g>WHdh!v8tq^+3sC7I#uZdHrn~2k*9mWQi4(;>FY{=$gKaw~vZ{X4k9`~ zX?!7rYH#(Tqm`2==i-?BoD0K*5P%+G`~1e;1zUtD90dF9SV}5@t3uax;NPF(>oo%< zv?-`(kUgC-f>PhBI}nxb7l7n96B=Xii-5&1*HQ|mQw&l^mnn+$X*;T$K`D6=n&i_3 zy4w~l0HQ(up@$@0hA*=~USE*osgk`t?C%(GG+K={?G8h4V&roPZ%a6b6w(w7jYgT} zdcNE&OLvy<6hNCh$cf%X;YEzeOXbo35*gcTx4hme@B$4n>0+j+$Tk-fV z17^Z1)R7D$ni!T#BXtFN#bP>YL9$D$vd`jL5CZcG(}EDNP&^9)Jxn|>3>Z~WDHd%C zK@f)JuQSSwQW}ErD{byI#LSVeWI)gxdzHAAWQd7eQ`4q&0ES#p9LRh`&Ee$;JhMoy z?#d1EqPl`pLDOmk@6o`gv?=T^Zod~r#v8fa=-5QQPNszmeZFRitub84qmyKBLkUC_ zL<)i^aWo2|(gY;D?N}A^BPf|RG!zQVp+Vqs1A&5&IAZ7%1gjK{Jb9g=IsMb?mQE#nc=!LzLUSjL ze`}OHe#ik}StkgB{9BS$W1DKk!KvJ#ScU|dk+w>5LqjA4xF8Vs)Gx!Ca+7Wv|1G6y zxHB2@F=;9Ll)54J=@)nJsANf?;#@19%I+Ll%n9w%87A_8qOc_Qof0m1M!chv?Hs8^ z9nM<-B~m%VAT8N8-o^=G4>d&~zgKw1d&#Y<(pnj>wNZZp{Tq%$XwciI{F1zhnwu$u z?&N7gn=$vW=9w)yW|f7mGpi>&MmJ43p82{upG=&QPl*p2hLP!i#ipB)hkS#uXNYOC z0tA~+wKH#Kv@;!!w}*pKe>z+lFd3oF?MSu^G={Y)gRdR@Iam!e(f&0Z)&|f3-*nig zt=Kjre06oP2o(#)jV4daC{7+0lI`_c2g&CP&$QWqJ#ycbyGzr=bK~c0%Mr8BmnQGi zG^r!~-T>{6-IMo>8>R}P(g7Uz&&t{#^H>G?-1=~(7Fn<9GRBs+Ta9~gWe^MgKk~jk zR<5hM`<|Jb8++T%mkbcB(|%x8t%xqIZfqw|u4r*?3?^;t1YZ|O0=b=d#?CbN;mnLp z%wMhov_Q*4DbJA7@)U)Z0{;|L8U#oMRJHUUpehuBXlV;oAE+;A%eVGEXTRp`_g-i0 z@B5H!t_|L2?X}lyul3t&c}DF8tpSfY4mI9j9UxCx-PhZ*>xcY5#mpbM*mbG$xThlf zsDQw)G7yOjIV|i9AR~i~1bxBiZ)?Aw?-Rzn`MrW$IK-_dn_XT2y_*oDl!gU;wvds$ z!Z+~LCk2h$Pac=(9Vk(rR2yU4uMi9UjVgf(XWmzpTTrkkV1{w!q164V`r~(&V&?>^ zKY4p)p?if!&s&Hw%?sa8PztysIf5oO7Ic4@;*j6)SAX0D8tzPaT&r$QD2`YNg{}^M z=8?3Jgc}$!@{S8O9L>5;&@bcIbtdO*jWR?yYl4uwKM-A2SM0EjT7utG1AtX`g zG8%itkVN@Z1zSlFsM$4gLlS5!p5Q%HD)@6QU!g8gWm6osLXPnarOw%s7p?~Fi4I#K z3d3P5%OyH&bp_UqQ)C0QjE&8YlMg33G4-o5xrZezzBfY#p;HJl3XA$=D$k=OIx>Yo z97m=snC!^ZHL9MIvHnXaA??qUVEIn7pAeYV_ln)9aNYMq1EoVVsOV?24H^PZ(cI&Vjfc24<(nmDm4R0f`zj>;z-gnR;u{)oN!bt5CAAXS3yHvh0d zDIl4c5Qr%UZd8jaBBP+I2yjePVe^D7IS^C0#!)x^DAgS4lZTV@(Gmc`3k&HX$s@^% zE2+Q_NMQQvopgOik9LcxXH*#_+7^x^dL_C2)O@B~0U;FemGfNe0wYodB8eBnSv3t0 zMef{rS&ZF%lfU-w+9#-eWpgu`O7{Q4tWs zj@Knu`c$P*gN_`2RUt=?mSzVWZ%k4WFwuqr23F=FEiSSOv>uZ5nnTvuUVp=1?nQ^=!_iMC;bGrW_+~7duUi+vUG!X+!Uxr3h(}J$ZFfSxS8U*mwvC<+0n^>JUqNOLypim2)gL zExoJjAL4COLPE323BO?yXF(qI54m*N!=3HnWMgymz)*iesAZ)xuqCg@qR#u=I@BM7 zamzU)!LPbgkMC7`1n<2=1H#s-URy_KbEB0HZIOKoy-CI8@Pr(%4>yOWr@f`>zf*E- zVyVlbhvJNdmiZNud@L^-x_#$tuM5AG`HYXD`hEU3`7Wt@`Ea~bn~w=YH}oePQ+Qml z!`tLL+7)&@xHL7q?vjAQ|pBtY=U&(6fh5W=F;E6E|uE4=yfw^^!U_W7Ur zSFgj>K6CFOhe`~wjoEt|a6_DNUZPkmLq+9qN}^b-V7!AtqBx~gBKVhxD-}eE?31iX z=SWC$dRP#|)5GR!vyEBgpE_i=G0EW@>n=G78)Bzj<%^F)z5%H`k`?{cCnZjPd0n)@ zxs4p_JvXEcj=>P=JEg0KwZGXssq!Y^)tKc?Hu-kawY>Q# zRy;Fq5?CG%g^#L_seMpWYw^&7IQF zqWatNImV6krx4suM?2dq)i}A8+xr)X;}bjMtrgi^i6LIgCbI;+E+4-{A4wY?krzz{ zERcr{2H;?A0A{~`=vgUPt0s@+F6Pmas^j7ndmFVxzT_Va5p-_UDo8|I6*)dKQa|NOm#jq!Z6VQc1|D2;D`fX8g(F{JZ>N{FGFY0pg+%TMOBA+K56 zyQ|w<>~(e(tDQ?r`pxpaNNm>W12+3#jz^6-{AP#T=ea#Kj>=@81f9m<%vSmIy%3B0 zbX5!WVILFL%&m^T(T`q>q>2NV4Ei>Q<8*}(OJT(j@^Sx=s5qW9iXT3}VBGSw9MXcg zN}H#ZKb#yUZ}-2gjNTlz;B=1$i5j$TAYz_*xh0HxZrqa>1D%kMpT58IMEce^aAig3oEM57N3!Qq>kogh{4$N})P4(ozuXwy>&jx2fbqjYoSUT+KY|DtKO9SG zR1AUQ?`PTMZQ~oer5V9Z>&XcTJQN$G?DjZEUik)J25neGfx@)R@n-WSeC~ilth{0i zZ1>Ja4dS?Rlt!%yi^g2x#sOMkL6HuH)<&>~&0reJv|<^AKc0DU?H$GV4y1rr;UXw6 zzj!}8Hi4iHsKV-)H2cGaxPbkgNC`ikN*S&Pr5znt zih^LJ`UHhWbJc39Pkb6N)u*qS`||S7k+Khhs^dIBZzN_8m8-E%|7?G2s}J*l`-2$Q z{6(bnrPz}Qt5MPoVp(8SiIV*kD;gynBc%Q$Vzk3-S|gywWi1V-(L!`-guMgWrDxL8 zqvR{tJ^^Hhhte5U+LMz9$J1sCHKM08R2~tpu@$Ff7>s*_jAwY)SVpW->DfJG#oRg1c%n*4W(*1#{NfnE+ACJ+A0=vbJIXH<* z194ea+g53dymx)9kd>l71w17Wcr!3ysLrR{`wVtKUiPOiva|M?81G^u)c z_hqCQNRlICp!S3lrA{tLHVUU238lcD$tRG4AVH=`A|VGs;sc05kX(~JD>`TqKPxsm z;9Zj=8^$y47y9E-fBW<>%Cj#-p}H<*T`w(!j$S_G5lD|?;C8QF4H&qE)uOX}zepY4 zOJ35uJ2Q1?p@CTy?HQ>>3&}*lYPXINLu9F+T(joU?a|50loE@%Arv34hbEIiW z7cLh!Y?ib=P7H>d(>~{f<@3u;4qYyJA(s}3?>TL>Phm3AyXqtrjj~ONYsZMSMKTi& zhk|yN30XWf8a&kN+&D)AN?z3=cVnA{=Z!WuS1$B7&pYC)%O&&HSoove9K41(Vxf$v zpRKp$iOLpF&mpdb47Qq&-RBlLldmt5mtsrr(Tm&tt1cZT{@y?988COQbl>O?c=hE69fZbySDEvP15|mayafheliHw`4D;kPG1DjT^`ZyZFI$ z&4=FA^$%scx?Hw*ke5SzT~hn*F3Cs>n$`5o=FZM|FuG9RO^=Q)v}Iw?ePu0QSK2B4 zSqm4`Mq(YMbSM&y_3|xiY9thypKC$>q=h`;snO=J2ch}qutXP3&d%19Rj?)i9|eKj zDL@KMN9Q(&6M79HZ}<1TioPFu8%rHIswOq;O$;IPkA5E0@p%{pe}Sg{`9y`h(~DkoY}1I&)??9=2No_((y1v0aK~pJwF}?%lycGa3|2v7lXp@!j z=7>JfyRQuIGEyE*E#9q9cg`KFONuSZZ3J&2enU_ni=IMk!y&P)>O$Gx>L4D%hx;Qs z@`E--6>tpIThf+Ktemq84TD4la34T_?YYIqJ+D2XqhPA`&X^??rBG|QsQ@N_zfFt5 zpgQvr@^h_uXJecQxGM)>IusW8?W|@?{DNdL*R_PJX!I7PkiE%lPhD%=fKASQxI#MA zx;+IK4Pi1K8-*Tb_zM~Gu7bQEUn{nT{p}W!Y5$S+K8|xuA<~0N8wga|-Q>l+7b2n3 zl<$Wl(jM-TJ21~jpwUjG2DEe=1|(X>(=qb_g%w8Ua#ZBi$fuyA&#({JJF6xjpQs2c zMFa}e*ke0GL6W&1d6*NwI?bu)s*nN-hMV%q{~LED*1j zIG~}+OenG|hy&m4;`(k4RfdrEH%b@2al%Mg&5|02B%6y&5Xi4)bX00O*)RmzT=u5s zSG!j9*t@zpO&ZM%sF7oBc$Az5NoHazyq|Q&O;p%?PRm}cDaao$VqkfsK5EdgS)>^1 z)rFbA5;Yj32r*O|D|8rYtWC{;6sh?LP!KBYzflU~Qj!{ouq)iXKZ^3~rLljB1Veyo z$<;MI#Q-hIrNzO`TTJbCu0x-SA>bC|CJo517AyyMe(eH%~%vvIk`$sc1@}h(B&N-N`w8BsgRe?j`aITYSd3fD@V&m}P@=#b# zov(Upe=?OE^MhhORgmY)kH&pV-vNhc81`^5dR@OA6q_W_J;*oB&3OoP`=PBkTOy0A9YB?Nx*U-|O6a)3#fly^qsCa2wt*Lsj-||! zP?d>XXwtgg!yq=_Y;C`w30==(jx<;0&2h@bfpcEdz}MtIXqq)W4X0|857vQ^^ie`3DKAl+w>TOql?yWGS?TLG{R+ zJLKo6yv3lu8>H1rq{5N%`NjZd_BLOWBay0)E&FzmCX5*3iTRE~(+UA4Y~8rhK?MzuUGi49C+uhvTjD z(>`r?*1N9pSLID1@~$ksat_vW0E-s+hPvcKF{aGC1@YnD?o&`#WV52ZQ;*%p!z~#) zcb8p9PEQ-=K+|Jqub@s*HnR3P+%OUtFczt=wZ-RwS$|3|TKm@e&m=T%xSv{~o#w-LOE-efUj)ZN3B344mKv6&`px|m^=LdC)M0_FPOH1 zyeFpi5k{M2Oc>`Pz%pSA!i5?Cbm69a?d1OPX#0T;Xfia}*y>Ln zfU#5yv{rcE@a$-t|4p&{1hQ|qxj8yFQU7o7m)paa&HUSI$gbh`fd9({N2JoB_Tjee z!LVv$SNq{EjhpfIx^WQM96%f7@QRb~a5;5t7=*`KAyk-zP=q%KnGuLiKDJ1H9?c{G z19hW-;q4GrjvsVH^R{r=X=70_wWq6Vsm){<4TRB-WSI;)$0o}-MAee;TvmxG>Ur%g zAR+vm9K91o6@<%Do`xCEGU9#{R#kbVlNUNS@+7(3M+kT#n8dHgdgX-=VdImybe z)|k^qG8G}edZ@~;^2>OAzfkk5I5r~Ufts1N2775F*vy+c zRliQTvP?UH8%doxh{g;Kmf|iV+_;jD zMOQEyJ5P=78w*iq6pm!qH~=1{(srW+hw~#Wd?W9OLCBa1ZN&JIccTXPC}b)__M|O} z;Fb_Ly$}lBWO6A1S>W->(?f!)^_?@*t^Pwf9FW>4BhKLPjJ!X(*kw9Fx@U}oWbBJT zADU&^n=HZYlGsn0$vLK?QVCf3;NVNaq9H`^klpv?$ zXfI0vrc7wwGVG=w4orb}UXik)(N5{LsnCTiu;jl15-vSuKSc`orC!$Pm^1c8l!FoW4; zmY8udnSII-ZeT%Fk|~gttbAb!e__F{EaEJTDh-#rA8&v?d*s^HYZ%y*a9(rkx6+C zVt8Ic;)5QV31XttYS8n`(Ls++mkfGhN*=l4jiAA33z5Pb@ey_av1UmB*SU3$uNSf5 zjrczNP$ex=<1?PD@J$|J)&{<|axpw65sF4%WJJakVuvCTkh%b@6dSsUj~gZM3JBc< zfw`fZSP0l08@kDV%ese)+pv~I4VM=3&FHJ6pn(2R<0h6v>JpX1j@A0vC@6eyQWIVu z9iNI#!V%F;=A0DcQyCeX3IdGUkX?9yUiy*YwjhZ6_w;B~EX9}`D-iNuS1r4(FV@Yv zL0%Ufh#R8LnP*s$jzDaIJAxp{kU+bD>jP*6e^JFU8~ml3d{GseF+D;V5?OF!m1#{{kNwTaRfj(E8Gp6JWf4Ef@ZDy%Bj zT(IYy6+3FfA87#v6&9X4&r?NCQQ2q$ud*O?b$vYOoFktoAB;HZ--*5@N~YFZ5PzYb zHG(;aEddESHf+4jJ?}S0yFBjcns>(k%`0_JT~1GwFCC~HP*6^<|JQ<}UGA|CD!u++Q`I=C z$n=gl8?GPXFH-COD!XD&NgOZvYIJSw!SloM#lYGcf^wv^vC*~{-p`16><5OSn@h}C z>H;N5P1$Dxc&|FWW&kLXj0i^-mW8b-QM0g)%p&VrV(2qNdob=1srlWi4m3?2 zsirpeKzy8;g+<`2lobc3tQ~opoLZng^CG2~s)s{*B3I*%N7sok#4x3#ZRv^h$7bf_ z%MJ=xUA{X;XoM7!#|+(+Z>qZ~Uq>U7s*Bt$bzj{`X4(oOvA{-~O2^hk%!j)?-xkS> zu2Zayq%eP<_BF$G7LT@Qda%9#@`QQgRBtlss&CzxP5gf0=}74Nn6e3)3548DY1qb7 z4H;0eGUPCY)ca{C5CXg1j#$`Kv;#y{#dR$Z*51S^8;Mx674?&Qv^eDQD~0kWoIkdM3DQe!BZIO5*g z`GtXbz%yc;sF((gVo4}hx%P*QC4oG#8(kN6nj!`dO+$q(3%o%!i+X|7F9;$E%JX|P z3Q}KM`28LQt%he1;Z<{=Y!%`^J9OnUfhU|zqQy74>Se`4I-#F7}-8lP)!gE#A z$5{^8S+*$UF7W?>u&9T;H_O;7qSFX%a9BtI7KmCqlhA%@L+kdDJ*&GZBJ}{Ed1ZfB zI$jrt43G(aT{KmC4R%AFm6l_5nu^ISNJv!hnXCv{5fldQsI&}`qvcD|!$XBQqc?#7 zF(~{CuvE}I+FkB8oWI1~DufE-Dx)hLZdTZv4EwJT$s8p(xA3S+*H7}fPIIo>&2=U) zbGtGH1yE+>Gg*GAm~XaDUeLS4Ki|xI>?*lbZRuRIP6T5KGy~4G$W*~Xr)$Lps{q-~ zJWWZP3{`CztI#02AbH6nstAlOqfu!gc_cnq1*nKD<&GjPdGyJwTd~;HaUPsDk^?uF zrmkWJojTj!+UmnY>wZ+;6}s|e-0B7WBDlm@446@7WXuFnM%ly&EZ>SY(g{9#BR<# zVS2m8k48j^!EeYAywVs9rqq@#d;oW52qzv)u;>6g-}viUA)HQ|4B@ouWX!X-%+O8H zV|0%#gl^)4*`b@hQf_AP6Ir2;q=o0zGA>#sRn7^41aZbQzWgDOpbs+y5{qVrKyn?| zb2D!U^zp33tt3Lm4%h;gk-in=83imjsbuTv1PvB5f(JpIA(503Jcwnpg9o`%{hUh= z9xPlD!8n20Oz!Cb1~<<$trz;^QGfgN(14f0!lisS!CzX)8@+tUy?Kv=bat;@%{qk@ zo#lH)F6tCH*}F3-7iIY}(|yfHTaJZ%r{kGa6nq*W7rpK+gF*THFfh+7>CHB4@%4q- zHETJ6gK)g+5XFRU+ebqF8I9XpJA>Y?=WXunjF0aOO84ZBm?QjukLS=I`8d6eStlF^ zv!~9QbDg$K`e|E;!n35`^G%it-Z{*|Jj({nji-F~O^dcL-<{hm8uqRS!_8^mzU$)3 z*!7=wvAilf>)ur->5+Q|)}+V}rh&mFj~#UDr)h`scGS+7m6$iOOSKf2wuM7f zv+YvSA<;A%gT2dZVnah2`N?PSPDI#@c;NTX&HaYS;e zUbGc`;!KV%EdO3Y>D?@`r$(E@-Xi@SE4<+hi_X)U3FIn8W13R{YM74BZ4M`FtT(S9 z`fJ(b4QRQ-TX(i6;_mt+o{ihPMs^jwh2_PCg(F8*+h5q4RM8jsNa|cqKBBeZ{wB3K z(Q3fM@ zuErHq)~7q?j@1cobcu`L8ALnSwgUnQ4ynUF4v^gvdrdgv7~lCdd?ABsZ}p<1m6Isv z;+RBm5x`5=a}WwW!uI)%y9>4mQ8)poBIB^$fD7Ge%H> z*i-ilK=S#7#u)q}VDZbfl&UyXbeW<^pSGjA=nMDP3Us$ES^z|Y{6h~(+B+{y>#YSj zmMYm>!~V7bN2Aq-9O*aT0LE`6ye;7zQb|#6_0ObVJ56X9Z470#J994O->i^+QoF# zf@GIgWuL{hAOz+WrUfBjp?DSqdMYdnLUc)0REkBLLIB}a)sewr-^kx(lo_Qo1mRcO z+-peqLB5v(L2v9;;#QI&CUQ+po6-RoazSw*^AR>kEfqW1}#v8fa=-5QQPNszmeZFRituf*6m1lbSybC1|Q4lEzqQucCh)NTX@U~-B z$giPf+R#uaFoy<#%MAnyf@^2!69lVNAx}gJ!rQ5eyv5L*{uy;ks*Hz|tmIhbNok3& zr2Ooib^U5ZKIj`+DY9jQ@y?=j>;e+Tz$aU;33 zXb{>mEN()Ucbc+?9tog|X)B>$tzj5; zebbcf-v={TX=K9or2#3Q9-<7~rg_aj=SD#u@l8>-?@XX2R>O*F4e0^agyv2f|JEpZ z&NfA2&whJB{wv91SVJTPY#+ou^~-Rk+@zbve@kf^Zcm1M zOj^o5rEd5)Bsf95qmm_oigT@aD!X%JF(n|B$?inw#Qek|y3kn=$vW z=9w)yW|f7mGpi^3GHRM|Jo9yRKAB@somav4>%-gnFy?N7yv8>Odxn@ME5MNPsdnbg zjJBu4@z!uK>Q9F&112LhG&+nY0}fBVpSJfl%={~KX3&{`3*;NV>9EgOv28}&=Bh9z zR4f=bnmjF|IC&`G=(%!}zb-t}W&`%feOK--O%uvl%Y?$t#ohX`0lL zes6$w$L`5{#tl;iQR$F37%d_1NDsTPRC^Z|2=ux2;YuyCUejfaEp4|N_u$GP7I;SO z1+9_$m;;o@Jr&tU1q6PTfkS2e10~9nYGd>izLiJ9lR$+t@2haS+>-=*0%jOj9!lM>sy{Ja z6RJOXdu5?}g+|X?h%wCz-%n5q6>08uug)9-`3-;d$4#K&&XmWs>gI&vh$F;C*Qmsh zgc@%zc?__Zue{j5OZ3)`{b>PXd78_VY7Lpc{%+eu468o#v1&B0ssvHqZVo0LU z)pkgte5!)2BnZ^(8qzVRC0(Jdc!Kv(so>AKe1*C|l}&Nj3OU9zl!3o3dErXqkTi#_ z5QX8emE{s0wz`6=6R(9wHNcgp_7 zp~(GtS&`_xX}DVF;4iE5o~7eTao!4DMqNhEC#Xk}nom&I$aUW7`2<=2 znzbn+4$jCYaBlM;B_Y;PWbAl)KA|SWD5ciBvMVYP#wRg6QMQEAT0kJ?^{dWAg5YOjAf`y9 z@cD9V5(4{E#6}^Cg#_S3wCE~;XCejnaB@Cc0wD7936CTzuA~A#Ac5(tchdD8J=zV9 zjh+sp%4q0q;YgxalG{(sXUY{2LJ?m%&&96eys(jYF|6%~P}A^G{7L?(lc5G02`Gyus~Dceq^uxMtNKG@3n zCMW<8K7#9VU~C}tK++hwFR#<3gWz-as)Qan%P^Luqzi#>4kL}2|9M?rh07Cg2pYf!BxgQ^{gB~Cu+0VnRxbME452_YahRa=VxQ3OPj;_2 z_2#_v9JRt`Z!!#UU2Sn%{P84xPgGa&`uv>viF`G8lwV z!_m9|Mnepz?oIB@Yme#ZmK-Uf5`QoZn&=`LmdV;Q>^^4GP}|HoSP`pdb1o%X7bk5r z>;V=eEO`i6<{VRw5x0w-CdTdZ-?Ox#7I6gdiQ5PA>ZG!i`1-N&5D>~^x3$&r40dY^ z9k6narKY8Kb^SxUZAwUJHaQWatdIJKT)Irl&h~J!u{nBR82p4V3S07eEb6?^twa4W z7`L1=68x$=_4r=3NATV|G$3rP>a}&VKabbh38|w4L2jWpskj`TkmL2?=J52iw^aRi z%FRhEby@UKoN z!__`>?;(du46*fZS#T#2xFOCsFHtO(p`vm)B~dI^Fy6r+QEY)qhSgFMfg6G-k$uWa z4~ylJ)5CIw8W?Ef3cqRT1#9w89nxq_^31wRPJ+m@)2{Nx$06TDc!YeK&1*F>ZV}{jk9R{B&&q8sC3L z3TURuBe{!tw4~~Io#qnxl7BEn(793bAfZ`R0g4e9=h38m{z0R&;wF9jOw1T4j`BM=x z=CHv6lius}0h_O?u1tZ1bX-^EF1#H7W{2G8xjn_vX$;P6l~3QJ^|()0wNT$9l!P^* zI(mp)vr(Dn4R#ft^bU}xtnTaW+4V#IpJL{ZT%5b)jehi6Bvl-^WYD)c9Ov?nG`O*w zT)lc#)t)1))bVlukf=DGG>RWSz+l|+v>eicxN;$xom}MNFnPPb-9z-|s0F8cG)UB- zeFG7IE?x?}$ARa@J$W(E3HkWx`#Zm<69Pg$?VnB;cj|^KD>~=AQ1m#GnTd~L+syU$ zy@=tD+HZmTWjhWkZm4g-_~9VV(fCuPQ$`SWO5^%@28AXrjUi9^tU6DM#sZl-1Qz%P%Eq!loE;UgY6y~d+fo+K zit@=z$vwThLdqw>1ghI6??y`XQAMK)8;VJb3)tU@c(IUF%5XhDC5e^l6MQt9t5#Ee z;?qof@f^I$a4z9#A^M*qWgo;&kMjV%k(hZ_xf<*A&-S;r`Y;c;KZtS7Uqnh@iam+2 z8YSHzmIYRoDA`Z3qEW)rBWW&L1^Famw8MnA5m4i@mWI=4A-dEy!H~$PntTP@CxGnm zP&%VZdvfyNc-l;%J}XAsFt4!{r)3z7dxVToX+O|gA z^|3-$iu#bJ#&gW}y0fd?+X!dLTaZFlok|(9HlNk%^>ceuQ>;)`P-ZkMt%j=N(}-v{3acNMW6%if2nvE~+LYffL>df{juU=1p7<=>S@ICn#+vyc5f#g?DmASJSYk zHsAyM$ez{R6h3f(LT$*Gky81R92o<(C!8pCazV0DIMrmJ4cP0#Uyx5A1wn#Lkwiic zg2V?9gCMzPjg1zGRsET&g|j5!U6Uil#WU^~`r}c5`}7davp4Fs~~q)XOm=;U~97=2bCKfUfP zgF$&RHt{mM9mv59&Yz~B-;L4sV03yo$w_9-u3k$P9N4lwRxJ3$%<7x3)*%YKvli0;pd=9Lhm{ zr9j zK849d@2ZnjG|Dz5t{o%R7RgLB=BNpFQ_xu^WbxE!@KCRFW0$GkPodiws}KJAZXJLVEU$b%RXpVsWmpIu8Ffn4V@#P#%(3_wC&*&%mfOW5)LL!+(p zTe6qa{{`mpgI)Y!y5>Xg>iUPWU0p8QJIE_ce961JBqJ?oR?{<^J3HgS=t6xrJvzG3 zmW4t0m9>0bX{YpOEnH9=iFK6Hp-42=%eSnl;gv}pXF>j?g*@S@(dMuR0rTdtL>ElX z&eoJQ0H^^y&`tqBXgWH#Ih@ct0r$5BdAq;wRrLMP+gR$zQ8lSyZ(<0UfAsU1j?cp= z_zN`k-`}JbE<=0FR+PyRdK6i`p%&SHFY1+ixFGlDhc)XDp6^5A_vq-%nc;Y7E&*UY z%lb$`ez9GOWuf0ra|3kX@+st1U9#O8&Bl6~4Fb^eVVA8R4L3?@&;Xbp!DlL-yC_#Ihx~u!%HL|OKiMopm3kyfWjOFJ#dZ&T&$Z^Ajd7wIf$n}kEe?x(JFD3ezaUx6buHm48ofm+ zWN$LtQ`Z_dAXum8Exh>-0z<9aQ*hA`CgZVD=wW)dDp*9`Rgf3tYsJ>EzimLK*{^sF z^qKRD_a9mB<2ct8B0Z?Ifk36*O`+&W(Y69|!im*~d;9S@obm9bBjUvGHl>Jq(UWq8G z696-4v-xm9L#f)~B?b9mfsvS%TnI#&T?Q^#AYLtTKtq?AO2$_Z2fo|I_1zk(3?c1r zlrDVZgpshCCAC75&F8G`YObTmAitW?QK{)0$8!5Bq|q0(5P!%$;w zY6cvUmx2tNk8lZ8*nguG#-$_$A;RXf#~^d}D9W>!#{MM|3<0VoSJ(6u1GFTU76*3) zi7A1nVhFeexk&@^s|Cw}>gPD(OFT)MkMh{PPltRS`EuB(0Axg9axMoglGXinIXJ03 zUF6}yY|{?!d+JYRJ@s!!Fbc*CRSY=`TTw@U0gy&vHf-0Mk+Z94xGoqNh&A^}{d;5w!xphVTwU5Ye`#xmB87erxvTf0%oMtAIwhOv z_rfM;Es@9lqnR3c(ZP7<91QSSVW@_xz#(ur*U7XzyzV}+anp_Ep|G4fU-j1hWGXr4 zFT;MSAkUW{jr*3q0}jzJ?BQVax-naT?m@n3Zq7rX+t*uhI-&v37#kJT)g`aZWUi!4 zLLAe38U0a0#|`%Z% z!yrZCF2_8lvAG0{?jEkx<0|vGAoVXGnCGfSez$|+sbmMK{DTBlO6lhvrc8iUa=mM`ny3|twbsuDW7i)V20?IxQNWai)@VfRzXfgGe(vE+L_yiwW66W zdB#s7dj|b!f8z{{6YWoXOTEPt&wtUauI_b9w9RKcC2X6`$(~~NY5a;xz|O26zwJCT z(DO2i)Df}Xuv{Ml+wXO5FinOP%q72>q9_(dfuis+u|MV6;AFUYW`*n5VFY!k*K*lZ zHbT}gI@dHpRx;?(f~*BAjJ_=c2JI*KgN`3vE9>q-i$8Yl0eIDM34ZN`Uq|4Vb{BA@ znHR}pF#(o|x|(mY^Ig1!9Bv8)#9$4}0O) z5qWxWNe$2aFfyi4c$=KtnG81$SLb)z_J!eidgpMwb$;5X?aq4FHU6r+DMa3trB}|u zdYsIz^N2qbW6I215FhUCKE=1^oqFs(9&X9dxx4H-WO}~PVUr#-dj)lhvQdA)3IoO> z^|iM6JTU7|=|yYbS|56WRR7Zp4@DR#&k2+(aguUq`G$tjUr(O6+G9d5_(7RMtww+F zOB>MT2aJCGhl}tB0Z=Tcxevbyu&tQ51BQNvKfXK-TS^VE4&t$tZA&|^NoQZaEwDFCp(_nh>+>mt+`@kZ(El2*e1Mclzug-g&9(S3qbrE2h zFwRAQWx^JO3p4)d(u&`G!?UC9jjjIVfnH~SvHS$GZ@9TRIyV_j_+Kv2zuX?aYy%n* zDSvql*)`lA$p4dFL=@x=xR%?t2Se6P4ixU%$^GNe_5=FLxRShaP&Wq_E!V82Bd989 z8|3hclkadjbC-b%=D+(yr$~+rrlq4u61ZNb2Y7XSB z9rD~9)vHAboSm4tl^6W7hs+qR(k4)$3?xx?Zsr=boIsUu6Ls#aE%5}^jUo9KP6flj z1gno=7x}jsAc4y$vx_IZept->QL5#?N7vmNVkR6( z23vSl0F1H^E21%@0hS8(D&%9)6^zEtQ=|LFLev?BBiS_$fJdpc-6+BQpo<8v;|kBn zJ7N$rW`F#V*qc(mi7wBx7F$`p_(+MrF!|OT8gF25NB1aI47`IN&;D%)=p9W{z=h z#ADV#pwuJL{)cm*j%&3b2dnoG;MyX&Izv^%#-YW&sdx{ST~f<+JLiQxk$))2&&FNK zYx7k3C~#KQ%_~>hYW|in7Ba#Gav`SWl}PA?wqH=KX4fQMC;Sl}$oAH*g+O*;mB@Ha zuUyzrh5*0IISrt$vbi~nz>&sYp8PCwUY-OR1LnWZCNK+i1d|YRI^{BFkD_-hme}SH z1E-cE$=@!>kvz@iQArN*z{xnB<(e42W%5@Axjk;DY@!y&o!8N{4S9>=7ldR&>*;vZ z?VAen?6?hzDHEEvOu)vtG6!@@e57pXbjehJoif@cp5`CuFWpEwvQXr=I^dBZ9z2K$yYoGE2-N zxFU?N7A6u9m1GKJB`aT8!e3ahD~mV_qe{c&PHgwb8yU48xpws$h6QD*TZRL5=1E@D ziFLPvJbsl>3RhX|&bBLpFPa#51pHIv1<}SJ@S*&1i0R$w%`5P-{Tw`QzrBBPI6eVSeOBZ%dn$nkWAjK{TJ_c@ z_zy(lgC3d*VxrS((DTdDL61(C40@V{4L!iImP8W58$pB779xc=;v?(;V$G2Luc6LS zo1ruy&)19C@J4(eeyEZbsqq<4R`@25Flz%}Te-;h8j-dq@FF8JrVu+6iGb7vV5Qj5 zO?=!afe8$7#&9G)bQ1*ThHhdZU~_EfCjTw#9-ich>$g1A9g}ZHUmXPn^oJTZu^dvD zs2p~z*3U*k34%agA03~HO~Mh;P3D{w<5L+Kn+gJq+K^p%fL{8M;kF=%`}g!{R4m1q z8!Hgs4YkOy68aM5OvNx!-{kSVhh|61VM%b+67!6KqHJ#&QhLU0Yf4S zu1$u3jv%un=q0mY31>*+^E3X>*;6+1hcR#UnU5B9RPkj4`(uJu`PxM33r9R$eNXh| zYKDCAM-^5TYcAMx&Wati;g7U{f(i>yrQW+>lI+e&3k0(-SF<2=b$vYOoFktoAB=#| z{Z8~PQ8KmOg7^#dtP#vXYzauvv0>va)zKy_&RG~z51+g-+U0Ri*Ss@gm6f6XAIWQt zruq*O&m5UlkxXY>6n`uum0cr>-R-1u>Q32TP)@J^*Mg&6?y(Lkz5ZWQ)i|oi^o}?i zt{>trQtSUJyIk2B*5YBwa-ig^(Y3V)&kx5J18Zvt%8^n~6Pu;m3-4#dJoW>_(9I=g zEOmj(+g?0)uR6VE04S1-2uBu{g{>%QH46)A>PR)Uu?OPg%q%PdU-ceE_vxJ)9G2PI zz=uH3yhtgg>fw-{$kn*x(RCsWF-$2@rAMSv9Gs|a;dNw>FuxnNUQY={4l{G|We0_; z!YI0&g;i7xvPT{>bW=W9cT+a>q>T)2Yz2{6V53c?W9uU3!I7gad%s499%T< zI~DMZ7$+*GL8Dj_%GDS*19@UMx-RTAMGPLAh6-C2c!O#dy@x8zhnBL+^m{Z4QeRs5 z{T>ChhI-`=fmgv9hd4_ezF+=Ahh!+ULSK|;&amOiCo7Z+DtoiNbB3i*c6OWxJIlS`$cVP40EMu>TP9wCzVIc)rAZo2-nP$wP z5#%)+72t7WLc^d*Zy(vSx|@bK2cSe)5E!qELk7qMzb=|8y#^=bL$t zT_u;QEuCxDiHYi=8E~dWX0>1~d1O!0#09GW+0HyoNt+B+?eDyb<)E||$x9wlMPPIp zjYQ83fip8#u^We0R95_^(x{4We>TG{&s}B#Y`%!sU=*pLI zs~7Z(;1XjoU`CmdF%v`?Wz*@yxGWT|blL&(&g^)B=}sf8#9$r`D$$Ig|mU(X8RblPMHm;TZ%FLV?17~NwFp_}+%cIc+BBrjiSq1z|2 zLLW&Bulr?Ov`ng;69Nh1jA!gL+(ISW41olFm?4l@G&2N}>$r-B-mKckvl6$G2pKzI z3s^?_R*+{Du;8kr?Y$CUG&6z+K^$)IAePMz9^^{(b1pr)ws1vgCY{aXo(^De^E}gf zp+6q=w@(iZc$q+0Ovf`_oP0OIUs}i;y?n^4hdqww*}Zl(>l9XWmhTn0s8i%*@6M!L zl;z7z_ce}Kj)i=uSAZz3wf8LHT?avPuG=W=U_hS&OeP%&u9>2^@r@?U9)J z)@}Pp$UmcTduwOV+x5K7ot^RVok8iI+!1qxs6205!d;0(-fMtBf8^uzz8C{Ev&d0p z@VHLfMgq0HYfY~mEFj%nky+UL8r zd~~_Vjbg6cAb;A$@~Z5tdsm&LNA4L|lOjKu1_qZrcF+z#(a~CpwXEgjUwnN_gPH%l z(dOm~1J`pY)-1>sB+}$H-S~Qw*;dm{_ZI2zC9d`Nypv5h6^F29$Uy6mjwFd-+>ce9fygVThMO#A~#c)+o{4#z*+FoBI&FykcYciA!WVi-qrOF zp1EGZ+Muf#_5tzBeCrn`hiUAPA0!kY%xZOp-lGPi3-yia=;#8j_9JFC@h;5<7wcF7yia)r0DGpRsIu`;$roFRE{LEK}I94>7A!Sns`aBy^V=FD(B zY&XsO{(`(PHEeOwDNCfSQW477YMwPxaoq z9WG+II?g@6N8&9e2t+&B%mUOu1>4_O`d*@!t8oRD_36&JV|BtCUE(-L{0L_P(GIrl zpvbEFNk8^+fb5poYr+x7_|C843mH^87QGmK|O=)>5LH+AV(aCN%wm{aUA6H35_xMMZn^hYbgcO zDF&&d%M?ZWv=-F~ucOPV_rm?P0^Mzk768#8|IkAcqTofKlD8J*SgK@i4g1>$9F6@p zve1or=l-pPw63hd0$xKNL- zV8G2pKgBRJo(p1kxlK0Puf_%W1`2m|tPIX^d76?1o%%Y(CDo?w^6~f|i*y8-lwcN9 z$mE9&8`GvBDVO1{qi|4=?_?l?~9oEn$hdTMY|iD$iZ^-F~8r;Sn_qOme!{ zC@WZoBRv^1OIC$@Y4halCSJqT5XiqpBd43LtBH59sD=@>)g~WQ3q`)4@Pfm<*&64P z3<8w`QCso&E(2!5D%26J>&!q`+GMg_*d&^Uymm1iwIJE0RoQ28EeL^mg=s+uSSX$a zfu0J>f)HI&6`fDy1L2hy%tHPyqs%C!AqcbauVc&C^Fv2?MBBY@^vyT zTs1?JEoaJhj% zL2&I1eS%<>qLC+WF*K(?T(_jkc%9Emj#ZwNmIzxaAf*GBrFY^LvQMcSexE8h;UASO2~?bG#Z%dxBa1nqT{^== z9#9mveC_`tyB`=1&ut7wTYwM0VL^M4$-^Gv9hGe7hyf^(${7Y}$+q#joZI*!w*JM! zGu}&XU6t0_^^N)y=-+S*LWAC(HTj3+P1M{JvYImbjN`d{he@5ncN(_wF3v28}&=Bh9zR4f=bnmjF|I2qBw zaXL*8IqiBQLVYwrxsia?)RAFk9Q>or}**wS{ZaSyHxVu5GWUeFr3 zk2yej+*6T#R6yWY8HhxN92Ry4kdeVgg1%rp7^6He0V{8x;?|S546w@!pihv}%wX^p zzJaGcDQMh&^0-9rK#B6C+88MyK@lRYaOQm#PM3R))Xl*Bw@z8 zGW%L`RjbU$#3zX%$t)c*B#8(+92m+@Vo0LU)pkgte5!)2BnZ^(8o40}v=vYA9x4_5 zIhU_c7pSr+4qG9|c!rn7VJk#oIBaFPM2D@ez+Mw@_oUky8?(rVlbo3PRhitw5*FW^ zA%oB<1Q~_pvO6+`KpaP=EST)b)HSM}leT!z>-=edrUc7(s=s}I<;UnxEG4*$|5W~@ zv2tTLeflrFcxS`EQ}!bz&^xKfZfLHicy~Qd;;e-4^k3h9Yw~Dr{@!DLX1*s^SI7ik}$s{J)ck$ zg%O@8TuKa2)D>8GqTPP1JmNtNeL|IZ$&<>ElutMa`2-aG5l8CR4T>kt+bxR(vN8%P z+fp+MF15JqftVtZ!spB9>Im#l5gUal780nLSzTlNf!M>z`Dh7%;9xCJA!FH^Vjzzs zE3TvhKOlkWt9R1%9X;9&j*XrUqsnOLZQ)3wSCZRL&1cFLq!^_Q|AA1%SI%>>>o_lL zBwh@4PX@&eGnC6f!ks%W^P5{?naP{{wSU7$gjj&pt!!>4v%D>zT|^jV)9PTu5H2wD zI?*698Ek|gIsBmkNVZCeBM;1uKsh7lo1g$Z_z14cfmA{0fuu1pN&ti4bM~r)9y!Y} zmZhW%g*F-$0m2Ti=M0R@Uzc3zQqg!&Mh)R6j;a@0W5QlDaVN0#ZD9BcKPpJRZYL8eRWb0B<9pQ}!F%t}fUvb@URyUi_V_g#T4disZ&GnN zJR!&H!_DF8X>Y0e@06R9Sn9Iqp*Z6hJ^WZ+G<5sU*kan2KT1#l&&7u{$@YQ#Ry;Fq5?CG%g^#L_sf9TEeBF}%YM8y8 z*8hq<#U}si?@w=#g62+XXi@!b`5fcM`cnvQr=y+i73geq4w@U?-oH2;pMXPIkau)*kAx|t_?ur`-f9NGff`JUCg5;RmXb+TOwcb4~7Uj zH)<6mG^>gn8eF`&{F!vgfG_y>BN$9R>z}`uurZ#GHf+t@6Q%JD5O6`He?R0gr1ODF zh@qPKWIhq?nzg;Vy1m6-XIHV>xwNES)!Z%Q-;2a%ojzdmRn?WrQE!c!!tu=xxzBTZ zY8;ixJ_$OF!I`b{>3g&u_vxw@>U)Heux2hC>y3W&S|n8*xMa|`IUJ`eL|6(eb$r}E zBr1+4jpBz7Fc`NyEr+xquF~e|a8bvcs3r1tf4hh1%~12MdzFsiXKNYGx5MSFSjIPHN6TMKO966 zjqB8Y3x2;`Ea>G%SNzZ+0po|0=){j80>%%=QW_OQp!oaQ#yk~ zKlJmG9S*RUW}Gk2Uc^WlXEo^w%=)8Q9FZd_N#u^=A=w;aho7bSq0h4263Fi*k z&r;!n^hok=40+OL)Ok`g7Rb~gu)sG^HkJkA?4+Ka6{V1ul6!i0g_J^qzpCzqez6+6 zJy(vO!et& z=01e@bENEppz1ge&>M-FXOye4PXBCwYpW0Qfct}p8#MVMQu167KY|(v-U5ZSqOPXouOfMnH|rS{hEHh3L`f37QsH-ZnAg~f(=rUkJwnDaylX5Y)~Io62<%WRcZQfej(Ck#bq?f4ya>f*a25HY z1(QS!S)z_5OqqJ=DSCGOwNfpnQqFhu>Mgk|i z69gNjLi%cliS%SlvBEn+X*=PaSRO6BlPkJj|2?%afMFlmv$~t2)DBRn4f!%sDqoT# zW1#kg6QxcrNHz*5L_#(^uM6BPWSV>eDF_l|iX;+p5F|c;7zD{RYizVR&attW1iWi< zWW#vI{X%~{>TjPO8az7|s_RnL_0mG<=;cEmf%G^&Zui>NfR9^PEjr8hi`3!0?orX@1*M`w&1@hDD-ZB`JCu0*Yv)h3j%;5ZK3i{m`Z4XAL zhm)LY)$HoEWWj-P#qZAXcM9Gj z8`pEYv2SgUz||JT9EDS}9WF`zl@7TFn;YZ)_SxaeWZEB3IY*k7bm4Mw!)8g_69 zVoUGQi`)IJ(dqT+XnOko-ahS*Wjp2)Kgfd^5}(%W%%5FL<|t1BsbLqaj*xuv$_}{; zTf&a_9~y0)-;%wYnnBYQXC6P;#Sf-yKJ>1xe<<74<+8nl6!J1hId+IId3Tp&qy^1t zdS-KHXFM2PsPCpnM;F?%FzCLrmai-0aD=RoKWpKF+DNRUlnzCrv0lDqO%1P1&NO@Y zlNR!Xr$(E@-Xi_IL>ElX&eoK586vk+0RNee&TS4Sz0QqYE`I)z?C>tn-}fr|eyBB; zQnDA;ZiU7!Q%8={%s={hOvmS86#NC6`tNU23zwliW-H3%2tA6d-cXC|zaNzI;eyxcY5#mpbMsp}G$AHiw}dBqcpxKB>^H~YMmU-@DLR_hhD z+9(OInjez|B>SK-%Upz60jGjWC^m(IzY5%@KW|cV8LaWu!cs zTD)7I?wmVTmlRu+Tdz0bdRh>O-w@QtqNflBaY$^dx=^;aI*3OM(ESH(iYnk3s<)&q zpIA9(7aG9eFlY*Z{@QbkjeA~uLPx<=?VT}8DhBXeSZMv#+V|VE7!0a2A0a>2ns+wF ziGaHb5@Hvu`T+TMR)PR<5!+=D~csgc2AjdF=f(v+ec`ipqUX6STI{FOzfW5P70`iH9uxg0F zx$xU{^t9C|0$fkoUj^&sh$0)>Y(5)Ta<#5jPX+m5fsvS%TnI#&T?Q`La8>PaKtq>l zr(vOE*;H4a3!BJyySTnvLzN+<{f*LvZ=5g^Rg4IX0Ep7|U%felD6pRmDd0|Qadpn>T^4N(}L z3)LD%W4RvhHYZ#?vu@mo*1tzKJ8Ti_!_}pY^N%Sp4n>mv9&%UjMVU!<-MC6d@=&u& z9`}!CYUD)+U`lRgmY) zkH&pV-vNhc81`^5dfkXaC(u2}H_gp?2z2{;D^9HL2d~X!uB1#t9QQZibW3D$wbNUG z(P39rZ})bsEy0^+`d4bQk^E6Y#|26|pbQL7_je=|i< zEQ|t0;bUTd%Co`AaP!Ox*RR6}>QJxMoPxU`1<7G%*?3vQSlxm=T9CD1h0(WVz@XW4 zxQoj_!K3gWT`TL(LW@6k?E&b_e+ho=gUgmvrow8jnQN#qHu*FnEUHearQANInpBjU*EB{e+v!^oII;caqmXENM4T%F%- z+ZTr8>7B#z*7<3lwma)x*Z8aQrVx2omR>mr>p61Ai+l@R@}U@0X5NDMaBuf1zCG{M zWB2iJONP$fW!E9o^MwwZ^q|=*s8f`U+7nj-1I8ltwYK;?FzZk0MQh($A9{gQ|I-T( zMHnd036v{wl5%MIhKA5zPoB8iV?rmVLWxvCUVqQf~Q*hV?EK;J*#5@+sp1~Ie6(R+4_PObT+ z`j=+WT{eNdqae3t69eJyG-oB>y8GsfMzUMH}Ijg+MY7aVDC1 z!xUg3z&rK`r63IW%o?b^xxx;ZdQa{G1%U6Gat-%Tk_(8P77}*5k?pa~>DukxpLd+{lyU zavve!5xbjtV;X#8Bp>Z$-gaa~0p(hmX9JCr1f_)Fj6#?wy|qK0o1=QQD1oyRGq>`B zU-pn0!&TY@DwKh=JQpbkT27!!xQRM<)|PmJ>c)_K3#UTIz=Z43zr_GaSV+LXxJAoo zV&5PWUD$q!vh$!!lR}3){qT+Fwuj1HgsKF`2ttMCCfa{Pk4~JZtImW#ahd|9- zmMw@W@<_D*;T)*rTCEa@g$oB%4Nq{rh+^>Zi8aVgyt{%ew9!PS6S@FvMaN_B;2t_Q6Y#a%P1T+GIFXsa*g?uZoEuNFvr&_gppOmwuM#~xZ(0~-ebZ;||R zbkL*IC4-*y`bT_tBWN(%LZt9Ue1si9tQpe(^~S-xr{P~OV#6EpefXhDTBOEjJXztJ zJi@FEd~M~TR3EpIt)FQyBMUDwB4Y}%Ly-tbT>w^!4c)}YjS`r^fY}Vdf|db@58VWT zxuKg_2-qAOy2*dbx`&I~uojSoM|9+y(N{-70sW!IO)Q7hB`Svv zVv}%0bdxzJ#rRZ4#-@S*qc&u}(}rI9k>R!=i2L{SXjCl4m>Vk)@?TdiyQM(rI$%kl zlGjBC;)bYm<{4I`BM@8Qjvxp!B!k6+G||CdnJEu#19PKC_&;Ya)+j?F3$9IufQ}%u zBVYdU|PtOBdDo+#J#@t|{#e4>0X;-r5k`j#k}T5m!8 zg?iQq<{-8NB>HSdgAWo4-UNAg;uss4k+Ge;&>B-7az#UIN^ zW!H#ecRQ(^x>I&s{;NbuM*Y8*Xzg;3bx`T`|C*}CQAMVA#MyBD5Py+c|5w?Seq3LT zuB|hzicphz+z99dWv zwxUGM!ivkV*UISm#6Imnp#K$1$I>))q?+2;1MzWY78Zf8E|gD9>Q%@b?U@%T#Z)~U z(i6EFcRadIgdv70C2dPjq(3$@Ctr3@xQZ`5j~Tit|8(6=`8pbrRG_k3>b|;>%(N9m zVu6h|m5!~8m=9Oz((-&eAuqa)vkt{R?Q4eXEFNvq^k97fmsZ4LoEH05WJwYsciQ_i(lpt})9?w2K+WSu}2YNKExK?+QE|SLNtO8jzRW z z17w0<7fqF3gWXVPrR7+ireaEN_}$%SvLawbP#CzQ(lSJjmM=*U4;A8!-b6*!cp2qZ zH)tO1E_WNwU*c{RLWOaa(G?CiE9^~%{a1+O&dtv>k)qesl6 zUU^~m<1vU~db`DsMns9hZ^#h5(ijY;)Rt{I++gH|aN@y)R{u+@vo*py^6Ob4oKBkz z;nH8a<%Mp79;16~A#@WT%nsf3m2zLheIhILk+krf8Xy-md3u z?(B?@?+i-!#6Yy^UEL9tN|g&YE+bHqTZnS{E=)uU*piJ>O)h;GM%P z%(HCJ+;}Q?f6YeKiDY9!&CR5@NII(v%S}tSFjtNc>|GCro6~;$uIDQU$wzb7Q_~?H5Iva zktVO{#@Cz7KAw@804ajn1SoovFX$j}64jg9Dl!VNKGvh5hrE{z3w_o8R?U+`X! zUfk|)jZUvmM=&b5w@>?{XH;15H<;PN3WOlK!lElNZ)BHhDK2dbhbUs(rKCfkU;Z69 z;ni`NzqtkNmM>oMX6oTCR!CXzxp#H_gJ-T+ur}x_hJ8Tu6G4~c{UD(LVOFa%^d2=B zU8rwVM@JWMwI9g?)iQEraQ`D6EFEa1Q8S7-A~{tr+KN8$1DJ)T`;x8X_Yz9)W{EvD z+8p*4>F-$K4QE(%p4LntSE*%)X-)yCVLCdu33tTKja|t-KQ`(OXt}~$ceW?u?)oI2 zjoZ6Mb``ya<;8`CBS%%7KAn$lyx-F7E|NOelaFX^xW7qlPPE*cSbDaFdz7IR|n zEr@#zlEa0qKX|@B9uAI<&YT&JhwY|$-(QdyriLvpI%SE}o{(-7X`Kn_zwM9#GNs14 zhpl+p%5i$e`y>iU?sJX5FUT(-vwCv6zu6zx5p8+m-qhvw$_F2sW-??8tXtfmzCzj! z8T6t$nM#dC0|47oy|-?Mi&(CXbB~v#0-4c6EIN})J`89kYgKBT}qNA0QDCgps{Oakf$BPr7 zEeEiDe&gp*% z9pXj4lYt15J2qfqnkDmvR&_|*u`OYVxmyhj)7))w`FHz?E`~?cFfhsKUZbpF8IJU1 z$SheE?u8I2Ck$Y~0A%Rf)Ka@*b_nF(qLI^0*VV+kSX9G+Rhyp#lkX?I;4p8t#zryOIJ_pL^Ifz7NwB9b}=2bAlao=*=KPr2!VNpX+a2BD4qp@ zo(ju?5M5Fgb%;erM1ZZDB93n%f0t2al+w_^uUwH1@_QK&^u}H#ZY3FFBG=TkDII_z z7Ze9FA8~SdHh5-{T-}u$;ze}@r-G){2;QTCPifJDyQ~NKAc~APa=X#7iF}<*3m5u) z%@SK6hx)TL!$f|N~TRip(wXEYn67{RN*~xl)xMs1THra zCLyI~xT1Wg9j?0mIuiMyZ)By&mJP-`=ipJ%iq3R%5BHmIEN^gSOZi&2 z_9s(6ky4I6P$)v%%(H!;*)F^vM&gyv2f|JEpZeghDQefX~=tHw6fh=WtPL$M4A zG9ztO05>d{1rYbtFTZeXnQ&wZw&{d{&cu9U@}6T+hIHzaByWm8(PxjN}>N1 z9rd?BzTuk=d((<-Gs0I_VN9r4Fm5z?T1Ij5kT_1Sk>5}Ly6{Y!4cH_1UAen7O*}V# zzP21O`+Q~cK24K4_BUScERn_44m|Vm+$adNJ9barGj5nFh)RdN!PuXGEbNWxNd$dv zeYjGKtk-lIV@unu#yz+)hy|WedqHdDKIQ=BaZg3|Q2~KpWgrq6a#+|IKt=`|3HpLD zVk*#nIsq$hpW@b&w+yh$3!qPsQd_JOc<*_IZ{Vp<3L3YcJTB2YP@+7kHpa>obHGR| zoOxe`)8(Ed*b^|rxbjfyepUU6@g`9H$=fRn-77SD-a?FNUif~3QUrH7c0^ljX6SzM z8~*B#n?S>zDUWN_%?ZU3>s>AV5tk4%k`|IM<6W73mZjsJEIfh7vNdE$3`u6`xKctA zg|4#vy4A zTOkU=VJpieI&5_X_L{)#m~A{FA5L;&>Q`lQ4@+2lZ-xv)rx0Wm7PZ-~*H<}1zC=f+ z5QyW*lm(L=nYsoQ9%*-G!5vbz&^xKfBJn8ua&TSr~B*Z$3j2%zUC)9)(rBt#arOK2To+#VG2v3yd62lX9 z1s0yzm|+fE=o6~MOP*AQqHV}FsX^h`p$is#~@Hu;BzOPb)oMjlxQqqOMH;0jiu)}Kx0+r%*$(24; zDL^j7ti~LERUt=?mSzVWZ%k4WFwuqr1{RVbEirtgc^?K1uz<7ICXEr zAN$3L!0477DWVb&$%5mM|H@$%ER*%`VfX&M3bqiaZRSubV)bm!r9|uEq>Y9>ECBH{ zivX56$CP8l?P8~ial8EYEN$oy-ni!LAwGF^QdvrT{n&U22<5S~w#koa7Zzv-tej)1 zY3W^E{}6AR5)ztCPWa_dD(j>EA(t+DxU)T+Y;2Am7zRHfjDowq9*a8dxpjcwwYe{9 z+;YxH@T>0B<9pQ}!F%t}fUvcy*VfG*JmLCHphfmA^d=RT!xM76KHMChp7xfi|4zxV ziKQ-!9*Q#-TIN?s^0B;V==Pnny)OJ#<}*Hq>i7BE8*hP1&k7$SYAboH?IH~Yy4lCCp#-m0#h*q&xL&(bx5V83LF`6)K}cG9)H`6yOA zGj0-C9u9?%#yqB8(%%UCA5;IUzdyY}3Yt5mp+)t#<#UW1>rWxLosM?4SE_O8E4TMA z4#y{U##<}0xe`OXmQ7~4?EeydE^T;3Zh9^)4xT)0FaU>Y1JL;XjVV~GCXcwbbX?nE z?~#@Sq04P{uOeUa4~7UjH)<6mqOFP?8eBY7{!F@Lz!xM>eAn?M@>&18cj$!#*agnOhxw zqaVE%Nfie!8T4%q#~E1faLhVdA|Lk;iHhS%qxj(i48|=_%ONd@tF(D~obKe^+x_hx zqBln^INhT`q6Y07hya++Qkc|H&y9QXVxSZ9@zeKr^7#Fp`_umEbaAI{xU!;i&I?74 zBbk|am2JA#_?d|1muXz5_FLe7*`Aq>8|oV{emID8vvkS`!cJ*iKhJ>i!?Bb`#Skd| zewIz%Homc2;(~jq(R=6sdCKa(-kx1QW`amnQH~Y>=|s;~aVA>suVOVG#ui z(=x}K&6n`G0}iqBiY>6+I~z5KnaBK!0{uOMW0J6hF>5MAv$;pG`X)}d-SZ>1` zU=Yk}Y{h9A2IC$f;~CyHmJw^zI0%7V`Ecj$Rj{5sj(Ck#bq?f4ya>f*a25HY1(QyF;vyplx2{#eq~=l3hN|QJX?x#Q8gI}obXN% zY?O*HZ{o6_2hhShL1{bTomd_%ypt=k*MDX|;M9Gj8`pDO?Y^}=0#{oUa}-X^a@oyoL4Kt} z?!o59xW9dNxH6gc$5YOcrjcB@T->nO@fJBT7;aAcoD-JMFE=@Kx#WdhS|q;bw9!6= z$wcp}lTGkPodiws}KJAZXJLVEU$b%RXpVsWmpIu9&sLfBX1Z2ZuE6FQ6QqYL%j^yuh9 zTNVc0SJv`%g@j*q61hwMtc43|Be9NBIuwb)6u!j;eUNg4Gc6iYFFvpPcS* z_IWG6^2G?O)+=hYQCc?j^ISL*W-LF~Aunl75&<_?H3Dgy$M_DI^VVxIh1(>5(IzY5 z%~93JJiNbfY zhFZ6$;G!W+#$%(c|pEbYz_O{Eh5wYBkO$}=bA#Kf=X+a1v<+E0XVyx zytwy5B*2;SeTH+@P$d>119`Yh?!Y`Bfkr!#8qm^h7?5ZgPshv$G^bOrXZQ=QMqTo1 z} zA8kj<-CU%M{5MKrTuM>{5f)-}j7akN?Eg`eXD^NYOC%TqR7$3v!bNOR06oYbB!a;1sJ zd3`GDsedzqQ7~SpV#rz8iaO?DyWWhPT}8un!N5S&GiYGCP(u`k=R&oH(O9m>yUht# zH`R?B(faquW``|eeYm=`asJZQ3`Gk49&%UjMVTpd-E>Me(eH%~%vvIk`$sc1@}h(B z&N-N`w8BsgRe?j`aITYSd3fD@V&mqU%0po}b-wDY{mE2v%!}Y4pDM`n#aELFv7Tw^smiiuB1#t9JLF^{zk9oC9=5M=`FyF zu{x8yYi$W0IKy9&KT7Dh!NrOmHKWE@l(vBvk&dOzl2Db2Tr<-86u}@i-)wEapb1^i zVvaOd<;`)*#es7IMc}s)FfTiQ-vV`%zZ9@AJ`WZq$$>qW{5*rnIQdfBxGCiGj$(qc zBu}bSgpHnQ-xn~^Fi4TO%Q4SsY%T$#yN4_FxXL^(Nd4m`rSP+Z{B8%sQ^^ie`3DKA zl+w>TOqo=ylBM3{)Hm|x4*5AMZ!zfa25Getsc@uxzA=CqqF>TcMK)$}n@zq|kQ332 zQKi3j=C)z2Xr{~6olcpcX_@f~Z9d~EVcTp@_7t;E<5yGyc4qzfZReSRo|jRij)>*w z=ro>}4gbr?Y%>a~JwDjOlX0CUifuHNCCa1hUb158B-{{P0sC1 zh8u^g^Sf>P!f-sjb2#2QKkd_YXT9qhe^uTTBJax5E9YQ6PNr-6=_2`1j43m3L43Hk z`xM`vcj~eGc(^4)=kBuWkm>otG$MM?>=o20%0}&pD;+Qvsjs!g=Yd&&N-tXb*80#3 zr23y;cqqa^c}}2QiIbE=Tu6%L8yeDoJ$d44k4e8^4CM|rAN|2EZ9u!98<;Jjf4B&L zD5$v)zX`Cdn79Lmeuh83JPlh)4X_U4u{26zI9~*lC|?!?lL@?kz9r7w>kMLMi=+4K z)}C7WOZBgCYyXad+?q`s99}(u*-zdPATgZ0w-piSV1sZ?rsL6IIM|qMLN`g;WA5|= zoK!QsF_^Z3yeFpi5k{M2Kv=s7BiAq5bser1M{sUBnA`mTYq3Q$m1#S7Qz0cXNd-i*; zGxqoU`%2Dt>@xeTz4m(TwLW{T!SurRkQq{Yv{8@`&y!nnrcHy&-?zXnZ)CTko0$^9^vYL3iYf?~r92HY zo@K;k$MqnayQ6|U+Q|!@8+npk?jr;|dTk_vWju>WKHkZ^?Z}D(%C$1j1{x&^N(sRk zg)mWiM~A#5NA+q^0%s>?Zsi5P>>)FTs|@E9mmpmPNmQMixkfD~P$k?%ojYqwJVA9` zNWO(rp<`fz6E5W6V}K-%eS=JNX^k3X=Ruh!zVnT940eAC?tl@QrmXyGjX7;3QxWp3 z2dn(5{bn%FJwun4R>O1N?CeB3mBVJmKdAXt92*hwK+Q~BgS}2Auam1)79DO`h6F3z zNFr^$B4d??x0)n-eU>g#!(KtUMg3;J{YINUNh&rQiB)i4| z@F4en9MREF$HTO@=-kGH~^NT8ccE+rrfJU)3| zNKm!9b!NKJe=vswQv1TUDtJXc6kY5xogm#a#z8XnMW7GOGHO(&Y`D~$qhp{3rwq56 zT!90wL&iKDa%JY&j5nHncw|9Lkw>Hb59dG~*J?oy7A_ocZIN7^p{ily&|=?Iyobsz zp}RA7&X1b?$AbJ^+@-uWPnC}XXI0(2a;2^2Zy94D<*tLH1zKK-gkEU-1=VVHP2zRJ zAK~T9-jOSzoH@5lcqQ|PGsQxi@&NoU=QM!2%I4-Q0!JEqdGfPp7Ij*vla=H-24w?@Gw^P>AiQ~@e zXxfInMez&JaeB|7Kkcub*&1*3r@e*V{E3_IJQ@bZC2uXri{dsYrc7ww(hp3b*ct7r z?Wj=C(|2BxvZ2!@Qvr6$Xq$LS75x)V*>87ZMM6+!j0A0ofvjUGGnvOVAnonc?{vs= zs!(x1l5JqhNX2!Z5@$$6rghfw%&Crtye8U>2m%`cVFt6yEHRT@(NVv_mX?61BvT+O zS^2^e{=$M?S;ScwRa(8={X_!-+9OvkU%?=tEOpD^pRR-nd88BTZUuS#Dxnmvve=Df zS7xttZhy+86{rls^tJy*oX$1TWjS;c@$|{R_kK35Y(H z4@BaF9-0YaqN4>pfta@WMSBdQqKe6{LqTsMBfbwmR7s1}_>3nje3M6* zwSliKUEp*_L@xA2Mr2GOb|?}7sSChLv7wvzxKRQl^3X0fEcloZx(NbvLpQMyusJq# zlfTQlhbKmCk`3RCzB&pD=npk+VlkvHQ90~bt)Gp81nJ=tLEaP{pNdVw5z$TNoD}0z z85x@j0*u-Kv8g2m=)A88-Vx}f9~o{7g1CQAk4D8(jJdG_A%DAS*=^;71uON&=s?^M zboCBeH@IcIx|_D3;q^_hU`wBco zbIyt#wc(GnfPxAOPlYjap1l=Vl?9=z>*GP^9Qj1~Km>&DccX8KlBx9;#6PHKjbIL9 zOF)8-H5+f!j#)?vc}uj*um9InHI6DWy(7+s>xcM<)cU{5uJq&jYIJSwfpf$0g}~Yx zf^x{MC6w3*Y%jc@5%bvh4?{PXn6cCaDy%|iDKCYm&84->wEDKvvqGn-5 zz>k!UrD^I&HMOw^;^WLLECOHkB=o&n1BRruEPqHo0m{QWV z^hEk&GjsA~2ZgKn(({<1oAL|lZpwz9w2{oT6+~izjW(5zt&5lsSGuR2?Ug+5Qh3(Y zX@Pvk*9_NLJlde?!TJKo6XuOmeSuL|ee3#Z7i=Ye;Pc?Vk13m=nLx zumbd_JLF~b5@T&V8AhEk+iI(Rp&e3*$k@E97);x7o7pfnLSnbdnHw~vwPSLJ!eLb* zsF`7$MdP-I#8gl7uE4`#l_SR}YNgGmR$n5o^Toq~1jw3-K|bzkOO2su;=dS#i3xZ{ zj1v{ppiwLd<*M*|{;d+o6T3kIVW%l#@X$0=*s{PIRI?~9hnh$Suqn^)(I`lLX~F%T z{fTsE(e4u|*j2-T5O@`wafq|j;rrz;bV!CmEA&Nq<_sIIe6m8Rpt6_iXBs3IymN-7 zP>E5II`DB0W4*h%E6v7Qa+7ikwFr8KpFjz z=j7yp@w8dCxZa=s5WA*5wNQ8|lYbb?F|OoZzU9olZF8!rjUg#$1F}lYVLO1cj?9feLNnXCvGK)`Tg+7uNUiZtm zXqi+wCj=728P8Z~oy&r^8UhLWFhd}*Xl4i`*D*Xd3%6*W$ch~#5i)kb7O;%;tsu`R zU=TqFcbM{Ki()i0f(JnyZtx(M%?=*qO7(Lt!7V+^w(yR?Y$o?~0E3(7nb!0D@uFhpoIqMWwbQbRsDX3HAWbgK*6qM!5O!qYe zk?l0T)A39y3O)^xi(dD}!Jxd^g{+bQs9DmRZPw!J3$tt1asmgzb-LR++wFzog8T~_ zw>P#1y z&$?J%m7R6(vXk`4Jp*e}0zP`l3^_RR9 zKhlk_H<@iU-E?oB{FV}ecE3=qr!r}!ORv`AOz7B7F~&XBfC^facNsPL=oE_WOQi1{M&HCs}-2a zxdH8#FI@6w>ftU{NLlZ>cX|DTXRcSUHs~sbeLyb9wPWH*evnXrFss!WdXE~6&eu1p zW25u9+K=RcY8lxv4!e$-e5`||1C2ClMiECOr|Lyp(Idwn=zqqTWPk-Ql#S9o-5b0Y4p&*Is*y+_Ed zqBplVKR0*ym&ZtHC)i)7HYZx{&8%lqfs|rpYzsI;abk}b#61Sd z;lfrQIM*K!2ggQd&J4%HcGJ8cD#%My!xk5vvP5c6NH-pZI1TOJb;tmjQe)l2Ry=Lx zI6dQi5`_ecs#z^Y|Dhnih|KEA>Hd0uTt~F!iF;C)*DD`p z++hN}s7|I*W6=P>_Ehh!+uFN@!D1&mem`V*~}r5eH(@jY$&6 zK|Y_*7=vE~EPlC`QZSujkUF|dQKV06Q60Zt%rqzT;hD(S3Us$ES^z|Y{6h~Z0(kn4 zf*enk?2Tc6vqc=uw-Vl#a1JS?DHs|{>C;eT?kwIVur;@n6TMrrur&%b4O|WRs}6C` zFLzvJf&!@wk`dyTS!WjNB4A+uyvxR*9h&Tf7>D&B@b{w*3g-E>_|yo+WT!o!H7 zRp0jggcls<&DJ=VWDuwnh}w$BcNs7fR-uk$sL;f>TpXz@$Q$OINL%gW2;8f7G8o_&Te2Rk+J8}}_2^1M` zLmqfro*hN5dNIFo?d%&((l+B6i3 zEf6);YiH0Y1m@5naJhj%L2&I1eS%<>D&&bs{QAohZ#OijzoBkPmDzPcwsu0#aC1S& zJ56q4i^aP|67&voTW=*P396eg$q4!t+Tp4jpgT|aMplY!*^@PD5SR z(!>3R@>*B6l&^KPKbiW86kH=8@r?uaP|FL{jb=zUl+O*J7xpfQ@zYJ9zk_?WxRG31 zG+1KcJO^8A!K2HA+i^|wp{;+KvWFg-Lrm{6;hm;zr@>g*vq7V>81EUQX(zG1q;nxj z>mTuD-!x_W_k~6#Y+o9X@_8Z3z@a)3e&A8x6lMF)1X^M_tfBcuKV`ZpYh(4e4MFbo5(69`Oyro*|~m3J`2Q z)y}+`(dKkG-WU!>{poONz+{9vw?o-7;JCl-r_H<!ya0)ZASR& z>S7Tp7K|HBo|aLZJSaVsgC%U9*M9Ohg=gAqz#h5p%H5@D;<@qjwdIJ}=WCPqX`0lr zztJ*&fy^&=;F*u-MnT&hyC?4%H%t{or2{zbpOv+DW-f@U8bYAYtq)gfk@cD`V{B== z)wl;&2C=|1YA;lUgH~h>XU-T?I({*^bV9LPpXZPT+z|n79P%QfzcJ@$!tMg(S>)S7u*Ju6hEGMQg~C7?RA=ajAqP3SDi7B+92M*h+#x z&90Ffl0aMW1n;3z!Jl*a3Uz@ho8qt)a*Su_G&fQqZJsL4VJk#oIBaFPM2D@ez+MfA z47P5GY;5>wk`q(EDwBIy!s2@~WDq)qAfvE=BMVCEc{8H>6CIgCAdVwb7EE?z>Kawg zNvSuw`1lw3{=`y(yZBG#PZ=vWhSR6N;l(=}eoxuII26AG2p%Om zZyK)FIrxj}yl3gSRGha$mr<9I^9kxvq~;UUHFBMIdOkrGcB*^=iauKEr)F&us^Qln zlAcfC+~z?_Lad|6*h+dnp(eyArIHmE*Xwjh8To{oD2(t#;ZkCFqOOn2T#7mx3 zhNOJL0mvty=#O6$grrbBY2I#GB#@O+P}!E6QE;gR1Y+J;!l@aOTj?1ET}41XL=_f+ z{V8HIt0rn@HZ~A@I5{6J0T8^fke*L?G+A*a75D)OOkcg7uJ7p4Zg6b$bQo1eLvIU5 z61|e#eri5bu7D7V_{w=McCGNjM&iYAR!zf0k=wUkCG3RX-jKKYYyYl&g4L~TZYHz5 zEuURP7-ds3At?xfnb(O1k;z~q1j*qK4M4J0$_8kWeA28fEobC>6BK|4AHj7wFg6f+ zAZd);m)GgiLGU?yRYH%PWf;p+(uKe`hml6i|GY7|(x)ng8g%6Fi?JQWQ+Opr3X`rR zZ%I-TFwuqrh6D5+g=9!8tLQYz@WWa*@ghCa0uDik(WoN^9O6S&=_0|{>nq5(H4QBe zxsL=NA!q;_kevAt_CsU>!?6@Vz|sYOoFNp5CJqynUhEV4{mJh2re|Cay;$vHVQtw- z;mSGu7Of?Ny&L0HiXYeaE;wm(=oGe>vorWyuM7VvgFy&29L)=0G{kV~-eRN(Mz`ch z5taD5$z=|!V418mlCaulQtUmfU^%OVo@w}jw#28+r>^3 z<97M`uC$?FTwb44mJ(k-J{|%>dF-~fI>gd$ZJ`5J&au?A^e(S|h__7%3C$)a0;H7J zpvk-aLoQwRaBFioSz907Kh&QPVg-qB!lF)lZk^+g!MNp|k>FR|smJ%KJ%ab%p#foQ z&Ahg5HtO+fG_=URk=~@@a(F_HSBLAv)6?EU_5Y{boWxQWMGwUp$2iR=^P-_!x6byu z@SifD@iErpgZ?)8E~$I$K;!;%x| zQgy|V0Le8kgXv46+`#%E<`v%j_}eU1F{bIi^7nbD+GjSqN(`}$*?SsrLwL=+M6p7`lJ?DMfaOq$noAwLb~4=?2o=dYI4K6-s~6FnS40uI#cJZ z>e`9zX?F80#djeB*)!-*`)g;m#v6Tztb6k(Zoc!V>oxX%^6jK+dGk@McxK!rusjqB zA5|Yy`&^vi{V;v+Is1lMaEJVxzdyY}9GW|&p+)tt<%^9Q>rWxLosPCPm!PxJHZ(W7 zwSQqaJ^_caB%3QS#B14PmY~<=2o~rAYQrOP(*tVqu)zTQOl<%f-@hRRG}Gjfi+J|b zS!n~X6HTi+6Va|XvUgXvH{a{*DwaFYgFw02xm(CTj>Kl2K47!|2w_9j24_|Q1S`?kf=DGG>RWSz+l|+v>eicxazG|U9p#Rf#hBOb`Q~;qZXX* z(I8QS_6-Ew-gxaHzzevFY1DJ$p1c_7gnaz;{awVa$!Gl2>Ece^aAig3oEM57N3!Qq z>%9aJEWb?SI{QGe5RICsdx!RYGjXXf8F^Q6~bSLaF5SRhk}zyjYu*;p2cvy*yyR@6jZN$&2w zJftQP{8e=?^oz7k((&K+$or8}eN@q?!iHkf;sW;fA|?EEDrLBC@3>zq2Bj;}QhkC# zqq%A|)h9lUnCjEllvnb!RNr49Wgi4pD?C7-P0U<>UAY|V^w0J;Hu^9RxId68@3(Yc zL`q+ZJ&CXyCEXyF1y+?P*-x>eQNo>hP03eRyiGod80|2d)(EI^Sxdudv=ChyVaYAs z`+DcLe+Anofb8&4I-^Q^a`M1<+DM^Zf1TgU2~MD1V=GR}Fc|j;8PD*pv5Z)w#;Mt1 zhgx|m4xd82#;Q68awFPM;4-+1e9?kQsdv)L&BB6Q>?I#Wx<4>AsbVqq<8eAsV7K@& z2PaW!ATG;l+v>gIOC81vSt;s6o*H(J`Rzy{t4^g1S<{=4U{$3BWk$2oYN#qcjToxx zYf2?f3st{@6xKT0N6MTOBmwW5 z9N93QaX;T5kNTUZhf$uLj!q0mbv=}Iy|fTIdi9V;AX@F%CWk!^qT7AsasbiIEf<}| z`$T^69`cIbU77hs3!(II$)jDeN<$~dYs2WX0{Q86ZyXHDld*}HSxO-XGdOpef_~RV zn}gBm;UuS8HM@E(S#V&EsWAl)SGzlSD(`IJR5rE-^}O?ytwGtwp|^B4wUyaH{=3A1 zyrDzx#MYR)%EIl&v$;vwcVv&i)fUCh31FxREy%BS$lci7822~N4woj={&>nc(lnBD z7mFJfJKiEE2E+Af-_D7PCKonNJTy;y&uOE53X_T6WhbdDB3IdiuWJKJBk%JLVEU$b%RXpVsWm zpIu9KP@V)*!(}3QC`(@3Auq?4u$BIUqm6SLvX|5U1?KUIE`Bgw^PzWn{X^NVE|={c zq_mVfmJo1UeLCd*U6PR&G^^>E^{uV(V06B|n;sjTZ_C1<`^s9rF11to^A;|sjl?=i z=};sZ>*ZV4RDMvi#awTm{ZCuS6P_Ba4|@ypjZ zXg1c;_va=O3X=dJw8mm#oPuc*~V$=qP?aG0_D zT!*}(HAw{AT-6AqZ64!0V9r}nAoZh%|FTV1!kZ)dK=0l%yvs;=EVX#II^EhnUY8VG zl-mg2Li~oHJ{CQNAcsR@Th)cKz12ZHM0Dm4?Z^+>6ji`6RBuUJKCyC#U1$V@-Ov;O z{k7*78~42SgpPu#+B;*GRFpi|ar1s_{QWj927~JCjF4})=ADgkBH*qZ$Y}TbX>nNO z+gZ()_yx&gu4@Tb(daEoA$ybAp1Rh!ftl_&zI=w{Gz_(FPr*e)n2g6pp@$g(I)7h5 z?#S1QjbVSYMP%B4X1$N&TvLctP-)GwKxc6v0B3iRm-SwX1UOT^&v33GcsTNKm)wSV zJ_3z)A~m3;+b|%}GMyb}EN1tIIuy>>}6%mqh-sz(~wWE(D^?E&>;9xTeYprzY;YVqX;on8Y^@d zYOGDoz)r_^5KdN&d>poq{0~ZDTuM?9B5X&LL@g(EtwT-Ujq>cJv44pKLx5_@)ipiE z04>R-#lh{gn7X7Bo{Ayh7UU)k$gdVG2dbZIAij1+d0eORdF0DsqXLi-fyucXv`AL> z0p{SO_H>aeO*GEy(^*gbn-PqH@j?|t&casIQF6~)CRuENyvW;-v#T1e%W^R`LKtT7 zUUXm}>KQaJU8o@n!*iip!wSn4Q8fQlw=v=B`U7?2MzsDtve{vaSRbw~ZJht&P-M>U zCNJ;ZnVC7)EvDoSRm!k|Sqd8OQ~uFRjlAe!ytNJUm6jN)p(=0)9L`lTEf24|Ppn;k zpgb6sQ|GH5?N6o=_Jw@9AUDg8#(hiQ0f%T9_HZD2-Iy&v_aNUiH|HVH?dz>L9npZN z3B4ADlQ(2CS5hV+j_JLO{y3rI1{W)O)QlQqQQ8JxL^_r-OF~s9=7AAm5SwqdwqMYM zu4gev+O6^hamtkdoP&AU`TG{Aqx_|Sh4Fc?FiDQEJ8ldSOvcHV+Qv;GpLY}!lqGpm zon|l*{#cfWMj;aogA|Fo9P^yU<`OWvd$?4OtIXqq)IUF(PskMcy$*(_k{zV-4-!-< zrJr}0G6gq`x7<3&+dAausl3IYzZ<00N~FS(^7+O9W{7@?tHz8@*%ig@szM_HYaQr#%~-4A;*ras4`kpbqs~E}P0m$QowonnuV9=_Bth$dO=$(YIy5pxN&S^NT;p zI}82jk+SYAH2>pQUNbj0_YnNq3x5v7AN`*2SWJLrqORtf?0gq5S zB>Y@*#acCKcx^%fGn3x4$TOC&f*Gf5)NTPk?1evv=_NHh_ru7TLg8()y)_xG9jean zmd*3S@$~lLc;no(PurdKu4?>K`TP)hSC(GdhV>k!(|Im{UGk9_Q)b?R_;7FcDZV}L z)MNMYa7%{HT`J$FD0j4Xsf(@?r-u#Gq3Nj~CNN+uQeSI}&jYjmlwP#<9qB_akm~>R z(t{BO%5wtcQkhNsx%;g?a2f-Z5Ap+(; z{3gJ*V&V=M`WgQC@-%EIHNZN6$I@WgP+3{NR~7=v zJm_X(1a%*tC%5FtpLW2#y{pxEuhQc#^R+GlEEC4L2(V1pf^cERKV3j zp?D_Y$$9b%XeI#|s2c?gZ-=OI{GcP6w}s128;c&a8V92tXqt3 zRbq;|w3`WiE@(pd#W{K>iYf?~r92HYo~5%y!Z{an@p-h97dkicB)Qy22zVqP^e5Ee z*TIi>GH*MwqJVO(%(H<;NrF;BXzn0P=*T-d?2p=hEkJPY;uEM1iJS#ERt8@bj6GQZsE z%`G}yV}eqfr^-ix zv#LUcTxqNMTgF&Ox$9uVP78&R&qr z_+8Fv0Cknk%~=GFH1_i3XOTRrz??`i8|W0;EGC$8nX^aHI~Gf9bBKXclk5qX0Jz)l%$6HgNr^iO!Q;oE2g^jnrJs72y6s| z8O$!T#4O>8r-z9IL?xL5S;@*5mhcxAED{xGVN_|j-2Fr&=h!1xE?>d$oh(sH{}U!u zK6fd^vqNL*U=)+W-r zI_{J)FGUi99-0YaqSI>7^DEIok4~2idNed!RQ)49yb&}QZ6Q*4BR;|oAl3}&zYTSc z1pE#c!!G%H5gXo!@52vO(jqlJgr5JN#1w#II)v{Y`O23rY7)0I}9f%vE&Y5Rek&ZxYfjfdA$dEw0fa`nhW-vvtmbW_#-W#pu)ma z=XtW*ITiU@+69bCSc|MD%5{A_=$s>;C?AM8>EDgMB}%5&TM+-Co;89wh%EsLIyfq` z!Q0&PeoM5=%8tuF z>maZ|((C`VL~ECOtbXQ-EK!LY%1CTqN?J$76?CwV+*f1vRAZ9 z={&R$3YlRoQrw|X1XaMux)%;sfc|ucyo_FAtT@y#>WtY|TlEX=kV-_xa@|M+{lK&x zx0wxNBP4dKoVh_$T016Zs2i9UhngA2Su}2YNKExK?+QE|Ryh*EvX0MpjH+HBuk*#j zfdt5!ia|c^YD@3;c{hcoUIVqlg1~rP z95O&A_;t}#={495byixA)oCiGA(qijm%IMP1XIKxl!-t2gR31B0v#9$r` zD$$IgWL9KI`81A221(!nW%NUylamL=(`MP?kUvW#c1?R~q3}{B|1g$gT*aH%yz;{A$Ky~6>g^Uk z8WAN1zac~LN@FmXQlA>M9G{(?qhCD**GOInCmu{_A)LL=MkBM(Dqx5FMpg)?( zW(z-^BJ1Nfk{7xOdW`O|h0slWFgtY9SIW&SK9v>vNLqN^FXN(RQstZwNDyZ{W0!e% zM?O+t^Fkm&A7%(77R?NSiq7IaA_aAdob276l!CH+nd!ck1hNV1cRHR)MZu>5a?$JFI2e>S zyO32905wZ`v&~w3qGWc>T29~~xbg3{&UU*X7UW;hxV^D8=N; zlg{$oV$;&i?I=eG_O1uR^=Y5)y1Ba@K5v^H|FbTZS7m42yX+)Aa?ik;6#2n4Fu3Hg zgLXXB?P$%%TK3-LUwwT`gPDKRXnlQ&f$O;xTZG785-0^uHyp z_4mA!O*j>Y$cfRl-9?x~Hli~F*TewSwk7cW^PbUv!FxS+VY9z6I=wm_!KmQgKJBlb zQDMQ~U}g&|5Q69mi>}1HkzJ}=_okwDZI3PFEK36Y@^8ZluU23x=LWP}zQE1YCF^V- z?qY?M^`3i|*FSjXdIf8Pu432+@Ck<_`Kd_-}A{dH<{qUGMqdL|V}DOSd|h%+RQ7sNdV$>G9QA2`<^4+qCa zXU+`ALvxzfUKVYna$R3jp0;wFp7B13 zLIOqAtQMpHP>^3lX7%KBf4x7hBii!BJ*msbs&{k^5{C3?9WS5R4J;E>jfg(^^y~*s9B_7sm9p0^Mzk768#8|IkB{Lgd}wI|_0U+%cd$PWta%(1vo zkFH?A%|t)NFf*PDVt27kHruNV1^EUFcXg}`&T)B~k_Dams)kFtSUmoxA{_xHC78t& zGWlV{#f&!@wk` zdyTS!WjNB4A+uz~;a*xiVFLNLXykO$bv5xWO}#8GzFvMm;RT0zvo+2o83ZZ?qPF7k zT?WjARj4BwMl>-jElQIk{a=tb%%`IkB)haK`z)>nAuz8nEeHV%#j_yLQ(;*UqD!ix zQY_jOf)FqZ`TLAAqm+gq{7Rd94GIUomjOX<>{a4ck|8E?O--B90T^;YaUeS*YW9mv zhrlz7kSFwZLv#A)*Da|syDrGqPUsnKF6eltxp=q8iQYkO>#Zc^M0Lw0_fOeP8(Vb) zw23F}$rHYjl_Fa<7;kOEc;h8j=dF^{P?xpzaDRSztt(r~*E-stOk=41h;JOQhgx2s zZZt!Be)-%GdSUNEsQq-){toWh;zn|5(O`*%^BioAmG}2fQ})m!bBO6ZCcM*>?KBt* zdp2lP7UMl*H0>nTmvk;9X|2D$?3<=+|Gv=3gzZa9Qa&$488}oY!Vf&^o1$#rnLtY{ zhZWVhiJ|6-X%w&~G9`#WEzwjI>n&+~7`U26+wQ zp891tQ*P2tmDb959U1i}(7)k0ga*Am%3qQ< zQFBvDvw3EptE0~XZ8PQ`);zN%$E>o@b!PR1$LOXB$1`77=aV(((@9SF{Mmkeco!RZ zM;`GF!k!_f$qEo`KGn{=nbGESINlfzM*ZnNge6;255Kep1fz=FjWwh4&Zn_D@#krJJQ3BnIUs~VI~Lkx%J^nEwWzIWsEIt zw;K20${-eaM(qWyk=r2;l*c_4*+&HgewBepWXK_5X8;))Y$WIlMg>ReFqdpz-hKiT zu=4gPZasO+0K2>Z`UEMpJ;V#V_q@h8@YE*-joVKim*^cRQJz#AW95qF{sR@xysyIP za!(TM37BDAdN6grs{X`y6R7^=?UjY@6&gKnA;vT>d_O@cRHV7rLO3;r{HDM9<0jB> zXUgMRb#p>-#8PZ@;JBA(BrPOi#=A26T5@%#Ly}oKWJuyj^FpMdK7sB!05qM%kVK)Y z?T|$IR0Uf}5UANTa^IIiTk!<%p;E!0bNLE&fhwEguoZHQXDD^flBIClqcn%D5QX8e zmE{s0wz>j)O<;D+Ha|{2n&iaPugc^emazEV3>kz@A;>5!7u}I51mZX{Wx-@ermj)- zob+bC_e-_YTpjJtlwkQz^*8UU{22X-r382JpUR&yR&ETZPk+OUcQ*W< zOP*AQqS0kO}@Y@D_w3zH*+6T`Rn>k$5qjRnzcL{7L?(lc5G02`Gyus~DJc}1&9pKA${9J|1O?#1 zM{r#Zj17byNE##e>sJC81fR24CG^NyhOsOqT?l-07-rgS4`WPLm8j(!-#%fJ4w>H0p=}hxkxcx=1kg z`U)2(;1D!`4M@&>2>T&!1>slVCghieAxpk;N2IH1ME{2#LW{^4+YZ7|w^@ln^#sqt|xgNxUfKbtNY@P!Djk*EFh_YyY7^U;Q_nR}u%z5xO*kJcf}>{06ic?{`% zpdw?aW(FpzipA&fL6KJ+*}JRTo9}gY70aE67DC1xHdtWNd!0UDbNzDMQE%<)S55Z~ z_?sPaujlp@N2f73vsFHQkJjTpUDZN;k5CfU%&m^T#gAT#q>2NV4EnYM$0-%nF~w-X zSrs9l@DGWK<4L3V;R6iDEl2MdzFsiXKNYGx1Suo4MXTju`%^{TBRw zxtNRb%K_tugE&XyPnAym2qIwoa4e-!F$9XgpJkJ`jc@FhxZu{;)HGN(B=Asdkh0qq zj=b^>z6{#1h*pJZ*}5G z1Z&s~rlCwLmSNG?E?;t758B>?6!0os1m)!y?-#`;5Y&M@IM%MI{4voDO!7lNFWKP$ zduhh`@(e~`temG&gAqgHFZd_N#u^=A=w;aho7bSq0h4263Fi)3Q_>iL{aZtx^oBZ5 zipBz&Is_K@2Fk{=K%8B~winNeV#zDX-MyEG6ib5XQg=(uC@r9L$or8}eN@q?!lGl+ zXk<^`i+Hh+RLXGOmT3$HyA%W~)hGC9G*_*r`oyOZQ+@iHxv9RtK*~M{s#bV_-bl>c zP%g(h{j>dzjXul+?hhhVo_rB0eJS=N!fKRsgIE?=Rib1+#fnA=mjRw5`{@a(KZzLa zFl*5WsBu|K!)de-U79z^O5F|nVr=piY@YzK!$aweD(%V11LJ8kg?dAbwqaglD^ANW z821Pn&+x9Xj98<_K?v+nwa~=yrx35Ps?LGjh!>%_46Y(yv|v)|o%C|EuSQ*bCLct) zKQJ|^VlnpPaXL|8xA-v!CsAo2F3W1$Dvg!bJYj{b6!jrbjpvvHtJ56^OqzwPZ$}DQ zbt+}Zx}RLJRzX_UCKQY|=`E1?nZ7w_CUq3hO?K-jmWpnyZ>ZZ^_RY952th5@c zicceks`{EziPN&lUqK4%Bvm|HigHmk83~;5P7rLA3YV*KM|HwL5UlV{P})v-CzeMG z@8pWE*MDC7`DLV3z9dJ+K{iJUujcb}UraLs{2L3!$S|4~-Fs4Pf5mPJZ{1%fZRdEf<}|`$YEe z9`cIbU76WK3!(II$)jDeN<$~dYs2WX0{Q86ZyXHDld*}HSxO-XGdOpef_~RVn}gBm z;UuS8HM@E(S#V(7C3Sa>^Nz0Y1$k!+r?Rm%sOO!pYz@jb4*mZhXY+;*xf5Gs>M9Gj z8}<=L^znUtC)jsnkATt^#SRLtb_My>4!Ikf8{_`w+2PV;+8<9jN18@*?jr5j)|!R1 z$BDskecI=o2zMhFO(q_iC%)&j(LROAMDMbbR5Z#qC9WMK))vW3GXz`I<;^a+GDn&Opef_$eFfN{^#Z&Eyz9jVa@6T=lYQN zJvKUXW;h<2O8^cx<*4V+%|7}jA1lZ&wM(%q^xMl8agW|!*Cm^+(QK@z*&qNdA9mUL z(QreS1`UAu5v+!gS3I$Z`{Z)PR;!Xh5Q6 zJRLJ1kfT~vZ~^Zw?#wB?9{Ch>^cnU6duP=I(Ex=cF_3muE5y7G>&iF~(<>$^2n z8A96MC|&r*2_s=OOKOEABLqrPg9mAuW@j9YL*+YkK5s@xrKXb&Ly+x=D69F^t`$A@ zzHUyFMl%Cynb->Nx7Y-O?aU$Bt4uN=BY!fFf#s3tHZ=n~98c5FG31xi-u(|sVO&a55F%`6G$|yoAn!(b_R`qDM1mne zwdCrWo??KObC|CJo517Ayy#2V;f>AJDsA9-j*or#RU`q#VNq%~3uH;9<9Cyn_wLNhj_amVvWb2#Y+%*``ILV&QzI`r7;kMuGnyraYN!ev0*7;z zOv}UT?h|X*zpy+QmQ&}e9_>%25k*<@>4MxWKN|NfeFq$(Vc5fg=yhYZ0NsOp)7+eg zK)0{A;r#WfZkKBKEgZ6ve_QP!v8U_NP4?oDA2`EOGrh zgrE-fS}vQ)M#vgQ=bA>y3h5*7F36E!h0(WVz@Ysk?-2B(N6NZG(EN{Ic@1>me+d5U zg+GVkk6wVZXBg+nV=)1iiMpC^vh!WMkXxj|JK$P+SrcIuNZ4ZCp4$k@VN0n2)&V@0 zsyh_hrR(Hfy8j~ z-dRMTgAKwpnT|(;;b3jD4&5YakGa$Lb5hNn^@3?D$OmF-A7QjfMqbUm`m>KbQ}q2K z|10uBp)4$Dv=Lre2qg2Mn~6+i2Ga}MLxwD~M;itC@I1LCNB*<}?(JQz&U=*}cbTts z5n!1x&P9M_!WM)JGydtqRqx8ledE#Q{cF%v^`P(XYh}k!&lAx?iFO$aC5+abIz5krow&8=3rO>6YGGjiz~@%2awGHv_TH9 zIQb5jQ`d$;c&rsdg-LK|JIIw8d#=KT$dmKr7tl-sFifBjd;t8tjLh>z~ z3WkB<+7X^(A^#o&BysE;WTH!J)F?X-$~5tvZ=72X->;C9to&+?Ic+3U5%Q}CBl4@@ zNyBsA?CeB3mBVJmKdAXt92*hwK+Q~BgT2nYYMoPM$#WVqioB{ka=4L1+ImICDh=0? zWcDk_{GvZYMU!Y+dWF0`OBbnOuOMCWHkS^Z`ooy!lnuMUXfZl98*B>YU?T7hW|G1eVMR1%G{929(HVI%x`NTzd1`duScp2Ka3s6N0q`i5 zwi_il!r8cjF=>|G6N8X36WWOJBkx8H?or58hU`gOlppCsH^1P6sYIW<( zbff=Z4hN+6iGvOo9P**)VwdRz>7FqTlCduWeQ1_ZqcUZ~rQRGJ12s5hxYgte9B>^n z=HZYlGsn0$(yuf&co`4=$)nN!hjXBgYqcN;3l|Q!wn(ndP}Q(;Xt8f9-a}>AjGgo1 zL4{Qy|5%Wpi@TK9=Be^g;H;{fSFW_x{4HZFq}WA-ZG%E#B=kbtFQ`_tYm%%J&JI1E zC!OUvXbe{u1bRoVWYZ7{uW~qhs3cP$D_Qx%68^%1 zU0K9g7*!fBcR$epd-lkc%U3Y4C(GC}pl6ty+86{rls^tJy*oXy1TWjS;c@$|{R_kK35Y(HmTvqjiAA33z5Pb@ey_av1UmB?c6%Y*NfQj zMtmQBsFD_`@flB6_$H4qYXe_fx**jj`2IuSMMh*yA$BMd0jUeXO0l7v__$F5XRq^f z>UBPaJMp2LATT#{6AJ;GV?#IjyR>1k9*21xaQ$+J#Vp)|lW#^}9R&sShZ;Ar7*dz0 z9CobM&qhHhXZ@z=_*85Xj)-nD=cE{)%E;JM5Mb1X?7{=|(vJ+c1wq`ur$?h=DaPDb zfsnslwd}S%xt4;WRUpV4qXTh6)H(AEE7B2&EpSH=1R0X?;{lq?ppVno<1+r2)TdX# zkjR2-lOdoZ$SetZ$t+ls$pilDl%ECpqnNk)%ts44s`#>j{V_qSd~Krig(HAfABetO z&5$qts=}&b%>{eTS+S!w{E-$=P+{R|E}rVs(UlxYQC%MoI_JnI$_J{Gb|pM(=iMd0 z8+}V`Y(e~kde#W$AhrY~=vcGyR@!q{35ca->R0Ggs|Apbyd~P@aZlI0Gh&sMq5dDq zYmKJ*FA~ojnN*QXXIm71EhCj(BZ}Sai@OiXqS8S&6J~5(ecd3 zCF%A5nySW8MW%Pe*>L?3|Bzb$SJ~x17t4W?uSVC_9ym7~UkI$NAt*;)Wev6$-p`16 z?E8nIn@h}C>H-y3A+#`o#Dn*$(`yERBFTtwWMNs@iV`&o%P+M7(yZlPZRuE=rjAro z8+#x=&dkCh@Kx%tBVQfUt%vr^in8PdhrEnlVyulP!>BW6TW!@Zv_mQp8Oz0}n}Yr!Fm1_*pxou-Jv zL(@=U%K~pu&7${ECF?YepbeEWOahI9)Rz{1zehoon%7-Edu<3C!HCOhEyZ8Tm}4m3gHYbhb+F z=-uWYbml#Om2Uc-YErJ3!vvZEXIf-d3)YfH_9RVQunLgv%+r*#$xt;kgSBYOfswrA z5mf|6m(i%SkUSC}WQUaNHdU5#flf;veKPA-EOxcRgVWjMzzx@xrmkWJRXW?>*yzJU z>wZ+;6}s|e-0B7WBDlm@446@7WXuFnM%i>F%B~5EI>~#p;{_&wjj$4fc{Hd*GlpC@ zW`}2aL&~RdEHX#}4=AG_@|>JJFrGHb7B^fM5bto0jd1y@ER9V5VH^$PO77*$ZoMHS zbX=yh8c3gL9xWC)l3(k(A^6Z9C} zV+*01_+WPErmrM3|7oGyQ(2*pq=nc0GA>#sRn7^41aZbQ7Fy@B^X^zEGXxU!VTM3r z(aaD?uHz~idPAU3WF>AT5i)kb7O;%;tsu`RU=TrYA7_23lWQ z!Gl~W?Kv0BW^zvlFt~Z1X+7T`kNTUZhX%ZiJ>?NMQ2e_I{?bC;=+#4BJ?wEb&+a3a zvyNazXYn49f;vS`_HIu~L0P`cbYF`gEN%SAcRHR)MZu>5a?$JFI2e>SyO32905wZ` zv&~w3qGWc>T29~~TIaU?{g3Dp(Ix+a#_f%*L2uVh>swpnm90VPp4<_0g#YjH9Qq@# z(EDNx(99x7mBHgWZ5#Wuqk8QSJMul>jI?0EfLWMl*`T@cl<&T2(dJs$u^@|vz3ahn zecHG0y0|oU{byY)ugcE4ciBmLU-@P%>0{1 z>+4GlTz|<+@gv>%dXw2!(@pp0>3>UH>+g9dn{X<0{@{qt1zZ!eB`==fDZ$G7=RKqU zg7FV}ecE3=qr!r}!ORv`AOz7B7F~&XBfC`7+i6u>c9;)-h3&TO zoe1>HzYQn6T7jvY8_;g~!XK=gJYvJXNKcpyJ_AJ738Iw4vMU*pY&rN2gq)Ty(S#7U`K$EeGOm8pxRr#=xF66${le`5iE4V z`zUH0+vhj#F4!VO;UL&&$5K*~tJ>??05Sb4zFsp>LYsnm2HDdYBdFl_>JEVAHRSUN zjWPH|z~Yx{DFxFh2C1XV6h->99o0o&xW87QyKT_|AR6Q!dPot#({~i)c&cP?4Evic z;%L5=@V10=NFmLGp}9+7Xl^Gbdbeg_Xq2yp<7fV=L)`Pr9akCoL4ln)78mN#6%4qU z=%*NF#yb*?vP`tt;t`NaDlyVk`Mh1|mrA*no)%!KW;mfHN(!1$DPxEKG|V(a+b0XMC!Q;Sn_qOme!{ zC@WZoBRv^1OI8#2(&j1J?OlgJ{w*3g-E>`Hc$YT!7-9#=_Y+=lm^WMFT#`YcQXpz8 z9^YlaOjw0Fa@k^j3+);}FfE=CdBc1HX^{nKm>O3e2HF;Bo_jg5cU2`UJr$ zMI%q%-fT{P!-2XbRc6-(+1d#`!_5U9?=+)KzE+- zjjR;evcY(38^#+iu{v*+oQArrxrh4=2g++**;2mN(f(vAdE%{KlaKhu0eh(B1?omK zqyy!1L+FLQ$B+yC9o)0UjpWjz!4eDSIoR3`SBiT->z$_Tp-1Kr(|b&KrzzWMFc$V~ z(5NiNd&X$mNvtpFTu9RTM|{~gP1*i^p^*vOmj|4y=MY*URmIF&mT%a9;5(pCjSBekf*c?+OKDrXp^CELbZ_9*mP@7zVcSa`;J$*rr>S{bh+qy7Z?Hynr1ptnc) zOY$aaZc1XGMfME()Bf6-t?@>G+FR(&pSbzXqpn6i3k2GXxra5+Y{@aJEOeb&J>fCB zX~OZ$*JI|B?bio%*@oH)_K`f(lqhh`1#s$#O(96$@?@->e%0yh_gWEmpkyx$8)0~(C*khdC$0Esvs&I z!0~!kmX?rrq=#L@Z1+MBKj?Go!SpJ z;~RMDlY++WCyz_?4wNWQs*RD>RlDcdm_;j`d0&Oo<(?$i6EMTL^kC|KRsE6s^(zEa zfAaRqLiY-dp0^NVnisyGpcHD4LN7y{nnHflU;S|tXt*=wajm*Jp*UiHj3ot2QzViW zk}%_4nSCv}>Ipm+tszTdNHR-@3`rbmUWhbwjGZH5NeoF8y4ns&luuQ#l>~vBT>~6* z);KU|E1uvzR4Vv$E?=Q8P-RmbwnC2a3~fHhX@-Ovk~D{{5QX8emE{s0wz>j)O<+z; z+L%Q?n&iaPugc^emazEV3>kz@A;>5!;K=g)mlT{dN2U;nC@lv;++k@r|e%GisXP-U=p1-4OiE&bh^Wf=b7w;=C397ZsN z5K|;l_Tp>4)0HKJloabT}7?COvNxT@=c0{OYcqnrF)~jMd zk+=G5|E_(4+E+F=lUd%D&n_a2vgwR5q-V0AKM2gcPBe&21{)zr4u7am%T_75<)Ao( zmNRm`2@1f2kKno-7#j#ZkTgc_*Y7uA5PZ&FneVIAAZHoIvebCoZln6w|dfJ4w>H0p=} zhxkxcx=0WseFcQvM}m(KG=S|)&V0b{R`MvrjbNJ%1T0;U>fzNSnm9~Qda+OF_b0p8 zn|gCT^kTKbW^XbKaOE6+i`J5HoJ#TI`rak4ZJ`Um_HuRxpX+tuKV>iop@yS*0gQ$i zPTiZ_nO}-vbW4sDQHigcT;{L}mdV;Q>^>$|huUTiwIWu}=3Gj&E>7BL*aM?=zhD5C zImeV^#O-3IiE+F9eU>)#?pf-M9-O>BsVpVFetbLxg!0&JZFM|@-P%G2tej)1Y3W^F z{}6AR5)ztCPDoCB4~4wjKjhM754SdlleP8H{lnlVgi+X%H(^odeQq7F8rs=XM7CRAN04$cS+r=hvTi6bAzzf2n(ksb9%qzV4@wZv3V)pr8`B$$))jo6YAqPtg zv5nb#+Cva=#(9Ziu?!WJLn(=3v4Zgq28m)zN+IlamI&MsM2YNER(e=0mz*A!E4YDy z7G{xu?vO@fl7n@ZoCJ|)r(NaeRv_PiR36ES{_2xjTov7KZXw5eFA3>>W3WH^25FQ; z+X?!+-s~3v75Q+|b*9c+)dgQ#Ry;Fq5?CGzg^#L_ zsX6kM^AH@=&G*kAw-)dryP{R1gjt0s@+E_R|NRmU$Q z*V$DpcOF{MZb^t1*&+9OZcmM)GTA3Vr!hFQRX%->*5f{1)k1xbP!iV6t&YCMk6w$UiUXGn z`nChd>1@an7;b;UKO`!SCynBV4=@-;mV56IWH7Fj$~%yr8$lTSuYSQ zzf9vgwci5w%l62jE0#erR1es$ZZs9!NA@i5rkR&(=oOK? z2PxoHxCqM2FWxVTO(3WPd2p;~Mg+G~;}E1|u+5&eN#D5E^^A z=hRGSJ6DH;o8>JV7q8z>vg z0&#X!LaHH1en}H~_OqgV@=9`d@8u!olVAeXZIkySrTVC%QH2e~q&0$a7wm+*7b)SV zQyCJjho&U4QhkC#qq%A|)h9lUnCjElly}Rt5dANZvJZl)6&|2B5;M;)mt&p&+5W~x zALaq~2LV2XYdZNNQu15+3Gko$)sLBx1C~Y+56r#$_!H zr_n-mX@o^+GSjul8%%!%+b4kR@K8FVN_%qhzU z?;6X9HENvt2=qr9_pIII^5x5_&fx!t?7n|E++G`uHYUB!wR4Mv%%2(!9_)3l>yl3) zUSn0A1Gy0|LU9>fMZRdkq|`g<oX z+O|q1_L?WGkd>l75)tQ7)<`BY_j%34)DM;SAiuVn4L-PEguTcqf)e3-9EL?De0|P)d`k zB~!_lky81R92o<(C!8pCazV0DIMqnVg5Y(ED0vDg2ohw9BocBEBtC!`1j#kiu+ahU znj9%Eo^e0lACLN*r-xCVy-{~{CZit8x?Wlc9ld(UBaj})!0kSAIbh)CmW$5feIg%t z4|zrJuFQO(g$8C-l;>4l8|a@r+9j(rbaK2lj6N%npI-OI!Js@Dn|PVs4&-14=T1}5 z@7ic{FgiV)SFa@t4r~Wc#f~Q^xS8*gceZdU8(V{V-ucSbplsvNTRNLM!PdPr z4vD;>L+-@Zn7Ycs?Z&gY-Pm_zkHF3r#SSt+vmFkRCBNDscVly7+}}JqT$)V#Amje{DjI+oU&7I?Hp5O$$4Bk+{?%CkDgyX`geVx#No_7naD1JTy;y&uOE53X_T6 zWhbdAM}2e!E%dkk@s{UD#&frqTNP()s@SIY)eT$?3w!yE%9b zbHqX!Q9oO6%M+C?o}LJE7Mtxcb}l-$OK(BGK2KhWExpGsZ1y)sr&p(=>FN7=`?SB7 z?U+maAP-_ld|I0+5p z@ElX z&eoJQ02T$Y-6;SFO-I}7a9ampf#5!Qm%r~-^!-q4EM@8$b@-V2^4QvMie~=N&tp11 z2czK6)6{=|om#jI?J-+XCP(N|WYu=ci^Y*A`_Y2jlONWsK5(uNiQi+RGiQe5VY?-O zj}_#X+ND?)`mq2Buj`V{)@U}?(`*ocmJhpZ{b;!8ra=Q>egvx_ zE+gf!)Z*RhbZh%~T~cgOZXS4)QLCYZm94d18qgqo?C3(^V$g1jd5ade-a0N`Wo`>tY%C6f@Cq*wS=o^^cJO%y~%7(U2EKcMT!@_i47QP-JXJr zhAI`Qh@VrlxHuE z{YxYm0#r+`uIVWTXh|+D4sM4|$8O#_LQ&{bF$CO#+@t~d)q>?f^>Yox*UlIZ$MKNQ zBVP_16@ZKgOwQ$?MY6gNFb5~Kr;A)NY1_y`XN~h}OSHb}(!a>%-Nhjq{hb zW++nVcaxX*?#xV~>!wq(iGD9^a@GR*lz%i+BQH7_Z*9Xv%O!?ts0thchjW!o%fsvL z6KglTpgb6sQ|GH5?N6qXV}2R@bU|*GAC3E#z5@=?Fzn$#^tv%yfbK!QX>QI#pxf75 zacY?ZM`++^q)*!;v=L7Z|W0Ow#{cK*Hv>L`CH zU}1b7EKHL5VF$p3oNyyL87E(A8#je~-cd|YmgGrwlDyK8J^tu6WTIh^B5{{vp3~S| z0!DWam+Enqd0deCC%NEvw!hcG@KmycRQ^GNDy8)E4pSz;Dp|@*PTdh4C2#AHpQrK` zgZ^%iRx6PTN6P0L1DGNDC64@@X+LXWANf{6PDC?CmHyh9+lIBGnJ#gb2{T?7Xq(S? zO4v4=lRd@k)A$vYfSp-Ce%rYr(DO2iS{)JlTPccSVH79|9~1l2o()ch>t~j@ejP$k zhk7l2Fz&sqfv(UHhnZ#LWesC>3-a!Q90^t!eOm?$+E4QCK|gw=th)!z|M->H%+1X` z1b_CzpTqD+ec<7meK}7aiwUqy)YW{Go$um>+#(I$0oT&Ynh2{v!s6`oK;gNKVEkF+ z8Ov9}kQz23wtye@!k@zgE~(+UA4bL$3U8C`t;ulhP<4K{Y@Q#Er?(Hs8|S8d+U~4( zRpXz^=ZDCn6tjA59`-ndhW6I215FhUCKE=1^oqFs(9&X9dxx45(TY9W8jfkEv zdj)lhvQdA)3IoO>^|iM6JTU7|=|yYbkv{YSss2wdJs4r2JSR{t#YxIRY76>ugUi)cw#EMD7Chi8*#F87!^2$oU45is}RHvzU46L-MS&+x~Wr(sK}0oDOL zmU5<04i)?_ltXh|!HQs^qVJz?j5GH-gP7Uk=sml&r}FukxnWP)r@8qJgU z6y#_&ad3F`0A@dVhk?X!^4?iQpo0yHs`O^-# zw|BKV?^Sx-Wxm!$fMvos7Xg+DTM#bH_@@gZf92%9@o4k@HE1$4S=;DO?(cQ>&n-TK z>>IAHkG3c3KSFu%-r?ENCjXxm^nbQFeAUeFUO{#ZHwXNGH0SEnONINE&B2gOsNfd2 zOXFs|y{;Y5&4ESB!}Sc>Act3+e22@aYr`Ns)(W9`CgI6>@(XAt0T`$o1q^S8sB-+E zBbv8`%T61M9<&+ffr(z7Gs*TIi>GH*MwqJVO(%(H<;Nm9p61g939 zL%812Auq{My;_vO*@>B3dBHDx$c*7CZ2}d_x41etbB$U~ph~!jI(OEVc!KJ>kbDcL zf?;5S)km<4{Cfl($?SM@N4xz;blt5XX2OwV zu!UCzz$p8$A{sL|SV|PIg1riPGP;7%*m-Jn-&lw`qi`g<#sTmsm9`rtIJ{=zjO&Fl zBkzep$e0Oj#Q2eSqXzdVWGX}Uq%9KLoI!bn3l(&e$)yBjfyXD$3kj-Lx6VvA`VZ!C zKx&`lK=__N6kY5xogm#a#z8XnMW7GOGHO(&;MPtyRM4BFW1t4747ZwGfdj5X#ylKy zW#-t7H=2ETWMj#r(f)^XppI*`AjjTbHf|~&;MyX&Izv^%#-YW&sdx{ST~f>SjD`cb z!vG%(taSwkZ$5?{ZE9sH<#l&LVK6v6m-5i{w!S=0u9w@TJhYXuy=q zoIQ%(u~=f8LkygnWKXyxf4d-u^E8)7B{{?cCtKIqzb?qFaXV#D297(gqiGxR7R4{5 zfNjO0leZS+MR6MxQzkTT>A=7&`avrcSEE&e3B2=)lntFOnF_FfozAVNU#5SwuRChz z^W?WXu_7TTGe&~8#6Z@ul$p%St91QNhb*TG755|A2BwTuT=yx>`h+>vCkS;4zao>@ zM7t3|U?U*RV0M`$W^&_t@E7h%Kva?`kd>@_VF`a>!LBUgEQ~4*m%E>6;5&Qd%H=B< zzLO3_n6iagSZb+>{%ew9!PS6S@NwksmJ79%i4Pq2)_VIw1_$|Kj9PimHys-MNn z2NXB#eVWPSrlO_Q=ON&qA}@$G27wRdk3&rFPOn?4MhGw6+P^RypV%62EXil~R00pg z=8?Fx>a9&&Zw#%EKwPg%5`rF@31Xt71wHoAA`LA3E73uZPL~XN((51b;fRnj6gKI6#>-{cWyZQyH57o?94_sZ}hBQmBC zI~0k4)CFLr*w9UU+$e$JL0OVQrWu&reM6tWw3|_`tM;3=m((C`V;Aofo!1t>iUthTN`hQJT zhT{u?wKW9gSagh4Taf#H7TGiCPy1_U zAOqH)_7-~cCvLv;DBWIoKO^R`?;nP4E-_=N3sm0r;=y~>=`{mDkz_{MTwe) z^;?;%$R43LYkk(LK>H`!xJLGOplRwzHMOw^;^WLLECOHgbFk1eFH(xBdN`yfay9OF zbe#x83{y%}>EYP|7ndr~lb)8giBUq4!_1s~*+Jo|w3QA$i%I2^#|+(+udll)UqvI5 zs*Bu>bzj}c;Ko)Ei3K*=R64dUVm@5S)t&8?Jnt8C@)=( zqpteal}vMpVhvFI1D^-?eN5Q|%>+X3rZjBhYC{H8tPDBK7?L3bcDo(1u&HPVh^mU~ zS|DtP_7scH;kd&qj_eg}R5}kWXmA?ZSi&6&MNkEdtb5^L-K3uGke5l5;W`W_&y3kt zTlEX=kV-_xazX2+l*b56+i{!Oz(eK$AcLl~c1(Qdg*zs8zA=WH8OB*OZhJ^f^)&Aa zJRDXz65)=XHFw>X$m@LZa3BG)rectfyV_D?D4KZPMPTs|JR`=5ifPa&mV|Ou>VS8v z1oFggbY0kKiWodJ4HdR5@TLv3IQtC6&!#-TN24J1rG?+`Q9!Hac^yF34t*7zafq|j z;rrz;bV!CmEA&Nq<_sIIe6m8Rpt6_iXR>$BuoTM9R(L?3u{jZE3Iou3EDz7RarO^| z=c=ZUvmCIqY*ETx@ys=W9_#TeW3Px#BecO`Aq7|{;GTp{dsZ zthb=!b#cf5nc&w&Q>EA71l3t-9`Kamw5X(4$eK3D~)h%8OX8F8KNl22ycip8#0 zcyQWC4%|?hx{4X}<7|IpqYn?Q`-2G0B+!*F<5n-|7r`aQV!(_tBV#6rGRh`DI%mer z9buC9X2%OmcN$?O2J>i8iDnEXvozne=nW~K#<9pC2|S>Te#moj^1yi7EL+?VqaFE| zugZAKl*~Cu$1|u(o6Avb|`k%ed)(F!a zp;CwZMpg)?(^p|dVp_`z`=pI`L-NXm8LpOb;+}ChVWraSH7M@qjxM-PFIVS`X z#2L>RBe#|p0txysLm;teW(XwLVMRk*}FX{-(;bP z*#s40PrlRfOezXK4Umgo_r}4Xyg3ZaGfQH#&02hYVRp@0PT(N8PIp^puRZrxkbgnr z_Quwrx9g_$t*!CO)}VAx?ua=;RGvFnv18{!pg-~oy^UEX90#+f&YE+bR`tjcg4;%| z6>Uo*`JQjGRPfGW7Uo$tXl^{^yKh>qxwbc?gDfm26@Ax(;rg_1-*xN97`E#_>tcCT zcGkViPSPXy46I3!A4~&-OCCFDt-)B^p4e_5JLF$|eM^Ixf757veTjkVxfEMeXR)c% zJ1gDfk#2mw$?Ws#rhD`Bza_5q_q>x$I2EVwsKe&&BFrHh(cyt>VgPE((n;Pw?-~6U zyw_tFHv1c+)2q`Fj0*1U)Bfri6&CyrX11^bA&9Q9=t|5R*`>O5FDq)~4)Y|!Hv;|g zZ^H?%R$waU2DDqgz|E8vU|m8si#*)L3MuP7_b#u0@XYlJ)&^b0un$O_e(SR65JB>T zgaU+Ft&1rW_OX)xt@GP zYs39@YICCH-pqO?6-X&o#x`zn3jbk}#|z>fgXC~ws}G#(kB5U}qcdlQ<6*mL-VYVz zrKw?yi%wZ0b+|4^^O8n!m;7Ca43H@`);(;+(^ihtGu|grNaS;>wC5WCP>^3lX7%KB zf4x7hBii!BJ*ms{rl$2zgUcdJ$IQc5RUNcZa zn}T`<+0z*#C_s)lpp$M`wm1&*`Gm$8{32lS%e9oMI8=0*qDY_CqB>nf!!O)lE70Ax zXaNun@((>EAqsxr={pK?JXNwchW*VJaWvmbcw53bq>!dyXp~!f*LrJjmasaDcL_kv z?c_x7)+|7cLK;mLkiY5>_xy6lRYtBQurtTvLOr^I0XGx<6vNE;u)TbPQn2WyKr@9y(QNzF_r+baEf@L_;lOeNYHE}O(o}%5} zbqM6&qLI^0*VV+kG=qX2v_v+&{C>g<4)bPfoJ%qYR0>3G#p9bPQRJroDk@7Rl9JxglOuS8yt5T8-d6 zI6kFCMh2ZEc>+bo8@b)+*hIcgriBZAzGjK7G2Uzum-LwyTjYHxfrx@gK@cU5MnO~> zilDXNOagEXzmAe=Lqnm!92x{JHxMWYi6e$SL9j~E$dk7ln$y3qZb_B#aFUfAt2`+! z5w=u7N(U}W@4(C2b?npiOSQBJcZ(G19ptv&N>Ykcw|26}6?wupvQlKr2IH-5z+x@w zOtKBLP(V8^F8;GQZpbB%Y`H1o2K&w$orN?Af4ES&a9L(X^9TU(&fu zr$L>*?3<=+|2~+(N+T1tFReoPybxtjY}zS+hK3*zAN5U9w(m@!C6>dAYTU$7bHy|Y zSQDB%Y5cBH@&YVDP`e}Y-$_=DZK@Fmr*em484_eh+NuCBcuKV`ZpYh(4elS~r^BTIlM(9N4rR-LgBkm2Gw+2aGYb8m=%~Lr$&7FKro&#hWZR7J z)m0c1Di(|zO`eufoIL3DVf*DC|EBOvn+@0__g%TWG)+7=e!jLGG5dUN@;*(II@0eA z(C*khdC$0Esvs&I@&@C$Q|*nJ83KK7eYjGKtk-lIV@unu#yz+)hy|WedqHdDKIQ=B zaZg3|Q2~KpWgrq6a!A-2Kt=`|3HpLj=|JYNn`~a*egYG)^7biiJ$cIjySxDU1S!o7 z24CYFc@V4}u{_2mLK*OCWk89P<3B?fyaLIC$N()Jt z@vh81%hK^q7CRl1%+hhGgd_@GZHFYvrz+S=fp3(kH36}3vfAhY|kI|o4N^lqdsr)Hp z<>vom?p@$4J+At|x!t4Ds3aNt`Zc*G;9wrk^`ocnyT>nSBw3a%$+AY4-)`TTz8W>^ z$GE#^tRcUR6$}PK7BH_Y0m3dMWZCR1u)85y3~V5~z&as7!aKkMiwO{xK!6Z7kgD&i zud2SU>YO^KzPjDx-%qc0&sXPN=lsv9s(|*@pNNhN2L7DGzflTfD;H^-w@Ozv24CB} zBi+V6xVfw6lM-a#q~liEydAoNzML6Pa8;3tC%E6RPL$R$O~(^#W2eLu5^TcvN1%Ti z_9l+=q~i&dw5uc~A=hzi?2&XlK@$?R(i2Hf9JX?#C(h>*(-Zd#MS4=-R$|GDI#iTc zL(>rj>k^9F_R5k@A;H9iM9fim&@E&lMj9lDN0P)u%puV1 z`S#hQcE*;Rh&e=6Q#W--iQ0}aGx4!xJlYOGRCq#qO7eBdj%%sNpODDcS07W~NA743 zY-~z91yw?HrQ4S=x23S6S3YAz1Z~p_0orM?V65J}P7H{J4SFd^8~La`t+h)*XpogtE$pgx zMw>OI160Wexi2S!jZ8f}kk+1u+w_Xj`sCil(`@a|zXuG~i6=4wfvq*C_rI`3B2Kwcx{*j^jX?O~!~t$lhifYE7=* zT5>7by3uqnuoJ|9rR!Eka+bMGD9@3%8%`7RcKhcG*f2Rc`HH00h*|x?gN2+>KER`G zeZ&%uw(nyIx_OoH}%9$wKHTNeKdgft;T6<|FNm~t|za^zE@RJJPgm+@ukJi z;>uC?O!?0gjLm%NMEua`nB#}v(=Ho&c>ij5RsOBlUhwgtVfYf?8NVhUv$`)@9PI1x zaac?@%nvt@Y9A?k`G2DDhSVq5DBeCl2B$(-D~8f=&}2pXe8-S61!=i_OZ2w zBBHnN5I>tCto+c5xUY3sk@{-HJCdn2w)3n16!ocl^UW9 z`qT;@HlIs|hwT^e!9YMp;`dgBcQMI^z9lEYr%d3cIVdoXukQ>{=x2C zCe+>2kG|zy_hE;uofD^@PAbd&L$T(Vu}@%cNF5(_4^tbUAy$Y44ic*iW9~eXmuVhe z+do`vs)NKG@rUvI=>t~abE;B{@~6Ew1sR(k$>jEEdw*}Od}VI!;rY$Q!J|?;Yxc&Z zV&b*u3(HzD!uJ__L#|-OhZsR|8uX0#;rRP|35->Ebbz(+ohaTK(!k?t zF^r0D&7>dM2?=yFUs~Nx|GbNLncL}C*_fO)l)A6((C|R>hE;@?NBe@9ib~?6kQ*m6H+4g0X z*SYsC#P?QnLmRdJU`5(l$Aqd+3%ZTieBT(WvMdUfp=%=%f)A1 zxVwAj**~xT*J=D`wP&s>053}na=3k@n!JkLTTEiuWUD5$9Ajnk)A^+%H^g#RT!Pr6 zZ=;eV?x-^8HI?ynutWoYc~Zy@OBU;JX!Vk{+yYjq%u*tQ6=*Fdi(alTz9mz_>qt>L zZ&i4|o(mwj5Ax)RcE0n+_hTbXAz$*#t}lDTLGDi%lJ7m0BQbXL8T80h1i?#|_Vz;hPX z{x34YmsZkMTAc-MFrSs8>SaMc%@>0fo}NjYI4eGxG21E3#7n62idLo5*-~^H-Wr!v z(%)_W+1z`Aw4ExIZqcQE{P4pEM?Qqgg86cmv8AMK8k|>!h0OAmu?4X~k8@{xUgiNP zcZl0B+*V!@j-9d)KajDEb$w3SjX4pTSHNZFn=P5NerI>NCW4BISA1ut`k`=>G8ZfU z^+ENbz#Zma+ssLn9++2TI<}hV9k2CsrL1b|!)h%z#{4%krL1)-ZOWQ{2}!Q5w4p2* zR+gHonx!+Qs`YPLCr(RMe>PKEXQ`^vx-5liC`jZK?JmbvLCpRPqjgt%pLgT&vI8zd2$TT2Q<|N20kTD6e-%Oj0 zCY)Hpm=}xy)B_b zfQBNf)SxiP(7$d~TvDkMa@vrsD``L7)qA(LdXIB~mnJ(BgW0-%MGgIKZtrbvUs)Ws z=~gwrt}RO{>=?CzOhZc7TZy8Eol5TYCVfEHflK9FM@vht+ zqq{6s+;~YfS9|xxJI!2er*n(~sKpM%WW|@Qi1+1|hCE?#b+LAMG(R}1B;rTXQzs=2 zD3UHu5?hO%qd6>z6SE6Q63?9$vE#JtJ*BXT?ro1N)95YJ;+7M{mPK-58pr4fa41;m z3vv31?X73!^TMm*<16C5xx>Pnw|92dZp?SCBl*=W=L_Gx+GbuOoUtfD=ws{4+KGBg z)SgLm`u_D8$Cn+$mfR7)d|G@UxAoqCb8o)8edW^8_R*E6yLZ?Av$vdZnLo*sIb?oX zTO)t4msnFfKEpC28?c3~h%a3c@5yapkIbLl-o3tS?{e~!rduvN{{7Yb$+VV3_jdg+ zz2*9HZ+VKmMK<=T02AM_Dq6sTVK;qhXMg`-Yx{<-P4C~nv8)J_eXrE^brL}ai!DFC zgaXwA*0Zz@4bZ4RzhqBM9;5iw5^%yNws#g?nK17xdf5%L66eyKRb_5Z$nl?}?Q1)W zLzVGYPS7EYg#NAgy;tn|FnY1nJ@>n(lHp4XCGwwLA4dn*> z8QPDpioK=L9PFn-BLGWQc5msbsZ2L*G(f%|QS=bnHBSI@KfE&EnO9f&_1>L9)b);< z4odQ#o7+lI<*tk3Y^QswfBMv^Q}@t<<&Uh0_b<&7<=k9(5y-M-oPPtZWUDi0%BJ{p z%M_(NbEFRFUGC9u8R^}h+Pu4Tw14eE-BMhl-7>Z%nZIRlAHYv#8dQzMF4Y%$%lo=C zk17ySO!ZHfX{vILp*%~vWR3NXc?+f8_d&>><x1#qMIh7nn>#aJfEQ{&`bt;y<)H zA6H4uAu=+Rwq>T$-YeeSeKONjnzKHABJE?V;_W&0W2Vs_O-*PO8-Yo*7Dvaz3aX5F zwSPS7fQ_{yz9O?0>eFYr6&!u5hJYtJ#;PU)Ph17Y8q`&z8OHS-{>ri5YN7}+1Nok$ z1_v6GXvZTRcM47Xc_#-jYo!pFG6M#h`hFwz)G^t|F-RafU3UDshMqAMzq6Wm@2*Xi zWq|#I)|EdvMkRlfrA?3>j|^PT2JQXZSKC~Z5Eg)* z99zbY(&mDsFtO#lpVpJDuancEmSlIrif5ZzsVr zjB44h*Zefc(31VMQQR?GAbQuTJSvu9Zox7cnEYBocE@n%9^4S^ITb@A$ykV2Ufiou4V&$abvdod+cS0OPuwo++`X1x3OknAoTZ% z_jKRY5<;(sPAew09__)bGvWv0PcwPA=+?phHTk~MS~1m7HaN^2&I>FoFHd(rx;Y#4 z1~g2)`s#i2!y_xP=mdxO!H#%~U6i+0or7vbBf!Ht`@4+3vXHk2#a?qmm7v(KU&UEV z;<4%lZ#apsY$;qznH_OVzsuq7%u1=cpeERqhK3h2sioJ-LR}&DTitpuA7YEW ztsOTsMD|m}*8<{0}QR zb}D;=RC`5AS4z3ho90a9StXz=D86w;yj9s7XZqVCO*@f0j?`Xm#eiAHeF=~N;9}y} zJK|BcU|j3hn>qAYYqrp38Q?$mdPZII$)}`X(|_61ynV|5qLZ)}(XXm@ro^4MpvZk9 z)_>&H$|)K>>rceaQaIy2$TIFBE4DO)6MUqDGmTce+ z-QZiY-`o)w$qrLzD=>rhdDYCuXI|{-+k;Mj&huX`Pjx&e|GG>5b&veZJ>Y@8eR*1Z zD<{DUQK#i*Z@%LjIbaRZ2d?cd8zQV660SDDMxyWq-VMqx7B9T;g6_)M;)&Rf{Fl4r zU-zgYm5%)sWY{;9ho$}=+PkqhIC^YxuzUSzuC6=lzQFrYZ$zAAy8#EUL;3wJP*QsPqKo?ZzQtc0W^1ARk2F`TCOl$&6!=M$-w1IJ=2nRY4x zE18}W^8E9?yl`I^P!YDO={-2wt4RNx{)-;%zojGYYfYR~o_>HiPToC|L^^rj)L~2q z2aN0R=wN$mv9)=)BX5&bcg$URrXrPytEr>op`plA;_q;JA2iymAV0fW^d%QxWbOnM ze`a1oq>8X3WVS)?ED*^YR6-2Prf`6*qnpqQv z-FBAo(ZNZFfpBnA@7S~ogsfgtzAdYl)6{LKPG~JDK=@5<{LVC0G89YgJS?coN;sW9 zxwb(Maf`28X_q>CI7wUSXGnM$vI*lT;=5N`Ry($%kfp3_7XyQqBwGpOvXIuG`>$5S z18rPSo08n@%$r;7lHcMWbB?XjdR^_rJz0Q))#om@#u_G20&b?y-HN3OpgK>%TPiB* zGce@PBmRgpNP_S!ZDLjRsIzcRLYgYhcdr*?_fNl5H2{YC^HdcCg4J3mB zhtvHyLZv{o8oVODqLnYIz+R4Y*=p{b2d|zHr!TC?>$oB#RpNU&ms5Iip`azWsX=2? zSro&VZwQlIsJHFYZhy?~yLo12Hj)-%sk#C(D92_+YQpe_rRbZ<_pmz{9_Ojqz4sxt z8PO-%YZ8zZ`t8QrMlZ8B{s zk+x9fC$FL;RhRakI@+B-+eQM``>X`$JpVm*vn!-Q#d8it7UE+}ABJr-=uEQ4F$Eqp z4}^J<_!@Q&qQBycm})_f4JmaK&ZKxE|-CqfO%9cElTbMY#@7J7Vs8C3!iAHwZCzy_O=> z)|$VeBVNyA(7ZBH+bt8yexw(DTAlM^+D4r&3kATGscYgD7@_{E0TL)>|Md#jBn)N2 zOb`MFTB)ViQh4poPW?A4;zFuXRr^Tm0MlbSF0Lhd%xG>Sr^4IFm$0`HL!g%swh(r& zm6@41(irxo7ktuGk}0sAtoFu2KVeanEJhbWmzKrd=XwBSS3Lj1^NRV-0-IHTBbv%z zU*T@IhP0_4Km;o!H9ls z>)4s!TpVnkUuzsyDtBbA*EiFg1eV2DvlCFmq*Re&vW=z+IrW&yry(Hd4XqDi%ggea z`LrRZ+TYXDP4lTjxLkvx6TabB|&hYtMA2xzDPWQr=+ z9gPZ4R<)Xu(+zNvo|$Km*BLW%{~fywv_jP^m>;;h&JYf`Jsv~HCM4Uo`u+ra%d2)e zzh=a*38Ui7E2_cIly!|J{U0_FWuhil=mm@BKYL7PuMt%yrnM{+EB=d{0i0$A0y4epg&38E9aZoo5bc>2IsJ>CG^k~ zd-My7iXHoz1(mr(NV0BF?bzc_=at&?6F`F^W>6UoW-TdkVb-X%)}hqmrk(na3Ut+o zRK1_WbW=z?GCz*NtQiDnD8~XW2W?fi-Z9No**{_)%&%2DOWo*cGrx)w*F3VL@X|b( z8*W%Q1mnmew*2Ptt$_=Vhx)?fz_qbVX+b>;?(8Atl-$lnSbJ8g$9tXeY0SyXtV+nM zWnQIHyTE;W4bQ4&O@FDHG&^_kCHmrw)z7)><$8?5dwLNWrC1gJYe&3>MuL|jc@{M6 zakwG%XKm(LPkqGqbj0~4-Mg0;TU*|1esph&N5oTQ$1`SCz^ z*2*?yU22oCMpJI=9745ZdS5^>Ww{vFEad|mUbBqDQRQkWJiWNdQ{jqS4!(@76FJ0Q zZrrsb`=wPL@Vtw6ncL}Bo;yQDk2~rt6jPRUN^I(qGlvSz9^o1`R0X##D6#Qo9Epn* zO7NS`oIhZNK<|oH zF9nD%Wd~M=FE2M}C1q}CbqFcEsPchk(f0~3GkSsej0<<$<3lTFJL+(1x)V$NR`t^7 zH)|hTJi3ptVn~zy6Bf+yoyBrX+w#97a7K=7f(~lYQcC{35}Amg9}tg+ouT=Q_#1&i zAu@p`kXf`NU0Nmf!HlU*m#_s?(q17Z*%reY;Kr-tedWw5DY)k!Ennc_^)ts&J%qxuv-gY?(&YD^yDM zWXhC;$_~C=2o^S7l7@A3ouTLrtO0TyFQEJ#JxK1e!U~85sQH7h} zkQv&1{7*?meL+f+CV)R8T4f+Tpln@GblQ*;(yBZZM3uh&OldCu0}@NlHE9e_s)AP&rw6cLE@{$Yo7;4RRr;a!>Q$S2t=wZ z+-)s(j^-72iScnMN)}(c+GaN@bl1J@aizkZ_uxsA4YYOi_GA>wLR@;ab_e4!Hm#oz zT-Hw)=9{;7cGij`O_jy@{_aPHM)p2F5;idqJP3l z88H=SdzSC6ZmU;XH_c1C%dnhZSQQVZr#tuG+?(%iU%7O&E#E%x-d+38-kZ|DTZD-= z34qKb+N@uwyG>wG%C0cQX4tK#PZ?|;)R_7V9I@UFb~{bm&wZmEsvrvgtA zvU%&?u9@uR@&f-UHa1cAQoZGv4rVd`H?f@`4z^FJ+*$+J0zP7VCl&>MKTItY6M!uvKzH{IH-l3_3UR;4BcU=@`JMz7` z)2B||bHAHrb)ZY>)KSkCuBMMEsyUwdj`I7KI&Kp;3XyQNAYUo;Zq&dc_oWYCpC2r? z?%#gusl~x!`Ec8}1@3!ErO(0xSQNv-LT}(%{P2o+Hz(x-BpAexq1lyeD(iWd>7)y{ zGQG40(NA`S_W*eUEr(a;JM#mbx%Do7kfWJ8f?SwZ#4*$$;$kTToSBRD#$3EI9l_`Jf+{?Uk zn3ZZQ=g>z1PSo4RH0v^w7T=f0?}~P4ZbN%7XQ$%i5N;2;)%c?LuD~Nq3GqQ&nG9Vr zYd%!n$2|6-<^YzF&m@(5KN`4-EJ0T05GCoPa8y1=Nqck$T!+oV@Wm{p>_DsGy^k z#!Q3|;cmCO(mT-&vh}ckFS)q$`qznDUNM=_Mhw;tX)ToB4qP}f))!7BoFTOrI%Gi* zE8|SG?X(74|Lq34dO+OQy_XDK)q{)`MU```ijRsuoD-^f8~ulhw5h5O@J70K3m1|A zuL^m}^VM_;ELst!K{fH8lYjAO_$?H3bfmy)hsI+$ck9EW=X#%f$Zr%a= zF;-;9#8pw>AqHslh`MBkCOU7ECd;75(4*z8zUR^+pJax(mOU65&>nL;&u5m;$NWmwL64p&A3e?h7{5~r`h;{^;2-8S^6k;cYejjD9|4XKaIwXQ!7HA zP(<#|ZcyM^JaQL&NF6=o_Zv$Un?s|48?_K)OmYt*EEU>^!g4+Jd|v!tQP>$lu9^Z* zuB-q=k*dpT{N))V$GQS&QXK}Ym59%;^l|YMrK%}dzEoXS0hz^xEuIm78QwfE?QB=4 zlpmk(JyWGiKF+e@4w5-j&J13@9)B;{dMk<(kmS9%-WzCk(8pV}dsd=8 za`7hz#89K9j%5VSB1T8pdp6GRJCdlK)-{$b|D_COcvx5wc){o_`Pf|E6(RqIQCP8| z)8tA996_rR<%w%DWmV6^73WisT=R7C`a$I{9r2F?Z{%5}J~%&mdU5b%=(hCFt4^YP z=EdIGJG!S;)cl<1fBEAw%BlaYmj94%RGitovNJzC?7mWbe)n?=q^3E)^AUNZY;$2& zb#s9`5UgZH9mGdMUY&ULq+R#4`n&(yoLPFU{%6YA_dId-iG%CPlo#EVSDwNsQkHf9 z;n5R?7UaKQI=G_#{f-ndh1L&Iz$q zA)gmSz~VY(herq7TZ^sDLzOnK75hUaS!foirfhiclSdLz;5sF;yOywYb=>gY`Q)1IfI{k_HJ)B8u1^t~VpRM1}2wrcP8?es4htvn^>0DioF-oicAeEHNFlp_8Vcw#)h%CH20RhEBjlE zgS8#?jPK_8wfWJ}V((~se{bz@vGdg0hp#UVZmtC_+eFm&83x}W-rjvA$YAm+5Tm~v zW^`g0eK>{Di`(_ ze`W#b6_B`P^*x5ww~I%*Z%bemT3h^9n8}%8@~u3rs)ZnlT82Jg7`i0h(S3U&L*P1q zKn38d8-`h?uND{YE!|g42&@T7;nx5=usK(3XWrgBS{&>ywzehmPxAAG9hkZX#3dBbY{C_bXaNf1 z{+$rfIO{;%%05BRw{5oN6$p)Ns}@`E{Fzc9u)hv*Chr6fa%NMk`y)PE;|$c`yNqK# zBHq@0h~k*E4+D&u%^lo(ZC(|#R82^m*l5V6 znEr}anI=&5+dFP#C5U4nD;(H7iEhbycUVfD>soAl9f(_?&K=m?39oZODnQ>g_ROwx z`D;O319k4e=2Li`8=xdlX2q_56DbiuojbHS7GCEP*Tu3ecAcBPCS(V|-aNEi`l}^KhYRaA@-_yv|);OOB>i zo9z0Ru&R3dD#)45vG6*#u|`~j?1GoN0;D@&?;YB_DrTujJqDwW36{!P2jUi}bBFZY z3ZVjvya~A#ZzYIhpw1oI+zQW|2aGxwpuzemJD+B(4&}{5n^)m^Gq6_NXW6>tt^~+c zKzpgpsqnqjqXVpSO$N!{GwEwUTmlhlWb-CGLIo*^Yn2_Lcgk*MYI4ZpnoR>EY9bn z+yZs($mUykojV}ZxyFy#^(^J<0Ey!mCRiUP%>2Kj%`kb@1-6cV4Vw4VEvZ8*^<|W_R_J&NY6^-ZSZIKwJV5YHIT)JVFI2 z0JREGVLg@|xmfE%{yMe!5z}9h3e;$|z2hRQ01^fL)#gY{e@&}Ct!CT%F-`%L6;5rQ zM7LzU3mj`d*R|y6`cR!awYd{s=Yp$&zH986UC;8@f~*?o=T2=th1avMnP6cEOvzCL{%*&Yjx4E4DWWDzIRkYq(C>o2hF+TmpOZ z%;sFNh!&s#^zVeNm-BgmNwYyfb?(gOTX>y2BGkDtB1jke)Xwb^R)y-^na#2AI=8U~ zsKF?A?E06uGPL*3Y+ef9)w38u_gAIhy}^xO(r4U4=9ITUXtK(bKYJhQnK zo;Qyfb*_oH?7W$?3dAXZVa#k^h3Cxx4WQ4mRmxoeVARH4U@x87oC@DdJ%YeGm%WA3 zKON-~+DmQTgh!|#DWFzkSnGrSYV#wezakZAMt?fw|=kn>*okE=UFFyX^9ozXZUn zF)pZ|yJ7Pwyv_}fM5=RRgz+1`iIhP7+zp#!;dL%?-9Wa*E^pJ9iisP-pf_yZ72BHw zC0ek~jgit%*u05rL)F8E&ADO`EwB>k-)$lqXB|MIP@TJB^DVs29TVzY!y|icneugr zGvIvghRw0?I=8Vls6k+T6F&RwI+wUIXsI@@idiaBk9l>j(PF~h%UK7Ir;rb3za)6a zCq21B$Owd9!*D`X#9Iks`8Oe!ZFYrc&J#wZ3lL%Lm|X!gR)@7~Q!6}m1{MR|HeI{i zl>ou2cMQS?ZSyL8KlKO!YhA-Ed*7t50kQcz5Suo6!oyUMgt%VWVT!j7#4hNwHcMjq zEK-LWy|(vFWEF^2Fe2MjiRrVEBGjw*FBP&5HgT`$0CAu-|O#-QR0Wxe%Ze`@r z^~Jx3eAgyXc){z+y|!H3f7z8Ue=R_~;<*(dGTY1wuXqEboKVG^NQw9oKm_YH+rmp; z;=+Nv3($%+dUj2mObfV2uiFePwnqo5v|!C^f?L8KOAb%?cJfmpNI7GCx?76)|*5MaHOUG@@JhK<25Ob$XVCu$vNp6S6NB1p>;U*KL-C=g?C|-D~15JBQ}10=NsTU7KOy zIW#~7=(mPlwidZ7K`cXc@4C&d@IBQdWC?XIeGQ0BXis%6PRGMkkOWY#02wwOv-euo z`k>E}Z%y}><5#4X+4jB+uMYa`z-CKKpN&+b?OippK+Jf+J=eK6oo>vvLJ-e|TN!P9 z9e{7J?sYCu$Ln6#m}_K!{%hhgyYA($1+fj)y#t$B;dO6-6jJVEm}gHkr>+Qz0IGWj zHrvAMULrw(yo+7drY{9>Z{jCd_d55k$M)zzp@#c9yY8KaZYU0}0q_flHJg3K!die$ z(8mEXjM({uSV0j$>kFvvb#7Y6>)z>7>t4#&0TBwQ?scwE$Lrq4;*bojuGsZ0bp?n; zD4}%jM~@jRQigeTZ^G8g`8>d-kuR8KITxkJvMhw2K>RhEP(r%JTM6JD%%Ppz(($@? z#;AJ%B8&q1>>Qf0I;>rbQ{mbL7K{5WTf5wq0KtNLs&k(@zNdPGETQhDuK}?MhAEpt z;bAIB0;pGZnBsjN;L>;r?6Xv?%A*IaJmYUlwf9+gb=YTZw#4+=mSrh_>=_U?WL;o_ zIJ8(5-I(&mxA1|(tQNPTwQ%U=s(8>)Mqo7M5U zH$cjg>)zBAArU}zuXE)(UiT6S!tyS50i3=jBnAkZ9@)-dY>y5Ugu$BB_&U4poh_~I zbuM6!g|z^iaUV|zYn*i;twMFLbMrc0_s*7D_fozNu?E$>&XwwT-P>3k)FHdvrG6OT z%xnzc9_-wg9y3;?JoD<_7(IND#z=y)CxBU&b8&hs%R<-<#NUL1h_^sYqF@f~+@6lt zy&H_W*Th|RuFP2l;2G4ebDcV_U4RJCZ`s=AE&y<9{TJF(o%_}CJ=LRV33V@h4S-)T zOgR^;<6$aD0;pGG#_T4t~`w`@i8gRDiEvCob#B9RRKiAb=%%~ z(<*G(m~AJJZp^hp5YL5M8FPFcNW;*avva{ZUiZ2hu8{%im|gDjmjJjm-UQA$k1ckE z*S!InC^?RyfPaEFbwxlH;GDB_1v_5%5(x_AT>wc|IP7{jnHCV69$TC%wnqmFwVcE+9y3;?PV?&CHpa?X2Vxh>vYd<5W0es?MoRom$i7$wsEq?EQU+S=$$UB#GI$>wd^^{-k)K`*=F%;8Fnh;x{0sZHDzAlWDN{1!yoD=)c#9 z;W6=O_u&*VK#Pmtg&4KH!M7za3JTmrs}CBco)C|BA4_ElTnLaV?7KE&;`%O900UH5 z-DYrZczFT&u5;l!*LTwjQ_FA$pT&8kpSk_&h6`w+EDjo4p>Au*Rw|gR*;x<-SYQd76GvD+Rh=@cj3Y`J`kP3cj4t> z-z8&Oo^bJ^G^>wBmI7=WHffthF-%(T3`kuJ&|yW<&ZMIRfOYZAW>%59*wu1vy>vh3 z7tU-V1XLH#Y_=7tivw%&@-BW|oVp}nW?)_HT z0oBFMtgB>0JpHo+N>*TvPg{vH;YqDld}-SFjN;i*R)5gCWMYa zEH>Irsg8I{K}>%y0P#4twV!W5JlFR3cNPb0NP2aj0dY3Ov^x%gIPs4F6q|T0eh?zv z=1?Py^tjVKX!v-8cwKiR$_MyGfEV$%Ar@>l^a{(u&i>)yTH4+=F`T?w^t-QYZ~}f< zd}o-z!FhXU=NB{x1chL%{emIpym)Q*)gfZQk3f?Aj(>?u|{CltK z{vQty9+qP4y~zLB=6dj_gWykx!Jm$TKOF~uItl)C8vN-j_|uKvi$h$T5C3>Q{Nq9R z$HVZCN8ulj!#|#ce>@HUcozO~n9D(!%R!jSL72-yn9D(!%R!jSL72-yn9D(!%R!jS zVVKKdn9E_9%VC(yVVKKdn9E_9%VC(yVVKKdn9EU^%TbuiQJBk7n9EU^%TbuiQJBk7 zn9EU^%TbuiahS_-n9FgP%W;^?ahS_-n9FgP%W;^?ahS_-n9E6+%So8aNtnw?n9E6+ z%So8aNtnw?n9E6+%So8aX_(7tn9FIH%W0U)X_(7tn9FIH%W0U)X_(7tn9Et1%UPJq zS(wXNn9Et1%UPJqS(wXNn9Et1%UPJqjWCxRVJp8!9f0zaPtKL_|-2l!nF_+1D1T?hDG2l!nF z_+1D1T?hDG2l!nF_>~D!Q13D$3jZA7SLQ_F-v{`WSyA}+0e)p(6#jjHUzr&Ne_y6X z;hzKi%H$~g`vAW(JqrImz^_b@!oLsjD^sNK?*shGBq{j&GD`~o9N<^xN#WlI_?4Mb z`1b*RWv&$deSlw?Ed_sHrc2?U1N_Q_Dg65YzcOVC|31L4Oq#;K5AZ9~rtt3r{K~{B z`1>+*3jZA7SLROP-v{`W*;DxU0e)rv6#jjHUztG#e_y6h;hzKi$|NfM`vAW(jSBxh zz^_cC!oLsjD^scP?*shGWGeXkGMftj9N<^xQ{mqS_>~z|`1b*RWlk0TeSlw?RRw=v zrd8pe1N_RwD*XEZzcRH7|31L4Os>Me5AZ9~tMKoK`IU%R|E%ad+Fd>Oqn-ZC#Pc>w ze0KBj>G`$AVfPMk`-R)Ot7ql^X>c$hPS2m+KJ2cXJ0;#%hp+D2-#a|&-d6mhf2X)@ z>*(gSMR(==YQN`w`m8u};dFQ9Y)6hR+*UN;yLVr_^XzK3_~C`lDe*<(GhCqP()Hc$ zs{C8eN6hL!DwhW!xBh2{JDyoAu5E5@?@AtCd8#k@d18C(*@Dc9N%3y+LU--rlbh=m zYcH-SzU!hm+v%R_pFVZ!)IBc8imRqDPm7fcr^E;9TuFSryZXrWonDjIUWfpz1?cnd z5HEEJcmBP5`qGCFj_T{k)#IPMEdJt^o$YIzmv3%5pIbXRxW3rhzOnXfB^5{EAp8B0 z=$?C%_^5apvPSpz#~1UR`?qiOa7ua*?~pbO^WuZzFN6rFS#+3on%cL-XNSJ8g}AxC zcC!*=n^nHFgX&6Ju#qIpZZAfuks2gc@v%-=pO0+sZSP7yzFq%I57RoT?9KH!6@X5q z(dRMRkBFCsbZT5WDB@!EFyC4zamFr-&kQa6(BYe+~aEK3uHQ0!R$v*#s z9~Ix!5xo!_mlivVD@WZkCyJl%^qaZt~+KRFfI@AbBKlprG{P$Ybex@ zL={wJ2K{@D_@`pr9Y(|-TBrZnp0e*2cV4(dCPOQQuEqCr)U!@e&%oNxaWL@D5#l&cK?>Tm@u_pRk{{7___;p!HRc~u9If(+cCXsEDXQsIUw&4xY9f)hWfpU^Xx}|6RE*%K%Nt1 z*nkV1!8R~Lmo(v+e+_n`1jKh@mY@MO{Jv}A%0Addvjz5@*U9vts!IMn|^^&i9J&RJhcD20g>DnyQw<8wa%b^%t@@Y2x4Mr7CQk!!_^tO1y(W1t z`yzbeqL@wFK@jniEY-TG8zMWaxD(+{V9|M?0k>Fm*7aDvygVUKSL;QD5vZPFv`Y!>L%gLQOCYYfZg9EH;d8g0XS-|nqV zhl`!3);@fFad0zuzXWYxX5;D%3b1{$Y!r)M-~@5$6clw2;_F(SFOz?Pdbq~fa6BE5 zcG7-raxu1@GpRt`O?6P{`L^^5ipq!Rl2+Tx=3HQP^gttQcpdG@`j|W5@Jua8O*_wKr!y=XP$%PMYSI@de(fM-C&?Pt1|N`S);u}G z4cp_|w_yqhp3U30NeI~>j*!{vkKZz9XaL{0Nn>lkYDwPRe$&d*D(3CmB%Ew8&ABzV z;8I|H@PS6k=+Ru)Ly3KGlc%)`c~Ke&Ue4-+4@l_QU}jZ*;{0KK@PWqB@IJWf=QW8Z zc{;m4%w8Vd!i{Fs`7cI}PQum(8~s~`J**Et)F>O?2M-87AKu39gC|mh8iTqomtg}W zuUWqbiiC++)7>VpSwx`znl z)$h_5D9Uzb-x88NhIp~yh4;aS8kNKQ;Ej9P)BTy<4<@XPd2&i0T$(>cxegL{LR?{e z@S(=w@IH7$x{!0wFd25sn7BG-+-G=waB1xn)jGJ_$uNaRXN|bUqBGK!Oq)+l2w@Bw za4oZ@l%(xbv^?RTEix8Msdi9~Y|e#s$%aI<4Uuzd=A8h$v8X_$!|Rd_Nod<}X~#rS ztMKKI_Q8i5ZNvNE1Nys;wAjt=12ZVV_UDv7xO57Nd!ut|jf2{JJe|Qmil+}X&W88F zfz^`sYb4ut8VqdrXOhVDa1zEgbdHTKn|?feVRiIKBW!pb?TIs~4<4H*cNEU9M#&U1 z>u3^`HaZ`VDrSr29~-ChhGl9ajiBL~T5zo-&o-zDJ5yuPfSW!mQVf_lkPxKwu6wJESu=QJtEDo0s3y|-R-xwHmzb- zA6$9`Mdd^Od2ps=(U@g63bYSC(nuLSn(KNfu@BDJyITdtD3yZM2bYdPQTd1wX83iw zYaMa^us--m<7jvv99*xRr?Wf8Y&uZC;Pt_!e^3aJLGZX#K7cx>)@lXy38>X0aT=IVPM?zY{6s)$h_>C`xzc;Fd>uG{-oH zo;VzlE^`}%9#n6)?<&s*9RYwaJi9#%bg@^*m>}=M& zPPN?_**#?fZGv?kT$(?@lgG|b+cIdzd{}q)u|D`%V{mvMydgWtb7*@tyJbvV9rIge zA6!~HMYRs@b}~$1=<1ksKbw(l>%83LVhJG(>5Mje@wk>*=&H1Rik1iHk~ShOmv0np z8xzqsM9$32n^uJ-KyfS*d0jH;5fptd?X)QBYrg!^KKNLpZFnDic;bC<1_jtomeL28 zPC-%kh~M1K%9IHpxQA<;4ex^kgtgnR*-p%)0u8TB4<})4V+Uh9S>~`hdQ7@k&69Of zAKdKS>{^q&KEbnVg>~S&t{zmACnNWv4fDEFn8F5>X$122fA0YJH^CBF`KsMn0V*G zrL|L3>)>uD!W}xLG|`9~K9_EU+&V8ezMBxjkmlzS`{2^{DOz>_x07TIL)#`qv<;DS zYUZ5)NU^9u-J91XlO93QcRRE#BXby@KGkR&-UlC@cpscW0k%J<^ueW55Y#^$Iq>Cin53N6M^|S zvbsk1ZPNUjN7b?S!5I|9-GFx{FU}^CcOG2&2PM2ukAZMBL$Uig)(4+zlnw8L2SnN$Z=-3S z-9JtwmDdNC-a@ef2YEX#K2VcD-v^&+3=Z#u6UEC!<7l?P?j9!-%X=SO`Vhq$Hc;Id zcje!np3Wv#zt2ebxs9^)!DIEi(;iKz-vjN@P^X+(zmw3pS^Pe?&2@ILtX#QWe33b2hjr4KHhf}(Ec#3&n2&&wT# zr_VIbhWEh%dfM&RY!7Erfw~(HPnTXnQMog}cCyT&b+ks<@O#xfSts?uO*Cg$$>jA3 z-px7>E**lRY6nU?QQpuxnsl#PP}1!@o1MH9*=)-c7^A+wXgLlbpA2ORpdV z^Nn>hq_oJ)nes=^gKv=TVRKiE_MmCFN~@3;rGen(tUkDO42sGJ{LvKPZZKPJf;>dsrWQL!)eXA3PxM)_9xIKD$Mn zNDXcStoOmCw@_?ggf3|ulYeP?q6E-B_=d*d@IE+kWgiVf_W;_S&+daK6U*y^OCO?G zgEIwpe4L;bD6sl{gLI$UI7=Vg#DatxH_-eHb-AM(CG$MEbQg-!9dzzQdBf`W4bpvX zK~1;Y_Y*R{fVLqTFw|dVmZ0_wJ?5?RaYW+Ovf;J=Ee2IBN3`1I@td2NuCQTZ@RWl|? z)U<3qzFfu?`hpNi1l$lwGm~%Q&V~tyWX8hLG~0sO>i$KJ#$CV-@t*;M1pX0nhA1L_t^qf^Ar36p zZsTVAIg<+1Be;I5eszu%wO)vs#^4)2l& zMBW;Vqjg5vE%ZcERbP~o-qCQm_4BChnYhgFlc%!z_fTWd2p!%iCob*dN$6%kGY58i zIhk1T8!Us6endg-)|~^T$Hxn5gaWPW*GYH2O-{I1PF_En3!G!t_3Je3ZrwR+no!pV zdc%p60BBvmPP+CjDC~AQKtg&H5D7#BX8z~YI4mRZOMf8wIeGxK&6#+X^u=>%xYw8+ z-Z7uF?3kM@pWP29tc>|{*|`?~il<*1PPgvhbSK2u$57R3gbr_-Hv|m%54vv~^|JfR z#MLqHJtO`lPrEc!ZrvGZI~lg{FipD8&4{;krfyW65T=nfDz9zM4AV3?ZheAUm(9C^3ED;AqRG9}ivV?Aw4e`{qd#Ztv0xvGzL{OaZ|Iczd{X36j1Y{OpAJLuXtE zq}ew>(Z2bBbf256^k~nT=IN~>Uz7%duk-rm z(m6<4AH`QT*%f~X!zPa=mk{LLMOK%M}(Z2aWBXoG*oG5-KR!7V6 z*}djuVwrt&=|?1knDQ@CPi6r%cpD$kzWG4oclaG~(=2`Sgn~EFJxyyUTXFE_;ib!v z^d86VoM!6s1cA@P4@lR<1(ltY10>{q0fj?LW9Dy#&8_QcsNKMsn>#@+@xJ*`V|IAo ze0pMibHd74hMm+mmo7l@=b?ksvHRwE_(J>ULygekee(t(kR#EGICg`XxH{&>+`hTA zc#?JHaq(-qx9VLI6YBsH}Fm-IjTMb7}u1MLTfZ337?|(}qMW4v{%C z7dN$e0$9bO0{3m+d2{I%B$cMP*@^ef85Cf9byDA4ItEGOBmP;n ztZdmJhGOnRjl$u5bAYmTdpFz9nN(!;a~c9SbY|a9mOZq-9+K{G^TeIhH#cFOU0;&d zC-^w8zLriw@^J@fJ5lcN#Eo=~TTs~T{F|NIuxP-&o0+)LaI~Q_{dSVf;d?k~_RSMD z+_k0^V(h*$g93to^ZMpP8h$o(&Y+$!fA}6wntcNV-p41ybro^!?~=1z5{D^dq+$>Aq zJa*HRZFcq(YU4o1HPm8f&cjQWA?e*YpSnESqj|tNyl*}t-6I!NcDo!PAw3GH6QTig zcuktyh_KQIQiJY;nHk<-tmXOOf`Mc|grJ6>OpN%y&V;y&iSIeC485Agct(kV#Vc7Qf^eVvm# zv~NBp-RBl0r=5SZa~l>7xOemV=F%TX`bO__bGEDHa*p2_GbYWxd7@72n;+{Q4%5io z!=+1*^zGnh?e}n)dP4ld_i)ne8=&%bKHh#GJHEcT^bC^L+swW(g^UL=v~NBp-T&s& z6Yc1tkPE}pTZO+UmCSqE$24qh?99L|mwtk`qkZ$SM%nPbIjHy~f2W-+J%(fk4lZ|GoS;@A z(7yRtBXoG*oG9T`td8AIPNoN!6t8bC{fJ}`4r+I+FpAE@Pe}L3ZDi@2CltJa?rEqQ z9^D{qCzIJXmo7um`^3TSPLfT$Au}P}BNtS5?i|1>gnttci$G>WM*0Iu-zaK#e9WQu zv`;iyqA-*IZ zHa#|W51F_+zDx7&X_ppH((bT-BHS2Hh=ggLdntuu9_;)E&|X`?Q)Z!Ybhq-Y0@ zJ4xnHEN()?;t)}%=Hdxh9E%Fvw|Qq+CNwN=(hiHu$lljt4{b~|7WdHAgZb9>V(;j1 z?egN_Xl=1~Wq)gNu(mV5xj2xIx@T&f8yA1SIJ0?WXMT9teOx@&eROC4$^|?<;E;fA zZ@RP0JH;SfgXHGulw0$r_N9tId;_X=4a7Z?r1f_D_lJz;pA=7YA5YgjLXsgA<_YWGKKOr9PK9;I;hzQ)rc_R}O8l*OHAhi?a4&BeE8b{;q=WvUc zrtb`!8B7Af%bEK*4NaS#@1HPVwo*g)^Qi{ZxcfQY!KP_G!+wX8f%`gdKbP)7+0Uoz z{gY+O>irwB_&(Lh8Y8~z!XohqIs&pa)7y42GD@cS71qmn(nCm!cjn@j%m5*%eG{sC z4ZXdczPHze3T>@4ufJ^e_jijYyH7CpcM>7E-}9~@kd8u9zjNI5#5w%CsP;8B_fEF< z^TBja8yonbcz^f3X*NJ12e*X(mlp_Vh~3md>~0Z0zYDbo4c@&gi#;%LJlcrG7aDW< zkoaKt{rSwrr&fHLXA?9?Z|XpL$H)wB{b;>E)#zTd-Y53AX?sITK};kdx@n$co`Rav zkiF@of1+I3d=RboryAFb=7Rxox621ok|hQam=AEDyq8Wv$^{%u??k!#Vl?h+MDN~M z+^?lvq1hGGw~@TR(J$^lVZ!`d#ap^>CWZMZapI?W0iA~8O&t_(DP&{es+61fQdGMd zzk7hC-6T(bw^7@p;^FRF`D%k!#a!1OJYkMcOlg?i)WPgdhCzYG>zPLDqVZZg2qjrY z{kRoA%OC<(I&<`GN(1X==g~#Y4%$x%<_y;H&fAQN)FCVS`7WyS6eNqaI!*4(gHF;l zD7*A54XIlOe<&_L(@0$emtQ~mK0AX1Y4pIv$e z$f%`bG&n}&Uq;dykJ5lcN{hT&8=Ls8ba?^Y~ z!*&OgK=5*2pM6F{*JjQM){|xs-_L1tbAZ;{d3uKZ?)dubGaAM=b1=3O<`3<&&os)$ zi0`_>NbIwlEUsMyj1occcwV1yFIw*hTAu7) zoUwwGf|y7^G%w)x*`?o*R6mHXY?1sW3IOlhY-n6BnhyrVosCCHx zA1}MIvwdyz^3BbwJNuXCJDYp^2fJ%ZUF#bfsXO18n7X#Vzax*cw}s>GGgDF>-T{V|#DS0(J-wy80cH2{ zVry%A@9N>&NA?f4HW&84=UemL&E2(!_O_3<=R4b9BsUiM*Z$#JVUU~i!_A|`-r@d1 z_qO}?_YRNxcZ%D#j&5FCbXU%=BH~-=|3z_I(WC3W=werzvy$n`=bZiW$4{L)wR-AQ zf5T{PO}wW2szPftTE+Kw#2ab98eps7BEU?i{}Rs@PK%Wbr}{6h7`f}BINRx->YtV@ zv|`|gR>Ulof$r%yZ}05%vgx(gQk2F``p*|)dv9y;?4~h~bRGBqlBbN%5O-g=vwQmN zm;c>P85`xq#^=+d*cDIC54Pu*cNR|$`Y#Z7O4B}AY;Eq%ccmFig|?vcGsGQN_VXd)+hTKYK)>^pZ_biSYX2VpIC; zGcu-hS6jXpKNt5z#5VBbg!*f)p)JD~V}YGhqbq`x7^0sF>U zQoz2EzNJ$v*=z^mVR8E-5~>-lJ=-#+eCnyE-X!+KZN=BRtMVVc>dQUO-eyRnIHab2 zWU2S#>p%O{r{3U3dN@d45r08ZKUzCHI@sQl`PJdh_LYVFulbc{D(T4Vs4NQZzIdlD z3OW}$r`!dwt)wFs;$`3>o9opwa9O9w50TKRk$^8VsFtC83aODg;(J%b3j^$Q@48fq z@RE{X@71h6i4OZe2mKd{yZ06!*_=PSeYm+h zKYXUUa_&_BdE)G1Z_EBSR6xl9{*WpRijq<9KC4m>&p&+nV0-VG+E+WOjJUkGy1i$= zD;HnvNRafycd82YXt8x^adlVT`+4d5Zm%DLRYO19Jl|WV3i$Cw znPjc^LTV$pT)!*sdSdVT?qNyjVIdMJ$0uavBvoDhceO!_x0dlhw`+@dAbqtA43ck_ zufJK`Hs9KkF0y*>)>f~KGvz-<6uJ#fI%bS*%&6_pkPY~O_k|h7Bhr50xQQ_soj`Bfs3!B(ez85mb)HnFu;ABdF6wCKUN^5~nYH_~58p{JmF1 zW#m@{&Megn;BITA6NAU zOAK<;R@#x9wvwk6Nv~X=wO!p`60Q5dtMa?0Ae|W`f*=;BcSZbIN8I0%MGAzPlOcFJ{<#&ub%%+`hq*PJnpx|;EFwnw z`71@w%0@IFdu^)#m1e%0V<6EpWZy42$N>GkPud^iBfM8h?v=-l*-6&(s%^(1jH+wBre!bky zpw@)Sx*3F6JciAn#!$A+Aizw*W)S#FQ|qkFAoRUxpRJ`Y+=y;8+|8igar1g!83MNo zh0VE(wqoFJ243fl6UuI|HiLj!EORq}Qs`*RWNrptKaM5AY?}eW0W?m%fuRSg>EtAm-bQPRTlh z`UgrFL?ZhN1}XXt8$C?{<1>{I)-fTdRsv0rM_LGa{kMsqX~ZWpu?v!dX%5j(WClyuTLu_3K+G*_@itH3EL}%CGe?)YUR??R2V+v zpOVKz3SxB^)q&C9!2Ri9El%y@yCHd>OPMo=n_pG^k)X2 zdaZbMw@=5X04y#hAwRjX$e_AEu|kFi=ZA(PPVTT7n7_^it@)*GB@|b}fH1gs- zY*{r>Rc^TDS%AAP2=%e@i>~o?umhC(*8_!fOQxIky_!SwN*S2PCZx&`Pp_4I`oYnG zQ(MX;7}?|65kJup`ju&WrV0^V150Iy&N@L$PU9xeiuVDeqY0?tFlwHdUc(%55S110 zL+mX*-q)DP6z>C!u;YE;8%b4wfEj)JRYDmZ*^$ z>*a2g#wp|eYtT}3Zjg?a?r9$M;9$vvnjEC#b`GW$H!D4bCjsUHMS4U*H)9QsP2%hu zNf|jfV{^Qj>{@y%DqZR;Xi7QAZ!>h+giGVX*fw~ngg?rmQkLAKEBNA?e#N~SGaYEmRDajVA#P#HunJ<<^T4KW z)x7Bg2_>XE=$SP*q&btP!v=CHdk!VlobOfGDcU@uW2ZYBfz;ij5|>r&qgPUj^*VAI zSGC0K3y$@ivcXP=8j9SA>)OalJW!H8&^M+O;HmyR=QG$|1t? z{t*hoA6%}s!Q;0<{j-kn+77{Q)?lvyV3kg66RVOCpa#`$zo6w&8z97PD-_?h;*UrM z`DsF2G%55_^=$)Ox}d@U7A6kgX*%MQf$BFbp$3vglTZ#uF_r|?=>pZ}GM%mg=%6f6 zovw~Okextjy*Y@LYIQ!e;JYW;={%f)WKuV1P8ZO3!2H{rC$wFl8mBs)2Pe=ZgvR&` zELSz>03Fhy;|JSi5jAj!bkw)~zB?S5XgnLr(Rmpgu zKm`Yk^S}YJ&@9YZ&kL4tz)-^hat#}%P^#W)XQ)ON3DBORiv{BS-a-!!4K$2d)V`*g zriy8^tbC=SRORbcp^X5haw=W)QmK@ChpnMzriS_~yz(eZ-4Sjn*d)!mnZNh~@p;`> z1m`d6%!RAOdcAH>jgoe@U9Hg}Y22w!j;_7JR3*!$#*W18s>7oscB*Q&g<8@RxnTJq z{z=EbsnW{v>U?*%T@gz;Ga6oxQN_xEI`5vp@JDN9Y>PrOlw-GGwOIp;?+DyA7Url~ z$PxwN-!%q&-n%z8i$ghM9qD3IwV-!(7QjZMm?cVybjw(HMLfb(gR*r03)DB1dNx_Q zoPw6@fAxwM+vu}87^RA>g&Vs6c{MG~vI2Mg1mv|DP6G|n(lvIT}zabO$uAs zF=l3_K1939^vYJg09T~wZdZAry2_k5Ww|Di?lgk^rar&4bsND0)d;4Sw$z-tDsA19 zB7=#pZk$(kN$jJIU3zI-Y0hoz4pd{8rn$u0R?d>MYuiZnXtowT^f7RnoV;?j9ulkC zwrX+~cCMP7Rc%Y9tI~dTpr-iHOl#kHNCMSRPML{b5}WXV^v74lL)lo=)6&Z`mR@dk1dhw(Q~@&CQoj9VHY4$;W)^~N z&^1gZ(7VjZQ4?rUZ@<(|@MykjO@OEv#P66}uh`r-@f$s1!2V7zpC9ydN<0WVVq;!f z10FVQAtUi0I@K{awwRjN&8LdnSbkPv#Vsv;GYdcoC%&20K&EAFW+BFpcQdQ8xs1&$ zz+$`2Ebzgmv7~Nhp-)4Hg7O6-^teR6&8!Y^@^+CBSh+O zt#(^Y#^?^zE-f}Q4^>SS_cAs!g2Pqd>(Ci1s<{juu2KlCc&<~TuS0jPNPWCIq?VQ~ z5G<&PxeYUo8|gzzLwDRreVc*FI#^FUyQt2ixTlncdgf?2ND)M)!6Tzuck*bcCXe)a zbFOMNUK(KLkPd3S_S(hgcIszR9NFfD4?fk{g)zf;WoScP}up zc!yB&Z##k!D4ahTgFLg5=wH#^V(3l$vblbWl5RQtHbKrbsr7Q%lkXW8T{eOSg*2yW47jC9*lw8lq=5398YBXR{;e4#ZX^eZI+T=4AM8;$e0ITa*CYoC4?IdT z;Tvf|B4AI)5+s1l6BT-Lknr%LGzsSq5;t&ut8y3}bfeGc$-+mC91Yz;x1kzzYY}-- z28_TA;N{G$)p8m-;HBpJBqULC;?ArM)yx`wWY4))84c8WXlvk}MjPszN`^Bg`$-dN z-YB1_6zNW+4b?j%KyM8)tO0=N;64j8QPN5q5T_tkzZU_2C>%;}!o1oY2jUTV&CAB4(F$J&&OlpRA>g90DPNgf<1MsN<(cuw$iq$fIZoyC!gJf`#4c+3hbIH zYA0<~Y0#}F=4D7HDu)X79HB#Tbo}J7jwo!4X)|H*@YnzvEZeHElzdYaB+}l6cT?_q8J{uRM+At`jUlcf$d-!@ZD39Nc zhMX^HTz=$?snG#!AI`1x>W-QT0l$$Rj4Or*N9ryL5)ACb z$zr@ODi0$%Gv24Ch|RHcdB(rwk0TU4-Z#?SI%mAkxhfz^oH#0haH)4P5Z;i0CZ1)z zzv`gHa$vT)ER2MP+vQ6%eCFp94lT5a>AHn@v;a_2NQz zMsDgT%Xi2W|E43zfzkPMHIOnBEm^1blpQyL$)8!Fmmv^1b5H(iqS;_=fHTdAVbso}F z(xrez$+!32nD1O)++w%)1!!`oitt$nc{GZ2U0h%he}ZR?GVPFhCNw~S%O*VUK)t;$ zuz=4Z=fWe>2fdkx-0BF-<>gdC9O0~*9pu~l0&=V2U&l6DmwYczfCKV{M}a@vM8rs^ zx&z;#U3e7ugT5d5LDWIUa@toI9gA8W`NE?Rw~bS-sp7U7UQU6;S*PCK=d&KbJeu*T z!qFmQd-*u7!D-9d%tDMG?`BqGa~YdifW>y3S>S_BW1-#NN3+A|2wzTZ zwaqMm$>wBaIeTS`&CJ7d6Qe8f_P&7l1Djb+i6?GY*oN9-GxI>+RB>m_g$S4SnXR)L z>l>qUaCF4Cy5eo@4#KIA)2?{)%}G!8MFWPIfsJb20kyFjP$S*m=c~5OaGyNI$i{UQK_hy8taRJl4pta~VMVOVx1SMhzb3npuIk zW;(Cjv`Of{@6jdX#ZOF&Q4-zrcQqCK(sjV9ADmBLSlIKD_s>zpgd4IAA zbdn}ry|2)Yd3k>851j?huu2}wEz>NA-m zNZ^xv1`f#^F#cxfaHcx5YZQwV~!JdK2CkR_o~oA&(P0g#Y2qL)liswu5YYwZJSVUZPVb!1f6)J zBSZQMzll5XroK%(^V!rSYeMiTaYu;*shY8Zgjiy)_CQqr#$wRZ&vZCQ<}V5Xnz*N* zC+hUGWxX#O=i>o|NHf^)Gscjyk9)&8eVaBhQqe z99=3uS0X?aX&M5Oix9bbn}hzWH_;z?>t(r!5heOnaesLZbE0wW0^qQi3H4&VYZ~8> zou9x`dV0|=ld#v|U0%{(ff%!VNkW1KF`E+qu_N9W8Ef6Yest~n(c0Ex z?=X3E&^hCSjzIJyV;K@u)u`V^aZmS+l!;lGeDUiY@s6C#QzkC5a!?0MfMij8)v9<~ zMmF`-#WKyNmtQS`F~Yp6h$4k{&-n9>*hp4L4O0tDU*zS=1MOtxk>M%ef4?H`NtQqp zWlNW3FQ;Glan=FU@@gdrs>Shp@n;?J#&oeas_JyH$0;?w($6ZmR@+?wJ88~~RoL7y z?qW3*4&7T@ZFeE&k9)hTu?mkEw%P6iY_{F*0-tP}P5O2h`gnBEFsDlqW}!rz?XHdv zJ85LvWwTksi4O`ZTrRLXiZR@yurhCu`g#cqj+A0 zL!zU49&!wxQ~u(PhSl6zy@{IDYnfG82}QGLgA}4ucXVudU$f*6-A&Zc9RuWCXCyR2 zq8kbrQ}2u1TOB~3=!@eftc&9+?V=-ebd=d>*PWr8s2RFs?eZrOpRste1Yh(BB1;tp zZ>F0PI&4M3#!ZOOiMyIU(b24UE}#OaOUJaIlslsP8ItP zaV(lNn23!PC7eO3MK_0lI;@RO8ptaqS+T=er3R#WIA`WTq>R!LpI+gicn;4^=rmP! z(UaJaY6C-x*f3E>u>lbjwXQ$th}hX1kkJNOnks3uTWOP|jDl(rIs!031&JO{=x9KM zsB+UoM;(;H7EvL(Ij5XYRRI7B4_hrlhX+F?8g%^JE8tN9+wwO+r{j`S%;A0urg)Ja zIy~elRnjTPBOu|^!cYayOwqwVI^ZPhCRIdF-N8Tg?dXX}fDJvwh$+s5RKp88#7763 zzEJbx4)IOZ5FZ2GaB5hLhiYhTYMGR)LAMGFHO;rJUYBo>Czp^Q{PV2W$GBC ztoZn_5>@b(%+S-`bVNyyF;3mX-cxnh+aOBR2}C`{s2;yCsN+<>XEGfcqetCbm@&pd z4M%c7OVlc*a0(z|l68zxz1-kHNCOQa$wi1ka6mqHAc1LzQQoHx8=o7XnJdFPy~OjH{tZAw@^w#|9dVP zVvnIF?ywaWGRnBqLP!XvJ3kMy_>-uyOh1q6iMuw^I2z;r;MH`Npp3EtkP|h-8G9sF zv-j})=HlQ{^<>oL#nHTEao4X=DMieq2Isq9TIX!@C>G$nX0Li7W!Lmj4|x+D0H~?4E{8m;@uC*2S7Rr$1WouUm4n3O703*7Dnp6Fea9>)xB>1TKm7FexulCe8=g_~Cx72E7DoqU_Y?9CHzW zx+3mtrS5wV54W%G85e0K%3fy6^Bu>*QnKrXUn_|(UJ>`Ul73@ArIqyYpllGB+;VFh zaRT))1%t`V1_y*`8FA_;Os$MWL+tWn69uqo8*w}wrj5G#YXaZw-dwGZxAL?n*#&biFoIhPr8 z&V^db>B?I@oT zSx9C!B9bo~cm|~y(eHW+6QD7f7K0H9_g;lb3F8Fn@(atZJ^0jS8`h}I*tr-nbMgYf zEVzb!6sXR;JqOi>d5z1+h~TeheB@S0=EU!J#JhsE;`ZLr;$T-k;4nX0w7cO8x$@+$ zQ>W+`pv5S10T}$clV60UfMQGd^FO4ALCNkLTi<07zUJi5R>XsubgHj7S(b?a9qX7F zAT?OmdcF|bds~ZVH;-;!TXa{>uVyr}*vZOs;*UEbJeV0{tcLp4iihZ;D$+F1ljO5I z4;>6%XL3&b`U(Jh#yD?|TcyblxF0{dxcA;dc%O`(hCb)Vdj}iLrVbUg!?{qD(+0Gc zCT8nr*;tB;HhHTG04mLEmMUoPn6|1Kthb!4D#SG9R<2wMt+AMYs|v8!W~&N(u4y7^ zTUF@8D? z>*B+~EG<-+aHc6eeW%vUozk1tru10U$;77sINgJ>Gd-#|b7t+L(iP@Tt#J(0>q`+xR6 zt+6vt95ZWDDVxM|1(f6Ss=0UuPKCzJTKr~5yptL{ z+)I@_V5wPFR>;Fkt7p5Ock!;XtKHMxm9w1-E6<(l%kG|@)qTCV(4E0!iUE6fC~-?K z3~}g)pX~r|Lf{KPm~6$MAJ9+9|1E<(7;X4>E8>AHTGUv>a!dlai#bjFCLjl1xyWT& zNZ0$E_%|Iv-f|p!ts&t7Xge3}9&*vVCh45(IUTIIcuxGx3LTbpcx__eDe~&&F>DmS z&Lwm9%-K019|@nOcy0skPLZtjpj`Z;{*t!S1aP7MOx?ToT6SFZ-n~{k?WSbqoU`80 zS|0#KJ288&IoGxA_|mrHlMoycgd)U?R)^kLE%$E6-R%{35R!>dUh<8O_yP!!5E!9w zjR^z-Ite@&h-99sHS03!Hfq$EwRU`X5z1JOJtEiU?KBC~_&ZJFVPQyccAC^`vX8dY5Jb$!-)SOr z%~%ijoSh~SG?%c`UB#pd+Ta;to~mR#%Phph#CeyQXR!@Cx+RjT#kJjqmgovOBba8~B80>_@e>^r~?) zj!hXf&tHMtHsq#ITT~wtfPLjMsWx?RB{~V%5*RAbyHNUHIDt^iC ziMcMZ->JJ&vQg5_rZgYA7mN4W7_xf#T;jr_~H#RVdM!>I&c)rW$y< zZA8&H=cV`<=Nu`%pSwj@i_=G49GuZ=NlEdQs}d~UvQfMh>)A3N#rb!E;Y?oI*sMG&{eG^9xW*)%y{7N%t$9$CxuylqRgWxYtu1_wd``JByL$coYeHT+ zBU?;9e0+NT;Md>2d;akBzUn-A`|LDtj0Z;n&p~fK9DYvC*H9+h!p?}Jh5})2?e9;; zy3*PZkJfrCQ(c#LZ#*^QRSw8{TevTcM9abR*AtO$6N6Y6Fwt6fy5ms`!^OglhU!L z8Pcs54IW-d$w@1Rv@F664FMutiFN9^XO;iWL_npMboi-7xI5i=ae$xl*$#eIy`6~; z*qvdw)zPu5STvjncLOo#8oiG>92m_)*Nyr#Xc6w(@6)guSg87unsoq*aJK;vUF&8I zYj8JBd1k&Le&#}czge7t{xz{^i>|_bB%apK8)lSy`utgP23S(y#zS^caj&(~-zjjn zbMrPeUjBlyL4VY88r(JL6;k%?H#94ji;H5#b@rP#2VHB*zxV;5jIs%i={5Nos^>K+ z-o}})<9jF1S;n>yr?Y3P7#sIAB>oOs=2~-9o&a1Rr)I=* z;cIY=;*J?=r(~xumZK8Ny1U5n+cBgQKDSxAFX9pE+b7ZX94;^5gH z&`S`y@h>0oxlreLt-`+Fn^LF~urEY-b>5kejgL6~Kn6A5x9QoBeDAs@{$wK3w~Bl4 z0eUc?c6A0ziJDDSt zQWc8(y2+e~kE2dz#wXDF$sEy>IuD}Z+HR)rh0F+47Q8qE$|?^O)*%!jGXniQoe|il z&l!uMvlN?F&InW1QWeaECdav^Y5-ML>RMN&8vo7pf1%z(VMh|x zJkeo^9Z4;E(oRsSr>c?Euml@GJIn&;!=yr_o|9~86URjiTD?zaBwsQSqxDAGS*m{D zO2nw|X5zsV@(wd-Y2pD?-C{m<*R^gvI)X}(uDOvgV5IGJt!rI2t3_g44Mx;;t(YHR7`6#-5&);!j!NY;_2?|{%Lj=T4Qbit~3@IWI8qX1I+sjF#k zitTJ&jOpY{Dm0Updt#=Qa{MN*cG5J@|HEmf>JGCr%=PO%HE$6&@q_?5e;Z#j^8UY> z;PR(ryoDE{)_e4%jlt-&Gv8}fSzq-4Ch|NA2WVkd4GZ8*e4htI`8OTAr3gVA2ZW1; z;7t683si^czJB5-;-*Qgb9Ovz#Sxr|A8~?ukD@0KchqYVMlNvz;skHouH*&ZG>MZ| zC+NgU$rDfwX?_uP7Gq19GX&KZFqnb-{wFiGxi&JU0`9ReCm7Xvqt?{nY==vUuhy#wrE0W<^X&;2P(wOvoA z6O0}BKRfUPXy4~;uGNdh?6fl1YQzvegoe4Qx$2$Na0I+`uEjpeVhNs^u70lCOv4j! z-A4;EIVUV3Yam~xg(SR7Gp^^L+@SE>%{TZ19#+8S zcq)BpWMh~=$X-*QlHW4r%?nT8{M4b!$#aAF2B58LuGs{A(MXx_o+b?o-SkDLTB_>{cS&^) zL02=AWiNm{fyqo1j}S(lu<+yw>gg4z0iNQ~x0h>y%D^_dnqxo>i_WB#KqHMl#ySDW z6PQj=JVKV|=r{4?2@54ps2@(v8Vue>*Gr&LkeOROYmkn|9A#t5O@PK4g@vb4SU`)5u0T@~K~b=i%ew0?G|Fqt&co)2MhuvxbdVywRNvtU>*V9UF86 zJ=U-YSi?rwZEQf@1{Mxhqp(P76gsM_bSIlm4w%v?oYfp6A#6Y!Fab@3{6OTs>$?sp zs5T-trX-oY3MEq)ELM%@2Jbrxs;PJh;ADliqzyBMj@4S2d z@X6_0=MNv8-#@)`_s+%bi%4*L@f=D>M`w&c;)|8Oc^;D+JL zj~?B7^MT!HQpa107w7Hffs3k^u1~YFVr0de%0&FfBzZ?E#Z93ShwJX+_^9ZO5t~^8 zf#f!`OWUwKu!i^6;csRs!!LI;^UOTPW|lB<+-8<|U>Gdi%`ElkVhRcL9T)x|GW)bz}y8AdjDtf~)Fd#Rp zQ^{+qE31yNmB96t&u_= z`wmc8_u$a{&*;w<+nQ%tGTcan=Em4lp)HgOjXB04A1-=IxQ0IxxiKIypkC!4Gal?3 zsnWip14=2AoR(2N-Kx?ccr3r52^X)#abcuskUE#2Wv*KROZ`w|DUg1XP*H%Ph2DY9}_c6^le>Fnrm^ z{~xr}6+bf1G={$SC!a`$;hZi!yIqYW({2Of@oLF-K4Bk)6`fAC0MA4*Y^1{K9z?u> zwEXklkrV$^_|Fqju~YyPJtXQK9krDE?|%!{vazp)io=wPZKo z@_@!4dF)B}q@DZH__H*eLDyIzBPw;VJi*97B*^pdF&u-rfW1g0K!E#F_o@Az`|qC5 z{@1PlaeXq=>ei+H?cL4lQ6S<2-5jC|lrabj47DetN5y2%^2=94|L?WKzlN2aV zZs+^@b%VV~L_CS-;r5FJQA^m3SKz5e(JBLyiI2>vJO!S5`h3LViyjrM&ny$`fRZl$ z+FoC|*m+7i_4FD!AzjZjSW0oCvf?_s))VDxgXnB?Z*eKx@+mNo}W^q?@is9rfLe%+#x1>{cc&Kr)WH z(kco&j}BEogeRRWJ4-6Vla-o7(EdP9t!?>rqD)^uD8s zhg<>`@krfmh8;SAsIWutaRZ{lzk~q5b<_EII=ELEg$5mchC^WC74PUSO+R_rvXPBg z&H73;Tw20W2<8bM|0?#>k2nn<{GTAZI+Sczy&uz#E&Q9Ym`uZ z>zpkzEa)=0qaD2wJk8^(4B6o87h4%k(PSLO#JnQo<0*|%MIMZqg4SakD@5jPhoJ&d z%TiHs@4?;k_ijIaBI`w`XII&CG;*W(h)B?)zw`)N5k9vhXVI&~9lZ!T4e=q7rA0r- zVrkJ+XlZivXK~!en`(nJIr`nqZeJXDwQ7?xfan)PAt7m7i=JZhzIJn_JaySyRf2+q zsxU*@E@`XsK-+P)s+7@}xm9^4w%@7}=8f5^5|0Z*gtt|t9^61S*y=#BXKz(L$8GjL ztX-$Pb|iL6_EyCsy5Sm1B)ZX`CAO;Kc5x}2Stc`KZ&f_38~KY68&;m%EQ%ct9j~T4 zu-*5KM6mPZbe)pZjW`IneY_Y$&Lp6SqhqLu{0buw>^wnSrv!2I-gFQQl}*r!udP=# z-pL!f25yet^AwQBx%++1%W=IJ7k;~+VJGN6f>yd#1H5#5=s>YO zi_dX@^tU>^V*B#S^^c zz(+-q{+7NRB`8j#k&CNRnae>0hZsS-=PU<_7>bVz;2TD(_+SYAc*{Y=gE*dTy|xi0 zemzjdF(JVL1Ia9ANNYV+oO=44y}&w`U~7%z9jLY~!aUi8%=8QFJ#AaBv~3mmhpNoe zLawb>>(`s>>v~YRMsotts>N9;UEpWO)6=T;N~^{g&ye__rT}@Ltlp`IJyg$I&I5|2 zW0%o6kCFHFUafSmYNf5bkNR#VDolywFt^qQ@ob1w!^-=X9^Ih^l>s|#FE!liTEyW} z!)~d;j=I#)@@_Ohz@>)WDr7qPQo}8^^Pah!px5|1koQqt^|odfa=;X`NOhTyQ(cuN!&Cz%R%xf=`35$~Z5*~Iw^}+>6LI?n7IYHaLe0cG z!R!E6N>E!ZC4MCSsCr7#N*u||yjp_VH;`mvO{*o+10%%*sc#_d(Sv3p*6g($f+FFx zJ14o-Qo=ZojkSFJh?T%9YZeFGELhqR1TFP?Y~t<_TE z596sG?-w9VZaF~dRJj;+sOD7~7?7W*$vD z)@GJ6+}q6PUyVmK$BhTvGk+rZn_0rdahqARX%x$;Y#z6NZy&aC&%p%syTwRHM0~79#W2-#em$cpGKA2LQ;HIo? zvzhJP%xf8M#ejDR6)5Op6sABgOnUO!fbA*QdlI^y+*2<0@x`G&z&VI+)Dr06Vx#C> z*vJBxo(xVsy@m?|pj0+NMuFhZL` zf$4#MkG=Bna^yf^;5vlJx<-5pYQ)P= zG{V3o=$a9mOBV*3zJW{AHRB-3Q%|S$%k$=!+MsF{Pu(|grv=GQ8YsU(==Z33oRJ1D z_2)q+4ct{NVy17P{DeU68yNY)SOA!Cd30glz>VuFN9r30-#|0>HDO+3OqSbO5WY&9 z6MOyhWRFfGmSY|v4S?UIe{G~2+8vWFe7jvY0jAxC##0tdY@pj&m^=4U`vwl71NCJr zsUTPP82!9|6ve)~s0L0Xt!MNbUKZKz%_AV%R(eG48yMLeLQtO9XtmDqk4N!@W`ymA zBaJ^x5k;zQij53XMV^P<-rLR#sJ+}pBEhc=+{843t*r9F9{LHCy-4^X!Xr;@c3FEE zhS`#@1gfBy(8*=vZGo}zB46m=D{LTY0-Eg;)@HG)giA8tu~ zjCBIg$T3IsP)ln11};5~+)`=eSlg*4M%X~Q@C3p_dAzTOMb%CNCDt(rpmQQ4EL?gO z)=O1kZH0x@cQbKeiYo`?WNso^ZmqY_DKD(Ba1H6u4iF}6NX&&2yZ-cvISs@PqJtqR zP*Rw5r|THCuy8$qA@w6ftDvxO6ELCOVyJc7fs#VhMk6d-Y6Kn#3#k%L(>G8~7;-Tr zSohRMVt5CjS?HU?zJVLpXO2_{>3&*+4!oZpw+_<%Oa?2U2Ex998&~MbI>%_fvBNLx zZFQdIs1Cipft$*TPv_|2249Uen z7wij)zid^rx5P`c7bvRP{2N0Ia#egd5x>;Zcu$`Zuh9n`#?j2le_;IGtKt_LnyED! z#|)c!`(a5m3T!(RMJjIhRq@{wv11R2N7-Uk-$`4mq6}Sz^>{U+__wKeiS2=qyYYQF zZ#ggGMIHzc+m1>)i7!T=_;^yhDLTJ)hCK)QZnra`aj=#hFlM_;(78}uX3pmYDPqfD z@Sq(oMjW%N7ZGIky1VCZcZoQ=wCygJ*OXWLG22~Kr_OhK(_M}*;q3Nm8S@HN z_0gq@zhn9rt~^2{;KmK zsz;+^Xd_o#89my89j}xg&0=U)z<5ROD|s=key|O#auf~v#b&Fb@jWNX8GXjko=CM8@@d!cC!r=VP@ccF?i#H+7|Sf*fQ_U!Tq8;eA`{W1h(yl@~;w(wsWNM8ITKrBm($iO+=3tFftUG-U0_+ zz)!l+Z&tE}W!2&Vex-z0TN)I>)e*`@@%Pg@wg+qufhyN`#cJ-{&?a9-N23jSqr`@{J(p=2v>SGym=}5o>14=PjV>`#Ytw6$t zKE{Y0SEz1}4xQ|ndb&OJbncjH;$$!xu7d$`N2zhDhfUN@bk=QX@i?6`8M))is~umd z+HuR7sPASX!<38;bIjbUvZ5;DtlV+y(Ix7iOAjf1;^OSAjo$Hpi>LMTW>;UnEA@T9 z{LaJ2x8-#)?q|@@u*@)m+w3|fB44R4=+x?e*XB*VB$rb zxl&S_H!Ki*Rot4rQ~?B&FHu28B=~!d27r>klC@HASm;8{-)($&MW*Xs3wzd}?-DvoUJx2=>@;=JVOJ4fL!34$Gqu>o zuQS~l>_RUz{Y8ge7#2Zw}l=UGuBB3sF4~9B(0e zDO@aJ7nFa5Si*H)T5>X*fxk$bvHHZe>-I^M#2Qyau)H z36lk5JlkF2?FifV#&)nMv^A7-Y)b^6_?L-zqc))5ymN6e4$FMSM*3^wmD$UzNQZAx z!DVFg=I6&4?17mw3!%T68l8~cwDyzu$yB^D#`+M=jB@3?apcAE4b&>FUCCKwzdaQ% zkMe#XY8mDIlja`5j!iZN8N1>nvSkXPX>4o_q8Yj3_>AWNu@z@OcEyP(&6t?qk6LjO zk;hmBE_0FEi;*i%L~2HP->w0zmmZ=jX5B~?V|$FRJz0!;{CxcS?!1Jpr?!q}4YVaf};UPaU&X>X8+P9_#)2i%; zZ_22~Y|xh2;KfCakJ2n;<5};@s>WDvnGIfE)Yy6s#}PlCh_5B{$a@bSpFg}fzkBb_ z4!=`5i5uM%S!2|rbAyb@~kht7c_k$KAwu# z8;z-Mq;jkf$$aZBnVTC1yLTiSsI>ESsao-d_$b$nzhsz41if#-#UD*Ezjii9tUw-I z;+A4J#D~)i5-nLC(!g}9%F)%=i7`ZdD2M#%xuI$Auxn+p1Cz zF82CrHwNWN#oelWl$_oTyi-ZBH^q%(?qIQ8kiAt!7@4((=5B{5EZ~WFYIcCmQt{30VMX>U(jZH>&>0F+H%8XC0Sb%aQ=LB=Y5Uv|d)mH@()O{= zdYTg#Eph&F&ATLXW5D8teN=f=oc|i>`%LpLd3E7cDeas8%t7C`f%-lfEq1OK-QuFf zE4tknk>3W0{K^k|MJxTJV?B0UzC7=SeP`ve0mL@CZuA$;d@#kMhbqf{VE}T~g`+pb zSApiQizu?ZfkVH1IXFC%FW$X>d$9o*?!*fzy@%HmUBD@?eH^01K7HO-A`_KK*%vL_ zBu8&>z(P{-mt{?>lP}7KaQZdMkFmcfUN@%bFNcR(RPz(|e^!2?*lSg!T!B8K$0wP5>Z?L4(Q3&x%hSt~Hr` z`qneoug;#DO|MUGPT#+ABJUB>P3@i+H)q?tSTZpWz$=SCn`CaM!2d4hY(72*24<1L zNxb|YSb8hdel!)YH5oc1L_HNH%h3>7*NfuF;Cdh7OwqYFX>TrB|6#lM%%naFJ!5& zM3)`wLExyT+e&o#ZZsd*u<$C;H#=R4zNIN%ZwRYx?mRxdpoXVG(~HKH=nJnBz3()d zRzE}Z5Y#sJA(dXpZ$Aery`Wuse8Xy+xB5ez-vO|vKIU!BT%~!6xh|^m79eTfqGD<= z@BbWPNY^Fqb2UG#QEMM>%dR)u$%_^-twcXtR3j$Nek@OQjhN#4f996GHu(XH^`&)xN_weO>@H_m9iq$K&uRVhop zW#f1&*0W_kj`QyVkK^|Qq>atWXSd(a6)4x3q%+&}&B8Sg zin7|YK)ULYMIBcQUnAdIZp^M;fB%}0m*B|emJc7Fozoj|o)>FT#r_WjY zv5Sr32s&GUa+RsPP_$GSCGXY~u5OiZwE~0!a#A+ID0vrN&xftKS#{f2#E0_-szjwD zQ9HJt5_PMTsFh<|mhc7#CIKY8QfpSvM5{)@f>9vOWCoI=r@}~hw?>?K_*Af!Pj_In zYGiGgc%2`%vY~-~a!r_632*NsQuGGiaC|H$ZPWw8NO%{14~X7yQtBnV`WOHuyvqod znnn-{n#@x_Q<{8Gw{FpvIdlo{BdvSgnvOa4>GNmFnc+daTU=-Csw&qA&53pCfD`XZ zr|1``2__@n-RkdPP`nFnb1J@}_=X0_a;;Gyxy}RT)vJE-0|+2pi_6bYeXl7pkk3D7 zQQJbC&YrD8Z5do#bX7P|3GB730ZhCyPIs zi1ZEU9+-e042WKx2gkoL_9ge4+k42dg}H3Ah=)l14`&pD_JQevtKlCmVez10p$m#Kn8<3RIYsC^NQ7j~#V ziq22WP|8+xzzgc>bwjD?iT#*_#v>4u$$GVVa01d9^mnYY(9Qt$VzQmr7+|Lw1DI<~ zdg{B`fH3vIhZ(gy(*dyqni$kZD<)g|q@*IC4x+5C@olo5uDl&CUFnw^lBl0?>mF(u z6qBtYK4Pwu?P#jIYgSCQ(-b%mldXtvbx$(@Nx2jQ_$)Ew0Wz{3sy75@us?7qa)0KU zCH@@kCw1zN=Zia=Or45nXP>P|rWi+YbpkUve}CqRkgIl0oAGZMZxA0jfS7K4GT_|KT)Kw};soX8rY^nzx9Xc=m#Xp^XRAY9Dl)`kM(Zl1fHq zcwuY3M``EIlO}FqP5ea0XNUln=eKy_K=6VNA7|qGWS~6j=<84l6*TdP;7t685LAch zzJB5-;^ycRKeOSXb4+k1enbrFJ?hS>O#%bA(PmIUz7UI@>Ez&>njGvr`NB@g7f_{X zei3!1RK?OaaCLJbO?7V`sooSD?qf!X#1t21bhm%eZ>~MM_ zf}Ii(R1YVYAjtZfdV)Z8ymc7*J)J=i%=h#m>sL3>Sb{*k;=%@nSB|z@8G-a7{=a-|c3_&QAqk#k(5`=X~5IR$i)&U(3>L_zD>JI7%oH${P zE~R#z5*L~v$PiEj0sRmhtb33m038AP)_)3XHb)a&m0k6YYKTII-x|_lx6P>t&;j)9bOm?Sb{euEC$`13EsWwEo)V!X zWYy%Ww4j7nh{p9Flq};l8?DcY-E314t?;+$=X%w}0R1%LW*qzh4=^NZhZ>pDGJ_F+ zJQeA?o+H8lEe#O^!&&nt(Gy1|fDDcN&7ce)PQ->WIRdc(!>hDS?WB!ocxyl+V>D|( z8|^_8uy=j#JjueYQnEnJ3vGVs z)qU*DRgSyD2P$*MG+yY=8c7!Pc*^Gi$-?33jPc}B2YC*!rw-H%QB#NRtgUWAKF{-* zr(39w3RvbLy+2R~`4cw$(MjT(;H-V>P(fw}>d>gT!(8y$`NbV;Lmj;04m4kb)L|1* z2UFM4Q5~+SLnl9&9~3%)Zh?B^?gLF7B)d@5;jHE`>9B)*3*z|7Z^bSNi$P9)M`5$qvim3<;O{mbxO1dP^;P zz~v}SjAxy#{(T~TH8CZ+(>sqJpFenf@9hVtkIwJ^#_2ono31J0of&X;Ulz9e3lJ)et9$~!?afj2UcO)0+8&TvXYgCm)lZ*1_LtKy5z zOw*)QOuC&>ji%vf$#+s)bE9!+Cgq_x6K3 z7qZ3e?8-??yCZrQ_Gi%6|8Oc^;D+JLj~?B7^MSoBn(M`RyLr&mg=rv08rfWJkNA&C z@^(~8u!ZUzuDh`!)gdU{+-w+;g1?z1V8~B>GxO9u$J)$NhF|Vx=9zhn%`9Q!xXmo_ zz%W?2n_247L-(HYqF}&7;tq{=lWZ^en;C+1%hnxrGfTkT=9^iBd6}#0p1YYPVk^#l zz(=2kL6ga3i=#ebGmH2y*WJh1s?cs(`q*S)SR)DAwy#z2U?|C)eD;b zKpm`rlp|u|1@O?awYERd#;3>SD?WidOuB49$l!`ky3790<~u8hSti2t2kJ~o-*Cm} z#vo*M#b?npyNY;Md~PF-Y_9n1b~VqGCaVTI-5Gkt1ie4dZw~!|i$7I z`8q736%I0ef_hzQ=%je^8&~*5%q5@Qnf_1+`ULf$%CPu`;yumU<#QBk=Xt`I zu&prxqG@^e>9v6yBxyf778MFz zP{;E!k7JJfN}UG_#78 zeDAL!AmwZ}2tr0w>Oy;hk%4rO=V7;ZH8TmHLEJ?m0R_xgJFHr(|cDy7R~`Ss%P0gC1pE3aFi^14V&Sh{Mjlj(!&2qgbihmz1M z@<0x5M20=AY?O}FRA2$}E;NJ<{#c-rIeKNI8B%oxpd8!z&m0C=J)JkCAw5Ek0sWDr zyr>==QJrs{hBjJ{*ibffzNF{D0ybLj7OK&@Wk%F@vw;Dt8#{G}+%OVU9h$9f+J~s1@qx~8M|`Vy zH2aWCp<*AYtJ1K;Cy*M-YXUL*?fZzogc-4GyYusOn6EMmjXZWIFI%j!5vy5Wi8a)t znxq2`DyS2mOZi<1Cu_Ggq@!P)v^sNSNCiH62utwob(Nhgdkx?yO?c#enmH;*Hhw%2 zU!w%;Tj$4;9&$q=dLwwQ$F&;YmB74}(G;b|QB2J1G(MivXjSCFm@a6Q#<4Ssmg#U<`CPxF1XKBSBZ3)z(_d1EP>WntyLn2j+et|{RqNmUj=IGDjxIaq5 z9Q|%)vo4OckgGN+yNiA?loyirwdg5so?Ygw;%-$59ujKB3~wu_+B_0=s2yd%OJ$MikB2vM2VOCyAxLeiHh;F#pkadCZ_=nywtHoX1 zt%}EXBeQs0RU)1)Wvj}h4+LcccdOz7-pI}Ex2hW9-~!dy(JAlyMmVT0!UIfDiR?xk z1e|lq3xx%Cs7#K|dW~#w;i=>nN+nkz8w6;mY=V)2^{}eETj*=zx&?Z}TTnFN*f$Wp z#UP(feccdys<(wwy|IQk#J*+!0MVcWHSm&eEQ;6`=lMl{&_b>}<7gT1=*^6P<11(#b^t`w^+mgQ&sqqIt?k7#yH!tyni6DJZ3x2^vnhE7k zD?=he8A<*sZ2e$L=WD>5A;jePYu+Y?PFndjJwdQYiRXrs7pk@%Th1g2B!L{y+O4yh zZ!j+~mxDxtZ{Q>*25d9h0klKS<1Gh1HX1QSRgi_96|S97Jr0M|>WCzq!uysno#u!jde`qdoqCc4g)`K7E`bw9f_0)9cV?=+dT|bxus+`)0{>blU z!@^Wh4zsJ%xJP#Cr}@%|{-h_LQbm|1Wvgged8iCsHog>5rS~tZ+dmJLp^`Za1*qz0 zh@%HYf6GP0cg&@TWQUvPf}KX%zUXhcAo}}lO)uoyp+tYF*7T9+&!#9q^4Ebzc-Tm* zKG80rca7;jS97Y>&;sxJ|DsLtR-&IRs#Cm7W{(;tPj%KfxwDwkM#lbX#eR4pXPYw{ zM+T@8T4$>(<7C)s5ydXfeE>~rV{7Ipg-#c8Tc-V;`|qBA!g>rQ-U+7ixPC$wWs#cN zC4MAU1)X{d3Oah0O&0lm-x6;I&7TNrkHN%)ViRlHGLaq_=_pzchugi{^%zW8Rz~yW zwoHkqdz6e@$enmDRk-IxRN2%-_ARJA1`}@?ax+pvdEz;=woHjXjORhTkb|7KpgEcT$=u{N`m;ofEzqlm_v zmS<*O5Xj%m5+;t@%n}a_gN4>(kQRqUSzybMD^D_NkHLg--e#5-ijUuK*iB#)FC6WX zBQ~>$`Z8BnVvoUuH9xh@ESH&-gN@oSH)U;`&5WSxJH~eB3c@>7S4Zc;~CWyLIyM+(8~A7nD9?P~)oMun`4>VDXuSj*g97P_+aE5v6Ws zvgo+#1&v&A=@pH4vRB>^IFu$=W`+qP7nF{L5P1nI&q+ZykmN??<3=vH^vcJV3qz8< z^6~QPLAhWT;$->MLF%;ZYbLFc3wCICjrtygs;%N7N-mTbQ9Yg3SX<)qE0`C~here`QN!sH%E!!O^h;d`@ZS+~f+vUL+!!AN-<-md74%r<rkm@AePiy4Sqxmei|DgMsOqz`v2zv}l!fy;@?Y5S5tk7{pz0m4B zl7S#zk3o5b(I|T@;8-_2T_lfDY=4UXnuuR3sAJ!L_whI0eSCWN{J|r({17BEsEp#I zAno|F)!p6_FU?+{=x+0`FFu@zUutPQMYYk?0-x_>j$Yz-uZmx2Xr@+f95Za@?T5wf zA+Z#(?UVHqfjy)h=Dt$BsM_ZSKJWl-ERJ_FYKuF;DzMQw57xBIY^60kyOm70^ z!25VoylXnYT!#G#`EIZHGIr9`aig}o1Xl?)XXccCiMyDm-9z`*QQKY0_{-kzJX7;B z#WCAm!pyPTUE+~p*zmWz)WZj{VsSCJtPOE5#%y-v<@V9)m?!4T$w2kf+go7=dbgQF3qO zX#+%7@-z#i0oW|M#fe_MBKTg`cIzisoT{COvG>t*Rg*Zwr&o>2;f;uK<(I?vnop~r zsj(4?7?%;GHZ}39vzlj0lP}ZdTO6OFR7|hiCXD>KKDJ6|om(2za%CE^bVQ(v z8PMhG+Dyh7!AAX(mX5vPFC7t^;(8jL<;Pf8Tuto~Xa<=Xl;T!GW>7M5k{Ow& z8B|X%R$FGmh;_~K5*7bNQpc=D%k*1RBBnz^I2fnIe-$WdNgG((Z*k=nMlMZbJz3y~ z&J)vm(f)N19fEeU*H45EHJJP6WwZFO*VJ7!l-3 zS8i$%_eL*ERFAXgz@TFE)2-1T-Q>}w+Ry06H85Gld+@m0&*+C3=K^&abg;k`xH9)%!%0>JRdQ6E z0X)&}5Sf!hdN$KbT_b-F7g`uRA}4>lSI@x_Z0=(5C4N}Gfmqx3(9kXKSk~Ga>QO zRNNY2d&o-0__270WnIkSn@+T-06NloOp@?MX+nqsz+}{gTC_2qciVky)3NC=a+yiw zx|hGq)UcFs%Z$%cyz?LCWVOy8cg;y?fTHr&k;_cPRYnB+e#|nHh&jeND&BjKw)h;R zXCs!Gh_j6G{IQl9V!`PWXk}s<3|#{B+eQMUs>nc+sl+lPT+M?nK<^UU#DFfI+-0fc zE)~#)nxWau+N(QVnr>9fExvc#OwU3Wpjh!%M!I}Oks0Ae)|0F(m1Jea$a2Yxtfs3c zFVrzGzLS~~tK*0xsGhti>57{a>gkTWP>G3+53l@mxzj=)DOfQffCh@N6dIi!;&&vsXgJHmELWbgrRYaWN#h<}-gH);d=%{vztNLxmE|D?G`u-(+1e@3o2iENodXc`+^gJ?#sI6k9EkeBABRW2%c%#U1g zB1$tR=J%skoJ8ccoF0GN-NIk8g$_uRX8tY_Mo=#@v>13({#4F2M8J*0EN_{M?S^Ko6_HNR3qhc(h zlUe)4SZhsd)X%t;5Ss%6>SQ(%tub|2Ybw^Vv?i^Sp+XFV5yv1KpH}5%fT{+d54cZc$ZRXxOVfY+4TApYbN%Ly!DejQ%Y|Osm@jL-JE`?^q5#VH0n~*qnI}lA5SxFuItUy ztyPW{BAIX9_3t?hgWK-n4r1+mT`J@wu)Elq*(g6GM-0bFAJ6Dj5X01aB*b=sPGT0C43f6Kv@{k|i5sbT4rHrO@18X&)o{9NR zVBD=LVcwXnD)G25M0i_O>cK=t zg#|njPnWV)tHpf9;j`{@ zkhqbe&vf)!d5XSOrJ@f|m7*=SKhs+G($Rp}$of|0OD1n5`~kFmu|ADnQ0a!))Ap^D zwvRQ$Atx^T2Xyn2(5oI6mph^*{3kJG6Mdh#l2p-9H$&9ZoxZQ?1x^3AwO4t&G9o`o zLyI1D(PFP8-FWPH4Q?sd%E>U#TVF5GD5M^S%;;sZ7caaM@3C^oDZNwJvLQ zoxCJq*RN52jQvG>A*qT_WeVuzwJu*s3TEP7t0Lt>tT;Qz55|D1aQ+^^&a%6jDM?M= z%#l48X}rFTk3m5EN4}XO`!gBTUXEays$00OZhC(C(W85BKFC~voBerRxu`3a*NcW^ zl*u44e=9!WjCO^%yd#`38ozJ0UE<1gi1>$z_!a6J0^2!7@F`3^4V?g476uI_CqFAb zb+}Vz_UT*CT)#R~sC{yC`u>fRFIchU^Wx@gn-@za0uj8j__Imoz6$*BO87iJKV6ak z*Xpg|H-J2m_|a6n)?^elk2p?9Wcoz&+>d=bbNhtsrY7z^xO@KI?Z;2PaXy=#U2Qcm z;s+B!-gQ72dbB)&zZ*q+bm?T&SMG)%TFAHd8{&g09o+(3Lq5*&*St+&$16WzzyWH z8h@urJS+?e&Q6nh>_EM-32alFqwO?%=1zmaZq%To?MVsB9oDpScbW)qGuFdBXQxR7 z%_ZzKIk1h5tKjW45z*#&w$YhuPusC9^_46Lp>&b1IotnC@)yY}B+v1r!RqOA_TKH* z4Rp{9lV79y-ujYs#d8#Y`I*5mCGV{$Nk>b`-QimpmGjywNne{AZzb`;jJGaw1BwcD z>Dby1s3+82E1@o9KtpDY*(AV4$WpXekE^%VE2!_B%fJ;aAQjJU@QJ%jRJpn^Jpk8Q z9;*XK-Oyrt09J>ESBbus1=d3}(iE>ZTqXL}=hLXh8ic2uT!&tI>&BJnYrhhGYtVF| z5*@r5)Hd%TDs3v!*Wg1E8gxe6tX@pb`k)dW+|GeM=55Vf<+mQiTo+Y&3y?IE-}}!Y zk#lY0K3DU@8ny0DUbGl#CHmQ-8Y5+3D(12HDB$p}I@LMc=*sIE`>W9{uUCyWT83Qt z2HjRyUC*%9sFc@Wk!Y0w)u->G^|SfBWwF!_)h!=jHA0G;fRtPD6Q=e!h3YDZfr4)otO2 zP6#cGk+JxOEq?xXYDa)o;9-h(%1)6Y!AUr$83 z2@PUhz)Nf0iGAY^V!MU67@J6fM1<>y`6QPS47b}GJeG)|(kWvTNf}DPq~m9{XI3_Y zVr(J_lX#m*;z3~$WU39aQz!MvVG}{bz%6`TTpAdgh|fe@yk)r7tNhukF=@3kF*cEi zidL?A``S$;5i5j+Q|`jS3g2?)$*@w!CK930T6eZJt2{p}JVe#6bPoD0!$G$mJXGpe z)_zQg4YS5Lo`ot`=^)f7dDotBb*+S}6(Dq1V}jLYO832d_pYH`=BCv(z%cZ&-JnWT zIubQwyYZB$>gly(t8r-21Unl9knqY<=6WVtH40`@Y=Ah^OB*t|VGfKM3Gc>;GY7xPsu#3j;&p!5oauPkpqGvYXHj<{@6Jc0Rc~ZCsf>mv<_yVn+2O7fb0M#awZlXh zB)mJ#N!8EP%qt+_m3>m5{_6Ri`>)=!oX9kS*nmb5#n@EK&4v2pvjHu0m?j@C9uGeBY>xgj;UQqnSL=Yx9_zMsAB{1FSS1RixuldLr^Pc#< zDTPP@`$D`|=bic3_;8E$ct=)$HMtf?|3v)BM5J#*_rL`7U_kWhJa{tRZY&_Ie-+=V z8mvSLZ6@2(wg$;=XP&C*mV9ooxJ#j`Q3Whek;4 zWRA#6od+=`%(aHDun=Mc)L6`{Kv5SVvw|U8^we1D>2tc2IoeO^)PE6gvmo`~*%awiJUjbrMT(Tkm9Ygtoxp6)uV%h3&#l@uO~$|JMxPNF za@$hJoWK~f8Jh>N&PTEmHGK<(PIiLJd+yx5b8-737HF8n0E-ng zv47P@?M<peAZ z5jXLK0JV^9d|eL(#NSMCkyJAH!V6pLJ&Nv~Zij8+7F8HnJ%EXf&rEzT_xP!<0(;iP z_sKxZ`KdxYY%JPh2={_B@gqV|9j5#GiJyp@qfdO+ZpfMV5izLuXxq7s)}~!^ghG~U_o2O9X;?0 z(l(|&))sstrg>q!h(I9*JDgijM6gvNg6iSa1i^leMGXN2L7ua#r5EIE&k%$y_?{kT z$MpnZYrdx+IGTF8BM9J=H0lFO5WI4g48Vf~&Ht z-cc=47;)igi!)jVNBDNSg1c%vEm=V8Oi{vd(!$QnVN1xW$yaGX39k^%;|1)y<7+m6 zpA);;rXqmhZ+i#<-O5!L1C*z`83%vB0}PGc`N-*t_~WTa-}M|3254!B7#Pl)H;Eo} z7~^FK5Qs+OlSB~W!-?21CPyGPV0hJSwTEjtBuzHB6HXRwL+Y8MQkD42IoD^ z9JZcXK|OuWT~g6;vfgsgp|NCP>q!>2m68Q;j>nc02OYe+kFAj`Y-=S8+eVTFJD#3o zK|P%@o?Pl66RCRYK=r((4nViyF>^YC^P1pnJ>A0A(=AZn%~A)i%wubG3tMV3ZK*@& zQ-|6MJ|%K%s#{P%4^W3rQwQ}kc4DK`-_tGh5p^(i9b{%_sY6<~K*azLG*L) z;0@KA9O5!`0S{%*Qdgu=Z>ePuh3Sn+9-6ozbMO-XJ`ul~m=gWzoyU*QA3VPI_Jh+$ z=l6f(^qqIlA3iyK>-^z^^ZU9_d`uw1+I5qSWEvRA#x{rMHea%#{E~QK_IxgsDenZy z1m4I%Hl_GVJHsgr4US}HzOlh~u8J=PCQ&#c}?B^ zn>uk*0@@>C-dU*lv#GqerQH!wlgWzLNM?I8-{~Y&MM}(UYxg^2WXDm(!EI@ zhLx&mcmKyEc{?g4*g|y<*WD*A{ZNi`^HFTXW|n{c}hJlt{M(AO?Hix zYNkKX&Qq$Xr#qz@V2?$&xOnj@Aa_Qtw(H8vpmVLc;!`>nR=sVj9%6)RTX4lE{bU!F z#7Xcw^sQ0)Fo2MIgODj#e1h^X*dJ&aa%3g*yQAv%pbo4&2-4&W)gNewUNI3}<(d7q z_M5st&`wL@+|!enF5=18VG%9b;DW&P2~r>Sl)s%6Pky7)X4?OCIl640pq>6u2>JwV zmo*Q1PqTLU9L3sGH!+qyVN57rFA8GD1GcW&Op5lSZ=lipGCml&a<`LLEU(;&&1}Wu zl^OnCi~^P8ItcHIBblcgL*ln+momKtVvxGs(LDp~J;UYw%ch8uX}AINa^<`nv-oFV zncSxuiD%*%wp_{f8Nwuj#5+#QKkuhj=qLAoo`{OY0+{IG(J18`Q81{YA}mO1<{9x< z(;7wx><#6DJg+5t3YP~o{@@eO9iR(r=B~_IS&FFCh4us^1L+{o!|3Nlx1~07ks^Ss zMIr$O%vZb1SR_1-IYGT9xQj#)JGB~x#(j;ri$nyPSRuUUEfR@%x|BsCx4j-^k%-t5 z&%@OF)r*T3E-bXmjiWE9E|0D-lv&ULe(XG5-cITAMl8PeeevbMCjO|ruD-Hy=gI4K zN?upd7YXD`Xy_R^nUXX*inKF#s_iO@(1sxZlDRPe$quO}ncFGJ9BW85J;Hop#~WO> zx03Zf_27u=d>cAwqxJgg#+{a`4VUfhRHJp&Tyx(c+elIP#kA| zL%gHehn&xfeWb2R!`5*iHQeopZ~m761*~1$ou8+}e3e;f;Zfi+LVL&k&(CVO=LmZyHuCkM5uLT_IHiO|}vXvM1;>Q#5 zHA=9)b$%@AA=j4ZjUep6g@@mrh+k}Fw5QadBm9Fn<0z(2#K%)wA{u!xrVCo7ajX!T zw;hHuXm?X679=(s4e=3?r9}hav9z?74@NmbEzuI9MXwtb=ZO!AR4sZGi>gIWp(V`G zpT%(>Z-ou0TJ*b_ZK3=HLay4RR3Q4rP+myd*P^G?%(Hg0in~=Mcu1%fGrX;!l5b^; zhun^{Ri%vfwyH6r!e)M8Z&eBN#%xuI$Auxn+p1Cz9>j#jNq!ky)t*Po#>IXdgrvO)Pg&myIgt6;sW@?cma zoeT_Cl}jfMTu;79&BLreP|+J*Ht2>Js^0oDrFt7Nfoe0}qzk+9q}9Ml`6D zyMPC;G!IrkV<%5Jjsz49t|KUHiqEAuw-_AuSJgtQ)O7tB+K_`T8hkQ{2Icegs$-u{ z!^L17OL0?`w9xE_*CKC;>qs7r46@NhkB_{Nn(+Yev9Jyp8a7JJQ-#XQ3QL z_l$OJ3Rp&4ioQb!u$4*tldh!7uTiNo)RVC!f?u#7h@KZWXIt_IA{DRT$9dn!IY=b<22NsPz&0=Kamj)dkhdK8*r*gm$5{aqya-n* zGna!16fuH!&sh!;4>on3=ckmG>SLoauBg0j%QoXcw7PKkPy(ydD|iV+p*$C zLS`~Uh6}ZF{h89ru@_k95^QEVQ&7Ph9R=y@5c@MIc$5FCLqSwkl~H%8>cT_TtKDew zdTJLr3xH4s8yx`I0fh>-{!A&@7z3(Be?}XH&U7gHlZW+s@KfmOZC=6!4q8u5XFgvt za{!i8_h+h{+KT?F!@^Wh4q+j&#qeLKiZCnsTYIEuS9>YquJNUasY}VOA&Wem8BN_^$U%(1JNIK8osURg&25S0nb5#nop1f1+}dE55C+jFVxj zK@@V5M~5cBm>M0LIO9s~F}S~T|J`$LkHMcpTc*VOLZm@&%anLftO`2y6cn+>e^ke% z#2Y|U#`hRZ{6K7COHlbne7S7Ikm@N!Z>d;YqGhd zJ_og&W#N`6_?uZoeVMB(vBzM-{fW(txyfA8Cb$>;%`8H}Tz9*f*D|omo0-9g3shG} zmkOHpbN#{-)~UzS+Rw!i#TN@MP;nfc5F4@J!qdj7r`L!DK{;x9Tz)}!R8xiZ4ihhQ z&$xb3+wSc^5f`Y~jn0bQAbV;z^?2?ehrqb(As`o&7_6QeR}F`qCa4gMj*g97aAAaC z!}zG1nJhZql($K{YFKfp>RkZhrPus;QPoS$$^|z*Lqn&OrQw&>@UL}gB#a? zj?`lizJccN+YCxA25)>Y0+^TEn-IRB*J^L$8VfK_Hx|ZI6wnyLZg)1yNe5wb#f@t? zz%<;F`CJM9xS?+6jy<`3*bpXAb;c49a`lqY&-;)X6%J6_hef~PWt5(30{gvG*4WPM zFlRi8L`38sgOR-<5G4s`qo>eH%%VR_=qnMQw%7`Kr$HYbk43y_#xBa&B z5atW^B9UNO4s?F%;GNn3y7fPR9Sa*hs=?2vlOxrI{BjhHGf(y_6HaziOGss~54PV3MrY&tVCa>LFm z!d6dr%!vAKIxsp_&<${rzQUyM&d^QLI$Bt{;TDo2`B?iuK_8^1mx4r{EbDFpiT?M1 zP37}Ex#4zzM2DXtz8UN>xQqCZbd-K?)IR%L!pIFfjl~1GVc~LEyBwJwgL2-GizUJO z2Ly*5DAGiy3+yqtaV_acb&~FucK0JJf z_bVK;-KC88wmXmf)4%H;P-atqh%wt;!pyPTUE+~p*zmWz)WZkmeQ_?hNqF21aX)Ii z^T~0O%_oJ7n?_*VdUwor7qMjay1VCZcZoP_dTPp~CuQYCVwbalJ!-p)05jk1_Jta| ztT@FjpzVOxkG?iX6k7bwL|658p4d-4e*OagJXf#2Ib0#qcR;d7H!I@N@^Xrykht>{ zd+O;m+!<6Mmq$iR7`rpC(!Dbs?7FHBc7{NAr~r>npxrQg3Ox0A{xFXcG0GCqdKO*V z!0#Y4?0`HiPT}c5+K3oCBTutWcR-pKFf&2N=Kwiw2b9nCg)yQ^&iqfH&VJ(>&hWVV>%{EMi znfW|Wo)U;^P-VontH8#!OkkgOG~|~XTYNMFT)X9gJ0n*l#J^{}YpK9C-mWf`?Hd`D z#s|eDK~nVli#0JlTGq%=WO|Dncmc@q3(e4eG!a$H2ly2awMG&``a7gD&X&#yA+7lP zX&vVSHivMP>$_sTcW!7g^-@cxB_~DYhq{UQ*GciJ>a3w^x1L@9qWrM-_}=3ubsIhy zF}FdtbPx8$a>UY+00{<~Vm%bwS?2N5qb(giwhHHGEJyyNtR?#qOGgB%m`g{wx;8VT z60sGZLB_R;$o+EWXogNA{?ZY#DXyoFv2+jxkR8wrGQS-hdM(NfYRG>*&7gXEvDz{- zm@Elohl>9q-dbyVe*0Z=`v=A#2B zGpN0%=2K7S4yvY4G`IYQFm*9%fl@E@bJc3OOT!l%9M_@f@QK1SvT-C{6^qZ7hP+m77l=Ad}=BT5BnPhB$ga z1Szi<4*AmB7)n*08WCi#(Yh~!TvPr1NRty80+a&KcnwsiQq4_z&>s*Tq`!p_!L{ov z)qX}lu7SxS-nIccVAOC%Kg8LuWg=-w$<`IPGWTA~N!F;b?GC;$+`J0;A+#79$cYlQ6~D;eX*;vJTCF&8f?An|GK*Lswq zG$BL*U^41LE!r5*yY0TUUyiuUBy!!$UuJ4p%D83ba(PPIIFJ{)j#_3St}-In_hXis zM9f{{GQ---$1O7vXBp%9V=Xg8-qHiDOss2&q&`_d@!LiM)O%tX_4xVd72Wv=T{ogbffB*#OGKqpo$ebba5l=NmhE*QE9}; zhPofjNuc~xrqm4bLLCF^JD}r;X9a`27#+tzOVo`o79~+Lctc*)AboTw4NHTOJ1OA z#Ytg0)Xnyl#Wyv1k^DxH7ovzDINn0~LPcc%2*HGF%e3TVHk!cT;*HmBvif=P+-%2A zRyh<#o+BNFaTT{Cc4kXr@JlS)9tv4B#2 z-nqCKhh@HEBmFh;%IsxUq{FwU;4-qw^Ydd2_P|VYJ3^Gw_>;Ey$yB^D#`+M=jB@3? zapc8w4ZxY^L&cyo-=2z>M|nSxwv6(=-C<9fy9Parj$CmP*)oODG`6t@(TrSid`6R~ zOm1#k%SCAHiW515Q?9j&3hn+*`JENf|WGD|9iusWH} zr!|##lXi377E|dr(7Xn}+>+5mD*bC5U?WeJ&RkbcHD-f0#7189 zfl-Ub=S>!}@vL{{Rb#BT%myFy#@4e=bw8enuO;)ydk-F;KfE}Gz?D`EFt#PsQtv##FafIaY{dzIB(} zkQ)ZKKgA7G+KoY}n)8PEXd=?>7fXhDM9@1w6@OH%v+}CNhv#>1Ke%%tyQa-rPOT`9 zmr2jzV#nDtH^rgnx%!VA;=^eMiIyx6X&}#knV$W{ACK;Uh<7c(|JsF88vGTwj<9vaw|&N#FZM(zo;^ zed_U<8wUz;RC_5YFsYHE&vf)!dWybfrJ}E*8`2GP$z-SQBXqK@rIGb5%a=^LMcEB8 z(Dr#J_Dub#TF?${$Y;;o67t5ND<@AW0ibY0>}mU!O54Xe>xHVm@>F$Kl1k>rfW=jV zVTo}geV^&(we<9TOHbcd^@67Vo4W83M2ivmEz4Igf@m?_yrg42rC-IBq=zYZN8L6d z?~Z9A_cPtR)Xg*iAom6!t1C&1QDYZaubbC0qR9G6Qu*>V&y-Pep{^udqL=YRSC3}3 z(NA>ytECnax~I?kN@SukDQkS$Cb?h?oxIf5Im+8kz9<{SX(MZ%XF7Q;^_N4?$xB|0 zQ!#FPt%{TjasE<{{C)xfd)4oL4`OH6+9`cANA~wMUSGz?AW)D>zGfr)Ga0U4bhA4D zIJoW#Z&{}qOVwAXwXII$IhL>Ao$PL8}K?U_4G%GAr-X^8Bb zx6>p{_ELq@twH;sm)TpXYMpU>~8cfG3p+%t*C`D`1;G-X(GJM zSP%D{ohA`9nWPI1qDXG!c8Er+wCuph+G!%9&GD=|tF;P4%$tsKqS9cx-dXaOLy7Ft zlLjx9G?=}2JC~45S)uyg`jYgOC+}5HuPI5VX)13BD`NFZ(wFAO+ZC;$ox}JmAkVUg>eWJ5qOWL**Bh=9ed+OO{32yA7ScxtG~Izp^i0Sk-12`m z-&uL#CXBXO-8T%-^irc~^)pz_H+VBxiN1=cw5ddw*9l{bhh1XqMaV1tAui|vxU&CC z^m*cK&0OWT9>rW2Re1{_t}9eb4czeO07LFvo4C)_{IEu?tCJTkZ(50dwy5S!P4;z% zch#xRGS;0duV?J9mayWL=56-g9ACvE(YMuA*E4K2U{y}?=+NYx+v?E77<2Lc-0iaNiSOqM zlxs}VnQi*6;hG0US#4S%UG>Nzqcu<_k+in(HS(?H#zA*0NjxYFf=snRcIu=ac?p{cDp+K1B0dw1 z@neG}kDvzD3r5`@Ch9klh>BLOdi&ZKW7ghFfG*>a1Z%n!?e zQ1vUFgMLeM(3PitT`Bb|Yd?1WD4qu_(_I1ODxK9tfYjd2yYhsq>ghEg6p)j$2}YZ# zE;$C+%iOfOLT|DR4jIlLK#5BCN7TJj-i$dg?vMtNc4b$(c_4IM?Ls9~5`3Gd2Br07MG;X^DZZPWw8 zNO(KH2gJ&7Qm+R@d0u^0I92L8gr(~0>jCltI$eIIH2I)zU4fQ4^yw8zd&MIy;azQK zS7m_A-P7mKk~2eQBFd*LTs^;J~G)R_fjRMJa z9x$&?yXAVYL3{vVt7~!j8LIC!MP8pgX9?RvoX(!DB5WC&Sd7N_4CP3t$`l<&c5Oc@ z&n5;TcO4xRA=FBFA2?gpWb3|6^Jczzb zyP3YXN%tC6W6^<%%nA7 zQ&rq@0%}Upx3}6MufSE_v#C8v?TcvG*m}G^>QD)7htktWsi)TsWk^p@X8`XoI@GP13>xUAjg>96%_DRWx z_#PILG`>)_lf(ds$<_vv@Gg|yM0`YYrGbxZiHvXWI`IKH0P=;nf{9KJzG=vTCtp}A z`2wmm%`c)fcQ|?gbO*5^40_<{4%SL{P(crt{}O--#sCf_b%Id@|7SHx z*q(geCS|=?%uXwlvL?GY$VTPK&CrD9nX9s^-cc=4C|q6?ohaP9oU}N`<*0&Zrz^Ot zw$qRWIk8*3ZgaXIt0rHi1tq*fG_L=kWErp7NPbT2W}AvghQD3G7i1R$^wWr&aqtI9 zfT58Y<@BM#(m$Sx^j*&pVStu~h=Jj(d6Vdg0T|#2jr^Si3FvMuKAeaRV{!yy1BO@8 z*K+%C8_x*V8j#2sFKrE|K@F@mz-J9|4+=WZoHX$~MPP93H6Wr5_0hU#tO1GWIsO{J z**tk`KtvoEUTr+hiPwH^fLg&*-_Qlk3QUOiG;`Q^Y6bQ5Id@5=Es7R8Y#K`zHlAcb zJ-r4xQ1e0q9eS_sV`C%>8*18Qa_9|^EX4UC9Xi5TEz)B($&Z1fADTaaHb z>!|}ZL$n@itXtT4x`mCWTd0l-ugqg(bPF5M_8oJoYl5@&sRP<$<(eA{OC1^&cUbD+ z7k6w8bx_3}m_hE%Aa&S<)S)+Z9X(YAV5q|;ty`dCfCrj7NOqy9!&%K?5@3f7C`ZBE zp*|V8?A*0ZLG=2J=$q6q* z7tomYEp6T*5^?1sm=i$@$VDytBEPGJiYT+c58in@9hVtkIwJ^#_2on zof122CbV{Ege!e zoxhnSU`TE=yR;3q;RykrbDeI>|0{wwxF-Scw z-myFglxp#rB~qkiRamA!(8g1$Z9JtKFJ8O~$c>S!Z2%lEyVmvx+WGVt2qDW4u?V^G z6`#E+>HFn(9zMRk-Wh~Ux#ANLZqOe{e!6=4tLJy_zk1JtG;_sgkM>mx4=vK<3w6ck z2EAe;>I0OGlHb(*fi_wa=bk=K-9LyYU-gUVPy*QNQbQ}nQ{SkxnKe_$^aUOTloRSr(PBjwG#4&8S zlI=5ueZe}xnh+5$KrO!;gz*#c&l6FxSO61oJR0>1@qQKk1f(*fP;u1Ec=1=$8b$}~ z4dsG7uO)j5m&YuYoTB+-Gk2v|LUj04M5QjYCm5-Afp^j_o>GIUm=PbWo96mrfw_uBoMJHQSURPh) zxb@_9>gg3CLzy{XFKtb8MQmPW39pSq|I0CnQCtYhIyI9a=`B^@KG9owDdlPW&nURT-4 zvey6(IpHyU(aNL@yg(<9`0+%1jS{SHowKEf1+@m({OFA!S-`a#-<^nGY-Kb>sc{q& z^E!=>r!-m>c`&96TBUKU5Sh0f2D=p`#Dwj}pv*}>BC@n-06doFogs@qs+O2mv@CkB z_zwHD>`{D3q-xPGu&7$}6k5U@{aGCMnXQsE3b{-8PMU?FTD3{pU2G`_twm39^Q@Uk z2eL_df**4Si;K+JTNRJ(hHEI1?M8o=*s7x30;u8Jg;FwI zDi11qtKtFP$Y1TZD$1;~NDyf3yyF}l2_EXm-#5a+ttYbEDv{lYgMf2R@nGYC9gxYz zA}>1YHL}62r;^($l^nf`9V`Sdn_y&vORuVYyEa^}R?aklcU1Hi$pf=;CO5>M>TRo3 zZ>%8>v2WQwPBbW?9)tZZ%FBU`I2TVq4o0695f~OmG`KZ#uwiuM(;Wb=dO;g4UbXpl zYoet}foMd7(y<;21|{p|@v%I(QQf(@RB-9loo^O~2U9N<9F%?niUzw73NKCZ`4Uud zlT1FS>2{zYN4)CLMT1YYXmDp5a@eQSa530lxRSx87A-r;pmdFX4Ld!`Gf5c9;7)(} z1Cqfdz}ICT?=?1;t0Ebj2YxUHR86q&LA30e%aPc*`+d#M@iKNUJ9j_B1p?~);{dvA zMnD3}?&o#AxA}sETifc;#}^8EM!PlzETb(&2Ww?33vb(d8`%0kp%$R`8Ys z9~*foG4^XIG}I=pV=yGY9qEn^Pv891p+^OCIfy_JBWU-Wb?Vlw>T-Iqabq~;?5Ji zsi)VVppN+VR^84jW76un5b zrjJB_HbnuFzg^){Kl8m#^@(-~y=zSOxtde0h8BH%LI#4r7j25S68&sZo#JJX1sc8b zROgIyMVyTN)rkF;>ytJ|h#ZZ;?7{r2PO#2aSH{V()gp=w_JfW5kYg5@AILoh_jm5U zd(Q1KSWr)J%QO@31hECKpP;s7O8iI+W=<&ziX)ks*HBQ~GLcNIY0E@BIEo_G$eEmPtT zLw|wQV-Dy_|gKNtHrtO>6$3KhrE39%6i?mTVWPHE#R#DYCx9+*@? zSoxP%9lkR?<96GchZ#NS4v5`iVoqnpZje2-n|eI0$6!3j8WNwWHxvT4=}o?LJl z;bc=oz5+Gmq$s&KcpL4mL6-~a9)s$R*C9&m(`ikKylJL3Dr59&(ngz<(d9s2n2&G@Rg+OWz%C&eyAe%7>qn&OsM*0*kf?v8qkq?48k|i{Cyec z@PNN_I|#xb$}sq{k7295g=;LpJWN^`PwDF#sR4jb+MSIuINxio*tv!SOv4SCmn-Mx zn3Xx<5~e$~$KViNP<58=F^CBWxf06g=Y39yeubd+7>s_y%P6CSE)wj}Un!M*HuDU* zeOP2~2t-N3+2|>>YUt?C68U2^cL3^a=B~_IS&Ar9aZGGvkT&u>?6%)_UO?^TE)ofr zWlquFkr4awqx2XI2Zr8bF#1bxk)T_Zz)!6PlO!&rHX@DeZL&xZ>rmq#k2alngFV6` zk=tIexFt6tjg4%vNZ`7__AAKV<;ti&kB*ddL%QV|)aS{6N!T!A@x>*nmcwXtvOEJa zJ8yDlz9b~FSjC%!78}avd5OQ0*;SwD%t*@(muIjDTmDNr&Gs0izMBmUQ+qj3B7KF4 zd?;~xSh?ZSCq>lvu#jkV`ox?FVjn5~;~%Osvrd@1j90qKQOgaN1`^4iAzB7|46Z^x zw30LEAU9l5b(vajxIEKXJdhhwA7gq9O7qCYl3-s^8;T(xR7EK4G1$45bfh{-_tP4A ztk``7-J=)>_4sh$pn}(PuXBZtxpB;}nYSO7GUaoVF4lK69X0Xa6R~3tNIsP=lIs*v zYib!)zbO7~DqdoHAWW|LzMQw57wHw25O>k~50dky!y`VP6z`hOudCtp$9%V+G_{K0 zVVkjz+3pgYE+|QD+g-xUvD;nZkzv^I zx4YEC2j#s9`Gl^%t?!Gw&ZzCqFG?@|*0FcE1T~f!xs*tL<9znaO>rnmKl}$ggic0n zcadn3z3%S$+g&2g+;ORGJS8Rdgf%|iMMrISQ5`$q?Z@&|aW**G477esKzrzQ_&bxK zvVD05#eVW%60^O)KhGW1H$`i#8=YCGWUpGnxIBZ3J^8OX?hG(h(H1LVT)|3r=|}^g zt2)>j0=+;5cyt17hgse#0Tg)Z@%&*9F?88SK*T6(dFxqp)qq$ZqVhC4kTxR5g^{Nn zAo4<8{$zo4)eBk?BP@bnF65oI`pI^Flov}UVqE!Xy0*>cz$!&?!fUk<-Ad|H*m z7o)~{!4omABS>v(;#aj|;e)zwfp!X^ix_`v=y|wYm`)+~>GP7RDwU`m7E*iiEBeaG z7`2ev$rz=3^s88hmVIyf9xfO9Yb2C0Qsiys)y21T=K#4nDwNOlg)yQ^UH(s?_Plcq z^1Ax{O>_MNLjt*5WK4R(L_`& zAK-62)Y@3d*wDm?x{AM_)^R>ya|ljwZj}D;hpn7V)g;Mi9&pWb)^a*tf=(#A9f_m^ob-!g!Xni@IIT;b;0+yFA7plD6 zGAHW0+0ZZ*rNb0m)ijIX5L9V;DBFMF_+|(k-wT4XUXWXiY zz8MrjZbH7amR#-t$JU^T5`CNzLCV_619J^TkW@wDBTY_Z2v7<@<24Y-AQxMrEB+Qj z1lO*wRQnnIxCSP`Rz{&`^wbO+!A2wAxUZbtqp5gVJbeFid!RW4_V0= zKNjz>tc$sLg1OA&r1h9ECYh8#5(R+Cs0+1dV?1wHyXJnumfexdOd{94{AH$wrHor< ze4avw%tv34VuF&mqhj1L6LFOh!M-1}%p_v&5|Rsxgi|lCvOj1v;fi56xK^K0TS$lP- z3)79NJ44TFHjM1ZcBo>74qfcX%0f2i1|?Y;F|u6pBGbrv@~`jaD%*&>nDWy#m0lY1(m~Y<_yzQ3>?@0JYVsoajUq2Z5kYXg1@xusi9h_0 z08I3*Ez^>d*=Pc@J$c6>^5YYbGdG`F&dYW*kOk(L7yW&KoW|a4TIxCKA4%Ri&#+X**ODSWAX{i#& zj9=erUUHZMC}il&DS|k*NthI?Lz^lUGF#%?+hp9U)9Q>urUNBB*cr=tp^-wy4Xr1U z>6Aof%+PXajr<;7Piv_10!wQe>ts4lC)0U48S1;)0P)JQI-`^6s&z73pVm;Rk=4n7 zVl4VJi<*epW?^f;7;9^24exH!T|{f7qXx=Eb-1N9X`KuxoJ1$2X_PK0wxTtq&U(?; z*CEw{xQ1$g)E78{L2ogYegn;GfYpqSTIol(Ap$Cob>_Nqsxce1B{qdle3;0j@ewnl zyz#Afx6B5=(X_FFoP(bD@kD$rnMdAx@c8`U#rfTPcOIXQ>nsN74ch$#y9@4_ zD_2g))28(5;=4KhQ0XsY<&b|5956N8>?_xArJZ>nPsQtv##FafIaY{dzIAhlp;-TV z!(~}<2eEbo@X31{xcd1bJhFZ1L$Dz_MhN>){o82sG7fK|fS?f>&wuJ4S4E961g0=tg#|njPwf}uTq03iqEpzIyH!QJnCIqWc&F$G#}a|0FYXIZzeD1E zBkAirNnfWVeIpJ6&N+0G`bLUAeJSZu%3%N`R!^@{^Z}|;Ho^2~TI*gqL!De_WPSY^ z;W%yNPj-lVpzWj6yh!sW><~-I40MRn_OXUIZ+j#1>${q0u~(ArJa)XPxRUgSa@hgIZorkKvCAS* ze3A)EdnM_{6p;S%I}ab<-t06$RzFi?<5!*LZ+jL|WOF6y29%YSLsFW2XmZt~m+{cM zc|FnHym~DpbWfl6mB>Va<+=rQz~~!Y)@tR8TUm&zjjVCExsbHiUk*VhuMMbZE&F&s zs3PT<9XLD355|CHjRD_-*x9vqO5ed>(IRhfd^;-r6UpxlQp$Q}J4pQPe!*I3bbg6U}o!CbV!TwnUrqMgt>$ zFcIWk2ZW(V%faM8dq&+*t4LS^-N=xLhesOP8{&g09o+(3Lq5*&*St+&XSPY?TmHf4 z?>;)eEf<`(?mc*OHvJ5!b#CN*A+I;za%OIzgynd))x!G!Q}?#NmK{g^@9ZGiNp)5R!=yFZqwI_y^E0 zNQi45!VO$Z^2;S4B_Pa$WN!CfUES5IyQ`~f^*#x_2yx`I_NwaYhx)8Td*GPBs)zaXyOBZ^KBz{+^2Rk;Z73~ZHZ@n!gD3_1$3Qr zBn=MYKXZ&uh+Ouy>6*Xno0H80X|Q?vM7(!ZP4&3Ph~ypd@_`H{Q;9P2=_qu^RwVs2kzwtQXvX214Dr5$XyCw9l+5-zeV$ zP@}7QT&-OXFrOR)PpZ)sE%Et*PpY93eQxEkKhX_{xuFFr(b@0j!Xl_dpRa8t`h3H* z5h~HOC!aGE*h+rayo-<#nr>W)z78tUH7T_OoLh;06IY^drhO{ChD!7`uS(#i0&Vl$ zeuxWu0B#W6CS&)JWv-gM8*THnDsMrO<~bsPX)`c&x7woS&P5AL7Det@FBeU8{}nxS-sq)p(H)%KW=v;79@6BlRbvjAA$ca3M==Y*s$I z{Z6hxxyB@$*{1Iru6a zCl5@|%e#w{yfFbdEo!&A-@(B=qP_~*e?3828~8*0@sV0pTH8mbwcg59>*d`WPt7K) zNxFkTF+k!K)+f>GZ$>KJgod!L$4hJ7=})8$;-MUff=whrBKqqGe3HuuhTCl-0ZVMX ziKGk-FeyK~-+!nZKA?fQBUQhNButWRB8dm}gCJ9FkexcIM_$1uLdt-}n@Gq+Ta5gb zZA5}iB%-2~tKP156G_AhwU9U1U~;GT;sztwL?RSg>(27J-*TJreIFL2`js!}{cX!Z zb+(Gcp62Nj(SGc$Ee{uf&;?Si#ypD;LP5#92!yNV=`|p9S}~;VZr*j!^I-uutu7FT zsqhD>MCBt_V8y$(bij*ubuif%s9C^3yt}a9!LWE2A>Hzby=y_TUTX|UuJZuA zI)dNjU;F^WR^Q_C(_G(Miad*tGoJh%zvSrKRGo}ptfFrjJYBX-<3^8>ecO-9vx!5< zD@H(2iq_$0>I`5uDX}-yH~gqH)bV4w%f)ar|Y%xJK~*8|tq|iZRL2vjf-{e7mQ9$IQ;YN^-cFQHmGcyQ%)@h(n~1eSN%F z=bg#W_&6T!l{DBZjZRKass3!F(l?<8U;=xvNA&7ENPRrN;q#KgZZaQ9ss<~OLSMyX z4iOgVo$ARPLYD(fQHHSWM#W-PJDDTQQWc83y2+e~kN8aDS^40Jp4-N#iC#UKBeGKG z!FKbN6k^_RqM8M%#^M7NofUN1VxYz{PoD@DL+=K1a9jdRimCV{1tevQKuonTVyX&B z8Ta-H&<>fxzkTj#9SGTRMPI}2FG+nApP#s))H$tw@TAj6EsQ>@YAE~kgd9pFCet}h zb?x(p>vR_~WM=?AGXXK#B4`Y-FpUA6n2h^wE+C*D*bXZM1WHUvUpg__EFdLJanyco z*h7*{;|pasw({0Ol9`1h=4Y4_9$jm)nfLjKspGZp&3qh)$(EJ^_rzq}+ePkM2B5K- zVE~_DW;{qnwj^b!{uHPF7tS1gv&5g_{iIF(3wBXC@z!NpC+fxVXDTwSG!-x>FrNzh zBIK%F6Ec3S+N9on@Ci6=Lh*#@?epcwRrW!muP1PeR0Uz4p)T2(`k?eJ6gJuM*F~@{ zfaT}M6c2dAMd>ZI8Lx^_ntVz1_al|og;J&k*?uBHX!2_1I>(wM;hRU9x;@Iye%4>_ zsY#2tAtztF8lY^V%|ImQqx#zsEs{#cTfa46y+{47*Ty*VkRh4)iHwg!1X@^ALj@!g zzeon!7*Pgz6?8x{@gqV|9j3ebiJyp@COcu{5QDpRV;SfU%+qV=!JPT_E?Eo; z(w5L3YspPIofk+E0iPbY;S59sOCutv9!@Di(BBU02|^(&aykU(F!ViL7&ZnoxQkJj z@I8H>^)Jm0G>#x}-^~R@P>#0LRaW+)^m)nI=m^3*BnZ56G(->@m!oaMaA34$IrLlDRh!M$}4Qv{?VAXmQs8z8i;$ctMHp^^EI3Aj_a;1$lSrn=ZOvQrq#mBD(R51!r^9 z#en4Le#YTvlmJ7cH*K0|{OOTO-}M|325e~`F$kPBX%angWTM2Ok-w85f!wXthanwGLPAdxY`8bE==Cic{z2GJT2vIfd|CUW~KqDD4WE)Ytz21K-> zK3aEzH6Rf^hhGCEd!}p+h=_y0tK`zuc!BWcAInB+h4+9t1ZoBI^oevy^)^!z(&;Xi zjU@}qK(b(-UIQJ_@sE6amtO{TA4@Gv@A6CTRgNx)ZX{W7;~7X6mPWErKb}(RpoP5k z)Pb8JI*&EhEi41w!ZOe;R7XWn=CRyBR8;F079n-0h^p@Uf1~1#|I-#7Y;;uZ1v>lw z?3)EphoHEFoHtmvu#BjKo<8=ynTrVM7FKE90vBF9wA4Yf3qu`FYYvkLJFIw(oNkzJ z#V$M7v=_2gQZLav}1n%LwJ zmnlymxAxFdU!+lQsbddx`9c#*MQ5vj9jRYUOo`>m+2!T=!^``R9-cfofAAY8Uw?Z3 z_}R(V&L2NKe{gbk@9g64MNA<2wd+0``PAJm3%;ul>hGRE89%RX+%g@*kEr`NO^2MH zb7B6PdS(1_F3dUai^0ToQ&0Tmc7_KH4dPn6edB@ynUe30^9R?}Tg^RG)~Y4BrNy>Q~cRfUfj~|C^^<~zn|XY7jCJ&;hpv0 zKT@wq!|?W#C->iZ=yp-57borJLA+xjPS(=BX;@Yqb@gwfp%LvD9FbCtbtT?mmbbNj%jHnJt0U04z@xZ#vk^35#5yizYpH?u@+ zWwO9>tVI&^gKUDk0h?LGf2HpJ(QIZO2d|I>EzT?=sadv(_BTeB2d@GRntA->W$F3M z5E8%e6ywH`;T4jj)hW<|^*}*epj9A8Gf%G}!$?w6HUav*Y;2c~jtZ|pleVh37SC5G zu8=S-IOX%5W7KQP4|1SPGmn=JvL(eu$G9+1s6UXd-l@mMRl@Qpl#J^_MTY9yYCW888@)Mto!$7%WJgOeKl zz2zTAs$#J~CI)yk>hr^YgA^-m=ZJo)tG<0y!|0H`eYv2_YssF%+| zmqwxF$7mf{oWy$1WpGe@gfm{CQ%6QWQ#;1p4!ia++JE5&R#b&C0!&U zwj}d#v$cUPW_14d^o7*r@sX0wg7%WQRiMjT8C~9h#g{xPB#a^Db@hFbR)M^3W#n}g zeUVV+Kt-dwYHu4!FrziuXr0fDpg+>uO4Yh%Qs(KN8FAkY)Ti5^x^bnk$9}4k`|MF=-h zgJRZ-4iYu4Zrp~|jVpjevg*clpAT)Jx^c@D2!hUVZLxT-KN7b=+1D-m&`Y6VAE~R- zsKY0c8mDbbj?zKAD<~RrCi_l(Q*(tKu0geUb;)X{ML6Y{vk@^ZJSig2ImL8(x zJ}IFn=@6xH5xu(l)<}J+mC+ofMzc2$X-ek9l=}FHN2?;o#B{-`G!7LaleWX)jxoov z{-WI&lyTulOqLc6K)}+{TD~v@ito}gg^pe~>JO$qWKy;0Q6j1qJ%yJrM}HQ_J>%O= z#H<@|x9H8%+qu%KHY*i~p4XQbvi7y;DSn<^<*br!RS6!_*NOpeD+mi)SsNfXa!(X* zRVkyda;pkV%u5QTTUElmAzM}Aas3dHZB?lUHxLsx3q&Q|szRchQ2aG}#b}c`Zn80o zw<;Oi?XRIswj2FfVykNB>2&_UI%D*>uG=BRTa^s(M*eayB2Kg7M;0C8DDN6+>|(78 zpYpyB!ohVQvNMn8F~PnO7yE;NcTR;#NG6v_K+FW)ojrJZK7Qej(x0ua9X}T_U?&?~ z2P(O>QOQ-XUWA6qCV*^k8&s9A;hMNLYKd2Vm;;!k=#5W({SXJLx3y8diOzZ!`$Dg^ z@`)CU1~tm6XTRhPdA*_&KShun43UEDks%RW`}TGq8eD@MtYfwM=^g-Ay`YPhpxS)B z018|sk$hd*X7uY21y{wd#nT>tZ3Ow26Y1iYh;bPK{B|u zU;eOUa9i`BoyF!l3fS1Z=LayLDxCigp=I4+0Lt`P?A-spW#=fWq!8UdGo5uad>S!< zYv%mp0J@g_gO#$x8|DyQwBPDLm;>nLX&V^r+Z2e5b`+hi-*07fPS{X3T2tjWxl|eJ z$+8l`U^}F~T97RTKkvRNy{vAHH|%dpW;&qPRNosZ))%$j-vvmsu{^MmX;Fr$J~-m@ zHRPi{#FY4J(k6!7wtU>mN$KQ}49jRa92u4PMlwlyl0b=P+q;A^-jF-NP$qy8z)7m1 zV4FeiK|9hs*>VtKBmM%H08Fh+?kj`!wPzFH0y;XP3PlIB>HFS|%Z!5)*8&bi>2S6~O=97My)n8WG90xBr2`z`KaUux(? ze;Z7C=BHdmnT0&_Q*pxfq3IS!C3+m?u%>_3dvkU=zU|l z&(@r3Ewn%lD_?b4+EUHO)9Ng($p(MTBzfLAU&JZcU#-|r2Zr}+n^Qv$OHiX0A7*;C z`Z7*|trk(3N#4f99E7Me(RrTTmTCXl=D$&o!NmLG<&WQEF!7-HQ|k#wK_M;FO}qg% zCHp}{aeE9VejqlnrY#fefsu}a)nkzN=uS!DmhD*d*K#b%0Fc|3DPded$%3@F__Ha6 zUrW2A{5UPYV7FyTSl{Pntb+2yb9il;5`P%a0}1R$`EknuNP;(ZTc(t?J(wnSx41KG z;++`&hIu|{`v&8tsn(#H^2dOqs4%_!dSR^rVPK@%`7nU5Sv-T#9^CR;(`5O z;q@4d!#(xr2C~40LCNkhm@qEb%o60b;cG^pZW@ULxjhCG);HSBcGAr(5nGw$kae`k z3Hur@{`?+;DR;No%yuqUYH3(`S936RgLHMVevZ$DUHiFBAgnWw=e3`UeQd>o<#2I3 zv;%#lIF3(!++<*|Q!@cHaCU+Zdm5R#B z1?Q(h(fDQw{1_FVmtE`Rg4+-xFUkvwTgS}GVib4t(?;dvRxYTY7m(yFkmR6zylf+q z3+ld%pZulsvj=b7cO=c$kn10F29<3$CSRBygB!BD245}+Jq9<>UBf+{*I6%^W@e+R zGiEOsoS#~`vJ(t$HEXvoQu90z44P|(bhLNC9e$$bLAJ-B>88og?CI73(7w>y~U7=3w&agRaW z2WFt*ZarBvJ=P_WOnH4ue1&lH&fJE>>M_`b7gn8R1%%}10!mIK6;np<77%RHT!AG4 zjUE$ZlmQ+kO-$sT?JSG9XsjFJ_85%p?dOm(ukos(;~$UWDYJ@|e4kAnn+*cHeOPQ{ zFKv{0n5Qiem!3s$EnXxNPyj$DIf_@YNQ7iiyhw!7!e}xED-k_z2V6Gdu7m6zgZ*HL zlRGn z(Dha`Pv=!8_JgNid`hiAj*o3gnH?V`f!NSwFWwmQ^a_-uGY4Q@sJ+w=yt~-K6(+jrKhH*x8){EJXYPD}L=UCa)Pp=E z+MYbi2Ua92%xZa^+|X7zc4e=cSXdUNf5r_EX`(`MLmj1kKBT*Jc5n+(MAA@@8=7vP zX7EC}p}xw3`fK;fx@`T%lc!(3Yvv8RSQ1SZbLzC@LwCXB{Uq)&xbiLOSap*Ar!DfB zQaPx&|DgXFL8@T(XD2$;{W~G$7wxuAc8E$*o2459;}Y%;?X4^c^`7po0Cs?*pmUB6 zXTHZ{!(&O3i>Ut?sb4FoW4o(~Pwt&Rd~z<89~KlFUvvuFJLE4w9g{Lg{jZVQiU%a%ZE3w9 zo&my3A=|p7&*Rj;9I4mD9_SM|xi2Rz=S5}*DuzS64I!~&({WHAkBWCq=hxMKe?qz2 z88qIdaBUqhWV=gny1wQNPU%;;-35g1aNAwV_+YyW$UmQG?5|5{ImSCRWV=h4Idr>A zJhC4)^6f75@J_7gCQz)rZ|d&yTvkK2yO10w=xNahE@H{zb$2JN z^H6^k=t1f!)WlCwZF$*mAY$B-okI8`#^1F)4^5lL{mOH(|I}bQg_PuEsYI=QA?-;0 z)C#G+j8QjUwXah4v>Ox!GRCd_8tMBU&Zku!Tz}i5d>sh}uRH2r?7CZ<`kMoN?9 zX-hY!=PWhw>Yqo&tE#hx_GFGyPxC%iw~WUDOGg4E05lCw+mzGEma6UIVK-pu2(eYe zcZ!Q@>7{@Uj(FJ`uyjPAN*K`P>e|eVO2k&t{H!Xn<(MVcl>tjf#HOU4b|?3Ss{>B} zxjZE`gOSiND{9gil*z31^ANk)KfPFOnF%9_1af&wivOx5kotLnIY#_ffuc}OOD&{M z0y)Knk=l73Q_wF#&mZ{4003&K8{;_%B>x*w1g4Z5RIKLPKV3SgmOfG5aFanH5v0!l zs;g(UKoC4b`p>(KtS_$?K`u>Q^qwYkGsBdZJLbd{FM z>+dI)oahkX6oAHSpqD}FS7E;U|BE4lZ`W6<{fvIx0u#CtQc(LD{ZJ-7CnHfOlAe_K z0$1VQ>o^I00dc!yNQQyXD^y@JztlDIcPNOQF5{1fh-|g9Irr(19Ov(sZy>hzeevws z`Qx`AJ-#^k>iOkaXX8bjze0jT{VJK$yzNlm8|u#ZwF=aie8T{qsDB@+mv|T``LKYI zRB{S3_EJoxu_M$+N9xW1+xx6!h#!l0Sk}c{Ji!n%Y3xSHVyZPEOaWjr>iSx=A)fbp zTx-8j6J8BlW)itBEHiC5N?Tv-1}-xpPvKh_1CHayhB1K5z-1=lDg%OjH)NSf#M~7w zGornG*fJAwmLZ;}H#?UwHV#jpV9ksn zrS;h0lKVsk>W=dDOTcl9lDDk*o|D z*}m?F99cwO^tb4G^1>Yh7eS5bH5HD$VAYD3!f+wTZ&~u9`HdkjOc6ov zc#G)^_x=Ar3?_VAriCQ4u>=O$`~{bMeonnK-inh?L6#7oF0Pq#q^B^x;&#B!>`2VI z%`5s%tnFTY5!kM;B7$Pv0k%8A$|m9KLm47E$^Fwvy|2$H!4;t>!PYsjDdmp&sM5Jbr_Ybz>FtssLq>bUrC?bZV zpcVPjR2g|zV;(;lzP^)-iF2XD6jP*-;Tsh}9A%q?EvJD(#yq_S)}RE3+yo#QpW!+! z?N6JG zS@vno4C=6ET+tNJ8lzb=gV)o^ESYZwjnXy6HngVHSuc}nAzU-%Mzj9#=Lm+r9T&3r zQcJD$8(3aL3n~Onw$i^Q$t83i>&^A$R6{msAU3np_*{c@B#qCTA^}Zcy)UmCVm&Y$ zy);CH5r?y0bgKK|NPQ)lN8W#UdH(p~{NDYu%kzVE7JK2^HGcoJ(i`DSf^M_&t(<4a7Wri2eEczP^wS8 zsXoee>Mt2a6G43%)g^52o{wL+rN-BvfB%O0kDFmk6kpD)Nc}>%X1-A$9%Ybd$@0EX zE4|m2o|5U=U;48+?u%uhH{+EZ5yEB;O>gH)uiEM3P|?yJ_LI4p;T`x!*=~6^MvK;= z1Z?TIcLMB3bOmd2zhOQ;_vdD=wM5>7jWeZJ!_FK-)Jp+CI?`_f>sm|M)9Ob(=K- zEG~CMyXhj%`>&C{&t6HooCf;7X`t_uMN3e5I|Y&76p8%Gu65nKmLWSvM6&!42awT~ zq=C?$f2K1nzw{juT><0_3P{Zr71YB`?gMPk{XNhjjv@iONh57&L+t>`r4s5)JK+v=o~!|COr4UG1MxFVw+Oh;>F zD+|TP+C;Re|2T;i`u zo4}6OK19A-4MVaa6D1tDsr2>c8_CS=A#sUk{S0blKNVKOohHGKay!jLed=zfFSxt^ z;`C(vyt;8~zu_Docozx)@pa)&lQK1+-*FDHr`&BuFEF9?PLnWAzSATg)(;8EPLq0U zN4=@A_LJyP+t5z9(}b`)Uep^Mz!-Arip@v3(?ob%upaItJ53^JGD#O1M3La8`R23} z?lck6mUx!j0{skXw2(PUH$&3kc4xt-7oN*of4SZ|_8`K}q_|-gv8jmKPY_3#3q&Uy|+yG!W{} zj8IoFpk;0IO*h)O2cSj__1b2x>)~V0Z1@!|`+FzsMaavUmB-q^F*mfhN_4Wr0P=wg zi=YyHreSM0%@8vo!^>%#S0SI?ICqorWx}e{Hls@PA!yoZo5M=<6`*NTiB1{hUOTu5 zc^y${s6?NUmE6eTplvqH+)-DLa@WkadSscap3R22POI`3Bx%Mpo$`nJON=3Xo4C)` z{IEr>=o*t(9e7%*`FL6lo|%6XA*Rte0F0?mlHD2UHJO3_Ro_eJIWRy|t z<{vI(sg2FbXSd(U6)4x3WHa0JUBfjGin7`-Bp1G_M;3KlH}@i4w-qHs$qhzB>C13* zbIZqe|)4?mDcvr zX|1<1)p~jN##1w1&Ujt0iKGk-FljJ3`RbBAOC3JCJ5|4lButWRB8dm}gCJ9FkexcI zM|PV?3_Kfnj}j;JwkE+QQe>hnEQmhF_0_;vwTVPjv~tzk)ovn*SZTR%3WN{V)iGaH z^_xh9LTlYw5!($;@;i^sNcAh9gZ?&f&_MkdXjN zE0ny5DwOwhQ1Z@$o)0rEr0cNPTI3H>iONT!ZfrLJ7&c1O%CW7G@OHBkU&yN=PdyWr zIgS)(@>_I*gm(tw%x*CC(>)lidO;T^LFb3r2Ev4Q>8N$BlklzsL^>z02)2tRs3ILF zoj2+M0TSMM*aKp1IjQ*>zipJ~6_)UB`ml5kjUeW;%IS4sKKaa1%bfTV5?{jm#7cP0 z+&2Wa69}89PhMnahHUY!9efb)&TZ*{C%!tEta&v~vDfe}=k_}o67SB@Ju>AHd)I>G zO$UvQ)=c21I)gK*kh!nD~kN4`lGZ`8mB-$J9u{7ADMo#_NNTqK=55NTW zV2|k4d9XoPQ0BPfOZ8+O4rkjGKysr@RL$Izu74E%Jei&IVhi(nrM!9-p7M zp$zm;mi)i zWZQ_3z=g7Np0e(`(}wh z!~02_`fuZH7A5vumrR|g7ssEeNT$+hJvp>Wr&B)X_eIE6yQa(d>5F2A9@ldDJ{cQ< zeZKs-%08Fq(>+Xw+9x_6=}Oe}EfhA{>0RD)cJJ)sE^z@Y)%X7m7nrxyX1po}X7Z=` z=V|r#BbC;LQlD7^nv=Q% z0X@*)0!=p2)fJhi*U$rQ(gBr6^PnJY4(+k#)d?{vBH+^lH=KcpU~WVN)x#+z2)Znz zo*7F2P-^~Su4#C^^?g2R1+W)dZ{h`RI;RwPu zBnTBjwJ%50!~P6*_!vnUI1rvwGhjhEf)JLYZGj;0FGhV^{U4W)pX#>yySvlV(+D9z zJ=(lt8eSN6#R3Zn@K3nFAs1=&P<`^r)35H|X%7*|Z^6B#4}pXQk`j<39HeoQlXoV5cv-tG3h8 z1^R;17IumVgRY!>gBO+fZqK*|1lA!6F}&#_`6aa-uPY)M|8}^hk<(E3%psY&pK|yI z0?5#k1th*99=Gv{`qLwozVA7r4A{~>W)L`Q(j7rSu+B%-IyQYhI3DP}2K10w1m z@G5ztZM;Bu?yzX2Sa{EJhee=RFi)RImsD>vxtalZSRmO#Ci6Ltv!)10s0*+pkS&;} zPd@wW5AMHv_p8tDzVqPGSI-{Yefa3{#YyUaGnohT^&!bL39C;s7ob~M%z1I)6X`~} z1vj99Zed|`3-tq9g&uxWsg5`7^EY4*hw!}%oksEnU7^w8hCsfsF!BXt5_3RZUYxiYI#65D*CZmd96IPPb6A@o8;NCpf5^{&avKmO6%^kfDBy1qPwP9VOuZ zk)R(6dR}*t$b#iujdH(S7{-cK{PtG6w#>v;8oT*qKb(C}oS$={{F-`Y{BkaoIq!rS1=VZ($W#}9xt-xb zdjtNVXJ6)VAoEA+53Z}XnweD7m40b-#=2XNW-wJm2+E-)QgjL^B~@_kj|#1d($O4 z#Zg!PHcH-=$_ciSALb%&*X-MeD(pwQe9gfPTCN|Ik zn^^>Wg{$jMx|tktcdCez@#`{|SOyZZe9wfhg>86SOmG+;A}NU+r1j2Cxvvtdko zj)(+l{LCT}s;O^`0Ez{Qw1rWmy(~SS8AMXyAa7J7aoS_BDG;YEj5w`^QS@ z6$R~I7Pe`JMU5a)TcEnA@{z@nPoz@|j`{o{H)1w4%7IXAVT5YJQMSyuXp8sI1p>#K zo>66r@gmezJUrQ^D(S5DD5w!J(2#@QJR@Tm%#TmK> zE$yd6)IDf{_;)$Bo&j~&4>HtU<_oZaYzx(2ob_#qo>BWAG%4;+js`&ZS)A{CQQq$H%V& zEFC-XA`ww0nTOji5<7B1_9-*c>It!fAx%mrLVLm0GSKQRjaF~K;!E1#eCCZI6?c)q zFna|b2oINm;%;dacQp)yygLt7-Z_0)Q1ZBh%hs0YTuRLp?*>vt$1izwV;YF)mPSM; z8dJ-WC>#8kV(p<$;{*taD>VZGw;i5*Rk4}4NOCJFj%>Wk(YaizkDT3O^MMbq~2Tv>~-(wI2 zS5PBm(q)=WceRgtO*9lFb(b1-3Ptk6C2C_&{^qwa6Yza`ewNSoO^N~VgGK?j9sF3U z<#=8RKg{n4&I4wr_SC~yc#8I2$2{m-$X2Jq9PuDWG_cp#eTwXLr~}RVR>tK(2W4D* zxVxr)I8tBXMC;ei4<$#GdL()ye8Q*o8{ZnKFSRl{P;c{f2Qhu5K0e}ctH^^f zXRykRLxsqs?J!i(_k^ya*=VSbm{ctqfPktsd!OH zU6@gt`I_@3+$6W$N*uBxWpI4{>&AE$=-|xbWn*jsaD~JP3kkKtML>b8h66%sh2&wu zPQp*84KG$8BwT?!Z2yhKJRPCnc}0@ww5LXJ<^yJ>>#OafYaZr>g!2FjFRQeTK!%s7 zDm|Wu^+H0ULNg`;8J=4*Y<{O^S_DGEMIR3@p)`G2B{S1T?-}um?ApN>5^BYqeV%Qy z>7LHR#%PYrxQ4>}mBV5^Ufk;NUteTFx=Vk}c^>;WnK<{}1 zEU5Zn-G0h5X5WL3MbZ8514UmjoEHgXkjPNV+5MZO(|VJ%doz8n0$B=fn^|ZnGi6Jw zOd^NU%Udosdg_}Ph>QlFju*^U7K#tJ3I8k2m*3>_Wvnq1$p$~~z6-sqZjCqW??NVC z(T`(#Ubye#(H~TnK;;%`Y81kEUD+ecqf3$cc zTBJs%>BmMgNqRCtiD%opg!=3zTn-ZG2rwlj5NtEZJ$y%+CtD6efYj=(Th9H|_YK){ z5FsN$(C#G5K_Z6Y{R(~AXf+J=D}-=4h$xZ7v#pl{rWkNV>gV`yNGCI2GR?zRfqu?B zeIj09>D3&Ng44jQNDUkx3V{@G6=>kh)2pBmSHcH!FQYs}={>ZXTcKh%as_lC18Lg$ z>>@%$%8h8CXxLph#?m@RG4#7Pj1T=C|X1xd#?}Pyct*PMlB24^9{Q32SqpFY=xF+5Jo0->D zaN9Q}ejqlnrhOCZfswX?)r*k#=%Tu!ePrKbuHhk6YRurP=yO zGNcz_!umczW0jjHp2KV3l=#DV9{4=5r7zRDGHZyX?VD28_OP1N-A%)+>;#vI$~UtF zBk8YCz}*V+z;;m);I~6R)Bt7m z*z&%0AeS>wuVTfBDwS<;0>Yqnd#>pIwud2Wg!!kl&`0jB(G6vdO7~Q zgm4++rj%Y{xeIzv6PE$QtQmoI@S2gT)OD+3APDgw(;VO}lQ-HYqi!&`6| zouO&LVV{R3gy!kI{(He>+lvTcVJHu!0uiB+3LA=#M1=Yig?+E`bKLT+Kt#B<-w=_A za8+RzK6z1)VSeR~sl0FWmoHEcCE zdKi#2QMyQk04g}`Qonb|B_UC&8ZBPpMIw?Mib9xKrOJq>D_JCR+bcD~>|TVik-bkQ zG7q<3B$)EUHL2O-!zJijw+=LW=IOk$#lBxHS)#_w`OHHlq`Zq6O1%Ja9VqYCMtN6L zmAWDoHU9NMfM_ZfO&-_D%+}CvZq4mC=TH*T(#3Azqa-(`ftGG!3<|ZK||AJ>tHciw#?tX=4NKJy~yBrDQh(w5MZXZS5i*V{Y*RiT7{Z9k@pj-4wYb5>e2#@nV`S*3AqJ0;b z$BMr8xz!Oj!!yV*4|*?7D+@-wr@QNe0d?qEpB9`@OLBa8Lm;e`i+1WiM(WoJlG#U3 zFW-H7d2;Xk;gds262Isqws+KP<5xHm+x%N$_(1D8bP@C~wltoj4sb&+AGGV7+6al0O>e>qaGi9OI~ zadKZyTF#4j#YRl7t(()^Npbwu$D`ss)%kU`59s7>cgfgsTep-QFl4(+5WBt(4G3OA zwAj+>0OTERyGt2=)!SWQ>S4CKgqcIPyTl{=VI$w}QV%Z*7~6(ldD0Ep?m~v#_RafW zhmO;VdVHP~Ua34}yNjr@c-`H}x4T4~UDf zeD3`FBP#&h1Oh2oeCFv@1ezn|fE>fMxicLE zJciyto4ZZbea+4xNRk(iGoMBKaSk+j8>7h+k8>YLm*+3;ji~9GfJ#>liSr_otHoZ7 z5Oh`?3cV3Gwo6Rnn%-BZwOjg)2jgmjlK(aUxvS_Y60hkjq4CnS7|LBscuv( z55mS(uULK!q}mkA!z?N%^(qiHZX#R_74s|fiJ%;q+UULDZ$q{d;rB-TT`O$dKsyol z^hrr(mCDy@*Hb5Lv=VD1ZPa~Q?YoqH?S{ok8#ng*B$hTJo?mvf{wUT#ef@REKz*q< zz>2ES{yq}E&3qrdS^Y)_gr6fBro?`4I!^kM+K$&kPMW6C1_e9RKl2?G#OEG|`%wds zmIwG8-!Dd0d~&@O8t)q}h>dsbf0_0+V>VHN(9(}Pr8)RpH9?IOZE?jeFc3<`Bv5w- z>l}K4Fgv5ZsHh)|RMq-{ekFjekz$Dd4k-fL(i#1fR((YRwBSKcffZj=$jznx_wqa`2uV}T|Df!Y7Q7WmY53zJGMUfj)N63d$ zI*IDC;6O)co?a})%&mRBf?FzBoGfxf>H#D8l%)sImvIy50nO8^pcJ>Cnv=c-HIo}% z+SWN0rENSnAbnp9QTQmzjcTCp+ZcV{%d#_TpF@%6U}2<|KM4>NS2O776Ip=Hr%%wE zaTC;^Z%qBUr%%<G5UxmA{BPN@s z2l)DAhVpf2Grtfv@^?=pSqNu}hO|6YC!M7<3M^flYwi2u*|YPDH1uk9)5O8Tc6h;=w6a^-uuCGrU;(6Ydf85#w!s}3UrZk{uk6@D%c^Zf;ozRvY)K6ZAHVpa4MZ&F@k54I%wTj$mVm(&atp}EeGrkl zj(P^O1zkrsHs+cb&tSMO=RzbXPTkHeq1$pV0r3oG8!#B%7L8grgXNK`zn@ZFmhk`%|X7;QC`jeo(Vt)Ejbjkn^oRgkWj6Q$=dz94tN9t})p zO&6`UJul{@E#ycCw);}u0k*q&Gb0DQt?i;S-9L@g+qDt>&e_GqFhKKV7wK=R*T=U- zk&fS@g3#!;(9F*tehtAt6(E}CcJ!f&uBfzk6@2p6C&t&GfBy#GR1ojjyTyJW{U@8tz_i=s~WWG>*J@t|8{qe5jJVlin1+aine!@_tY5GRXUWBX8~+ za4PS1ZLoh6VjR>+ob%QnEl7iolst`hORgfuNe~ayFn{XBJ$dM zw0>2gRy{Rv#fg~BAnzY+#i8+yPfn3uCYVQj+_9%Nu6Ox15X@|iU}pIG?!81!Z*LS* z%?OP$498=eKsB?iRL!vSI+v3NRa{$8&1`E`GuuY089%mxY{oo((AY}3jm~=o+=e|2 zpq<$^*3N7L?TmT4=Qiw@b1@PWZEZn2v*q$1$8EGTpZn#nKfVN*!9@X1JJYCC%V}rA zQmyGJaGRi1tFR0ps4|VX4OD5doAX(2vpP*{XQor;@&v8b^$nwYZd2;S7k!ZJaxM>U z^c#_fYxiD3C#o606IJ>RfZWg!L(o<#{c{p@qtBNP-OPQp)sU^)u^amC)5hmX5xfbk z_r+I3tatP#SAs;%$oM#h)K-lhwo^YGsjnn6%KHy5&mUi$-@AWyd48}CWG`I1#_twg zdLs-r=yojM%4v#9KM|{mSQnzQXLsrMv5EOMEFT}KFEkoc+k)j#A(Hw2U2;Qi7>cER zH{Bwi=rq=Bl+;Hfm6m#z3=7y`a2PL@R^Cjd>&+};b{D&;K0M06(URqT@s`YgnJ)ju zACJ5Ir_*TGDD>_r^&!`6ILxY@J`WWw?EyfUn-5@8dFqO{sswiFw}AoxNOTWtxm5-5 z?=V|c%4km4uvRe|nAm=+N|-lft4chsA0o1?D)r#5OIUd+0d2$qx2m0Zs|ulV{Gq7X zu`U+yrJIG_49`YI_lmcw2q=rz(A-@Ug#|njPjSA0K3cV#&cY2*yj4ZySmx&TTUELH z*G*unt5YQNi)*F$1(xqynV){3y?@Whnf8y*Tv1SnquL8ikx5A97n>BR`Lacp_CnLu zDOUOEf6}0@Lgj~XQhCZl6Q{N9wzIFL(_ZX!T>8g-w3Hua$>-{{k>1abG1mL-A1@o@ zJ~1wvgt~rd>?Xm+<=$yN1@p56Y5szwhyPi{2m_Vv`n5X6nm_$d0XwdGLl-c(G*=G^ z1WftHMr5w*`lWpvP~>uNBJ>o8lF@ah@jT?0VM5n0a|wizzuNmLrc*%4Ro9u`47z?z zXAw_^>rC}U55;(Nv(ZG##eTsfZ}H)G{d(4Q{aT$`aiM>D=JJaCa%vJP_o8K++-0Vw zkE8K7qVKwyoVBv%f%Y=f)v5iKh&q04%9oj<$oef%O6nCke~kz239z7|r<;Zbm_Pdt zQ0coovcC_YomR}T@AAn0PKK}-quHE%_-Kgqd3%8!tkmeVMMv zXouD53rt(Rd`=Eo9%+3y@2_wHEN=T0VW@r*3Y=o1K6SVI{&QCg^xi#xGJam&xb@ul z==ss{PpePtuR|Gs>dp(#Umx%P$6H75-#q@j(@egsZjCp2&153<(Ocp?>$>{$QRapV z`XM=j3kJqJ+BN-uhcF?}CVp_F-fS|8mP{NbBr=_)W$yR;&i_#A{=<9c@7=w8_O4cb zo?dU}d*a*IP4&HzO5Ywu+Y@l}R&id`ruC2mWmtmFs}GL&5EsqAJ}H;@YtkmLeYmvp zd73hmJliAg>boPgk<8p49G7^uy-O(L%`NbPWhFt8ay!j|cas1UQzU#O+-Xv#ayWRK zfDX)Sz0)L2lkYT%hxJ23veTp<+wt#1Z7`&i6Jz#(4I3Cs^xo~_+=Ob;6U^<44 z<-)5|BNrC$-QFeCWb~$xl3zZ`wgu~}Q>^&a|5S;x$s%C1a)GMV+?inr68I7R3)%J# zJ;woTQo748Sa&0erMs(BBi$8@sHNA`K)T_^mnduIC;``8=Y!Lx=w6pPOpWtrojluV zV9JyAe3UsohWH5EpH{i{p)2jTiYi&c6#@{i%HSNd3tuqvFj!Tu4usYw^f9@E+a7!6Ot+juAq*j&I_91Gmw=$)CdH2RsGhSd( zO4}k#(wkLrG}YgXRJunEVO@`$*1EGvxD+k(hqn4nB*7&5>j!+h%7}>DZ6X0xY`ux3 z4CMgR;Sn2{)p8R_m?Yao5)bMJL8byBJ9Scz%mLCCF0Pm)^_xgYM_arua|^(yD)OS> zZpTP$SgL*#iTG&cs<*4%L=v%b;7ufVLN7*a%wrz1pKHMlNPa9@bkRQEctkFJRnFYBF$uyjeOLG4Ba5ox|joZY*3 zcdjo@YC=pw*1PF-h?oO0H60>~)4=&5l=UtmYzl26HmjP4YNI=Liu&rrlZ-FxeF7{N z+UmHcPhMmvh-^Wxon3H+uemMCEISAWhaE>gy2d?m4P3@O4Ea*jDZHj}Zz79xP zuQ>)O*ZBb6z3L~m9MDr&-wE^6T<-%q4ciKTauKo@d; zs&n9LE`V2RF`h9482xOR-4JyAhWhJ~V$5>%>;U`ql_SG?g&q$#yr2H5KRV*DDP&(C z^woK1GBiHog#Y@J~Bf7Y^2iHs0WAwd$7m$>O4q&i@f1?#lfzqp3I44`YI-K z4|0)1RXdqO{1T+1)bAu{$l$5Vkfq(Ip3D(&sS3qi-DFP0M+2|aPgrUNQMHpfVk~tY zY&Ty?vDpo$(3&EJRuIpXXhkOmU1qllR5nJTC0Go-8>ngJjUe?^e5L|2vuU8Oni_po z4O2l!Yx4D-m`M8!;S}3EaW&;OPXx3yLyD<5$>Wn07|U6pm@-eV8_Tl5%?- zF4>~XX!&bGR2{9Lp{BX<)dV|KK`$&WF~7q*%7g;5WkgEgGFtX^t{^a*SuETam~nfM z{!gF*8hRNT@EK<6gQR6MQYDF|%QsQ3yzo60e}?yy00EF?TbDzfs29hdsmP%+2!LAp z;w|@@NvvGZ9!J=l}JdnzV?Ucu7GVpox0~>zw-A5iOibMrQxA*Lsh(i3fTKa35AA6F-srao9kM zajfA3l8GNOg3@-;SM{~AuyVN(P5g)zREO!Ve&Q$M#_h08g1*geI1)|#h#k~>w6!y+ z@OA0;jFdF+3)XadoIOLB1(F8y^a-gx%`c+N70hLXNQ%JEbt1k17j>*=ffQk8qzE+} zVaa@b;0Wf1TF((?&_iohofwm{0zOA@;~B^bW=2*}J)V6{1{s*YseIzC%WecLf$Mq~ zIpp_s0ZV}I>)nJf*VZ7DYU3g(R+}w>C9Igw@ho8(u!Ne6d(VvYLf@?0M0(*^La$hD z2`qtItY$bUQWLEDEWzASY~S3ktYqjpmVl)N_$Qb?e+Vo=2ZCV<x< zf4hP-=)OAS=Mq2V@DBv2p|QlEkqi16iTcwcmA)uCVhz~RK5P&;Ytkfo;<q88_O8<&3dTg$*fk!fI1A5ukY|fP!%!*8N&>Ht0JERZkRWcF~ET83-^#4 z%+mz}D%LXSlemCCaDDH5)>z9h53~&PP|LvO1}-pyB9J*~8Rldg74)s!^M_T)AJFIP zQe<^mJSzPVik*zG!D|_Mr5`Kc55mhq*Aah!s*i1j&-Vop^Nf~(%L5((f6yGm@Q2fy zQzZfrb6!KE8|p7`n(jNSkQ$0!HUtD>9UsmSfq2E`bPF{bpVsDd5(J@7bRI$YnvMD` z)-i|z(v8j@$o?h=!3?1=-TJ(xzJ#ORQb!?{$4y>1b6Bc8!g)8Ue;uh`P0We)$=T)Q z`NPZmj~<>pIe+jQCtrVh{`lF+*Ule5Jb!R$s0z9p z`}Q~fU>i4?Z8%y=-yEDedvgLhU}C{$Pt<2w>k{d6L#voqecn*tKT@wq!|?W#C->iZ zIKI*S|Kx#}l@&h`x#3i5YW*gbfq}x>n zY-Sdu{6S&d8enqO^`JMxK0=#_D8u!j+w%3Gq0t>yChw-;_YwLX+ec__ zWpZ7YqPe-7B6V>gorCSyQg_TN$ARr9v^ce;h)ABNKU!HkYZkUIfHn5hq3 z`htfiPrrJ1I@J#{)LrHau%Rlf^*vXP+%A1jniTgZM*|={eS_rdj$Cfldb|{o%M6Jx zkH1E@-c(;ZDLm&mCckcY&dKm~Ml2EN-=z88a9?66G~6NjjoZK)dO&>ijgRKnOs~vT z>+wt=>ze*36~)**ZQ5<;2R zl3j?)0~&w&Id12pH29>xjK_1uGg%W+xA^G|_Ha*5DD$viRN8rn@`ZSjNZI=^_y!Cs7FRWQ#;1o~U`+YyrvbxYP)X z7m0{6$voVCkzlG0=cHCIQWs`9j8261C0_GDt7o1*8L;?zw~&Je3}t~7cl=Z~f=#aO zQTMb&iaYc4DyT$WuZNOSt1nxe26c~fxQuOaTI&WD2W;+X2bEZ7Z3L?;?k%aL4$ zOOEMpB7~%+MgoMyb-#-q)SW#>lJ&KZy7>!so9WK(sTj&NNH@6H2x=b}x}I-8-4#sO zwu2vQwH(hY;fFbBXV3>ZBlOwNC;z^}Q?&0|=AmPRG_=)OFGoZ8_WHU{k-Zjm1hc-C zagt+K!t3gXBlQ(dw0`aUP;x}6N1`{v*n`$@d~2k>)XHd%dgCA_=CvCiAMvrgaol!AS&rrMX=qnWeovu ze?{R-3#D6?jPgeBDchnEGm5(=oL` zIA@+-MSW43c-aOkC0v{aHRlVsUv7aagUb(dAd{5A@%gVG<3Ia4;0S~THC3+#@HiW{pVb_si zRF%$OO|nA5#c8i9eG6o`a3Nu@v@Z}6>Z>3=`Ag?#58k-%P?#-E*FVQlCAp)&j+Be1 z5ALoOsB1_3(SR=`)QUO#Jp0kWJ)MV*(Q;}Ms+x7w$q22S*~5eWboOQml__y@v#Fle8$2AGR_i0-3aS zQGIa4CvC`KeVi%r*Q8Ah#mkz{&W>BTHO)4uK{ic4Hj+uwBbXA;`t7xq{iLO`^?+^0Edh%TvK&OnND#C;$#Rg0p(|Jpav+HsWZ`lU zQ6hu-<($~=7{USPe|%{8R&&k`NYlp8$skvUEmmc5iuMi)yqB-0=itq94{@!SaPr5 z{41Q<`~G#GtvTKTzL*}r>T zmwC21+~t@B z`IGVU>c*|-#z)T|9T(IV(mp$h_r?6;x+&RWHNQ(dC{_xca8wlsG9|C8;PxV9nOM`l ziS@upThV$r+<7LUooG20W&X+SMVRngz{}ElI~>;|#s`%|Cf(8qW?m!v<2pc1zWc-HR|`T(Ft-_le{G z2D{sCz-AWlU*YP?>_wPxe_}Ihx9=Qc&EE~!%pxc(b@xFGD<9K5Ax2ue82a!DvFlQ| z48(Rzc9*(1qO5?hxG_5ydZUnDPN?=WKv{u+a2d$umPRgDab2#aN{$s5jZvTnn+CPx zOK2>&tZFRh04S2b#iJb2;Vjzoqa5hpmPY@^>O~ljvPH)g9zwkcH4~KZ_R%#Da}q-BTa1&ZsII&m#f~YX z;&T4yiikDqncaOxnTEyXvneoTbFpsmwYb3)O>gU4H1b5 zr-*_VU(tNJ>O_P(+zotZ3B4j;SQ}R8Ltgo(aWBG^?^DO>MTp-3Nc`+r+HVj)Ekomr z=EV6uLh`sJ-^%wL5T0`Y%TLkL+b(U$ad?Rbf?k9x-;O|NI52v>e73VU$KoaSx+%0b2ayQ9Q*f6%rV>^L;vBEkzWoawayiM?+;EcC%M3SB_qcjwo)!(*9!;R11)2 zkfBk|CS}O2qF~2vSQ?oaEE0UH4!E3f)Q93lB0^505N7rwjGUdoO_9Rj0E!sqB z-HR|bvc)2?YqmDfEePFO3mOd;TEqMd*JggGERSaWBG^uM-uWY#j4I?rn8S%@GgExP-gS$;68h zD+@-wr@MkthdNe?Uve8#!^4OmUHgxb`n7^&wv!T`+&h2xL@Y@x$PT~g+0tlnrYP>hYXuZ`(a6_65G1G z%o^d7p8Btm+KLAxpH!FMAdy+ObwUK|^r2<1K_0(wOYJYD-v5XCmm~F>*aLkQC->!~ z<-CX&d6aIqb#r>5FOI+ZcvQTnI=`;=uTGY`-S2&E-BNbIknJu(?D{%1II3Uab{FXR z4!7N?ukJFiL$aWW&tTVS4d;%0o z^jj`WyFuGsM3u$s?oPhlCE~2CaJaG!F~gt-ZFiBNQSSD`Y7JEph4ES3PgM>BjT$e&~wI=fMwZh)Pg$r)yitrRtTS! zk2TIAq$H0oYxLtBX!2G@lP4bMK9DY(g@uh;aZ}$LaaJo=>rgV1s|ksI&>L|Day9pK zgr=+B&}m_V68V)b`?im+i9j!GoQIIQwxQ&n_s&qUywGU|XFFk|>4bWJ= z=6C#pT8^PW*tm#rHB`*wE(PU>>iKqI+JCIcb|QRXdIA1=2>XaNYYHNgH+VNc%43=eS{U(#Ey@K8d7_b>eURp|gYf`sQ;H##8v==;%|4tc($w&S%Bd1jD)F@ocx*)^qeV~@UlM+Nb@$Kn1>!?$!X z))trs#_L%KUvk98JN8evB50asNZqV72Y;(3sF9*AR-FI?2{r-pGYln)6lg?)6%oCv z`oTz5tsm%b1LzuMp#v0&W+L_*t+llIyGM1z582$utWw_DS9mqVO_qJqG}A?*_RX2k17YOavi80(lI!zJU?@5e3_~et_u|D6Nh8;& z+@AEt-yxcyF%TrwrOMFg#{r!zo1E8?o>EX~8oiuJDIGJ>`wmzs+=YRd%nEwtC=Eeo zwL8jWB$c7f{6g5sd7(&BuK-tLmgt10QDEuXTx;JK&z_w>e*4kmi<7UOU!K{UL*h&p zQXPImKqWCnR#22RH698hi)4xdlTp{FC__B&x9G-3lIJ8fY?(=9y|Bz&-QrP0Rfa7y zAz4ZAlD0#~VJ4yQh&&$+TV^7{G9cJ@LzbCD%*Ffo!PX4XUOsG@iHOS(&mU@;q3cpz z>EfEy&+y61?}0369SCU5<0pexbnhc_KDqRnl)wn^OQ3*R8wE^_CJ2=}k+1I(?4T-j z4K1tIn_4#@4NFX+`1r*SZ6IP<>&FinS}}vsMRo!P!yN?(q7Bk@1Sj-5AmbR!26P>J zB4eJ87z{U$cOeoKr>?h_&}~rTCNk7KgIR_QrnWeB*0?xz)+-If1I4K` zq^=`<1q^0`g)9j3aB0YUz+m*)83x1D74*)zh{J5SH*@?0CKSGV(?XKs0E?mP8Nc9? z(a))u##?bRDo9sM;_7*fFUTFRM>{67jL$U)yR)F2wzNy%}SE{wL+{XM4Z?V*C8-gn9h}%Gw z*1Yy!(@jP@vu3`RYpt%)l;Jj|PJEf13xgXjWy77QG*<^{2~)k(J1hgaS(6Y$(2Of> zums)c^QA{OzS?TYR_)l$h8}K8c+&W2lp#ANpZC`L;;SLnJ9@KieijjoFBN#|ha>fs zWJY=a;pO?`i}QQ;&o0jo)`9E=-F}N5`lUC*IR)L0kY%KRN{G4eKb<(&7x#j zzy{kzO_Y!8vq#(;>?WQ-<>oCn&AQSnpi>_nW#DMZ^1gUW=D$pr|Kg9wUH)ASce_zY zXXeFdikI(^S+&#Wp`xWd04Q^_pJ(m7Q}ly)t4d&(ej6wNfJFDOmRnT-{|>WNrHl@` zhQ;Z}z8Pk#0S!!Szf~p78?sd;9@h^M*;bW$aMvZQ81H_P9d0w*iMOgERBjhYhuOWv zTU7*1#Y_Pk;qgA+4JrIRdNoy1?7h zcX?!gA3*yiJ`f?b;BBW3TdU=Gp4WzDeJ$4bPsX3R z^TPAj$Ip$Ao*&&hdjIC}=UsTetZt1rRpE`^1d~9DPmI)`k1{u0B<2tLff7jPtNr)X z;s9ok)DMo-n@vX1l8M8FM5eQ}%>9X$<*i&xd>= z=A?xQR{p^eAL1fg`=ngruSuK0_Q#Y~K2KAIl4tue+3$|jMly4Ia9rZqW-BmhGn~hT zJ57Qj{fKs&023c#r%9R0;oxloIxsI!!DTy5!Zi6#lXzG^BqTdc>aoRJGFthbnnFsp z(}dXjz&k{vwPdGcJ52<-1?%BXveP7jru9yfJNHu|6Ce3UO19HPY+K^l_Aa4}cl>ut z%7yur>l(p)$ux4`26AEZ^oe-yrnF}0km?PRC%Q>tu66Of0=X~eA|t3%-|E_H`&8xx@p|V?NUC|l>wRc- zkQ&vg^Fu3w*9tCCp7-k1Hvp*%b?P_W)Y4Ipi0d}u(@>{A=hdm-G~GKpXluXB^?LV=Pmcm{|ciy-%akbH9ri{mF~6ws-sd%H6Kr_sZ<8JlB=XX z-+A!ptM3Intz)fp9eSHnYmT(?ZS}SH0$UworR&by*eW@Msqf@&4z_yuL6;>@)Ti!t z&$+wTM?M)puWsBjc=hNwqYcx#uxr<~jQNg>@r_!I7x@^^zh^*Q)t`(qD&G79N>nx8 zsP^=<@vGUG>OfZrj4w#K(!7U6sCWJ))anKE6CT zfB5T&~`@mrbnSL=sG*zka~Sxr1*M z0ac`+>5*U)Ng2ukro)$vz^s;=NWvu9CX#qiKL|1v2-&HVdSthW6rT6a=v z6nl7p?`AC`%kwi*5z8l{zYRomV?}8lv@wcU(SGdLhw^YatVN1h`8_;A-MbC6tXrdH ztwEyOiZONf`p&igZtMF?E%bsbTDMgXvkqCGB5i7%K1C2T+tN1i43~`kw+sS~a*%){FZ>=)vehMnAe{ z>Udf2HiV^BucV!}C#DP3AtKm7>+q>woVq(t7QGG;TT4vM@6=d{WxcxV#3z60{OrLS z_Z=>RHW7={ni;E&HtckLiu&sC?TqBU>Pg39_4AOdcmMRsi|ho^q%t+00YUG=78Q8B zJl6qb?W^&M74)u8?Uykm=v`Dr=GQG!HeEstQm*p>ygN^*-EFW&4(O???}YhjuJstN0><-o+!yGWo5IX{8tZ;-u8Uvr_nQj7778OrFL0)D3cdZZY$ z9Q}EK{SqnI@b`;2U|&XgBQSybTy{hK(GiDDA^ZBEug*J@q45zfW9UYP;l52HJ}Ma1 zpN&-d8ub8CU=Q}#UY!U1PCDFL(yrRcoJgi&G6&}_>PHVcvv<{#nZPfk6$hC13GNb2 z`|8OY0hg*!+|^CyM0^}{GUv82YUHXXbHrHcJlJkB^R7u*pCW};oPG1ricSm+OwrE^ zvTvi%5-f(~7Ug^dd}w`&^i{DDe5PV$X8L)75k_BC!Bq6o+EFkvwm?j@z_qX1gODV1 znWa|JvJ+Tcn z*QdC1fc_^;PimSwc9g;v3&Ti2Oz@;fHVvS}fcXnCS{o-hBR_Km#+)FYR>wD7VBS)j@v0b@$)B#siT{42(qd7{v>@+~j^_MdHzVC^0(jiU0exZ9!J=l}JdnzV=;a?6Nk3xWh~oS9rC)ZdP1;Z!m*`}OZHyxqZpp+? zg4n6UEA{f#%$;1yC!H?m@kI8{(;zz8YI!t%<6F(6*2cO%b-B2>|BX&^l(Iz_s zjSk$};?r+Z(!gg3x;;*pcVbDy`qW4oNcCxc5q0lM?hek7DUu@a6X=L9K)Kp_8b}e$ z)2lcF_ZmdU5pcPh_M6Qar(7NZiSZQ43ixh_IRcF5G>{dTr`L~XUz0&D+rOz?WtVP5 zCt(R(*E^S?@9P2F=KIe82e?#qJ%#V<`w1~M*VZ7Dm|^YW0vD^Tr@BFyeJb-go+V5} zmcYI4!LbBXtQK>+s3JS`3siX42CU^`P;&n*221MMUK3aEzH6Rf^f(d%~Wn-sA zYe2*v1YT`C&1nqbllDdmhW9Lem<9@lsZlUUmsH-O=q+h1W0(dqhH0&gVa9xYmuBLs zko6Q~4AWW}!?cl%!3}62V=zw_45(PkpmVAEauD3GoLatp=pZdau=q>)L!f1thFS*n zag}^I7Z^bi$aDq#ftyo%{xA#q!`81jH!tWMlWx!(??HLHL@D`YqNmh)v^HY;`aXEcGQE^_DscK^HeP$tOe~c>Z;y zel;;CHYaD7m*)>J?>~BY^5p!%Z=8Jn>G|VlCto{%{P6t2$=SWLi@O&wi|E(5Ltc_; zbfCN8>^tK8oD1{U)GOncb79VTC(H!sh6lRt_{;4K4;mi`cdmVlorb=R4Lt!eft}~@5YT0S`LK1IXHFp<^*)W#1a^*;cFg8$W0$Ef<`>nYe9Sy zWn$gGuD*YyUXh03?I%y}zw^+k=%ikpw3`QH>f3TO(!9Cs?tdF4Z&u|5TVJswb+?~a zEge!eUA~zmu*i>mGYeqbp*FLW;lXAWCm4f!GuDoX|`DRAI-QwI%5#Dx-O9Jzz_r3v}SpnI&TDiZ-)cW>Oq= zCEv^<5-fH1AvQD4Jad|yMzXYTjDQRWy0od$rM)aYpBY3_UNf8v9q5}OiCVnFL{hY@ zgvPcHTF(NBnt6JaL~UMU12lzM+CCmuh6Fmb8Pchh0~|(qhQw;ILn2XH)*)_`XMtMH zJYG7=mKhgq@uQ6T2 zVVfZoU39a*9#s1l*TgNWp5kC~Hw8aZ;yyyZWBUlrtW3^7eUiGkkj}yOYpFZt6vV1}44L^3zfV%4k8R{0O1$M zuRA)qRqOFmL?<(tyJ${)K0`QOXS?21Uppy0=KzymP7m?2$L-GQ7YZ4^(rxmf#8PND zAbP!g4jPY!CT7mdrdMXF^>`+bach=rKO(A3iFtQw`R9Y}mHeWp{&A!#Rtsce07?Tw zNqtjIrIqc>j(}=ZMy0-eR72~Ky?qIx%xlRm#N`2vKmBADjgs%f&TUA0*lZ9?MBU=2 zH`v2HIibwM+#86?PA4ah-+ZNuL;?@M_q(fDBzhQ-3LNnw5dx@IA(ntg(LCp`fQc1J zyhudINfg36*&>mMr}#K?fEl0JJjvN3UL+#QB=fMFz0Fn&rs{A;YW4Uvk2(?B7h}x= zt)6-MWWeI<-9j!0F8@c0yO^Qyp%U~@ngxoxnNi%;^iNulA?Iu}eOvY|YL-FW`qjHrbN!~FAgNo{s8b)3AI`WE(QkVu zu+nYNAvONlx&tu454uF~wgV$;wH(hY!H79%3z7!2Qv>x_>Gs`Vp=jTA%ws8dllU;> z*-W!-{35j1*L{lYb*O{Bw6ER4IIUqJm*XFf)K@ss`n7Yh98qw2A4nCYu8H0jCj{5k zw?^tqt&HZVHx6P(Uc2$}5szC%UWz$`Rc;(AL?&&AK^Kdu`w#D(zjycY*}LcCqtol+ zl_ByCgI2;(OMS$oYSE7bR4pMtY!^roE7M2U2I9jHnS3p>T*TL+r|?qe=+EM~XNrjG>^z~z&yj9;b+j^@Cz?^)m zN*R5XTUB62`>iTr-jJ;-@wk47$hNA~gNp*gR;xTm_9We^LbiLrttx?gf3#bbjPmx^ zP$uP#{w%RoUByO}qrAmWxZ(kRm0&z3dzG_5Asnk z2nn@N4~LX~Izqu!Z|DLhs72Qm+csdT)Nt-XLQNORNs&u%JaMi0y zUs^J3e#ZrUd48xsNVtmdFqEdRYGu27Mm#6Gc0_`~tZY{+=Irxqq`0T^urXRrr82^m zl`|t5p>AW*DE3)b~b; z^>MBDkN|EDk_@)8DuSSqkVt)S#3yaYVSSt_@z*InVceWh?4i+cEuG*7l1 zgaD~L6N4p)mKG#vR3onlmxBly34(SfSq>60bOp;nZsR1QNVXhAlt|*)*2@7?3^+&n zxnMiwmq;0cq;xW)&;57T^FTjmo<0#Tu-+x)pt%elY2f&8AHgDb1;RYgz|D;Yt_BKm zC43;1Yy+P;Wgq_12zharJjDGHiLJ4G2vyy2#)P#Jmu26)>Mw?S+b~#uqBCY!xHa zjjjOq=-**ae#9=;ebTdn3l&!t;>!!Ojx8+X?h66Ag2H!!W9UW1aEw$h`$!1raup;4 zTyWbb{uNGgegC@8)*NpEU##N|8VTKAb@|*<&BxQ~d@hY9$m5hr_k0TPYdi%zEeHTt z^lcAq4r@7#K`wuRt-i=pU@O21Jk#44COKXug38o&vf9h-Q;i?9UWAEv!U;l~_DzW& zi9f%da8wo2a@@olU^DZ&3U2$R#1F(K*0gV8JuuQ%uzC^l9$i#dv>c1_CpB*Sri5_; zFH7XKEe(nC?=~6>ZZE=w^?icIDmPC&hu6L-@rUs|h!;^r(prv2nx>V5joZE{Wo-|u zN!{(wjx8NhHeJ4%B^XJ6eFE-=(i3urIX+Ma-gw-0rb&Tt^-LLlwVPRwGG@pDyRmM< zJ@F(Swwa}TwwF!cJ$rIjcW{34=rOMsA?sbcvxT2hvAn(AB;$w$<(pYZdt1KsVRkOF z)c3BiV)r6U4XvcHZsIx8%`6eNSG1Ysl9lp`F5k=|C@kE}c7EoyjM%d2fW{U`YnNFZ zk7L19Br8;|OWh(6+nLAny41y?Xa$5tTbv7>UT#6^YW#&V0=mT-HRxFF?kd}{1RInckE$4f`KkB-Y8!V*GF!UX)dYCzmf zA%$V}62gV;$=2y?7v?%A50R_h(1lD;Qofj5lDw$u^yVdm>kub%L{Fe|_7Xx=T#k;m zS*nm&Ys~Gt`gwL%^oq;Zz?6gH@^TCX62eUcmZ8FYRW%MeQaZX!w&37P2%#6@!piRa z(|P^(f+bZfBD6wg1mg843TvlZddq@11tP+Q{f3A|gh*mm9y|T%suL0FH#6{^CG?7X z0XCq|HRYd1y$DzO4s|G1XZ0e)ZvZ5I6CaTfiI?^%#J`lG@nutSOFn&1I+!Xv=Qu3C z6gSnwR@wV%q{dNR4|);m+s+3X?if8?px^TFCq+BlHe=mDU?o~sRCtIsW6{rtxSIU- z!tF&E{YD9tN(G_O55-ruM1sKzd%3*`qu1!8(TX(|1qQEHI{xt}p5h6MuxLBqr)N$n zqF9wPv5`F*D)X@4fZKV2v{$-FB=7(r&#PD@dKi#2QMyQk04g}`QomyIyN2OJ1~Sh6I_!Wndklz_ zQ3F0PTMr-Z^ba3sV=X!eb z47?Nvmmxvn;Wy(v^_Im&Yg+*c62y(gMdo)%lbqq2JR$P2jYyI9m3?#fC**_8eB8qm zbJvU+z`g{LEBkc42=)A7mr|mI(CjRhf#?eYc;k$F5l(&QI#v~>|7n09ThdYS){MxA zw1CpTL>e6gz4@lTPE>X{x1q7xHrs$DfE*jmFHi32FYV_Ck|ci7No?<^*T%1KB)0jt!Yd*v8Bw|j`WIUo&rxz5 z%+%~B1oem4)z3FH)2cZR88(yl!%_zFe$9_HA5BM1{ntos#RC$Mx8USdWZ0?9x~XL` ztkZ{_z zBxsboz3IV#p&52F+v)$7r2a1wFJ`&(?~iQV`!Wy!n#WIGk)P*kr8t0Im!yi1PoP28 zXfAd|4AF@AY6vtURmIiU$r{%|9sCm7-04eS9fB%9%sB)}@-o|FLQoblYURf{(Bv(R zCQm%h7DyMZafS!I5p`jxfJ#>liSr^USF0B`np=-hSeYB6ys)wA4MEtr4oc*g`Yv4k z{JZPrvb-00?|Nb5HiXo*q1W-=M#b_VY+Uz><+nhpgJSt&Fa^C4b(Q-kf9d?}!5jA- znuCh@b(JPmN6N+C;ltaB810N^72iii_*36UZ&tt2S;DWI8|V{k zn4Kf4`=+CyFRAT#Eu^3s)?A(`8U?$a15@8oL4583iQ{_ct-~!3lbH_&1`K%PeZvK@ z@os^cOUHu@m6k5qgMqhdf*L8>62*R>=n_y$#UxO72J0MpfiM-LzILl0j8xV7fqo@` zt}Wj_Dy56$NPGhmcvpS*sE+s{oBNnm>bqhAU2bUO*pg_ab5GCAODclmDd|&Q_0OZ? zHP^XU>62Bdr|IvK7ZY>2VeqqL zkn*J?LRP|n-bt5^L~J#l?kxvnz|s*>DygSi+QVChJe(qRgnT%qlPH~Q=qxzU5iX66 z@MY=gCd{O$3G2-v^?-aR1+vIxpa)zUJzxctV)s*X(zl>yatYn{mIZMVY%b0X`VIr9 z?~9dd5pv2RM;KLeSAt0;yXsK$bm^#C4h3&aMk^LZ>Ks&npx6Td^l60(&|}B(>62zf z`{`gVJwT2#_2-U0>6hcV!9|BHH?@z3TU_*|2fXJvVWf5@zK>&#>Pkzqtiw$kFO1}q z&PXung^{+Z(+VRs_pky?{X0zCurN{={r5@J3WSjrV$=(>o|#UmW&nG_$OYHum%vbT zAaKe<<5Q|x1ua3*Z3LLuQhDfvqQEh)LWU+UG`>Po_* zJ~~o&2H4)GCPVyKy#KN;<|uq|+}iEwZ}wC$7@Y`H6qt;6LzAG?9eOhi})1p98tGLwk8qHDtsU>>11o=%y)f8a6` z5tkvJcgN-CrjrY;OEpRT44=IG9>{`Lfq-UZ1T=$JbnhdcO-KS0&{lqyXc&#V7Q-~{ zN9nUsz|`OuZj$B_?4T-j1ud)C4gKBFA`Oes4TjMuKeT~}Wo1Mx1BO=2V06AwF9Twy z%FcU8*U^N*fUZMFl1}K&H8GyS)P~3=C{A7J)?D_f^uh~(MtKG^4H(RZd-a0$(w@O= z8W*Q-!s1lj;?MsMdlf+1m084Kpg46?;q#0AkYOPU{t2ef-vb7tS&m^aOkF|moQpXO zN|(uX@gFdu@ZFmhk`xD63~IUf1(&0KPQ5hVigVO9z{4c2p2zrt+yQ&EV=_pn*Y+%! zk)+U(m0%Qs?Y>lZfbAe+(JMpzdwxc?bw}pdM!G}&(@4Et8`1BaU0jp_O|W@VGmX3x z^viCVys2Iv-xf_1{Ju)Jg=T)#@@oiwK!9lYHYO6I)~3ZZ7?C7K{pXQ-eb8|CdP5I# z<)m@s`H9uswe?UYp7{xs1E9YQ}W;LfEEOHN*YN4cKNA zRB^39HKRM-K-x|E!?|JINHyce)^xJ+elU+8G`3=Hqccwdw?T;}wJUn}HP+6Uc3H@K zp>~G*axO-KqOCP(XVzR6;knH$Thk;VlWe>#9c? z{7zKqHvn=or~OA|`GTHP>32z=Hq80b+v}^XhHTZqZZ`DYr;U#*5}p%U?~AX7Snue~ zJWk?}4yv*BB7F11k@`w9qrCs{^8E3|`MvvRm*)rTK=#75Yy5_Er8mNf4BdU~TRBZp z=_g_}5$i%!_5?2dJ~lDmb?oCK^@T=bYFn@zDnv5hze{e&4TE3b9>Ib07%?q+H$K3_}*c*s+7^3u3@bP zG%&IKR+TVs$X1njTt7r)TUF}8U6Hm+MGm)Bl^Yhxx2g~-Gq;EE_)?tlhuqS1>Ai5P zih#0c4b9y(QCPqe@zj1X&LtA80qaQ&#amTGj%98xb{)BS0RMR^)=1{}KFIt``yqtI zjLdJqL7>7p-Xx^*iv?QzBFo3NNX~Q{LKfAk{18qmhNtWNwXxlHI{DTbq<)M}L_S)) zLy+Dtb_l=R(vNYV_gfpipJ&X628lp2sUMKC|OkbD#|5NvFL6#=R zdDxtuMPs(OT7id>F+7ySpiGfq<1&5De?7A|%9haddruwHFB!V!LP=s^*Fu_8!`p#VwnB8oSV5Ghh5criuNRp-z0*Of1L)*2Lg@huD>!XE3@*eFSClyH5y;p?^jVaSD5?aIRed!up ziLy3~zs>u1sBQ$deez%^HwjfZ#e3PiwojhSpS=6TjqCHr<_9-sw+}vh^XMhbc6^%M zo{w?1W5B$@Okkph;S)3VGqcDG7x)Ka2(A*)k7(EU|LsG(I5zQn2kgZvt0<9){g_0g zvvisIO}Ow;iCU(G+(zGIzd2*%umkMX!p&WV+f8WAGwd4nW(aYpCAW)_urD0YL0llK ziYf;CrtL?`y4vtxh@5v27&wgXZ#*&@ufa4{aH9M%5-C$ZMj5Glh>8liJESPv7 zktQUng@e;`%!(1n5_`tVB27TF@<-h7x=?XNu&uFO>MUtgSatg8wfdtk;cQeOEg#*x-qg!27?wK(&?;CT)Kj^5 zF)6VJK;k%*I1ma(?#)W@Z)Q|Dw%wFht^4d1!7km@z9Y2i(wrLBpsE7z&7 zojUbag4E(WMA@ufft`m>t2*`d5>=7@nAnZy!jFyeX3ph1gGU!ITV`)4^p)TK2=sNC zmB0X+^N2Ey!T$duz#|*eP3{9ZA6B4ihil7YQvnw2a~iVhSTkQ{LsnrdP#_5!ZEZ~T zeLJ9Dcm3 zO?71byp5Y`d@c}+!t&h%_EWP6i#I-aA*xzgtT>|j>#_3X9Fb&>8-AxaXCN=X4U?v` zhTP@0$hFAL`Sly0J=Edj$j4bLtOnz)M|cy=N0+zG9{#}V_s$;OdLVjZo{YER!dMV0 zhu^KYgXOHPgHBpV?m^rZlNE&Xx)iW~b-?;eXp#@pRfI%bFNj#^VeKFi64U}rBUf|N z7tRu^W)KO8R2D=61Gz>JsX&NsozRd;fV76L)0-qeh}h`p*KiOCV5Ip$#KT83z1}7_ zhy-kSqCYGVQBl+lpdi7BEsDoR zAIYUOvY=#EIwY#ty@F{AWD#7}oF1O&tE$dXTC;oCR!4|YJ<&P>Bql>4fSc+^K03Qs zFRsN+9R)Wvb8*wOURN>a03GG!>;>p(nd(roCjB`TMe7v|RbWeebO4cN%}~*L*H&MM zu~#@LjhianfT8tvHdspILK|nYLrg)32rJ#9-%|_jFW=C+OMf( z*tFi>g-zG0O~e{~%-1zcMe7~WzB+VT?}fr*M`iC0&^yYTF+k{DXF0?-6{O64j)0W&bfD%A8tEjpBItkybv2zZ-$&KH z!jn--yfh;WpnOUrbxU?@{=qDyE&?d?oY9D8x`Je$QWTiHUR5}{70`0ix)TVi5!AE; z$rMy0$xQ`8E1ppSY|atZ=hB7=sdpYhSLOWJyO|%>w%D_Fz9QNp_$~b)+1^c-aMUGBN9Pr(@-?alSt(b&dXdQ~|z(xNiYl8iY0}3_;8sex4#AjX&!QjESRaE3-wba52Fijz|ieia>4Q8Y6_f&oGZepL>p0%}&#>1#5P zRr|VD%@a96hC(JfG;0YMQ&W%>y_ObZYAqO3S!*eA6q*o_V8)Nc1rSP4+dM{HXzKtv zdZGws0_y-)BRU5df*Eyk+9s^WXhrtF+g3UCnWWg% zC}+h_o%{EwVABuE;hTSK=|tjag-mX7mQ#(c@U}U!{l`H|9nj?YS@jx z;Q`sFj=l;u{V&gO;#ANw-N#<@HHv!&eITo%mZFUx$o(kTK#d9K-~%NazamCZM_hNC z2XKOtjqhOv*=D-Q-}nK$+55(i1Uz()v67ANVF&pd)d$3*I<|wJXmt9GQX1$O0&kDA zOL8nq!&*=pQ1)ruiK1(yb!9R_l!{P*BUn^~wV)#8;0V;bL+6ZEx!PLw&{|V-PraTf zt$@zR(CcZ@3f6*FkiDM8SOU+D3a|vK>b1(y`}G1WLEW$KHiWoeH&_C7a?&okW2;zg zJt!<;)H?v(C*V=l0b&V!T;c!{OX!a47OU~`Y%F2eVF}yskf#RN1fz!~$YM1e*t8tt?43lgy9MHXBDO}`T-eW3eV`& ztzpagt<2P|1Xti^IGuO5rU)-PwrXr)jS5&J?#5UL=n7hLnsU5sIVHv*b~gqi)R432 zWc711_5#f+;hUg-bqFQjcu`yHQ*1pSGOQJTI|FIpeRWWX-E2Ad0SjudKw8XV#PYiQ zfl2}sQjVWKVBv?NJ*)vr>Vgdg%34&ho&rzgZzq6I)KNU+jUI=B&EXIb3bbm%S{+I7 z6G@Q|2;`1e<^ys-haw+fqYhQpfWk1KyS9M0VMRW`!yWRibyMI20@kzt4P6u}Yei(Urq-hNF+f@&i;7!K2 zI0_%nk|dYLKur&InhC2yxZ-wcCZC0t6xD}OLnRu6UQkG57>@;wA%8(7{(zpR6t2!7 zmx1487T^z5-AnvoJg&=S;Nw}QV{;j(lhcL)i$M4SD8oS8sF(_VfIo~j{y-%|X#4?` ze)t3#dPjH3W$2Eb(vMN$4~36|t{wcL6S@*y2L2^<@8}(-n&pgQY9TJenmEqu3V+}@ zhQJ?=b54~8fZ+4cb3?z(eio+bro#%#p>VPxg+Sm5`m;KxYuRdkoSV}X9Q;KOI|D(O znvMJrD=`Re7e`f&cqKrCf9f&tjxDJv;mDU%qY(H5pH;L_bjtd}8T-3|J+Z!ZdU<*F z@bdoa4{u$ZJ@~;}Z@zW*=$%`yojrPZ_Tbj(z0>oPb03RvYur^{5@~e6JLPP<<$O$= z@`u^8^QU8`Oqmlr$m0zUc=L)+))So4_+U?B#v35~H`m#BRufS+LfD_NENU-Dpk+?Y zB|vK24(;AO;^pAbQH^5IaZD`z{QxjlQvn`2=Js*ese%261NLmG8Q!_Lxc}>1Eb4GF?`$7yl> zLzF?1kDk}q=9&%jt_a%t_i*O7v}GhkZY~?rr1e%kW-NfZ_1oR zz@Cn(sU&hn!K?=JP{mHdLz&7nT;E9$S%*u&)hond1k%KKBE)GrC*#uM>WR@86Pz!; zmIo#B}%Yd5y#PLKuL7; zx<|dH7Kv^tNOVPO>f%V4b|CkVXORUE5?8;{FseU|z%9}4?Af)C<5E)B9^Y*Y(Vbn_ zK4#m7NXZRwv}}}ZP-$tpj}lkrW1oKW(WNBh!Q>iU4{lsrxQK7*chs?<&|z%bPAl%A za2P@`2^VqX!h8o^H@&)i0Ri%{^pGM|`?#bMcxCP5xRf}#(?_YJj>EC(i?=>=vh28s zf_oTxWeq!O1Nz|-eKrqpOn(Vt0;W69hw=`<3jA35F%v)=cQmvyux{RDLL1_sT@b}Z zq}Hg1@C=d}QWSSJ=8>S2^26O*7iTA2!q@IUe06^CKA?qKN16zS(U#ZLeHO{9pboz4 zYfU=_W?;?)vnuwVX6$P;iF?2m}3nzsNJs^%rkVU7ee`$A?%HQlgZ zCE{zI)#PomyBLs>)DaZ=PkLDqv4n5~^Z(M{=!2 zf>9Kz%Aj6P&Dzq7Q!szXw9pzG1iU-n;d-ROC`9^MMJ8NY#Ji;+-W6&0e0vbch~oyg z1I)Pyq8s2xdT6kw5-u&`+)@zdGN^jZtsg)nlaKw(gjo}Q9MhCx-j6@KZP5TUF`l`F$F~!z&jaUaHdc zC4kb?gL0v_1iKa5wS!J1qgEhXBWw53CESZ z3j{Pmj$`XfnY5C5DPXnm7p=F-@<6vJ_lhD6J~a&?v4}DspZMZJHPq39a#Tyt=>Ez9AAEsiZ4S3 zRFpP|F@j#w_d-vz+w+lpFC>f{eB6c+vHJw}n=?kbTywr)ft%O|#O86?J+-9tKvj<; zXeoky;eejL0ByQ(W{JKQb!^D&>RLL?X9fE1a3-35j3t}Ip$V2~w!TU*a~9W41xp!7 z^%YFHe|O-(cc^$}?!X3+@Lh;A6ZYmUd@QNZ9e9vYfzfVC+(E#GlDf37-odp;9rqE5 zs)g>rLy1Z>i>pMRpexC*k(3q6pQD3AJek=hx+{x6Cq`csJ6Pi((XFp=t`*?G6(*zG z`cl<{aAk4e#OPU|5J&;g=_@LrK-CN$v{J3+Rv0` zIBf+hDsdo80=?W85s;^1`L(1SRry+Jpy1laNa^U0d{A**S@qT{QEzRCfYixp%aG-j zYCRtxpkmi} zNDoI<<;=Bpp^jEJAzVp}+a&^08$Z6Ga18E51db8vW$zIIb*=(vfGbepi!Ra^VP2m9*S6qql}>93R9Hw6arnL@WLRz>0AXG^^(EqM)0lia>3@VCB*IqjQB1A5#F zQa3Hy&^vZT674oI$tPmeTnK9Wrhsr3UdChocK^(uxXrfh4tG5dEqhA~9Y~cc3VY;& zYvf*p0qI=?jZ|(P7>CxrDe#B>K8OS@lG>wLbw&M?dl81Dbzrqpb(?)vQLyxN*o0sf zz(`#Gq;R(cJ+Pj{7W{@21bO{&q4x|4ZUr+x&!`-yEHUFGL255T->d^7P77v%0bOIE z^&%wAb$7N%TffQH7XQ zw)k%=!G9z5BJ@{z8ikN^Tng~xCL`YIyOahH!|2Gn2QD*3A#AMYoW$q=L@qL!l~q!{ z;$_lXh`*t9JX5=j846)z!O1I7eNQsHEi$_T#pS+xp<~LQU^}E&Fwn#`io%o|i7AWk z=G!a$N&jaQH&gDTJyQVIygGJA$GZoz{!c+H9pHLUH9MG!96F zt)$>JS=vz~!baX90utd0+?k|pi(g&sNQC_AN`UVqqL-%&B}3MyVK2hY^r<8DB7`$g zNPJr_-v>x_X@^2ML#riUXZj8lo^v!Tk3WG_?+g?vzEArIOYq*bBPbMHVf5U6dZ%Hq zWc36uKMhmcjP1+>GSP}ig{2%y@AD9OpP^S2qV+mH^FQv{lfQQWX{es=^I3=FHe_ylB&K^Bw_CP1yi?Fdcdt!82*`ljgi7a7jrgdiWf55!+vpPDc zqzH(O#k^|-^DYNe0>y2Z9D?$+%fzkv!bbI*YpCzHDk!NzxpZ_WN%bOZEH0fGy=+a3 zDTuscr~pEus$Z)OEz|Z;q9_8VD2RCw})*&?PF zVQXVX8OF6U(#}v2=?=q#k(la5D2SR6E@~BA#J@w7&FV$ixk%B9?;|;LtqycxO+joV zPVQ0=skBdyP47?;E}}p|Nc=J22t{O37zp5xp}qw6BJ54)I#Ly-`LqH*fTX2WlxENf zkb;ThpBJuKz4>}mCt7r}QJ4qkcw?!5ts`y>@xWAI%bQs+YLvfwK3GE??Ul$|bQ1f2 zXYBhEl-V7X@YcPvhZn_^L~#3zQY%vY#E)w(?p^lb`Lh%*ZhZQT7eW8+n!;0<991Q- zk)B9^VE^59_FWallx&WDn$4pAvZR4d6R&(mN7Yqhe=%e0;svpgx2p5S~aOa`Q#eR2F9_|8mc4gr% zme-^WZ%VlH2#w`x&pCQlHADl|qWnJ(Uyg!p|48*jY%Bs`BM5-cmY?UgRGj2;>QjRe z3uW=q5okqgY%CIABS?HX2y~aR(E5pGQF$v6ISDnYHg`={n>z)nwt$l7L-KSiTCa0! zaq`6Q)uC`?#iqQcyU1l^^iG0i7!j(hk zSfC*`b~Z?z+mdsK_kv>i2oMqIQR*Yrn%L+R%XbQ?wuESq&arTHm%J7M@y=nl*Q^$+F*{C7-#D{jH0vmTFpONL5V^ z)9wV<5r7n~)^#{=A}VIN>#zYV>yU_jald|qLYSQ3Iy}g#(4f<`w~-wcuq{6fgWH*U z9y$hAVQW3Xb$BRMsiybgI*6pm7UdDr!6}|Z?XqBtM<_;5mSRTCsE8KX0zRO%-RW7e z)t9lg_<&;cEKrKtPfbzZL=~pNqgr0two@uVQczpgP!!7VqeD@8RV{vBEBJk>0^7){ zx;PXn&JQueBIId70R%{A~lMku4~ zAevyh`lTvEy^kyC1ka-jDm=X}6(kpop9>9UV(MQDqp(6uc!4@bkgRuFiD4<|%33nh z3t>Hd+ayWu0m7^nY%W)%rVZmzVCvaiDc{FW-#L5qiPs;U-}=ni<>^kwb3c<+sSYX~ z2BCaHgMKftyYmlcfPTRlS}+RxlNo!81_%WQn}n9s7AzMr`J)4NcY^FLYBEKS$@eew zY|h^;AO)(e_WaE(XpuxxV9@GZ6lIF$O^B|o7i_~%bTfghSGk$1^CUT_$}~4)BP$aM zKE=LpG26P!?&B?Nnw#+umI=nbnc`*wHg{B8cloXa^K=v4jEA^P(fq#L3>A4n`7?Cn z<#UP_)LI0ZRuE_=dvs!EsnX{TrNGcRL8>aXvluX9^ep&gAWmN=*j80)t6EldpxG1} zL9`lB4htQA(M!u8(7eub{l(+kgufFcmI+H+jKT0svH*jju7cJc;C1-uf`_$4(FKLU z@Rul{^29SS27{sI@!CYJ;?!;_F&OeAAf`BVurU~Fm8`~KKyhjz(kXx%U7RW^pCJY_ zC=5mvr&8e|fWeF&2BV5on~uIMw-^k6Yh$~BE(SwIL*7;x499W=21Dc(IOkk|!*qbj zf;xmR1B7BQ-J9By6h~n(qaR-Z9pxiBGku0VHD4EJrUiI_NL)1@Gnj(h2{Bq@GI-Hy z-7`-SE@R1VN_8j5u1GArCjy2?)s?;I2JLrd>=U^b{p#uYc^aU>GN(^!OMj6)KfhCC z>2MYn2!nbk(H5F`{b76v{FDXJSk4>sCZSBJx)RcW4g1pr_WY#fzH@PL|J8@{!`**U zZ<4~QTvWEaqv{&1oIrj4fZds-eG$jzH@=fslwaPzd}o^W&B36$+OVC#CVHGe-mI`1 z`fj{J-?VK+KgHwNXboM)05(&3p?z@1C)G{$I38XzNtKg4PQdc~=njD4>e|ltMr@+T z@i3c7+D8NP-rUhfm#1`<_%gb8d1n!3#PEyh{+*^JBUEM9OlPrXx?I)_HLue-xmCs0 zDb`F!%`-LdP>D5T*4CoUbb>ZBWo;#HgGENY1FUyV`Q3N{ZbMxLtvB+?mei z&QK?(ZNw_t>J)dTqvAgXw;65RCIj8tF;GCMmanA(?u=8aH7eXjlxopT#cC?H_HY}n zD>-xGcwFZ@+$O@E0YymI1jVbsyZR#zYzy2b)wR;Qklp265V)b54ez>YpbUB^s_+aH za)WmpFX+4#{#*%k!!ueVuc@}05~?+J)B7hJ09aK%;S_;4mh`6hYKrt4y-D~74BDxz znh3VT{?m+oPcWmr|M2qc(fQfE`=^&@(>jnLEm@wpcI}94`l%T~3jGo2v18`E(>*>y4qT86DJMBv7Iz{#C z)2~y4L%u>j`$~L2y~ezqJ=BQc`xULRM2wSa(HnrBgOW>LUQ_@ZuZv8o66a5K{o>v7 zp#p8rUy;FxW^Be_dPV%}S2^OEu3u{lMQ#DZG`Uz}A8r9z-puSBb!_!v2Rf9jx_*gP zX+lI*hPjqdviMGpl(6d;-`ns0)Y<8S7w&6BOg(4Xg0fSR&|6Ts2e0@PI-hOVFAgrt z@ua2b(Ibyn#KW>ksPqm>D~hR(U!vid(6?bt_PkD2&Ld|))$xn>$=X^y==jxU+2=oy zh`aE}0&x#Np=A`EObq^PIzWZr@`!$40`1(1W3Vdud5`FKA_#lhi1|s# zh-vC_i_SF~U#B=+Odqw9jtKNFkP0KTDbpmd+_*m9{*Sj0K6~@% zCC!<9n%$m{an5AG^ua7)o^_r5%q;T41wM$6xYyA&?Um$Yzjwf1tg?y{nb?m> zL^?~CsbBl#0Fquwt=Kj4P4=5JMh-i`CN10?xmLa@cB@-(SW7SuBXlZnX1_6GW692S!0{5znjKV2epJ%3NE1MjVvzd)V0&Mw*bQ6db(H>Jd@!vPcsUtvu2M26IiKB+`V2O}>&*OZVu`DT_2V?2S8k zG}7%a5cz<4VWjbZ?gD?fDTy=zqq%}e6UzWlD=LdL9=2Vg+1il?tz5rp@Lxa|_HieA zpHUVq z-T^47p33cMeFwpcqSRe_!Ma{ii@Mtj>Tbb`ruohQiwC^|IeJw9Tyy0s^FUX@*baEt zg?{tl)vxPbfxV>5?uz_-@vMukQzuVO@ts?%PQ4%GMMpmph17g!uFS_i{pO=fNdj)s zN0h;OXpp*cox14A2|#L&m`X65TBk0o9w4qu7oToboqCJb4**dW-?^7Jb55Ol4{&a5 zd$Ttb`pR*5fxa%Y5*U%WDC~8OTCg?&_C*NInQn3)$oa4WU6+1<1~lL2H1gE3X1>fu zo+6Mds!x8t)4=0ss_zRTt-@Mcybit2sWpP5qDv}J6CloT_|R+wsWyc^kLm_*@_ageBAm?5AcC7H@p;6sqc<*S3IZ zWwGLHuV0UqFXxCPbKLMd#W@3c`E8gqoi*exw?*!PZqBdY`0Sx}!U3GM!fG(ydW1K@ zd~|v1?BNf*e(&tjtp}nv=E?e2To?;N1*i{7rIOMFK-CITq#s4zeV+ZR1J-9s>w>7c z+KQC&#ntO?O=AN!bQMNP(}w8X4bIrVo3U_@8bmq=In7mPU0o!lX8A!RfJwOiBYin{ zaJ2UTV$3xY^UPzx}%d;>l8sT658EI)_@L@Em+fq`5jh*Tg%w@zrtuZbYCDF`Aq zI_lkrxx8H&JuHJt?%fb@I^BV83WA7-k7jzkO>Phg*ovQafg89AYKdfC$yF){A|5E3 ztInE1WP|aSvMjz?r(y{GZYmud)r)I!QwPCK&0JgwjB2}ZS zINXxmntw10hbunO@SM?zX1ao8o>DZJj5o%Zesqfg$>}w;1U2nIG6mI0@=yUbjUmNJ zVK6fi$?&OOZLG3lqrn*uOZSU{gRo7wF5=&SI*4XlHyhUAJL~4>Jef z`e*Fl&RF;vbqk_^8g#I|JPp=V048eoW)39NSFxEbd~Swa>*a4|8-59Z@>6Zlr4`Qv(P<<02)UdW@F*XpAD5aT3$k-d%zj^*NHiuh0MvqX{(N{lwG< zP|HS?qH0(T>?dKaYYHzb~ z-Y2a#I4|pr&3cix8^Z$HJEo#_}j_*fx6k7)i z_~jW+oC+GB``Bx~Mr$hrb!;UYKal%Tuz?y&n}ZLOYar z==HQ{1!DC4^(@8`_&s6)mOxGUw93H!y2ZAmQ+x#Apo`T;b-%vb5aQVygC$TWr|rTj zRvUQ%mK>F6U#3{Cv9W~QM^*V#HB~yQeWzrXh=*_>me43HK@_WDrmi=HU98r6Sb{26 zFH0$Ra%=!agAmp`LZw}vg}w=z?=@IgQ%Zw6N|MR?h3*W7z}SK^ub~3 z5FmO~nR&^P%0K8`V}Cqj-xt^u%i~+8mzQS`FYmwp@YcoIgCD&0=38fv-nsSK*`tSN z4{n{_J3T)+_rVCa$lW9+u{Pw+ECogp}P7_GmGS>PuoGj{lNc)%8GXj1ISDKJ3rJ?WRe# z8wbad-wh#TRVDXuGH(9Wa}!seP4H-l zcCs@**SU0>65Tv7xYX5dilbY=!Xo)2>AdmQ#n}lztG{;t;j8n5_nClH)4y+~L^ls9 zE>^iYU{*EUq?$Eqwys26`|+8?jyZ4vDyGEni)n|5j01p?suGJ$br|A^1r#cm#Rd?k z^vbBdG2#_gfvoMAN^3=BTLUUvdN4)KT36z;nPs*JZDRQ1WtO;dQW(8g;1i-8Ff0L= zvsOea5is2RB%NL@r+N*oEr#3LX1Ha|P}z)lRT0>XqPeZpFC_vVl<76pSy(gd5T9Hu zu@$$lEq}RI6W-{@J@08zKHd%n77y175moVACjn>iog7zT^&09u)LA_t%3G@^Wi^J$ zh2B!%*68~c^jn|C2{*Al~Tg;;qk|EX89O0-6tTqY27TLV(9Cl8eoL3nJO3b4`e3hEG9Z z^*xnLhYwfp#06k`4)Ge;-qcu@o_92O5AMAx9(-bi!sMd*Fvu3&Hlmwi%u>-6nh(8b zMbTrJev#-r|Nm$P@B{m}6g4mUPL0z)$DWx#m5bAZ@7t*udQJEr!*yw~O$I(;kAc{uT_gsAg6Eh6QP2Mg50iRn)G2cgCJ7+&m6uTBv0c zNGmD=6}i}71b_o`?H6x$R!uGo6HPQqD)tvPKxN^dm{vnKFBJt(dzNy4;XzVmmb@wS z7XhoYC{TaWqmA> zbhj51mGa4ZnC(Ck?iB?{5^4dDW;=$%$mc>_XkPShvS!obCsezMeOqB8fci%4R*a1r zfk?2`@Dq(|jGqLJjHXbuEc_}+qz5I<)sA^FKxC4{rhv3eVvU4=yK2c3`w=qEY5e?kBDCcx(Pykm z&d1*DWKnr3(!jcL@;)`B-b%4Qp+axoB3RJdzStN?^+ffa7i3?ef^eSfWgwh43^A$p z{?51mntBQYZ3XDH!2#Je*a*><0{rOos4~29=mK9xaPHy z>v#4=Tr7`Wg@CXpSYhl6Se2jGnZQvmuEnyEWLztcU4^Kyr)qN)yvlv1P{j2y%HX3z z!=ID9L_VMehY1QF4Mub|bH43On&UDw!iSRZ=m4=~C9aPl2ao@h2@nHpC2fzIVFoMQ zxl$Dnh`wg$vc0uJP(mHO%brq;B_fC!{9jYFkh%f_(TFgk4+-Sf^ zUNCX_Uw3d0NjUCv|O@R)!JE~A3WbHHXlHS1Yd=BqiT-<;4 zVX#}GkByrND8zmAPez6hNPT4w`%nagPpgL^4Cr2D^G0mbIiXB!jlB=0%Mh5N*>Ke~xXtEYA|ug&W9GwD0muTU`W!7-oO|>U z%~gDcJvCp`xr)L9!92yw_p}>1=MOCiG`mN32rNfMU@qoyRbEI`a z&mxsADp`_m`ygqorrIzRimkNZq=!05hTSkQLid@4BJI-nA=OMz;CvAAW9=eAgL>a_09BKyL2#XK(UGM@!Cg)v zz&g-z-c&FL_aXx32z9kXPMv^JxTHP@3LZrqIO46+{?Ma=@h>ZMV@-V`Jw8+q{SiP+VW7vb_XLzM{+)42QBB>HU9cqF-@zvjp8sbPKHS z#OM|n(X|+2hs2B}Fk~Hh)Y1d zqtn|xxS1&~;?iQ?iP5uQWq?1WS4MFW2UxMr)v5r*tZsLDnE{c`&oAi!vsq?~Nw*YC zI$9%4f0?JT5qZ640o0teB3g+k$Bd3QD>mX%Hj3N6O}$=+F=24Q)z zy2kU{n#f`C@fXv-$p2}*>~>}cDOXq9A<_H3jVhvdHRO(G-bXArwU`v@4eKMP9-W9T zo>*Tu$s~73^rq$ls5$-?SO>Ad&Hub-RlfNG@U-HRgrQT^uE-tFd@&u&v{1|VJ)Vv= z^|Vy(F9NWa0!HZ!Ka9^@O~{Sti&M5BHb7N(PnQ>Sy4L5j8R;Doy|L=}3u=c%|MR}T zz_#qwIFm@Irgup6#jN5lHs;(_Z?~X6NHZCiDEp5NKFMZ!>!rp16QdIgBwf`?JqL8+X$5Winy_Y7GA>mUycI}KumO!4P*xrtYEp~3 zvRHX5!OEjmy}FB9%4XzAW&y-R*1tlRR%lH)9bHmv#+6kPE=D)#63jW#Mb*Y3Ye07- z#+7`oWxb*@^fa51EAz2Wzxn7=7N`wW#;t~v*5<(tAIS6`)`)I0^}_QYYb_<4kpmmM zhdl=V9pe~L9esiUU}WhbNxaqQwyt#mh#oqgVl%EJlJ2k>IUx-Db3JLMIwTH!1JR2% z!LAYEMM5Mi(7zA+Is+XN2TsOTfB~&TqWQMMMwaC@nP#vEAcD!^9|-`b5p+l#IMZ6m zSav$pkjPN)EEfYk>PVbhoN5+gSDdPm5WE<-;$fs9i~G+r_V-c+nyw3)R*`4`wq<$dGeCIR^{1+ePvL;sB=+ufB1nF z0C$yFChYb}=J(tL)jL`F1l5~OSy?k_9ZZUU0a(vvvniNw!Ja(X1+0@TN%&%ZgB{*} zY<_U#Ak|#4s!lA@;B_`B{)NQ9^7v;7w5RwN5Oi|<3ykU7&V=|E8a>JMuWV@P%{M9j z*>LlIqLc9A&h1j&u)3KR|2$4)xf9+@h<^cVBQOoPZk<Yq@Z8dnyTa3!dO<;s>gwN^sfwP?9eLLeQiRt(9N#Rpso zK43=SJ;$_?Q`eh7yaBUC)JCiNex-Wvt+IOWX@#KVK00h|7TRL>tpvNTe4!_CBslelIs@Y z*(0dVPfv*}ZD+`!M4wJq_P}m_1tT|xU7|lMC8#}GCHm?4!H$i*?BC%?j;c<7T;=Ru zF7%cZDEl42pW}@g6iVI{WY+37 zVCAC+29`g1Nn@o?v)l8r6f2F~7NmCzPnk7gkXJz{fy>W7>S|!b8y-|mPhwXGnc|yH z59Q*kdjjrLy(2ML_NwAXj>KuZ7l^^fvP>KB;{4)Z1&>&c5xu^_SWqQ=TFW(pxw7A% zu{=it{@8+M>9Pf~&6*++h?nfkxj7M#+67~a)t4?mkFA<-Vg8;1Kx|d*hGAIDnAsoB z(r`kv}3a(-q}y^A_1`BYu1Zwrg#w>;Hu20tZ4ZGM&$%A;z3%ac15Z6@|rW8%ZYfwfa9lYWx);9faEJmUjJzew} zu_YI|qR9iOh4d`Z1GAW7ST+{5uo2Y49MFo|<4uw1MyoorQ7r)*DxphZEDg#Oq(fDD zSuLiZ7`=E|B~FE{rrr?jhnRtKc!HQ;#>*Q7(!QiMhgQ4O(J95TY^-AUMijdPk%*Ji z<{>Lvd&oor(o#9xnY)q;nPDH~|eO->_DC#Fjdv=+;{fn{PRGb&nwVc^#&8 zffpuU1eOwh{Pdl(N1u58(K)9TU!KY*Py7^Dr9T)n@<|OEzQ~@R-^l?DgR_7XOZ*)U z#v9N)O+<`>pR&M|>T-C*fx3wn`_lvV{1oY3P-T)T)9-oaS)M-Q993VJ8oO@BK2K*C z25rxMT4R#-_oyw@&rb9>fpqw*;Bg$D7b@^Xk7ENbl`Y#xOA?Rn;727NGM?&jJOE~j z&2J`ooPg!|p%uWn)t4&Y2~G7l9w0MG`?&3BI!$f88DIuw-_U`X&ns3@W6^IKLBFB0 z$K(E;rX?d3%AoO3m}qI9MrSMnO(O_2Ip|D7oW4%Kt@_tS_0DQQVtfoK{VqM^Pnq9K~fWw4`D2HPlzrmzgQ^&lEm20P}K!8VWv z15hB)jSe;8cN9d!p&kLzPzec+UxY9X^?9G)hVX@HA=MS-JqoG;a(`4Oug|il=eNs~ z*E+b3$Y>2*V~T&LglY}h0D4~Kuv~)C_{Xhlmh`-LgW~^8kzP|^5@qrThE_WRjoi|S zHYT6Uw(D0<&(EjfH?Qa|@Dh7rez$B3z-eRvIDbZGDHvx0zhxmfm=hD`=GI%(mpl{l zVSjPJUYNGzyVsocR!Me{^-vh7%|&hd^EWxjsp_l0AgTI^19o@X3OHn$X)9oWoch|q zw#Q`W6G*5jP)=2IbvpMd@CZ!bk6cydjh zqBJ(u`FN1dv=!K=!cgaLqo(^lf#n8me7H)8l{i(AaR;Z*$sSO;XGw9NsSH*_G( zgL{}_-n16;rnQ+j)X8Zhu}Z>P#k^^$@Q{IZS_{@$gKDspmbUzrb)`Zu)!l6E6oR!9 z))C(!`vHKtbS|v3RyAO2Diw?Cd|MOMMwmBLvXX8A&yjBd>m+*uMsf#$>$Fto0vmww z_Jk_aPc{GxPeB1YT^$t0#KDpdTHz^*_n9=Z!y5}Ki6=#D1?{Zy@}}B{>=L-gmfzHN zO_E=MJKRguma)ze&rqV+AI#YIh7-;E4=>LiouA#ie|mYgw^nFRT)URh9WON{2=U-; z(0(z>NlJagXCv8GW&wy5=j_1&{N%rF&oqDHHk)6+@!3Q1|882gl&q1w89*`F zmkwf;UMbmK0$3#ariO{VNs#(G-~Lg!u;Su`al*8jePzZXhoq#e4oX<4YO~=gY9lRo zaqJ49GHxp>g%TodJgdeo3nAQ7>0P{V7hDA zO)aH>A$NeCZ2Gg~fwhV>*aFgEhC%>y(1+UTb)9CBE#_b=n1f}D>~aRvGM$6kPF|o^ zfG}q~ym`~yR#ajSs`g{8#U5;J_F&cwwZ&LPz^$SXwtxYc)~;zk*4cn_`f*@j#EML~ zuKYynS{AaZde2Uw&fOtk^X6k;)d83>-xzs##4KPsbKlyaB&1+QSZzUZh>Km8z*A zC2Wa{37|}0tiHbjnnVUN*&;6EK}{(uHy6w%6MZN4mPJ-js$O z4+k&OZmlEC23MrQh%~I%p(JE_y*vkNI7V}_?kp0r7=2M5&&?uZThz#cGKoESJ6R^* zSw!Ma5Q%fjS zF=);pyLO#3pr6;)yk^6^L##PI87c^S!2ZoFMh%XSo&sR~K)=m`YYP_7fPX7mFr6cm z%zh)!Fz1ZqWxgK`d%>(RlUZ*SeIr*jH@6gF#@nA9=ZX_YTa(`lCBaoB_25=+#ugh| z%KK9XY{-?@#apu#7ONV}u^4kctD>(kph{XlCx2$YGGmc$If#4?$eOKE?4oTfqCQN_bg zvsGGGhIEWFfFt19`Uh}2c={s>JhgbXo#5G)#b~3Eq)ixMP>w7eJyq`(cNRsq6BOAT z^mItCgO*Z}hWlfyCq>7pdOMs9U^W>FK^d=f996IGD0z&#ISIyV?&?YmH7SG93}g#) zM5q8rEz=#!0i~m<)Z^MHEV{FE!-blohiCenF?s4nN%|~Y~caos(f(0 z0%2dLXKJ#Pq>%&Xoe@uzmYQ#cf%LkCyKe(5%6=S64d0g44JbFH9pH>VHA(JiMFP)!+oa74Mc^wOl_RjUEvo5Bk}Il2 zorO2Vh};zfi6PEu56-dJ4-?E2^7uU7uYv=x*KqD=n;lp;Z>Tm4z6L6T(w>T$1Yb5<-7r}`kDd@ILs!%f)7wrBEVAXsgwTKFp)MLAHSXWO@$F8 zWWzGf+d3ILW*?e=ASW3M!v~C?%nUeSYNPYQsD=hWm)*&MHIn1wfy!5&Eukrj&XZ6h z0@XXRPS(>#)W$|0-H_Rkd1_?x3l9`1Yf;5|P?(^S&I+J{U!5fwSV04IAG0mc>QN<`vAC?G zZJ7@UB!v{{po(BQ@IjFeuptB<7+lImB{AH6JmrnrG9TbU3HjE#DewUS>j@t+GD9{% z7waq$EAjyzv{0Z`y+7+X0~~8GSt%ey(7UL4&;#Qgdv+O+MTqDHAp$40#$BpW2ZzHjPMc9u0v%@13u=)PdO=FaUr-4( zV2aSIZ9xsZBBTInpz2=fY(TH)^Pz@8UC(c~gcuz_4b;hL!{Bx(2o7?8t>u3ij&WD1 za&-VTtZk?v!?JeFjk;xV{jpOPw^mStD2t<+3M`8ok3FbCmc{We5yw+d!vKj12F)0$ zI)7b44SY^pr~$uJc3dEU8wS8kz%;^s7N+K=t=0QcxWWFegdDhxeqyXQ=h3q|r)$}2 zew>@r6$E|8Pjm)$Fa>}4Ayz{V18xqO{8dgF5xByZ)Rg|^OR6D=vDRTCKbykLi2eMG zy-2i8_(rE+Af|E z`*3vnDhXX$H7A22K@8LB19BRa3Ieef)F?Ga*LmtHf_8c!IB;U_vaj zPMHv&=!u&6BFy1Un-`pd;T-xEK(Rb}PUmbLYv#-BoUKZPH+_Ocx~CBe{_3_W2`PxQ z3ZhshkR}YGs1u@N7=b@u3;Td$oxdU*EW*6F>|^OJKQs&LEdO|J6f$u3ns*|r6_nBQQBw?zZ> zgQH0E1>VkQ+tTN|^hUVDo|``t-3XKkLj>bG`}Z^UN=4Br4Hxz$OuX^JKe^66R!>OT zfMJgov#7o#<{0SrRj;}pNRiU^G0;QtMso1zMzLKBICI5asSnaI`R}Na#3+Vpcq+IU53L)+T{Lp7JfaJVsb9)r#ThzH) zO|4CdZUNLJ7TvBcf?0?h-bQFbbPI`YMYlqrV%y2C;`(mbgy?XidX3z$+lQ`Ta&bB3rcI$d9Y1PX~+PHVB5>GsWjNfu$G)q}NnK zf|bZ_s*llNk==$&vKt4Ul2g}G>OMwG*-m6fY#J2fZ5VR~(+UB!H#+*~(czd`W{dYW z2;N)iGQ03{QdbN#G#WX^l23ulTM?ziu@VpN6?xwx=ueyix7al_5~CX)T;2>-L!+ft zj5#PW9E8-<+SLt>xNa@T8I&ZYN71-*-RDpHr7SI!6?o%P;m%plpB`EbjmDJ+KFjA% z`R6jX+qr0HG@>tR&>I^4sBCD&R{}hLDn?%fpJc={v#mExuAK2isWCqPPjd6mP z1n7Dwz<%YQl?-@(^n%7PZ?kdUmtmL@1UYF7A_+FanmWzW^Nz;vo6;Gfx~h0WI=4kP z<(#FWYd9Z&47H*Nq4(J0HmuBF1b_o|?T0QU#$T)^v;~;XSh2rA04h|-05(-oHrG4&wU*`n!h@vB zEO}GtF9KF|R8e!`>y*?|I{lgSCN1|D9tu^c<(j%dwV>{4H25^4tUo&V#M7j0DcNYT z{>12ur!uf781QW52@}gmgOd01D=bHgyf+H+UPguG2$C9f`kJX`RalPdeYvqkeZR3| zjgkO5o{uPckWv&;)Qf7-@kT+%D_c|-XF|0C9v|dMW(mYZWvywwfkyIlh%QcJ93=}* zqD!KYyncKz-#Dx)%dwMHmLs*_kD)QHY)omSTi1@pSh>n_X;)c}D+MP}A0IrlX^f4B zB&jOPWkzM|4qK5(y3^2Uq>_p+6iC9oq5w&PkH%rAQsDy?M3Sfp)UN%PAz)xy1l`JX z!wMT|{8Kx?M(%2CWMuuk&BaE9zb{B6#7WSABBq+OEc|Ln1az>kc36q96!AUhlP7J7 zO^s=p#2N{~JA7ADLJS8%{r?}#*!Plb_O-KOW@Cc;QLF%ON)Q#nnIFG6W1pxdw#ECP zhqNu#V?V;r*ypbkQ&mriJ_sQ>AA7TtMdhW?XEmPdvHK72oxOc>`OX_>^Mm8-)h0}z7_{YhKZti+X7 z={~9_^F{z}r3X`syhaIibf{P_^4en5iQx+uSz^UWQOqJkP9i7D6mSxo5}R71OfwH1 zj-u^~lekt)v+ae77#*O-Wfo&qz^{2#w;Vc0T*FChEc|%2)GLr4PHe<2h`a)|@cu^h zYTlztQ;#gJoLcxsg2>`K>1k;RY))e9BFQUN5WfOfJ4p)_LT};ZHRu-N3r*)Fa_*fR z4{{R4=tP7WIMh-`;!1L(0V9!jR*=)u`b?`{GZNSG<`6IvS6Tb)$U4oJSm@VXhbGtbwFY)KO%e)obJ<5@bzE_Z#`3D2 zF3&BoJJUU(Ol*a{bHC}s{Ck*w40}(zEsjLLL`I_F&3zFA@gw>s=^6Iad`-Ve5~dEm zZli(d4a1jbjCLdEe8Pf2t!hf_bD<2fn$TYU=#e7X7Y~T6I7eC+^ej@@qLL-~djW{)g;Gc)Xs?&dWHbBq`*v!PFP7e2jasdWrd!>2ABef zQ8iK0`#r-C7J3E`QYx|CO_65^SkhH^hFB~`ZK=>Rcqmelc58VCydR?R3>sHJnRs;Y zh^IQcT-9RYtpyXW+}-NwqAUyz8m&<}9UU}M&qJ;)I-M9j3p4`h0cv-ORk*S$64$E# z-MT}i*E%w40PHwFd!<7~dPyyI+*+{XN|v;kQ^*U73Lv6fF6J7o?xv$d>XFE`RopE` z2j~!Wa@sUxC8ykDy#j^+5ehXi!zpZSOen*)c5qPTMhG`6UXB^-=QNC1$Q}x)OM6_Dwu<>oq#!lz4tv%p+3_CM&X+JzQ!+rw7)YQ z^*)$&-c{&E=Lc%QC;YI^0(;iZSNU0B1WUlyT%7hatkIcz(vpCyxP&Xa50KYN&}OE=X#utLGB?H_=RfZ8 z`(hT7J0u3=cVRYKm3v?yVh^dnU;3Lq5>3+<=Rw;5rmMM<^d3?nc^&MnSmj4GMM~e9 z>HSRakmySvfKXgtrNFtXi*6Q1x3}mP65Wb!etJ8t`C&l&FpS6fhU5;3fiL*Gd17=6 zscqX2`sDQDgm;_1c>NKvLn3Lmy90-|_5H>!3NdOzbhCkB54`2}wpHZ(iQ8;`{l;g% z`VD~tD@5;*82IB_(T&<6F*F75TGG!%_Y^v z+N0+h9>{E3=dCT;-kRPz&o?W{MO4+7>EO*i$(Vq&m&eB__l}M$U$Pr&z#Sh5jGu)p z0|+W@njdEp0JDnL*XxCzjIo>aXhtc3O-EmA4}E4oG1FR+d^)8SL8zlOO7vIT1)I}? zK~f?w#w|dcvsT6E5#^fEapzBIHOh{0N{xbTW~$ko zn2DW@LIZp=y-=E&SgE{SGZUTi_D(_3qP#shEhFWp;Zr@mhexX#_tiRQP;M>)l;*w& z^Z}u7JCCx+{vo@e29&-?{m9LQMwqg8-Iku1;aoY1b_>flk|Yf<{C=UQmZPGNYKi0( zBp@g@;FYrEQnDXW$UVQDptnr))S<;_bHso>6FbvOkJd90&Ozb$eLZXd;Qz(F7U2}F z#(|w_flzt^(x5)L7pr&_iUIpci}KH}*KE;EXM|GGHNuZ~3aF@!^20K@2ine70B10e zD!*%otzx{SVuq>reH#ow@2bc>(7caW0Bgm$d)cfUWvc3P|_JL zWU;?UKvs1(dA&8?l=};h&sb*3)Dl^ORiR7kM}L4`RdXhha7^!+=!;p!UjS69qO54> zHMun)1QuR|3vfy#OA@$|j6juJ(>v_B+Ldq=CWm&k0Et4WeRL%1buEzO_9=k`xi6e;8cCqO|7sQ&5cUceSD;Z~aV29<=|WKUXm!u-;=1S)w#p2)Pvm(( z0q{g!1fx1oF_&4I>Fu%@kWCtph_N{KJ22{nSOh;nojd^g_aC{v<9e0 zaha$2Py~%ZWrST6J5#q>cI;6gNdxM^t3~D*M1qMI(74q_vCA!BHOk(-pscVFxcjR~ z965L4o7rRCTx0)v#{OQ4r@MRG`PMytg}*NbYX8cKM=MfSnJF9aWag z!||%HQrI6Kush{Ca3Q=2>ReRaAGNg~hV0U|=DB zq<*HwzW~rz$dRnNS1tT+kMS=gz7_uple{gNvaUwhr1%#QbaMO)jOp6Wg!mU4eRpG} zg3rDKA%(y-DgN02bd_KfxL&I3kv3js!L<12LDJ<;crzjX1*|QSZzdhGpOu+_Re3F@ z#XpZfIYIS%ihq662o2|D|m95^?0?>3Sa=zDP2C35-hk)0bB|S zVEKZ(Xm(mtkW0xSxdKQ#YhjEg0XfdMuL3Tmcvzf+&I4mQG_#X{9Eccrp^#ieYVvi?EA+@h>MX+PQv+g%<<59?O7{Y{@UC^wfK z*KvvVKJhq{QBV6>?rGJC7MW6fV3A~Wg)+NhI;A9egDZkjLF&sa9-^jGidV$N9)NVp zEX)5fbuYj#iNw7?u|l*UStRD)0ZjvX(-g2Nz&g|=!ALMSFW(2(2rp@n^=Wo{K9+*4 z;iH|%q;;2s-gJQ|KNjg51N=Sas)=IkA|dHKZ1GK#h;s4u^#R^Ip{n??TZOdv97z9I zmT3b5T^YTygi~tyJTL=zVUQq9{r-&Qc@pp$3$eApk<>+%LwYPKD2@v6A-f)Vi+wp4 zDg;uy7;Ulo(v|Bm5q%3o^USKI6-I8c1!sRaOFsi171g0n6>D1`hw2&|D)Vw1GkE`jcZ7Q(KXd!c#BJvw@AmdM5e`a5nczkw26!21Lr)jEX~DS+Wf<%;2-9ISyXbIrrFt5ol9PKCv^+Z7osizq0rZb2UVe&w48&@C|n9g zVX_!Fvc4{Wg~|l!n4D@=fL(!VR`g2-wkj%{_uwoJP^tx#Og;r$iMn|MTT$ny?L-yE zs+G}pvZgr|X6Trs4OUSHP&ie4T$hg7%IZ@^#~dY2CB9>hUFa%Wbj-B}QSn!Qwso%U zF;Q2VjwzlG;u_!KE&$J=5+gSXcHzrMs$%tb4xFXl=1ygYbiCTbL|h^**oA z79em*)u}gb7&gJB^U)l}^G^e_m_pss0Iflm`URYx26nW_7{r1tvn_Q-#cpH{8e3V{ zOm1T62_~7JhZ?|+aSSo}z_66?E}W7EYCM|kPTH| zmtaPOYyb0fB4W_?Tt3_+?d!O9R9!WgH=62k0x9xW!Q9yk+)yfkPdo_~SXB)B66w8^#=yM}hvDA0yKctF#s0d3Y6Xw&A_(6*IpXj`X- zcCA1gQA4YNY)nY2hL%5s{O(Vkoj!QszDh=0RYTi0L|qxs29jkAXtM_W$IyxKH+L@H z`pn5vyv{5@8!BtT@sbvB0~lfGw|^S~8m1#vSCk1!@P-TSLnmgDKB_a^XW7&9+vOQ< z0j48zTm#^k%Hb)&TY)$&{$^<96J`;RV@Ypnil<1g_si7*bAMpIugE+EAI;()y{ zZOQLkT-<;4;rwv-pXBSPur?R9?a$wH0N1In-vVIgCl1)%X)EARYNpvUEw&$3SKVai z6G*XDUPA{yFfLAQkvD9r^ReNbD{vqb)Q0ZzCOaPw;F)Cln`zD`V1MRFSKW3Bj&=c8 z$fi0U59FD)0{c`LddfEmY$Z;PY2Yp%bA%bJ;N(-b9jne0Q8VEakEbVJI99& zOMo7mowKsoIpq0iW3lVWR=nJY??KLx;|=C%V2B2KxJDoQ^qY?^6=j7A9B6h9s3I#g zHqlXysb*=dT}8G&RzQ!aBBRL$**WW_2lS`{vXgMxZ+%%h~?R``1b2Lf^a)L748iohnt za1Hjf^>c&*8T~^KAh1~So0_mm@@u%Kt32E4>fEqDn6d8-r=Is8UYou+HPi8FAiZUgd z1u264Gb6|r00Ammp9>!zafA1t{=3P(bPz*TQ?k1xu|gHb9z95YeqWCsq&^%87OBeC zo9rty7C9xQKIt;T3RP{kQ$<5d%UvA10yvG^*-GJtR~3>hMi?3;%VSqaY^xjEzGIgq zWV{n-aqJ2RI4O1o#&hkYFm{DT-gQ4qn!VXQD>@!SSO(lk* zei>FYt^@ZoVtm6eEVLZEhPcr*!=^PJCoQw?X2un4LD1<=OThj#4qf?W*j5$53UHir z^Og#Sro0Rb<_BMfU9UVaTD=V0reB5ygK~2@-{SoZcnqynKk(p>pry=`YRc?2jGq-mXIg|TnLC!2v zh86-`IixoT@5f^BYtz&!^2$jx-YKXaC+N{%If;I#ILYV9t`XE!50{;KG(wNR>Qe1Q z6eD7$*Vhu;6cd+;t%3Qm9?BOv*b=r6PipLQH@KGIj~#zE*`vTtRvvieCaenBcrSa` z_BojOlXsuEaee;S{NTpy_Q7Xw9(_bpF`r>i&DRJOGhpOkDlo^x0F4>@xmoPx5Bv); z{8x$SJNqpZzYhW9V8&Mt*h_U*QbHMfvWr-A?Xo2>amWWLs_7rO%Du_HJY)1Q1kBb# z-UGNpwb^?>y2~8C33m|Mt?MTH;z0%k45W3j_9B%nDjB$cQ45fWYI@J+=iX%88?TG@ zutUo&(k{l4GRNI&Zd(Xj;$i~WQ!Mm=`z$mt9PK+sZAwE=NLUJzUdf1wIC^2|2?#eK z^aMt8?V>33gvL$2A5u;K`v^TYa9(YyHRFu%&`yO)Y3T8w@)GB`DGEIS)47V!6U!7( zTPh4a9y(s6-Kz~f8-SCdOl5kdJ%?;Ku)#T>jm=bU1XH;@q|@{Q!~tYuKPf3Fuy1Ux zaU;0K866SVE20F7OWWB6_KkjcxIud>n&S!31vFqxawYn3fizQ zMW17qocw*fLCwj}*RFE{^z+)9*KC+~h&9J213@J$`#xa*W)`Cm$45^=vJtwnAd&9S z=qK*78?*VfqsO*TZEb;P?6;z&);U7Sd^qy@bIwR!E*xQPXN|ebgMH++iaL=yo15F7 zZN|~_#wo{Prydu_!e6;rSlnXy0RSMOglPHee$rZJC%Kt5`>6vq~B)NWOWo1cHU{j@ovOs>{n(i(q{*e&w*XDRa#jcNx@5^N&v)geTljUi*T4LMimQ? ztR7WDqFQLHs1g{-wS-tLM05{@#!RBC6~YVpAneYavx2B%L#vj# zBsXUO^Q_?|ecG{CNh9Sgh$xcXm&Uw$&gjXi1S)v0)WAHqwf| zAM0vP!>ypZg`dZZ#uX5Ah%=?s8_Av1-J(%&tJU2iEv3aOZavIZwYcbLd{Y^WHLJKq zy9Ci$#T&&c7EL|2KInXVVS;c)EO7wXF}Er;hqV9Eu8Mc*QGX^c z+rY!f9r`e`JGTo3N^=J!KiEMXlA${`)Y|iPc#C~SRB2U#@s0P%cg9~PAQAfI? z>9fafN)Pk*bs+h^l6`IAeket`U|-vyO#sB7ybJ*h#wB4N(zkm4zV=|6Jd509?(2Y+ ze9i6aP5!?2keN&^R@~Pcd|&Gd))pm%(E%BrJjmp7v;ho32rKc9==`;!0Oo+Qxje)r z`q+nl>sOSNZnGdMmtb?{ATgl277I?grD(?$cjw+cok!~~k-XScgD~hwdQm$-(6S)r zJTQjoMHQuWz+By$k zQ4jp~H2&YA;k0!UU3OcD0*nhaWcRf7fi!@ZMFaRsck?P=6VXv(ScB-9we zdU{F_yWher-A}O$^-+{IMV|GL`H^{kgY+P;71GdIVNk-0+P0r!>-mtj?cglgZ_HSP zsfr0_0R?n<0;xwGF^vfF!6;NsRQ)&HxsQe1HcomY8;D&{*L50v0E7ip zQUgA87WmMk1$BpJQ8f4f3hJz5GacEo-Y@B$g1SzD57dIXL;IoH-UT0eRZ!OhnhmNe zT~OCS@&X`5P?i3M0w1_B2=D>FcXk}21wa7y0y_1dg~_^Uk@bEQfUrNTK!_f9Vje%= zq?cp;PtNC$mViyK3M4o5TbW7yW>uXWRhB6%<0557OHHcmacQBT05{KeY zqFEjPtJw`2zRKJ|AkDlwcaVcpl(_>Nrf~ONWkOP=b>0-Z0}rSuFxpLtI|$g2AEK-8 z_!9V{&>eVSMTur{52^3S>bOJD1E>QWiNd5qU>f{ELCKL2SN$-<7aWOVA4}5;NH%K| zB|4iZ(d7~)sQHY>E9`2OP7@_M>Q%PJEBGBV0z}dA3ca%8G8?R{7{6p?i`f$VxLyEQ zB$6$)vy2V(wb>E^gMxaq*%H|nVi%|Mnk~U|E+q`QZ9oU0-UckF*og)#DwU+D2UmNi zB&An?g;kP*9&;?V#NYuIsv4!IYFz^seS|GRg%93TfCUFW1YohuI(9I zS-oi)gYz_;G(bTXK+NZKuGg_J$(}LO9$VvoIX8-Jf?fg>Clx07K>5C>WKXSgkQZeLP^WEelWC3-B(q1w?!c z0G3Xn4GAg_(`{JoW=%f6}2&i`h_s=$QU|21QQZ3TYA z0&D2A45Sj-lcQ88{YCI_4d(B{Gq|*i-1R%kmnsatplj}V5`+J+|7FJhrthtw1{_2N z*MM>L`I2EWBsWJOg;TIgO_R3}mqf>q)aJ00@(LzjxgSVsb9iYuY8o4Sa|G7U_mReK zvo}uBl}T!Ic$qh@f%45kZOp+B%?LGtUL(VgkNnWcAD$AY=I!xO{m64Og1<15qMKUh zlmkksC@6;!2U0L^F@9D#3^)~_5-57fV3))6`uQM!X4T0BI+D=^N|T_|OY|Cwxj|hB zngqIr#Y7T*TU?M>Q*~`Lg!Th2sC*ghe!zq12W%=|sPogdU{}5jOI7(YPGl(-xN6(LCOeumj5fvwoJc<|04Vil_r9ze}_Jse>3~*^KU}V=8=>H zliiQB*Tm7@oxe&7`9AZ`%qHW!nzP6M&RGt- zDuk1NaCZOIAG*A_b$ajg{Ny~qZP-3Cb8>oda(VXf;`K-B;2XYg6xWU0`^NWallmL% zugw=*lL~$jUVG!WjQ`V&eK0sgVA7Pk#iA+3ZxsJ*#=f06J$a8Z5f}5n#J=wKMl#KVyrqluimBma=5>scEv97Yd|n?ge^f zFJxujWnL&W>6%{1nqs~e3QQ04LV;g*mmutgLVvhpE~L!Cm{_@++OU`xO5!2ZO&P#J zf_??Ywb@Zhc=483&9xuOl7fFzaaFii`2X>*!(|New(;1_|H{@dA7!N8iTM z!9}pIAv(tKx3$%%i{KrSH;5GJFa|{_a@jO#KJl8hX0UDwt7fItC zpm2tMeeU5PnmaQ%g)?h~cZkB7)U32hIe9~0jIWN5JgkFsMSq^DP7{#$SAcb1>Z!1=PMyYJyKzx)I4mg6dE=@57iJ$7ocS4-FEa zhBN^{f=R0Jdv>UgSe7m76B1OmdYF^k1vj34^ zG|QETcdOA0{cltcO26msll(#H$3zY-F+8DlxedGMS&=|nb_StsfApLjGC^E}#h{UX}Ppf1WM^5D;ZS(mZEIxaf)2k-xOyz)hQe zc!RPZpRxA_gmM!8fKW%o&tg(3{~fl@KxU#+dTh|CmTM=`Ve3q0iX`O((!^rB4qK;} z8Tt#k30P-9^I`MZugz}4*6AgJ0ikrDrkUvm6U1Zd&+#4kjtV+{ZsX-` z(cDH|&5I4ceJ*870_U@n`JDI;&8%OBsoc^_<`>t0#Wwu_39Y(p1>5w z6ReB3K6Aou_2H!kLned6^`zOnea|9Yp`ljVACeh_;`{|h%a_`?eR(7+#B_(KPO z=;03o{9%MYz!@yz43=;POE`lioWT;#U{8Sq|F z;Md>`c#k0SFT80l_yNv2t z;SAPr25UG2{$P}u+ZxVb4QH^1Gg!kJK%dw^pV&a3*c^iyG{;~D%`uoka|~wC9D^A& z$6yA{F_=Mf40O;ypV&a3*g&7yK%dw^pV&a3*g&7yK%dw^pV&a3*g&7yK%dw^pV&a3 z*g&7yK%dw^pV&a3*g&7yK%dw^pV&a3*g&7yK%dw^pV&a3*g&7yK%dw^pV&a3*g&7y zK%dw^pV&a3*g&7yw9pQqPi&x1Y@knUpigX|Pi&x1Y@knUpigW%Xo=7#HXXD?=o1_0 z6PpfNBJ_z3^odOm_X6~ZO%G=PePYwY89<-dK%dw^pV&a3*g&7y3~&a}CpOS0HUnG* z=o6a(?ndYn8|V`o=o1_06C3Cgn-Tg0=o1_06C3Cg8|V`o=o1_06C3Cg8|V`o=o1_0 z6C3Cg8|V`o=o1_06C3Cg8|V{T=o4G$6IMP-9P=zgR@V* zF?W7meBG-LUjNMLgOiKPvo|0?_3H=L~nx^6wrnNIJqwRW;Z^|ttW z$pSDRwg!x8lrt`SV`21W(?5;>Nq@ z*X|zfehJRx-O-u6!sM^d|I(*_;Pk=iqffteet!Dsodx{V;&CzLt(VVEAI1h!t2CQ` z`MvBPe*^m;X6*LK$^C~vc=qV*;SZdhT>jw6z0=Fnle4$qID2&e{OsW+x4rYzH%?9; z-aGls>E#dn(8<#ePR~Dc@ATOd{&Edtho_HTy*N3(`R>IN?8g0j{15J5zH{^X;z`ET zPbW*~=autwy<`t+&$z*eK5ZU zb@}+wO^*wG$R4}Re{Jz!u|;ffzRK^~)sgOJ_|<-gZ`ANn;W^poU7OPfZ=KD+Ej@l<)sf^SS@OEu^_EzQzByopIdWal39GFaFEeubeC?zqF$KQbYNr zR{s)zGmt%e>-^;64Pn~zgPYfmUSW)Hnv1iO*UsL#nBQgpA9L>lWY=*Wh~d2%4uHmo zzynB)uP6zIEs=f_HN*@onnRHUhkuD6DPl-fl9i_(F%QIugPGCH0|`i~ETZ4~BdNSj zQmZ6#wkq-_+N>kT@!D&zWm&Ra(%Pc3F3X=dwiIbyu@%d4#UCYQSGknaefxh;cb~qu z2ePV07=Z6}_xbku_nE=|wT-RIla;mAodK(Vy)@c+VQ}Nx&I;7Pt%GO^Ss3hJA3Zl- z7oR(@Gk%`SD*wyy_6FNSi^bv%e0V}n{jq0pkNkBwUa2mL$t4-d0_%iEVX&W|R5&D-|Go(bylFFv?4;a{ApKHOvb zssqzLz*8m7JJxQY#v)|-d!xf0+qea z0=CYASSRpx4Y>o{X>6U-Vx7nwC*71Zb@IDy8)9(GU`;Dj=OgphM~wD%KC)na#OQA4 zBd4v87|nh1RmQp*991XL=YVn6hro*BdRwZ@Y1RiF`?`I51}E7Gc6^l zur^*@*%)1d3wR)}GxpVK_Jizp+YoyYtgxN&WaYy8Xd(fJ@%hc|Rk=|0Hg@Fv=B2IeF^ovw+z`qhz0A33>KIZ%qqKUG9eD2Y z1uz{~cGliKR;$63irT(7u(V(J8dKM<9dRwQ02@_1%s=+P{|9XWe3uW6CPsDKI+gy{ zkUpxQ{a(@2V%S*^e~TBv0$v}3U2)_Gz5f(iSxLoB?U-l&WEvY(20J;u?{!BG{_XH< z-~MOcn&+dl_FNqOyQBO6`SzQC?be_C{rlhY!{3DGSAP56!N0nfEi&F*FqmPzD<^x) z{LicKX9a)08~!}`N(TQp%}s^HBkb_}!NbRT{ppz-ruOd}>}79f)9~FV28Y><5QZK2 z#<2H9wU15hT)xD=evtK^7)-5gOm6Gl(F3{Di`~qQ&kuV2sp%WQmOpUt(BY$lV>iFz zrfa?4kM(-&A{=Uc9ju8P#JaO=@LlZa_IUO3`3Yy{wT;!a^W&Y9)5~kC&@sYhY9nx# zoxF+de_=d^lJK=lP!j$eJQ`Plop5rNiI3KRRA3M+f@?6}pV|+@;CU(jczXh4r}m%U zgoXJIu;G3-%{`-e0OA-wHd|lYncT(^@qGf8<7_{`2tFAX1rA}=pPd3xWLLzBFqV99 zfK6?WF8~?!W4Y7Mp@*Xl(>(mVMIcEBFNq(JO_ z40c{Z9B%N^Km}}G-abE`Jr0DDhqz&tLvr#EtAD8x!%zPLJG8{zzO$nV^wB&EqscwS zIH0i2$@>a@=*5GK&F|yf!CB)Z_zR=ecdlH~C*Py~>8klB_Mfu7OKYPY&VLqqIO=&* zFvpa$+X=JVP3+kG(FXgRyqz83L|eJAxqWFg8O-cVw!xpQX26qw^z9EU9%cvY-{Air zc=YYC{{gxGBip0Z!BKV)$O3Nf`4_gfH@9|w|EJkaOWT{H)$?E-^AoIry>UPI*C1N` zojYKE@yjWG5crpE{x9Me{x87lN|+pmeu4?>7LW)6EiC5>e!Y59*uo%>wJ|Lun*U8$ z$o1cjTgJ8Gt34o!`@la2iyL?WqhJA;ABF`X?+L&eZUT{ zf+T<=dT9g*9dFMvu5RnQarP#Tt_y?d3*eD~(y}*jwB^ZH3&G;nQtfj({}t|q6YP-G zhtirlAe626ORa6eq~fN`c>BdM9BzAcZ37e}SmuULe~?W-yasp10Q5GH;3W9ctK%!R z-D4@2AO&6x#;5s)kO)Vv{!Nb^ei)21z<+DJe@Ae~UwD3aeRFgB!L=6^HxsY23ZlLN0u!z9I@C~B+6546pZ4-aei*tj)hf@0mm%aeo?n{?lEW6*6YPew+%V#Xp5xHj8-savKv-kk4I5vXowAH3*P0TT0p-#HGau>!)i+ni z>*2fBlG>a2g>ZiB^2&3USDznG%!Sa|@L=~aK5QdMSpPQQgq@zL?@jK=zL8BmF>cCA z$ZukB5Ai-`2Jz~dCpX5AZgRg=*|I15%a3uN82X(j7iXv7zw85NSkQtq`eC+z^TGxF zcUO%7(fMUsqvpWux7Gh_lNWfO@1Sg-{fm3q4f9jez^E_@{>CS_CTo|#Tntl$8={-_ zr^GF^16e}bTI@`Xlee+s_694Uk}s@*zjUy+7sYMxeHEl6U>M%?Q{cyOP8JAKp$VIl zU2e76_(mIBkhWnfZ0afS5Uu-oa1-0NHChEr^!c5W`@!s}e|52LS@^Gl1cdsKMzM)Q zya_hgd@0Ch4Tk9*6OJ~xu!o+8>u>YYGaGB%>*Yrjb~9Ibh$A(KfiOpf7kj?T1#Btj@S*bKd1}{CG1Q_+mkvd!~ePt>|*^Dgc&@r zI%ES>ktUt^I~Y}dd`2XaU|V3mGmWANC#V6b=TuPRUbYts2-aI0!0!PhkGH@qjPO3l z<@wFa8#6Aw?11xfksDIH)Sg&O;Ga%Nfl*C6Lrke#c)uO!>B-6q=)i-R>c z2KyXj6`ZUKYgcCX4vw&c^;RNu4U#-K&W?yxL}|dzI=9q8zK^oK_2+ixXI6!G2f;S^ zn;$dd{PITNmmQd`qBuQeKHLola0NUX+ZH5n`Z~lO*Xew7?H9l`+vd8;x5$ccr{JbgKTQNv8qUX1c)KN zxzmoy++q_v1IaF^cF@<)F}4qO@&0j^9U8B%uWju>O@#Y#_%9pdmsYrGj{|#A@$D~wI(CQOB@w3jAUoFbl#|Zu$(C-=kh|vTTXmSbYWb=Zu4 zRZv+8bitx!;y<~$#gj_A4r|J_`Y z`B!%sQ^mayzq_$cwAPLbBwPK>ynIm`aA4En%&9k?I{ilIW_hvx<@W9Kr*5C0zy0~q zh56H~=kGf8-1iSY{*BM}*gs~kvdXxd&q%4#32gG5^+l-FlX)*)#_d#UgiEjgn;x?Z z$pKy++#k{EwfD~@+-GXTNkJzj2LzKku2VfMm zy*96)n3K1#yp5~G=BfX#%4%4k$QWu{!7KT$#wm7$EwD1NihPwu{qr7s zea5mdY)e+S$ov1~p}E!Zix5m%eBh~Ppq|(gjbIVx@D0i)_@1Tj z7JWspeB<2Y($?H~3*3V_$Z4%@yb-#3wnZ^0dyyR%6{Wn0l-Ho{)UT}t4)7Ya_RUMj z;rT`W)Ix8OsZ_zYogT8-Yj(T{?y$QrA37|oj!G8!py%>2A}g%-4iD-WCKT^Q=FRrl zM{!>Y@vkcHi2mkK<10Nbad4<73z6 z3JoS!Sr6~U2%P#&yi(`X({q_s$3a$Jk>TzDC%sZ}(qQR`C<+d6QP_lqd&&#CyML{G zB-Y{gCf))>^xw^>iF=b*sZ^tx{UawH8-TfKt*_c59R~H8I=D+-mZ;I51v_8lc4~?9 zG1ELQaox`}L)czOyYfh72eT^?$gb+6`Dc1{B#Pa?E%-VmFg+?AxGX?M#j1WIWBmk; znCG!ly3(7hFt09;vH=_H`H$T`U=|eb|1Vv2n5asue4emyULnq=K)GOyhtBxw9(1L6hxkO@O2?lhmcR0g@q~DZ=!Er&;R9U9-uv{MyR_J{ ze+GKb&T;F8Vi0GMu^OCRdMjM~;vJ-=<8H6+h0DCzW9i7oCKPRsc>;DrWa5{;OQnV& zR}8iR&t(Z#H<6_Z-~oids(|6Grh%>- z#-C49)l@v|YknX4p_$)+dXrrnv!>=~~J%qa=0T)f)=y-L6BaEu)t z+$i35Fz~YFU@>8!$gr{(48P z1?ZyPg*w&f(Pgs9pnpd%qFzow<=(5GgDTd!@ula+buG=Dsi>dh#qqLHznquypTpzj z;XMadMSsi@yJVPAkw9{*G?L1*8Jg&7kv z-v*zr?uZ6~6WI^P#T!C->B{-_5fqEw#$Gr0ULgUA)rYTASIzOIf%F6WC-?o()6hpY zwa2j%W4>6Q{o5Y9Ke3k+9B)8!i5EvovbO=G`TP#=g|n}`piRj>4YkId4!<*51^2$} z@!9Wx;?1<<|I0pmywmahoSL6xs|qkF=X>s>h|vhTow^OBA$C$bJN)Le0x7-WcQ@qk zMs{fNz+igc5B-ZCH|=KFzwNU#op1$t(fU!gqA0IQ@E-wwrK@^vlm+`_pS`&Y^6D^K zXZ%;)HO;Q0>PwIgN}$Sl!t`d9sdM=?m%p+1OWVkpZymCa_1WtZ@c2)irr@tS@lqQG z4!rM`RvXP{wDc^49yhYTH-(yQNCaA`ofK+zojteMtNbI&psj0ClrzBoDWnerTbh7z z#6+v&2lm3oktq}>ln$JNL|6FU274BJ&c!no@}1a!=&@NJXtl#Y zcgs~I_NhYXYhp%iA%Ga3Xxxg>9KFc+eJFw)GhE4t{yRQDhlg0|hEdUQ$M0A^iU%4% zcp0$x1hgrr-aD}L*`%rIBrn07w7UF2;_}p25TYu_r$(c-#@H}LA@-SmGo*nk zrD!;y_7n=yrYfbgt_uQxx5r*f;z`GUNYMpz4BB&n?;e0*Eej0~h%AYeR@*sepX#$W zkl^zD2pTq;F^YVyi<$Yu_SixnwM|RX=Xz{NLzw4X&=7WAn0Y&F8_v77VL49&dKH3} z;k>a78?o)vVUtjvRiSLkGYMnf1{C|9@TtgWy+{zlUsvE^Z5QTJLZ!T$`xxvCdjX+ z!mPYtjzc@Z1@=9S1BiKdE6~xivohvmYUZve`8=k4akjJ%eDOVt{cB#)Us%%+_R;g{ zk$t9Vir^DaF`Nd-3^*~fPAb9yuTVauN8g9l)jQd(=|jR;HOQ&TD2Ov)d&(bfJ{2g7 zh2%Pe!BmaA9nyV}O0`MM$0_oeid7EwDk>DASeM9a!i+`{Ap-l*%x2{JVw1fwjiq`z zqdN)SG9iO?mteNZdKo_VTueD-;Nahpx1$*kA7K39MYYY!(mTX9ZwI>8QzTDZIq~Lu zPFy{)y0&^^V{>xi;^zA53Fsc@^Vn{GX>Dh`c;&=x>GND!yyN>$+;hT~kGDwyw3UCX^2v>ND1m58V<^68x3hi-;5$Rn zASOZMwF5}M4RAf^{nvdqmr3w}oV;4m)K}FG3wN~q-YLpeaHZwq3zWsLLW~_cm_?!E z-n~;fgI#K}KmO=cGGW#F; z&G07zl^-ppV4`Z2AYlu~3NNFP;*EQz6A!57bd1)sxxz@nTY+jJ@!gp ziHhMUU}eIjDd{Ctcl@CLA;k3+&M;>Q@QFi)3eS4O|7<5KUfT)|SfendKA`nDMzQl< zgR}8IO#1nLvBz%maZ?X@st0&gWbUY9gs37&Fnj_Uq@MJ}qoNVsmGzkaJyIv93NcPQ zk;E1V#$%5?Krlygj=2v0_i1^@2x_NH!hf~jn>V*OV3+QIjrE{MAp z^cBch&(WnD_w@$i@w1uXG^eL6F}>JJO*8CQdd(IDcO&Pv^e-nUl9Jcw-ukgMQ9~t$ zB0~HtzBiVM&Bp_Kf4>#zq^1`|+v~FLVEUlZw$xFdw8qb!H^FHp4!Nm_FXI zy<0DV1;hO<0Eiy-XQOiR{0N3pIw*!bkCqck39kWLtQ8g%Kcg5uH6)=Mg>v3bj-)K7`aRdocL2l7cxVEI28n!qNr{xV0F=Kl@imrLvC*Ph3)nF!Is{hUjVE z*JD56n>FUNmo~*)Cn5ke*LW`q2`$HLJl+~4ua!ef2lPA)7{C9Uy=L7zeu#mX?_MS7_Lk#QAK4eP5V&L?JtGm7D`9G?RmzC`TyU*Nh6k{OJ$Bp20QVNR=AcoegXA zyHTfMKCN_^;*5$8ie^3CL7pwQ>)n0UIWc? zAG?NR7pzU{yCr^%Dyl<6)jS#*3@Oc4Cx9^+aXpPbVzsS}@95MX z8s0$bpW8ma454nuhj0d^N7yzs@?z)>(~`rMi>pnQtlKybugY!13~rI=c5+G3UoXX} z(TS|GL-y4ebGoM(jITUiU)Ao8CMnctH94hnY3?`bGb(`nE|XX+1qKW8@|GTZFFWd- z#o`Y1kWp@+?8tq2$qZ94gI0D=NVs?my9R1yw6jvb9;K|KFbS@KS;omgC&52bvjyyZ zouX@k{tPK{_Fwiqq>9kIc!OA}QF~v(ZU$b}!C`+y98J-=$6M~9zg|6I%oC?E%yT;P zNm4Md`iMhot*KlV;HRzJe-d&_DgY zGjbJIJW8$L=QY@OGp>=8bv-SCL4(pc=Va?V+qrVb+-L$E&g(JfcE;-$=HLyd?W=Py zjJG$&>rQ3|E&=|^Iv<;dO!0a^=S1&fcMiVa%85!q2-}59AD})nj|zX(Q_bb{&Ev4z z1)D`c#U$I+9bp^gt39?{n3T>Wuw4R>Uf21DefEyB049q(cSAP#O^{#lITNY)EI}Fe z_kH%Zve?xse0Ql^wpgZr(<}N5bTzh^)^t-^Hs!V{6 zu&OLW6PN_3!Y3tzo$&%Za17nO#^!@D*Q!3NgDrv+~=@9U1t9JU(0PW0a7=b2lIo#p@?6FgH)w~3X<$f4+uvh+) zbLYt!7qS8+i%Vk4SB^E*A?lmh0nu@MfxFxA9{FJUiAUf500THlrrrzV7H)i74m6t; zbK*e`Woua+!)9woy)?oVwhQmrRH5WW0bJl-K9@t1X-x&q?Zn)bbJLyo-|e&igd&Gj z;F7L;d12^=S4RO9ryZ(4?z6WR z#Cz?mm!@P@Tvc9YEJJvjWHuRAajbJkp(klt(q1wwK;D<4fN9_Y0ADrwA;;AP~z*%H_ycCb(RpA8*?vJ4+-DS?gVyR^82(){{3uW zaEk1bbIkFT3pwO}*i)l5(*_>jHZ1Tlrw5bu&Im|*=IodHsy%=PoiE~Y>L~5YRY%T< z%5-OJ6ifP|oQ1t%X9%MVs2n=>dl`2e(fKC%pq|DvdaIP{4WTt?ndMJo zrP)>b+s5%GYhCGf=ktXD7^Xfukp}yRlRdW&PR^d&Uy8GjcEm~9 zbNad-)o9ss`trUrUWD({WKo^Q-iHjoAx%b>rMu=w8=uv zp4;m?c6!0=xxEL~*UpUOF>3bQK8tU;lk&Uv0<`S8{f($XDZ?bV(Vmpgm?uHbo;#p= zPI@+WjrDh(G!!*^UPBpo)uWC@+wiB^nC!_jYWCc$?WlnQM@{?OuCPZ@*Hla7?74x^ z7krW@W4=zyo}14*LN5Y3F=1;RiZdGqLUA#~lC$S#7vFWy#G(=J0z4K4jha2LK_fcw zr0jXUqO8g0`p^Jf%7v$9&+DT!ocIi_EZulgl&)WcpM&so27YW6HO0WwmecOzaT9qM(?UQ3p|0eY zq^d|?aiBKXc6$wOZD8Rhc5Lyec}lRq?;!l;4Dallr-X&{+a7aU<8A$lx|Pz2buxu8 z56Bk7oq04v9gcm4Cka}fiLL1 zVeonoNuYE)irZLLnz13qLc+B^j$!TGKw{Mf?qLke62pRr4P6`=U?0(q?s1mTtPPKf z+RTk>3!i5lZ2`90;T0HxuyuK|A_62k1Fe17*N72V>pV7l=Mk$qoh3&Sbc4zm^203t z4c+n{N4dOs4=s%*=P%ZOd3L@cL8!4)k+@KQeB~6U)M)jcEAWsA)UV0ETy@KmZ)?!+ zW_EmWAoQjNTeL&!=|BlcFRyo=4g@TBFZ%^O)g-xL-4`oo4_qt)EUh8WBlB$=yb`vEh*S4SKM$#CV1{vInTe}CF~Ew0&wDj>h=<5*&S--`Y#oiu`~Chf_gRk6iL(T3SdTd%F_1W3 zl}U^Y)M<2W(M&cVCg?r=o|I+G1pE3h?47yg^L(fZ2?b)yT0W@=rr9+T7R@qrw6UKf z5*r2()NxqS$Nia2@VxYDHJidRF(ew=8TpWD1Jf#EKLz48||5-HBw0BxH?T~?1n z#aCD36H1LrqPk3K-*fi-B;3$B0oM+9iLS{Xx}o!aVik-!j5~kB9<0<>B;m-c>nX}i zTLPb5K5tEB2>JvjlDnvoEGW)tXu4hFmy#`W*f1uWAA0Ohy3t%;B%~r41(VM>Fu|$Q z_hHUZ=+@MC%lrDY60gu1)y$V-T+s<+z`Itcbalu6rpFc?yC^U@07pW7i*%wCTrQoy z50s{Jzoz|*w78UlTG%_vj%}^VVA2$rV^M< zrS)Jbe7O8UmN@psOoj&=vNozop;~1JlA=-y6v>r_+v(82OKw?GWldIUO>t{@5{ph^ zqJsg&%~|?H81^+;vZMjNm&ax9EH64hxdGXEHy|4}o4Lu1W)o8JK~e-v1!S!k=Sg&N z(n5J5SqZuKIDITehz4j|gk=lRLe!0a9k1{n)2SZu*zFKSGhaP*jI7yuWak~a2`~6S6nFi$7 z`)s8&kY2&`;70Ess@;(k7vOY-B^od*L=sTGI)=wSIF%e3nJ+8Ko7!#aRl6hu=0+=o z@+83-JYD}}VP3DpWakR6D$3q>MZFHM7qT^Z*^bb>1mP;DE+edY&2B` zIHpDIdQ@eUKsawbO@;iM*v#yIX)F+*gs1XPZJLG}Td{9t-kGviZ>Ii}R-~t0{i>}J z<%BjKcl(iZ##Z=dI!xgpt5H9nrOYL^+OZN;8DV4PVpoF~L0~T>Ywi@Vg&*xAM1#DY zom!Mq(V!2e$D?}3@QVq`Jnz;09A)ZIzTRVRA>Q$tA?TGTsZ5%O>be#M^^u($*Yq?YN4N=#k zP;XqVu_)Cj?FAQ#P%I76XX&1R z+{CI(N!1W}*tV%ou1!gU@G{+~mmpYGDVg%X>ryf`l3SUQ2IrIMHUf!PU5roVEgSqQ zFef~RdB9J8m}Vt;uPFyd9zLS8koT~=2X|#;A>AFw@u8^K&A?ww_0UOLxZ9TtAQeLh zA&3M+Rqk#@PA{(93T0IxPKri{ieVEEDM!Z={8^70vv+yPk0-anZj{Z9$d|$ouXxF( z=kudWmqtzoU%Z4UpSyOP!Wg25k7^xqFMIP~v6T)HX9+72sKhKe}x6P%^OYjEf&K$4g0lm0-`8p#A~b2bz_1I&r zkN>XABjB(H*DhQbZ^OLY@!;q=|DSRwTm+9Av~+-seLVemNyBzkcz$uTy*2`vcTvVZ zMUGq=b>|;X%2_!YC!fJ+fw45dLJ)j8DXhH2_P$(ZevZ%9;LiZhfv;$~+?&Cxmp}dC z*Sp3*DA)t#HurQoMpnEvO#7-UGK?&x6bk#9PN;u+=Uv;A!42Y%svGVy&J2l=4twda zRR=8B0C%uy47BN}2|QBJ;iy(XC^f>ty0p9DHwU`E=nUO_-GtTY=Iz3~fZ-f~A@I@x zyE=zQERr)?=DG1?BpmK8>V`2E_ovSXZ9Y{JPP>Vr0GLfvfVVl>;Ds9) zg3sdxvFb#38vPmWeExs|GTarQ8x0LC_&i=q;JhA%X467>Y2aQU?-CgFXLT3dR0;e! zH5pD|E^<+e!qwUE^LQ=ssx1}#dAydZdejQ|1uYks6Zw<4GFM1&J=Eh{SwlUrw^ap!*QOAy$^khF* zKkI7vG%Gd()#2yyg0T9bV0;W*@auu+@q(X^-Zp7sMnlB~D22v|!_VV|!Sv$4nOMEf z5&}j89_&Qm!|;8|dOw4#e@-hl;#&sPd;)&yoS{qWlQdLXz>Ud5+( zg@;n<>(#i`)pniv<_Keuff!#|nOxl(i;#OH@iEJS?>%*vzyPqor<)9#M9<}&@d_l6 zUs&6C-c6l!#S`B55`IrH1d^xwv#@70e`jc-4FB!W^j!ofuhXP!SR^o|0omAowZZ8- zc@7A#-KQRxncSX~+3)}mbRJjL@>_WR`6Q2FWAaKDGH$ zEENF{Fo-!ers*ya31Gh*8vcrJI+%i;1A}RhZ>Snfd>#|>1p!ygtQ3{&vl`hsI0G*? zR(pKx);sTUE&347U`R1O9}NLWyVhhBh+_tZ6(bt}#xyq4h-1H$adurBi4WdE4x{Lu zgP!!EHGFImm;r$!O z`q)Q9qpJ|KEsr)X4yO742Ramy?}+o~94Lc~A}}8#BMeq1lJPScWE9)h@fk81mXgRwZq2ek=mi3g zGLzbJURC%_i2m?yT>*NE%?)l3SAZf%2qk{_EJuG98H3gXyl)O!*zrCZxeq(76S%v( zfz`Z38i^hT(>)(Yb5+^kxT%~@^k^Quk*o=JTK!r-Ff}b3ATJ@%OyA(a_3V8_sMfvO zBf;5O?(xsyd7e&4w=+EKEB!zjT{d)}V{85NOc^%ws8SW>UZbWvp)yH6KD^_hBKW8; zeY_m54Xvtz%I&&g^D8}e8x@;QXbp*mbs~HtROe@6r1Vk~sP)HrfDJe6FbaE}Ib_6C zWf(G=Lg|EDG(oyQz~3LyHoFW^8;qxC4!d^8{P&?$XB6 z9J_t+dZMw!*9`Z^_$h`+cyNY6Q|;^00?Ri{iE9T?8p7(w)o0vGISSDy`V=M;Wbp)) zh={5+)$zD{_R_S_G;EsZDfSM85g70=y>HIl>Z%=l$?F*k1WdMaD}YbmXiqL`6iZTL!3S zG%v7A36jYs=~Az9y*q(Dxt$n?qXK@V1|YqMy(y|#HK@*%pG#$?9<|)O%%~$^T6`Xh zVk;c2v>`^v>SOy<{-D+*+Kz62XE6_4)oPQ@YT*G4;cP3Nq#kIsvE3iP~wSFWpW9TJP z4=O;GiQ+oUA7Bc!S_N9Ie=_tqWpl&WOQg+DL*zDoO=&FwLxn zLBtP}S{eGn@OMzXv*@rF-H=LHMJ`zl*3)PHsg%hom$t_I;Dcd37;Sn*ADc=Zk>8%@ zw$Cp^B^6_zN?D)sxIB9!bC*<-(0C$ud=DgCUlTLBAOln@_&6`vV0CHhsS{3VoGY&9 zC5(8Jn54JYD-{eYI0`@JEf4R}0f+^5YA}}!K)5G}1tA!`2Q!Qj!(GY=FLdz#vAVjB zK>Fbc_Ot!K7(Sd#UYH>RDS>8ve47S0lnys^LnF`EMPsKfKlqb(GO%Vea<>!5!w(S5 zB+4p23`O~y_i70|&E^MpWJq9gPEJPD3=+k7?M;K%NON?$0_BTbbLx5qY(L)*%woo2 z^NV70@In<>oF$^czOE?6gZ8n@fd_}9=7*8k8XhlFn1eVYvt8kgLIf2}cmaEZ71X$k zU7mLHjf9tKP;4Ja&fZAbI)P&SOjt|J(D-C%#!~P+_S_n8icQoF!I5TIEID)Hk3v1+ z;iLNFX|RbkNqXV)8p4}eQ+^C|Nqdw1c#l0wv&2rJ6)lhmz8Om#c@7 z=TodIh7&`J6sK`_D>!~8_-=Otrd=2{^X1TB*ZAgy_*s~c6+S&JzT@20w1UzO)=z{6 z2S)aAM#$jPUL)TIA28#55NgieK?gXpu~!d?$-d$Zd#6YZx73`T$ zrOfmNG_#*W?&wobWl#0C1H;4b;mYLV#W;8C>T!ez{!0isf(oxittqclUKmj-!B3w2t$)G_hgE0WnXh2!Scp;V-;WHjFcQ$eYKed>NMRO$<;ISCk>~FK~La)*p zW34;pB4GsK6Qllm0GGR3`3xgUu~`~{qsYQ%Qvsc!wB_WEHcSLXb6www4UA%K!1=&bG`dr0gmq52S%~B1}^E@ll&sXpw1ly zm6jXlVRSLf)PfrF(Rw|-%r!dFRw?S&7a)V<)qktco}_q{R7kZ?kBwq2MJ<|#Njc?reyV0QPyz$_yUZa6d@KE#Tvn~W4t5Xzy*%ou;+1? zJX2kSMzKa<_5F8}0cyA$9mN{k&UW^E6`2;9nDt#kX3@f=&?wdzn4aIBiB$q33Ti6e zRD=!qMB$5j?8Nzj!Ica7>w(Fd!EGIZcY<~?p9qWBbw)r#F*F?g&}7XZbiRmVvz>xp zMki|qKkZmRwU^!*BdAP9>tSPBA~abu43a0xS=clFd1o}lIFWse7@*>cyNqv6a#77a zA>tgIE~cw^B<5*G?(e#}#{H5zgflcx6D4NDFDENy>6MdW032kYBRSg?@X4vU_ZLU& zm&eyH_ik@j%rB?r-tEmDjpG{>;(bUJO>P^+w|7H36a66eg=L-tDt;>z}{dGS44wA?M!h72@op{d7|9oxZL|HCpbS zzP#_N7vcNJJtVWE)Z9B6kc`|rc}?+n3aNl8V@bznzfRbu=H6{E95X&iP|SQU0$js{ zC;0Z{+`A30Z`hL(4fX=G+`Ijas6vU^-K`VR_U?gvU6Rz~+`GM*pPI|YZjKjUg7LU( zgN?&zSi?1xaZjG^1-FfcnN!5Pes1EE40W@dqXr6w@5~O($CVmhx*NHRI&ak6yV>6t z{7eMVa_`o5cF?X{?%gaJkmVSrV)})gdp9TW1w0dr#8unYuFn;eumP{$-R3_WLTkX5}y4CvBu!(vG7sqY@%{)U#lLhx@D=74^~Z| zJaPic!mq&(pIChcer%N#d%kL#7r|o2*CZGKK%6Hhc_fRgkX@e$A<8PPczuR3`1>pwl;Uh=yaIIHx+fT4xq+|U7!uF zW7y@0g#pwY!{W)7Wr45$9aHSX#EK|+M0o9@rl)-O49DjQz~MFc$!BZ`!H{sRk7HOH z>o#g2k4OI$W0NAJdLvmL5W}*>u;5`s7e@xzM|$iuC10zl3>deW8%H?cI?TY4cA9DS zHGI%U198@QZ1&D0R@Dl0gUT54!z}&{e6~yjM0qbCS{hBxU#$P~>^wTvv+-4p9rQhx z@W)q9aaA)~edo#*5p0)#xe9;j8L)3_=*^qi@x_7An;LA+d<-A>*4Y0nJW_O3sY|q) z>0Vm}h98#x1$?d+wqe~wf=y;=etI;46QW}ro1f*N49OVMNM-Lu=X^BQf9T2KfsS;kdd%SRaE!O9%Ahl)avf&?)vgqSIuyA29QKb;MC25GJfhSBF`F1Czj7 zuf8!jUy06h4nEu%27^RqbEXSo%tn1$=hH-H)BUH|;I0ON#qlbVYhlo*kwC{G+oG9l zKul2iG3h*W%ja!o3QH$EAx1)h*s_*SY$O|oGMZf@VWA^Q-Oz$Vw(5S<>V=3pXS<6o3jgB8U z@gCOo6lI1kgU^o5FKMt_02fckhA)#mOtr9}c{EKs@P z4ukGu$$o(tbWsq=&CO6~Qt;RNHq6k@rB@n@Q2m6xa1unLJ$%?37LCm_3$EiT*(#qu zT3Asq#4`(lTC3(5gy^i8AQ~iN&`m+Z|G!S`7gStijhSvuaeH8Nbo4Y``$1`RmOc@N zeSO*G5;Kp>#5ig)9EL8r0onQ84ah=#Fe}D4ZvwI;x;SYe8uP1=tb|;nZshp}Lx z9jR>*lAZqt7LuKB9g>~T4awG^N0F;Tl76y5Q;I?H5UEVNj>5vccrPeRVlAb)gvitG zJ{`w}sbVqy@j*l*xO)t|`BDFcU1c$-*&t~@gO3H9#jDl5W zovf~oUsV3)uAQG=1ON3tTPY2sS5MuXK({-RLa1Lv+tUkC0XS9{zuR%x9{b=_avo*A ztf&~(Zd0$?B^fZPwh;a!I$Bo5|8;t9KrS8XFj=arBX?C%6~8O$l^`QJ6i|jP>@U;R zO9+0PqS!VS>I6uUTH8V}|8bw))k3R!BdStKYy@r9lP$Kj%Vi;i;*0@_)vylBU@K5r zo0eyYwx#C(At>_8Xj@?UJD|IMug8{hNE@$~Z&AMQV3MyMtAcm#zi96>Z*@q8D;rxo z$^nt$BH%TYPM4bbFQ!LTHlXyY=CjdM72uet{1?GQKJkV{ake@tPr=7to`XD5?5x57l)oigX61TtqsqZp2QL6EQ5WoBerLn(Hc* zRO_GzkIy<#hthEcs80xRl)ft1wSKb2F>y}i=NIIRj)3!elr+(m8_E6(f9w$pRH9s6 zyVxCR1+^$0R{&~f0DUDnSl1)FQ-^{nGiq7zuckA=&8_Q7luQj#*P>8wTuoc3Mp3*G zk{_r>r`jZPS(*7a@3wbA;g{&kH5N^%7uD6=&BSd^N%-n#=2*U>V>#$lA}mIOZNoiCRSxis)oqJ zg-vyGZAzLb@OUv5xe^4cDkYzsxjdHpP&!nrrP>H2UUe~k8}L!LBt3*Yhqu6uewb!& zns*1E8&5{gPz}*vuM%31=p^Mm?C!x`8A(caheEL>a3@nOc9K-?_Nn+%F))k9m|&dM z-L1&!g`HcWtSZDw(Rfoaytjxs{aKG1(sy~u4>z~M?ywl-v9Ob|7@!~~o1XZ)+)6V% zsx`yC?9GG4R+>TV5LOMog&YHwms}AqZ$nP8zoQ$RDuvj7oUL~*j<)%D_0gsApg-6% zEQgY`8ew1UGot0@!Rtq%+hHwVJ>cvVRS)G5+Co17LlMm_AHiN=x*dX}Nfs^G<#>*L ztjC_ny=~N!*zt$qEuo6_umxL-*@s&2x_vbHC^p^7s|sTx{pvA<)PiqcYijeJln2|= zDDr@Owj!~St525%lKoweJ=Xg8{uAZcqw*OPE`mn`d)jOIc=}=2hV82GtX_8q+hg_e zdH&$-%G$>2+WGO0x{fJb`H?{xJ@zSbVAH5O|9Dc)%F#Ia#zqT_iDv(`RTxMC&HixF zzAzqdtw8htjyseh^;$j8x!jk_%z?M*Ed?3uc1L>FUjFolU(c&??&)-ltZ;0Y_ElG8 z7+Ff9{%1O&-g(2m8}2gB40)0cd+DfF2Q1ew+;^~PazYXC0z!wQTJfR`mEH}%hL&kv z{uiC0o3DGcI^DcgM}U)XYLqYtymZ)1r}&69bVkdF=oGi7yQmk=P4hX-3p%v}^NP0v zYpvR06(Is$D4dzqgP93$L+_EbjR_2&gXUmpJDzLV>p&h*8Qd(^=KupU>~wa_$LRAh zollj7)9!I704A2^a={B)2z-#=(9GxA%f9<3zRnGT6CfPeXZW z;7p2l6vUL1sCA#khlq4XHw(VK*AlPVQo-NeYq_dNt$<(9a&bA&KXM&P8tBAj0lY&+ z`CAJLYkl9~#o{f|0+bBAz1Ma@XO4e&9Hylv{PtcD1ScF^;Eat>E@*fN9B+*W2NxM1 zK__|3QLkxMY=)o1Z|?bt zyTU`M^z{Ua4cm7{<`KpqIy%0xGPw$BmWTNwnY>vRe5^45fpbH2K5XND@VinRcJD<& zL6VZ;#~K6B`F8?73nB33zA-DByy8k(~sb4L{Zx0?Et$WnnK)TWx%HWjXfk z&~RP^D6bKvLqJ1>MFK+{kd5s}9W>}I`cuTEZkk%e`;Fme8&Wwoh%oWajaPZaD0MIwNN7uSzZ3W4|1l4U2C&n1Y=Hb8C=qNIC?1 zp^z^KxHLv$&Vp!bn9ac%_(rkPCTD`)$}r$VT_MYLjcmQH9m3${|1c3BX0m0 z%fZ(#Wt?4?PU2%xB<&aU_Kru2l zhJXge_A?n|6x-JE88R8=Q7f|jz_FahmPTw2+Biw(<7nGMFg%( zoP)Pq2o^FSP+BibNu-lomSyk5=g@!YNJK~J%{!R_ICP-L4>a)(bshxdwX z%^FYcKxYEpR)=ircpr`QfSuOq1X}a5p!c!D@XyE57;!c@ZYrlRK=k2M@Y3KjWdQOL z0?mKEoMbP7thsnLb{zfVqfV8%ILD8107rIr`@4SRg`;&oiG@Q zgi5L@FO$J+kB5qIX?^MA<#26iRTWfj*A1Ir>9O0W*mOc`)QpHEUo#`p4a7!Z4m7!w zxT*K$0XE#M!zk=^=8zFnm0`$k7{TLbX~6=@T;3V4fJyMe+Q#!vnI(gvE%dq$Vja(9728&c{*=r&)5=k%&p-?QIc z3gmbAZ0uHRD+jVObz6a!S>yaWq1hWzF08${me^?p7D&8NtP!hIYS7R_r5;p(taudH z@3#)|%1?&gu54~t&mkg7syJTan$lW=@&Z;U1?rs=LVx96zWV8u4J9Fvo0|DTxDgUM zYTz|Z@8euLR8S+B$@q=XxK?j5dzOs%vK2lt%KN@Mt}TPk8iKSSKwlXCOsaPl9rlVd zrZX7`PeXbOOR^Gf6$KUhsg%hoImVg-gikv4j$+d*`W96Zt^5W&w|#yYDybOzRLc64 z$K|%N%w1ARLgR_t={}HfeND{hf(%ft;N!etgVm)MsZKa4>ZcinmoVZ@Vv^oouT(Is zaE}fvEU;69xnx+uUAItZ0iKH)#)#o=>Vy}N@LordN(L<0&-Md@%5XAyVTKH(RGqlN z+y@t1BX1v@_Vp zW)`Oh{e8g!H+FntihU#Dr5Y652a>ZlQnpT@SU(e1ZW5-))uac)d@?j+DR`d96q~3U zf+NkaSaRmXABB3t!$+k|tVz-fpVtuHbh5ei$o6P;ZM-ot&?Rkr_TxSFD9ticOm`h- z()?II9+YYlRUJx7175BkMxIZxsu)g^vOp85;IrNhn08^*%$GxRU*nq-;%8w(qEk~( z&Bb?|yJ}cJDcC;|8XOqe!xVM!;7(wxEILImK+OL3FW>?&f-M`6A7! zKd(=l28@BG%4RVUnW=SE_ICw)=2M9WXp{1y;pWZk=a4&k-g=UDXS{x4?p>G1+gImg z650+3L3q-63(kELnD`1(bh}U`FI4g7Rq}$lh0QXFZceZ0EdpwnWGqbT#-4-h0G?an z>qHO&d7R-cVL^;*f!WpXD9fwMig};l6$b0gir?LkjWRfLGR_WZBGO;Kaf+v(|Gtk7 zYwEy0puy@&;N7KKwJTg7Q=$@$I^te}>Df2?>}*l)uJ0XeclUKmj-!B3w4n!Jt7ZgEeclz~!!1KEsGoY?emgD6;U`R6u7a zZ8^E44HLC4gw{O70;5`8u+p`3F^;W4igtX|x!4~$}MjaTi{W20DGuj)~Y z=3$bFGg)07zo=DV2{I*J4i*^2+7?5#FaIPZK0rkaP!b)*+V+TZTvc9YEJO5~WK|iQ zIL3Q-AHINgY?%K65B|$UGv2`J`|p{kLPxQ-7A zdYvT%GQHSPmjG}2=hWOg9YD|L zYngke?Q*FE-a7AMpOqg(zuUs7xp#YoIQwWnos@g0uj^5bmV2i!Uvs^= zT8r>^+FPgQ-pPPu4p{<*B)MvvEod6acNE1Uk6*I>i(m-DC#OwGMpV0i+biN$D`!N`7hY%pr>y#}p}e~+c?1BzrP@A2t&7CniGq3SYMIEdIsPSPJXoJES(mB52ML~{u@ifb_z!yjLhl!~Zc|>^a zq9)S~_1j!Hu1^4dufb0~V?zjrgll~q!`fK4Q3H8A`llG16d~0c)o>ehI~+^qc=nMV zJ55OdYbpcAcayolh7a0kAkI3E&E9#$s#<~0B~>%5Apn)%dD}cT5TpU3yq6CxjV9+W z)_-|+9-Zph_^QSZ`W{R8<142)aHG|Cu3Ql@cKMg9&M4&JZ4GsLGdsRG5b9J*HWzP( z4}5Fve-<7ox~kMATFrE?tscXV*8BoKR}0&)ZX(GbM1{>=!dAAMdg-WoFFNO=vHn9> zwniJP53XI>nOj}cSIo`oK0mZ{KtDj)>)A+)UEn?TIbuG@Y(HS;_v(nFLU>4tbaj{& zI50+(;#lj|R|jV>(TmRUz^@J>vx)J2F=nGq7yLAl*>wLYiGjrNDw1nq(3cQN8)M3@ zWLq?o4TuRUKPH`LZuz{;XkqF6MSLV%gU7{|wR~bD*&3e*l+o-O2}=t`V?ReEww}^P zI{8_U6cSg6lT*(vZ7ZRHVO$2uGC#L@VFGho2Ycl&=-|c%=J|ktxUgDh(Z^?(X>C%g za+Y1GiON)$N$q=1yndauEV)(5txNu63qnlePP6wDYiQJA+_98%)~sc?r1jH=q3M}{ zJY?rnlo`4VK07wQq`__hTttU2lRQkdu%Nh2Lve-blv)O_!`vbMwLyP&kK+3&_9xwF zt}hZ&@s)zfXB-&L-0A!F$lpMHH+vuc0yLGNpzUh)UyLg{fed(8H0;q?fc*_VGm40W zI@#BWQgFF+`o2wRvX*Br(t2MMdkTA}kU2(jxvh=su^?F*LL9duMG4hL{Q5Hd6B3zs z)B5SuqENX~ZW#0zh(Q-`b2Aj06#VtR4X+>O(kqQcsD8p;I0?#}J$%?37LCm_3$EiT z*%tgkmM@J1KEyK%fm*BP7=-Aom>?P?SJIWFf#%nV{ep^%tTEHADIOg+O{a)Gb{a~f zv-F8D?CZ-emza56=8mH#1D73;o!{MnEW`)1Vtn%^AWNc)lNO>ezY57p$Tdq3uUklV z{vTLKcD{8;c0M;GTZ3+JPL0BHBVPw`dGQdbOuLT4!n}AdC`)23rMZO2)9yYU$Azh4 zp}?&6FM83azYfeo%B4lrf5Blf%jJxExkEf31CFwj`(jYDLDGH(9}6~%WzjB|!)nNB z942#Gbu!W6un!PTETv2qTs2ziVJDbtpQSz*jQY$aP@eLX@ESC`$}>{2CEu0-5q`#r zLa&dmjRA{}14 z`n?diXg0?QlZE5O@YK_@#Yrmqc9p-mYe2H!|HPZ!I8w;(`+A?Plm^nPr*2lk?~bG> z;hh>5!YgP{6MM7noXjG7_Q9#-Jj#4oQ8B9Bre3v6G7zgl@@LVNp(p$6K6|n-uh(I+ zbA?wGRq?x`UcUoa|I2js62kpXQEa1pwMY#+FaL(>9jD~J5U0VUgu7bsgg2rpmBdES zR+VP(XpMXsYy~D)MQkoP{|};Vsri2hiu@5)EwHR^nqH}!CJpQF_1IDlN5!k`(k=jWdlmTYCaoHRRNAMtX>3F z`GxX`0%A}WnF{$gvFTOMXJfN4CSsh*pg%rKEy-HFnfgy!k)C$-tF}s%lg5NZ+PB4P z%(8E$!xRo-W$mh&UDpC@$+3OTjzX^%>XmN`0$T3u)S{G%27NHy|K%OS z&2yEiLg{{vGIc0l@3FTK?|97+&I{<#N|aQ6w}%+O>YN#W8VC<>wdVjE;ai^(bkI{}uk&BNnJcxw>|- zJJJejQ97<*SEHnXzLFeyrq4cE-(4v=2qB^ zV@35;6%#665R*+${9SIP86MU0elL6TV6m0FiyguO?>ElH^baq&B3|BxoML}RH?XZ@w9&mO-qQUw;8s|XG5H(~_jdM?@V`PP6!?druB7;sB_A{MO@4R8( z4R;x5hCE4!y>wKo1D5L-?mO5tIiU!60inZDt$0z!rQHp`hL&kv{uiC0o3DGcI^De0 z^#d3JFC8}1DL!HiozXHPI>qhjF6xDI(|itjL8o?LUWgJxN6*y#tDO&hstbiPvwAQy z;ce(WvbHgS;d9U&3~k4A^xA(Y#M~My4}d&_GPqf+&jAKz*mq^j$LNRO-V1K+L`NYn zwQ(LWvAVJ$USk&zN)-gNX$tV>CL6qPe?#!?y_TT4)BPfFliC1l@;BW7>T-Bpiz%OXIxSZ!7 zx%NeU@~om?uK@kRZ|}7OqP}nNV)1q>ik^k1N?r*q-`)!zKW2~C6l1}+_kxQ$;b2GS zrTGKtKPYhXH6qDqv&1Y5zP%TO)ei;ZQ{sYO54^n>{Crd)7m%rzeFw9#7&HsMy%z@4 zi~nX~wWO9INC+4Wc(4l6@%vI#2nyaAaybCem67(iww+Q`r)5Eb!^3w8b=WXH*2Phe%RvHvCv)2qaJUXJOB1{>~y01?tt0VLtEC&_#qZF`7l`;`o|iEUeOKOB|LZUWC~w%3?txezR5LZGx>mXgRwZq2ek=&`q= zQ9W5p(K5wzx2^|0#pVXLhwDL+Z9>T%J_#M(GO{&m%-`?x&IG)z4%yW4KB8%zPM|f7 z1A0OShJQX@h*Zv)s=@%#hg1DpKQL4+n?7DbpgEGNB#+w+7@ZY0%byaNvGXfRb4(38 zA>GdQu&?w3WpvqC1RY!Jr)SErnMakXDEBH<_PU`m;s?~^;DuR#t(#Ol;+R&;h zsN7`Jb->16`M|I4auf-ImoaPs`K+G5^QSHM{)}U*l@EBqp;VRLq<$h zh9LtXu%t*t<8oby!iSFIrVfTZFStg=-gVO)(5sCP4|kvuah}IE)+TGC^)<*c%!y2c ziYDBp4XHVH`{4CNLyE5%?xb;TsAv>=hWFgQE-kQp!<4vo0Hq~Fnkt9_duZmw&T1!x74uw*n-YFsUSMKGjpHA6O5(2rYnJo`&=mmSok~O-nKRsg%hoImVg-gikv4j$+e`UtJIGOqQV*DjL3&`Z6*phMFFiw?<{X_4(K6 z_OY48=|O*AaKMcnmY8DSNO-9R#rA>Z?2VMI6DZctgf%_$1#HOWwbd)4`W~e3h4FZ6 zWp(Y+&Y*u=&nV1KhGr}U&l8zq6Lmvyr0J6+0b+j?>In}Y)gMoTO{_`M3!m2z-gL6L z^~m;Ubq(IAOM4$0e)i)%_9)FVQ%rXqX43pvKOU575>*{aN&{Z59!8!|v8otOG^UfB z#+|jG=s@@yA1Bs;X%|M#d^t4tHNH6^eikMqIyH4(J-ukY=YdN{H~Wdu;J^q}&IlQN z7;NO*-~*<)Kowu`P;>SUI>3?5EcK9>>?_{jo8*yaAIl0SBsFFmY~`H)w^}d^KffIL z(&Cy5bj?Iq4ED>>{*a!@#u`Ky>*#K-2bV9>0(&_m0!@|8Vj?nA>#FSU3iiyW5)aTS zIZu;|H?yBZ?&t;UN!p$9`h~f7T^?^=os&suJ0Jw%Nq2aOGo*$19#nL@P$e%^@#a>^aCNZsP^Z?y#0!Iy=467iL$#qb&WF74tqJ2JbX0 zes@DQY|ktjI+rh;g6*azt0>8y8Wtgjovdbm-$#cvbzmRRV09(%?ozGV6|PsP?v+7x z#DtL#RerP2&KBkF`rg4t4t1j`Iw+^hk}&A^x?>wAF=Ga46bvVK;D`TsZBTvn9b9Ch zWWTQ4cdmsb7hr+P5~9vPcVf@YMB(+@_OlUIm1XE?qM-zq@=KYe?85t`n0*(=u2=Aw zjos@mGrx@&Tm_&M^?bIa=u{*VXBb7+E|flmR}8S``^d(Q8fd2Ns6&S$J2cx0I^;eW zG)QGICswhXi6Hb9wFtP55h}Q(1P;+$L+jKBpYe#fvoWwd0nfx@2^`R}6@Qy`7kZV> z7;B|sX)C8te?5RpyYjhNyKJB>WMFJv5gi!C z+JN(gaSJcMtui;874z{yOrN4y(PnE$`!u2%wufH;Ox$&MJ;FeF9Se+NZBIg{e|J}a zqx<$ShzuL<*&4W{XHW8rs9YipeJHjlk;r?<51%0D_KUjcsQ;d;Ucl_1z2HJ-D@>jJJeFvBtn`W7Zq2mgN7m zwDC;uy;Q9c_t>Mp1aY_^idMcY*_jVb)(o!f2)q-FivooB^_^jCgbW*AeQ2^~5ISGP zvDr?+FQbz+gP(Tpr`k*Jj1gRFQ-leXCx#|#hC%W~ISYG6kUOIx1~B^;F+jytOc~#t zgTi?vCEzl*PTkq3 zK$^ZA)3(UDcYEPDcy8|9xwx(~wA_j3`6E&mmkwakLnr0l>HB#;dfR-FyW_*&mC3EWbk*JP7IrnZa?i==`M1#EmE%$DJBdSo!T&1|ZPoiMZ=%?o1 z9Z)^PKO4Ko+`Ud3ikf?`p^SU-ZNn{QZL+%_&9^dY?%izcsDT1-q+N11Hxu*J5IsZA zy&DL9!6$h#=IgZFyZO8$^tD$HzF0`uTIZzAhJhNJ7-Gq}ce9J%BbbRrBi;pgED9Pm z_g;fm#=pl>_6@hX-Y}SZC+61?$6n)>oO`d2((v&!l(zKpNr4TIi3<4nKSHcAczP^+ zl=>|s)wdmP{hS`YRy}5Q%TgsDvzk76I8bZm7#?IZ%VJ~E?s;pY0rn<#Z1E^Gz>1(s4}LiaKWEsF;*G8i z@N~3kZVO(bRmjE>3OL-;ho2t&@TLO?p>P$3dUL0F${riS6F$vg|1u1<+1c9M8KX1v z8sAjZ!8(8%A9jHr#*oVK33I18V#P}>%K}aO-7Dpru@4g~qT~_bwTqgbH`H%)Ss{(s z0Cp;u^AHRP*ZMeywR7W)Xje`H+j^rKZjHFMto@N5J55RDXetB7ZRYNXaA15#3fR|( z0b1)kHhbq0t7-*$R~G*UK3k?iYI!doS{hBxU#$P~>^wTvv+-4p9rQhx@W)q94R32` zo}1b6#eqfpedh|a&fg}q7yrvu_>0vxj}Lrn?0*&>DY~lErNvRf7o70$ z>Amb1@VQ#phIJDOHkqaQ>CpsESYdOQnxEyM7^x&4%(mJa9#D0_lLtNLxh@SW0QpCjgj%=QCjey@%=Dg?rW73u0QD{x>Etk>3G68*Kd?k=S}l8|ma{scJ=> zyp39caTz4b{M_b+3CwLB?3KTugBu%|=K}&1LnW{(Ek7>4>@uxQYE{m%vnMJy`Hv4v zrtu%uWm5Z|yEuYpS*2we)}AqAlDkXSscC_BZ&`nq8Q*;>haTLLhvj8-1d1sw3_kK z=q!CA4Ey@B%Oz$Wm$~Dp$-rd?War(0EFSzmAx&a5n-C3tLwqnR#y4*QvLw1VX(1Z( ztB|aOyk#U->o%4jlAU)#viY70P{WDBx}HjYNOrz;NOnFqBwK^tH6#ngL!>h8ItmN( z;=Q0OiM5pG5+YB#`*a+4WuDUn4Y%&moI}P!3l>{&SPYlyC>uqknp&1^@UdXCSQhPa zIjn{}id&h}63IeNCnT@+0iubel*xjtMqi}@yvLUYGQitExo^)AKzY`?!!XxAOMNaF z^+~#(ECS^zPYJK#Wj9qDsphH+;2DVUGfotGeROray<%HzatvY1ccZOPaV)jiA3mgq zqTkK#8r&Hgik>)GqDY=6%2Fe$vf;J+(hEU>d31Z6WOq1T3{O2hTb!g$e|jK(QHq33 z;Cb1*eY^J9*ZXXxG?2=sT@6j3+Z{=9085wb+MzKe_fw@%57Mez-Ps4HlJh9@WktoP zcAI+DF3EsGd9M`8BRX2vX#aJeJz1F7>o8fWtD_BdlU)e;i}JeoB$i6kh5co^dIW~U&H@0?^1Jc!A1Uz4;hmFp5*Wiej zM1L_os!;n1Z~Ki>jt$+3OTjzX^%>XmN`0$T1&tVQXE!UkBRhvgd9&RPeaSiVabyW(L`JKnksa}Xn z<{U-xS-K}6H?b;HQZ+;#wr#50RVC3nl{8V{@nR}+B?wkkNz~M z?k->@8h?UH%jbBpPgFqZqc@%yR*TBAQ!1g1x|WI|T0r7NRQR zV?Fjn?ro!<#NOqH;Xl-Z*X^UpN3rQvUR4+qhM2Euw8I}_2Nw79UgQ(d6W`-zdiW7+ z=t+67EsY`%$Y(1O8@Y7Y*x&WoW37+xKT(c7DxX2&B6u{gC;bif@$|#44ck@WS-tKK zw#Vw_^OLpBjg_^H)wT2E9d#Xd?%4knIk0Keoqs$jXXR*|d}E^p#thB6$s*)j?#pFn z@3D48yk=eB%b))6>s{j~6bpiCoO?PQBP$#mrhU~F8Fad^pXr2p=MDRAxXU;*Z&Q0@1RMS^A=+q8!22n!j=$X2Iwez7*b)hiN>Lo$dSmARH z--g~JYa0_7J_pUg&~`l6ve$t;pfb2ws1AXTx`VDi`>u?3Q}8jJPnCqz?r|spX44el zT~Ic7!D@)`+j~K*I)PtExow;WBCoD&h==Cq-0A+i0$ihiA=u9ze0#4YaLJyA^3s4e zC=|e}-JIaEN<;yO8v5Dr+j}kXsx1}#?Y)+(dejQ|1uYks^Zb*z(#n!PVK{gqUQzzm zf_kXFZ}4LA;34AT50nhNz1Ma@XO4dXzE4st{PtcD1ScF^;EaukGTQqvn<=(PY&rb) zUJzD46pT-a3w}ND_FnMwQFe<4f(y_APa35;=qsUQcNk1B{+o%_n9)}2?4o)IrvKu> zP6R%@2chmPUK#DIOrVxxb6Z6<$S;lsb{f7yH7OrnI;f{m_OCgz5;OLC>^DR4(!iDu zhl)Y-b7Bs#0*O=%+8J)#24r9X;p3Oczzn7zURz%e1h>hn`1G#uP%3@Bdem|xIWcQE zzB6)<>I%r%!VIp3_$1K3HU68z?TE_ z!2+LdO53nAMsPKsB8=CMqS-C{SYrqzPxohGuRHWnI90fZeLFOq7Xiv^MClq92@G*S zHntyiP$%a$f|&?GYfG7Rzw{iR%f#Pw!N*S^2WhmVkneNZyn|P~Av>8?6b*38_Zw%~ zc3>_Cvh7owFGc6^?m2VDwN68+h+huPhQ&7>Ou^29xi!c)Bpm|1P{A zg>c{UXyf8wn*TptXvO(+4wOMg5txsWkv<&sGoiUh2pYt;b$o_QMtRhVY(H=;r^%-g zo6G(425pYT8(?$1G4DS1D;a3xoFKDgWVDIzfW{+>n835fN*aOL5G-WMZ}u%GK+b{S z4|;*XqpYQ9nc}%y*Mpv7bA#K%^`OW$q2vx9=a@B)3_26=wmM`}$NOmHKJ2tkC(xS4 z0X?Au!#^KKW5n6uxT&1N0MU~;zt#^7Rm%p*O9%{Z^!_0bo6O!vjHz+W-$=}Gme=vk z;CY@-NVhXQ>?{318C^DXpgd~*^h_Bx^NCQDdySIogvuoO7|ix~s0cplOCK+XYeTE5 zpmMuz*n|N3ZB%SJp|wVE%g>xhX>$ycBz2L@$*>DQ@c4$a7a~ zA>dm&I4Tvp7GyF3?>cr3!wm46Smy{!AfETHt`7FfUn(+Q0-+;kbs;L836dW3jG-$Y z%Vw7nB$F-V_M=_SjZY6j@Rz;dU6eTXY+s7NztE7Sh8-zhp5o7Pv>~M~gl_X?KH;d< z_h8aZfG!H49)sD~t=3i!WM}HO0xh$~`FBFIH=cAU8Ghg>WMz zbkx9Wn%>8`bR-4;H$vlDy~XTVGTzHp_{1ph`|hN_Hgt|SlX5Z{^o8Niq*65TP!?LLX8|;&z8B4+QM5fq8 z-4GmU8Atu2P)~UHsFaB{NqXV)8p4}SHn$$x9<8p8Hzw)Ofrg*`c#l0wv&2rJ6)lhmz8Om#c@7=TodIh7*nHVM-{XvK4&Ry8+WKjGFm!s1+{0IU#-)CL}sF zbzv^P( zn0$PQ-@rzto4BjWj?7KL2y@JPV?8dA+33yQ*#s?Q*B{=~oMLnNwDLNI& z#OXVcwF{*Wp=}>)zK?9|r~&7ajpkSog;4+r4Rz5db7s&WmBE~XY+h<26G7-JY7vAw zMyTKtq?Fh=H3hJipSR#M9x-<|29_t_nOH0eaEvqA-)7x~UZpd}TB%st$|=-e58%?S zd~PwU11osexaA{Cu~`~{qsYQ%Qvsc!wB_WEHcX_C3q8;WR!5W3$^|}6b26A2Oh57H z+aHMC1hR}NFp9MS=L_Q&UVdANG@BLk@dKk+Tf=F4_yxf1;@`tmHB$J(VAWBq%|Oj$ zb#?rriqtq4S;)KJ6$}X9J}`>4HQkb)J;^U3qmswGVp1sWF|6*JI`}h!3hoT>dh+=} zf7YGr8ydwLSuD;p+NZ}xv9?~-qZZA>BojyQxzV8om^kfk1xB&9#Zc|bKS_xXP|?CY z`KO95==ky11 z7QH!Z+t2`mPI8h(X14Onk1`qXI_n|G^kM@(Q8LCo_LvJnoiZ5Qmh8-jCTj*)b_Cur za-sks-a2uPWbhey_j;DOM|n!!&yc0ui>cg9G*A0E>qp~;$I zkUS4P3wuV8JEI}SiR@b$cU&3YoXml?XRbqyg&gj%-No~qt`>uYuogUeI$Gc-5zr8 z-CipWo|}8$rL%L_+sz~ekVnqF(*g8+zLvRn`hI%q7Vasx_1;TRhEa3xbZE7oPRhO0 z*Y&7I%e~W=m$~;6NF@+d-C2y9dnW^uk$WevIY-XDXDsRX?AHhFjGBA5!Ens@BIj&a zXyw@*#Zz1v^%eRxtD$6i1iaYbNA zt15Ety*^~zlW!ZEGi#IG^=Q8pH}Oe^y4l!K0}c1tHDzL6Rmq^8IbFlUC+FVH8ouCX zB8ZlIx3;r`cC}YufISNP_0-(E1(xU8XJXNacL8L?_>i1?uR$y0-(xBJhFe{4xT_vM zHTPa0rQzdeC~fKGlL8wa6B$-$>+rK5A=Vf?Jr+Jn{T33EmyNwkH9dT-dfDohrAoeR zHGT5P2`G}k20wi4^%?kK;D}=lvkm&o!|-V`;#3tWL{hg2NK(=#5ifp73b~`*E;K#=4Cf$m7vJ#n>c8qBpAH zCiAeNiz5T0X0~*p%^B#u{+TnmFfjRZOl*%r-Y17d>8k4fj5TRu;dPIy9$gaWZ;EuYv(wuVdq%4l|t zgoTbIjaT+_L}KeHZKRW*1xcZ!BynzOQV9XkLILA4NS67z%?lHl+d9}Qe?bQ~HZacz z1dvsnLFQ#zo7AeDWoJ)RZr%=jSTc?Os4kP*_nh5yT)?vA?$R~+!`@G*b65?G}^<5yUnlkpDlW3dOt+>uzv<6ypI-()l0`>Plc~dpU}fnO z;gZysT`n>6xXc|#O$R79AUp2{Wbydm326nZ1cGh9C*y-zF}`^dkR{Q@Nej`KUxj2P zy;Z24~VLwcO~-R!QxouP5&i4BX@zC)rcHJB|MUc36e5V*~F9T$H4z7J<};Hjr) zi<8vpPv12l={4|Q@3WQCKzjAm%_{iakrbDz$#!Nu&Q^#dc7%?7a4I>EGGA6yjB2;3 zSM8Dv7*$&cfgBCW+2_4O2Y`2#WkV`4(9I4oK1O_1IDl zN5!k$y&Xc`cGPso_6)Cwn~)K z$avf%QL-1&+O?L zT(}a25Dij5%blHClv2^452k0YddG0{T&1c|x}T#=9m>~x>@CDQUNeOA0(!I(B~{<; zp}JKqN>cjSpLeCXu0lzT?e*aCSqJJ+I<5fq3I9KH?;30OQRNGI@8cXo#YAC;$I(`< zKuArZjiQZ@b4lCAiDU8tk`Q7i{bhR2ocQc?4sni+&v9Y{SMoFj={64`G#?lVKEVe* zpdf@A(c;;34+t7G%!TQ0bb*2CJAi=U4lmt-5n@)=zy9_5uT{0`vEBW_kbw8UYCURI zt=Df61t`{_T^=Xlg%kVi{rrZSQAxP0M@b6Wvfx+vutzLViE4G{VpqxvYEde`0MgDN z`eG9DC=h$II+QP^Gr+^8yGoQy3(?i0aBo~ATTV5~+t`ks{_LP0#Yqv75ndGd`zr3wT;XbMS-$PAGDwQ{sojGZgH4;BgSAqqfjnt)JgIn2WIa!wFQMV)wLY|{r;6~rgdvDsggO4uuHa$=cZvw(SCP{fr z9Gx7_NK*PMG>R>OJDF;+)1va1PsNv-k=ZoH1duF8i^Lg(or_pDZKO%DxX-K^-|>iZ zG$ixW5jUjo>rxnZD&lSwa>PCWTJc8~#AM48pUd5sR^MwZ!)@Z#lOsizL9IZWhOlui zDo&`^U|!x7m&CSKhQSC(=`yp>q7x_7-%>LkBXP z{wEfTa~sf+yz6z!3uqC(vs6Zr0rJ^~#AR-pwmz^fem@egE#7`OqI~lFli~lA?0vQG z_-PY9+C!RlGd`G}j52t;_I&QAyVGZEI>1o4TC5(~4;TvMv2x3+R>ZUHz@{~K;r8@7 zYbWF6jEw@0S(()hq=%f#?YYXlgxT~^jdQGO8(HC4TK3agX7J@AK2gPbZ-#vx?=tob zd6J5?OjN6a%iVh-k>cjc4XQy8lv$50X~jlD}?K_x^qgevfce}V$~hKQ1YTn^R546cT-rDaW!!x{ThX-Nmk#A&r#U0_@a{)Sed}BL4Kj-K=eW(KTveVUQ?=nsf=gf8_h5FuY%#F ze*KNNGBIQU!x)E8LkKdiH4eFg{{l*o*$V*2a`^R`jJ=!ENgVbC*?TW<_U?f_IlJE@ z(C(0}z?T8D+Q^$~5*g~j?8JSyyRX%eJR>>MlY!*>!4GG1O>dPia}bQZcOtG*cUZ)G zGagH)Hg_JM?3e$^KeTH9IUC9lqYBJhF%HEAaLT^8$O`<=XNXZP+r?+tV$@rmvfmGq z%S+V$^*G!1G1>No47N#4&{;Az+thEs;<1mI;&YC=_eo746LSs@5_YRHe__Z?2H5cT2`1s7Gl$KX zX&Z%%6};Ce1YUffJaiIt9DnLy+zS#bZ0_~;dsIhbni}ZU#=*lqYE+zO5c}+-35U&N zX)bP_yn^{ylHa5|X~G9bBs{n(dv0HzKCp5@@3@8sV;07uSN%ILwg;@crkpS0!()z; zDX|0xl*x!`vDD*@V6kv`IY00~xcDL!&KW9ia1fbXiqqO(zq4|}1OTouj6x98#_bOg z20BFemL8AX*fof-Eu<2F8Q?Xs-WE>4@w{Ve%K|B}`9g$_ysxXHB9$N+a#*#t7Y}NS zCleymE#$3+cD*IRJ-yt?f>~W-%2&Ov^Hs%L#jB!*)sX6e06C+%_>so}Dst)KZIF;a zX>qugUafG_qxeVhQ{c+czj!=1EOMu7cr)nClC!xc<28efzaqzSl!($4LYMp)#_3Iq ze&^0zA~Z+Ing z5O=|2;f6~7qwztRuZzY`eSHXv>oT}DtK5DxSk*@^WH6H{9UKg$4*}k0&gv~ItW} z8p6|Hh&4%h;dl-7qI+l0-E)3(YkRS?m-apk{o?&2aWC&>)+^l=z@&NKI3AQ53bo&q zJ`H%eI?cRLV$)_kF`P~wis;!Q$9fNFS_0I}wSQCcuGX&qemufy88A@j0djJo4;qIpKt)#Vp~L zl}FMw7vC)0zbr+Dl*gA zs_fq?^~{%&7obfm9yGI`MDFN+Ki}Mg&RdDRINx2Id2IgX3ybrY<~j-O2819y86bSP zSx)8@r08-~1{5lF2i3Y!m=<_$J=B&~7pR35;%~way!=-Vyu!voMtK`ASoSM?m$Y|& zr7vJt|9M;ao$8hM0fkO8toU8V+3+*Rkt3=w7*^df-QNDLa3`)sg)cWE)6nZO4=w&M zM!}k*Ux=#@Sg@`Vc)3!WE=B8OuMaJl{<+sg7~*GRaeq_o?(Q87a;Q$GC{Rw#Cpi!< zPR&mXJ0y{qu?K0C3@>-!nfvr1MoNZT&>?YsPoLe{+uYt+oL7M1VTE(2s{(8%Hf|L-?MCTCm|}o6-$xdv%|I+BjFlfZV8Y=14=SVuG)QGACswhX$slx!T7;qA zDcenfM;^i9`6POM@jo#)?FA>I%BGpiluEFME!Mwmu}^= zd2%%>fffK!ieYIGTNzY9XBcgb+)<)LRu%Ai=a}v8>Te{Ste4;uivlmdR6*@x&3t?U zigh4QyR>J02GI=H!*4()?z;OPAyrjJK(Q8+=6m-`A&&3c6Hu(h#3eoZ_NSpgr7K+6 z!3c4lYwoK@+ShQZ%LM_&8d*AV9(tc1gJLcIY9KA1hsiy{c~Gno8e(QP2nkS7tdXxr z`QG{RZ{WlSxNL#w3<||sxTu^gn&tWTqDb z_(a7R_t?FFSInQ_f+~`Ghi zO8efnCm8~BSQP-Op98;~ZIrcFp0760t#!@T=HBMUV={HPH(8zRf8gFfdrizF>sFXP zh9O@sCpj;=r#8=ASX{Z>ySrR7znq(UcNh08PWYhk(6mZ}|L$sN`&|ffLHXv~+`GGA zoIKwZtmUvY?S{Qxp($A=h92m1uB3z{|b6zlbVy)PHZnwo(tCn|t?24Icj4+`Zyw zhMRlusEm8^rQsI)?U;wo_n;+YKI@@w7xv5m;(#|GEj+$8m}BhRyQMHZ_?Zmi<=&m; z+@M|Bt8c(x{G3N!U~cZ+AuBlGnOwBTy8*#b(zv03WBwd?;78UO&9@UoH2gK?rwXmG0 zG`n#yef5Rn`XkSWQS>UPGJ=0N0Dq2)cj1k$1M*C)YJLtT(Z-n#{M|Q(KO^`fn-2KQ zhLk-9!b9(7^4wDZYO{Op?Ct`^%p3f&mk!neH2BLt(U#XS?n=hOJV41|HL~TD;+LFp z^=@YBL~jwMU3Buis+-N#TCpJn!{W8sj#cg4I3wDXGvKyasH2;$!&WZd84&LoiCZ|~ zLL+71_%fCI2h0GivmaNz^Mo}mBHgmGOvQhKW6KOgt?%XMo!Hzv{do5;kF2Apo`atn z?x6En=)b;kP@195tv794gyE9PCdq%fK+BR2=!?V~$EeAMUrJf=aAM-){Ez@bT%TI*;=jnBx36UWs*41IZ zAc;}^$%M z7%R^4o@^jYF!|NVdV>2bS^PHJTQ~13uft zUo(YmP}*2Kzn^PX)Xus4)^96yC;%>l^)f$s_OU&{ZJj(v{{sr#IN-bl2#6n0XD?>^ z;sUQtYD;6;xroaA1N)7yf88H#$99?ebx*oY$Mnlmx=WY!ANL@{Eb)Hvc4iGtzZoBv za_(R)BS$juC+H(tyNfoU%iyzP_$32%BX|)7UuJoj{fCvr@%)@~#VUv-)8ooN0OCh= zGB+0qtN5y?DO4N)XRi9a3;9c&M)8;U1Zb{6!P<4^znNB4kqmt23)xz^-{6>0OeS>6 zzA8z<v!yf)^{b3Go$X&?Q*hfZTK1T?+9X%qLcFOb+>G^xySy+ZUXe=!EU~WmoUh^P0F(hp_-$`z-T$ z$eK^q_2dXB&v87ZpOz36a+ANHcqXErY_&TqIcHVs2qI5+w(gLQsX9G=`t zVVx%yY*zaY7hwvbDw|%nFTD}Dm|L>XOLo)2DH!$iNOO_yZEY={>ix~75h?!Z!>@Kv zB~9cH$6}*3kwHE6unK-TCq)u$!(z_MFFMObH#s}~<-HMDnYEFRcxENp6O(^d`WV%+ zr8iyD8R)q@YQ$wmMf|VF;th>;y$h42I-W|To9sf!|EO0N-$u6?T>oXddI=%LD~ioL z%Sa7vBWeeAv~a(41waRm2$>7ky%+y77KaNoEtsB>N@62yo3>W)=EoE-m5n^Gw1Ebp zUf5l1z`N&(?VZ1n6(O?MFopkzuqfrXw;BiuZy-~e9 zv^u21*@LA$y$NeVJ0bod4OKayGOU`3I&^_jYYq3R>>jZm0xh{ z6R1j*%VTt&d&WL{Kfj@7R1z-hQIdi-x25KBjqTX!&o-?``3C=a=`U|9Qn*hl|MyUn zl1k+bWk=oe_5}VoT?rO^Hd2>@4Q^$hM53i+oQ&FXB|Dq-ux)Zo5*ST`R8c??O)-86`KVix1|d(Yl+ivpT49o^6AKd+KyW# zF=ae(_S0Hsn42ig`cG7`-kV`x$GeO@L!P8! zEfdwM;Bq(NzT!<|sZ9q}{&SUxoOn@&Nw4E?4t2k%rf$9K(W<(6r&o?(2)tA*W>S32 z8aiubRCG$dr)$iM)TZ?u@q$gQIA;(ggpQt-JGQDfeXvHtnOQx+On4)Dk8JPk0r(s= z2SeNOys+1S4Nyrg7VC3>W2@x5t4<8qXS#KdcUcE=G^+ksxR9kzDlx zzmRg&Yb0J#~C}M{_4ed{Ze0j=#206*TH%A$B z_((!4`Xtuq*?Z)2aiuXY6@2zy;a3A`5q?48$3r~-2AUpuAgS`_HG1}584=xmgBOdJ zeNiF#H#e0rkGbPae=;&=kC6n@v-ilKc;VozK7}Ah^Ma#Ir}~b4t$Mqsa{7znNTZssHKICLaxWuoIyV??LD~i#IlR zH};^G;_Uf8B*mUMMyxb@ao?nT?TG`rNqKZ+e0h!g*c*wjQSnk@Nl!ynB7ClJfP0dt z2VOO)oaY_9fN=OFHZha^cW<9LLjq#SmOL&S6cQCvZ8CWgk5I zbQF(w-|^$_jY~rhe2&6~#TT8dz{&(}4e|>u2cj1W`GKO#9-GKhPqi{+@eI7&`Nh<$ zV0futf8(tIhAdzh%iuk-YTDKAQ&UjH2cJRDP$`|`%Z1{ zJU-bk|C4`c)&A4_Fk2QGVpM_om>2bR?bQS7NT1sQ|qd8(KW zDrk_f#P`NT8C^CP!FY7`>ESkN7D#2hDR1O(7;;&RgqtyS%75JLkK>j67yjqz-I#6{G9eHiW&5MaZfb(n;M&Kx#l zrfn2*R>jlo6pCdU6`_-$T!bAdb9r zwGI)!rN^UIv1@fy!q_!{8Q?Xs-WE>4@w{VeYw{fZ4^uW@h|rPub#+vv5?GNIMyglE zlL?XO7IJrJmuG7DAqa`5B;JzXo?dP#_(*RRuZkL0L#hV?kmUEL_|oA>}L{ z#L)yOEe_Yxs})Xq6#poG3S2q%TDsypa;{Sqdm?D`FQdgdjfG=tSczR2rcV^w{d!rq8#Vdupaa;JzaIPna9>i#P>`mcpSCAE9Pf4j1fjh*&38Z7*?687slT9HLU7!7d^6=TE*OjtDvbNldW*XsNnZL(qD<3)8`xd z!tgPv!Cp+*E3TLxGEn?!?29$WCVedRWYwQ`P=Iix(||@-UNRubnoE^t@w2Jdr{1o7 zE9=}PS0p^0$VK;oqJ)>k9$kI{5b7hJMxdXX{{POGwn5pNRH@(y~X$gsjO6IM7R z4o>FDu!8^HMxg~57c+|q!(Y@3FAO7Zvr3ZEfdl`!81+0qIVOP0a4`j8h76`mtF$2+ zZm8sELC68}b!!V*;vqa9le{1?M$~FWdu*tAthDS{CIWPsCWr;BfT$ z5hz7?WOEqTj`Wx|%8DkufPJ6}J>12T(QbZ{@KP;{-3@8%jnrF*DAv!UH3`wc1KAKC zr7)Jr{X|e~qHYMzG^=7S5!Y^Xi*Bei;qK&l;+iARmHGUaPR3V{#CNHlaC)m8#G0hM zaJ&Y3(Y>?h?m54?wY}Kcv(%;E_3s~vdwDOjUg@p?Ce8cC@u1XDsQsq&X~4_XY379z zn>OQ#l|@R^qDKe9*EpQmfTkrt&3uc(eT^^9h@a7dL{U@!{XD+n{8HLLnE=R#DB!@z z8XklU4h9?fHTZxT`-4z(_6jDzk*zEpNKEz>Z`C_(YQQ7U-j)+iNLtJiZsmkqzgf6{ z4f#@VO%Yu)5fuY|<&XsFnQX2>bg_%>=6Z1XM_NjMJ%=8!RM{$~A~T(>%KoiV&wMG# z0B!QTn4=||*-s*O^sDBZd(e3+QQ_vhi!+bS-+WI;{9z#@X;AfRVGRhZ;IL;{M#Ov^hL(vReFMjDj^) zs1I1Mt`c~;QbUKJR%}~Cx1Xp*;g6+h{%|z)9+xD{=Hf^iWv3HH;3S@Ms zvfw7C6bBO!FrO5A-^IBb6g*~gx4#TVZ756Y5C9ioB{>9@s-Dk%DSAUdnCLr^Z#POG zLfbypd>>iZGXvf+jKzep3`GMb49vx_QW+YgGL%z^El5pdG63aAL$i!WD-ynJGJV$yu?eksK9eR~3mwV1f1XV3DBtb%%X6dv>H zPbnLIHA0-!2-BSFt4H#k=C%q`P^^)qj7%FfI_jxmUErG&_pifMQ`_t`{j&t3Z3Mv7nxluEI-QRDi^&s!ss+v z#@NZGA`IXY{d(MEXW|}`UvA{D!w?V*z=&MdJMdNBdNl*WBR!n$DI5RU>WF#jS*$^k zHAzN#_fzMkS92ua50BXq3R#mTDe%yku-9T!*+&lmCI`>|kS$9`VKbP6LclWbV{%7Rg+22&dQ&VkE&(6CWmaYve_ul;|?#Y*i=Ipl{`SRS{yIt5b0}bdH z7=c0h+z-CT@~B84P;hhamcsDhvpgC5b6)P<{(bMzcV2y4r@T1cW%tO=z1yFJ2RxID z_V6~)v6#-VbMGB#ZTx#IWj}DMy9Hej-#p4n+ zlsf$4ub4Fk4=t3B(#=BZE9n2EA6Rw6iK!k~wf|+;y%g%(FT)>+W_=v~xK=6#XT4t* z!D9`x1NrOCNHiI>t0^l)(zOX_QSpA_QEj<>K&+lp3#(yEHx5$%FBI1wc|PTmlnX<<~c<9|so_h*F zZFbL{-CdxVd4pf}(!n}_27lQn+VVQaUCCHzSd|=BBU?@>evMzbBHqnRo#-vXw2Mxj zx;3Pd*DnSBUWPyY36Sj%EMA-KSk=b5jRwq<(Z61?SrJmRP)E0AccE&(XC!Xndhq@F}=;bi|D~lpT=>uum!s&5)7GT__sIrV24u-cWL-{B`GDWdK-%KG5G#n7td|( zY~8v2hJUCow&vfpd5`{rp2zFk@#C?KFBUV5%!;_*LfyH7J;y?4*La3 zjOr(sM8Q44<>08hMsk!*!S}_KEr80Skm^jGLvVkzc zSljbCv$U=u!^sGnnJ|^aOSGt zb3mH^5}yFg6)0G{&ipsiiYk(U@4ZfpW^Boe-{6>0OeS>6zA8z<#uPcxExsSk;^dA&M14vOz0`(drCq`S;9z!Cs3TnCW7P&#o0$_q8V~-T|l>ckb?; z0uMC9(90{n2d$Q+(OLFHJ?y*7E?1a&TBhKr*;d?0FgqZ-?gwP)XR~}V_i}lg4ojuH(7^1zeBwpEiXB?2ib2gr zrivYMRV+((xf<3{9#t%z%2LTfDJPT8GNKXhU_P;WV{*thqyMfC@V?;0L3v%KWp^$EcPqz3Gz9 zK+M!?;45Ql+sIS=dMw`1Sl7ES*~!LVRaNmzF|Yp$T>oXddI`~fuP7Gk?4Zo%{9WB|&W#KO9DtZna5i+@3hD`S_V{y1ZCjyA7R1zCu+qAWU_eOQpisN`C zZnD7wskxWEhAI3%ghgQ)Z2^~?GbKsB_{Whrk)u)ZYWad@6OWVf>y7H=0cJ22&K@l7 z8Hd3UZd>kN($F#3u=s~GROR4DVby##o2CskCf*SahPfp)<$wivhmI_ZOojXhp9+tC zHaD9nLY$rbgGctqXPG5ktGC+yH?K&~xcbvlBkI?`Z>kV>#FA+TdRR1X%&GWUI!xgS zR@bgt?79N3C1Lx#6_s68z4E=V%JJ)-W(&D-B?==N)Dx{&_G(cgSano<0gDugXVU#& z!8ZInSF0+N`hD0s6js~<3piZ9UHot)UdO!SjY4=6(4&C9u|=;UxlI@8}vED9yv$yAG- z7M1@|FTTX52?RdOv|?cG20{yfSAv%kih|9YZ-&Bg=oj2=mGk` zNIZ~x+w>!`WuXt)#jOjcA|jp zw`U9iyyilE^p{h{*gldkavP=%f}Xuc{yuRY`uYZZ_Fmyv18EU{LE*Q3Y*@M@Jm1ZyQo0P9TaloKZjxVpd z_h{^~Hxgf?;-$oro`x!*`T1+k?(FV0BdO+j2QMHTeu+)YWdGgUXU-78ZT43}c@v>j z{`VA%t=d=X&2>UR24Zn>WAD;AXf6Zv&q(59O_0MH2@2>NBDM^6IaE_RxEjJ%nKgP? zBSBqwC*W8J#E%JlFye0yPaEDI5*1%8hsU35NbHav)=2yyEDp%xUc)00(?sz{3Y-@q zDriKh4T}&U4#>i-hB~ZC1+lm%YUm?9wh`ILpe2QUpX=rwyy6X6sa-3|nmAA3av;k- zc=qWi9`C;63R8beTT?pz@Hq+_7GHF-0xJ`^HOMct9Ee^hUYcj!lN6;xie0H>Hy}>iwbz)iSruO+en2NniS2P$6D{p7eY_m5L{`&=#_fu6 z69VWrak=S*){tyUmEZ@#vM`ULkFsYS)7tWxG2uvH5)L|Z*o>LBQOLTJlbu!33{Jr{xym&3IfldL zAvG5_PhP=%NXc)~oiyQtBN855l|8pFPajyhpm$sc$199QuX?vwEU3GR1kV*h%fY-!&TQ~v7^Ny`83#7#63lTcn#cH>E%3pB zgn6=Iizcx_0ZNO*we)I*lODxCik|{kj=h$~CU8YAca3xAq8@=~ULs0U2wn1H7^gQa z`kjk(LkS45v5UUlJCN1VEg~(m#`)V6_C{0-J1?$~J4Ix{iI0UGv_AG1COH*1pbD5} zqqsGP`6JZq%He_?4iW35s_jkV*OajmoC#P|3N+Xyg#P@=S2e8pc*c3S?4afBH`(TLifqg2`EZQ-nX2@X5v`QPI)pB(# zz@rLLpI<<}4o`{WTnoYu8C;8Q!x-aP9S3AElPMh>45beN-e%6~EnW>d z#nDU89@fSe`uqr#BCLna zVce{8ue>uaprQ%K0vV`64|lO-w40wKyj06#cS9O`BlXrHiuE&TwHC7uL*t_q#uB-o z2#QVA4Z)d~f#2|5swbS@Y6wq*A=V`2h2u5Qi|(C0chC9Ft?k9mo|RJa%`e_R68G|6 zX1&s-F==R6d0Fyul^Tr-lP|9^y*;Cmn<82f?;FR1QbVEko6@HNFIT6T7fNi}j3?3s zJGH%aQMIGM1@**Yac*O4`^jAxvM{oON*yP1toMK>0&2zyWs`MBNCo8AKl|=ih;LE2 zukpnh@iSVGC~E4f=J6Hhmohl|AqqG!vW5pCgM-0FehofghT(xwbM^`*z=31WMfTQ# z#AILbIS~)=$g{WQgcFh$vxHk#U(%-Xn}z$=kS_(-6wx&k5hL&`ha^bPWOEIoi(Pa# z*MrMH(o*{CIrM;~%2qKIndxj*_HUJX=1a*7(54j+#-p7??&w?Rn|siCE0GuHyNffA z&EI@sasJX=C!yVd5QHav@Dk`&hYrcBx&}<<46*2PR0dSSv}9HC!gCA5GKntEujp+; zDwZb*3vH;VJ0Nvq;~*=@>p(1;%5q2`AFc(ktN*;MuAb_Zml{I5jI+@jofYn|u&4y( z>((HzxKG_Q8qBI&VJ!YIM!}k*iziM;pAuQF)TT?(`d&(-Hd2aM;%<7V_}N(8-&DK1 zdk2FYs*@=SlvDFb3E`6DPg{9NBxdYE8YRQa9Uzs&O>Tfepb;!)ccoShn+4m6jhl(m z8@BCdGi=&cp^3OBLn6RTLxWDq(e zXF;q=J(lE<7>}4c8zU^LDXL-cXbu@;l&d-tp${!%!gx_AjF)?(t4o;}MivXb-eC_Ls>;e7W8 zoDIbq;fog#prBYIzmHQ2QOT~%5rbka{%RmCo`=a6j_OlyVTjV6Mt?t1P^_gK>U{Y( zaN+}8wtyt34>tW^rXOE#OeiSU2$#LbyJz##!bM#EdKzYf=RfZ7Q&6lCT0{Sx^#HZ% zgF>;!mUErG&_wnYj*Zb9S*T$gfP!L;tR9vhWpcGCBa%8RAk&Khe4_NlJ$5FrB>Ck= z{yKrINiItt_?SyvA9|_xAOgW z3R#mTDe%yvyHO$ z%2_c0NwTt$Qf;81>6)$0z0HlsapPF-rdDLixWO5ydf#2nHP68q(Y7Fn{#vT?t*dh zeD8co1h&t;^9c-mz9HHcEBDS{PpM?=cy@VMHz(d3wen28yFJe__ksoN+`Id^xcV4B zot1m%e>adCFZa&>ydIm`0m_cJ95SGVaNj2CeM38~O6w+`C=aGXo9iNW0|zVUKxgNX&qpd$$yZ2cPB1 z*q`%q@AmI|hd%AqH(+^zU(e0GJ7fi(eI^&}@oqqHOlR1+_l~qS{ymnmAGp=sg8wFR z@7-1gK7NML)?Pj-R*0l)6VRgK{luf%9=ro$^^{szb+_F( z2rax&Tz};G5L!?{l@a{I0r+!VybEu19gt_Hx6IGMB-%K$fxr94@Mi>nWYYnk*^siw zKzQigOrCoRKy7xk4t(_DK^QuS0=^z zZe~T4-Xcu9=;WzeLs}~~gkV^_HruhPof~IFyK)BHHVbuh8^+BFRr@_7aSJD3YorVu zU#4>ZfEl25_T#E|p0K7xq}QtWPjGCRfvEMp{JayJd#4}o{^gN%6xDO^Q^Or}9t-`~ zHx9~Uxw-YG4e0;9N!cX%FPHq@-|muNN9 zZLWDF`+>!$aa=8I!LHy0LuMKN?ae*d;S|GN8vb2LianKlR=o|y`51ivu8Ze3ced_? ze(X6kW%oSeG^s}-=_B#C%zThtej@BYPp|V#h%5qOT^;rdk{H!_E{Sq4gjt+}qwYS* zQ8oqN*E0@NHZD}?XPL@o`cGL3q_#I@xfYgvT!nMlb&mIB17U*6kDborH+-3QI$?wu ziv?=Q?tEg9Y=biaEaSavG%Y3E`fH}J4N4np=l65XirP7M-};fI4h6tvuwLdT&px&X zxUG}t=zll_4AhO_MHGCQP%v{L-4{CUttDag2gQ;G%NTU{2Hbo=E}>S zLV{XZyNyjaAS5nqcDcgL(=s29nom$}Kz4n(0a=I- zX2tl{Z9tYK7cVWuQ~o|At0|9$2X*vYr4@|s`>|V*ACg`7Lb7zl5!pvMkl=Etfj-yt~iDaRF5< zC$WM?aQ7L7D^RgRo{ANyQ6&7hUvSH^Ek6#qDwZX?Tn+0ekFwu8)u<|pcQBt=y)ilD zo6&#QCv0DEVxkka+t~G7#y+Gb>3v%KWp^$EcPqz3Gz9Kr6u{%S??;7r!2hH#FAuE=-o{>L`(JvI`;q zqnUN_ZBCB(Wx9F^A;l|-1uNc_+0?+rA*HOXbTF*Al}HJP3p6d5o{>soBW#iM6I%eDG5MEj@qiretKZHeLoqPe88B8$&wKg0O|2PsSax^MlEnm=C;&D=by-~e9 zv^u21*@LA$e@H`B4yX*P=Cj!}ZJ;qR{J4^~nE{pf z*^V+;Kv`re> z4YuLuxms1B)bGRAp?Iz#ktRgD@-N>memD}ZW8U#bA-oCb(Mpt5qe#H&t`-HdP}1$4 zjMq%~wrV)I_<60&O%+NCwl|kbh? zeuWQv!~&J5R(CFTrL3S9rSc0PZDaA1)QgKrVBJ9M&FWCTl+FMTm+mT2GA%?`i^9Ee zjchs9C~spscKWk}dK4`18`PapMKR(V{O6^=ysb##KB@fQLrqF5l{b`~RZHT!5kF2> zf(4(A)TLm9TiIthS(hK|L)9*296Ol0naMy?mBLkiZ`e843-QHlC6+cMU*vlNdJ(5G zB~?QdaNCu7IMCdyM@eqmt!$CQi@D6zBsf(m`QpsgF;$t8rl{9KS*@08BhY*`#rP%U zqi#tWggobF1~>X{-h0!|9ei}Lw>j*u?+YmJF-gi};^^dXMv~HBp;2rJ+{sjnofeh9 zycb_;M$@umOaRGpv`Cym*tv*h(?*(95Dtols9H%+{B*<(>HE4A#+{0|J8MNoG_Gdu zC0P)YEl+$dcW)W)HJ0Hv@#@KuBFmswpiM*gyo-g}WW5IS^0wks>wC6g80nq0BebQx z5Z@mQ=F2UhH;h7;aV_6G;OZ1T>+M)yfad}V#e8ms3=Uvj${1|UZbJo(9-t45!~?mv zO+OMlJ{aB!p~b4n$l2K9odvq?ZcPux=D+f$jX7b3nVNGgAKK)&YymC8cb3X1GC)4t zkhsj{(BVk;9n(eHhmh$D$`JLzyAxa(k{aFJU%4RO1}0 z+D2A5mX`grmKhnW|3nq*y&3j(yvx`#1i)!lDyB@8on|E4Ca329nQ!exP6^oe^AG3zeS{W6clJDsn^CGorJx6)L zrdB8~LP;W4k+96_Wk{Huy|=xy2jFwi91Ly8^TJ*SHb5o0Sgg+h05iO| zOr$kvOwZmUm-eEgkeAvy4V>88*pVX@&Ugs_4amgOTt0a;_y#$9uaGn!-G3>>$(99I zb~JJ1?7c$bl06M?L|;{ftEp;dpU}u1s_JmptIyUKiE0hnEFCtdXEDyc4omL*Ro^OS!XHyRi=D?yqJ9 zZto)*FeQ%*yNeA-Wk0dK^A{4Mm>$+hkrdeeEbcWt0x>$pA1QENgs7kqr8X==fH)uv zw;Jl8#Ab?DDC7r`X|?A?@3;;=8s-g|kocMt4|y`4KD z8~Vf^fp&*%1-=ZJ)e(8mPTY4pPu_4wa-=5%$@fFA1Dk7lt9+S*U<^;v>=W7c^Yx=?9fvgX zBCjKB_=1oSiW2Q z#7z~`K?Mz3mH6J6D5J~fBG}l@K0Vw<%>t=Rn`*B$(^afYiceyD+|`61-KCG0!XQt{U$Cqz0ewxO{o%m7#fq06qT2qXcQ`adOdTW)|StV2}c5xaL}27=d`pi0nPQdZJV{2>j9Q_YdHeZO)k@t0VRHPCZ))!A87EdNbrd!Bc4efeM zf_r*7g_hHs^Xglkqs_iANDPtu7bI6nm zP%2HOK!aUE=+B>gRl}N(r`}MS0==lk7s4MQ)K-aUn!#U}r-NXQ;hy*^1=kvU%*B!k zCR^c%QNizhxV92Grw;=3h2djTgT0urS6neYWT5!d*cYqeZu)u=A4@%1^`{*aAROs5 zpwX3=3`nx(Ql(k^Z0hx?w=1(ZI(Nwx2~Q{dxl`#4L&EhXu}2qVfI0&o*9DiXnqH*L zgwv`lVZ@unw7i2}DKe~Z%tQ_jiG!1QGIHR5mmNKbmD;V!aWS))F#JWm@Iri$BR%t^t+)&Arf{+8`>+qB~&b1)ykioU6T3T~m#{oG( zktrP<45jxsZ!>507O_5gWyV=;iKqjK;^?Jk59<&|r}1g6Wys*TerilWH4hZ8lT>hC zbNIqNu((QC-K^z7{_#YNg$xcypC5rzgkV`YjANeJO1{X7CcJ=spb98j0g4 z2`|;M*xiuE-blT5h+_RrT4z--Hb$q#M=6XYaz7Cio2VOtGtKNINs#z1)e}x{m4jH5 zloyWIKrgy?_S`+^H@CJIJA0P8^fUSWBXKY9W!5WQ8k6QlegDq$PF!R9cSa*O8W$@S z8j1If<3XvRQ2R~k(}0(&)65GcHf_d}q=NBaX#_Radq5KbH8TuueEqZUeuXO%vcySz zi)w|7FV2Xc(Sk%#Q*WKeSDat6s-Ln{;zJa0U}OysLIwwejrNJts5yHD6X3wH z=OTOSKw`45cuP;()PP5xy)7r4khGX3+)4*(429yGh5Of#F9p{W(KQnhBk;>?I8B=b zNYCb)iSFilaQR1CNC>hLO+MY}GE^i>K--Pp9x#q(7l6w@t&A8**_~s=BQQG&QtJqpP@Bia^<`9^} zssK<%4g7MpQPy5LD+VC%0A(Xpm+}k6HCvl|n;VbGnH_tR)ye(`?)|gZ#IBsKwa&1G zoqK<3^UQ_CmCL=m%O%`x*F}kX{_NblySQg@!Uu&nB!%ioE6t_2yWsfjf}4AHKO85| z_s*9@2;V>F=HB@P20mY5?w!A$%GULVZ>2cz<1Q%G-#}=&XCXbNZ1?iM6-?@TW~78YRV-n*>~eEbZft-X9!V8d#GQlgk`4jg(SO)vh6S!3|fLis4& zETq0r>QDNHw;N7Ob;H~KmtFT#s0zOfea%7}F)c0f1UE{(u^|M*;()qq|zQ-!l@oaPqZA%D{0~xp9UAuRWdR zTaq6z1GLV5T=mWq*0hN9D9=gFHU~re1jm*ch+5yv&pWZXclz<}UmjUUQ9TDgHQYhx zvCw~g)#JUiY_7M zl59)D!>6~2Pvf{+*n(XX35Luv{M(y*u)`^ayEOc}l9U3ziMOFRAA|4Tb@ANh&eomV zPwvjQw#^k&-N630%zThtej@BYPp|V#h%5qOT^;rdk{B-6M%~q+v<=#rPDwtZuq&Rq3-7xI@-J@J?L1Zb{6 z!P<4^znNB4kqmqn>i4ds*~dUU#p`|P)lT`R8cYfognLSEr_?(UrecXE_)5}Y@-3tBDbEo9jf^|0?Q zyIf)BX_*g4&6i7VKz4n(0a=I-X2tl{Z9tYK7cVWuQ~o|At0}k2?vj~@Q-7YShuVROms$$%-Y|D>Bu8L*JE?2`k%A<^SkFS)INoN^3iFYuc zSiLbhnQt>Ag<3(HTprds4 zv>Tlvor|S$Lj2Q*UoCz(78|XJ?BBFo56Cu>Gs7zQ<(!mQRf~qELSg3`F&Knt!*;G@ z#WO2OJj(pD(#NQlExqZI&Vad_W{vnCG1j`?^w(qYhQ_+yg~?7f{;H~qUy6DCSK#_D z)749e_IpLK?rCD2a-{6Fk&X?jbK!pJ3V`%zY^`bkG8TsmbRy_&PbIMtwoO|rc(lj8 z5m!sA16#e5!oXF9rc(r^%bA!i8`SnKi^3dv#3TF?N_KX8^ z+MAF!IO`Ra_~IYZP?ZBJ!>ajgHccC7OhIXT)MT^}jqFTKh5U!y%xsSte83+17Bobw z@nwH}mRZuZdaK=k^NRG0t3NF@qJANKi|YBGrNb1STn6>?1zbzQ_IWERyQ+HSBk#>1 zI6&3VKg||0Qi(z&@{rQRZ+!jh$K1k@5hohd6RlVFYEkIPO=Bu-a2=F>gt+$NnKbrl zunj-YfTHU$Wm9m0bl*%uFw3D)i>!f)`Om;Um$B8B^;@_!FC zDXCQ6P&L_-v#u1smMTKFi6v{9qrdb}8f7!PL!62AZlAuJU`s&beNQ zFXkLW@VLyHzF8I+dh!1Pm|cGFOw}RHfvLGgrs_%9J!YP_35g zq|kgd#rP%Uqi#tWgxp$#G$r6h-_3h(+PQ;|F7`Hu{q_CDb&ttC9ur3=hcj}I{tAs^ zOW;nXTI{r_{N=s)QZt%b29%`h4T0zLi>-~F%_re`KG`QBS&kNoGYC5uv25B%lcK?+ zW}?6&&QZLEpN_a8eP5TtaB~rN?Wu0Yf+(?25R)xWd@gry8SXWf;WqK=$&n(#QA1(w0She=L|Uw}9R-3SGvveDi>-Q$wcZ z+klvDk_AA_=T^wzATV9VU~3`TsZ~s&#Ro>>f!y1sABimseZVemT{ta=ac*qyY;B)j z?DnRjx|{f&1-kBTO%KK9zw)MyInineHuR+2St_H*0Qqb~;xgANYI5TA??>Xb#oG@@ zluw?2GW?&CE%o*tKW)Ot9LMj2>A|jpw`R8zO!^=MVyywgg8lXM!E2nf7XEM`)C z%o;jtWmI%ZzNc%*XM)*Wl1+Bne(Pco2t0+XL3z+b50nzlOz*8_@bKYJ@gPYZwa7WEC-ck6tc{ z-Jd)%v1l!StSPxWyf9jjv-ijkg$bfeuC~c}RO!)?@n!G0#U6Vj@ii)5N-XI;^%T(j{55BHcK3Rc zyt|cp169+FzX1ijfN=OFHZha^cW<9LLjyetkJ_73F^W-0mniheoWwl5r2Dl z+M*`Dnp@ck!ZmtWBSlh>`^(~9{Sk;Npb4M;G2>AbG@{goMFP8%@^dc{d|TP)v{fDhAl?D)hYY^Ai2CGwn>q! zTecNyKg8*87hlL=n_9L&`(dkmbyIwfdCA??U^tsi!J$kFdaIXZB-V2%_|Aw3JnCAC z!W7R@QxAGj%qKU~^`OWyRB}ho*9n_Upt# zX_%Gq^Hnrp6@Y)cGZ!_Gvj%nKsqlo|U6IR!01QPJ9yE z505!Yro<8)P$na$#Zr&uC|0A#wVWSBrV~1o@mR}Hd4q$<EA}2!8wt3=FbY9T z8@E45JVSb*}aTC5Ty8CTP-=36bd*@>WB;-jd*+UQVIq^d`9#TJ~GTtD=V0km`W|IitC1oXNt) zEfNy8c-W#zL_?+0;&3g!TH&Ne@sHxCz?CCjc;^MOTn8(y)N(9Gi6~7WbjgoloZhtP zcN|qgeV+IyTe4E1Fh-zWy#T9a=h%7kq_O49QM-D0tr~+o$ zC~nPR{s=X@a=2iJL&Q3%YI})m%2)}&93o7OVH}wP4R#5kzjl+aKAw6*X$thB7GDT| zgiu>0rfCL$U7il(!xQWhU!~w$gO9mbGQngk95E{Ry${!xA!kED`XE4G7(OO7*oz5! z#TC;-28ut8eX-`)q>rVZtoqXq3J{JP8_?*=t0DH!v#Hmo-mZKr>)a(*Bs`tSMfZV( z>q}yfF313N20pF}E?G6bNL|7yL#LsH5pNRH@(y~X$gsjO6IM7R4o>FDu!8^HMxg~5 z7c+|q!(Y@3FL+WIQCadb9jKor^WpL51X#%Rx&|AMH7w%GEjvc?jnwML)RzBME_N4ASBHDyT)`hj1&-a zUtYN)ev$rAuSdRG)C=FhtJBO2B{prw6X}Ab4@B}@32LnOfF=TJ z#(PE+F#)7@_3NK~_bbG=DBRch;*9tiEl3nK^-t&V73Y_%>L(BOhbZ8{$QmAm3=Rex z`8D`}8T*4!bM^`*z=31WMfTQ##AILbmY%e!0gpU;TTVD3X)#N%rw8X(|0I02i=S*(#yu!voMtK`ASoSM?mqdsx;`OU|2u@1u>OXHQ zzhuSlGR}r1))e0Qlt<2MH#8E@I=7IF_`_IW@H7?b0~V~S1YWMxrc2R!dwaLwZtUXJ z<3N8l7WX&R?(W`cFH;mKr$$XAgo_EiTp?})5;OK7jgmp<4$|j#2*p2ZLTMOI33>ODf0mT;k$A|J&IlH>7zFw$gN@t<*bHnM{-Pm#$*bM9I&BbNOWzstG{gGy&i zwNkOPjf1GaPVmyLd^XR70*DT*1X=(@DTbv%Y(*D7+Y0CmqpguUu$Tz3nyrX7WEWz0 ztV%jrFTp1k1ztWKlphx>uiZ$F|4%`&4mfN-?^&NgG{g1qOO3N#3yFe55W_>7r8{>C zDAr=qeD8iK#PNN50*bYmxTI&_{xtLkd!>cWH;NHZP^`s#H9kEC#ajH;Kw3Nx(;ln% z5iW#cjS$1|<_Rd)$Va0@biVu>IPn24TR@U16l>`g6>4>V21YH;K+-j#pjaa$c#rq+ zsZ2P}xiLD#*RQv(f`DR;&>H&htOux79~6o;ww&wig%@emcdzFT@!C&o5Ok8WUSyUn zMbCQs^DBHpAk&Khe4=8Ed+bc8MDojx{B;6ZlU&w2@G&{nUBs5r3JBJ4r4_oamSYDP&D@jqoUEanBy)YAuNp^;>3uN}8B9 zzIlm3wA^UrnG<-Ls4xe9Is2;9UO6iUAUWL9C+FtgRe9%?%e}kH<@nxU;As>sZtmS( z+_O00gTfn&?_k zaq`4nuvI3XoqKl|kE@UI(^j`EKv!IBFtC!Ogw13CYO4v%fi) z9y_amDdUrz@hPi~xy{YJyJUFJc(33%tt*H>-o%~`9`}{GcbDGqVb4l5xF6u<-rZkB z4GQrbXEC43skJ$~+}yiIYDwe;z>OJR8MGa1Coy*tafLA$h9UmB^i z3lfGTriL=9$MQ+-Dxqhe$whm-8xR~NjhlP#NNeNYV=4QATiq@AZzA{JZDru&XIzOn zx%WVctSS`e-v5eOWAM;I`6%5iq`q(aPx|)G8%|7hd*}X_UH4MR8efJ#63zNJ{Bcth zvGG;=1vuhZ!|Xu*dNUGDM(t|K3Xybe0$NnOpLkTy4v5uLYCYBEdE+3W@IrC@k>>+0 zw}N3B!9N^;KgW%b#JliD*8zD33U7W6Ceg;34gB3VhCd_tBbyHR%!ZUb2Es${X7b!q z0BW;)?(FUY#mpQ0vX>6l0W|o_KGBA0vlyEthXH_EvRaL7Ii>goEx6*{OslN72-7Y) z!Rpo!I$E@UR-p8<{ei`6vmLA2ShvxDc{2LfD>mykG7EKdTXvV}xOmS<+`>r!8z}?F zm#N%8z=1XfjC1zms&}5SrbVQeX||D+Y@5dbK?Wr1d--`MHup|H-u=rX>nN({;HQQ= z=sXtsuWuZbW@vNkO&ibMoei%}8 zsW+EsSJG{+d8DhEK8@pQVGDLmBpHI3Vz^7#m))jbKC9k_;(QFgf7iuxn>$-~0zknW zPzrjUao`#vk@S)HTc*=wm!Amx&(rHX6Cy)OtgFL*K@ww5NJ7Cz-PIu#Of{=ZKFF(M z2L((Wa8SYb^^C)mjSCg}S*Eg?{!^9$sqIZ!u7zb^2FLoUPjQa-WCLM>$&a1R<2QVn zcRK$Dhh!T_NG;i&PYjZ6@b`gbymyVJrG#65%@nplX=CmDemE&Ctx!8p>{~yn)VVNH z2Dl8?%lzco$MyiXb@Ckj4=8Zsfb$X{fNkQ;`4$&=ZBknr%dQooGS_A5*F7mms`QqE zemzQe>9YRg9)!5d^mb+qO}`l*mU8Z3EqjqSqRo}aLyFp6SkUF-U0l9h){$fQB?EOM zco79(W_g(Xhn2)#87jmh+gHoibwE*(Kbwr#_Vs$dB7RgSb90ffim!T_Ld5}a=BnQV zlyuPgd5NC(LOCps8Dp!0kXyUu5 z;SXN8;w#LcOHYwr+=4>0g1^D9VftaNywY5R>L+Z%3CTtmd^iA$#_-I@@AyWxMnCum z4M04z7-+0&&VvxeiXqvc6~oGjPtD&m`vrR~a$u&5CC-MM=92-X(OLFHJ?y*7E?1a& zTBhKr*>o5QW(Q=~{eUd}Y?e>v{7eJH2eV>)>oy?El8cuX;wgV0lGT(O))l553PQ5$ zUP#t?iAP-!A|MW8@e&gvKP0nP`Gn9gZ+=|d2h&1=1b`NZmt$sym2wi@gNxb|7*^N=;4xdO^Fo|<0EWfH!t%(fksfShY%Q-3i;a!@uaQRbhOK1Q`{=}nh(1{gf-<9UtbDSkZ`Z)mLRU6?G@)lnkd zWEVpIM>Ff<+ngft%XIY;LQ1?S7R-8AccVv4Yaq$YGm6yEH=-Gk5)Kz=S^!a%N@62y zo3>W)-l&cm9YUg1HuAvIvK{v2N~DAo{vX1ku#C2V%gvdRBwzgFNSw&gs8-eT1x+O$ z&*eU%_43f_kP2rHKIj<-!R;~Lp=s!`s?cI@4V(iDA>?+NHdoXV6`4N<^tSL(BC#npWrU!{EE62EhI?-aB;M8n_q%0GGjwEWvb z4lm|1SCimWrR0kUI;X0lo>RWq;oYKrko$Vc6hGzhubXDq(1bC%#n-_3h( z+PQ;|F7`G(Pz`Sa!aXKQc}yIg9L`8m`YXgDM*dSY>PCY*nQF1qqVktd#h03q*)+xk za8^f)#2JL0i&!>oq)D;3&#W2W@tG)5UHa*W8`Af6DU3T6aW@Y0VImQ&_#+EqvgL`- zWK!~?mvO+OMl zJ{Z2boA{jty6$dG55?xc@}`YBksNQW8jsUsae(_S0Hsn3Ixb{U@qe@6E8U<6UyAhA5dQ zsaVTIwJNyWO}MXk)8tq?n1E2RRVQAQVbbgPYebbMQ~aWuy7jI{tLo;RRuY^MBZI(8 z#bPGK$E=~VRz^jq^*X6FFFc&sh!in2|!>yd;ZDI zGakY}aYyqpTFE`2ZPz~8BSFsIBR?A_%}4idAj9E1hu_yg&fY5|uFr*RJ)q<`>}hx- z`l>2iO%*r0N*c|%s0!INx8O&MeR$eqeT8&zA$VIM(D z9TVj2J(|oQ{+r3wmb+)15O_4;!A^ueya%D{EZ*4M-PnU#inHhYkQ9637_rjq#eLlB z+7ky%+-iK;I~cLY-bj3nikA{gdL~vS!so9!yR*C3jHKZUNH((jC!Kgj3U~qG@JnoB zCj0N+K68c$ZnM7<%6p23Qu*IgEVgQ2tvA;R0nyRL#f?6=8Oye=336B?L4kKeRBv{0 z8HDp8Jar#MNkNj5^sq*Py6{fGu@HzK6Zl}n-yWVe{y;|Jlf*3&NqnuvGLZDJMq>T2 zI3SCAaoL)pfLfIkf6RCk1&t`RVG#nv0pVt=p$>)A_&JsSjl78m8WN+Mow)CIPY(o~ksRsC zK=S?2>%iuk-YQ?_AQ&UjH2cJRDP${zZKpPO9-r)&|H(hJYX9kdm@SJ8F{;44731*O z0`9UeF0!-h^BH1P%XaY@wixwRr|kEGn_9L&`(dkm zbyIxK;ae^RhcYSXtzMRqSkIy0J0tOWG^(d-DGF0OM@>EGK{21)OxJ@V%TUQ33FjC# znMTG=jAds6-d2Y!>UkewG|EfsRFQ^R8Lr0<;Gg$35|sy2)tDfD66f#61gKiJe1e3) zFn!Zf$Gs~&W?a|gD3Mh>aZ|-~FJ<$+F;PaB%|)=Woqc+^jhY2gnKsqlSrrE(&wnko z-ocjG9(Oh2M|bJtE6_soLrw(gvp4OS@=2@Ol7} zY9j+|__Gd^aL}2;htyo$Jb4B4Atk>_chZCpjz~OoRY1?}%hLx|F6bTCFmAo--nXws7jk?9t4cW9S)L-V6?3M%O>3GV6T6k1Ln;BAGL{Z{d+s9`mvdLTfJ zMspP|-UbN?lop3;>D3A+J&J!6KLxHFafi&)BG$*e1%2qxUBhj>J6nS(2H7pA^Z_S zZIzg&8T@tcbP(61_$me08hp&fk_jeT;fPVe?|rzo3^^zD-y9VIePQ^R)L<_r>=jo` z4;ctgLk1t#Vl_Kr+$0}MJz2HJI4D3k(rK_2S6=e|>(P7aaDc?Ksn@68u6!%&+$C2e zJe^SJK7FW;!Cy$Yz9jbOf(%e+;N!aBl2y}-l$mf^l_iXLlbDuw&?`lT6^@y(!Xa^R zGEasT{O>jjEx@>#SxgxIqF#8xd#%NsKrEP0NqiVkmi4~}amso=ck@P&I|ah1p~XjMU@3%iRAP@5;Vcm4tj z9!FOFD|st}!_ntQpcLVe&0$=#O~>kLS7k*LUcf$3g&yu=$!Iq}NqDK2#qNeQ_D1Th zLlo<0(kdU6r;>0rNpuA9Q3_*;+)o6>ChCUZOtUJM#IE@+)e}x{m4jH5loyWIKrgy? z_S`+^H@CJIJA0P8qz@(DKN9!yUS_@0r7>xoCtI{lhXSY(?;FR1QbVEko6@HNFIT6T z7fNi}j3-tWDNT!>Epn{)fTkrt&3uc(eT^^9h@a7dL{U=@&*Ll3FY(zeK12ZrM%M5k zWN@H1+ z_-5h$HRMadHAQsIM8pXEN>gHp7LcCF<{Cs7yXbDN2bX`OrS!7^T)`vp79ua>!c@^c7#`i+i`AQI%u+CS`ggS}5Na8E-i+bA-gyjmtVjC** z{sVl%a1DVzV#~5f z9%{^Buw`9M@RB7reYo+)Ll~Df>K`%7-5^WxU&kndLd7}mZlNUzcDf9U8hdDTEb33s zsLkmS{}rDK66ZLbO-PCuODVL`u0g0WtVz~7_J54g!qPa!VJkn@h#da{3KbFmN%RJe zrXSjH6wg8=MJ#@V7r-Ql4GJOiIo$YYq|c&&BA7ceEUzl{bKj6&vGNJ?Ec{dPs>v<6 z&qDd}MyW+8{lpv*N`7)>&mth*1}%si9PmNvkY6S(zJkZYVl;*}AeTqedR=Lq!e}5C ziTQ%f?TB(RIP1Bbv?sj*0r8Z^|Ce_^2KCU{-mmhU(Ly(hpKgJ)SK?egYSyq<5cv~o zFDJrE=^P`jAwG(9$?O<_1FqTH+}qrEOhQraO;#uSAGr6=UK6wBy7CVO!SP;+Pb>=J z0mXW1SIhDEqn*6~Q}#yIt!3AazC^qR-xh?{L;@hU#Cw=?iSy$Co_|S1BWBCtfeaOh ze;$kfuBq&jF@Pl;ir8t?UXC*sw?Eq}P54Si3&iin0)yxr+B&Z081H|n5@Y(I+)am` ze+yksdkjAwquu$%yzjjJ=Hi^Tx0_9rI?3Sv42|RzdERTwDvbBGA3VZwFWY~Kr}x*= zWw1LJ|HOUe+uxME^6l@@M*U(%ZlWHAetay#ARyw-d#PE@gau5wdT5lm2~(0fDU*{} z(E~2axNm>l`1Y?7w@lXgzP;pLL(d)n1hDQ1nk4V+BZst}uz^9T@#?=Y&d^FFvY;9< zk2`4fI?YV)6|G`Fu_HFC=PiElqjk6cJ6$}N0&Xh+0qP;iM3Jb1p5*8clC+dWa|Pjx<*I_*Y5-NbXY~b?w{8 z!*%ZY`FuWujYrSR=kq@tNRFA$=YK!+>f5k8{e))c^Vy{2q>o3C}38sA<{}ptAyoKkY?yh!uJA>5ej1@)XdoC3Drhm{W>VQXD;6i+B$=|DrbuLpM61 z>IRk83Jx`cJbs(~SQXC>PfmG8#Vpm4ZkRVKRB>o;#|pZJieW?zTz8e6?h4vJ!vz95 zPagpBFWk_svmsZ_GbM(0o299jhjysJ>;oWUyBOBNaAAE}Kkvlm-s#7?e|cnm$|KTn zXv)(@^Y=Fn&Tbmm$QOwh9GNH^X^yfy&yo|HTW{L9C?{w2|8mJgl@Mo9SRaH~7@Ksd zx1Llb5*mZO&9$N<6hF~yK&&~HPhF`tpf4*mzU+j6+TT^G-7?rhz; z{p9X^Yuns9FY2%P=bgC5jKv%&Ic=GVhxDl(iNE8fmF)TxVgGr0op*IWd=c$@Op&Ad z+$CA=6_McPsJmgLTC#Oc;R^OThv^m<+Yd&+@N~{}#`4riZ4jOC6k{QuQ|1a0Z)U#d zK%{U9n1hNi?f4B}_6{UDDCx5};@6NewPtrz*<5Bmc{DkO4}8{%Pjh7sHPBs@m$&)* z;n483Lv5bBb^QWU=i`6kzTQuseQXb~Whc+k|A6A~4Fx7q40?-};HbN;r%;@<;pewB z;+~89JXAyp>C96^wrkZdeDd+lbBo??;18Xb!b3;8RhRW2cPDYz>K)uNoqjhyeCFK2 zgdQ>lcH=o6Bl2`sZ^NQP=iKIj#T8J-{#x%&IkS0baegDF6dpo$6egPIi}rk0Qg`L3 z?$zDJVrzXeUfai4ba}nWpQ&HzbFv~}9zCx-tKUx`cXPE=eMKTk*L}bRC?+H?o*CEC z3YM?C58CNQ9o4}65{m~_JVz_EySrBE#09ra-3MHGvmdhd(qJO0cY(MERHv#3M^yZelSN^soV?7<;V(}0tlpd>eyAlQZB zmLrE;P2*Mi_X5mTw2pYgzo;y9vP@+nJ&hC3aNl3O6+7fx)mHPWpy@i#njW&&ly_q} zQp&WYnqygAuBekz=yMw#N0yG z_bPzkrSBsMx;6N%2*zALZb<9DjKrU{7ppfudwFCEXqWecPn3W7U&i8pXijEO(LEfh zy`%%p^u}1q=Qh^)Gg!{yQTuHa+YjFP8j4~;N7_T08P)iF&zANrIa74(ou0Gq&gOLMMA^*2wm1gIDP$k1eo60(r*rc2jjHzlFGjF#uy))|w5_=>YReexSp2UHJkG(dLd%fM z8w|cMZJ{@9SkM^NPe3hZGHEd2M~Kdm&ry5++Z>~4m%{&>dynekz}4=*nN@)%D1c6- zXBkz?l}4TY;unmNh9}-Z8GoL2*BmQUsbWMv!#HsPMUURe z0GbW<5mY-gs&Z+t5nJbi#U{9EX$X_Qo_p~drQlb(BvWz$x4T-GzBa{o!d)+z(b(b? zyGV5IBxr1~lR)k!8(;YyS?5yw3Agfr3dX#BTuZS@!nV|pTG~Mj4AJ*Bfa13!ajZ(a zJ7>IBVBmSzegb43dBq34T;fu?SRzn&RW{P1&gFk$q?CticVl~UtwmSs!oG!#{o+)+ z{0F}Or@z~@-UUF@*bfDNeOvLuc3}MvK+Q`sD>#%ORrq9|A3tRjodw^G)VvVfS zuoG1xm2m?9Q~oh92A6#dylL=->SE-RwlCsK8P0;%DW~!!S*aAT+!a2%_9ZDIIWCsX zU`>Qm^^$usWYjg+K&i@?Bq2s{;;W2?2F+X36HvqgH20jEd*wFD2*=gt*t_JZ}@Po*c=^dHPE<3P}Cmi~vn=cBisTsiklV)I2?ql%QzfoO#c|4HRFN| z%;gIeA7L+&%NCd z|MN(^A^SH)UEYW#)Z+fbDvT@VC|E8^( zVO6_sj`-(C;;v#jopgrGPY$^MHsyMXR^MF)^QjYm7>U=F6A%u1pFIC$_&=rla~(0a z;bqR-{X>kIF@q;s&+Df)&u?$;oL+fdh$pC)6&!2AwGU;dkY{m01gLk=O-46a~pttutJ_CjD3?}%X*EAGOv)$g*#FS( zo6qk}u2TOst>j-?qs)y9Z`Yc|Odza=(;dwiiyvo=}ty<3^7>sO<7K5>e2BVcbwzjIbeXvT- zGE3MSqx1ttVm&;B-!0oadw{J6UChuRJuhsz;KPrStHs(a0PMrPDPu3Av7kp5l52Z$ z)o80l3qpra0X}jFJ+iQnIDgc^a>(PiSI{F13&~5iRxIIAxT>12rCvi3OEOSyNzTZ9 z*p>=|ol|Cx)lpvy{2#I(rMA!bjvAMgk2W1*( z&-Wo?{s=Z=so9JBaLKhN4j8b_@n!EM#-4#A@pUrNO04PKb}G3(f6dvQT?j3x{~~LD z>WO7M2*r0uR4+C;ll^yZpE*N?#@U|22p{53bl|IgjKKxuYe<$?WIwOg&D zhvi?kX?|l{w(+ZMsk;7px85b$vOtoJ)!37PP*dvqB^9c!YN|?)c3?$7NJ4lACIp7Z zgaHf*FB3>I37I?wNEQRbtYIb$nKcO{nJiu*gy%4TJ^Mb+JlO<>jp2ggzjVnw)K3!3?G4ONb9bIGe?7zkVreZpGQMj8YA|~zlp^FlF=FW?!sQqtUw!+_S5|JdB zkCCVOdNzh(*G1k{Cfy>COgzLtV3F+qTu{f{l_HV%e(3N;bD$ARP7jn{8qgQa@b{el zWA-lG_*;kb^lI+voyd>W&fo^LB0o{8X3PnS#FB2_&Th1YT6{^D3!&X^f9JaJ(p9Vw2+;CZkk!&5)>u{1A-DY54O^4Bxh5jMcDwu~WTQ3;JkQ&ZQhW78&7U(epv6*1lA zor7nGXN!NgKyax-T9QU+)yA1+cON;EE}a%S^PjYFX8GQ#t0Q-2E$m2MTNI{l43oBD zE7KKz8DrC(5t@!D`0a2|pSsCSR}0%B>`G&|v0ZB=mfM0^FM};@)kX>WzT0E*#$`>{ z$UHB)ry-k;)y3gmfNUCh4|h^jTnSGfezMnE)FPNubZ09w_ZNAqvrX~)SmhNlI`el^ zF032mr$aPQ^>|~ zc4)(sr3>ZFKK99y`jt73xYyc{w0c%+)kk^?HqSGoeuEe#y4p}p@2OJ^7uS=#zaNuS zwE#oF%W>GK&y}uYNP}IM=a1rBi0>9#lyA$W-xxWv=@biLUyc@b=i;qO)uK7a)GaYL zKX`n+_sqfYT=n0P+%}`00u%P;x>+pVmrJ^?NaIWi6tke$n3G#@c0vD>?A7^>LS=i% zK3#Jr0xvZs)>YuJN!cwZ(|JJzwB^2+l5MibE4Qs`%O+)1L_y@h~a1lrFIb3te=BTF|w& zk`=;hjWf$pqyvkCWP!3WUPpG#S=e&1B?eEMRwj#t`7 zS?5|K>!K&o4llGru{Tw@Z|~=JC%Wr^xp+p$Zd`fz^|AySY+veFop&-t@tcNxhjt+K zwKn!mPLw#a+}V(^vwpg`wJGlh>E0d=IeKHI=!VL?hwHs&cOU0z*OF)bi|q9wFdD1< zUap1mG$Zs20<-ZT`>|n9F*2BG19Nc^TS>gmy81Z#5zz3B)!j-m)RqFKQdhHPNbXT$ z4_7@wscqQb*TQ}@Szy&V)NeI+CNbp+CoK12Qt$6OKw(g4`qMoi#MVT)GdtpL9AdHQ+Hx|)p?j5iwqa&qi z*-ylTty-s{qn2k_=L)nklWesW%>*gFp6I@&8L^_q!;9%=g!|y08WO)WQ15{rao_4R zC45$$+ZRB*zi7{*t3mv}h5Ea5+($p-uJW1_ZS2BbHV}VN(4c}%imrCaXjIVt0XJaX zsPON;ZeuyKMt9W{HZ6Wm4%_(MF8U<%{sy+IYI(TS2Da^P0)qHabPA1Aq47|7IZ&jr z-&rjmeF5jT#pe@nDo1 zY?NbZ=6dsv_eQ|scKkscH5{wv!nw~1|9bM=|NO5_veI-q2QhzuGy&m z1$*l)e!Hak#(Q(bToQ-hxp>_`Aa$jf`cC$qMDP;lUL58sT=3(xS!^L8wXEB}qG^h> zw_tViP5abuR7$Tqw1ICgnALztub^2bYoo$13JYIAZIH`W7Ujf~btD>|)gZ#imD zR!!p8o4pKx7g5maRvKEm+Q(Yix#_CH((YKE_hHg6I;jW-olj+bM}dueg!gocKT z3nv$kC;Lw?e$n%jqw9D!7$RZrzgKOY6($1#eDP27%AOjl$**|WCx zhkpA}HcYW{hVd^m-!g{g#nfr{2AXN9D8GZ|U(Km6=-G$53%>gV-50pCA2N@p0uk{C z=b-c4gKUQ;0G0vp#bEWN^YN>wpcpuYXfdl8$y4@R{Nu761o!Ww=_mY?j#2+{PCe0Y zkXLyqdYw`S`;cSfHvn~crgJaN%q-4ys-Pmy5kyPuMH-*|=o)lCDWN#b<_i49HxZ&= z=YbmgR}Vx>eE5V>&w&klp8Z%f-uO(T_}?>CBeaj-0&Q>Mrn>7X!=x4c-x!V;qi9LL zwQER+87;kv2|T=~14;VEL-hIp@MdR4h(Np#wKrzNN;<5-_pbtl~8zJ1dxa_(U&v6>XsEbCZmQKz5*-xAoQz;I5+0 z*3@nW^r2{%RseSuZMJf5BfzP489HJ9vT8Xyv_Uy|%w0vBtzNPTAlo|`$?=xBtjO&N zV7->l_hL1N-tPD+UCnV<(Po>r+5ShbqRoD+=P}_%RP3N~!XEWf>n2K*YkELDdlhZ2 zk*aO~heW*usbz{y|L64k#A3^z3`~uXK_3VyI^VK#%b3L}?t>~kzRk8ztRbzoj;22M z3`TPHDq0j*9+9c7wc=ONq8Hj_zJII$xA{FFkJ(WqgOa_97AfitRJOHjuRuuSPSJth z{bIc1&As%(#HjU(NB6X~=k63`-|JxCNvLSsR75Rhdir?V_QB58NqGTicM}9lPtX8h z?-XVA^EdU;TdUZgj^8QD{&lA`Dw{vzdqdf0G5}6_hx=fv;_eh>1@JfX?JQ_8-90{# z&g%bZllOYs#>=*{CS8@LG!`KcwVs)dZQ{9`M58VZw(#;*y57J?ZxW5Zt6VdCxk>b# z{6xR}KHUsY^zHZcwuKj;=-Y31mLI>+aPA7@Hn!R43bq)LkIn@r{Y824xuV=_KN&0N zZsez>$`nrYOUL-`zaM|t`+xa~zFgmCOD{Xomp|C^obW_n z{`~qxzaJ~7U96h}@)LcrDs4{m#SdMSpXj%EX3On;FU~2xUM}mp+ln}4ex!KsMYQzg zV|6p2-FhNF(YH1B3;*^u6Q1ZhFNIru4VwJ(Yx`Zy#vqMda=*7bP~sDPTL6D!(B8u7 zJ-R8Ddx=mbKGClf#hd=@_KbOWPuqU+iN1NWvm&H--fq7S?ZjfefgUU+KhZZj`fGq} zt7Cp%c%pB9z$s9x20(~56VpAIg;@oVgXAarmSS10J@fPZ>I}6LpXgW8_0$T8G68~3 zarMg6aiU+z60`%_tm3Us^i42$ZIJA{*1k$gUik2&+{aX(kkV7;Z`a@Y`i!kn{jIOF zZ@ujf+N!=x{}i`+KS2N3>mHiSJ6qft8y)Knf}obBxEQAtYb0x^RFev|zl1;IC~h!# zvpRRVe0BMS`gf{Zx2q|c?eY(ghp$t&UwX}O?tFP~%Yc4yGyU^``f6qzH_))@YAs%) zZ~S6|la0^J(LV$Frx<*Y8=QLG(JO^`afUZuru%u0uN@vw;@_eOepI_CI*}CoZC7t& zZ)D-Zxbx>T9ILt6E6<9!TZXdPb%10H-2DX75HG6XMb7r@($i zr&I^Hthd^{s#4vUJTt_8ZK=l9WOX~pM**15t82Ec9^&H*eFl?g{%l~yj3)JqZgV4TGrY6maUzWE6-NH`N%kak!kR=jfLW0uHpaw z_ENDZ?d*Qx_Dki3QuS}H({Bci4h!!JJpX$2+Dk+7{2la94e|o%in-D*5a4n8u+w$D zdWL)so1k=`dDnG#<1#3;+H@KFL>We*LDte zAKtrqytup9WE~!mg}q)?Rwct*wr*-Z_c<8YMx8Hi6qfB9%=?GR{Caeovz`=vC|NQw zk0{hh7;aGIkRrT__Uq-s0F*B7x8N<1N#^E8f2_}F9>_RyY+2- z_O;Ssj0M+6v*;CXt@MkZe0JyBM6Vn1Lgx-jSjD#L<@z6InWaXUzYA#!C-oPkL>(=S zdkg2Hb>Vbq6q3f-bnt1kvl#u{JWWLQ=nX7E%rVx0H&Dy=qWa0D*W>4X&N4{9F}G5S zVW57kqu-m+iKfNnf)djCvDP+!4(Xs>TjeU^2>gI$=2IlA^PVJ1?EVHs+uPzd0-7L?;Ig7|oOW6VX zUDjRO(Juh8{<0lklmJg2T`BI^R`C99v5Sh&IJH|{%6&_f0$aPZWwe!da*I`IK-a#b zJDzSwmzs<1`h#QB(Um#JIb@1o`xFG2?&;RbH|ATIT~BxH^mMJ<3Xgz<4c?rd?l`}v zJ8s?6t+emj)BSDco7{oFt2Qp(pYL^bWf7*=u@e5mjM#L7>9#a-ZqeB-nz?68sI}>5 zH;rMHaeA=qL>kxx3xmtqR4Xwv+*oAUjx-t9UpugMpjFDttX`jyl_!yflIEQ3E2Q<8 zURajhRaFS@IgZ?Y8#NH<yxAHB^%B4Rg0M)=jLO7ZsGmSjqvxVOT!KJM)-s@oCjkU?iJWcGSpjZY){hn zrBs)e7JX>&qy4+p`H1)Q)Kv7PZhO7sJ9qzXPW@NCtLag?Lo)u90HfE-`QBhbE1+r5 zS<9hhu$L!pH#}U* z2@G!Hdw7hSUEsP!F{VEKJ(Z znImwlSv?idrOV_h#R;R^kSYI3O{i% zlZOVOMLiOKdy5+U_oF!^^3z0ryEUpN2cE0`D`W*WO@UGNqFRKA*_!w9Vf5b;v@{&| z)~s3ciS!)HrLQG)-dV1*Jx%w&BzJX$=^{Vx3>Yv<(mm2@*ROWCE^pKv!q;vMYYySu zThE`Vn?m;X6!nhh)L#_bHS5XYo|yH;a_K=3!*g*qlX5o^N$CEfP5#-^#NPyk^Pzdj zbV(k`dDhP2x`6Vg{wOXr)}r4SkF~PmS~jQg9iLNtiM{J1ButktpX;h|a|qjKe00^F z+<^A^UVByjKJ&FqvEQrBcvpuJ*QM(tY@es&{d)Q~VGnpEZA9~2V$Dt|CHbcWN6O*5 ztIzkeT`elEi#Q86$d-Kf{Y&Kgzxuho@?GfKG;!_ezTYrj#I=_D0OGtPPX$8{a%7jL znpM9pAf0*7wdoJum*S%zuNLDvJs#jmQc|RvW1BPr>Eoqsjo2)>Vi-7Il+O&8#y{zG zRsGWzn}s^2Eapp+spOg5c0T00B)wrXXVlNPDEn(AEY?egQ(MPc$$V)MYs8I968CG{ zHZuX@7_&_f-@9B~tRHT^qv(79k|w*-9@6R-cUCgYXNjR`#^KBd7rvx zc&W|7ru&MXI#T~ZF#i~ulvH%OQvGQ)<^EQ2wah(vF7^7>$+^j>)kb~AQ$`O7Hnw{6 zprVM9o(s4U?f0wmENpT}Zz{p>aVzS~wRdX#v*&t|c(hq0?pOB?@6KB!%2)8KgkM9$ z&K&Jor77L`v=qQuCYGs~i z?%tKjv0gPSbfL~8I|Z%2GAHJ3+3F9Dw`I>`svBh zL~c-`X$baXO&Rr#M$NBgyqnrhcBFKaqsT`XPc+@;nr3Mcy)itZBc zBCOS6YLm%38o7;v|Earz<7#lMTRQF(#bo9-Iuk>ni-nsr!w!4%PK!+uRpT}ZoEF(G zQCxm^v07%kJf;LgCqEa*gu0`c)?e{oE!;aeq09AXjF~2+7c*xr5S{(3fY1_NmG9A-ud7`ABu)0g8b?b{;P$V%DEpq;FuND zQ9GEsQ2bX5Gu2DZR(!bEP${|AcBrN&A1=hNJ?X>Zv~zF9F%_QU=BtI7 zkM%spZ5%S6UVq-8SvdPJGpL+Oy8Q+4)xtbAs{F&S)$<-%&QA2dZkiM|8%y(FEo9B& zw2~tY;ywfqI>XVqj0-JS4)E1NR$IS~%v^xS{yg{9LiPt>ZiP+cJ`5l49;XpYN3_iG zF|VlCJ8o-PRK!*5NUva0R7x;i(uj)jS%_->d3)!0`-FHJhex_)EdN9s^QwhU(`?vV zw{C83)SP?SK?o`X2kK|pPAl_HCs4~f_ZJ^JJUFJ-LiryTn_QvPUGsQ#sD6pP)=R9; z@a+5d_V>BgxcD)?)43+B{DGBy!I91d6Zzc@e4A2q!ZCSi`{X*^G+11a&428Y(ZNX> z*7?W)9sFE~oj!I~CrB;+$1X2`k6kkQ`ALGgzJd8^?$R^nuZOsUS)sed;yD3@ANbfM zr<9ir+F7v0*FKr|<3QBqullk%?oQ&!bG@;cF71!9i^5$!5qWiAU%DtU=#AKb%S|GZ z1OqbiG+&PsG4#5~yZRf9xuA}@J4GVz{m|iy=0L+)-ng_EOIi;+p#Cv?8E*Wo!+Cl& zclS=@M`~wq6Izj3GBmOs0pYcf~YM8wx_m%oH;gv#s6*NuwmdG0d5h%oeMy0sg) zePQdd`(HTXl8h&a8zU$!m+k)YCFu;g>Ns-5H_jh=bn_2lc6& z+;p|DEyAueb{pHZR${p=nDsK)(pG86G-?)KczZ0~Ag}2fnde3KG)&a7x;VTGFi|7# z;ZBN*JK^aDE*qYxMKGu6(pF~hFY;Drn*z^>xFV^W(qTOjoEj;bsCv9FcR6=&iHg9! z^(&5t)KAZG_rl$Z3xkAyneSN+NyN8D>|NLoFs zwdx~11)JxYQNKalEN2UyEQr?4%5BB{B=7IXBvmcI5b$ywHtKVwo1I1JV~Fn-ThwXG zb@O%5Riov%Jlq}D7irZ%brwdjo2>tOc)V0J=a{-B=4NkNNN$@^Pk{-0bKPtn@5?2P zTBLEN1d3VEYs|?lIJ=<#N%ro1N1?JkWPd2UJhagLhf7U~by3|le6O^`Djv#D_qf#+ zY$x94I#_YRg(ADFKp8D?VOJ{#BWx@J=<@G;$k?!l#r!Q9Pc%^NW zb?!Cl>3e3mZ#UOF#WuInMjKsrz}!5eV=JyaywoXy2HTf<6`FUmlC%9wqtm|rm=b4} zI~y`~)=wALHs$@GeQ^)d{7`g7W!}T}UbDN8^R#Qpv;IZ){ty_A)qXG6LV21I`gQEW zo@?RVF6eEM-Hn0@fA7{S;8osf*TuFx7QovN|YSs+NXT#XTRZmcA8}|3Pupdnp zShWuITg|0OOnJfy%YB*D`}+=1xMhG?SKPfThUCAJ_WBQ zy02+Qtf=wuV!9dOKDeib#4iohd!R?$w>nJ;r_qPDjIoYGSM}2!5%qWHxXak$uJW1_ zZS2BbHV}VN(4fj_CcwT%N8=WkrQrsw8x{U_LTxN(*65BxBYNLC^>cFA#_x8~+td3S z*siMO;Zhse_IXVZokHVOXgt*OV`f{T`W}&F8N`hcv{V<~@?qBA8;?|qJ41gC=Fm9s~EnV$nt?cJii&y&} zE(0~gS4EpQnX?k4)LtvBC66xWvvnrjPM?|gBe z*z4o(G~Y>%T>_lGjmi!kP74YlHaje91BTEP)DP80#Gp=w&ip=w>GW2iOPQ2gDT^9wDbjbEeIu%T9R4OQzh9Yd|VhH7aH>hAjs+r5 z)`c;F%zo-fy)jzTEM0Uh)hG6Te4sv!c_}v4M%q+Z8<|lUi}JQwa&2Xu>7n+gWOY}( zt(LH@HnVM2Ym-xLHkRXUHN^t`@33(-HnxiHXaprX6Iiu6Qwppp7U{gPtSUWfEJ2yh z7)#gbx&f6FV(LV_xm-N19Uksajuwwkj`nuxQqALi`nUl7Zs*E#`k1l(>R#*#_5Ykx z4E2I~OX97>4+(E{%E5mjz@U~w^4_t2Z+wj|9;N?rbb|HKKc7>d&MI7-M3h#4vM~CH zSC(k@pHqu3uhK?SGP>|_xG+H+kZ&c zG(9zWX751%wO18??O^hv_p`UE^OJ*J^OxQh)RM1JufB9kx!ImOU+uLE>pogs0aacH zuY+8DOxe~5NsreBdz%}pw@AZ*W zjh4*0>i0AU_HR*d9=_2W*he1AjZsstrA@zOPW^!XGE?Lw@~P=oqR05+{%jIQK_v29 z^{KAb;<|U`b`&_c{FeK=q@t6i>N^*$1@`7@!-?nW2(R^Xt7~rS;GJJ+WeJBDZ&hHB zf?VEzc+MV+R^L9SszIsN(o^TN8ei_ms-sTP)GPt*tOz==m95lQ&8ez)*VaB&Lpk5J zqv|U7wy!O(*+!k^ItKKT*~(t(9}d(7dvUkdSBK{pPUyCJQSot+D~AUsJ9`I{qv71} z46Xd`;Bc)*yU166JaDTiTYmqxK336DUO==nK$NE=@GBYq`QynprL5=n4xTB#%hEtY z{WpVp_Ztf;Z+MW^zIj|bebPw#E=>Y=q$_9H0`j9S0pHyu>Nl%54&T5e>f#waabJH} zABqG*roz94s6p^|nA}zWHtO`!tR4fdI_(J*tLGMiKX+=9a%LJb*IMWwkA1`zR&-(3 z8^VRD8H603f0UqpvUhTQ#As>4x`m)nr0Jp1~8{j$Wv7Xz$vcGeDJp45Ey5a5Ra^=32 z*-aFiiDEaQ{yMR&-Jig67KBEs?pAYN#>NK3xuKS|fqOfji`Yy&AWrTsm4k>N5Q3u6UwC`!*?W$C7 zYp79cAT|Q_p{=61&Rbn2QzeKX)`HUGwYu)U>C%nGNn#mSc3*7(x;Awb61)6$uB3jg z?I|*9<2_h)2vm+&8>wJZU#xETbn+@_Pp71r+C+!?D^%6huP=)nwnBbWv%b7m4Trbc zDW~|sg-_Rvq1(1@rM_NyTD)+!NH?yNL33Gs^FZBX3FCwz__@Mylt-Pan(Nl6G~Q6EvS**`o! zUfivxB>TF~3WXTTl%?% z<2z`k7%2m<+lbiO!9(#IbD1K)#MPI;SeI$&D4SPbNi7s4U_%gvf2+791OElORe@IUucb{XFjZA{q!hpcIhRi)1^TFmMpCDrJ0`XjWfp7s zXoA3v$nt%oaHVNH%DG&9o_f9Wg8DLsex(0G_1f6Oi$nCC{fGBn7%}f21xlwC`EuH9 zJ1+w_G@#&tR!j!Mq35rr?vHdw%8SLL1~S?vU-%_nez9&u;LRbv`5^iFXT88LS=-g= zh3iB5KRnBp_Z2iPm*ktihH9bGzh0xq1w#4$QLKP8Hg!5RRd?r2!Kv)cN?kf%pUTb` z1(o|%jsYF0-&~3lv~z4JM?tY+#9-Ps2~pG&^?RQ@eENhIdpiEDelc>yF|SEI@3MMz zsmEi7&*(=g^XuHF{+H@?w2Cfm;*o-7_7Y`?Ix|9n%S#QGu9dw2Ab)ja#@r_&Gj z%WftGs>^ji>xb`7uAMxq>#Cuy>2KV+fp%PvtB;+Up>@zFwQ@(<}N>4vIuThXA4rp z@TLo&(*GM!>FpC+ov>FQRp(1DHmjJU#K;ev)bkp-)I^QS2S54zRhoF-eg4YH9z_U! z_KIjA$J*L&(AFMq-Bwy#X~{q3jRtehe6g0FUQ^80Tpcy)jKKTRnC_WI2f%u{L)}-u z_=(9g!+H8w{bCY9_VbsOD(&NZ;4MB9>TD+wO=v(?zp;Gk<3;cCZ1L|}u0lfVg+@lk zu+Tz^D(yF3d3JK;xod}Xqa=BC?__Uhe~(Tw7Ox#mp5D{nB6;gA^f~7j$m5qD`q1ZI zy2ycNVN1us=GOZ88-{1LmhM`-@+=)-(sA62FJ9dy*la&`=hArl19z^kZEQbCzbrp` zw0-q(cd~yM$#G&A>o=%x8K}200#@G%+&`N(*L9XyhlcO+HR*YMTq+-~u0JhFhA9}`Ono)-PgK%Mi>;}Sj? zG5W*zz-HM}(<^$Xdi(Hg<}{U!gZL#h>c6FZS=>4FjUNauCu%uZ_kgrw>-ct7L;<{3 zG-}?n_3j2W6#+16t?D8dM?Jp%F;HJhZ2(rh8A0rSrvpq*1{-84SC3?d9?2E%1~r;{ zJzqhMIIK13YVIcQx)Ey%lZM6(e%__tG5pK~eyWIww++!!5rHrQ+!zC{2P`>Ri%XA% zh+G_u7X^UZ(12Uqy0Zb?+x#?j?eLh=I2CEeldgD~psRWB)6Z-Fk_X*b0y^z(7Idz$ zS+eHfk`J~ufizYN?r5bK$W=>w5&Ipmi7WeE_q z!&xA@=4OEyD(nN1?@7%uIQ(UU!{r8tGAFV~Tm}c6wHzU0Ozzxh(0iA9?{FcC-gs0+ z@r$6ia!Q~GO3)Pv6lEpoi-$+MeS@%`tDimy>B>S5gewvdYFD!$bPdgdFjUwFp`YuB z;^8-~UafAqbkp$c`M>-<^<5QoTZvAGD-8^TJ%r~#7Jw^}!qDJLQk!~jif`^Dih=## z>L~#qU4_i$!Bq+Pq&vBVM|Hw{VC%jnVaD#=xwyD3?M6xJ6wOdpY8 zQ@AQYr1n3HNZ0x-B146fh(wzz_5T_yt~OZAauW;C)%?!PBTaU^SKU3_w6f#D73v-y zE$;7JpB!x`Lnsbulq}ay30~=%eh#l|61;wvhT(%-_ct(%-O2-P*k2c?zDVR11?-vx zusQ%(z`EvV0UIjZ4X}S=0K3)zHqW6fV%K0w{XhfC`_w(dOIe`IZ2}?4uAdUf{vZ$8 zbqQqOry+aK)};orv1{idkgetbN`|8}Q={}+-)8fQqIF$@R_%Kht*+%+w1x`%XhqqS z#`oN_5A{0+t?Lb1TlN*7KLk&~r8Iq%f+p>T1f|;hEJ|J5vnUM}_E8GjnDqp0 z09Bfi)w|M-2C5k@Wx=|!@V*9+cdO0edTT(|F%ySmT4m6OLYUn=C7Atw9CwZ6}5`e+4P)=ddowclB^x>je=8Y=9g zwOvmAbpzDR2B@vv!$Ncu=2PL98Q7?Z??Q1d{TuWdDDE)&;wb?p-Qe;LUPi$g=R+EB zm$o(R%BbniaxEXVUv4q3uU4l=cCd&MiSY z{r!QuQL_y$HZX1J)DdqZh|~w0M7yq5hs%;g+pOQ+lB`02O^|?Q#z{bcU6Q0yz$RG8 z2Q%GpbyVd~rz}+w*`8mG%MKE~Eaog;hfh5@FT& zjK}KIRIqB^+;?8pT#}mU?AvyV^gK82>8deyZFlJ`jHFqeq;eswjN4~s4T&|3(D8)n zSO)QWVPW89=400IjnPGO-%o8TAoQiFV7%na}WMS))Ul>XAGiW))Q|>_JctP1UB;hb3cLgoV zt`6;p1+Dfz4_epuJZM9uebDA-S3^4zWmn@*9<9q$$*$(j%wm9S0AZ@RDNH0y5t$Aq z6w0nHJ%*7K6~XzLw=#Sr^TIMXBs1$7O2|)JWyTh^FR%i_JP_lyBcFb(oc6B*4jHE5PyEGp+Il4O8-P<{t z&@kM=@ubJx8ZvnpfxAK$T)k{s@kc`v%wltcNlSSLxoSJG88phKu}(O+13) zlnA6dgr+8{FcCB}XcFatEp}b53?u0S8k*}{s|_^YBZFp3(l1J+Az_D+Hx4=Z0)epU zP?T6~YS;7FbPdmAGgP`8Hi`GpoTwUxQX6OT;M{AL|sU2Y5`>D?NgFItxUw~F;P|dI9DCeylAy0V0cm_u!e?mw&H~c2It$2%(mo);PU`|Q?EPy7m86iq z??miQ7MJ6T3wJj_T2hO{JB2{1!z5m+rWFKds}P==cET*|Gpl83l8uF@OMGD@6;G5d zr6x+9)Rep79EU6%M(`=-&asGkM50!=Pw`N7J-|aXRN9AX&>cegO3^r*&Yh%-MrmvS zHR)9zv6z4)95PM5KO~G?dVv< zMIupC$aP)7BR5pqM=k0|_jwji(7)5gUm_O~KZ+J`dYa>29#4Pa`!6 z*RYXd{08VX9rLa=H^3IlGy4kLC~ zxPsHuv51p2))Bzh$FBW(xQ0sma1H8Xu=T4y#`~uhW(_APHk+DoERR~uN)isPCf?30 z3~BvAGkC1IDNJn547`AsL^2AxPhwt@ZRKWasoj+i;H-v(Jq1}-#7k;?&%@TWJP+GY zX&<)jiernHG{(Fn)|)(5d*LOq?`2Sm^oQA1GgDq^fC669*yYABl8SjrQPcLBSwo5r zBXU=&1Inx-UJ_Y0RS3Jb=OG#@?L#z~NWx95#Y-AvUJ`IBkJL2=Ole6$x0miG(33Q9 zx)n$a0jt?5Oiaxzqawh%{1`^ke85_>g;EH?ooh(bVMOj)3IqPbSj0^lYwzGqV^V8z7i2pB7RR$gU)#psgfD z7+p^Hs?wap)U~5Qbyazx+Tfxxm;|0tYYAl6Wf8fl@jnYx&-yG-BTD-~?Ur3hAzcfL zva9Va7O%?};n8YtmbBl^AU2u4A#63#g-HN2%dVgsyzEkK7)e`ngCkQ*y>)|0SRrK5 zL9rn3B7*R;h~z|~CR$CN<$1`4O8bxnc?D0Yk}evpu_@K1MaG*vQZcznIHk(Imq97g z9|BbKS(vb&89+h#yX-P*7)epQwoL;Db=O;XJ~q9Fp;RjeA*SPV6VC?;xLh@DX_YB=fN5(?SmEM6Fj9#x@ernrc@bU z@)%vt=P3zCSQ(Mx;DylCtQ98UX9i6$rMm3$W*AAuQ>qc$^+yXPSx*5rMT8|1HHA&r z@H{p{rG0D$+fxKqWlgCr8?I7pN|kda4^B){5>BaduZ;r`E(Z{ur-hgVrc{?*dJKb3 z@sw&~S%%!3dn!X>3V|6fT;-V?B~1|}i9}5S)3rGd%uwlWz_g}RmklK;Hl-TzBacZ; zMiP#kM&8&r95oSz3G$gGOE9In>{4MENyUYsc%0RDLAG635{Aq|I-yen2w@)U88$iOMAx-Q_c8Y=B$wOv*FNyBDX zHr%7=uv^k~JZ3QkNjlG#6o3IYEb9RUe_1n8n7EqReJnr5&(DyIg0s@k)<>XQTMslB zK&(YBfLMYy?r}>RQgaw1PZPW%K2j3}{E4aS10KAg(mr@msD{*rm*<+1w`D^`%JeIb zUVMkrqYY{Ies%BgZtl(qNo2)ukU%zpsaJDSn8=!0>ZNZoau-A_yQ~;S()Vll-n(^o z1K%YfeEZ(BhWs2x>S-cZL`ov#rqZr!ejd4@(%q2z2UglOq@?yP zm(<2^tqr2BV2gqo;dLcJO`2J{{REF!l2MRuQK|d^jph5d?rpHV+zQJr&rwO{8FiV! zrglZdPHIAc&9$BbSj0w@_7MyAeB7s0KV@J`3hDEi9PhEfUBLvU_cz$QM_n3jw8v&0 zJW+HCzR?592KHt z4v*4MX& zgURt^`^6WpZcko1*?#QKrSbL$?p$Bn*nW_HSz2`a>f!EW|E?>~T|03P5}86VRU$85XI`a2d}4P8kF zSK}uhuJ{XZ*>E+l&dA*`yK43d6H7BgD(E<^xcnJL(!7pSFfiz@E1;}~SPhvxjMQDR zit?+7zC^}Nx>&pt7@2ILk^&Y>tWaYJXS-ceXMrOv4;30%dxg!dC20A!ewVr`{fLvljeZP)hrk$vYa?( z_Mo&_)x+GITkF*J$=U2G)E{Xv8EWb;!-i}iM&hohMHyB^fm+ed=TYmL zo=0t{w2#_06sQZz?hhLZ6w@DO+{uG=9c94_uojkgf*GmeD25nKP@cTV&w-!2CuDFA zdrdCghKcw#b#TOdefEb#+7BZ$*dQ+m5RB$nS43Q=743c=rLOULl!i+CC`~1k$Q-NP zA2x(1rWj_u%7Ya%o#cfw`|3C%lY3_Pz10;>b77)tW|#%zawMaDtiVVrAD6p0#Zgw0 z^%R`CiWp9f>v_DohUf7bD(&O7-M-Ztms>Fmr|7sG<4hi=_|7HixE%9Z2BJuNlu=g` ztfrZv6WrWFG78Wszquu1+5V`*B&s7nfsqsAvPtWo!JR?eXgXeFWPB7`BnM~1} zTR3mB_*}&lCh5&B+?~RwYmKsXLJL8mydiKWRYdYI+ zP>G^863KZAuqh%ek*JB5uV-`~o1xO(u=#H+Y#PFn)R8(D(z!3XcB?_Xf@=je!~K$p(n}R#6uGklkRDtIaU{kcS)g{ zOG&D=1ytWv%}HS*$%%WW#nVeG7fiL1jDn(8KDlb&4w{}^g+KvcsUfLPK?W5OlbRsl z(dxQ@M{B6Gk5-gRLdx$aEub1=Qgo^{=~f=J%L~%U)#OVuYFgAfh8t4NQDLw!g{SYK8Xiacdfqb{NsSrdn_wvnpaLk%d#`uj>OIyrI(Fz-vvlk}i5{ z8l7rQ`;|v8zH>=B)tZ(d1KJS2nySJ?*394w_z$ZtS%#5R&VPs)x$li@NZVngc1jzt zO=b>k&~?D8BCZk{H-%i+{5*0)rMn^5nrbCoG;*WV2k25Bw5#Ye^&Cox#3|Pwm%>x6 zP_j6fAzn3Ig^8z`;T24^uDXO7MpF6YDr>?XPgP0g85PQav9u~;ERm=w#JUdP5gRJ) zBNpsp1yil0kcMq^sx{$R9=Ox_V4QRsHX$~|uO_ZAsF)dk0c&a1rOwDC1*|38T+V2r z%iRFk4>qJi0uLkcG^=D0Yl#e-Lau9i9=V~?K62Y_aV^%;YJ#=Ix|4@1zOYG3Ok!Ug zPoprmLyT(L3KK^&!zf_yue!t;Mp8LzDQeq3^J_@pVPx)#Qj}jstR*sS3Z<^`d6b4q z`zX!GuNG@*HNjc}Ugg2s4{Hg$nt^N56cDqgg|&n*>yl@f`0s#OYZ6hquY{R3B=ImZ zcf~B~KZ%%2tJ?j1^Vl^$kJ(UZAG7T;t2NcSYM4t=l3{=u<4_*19d2flG7y1WaV#UG zuALSJ6X-aRjDo|SypB^iUf2g#Nn+vt6_nXblgTO5(Xs5Bh{n_e0E<=6{47=@O8Zz% zCYA8C$C_#-h4lV4*>(#2%0hPS;=&DTaCta)agfzU3f>e4VUh+MYj8p2YK{w&1x`a< zye6VHb;z;SlwvQq4cR`7b-UsgW`kJ_Fg4K`(#7Nz>9tL~Z zH4*ix2>?FFy5{Gx8Y=B$wPTJo)F)Yvwf)LNcATGD?H4e>OskxE!T(x<3N7p}i zbgGs8Mk){1Wm6E#o55Vrv}O;-#%5t1tjR9S>hDGVD>Bu(=F)8#N!!%H5p(z0A0~N) zD4k|*6>*(N)Knep8lOjLsI-sLRHDICt!swx#1zAX=j0NLN!A(R742I^45!BRd{%W0 z&*L>z+Q%!}v4yQBYpQk4Fr1=Ot&B5yn65G~f)awNm4-?*<7q{WhZotg0Y*}oQ8lH7 zL8nuhO>+PMOtr4Lq#B8&sAVG`e~s9)I@QWKiO1&}*Vx6DqzkK!9rI3KRA-OP^q>b&+cX$xLdI=V{Aw`D~Jf0Wj8(_5e6!DVCzA3P} z9^kGfTQZ#3$W#5l88LI+xzFb*agv7iGD$ z=N&5vdNyD3(?V1F5dc<~C&Mf&k7%&ow{=eg)^QuK_PCi{4=Fp0*zJ2tZh-4TcfrPt2l0x27%x`EleeF zX0z_nWf&BQx3}Um8@q;aCSK9DmV`Y8Syn_;B2iPYbuG`sHdNY&E!xI~N2=DD&AK5f z#m;P4C-GRtUyu_X%dqdwD7Ya&HBW^J` znOQ@^4kNO7O^I)G(Ui@)h^4fyz0dDoUEA{z4VCsGnoJ}iPJiFZtcIo3(y2UBF-hsc zCabQi)#0)vtNP(Ijf{F)I7$exE>(t!{(QiiiASf>%%v%)QW(W_7L412t6I2Wm!U{UZylhP9OH zGX-wt@w%2ZS!#F{8J(y=x;Wk;Y&DmKiL05x7G+nLO~XivQdNHDwT2OGKts@mq#j1( zuAoKPRm5ak*WTwr>)M_NZK$*l+WhQlm`tMVYJA3{72my-n|aNf<4A;5CkRtbZ($;7 zW-vwB)g{<4lG-Rj_3J`@8n!1DAy7l64$Lwn8 zO`_~-dyB^_W;x~NPW#;qVw1*zu+>BtCaz`%ThtA9sWyzHV%WCm21iDniuqOu*^uwU z2;3F2D7T78PV3tDJY-$V^N|kX1KCY^KKhEJQuqvk;9a?L#z`NFq43 zy1}H7rZmNCVAiQDQaA7yT0+QIBJT5Wy=LS2r~6go&dmBGNgtS71t& zWEAv;#KWh{so_&cFqpQ$r&md6VNu%^s-S_pA)+2NF~CFB^#BjmP-!2ku;fEis-%nF zaK@%olV0T!y9#%g!YS3{`!h;lm{&Cog#qr&%3Ls|y5SOG7)epXMLfX@%^z-6VbvSb zau~^{iChsEi42^odR-Ur$PJb5hTKnD`PI;nVpFPV*YVI@gU}U@oTdfn8@rl~!o=0g zunR~I8!jJ)k<_W)5s<|shln%G-UPxnB<3(ePZPExHWJx31zXqpJZwXyyTR6)Qr$3Y zq!_;eI*LcEb-|L5%m81Mfy~H~0TZoeq%hGmGq8dw)eV;u!$>NgQe`dP=Tsp?N zZK$*l+jiSpiXd7%vGpl||~N^&vPR2?@N0 z0cKk3FhfJYYIX{fr)HK>5nx5UB=O!gAK7L=5$#+zLz)g_+^(fCXccdYxJe{xBB#>d zuop`nu%XgEV0}2^Ge~Pnm2_!^RpTZet4n5pEf=fiH4Ii!5^cF_-PA-CCX!|bQ_x}} z8SPMEB+cp^g#&|9v_qhVoE=7J>D#z(NukUtA}Tfh=QFEoeIBTx(n+8ulaBf$(;nV5 zY^7A6DR3)~*RjECMxQBo5rf#IqHSTTX)H`!%?!3EyShXgMpD!a&QGY;+@TFI2-=X( z!-zc1>?$HLZEElHpmlA}gEmy!2Q5rCe(F_rb~Pj>Q6+4A#-nvvdIv;M0Gp>X5QTb% zwQsX*z#>?aTbM|i8B9@jb?G&Xq&A8W)j>X`%6O>S;6k8=JRe5ru0TcERYY#u)c)sz z>RO)%YN)gi)O3ys&}wH_Lv9jfS9?;F$15f|rRw(Ke2ZH>IXjN!S^+n$XTwL~zb>ZVJNVI&n#sWz51 zeBNFrZAeUE(iHES;|PXHQ$$H3QB#{x*XBGhL#4X`)0$G#X$9vBlY*VB*C2ee4HR{w6FcI4L z*7AT;k_F*FYgDW_h!w=CCZI6cXl6L2u{Sl!$=-akMd;x ze<92|JR}~oIt~D4T`vI4hKl=`ZCB)e(uS+yBE`m5lfDJewJ1r#A=TvD;)sOAs~E!p zBKWj0lYR^#*d@m>I1o>#Lc6!zqCyCVTpdRAsY2LcDDe>1fdC-vx&c5qRJSJY8&Y454Fz<$x9!BD>_=S1bAu{p!)&2+gb*&HZ8!GPO7p2+oeCM}p=wdPx>r(*Q zUf4|Riy62g3t*&bLJJd2Gpmw+5Fyp&)G(ro`Akvc_L^%$ZVw}K*BTk-T8GiZV^v20 zz^ZG1fYnfOAFIj45(4%QZNOqy6L2knY%jDX@M;FMNmF3to)%scM6OG)VNj5dTx(uY zdUC2@mJNwMjLcn;3$v_4Z{m@wBLE=RwLd^^sJM^ZcDvi(u|aDnPBAKCfE?ph09xyo zCL!?4u5@tbSBDuhM2d##8VD2C-?C!AUXBx0)EkgxBe)lwEEP zBWixWwMH5HL2t+rVr1@$UYKtkW)zQJ9RUEnuKfXeL&bgc!sO#8W98&q!;TW=TjNy# z+s)?KMLGx@gAbJ83b zznVV7B!rpeThv2datS<)s4aTPkqSeVUt*O?GLaTD(+*|&!UJQTpr+FP<9V_$&jVu zUE`)UTXsT_L&BUkfd7&y!<$I<{EzskD|9Y$2~aO_Ig;aHeqLy{09 zFRJMUm9v5wR>qE6(%uKvv1@yP+E8&HwIIJlzT-+$c+EV)l3_>1Mr0X}0!{=E5+6G&79SjW9e$T{aFQDvEjW1Z~9ZAw3yyVM9g`Bd`F7a7-40R7Qnb z((VT!b&U@|8Y2mV0mSFx8Y7CX{9d)9*nrbxAjj zsN!+i#r4*D5NgOyv7y?%}F>l z8+m2#xYYC(CV*y!OEfjRfUo>_a{mWB);Mq0b0jFKdGzcAdE%BW3b)+91r z%L8PFiu=d}dojt>?2@52#inMXP6Ys6lg_K{za%d#mq7-nwnj5={1a~;;C8afN)!i@*v)_hGZW`^r;ra zGL{qBI0<3b4FJNS;y#3fZZL{&Wlzm68G=)6YBud$fMI;olW?LoEkOpraT?Yf7bd)B z24cvXa9K8tsA6VB#LB&IXhXUWBlT4AE8{sW=@Hd zLJClfPI|;_DBB18L@lhFN!l6V7j1pVGR9Nme}G@t`T)P7;y!-c)y+2JiS*IXjWM2B zzX8zp!gyj|%)k{@9vG>b@A66`6)~R1F5`v~Rm^yb8n@S68xnsQnY-4=FxSc$Ph{NG z9@n!!z-p*?607M6z{nDGQ$3m`j|&H;eOGr&x0hsahKxtjCBgx1V*EJUu$x?x1k zN3NN#lwXL5SWn}S{KLrH6}d3W%2-b%Y!bPy{Q+`A#eL+W924$t?J1tIVLioYiUF#O zR{>}-^GQfk1g~ZQjjVw2s_8FGK+O!Vs2??U2{??Xs3y+pRfRK=_L(cLi$f+*hSUNy zE0}9#6eu!p60xoa0AfSMeZ;1dk6$?}dx~dlcu=XPRNz1WTug?_Xi5byXP}!j2ga`^ z$1veFGyKAQ>(X-=QSnli%#)XeY-^Q}wTVMNW(sm^R;KZp(4 zLyXj25exIJj5amaF#!5dt_J{OL&bf>CKF6pCOi4o(56KB)^;uc?wa(Lil7*_U%*j)RGM0HX15}Vy&t;$UET0xe z6q=ePAq5x}PsT>f-XjH>gcX*-(j7B5`GY{ZETcj-?&m@38J`DfL~$RaFob;sD|>2| z^wDK-Y-*NsDUZ_Si?HK`0JR_-spW)<;~N&i%cq6%gr;UmNC8a6Q?rfv`X(8ZoI-?7 zGmS3G$WD#t0YY871B8Z(cS9%@_s!HS>7x-Eo0^Sy6Mz%boP<-ekypl{2Uh}&OHFTK z>adxWxM*s2SwwIWPtAH}?Q>g7lFkU3Xj{50qc)MSsRJ9&@&K8k;yyCLK20(;OG;^Y z#-?VYP6YrRGe}BH33ijxN!+Ng7^j-d@=D_rP0cR5bQ(rf@zm_vY^G*OY9VH)TJOq8 zO(bj*v#u8aW<$k&%(km{?Wx&iLu!gm%_e;dpo^(Z!in0X0L=P;EP@fN$t?^loZy<` z>5W2Dv&$~Mh7nafH4Ba0=aw}j`!J$AR&IZc15a+2Wh^JMaT3C=8vuku#k)b+o|+|n zbjFQM&8D3TFuc-qV5J|OO}{P!v5{Re2C?S2FyS?`av8BEmR*(&BdQn)VtPfyLuTg- z`@$r#Fa@`Vtb1AwTTU{b6A7Eduj>JT-%#;x__e2ImkrM;Mwo!^0?=KR5+>ja`T=XR zieS`go(ltonV}X<%`Ur48%9*|)GTZKUbnj;>4%ZKYaNt`7iiL&bgE z`rY8<-NomW{jJ$$!*+^I%_iIn;JemfWJ@OKRq`!$wj&39qK-zhz(xz$CCClXR{*dt~v?rvqQr)3=h0J*OH0dhmdedPMM#3z~d z)a?hQZA|VBtHml7P&Ne!=1xS7nu_Upu zE|$KIqd=j3?TU;7)x-dgSkD7IVk3(Ch)pIKJjFvw>0Fy?N(Ij4fxCk5f)Y;62HCbk zVO_<&R^3VyKYP17CzI_mI)-!a8VGjvs8U`zEwm_)IcnH&on5FHB^9GIpkBNgoYdQNA@^1+ZO`js*nJG>`8&W#fFSDKks~ zH?tlT%(o)CRDQm#OyB3;b|qvFF$V5hA;WwtqfL>psXi3_6??@5hz%9@5$l6eka6s( z*%d>Z66IUlxd6EMD|os2)_yqy>ty+X@vG@GOnA)`2;u3fmQCswoBNJ6UX4{ZO z#7Nu~yD;0z=u>3cBz9fv1MG&1``889HZp`~_mEc%eJXbI;u<&43LuMVQ{o;n7fkW; z7+C-#RkLK6V47mv*kugB%Gl-UFrv1oj3dVFHP420Ax7q|ScQ33#*x zMpW@+?3!RQHv5#8WEJjbyP_6mSQ$HtgiU5x*Y*Ilq2fMjL4HYmYj(x3qheFDj7I@j z@eNSIiCX5h40e(B7^9jC!vy}!Fp8#TS6ns@BdU06He%WyDaa)3X#lB=3Pr*uLFyVG zfHYLR8%XV`*%iZricQUOE(IvH=ug5aSneemm~5R&HRXkg{h7fOP0g;jq#H(5@ziW% zZo?gR3iA9k8ghOZS*MytWn?EZY7(KY-2p;F#eIYZO(vLGsJL&YW>*Z^DK<45@g@K# zra1|xW+Si6fGSR(34+tiaEYd7S6qS(6YJswO!0WFXa8#dijP}Eh7ThxIM8KISe>;4 zQkzy})Fv`(5}B^$0Ww3yePn`dsbp%Fl+vd*v8mapuK++Xsp)QYZu`pq&hhbZNi7cV z+&{c>>4B#vM<fU z>$91fC8>q=u6Tusr(HO5w<04ok+4b3x?TX74Hfq>i;`!^j4}X_`Wx^B(V^`(@epuGM*C&n@Yi+2LOIU#k=9xo|+|nG=5`*3Fub< z-6aEEAv*%TX%?u}JQoHEGb^6a)a5F-Th)vg1{yO% zF=9NCkb(kQ%y_cxJcX%Ql2(Y{sn*Rh##7^eP&d2Q2lx#Y_wn0q>uWQfRx#s=^(g>t zFN`Pl#SC1L1#p$D`7R7nog!|E^e2TFPpdBDh7lFDX~k@)sNo}d7`B#4ULjT;^CAV} zpoXpGRT<-ngiYpJ*Zu&jq2fMPlZhqlRoRRuQcCCA7~=`J4?q@w{Z2@N0lClWS^T-W{pxuN1d za#4;6L2FOVt{T>p2(-ql0JN(Ns-T$R?Js7BoQrcKEyk;+zc2yiq=IQ?TG)?r2{;l_ zQP#?8Tm=)DIM;?uAV%u0h=u!F83k%p#{j5{T@L`nhKl=$O(vKSw)Pays^LMUno@z! z0C2_Mz6&yHACs}@%n9RHlVh0hnj&PKtb!|Lm!89jnxAj20Z3=P0*e4fZ^#m2WbTSy zm~UnDs8t;S0KKmL0eVBlee}Xa=C`bzso7OSj}qlu<5d7#Oq0q@p61mV=^W!#Q)ZZe znpu4e^Q}wTVMNVOqt0w&KZp(4LyXj25exIJj5f8ZV*nu5^#DL@sJM^VuKCu`rbPMH zb}j%ercY&@3zf; zFx$%LQzUG1ck5aoU^i6U$1cdWks&;%hivFmv8mbh_EWQ~o2W`%5KhgqFJ|D1EC7+Z zc3S9Cs1rp(3Ql`kRK^jcM*75%p@ce7B(JcC+!d=R&#uXMQcVQ#dDgQ(kJX6cK2}qS zB?4HxhfGT09y04%KFhA*yP(A7G5cx;v`JH7cw?p!JL~ zs<|)>MrYP{LQ}JAE*pmtRXjBtF?){`WD@o?fK*0>B4Lvtb&U@|8Y=FC)bA@nO0M&* z*)_w0icQUOE(ItZOFtRQJ$t4wJuQ?cG&Q^Cl5QB7il=59yZ22pCOL%&mG-olK|MI< zvnC@uk+4aGx^@Q$4Hfqh3X_Zv--000OwE!$I*rDrW+UDN;KVd1;nZy8l^K}D=~L5N zm|%ATXo|U1p{dz5mtez)DjtUQ%-ZX=G-UWN(oS_-Dx)@$S(C_gEf0_xD()i_?8PKg zvulRh6q}ljIu!tPg+WqUN>Hgvzcm{b7UNWtS(tE|*;W)y&91q08b(y{)a=G=re;ZM zVKxQFxJ&|dkEY?KR7PqdVUw73y#O#9D(+*pT{g9+X4edVk0iHTe=| z4S-q$_+tcXati~CnIRZW&91rh8b(y{)GTyBpIg?D?8AuOHSIzO%UDii<0OP#HvkBS ziu(}uYhJu%wWns+48bWjHJf%W!0;M8!VwPArX|P#I0muixG>>0#R}O?C5Sb#=CW)U zQN^r@h?RTa(1vs$M(U~JSH^Q9^Ct1@dH~=zRNTieOh!Jql|41PW_V69!UXgyfG%b_ z2?-PM<-Ma;^IVvqn%O27P0g;kOdCd2@zgA9;9j@8A?b&ayDM(tDXfg`L zYcA)85fyDD#cU|smi$C5#y#dL(MRZS51Fm@OmoxQR^-NhY>Za9~BH%_Jr7w3B*Vp2cldG zm}_MeC^Bynv91RIVnfA!#HN#vk7;F3&8{0BRH`WzI2QmH-vyP?lnP#+n~O1iH93X} zuhWrlU3v~9syKJ$CuTDZ33s*YAxnsnxhs0nu6A8Uk0N1{=ymN6&>Jf5qu0k~i+pS7 zQKBN*I19jbx%K@LLE<&$kbGSQ?n$kaCh5&L+jMhcy}wKPm!=m?7G$m*bNo;u?td9 zWC+jhA+H>}!c~-`gBI72p>e?S*HB{WkYAUfr0PFOS4OuE)9<#0mS$2&`*P@zm_bG^b`sR$+$iidwjzm9e8p*d%IQ+XK{w ziuB)Et8#00zfu{;m85N2wn*^zAd;ro=aUZ0h2#k!j+uxd9H!P^w z)GX&VfYOb8lnTF1%L$bOQ%!kcLTP3&MN_luF6o95RXjD@n6Gb=G0Axv(x{B=M8YN! z>e?M3G*rABLaB+?OwFzvvQunoHsVbH&doSR@*yi6w2ep!)8|Hl;55bksM``jQ?n$b zfZ!yan)Phd=eCq2oe?ro`*%Y|ZEAeaQ=2@?^T>=S?jsXy^(0fXq?Be!#HMDWPUQi* zaghONT1vDb6@H^ODlEpSCbKX}X=eEoP0enI=uG0N+0EHZ&63nY%%TEaPz%Gh?}m)j zM8c-BspkcN*-&vGv+e3#duo>S(Qu7T%_e;dpu5CC7Yd6|D;&a2%8U`L$t|xmg3;9M zhD)zuL={iXLW_ml(rjKLi7kY%0H#n1)K}V&v7AWQB!pcz00@VQcZ0A!HB0(v2*;*o z)6NAL9_LrW!jamvc=2TfgIIH17-Y;W4I|dXhRd>HL>03pB8KdHLrW41@$0nQ+vj7P zf@M4>5;lon*8>2*q2k@}YfsH?7@kv%FaiAvpo=er5)vlh%j3X=D+Wfb=D9FXm>Fu( z)a-`Kv|&UQPtCFh?sdBxl71Mu+e6iD55VoNjO|1QPU6<}0l;mjxQ|<~k(5l$ZWy*x zY-%>)UI5>W>!NCTQ3);lMr}fDjAG4rVM1(XC`OE@4VQDnh$?10+4kZmV`0UL?j9#; zg?YCteqr4#V?2?tN&LFj2lx#Y_wfsNlm20?vKdbsnDNB=6o7WsoUjV%P3&7Tl56C| z7b8{kU6}Bn8B)>g-G>4JcEhlqVl>47ImUAUwD?A-T+o_V zGoVGa2F9zVzc2wcvrG&7Q7!?85jCqH70xzF8$>wQhD;zv>aK`|xmHGj+R!lo5bJsX zAU0IoM{F{|;BU=t7#>urDHS*u0JoQK-!W@OU8%)~;gjaT_|@bXCcKg@4Hfs%i!x0JTqoZedXy;N z8m|J_Zo+&Y*>{?KaBea-a>bJ|#Ovm1p-Z8BOF|0hQu+DTnQiO^u_Uo@Z`&2IDBo_% zXj4rL@cGvB0FT&+;yz-N2_}TCeMQR~_spf*(8 zM=i*(k%2sWYL@iT88$XG%Xk!kbr~f>crl{%Bejf7@wAMytLDNmfj_f47){M?x@;Uq zR8&=pr)DFT4e80S=|sW`LF%l3NcwmVh7`t485N3zO@h=lJ^*Q`csG#RQ?sOx25D?+ zmUA0GDZcJWI2g;lH;zQOE?}5y$_o?wPAE+?m7=NHO_y{d5fw$fc)Yf8f?*?Lgod0S zM%Fm+oC9B2g2OZ_BRi2%lL&R~4iFkD?jy9_F)TIFnyJ}MLw1Ty%|@IAz=MH{PL3yS^w583Ey~D^B2b)|9m@j3tB@#A?O4rT+m7(H3D%)*4|EZlX z4Q+|>9)KUFC%s#p+rF~Db9_8}vwGw34f}^zEvW`?tGMmTdx zFpQ{9>PGqSP7TP<+G_BA;EbS({#7Ao!|4I}_= zjOKr`;b!PYtZx9|FsO$<{+e_y0P5_s(pE>jL{W972)Z3$vA1{7*va|b-r}ifwyf(bm{r%{sXZ8=D z+S%VeKABvzUv(pErmq;p?X*1}o-OP(l1Z|l)Y&JGu3R3@UA%GOR(1aJ(enpac1|Y4 zxp$o@{=bKhPm2GKH8i;^Zqa39(cASpq_)nVpC6vNH1}IqZ6)VfC0}B6T&{IwUu~&1 zpQY9ps89Dj`8d_$SD(MScwGl#BvSs~=e^(7j5pg0>eHC#4{xYnG>YP2dpzPL;|$qw zlGHVIyZ_$Zy%&ZzP(1EGy!XP0lZp|>`lbMf=g$w{c>MfR!}%wle`=)Hjrh|SUuxv$ z6jPrc+L4~#7oML}uk)={KE0AYuvPu$QOv|?RrW(eedzGT;XM6o6c=L@kAGUs(r=rD zZ+rat{oy?QE0muU3;&Sh;@r}_UhS$^dlnqdKe)RaKCO6fG~h`YCbZ03{)~EEvF4H6 z-F3Bxb#|i8RgbQhocN2ZrIKCEtTicEYSm}y96HvkOSo!&QrT+@`yTa?T4bFGY6WuE zK}+U&_0jsddiAZ*JMR48htv%xN6$~3g{TsGJgWXHSZC*o3w?<-cqBm-JFM=6*~+cV1nt&Ad`aT=|@iHdYj1uObv9 z^T^AbOKP5;RV{+mQ)@L>v=)5gBdQ6RmP*BPr74~>@>htIi3aLY^^kG(kSau0MupGP z0&J@Jo!#AP6{&M#{^>ewEFAVO$d+Jc# zYM%W5z5V^++2Y?LUx!glfTRx=Zc}RSV0ZG;_Q`d!!`#Jrsx97OUbugFe7yMhWanVJ z_%m9UOTDWTNTXfdF}h{zCbC-DF}gGuxTXU;T=mxm>WaESv+7N(T;*N~o8S()3BF^X-pIb{^4|Vrc((ZWC~2orLSmrr9Yru)CiI=` zUE7}=m)5yK4a$v!ngACiw~e%C+p=xJuPfhxLLta5*kr(1wZzaEjp}>q4W4VBm^@SA zfqpTZr+-js#&hzw-ZE6(YbL#5{ z>N7mu%nRwXLdW;TrUCH|^_j!BmJJB@=L$g=hFZcKw{9qRi-S2=EM`qqG;rXXyAi2b z@kiylM1Q_`G}(Rr%E{j00iChz?p>K2kDx-S1#y0%wW{jJfy7r2cZcWSzkhgm^!VZK z2q~73yN<>T`8bu;Hx1OAp%YCfsI^guj$0>%)~S?Qt8e4PI@&pSX0mvEvU7BzL(@sn zOi;CVY9xNosI+9ell_w&Em?b`g`n}WD5IbH)i8ZDpeb!OeAR~w(~TsS#gUkLL}`Ky zgYyN#T(Nz*ytn&OS)Nd_@K!E7io$5II=y;Txx{D4k?n*38LXvN-IjvfNT3}DX-S(0q11Ms$v9GSTGGkbdp zV)W{=VtnlInI{jQKDoMMwN?{isp~;aZogZ%K=9jdF5iB5Fxh_g@Wgm3_3A4@sxt~y z+Q09mYr}c^SEyg&*^LnD&#M)0P4izut_xxm9CXFc)5jwEz5E_QP@wv(5+V<|=RoW`QYoNY0tp-V3^}75VAJI3c z1N^^3UQY---)L53uGBl5UZB6AdNxiyhyHr=-jskv>*t+b^lk0!9MjoC72ox59wDR; zZ?b-qv<_)wq{eg=Ge=j8Eb)hT~H zJXighM4Hjbk85?`!TtYs%Kz`p!D4=k zy6Mu53H6xVoyZ z9t5?f^FnW}PrGLH8b|8l@W;dhFX|nosOoFNu#AhNmw$mkw>bhiC`rom0<=z<2ZjBi zqCw;~V!*~rePMu3+YIKtrc;o=@BcpsYSmi`f|qR_9q#O2Azu~?Leao~@aUQ2;p?J* z9Mnz0)2af|Uw`6%jbeTrfbS0Vrc1XBZ#u7Y*{25HZ*|DcD{5@|zzkkQt4oz0{1V;% zn@bH8Mz3`uN;eu)!K7sxsGk$aL`k>Gyb|e0c@6|iM}FY}dnT5BX)86&>_Jp?9a=Tt^2f)EqY@9a9=$Bwj_B79JBj#WwOTBX!97EF530 z2*s}+De9#2`d1?^P?Eydb}&LEL->eMk7*=1R-WAuVmaxKrQv|(5Z#pk*y}mvqn}6%g#Ua^%{%KPP%`57#EePHpWEW_+ z&_jRh0Td(A9LMmpY7{QKDj`nsZyW7Y6be7i2rZ2a)>5t5- z&A3ftQdELHS;+NBvC7@{r1vLWjK&^T98r$mYzMmYw~7;{E6ajhfAw=IzjnoOEj ztvE&R;x+11MF7KNWU%mQI<4Khb#tB82AAfjzha|IeOos}yJC@{>+@7zu(Py7T860{!`E)Fp@;;mc};8VmJ$FrU+5P~SGE9_ho9w2i5Dac$8wJwM8} z{rJ|4@0jbf;55Dp@=0GtAN3cNI*|M$K(rlF-!s?K`eph6Z6oOO`iS4?NbJ)+Y&GeZ7#`o$;{R}tFwWAEv#{x zm+0w>xm}hFOSi}MIYSrq{LDz3;i>N){y}dhXsVt5Ogfrow4);D4q2U~zOUQTZv{|L z4E1Ei;t-)v{A$4Y1rgWHreo@g6!oRusz%FHgf+cQU54Av$BHff6Xiz!@M<}lbd|<# zM*TuGy`ODNrhd9N8V2(KPYKMa3o-T2=j@?3A_F}!JX`#`^WtDmK`CAJ?UV*O3t!NT%Uo6; z9)8Fhmk}-lPcfans@|_y%gGQeSZQ4s=OT99IZ3`m4C+)^A`<(nYy2AFOa&|IkO9F4 zs$+Yqz(OflbZOx#*Z%y}87H{ry8{h^H| z)aY`u2`BFumIfVqk%GcO*O+TUG??V_;o~nlIOTahbLzXhZ^~?-P#iSOrCzL96^4CgX%ky+eF(eMU)8OKE`z*E$>lE} zDHsGdHoqL@!loCy)N7ovOLBYKv8db8?0SbT(^paVwcUz3S2$DDV4=dTIMuN#Y6R@H zDr%nEwJI_XDsO;DtzyQYP99J{(XGjAq%ieF?_t)Z)!PUd4Ks07!SIOHcXlf%8tEdv zrY;3V^W5$Qg?UsFXJ$?FyS$u*zI~t`D{7`{6Dcg1uhvc37mC|I52D=2zPK^;dA{~b zJ|mxXhs$=;l=@utG8z4!UMEQy43eGxVZ2 zzc6)o@ZmU6zu%w5-5YLt^y7bycK@S+dZMRDubb?em^bxsdcCF{=HBx5`*ibOPrBz` z12lCbg_p;iQN8Q!Try;nKG^}OS3vK9`i1%4(yViZ(G-HH*QI9EuYlEkO8jE(+>SK$ ztfGEzpgyN36o)qy9aEPRV9i%Y>Ixeg9c?xu-aE zF0bH;102=KyTE$lV-R=fds8Sm2-FzhganOvwqiszEV;hGN;4?KQ`AJ z7J}Yu`(BMXd~I4;^;1Vv0ao?_^(S&p@#BGOT_x97h-sU(@9xG$)Yt+S(WOaj#YHcD zA=0^GJa$+70}tEPD}Spt^@RH}dToQT_51VvUZCs|k|ncw;o1Ox4&(U`|9X zEC}kor=ehH(|a5h43pmdSN5-rVI49_+RW--PXXX1$6>1B`b^)sFIrd;86}(S)ObYw zY|dRDdXvDU>N-|!x@-S|RS<}#SM9lhDrQTme?90wlQ)+Vw4W-FI;x3>WH&;6%Rqe~ zJHCsl$;j9}zo}`r|N8ei^DNp%mTVR==kpq3&gZBP4j=DE%&}h4jg&*{;`@7WLPN-T zU&{)xP$WXeX+QrFH@*(X`Gu*PpKl8f)c@U&g{LOjWbU>S7-m%yTSQ^@m+Z?}4`FgL?V274uLBRvoBa=826{Uh$pI zoT34b{9|IHl0KlfKK|DqzWYcg0qJF`Kme77`k{VeJ!lEHFIf!Pr8xGeZ$#A8jozW- z-ms8tHEC=Nifd5&Sm7<2OCJHH?bIiF8lBV4@sJ5qd>N|`%=JcX`Nn<%J)pJp6%4&p zAl0{wTbx>c>%6+nH18gy8$GCTQC*?cWr(vpxU8D1#W~}&PM|u_q)oEWB!k9kpWn1u zKde49d?K&S>aJ$yIM`fqCivi5&A5DoSrTYj%_1A>Ri5Z7L%i!fyB3s%3`0G8Qeh(;te$s zqW*ZG9?V_o>X;|5H+PL?8>cjI_f=mz+uQ#!Hd6oVK)p|@y&bhx?Q4C9+WjbXcJit% z+PEj@!-k6PND~WCDiwZzuK5U{_(9fJq05Cx?^y2=ECq>AAf}eB3+`72i!VxB-$4AD zIeCCH)}R2_A4O!;**iFy99^C4?(LjR7I(|XP_??foFC9ZBsgCPPku@9LiP1F=ZDHq zZL5DL06eEo74}VmBMuJ6uA_F=92;09Lqh%PKs}T~c2Z59&pIo5F^+8&_y2d%VnnqMEkYWo$R^`aQ4OyAspp@EsA?}K$zwb>eVd=9Ca zX6|FFN29f4q4<~Px`Mk-X{K20XsBYNlp9ty)S4g=7LCN%==j<)nyV&(JTqJ^mdVAJ zvQ@T`WK>RPQaVEFsIIkn>iDE;8Avj`GtfaBsiY}{ZG>+Y93f!|Yoyp4tB!`NwpTsn z6lRl^w4chXRLaG?sy#~%hAt0gUa9zkm`VlE%BcWxc?qD9$*oi@!J=z_pOuRBu8I+p z6IK30%zcC(VWB~a9nX5nU98gPwLnMvG|s+S~gJuf$j z+m9#*9GFFkF_ZeyysT!w*4!)HTgc>OHcO;v$WUf$Y?EA~ark6rHcP^uGMn|p{LE%Q zq`pBoZJ)-J&e&FQ9bV2dczNgG*($LWzj}mDoQCK1uSR!@n+C;?q-oZXb;~kA;~7C6 zq>Yf9(2G*G?%}_(3Edau)FgCEuqdeeOz76T@)NrMkgf^6L+inu4pm#Met^lH$Hr0B z21FfN4;ns0W!-qFHW_;^UPs(JTp@LclcY zI>uaniDrparY4hB+HxB%sK1ck%q~wfd#Y5vDIdZ{jQ;Bs`Z#U!*5F90D;!jc6^(1I zO36)A#sJw}Q1b{AOL;8^s!)>Wg2-ZzYo1u{?$OO{v@|h{v4{{KI(*TjnB=BM{dEDO zAo7+bIA{T=R^YA%OW$mb#E8ad$J&Ix5T3g_mZ2C_CY?-FWV!C(Vojp=eeOX?F~ZvE zrGx}slgLl_D7dI1=?gn;(C~eEy+#`ix+3XrwyZ8=BT5tzJsyR8L;j(mjSezb!!YY) zvZ@m41lZ1X$Y0Gv3oY4IrgiF&z8DgjCo`^4rYdp7x}a{|K79V<+Vdyd$Hk`Scz8~o zy)-vGL%gk5y|A)iSIA^nlctV0?>q7mI^9n!q1^g~u1;y!}vka1w4JJg#l z-7>uC{HOH)2HtP=f;C?-opN$UBLpo5`F)aRC24jIEl3(9!s7KdyFrT59QZe?(*1>t z*Y)XP?AFnj@3)WY_=}W@C*8^B&5fDzd{-JPE4jPs>%ifaBZgon<>-CGdHPq227@Mp zcY3enc1nx!L*HZS6ih|v*PWD3V$*ZdG|@;|+qji}(FR0xRN6`2rkL+;y{vM+(L1`- zoA6Fv$sXoSgok$mC0D_MaktI6Dm65XfSN8a3ej3$_KsjQ7u zO~a~NYq@z|erwr0qGlm-_F#@*es?aoQ2{odAz|6ko#BUjM4WOOizm$7cW}GIXfXR z7aj;WS`o8V*>P`oa@(EiE+(GOB-hPz>JJ3>@<-wzRJABKsy`h)5La#Igv#ENy~BgW z>U3lAv7PIaqbClJt}dGEp-P7BC@_}^WM@P>IoExznCp%Vyq{5D9F_SyB<^=BlMZrfgM1Ct421A*v>ImPC%cJ``Lc(8+61;;VJnfi9Q}rEvd5GHm)hhb`k@xOVmLFBUcz@67bY?0uAvr*p^~EK#T)N=4 zdOOoY02`4>XUGuEOhS4tk0hi$nR8}xoPOo>nIXepqdXK41X11*eaOrABeI*RP?9F^#w|PyXyO`*Qxr}bAQi_SJvvWyG_^GwcoY(r*^SM9PjEpinkQ6 zUfl;R#ra-mPqEUuxS*XZ9VgU+yc3JjI)T9G`c}K+u2@$;Zgw4ncIdkir+mjMLs~W; zB?z@cAU#zm6F@ldAK%|fECWMaYD6_IBAiviDN z#WBTbnBX(re~yVGndz+HiH?eLGMT(FU?N2%E%G{=sa3-n_z5`CxejcSP5xN-sRG;L z#KTtDUW;-)Cj9+Mz=M5pg0~!Gku3+9=ZC$=5zQg^^&z{8;{!=mLZKAQ6)1Gu?nxL9acN&$YlIrMrQy4~+Ze z8I+7!5vtXc1EBVy2eRJ|cwINv{Rp=&dnD-WuKFf8@k(Jg57A!zl&Y-dqU5j4dg|lr zdMX|dWQ7r~Z+YL`O%|Wr<9(uSau2z?cV|%BB=}?1-O!Jut(1_W?c1?Rf>i#jvBIR! zGkM3nB?lbE$nwcsunL1ZnKF;huU(>*1Owu zf>f)X zOYS||xS_L`T)A=u?Z$aB?T=xmKTLy-JR6z2G|R?I-KEimzrv+wOlf{H=1tGeaIu=a8|z%bjHOD?*tf^%w6ne9+bontWr)aJtNpCBW?xl+ zpe3>tH6kU)6WQj*v5~Ag8^^E1N@R61Wg=_d>($%ujW>evdfI|CtAA-Xffb*PnZW9c zN@Gq-VE+Uw*OO%No;J=g*5)mcay~(_RW+o!JKW-|ay~(7OF5rdA1~*VOR^h>cFd|b z6(oO*Rs5Iq$h@dU6~*ba3!<$mnvI?Q2v%Yw=(JP{aS|i(37CnIT(m}9H)v@C5SxTt zp>kBsc!~aee?01Mof=v~eOk@Rc#o2NU*74nfzr|Qhdd`4iGt!fd2a8n+=61m2QpV; zKKS0z(~2f>y|znE(rn6g!!cT1phLaxje|kC7G0j1-Xv6N2Io#uB=P!aYcM)B+|4R_ zO)p=o7YL2Vw(jB##n!CK1$nRy$!u=J6lJ&(usuG$Jt$j|{C^esz~gDpMz@^0HFF6( z$aNeXynpp70S+z--HW*N^(7s0H#$GY{jD>@rQJz?Jh=osF}>>4a}`ZqwMAYGh8vUq zWx<9&dDnVS3TK;tr}FWvgn}k~?dSXr><;z~!=tXC3kb9dZ}Cf4F(H zv9VOmNvm%V%y74+dk8d%pX-vGdYnCpp{%Iy&9@ep$~Gl9%f!nZ2I2^9A~htRpC|XA z>+sPFTm8+^skO;ya_YX`e(lGyCD++P>IbQ13)JS937f5DmI@~zAI*B!k$MJi5#5Qd zXUF>wj5g10$~ja28C>Q2x-DecH79yk*58yZ8+F;DFlVX6LAv{{F3G7O+RgPeJ+cl) z=j+4k(b4(Vd<}l^thG#-_hlk7@_TKZV2x^gPNt*Z6{j9?sb0Trlht7s-kLm@Z9#sg zjjH0w(Z;YhPya3P)su6yJ*)KttQH|5I|+cJ$!KQ-h9GyY?@EroF<>ZH^Z`;EdSx4K zb>yg8`e+`w2>k+lkxa(thSb{ulA^yson4L***z*#Cj9KO8j0n__DJczzaXp#n^}Nf zyZ>AtO2tPsz4;G}4ke#_mmUw|?qT_%KFX@u)_K3Fj;+l{I zZOglCUpjeX$uD-ucef{t-BbOIJ|8$(J_kRyt-TXW)Tu|*?^`8%3Yb|xKQ}jbB+PC8 zuMT-`dn^DnUNx4abvCyUk2!lq(_#?{6<&Thm83(A^C)6!Ic} zlQ#*9AD$YL^;*wW(0iMzNMU-j$ZmaTZoyvN9LSf8HnTc0s3zxrQ>Z}Y7JbYhZ!5?h z2)3~~>~AsU34VnjY9!iM^6uqvzA02HP&e&Ss|F;%;!xn>?jq0WJqrsDr%=>{g(KhJ zCAXuBNI=|fOU`pC-4=LT&W$y@khm&$XC;B=K0!fVjx7lt(uOW&|2sSZwMRuoHK^f? zEW#y^7j;kopvn#_+%ShI&1j3+@HKd>6r+wS`Cx&rUEW@DjPeMrXUe5s2Z z%r&T5Ybw@)n?fmVP1{2SPn(evH2b}dijX5vkIfmfVn$sPmNuguC#(HCyIE5k>@29^ zXc#basYf+LI%|HDo!kmYIP+{N3X4X5dmaTEB>Ab~&8FdJxaS(?3QOD+jl#}wsjMJp zxM?_b2s2l#5X>2Q2&cO)^~r#lUGA>$%{Uir3=T+m83z1IF0UCZ3JglFErfUJkBSN$ z5-zD>03hTxEpW0ndwAo(aM$?K8NT+jJ)Y097v zkLh`aH7t$J*C!f<TQukPQvfH8SrDDF4-<0Cmp`c^4|XzrEYld{41vHt3Np%5KxtPx zXu{7H!9fBRg+K(T0*^oqFH@T81do^xm+Ch;GZ`T*QEbf51M>b3#?}?Q6c!^BW|N0m z72uzmkQ8AMzt?xjcTug2x>{k-R$G$_aYA4n6GRJwp>Cx*sn zY7v5P&3S1hk0e(O`jh_pX?RDqKj|&>=1)BRSvNVy!O|I}38b)%ny;FQXI~SRRFZks zOP5ayjmC^7^?_UwSd4dK175?`pU`E`>yjT&GMENlfx+-ca{s$;i@U>()BI-R`e6*S z;JRb5#yRbrJ)yf`vPK4R@}`2^6mB^B#w_SI4HD5@X`EmDCVtuE>sHHpKGFO)Tyqd^ zoLq#jeeiVzzQ|jm5-%HhM)r`P>kTU5f9G2UUK?o@6j^(VWAU-Z^z}!Qhp$`YV5nhCRJPHWN=$Js~Zg*#UcgSBqzh&$Ea6GwvINm%r>C@?Ky=#qMl{W;b zy|VVw4lL-ndNq2fFBC3QU&$?5eEgf85@JZ?@zJFThw7|sfpJiM7D|z zMkx?ymf##^)*OMm*sGD=U*{2XVH_LS?9-_~zgT~%7YbdkKF^Z}3vx7{IuKwzOUh5* z5uh=gzBd#xC}fL7XX#-oCX; z;+IgKKv08;`cmuEQjx-Sy4>?pb(+YmUyIxOaGF zw8ei@EIyp=X&xyVpJyzF@Tk*8qagtw0;znkdI{9pXkS69HcTW%_>mN zpCaO9|1W>(BDRigyY@DsKuU2FPB`+KpxC_p!$kwn{ zqh5$nkzKATvbDCli!TseBP!)gfTHxrvANmi6Dx=&|Fp=zcE>yTuvp0fR41Ro9x!Lycc2}-us zwogws`wwI>LE53*S}-@rPsTUPObMhPfsC~n3u6F|c3ri`2lW)xfG2VvdR`r$8W|h3 zoO^P-a21-_(I*1UWoH_XPnuygV%PZ5hx|x|Z}l4Jiq5SN-Vl9gs;|)CcPP9sOClnx$yx>E8hyOd6vN6T zms6FvjnJY?L0*u_D=`N*+hcO=&0;$6EJdOJrXV+F+AvR6vd9N+Mp|#M(F(SZKQGAf z$j!3X#5mM_oyEyUkS|W45ahx{?RbFiYYOs|$Svw`{Bmc6T|9VQe6Y|7lSPAO(&!L; zTGOCk;dPC-b)wZmkY-I0n@R|Ik||>iOc2RavSkAxcaVJj%KqL}dt{)Q`jKm6m%P10 zo}8>%d;)X6uPI|KSa88AD!xCoGcr8Zl`JPCRm@>{nWt@%F?w1U>~^Wbq>~^Ig9s`U zakqn|Y;Ki-1K>9rhSVe1tXwT&)(mCpYJCaz5_mZ{c8h|(T4BI7iuP9?%kzCqjJ6aN zN|`ZB90UP)8=Fy8LAr*PvKU=`FRonRL}ve0QKmf=HLYGvfru4*ifrN$=nlV_V!a7_ z{SsWd-+_DgxArd#$0y*f(UKeiKo#*&blr(-ZG8%@e-sv>UpVPAShnLZ+IUNaB;-Q0 zaKscyD<(829BIhp?~Bicbi!mVl-@u|$a(}3)=-L-^@vYlGKls_`k${K;(C?VDEVv= zo%IL;;G3$xk(&_spITv#2R?7<0^e>#LPOv(NNlDdHYrL5$t%E8(Mg^7c&jG5N(J)r zT0~1eA*mA-W+!!GC17=QQYZhIO+)8VcIwi3cvAQI`17S8fqqlNF}^vdu~9|wXb&)* z9)$~97!Ud3_~cdW6XZnqnI$g9t1>om6(m^ABD-_()%aqOpojaNRQA03cvWpswwmbcJWJajsQ<$ z`?n@fz*w-k5_ZR00F2bF>UZ44R)g6_ekp300E^s$kSh5MLT`B|T@akDsfA8mt@1XQfD)`EXa;;yn#2%WVEMM9rPcB+ zl(=KgD$btT*UY32vKA>H?B$}VW> z5H?eq`#5MTnu?X7>KQ{b;B7rINUUt@`~hC8;=7h0KZ+w5uU_3J=DKt)E`(xn3OfJ7 zlO08n1+1)R_h4o2zt$noq34@s$ZHsR*7|Hy0}W43B_?mVwxu=JzUIA57RE`a1y`kZ z(_Y$>RA7VY(FROpFS&B%3W|DIpiMUvX&6UA&&fTQt-h&afeW#|lZz8vkEVvT5@@>N zXufR6m^yM-Yu2oGTf8B}sZivTNVX!L%*NxZKgJ935~wSCzpx*c>4#s8$Q$^rmu-Ag9U057WQhAvuPv5EvDxbNspT znHBB@m&1!$*k5=OHOFNrdppk4^qlR9zYkCy$zWVmPd3mA2Do_l5n6x&FEm2W@GtLLfvi!M3 zhMFcS8RE@`$giUDfI&s70|JtC8CT2aX5@{Eh^*lSGRGk3xIg!zgRwmrBBvA;N`^D0`&G3nJ%`kVrvQh=0X>74b^SVbo_;!-q(Yrn1 z!8flCsFc$Wq7S`w0+n3YOrie_{G0U_=;o-RwmBh&$HopdzS!ANtzzDNC zmo|z_;uEXP6_cxZdFwCb9h}9!j`OT_x`6P+(mD^#Vo+!Lo11;O``sV-?MB!#9W0nr z&Ji0BIVZ|$zz*L#M-~pqTsEPksuH?FA}Tf=0$QL*JU@abNj`-Qm>>!<`d{@ zmQ(N*1X|TBxcby|PYUp#ISGZJ4mqI^>*gmEa;dbNWbiVT=OF}3c&1^FjD zj&E)cdV6l(*xnu=-yW2n*GgjKh|6~eH)bajI#~Y%JyAvh*c!B{a(>*pTQ}V77nlCY z`&L}AcEAInXV{Z_;|bsYW#rJN;2P-r9t<}oeZKG3)7wtMsR z-x4_%_-^YaNXDUfV(RM*av69Pazr0#T%TOn8M2oe-#zbp3Nd)MM=xykH%F(|CL?%D zxwl{Y(f4Fp@P{xLhL#>d(uGD@YLTf~yJweB(NC(n;m!G7hkV)l)NimDkHc%Fn=nQC z!sYL>zQ2o>Yu3B&U0HwQd)_RVow}{vj+7b;AO7p^QUZ?3nYcM)rAGwZ> z&LjH)tdLa8;4+cG&veji!Kl6F^ngSut6sGo!{S1ihGTh4u^&k3_WUUl)N?X6vLz8lg`QgSwx1^|y1MsRmYE zmT@jXj=U93-dPaOD^QM@ws!xy{&+YzIy!xNI3BiJ9R7)dJS#a&x(Jn>R!1m{x~KBR z4jE#TYVDac^s=H6c%03p z&9T9EgZiA;?9yOt)><+-4R!#Cq3ZH?ON>N|e4K&Yefd2dg)*Y^3vL-uZ2!Hb_iB9k z`?x1ej#Hm>fG zv?GmjHjzs3D1emAR-+$bN>gtt*^ z2HnP*0r*AFWABcGj-)~Wa-J5M0n#NbV4V0HAa>?+ZR;+!%KqlFbU{9k!(knOgCsNG z7_}hTuW5j%OUCCf6&dg{smUzQ2&s}3?W$A=5QqF-4$w&M*#fj_Rg5OsL7)!^U*F!h}3{J=_O*T!l1d3jIzlzVHKFBRgpG({Emv$AyBW2$7!3f%mmT2 zx;xWA0~PjbDOX4A+pWPf=}=Rt6}3-K7=K_Vt%5VnD7P>`F)pwUdDT2g@2^XaX&w7% zunb1lUIAq=(ijSo!JvpK35qwP#yTVA2m;1@F{f84bwc1RI4I`wbV2`;19~@hE5R`6 zW|aZyFvx@_>0A-1E<&_>3ZfT0(%ZyxOT0m^fM4+J8zY<;pqEw+!68@j(>S_rl!W6` z7Jcr7{^)3lZaUn%MnhfRfs>yoFc%Dk;t&^1>j`K*IN&WW$EnSsL0jPE9SlYruojFu z6N9v1c&KPN%Nxwm`%kLJXys0Yki)3(OXFPK5hY3;0!0XS1Jqtr5qTwvqRGAMq5`wZV9M2LN45tYchNwFOAw8d)NT3y&w(pH%)gXNi@sbyI(LBHg6F zLH1ML=aN0-@?N(|@Z*9AhRsI^kHGv1&nxU!dMP)TSgOynSgiZJ{`aOm1j(Xf`Vb8N zd(+KCSbKv_psPII`<@kTCbB-t=YgpH*!$jeGXcUdQ#c^qL-{R1Mv(vkdX4v;>E_J| z_{Ec@*9oGNdM;*Kz@FeT3wKrGkF6RQ;l$L++v3B>-m}Bu&iY`q3E17o&5?)FJRIAW zks&8Dh=LO%%O*vj`3tmV` zP*s0eTAzTPVrt_HZS?gqalN$AUnc9bm&ts-?ErbX_r35bYD{i`g7eq?8n!dqnheLA z!@;OO87@^!r7`Ppz9z)IxORX&YNInsf5MG?&iiie^-E3L5fQwCM+YlMgO)WZN5jW2 zOSABc{jUn&OS}QwCY zWFA52C}+c*C8KFi)!k}%DKX5D&xJoWYIl^B&*7( zm$$qw+MNR?MmB{)q{ybMm>${GMXDj1k%e>R3=ZS)QiAKr{?>g}V58r$LgF6&UHPrS zMFaXLeqf?ocZ6j8hb56c=9C`l8vdmX=rpHR3At~o;2368Kh|zl!bJSmlv8@(V&l0>N)Dt*?s!J&p(e$ut&$rJ;n*E^60;U% zV<=gRvR-P|qArn=wZKUD)Ptb3inP+hAt*iISd!ERZ#qKptC<^S`9WUsL1kTX@xf&u zmq+oTrcfjTva5?Bf+Xdcwih@i>NPsA1_@;&FHA3HvyLWGrt%}{CM?Oo&r{%uTJJn( zOQ@p)Vq{tUep8l`8m1%zlB=L{iQJL{WGGLitGU|jE{+}Nt&!Bz$!A+ql~IslwFt;Yyi9N=?1Gu( zRq0Jb)jB{s1i~zfWsLPWvBS?IdqtA`aGJG%%dR4Aa~|Tj2<8~rIR{XR)IyFw@n?r_*7L-o2QNfPC6w{tAx;m=-yanB?Wi&Q-R;YEj5L6 zKnP}S4nD8fg@4L$Y(tHD^TQ=A>7BYgxjC;^5qOU&OPQ$Z*V8+**ahq4?Hd>nvu|jX zX3;B>`F!prrSn4GR#zXv31<#NUz$!y>;-iwiM?1YHL;fqxXO)wV|i&>e@lG)vGEX~ z(POvw)ghhD-uEBVN+{FB^scO_N^n#umC)(WkBfb83V03FAAW$7#}eAS zHryDVn)DW`|C^T(Um}APl8%TTK!8j6ol3))vCSU{@SHW(47=-u?T0*}5=g1nsf5NJm;>*fNfky@?Nz*Q41}$=G$jlF z?4ZDPD-V<4ScFc}A>~K&Jq#qw7xuC|(eP+2Zf^w+;@!Jt%7h>&z(S6UY9rI(iF5;Y33i7_J0)7moR%N;U3Qsvv~AwMaa&m_Li%J0XaG=bG& z$x94tX$kQ!o)n{bzLPw?2k*|GS~-#sWAx3^<->*q9U$*cJMh$%u8P0PopJpFxgk%v z@%S6z2MC0Xl#`Ho6I$GZHV7;qABuWbK{qavq0^^uG;`Y*Uk$kYalP)CP?oh2y1-yz8Bsl6BB=DU|h0fk}RVWx}GlSTC3^`3zh>i!O@A>}Da-MaMf%3xO&4F9tlB6(=mC zX@bvn|5+xEWafIL_tR2yCzHvW0-ncJMOJBThh3&T} z+appS@+$$a_Qgrwa+<|wUR$nNq=vfj_%ot+VrE3@NC+OM5xQf=9Kfr7BH*>=Ac51Y zR3#KYr3iVgIhItbt7u9pymw$@Kb?7K|B&LOWpfh?7(0XGj16^0aIjSLQ*#~?kVK%|geWpR4^ z34f^a%~*-BLeX(rf2hh1Pl-O*sjm_tc|Ow($fVaFsXx~OhgikEx>tf8GZ^>FGbkCe zB2=p>Bj(J%9q`0%tospeUp7k6+08_Wq^JtJnN>%n=KA$Q^4FH4>Y=)*ipK<5W`r4c z-j{ciw({5nnEvo4o%%R)s?F`8ZGAjq=T* z^2yOzz?uH$W*^>8?nhPUe;=#XDRw3DZJd&XSQofe&O*^Ka_x^{CO}MujZ7PvyEM;k z!<6+98}S@GT)IJ~2Na0s*f~AkJ9h6MPfF8dV-jT4b8hA0WHGo6wjrK#8M)RF6Py~m zx*6o-nCD#8_CR^Wy;Ni#R~a#0j#;Th4KQ+XODe5)8Yb_?I$AJmsS-5y?J@d%!d~%h z7TTh+L}aejj#e_pR~W!cWhr_@N{&~hu~OOBVWqMk|%fo+ymF!8f_^c0{gRIS4ASHc*WUFdiR!N^AwWXv_tdE!U$tBs% zL!Zy&YCQE|EBRxr`oE+{=0z>4C{D9o5N%b_;OwXYP)t68l^6*+EyY5d#7KMsW@024 ztrMe$PIMFhLatIdiebD;f4)B+^|ww9qeA^$6j1E@@{XSkl#ZT1a!u@t78mGHuY2QQP_9MSXQnp; zrJBLHQxs9WKH3_LP7Qan%3jmU*XjjAo267-XiT2EkO$k4%;xr>ej)z&_MmJ<@@>h7 zGv+c86n#YP+31#2cW1Z<3B=kIv*gKj+@5nH? zo?MnLcCJlc42Bz%KIg?HjU(GY3(X+E(?(VC?DAg zCZnB=;cl;UeOCy8yfI)nSM&kWJN)X%QT2G+Ja`fM1^6PFjL!|Jw*w?ae}g)^93!%O zRHjV$*=02v%L{&&(tUqHSP?d}0KInqxjvMNkB&~C9*&3h#(^Vhx#qJ?HxFZPUW%j)b|*|J5PSZI1;2VyvzSwa(@i;xT7$of0}{-&BKS$fsN7CR}r> zCbLJ(J}IA;-1=LaZ0{VaTas<6Zv=lKscX?`LEgh5x9!?Q+2Zn`B-|;6XzbTo)mQ*T zRF~k|mRdRMm>Q5u69OlN!2>Lbei+$%b~xNwAB;93tA5;^5D@`1VT)nXI0d$-c@&@3 zqMTU;GJx&Eo#+_h|7g{9FyzmSRQ>+;;1Spr_sRNzuzS&LwCT|=RKRh)g>$RS% zp!YUak;3$5nB7JfKfYYFnbnC|H8~fYLIo;y&j-^aZ!5?h2)3~~>~AsU3B9Bl2Y35Q z-pxGDH-$ZYAin?M;ufqT1)Jg4_8EZmzyQ4{8ke1DhRjw&Jnal0)!&!u!*;B7fq z*6c#!s@z?ikp!9xEd_Zwwj^{&8@iPJ@9_lG9u*a(P{TT`T^90qQ3nM8s_d}B4ReT+ zLl$m%7NoT!X5$reZC) zDU{OIv^`Ytv>7Quv+vPRa0KeHIdj3xsB6N~X0+pEwSQ+fYqF%B1vMNE178$2fd)x_YIw70_!;iGhPlELH$|hcGh8Yw$Qf=LP94I` zl?T#HKh*3YobI~RCj(}7d7ljn@@AZiHU76k?+*A~K?xw1N9HNB*U z0f3O(w7|*Q@FBQ?kHg3=pVMQ9{1*1<(P$jxBxVvmMq6z4j9?CTYSAdYC^2qbQw3J? ziM-JRW?G{F$8eRAbFm$}3~0AF0AsJkuE%O%F(sTeqQb1uh$BY{yp)J51_LODo3=2J z7_Mim0AixC<3);*Pjt;ke3t8ZlQ{S2S^vGNyaRBg&Hf6vnpTux@wBL~-%al9Jv+C) zu6s{8my$ZKuouA=gr6h79PkE}QAiz(w|C$jsU=-N4et&umxlvO@qF2v`nwZBq2J|@ zj~C>Y=q3XHFntdk?{TeM5AMsu^Hj^_a^x;xVL$C+n*;_Nd7rOy3W8GHPKW9TbNMSt zllaN>c}u^XQWa9g&B(E~wwsWVSduC8^i<^|*PJw@o3)q-YH4D?|02dU7OM0-A>13Jm8pj9(_EDTs zo&rj{5^6Fv`O~Xk1S|@H2v7wcf*KyFG}Q^-GaoM1&vRxnLRz9^gn#Sr?_g|QflOgB zGGRb@m{kD;s;OrfW=*f}knf^e7j?VBpslth73GA$IwlCrNTmVq8X3or9^@|za$7ua zR4b^?42>_=A_U=@^KhQLYS5qb*H6Pss{KiCp*Mfx>Cd{!HIGlpytRFl-*C-A zc%9%PeC>m;Bk-joIo=YLc-hD^vWEm|G^m9C$!{5WZKPFbWbVVJvWLf#tJkY{m)CbG z3TLNNYvDmxc2A^$Z}!305qgr1C=@X|CeXaw-Pzt9t{<*sc+1xL;dpZUaJ+eL z(x(I3de<7iDsKo!p0M>N^enc2wGX38s{hwBmt%aD zeWOAO<@J&JvP@0CKlq+?xb6VA8|oh}FefLBJm_<28y??|(fo;ad zAn^NBEbn733XJjgv6qBrBWad;tx!E2=48*R4K63>73_d#L`OK39uC0us@b2&8k_%|@LB zhl~sI)AQszvvg3Cac|#RCGs^Y@e;xIVh}iC%!@(bgslkoTv`yZLu?%$T)WhmZ`m3Q zs~dFJk!yDE8;`cmuEY4G-Sy4>?pb&*Y;N)4WdCquW3;oY{t}v!_YTjDw)lTm(En^} z_`IpVy_)P9ZVmW5Fu>xYot0*rKyXju8=%gK|DAYMYxwzjax0!| z02b=b0VVWd8XgyR%)Gg91!`qgmQqUlud~Q5)RS6w1PJ71`ygA}i9M=2QGUjjj=u zawb4g`s46i@e8#gi{vjdO0rEXma=wv$+{fNitqGzsgz{%%$@pC9gKxtP_3pj1ZjsdjD>IeC*zxCrUcTDK*m~( zg)sm}yRKT}L-w`0jC5wKAllHbwv|0$#tSMqs2_a81Wvqb-BBxTawEJ6c z?~o@aYZf2EobPMOSPK?hu!<_dXS(9c9DT1VSx!c(n8WZgPunD8^t3SiPzVZ>PJ%oP zBB)Hn-42#tx>ZIUM*NLNU(kJ0w*&2uZlA5sizc3u1fICM^au@*B!$Z+!C$6;hA+-Kc`WxdSPue9Fl8_3~x)IZ#R#Tzx zi%*4g!elCR!0!8>H-`l>95p&5J^aS1QFIqik0+;Phm2Mc1Zf4 z=k`fHTSO;4f&loYDsSZG1Eu$|73O!~^Oi1zAN_>E(YqkAiH6vus2C)#082&ZbmHT! zn&^zTV%y=IO3CR2h1ofsSP57iozuxbX7kX+nK?D5`+WTIQjkEusnHnU9MspSf_StC zn9hz;B>uzkxvSVG$cgSVOI(asWo+ImNU)kkcH`o!@wFmB5fAzqnW|WeF*sT+=joP=bm}r#Y-kXhsQDRvOYRWO0{9IsmJ-k(=q;5-673HK9SwXH#_jAaG$XdM?oQVo_aI^0PG*O$Uz)mnEVp|o;_mcM#h)dW52`pT zexV8+!!tzpkpv|i18VR$SLi<+9|d{vGmn;7ZEC<2#Ijo>s{TQikzz|QlJ9Jj;>R-9 z>@~935l{+D&szOcf-*1#wRLKB04!l;U&RuXYnn% z`_B!>7Xn*!sheX;U2|QMo-m4sZZKjD6AR*JhoRd`t#j&XmAAnJkYFW2GvJHLB!<`m ztG5*`tyXVgDow1tHgO1Y=gjIY25Y6lVAXD6^k1wlQ!8<(Q0#I%4(X;52vVlBs467t zE(U+nhEx$yq1Xs*=TE)@;V>3og?`*JEc$RgEZQ<_HLOLG-H_~I>nv#Bx{_=DEc5+^XU2!`!<0?X z!XfOXv~1%smYS(p8ET#}B)>X~?|{{+xUOZ#kKzrdt5^4lxh8XRF(RWUpY**77)G77G~3iW%P|f+3XzVOA!j?jqRm%erGmaR3kh~+y*(($psSOSbX>x;TwYm=vf98#R4a?huiSgs2 zZKkZnZ>W#L<#-*_Fk`X;c-am~4l>KbS&(tGygRQVQO9DdogpTO@9mHiD1-{>!91f5)I=@?%|6OX0?#ul+bbbTA;60 zWF}gwjR5<^eL|Ed!ZJY&Bj$OaXsGLuFei_GAS+{L3SLEF_U$ofTlV&fZ>`VU1Xv6P zQ){#qyE;Zp#?9~T=IzPfBkJjylfzE^!y?3 z9^NC$O()46z1x$@O%0#I?7YX(LVH)QWV6nSj^2`NlExQ(;MrhlfL`>vHx35n(_P3a z2{oGDjycFJKH5&NS?dW1!iALL_TF7=Sxn?hXx!f19`yFyys^DKKE6HhJZKjq!vFVp z{x7~O5W(!JPfN41Y5Svd5wE3v**9G=c<1nd=Na}S-gq*r-J2e53S`mb(Xe+t7;a4Z zmu}Zn@^0Go|L$T%RVM4+6?agPd+rOS=ntlY!6lCgy6q6)h(jHf-w3!p@LB zSL{$iG10Y4-ZoFxXc;r+Hi2CjM=xykH%F(|CL?&%32tjs*MOR|o z$j(TrG_9+P=v}*@{`H}+OTOfr&k$@8$Ke^!O&I8WfxD@OSl2oO%+x~A&xF(1 zJ@>AxH($14ah^dBvBbhuEH3@lMd46E^3{X}gx#!8({t2dbiO`O9UYy=wSM=E)mZ*J zytneyZrexM5BR4125-wKqhb`N>P1_zC$9d}Lpi-w)t^piz1ubRDTX?tr|oLwKut7ZFE$({m6InK|` z9XYBN%!=B94ijpT<`*==N67y;dYq&Ll7eu+Y_ymqrYzaqO^ToL#WJ1cifO z%Z`<#rqHnwcKxULcFjNuZ7S;N)VSM_>sd1>K#n*NlcD(R-$)pY!4Cor)Ye+6;!x3b zieXZ*&8x1thhw*OttOu?(8IPE0T2!H4;7MdSl(?Dcm>KmXvq-=H;4T#3ywxZcBY+S zDBy;CF5zwo=a530f}y!nU}$b9CwjMLVQ7?3$ivV4iEpx6kOTyF=4f21N4H_Y%|u_t zurq$KO^!FKmePXy3l#3^SQ+<#wNP=DSUeV`^h?BS^7n;jd}0twYER8uH*Ibv8yEA% z3`CH092QJWo4YofQh|6F%-w2OnAz@DZ^1Q|XFuG{nCb`^e)fFtqkzNd$Cac1| zv?sjcY+j%qvU_Lc&7vEC}>eSQdoLhv8UelKH@gRD&2C zK8XBvMw?N3Qv<)E9i^JUEN%K-i*F)d&VZmBdzH8qY->sfU>HQIc$GG}?b0x~W|7?8 zl^fzsbp@w_X4MGMqc#9uzv#B=(_P$t4~mW(wcY64M4vUGKhT*D%98q^#UIZ0XjBTk zev07I6+{YxDseOlqS53bQGO7m(uRgYc_oU58weC6J;2B{qLC-BwG61gp&m(iWBLFLxL$0ykJ$WWx5P11W8f){0X%|4{DJyk;v*hQQlx zsXG&BiPQyvHKDZ=zC*X>saL`lA4c|`9S(Qa2cu0u(yA|vdf7w%J>4vgJUEp*6w8pH zGg9e=ko7iiKnMbLPyJ+&DR=4r&FQ~~Tf0O4Zd%GdrESQ4`UU$#o#dSKh`T{2Nt!b&F%$vUyzq} zf)%@g6`H;^eZ>Y}ft|()}MGpD)O&$DF-;P&m2XMT8D+}XHwGXgVFz42XD|OQqb*f2JIh6$^3LEoI zIAA=!G{DU?Nc<`Xk;st4!q31U^I#_dP%t^x+36%}C-FvqRB-Q8oO-g^ps+^1R7Gg~E!uJ!5LOmmS8RD!VkRS1Pf7}FG&P;h+t4>Zh5v1Y#T%C1FGEIj}NgQJmA=6NA4!WkqbP`h%g)XbJZz{AVD3tP16>KFzp{Cc6 zfjRAdDU20Qh#o2x{5qGfQ5R^kDFIs{$11}nXr|K*3ICoLuoa>(0=BYTV!&2cU^fh! zvKPCDBJWEIV(O>L(Ysj6;=40sQ96YntFk~K^ZMdI#Yqcf3V}F*Oj$5Fkg039hwOTp zN9Qvoc)pYUt^2AtM!#bv!9D!D@>@bn0rW4tcxSiP2%!Y%2{)z0;7!xjx(0tqgZDHY zmrL+g=(5@}V(?ZzDmi#huaO(P)9VScu~XF(Q0%dxe`>cT5nlaPMAGXC+<17Bl91~t zLOULC$=|YDWd*C2N>(&Q4~}mOHa7(oa(GL@QA>rqJ*4W`PAdG zs~S~LI0W?s6#Ma8g3y$Vih?u=T8GaU7Gqm#MWH6fY8ID8ML{|lhMSShNhRDch04GuNNYSNLzuPWupu~MFJ z=nXz^o*@!RW<4VcwsdUT>StdGV=ghplKw@J|_BG^pWdUI3#dhg0_!6S3jxmK-gj8eb2& z%wZKQleKBsy@)0L-P0$NbEp-$dN$`$vUMS8t7DH~gF9~rEOU-4$H?2oP80KX`S-4} zp*Kq*FG*@kiBCT^9s)vn?Dn=go}t;>!UU`wvDCEmuB^X__e}`}O_LK5T1wvNf8^5T z54X35yXzaHvqSv~Ay$(3Ls-;#uUm)4V{mRcG7|i%C-wMlwMX#Y`)EMeT4A@X4@ey| zN9w?c>>KGx%5kI6Ma$Z7V|Z%PTd4kTN{wyWBxY-eLdO1BDY!^Jl$QY_oPXCyEm!zLR9BD`n&5BxLmQ#$K<=N7)OEB0-g^EB-HCkLSqdKMfc5f z3L~j@CB8}fVT&zI3Bo&lCve%yt%xF?;WYeOM?TRZH)XpO$fq$KOHQIo%@s!jgc}Lq zVj6fscvN~MS;joV+b@6HwaULe!3pwj`A4tA)jo62AoL49o%TQH0 zoKh$jD_GB9P$+icGK2;+brKSBrGh9?eUd%t91Te>4-0~Ld004Cy-wUzfBAJ3Qkk*w^mmXy?pH8BR~7ILijw2(147DuG-l&&5&{^kIAchYgD zu3Oc;6L}k3znX2zQtmtc+VuhQg`{J-CQkroWs2A7;V#@_; z@&bzkcw+4UjL*Lz1#4yU$Q#8LGtM(ojt`$2Zj--SSRuhBsS~x0h{ZaoA)oD#dp)NoLsT~ZBZfj9#*ZO2MDJno}N2-_XhvChv>~w4^H=J(WpWD4k89!yx>$n8o6;LFCKJ47C*gz zC&lj%c^~(`P8TP2%aIkGb6zNV9Lb(Tt)Gclahc|I>bwPiTs9u=b;jj@=ZAwhH%%vg z1`+W5a4e-!IRr|)pJtPHjBo6grW7{~)(r(b6d$DQ_BcP3_~sS|ZCGT1!nMrsWb+63 z+<|~tMa34_?tL3Ih~tVVtyY8a;3RZdP_#p#)kxN`8B9}|Ry>1@*m}bD7;^P`Rlix^ zrA^D--hOh`%3fM#IY^I)7}j$XDsuz3xp4D->jJpHFCw)>~ zCq-j{EFA(1d=JVd%PyjRxOhf1OP)*a>OCiFfi3-D4c;7Gn~MDl~=>sR*ouG;eu z|6einBM)?4@_&%34`Qdsd4g^fW}Z~8#0LE{{msojya(Kms%X8qS4Xr03S;BZ%1!Q(hyX#$_!{r_oY$X>F2~k;(L@uww$~4o{^sy0mxf z-anq0CDbR$efkq&%wufDX_*G&ijeUP?-ol43T#togbK)RJj=Y_85IpV6XT#2PaWkATG;l-zqd|(RI9%9?)r7ON{QD6_hiR#R2+(TJ(4zNYLR zrB#zZg_PDws(7^&<)CVF5;*CdAlRxEVbR3pKM$g%cY@M((mSy{T6!l}blv{*`p-W^ zD&&K4< z7L$N`O^#X@uehJ@k4OEjQ$tH+Z^oUuf;oLwQA2!@i&@9Z20};AAMy<3K2Zd`huqz} zGqVWT@OjLGm@kxfw5>Lo19^3qtkKlTaoaFjR-iwnYslth9r7$} zi>bRT+;24dw_+hV@$a?Pkh&DqhkuwiQW}=P}L~g)VOw> zSX(4B)tI5;H+!aznpu+P_Rk!p1Re*#?(6N>ek|K@m-tB@@~2 zt>jXu8V6)J&J-Q;q7HcuwuT+=KQP)nw<%{i{U>mb@9E+v(={J@SJvN@?b>qLo**xU z`ZoK;4tZObWRwN%W_o&KdwV<>ov+WPM@Q$|@-XPJveqw`p(|=2J#pLX4%~%7FkdSem1c1#=)Lrbp;m=hr8q|IFRDYw-NBNb{MqstxQLCL&kXKx^4dI?ULVmqN z?ru#J0XJ6-0%@Da_yL&n*6T54Q2e7dSqXQJszK)AQ%1_8sm;5!$@b2%x~15n-U{%Y zm&pY3Hw5!(uv18KI3%`JTPWLG9n?c8qCd4GUu{!W0mo3?C2d(^<%~mU1cTK|`r^aL z-m}Bu&iY`q32F4>=IH&D_QIlZ&TCKUD441}8MC6IUzep+G-wO?N}C>oL3L)M>`9jvPC4N9^aM!hlt7!BVwUFJ(Y|mV4+<;)+)T^W3!UjBQ-JXGq zfiM}DjY5SfIpE#%TMP1xe5=?T_O~p^G!3WVD5W|u=M<(!{x$1<9Os%sq=HHt2vpjg z8&PFoLYALa9b(9vhu z0`|UD6Hrf76qq6c>*GcD(hG%-8byHXDf_Eny&O?A8?6gLlQ}vm1$n5zD9lPOSWsrO z;e?5+(qYMsxTMJgi)}XCbdG6sk7l{qYKT^ zs6*f9&6sCm23Zq9Hk&<~79XhZZ|&v`X|yw-Mvk=+QF5lw*om$1elv=}Q6MpqAdp|1 z$H4MPd(@y|(?~JYqYHC?C2BH85n`w`R_HL)T${Q9Dn2$h$B}b%gp6mSfaE_=8sk!v z3`E#v?(W`*a_z;~zeIu|K(*xRnw?^RmgLexa5L6Gmod<*VhFeexk(H1s|C-28s|9T z>#{ii`;0Uab6$Gy6SIcFbc*CRSh`{TQNt9 zm2W_9`THs4HOSRfG+Y-v7>FtZ1Ji|?qA)xcsyB?ra(%RGzHs$z_25R7tLvd_+Bhnq_q@|V75D9Y-0k>~WDm04ETU8kJ*NIx&^!K?-HQUB9SMqP9;-rj-vyi2-p z8Xn~NwsJYFrp{NrslPk%I}yQC^09(EJ+>FXKTF>MhiF_YmxCMXC)N*rTX`sM53sHv z(7zWtJ;?W(n^OpsUVRiNR*&grnaq`xNr+?mETdman7E;eno(nIN;iQQk;GDFX{gFX zu2yMtir^tO-`m=LMFU4%*?Japq`N9_h%+t@oU0S;BEZ1lz3lvb3)E5mQozFaI#`${ za|H=tixX-=;ANbAtu;;x`Mjf;pe)Uk=H%BL$xR*m6Y!#8kRkDqV}Ga7Tmqi%9xm0- zRpxO)+Ft}0$WL`JB9)vVl`lvzrIdNzVaB8mjBlDKp|(Q|Al_>_V*oS6xTM)fQ+raF6Zu?0PDC?C)&A;GP@BRrV6pyKE1K&PBm=(vszHC!Uq21c ziS{SGh2H#$r$6f^*YvstI_5K;61GkAWl!<;Y5a;x!p^)OKkYm*F!C~rT74q+$5ITX zfl**6{F&Hq^=xo=xN&-k+t*=@U;)Vj=-0C@1yIe`_7XGV-hSIbv57Q_}_RXZ_ozs z1J^Rj+6b#a!Xnqy-P~i1`Sy<_4_~xV1(-LiFlIG)@-9B-bR^y#>>-nGWB%EyQ3yR!7s4y?zO%q`;g#<(*37Q_el_MYV1 z^S*lQ77zDi=-lN@DHdHqrt+-?eS})eNy<-pla;^&W0Ce+dwial^(XY8wSTn_vp}l< z*E5$R9F#`_f18qtu79t)TaxZ;W&II)j+o;^@6*Z%;k_#rliV zKbTzD8M2$$&6t~c@}`2^luaB6uO89lC+`T57*5{niwJbEMY(n-l~O{d^p)Z+}IfH?5e+nw&K0RGovm3pB3~!+ZsM^ z>Tj^?gb&S=??Q74z(Cz8 zU_?7aljDMp7~U4HJ8dkwe3?22c*iEo5k~cr@IqFJsp?WR39q%$g78yw>`pXQ5DrUu z7G{-Y#Hq(s2j)C3$g4Yfsk2ch$>lym!XtJ!b+t73#7I8a$vo}IjsnWDGS3HEH3?b? zK^hW#UL>#Ukf-ICUad+H?8KW}dC4!k$czyxr}QYe03*r^QGIR}8nv20Rd5r1?yN2G z0@d{)^%hPA!@vZakFY=T&oMv}$G$-)y0k})vh$!y6W{qJAZHRbo>O0~F{ef`6`{Vm zT-8^b&$9Dec3m1t!)xA5cA}ihVYA{FYJC;QM?_puv(sj;efQhl;q@H(akQK1nHhJ1j*PJfj+eBsMVOV;Zm=R&VdY0Sxz;% z3I|+=ig`HX%FZz!jrbvD%9a}7W#rY-@rQGu&TF+Kr+K`r@$#Lsst&leNUly%W%xL> z+&2~Pp|VT(t0^J*$4LHNLB1pIP+q&IDn>!Dip5sjwXNoF8RH>Sd>b?hBcT`Ce!;ZN zsY$#~_-NkF?5$n{?aa9q!aJD{O*IRfj0fO%Inn^?DxaIP2#7TH@Z@KaD{T=In%pNV zf!XNnDd;PgIeQedV{@G(SKAz7Kx)!{gHsgpg@PQ(vs|8)Y~5!6tRT0>?UcQ! z#c}6#Ha+U^!tJPI2J#jqE=U2Upl*M(Am1LhLGj9j7A+Hw{f?+PPw#m}>V{62ECn=E zM#sd{Vg&ut3KCAr-qMLx2|<}P6KpCBWQnEBWM1AN_aE<&l~kqTaU|Qpl#zz(mXfrx zU^nRsJagoQ(P2aoXe5Lg%r3LkY%1pqXARHn38+dk1+tTsuPos=ELajLWMMUFINW`s z0RioiYgVq7aAyYp3~FFtNPfJBtGe( zl^~`%ttLI+8=ds%bjhSAz5fxP-Uu43z7Q$B5g)e~K z!#7pYA~iqbr#|1yBh1^tr!8GbDpf=$aAJp|5D=M^hy~!A*wjt@tyKdf7!VeZ<5M?5 zU~cLr76LZMrf%|Y*+g-Bd<|z7`E2yjQBXj?sc{pFA$^G|V8>ehbQTo8H|`$_@EnNMys+ zWC+*@GE0JPG7FX@F+|+ln_|c>#@y9sF7NF4%R>iXFA%kMw|oiUytvXXccBGoC63LRa_4gXA1}Mfp&Kq<=E{lqi{6Z$bP* zl{JDph;0E0I@X(bnK|lTD6*&`na;K-ek>!6 zT_cK}?E=al4dFSx|6ePPcDTn9RC@ovrpgdiWO+x(hTDhuMQZ}!KF^3n?6bqr%_UYWb%he7rtC8TyjMxjOaMiZ5kX{O zRoIFWwF)adicw-rE636@b)=a#u?OF@XTbsIHXhJFx8Jh?Cf!B82ZZ?dKP}r?% z<`zq7O-# zwDf8O9jKJy5@;5r-n8IxFL3r;$MGmOMZOZa6`Uc&Y3lIf@^5xXhCwUzMMY*rfq7sI zPGwfeXUUkf&os;mMCS}op-^+2C*&ErBl1k)09w!G;aWG&{zZO*vmCIqY*EWyiOs>l zi1m&vXRoM^k=o#}kOC|atyacC-Q7RNNN@I&t5){X^yVPIdJFozE)E%>6a2Vn>-1Xu zhWb`oj@K~_Q*y%}?mm%~0V|5az#Y{o0Fk5R-RbF}LYy_4ASQ?~Jlfsvw#Z-NZWTg> zbCuZ@4mX*2a@0&Ak_R`hc1cCl%#!?i$9Pw5`A}U6%-pUlK>?H*StfEj&lL}yt&wN+ zZu38M=9RxnH~pOIVO}qX2@C_y^vJ9itQC)%N5%N^Q?0y2A={bKl(flIHNUi#S3IJI z!056%l@^Le;scqGa>J&|(!Ao)N3u@EVpqp`a%vO@4wbg4Vir?6)8E|e!$s?URNWP( z@@1Ur1^ptq#8?cNQD#)k1W{Jql>W<6_Hk#J4P!YUI)49rcI;^HLvfvk*~DR>oy*|*2&HT}KfTPvc(5;tTDURewV zQ);P6rg$AjUJ55JOlT=wbDxct5w$eUadEywUY?c0>9oldF8!ukUg{?3v4+PMQaAB| z*{PepQf_7O!&#}1q=onWG7efcQ_e|&1aVdwyWG1w^AQG}nF0y=FjFA0Xl4o|*I`XV zpX(Al(Q%~Y7qSYsk_eeQU@KTg`Bso;RWQgPf-WRBU1Opc&5Yzh5Qm#Qh-I^r2f0$( zbuM_D$^ANjCCu|m>-qk8)ZaQawBTi$<7M37xW7$^mks2No9B{XhtZV!5UZr<459v|Nxl1Vnq1=9?xMu@^N}zi~*W8 z=uzeHxJ|dMNcN}PrUnbTwC&5jH`0O?1Kz?s!ycL&PiD1y(^*X&dK$ba!QS;?xH0Kp zx?NAnyJ^?|yNeZ7nXG$P+(AX|d9Wr$e=r>kE_qDQmoPkZ)V!1Yqi=3$F!OI7ZEP$t zaQ$Vk#V_r~_nYj#nr^x`Pya1(y}#$4Y=Tr~mzu!qg(Ldz;d;O$ss(x5yyxk^;JqHb zu+`rjom!iW;Hlu=e(gukQ(?iM!Q2*BA_UnL7F~&Tqh_skRr_YAO1!~M=jGpqlU^N% z_cu3T-13FX-c7xriKvH_!vn$g4&#i@GHR_uujpwNq^{>Ji8C$!$}8hdiI zG3?FLf3eaV&amh@t(`z_Qj3stoP>txWVEv}++`0g^9Hs5atC=OTCMP=?X6vLc6}r- zgzZ};dy3xN;{4p)k)vu)t#DO$k<7V%`H12K`y15fMC-ks^-L>}O00}+enAO~Anz!M z`xzuh2wS`VTz@Jo%hK(9IRk=j>{a4ck|8E?P0gCp0T^;YA&^V1YV?{#a(7p5 zh&R<0oC=y%BSa4apVFoQIPw+p9uyrnYP-?7iM*Xm8yEWfnkBZyaGx2CN_i_vBBCHt z5LAhyQ4ozL4~g=FD3vx1h0*}!hj$^asL(QxK%gLmc7{Gd@Jbc(B%IU@P-s5|+ zQdG+ZS%(wY))X3`1Hj z9~(k1>^+8DnD5|zTii-6Z5k}Gbe@B)v4B=A8+bCG zjk_4?u9!suYeH)$#@|{s4-Nb9?@3;bZOX`lQ@KO23<)|TeU%V$EO>!{x~G0J$dtQu z+wetn&+}e#>#DX^&TDnl--Y=N#~?N6U8Vex zyos8dQkx%1t{U_w{q@t^Kmgy-nC z7mlYsJ@b9C1Nxg?_X5IAl*kM6Qr|<^Q^c680LA97+SxZV+L{c*vAQ}~gsKJOhRM@1ij&LI8hpO}0Qr33 zd2Kdek34qe?$Y++x%J~~&k?iF7bWl0`%=dNV}ASsnP2I^ogedfLB}1tU*5BBm>P&m z2XNf~R@T0l8GbP5)`u%?(Dj-wYi`+es&PMD8N>q5Q+q*cz-5kQLwEZ z`UEWvJOMSNM&yOQ2Ty%kV4QyPxJ1uDiTb4a7_R2i#r`HXK!-E$Q$e~sk_1Nr<`|bA zNZqe$Ki(?i0l^X4{mI)aOWi9pdfq~eXFR3iRb(QoN*nS$dCBDKW+joXQn)^ zRVOEuKrFeWAEf-LS6WKKoOk8+wc@IoYp8ehx{HljN@7YfO@~ZLFv`4;yO6Y$M4_wg zltlTc3bvAHz=47V81UZh^e0{i+fng;=40sQ96YntFm13K&B9g6UdYWlLMK$Mm2LX zJ&?hCrUcJ-vcGj-702jztR%RHe^-8swR2;UKK%lYU-{fjW_d?GyNIyrrf`-?#S;4Fb)rLLI%uRIIsBn9E!(6dKfoWhqShePLoVO(%UL&35TG=>eLYv4)H-% z>moso^c5~m!Xan?-H5|fT}^ z5j@?JqeWEX>mipptb%2-HVwOv*);SvbEp-$dN$`$vUMS8t7DH~gEO0l>r9L;VUNR+9KbSk!6Ptrg@;qbgYMUSE7F0q2$@Bf+nFQjhOe zdj#*jj|POT6*lKzg6!yJdi^OW`XI=S^duEm!xMVEHryDVf;%DA|4pf}iKQ-y9SRvo zL<{nvylm*!?K8bD{8Q#LK8EUd`p4wEwC?%C@pkP#b{V>%zq>wx%N5&vO#UL4BMEpu zB#=r)Le?GlM3(TF;)S~)r`Bdx1`V&ZQC!_7HC75iq1H+AO?Ozy_hzq+-WirCy*{sSl7BSRBCPYX@L_{&EV|%H)wZiY>@Z&0=U(hx~#6VTd5PQLi8o zZB^!AaPjfwQ_?j9K0z|aS65pgpY*@Km#{HjkG5>hJ`=_G3{dcxZ9Ir1ALz(fx|xsS z6ZWoN-M6RPo9}h@6f2#J3u<$w>lhkZggkR-u)t>bI(@)q|IKlDy)cx{4*6_{-0L|# z#nEX^&YC)(zDw(IA6@lAeb~o@HFLY8ukxeUBB|oQC5yeyK%CAdhP2{G8@KcKk}FrP zU=usYFZdscisOmV{O|z=sJq3CfW^FytdTavMwU7^Kgn%AlG z7W{EJ?or=>=ZAwhH%%vg1`+W5a4e-!IRr|)pJtPHjBo6gxZ>8(<+SVJaPtbq2PwNf z&XHHXhcAOREV5PMT4s2%`2&3JKtQabVhe2dzKt5haYd9?tC4=tkx{r`o}?WX6zx!G zHIg-K2Gdlg70(d<_@X|8l)MQk;Z?W@#>)@h-yT~)P#@&M@ph*1OXc&%Yx#4FUQ6bJ{6E-uj?(4 zzqHgzZ>Z~}Xe^MWLtugLLD^Uq2-#5$q0Nd;Oda-&=$bs2+|_$dNY^BoKy};XZAhg) zYG~A8LosP_3HzH7>#CAUnXc!TB(X|;f{#{r)oQ6vd^BRIPhV3W$sQl_lww{NjhRY956t+bk|ijPK2RrNKc5vQf9pF&FOBvrgxigHl3ISHKf zP7rL>3Sq^rm@}jFPEguTdMB1gOYh{0uG@c)F@OR4$yF8P?)m^eiQzyr5!)RH7{`9&x4hH2Nv4xlEEkX@uaPAZZ{jS3;$(Vq5s#A&C5FES=bg+cUfldH?HG; zWB=+^0>N7py6i2iZUD(_h_qWarmv$%p@nrTKF+J%NOt{IDw#SLVaAVTH95^wh z7fqbFI8S`nX`^Eb(}~^{cTm+R+tj#roLE~VGu5~(f>uX`o01oI$eq}3;pWlC#?txz z#<|NPzY4U44|a3#8upBZGNQh=-c}?kTa?}m)z!Qx?}%k02_&DHC(p&!-lG?``kSLu zYm?FB)P24E+K**B?h-%AgB%i{*6hk(vz5$H)dV)AS;IO~!(P-O&%xHPWz3{0&^^AVi=Rx_eCS0o5q1qqGi1p|M`RWlPnO#u0tW-)f;wcyhEc?9J1E zOLWEL>}<_g>DaQ90RNecb~c8)z0UPrZUoL($PZR`8sFfbdlh{?)CNo021p$_sundG zC)W|E{G-cbGCl`S!Jnt4|NaK`a2eKP_o7UX(6h*?u~U9f&ie{-PkvgncK^9P6n>A6 zPM;o*hwZiiK2VVFZkJ-&=x=UY7@z~krTGqdahGhhMzgh_)&v1)S=eRkqTy~U4H^LN zN3b43Uh||u-FHv*H~M^(U-@hVR_h(L+9@@TgR(i%xg(krCBNPwcekdAfSaoZfwawI z`~b{(TVG=g35tKzCM)62Q8mase9B09G_`rRHrd`eR<{&e)Y}N&LjHzeJ`HvXX%L6R zwrUGyd#i(ba7P%Ts$XqWRRPCP-6d^VV&#lOXuzGJumOAk`fD#N8t1(Bl#YU_+LJLW zDoUl+fh+z*_A70A3$+ zs5E7Of44FF9|-8$o#ffQXCgt@6oQ&yYvcu8avP?61RCu`YC=n?wjj|mu8!FSWPYOQbb^LA7tjTfKj6ea6M&z6|9#die{ov;ApO> zup6mJvxf?d!mQ+i1!XoHPMFQoK!jO49MI5Z4Un$cIGsh4GkP1D{B0LEc5A3IgtEU? zyYQV8M!{;D)Cx&98w+UmtvZSf>Z=(OmCPV(BFJX5U!Pt#Wntv4-JBtfb_UeQu{I(~ z&VVF4u@&C$G783)2b+Fv9s|oG?NNhxxx0HK%C#3`{}Ks?0M(MKYj%nOT9Qi(!OdDtUD^p( z#Sm}{a+4P1R|}p4HO_Iw*KF*_Aq=4xU`Ck`nwaMVSYzL zK30&Y%MXqFEPV$YqG8#?q3CsEwgA(Ee6P7Vg+R9-I*M~HYWl&;GMOtWlMu)BSw_E< zFmZ#c6%{q3#@du_0xu$orOeV$m5E%f(!hxD5S#C9ZNH)kThC&S^jqZ(amK}ga}jY0 z-pkJ4w?G}`F9j@&uY-kYvfpE>mv9#!M)sZ^4tLfEqfNl*KTf`QN8n|ge62N33i-UF zn4m1pljfu&9RfJV@q{GcMZ+LN;vvWWPNTU5Jl#E9s-LUOqmLt%4s(@`CLIx zL^DR!{yLa9J=Timx}*ZcKlZvn$9%?9!nSF?>?z(pjbBko*qQg^r=2GTMqWmd`b4a- z3gcbWxd3;f{n5P;b_()iDTdO(C@>WMOzgLMHn=<7IK9N}>o9^j)NAQ)$@b|MU3_8h z#aiCbRl@Nb3$hw)F#5JE7_@KVNQY%o%;F=-!&k0_35)E$VgcXmgRdjfTWDxv zdPohg{V*~n(0JS3+1?$lAFkwg%hvhfcyjx2ym@ZYr{m6g*BZYnA0MLc%F;_au%4rJ zI?p%MCGU-KW%ezI5AN+f$+zcy_1G;Q?#a-(yXX=*l{QS+q+(~cola6VYFAtdJTMk% zueHbLiCKR_4_f%^NY=4frf;O^%r?lL2k+>4un@fz>G-V5g?JgzKB2vTa;^e zG9C?vgZ14Fm?lYQ%$+*RNwp&sg9Bwj-X7EY2%}9h@-aNBU$^=QF{zJy?VrNYsE{kIXO_HQ1n1KUuxg`Sc*kn1vs9q9Y$SN^a zU5aMe$nX!6uV2~UyJ`>5ARa2Fek4#3erk^0iKYs|VJXkTtg__HDI$(Pt}EK))t$W5 z*{GA`avve#iQo-C0P39=^1)8#X-9SxP>z*(KG3R3&`JodD7Z98Xe4=EhdeFE^lDXt zU?<+(%1eIPMP`gpscT0#-?(T04x1IfQ0uEWJ|g0Rnw>U-y~2?>?YpkBR>gJ59Yv*;z=?ieJ8JzZIN7^ zqRQ}bXt{4H-a}=V3#(HNsBRM1-xcILsJ{#X{4qtT-BT5#AXrtYLaw&e{4HZVq})@u zZ-Yi*B=kbtFPN4&HHr5LwS2e9Y#z`}~)#ZucGVnAvUS)HgEeW4&n@+_BUB{{?cCtJ7K zKP$+saXV#GnK!|x^#xTv=bX{1$q1^p){_t*q!ZH-05;rL4_Ghc($SI7=^<|W=>T^t}&m~ zEiF~wi<=K9Vc2^$lj%)Gi+VjoyWl}@Kcw9XkroBF{Bnrx-suyU;AZ;{ zTyDR$e_=R2u|3{glK1VY1|Ev7BXMojN1NbZ5Q$HEXeEfLj+XQ^mk2`f!CKluzBfAQ z(dm*&kJ;GJK54PBB|_rU8$pBB7b2xM;zO7KV%?Db*X!b{Vp@9hnIbm55uXp=R7H!_ z{EQbXd@qkMZv&sUbU}EA@DzsL$cW4-#12IvAawy)DK>QzA8ysa1O~WD0dpiibrS^U zrfy;(U~_EhCjXWV3%AGD2qpP!^wCjJK)+*A-?wFWgeZ7LKxvr(9(ADM0of_QvS&ql>kjJdH2A^*A>+1+?i-EJG? z714>fCF`8Jh7}nI#1?oU2!ad=w2KUVMQB}=-P+@~t=RP)`UAVUMHv#=a5WhMHiFEO zpqtEsB}se`H}~`#Y~&YX?&`A`E$FE7O9T63gI4**MCl7h0IS{}eYlz;U;L=Ts$$&* zyUtm$qjvm}9#Bxxz*89rTsYHaJXOvW<+?u}BY<{)WXdOH6V?h5ulGhqd^$!xy99dM6OlMmZKbDck zt`Wt~cIge}^!|UX(c0l2OHk?k|C%a8RFUN!AscQV;uop?f0bR*wlpw?Uy9^UqkC)j zpBs)Z1oqYtj6-fMrLzXx3!i7iBKFx~=;jhDmbyaaZ7&|YS4q!I07a1zL1bZ7*ou-? ztFTa}jx^II_CS7|S%pR5tJD>vUY$0rhtAB4)M9EL4(*9tjRzjxCc+TIloC~Y#BtGJ zPFgWPp~+!qPQLD-a23CT|Da`>@)PT6%9fe5k;zR{LnId1*ksbNeG$8GsZG)$#`FGI zE6sN{^OecReZz2_&7)0P9;^?5ykKrf^@&zn^{p#8;hl*}-tUXxz7JD2K`Vh!yD2T( zc#Ne2DprOXW`uYNq`>a5BNnzb?Eq0#aa~Jq|Hy8?yd>KwUPM_%m9hXV?5gU=``p{<%qkvW&=V5o(9r`Lb;}B=5 zlYTicUhuu5Az$o}424$ci}K7FHeC5+g;GIfFZWT&-Z{fkC_7u{0eQyeM4TxMKZ~*mc*^kG61CY>Ovw$e7bKs}ihvbCVc?GH7J$gn^2O=lp+cO| zn?Qi*0?ov55t>K4%iYcsR0tKuRYq4h++@d-BW4Pb%u%YS6fJ`C$E4)*o#tG%XVjU% z%l@^jm;)7LyipbLCgpM&>3A*HyS+`=bt92fn zHj)F+C{10(461a#zrEduht~Z8r_Nd6s-C#hu6!A{dO^PkE-@AZW|SEjGeMM5HZcOr zcV@>6Om`Y#B?j|oP>E&?r6lRz%^OmF4#y&cB=CST`XSHh$phnQvuyDUziSq|raiS# zcqx;A7|SuP;so22<)&^Ikorrsaij;=x2KA>40fg>X7;GK5Qi>6RC| z33`m~v4zl0d@wt7(^rDaS61luiLB5^(!%S085b>+D(8eif;i(D2j2J{=KT=K3xNcE zm?4l@G&2N}>$r-B-Vo^HS+Rp8LdFi*0+x}!733KOOszU5J=Fq?W=8NJh{FvY#Io7J zgIuY8&SeQ6EWD$$ncUL>3~ruhS}*p;qyEm>p#d*rPkF>4{J)prFD>MaUN+>_!+S)! z=?ppDyE7@>Wcdoy^BxBQ?O(f^H5n_|NSnd*(%z7bw6UGWcRQX*MZu>5a?$JFG8mM% zx{y^805!XvW}CJ6`oip*wVc2~@L;XG-1bq}i@yc=S2S*K?+$wVp1ZZXJ6_)%l#i*g?Bq|7ll`mfZ;X zH(%erKTBNe?|CPia4Pc) zO<-){hz>d26k(w&Sb6uNXY^n2UQb-w>2HtDZcIloD!6y8_E*oSu;6bnvxOB1L3D*h zS7P4CF4e6orl?)pg`*>6miEiP11G#%hZ&pO7q(X}aWmDjiCs2WAiV@_F^%NmE>=ic z@40t%{ex$&SFkqdDu#VPPA6HX16sdmksl@$Ak1oYj^3jNql@*8>cr?GuJ$8&pjt+* z4DNrVgQWwFG-^f>M@CtivBDe9u;@IknLw^m z0R?M_O3pyZbUNDG8cx*t%LREoTCVW;?#@KqT_4Z0We3*Cz5*svEiNn^JE7YC!q%jU zKz0{No$JX*6erlxi~Ibx-Q@dgX%;O|t~3%O(T2z*@@jia8iM=Kqv5Xb{Nbs-Y3mH^<5#s!KR z$M*S+y9>4mQ8)!O z7=vE~EPlC`QZSujkUF|dQKV06QJoL3B)F&F$jR3Wbhj;907QfQLk~%A#jPI9VudlrU9`9dCk=C3=%J-^&>m61dU?97R{ zP>-%)z|BNI#V|9T3u1SZ|qg#R+1qma!pN}(g7H9L2)4S5jB^D!842G z>aN@nFRCjz6*R3z@E#nWQr8GB*R`=jK8PaYjofZ@Y$9JL)53*5U$ey47%t?|sFZi1 z1R@F|1woWJ8U;~l0utU1B+74~WZKYBC@_Zxfy)g93W95A=o191R3T3y%8hR}G^c-N z-I6M^>w;|Ugr4E%f{u5Z0dg+5dbdcaK9AhdTTe=<>PAkoirOO@dUeSMeIqMHwrnuo z-GlMQE3D32C8wb-Yw6+s%<@K8wv?}Ryg!*r+vu;te#kct*h4KZP&b+(J+pjr2)(fP z7;>S%gL}5Pkz86dSYhEj2V2|WN^$RJz0;IE^vE1ydXEY3G-W#t#=@Qr8kNO(&lpWR ziS;EtuYbf>eAATe-xnI0uzhJB%BO@V1BdE()mEqsdBitG*}gM@mRJods&NxT%@xxq zU`=T5r185($rIZOunB;$wtQy-?BMwgG4#hGg$c(gAk{cQ#!Ot54;-30tI8$!Y zP2<0%G!1tqLp~-gWuH#DR?#%pcVpFsbHlMov8u2KGyyos8d zfJVJ$(4Y1<&+U%4`_tZ1Z}HTvFF5YP$}Gt(L7Or6u;!U9IcAlGt~0AAJVrN7IG*{s zI-e|IEZISQX4ky{TPq|hUhNx%Jwr^B6(HDrs-1Z=qn+t+ygeL@`qSacfXN7TZb!3a zKw}2o*MiuWzUMyMv%~Hk?9r((S7Tp7K|HBo|aLZJR-O4*dyz2 zdXW5W;h8oYut)B@a(8K(cy9cBZ8>80`HJLynkIFm-y5Lav3v5Kal=$WR62m;{#jWE zVrGUwpIaZU)FSIOUB=kbcB^p@t_)&0FMvKlO2tNPFY=tQTJmz=z*CBLmd~oW8;3xSlfAz;rpyAGx$F=I_gyM(;xMWe7w2*`u@5=0J$<>36`G!km zakT?4k3GQ(NefA4>5w6bqe#8%NCD4=E41}~LP(;})pkgte5!)2BnZ^(8o40}v=vYA z9x4_5IhU_c7pSr+4qG9|c!pti$L7Ei9kxOghQn5tOLW-k3hae}+0Izm@ZlsUrhZi> z_ppS;_h!f-bP7R6VF^z5dEH&2BU1>(ab(JZ$&O52qv|;+{I&kT$?xB2f2IV>cc#B{ zU**T>Pb?+4kN;HujInZKIDPsXUc9s6_musMLy;Wti$a?7rr~OxgV&t5$3-t&oqMA5 zo~1)NZ>?`%x+hO`-U?ktT}I9)s7H~SPf*v$b>8Xu1XE zlutMe`2-aGiBXC4jDl1Ny4(E20;PatVnQIM9Jo;}AP^G~s@K6fbtN$nQwS73UshrB zge^G`Q;6cI8^I+l!4?pRJ)E46mH-G|SV#{^9!XYQNrenZA~FW9u7w1quii=5cl2mC zI5v7Zj4Go<+rp7VuOzphn$MIgAcP{ma-NG_>%6d$crmPvno!g5P~^_tmqwTw@(M83Cz4sG>A+F8zD#zf2dE(Rw-*xX12DpoRRZQ zPyil$1lQ%j*g)ukq%rb{y}p(Xg3sA2^L>>XFNAXaF0KocR#;Lxvl{HX8_7x#SPcg*7H1o(W1X_6hy|WcPYg56*|49oA$R;K~U; zS*3$o{CmSuCyY}meq8+~@Y)iwNDc#RFK1`)xn39kD1$)=H5|fXc-{NhAM zx8z6>mH4{JWe%%gnXFC2?)`feY#~tF%%N7q>e-x2iPptQ8x4B^8(eQMB#!{goMXx{ z;&!pq#JFAlK1&-$CP5@CUX@gq5??<#9s)vnEUm50@R+Qy4!tZ0bim3vmYSB{)%6eY zwkaW@+2llkl=9mYdAonerOO`f?hGfJTci7j!A}Th@gvS`(Ce_M^FFr@^~Ye`a?VKb ztM1g}d(|Gnd+*SIu(f7hTQ|Gv`0uIEBKsD4lZwmX2|3;vZVk`E+-Z<8Mb zL~|OFfbk)LgnBK>j8Cxz!UUS4`}#SBkyM)!U#I=B#a3nx!rOf(a9ISi^4ShKo^4hj zp2jpRIe{)!R~!k5QQ0)`g0QIcN^%7A3U7Y=ZI&v(9Tl7)|DAvJI$G^B_a1Vj#1LDt zGd~j&AA*Q8&Px=FWvHkeO-U4s6^wVV;6$<5Tb!enqxUVu;tW z$*kty3fJD2=-X(+BXarJnmlYU0EcS>F#G)@&rHEuHF>l=k3Lf!7h~9iYNe~dm;Hkw zg3gUv1&L^@B8OQn9z9(CT)Jey7h*Q{Isg2`Nvrkw=lv z2P!g#YGxpgx>$Z99~61r+JSxD-eRw_uUPFs_ks5!v00}N*zA8f-qo*~UFQw>TOD$* z=l0Y%DwBN@bQ*&*TjkUDLM-mnRV~zqeN0#rs-s655A+U^pIE)Ncg?;Z@&74i{>sI< zOWxo|uSHVDflCH`o5OMPofYmDEv}je`M7^bR2)wl#Sb4~Fm8ET4rxJLB`ftxiDORG z5_y~dZDsW4s0F8cG)UB-eFG7LE?#ni;|O?e+>;jrosf^8zQ2>l?{|5h@lU6VJ9WdA z6`gZlD0&>p%*3l~)3wIWL@d8d<2tq9g5NL44fPEeKODrlSvv6}h=B3Kv6M!|5GekB zmQCI^zOh^4f}0gbz{@KX8>H-Zog=S&gD-+z0adibi zMn@QWS@yu@HK=mHKz4X2ol&JdJ$Ya}ZKhBodO9zmUSlgx%P<)C2pP}tuCa_*qs9e?z>aI| z68RM3HCEL*kQ?zL6qmtOAsbVqq<4HPEV88e=2PaW! zATG;l+bWHf*F0f`tQ7SjPmSl8(;9mxpoQiiNr72Vnd&CuZ#EmReh8O=(o zp{n>aVyLRGDU~=aRQ)PaSSP9C*;15?s>w*;gm;2qqf|&Wci6&1&LEIrg?EC|cEUTc zJX&}sS7fjMe1=l;T8sb4SCCTqk{lTWwI`e?b#g(nQ8dRS+%NXWqyEm>p~15|x@bh9x*p29URnqpy==%Mko!cM z@E-Ew-d&k#Ld*A<1?h2s-Tt+!0bsYVT6FY`m*oh_BVDpVLnp^;!|1aD`RR3U84SwP zv5A-2EkX`vaN#Tk{cetS2BWjXNlvwDcJ*4a;J_T%KQT9|yF9n!ByVluRJL~q^}O@- z-9g#Lp|^B4b?IS_{C9~1c}<7B09#|~D$D%s#uK>RxOVLt0n9B5-FDb7QU&?74!Ikf z8{_`Y`QgfB+8v9?HNqM==~=H{dvMzTX**&%mfn}z3&wzgI-_O~uv5%HBS^VWCj zAMNJgHOvtUWkmgKy)92vws?B{+AvRt-fYC-a%$gLBrn01-V>L0`rD(k8`IJB?0vm! zwZE3_m`nU14`N7sTC+2Mb}hLaq9!04b_wf{S9HjWuqAB0|KMo*!nW+?G(T713kj^| z@q=CbV7lf*@9O%8vRz#++dIfP26YZ$6-q?j-6a`mL9?2k+uGe74@MX3yXlG1#kMRA zy05I|OYrB)I4<%REnH9=iFK6Hp-42=%eSnl{GeuwxxVx}`Lh=Cgl9%u!`>qOvqTq6 z&d%19bs6HRGXVdYj`p^OlV0bhuH>yBsrWX3->c~Rp|`Qru@mYwzOZ^E$o!+9$8>xF zM!{dCssH{KwQw2QW45A9j?kmXnYL5@ll^c(?#T~pHXgXphs5uR(YbTO@vz+zz()%5 z%k5Gu3;ix1YcFXu0ST|{lAYFQHrCT@5P+5syKMbvxY;_va<;$K z=dJw87b37)uc*~V3HsbzRV7sptdV^MjM80PSXej~W-LG7Aun!C5&<_?H3Dgy$M_DI z^H#i?@SD!RYLk`l=BR399^Pf7Jds+w+nDa|ovcfWEy`^KZy|m|P#=q)LKws$v90Pt z+1~0P9^4Q*+agunf5;Ep6ji`6RBuUJKCyDnE>ug>KZfkTe>mLR9E`RhdcIM%mGZwg zprF6@++yRN*PhT(FjaeJ%#w;ys5RVF0F!^vro~`To%sm)UTfaj7$*kfu3VTw@|~<^ zOZ?2C3j$+k3gfHN)2e~)(l9r zjHhGf16rQX$=K5n66;1@g?tJ+`V9Mky|Zcp@`;K7Q$(OZon7`rp{+&{;CjmbDp)T^ z6wPOY*+{b=6&Q(G$pr(-Y(AW57E1#WX69hI6+wuvB{%Ran_Noz6cUEQ1}jb;Ya$gws& zN=}0$GqDxkZ{E&04lVYlix^lQsgD{oY!)eodUavuuS5;TC_)UC#tI#V8f#NCFmI6x zLWTV=N?}||QUeiog}e8+qda?Q>|Y|m5TIIebxlt(KudCIad7h%Q-}YjPsI>$3v!bN z0p{SO_H;2k8Mv&~1<0qfp87W< z7zN{nDu$eet*B!jw(CvE*;NhKWxJ4?APh5jFFG&~^$Z%AF4Pc(;ki()VKkQOQ{Co- ztD|+}MzsDtve{vaSRbw~ZJa-M(8cmlq}A^xFY3J@Gp(+hPRSwqesglVL5fa>hb<$Dmmr{#eBLTx5|&keM{c~ zhiDk~a5#Ehe*sx+l6DXBO>=V|0^Ppeic<^X9if41{42CcAg{?}uB1#t95vdadKvvm zLdOj*R`jSDHO8W}4ZMhSEM=C2s!ZfUlLkhFL2SO++I~S3x}L=xX|BqfV6xAbO1NH{Br< z4TBVkyBzbJ#^w?*x_h`%kE_h%g493%b_Jiv?{_dfmFysue~_R`DgC^|lnJm(0zzuL zSU5*>y{SWffy!G9`ny3|twbsuDW7i)V21CPX5UR&xJJHRkW>654#mQ^K~{oa`xPpT@7K1nkWE@!QVR13fRJ$T}jn&)}(kcYFi6W(wtm$=>c{ zxT(*TCBL1bC>BP6qVO@XKkM1xWVm&1h3nT*1a+v_N|`0Qr2i6SkOcjbw-;nBSYh;S z88B!+%^!69_*z-_^jZAL>kq-Jj)&mC1MuH5_)mS{;cAG#NFI#|uuRm|e3PB;;)UEI z4c-CQ(#x6%t3bj}Zp^HFJbCQu4KSmW;oS@P;Q;)1OuCDitYHjRkC7oYB;9}l-= z;h8lvX$@9V**`SCPW+Im4!eukGPpQ(=5v_ z6b1RfBDpO`{snhONlAct3+e22@a zlLo;>Qk1bEkG4W6o=N!FBKbu$lK>3VjRJ^bxnN#4>S&&g4}T9m-qiJ4n@ z!7qErjNvM60u{`fs5&=ujap8iO1OzSch;78g6gJ_d<&;S$H0W^(SO7MNgVqInds6Q zHOkI|GEIEv+ZP1vz6dwI1VBK^8NeEI+Q=mR--sW!Lm&ko< zOOW2_Kp!sc0&2xlkw1!QPT8;vj25F)v%#iN4kiNMU?!OzZ|-Qf|BSA?HN;Ffk_@)+ zssI>eA67(T1_w(C7(=)-`B-!Xqp|bU=)SQKbw=Sxc8vqzQ7UaWO0Wzyf*IpPSdn+c zAY{yhHe&q9yHSIC6f%_|d(sw3ZfWJs2uBUwWO6A1S>W->Q$m8Ojoow8?f!!~9FW>4 zIS?!_T3v2rfxJB0jR*o80bvHS z%PcXIT+xHSU`tCtRFWx>m8^VW34dY1t}Nm#j4BP6yPs-6Kzro+)$14plx1ug{4=Z& z`sz-syA|Z|tAtXx%3?Q`U75Yoxdo=Eu)>sO6b>61IaMCH#(YwVeivS zCN~u=>hlnhZPcE8q`wm(;GZHdh&Bd+59N)Sw z)L1)FCZXaJ={N-qMq7v!-iVK|1Bf+4`fumfIlfWEhBxB-@I#ffNR7{UvcflcgjpN- z+R7!VK5iphKgeK47G7jT#uQ?QA`y_f0IU=nx`~e)B`|>j*rA&sFgJ7)3jv#BLpS-m ztfRPaJgK4Ex1z6(f&%(Ojhk2wsY_H2J67vwqo4#qAg_y#PsJwTi0CGBPKxoVjEqeM z0Y+`eE<8Xl{m5`z5XAj^dNeAQV$6*d2>IJp%Wf$Ux(-+psN}WLfw&>+oOy;7=?KIY zxFZOH3<RBV0gV+*~ zpkvd<+uZYhL$u4|p00UkwCiD9SsCj8k-XMus{bPK%#leI$#k|w@z*j^*)^ir-A*2s z?v(un<@EZ0EjZfc-abNh0ouKV)(bir_5Yfx#!*G4cf{Fn{Sg0|Fb zaA7#U6j)nBP!74Z$CGOY{b_&m93;8=)80~V@zkv^I1UB%>o%(#*ye;rl#||ma`oy} zEbl~Md*S_zn8&_<7`nN{jHND6PV#+wPXO;#r`HSsMUoNW$ilL)6(y}^VIfT&sirpe zKzy8;g+<`2tKW|Pf1{b}fV3=uo_Uc{Ox437J&~(%$D`{+7-E=GqDoIAFTD`ZC{FvO zW{amte{5z>zU-iIRT>j*(}AL*%NOhQL>@JCQ+{gQP5A~IkyKseZmIj~Ml#b@5Qzmg z+EhBWE@D1h$<_JxN?!B}H~EaO8LqQD;=fQm+ zQ#L^}fsnf?4cmB^i^gpaiK(9EU4e(=svJ3QL$CD3!+`|I znu`PE#>sqJqz6MZk)n zFmOkuWr!RtUz{EuD#RJR2?U5iSZ;q9NuYVOyWDL!e~G(Q2o=UvMpro8tgts3_Fo~A zIZAMDk+nV{mgMuD=3KR#>r7zgc4Z0*pv=f;;ui$5#pRcXQD__FdA&RQqtLt(qV9}- zhmJt&&}P7y7MUuTK3&V6q=^ew;hkFMX-e8;sG6VJ%1a(mMPPIpjYN?J*j~oxRzYAx}V|)4gvD4AfLmr z$RG(kpp1UVb9(Z?c-kym+${Ix&x7P&zAED_lYbaT!?==r`LbK@;E>RvKuh#}WM-hH z>Ig`IXkdxK!YUHP^dc^1NRiXji2Qn1#LNV|^1|%LlMuu7_KP2lh!TU}kRf=bF&Ip# zEt}l8*I?v@aN@y)R{z5a;eInKgwtu0AzaMkHZ3RE62uwLIPk{ruxN)eLm)vPW(Xt}%?yF$I_QGe&`(14e*`^z}s`1cb0rG>oF%Z9vqc#lXXogt@tcP1s1EMH-I-s5PY{cBgV zCSyfMbGtT?Cg1INCKUyr2FOLPd&^)@-WmqxncYsa&02hYVRp@0PT(Mx+dcsY5{pNY zUGlGJ+}_?D^!7b>Yj=0NzB?%0lRILL5S8b33BP+enNOfU@;beZStlF^v!~9QbDegb zc-zuu9@nH#C?ntZO_mDYIn2U5&j!tnr+oKq@33jnWptTsUZ5g*)kPi7bU*FPT=0A6|wY9>)^;f(U zzq%V=Z!-IQy6N5`{jFV}Yqh_6Mui1`gPAR?KnS8MEV>f&Ms}&5pSbW6wWDQdhd{slJ8;6Qb(pxh z4egdMUGZk>;VxE4S?{@bb^U{9u2--&=qiSNK=TtpmxugeLIJ|8R_EwFYB0K3->6QE zF5+rGk_W0~<$K&`6_Z6mdjys$R4eec}f&+fLJ)RsDm6(z{t=&y2Q)y+!&b zR(Qi17M-Ux6UbF60C5FT$r*q>Ohn4Q9{$55 z?=6UX43fizZ9H(HKOPQFjLw}Kj)(20dEZ}<=ck4(E;?n2)Si%z!{GgWhYXM@HP$_B z#nV=f(=*;DQAnVun)R6S9}DtJ$gG~6?QiwRbwpd9x+iscz4F0_rkM=c!m%(z2EC|G zr&43l0KoQC@2%V6B9^P;+~aO0z^p;EgUu{J{Zp|0y`}FZdbt``P}!L7?wza?-slqN zg`Ej!K_J?}wjB^Aa7Z2Yae(ZW*lWTO`(;Zq?CbbK2G!o`MMo#X)>{v=F{q4us@b#L36515hbI6{~7(oGYM7R_e{z$7aJcjJQe>mLR z9E`T%@^j+?`9eZt41N)?_~lwkRU9h1Oi`pyYf+tGs|G>!z8NH6FVNk#XaNun@((>E zDMa3c!dnV*GF7s-hy9%vaWvmfcw53bq>!dyXtc&Rmh1WQT>@xxCpp!-Jqu`~kYNMa zkiYH__xy6lRYtBaurnv(LOr^I0XGx<6vNE;R+(wf>=s-By5yTE+|{u%ILGB_N)~kL z8x)rm!iP2wg8XxljsTMq9D$8#Q$>StG2hKV1j!v6FfncJEEx}jxmyhjGvD0?;$S|} z#qfw41|~V(Ym^l%!;zj0nI)^jy|j6nb{p4GV_^dMw`k;a({(lRE>~dWP>>%ayx=fz zw#K<6gFvM~)K)ycnT46K3Uwshd7Ex*IS$+Mn#FX~f@GIgWuL{hAOz+WrUfBjp?DSq zdMYdnLUc)0l!^uGfZzy$I&+a7@(&qhMkx&q{K^$+cD|nhL2v9;;#QI&CUQ+po6-Ro zazSw*bCDvK7}IAK$<{vKIL+S5_OiSOFoDqCUp+Vqs z1A&6z+8O!;!75eAlZbNTn+?tBpH{b|%6K@*N{&^Yl$HovDj=l;m!)@Tl(p{`xzOj4 zJ9_I$xlrAN$qrULdh7btihR&FvQlKr2IJj5811s6Gu_g|{b}Wmu52k^>v(@MmA26< zP~=0taljsGd4amo4C!g*lSAl*y~mIX{T?DcipfX0X!8gzZZMQa&X_85%atufLD@ zrYPHYCeRYAVMR4=VyL-d8U?Hg&7CxU*C=^@>#8QumS>B}!jO>vNwR8eQ;j${l{*y6 zkRUVCR;6PWK-^Qm3}?zsx@r8kl&0a%WXQ*)rR-DchB!9=@%u+5O9B<=TJcnN=g4AC zXqV0~kq_?RWr?4OcT}>SBL<*EDrXp^CELbZfDg8+l;w~HP39xF{>J%&i+%$phQOo!v`;b7FC4p#Ctq)gfk@cD`V{B==)wl;&2C=|1YA-&mjPz1%nO)F%av+fN>s=p86go>Utn zZ0F3C$QBqvg){G~aJt-+1bYHz7*`%l-LI-Ya=(7B3Duvxy|U1~LZjy`#F*xV?yL|2vZ#LTmI^gn?S>zDUWN_%?ZU3OR>>`<6h<^h9t~*S7x7O>3AoLu=oSQ+^ecS zJtA^R3`u6`kRgdX($E7XN5qmCk|=bw9g--Ys$eS#0yVovZb$-c#S^@TN(Fz;(1)VVK=PcGwD07!F%mF41ADE4VuG+8p+Wlbo3PRhitw5*FW^A%oB<1Q~_p zvO6+`KpaP=EST)b)HUiRyIzlFadosmQ-b9?)8DzT@?-QTmJ-~@e=2{*Sh+ErKK%_Z z-r4Yb%KpWn$o+X)k?6c>xLW7nHRs**%DS7hmL?OO_beS(it|?JGU_sNK0!T-)O>=v zMy~Tt&nL*jPL)qU(ML=D)T~W0%9EZ?;N0dxNHUr2gB zp(Y9=JW;rm7@nvrq=YB%s3`H0CzT;7pKuuR2`Ks_<-qSx(lZKDCFpMR4-1q6l8Fg{ zm~!AowFs`d`ni>!QP5QcI3~s>Tmc~z@s;yj>;fZF1tN(T!&x;A z4@K_WeW_3ge+HAh(O>&_?Gx0#vbmYe^0s_-5n+@~TlT{ldkD@En0cLO5Sa`%LXaH( zP@k5qQdWKl*nnns>4UADZ-N5w;3K#$2gU|M47BL z*u$do0FMk<<{VRw5x0w-CdTdZ_gUJ|n+=j8q(|_tN-9fc7SSk3CP z$s_~35;|bz97|10@9O%8c-xea&}?!-a@u<+ToXA7wt{W2kRvV+ z@7CsH!q5%<$>tOuSM2gO`O8?h&4BSCfrNT3iEiJlE{HZo_w{oMBdInezE1mLi>=HY zgtz-n;Ias2<+B}fJlm{5JdJ5saspkdt~e49qq1q>1z}O?mE;KK72f>#+bmTv`~2_x ztJl$LpSkytBPE8|`Y#oNJ0uObA0t}A$iH+*qcO?T>n=G78)Bzj<%`!L-+)ve$%_8!lM;Hdydm1) z+(u6Jo)gjr$6$!`ozm55^Fn|7n}g&7N!OV=Z&lY$Y)`YBXDRn>eo*atjeU@OC+S+= zd=x9588-e|m!yGQ_y7D(zbxa z(t@~3R%+!BR}i*D-sW%j5WP8S!Ra0i5;bVwK%l{VmZ;~(J$W(E3HkWx`#X93e$V|G z|8%;zQ#V{$(K+XZqQ{ZUFSTB7Nych=h31!OT&MP1@cZSsp}qm*hl4mbOQ(z=l!B-; zL~teBA=DrNq{QWGOyls4Ax5NdvzNU6lnsd?w9*PZ8cDv4zSH8Z*K^qp) zsxU2cyxDvipF7|XE3eoB+r6_@r z%OKb?(^$tlit!yt0k6VEP+orVepYM(K^@40W9^#CFL~wl*L0E}`gzF?2iQw9&X?yf z0%PSojT#K0vB+(D+33`0@ZEaBWCYvG|kF(}T0{GB0B z`ph~{ipBz&Is_K@2Fk{=K%AY})be>z3V8{+yZ54yQb_Pu)xFR!RH%Xq0e_kouE| z(GIf~jer`LwKSYY3(=){la5Ds!`@)}tJppPWQT{+8CBZTlLyArW(xJ0F>wXvHMZik z41;lxkns%f8q0_^YMdGZr&k<4g?Nosbq?f4w4uOda25HY1(Q>ae|?J; zstU@CW~J3oReTyTRMpp%N}Lv|eibRKlT`6+Dau9FWF&CHJ3+8fDx{h_Y@s`KO0dE^ zL1{bTomd_%ypt=snudAp=U0$Y`H~zN1GOidD0Ol{vQap*QO=)03W5ZgB8h|?1c?tI z20?O7_N+)rtm@Yf8#X%NU6Ugl#xw30`{Pl6=j<@bv&YR?d?@RBX(4pSbBPwBn3?HuNZ$Uns{xX?uv&D;BVDpVLnp^;!|1aD`RR3U84SwP zv5A*i>LCX+xNw$&em6%ugVEXHB&S+6yLv5IaA0#h1Q%wKSG~zwTR4^N-9bI?e0_IN zwsGh!oek&8m2pVqH68K-Y>la_EZlC)&m=s7n}loEt`R8RqR>FP&TRRwb;#Y=+!*(F z&JR~6)Bbome~vVb* z64#CqYl~zi8drp2?5H($^2!dm3)?I_ceJ&&Qq6U#(ft+yoFhM))Q|Cha(On-&iCs!Is_=mv;Kwqq7^+(e&(ny=%3 z(={J@SJyw3?do#b-a%dt5w+XJ4taN%WTXYnYI<&KcXvD(U99h>Cq@_BvM}hrvX-wa z%*c?xXyJm|NUWoj4n?A|UcO~b1c0jIO%n6>XNtl`(8!g550}0j-61G8urFgkoiYHkLmaVjDo*NQ~&)f zYT+`p$81HJ9HB>%)f;O65!V^If_%6j_vD8)8xLIQL*ntpI zThf+Ktemq8jbN~vNnd^p*?<3VxVJeNZA0{Y;{tu3(wwuk+l7LjTHmGwT3b4?*K7L_&- zsIgTSJv0r2UQ3g>RfN5>~UMR!B0tOVIFL z^KnpgK5s@xrKXc@BFL_Ad4E?or%9uk0X1^04Udx3AjwQ@h4*t**o%2P>IRaG{OKYF zmPhKN1`V4UZEcsuQ!nl;A1|sYV zH+OGGdG^xSzeIu|K(*xRnx0~SmgLgn;2f4u0Wi+xSm{$S1l)q$qyhQWg5^Nl{~{ynnU zVT)KFt}bnyztoMP$e7R8)RA@UU#3`eCkc*k+7UPU-fu@GL;#~(G-!90hXvV10UpsT#uvRqF1=#m%2K{M&^BhbO?N574y~R_v zzTmhE!?q+einjTTr-W^@IoVUpK8;^d3D}wS+7{EQ|t0;bUTd z*0aIMaO>O(*RP`p>QJwhGE3GlI#-z``}TsY1uKlcEdvJ4USU{V{%PLt=f~H|y5G;@ zPhNj$VPWAR`0oJxcMSehAAfYoqcH)NiMpC^vh!WMkXxj|JK$P+SrcIuNchR*y3J~& z@aBXfV)p`!r}C*G@~$ksvIpyNqFuAg7RiTVOqqEL;={fDXZZHKQ;*%p z!z~#)cR5pvW!Kr$V}~)&^!(X#s56wG>;Wqb7>m@`+T!!TtUsj}t!vl%&Uyz!V?#@xKpwZ+V_G6OMcGBb+=t%; z*j7y30Yg8-A77q^Eu{unhw)e%B`BPg4AjI~shE3s|9nfFxz`!Q%oa!Q*)2M?<`32X zX%^j4R9qzQD9G__;^6S=7EONgjsc0`A&|@? zoQY;e=@dBRuOJ^-B)8?rpLW2#0~^(OZ_wi|^R+GlEEC4L2(V1pf^cERKV6Xj>nHb( zM?3d#LX)A%=5~K_f3I_Jfz}G|9iAWU@P8?mA4{$sZf%YBChGqi{Nv8>r8B>K9oaYB z8Svj+a7aam+K1bA2E!_^#jEeQlDv6XHwRktUYGWnMSg)m8|3hclkadjb&iqL~C>pl%c}yd9#-@q>Uk9v%kr(wpkl()d(yNxqrOdjdvh0cvUNiO#h0v^dE?SXW}b?~E| z%-fEvD4<*`^K76|lGMRl!5IZt@qxUhL!OhPdbKElvlBD7@`7LXkQu{O+5{?;Z*g^Q z<{Gt}K$UP4b?&S!@dVXPA^8?gg^qydvyg^=tz(P5{>y$<-OE8a56s_D#imsO%EDJ7eek0LgzU$j`@J z%4_pf`6zH!)y*qc+G_rmF%~k#wn3pV5_+NS7gVd+HHp^=e~d@9y|wEhs$EzmyjuCA znPQ>E*iiUg&S?O3mCemr1dcTJ^5kcc^YSF17M*>CD{7I*4FXdxbM`2D$6|?X4l!_Q zQh$TvUhT)me)5!j`XBk7POL}>%8Zeq zEisUFEM+G1@+w`w+aar|LdE?^wt*=l71w=AQpSyut0mTh_pJp<-sc2E3 zhulD}5w(cTK0^$9uTJ`tiF-9_`t#!W1T)Fmp19jo=TQ4rVQ7_`LeqT^Gs zNjM_9$()m7d@3VjQ$c`H8z45dqyWS8gg`I-$Z%T_#Ql4EG%A*2%#9TY`P)^??wORe zjW0YBC9jPR#0^pB%rmS=M*H zWR?WIWELz@_i1@z7R)p9$1!j9nU5B9RPkj4`(uJu`PxM33r7H}-V=Sfnjv5ORfSc> znhW-vvtmbW_#-W#pu)ma=?GjfN$UIHzTN~@ksJtJT^|oR=g23@hpUse7cLk1#=GSA zqHl?ksr44bKd5JoU=CtSK!T1<8*llx-4B%Q=lzCgm&ZL_^UfGiI(B5MI|STN@MM!%6zZM+ra*uUT>Gl7bs>V@8rgy~I zaQzVfkXrv&*)z!||oS+8Tm#2vSoT+tBTW_cLN1`~G3*<`OfO zxf2MsH?tpCDXi-3i(5y2lstU*#yl5LhhzCY~x9W45(Nca+onBLkR44J7Qr| z(GCz*71y;u_=y}_c-`6o(GH~pCAL6hO+gz=xI>``s(_JoFC44@{c|1iLVAg@Hl7Tl z&X{erRlm>HKbz!F|V(`#3RM@h>8&tDsk$0~|WcocC1*tDB{CyjKG`&kFji z8)yF_KEYWI*jctH<$?jO*yWJZb&>aG8GA)^8lepi3n{KPyl5{J`=Hl<(G&-XB*^sy*vDa&b;TZ(oMfpP0IE1lt44!OpDBF!CLYtFh}Hi zQca^;`c5tLG$m~^RLxIqQj^Z_TEABLOAhYLJQ&Sb+)*C<8Nk#a5`-=giC+vmKVASdW`O| zh0slWFgtY9SIT`2_lc~~N7BOUei;`nlPc$gK!P~q8B4Wv7%fk3EiVKT^kIfTV$sYH zNUq~58hS0-$FmZ*k_Z_)U<+7A`c{x<6tL{4T7c2a2p$A+xWR*1HamEbE7i}rcqVct zn`C*{MQ1a)rvn(=JkPXV?2kwNowGv&UdHY(8_nUI2>I^t{K>Li^XQW}V0iHg0Axy|g!EW0Y;D@!gJRQc>_}fL!#tw+sg5tuADh z1VGJhr`cvLzP>QKW-TXh5RNk!UB4?iHfU}<<-4~Sl(to9 zxoPQi@CEv=2g9vtpYOV517X%WvI(ccZ?vt4 zA9T>+hz=m!6k(xV0t@o)MbGHJ;Ju!>wA0@no!yv@U{rALTJ5i%QDMQ~U}g&|5Q69m zi>}1HkzK0iCoZChZI_Y`?U#QCPI$ErQ#rSx-SVX?-b_8*#R@6wJ@>AzfAGxp3f2Z) z#jp>k^&rn7a6<$K&`6_Z z6mdjys$R4eeM&AY|3O0O-7K+ZMq9()BK;F9yx|Os&eNI+G z68SqteLT;W9atm#3YbN;xUg{Sgo@#(!_u$YtU7IOPUyXV(pw~* z)dd3CT_km`Cm&IqV1J9+oM^c>vz|!>Qi_$ajT;6*feZcd zaByOD?%Z%ZY&XsO{(?L|HEeOwDNCgGgmhd1_wPGofJ~{e?qMsQwsM@F@ji({BA-*G z;o<(VAispn>dD#uR)1VawB@OLQkT~&AAD$<$&f9u9&$YMZ#QJni|TYLH5Ls3Y)|#x zx*aZJxjN20um2H-k48J#%mUOu1>4_S`d*@!t8oRDjp^>*$vWYUE^!e&gJ=iac2Hzh z{iGlJI6!tw>^0$tg$VhPd*ti*LI%~|>P1H@CsEGDF-5RY%icbQ?7x3F+}j+CwjuDm zae)Bz2;1j3?k?COMByOVXU9@flB*i5FF;JchOgHQl+dQ2onyWjwJNLn7&@1yKT_|AR6Q!dPw2v1@e}HoJ^JM z?O}hXMI6ny6W*3^4k@H57@E5ThUQLks&{)9hDP~99)9MpJH$P|+;NqWLR1_^!KtsNpfe4a2Heh1fR1+WW#hS3h+^vR%neT1`aWJ3gVt7Oi1CyNY zHOdN>;Yd%0%#u~%UhHPLqur`nJaq#3w`k;a({(lRE}Av6B=%diL`~%_$PW@;aF{n+ z<6M$Kpi&@eD<0ovz)V<$I+~A*i^I0OW-%SLAlao=*=KPr2!VNpX+a2BD4qp@9wr_b z28^oce8eIM!}1RqWkx9t4g5-rrZOEU`uz+DdSkB=w~`Dok!xz&ln%g<3yK4okEl63 zD1m1d$<{R_>?vkz{TwkqR4n7w;LUs$k)lVaG}rFEU`5vJZ+>u z)FSUf2}Be`3W6wcGzy~9_%quMtP1%JluR2M3I*oSAaJ>XKtV_xG4u(7Rf1g@n5PD(n zJYER_TtR;a_iS+^xwL4o!oqnDwzk8S;@;1Crzv~rkvYWl9uqK(JJ45iA)x}tcw#K< z*`QHbjQ5Pu^s8dHfz9GuD>ie*TU8ELB`%=k=<3*w&o zWjIrA(oN&Pr8Et9CPO|ZEoGllHzH1*ct<5m0u|?4@ll^ha(7)j%ga*B9 z*5qH3H&Js_9_Qo9HG}@NzjH;dpyE81<*al>w6x8X6tV zmVvkdb_dyGf+i2-05kGU-*niUS8SV^xmbjX1>;7Or)3l;{T&<+3Vk7eTX?3;2JDgh zuH0RkCY~EVUt12b3$N2YNM4b=Pt&9ht=f06@r-AQEUtFo%}?`sLE9a>C+`_IOcg|> z132!Vm31IyW(f4T_2Eh_vR>0=j4f@q8u#GJAQpH=?FFro`QCNYS?FG& z(eoB!O!LC`6O;n(NWMhK!9WH1Er0dLO`zeiI;haAqg|ymD$&l ztIb$LeQexawniz5A;~NqS4v2t(A9QGqI{}?tt1H4>>9Zt3A7bY@E$4^{5hAeP#37O zDGpm9$9RTecE{?e(j2xz6o$iAmP>Tl>I&?IfjMJiWy6P)oS6DmncTw?7T=p8gU~4i z8HGhn_IZu6Q?f{!BU1>(ab(JZ$&O52qv|;s;O@-=M*A}*SiUp;o%KeJuJ3XIJQvtVuDxZL&kCytWS({`W>@~7@hJsQCDE$iB?ClMJ0kg^a)ksB~L0tQa<4@ zyiPQTOa>bvNDhB!0Ftdzk{>~F2rXyid=nIa2Oq(8 zIgly{J&-iUFZ+TcbP#;bUYYN!)F5XW#0wY>z#-@`8g;~gLwu+zT_lK+zQW}R zI0Ow~JCidX@Vk{fid#WAmI4S^xg;#9-{?~DOi+5UPw4k2yVsj~b3XKJRaNZ0IRUPm z;FDE4sKviGl2a*uT>YlXYfD^1VS70{gU|K4@JAU8La5JUr2wS^8?Imc4d(!09;A>KA6Bs80xh!MoM z`-fb*?BVXtaI(2Ix_=n_gfI&3`Z_G?yw9ye{V^D~oHG*qsyp@gUbRQ?-a9lPY^|Bs z*3D)*evO6}*|*S}R9p^E$nnN-Yj_skgjD}KCC4U~x-5Dq&e-oOg{G2^%|?3Gi#G zP-^;;C^xV^f_a6Hsu*~EkKE^f=kN1qwa;vJl^9|xcIH3qMBs+-nt6$0u?)Sv9!*IU zixrG_Fh~@;9{ox_*%A+FxKcrs$UbGIhsAQq>0!A-4Gc8H{%i;i@-H3IXiW0dx=T)i z$g|U~@^kBuZ$K)KWJQ1VNlA&=5WQ_~BPV;$3F&QPaBuof>FQxkb`FvcBwc6fyj5M` zmann(r`gT3T#kQatvn|CPSUl!`6yOAGj0-C9u0+$s*kA~?}rz>7ICThez*`)LH^y} zpWYw^&7IQFqWahJ*~X3arx4suN4q;K)i}A8+xwS><5TbienoanVu;tW$t-;feHdfn z>;$qz-%T6djw|n`A2t|(n`;Bm`2NuptW}dog6qp|_L<^3+3`yT`Lcg7M9{fWs~{0= zRpij%;?3pHrAr2U!M`8DVDdTt{Jn&Y@qDylYv!IPjczOL_YuV zy0ruQy1m6-XJ4_}d1xv0W?6)cIc%`Nr1v^~z~=hpI0n+Y3rO#8b;!M*+fy8!#^B6W z`Sd+nkNb303-vuhNmw(tI{F4bdM%PF4qP(m+Z>KlSQBSw7H2$8KJFh96~~iC@xuog zj9Z?TLs}45$x0pA|q7BlX<4Cocv% zAs;`5d<(Mu+b5s#Pp6AJb;FevopW9&dK}5j#H(!6wZ;_qkKge zFM~ELqE%s9=6JLDGCp^}Ay!_o1-5%)?)rk2l(Qpiil-MtrultO~Ps_un;k(Nh7QnBwwO7&4iqY4{}Ns9~E--(!4 zl~l@bJwGLhmFg3GG@7ecQ+?voh^an(O}JYgX>f4;xH-G_O={is~~myptzVoxHhMoBk_Wr0;CO7>H%Xq0e_ zkouE|(GIf~jer`LwKSYY3(=){<9ajuSl&Cg{j1nM0c3}V(iv6S(~}3r(`E|wX>y2*qV^75SnClTz>O8B`|+ z--mR6U}{puV(iD0bfUn1@na57qS8QImesaZDzVo*VTG&|^&w9UJIDNHq>xpoQiiP7 zrlp3~>RM$Tttu@jGn$oFLsju<#86dVGdEQIDpFV{sp8pEl#8m#NZ^Ebf?%UmNHurZ zLU-zvV1;*r(ssf-u{>IMCs$;z|9pm0@>)y7kgp)6@+CPk25L_@QR?J^WTS9GBxJ!a z2j=oWffNJ@GDQ*zIS3LTKn#N9nrYbRfOk!fY#7hDU+j-Z{hhPJD9;{OlYJ=bdTAka z^s*t3K<*Q1!h6Vzdv|4~2`%4a7Np0Kc>C9`1|;6XYSGa`1)DmMN4jK#hE9&xhS6sQ z^3&_yG8mMnV-qj4TZA0U;KEr7`rRDu3`S>%lbmYR?CP~-!GUoD+g+aDk=c8`738fg zoXYm@pq_WWzB?$}IQ0L2oXu-Gv9?HNqM^mK^OFy|S?G{gcF0}WX5qP`t*w=d{jCckzOo459G!U6 z5a>EWKHAN}YnUSz%82^edRv~TZ1MCiN3hjw#9^1-f_!6M8_(J+cXYavf~X0$czEz&CsK=d8`IsrlXXe4MY;8|l{AC?0?v%r~;0mdP~~!iIsD9p%Dz$^7)S;`|lqP_cjNkZ3wh)WUXl0 zbBm38UVB1E!Bp*?F-s~+o@=<`Fd#!nAW1I-ME8+6eVS(StYPQ5L zNEUNlOSp*eLWcB?r8Fepf-Bm#-Du z!~RZ-$h7~;dLPHRrVy#1(t@xQ0|8yTi@dP+d?e_aLXZu%Mjq~xJ21~jpwUjH2DFrF z0}?Ic>6rO|xDa=j=fiz_fvO;{LOul@eTIF&-dQyP`9wuFHALX$KMQCziU8MB_E*7r zIihGj8_Y(U{iwi5%t|g8P-gStM6+1Bpk`aT)?#Tp9MI5ZHbl+k;>--1=6hXS->sp_ z5Yqle>B2Wo7zwLcQY$3ce9qbKcslCcm(J(S=&00mvP}fpeD>SacwMfo8vm|tPLoD6 z18U@08y+R6L6Vu+3hy_kcqj@aCW1Bcr;8X^9;uHSG;9_rhI(~j=C4Ez#wbDzmBtDk zh8k;AGeD28>a623Ms+e{P60XjUzEbQl%xhC>1h2gnS ztzk5l>r>t4gsZ35jT_PW_sC|4En6C2ZaceiA zi1{i1Xr@M9bTHoCgZWA;4AoE-I0O#o2AP(J*WIT!pZfIjNLWstuX?;cnM#g%5gg>x z1-VszH11pa4md=^u!qCZ>&9#Wx(E5Dxj7GkZa=gYXI%G#*JLtRQYIme>Aj5pB%$L5 z7b|+yj2dH6+6G=kI+ij^LRBW_wbX<`Y`)prenAtup2Zw#uF9L^l#2uByrw~@Ac1+= z`TG{Aqx_|Sh4Fc?FiG~?d^LT#JPRF?JcG$N`BK}sDdh8xVuG?HPpVS@uLM*@yDx-H zGz?NC?sCj?8k4Bb?QPk>)*xyc36bqw3QTUkHpY?2TGTb`1!u9JY zf;!Y|>9b_-rJ!6P6#D-yY|iWwd3!UaqL zI{^P3ga35Mfj!l@NFI#|uuRm|e3PB;;)UEI4c-CQ(#x6%bAp7Q+~{xrc=FiQ8(_xi zAvztnfFBOPf5+%0H9Ysj$e2RmZL+sJ8Ezh}&hNIJi^K8s&f$3b!n9A@o%L>L{8RbV z5P4UYUfF~7xMX%6PlgxChhj{bc?;shz5QqS_PkS%-N(Z%89H}4Q;KERA=C4%51KuP zIz#zM-?|bQFczt=wZ-RwS$|3|TGy`ip%+N?e|r9r2m|FgfpR5IQjRR&*pT%nkjJg| zn5+vnP+m~0(I5QE=7Y@Y)qlAJe^DlagSijC39zl0xC4fMhCjYM4O>bLunyy~G)ho7 zg9NuIgQSlj@c#LhICHNvh?y;p-m_bKYUv-U{|mSF?H*9S$Fl4v{TPt1 zkxt%PM4*EW!Zn$WM}y&DbFu~9Bx#Siv-fjS%?L&P`wlBwkoUyYKEh~|jC>OJ>Q7jE zoaod?{!ffr(b(i96QYgq%0eKSM>rGh=S`p7*Dp;S%?8N_7RhZn@~0he@4!ZN-W&9| z%Y3bi0Lz4NE&?nQwjf-X@lTfu*KIq4A((yH>VX1Pubr8B>K9oaYB8Svj+a7c28#?5$p-8_tJ z4xkNkc*V(gxSYB+48o(W5GqVUh}t`jxgZ}~B)^De5`claQNZwah$_bqI-+@7xa_pC z=#k|JmF1ClSGOoOLt-==M(ffrBn0NzWEqF3S`wbiDltV}vO8f~8%+p5J4f$CQ3c_$ zl&4|Fv*gSvBCbD#K+C2<fJNjX6t9KHAB=?Z}D(%C$1j z1{x&^N(sS<1eXK}i6n37kmuy6UM))C?8MBiyx^BTWX5onVvjQ@y-)_y@?4}GXgPr@ z;U?s+&UcEu0D+0~4HiA^#BrBw--||Kc_+qltZkOmt!UCCbi&GEE8{^7O+u zo|9j#F{h1WDnfqsNJM@WJZX5&o1L9Vr*hb=_y;w=ien=p9;lgVYp|C_cOqRX&)lWs zT1p*mBpEtR4W|y7@@JZ8qAWCpvPfQ)rHhh2d7Ha|=SKFeEkSyx1AVxFhlA;*y&=1E zTY**Nk7AlrHtYhU#pu**uql*-iNH6QNuKF^k=DQMKcnkziYXHYwKAs}Y>~FP^kGFb zW^k}nuqhRqNIn)_!D#F}HM(ysM4eGMl3n8f^=x(t7Tb*oVqM`Gc}EOF#!P4<#*e%k zHMmD1QyH=+ZIR@b28Rmg9qlHQO9{vVk58Tw5>##Mo||s>AI#x^)V}bo3SN=-M;E(H zCrI~8}pw!yT8 zU*wT!|HC;@$F*7|P>UDrb(qewy$>w~kQ2bQMRIk9s)mh2i+xk^9xA(p?#^65y^v$_ zPX+mTYA=HTe@swn^HljLa8^~QkSlF9f6EvPDR&(lEzt5xB=kbtFQ`_tYZ9*${uqyU zdu!K2yt}YUcqQ{kfxQs`ewT9^KwV{Xa~6RkjlDeiSrnQ}DqLW11&;w!E_3!MddFgk zZ4NPTYAIg&I|VtGr@1^T$sry%8K<+ssw%Xb{7pe_kJ~A0{>*Xbbu?{5-lF)0Jk;$Q z3-YYE4T>ofnzzJG3BWtANZHWolBs||KfdU|&(MDIG^0WPBfry$6$wF^F%q;T2C|N& z%w!(dfV4!w?{>&)s!(x1l5JqhNX2!Z5@!fnb?fEPZbT5+2naKnU1o`yX)94J>GnT)%pq#62?{sPn4#>Q1b?73A@&gi^T5VmFpu znZ44vU2{nk-r%@j(yG1ToRkf*yp>LPZsmUyTlWbh>2FW36uh zJ8IOZxiXRv-Uu3uwh$@25g%a(5Nn3?-`*5$)ziY8ZxpfNjrczNP$ex=<1?PD@J$|J z)&{<|a!IO>+sM|hGnkQu7a5T;h1j7;1f(tiE5(Lx;^RgMblvOaGlKAYFo_S{1cAAs zn^*|g92>gH-(}6hh2u#L-M$rlbrclPA8Opha!6gGa@et2KN|(5Xu|8F<5RIoI3l{q zoReaFDkEc4L4Z*kAU1Uco03PiEIahlj|{g3LEOKmN26jX#@twekiT8E?6y6*mQ2ue zgS<965I01fGtaOh9f8;acLYI@AsJX6rZ@=xqKbhc=wI{~RiPQvW0WD01=l7+Ku3^S z67-T;u!J)tkq7+O=@Y(!{Bg`%edePD9aVhU!2Xz^RlYV+`oa~k%3i@*{eTS+S!w{E-$=P+{RI-?00^DvgBrt7Ll00)dtFM7gey2c2`|6XnAZ zC;fZTw?xU*dJEzo)U!q~2eBm}K?g@=Hh7zR-fxI@dEC=A?~LwhF_klkQnTx9hd}*5 zlGhqd^9c}+&oJt$Z(YeWKS1#heIDHR zF=Z1p69~DR(y)yu88V<^WyoRr*)29jLtwYt5eu7&c7Uj=xUL1lPvjM&>(&m4Rw*5* ztOedx3W|Jmhe8ol0VC^PI9LJt=Q`wt(qy;}!^tyaw$)bsLOY}qk+EEi(yC%GZO3h9 z0}q)4fDD?_+A;B+7tWC0nK-9i%rMTPaoa;;s;7BZ;NiF`N6y>OD}C{BAOW(bVvvu! z+EU}-qKV(BfM>)wQ85h~#gb62N*(ZT0`kOebY0kKiWodJ4HdR5@TLv3IEz{GvnkK- z(I`lLY2o*K6sH>Ml{*Ap1!o-MEOq#P`HLNrq0kC_QJy)&hAW?}P%5bG<@%ZIoii+j zva@v_kY{X8#F@eXq{Z@H4dgs4=(BE|{X^los_Ek_2kb0clyYatBDmvKk?x?g3k)!2{)5AlBIHNaF5#L@$IhB$&k9L>4ohPUe zDvYa)u5h?nVQ(_*zd|H)ln{LSnP$eTk~D^pMaWkx;|saG!%gU&X{ z^Llsq2c3D(U!|LVry9)b`9uqU=<+SnWrgflc8#UYAY{!L=}P2 zWi%=+B#*=gs{j>|rFqk~KACkZ7Q0&K!D%Bo@XXTGRm`AD=lk2+eRycyA4nDTs}6MK z%ed7G`bBVwu^2F;%*dDtqKvXBnd36Om`Y#B?j|oP>E&?mwmCL5iVbq@s`OyjH6*($-R8pt#@$F(4jy}1l~tx z23o3)fE0)ZmKZFoB0)?qVvh1$7t|&4>sb*q6Y$ClvmZ}F4Aa{$el#LV41PmiqAHES zU`lP-!p%o!2qzv)X!Sp=5big#LO7i^8Nzwqvn^X+!jZhtP0(X>k1d35;)B_to4%6F z{HJ9WKamysNLqMaE#snPQstZwNDyZ{W0z5KM?S*cle`c}(1#fUiA6I*Ai0jKXy}WK z$j7r1w~`1MJ75b~M*3EeXB4pD2*UPM3ox1)!Gj>qkVwi19>lWQ!Gm0>e$J%@&)&yf z1ZFe2rvn(=JkPXV?2kwNowGv&UdDmi_wYIo#&3I5VT-sojRUOl`=q?^u=)4e;B z(oL4HFg@>aw9x*wt63+qqN8ogb}Zz(9nYkq;L`xP=yh)y49Z)>z&x|tX|`F5uP@B5 zS<49=1ULTO<+hK)E{FyBS2S*K?+$wVp1ZZXJ6_)%ld>|GCrThspK+jU?g$v^L6 zc~y4Sy{k^sBlirfNs%8+1A|K*JLt>k9y)5?mi(KqZ)q^|pF7&xT4CUNF2xqrX?^ca zoz|tByt*4-Z!-IQy6N5`{jw|wb}H&YLHu|mpv&%LYbA3Srtg0(?cG3*1Hp9s1n?}rHm2(wz9qxY!6=wf}N zIx)J4tNlnGsFsn|-L!-0;^fIkI#@c;NTX&HaYS;eUbGc`3a<>!v`YAcgwnfNV$Y1W zhCP_#xHZHIZ#cuE^R#9Hxk}M7%4Y!7Fdgk}4JYd4WoH-3s^tWEJzB2t`0ma`++82f zvto%)So0}7wVw?09NoRF|j?Y0-=X&xHtqu3LsLhF% zdo$~qR3N2T8QZwhhn(1Z3*sJwmIh^X)DL+8Sj%QBv4dM7!l0HyG#DDAispn z>dD#uR)1VawB@OLQkT~&AAD$<$&f7^Q?b-S4T*Pv9oO=TG z;zt@nw1dqoK>bs&{k^5{C3?9US5VoQ?(Ute6W-_&7r`@#cCc*+1QHxlhkYC%yCwFT zaKzlWg!s;{;|m#7d#e{6t(-(T7snLALN$ZB+!$>R&?9W0-?+PAix7o_V4odJNlC8i z^?Sc9{2IPqGf+aCf_e_w(-|YEVC0AcdwPg`A)zq_zX(|TaxJAQ4i#OdDAK33sLqF1 zvIsNzdV%h?MGJsvkbmeQg{K$DTMBYARkF8-{T&02#(r&yyI_4g;cW@$kV2Y*p}9+7 zXznDZdbej`Xp}GH;b;E3L)`Pr9akAiguu?6hzs@T3I^Ov^ivEoEeSu)^R>Q)ydB)9+=tu!;diE1t43DT`V3O0lMp?l!9O=oBS+XkJOPi-@H}N{= z1@dpv$myo*YT{km+*`7w{vhE6hk3I#&LtTHDg~mp;_=Na%!E~_qxpmpaoCpEET*Ft zB)haK`z)>nAuz8nEeHV%#j_yL!^8u_fKe5lk5~j@SpFfS%qXRyfnT{I&Cd5TAn1*~ zO592^#6+&CX;V4?LoO%|q}7S}$Y6S1W@hN=Qmy7M> zgD5iI$n8eQCh~PMEnMjHHA`%b(Xl2GNu+n71R@F|1woWJ8U;~l@{lOMfs$!ML!rPN z8U!vk5GV+VBZfXfuu2v3#9NKvQA9!B+-y#N^G$V2s*Hz|tmIhbNok3&r29`#WEzwjI>pf8?l%L5ckwC!?a~=0@_?eSl}>_2C=>6fWIIO;K#5e&Fi1Ub(UEiobf&L9AAvEY+vnKzNyos8dQkqi+`azp9_ps)fEjebDg|0KJC;T#M zns7Yxb#*?OIHQC?We1ycxC-)W-yrN6Vw$V~L&m4tnKv`qnGVO>!@;OO9j*+RjL^{N zXtoSAc;NJC2Wfxrr6vy+ge;M7`liD^b;Y(Baht2cm{74`+-UN&jN;@GZ)|^1?(uI6 z&$QWqJ#ycbyGzr=bK~c0%OQ5*-!OSa@;*(II;_3tK%TCu-ofu%TY@e*9eDYpmh1_% zJ9barGj5nFh)RdN!MIYx4xH72nTIR{`rP_(r50JQ=`zNawp)#RaAgn+Jfrr4*2wLU z2g>7~itM8T0>8>YBr@cvurq*+3^o$<1!Kfipo4S*R^C3vttW38V3!v_pCF}~!Qjh% z15bTY(765Naf#l666Hy?G19u~Hz7;BNT9-*_f~OxHO$$jBy4ns&luuQ#l>~vBT_g8> zDYO+&@E$4^{5hAeP#37ODGpm9$9RU9#bGN%VK{7MxkQJpuE1Uxh-^TZ!~SrR6H~t` zlY3ag;(Ie>5ITh*qp)0dN2U;nLiCk*?go=k$C+orAxu&YOnKDjipf^H%6G>N0XZ zL1jT|K0#eWDp6X+G(DeCQvtVuDxZL&kCytWS({>%Cq19Qxy^%=gjh$BF<@6pc%rOe zl+v(kku8WJ>G_13I~d`Kvd6^mL|uV}Cziy46dcX(&__jymprKqN%@4skWWC-pMVQA z57{y@3Q{HLZu1Wdlme2834xe$;6}B8K#XrUfHST_Y(_>wR}tWt7@L$QY{`L`!Zj8W zfDiFTACix{$~>H$kCp%ko=-?GaXgZ&xRMI|fCQ$m-bvSY^k_FYHhMaYDx*Z(!jVL; zB)6ZM&y*|V<`Ezi@s;yj>;fZF1tN(T!&x;A4@K_WeW_FruO^W<`fLBLeS+FoHaC-5 z-j>fUB8;*rnPBs~P74b&uM-U-lfgy^lEWYB)3Q}c@nI)a9s|J z4TK&@8YB1XR{|IWpR-ry`zkfaS%$GJH6FJcX~eMOwaJw}RVmbjT;0>iNsK)}kS7+VX8!vv)l`-FafvU|O$H|ImoRsl!$ z-kbne&f&LcEg9_H7^hPFxW0Fr!xm>iVS70{gU|K4@JAU8La5^XLBwkS{Em6H0*&XL%&A`EOU-2$B5g- zP7~vH`TMS_rk@qBN-9f$(fvdH2_aUH_&O}=yw9ye{V^D~oHG*qsyp@gUbRQ?-a9lP zY^|`@))JadAJgljr0779Tj)(HE{7-Ncw@LVJUi_zRsTCB$0nA#EP5!;I7SaYmKP1( zzI(pcg+I!C#>Y_oK7X5hm(;y%INq(z$AqC9`jgEmJg(T~ZSp;rNHv6Z4|xa}9}-BY zW=Rw~XLTyertlvwz|a)k*Uu@8q}r7DI_-xowlZ@N-tIer%T{JZWH1^=BfwRCwnL6* zn-z$sF%3&jpi9*iM*{rJ6G~0ezzf2n(ksak%qzV4@wZv3#4h|BCjXs(^*UPZGxr{H zq{I;0n7yX~H^dp|C5pu|R8)?pB#Olf#yc1!ic?A@f`5s?4MCL1KFOMNj)WwqhXp}A zJ#6EOFuGdWnEXqJG#Zl}uDj$UY>1t9m7iOOd;?N>BrE!>PfDEp@`h-Ga~nC?drn9j z9D^a!cS=_eYkzZ)d?4vMQ|GPf0=ImPtv}6fo~7J({A=rj~Pj8Td=1ys7QT=QAY~#lIQwVOSqurgAYMk84?fpx`@u}VM_KIw- z#1OA#lUedK;~r2i(FfFqN8}6SVS@oUQX7EA_iswUS~YnjcQKEaR2^@eutdJ>9}E$6 zZqzDBL|YX(G`M)A{JC_=fG1HW$V0K5ZV>~;1Ps~u=Kpxo?SH_L*&7m3X}eZc0csw8a9Jz zDAS5%5Nw(07~-2iN#21J@G4vc<>eReXT>HE)PX!W)~>1iet#^uNhkTCpO@@#fW0*1 ze0dHdFjmgfsKF2#yWV;T4*Ho*zUZGE8*6lgp_gS3Y+i#Z2TYEQC7e5Ct@`2Tp1c9Ug#HTc_bti`);IE zA5}D}usSBqo=A%e*x!ki@YAW3;d*{b5-ZgwC^VX@R#ScA(}<})eNA~KPYcoi5-Iy2 zs9NU%dLuD&v|NpK`se%G+kKb^+#g7l_a`2{gp|G%dlF$aO1eQT3#=+pvY%o_ql9CG z)SpC*c9>0T1k|{!rQtMMh%Sw=ltSGNdxPn(V*3P;9Ue+&RB2C79vDxXDb%BKpMGt_ zyvA0XmSHgN5i*|PU1J%sMvYUm!|4@=Pa$4oRhUh{+%vQpHCJT>eb^P7=ER-H;2 zvZlYj#R^pgWk$2oYN#qcjToxxYf2?f3st|06xK21@4Dk&L55=b&?%q_lblB|Qyh`&y%QS#ic_cp zU%5=>PDQy6f+FybFd_^vBQTEQKY%01z@Vt$cwxMO+TXL*`>xx5*L^+jPQJgdGpQ%h zv)*St_w_uV^{h`IB|(BrQA9#cg2V?9lOVY!yH;dJjGNobB;a0?qZ-C5?&tgCQGe_7 zFev(A)boBBe&p;j$Wy1T(eZ6}!%Y>HiF%Qz?NW49(R{|1mVWsHkAUq4a zojl$pYczFo+%}As73fc|d-GsW-WgkXncX7PU#uf9!_$)RkN$tngs&m zIcs-mej>9!YEzK6wjh3O zx^X$Or63mqx5gY6bxFxuEy zuHK!i(f#w=Yt#CRu>9(y-5k7zIb)%WsIRTJ6^Y6gr6)q^9KCn*#=I;2`oh0J5TDIdZ@sm8rA@OO=uKd}pgsNJFpRkL&a|!g3ys|@Hf~{dE z`j3n@&uz+CPX8CU#}9V#lj)idy({Y<%64tJY)_DvKr>xKNy)prB%>^7Hq$d3+uP&8 z=zM)PJvKVumWM%)m9>6dZc_U577nP5!a7RpP!t;L=**enc-U?W z;3Eb3<#s8SjeaZ)?PFcC)f&ylewqye(6X@0)k@8q-^KNary>q;7DYmG$5xj-`4Z(aYb_!__hs3sO3uSw&gL=eV>G@%s zstP!U>Mm)^5-aB%LbW3OQ^=mP!{N^QV6+L@^K}b63>AJc5CHwP7Zw}my!Moif~ne* zF)J$ixhWa1A?wKx+VmI6L`EbhA2a21W-q873{ zneCZtjT^9N@shW&0gqa@XW(KWOvYuSP+>|Ac=!CSg1ji-DmI7xtrn4K|Cx0^j&n^R zQbDB+1S;(wa(C~=NT@XBTX~4IN4w;9O!){j+R4;}mTt{}M9a83W)=`v;_lLXL}c%Y z735XOQqa+7*aG&xRTEH8R1{c41YUYsz^G9KxSq1V3f9XJMf2I91wf|O8Qh~HpDZv6 zvyux2l-YbZ(ZrJINHW`iZqg11G<4ae;y@sp=6hY-*sY<;5X$~W?ZS6X7zL|YQkzJ! z`RIZ@2z3b5S2HFmHG^yuK{lT~HZ49--`~~E8PaHGK#d%0BckLCNHP;!;r-^F7{#H* z{$vpY%OmYkgNDr_#ZZqf%>9+9$rwe5q0(5P!%%Z=>IUYNVn^Z6$p4@;#-%1T5Mh_O zyL&szwU@^JB@zq)swG#~>=XmEB$pO~o44n4MPs-shJag;n=~N5TJRjGagHOt9Bi7i z#K;$r4~LBoKt={8-{qi9vU(0M2Pd^>i(CZJIImA0h3XBXv0R_(Hea~9p&s0bHor#>J8Ti_ z!_B1)`Ac6j6lL{$$xC{7WtP=-*D2XVzZdpk)-LiX|I9cJI$!mc{$wgS<_E=mx*#u#0{=iRMd&DJNhMnCBtBtN#Yzqfbyll*^*nZI&t>e7eUd~a*}6;0TB7IUP# zDxVW)TnWHAcrQDD-vV`%zZ9@Az77_q$pL;RXa+WcmvQp7wsBI(=N-iaWoe!?CvF(C zhX8-j33$;k$dGu*F~8H;Tmqi%9xm6AeckF z){?f@KHIA!s<0+7q8?NZSX#DEu*Z7unHt>@ovv*B;!vfPhGhN-Z(u- z9|bPphkfwpD1k$2cyma zt^^(!i?r9;p!GT-VVz%pT+ivY`n zEeHo@eCbjJx^-(X1P3oiJ#+!6diCU?@o4MpI*cBgtZ()wXW_A`1=>M;VEEu@i~mcp z^i;BcxUn(XnW+D7@Q+)=m(Bd{Rb=;YYrubV!8wlxRg>2bA-e+@gB;#*@(GtyCmlkR z(L2~!kSAIp6wf7mY?1sTno9r%>P`V8+98@87j(q%ws75PV^NkpMmPtfD=>PDJm&!K z*kn1vs9q9Y$SN^aU7r3DT1&Jb{Jb2y6HOI_!&07w8D+_tQ$!qeTzsw{*z!b4;&RB?xxn&8@uT zmtADW2$f=wGtGoh2FCTdS!mR10#(6H^trRP#0yl{htyj*6$}FtURvQ+MgBboNaENx z=tP(Hs8MzvRB7Tn-#oV=r60_Jtomw=Ic*eE5$daltNN<_EW5}z+oh2-yynehC(5ZD zHY@%?t*_$vh=>bnX4)F;rRT+F-PGwMwURoVNFsf`qGOeXw^}57RhBJM$6i6Y-M!GdNf(*p>2~7vy8n9gN1#Q?vWVLNpkKBiS_$ zKt!pm-KfD4xocQf_Kp~YjG53zjvsk8YH*K2rZRL-`l6Y9EJ3MYnoKSwpbI=dd3H!r zwYGg`y4imuhXd03gn5K34talcv&$qwx@QbQGWJEF56wDiG^T90)a#>jpa!Q5r)Au-Y5JKp4IkNuZFC4VTJH+<&!hbLYwgb{4PftKwaf?a~1)S#vY#hERv!Mtcet^ z*^=-aS01Cp>{u+d&0#o5O|mCklfP4tqj{FgvyvR*fs?J<>|Yn;wz!=#TxGEv%dgB{BXj2-MTH<{%m6rSWad;w%=g?ecB?3&ZhAxbw3tuh~-#JQQ0; z;@YZ@HsP_Tz`7}tko3?>5K|p3=?N?mxUPn%sABS~(MgX^mrQy*U5+Jde}u%RH-ZME zFGNaj#7CF_V%?Db+qr#?ZxpfVjrczNP!%mw^D|zo@Vz|3ybXM9`9eaBrZ+Moa|*FT zQ3yy~09J}k-NeU@8W_QVps1aYx(NbvQ#Y{?usJq$lfTO*iklO%aAuKjMIRjn1@wm+ zH?b7bm#6}EtkutELE(E7A&0y!IzJVg1QF3q=17Y1sf^4`1p!8D$ZkBqEd9uITM)$K zdwMo1mSW6}RS5ap)yQrk=D-HSYoil!L)JNS4J$Gbh%N9y5Cj<#Xcq{5khIXtY3ygGg8Wg;U40g#1szp>*}(popjEyxQToCW z7gygCeYlz-U;I^tRmHjscAc|gNA36{J)oe%!c%GYu9kobp`ub22&^m!UELoKl5^x0 z-**{)@yjM;28i)7ciqU&}~i*N9?gJHhsupzN3l!6pdoBI*7A zT5+_)J(i%-`~Njn4N*mwcZ6)XeTaWZ?fwX+3miUZfUN^KfWS~;_C?IX72H^w@2up-0Bql9e8X^^&7)0P9;^?5ykOpt>T`^? z>RVSb&7V*HfiHslKBjDfRsx}RQyRAMOhW}!tPC|wKfA@IXbS8OJ7Qr|(+&_-71y;y z_!%5qc-87YF%G43aaZVBrJ%@1cPbP?6)>`%g@cu#f38FBriU2o;K?xRjMY|K_X|x( zB_d<H1tKGcM8#{+D4v9JRRj@#ydh90cBAXU zq$zUn&@xomvcMZuw`l2<`emIL_h=TR-n8(?JqlyBk&Q)etINW67$x$E}DmchzpJD}kBYl_e;E zG9$}Gs>@5oLuYH`MZMeo51o1CuhLDwLmxQn1lTVO1J3lwtQV{mk3##|j$70;tEG_b zOleBmWU889+R7^)QA1#K8J$WC#Ut^-DnUhLX~qkWev)@)=L^hs8fhg4 z^Jr3u<_smX^yucxG|A_1JThnkPbi}=^4vLjcsy;kEpC*1^5#1Thp)nLok=#kIXn|3mZ1%f*35UGC|BJBC=@RQe2!Qzn+ybGX<}rF#GX1 zw1Rqj#E(Wqi6L&tLsVrk7)+_9CbYqOG$nZ{oVYNdrEm@uq9<$9YLVnOvQjvmHkrbu z-*n4M-2^?x@Yq7?CO(**y6G$B-iG@`R_Y^Z;eEf1gOn^n|3VY&Kkbgnr_U87WxBCSf+uP$4+k?{maz~5^QF(p| zSuNyEU_SB*dR~kHnpyOya(LXP9fEFM$;{8Z=|cy}_kC}q1uF);h4~|GCr8`J)!+qEuJs~`WfE>=`!vhH1RCl$Hp!I~8P!E`XV=d^iU?J$>mEE% zHlhz8t_R$8Ey%kUJx~7y@AcS)t^Vfd^xAX;PX+h(Yk&1T6&Cyr=C-gBA;_+<=t`^` z*|mCp>LQxh4xdF}UjFSk>D38%e{&PYEnm3o-PEI9tdz3ebMMOf2hY1+!PcOg7`A{~ zcT^6A6Y5WXn9zVQo7EY5jv9>4*C(oDqw~1dj}(Dw9l1QX|B()s4K&KA8BH8foT?XX z#h$zkA+Dn{^ES!vC$!$p8hdKAG3+hU|6-*#oMF**S~G#%q$Df64Qu-1@+s(@PDeW% z!wDV5L0*qmE4*cUYa-6BPv_OLeXC@50dG+)E-V~9rWVW!>p=^uyGZ6-zkEb-g8dEZ zbE5U$%zCC3NF`Rrwg?%L_ZGzc43Z;+tv!6MKOPQ_jn14Ij)(1*dEZ}<7pJBzE;?n4 z)LxK|yD;*19Wp?s)Y$j1HBVa+r{{T}L?y{>s{TVkehHb?lhgf;{x8MAQ(uuEaG|)~4G#$LoSOy2f#i1OS^M+rjo76j@bY>Bkla=x&L2RYfhEek`Vd?61% z^H&|>e!tvtl#xUT?98#aRFAG;z|BNo#V|8|sZEaCOT6N$ZSwIJE0b=hZeEeMHu zg=s-ZSSX$afu0J>f{Z|qg# zR+1qma!t*e(g7H9K_QU&sG3W|;F?8pcUNwRH`NuK3Yt|TL=TQnsjD^Z`g<3*KZv5^ zjoNN>ZX$0d)5eAVzGjJSG2Utzw}M$z%DYe!5e1Qgph_H#f@m~Nc!^|)w-ZSI4@#vC z4TS>l&>(QRfj~hB?F@Z_;FY40CvP?kr@yHlNtHQuL5_ApuW)lk$CIYi>U%}O^hM;SoD%sAFTGc_` z0w|Hn83t*|_VL2+=?ZV*3;9yvdEQHIUDejgd99B66PVv{98!bcUga;zo2a=7Xwywu|U4*dmZ-3vTZXWR#yj$P_or}*+|qWcaX(xc!~)M#dqHc+@#NiNU%<-Sr#SWGEd%WG0_YR8RQeH1(&Ik&yu$b3sZR?Ur=L77(KAq@KB+#& zc0EyYOG_9DbU5?A3ex3~BsdZ<$GH4R>V8%Gk^AzUhz=0w{^aeIrS26PJ#Qh#G%tKV z!6?)-l9wTF1NlvV_s313;mnlBwd&-A5{RYRXyy=*dZnc#%z0OCUn{Q8J0+Q=<8nz! z6uR0@Nt91ju$2Ubnq5N%=CqUq#)>CI50wgjoy*s#3pCl3fUS^Yl%X{=xr#QzP)SlBEEK>i(Myp zVNBGbV} z3X;Pg8q=~(%9dwHWzotfw4IUjO)vnSd<6I9z}Z0Rfn+f;$p<_HpR-pM`zkZYS%$GJ zrCbPnb2w>8JA%rBeuk0PCU^Q&qX4}S(61bRRVhb~mGXqc>n3eR@`fZM0T*p(U^vj> zQAmcg%8X8vOg}uy&XP4vN=!He9Y&{)m~e;>RjrEzHPTmrgyV1{WjsiJY-N9M@9ro0 z{}eNS<%-p%Sc-%gA!q;}kX-o?{zIl4!9E)ZSiazQ9eke!EgU8oz1Szr`;*h_O*1$z zK2Mv$X$yM<;L17t7NaHOIF%B|^}X9RBsmPQy_}uF=XzcEPZ^^4G(A&(RR^;m0oJ+~pg`|y+9W6xxEOU-4$H?2o zP80KX`TMMFn4FxvDyc0czJ7c>1cdU~?QM04rG2D3Ou)(!OHE7f%KC?R-;_|$Y;wZy zn1nf#xBDNtbos;Wt>I*SV{~>He1$M7Nqij^b>8dNq45};TaJtbzv@XnzFX}Py!SpD z5VqFLZR_SR9j~F&5!pA>lT=&{Pw4U5aASBHE{9bAJEg`ZmbxT%C}b?G%x{q7V|m%o zZQBp_y6~SepYbtNzt2A=-=%dgAC9+c_c3AUhW=!I3YROk`Ivl9wOCD|yN*GBosj24 z0txkclF(QKL(zTnoWe+|U5Rhfe%NBmGlK9=-w9k6!L59@LvG1-E09lPI+mP7mzpb% z2E>?b8hAl?RC**ijCq7NU;Z|0Rm?vBEC1+qq}pfhIplDOp|&x5PrC>rWSmzh7Ryjo zIg(N+7AqLfU{ENwKr+HkXNkZKL6oRIWtE4;a>?akxk3#Lv~k7UrzZd0A&tQ#KT}W1 zNl>eKe#rD2ga6StNX;>9>dgURx#R;$$CRrwI!v*;_|9pCbIy84m zQ;X_f%jX$4)}KOhI~{FrEyHA^9T;wOTmQmvd=l=$FU#Rd4DnhvnWa>Uc$zBJ=(A}^GuY+H$cIK4*K^)o%xCcTAFo>7 zx4YY0>~(e*E1iqGLQj@O$TNox7MSc_rw`cdzd0T?jq6(-@_^^`6i25qIkR;>eV5kb zK3(-feb~o@HFM!uZ}6knBB|oQC4;@qL7c7>Vpm|g{c-;zQE@zJG(UWR!MNpVJER45 zoqLG zNWk;Mv6M#T5Ge6}mQCIVrHu-mYo2CRV`H6;aP+e6fz4|$<$#xCV+rR0*{hjSi8X^Kf7`5+-gtdoCq-j{EFA(1d=JXT zvOvgAZt+3U6L~4QxA&5eo=EUl)xFR!R?jYZ`A^=BRO+LKMh!L;lNOh-zZ0>pDyfv| zdVWa~tJEj>XmnStmiok}5lemgn(|1VmZJX!QuRSlb%H18vxS)(uP;|(gZ_j4&CNc% z2i%XUwSNhzd@1%M(rT1)gIE?=Riacs#fnA^$4IF^iJ0v$Z_!Aoaal{#X|xnwx;Ghk zwAkuTroW0E6F_%(DxJ}#y>s&Lc-knT-gv#-r$5|b9%Cy`%QP5Qgp6l+$5=+J(c;t; z*jXWWT@@vtLOjN*HV1kmZiM18xXOIdl1XWI5apbFv|xkYhje~mW>V#1?8oEuMS(ry z#~hqQWr4UXt9`39;%7FlO716nSN2eB#X;ICMoU>K>O)EmyT<%xq?A>sQl_lww{Njh zRY952t+bk|icce^s`{F_sp?me(mF{Mua=@5R839-C%qE{8@0ma84rf#egrGM6O^`- z-ihVW(mT1LYiO9)e|`n2lrPDVIZ%7biP9$*Bpa1e&4g^gUS|^Y38W-QkSU5t$Vrg+ z0Adm(*Q~M8<~+v+Bqss)njFbyTniO zAcw@KHM{a>w~|YtY67xh*RT$GWrw^3Tf1v ze<<6v<+43Nx`wh`0@Wt(?vjkMpxI2%Y;13j2cz@#+4R`xd|Ms{JyzEGbqOTlQVjCv zEgVoAg>{tHp(r%g%eQQ)Tu_C6*~Jf|_|q2Zgr`Ou!`>qOZ;7s$oSm&1>k{NsrvUyl z9qnukC%w+~UCCQNbmMLQxmVHWL+@azqsP=ks(pwdRQ}QBF&&?Sr{K@i(tm%0dbkYh zF?&&_N9bAP%-AWvWItSx`}5P9wTI93q40Zbbmq)(JT$ie9D}rH%luAf`H_PBa=R4E zM!%irI1KG$U9#00&BlJ34Fb@zu*=p(!`)UIiWT0EU_FGq=7~k!C#U-xeLl*syc>bl zdPl8xO18=cOt!hADz@xfCA$lF5@&H?Vd1FeM9JqnhgoO;*C4BYHvafiir`NO>%^dABy*-Z@^k6kF8W2;M^ehG0GxJB2ieLt^VCe?yL_+n~*(UC&xMS$xm z`>SBR98u)j>3kH?0xnZ)4PEkNfl-*1Tri-_=Cj)-eku@Q)(!_WblIiiKw!K0dtKbv zt)a>g%Kk>}!go#>1*=(7D|`#Ylfn% zelK}R@2x( z9%A#ot?gGdVe47Uk?yK|PMmRZ;G7qs!Fg5C_ptJD;90;!PGLOK^ zIQd%JI4R`wj$(qcG*6n7P>>mLgr7M9FB%3J5)V1%cN&{Zz|-Bs<@&kGJT6H4lU(rc z`}aB+kxEXG$`>S`F@PC< zT$(*kTi8dwU67N}j8V0}4(7JUTG3pW<@HwAXgm305`i(F@szM_HedD>Z=c4ms3h#n z`|;DxO@Wb@QKUW*dkxt;h0efaXL~YSr%$CW_|shdebooxmkQx}$Zw?>iiJ^NDEyh& zpZ08UGTb<`%3a=(OV;pot_1y(w-;nJ*kJT+88B!+$uDmF=xSMy>RJ5p zs}I8Mjf?PSAN)BAe>8A~qZ_hFo`^}XOw`qUlaue_mE58Y-UqH_lr<6N1PMQ@F}?Qb zEReemZfJ*0-$ei#|jypB}zyLIdQa6G+ZINm%r?bC5*y=xl(RNfGx z@5<84JFuQhvfM(uef^gU zniF5Tu3-+>lBccom^lcRQNB@kAAS;GTQPA59Q_P`e03VOl$u~2!egn*A%@+c?4lf! zUOZT-2;6_ZInLed3}SAJqxbCIo_hL=^`CHW|Biy(l1&^4uYQ0zuvoRYIi6-lt{mI#0=fDE(EMdNe)lS}d$={=zqx>`$?J!Z-2se24(~Ymgv+Uu z4k60u9Re)K6Ri*`TtbN1JL&t_BKbu$mjDdZodQO*Lo_)q=!oHM;kwhtqKD^U>JI&| zI3W6tO_n2!>LuZYtP)ez#R+fea3g&`NPcW(e{b(@o+Uh4%>0!=LHKz&b|;!D2#2LS z3p2_x{74hlcOj2=@>1tUog|n02no+j_ayif$ThDZAMIqGc4S8Z)dZ@7o9J_AZHX7C zt`DiVa4HxEM(i!vpvb?+07)GC2A$~A9yQ9&gDOpY=bPs?gj)!AzFGCv8gtqxrXti= z4_EcoT9REyhO$T_X?V?>$xf70Ic!$^gIZt3@evUh)XcOs*ek{C@{4tql~vJcfl5?b z>4_xL*DE?!S$L~OvR7r=MzXexEUt8523HebMgA~mIAy~xFj|a3 z%?6u75ljTW!Avp}Z|>7>{}J7HYp9tBBpGbsT>)^)KCFnw3=WnGjv>M<$;YBQ7>%8$ zX7`PSXfO&#vTGcGh*DX*QG>%{Okjflju?cDnb1a#A9*%vaF0T!GIUS+BB30qDLKc0?$vL9gD6id96{8?nRVS}pZL9fP#(2mS-v*7sNa%&O zUofra)Fj>~{1M)=?5$qS9xNoX4yiY*Z74&4-{nXHsH=Q#&LSYv*u#^bMMB|1OC^OX zeJSWGmpOYBvtzN;HisCHnq*IK-b=nykfV8)%d?Uk;(?Q`+w5N#F0G4Yft`X`*S z-|ob!grLlr3EBz+Sz;+Ona90C+TE$&>5!FFrQ&fU+rgBPhU=CRXGjqd`CYvU@`~s% zA_#0Ggc-~(v(!v-MQbXA^N4_|BvT+eS^3Hme#3$tS%fT%CJl$XpJ>z^d*tess~EnM z6>8~!T#1H6LXuZ^V#BQ8yZ$qCW0l&jzU6*{Lh)r+A_u+@CXpx$q@nVJV;A_hl!hIIn*vXQN^%#@twikiT7x>=q_HQ^dB_NM9SBh#RubnQK^)fk145 z2ZA8TkU+aY=&SFeH7=*g14n3~^e6#_L^fQT3;`QKW=YUZX2Fu+-Kw0kTv{W46mwUf z#b`lCm0vcnKPG6EZ%mZFaKy#c_e3AAX2=(RRbf@J?t)$Ctk_XI{zwlfsIc%<+IMfpp;a#^l=k}0z@LnaoW&$XRj0hqNtHM^4 zv|5FQGIgYx+Smj6ab^`3fv-|m8sm(dPn_S3d;`qq_>%%4yGfiHslKBjDf zRsx}RQyRAMOhW}!tPC~G7?L3cc849Yu&HSWh^mU~S|a=mjxD@ub)OiA(gCloU9z!* zd7zz;#YmkhU}QZD2P;AUT!-9E4>8uklVLnFR$Fb|FEk;Q?i@3WbSCcHt{A+w<94%w zv~U2B!BScilQY!~tC!QB8HOwxr#&R5dWLre3Ws%$M7pCN&>fFZ?;?-+^5H-NWKYGQ zANRCn#_DYiZ`^R|3yq8z5*4pOqj(a^)rc&DK%Ll)t_zc<$iYL)P+`jgZ&2N$Mc=)O zkQw)A7Np*^@W(w0Xbp|Z9eOJ`Lx{7~;m74Kc1VUnEA&N0<_sUMe3C^}4~)U7%(H8O zeb#E22FZo!oZ%@HYEJNkJVSRxMkgFVS}yOkK+dy*UhBr$KQx}Jo<7cUz|OKoEqBHu z!5vrnbCA3@%h@Zc(@1S_SV#dDh*qmtsbLu8RqGYbaeYFwpGj{&*}Jlb*0l}-npc_2 z4bSW1kO4ZukBg>GufcDqZ>8mUou*;NWCfqi%77I`Vc?F+$`CnP-j|*pD#RJ1NieF$ z>$2$ZXm`8YAb*LwRR|T%Rc2Q>+^n!S8U9}(k{>;jbF9yIn(wL|sVjk*+m$6KfHEV? zB*LziUMikKTO%*(-R^%1&8s2m&S(^%WyB602At`USudE0lqai!H6qvRS~ROA0NKuz zrld`#s*a-7Qc&58fZj%8ZJcAj+tl&Xh4M>m={Y&KH>NG}1~8 z=Fy}Q%^8L#k$Un;K8NFxK@)gF8GVuG&dI~$X|ru{B*r@O4_}q@mc>5|(J-#$9=`0} z8&X17MKsVx;C^IgqNUmhXn`1DiNV4u6U2-nmORkKN%HGi88cJxDhjh7k3%b{w@3VF zM3flfhD^aLi@{(@?b*WJM`j8qE=*`CoW0K$S8x1{tQ1bCO{Q?^H{J45H$jgvJhqU! zi4SI{Zu&~OmBmkFr9P4tURTRFXqikoCj}D38D;Ep?*{|jdtM48=)+8b#G;ugkX*;r zH1w6%$j7q^w~`2%J76nVM)_8dXH>A@qLS^t5@0klk_SPYp^%i3JcwnplLxs{+I22? zo5}q;fFaEDO6&Rlc+}rIJv88D?C~;gaNPG2;-!VW(aVRtAalPcCY>U8_U=e3CRx71 z?7YX(LVH%PWRu7W_M*gKdSPeCp0u$k>vubzmx_X=0eaEv-aHtTFLWWRBmio5JIyg` z@%@F_HETTqK{zH`i)KQ1sbw*de?jB+=Jues`vn`@+v5}4gVOzSM~n#n|HpHfk9>lj z#$teG7CovQ-n?u&`^lm2``$#-8Vhj49LRf(Xe+t7;a4aeAkwb zw(U5Un%tOqaqiE$SW%V9x_8B$ROFrqYf|(F)4|}9#{}KFB<;L8MbFcJ!FxS+VXMD6I=wa>!BfG#{n}qWPlW}4gSjoNLfwMs}^9 zpStiCwIg@#(0Tc{mNMtdIej9ZerL1nx6{V z{ZohhFrfiqHmfuA95oo7uTNCRM(1&@A1MOWI^r_Nxk-k6q=RJxjWTLR6Gs%M>P1_z zC(h*A7ZSWx)!$EOy_+@m)M#VaTcrQRN^dyBqU*F~0=Y@ioZ}RL8m6P24M1^ruJ20j z`B8bVN2?Xyvb{ACXV<6mYT3S3vb%s6sTLO&jviAn{Dn0|h3YPnIoB^A(cW->gZi9k zy*IO-X$4Y=m9Z@bbnxDSxSv6Cgs`=T&-KT{!LiYqGsE%FT;_EIgr$fQFnNDLUYwe? zxagEEQad$q4cx!$kO4BK#=eKGdD@CNJ6?Od0IUd%u~6dI{ho&Zs90_SyXqXQe&|I!1h%4t=r)uR;c6L6Q~z|AcAZM zn^}PRr(pXBO5aWNawV>zvNqk`IbIjM(KRlDXOQh+`wmDXIHV3+9H6@;_L?B#m~7|k z_(}%d-s(n2E0QSZ5|{$gkD%GA6g7?=^BZRuY!#w#5Nz48l2ia!g{^Bh&dJyC?V5oS z+Emmt$dS&NL8JSCokZ&j4E#VweNK-I0_XrHl z9pq&1wk!;d@`XJ7%wKhg`~7mqQAQFWurtTvQa!qY0XGwU6~oMUN1;){%(N&L@=X-( z>R1^Zae0Q41xbC4;*x6D)?I3mk1j!ye=5=uU{ZrwP9cjQZEhwT7xUc=M3CIE0Ta{a zuA2yuKs*fQZZ#~-e0QtkU99EVPjoRNqK1JTl7=>1OC^;$7O@ThlHBHT4GxH#p3jt#K~NAW*3gwHJ>cW??3*LLIrH&wRpP1NTB+ zvzU%rknGaB?6bHQgv7kUv>+rb6wiV{PlaVc$S$dgI@F>wAGLtgs~W-Z_Ze+QsSOSM ziguK;AxG!zo5=SwAn1*~O592^#6+&CSyMUyLoO%;G9PtvL^%T2ERwssaznhSuHaPA ztQsMDaD2)oKGJl^2T^pqQQM8qP2}xl+PKi)*DSFuCcJDkGw$GBD2a%INI_5~jz&Q= znt+109jikA4@#vC4TS>l&>(QRfj~hB?F@Z_;FY40CvP?kr@x^dNtJPNlC>PGIw@@t zw$wmM2QJI#&}eJlD+;DBBDeQWBo$0`7bYiI(U3}$r>@_u$OnB-R*GubV7$EpWB-pXNpAaBdXjao;P-_MHi|#7bCGjk_4?u9!su zYeH)$jo&qDp5MEwm&TT!LiU^;4tLfEqfNltUMC+HCI6k|)!3#Qd2lLsD3&2XXQZzR z;6{D+n1N;!Q1{d?gG{+gH$DF?wQ0CD8S-b+QuZlrLnyOUn-oY(58KY{rT$00T7 z?N$Dgyos8du8BRnq^g^r!=q!y+>bTSY{~IfS?D^mcES&%rWcN9zOKGcmhddu0sYOc z2f&AmPJ2OK?RyA&hL|QRz>)D+?aZ4QZB2*c&Ea6wpAMG?Oh;&NJCdyf!Y?^hp zu$;Wfl?C!m-|Mi4mu;I7vARlQLe+wCqsh}Uij#+>!uxWX1LSWC&ug;*d*rbzcbBFY z&yAn2Jx9zwUzxm5(@P!c=LQ&e?0$LAxM6A_Djo6!W4{8jurKD#5SVl8!^L+6t2Yi1 zVFYYtxx|32uE1^>*zi9V|MTIbAf|p*7Wc4}#dl}OpmYjBMrBd2Oyxzi#6YGHh!e<^ z1(O4rx<)l~QW}*Oru^}Z&Sy&Se5d+b4^?rD{=`axyZKM$PZ>Km2I-6KX2pG*HzO zQ0&ptKQ&uZjPaz`6F9edl9G_?D7p{qN=Z+Y6^vTy%szw*vE?p_>4`!VMtY(wmzbWY zE3ouL4=cW8g}|kv#6zAmhNODJA*d&y*pC#oH}9Z$(!AZWOdzYGpt3EsqEN44G>glk zqEHhkV!nJf34#47YBOsl5JPlB@#>CHGxBJ1Jz4@FI9Myap740G;Yup-6B2lR^$xne zV@A6nuu{UrUa+YB%OU=jaP8zZP^V;N2pK27K7vinP z9DXsjqeKd`3HUXLydlX*z(pGx7+95uw6=;)lT1If3zi4SRqJqfVCUR)eL{~klin_} zxYB{?uFY(TKKGuGa0oh#P8~7f5Fe^q7YWW@Um>?~0kw~W7$Il?-SsQ3WFxPh=ygdHVwP? zTZ~`}f!=10U`4K;&AF6pT}ayK*aO(0Ofhj>!T>CDjw{E=+r>^3^LF|BhO(j8G^Dz` zDyc0czJ7c>1cdTfdRvoM)ghMF0m*a%R*qO|T6$O3Kg9c{go0+16N0yXbtP~2KXU1E zE!$he$@<3V>`=c#h?OM14vRYPb?ce9YdeNg=awTQ!LNE!kMCA{1n<3%2869u-L@7- zjmE&70G)yn**DXZR9p>D=<(WcV|aSn+g1JVl#`QK>XO)@kg=at!cxh{^0J}Zwjbh<~fCtRJ#)2r2VkPmS+UvoxT&eY~@xGyopdp^4Si#CEKk)K8@*EauQu? zt~eSXcobl=Y2XFnQR$K7Fy;~7eEHj~RWbYgul%Fek!qj0=a9oChT8f?E4Y(*5k$y1 zuTU(Op{jBurBEzZFrLAnP;4nF1olkQE`kW6MD;1FJS>(=E)UBUYG9y^D?}oCK9;(yr=rC!pSd)E>#o{%T1HyI5KiV{mRI$9vBY8G~bRMEXwY>S5z= z4v-Hd9cSvgRo&pWUCnOFQtsP)q1tsD`vCb)(y_dGDOOxFZW34?35AcUm#Kw1H|`HF zczxuq=KbM<{F{G1y+I3_JEf^b^{?ggj2r7uA-SE7wzrn6?!@wK{R_kK$?fsxvK+3& z5U*vESyGyfeLL?>@1pmn4Y$b`$fE`aaAWNN8sEPm1vJy-k=(^RMpAYB3PHZ?e;6W2 zZqzGCXjYXuG`M(U`E%)-0bh{p^A*^3kf(lhweX;-c8+uiLg_By+ZmCnUoKYacBxaDa(qy=>q9&H9U>UbV%7kQiiZe{f5s0XKeG-%YI zeFqVME)G>Jh}^i67Y{lii=W=Vlj8SB?$7vNr;C%i;mC^4IWH7Fj%4QIRkmq|SAPP5 z7ME#Wr_NjO$K|+3eFL5!4&vM_o%k6mT(@Bbu0GjsZeS1w}v|DO?90VjRmrF2rTeD zC>zTHAv>|Dr3XbTtOo>mG-+24&+>Z6864K@^$7MHNU6RF^* zQz_GRTct6y)K{RD`UHhWchzdCPkb7&)TgfrXUnV<{V$NJ4}z)_JV9?1W^O81VuSvJ z{msojya(JLNR#)vJn|)^@}=05NUKrG4PsefRf$sl6e}7vJUx>ZOedd2%yyVhYb4aT ztflERT8b{6uwckc_0yY7e-%3>fbQ^AI-^T_=j7q>v{^#EN$%5cZJ5W{iqkR;#uXvs z8Qw9L5o@$KH3fDR#+|JP@+rh)tZH+hH{wPpE`zJg7cH5Tc4x1ky4XwJhje~mW>V#1 z?8oEuMS(ry#~hqQWr4UXt9@(u`*{Q_Wu>SODK%bW4(v{M9588`+mbgUrK~!YGG%RE ztJRZL>xEpjR8>%BbStf(wG`!`YH|`d>75|hs1-6+ zJA9$1Vv3dC2};{Z@5J(G>787W-To7xJiW@|P4B*fRLYm+$Q-CWsb)ep zV6W?{Dft9a5+ukJMI_`TNPGY>36g8p*l1lXZ4BjF9CEM8Q4Ql2_w)VnsK0f3Xo&1s zsIH4y$4d*Lqn8hj8HkN-4~bgh{p7yhJ(;yc3sKC?^f(G{&+3(c!dqA=I^^*#S)-|w zbRT+;7ZJ39scYVgKr00n9B5 z{Q%++0Qt2Jxfh!o4$pzwZE3_xJ&#b4{}I+TC*#Eb}OOzVTGTt=P;d1kwm_-LtcWdVJG^Jj5g10 z%2`hT7r4g{cJY(xnh(7z>mSN?ZMkeukPc6INwoIeU6N52G@I#}jqUC6V06Adn;sjT zZ_C4=$I4p2E;lLtc?$>BMqwSLbtnpr_3|xSDtGj=ufTij?0?!qo$%CXW7u1y|1HrK zle4omW0jor0yUN(y*dR*q3LL617-<#uJ4k!`R86mpAWUcQcCC7I<0WYH0ki@F=**en zcxY|`C?a;=+;%kg7vKgs`A9*2xm}87qu<`Ph{Mo6)+Jl5(QNFe*&qNd3%hJxG~AJ; zK?C6Z2-ZW$Yo1uteR8_L(dVQ5%DWL*t#{OFr)2bl3rE!ou+@i>&v(dut!X0Q=Bhy; zZSxpE0CV0(xX!<9la+Aih+fcppbVcfQXWff-mOiycaGOB#TNC}pLuZZ6Ug5X%*SG< zklS!bY^%0VwzoQ{hd;P;S6PrBwyCOsW2o+uwk)x7&LLE=FM!idXbOy?X)i1`&Ux)A z9R*XhCu3GroPob`QC`_hKZ#ner_g*i74^Z451Wv`cQsl#f88olH$= z>DCNLw2Z4`W&t@0$x35%F5)AvLY9J#KEoEU_pO?MdZMDh8X|Bm{C0_+jv7UP>nZ!I zV7(ksG@lJ-Bh8*HFbcDh3kHlIYIg>KWVNfQBx+R2&FI(|oUs8@n}B8A93L zs9pHZ38P>&OKOEAn~(fmS(0-0rSJ1*OjK$H*(QQ)KKpH2e4xI+tD7^V(aeAvIo3u* z$r+GjCbq)+@$A04G#`025Mj>n{3nYTSRQGQ8Z>McDTaDWXoo7mgMp|rXkfZf zQxt~hLiL6fmMfwK_Ni|3g{z;b2REY4?~%g}Tg3Wsb7@2VLi3k~qO5)|c}efC%(A-f zIwhOv_re~`+C@I)f10UL7afeZci?@cWrk{~8XN+LbB!#^!|m>q>o@*Pc{r@5&R4yq zKbcC7d6kiTx*#u!pxX}}#o6d1xlTFcHJQwn zlu3xAbyg8P9d<=^cW?LVF1Tsj>FGsJ^2Z4iH@I3+Q8Q|cO=%l=5lJj%mWHZKf`8w?*TINXa)MO8Aip>0G|~41F(G-cKnXzw^9tn!YD8l z{!Hvodp0;3Zk$=>_H_io9O|`nCQt?`u~%arWsox*v*%@Zk+&CQHP~SEZ5c3VKgl!g zA6+f$nLdj@e)T~ZWqJ|*?1Mi?;g4E1aowO=Bu~U7SSIRfzRAgV@k(yd2JZvcGRm3= zt3blfYCK8vbn?`dYv45+mMs@Y)X}V+xJ8$qBc zhvVrT!|~?1X`hZe>s{0Mr}BmneOH!V-huTTC%{F%p)UDQj4Ly5L43Hk=M>+b_tj&! zc(^A+=PuuPvE-6H6+2|hRQ^Y)wVa~-WEWWBfw4$?tvxA77n@Eu|({hwxY$r6`;)f=QGwGDgt*&o{@pd!0edZE^IT-P==7f3f}(?(N@E zkXy2e1L4&VF#E|n3M7V;_tqi;9c)mp$#gs#3W3$+DZtnD7bWryrn~)pJRHpDnYOlZ*Ju!zw9D2MyTX2X||aV$~O|# z=VqZ%s|i#EH__+L+7d5NT^~|!;Z!gTOt|Y2cJ%KtKoZBkK_|MjM~$-cph^?p`R2I| zDg9s$WYt$|%xR;TicnuY98q6|NE%-AX0j9IR1TXJ|De`aaePF?1vN8m4fZ=SVjIYW;kWTFECn+LCpr6 zLJ>>^zQIf~6L0R*ZvPS8cWbDb2qYP7;avf6%08@!#taUY3U;N^z&;k;!D#F}HM?&t zM1xT{l3n8fM3lz$or$4T_y?AJ!1%xu`dFBXx34qF{x)9 z3-HGcmwJ744%Fb3;Z&2WaKLq_n1@5I%pBv)1MnW&N{eo#VrzY_};g9eRMQ`>s3K;_1iepHRHVBKhr3tV#&VjG3UVFpwpdGLv~+snYIF{Z5Ch zq$(AUBiRn7j5J)gl)~e@E8ZxmwO2%k5kX)hAH0qv2iSFU0ZP*%ES@K0C4guJ>F8*T-8{3xL`uCmyT0%k8)IFAT>gx5t~y@|r!>z(cWhB(AOcXcK*tftNt^ zU(#=tkocsBR)U!7Xi1MfwMYXC|7vv7qthjmp7j1le0n2jF#1BI^hSJy2_V)D>Awwq zjs*M;k9A%0jUqO^5#NU&s-i_|e#VOxzL!Usw}G!MUx>(DxGy4bBO@}W5IYowfYb$G zrP$O>eB7vk5e#sJVJ`oI#HVh8z}(bLECg(hP2J@0vWenc*fRyZ@YL;F(MLx?0sW!I zO)Q1{h;)bkq<{DOHAP`&NfglJnB+xDp`s&MN zjhkljZ0dxYh-+(bjgK-Uvf7NF4%R>iXFA%kMw|o3JXueL~aFE zWkKlb{&vfuQi(LzeqfDWKl&joo!M4wTv`&jVN}ui)bP1scT06zg8UWaE~RZ^!|TMRYO#f zlL9vN2#4x3#ZS9GS$7bf_>kbN6rKd@m=cMZwE9H|X4AYcDa!AS-3HY(tM{WFOtvrhT%G!N1L=fSRVj+!Mq{WpEcU5Z`}x{kw5T7 zaNoz2P0&gp)NV?{HlAsyfQpr&hAFMbpHl$@)*W`l!ltGjAgU^^Yl-kPIJWSr)qP@A zO6TIPP{<4$k>XB;BB%mJ*0XT167SKbv?FQ+{-3|TZzdq_<64DSjQj;nLz;vsp=mk$RLAbTnX z{kW$sGlr&#*Y^gh2}DK=iHg^tQ9KFdYQPS8%aA~w*p03WlO`|zDE%NL-k`cgzHPnS zI2IvGoE}A;M2EU=c zm6qdmnuZy3`21{E2COIw19w!X07QIaH-AFXQAY=oiT)#$v#XGNWQ9h%)MCSnlXdKgm0@^981WjkFSjc{Hg+bB4mjM&_2} zb2uIuG=V3S(HD8{oIE_9Hrf_9T`%|KPigWGUzMkk#Xk(uFs|etzUUXbT%Q&&jVcyB)^`OF*60PqA>gMIJAO#d&G}MM2R78$P~P? z7!0P=Qj<*a9Y!-#IB{X3l@#tbvQjvmHkra{g9of*{QyHS`(A2h>L%zhhQ}6CH}S#j z)Jo~WCeSMMh*wEXfpY3$MaHAurxp~dfl4`gYtzgWR(O!&2FbTW-Y!@GP`E2Cm;yN zut3pF=q|M^Ch{+6+}_+C^me~sV|#miVtY`!U+#zzAu7-7@^b5ZbqSZieB=}Kych#C zv*=Of@aARH*-s9A-}gpZuwuYlm=Cgt=El=`?Y`;JWsPAYQGgEcAogXv&!$zy_UI|MjfRYyI&kbm{fEe&S=3q~6o%M4u4wOBj7 zTkNQHMl=n)Ag}Jm_nXYVnr^zcNdH^ndVkM7*#xOLRAPRq34I)KL?10&=kp}VyB9rA z{{`>$*oCeB=IHd=bOcWY_x5Xl^*j|8{0-)|uo5B2uCVAztQ*<2nwK%!&#FW>+upG4 z(0Tc{xd&>bOd+W77_VK2g?Q;Wz>u& zjwnvmi?(7F^-Xk*x0r2oZAZ#cuE>$GM9xk=Hy;uJJQr=y*X z;Y58Ax$uLe$m`K+g|}>PO~l#t>AYIDZzP&{l~@_u{DKk|LEc*s_cKV25VrR4x&C-K zI5s+SW;h zs{TVkehHb?lhgf;{@a-=h6P{G%c9JtDZOdG5A5i5|?W&1=lGa zQb*S*iu7qMs}oU2*H!O<`|AaI*cKxIqCx(lLXwv0-QQaZay(VDH;4VL7I8G+PPkjb zIi!%LU})|U7@9lC$=+>Q7#igZdH9*X>Jaz)<&L9_Btl?kj>V;VbOi%$Ci*Idneki^ zyGw1d-hNIl$Tv~At7Bzw#N`=E79{nxU|dp*{3riZq$9wj2D6+(7C%}f7%Fo$sC@tG zUNIS~n88W7wCcMVh#E zOmex`s4G~GBfS_hOIC$@X;YeZW7IJ(P=AX?PB%kW6YtXI-jX%-2MISg%$u!oF3BKJ zsSvdnkDoGNCagjo%_og$;8z@N(t^BZF&(ua*`;;aXK^hEiFt)-K}c9Ao&|v(rXCm$ zjG8D_i)vVc%RSU5zgAOF^7k2SMyU-A{K{qNcD|nhL2v9;;#QI&CUQ;9n$iInazP=G z`FJ3Q8q;eQ$=zMKA>LG1a4Kk4jSxMe@F^`C2l*h1jyGz%(YcAdolF}S`umzCw#6*@ zszuxiW?(DiT_}l&f=EG7C5}cxG@3jl%Kt&Bv}q_51%++x(k_`gyho1_c!vgo%MAny zLg9#^PY}FPH1g!l&EfPn9jZrCWlmj?GoH{Z++5M|q$$n(UQsZ85xKp0BB@}iyD-TJ zp1pMg`kEi~Jy|KLWrOkd4uGzfS)aEmPD5YTJj4B_L*;d@YAIjqmi}bwCsJ^Ye8~44 zu!maSpl%F9I#fO@gkIRYAjVHOf%y*Zx5bU*(x$;OOXoS*+73yIyFcrFP1!?_tRXgs zN2mzB?vi!v1bvJGQU4ebHegw{?PziZSyC!j&>IlxW)cam3QCeg@) zQ@KO23<)|TeU;p>-Us$-2LW|Y{W8duyL8j@-%^`~TazJwCM{*3(l+Ei{X6qMl`IKV zf@{T7Ih`YmIiXWJ!&Dwn6qe$Cm@)XDO15*PR&|iK07|5ChCy1geY^$uU||sXQbF8L zN6D?L+FFOcQGWvS8;(P2(A&Ez|B}3knwwIaQy%n#F=Ot>nrF7;c&jXQomo5KIlAeE zyGIx5O1txEDFzQc-%LAq( zG&MScrvnYHIPZZRpcB7$HMz1tzUg}%_Rz9zGcyN^P_4{RCuAJXNOM zpS-=Y)V)HZ=PksT=7sMk7zLb>yxBwGY3Vop-5)oBhBH$h*Q%2fN+7nGiI;gKEhS;j zyK?(laaC#cIO>+Bak*(J$t)e0OG={9Wpwt4DT(r_3bvA8ipIAw7H~*>p zDP!lxAbt89Ubvv)_muq$p#&s6+{6-tH%(XT8oU;~FGW2;rQ>o5-U@$=wu~6Ol}{xH z@7XnSgLis8K{j@(dIE|)TKcDEYm)B6>tr+P30!G;l9G_?C_+0Ca3SBYTV(~KmP%Gw z&dTm6lU`4#xr33OD0@syPt+As(i6B;lz7OK#=tAnQT2pFP)|UyAE^g^lSr>9NRyzu z%@-CJ1r!q#5-~;KMzgr&iI_s5i21S$o2P8aiI_qZi$k&|Jp-ckXD8&*ZYK$XlujQ&U3K~oJf_3BpwX)NCy7M4COMAaL4w`V*KtK{k?y~N2t+Oyyl5WCbPUF zpIt;4bu+U@0&iX?Iz+~oMhcR{9~#rLO-d?-4(zg|X!OEX&NsmTc=8e4mjkVW)C0+4 z zWF+9CeGQC^K%drD(P@(DhZO;C=%i@a@x+8f&|%bN#Dqh9sH%P>sFA+HrAash4d4Tk zD<8sth*Lp0mI4S^z7XSUK|K?UUhEU*{mJR|rkh(ApQoy!_0|`KD<{Ncl@99hABf~s zN*q_esq)$uxdFD9vorWyuM7VvgNp_=9L)=0G~{sV-eR-}o^Hv}BC7HAkjora!7^E! zhTX?(8hV>K)Iz*WPEv%N80S*5b+Llcu{$e*F+2h+bB-&=$lJwE6Z3ZY`>wK~KdXIJ zQd>%V{rGqY2<5Tc+v*TYySIf2SUF;;Y3W^A{}At+5(=74PQ)nU+x?GRy8Pkx)^M`E zF*-X8zCsuUcYPfeb=q}nOGGl>u%3z2^gZ|-83}&XlX`r&+9P=HeKa6!t?IUQbB2!B zkcHOK2SIM8C#kp^p3vj9;l}Xvw70AJ-zhaVvD788Lm^|qk${@;$MUkF+qNITl{ zKI8qwihYsy`N!nDwC?4@@pkP#CJf!spR7;ea>X_ulkce(t0{EXG3BWf@_a}jp_(T# zW%i-528N>h<~fCtRJ#)2r2VkPmS+UvoxT&eEP`A4Y=_*E?N%V4#&j$>i7qu)91Vyu z*);Hi@TfGgK8$&UkLnmi1fl=RKYAUh_L+NvJ6vL@t#G=2VJ8ANgpBhF#bOz%Do0WZ z#bO2H84L==gusJWgGbmKQcod!&JKq=>x0oIJSuhFg5;AW%_C{R4MCKsK4q1M#d68z zVYz}E7-(a|KX*uDFv;P1N=|~xGig`#;uBDBKx&UsPazvXuLdzioYhd?)Ex-n8)|8?Fs2m#P2FKcC*91(SkG`@c*1#8vhktBH@BdI#xShVILFL%48~4E8n$aY{pVJXEyHnU9l?`yYvl z<4L3W;R6iDEl=AaEvTzxr4DRzF_yf|KkgxVbJT;=JsLD>(7uC!)0;PRIrT~Kkw$J@ z$%_Y_ki}2$-wCq)+b5s#zfKn?b;FSropW9&dK}5j#j9-7t;VZEXmOe5b?UqY9+%Ug zPt8Yt1D+oa;@m8qGJ|j+t|2i9&2%~;&kx5^8kIw!#QRw`dB^z1Ziy>y!k`-NBf7MJ zhvI{j-JamcE8oMHK^qp?s&FlHJlT91pF0o`tEkul+r4k2260>wrO|4*@)S#sjN+g( zv=ZpBplF9eYa?01W-v`Sn3xw?C79SL?ke8BsdoKxTg#>?9-3$F< z^|q6j|K#0Br9NtC)L=s~X^o`Z2_<&`u39biiBBVz`t&v7 zY?+m!{{>R@K~QyqC+LmB%njvAY|wwOzq#3m_kjBYf4xCC1@a}N@}=05NUKrG4Psef zRf$sl6e}7vJUx?^Ed3;6w!^$dBcaA+ElsDfyiXLM=r zoIE_9HcO~C$bI_59p*8%;hDhtGAS?yZ|AG{Cg zp{1-8^&zE(U1NSTQp&1RDO1+;AtYF-s-Vp1R$5I}#itQdReep_21y$d@>Qg?PEy6I zr6>nglas(n?*zd{t&nE!2sNIHDOP$XC~YUb6U(EecXCD7?LTjN_Z6g4z9dKHK36g8p*k~Is`JhleOA5KyJOg=1lnL)A_x0|{EE8J3$2>@nBk}gEUI|FN zg_WYCx27#3B#(E=8cm%Xw+*9Z1^UzL-aHtTcg7Z8X1546n8CTz6!g13+8T^b4<|X@ zs@c_R%>sdO!|yK5Ph@roL_yx#f>buQ2ladBC$$K4^h3Q1^iaV)llx=EUJ5H=El9_6_8g5#?JGbCQ9_x^Mu-(E7MjIQ; z=ldJyE{pss&=x-0&B1G!GZxB-`r3M1k*I7@dY2;EYEI;E1+{N1l9ys@@39M8{ms$o zwdrVj`k~%_?XP7!?h-%AgB%i{*6hlk-AXQnstL%3UBf!$l^yaDYz;fne`K_IZd1;3 zvc%^(o^hJVltuZu5pJhGB^cS%NB&}^n>Hnz9N zgVFi=YF4=00W@A6i1_5YU*k$XY;Z7+H8UXJ{ zupUBQ^TeX=lhgf;J|E>*-i^R&y`xq;CF9)Ag`;84^79>XUu&8OxVdT&NZUNd55Sza z5w7zu+hir&IjRPkhff(PkEJ&6)~4G#$Lp42i+bygM%=aw0{I()`B>}}k{k|+ZPgaa z_Erb=5X|zYcI1a`sw&_Zs=K5uORSu82#sK{R>I;s+#)yRJ1{MWeT1_bMLAeJA14Uby4XW(KWOvYuSP+>|Ac#ApjD#(lStzvW7-)a$=_Mchz<2ct8 zA{A7cHWSU(BNzzi+CAj%-iwi-YYIU&*cy4XOK!)Mk3gfHOigGh)dnP5#?>*ifVdKO zm*&HLdl9)HuR@lBjy}T{u=lN+fO?{$n-mdv$weTvfsPtQfa@vyt6;qxQ8b^OW+Tm> zEHDbQk_!fu*?c(BY>EaV%-Z3AhAz8Q9MIdsHt;_AUKclZYp61WvcFNg@SPJz!D^P& z3Q0B}`P&0eM?L!zsIO*BRB8sr{u5QkdMl%CyNCqm(z zqH-Ks>`xXkusqTpHE7r@QVjLz!rWhpnv79|7%GhwIt(?}rfz_$R5e-0<%|TzW-fk_ z{0~ZFTxwDS5jK}?EHx9H!0`Tdlxr`I{YxYm0#r+`uGuLDXh|+D1UDy#7h(hSsu%)p zL2lB3{A$5-pvJid;%hzzixrunD3 z%@?koQx9%Lo8Kdc9kz(|;pWnY{Dt#a8j6bfz2qgmyD}^0y6co|;(4q8!K_{6Q~sx! z8gXvV18UkCG1e*$^tabU6jSSy?l}Vqp{*3V$Z{r#%~-3^&d!bNf1iU=Ht zH+1D&+vM#9Sq(NAeOm?$+E4OK`$t#HdZy3fk6(RoVPWAS{MiS8j=~=;`*vJ3UL;S% zBv>ZuYQD+IckxPY(FX4W*D}hQ2&+KCx_5h4;{@|`^3;`UdY#<|>BGMT{IC!H9F^`4 z4yoa_A4bL$8gG-G?a6TcNF~2px6Tj8(>sRa&2!T}9e38drtweZ4I%ojEWNw~>$xP$ zx6maYig9J;Er<{I_MGC|^S*lQ77zDi=-gd$37N{b7BssBb&9f4zrYF)j78dO?eTeH z)}PXY*8bH#%mS(YPcJ?k;h;PcD3{|jJU~pZ-F0Lg{ zTj?>o5G?h$UkfD*_a#`f9MFcw7pj?yb zcr+Le)+ZYLdUAWc@!g9d^4` z6g1li?<@q8dDzXw2|8HBvm(Bd{ zRb=;YYry}X793{Z5t#~g6t`{-h7hT)hlf5fHF^Dz?hYI#1nj7b+1l7l-Br*r$l)C) zpKv*KZ8(G{S|JqAC46j={34o500!z#0VCQWnj9B&#PGIo-DzXd!%dtDrR6#f_4d47)R)v5%+PQ1C5m;ADe z%ow3k>~RLA7s@~q)#qlRQL7151vk;>&e{?$P+cEVZ{bwv7?^N7`u7+hiDTcO6J6S) zM%j5#rHSu+^V~*+tL2p^3Ia;50M?k(MlltkzIr&Kz6y~vyynehC(5ZDHY@%?t*_$v zh=>bnX4)F;b-F{XiVo*FLxB}eB$2*e(Xq+tW!WNi>=mR-p5{F8 zs3j|~iu_^BaLR^XV6+&6nhiFEBA5t#gPCL|-rT3%{v*2Y)=)DMNHW;My8_^peOM8V z84a*hcrf``bO)oc^VICVu@DVL;YfCk0}xRvYd31JGh3H=NI2Kt5rdF16WYk}BhN++ z?or58hVDsUBpH&L>BBUcTuML}cz*KikfdsD`^KJM)3hg8b4vA^FFG{CwP@ zymn7jjDlcQoxF0jt>$kT;~}MY4-a?HUWq(-Y5N7!YEDh!eZn8%9hBbc)zHpdSRuTV z`Q%Kq&|*^*ewQN+psw<{Ig5ZuV-HV$7D-VB)}kS6{#CKU9uElr;Lt?r-cgoC%jnk+nrdI5R@4+L0e%UODtt3^SDAH+c2dfukOT#TR|Q_N+^x1EOux66}M&- z6;zn<#&XC5IBaC*R7K<(^GV&(a`m&g`G69Jy;n1t-c+=-dOZZbDq{ohF zO9Ko4YIM@0(_`t#!W1R^d+i*9c%ToSx|~4 zye>LF6`KSR(M{$^it(w8%uNLWMr+7!Jisjd$aGr}#N&HlV5r$our zdJEzoR9PdqgV+|3pkv*}+gy3SAv)yoNY^|wVzrf_{~yU~ji&l963-l2RFO<)TNHmS zBaK}nikme{RFH(!Cc{sEu zay1@!bejl63{y(1+7qZD`q(2~<*4>V#$z*c@^uG=tNLEk3->|`$rFZY$~V^2lnpa! zBbjMyh{OULZ6+Pt7cmQ0aAVH1k{A6M8u^TG7_PHGa)K;+f zW4?SikO0|JG3duVZJBX!)5ITCATnY|RJ;a_;z=l1r49Hu0d--<--k=UhBr$zsOH;mIHQ{Eo!+d zpE>BvXM|59@6B@dit02{8ypr=fCZw}ijC{H3zweRK=ke>dsp^Q#OgtS^%nGbT^uq% zC-`yE)af<&4fUZs1Rq2CXgU{`4ebv zFg)7b?l#C@;%*f}g>#kJ6%IFPledc4Y|)pv=fJ zk#_Y`@zB{Cc~S3n|3hb9`Kxr(?^F--dO1vo0cUz-)(h5(NA@C3T(S!H)H01{6*?_F{bbguSnTQqPfi=fftyOx zR561oJ=ove?88OtepKBRrt)Q+>IMBGxx`oum{Dd_%mh(J-3$w6&h(SKGdo{k3fM?1 zF_=e_N;GH4ZDaPDoHwQX9F9i@P2dS-^hKUKCl8OO&9=o&a!>xil7INBJdG^=VTgut zCHL@U_uh~aIt!&?cMafvWX3^T*sviN#9(2S2_gn&sa0Af&B;mf>sc8yQ}8MZvmcMs z2eS5vAB~6-L)?%lcx5pdOsST0#BO0B(9uPR?xKPs2%%&b(G=9{67kXVEPAr@o6%pe29;IRjo z83TiWF?a?TncaweGe#Kj1_s+;!^FKeZbaN0aZa2QH*RJ#qwni2m38Cnan65EL^Mm` z^nr<1QnUELoBo=mZkitN_Siz|rWb6Sx>di_^fcV}H%fgBEoI-YL80Yy zGqYGXI=(|Ja`w!{b9s@36M8`|E%mDKCa=WOX5QJ(I~hHSS9!KZaFP<+nHPQo5(r~ z)s`e+$>I+d0xuPjwIS`IyLjKmM(+*SWR)3!sv&IKX>fcm$A{9~k^kTnlRUm+t z`1i@Ay|uH^J^A3~&d%=nosAy;atB9+|KF3jydU{|Rxfe}%>X+p0^VWywEumk+QovH zwf%AEWht^@$hR=B<`2#7?w8yH0i|{KawA=J_(G)SdSkY^KP|a#J9xO<)$|RBUHs%? zs$DhLb@!GFnaTqX*2LIP$^es=yj;<(+b`^}3wBy5{wQ>BDOvar&Nnxg^5Obty%m4y zBER3{kJTB|-Q(HcdugM8;GArPR*wn>`lV;{hQZk`NM=CJ05uYkId`Y zo7>Z^`IU?NbNN(o_s+_H2A&EduaSqXREdzxu2Rtzi5tOIeRw9Jo7hOq67s(MH>XLj z&dXOhx8!xpH_I?p=@SDRU+eMOvm=WSET&2+70=yU)xQMZ^&)#i)x}aP=WQq5+L#}xit6nAAg%W&+JTgfv~Kc<(6#2pcP*sy0S88H(8ST;l=^5Z{uCTa zETvTIJd1xA(R%k=?4|kUtb08Bd#d!N6c`()^%GdS6k$M5W*(Ph%XEK!eRH;#Wgc35 z;wzJ3h4=4l?*(hu7nUg69cRSJPWM><__1U6p0$CrPKNF#a%c7OQ8p(y-OOSR+wT3M z&#fX6#F~pu>m#g8@vR+!KZAs9VHcmeG2NYQoSi@M#B6uwhj}YG0T!rmJXm~tM?4mr zwqSIc7HJ%D&U~Kuy#=w6m{M=w18SbOqRzncK9NS!w5j_09r4!0qTahQ-JI^KjJEg4 zDFjacv-fSI(B4}Rq9B8YIR84?WO*v zRIxT8!ru97>5WWwdy5ktt!ScmC>xUvi}s%*Zo4*{U0>grZ^>LyUJAr8NI80xde3)i zm((Vdjf14tT@FbZ!Br+d2xj`L>HRt%N@&wipGdsZc{3<`awL8@>Bl23#wZH`U|&{N zDnqA!NIkhtWs^RwWp&!gtF{`NZW6!NNxp6CD*)I-{!}H=;RW$`I^tZcWpB-<+bzz~ z{7%H#l2SqzY1RzQJM|3BqvDb7!;NNWtQDHg&-}`Qz~3)NN*P0ho}D?%OZBoZc;;r< zM=|_@@3$#&xI8vZFY#MRq+6YpNu#d7rDPE?BaIe=^QU#89 zFr-daG4}$;p~_rJtD#8ZZ|uowck6QJU7nSe`ezX*IQ-12IhW`mP-74(it}$7@Do-{ z9hH^lVw<+pcji~HomG7Jc>L6Y5m%+lUY~18keC-UEeR5qY@Q_nJ;f|bg6uML(abE0 zI})(aqbk2vZAbj825m-T4bJ?^voh@baf1Fb&?u9yA)&NbrGYG+`1|kzN`}+|^&7485idudh8af_)P} zdas~3LpiFWQcCpIo0n#nf}Ri`4ZWfabEc;yPScudWr)ebRqjzzG*s-w`LkZjOE&xP zU!tN~&Xkh}$4V!gWr(mD6RS#-ucx@-pf1flc4ev4GNk*S|2Eb%+uoa%&!olhX?(-* z8RjhbRI(w^+FT2nn%g-V35R-1C(Y#TDGKP&4|a=}2tJhzOGKHv)VCg#h?Pt}(vm91 z*9{(!i}>kI;CU~@x^1mR@H#V}?#cTb&dJoEd%N`?(M%|5_9Nc$(;kprGloCbTw@7& ztBm@cx}TIey6=VK+UoZEWO~nvc*^eX>f>GcLuKMiLl0rsh;cI5I%D^1Jh`+?9bD_Y=PrdL8!E62y$&tBb;dY8JUPT$%P* zk{oEu4i-*#i+|Y(yf#bVvD~|wa_M{V+`GJr92`DB7|o~erH=TzL0)%^zr5#VICBu2 z4k_{QTUmE--VBlV+=k{#0b5sed2>sss=*(wB#9vK)E?;@O%HQO=fPh^4pAW!zal_H z3S_9`jGT-l2Z`(iqjdvWC;_}XdjYE%pStR4_8BRz-~fGul@co;C?op9(1WKTD{!ix zW>U&(ph$bNV~mNe)rnY046?)7Y*neI!hQu@ zJ?xaEPKQZJQkNPNi4eBPD!k#im{algQ|a{X-Xom6_40GY%0p*T%krC(Pd*c zY>OPP4xyeUF&@Qj*cMSKHf-y2ksG%Ch5Up;H2?wow2oKeA4P3q_NoTlOJyvfGeaJu zvk3AUiw*V#gEDGCksFysV4980`e5`%W`Co&bF%g-T3A#i_>|qx6cPC@O}C#c_Oa~W za!BxG`ER{fdOJ5B?Pq@?IPdKFa~A(vDcUu{nNQ^Aou#V^gFn*EyH1DMypbO5A>j#1 zK3-o=j3?Nth{Y4^Z#3P!4TTlF@);w-Yn#Ma3!ntq zo7a&6F}Fb{1!*FmECFfkQpPB%Q56z^Y-coCQ#wGIe31KcGT6w}Baz1#e#85Qd~&9wLagU%UWx!#C(`!EEeShZm zdf)M`XI^g~q=4_{l;g@xCGd5XMUa~heqn?Ik!hGmr_RMZ8JTT}OA8Qhxa zNXtvz)Dt?@&X9fZ(E!@FYWMuRm+6FeJ$Xg;eOV>N!|=2nUz}~uuIzVD6#tB=ob;!T z#1FNO3A;rZHO2Qd%Z47_x!PTnf9o|DJcsIU3ty9uS>5|*yE`g;oM9JEj|@hmb-^y^)X|Yz}}EL zK58GPcFqqU3ude*-1EcwYVnTvlkolY9&>Ox%~Ffvr@hyE8Jq6Q%aAJ=@r4q-6XHk1@9#x0R^rhf z*8F#(IBQ4)=dkfjiS#2oA)am~eBo@T|7qnLkawNAa2+qiyvL0k@0Pdqxk$r;r73>{kMEnW0NLx)fzzS*=p`jPPT z8k4GWaLMCuhoDa5sK_fuC(yiGd|&vHD4tB5=0|8?koWnl9nyli_E8D?b9eD~!`D5q znad)0#-qogHrfvmoH3t(Vac64t>*QEPKni@oxclbBHGu)KMB81$J4r}WKAx);3zte zBymfXUa!&O!4i#^;;EYKuq3e#i&iIDOD*6mm1!k17=c#zpx{%J-gD}5}^;8s`F)n?f^5(z0=%H|LSyWYbxIZo^BBKlrrv# zpH2i{YMyjjodj;spOvEOB|$&+7rho9pGoUDD?Xku+sUJalTfD>ZI(_aOVLeu(*mgh zI>BW67gO&E(ss&Jxw3YaQBnFz`^yK$C)Xr< zjkEET^YLL`s|l=>-xdLFSurv(bceO$ZuRBJKy%-T8F;%8a$MA7f2*J=3VX z1PGnKe^zE7PimO(aq;f%I~&4;z6X)Wx1G6FPTL(j-C3v%+CfAA;-ib=VwO4~ zrw!SGJTRKbIA0xHb(wnFF6Wx(jyy4tP~$y5bsEBVdtk$&$n)D znY*0+3WUe+Tuh%#D>-y;RsYg!*OzWIgir!^O!x-pf&@7ehi zPt0~_?Y02Et0UgqF2$zN4_~>)o1y*CqS$VYW^X?Y8Ubio*}c|P(`+Ls1nvX)sZ4Up zk=Ry!q1WEmrFm#WB#86R+cZ@<$55OlZCPWzL*7Ddc)b4{aoe@o?E3n~d`srW_n2#? z2-275Q`*}WojR{QqsxY=+S73kDu$(L%)UBV^=EBHjC@q*aE$zVYrC^IPn2_4q#<*3 zW>9>p(Pc|zfdqtWW#RTTdW%+=pUG^`UDezGV8ubaBYV`kJqOnp!ZbK+L{(;xRV^a% zEgkWe=3cQin{In1(-6!byK%cdwI=>kqw{g4)GQ*csk99}mG(~Y_U>berqZmnnoXpA zU{Sm|rGE4@+9RMQo_wNr^vPOuKhirpgk){$A@!ADmJER&`Q~NwUMyKOAwYQ)CifZE&NKyUDr} zWQW7A>+wW;|CYrjmo)kXpeD!K*io8XkmM(}nD;y64Y%)1gkjl$_`i>*nB_74SQ!ng zPl_ciUGnhPOPY*j6JkkeET+Sf=Gs{pIBb(5O@;l}B#m)nNzO#rvpn2=Ly}`J4*zBn zEWxOj`Fh1qQw%McPiw^;wwWS3b9PiL!Q6sj(lhzhLgYxg&#@-H4#nBTk0&k;oen^Q z49vXCku}NG{kr9FQhUBwI!*o3j?gh{@xzUd`nws7HRF}k3`-HV;*MGN159p(`zhjU z6Qiri=DOsAfk}1X%ycD9QSy1Nq|uN)%k@Kx&I?zo19fp@vip0?Wrr=!`jqa{hW<^g znQ2)4UE&?xw>5;-tD)0~DO->Ba@Gm)L*b{H99(o`cjvl%Uuh|yYA6~UdJgA#29}qn zyB}F!9rOk?OuhW-{nNdDBeCFgjQHV>c!OD#w^o&dazw+!!!rB392Vs5L7~^&P$gvS zQ?KIW-G1=lhQc+JnGwhMyNrG?;>L~AtTI*S)OefH(C}g+we%WUs4c{1!-(=Bw$R(! zVMEit-bgqmT=m90=ZYAdlka6W=UY!5HUE?pAs+{$Su!K;pmP}6%Q(%g)~OVl%PyM< zYGirFos6t1lRbnKmhhq>$&tz}$Nx^F`w97U_iRZ$SJ_Mo#{Y~I!kg{?ypUq2GB-#y zSEO{MnESkG&Ln4*^av?*j{-K#B))b*yfL#k&h)oKnsOq09I3h7@&U7i`;wMZ)s+CL ziQnmnN7#a~tzT{C&||I1LYFZB{u{3+vTHv1l;mu>FMH~@Px)W8681d$mDSD)ap%n` zYW0cOzZc^ufS{zKl+VPzD!}00Z1ahw(!YiY+(CUS<}F#z8@gubSA0W9oFO|*l`YQ< z+Gm&d_I&!8p1QZ^_-8)%rN@pPdq)0shy3ea`Imi=1AF`OxcDYcg5{%5$<5q+$2W4o z8iEg8n_bpNSTQ6FM7NhamxjJjJpc6b&0YuY$bY#*{&jCwq|&jUoQ!?hc-y&E^xyY8&}Jm*Kf5yjrEPcL1U-$Oc=$II}yDE)T%e%k3$CdTRkR{6jf(_dA@FB7xrepYDRd1fl_ z0xABl#|Df570^I$iDxN;(hxfRdmN*Gk$A!BZf^8iFqw3+@G;%^y!EH^;|uCPZp#1Y z2pR6_l|aslPdYNt*W?qzY4T6m3D$Btk+LkNl#3QpCKoH*%k$6o@xpyoKzZ0Ir}yA! zpGEp-)L-;y|BW4Se`DgL^3(&&Ve;;kB+|+I`VM0{*kfFK`@8dv*~a?bro2ruyJPOk zwUSi-6e5|nj`;hW-Up2~BgmJOyZVdHyg=W^C;md;UriNZC(AbIodqJ916+u;aVNOT z+!5b#Ts+jo{wfOY-f^+e_r*-zz2;uW7?z<)jxj7l`=BhC(`#1_WIlKA$=&((wRL&P z(BAsibnjYs;qGHuukiBh>U_KWNvHpOap!DvbAEl#{x$i-?b-WkfA%hMa<;uu{>fuV zrjp-f`_T5rOzt-3B8kPcmb|{4I2_1pkm((#kaA1vOh7=KNS=ndrMM`M5-h|wwL)mR zknlan#ha3a1S!C3P)M=cC3Lw`(TvO6=(f|wM+eW+;2__zX%q-qy`*$oRxhQgmx+Dc zJIWFeeti?a(@m8G#Zog5^Xjq`4yWot0pX7>G)tWwoTRDr6C^y+bE zd~2*=0tMiv``nFKDgmlf6uhORqB;XZ23*9Sat28dzNJkpW}4KczCNI*_Sl(#VoKFBVDGNtK8KKzW63vPpYsxdFF&fffl;Q#P@P8r*z^% zPK$R_gT|(kD26fL;3qj(Z_}sU{v*5Z=9rn@NE(Qx>@b|$`u z-NA4;kIn9#4`rJXeUiN*0lA|T-tM(vx;)TnDP1RsZ{$qK_z87#d}eLrNG}_iYG8XN z7BQry4X!c|lHDfLloDwRWq$GsN>X)k=ZXET>C;UlV7yN&nl1|Q?d)ckPlFlHDHNHD zk1>7dw^6S%HGWd>E7>`aBPmZ+)70Rk)Fon0AGz{NtlW+Kjsae0@KN^uQ%XSPwUs0X zzFw9$7cS@8m{P4#<;1vTxo_;ehb=DS<(23~7!3W_9q~F|QLe&Mu}8_xs;Io0TAPw@ zj}svc*Yt{0HVT=e7ZrY`Yq_^3m3_h&mx*lm%yVTTd+f9*JDGo3YZgLNhI0I_i8ds4 zMa*sT5vfvc@e~#jmW*_a9*M|_e2%M`p!D6bsm!*Cz@^rV1n8FhsgAg}ndg>SNfY9w zBwP2{U+;*Ad7S2r!gB6hWz*2)ZIXQfyZq7IB`@!vzI|i5KV5%fXLoD5-#yVi{>U5N zcE9+Vj(81^LH)`^Ww)#p+;A`Tct1SBk`37idCrSz8&$dt6aX)nT@#-zmi*`cFJx~c znm{KZY#{7jBQq<_q<%F3eJhcs5>0{WWHmPy`U#7?WYN0ty0qQE`YdMzv@4!_`Yto~ zoR5E^-K#HM;BL2?yt0(A8dr_Pohq*I45m{|5NDFg#$lP6Q?Vnfna?aNEfxQ+Pd+5s zhJ!~lbGs>Mv5$x78uxkPcHNX%pXMX3!9!-%vlEx|dEvjs1_mGj$vw0@9P*+HdZew2 z3lx=quzGiii6V{+I#?ZfA~Or$QILqNCWCr%+o)tUaPAN{4g9y*SxuELlhwGLi)sb~ z_B5va?++)NXZyO1%9 zDlkVklHX?6hMI!xU({ekzp`~~PH)b3*H10gjw%&9GS}*a@g+QmQCp zvWcd0IrW&yrzXJb4UIbs^0ItpKCKBV_xJd8Q-3NKF4rK;pDym?)@rVeNkJg!m$Nf% zPg+?!XT{tIGz;ZMK$;0Mnd5Sn2V_U4s;8vT5fXhAO`g9j|9E~9O-O+-JF2cgq>m&P z3F)Y^7D*KA&|$M5;2g%!_h?N=kzYE)J-*qg&^=Jq4;PbDs*u1BAOC%J^;9EY|EE1S z#SItvdCfu`D)2KSpr!(lDXMb9z*Lhn$f^NO(lhf6@-joy{}a0mG(uG@=pWd+P7n^b zJswTRIwV^Y=dm^psCe-a_Lf)fbZ*UX~fJ4AegUl~rNxQDwG3#`Y5;v4*mDo0h(4gR|0#y0qEan(|%pq8{RieF_n&z`z5 z+r3F_o+WS&O)AhGLtASZd-My7@*VryjLKXjBw016X6*5&^9t>`37|$16R3;^vlf)J z3bV@BW0-CViAUzgA(%CR;EXY(0$Js)cT6+Q>>n`?=GV%drE2uFncs|(w#|dN;rfL` zFpeZ*OK%?EMe)CN#H(l|I4P22LBk%08>5&i zI7;c(o)F*D5vS^OZ(W{kY|OW>?k(AKwkkGBP^ah$@YwBrdV!s~->#l|qf!08cEo^7 zb%t+B;(JWGUkuT`WuIB2k^MokurubWcRQyv)*Dpm88W2w)V)3N?YjTsDNS8_ka8`c z5HBPQwdjaOY#8bh4aoe7rmeVt+!3$NzGg=$gnjL3MAnRd_Fe+nYx4Zbr!kkz;{;hlYnjTEv0b7D%TW{Y2RttFR!m`>9C7E&y9fl=JY_$N zWP0GSke}~}5hH}?}5E+4`(2+a!$96dOB- zP$ik(%PFQT7c(5uoJ&5i;WW!Q9F?vX!_y1vJms#)<=~6hI*~){#l~GpvOld%vhO-` zhrXR|LC!i6$E-UURLYcPl@gP>q|Bjwv(vbSHC5iN3rcL98AswGxf0x_Gv`kmM~<7F z{*Y%fhq`{5Lc}Hl zr*A3d7|ZxC?fNLW-6yn zRXaPRJch}1LW?f18$x>|(lhzB6+6H-L(Tq5~Tdo9sq)6mnQK=Fkpu3EZiOR zW1cf|WD~Sdi-uBi=atArc>N%8V)SL6K_N1MI*=K(BVJks_Q8y)b(df#TGbp|{ErLn zL#rdzcZ^`o045TdbNcXnIUA3Ei+g=K{~zMQSY*-HwaN|*vYw$DGp$3mMKw*o5W*>;=e`} z^*L3ZJSbK?MHUq6RJlFN9H?3k(163=^IU+2$C%t8iybTf*K@Mf-n~u#Yk3Nl@R8_k zNJ-x4pH%zsQx|<$Fqd!}sf+J)H!isY$0VC09o_*&&pl!f`%gAARg?VB|{OwmvbNNmNN{uow+r? zlCrZ{MF9XS6#vJ9cx{%G`A7Z?(&%4Y>bkpl-^NDoHE;=8|C8>$nzS9N{vD-nDcCXs zE%w_GEqrpVBmU1u(%ahE=$?FVb7yDw{LV(NnHv7;%23|_;ymxd^mmiXwP67Mw>4`w zy<%4KD;@Vi96E>cLE@|VYoELOhZ{j^*EB*ABn5XHv(5eK;c{2gAH>~PEjHPWa@}=r zxsa)_=R9~4Wdltey*U}VvJN@aF52mm_+`&!{d8eIIN#h{%8xWXD=qGqiG4W#ewZ=W zJ)ZsDeL#&aB87{%=d$RZFj8)JdzP=b&a>B5*Y|H;pXE=iqM>+EoQqF)&feUfZq2V; z+@H(0&%1Y4{e^MaZnnYNT@=6MdHm|kw!M^Kw6}F+l;_y;6Z&TnT_`;P*y4R+GV@hrCuaH{1$d; zzB%h2&;H(IJ1+%7!pZ!El%}4&ntn<41NZ0GH)R*pdeWTQ82-XW_|E-1+k3hW0+M!l z!H9Zk?v69!WT$(qfBe|7d(YZ{NJq${o()`09aGHac&3|~-#658lem$K#F!s#P*}cF z=-jA*Med7F-I(sqHqOqUcw)9Y^TTb4Musz7PJFZHzL!|~3_O7C2{>4o%`w{sv?zXb zLA;%l@&OX`;>XbJN-~x8oXd3Lg?+<}fedLzxTdxgA2-nn8Uf+^YiFNSu~NvO#-Qc8vtVM?p$h%ZVb3_0PNjagsr z`A-k_ULA&@FYfPLKc~R%BsQYQC`cm7a)A^cEVtis5GIk8xfi)aSgTYJZKB4Q#>dl` zhU`-n6Zpxb+iNYf+dx+TIpVf!v)T3ajro?$rS35=me2$#E6rkz*}E!)_woZ1+f+90Z!SN=W(j& z0aQ?mrd(Az>Fskx?4tk0LhMdb>7cU!JFmW*I54ktt}kD7%~NXx4wu!AxkOn=Ck5Sb zZ~C}pTP!sXuCEZS>L-Y^#eSw#mz*_rWt3WW$Qw(w535i=n4Er_b-L!_M;68Xaj8h9 z4PNhxM<$R(M?7+)&br}|>lU5x$Xz~jmG&|tQV0BJn9UVq&`$UJS>xtE!W^=6crsn@ z6c)~F)i}hMC*;z5H5^J7vq4{0kIE2Uw*IUxcO&nLjJQsBJGFsg())0<4LaBwX~!^K z?F{U#zTn9lxiEUH#OX3lMS&b&5WVqQ{p?6HD50Z<#@xJ*EwA*Bbc0F{o6xHUi-ms| zUEDeS>&PuHpG;^X1{%<AX`n@|5Q*=@b~W!cRli#D7Nq z#iQZ(QOwbj0;`qJ5NwYng3?#w+VMZp5%3qlCOnl|LYV0_}5(pGNO%w+K<1v71H= zDa3gWT|YUi1;66+vAgq2Hb$QQi2n^cZk$>W@`NICcXo{e$KsK@;C<@oA-`W+D&HI+ z+Yj0iO|HIu$S>Dp&*jDc1tGCBf?O2^j$By*h$2;+)$q$RfZS+u>HVVtdg9;MK1IXqc;$VZERSUb(iUpJs8HC1Sn*6y5#m3wDn`Dwplim4R&(^^Y&<>vM99U7i||vpaX^vpdCo zaEOeBd}0NO2-Z@L$ve;dIDJxHakU`- z>WKfANO*2;CKCsu@B%jq%NwbYq|9croD-zUS&>VWV6m4W3f9sUa>5$PRH_6^apj|p zWDa|nHE0@;-ok?TF*F6U=qtB#Yi@_Z7*Vi|Omi38pe;_W>}+3|?l0vYtKN!=4oLD| zT;~lmGw9zAG!Er17fIAQpYj^YZ0v@>^&Rn_bo~EamQb%B=TzDbulxsJlJJ5 z7R_oOoyxl+E zZ1*wWZRrDz*xVyNr~6qsVpAW?d7nHhwm#FJ>wk86O6Akf^iJN^J)T9{&wTC|UyzY8 z`=2}ok^Rrk1#UwfS{}qlLSB{l>`A-s@$B#JYjc{yrRtw4VxO$MFYaEsBqe*Byi@$d z`jyS;-d>l^^E+R`uFG)oZ zp{h2Db*+pjBHH!hY_%679FtF%Sr4L@O1>oic}FZ!)uFN-Gab)Hov>ulr-E7qh$KqI zGsdMLI^yFUvCQPlhHyWvo*R{!UW6r-MAhqAv0wQ!mt}I*$)-sWuXNdzIyzDOwCAX3XM48( zt(Awwx$Xl2t(bcujmWY=rP|ja#(vk&*jO|6 zhWg&JePw53w!5^MJ>$E6YH7N^Kil4)?`$va%{HG{dg{h(_vVt%f{otZr5SvScysr> zm%->&AV&X+~i{dTa zH%BrGt^+X(`0AQwmg%ec#e05V1xPSv-Q0)a)y2QoVBFOz#YHaQ91kc-DrmNsVFdo#awHNTAvIlKc&nq=k?vfEau2y1w%UH5I` zt=)?}*QKomaV=k*@#@%QMwhr;`!;_kr_tKJs(F8pcvtrwQM{8@ge35R1-}GL8ueHT z33$>UErYq1!Mb=)_g%?k;8_!r!oL9Qz~o$^oq4{!Kil1!ZOkQdkNJ7ThD`l5z?|M7 zz=&oNu24ktkOKO5h@6-$o3Vb@C-C~V$+nyVp_Xk`Vhf%>Qwjw3*8vF%c_+A+Gm~PS z4ft$@Gf;zX(~fywys7&T#W86g1{e!ls>!QDmZ}J85*sx+F_y|$AIhx;^xO(r%@YM9 zQYzj8LH&NpZ@~kTTj6>0@}bnYoK*mpLEW0X3eTGX8bF`byDD4B+?61XfxUEKaw>c; zb?5*qP|YiQQ>Cu~aS2oh1Cux55z0$JSgY*nfVU3BE#$8zKVteTTxpu9%dTe~UXfKG zPJudiU~(j;zXqyLZNts|HLw!IF=WXmPoi70+9q15b8YR0*MYbN>fC|Jo$xvrq$2FQ z>^hgf7Q{7B=MGFhh1aKcG#p*nYHa;{KB^H31>?})vbvkt^9Q0ER!zJ=Gh%S+*C zs{NQ<=Tg27aR$`6Lz83Sb#86#um;%$FLecoLtyV6n!GAxsYqewppF0)CK5$hFk^kF z78=rXD`YiK)2`O-_aHr4AinoeNN)h2LlIvE;R(y>w*qCOkrUR|2)l4nVxmhXe`v>&WCs zOn*fxP#v9`J1(*cz%t~oBa(Yv$0oPJ^X4I=&eeGaJ8$N!0&xoThsGwa!t-W; z2GD2OD&;N!FlwR`u$PWaPKEEK4nbg@t2N86XX$GI?1B->l@=^e5RddP?RlIc| zZUJ0sZ1N+fzw#?kZD9T8ehjY;#$uBrG5xh+8QO2|&^QI4C7V2nZpmtQXr<1zEjhdn zWXV9CJ2ANvUgv^TgngG?=knKrxCZLniOHw%I@d!<%XMz-ijV|=I(K4nEWFMo(&Ndt z*yU~fQULQhbV3;P#N=I}z1dTlC9ZRaZPmFGlXHb4nukl!ze7UjPlI|T0Id(rEl*6o zh1a=5LY=Gg4|c&z`8ps$Xl{98axA>gt*s47P`d}eo~5n;aR?wX6O&hkEETE8oH|!) zF(QlPtOIci)It+_ZiP_ciM*QMh#ZQy62vi(H&0A%h3Cy9Mx6`LV62;+H#1g;^5zwj zSK)axuomdEY?X2s2#FKgOIJ)zh3};fK`qp|^fdr>q3U48l@=^e5ReLNuLh;st zxCQ)m#pFj!e}yZ}lzxe<0&xn&;uVu4G5r;gTSQ+zGF9ZGWn)7WQ3soy%Vfkgfpg+!d2g;dQQul9ub-*cBlO0Cnz) z$+7S{mq-tmZLtg9_%$IZ02uU&$-6>(v!?>|*12rxdDOJL)gc7hn^#TF6^dvcF2nvE z5xF?)Ksp8bxvM7M!t2~ot935r>kwx^KX=vSSa_XVTRW^lcHfq|0>mM(_pX|}DrBih zJ(jS})qIkoKEy4MTdmS_D}-At@NVmXVx@vMNd@ptAXra!fuK{rhMJSUu;StJ9L0GHoy_B~O#4Y5nCO=~O zD^h{#nx2h6IlLmPK%4@xc+KQUOn(hjpPJj|{u)>b;ux^xHIpaNExA(pdGlX@3Z10# zyC1gJgX=?e?wZM+@H!V<4fNe0$L#u-O#@Wtu9^e7oO~B?L40_GvU7@|%Q<-`FT*ozb!8>kR@IrO&n#s9B5zWJ8ynjdR z*^Kp}e(svdx9~c5+-jXm`8s5+K%Kj0axA>gt*s4ekX_zVKMXLY!yjm=Ca(%vDpHRn ztaCNPO`^V#4`#o}d&eg|xkA{*LT^NJ#aju`wD>)UWs_aune&8E>FSuv&YU@`K&*n= zHK`SzIs-(2ZmZd4YnQtc#Ik(Hz;Dndufq3JhY+yVWpAVOPXpYUDD{UBn}@r zd3I$RyCNilF9bxeY_cu94A36EY%;LW9_^{d64$(w#x*Y#)|O57 z6$)z}Hp4z15xzLk;YbM*m%ih`|VI8vT zU+M~g0Kq+Y*<@HDV@1j{r|t#FFnWwIV#fMlmbFaJvXJGl_#06f@m2yv3+B+vCdtU`6~vdOUU92y`Z?6+*~a#w;_2KQ8xUEzDGLkL*+>U@e_0Mpli*o5{} z>*91gOnFHF^$L*DAWX5=2Yr@&Yr3->TzT@#TF%Eqk_032z+_8IpAA%_+Kij~Y+xmT zcW}?O?oFo~v)Uw(d1m&OYU?-rd4NmpD`4GgU7(KFy&yfH|7snx>t6m^0RIB2dj}@7 z!s}iSDWtj=AV@2u&#s3fDG^ZJYhAI9*S*ArJ$YA92<&<|nwIR1T`x90Fd0~AkM*>OjV>Zz*2~cm(TS>k4(e z?yW5j>QJ-Cu6wB~KrBKDrFB1g$XJo`%&B`LjFt0wfJ+_n!7R(VC_R*AA@q3Suhwit zmc?5M;2q4Nt=rP^x_9+Z>R!$&5UWt#Yh9m?YZo8_^jm-sZEyUmZ_Yus>q2$B?zQD!AtS6~cHPTg z3t}6pd##(*@w(ST$`RGQk(7XSuXW`*UiT6g#_}$9eH*_fWC>8+J2IWY&>ron*b>*h ztF6_&)&=aLu;yVi?Bfw(jk6A@~3oLFDOm;m?SQV^$tt-{>y0^AS zScmMom%0KVKro@S?n@6DD^i|0buU1M(PM-WGu8*|UhCrYP-TQHhsEEBEQ_}iAX>2Q zwQf(x>)y3Pse3uAK&(PJv~`_2u3dnLu-~$^%Uua#8QN2=`_=J1)gh#Xx|hBN#3ocK zSQo3~VaiJas8@Cr;(Z?AQb$41XUVs!JM^?t_i|PNcm^Xf`C4^=2vE1pT{W)FYqSd* zv*`rVjah9Ht<=4?Xv6D38iwlLv6->L>t2uyP{+E|XV<;_wIH^ky4SjC9j|*mq>$=f z%{;rW8@nPT0;u0RHkcM(_Yw)h@-BAW8^0zb25^ryI9F(o_7rM~>)y52>R#(Y_E1>! zuo?F8h|QX_4y0A6?zL`W$Lrp;BdL1{tAcf}b;UYf_tq8(>yTadQdfYq2kpVu{plfN zMana$?$t3l!dN-$KYtWWplc>*S#Pf?LM-(bLBMd{j9G8Jerj9 zk054Ej&(kY*|6)rS9AWDxYWH6&pEghz`gigh-ouNb&kn&xcGgiX7^F?NcZ6wcELp; zM*j`OsL7+=6u~Gca2>54*UVlL7rKweG7ByQF%0^y$(XplODTX8I5Dye#4O~y)`ja_ z-wjl#TE7W=H?S1MG-T4&o$CyfRyziz2ySN5;e{ZE{}k|H>-zOK(|x!?2iLvyw$Z<= z{rm0WvF;_Nf0Kw1UjQ6-wQgd6E1Pj%A^fCf^KtR+?mOeyBrOSP|3i=v%*nyGB@u!r z75hBR{d>jZ-FK(s-m@0KzqUAV&$jMv5ANBX>J2Z3WA<$7A^^t%3~(kR3kEn34WO^< zm>v`0I151x|1uQdOg463$iyoV#jDdx+L+%c-q3xJWX$2O0z4VB{cVUjlYu?JVh+?} zh?M@@wFDj!4|m_lmjJXhY_cZn3Yx5vH7Qs&GFi?-fJ_CHd0E%42Qx2(4o`g4EXQPC zyrm$fq1@WKeLav{FEi?5%`iW=<}3p-3u@S$CJbm8paOCt9WwfS4Re=*n1yGxo z9_vs7maJO8{5_Vw2*fC~$6D932LhIthOln=dn|7uh+)uoP3Ms7yGX&Rb(_F=Gs19fq26;O}bX!-RobqT;!z&+Wzn>}c}<+*F_KNHfv&~JNBVA;jclY|1d0zfkr^$b<_}uPVkbLlq5DR|~V!>oXFEcD`?(FR?dG|Kp`zG8^G$*eV{q8+A zPQVX~Z}Ss4IA!kae0hyPPzc&8PHSRLiC1@D=_3aG2*9cM_-Bjf*;l#m?N6^<>mGkp z{=L_=|Br_U4@)ujUf}+0ecAidf%m6F?@veGpN_phop^t`;{EBW_or*U7y7t3<^S=r z|HlLWkB9yrkNiI#`+q#~|9Hj!<5mBU{ag z{9I1_Tu%I4PW)U>{9I1_Tu%I4uK2lJ@pHN2=W@l*<%*xn6+f3NelA!1T(0=JT=8?c z>gRIR&*iG0%T+&@t9~w5{amj4xm@*gx$5U~)z9UcpUX8rmur45*Zf?r`MF&4bGhc{ za?Q`>nxD(HwVwBywNs~npDzPH9{@ie0zV%CKOX}>p8!8!0e-#;{2btS8Q^yr;CC6| zcNyS!8Q^yr;CC6|cNyS!8Q^yr;8!L@UcJkV$p3SIUzrp6e;?piW<~zr2l$nFk^lDr zer0Cl{e77l`F{@ZE0ZJt?*shG^vM7F0KYOp^8Y@-uS}8rzYp*$lO*r&%Ph(NbAVr& zC;5LL;8$i!{@(}qmAR7t_W^!ow&eYNnJ)Q%4)7}zCjajP{K}Nc|N8*HGHLSvKESU` zoBY2I@GBE1@9)dZ$^UbJUzt1ke;?piW>5a#2l$owlmGVteq{#b{e782`F{@ZE0ZYy z?*shGG|K<`0KYPk^8Y@-uS}) zAMf;EEbdw_@Y(gfC#TnEd)?c_t*3A4E}oSCr^dmAI6i%PzSmv2`o6WATZ_KwOWA_{rZx_$E z*Umq=zFe~Q!jiQ+&WMwp?y>&yW5S3=_-j4_Di-ZiJC(r=@e<$nl@QmLmu{9~gzS~D zy7*uR)s?hhElHT&UWig{ZID>S2ReRzo}X{ex1=B6s{W;iX&qJe`ZBb*rcMAll}4Y& zXx}G3&8JiC(q0kgtNZy@LJ2c=S$w*0;fMAfl=9u5bx)T620KJ3Jc{@FHAskrKm7SL zpa1%f==s>VINO|E+3%hx{@FuW?LivwOIT5_O#~9P_|=X-9!Tv;C&~l%lIb872U52# z&o(yZ+gJCN&Q159ob5g~SkHcC>B21EO?wEl{g*oX;9cU5(?V{&&%8AI+X+MR8Qq6k z6_f6$)`udIB4iV??-RZ6XX~e=J;}D|{(S#t4>D=#j@qIU5Ezz+_$5R`gF-{EhBf5t zN1_UPiKC!9#qFnWlS$1&O+WW({j7*_cNox*j}Y;L9Q7y;Nxs-^bC&fTEG1(MS`@Fi$BY zApQhV#;8!n>tJP2rW8>X>A3I~t$=glf$seQ1yI&SDwbzXh@ar-Y)H}B$mr~_nfcF% z632xSojVN7hU61www2lYlN=<#b||ez%I%%pm2pLWf}_Y0MUi8p$irvn%NO9>Vp6CQ zzvPi|*9EFl_TA}>g|WDFq8gdAt(=~Xj4Aq^82uhClj0+YvR4XaELvF-Q3{4BFa7Bl{{V7J=dyIGSFiXnKtnOZqJ{ z_7w|$j3UPYPDO3j@aa0#rIAXfwx9AJ1$C4S)0iERpa<_YOV!@wvXEmqmKeWpzMmt^ zeHEr|h5InK>teMi*Va+I`M!-)f%;me@mKhOA>YsBMQkLMTX50QGWStT;|fQjhjMkg z0X}4q%EcRny&#q1*H|jI5h&QpEeNVz>ci1Qd@ZJR1yJGl9NQtYBF`vKXJ38HULJLE z9bGfro8jv=&;={rLb{HaPw_37-W3*xU-aw|yboNdpA$pIBg(`Q?Nn>AohZYMW>RaOa!Ts^>43T3f-G<~~rCXTXk@6_+jtxzM;$<7D)eNm{UtDBke8hbk8|24m5ZfMh0pV&{!D2v za2LcJ`0g6E{&Otl+8`XPTnmLe8uqgLSYYxL0EgdtU(;(d?`7YJPh1`I+Y{nvS*o>B zH$-;EZ;ys4EIJPq;1-I`svb+1mq$cr1`W8DdC{2!xD8GjEgvOo7}_=;HuO2MZ){g0&u|yy+IDMedHoW(3-{!P8+k9f_sT;H1o8J2+ z(0y5FUTi#>K>@Z=7offP6;2SBPC-$(g|rhRF|4IgcQ!jWD^rucKndkz+S>dgnw^zS=-FVkS+GvPmJT9lAAF#2G`tUP8)1dCW}eP&6tn5T-GS8y zm;OQ7%h$Br`$x?l)(0Odlnw8Ld*t01Z=;Pevirx8r1JXU(pv~NFg%5toiUI@V-o9w z4;2Q7_rZzcWukF3TVQvOqlp!)KDhKDiZxJixkF?VtKWwTp$paTp7z}#_Q99Qk7I{0 z0%jjvx(h|=7CIMEzY}E}%x&%#Sp7a!*jy;%^H9@l-;c=n7&KtPm)QrGzCcm7h07f! zYgiwAs8Bh)4_>>Mo$b$vuhm-CeRfZoK$~Fo!KL|AlsjsLRLpg2qC)%NLxsWNeejy> znt3m~WlUTh^IK*gTv|Itwbq2%(J+NZXN9U8t3Ir>Tw1QxRzPysfr59seYqIf#H zLCjbf+sSHY)_8qz=@by|UO`d0g_9jEb66jIq>wUtG*>0Vq4&XUs|(UV z@N(99aOoHnm0NRcjY4_~{;>1lBZZ^keQ;3X%{-mmC}z`v@8zsMxbzQ-;;A!j?emVF z2OlYv4ex_{bfNm))0t%V&9wd_YFy&_!Tq1v2bbw}LKDu?&M z2My1IYky{Uxd|&{o?Nt!d3|tc{sd1Rp9-(5RlskV6Zl5^;A4fs;eGJhDv;-(m<%m% zWA~GZr18#!OKYd7*23M|q8oGF6DYbmCf(1b#n(75j|j?0&#-<)6e7#!XQC(53Q z#}e4^Cb`xIZ%c;L=?vO1I|V zBCrWh*ViXo0&iITJ|W%b=9P5negAIlg8Y^;i$GreE`5Qb>~YIPW5LTv4()?a6e@@J z!G}lI2PdqIS$j+$T$(>cxfT+~?(V1I3cDA5qA)nT4_>1q%6oO*#%>uCSI4ZJc^+I^ zJ4Lk??shaxq0w0(ZunffmTlv_yft(cX))Hj<;>_z0^BC6s9`nQcsQ(KXxoH{wjt{= zGjHw95ttK;3RL(!v~5B{+a}hHY)8u+hNrJ6v<>frj}rC48CV+|lV;G!!_%cx5Q6xM zg|xLvJC=Q@EIs`@4gYdfA*81#9o6J2rvA9CISIxUtlxMRW3aq7xe$BeMNBROq z*+bge{A!ye0Q6q;71I2gvmV^1Mf>2*H#B%*hGX&eZRrjaWsmr+dxls!+@bq6X?_hL zEXup%Gjk>tsIxKmZ4yqlV%@*ieBZ|V_&9S|AAE&$51TD(RS!k>!L_~{h4ml}VTa)L z!Btvs=WG)ZjM2>Vr#fq1eDG|D<;tyy*f#V14ivg~8!{a3X!NXdJsi98C{y z4Xk_7r4Lc8!NTR9Eo@@v!By4k)Tkr8?Lg=oM$oOvEx`T|AS7A|*`tYPQDR~0IU_rXU;)(0o7jCpcQ zA6%M0MY$*vSFEcyTK;)(0^jI)@KuGu;eGJhDv;-(n2gTb*ezq?>X_d$`{2^rDXO(5 z)Q*NJtPj3Qx}QzUwsBtGQXiZ}Ag>QDZJ(m$Q4DU1^m^;fd(>`qXh2Vkt`gBUgeqp{ zP1AM+TE!xfhqg(Npy<1`)1s&;yYffz^i_qn;eGINqCU8e-E3f;K>@Zu$D9Y3PC-$( zg|rE4-QB>YQJvD!o;@#G%(e@}OL@qQ_Bi*a!m1B#Z&8|dQ1Tr%<5>mEmA!Y4SkGeOy zZ?BQ&*Bnv8?Pyxf)qH1zDIi!s>t6I#5<<2XM#xU+xJTZN@iraj z*?sUxYH%B1^}(gLP;9_D4EpR~66=GnDGUzpgA--%qCvDhfZZUDrU$nMRv%pY5XBm- zDY$0~n^^t6M!L^!e8hcl^7_I3pVO5xHNx?a#zFa zntGjmVw}J?+6P}#7#!XQudM<(2fA0YJH^CBF`G6>2=88WY3&r%uJwev}q|fPRx<;Z1lkIDYFRVVd>KNjhZn*8V8Mt z&q!yN)e;vHeL;vM0&WO3vB`G?CdHxx_3-RcwpaRm!1uGZ6-L$q$eZAZ)Am!W=6x?{}|cT`VYtD9YSlGi8rfb5)ceB8#>5N)&GKxmKu ztW%0mNxEjuE9@x$W~Vf)rBSCg+nj8harA8?XP6}Q0{px)9hZSUoHWtqPy_d+(JD-P zck?|Crhwqz%st!&%@BQ0!Ydn#Xj1}#T%9!32B^&i7jzpNlQyi4v8`5}RE z?Cx?Tso6c*ob--{%PpTmZBOfnf2ur|&2LAIK_PT_qnx-j6RTr4j-$z-sYLuP%OIp5 zQBb>OYf$c4!m&W>`eo9cZ<8dw@(_AwwNN7}Ur!6w*Ky8nhGGPQWg2$3jGi@Zd}^F5 zf4r_=CSCjHU94FS5K-WJ)CtjmIsDz(F&2)#X{g<@h1wk@bI+jRUSW24$9!^R9dp9U zSPHB<=Kq|hUm8xgjN){~%4T zF^I-ekfL3#69F7^Usfm_-Zuv*i`u(e@^dDMTtBBFaLd+_Q?{QoIl%oKt*-~9JJ}q1 zqWb2c`r4X-W7pSY3c2+)4QLy%&YZT$9h$feNY}D?MH%Ja@rfIY2HgC4iJSBXN=zRt zql@Z@%kRLjILG_u1Jdl9vmo5HrWIoBt}%mx@Mg!{!=+1*^xb6kEiHd&u09~mz5xV9 z`FQiatwFn7`x$c&m!3h=x`m=0DW7=Xd_cO-%~pD_wM+AKcFD@3f#B=BzPWS`lGd%m zrAN*o+BY94lnw8jgF=t;cfF^x`^#)P&@GnPHw3~Pab9wo zCrHV_sxe2p~L&;HAzGMLo4Fg4QAr%nD=t~=F;LR(Ryf2w;c^zXqX<7?tjzb zO_!-hL^7mP+GgKe+CNFr794k!%%ND^kch=0qGsmeIuVNit5{UvzRf#tF1>=J@@80E zLiW(U`A}hTc;9?AQQurEoQ=6ND8M#tbsn2{R!uqvN#hm{cf|bRey&hByl)N&JZkS| zyEk)j)V*Udcj*}7(k@AVpr4LDWxH)o<>YKC6YVxvS*EZVb&7}`%K--XY9<@C| z;B)Ci(lu^gsYm&@4(sgPhD8JF=FGWt=?^6Dws5qgWDeiMNwaT`s1Lnw&Y&Rd=*+&k zbP1BaBWB-{vSs2IzK4@$-vG*@e7vREHztY9zWI=bq75w+t@$3_U_Jovn~zA>zu6L1 zRYc@@bH@Bzqu3avl6l^INW<1f7Pi(d{e)4 zpnda^!tC(A`D(+yxlZ)h{bj<+Sms+a(3yR6=>jDEqBvcxbVVDLf9p7fvtLI8&!VGZB#tX+qYU z>P+}m+CNFr);L?c^rHrapF_5Hda_1-Zx)6^1eBP0&K62>YGc)plE!|I*ZEgtTQBl`}tU*aCqMwSgzUL&30ub z6{trr{aktmN$b|^+tIRz*4JaweQu7pqx$AL6K7YTmhg93to^Jd?qOOW(! z;b+ZvXjsAo{?NYpm^Awa2)vn(H{ax%G?sZYuF^9|T5mG@#uPFh#L#=%$E5q;Y6`0_5K;Mh z+NhD8WoBJZ!|o>54Be4(ht~BI(mirsWk=-z5m}K(&(PACrf#T;A^d330 zI{zLymh|zy`9xuMc;9^O$ol4lm9gADs&6h`fTUjwr(@3pq~Qy_r+uOjI=pXQBLs3J zbl;}s@$6IEMADe|v`dR8X?KXl4(mG8*4V%f z`+G~5XS@4Lv+XN88?)V|&FRhAu6)!zQRDpG;>7xu&FS7=_kwt|`^e_bmD6}uz*!mF zr0X}C^G-2H*C54n3yC{oPQ}-tT30~a`6#VN?cciSdapKt$Hb-Xg?JNy$iSVQ*KU#S zK~lU0!yPeyUykZsp>XGzdXH+Kzf&vvQSnIk;TT0j>vBAu8Qp1s+r%1#J5oNyM^Jq$ zpzTc&`i9&K&C<1(KdzO3NnGeY8moMW2;9ec(8`1csZFSmTFa~pbU$BFI2w09hg-Zf zeJ9xNU=qUa&g`yCXlUAsg{B=fgXn&~q5w7SevWsrQO(op@dTS8P9yW$0qGu;{d~n5 zhHIS#JeGYe7T;GCvc`z-s<3GP2zs+H7$j5tChO%q=^-S=hu=B1V41dA{CZUP3VM4p zeQ&P_71~-+uh;wgWxc<@T|Cyk#N6LWgy4R`8>NtrLQ=mqCU@i8+rLRIPpCFgZo*{S}4o1?zoJyVKX!rtJ+e1u>C;U?p%j z^K`@t4cS|St~LCKc)0sUzJ{UYF&{p`ymnxPhS{xH7s(wFUue8uRcKu>UMm-(B+Ilt zIwrI+h`@Et!)I4$VBM+()*UH#C{njdMd}b?yZJ6n=P^hYiws=ryw00+4azRP+7PKr z$RCQ!uPUT2fXgo@>a%N{h7VgaNWgYabwZujXP4eVQo1!2chn5x-mWmZ04@)R9NI|R z@^&T}Nxhwh&#hYc96y>R+V$=47X|x z&b3H9A&}vDoI>9Md7M|6QC?0%-1un_i$dnrf6^&P!QC31YncH&g5mo)ZEntyC*0&l zeLUrU4wFD|1Ll5CL)TUz0)wZ)kbr3 zOflp9p?&sMg|ac?yDBUq`|Mil%_3ls2!hA+`s}MT| zK%X5Xu9?^KyU=VxaKGU7*`=e9_IT7;E zXuZFt(7j;2?`e6mdvT$9-#YahQ!*1tWY+uAZ%C>S&)*#qduYACrf|JrKIjoX zA7l`L`2e@iE}ep;atqTtQtr?``FFd7gwH zoj>4x_BDmy1^VnYlA3ukzaLFp9`j*tpIw?gNx>FocSL;Q@misE!FY|d1MA1gX?R^s zL}L(v>zdbRmrg*^wFTB4DR(GRw?;+k5MeX3bB)==AayJKqpZ!Hy}c!N^{S)fMC<02qTjuzrW^3% z{g;Tl){AJlzW3zx`fRUzr?~y}ZQX^Fi`~VO&zundn5PaBP9^ydF03O|Kaw@^W2goc z^7QhW1|XZIofD;(~arY`qt7z+w=YTbaVcGxv|Kx7@$8y|>?A(?VVnuj;-c7cy;){$CfjKI<-=T151*AbzkTUPoKf!&c5k0%l%VGIPfnakA4r);})Us4&s# zzsNDF<6`0TG4Ue{Vl|e5?(qlbo1494dhIn7W&dwDOg=~4dHVM5@snTtt(`oInPAeu zc%Cc7e0yW|^m=I`*~G@@(4^QEk4<;y)61K)#|Hi9h}*90Z0}F!+q2#7f>hj##4Wk4 zcWYEz_Sm6t`8;vE{Lb!dV|{zNC9PO~I-cL`gxsrW+%4U=w>Q7K-91tKvquz4FWLl& z!nmrZx{4{eGnJ04~PNVqxilzY*KV5_q6n3{D2;?50;_=_Cfs4dq5G3 zHtC1Ot>-0FGhBMQOex(k^TZQRyxvZ3;k(E-yD0zBE5F?1>}@gBk1*>0*(W~n zS{n6N#9zy(A1&?e@6I=58nm}LzcQ2mHNA4Jlnz{GWl~Y)!JTJrSLH$Hbf=P1FN&FX zF}TS3a=DEEQyrTRtigcmaj_FieKE zZ1=IjdiEDQ!x1A|4dCOgA>9 zi!9!^vC%8yO!1Eqg>Frgig;raGb;NtWEgH(eBTd#{_BN`c&cTYb11(|jCjT7z>>m- z=v{p3hAh&sAmxu9)j)&*6dmz|_;h%cgi#7#M~5jtKh;o#^78`=;srj28pSC@5j|9V z=>IzaLsKOpDRW+wznwfZ)q|ywor9A~d`}-mtn%uP;`FRw^`JOWFxgfA>}jT1E~)~| z!~_^OcI$4SlCZ8L3UNHmYG~BHy$)8tJ7f+M(f$dnUB=%Yk%k3Y&&lgH0g7_ zUfa`H47~gfJ`%A$uX1)$m(fcHRi$Tw7sD0v=uWB1>M!{JILh^Er=n9 z%R1P~5|N9w^NbY;*C8g|&BN-RHC2m{*N-i;Bv^xBSrSa>5+SK^Hv=T{Y%r;9(|WGw^(+u65RC;QLTZTDs{IuF}OTQU- z`e;Y6891FcmWWuJfyXS`jOquGI}LX;aQbm93HBSiL}(8%4xnJ9H4HtVcjk5r7)ed} z<}i}{a*w}{Y66iFOqI`_7bmmckbE@^X)e;+b6es>@woy4(cfltO01E^kRrqb1q>pQ z{R{@l`weS7MM02vEhSM3Vc7^F670(R4f{$6YW}nkg8Vj1ll8tsP!>;?_^-rZN^q3k zpIVL(*PyYuGKf{YNr_7x2`m1q?ek!5Z%pwX?SXTeE0MMW`We~~mQz=Ue zLA4;*;0F|bx?=ECx1`OW(Q0J^bhb+TRG!hv*(ylu6spwIDyVs|c5;=@oVx&Q#2npL~3`^iM6RHML2R%iF z4iMo_$-^Hxv8sz|KvWCe3`+C!3J6vDK?b2Rcvb|Df^=$*#827y<3J;IpG^{Ufu;`J zT*H87{Z$QZy;{7o+owZYu}OwJ#Z@B1gHuDz5hr)pDm-9}GS)D?*h<3RD2hLYuuj_K zFFHw-&kjCZx4!E;d-BA<&UO=w^|jg%hhn+AbjRev^qBS8w%I0sFBl<#|v0h4jMV}9yZDfs9MQXgDYjO zUq3+5kY5LZmDZI{Fm>-ZEi(2Lk8%#pDP>?D>yRo#JiSKt3B+fTWWh1kW+E0p(-G>G zX>+Cu5nT;SWr!}B69mPTU57)D0Uo5I4yctdY9moYAZp9;-p5|Lc;5^_WyO0BBkXwZ z`9@vWjCk++S^=HvLZP9>v>xxPxT!z7R?Ens_a)y3T*FPP8K&Xpqes>UGmlI#{}=P$M@ zGz^CIQUZnO+#nq--BvuP!NG!uY;ur}+gX@an0z&bCjsUhMLM5BH)92kS=d(iL!}(h zvAGNjt%fBC-70mduAnL8AivGfrGiU?q-$-9xyDN+{80>*wk7vcT&o5!bxof>Zu+n? z8#S?4h)aH31E%PmfHk)kZxAOk6l_`ES7EGTMOg)U14{i%_bJSDpfFSQT}F(_f<#Wt zvVlzjqwyTIbl_C+M#Y;tkWfIngPvZ4Lz*>tI;bejLN8X|=$%}yZCh1Pv@HD5N_^adcKRh$L$YWR(pv(l z3FTcFm6f16ou}F?oenj}Sa~0iJ(BEnRqVk79>>#~(u2uP=im$^lRBU|ok!oT_!YEW zU`9oDItNamNvOHg9pFx!?p607_drD$OZ8)%?2lOK)`__<1WoM>WJ^6H1 z$-cT#KCgZU^4m1UDVzW!SZh^mXox`06R;rys*p?hpf_aD(E)k)Wd`gF)F?_JpX@}x zIw`wOrK?oFI%HUURTg-~*HLFRB7BX)1cnt%U_|`WQZ%{!jSU>&dX7`}L#MOlxL{$< zYF@B_1BMC?kZWkNQSv0JpzGa{K?0PgsA9po-y5?vp$3Nr3dYQ9U-NY$LF+~3D;1Sh zzD^Zd3t(AJrHWoEl`>&7HPouEp*{nzJc?2`plg{@RJ*2D#kj0TH(oA2yZbWl=th+} zAmOak>*mxbX=mF~8XclW;qLPmAF(+SFe@Z%b|h|B9v&sJQ&zL1Rw$0jO#Vs7y{Xcu z_3CtMD_q19Yn16>xb&Gj>hQ-kQyPQo>2oLnQH< zFFWG{w5v?-jLR3`@)X_fDretzFTC<w}9iAs%^*h+E#i;?lP;yZ>idLX!mt>ZA&dZ-&@qSvOf0SD^w3*rnnuR zB6`9PS;g(LxwOW)QmD36QQeM{>}OW3NV#IDMv3@In*(#0hql3$_3b2nQwQo)+)R+d zzsis~%R^(RtZ-LStERyg|EC@Cjsh&%zOu6+pTF3g-kj~OpIVyk%a<>*BQ8sOv&|=# zp1LvHy}5L4w!1ys1RJ;YY1lFcm(womS2m}6d)<4*>$~12qMDX&&sch?)q%~;)ZfTTsWNBmTKu`5g<#uz1(QLh zE)+u^)Fvfs^-?gKb*r4HQDoR{6!C}p)+;vmb^Jz67_h(7OXmkz&DT#UYd_=@Gy(CS zI^{7qwwUVI&Buz{}2tqSry>4_XxsP-RjQf9ufj}wg#J-1BOOM*Mr#2%wt;# zH!P^Z7BnbqLv65`IjCx^xCypdVs|OUS%R-aYpjS4;W?nJc&1aJuS0vTDEoMMNG&c~ zU`R)ElxW;Y9a0+F<3>Z?xRHhR1npFxN3l;S4b>GsLu=+J{3QiG4z0l>x@w)%OGdT! zDSLM^q%o$#i(?vPTnhhrAyybU2jSC<* zTZ1E~H=9YNNgd5qhn5D`eQRNHNnW?)aM{p0#7XDXJ+ySQIQcUy4{?^??v2@iN(TIL zw5Col4eBSDB*4fFSGO-4|>%ko%E$Vcf)6@ps z;v|fB4Rv0YOlsMU?omqh6#A)yM5^Jq*dT#GNg=DMUow_TUmgs}30~1b!U2z>O!y!! zNO+q&=OM9R)MM&qM(~QK!>ZYM9SJ zLS#8qWg&qmhiZavq`>Uz=``w)>`sAd$WWJ|8tPpoa`J8l`bNt{t=cz1FPj@bUd>Bt z3os-35SauVRsFak#@%5={V4v>MlsB3yr`X2VHkr3r`uBHC;+@sW!D_ui(>hkPn zQrR4#_=^QviRpY0e44N=6X~T@gWhTr+Lbf-7w(9N|42m190gQRvWhVUQP&JxNmT_q zboJ}N7&J-oJ4C$9Ss@B9bG||=g`Gd<+I<_Asl?UmlPj)p$(&`dQHv|~FRr#2%Y*ndtp=JYsv0IggJvIIpwN-gOS2vNkt;+Y| zV)QSa{3CWK?ZDAnRTZlxZV#37hv9v+%`NRf#BNn_sI3zr;o9t&ii2()F^4H^L0Bc~IkJZ0plZO);T>F8{C5h*vWMcbZ_{AEx=iXvuERTTiON}{ zZ7$h#5LeCkjqJ&`QGTb^mCgC}^~*QcuWs&Ko^Gyh@9b_Z`F>_U8y6;DhT0s5vJYR6 z{)oO3p7AEA?uRsg1HRwW#Dy!MQ;$w82}Kl1to*jp?p1 zd`o+THOl3hkOSa?K)4$48gPRAe-|+=}+1C+sL|$ zB9hl^+E2W%81NgZ!8rYG8Vqd3$$Y#oC=bnPc|rL=O%dy3=kkny!5_y})Og=Wb?dD0 zKH5F!YRqoG0K%o-$v}8b0*ZK+EB&^G7R!Oz^0F`z8g7;^(C}3^wXu$_$3oj?mFPjQ ze{VBs&hV2kVzay&{!)H}&oQQ|768;cY@VXRvpMx2#dC>^L+y^R=d}Tu*EMsH;lvM6 zabkP+%eg_U#1|ga?U>hy)7@2aJ0>H|S|HId-@obzQh%(rdIhBPppL1Sbc#x6@}^o( zqP@jWF3=G(OVV{%GDcp#G&ZDnn|>+*`vFNzjBY^gh)?PJ6!9)=&5G*Bl4B&BUL+Xv zk>S8ovX|kYlIvhgCsH{&;3wk3K@|w1bJSYkAdT$Bg#!ori7?uOm~h~+A&)Djb_WRb z!7Md095`rCjAYaEk8)aq0qAI(1LlAwtt?*hD%Kut8)c(yt1&4^UcuHE#qee$} z)Q#yQdq!=P&8X2w^r+KyYAP$c^6S3xVBSdGt2UzEtESXAjv6_0M)$HBG*jc)9y!at z9iv8t=wRB@Hg8=YFNbv4U=CH}rL?RA%uR3{HJ~@P&o7U&^UEcU!?^t|gCDYk(ApHn zrwRyXY#lnpWT^-AR&DfVNd41NG&uv z8OKD0Rdbdyf>i}RNVmp7gf!eoWl+=$j}E4rH)e-@d!J`f&gh*)F78_2xv_u!#{PPK z$Jk!?q&RW26uvIsWR;ldo0$ol;xdZXRJZ3lv>hY<>(bDNE-QZsQX69qOvYA0s z1}VkhRxZvaC#bjgxn}KQvDs$k`C#2xXt(!~Xf1M!He9OR+xtAsRW`Gf2Gq{ZmF4WU zDR=VieID);ZDz#V`+NjAt-^M!%HQeNMEQ9FRA1GdqaB?Zu;kaJHm1QVubV z(K$Fe;`>ZTVTJY}T=sF=6>lyxYO*h^))E-j99aWxZsez9ziP5?Y!9@Jvw=1`_G`gN zK|adCt)YEJd#r9|8(Rl!!(S?KO>{oZL3=^)POH`)P#b3hYOkRHPp)be6kIAUY_-;K zpFG9L#)DaED23O2j4PV}6l%C{T#j+&czX7vPzegJh1oYwtA_isy8sO`JXXlCb^beS zBjs@4S_K~FnpuvxCOX6|H|*kbHb%UrFyibGcaSL~r$>d@-@P&0m>;amR~~P@Fq>|k zotI#lI?6p{4MR$^$fn=M=q(n`HGQB@sq6d30zXFzihis86BJ9<;??i74$S>-qzn9P z0(y#mH_Ncb{9JGCyy_&iP*_aXNg2Zf7Hvlob@Y}VmhvzI3w^9^Q+tXoI{ie6L0Wy z7k4(Dgfij>0(V0-5sG|+pUY&5AaRgBNF3lnqAZEgp*lKLq;BvV+e39@^2Kj;p(>mR zi|SfnB}6o_=Hr?`M4wk+UK4vhF8g-x6~2R>zMsj8M^lF@p_~u2&-IPft!)$Pt!+w^ z6KmoPYQVU1aS$V+G&!*+-m-7g&V1H2S%-x%m5L|CZ3PaL)r=)1#1cF9d94#|F;SK1Zk~mwO!<;Bw zI|n$d&+A`AKPRp$d_#H)`2?foAKAW9pc@nGj5lV%;@53p!`%HwI=pCBRp?FfeZ1?f6*bayvOSU7pU2%34Y^Bk%&XQ?^*xPRnIWm0jjJO(Xl_oMs7P;D=IE zaRi|nn8L-Mb;P>@-GyTGZfRq-y%#+?Xr1v{#OC?>5;awqPo5F?c3(#siG}G4Oc{#b z>4>+aWWF+Sk(LuTO>kx_KoI=P7sZSKVi`idj2jOe+7aj4s zc(K>2>UgmeELAtHlZMeo+nooCtIvy-*j!4Y(~TE4+3tMIx3=9iyC1O0cIRQU>2~M& zWZi7yw>#g*bG7KrCTZk}8#QdQ-BrwMN^WbAUWL?jFQ%cOE-S zD8MJhwuo>NyT+sQvGNp3Z z?Ej_iU4JFNj=Ep>9%tfbDawm}!>IoN6z#<9-S_UXlUh*G3;vYbOgun=e+Y=KA1oR|uFc8U{wYt}()~!~p`c;qfyn#%V{p(-V zs&%Q)?_0I@gJZ5t^3a(TD6l+zwK|9pS6$+ET6so@RxJEBjS32od|bqUujtdm*A8e*@~d zey|pFDUbgN5PRJz!2#`Ae`0K@{VdP?0kGB%vqbY5wH>dSM3Z|yQvV694ZRH5rx}az z^#xWKrI$>A`I5xQ_R#BqZM)_9df|Lrw%3?`drj2I{%buEO%f(DB$&|(27W|Hny6b$ zcC8)Y4I+>#tUHo)RGRdvzdz}tc);cqopOEGti-MjdqJe=4KyN9tck zs#SYaH8i26Tu-ClyTisU=sWXSjJb3aFo8p&axW+)SmZAqF_emq!L0~Oq{f`8#7e~F z62;SOQ5&#yWEhGW(Cxa{nthLF>8S0vt$|G-)PSWUBTrmUiv|~Jzgxl`oLK|mUo<`F zS)!p*O6N@`iC#nDpLuq(BC--g=`|LZQQj6H;>+<$29&CJ<^Q&}KoQ^Ci1<3_2DO2i z6)`Sg{o>kowp)X=pJYFf8RI;qM;go?-Z>n#&_J^?#&&|*>77Yf)^|@Dm1ensvr^6$o~y#NGL;P?y!!9Eb`m| z03f-OZ;gb|O{UkQX6}w8jf>@DGMRH6GaAm;C9&4LFW-K2@!)lHWz;t=u5R1gHS&X| zOJZL4EbkTd()fiAmRI~CAqJu`*E2NKQ+$hvgG4(#7FTBc(_}{4Jt_GpBs3qMsFw!V zp7M|(er(?M*cWrB(%uw7B7Pe*B1WQwuNZadT@FJ$FYYQd@cnUU*fN8#64{ZL;w8hD znV6YKTi9MHVoeY_!1>CihAlH0I~fq{yCKU=A?EUZ9H>^MhpB%F8o114EMtMYmE%y`SrZ%xBb-|lN3s$vK1zK*pF zC>@nO%~W7aWV#l4I(g9m)1B-DFC&$;j>TR6Wu#v34d*v+UtYGsOLX8tm!7Zs5Pns? zJbsZC!tk9iMq_RSAb!0F>j!+e9C{HqyPZ*~pPi_e2Mu@92xpKhC&H1Juah7z68APE zG5YR_dU25V6M@Pg@1Kf$1fAsWyr2fIIE8%Jot{3n6$iyNLsy)b!k|9RDTwqSQxq!^ zwi&wOWE^Hl%GuXo zT+G8K!x#79B5a~N9DmuT^$^Cbgql-0Rxpw7iSyNAat&_vMNWiA&qU+2xmnI=u$ zXWSdBbJohMGz{&jGayGja^JaZS56 zQ0=mb)GpK)b8!Jnr#4pYvZ3-OPh=L5$mokfvDp-2)2dyTpkOKwt8OA=?$HE9W^o`g z=0}LH1=TLggvgl7PxteQ+SkmgT{dO43so27nv!cGZHUZG-l;YqnGH2e^angoDHilz zPuX|eeM)DU45;8_Ge%ngN4L$+Z{WtMm~knj#&Aus3zh>Z_*P(zHj zTt=@5{pnktUIGj!?5~~r=}3LOIIHg5ySjLAsV{K2eRVOc!xwT%C%waL>vgb(UY~m6 z379NgQ{S(tfLgzj>wn^lK}mZYdGfd3rgC?v`sk$6Qo8lP+&tUjFic^IxSu ze)Grk*4NtkN!^A&zX>i{<=vrG!g<52+U2uQ)6xX(@!Wh2SCEWmtwRN%lD1iLXisW> zp`E$X?Fj37$=<3;Mn_$_P$>`JaMvg}-wCPoY)`EX zROxLZm0s7|xfp?ku^TJZ+jLxLctiP~TFZzI1N7J)7>9_3ZuYc2wU)NV(-uT7Eh45| z?HQIy1$%0(Qci5I_FQ)PUR!&qi@7)G^)~dLTACUk9yh+Gg_aZl?8&^7beUwSXPvbD zw9HkrhQtR2qVCOhc%zQ(NZEc`Tl@Vl^wV0ddLHwRRW<6DG7O#LgTvcz&wGCM2M{TT z4ol_RFRG^=i_TXSj@q0^`QZq_RPsd{)z?$HQ%C^kp){6tR&m4}uaA{F(bZ6S)1|S^ zp;DG*vjgQ|ag;K!rc-PhScn<*4uoU;C`OIo{O8iG^4M zKY02J>J$6BM#i6h=~GW#8$UTdd1`d?BzQR&6ZbAQ;rqy`|T49HjS8|p_RMc#5iAV(xT`mBQeS_vUY*_{G} ztmOEHdjEuuWr3|J`{wv-A`{rBQr;@#g4Wp8kh{4ge8ZW!3GU{2R(=LyUSRGt1uk^l zoo4s3?KCA*BeET>7Bz{QOq|!Qqfoq^reGR>rzt!v4GGRpQ+n)yXk%e5J&6j}O)z(w z7-Sa^v{a@=62Fm%5EH?)^=;Owv2I1i8SgA(1}k|g?<>l z>Wmlk>XYtj@vT+5KDD0&=H?S$E#4aquv&bpf$HIjVRb;g=(WkKh(4p$;&CwTKb6Q< zi<_Iq4H32Ad1b*f^+MqrQMLFwWzM!*d`YXfzTv9HxAt>d(uoh;pMg61p5>}~?i#K- z?MhFABzcQ!p~1-f1w@UZTifS)e%IpEbp@wl_U4-=AgRxLdbL#Z@wA&>nXD&p)KCJ? z*hL*hU8657bEvB^RLqaDsi6R#VXA|tc=fl8sho3``eE%BT`OiEba5~yD}Kqx<3_E< zi+ViPU!*1|*6OV%uBC$FKOR*SxAo_@c&UU5`7wabN`F`@BJMHCX0PQtdwU*~Wv*W= zE_+>%OmtWh4$s*r^0nc{_}WwNU03?j8Qo&?!PVKtz2Ch5#>IoPyQcHxt@E?GF%cY< zc<(GX)sYjafe-H*C#X%+3N%*z!HHURTASk0UT;;Z>+0^!r)IvFgpBc2VgSmxwCQ2> zHzQSU6N6ZnFwtIjhI6Qd;cXN9H<1EuNY@Yd*=@K9RjF@jqd=v;{hLV1&GV zGj@N)s!k+^sJP8AHjxZ>_PSGcJ~XGFXPrR7O8QKcFMj(@neSTG2^FkcqhMui#)QD| zZEYOvL zq62nk*luie0KC_>(~EG|5QDDJYj*7-4Ck6*p(~+24OWD^iu*LI4=mLDNLUox_CXQu zCIO-=+pJ*)u7GJj@pbh-F7@}D={efJhOWYWs6J_*cbHM(>653)86a7Kdl);t&hM4} zE35Pm3f#@ay-kmozu;`JAGLx8cQx->^}c0Uv0hvZEAF%3y*ap@sXdzS1E8v#;J99s zpP_o*lHyJ9)-#@|P1V`>=`N;L;mu7m^1S;A^>@&+(3+$31Y$Aq>Rj;Eg0F_3)h=T@ zLr;!QJ1ryD3tx7&1pI#Z`_}kybBL8rd!KJdC6vo$qFR^Dzu-f3r558EF+{)5y_30+CQD6bv|DCJu=h^p7_1DKW+rOW zdg!6Ump9f_y^}dZDP5ts>zmAl_>ha`+Mb0)cws5&p3E6N>GL4EvIu%d>5Ran<0?Rv%?uWCT(j%mkdP(YMz% zfcq5GR;4=sP3(W5HZ1fbiE5tcu*8p~6+Kxe7}ZnPNT#p^^##y}X@y8XC!zY>I}5Ge zr!$f-xrniPBkL?vzwaet)OU07;0k#UGic%B0o4*+_cKO$k=Q1pBcNy(9SIGQ_E9N& zUF*u0&03Mz#)1)ZU8_Sf=+YO7#I`9DvDdY(XkxtUy86}v1Iu_5l^B3gsH^VS%wGdC zllgy^nCa~GVeSP`tG#d3AHu{q^fCNJ+E3bqf17V`AWz)-1nW#aJ^pM*f>lYgQQzq{ zeO?r{3?;-}yQRzX#V9jQ%7391Hz^_;ob|< z1*^KDtSykesW#(PGxUm&7s~H7f-1{BG1HG|Jo!O!?iwf2*N2I2jJ`Dy^0p@5Lu%P^CTCIR}f)NXJkEZ1Y*dY3$R)8Dma*KX$ zKuz_W!=~T^Iz8euo}e5znA!$V2Zx!b2W~)pHy0B+0ADn9SO;#fwf|vyP7E#24Ym4^f{;AH4`(D#m>PLP|8O!j3OaM^ zXALlYqJ<66D2T9zNTV=?8U^yZxuCE`95yJNMq%2kQP7Tj@>kz}aCNwVa!t)>Eo)c{ z6>sP+8v96(iZ^EG2i9PI1l`yDJb)U7*)(AdYg@Om26Y=~6gZ8-w5U;-QST$Qoow{` z>{-Kk&mj`R2CB3|e;E!RzC#BTR2z{bKbAJ&s`lr6&bCmq@p*5~#vl-;=2D#AF@R7DH-i3i| z1L;{UL&}ZKBx&kxXHz6SoAyc1G)Zblsat=QUS9U7%=X~0{LksnrkkE;`7s6}wA^fZ z;aV8uNQE{vDm3O8TRz<3pwRF~BR2*l1_-qIt3IG+K$VtzFjJpZJ4Sug29!}IX%*Sz z@TtT+-K)}cxVT_wkF^h%sKjyVq-mNuH*NI4=C9t=euc~!+%DIAm}wQZ4`11k)XWhPm6O5W}E%xx8 zZ`?AEF*SfZKR64(w}JNPYDrPR(V>S-iSVN%VR8Gcg2bCIk~0CocJ(C~R$8@H>+zBW zIMx3LEe*ww%rlLn@Ay@ma_8t8f^RxDwmZ~FGVOLSet`~{oZaj*vFTY^Y5`t}V%SKf z+C7MP@o4)eFqL1NzM)_D{&}Q2mI`2EghbN>EwA>Wzfvig6!V(;>ysWX2kcFSfjqAj zy8*WcJO9Xoe+kaMa)o5>RLqycqFur>F zWI!y$LV}Dc@aPyxXFxjf(HT{wz?&Hb-hjoI=9XX#P|^*-N3Pq^EK<_VjFOICxi~{_ zvIB{dQ;tQIi?g|v6V5;yk41NI0uaaL$b$}z{D?;4xS0{hu||}uDAWmDclDz7y`M8Q zrwwnBZeWteI(8aZZJOmorb)W%dej{kQT5_%>EeQ_J9E{GIuVC-2)!B)M5hBovek?0 zsCscWvtWpNHQx0|Q8-VTkgZ-^)9XYnD6t4U?s_z%_8WWK5)Zuu8sbs9+YCE&0#V@% zbT>!i_pcxT2;Fpki4N{fMqxn*p7+1#fyG)a$Ma5LVSc83K1bJh^jT9;~AM73O8BiI;??LHAgV8 zB4u=>$sh>~<@}K!F81=mk4|WeDsyno6s#WOSRpcz9fok^kn=G|iSZ$kpk)IP5wwDW zLe8u-!8N^53L3$f(Y3Mg2Sk>ZS`R3YNQ%2v@rZ7^h7yTx_Gg8y zim69`EL+^ptlF%tiidSGf5~q|JrcngOHG$CU-h1o2+krooq0SB1*U$v+#dvjbJ|A^ zr-+-OinvS&Sk-emf#56>#F?k}5MJ~)thCz43IsP%)%Z-`&~@PE+m2(6XT5V&-bQD= zeuU?dzRf(IH^M3GZI3XJ2WlZpKk)@s_w65Z;WP*OFH-iLpEmZ`*U1CtPXDzZQ}gtI zeY;-J1xr*-K0mj?GVhWax=R7|rx^f%H(lyhOMy2+b>;R_z)f6NK0i=k^CLZL1C$3Y z5){0##pN6DL4XtT>htUwKM^+vldL z*a`Yipp~K3059D>bfDN!6v>&&2k0H84(S!FuAt4jGmGm%z_QtrbG<}94vT8!yuD88 z^!!ySNruX?su)oB`>>+aXVuN|hWw4k#4Pv$At~UGMvC+eE&25bV+w10q#Tqcn(F-% zdZGe6oq|h_za}yE&RDODBAi%TmSV7)WO^hqN|Q#m$W>j^WXi0bzj=cn7L&qvIbZ@S8{4wz<)^ zbqM^Z$~-IZ+eEc~bJrSfPHhe6#}jDPavnj4K7KqSt=imZ)fnSxi4Qi6KK{Yw1X^Or zfIU>tdzXP?>B`5T!gU;3d8n87&7)fBxv7=*@;>Ujxu|d@l83pqa)^hj0Q2&`MMQT% zIV3hDuI_Uqd#T~Z)*>D*HJmLh*kN62I9sMn$6jib_+b zdcX{`D0P|7%@uqPtR}ENOm5%BXz<3)ZK!j8rdPcn+=AQsEEtsV*8SWjfa6 z$oXax$Q`-WQpwr`p>f^qPjaHpYPv=7H?sl{N!KUG*yQF)m$VO6sPi|olHt*2CTT8V z)2a$UsC@(Tur8Q5Y%?o7Fbx)3-$2s49?llWm{L34WI^4z_r}G$x2_(2>tcLzeyzP{ z4%p0MvRlxrOU7Ed4GAgM`I}kBdYP*$v2P$Lwy=#vaP6*TChZlS+61>`t;}Xd(C-~( zZU%&kD(EumTU|hXPq}#{p_`L?%H=*zk#KuF9p77^IygEPcCx@lB!e?g@8QDB_Oj>4 zSm{N|G+x_8)!%d12yOvxmh5l`khSHTB%KiZFCg@FL? zb$Adczj6;|V7NL6A zh68U8UFAr91K}H3=Dy00Ljclp`~Sfos(fQ_TMGLI>K>g=EXO=TIslK?)ytj6D-u9L zj7D~cZURiZ9gW8e?B(p{pL=yePwE?Z2rj5Et4amAdd}=8gv5f>mSEZMM@UpyE-2@u zH2);G1k3DAAt*^9n>~eA>zw^r0f72OJ#x+}b5QKVRlZb@$SuKgBa>8-=i#vT$~gve z0lB$oYHxvGIk<`a9=nh|`+f2XVuqA87){h9HAEC~VDwcEl|Zl;i3~3ByvAE33h^Wu zD*+1~hW{$kCyZ zAJIr7w=fzxQYm&iqFiC2KHm4kB5G4T?=(xV7bR!H?{+CtY9P5M$ZKsD?w=F0s#1<0egoO)>z!PC1Rl@1|2I>hzE`|iN#GE!A z@1S-(gna|oq0bzt4l?|-1s&^CZmnZ@R0U;V?$6Kc0k1@T1J|L@leNoQzOg+{IbiUC zy+D`YirC!FiqG@~EY@hv^kbApX0KBAh5E0N`i+JP_WnCp-+Jfj?2U_i57|OPtRd5v zy#V$l^}_f$iU79$NZ2U70F_vJ)# zUX)i@LfmERe-I7IhNAjt)VwLWzO<%22l;N7wTj?jM@9#2cLkb~s>|H@Tyj#$i4liv zcO~PGw%zRtC1j6hyDOMEbh|4&G7TI4c2|1%eprsPnel{)L$^n`?xl?%IDlc+g*m0`ED28?s{a89?LFKRbR%EtC_^_j%?-n zGLrR~$4{Q)pXc^^8@ETd<0_a1)OpcKxWc{)+sd9cJF((266Yy0p#l@s`gEK;>q~3BP0e0_aaS2H$f%;cV!V1OF3{Eb7j(PRKXw z-_o}5jo9|kFM(~l zqx@j`l(fBaxNR@O4e075-NLeK@c=&(;ne~!m9!_# z6NlWbtp5I_kL>}QQ=rQAU9+0IHnIgVqpZ4!Q!`h4S@o}@=0(%BzfBn`*VAElmW)Si zBAbt8%%!6M4;(0!dwZ$gi2S7^MpcrQNv;!xZ$JZy6ZbOQQ~K6;oEih$VJhqB=i% zt*jF}E+d_vd3q0iqN>rHs4c2{T)J*~OOUW(!5zjF)$P%tlONMax3@I9J?5Bl<&Ik8 z)DN4e-uIjdEgq+HCMS1XMz!NhQ#1&s6f0>r*`Epf4?2OI zD8>6yAS$H2S)JWl{1-hNHcoR*2Rq0 zyf5E=bn)Qz`wuS9zHxDN+e%FH%+;kfuY1Dxih61MLI>e1{ty{{po^J-U(e70Q1LAy zIudcv6cH{00wsn>Lc_|=`O zN7V0iW$)9r-&1M74P9{x*|H-v!mlcYXa=r0F{2rTg2f$l{1qpoG(%RL-Jlhx5P6St z#bJS(fh$f%Y6f}#<5_WnDrP05it#9p{J3VW+uRzNE@^2 z)5hoo$ANg%4AXh}kW5BpSt}=wS@nuzRzl*KFtm|6#yoz|&@v@5I_d4FHRM6?v_@DW zvx+1#D=3j6XZ=1vqB5+N)5xr-)W^{p?a3#N9>pc4R0!Z{jZpcOr!`Ui)%@Ji8dHFU zQ3*f4^fjn~rA#wgqp#{cOfY&2wnIVEx~P(wPyG+0ucvG)^tpzxhSs#&=Vf1C52+Tw zHEZgUqV)y775awWPpbR|me+{4&no{K0}_}^h`hPFAHFN28nQuqVzbE4H2_lypCVcM znaKK3RyD+W&uqZbAZ#k<2v-;i_0y61dNGf@bMNZn!R5spcWz%@4C^CC?hf4gNvu!U zr{eoMuW7{d>05^>a%ay8GYL-s>jk|EvL6vv%)>-`3QJwMN@5!|Ft^RX}~iOzyN#n4CMZ z>|DoZ-qq!X`rxEeX4SGh6>61RRkof|X}R6{vwT_;Mmzib0tlg;is2W>c_EF70es5RTUnWh6rz~ zDm^&q^(8kDj6I9JRUs((W8JC>xLqY1=$@3& zO-HG>ytGd+Pwg;xFo& z)_~k~LpW<>&-1?MOtD`M_qC`8A5YJFVsmFzk$NFEoSowbXF$6@8yMRkKE zM(_w%#43bIlR=;GF+=^sNc|dh4S|iEvP(wO)YH)kusvdYdvvO=3$l&so>e!;o4Q`I zFayAwo&JLQ#Qsi_@uy$<)Kk~S2A+>@p1gNM{rRYJI|cs0oU_IFT4&tNC`(MXYZ9T1-pEtiU&jC7}$;@*fVldw@tk?{~4D^RX=1x;!N41?s zW)yeT4FpahbSq=-G$m6b`dxz+Qic}*0>G(-eZ{G3+!ou#GLm%$+79+8odNO+YG4&;v{W*+3ec zuWcEyu3ko=`OCiUR^PgL;4k)H8gjDt?%)!1&^&NN^}U&zxoV2*0I2WRgr*pIFTEsv zj+Xj@BeD~*PNS0awY%4LjrNNJ&+%RWLfzVCm=APTs;>NiMnc`S5$ZAqG-cN4fQCx+ zxu!?mmFU{_9&Vz)<6O2uFJ$SjMAsebLExCD`$}~3yZNw)D$&g00T=4n)kKA%}MZGMJ0dQjUuPpS0ORiaO6mmc5nmmY8I zhq$-{;GFuH`^L$$l6zueQ&z_cQ zKAv{7rwV&Tjnk*Pz*s|hJ!5}$gmsF`wPjAPHRMXSHB{F#Y;|B&PxAh38;t=O_w3XU zYj^!x?fcQLH-5VM7E6^dAz#7i9QB8_ z0_7f)Y-U@&S-9sxSyubS;=@2cf;H{|{w+h(&<{hLTeMLSo$UGFARh?V@vExJY2NN9V7?cYQ)6x!=f z$xQ?m^xFC184ysv<_jC0gMP;oUfK=KkZrMn64((XNs8VG?zI*yx4^_Q^fgdI|45Mx>y|yFETS z)()Ihs0W0T@J{0%5Oc>#O%I6XTlvLQK?(07gQcz!#1xfz+B^|1H<$0+y*1sSEpzA+ z-iKEAx^W$I{L?2-lQYA;c(=nbUc75d2ZDH4JH@_$cVpY*%87S3_B$99@1j>sr{4XR z1<87?F(A3m1Mbym?L8*h}j7 zPbfqR*q7qHKJQG1#z!9eAS)Z}j>=kw4U78Ikt*MS9)Sty!G!4bc`$5y!>vW_>YdDm z6k2LBOZD^Yoy;+GiO4$+GVC7(K*ol2WnqL$4|DXK8nsy+)(OPbifPd>3u_)(i3!^ z0Wq1bSL+8SRHu7qfzALqY@#z0cbDuoY7DS7jRD*}EpzI-`G9cs!221sJ+}d|rD=_> z2Qk?qCM6vKwTC1#;oD?qw(|CH>B@XzA&L1J;*^16vSr3c+;y@u>U|{cT9d7%z=@a) zkR>#B?^p()DVJdYpCx8IKt{Hu>HMy%KZKcMXqNbkw4b!8{}Nr7@;w+NlUtvKovEkC zpY6!P$|q^1PhcSx4n@dayQa_hH>bmoz{T(MY=_16h4SMr`+TCmEjy9@c7xruGxb5` zTPSR@6D}+uT>$%vkFT3*GhQ|0tN3`KO#kmjs;mpeOpCJp>_9#&xtDQHiXhYFT~B!Cq9c7a3+364EjAPJMoF+V2jEZ z=;YwrmKo6QW*had8tJ<3 zNZY#hSX*k+>Air82;E z+V$wFja5nzX09A<24pT+^!akMEz}VNaRSxfmzE&t5HJJ*{Sch2dypcmiNpJ6FzF8M zp-wT15dW+t39Eb?2C(4sKAD@V>G-TOnPcR-N}t>UO$b$X-8-ry3bQHhQA|D{ap0>7 z2qHT}!CkkVjx4~5E%Dx*!&4!Yl&+e5g%*_XUeCP#!(|z-`iOi+ZO7}5h=k7<;tRT8 z4*F>%%sBYt5nzx^2aNyI$|4zmsXsYU<-49U!T>Ez5d*_n6OriYTmt}sknvx6B~%}b z)S59l60rfptNzS&Dr6tJ#j@6bLdNLU0PM$P45D)g)5)ZN0AkhxF&SIE$8EqK@8+WdmU(FJAE<-=3F6%F)L|7<2P*yW)Iq4Y!!^N~ z#l;;fM;&+NH2N*|>|rT&>h8Cuz#RNqrT%@Se!VayW@oprt}gCf-MN47?BT`TZ=SvV&c%aA zXK!6RxOZ{)?DiYCFK=Du1R||nci9MXP3GyBe2l-Ko*O?~i*d>e)w<#6PB?it99>WR zRXM|>h6Xi3rlhI)aVqPYqrP`dy(VU&ZE$eBV3|k{2Ua0>ec#r`PXo~?%=+t?yN^pJ zZc0o?B&?9zO2$+M+H{ab?QyKvqGTMYcsu;h)sIiqbKEd|@!`WeZ{G8_MRUD4k(&qk zY6?)8HN*C07Ghy)FR%`lQw!C&w>vt2s&6>Zv$ z4g>;CS~K>8KZJy7VUVFRZFX*yY0N>kq_}BI7-Y~NNCU8bTukc}l1>vLrJC&zG&_e% zHT^HBRO7`9tbo)b;^GDH(58=-?+>(&>G8&UzdajG&>e(p!dHCGTuI;IiqEw}$X!=_ zUXS_%Z8DDRuK1jR&oFJjX_GJa-(t`!Cg}Zve(TU5Xm)NTapCEc(*1*Y^3}hH9&X^A z+fu_I#na!YteHLWVEY8k&h3Xn&?ji#zU1?||0(sZW$pSohPCrN;Y{cXX$Wlf-KWVK;I-I4<(~T$blqC)MHXwMA ztjPf*!$T1z({K;w@u8MXwzxn?C-3x?k9F1}QGf#OtDV#k(UpR@iQGjZ#!f#Mi$noW$t@BY zXkvx%j<-k@;we8%9bp});U{bxU@sCGTjF_GevzP6HO|gaT^_xvQD;F1g&Wl6?Vmmw zu=t{@fcKe4CsR~j*DW%f>GUCV1k(S~iGN3BBgk6N&~tLKIjn5duCr8NO1o3dx44>3 zQOO*=ve6BxxdKp+?c{iQEFH>6yhp_Ui*gCnZ*9a^2Djn=!X8)sA4XkGscG9&cv zM36)Kz<||_dU#!6P^%NY&TtcvBI2cg`@q^mqPcK&<24VfhfTn+jyRp6xdqTfe9|8YU?%K&>TSzD^ipWpN9n3G?2HGbhQwy%maq z#-urq1baheC(B+3IN*erGLCcDDukE%=}3K@60C1su%(Awle3@f5f`Fbjqi`tSEP)N zlp1oTmL(W*U#E%`M16EZOGGmV=XAlUG>#P_6WL+#I|@7H&s?G#->5tH-ne-8*43kL zU5ro8uMxiPd`M(z*{?+`O^}dfKXjTAG8t{g2Sloty%3A4Wly0c%-NsiabMXgo8JcV zB~~EO&feFo+NA6*`^!{bNZQx3ryOIRHC@ZzstPuUjI1v*{of$eW1hm=>?p75O-6L*Q1y1|%ZdE+C zo35cmwwwJ~VXJDcoR4Kwsw5=rt%?VDGk?i%RXxJNDXOufQ{MNSaL``H&m0Snw1!$-_4ZLh{4y(#%Q+rL^Yzpds+Yj>q6BWJD zsc#tKNcA=~syEgU4-yUPg3EsPyJf`_kb{Ba*Qj>wVxc&}P-5jogHtC5J5Y1|^w8d} z>jmAVg7c`_e7bTJm?GElCCOeixQ$WpveRV;9xPOM?k*La$93mh$AeAXxwD}Co=`Nn zIM1N4D?Xp2mkOpzUnd%zf`%Lw>`@4eOiZXAS<&FsHRQN7l0@@#8ZHL=%U&|r;U0Ij z;M|Iqy<~8q8yMKv=yA7`3{LHrKPVaOip|{to9i*7WAnrh&Va7v_y-WRhURi4b{>BJ zz|I%v`5{Muorj;{0s(XW2|zhCBj6?86RSE!uN<^<2$Kp*FZwg>`a$5B za%y*_fqpkE8vIy?uy7TWhuJl=xQE=S---UTC!aJ$m@Z`lsWJ9KzA|(nd?}(y?_V^x zf9@+oHFG!=pqiiYPC#7+uQXUpGrr?4MI<}iiT-95X~&|!1$7$UxAa1<9Y*w5YE7Ss z{(Oo8B!4qdG>#6C?h_pn`p}r}b3Lb83oWLn&-)Z_spjKpcZw&nuLn?~^r_Apr*{@t z+Q`^nz1VMaDsynHum&_{^DL-lYbfJn*y<4lnB+1>S6D-@(c~V3yW4l)x%heNF<5w^ z9vPr1$|5zjEBsFWsr4*UP;d)V3qMt0*I6&Y+6NJGTc*PA<|g*EWgBIEUU&|zEmPqS^Lfzz-j(kyAU~EI z!XjlOw`D3>o4_=#y9qw>sg$5Y25e>pGLo)OE;pBZ;g-e^GA>kgIAAj?8GfYAY?vUB z+Auc{>w<~HHnYM5(_o?X7$m*x;cRh?&5S1u!H~KVUPN;?<0}pgPY&YTeJ zFuX0WvE$$rYWkIlZvpmr>T$O}+;-X0(Avb|hzAyhu@OtzXn77^KJ!=-zOC#u;q1`nySzr~dxE_P1t)d>r zNq9P~vtBgK%tqy3wQZF)#=5}Dm4Umg-cC1QZ2H;%qH=;k)3eYR0|bK$0Kc10bUg-5 zq?3CLW}dJn=@heC4d_ZY3T4nAzJcZMvpk0neE5{xV-WsOh0nJwg{}7V z=Pd`~nTJUS?jB#2_WKbYEl30949$3HCJuzSW`ur-A@>-}>`j3vNjRH5g;oun{aGP@ zEYBV6GikogfRs~gxXA6pawC(pk>}yCcZr#VM-;WkU}|rHWjW9(_2yE-Lhd3FW2aVZ zrs85bFLbUT>_s92O{@?m)}a#OiCq()J-f11w}!MPR%9;{naq&qVfjUZC^ej;`aC*P zat-NBqr80XdIPk63gL*Ij3-&+^#g{dDta zb&ghH)0xqiEQK`n!(bkmBJ5xg@!edH!G){6w8P;710>{_vC7L0XE7;S5#Pf@qV?G$ zcOuBS+W%aCZyBF^D^pOpyNnmM%F)UVXATnSpYa2PzG5UdoM(JUbJzpiqFzGmjLw$z=NOh9or!DfBi)^ld8F1SY zhBu3P?&&+Qoq?>~){~B<*mr$51EvAvxop}}p3D}X)VSYbaY=8M`VKFk;D zzeege8tT~l?_7QBovX7qF77=%iu~})UT6D~dSUz=MQ2-or7i*D2P5?>lEz1>jmj!~ zDr;`OA7tW0^?TRUFAJJ!l^e$ln~D6e_%&O2061dA{~oC=dq7&eO*X!=j9El8qWafnyxRcX@5e#+mAID+YT7A-4!@psyTC~ zbji_{lPC_^?n=f-+nub*gl(;wKM&dN3T6)7?h21g!-l`zl^(tymg7t@o?LOrb{CW5 zP9Aji%@;gTbcyt0ChB;E;naA^o*<$FXYy50cK*Xq3HvKHRSDtpjyr?{l z4y2ukaqi@4{^kmC+7+x#(9jUL6K~8pv>yB=e$c$4qu}l z`Si@VwWi7kZoG`k;pdi5o1f{i5sDZW8KibK@yjlen-ekW(tPAw4_(Cg+qUPS=>V`_ zc`nW8JM;8OOI1}$)cO}v>v1cj4l+jF64JhkpO0C z@sNKb4BD8?1M_`kyq4a&uN!RRJ@KcAe=+f*X!=o(4W)6wYdtYdk~K19s?kCQend#v z_>+<9T0X!Vv{0+0u~}+6MBMoICw-g`*qp*uuJ4-l-nC&bNF1}70>G-|7?hV7^!wSr zj+$3hR}D>Mj$BWN4O-F{sMY+XqW}^dG^IBnm9rBSMk@Z&5o0U5Wd#x?X>5CuQT(MN z169m`Zdcbg`z^4*)f&PA{NpK01K%gBq#%7Dmn2Dxs|ms-;iZuI6o3Km@6)och5N)%{L6xCDNA zAdR9G1dk~%_aaE@yZO*?6{Uw+<}PGefD2hLrv?$^A|g%8j?g+lrbQ4z4uMQg1i7@8 zn^pw5aF9v=j1L5S>Vu1xwZlL8wTs(#U%BHlz@?r;_Px=k_}@}a1Zh;KX7CabBvq04 z(2^4!0+a&Ke7cxZspBS8rRx6z5J4-^T^`-4{mg#c0+VIFZ3AG!sQt`-sGJg`BV|s5 zo|I^Xx-*iw_c~6pY$&F)+y{6nSrPt;2TY)@Iq5I;8WuP{D~8`WY%iv<86+>OjJ&w=(`9e@=@P0|f)s}8uWwuOqKD0p7ovzDIo<;Lq7^eV zcl;v+6QM2Bf|J=;0@JV=%ih1}v)s?CXU1E0mRm+NN}nSGg$Wh619oOlVsPvvdnl{` z7j(p9+e0z#0NXvT#7k6VZD*b2{$-?I?+xfTZ(m-vvCLO}z`m+p9>2&6SoljqTt+u} zCVuM(O9*_82+R!n0xj;KcK)V*cA{P$G~6#fe0b;0d!Et`a^*xg^86lsD(+etI`iEV z_2MA!C(@Qd-an$Zu>DmHTyYB7vLiIYqb`ML2Cg`dm(fW3aOjGYQJNt!zZm--_JmJ3v}$#4*Fy_uwUT8TCd9D`b|D zLT1^ikXaGm-ou)xPRo4V*Z#aU>ya@W{y<0}6NWaD$SjRSX2{TTbuu~|>8CZ+LGU3Y ztdm(rI+dgSnxcYRCrdiOr_)gCQz~&owM$ z6Imb1tA<$bnGIMPgiYlTn!-@1pN`bmixcb4y{n4{mlto`xqWpptg{$7+*?11bq@Q~ z6HmZT-S|y#->>P1T5pgmhvG|3BZyPeK+5?e-#tugG4;`j`l8U7+SV$^3X#gUZtgHN z>tAn#rAKZQvi@S~!;vbtUu+o`5ka{(%xInt9IdGM*2VZ!H`VysQ}10j|K|prKqG=L zb6tk|;G}{?Tb8FnEuQ}>J^PzKp4|aKZ~54Ic4d!l)$ZTO7wb?LV^;0*aj5k+sVa-- zW-G}ojr5n2u52$I3y7VpZ@KAtm?xe> z+P*Nvk+yGXw0*20PE~#FAqL&NH1z6+#Q<@&f6Vz$V@eTypX=sjXlR%r=IKG-*Y$#~ zo7XC;yj?nxpQfQrBSYxDtZWtII%3BwaDz$nX=Vx@0J89vq$^iI`m1k0xVoifuHk3A z1#M#$?B=yeDe}r)NxEvcR0~7i}v69pHe%^+4VwfI6KD=&Va6P z{sF+wy1SaI)-2!5Q9Tz~ypD5b@--XPpW&wQCTsO@-Q4tiP(4*Hz^(qgtz5Jf%PY_= zP3quWW9Q9fDF+o##72igT;37R8O^er}i2yEw+!KXC!badhi zUIzKBx;fs|Wsrq10zY{A3+fa5i%`a&e(6(BT^m0+K6z?%^Caef7(gGXKOa@@tH6J` zh0mo7%umtTl^fNssy{hVuZoPK9I+T|_5|H_`g-#XXXYkKSdM4K z%toWh;|1PMQ(#BO-D!3o+fGw5HKN}&av`&d>S>YHE+s_)G}bv03NeE<#5 zG5qDw2XJNNy&Wa#XeoIZP~}9dtEeP>HMc5fO?n(QeXi$+EouceKJSUtQq9NHZX#8o zs;EhMWxF}m1;!f6>lyp2(=D&?N^_Ysp@v*B^J7~>bv?sYr&3<6WQM^9o7m~&hqc>( ztsMSn7aqUl<9ef3<3&BL>n~ChD{J-F6W0>4vijpuMX_6dev73_m@tH+{;*b{++&i> zY|D2I_dF=eYRdxYu16j+TIBo4jlHxsObdx2$q=j3m*MEc&LiG&;_=ttsTR5_Va;dRr@h{)RO{8a4#Rsd5t<#JYr+_PSGcJ~Yi&{hLUEM5OD7`y@AfT&q&y(ng^hmwfuxVQ6mc zXUD_wQ|kIn^9ApOv5AxnjWB7nHq@A@gb&I)jq2Y-3MTP3k-~%0AgELuRHshqk%vvh zk$r)6qCh<{F^FAO`!|u8i5_KF*qsu0g;;CeUK08@k&KFVu6n!PO{5SjC=HdEgv<>_ z^=~2>3hi~LQd%@CfYR$E|Au!IFk-WorHJo#F_h1HBS#He>9drECCr^_z!@YRd zpSmbU9K^dOG>xA<@m=a*vMSfMnCv)_~+b54cxHCE4k6 z`96fLp~dBAsJ=f?WZff|dECg4K|syY zj-=`@M~X1X!J{MCmnBrGy779POQBX5lD@%?P``gdAyUA;6z}zUXEHQC^5qDqAqTso zvPfdXqW*NG$~U1$U;=tDA$olt^qZQ%EI{q*oy>(4+D+!6<0941vv)Gb&}Fz4M_vxP zCv%2bxGPo6WF|J!T%&3%I#AJB!693W)L7=}6UJf~ z+<>k~J^@C>RCJQ!BxUPJOtm&*DmqCa-#+0aWhRjd5M}N9BCe^oVR$c~`lyVPRE2eZ zC?kE;+UTRYhBBom=un1hO>`>N4^GHV_s#;H0rEtb&P?2evg@cZz}hqhaAGp@yZL}{ z^}vT2wQ{(GrjfetuQk!0eA4g;Tfn#Oy--%qUJprfA7wD48c?oL=Gs=?9!jxw=6_)k+lvUV5BuBkWpIx*SWQsDSPS$(^RClqMx-mwfovkAihK1G5Yfaw*PXS)ags&TnSEEzj-R zHC@KPo}q>fJ+9UAeKa;A`$G9~mwhhL*AqAw--Q?*Y9H%-q$^R&w@}z*C)~V2x&ZbS zAAC2}X1r|#7G?W|1YO2J$-F`R(MhFl53@7P`ujaKk%${~ z1oPPf(16T96rq{=+Yv63Dn@PE8nEA^VJVO?4qezd6Tgu0IU*qO#!^&(Gx1|GaHNS} z5M&JChpdU85rXb8-StoWLfkOU8b@CacC3k?5rck@;;W!P#EQ(Z@VOq9FVM-sw;ef% z$TNJOwPB7*MWv;+Zre?Qis98KR+$sh<+$NR7W-_!H3=_d#q_dR`|^>54# zG@c+(-^~X_RF1aMy}#^3y}QsQBM^jXOc1CSIA|Z;e&dZ>Q_VqS2|`qkwwyYG&?rX( z2{a@Kvy33Pa_3HRrHGB;P#@mXgw$JpUb$`)ussIu$cQ9V&uGTtITX8^h)u?Jtq zc7}qxZaY0$z%MtIcxe`0&{dPK(1H@)>zUVoxGdvUACb?f?Rec0k?{FKd_niiL3w(Z zaq!1mfI-MaQNr*iC#rncb4D1Tr72=yIBOyjJ^6$N96`w6MUa3l=jwxzS~Dg`A~s-n z)o&!HLdK(eBWn#PWQ?L~fD~$AtpPD>h_sxTCrucqZQn%(kb)%?YYoU~Lw~gH7;8Ww zdeEtO`~{-2d-B$Rj5sj7D!c{|)e0LSwZglWIcy@ef_eIcyQJa);uwslCmSGHpbOA5 z7@?eGVG~If%+q_I12s?a`DIl1v2l`x4fQHVpF;!5LbQ(2U1G?D^5Ypv7B)t*&_AAB z-GWY}x(hy*v);H7)-7xz-NGi)Ep$gkROYd9x`hojnRZQZ^j*A>ZUGdM+l!gRAQxGdoWm%DM%r&f}h?4w_vU z>Tuq3n1tAY)-36|=~^7J^Uyj4)k`EvkmU}*!smQSw@|b3d2dQ5li!<;biuSxl+nMj zdh*avgkVAdkwaZY?GRfUiZuEy_3UAbPokq+T02|)`$+wIVM@%;ZeLwp+`GDS|K8cd zi@V=Ed;6V>2anF)x_EH!;;!uzpA(3*cHLzoU-IGmf_iTJY%P2#KU(W}qr2T?9dFP^ z2D+a3t8#`%4GkpC&7phN+@%A@e0v@k# z|1F)+DDfVVFdYn0yBhmD(e{rNX4(IFAw!n*R0`G9k5ANd+%P;S<-B5!>&1!OJV4W6 zNk5ihS#x;Re~gN^qf&w`Rp)Ts-7z;LC#m9ZW(62>{LQS|mFF>TW+lUqw3(3_mLFj= zE0{QJGb=nW4HoWZR(f>Oy+>}1>FnZj*gh10GeeMWb6G2KT}e-*yOwFLU)XC0Y-Sne zWv;F}?q*hqt;e*PRg#bPHZfo`%lI$X-GVz5+5{*)Gu= zgHo;Q1%13k6_8seSKH3o%b=fO%Jm1*j)nAi11NTTd^|!HzT(s9!vI2_JA|x%##_)f zHtG+wNI9~R`5pLo*QC`ZUz+|v+u>JC?E3@#mhBI;wUW5-^x~z9c=FZ1h*}reQo|s{ z)8DA9nQcG&A@O~Jw)R6I=o7R7Rfg@rde^dc{T##EOE)pLJ>g8~`sD~shh{S++K;}0 zMepFZU-s;BtJdQsi(OXud()x$@$3nc;|2&HiX)k)8%N?dc)6t9NxMSa+u<4+uJ2!V zMU+g#9hlcEC&u384CkMLHK{H&60gKDY`IeHGZYa+>~kaaX8ZQbV0uFJ&m+~bSO60v zJW9M}mGG`vHaMUHsgSvKPBb>Ni5nD)CZb9g z+7paSq=P&U`^{C*Lzpj2dtBBcQGf#OtDV$zl9CZ}7l{}<9eHYtjrWedNMxXi6~a5- zB2kDZIhRz$OGWVi42XTF;JwfyHjn8-iZdD;{g&X znWI-W`XP-Zb6X>sV-0CakFdE4s~dH_Pd_-KI^R1DG+K|2#C-rsx&BC7D^)vOwzoBn z)_rv&_1%16MAePkwXJU4QgvuvXSj$-QEzplR!e%FVLrY3`y&}GCV)f>2Z>B|qnq34 z3J~>2T4sF6RX1uuCKhY{x0KTvZY>s1`Xf=*zHeLhq35$P*Nb;7Md4Kx6= z?B$8r-cZ@eveyF+yy{WP1vR{Zxj_ANq`poG*0(O$(nI7fBlfv8dz~IpqWb$Jbit}Ljuj#k*jt3r5=)#6>_(#lF>)ms)k7lxm#7iydhgv;c;n*@V2VbgAZat zskNDXs|u%NTt>+ZC+*7M*fNu)FLeckz2)Ir2%l)HcjFD(x?KhrZKz6P4+ zuQDj?iqE^nxkgPlKSvvK7(HOT{Z~Ap9vwu3`gwL8^-rhaVz7>KC4)v5Y`h&LgSvr% zeT^P>f6z(Vgvn4G*P^uAF1r z=L|!YU!_uIs3)sR1i#=v-#n{sjyL4bHzsVr4`8ytc-2I|6#QtUNMF>FUyqQclto!e zC@cWuJe8ZU}3Af~UHf*j^_M?{amV-iq z@8G1&TT4j+dCNhJjifC4`CJY%P{at@9cMWx#L#0{4r&`GI!SoTLB@tSo|Rq>Xu`kw zIjWVTBO#s195P&}mD@jk!d_s5OVB~{sXHooqoW`z`Rf!qhMnUXi9(30R@*L>ce_J_@TPqyA0G*=Psi& zA1nHspTlx${V<sPE>(!c|Zn!a`$%tOuO56-?`xS7{r8r+$zq(3w25T zbWs+msa@fBa#he-q@a*ADc=R@azAR1L6V6*ZJ9_9%ybme!x0Xxk^{n#{ZZR875o-a zvVxjI(zGmF*1=cM^2KPd&3vgn1`F1w+>BIEUU&|zEmPqS^QqqEz4AkCpWFVw z9)Id3_0(wm#OagSCa3tDS;53%n_1z3X|T|G3>JKmFF1#@g{Z$7dzhcn zavmLL=Q0aA5M%p;@^q#`-HQ`W~ zi3LqzeZa(1+cR!H?dlolz~>YdyU|&(>oGW=Mrt?nckWfm~2yuzqR`u4`%k znD-zm1f!#4Cl{PLA=sYsPR-3szFxfR1)W^b^ezDLGPnGA3i_F~t@U!jbqtXK%x*u# zlVqXtaW5B)%g5J_BuC}r?bm~H!A*vfT@CrRYY?=_mqso)MZ0Uz<$|`ypne{d3!10X zI_pJisoALf`KrAX>7BvKGp$@X2nO{>32SE0^PFICYQGpl!QiI-KK`flEm+qdzVnS+ zOZ{F3xsy~~nJ1hHT_Fv746Z{1I#Q28_y(50<1#vK+d=q46+Yjd9k9pXIy4sGrs`HW zxUIDqXueQ>gbZH|1kXd4QIElOXgI($+>?1+A1axo6md;^dXd|Q9l{H$&a(5qV;Edb zWcCw65<+VGu`51nQy<^W|OxqVpn8fgyMvBZMFpp}^AKc2-?WfklD z@&@RrSX`2O4CY2AX(P|WeoqvPK+LP`MWVp6q!$S(VIg;sh^2**WCC+(f&PdMCh5Xr zY9rD#gfn>|D};$X2J;`!7l~t3nP$JsSc=$-L?$!jd6=dvkqf$jlu&(MhK{PGb+#eh zd>ZNVrsRfn=^$F3gw9IOVSM5bWOhSPk`o(FBbnXQ$m~A*n|JSg>()0O-Foxx{cqg9 zd+Xl)2bX81|4yeqVdclnP+nCQ6%kHtE4lfU`W){Cpq!(7Ap=Xw4{4;Fn;PXDsn|3f z(v%*dP62X5Eivi`M~y2>v(TBLRoHZ9to9G7}WEITr3GDi#c^#@-d^nIqWgG4lU_Ob&}zyE%IpCneE{q z%jXtRhIc4$-cirJbtnX7?Y3TaIKWV`^n_w>XGv&q=Ak6i0uD3bv9XgfsrHEvUGM*E zq<*8Jj=lfR)wkZcI(y^d-b1$h@XKCH`;vNL{2WC~TmRV}ks*4?@dqRIE0V@Xsts9_ z^UsT!KBIa@{Jm@Hmj%tV%8g@&%|w1!PG!eqkbdL8M{3I+kchWMsWb@`Cu#~|w`7Q} ztY1|BcA{Qjdmtrnd|yr^=S9BA1FR++eO!+H(WrUXbbZ-PuRrFyJ)AOS-O`)@YRGn1 z;B=|x%yBzOv*aX-L$f4eI^d_OFr>-xYxhC{Zy zm>kQ_E8=F=DYh>_>p|OH#**3V?vB6R72@o%ZFjZ2roGw^+3qsH%y)bH1M_h<#GR+j zGoba$GYL9v{@szSY@bD9KlAv>bNusMzxrl>g~)t{O7=3TVl_c>62@7i*fUS>;m%$C z;;bZXb{?sDy7n#PexT z4xg_s^gD2@h;e2b0PeZ_ufF}@>eh1W__Qg9cd)m;roxMo^@XUP{MyCsyRY2wOwQKC z>mTcRrZ#!6?wg^VLg*sK-?ohcXRcF-fBK}QswyRF{R^oj^;SsjWsED`c-6j2+tc2G z?yBVZ%zljwGRE0?SO3B99Au1o-1aD+>kDT@m%99)K%M^`zf=?QfXZ=clg4p=bQh{x}UtQ*p z%{2b$SmS`#dSV)=*YXlMM?M&8&1fM5KO!V-{K-glEg#?wBGig~Mk(Zc9N}%%6IrH; zD<9S0pY(A)U~>vrxxQ=Gd)JK3V=JFa0IW&|Tp6ok1F!yd)V!*?YN*+*$4|YcKdfEd zxq8&M;e!$L+plQIp;q&kjsi%KUpl1Zj{Kz~##RMYv|QLygp4u5LWh>WbY!54xpcIv zYb!IV5L=IF>8NeOsHgc$N5-bOo<7FXLF~0S1De4+X`~}5ok1P)-$*lPp5Cmstb|co zQi~3v#|A+B7k!kaD*+9!QceOni^P93BmSdHAa%y!^JMhxz0xfzjGVc~db5U5$Nh-s zjB`}YM+Z=DQ0I}F&pe$ws40Cy-4aBQIsxqmPul(17}lfSMYdbjgFS&*_e^i{kVMtv9<5Zw;x?Rc->SYeBmNU@Y*%S_Bu zq67=|_R-f63Vj*4%w$|;K(OzIEHj0eD@^TUtvoEoGHjX2ILi>v?Qr5d?G`bW)cQ8Rf}~({UW) zQ$IbB7b-Dv@nJ4s2<4}9-F3@8lsHQ~d6`D!MPH`i7n@dm?a7O7Pw2}}b2>$r_qg&? z)BG^Vdsb6RUd+!B-wc$Wu4Wl|apk8|>V)*<1*%qp6ov{xzHP~i<~N4C5Jd#Z@fOk- zDkA$w2qr>XrUfUnu>@u&a?A$==8Hb}{k(c+yk+OUN}nSGg$Wh619oOlV(?2WKmI%_ zwmlT%4zS%38NDcAC91NvvrcmVGE%Si2K1Y^FE86z=Bqw5UsW%UUu1R!pkb#OB4X#kbli}m9RynpAy5eM%W{@iftvH3q zD@^jEO)J*+K5)g!NX;Pcsc@gC}i?9 zk^yTXh0MHDAu}Vsy-&uYI<2`=$jo~cG7BMvj33%XB#|+XA2hU<)?g#*J*R+9MyH+q zw1%2lJ6f{mi4T_m& z2U;UkjOA%fRE)KnIa(7HW1(*bY0W&PHB(oIrJL|$PYFTiXw9OmlbI7AV;iNf^spIP z)9S33eSJM-d;r%hs7s3e@D~UMy~R}d4J@z0>rEIPrpmv@phD(U8O#mkR6{msM{M*( z9~iX|-k&Q)L} zkz>2{lUP@=Pd)Jjl+a&8yNZ3krXOm(L9QH%FEu4?d2;^94{4_5nEL2MeNkvkZEKZd zg-GRFH+LBP7DR6c-1@20{x)Ja)Q2NgZok<2wTK92v`!Cqe$(g;81|!PMduB(9;Y3~ z#3E9Ca8dyjEz46H$n#&NXMgj@^LE^2O4q8m-+T9+hZnarqJHbny*I}vp8o;6*|-l__)C2jBIu%FZjp&h{<%T`q~I_d@{XKHyDikGEI`p4d? z3g!*jstS)wLxi_gl^%TP{?&|k`AJr262{)DVwC)IzcUqRGIy)WFfwZmt=$~aSilSM zMBDO@z9}II25YOzcrnk-$CyRUsq2aALP*m0o|E(~B1xZle1(fN;wY`4>@Z14(dRmP zEh0tVqEpe=Zo+lLM}Ttunaq8JLAJGUvcAQFFqPb=g)qcI+P*Nvk+yGPw0#36khi{V zn+n~$G;?FX;;zB)#2D!N@{>fSH7ewZ3n!MmZH?M^i5{9Qw=AFpIWK!)(<&D(~#-Wp!xjM&qJIEJxgE(ts&-2`c zqzn7y5Ong=7vnVJJG|J=sv`A5T)&i~zMnwA4k~GzFQ9$^v2$qcRKA&``ui5IFDl@> z`5GvnCSTf7eLa)42DlyyZ&{}qPt`%EHmN+W#?CnojEvrfLR^;79;V}ExKe(ooMYP) zO8vt~{Tg)*0nLF5OnFcFZry%(>q_6~asPp%6JPK$$Y<5f@un_=EJP#t!P8$*pV(i7 zGXC^SpL*)r_{s6fQ=^+F??ogL1L!05=cCGf75Fa~OpTa?l!5u;Prr)eCnxGvkx{ff z;y58u=@ZR!KlXhr9G_36Mkgcc>PI6*-hDwJN3^_vwWVG-1M#VHE}A(b)Sc@66FRyD zwx)cXbj?DLgC<3C>PadhCIEJA5LvbQFj-zWVZZnizIR`6o6{OP^Y0 z_M{wV>+jU|srPO$cbW`uGuFc$XQwFyO(mym`Mk=yg`CgaX)>bC@vQVt zgZ2Wv5hy1r4W`$oYyPq?kzGX6;DwO}v-j@c5-c#hDN9t}+plt7M)F?s^q!J*S}FMr zcNOw%8kM9k+>N(Ow4q%%#0M_`p>Cco(M!_ZfG#7U?$QW#83WqZHUm?DwpmwN_G_D| zu6Lw*iC)n{C)KVJeQD*fhfWFR=|S5}eK)6VHkT@gPa4aqtwdka6t6dw*ETOAKCLfO z_B(cKa@{r;3eSWq(an9s08K9}nl?W}d^4zRUS?F%L{K`dg1-uG6l(1rXOIsxSqUQyi&~c`GJW1ds->Ebr`?c>vF96p z;2G;Sl-D!%S5H{CAs%E-4K==sg-OBDv8|!Ho?)v4t9p{lm{>!nohEwEs~^^G|Fzop zK^Gps3Ml; zepWXo0;f?ri+t@l{9Gxgm9RWxtUC50v9q=I4^Gso)7liB_Ij&Qtygz%J~i`&9LRzt z+*c;3&CyhUGg9RyG>CNxFYR@wU*1n83~!s*zljt`M7n-BP&(X%s#N$fY$7E?BTOn` z-s9aw3MTP3k-~%0AgELuRHshqk>Z<335m8RN&hAiGf{~^A#MS&u)JwnVZ&1Wn@C1Q zJ6FA3?{j?$#4HuLZj-ertPX`*8L5GUc)7d4A?>~t6M&f*>rOUFs09uUnq ziX^wUE1$pK)gHanuFi6AX%?91|;`+z`Z(p z%XRwd??c!cT3mjH>U&F(R~69P45<7V1e%yWA~30(D0+vP|uI?)+WOZ<)2J7SAD)6mr!PxRf4-r{{-;z zE!1@MFGq?n$-(bNu&axTRW@>v7J8d=I#;e*c6*q=0=X-s|(uWN3Vx4)(4T52-WTXte{wCk#OGRM#* zN>VV_6&T3i!6Hb|`X_USS-L`T*Eg99@liRGw0v6?oDE$YcHNUXBP)F#l$*?`u$wgaW$FWQKRBT}-NB|6 z+8Kb(Oq`g^bV=N2!-u#xjRE|HvMY2G^0DqUhBH?We0Y{l`|U&h{}g>-5R z%s<>rs{>~M7s}>bssCD&nI;B6Otx~6ME?x&&7hcUo$(Qk0_=NxThQk!Cnj523Y>_^ z0G>i8@*T?n^spHQ@L6KU17u_?Ky=Wb`a_r^hh~YtNc%~f`r~(tZhfM4rk);uwj1S6(4*z)n>eE24C^=b<06^i!FED53-Pwg0`{)u0Ro3Mw;xT`pk@d=f6j~=lment%XJu2-C z?2r?ZZdATNCkNklAarEMhjbi#PtU`KK@i;cbpP}~ z5UB6wf?^((qpjACAZ&y?KHFbp%12K=t>fB?vkM3_(CY1SjhrqzIsB3Z3(R7LxA8d1w!Hicy64 zXDvzCoPORXY;!dopLHf|!uhPyCpSkE7U!YLu6sxIL;=6>RO0Bt0S&R8q2R9DPDd8> z#LiALK-D2(XH7(a2CA-_e1#U2@OsfYUU=0<>oaOQUUx(*e7+E0(EV~yo*rf#{P74d z2-zcv6HFi__Zz4`IZ@@io-@J#Elm*v!&wuN=n+|^0D6!aHTA(rtr?Re5gRbPO518r zh0J8F0fmgwkp_|1l0prvH6UgU`FxY{JOv?i=!#ftKt>z-qjkqv0}9cT?~l?Gnzcc) z)_{ySFuWQP38DFfo&(%h+ci)t&^62%i%{=c=CF>`3g+n(?vffDLbF6zvapUM3+qnF z!iKh_`!Cej^bE>f z@Hvl*J2s9wM8zGbWso{-Q|d5xbsh8G3qIFn-2xQ@+_Tg{vkOBV&U+4%06VOyudTlo zhwMDGPC@k&NfKna1D=*Y=d-+pnvKtUvpmLe-*lu4riG%6{*BdD&2x+%q16faKX<^h@`3%aecK{rQJ{l}1Cg<#~*oS;_DtZDzw7ne#WZf{DX6v%&+@VBv0NrAH^- zd*s%b&MrQKH?tV&{vWV+<#c$!W|m=I=IXlRZf1qp5UJd4C8A)50J}%CxmnrZETElH#T#2O(p>SmMz5I2`T$igg8ePqA82hQapCEc(*1*Y^0i+?-4faMls4v20ywv&hQM96 z+0Ghmwvz^xyX13YKNNyKL38kVr0wN*53JpMa5Su)=Lu`V2FNhmzZ?PE&}^ne`_ZvD zp!YSd6noj@%&l6Fmn@vgPVaS&8z6irj%1#0Jc-Bq?aSpHH5}g_;>hq&M9DPVuP6A0 zpK^w?&&Q_oWvP*PC5~asm1>_M;ytGApAb?B{_{w6EEd4T2#;c4Cku+bQ#mhWU>0vq zr2hJ(htUChQ@J3|YsH?z?ZM6^r&#VL=-&RiC8Z;(bfG=LNV6M{pL$JS`*3yV>QT=o z2F6Lycr>crse!#n6rh0nYKQ)?)ZC$K49$?T!ac@Lol_Op3>Chdh&hJ6NMxXi=QZ9U zQHZC^M=&YBbrmfWqUe>><35M`F_%1KNvy>ZM2^039F&w`XgogLKEz33^)S}n)(PD*0l%eIIl1$eji zN~wqAr_}YEPmWKXI$UU3Opp$iaFi}rrb5n z)sMe#we*MVt*T`7QMRgKl0xoQRWNVJR#kXh8X~-{s`TJQOenRQ@T7#?ttuwEVBU<{ zU(b)ZRpoj?bA4Aqq0io`cx*RaLwCHbst`|NyL>HwXs*oIX35^Fcz`!^v;0=oBOKfS zja@WH=z=?gg4KIYIJk*KcINRkCYbu=a$jdan5vHE1z?BDsn)={oxaA8HuK{BXaV_zfeR&94X$>7$0`2&){1$wn$sy*CUY_3;DGB!{A z;0)+m0e%3{a%e6`V&~!aEjtGv>$ucAJhQ^#n`4q6gq|?xp8(K9GXfG&4sT@Zz19SNq{pgtr{T*a-7q2@=VSTLBT_ zYXx&T$UqSzXm_0Dpb$gSv0vE((U2)DNv59>V>!s!5XZCfF2Tat^OG&Am7@#)bSAUU zaJP|G&OCj>@TlcMm<8Tn!4x2PW2Wwg7bz$E+eU)7tr5I+pr9EZd>&-EE^>>g7;x*l z$Ze@zC=au1VR6rPNenM9`kTe12f)Raj}D?g;Y$%0wo>tv`^r$=G|`Iw%%zBK zZlET@D-Cov$3*nE&=l3aH}P>!^tZK0I}!a+n=#(E^g^#4hF+9f(`TYTpP~TC-`a70`eMNWO~#DcSQrV_d5Fc-oyYDt)R8#)TqI#{TNWev8$q z%(1NoD7emwZ4G6d3|l>-z;C>i@dM}RrKoM0cDL`obMd%)3>JR0N6v={HB3H#!3qKrDGReTouvl3QF}z|hDVzj zDSwx=yS(Em{$^G%aoA>7cwibVv>t=`6f8aZAPb~V1Hsu>kPZ&m%wonH1ybAeZCnQ-24uHAE?)i>Rflg~&$w+@ zS1!+rf!Hng2b~p%L5|dJTcdU(^%%?t*#hH^L4Hel$f|WM{nVJE=Wws%B1&H)diJh6 zqU9c>qhl)L<0(xkF%6yE zS-gVtBg1E*;Q-TcN9Og)$vm5)LoMfy9l3qjAxxm^OuWZn_7g%9LTdZ4?Dr!)n&oh~ zdI$ZeX+~G7YC3ysa{I9CFH;~&63%8%p_Q0re^!8@T#<41?8+?c`-T^iS91HX>_yUh zR(KxvCpLBwUiM;WxUv_C0&sGtXsNN45*Bh7i5NTmpR!2o*o#C4nph!BtV2!1v=C2d z;O4t%n8adpr}AZ-y+~wiiRWQK`_OeV&y|5bFVB_fNXa&&Tbx6Ep8l7H4W!D%G8;EF#pdb5wx(91D;Z=csay4v_qi zLgn24>AWFL=@IG_&|^^72K9rZZl?{e!rqsd+8%?8b6A9}|D{96lQ8PLITh;C)m|Pb zk-fr1Ka?0?UT(OENfGruJS1A4J#r_4e666@l)Af&m$u5$$_*C|66v2IS_XRzE>k|V z)H7%=H(XM6nO1JNIJa0lksDGQ{<$85+B|ZxBv}7~;84E^(Gjr6;4HMHBh^WUpSH*Y zzhr=0IvIY3alnsy?#)6WC~LQMq+^NS_A7Phtxyc~EEL<@SrQsNeJBa_fCEf;DdQ^V zDb%?9p!bCOuaWwVhC24pAMi|fe1FK6A7YA)FMDC^OX`L3a};50{gryeiRy!q`V~pz z`&ApVw&$CJQ7Wq6yQY3w&`hh`IA+*Pf4eI^ zyj!d&J9OEGIB2_z$?HNVm$78_y1V0VcZE0;U72g?4C#CmcW)2ydeC;4 z0cO72g+0~wNlu$ltsk8>|E}$IxHyMmKm9L_*`DK{=l1Fwj#;Q=FXPMA9mDoJT%1G2 zp8i)KcLtcMX^Zc7xP+DN+K~=EQ^X+IFi%kd9-Tn@Vb-@w00o|TJb#!|4Bhq-5Hadn z-hLL{H6V`SPf>YVzld?_rNT=3(O~P zeqwy>srRnO;1qN8fTbe?Rm^~HSJzf%R3WwsQ`R4S1m{f+U!k7nFC7`1;(GcxzCf+^ zc{b1t{{PgyTd!rwQSV!GRa0*ASbmr_QoLDAexKRoXk1$5_yY^j1jZiyx4B5 z*O((BBQNnA|A@?KKOITw3~I=%BhBEYdFY`?|jEE4~jr}s!8 z!Py%f45xHk!@|hr%r@53FA)L+eU2Ygpys0sK{GI={Gdi^zL`<;F$XoJPpCOS1gVpP ze(*$XSLfY^)|b2A3N7qk>^&fP{^}B(FL(<0x_Wi`at8%0{*VwK`S*!>o`!*nZxJz)==(etSjO+E zEKS-y?X*aJaH3uvVSCC-#`v*$hh<;Pok@GMtU_yzarM)@mOv5(fW@dw*~u8shy7XF zXP9?JE;EH(*DW*jNhoVC3Rcv(WhUk+?U_vn9>-sa_9K^>jH`?Y_WhV;rVw+&%Z$tx z4irG+mYIyRjPbm?beOF>teDj_Nr2$%`(LgRV0-zIgINB_=*TqVm)E%#xQm_0|keUN$j#S)!B6 zPeBu31Hlk3KV8S=ryECJc;%0Y|I`T4ys+Ayxp+b;vTJoa#jUg{Y5kYdi z1@vVOR4eF0@P7c9ScJArOHO8E3Cy;T82iXL`>ao7Kdqjh?%0Ve;~T6wM|ujg2o<*@ zc4kjv@Um3d!!nD;M7D=w+!3}r!OFVUZVd}OG)83IwO9W%QD5i{=&#d|dj;jr!J!dS#UN%{DfD_Dx+g{`lVGC*!;y_8oE8V5{5M z6{nCbJ3=FD2MS*3BxLN06Em9QuQ=Uzg`(%)u`5nSX~rzx`%x=SA@ZK)io=08V^^Gv z)Qs}}(N>&29)dnA71GAIR^&@9V~1&}5yy;Q--DOXWz-ubtdP-Sb!bzkLWY{u`D8q* z)0#Vl%$(Zd$a?`OWOCUP9oB@QjU+O2Bas<1v|L)FzlZnJ8tNc;=K-BebZVz-jiP2@ z^GGK%k90EN4goa0eSkz|S#zh8nNx|Br#0J{)=;UD_R&CVKrt5GEm;#W<(h@9<6^9> zqcu@67V06Wli6joMmx&BH@Y_Z#YTBrQ`X7sh>vlN(ly04w5HW8G5h*Dq*@TyP@`Gv z3j)EQx0ouwf#o$j21?1#1VAyv=u=gGL364M=7w^rF&nfeHVrYGK7+;|pn-wHXYdNo ziLDRiRb#C8%m%;FRM!a);d&xEoplvhyAc7m{M~O z?G~19V~V>~m5h$MfswK^Ih)J-nc{9$1@p#iRfWf;A;R0LN)H~of4TAtv6ZH94Jl#4 z=5AFnN^VYGvzn!9V!7brY-a4OD#OUEHMDkfL}LLj#1pnr2nH8(8)I)(887C!S$?b9 z6K0jiVM3C=_nf3}9!dJ*%HBFaQ#pr}rrO zI$GA;^c}s{wwKOOCpUMpz6Fr=H5att49QvJleuxhqK*k-pD7J?Q(oUeI;(+C-JN3n%hhbX>i7Gh9i!i`em|eXr@wXbK)c z>_+%X(v2%1{n9rcKEA!#S%7SQ#{YPGsZgMs*Y+Z#$nHwg&8+<>|Ep$}U70IM7wBa? z^ln~HY&Wlk6%vN0pDsF)iOHn=i?(gjrVJQ;W6N5-d~vG_QLT|Z&vTu;7WT^_=;XEO zip~9?iqvCv;OrbfI0L{n3hm|Q?;v&#t)0p@b5zeo176>9Z>TKf*#JS4Z}h0Xo;m*n zxE=~`S*ICK)oULUnoiFTi@;i`1Nl{KbST7S8SP*?X|2S*CrFuB;q2;vPt>nb*AUpw zDT7aG>gnjjXS}%a1$Aq>t&1BA(FlIW7;5yuIMN}p(+`?2q_;rMYTH9C3w@gGhU zdDj7f9MN*Ax64>U+Nvp&2Ncp_X77)ooVb`80 zeZBdXGjkIp&hf0gOEBY^J57Nd9e1bMe`GsN$<&B`%S7(}?CI|`1=ILDP2pi_NN{$V z(qjkeE#>!Y55d=A2ip~Ir-@;AM7^+iT7uqE4rp&v%$+8~+l=*a&)I1TL4yx!cGeFc zNQKTSwUa(?r^$#m$FtHq4XSs(87Lnb##N$k zXqD(Uy|#H_Kg7j70JpTBi8s|-mbq&3ZnVv_uDk_Enir^;8qE8@gcvfkiThN~4_nka zLm}cz9#XB;Vmj-FRQ)pPo90yK8S65X*E9B4M_4y_rMb*0uA)dr13R`gRM#_X^-TH7|5_pYs0)u@^?|xo>*=x{sP!MK36iyn>)Q1mLGqtWDvI6uZ?{;g zgn47K+wa#3lzU9FnQi&5;hqO&S#4P$-Sx;JqeVWs+#EW)`%{L*kYtIk(wE`r=9Uj1 zpI<)s&A0AeK0Lo~dS2dMp4W|uz^TMrW!b2XoJdVpV zrCP7<-h66epNFk$uxwUhh>DC={mn#`o6sQECA_rPodjP<&KT9di4;ggx_-D%a^pv) zDhGZFn@Gvf2$M>f_jEUrf=Rqhr0}3L2rAVE)u~f@ZT2cu5HyL95r z!>59!etH0-T`%Y*yiwGfCc-aQ@eO4-gFCyjbQv~7eE-gyQ~T_cDMpvvuMItlL*ZJ9%t@IJB<-lgl9 zbI{J3r=LDY&J4-o-C+Rr8ezq|wsatfcXcq?7x1&Ny%;Me-d)=7U{Jh^UM!vl)Y}#$ z>$S#!wA&nF za;=urIWZ!^qp?<^ahW~xrR|3L>xm*va`3wm?86>y!ZoeY9B*}+d^N7VRDX0rAyUA; z6z}zU=X7j*FwD)nKR7N6^i@5$y|sJf#L3-u-FRHJ()AI(&s_! zd+ZIlnZ6H6H>$>>0~MVW4B29&##$OR7Gp7_xgi(<#8i1UM<*%nYT0EZrdk>?RR>7{ z#edk+<|SpO@b8d2S_iu9nA#UnCwRkO0)14pG|`(J9wcRcD0NP&AAA~~-ZzvTVzL`h zOr~?1?%L-q)#=_@XlDRAGjSKnE~CZ(OVb#@i^-_hk+?>gnsC5?*t&o~wMJoeP+m+n zk4T9YM`7{kR~Iv_4t363;R|KAw({0OlDUN>=4Xg+2E}BHl#iG?Ui;p}$GJwCD@%bB zF&V&9nB&wv%K$VsGYsI9#Eb{X$W}Bh+ko}sYKl5vlnFlGXCmMDAeChaFJ9oYSRl_`#tJ+@G}|9v1OAr z@e3KBBLWg{jOP0Esi!y-UyuyQxFVhAIrn2t{EQHEhv~k5;uqrP=o7!P;h`r>*2K?< zLBB`Qmy%r-0y$Wr@&!6M_@*TXt4O|Jo_>ldP3wy&%^jW|tbp!-o+@WlK{x2QTABPpxvMs| zaXH$;5rjrL8ek)kAS^S2;L6dq)JtnTL4Z1fAWooOBzt8Ef*v+Q5YP|7$+`zA0`;x` zUzn7J_E2XSMTmdak^~SBpZ1yCLd~Y$ISR_j=~_A1=#y!w2c}YB$|<1Sx#J z5MR(;4A4&_VaCB9KPF%ha!;Cis6RbX<-49U!T>Ez5d*_nry|jFCghiXl911dfH&0l zCThc&9EsR~;nk2)2ss8_>sV_*A!8I>1Ef#`YYm85gUA?A)C6>bV!}32GV#`cj5hQ~ z>z=U&6ru+|IA`QNOjDV~qq3p$)_{ySFuW?f2B2+l4*A58zzC@o-gV3&QY);CT7kQy z;sN3+Dg$&7mMp9y$-=5rvH*^M=vwTdLsa*%(!%s%czPuZDDG_FUftXI<7UfN}!#F%K+t(Corc zhl`%WB)|@9w1*a8FZx@t&(1aNby+K_mq?Nz%N#(`GSb0mE#jSd)2>A)UOw&#OnOcoT@Dkl(W?YhfGup7^Q$;a`_>c!~` zwK%4{P_0vr?vj&r$}#oCUz9UEYG@#7Qhs2dcNzbK>*`H06K#Wo;|0s9^l)GmayOE$ zIg&$`chAvxAD2$tl$ef4So8%3DNM4~zq9+-6pIU>Ypk&8+n3p?i;mD}qaSp+`@9=9;VQnd?G~bdS0_ z6;N+^cZzv$dv)b+W*O$?dXc-C6=F*?n3x+DwxN#L%rgGVb+-o`qRwhZnuB3$AVG@` zfOM^i(W$9#O)L*yM;bKq_|q4-=QBe{f`dAY0&=uG#mQu6)z&1gEzmlWqpdsTXshXI^)Rhiq9RbNt4Y%waHiJiqAEA#RR=S(C^s( zKx-?B3r{~S-9LyYUx!7sIc#ACa(#k!w$w04@ib(&X8M1-?RnoPXl*|ff<8gJdHe7E zPpNk;YuC>)texixXTsV)xBUt+nhwooO0*vx3ya=yF8!KEmOHhXu2{&j!rz-wU~=35 z;X`pG^K|1${4PI+fsYM84O}BKJQPtf4R>H(ubg7%)}uIVZ%B2ik$5GJVat_jpCRmv z*%|Q))AsL^C4}l9C#qwy047FwGy{^trHNvWQQtl3VRXRWR4&N#TCt~adpz?GKKY^n zV5pe8Dl-%dViQrN3+)L;CelHkhr`|_=LOVW?jliu0`9Aw)SFAm2)TB%pI6^gw` zWNeA&VfjUZs4!fkx;%PSqt1d3lDKuG%QH_u9kKWZw_pS7MTTo2ucM=r3^Do2#&smG zGf(fSYy?>g`u0vvwu>qo*X~ZWHB}|$@dqlI%Ro^TsQDp{By(#cnPUxUN{>*dfXnvu z0lyy{QJwFcwm}=M_xDHISgG33Ov*exR5zk`Cj$ML4~(d~ajmh(VXAEa=|O$zb%yJh z6an0X{=J7ppt><9wOJKoHL#q{a2HoMt{o&Y)s6le7y4$fx^a{8p`EL4+yO|3KF(D) zZY&m0`Xf{W-9XP{PUDZ5`>@Eu;;VIwxPn_J+z%mc1TuEPq<<-K(EW z)R!s2`t?h;^pHC^*-!R}3pM!o#OB+Mv>N^wlc`2hKrG}iT z`8rP*9e#L1qg9!MbGl$v8pjHeQ`un%yO-QRWKu!`3jct}(y{@FSXz`OH>tE_uAIHs zS;qCsQV#V!B2~+NfkoA_r_d7S?9cMJXD7_rvnxwKX*{c0vtG4HsX+ENsl1T1uVqhZ zPk|%MMD|uy;3272Y%Fg}J=7%~Jjzy8GWsZ6)i_BZcdIIxH)g9UJT46p-d0t5@E|6X z+M!qqLO(qfcdLqtt{|t)rxJ1R>6H<3w<;dn&CKF$RfTvWEXI0xYFmleMpr3)GOIRg ztKtFP%*{U11r2TP-5!i?P>mg(^1kPUgPTZXXC6;uf~jAw4hI23=r|q}Oad~w=!Dt_ zewC9AZX%W3#;D{vdf*$V=aTmW~1g zxY>5C7Y%M=6x{Vjo(Bunox4i~cX8eM#_?cNciuG32bhAQ!EFYGwJvT@JZ^iQsZBnp z={BGt2Xoai%0l(TiU!SvqB=lM!_#TF7_6gg$)FW22g#srU|@~(&TRY0O9nUg%O8{s zqE`#1GjVURxsC!dHc$NE4Co5yzXNDlcNlPG`XqK9e&4e5?NLB@_!({*Z_Ym+pljK` zvyzv1yE%l5Bc)P{THStgO^A&SZ31srOIlNel3F z3NbnUI+cl`Jm!Pw=1j;%GfBJ#{_@=*-7m zifAq}1THh>)c#UL>bqId-{QhmP+Ip}+@tSAb0`#zvv4Wm7Lp$QSoF{Q!wnZRoMdq! zdnuxRUesK0J3yYp&k)}XivD&9-_f&a$W8SFiH~!lKXY578N5XFN4<9TNJ`nC??Or&PfE>A>)=`v~)n{QPrGMYe`APKK=(QJ6{Y7YV6@tVF%VHF8^~{hj;oTylF1 za!XVTKPo2m(?waNrgnvQ$W=k-k%FT3!AsH&`R02Q?@6UoG$woIf4W;zP#;b^1s zFexR6CKCpd+cFjW7E!Y30|$w2rbTj+c4nEhj>$1S1`F1w+>BIEUU&|zEmPqS^QoS1 z7=WKiy0^T(H%Fb^mZ@ZI0@JweKF+*qn@(+*o10!BBkB6&Xq(&^%ZUN`n_0>5Xfu;E zm#}G7Spa`CE0{QLGb=nW4HjCDL0TLRvOs@+lpK(3t7`9=BQ~>`@e;R>_`6SY4KT~| z@ix6ZxnXW;XgLjY3(w(hW`(eQN}E}Y05?Z-#AcSEVCH7FKgMRZC(JX?#DK0YUqIm!|f3dZSQr9Gwt5vEVk+#%+x@u7ePFsbZ}BYjF`(hi_fa zxGgB+?eGUGcB8Z6FvyYGZEMtSq#lF$AX{MEw8hU81>}OdP_myIcMXS+C?EvOrwJV$ zJGtQ23Bd;3Yi?%p=(y_zeY`|PL_BajQ~zKkFH)yq5gU%%%`nyn$%KjsZ;+iu|bmFY3K1>H4jm?n>& ztoNXDJqEX~yM}){t+QS<%}mzSw!xdLjX5YL7Z+Bp>;;26&D!mY^gQpgR{OZMUkrm_ zQ2&Jg+4d7%k3rK-liXu4^Mo@2blGUf47?OZw@P$kI#Q28_yz-iKT_j4Zf`>Pf>Dn_ z-P^@k$~@h87?0mGESGcerGv1?pzZ@R(r~|?Eb}ukXp|slIR6Z+lY3H+!68hb>Z~dt z^aUS%uKqQ^Kw(7gv~9$&E5PJ*~;wj%?!50%TcX31wI-*Dwbh(kqhZ3HL9VI!jp~+r?G3MzVCEXG>srZ?pX2a{Zsh;2Te$NCmJX2Iked4ffESE%Zp2gxk&Zeg^cd8G&<~EZ)3C?; zmk6PQR$_n@*p1{W8$%5k_WLlXP41#gY%*1&=nwu*cv!w4@`|Nrs=c$O9U$IQE0V@X zsts97@&iRCHmvG4{_wi`WkEBoa^sj`b1FY9eu@i+7)QVHUlX-s4@kt@WP>cjW;*H$ za>6(}Z+;mlr0D8jPSne652OT+@5@ukd693QAjy^;s50|wj)eMf(!6WBzM7_3j_>wk z&Be9@#%y;5PM2!V+$mjhwB;mf4eI^JP|9( z#^Tf7?qYH*Ys-I+?Ji@<>~(j~-|h-=CVGxyf11TM-BH_JCSK&b{W#lQkAQIpw0?Ag zlu>B)`>xmFE)x5h$Dh8)KhGW1Hy*gOcX3Z#fn<+vR>q@!ze98R27re~u}7CMYE0F9 za~}!rn&ibSs&wDE4tDx_j*f?Uf1m<9I)M(u94YX2MuEp3=Cta)X&(?VYNW?t(SWD5 zf6V(3$kXzqn+~L%h;irSX#+$ymp^$R-SvV_#5jwJ;CE{)V%&AL+jJsE?O2GWXI*dP z`1A~v!_%i>&ZnJ-aTb-sYwkYqY4bCF0=1_H?0Kla3M7zv=4#?+U03*OldldR13;$` zy2-hS@prB6&$M|wtUUhdr!6^IDN*ZRNC#3svqI`1W7LgT?W?ps?FYs8J>1!^kwC^c z14z9+8-IJ~9ALiY7?jWTg);(th-m-zKY%)W-P6}0?|J|JX1PB2Ui5~?o6oD=bi>4( z73keGEk7(>Q#vE|T?KT%UuP}*wBsQE?4s~%K2k&vk9mEefla1U;vIaMH*5fXyjf<@&B! zB3B#sg2WN)kuT`B%ACbMKC6G8G_R_z8k#7WxSkG+s-!PatNBYu0VFtRD);tMa!3Bs z5o4?WK1;`b#L|(0DrP|Mxl2bOw$3Djs=TFR6D8lx!zZB;OGn10xSl@7(m@kIt}al` zV1}NmAGdV|r4y&y`gwre9G>2+wnP_ZqUOkAgVhBp{>zY4)f_8<)X$5|G2*`t6a_L4 zEIhpgas~?{wetq1pkE>c2>KlF1ys#P2T*=cp_*@{+wQeXXe)!7(kIj$AcEBSUw8Ga z76@XG_dazZ$khc}UruLERs^Y^7X^DR>a%L+xtq5}NAd~(XKR~F73W^{X8DG*Q*S%0X*yQ_lC)kJ>j7 z+b4eR&XdcBUwG@`)%jO0AK$TV&huB7;9&Fi>UkOl zD*jN!NPx_9(j4NnN077!+F6(S;6%MT!uFJvjPYah4$Hon%jZzYJj7$IKQU5T0!b7A z7NaiJqK)x99;p7av~YNiTxJTn?nki9&_T<{WhUk+h1zl1(b?yhYLB43BbS+stBeTt z{g`E@5Od`%6rWKT%@rQTk;_cRS;lxi9HoVgM7qcpo3I z{8SHviw|n~s3$Lrn7n|lPObPF2nHxW&3(imFLC+l!jTtKeu}+NnE#NxEHm=r%1>uN z^?^NCdGZ2PD?tiFg&^OwMep)?0-LVtd^3lSaBLjsA6}KaHW=~=^KLv?Vp0@?pMz-s#h@4hugzcWlY=v#F z3A(X~h}!;76ZM7Ofd1;8tE)Da`JxZlH`Hs>S6BfHe~FA<&^czBJh$3~W=H_OMg(TW zZEVApxJs}!?gcj9qauN;4+r_oG&vLgXE7 z4$18E6E$+h$w4r@-;tT|d^G;3z?5}gcHSM;u>HJV}@TGQ&R zS4p)1u9;Dn6s<4t1Vi7B>$3P(ORe%7SYETrQ;NV1INi3&zgA{4`0(pIHkcdAsm5&3 zj@Zoc8YAIjcCE|^o8Kmj57vkBsxj6(W&@W7VN+#Jb=TBSChE(@Jo4Uy$CnSUF7MvE z^Z0UHXEAd3wAN2z&B8u@?Ha@xXbUs-otl2A^#-|eD8AHG&eVKkpm!7d@TAfUzu!FD zTIEVt_Y?-eb>A|ezV1vvnOH^fp!Z^w=wosRWds21}3#O zjOzr(-Kq-ajoGRSk4r;@w^fxMoOJ(^8;EQpYVU{attv*z1sp7A7IE)^$G-`|a_&}@ zVPw`CTDv)-v49uiDbFjxrz*KoD1W0Ot=g=uD&xgGH_LBT?XF%o15pK%zWhW_FQI(T zN%~B!7vNz=(#P^{@1>-`Bq2qgy_9ris=Rr@`ZLXKFP)(*b>?J!Ga&10 zKg<(PpzVuJ>~z)U0T+iMj+`N0^{~> z8FcM3}(P` z*j?Ho;4UPc+p^XmUtAAOu=GpyJkMT8x-#Fr%mNBJdCj|4vc2W(`n?Qi=lH=IplPrB zEJrwi{T&3ap|w-_W{&E)$l~>NehdP5oqSnG^=G(gyvbSvUJiw~tkaCA>i9*ZQir}8 zPH)G0WOOLRWf|>ZI{4lvWnqn?H!UngoBH1q^=s5M1h#X^;8U7tHy3oJoyNhAGq!CyKo5fIyCDIr5FK$i}6t$>>BX ztBa=9M7@7PN4LP%l#g@#bt)6sVu4Ub3ax>xLUDb)`Ia+t6D2Iiv*Fky6;_xRm^)2@ z9UXV4*?(j^P07@Vev>kFnX#+^a>d(e3a0UQn!>}wf+jy@1GFvW^Zs2(@phVwXmdP^4!QI;wD%x;4y3`P zwKG40A8OMzf6=Vs_R4564YW91hOGZdLr(VIO>;>w1?YS8x9l%TH}Xqhijntrl%xYv zgKqg=#2S^P&+U!3E21_tVEj2M)TNiC`vHxFy5{Me0Zo}T>Jp%B)`+2B+e~%6Bh_WL&&6|u$yGr!=j_7cwZ8pr@3|^vb23LHb{(8$YS3R2zbDeeNEkM$& zX}Tw_=nwuS#E_v)+^2ee*rL`=zAXdP`jUrKE47%;x*=5sw4%moeh?UID6ePiub!~t zM#`K$YJ8P$YpAYg*y_P5+OFaNPvV>#jJAv;Di~IMzjpht)xM9q@c309s9Uw3F6)6> z|FN1NS*y6NUGEVj|H-7H*scF|i=|4Km+x|Pj{5ytfpU*YHnT0?HQe){EUW!&apCKF z5&N*m2h~rMPOQtVHIX8eV3p&B~BI^7w-?}i_^~9W) z@NObRx2>{ z3)cu@0UALxW7C*?Wj^^V(3UxL3GX8-;Wcyr5GDTUr_Yfy!-IIY`R?9z+6e1sTRITL zyE>Tc3;2JV;Akh_HIZqI0OH-nqW$;&r_|dPBF79`B0ExmIbe`cgYCq0BB5+{a(*!U0_ z?lRazV=S9afa(3&M3rwskH7@l0wt$-0+Pv z7g2%h!u3Ved5BwjF>TC^FqEi1iq22`P)7Qwh0#ZK4P}Rz?1lnjGA$tK2PaggJJ_^9 zI|I;}iMvpC5j6%dPY=w5`fg54W}fxR-Mf~OEc898!_=Z_jczC}CfmlOq;aRDi#8Kb z10)f?PJRMN`CRE-JvaSiN3JBYzg{yZu#oeIBIK@J(`Wok znU@Od?%1Q~Rcv1pmWD zsqJ*#jK|{RvHZOw6aM>&D(gZq(^}cSjDeDwhx+zOrEU+iGtK(@J#{J(H#tuXiTIg} zulwnM`r8REk}5`R+8VImqhZ~Wu^ip>I1|5+@kvg6dg6~Cs5uiqCIb@w&oAIO8^QGo ze1SFbGeXcEru+VhUx=HdPy7nf#v_L_@iSu3?@{!nWLJeylC}Wy1^QRhj3}sYI&u)n z7tGU7QKe~p5v92!4m#)oIGm&FuTK`Eg0zKekF}sCojxH(MFezu z;D<935iE>|pnEvE1VNW&^b-WCQd2&#b~2$!QR;&QZ|BM8)TGy}{2ug4GH(QWnjw`Ug@8A5RNXp0Vy zUm5PXghB%N6U4>nf%@2^cfN9XrvpSlzXc~t9{>qUASFOYI8qUW4pCTd3LeyfdL=VkhU^ zgxaYqCts&UCA{0Sju^0MarTA}*XPx4y6Fg4_~T>%VLBP0Og&6F`1}@R5VAs)GW_X@ zD&O~QZPv$3X%g@D4rhp^<99Z$Bdec@Pn5Du*ynj8* zAI$IgXh5A6^M_f=AC|7fV>w43L_R5NPM5 z(2PO-1x(XJ9~D$Qkz_-TK+N)85fF$MeM+}f^XWxzO23VjY9f z;4af90ux1!Beik{YzZ{^Yp^twZ1h{|D1@Hl5~m_*=j_$LPSmd#=EVB^&f~|I4<6rp z>%sY>%lp4|{*8AoA3iz%`sKq1m-o+Wa)-2%-7QkTY! zjiXysd{NHusJ(%tx%v7|@0Ifh*VUV1Cfarf#|xHI>EXcY-EJg3bR>r?Ym{K_0xq4s zDe)dLF&zw1uN!MclrA7;TJ~QrWYv=9r+7{M=tRB94a1>U&TlnbFP_TH1Aj(2lXO~n zb0CdnT>myv|28S!l}ZV=RI9^v_eV7IMr>vUEOPwKtlFdJDQ;#Z!=uftk~|1Tf31^w zl*B0hW>zq9+-6pIU>Ypk&8+n3p_fm4EpS42(V>yt%wo8k?h|MK4R*Kvh|MemzRcBi z&)v)lu_YQRa(kHHV@GUe842dP`=i;+f*@@P1ZmMpueLJhDZ`;h`nM(yibaYv^Z3&j zxaTv2Nc?EEJrx)+5T`MXG72{J(AgA;)0Re@*25_QPHL{RzDv;3Htn#e(Mi;n+qNIP z%Ys9MbZTLgBca;T2-TRQY?*P>F~O)S0FE_1v7)u34Tz)1DkxM6FTV3>o%Xntrg{PmEE+xcsu)|u~ zVT|r;HH@IGKLmJOf3&i8dfVG>Qm%W@%6>Wo-Gi2(_OSVi>WQQ7`ay=e^L*iKptXe> z;D))wDv;aIgQmp&(a~@aevT6A*F4zVsm*l70-F^u-1f3Yx53mYcqYFPJQUX8 zd($|~>n6<)hKJ%xrr{pZFR>pLbV^N1*mR0%t6XX@UI}E_ily3r2>T3jM!XNT{rhy~ z;6F}O$7%sgj6i7tamv|}C9Wgtm(+JpdT1T6HO#GUHQ3F*?ZZGL`Io-9zKSl=px}$W}ww$a+m{L(uvSPaJ7oG zdMl&V8?pEXx6q6MiaUD#pH)ICCp=t5io2Cj-1RVw4z1IW$3-QND|gx23Z(TkHvkb` zp4`$)9{rd`BD$3k(XqzVawPwm0yXAq03qo#j)9Os+KB#=p9*LX_KXspRU*u54_8*Y zHrz$IGCkOX_OLrPqRPkB%2qzEsIs)6J)FnPsJHY{gG3Qrxx+=K_7NL5 zzc$3$!$n4lT;p{F90LG9_BSe#w?2f`iU=O(s5Zd#KgBeb81tnm(&; z+A zs!B#5Wve2|39H!~cdIIxH)g9UJT46p-d0t5aJRrvYQ5R()^FbHyE*9 zErav zoHI}FqP~Ecv~6%V$r+`Yslc@m-Q6U<{2wYM8u3)<6cAk1?J9`Y~QdIym!q-WaC< zxa}V(B-9EQ1_kaK4u^u*s6330f}N0X?c`yHZzSgF00nowp%0j-7JaSjtL>vZmKPO- zgo_9Yue!92+G&T+7b~HvbSESgv(UtVWwhhzdcl-3jD4=f z+-~^t8&tjwHD*=W;Ai~jpcmAw>6ZLC$b=2}0kSR-{rshVI8mgJYss%iaI+xbkuox4 zl_P{k>irXX`T{(i;!KXePGw>U9m_n70q2w&kRs(YO+U7rNs5t4q||GwB?|{gP#6!kRc;R(C#_QK_P~o!i=wNs^}!)Ee9DT;&@hi zIoP91ZU6^BKbI#nbU37wnIRjE^mA*YpJOku!6oRR`3xS_z|ok z&`9Oxh3C-PHx>Rcp9iABRD@#ffXa7<0PM)^n@ZLuSdHs$f(8C#nP5bcdlBZQ7Z^#p zJ~{3t_rh`lLH=e|GW!BTSfr6tF#WM{H&> z?Ims>@I=Q{zzfb;axcQt&~h5<7M{c1%nD(PZ6tylU5y2|zZ_E=>lWO^YOGs$&M}UW zd%`^PJQ3B}(Fw8dQn!x8c58B%y4*KbK-e5c??TV}HmF{X&WfFYa1+Vp%+tGAF`!CK zV|*_{Q>zMqv1=^1>1r(JMtOtE-{{oXkMbtcznRC=dJ*QMY|(MkF=3RUgisg6GWc=V zfOsDQVOW$ZRm#Q-$+aQP%1H<}wkO+AO};V0umF*}df<5pVN_DSnOTy&nYY)2;N2i0 zT*WvU)Ri{}%gn@q%nB8kdkJAuT)uKlIl5T4{eF>zaGil=S7EN@y8dL;eyBG3AgKiJl@k#fsjyjx0EpM0DD0c`1eOyKZtOQi zAR=6Y8@-yZ7v7svWKT=v}270fk7)k z%YQtJr^*V~@_llVir?JW+TQF{$!*56zf2QIo`;WM68b@!+>0=^w}1y6@|0R6q{N2Y zMIr`JQHszC)MG6Y6lo%Rk;o*6tPm#lA|%8U8o2rH8$N1?Ee916#a<*b=^@X<@{0sf zez*aeJ$fRPK_`<49ps0bNV8|2PAgkXgC{Lb#inO~N~rRV4wak$aT6);Hb!~ZBR~Z0 z+tJr|GBZ=L78$e(yH4 zOuJu#xQVK?H>OHE&?D;0`Pi_9nRb@Bdrvl0aaxcdt|Ml&1{IiSkUCtnUdZ+$GzE($ z=v~F>+TtSfJNS0tu#cYpfP8FHQnc2P?Et#gpy(Gh)kzRHmXoIvM35k(&(UmpD5cc% zhg?btCJhiFZUE&%e-rj1oQ2MHq$`XXQ|_F4T7K}3VSckItxaFr-y=3hdT7E$E251_ErvC=(JYKCQ8&r(7z&Se5B-v@}1gYPu4GG^`iR2>*|*U&9rKc zV}{MC{IGP8`7_y3$s0iZ*F^2u0}_$9vIbB<#3Ac~RD7cT5KgH+(C-t08@Kq31*=xK01MRvc_#B;@cV#Uk`x>q^da!ysU8@ z)xmFFo4ajqgSx}_fGv>ZRkp{13NK^SV4NdOo_RcfoW}_pHCL|RL+V~N=)il0ishZKah(*)Zyc#M z#qu3%MI>z8X1Ll_%&);Gg7)Nf!p1GyP9zgBRvWYauI-Jub?rp1-noB!WuAW8l3A7V zwchpAOB=1k8UTLXr`5hokJFvBacjR%LTO`H!0V4<1Ju`FcMR0$dc#@K71}=nMfmg3 zM{nN0(K*7uJ$u80(C5`|x?zG)xwBXJz+iG?-*aFdIx4WwJs$T1MXBWZuEzJ9)3-q5 zW8*`^1-9{?{o}nQB>hq2|4n}=4F=xq32LBd%PV#r17WCHqqPkDh>+Ov#}n1HetWNBn@zDQ4yRu311=8}>Z5!X6~)@G@z|+?uZ-6ZOxN z<~7$uEou1=kRwC~}MH z2s7~`mP8F%aHJzNPj8lD))v9x#tw8RaH(j3Qm7t~4yBwdavSLZw?+@x1EoNfB4&s3 z(zmE)a;r<*2B%WH?>T+n7S;D<$f-T5k-pD7ojacnJ!qTg`N{e?ksChg^@era~?Z}qQ}lKl$Lfc zUkrgXatA6?F*DTPL7EV{`n4)UvmbZpWLIGx*h}Qc^ng1!g{IldDSa*s%p}ymGIwE5 zOvGy=vp2U_S}>d%+e|Nnt-OKImCY zYBI);&HFF=VvaG{^A!{%M#95rbRtAiU@_`aecBk$;|qk_^3S9_Fm9PCWW8>gi9Y(H zTc)f@D1TAomYJBWNWZt_no^=Gu#5=y{g`E@5OY7>%glb{GLsRPF`oB%rtnHa zBriKuKSL)kzvpB@yGTG|9)CJ|MGrp0)-8`;c0dA?tC;%n%PvyDn5TE)7jUXYU*9Fz zrjMA(Lx%)==UP_nItm95yk-Y9EV&!#_$3T&Bw{g-A2YOU2BS-|7z~D*Dmw zT*zU7k_8>Ce}IG{bZ=U6QXI=-X7~k|pYeRPR)GMRBpU7QCc|WYO;;zA_z8bmW6!K=rYJ}}T5e!`+GIqs@X^rG+P@pT)3-f;L zij(o0F)_a%wc->a51sB$b2+N){bN_0jMP%lO{qyGSszGlH4%>pLD( zE}IM32B>D}8Fz+rlvB;@BGt^UQ#GSA9BlFT5#UubriU}YHcmCO1H}&51A`lcR5M|0 zBiYQ($Y#cjt>rdu_)srF?TpTQ8Qg|E44&HvYiD+mc4il8XY|X_+~Cn(RJ65o+L;{{ zPI+#lo%zg{zVYy}gBes5;JJ-Zsg}Ff)HLn`+-7#+xDD@KQ+=Cf%57Yg7P&c})y}Lh z%G#M3sIElkb!br^OjKFw*)puf2IXEl!vH$q z@;gm$!a3Exl(3A zngBqan~yPzNJg{Pp#pYE+dw%0C^fQZ=dh=;Rh5j6x`xS_`cZ6E1@p#iRfWf;A;R0L zN)H~og!v3sZXmLasJ$Pux2hN_%h_<8eW?Jd*;`cxlv!(N?XHQ&0$zxxJheN*R#m~n zgaG4iRT(+vxtX4Iw>;mr)o+UqDy}bp%#Y4iGx?u-&&vGt1E<%ulIiqtQ{&ohmp!v%W6PdQJT6bx$GhJUm&7b}! zsQF_C3@pvnL*fEv*<-H1&Q$w0qR5~Wv^_>1B@16?YA%6*D0${kvboN*`C4I6Vb`zO zJmbmkI#YPjL-Xk-Qm%AguJr}D#V3yxde<)vm+kZH;^Uuwy6B`~CZY1yw;km#Gc|o2 zjlY4UQ8$yb@6q!>+wp6CVZSAUj$d0qyxUIxrX%ipMK;8p;|XU$M^85m4IqCG9iYm0 zc~mzx2ikGP9QiJf>hDzOd)vSIBor`vsLN%YYdl}aZ!VEKmsi+1A>1RQLzyniXphym z+%d?{%HcDJ+(}XCyLo?&x{QD(LIt{fKytV5Ji7f@-vRR0L&qsVQ5iNlPTsqz{(Mro;R1hP!PJOONWFKy zNzgI+k5AMaBBN-@#BoBR(pj44{;=;z4M5H$kgvWs)ek3%yzPMajJSE)v@LJfIDeWAJcv3Ht+Y5bk0@US!_I6F=0u>=2> z3aq{j#&2-UohHWK`Xga_U9yx}>27g0Vpn7CG#Th-tcQEfPE!aPp`E6Nv7hozld)}% zXP@d$vk$Ni$c6LMb%w61uudbG&mWKulnbvfj9i$#cl+6m`X-oyN`5nibJY~9d|qEb z#jpM+ZAwSKz}XyqYj2x!s#bGnhF3mo&p9Xwa3U7h!40T%mtL^$M-)nT*B3^*%NS8h zuQ3aPOF*wc%^W%48lZpdC%w~V;9i$}8qk@wty5o{@?T!E>SRk*YXAitYdA3OT^w!m>Z)tVvH_dE3KwAy<4+EI=3Y;y7@AbB$ui8XIUuRtj43IhN zFC18q_)t=R1$kuXCikhHA9m1nfs(bCJcU}R#dOw9p(?l!Gt->!0&NY|_l%v^u~vOC z5cZ3x2V{NT>`hKl^xyK}%#PGMtN|i`r{L|?zA?AsJ-5*l=9Wxn@>&I7$~pSq@hbN zs`{IWD)*>CtV_sguRCRZ(X^Vei4>Sbx_&rh3MLOmG*1NPepP2|cN^8eiIfbDfa&;| zJr0)kZz2Vgc$-M!L1_?FDiEqur}W5S6Or(9wliQ~?31mDv5CZVw7gnJi5=l}j0k5L3%6HpMUh@}(J-?l_l z6D){=j3Sn`ABRn~y&*bv?+g{Q(vhfB_iDldZ82Kb9wZ7NOxp&h?lo6=1}-|YSG2A* ziSfhVewcR%s!gTCQ9rnmHg#sSshxw{A?vm0V<_v@y0iW+p4xRyk9p^Yc>%O~jTa{Ir3ojUL!3+E<4z z>wV-{tZS>|pMLrrIYA@~daV)e3SVdz*se2sd9D4+)uw%6xDG+h1h z;%u}l4C(3`tU@v^HJPRQe)dl07{C0K&D{cW>7C3OaOn!gecxm*#K%!5b8QGPo6e5EzqT+dL279FkV#9+wmwh@jn3N6NBNL$lF9*pX%=uE}Q%x00kYG(9RJxrz1 zk%Y?YA`cNs1DI#(w!n_DcL3d{gz zLf`9uI2xd#m!SclB&I$%ITJr)2mKxmCuoVeO=JjjRMJ3a z2)aGafg#KzNrQR%DXKoLFQT+{&~t=2Dn;n$2=hpaV4mK^5mv<4cR8cEp*C=Yx$B`d zr{#?JVH^5YrF3g$*u&^?}(C19JakNxO!4qbM`UQ!F^vJCWN`RCa?tR%lWv7iq+;T#}a^&0P`y-R$E0ZL0{a9{aqLj?WKVw2p6mE zl47-$V+qVlRo5v?Fn1K&Hz%6vSpt+60KvkV_?!ohCFnpfECKyCoGgHlCv1qn_dd+N zLl>ztj3>lD>zD$Fs!#hIZlPw=^UfTOk!0s*fon>i;R06(MR?tNt78khS>acFtcx7T zD!Ku5C}TTAIbOG&jxn@mHzu8sIeZlJE%lcZ^*Svo;q9Y!(C~&2(&yD~y6Ff~_UHHcS#KnL02=M zL;@K@w2RX-O2a7JJ)vuTvmPi7nfJ;Vz(F2;eV=AVRUvaHW0-e*aXw_b3n3YU8&G}W z9&m$sI%7cD{6U|@8T^6jdlxzjpk<&(OPxO~A}xb?I^Yk~m-B%U6@e_AmSKT*`v?{M zfIqBb{y+sGo<9hcez;~i=5k@EWmr4@z`PuElko>v^|9;l`ITwk62sEp_@V45B}te_f-BpY%BVsrMQ&*_$GKE3G8>14=*U;RzD)PNvD z%|`zg>lp-Iy&-a(WkHC@(on+DZ>gsc4G4}=2oX0_j$rCvC+gP=b7FIT=dtc2{`lTo z56&N5-v6!hZ@hE)@X7htFCRX*ynkL3J)|}6E-(2dAIC4N7pE`O;+XQIwXQk3V@_Y! z9NoO)i*klXjSonN9!a0%vcE$YojFiC3_lSu}zGcR?(k=|vMNu##rBe1^&r^2wqZ9QaHw<5S^yuDKA53oy|6M%rQoML7 zHxDYCD#t!tQpFB8Xpdr8dpeKU%nDefwwZwmfp_hZ8`Ns| z9=q$ACK}Auvt)R*nMqnv-lu5tbpB>mFmc>wR(N0PkVuJ1sQ^?$zF3& zb??F5%Xe=-e)9I^^yK3Dh|Mg9yI>x*FN`i$iAshhaXTEbnPtG2xw`JTn^_^YL_?+g zou?hQS?&`)L5CVzqpm z%tUFmrmQNAa->#U7_}O6lr1y%+e_O+u#b>tU<`1)X}ivQIrvaWz2>e5T}0}&MXX-q zg-lc;xo{G;1;Cf37fY<6yG69<(Ed zqCP^~j3~S7L3g{(NExs)xSJw6htNG>nhmMnv3-OVRwg%eDO#AjDM}X?;yKu1E$#4- z?I*Onu%(DVo~J)rSvz}N?)nKW?5D%fPe}hp&x7;@50BpYimr;(4>Hu9=L=^;S6J&C z6+NR~g`PAe?vIX!gYet4*F3!3sm*l7f|qj808|=kbQ??!wUf+qj%V_er8hgWh1U{1 zQdZEtN%Mo@p~RADxJUGOv880P!8|FOG{4k(yb{Q;HA}T0QMCGDpVFwep0-~G(-^9M zoT!e~0+<+qQo2<+uM{*iODeNgE9$!^J+uzkn@R|IUMqGXZVzEDLq$bI#M_12zgt;H zvA8u6b&HLAOGyj4i$n~djyUy&1%}*3B12BB z5Z?0^i9$R*l|`cRr5hIGVwetzeo^OhYO(9%drE!=20g?gXk^N>Y1mXj#zwy zThJBHk+LPKxXa*Cm7%!)NxG*cP~4fPcR{5s8gkw?*FR}_5!F2|++}Rb3*vLU8-R!| zA5%I?a$~CdvOqT&5gltxQ;vj=>DN>!6x3292125`-@6U%&Yo!utI^{w>eJ0%pxaD$ zc3_epespj{Y(%w>OI^=*m~KmIpPkT0X$zUr@`9>sv$$w^0cs!f;L5O2)k{`;sGk>W z54Qs@I{XgNM^PW8T|$cVtQu+`^%IHDbG464%gNI|N;`DO{ErwmJx~z6CK?J-x@8SJ z@&ftc61^M{XiUEgQ{&K`=NJ19zz%)@$@43of^5`ky6B`J=2Xo%yXWVHYvwo~lK)WQ z$=Y`<^UyI;_$5D(f(c{oNnV3ixk3XxOUVd(faAsBhNX5#}QRm%pTPSu)Un9P!z_Vtzv#|*9M_>8Xn z8}x|kdqlpLS+*fu_(kN7AhhoEFR&7#FkXy=uY}r%B@74Juo33SV zRRuDV>c_^Cx73`Lf)K9bsH%~@Rh5iB%2qY*(m?K3RWNVNR#kXh8X~-{s`TJNVCcU* zORWk8@nCyDRNSp9X1mb`pyo_I=2lfe!NOY=kMd?K1aGS<#FPADTtm8~flYQl_EyD1 zy_uWaQ{V{x^Af2g(B0)HdOGiY&xr|_k?_tuo(2a~zg!(^4}$5=^03^4bQJ7_gi9w68=$R`hXK31x$ULFX-}6IQ7!s% z?RYRnyyG@{6@-Ml7z+1>osq*XFF;j#wEfaGbp{{nd3})A>gUB|c<0ElsY-V?5L^a@ zg!(FokNxW9o%^rf^C--grt6>Ud8m$*gTQWycI}{FNT?Na_IdWBfqyy;8-wN4Bvd`7 zcp0IUGY1)=xvs&!N88W-CzX>CF73BKC?o7r&3@AYbaR0R8K5Vga29mQM0M+#z?ef1 zIub<>zi%n}`~ro`T(BN~hl>WxNhp{ZdKHk+a`<^$4eq|OEV-|PdPyoLxr&VrO$=B@ z`$Yu5x=`xn*~eg$w7)^+%TNJTl?{Hze-3&<-I{L6pMy-;fFHo5fDlPc&`)mahZ9Bm zxR(5S1UIED%D`l%`J)?F$V=7xCv?&VJe}f9j=xT2Vkn>Jq2TLGY8Z+im+&akO+U7r zNs3O+*spDlXlQ01y2DCjx*ai=gNzb!JnLtwm~SvIc#^V0^>cJMq?4J?o911*{st(Z z%+pU99<^i$v%vc+n1X8H=upT>0auX*Ze=uZ9Z*Q80JPV)t>%^&QJvt*wVGQ|ug!BH zV})wk=qSjKXryT~Pv?xN?cF)pas`-`s0LvtEk?3W$LXx z{ppVkTV82r*xZ7j6;<5kg@D?b&zQRlUET&jS<6cY+%uECP;oi4RgAW7RI`V}WefWM z*SH6JBR{%&00Fz0Wu(VmsHmSve4Y~mt}NqDgn;BrEZ=q1K~S$DEmAp+>2W;=7JFf`{o&n?!W#*?wr zdI2C_`zLeqtCTaM!5G^biaZ&%dRVa&J5}^2S$XaU#+BNOaDV6iJC{$p7h&N?dnA6E zbW}x$)i z&ywLs+RVm@1o@j;!NhT!S>b_cu+VxD(&BK~%p}8(v4@3>kJ?zbU|h7BH8+mJml(n2 zldKVpw!0CVS;l{vn;EeeVZr@{%}g??%9#mim=nEV2sUbC-IBG*HzKaPMPsh*4Yx4{ zG=^&J=!DpJsar*2yA`=hUG8JPfN%x$a&)ay#*lq4!c`=fGf(eg#enfPN672JW>M|< z$~Bf-tvilCVU$t%8=V@5QI7O)E2DoS^&-qi*`i|?m~HEzgmA9;Nxy^;V!rl(c=8T} zVf3KN;IeWO!je)y8f3^M8AR*Mgg*bUO zZ~w9X9TRv1Qx+;NcM`%`TwFe%Ii?(4tlJ*cNJ2PI!Sc*in9l%&Z$H#Y2v=we4!VTU z^&(uk790oZo_RX0|6VlNxFW)t6*2=5uRl@PH}M1Bjtws&)O`2w4H1Y4XI%|e^{1;t z75-d@yMgZ{p;zV$Yr`6ROKOiW>_xZ^ed2N2OyAoHY5?nO9698{uJMTK0Q zWcCw6vO;Q`vF!Jw+)~b}jEM?M3#o0!GJ8`r+Ofuhz@QbNHBgS zh%31lVfG>^4dr<_EP--CiTQ%PNEE1+L!MG^PS+FSA85IYL{n}kXQ<@5L2g)@hH!?Q zSRqX8MVSA1zDQ&nY?!BF3kXM)$X+Be=^@XL-AbXD;)~IHW4wqctx^<-4Gf$_L zEvCVfk|lJ-bNT-oRo-RrsM;sZ2@rMu0JN0m>0M9>qz>roI|1S>sxMr-esgQ;bG#dX zmX28zXn2!pw11!zI=6ssFj_iNDQY^VYzd-H;}{63+ijpsdkj5t62x^>rM)&)+JPQX zTLAjlu!Wg!L@YR3gR~rtu^>U5$IJ*+V4@H3xM*|!#2pLhUdnnDD%Xo}W-B0_1aalLm+o z*PuEW{Y}`5P`|&s!;w@)8GhQqj|G|$bE_!B?=X(~Q7^)Es1s%FyPkQBxG7lb$XHP- z#`bm=j0R5+1*0BytkHp$GKqB(`;UqGjfP})=p%T3_wvCbwj}X$USIpFdU^UHMPFP0 zSus(fE`t6QN#i3WhpfqYj|RFh@ei-7UlugesyU7sHmCB#LV&*hlbh_&W%3X8UlX-s z4@gAb3KB)xm(LPw8Ij_@oT!)C9!Oan-;_j%%Y9WQ!&XFc>Z8Ul8 zaULgZ)QTGhm0l25DYsLG*@DW|G67@t!-f+!uAN+M2r?sA3uu}bG*OBCT9M~9`R z9W=f-;(QT9YEUQNeu^X2bD?5+Cu}qw@m{%m_tcT_oL>j!v)2)bgQ1WP%m zq$jeB6BkITADr|NKVWl;S-HMzR)g1uJ&!FEu}ZqRm`|dzA@$Fb<~7$a){{wIeY~At@ zWdrnmV^EYI)kxoGp3WWBltZEBd@WqhG1s+E3ZgJUA1#>|dxr~_8oO=7O$4$#1jLgF;6F^oOofAgt!Hsa!(%~lkJ7H}^ zLnZ3XxJpUWTo;KRgpu3f)?&XUikVbB(3f8{^G(lq^ zmr%DVL$e=u=mfuifKe=({SxDZp!B)WGZVe!*_3HHH4x(55At6%o?-TVr4G0zABBCYI7IqSlW_TwG0rkO&dUb^DDK#16$L9T) zeKD6$Fp$qk4-#zqaSDdUcTgMfsQ)}suZdYOlQEl7-v4M;oX|vUE2Nk4z00?eV8%TDbo}~G^OBc~Y=LSf&n5e@ z%{Efan5TDP8){zXu}xIPwRLw^ZhKWTTOrj<7~4oTV;(-d*ezO}kWQ-<5LI`LI#HtaVtklJaKQyFQqx zvedKnvk@EAA)pmfZ(mM7eoIZSKl|Pd^B*^1lurmRr8kyDXC>&RG<{d!JE_3YmgT8< zi|4;em;dIEAE|+N>vzlhaeGkBs$D)0wSF;GYVq9cBV90*T6MHH*;`csyQFQP8~{W% z4KY*Eu43snrnp;G$>^wSnABD^?y^(vR#h->%vM!+TpA+0t*Z3kp-WgZ-sLA*p@*2_ zZdEZ<{;A)pa$z8QtIB{fYYna4HPKkW3-N^ETfs%E`~j;O#q6yrBgZ^9A7d7^jrM9L zA(`KMPUg3bWPV#C^BZvxXrDB^`vNW&Qu#f!MV8x0<+trs`2jgmdstlOuZ8WlGt|;; zoz!m&q<-y(`7svK`-L%%^nP2T_hXH5N{rhkU4fEI&Ak}dxNA6^2X{i6KiBnZ7is>? z(}Cu%>kWOtM5Vb~C;Zz1DBAXJ==!yZDDt9xohisIJxT`GnMNuSdgZgx3F>)06h|&> zeW~mEwQ(r9>pIh$D(d>RO?mRdU1z#z_lJv(J``l`&|7@y*O@-CUB7l#T==v8>6ObX z>f5pb%zl+@FR03O{94$uS3uu2Yq#&w^FY_}YiGYDf{tGcQ1IH0&u==Vq+XE?aW7xv zfqTMP0Lm!aKdd{Mc>b*Qx~`P7e3wV{_Z?_&_q@89ADIFXJtN=tQGJtgT5nE5!OT#X z%R1LMzP1+bur zP=PKV0O0LAk8VHKcYwV0u+1qZXP@!1$rseE>9#JLESL{o=IrOx&m68*nSSEck3V~T z`pop?*~zVw7(iZEe?F<)aDhL)U}{7sq>LDY(SLlR-Vhl@OD2vJ5|z%kxHl^Ow^V$a}#we$FuS?2#a9F+i40EDaD;`r;!<`J#9BZ zKH5&0J59;dBkVNeC^&DYDVWCJX$lWZLxQu@lpZ_qZ!D~(Cs7G}^LCmTd;b*dG&$H7 z+i5b;%~%ijoSmi+H2HyFd^2H4Yi=@knv898Jd1C>zWhRH0r&iOhsuTNmFpV89E790 zNG@!ie#+jv!@7a3F2NL3^4qU`-bISvol*SuTw<{$zP?jFFQS6=-NIHQ@6fh)F;p8y z0G0023)cOJM$+A#k?t}^G^N+*h{8H`EnsATYpU}-vqtZA>914YS-I@b4ntzDbqP0B zY-w}d)JzXDIuRHF*oIJAZr5w25TBz5Xr2dSx-DBskR9}w4F z#;0AK`T~3>Kts>zop<)jT-qy;CK11>-gfj=fBiA^b=H-@0GacS`sRNHd1UA&_o<#A zcF=W~Z_og#zT}bAN-d_dZX{I!ap<_{^If2=q57V&(|Xp5H$s%zXlwA5ZfmH$XV~fy z>zd(nGzMhcvsd4*-5hKs@1rhDe$@x+R;{PYdZ5;StR_;{%C2kIdqm2AGO4I|>%ZNi zs!_gD`%Wm~gEDqi->;P~_n2gJ-10reJrBzA+p=i7>ybll>kr=&zRU?K@*(EtK`5+r z>k-`q^Wo$3%Ll*t*4@j8=l4x-%-fsux-k(#P4!&EW69EE`>z-9s)R#Qw5tC2M6Elm zO(ANpw<@K4b@%2|GhZ!1N-N=~A{)iNUfJOJn~5s-s6nht$Z4-T+iw+4t^%%MY$63F zk**&OnT~(KszmrHY$7E?BVc;GeS-v18JkGKB;F=ccu*Pyl?sIF)G0l3*hHGEyyTQ9 z5z_umB&MTBS^f*WQ)IJZY$6#S?OgTty_-lOR*t-h)R<^9HdX&7l0ng4cgoI(wuv4E zp^8{K5&dmPM7QvU-q4-Xz#oBQ&o`l?qXQl z>$C$gT?%!GaI)TI(jj7JiK+RW9xI`&S9hKG*sosRx&Qh-kBeNJh~-7wWBzB&s&jQ1 zOSG>JUDo?(ko9gZTw5Lg^wa0a38G15{42Jt@q*r^Eh-3jd7%T!zCq6et)O>vVZV$4 zLGKd1Mc1a?jy|YwSfp&agczjU=L7fdppi~<%`qa4KwU#8%wM2--|}SGZRB~+~9n#mc1X`w4f=PAX*VQ9c7iK&k{{y7#vQ%T-X--fwz=&AVg zw4ZbkV3u#VKvcH#Dc8ArZu-fNl&eAoFgGfF7W3J^?l0hKl)Ls%m-{d00;l5Q=^hUS zowUX2M(}2AUnoxQvd^Xba2Ah0D2Nzpz*PC(3Y+#M_x;?td*|x*RW90aDZ=YM_-?7~ zblnWT;^T#K{=c87vRD)|Ez0`~*}RN_?TkLw6roe_+b5N(KFrQE|L^zIsYKi`mYjU? z1KEbi@Tk9?;KHe5)TS*2`#tLSq%#pS&_$0m@e8@1oA~ssTEaGzBq+|rj~M|r-%5;6 z#-`3yL2)L2#tOQ_bl*Sm3vqMwSYgMN?7PW(MS@qHo%Bn|l&W;#RA?QsT% z03{8Z3nOVj)u;7EG`ItshfbTZnW0jIevUAUqzLBeJ=zS|c?NxbpEJU8HSIT>GtQ_y z!j0!_ysW^$2Ia5mIothSly24VK8-edSY4=iC8vjplbA)Y0GVl}$>v;YVeK-(@Xmem8t5_BLKmVka6P8LAG6J`r^Xng>v z_hI%Ox=5X2JR$yB#}wxHFCr=O*{6NZw@|a`d1ua7&H&6AE^viVgx9^ddbY3JFZaRVvK3|A5=)O8A zVh>Xe{&)*D2$|TKRiJ^Q{`5qZFN)4s1GF@S4Gd?UibM~ZbqM*pe}PcC(c^|OIT8T^ z!>fKBIukM;-5ptLKp}T@%t7RxQIya@;aF=xOdW*mk(}@a*uh%^GVahHt$W5AP>3Gk znCEbY!ZO5L12Xo&@T#~!&^5rPUP20ncP)LGMGA(QQ7~|qRCEXQ-U|3ZSjI4mWDK)j z83Q#{@%RB&g=}U{#xU!ZG0cQy41Pc(8H0H`V?f#bLFZBo{?JIG{AYeY34oS?o>^yH zLg~7Vfel1jhFPp-=njmi2xR883^Qs@9r(jM<_|#l+*&FQT80%U{h)iP&L38B>BnMj z`Gff#KfTdKK+k;u2NoHBa8(~G&<7ti)oB@K6)gjm2fXF@gXS29KV0;jDj|SCJ1gDr zUQ&Mn)AZ0`1=Ua_*^na;3+cr5MV}%r)qHxL~=?_ejK2&cRFl>qPx}VNPt%?>yF>#2?>#>%sY>%lp4|{*8AoA3iz% z`sKq1m-mOR=LOM2TI25W67Ve)y>IjEmwbf2tX`bHP>WE?+tu3Z=+-)>y^g6n{-T`W zQR4$iEAkIf^y@wU;JSKK%tYG=;dsGvDm@&2Q5FrWt}Z)=n>cTndy`9NZ%PGq);5boLW@MzX zjFoy5^>35n&8n1OOBFj@cOS<`O>d0Y%nDfKr)V>)w)q*qnUxGb(q=~L+EXoR;BRIH z6US|4g$Jg=!rjbDk8XDPvFp0arUh5?Qnw%eW){QU=5#*pMpY<0BW$lDHnR-)GFR6< zcQY%*)-e|9S{gzcZ@JAZBf(sEW1St2IP(Hd6_TZWYvQ0-q)VF_UD^xW^O->;A&8no z?2VcWNz`onpv^pzsF|mCNz~eR3DX9}kNfolb3zb~Z^OCO8C$)#ywwD~3~k zl;@FJZEnv}^UGEs@-+)3Ex?TbA@{P{jY+PAPKZUulg$!P;j2G@h;{>Y$l&6hfmhnDgy zTk+=#;}#AkcU=#9Bf1_`UxAra$F1D;psTKN(2;Uw`UuU@yD8{>gnq~N5t>_>TzL9v z>Ec2>2kTu+-36g5TZ#w>y#8oq?d;gSlb zM}&O_IU`;a+x~sBtWf>qM0Kndz{CiY7U~q`Y)JuU-hC)0vY3n2cTaj~9k4f*5c0fM z>_Xfg&s>Ho7%SzK0`p?NMu!SW6H&M5=?(O7A}8c|IAlk15s3MMy+{=Bfct(Y_2yF2 zLhd3F1E?rPz+74aQF1P6FDL9pB12BB5Z?0^i9$SK^HgjBscpU32(uT7j56^&EWbz) zYaZu7s~61&I)gBUkxC~*2f@`m((0L~pN?33gImzS^Z7rjxa+TboJWegxl!D8^rT*( z*Xx0N-U$y^QQhO*UdFbWqj#x+=Xf^&5nZ0qXUM3U(fcutM09f_qGOF|%8}49{aR>6 zt+PlDLIPY(;m*PbnS3a-)3&QvxnNp?%0TGALkoe`#7ie*$LXi zWz39vYadsj_Aw7GIxcdxkE^8iap`aov-WY7ks??7xT5x)a@xbW<>YA}r4?~m(rkL5 zAbMUK3R1dd4LfQB`Qdzr$}jy3sqeznICSUv#l8cugCF=kcCUE&u~Dn(q7#0YgLVad zpwm^K{X+5|Dm+>Hu4f)#NTm*T9Nl4lKDuU3Cc)lN_sO!?qmFgqmwXU?v?sBjOw^Yt z(fajEwj7Z=#Mw_aG=>lZpt?u(or(Iql+lrTL(bHEv8M|WKRlsvtIWYUXRykRV};16 z>@bXP?Hy`c}7s1{vUY}nW0w!F9o+3bF8-Q2*Fi2=)K&(pzzDVIH!bx_S|Q}C|g z%WqKmGE_iSWrLsb-+o?Dx29Y2w;vPH;0MUMK=kvM`r$;8KCUId9>LA**0Jcoi~hdx zczVg9l!H>vsop=KlQ!V#6lZe$bt)4>Of*jAvn}~ix zYT>`3vMITcnuFL57SNLA&QH2Zb2Q_bc#a zLn>T~Gag-tn9D&%i8!8>UJi(2zy;9HMKijO6e*p|eEyIeMfy4O^i%c%8(e}-#sGy- z4ICW`T{%-C4cx+L;Ci4CRb}O*fa|DEaG~?C!J*Vva~@;>O&fg>Gei_dG}5#!jHZn- zqHH0cKCUwm5Y_qKX=uSUo&Go>;3BHGUYL68AOxhooR17!UTFhD(~BW8Kn8+NkRSwH zN6cpp>L}3%c-$j=q2k(BF}k|ZwZlF7cgT|;u!~Jfde-hj#WhtK>V$wx%eYe^Aaypr z?Kp;BL=49$^|H@|fIe3NGQcHZROlf66-;tN|GH209B&=I*yI~@UOi9;qrM&CNe1(ItS?_Dv)cd)haV9++t> zNWBPYk51GTk^{n#0#e&I6^x5`St0F}G%e*`rdLNn-xou=CL1NUZz@=y5;Rh|dEq&< z_DzL9%%?gBwS(kZa=eNOHn|sJ$=U>~aoruxj*{lKBaOeA6&OjnJ~{3tXOf&ikiVIg z3_sFlHclkS-^>aoj@!%%4@`rF){8Kof~7|fn^`mZk1^>Aazb)r-GXt^W+pk0j0Egj&h2Q+oKF6gqnmg_%XnI?Vs~Og$l#y z5ZOrxm$oO{ptD_?>zo2a?s`KXGEqtSa$!mGva8cukPvQSoD8@&$!S9r3l*1p31L!P zzHv<16qk3b6kme3+iWwi>?+LHUE`o5<;r{|TY?rGIboqo2wgA2rIp=rWKc&K{RL5#iE)Lxdti(Da}kTl!<{t)%A(QobX9gJ7MJ8UW4VzD z4drGDOtkykwD6T@-ENn=upWC5SNkiZfTTvJpx4P^`%~B7S$InUB9^{NWExoK(%yq zl;p=W($X!BmX1`4nvSXENSFb&oW_JF>d5NxW&3Y6n*Nbpt zDOLPL#Fpdgg%}Ds?hp zi4kEh!dYj*Xz=uKeXv6vde&ngL(1L?#z3c``j3hFjfQ0Qt#=;3{m$d_yO$3hu_cM0 z^BUS$)yvZtDH_`P&x(l>brJNhNE#n0Ib==F4_I_z;vZgDzbt5`RdXCOY)<8e#cw!g zvID~rDE`+(?brhnk+-L&#E|ux`j->+GTQ?wi{tz9RB~P%E5ge8b@PKozUNHThm+<# z)%9gJ?PJJyyKFG^e6Jd_-4zH+szY-}b;%`0PPRB^yDJ%gwC!%3baBjfS1@zzc2{_0 z8aDjxuJrIh!05kh%f{l<-tL+VS=PMc?5Cr)yG+Q)UU&EW?XD1Kh5h9?J76uZY3_id zwz~{9^W84G-4R+L`eE>>c7^Ky=-m1DT~EYSBmgvzKYfvZp6iw3I0f1YsQBn~EF;s( z$r{bYu7DvL5nm611_jFP*mN4TO;iWJa&7LqFMZYTeS5P)C3$of9mYA*ezWshs*r==AlS`I2u3~-zD(2e{)ka?( z<{M}yk`o%bu<>_oZ^V^rC*q%e+R~Sm^0mWy+O#6fGbe4_Sc!FzHZFCaR{JjgUvCGj zzh>E7eWW9EVEXpEKv#_Fk75UDqv=mzpgz|d&WbMi>W@HS`YiO(oA+;YmhkK51{er_ z!$Z>N)o!|BLedK5Z-#?G!M+ZE7CI`h&pi(JZ|2-rLcJw<@V|GU6UZ##eK$dDB?pxd>#l^Iott^6zz=C0tfQrm>l zNyJ||GD^ku^f8tWq9}5O>Ims@N+(e|*U%6o(h-`cH%l=qVZ@R}u24N-2A`^>rA`*P ziu8cy={-IqEwM#_IbFuTpMQ&0Ps5l^m)= z&C|J~YB`kuK!TbBgpoQ2WgsZ>066-zMhnpC^hvX#!*np09stLg`g2d8$S>!k!9hBL+5g{j?YWW?IYX1MCCf${P505H%{e_Y-Di$w zbTNjsWDY6? z8D*ovG~y73N+~U7vT^mV%qVOTlODCS@F)X>t#_PE@F#tl>4mU?N+%>qya{JLu~;WG z2?ge!&DHXK;`TeI_g{Yf{@L}Y0sY=iDzfXK1EID=rQ~LWtq)|egmXHwb!H^T@jKN zNhAeEt!~Oso1=LbAT6Gm)(KFm9$6sxr^b_{d7#b*fuYsy5!l2a0)aCO}wb z82iN>HxseBhv;Tlv3#DJ2@sb#nn$DXp?uJV%h3UUCcM6;BQKvfw4mN2(DaHxGuxxn zbmU}0DKO>mi)z{r(r3khDc}T=lQf%PdsV5uX;~#3`eT8^nu=2c%3+~1f!xx1B$i&0 zSY|A(#$aeNK@3Lb8|5@0xeBzT)nYIxuY-ZX7+#08B$?2wXJQ-%lgVt&%q>DUc*Uu` zY|W)dB?{>SG#Bvk{L*x}q&bbhW*%H6}Rftfy?oBNx#W7e+iqL^> zXAGfv$!5V%ipQ5DI}28ctHWblLGFwgZ84cK4gulIfC*jq#Cqn-?n-rM$Zkk1(lZz@ zR5^F&9N@z9?{w^ykbrB1@YYSCd)M8mntDN@{NNh<_MgFXsU}{ z6{Gpak$7&7^iz;#mMT}3Eibfh0Mk@o7Y0uAg(Goumi8xfms#3(o~gQOuy8il<3#di zgVj`)qZU{*)8qJPO$dk_R~P%6$>Rig%^aJ*nB{RImUpl{R42SI#7vJ9U^cU~e?V=a z9>*r9mH0BYce(0l3?L3gn3?b2ot9`qRc6if9&4sAWzA4;xeaVHcvW1zVa@cVteGuf z&4ej69k#J+t2$Y^wN=B3C;5n!nU15Veiq}x7RLkNveyLWwF}RIas)fqt;Wo__ zw;4>8)IZ-uSq+1I(b|I7 zd22)OL=~TbL2mGB6hR?DWpv+vlhAI*^fVJNHSf``DwGF2O#-mnGQq z*sc8*xwU)R@xk({xN_r><*6Tu;8kL@#!z3IJu6 z553@ps)#+Imsa&wA)Z^1yf!OT1;{Z^&Gkc7z75zl%LUt3qWQgRXntFd=BI{F03)P> zfKyI>9WBsTiREWHf2nRm0O%`a`2jd7FN0~~H0W+Sd-~SaQ2k0e5rH+X#P@S+?D74! zitjgL2bBZ^(Dkd86dnd_T(lh4qyXoSSy89}W?`N)-FlqAtMAW1Y%J7 zLj2q!@xlfE`FG3?Oy0lsW4oyVgRs@dOM zh=H?nC+t{`X6x@DR8n4~iJ(ZSNCQ#}G!RtUVi(Df#EdjCQ9T^Ij+xGTH-i^xBBJpl zO=PgCNpK=fY}h^iT}$s}0mg1mW~A|9Z?HKQ7y)IMa7@b3t8JKkOcK0E69C-|e|W)( zG!dhTr&L*;CogC#S)brVngHA8Xx5r?HA_MxjRosax-h+RT|zLQRP~)(k1o7bbYXVv z#_hr0c^Od#K>78ya@mU(!t@JN<-t-W-6&6uUR1CS`rxB0VOgzS!TQ!bl?zJ-(Q&LA zQ0gw7`D|8H6%zwBDC#a_MGN@O*We?{t*pOf0M}IITcv?_U6{7B!>moOKt*E*N>k6e zxH@&}a~{AA z2l~Y?L1@l(llySVhYjf3p-K<)M1Z$GZLzFQG|P1{mK8r5sPmlz9$oc4Bhp%|6@NUr z&XIyaZ{DFMPF7cY&ydwXta7I#oat&O4n}VO8>yFrH6MP~W6952TO35cY}2-w{#c5n zELmTdE|-v$e`b+j@ut6>LsdyGxP}hmu&eluRQYm=NIJ)jzf)W?P>|og6!m;XL+*21 zj&AE>RSj(iDX@^OM`RPs`{&nB@BQ7^@0{Mh{+jBId26_y7RG~6gPd|o7}jEV z29=jOg3MNsU;L{h(HBZP1yM`2l_=#)t2f-5p@RkBX$7^!<05t;ykAuD?-nB7qXv;~ zf}EDB^SFk*dHb^&V0jRUU=maRVJ_$9%aWuIKZGC>6VwAtYls(*4}SeYA_->%k%&mV zAQBm9Y6J;dXL9SrhI|i#$VH@zr96oE=;%SbeWX2TmDt1xA^|>H==Cm2gGj_yxK94S z!RLYzL;_H>RGl>sRVc|<$4LtXp^R8MBKl(*5gj~6tSlLSd&P>!P9NFGXbH;$AhUXh zyM#AB4n4VQ^l9JqRZLp|Ybh>kNe|C1e_XLr3PaW$y^aurd7^bF8q6*(;HHMM(!o)? zxE?okP~6nQ#nr%Qc1@jPAgx!Hvop|9p6UQulW9vHtyeKrfi20=0YnO5wyb8HrFeZI zhTh<$G;XSP1CG|)`Cuu);qp_G_O|N~F;?mjVQ9VMq(ek!5L4A5B0DW-LmsWSpTef2 zX%jI5s5(2k@;Ybg4roC8>dYY;acg&$Vn<(ngq$Gm2))_4JIxqE@2HCk0K6;> zO3%Rl^SsXO6TXAK83TmgvB=EdG?23NISNuP(}9^gXrze_)63TlyL z;0EVQm1=92l`5ibAraQ6^M(njcL71y~b)g2m4Lgj92`O9F;4Q?adc;ibINjaYVtUfP7P+uS`3ux%N?Exr^^eu7tMkkVWxt z7b5 z8~`pwrg%}dnIra*I69!_65OJ?tvKJ5iZ^qBv6N}B-ex8~P&>|NXes{N>u|;D$YLszoPxAQEPoe8kzjxjieFWNses}s%yJd6WuK9HwXdUT zo+t@2*q=w#02otqkQBR?9%E`$jH#lv)Hn)F2>1|+JQ6b?6sp<{@HC>W1Iim}j$Y>g zH99~~sFSlc;XOvHvj6>tqV~gd1hd}9OaStNEsWY*d*cL;(S}A<{t(J2k%OCe?yO7| zbZ_7i=3}(m04b4}fS$SCBIri41T%%?sBJ$bm{C>s_Y4{!*G{1UA0nncfLaF89dvnq z8s^Hbr{ed}K50Mz{9V$~CQ{eMN0uKdNTe8XT%E;S_U~$xi{huv{c&_-?kO#m`J*f5 z$>)lbi{!H@KfZ8t z^vqlQ`a;CXqL^r2-XF=`>j*f_-{mc{Zanz@8}HjQp~%RI`01lWRUanjH2*Kx)T)Nv zFkB9n#UEE$9`UOSoH!M=+Gz{Ha*evcS49jwd~!B^B=6{P`XpQfFtQwMr7f%*)h%^AINwbAs@8mYObRRf?E zgzHIXWZ3oeXa%F96%?;08%vP65d%x0rhHaq;C`J>H}(l4lEx~5p}4tfD|HwVEKRmcPO{k3x*GfSWjVmtnDCD0g~kd z0_=gIRa_sYG_a0?xGS*eRzAHddcy?<^OK<71yQG6yhYQOJf*RdZg|% zzeOdH@~1-hVIWRl!w=mFtO}8e+ozfGyDm1MPF?5H!$~C?gI!QaW7v((SWq_pAa61m z_yaXHw>qoKWf)E%mqGsI;SbcwS;K%uAaVhWVF2DTY8VcG7<~MJI)KF=K1#i7{mr*d@4s{XbEo(3oxXNGMS7UlxSPBLd1CYU`>l;XBc575k+N~h zgi;-KWH+6lqmHUOex;t^UgLwBB85jm`tzRu=(6~DH4$|qgu@xjs`hd?k6m*?f*v|p zq2BJf%q!~U;Mm!lV$&WbR#T1u+me_Muh_M$3(KDquE|KUhH#Wlm54<6io z_1^Nz{=d-!FGY*1di&roRu#v{Na~)nt`&c}h`y{!QPY&M!!_E&=qRfV!4GB;SR@t9 z>TSYg_nuh5Lf@(82eX*q2MT8MTA%ZSSwzHn!7MW1)L6K|EH-qu%a3-wHbJRrp+UHt z5zKsW_n`%#ibu?g8Nn<7e3{<$f*Z^twiTvd0Nt%=M$Gs&BbWt9Fjw6*%_o0Ww4u|% zsg-DHUsxDq81mAF6U9q=f_pww5D7~tKvDB7aunNT$3iL-kbG9I z@St`{WxGyPS3Ai(=U7a>i+sy1e!W9s3YgF8!WZ0?STY5-5Iu08^`uPT?s;DoSp|hz z$6Ajk3>mg&N%kWmK8u?FL?!0n?<_=tw*Vq~Kxv3VQeVF%2pS}=b7~cT<){R$1M*HO zggmW9yAbDf5YkW+jt#pj4J{lbpVjPANQ=5bPj8@xCv-xdhMnuICK7sw&UO+Jc)*+s zy$lUq(poa2Bt*5Oti|vFs2SThIRuBk_&Q={|N zbc!RPN&$V8WERPQkjUyc=+PSO&Q3>o#`Y4enXKX{I z-8w*uOzq>EO5ioz!?niACw-JU>Nxie3L-^R+{4%_YuHg6&=1$>vw6S={0hVbTz8(2 zmK}f%_^}R;=zy1Y)8Zjp(Jh-oJVYI|8A*eR)E4zno3UITT=T^}5Vz{1v=de3eIwDuT;FBeULzKVqiwp6|T0(p5 zja105r-Ja#ARs7;KRcq~R)K;;oIx`;4rL;%y31gL6g%??R1&_wl*baqpAZw~;A@9M z)q)Ql2dG*VKBD+FBEA+(h=s2OqtH_4;Lk$4PXwqeSROihm-W^&N+G>ADbEWf<$!F# zDETdLfSss>X8EBAGBV|l8RV^&pRPwW+Yh+nq#*vQxKGn)l^8g){?ruQ%}Cwq!7?5I}>Qga5iU_IPk>#e-w5u?Bz( z^aF^564~RRz(vbpHHgx~=uq&dj7b@K*xtb8=l}{XnxP>QuDx3H^$A7ko%D`m0DYnfBLM*I^Cm4@&p3?P^`Il6=s3I4 zmfh!dHMqG@+7L|T5a4q#p}1Onp5z+*l6eiyAX#>if!kHX> zt?JlNxK{uqSxe{T{2(Elehi#VGC?ponsrgJmi*WT6+CwkN%aj(S!dQL>963O6P`Qp z0VGU+%7CZrzgik7^c75Z5I{x@qrKp`gNO}11b2`MoY*2W-9dm7aWq@o9T3TY8Cp*WvBZ~PX7BpfeYoU!Yv2pubDp+|T_>95x;U|mr{uUyAy%N$J%?Mk z8c#-~H3YzQMzP@}tgfqNyU~orm(>+{GGw)21^(z_t%Gb-9CWG4?VB!czxLMYH&ZXd z$b?F$f#hCwQq|2VYm;_)VZSC z5^+0g)8zI|F=;2T8du$f>j-mDV|RLrOn*i&i(n*E|738tdTU&dAV}>+7%D9$xEIW7 ziW^>02SF>@>xLi9A|lQUW|0A>#zN~wSg;(22vrv>UOTn1ZbUdQn0eSp)~y{jNRj!4 zGlE%w|1!NRu@_-P{gGf6&lpInG3kMO5CfeN%mPrDtL}pcmXDCO!g_;p?dS-x?NYb# z$aZRYT9>-eqRw{Bi=Be9cH@Kse7P_xloJRy9$jvu=yD|lLV&mAWw3lOuXcQ68p~~l z{Nh@bIfbC~H~Pc5Zk0X$+eY!qY2o zwjGJEV3(O*gzCy~LL!t`6k5&_l^YUaqwf#_iEslhbn+G|udenaLV0ziz;_bSE6|0J zfx>PxYV&##cCJqysTU!ffkEQiFy99(Gj4lA_$79t4|@@IuI~WzoMU16v1Urn6%Nz7 z*Nd=s?Fg8H8;oAM&#Yx*)2gVp1b2emW^8X>kcpNg6>=-kf=`f{gZO;}war-Y{d8`L zUSvjSMfFq?;I#n>k=$l1kar4=7Wi0T7_ zR&f#+JSP#cr}}##5h+W(5!*=wC=<^vEAAU~d)i~NO_>2_FHDW;;F9TE*La-0hTOMq zYPnNn36siVC=FoVg~HH5B||`LJm#Gmy$Dn)cz1Qjk=l6mg^lSq*HFi?Yl?E|=unbf zQ;$p6C@vkT6m`0$Qyd9hQ%FIS6+;XNiQ1;mstqmE4p5@xMU|l-HeQu>qpGwW1(CX@ zhJv`!g_#;JGVh)=84i`*?~a1l`j`=QB2r6xBkk(;B2+|82p6>mE|TAI;Q}=kkAm1u zk)n;bdMCWzipdeC2aw#R)G=PNIZ1IR@*BSOA>|N(NQWd59 zv;jZxX9>7fl_ynSI#E{KHJFFwcw=nCdI$GthJr4w*Nd<(EEsi~-WQBo zr~^l}YRStvoQeN;A^vuTGW+^l=Wo1qe*Mnry$5Vc;wLPv_F3`F@+k_fHvQQW;vw3T zP!>V|^_s$aOb#!LOWlvu)k}B^Q-lxkKU@|+P*F_H<~XF;tm-d|i$-;I9?%c-u%`I) zg&5fj;vsJVY}s=G<@> z8FFei{BRc=e9y)=IZ9Qp#G^3I33on*{QryS7NXI!!d*aQWc%FawNX>7L4QQ4A3c{6r5kEvtM4ZvwQ(#N zNM4HVsg2T0Do~%R4I@R9w-`G;@gG1De((C|ZOQ|kLHPUQ7c3C^xEPmPCJ2gb*JXPgLU8KK3IB@Tl>Xr&`^(!sPs{Y)ePsP@u%T4u8ljy*H%TWn)ju_c{m{YSxX<2SP3^B6V)Fno$B1~oW zrZ;1+;U@JR`*n1VasdHuvYjGLjRfg-Kv`4J4|5N_XsmrojHK%P++Zlu5GduLp)&;3 z$aX-LU{xi)4ABJF)h|~W8hqS9CtJpuJdF7c?l@)?lLnI$@?7XJ6IcJrjKT&nkp=1) zL9*T?mnUI-f~&nHPr4-mfs#OCsl)O4Ysz}&OBTE0)*e&_W5%dg)*yZ+hJ z^V|C6kT8=qsm{w54}4xcyL_eq4~%}O5>oQ@g?O9>2t{X+LrZE4T{Ei^e{v+Aogw=a zHJPKw?E9BxHg{ZG?=?2818@{Al1K`STHO>ynWK3d$ExcLb7wQ%OeE_yZU*~mWfAF+ zi&8Q=3U~>egT|%;MA` zz+g;qYC}acO>wGxYh$;7Qw)ZRhP-Vs7>VU542H-nn4EI~4%1Oz=J+Z=D7LP9Q_D$l z3>GtlAszrLFWHp#N%8n{WT&(YJV2c&EgrLV1-Ua~w8dmH3daDOx@X4(XyeQ7N_A(* zZb&S$CxT_%)s=;FT^2Du@w*H0a;ZhXdi(4w574|~t@#D<-0~*Nn&Fo!AvCfrbgsW> z01m)ub>Qa{eP)9NzNTZRjw*q-f?x^mnTW#9&2U(QW|k^vd7OylVY|J)k(GO+xgICL zY-VZSorS8KEVkPn!cmDYV|$l(9$}_agqivN-D!y?6kyF%fo(k2OjpX9q2_fqC-VajsClLZ9sp~GUeOl7Il`^2N1N#sZD!8eYTU+V1L)eqregALoPpa=pRu&+tIVD0 zJnl^Ab7!cNvo_)tZFPn_(^0`GhuaK3ZbKbFYH5$#R4Uc_xX_H2NvYOga2w{krelEH zNL}ff6Wxl%Z4%rWP=thQC7*F6cXriJ{*G_s@d^JaUi{5O%8P0In3Wz^g zh#!n*ly~o)pWZ(^y>s{W`RTk4WGnI5t-Z0{+BHaeu)Hd++<0VpboGed4mLL(tTWl8 zmoCA4qso;^)op#9mAc&XBw>J}`AGLgjncTPL0WpE*? z-a?dj8M`L_B-OS%Czyu~dT2zhl<=C`<Dsqxf$y@O@ zm$^|A4ZL%I7TQzjGaU+7(P_mK^6XkGuN^-RWhI>efIKxHVxN(OX8EBA>@sZw6#)P* zWBn_unE={iOvWvCsEP^gbq%ZK8&b`{V8mTi4^7j(G)NOHI&D>BGAjdp4*AG?s?qAm& zU|fmj_pYJ&$>+}?ok`LBW*h{vEm^Kv^;wbnN-RIq`AfcQ1|(9W7qR>RoRqgk^Ezdt zyY1{t=sH97>jBj-KbS#2d%*Xb1035mmhJ4JMik$VwZS?nf2Qk~?3ND|=yU#xW@rpXHDkMJTgCN;_}AxOYy{%Eu3w{vBDbYxw*zElGqZQh zu{FH5`RChz)cFB1VVZJZ?*U%bDou%K$}mR_C9CgP3&}eS?>SSsx8MKi)7!5-f7c>n z<~h^0;5pNsKiw5%_UJ1<^sZm;=&oP#oq@fj*wI%_A#$P;suKN}$4phjF{N);x!TJ* zRXvZqp6w@v>G&o4WbI@D9lzRw$4sw@AJt4r<&hQQ&e4RCpe>O=b~14Ytm^<3f6F6z zK+-_FbmHXqI+9~0Q1rYjMYxD6X?J3}x?I+|#^URht`(A+&}Ecdh4@0dGF_I?#)(LW zIcp^hp|wtWg7o17-0#^D2K5W^`J$1hqAtKN2rgSw)RZ z9L6LPouzr|w^5~<;e?n3a_f6d{MJH{!w#@d4>ykh;kC>ufFh>Ewrjp=p|XDQhz{Zc zS*J)jM_;QthF#4{qHYwsw%uKKy!GJpmh@7ZQSzb-A|&+3y;Hb&4*`j@?crSoCyIHlX~5i9&h7dhaoQ zd&T%Ic*H^i7uc+=Y{IZwRf~53N}8u~ds^E&u%ak+H()p??{e*mdeq%sQFj?Ds?lqj zN&w$kqE`&y8bbfrC!Ot}@4A%NsrQ;LyD#$Z)w3?PPQBbRygK!MtLxNzlv@vwYRh+) z%6#NAZ{9!GB;Xc(L>UZV_|CFtW)+ay@SRmhP5@F%#B|TAR^PE31l>k*0RiH=p5oK3 zsZ*C-H!(Y~A5*(=58CRRxk;UR4{&a5K=Fn_UnLH&(ARZQ0;4h)gS~<)G4zXHf_S9s zCimfz4;#>R9R_H?ZhqRLP@QO&>tYltfj*+rNhHCdt%Iw+XGB_qwYK=<$#qU43<~oO zHF2`K+Ixnq7GlL8POc+sid2BE>u;o94%U45S&t=SROM%_5f7qYwrL|yPb5WBmPD^h zmwj0BviO-rg2kI2KZmNOj%YPR3lC5M(kgx~``osM zKEretx%0TTFRC@gVM4YZkxek~pI<+{_jg~vb9(>!YpOTqt?_zV7!N}A4)*!)MiAE; zDhM)LX=5JgcrUj2S4W~Rly(ZDmTD_e%9mDexHXS!s3@MYl2$Bw>OkU&&M5Q&Jy3nGz$ zrbdvcKuB(#*pT}mLO16MyC?FzFe8Zg=qS;B`!{BJsSgrtE*L>1z()(c-bHB;iP%b& zrT^v#A^|8`s!og)GC3gCP@9EV;{ts+SpbipyF8i30E>>lp02zOC00Vr!mg-L6>*LAj}PaMUiY$4%WTZffD;YGBly zW%5D^K$Y;!Oi7OP`}4 zM-*%d$TtQ0%CxhZYaiGp;+4=>IS%9Te8or8I`MB8BL0lJ2T?!`PO!Z)4Z6TLUm)aO z=t}WsjwI8yHZ$7I(}z1ykxDnS55LUScilFg%GM_gI8nNp1Hh%o6fep)bHqLp5gS{> zgKqJQ;>{djEM*$3x0zA0*%b#z0EQMlYYZS4Arpgr(&#a?wu+&}aE49|*wQ;8P<|C1 zQ!($NY(0L}R`IJ!FcnY}h)&cfFxz^a1GcJjfFqa{+k`F_)}}BR zo1jsTt_Vjk+jy8s2Y@C_T}n&0@7%f7b>J~tx`PO+qlF=uspn!KT(Z${iTaLRBXlL@ z)Bpm|Op%h#JVq-Yjj<3ODykuvNfOh}-lqgJ09K*n{5^vPNEl6_0Usi!K7d*_fJ$}r zTR#nRW!F>jduX3DAOOy{8g0sTU3_Hup@Ni)5yyo*i@EII)hHLmPm}x0C#5j$g=Ypp zrJH-IOYqBl`CM^wk$fiQm%I2qlP>`fs5;g7yA?X^aSu$89*=$A8s7~uEc>kSMTf7H z^Z)un#L1$VXkOm$q4T+=o1Eg0>!>1+5{REZN>ue>a!&LAa!svj*iE>nL8@OzgVA&! z-H75>7dUY$YPHkHUduIF$z)JNHL;bo@gupP!3JuqRV{pgv+;e5AWX`utW90$fVJ@h ztf1ITFUmK5#BM6-vwU-79*4E@1MHw&qo_A>$Kfh_%qCI;N@)nNf@FD%@`d05z29d!J{G!ii%KzBT(}Wiz9gDYJ=&aHBj>)gCh*{X$1;w zQ0*>)7)ad2wsj&oX8SP>Vh$YB-LnZWHz`(Y43*@Do-mxX$kxEJ3cF!V=KC;fVu;cmi5v2bSoIF#C30q>eE>!T+qm z6f#C34@5pM9;Dt++MKTu>*e*roG%VK2ziEcxPmLfD~_!eTbS`!YK^#?!WW>NqbtWN zmeXPkc!63CQ4&xBzB7JdA)cpMC49&-tPU~x#tYWy9v9{7OiJ?5l%9>~71! zk9$xH}(l4lEzwqYmB?6w~CvHaG>K;Q4?6cPO{k3x*Gf zSWkQpzQ>;7La=;5fITp@TG84LqpFogVeI{NUoNxeJ1SOC=~k$}eF z?O=KmsW1vzji3`ekH#>R(ik!ac{_cZW_nd2gP}1DfS!uEB8S3MqA}P7^=J%(qA`>& zD4WY5>Aej6p>=zSrnAaihQZ@93_h2EIyq|?UJ=M(xD3Nc+#in1(E9j80YPTZWvEp8 zVYv*_vLKhCHTZ*9`jJ^c01k8k{vbDzo;lGV4E`W9z$ll2iU+)I@CS)wDE#51=QY_<4|~kFLWC%AuHKLk58W^%PIpWw4&;sm{M(3S`P zejz?pQFO2OK~;gn<6zJ+CwUq5k1vbwuP3DLhHyZOSyf*W7jddCN>SULE4#Kk;y=gR zO_R*t1IN5$HB!>F|86WhuBwXwtCJS%_21F*|K3rwj*2Az6f%o-iX;uRs<93r|! zgq#`OA|p;Mh8^8vWAD5F82sY6qT7fx`N8@d%#Use?3>->2aawL1h{^53&3EmtG!@H zw}^%1*E~#W4im=`sm+||79hoJl^-O!HPm@nGi+a|H3-ql!g&pQ^p zAFV!X<{k-NAYpYwN~SNmE5_uCZqR(`MZ=FBL>q~NZ(bWhqT3%U0DfQ}mt^xI>;n<_ZH%D=H{7Trc5owS8Wk%aT3}7!Uw3n zS^Y(bRq_2r07>yI`GV;$B36|+0HHxC2-AmfF&BdEF9H;bsbzQgsJQ1a&ANmdu>Ha` ziOyS;Y0|EoYxLNDqhkBbam7wCG-&{7kB+efC6$)DIC`YLQIYmaN+)|-9I!9_^E;ul zPO3w>+6;0$0L1P_2j~)Ya)!Uyd-ao}L`U>DhC8C{_$FZ>!vq{kakRSV52j06pFp2V(zlqa!8 zLRyCM!^96zA*(6rLj2J}{17RN``js;*$9|emoc^>KUE7RmmxrkUtWlptBLLLKG72$dTqZ(W!DmCps^)wg%uZI7m%_R5ii^SFGJc&3y#KsfLf%-9vNs{AeJ!-;N*OIpE>T})IsQ1t;~R{<|^ zJV6;iJKbQ_Y<14hIDLF?<#ScaJB4ga>lJMG68{yY}MtYj2${ zAAMF_T`VshKe8)s*Q`W2o+pyx;FSQxpeQ|mrmze%Sm7>}YJk9or4T@d68ZZ)NO9edT zgKmsrpav)f@!iApXA;;6DxW;1O1VPAlLOF}7CXCjOFEWV`!@2J9G1%>j zBHxN!veDUu08sOUD2Z#W~mMTDFg-6A7SEr!@3G2Dx>u_vs< zs#f@rqgxgjt}5JNqUW6G7C^;ZS4-`XNV2ep6}wW2OWwbe`~Sw|trXoZ4ieo;n22k@ z+6&PkI+$!)=Z$NRwx@*E?F($N8Cx6B@sP7Tjq<|)dY-@1^EDC;#DP%}UT z^PjUy1RS%_s_1yL?vOaHb)&d%*(1a1Au7j4Yz7p>=f!nw#L-8dOD%72 z#-OB~Nz79iwcOfzTTDvYM*}~5CGB}BKLg$cme)A%|McnY*Pg#?Q8!c9zAfr*YS@Tt zwDm`Ehs1GhT7T?FfNFGNQ$WwDs-R{4Ox`64!$;g|R+cK!?QbiR+w4>yQYipiz4XXS2H?!r|gcJYjhd zcp0_{+)5Uyxt`!2F)f7O9AdWoW{Xgd`UxLjbk_%gE4uY8_^ZoRla)CADsSTC4vBlc zLCRIKLn51W8GNG>Ly_F^ESLZer_Nvt#X!`0dZTwd3nZQbs3rauSO>AdE&P07Rs5zA z!Qcr`RCe{}fLVISvtTAuz9Li0Hs01$5~dD%hs3GK5opQ)qt)!b7H^U5FR}ntMNtEI zYKAxB`-=dQ;#o4aLn6Vd4zMRu$P>0xBB&VOUj!^hrj`$)8%zrhra{?%bnr;W&=ok%uIJz0Bg968;``NMlS)KWLkmE+Ag}|GJ2JaYtsa81K2j$ zfOb(gfR#6=?A0vl#$)BF(TP>B(?y-)OsF(~%_x&h2E+teO6Zpiy0if`;Y{;U5R~nZ zxbaHD)#wggB2UhANZjfg&^?K9qn~RTH&mLNWiv`;KJuA2@1N@ewJk`!b6VQNNn@P{ zH!Pb`*IH^eqXahg4|{CockCWPcak)G0F0~yBuTa!)iWoKXV{DzjigW5jFJ#$5ph&y zrbFUZZXkNmCfGF!FVYanra^UJXCy#}#H}P_8^D0pA<=!?U?VvBja_%?KEycv!w!jC z$+R{y^1`o$L=IH5=tR}P#%zg8i&LjFut>-d9)|$4vEr#Xj)S0^=RYmP-^mqdzW&zv z8*iOozjJ!;!6}#B_(@x~`JA}9e413Ynf`7yexoeA{&ZFGJq}1^p~JfvI>+&!To#|G zDX8Xw9MpDJ2Ur^G!H+914u=i$`6F?IzbGC&=bfao$Y%_ItE`V8LzmVYpGt{}Nx<5jf@1V2VrWbenStAhA{GOhn`Uy8=hUyQtDf2p*75^fz-jvN|V7{um zs~!V#R{V>Jf4KNJj|Dj^{zU|x9seR@PHksK{ELmg=R#(;#DO+rq)(%AKzy&^!_9fI zBEg%SeG8Il{154cf$zygLO#I*7<7>4BFRo`z+@4Zba zWot*j+W>Ok9N5||w8!pK!_QdgDR!M#65vQm0FVP<1DrAcIlEs_=36;Oa$}fp`$baB zH_MS^T8&qjzv(qca#Pfx(sCrFZXupMtO`|~S}64iabr1fq0|IEo_B#VWwyotvI@co%s>B_tAPPr!B#z*r`jH5 zWsR(Z@WgkW9{A#$djjcGjmNBu5V23lG4Um0@cl)m4R~>3aj=0$ICN0Yse!Ll;!XbH zLXs)ii%y=J74iCfh8Kx|4aSRjXG%=*HAO~wmtuw&@d2)2t9mdJ zuJ*(Fm9UfK94`_;T3ov(l=&u%IAURS4W`^vU|wvF7YR@-wzl2brJ_^4K<(O6@fyHL z3@^#);FV0VcIkKHF%s42`J&H+EwLDu4WJf=^3Xvm!?0{TYM~mv2(+S--wcUv@TxN# z(-N>L=;*g!2EY^y)j%)iwae-;1slZ_e3E&To#Iqjun~}BDHFudPj#s zILdG=2VFmEIF=g&N6GJ?j?4<~b1a(xiIUK-o;h(m!?A2M8b9S&Qt6Iw8Hhz10;v!+ zbccXt*-+mL|8)p8xITl~8r0zH23g6-i-Wzjt@!-5)^wQS zmZv|jSP1b2@!axe34|D(69&1c_G(WM!)$c$dSFU*u?c|6H;%+}bEKbwDzj9Xf6ueb z@@(KbuCChhofYo-h4=!UT^O~!X{qWg?H^EE=y_+R$BCrF#^Ye0t!)x0&fQS9TL2blAG>s3@l$OB`pbVDjKz58F3(-ugI` zFw0=u0HQHvu)|1nD;A=GG#Egxp<*TP7>GtfJqn_s5)vkU5yCW}dmy^pzXRb5*Fvge z7aKXatF&<$QHTm3X$o@A;vCa)CR^{cnf&hq%pr>srCD4t(F%d=_tDGZkJGA7*I z<&zB5dlKhN;5_OfIQ3OTfN=i&NIXAp$@i~0+pRKBo2%OP7w&dIk*Tk7#-{tJBk}CK z6_}7^=B0iuqJ`wwSsLqFl=*)FK0i-i;1rFzYh-{m-65Gc1SZ_VLjT(Mc!P`4Mp)S?r<1OIa z(9x*??Gc7^(|Vj6HF^=ML(Te3#y)yAVy)raw56Pzwi4&YEv`qpQNzz$Tn+1BYid{C zW8O%rJ_FXFu7fqB%FLVAW8Snr^QN_u;xW`3iB}TV8s<$)g@+uh(|WMZ2r3t_g|UTT zBPaw5A=L7>TE|HtSgTecBs!My2+^Rg}ACGxFc zoyVEyV1VniRJ)O_U@gue+Y_2hKidE-xJ2G1g9g~a=^31MSn=<1kR6%oI*DB|*Q|(b zpq=rAF{AK!(^!+g5V6fG`;P zMtUXZ=M7=7^$3G%_!NSYF(zHe$&slNU+})JPHjEDU|Y%;1XXfbVbmiFyMnV(H~rc3 zz*<8ZZ2SDN-yKjaMp}tE=oZ;y4z`Lp$U6Z}IfHox!S-XaK#hSgcPtfFVh?`G*p$Z} zY<>11&tkkH;MPzGTfhL!YuB|O>wLf&Aei}4q}}a}7(CG$Iu|jUW7JkdIqTFM@O25%2}r;pU>Hq?yjQ6?k~t;btlmSoV?zPVYe-X3a8 zkrkC@S+|fkfBQh|s0(ZXJF?e3yV`uuu50#__JP(;-#miWV`Gs^d)tsM<;)a1&r*dJ zp#$x&@e>CCjpWQarvK z5te1dmcbW~KOo+>d-7%Z*t3sby}W#6d31GgqXeqoV%lmq{C4s@?3Vi*hP zko(t;#EW%SQiB!;vWryTYu*yr7;{{05kj~ExnsX3zPb?fCYjo31WAp-ys)wC+6j}Y7`LNI4Zr${%t7hwZK2}LkKZd!8zpf&V4n)2%SPP1^o zWxp-dYc#GpzSuOp>iACLvwBpQ{GsY}H0s_*wFjb&T0k3A$A|8M4rrU|_;FGl-y1-i zS{<(z5P-AW0LwPj@guFY{F>U1XFE@(@A0M$47Ws`ziEJ4xvdpYyD3X+0DHPancDzl z{v`-py7qP-Df+U(XMvp0SOBXR?Q&BNU{y&sFDanG0kf{^pA&E`EQ>!5UFm3;!fDY> z>q+Zs{yEZGG#0ORuOuz&^!1hGyU3brKkG5*=d1;fV!P}R3r>G6#eJ6Cu}harxX=G) zk)j5tKb`|%Q|Fy3f`^_y0A;u+?Wk%G)O!Nx8cbCUH|QEr&=B6k+o zmY1)-dqv3hCns^Io)*VLTfKYP?Dtg4HmMRmt|EJoFiT7GxXfDKpE(lSQhBF%Yq7#o zRfB02hnqAJ_(&R0vcGDmUVM2W65VnT`6eK1u}bS$OKI>2X5loM3~{Ipo5E+MZXdtORIEA3F8Egd}laRZ)uJX z?<4EqV9z*A`Gf`YHsWOY(K0ZPn~|$j(h28Ula!uTH1I69ja$tM$Cebrob^88S@M!E zs_RKoDyl}El{X9uxIs^m3exirtTJnUhG3qM#~1K^4IFUO+T~Vs%Z6%~VC+imQ-5V4 zDBB%;*aHWHG;T`8yy9nY$SuZ|Sn@VsPvT!5k%mRM2X+m4tf$@F(Z1B;nVv zwGltJ5Q#_QJ^TV{ae@FAYSFp83Id~|74Pdvnq6&Q*Xm4Kysv$D#zXTk?Hy#(MykI9 zzf!)h13;$CA}U7!HL)dV54c`of=P#7WZJMx z$9s?KrAA+ITq8!Oxni|s8EZXC)e5kc;{*(0TY&Z+RZERtgt>r9S$j>)p`vaPq}K;7hyhUzyb-gDX`$f z)RYLY)OwWE2(0*D!$jJ3eEeSGHw{Jr_9>NhX?sfLY~YVIN)ld z%fhII2Jo8M+2@TER)B6xPeNA|T_&ML1n`cmvnAx5s8nri;?WJA4Y{XANRQ*bV6F0T zF)p`xtBj6c2@?3tg-8%kG2uK|U{?B2Nf~))Lj3Ges*D4ZbP5kJl(nj2J!k+4T@Szz zDycZy1X29jBE!In8t?QmTZUF~eb}&DM|hqOh$MvuI;b+OTJQnO2ULU*Dk|NS5}Q$ag>k)U?A24@3cH{vDS-|% z*ah`S3B4jElrQL%(t$20gc@W;2m@*WSrhstqq8lzo=>1p$Dv6D5QJ*$x7!3;b3Lzr z@}LIlx|RY6Lg=!%{$x@XHyWrx zmBnEqS;yc(4Z}$QHR!T9`6c@&=)krtZVQPCj_xp=pnq;ZB3{=}gPhY2YQQg*CoT}c z4O>95!89U%9;W85t<{H6xWWIfh8(1f*gf`BHs5Q-W_ePY?=gbDkS96^JGg?s@(^pG z2fQ4vO88~L51yp1^sih}3qjxq312)7YfnrL@}jlNC&Z2AP-2(x z1C@Y?-^$%NyR8Sm?jaGuz2Q7QLM;VN6BF^pBawI$JAfjftWzukzle(KYgNaF7)~kl zksuSPq-y2(2F@m#&=)wGwV~{|k}zoC6g+njNh<5QgIbsb&mH)n#B|@ur($#!;rnmA zZ&xt+&EOv{nC>8eB^X9~!Epx>8$wO>A#D6aKxMjv0Gi-vwzfMU@*lR9_z&-xyv*Z2 zsL@w!2kTUV9?WfuS(yZ}^+*ugLK1{LoFm=9utB*IcphSiOJS5A+}T^gM}xgG9UTO zoA=Lkp~br3iCxu(zF$`=EHUuJ+Ak~_PBc8BzC+wSuxrn-p28DrQ&Y0u5O-7FR*FSf9|K90q*KgmseRk_CgesE)A2Yo^R7)!Ik4^$A-0h=0EjpQP9#<7M#OuR$KL7Z# z`2KoA>IMu4w3t=(C2@hQs`7)j3s@!Qm3_M(`9^Z|FitUT4>$`lWoZgtzn%RD{d5^A zA=Cd{`lxAppqe@pzjq{_W*g(p2M_MPde254wjx&*_d_C*k~E!S{U29V`8O8v*JCLr zH)Z{>Reqo?m{s$f=oUduQqisAf+br8#UhwQ@{@gw=0vxc=w5WIDlj36B?|>+M7M~L zGoxE%#Hq!wqg!n3Y`Y=t)eS;X)3IEH&v=U&(apz!J%Va>ldF^U0i#x=+$|n#=R~&v z8|J#&3wCsiSlB__b}CCDt@9*iiW$)@0FBuyue-1EdNBLhA&!*@ZvVhMeY*7+Zd=80 zdy;)_Q&@_#q^x*ODnNGgJVGkB3ImB^#0k^KXzP*PwuNLj2t4IwFnx^Hx}AvTOc}=8 zcE~3SR%K2hfcC~zh~mZxx6B^zZL4^1+-07!-|{K~4UHsj$ibM=mL`+fXO%br9$aWJ z0pBfo^7xd#Z&CCoN1;@syOKL;Gg&>jWwtis6=QA<84jY>*-?dUXe4#(LCyuwpI%W) z+vCo4pFi!_y0lPN;0P=tSKf~Owbb{hNAe73sK@WfRG-X>Yb)gAtSkFMY#i5Hy}WXFCP`*c|hVR810zcV($oCfk3dPj2ok~q5 zVd`)RHOl&BN(WF7id*v0DgGb&Q6nU?p!g45J zP3;qPuflRL@5>D>EuOeVMd^6-3QN1F9vyE`bUfaoPH`r5Q9+fZOfoqT6IK6)#yEgR z^4_764j?E9N|-9kL9^iY1-fcGy#%isw~%~RhENB(kBg_1}drefdNURS5zQL z^wBu%RH}TSfk;3Hcy!ml)Xp`bipwBt#F7f$&XzT;AF}|Gw>WjAFQ|ZMT|zlj$JXa zz1T%c$ErDv@QIP1NXbDdHXD+)G^ zOR9lWE|~IT7ZVVkI#enk9Q<7*cExXfNXwFfaw|GMe(YkR!hy;T5ScFW@z9}&fWa4@ z_vz5^=M67$@F;wPqVUmR#HnV^c87z!C{}?^CP2aqWugPbhLt#Y9K1ns@Jaw;(4jwT z_6k+p;cpwS(tR*b<_!Sc$`58B&k-fm(V=3y$fL)oQ^PYCd5RV1WrCbUNt7vX%O`O? z!_Lhcm$j)8FwH`nD(57QhH19DP*I};)VRq{%*x;_#;brIWmUHx+Wyo0sDP8$c=++A zh$d+`iJMBb@aB=l&7>B-(IB$=PF^^x2cMJJPLbq|DTv=}N(MEY#1ZWl5^xWb2#Vz1 z?WljFCh_eVk7{%x!VDbh?v^dPx153uBXOg-(ExatofY)7>_5+INf<`rsBaDdBXLu- zAKx}$yS%icgY6Sl7!iuT{2LIPbzSjjh`XDBLEKwDs_q8krtBf20Jw+}l64@oYm&gs z&2=A*4c^bO+KD@XtP@NWiru;H2|Te4_AdSAK`s=W3GD0qQ*4Rso>$rYi)17k-rOe{ zh%edOq9?`U%aMLtq=W)rAX5y{3mx&*g`nNYO+N2Ipw(mw^1-@1CGn*rA}ekp?G*H6 zsccorlEOU$0MBYm5zkD~8%fDYA6c7a!Z>7U_qdwmn2a{6b3H=@S~6f}B>`$N4Vj+7 z2bd1Lp=&Mr1=lkKkP^psFIb)-Vo8Yu4{Fds0Iu&$E>yJ&$1?;d5=*;w&sMV*JfvD& z0cGOR!6TXK>~mF*i8m@H9!IGfIYN!XKmabmN9lBQ(CEy_;?e0wMW-tPjX(_`CWmSj zzsHNbip0_MzZW= zHys@^oWjv7?jBWfw|yj%Ocb@Ip({D{9_tM-M5UTJPGRd~LIt+9x6#UvL~eAwq$yEt z4IHE%iNq|#`b)X{0kqKtNRZq{dgk^|(1F!<+@m8J1)p*XJM_=(N5q>3=8$WrV2)_- z{eV+w&-8#%I8vFPzXa0$&UMuLNYQ!Mpc~!sGd2tCMZ4UTXMt5hJ%BajY0ttMovSD1 z1YE-<+@xv@QR55FBW{wl;`f=#dY!bcpp+x6#WL`RB&!|n8F?AGL*m8l*WNnic1UEm zREm7H8k3ORAu%$E@VfT8$H%BDaCn-avkA#Pq)0**^^hWs7_c&E9TI6{?|B(j*CL=> zQbTLeOx`YHp?Ror?-d_t5bKO?%uL_bt? zOHm2^n&U))nV1pXB0|oLZjlkE7DMci81BW`*u|{Fs&*O#LVAb9h|~tXQH3Mp2WPX(9LL*Fh zyRMX;C>wk3Rr4?4RS9pC+IB`19ez2 zz*SWKH+PjLYQuW?}KS|D&wKo->Ba%Y6;g5MB#`H+n*FAeL)~pAUGj_!TdfjaD?Z=+>k6Knum3Fc_Iywh_0Y z?V$SR`-=!zW&B08b*jZ=B==05uB*>w^y;rsrx>}qsi>5^-{Sj=0HoqsGPOiDPj|2#nc-;!5xqU_;MeZx%I8vM1eaxhLy28=w z(;^Afk=)_}!k*{#=!#J&m2iuz8GA+-in2$md!8=t6n#P$7g8N%9>4&ez)406d0HA9hjfT)#b9O{e>~K})t_CyNjaxPy(4iWrRHi zJ6E^LJN6imq**hjf@-JQLFdftqS%!dusUVy%1W~(HVUmX%eY7{oQIDNLiZw>8j#;Jdsq8tE`MLLInIs@lP&` zPt+7tb43nnJF5dMO#!rjTzQ#7#>nT7#0~zU;DB8f#-egu(-v2CwIcrXNZjP>U<%>Q zQ0J=Z{-|v{@f19JuM{30j+~+T_ij^)S@ACd^i3IW<`iFb*VJQ4 z&We9A@xAz0S!hC>OZLQ>75^fF&W?YPF{idOBmTuk-}5Cc6RGmz1ko+Osg(obdkr6e zt}IP}BOfZ9&WnEmBt1i&GvZ&w+OWV=H#Vlyo@{KK7yknO?9~U^Vx=;WSku8p;_AVCrNHu|&$XPZ?Y*%3uk64WO;O z7={hk=sJCkVXsdNJh?9Fu;&&WWdPEdlwiSq3gB8%0Qn29(d?{(tVsySrIe5y1JW*9 z7^g`j3YdyM@-&oJaS`Pqhv$_Ud9=ATi zT}1G-Fn6W0e#50~CT0Cw19PjgetYoAOX+bbhXCs~wf#+BvJ%o<_FTs`+WUmgrF>U< zS~H?Yrc@tTWf@(u%x*`gEKroCQ)-eopi>Tphqx}{AzC`6dPQ9A0Z6AT+MYjY?giu} zk-8UPE5r&?WaGaJ%IJI76tH1n9p;iC^%_sspu@u$xLkPABC}738_R)<%&JV<_6_J= z7YP2b$le&V==`f5nfOg0J@H+W2)_7+TG`;u6RL{OKTAs>;rokB8xZJ1Vv9jdIMi3q zuYs?SIg~iDBe@*P71B6f?Ytk9!3`)i z9doUjdX)krj1gHyF9j*dwO_yU@yFi%>y zy86`6F$axPsqc`d2C$YffT-lFKRY663o202J!R>b>iHn4=RNKMzd@YXg@Ik<@{y`o zL+=O)nez1q8zfKxejUOOu5DqqfHnBML0d+~$vq7D0BGUVyauyVCr1afxI$fSfVLn@ zW_t$3XF#xu!8?J!_)-^C><0RqP&2D9az$NS2#FzP9~jmWe&Y5!r}tle{r;I`AfMmf zOL-Qiy(T4k#e#k>i076!OF+NqmnwlP-&lwzh$vEY8XmynRcBznL+#or%dNVC@`;-G z#*uh#j`UNYWtJ-Q&x4j(o^#{XRU7BVLY}8wp zDnrw0wtT{}46vc5&jzq*{o>h;ZNnvBxioI;s4{cbm<`&< zYB2_ALq#GiXtP1dBMi`HquW*)Yf{`12ig?dieE#!X$)v1A3UJhiUVy%5435i#%4jA z$~Cm@q=t4hppB}b#aw|E+X+CZ|518qhs zcp~52dGOX}Z>`nq%nG!jvKA&@(gSV)TLu;4cOanQI#PA4Oh|(_Qg95g_?XRppAt_j zZ}78U21KMzoDRTomBVv_w*hh5wY9bS2}F%7>yj_Mt0|r%z1=@}oo)5QdAUMV7o}H< z?fTW*XJ>gt=Tp|EUlh+TpXJ#!{8S}KN51)x?WP(a32+|uFrE4;B7i!7ek7isx8ye; zJh=Pnz2%ktf63QT&NoACu4>yK9{K_=3|Kbxmm|6(#ZMiHXXmYeifhZOAD0F2=XcNF znYjYT_2tImtBnY=olhjiHsDTek!qnkbDfV5@5};}>#OJ?IG+IEndQna<~g5;{e`Ez z1Ns*Z>@(Z>1R&466|ml*x);J5t^o)DaB{p2MZZ{ASdefH+|^@_5D!tDoC*%#sS0zi zK(N}xoQ+S+*%T6U$jP6@dHe#ijV0!6iitU!O2izn$UZSgePG@qYoG^PTWbLVG)t2y zG6U!#FNOnpYO`}TK09aQv2*17XqvFb;@6XHEIS7_n8$_`7U+>0edIH5-aj{#73xc; z7U%(0WOOHjuygz>vhlw2cOpkdv zsU6UhZ8Vrj%^=bPyWFE6hyBF*g6>-JIT)yC9UcMzH-*z2EdD(X)+5iK7VFu%BCuI8 z+=4xLSyb)wW`dc6x()RtmXqJrgw2xQ!aaChRBicLr@ud1h#!ilo_Ft^pWZ(^y>s{W z`RTmQYIFQ%)_$Yh+P#eVV0l$sx$(&I=<1R9QY-Rm!B_I5b5_e zc16gFwPRO+M)MqM@&S&FKaW^dn0f;IxuUHxLnJ?T1+X(y)z;$c9TxL1!(MZsYy(Jz zbZjUf8w&X}tgb$5d{UwMK+FB~GF3Xx8|g!lsZ@2=S+^Jo5bp zU7V)723gG-vY31?Av=n-ET9#@E)0WTz#>d$cVMx7cALammeF8yvdmW!P&1b~76{0T^vwhHU}YGcUBf zr#RbH#5drvBc{}<<(c>i)6A^VTt+bdsx1XuROV&PCT|krQ(EVpuFM4-O4%vTNSYtW ze$%}RtD4eT$bP@F-ATzv?-+8X^wEl($x?j7g3fcf-yP?A%LqxdS>b%MI;Vcx07BJ|w%U?bQD&_cCy<(V(GD<6JI(ifPjItQ>>k(vQ;GmM@2+| zjB1M#g)x*j9$9;M0mlit4DurKXz9U2PeiyG zp(irh)Gk<|CpPXLLC37S`ewQaFNV3H#|O?27kVN%_f}j_$z)$cU}bZH2|;# zWMd`*rK<4V>(q_UHEtBwxS%8ANUK!7W;-Hwet~^sA0BQ14oTDn${MCa;BIMs)^MX( z!gd)R59{uz)@oqx4%F4tuhi0X>&~yMA2%B2R^PD;g$n7{ z)i+awyEAq5U4sr|Kj!M{RqG3j-2zOQR-k`P+&9pzlu<#qo3cCyvaDroH*KMR2|}5! zRo+L6K5Y=&CQJ+g68Vfpxq8tqH{~c7<6xl9e-5g3{Q@{4*Z{R1<4tND!mN~`ocYqa zZUG!=4L%EZe5H9NfL7dML!@7PCHWk)=H$UFc`$NCdTmG4wc+Hr z%CZCit&ojVPU22IEslr3dWYq_JdvEaiev#)D=pCDJn!`>e&$GQOXZy+u*C{X6&a>g z9B$Guh60vY6%|H-RqT#gtMtnYk?6C7$Txvqi&a`@KjiJzsalBS9*T{*k1EDyYwVE-cuH^yjHr@@R;ySL zF-*XV@dZI%MpOyF)lwIDQ5;nw7IWyRk~-689j{kvKR%w2zs~*U}D>E%1|GY2eh>C*!K|SG+LoM9#08g@;>|MoVv4YRSc_b0UPo6pOY5Pi+d;){=`mznevC zAH0R#6-%^&(qiezs#|5B7TYRXZ0RcJS;doF1iq0xN}d}7S_RlwiTVIdG&~E?@m0$z zZVgRze+ALX@d3VCG)coM?);t>ZEHbR(91F}vtt!EKGNEi*s+CMD|NRptm1Cc-J&sY ztLko%70x;opH|0RUkI z6~=b}wwB`O79#Ofy@zT*ElzO3LM|U z9mWq_l<(^RrYW!v692?>l8qP?Sk@s9UwD>TD1Fa{su>hZdU-82N1%-wG4@&KH_0%OHy$+@7%F@ zwCxfJi%n^|=m0^>ySSu7ZVIFkc2QNSr(9k`&{E?UE@}a%?TW`~la+Q1;DxGuYd#%l zeh}^hI_6>@_*Fzjs4~_2z`hkTQs-xl!_+tLH(7UL%tX?0Fq^Qu2(%DweWV4HGGimz z((S(T1HZj4Bi0Z6j;#h+)OYOkQ6YU!TNmIf=7Hay`oxQEA*!f^Y7uw%0l*~y#)aza zH3J_=v`oPVAExF=kj*AJNZ3cLj>NwPsrBGGNPaKzqXsp=)65$-J3A2{U4FPEJ4{YG|YOAU$DmdxEPn)ys<`Sg24&mHy0woRKp1`K7bs#s5Gq5wFkq%#J1@N0_z=L9AKw%Wwp@afL zO-@X1+507Roox`-QL|K+MWf6GI`P9Ttj}EN6my|`VKw-Gt*2de4}6d%K@9MLx&TgX z=nXyxd@wio8ZJ;b_zrxaJ{n>T#4o7pItxDZR&99La0PXp2R`()g1RdXd;kS?Axsk# z&L!c>kLw<0{iL9-GvEWYpiWH<00aFLeCSOT;5Ud$f5U(e(ijx@ zfZsb$9HIw60QLep^`D2yx@(d3Arye{KWspV{`e`IyR~AoJSolH7+^z~AvkbA<#`T> zaHWFfLDzs0{kl%j<$w-!NnO!kxugb^kh{FnalJghLM8phg?NEzoA6!GuxiC*B`;c= zeL~z=4kb1VKY#%a-UUAHb$=^&FYUG-{JMut)RJh5Cdg(bx^U_JH{Q3e0i(k@VX`E? zcq9@JaR=Z8ly!Q}3DmMGCtud;38E$^5u z5aqeHLZXCSS#_BWup2c#XJy%J33*&+02b60aC)oFmQWZJ)SJ(ipw7=)h+mx2TegJE zxv&J*`dq-4=zYL~itU(^lpd6%Q~@j|B`LiDESL|jZUex=RHO9853cqJwgeSEc+&tD z68KPn#k%O&2@wl2uJk3C)w`B4I8S4e1{la<8=l}DesKJ>O$|HIEZ4=UAp_44^2Fyb z3s*?Oh`koHz#merbP}H}tE($v$ZBB=oCs4#R?e|Z{ND?4bGJ9%c>VR)PVdiq6Ln(^ z^Mm4hmf!6RGd*~!v5#!(Q`Xq$e>a!k>2HN#i~q`LDpGUA3LSLN&PV*yBeC~OB~#Y@ z6y`6QzWgZp_Fi2!v>kI46ExJpitu*NZe9`p&q91x#z;M+$vb}`HD>$am9L!^&~PLR za4Bg6UPB!%&3}I+z9%Ei6idoUuj5W&o4r z)L|`$C1h3{hZ^;OvXeS;g&rcH@`%?A2BvP_VaBg0JEV}}L@t?rC6bQq)6)ASoxYT$ zgLWa%M;yb!w0&B7-627jTJ@Hx)90V^yJGbQ7PS&x$1SQ)*Xb2qC$p$is4Ksy@Ucgk zvNP~507&FN=RhyO((w+f^!=cyJi@kR>3x=tgo+?br)ZL7fq};!LmxpW?U=SKy(xy$k3DV;2Bsc+%udVd!Eehl1VEU1?6C*1 za&~mJxz2WNQ+n`n0Np}@b}i!6v}IAdf9EmX_^aUe>yF?kL3(HoQV*$dE1lY|d zOGf;dow&a409{zN`Bxo*!}_cN2IW8s7{sNvvn?{#{|jJcqifc2Eu~Do9W10r2}Ls7_K!6H*(^UKCn!t4Agg_8Ek~k--livQ_PcONL4Iq-;Pfo1Ge2@6<c~sS?HfHL}Xim z?|5KM`7DD>r7{8N&!dP0nEw=>!KGc5uHSLFO3KMN;SGc12#No7A^uwEt)K=bhzza) zhCsBtwcO^2q;LjyDb;Z;>P&8P_(^$P%PwArkRT&+n@asd;yNG(XC0gWxZmq_lZ&0i_~5>@&(? z)PWStTa8~-4l@w1{0(_v>zBjy_W2-rX0<64Q5S$F!OR$UeK_nIs<}a3D4GPihB~{( zrfY@rDqrMnF(a`CCvNPM&dRr-A8=?bjLMg--w#-g4=P`%^Ru?#SH5i5rt)P=?aAOO zUj{FemIKg?$t2O+%9StMN#)C+lW6rF;*RkuU&d+vES2^C+js8V>bK}qQXtK47Yf`JwCT&)=2apz}9Lq1X-z6TqrDwKV%m>T=LVW9uZ&@B)y*zpAh_0#NPEq`y3-Ont@_T8zcj(AY zEJCTEPcg{v|DqvMK;hMU2Y)#_?`e7-mu^K}^u(^G=)lK^T!MW}R)+W=7h)Bca*~3_ zrF8z4{K>4@CB0B2U7KE@8!xm9O#7<1a^n$c+DByNq<9cf4<%AwC^l)97n=93dCCh# zriXc<$UmPhLEH<){&3G+uxA;n1;t*#$?cY;7xM9thvtPsxfHP?u9P$UdZ02Zo!SbXCU;uG6jN&+Xmm*`Zw}w z{1xJaopd=JgKQPOFQ;Q05A2-%`|_>2FE45JF=iER#T1{=Ef18QK%ddAeYs=X6^!QA z5!{N02ReGo54Vo~t@u{mibvOR${fh9<7>k8z>o}a2EI|EZ>J$}5gclWj&b~TYc=K~ zc#q_5)f4hSvVv8#HsKe{YS-hiC;f!HMY$t^?qrp48 z!kO&!KJB`FY)(!B%)=DVYysUOJEo-!Y(b$R(3zLKC zjwq0uqjMv#7&oH1NR}rc)nRVlw*Y6$jqLiJY&Ad)X##)*lO5;>_o$CrmL2L76V%SI zWGDYG5I=DZF&^_3xkq~B*Ta|$(1X8f6ZsE_zqI@dkwl)Exrl!TXX#3rqOskgJD|f8 z^`Mx7aB}MtQHwukSD+M~vqg09vK*=yiL3jpyA%JDylB=d5ARl!7y8~tPG?Y;kN<0H z8s8zlefa@s8sJ06-!9&Far?EmPM6>I>|<9KOO<_HK7K?YC+Y@x8BWbZ1c)CAvibkg zoTK;o-esRrjENPZ3hGsfZx81=MF1kAtVo&7!+n?=+*BbkXw>FZkOMX`1InPN>VqiHit8ROO&WQO5FUPP=jqIt>JnF53cMY%lR zauaSXZk<6Q7!j(_Iu+w0DW1!u2K1Jge!Km*%$ZoeuLW;L)%dH_cEqM=#RYQ(x47Z? zou2jGvvc)taR+saOV9e$3B#g0n6h6oCpDiFJlJaZVNDyzW-~cZs*Et{9eH|o@eNmcpDPab zY1g5lZG)$DQm54Uh$38@TiucN_a>i{JMtY9bV4=JN#N#cUTydtjIkqu%h~CCPJPFk z12#ojRRMCBb&$%LS0e`Ec6-qzU4sUq{M+lz$r_uDowD>h?|<;tXK%@WzfWVn4IV1s zjr9Ny@5tUz2Ud6e>nh{F6^v_g7yPpKuh;|XUlhybw@nYIiNpn)+COzu8`O5y%;hi) z*@R_4m59-li6`Ia!B2P-brkDxqBctzsppUA2f0#;fA*vhuik(Ctv7DnIep{7@{yNc zf8~wUxBfSK#ok(vqx(Va8T%V)Zt#Z<{GowAwD5-x{?Nl8w(y4m{s3pNhBH{h8LZ(9 z)^G-EID<8u!5YqB4QH^1Gg!kJY~Tzwa0VMVgAJU)2F_puXRv`Y*uWWV;0!i!1`V7+ z182~{88mPP4V*y(XVAbIG;jtDoIwL;(83wCa0c=vaAzScoPm7s8T);GJI0M<58vHXj1KE=p{07dTgENr5qQF0cGmt%k+;3#lUho5)fh@TL zzkxH5rB>iKa0We`K@VpjxrOd;wr~cLJOzFOXRw7ckl;P|XK)4*!Uey9GuXlzY~c*# z$pV&a3*lgh{K%dxb;ckRJv4K9Zfj+T;KCyv5 zu^FI0fIhK-KCyv5v4K9Zfj+T;KCyv5v4K9Zfj+T;KCyv5v4K9Zfj+T;KCyv5v4K9Z zg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%` zKCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9 zv4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXd zg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%` zKCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9 zv4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdg+8%`KCy*9v4uXdgFdl?KCy#7v4cLb zgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl? zKCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7 zv4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLb zgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl? zKCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbgFdl?KCy#7 zv4cLbgFdl?KCy#7v4cLbgFdl?KCy#7v4cLbhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jf zhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~ zKCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>B zv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jf zhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhdxnOXFB+-hd!}~KCy>B zv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=jfhd!}~KCy>Bv4=kK z|6}h>pe?$-g{Mbt6mRoRjUWtR;w+Qn*?I+ zx>fI%^w697y<3v1F$)>n*v11h*kCinIGBVuAq-}4zz`M=*bXcVI0-{A1aM-L#bh$9 z5OVf^hW`wEpS}O{|0TRux7TV*Rl0khJ$z@MefHVs{8z%5cqNR9SHhThC5(wz!kBm^ zjEU1QCQiecI1OXsG>nPUFeXmJm^ck%;xvqj(=aAZ!nPUFeXmJm^ck%;xvqj(=aAZ!nPUFeXmJm^ck%;xvqj(=aAZ!nPUFeXmJm^ck%;xvqj(=aAZ!nPUFeXmJm^ck%;xvqj(=aAZ!nPUFeXmJm^ck%;xvqj(=aAZ!nP*q`zHqu3lW%4Z*MjBo(h6 zijv2*cbwJj?vI$m_(0w4Zz6?l!;S9N3Ys;~>Q1LeVyn&8i*}nlYP$g1fg^^L=<1R> z*LD=>K}=6U&P_YD{hfB|1G@*u$Fudd<@}{%*ZK1$&IJ!$KX>N-@rmoF&5uG!?k6Sr zjIuu;KltLU`R@Ga#XEa@^P{_yaHz>yV7d0(V!rRJKB72|f8k5ix4lLEu94cB&35*0 zFOC-bw-&SI?b-HxIiD?FKU^H`>@D_}8hd;5!`Xa)dvdg*v}@%;UhbLz~_wkFtF-aUVE@{m&QPqX#p&l|~~HilQsN5_j7 zclKY}U3~oY`QG8~BL4otqt2pduAkPRUKyVVZFzR}eAbS6TAkX`Kbw4mno!KOm+v$8 zNbe2%jeE;}^WL(*JR^PAA>1m_?%D$PPuRxgZ zf{xrjPHdg9{|a5wmcc8FyN3QP>hz6SvTP z!!(4~)knu?clHnOEN460$75ywb#H$B%J|`(94}si zpa6eZeJYz@CDA9pX{7GIaq#T^a&u#R>SOcc<>V6Lq5QA5jM&Z1kK?V2t4f`E@#xl# z@yf-ss}HF&H;(S?-Z}z8;(E0>>%)liZGkv_ ztg|5&hq!Az2_@O=VVKRLZ-~)D=_40i>LJOckZVz@!s|k#iGHH#S&c+k-p2mlCJ$H; zZQa$#w7uyJ^u9(Z$4Eo3W%Z!PV+UrR2$*eP%r;_XgyAP4pG}OsEdQt*=?us^OI z)L9gSQyaxksTCcy-+Si7_z87EJvP%%s~qWYytuV z`L4#~(Ma6 zvDGh7e?tqfDj?y8*L?Kw2kJt~tNohe&Bb!AM?Pcr=X%^s)cCh<$n&_>n*kA5w@jKZ z(FeM&W3Enq?gph?t$AoYzi*^AjH3Iu4z~3t0(a+kH4ZMW&6oONIk5KH@nZM(+UM>p zj_$6#vN+mb>}o+CaQ6hj>eH?aoM>;y?J-+7KKPv@by3b5u0Ds7fsHPABWL9@Hj~Pw zw+@KKE=uKU$yvEHPwmU!JW^jNQ5YCoYq{8QO$&v60W-saqje%t=+X>3psQH{5nDk> z6QL_);6x+QWrCd4i-vWIE(mubbhYHHTqf}S-jTY*c+qUyqpx`e!QrYF?OJyZPCUw^ znyvQe@8;u0Bwd2DslcHbJtvuFswMmVBXvclbsH#cAa}FivX<_HH9R;8HDfK|A9Nov z)wM$+)>_?L3Dd}bMT_{Mt|B%}I2ko(Eh9NAbf&O4+FNYz=;A?&Jii)%boi_w$(>l_ zY&CM}hZ6DFhA>X&D!3I7ZUiwalkw+wjFe3{5`RVr&eqHgSOLseBT^tBgq$qEP5bRz zN9suw77@$TG7$Ekl96$8tec4c-6Ms46Z_&?hHN|cJa!;7U?1`(NLhJIRM2Z;QNets zd4V8~;<-_#OEkeqB%26IWL)dQjB5kL!q&th1I}8&m7e_s>jX_3;6ZHm7sy$;j2DTh zlre5orXUmCprTTNoRzB>l~Bg1wWz2>60)E2SklF?VC6Hps=%j_d@0r-TjhM3F<@mg z`Kr!pu)GUVyp6Rlk;JS_#-l{;weIqa<`gX7QEtLofaP8& zAuEen1&G`$VVPP6;y;#}AY`l@Ch`-xmuE=MzfFh_SnfrTvhtY7K;&Ks({!$FLO8{8 zFN&6xs2E2#gt^xg(1XR%A_Gn~;7YHJ1nUG%Q$Q0o`wQf(Tn1G{?v*h{bFV2lQY2SP z&dOzcNkk=7uGCt9cPuKA#H>sP@x(fvu?b)CajER??&WaMfTZx0GnPc_u{#* zQWkf(*ADwcgww5&uXlIpG!jO!&P`)ct6a$%vg$iWFmadv&O#2Xbih`}E2%fzL@Jc+A_OiG&RTj_GGMAD)7?v0*H`F|C{gk2 zptsybNMNy4jN!pcXmVJAXCnzS%t0EC0;iWV? zJt8+N^3T@@*d}AS8PA57v$$8jjO8#05 zdPaesm2A`0Gl?Za4Z#&Y4MG&?;aH!Jeb*9h1qW4Rg6 zhL^LrSH2?4%`Sm}_&h-}C}C~}Of?G0sE2S)5qJsq@OacCs9AdfODz?Vn_b~@vrDi( zV7VDe$xCBa1|m00cxKjtH37@b5ISBGvlGfuGzf1 z1Z4v(H=~GonTnBhD$UIEplL`1Ge;ud;Mnx%-nwQE%DPjfB_=cHPnEzacXeGi~$bpy6L@lW|@!ae(L@yD} zT6$hG6UB(!EVNscx8%9mWyor=+>GJDOK9?1foCHLGt5C|%Xv2Az)NRxTOv0@bT6z2 zu*bu4GnNZ0o+CXMuRyfGax;RKm&n8kA~!?0X7la}1XwILqlkH#ijnoIG&gH5bzzaU$bpp( z*wV3p;GLps3Vy_5L4lr^%=nU6%`?t;Zr0qB5=qw5^O6;#6e?M2vF4H)7NtmPRw{eD zj99@lzF}q+=4Qicx);_1*yCZj8Ow#2((Lqz+^onyUn5|f zjOAuL8(z-hUim7P0p&<+uI1zNq&_#ZxAKX)87f(dz~)ju9`z8pfSR=zu+(vd$jz>D zxmj~j70b;~N>&47o{Lk67FcdZ(DD+QI6>rQ z2-j@hHLn3+xfw;w%T$c4*QB{wbLk6axE zIPlV$+?L4A5Z&qZpt+Wh- zL2@92WL7eJYoCmI2JikuRK~AFZg!2!&6SZ;<=^3s@x+eM%=Aj}UGZwo8;f@0H0fZft6(X)~nUy3DblPjHS zc*&p)DwC|GXC*W7Q%5OmR_8hg$(ff6Dvj8^K$@DB%0wwK&cf6xv>0TwGMt4RSm{jE z5__{vu0q)$dda-m(zBA8D5i5WY&DDg5hX4@IrBb+lcX(fh9W#z2~A!r@odCBnWfOc z$$K{Az)5HOLUnG2<$aj_vIW6ea1-`;Skl9DVWl)XJ)N7e*tQmErF4sg7ZVDk9~L?~ zVZ+L4cJlN*L6Vppi69^`Pe^)BGHA`xyDK{CA)FHgUIPCTQIDWzr80iix!L=1heUYJ zEYVNato>A&@4q&Sh8=ux}xxvH%?`iCGPFZiZ!hEyos~$5SxNP3IIErATU4Difu| zI14kYz+%mzz6@s}2TnR*YwBj`9w5`Y&8o{c0-XDMX1tY%nE%tE0IY%Y~KF?DVMItjIrG zBcMqM&CPf=tepPrpO_~|4wGdZO&1k0Pe^)BGTTc< zD^zagpL{2t^oMqV37VUsl&p=+%0T623D0DHUV+RN&07#URuZ!sP`R1sNY2G85G}AG z)jC1TN@U^$m75`4(|LCV$_8j|MiH|znTo+HYSN_AhX)6{`YVJ?JANSJ`_HLIOfa9lw%B>;<>j%SiI2yB<32c>{r=_E!>`OP?YH3d_mvNh_U8J#o?5qGsUEZ>R-tk5M>kmXcXp(Ag})1j(chIl(1L5; zf7w4wvqyc;c5S{r-<$1eJ?(~s`=t{}`g1F-UF2Erh$3<|UKOkHYBx1rbpVp7@#-3; z#^3W>tkw$n2YbCw?97$@-l^1$$ol7_&BSRx^t)D)YzXzhS6w9a9_vO_p&stNzvZ7+ zYQ2GcAi14SHW`up+Ig-EiB;?)so!XsP_K1T3w$P*(R-{5NoTciZ~Zv(MeTUqZXK0< zMH~I0^O5k^Ru9_~(1HNxjfD%T3g8rxP~IQHUw2Uf_X4*5v<7H$mDUYUrAWqq>HHx5 zxz!>I+yr;b6Jj;c^#>iu`$DD$z5Wo~F*Uw@MPV%}#gop~JvPuJUtTSIwcL_Ae^t|4 z2Y!u#NB8lEH3hi1QWfv>YWjYrz~|^2E9e0cfr@5->C+Re?!Xf|L{w5+BRng|Lm)J| z2G*JYuu!FhM?xia9~1OLK~VAShJuW+6AE(bE+)vSv6vu5LCQ^xf{d`23H*{Enh5#W zm8%+=c>y*)hJt!ii1UJ66pw~VQ}{?wX$l+_(r{?>iKBsNS<46= zA}Xo7tO)to7#eC#AYKttF+38gOyF@KS`qTGF%$fj6@@irhbw|9Lq-+gcGqCbO60mg z6FW5Y;9wE)9J59gWEk2cfQ)K|JUV)C=xY)@RtUmoWaKznaAA)kbqJ}Z7p@@W_!v6s zO(9u92vIyTDox=d2e*Qd<71}y-W7$N_#&~REn>jwb{jpe%VsW0?$FVThegQqedHM4 zG=P*IYI(Hu;?dtUcr5w;6(vM*>-MCQ__40;3h5ng`Z2NA`AjcKFy!~m))v1bm^&jHbxe^z}vl#1_qY0cxJ5pICpGKGX z$e)35s5iT+2)^tjtoLL`su~q>Fa8Vsnx%J2o6E&J0Ic`6Oys;-L21?fc*&!vpr}`C zR0hv>7TkNcg0fa++{^z4|EMD3urTA+I^*Zkb`rHF52>7=Ye=oupO85^8Wme}tJDTp zcbDD!x`wn`ZQR>LsUvLY&Ti+l**3PxlaM;USJPW7V5JVx_!t7}O#v{dRK;VVn!cYY za8wZ655ly;(x?rAZ6J6;hlomQYXocuCKww-L#+t_3sp*ZBvexOF+o2Rr03Gmh%8bf zBkY8NoVtq%a_T-PNV$nokP-GWfv*9JCPF?oW&&a}noTR<7DobXL&=dj^{Z&D|FL~I zJMM6*gYPTE_w&DswqA97?n@U#a{?J{cD4GV#S9^J-mj*&W)S5C*!UO<>P;ce3vy9B z8Y)fUBSEDpa8yVaMWas~4MZzKM(7YxN!?{d$j8ReP-_D5ija!okx*psO8bpi${Oc;IV`_FfW=rtoUi$U`IA$dWV~SOssXVege597;<`= z2cXkOGmo8qOa?Fyjvg{)Lyl(CTm?_MI6{Y)N@{BfJY{3Y#t>3#0)R%95*`zk)O}2V zqkybqAO?+d5qv{lCUXd=qP5E7bqr=)%=>k!gYPTE_nxn!tydlQI&sMEAI=EhylBrg zu*32aiSu@6I;$=oIHSZW3QXu(Wp`D8*{&*rFFOhAJ=u|}Mn&9<#NoO(6tx4uj>yYI&YKmKR$Z*0fEg3>YK_X^ z+0KG{?^aOOs*HP?I9NxFmt^eN={Q5>{9Hq7wZ{4hxKS}rSE&uI?k>Ccbq#5?+PJrg zQpZN=Gdb^Cd+8P1dfeFUO&`@gvz(SQ09toAyeN4~~1?jmo)lZOlA|vdCf}FaG33BQ_C`h@9 zQIHY#GJ&rFizY%oHf921Gm1Jhjs)0-k|T5KSJ7JkWBYJ++~HIQ-&ct5=YJJ#z3TYf zmoA3p1Txy}N*6MkonJ~oDiS`&y@gj5WVgentw9Eetgd~D1FL_a|oV(C(tjtJB~$H&l7Zwko@ zLWts#QE3VvIk**s93L|U(NBO(J*t|B9e7L09XfjPun58W3B+;?ZyG>K54Aj6dhzIQ z8a$Q|2j)d{C)G*p$VN=>aMO>8wGP%#AeRI~PEYdyboyxKvD1&q0OrBbL#AxV(QLYt z!ILhI(BY+$+FAln+1Rl$gw&b2LE~Hm-;kHd90ICnt@3ys zgBcg|ex2&z`wH>B=c{PzRmZ(f9I|hqU$iT}Pf#Lp-o_5ui*#09Ja9&dRTP-I^b-`u z2M>1?*?YG$U7f19SBb;+MAdi}W5?z+f%9lbDyt;cPf!>Oo3mY21YdR%)_bxeRgH?c z7m34lZzyU9fE|&SiJUhpD6P6!KLIl)=G7XN!LywO_uj3btW_EJGI6ku7%$1#vD0ye z%K5p5)M}0O6L6zqo~}|GT-{xE@9P@UYPE516Qz#L`hEflsq;Hl>JZUeD`2G#5j}>0 zdQ$)lDpm1VsHX2{3LF*0_Jc5Oyl!I~2%gX(qLSJg0o#EI#>UW4YXZPRl@cBamDGJq z&<_Raxir;Jka;2_?1X}xx{C>N>OLq)xrtGb5%w~HuK|lDLOwQT0%9|YIx~(0*oKlL zbLv;oTK{AFaCY3`R0rQzi0|ir6>Yuh_}rH+hUNq^+U!afLxj|MU$huv^wtcbya1!e zP*86QabA#%;?YoP3Lgn7O@X6Ax+ofb;%FdR)-pneh)U`%D?&athK5=bh*yME43C5= z6L=hmR)lF4fJ{BAnur~E zOUWHNdhxIb!TJfratv=8KuQm_JX(73=x-W4mJkQ#MROs zC3PPY;3y#L7>GgRTm;{cm&qIgs%Wk9cpZZo7xR9d>frkd@xAA(XzNwSy-xHKY8*{&*rFFOhAJ=u|}Mn&9<#NoO(6tx4uj>yYI&YKmKR$Z*0fEg3>YK_X^ z+0KG{?^aOOs*HP?I9NxFmt^eN={Q5>{9Hq7wZ{4hxKS}rSE&uI?k>Ccbq#5?+PJrg zQpctGegX-p^E+1R5YbyJV5JTbJ%)gKQveJqRqDC*5*`Va)O}3Q4+ZJDG}TX#c_Jh1go2#9iwSb-J}5}JiBXUd z_A-I50gEO=J~n0oVl#?5GmZqQ~WP|6}`bcHH4q2j5qS@8^FNZN2LF+?Otf z<^(d@>`E6ygw%Oov>0Oa)(oP&0HeoHP;UxxUXY98(NJj$9|nj@XdqhF zGD3%lO6o2vLOwQzhFTMdSAT6*#5ZyG$7 z5C`T(b0^hF?8rt;?{L$PiM0;aPau~BLrzcg0Cf6j=CRX{$pGfT(L<(e$kA-NlfjcN zj?m$ylG<7VPubY9F@)5b0H9H&gvUfBbsrPpC?M+?h(Y6A1mBRC$s7WzXsz;i9fKJc z^M0M`;QI>kz2~cF>s80SPV^I8>fTRKB5~fv4%v%zR$V-BMu}Avn7Z^66vhV+cNE!s zw=-Rxs<>B)!}dhgcot*F<}`uxXh$lmB-T$*7z>-TT~!2Mb`sWmvLjWEintev!*y>c zY6pNFk(Y^_H!CQux>!E}GbZNM8kNDbodx&at)Q$`8TT@Au#Oln$=I>eafZtIxrWqg zjr9|7qhg+}QX5>|U3Ty58q#XDac>i)j?4A^1QJr`cdXPQqPJGSN*yA43<33~02oxN z;;~Ro-_H~{Dv0d|VcK}z#x@W^yHqaY*fWddIV7EOeFY|I42W)yX190{-uB}eAeucEd7 z$M)guxWlOqzONA9&;Kggde!l{FI^1H31qa{l`e(|sq?;QF~sPt8AN#jMvtMO-W1}z zAQ#1>q0$sS5>%Q3M}>4zH2TERK(wr7gbopv)LmADd~6I2wI&d+2&otz2~{TWI1sG} z`Pi5Vh<<`F#L}fO9UZWvZyB*eLk|uX5m-NgP==vR0?4RV$fKhNhrTAkV}&4WMktA+ z1sC=xQiqUgdf^H}j*p?E-V~A*gb>9eqtX;Ua&RjMIX-3zqMrbndQ>$LJMfm0J9PBo zVG)A$6Nu#)-ZX%e9%^~C^y1OqG600aMb?GN4j1L~}D6;o%XSzC7ajz1G?TM=KEXIz_X#(fbj#O4jte>DT z7B*+QstCU9B&_#jN2(eXaW4{w>)ue*4gfnMFB3U$R!~}Xv3>$(Ow6k_DuZV`3+}yJ zL0PLZ?q%X&9Wh>#v16y>43+b94XM={>nGqw#XMc5Hn_UG?B3Tkq}6KU-X=;NSL*u- zB&5#oSgAuqZ>@lpIz;ps0_sfxFsM|;W1*V9pDA!u5Ze#JwDG!)Z6J6;hlomQYXocu zCKww-L#+t_3sp*ZBvexOF+o2Rr03F9KSAb+jIa|5a_TN7$f^6FAmt`TK}Oii1il6= znh5#Wm~J zyQ~QL*ccjWO(0$oQZYOds!ZTdH4k6X_!WD!ZA45mIDI_ZhA&N&vr73*m z;8qZFe9RO?KLIlJsA?j1;4LM0=;+16A_VIv5X&*VX#gob)beQQ#iPGz@K{0|m>12R zR41_`8!^4ZO+O~qI#@q}ToMd9J7$v)PCq6CmnX(~Av*}I-Pr5ilhnGrf zYY9AMW5>o2QfmT$MwJpC6P46`On{?+tYaVsjdKxvLtZ9x2&kg9%HwqmW?ansb*h8! zE5!GnucED29rrrXPjIDsKS7Dac^f-qFVb0c@xU1+R#9N;(oawrA3WSqWbfV1bakrY zUL_9O6IJ6`j2)ZP1kR%!sjQM%KS5zEY|eI75q#N6SntV>R5dE%UL+3Jy`iWb0Cq%P zCUV}aptR~@{RGUIm{)652G4dD+%f!JtV!R|{$4 zsSduc5Z}-LD%yJ0@wqQu49y8-wAqy|h6t(izGyMT=&czzX`p`hLr;=CXi#iOCp z6h0DEngU0KbWt?=#L+;stYw4_5tYt&p-lqFs8-0MqX&n+Cc$HcAZ$h`iK7J<_9#+^ zkZOA23PO&Lp`+duk`;sy#UrEA6h3lrD+oD0W(uO80GWDJH4!`TmXbSk^x|O=g7p)K z8*{&*r zFFOhAJ=u|}Mn&9<#NoO(6tx4uj>yYI&YKmKR$Z*0fEg3>YK_X^+0KG{?^aOOs*HP? zI9NxFmt^eN={Q5>{9Hq7wZ{4hxKS}rSE&uI?k>Ccbq#5?+PJrgQpeT$egX-p^E+1R z5YbyJV5JTbJ%)gKQveJqRqDC*5*`Va)O}3Q4+ZJDG}TX#c_Jh1go2#9iwSb-J}5}JiBXUd_A-I50gEO=J~n0o zVl#?5GmZqQ~WP|6}`bcHH4q2j5qS@8^FNZN2LF+?Otf<^(d@>`E6ygw%Oo zv>0Oa)(oP&0HeoHP;UxxUXY98(NJj$9|nj@XdqhFGD3%lO6o2vLOwQz zhFTMdSAT6*#5ZyG$75C`T(b0^hF?8rt; z?{L$PiM0;aPau~BLrzcg0Cf6j=CRX{$pGfT(L<(e$kA-NlfjcNj?m$ylG<7VPubY9 zF@)5b0H9H&gvUfBbsrPpC?M+?h(Y6A1mBRC$s7WzXsz;i9fKJc^M0M`;QI>kz2~cF z>s80SPV^I8?cPsNB5~fv4%v%zR$V-BMu}Avn7Z^66vhV+cNE!sw=-Rxs<>B)!}dhg zcot*F<}`uxXh$lmB-T$*7z>-TT~!2Mb`sWmvLjWEintev!*y>cY6pNFk(Y^_H!CQu zx>!E}GbZNM8kNDbodx&at)Q$`8TT@Au#Oln$=I>eafZtIxrWqgjr9|7qhg+}QX5>| zU3Ty58q#XDac>i)j%)S(1QJr`cdXPQqPJGSN*yA43<33~02oxN;;~Ro-_H~{Dv0d| zVcK}z#x@W^yHqaY*fWddIV7EOeFY|I42W)yX190{-uB}eAeucEd7$M)guxWlOqzONA9 z&;Kggde!l{FI^1H31qa{l`e(|sq?;QF~sPt8AN#jMvtMO-W1}zAQ#1>q0$sS5>%Q3 zM}>4zH2TERK(wr7gbopv)LmADd~6I2wI&d+2&otz2~{TWI1sG}`Pi5Vh<<`F#L}fO z9UZWvZyB*eLk|uX5m-NgP==vR0?4RV$fKhNhrTAkV}&4WMktA+1sC=xQiqUgdf^H} zj*p?E-V~A*gb>9eqtX;Ua&RjMIX-3zqMrbndQ>$LJMfm0J9PBoVG)A$6Nu#)-ZX%e z9%^~C^y1OqG600aM zb?GN4j1L~}D6;o%XSzC7ajz1G?TM=KEXIz_X#(fbj#O4jte>DT7B*+QstCU9B&_#j zN2(eXaW4{w>)ue*4gfnMFB3U$R!~}Xv3>$(Ow6k_DuZV`3+}yJL0PLZ?q%X&9Wh># zv16y>43+b94XM={>nGqw#XMc5Hn_UG?B3Tkq}6KU-u|u?g{}MR7iA3vUv+bFv|L;4 z-#XY{9Ifrn?=FsJB3kG9dIr`IUt28!hJqgE04&sM;}KEMpr<)-bbQN-!r~F-;dlZ6 za0kx>lnxoShT{$|2ifS_eChGL?h^Qs+Wrm(;NDr!^!9b=*M%fJwIe8~DXd{8tu6FO&tI%121PgRXPp!$G; z8V!6rs3Y#7J{}9i3#B=tjCRA?1hRDwp>-&zXJAbs%NsBWFf{Zqhdgh{h4HAUHHVK3 zwdTOlAzmh}C@G#>^dx1@XN1cD&cNZPCl_m>uj?lZhM>MC z0tod|&SR)2mmy4qqlixHkh9)&=R@RW9HqleExEM@k=C)JV@Rns13;us508yn@?K`Z z(LmQh5Ua+y3bB_j(>Wy65nBTgbr5D`3n2r6%r8;)u7kR-3ZjTrB5n^X>WG49+gFb6SSVd9{Yr zs*Ig%bE9Grx=L;EaCh0gmupC?)yBO|G|^!eOfr1pK-^BL79=5=^LG`k)g0>|^W^hK3&Ikmn7#Fdh}P z=J1iB)*LuG#0#cTDUJ@Z6)&T7$fzanv{K}wW2mS$gM6h(1@VZeGlR#2Y^BIY$IL(- z+=p>Ds=uVs0zGw?Q9D%hVPTPh_A>~j7~Cv?kb0dwO8T(qZ5BLsNJ43ZmN4AljvEyl-{AF9}|ld zw4XsN$vrR+Af}II9yR@#3}7BSwor%lMVp6ZOKl(YU`Wis;io4TYaz6sK`sl1puQ#o z2=!9VW2h&WAxwm$h)(N}^WZd>A@VYg(&46-+**T3>)6pTq*R*$AX2A?$3`uAFEik1 zpz9!rRpVTR*vpsc91`k?tpSKS2s1KYzXIR6MKE6MLG%dA2K=RTS-K zD2#^9;m&G;PrJ$MJ=&G5N=@9G)WN>D7PVtQPv^^2&Z{+~R%NuGff*I^Zk5{L;qJ0~ zFV~P(tBre`I@Cujm}L0q!8}9e{9Q$BHAnjyxN$L$*QpLJuMpq+yo$D7b=>PeJW}s& z`pFIt4tDhu3)_qR_JtErpUi=*T5i6`EsPVXKZA1{u_532{&1Mu%_&n@QrTNj`B zM)kl;y9YPtyR+lv;&8k-ensu52ZO)Tf{uefI=j3({JR$x^X)bGf1jNn?aXiPE}lQ! zy7=U1<6Ys85Ys&I+fJ))jzeO@?SiwBlU7;d3WogdMYDFbX$9|VY#)g z-YfaJW?PULk*Ae7^~BW@r*UDF@kdft9X0*=74?oxS?`kBdTDUlDYv&Su71yo;&6i4 z$hcs(u8o^J)c?ozlD!Z@Yk4i!xR0qheZ~cmFFvuM-WFJT2J?P)a5%d)zxDEB{J46H zdNlaYI*e=BKk-g=c4xUbnlBIZgnvXmq|QBS|M=Mr{hxpGL+ar@{qEE5>{UBMkH2p~yY8*A*=yc7LB|!pW#PYZ^cx4klCYI zY?BgchQ^ENQ#(OVGlerJ+6JH=16uKPr8#J1WCYGy#vjCyR%D;rZ$pO_6ly#d&bPpt z#u~CI1=V%XaQ55WB(O%1a`FH_8fysC)K;692&_@GtVEh2jkS#qSi5qKz#7YhlMy)6 zSSzwm?KizhV2$U(NeQfJtZleTIvLll!d%7T8bQj*1N>;LAxu+SU4?mx#WjkSmB`FS zy*goWZLEDte`|(E3(h*M=hZ|7goyR8oJ41?HaX)5~Q3wz>mfn!Zfwj zHEIo|XjzHO8mj$>VeOItYv)|Pagka?SuRe(oL@(6SN%Lv6X=if3&I*UVn))KW?j^D>#Wl)&8OE|^<~35q3d zEEir%Ggk@B75V2jglUWI5_mSeoMr|Sn7hnlZUYh$Y(XVxS&4uljX8vCX0J_1Kd>d0 zB4%X*j#?%Hb62`xu1R(=%&}Z}Da~vpFjwTC+Ysh0hB=-MFQ=Ko1m>>rn7agt35Gd> zmX!z?(wIZIX7;)SX$XcnikO$lWFG|Prd=?1iP|)0)aHTZ!b@ppD}lKp z|J;Vu_JL=^%V}mXfw`+Z=BCsJf}mw30){l^5U!cMrqmXKBIac(?jedFgR^xcpZ_+S zdu(lGxv)~&eLjJ?BLB>W&6XcqTX{CToW(r^^6`$;hS#V)1VPJ61Pp1+AzU+iU8D98 z6frN8-$P&(eElxSe7S@p#)ExwlKSE$%u&plJQpV>RiWPWdkFSJ5m|}a8PSUg^A_!d zE-KrwVdeCD2&}e`y|R)NfSddTOP(lNP9k7vF!zVJwO5mim{y(m{DUOsWU?8E zURzPj70d96y)MHXMKQ;6VWsqY2o!Tg{+SJ7-lCY}*|2i@Jp_t5q)?yOu*pwQ%n`Jl zM8L3sIfQF!uPcy?pqQhGS(%y)M7NWiJ-)NIr%#e~N69&*)EVd~frHj3e(|4=)LWKE z^PPQY0|`!>f99W$#(!P^N96eTwS%MWS$n=49RGsDv*X|7=T2to_}4i0n?~w=>Dl%2 z;FZPxYG8QVS)6 zEhE;$ULa@XGP_2pTO@ywHffJ2HmkDxGFSbu7gpxtH6M3;EXXP@1@Ar-r_94vB6E@)t zTc)*a&XZa!W_(j_>9e4TS>M$buRGW9R^ERnmWXd^-S*VIeM zS%KU*%;ziQQ24AM$(?h_*=pp{r(k$&Ll~zk0TjdV;6@O$GMQ}iJ4VVT9C&|MGC$pJ za06C=HDqh#o2f!VP8Q&%*UE1lsV8}ssbwJSV<~(h36ODetec4c#h$kf63Hfl5^G|( zD=TGO8xY>H5L#rwc@MbK85hAiLDL2VcWm|-$XU5eY$u{p#<)$Hg0*^sib@4?R<2@H zLK&ylqM{N>%*teD3J|$h!ZNiC#D6UJLdaM-OynnWFVB#ie_t>?;6xk^JEKdv*A&o$#nB=I&U?U> z&b$XU4zs)#i^V~plrQ&<;8a<$~FTxLxqc7&*0skLA|#G(>O%*tfe zLMgt&)G3Hg&H26vUm*ilwqkCdsoW+{p=4BUUm$1YGPymGdxean)FfN^rm*s1xfjEN zmCvkx0-r|mrC5U%kn?HAfR)Xxf<*3xC{D)yFWzY>M_(kz$q;yzs_AcFS&kv+JPbVP z+zVx#B9AKSlf;}%Q&Zz1axZ^phRD6H!digkUML|a3vko+A#$&TWn!7D5dX2<3n63W zFp;0gy*xv5{(ZsRi6Y`?cvGm9dtHMlkHyg<1I~NEmCn5g)(M)fL4?P0oB}y3mx=F0 z?v*ialc!+KzD7l*0y!&JF)E>q6Kh?iq7q5W$pjp=52g4DQ>V~c5S>K$3K?*+LHAq- ze4)8l(-bJeS4+;yWug|5dxfruQj>h;gK9LEdoe6H`L3FrRp8S|z7%VawQ@eq7_hRL zyp_nk5XI@(54$$3WtL^a%4qg(M88{+ecBjx6*j$CzZ=hmm9n_=-K3MJaNdG=`0OB4 zH%_i=MlNE9iOQ8A?;3a(+hLN#tW3tE--(|&`a7ga?&Z6)FT!GgC0{6|6NbLkY1RQE z`$~AGwt*D@%f1jYCkS<>cFn|oBKz_T$r<%N1*Igyp=V%Gu7Ei5gRB@s}&r4>ayi~F%18XItf>S10OV3MYqSJ53J0eidnZ>B^ zL{jronfN5dS(rrWXznex4sl9^vycNT9kA7TnADqXA{EMZ5ds&s)MQp`=~>Busg_K4 zFJWC@A?GMj@tK&n+(k%Wu~dxV!AodzSb=9F2{X*WrknF@#(|g4WU|DqIYjqtEodzW z8#yc;W4Z8Bn*H1l;r9=U{9QiQiU~VmEFt6B@N$~{FgZ`ilq#GjO;a#FPe^)JGTReO zMm>aciokX7FP4;D!ZJ;u=A|-zC9*RA%mS5_t;70&Wo0NOFO69lz76jMOL%71fi(e3 z%MdzV60;f*w=X?MdM1#JrR4<_Br80gcax?!N5tW-Y4Ru&?MAa$tQ3Z0X#L;GLn0 zs%I4FS;;m{J%d=yGtO;p6|Ux+RFo>v^OBkPNkl1>b7nDEg|H|^Qu9)oC?&;Nm|2Au zy97l6EH}#}3pudT0ox8ZYjdkeHppnPHHVN*6x}bGC`ROFA?GM>$ydatB^b-i7#^(c zfU&lnz_XEr8Rj6ficy3^~y7hly|hexYR^90F&awJ}Y3NAiRNP5=K zz*I{{Mm>acioi>-hsUEHLCs5LcJM@Q=AZ4OaBqtAqQSM6Sbt? z#B;OD5WTQ9hmZ}d&P!&Z7?GQWoTI!YUlA`uR*U6k3=dvHlh+D78%da94l-NLvl$0o zI+NQHxf!B+VLgC79#%JFxv)~&cixD)S&@HMH+vIapC#uBlEdUkyaM}pe4dc>tObE7 zotsUuie835@Gq8|k<_eIm(32I$j$uowp4C*1=a^FH$y2|X@H%!5s{lEJX7mjfi(fk z%@8_X60;f*xtZrk&&4mKSB#0s8b0h>%*~ohU07r-a$tQ3Z0X#L;GLpsO4Tz8^t@yy z<`YpW{6B`ZcLlyhpa=8_o}rATU4D*MGGDbB*oD$LEA3#uZVg&bJv zfNcky;knrqqL;{icx);_1*yCZj8Ow#2((Lqz+^ony%gvgXQL!^ko((T&aj$$8%YbquHrMj; zc~YO7*<1NU-3*m1MPPF&ACG#7TtLm*3s~y7LgZ%t8F(r;Yc8r{xfx2yN@MS+61iEz zGqVn4u2^n{(6N%(o2f)@<~h=H@eAp-eQk;dERhXMK z7g(_sJd-Tsz)A;f>D-LqommZ{m<Pax;6AmB`IP$)da^UlE(D_E>Jl@L+9cZ`}(# z8%da14>DWMvl$0oI+NQHxf!B6y&in=YbNXPrIgYte@?zeKP7H zoKplgm-g|fM^N)pnH@ZloB7YBP`MdgIz{VdC?zkASs94jEa9112i62EH$&)nNz7_M zNOqvyz#ZFP5xLu5_y5C4+M&nPe?JE18L(I!a+Xu3Tr}oLUSjjo7_Fnwpi$ zL@6=O!qh6X7(^=>&O#2XbS7$vz1b#Lp==PnWZrD)S;LHvH z1YQFF5>bz!W~DNI)w$XG@tGIlIkQyLhk(6Mpb>{qa?$`hZ6lqVeT6`p#5$K@O~Af| zkjesdtR!YN(773w?X?_RlqX;E`a=4892Hr^r>;x6S<^#+#+4!m&WE;(fQkhK?-W%K z_0d>RpywpBodjaZ+T1Ew&71BinPe?JD_J>8WjZ6dS<^X%Mk$h-mC8gZG0wuwDzI4d zX<`}9LJpjCz}D0Q&^p-vO)Bcd9$TwB{NZs%FRN_kZ2u9*K~xUxf#QQlhF2r zN<14$n9fqjY+27{99Zd0ZcF86i0`D4t5tuYYA5T+d0<&EbpRz={!O{R|e!7 zsxPVYISUJgb|}EYJMrJ~0~z0cPCa6R`RujD&Py*ZkM&G^JpP-tr;m?!UfQ3ee|@x= zZ*N`Pcq;kzv-|oJs`K5Q&ub$skM8W>nlBgIv;Bjkz4>mV+pkm)S`w?!IQXL*Ec!b; z(!0Xng~RBf4z%D`M_=~QKYZnn?tJhB{NKNw{zpe=zwSpL{E7Gdp`U&FiQlI`c&mQB zdgn{K2RG-tv*YFBa4lmozVBlP`^U@GvAVBG4aR4$-+$)p_{0mN6RTgUPS0-b&X14B zkE!wak==t^TOSRWtv;ac+lFt~j#n<8U45&1Xl8{yetCYVKg)Yct!$lun|cofMB@iO zJ3so=e1Cg)@%-WH!|LpvCKd5teTRBLQyhW4=X;CsgX3^C zP6Z9R4+O%kJ$u|5XnRdd8vVudhvRYncavy9$_&3=jl+QqzQ&K3e`=;X@Ui0O4Ptf*a%WBK5j@C?0ow{|k#|=!qG6g36T&bWWYVVN;Wf#Q`0D?pNP4QsWzo{o{k9 z7k9O3pP4V`$BX3{`MXIXVmkQaK7VXPu(_2s^@BHT&INhc@%TP3FF<&~X#ebZnMl|vW?M9?ZtGNE^KT!|dymMQdcXq7z8-&i4p=Voty*hQgSZot) zu}y+xDD_#L)@!yK1YgPo+<5BGYe2vFvNp%ApIynHCyA><<4qb}Cq^`(x4u-JKhn{o z8FIFm-+Fnrvs@g_mpcdh{~`Ky$=IoBC#$(gwoE`-sK#_cUN-oLXqJDeZQ zcXt=Nvs?NfIz}`bzEs`nL1yZq?V|&j*L!+V*w*_v!2oJ};#ZEM7_KPGssD3Ded)(v zKU^H`z?7fw>frVC{`PgSnwyTyf1jk72y8wb$O|7*TVbr!3+lFBBW8<0+Zn`6y1>Jx&t3l-ogILZC$B#iOM3gxv7QsxAK5?9;8^TxCi@MO>u+I;>1LIZD z`PId(<-w7)(bk<~ZOE4o4)i~6&vnh^c)4|ZzI(hVxvMvz)SnosbI%=YFLvFOi2gK5 zZL`{YM1a_>!#lH^ceY>BInB!X6QluYoT@)IQtH_zejXfsc4Hj;JIO688PV)lAB}56 zP+;n&tQa7ZIc0_E~*8%f&l~N7|+8mOAw@J*~#g z|DQaX=(e>{2pbnxl{)pz@e(RHBl!0hk8Z)gpIm)Now;GlwmR^gP@hv@+LV|Vdp8%` zIwZbyyrx$m9j~6***jj_-r2Kj-0E<<6n%VTYVGKlW7Up7$^6F%3##4#p+Qw!!8*D6 zWtp2ERS$1H=vOiQFC+C6BXvEs#A|xTHw&a}E2eqzJKEx{zSAumk6jAL^XTJRf zUAaDa;>7BPpCC`F_m023nIO#I;fhh(l&z>gJ5o;J+aGs|oV%{h zjC52v1=aVy3XsX8`mvGvSTEs^m)m-KPm8a9_6I+7@5KKjE9z@|iSH)o_NSX8mGQ&L zUrYu7X*)A2acDL;un-1YqS5NRvw8NgdT{H3@u@T4{zD_k7f-|dY&Kn*R{hx(_0e9o z+PkwmKKb08-N~RPpA3ONXW~&6)ObUxA6`-RN_88!yM!xKX6tSLN&iT5&rS8y`Z$Qe z*;u)wq0eq~#x#<8)y%6b+k2}!AdKgb`rZ}wUPd2#&N@DA4|h5JT{mpoZyE;!?!)RoQ5 z3ZI-WUtZ|F(%Ss>9lswB7Mzph#wn`soE)Fn8l6!8Xk?o%0cu1YwQjIOl0;%=8(dY{qtloHp$yXJgDx! z0YxwP&lg8(nMXS7*x7oFrgyu%?)>$)+649F6X!1I-0ltiza~kIOidpGLyx1Yvq!5cAuHn&7`ALOYm@5@>ySX3nvCWb2_ev4@;P>7#&8rQ$6QURmtV;3Q$l z{C>rm?yAr4o6-rI-=iXgi%&Vr|0^T)$pN$yE z-0vjR-&j$f7{GRZPjC+$F&Pf0&U~XD=HuL@{`^W_(m%zbNhBY4#5K*U)Np3 zd5-OwoI2Z!!{y7a2%2w)RzgPPK1%&>$VSnQFh2eK{^C;y%L$qJZ{(YyNV*B@iwSQ= z2ZPb-Thys$-}JzF*Esd>N2_Pl%8h56jS@9IU>B z`m8>a)s&Eq4G>*)Puc$QAsn0D9P2ZfRtJ827*WrGy0qupjOX=`d%oXOnAw}{%qC1q zPj4%7Ym>N`Alvc%di9pnBf4GDBr+skb)#fT)G$I%P|-NPOOuK<$oQwC0HO5FK6~%IUH@{0 zzEktXmfwC9OO0Lv( zV~y6e$XLPJL-t+9c7tx(6H0IJ?e#k8h`MTIm*JP48)nzPp=r+j6Y?9<PG(M z>S-PE>J-{^rhGH}{jXQt+fZVMH)le6ZcnnU2^zlKU3)SAH3O(hx^>A?@OmZlO$qb6 zy2h?=J~bVN_uqf={*!QX@A%H%p1!!In~+z|pHzQ+l$WY>I@d&{AovB3FDl z`2%Wo{3KdVMg$0_Zn>EJmqzO2X{baP7VAWkwvho8A=5AK1y@Y&tqO=pw9 z$hPrqR2w(ck0PBpUBxxKtwCHe874Nl4MeqFc+Rdao(wejS!z(H)aUfcy)nbVyCRqy-#<&a@Qe&0yFJq@DGg+BcP=}SAVDRgr2SEoNha!0d+G`kF<9~O?w;s3%_ zq@1Ye4OrHgVgwAXhz%y+8(KD~Cd%deGb8nQYMVyB@qNudCSr6b}6bPRu~wsJ&v@2!N}eEO}?*Cd%V=v9&ty|rAMn{Ycgr+ zv9X^u;;+s9uS<`*pwV1w#>$d(+3`{@6*B{m`;Y}ykY1haygL|vAcNryDg8n z<=Qr}omStxk{=Vt8l)>@c8wwaxqFVP-E_zQJ@+DVXxW0NLTh|UEWtwkFp~NtVkfH} zV})J{6h-~x#d*@a!yB1cpY!+z#1dnAYUrawNl$QVuSSS)-5J3lbk?kz5JWQt!GDbm zX*Q_45^cWfCO{o^?Hx2P6f^}O{W971om)_!*DoWOK2=-F*X``)g#E$m57>^=(bgzA ziKg33N9wPQ^1k=hWT6n_Q_txuOp`1DXjWqP{$+PfU@ZvGRhpZ1?p&p{zUaEy?-=FB zFu5`>E)Nbrc{JbN(PtXtbK##R*%g4C8hSfLm*2eX@XCPwd_>7SFX}@+?dNB9;Bf|c z`(pfP_E(hkv%{xc?oiKFQRdSs@Eb?@y_kH2SK$>hE%!Y{>%Tx+hG;=7Vry)l-`${x z$@TS?lL4hP1Ju5IjvT{I)5s3*1g_~>*uET;S6F}Q$G+?R@rpv6TPwYcXGf*2?$LlI z{mzOQ?XO+NT$SmiIa_T$$tL?q5R2rcOA zq9Uspk+Yi=t;l9shUKk=pJ<*d{ZyAo1T8S^xpemq6^BH~(QLX(_+NLBaKlI_4COKq zfXE_Sdd1X#DMZ2Xotxv6FW$MCAEULdh3EZsl$meKCJ^{(!%HF)9Ok_U1i{8z^m`Cy zU1nu%imp=6)jQJ45WFS>kBsR@b=>ntqR5AI2mbU={nUqbi)VI;$_?=%oYuI#KAmNq z)Hm>KnkKs#q}xfdsOrYlZEo8Kd$T+HJGUY3>t|7%0D07hB9KV;E(6YeixoDYw4GDe z^z4JxZNMf}e4)|!>y5(W2ctjc?U0U1;~NL3ld%QiX(*zW3gm1o3a!!(z9z$>Ajr^(_uZWJOX!GFA_gOgZn zuy#kWZ_=Sd8=>u0Byuq%B4z#3>J8*{!D;K!M@Fzlj7c$u@(@E>vz-LXqdSZ3omX2` z`^1dzL~1q3Kv1Zj?>@8hYTo9Y{BrH}{4l`%1Q6=5N2#y=2vTft3_htBYyB*d>DxRP z{wd#LI+{3gFh23pi4X}bcG<~?8&}T(lQD^wbF=GMnS9?nqSx`@xt=Wh zb;E`h7EV&#y@|Z#qq4Vr^dp?)>}hfvW4w}iX>u#Hl)`SEPoyw&Q`v2xA_rjE>9a+!X#(_W4b z$Bp`k`^44XIZBM$l;%%;Vn_E_^ZL3Q9GPe>u=SODc-ryx=5Ba%b3|0m+f|Q5zV?wS zh5$%AiT|pZSCY`Qk5IvZc9-+k6?h!c-s4EeYv)xt@VKy&(%V_) zmx%bpwQywJJU@9FE-1Yi*?o6@S3egUejhRWpjvdYqZfhb6Bco3!VL(ff!XuN&67K7`*}uTAa=>e~fv5K|W-v2nBdg%$PO zp!PSr`VhC2k0RDE+k~d!^Pls>|cej#(j$8!xfN z#*F{Pw-HE{ioyUguLlr+(2k-1Zbe;|FD#AQl8KLfPm#bjkh_j_sX?P0yFp2*@n>}a z`=YbB+;{aq`o0daTFC_OXvLy9ZB*-T4fQ{FS__jW9f*DHqyyy4r&pUZ&sML(l4`cD zfRXkch~_1S_^+C$A=Ez^sdYK}uWR@|tsgqL-~Nkhjg$D?zHwnYP$U%1TmzKo;{OB9 ztxq`Sn=i+w$EQC1sbBH2QiQiga1kE#6s3B>7rMjX+-n!0UM^O0A`;iEd)PS<3zG74@rz^c%Z%AAU9fNd0~}L~QjgzQ6I+ zFRiHOhqSBTwBJi_$+nQ)aNaH*{jFLGymmwNk5|;!3}|<=Ss&Vylfgsfn`~uWo))D| zEmWfH9N!Fgow@_&_^q7T&RzYCFi&rJq>cU^O*I)3z}0DatIjH(mKQQXTN*0t`qkeT z=IyN*O(Gg>%sX23O`ak8-I$L^<4}`thFPC@Oa4`UOMY2xj@QLoa($n+i+SZz_bA$mn+m4#ry}j^tk511+X~Mu<-grw- z&qMCLx8Aj9B%u(YiuC4NdU_smE~_Wne!qDNN>9(_3bI>p(doH-yk>61-YkSFn;Ytm zF4(g--~O=0vXj+EZ=tLph_42SPtQXRW%&4wxajm;&NuOz?DSmD-k`o+Oa=is>t^xk zxmb}7r{`jxO$)lH#Mgl(Tv281Gps+vr{|#>QJy{sDzbF@UTyc4^z=N`H_hBd=er>T z+39&mQ4A>eVs|}AAQhjUN2(@=Yn`n-?2O=NAnaZ2OI_t`*l1b5Rj;jnw&TW2Y;o@* zJw0!kM`57Xc6|Uvvj~#C%%P;G=dFh6#24*kT5{RxdHefl(%-w&^L9WYkQ8wuJw0zn zNw@ADt=LZQ0bo*8O?rB6R1j_TWgq8}q*G+h>ofw`@JyeoAoV%sKssk1Bn$Z$nZR z^35Gb#Qw91vlXZZMHY`|GEQP*H}ch_TseA?pjEk z)ac!sJHAF!GQKsuf%bR%i|%DecxY>jhv3Eyu5BG*Rvtc5o3kdDI?W@i z`fACzB(lAy)X)1X?`4zQh4;bh^X09V%^yCyF%fC#tuqm!(Eoh4zWQVq6yK^I+Zs1P z(JR+nfz?! zd$IQI0p{1uC!yTqCd_X^l@3Pq-EiM1TVg)``k{Vr;u-yI;5GfF7k8Nxe`hWEpb-56 zNb}C1C|zxbAFKA?Cyzw)NY?>nOGnEAl<_l!F<<-is(FkJ>;OxL5Ny%Y27T_O>UW>* zgF;+Et4DfMcMmFr-;P;f?l^E<9x&ZFBI~m`etDl zj1wk)e`BbJmHJ9C%xg=!6Wo5@x(qkM7B7@uU=nQ?9plF14$;Z0Oq~)Vf6NxW%=AzD z`P?rgq9z%qQwkjYf>SSg{mzhEL&(Qxe#45a#fy^Bz4 zL^BC)1hb0t@|0d&deMiTmaF}$tS;rW9x@#hZFYiz&C7ocQIcpvpwoBibiHg4owa#+8`pa93GABF^a<$8x z3twg$$1h>^n7bAD5YI1f1ND=_Qi-#>&VEf&Eh^QzJWy1Xazj{DdKsBaJ4{sdm)u0P zcQ;hS4t?FKq-S}lzlHI5-$!>Hl z$r{y*HuqJ_^svD@0J?)jjn<|Lt6!N;in1bTxyaQ_dD+_67G^g(sfgBP+mFsyKy34# zLQL9s-P&U_gjgrvbw!xb++(HahiH07=UU$)OtD6FAqW&e{y@a4Y>UzCbZR)E7 zK`-r!Rr;-iYZaJCUXkMZpAfT6h6eahCc^Ko7TPZbH`R5ujeDXGF;UtlTVdt@0JBKn zdA;5RQRQ~{QhzS}MM<6MS6oAhTmWK)iR1|GGH`{rj6)P0(#hB7W0BY;_aRhPKgqu) znQ57}#dkK-WB|le^sW~#!4G0h-4@A}{Bvt|y#JCh5I!$Ah=RY=am}>7dZ*1lWleg1 zt3Rm?qNOlc*FFCWe3>HhOV&T%DVSSp)j?^19{ShVo;#;3^3kq7ohX88P%?7Vi!xL- z=v(;Z?gl3Mq|niQC_OoZ{RzZ3s?-}R>>Wl%PPfk=5E(rchkleo;%=efnsn5^ z@}I{gyQ_*OFk9@NzS75)JvnG^cToCu_8Li|^I&}_`ipklaZo<5v-zeVLS+@4%f0!?+{j@y2jE8SKQlu$WOs{)Vy8Di|=0D>*!4gOm`cgjB zXN6+>@)J3l-X)I~cK)fq-WmM9D@6+a=Jdwqo$Q;agt;cwceq6p2uviK@Q@9?Y zM7+@Jf?yZ)os>FR)KAHidaFXvmC|uOKwnD7h!l6G@E-YL{u+U`t8d1aVk=aCLc9~+ z1w4A6{0Z4~{aW~l)u-W;xy{#cLD<^S!-!bb{Y}3DG{iM}D6~fAjiNJ~DhneQ3N@6&ae* zJA?+^kRdGWwC-%ih++dW%VK~!Q#0L#1 zVLp`_IH^3>8F;`+REi8f5UKASsZV!ZZT1fK7W>QNHR$BgadG?3t>wGyIM<40di|Ei0J2eD(T zJHLWDT(;pZ70=GI;e5<~2BX4at0dbh+|hD;U-OSiKihSXSs#S-7|WNfdRW;g+#{!HwdJFo{zyR&QGf78aEJTqza`FFma5&+vpza&44ps#B0p_X#h6rj0J{)iJdP zcy(ykS9VT5z1>^r^*$bkV>{3DPh(dyc@iZV{<9tGeb@Ss1i9KkA?s;>oIts z0e{-7e{0G&^>zJ;e5w%L4O{P)HN*ycQh68c|3Z!@R%`O zNQb}!Ugc#`QmR4IpU0!sOzOWKty;>}?7}c)9MVYxUhG8l!!H?_tHZPT@ocG^B@T{U z+4oCaY4@S)=X_~*<&E>J;;)4fB7_s8lunM+U!&@!1m`GaXb!^HK6bEwtg8jhKd3}* zNQLP8sX7C~-&YbVGd}f+o!wocxGm{!Z5OknAmYsf*^k`u81@qU5GY-?9ktN&{le@T(1a9obG- za(sGx>eHY46(1|52u5jjM7&#s4Evdb*Gfkb`r{1Eqxsu5F4GIte@i|7iVr$IsYfOr zra>rN4<`N|6GG7F)tZt8N*9~yU2duO_$}|73LlBpj}+tYqah0Ut+m($jhMt^#mE5K zVqNU@=Q_yldn0wjA@adfL>TAqF6QYX{gU|hDk6$Dqkfe7bvg@Nz5cad6=tQyInu>@ zP)L4G_qd9!W<{0zTswvjC*H2=d#I#|Jh*n5{ zEAZ=wQ7AbO$d7OE;(B+(H3v7?)z;pS_Gd65zKd^HorgBT)9UZ95N&jwwbL=y#HV}X zmC2y!^+C9w#cDNc}O$H2Z_t=QuSx`WKQfkepw3b!6(qj_i zu8y+@>P5#dLo4ZQf4-*+jr#nhPvCy;LTubnL`JN_Ft6 zIG*ZY)=P?OV(azOme0%Bt*NKR?-!<6 zAWSrc;8N{-`3Vbj(R2-Zd`6}E<7sRL_K-ab6KO&GwEDpn3CJ3^BokCQa?_Wb8_$|6H#cT)gJnkB=-hW`b8B%P)KUpc>F;&@~|V_B=(k(jl$Zs ziw*1+t(g}|p&HQ9{#8@?R{FI9yeBpi;kurlPwS6^-EaS8BDM=rI*PbAQ>Cd?R~D#G zw?5lrTxKsx5A{{9&okDu!`+hykbbSYQVgtK)##Ozf z4;+Bq0Dc)JZ`Z|_#A20hM& z^elav=x0GTkF!d{Ki+QKSX_;G9S=9 zG$fj-jd;v5<)eh;Q6cIGG?7x5{^Czte?mI>!@furWh^2$?!K{{?5&wL+@ub#PG*gh-dh4Hzj3;4n`p^>*K>@XegE& zdRnQXOMV@U;)M9V)~JQBAv2b3@4nkJ?#P=O?U%)1_k%rMftUx-Y;?dJhMw}O&khj-`?PI z9imykqpq+blCRXXkmVlOP(Mt)u|&!fzFU*LVpH6P;6~%0<8e!a`g_!!a9Y&#c^d4+ z8rEOk>Bmkbh-ym&HN1Yuxoj6LH&$&B=O{KdBnB59VYe{ zinujXaHFY}OccHO~spVHv{P7k=~NYxo)O=ZT!s?M|xJiB`-D$w#1 zM_*=t5m}?x*XB!o^A=uSTsvOu-d_9MoyF1J?q8fGI(XezWrZI)DoPV>!pfbzI^DwF zGAVNJ-O)R=igY8aeyJP#-eV8-L><0wo`U-A>Tehpt8B;by_g#zV2iD}_Va~kqwr8x z_YK49->oR@d72*Z(Gkpa0^eJ&P3{Tm+of(0Qx_t!akKh`74_Vp=r-%l@IFN!%IO0W zc*2FX&#?Z0Z>?hqX;F>n>;Ne^4g!>{>X>#Y+o%woy|C7UcmJ{P$rSZXukCj>*ko8h zckdClw$i@3bG(@8()BAl`!6*wAiElj^lPL=L%|0~vHY$`l$J3&TlY<$Gk+6rxFWP{ z`uVcMB6l27IuH~-hg29hUSdlJ2XJj4yL?|@9)*DcWX)!Lb`mI82g>WoO$4PcNOb^b z5M_^;##4M(CmrP{W_t^J?(t{C-~RC2vKa~bK3#kA!yr|J8D4GV!yUEntF;g z)ANC2k> zJw5gm>rmd3?J%Vx+oxE4#y16 zko3)sL;9o_d*Bn*uYSiae1Vc2HO>)=`*q@FO_Fsq;nPXAI_y?k`o}@BK7=2lj{B@KUm|c3wQN;=08!zYvo_bl6Rwg=u@druP`sfm%z zEoC3atM}y6v-jp>hQDmi-a~1^xLw|tOV8dz?oq@^6AH;|TD1ozj+Eu&)($doNyI7G9XYXR3&Bs$j4NM(aMxRf4bIdM2dk@tpr}6h}BR_;i z#AolJzG-4FI%o(P$j;tFief-1rt?Essd@)5J$sK-O%ngk)@{?>L3C$B6>M#==P38%h`JZ zq819(`~2S|>@j#@(d<#ow~*YS^LyMEuikNeV!wEG>WN1#=*Obp(Eo#{us^E*7Y5zZ zM*I}qIm9l^X8yOrz-uy%R1;B R}wu9A^NTCcuOJv1}TBeQ1M2kxOSENUYx|y9(f=Fi|G~`#xv}+! z(6C2%Xlcf0Z|FyDjt>ux7wDUL8Kv_Mb^|E;ajbOF)?jA+Ff`b^`BKY`#!T=9Ex&2{ zE@4NM6;Xff!nnzP_iSij>k^E})?S}nbifu>Nc})YZ*REO2(d*kSPwLa zSKmESmn7%4-pktJ_Y&Oy5PxVRGh84~2=8vrnpCM?G_C;RKaamHlc{LeaBx*kA?pEY<=~~ELOf%J+?J&Vx?COAOrzxzWr-vufvC`+#l{nebB^% zZ`t_I>5mlMb>R|w)$GJa!th8}O#d|gt`;_+xkQ4!%##0lzSI&d8--0UbtQ1>Q zeIxqLN5=o-uOH6$x1ZVBJ6_w~@k>naryTyG@I0s;eyrMmpF9$U5M3smh?b6)11Q5l zZ)BE^273DT(hX2QS~~X82U9Z(HqZB!z$X2)&W{Pf=I@`10Mv*#5p6A6^GN|0*iaqH zQa0^O9r8=F^P}(hE3(=7XYeQ4G7mLFn)!)6$(H?In@pD1SY0{*&F3Y5Qs`{mh-}KV zBFUeVhLuKgsh4e%1v5qI=Q2c@`HKg)m-^k-@%{E+(1#l{%SR6c$Yb3d*)^$c`;rVqdsjh)Bz9vXi)vrem?gLiKzL?>68LTKhE5XK78_*=vP<2 z0sjKD)S)`sneaatDtf7`?anM2LB#bzTKyaR%_yM}%qr5$Q+jdfMXxc~q=YQ|9xs`*e<}?pzt}G3cKV z7G09T+ee{A#ovTlfBmqlzbNB*>q74~_JtD?ec|K7$z##jduAk`ILP+Fhj_-pBc7cM z=%Z?vgb;lzhBSjkrT2$CeyG1Jyf5fb6g)Fs?eXGs(_%+N&ra*s=uYvZN$kyH*VUQb zP}V$0E!WA1^R5NijeA>=)%C$nHNK6uAS?1qbSxwTzbnaF&0FxlX(ic>UuGrQjp~x@ zM%R+8QN3?TRzKV!YST7VSp8mnQj`@z%SEnc%FEV1U&qz2ic#g(6D{CM=Ul#atIu94 zs{Whz6k^i0?A9KeA;dcQt}DWf=3xH(xT=V|=$z$SgejH_CYx~>C7wKcM3)`lqs77I zv{ZFifp_gYh0o14e2SK*z(L)Dr`5N`avVGhy-cS6ANv2kKB_|R&hIXcbm_8vbI!gW zFzk1u-K%pS_b-}%P+c2OsTa)|dE}nd3nBMJRcC#}d+FF#Oe1Z0HR;nF{i&zV4(`(B z_V(gcdcUcD_6I-orYiY!D{3~hl1V>x{>|xoGgAP(zStEax%xr6VbG?&^(6ZqW%s$_ zYE<{ydy{)K2ij0OjFtJ9qtw6V-wo*M2U8~FZgq6!+H}R=lkoZ6a3&RYcX~i7M_88!_AD`HEqSciab>o#^FOe$JL!^r70u zz5j=pC~c#yu=0O^S@d^C>UtN~)_2QSH2+6dhLPjdx7aMKuR3t$?2J7f97;n0mUOpX z^hNtYl+B_-b^~AiB>z-pre)eS-`Px)0T5I3mSm<$MSpy8;_J3ZuH>J|kWTx~)*QwX zMS%LMr@kNGDJBOvL-#h)t}S`(KZaS`(totyE!NC$NvIfridJCdso}H>ng0q zKHC?K*u>R=?#jO=cJjcTD1vH`Q`(Lk^`Z<_4f2`w zPLwxR*ei~VoNk{#AToL?u02pn+|!SV|111sk60xVMYY*t_w<$SMd`^Q^frrs^*Z~o zZsPW)v4gG)%D=4rm-rLld}#B5nvM~|^rA>tT<=>+H_A7PH+Giop#3O6El)4w;oFWB zX-XB->z)+uC@=S2n}t!mJ^zgF1WVW!`ckkDw~FU7+n1ln5%(_f(@uYNv!elhSBez; z(Z|lEyO7w;t;cwceq6p2u!97hDO?XxV(rlD#oiQF?5HP2f4o>~T&qIRmD1gvT`|`6 zL+QF(jk`%S#xJq9@_@qCf{IL4q z_*#b(%6RlR6)o+jmEKPN|IEEvjGafdHtfA^noyX0$qqwnD-sbBBx)3G+|Ea2J5C&v z35b&rJ2_XVwd3xbG;z9(yPepO3kwJrOhR070atLr1qwn4#1S(GftV!p!N>X04~XdK z2SgAO(*;O~f7LtG^scpP)v$fI=)|`At*SNDs#-%mEB3^$6O}(jz4%f!q6IM~fU`Q1 zCC(u1oW-(fB~9YuJ}cF@iASSZUHaWf8q)W5DU3U1ahF9?v|GLwe>8|mvOMv=+`VP+ zQaBu=S4bNImDu+f8*;0&1THdP-Tsryn-o4zIXf*ZrXH$&Im zrFBiQ<+r?PWlqGqGQ*CK_mx-hMEH(e88HpylMRW>T&t+D2H)!MBlX7Y<%ccGv8M6~ z60X9>l(5;y;*DJgch{cJeRp@bd#s;7tsCC1ZEme^p5EB$kK^JO`_GC4o7UWg%j@=8 zJsC%PY-Dh()?S31%WbL3ynx>HP>pl6XcEVzQ~9En>ab!@i7n32TNt zNx@tus#U<{w!?kFoyJ=FXuL>42ZVy9I`N_elU~N(6zcv^OxS2@8aiubn&_0iPnVb%txciz@`=`Z__)lwWmT+6C}AcpMDRj%W4XIsn>*yb}?*7B_~-zgQuA?~!AB(NV}t`CZlAHx5L; zrnuc6xipvT5#4)_{GKOGgzhiK9wisy5kWo4R%Guza$KIc=$nT6yTK=vvAzwI|FX`x z=;sfU@v?c7x3|(>G0!R#*?TXOujZzL_uk7KHINqJ7i13J;rUn4xck-(1n& zw&x9AES||0kR;K2FL%JC-h1RE++o)39K&U(D954C$kDxUFj8P#K}!bFUG&kxa-zAQ zd+(97hN0j@RuR4T=;_#Z2^AQNM|tC$62yFdXJZYr507td{kso2`3ox9O_LeKe-pWy zpxzY@1AS@0gPjO{_#1@wW%1g4XKfd1DYnn`At@(A#9hfoE3fF$qpv)^zeA569Ufm= zmiH_p)~dco#Y>3~yemTkG(UU8_SViWLK6ARl)CMS+sP4Y(>EX-eu?;CNX%s4U7Kgl z5W#KnD51QkcqmmqUc+MRt6ZfN1UEM>tnFR|xI7KXS;{&t+q$en75IaHf`ZryQM}s0 z`4B!o362IPt+PxwYb2-(&x9n_5cpu!RPK2YZ|hLZSNwQjxd$Ch@PZ|Zq`>wkanEjn zVp3V6{+Vzo3Lc^q1T>nk2m#`NaMNw-;Jr0z8ZTN_F{gV|xV((KRMF)SIQ#^%kinA_ z@p;}o-bq&=I6_{$KHr_MJ-U7FiTUniX|nIWd;aYW+|8h))m(hf5x5)(fe&my#p3bq zIo5#duXpB{JowaCC~R1K(#c+!nZT_;G^}wE#)$htp0E`U4 zFO0lbj^BHmOy00ZQltmaPxvw^TXTKrks>`%uDV>C}AdvB^IDU-_nO)}OLz z6U5k@Kvs;zmPGIG0cfcDYJwP>Y1{aWSd86Lr{epA=5mT06WFFjvTE5jv~~a*CDi_} zC9tiTHbeVisd9B|_?)6{sT3Scq#*f}YuJ<``Jau5z++cSQ8&e-!OS(H*b!6cc|gr3 zH`0Zk$SPdR%0&{!3SK7Q&+3p-J@=SR>!2E6k0km+1>m3eqA}uRazv?|S~8*u`p+=| zs+LSpkPsL`QsKe%VtlN|pzH)-pl5IFl2)`%t8ph8AWv!#yg3A;%PSc-2^j&TZJCgYK!q4EYB zQOKn#gs~(5;EKQ~1Tk&g{@~HHwS$Cj>G8;oT{G6B1J>}O+Z2+L;hh371H2~ITf%YJ zp0`8quorfJn2PyAgpRz?o2{aiAU9IIWa)V#BC_rYx!bhMGm-Kz&JTa^gQMyl5$<)< zEd(Fwo$B?hVKtd1( z^D(s3n`V9IBHa+XMFBQ;*4VuRSuEWw(sYe;+`yXa*-t3!4ORXcgkM6ar4rpVgJ0LXqt|_kUFy3OTx&3xizO3uw!#sk zg71B}wgfpF3Zffb!MiZLO=_?f6ZYa2(?bU0ug1PuQ*6?wV^`L0Xa@xdM>-8?bmdhM z`{&u%?Xz31PD^yTOQ}d?I+2R*0}0nxa+fa10JR35*9DiXlD9~0!YM(gp@b2CB-Y70 zcq=t1sp`WDBnk%v3yzv#!9jIkGK&Ta{5j>mxo#J;!U@A4*b6UscYe+Z)E+fN?LAcJ zLIMZD#IPW5Nqufi6pGBAgHp_UHsn~bl}dGQxC4I!c5-;2#lhtbH@<-O{+5iixU*j-SE zy%D?X5XJh5v|4i-!+g=PeG=G4eTu?ZBG<#9*syL0&NMk4>gUv(aC)f{#G0tQaJ&Y3 z(%tQ|cb}WDZ*FYuTI%Ai(?>??9@)*TSGqVRjdNw=%67H#WEXsBa)z*7J(*mGoeoQ@u<0luy&uBrSsHq2L{EV}Ua(1haQ^0}D93F%W z4hGA74L)GP`XJPtJc9{v;MjA_+FFp9HL8A+4~p7P}hn zq6_sa)J=J;uUik+rPT#$!9w{E2dhmYT?h+E-B>$FmU(=^vZ%uM@(~Ew)o-mUzo52s zSn<1vy`dF0y=J#Tq3~*U`}1XQ6&+Lm7^7fK1?mG9tgQrItkk9p(fXddqbl6x4w_c7 z>iMxcQB}L!a|eSQDw8P+lv6UKhHx?2iz?=XIrlnl0}?ZKA!W&+a|hA+ka%Goy&42H z!63MmscpN|N6u&73gtgI-;Ohvx$p(_VsszM-aDVIaj`1PxLZOc!7dq&E~7|=`kqQ z>`?=0$vjNy7Ve-}jfJehV&?Se{4D~CHJ3xJFMq!zJ|JZaND_r&&0XS5n@WM&SCoSy zpja6ayvut9O$cw>+Wxl@@coBX^US_|Yww*MQT0Jw0}6_jp*8g1MHf)3J}49`H=XP3 zg$HSk0|URci{V7u4#|0!fmlDcq{4) z^Tjn^x78NK&D~+WYy99Az9^?VJ3H(yUE5Ak?%n+;yx1o*OqP51reythcmYW{(N(yV zo7@-Y-aS%-i+?hA`*S3}I>=p)V9!W%@9nVpo_uad%+BkdgPaQ|&b`~gJu`qj%B`^c zfpIBt@x{4!OJTV2MV^fPT$X#czwg~57MFXs6B7HJ)e~v%-OiE~q_dw3efx546X)LB zwbk+Ov6TJ5t!^j$m&m=hOBwk12}V05_inYIQKAZHWbYH!7(BSq_$X~Jq~5iCb^pBc z+T+vidFQ@YU3DEmU|)hC4TXISeq5u=IpX_t5j^%_b|C+5F&a&#S=Ce&B59uqbfS`d z;!(Y_UoD+#^C!AV)AfUD{+Fq14qpxLUYk(T2>!4ievX;rSiK*A=sF;87qDY?&Tj8) z%+5k5TAta!-#uga8NrW!xgcjYG(y8bc<9wkUUaHo7>b!U__CJ{_5x_|%RbSDX)K#H zhXE~Gv$|>7a;o7M2uFQDm^$f}2;DAP!M>pF%~e~mAp|4hwONi;ZS1wtfO$Il*DJOt zLTV;z>9*|7({c5|k-ABe05(zvjxSQVe}w~W3>fFE$5rn%VNJ70&(mxpDgA670|Xh6 z*qzHSIX>S#{aE{#hgVTl&%vjLJ7_-^`q$SEtlTiLjxSd)Jv?cwquI84?`nfJUw_-$ zg~|Aa#(L?$T!g=jEU8;4hUH-X6VRmSVs9?duunF5oBng6;Vz^7# z$ndCF&O{R}dK$&~7>s}Cg|qXm^*c77*qNw#y^Fem{iL!s5Ugen( z9a0h<9d-mwj1FF&(nRS)y5~!Tm813;(fUZB3l2Eg!1r;=7C_~ZfOKCLDw}zKDpH_k zc~g;tX4xm_l3}bmE4#9RFu~-T-RYH_$jkfo9J}`0x`mjC1=oBQpf04_sxGe5cg=q}*4PF~pk0R?UxaNYw5 z5F0&pUKXld@38D#MCIY>puv)5r-RgG>gPSFTdMSyg1%X5cj;31<1U1Encgj|q3IXn z!&1&FVJ(-g!h$X**n&NQh1GTB7=Fn>-3VSh2q)t96Ax`|>MoR%mw11bxr_b4jl|s~ zT5l=>KxVP3Yke@uJoW1`nVW+|RD9Lb6eHQ0i86{*wo9ru+6dW!^-@B3)YrJ|=_V%S$&*r|<8(Bi1(Dx6a5rONN$mMn> zZec-0F(~INXxk@>684|yu^G{(FGq4z@jlcGjbf?$Y$t=cz;=10mL(jfySz)TnJID z7?KU0Vwg3`&8GU6Fn`}`kpnZGEpZlZS}qarTkqJ}Jq3R2NS8(@*^|a4X%D+pVWw%B z4@WH~E;%5(>IYddt;pdD+?Lu#{D+kTNtbyT@=TL&XkyD&|CiSTC2@;-S9jnvFyiJLsxd zlI&77tfk!Wj+KMAy@5$Kuk{|`6{|NU2YoY2K%QL%oYYN=h4L6{T>B*RdC;2AQ~{+K z&yHT-6mlg@!4q{||M!p=g+4RCxN#2pfzNk$0b+J@d#lIdtlKxbN1eOVz?~mahbA{u zaOaWr$R!C@IR;VH!7j?_b>F39{sd@ReKKj62wFWoTwSEQ>+2g&_WtIO$mSjTr-EhRIG=9<`~8Ux<0F#L5KWsXxZ6mk{muiejDng=8-yHMDhhvIglty}h+N z-`v_bHyKa%5HI5uASE2i(1`$|Dwf1X*fy=L;P&EKiG}#9*!a7aVXeN7w#D8cAS?>Y zXfwE6ohecB)lWz2c#1~x)$-nk;>wiS-q{&1nml>vHR`7hv2ga_BicByv>4rS=l%k!Ymjntcj zXS`7euLHVQdJVy4Z!Xf?AZq5kT4rL|=#^a0O!b z>QG9I$T#B|;NjG5B}$@&XqFtUv<@a~s!@pDV(t^_X|ZG1sZ7;+lpo5km;Us&B1L+o z^1p|glvpY+V5i=T{Z6CGQ`B$bm0-cJOkE1rxK+HDv$Q2)+d zh%aUvB(^o5ZP1iPaD>QfF&ZqRZA_o0=48UMyv9M}kw8k}l3v9rG(wqNFLP zR*TilcYHO)_&MaGZb=-3JmruAH~KEwebe?Gd}L#HK73!_C$zW@2zQ$dCf(Gc33wW+{xGzyG~U86kdF35X_M;0D(MRsm6vd7Of#!j%0~52s>x7Y+6Z^Du;@y zv0MFaBn|2Nx)jEpvbbwcjB4t`ZX76xNtP$xmpf`1ph);0V;OE$w@eOaS%zkI`Oeh! z8p6ZdinE!+vkme%Iud)XM|NlmRt7dp{c@~?H@AS^FbbW=wRH1S@!r#MXsAU>DcVpVoy@Ynxl^o2NH+dQ(yS zn)tmLy6!HmYl^iu+_I&QUyDxB$Ju41uT5}gJuiIz!WE@43%;4Bc(H1n? zt{~@fTdFcIpf^2K;~XtoMp8JImi@Gp8HTe_0jbXxvEJ)pU&gzHHA9}HU@jBYD&TV4 z;lAKbW2ub?RiFbx!BU-gQNp2J#@`g`{!mQaYWqg3=;WOi5*(GHVTppkOTlCi)NT_J z!WuejWt!-ezE78!7p+aJDMvNf)B@#&C?UM)*?aqX@uCluNZ89HL`6;a-rL;T1@Jj| z91PEnXY$(r9T)6^C@|1(9pG7>yzLIT7B^79y&z(G?~!AB(NV}tZJh>AtgmeWIK=#n zhw$%352&0NmJ6)H1G@KKCTTvp|8U*S+XgYbD7gq%cJ$+o|2tNAH#c%zUT)Dh4R1kj zRhhFjyP=Yvo7xMXam*HP%I}v^91^8Iya31V+RIQ){Bn907oHqV?Pg4^OzLU|LRRQY%<)H0Kt zShjoODy1O1uyJ8+_ab2FX=Kk*ZBk^jMuGzGgz&EXED1x4UA*-8LzQK^StCJRcqZUj z2*k$(J{a-a!`+4$U@Q7kL8n(DEYr;zDUt%)pTs?`p@XvpO8KDvNrCesL|6Sy_V7o8l4w@}Cj4PDlJ zL=`24XW$#nF3`7vYp%aph#?CY#yET$LXdH-amW??3n)PrPXHWC;n$ZE)^18Cao87R z?Y+Fks|WU^tbUh(2tt+uBLikNBkz^t_X?pAVUMIpPY04ufFCaATDMfW%t0_*w4$@2 zy}4B%qL8f+ww;=9JvP~=|100L&H5K4#wIY&iBbNt>#GT3Y^H7FGh#7zOPz}E51Pvh zVw)DpD$lk(Lhb)r0^6EtGqfL;Dp$9L&nfz!O2NTI3VN%TWF%HoB>%G!5qRusDe9(p zH0q*86gy%{LJz3fqlX-H9raL|>=?{PSKkMx0EJ zpUUx^R(YPP$@i29Qg<4Bs18BV^bHPC)Pap{ z?bBkUGOen;*2tCUAZP4brfY1EJF94Ed+6ika5=J?Ry1yxjhnw1sT-u+q(f_r_ttF{ zB)(RpC(*=Br!8L^6OIHX;oxPCm@(5j3Rz~zlCvsSGE`S)1)$^jRR`l<5L^>;@3QGS zyw%3R!#!%6IFH-My2&*3(FE?v-63;}nyDKnuNB^+^nC6eYQzN^mlV9YeNBAB>OsBj zI(VWYpHheDASVhB2-dddPe527fY=I14R@~k#8{$ajaY&WDrCeoTk7dS=_NKup3gS@ zG8vDx43#(7h(a!v$dpN~ESc@C0B}WM6oQyGZhw$?hW1Dg627I!BR6*4M^3nVM-rqG zfEnO5vECAn!}h#=ecb{niTOf=j=a&Et)i768!O4m?1_lTx+mndns&V-!o6;~lY;Wa zWVGyes@JoI)sX6e099(ul{tAGBqUH;9ImBTE1Yx||A^lOt{hP}%0Hf%`(?K~!D%*b zoNmjJBT7>Uo%1oY)0<{}C&4uWZ0xMDdk3;ux>=+p);Rx!!roxDu=V0Hxsyc}?091z z_fJLQ1gH=N4uU_Y?&&2Lf z9R=N_7GDUzgiuQ*x@iW#?(GgD%%{Ff!L=ctL+@51mlsli%I z*o#+84;hHR8vA06?^(0G)u&@u)^2DA1qeqv4QO=brO&^U=T7U6QflQ&bwwXMZ7M0E zY}H>I?4M_2x6f|5`mWsNE~O%o=|n2J4Adw10{_3 zBe722!CR@Uu)VUR-7aQ@6NW#q7hXU@1cl2E!vsT+jReX* zHzt6}a4`j8h6JWei?o4_9Tf5t4-c~X0QovRC603~2sy#y7V5!kB<+9w=TbsbIgR@I|^| zah2c|7mOkmSua);bUW+dFnj$7lp+MnN@1McmZb{8+90TC!m&UGs?ft-EEw(Pw-GPZ zve;cvhrJQI>k!5IiL^!`8gym);~N`i*VZ?m*nw;2NM9mPZLVKvO7{DhsQMIzu|%$i zL9t=o5S(dN#bTf!>gUv(aC)gBJPn3e6O|W^*FaCYyM6ZVbMy60=uwyAhJ0kC?vdTh zdZoJzm^2?A^FgVhQ2Rx7y8$m(rEYTtU%eix9Nu(3Z*0S7Wvc@Q!<7%cNO_<&JC8v=BwIe7*X;J~rxn6Lq82lUTPf@RQRezP$d`;`ig-1{h!OaeLK38BlDP)a#WuQ|>cORN zX)XPP8rLNSmMU4rn#fFRtCGKM)H7X5UVt_lhtrSQPaNXXKz2haqi-* zOG3K=A&5Hr;3e%t^cS=bFd2{*UBN5tti`Uz4XAK0cgOm=^~_zp{!o8sNeM<8HVzqPLX`k?jteGTx=unFxV_J*sll1`Q2<>(~wWV~#; zqwlS@b8PzgA7h2V(-f!=Sg^Jdc(GEOE=23OOGnsxw#F7v&yUrKs@mP2JN0FX0_Bvb zi3j0QgO_pHX=rckLdufinb#x zXsE(AG(IWNCg@L!-S2qr2APLs?%KIoR$r&P5&-#H>f^uRN%ZUs^yQnb?^)8`}mgHwBD#!yh z0_y+4-C2&TV1p-e(QfYwG_hOYZ%JoiQ0a`RRxFmbb^!I)30}ID&v0H^+>J!kj;6@n ze%+#W{C{{eFk&f#3g`r*-63~iF%e=mTM^wSH?m8kDJa&#%#Ux>_wsY~*RI92=e&;( z+w=j+E!vLlS)V~PL*U`pC6D&+0aZ;DzL1h8^6oDeaRd}=wthxEdy!uxW@YoD*LwmT zy-wi;nB!Anx zCl`9@#rlH#=n$a7fDvhuf~}myJ-d;Mso^wH{Z$yC(k7;kZ!U6Ctwn3i;ud_D^FLOY**Y;D9yb;IX`oL<6kEC?yknQ+q_>c&Aq#$dlo0WPV5k@n4{xp#MDo;;Cve)$NL=HBH5Mm=;yB*v!0|WP2J~igw zS0Fv~@Wr`zOJTV2MV^fPT$X#czwd4O)~n|p09N1;>bkQDe{t^JAuHJ6iCkotWA}bJ zH@g4?dNq?5odQssowM6J8z^Sp;LBb**bAV+FZ)DW7IIA2{07))&G@Ee z%c+K6n$;f=Rz!76gl-qDn7XGSwG|seFd|-?%RT)p)0q_LoeWWA&N z@8SJCw6EGiF)Ro3pMWMs7khJwhFx!U%_AK#{UVO5g-zH#B1sU$6vJJ@MxI?Z9`DS+#=zJKDj$Qk0Z`Z2-z#-WN zI^Im#?oSMoZSecRGTFVRqa}x1pBD<-ptOlregbxiNGqC^OV{1qt(r{%a2cYT`N{1^ zcLBF`^1|*9C~)I|^BN#PxKrIlotHh6TJNyzY9T61U8a8ClPPGN_1z#36^K!5?CXK6X+;sq!1o+0 z_ZJ*9N~qOsvad)|aJUqG50cKf$UZ5HzxAST?mPW;48Ljl8ZjZfD{a79`0+ zG|QceqJ;fK`uKv2cNl`&r^biP+xroLmqaQHRId17(9{G1ma6XvgDyQq-Q*S&S`_>Z zzJ~6Hsq(VBkf2uAZetxzNH)6Q!;}5)-=-Lz899z`WHa<5;}!)Vo=FUZJ7jyQPPq`G zSTQ6UI>oSZ659{;E#Z5?UW*)<>1>G)Ag0C3)xK&EHLd^LgI4R(=p=j6Tx;*;PLv>s(H`$OLz7)>q~}C7~cbm=xn%wEr5VqT-ju|+ zXI^#v-$Py$`po>|# zqAHnQ!MpVOWRfu}1+AVQt}fE%-7f8Ks{i`LEent(P2_W9wN{(Rpq_e&K)0Bkf|j9I zWTuU-Ik!py*qwT2ZxoL*eXsQ1)S{&~UCKEcU7+b=0OR zej(~)Z_W!aXJ>%+6bB(AqgUoS|R6YNEJWSyUwyRyW-Mcfm7KQEe zW^C-LR;3DW<)0D@xpF0nAR5#Y-OcRPqST58Jrl=X4VK~Ox!fugEYApG5zkYjvUcU` zs8fgX+(^Aic*Yxr@H(Kgl_;_L?ts;8ElQUc-0L$kzTJWE)?IdrQPl5CWp1iaVz9jd zJ)!DA9ZKN{l1h|I<0!mvWSzaAUr{p(3D@-~aSiA*%IZ&evqvlviE4H0Vi(E^YEcSD z0BI)>eF5zg!_Zu<4&|Hi3~-=vW=ADjh-S&rO6y>9~G9^|+6!5~8x?5E;s8fk9+pSMg(7ag6 z+>QjNDkWWVGc@O^!s z(39L3-t4tCH-CnuHtu_4CPUEWO67WL+Q`%Uq#tC0l1T~Cw85v{3*Qn z(t<-n)%ct|l-*-2!>#I;$>A)^(9E7-8lYZ7cz9cJHgkBk!3!^N97U2u70Hg9DD}&+ z65iYbdc!Dm9@o;%1FlYXBh*uxbYp-RDO~_0yl#aI4g%A83|<6dIE7Xp9jW_LPn*6a zHs2WD*;HH&lCEjNUlYGKL)YDdXfqezk&9EcFyw;T+UGE3gI7f??kra-lWj`%tMqDn&j(oO=^ zr{Il)7=e>qwd^WD_TD4M^#Z?;a@#r$LSA3nY8uVYMCksPi#W3PUM6wTHx2c71GgQU zTJ2^0#X}ty+wu)(H;xq9doPo(=B9%8-pd>{kQU(=WDefp`S;h$3RWgAX&+PFC2_u_?ItA_-tX>H~P#`-r5ej_Z~@W z7z)Nw;>gd5-h1Tlv!_i>v1dZ~FJocgUr@^no5e5u*-?ne4l3^UN6{xGf$fl=l=5rOL-^ zSZp*BUTP=^Zf;yy+r4;pqXF+M<}Z`@SXN}SMuGzGgmCXt+=GI~w(LAMhXC%{*8B;e z3vfS+>5AqFm+5AW1a;w=fMX#L9~1as#BUE36fwY7FN(R9oEKcy9Zm3pC5oiL_9t<# zTpu_+H;h#Oq`-L*qJoDgrC|{Q!~vPuZR)_>eVW%O&E~RD5%UJc1nzG3pC0jmO=x@& zWFdnmDdO|Id%TmbK)Fv*?A=+NBXBto0w36Z%Hr{2C)}|HQ-8fP$Mga96$%>`pLDVp zW+rfJkS{tp5O1N74;s2SPps3O-tSA{8TdxC3-qnvn(J>qDm&Bx!x)E8LkKdiH4eFg ze*v&~;t7Cb#!AE)r@oZ1c2hcu!@eME@8vCCJ+LRfXhWacCD87WrNGF5S#9J^JUc=# zt5=TSdz+^R3ie2f^mHKk1o+`%u60Y5%Nzug!nF@k$W{p3PR+L-o9xs7m2cW+{YFUf zm8FQW3CvqDVx~^|qTs6uVr-^u<1=D0c1xX#?+=>G3u2oV$tusbJ-(Kc&p^JGz_w=E z4DE-d%GIsmb6m>|XH_Z%2NNmitzMFmNJwritDlXCz++cSk(uHlGB+A%QNvjsF(sh~ z)NFDiT@s3{!nL4W5zAP=Ou(PjA)|WU#~}A1NbA%hiM~()_~*T7j5wJbQ7UKg1pVij z098vSC`bqlA*t}-dh$HrvY*|+(=w&(!V@x_FcGaf_UWNIY8FUk zT2*_?YzX5+wrsUF9(8tT+a%45FXxt3B3udc=0QwD5ZhE1$qWv|9jKOWM z*~^pthn3z&ozT7bU0g?o2_YR|!>@IigoBqkV#ZACC}f$e@%WjDY3Fw~*5J&4d~@sH zArdn(%BKKy9KY&d+zWzhV(yvg%IJ?$s^G0Q4j%4N^9TScdAE-y)6ho~xF>gq%q?oB zZk)VUc#G2Wxp$}tfD^pAeNBAB>OsBjI@}s71Jwv$k~LxpHmHyh(`>0* z)hu8ZzA>3#@DdVeSMV|!kF^YyH`s_mE^+yhr5cP-0JtJB3PDU8w?9ZcLwlqL3E$G= zksG`2ql?_VE(uZzzzp!3SZ@i(VSC=bzHWh(#C#z_N8aeoR?$jerG~c^j#I6kh={Cv zLhd&0@(yizH13B#_`y-=i^6lSo3262=}j`>+a{xBzf--QHLQkI4+O|*&4q=mz`2(2 zLuF202MGz37Kdx;)e0w_#XpkE5%=Fbt}$nOw>#n2jmM_ja^#586hh~G4DIx$S>H); zjQ|@vYwX^EES7E-X^Az?KcTQUSS@V5xJ>S3kp(;68rG<0l8s6Os(@L16jvJj{v_4A zayVhz5lggFo8?XUW6D?w=zwJoR^k94(|N%vA@t{WzQT;Glr^7;-Jv=Px=Ag*5Pk`v zmP&Nf41Qhj4uUy`d+NIsTx&3xizO2v2;mrw!S?}fP==i28v)*h;cZfbwV1FMub3V( z5PvoH#Tvi$SpI)Hc4h5`c2IzDq|<;#S6=%3>pMynV*flFyM1=c)pzACcPSN#Z11O{ z`#{3=mE5HZGC-|?=XJp)tK=*O80m2#oK#bJe`CaiE!9hl6b zVFiCq|N6;!oeJ8;tZ>5c2lm1X-ffMWM7y?OpEz$DAsrI1VMQ3#!{V`HbjK7`0S30$j2B&T^gm?@MF4u;bEn{r(zgN1Nmxz+6Io7C#$ zbqTw=S)EQ4vxlA)PGuaO!i5jtR}5l@>I-86s(G+oTS*1`HH9zI4U4M;A2S0}M@804 z#gD(C(148W^&?P;qNk;VyWzn_?u6-$uMt%VKvy9ri}- zu0s^-C(>#~6T>~wn~?;zQJXJn{axm62zLQyl}h*deYtP zvv;4HuWxQ_?WVXP9~r58WH+;3=`I5%&4)`G+&AMuu+xJL^fdGfNHa6;5#=5Q-z{XfcFe+T)JaZC}fW*9L7zf!C} zq-T=32GPYfx|{03rEh61{nZqDz)~fvSQD9PZB_EOje4d_Ne5^XLP5|T?IdzW-!z-= z!pmEY3OC!?IP>W2?dLbnU7U4EXg44PQRe{RLsvKUUMG>|>>JVw`yjEaaYIFXB`;jJ z9;{2NprW@niF9QzAa!HyAX(;lB|}3|N*d8#J^}%|`mJ^4cdA$3`xf(qyg$e}Ws zqChz%Luv>Yb7a-vWn6X|i5a_)vSiS?1LlI-=?aoSs6y!=jS$u(xASYsY4+g z5;L!V_OYQ_=wg69zGnva%)ns(4e|o=wP0Dsa$YESAeEq;SjBQ8gU~K&3`4z3Xrm=o zwae+plMbml5RLJOxw9Nu!3Iy{VpSfnTi|a=XJJt3jHy;EmbP{P_16hrF0}F)AWAVT z4Pq&S3g`r*-6416C{ZqUhb@3t1QhFl&c`?Ed-=J0*sjI3p}dbzK(S`i>3aBH3Y{u~ z!z%scD6`BuHL98@d|}z<8P+`SzOL-(-~|E2%J73HF6!Bf{30cTCKVLOo@qsW@%Z+oXn>!ow0g2k-E zcQhwK3R#mTDe%yfxW`8jOgy`k59+VN0F|~-b$s(2gUFe+l&8qX_)&P8s4xe9x!5Ro zUb!d+pgG*w$mW$Yb9v|pHadX6W7fBqlad?Vlk+p@H~wXE@9t_`)6Dzj(%ic{x@U31 z3q^C??8^*1&nwY0&rQR)D^=3myZaGOo(Ns5#@byO(*C(L_bw+e@cA-x@ACOt)U8?^ zGak0jKE^0?i9mQZr@x(ym-qCxZm*K&-sRL9KV6i2mya7rO_qC?4_j*{Sa`sn`Tf(PQL_Un%!)N59Ou zcRO)bkj{QC^zF+xi<0Kv+qKp4@3EBqz^!g4baU@;>Qc=wxI>Wi{&h}TXqvUJ+EJ;t~q=)w7YKf zXas-Q4?oAOSoW)>Q_Tu>TPO8?{GscBJfq38v(Sl_XEyM6&lrA2@S|TY$e9f(dkln! zUd`l1rvTJu=j`^*28x+C__CJ{_5x_|%RbSXj8hHuz&OiKP6C?wfOD&*a7=wbm^$f} z2;DAPF}9!p9G*+zf>VqS7lO$Yd zqzoKiq;mfX2ih1g&RLJE-f6;`W|7V^*HvbGFDQ66j{$-VNbJt#mmHt(o_?(T%fqWE zs^{QS!yU9A3;pYB2Uczv*vglymmZ!pw$c!%x2vXZzW%ng3xJc{STFsTi#}-b7K&jx znEwPcDZ1F3OSIPMR@Xez0mU!kxLVkR?IV%|K}<2+C2VB(QLmVr>S+|`V=(@m7tYSN z*6-MSVrRC#X^xm;D2Hzd^Fem{iL!s5Ugen(9a0h<9d-mwjILcerHS%R4xWRf_TkpeZ#n~EGX%RYMy$k}yPc4Y%$g30H0=do+QE!&;{ zfkUzlB-BjV?oSMoZSecRGTFVRqa|YF9Gx2$2%12+=KYiPZ=pj_wSGn3U*zb=!xIY>msS3OOk;s7{v(f6B9JtiS^^)9>v zv{ayA?OO9+O)H8>2EON5xxe6;Q9>rP$-W{EkO7Z~P9~d%Av5s9f>EpgsGP#z*xX zVbCR*+=4=jg1^Dn(ETt~Ugictu~Sz+VI5A$J?DZC2Vl_{o*6lgZ)7v{L%hEX4>X8p z5(AA@O}P-FSTQ6UI;}Kol$%ZUE#Z5?UW*)<>1>JXlb(c|mP-UixnpPd6!@(pT^gNa zPnwH;d)TE4Gfm5UIBGF*$pP8b#Rg;{KA05aTeSgMkzBmAkWBgekZec!N?s(^?3s*^ z?CQT*NOmf#|$nRXk61-$qmC@WH}-f}ftUUv4GI4+=yQQ=lo z@i!6^pP^z0Jr$b{g2Y#-rNjFmNW5l%YC;@qv+}H_l?UT&sL2Eum$CD$VG~?OPYd5lQMVVTrwiG~b z;fK5^^qKj^jdN@6Ri%R=%hr$IYEq7F@NrqIqK3s-Ou#vrPa z>2=?w4+Cv_3$N5!y&EQ@vpLY}>EY@k_0Z?!evihOSojnZ40BG3k4yk5MgJdea5%0j|cVdsblH8r6}Z`tw-bUs>1N zFj?%yQ;u|_uMpA)_1?1A;b~lo>W}g2C4>~ZC^ov^u*%p<)C!B~HB!Q%3{4BVXT*}& z2-~K$6+A9ru8_Ufk%iXS>u6gH{|{kNSVo({WqVDmq*qa$RzDr7<0%@IUoD^U>cr!+ zp4mSkxiWVB)FBqm9t`gphdK{u26G51;S;F*ejKWDu!68^KABC^3L3-i?kcK+?rpYM zUB66?h5U!yOne?Q7{G4&DrB>p0^Y|bnWe5)Z>jyitVqwe`qNw^%AE~lg}hcM>!v@_ zRPXf{J@tG%OyS9;t6jC-yEC{Jh3)fZZ0u_DR=!_!HjtfqN-Sig62*xJkS;8exNLfR z3!*_i(cR2mEea7~NR9?Q6USZ+mf`2Q+$xkzcIcpjDqn}KL&4$}L8_R{9zHixZxWvI zMj^Zo=xilQtiC&7wX_z6xw>>V07PS4XZ8D1nVTw<7;JAqPpCRjhf+9#Tc1Ega@3`9 z6ka&8&fd?js2PQX>w1*9)O+)&{)9Js!~%?{R<|y8p{$@5rEmmS+EMD&g(w17Aa<_~ zrNoGQGoAq+PTf|bBwC2J7DamE8re!(l0PkW?7CqG^(ZAQ{Gt4M=}&JfQlwWZ|9hxO ziKX&}va|P6$_V`?UI`Wq%haV{ja$WgIa-$=tfQ%2>Mcp!=;*GSi3~JVDN^P4nw?9% z5MRu3LX1q+*X5o-H;Gf3600E!xa~@vtxbvAKmE0-NrC3YQs#CfI8`a>;!M>szcM9C znyyxh)y#K%HO2Tjm~JP>y7YGYC6pv20pNlUAzn zCq1sj#k$n*M$(YJuS;RLIg7h?S65>X8-D^2lVo|~eYvBS0i5G|jAgi0-7-0xWf_{; z6HEitYX}c-E6!#P&o*#O#5R9DO2gg1!ka1e%dryP+yZ*TD0Cjz(#->|PIbUn>~^Xq z5M2Nyyl#aI4g%A83|;_9?9rw^I#Tzgo;H0;Y+dLBc5(guY2A!-ZF6gV^Yq3}Zz_sk z6Tde@*WIOcO|j*-ylG`l=-d<84BkpQ)>B@;Nx36eMoa_wWJBUIH=&jM-$&|=*~<@G zluxcc8U9oB?W?`VPpk0JiZ1SEd@SA>WpH=x`P_GRU*H~lRvg&0<}O@bx6kUyIEo~h z!LeF<5ppiKr7H6RdecKS&e5V}B!y#X*-uNEkzhwYTf};=hkY6E?Ea6MZpf1q%w?il z1zc`B+!x$wEVXfK2OSU!mg>Zd5)SP${-#j(hhpkh+c#Q8C-3y~5e$Ksg2_yZPgp}| ztxOZ0()Z~S^P;tBHATE&Qwx+AqJ;3GXYcLn#fv^rBH_fW9$+TC1wBVLw{`)14ju=? zv*TIjTL+@RKtEvsw2KunL?yWvH&DR7rqC@iAE-&E`^KT5W}yNP=-zuI+q}Rpq};Yn zgOJzPwwgxsGZDIfUD=V9FN*BFmq}doO+)?N(3_yW6@;x7-V|ZD+CSp*1Wg_Oviq)r z?!A}ES94Rrd+%kA8c2)q3o-`}@%$^OG;={}WkC1d%Oj#aZ}4LAyb&8D|LUf)hU!O| z+i$cH-FuJZi5Ct=N|7rl#~{=TwGBHBHq`#ZN(0b!+tfQwqD0fZ_efg9P%w@XM}AK9 z-XnjXJ#DyLO%0{Nvqy2RlD+q6GK2VUB3A|W+Z0fgE2F+N;K5FWKKu#P6@ls-BPeU7^`B{Sl+>?YA zA;JAs+&+z%)+LC&0pajV#Hd4JCj0K%JadK!Zi`0=R$b%S`HR*%)-liYdr0 zY+P8|y?Az`LG~=`5tm+5AW z1a;w=fMX#L9~1as#BUE3)XKS-HMqTxN$C8}#u}unAK%>ica1zuH*2Iw3T%H8_sZ=E zOJ09!sDD!6ya-XjLzL362m#`NOzbvwU{2|#Bz;+59_$;GdGiLv`={Vp3Y0_O@Ds>F z22WDN=Xv*d2d{WTX0lIFGP^ifJE_hQxE#o|4{SeW@pwIZM%14ouTY@D!!h+03L6%m zbg~y_CU9$zFFH98Z=sM68oI1Si7HVF&%n!_U7&9T*Ia+I5JMI)jB)rhgdpQuy|3#8VDwZYagPJtq|=yHQ#z{vQPh4zG<8Fr)=5;F*bpDD@N=xFMZkd z)dVp%)3)&$u^79hPQ~{J&E*t1Ca_J5WR+*z9v?fEd=U4w1hzHPW@tYwRjzIgpHujj zO2NTI3X)H`hD|9H{A{G&j3z+3T8g?U9t~y=X`YoMrX=)$noVw`OG1%Vxb89M?~fI{ zOu(PjA)|WU#~}CNrgc)TK@~}Dz(4P;Bq|T4YH0ZyII3DQqC*(aY4D*s1O*9!AtY5z z9d{Lw&95vs(OWma(jR4ei+6y@_NZTsi88u$E$<(rE=DTTs@mJLa#Y4j#t&%4rvTgI z&MI2k9{PAWT#l@!6^+|v<0b^qZ;*164y{F{hE?Ym8MHxcEzHt%+VZ6_;YeT-4qoPn z88fY;kZh8rh+BBLEh;q)I*$88825tUnwWc+P1oVAHVz){QPaeE-ng~7yE#9z30a0& zlWFLq37&V!1oW0xxW{>i%zbO7Zk)VUc;C`1({FBH6W_4s8X$qlpK8ALB)2#1Yq#HtM6kua#jomwt z#nR0pEwRS=ClvMutA(u>m&u(hvS7#eD$9l&S}i3Tl?GG+v-l{kH2D2Vs(0mZ!WM@} zv{ReqP5EQWSP4l7EGh*WtP(9ykdID zK>XF%7pvuZe9fPZU0J)K9TXrOH#VTrm6txc$i7t3CF-vY_Rq7i+h@02eOK;smr{|) zbRreq2NJHY;eYFe7rd)4OAJzIS4cxVw4pvXCV%`2B^B(~6uw9|EUpr4#J-d*>6=&whuP~# zpcElkRtn?no-L)5-Lz2AgkymWRH28v;L&cJgNiFw&`I(aG~FSg`QMAp`-ss<{WjvI zS{Azt>aaIrcO9ZwKap0wp=#jSe2T(YBG<#9*syL0&NQoHQT&FVQ*XlQrH1e{7-CIS zUN~L@J?ZZD*}Kop*Ect|b}e<$JHGnJNZlj5ne|F{88B%+Jm!N^L!tJI>UINOu1+&A zl-RTyPipNLs;&2crUgLF{FuUhL;a1A}f0ViY4)P`A zm?B=yFk%FLrH};anPjd(bg_-@rh0JcTUtv$3BUy`RkDgTk(t(3C4bwfXS$TU0BwBZ zKzp>4$Qk{=XY*Znd8?5ZXFD5b9-Y1Y{KmP9vn~nk280GxuI;Py7qkyB8ITrT!7FT% z#IDAT2630;v4ZQ?gLP?jD|)dI{{l#)3t<7N8|${qGLJ7<_Q6^lCxl!C@{Fq~1Fj)x zja~iLy7G%w{4Qc|xC+ar;<^r}5GvPox|8v;X`{Zk+Rm{R!2cMdU`++;0~V}Ji7ZxX zXcN?eWrxr$HFTG&)M>@6o*%0dRkgc4cQDAI0_K2BQJ|cXA$7}H=A`Pjfyr;wkcUQM z#xA5R8D8$7#wrFeL~TBYs4YvW7Gs#L(6$q6H^b5!KHE=b*tD)fvvJER^1y?aQQg%2 zNwNDK&)p#Nkj!0s8LvWnoXY=$7hDlkHuZe&P0_1J=EA3J>QD$z!z{|%O$wgvV~_8d z?>#dRPvd#hC08hyldIr?RDyD170ZbXLc6Fj4D~Ldjh0BknV3OqS26>}Bj(OBD+u!x ziCnbXyIcoYwNrmfItznJXH2zXv9z@VsJ~9|(ye@Ej8s=%=|ObFQ9A&l6vNUWmNKY- zPB7XXaz~C5Syjlch$f&|2XsEZQQynY)nB_7*M{;wJ^{s=O{eSOXF9;TeQdTtvpk@n zShGp1=iQeJn1W)>CNApPi~J(XeeaCIV_pIu26g4jxu>^>apc{-GNPbZv-xU#dJKv+ zd(=Q$G7nQMoaUXnwV@UG%^X9WH~-&b6cHhpL#;1=za%~&WsCOYAE-K@_Tw)HbB!n| zC{~8c-sRmh=uw$s4+txeJMbXz7yt#u%Fr76@1hH+RUZ_Jm7C6W_CgbBWL85%GTn3} zexTqbx#&h_*^+gwx1U#KL{w+>qaqC86BXlvJa!@MBDp00KqHVf$*;W)KMagX@e#@K zhcUc@Izi8z-;zks*CQVWhf=l9F zMN(O!{wfSmX%kb&H_tJMmK%&sN^#M3U=FJSK!rK*%f&}s=aq|M0Gh*%jci^i=epvC zQ~LIDQgQ)nDmwpVa_{bHT+__^<vHZ7pJN_{1`%Y7x12mp6JL#${L1+`Bsh&t2oE zi*oPsaRaHza_{otYvtbE8RKyoaqit6mgRp!?p-`)(>z5~z?3kg_SuUw>+XC4ZcB6T z?y%l9zT9um&6v7N8s7W~&s~~(cLxtI_M$|CJAf?r?tZ~4bX=^6cU@VtP>(>Id-q5U zF8;~ftxQ?;AkwZ428S`Y4YySCo;-)~l^f8q^ZG|D>)J?j?{;w73~&?GjNI9^4EO)Q z^B3pd?Wp0#7kM)Fb6M`){=T>A<6eCQ^+0zqpP-iJ-W{@n4W7tFb~$$Mmvcj$dv8gr zoxp#9-V60V=hcC^&x3mv@`~;(&l6w!7$Vxlc;aBey))+jv(bWF6y^#8M z+pq3lbY6RW+P&!9_o}O|gPigu_|Z_<$Kb~`x|}1vPZz;s4`v7Q?-rxcWSUh?MIn;* znLsBh*(V;=mfQQ)(y6u-a{RizE7z+l*V?BkYt29IOkSq0IehhGd_|*iBlyF9_&KKD zk3V!BkY_r(%+5k5+Me0d!8BK>rNjGRdf0^E?&*EsiZS5OsGhzvTFQ_w-}!UmjjXQ9TEr8t$O|Sm+MYdz76I-0Zoc7_U3XJ zm4uypt7`)3i0Kz`TrF(E_7Mq&%rgA#`7W$*is3E|e^--|!>XrIoR7iycV0L<-&((8 z^NF3=`ldNz2!sZgY1{!mQr{5fgY5JZW&b?A$}=Il2t;&r*by`_Qm&2KqeCm0M*KRZ z`Jj)EEfg?yz`+K-k5e`&y5N_E%4Xi5iWI0>-c;mTSoUSGPMwuq*+7_J^3Cof(D{%j zdV7vt`)yYNM+#bqiCEA~+3rsak`0rn>|WE+lEbaf3x#b^+C(cqAvG(_%BAaeEik}k zh;HU5w;$aF+}6npyFZ}7jRVem00Hu2;k@jb)Ov?yR|`>D>N557o@`oOM{kz;R;h27 z`o~px&f@Mt7#Q|{UqVG4I+zugU^{%n3 zK#W>r|HikPRuqv8eD8H)v?5jm^%op7N~qOsvad)|aJUqG50VyBYZ#=Sl)ZiFMc>?a z`s-MU5rONN$mMn>Zec-^EJS-f6GaL8iS+SB;V@ga;dn`;vOwjE4+dSry6*^sEB&s1K|$Y zUaC_rgtDkqAX=l%QO*;PLvn;!hWuDeR3!CD3J!K4`9stw4BB*RdC;2AL>q80&3Jb78iuX-#xq(Sps{xz@}kgZ z<`*~4t+_Y04u-I9-)Nnm)u?lK8W8g%>d@q73dB6J9%=k&tgUjaI6RqNLG^omGU?zH zw0e5Dx=1Dcr7tIu&yCeuZ6bqu>R}c9Vs=W!Jz(Q!j^)e6wNuaRjlPLV-z&X0wP@*0 z7qkbgsjZ~$jE(Nzum5?h?ys!tZI~?fsv}3b(N_rRgX+5Y%f)OXUiHU#^%6pgR}_mp zajis$@S(bUE(*+m0;GgP8JZS!&xj?l5w=ZhD|lgzd?hiV^|kt~VGREdVNv+jJ%h{t z1H0&_BXvAQqxfohwy>NlrO$X{;)y^n%iB!)sY5JWIT#^k9IC8xU>XLM@E##-dA0g} z9IA3aWmq+z%%*7tjbX{E{XAyC0vzy8zf6sV{0BpY+di3_JSR*5y@u{7cpsl+mbzNK zrS|`_B0b~kPjii^IpeuQBB^@*`FNPZ6YQXVK7(sf*gkK@#;!JR<&DI%Yv`X63%PP7 ziXa-)6Wz`1)uQl09@i?e!6Nlc{QXz33_s81R-t%)8c`R3Rf$3Pq{2FE9ST+rBdzFr zZlvBMJmZZ*cpcE$N|YD^M!;&W7KL%Oy6q0)us*bd5T|}$DsxkX5_>ur&=aZ-)S(oP zAgM&TG>$%TjI6Ww^DAmbA>q0nC9VOzc~pPGn>}KYNK~s^7rRhaP>WJHf-CJP_3A;!M>szcM9Crh;m<*fWBTucjD3hkVp6iGz?^YY?Xd+~~Vx_f6Y(@R5z(`S5*x zUqH!)j_x)Y%A@MYpkiM@=VcaQ;yY;xjMt=K~R^t>V%M#I;$>A)^(9E7-8lYZ7cz9cJHgm*mLp@@u8i!wwmGI^k&>Kdf^SG979&mNS1H#$6 zx}NIk0wCdaD`ap0>ypP{d+yg$K+aaEkB-!Rsi#ff5?dGgfL&Zae_A)=T-)4Q-#ops zV@(BB4618&&|dKF#^0;no1yFO(z>SD@>|}tGACAG>(^)-4{uWL$dwV(Kt9=!xXexH zIQ{pLdSmwT!xrUOQ~3l5SK*`Gr18u7$Ks7$2Y1(=&wY1ydXGu=5mS{tQqPJ5o7UWg z%j@=8JsC%PY-Dh()?S31%WbL3ynx>HP>pl6XcSZ>nATwQ2{^p7r)}A+bv3Mq1K$1l7z1#sk9RCW2BPubv_Z~??EF4^6*@HMR zl!xpz*iicqD-A~1ZcX3l$7%gB)k2%3H4FvgC~@THMDIQF_u12i2n0vT=5jnW_9@wW zk0vvS|0Z&^rj_CDeVzS>O9LM4MCilcAha)w*XBEGyHHE9eXbAtzM>E9UU__f2MRqp zJifFn?`B6V!QB{KLF!0-kBXNPBYPU!0L{!!L=6ne4l3 z^UN6{xGf$fl=l=5rOL-^SZvk)Dy1NSuyJ8+_u|Qp8Yz+j+n>ZeF6+5$_eUuo)ITY3UWBONAxddjgaC0sCeCW=u&foxf_luG zLO1W99?}0qw{1igGI)|AKF_$D9h>`fy;qR`@r^7S$MpAjx|{O>z$b% zsjpDju=u2ty)ZL@TZ4Sj$$@wag?!M^Ww#A;lWg8NB=HP9?wnEr;YQFW@ zWS{=8eA71TPgzic7@NSn6=TX}K6vq^zM3G$X4*DBBNk(~)T#LXpt-ytwrP>9TDA?X z9Xz=cYX8>~*w##&q5ZH_xwUkeIX`Nam(Fq*@|GXEC5hs%)O64q` zp#K~bplZnkMbkHUa6NgRCN9g}CYGCkEuPvbWhkE1QNI`yWpv5ZfsJkL(?fOCERf2y zs`gqlUBp+$4`^knu|4jr!jJaQ$IIbzWHqg5+%6k8A%K2^l$%~?tw_p)Fj<&m>GrKO zTiX5l5V%SQ*zju|CgI>^j+imiItp2K4CC>$=zijYdKz>bzv^Jz3xaE6?p-!rhqu}| zc(_MR6X$v3*5>Z!{LCg~8D>qUp^qkTPwozxThvV5IC-t`7NzHN?@*CPcJSu*HSrCr z2lckAx09H#_!1TtFCc8N)5C!HlKRA0qGXL&f(@0e1V$liDzh!^dR9|dOUJt*Nk~gK~gfjQvhaw*Ti~D zI1bzM_VxA23%fr|#e5+`N8aeoR?$iT2s|%dpYP7s9@SM;yOX8KzWeU^w>NMX=8o#V zG*LYf5n1>VW_`t+`S{R^W8ICRt@pUIz&Ylop3; z>D3A+oy9-mcY!NMfjR=Ktha^M>3Pfq&pbzzrVu*kV`!&0&H4_IZYaeuyAZIkv&QZn z$YSYck)~^$;|A6c`4b9zgVn;;i_7Fr7Fn?4jlGZ4M>Z-Ar~+o$C@!TCSadfGet(kc zT{)bv#UT>y)Mj~+W6D?w3FZ)$0u5FPp}%^auRasILv<8%lUjTs{1QSfmFT7!{JP#9 zz3x*MKz)~jYYhf-v1EeIRybl*@VyV$mLTVZ{SEKJ@HVNzT1?oBS49Vzpe4 zE?#{)c4h5`c2IzDq|<;#S6&sdf1ZuqKD*`WyKA6RGGvkZ^q^cjcC_c4M6y-)9qqbIAQp|?u8f1W{#K5EqS+Poj-TVqFn(zYlfhzQH7d+ZcF%rjbBVMXyvAduSdn0z&A&T`A zX^msPfB}dD-l$Je7)#`O7!(`U4Z)daRV<1*@pI};IK5N}Vog+DI9>xi>F)N~yU)$n zH#fF+Ep^fRruxW8-6Ol1^-6acFljzK=7Um0q4taFb^~6nPBSl**t8l?tSn-hrmw2x z)=^Yj?*UB&)Qsl_$b&#X`0@u{tA0%3zVed`;%BrVQPkA`J>zGbU6iw1eVhUgWajW7 zWN@(cM%JE`3XD>93~H1C}aT#hS=WYpasKZPYVe zN?w3AzHy*E+DYV$e*J8|3omaq^5Se~cqI*5Y(Z)wkwIt|HHM+y9lOyIHoDH5feI`|jey2@ z#N1hqtYCvDa?$L&8fwC6rutjbSr}A0W2zO4rL7%6{dIzuZsoJNlrd&2r36|5L@9=) zK`f;!e6|(P2}Zj^?#NN13Kh`=6zhP_$2T&(e4^KD@5$%xNxK%;p7XDD2`JWVI`!qW zPoZOVj8MN)mpt12C7@WdNfUYZmy0+8iZz?KsAn(oi>wBEXA~aua+P$%MHagoJX#T6 zhu(YV~F2GDNRMw<^mM@A6*7wXS)wx9c1f7b*gZl@nLaHQp`(g<|EVbDh1= zMAkSZ(TjO}FLM5Zf|ul?8<}NGSbmhKfR}lPK&BT1_(a8+@39Nv1Ibzaqmy-32xLui zXm7(8dF#bQ3Vd~V_%BvR%u~m(ok-T8$eQGDd#6dEmtM?D)D%$2nlwp)hn~beK1N|2 z)lyQ`UlY!_I=;DB16|L&l0$CHBZa4l3UlC>i;Z&Um5X8k`V43`UGqxWNI3VovIder zDY?--IX`oLBR%&X@yc@p=I&}-yUqLM(%ic{x@U313x(Gt#k1@33bdl!ySq|x?%n-} zCr{*^*Onn2e30he(Cz6VR7!gT_o?xbN)5^Qa3j^Z5QQU7)SbwFS@AP?>#dxNOgOo^ZprP6Dm*_S_CD{ zy;};yji1OMS?=AL&JEhNUOnntN4a@^XJZW>jy}G*_3zLQVPrQ*Ia|fKcZaONvrpt= zcC^y2fM;0mJc@Jg?Z9>Xdn{!?aI4!17wX|lbMNg^20ng*(N4*|TM;&=gjM8hKmUEg z8iNNn8Xu+Yh14gVU)>`oTzh=lb-dd5s;jO8*Yy(oXejJs@Z(xO&Jo|Ii{P;bvjh2e zi_vH@&8nuN5J~$?pc9qs6OZbZ{c7n{GhfrB>3VhLTKghpt@+2D$;;F=hp(QDuV{)q zM(~II@N>)vNxdI`=sFR#_%(OA6?TRXEvnlF%TYlHIo;e0#KWs zv)elxC}!T^%U(Lz3!uR-`$TIpPBql)eptOlTH$!E*$uq}HQ$e^4+v8y-4dbOMJw3u zX^7*Avl_c1+W|TRBjUALj#X{!wb6iiI{McuwkSesCTi)n9M02m^}&(4NfItJQU;DM zQn`PH18oc#=d8z7?=)devq-l*muj`QDb#OqY?%Rx-MRddETIZK+R^>JDLD|8_a(KniO5^ z%_Uk8b*pP0>44%Faa=8I!uAmfhRibj?fEXOaEjqB4S!dYVl_n0s;5z$kHPqNUN}47 zTEAoSiJjT{ra5AYp&Y&;bein+6J`HAy~;BoI;12zI_wCVm=&H`qxO){dZiybozmpH zXN*+Y8u&g=*#Za6U_2EqiB&+X1**M3{J zJE4V`hy~4*?f%3d*#>(8SSGvIbhPAf>+?ck8}nw@OI@aZ-jhwM>+WVrUl_H!bgBDs zHv)SBRqqzo(DaM(VJT;ou$D_#Q3rGxe0B`KWT0*YFIezpk%!q2+(;aMO6zzQqlf`< zR9)+X&B*<_Oy=ew5fxwcG=+)-;LJteyO6&`R$RRc?*J_oC|JAJ{8!V8B9ejcIi~k7 zIA)ZP32m~kNK$aP6n*bXT5jB*l)ZiF)w8+pRJx9dTyAIL78WGQLNv>rilT)5MEdx$ z`a^9Q950Df7N}hD!Jvt8;gwGZlDhx60+7@_D}aw0+;G)*gh7{{qHb~v3M~r$246$> z!&G_2xd_!yScemmjV}0b02YnmnUUl8Mz%yh|#Jz0o%@>3gM*Q7u|}(*^ATd!eW#A;wsD zXn=nntNSbKdK)H-y?DxzZuAvG`k=ZlKFX!qa#4SbS1%!?ctx?!{X($cu*d6_bS$ie z)#0tS0BHYVr4z)Rrv3L=9m>$O;B9*>iH)#rT3f+u9Y%NoZd@CorkSe3)o%@B_egqMwe`@f3~XtL3@YRT@JJG8Xb5axT30t+W*Un^o*-N%{8J{HyAsfgk1N0JWSy+tE*kL*mW6Pi^BGKGd6a$c`Kh= zI&Zk9o)QbWawUo&8q^cr&Fs~p)QSc@6USZ+mf`2Q+$t0-&*(>{Xh*+%9d+tZo*St* z3D0<=5MBp#wh|>)-yN`8T8k3j`fz(UuCw}msmx6kN({C)peIxvs6#0n0qPT|tal$Q zSC__7c;U!8dq2OTW)u?c)uTkkul|HLd&DA1G7Nu|mw;Cl*^o1zoQ6Tne zbtvDAXMl%Ox0NV~7NV_1k)F6lwvuX;r^Sw4H|(Gu1v?Ldx)W;msvpX)m;Us&B1L+o z^1p|glvpZnC_9P_d(Zeyyb>%JmZ?j@8n=q~aw)?&`si0ro?K90&crf*THnE$|SyQ{dHfjqM&)Pl(`)VPE|^} zI8${jwNW}$tHquXbbK|%_&MaGZb=-3Tp$~BIRQ8NF4=w4_8oj=V|VU>YKS`^+-)+H zN7a$Zp@a;jKYK!K3Eaup6T41S{uEw(soJ`029%H(w}xank|oX{?3~52X(dfU!adep zRue^f7)t$aBn|2Nx)g?+v$!jaJ=Nlm1~EyNC*GI4Pj#~)J~A8)H@16>*|=5RGC7=O zHkze*c0-q32@h{8&Sn_THq=|(qK(5Z$4Yo}3+N4_(0N=-HxIZvvF4$kG&1Z1D3tKJ z6*4#oOy@D!8p(QG;P$Ng=t$j{dfN0Yu@~GJ{=FHx?k=rsiY>q8O)GPvR@-=ZB78@# zjF<-U$%e#bZk#rAKmGeiy)k?FVT&+5rIiX@rAvADj%(M|TDH)dHu z&gHgLWhQ#l4`W-p?Ywutp@KrX8s})yGLpivwCty)%rKmd3P^pni1l6%`!e1otQqno z1#_9GRsomW4)+CjniOjX9S{nZ>coo@4(&4jrcn2XV(M1gH(Es}@3fHM{#=ezt|0JI zFquj532W%Am1&|=`aWG^UbHqj?_J_iXRxUS$_r6Kc+s=>_VwaLA1INq&gw-;H3m%g z7<>zQj%;r20{9#}4u)sPv-)ox0PR8!WJ#{YJuCSaD`f9Ia%?X;3VErm)4&O6kh*>D ziTN21;eU9J!+V;MpPvi=gM@X$liP8_q^PqZyM_F z29F?aZ#G58%lPw$I()VD8-Jy0=rj3hZYp^1z06SqX%T)w=HMZoe+3pZhPM?i-Fq*O zi1xg}i^cOsY>@mPsCgvlD3LjEY8d@Em&x9HBu~6>@F0i_#%}mLU^#VTBNfi&I!E{3 zBWVpo!8l4B`8m;hkNkbsAlJ|bJjxr@5-tW!_uiw)4C23uT)oUY1TGDDuoIyVe}mAz zEMA-MtnETA#rC;AB;`Ev;jUz(l~?qTGFKko-=Rm34v#M_uZY`{5o=Z7qvEB+2cCvD ziSXGQwzqb6dyToV_A{S8VL%#6T(sQ5tr0mMiTtR`n@tZpTZ|C%ZhB) zNKhA^2{;x4@iBo9M*Q}0w^=197CC+vaEgE&19Y=Silo5yCvlJ09}}%9WsUkL1K4T}&U4#>oAQwKvjbCfsbb4l5k|I9OyT?0t z#Tzn{eTovN)35}8wW7`uxE#o|4{Sfh;_>b|c1(6aeTBk?#V4KYg_#N58sv*k4#Zn1 zoa}bO@b)$AieTYJ~LfCd{zV+B-pZ>3W(>Ci*Sx|x)o4~vkBX*gWz9{%= zf*6}=+xU!FjNMYF;`@W<@&dJg)v_&9`@vDVO?@qaZCVq$ESZ>X%{O52#9OT4bBem9 zQgASF|Mpf-OhH3(3I#tKsW+ohy{?v`Zi+{PnQKI`Bc|%}fSOHiq^mxWRk$Z97fBfF zmkIc@I%HJO`^ZV_)FO#a=m7ZVy_G~IfvH+NLC|UNp*jQw34tLbRZSgt6_CZNS#BU> z@Wf3icWuh%7h|G~E?vv}$Eb^u%CxHXme~;IYST&0_yK_Jac31RZ4Z6C94<#z(~8E; z9vMZ78D*;upx+?nrWaa^4vmG>iijD6)53I5rz!3gNC(*P`w1rD;AM`OG1EE@!`Msy3Cu)smk^_%J~5UkStFKUg9;fj z&6YY6bhewcfTj2r zu#gov*WQsq1SD1Fw0$(r-4x~^<4_CH5kmrk_kFn;fPVe_dZ-(f}Eq<+R_5=!tge! z!CK%V*M{YJ#q^MY_^YumRk;rtSLH9|}2}roUlDl+42BzK)HZ`uY$e?<8=oRNn0F>2IJqm?@MF4u;Z)06F-aTg`U7 zNv%#^m#`g&pgC+nMnR&OJ@l+_hB!LN3y0E$hg)u?z#^zGj0vdb!FFvW73|j(zDPGL zt`h8qEoEuOj)mYbd;JKMBHXemjI+0>QaZWat1FstERcaJ^l%qE+Km&8;ffV>*1zkd z`;y}?HqRbLBlX*emugw;E~vxah~0IFV*Nx~qYw=`q%XkG_!Nb)M6QQHv0>d1oM~}7 zjBWop^(LHNY6wq*A=X6Yh2u5QlkRSxz5CpJeRE@L*KVNbeOY~Er0$X3%zCAZW6~7w zgb$DTpwv*P{i3?vfS0S&%nKzpt;Q28i&v^E*V+}VHBSk5CNEJhKKvqGIda`(e8q^} zI*MxRJ)mg;P%}TKa9{b!1@SXlkSJ>E>u3Covx{DBFxn`ofsUBSVme$fw0&oFKm8@b-WTv%M$=^2WnJy(ApzUQ47ng32b`m+G zZ=TI};pMGHg`4ecoOyKi_VXL(F3!3nv>Ol_RJpcq*wMnEadjY8%>$b}>k1tsE!CHptod%TU@(~Ew)o-mUzi7qpBKAhV zJ@-o6fUy$P>TmPO&h?lrDsEd)dg=3%8ih?s& zn64e@^J8_Qs&=>M4hA_?z#Nb%3Y1eaq=s<0JWTGl-bh1xV;54E3@>+(xYLl}WfhVK z5zMf>xYI7MomjgWmfrB$elo+RbrrgflHdW#=>Ft_rn)~VcE97f8x%YybJr+y9lk_4 zpn(gpk{p7{rk>BeDT97oWq@B^+nPTC zs3a3HtRJ&F9qKu~y|p{v+}b#&$5ik@DnU81iseKGp0j{K>c-smu}^=+_TjxWFb08 z)DD0s#jrGpr3@;d6O49;+>xV1Ru!@3Mkf4PVwpjfksi+c7VzsM?=4NGVgR2geZ0rcxe#ozuMk_1T#~mR zDP&FZYj49BFC}!YO9}=N!$BcylGKG;9K&`ZKSq%?$=~+2cic-aW_5;-4go5WGp9{t zqwK*}PU4>3$i<{`TTA^l;f$-}n~Pjj=a^|8?J|};?@clU=CCRNR6hrPx!5RoUU|CO zWIs+edLz!gKRG{he&b&z_wKI7HO;K&FV4NYqk9%-;B&d2xf6gO^WzA5Vgt0Dn}%^$ zs-(Gh_amM>k#}BW?XC<-=IOrCqAaeQz`*AlqHT$C@ACPYx)e=pjcj?piSu*#?!uKc z_wIhgtB>*17b`Vc?p?m@F89tHCZVrhV+$y0?p;hsLhfBWX45=HRKS!lr1se_H@Bs^ zcb5$B8XsRUl3J^wCvUBh5a^gEsoO^GFv7dPO<2<`(o0qRw>Y-UK-As2{F3AI-P4b?e|dNnMfDtfYPf^; zW1)Y2?SLk3zW%ng3lnHR)&J!p{AFZS?S_HOU#?zycmgul?}EF{tar4+0pAAmpMWMs z7khI#d=Uf+bE{s;de>)<4k&&R$JN3nY#)(e$SlL(p6|j6rx@|_vJHM8SSGvIbhPAf>+?ck8T}2(xWvC;^@Jj~jM(~0K zUlw_o{lJaH-3P4wmJ0xx?@KK94X2k1n&nPKQNn&AeSBH{AxI?fl1OEN$`v0B8grYZ zIN9G323>lJy2&jlv?%x+d=1?XQ{@%sBGhhU9ZpC#y5PeBSTu%bMvmhf*%JMbZ!`e$ zOk$w1swo#jSyUMd8Z{2zXrQFjOnYwj3 ztu}@D9^n$=la3cdtEY#ni&WBI`f?Ka+*qyECNf$DznGokDsLiCT)*%4R#Y$> zcY9`U6pu1}ukwk<_FCnCOMX{CEI(X7~H$=B?a zB!|}5>bHimHwXxeLI|r2E(@4qLLGHDpnf`1$5S*azgj-y;e*FzJ+prlvvtq_GZ+hJ z4~F-QgX5djH^G*HKLpu%zGdId^}9y33gCFpTV^# zY@at{V^86x58^QXuUpWfeqSnc$bh$b^B8PzKu@STP=``Df?Jt?x;iy(bl3!Ph2BgNj1vTV#lr&sY}5cw~F_2v@SnbM^n2L-ED>g3c7A4GSF0|NR{7fb}scod@-B# z3asbX<(@z{iBp*pt04-w?Mi*+N_9ma$8X-^#HHZWp}3VT3Yr&7ncI=zRHdYgGgZe@ z8>K_FTI?l7$5&H~pF=+Cmc&8G<0@sePXj#NCA)9hzJrf!?9PYp>-&R?Lt(hvWGIiS zBa=f38A^Y4*T3KcJ__8)*b}=>RQ?q8;!D+t9>lH(2o4&OsJHU?{+1Jm1iIBP$4Yo}3+N4_(0N=-HxIZvvF4$kG&1Z1 zD3tKJ6*4%0b;)C}wGj2Vz?z7Uj?{grr%m4yTNnC(U0gqZS~ufd+uT~;JiW2gn+hkC z$HfnI;SSyy{=FHx?k=rsiY>q8O)GOkXO!Yy;iFC2^_!GCa%IFckWV%wE_1D-#u|L9 zzmL=#vzH&XD94)0CrG#oA1&=sZ%TbE-q>|;ckTJycXy}v7_oKSGlt!J>REAM)0(?* zdEGv%C*x?3jSP-inZ@*#-L*RBa$Bl0FQ7L)RO1{iT1HYhmX`grlo@ilsLvL$-s@ps z#=EG+Hbm(>Nx@tus#U<{w!?kFohBs|@&6Wy$cYyvnDjFK8d1f`RDUR@Znb@*Rdn)B zFCW1Wcqy37r1*q2bk@o=(J6hOE-^1!n^sfA3pTZ2pFxxmUi9p}eZ6?m2TCNYvwG26 ziQ$SOd+%*-?E?55JPw9u$Ft104)k*{@`Z_h2LVs^4jJ{D=PPGj~SaC)6-Xm!ZL%}#o9QirXdyo8mR)N_-a0NNw z5n$$m)EZ&B_a04V5dTf&Y7Gk?CCh9weQCghod|vS8-(^{@!EW6Z5L`Ow$Jrp-&gd} z+$)do??9nPhsT$eSH!Kzh$Xmtg%4_dkBXNPBYPU!0L{o;gDVx5cA`@}AsqjBU!z2O2gxZZq`Up7oG_?76S1xfe%Lf_HegZB`DrfxV?|bBY5|w zJDT7HOB6|g?N8z!SFf;Kk2UI_6gV$JRPYd`G%P}ZI3N?dO&yrytSL`l)?3L!Ma-M} zHt(MvY5u9d;VTj!1X;-7Ns9P9?;h`@E0C3S=9`pdb&kO0KnQ$b`zaQWch9k7?uknu zP+y_2Vev^Pdtqh*w+8v5lLPS<3i+U+%d#@6L@7K2FL!pKzZGPLeXhCw=A#NQjB)rh zgdpQuS!+|+xVHb9G7Z})C z#*8sAFc@qY#w?7zuxnnP1FQ|RV8FxM%NQ_h1C1mf$Pk~VJOHP_pW z?Y)a@)9h@kBl?+=dG{_z{pOG;qsx{eY##Og^wu(F7HDN$)Ow8;2K}_aoC9iE>)9Un z7m-Kp^zm}IF1i{Q+}x%tM5`>FK)*#QO)s?;Uvc!KIZTb|o9pUccGNtJ+(6Ac&ZVv8 zCx*m90@HBNnIjg=xQszst@M*5o?A=Fu zNCQ^?RS&KnY*@MxV0R~17wMkc+dbIY*_~V*4M&II{-A;NmFu-g)OQahMpmgM0H{z9 z2; zZtmJ-%aS>bd07%$!7u~7Ce|Zi1K@ej_V(z?<{!plxe%!%@9Py*v=Ou_?U<>aSI;C= z*0hk9dbI0Z3F$SX}C!K52XU1oPuFdMtmx6J{vw{|m4HcB-Y z>-1W%#&dMMvmz%mtO&L*i)mo19~}~-nj^!jH5K4BM=nyp;wI5ynmUqZZ_U4_Y;|&& z4L<^-2#00m2(B2>TwU$Ff{G@*fITt=Gwy<)c0>2N>WhAKz>Vu?=G12rr_@%&Enz{sE1^cnK8N<%%u62A)l0rm1>A;(14e#)4~fiHZCTU zRu?HtAGDm7@~4p2?OPWEbn~Zte*#AJSU;Sp!^Pr^+_5Dl=2>RoTC-%rjq0UV=7l zcjC6K0CR1X|cJ4b`wI7p7e*8)a6fC)D4(K1+RIK zR`^H~ON|%kGLM80b-4AGvb?&3T8LygGRveYrYX((Ku_w%UfX4b=ame_s`M&+Y%t;@pV=X~_VW_b2H0+Of^MM_# zt^}U0*3d^#dyw6XZYg5F%m&;j)Lb>IKOCw@i&}T>cgpJ&9hB2U-l7pM^Tyf-Wvm@0 zF=GKK%Z7J%5ElA2>hzM7Y3(C75Jl`ROjqi;|JZXgEWcsfezw5IWeu85LKYDleglzI zqk{KIvG|UcZgBIEE!`qO;3E2Phl0TcSV<1SWmV7TMvCIAL*Zl|W?eTXOg#SqZTndB zJ@avI8}RU+dnr+X(GciJr7|=~Wf*6rw%}?aQ$Xk|YK%a=D5A0w>AWK5qlGUy^o&RB zILlBF-cw{sk-cGPyjF$4mqgL+x4qpbIs&^Ve5{U9 zK4p~=waXsexeHU(MCl7DYqJ=|DrK4)#mfAi*IV*rFFF^AMVUW})oZ^s02hy9W#riina4s18vDpm}2ErSu7)B@d!|cnRL)hjbdd!%F4Yy%fG4U&ITi3 zif@%OmNI0oMNw62Ds=FkJmq0!MFm`26v5+}1!@#4XHi|k_*PWtC{}Jc_plc_Qugd# zK#E;Y?$AjtipW~AbPDU+pU)T}aMOzo_(av%{lqSutVpiv2SD60nwqRhewqe6w+x2$ z-b_YFkGqPY@n39?*hw9m?L>+OovcazZO=c|r}W+u!BLx1OweR(5k*bbq*)45a9Pq5 zp3nQrNPRBD$CWvoi!IPSnKSOQ>41b>cY%4Ds4}B-xi~6lPUWH;fZoI_&=u-ZmfpQT zw{`y6$z^l*?o#9K+x)p)diU;r-P>`(fWmtug*cgay%TVWa9O0!=F+=&_scwc{^>ks zu-6^CONNX=>D{~gWuC#{%ok7FIwLvl$k*%eTnceUFJ)G?2ta)IF6Y*s(?xgh^6v&( zlij_`KOa7uQbg2@YpSof#CPvvMl$Z+#ow$RPZ4DRGd`)GPcbD3oi7yK^#r*Q>D{~g zSvrhQY8Q*;i`Z2Z&l4wi>D{}_Z>PI=_XD!KclQ_UcFmsOi%1hc3naBGzI*pr4FZ4B zHIDm%nPWzJ_g)j({UqNt71OTk#}nLHPLIqN1?ug`y=@@EUv!?9(}Hah{biw~ckeBS zA@DN=B)fZemUEMK^{L*vgZ})Nxr7(ry*q3L0MC?S3-1ClU?CvAd#_n5bH2x}?4x6K zy`X!-w|2qo$vH=T{XD4X++C9ro$)g^?VP*!z=&E?=zRD74MINOGO5le^;<}$Z~OJ8 zz4J92_ykCa58Q5&q~=*muJpccbGD;)it+Jk~H9>AxXHPm`&r8jDm_w+S?= zlKsTvx|Q}dYW{59N-#V+w{Ww%ZnJJp*{uHK{^*tJ#`POU!>cNb8^AwY1Ap#T--tK5 zM(o*4)ZztriIx{O$oJ3?{tV!c?p%-y8?NlJAv`oSqnDk9p*H&$_Vy>}%)IEAMmtys zQ1q95rdwHa($#{6Y0+B3)srn}D|vxV)cb^2Ck=`4+C{C;rZuFrVM7W=Bx_4fYuZ@1 zQ6xN_{WFFw%8=THYQ9@>cQoy98mMEEl!(c{V9CVMkF|g9@o6{lr`?eI=gR> zmTfhq+vc%>AQ6j=U%qx@>)_l|^}jr^j8658erlzIzQ;oU_06S)TcSh%KUaV^VN%ZmoC!!5uvi}lr|CNj{;IeR#8zd1xOQHX@KZ)8KTJi%-)j+Ft>GdW%Cc{ z;KoSu-hcr4UiYl5O=`QDWmjsVvh>Jg);*&OFwIUuPs=&HKu4YWDD`6V#|6op`Wj&k zjUmP#OSxER)^cAe+M&ys+m`O6aAwEmmqg}9$bucdEIMJD4_6w;+hM()9I>fuowqk~ zzuQme){%&cuMAJ2;lOa_UcZNtzs84Ez4rntG~c! zMhS&bU-tFV6gZb&zXwf=xwQgP&&l#{#yqQd$855FHW3KNL`QCC;d(4c)(%mTI~HXL z%|hCJS@DpN!RIBB#sZZq<%U83oG|HPIBFJeheC^zzrnBJ^}}3!rKJefPuL45q!``d z!@;mx=u zX))>gEBEamoP|T^K)<(;wVzaheeLXWm6>N{)i`Q7PsYV(CuEoXglxC4S%=K%Erf#n zU{;QAxlPE5^x|C$$&BBmWE;-e^k9?puC5hYlz6#Ym7kJbzRXgx%bioQ%eg7pnsps! z+dG6-uF@+A1;s<6GVMAG3-jWGq^!ua%+6IoUgrC39v7yHNtw~IK=ZPPtYA8r*veE* z3~Dw+yI(7h71zYFbeC&kHRr7OJyfHnsJ>PhVwufk#otEZquU3UARD3YE$KaVxIoNC zyB;=FxMfr@*FJ0eTxqw@Tm|KAo(;cTp|JO)VcfZxbsu$rQVl=)aOIjU)EZ#*Roa9eFKq^Xcn)SnO4n+n@{ohD0lb#&2gvI`;qqh4KnAwy^C z=9)f)KTTIJA*^^svEDwJToEgKEyQh9ol8{5|ISdI>|kj@Z+q$z8xh;Mw1E@Vt|BE+ zJ*}ni4B58S{69oQVXu1!Dch4wAx_eJq59Q<+Q_k}?$z=g8Xr6<>y>?1%oYp7n88#! zJNmr0IdEUcwt^!3MH>NMrmy}(dQ@e^Wmq+zEv9h+i|Nuo6>&mrA}P&L`-h$iol5yf zG=#9vmS!1AusoZyrS6Z<+LoGHz4`ioWkq`Ss6XzyiE`UkPSL#GPSqc#(-fX!o7&ZO zc3lUlCCB!8D^~kz)hpk%biU>Gd9jiUSE2~AL5AsOWnnE!DgosBg*o-!^z7Au4F8_1 zR252pI?B|ce0re1TI9|H0M9JJC>RJ@3$F;hZRHM8{e6VXiTeKeK zwX0C+qS9I{ezDH0suZd5dykz(KLOV!Lba<#?yUD!vDCFvpLIkpAI=m%u+X8Z2o+b0KG-cSwi1q3fq z%N}*fr<)HQT6oajWu8{2Mkh1wGJQJraWd6n*Qm<>p_gAOMa@EH&WNAI7eBJ1pF=s- zNu5F3xs%Aog)FI<@=NEO{$LHD@6=AAl8w?%a^Yvj~8Ozo9vsdqIWm(*LARUm$c?lYB;26CJX}DujWtasQ^3MFqi_0WvxEp&gHILW1hm>^iYj+x)&K) z=~!C#lf~O#59(f^tPv8KGh9f{rAI^ZW~#G_&zmRrE*bxA(d+QTrRbSiIX8t4aRFw@S_>)sitN$#?^iWKfZB z?~y<8(!ueiObK5VC7IEtgCm)!rr+Kpc@0y+_>?&E=fvB4_P)o6Q(WHG}ZH7W$w{cA~6nd~e zeDTPF__Z;D1Q!@wus2Y@K;=t`FMAeRU4$>*ySKZ4U|7QD2|@f?k|8?0g9r#8za(a6 zbohat^XG}=w)iWdzKK+-{Cky%TeH7jDY!tGys&8-)O*X9xf?w~Kh{Wa;CYDNq(#mk zY&;440LKc_9izy{8VT+~pMcLoAbw2D2P6J=jJDPYddn5P)Z?zgjv$_deyovVDcJk7 zq}P8RVU5Ec_5Y3<&WkV=G@|rPix5K`kcHhx9lS>w-(Jm!Hq3Ayt}0ir)Y{JZ*wVk& z^$B>cULp@&AbrWLH#5( z8x~)5GzTjab8C=aGP@PgBiy7_%wte zd$h(!uHe6b5@hiLAh8^I{X_=swsbNVcXH}MJZU9c{?vhd3xF@f%xXs83mcF4Rr_S= zHee)2dpeQ)DD*mrrPh!tml`N0M{3_pO}0X;@9fs@Q=`NB|I0yJ(QlQMrYuK|RbswX zW1psspUO~UwQQZA5v#Ewbu5k_w3MeLHm#E7Zely+qe6W+Lu}Qu9jqTnm77~7=N!4^ zaTjhN0y3O-UPf z4UoNGIijBuAs%GzD_wirqkeNpl+oo|xoJk7j#kD+t(OZ18o}FBVIN`Z*&g>7vC`V< zW_MG<-Um|ru*=c%$d)7?5gh`N{p;h zO8`)zAjX}w&LXmhx^X;mI#M5@(mAp520#=_sgKGz$}PMV3|tXx3PDaA-+z#NhCXVo z5NAtIMx}Ds^V&MX+%*g{z-wYX5;g#y_iS&sgOtQ_AyP-)*DI)KBjBQlaK17=Ur9Za zP+8MLUh2`VcO|6PEa%Nfav~A-{F$1T{cg3&ZmgD7ZwOGKr@1;VUS<*!7%e`o#h4Y& zIy?V}M}ezH&l?C#)~l^^isj5kGpC>B=%Pwn2;Jq!@SNVb)9>6#H$*N&!EEeK-!=!x z-rDWNT4s&&FH^HOSSzegaUG@8i7i09F*P>bQ17thL!|*%Agnfv8>fJx$wTBF#`jQf zSH=t0bBIJpt;iemZp!WwFms5oWGguGIxnCSQh()MzWTvb4ApSZEZWW&!h;ZsRN^(w z;IC_R*o+7vF7@-&xYpof?kt($Wh;DQRPcL$Tw8{onNwE~Vd%av{F&5%76|0(v^;N^ z-ee&B5;gd6qt;gHN#fNHq(aulVw@>J_@vW-6jxsbXaBrE6@513>R7qCyOgR#W)r#7 zec;0NwOr5zH$e3U&)b5_R-ay^%nPS^UBZYri8cBTdZjw06;9i;bSA8=({)tPDksV$gBu(`YS47_kNI<(M}!F0YseQ-z&D#O(j zq!}`VGVWy!M3ptCG&V!gPq6yFj?JC=_7G0q$&eavG1f`;WCxO&LhazgP)rI?ZtG;S z5T@|lX;FPlEsu_7pt_ikH9EG$mR&HjiUPU$!%{6Dw^iBUVoXs2ft}FylSAKM-J0a-%rh0BI#kK*s#YCTxfDW)NfEd;WVk?cp7YCO;TU@ zyasyFgS`t6UfkN=nd}}2GF`^=sc#*qhh&kNF}n0jn%+qGmLZ>%ij``JYS4g}tJA^@ zH8w6L6BY*Ei;KN{n&w@bt`(Y+YczODa|5X2n&iPxuQ1F;7-Li;CBTx@6z)ErmE~|PzH&XN-0U<9LAY>9w*kRW7 zT`Zfbg|>aH`JVZ>w+*-;%f5mrz-Y7?M7Jag8l*Ce6RTLx6cGA~8Y56I2(7GyMS20L z;W%gQ1ZF&9$61c80N|NY z=Y|I5?rsccdq9n1?aZ3!bYE6?bn=22#oC#<X&QjE^!VprFk%0)Q<&2qIzYO?`1ef9R%!Pe%J z`s(mtG(S50*h6o-H+>Yhk`mv&SLL0T&E308jk|gA=W^-YyZd$i%;lcU3*WuFM7S){ z$5-jyyZdFHJ<;jB>!pWFhKxb!-Mjl`p26e^U3B*@U$2U(HM90`>~qfSThq9RcL`V0 zyLb0Tymak3U3B*@|8Af)+15)e?*yLT}o8F%mEZ&r_|hzgi8KACxDFCuN} z-Mh<%7si*MLVE7b-6Y~ug!t~=0W&I z?k*O#5M@iZFz-p2uSbVYu3t~@3AZU=vZAZ z=-$2CY!9bHj%UxYt;O!nRI$K85TY}4We zc!{>VutC0uhVW+qe{@%aFKl=S4I9EkQ!{$mSr}@we_?Nbg3io~erdFWbpSMN~r~ymnD5_KNy#t`5e#z<@zP3PvPrOHOOrShrCm zJe~bBhArAi>_Ro)thgId`^M~pL4DIe9h2PGS}lXbUE{_D4mL|0Fe~s+;e$3J!8!D} z=AEakaVOU8zQ78fv~}G!j|~KgSZw_AwHsRp=bozn<%wl={iT(~9LKwqU^u|BFcP|LP9K4RcowEm;;NYN#;U3$f@S=CW@ zsvpDWYGDi3O(YqL7-MsnurEVXuUJCXmehODIUmvY?|Fn$i}NB zYkXoO*`mJ>B9p~64KH1!^&>)M+bL}#${z)!h^(TbTuOHnWL2PmaT%h>{LJ2y2QasF zbY=4o=-|dk@_ImkJ;8}LJ@u@tO=`QDWmjsVvh>Jg);;4k9piz{O%UklBhib^9~TJR z-Q{8RHNqMiLySL`avR9q2ULs^9IWbL&V%#aD)>&~RWlbFbg;I3s_F@JxL*{sOdAp|IQ4+kY{u=*2R~ z-PlBi8K}R&XGRHyP+#`-(iAwCUcU!Ti}!RZAoZNA_r;iJ74H-Z$3#bNXW@D*NY)Nf zkvkS;3C%*Jm1PMCBt95su#L!m{<-{9Bq`eClV(o%%lZR~{; zQjG5K;b2%aHqVUwj=#xvu#XN=6b$jq5}@5xa{@wiRtzbI#xS~U8$T=T7c^GnjG6AN z@rCQuRi^f&%6G`OJ8xTTkd`TG?cwo2l)Pl3l*cQnJgPQ?kpsDcPF! zR_2|qAN1f3QJHp~g@t+XK~h#^T4v{}ATRTMHjfKa#S9`@L$c}D=7t4L?Nj8Q90lcVo(;dPnEgB3WOaas@LRns^!cqzlZ%^fsM3rf zEcnL2mB9b$e;ogA%EL(CKO$TbO$Fg2Jhn_ofodwDkL`geHh>51YhmAo%KOC~=( zRGX!lG%@WqHGytANO6s~bFX?i+1nhPdhc9v9%VjOOpa>W+8a*^0}F*D&1}t0i-U<= z)Ssv42IT6YPLrj&I&!~SRmD$bdu<)uti-SWG+n)f$j2*+#kqguXN-U-+fk3?}Z?K-rqc>2&Z>Q;wTIy?H2w>ijpK{kLAEW}+m@w~eK zkRDYTaT!+4XNzfEz+#f!0RbpmXcQ4Gx?AY7$W+QdmS#{rpDj&Z5+=y#(l->*s{7-! zwxy<4Z@&ItS&^PS>W{l_qJq9~fLwbf->Q<_-hZmv;;qf7KTM}7JYhAptL^LE9i)~V z+vlxV?W?f=d6(Xn5q1`fOjF+bWdQ*xrDj z&~%^Cq^|N}-KDEielJ}K7JQbeOTiv)6=OMBmmknk)h=bg4qn~N6rin2 zks81E*tzr+;;Y$i6X_jAeMEi|XcloQQ&KfVfwWz%r%BUp9jieJt6KEq#ZuumEI3su z`RdHIv8pm9!^WUmEmbq$$ki6(cVQnrmZXP}=hzk;qaTpPo9(-UZ=W1&c|$eCFCaW< zZzxZzQ=^j^HQBRmA{0SUy71BasrU)2@A!<5G9>qoYkpL>I~A( zokTV+WJ!FMB~Ajj_Fhb|c+Tk$2GW$iZ%bjixs!CuSq+C5D@$DvldVtuU2aoaP3DoI z-R&W}Gu)}}7_E2O8LHLg$kg>3Lg(G8vs%O38+zrQ%ds791q;!hrG9IugppgoZ11so)%y9wB0w;-Chy+VEaKs}c0ZOoI{y3hyq#qDR$ z>Bl%XcXqdT&cTc5%@Bp&-Nawl!PZ^Unn$taQQo+)B=n6E`51hw>C~I@x<&ZDt~z2K zkk8&oT;XPr|5pR`aA)$vC(5x$g0i`gqaVGUio%T(DZ9Z&b!UbHni{Vr+ag*~;0@j{jmI(p{r z+3rpBQXdUx-s-{3gg2vpWM}sPhR;EBFti;nmhW|dX&3TmcSu^?L;=@9G*sl}Xg{xhUvy5Rd01NpRVt`x{`8-&=~R(CcmRQX|Bm2K_8l)$usGWTy3O+zrEM- ztAVzNb3w6UlAyS3-S{@l>k(d-bSg;~C+uOTh)gn9> z)=04w?EP8NYj^(MhQTqc{w^aZ3K~)RrbUP$4u~|}M;(~hrz%gcz?IXM0_70+_z8rN zs3k@0&zt5Qyy6X6saz|{5p|K6%YiKW_}=sEJl;chSHtzUo9&o(P(Mk{hQ${h&B4mV z+#2K;jUI?zDC7qfUoEDRMwBBnIOTGs$v57-cG`_nXqmwbV|;uX!jL^$<0Dt_UqA`6 zcma?Y7ZEpDGmcvQLw*dGu%&dOFM1k_;*TTjl{_t-F7|GF| zP9#4Hy$)ijHKfX=28zkS$Tw4ytq|)wyS4k&=&=6(a?n=v=Ma>k#ws!2su2t3rL)?9 zDnpIcvUPq&tj31au{eIvQl667v`UtFu^sZQoV=|5aE92bWjk0ukSaH~O3pcQ%jIAt zlY?w8SBWWygHI0B8_}#@Q%h0yiboGK*AvA~*^yEK6HQhI4?`H3NMkiaw?bmoW! zGcIG0>;uWYgp23h($a&WXoO>Gj+gik}drpy&1tX~61_dT{k%!{i_5u}pXm#v)jUIq-S)-9w3yRcZ+UDip-H zv({Ne_K*sWM@}ueGa1hm#l{-|Q7EMrhCLW#S%85nf=waFY2*73lF!gbtrg;I>B-2= zU7KuKGBs70C?WBz1+Jj^ z9tEx*jg&n*9DA>g~#S!Fmpn2&ol$W8O{KT>@qf z8TQF!Gj#Oc9 z=dhxdMvKXugHMo`chJwpEO5&3f9s_ekZfR2%F>Agk9e+#a9(|INDL~&)fA)|GK4bj zWeqG7Y%Zdwu+e;kzOTct#Br$wX@?A{@qF7*q92nAl9@v7;KNW%fe8smh|un|a6G1# zM@KW@m^)DiE{a*F*Mc>kqf-)FSiZ2Q6x5FniBZjw;nkW7@R}nRDPVDvSU`cpn={u* z(N}Pg9A?9hz$n6DSvi8UKCrjVbG}bOMH61Y9+`p}cQNH@H=jwIQd<$b1U0iaQn3zE zte?qitD4lT26pm|`hIH05=jp;#fCkG;6ju0p?-tv38zVwOsq-j3!m3OFM6wip%3+zb)xa zx0QM3YspK{rnLk^c(6AFmyRxO9YE`?9=*7@KRN&8;ya(6T)ec{+(Nq*~a4l{$IyhaNP zp35|L)U=3<)ep+(MEx=TDhMwe?>52P{Ta{OAQ-uPD1u*|9XYqRd$6^$JGnR-!n~}o zCb_Y(|7VC6mhv2jt^8OcGXDj1R0RJM)&sLQ3}qyW=S!0+R=(ZK>ZpRjg6;a?nXm{A{t7c7=CYvcIS7NeSPXLuFi4!eRI~5hQzll+G?%b zi|q?+V$zek>$L{y(-?A?P`@UP@>_1ZMfvUT(MJ8whRmMwOX#BO86Pt4vPi9!Ojy8_ zYlpLj3S3TNMGv?x`Ln$jHn1mRvR)AezuaxJ0ITZQXjz^G&^pir>IuA9cLdFnhkYb< z*PEnxGl7do?v@pGJq00U6hfuEgu)p9@MjW|8h~wwtAbc0{ZO1=nbFf|qa7&G{Xx z$-vjD@TaZE(h4W1tGBlfwl<&KyZFr3!DxPT__2rHc5hlHN_?MRJ*jc|+~>RGxcgk8 z`+WEF-tH4sF9_aV>3pf+QsFX5pW?;$`R=EA2K{q<>HB<_4jGH$`+WD)Jd4S*{8>@z zK)zr-(W}{|d)&jjD{ASLC=Za|=ew)(0?xe8mw!0WobWzh{{3S2`R|3AbTk8Iw$kQhuLbGg<|(v%Sx6 z7xt0Eo{Giy`R&)e9RWG_cTPREOO!qNi|_MW0>V8&cGPKqFTBrhf8gQ2KL2w^{1P%o z5Bcn&TYjJKFc#zhawXX!y#&uJ7R2}YHFsq)0ocHs_)J`{=-%ghC;hjNhPNDQCB^J! zWS&5Y@AGR~qHI9UCf?)&WLc2FmO3W_tbEirN)uUcnN-IZzkdJqrj7138{?*p?(plb zzX?j#UxYt;68GKk$K8Xu6Z>HWucTn}=VK#isL|7GDze71d`jJ>(CAC}8;@}-xP@!f z{Ml;h>b+E0H+n=Cs1;jr!Dq0a8t9`y(d2G(^Iq8?{oesGkU<3sMd*;JUa8vf23d8wNj zsDCHDR%+KDsrH}8*L$u5@{4HaV~U-G;0Co1sVop0sLpBufDg&HxoWPUvAH;f4*o&@ zc#X{FY-g;@j4A@r7oIo~T0zL=sNN}z=c;zr7O05O?taZjU3(+(t3HHJ{E9iN)~ul_ zma;J9Y1ru^u^*G}Gf)FvXL)7BKMJ6cSw}^@l=LR>(x6fQOd7qP*?aN;X3LJQZ2keA z!xt+|Pcbk%TEYfhvYtcdq{W}#ZYJ)zbDukUjE{fpW2fauE3@z!{ql?1H|Xe6N0u)( ze_Tkgv6A{)X_=0p#vh-#=*)!fC=?+1P=7MYOoxUt_J=*6a(?U5#p5OKXrk(>GuIw-*V2MM~Cduu&yE@v*jXDM^30F zI}sU}SzP7q78Su`S%T3lr5zeIv)Y;$@l*gUI;f2)H^mxe&dQP%hNourcAT{=c^v#4 z9&XImnw1zOQuWz^I++DOcXDzt>ADY&yO>Fb_zK2FQH zQ=?g17*N@T^gG+Cv;QDC;A)y&Hbt+^wsZX$g{d#^P9GP~G{zpL`k zQ0cX{4=zD+L+4ZJ-E^p%uJd-&m3B9kc`QdunR{u&GFQ*ploXgq!ryK65TPehabYuI zvFwxBrbqq(+ZlC-S|6Pdbw(is$<+ziNzU9t7Vs*7Akz0I2nuTOr@$zkaQ;sN^|tbA z^`6g;Pp0bEzWa`a*|IrQ|D-sZK}B~wTW?B$aZOm#G>fIyN{Zhm@zmGPrAKb&L&hYw zrmn&9w4hK-+Dt`QGkeW!-|8;~vkr1~Q76~V7Do~s-O~LKxx`j8OWN3?)P9pxf%ZN`6%vLJ@zk+04^nGuiK!9I{ zi3s+6diH-&Fpo3hGu#=H_Xg1y#wF~AgJP%Vn}|%JEr_)osKrcP8bk|a_v=JoXz^dh z%2*!%ORgVniUa5C{}omR+N1#bD!ph^ZKc9&>}18hHn8?!U$p?JKN6%gJmo}X{CT^3 zBe9MqRiKprg>;Rd;TNfNA@09Wskm2FlaWQBdB(8s$AkT zx6W33jFUHj1qf;DT;!UGDhu`P#2clK-x`Iy(j|E%7f8FVb!pC5<=#ez#9DG_Y(Pn% zcY7P(^ET8C5IkuOuCGDdx@1TZ^^s4&YK%MSo-SuI-ufpKu-S-o& z&XW#MT}l^A1m<3qjg+W!`S*e=<#_G-*`D0pqONrjdttkONuFx|O8NP(`EJpA7Z{o* z4ixGpttlBz^12|OzV^?&EX zz;3u=7g7Xd?8@NfQv>y&&||kRG`@Xuu;pC| ziXVh{$XJnm4d-C*c0T~ zso&=}Oim@sK2WKROW6~fq$(#~Je$p_TY*JMuLS}L2kGi=5-w<49Gi9<(M{6%O2r>D zHz6;(FgZ*X`t%;QJIGz?^ypN#9i*bo+d_0Lh}5IEI<3CrZ4(8_%fvl99R=B;)Nc>P zhqZv?aHHrbxDGo7ZdB|}QBWftwhAa1l*VD9kb}LVqmX&Cu#rMJDXH%ms5fPY6y`~B zT}uQh{q1MZ>91rVK1#uVUbL#86*t|SxQ(bmsrLt&2NbyXPh}*d%~r zYDWj^1M-|`Yzv0I250390gQXg!%8=)B=y05;(yEjJ1-uLuBiTJ+*5u@ zVl7)bcI~s6O@#H4bj@?0SjX}sNWs9%;C+a75==Rsyr+#0(~TGuMB&f&3kP&+Nzdw7=}%@j}^ma4-H0h_iS(XCVQ!up7kxEiOK}vrJpR^ z**$>SdeFrT4bqF{TP~oPFymn`k{H*6Cc9_ejC)b?<6d61ORI%^6e&rA%OD*h&`%bU zALf}8J#{cM_UI=IJF=H-t$0gOQ`y(UYxGTwZefq4v>&@c^NpZDXYXRHWCGC%tw6og(5XI^pB?IcD zo-FM8Mek9A0&?WsTw~?*lZ7M`UMjf+l~E2Eis-71VRX&9(&P1NS~CHRWs~VA3rTLn zG%_)MoBTcXWFh$jb`wc!lua|2&=#6sR&Yg2+#b`62D#&0$rjym(t=eGN)TC6ghjj+ zqHcfQ+}htffHIA}i;|S_t4+%Ix{Yg^Y1xDI;fqHy-fbi3a{)D{K1OC*iJyChT3xO$ zR^yKiQ$#7O{Bb>EU6*V+@cSe=)rQ>k(BXF{cqi6pH2gY}eYdziUi zDY$}|ys&AT<@?K+x%xc9ymU$M;JFO1!;V`A*})4KedKMz{+c7qOP2(Hp_3q{HxR$3 zW}Xp$JvP*KC+x3ybbIIS=Z+vAhk5CeVk&S3c@ivO7OWu}sHv|WQZGCsd<9)Gy|XCP zSzv@rbh1|KWUUfHE1Ef zs93X7Nt;lX+~5ec_yQd++<5bv-!epY&cO6$Vk#WMl0AnLGh^_dKmoLT36NTj%;1t+ z{%u>WNldOm$Tw=2F9I^kVc=P6mLEb4d>USEVAQ^_@rZEJIlv`LhdL4cC>-16l52>~ zP(7$8jS?$c^)2MAE<|)^w|1Wz9oGLp1HlywcEL1lDo2@BcAryb)Lcel_@{G}S*=@V zb>zxyh#kvg3oYttiB9WfnHSw5Ar7DfKMx1Zt(#nQ4Y3(!R|z|9ZC1Lvk06hUm0U2J zZ7^?{jLWhk`hB-Y@Wf?Pu2J^BNDrQbFm5_N+z}OHijKFSq3Y36>XMA!ngL569a7`svUL>PEW~5Bikdop3)EX;`S(9Zs+6<( z^pGmt%N8W;p7mb!L?L4fG&3%1#CA{axvwynMvaVNJjnxPL{i5BL;+qGg^h_0=lpN! zTb+JpoD`(hn0^cKOT`kU({kw*k@+|Mn2hpr;Vn>OGanvOhYd_M_w4WQykmDXU;pn| zE}L;7!|bVgJRK)e5A|@h0yj>CK$rxB#GG7$T<}-x(wc2QA*r7wN9TJCRq3HlPfD-b z`0A*-ee`A0_)0J6Mppwak2k0sRmuogi!Or_7O?;DD2tx`>Z zRjDk-#PocD?vd1WYmgTt27jehNp$j`)#1elaciZ#C?j4K#xE7>7})HtynAt?}j z5M7~;p`M^RrNAqQwE1o_1r7%Xq9)g)VFRFg54^Z`W%Car|I?(svy7ot-YeDB+gk@) zn@>We{no)~esuV;hu(HCw-?v*vD9u#aZS^=`_Nf@#T$)zBJ(}(PD!v?&zt$=+#Rw< z)dhYylzNX^<9AxitT)!E(4eQTtGCmg966AuTVx2A&c^9k2gKkRW4k!>?CWI-G`PAn zwpwnI6FAOJ>a@k!oqnySHg)?U%_HlF8?62c)+-25T{~9?ygtf-{ zyw_Di-8cjETm1+&RPwRmfH4plp9Ypb{Q7<5BcSn$^=u^>YAXU$uFLKjk~vD`ABQB5pLnuJw@g4^7i*q{uqlcUbc+ z3xHOlldX|r+dU#7!U zh59!`>d{Y>x3I!5M=sQ`0eqn-L1oQrp|23~vl2*xM4umsi({OjAX$+kxpJ(tsh5w- zVT;dpF*QuwOrX0gY>%sz(NYd<<4HhlCP&{w^RBSn5#r%MNMgS?=^^?Sdh8DB-wvsf z)sg+x+6wZDS)vhR7dH$9Kdnvo7HdA9^$1EnFbPZ}oLrwJyn?K<`2DA3>q2A^W=IA< z!l(CC5ooL8qwZGBE=A28kz^P|wGd>o+v?}F(Oae0s;k#urf4J>-AZCp+)tv>oe=iSsmzsM?x`8h=507?Evb9L0Ol1+R3 z>XPIet<5npB#vKm@lpoH|2a7)ky*r>i$gRsGCzmIqTDF-U2pypHYt)X!FuvdzUyyU zO7A;xfX^q)NkF7`&`u)DVc`>9S!gP1yZA1ut%tES5L!9`ZTVJkTbVC8NKe*D;_Sm- z2Y_oUSk7Sywyu0ztMWPZ=9T~B%1}G#%1iU6K?ITGSmxwt<$|K#F3 zpPgL1wAkEoyGa8`U;9tr*1hys)Lo;*#i)7gSNggXOOBiBVXtWwjL14um|MX^zr~p( z?_hObJo=U~D?h$;+Mhtv@g}%>3Al(H)Nc*dKQ64ppuRP(Je~?fT(gzD{fUmPDI!x% zi7i}KUp1%DsXIMv2sX32B6-T1!ylLJL3lHd#`d(95jV3f*RT7kKOU;b${Km?ho+!d zd;(Apnd=?Z1{30!7uf|lS8pUvWUAQjdRGu-Y$M8DhIPp`y}6r5oMzQ;BA~DHupEaM z6KIIfeqmhLu!&bSlNe)jrA2h9ewT`unibrgwHRYHxTjaYqpQa&Nj*^XOc+zp;j2mHr@db23XZa>bSl?BOw z!VJ=>0BRI%cQr~{j%UUmHHx+~d&wk#?Cc~pm3kCybFh?;pJmwdPVp`XGLsABC|YLi z1T4()A00*8`Llt>ga=Wvf+}+_;9=VZ(}6iTingne>SO=oWp_Dh6m3@`7yToDi9IBF z{8Q8@T80TPxGzCvbU>KeVV@dB%W&H*%x|q7A4SV8=!W^>wL)1S;-ySUqd|@x z(N1B1E7`KLNOG+&^zIVxAl(o5(rJ>YWc9MX(_@b_zexPjLTF9?RPkb7emC*;N&XZtCZdtzmi!`aF z`6_gbmmOu>4X(<8wi01>sJhWdTaTcPnwlhv+G%uzmnCOfso^i!qib37X?gD4SS`NM ze{Spivy;o`M&BjJ-N^YPy!b}n{k*sRgn>pCjG8h;tkDk!pmn|;aH(*aq>u698-4fF zJcIs?e#~OO%X*g%8H?f@efQHmi^0*b>y5sA!TLDwRYx$pR&$3}_D0_&RtvtvjlTN{ zUci|*`tlD4niJmW%fDajM&DghaIhEO=)2!#b&zwTFaGEvC{~zA$@r$e(J!-Sra@(D zHS`ORvaGwyh!^HZLHe|g=VgQ=ztML&4g>#unF(+7y+zrhK16fWG_f+0BChw6#wEYe z_m~Y12Kf?B?=NlOSOjPC8~vKmDu8u$I{h2{c40rtU=Kgy8~yg{-i{EL(TE>Ui~|35 z3El2=Zg2NsYiDea9)zR_<92=@TlQK$XA@J7G=f#=xj1c2GMiwHFTypJRx zbYrXR31)%62!HfU z@VnuUy9skg_QQH+EZWu^89_sh9*k3wB}}N)oeGVGkHjha7OEj=0@om;p$ z>RNrJx^evmXjrW*Z~*^s4g9%Vy^m-98}~IOMv~ zkwP>%qnDk9`8@j<_Vy?ETNKexjf&8Yr08$^e7ExEB&>xC{lQww)$=xIE0KX-)Hg}D z>4r#n52RLV)1y+_z#+(x$=lM?nmF416v!ash%q6N>>PqzGtbo6xEuGaIJ=oTBzMt)CJ~p7x4w2`>)_l|^}jr^j1Mx6 zerqL!zg$EA{mrF?TcSPkRq7S%qiT<|2-Yyx+t=3iS8TpeJ^R)C%O&`ift8UWM%qOL zJPHpxT{638oX>>K@=kZRq?4^bLCg!oR;(LZayK3X&*SU8{Y2-XWX`1hgk~pUfrHwIR2B&7 z*Rxsx+P52Vi*2r&QfO>W-W4Cu3H-3s=4^+o%#11m$GIA@f{@Emy;B&^fl9#%a6rBL zH6N7$6<%PINugS^hAKL6E&>OLO&Ibt>~xXXk4be7RnT>sS4RA!0GbTBBy`-wTNPF; zKK_|BdOx%G->t*_KrY4Op@EPUpvtqT*g zZ_tsaK32Wh{Ba?{AFaMtTBu{F@uzSu78*F+n~ZkyH0I5v18Sj=9i6}u*&HDab^c~^(}^hdLAF8NitqVLT1bDv!Ot!COe-S zfLL7R?G_cmV_AaHETtVrAv&(ZZdtC^ z1c|b2R-oZlTk=ihqtfaCvz9Y!8+`$Y)R(JxcILL2c>R_8_7Bd&$#tMdd*$se&XFiv zrJ2@N&E)Q2ZP^*!<>_X0A-R~BA6#xTx-#W>mrVHBr^vuePPbwHrO4?ndpX^4l>EC& z&95JZXk z@Cjl)Rcq`dXU=7F>~_*ShDFBX=$QkIvJl7mp~1J6SF5C>z7Pkl&PXY$|1?zpq&S;F zNq1a~KPAApDKZ1-xl$IP{jL^Evc>*vx_H}fH5PFz#h;CNb4Fj^WE(^6U& ze>qf-mp1u2xpubnXRG4;>Fl+$6S&o}`g1{<6oLY;Aa`mN`K7kqu(VfCcv$aZ7{Rk5 z$-8kxa^vELUd(Lpg54x~k*eL+z92=H9B_z8DICDpMbegtVggEvLjC$c-J4}te9=Bv zM-OMIE2@me=!T7#Uo{wAp{pw97Y^&=e5Vc#PYs$b&!g(sNt4;|=o8*P(WNUQ`eM;K zeQml>hInzZy}7&f%tV*{t6vn1HNGRTRvGsVDX7IvUK+#_ z4T|{lC0be-9NP*y`t$avrZ{lE{$F8LpiK&lJ8!D3RG8f?r1?9tZ=X3Yk4_%~*ys$2>XQl`!Y z(&TTHI(};u-b3?jrAzWkE|4}?>ypO3cdT|A6He7j2c>4)TJCG~ zP#c#l+;3!SdjiBZ=86D6yBu9?KHk2g=sf@lon0Bg`P4u?C=5uQ^(F^f@mK)k&bWlr zL-tnlPIbp@N zw@&s=qEZ`|vZvOJ%gM4#rV1=ddN1HxZJ4j_CgEuahBI%Snafuy{_J^jlYowqhwUbD zmpVN<)oqigme1HF;LIv@>8(zyMZ9ey&J>qqCvz*{_r>2HDrr;~a2%$wI|{DDPJtU0 zcePYf{Q0|wf{H<$EAhQYP8=2<3BY}_jlIQ8RrH1(KYVBjq^7ZSHN=a z)E5To%eyo1&S&+7-hDeyJ~_EK**!NI-Eb!SPw58THqIZHk!AXGNYtA!oeOZHa0rKqVu z!QzF9uSw?Srs{S$g12E561i=`U|uckC|cOUYG+5hTG;WkfyStVL&vX&&l{8wb5g_1 ztA*XzsQnL7t={dERqsUkeY15@vyn6mfFwq`kz(;u$t9?a62@o~#*y4@BF=Iv`qe^` z+c1rsD=q5PLV7`W?0x84*(7oaUEtZ{G-Bb3mbgBq84YsBxsshVR3MZfvZM%$crQfV z|Gc@izj*+48haN_%9w7!9W1=sv_@RFaZNKWd$2xy@yLSw6+41H7f^HRV`Qe4__-G+ zRWtV&@7>$ohtxv#KknDb22rZbwl{C;ljK-0F*~Ef5A2*jPo>7ip9z&tC1K?stmS*X zQgHJyd13S55=;Qp3A)}=qSmx8vKKIxAt zi$aZ_Kt$@kbRkZ9ROo9*U2MSxz#(l~?SFc^{!h$s6NHkeK}PQ8n|30iybf8bZH%!z zcYNdGh}x%SiXiJgzV|$Lp#cwCR4)u92i>WUki&5CRY!BMG&Oq%`AMS;szVF;Ma5b& zUFi6R9J#@Xc<}`jF2sWrZ@hW!v?z*DFus`>3x}{|&*8+-82l$t0WDtwq?RKyxa5|9 z+m>q*qiYcIjoRglfQ)h&c$OOGhY$mwhN<0*+7~t+xl5*YfJ>GRVbs;!qtJLEms~?^ zhU!5*Ig2iiI4Ka+F#5rB<25MJ*cT2UC~TPvEjjO}R$L`y!&*RZn%-qtfoM#i+~b=;(H)7!@JI-D}y)C_S{5X#7Jy z>m5_VJeK5+s2EdZ_CY~M)uW}nONci05n-HFX1`x&GE{FL7Mc z9&k85QV$$)`-49{qzd=41qr)ny;nU^$k+nSjLRCaKv4F4)ytywe3A#sh@_4MhyuJW z3LBT)>3W(plYVau*XTn(hWMpoiF#?d_`0Mg)lt?hyw+)GcWNO77}d!+B5uLd9CQoG zWiu{hm_3Dv=ibr=0otUVyw+}<2!SvO28lVj1bcFx=FuAc&yutAJ%+0EkbAr4Ju2;) zA(PykXNOg-+eqxFx_$Iz(nw5`rYEo8EWob*yB>1olp`4LL9k$amHNJ+$kr;=1Xz{I zVoZ$B7wDc%ocECo4qMKu6g+oFOE?W-)Swg z-drQ+Ju^AC)eu96@ImL*%Zzt|VF$$I8Dm>G^XyQk1R7jj;`Mv3fVcxlgGu=7_LQd` zXD4;qV(d=8)?=Hx{ZKNj2}V?Q`?xu4_SbPY?lSAFe~p|U!dhc}-s>u%Zkz%7+kH0a zHt3gM1I9pLZ5lYf1&dyIrKakr?;~FUjaRIPE6Gq>5twpacF&LuQX(%2141Ef=sy?M zM?62246qs=npNAONkn);5lg*H8vK0^D1ve znD}9-iQ1G|!bH7sM~V2^SFgay*aPPKlTmH+ZD+JmzS;^7H2Od?K5Yor@pN-5N<0Z1 zmTN$^6E17LdZjY&u;w*diB7ggifsq!+u89%+)0tm zKip~a569H<=x93s;8Wx!%+)6c>Q+`G{*TKgCHis^r5?kAzj^g%pGhD zPg-fj+zEPFXM9o1U^+sW-3uA6JIjL03A>oAlY1b@fj58qar| zXOe;!>A%=?1s~aXRnrwb_quCtg1g2S^?hT;`^j^9p4hAKHQRTRu_3_eA>U(`mxX^A zz3)C!*RNj(@4HvGe1p0u(}el{)5)Zuf=8xjA}HrUX-rTl-?q6#2rBVG?-<3z>1LqJ`x|>8P*-VLBnKB60eGdTq6NT-e(?Ke@QL ze{gYU8-{A`pNGc<;NP~+J#DTTagpeApZP!c9-k|eom0=LFHWoq^&!nfCwkpMvvARM zm-%(|n&yvu-Tz~#{vOG&zKN)$enty@Ol;;We{Fa2eDK@r)K!z+?e;Hy zNpuFkLA`wa+Ug8Ggs0f6huGmx_+dS2${z#&E_I_TNqCuHGI`fzbY=4oV_ME+qO>{r z&%?xkYxe8!)wAfYP;=wK!0txs$IX>fgO;*=)sQ}^QYg3DJNqwIuN{3+duNZVDmB5O z`T_GY6Y`bbUDI7jrkkXp_^p1bE1O)=tlaVh!%b8D&P6uCL0@g$@mycwWq-FGbDPVl zEpT^Di*1pt2IS781@`K@hN^Cq%BCJNpLP3k8L5tYS(DuY7P7?2WMFHdsc#&rI(L^V zpE{x3UAE(T6nyTo)d+>DN#iF6>WI6!yZftG=a;30sJbU?32qG7t3gVtP8Zn#!r-Bt z&E2hMU=-r$koxO^A5FRRhtv8p%=-KB z#$+2IM^`-zP(Ro?xHKlXl+}dh2>xH7du7n#iiOJK`Z(zc+-+N>YME@5?De%$^BS#> zP~SQ5&p!30xDhs~B0EAeijk8%gzQ&uZyjuHKB-@UI2g^34nOwL+wM(=`eqYR=bCFs zPae+9Gw(d%(Qa8xuBN^h3Lxv7_08ueJMVbvVBa%u=>72@v#3HhvkT#4nDl9(Ixc*% z-H2YTUOBq4+KBwM+|7VHmk-sqLzT7vTb@o`&}hgcnu*XcY_Nakkv&lV1H5CNLz?Iq zkVypD1P9O6oXR0C8S>{_3uA? z4WPU1y3J<6%0%`@SE>2+>T~M9AE>JxVmvd9Z0zO6#+qM?m%-xT1PeOrKXNDw&CaEI z%*&IFZ4+b=ty-K&ch&VS-GOPjbhl=cqJmrR33bEyz5V^g?WRl89p@2uSL;a~zTSw~ z0)n@Z_uVtq_R?G&;BnJU+Bh>R)rM+Yz1CaA?^#|fG@c{1N|`Yn{}c2SbIh>gHV|2( zSwCNQ^zMgE)%bcyp>PD+0#WvF=_wifUx29!gi%?G*C!YLBPK;D7b#DbuG&nB%S0;b zo8eH(h^h`dQVj3}dBZ1Cl&QA`Xasi3%$V3z;69w~!y18ak=1(Rc&D)%rBtr|YxOE` zfo2#(8S1}6y#o8VzC_R*I ztEz8;Sg6u(km&J&P^~}40!XVP2+zWZ>l&rjuWGJf=X5~QKEXj`c?NKie@#VBl+JOf z97n-MO9qj@Nl>RAYt}up_v8U=_Hg}M|HIfLr(GuXy|e1&6(5i6y~BJ`wg24ptN)dH zW%bp)-O1)tdk2eqyPMB#oqu*>QYLZeBcHotwk-SYC!b7z9Kv$%xJ9@?b+$2R^YQJ; zg@dOI)x?h$M&>N=vW+X@aCc~ZFz2o|r(h0Ct)Awyqxm=OP9EAj7~k+PJZ`&nuy^sn z$-!fj7Y^1}#yL3^HruE9t9wQct@3YFhjl5#=;|YXXZ~-naFrUKeYh$uNZH2Q z)K!%h?JnjqF%Au2NNYXcXoFUcDBy%z)i;Dg^QCXcg$NPue%nWbAApAf7b7P{H^Op;c?}L z?2{X7S6w+ew6S#i;<=}Gb|?Fj&F7zgW)tAC`N(Zc%bO3~wzhg=^B(xe>T#pZXZE%y z=WmDg4qVh-NwJ~6W1zl-Ah3QXu)5dWZ%`kSxnjs%*tiAA)F<`9#y_a;NG8v`eX_m1 zv-^(y#kL<)PoDAVqBTEI|E;>2Ag#&{Y;Hj@QB|Twk;i=8Pe>cmy^ZloJ)GlJI0d7Z zL30#rRg)){$Gn2a67i0ujD`x}?yCTw^VPdgARaZhJQ%277^wN+J}#lGF~NTe-6fmg zxmmq_^d;>LC~W|EevD-W;*c;m_CLe93|9`k@pE~A3yzMjCrK2*>zIMpts7rz;Z+v^ zqSB3qwOq;U-vjkkkOm;h6GF@jUT_$(G{9?pM<^Oxkfl^TmI-<+XE=mfh$i+jS5N7^ zs)icWur{D;YBy=sjY(0knZacX!CTZDM*lDY!P-djF{l{8E%ShLq8HK=m2i2QM9L19 z7X^SjVF0(dahnC)mxTW8!rnf(aR}pEP0F1b1G=W}J?+=}g6==kpj(!J&ZwINoiA(> zbTPw8&@lx_hpP)t{X(ok-m*oWM295$mWkYYk`@-Vs}{9irWQuGpr}PGjf3F|4@2t6 zRyx?!RaO)XS0pf8H!xhf!Op^q=DEsz=SD>)Rw;@0isbj z2}EDqBoJeULm<*QsbO7L5V}Y8sTPMT7Kbt=l1N-(5p}{M>UQ;IqXiUExY6L=FvT+m z#bZ1aOGA!4-!UQRF$ol9A?Wja7q?4;u<5IxJ_uoCAqB!?5)c|ylOXg3O@c6HI0WGr zN7>q!t83S<9v!~wzx&>%T_%DdOe-<&@R)^RbcP5NNCNN}3qVOA8mvUP)-S}7 zo0%VkQOJ}ZJT3tr=UyH*KJxC3JFSnzM(rG3+&X~JR|l*}`)Zhw;SPn(0EaXp`^Klt zq42l_kw*U{B7NzTh>RHy5y>VkA}sz@i^by>i=A{v0`z#BnR(EHfgkqX~zAqV%+PR1g%E* zBwBsRlW2_@?uXVtYtg!9(VDFy5~yn|Q17)sJ*n14$1{N1S_csfbDow9gHP}nomm(R z{cMz5Jt4v9g9dtcZrov^w_K-#QlU&O+`6-=NRqZ0rIbb~4n8>zOp^A51f@p%BuahR zlPHZD4pGYFL7QgHC=!MqS#v5qVWHYVr6gESEZlDac}kretz`qUu^Bx0O|u+Fz&0}H zFnf~6?9B2;{~L|jlM>9{V=#MSW7T37o8a%d%l*muCl{ZdTv`lU13KShHuJouk5q);FUIU}aol%j&Z$UjYNra2er1r-` z*^e`r4gs3>Eg^)}=LYI!rrO}Bg=wZz$Fz)Cq~2p4?V371T9G(fyV3V^vWf%j85!Ui6=~?5# z8iMG)hg8n9dZk39XEpk#f$B@225QW72-MJTrpFAa&pT*s%ZXA+rAVzbUY(~WI;4=I zg>VF>8AUb*7PhAHgs`ib!Itx`e(?)IX?O2h^;%>eRXOh(7vT^j_s#!u-qkBRB0+1k zPlMK%Jq_BJ=@7Kty=z-`B=WAUK54YB5cpz=WKuy3U4!SASQe(HHifXHnZcCvu6}t8 zL22=;nQG^(rHFW5#Q`-g)agL9!1F53yLzQbBv6h1X`uSjr-2$X9Rjt(xz+WqZFQ2! zySl2Q@p{ZAT^G7~S9c)-{b|nQEr5lsDPSS&YG$y7*EPwY*sq}>D9uO%Lx0@l;xm)& zovnijG{f!gPYU|hxRQq;SO815pM~?SUhNVIS)+RzvcBYL$i_^Ekc|Rj#*ZCQpLJs4 zwtfju1{2z(k$N17KwI#Nv%$o=1aMh)x#W^5SRt%vW&q{FU%zICpfn@=MMR5uP8J=U zSB-0T2$H!_KmcFXt9sQ+Bt(t&X^8r=ry&|M9YQqaTaLsTCN~@0mM!5iFsW1;seHkb zIG`n0Ca{_m1nW^v%?e>tGpmL8e#WEvWite&cbV&()s18J`ep$whNy>ob#A}rz#5nC z5abE~$?h1OKlO^0NU$2s(_r=GPJ=aOIs_}@Pv&Ese#Y^qZNXB8E@_OeEu6NuGAn9n zv`F8|;8v1E57Mi9|g|MKR`BEr(AcI}MR)(PTZUfD=jpG)Yx5}WIS@g?cGcMR6 z=$$4uy%Hr7n@06CHhsa<*o>L(hfSz^us5o0g;I_(X>gw4pt5kM1@1AmJUW^I+|G5r zT%>CXQwSUKa*LhWwe3bQ-rz~cd)Z*wiA!ZvB!cm)Ko{MFdP1%WLW1l8G zp_7e-r(gIYp_Fq$;y%=+R39pL=K_MA5mXu9xhyI@vNk|N)z<(G)tKoJs(A-aXlFCN z6L>Kw#TrnP%AygwMC5N+QseuPHuo#+V@;CG0@swG5OA2;A@qX`aQ#9Uf>Lo0Dl4yY z$I<1uEQg?cipSAqQ7saXTF;H6k?X5~MsCb>h}^uW_ir8Kwq;0ICu&-)G<262gzcv3 zMG9{MEOt#b3Sn0>!!A?guQw#^pH#*?r;4)Fds7YLb?X1<+pBmp={HnCTF%AtKX-u<8pAX4{%1 ztT#2GSQ@p*q|K>`6$orHN7$wxmkGC7@S55b!nS4xUZyT-*{_ZvC>7Twxf-ZS2J_of znY$Vw_A)ruit3Vp)RdR?B~QaPW;zL5&h5H&Ny~g)5~)oZtH-+hqQ2~Dh{jAOA)53$ZkOsZ zUzfxvl}75Sw3wdxIu5yX)WBt_S-_g=6vC!13&8sIF$AUE0qe967CE`bWjX}OeF4j* zibb_a%SQWjDD2Cg25ii92w2XA#j`Q541vk6SFSzJtEP;Fu&c|$yZVJR1f|`*YuhO3 z?Idi`Eq7v#>v;&0`+}D9uA(BQWutu>w7%?V(8f%Mpp6nutgCUoYg@!5@~*8uX|(c_ zmAZP@cICp}btO^MG&7ho-W6yR6g74Cu3nd63GWI7&nUITc-IwCag)_Q3DiLPBv28i zL!kEauE3Bfn-F@>5Wn8^kMjRM$aWP_2e zrMzKEg^-nQ5X38oT!HY4sL09co`$S1c^a}Y(;;MYe5==`3cMJ!VqL0Bi-b05 zq%LuZo&|BjmsmI7uO49mYU){t2yJG5mB|LL_!Tt-r5WKbBHDu4U?5irQKs&KK&2O- zb6u({qG~1}HIWStWKTmhW;%pu%B%8Ss=$!JDINopN~MwN%vBNxyyP+jFwhU=8G)5KtFDM@n1Iw2Sbe$EV2zm$!J6Yw z=ati?3cMJcVqL0)E@_MwpQa>idL;;Dp=qjC2!o#)G?^~d6~AtVpj6zYib$zkSuoHl z#O5?(T~Wajkeb4#FL)Z8G1L99>2#^Cw6#iDmnx-98l3!7Z^ABBYFWaevPm88aOMk~w?{x>Q%%0wkdU}+ifWI5)D%{I70_6XnGUg< z=RE(mt!7whYmd-ox1?HW%&v7W?3H%OC8@&!&z?g~B?^IvnPK;b>He7&zZQm|^o?c! z#Ky+mb^ydGG5}(U6=T68X26vX-knh z>Xk+>KSSwZI~U?Eb$WD)nhPPZEd1h>+t66}nwk{CvStQfc$1MD5V7J{#SoM}VBmXt zNon%db%rXTEAQ6i&9U_)F>QPUr{#^@OV92DJx@eXJ?y;lo`St;e(mT}p=tMqB zisQJUo-3gA@Mw6|VB@wFe#^J4f<%SL?OOrCM9IfQMM_rr zG;)2>)5wjP4w1`6gmHfOj~vjpMM`<lB#CxdRJ9t>xmh`W2$qpgizGqp^zl!RVjv4jS zDD{O;qcmnZL}?TeCmpW(O9!fLol>q=X|OJ{C50OntS8j!=vW?D`?*@yubO%l!m0vq z_0nM0ub3ey?Sa|0(b3;u7{|vYkB`NbI|R*rG0S;YQMnS3nmEP6-*68lG-hL_lbGdP zt@<+uu5G!J46dz0X}Izempa3>T^i2`v&m`}sit~`u%($HmB~0A^Xq2_O1ov8D(_lL zx1@iKD|ra2`(l;zucGoL5N-;qzW8aZ#!QD;O}f)K6#i=mux$-fGVK)Um4@sw0#a#W z<9pTiAlxNg6%oxMH&N8&toyMCYg_I^+W@fFLUbbW80;P4S|yai&Rth zLfBFP5c&}&fp*3DSikaypj3?19?sb9iDa=F*8mYzb6bi4$6T+?F;U^uF=GHUR(-yNX@JZ%lU3Jlr?_^H3D73JIE!3U1w&ZY%#h3Fhy6Mn8KrD~ z80m8=*2H=&t_C6~79f-D{NX&Ts00c~O-07O=xNl(Ooymtd<)VX=jMmo5-1`+OsJCv zD?fcno*yO_CVq3v>nj%b+jwN27?IbvA?q8#@eI*(s2RZYy~7NF6+Kp?DLKDh;vY`M~5AXZou zF-3SD7nL(v-IMM#kUR-bgz0|p{A`;E23|}h6FRwt(k6+|<44&W6!s2QI@X0E3P&#s z&_vl1k^d=??e!8(>U-&uh~s|k3<1zPOi9Gb#*!_GxEU{7YWi}VVos^^Dxu&Ipst@$EAj)Z3~mouGXYl zY0&cJNz%sEBmsCx%ZAu2c1?{6fe0@SoT95$Ojql1ze5Cc9c$85ca7#zi{> z>C*(Ss8R_eoC2?}0~)+B)BV8fbhRFDiE$|L6*&&G@HM3> zgk{al^)mGj$Ngd%f>L=|10wg*2iLf4hoE|zNhMLO5(qbiTwnY&a$}}L*3c0@M zY2?OChsfnzYkb6Ys+Nu?s+LG~(r_(PwL~sV0F{#fi&0ayLRkLHjyHq<@VIcnF$3>E<77xr*T9qN*hjZ3?Bn@M)CBOou3qBI2}najKSpA#>M?RxL4F zrNLUJYKd8zfNPp_W{mK0Q@TP}Rp6~&lCD;uQBW7r1G7#iq7+5!^=n+jLr7P;T8G(8 z$3>M(R{iu0voX^lW;wqSdse6826)MXYpYNit|!~XleD5R!e+Y;flZd>xYdsH zuB(?z1rw8T0vZKK<(yNuY*f@_=*R5L15`k1kgy|5g zNv9G6*6Fwbh74enX{ShcBxF~Q_Is5adM)QS&z;}e-yiYg)(D()7r88QO^plT1!i`3 zW#ZyhQE8JH13jE^+p@^{SfE#kTW+@zoNT%IhpVCzCm=Nu7YD+paT_xo;x>wlIS=Gc zjsRW^WMv_7t0x+;e5q4sx7#k=!R@k0O_V&%EE?u~tY3RW>?u9`sMmlfMLdhuxC)4% zT7XBkKaTUUqSB{TV*s>|^~FzPHD)@*YEK{AmOjaRtgBZVvdb@(k9EnzVa+1fRKXCI zH8bQg`QcT+4o5~Qn;%9RoQma#fm)%5?TcC_KfEd`fdW!f9@ZBZg6HFMJxMG1DPRQ@)k&YF%v$o`_(W)GF;+`Pb#-!7#Zr0oSA`Sifq@ zTnHr0%&#(Ct*d^a4MC~AS=ByN^vKqw$;HR^EXf#hM-j5)rv^ATJqIq>Q({=aF``QbghCWdUj`R|vD8nfqkAT37wr8G=%ISF4p?;o@MR^D@Au zs9*_5O<~g)I*rYk>3-OBx>{G;f+e)86;UJ&OTJV|+SQ6IiGz;a1Xx6xDiy-cXXY}Q zuGUq*PKKbA^FZ>B*FfUP(UeW?#+5n*wSB``?)WJxPXeK)FzJh(#$?QNh)L$mCFp8h zZOfC;u2xiMG&E1N9#GoCimpH)G0kgcp~K};lJrA#-}$Ov6+_tk!=^Lt?u|Qb-}y15 z?|gZ2>i`NT4tDl-@j*#t&Yp3J4ncTduyPq)QC$+qHw9K-12kA;rbDoDr?2?z`8RD- z>uOt<)Jdx}Wckrb58670)9TddB-MzXwyLrNBi1gL3MHl;9cUC(D2dzASrL~zmIBqn z&{uk2+o7UxO;n|1jeyMZ1}Y$ti!j{}xqsUt7kDvuF<3i#TCF5>*PMsrguUnK)dVe$KErwcR+uzSC8x_4}>Tc>+e=5!06Sb<<6Y^UgO*F;51R{u0? zed*J%jhPN%n{$Oaul~Y;ZCj+2tx+1We3jC@*4ds^Yop^C&epM}!j9HdsSseDSyt&^ zGhp@WWC%(hG+@1R;|>eff1Gz0O)(u|tq7U9C&PNmQl; zq^3e)U-mR$W2TdU<(#bg69=npnUV;rtv+e2@|8+mvD&VTyH3`XvoJLUD}*J@%&&4O zCck8cptMtpDe7}9a;LKv*X|HR3%n zLy6Jqde^qRN#tE!wbFRy$1ipDuI}0dVw1vPVQUIq2)mlu*_F!%`=vGnr5pk>GQ~)D zQ`FMRLN>1VAqegp5Hm$iYoa2jHKThPvcBYL$i_^Ekd0DJjH5Z?WrN!yCp;TWXpBZG zU)&@t(qCRA)+JENvdaueJ8`)bH8J5Y&?vaB$q0WD(H6`G1GzK0uwl-uCq&guR{JDG z1KE=hMVJmDnsOva&#IFR28K*Y6CMMTN+pqc!l`8v2fXAm1TfQF=T$PDNK`U;ap4r* ztNMQW`UYqe0M?_FbFX)qd!E&eWA>hB0lnuzpUmO>DUdoNurmJigs6teYMutGFLxTO zG1DPf8GnL2!}^&vw*yPuuI96}2>4q_1;OtHJTi1-mCq)d~@j z&FqZIG?N330%(ex$(Q4L{&Ho(K&ud&;;jS^in$?ZCqxBHR`oPCeZkY%jG69-O{bY0 zcrn<-n#n0;(%|Hal!VRX)Ur4L+3kQu=jBqE^gWE@^vh$2edLt!hszsBtv`GN=MSAy zO(|4Zrx{o05Tr_1>yY3*Au35aVYE&I)0a67%$VtZ!2FcsG~1FSp&n^W@*puob>6rvE8Gc#xTFBmxbr7#4gVjDkgaC>ZH+2Zz=G;Zs&DGSKB z7Kb1fcYyM{TJFdxDnL46^iBiPmpTo|nCTFZQC5gKjWzF){=S3Cwg5@2B^sA}>Cq{R z%O$lqx=o17u4TOGw^vV?vJ=9NW(H5DOBHAo#IoWp)s>lasRGf$pq4ul3t-F!wW87^ zYXdY?eGSl1jhPOi%K4G_y2a^IJ<*mPVO^?8t$_BhqW@BPgjY8O!7ims2x62ecobc;m2uj684u}|?BZtQ@gDN&Qjtg@Lq6JQs zEg|GGyrOC(kZlUKzVvC>#!QE>jbdZY1G`>)1z`;X`+Xd9Y|e;F zA+umL6)A)@%`8G@x>Qg2H8BLG;x1KE^d)ak?juD<+fxW-H; z;mRCk>$X+pbg7YH9<8CUQ?Sw*w)Oz%hV;E@T+48 zO2xHDuC#2|QdE1yIyh$VFAys{*7k)h=Ve88NkD1}w!Y+P*v3pJVas~i&pN>|UzbE` zlg4V1x+HR4Vk2PLW$KbnChC%=82x(5P{q_G0gZzElejJk6>UM!3gikQ%0$}4MU5{r z#+=BN5eP#5Fp7wNMI2bGdu>u zV`hL1v)~T~_!wXoVKHO|CWLn&{&t;n>UFAW*YjL8-8ZJ@T}rpNsce{-N>~^J9FLKi=TYo!HRqfk>s!zHNRA)fZb748U3UZTIbd^`jj#r%I6dPUTZ(9)S zR0ay${X%v^s==!&)g}S~0hMAylap2{Mj@;7eJlcdz!J^Mk!tX&7s(0q>Pm=J&D$ZX zPr40RB(4~`-=u0V%!Tb-OFA*xTh4N+GhG2m3L1`A4)IR~~bsvW5( zoZIFjdFsVh+01B00n3;y7Mq$GV397>Rh3I)k<`2SbVYB4%l51)i0WBjQg#^O`7lNr zUiIQIfvB!{s_NYiR(;ZKup;pU?^1=iuy|_fQnjq59i#YdN}i3Vmam%8hjN-wXfoD{ z#o%WKO{7b8RprfCB=zo6H7GV8Ef{3=#Aa$B#d2LQ!V-w;!ltTtJ2v%6r(sj-Qe72X zB~zEGZ87cO#IH{B>{7LTHa9vMImKf09Ttt$Gkl~=bycOuSS0oCQstV>O#B5gJ%QO} z(Db4tfv7HEsyep=Q=fDiFr_ZlRY6HIb*UQG(T+)+jO5v+YWQLd9KWw1pTF_gXyfrK zyO08}k9Wq?@o;PW#?jvT(w+N-hXn7bz6cT0gNWqae%exN=%({)3-ZN?q@I{e4VLf_ zaA}ufaMMGh$#i#b^s433z4oQA=F5b#fS|Tb! z_mrOPs>Y96AnKexL26&{iu zv&;?<%&Mz^U{;@a8?#{>(9ae`z*TUOOl_;qwMC$dKltX^x9SwY0%J@gID#27#bR7D z+n4@0hF~vhl6OAo-QtR_MzxnED1>#cjz!CzLReua*&)mf0fDf(3?9JQN4)`1}(?zWKsO; zv>l7oo#Iy^E7|eOYyiQpx&jD(^@+Ffiv=8=0rA&z2o{7T69ocYR|L8^XUUTTfqpzY zY8h?CVpKCLI{heyT9r9t5!Jhw)vDmE+g+W-W05-*x9E0P;V#*6%d7ywt-1yXZuN<` zaf|HtY|pMfBZF7)mrNb4j^#z*dyMnwefv-y-?E_9J%63zk0;QK$zn0FnZ@28!ceS| zX)L08Qe&haaX7GSR_uxFS_+rh)j0;vh_#st7;#?uRifMe#7GKpUKe0$xPNY zML>Jp0y0!c9bW?ao&+eXtT<8`p~YfMGehc!Fr=!S8jGmjd?ur&vc_7S+hdVA6{~2h zRTxcntTGEgu&U}G!Kyy-HdeJgt~2_T9ujNA zvD^wn(0HpbqwMHq7J#5v)jxt>ed12^Vqr;*w}Ktz6>r7LA+W`1Qn5+0__P_4_z@VE zA1^uRQZvIV7H>gF54u!*OjX(yeIbO#Tab7k(59A@yPK?tyiCFgz zPmZ@RDb@+#6>sImwF7tQPT%WHwM1F|fCZk;8DgOvjkimTK4S?8Gs7=dLk1x|@EcG= zHZ(qI?XFn(!?CNgh%CzWbrmtucFB)E#n&IfuBv?myZXf2*hPY?q0dgPA;WwaY)uWZ z$E*#pR>tGf(t>x<*Jd!mPvoGTdpJ@VOUB2TIy}H7SZ2JW@^mbs24uem#pkRcgS`78 z&ie7BK-i{fz)Co6=^w$WKJhkIT|vhHShTHm7!x1)w?wp4D|vf?Eqo#A7VHWiuJaqE&FwF%VfM5d~F1eyB8+sM>n zC$dj%o<04f3{OFAGIh-w7Y+d^PHOV(nl*mZ0#WyD(c>CMX0e#l%y5cz%`T~Q8jGmj zU9-n$(=`iHdzO+g595Jn^pJJQkJJRhIx(xR0)kn6;%&@^MN_$J7Ush?rKYY~=h`CB z#jkMk?3#57V1Y5F5j+ECnlYK0rd%zSuKSJwD zek>=@u@l1TG9VDvC!PjjxodVw5S&b1v+l)3FpS^yarmLNNb8OOz9UNZwR%9>DF zHWpF6SrY~=XW!83bRUbHDIzDx^LGBc#n(7Wv5I(2yHs$SKGb~z471(jxS_3<74HIwUHe|coJOe9t&BCNu{%q=+bu6JBzRUQNb)FBpIzDg4sGHb~ zqxhc!<7rvt+*qQRH{(gxa|d0sAgw2UyUd%H{TNRmtSfJBX&=F_K5-|0(Vd^nc!K#b z=$aT$)-^>y>&&7CgHwwH-joM;}Mojcj2$brH+)kscAM1%*0FouE`bUtfPrQv>H2$ldEsV&FXgt#H7xoiYk6Km& z>cH{K$T1f4+8>#+O3$%~8XrcbCdWMJ)mcIonN!h=odzxY(W91`1t9TO)jxt>ed2BO zwnHi!@6>oJ=uuwrR;(%lTbw2p8&bum%^?+^G@2hX%8bRJW`i>UGOR_S}p zgIJwCWRW@*v1q*YqfIR{8$b}Nt^k5qed2AzIs=R&-U`~3SG<)M7XdC#pNfsQ^2aUE zb*cl8w~RhxF|V277wei`RtY>7Q3GnohVG|)U9%wVe%Rgm(WgLIXRfSjAHl9Z@iult zqU{l))Q|;z%3MRXtSJInoHpgzQG49hu4{oSvcErqBbBjaEapEmq@p!sm8WA7H6SxK ztZLR8vQ8JW$lPh1_2Wq`GYdfCtg3$mtNO&9San63Bg)N-%YrOrPLHiii$E4z+B9BQt)C#k6LIT&!z$S!L*0MD^~PJ>F&4EXcYaBCH=f3WRk=SXK83YW0b? zQH#VEN7wAKU`Ls{W-Tj^<=Jj)k;#l=G`gW)0fQNI?c+ z_W?*({HRcT`R$>#rF=U`4HHiTsoXUS^I=p9Q`f9*N$n_I!Rb$)U9-0Dnvob^ZdzfK z7fXDaSt!N2WS;zxGkD~}*l)jNVved1{dg(4d7 znuYl=gqpf$4XcTO6MqlSvuoDy$?Ujf^p@+5ORQ^lMJ3o+M8ycj-d(dTl^eFK=y5AZ z>WR!Qx1}q7)Fu$tmEg8Ck04W@cpI6>Ud-pMS(p@qr>Se!xNrzSBd&33q(ts_S@7(( zH42NTx%-9AgmulXsC1eLoO*ZEo(Srw#WAaMdm<$Fz%$x4yW&S`0u?(ktF8iqS$*Pd z%!X}B<*r$n4}+_zYu34b2y|B#JUe5ZABl*j(Ny9UMsBef*UU0ttZNp8^vHm{vr6yw zT6Cq{Z_R?(GYY#HPh!Q7<-`pENr=^DKp?D7yba-ylbZuX?wW=9FbJEvX5DLsU>IjP zd3MdZC5Qkx9)=mm<$8k{V@<56EE|ic-mD3Oma}hYb-Is5YAnf54#9prCs4051gk56 z;8&k`8h+)j*%iTaG7%=wbw!|i9IXOj@zJ+y7M08lX&kkT=VCFcnT1}gYj#Ct+E_&O z?u@nS(6FVA?&JZPXB2law$q9q+X;kq;#OS)1h@Lc+qgxxyFPEt!lYR6HFeE8mKTBV z3EwAX9iK4c(A^n}IS3rZjPYWD#>`NRF`iab&W%M>?*!VL8YM55(q4;ieRblGMPiTG zih2<=Z}ww6foh%jRke@cSD$zrzhQZ^%y?SC8Bf-wLO^@c_niRi;1;-|8+@eFg(H>m zT`a~lGo)gSrxlfPV-eMx@nlqd)>x|(e=IVmVik?GevBtjt`n=O{t>L|6K`WRBG$@X zvnx2`30YbMvRN2U$fqrUcB%qLF5|pdOlyB2S7qH;M2$zT)W_%>sPQPPlYcBSce=av zV?C`f3qYzJx7x7mYvd5Uh0ZtO(YViKeI_N31FW+AMG1A)mH@7SjTX*JBR$ z)6DRS)uTX2kD_u^JxXh=RBow4Gg9n^3Xpgo>}!wtQJ}aD*b&=W0Xt$16K^Bd8DQ`( z9+(u{*Sd;Q+Tz-Qd+g4%jb}_{N)yivZXCb+g%*Wn%EwfCjwN=*$6Kk%F-yD!d1r)P zY*%~Cj~)fWx-#XK{t@)*6K|up9a7O+t`cuyJ`A{C@m8!V0^23@`&g;S_r)wN7z?># z@X9DN77*-@cg*wlf>SeOY;Swak2ZCT*#Hu6)fGSxt53X**wlC{ zXj5MCR$g2LxcHq=o?WvRY&KPsE^etf#^^H^^O_lcu^KW6>5(r7l!gsmQ~A1PLE8Ng zZT;v|eEkvZs@g}et53X*-H>RjcFi6W^eI!<>V}2WX1-4=d2+g)9FGMnY*kZAM@i$fw0cZSk*s*Rej=Z zth(Z?ek6rLVimllC6&XV%%nl<{0q!^3$yB}lx7>jAm?2dA*YZip`h_T*XvnRUj zngv-sBdl*S9@Fk~%#R(#S09P6s_qfg>Jx9H7700yu34B51FWfQ*0Q1qSn(U6JR4&z zpEYk)(akd+T^Sd~g3+1f!C2SqF_n#D5!JhE)}Y#qU9%vpCrDGbuxQcAj|v6CIzy|f zd<3NW#M3}3cg-FXEGSdgtZhjVl;YPtd3Md(zMCDUjPhc!{h7fO>zX~Ll5Q-bdUwro z-R344gPfiS?KF(~k)1$TCqh-dBM8+eo`z7lYxbBRJDIv>4XcTO6TjHWvuoDy$?Ujf z^cIV;&kUDX*X%KsU}F*0+t$5XXj}TPcN!=`0wncBW~bXyKWY;Q>qMric?6mI#M{V3 zwxvE@v&RIr$<#G#Tq^{irz}WvONnp^eY<9j!s2NzBePhHe`Ywvx@M25bQ+7O-kq^0 zgF0j5DH1xj$0B(uX3>1tkJJPzc4Agt1q8GD#M_t+Df=EEa@XuJL25E}%{tc>fo{U3 zPM+^NButyQ9{se8UMXy4?ktXM|gf?R4Ca?F7QQQs>qhAh^{h z-o`Dmk@V@Bg-NmCYwDVHEFl8lxo~??|5vB(F>!F`!SwCSSNl}?IZZrC*HnuYl==$aT$)}=x~8_syL ze#MMDG$l5$mvN*rzKg~DXO<;njHlx&Key%3ZU^amEv}dS`r z*7{MPjx!rT>QU+nAc)l`-bSo5z~Ei8#|00{Rg}^e7XdC#hKeXkX&<*h*QpL1zlNv9j1hMK0Ac)l`-bSo5z&PTqpiOzjTX}I2;No{e#l~Ct;}+;T)q%%bMxU{m z*UaKA)-`)vCGc284JZv88n*Iv&4RR^HRM$6qPtr^`VA|$X_J{#lvth+U zYset4Csr|;J#dWfYft#`q__px<7`X+cB~pE-o~mcuowU<*N|aS?9j)Y9$S~s4%rj# zyP!GtzP7?iY>})-`)VW$0L9toQ!VyS?_LxxE&xMuDuJsP#Z5njK>i z_Jkih3WRk<*p}`Q)any&qZSD{j;>jl4+E^JYu2(*2w0aEJUe48Lz|%%dN9kuk7JZ^ zVJwC}v*?O-&7SaLM0sP>yS>)1KJJHmK-m2NQa>sb2lh zuug=kdPfkdPdp8wa@Xt$L3T2A%^Fq{0q4qsXJ@Qo*a++ycuk(+lhIo&#y+#~iFM7M zPzg2`QN5dCTPinfJ<-$>yQCN-^+aZ;+fqMj6A0@>rmA@anfk=r$V9fKK3%gX1hvW3 zHEUc{1fa)!J6w$eTOjJ5qFy`l#3>`QSd4#WIK{eVPpEVni>TgRv!`a$H49REV%9rB zH}76i>hK9aQWFU4#H_js2xj$(w=o;GDV4isVLohAYU-MGt}O!H;}-Ox!Xm))eY(~u zGoAr6a*G8PGea=eHG4v(*H}dL?wUn4ma%IV#NH1O_G3ALuucf8%YZ;wpLiOC<*wNi zg5YH8nsqNOf?=HHarmLS4Y;2>rk7Yj0G24ajgaYAL;SVZ+^O&GMCeM76$eJoO^ zrp;(w*pKG~>UH8*T>%8Y`oz=lD|gMF5IiRnVFF!O1iCoW$&)aFempyB8PCOHR5L>@ z)-`)VW!hLo_3n(dDmd$QS10{g4eI;v54x;c#`$3G{gEM z2Q63W#2<^qsrW_Xt{>wGRO`gAs(l2%`o!D#MR${J@-dMaPbYB3lXXoI(60JE4`3bK z0!eh1$etNR>dAi!j3*G%gYo3eYBDN5XRHNzXSBz~V(m#k#uK*yJL9RPe>+wU6K`YH z8CY~?vCMdaNwHXKVmu*BYlrMf{9Y(eJ`?h33!t5&*1FP1PcvlxqU6;FCGp0J%_ zJgP`6b?6A0^yvMv21$kiv_MlKqEv?wcg&BAD{4#Qk#k>?ybjXn+XuMVFITlglW2)5TsLU6^7LHz>C1jD=1EDAt1l!e~^rJ_e zWEOyA%Bubm^y(9DqZf-bI&hVEE9g;PNwQc~1hy-_PsVC!6~jlw)M&cYNk*Bm7?fha zUE&puw<>AJM^prA3-R$*Zi2)StFwnvr8YwwkGFobDQ*J@V$~Hu5UWqTjaYZg(b8lk z-U`~3SG<+i3;`~FCsf2CP!;Uuk6U2vR0ke!8GXiLUL8VK-7jJ_m@TR-{~2`6hNGIhqO zzOgl&Oa{lv%3x`0cjNS#M@M_p#nH~j?&fH3acg*Ow6|{kum!G2zJb43d6KbYEatyI zDp4v=$0BM#W^6!e)*7--7qZB#Br;fb5EEzpcv9Q~kT|RAAHk|V@itanfyL1^ds2|4 z%;~XpX%WbtaNcm7rEAs*YHTBn$5_UXv6$A(Vl37*ds1cSSVZ-1hCO9&hE-xdx|76C zX+hTg5Mlk;Q6Q`{!m7GQP^(Y8janqWIJ#y}3U-vKYu2)&2v~7mlxIh+<+I_@^#G1h z#)YvM{>-8))-`)lW#d>x#e$P}qpU&gIz1WR!s?75i@@fH>F+AgqLUvL3N-5ksj7Sg zr253uKq`05o)j!7Q`f9**$|ZC^e4}*S=)EB!<128ET*JnX&uH@f_2TFR7p1$QN6on zxo&flj6u%*5JvsTP9Urkp{m{ygz6JdL#W&}ds2{{OkJ~v)kMI#iWO#kJ8BJ|%#F)a z4uaFnaEW!zf{-2rCq+~&mU_3>w)9``Bhc6hlFkU37`5q?AGL|Ey?tBS(!3p+hKaY4 zsg;nBeQMoCOFk(bx5A{@ajU6o*0@%7fSy9Haca7F^&?PmHhUMZGA)tC!I#*pOuJX7 zSFTPM>lhHd6+OksES8`&vosg$ngt;}aO&MPyE>b$S&({0n8otlQ+}i-ZU+cv)m1<+ zt54jCS!|c8yfq8+VQ@8d%{tc>f$ow(q37GN&aXv;RYQPBH>#%?xy6Fu4wGJGM@F!& zSrF0#!QNf7sNVH$X>?r)#NH1O_G3A5LqH&`E&~E#ed28hV~Evx5^~qo!{QjJ(|y zY$sFKtYdi*_#Shf#2&za1|&L0uegh>4o=!Oz77G+@W)Zi7%vuR%nZdCfdj;!fkPAL9vB>%_0BeFVSy#M}4{%bVq{*;6><$+~U`XtOY$ ztRFUm5uuAbVU@{|Hv~iMO$e zRFZ8sRLEVkr*Os-va|?fPxv+jAj4aL>7Fi9xyEBG*6`3vTiJ*#v@m1 za`Zh@!lSHC{;|m1X_WP2J%MtaQC8JIf?R##ZRCbUS-FenlwduXXo?!D#Hu2o#hFi@ zU9-riEr7jw$ z>`(}ew<Jx9H7mGCdX^;|c z1wG0u-ilR4V2jhFV&kp&bnbY&x?kv0SiA)xJ$AP7@mA?=%z{{ucpub}SN&*H+y?B3 zZLNSEv4)Aa5!;TY9uY^5w=gN+Lm;hj)ehWM{7$IYc&k`ePMgA>yKbp)jon~ zYHDVT?Q2*4cv9Q~5Ui^DN3g0-yp2`YzSaO(xrPjrVt_TL$JP}?qAbpm@=lMfpSFP3 zsR}&CGJcF@(>1f26xKDn>cx@r?wVcgvTGJ(oe^rW{p_kAI|_t#)sI`cM^LLzyp38U z!aBNUVLlA7rmk7biXvcLv7iYR5jm*weY(~nDUMObg|T3?!+@$(XbILeyQ;ErETVdM z#v1gLk%A1u?gxsNP+(TzwWIarB`M$mxktEJyc9jM3NeR{h9MAgmLi zs@@TV>Jv{xC?xmf?#X0)d1rkI=EK6MscY7-ng}>?nv>@{w1!Wbp@(_`9GClr;DmL} zuBrqZ3ogBzVOwg=No@j2J(1~wKGZ*;+fqMj6A0@>rmA@anfk=r$P6j}8oFjz1+~f4 zHEUc{1fVB;-PGb@Ei^&aDb$sx`wW9}b9Lx`T2aC1at9$LP?Cx&CwRYL#rOw{5$d-yq zbld4iTLL9IQK{+~L8U(NHY(9QT^q9GV{*I{v?a^xA@DnCK@chqw#ED+PUe3B`MkmN z1ess6DI8~vf?_d{nc?i67|v7@j78KhVL1DOwHI=n9no=?HU8S za^pR_JCo`9QezVfw&IN9hKs@6!u${a()&-!^_GG>Om2U0ck9~D?l#1qv!jjSwZ(ge zlj-79Y;b*$IryJ4h)n?UdUC`x`1%q){L?S-8-7f_&^Y|^;zOh1=BtK#`{;0MeKH+gkw0~lbtZ2r%z9d19?VmHm41>e5Hf#ZZ{y;icjrwDw~&Ju z_pa`245y<(?~Xa!x9Pt*JDJjdLzl~}(G{W5XEBFG)(#%*59Uty{%l*GWZ(87@|0zc zE6p6OpO&W8kDAsa^g?*^}hCmhlHSHXkaDI9OjU z?UM0kQ|ouHkY}~OyE%Sxa3k2`*4gosrA;b}Fe~mazE1z*^h`v+%cTYoME8R=g; ze|2lnhu?JhVbc0vf?Rw!+xN&O`J9#p2mQM@H}%ozcZ~+`Gnv41RV* z4%9<8F30>K$=}4d(LQ+vi=oTRyVzo=?oB^SWS`R~uWCeQk)WACj31Qd)hDlR#%12O zLGQ9{@#Q!n57ZWf+(T}h?p+-zj;I!Txvc-Mho6BfU$(DqiKwHe30kgU6P(;d`Y=`v zhI-VJN$58DFDLW&jo@zg>i0(eCFbhv%DRyhG@|y&#YWA?8jGvOk@QE}`m;?#);}27 z4S-@yJ#%xuY0hcVhp}oWSdZFF^O#kXW_AUvHB~62ij}9KZ0HMcQ2^sHfj-FbdUW569f!fxXTp|Bld~j=bGFd!78t$yqzXQKq z>7}lsvf16vbaV6CA<#`#F*;ooih|H$={pK?)b&Z4+f?5!Tk0r1r4_0;Ob%SEhN{$n z>~zUvKCMHgIUh_V`qg-zJ4;3yfB=e3sR{?X+c5AegF z55FOwYJ|p=%f%o&niX8Qx?O+jg6yb1k3sI)Yl8-Iy%!!l3LEHtYZKi^KHMYUR*=ta z%}sm|s>aN|2kd({t=(Ad{fgegoxJ4m&&d}IURaeJtYcB(Wec#oA80setDB3&J$_mF zCHVWrz0u~?jp=xI2Ws7$E>OZrIwKv?kJX)MghkH}IY0gEgtujH|Lc1l~9Boa9 zcFBsaB>NLRd~ncIIvd>Z5LI32qgmV%T|-nVpgHTs@#a%idgGn?}wLpHl0J zGc-Loh_ISfBNx;oYXM8YH7(r}O;NW!-|CTH&~rd;L6>MKr4@!csM1=e)!J%{aVe&& zSpbvG`srpXBla}y#6V6kJr|74qgYV)H_zAhw1){feGrcIkJ6)kdfaNwn52z|Yc5&_ ztSuR~^T37dlQ(KPA+0`8C@!976&`dJ<>yox<9)lAFYI2LZV#oJwCmf~w47Z2HY|grSl|dhVOPJsPO^&sETKglglzN`cvYh9! zt-)aJ)@o}&o?-uMl>lsM)i3H1-5>$2Q*CQtSC^uUnhDppV_^#DTfgb?(df$h=6D;n z&UX}YnnS+JJq?hyu64CzykWjt9?<@8VAUN0_xa2SL@w7ic`dN*P+K~7jRXHV|E^0w z0@>V)TGqEV9!}tFp|FTP?DWvj~E4!1?`mj3K1;UN^t4HZy5!j2iQ+!=)@L3^(4=U%ixwm@-X0S-D zLAn^+ybh9*z!`#1z zK({IaDJXFXMxI2qQ^i5g`VcJ;sf8FU75i!pow6FlcSWVZQ7E?;56JHp#JtzKmPIc3hp4 z)sSrXARV2owlW)fC(bZ-~y$FVN7_9Jd+xoEL{XkD*cuvvJx5 zp9c9A54k`+R4y+qRdXm$Mxp%llC|p5O3TdVL*PN7FLHgku@0CJKW=g^Or?M_)QhyY zX}07Twvk^dq}J)7bm41{Kz}+%^Q439r_yF9PGK88a7%F74EabQZzFU${j$2isoF9; zl%~x5Bl#_dYW2j4r%i!!mY=69q~>CFkp&jyzZB$7%xLtPfyJ|w@L95~7vMI_(Hlo) z;Dr6ytRe}d`K&1~ys@wOEc(H zIVZ7aj{OC@@}F*$Pd?Hot7e^=q^JeEK;=3|z1*!AT7P4+QPW`bII?_!Z0O2=MNgPE z9xp%N@gJ-&x&M>pGQe&i!reM~Y;U~t_&OA8C+pk8$>UIzo};C+M@N^(JM3?6Sa=q> zd9<}PzA};jGx*D$(HmO-7U;mEolWsCzd$pb0lD?`Eo`S*)RqGa{c_3|Q5XV%!WXwz zY_zkny9rG}JaVoWtj2CAy#@18_S~%ogwicS+iq>&*N_-*YGu%zxyw>9hk3+$_euV+ zATPIQcWZc!c5p2jZvPm<0n82*IO>ggt$e$go zvpibinpsX;w?H!B^17!+;1oS&)u2vg0L*h)=(=U?a1*+U(>_p-I@r)VsJ)G|vR$4sCrSvkNRSLJI)*ny>{Xj31{}Oe8Pz!oEhy5Bo0Td0qiUS^g z9!{yUEIFqP=2>^Db+PV96JwKPyuU}z=TJA#_e7;6mwNI#R}~6Q1#w-bM$m1j*DV`D zdxi)~x32^k!zl~d7EgruT*>;!dn5(el7L2N+k$*bWR|{JpAF>IFte?=69jlMH^pU0 zb34z@8B(U#j*PGwj@v!3dYp$3{;6y#;Mq<-m6lyvyJr|fBK=YFfi%}I8IYC?Wzr%& zM97I>2Do;y-F6@Y2i7$y>R+a1jgqPWYiygkf^IwSqg(v@tBw5bcGa4+4Q)4NUWmFe z##KgsIhziu6Owv!k!f`+t(LD(NMxiXGKLq(9Q_>=ica$39(f>-q=uv`lR~Q{N-{P{ zp{;#w8-30kjQq!*-1G)yp!)~&^!H@vAWlIcT;*OAVK9&I)|iXrp}~W#))-GW@TjG- zSLgdRS=p^hV5OxxXKK4@pQOEJOH)(dB2|yFzf$d22q!Yw^y*n+d5U7863pmYY?*7m zbxOvGtkoF03xh(R{B+v3nQQVq%UN%svemu2l}lLigj00|Z|k_#4WEyuRoA%qJl{;& zB=gZg8D&b@jSpnBPc)jbl`XJKEsv7#OHUpu8`Lt*Ec&B)WeND1+c8{$m*nn=E>1 ze4LeI#df7!akSMSWpRg$U9Zq(;_7}{!biR}EvZw56G>e>u{bfA;{qAIKo`jz|6*N) zOffQbC~;H}SAzVRv}{NS!_FF0wqPbqCuK!kJmW(9`Ex8JBxgljz^qvjA2l^A5(8DY zJ*eK@l?(~z_S0!aULl3ZC$gJaN~xC%Xf!jq?quNvqn1eCo0d>;=?t}|JTZqZOVS3F zi|D8`S=2|~U6A`|&QxzAJrm~brYQR)z5R2?$c?mDMMAiyJp#}3g}Z!fQR+I{VL?7PPcCM%L*??S z_FY!Fb_$gXD|4ks^roy_lEoI8%t@|#4_D_V0%@K5+OAI*I(+m)Wdh@KBW4@uhHilcny{5!iKJZBN`0A+%yQ#|L(&o^vmWd zuk33AHe6;O6c#$f*8ZRaTjK=UZdRk5O{DVBq91J#I8E8ZkhZhQZby`1l70QE`jv-b zc{`6nOd-GD0l+KLWZm$TeY%zNC@<%Zea6HJr??(Pi73K_{#wjkA8eCg3Du3Q%5>N4 znyyx{yKK*FkIub1`TZh)Bigc3w#y7@SWd>`3BhWF6`k@6E6XoRIUWBY0b)PdgjENE_I2mCizs6Krd@yL{v#^q?%5rsT};+>of)VUnxW* zG3T5_n=TcTcfJb2#`KI&ms63;ANE)B{(V17h*XF*@zn{V!DMqG31ySEU^cMKACb>DQv=v6w~Q! zmAuHZpCW&gLVDL`D<0H~r>%%V6<8%uyTpimQG_?@bP2Fn2Mr$a$HZbv?0{Z+{4XEA zbA%I(^m5%a{v&y8T`LKPr!0c(LMxZejfjl8ksUhB*xPpwx!Mm2T!Wg!!%H-m90tYo zloLJJJB-eQG5RuA4$LXk7K_7%*FBoPmBVY~U@@3mn4g<#QC`Tm=P`7R){@H@vSc9j zTgJR#>|Uks>XTbV@$PQ8(E|z>^%YtvO`O%)LtR{@=ZsyQKqb(mMX+Eb1E;ct%Xl3x zu0BK_9NZsQTvh#@xNwYqhEkRLyD|EoVwniKI6ODT+4!lCOzVpb$tD9pf*-c%Oe; zxNLNl5XQ4X8|f-d_hpQ!$(mwlj=SLEcM5VB9nGUPT+e6bfOn0pHd z3h(H}Pma?sQq<#JX3=zOv_QY(XSeGh&m;g;%aEfLM9C2`rX#GM*?v3ubvkYu4e+1BbAn8@;$`P?0>z8`C0{IChttvGW;G0CtLrQAX^dtGh+>Y^Pqit}Y1>{>B!A=qc-wSpv*!tpw{ckX zT83$JS-=8OM)khQ5%MPmxyOZU=QLS7OP=VRW)v6%Xncw@t{vgZAYUIpRp-MbN3=%* z50LrOJ*dIxTM&AUv(uXzrc99{3j$hiUKvBkT8BgKm-ZpO*2HJC zzoC~3?k1#(WHC`ubfXk&Rx;GMCsn&%ZTg~&jl|DZ)?60}{3F9Qb(4!vMNe5ql9;kP zlEM*E;&i3WQ{qUq8*nnbBhUdGDW@v@3Jv6)9!E&Hgbh;Uja8!J>g`pgnBrCxuSTgo z7Ll1s<;?5cv*KW|^gvHRRMMO^d5O(b+JcBoC9*8s&SJdU^9;5l31;;DIWraMTQwsl z#;V$bq~1sj*Qdus6Il#)7qRoi&!GRxH`f-*u^Sb{C4r0!>F0v1p(`?oOHySC;?j7r zL0lf9YFevxGW^M3Wk_!0B z2*fz!d`jt|so`JWfKL6kEQko`ISzdxy2l1|NiZX*a|U$jTk!$CJxFRmPbfWz)1i8c zl?Rx%^H?*g+|HE4j6O{sle@?oXSqw_IOh{e4~^E1*6JC6EFqjkW-XF}vScmNc(GZF zJVceXm^Vt+{0**zj|rF3XU|iQ7gKszg3?1CByr1kgM=g$ADCYkX<81|x$iA7I0Q-m zMtsPpxXz%9f+;pdlO*JcKw63>jToPz$%8a0nguaVwUMujFJ@OKnwc_{x5q1CXFRdk-DBh6I9>xXlm~a;;C8-jss;HD4@LntRd4|HEhwo9bKh)@iD9%+ zTdYOhtWXWS5FQ&=DpS-llOz?V^KmyB2{7?8GG(bj@3p*ib z_&|KVUV9&rk)*3Fs>(>iY9)#h5gD_Qx5ghD%IE;=;0&`QmDQO@5@4$|Af;6Fwm7ta z%dXc<$wP95|HzXWRVa~_DAg0|F^q{r5}Ai^L5voiGY{cks%s@NwxkY`p;8hPGq@1o z4C3n~+>?kb#NUO32#eQ3JTF*j4E*_2;r_zKYwYyU)K&KMOXL$o`$fvYlTOk(zcDjV zSy91VoxTX}wsH$F1d}EQ-y8Jdx6+^6z{uc9-}SkjQliblJ5))=L`}b`q$G(ADr4*=lQ&iu3nh8vm$_I_>@JRph- zMcy2DzQsR(Zf^wC=sn!~s+3N<_x+2w5=vW`!Htbn5soUR659TLSnPY3$7`UNm)PAI zP1d)@kB{;xn#M#iiG$yZwF$*EHE(Zn?AzTL}%KKxc@(0+*A z@Ar5h09)@jj`nsNPco@$^25pc6goV1*#iN{C5%0+M_MDp`8=WGDXG(`_{JW%2i|!n zm7h$FS8={E;5XW0Yc1S&2YHU$x^qJFAMVtXz$N8p(>)9%+gW`~QW8vYo5>DBup7yW z;Qp2FS(dZz_AOoNX5(_(3RedChYZ`Zm3pel8(HwEx6k|U6CZFzUm;5X@m;A&m zJ@pZyS0vjBOIl*stQKGXg^T=-o>!2U4es^2qsQQs)XCD}tGjvJNF^VP+we4%uJ!u( zWjvhRy}5ag@;(nq-Q;EaiX5r+^*Q9w=5RV(zqGrzJ%qG-Fn|BcUVYEP^R%Pvo5+o) zZ=jvQRWoeNP6+7>*90t&d9zhR+Px2iwtN4z$G!XpJJ6>}(~bIX%QNPw4X2RyPRF}D zi@R5+SFTPM?;Bnl?cERSvM4Tpsu;G0^+fg;2dYKWCtj$3_2{5?#~cR)`C8oztTbMr zy;w_8L@r%L{NoEVxiPEfMCAeUTON;Mc&!_C3~eH5TaPQ{mnT>Te8D{-u43{_9xoF{ z%#4+=^|@ItanZg37S5$~Pl&uH*9t+ozxJZaf#`GDU4gf(9X{9}%ny18i_^WQ56I^_ zePoAPcs&8BDZvDNRX|X*yQ>|p3G)@SN8gRuaBZu?1T|)Aw!>5Bt59Lzu8&il4KNXd zrXH}uy+KZDT%cHt52d4S7AxwPVPh%90rJxx&t-)vMLRIbuXOi5CQRhYW`*-pI?QnO zeYeNcc!9DAR*`O1qZ#-K*wGz5DT4FRe@KX+#f}S@^!i`OCS&kR+^@z-8AlWx;ggBh=^;@J8i( zbLGL3A_2v0TmEIi1qS&+qD-DocLO5lb*6mek2=thbh?q<9xCm-AJ)wR<%~J=$30%x z4V9U@eQ`#TUETG$NuGEmVW<_N+ZV#R&Gt|9dg`&Jo|;bwqQVHb+d1Fdjb1)?uoZxQ zj32B#LS8p`t=GYdw_x&o=_l!7OOJeiu96^K|LS0+T78zu+j}Hvg6>~$`<|Q?SRtm! zDwBAp+H#*NAo6q$;%iP<=n`7l&!8yR?;r8|r z9!(wwRpx&)SDlkjif7yTl~v{`lyXsxcAg{C{<)m#Pi4Vco}F2{G|SGHx(m>W7^4$pyb^JSOgiLG{xz$wN!xu|BpttT#B!1W>oH7j998j)2|0^Shg6CoBK|m-T_i)4nP@ za+;$V=0R>}9^~e=L!7Z&6tgr@#fxC}`b|CZy6pPc8}3{lEl#Gxy(!a)?p5cmn?7t_ zwL@KOj<%*lyDqMuZP;}2rg`G*;|-25m0b*Oc!-)u*{RuebVSdJBZi zUmR~fHR#>hC*Rm34`=re_l~!=7M~n$T}6_v{RfGk=#!LsoH>c1s%ZM=JBv$Ymn(>M zo`~#BZ?d0VN=-g7Paer`!)LGU47bM{7pCLs#$$tj<-ewNTs(%Wq$Ky8to3X9;6RQ>n;M}Xga>KHJZ@< z0$%{~-5!^61#cjgORp?j*9FrJduPtd73PA#c|raqneJU3QEvxKis2S@b}1k-d(_I9 z@@JQRm+l=9tNW3HR7Kdt0`!F^t`4D8e0F^4(r9ng?ZLs16=awpJHtd4e1TjH-EZoX z-R^`op2gXi1YNIp+5PIIF$|vWk$=(MT1+;ETSIo?VELN-xozW}jP7ynjN%?`7g>>C z?UC1arvhN()z_PKuI3KnF;g!;cy%{q{9dQpgdQh)TiT;#z_lmkOQVneE=+f?oNJCG zyENYb{X$aLq0@rAhcRxu#YEZR@L(i7sJrt$^4U%;76=jb{eE4qS~=^O>dF0ey9{6x zb{HnLU0{cnNBLPDu5(p^6sE1RVBC8C$4*^G#r&C(sXx(Oyf&tfIf(8@$)Bd(1jRN_ zmSnxLbD8wsr72RH-Yl|P9vW3^cNRZBU38h%@r!Cw`kQ-QDpp$ypA9u0RY z<56EBux|jv=x%?X)|bchCQ0d_x{Wy1?P2no!K-so-6R40X>Q-pCl6*_B2IC8AUe;b z6EV=+QaaYmi^NR1zc8aEu~PyH@|N5yp;Fqiud@3c9*^3iWU~s@a38G4n|fIYpwbL$ z-Y}ym*(hu?nz}cA4IV4aQO6Z}Fi_WQb!Pn_c5YTsmGwh9(t(q*c0FvtvKHi1{mj8! zp{jLeVjZ}tETyeB9%_2pj27G;hg51B)MHaFS<$1eA1!T0T96@bt;hagKk3#6vkDqI zS`M1Iu%m_|opgVbS=^dPD6%D;YBr-4v)gS!K0cp?8YJ|o@Xgxvv%LG7$`zJ*Q#49D z%PVC~Im?@dQ->f7Yer_|MKyaS&*{3bCoh`Wj5Ofo`rezT(MHpNpqJ&Ke~IB0z-FOA ziLv?O%_{%2ySg673xOuJ90Uk4O#_{*3m=j<@G%-$Q=HoeyZIjZwcNW$?cg9KF%$GL zdc@Y;WX$MJ9fH#Tj!gA#P4WwAR}biEjYJ&FYm7{b-Qd!FfVhAV#@?B`9;=YWJV;^I zEE`@(%)1O4Q7muTQjx^+ddAk9iS7fmkM*@DypA`QP3CE%ua4jA!aMLrx;$QCpy|X0 z7B2C6`|HSS2Va$1Z&#hCxSf)QWxt2OB$nULc|1L()loP1cCWzmQH$yV)x@6mRe;CK z70*IE8}**yWa{B(*ijhy`GUMW_y~c0Oq~K_d+3u`4x9vx%hII6JsASG8=Xs<1O3O{ z98Y~qs(OWJf;<_2+0xI)l!WBfuEa5(Yujj%xmKo3GgDcQX*9>bE!>ZcqoA)u0iRq$ zK21Nuft4o$K^8&C`*Dx=)!kF)G$!W@h3>IHjS(d?KZ_1LyUZ&K|2_7*YEjG_^CMrT8Q7FPa^^f#&OkLtj zpBo=~BJHS)&= zc_3UjG79P)gY+`3`b=$xAlztPOd?4L?2+6rbxm-25#$ZPe! zU;#fIhF@pM$AXA}l^g;n&Fkkc-+6MhH+^ulw|#Xwq_?vTZqxp%Jnm)o(%6evU_j>X z@OA!hkXtiPf`55%(?#Y9+|NJbOVn6T2pz*LoTg5umtp~1X}ZYpt~p)psuSxXRcUqq z4!o%r3fv|Du!Lp=^=s673;o=o zA>nrN?9+n^30yx^N$lUl1x0tvQSTqHQK%+kQV{t4mWrH3ooXyO@mk8L)*y4C+N9dx zs-dctUQ&l%?6pAeuj;6BVT>DScmCAzKi&NDcmD4#$k}x1U;&%gsO^W7^9yd-&w%`j z?Bu(O5Ein*!Y0$b@#biAeX<2NVbW{yHXdiPnt7J00LTp#1^I!XQxiYogeqUa;{5Gv z&*nE7lFvTV{I9aN&FQr(qpJ7cuK0o`AOR;5o^meZY77wXN9M^Zl59{}G%2lIt^gic+<0{wsGV!`_4Dabi(tNJrC+Eq1`CJ32P<0MjLSIhc zvBeI#Y|h_+IvI7@>LiT*KzAD*ISF{KC(R=UokV2}U}aW|XE}b{tI|+HK9ht$ejbx! zBbA6Waj8TsxWxi<-ejX^>5<$!`N4$9u}$wWL*K8(OPY zFU(PqT@I+oLUawNlvfMLOMi@;%l|^I$Y$~v2_;#Ziunqnz;DX2Rq>r}{iKp?5J5(> zdFn@1GM2k2MT=@mQ3~0y&XYH%fl>%4_IvS^te*DaDhwew znr+n>AClh`{PytFNSmNx-;?5n>(E?neFE5AW~H(C6iC5(dw4`*N@x-?{l_ z84w=X4wfD_2J}?vQ-)ZjU-HCG8=uDxT?T8nRxd;nb|Csp>&Vf9ElTf8S`nVrB#i>= z8oa$y48y3Ub~Uq*-;?4-iSE5P-J^G~AYYrPD}E1dvd3iBo1}EmS&TvdX+iEz#4t-% zlEepX2BMc?(tpux^34+{gek`meBV)!FAJQa`o=GFM!Aaz z-w_@xRKY~izriiu@LfXodp#bATvXrcWOkBy} z`+MZBXv<4sWb#pwEH-g1#AGP_$vE9Ch1?v^uL&_hYMj0rMa>m{*`JCf<2P@&)0d|WL(zs-;3a^hYJ8d)b)+jguwmO zO67RqdyCin1GI1_&Nj!o)55x&3VBG4ZiI>r}!9c_JW)5K zq^)kS(9pmq(daX{O>T*10j#mBN(orcoRn_)()_4SLeFn?od8&JYsIg}niGs>!i(YvO zgsyJ2H*NZ*sig?YR#NkHbL8%|*rCnSO99LMq43S5&}%aZ{ulCs=kN@{2T5F#b(?Z22w$&x)q_CKZrU~Xbnc_y_zKdV;Ks2MXYKI7=J>YKq_dg zN2vqkTC4<8K_iy7T%qM4zCH{l{4b(O1+vCCT0+;3{I~Ekx+kuV_O5xJ(d7V%m|V;W z9ZWFb9;PaYKR)uEUTmRLHLHc^X;r10$c?9OpjBFNjVHYLduVew9j;%3qTp~kI53#M z|7EYfheWUh>lM@kJ}*t?ST@h{Z7oZuBlcr2%>Dq=uL$p~x87wG|W}}MynBft4%b|U#?uq8I==R?q4<_c@4pf2% zvOzoeT*bV;B3jM(O6YlPLF;ob@(a#;e$n%%ZCX*R_jXVisf+FHl|5lbtZQ7i@LaLv zN1Y+v^)+c0v~&oQDfM|AH5Ecw zoVsQ0Fuwq%_w)fuuykZc0b+qFo4b2*W$yo>M_xnsH+IUapLoXltnC0T-#V2!b(?k_ z^k<#TWPX~2T5w%z*YVQbO5q8zAMXPW-LkBPwiR@w1>S0P-8$wu5O(f_?K9}1TOP|7 z$1%E%Jlq+Y;k3og`|vZBU(AFR{$zGOy?V@%9j6eeD?7-0ekR2ebMI=E-?*TjybjX6 z{!BM&nIts|(zHl!@o+~$N{&_NAxQM@k@q-jG^oPc;^hz1AMcSAU`ql91?m)kE`77+ zy`(^;Js7ZHmbH@3QS$uLHxCXS6ip_rKe7b3PvA+^6qg~1bDpK?DbwT0C-(%AIX;J< z6R8CH%`>ThXDpe<1C^FtTDvEk28rcI$p_M0zhppKx|2zZl&z5yzYIKWBJ2w^9NtV0 zoxX{dR*wRIxS-E&XOaU9gKZ{_hs6dbEx#G(B4ru=xnPFs5ETk>gYB2I>7b&D)Ek7H zS}k9nkT*)!Yz!|MGe@`>+rc41S5cCbyIPTSWm0IhL`lXbiCM)?2zM&K&1rGmILWW{ zw3^|`re>HrUs0(7(sb@+p);TlvGoEn@`XqEBXJkWLxTrB9*J|Bi#$ns#c`^8u1=Gc zB9pzA6PCpSJv?WK=wUxGf)9ZpH7MUm-D#47-rHdjH|6-lF=G1Ch-ukm?8)rVR z^DJv^7Z6UC#{19=D7ie`-X22l_i&T@8aGl2Tr8MU&fyc_IcJv9Kpn1kjtm+Q$pIJ3 zXrWkf?WkWeO1>{Wfv6y)Wt&;_M{~`5eo@$y8uXl*<}YRP&R`BK&kS!NdT8>*-c&p6 zI4LIOjNLi5X0;<1g~BYWQ|TpnYi4)lkSBW84>3R;Ake0nvKQxAHYA8yv{t`tckbA`NchKbpBIs1vP$lSMDh(h1BtT}NVK$2&BX#LnK= z-&+Vs`pCDYC3dQCBC&Hc)Gb(B@lq2yE|4)?c9GEWFQzAS>QJfect4Yt4+&w|sbR`C z%!KKryoifuT62O;sR#Pi}*Ign~<#IicX=rY96)sHT&QpChOyXS=2gUZ!$9gkXsLtTg-NaBn=^ z*%-ysZ?I@O*Pojse-;x!E%J@uIAX_-7IiEP7#ZlMwHvE@&x+o{IZTaem`8qxkP)=##mr6n49Z&uVQ;X zSCGHTr}6FG&B4KYw{~~;&hKuPj@L>;*75&~dWVbk@25M;AOf4knri1K&~17$s_5R# zWF#ZB1+Y2iTXEjj0S|;;uAbc6n~M40vR+@6vOb}jnGEJh@ARC(yUop3C;8{SIog^I z*}Qe{J?k~M#1- zVv9_6>qd-5dU&&I6QSl=9m$hVJD>XXUdHqATIn|2B7Kdyto<8*zoDNk*KB4zxUu=c z^}Lz4I#sMam?<^pi?lxskvtP~cwvUuCA!ht96#A?xz3KC%7u4fo=BYc!arzh}#On9jnhyJis?wuHKz zhu;g3`CNoE*y;Gn7Hm0tclOC!^Hmq`+1;7&4eoPTneOl!Iat7TlJj$OXU@vBrowim z7QGW77p|L!pC}D+xJBJhxcQs4UztX(x@@JnfORB4SP;i6P)JO>@Wj>O-e~je_|m1( z-pE`WM!XYpx=QQ8@)GyS2MY4)=rD;1Do$Dh4tucxOFr2nquinz&rED*u-&Yv>tUlv zLrHBC`JV-ORc>uhHilcny(ZNyA9y5seD&0XJsxM<;hS9D+!+}M1%Yd`E<}rC!2vLa z>gMl`6qzmZF$MC2FK3{lRL1Q3f-VEf?SHg%w#Jut1s=y-nC@OV*A&Y08(x6kAO*_q zMkr~kxn>~ynt%j=vx!u0&aofOgmRj)rvci|Cc7O` zhDrAItLj(6JXH>W+cExr2LP`)4xO6o<=nB)m{@qjeKLR8Gv7OJob)kbi@}q zA?Kxm86X|PK*kC005L0{nPb0ORol1y1^GlC4XXs4Ol#)4M$K#OFHpi0qJi100r^ys zKra)PT;&;}N`lSnc2=vqdF>FtU#lp(H8uJ0ocym8qLG-hfoP+9tGU_D%j&JN&Q~GW z*oqXo9%i|||Gu9kL@LCZ_!6hgs1(iqx^}2ALwjRrL63ulr zOe;V?8`C;so^C^)iKm(ht7v?B#PtVe(P}!AU(C}yLWSJaU zdrg$dk;YJxOb&@^f=rIYlY+I|V97WyhzSY`4*6utS*5TE&bPowFjR-U=aZn95%oXXc)9 zMTKp@j+ay)A`cGkk1MIF-c0C%euk2b`yLSbVdo>Byt=lzw|fP;nii`Qqg>;&-*{noeA?r^X}b`{vq2l_DoyvUXGQHoHd}dSDRJ*j z|Bmy$X}bU%m@yg0sw!@@ETzkr*>XlF8F z55+}IRiH#n)d>LNeoPv*dpfzKwW$>3edDN~?E^EzmNk1_N=Dp|N!yju*gPnPlP{vJ zr=rTFdwrEdVN*@B_b$k96t3sZ#O!6OtzFsLcsPNS>l|bZgG16E;-}GQ)bgeuM#2#{ zrqMgnUP7IL$K?W>)ufQu+Pnvn5>(kAmc}QbTS^Uiq3wwJ)Yx#)yJJrMt4HCl3ho!k zTb%EOx1`av0S3-q_p?lAyfYo`ZI3p`!|7Rc8CwC;W{>~O0g20Q5RBE1{dfxy~yaSdv zRlYS!90ye9X6{E=E1zpr#z@}*O9Z#JMM_i@O~zx*O7=HGF<#jt0meq+XJLt1Xz`B> zl{Z>^DtgK?lEjpe{Gxgs^awpdN}O($+t-_?#E~k+xmyP!&;c7Mrz-pk4dk63M@YDY zhK*IC;_B^HB7j?^wRRi1Q-35fQ>mPJ9S4-c?94Njwjlc^Q;`JgJX4XrRjo!;rqUiH z^+sa2K0PLy$YQX@GUth(Cw8NPxFnEqA^lvCEyBkJaY?EyL0lRyHi*kZRLu;mrqil< zT1r6V$Ks-z{8@3Ya5j^1_AVK0PZG&!GS@${NfKs?Y|@DFkxd??zC=@`MQSGKj%;up zUQBSkINW)x4s7(Ns*re){ZxL5chP|U6F1Eow;jGR9(fSMT5Le4sccijzrF#zEy%tJ z=#pSYQ0Fc^$g2?*&?SLeX=2<7Y(nWloDS8chdjWvoyVF{B|M4ScO;Y^*s`(QB_;=w zR_=U4>7fzDXstpUmT-)wHL+QXG_fpMi!@$r)*=rPleNH=@Ua^~fr_}&!xEGp@*pWB zz$6qO1UBe|Gprb(18R^lDVntGMtneu#zLxxaN5hWUF^z#csq>TA-D4ZC0U64JyKfLMaUzO*hX));m_FsAVQeDxy&?9b7^siU+k;mEb@m>7#%ImXr9O@PX%nj3gI! z#FX7=_&{8(k-N$&9M*V9S6ft-k&9wP1Y{#viV>o!x5ghD%IE;=;0&`QmDQO@5@3__ zkhjI51zdI=X_Y(#;id;MBqC2{RG~yx5`QuinTK#ej24|U58+>`)*&hnQHMx9`hZGF zOw8ayfHR1%lWGD?D=T1 ziM<#AD>ph_FnM#_`4<2DxxEokqxW#{t5Q1c-uEx!N+@k%1~)cRML4RMN@)B0;z|a0 z!}SsJE|1qhG4Hs$Gn%Y#jUOL*H#Y_)kl&U~yy%wjR`8~dE@_###+r+UoXz(E>UTev z;5S&SBV4~r$PI0{5qvG8j>WC{!f0!>F&!MJ|8H7Ce4PNvS2`q)el#&nx^MULpbx*5 z8MGfF_xn8_2*8Z*jibHY#*<7c+Qe|OK7|gCUG_i#_Kro6Pw{y|#ZywnshBc<-`E59 zz&r1x@{_6YD$X|s{6<@Bt%dvUAkT3#co?{({A{|1fn+OUq&ioP}sn6Zv$Hw69cIZ7%tVS$bO65WOPVPFT_s!)CP> zF7nHJUO`?qfam5nPM?VgE$SSpTk^ZKM`{@RVBChMsdTMZ#~V`5xY|IjttZy{Erhc4 zjnl9ZvlG%gp?M#)CSZBYo2@Fhd$Y*qbd3kpre<5;H~DLid-)B)NuMfBH|oDF&zPq+ zoI=_=9q;Ze)(_4v-Z#89+Pi;uZ+lT(4^@$jOHEx~k?;U*@HRBOn;N{y*Y(J40tt81 zo1m8NY}`Qm^QWSiT)K|LEM{`4R?lrE3McSqAR%k@&wy}PY4aV zHtGk+FMGUA7%??g#5PRMa*vDl5ioN1((tL0_vBh5I7X>gNI4UIF1vOAmbJtD*8Seo z2jrujKGH814+Rip3~V2H4GFNjVy^%;ErgRziN-mnE`#%gt)upDWnokLGYi{X$ zc{h6fC|>C?hesJzK>IO%;PME0-Qcxe2QJuAaCdvaD` zg_t6dS6WN$S$(l!t8?b73UW;v{Z_YCLH_MJeb0wBS?A2oV&0G}I#wb*>X(RJV!SGrB0=_a#0U_dO#PC4U!X!69~wA3LhX6oGV=H2Jk z$`-}oF4UHBpIeD*L>{EZ$PU)GlKiop`&=1&U_87pHM5TEjJTtyhw7A5GL&SQd@$G6 zqC!iZpjAJfqt7SY#D7d;ThvI*thL6`3N$-CRkl=?vPWt#Q)(EiN(WS)(Ny+3a;37W zm?D*pZv)PDu+2p?`j>7~SpMytDXcnZYW?|lb0vF1EOxM+XCrI!7MYShF0#=y1l$pC ziD!TLl0GhKM@b(aKVQ-(hO9adT|QH&TMawEldJwOESG*QTo{4VOS zF^7L9S7OBFG!zT-Bu4xTITIr>Xv4g=D9UD>KCM(pF^pB|pB(OuhdUdipir-!+@hiW zYtpu#7D>l%7c7Ud}Cg)XtP`ytb*BW zHTvY;U98OZ?q>5s{Q2F@vXhnXYJHeu_Q^ugN7TM5J93)S8Mdproq3R(*A8*oZc)rq z1&S9z43~UUkGw9sKK6z?mq&|}>2Pn#bfO4kacc{@>z$jS(z?{e=4fj=WV+C&dHrm| zvJ~V^^TgTD8ysOOyBOT?5H*joQ?u*HopmO3Wgj!Nbu7hNAY}gHc=M@2@6JB?#vXY% zyMMTMytTFXfq}1cgNeop*(?8!?Tq?UaxSdye>wJ+3x%MB&CLEhU=xR z-QB&-@srKw_3Ze`&U{V&;MrIinkij}-7WczE_Se5H9jTNQG3RzM_g>i?;5fUR>6#n zU+UGmVzT4w+uHX%%^YW`R zXXV~#`^AfJUXXuDrh8XM)Y}1*Vz@<}T?&ZI9^ec_3# zLnsxW9bdXM+8db<4wT9f62h>{+MpGk{8&MT8L~4>WWg86MQr${KH2R~XyaL&jY-h; zdY9d=4)b4eB-y8Xf@_U_X6WW|;OVOicz}+Y1OQVneE=+f? zoNJCGyENYb{X$aLq0^#F>*m(CP;B#LN!A-Xmr3tknj)p?%`m$SE`EHv=rXJ0ch#gcIQcFpx1a?VO_RK@AP;50 z#`b8qQyGuiq#3_al{4`^`TMlaJf=5EN{Q;GY^^O!00Swbw=m*-xQEGW2CvS=hm)kF zTsZO#eez(|CE^sf2cq*_Dx-nkmeR6jUL)fm$ zsp!Gtpu5t6LRqtX+YlUz>!BO+sea~Qu29uFGqDcbRF=|K8xJ)-P2z>kYLV^O?-8iS zrrfilM_oT!+VyaO7Wk1L>?hqUX;wi)N6SGo7k1Q8q?4|2GK*Ui2}N|)X=pQAF}tl6 z1H$|hgv%FH)l(W2PICTiZux4aNkY?6*8`4R~ce@6FR_qiI0U%W}}a#PAAWv(TW#*nIJjJwtEShD@8(au6WIG!1mJ zE__Jdz!yX#n+?FoujSr7Y6k}?iJ72}(Id9zGGj(}>JXICmuCJ#+SLPkS|bt1@){%4 zVmG)cRe;6SqxrE%)1O4Q7muTQjx^+dd6Ua*e~?;M|_r>`zG`B(O2j1b>$t1 zBVC@aFw}Hn1S~8U3zv9N{dMHEgRe?0s;h2PTu@0r==Vn00rK-6Z%}E4)Xlx!EAWog zqPjpev*$KZ&WBEOc-U#VT=6W$vtjQUPNqU>ZmQzP3i5LC)xeLbb6|WA!z`A=2X-Gr z^Jx;e;OJb^Bky%b0)A4UTUfDxoUx@-LPXvN2 zf|K{-9#5^ir_Py7&c_PfV}TkYs%Cx`jgpO+4fznm-rY-dvX#YX+lUP6qdcQD0~ELt z8VThS0irSp0j9u0kmZp|-JIY(^U-4SJZGXJ1QLZJ-0}A#y&O}QC{wu<8F4|mAF6-? zWhvPE;f;J}k9-BSx}fd~O|}M2GRrB0>Z*n4AX2qiOBM~1KQ71v;kuDg&_ww0Z~`gl zIZ(6uVr_;X+-P2)Qf=5RV(zXVUJ4yS_ygZcYk_Ue1k8)4}+r4gc36V=~0^rps-RysIGh_yb3sn1SH-g{+O_bJxn|lB4+--g*>zNuGvZhvC;5_@$89km~e2 z`JSM}%OqZ-IdPG3gsodl!u{m82)riIYBsW(G`vXL;yi~u>-24d-oc~vwd4YRI1Ine zh!upLY(Sx?l4A;ocatl-lM#Ce{pCAPj`pSxj`p^%PKWe@w!v-MUzNwb%w8IM@d^xx zEiaoypL{sTt(hmmzdX3IQC5|>FsOJ<`*53&py-q zuQI{S={30Ton9pd2}r=9gr}U#xEfUvkO))JuMWu6UUNdm#)Q zKjnomaQsGuK9>%bc;Q*(=AF?S*5M|k$@=zi@;E&GHMejJIXK$c6o0lGYL5c_PkE2R{O9?zjgB1-gxKn*2ushdvtVpyu-#Wr~sp_t?`wKBtT|+USG=W7T`+ffCDSn z`ZDo^V+aC2ioEx9BWXU@@RRf8zI?6$RH!IW|&>NHZ>I zz=p`{p#<`dUSdYARa25zAV(6^;%NDABiXeP>R=gtdf1(3@_scbHu>$VoakTp`(lz*k!_f*TKP3cMRqx$A`8(q zpi*8fATPi&dM^J9xgwj%UxYNtnWfWIj5#rX#%)tC;LRv2cuy{1ec!(Kl{)oAD#SbBhly-K%G~V01>7k9_;N7H#dO9dahx3 z+`V^s;vaK7DJ1U=V%?Y})wTf1b_h!md`_pWLVcLZQmHd}e;4E>)N=e7g_X55uBr}N5WRxA+-93=bX*4c%NR!o$Bm&}4rR&^tMq%G z*lFYQ*vaQ$?bhm%NFvIZ>anr~Ta@0Hv?4sKNg4&#HF$gF)>1o1DdZ=mgb{m|Sqk#C ziMrxf;U;@bX1z&Dhoy@#=szvU-H8}x$x4#=pv~^l`v(O%A2`|SR%=FeU$5e1!+Src zk$m$63Sr7I1mAZQUl%WX5mny(#`WFO}HVIVeG#b27`$gL2L#ym`p%#N0ndVCIcHlw((u~0__BPKyE#Ki{M*V zbW_Z7=;OhmQ{?hB1_Q2epn~uPCSYwKbp})FVcT6wdH@r65yB}JO7R4O^)DD+|%TCN_ z)k|2g{bpQmN<=C|>qeFRbeal%O?WD#3MNva`16y9q=yU9Y0|^LQmsL-LsI`aoRc&T zDXwY#zZbzt4;KJ_sLLCv`GEVemCEnH_ZF`Oii{32Q^V*P+>iyNORO77v!h0s|`em3YXfhuxyLF(jFV-}gnTh1G< z)R6-k7X$A6|t%rv_exfUydQ_zT|Emu6J zpczptSgak%8nY~jdnL(#3qPWJ;_7Jcn&%N+?Bpt=ZnO@M-3DL0!xcg&7;p_!6~rGO z`A#pk&Z(Le7dB?5B!VPZub>|Ad1W%kvUyf-Ygsz2-omXkp?I}v2y*Ad>TM3yVz`d0 z-NMy>p|(s>f|yX~a4Zg~(C{QFQ(0sax{Vy-4tD)f<)RTxcTIFt{`IiBc>_!+G(wy8 zlWstH6pN`L&V;_#a9Q+9b6K?EuH8^|+Rjsg3H~-{2cN5$7nnM8dNcDprr~xPHKSXj z=?3a$G_+b7gy`|pHZ3UD8xSb&)U|ePWmk-MLA%BcP!{<};ke_&^)YD@v~UP>DGk$j zfuUx~M~0fGh}y_4Z}mcxZ%@)YaJ6c!E7|dL8GG@TwZr`0mfq6`e0ewpy-o=4HRsbc z9Y(5Bfg+o0d+dAP-a465bulhn>0PL$YydvS<0py zzn}E$dno}y*t|G`G1v$G+>4*8jP@ZjgPcwnHxr1AIaZXvlbvN*sa8WGBQU`53qSmH z%~Nr~a>}HetZtE=6txbPoQ9AX_h=R*_j9ngXOC@^CF>iNdjzi0OIXCVbRsU>#@m(M~CwK~JX6IQ#o-jG?&bb4K%jG#R z-BA3=r1&GvpXVpp)LABkElDt=m19BqC>p4kxxtN*NzPu9P)lv{sE{T%8Bxn8-g5UZ z&@#iD$)VFX(ai8D(1Hu{*`!^Lq^I+OPh+}M*5LN!N8ub-aT)%#;C<@A6xwiO3|_X& zv<@Q6%QIl5(eia^6^SMm8*%1M!EtNxhkNAyEP`_9Dzd9Q+62G>J2=7rOSWDGN7sA9A}uUO9Nc{g4}qV5(YKZu%WsCG2^q z@;j4rgz7X+DIVFA!LVfYUne_PA)825(@R@v#Ur{}m0U)rVoOmI!=_L~YSO^LS*i1a|&lBAlVN z`OxHvy{Ud|aZJp~ozkRlzG_b+(!v}|Q)wl2^Hm9EL3n z@P5J^Dy%cX93Wknb7AGZj zQ%6co-Tqxt>LX-f=Y9zrEfbVeQXnpzag8zej%PjsuW2a|moH}u#0O1Gfy6jfhnm@U zJ6#+3w@HOtA%w^savNAe`Id`kG%)X?lC7`nT}jCU7tT;fN=P2~u*u1T7^!Levgm=s zO|9tCCs}aDt|82`O6!xuz435oV`SjV*h?N^=U@LbB3@d=8^3YHP7eo+&2!V*jn!Rb zMQ`B|UV^$v9vVCtRf4j7xx&5@qb-X>KItOQdanl9#h`!p=4Saaj9De9Ms3mci`>Ga zZF|hdPGBLJHTFAKyPY6akWXdP_V(`P;NZPmySsbmcQ+jm+J!7(R-GfNnAv)ezj3ud zc&n$`Eip`+AB8JAH!ua)h>FRlUAHTG&m11`yj(qrw>M?8SER6bmj&JJ&2_Dw?6ck+ zZB2)4*7}5IyxtZ~lmFb$6;-RX9^CK{U2;diV2u6bbTE0vs}{Oz^F$t5>nWFH{Dtc| zOK;`hJKow_toZf&-iqJc53fy|V?LeqV4nW&xDCZyjg9wrJn`r~oLChR3gJ=1F?|AY zoxR6-9j&n-@0%wVXc;r4o4~vnXRqxHx5pb7rek>AcyP1w$1>AqpmJNe5+TU0a>*50 zH!>@d8cpXm!OyPkxkZoLEcuk{eg&9o5XJrn5VY|SJ zKy^10oSP??=-Kdai~5{kyEm)8vWiSuRi!pJiH)x*edr#4(He0J7oNB}+#7A49bdXM z+8cFS=7pEHUlp6&VscuJNU?xvuUu2I9{H0V*~~1d@!Z39c-qNwmK`ZSt(g*DZ@`}p z3pd`$#&B!6*JQNi{g1@1qVE5Pw3sW@G41ig zQFkDMY$vy}fb~yXZXYdOo#^H1umhC~)7>lQntgkAiwn>hWIMTi2P6`VQkQ!71_f!zXBkb9keIbS8t!z+qxqwV-V&xENi>N= zV>$XX95VM7Udy>P50d)__a$*_Bxy8W4f$Qy&1zmNz_~ML!%{sent?YH+=^jVJi}4H zQ+3+6#s&GqEZSAMGWZ2+zROkS(lIjqndqPVdEvM|(Mu(br)EbtHg3$UaEa8g3-ZYX zMv!m@IXN*MswL}eom`j>Cn8rh`quL7zw76ShzbWLy4-8Dm0XS^z8KOLtLDAf{gA>- z=`s`$sJ{ibUFo4qFe{ihGO@O2-q9f>fe^IN)H z1-nHT&1K1u`BXfXrj*!icIiWAA^#)c zm{E9BjlU8}ku5s)v7b($plfp#yp`N&iYH)bLIPi9CZ=ZhH$B>B5q-KVCd3ZaC7wz) zRA-1D4FImcwtpy#jcc{t;M_#LYeIdXGoF+s>_Le?(M2+N{WOD1ml!D-Oo_2kGJ+faAt@8ZM|m-dstQ&^TEgIo5SjdPJDJ!p#gl z17@^i?T2|0^p)g+!TG2nsOrMR9aq%J(;Ke(9tTE0RFMDE@BC$RZ}$p3ki4iW-ExQf z@$wF=TFS<{XE>R%kW_d``@;qK@)T*=0kw7+((&>|KJvoc-BR)i5xCzWLb-H;i&Z+$ z_Q*=Iilj1*OWw)9j8dPF$3mbwqFu&ufXtun0e4S-vzfqi9-gt983{GGN(SS>o{bun zr1&zz%49V$6NtO|>$eJWnvUjf?$A^OZm>pV#Po#&DNT>o@?AIksOHn$R2_J`Ep}#{ zEfG5asK&Q;!e)qDo`)yEC=bj=3i1!}u+;M4SnW_QLxjx;qvz9U<+^U80CSK1GOQ_c z>0gBU@6pa=#NJJdnWw;p$Evf$i2SQw^vopIMUrxSZ|U?i`Q-aI{M~Ix zRx_1ybNTpor-9`tocx!9ob{Z#Y^_b`8xJRNf5SP53WGyy;t$c2XtXI%8}ZT~y=Kfe zP*zhy-YoMSM@m8~-@|hBFLA>^*DGa5Agz(!IICYNdvnj*uq~nI|*enU40h zN1Nl}bhNlxjY0#|NO z8DLheuzi8LL!u{~X$?H@u}GP%J^Zfz+J#QLg4f#ZLZS3C~crhX-@in z1SnvSgyd1xtCUAP0rGbD=?^ZTVb7GNwR-22CxW=OWC2WEN>VxR+U+aFRlP>3ve~Rz z3W-Ze+H#1L#NKQ=*G@YS4rMPgC6VMZI(u88HNv2zPt~-Q2!m>mk$Pz_Tq_4m1kJWLE{#2C&53--i z7yFh1=)Z9CopGz-ixQFn#szPht~NFJ>l(bn2-eJVP5x5r2g-tqMW zaj;X@6SCmZa(-$?lgDm><-znt?T92r);3F0(uO%7Ur%ThZnV^_|W~MWfFF8P%+FCMY zX_2~U-OL@Hx6n*(1}4$WB+MNNNg&sgIeR9@ocqW*XNC-ajTJxx%A+FBB%(YEK@swg zP+};cN}w#oe*jC6l0s3z@XHLttBMM!{wl-c?sOx8gGj6@lW+mt&LCxE39V)fASULqVC@nH1cE z(fMdrghUES9*#C#NdG9Usj0=I0PM9T}BKz#D~i4M}!)wD_oj@L(l*|AUg9Q{D-&|_+u%60QvaS zr*sVJ8DaEdpV045cCR3k!n}*$oY#M5tDb$Ks zJ(+Vc(Ylbd*0K8`DLhVOz%r+}a*ViL>@+fNm;aun4V~euR3Rm@Cl5uHrNq~djSE01 zkEOTOc~uogX>REb9k5cwQqj`8y!s*BH$@~go1Ad#pO6K4#69HF<_~wai^=9_cweDD zA;bz2-;70_^4z)x3pT34vQCXIUQg%2r^ra~tLoI_det7ld*{#qzqMvwTQ?i^I1QZ+ z1i69Uq%1EAZM19@qvG_mw^aV`m>k=zNzB&{g^WY&@bR>0=$4(cy)OJGPiMUF4Q}$u z2ioSJEgVzCUpy&jH96pIzKcQ8m4 zo7`@jxKcrs$UY^dhsAQy>0!A-1q?K#hYidk|J)&s#w5?Ly5vNtJd?JlDq@vKlA^y# zQf_A%GLMQI5t{sl1@8vj#cj!%^3nDsNTQPULHB^=WofmU36&udNS| z??he8EpM)y1m=f*;iL7()I!Z&7m+3Pu`qi%t^O5#Kb(<&bN8p$$hf&v8d{Wpn_sWp zSbqw^?R2=ay#k$$cA>e^E&U6{_{7e5YehC!Vu;_o$t)>N%f6TByJ^jfaqYY52Q?1h z1(gG6eE)edSgR(FxGi*?M!6(6+-CPG@)h@B2rs!&o3G#y$9a*X!NnKkKNGJR@CC^e zixRL$UrXfk?)iHW8{_$C&DQiiQ5xR>1rOQAV@UFWij1b3`Ak0l@yhjmd%C^FUT06X z*15P;FT!*!&z~}%F^3Hn=y6TnEMrVh7t3&Q_+@8Yd)COm!%BSnm zdf2DSTBz$0O2nF})zP=Q(QAQJap021-sT`q&w9s?G+MIlC6}*V#+rDLPr8Rhh4G|O z{BQvV!|0W-tw9TJa3(c@i*+@lm9Jx5h;f8g9$Y*&}HY50OtJMmx+~ zGy-Z|*3xhqEku{@%|Cor!*235>@xv$hlkP$Roau2`^VE}3iX9T@a~hAU|wTOPRlSD zSA>jbc-L4)tk&Wn1Qz9e%Qe=XAtp~CUSsPv2YMr3gyJ%|jC|38NojZTl$(uvZR{l< zM7lpPGbv*+_Tw=+QDCq5F$E`4S|BdVD&Hz~@L7$*)_!uu+Fn}NI!H^!Xdx>_eMqTc z=a}Dt6tb#R%8)hw^(|JYDk#&sl~zMl@oB_RRb3N|c(tL32g#4E?eAT&=SluQ+03t8 zi@W4&NMW6%if2nvE~+{wffL>dg0))VS7x*ZfoS2KptPOvPArcW-pLi&>pyX7tHo`n z|NJUaDqoT#W1#kw6QxcrNY*N6JrXh?c&++NK7|wn2{JVaK$IXzd;l>Bl51AjXcG(d zdh7G9$&%~h8Ta%3@vy&rx(GVi7lKe-7n81+2119gDR>0Zz4yQZecC!EZ-{< zgLjix_U=qf3>s))UPY-~8$?JR?vf1}I$2&D2FnWcr`NqXl}Jz{F`kNAZ|vgJ$GiEl6c+XHd;MKfW`_+Yow7vavAg63C6b zu|sahwwS8Q!u^JA$JcP5vVZ*wfw|4HIVx`5uaIBqkh`$CG45}lEmkJe{&>nc()6SY zmuSbP=Tj|mVo;2xeKRL6mF$h2xVT7M&uN3t6s8ls%T7|&$lKJocAQvSBs0}mCS>u{ zaPUB{b6t0iR$=y19rF4Pxf9zhymUAkt(@kFNv?gc% z%vK`RX>Q6xKgj?Y7B%mX*LBDp*cx`c|G;qT+?MR+)C`)Qa_;dXUHo9W;zRH9>W92t zTh7}Pq)mk`fohZYbxA^6P;aJZMmsy>!SH-_H$6H$-G&Lsfho59`701utya`3r=;(;cf`+G zexXBN*_tK-Zmw)0);5pv4`9w)r^n>_myy3{la=u1sBC2J-(@5}8e6>EnC|Qzt4fM3 z>a8d~JDE)5Hw5!B*eS#|91`2AE#&R34(cHk(F1z>!!}hFa17;J(v~Hb&-n-qV6YvU z0-(S4j>X13uRW!sV5;_H%#w6l(X{J84QiS6bBenuWbmV$~t!xpe}R*gqKv8I#?B5*z?V3P&3 z8byHX)%I7ydO4z~j_0z}GHA;5vJ!qV13jAyhscu|Mq*ZS!4qXRm(A6)EnOq7svQnE z(Pd8|4t%eRdv+_RGK93hR=e<>6Gp;nmedkSHXm7=on5DnWh0PZP3ULRJY@Y4Wb@&H z;&lxn8F_CvrAecn0TptrjT0rM`;4C067MGiqb4eBKGHI4Y)1Zg5d+I3?NNb-%_7B6 zuP)5}<*30JMTnu&SfayFV{PgNq)N?4fxJ**|ASH(mzv~-2%FFTz1-a+D9>IR`P42ygb`EuB(03<|U zaxMogl2v_xIXJ03TjWMF+qA>^p87LMPyL$_Obz3Os)n3}t(apIuqFBFE}Xm_IlGF6 z>w@#T5I=FkD^(IVD|n@bz= z7yMc-M5_EQa!2p>#8kQJI%UU4`aL!d%cD;JgnKkoBQH7_@9e^Sr4>~;B@X9mnU;sw z-6u8=y(nMx%c=8KZ|qN|l4DL~B%jI1P4c5*-%@wLAsU)J9131HWDC$e=%p!Jp#Rpl z6=zuYgEuBJS5hV+j`6*W{wSj323IR8YDSH=C~X2SB8eqW($G2+)A|&_AU4-*ZL762 zO2849ww}Zs>8|qUg&9`_a1Q2Wr|(;!j?$L`7KZ1+{509^^R3V!;kF1&#!1)O#!Vre zcN7zpqTslVzA=Cq;e^o&=B8n-Xs%0INBE4_CEDgQoD#Op=44MX`!syTTENb{AHVIq!1Lr~6xBz>I>T~x z46JD~tY9wr%@{*5FbWKXkBR+h#|9_G=*$YYufquDP_1QirEG+(W^}G@gsf!HBNLAS3x(I*v!Ji}WM}-`i^S+DZv5){uM_t7?+5R_N z$qm}z9B?g9Ssh_kknpn`O+%kSp0;*1G_GYMVl((*AN)B&FRA{yA4bL$3U8C$ok_8I zxRl?`+vkh%^tNKWb#B_H?aq2vH~uPru8+RgmR{L~^|+GRW|u9JkA}E1{T9TBdwWmu z?K!6&+r`5@89H~V?w_LEVI6bC>1LBS6*lxOrcysbV8B?Qy;dHd2WI^#y=d)U??W$; z^8fVm)c^f%$>fElj_ML6Ap5Lyg#J&;YXWf=ZbComEgc0>+F8S~xxj9Au6an}4ZItrfpyJNctu6p8Fm!5x)(Y<_&JMTv-(<^ABm0ZdXt+D6|G&XsZWphf`L|b+ zJ;nBb|H}o7&UUDB-@H92$`*Ikf!)Q`JJ4&g>W(ZlXS5Bnf5piqTuxmZ4&kv@ z2o)~Dd<2+WDLXjDaYY_qBtMVl5`clKQ@}Xw5KWE?I-q%*zwWfL=;~9{Ilvs7Bu5xj zOTs&{a!gg{F=t$R3rGlGpJI2SsDf}=O0zJnEID%m?l#u)3VFDb7CJZbB&pm-2zVru z)F;&b*TGM85^p=QqJVO(O!I+SO@dN__x@;|_+<~7F^)=;LisX~mgfTXK*BOiIc+3U5%R06Wq#Eh-7fOY zc4=wVKj+P4C(@}LHY@(3o?pfB5dja>^t3hD%Xgkx{bH3*Wm$B3!xJf40ymNb8>fy_ zg-p3KO*B#Fn?hM64<*^6q)*=F;(=3t5Yn8o;TIS!+C$9*n?eyx0KP#_GVgrBa{)2f zDe@n|b+?);6Ara>rwMG4zBzYcMHFUmuoRCP0&?_^2Ujo}JCBX-8w=6H$REkBaDbj{ zW(}5J$osId2l$VMAY}A}He&q1yHSIC6f%{dd(sz4ZpB9W&`l$ zGu`Sxkir3Jeg5_%ctt)GTKv;j#Rk94)T|LNAp4f@w9o zCdoSCm7DO+WcS>3b3)tEOnQr?v$g>3`oKMbF1pt1RA`PIf^tmaE zfJj3xPi_`j?@R(}0rRtGgOyC5OM)qvDSH&XV^f_ZSKAa~Kx#3X{GE&(NwZuYm81|4 zoNV1@|0*N5gzc2IbmF-4Dw;MSZ&A(#DZm)i?b|Z);;;>hDHFM?aE(%5|84szg&DD_QBv;{U>e1(8A)T9by$ z-A^>|ojr2Z+La97$r82nKeG{sp?dX(PUvwf$g5NZyrC#uC9xaJugqTQ>=~u3Cs;<| zuz`_Nc_LSsPwJLd%Adu{2b5#j`81R1O-W0u&qLsR3cMhi7zA3wABX7PoxX6Tym74D z(!WrQPwb4hR^&5#s)2_>^GIA<)z&8J`Q;a?68 zdQ`e(&=X((2oG-r4O(A_6yAuBFad{zL|xFH{$#7Lz%RQjnDY4 z&oz03c^mlJ$^}k$K;%MSWCX?(VuvCT5W4`Z6dJmTk83qBf&tcpnS_UKg22?!O)La# z4h`Mpzhzy*?MH6)JcmUoJY%P(`pCC}ua1HO`a^}ASoWz)lm~XG)z3yj5z#7nb8vhr zHVGo4n@o`u<5M0On+gK7){tCyfL{86;kF=%pYQR}s91_IH&h|yzb;#LOV6NON{k(& z-xM5(YogAXXIOy;f!G2+2)rOe0__4vA0R`tkJIFfD&LIh5dsVeEVwoq0ycuo60euc zf+bQ3d@*x%7OIcZ$RCEh)n{k4pred08`vKmw93^cN?kbM;p+Q?FIO|R)=O3v(OUq*5D(LpLESTqd8IDSq>;U6R7_O@>-#(`bDON z0+T9|>1>PQ*F4bJ6{6VPPOx1KH*Gh4);H9j#f$6f!L_yf&lTefp0zau$OI4xL zwigcGE2Y;A07a4!L1cbe*piY~v#^k+4m48}dmuhe%)%n@RqBcb;1klVhxW`1)M9EL z7VQaKjUPO!P52>(DJ6C75eX6tbILPH&1IW| zJ!u1j8&g3f6xe7o>Cn1}Ubu5^k-X@Rz=WL1XI;&3t;M4)njWk!fIMM7E7jHOw6^M6 zSCHu54f%c73GVt>Z4)#T2)UbFvyEqIGN58*$YHuE6`P_Vu>IH(3+sw@fT$|Dt^~s7 z8%wtQOpY(Sa($m@qtdy!q`+w?^3fg&1yBWyta{;K1?ZpekXO)4j1hz%$aK}r8P0x#w;&ci$R=c8LgRN$f9xELt?6#^ zhCrU!j;`~Qrij6P(@=iP0&m)Ii?f&|ry?Ytdo&8FzqH`z9;Bj0Gap;#$&yZ_$XEI* zI6;WB)Zw4YU+RzqgRapRbTTLS@Y*L!lnN?)xqU{T&Iz8PwzK0rAWztwh%>qbnS4GV!FSS&c|`P}YM_)#%(x_k~VluG;FMDifHv zU73OcC^NE5q(;0-3_9B&FYDdv9(3lEf1Pgn9XjBw;}O+EGvGvz%xb|(^2nT|p)shE zCbGq{6tbNtO-Y*!RXsy#bDoQ&C6A~fFuJr(rG?~?_+S~JBC<3MQvK0ZygGw}KD4C_6ZqAVM^Ee(EG=T?{!4G*(PVOI1n`Mhb?(;MM7w9rk^qdgv52;Ia7lS4OkC3*RZ%Pf8>DfE%F@VZ~ZMN4PODIt&`PAg+!bqTq( zv=B(phZzEiMH53HxehBDdiyXH_DepQl(>~d$k+i}!4lH9f;_E))vJz%o@dsq69J57 zLhvAn!wnw9vdO`NTq*837tCg|PY2K(^E}gfzCRxJw@()uyo@>J5!N{FdlBcQfxO{s z3SK?zam3Ev^~+gDu&lFuw@5*qA}4#dMWvt&UuL$iL=Xl|Cg1HiCKY*01N5TTyB8G;VM040?NBI@;M8AKw||_Q@R~ zB1Gl6g%xv?2kno1oZc5hfMy0gDjgoTY5i=ataSxb?ApR&@_pBgG;hU#S(s2s-F@uIq9I&&XTQa)meUY){19^%*=Hw{M;7$$AURiwg@!j@Cm| zI@DL&+n;84k<7W8d_-}A{Soy!!FsP}J<|%L6f0w!kcV5CyL}U z(czgh#kgoU&HJH@ygW8+vC%0@q~?Tl7zXd}I%I%MskZK6DxS6?PRDqkNF_-vs{TVp zegT=)lhgfCe_Tbh`H8z@mscwvd}x}^ku9*cS^TY%*^xmns*|zQ7%Tv=J>`4rcDRUj z)N$@P0fxfy(P#&oS%CVdVEcP=*Gu$#Ev%rjG2PicRwcaAH75ba?54vMU*p7djj z19UgXUK2ztRLJA({06>~LAAGh(b0+|^0^#L0W7o^<^g(yedagrF4!tW;UL(uV<{=g zRTUlUG;s2De7k0#gfij!f8nWS>Cx*zzz9!#M=_ip&Drp z49%SaLvtHB(YqxHLsR>zKYr#fJH$S}+;WwX9~9V`qhX;QT|tAJ34V&9XFNY5y31{{ z*{ow3`D+yJs#qBuaXFfjc}aak!6jWP9{*Doj{uVz%wh_e{4il-+T5h4=gP=;6A(di z#~MtGA!n!ogeWR-J|XB>W)a1&4mK70x9Z1S%Dx^5WrL z2K0oLs3RFhH1REMO2MpY8F}MkJZeF*OR2KY;#v>_^Agj75U@}@3j#eQmIWcYq$cVI zB4|@V1cv4B6UvNI8@%uucG>Jt1$K7yj-jofZ;%F2^qj48tmMut>Uqi{Xp`lP<4h;gA8weDHqn)8o5WG?_^5h+w=Jcyo zORB`K3$nEndWM@TI-WFnxScKEC6c2rBe(XBMt7kXi(!WPe1IMqHNxoKufIo71g+lq3()V z6tKoOchdN`TFtWp8idbs`wQ~lQC^L0su2gra))9WB6LRjD)&AmOB7TC0d zblv!GsZGW9q~K%HV)iL*L+-+@?#@xklEB*GTJltO=SX6XZt=(-8VGhdIIPjwgRp0aX|e(on@_dV zZ)UhXEyi2LVA!7)D+8t@^l&?ztOHk*E5rcI&3k&CYionAEl%!04+h09HTeV30M~Tb z7p|B#BhKpbV&SV6j2lg!=0TiXl?rd!HXB=afc#D7m^SONM}Bst?ov1LT>JUTbA;^k zbU$qXJzdRX~PVCZe6%igRWO}X=6*%t;RmM(u)O- zQF~r%z+;Zp#v81I$gm=2xzEU1~h`smLxWAn+?4L?lBF3qJ$MNN*=W zUoZwt1v)?{V5RL-+^`vvkOiM8OEy zim8@-FvKn-E+nbZ)pkfy`&0>AiBPE7HRQn@*R&VfibtFtbt?FCE?1$>(_~`~>>4>* z8Jf-@s~ZxA8hK#Xh{8CqYs*C**!30IYXXtMRy~m+!9zY4b%@o!DwBIy%Hn!6q*1yW zL0V z)da3J3kbxVmtxNdK0c#RaZ8-}vI?7~Y|(*OjcXirqlWKpp7FS!JQ$simH-G|SWs^9 z2O3%OC9=5IfwpGNv;%oK+HfTmG9XdtNXRC@^wrzw`j#H;nuCo>hf$@~Xj3>6c`C{6 z$L2HT3J9TytDI+J7dVkJ5Q)4P&YG!zD017*t3BL|FqyZxYyY-=g4vfoH<4M+md`dK zw7Mw|AH)6M>S_thypD8;Oa~hwND6=GnU-u)k{_yjHE@SOIV0tpU;sS$@UF{&vw_e9 z$ztSwod*aV1fQ~3cJ`HKkg^P8SxmYR_-1j^5O&~9@IFA^6kX}78wKcvfPSU$s|-0( ztds{F+#80bfXG{;j09Y?rhtK!d59~ks5Hs&qZ!|*IS^xHMB)MtL5J3=5n4|leU$>wNyU!guB#L5P}8H+mSb8Fx87@S**j0C@`PCc$y?E$=Z4h`^I zYv#3evr&)RqM=3h4fG}zmctWzyitsb)6?Ek`M+awY+|WPVuwP;A$It9S~PUa&e>iU z{*$LOK7{HIy4&R2wC**FOpf(l;?o94X3UJx2G^VwZWOG^67fC9`@<7 z7V5%2My#1y9et}Cy%tCn2QF#sZ4TnJ#R^N-C|e${pL7q23gbzm_~8N!hAmIaAuXt@ zV8{%5<3}2p&AsIEwaZwy6Y_3%yNBS-Q4db{sL`lF`wjvP=CedkZd}QWflkQcr|<8i z_?_`9`K)_7U6|B0S5|b+IicubBr_LZXPXKKbTjb-h~ZDY--7#G4jbz0F@D&Kb2R>} z(5p2nJI>*an{BS6xK`{i%c|Xf0XB*$pEzKxy-1Bfq;Gy`S+HQ|?zQGA-mIpj7PpkWzir(5S(xm^Aal zg|LAAJ%|?zNu>>FNjHdPfmJz5_EW5=)o|+>EBR{4 z;2k8NMvQitw`c^^xU8k&G+Kx*-J9f=dJMawYVtMgGXZpmhtdgE+LM#}$J1sC^?7b1 z6PzHJ*VvNNG7QEQA>$d|HI@;pwYYi+Y%wc)>47|fc#W;w9O#XB5sJ&;GV(LwDYS=mE zcOZqVDwQ&1jemWM6{-r#v~H!yF;rF81S3xBfBrS3uuf9Nv!y5(Rh^T-3GW2K zTCKpym|}%@g3@-vJFz@kcqdo1z5a8CcV9(H!pFv;cE&Wf%G^AZtwc#fPq_B%R0;ViX7tI@hun^BF;$g?`wbf%*KnV*fBg!9tIe`G zDsJ7ckYDMLyRf-2?r)zhRwmQ_c*;4_^rQ=yXve0&qD4*&iqW)h=ES9vy^#|a7m4dR zZSa}GbfS0JNvaxon;O@S6KjiPrW$j!1hZ%AEEBSLYB+eH*SW4sUf&^iV!MTx4o9Pv z^Zn5|OMLZ|vxQG|Q}7!4h=nqudbZxyNtCxJJzMov9J9;K)){k)1IRZQ$*Zun_vnS~ z{?_pH#&kG6eQ$5S@@w9XyTlLjAcn-JH97NVwi2mMb5kDr$r8I^@s;Ft9dZY@h8^!e zFx)z~C3`tFgQlmPd;CZjKbWrg(7U|)A#c}~^Y#RJDb&}dweRbagtVaEOwWvVcE*F@ z`RZ{B{ZeLet^ysF=_@f#5!Qx4Z9E@cq!)Sn9~p`fFcv6GO=S zqswDDJ_n=V&(qX@e?&c8g7xUVDAObKD6(uj%77cHX*ZxKoNlvBLZa zRzpZDo*2}9a=Jh2^HzTOD-c+%R@5q|1buEUE0Zew*2$g>rrj(qEG!)HbCzG|kXN>* ziGZ6c8-cXVWBdb{^ESYB{zaRtgf~a@f!;m2f0vQ`Xl(IrW4g0@tSTwCsJHykamOMA z;x`2IG1w^tIUEw(sx9R0tq$tJ9bx9d&L6g^s(@oC-;%a0v3$-)XaIw)$a(rcrM+XZ zanEZ{=_r`0JsGp4qU5;><8``a@`E-#27~I%N6B|vJDs(0BH*qp5NY3s)Za;Jw!|+; z26tU)cpZ)2q873@neCZtg&Q#Gxf4JLL#^90aM2JZ;jxjgFeL|^dwy?5UY2eZTSb3c zgG@7D@xrd9|C#hYj&rR>WDS*OfPyVI%LSd~fk3O>NnX)=ITBi}Mppx=Mjq^vTQTJ$ z&}b)O16oS128ovNbWATGzryHTL_!`ymV$~t!xpe}R*gqKv8I$15ooT1h~}r=`IrZ7 zHHrY&tL?9Z^>Rd!2~jo|MXTo-x;CbhCo_!1tmJ|QW#;c07v>{+g*z>Az=VU_v}_sWe911t#;u%Cya#EEU6`uY(BC!`&Jc22Km*5j!Ml#)(=58A0E_Ev_>T5 zd%G!38ubjQkYjC}C@Bq+^u(5UKOWt8HK?%pNXx9T8TsQy3@nedM+F)-ixfk>x-j>b zqXuIXA%;q0i4H@JwW%Ahgc=>GZu3ztju<2VgHjlmn&gEDo6jD-+}$H6&t4k)mq;)K zsFqw^u~Q7tl3ZE{ZcbsQ*}X|)pijjRa0_yi8st|Co&)uqYaqVnV>}k|kS`)%4jUDK zga}N|<)B5fs=qD=C$(pb+3Xv+m zi`>z>Juy|Tx=zU^`aQpaSxe*z_h_a@UUV?t*@gK^D-6|87B~bB=W3ajhu7UFHdmjY zulnWG`KmYeCsPmq!l51HGa0!_el+Y`>JB(WL$ilN!Rtz)yUhr?2f3!XS%pBi@7szq ztj6@lMCMA$B*Zbkm(d?Zbll)-MMcf1(H5mm;6)^{RSla>WU?_Y{>`yy3I4MSFR=9l~MlgqJEurYP6Kkrk1MMUTeC<$ z8sf_ITM!@a?LEb}=bUKCJuyG4PbVYcLYfEC+}TZ06JKsT$AZ|I4B02 zlM!^2q&?MJ(sfUzo4h}y_TfjHWaP6r1=p-UUG)7U|NCV1KTIhH z(+j%=duwmm3yn7XD+``vu5u>o85Ka?hZo7sDf*`fxVLYkl=lV|cb;x_0bm(F&IN#F z{1$`@Grn|b#qXZt>~MQ?t3SC9CR1h0Pb2$_(P+3k8BX|LF3`W+E?&I}jfiT0c_rCX zY!BrBlUX+<*nm2Oo3{rA>m~=*ch%(H@o@V-b!A*l-aLft4xkOPf5piqTuz;I2xc~B z(9xuFS3b^yJk|=K!X?-U^>G*E@kR3UXf6R5s5%9V(+<((xS#`?xB2T%8;hrcolFEIA zfQM_)ESAcjE1&2j-gaa~0p(hm<^#2wR3E(Mol&qAAILj9HTF=aX;YR4@#T$g%s|(Z7cPNi6#Yo#@gU zHOkJtGEIEvTjxdrnI_Jnr2J}yIc+3U5%R060r^!pN&R!)Om-rj%3-tOFY5VK93K(z zKuu3ugT2zoZDw61S6LRFUif5!6>cPfzP_eorG+a=GV>K=aoL@rqDeF*Lq#4+vPJ6H zOGuZz&BX(!{vf0|Wy3EpTC|6n2{wfym;ii(o@6H8)G5UO5nOj`sF^rO64=740^pQg zSP_L84X~8|Ve)uz1*5U^*yz5o5Iv0ik?aZw;6y2{U8}(Xhm9>56V8l08iJ6~6WWOJ z1Mfx+?or58g6>IQB(^yNnDSRD=q8g&3FrckPoCowRBh~>nQrwTNa29AKFI<9J%1>; z*rk&o-7|(D8T$gzhk6~=8dLH}y|)C%Kn+f5ZZ)Y22V94Yxj*De&#{@)X!f#fKunQ` zgU=t%fhw+5f*gpTd7ko?Dt{A}2z zymC+F83l(`dGkuuwu--bh=-KB_V;a2C=7&NDEkG|YIaTHb;2Lv6^h>aRctgO;g!rM zXNrX;;{o_xiZp<_(&wfu0wN8)Jh@rq3na*96PQ_j1d|Xrfhm_MdlbE6vD7w&7?4^( zR3}mh-^s|4G|S~tNec16$<}T5uQGB=*iKpVXO26sqG=QI7Uf(JnrQ&mT2Z;=Z5erS z*apRv3FWkmxb|C7lJL$eP&ZV%WGcW+8Eq3!lM(cfl8~@c_FJ7$l@OF^BSBMQAW1BF zBJ;RX#Z|g~yF=Dum5QGu$qptDG+eusW)Fu2?cw0>Ov!75j}bv&BOpv*cAlhWTuo*p zbpN0TR3(}MS;1kR_x z3!;fZaE9{7A-Z>`FIa(>?Yr=}{g(cPVtis}ytN{q*;5TX6q-lk+N!oT!Fm@WoK-~7 zLo-24b(Ek7yAJZp!9kBomkfI1>mT9aji5p63z5Pb@ew9~P&cIhYp8RqHoWpd@?&fJ zdspm{_vV>jx%za;H?z?2MtmQBD3cbk@flB6xF(M-W z5fHlotP~o$iH~bFFn|HJV_0%0JaiKTriN}}Az*W8=qCRy>k`gy^U?K*Zv|f+1qJkn z3OBLrQ`<$pje>*({WXHTIXFHQn*Gwemfcbz)ZDd!FXToOy;7co2v!@Poh$G9=J0 zaP-w)nMJ&|;{TkzSfdOHEVwoq0ycuo60eucf+fEB;Uo2}8y>BZKMZ-R&(3H;M;Tu> zus=Fzm8(sZx^Mun>ixl&s~PgeuXR{esJmd#IZJj_jz7=?3MvddrIHlrDCq=z!Bt5R zx~e|zCFj5=%7+3Z{X4<8M9I`@3*s;8vPN(Rp(P+e$EJz5;UZQvguFHQ$m1tn^Ui3m z7Sr!~Rh~3ef?z2__wN@g^v7J!=uN24g za_=hnD?&`D|5sEsL=~9c5whX-A^sw^{$FR8v@Jy#ZYh$l2iMl_KUa(|c-Gbsj6-hi z8RUvVf7;(X14*v_w71k-JaN5lVM<9|dqgV533}3~iuzYf_acfMdgi3-4hmQCwdXNSH{};p z-IO&wX#;~BQ$ZvY*l058(7K3TxYU+t@8U%pduJo2`LtCgpLI3EwHA-IXnL@^0P=)+ zL#i*(+Nx_^;egysBfsxD!CfD#ZGvV3A$OB&w(%@Y22`vJIm{4}0Rn43cErNEq8%Wr zO0Ihb!rAgOIqvYv^?jm^O6TH|8Uh3I(H;r~Pz8*vdf{N*q(0vvub`J0W8+Cb>a^Kb zQ}qi?NI4>7xuV5^eqh>;-Oc*35fZy)&Rk~3;7uEDaTc@WR0P^kDZwSsD5(C@f}eY4)QK%IMZvD>sSK}z z6NETR9sarer4C6j=o)=NCo`hJ954o}GS|pw!B{G01y1J#Pf?@hI1k7ZbVtOQ!U4p^ z@=gmBK8?V$Zk+v#_ylJ;U}xTDBm+`mD4RuhTS4cQ7DS zjC{`3m|R1-M2yxHg@HTDTL2?H*Ste2=UL^*dZIGAsZgmej zbIQL?H~kJx5_UW^51IiddSq4$R+2~NBuyhAwY<9OP{?+oG$m~^RP{<AgmU4xTOCEhX=~gUub({yMjpV=!b6r<4jVYb&Z*BGAp>;nf z?+RV{5^nW^ei2+^EC$TT6EbFkD6MXCac2pE<$IFj1*U+Fuo8oLG^j*#hLTz8>E?(uAGH3!1D1#sJoSfW0o;J%CFO++7pV8@;uhP@VJ={% zcpsT?(H1tWi3KrOSVn?~2eZ^F)lzJnB)^&zF*5}(Phs}sF(?J~_KF|1h!V}YAulqe z#b7X{lA8aGl%$1l;=zOx!udiHEsbz!tVe!5DTGsLlObIEOSiPpP0*t~9$N_A#0Qf@ zH+7}d%;Kk#LLW&Bulprjv~;GN5&{Y0v@-U*@mpNiL6Q~%3HmTYAhBp-2qf3BJvVcP zK%Y!X+)5&3?0~Ic3F%uwo>svif-t;9@>4z9U@;+h5X9jI4`SKm;6bhw_nZr6GufvD zXpVWFX+7T`5BuAv3k_bzobm`8DE_^O^U^@x@HGXm9`-n5XYcyutW#LlS-xAOpiYsK zz1yNvP=+rv+gBn8gG!U{b{vz6yrltp(d*tY800tEkW~@@HA{N4&02V(WOmI;Pe2en z#O=1uc5|UPBmaWN?X8_bZ_i6dJ3Hg!JA>Rlxg$h`|L^e}`Xe8w_r(yPnL&?AhsSN& zH1=n%c5TzP@4IHCc`F9Y!aU0c&5fsg_decbx#{*6+O|J}z3V|Sn)dmw4Q^Dk(X8!_ zs<4tj?P8s(OxC^2PEwIO2G+#r52k~`C65VOX)xBdC7X@6L;ls(x73^YFCC6XD-2x! zl-J@nbi?aSdS6X9-CLyp&2hcI5b_SMg{lwE5ABMg?WF2xhLBIT4al)(P zFmZG1+}6qk?xrF`jMh(H@?aM$q^$PbyS)0rG1tr68dMX*7LcVjUz>i~vU-vqMie0Q zW_5<%qXxtC)s5=t@I0>dBb`8%jyyHE|M3o%4K&iI2}K-{oT?UW#h$nao`t46v#P%r zQF_;F?5W|X=q=L!Vud%HVZnJ?J%QY$+yE##dG!>OOsB)$5uiRh*L4Md$y?BJg*WbO zPsH8z89W=eZ=LMPdJD^o3kye%)-P0aSo$m;b&+(|7HD=C$(*anM-(U6A5otZtoM4> zGp#^Mu`;#^dANm1K9CXn7$nOPwsHTt{DydTQQ%VWb98=bO5Y6+Sg zHQTx(cFEs$$N-sAZQa9EJZ(jsj`2Q`N+O?A#o^)pAtS$l%<9SM{-{5$BHH}K-LcE7 zl@C5NP3OoKSj#lt&dBV@pcmE2SZWLw0N9@Ly>&ZW#5(FY_nZ(a;rIwdJJ`$u)ISB= z-;=vuqUUR21(l8I&hD`);f=0w0X&0f2itd0WL5Q~A6p!tyE*onAY!3H9%tt_@Rbaz zz2%FJRwR+n#T&BR?**X5l8dwh_@x2Lp9QB7#hQkTvL1PEZ-@xHMfxyy<3v7 zH8pA)xEk`89b%tfZn?_Hg#~uzXjrI6SJ2>Qf}djO8PBb%+bTQF_vDQHH41lCtPGC0 z98Jl*q`pCMNtW%;h5zKAvUmiT)L<4<$mE9!8#6Dvw6T?ZHvthOcdWt0=-@7M+p)nt zqV86~!ZdfAwdyE*%7NrlU5pb^!N5ePd$qcP_k<-ni ztBH4cN^0s4B3^LlH(TLcqCuciAu2B(-eo{fScy8a`C@s*HL))(Y9V>!VmxX=vP-G5 z&*EAT0`n5nf)KD!JPQInC6)ytx}+xB>RI6fT@kuR{yw41D7B%1UwKNpo$n_=&>MS| zxRq##iCj~$rg#8`Tu=x^NV-*>vz=`4%p$tFD>uZ8>NT7SnpGp59zpn&Hof~tP;|VJ z+YOFQ$WEH5!c8{ku4jHcXpu(;0mkrmdRLTz-`q*#-)c3_?Om;q2d*O- z`R^#N#x~W6gJZcvu?!J9BYlkXcI>=i9C1N?lAT3xve)gq- zY3pCk9OJ#@*6Z3@I^d>59N^OoA=m%}a?8BO8wq%)A=DW_U zo$wf4H{p2Z>rwN`4yeP(`Y;G@LH(<8KG_>wgRp0aX|e(on@_dVZ)UhXEyi2LVA!7) zD+8t@^l&?ztOFKDc7XQyUQ#zqd621pMR_1cXOq8nO^1EKifJ?AtS&DWzG}g^(d20! z#K~1@4HlNL`C$jh-(-$yvmSfoXIJVjbra9EpRYVe$Ua{ey-(ewjsuOCJ4xKe|zS9ED(OVh2!KDg401&&dB zUTdUw$OGfCPepc70fArXAR-xZSoj%0MtVC5`hqcFD$oJ)i2GEqD~GuCB(uv2ptlpE z)l$<#JkNX2Yh43RU0Tq%{iJb;-hm?ZsqSN>cU7?2CT7tJXWCb5w;=D6fH}sM2V(bI zw;#D*_o)fhpR~QQ(7i;Xr!B;o=7jG@7=`y8$<3|ba94lq1ZwU~Xl?a8J zT|*wsaY)q5r}h;)fS9C$m(m9 zJtrlXRP}N4j6i>;2+wz_zkP3cj?tf3N^lSVDgP;L?OzBb;3$tg zylJ>v<=`*r;XO--JiIMEj%N?(mhuS|f3&uYczDmUAU2;+7ia3>9iLB-g`F~=fMSn^ z`l;TULX0OqpTN1zgOrF^rzZI09uM+0yREID)l$=mYdu_q`z7)Dgo-;D;Yn?ek>N>w zg_!UJo=`$TH$_+a>P7*2 zA)sF={3=6^6e~>*INlm%B;cYo1q`gRL0nlyrAdY#@nKM0z#-_+>M~-$AwE==E+W)O zT>&BY5phNc8o&ocXFi1g5Vrzt1N zcCpjQxLy8xTiVcVBM(KDrNq~djSE01kL})8g;<)sEp)(25lcl&@AB%0aNiV>&}?!d zKugIZ?je^pf4H+)1skUcFy*? z@Si-L@gY=y(A_5Argg6=#yge!m@srhf3i7+#}zxgO}_1kaZsJxV|<7wp^7J|HKH>f zD$Ba?AD%};Q*>89CqI%ZSK{il@3+{>j3AudcRZIxa4VndkQH9of?lYNPIfmMX$T$wTAx_P-M6py=AdKpb#G5GArtN$Fv+Ty%O^u3!fS8kj}?xkF|flN`RT>XH+o@=V%g zzW6xg8<5H)Nzq>=spSpP{pMzJtoIV1?l+oKP2C_>C%>*Y@?in_aMX3C%3GDZt-Z!p zpJq2@DKs7;kShlLX@B#~&UmX2k#%qJ#7(!~xcnoH=aGBK;gAzt$)vjn}igJ6ltPJJzHdO%Gc)Hr}cl>?al{_2ZjuvSeT z$z9B&B~{19EB3N_iG0O97{W_#)ICTh`l zv}SAio+yoPfP#l?<1r-pKqW*|&3q=G|9Iv4zCGREVz09&TkAj%g4)flYe7(FK4T6W zEYR7#N*}P<{c;?dZ28jJA>ZnddmOi?FgmrtnW^&WdbA$)>9Q8;!ahc2R4rxJMrO#7=O_m^RiM-q0?jd+{ z)PvJKYBXxlzJq|i)`<4Rr(bV3$CeShb%YySL=eAYdkE==m0D=RwZoKW;I zl9`K_?l=@=y@Yly)3~nQZ^3;oXA2?cvd8#gFV4-suo+B4nN~bQbQk?ckpkWt7r}V>#rwsf355DU9vpAiG=3S} zI)ghp$q!v#vcm!P(v0)@8H~WVCQq#eBZkIba!-y8bvnY)%d!VHufdc(CdY;nK64<4 zN0NW5$&;>Dc~UeM$kZXQz%@`dlm$X|(oWBcn#iljUA;SeY9iiWt?z|?k=99kQnBwt zO7&4gqXz4XNec_u--DP}l~l@bJv}9fmFg3G)VixyQ+?voh^anxO?kJB3(@}^Df=L( zI?e<1Mq*|)Ukg3-&-S;r`Y;c;KL|-Yd>JWyDfT47YLs+?SQc27qhvqDidqeq-qMn! zv&Yut(}>Xy^A?SO8ke;+oJI@LrF)aCRI{J1aqi@6*k=Oh4iBXhs-gGl!WW+r7U#(q3TCkpHpKc?U$N(;ngS>;>(Z~g=;WTmJNDK+dI^E;42 zR+UN_vc|u@#R^pgWm>n=YN#qcjTow`Yf2-I3st{{6xKafbMhd=)7MlH|x3s6FLGsgnzmwaN*R(ElNO z?<veVIt{-~~tN!LpQp~Ke{JOXJBF7GVgD^iDdlUMfcOiUdbXkcDN z0pXbx{oyXzprMoHwPCQVK!1AO8wP{?WN6}Lb~})R8Js&!LBE^B?ZNPLF-fUb&8}W) z76{CSQRL2|J3rBpceNmut(`$N@BH}AAa6tHEy<=r_RNw0HgzCx?2y~BEvBlnaKCX4 z$JOp%ze3<@vuu6}=<7l=@+%#37dAJ>{q3{G%4FIfPdP`Lo^;_-al_M0{mXz~!d5xKuVmKG99VYv>~u%82UOdRr$^-lFtmK-FxIF*gmh*i1&g zxkz4xt-VJtZ1=Z@r#Gg<>FIlW`;}kwcHAX?kOwg&KCQ`_KeLt0F-8zb4O@#x9?Fu} zb;upq8g{(@z;Nr_mh9#9)NzBRH|OLdUHo9W;zRH9>W92tTh7}Pq{T^fq!*o=&m`~b zl7zIN-b~Moc6P>t;rZ%rdUSZcEf0f!R#y6T2_#`t4Dx3!Tu>W{b(GSfNHkWfy_U;Jf`DwFbe)WP5t*r)WaoMkKT(iJwlHnXWCA=(di${ z$ld8-&Bp!b`jGfNIy`fx7#Ho906w0PUu>6RS?D+SbqvGMzP?MgTccT9PcuOPS{8QR zx@fqMkAnul{0LSAwxVh8SZv($ z+EY3TrfN^dEUD4W4uNzIn{1kBxkVDUUA~mXf?TBQHz0imjr*twE;Q zlK|;Ci+Fm9Sop zD6*&lT>LL95f?K^w7G1&o`dL;Co_!1tmJ|x%545SHD+7d;eZoe_7vj4_qw=ew}L7| zNc(HG3*R|mB&=phEsE&;GsK-6JT^UK;zCNH7GbmRw!2Qw-3OTv`au z2K1S;3KMuLhJag;o75n`TJRjG=UfBvH6PjbJaMidGt$&YfcGx1;hnq_q@|V7*5GnM#$Q`}g6I1A_ z>y&Jw-}9TCwM3q9k7jD*MF-=ZU6`-5!cYxmfkWVMu9j(ec-?(s^YGz()i0;cSG}=6 znM#g15gg<*8M#S*H0)dI4mdxOIrx(B(Yxmkrkx9{7Evr$K~g^%QoiOiLh zNr+>7FQY$-=(xevii(<1qb*9Cz>7#?$&)m+&cx-_#z?t?$gm=2xzY zT{?)(HCx-QXu{T$m?Pa){=6{b;=nlp%W(S$n3tWtZ-F{WUkX?lo(J>OWG?K?9f(=M zM07Gvy4E&s3hBI~n4l!hljbCOrM~)bE0o7XLoY+(M~;3@V{-`@-Ce9y6DrfVAni|b z!M*R_?O>c#vV&B*Ai zFhe|-1fNt>GH{K2J0mB88B^E(>M$SnClF^I0~V{pTG3or#Aw$gp)|D3XE-Hno6X6d zV)kkHinV~9c|U&Jd4cE2%P4AfMC@a>WU?_Y{>`yy3I4MSFR=9l~MlgqJE!*MC zM#ySr=w|9v*?U+=I&N0vk&LW+8;rUw4F>HeIcDQW*Ym1-(Bh9@br9xiU4%dT;Lj2G zqkt0ASV(}Sqpsqc?0grlucHCR-|2nBh(3%U(g9qHI(j zu)=_`Kzpq`J`c?LQ+m${~S#B}`Lleb=vTDE&3$>1#cv^nx*z zJJfyj2fwrl{WPy-zJ&gA0e;B{xO@Mb0NaX=JK*SN_~Xjcu%+q&)*(EW25AiEi(nGv z%L2EyjJ|)qATOePM5R}ElxlXnD2^e69KSpYg% zqg<2ecsM8qo0AcAlcYW7PT$8#HCNd(0}fK9o4h}y_TfjHWTa9+w~_iRe*9mv{&dki zko>PmAkk>Uzp~&-<|=2R2{7tQKKbw>xj9Au6an}4ZItrfpyJNctu6p8Z-}T zT*Bjvv%kW?@=ca^_^q{f%wiuiMiXI022I*GR(Sy4c_R;KwtttNRtd-lyckau>-OHxd) zQYAR-#LTU<;FmpQ#yBedZ_-?YR22-XbF-sS$qAGRH&N$K+7eGtUFVZ;;Z&#?7;l3j z{~iJ)vFsakqDyPkC_DGcH1VBpof`$XS}y*i{Az_cZ6s3>@~f+5epSRS=G)IAfs@of z=gnj%(y1IaEB>OMU&Zke0T0ylv^CgkHS#LC%97_a;?pEp;YJeZ>uWkzT6n8TvWJpv zkvjGg(j{+mN`B-(BMZAk7S}qx1*jrb?{W96FF|q#{)+rTNOQ`DUtqLo4>c2P3Pmsh z_y#@6OuVV1-TouE?$%HOUO*P)~XCi@aGO3h+ zF7Wu|IX*$v#?G1PR{wz%4oK^h9Pl@!4+R&ybP}X{#tA2af^VVCmCJ(Xt^99HGcD^=Sn{^lVbQtsN{w?UyW5PG5P z7fh?!HHp^=e}o6Jz4fagkX=|Kyg>ORE1M6d7j_GVx-c0J!0%F|0o0W~H)RnJY3Sw2 z&7xW8cBag5;@4v96-z)hI(sq#Q!Z2XD0;_Ysci}|AT|GhMCd&EPDYNTSuT%CQiumm zwr;b3m62P*cFOuYaol+oO`DLnDCdF@OAP#$ye%Uy4%?uZGNGK7ZeR+<&iEEl7YV%c z3e*jiE}06jB$AfRRMQsnlw$fvp_=?wCsZW_W!gy4lo&`7OPg?*9IRWg1|;Vn855jNzEiz)ZovXM+8(QngUtLN>>*D z7Z&WwB4nX8Y3~BoPc$-Hz z;bVE6M&Yo5kyCjhSC~)gmR8E2#mfhjW7zpLlj%)KORLXA;Cu?aAetBiXDELhqI-Ax z{1tfFz6+1rZ|Pqs#wT{hTPyOJJ=MTNp?M^(t!ir%>0~T-%8-{L5kU{l1TocVHR$=} z;GjpPO9nm7!iEyWEE`KA5#f!XLF)^V!W;1sCV)^kr2gyFI>$G&(C|ilAATs47P0Xe zPgb}lk1%foUt75lbHRfb8G$i{*r7-S#4Z3Ug@$h8<5~^0#$nRA_{S{ap_?EuHFOgT z0h>cZH~DW__i*q1N){=HH~CiZ)lpDDf2eR1%RY69^1u$Y`q?N*Zp3@<_XcRoQs`m$9u4c#=zt&+@ zq3(h`=PcP#IsQNoD5x;-)GBfhJk@7{D>;&)sy^-|=fEe*hXN%1JHfZa#umh1)Mbs} z4nj*nf{slSZ{@*nLkt@QPu?1Q!8!;mk@)(5r8vsv9!pU1_5X^hhNuG5J3=i+itz={+8Tm!NWz~%t{C*E{mnCwzR|TJ1AVWoe?34d!dEoF->>o(}r7|eTL%Hh{SV`MnUzL z7W~{ZpR}E2^O;Bit;&f_`xGSzah5v#bNNdhl3>s^`hrg81Rq}eWQkHiWiPkS=+imD zQ`B~LoCo9yn-g)SZ~(2x^6;!1Xa6ET!C4O2nYXCrwvgqZ=ZW=!Bxhe!oknPb!$JzM zK(tzJXS449Sq-f2esaayUW!pY2!P#;j@QK@19XC47j>0h%^_HymF56XY5uiDZ8iyf86SpfMIsuF&&3w2Td~;HaUPsDk^|4nbzQ|YHg&eYwbh4**8M?<=|ES$gj>C! zUj&yJivct8gp8RWN~@b(+*y2tN#2tjFEHC_gq0Y~qd_H_GnC9yQ)kYQ^7A+z88m?h zl)(>qPEPJ0Pn%_n=Y?2D`sJ&1-ZJ@zAsWV&)XSIIdwV|uv?WmK`^bcgwyQWdcHp%_P0+L z8oZ1-NPWEn#N>~}l zqv9jq?KmbCc}oNIqSw7)FvxGRA*&<+YL=>Io3-%z!t9!ro`4{@^Y6CJQA<88BmaWN z?X8_bZ_i6dJ3Hg!JA>Rlxg$h`|L^e}`Xe8w_r(yPnL&?AhsSMNKU*nlUBMJzV#%eF z@4IHCc`F9Y!aU0c&5fsg_X-vM9&Mp*_Mq>2P>iO1zH9%D&Q}k}pLVfMRVM4+Whbe~ z9Rq7(^asE0s!Z;tEz9q(i#q#{ETnEy4T^946W-`JMJKZ~|!=ic?YVb(^@C%sm$x;jCWbAb`Kh2y^L`jnfY6)O8G4Tz49{0Ls-wg6 zxYmz!0#!Oik1jvFOPhQ2vdbStyx`Dpw!*nYgFvN1R9-y1%YdG+5_Kexu_nHy zO(~c)EhBGSj7Kd`$d}Y~@!K@7( z^7jd4MyU-A{ECko(A-{HJw?g)6Cmh~y-M6lG{i)%saR7y07EV)1Tr60vuPMSvxu(l z$_??NdJU(7X4MF%M-V=xMWZJlLDBI>ZZ|kKk*||!;X<9SnPXdwKMqCB_>%Xc1R^y= z3W6$eGzy~8ESqsA?g|6>HIz&n8VUvG&>(QRfj~h>95M6>f>#Pgp1ecToc{S$ORB`K z3$n)(dWM@TI-WG$%FK_$Z22ycCVd&XwRb!!O{zLM$twC48hUlfM_eN-MYe1(-q{7v z)e5WgmdR_LmM!IL-PoT@<&NFut&h6K0XwMW1?omKr03_)_MsQ{9w0&j z{T=MH#kJ(pqQMFa=Q-Hg3|9(!KkJ;P?4U>H5L=w4Y^Fh5*fT++k{It8qiH6wx}>M| zkNB!%EUDchVlFb~cR2}3kfD*BsVUU(& zA1^qe_HQxpg?u@4jQ5gTuWM`Ryw->P3G{C`2BAUkirO#Ho2a-c#Thfu588~`hc(Y^ z$ug_Vcb!=~;W4^y!tu=4qvn$x@SIQf2G=0$8Dg5O0LA80?ev=&ZcmHxRxud%r^U*E z=?FdC4kznCW87M8@UsW z2ZX(lzsVfaWL#9RKVNwcu?zQx$?Kx`shiZHq}@Xr)e&fSY@fWR-7qzf zIvv1q_pGdaGjl;KfQ1f$KDRDhsX^B(y0o#S=~iPOTuzE8B;{` zTGzl+mliZ`KWSW|cc4gps{0tZalgP2Dx7Iw1?lpW#QP**j&bFI*!|Y+NAA~c5K#R| z+bav*OEh}gLX2rn_@z#YlS2S39hzu~U_*a_6!nbNpc-keYlVh=9qZ&;Bb33J}1 z+gFmS@u|$XkYtt)8Io9v)Rr-33$x)$WJpq@tL>1a_NfxK5}{DDYsiB+?tLk=6^}SQ z>QwOOT&_Z$r^&_~*fnyrGBmxM_w+|`2X>7pj03y2T;zdWUxB?E(EAw^v&hGy4zc=I zWpWQoSzK?1G)h+^NUJRMWMA5e7I~1>2*f$aY70gmWc4-5o|Ewh8T4n0@O-EG+xM2| z82yQ*1o!Zt@}JUHZZy)T|H26u)cm{J{)JFNBEa~3LY0HRq=z>Ro9lGQ!`o7$ojssi z$|qEGX>A!fpHOE(Y(AmB1{~f=9rxq&2^AG^8z}P$DE4TmpX#kC#CYQK37p$JNQsDb zYPt{XiV06@D`>UUbmF!ep45oK2v2IuMTRH!6=K2@ctRC<$&Q3!_kH-slX3NVEXE9bbU*YcFn;?rNgMwYP2aFi9D6$_G9yza)s190)!&2 za-NM{;6%zmB=TZd`4M5J{-MZiJFgD7$&k0XYyY-=g4vfoH<4M+md`dKw7TgRY?wk7 zI`cZxAu=6ogdi#Wp=VmMNtt#k4{62i(g#~9-vk5T!H0KU4x9~y9!M4wVgxV?_S6Wf{h@*m&IRq#^9^>fPjK7rgSgC$N|OvfLckqr+m^V1L(rjh>WBe{_)uB8h)^SS1%%v3#2F!I0N)v% z`G9|0>Z1%dynQwhuyR43AfW8_IEC${>%xC>ZxBKSM{@!gH8EU$Zy{O)qgzt6h-!S*^J<97M)ZD~Wdr+p}@ zEG52vY+L|Bd2IK#D#X(4ZJ`5JidZUIdY4x}g!`t5gl3Zy!W7-=N*-|!xwQGio$X?> zIU3$qct0TwDo?!`i#q3XYv1!2oLh>F1iz|IJ+4>n0laq(4e(p5d~GeR8qG#MZi|K% z**DOeR9FsA=6<7v^*Ejwp>UHDI)&iD|jKj?0g zZ_~Qh6yu%BeM}g-p+DK2!sCh^-X=c;h-R&GdyEh9BvkPv2CT8Z(f)!G@e^o@?&{~{ zM^fcVT%Gp)7F(GSgtPmO=duWH<#Qc!W3pR;cpB5Oqy)OuTyZoY#AM^Z3&Nw)E6FP6 z6<+`NTd!5_ct&u7{8#$b>u|Zx)O*Nkj-j?PvWvHqgj!6`Y6|{FS zNEDmguHW4&4p%CO64|Gu^srbiIz22`sDObcu83Gpt(^RGhcp_KyrAlm6QS}<+GT$3 zIOH3U$|Fh9UnMEG!ARTM5N&X7CdYa&@o9sjaaAfW=;~$rn-51_XR5qaRp2%~&2Gw4 z?kfDX^#SsosB3xiQLK1o+$1nR>eoRkmLiE5KT4nnS3JJmFxTVbbE`v z&Yo2zUI*IZfAIp>6;hmp)&e4TBo)wr36buQDmuHJ9KeJ+O$_4OD(?8UiRI%Ndm z8On5Q$#w`ee8vxlQW_LPpq%%!Y;v~o4c*d=;>JA>hnH6vQjvB3vU=5qWG?Zz@Gt8iD zT@x4u|H-3B0dI|qV7&a|{o>FBLVX|)j<;(Xzu3I{W=|*iq038lIKW<-aXvqT5g6Cx zskLATjh$hQ^{OSLK)&Rj92@F%grk>b4{Tn8DSJ$g4JCZ$K%Si(>ts3kTTPzyMOB^@ zjRi7w2rO_7lnrHpkR8>KatM-h+d`R^&x)GJtH@owJA7&)-UQaSP2Pu;>Z6864b~Tv z78bC-2PxsFQYpjr^pqr4s!vd;byuyX`oyOZQ+?{1sj0p{N6J13s*dvjy^)xCQN9*> z=%4LxZS`RuaDO1}*lF_Q%Sh=)M~ijVQb6a9VDMdjCPn$ zYXsD|tfk>JT8J*4aC9*JHS9A1bccu1302yYll#ZhW(xI1A=ZX@jV(DX!(dzyGM?dG zV;Qkpi-Qo@a(lJsg9Yo!6NuN?y3K*!h!>%_3@#&Iv|v)&ojHSQ<1_gn(*1#%Ng0c= zACJ+A0(-@eDL9GJ0&!VZ`BtGx&Ov%;AuC0FNU34xnBRdEvZ_?dkTt%kDORW|DAT%? zRzp?sX~a-fT~iuyTsHY@NMW6%if2nvE~+{wffL>dg0)&9&D`P(4_bpjwD3+)+D>>U zmPZTkUy>tZp!Sp#rA{tL)+#4NLSATQ6Be3$3MmK@WQrsbau6gw zfEWbHH7jhi9s_M;p2xc;ONxtU+|T#N!~XW^BIslfo3VH?>3V4(boiQrM<6|pkK4O` zIpE_K*0Rp>y&@lYH+g06&cuA6fd=MPl*+XM-cBCwk_{RLDn=8YY4JGRACRTl0yuHk-T|N0dISDR&XR8GZr*fjN5I^-^FZjAfeXN#4|v_GD5 zjx;^#!lmMd$&yp=Z$eq}3;ibdTXytr=bj}iAJ>_iS6WtWNhCX7UjHsTi zw{;TbElN)YTg~5b`ddivhpe&yG^9e0Tz52h61)(w<;x4Z9E@cmF}ETwB1d*o>SRkFE>A!Pp1IeMtNs9iBN;j0=4Uz%xPRe3td`jQnD|6w5-t zdF)sihW7PcvfUcZ+IpG^0?@Ls^VUVf-Buhl0Om)q8bVs}#Gvkz)BRDOxAMzhfxv3D zqE@f(8q80-|nAP$Lb)fV#hRtNR)cs#{X0OW^lsw&_Z z%D1E~ODv!B5$eHUFUSgQMbqA~*tqAlr*srd)t-!5QqhCwT&Y!vJ^4YK9)m%3=A-1h zt)0%=IMEw-Wy1`T?<6%_;uj=?yRJ06jz(`$3)!2@_RO`y4Tva0)nYdQ8-`l9XW*hC zOu}O$UtvlPIFmW=&B)8rtzxU_Z)=cgSl813OnM*3xmF|6fl3<)RN9^76}^`uq0(yK z_eZ2X*d@1O%15BlPQ(VZbQ>BZTEf#Yy@0q9cbDfQ4s)tMBM%`3c5^@ zhJlV{U0r!DY$D(5;-1|Ksth6RuhlMm=Y)~4nkBVFlFdbowpk+ziVX6r2_2Q1hpZoh zY(6~ju{84DZc39zJp(G_SQ{rwN`oXlu_fM521ZT9*nFgA3zhzO5d+I3?NNb-%_7B6 zuP)5}<*30JMTnu&SfayFV{PgNtUicCDdrT1!{=CAA@V;cg>k7#UWl;y7?78{dj#d# zOJn~M35EdGlB+9riUC@ZOAEn0h18Tj6+^%+$W3aHUoChJ)N_s_zUJ&8%%@L>d=dF_ z*r)&`L|}3*2Q89SeSkSQsXbfd;X*LZ>oZAD{hJX?4daEXhMa}1n4{#LyTaOlfswZ( zXIB+mm+3<4f%l>V15strz;vO8C=Abq>J5XjTu*cx6Rw_nUDdb|t$&Z~VAvwohnq_q z@|V_9h!px=7VriOuI;m#_Ne)cLA6_9s&};ew~+Ga0!_el+Y`>JB(WL$ilN!RrbF-FqR>J;*i9 z%_;=Cecx7`Vciejn8;j7nS?kVAXjddw`Y9`-Zay{ zs^1C7A4PQB;A%xh&8X29rA^>PB(da48d_%}SF1QMA`D`4&DM4+ny~dG=16yyKQGL< zIB>2Curm(>gL&EM`xdC9^re7>;dwAWO?Gp`uk8(;jFYaljhjL`?6Kkrk1MNfyy+tOXoxG* zZ$W&xxAzp^o^$H4T|C^Ap>vlrC9?{d$`=MgQ$e$5Pp2q9S$7*4FcxU9mB;6SS$|3| zTKm`g&^Xf~N%Bz5h*sZAHf&aP%|$aph^)QuP4q5FShG`xZ=~`(6$KTsOnvolmi38zP1DM_99RU*k$$M88fDYCu*JL^#4vN9%WCYzL zX^*+n_ivaxf%TFWwi_vJfI~h*+UoOzU+%8_d35|$qe|aU@Q*009|C3oaJKULe zsYAGVdr+`$a&S+-YI5&*xP70xGOi|X9#Y+bnM~R|qDp2f4LVgN+2P8eTvNs8%J zssx9fn7Ne}{IZA47)PZ^p?n!g%X5KxpyUL~gqx^yCvAx*sIK$Lw{R*L21eu@{q5-A zLx3bKB;a4%qGdF(Z_tS@Y(GcYxmTvCMn{0p9qNGpW=Hh`90M!GG83qadhZhbZ(QhMem?ZZv$zyN!#4WqlI2ib57D@?;}0u zBsBaLr@&8&0-eg6it-!;Mc^M{L>OR3U>wDN07sC4K~cf+!T17|-`e~8-LLcA@4db! z8Gp_+-bC;IuD$kp?X^C8t*!neIUJDM7rs@&EAqbRVwdRz>7FqTlCduWeQ1_ZqdUlk zOT9ii25NB1aI47`IN&;D%)=p9W{z=h6bY1iBHI6O4%Bh2RtdDXmyN@P1FkKSt20zJ zY#dtbn~L{P*(G#$#?CpwJPh!$ME;>5KS%9l5a5pqN^PDh9|g{;3Keptt>$kTV&S?O3mCemr1dcTJ z^5kcc^YWzRl8V{*rcj$_hyy1ubB$vFQAkU54pqMhDc}rK`CLvuw-+4vK zhEA7E1=zn%+r-mE1^rJql8)Vw-|EDQgrLkA3EC0^S;talGH*Gdz{O>)OMbgUR#SzF z`;lw|Q${MT`;=y#!-CN{MA*<*M7t3|U?U*RV0M`$X8g$0%UK{Q$rQ*+R=%)=zp!AD zs5lFwN(cAt{qaT_xks*Ey@~-vSzed^K_2VGx?4dWze*^Dt1NbB+ZA4hn7J*uqv)an z4l`x|95ym?syuRy`J`rPrTSUCd_Zx--lv&NZYo+@eI5e-De{77V-WaI{y4<+?)0XW z>S^-Io&5{L@o{+Svm&3_Qwcl}n@8f(s<$?gM#gccjCvpvAN0^n5EC6O=&^?uX<*@B zi4J;nx@6Gfc^CvcIiZM4FBBO@Ga|;fY4VL0Wk_VfwaF0B5oDGGy<`?F z5xSnl8U6Jd`NNpE`picQI;!}xf&DQ-t9)&u^o1iHuD(0^ay3J~_^S%5iZvJPIcLR= z+VDqOKtY9tr(s9VF5t94(0H)=spv|Mq^Pcs2c2`|6XgRDC;dCox5UO4#6PHKjbIL9 zOF)7Sj>>HCHut3pmSY>6X|3~s#qpALj#4|@GRV35d7R6u7NM+ZEVs|@v zSbBu&xct`z<@EZ0t(4m3-d+RnO~KV_2_(J#UsKgMs>t+?I2*1X;vZ7$|0=uEkL#<^ zwY7)O4aXM(YikI~A-85Oo;fRT1hyC6&xm>KhlZh>OUzj60+qMDc<^3zdd&b(BpDHo zEG!FKQPOG_7Shy_YHDK-#K)OgSOmVh8lT0aUWLrjo_Uc{Ox437J&~(%$D`{+7-E=G zqDqg~H9@=1J;Lg4^jV_|?VlcK^!$kQ$7bf_%MJ=x@ulZULpS9EbvI>0PufUk+6p4E zz($)&$JRy6hf8&d_AXv@v9}71gxW37r>!#iw67Vivv{;c(}VQ|kSEL=r+UDstG;z5 zC%j5ae&6T8eIHXcK{J7nyD1IZxYm#X6)QsyGvas&gurgMBNjFl?Eq0#aa{|9pTV(( zSFP<8tx`G{m$Wd~(ejM7NO6Zk5mW&q>s~lmH>uBd$cyMD#)?A?qt2LZwN<~+4yi+hRHB($8BcA*a(T;Draucl-7mmO$-S@a$%H=HvtIEwhsFbOmYQeRqdzn7MnqF`6eW7Ai`8HYGa9ll@wLWg81v_fB$ zXGT0Q4~)Sn%nJD|7z6#xg5-jC&ae~;HOF{Bo*_FT&J+e9EtU^gOajlkarQ6b6P)FM zon?zs?sCf4g+)E&Jz2(H5uHY8gTq1!ut3yWi!oomwt>~%NA|4lrqI;uphQ>@7_W;% z2FL`zE}AO62D_ooO3Sf2O~sVl@G2YmOjZP}2nqvtRJQ;`j+QS;4-XaMjNSw>K!oPe z?sB)`{3Y&IAygPw8C~IUlO0cvm?=bZ=VrN6_=WCso#tG%gLNh_bGtGH1yE+>Gl__{ zmtQJIp{#kL*dBMnLLFA$90Gwan9$w8>C4 zKed&YJfe!g=rS6W7LrHegY1xU-KNS??kLidN1w>L6^mURn{#T!z77RMrkB=CST`XSGW$;0DmvutrN!hs=NzA8&2lYbaT!?==r`LbJY2nk*J z;=s!nypPPdXbT-S#DW+stRg`~$86IS7bnTDW<|_Qz$-7zemn}Lpx$orqY+VJ@EbA& zuQUdODfOvIrg#lTUI-^1Otcch{d!gir_&}wIO|I`?}*gA&`r=|bdN2BZsLR4p_{%^ z?rXS@XN5kJ7GC$uxM-PFIVS`X#2L@nW!^(`Yk47%pbs+y5{qVrKyn>d(a>wrK9-fZ zl|;zc0b9T_(zk*SAZz3#1pL3z6iStS8bvs5+Pti>lv zX4kCc1P;QH<`Prix~+5jpV7FzwKM4Ldj96l&iL5Qpma~}h&jUleLRQ$$j9h?F$QR6 zk)z7sah*2LPb%6LA`MR95^M53-;A_i!GKwqXV{>*@s#hrX{;9JyK{>bMc?&cxH;|H zcRlCxw(a^)yI5Y8optYu6ZFVE18Y*`2h+gdlE)6(p->&I`B>YI`CojmU4xnb{L$v- z3Io?)_EP*AxfAGxp3f2Z)#jp>ECrDbqFu6pW{2-wK zVOFct^d2=Bov&|HM@HvywI9g?)iQEPe`bPDf`qhZEW#jl3Q$S9sgb z_C(xWpT-kyd)LUWqPMWTxUg{ehzPy_rC1r;BF>P!ry%YzNDdda{_wf}csMvRI(>RL9-7mX4jCX*YOH(Mil?m{r)RuRqLAbkRsX&qzlhB0$*KNke_Thj zO?9v77YMwPxaoq9WG+II?g>_lJbM92t+&B%mUOu z1=~MR`d*@!t8oRD_36&pqjka?UE=&+oO6{xw1aIsD6*=4(vN){AiE{@nsCH1zVmDN zLI%~|>P1H@CsEGDG5OWgS&tVdKmdA#?eiOV7iF zN@!D1Pa}IeV+7^Ko`!;xKA+GSgI@$Jez}%XFr8wMI=W0zq)*#X-OPR3dk-gHE70Ax zXaNun@((>E@3w^Mk~bIRXsTpy4g1>$98KCg_r~Rmd@JE?3FnYPnu4KeR-8M__X>E- z-Q;-h&MbJ0@>P1_SEJ~zH68Gm9pau}?zqax4+`wek+@Kgu3*5;L_ftaGoH&?w^cUV z1+F09K;f>AmBBeKPgAm>Q(vdJq&i%_bUgmYA{_xHC78t&GWlV{#>`D$T~hU(3`CIJ zaUdpUZacOmEHQWM#ll?j-G01_;Sn_qOme!{C@WZoBRv^1OIC$@Y4hal%BW##2;|?Q zk<(4r6^3_dbH5>IU-JEg7aZoz);O1B5U3Q0+KR__888!8p^l`BYhqa1JO#Vd734LG z>8J(CF0INwi)%p$%qvU_Lcl`tEC}>4@xU-(R7G12Dx)wm6a-P?XcR=HX~Ih+$NWj} zUqi{Xp`lP<4h;gA8weDH#1TWEAXudedEz}7afakghUWB#>XuZQT^D5ICiDz97j(SS z6dYN;Pb5e0A$RqTB_&676DAo!zd}*SU4M0se84xdQe?{p{j%-YPi_by-Uf z_e16Nu52k^>$d)6DtY1+DDpwyIA9O8yg=P(hIFWWRtUYY_ZV`azk_?WxRG31G+1Hb zJO^9b;Yy9%z7v(}ou=%eN9GWsJUVq1*{3poiu*e zD0yMXh5#Gn8Lzcbe**m*jzVbA+oSv?c@s4^T^)Otgdi`Qr~T13 zWA0(iGh1@ZDhpj_R!?}0ZkljB^YxhdWcy?1lRf4eggrw{lNBJ?e5##!Go$V4aJ)4f zjQZ2z%7DoTb#8~UWuP&4P0q7+S-btT(f3l5Q2-6_O^3Z{#kLvYtE-Des8}#=GM&jIpKdR^uLA8N>q5sJ);y zavyVm^0=oW`>24xuQCvc3^^q13?L(ejRbwc=${I-U#J(IfR(pTaqG!j2H52V&?iW# z*l1EzNM7L^cBLmdXUgMRb#p>-!~tBg zC`?*N!i;xi_O;|{O!iU>RuV&!SvoG4kVK)Y?T|$IR0Uf}5UANTz%gfy1B15W3Eo4c zfD;qwP~B9< z`7!zvO9}4cKb1detlSt*pZ{?>AbZx zh64b#>?r0!;(?^+6KcAQx{RDpP+5?gPf*vmboqpu3b+ka`2-YwwA4?{+7zQa>G=fC zZ62f~#5#%&0=rVe6J-UXl-fLQSlE&no+w0NgeS^!iQ$R50t-(p6(uP+n%^Pts3`H0 zCzT;7pKt*32`Ks_cjose=@|v75_GruhjonlsML%?O^i`3E{lvpO`!1kvI?6gY{`Mx zteQw4GLvBQCmP73$@yprfZ&A%Ef9!FE|4dZ6<1Or1CsO@M9*@-^wqoR`i>s$2FFHE zhf!seXj?dv=#}L5Q}dZ}h1@&>ls0_jJQusbh*W_{;>B=QO~XTxyLVpZ*A!uy$s7E& zf7d=i?JJv`$t-WnXBQDh*~|<=VCHqAL1Z%62tjiALw#DdN{J(%0d`rU5?ap4`6lzC zl?nvc<-piL=z*j$a=-rI0|vq8?3MYxN)2+BVJu6H$L&TM!jAt#c0V*6p4}LXwxH?f z^$S{;eUUZu^Xaw8l|EG|z_$tLR}R0bkRwM+In39)b16)EjQ)6%l7NXe6foQg18J#8>e#AsR8jU0^;1F~ejXGk$AwE=F@;>Xo*0Tmo9s_vpt+_Y>plp z20tN;g1f#Bi#qRf>rj6T#x3WJ1i$J|J-%1%5xn;f4G3E+?6tK-ozlnj`Zy}uBKuZ) zlZwmX2{~RLZVpdPdrQ^-PRX%}r7npciZc#y68u(6KAIN|-MMq7*M&dIe8$I6{a$~Y ze3#U{d^p~z&BuhH8~T%tDLk&&;cfCg$D%X*via7c)RZeE{kARKGPw$Wt$a$=?!Px+$jw$s(&qS zF>b6sh2VBN+Sy*I#>uVR*}pIxAKw{ot;ps|4DnhvnI%utky2TrkE9K+$qW76^aHfI zMjka7fSYOq(D?qL6s%Q~M{*bQXi3%a3k3O+e=tPQxlx_3;SgZ|s>q?i#hc2XO_vP# zf@GgV2(Du<<;iFL^Y;=q#`DpJt(kkGG`;}>9*yv3|A1gMBeIe_Yl1~YQgCq4H7kI-$2AX^ICEX7I<#llNSS> zkdL3fzmv!B_uQZMPp6AJb;FevopW9&dK}5j!~@&B+>(sd^a{-{)3{FUx8V27aYKCr z#t#Q^ZkA5`2qIwoa4e-!F$9XgpJkJ`jc@FhxZq|z4~GOEiVaeBdyHR7e1k88HY}o5 zVOnx;HsOj+zJ$*maEO&xY=Q0G*{DGrSB}!CHLgG+bl@aZSWu)xp|ugLVKbP9GObt! zi(~Q5WF&7#3V0PRg7Wf<_j6+t2Ryp-J6dvQpiBlxT8Ug#HT@gpP^`!1wZA5}D}u%Vc=xPbi~h zr3}|?nZ{7COF^(weS(ihbJc39Pkb6N)u*qSo9g>>r0j#B>KG5u8;O~l%hgz?f2O~+ z)rWb&{XxVHntTx{eJS=N!fKRsgIE?=Rib1+#fnA=m)6pe{dC9H%_46Y(yv|v)|otk*oHP6Lf@?NC-15=YK7Gpmi zr4t2qiyw1v5|swxvaGhP(pX*FsCw-qdscVTs@HY20!*+%R*L$Nr^a*4_JVZBfsAg1 zv*b-kA*)WM3|Z6rqG5%qf-<97X*E<8pGFK-^);mur%ibJ3Q|}nsp8pEl#8m#NZ^Eb zf?%UmU}Q|O!aG4}JK>#J9xc3+E4p6)dF|(yky81R92o<(C!8pCazV0DI3W@WLNnX2 z(B$JtL69I*B$1GVAn^gjAV{uRW24O&Xe;vq-ZeS0VLao0zCRxIw@(exJbPpQBDt7# zy|fTIdijt?AUzIy+r4%rz}^;Ci_Y?cB6WB_c}efy%+#TU24+<>be;v?PM+wJbs9Q3 zUK>WA706God+T6Oo`_An%x(vAFoScaDCl7~@$72sq z3-Xp0PGxInP|rI*wlgT(IP{jz#=)pdaBk!^9r8kKjj5|F+-^Lb+l_r|djwpzDCQ`f zn(c5L0QtWiavwG~#{KOx!Ml z;v(@qr;YY0OeT6)oS>pnwkdJ#7_qiUW}@MGyRCJBn_KKbUezJ@Vw;8Mk2W_~&i6OZ zIpV8JCG*x;_#@pMyoNbqp^T`Xt+(Zg$`(&gI_*aDi#ghKms9)tB6%sc^d7me-QOCW zTAz-lrylI>)BakvV=nQ7JcuFjY0b|3*|kK9+T2VZ{bU8%uuE8nys|@Hj4ffu`j3pZ z&TYwFPM1$eU^R~)=;8;{H6MCc)<2Z(>T=oML0$^+bxG~Jx+EhlXjaqHn>#z>!RUN_ zH$5^s-&)g@YG8!L_d-f6_vp z@Z@N7*juE3mgs`X+1Z-1(zazM0sb=`o!uNlx9J2n2yiEDERX<_21v57A`}3%vO}i5qcCE)^=)1oIUwaLGI5FYt|n=*N4RK zk~mm!;AzLHvfGJ{CQNFo;89Th)cKz12ZH1hWF-w;!}As(@pt-jcR_V&$A&XdocP zE7kOUN_%dxanEZ{=qQ+~y)$M>MJd!ede_h7@3(0&7*uCILVmV2?`(_{0}<$+S0CTb zYPQ5LNEUNlOSp;soe`LUrOkyy$^2n8A96MC|&r*2_s=OOKKdFY(B>g`PGb$N=+vlh9H~E-oiBgo!y)! zjb;Ya$gws&N=}0$GqDxk&*dHY-KX0er8K8d>5msNusl*9HE7r@QVjL#!pvWZ8jMkd z7%GhwIt(?|re~sg z2bhDC+S5g@7)2!Di+n2UsedzqQ7~SpV#rz8iaO?DyWWVLUDa@1iuGP)pP2`jcEOQWV6E-u|8Z~+Bkn{Ylb3)ejj;p z?}eEublr4HHqq~~aaaL$`cL{tGd1#}gYnK;n6I?LPz_aqL*Q_(k!g8&-F^!-42^W3~X@gM8E6oQFWSueahm+2@FLQ&lvxt0GLef_TAv~q#O9l= z?H4qm>sic^=Bm6YPPr0*b1*MEf8PRil)n_PFg_0!Cdq!AABHlblX3E;wsBL)=N-ia zWl5e?rvP4wU`Rs;g-kRIQY7wj%ySx>OTg&v;YvNOGLH*V|01|Rez$|+sbmMK{DTBl zO6lhvrcBvAqTbjcKTqW?2L0V2tyUrxj+D2sUu>AS@g!|`drjL zor|jf6CIUWkl##E6bqw3QTUkHpYm*QGTc19!u9JAf;!Y|HMh7R1eA9oy%JRhnRyH1 z!@b=n`S!e1kKM<^Eg3p@Ia7*d*CEsMg;vG%pxLvhla!zItt){6W0Cq=TYMgv^{4cr zwQsEty+Eq}(+dtp7%0yPlq+$Pa&Ya<@`758{@|B49$~h={>uew zU5KL$J9Q;Xp@Y*mV5Q5uRT@t1SJ3S!}@=iP!6US&JNjYe8*mBv=Lre z2qg0$XQB-w8kn>!$om(`9Xax+9dK{&dUf9G^tj7>t&0H5gmEqcEEBdMT$u4sm(~Oy z7@iqzZ*27^55YjHV)-d#-*9ttbapbD@PAyO|F}JT*#nw$D{3s^p$ZXdE)@GIe<3E;T0#};d1KQFbGe!LMWa|_~;_} z1vHZY4AhMRhPOjhIeyR)&D+9dr;SAqQs*soway+`m3CgHguon|EJGMoOTu$mC8ntJ zpggUj1PS3=a`a9VRS+&qc^YOsOU|4k;`-y7DnOp-15t^WJLkxTA619jgk~xN(jCpzPUr5m!o>MD1oyRGq>`BU-pn0!&TY@Dws7Etf7$Avb-yjoRTBAnUc~GW_?|kdrrr1un z&CSZM)|k^qG8G}edN3lt3Z67P=grPeq*FO;R{VpSU&XNz5f9YNv^Cf(kJ+Wxb#j$u z(PM&U?yjRW9ODs4APaJY{NgiXCY1|ef6v=QS+-i;dEqmZc#*^{9%}M{+nIwJ$bO=5aso>% z-rhriYm4OS3{?#qhZg&$;yqM$+3cJHxx)Y-OXMF4@^f*Q^4dIAJ_?*wb@R%Vwwk|X zjD?iD4i6HbP#6il(Dnx9qb<;>pN)oe5(;g!rE&Xh83$^-DboYMg6Dw~_L z2pnnb<;l;YP*6%S8?_YLuqWZ7V9I6A9!2k1EV0cY22L%-OMkl{hx0U-MxAt zRj9Zh$u=-$q~f|yY1TO`7@b47F(t2vb|ZqoMnIUs>@rKt!uQJsY9^K^AS%fe$Vyhe zu!O&`U{@A#7Dkm0E@1t511Q=fSFc{hfTAp&OaCB`bz;2~AT zdxz5BPOujv1pHIv1<}SJ@S*&1i0R$wjVti7{VY6gzq5Z~I6l5J-dd5*?5PAEh|ME$ zY1Lbs1o6ZA8boIiW$;2m&_gppOmtcedVVE3=+WttLC^k%VOb)Js~s3S;;D>K@riVt zf(D~4L<(=jN7w@m-4Sa3oLV}N`FES!y z3b8|x2uNK3R*DVX#K(;ico~Flg23F+O)La#jt$-9@3QXUX8#&4E##ZgS4Tks{h`K9 zEQi!3Du*4b^|MisbR_O5q<*KwVz1Xl$ERYGa71*IIVZ*VR7S?8f&im71jnYfL?QIj zj|{g3LEOKmN26jX#@twekiT8E?3Mzd*XV3pe{FOiZiqT(o?%5g0;Rs;WyQ42xGvtfEs<5h9bHSc-R_v$^f20KzR9JW#rUUE(P8Ff? zVD(ecl^jV?T^|oR=g23@2O>`TccO2JjV*|OP|q5{9K@D@1RWbT-sYb7b=3)t~m3dW*+zf8lM;${XEYct0cNu^$?SZZ0unsS8xz_Ts^N)#)_@K#^ob zII^%TY(525mX6EF}4hmQCrRPaQH|3wMyD1xb(nd1VRuG8=HriA=wk~2m zT*=k>_DWuK*?=0`_i0}3$b3m{LJH%|4Zjk@YvS2E3O^2zV}Jh<;;$|h(g z5OO!AVH?*PGN58*$YJ`~EjC3Vux_^_7B&^_08v$OT?>TmuYHQ;XK>u%Rcm`i8G|&%B+i{!O zFg8MBx5}9tG^Mp;;yW*#A-%JZL~CXkXVJLrAu-j{yesfPa7u>|Fan+t<3z6Ai!3nCf(mdcP!*5H}W>Yc!UQY0fd?qUbRs@BC zJE~g%B1g-Yq=$zJaYkaPEZRKUUG8?CphBoHt}?p9;U+tt95GXfWR5Z__?#*F zeXi4-tM)T>CNOimG6e-tX5=#wTC@C8G3aca+|#?uKj_SR{wm${JJrZszvMwP;7p6m zYQb9a$eyI3F{qX%a=g0gP{?-XX-e8;s2Y^Ma?n+fyyOv81V)$9sI-tg5+AGrR795M zg{Pm$x)qCE9pk}iBRTLhrKzi!L6y$*x3>E5(7Hd6D(Y7q=*pLIs~7Z(;1XjoU`Cmd zF%v`?Wiu>xbcUbg9og{$)15|GiNQP?RH7NfWnb(lMNI=ykk8^+WRL_NP)0xGIWc*7 zJZ+XOenyh)*N*(lS7p3q@(<%^7*}#HUv}#aAtAqgy3}CFsY&#GWX45X=&&Ic#9(0+ z31WH?!7@z|oD3wtniVlK0k6C;`|&80f_l5fk48j^!EeY*RHZQ(OsP*z%8}2Gkh~C1 zJeX)Dg!}cZ5KgB}hH&XG-SR>=L66Zrwh+3B4`zpM`bu#5%E~N$JS+5(wD7uL#zo7d z$~hsBAkKKkE~6BhTgwZ91bvtxkXSS`1d{8xiiTc`_OYzQtt3Lm4%h;gk-in=83jy@ zAS_Rk+`$82G&6z+K^$)IAePMz9^^{(b1oZ`%s~Z#*-Y-~00uYDGp*&1sZ+9W9Bmiobs%D$D`1-=^nzfw3LGWm<+crm?d8>a$#z>V>^S=J-H+12>+cdy|g!Y$fD+xmyT>%nky+UL7&n&_hK4QXd;n_Er(w2S3c*;)6lI6;rx zGq5H_elQITE_v*rTR%;^)##|r&B(v_Ub_Y}|M{cM%@qc&zwD*>v2J|5$!x3XrhAL@ z&l1=Ad)~<=oQf-AkYNi9FC5WXf*WEyX3M^Wylc@j`Y(8|M=oslw??Pdrz02@+}o%9 z)iWwA_#4b@VFf}EU18Ccm^ZRZH5G}fw(K$`_2sMeaF;;8{JU_%t79;6a|_xnU%2ee z)T3Rjkh0!$@5=fI&s?uyZO~N=`+%fXYS{)kOk;=qAfW(ZR;$zW9yJ)9uWwXGM(1(0 zAISsNGIDuv|HBcUL+U{i=(QE;?n2)Si%z!{GgGhYXM@ zHP$_B#nV=f(=*;DQAl!IjQ)K=ei50~lT-c8{#Nl)jhf63sv1z#$TGcI;69jX&UM$Sq z&ey_f^6@T)N7OJd$?0CBtY8_A^km2^SrzWZt}q%drA@gdkbjFtPB&dw7~bWQOX~L% zUT~NnAuz8n zEeHV%#j_yL!^8u_fKe5dV$r4$K)%uomLY$aQD&6V(7>-)eAIF?94Oi}tJ+0|d@lon z-q@?ett3NChMNmI-f1r1Cvu|qkh^-vl5(QD<&yiS?56Ql*RL}2WIo^#u5gNYrzv~rkvT-q%;aTj4+{S@WjhVV!k!HpmBo0^7)?8g z^(CDPNm}3dFZ-q`+rKX~GGY7Dl9bO5QHBJl&J(^V%J!WJw8UyyQH`4zYOa_@0c%2Y zCyn1VN}k_dIUx^RMhfy@Nmh+*su2gLa))9W5@bf&DzP0sg7q_rfViiA8P1fObkq26 zDNV!e$&impOWCK?4gZD&Cx~}cvLsM(t`$#ZcaALPgm&o+6L~;U*b*oIhwOf6I6S*C z7;OPi{Q3p;!KCDfcT}>SBekf*c?+OKDrXp^CELd9*c=;QI4F;NvG9!dl3Q1$wK867 zqy7Z?Hynk~ptnc)OY$aaZi=|O3tN_90Q|-)~8EsF8gIM4hwHLHT?qd#69`{sa z9~BVzRR$uFA%}#W0c2#bk)SUal^&*kBLNez^7biiJ$cIjySxDU1S!o724CSDcd_O@cRHV7rLO3%7FdVkBT%yBPS75IR_+@7%b!kcIkPjs}G4-o5xrZezzBfY#p;HJl z3JW+guPt`^FJa*m9hpKPjw4eROm<}I8dcB98L9E}jEfuX&y-;KPWHDSto#`LiKPU0 z@t?|{G*)g5r%!*w3l}u}p0a;&C;@UJ_MDzisB`d_)OpjeS*7E0ao!4DMqNhEC#Wn) z%_pd9ZfLH5?0c$MRcNeLXe^e2bjI;@|3p>uq!1zQ3x?g z=_LtI6rwP~6J@!?@I+l9B|L#gMTwU@sSHW^gaeRIK+&HVl}OJha04WRF$A`yW)x~- zjA{XaSWtB}z<{J@6lwy6&zDu$Jgy}NVzX)@eaP%nbZO~7Jsn1sQKD_(NTOGg+fU7B$`wG2vWEXaDB>&Ux!84#2O^0V z!&x;A4@K_Yd0C9zeS^RD@7kvbL^7G>ZTajX!YG@P36}f6(-_m4*NFy^$zUS{$>9$T zK(bXz@*}vMk@HPZ03Lh<*X6+2KuJox&0dgUrU)gL&@f19(5t{-ck0%8jV4@8L46L$2T4qM4NroRPc*9;pF~C0P zxc7vBL(pNAWyF9(e5fj2Bp7>r1sS)d6)w^O4nYIh&g9Go{B9+W;#Lrjr2qm}F2vYc z5YGgq7yE>Mf3kbMBAnKDT*fBD7jL0(hTB+soM* ze6H7pKgwVbLJddr0vHW3oVqu;GyhEwMz`ch5taD5$z=|!V418)#Dud*m7TJ?mCzYkd*N={efKVPw zYioK;tHUU*MwoWM$~l&rmfn^15An7sA)(pigygjMP{`Z-LoQwRaA$is+1MOCGz@-1 zI16cYVoP3!MVFR|smJ%KJ%ab%p#foQ&Gh+iHtO-q6tu{` zmENS{a(F_H*N2=HYgtz-n;Ias2o#Js|rAAg&rDrTSm znZM6N)jqS?Rbq&3R8W_cC>CBbFHtO(p|{sVDT!jSg7FR(oG7*>9?~9y2%<#xDJwlJ zmSfD$3UXLx5b}E3waN-^;dvSIPaV=|Omd*^l9M3v?6j-=+%d>EAeBe57*n5=)SPwE z59bbYwD-J_emDk0r0SJo({bNagBiv{00j?o?rqDxka&~7j+@Q~fMQibb{Hwn|y+I0^JEfsT z^{?eE#*Ou-5Zq2jJKHPOl)05V`xl1e<2&Q671?EpAzsTSv*c+y7_da&O&cDO8{bVo zYA^r?YXi{u{tYQut0s@+F6Pmjs^bk9mdKajar2o4gvPBiX0kTJXrp0x@5o? zB>Nmfz@CgRk*ya9i+Lmu$lp5o{<24}X) zr|;2v+^4HrsP7ROb2v_C5kqRRW5m%C`Ivu5R2)wl#SdTo zK5lth4rxJLXTozA%LHS|Tm9`GqBln^INhT`q6Y07h=6BaOJPz+JvZ*ji-Ata$4?>O z;!FcVl27}m)5V>-;mV56IWH7Fj%0qR<=532_3?*4??DWI)P4)xFWV!Bjw5T`82C#9 z#t#Q^j>exVoic(*qqI54D`5O^ETvH~1d6|(Ws|p!Z|s(41UC(31uw5qY>=|sWBgL$ z8+;kGVG*qg(~^6$@oKcNpyW&V+yRGJdBqmk?wySq#Bt>)jannvv+M{hJOd}8!h#|l z3ayP`4V%FBzMt&J5s=_a1oT3U%a0in?O(p^59szrt(YS^QtkOKt*t)H1MUwZRGxehDSavSB*JQxbc0wHSXH8AKgEhh2@lIdIpb~e3B+iJ zS&K$MjmugZPNRkB(!5ExqN5b9C0A`!#n6ojMF38E`^cWv-L#B(9j&*JuVDKGkR2XM zXH;oVOdcLjn<>;oG1`WCjjcE>!(iMaWIV&W#xi1!8V4b;>Tr(kV00SN*S`IHzC0aRRv{6v(jp)Dn5-Es_JWk5@$6D`3h24C#mAuQk09T z$w=UYcY2+@%49XL+iI>^!Kn`Yb?i2<6Zj81Eqf^64PPJ-w z^;)vvz__*O&hdEcz=DFjrG-=3+8NaI&X4U3$~F%D|BthIO^3V?TVv`f3%47Z<~rmA zo~hJ>nD(vh5g^;5n4c7r|Jxz=VRHkX8k`xfOs4(ulyjtMNf+j~Ne!}Wdm7XtCkDgK zY2VI?IUaWyCoV1$-*eh%pTcCKcf|=R8fBXj*Nzcui)1Dm+W5}-#TRzN)FH3xkbAMs z!t+O)n=9x0o97(y)n%s(AL-`cHOvtUWkmgKy)92vws?9n*lM=NxE!RnAYWf3FU6ML zBNw*&TccC!)6w+QgS~y)U(0sPC4P_xF(f{%*_l7PmPk>XpI~u7h7$(y$_{xkwuBw) zKQh`nw5FDex08~0Rbje%&eXpYLhu+3ghmWY2S9^;( z$o!+9$8>xSM!}z_ssH{awQw2QW45A9j?kmXnYL5jvnC%Z$o=_Y&HBUV`jGfNGCF;F zI3Aiy09X&ZKU|PsYL{YJ=(p1xhoOB{mu$C2v$39Lg8;OA*k$WS!wp#)iWTNZuo^;M z@x&tTlT-c8K5yk$z6gQUdPS`^Nh_*Xp@!j=BR399^Pf7Jd#?xTc7TnJzAF(Ta;T-;P*0_K>UWFJ{CQN*oH%5Th)cK zz12ZHgdqAcxa0?IiYnk3s<)&qpIA9(7wYG$6Okg$hXClWJ-67n=d~wv6in6L8MCCK zU|MwlgYUO#F&I>5K0HvP~6tGodvlkUn{nT{cQs>&3?sebOdqM2mT}L zeH`bSLZpI9vp~U?o8^Me@<5=~?jMzKvkxxNKpJ5-ccUDb6K2Z@$4G}mW6>yyev>HW#>nZ!IV7(ks zWCM}SXXBw*H}b;*BQYzv5Qs9n1YEEf=$1I(MVDPh9QbY**LQ2EGK93hQM&Mr6Gp;n zmedMKHlMS$8@rB*WuxUhMv*^W#K7`Mebk^~vq&-2s|z!KC2BB65n`w` zR_HL)Seu#whe&ni9fxzo82N9M!nl;AAVk<@WbWRE^6aIte~AP`fNIIrH9f@uEy<7N4^|3DgYS~n4HT&i)3|wT@Fra zPZwvh{l1JK*Rg_pD(k6#GlEetUZ`ToS=fp?1{A6qkS@u>HzH?O(QsWbFc9?&8kjEB z5QX8nP_1Ew!wq(iGD9^VAc}(q<=J1BQH7_@0^8}Vk->QP!%`?4(A$~mWS8f$2V@eu{;=-Q|GJR z)}Krz$GpNwK2?z03$;9^COno(mcO54DTNXJrUNvO(1t{G`yL>R>8o2~5^G@k_-NQ|857vQ^^ie`3DKAl+w>T zOql?yWJ$XRD$yG|y3!^|$ z_?XzA@@#N2+&sO)_3IFVI@D{qyf7OfYnY*H8X+sBkG!oQYrzVmZ_9u|`w89@=ttJd zx+&1&k6nG;!otEu_-`-#cNqRt-+j26s4bEwV*)G_bv55)=eu|zw@8C`z_s+UCc-L^ zux7lLpVjE8{WS8F)oY-gHk%Avzz=)jzr*tM;F21i`(b3rVakW9^SfjF{BS(IdpO=Y zH|^7QXT570|5V->BJax5D`#OnPNw^aKNw@m%v%s2?(IGabwxHS+B@~weLUQfp>x-H z$n<=nfjT|t!&FI5QZ}-3GIw_b28>1OYi;p)VAh|~i`KriKJ)^q{!cGB7-671Cs3}$ zNy++GBblSVs9qz}$!51lU$g+yO&B z!yjLshApKASO@S}%GIcd5*^Odf-{)%bb-EqzBSI=>kMLMi=+4K)}C7Wi}iou*8c4U zxhy~Zz&?s!3N=)Ovj_aaIi7igl>|w$K0uhIH?*U$6j)yPsdGN z_ZG;zV`?8^v`I!@%Rl<`wWo?!f8>9DSpN?bv4KFoILN2bMtEf*kj#TfOf1Oz7s(wt z@~0heZ|{0_-s|+Z%Y3bi0Lz4NE&?nQwjf-X@lSX8Dza<1J=oalPaf)Z_Ai8?XF#Fu zpnxpqJfBXko;)}nZ9lXDZHgvxC8#AnFg!Eb=KoSGKZWcYZf=gwPSpQ5_{Z(x%kXdC44ij(hfId#$?I8+=)Fw_<*#8{9gTOm}K1T!Ne4Aw^%$uFRp z1Yn?U6fnFUqRR1uj%eN%E<0^3%A!Z9=XH?W1uu!N@G)kHz#N+_Ll{*{!gE;0CS3U=aZjeH1=mAq^5zbCUXJS3q6E%P%-qTge%V823|DCrs8GI< zs5&=ujap8iO1OzSch;78g6f8ld<&;S$H0W^(Z9t2NgVqInds6QHOkI|GEIEvTjw_A z>4$GTC%;-_P8-Qog#7Bki2N#e((s%&J3En1<*-@t4{ClD$3{dvP&3okV6QVvRwq}h zEV?VXhFxH^7@e98HidF95%>l( z$?SM@N4x!dblt5XX2OwVu!UCzz$p8$A{sLqV5xAM`qAhLMq}rx(S2hf>WspX>>3Ba zqg2{%lwfCwmEwus9)plE6WWOJBkx8H?or58hU`gOBpH$#=|eY}TuML|czp8gkf3UP z=k#=||40r8r1nV;ge%hfqKjRo6Qp~_I7r662=t*@Mvcm(-f?VrAm^3l_0cgt_a{SW6r9oK3>4p#3Wz_mqkb%v^jjYErlQ}G@u zyQG%ur{xDo{zE~2F78rZo2SZ0fwQV^Ub)g%^S6w#kaE}Iwhan}k981ftg6D9?n za+$M7(K{APY;%Z#Qwv9;2wstI7vylB=JKc{hj`#*>pJ^a1-UbBr);7Y$DP;Fv<-QS z;uplRKv(HKgZ{L?ae8OG)t~m3dW*+zf8lMe$OuE8$p{(Q;(QZT#*a!$Sm|bRxnb^)ukAU!?2t*~B0$It*7nblB7VOF* z&cdkD!F_vwynzDkk*im)Vkl6S*JTLM$2zg@R*=W95=!AJi{06F#Vr{{1r?^dwulym z!$w9su@`7k%5cp93IK=es^w0{tY(EQ++wbgO7>M{s z3WwvGX?;AB5cJSY5EGqNgPvcB4tjLDWYCjd|A-H71Pw-8h!oz4kFW!XHADJuL!Cpw zstz_R@O9)zSNHYy?2>omnZJ^+7qQ`u_&)qlB`s3pGoGyQO&(#^2EMj(f$ufu%>lf~ zh>R)34n-m$bpco@HgppoH%j1T5V{Ehb3-?=5U@Embd$f!x`YeIqv!414!VBJ!@`np zMqeET1@wm+H?bU2m#7?etk%y)K@oz$nL}O|9iNI#!V%F;=A0DcQyCeX3IdGU5FDFY zvWNtF=|_g!f*|hS)1y(b6k~3zK*-;&T6SAstn=g7uQlYg(Sf)j>YRCo73m1X7Pun_ zf(*$R@c>P9@P|~|<1{6x3e1HbrVNQJxHcIAI)coSpqI>oB~EaS1~F7Pr$0dxE|&aZ z%v*itqXivReA&SMn4nd@Hc|S*6|l<7Jm{PwpC})QfYAL;^ewTm1@RB+StFQ( z*b;JXjXqS68|3m8q9g<%Euc>MrRb+ZcoDJ6x@eir>f0bQQxAc32 zBQaB@v&dJYYikdm8;&mo*47Y|LvAglvIg4=?`On3_Cv$a%_U|mb%Dy;UOafOI=yB9 zD3Xi_M;4Zatte?V3kzxLNHw*w2jb(*EGzNEyhtgg>fw-{$kn*x z(RCsWF-$3GTY4h>v6(sfvV+1^mrIWk8X<+`Nkcc~o9b@LhMu&M%(N9mVu6h|m5!~8 zm=AZ(Es_^qr&w30cwk08?Q4eXEFNvq^k97fSt!C`|}t?d=9QaTrxLPOw! zHkNRQLJ?E}BkNu`SONNHJLE<55@T&V8AhEk+iI(Rp&e3*$XG6DX$_NM+K$`IhOrS6 zyH(EIped~#(=0}G!Hf}gj9FfyUCc1fqH)_pVydTkSK#5W%8`E2b1{WWR6$KQRkR7`_Lu_TnMeupSe6Y|7vbY0kKiWodJ4HdR5 z@CMZ^`cIt4Ju~&9JikYyAoZn%-|taCYp7T5&^|>Ohd4_ezF+=Ahh!+ULSK|;&amOi zCo7Z+Dto#1C41)#OQG!S7!SxZHYegtVE|f><>6U3&i+Mwg0md3vushyxzH4@R{+;J z!yodVEMu>TP9wCzVIc)rAZo2-nI7l(=AO~K1^=(efqf;h{pD(VGOLYW!S;=F#qQxAO!QLWOaa(G?Ci z+41CvnL;FUloF5`KVIZ>o#tG%o9j$q=5}QY3ZTr$XCn3LrDD+8I=QEJmw(Wi_xx45 z>38UWvyMN!C(sNy(;~B4u$DZsCu!n>Rd}bCd76?o8L9>r(`MP?X1OPSNRxm0s*JZx{$U&q<4W%3%Wl0PBy^@mU0FiGTmtVSGcMXf zhYhhH1`Df5umvwN8wa{LNq#jeVrBwfd13bBQ78rVc8ed4h!TU}ke8@RV=$OfpPG;c zXheS1M)l}!V?tB$liohEXLUDyHF+JtREx|IPCS@sC4~F+tPoD8O@?skFWvG&H$jil zJ+=_Ki4SInZu&~OUH?9w75Ye8c-=4KqGeL$oDfJ5XFOx6mJaLW$*tvuK!QHZ5J)VV z83M_5T+hwC7VTqMiCam8j2*BAEF*m@$TJF2x?0h?@TodazUP~f7AzPr3-b&cG&i2|-8U`T!d!P}dqs@C>%nky z+PCj|zGrIP^`CaJyed2E-W4b4k$VQ#q{t7Zfx#t@9dzq5Y_}R675+&6#rN7ZnEB5i zZEmhGaQ$U3#gBF4>rG}`O*h?Jq<@yU*5C6^HsMrcXaZvdM|9TUh8Tg_c8qA3#)7KM$_+=6z?7cP4<^=KC>q^$ScyR!blGuJCv8*~-JKA`!DpiA<8kWhdy ztJP_Gj~a~5*Egyoqw~1hkK}=B8M!>T|KSdn4m8rJ8ATkCoT?XXMW1xMxwkyT&#NH6 zmr!~)OYF(f=CHR&|HKM!IK!gzv}OXiN-aZ7a}o-o)6v;YD4ca}=n8iDksjprXt~1M zcD5(t?)o&IjoZ6Mb``ya<;8`C!$(vMKb?<0Tl8Bboz(>b*-*(6VnNnlj!&W?PAw2?w4vMU*pY&rN2gq)Ty(S#7@Qi*ZO1_3K zWKiv`UUal_66IVRQv?g$g?WG;Vf*~X-342OC>#X)>{v=_<`UyIf#MX%SMl|lffCvj z)YHhG&KNprNtAeCrUzS}`^_>hvkle8W z6JyC4E}O72Z63GE34*y>FBWEAyRxQdKiJT<|*10vqK>N7LA;4x~?$1OPhOC_0KA)-%ohKVcu+wb4dn)N`a`Yczl-uGhr3# zNQMzjd`p|BV3)Llyk;>SwIJE0RoQ28EeL^mg=s+uSSX$afu0J>f)HI&6{TWP^+>QE zR7c8Ai;_bAE~CsSrJ;de2{8kP*r3%YiF_{ug5KDx#H}PlOyruHHl+hF3L!l_{!3sBl>GD@p&}|K>Lg&yR zaJhj%L2&I1eS%<>D&$E-x$#Yg=Je00TT*3qU68Gv&@Y5+$jjR;evcY)gEOc5}VRharISqALOAq(wl-IkmrF^a1`je^T ziC6c@2Yus!J=F37b)y;5bINCh& zmDb95t&RE<=-+S@LWABO9^`Yo6JXV^&${I2SO?9E|$Y;mUx? z2z73U@MOTjmHo7%+LF!!%9TR@C(4xt@(tf~*h4F}%?MvzT`WSyf^nnC(=v*a2gPxE zo%Vk6*M(==Y``A5@5VO)76b-$|q$o)Fv2Q~<7KY{8`-dbh#E%Z_|7#4fb;enLp1(A9QGqI{}?tt1H4>>9Zt3A7bY@E$4^{5hAe zP#37ODGpm9$9RUB*_$+ntq_Ibu$AQ!9k#jxdrcrJnYtyiIqVN5IWhICGP#E(EWS5G z2BA|3G71YgGOsNTb9E%mktqb?I5K6yWJjj1QT3b*;F4LuXn&>z%XhNB{b1$C=ua#q zxQqW({-m*TV>o^K8(zG#;rEpNi$jsF+$%5{`Gh(LuQ~7ijZPp-WO20vZ9_b6ouuaz zR63;dcKc-Y*}v1bCpvE$&ewDqbs0IIpt2w}pP;Uh>%7zR2{jdP8>sRLDEerrpPIEP z!hB@p6F9edkdhGVC^81@N(oPt6^v4PNx~C_D2(t#SuQa=QCCO_PvB8e;w4WiLsCBB z0OS);^e09o(lZKDCFpMR59=7U!KoRAni!*6T=GCnkx1e5WfeA0*pdS=g{T?S&1^;1 zACx?roR5|O2wqt51Rl%QkcB*vthka2{D1_euij19cl2mCI5v7Zj4Go<+rp7VuOzph zn$MIgAcP{ma-NG_U_`1wB=KTc+YzCr;i1UgJ1>ig-;6Z-wh?)9dZTNiIplXTWsUlgvK!*9`A zGLBOzeq7(XEn*R8Kw*11JA=>ly6{IC3__^kXkGxLA%;`;CU)R=-E?$IjucUeubW)v zunLyR+BEFmzgNK)Ix;zjS`n*fb1o%X7bk5r>;Y^DLSxX8$vLJRBW@QvO^n;+@3XX_ zKV}o*ofg@XS0|OF#Mh6Ghk#HXyREGbv9w!T=zx`TEHy2?E9)QPZBs%*v&jjeihgk= zZ}ShiblJn5?cro&bM(+K_z7VY-1T)>)M?MHEtPiT3+ov_O^sX583}&XoqBw)+9P=H z9U2g}*34_`W}_ajq7!J5eJj05#pUpX9Ip>Iho`2!rRsmDG05Q8|>7C>ARi?_j}*cT1Q@(r~4MD3N{2N)L)CWt}}JssxENbo@O`C5(1AsgZ{L?ae8OG)rWw(w|M;a7v8q)7)Y>R2n+dk z(zU$#C{{c(ZW34?3WblxJf=208+JdY{#Sp0dczqvcS=Kx>R-!Sj2r7uA-J86cD7fb zv(Z^-Zggk=!f<>XPG?0nS7M0QvdJvN%V%*|E(u?vPpJ(L%nS0U!2tYBZ2%hIe@+V4 zs>vfs@;q8nbzF>LZ^V|!m;8evg3gWVyhOBBkwb%vKU4l}x@5o?B>Q}~@FntD|NOm# zjq!Z6VQc1|D2;D`fX8g(Nu=|EN{FGF`3&Av@2a)EqC0&DS`R2UJJ-#!2pMzOV1Y^R zb^3tK{+HuXV-COBArE+NPjPe_gEL#@)AvFw?$cE*)b|J_Va;4P*5iKkS|n8*xMa|` zIUJ`GXCbxN5m79WkNJm0#qp$3{O|z= zL=DtMjeGK9pcC@()Ax7sg#8=)w0}BX+^HL`tmvHcLeb+$W+ookrs;z} zoD*zD^UE}@Q~ND&znq4HY7G1(0po{*I5$ftegqLPemIuWs2Bpp-_Nqi+r~F`OI&ad zHr@(eM}Bm5UvJN@AM*bxX8y{Bw@Y7Mq1Yg0x5qg0$~X8jXu~2}6{aQkX5%%BoI3^i z5+mZt8fvNmtVY}8=F8-2lC)pyQcC>UU}7+PVz%PFWKP$duhh`@-#+ZtemG&gAqgH zFZd_N#u^=A=w;aho7bSq0h4263Fi(uIgzTHaduKq z&xk_DOUZq`7l#x&g1@Tng?_P`?c}!%??OuTQAMK)t7FmvGkP5d=7dJ?S#;r>`lm7m?DJVoxHhMoBk_Wr0;CO7>H%Xq0eiEgcC?cWg~Qff(&DYtaa( zaal{lX|xbsnzvyI4#3q z+#_T>!@I^ZVvQQ7J_2><+>-M?OFoHsja79HKC4BrLRO0Ukf(;7V}27-$f{E*Lsl)c zwsssbG)Y4XRRv{6v(jp)Dn5-Es_JV>B~A-fzk(FjNve3Z6y>67G7>o9ogmmK6;jO| zws212ouIUx@J=j`7T(Df+3P=_p_IJV!@DmdrSc^?G6rf-I8o~4f@Gs`LL_9v^M<`J z{Www(B*+v=B;+7Sd;l>Bl55u3X!F}s8;e8UH94|jJmY@8KOXhBPYn&8UGtdsjE^fB zwE*xhaxv?AX(4p<@*$5vY%SFa@t4vZVv z?i`Q7F1H1FOADv6wKJ&aogdp7lx-aP{~u@bnhtp(w#L*|7H&5jq5(#?9%Q|5ZI8g* z7R4L|SF<1gw?pp3=Ek_ceP*~anfAw1&XK0YT)0F#wtBtxI58M*PWyIFT&lCTa^m75 z@ja)F_9;vzdRLsFqEWUfaqSqfwn%28;haB0$l}S-;E`VEhB-Pn@~RHG7uzg6f3&%| za=yQLPQ+JDt8m%r!biF}cnx#JLK#s%TW`w~l`Wp047Qr>G3FNqJA5qp`XYHLw)7sk zu-)Gpom!ucrl%h4?bH5Rwqq{wgFJ{K@oCM@{Mof+jxmBjYS;yr20rKWQOP zcyhEk>_KS0IV{lyle4omWnG5I?Igf|rlYf)!wI`vh#avWy75+j->c~Rq1IT+)H3$) z5%pEFwciwh%s={hOvmS76#RLb`tNU23zwliW-H3%2tA6d+D>^7ROl7?P(kj`4{O#R zKG%oD?~&2z)5G!5Tmo=3%Z_^3{o#WAQo9t(Lcg8n#=|z3Pa&`BlI_-LHrCT@5P+5s zyKMbvma`t3!TbnTL&z(hSj2sDs=wLit^CRtA+TDnsMSU($SRt<@EsyVeUCW(NX zs~Ul{&0~BA%y}E(6#t@4R>GSj`ati2GQ7)3c_g)Xw?5rDd$cYowkWp|ZVBQy1og4# zDTF~B65FaSl+VY8&b9SMAzB*yK;(RFRuRXWexaYMe zbQDb0-Wju`VgS#Dh1QW(-*3}mFsROag#2u4-q{!@0`4jq4Jal2+gZ()_yx&gu4@Tb z(daEoA$ybAp1Rh!fyg)3lG8BMx;+IK4Pi1K8-*UG*pm0D40&fk?#b7Rtzm!LfK0Pr z@hIR-jkR5U(K_TmvfjsWt|>$+s5BM7VXNUUH%qIX<$-{x-Ai87djS$eO(BxP)5xP; zau??L2sGO9)PR;!Xh5Q6JRLJ15EtU^d^qjMF%;z0$fuyA&#({JJF6xjpQy;8h6tRC z)H}}tT8$#W^_2ZpuwITRvLVXmvvJKD*=wf-`C);Pn3Y@zM44RzE?DZeEpfn$F1w64 z@ZB!1@77Rd2x)(#bm1E(jD*!JsTGoJK4)!L^Ukt@SEKWJGde0YoopC_Y(6~D?4C{I z-`UM+(r9KtjT~#kqvSM5G80?j{bXR&M2yYnwCoaFkUw6;!173a)SzLrNHNr_3p0Nu zYA{9-VyHA$=rGh+o0Fa)TUTwT*s4A7EXS{&TG#ndI8@Kg)|w;(rZKz_AgIZ*u^M|{;;<4b>0?~uwxW)C*seDsXIIg1 zT`({Z^$Z%AF4Pc(;ki()VKkQOlikLItD6thjT_PW_sC|4En4;k zy(W{nk}?T#yoSF{-nF&_uai5ynVOfiy%G7NgpM0rtmsiQYK%o`8+Z}vSjsF3RhgIv zMub6ZzS-J-K@+;3#T;p_%A4YpD*-qM^Rn~zEl@}KO92bx^I&0;9EK1I@C+~+CtqqC zH-&uOQA|*lT#8MT#))Fx!`xUzuUp^ zRI-Cq{y~B&rS$U-Q>N@5QE%*!pQrK`gZ^%iRx6PTN6P0L1DGNDB};ii3YVs0&I_9&k6Lrj3RYJ ztbjy%O@=;IwcqPpZ<-ZLeltZ;EQ|t0;bUTd%Co`AaP#yE*RMke>QJvG#ooanHaONW zOt;W^_5kG>d0Ro&f)z&JmH~tI6Fdh0k+rh!5VZJXS6>I+_b^>fD$A6{zX`Cdn79Lmeuh83JPlh)4X_U2v6M4~a;V^cp&VL}9CBF> z!66VT0`H%1jWhQ;gP7Uk=skPIZgJV}38)viSpP@2GnigDJETHR-d>Q~vWbJks|PUq zNjMB-Xe2nWkc;FkMFcw7AY7B_cr+LeHYS_UO_KJQJM|DJRb_*N2Ksc|)Piq;ygR1$ z5k{M2q>?lZSep{R@s0>`+0xV|y^H zTF!||`P0eOlLyD6?T0p?P0>WI1c+fy@MJ55;+cexE|OnBGYP;z-6&vqJ4BV^2OZJ8 zEnIfmSoEON7zq8>tU?RMVAe#@IW}2_Fshb>=dwynQO~O=K|=VJ9K91o6@<%Do`xCE zGO(Hwe(`ytlNUNS@+7(3M+kVr?|t$jq$O0yM>?6e9a&L8xmM=cK%*q7gSSF&fllco zZ|;!i<)~gQO5p6o%&olOmpx?0aFrG@2~Nq7s5&=ujap8iO1OzSch;78g6f8ld<&<7 zVPJxdieMM{w-_LaW8WYXU0S0?*?CZ=iSK;toPgcO*Q*cMTfA^kYI%yNu;e;WUSKg zR+D6}&eBC{*eghvyv?Npr~V+OIc38xFj|aG%?6u7IhY81gPCM@yfE4=tAE?SN7vmN zVkR6(23vSl0F1H^E21%@0hS8(D&(Wl6^zEtQ=|LFLev?BBiS_$fJdpc-6+8kv1`~= zu(!t`WXyy%V*JRvQG* zAhl1#TH&l9?~5*WnNE=I8RH-s`y$YXW*IdqQ#M@c_0cg5{2{Kpct--b4Gz zQ}XG595afs>N>A@9v*?txtx8K>nFdQF;=wn4bv!@bx zAU2Q0rB!cj;<{nRybBS=Dk12hnII-QTF`?KTByij@+;9nk4~2ideZA3@!^f2!DtJS z!W;1sb^x(vNdN8JI>*lmt4Q)p?;ZsGy3W% zD4;*oxQXSExr!q1&6$BWy0bYS~W zJ?*6*8Ey-LxPMQNM#WN$xv>Hvf4gegJu~RRk{T6aLtYylh#R8LnP*s$jzDaIJAxp{ zkU+bD>kG_+9;Ue(<104#MvF2evf$ce2nhW-vvtmbW_#-W#pu)maq0vrdV3$w&(s1_W zD#?M+)%Ee9bB=tXd?4bae<%8uD4AMsLHvVy)(GYxwge>T*s$?7_q-pEc6r>>HSdgA zWo4-UNAg;uss4+^Ge;&>B-7az#b3)vW!H#ecRP7ldSL444#8RpERpp3e=RuLuG+Sp;=BoAOdZI#KVea#iK2n9;`2bJYn8A z)q_S|^{pGhEAsn35AOSzvI&|AgxpPO*v7Sn45(Nca+onBLkR44J7Qr|(GCz*71y;u z_!%5qc-7io(JG~LaVZqc!&;=cL!k(&fRS}C9IOESvmNpxdWo?%o(!YTm~FLHzt9e; zL}Y9p=m)0lxXo-B8zHe<<;)G5(%Lae`{(l6VeaCep? zd^Y()hh!+ULSK|;Mm#VNjKL|)3i&J;1O2SUu$$th=`6eqOQBG6j0fZyvLoV5VE|f> z<>6U3&i+Mwg0md3vushyT{*t%obe17w}QMU%h)TT(+F*FSV#dDh+4aRZ3D5pkL+39 zO~I(wK`l^e;;I!W=y+WmGC(Hyb z94%jx9v&*h8NCSvhK17;;7p6m zYQb9aC@mOl1hN#eoq3v)wvbS@SPn{ik-X#)RRl(t(WtbLJQ5#dhm`9!RhDuoO$$#y zk##E;yE?{$(?)XOP-*HaX3&o_{jIG&Jhbi)_>Ky{X{TNJGH&&Pei2+^EC$RdGcsm^ zD5Gp*1eWi}ju)5!Ho{5_=Fy-M%@|5a(jrE0q)9%DW064;ct9EbkmtnY;qkOtwm9Th z3bAXFfBC8`jZFSw91Y`2?&ZsFz15tdGh^yZ7sC6W&lObICOSinxP0(X>k1d35;)B_to4!)+Yq*bRg+7uNUiZtmXqi+w zCj=728P8a%rNg|tlWZg}1QPUNhCpJ`%n(Sf!-|G+k&oal`B+xsRuUm&2W$b$NZ$(b zi~2^<7h&u-fsbp&Gt z`DZk4Z|w|vyPm(fvok)nGbr7YJ7SLTe;?1GKk_kpUyK2oS>&iPcwDE=vz3b01x&Lg z_DpX=8BlirfNs%8+1A|K*J7~3ASnG0YSKbc!7hm7fVCFx6 zw7I#$!1b5C6hGFDuQ!=(HQjV?k^WiYT7S{BxIIA zzx=y!!mDF2Q*#U2Ennbf%Hrz`A;x^2FnP3#6;jrF?p;~`;F;?ctPQ$~VIPp=j?)p6 z(YBnDA0!kY%xZO--lGPi^YxAD$ml$-_9JD?@`Cr6vZ-Xi@IE4<+hi_X)U3FIoZ3^C0~D2Pr+XE))F2*cSW z_q-S^@_Mvf;cYwH6LEKa8c($CT_d}S-oopPY=?i=$b*?8LQJi3Z zliHkUxi_<(Nd;1hm9Z`6#NJa7_ZTFH3tNBqTz@O? z5~(YCvL~w>#a;5Z9Wp?s)L8eh6;E3^PS1FsL?Ow2uJQK;`9)+_PfqnW`{O#IEsx)y zy1ZWb;6u|)hHQcLkQ>xjNV_3}UQ{Pisj+APV0)_f*6nZ+%hhr2dHs(td^Fm@W)`6S zDcJsj()SX*T#YNJtWS5&9<3AJ=n@yfGl+JuZ3jhG)ld4dj{{`4#9kAQScs4xxktW+ zFJw^dtzLArauVfS98=hXsQ~^0^a$JMH|{RjB1GXJ*k{L5Qc^H=jEGmk$yf39nt>A9 z6x7qmp3WFS1!7M<;H1wdG{)c;0gGR*r4&r37^IFaQxxgbT2$w|SzCmee62ut+oA

    &5?@Sn;yS|-%+)`w>$c_vrT2W)@9Kbj}QcFgEw=>=}}J5aN0Ooh4S+(q}ff^%1G zxCo+E^oFU+L3n7d3cH1Ptrk_$GOu&cFygciz4YO-Z*2dvJ8ODXe_2%c&C?53Oxx?H zt#{roZ&4s`(xB7n1tqGX;~e@hxkp-E_5m?=-G3a9-1z#^*GIkfwyw=+F=FfdqrHw3 zMkb_Q`JX?Hg&vv5R5#nr+&IY;yN)7jsiWQ!#i1Uo?Y*f3))2!$*2x-@yhtL}B6R&;pW9?$jWwA|yk-kkOm zYWX`q7g;uW*=F{_J?@TT?A~SMsW-~rC;#j4-P;oPwmyB!`yBn5)6?g#&<#D3zS6Q= z#|vWRTIm2=;@R((AVONa9_4y>ow_r;)zAetOH)rTo0Z~kpxbPfR(oOd%+hHq1%pm4 zio?!&MN!B&G#j%O?oS&zjs^TzUlD)ZvMU1@obXs3wbH`@b4a&(ApH>VEcUx10oqoN zJXwh+=8BF}+ET!)<=_ufe#Xh?2 zkgKQaGihftP0J+JqL;-dy%$I6oR!D9z98xGb0_AB1gxiSCI};~G zd%{mP|Bvcz({{9MYs?IC>4kBRJTlw7KBw5vC~}R8LRK*&wZr@lsKMWWUxZzH5q40Li%{Md zw$}qQgBM4{X|g1{;Wy-X0hYG775}g&ai}%T|GgYJ_a2-&>A-N!) z%^<11_0aPq&s#69qL)d^!nSFKKZar7Y|q2dJzabLf|2Lq^!D>tW9WKsX_>$zmVS5i z-Vw6og%L{`W@2eQ#U@Z>4Hbno(I^gpy{ezt4U%C%m8q!~q>zK-MVa&jP_}~m$sOZ! zbn2+!w~;1Pe7E+;drA33a^CE5enPQ8<9?WmLT5xCG#xqXR_UE4OAlU%RTrI1nIXn{ zen!1Flqe=M`HD>XxFXFrUJxJF5VqERc|@*y*(~Hd$QKsT>FS0sl&|?f3xvl|VVEAA zF6twlO!CZnbt8S-RhtC#-$U1GH_W){ks#cus1)}r8l`>Y5W$nin@1y)Y+Z+E%iial z9mUdd5;FYt<&`}6B+}9pe?bf`D~5Zn;*jShd1BaE=~^~j-X~9)*$2(TI+=I&t6Jq> zkarsCeLkZB>L@T|xP=qINBbqCxbI)z`&N^s_l~!bAuIkkBDGB6fq^ z9da*psF%eBf+dQzp$9a$PsK=(J@L^{;*}=_~ z%VXReHWP^-QS3g76jD(Q%2dG`7FpBk19cDus)wGlI`#5-o4r%Tz;d?Cf3F|%mSImf zP;-D;yk1!)f)iEHS@On|ZHgFhFlGraLu-~(p4f%yE`7m&nLlh;*fjZ}!UUcwU%SD_ zuFyeP*$j?nU%1FiTF987voX>C9N92Zg*|`Ocv5TzkZg*DRQ_}-3iTB1zEUaepcARh z!EJ#Dd_I@=1+@i1TpE?0t3c?D^m~XQNRm+PBgK%nZM;|B?44a7|@d`aa=3l8Yf5!Q=@PNCZJt7DJ0Vs8wCPRqxwO&-6@BPZz)L zwx*Y=>5Sc7-Ccm9xB#L=1u=kZf~YJiqOw`yS}ItmC~oX-R3Ru*sPI202~{K#&4Yx0 z%9;9&$=hz|+%M;zdzSArt(q|-ca|YMB1s{)#KW%|#YU3CKJ?HkZ;5Y+8zW8tf7eFR zAKWg@=Vb)k07-CEb{K%7r3WrObS#5qOb?~E#Q97Q|1Zm&*N(lJvqs)b6~!K*NIB?4YJlZXN0a(KRwQ-ONZ->AayjXN??(#@#2fhFvDO9@ z(ip(OxE%&@@cto(W=SS=MeFOB4tQXg4?Uj-w|ocyZI-5ms9We>XkCjEc8S^rRnsps z2?2T?o;Z-k880}w;LTD?sJm~RAe5))zhg?IV7JD_3^js5K@gd|k?T1=eA~kDPS=>e zRZHf0OaqVeAx<`B^5!R{;ihbsW@sQ`@2Ts13OZO{gE(0CheU~O_DY*{v!-bGV8sPn{ zOkC=Z%8xma6Dkm+%1jE$f#BLQ90%{Om)-NY`&ugV*X7Yvmmku=?Pm9gav&p+9(0MU z3W}eaOUJYKT}s2!y;C8qf^~NXJi2Kd5cqoK86o&tZHi2;T9Mq*Zj61)k1T|_;Wkca zpS1fYZ;56bK((Yfe?2L1BqxpL?hwU-O6NYPsSaJuZ;VJ5Z{#g#m15;zHcOuPJ$&yiW_z9?)r=#U^z z^yro+Fep>DnRyDOE1jxUAX{c7i)H;e;x-rN?&dT{_PaxREYS9wQis{C6Jv%93?;O& zO20{ZL$Xe>LzTd&n@Mg+mMf;5%cMFTEzq(69@RV=70*Ge0xPUEimLfQ^Bp6^RKPB} z7&6-9cUH~7W&9{MGw?W1b3*Z?ou@SKnRdQCU+!SPU?*n!+p47{@`tIK%%dUXHYBSXBi4AH_QDRgX3{W$;X00p87B@bf80D`oHOPBX4){qCAe$HxG~fB za#M0&P#qrY(HEBH-xzUIlESX`YgS~4_2*8|4H2Jys$Kc7lXhiE=b9y_bhBi1WeLwU$Lt3f0kT0Q)XOKxQc%2nBy5hZ^6efxijM`K&6EJC2=d*{*b90OpdEx7a*(~!U2ni(@ee3Uy=AD!_8+SxRU z>_rbDNK)bTEEu)vHHrlBDn8Wc2ed%V3M@CpiqopA0eQ}NSQ@5LV6WXdkmu7Xm+POz zdX@{&;NL=DV=||Myr1t7l)}_W)p#dT12rhS=wqZoxqRG)%vP8f0}PBF;bxT<9Q)cI z|8IFOO;XGNX_YIP>(1&%=Hdb@TEMJ<)ry*orsG8imO|)7W7y)_+mG94FEF^YE{{%X zNSPf6buSt@sV6A*7)5HR$bp5)JdFZtxplNqyZj5Kt-#M8&jP=3V246oD!NUggy~|8 zsFd6o!FF!k0+;B55j^Et)YttRHnXRY303S0$ z`$@F``UYc!1wxJUUipy2R?jXPolxvj|Ih>W8oj*C>0NZ~)XN^3({F{0`8&4sI~eO+ zSUUb#@8dKV_FK*B+4*sZY0uFs;u)LiMjD_q#W!cS$+gNZ&r2jzH8ig?R14MJD@Y%m zrrbg2^42(G>tZ#SpaI?mU{ZY?UQM?M+Q@P5-EM92LKm!+Iqy*?KoR7@nJZnj${LR@ zPaFIa#)^Go3p7@Qn}2TTYQUY}8P=$!yK6nkU2b6)`~5kPI3sne0g8Q0k%v^IZpi^A zA04GnjtV>4+T{ZZwMOxP#^Ot`iyB3is(|D=RZZ8bDtYy8y}Wof2_&l2TAz4`{btWd zn~~w9nSw@E3jz`cJmO%9Lbgn;GHPM@0z8jr_e?!bKz1OX+2GV9IzcCq3NM{hXuT|l zu7-5wkOL}k*UPX$K`)XGKn8_uC=}15Yk&(2B?{_gXGG0pz~yXUh1c5oozS_^4tNJi z?^J0rlrRoC=m@Z`fl==#hn2&1q;RS0ax!I8Fs$ua+<*IlvDBE4i3>tBAG`EK2K2zu=}TVRHuw`l223G$28xzs?V zT{?Z0jSAba;M1?v20%((dfK&ETn?G1tBamWa_Crs)p(5&R~PjF3&0IY`2t%b$3mPK zFI0GBeP|es_}f2hp7WYv<@(O9U;UP(jn-RY$7ncYwBstISV%(VQjsO@?E(zSeZ)H= z8kG0A-}S6vGF{K9s>APkHb%4v(k0tG?s}f_Xcs)=L*`MrfgkHJz-!=D`uET`e0$uh z>BRZzGqS}tb?L_p3Ij}xX3~lW=3V*toXSGeRwTQP@tJ8mLXQAB39-ZI^xS%RG zd0nuTh?-X`ual?KjV!c2Z}Q8A-Ov%|cEJ!(bE0hGR;Jt^H!^Js^=+m-G}@EdYK?%gUTC13*V#q)D*34oq|%JyfL{D$#2ZFe=(Cz_+$CM*H{LhQ>y%AH;X z5MTBxG>X>HdKLl{%8C7AE%`=m?Bk6(J>1rKyX^z`(T73BrnbUvmliX>ev}K*C5z6G zJ)y<&(kaz+A=d)@QRkTJNsF+`bxRyGEllhz*Jo?=yqx*dEV-4 zqH7}LUN5939!C~o2o0VUaKUAW|NZ7!Z_^&bSH!6>(~-E#e^02+tkih7DQ?S-JB=eN z%;lRlz1I>%m_9QR$MLJ)nrCV!>~`re!=a9a1K6~2)%65}iVAy4yqg+-`u`v1+LC$x z|M0_~#BKH1_CIB_e{YCp9Q=i=jNBV7N@~YhyfsEyyeAafPmxDdCU+V;1W&EPEIBg<)+?mYGhDMkY;gh<+HR?v+0b!&*r7#RYpp zkw0T%H|H=}mQJ(H=hfr`%-e|3V(`oFfAHY92D=gP&a`Qy!;UK_R~oHQeH7bEk!~vT zAFYbNK9UcsmURKuRPoYVuisg$Mia3+bWpy+wToUmt%6nelXw;jkfMe4Zt-l9+e0sP zky{?ve07fuD^8C3H;bUuG6R@Zuv4`fZvMtjIbF5>u@P@DcNUPknK z<;cvdiLn{~At!T1{jTD(4AX?rE3z3klRnb1MWxdpDf8X>LAq`O|42YPR2ZXa>xY^I zq=3(RZKZ6;K~D;=Rc1IPfmENKG^)n)ny5^Q`g{W}Hf*i8cr7#oDx1Z0sw*gsx53Cf zW6!K1T01r}KoK@V$YKx0?xM&}Dl*Hpm}zytx@gD&%?gsesJoHUJ%KqvYo*-+C{aY! z!%80RUDOYpTi`zw2>_+eFht4bg`KK;<|3>m11{hiBQP>-kW+po?91en(ki*D0Wkk{ zoxDww>=-b6jR2ELv5+R%HY^MPJ6)aQfqY`RLfc**665L)v4ag(8wm5o2=gUSINR-n zncMEe<(gHP4^|CCoMcn3I3{Xm^Z%RNG0(*v7t2Tv)-A3-rKp}5dXK(^|pNIWd%96ryeb~Uy z^UsIsrOUMXs;l~TASku%WD=NT{7*#_m3`(41;B%?QJ@>&JQPL=rA&7C+PrW07_3L( z&_5p_8DBE%QE6m7iYOM!XA7vv+f0S?Z5i^!9VI(~Qz4yh54|9&lN}Z83SURsn77}? zXk2SBXep|W@jsD49k$^LR`2f&$#%(s{*5L@x041)|9b^}sTP2MYg%w^}v?wx5sv*X)$#%>P;Tq(fQBOEx4-|R`cJ?3-7o*=9my(+T|tqk zQNUyQDQmzi>wjI$*r_I}tQI?`f-~aDbZ-g7{%9cEN208G`*p&t&yKfX3>YPxmW)FXGp5dN`pv?<+ z(wnAS2P4&_IIAq=jjv#b3+ybnF@`(G0mwT`h0J#hKwA1?@6X8T(R>;^UWRTP!R0E& zUZzL`6`2f_&sC&qRu1%urAki;HGB<(OV9>w_gbS`1C3?K?xR&5l0$#$!Z=|IY@EyK z)SnSLq=O$?s(Cr?RW8F=g0$WZJK!WXhP#sS*1sji&eJ}^g!Ar6! zgr38k<(M*WxdN*$Om{MX>%>>8|3OyTag=w55s;E8Hi;q$RAh(rIwW=mJNGC(tB4}nn@ugMNRpT7NN5B_6l|END!#-X?c z+7gZ8M)*qDMOD&SuIeb^LC{x370w-OyO&0xmx!5=))tE~V)T=@1uBfkwDbvT5vH*u}@(RF$8j?KnLHPy$A&0&>86mock_KH5EA+Senmhy7cEyXY+}s*FRP8k_64O78({W@S*PwMqZ2gBm@Gtbvx{F*u=S(yrUz zkNWa2NBO=9;=vio5a-<{J^@tzHT*IlWV$5kmY4BRvGJ&&5{TPR`Cbuq(GO?TFqQtP ztPKpw@DzOhIwp*K{|u>_{ZI}7UX zkHQ^W{^>K1+OVdeEdSDoZNk$~G5?s9g18jko8{WX#t83Ci<_D5bCS*xo}@2`FfMjM zloF8T+8=yH*VhQ8*#X_q_SNEAqR6FFrL|)GD;i+m)2J+VZ*e^sm`5M->*bx6o`y_N zI|NWJh;X-jz#|2uRo5@NNYT-S_k$?*#5M*WcLsag7M}f~ipiN#Ws~6-mG!`kE6aAyrTL!#0f$fKq0mE%0P$2!-<^}n>Yv?7V@~lv z;6CBGC5Qg?&f<5|9DB)o7wDvLt#a?eXC9YT-D9>MXJ*>CmIO^xz|X%7^a zvON1W)x=E46w=#a=s1r)3;hBWP>~0f!69nYFTv0uP~Jf2O0H8&Ff!fsg4XAdLmu5I zuXBRts>M4)u&t?!{`;WAKOVSL%?Isil(uM+>WOgi#E`>I*K3ef$OvhY-kA#p#;LAt ziYHKco(7awXjoD2?|?^x68V1heuMts>VPAHJRVZm7l^k)SPGcYYM3VJr$iTNzzu=^ zYO)Vo>pijJeYN;wlxH)9MKkNyx5_qrt+#Fl)8>pIc#dt`it*;Wh_j`##6=6sI|`-^F{=y|wR|Hrbl#PQ@noOK$62*M#Rn zS&RA_xf#$5UGBQh`ofSp**c)pzBg@$^b;AeHExAt1^1%&+rXpxt#Qq`M z^`~-!%c#HAv7q6l&v1f^{cbXfdOXe&gDoj>$T&+@bF2BY-;E*3F*1o;C^nfQNmS$( z@m^jQ+ex0e9rbFA$W|e?`&~YgAM;r8+EK3)%q22`#t>k}0pKqMr8RFT`u)49ru!Sa z^<&NKSa#A_NVCryosm3Xo_Xj3`%Uy&F@8NUKI@1nv3#aSJTw7sIhnG_4!8eQWXeQj zhVUoHeTP)XMVLBuy)&|;EISr*EM(J^F44sP#Na!$^t+#cuiNM&xPA9_OURpFb~M<- zu#Eit{G^K{^-GqcJZQ976i{p)MY5?#WDCM}d#sm2xv3bTF5QirVl4YzE!-8lyg_y} zKm%C@)F{GUgylP3FX$2_39!NGcc~LRCU<5R2*62Uo;-L9Ha$8U(per=Q2*i@m%HaQ&8F}4pG!&Bzbv7*rD=cZ(cW>^HVzc7=! zw|MoEq&JIUt*MjcIX6U{4&UtA?^57XDN6LY7;zsIqjtJhtIp5|ovKv{%!I5w9KUrN z{tAt!#{9QFN=dc-{s6na^4C4zHW-jQ^J+dNr|dX_)@fucu2AeHihK+!3q-mS6Q5YoB=(TQhp%-&Xg%HUid3sQ?28g-^|AzU*~0bwM9+tBJ^g-@0%Xt> zSENG@TbO7Nul-cCPJxVh5S8thLOv9hr|#E|@zuo;_&H8s6lH;1s`MUzQxKGN?)UGP zw(A(~fT1o`kOR#m$Pv;d%Op_psTar9D6k#}<#8?h&5Q-~<`*T4t{9`+%n>KhPx|DC zbIzF(#F$}LK>|~)689EO((58%_ru!#M4!(<(J~s6ZFzKt_@tzk4paV*)?w-F# zR7LPgr!3A|jnXd%40q(dd`xTH!3nvOE_QykdA4EMTGE`qo)mn^{N9sBOW7fcEu+Xj zDzYLxT@vMeU*0jlFeKf#pEO3?m+ymkT47k4la>_t9wP&;gVq@cjGXw6Bq?+yb@sYc~l{)sNUt-KAJ?nc9FO9D1qyrB%}}1+`0|4iUKN zA3%R(VMt-fKHz}ZCB$suN)@Cnu$Z+H+Go>!5*Sc@!nZlX1ZG_TW+m|5C&78%Ev{<3 z{z-7v^eopQhjjW9>2b%{Z*Is1sF10a4$5yc_)%rE_bJgAQscobSv6!hZbRC;0+JtC zfr#YvmC%>hBDl?9HDdCD%Ak*6vxgNqxgj6`S0JuoP^K1(7i$ABT<{T$KRR};3&R&d z)a{f-Gpqt%GBNCz24zZWIFpeZJ*^Bf~B6sky*KbPx z-m16))WHcs&9ck><M_79d{EgC)EpemuconPYjgySI(?RhJI%s^!v!;c!lUO$Lnn~IrvVF<3RG}`UqaeC-Ql1O5Rw^nDTkD3++*!V->D7MsBs$ibX5c$& zr_Xz&yPpj;X9`CBy2fyF942K9pJw?Z=2x~+LJyN%*1!3UpBW6$=D&aK4RXnjclZyD zOiBmEwo>F4Qn@xLTj+HGS|#XQAlb7{8W#7|uvJ8o=j)#fjL z^VYdmc)>DK8I;URSGAJ^v(%|m+mx7#&v46@z#+0i;4*`Y*E%4_Hw2(>Gef>Dg3GQ^ z5+JyTEPZD%?ltr1yP{MeXk+kh_<#`o$>AH9=cv;CdT8vxSzaw|_sW4fcw`5~x)Iu~^$gK>H3thaQVV4oRLBp;$VMyE|LkC)4@j03*{!ZF(I4r`jKz*3rBoz%Ww|RK7S>3bLuK z?kLCvBs46D)9kNSkRRCE5H=pB)3oc%E?}q9*-}s{eH*sqc7Lcx2 z1~==B+6tdh>;Od`L+LQ=jd54Jd3w6|qQ@sXz2+DeqvPl(p-mZ%scRI^6zbD6KqIVF zlHA#2i+|vu)5QI^=K&&O}Fi1OXmAHz;pY zMDuU@Or#^k(K@}%s95d#>u)XST4bgiwWErcs=pdx>B=^scJFN2_e~YLtTj zM`h4zw}X-eKW9c z-s8N-A!NhI2-_4sJ(ko(V%qWauV6}^-W0roTMfd&7nvfriXB%e6k!Um9u;`!v`RI?fyT|a` z=C#nM!cr$YL2p%61Ys%j!RfiGI0mmlZ#m<7d+!*hkIj0o8s-`bSb^g z_3aDveqXD4(VtJ572xvuSk`>=z3+K_-BfJhd3p{z&V`u~U`Stx^+7=m%6@5!EA&-k zZVH8cu&3_4N45(3w6Uiel_qUZt28`eV*tVMqc(jtzb-Ytdok3s_40Xm*s;wp(~fyV z(n2bn(fjCQfC5-?j%eL8p!s+GFT)Nt?!nsQ}pjhp< zc4_XLTYuiZ_#CNlN%UMbuTB0WcoQF=I4{g|u3_FyqV9aN_gn8BR37nbjKF{I``;#~ z-}-au;&GdnF#yOSpJu%1G7q2~-}&F;%M1>2srNZK>EyOuw`0o_YqVW|OtFyGctAz2 z{NAm_$T$IBFa(UDzOXvHPIllsw-&2=WcnShPIWg69NYs-5_y~49?NoJTipOO4mqkU zN~5^sxt#%@CzmRP8lqfk{apO_esF0RZsH+(-=@$H0|!Vs)#{GAm)R;E6(v#(6uVz{ zE>~^}8gjr-N7>9a;J5%PE1)}7=g;enL$P7Gf{!te4ww%d(ZqjS*)Y+@hV5G zggFj05XIsYRfY3_%Wg@xa{;fyxeg?5>je4k28ykx$T_4NEtP*FiGxmw4j$^fgAG|L(@ao zp{j!H5$yz4%snDK@7wBmk3!YG@_~h=;z5UAd9`2%3n_&`IaU}VcV>lX1KB{L0_$Yy zbe%wrX@#ok*CCqN1@S;Vy=I(S52V8|Q+9(n=ZQk!&}&ZWMAcBdc-^^~8p%cKU%oo|7($E@jwO;hQNqTjWTCeVI zW8-HuL7b3b^rUSMI3Z(l&!Kk~%M1qR^jqJ)K{nd4!6`B_I6Ejdog!&i|5Y_#-8ZM* z3o1lGEC@q3DDY7rOc(34boJEPLk?JRg%|a^pNI1fh1wi1oIoD(_ z=z0yn-zMEO5*Dbt334T)H&e*lG7}|yTc_sH8~AZTs7Gy67R?-TKp#Y-Ksik_7;{J)>YNxGJ)5RIU4SI?cyX(1@+Z0=wawPi{HjG=fiW7`R z7$ghcazV{=_nz8O=ksOvQ0kpkCJY|4RSx4RiLd^cEpw)p8$GTE;UQALs>@*`i~)R%%IR37zu5E?Qu_aFWK=Af{ueI`R-N_F z2jlkZpZ=E_7d;K;qkU2EO|s3AfY$X0r*9X<0;y*X6}bvR_YFV=ktHdn^O^i5E16ZmCJs(=f0ptP~zSn{LDAeXC;&A(r&|_>o_On z^DoibZ#Yeg{l?$2sDP`coNJu6N_I;e2C?Olpx;Qbi4<8+fvyyIrrzp4VV`c}{7LZj zAhbIZySdS$(L!un8@ewY@A%LDN$G4bA%9h0`X3VKNOFz7&8-v*8uS|tp@<}&)(2T( zVD-)r*Fz4agLjoZDUY67%x8e$7_W? z251xx{qq5m@g-xLs5HWH5ykGNNC61KK(izWxVU2ozQJuL(;`TFJrjzbunGUO;P~(| z-aa55uXCy4HwPc`8&IUZjy_5W1JQSgn(Ubx3q86>(RZEHIpMuE;YqfU8RLfU(cKkm z;MsHx^Z5SFC*PNv(!H8tRZS3MMk9rNDmfU|5Ts7^%j5MbPysNHMm}g8`F*DUQ-+T% z|EhlC8ne`TwD+Pb*hLNI-Aoft*S8OVBiCfSSWZ(QDyLN z%kWsM#D?uYK?Xn8<1&-a94Dte9(X+S=nHF%SP_^ZZYI}7nbY&2W>>2$3V+~nQhA1L zjEMJH5s1ClJKQ$9tsplg>GTaK%)_y*_&c%uH!!-XF+OX(-$%>TM_Wvl3hZ`EY(_#M zDlnC)f{DW%PO?vd7!4f?Y8Ni<;H`q!W$9q`TA(HNeJf7ri>H9=bia5s>s{q!@+M!t zwf%=q1}EjjSE~O(R&oo`+p%TYVdTCfQ*07N5~#>M;g_asl@}LWrgNQ2d5H_JE&>%h zVGn(A!42tzfoFKSEd|l={S);0YKpC($RR4SHlRh& z5A9@H<%J;R?`e$qaOq76zSk=Mew!5e$Ea@4!@iDa3Xg#1lP(M!;6YD^qRhWnG31-7 zD)ZgI&zwFezsw{Apk`r0(9s!JM8i+?@Hb08cE{(2d~?A8-YV@6wg!SM1%>f~@scAe zAcvD>p<2Iv=4az+8nWXmZ8J?nsZ5D`{S5RoP?k@9VL_cBK}S}0gJ}&$4emOpdT69} zfI0ukU z-4(1JDM5_+pf$1${GM>FN>4(Z=<`Uf?t}z$7p))28YW3U zUDm-VR=C~#nsn=)-!#)!=jTCT$7aJ!Lo+T7QNpTe*g9WH>lMJt;a#f&S`yRg zkCX>I;@NiTap=(*a;SH!6m*FifEY8CsaX=W5cf~3=b>%rU%I?nr>tB$`!QK9+%k`I z`Ktx9XBlcfHwfn6Snyd~ZZVIVx!v`bOiPyRcC&1z2&o99hab5d5oHK7*r)u>yiCiU}wi>snE!q(NgR- zifpAKPZ14IT_^2u#^z`cFTkQ}^;YSpch)Ma<*~Ci%qjs6D7+tHxD4KnXRk3^W@?pfiWD{qxTg_g z<25oPXLC#lF*Pp3$0G+G=X>)8hxty!Ra3esyIrHq5I`YsLVT{V^Z(eFk$Z&?quyMdwE5gEbC{8e$JfrE5Jj?)-AC>LO zAf?>mEq1&#L95{iP<=?T)fB0qA`b+@M#u5O_iJCzcq7j7B=wzC7@1l3`|s=^l^Bs23rKis8hNqX zrN~T_1;c9Xp=u1Q?@vYw}4NFkOnO__snRe`{R2eNQ#S{y2n7gRRQc_CNY2@fv zSJKeh1v*M;p2rCgcFX_1Go%joJeX5NmW*>w8Dce-(CXf_(%TlEa{_@IR2_&4tOUZ` z_z-MB#apigZX0(q*K^ij%$Ru&Mb8aAt`nRvGwG)%;%1v-hD%_@j*|dp5QbV2SvqdP%rCM#IwU~%XmhsJmyj3(?VlDDI=sRAX>P~TdTaOjDrvhUe_opY1CWNDlelp z(+c0sz8S#NWPKZDC4`L&O`|>h&Qh;+?gl&4@XG_sNQxcX8BlW@;k@QkEL4+cQIVMa z!_41sUJ!eOG*hq-47-8WxoV~?)29ThRkd!I6a=cmEjkquOdG`pl!j`R0}HEF+=*{T z`$k7QH{RJFZA@#N<7Qu&HB>I2mS^3W|NN^q^M(U5 z%#2{0)7+5pUjG~SOgohA)}EN@Iljlg$H(=iL8(5l;G6_?BPJFKhKaNU5w@B07mvyb z5tFz7^xHo-<$HTYJXkZdby&vIphP)!WLG~aS?OBu21*{1dywm?hwNaoPadsR6*Ft+ zBg&9&v`L!bR3ECRB-JWy3^q$(HDZ|=@%F{A81GfyA6Er0HMp%$d6_>am5$_^k#~5G zV$V>}K@qtzaD}jbdV%k|F+LSR>-m{3DI_U2q@u#I5cFE>CF4 zRyHXLgw`#YvNHO|fb-EGk+02&`GWyAf2{wnS>(PQZ=cs15okZ9*v}}^M@9C#)CXym zM`!5h;A-U;ymV~+*f4f#euWZg*dIY*N^cn6qnBj7>-kioV~FdO6o%C>XGk4L6Jac7 zQ&1OF+pT@$;gZWh*Q8^AJRhA%d%C+ zz1sy(;F0^GYf0sdHBge*R$XLf-d5JiBTKA$=j*u(hx47OHoE232--*WqOB8In_q`L~Zhhh}7>M{Eqe6;q`OQ2{EI) zU^W2{C&W-;HIB1Pso1&fz3kVw17e1e_)4SL6#Qc&`=DtSVu6hjn>{a)Oi5ArGk%$G z(W07Z8EltlV??ohzc0SmUm1(fK9*VK45qPs$QYhcCj4qXV=yRxIMsKOtl;J#+i}${ z)K!fzKbt6a14ZJgNIiobuyEHbLh?zF$;0iqI)R@ zu_aHQK@M<3g&iA>1|w7)r&y2~u7+fjaJS@(%EqV$N5phXF%5QMYy0&_RP13z6(tWjS;1E`2vk%W7y#IlaKDXZ4V9t-VSjLQ@V#j}CT3L{3l_M} z>sanLkA+c6!6*M>=#Tj-13pei{$mp zq@Q#Jqb7#>6WJQYHYNVP8$qQ5f3iak*vuXSA|80dvMkca+-9(LDQzmIhzEGk{WA+6 z(#20shj;1Z=JCRi1J?bQL!s2xsU5*sxez<^^mu1fDto038Kwb{jVk$;M6 z1AqUL^=y&gF{@l(-MdZ@E8` zUlDaF4o+{I2jbyN23)qg;|Mn;&7w?I0&@~n<&&4}0Q1urfof*@v4AMd|D=2;-7){E zj$bNDxH9mV>ZT-@t`2XEz|`dM)3+oK6wiD&^KXP-p!HXyzJ4X*$(uJNKny(MQ|e?L z^RdTz!*}(cbwhGraLYYf*rxbEe?p^JH}lm9&)VJcUy*gw zTVL1{E7&cYG*a=-zW1>ymoJx<&W;gdhLRZc)_@)hWlFHu4!m#oT=D9V?cup0(7}ym z7_IbUSKBOJU%+AHJ5JCT`Jj#8UsgWO|FS2cqJz$bpl5#0|1zw9QGxq?+vZg-+9Ey% zlxOF}aS_Gg6FJk+ECRIHVw~l zqrYizb^>B&|CAhmWuk*^Mq9;86#Fp++sDWT;52KD=nTx8)kg;9Pi4EO-s6{g*2z|o z9C`<^>>=6eNqG#}L8mS(a!rJUb(FAt!9L#yfoBEaPtAJja)u=O+~m~;U@m;i%zl?U zbMdZ|^1iVBWF>^2YnaYZt*TM_gkVnnyhmdMD!6?lLiAP&^8AoX`ndNM2m;~w8pUeo zHido+B$nWaVh%7I4VoOz8N=*lM?dP9ipP_Tu;XeSGsy_lJi+QejQBtlu`&oYSMsEA z3^H(>)0cg>FTZXnDqcR4b??af-H#k+8O+UJ-)}oYcG_{s@3@fxIY_blDN;g3meV&S z{Sf%Q3=TOgQ<2A@pTEL+wZ|I8X6Ytz8&qZG1MN5V#I=BMc?S>u zAibKwrJx6aX|e|%4H2=9n|w3aKBsG>2gE-6;Tdg2{@bcRO<%KJ2mBfj0x-&^6;Heg+geyi^tT?SwN6jl%G=ssp>3cgAbvuz0 zME%rln@1k~)J;2ES8UPlS~bO*OO?5=d`werZZIpJ=d^@P@;viKy(wvp8Mc>tWv8lz zlrApvM;D_$Xw|%Q`eXMEiyy$h7a@*Z$ZPjp5%_fBd9hYm%bfJ@;kS^e*D8Vv!!~<% zif>5D7hIQgLbC%#gY%hEx`+REMbJiaEA$DuJ+Bn1!r`!okB5i-wW@6TnWR$OE_lZ8 z;jbQdxPMHO^)efb*T*?=oF>bDBf~#^WBGuoFq7Sy88aeGSG+TTDSSP!duo+A!u7lZ z;VIL1yVa=jXScdPoVN}D8yAqh=%}O)w%oM=Bk%5K_J}HifVES#lHViCU@*rHn{Rrh z%qDio8dx4DY)5^Z8`9 z4WG0eM8i)^%w!?a-~|C<>=T1DG?*Z2he2rp1*#}%_*Ax(5D}vUQ z)}XWVo;WpuT26}~CEzl&)#8yx5k;j0^oM0j+M#wek8S}2a!Pf78Z_ua#!0JO8w`)< z(Z`)msrEV*1JUphl=N+PPYuNn(@?Enee`RV$ zLH^0`3g;wQ>6D`a;1USi52SkQGzTK>@h8#qXo+N>8ZPTcH~C3kbPS-hzJ zzx-f0QAG{BMb6nVXzm+<<_5(!QKS(ITldI!FxNyk+;e$%$iSS>=JbbkC~hkH=Jw2h zwo4$L-t5`$lI+tetKsjSnJPU)niLzHVy9}A_*|!ahZKq)puD13SCMs=v`e2Z?D8CR zKvB60=S%`r1AYVGdhe5Tf(5}s*8ScOy&5C3RmJYj!O{FD1gn@2Dv$UThOG(T?voX= zSz4gFF1jOM?HCK(Z}=&#vP+VW(iNZ7M%*+itAMoc3I42^io_HU2_=tBQ6x$5&_Dk9Y8R8x~>W4QZopP+Ux6mY5ZDK6O zh1EXMvVm@{doYpw?07jUH=5~P6bpjdIaK5}pB_3PsN1tyRPfrT&_|J| z$eE$9xI-oFC}D%7nU^7Mj6m|oc0s2+!>yN>=9K29oq9*!7?I>#4H~5hO-^(zYi#L1NO+K3Fm5>jp?1|JAko4$Ol`}a#LoUsbQ*2>#JDe|PhUJ9c{u$PxZXR7YZ#s2M&msU7$pklu_0p?^3=mij41i*3K#0kCB zwqJ%$Gj&>?C#$hvkP|bm2{ygznLsxwl+z|p?sS>(M&rsof}7^Gu^ z+Sq**OYvtOYNCo`4}A@`=MQ_2C6X*VwmmgQwx^V0!5HnPA`wDJY}Cl;08`;y51N4| z<=HbZ7>h*bT4f*k#BU{Y-5HDVz(XCbjsQJNU9}yGw5X95R+rwc4J!%6On2F9Ypt%! zV?~5mI{6meXC5J=5!w89%3nfES1c}}IXmvmHnYAV#&ih7M*SFJI;0lW-A=$D8-8S7 zZb%Ju@MF^eZme`n#u%${zWaXF9{B~g+JFmQ zNyOp>X3Uc`Mv#p8^DKSv&k;4^pBvVni?gpzB3E9SSj=Oi)#?t#c2K01io`tNWm3$0 zOzzB9V-ZE2yi(N5+vojo2L7uq6J~w}d@X6GA4xi4^IpTWDXN$**CB^` znO?9CmO{{|N(ahTzu{L%$X!IwagJ4Q|XpxrzYpVD)Rb=}sgSEp(SldmpKtG&EMQ)k` zezIEuRDc#z2Y?|fBczaOl4=4w6hjW!09PA!7}y+uaN=jPfPCvN|7S5=S!FKaYCMP#=f`^F$VGoo0gf?^L* zq>PHJ2!H0gLYNWK5wY6wHiN%woY(^1N=D1ebH-kjLONUZAuq#e|Dp_NzrD?TCXHQ~ z%)>KB{qf8blEG|u%Yb`tGZ>9RynOLI+JYK>Ca+Ea8oX9hB(nk{jBSE7-~eNA@{*|@ zZT$PD_U8F5yB+(&W~@pB#P8J_o%AxUSj99})U^nrgr5fFs*ZuAVVwZhkQ`OMZ!TF2 zgmBn=pl)?z% z{ybjnIK*RSf8I_blL&6D^=nYQY_A+Q_a4r;6)Jv5J+cPswg|24xZ^_^Y6 z`YlOwBw+hTECz)Xn@^EkD)KY7H!LUQ-2&gZfZc8bE*X;d^k*6(?giG#J0fCUKlAQ9M7)g=NClOv%H*ZfU7{?>VPCAju9l*90Pa5;n3%63)o?Xv z5hTiXg38ID96RNU#RH&@jk}b6@b!xLuq zs~*z@PV-~GhWirb_4cIU=u1}FpUE=^yX9CX+l&w$PqAw$vKq@{u;U(Bfgo^(?UA-J z6)(n#B|v@g2)4P(38<5|`TSs!n*k3$s@Z;-Y~mJ;vtOhW5IY^=erHi^1_f?QWU>4{ zC}^Dbs1vj?U82kWn|TAE7Pn@>4oH9_A5pC1Ro4fud+1!BRgMi2?`XbP_QRrYt$FLf z;&*JNBKHLjEWy;Uc{6(bVhEeOi~QZ$Z#o)K64CyynnZI$i5>5jp>SpdN|Goxfgv)knR9yzFGm>BuQ#Jq2Pzhk%Qj@NEmtIcf1^-SOG zG#1ph&`Cs{F2-;NykP!Eox$%Xy>g7MWjIy3LObjQQ5JYmm9(BM{x*|Gr@HP?H3Sui z8h&l5K#r9L= z5fwQAA$j$7@za2PlFd_k7Ze7ULf+=nQ1xEkc6aq1X$FI5uzR^$9y<%~C=jDYR|Zq( z)J+%oLLHRv36Z)Ooa|!3Cr;`NCY|2vRL|cAh25Ly>ARkzgeOGTV84VTptdpIf$x(d zpufA)3ooA#U51|7%Am`FE>WD@reLVh!u-Kn)d^8tXn&YSaYw3u`r*`G;K71woC0xr zFgAW;vn29PpMb{d0wBr)H!3|i0aihMSNInrlr89X|8)Liw;cCZq~lsH6U+)Qps%E> zIELiCG7DI((E?URvHK`+eIoOjwF`OzcP@HN(3Qbi*OEAP1Nae^AWc6qI5U%0xNxwN3szXR4qsCq`n z+nVd4HLTnSv&Y4dz3|@qlXJE&Ga#qb`<$F~azlSTEiT}AX7ELcrCCGcj$jabI4)~dRI z;wFwn3w4LcDS1-9-MvZKqST#eQRXXipkSiMuRvVGWbsPFo`i1TpCtP8Nnz{R2OyxR z-sjsQNT=IF)$PhuslFg=P=3%!pL4fc3wN!0UGEeF@^1g;`rpYKJ6IkFjG~jkk?*7^D8KIoX*}N3hLFPTmwXW$(&| z3IjH_Bn4a{o9!4Iupo`tV&zgSG+t*=k@d0!agM5k*Q;m;X{G^}C}CyLdJnbE`w3Vb zk_?L$m=~=|-=wBbVAn$}_(#enX@*mMs5MugQRb8rP)4jg_;!@xtageMP$qvI{mvP= z0VwIO{r4Y;#*Tpk0qYS!$)Q+a#7d_kv0eu27%)16uE&6j8WYlb=7!QRtR`>tco>F! z#dY!rd{|X_6-Y;fB#a*^6U4SPFXO<8)i5#IK-_S0?!$j8{j#T@@djkR$z@p#Z~C18&_OdxgWj;TSh z+hDR8ONY%#XTjajbx@WBo|a5^g-!>EYZFO%C&1WtFzY7X|r%v6W>Q*Ewx`JzfJ28rC1=iO` z@}#is?pQyrQS2u1Y%^46?Epod#)t%_kQs1k4L(6515Mci{j+^cCe^d}*5aemM8)YD z8-uLt9giPph&A2+jk>gq_2dr+Zp%%>Z)T!ghy9Cr`{mvACyHO5lbZ@IUK*Z!i4c#qCRd;ZRn+$mz#xzf}XpM}Ctv$KchL&-(LfvdeLlDm^>4 z9jA&cL*JQOi76=M_V7KtUCa)t z^?nQ7pgwFW%)VL;?&hKXAGdxyuhE3GQ!Eg5+@vD0W#hPa0<>x5 zOO`KRewFQ2Ucym?AU&p5zWA8A4c?A1eo;R%gbO;d=80R3&ZyN zo}kbCviN5Q7OT&_xx5pka#7murXtw~TIJ-cLLQUVj;r{&keXWK7U$R)feFT9d7TW; zw#y$tLA%9#&?BJqqPJFr|BF6`^R*L?{+jh=j|Fd6>Qe;w_?v<-qm=B^MVI((06L>I zr!D`~F-wiv>N9MyUk+WqKyS}t-EI?g4jmg%!D8N7ecV3*nzK(i?GCjO8MIaApY!a# z`~hy}?~1?kqUna!Y~{P2>q)vJsW4j2_EId!g630^KoRAHI>VTjDGh58X#TCj`LSO) zWb&{LH9n*?47V&;aS4Y=Vcs4XbZ` z(EKw~mU+94|C-^AzvuCQk5qVBJgAuxb$Vz(1@MdiCK?+S;Q}3FtIfD&198H|WS`K# z{Yhe2T;w0T(LfSM+i2Twpa6*RMka~UDK?EFDO4mDH1vi6*N+P8aNCjX2iXuIM}&O` zm`P}%A&>`@LAq>$o+oGm`kX9Ajq#SsvZ)`(UFK0!@G!c0i2)(=mVewrD!3&W?Rb*| znmHp9jAtqKG(}EOky}-}d6z^;VTQGkCGs9wMbIOc&9ibtAJKI|jS&NGtA#-10J;U) zBpn!~dcyOWkEOkmhrC^(cxPio3lJHvM3x~q(kkx-S3nEsNzy6rCg(-_As~FnC)Wp? zLN$u5sye||lCV&JXUvqw@P_%Q3YM(`sv>y0Ys_oj7?Ea-yB8ge5p7StdDz>OCfaUS zA~XEe)zT*FPUsF*58W%j<~s;sxGL4vMQw7_GdZTZ?y0{vd}g9nCeAZH#u?=Ln0ebW zO0e{d5Hk66JHUoUNcG2 z9910+y+YST5a-cVD?rtD*d`#6Qey}RdnYxD0oO{=#DT;zY<%uhOaLnMAMx)iet+J@ zU@*$x{MH4s?v*hZz<4#nU~HqI_3`}Oq?j8z>^DC}&KselhGHuza+r!t z^-HHu_+hXZ^ov6_@ZUxr?{-1O^tW}@!to*77NS*&5k_G$GD_GkMV+J@z6h6H!ahR9Bd6}r=x_0VFS=Ch9+qEhBhb?A2-^#V(4>Q^U^ot zGe%#jpWFC<%fjd<`nXfO6kCQsl7a^Xwzu5VLBU$9+$%4ejkAwrhvS}e%V&lwIxHV~ ze0L5hUgrx97Utd3DSsyG?YKt>D3wR7C}|X%f)Sj^{;*p5yk|N+{MrDmh9~Gp@?NOI zO83cftqr@lAkP_kK-{}z6XtQ)>x6UK&rn{nDc9dJz~lY=LwCr@(PFT6yvzXO`v}N1 zQtU;FT%aQVI_R*&r+fkSA~#7?;CSUAR8T)dT?*YVIAAh7)+IZ}f9h60!$KRy6>cwGbS${x|FQQj za80FW{1?Od zDPHi(O$8KGE($~t5y2bET@*&VfTBk61`^P5Py_}YRQNwn5@tvwniB#W?XPys$+;!w z{hsH2-skfDT4Q-E!>#MHoF?;3pau0!4GmPOA!I>}A3MlZC6{;B?0NV2TSy;+_PW%_tV z%B(C=AC~7rJ9M5nU5=|4eN%K2sEd43L1Q(#`T%K^mPWMkAL%>JYG$`cO2{=k#9M3x z#WN@|+c9R)u(E)Y()+*l3yu*&|NUuyJz3$v5Q5gi5iZOoibBXY(jniwc`Rz{U$%pD0FjWR5d++fL5=PWym}748+5whwZ>lu zIuf{D9@M~zGE;%X2y=AXCPpg9Tzqf z@n7~tGK5k=xfACdSj4UfxGXeCivXKDP-&#g&nQ8*e52bFaSJqr)XmSKo2Q=+oPa8( zNw$r-U553!H?eFxMp>gxx}Up#!5hZ8sO9ZDxn#csdno5jEJQ8ER8yn^*p@+9Z z7kM%t(A(tE5R0$#YZmJT_8<;Vm-?3bsCUsBvyXWl06vzZimP+432S)R_iP~eX_rH= zAD9s`MMF+nxuJI+y`J5I>zsIzPNEAf;{%@~Hv!JSEZoOg1KMhj$Z0opg2q_5ZySs( z&OoLIo(q)cA7lq3-}b>SVuVq#Q|4*tb76G@2X=OHOhC7pVv;Dbfr?5FsT1XM+hz6A zJYT4XgQ}+<`rOP~@xcXs(5P5K4@|Ik%mPQxA9y<-GDUx@&2f{h?m(?zl`KmXmsH znqT)K*S(kU?JeI~`O}1-?0hfp{pug8H&P$M|Fw|E(y9y>wDIdA2V8LhmoF;j+~jX_ zhJ_kkA8~={n6Fj!lR@bU&VJui<~XNQsa0u2d%2(-8>Exw`g9ZQg2eqh#beTI8(-Yn zZr*-5%gO!kE5VI+PbaH`hDNo45)-}f9A&M`_NnUIpeKx3Cc_!F*J!gH57y{l`vEqd z-$(xY(|&}j>=i_eH3Nx)r)?UJc`10P9n z%2LNMxgohwu!Uc3jf?t}u$t2u+~i)sS?y85ZjOG!r zuG74f;uQ%2Stz_fr%EH}@k?9S=XcBV0NE6z_lu^haYwXZ$O-$J&XB7!CbF21HJB_6 zzP)#^ko|chWi&)j-JYHygAS~|1)XXmT$p%@SxB(Ly9Z(A-WToOg9@Sokzp)f6S$2ujL#O&T^2>vIcm?4) z={eW(pl+8;akn-(PkH_+rmNDV9i8MQAynO@AV8cCYbobgkxH z_wDhlqGP;PzTN|Y<@GN8;aU~CE65BH9 zMj2&0?P3R(+^NTue_y?8ozg_-Yr-TS^Ve~NilvdY4nmf$9>4wf<@LYuGg^?kw>FlN z?3b)2qQ=Ck?57xz1}&tbE_;_iD}1snI_x2>nTwjn+db>01JDwmAxeM-_)N|X*>MN~ z<53Lm_IN%4m!#Vn6#yT|a+MDtrgn+! zqu3iJmhH=*$-MtDFNbj0PH77T4p>U;HM6+3(8z*QF@>@#AKmO?R`D3VA;6-3^WeT5VA}&&TzAowY+g}T(Nyd8kZ6O>ZX=`<+aE@M`@@eu?0UDyJq>ypyXe&+UG#dl z3_V}>1?Mc0CIF9M`!m+Q+*@+*`+bpjU)v$+avzYPv$0p2$jfxTO`t`d#_MDJcTkTb zk;;3IOOLFKQ{je%FXha}5S$fKIl7q|S`&b5xxoHCmIvGRv3jA?Fs85ytFdxi=QAZL ze5+!38p`88NXOq!io2-&!i{FgYZiKO5aODREk6n@2kqr(6WG|z7J(-vzOpy6~c=`6xYFn zyXd>(LLo|@j0L}6J=|Yy*b&iR-P0Z$tPoB8Fg4`A-Z28_PFVGQ@+mw2#c>}A(rE&- z%M^2wBIl{7&75U|L>?$293Z{2&d@Xo>R){L)Hj8P#uVM$rAtdiI?%yxmwifl#Ogdw z9o+=ggs8pHF2lZX@RXZCk^%J^;+WMTYK&3i>3A%8>e8h;@W4xeb}_}Rk*?-oNC0nC z5{~EOyR4zsP*^y>0UT^78wA2@JU4l20cH4;qb&A21xI@LSPp1~*|j4QtuVGnO4s)8N>(qSklwBg?PazDDFlKcVrsH6v{0 zt?YUlD_b*pX{=MNvk?zr99YiNe+mG@?Z@2Todnf%aafKt7CvoJX?wV zamj%lG9a&Nk@WKOMb{tw}4Zh5vayAda~@8m*Lt|oYvr4qM@I- zLfp?={VCb!dC=>&IMXFF1Y^-4;DZsHjKBxHLO~2g{_OA--lU!1iic6^kuXbrE)hH~X4OeG}9^$SD7;s%7D0lh# zZq2GreO-DlH^HZezTlPmTC-|12TOb|c;#`iV*Qk|knREhWuI*0n;k$o-6v_}?{cdl z#UTaZhF7dvujTHKNL3zdH~tWN7M7zy%6#8X_D0V(7dm#>Y?lS;;*Q80 z;*W31$|Es~2mWtx7kx?CDJ=-^nA^gvoud7x^0)VXZ{K$k790LfmtTMX(YFjo*G;i% z^vr)X7Vm8ycov^w%c-nf9cs$=zVTp+(Gsb2e^8TU>@1N3&uJMZmS_XT#8YHFuHJQY z&FpJJC;{}ToK+!=nNtC4n(;={65tTO5jtN#S#M~#49|9gI06_*rR6^7HgXXMW*ciHHX;FCXR;wBrG=y=hf+U@v>d`*P8F>zS$ z)Pi|<{Oud!%PQS>Ziqp_hTrW{&$;D^%tRH!zS-5B&2x%5RnkuV@yn{SaBeZDn%CqJ zEqMH9FUb-OpI0HQlEzLOH$<=mz_3mD>eX%|ZgwM9a}PBoSpGNpu$S8KK^e(%;H(j( zKu3hw_EAhBMRr3@H1OniA4Kj4qHwKhC9jEdfxZ!uBEo9D0a+Tqa@H5YIZzXRXP#EI zTTtu?3doXPZWlc`JKy_^+ZVyjs_uW-_s+$4_kJ(-tta1#UwlE8E?@Rq_oDFt&R8*{ z2anl!+v6=e%(!3q!S_EjV&=+^W5P%uyOoUtCyX|lP*1F!%q*Ks2B|1Rzbh18K`6J; zIUTrZD(9noL6!2j2RbgOJBvgPdC<0t=d?)DLr_Ww8*wqiGvs8@sKs91n_`@_n%oE6 zZbFC-g?I9yce2R?9iK;}%s)+%EW7K`863~S3O;15sG~7Mlnt-g)dWAYZwto+ETNu7)wo)s{&%mq`HUt9|1)SF|D*#E=RuEh;d{?Z8&JZDeWS+8Q zzJY;f4SdEU(j|M~S~)9Sj-1K2JRP=5>yx!Vt#C2I@FVrbKaw@|7H=w}oPo zDYB7@LdLZgNz9y7e|=YYneSFvB02Ip1dOpm^Hb92x;-K{WCFW1S=uvum@2F8@vg4z zusJjw)(NsO*!5yuC1D+?Iu@e&Sikv|U}O&|Q>ya?PlIhN0&7MHjRgm`|Oai zz{l|F=-Hc$!yb0{aCtpxjl}3~2pax+;@mwo@dHirPh=gf!7| z>J~I6Ax^tAkE^54!h#~6*8}6+DaCgNc17IMs(}3$7=PNm zPk1A1(N=+uKE~ZCF5-?G6l?>)uyGkbD6q09R9fE0zyC+0MQMl(`XM;)dift5g zgCbX{C|E8o-0s=NM{6GAUlMxNf50vG$Dh8Xi^Q_-6J%XPkuvokt0OlrImF#BOeN|~ zQ@eymm3u-5rH@?E{3Dl3i=;wvO}O+B7k7NdtE2Nt7pOodiS+YB16k|418N1-X_6|x z)gc40e0g7c6(y&BU|R_uLyu|GcNc z`#5UDtb^ZE-Q@lvMf0(a<=3NX+m_ig{MB1lW9hhS$=?@z$KQM@^epd;15ZvCmPf~6 z4ni@&rs=mq88B=Kk_2bEmT-|AEnSXrVog9f?=RzKg1*AZ@CHj94_`L!uaHe~d$b^KoXN=($8#!4tj7V5p4ofEY10yUoxF19Gd8lhs$f)CM|oi0qa{#);D94d2-3iTNPRD4}~3EttS?J`61z7_mgc zm`*?b*B{r+HTo)xTk~Q`z60l}>rICG2*n(t$bKqn*|f}%Oiqm74e_9~ORq87!QJk8 z&#wu3ypAf$7hP6WNd~2voHqWl8U2F8f&KH!W;_v6N~AuqMM}#Rd+~S`YI*!*?C~D?6*Q=gAwJ%^h7f zR>gr=M;2V*cv0WXb~(xdHp{VKZiSz|L$Mw5$t(O02O6XY^%3HcyYfZYgOVqHOsnw> zWXGzUzYS7PcRe&Ts;`2X-AW9Wjepfg1Kq_rSRh<#}wgc{Nne3^68b32f~rcy}#iVPLt zbMsb0K{=kk)h9V52MUB6rM=Pu?$(f2kZ9R9ZG)(Zv5ZAFz%ieq?C{J%;{9ASuh`CW6x&Gv!ZtRQwPARP~J&R2sR$ti(4NZ>^x#_#5AxTfBmL z6duF7>-}+u>V#*dX9j)2U4L3yl;O19)Sp&OX>{+Rbs+3@O?XH8hzu+&2>((Nv#?#Z zR#r2+0cwnMxjkUP8YLy(z#mJZ1x1P+x=4{rubsNq_cPy%%BPSP6XSH6Vop+|4jI2S+zhDJh6+^42Hr{K-i50}$`@6E z?XQyD_DmGq29A^}Uec^oWg_?uD|t6PpYXC5mBXvys&(F4>DK8d{rkh~=s$nB^{v*( z6(oh*sXFf8!t}~+lOCv`JMX>GEf-!*ul@SUm5>bm^M~ud(Wt-H7UO4&Eo1F#ShQG{ z53sUjj{C&ddwBoC!-%IJRd2sUHoY>o3>d0MB(Zia!)~g`Ao-PJDz8X@^tD>`j1oVXK1^4?CP=I42*e#>fPDfO{zg3#c z=(2H{9XQToL3H=F`wg z8{XXNb(I5Za-fn)KjvWJcmq@#8@dNJOfL_86uxdgXk&i81Dk2oSbqNXuHA})F}jI1 z!pIu4IVM2e^H2Wd@2hiP9|^?1f@{Jqx36k;#v|VMqJgmz6EC`r-H-qB%SV44i?p)i zE)$3aY2^(1BI)t$1cuOlx!7&f7)DJzj-unu+~f6F(_ z3cK*}SdihkPM0NQJccB>R@LbGDENzc8@Nz%<^L!=k5f49%A!ip$G9xHZfBZJ%fG$h z2n%Gq=vH<;{`kwm?~DZ*FQ3iO(6PX47A-go@}JFegZiktg1bx66og*vrm4B(Q%GfL z*jONALLy7U-;;k$c+)&~_lh{G7DA8}bUX)pQL!&I)&HvS0OYUp1^pbLR?P(VI|%<9 zcVeM(l25*iEsOBspV#_VHUKgDoz~xc3}?;?uGsW@<9c_^+V};ei`{zHfmbr?O=5RX zDQ18o53$01jmIre5^k3rlI&AzeTG*sC~n>nTtqtQ2XvSFp3vd_w>de=m9BR{hQd&7 zj(oi}iY9kl)S&1^zDJTsy-hyA-3I;F@leZkL|MVFgjmwD?_G!9ZT!Ie?HgobgkH}$AUC)Ks zIU2X&FEacE@iY6Rx|xR0Rj&xREX0%-ez$%O(%3*o-eKc+N!31+oQ(J@`0`hOA<+(; z<4iY+pd?aE97WbqQ5w3)JAtR6uZgwL_-^Ow!MDc9m|ZdfR6z(V)Fk5s8|D(iAjN(;?FWr z>NB5Ij-_7Hf#)d;^_tKXMdx{5hWX-&>z7{bUPDeP5jjAJw_Kp{js^OR35JbzkU`hQ zbV$eh-M5-P__jGo*|XNdfmb*dsL1N&nY?__Rnq9(MaQ}v1BYRNtL}+-A}&)_(dr8e zQWj!CFK`S&i~dAG!Fmi#qzeoAeTz$)-~8@7@0rgf4%;eVVcsZm=~Nne8-ZjS>n6{6 zoTZ@$x^n)~35|=vlWaWz#{021y}xAuj8gkQ`CnIY-xC;p&Oek){T~v`Zq4JsfwX)R z*JB&Sq~e#2%BJ&uwenjthF_b3Y8$KZ9ujO`%Mla`SG$(anh+tauLsV0Fd3}L*H@13 zmRt0oZ-fyZn>Er*a?pX7B+VuReu83-QKX8BLYt8&$Z=D*%aVEBGw>~B;WkI)i`Ked z3`&+g;8h5FNFxxSs<%w91ko)#ugzzzyZTFM4y_A~V|uw+v%2VXJttn=P#Jf~xR z7c?D}K<|C3qI_ogOm!`}=2J`BeAEMOZTymO4P7hk<*to@6(`DxCQr55Nnr~O<8;H~ z#Wo1BvNMy!r`DczGXm)s*T@pG!GVFa*91se6a$OhtyC0-(FWYQovUE6qDHBQ7Rh$M zdc}}abYPP^VyRwy9G1S(0?6B-`~x*8`RA4oPXC|QZ*G78?&4cN z$>rW54Rj^&x9FsK-W{%&K;-YZVikeFR$#pxYftLwa~|X9#@Y&&@t%2Dfra`)74ZA_ zjbM?j_xm0>%g%9e;6-qk3AAod%vFk9Mo}XJUnBx;WxFgTq!=0* z2e|1#Q;E9PC<}n4BkjQD9W6lGZUfyVppM2OlCQ|*TV&XUkMUR?jU_9mLxL_{#*h!@iKBG~JS|A?==t z&QHB^>C0Ym&*`&x!IysSVLSZN&%M8`2iZOO4okd$@X^N7C?jAV2K`w|+Flux6m60} z>Y4?&{I?StN%|aGlWWbGO(CD;C&P-32Z}Pa|qt?-h0i_F&{1)s9rCvtG zG8rQMkt#LKE0oLC z__xsqy}p!G1f32<&GQ^3s-~yQuelV4Mu&CCYIsYR9sx3@qrUgWwPbM4@mcB(;i#Lw z-&sTN7Mu>I3^!7eK8_E99OWzvtw?5#`<3W zxGfy|56|BkLHE|4;6;zCGlm zXThUjSq|&3(}6WGyWpPXum_o&KbP%`c*+??U3K1V+yd4eX7b~)aul61kGSpnax$-RwlI4Ra8k=GlZ1iTdS zoN^ab@?L@o4m>?q8RIo9oZ+Z&Vr7aa#T|cdk9oVC!-fLBOrpNyjlY3M97z&7Ff@&2ppzA4ku8#QbRkK1N(Uew(cDG5KGvlzjI0Rg9HE zuOJn4Zb*$kPK3BU;C2@BRXSQHLB7W32~Kgcz|r`DW|TYo^!@UE^R>Rk%*AMdYlrm$ ziG|1jCb=KKS>oLyd9rZr^kxp$r7Rc3Lrn@UySB=6X0M-#_S%^pBkblksyzUi6gd%@?D&sB2Kus9?(i0p;dKx?ThFXZX|0X`sfnX z+K41kEbPUru9Jrm6`V`LMBXEDn%}0WcZ0R6W)2={l^s$%kZA%MxtAdh)y}!jxihbp zXSC|6aJf|rcT6`H|EROgjZOesYUQc)$UFGAM{V#DuwuAcGKtBHSD zwH0)z4SS)Y9F))Rhvsn_-RoctT|qzP-3_khR4JcAr}%C6JkCYQMM-p6M)(d`1J`nO z{6VK8;Vn^yUppPcJP16^Uj}{O{Sa@dpz#{JLCxNF!)CmgU1%9TKa%~byjpUAx0XHaU0z+7Kk6;A?pHtQ|lpSZw!qx z|BT#ah7`+hEn{eurO_m}%#Zmmd#)WBIS^6hE?0%wAe*8KZI-Lk>cRm68T1*CI(qe#@+pPyu>;_WfOjUAFU`}|*+5=ZF2 zsFBpoApcqCcTRW`h?N?pLrypNZ6S}s4f3Sw4b%2}XD!U8_dA#I)y00Nb2}XP_XV!EKR)>5 z$wrg3a&_QwGUUVxi4GhB1l_L@i?9TWiKWO|D(bFBJ_y^YcPk1*4V?{a9N95E;O3WF zV~EVJnTYq?pUk{q9*=oNG6)tTH3meihvg3n2!qk7oSN=A7BgV$Fj)hM32uGT|5M&Q zzW<#iMk5hfxiEyZzA^#dArm9wPtG^OxICP%QWa2r?RK)ux!B*J}_0Q+Q6-n^td1MzU`Vh z$|}w(%)Qw28S5F(JTjBZ1NuHNV(y(A&b{R76avllBP{yo6my3n9Z)~M;EWjM#UV)1 z3vKW1vgM?M8@+J3AF7J)pAE(y^E=tfc3GNN%XD?ObIpR3uykb;N5=!JT7HzI~ zJFdj~N;R@cXVBX{v1q1?u8SOS?S*!DunLb{u?h?2<*NKZ16>x!U@>Sil&+&t^?YY0&(EUL_C2h;-$_4RmUZVlMl1EZ3u*6@W9()|2VS88xBduY zd!Aw%DRLSaZii8l4oj;X&;EJGxH?IRG{#Gfp#dNVUa(96eOdunl?UZr8oEY-<)Z4H zp4E#C^`*&Em%{(DX0`FL;4gz75MOgK5Emqeq_{Og563-sD8+9Tc0d(;1z+RWBDp5c z4^U%ehv9Z=EcwZGt|V(*)XM}<;Iv)98L3{UdL-F2H6GN%M=_wIK*F9Fv~mI*))gL-U~9KYHe~)5!u5IRSvu;YcN4oImH~I$RR2Ui+DD0dpvJ?^}k*RWut?@jSqN2 zs$^*U!XxWpY8--=Km@^Ot``=-?uwvTuXLZMUTxSKh8AnRAl+w`&!AI}d%L_nERNX# zR4Ka!JD58EBp*Nz#=U#oH8Tbv#bGT(o*kSKd&UPpBR;^!Ol7&NC%KvPLE&_Gm*tz0-mRL%$E(c~WM(uI}i;L9sezy%rV%lJ3M z+rhLTS3T;zV=d`rAu?XT6(_vodC9U^mvWw=)k}@8ezFW_qzZmFZSR>9;sC$U1%q>j+dHReQs4(mf(p!P#)0$6xMJ46X*P+G=+2o3BSmmB^M zBii_wE!`<@2bS`i;&UG76vc`f{y3PE(Z+*4Kt{WZ{r9l~2=zza`G0%a6X>TPk&p{9 zngQ-Ph~bpGXagR(UXk|tqqg!PNM)tVwE_2D>-FENxTGkJXqWc@6IaoA%sS5uCVPy4 z!Fc8__TRw{BRTVu%Q?nr=)XVhuO}-UcpAzwnT9q|Od>_%s3-&NC)5Ocl`hyWS7-Qi zxF-1+=A{X;7hjK}wW#>|iiz1`g^J0akmg0fMpR_^WT%iq2S&vy6I4`C%ux!O7hsK( zEL-UwEr{hET6kxk#%0jy4kRkidX)#YOY7+-y2k$yr_{B}tCZRUs;Eg&JY36(7If2x zlo*FjniaqB0ll1*@plJq4?H|8jkCr-dyy{m&_b>1-fOo*qXp$bcf1ZQ+~LtL)OcQD z9!|Ft)T{;2v)+fLJ@x?@D*#dP>VMk$vXgXSoV!ffLz;k6r-2lPs`qi$Nvge5{p~c; zEXKu3P_Ve8eNJTsipi6Feo!>iXjPPPYb(iic5xO5j^2UE842!gSb_b<^4+vK0g^F>9_L8YPZ%mdaXt^5p) zd5;X#h7aB3Hy$KBR-((2eKGF0528li<>=zrm>EXA{KM}5_!Zgaz_Snp)~{l*RP-pE_BL8o75)6F4G5P$u38^52{aO&k}MHfklqD0XF3+Oa{fA|SF8fRV#yzin3Lh5CL+=RC0yF4Dh!vza8k zGUjEE3Ai#SW*bFPsi@M3T8N}%h|0NzLhQc8#A=@B$`JL=5G-gvEid3AV?eX~;`E}| zD!9iWK08hbG97bU@ML<2@p;$!+xvdzs+JbyAp2j(Y<8vp0dMGOsU=?xf@&n zIr8(sl7wezfb^(JiQmBRMIi-_xy7rTRj#*ST3JIixhMHF3bRC)K&$nj7m|+*xEg4b zkS_^aclF;~&) zz+K+M-6H9slioEH_#9QFD+gS+d!Cf$I%`#%Id$|UP7mEidU(--e)5!ej^59?MEb>N zgj*yze&cJbY_U?~=Ie|wc@&=QdY`W--N zqXSABb@rkhy3Ah(>B0)d4yIY22*huX>F9+ybgH5_FveqR2$I1ek4|6sd0=X~88YOg zp{v1}E`u_kbom;W@$irJV>1S{7(g>(gzdSx>ZEt8x!Q}vW&tedwM!EuL^b8=1gW?b)d>(3>9E&fSoQHro;?d%RrJUpIaUo4P=5j%93zQRAX_NVOfT>f>=Cu zO$h43FtHmbP_d_PoodaTHLx^O_xr`o8FK0;-LNbJHXTq!Hvn+Rsg161*GY~jpLk%- zJJTgK1iuZq2uPrcQ#(%B=pcY6HfBVlAZNx*wZ#6Dgq;S)o=>^K1hJ_LcU*tr z)0rg^E=I(Bq`vq^vc`dz^|>Zc-a;|S6xm2cAw^)b96fkc3R0u#(swsNCA%7phX%Nn zCup%{fEhc?+8vq}Z^+J=G>0T;%&U+cb~3Y29*xnu6V5uI&dv={xAAk7`HE#8gLBrw zMAbLDnxpO}dA{8Qr==`UC|;(UoM8szcF$;+H@O!;S@Ie|aX`5XN-eAVIk*e&@i_1? z@NnY==jfE_2b8M?NY<@h?{*2QM*D@?v++}JyYGxB0k#4}V~Zqa&Kjlql)Tv^p7The zFUdZsOm$u_JLC>*xTj{GaDysqSsf7W1t0I5F2_5UyPTy@!L9HEM#C{X zxY;@rz3{qcV8;$_o4MrZsY5x=+JH;D`=OU^z8Q3@gGGb`QVxL;?FpZ&O62SpLcDsWTQaj z8Bo|~2j{}~ziF;YWI^cX06bH5;5zm~ktB37V2cCBk`9YeULE|$gI)t}i9FmfaZ+Kd zpIH0+YA?JCyC>FR8ybGMrgf4z5rKt%|J5P6Azk1fba1zOst+l8fJG#c*X53Df_F@1U3rifp5z zzLe&RFnqI@o5?{I(R6v9_-Qai#nJHXkHC)eXjJ<1Y*MYR!9F9M<=>(5({VUlMA_F#v7f>_lF=etdOql~c=I(B;fg zZud+EhNNAbcn&r>VWVFL-Q%7n!Jesh8TJw8i%`2E4wM^mfw#EH9oUxu8X0tzACToJ zA)f_vX(wGtHo`VG8Fp}cWw;!}tBL}2K%`#jY3P|*EBcc?{Fw}@QA18E0y9PBE+~4LKs|vOVB72>1%y#UQB+~H zHKpZmW8)Fqeq0+rjKyW_o?wS{dH*Bj_eT~Ry_&EO&$p5CSH`PpG4X0nQw%Un)=^Oz z^eH*6>~zu+x>K61JmQxuLtBv>ng?1R1F|$mpZ7OB5o&`jeC)#FkbD6cqqVZNvSeA6 z>vi9iA@N?&bqAC<2Nz)K0N>+&)L%e~Y$WnGEFg;&x9F`4^Ig;jK)DE8@sO|wPlr_{ zXnerDXoE4b?Xh^?95&i97SDt6)W2Myx#lViugFwoL9^kgGF=%5lhHQkZ5|x~HHty^ zM)wT5hL20ZAt%(70r?hM6OiPSX}jw#d657wVx+Xer~#9WXY?TxD8gEErkz_rEZ$kld1l`cO(s`%%?HMYSByl@Cuqrg5!;KkpIjY*)K z|MTN-ylceF?>m3rK~6a^WZ! z+xWPd;dfcG!>4qCfhPqgo&mQb5WZJqWyK0vDpawzb5KeM6kao?qPrUx0h&P{a(WH=YAkSl_anMJ+YWhH2rAMxAfoa@{LtS^x04%#a%hiMPK z#tkv<&hYDjrP+1gI4H_^AtLPpuhAg)f_*Q-7dzjq;rnkUnwRQWsByXfZ}E$Fz554^ z;*cz6+W)sDXjNE_S9@3qy?5c3AZ^4-&NY_~)lP9f!Fnj9KiV(8KKHB$$*5OyGy#Q? zJLrdDiEy+-^)T#(zwV3vl@^YVc~C6ecdWZv4T|F?Woq`k-6k^{qs<%qFOw7pHb(nR zj8QJd?4(Ghp`!s^-fKeOc=$|t5PD;vlY!eb8t4o+hUZ;B76d=R531 zf*ET}O~Vb}Yd&GVMz^rIE%LZ9?VRw!f=hI@q)}2U#?q_};TfPH*dA6E)<(DS^E@xh zR!>>(ykDsGSu-VF{`I+4(5=$Oe{@Gn{1p5$5yw7{C1ipv8 z(rQT?AFpfdWo3&_gp{o%QRtEZ&-^+PS^9k1i` znkn%?xg^VYu2z-ib5<0~%cggR{u6_ zw#hQnA+vUDYqIM%u|jQa&_n4+3&4B z@7XaI+mT8|RkXS*xEr#uNJf5@Q@*Gq{FWz{mtfxq{+XCx*!my#D-&&(AMQmrVP(7> z*Y;kzmtFRz(Qy3i&bh7R$Y{!#4!o$9k9yoP z@~d0pXPehlIjo0kq1X#G$7*@)ocn@HbT0P}F_1`TpnVqimd(((l)K!Nm4)FJ3YBFj zdd7zY_?h3{COkZI7yIw9jE7MK8>r~_XYU9#8XTc>8OzA;|0GkQT@W%U&FPO;$W-1HbPt?D>Xl^mIYTabC43F{^7b)&JJ zA3NT9!()}#g0w1R%|czVbY&&E?S2lL{43{ggL}8j(0J{W?(@VN8oSK4HbLkGw!|KQ zVh5q+v*blnjS#y1U)TOh);chRG$s&Ar5G@Xn~+$#Fd#lCfrt5nE*j;$vC@0LbG;NR z#7n~0yJ3C!(h2L2u{!oTj0SrQyH)pKg^fvnoSk=Nnh_f5Bhq*tL1yzGs(fxLiP@9BAQKRkrNbij3QMdg&?5~ zWV@$E(BlU=-&WZUCAJKHeBQH_+?=&sQZ+RZ%4Ii17`g}>fFrYAw&C^Sz;eZH&-32Z z-d8z5Dr{K9C(90ymHr3Frl6%5WP{#BpxayNzbPnPu2+i$)+xG&zXu|2o2CMX;gF(F z0#cGT+92ED@|90w#fELrvbQKhsBr@`!j(vGse?Ue)!LwY{fs#AHCtx{w!bOs z82ki{HGD5N&;I1Mqus+686A}7pRCLwr7xL-a>>L*d`dAVC~}O7N}F~gC|$V<$euEp z2KQ}qleh^!YT)*jmiey&-5Y~a=#W!qaFcr~$eO^CW`@3Z3{-H^<*5Bx?vmuQTcO5- zNU#ik3V=_E^gfSV8>M}-kt_%_m|Bvn zCkhe;Ic_;XR*GI!4qc@@?x82yXps!K;+p50@CZ1&M-_WWx>DWjo-Dij8gg{ENWeFK zL^55Lh1>?(aE!ksQmb6!ImR{h@M1}gEDHy{l6;o#qMO{$(V4sf_^d0;vS}#$rJJ|P zHIV}P?t6am@P$%B4uG01vbf%vS4UR}a{#%xHNZ`0)P|3OMtyO-s1sVW&k3Pam)_#n zLq@ZZ6M!{SFfF)TAV7e~>HK24zovRahS+mfhoq~4hRDyqooKP4D*&A?5?3Hc> z+Jt5(fI*f8xBw%>yWH>)(Amjg;XOXp4t7D%vj75^qXz;qu|zH;CzKbi6(cAH{L*fl_blu z7GC$eDCu(@aMkhE9o!4_L9b3a6K=9H-BEf#c=?KR-5*OpKr_cFH_WRyYXu#z8hW_^ za^B4xEc7uXIg$GTMNxIqF82}e`+C5~Yl{uQ$Ll+ccHiqUKcC}dL`K1vzxoS_c3`(C z-DIp2DJG60>p-as_&cyi%}~#k3_f*|&x8){@DFRvAP=8Ak+1Bm*vX?G{cw&E5>6uJ zA(G?3kbvyMh@|xaih+8-y;Rg*X(De>iazlr`pTk&1w|ZOqOS^y4??zHSZe8sclr6V zlVxW-(6P}W6;!^UjgRM6(Dl+CDqR0GN;+KnLBCLqx=0mVR7TnZ9=Dw<1g*i1wey%g zam5NZlRgdo?&SB3xXGjjzE93LFm66K!OeAw0a_6)71hH(#MSZp=iTQohi1gFLEtG&xzQbYxcB866UXh8|R#<_>z zD<0q;5Vr&2^!8U&WZ{HTJ;=5nSusRyEKbRmluQ& zaO*s^aQ>A=PsH04?K17+F27C{-m!{~5j^GLw`-Rj5EX=LBXprwnmKy{%sL=i2SZi> zqtXT28U#jw{h?&)|BzS*1{f%4k1$NzC?*xtCQ-GK4;S4JW*Seqjd1tYO$Mf!G&tvT1v2+_b_FE%7-r93~I{D0j;jzYK z$u~eT5K8Q$qOiCsp4Z05T(%+3lLPu1C3Ja^x+-EH2NmZu^cV9kJ1<44Il~z!frKkD zobyftjc+!+Oz?>0i>iPQye?857tt#0fCyrZ0TIcwL#u>(|RUwYfDx zsE8j;EfXAqiEdA*x-VQOZ3~8tcwV=26HvQqnHb>2$1{-FwMKD7X;=fc@xT5yIY0_f zJzb4oW=|;U11HG3Ad;z%Phs@CDniSF8(+o6Px%4XX@t|}K@NZQ~%ESl6ZTu$-yQa5w5Pe)l{yGRvnP)_6BU0r@BL;zb;@*t45l?xCxbPtmuXrMEbUh?l5qL*5X5ix(KGNbb3h{L&XxqnTB<7SOs$kmXsU zm`=+Uuw^+{Iu^zV2VT-!V2mgMJBV@zEfQGNH+i%}l0+{#4j|Om8Dds9#jOzqQb!x& zQGVSq?_v#`F@K_UHz>6_CUBb6cqE zDWPK-Ok6e!i)CGIT2-U)4s`vW5j-Tj#YYvG!i=9*G7X#>=<)AzubQatC-z{ld3;%2 zXc-KnM6;=%e^Q@q-YDg;0I|^elpAtP-_U|_-!ykUpLiEcASkLd-WK7Rk+BvP&s;dp zJ6NG&(qYBL)8=f6M@@XkJro0rUkw%YVCp*ZFrtRvBh7{AEWXm$NA{P zzQqZg1a6{-PI6Rnb?%@Ho?rJm8lA;~g&QJH6P@IK_|Dk}y!&PX`z^3FjL(Q!J1oNn zdz9C486Z}};*n#GxfP|Ygt%~vER$Ed0H_fCvcGrTw<_Z5 z-1wkANiDb4yA>k8d&vggWr%`kRY+|X9~3R<2ZC=rUM%ecs*(7heabd|1JQ+kO8U51 z)pa{8i(3?N*4}RH*nl(A651C;tbjvZs#?6uJcr4qrqzKlqfH?FH{jL_ z&AO+Ac1}rIIi6Ml$jW=!^R)J*!T@uQUWfHvEmC(^;6j&wXC5qavDBw$aTTa4{cjAk zRl__{-3cRb!pcre+Cz8nUUpZcol~c8wEbM#JzZZ1hJ{>B{1Wfwf=m(k9#~i0C@m50 znO34yFBjx7Dc))nOu_;&Y`#RH%gS-`q;X)ZM9RpiW90UTv9<3eJEZ*64_^Pd(&&nu zjQV5~DV{Um1o|Es9CQ+jTy2ho(ZKeB-h-I^} z_#i$A4*&}?M5EM?-%R0RWvvDXH<5m5xuC}Fkp#xCRfZc_=eOMx<%rc4pjCyWVu?H? z6*J_5FzeFex!!$@MU`V5tC8dO?DiOjkKI@~tXRx}b!^uki? z4AEZcg#`mZ*AhFjL~;y2D4R>I1O=O0+U49=1QfO3cq3DC&^p3wDfYOK&}qwjhkH|zLBbDoDaM}>Bj-^PFaju8{TyO8!iIp)9% z-&T`RK2I?)FP=6CtF=gWa!VILkfh1I5qjw`=88K&eg@P;^v+5ZzYe-hl4U2StYG|#UPyoYa3c*)x6Df>Jtv$?qpHXRRXZpKjL9}Cs()UI5;>|6O3~A|%GS#cxkEiS zVg;pWv5;jK^d~g(8~JT+*sTV5K)#U)SR>gAl4pn2zI#}KWYU1{>V6L+NPbkk{Sw*4 zZaQ+@xHH*j0;!!81HRXGAO@vx$(EBGh_Y$KtG&+hlLej14%Lz;0U3U~#ru8hUQhEw zIk+l`CSq;G^_i8PT2&rryGN!xXZ*8=j*6+ktZdca7fRAd0(Y(KjQ9prFSV22g<92q z`ZG^ht@%8X>=d_iw5nRNpIgjXrCJ-YGknmgL!2p&7F5yuJjdUyu@NQqH8Z2!ZE%I} zkh#D)n=q^cS0uqwe?&3wS90h%f>>^=3~42>FhyOX$P-|{WQFh+ea|ml-o!^gL_@E2 z$LNbXmK)>OD6J6Up?r~^j3P&=qqFsnQVD^U1T2>{8ulOk z+$*+D`e4OuYj%=;!1pWBUYVXQf6~ZoIf>*n= zl*$yeOHX+B(9zI9jeRRUVNbm_h>DJ%{w5pEz+Z6jEbGg0YT2-s;kOC?iW}V?4AlF5GqIeP1G3Gg=f0TsBxcqf-CrtAYcwa8i1at<+b53~2^Yqb;X7?KZEOImC%=8pU z+=%vaPf9E31JZ7&XT)63-uVfxSgrye*+7nq5_wk^U7bX}Rv^N<$JEYey&{NkdhIt4e_=#K%=YE8$PEXc zrkVb#Q8h;sKbJ+n#B>bnsTLiM3ocW~4WSR{-N-rxXuiZ(+Y){Iyg1 z7(@#68R{Hmp6^5M8UT8Wq$gqrrLN?Z&qT~?0%E-GdI7;^1&yXqf5#m_43*-!9+Vy)>%^+;t^vzL7U|ruWd13P_Y4p z16TT6Sbk-QGJ$>i&OG%~S)+6fWoUpl*b+mpI}&pOnF)YN zdH-g?TgI8{2{-F!q~evyC|xy~*UnH(14Zgl$n8sJpJ@Bcv!buJl4YBm*Z8-?l!p5a zQNBXgAz?Ld4Mg&gi3&J0JTt;mxHZgU?h1k7td!{oJdk(lujl(O;G3 z_W*d*3}>Ne3zjG2yVFw}2vB}-6S$eYCy`epZ-^U6f^Fn`HUaAEiN>leTkuyFwd&_% z0h9wzL>2(GR|=dzgYOK;Hu6%5x`N*&%=fs zxnqZ$8t7|7TbK@Rm+%2c6R?}!K)UE_P)4ZMNqS*{d6iSXh@D-V2tXN8#|l>;bJqV~ zU!#Tk@4dANWTylB($yw5r-WjlQg06x)hjC(HHUW7J@jYr%>U(S-Zpzq8>+Mh#nb&^WX89Ie~G4wcyu1LXa`Le|Dj8(5XRqS6nDOD@-RP;VHnF7#)~2 ztBS7W*3;#HsjWVEZu|5iVez8&utBFnVFRa*ZV%G|FO}gwP`f1Uvmh=W+179&wUS9fUd6$Ne!LP!C00 z_||lpJ^VAWejN`yu>KOGT)@g6O?Lb8{mJI^L2Q;{4xCXhFu~eaB4X>mbwm8gEnq3h zpl`W%d!81S&wTv)lZCM8lC1N-q&yDJ_A##>&*T0ZL@NU`yvt_VcN!aI#PqvtfRm+L z*>%5X_<`5WyYgQu4l)b9_gBe1V9)GSR?dQ^Y^3-|aA_D-Bgd}4<|sF_Zs13`aGZBp zwj)0lyzw)0P0ClqW>}Ec%wE*WiQyUk#?^ENeTjqpDwi3o{>=Bq`p=0m6UITe4m1Wo zW1Mkfa{H%$j!?hsdGKZ+j*sVHaBipDKF%#qZ0K#59h*_&S>u0&<>J-k$Fb>+)gx`w z-EF_fvUwP-KH|XZs>Pm0C$;(K2bYpm2X;~qnK%x46a!XaCl!U#ekhOx-H2&^7&$wt zjAO9YKUr1-)Ml{rfpI4Wat&<&#zUJ$`JyBrxDYbnKsYsCrow8h*mVK&Vhpju%A|MQ zs{_ng8(xv=#RB<69_ZQTd3DV1k`-{ZUO8UvVU56fiBf|%ArQM(W+0Xt-k)iP&%isk@Gw@4FE#Xs)f&ROqohk<+14`_DW%5gTk zYW&}_GA+~>o9-Xr8KOzXy zO_5zx6!IkFqNGZ?lGy!PdtSBD(;zryW- zN*V*z0*E4qZdde5wf^eQJbR_};^Vw=X*aPR0;5h1hMm@pDyy?tS&~WHU;9ar`9kLv zaZ)Ul2k+%>n|52AF3)m3GJBiHpno=9;#t3_kk0kl9Pw~G`WBx%dRC+8xp%Su!8b2f z{nfnS*I|1FEL0+E=tN$5WQJ%^nlC!~|Ficka80FW`g6o{NInd?0p=V5Q6k925y4Oa z8+4}aOsCtKw%cxZy508QR(f%_-A;G2_BJg#;#~v<70>_zA_xfL4HX1$co`KG6-7jl zfR2L~@PerLf8Qi1iNwi)ghpp~eqGKvIp<5v_vHQF@4Y+^2+P2}<%Y*n5Tu6fOUi_J zsJg>4yUpUupjV@#bLb-9RyI@hbZRezPJ29Vi=YztgtAheA%**oPDuAW5w>&e5~y-Z zg1#e71CXU1ov=P)BM8lI7H{AxLa<;;CvRX+OHu;!r4=Dbz6~sjq2hI@$-4q&!x`h( zf#gk&s-An!8Ph?}M*hrXwgM#HL9xA#!8=j%G}A{@r_@0^!wnCt-qR|dI+aSa%2;u_ zv{Dq$912_~2hL4HxpcbpsnbruYf_1PwQ=St^I|&-Rcgq{vk_&ZL-iS#63KjKKdB6~ zQEr6S@5;)bGOsahaWk(4!Ewt&w7vIIe2uP6!P6h^AQ=wq+EkdhHpLWINRfOhrXKXF z8$)UY9Y8yUy#_}`?O=`!ev6(f6Z;J6fITYCTc=nPq1hJ*MZJ18nG9j2sGDryz#?=& z2`+Lwt%n682qrD=(aC&j1e+^zb&6d){Z1QOlD>i#erh) z1}dgJq+PKmphQrkxFhQV_G>)WU~|X?c9q9LVLSuL*8;H0T~m5oZ~E6VX;JfmWpg`O z;Ho(s5HD?I>y+g__LzyjfBq0>q;D9T|K&F$4$F>tp{sD}Xd_e_-aP*iNpfK80;010 z_Q=^32XZ4?5FQS_>6JOYn%?Ho9d^klWlRRQQE&v<-!Zm(H@qTbE3_}wtE*va)gieX zUhY$)s2aD4hWH={&!_^Lq7N(P>a6 zzB{3csR&5rD(Neda!42!g`F8!3*<33goiz{+~S4XBGw1gse3+n^4;367ksblgY(}$ ztFBjLFWsCubIji)HhisMjkNEvWVV`4fHoJc5^mKhcPKV+aF9Q)7lOc7*j%Q0%KdPB z`?$abQ#a9Z6XyfNM1f$<#8eI=K{$4=Qz2=E2AOnvb-#M^!Gi9K*0n*t{tSBjK=+yZ z?>{uRO%_Z*4&64b)N8eO;n-w|w%1BeNimz#%%+aspem);dq0IQuF0!#Y@IrZS?`H@ zhj>3;8CZz$*(bKe!C+^@7VrJCG5F1f!!BX$)F-<~8EwT*P1r?OX<$I{g80Bg*yZYk zrquQ?7tb@MsTYwZnFeW1j{04Y%r-Cp*xk{2Vg9Vd#0$^a`vOOrT@^-qkv{f6|BEEQ zVj97qan#QZ&Y`#s6iKII=0eTOe7`*bNYY%dKBm&Db_$Av@Cp*vb<&H2a_9>|=cgBi zK?#)b_E^Jt{zTA(3oljJ8y{Ap;YAO_&Tz222>#)dJrDoQh!g+JW{~IRx2`yFm*agC z1YM&z9YtEGm{{??F?$0$fLmrGt-*4c9=Gk&Aq0>lfDmpda!zQKPo~_PlFTl2X_nN6 zXx2GDnu6SZz3oPeKqUJT z2(W8GoeO)u4I~vOJz4|_pi`|2I}^6VtQWfvLxTq4i?AI0mA((t@9&Yxb8^kA9r$FU z95;Fix_$kP)hdchpvZD62D!-2d6%-9OtMm_XM`*e9G#ZTBAV=Niu!%=`a`QxVg8_< z&zF1b{B3ZbWBdM%tt5}10dineRGXlpjN%|XSxUtmk=-Y^X(aE}ro_#PO$!57(dG%F%8%Zc0#%ejou zWcJ3GZA{IyeNg3>FUIt`?U;S7#*Vf2!bX^?f4}@+#v8FS|0C~ZB;A1nHsvO0*g|p8 z`jJP)6n=L{bS`s$G8V-ZDauCIt8-NI-HTmXl~+6~=xyV=*wXM;_HRv))Gkt-_UM3~ zkSx-szTyd&fS)L`B4i`9>|>?p-SAjs4PkY5adT{f0UOPG|6weI_z|m{^p%V2S7klP0`cI76u{HQHdt_Kn*RWw9f7NJ1eti9O8rki@Hl*If zh8&|fSYi%RF%K0T;4a?wKILAg-YjlYr%-F(iGS~``Vxs3bWLdp*&5s;`czpQkV9vx z4hqX%7EZkf+oMB)N5F%{l7k-CBZ`gW=F|=H2JSqF{hb2&`g-;H@g0&}Itd!&AG>DD zbY!RBLFjb_;Tt$_Wp+kxP?fvns?z;ZC!NQN8yI;AK8-c-uy%EM@xH+$iJy7-w>BuQc_W!c5?>o<1{mf0?Z-1I4qDdIsF;PJ8;QQT)hJhu0hw}_4)?_%7i*MFGN>Z8 z(|24zLlC)WTOhODNvx*f9pvN<0a1f|N9!J|{afdJ^JZ3vc)I*|%&&eM^Sl4~>2GHJ zw`>8$eL|7gVXiNI@$-X-K>t~P)}9J-(cr$GR6UmEi*Gy66EAgx@+%bEY9zaYk2>E| zBnUMbOg34;cCa6VJQh@E*`-X=>=O@e%FPC`!znK#GD2DM!?Za@BXaTkaZ|}%2VRj@ znxrA;kKpExAiY$~t#_Mbt)3;o*^xu9_P!!o7qBZZR$L*eSMQvJN)&Cb6_WEzyZZh& z`|f*wPp_~-awG!ptdQVw66w1;ms+yy2D!)hS7SR@Xn%ubpgG?BN4q$H-c+KN0`Hkl<~*tGHF*hUQPE;5_FSP z<~F@WJfF$;+bQS~obc*(I`500xZGt)P_J)=!d=X_7sXC#vez=e%na=OhQoM!H)8`dwv`yYdWWH#tpT_Uw>cpOoQ|DQKOx zK;5aBH3Sc6AiFa-gZ85TW9v=@%LZ-ie+%5qdnoyAupAgq7CJ6Z`{*WbiENT#!*!at zV+QWjA4hd~>)h%%t#1vqv84GHy4(m(GjoLAJJ!a) zF#*$%P-HTaZBM3(0z%C_q1OZ}7W^iyZ;@mXNt9kc05detLV31uA+zgKn_%|&R`0d(tsTSHSSOR zI_NA}JoCu)ELX2yD<}`kp*MvN6D;&I%L^NJ#De+ITeckiC$-T_)`gBuC42a7wH0vlAiVj{Dx1tLGN`l(foD z(#j-5#$bVKjc9iSR$W$v)Q?^-A9l%f>w#hQoZ95ZntxisnAg!bY_!Hg%}mOe@<9Aw z9aQXM)$Tn-7n=h$ba7tg<1uWT>r$`&1R8GfI+l<3y|y``#y!_%@IQ9zCcyfcH~Z~d zH~gUQvb@I`D6#afv;N21-?$uA&sAaJC zPX+d{nhJNFe7C9t2vW;rH`pFJdwP{BURWXM5}?SA;*rhn-l`02>U;$;F zvghlDGESfu*{yo)Q5ZH?{VDWtUUSc5I^B2s*Lv(Di$`^nHg>-d6mCFKFesPK4O;4U zQic1SeDPB!WMfX&>tiRgRTIjUw?&ZTYG%6$R%+G=E=X|Hr%smw*G%mBdb6aBg}15* z>G&BQb_t3FOUBIiUhk3NqA^^`2*r!JKx?A!Hbwo?Z2E=}LP#^u8yC=da~o`YYw-j4b=bXCk9nnsI*30kXt_ zGn$)B{FYRTTSJi~Dh5k~s7lmf$#pr0Pl}9-!^U1LkTPcp zw#hcqb5(zVt-Wln>IvvRd|W|Sf-pobv)BL6c(^`bn{4<{%L}u%-nG93wY)G(o!VOT zqc2yXT}Ei5WO?WosI2N$V4w(zr=EapQ98YFs-~505_Ni3xXlwU_P!eq9%deFnv&V0 z&chMM<%y`iz|c#uV4IgdsXuZ&(zvj=-aC>^cJkW{IdE*{tjYRVO>qY)QbolW_`(bX zLi^nIx#fc%RuZT(8^n_vB`v@gazFqJ=2hmDSBvP1=UlNtQF(>$L$9uI%_*;4vYx0U zK?`69c|}UtW2$)IPs>r&JUgp#?gECRT>WcWWuE+$*KN9lhNd>uC_Tu;4kef~NF?I?o39#tQ{|J!XcBa#$ahuQZc^M0id>^&mJoDKnt>Jwi*i4cqykMI zGR5l!vhl^d3CBrpP&ss}T$;3X8XmPz!CT{*JEHZ_6I}xtur5I-9WTYZp%2@6E2Hml zstjy@Ghf;=r8EqDAS8)E^-$2uW;Ljkad=$35CePoH<`Wbb3)ncR3$$fd6=AyJg;B! zQsqa*`BId(MLteETjsLS|1;OKk?B%RH(Acb2ItEPLKlW=Qh;(4`yP=4SGR>- z4ZZAH8a~ghYf2({%j`y3&!~hTyl?iRZ$N8G zk6T74IH^djiTQWKCq;?-da^upb}O6Y3luW?KXp(Lv%@PM*hI`)R)&N%tA43g-(-SR zURZNnM`Qj<-wT08tZnTmnMZP8F~*uhCO?dw6t|5cCE$3&R(u_m|8KV>_rsgnbg5QZ z4q7LtVdGw}UhTJ9(5fhKy~M-?wZrD@iR?UcL~)6(j%P=qXKk&^z7?H!~uwrt{fwyPN zP4bveDGr>Zhg3`w$&_AZGsYj1pOW@Ju2Yx(OV7Wag~k-nJszJBR2F%j0cJlLw)*dg zIP0}2(D2Mo2#lxGPvAPL=X8qd)$0Nd$!pv@>4%ENk*mFn0Bu_Y56E`E6C^hH>^GWV zqrcwsq5Jj8om1nP%7Clv&Zz5?w=(;oA=ohPT+;07WW+B-y1~ zf~|9Xf9qfllUeL!K^Hn!yudZvFP^!;oRCyV8-nZAh(mn!vb0m!5Q0y&t16>E3Enf~ z8d$9~U+4m!vOvAY6*)pdcAM6DrUUizYS(SD6TwGB$?Oy1PLNNo2*J_N8V=El*$+rV zGhFNd-0VWRg%g_33mwlGgkEX*B>yGfB7Kw(YA&m4~b+CH$& zfAKRH&iLa0275|mhmYTO6-r@ZVEkFWn_!`S>aeqX=CauHv8>nG9>qb*vT+*wQt2imDQ?oN{& zH?dFU6t|lqJNrV~m>szjd?#2Rr$`~U+)qx572l>irl3I2C8pQu`sBojr(s%U#e`TW zX}#*%0{O5u!8&DZ)KjNA=%d&ji(J=7< zyOl=9Hn_J8_69FugLyjb#R_M~jfxQqDYmv>#?8o?+54SO-dZ^F`SGHW22VQwUzg3) zDpUREQ3djQiW)@%xxfyzcG%|I??1{Q5Yqo~yB%9b$TZicBJ-LsK5Y>Wj2#Q5Z3bOt z6sn6ApLOqs9aIxr<6cc)oSrHDM2r!m;lWDZWb?pG-y_5FcipaExPS55&bD!yt-(64c12mFA@^Q2VJ!#Q4M=d>Vrs3Q1s))AhU;t;q(LFzbu+)Tw|Q1s+}aqf!7$w3iVrKwo@Fal5L@4 z&ML}%dcJaw%zAcEr>>AZ07~-4nV-MiH1o5s9gp7eR{eL+fAlWsEyAc z!5*s58HgK8=>p)2trFIH#EMVKYQVR}F=yG5M9(@T47VYH#gW%;*0m?L%DjJ$&j1b_ zinGvdr=wA)6+>}G{_iM_!V^lvU``Bg5n4L0&>S2N z>!DbfZk%XbD?lF|4aDBSEWsfKWD_ZGs0-wk#7 z$f7JbOehw*n#YzRv zoA@M0DGu7<4uI?lNGbI?^*Uv%u+w_8-(w(c)~9gT63<2#O|3_TXgPP&9b4nD4ZARG z_5z@O?vfRRL-dVpW#dB*3bm>Z_Z>kX+3mYirP<<@?428lT}l{rt93t3X9#h%$_mib zGH2=A&dDq&PO4Dn2MWiz*`1zC-SGqT4mQ}&4@+O*qzw4-;BaRdOTYfZ{oYsIm%Zw_ zOxQzbO6U3Bm3J@+k!7+>=|I=&)a4<&#_t-J7TTsSrA}rI=Q|<_J#xcaFNlEHz&(o>c}GC*pFde5o1$jfMiethbUvm}*Yhou9D zPj;D@r+kXbrARgvlNPnW^&;2BVtl_D=zZ}VnDp2@zZ{Uh0AgscGf$oFDs(=0hv9YD z9bv}kIM*dO2Eyw_(N}?+&&GJ_H=jdAzAHxP|zZ3 zRiu!;ve>bp=H{xiXTizPCf~Lb-hVWHcIAV!>%aFKqg@I7k1=CN+iPQ2=9_dW-KV%N zirl7R+L68RuIs|kbZ7`!L^KP93q$jzCuTrw1>|et+WcTlHkSK9)CJYwP_ee2(;QLE z9bFNS$`yq72n-V0X;I1SCOR>^i|u=DS=1(axBn?G+@R+`;flskt+K-l$EZ_k>XkdZ z=DTa%G$(>P=mj(8{2gSsbA#$ZlN%{RlYH+$8O*0l^AwQIj*gAO*11J4vymt#%dG_D zp$ywlWFQ48GW@KS@(!T9I3j?5hgehMQUV2;$PKwxeq##qL$*W4)M0rRYySGD4<{Lc zH={Xk8Oa|>KqH`E@N755fn3ct=qrO{}F%s=z&Qj{=?y;D&oTl3Y$04S}pn5>r{6jqH(W}zF~Hyh{=7Dv^%<*07@BhQFd*{;|iU(Pj# z)Ce#c0c0hir@Y$K3r8Z+R=qk_jEka<-bBNC7rH;7*1dq~mZRXHUTPOw2vFJ`$JZ*C zM&%8mn!?U#dJYesN4(HPb(QbQb2nPAW8eCx%VZ@#pUZLWGNiituizUfE}bIlko44$ z2*>K0BwtMyxzFU%XVW4;E595yJcS z*P*@$DT{A@)1)cvHMSbU>tohQU`v{-dA}c^^MZ&-xiwpxxP~QD+?U^ z%LKg&%qBdD$P(;P?GISWoO8#zDXp?=Le;q8hRhdZ#ps9H8ysfiTgJ-&tZDek`$j{t z>qEaEljHn+4F`@JUpFx;4HQ>Tk+W0``lJb@NRsZ4#KK6NcheoD+}ptg=~e7<#**-i z(5C@~VVcuWPYy!L?znD3X-*JqWw77%4&C9dqZ?)Q0wmi4^2gkuEH;5W@yni$+0Y%Uj%%M`_ai4Icql|)#$usK2pgdcePx#DvL@&_f2F9=@=^4UvBCB2l~CpyJm5!@|bq`DZ~DO~AsPl0kM zI$$k&KGJ>tNOyqoYIDSOvWc;s5}nNv@#5qT>WXDcG*DKn(rwfHkw)vJ)_tTQa~(L2 z16!GX(aA)Li>Js^B&;?VfOEiX1~Eucyr{*c#qNrO15J)qm>B3mJ7434i4n6;9i9KG zv&6JP@^PUM<8zg&hLA?D8d0kMU$wqqOik8^)|Jo?Q0 z*P|c(XzBYY)ZuSk|JJduwMW;fQ3Q8ED7Gb>70n%UN_s=~#~lB;V{XWfh%gDZiQewW&%;&y1_kU37$=`~R*-n|oYVVb0q1W+?ZW zVVD=@s1uV{jxetYd`+4YER;6ZIA_W3`!AmeDXs*v`%TR{=g+*62e?6=H;h6ahDoy; zLuPYq@6(5?e)-M_w8T*KRhk1?wp;l_G1Xn4+-z;SKV1Iz3pkIiNs}r+U@0LxBbKs*si`&`%9@SJ~6E zr?P(L8>|)lo)#T>Ew7$7JzH0#$GQwihWB2OH^CR15|w zw8||C@Et)HHdfrq?on!0aU+q=_N?=5=LI8A(>223`SK0_c3x9p;p?);WnNR@xa>;5 zdYtxk^X%Y@kzB|9{FZWn2xN$}h>;NrB4}K^YYG(Y-Bv)$n|rQHy;=uH4@jv5*c@bu z;u=XAWw>V;pv?=>&lVngVAym_%ZQ%vFEf{#Z&(~Q@Md8PvXNfznM)(LO9cP|i`&jk zm?w@McJ3ei&aFg%>GTZ4nPmhF5KepJ1I^E?%}s~HqQXL{K!$R)@6ph>k@JK(;zP`W zksVVyr=AwqkjtJ+nFI7b&kP|}CEa%03iRET^0LSl-^+^Z>FvXznQA3So*4+kf&Ed! z3nbL->dP0+JKZcakD>eD${-(;R=J;4210JbuqU`UeW!Fi=nGXyt_~|LF0UD|{}=!4 z9eB-vRDeS+6`Omg{@XAQg}tS|chCDF{btfK;+whP?*0%8g-pVSD}Os_NB8#cSoAbmI(6 zD3wKS^Lrc((v*sgf-ZOhM*L2m-_ft$1R}(o8F_vWU9QtN{S8HNXSfQtA@nNn`c;@& zCu>fh!JOy8f`=bJW{q%KlNfk`tbT3a1hP2&HftLx4$2oYs2G%2L*<)FI^BP@_jV6R zIgjt5%Un~)VbVhv$}%QZd8LfW;4nmgak}A-N8jK79$xDb?69#4&I{8uc%Zy6lua)L z|LLRbH;q6k3;N?ivY8)H9M_@}Ff;vtvWMcJ+ig1)b4S(=jkQQ>ioFQf#h^j4FdbbQ z-s6@Gt+jDpbzWK!{)5hwW=R`c5Q>EY7??HGC~OCyX_fgd`UVmlY$Mc)FOJ;n4Fm0z zAN0W%5bT7*x&vERscF=$ZeK_cD^s4?PaNYL%OyXA8v@P*4OK zlPFH)GNolyhkvHDRbhGtX~XY5oY@(@^LwBCuwE~JlTM#!ZUi@Qd4g?^wOx%|L;cO#fO2?g4kahEs-E6QEm`6d^D>Txy-?VtELx$tSyL?s(J))0X5un zR63~N)`>2RE;KeH(Nru_o`VNTzE|Db>1#;!4202$P8(p!q8S?W3;4Ns<%ygA z^fjT_h`f(VNBxm3b717Q96qkagGUfr>H0hYT;s2v70Q$D`-DAbegnMP@1A1sX z_VSzmke{FPg~#(J?R&%v9#ru^XTJ4&BRm!@j*lcAugy}n)WnQEp|~CjWLRUWVYBq- ztL$9ro}ylTebNKc>3-BXF7T-zH1vmVgD%=9f)r&ec|d9upZM3SxBHdSD+zK$&*RdW zyV72#MNx+Z8G$XLxRIITbD37QIN%WTbSlJ+VBo)BWmEnCV&EI8_D_qdrsGDgnX&Kv zL%+Bc9T#{g;6X%<`$6GKw*^zve6epgZ$g9Lf{|L~y(u+{b>1r_OFUMD;BM|Xxj(r@ zgd%`d%#A6xfBMO{7QTmR@@jv)v*e$5{$Sb6+;>+}Prsk>_LCWFW>n6&@QW26F8|rW z88>Hr_RjhlDR9m3u2;XKUupTGbAnY1_vFtlq(Im% z+9=pN)@B})t+8M$d&bXlT|bb%!Odv7e%?rCkwgbx*tVD~Y?&0NrAV5gOD&VE6k@8o z$aTA4GP_c>!>fbF?$$j4C4zjhj-GuuARbt=kjN98+Ez*`%q5m>_5JrBXOI}_KZIS5 zd12%Y8B<YKd%nlXJO{=|h6zSiQrrrPEHhL^U>Ssgx-(hNPJ>{unK`h2o~#3f z_2b#>t3VPZu3!6$nr z^cjjDyDuOc9N17)niz_$6bGwI0q)P!V7aJQ=8HAh;&@Dz;;e}k$1~|b1d}h->sMeu z5%4_(?GM0?wwqIsUqO?{bdw4JnwKQsR5xsW!^=%BcuS_B19&2A70-J%ebyNgYh1}^ zL)zqkA83lJe;wyz1dVF>;(cVj1A}Id31|u_4))r)RLqXZld?0ygNnnXIdCysuWl1O zmURg_+>gi(3D0^R9=B~$rZk&ABHk>=3_=HcgUttUQ7ZT8v_CGMAW&{A9)Mb=U=hXSFa zL=-FD6LH%!F#-uCN<6d8d3n$| z=$yeJFzz4Eln55M0uwOZr8pBf@70 zaNyP=3%Y(esvJ)p-A7fUsT$WFcFp~;vu1t7#?Z76hS};^4;X8v*>?AwDIRM#Z#(te#fcIxbDw<9aBt*83W*X>r>XoJ)6-dR6b# z3}JIfE`4%zH%SZCWC*)SF1_8QQP!JBB4y#?X%DLhb!Pn@ABF%1K zg#ezK}hn{7NGf=%S^Y+`7qm6KDc=P;6B*}r{2GVW);Fe8sP*$#`Vm7&9rCmwCO1n*N z@yt2$$-iHUj%VuB$ODZzO1ySo@>q7^yYb8)`^xg{ZFU9?6Z3hs@i{M;P)n(f>@Qb= z4cd-Rf>ZtXiawiKO(&2FQ75VLPWMU_)~k<2;H#fa?RLd>`UIijt~@4j^aHXpvYXVo zulHOZa4E1ZOsgy(zd53sK0@0|`Fzm$dTxnqj4Z=&$9C`X4Ko@RiF0%oDSmDC)+bHO zP$k7djn-}|=KRdl?h)(!Hm+ZxLmZME&0V6l5^tL+(ntfCn74Cvsp-(kfSn zUvY0Hc}yDHIbn;SNmj~yS?x>3@1^HGjh=Ab&vwivDGt2F2Dy!XGoMFs&=;BoLUEyK zQ4PQpvv1lRR}B)v15Lb3xe!U;p=$*gLZoR?nPfLe?v#eNvYIOS*+{L@Kp&DP)1b6T zH#y;kttj?F!H}(l`On)%=azkwfxHpPMc6$>0QnXX+d%HHHc_tH}* zWR}Sn-*#?g_kr4NuOcyGp15rqv>bO#(b;%edFlM}N7GA(v)PM*H?A_~{;b2c_gnB# z4T9GYQvt=DP0;CPsC_9KJ=fWwF;Xjn1XS$Q4DsP{mqE9uo`c>3*X#Dw@#>o%{o*`^(RUtc|`^7i3&-{`{BLy=0REuQ#AW+HbwtMRD6HvXzP{rAvW*ZmCqg9YcLDRnU2C4z1ZN-oUkpaF85&%QV~xV-L|~ z87@_*1J%Ihi@zLBDlh1asN10Yj(NRwF7CP`M1A3zUCV36T zB~c`iirJuQl2t`7UY@7}St<7$8DSNZHkYv$S*)OhyH za2)WnVl9saT64}^hxJh{u;AWy#*Pk%q0XL-Vv+`ahJV2C-#SDL2L?>vgBK#+5cn!i znK#?NCc9b-t@&}nRQXYHw(5q*Iq%{z8m#5lg{_LzAM73*=k?U7jydOX*XOo(lh+l` zLYKd;_taeRT<^Ixx<~N%D_Z4!CdqeoRML12-jq)|fUdXBBT>3hb$#;rfOQ`DGLWwV z8z}g~cw&$k=_XBHm!~y)H+ti{>))_UFMrZaHqkB0%gWTzO(8J2BuSbbw2Z@>@WW`b z^*8k@a4){zqeliGGhh1GUZ;;yIz_8oCEeiyef_9Zf((tCGT99lKT~edW7jHS9#f)l zSc%!GCjvsgT#gF{=lcH-C}PC_0K;V+TZ5@l8J;ucqnwRjE=sCokm}4DC z-WxwK*&AP{xHgJhrDAfKE_Sc1(Bm{+Bq@nLBkfhhhiqkbxKxAkS%xT6dQQ^C)~O9g z$?O{Dg5;i}l3{5~`JtC`9Uy&>po|oV&8)o1*P|h7%HLl!mi{d<}Lsx!R zPtvlL*#!fFSTx-v>n3Hu3%yeO*egLePyR%Rt&b1cOxHT~BDOSqliOZ@9IQoDDc6Zk zyWI^x?~B35p7)z%*)knD@B1n6a4uw;WM%MnFM;b2ux7Ff4erFU(d37<-yOWuE;3?m z#`!e|$dcD4G`GnFnW+@Fh9XI*P=ieYeHt~IvWSD=r_^(s1F#?t-QD%$HK1vPii|Ir zeaqkfAYS1Or9r&MU>^B<>*hN{X>fJmbh3p8S1d8WLIoXtE<`6U4Zq<56-^k-UF?1? z#1PfdVADvs3o4p4$alXvEEwSp5HKo(nQ*?(tnfctJH=={GJUgGlOlfB!+`uCM##Y1AYUE2JwU5m%ajGI^4JVV2W4AE z>kAil3T{bS*^}N0!lf*%LdXWwA}XPqfyWZHhwA`nbxKVglgzfVk7z@CTZ23)vLd8X z_=)(Ydu~vQDnWPzC^|J*OozM42G4S+uhg;C&?r|CvQfG)RI5yLfvT!PK}T5bIDN}5 zyx2_A=?jvIkS&UnZ{o|I^Zbz^6gfyL1l!=YHo8l0L+x}I6+ zk5r>W|A`LtkYFti!yc!+Tu#UR$&PKh^^K$J!_LO_>#}Sur#VPqwZSX1@kZcw_AQdkXBu7b zT4{KE$mY;n1Vopp0e|vC2GyXD-S=ZR{2lUV7-#^?$o|X5U)`H#1WP*mlYFxMwQ+sU zn84@|#qFoaK5%_JmUF42TY$UrI_Sx1eYXbX(3Nsbeak5&PD$-O0Y~V7SuaT>Y ztBTVQ9Rfp%j1R3)iC4vRDQ07ZCNgIt3iG8cpb6Z=R{DRYNK%!CZ&BnzHzYo-hb2}| zXiB2n1YM#f(@H@)@UZg`z-I%5nc~glY#{7dckR6+2~FQJ;^;4VyRVZYuMLhM64cKe z)l=MAikzlm){S012|bZ^daLJ6|5c+Jp?kg-xVk$ajksnOmgF_V!h6{-F#=tg(&#y{ z)V*lUp}^zZc0Wv3&hyv#+;q=+AEPyJxd1?TjloFdIbZ=l;en-ky$a1CW}^zaP0&-q z*J~tMg6$q%f^Kp*T&LJF8mjzGc;FqN(4r`2mWuPG$?Ro6!}07Rin*h+1iNVKR(1PA z%LHf7POJe}$>r}9zUpFLCnqH^hsR}zGq}gX)q+b*Tu{5HgWVrIMV>>j6xMvTc+cRb-P+Lyf1l+wYk-yk>OeAwqMuOSm>E0_ zcs*$uGrx~$yJViV;L<$yN2E$+(6PV*9~7uOdD z4g_1M@Q#hjo`B@t$mKp8I}}ezman-%Jz8NhU{YI_$#4@opo{y*V=VC=szV^L1CDd1C!6i$6*-3#r1y0 zoKuX?01oVNSYRN#K)RtwCS`n?YZ=wz(gn@&aeR$$qYF?9 zfvwXQE>?M!yEMCJ!a{fiqAQC>b(3|@x(Hxl#7ufA1f8Euxiy$X1kUqRm=D8J(?gDLM5DHYLOBbkL!HFCoB;vTfsOLL3(%(0tZr6u?KPJZ=xG)g7 zT>C|T8z`=xB4??XORSyo-0`A+cv31lrbq!Pzt3K zk7bx?s8eTCpM6tb7zi3|Pv6HD;TypG5 zG*6i;EK#hL8{QXq@|a2ySTlLA0`JRK-C?UdT3N#=hXN7w*d>!C+v8{b$~xF3vTv#6 z`Pb^z)Uy2=0QS_OjH;4f7|~VeGIb)k;=t(YH9^;HitC`rXH?8)aT*lV=xD>LbvYuz z8Ov<;2I@Qhv9{U zd44)rcoW38CYjR>3^U|jj?jL9Gnkd7I?vc5ESf-px_L+=NfS{;p* zc!my3a5;2g%ei^t@~{@=XUaM7fBJl*zVr%xCeX^EPfLy|a8j^CvxVLZl`@8hdq&x! z(8~B#*nE2WeZ@mL_vXhG05`2oTSGSUlsx1nvFI_#p7E)9>;$?#Pke%{QV5~{=u5ddP*yC0a za@1pwA`em$bxbbZEQ=3$=yk`VT^%Qk6Xwv}+$q?Pt^Ddvs3p|NQzqb38Nze&lcV>5 zi_}6J?zRWY38eM@|v zel(>~l1$C%{plup?Kf9YHSW1Cy7x+;ENm`wPG0$c>GGfcA?9cQ`O}y`{_MBE|IZj1 z%{dW@8puh$yIasW%4lP*{krjEvY6k>HFu@}y84G$*d2Vn6`-zDuj}QCNkj zL$c7LSX>My2mgPjN2?NFt8q`K_YViZTAxwiXKnhPvj5*d&1p^lz7ZS0DEVa%seNs* zamxf7%@lW;A`Mhb5;;8Xj-->Wd8{K-S%vZtGmlFmRm^^%lxPS!2(5O90=0sjf*#ig=ELrQ9V;R>kBpAem05 zsE-OWn7RIyDjW~jI()v9zE7HD1z_j?22$OVfmoKQV)A{rLYY^rxX8EvM?jgwF2#KJ zyWv?X{CL|G_pHGxKa4$JxQ2H0#TX@D&E0Qaqimt0%^8R`SUiWbp% z(8<~2mgcr*;yj_gd=C4W8szIjy+)2S4%tT+;J^#>Cz0lB^ieNEr0PB zHmEyz`3;V{O}f$bm*g2ngEV!{nHEy+zz%VPiHSKuamOf9O~tJ6x;WjC$-E8D>`?*Gcd^U4MdITGT0UiYT)$iKB8K{q z;J(wnA^=h+*iMry(1BR{QzuN^VwI|%2(pA};4s&*kLhr0V5@}rz%$F+Zyqwxt+XqA z(asC#BS!wu(E9c-doiR4fq-pCs@ z*3IV7v1Q%+@V#qi^nnk}MV}m&?*cS<`Z>Q_C=NQq@~9ZB9Q|x+D$o{m&|AoIk9g@3 z@ey$WGoQhEZ*j}?JQRpohSkFP?&ZF@syuiNofJ%}LURM$t5ue%mPc%)=Q?)@pk2?+ zJ|Gx62JF)>%*(z|4}6_h4#mDWaI)WmeKF)tvG=nNwc188A(o<{EYWe++K>5Q8DEqsiPN(PfDx2;=I;O)RaP-(HXkX<5MVHIKXxI z6}a|z0T&3J7@<}62;#i1N;R?K`N3(vE6FFpoxEvb2=y0+j8=3d<2aPE8DZhtMqxqj96%FI_JM@2Ijy8b@QGanWi zXp`B4vT~mV>aD@hkh#vIMY+fYik8#$dGdT|kK4vjojlX?2K&^>kW#uFv3DvanP72= z73YFJ3uYAtwezuV(gul)e&)@(_u+e&Wn>Hx93GK2;Y8|CdSx8gf3eUjvooL*n(mGh zonkZHJGFy`6nY(-EP&!vK~Gek=LKjTM^X4?E}-SEbK5s68I$ZC>;TKAEL%yr+hW=OGm;Ew7*49qqwyc{9|M4 z)N7|95A$PLGP@;uPR@*E>W??ytyk+r<>RqyzMB5qoV_#GzT5oXgBeS{Has(8-6Zrg z*!Jy4dp5beCc$yDb9=m8H(D9@Z<&&y4{mn4XOiP&u^;$tHHv_eZqa?QJv9V&=~|+Ops$As|+pOeELir zcV94ZlKR+ClGG0Dhge8b*LyWW=_UvX25WNVNN&GH@Tug4vH&uRxG-UeL>HAi)?Rpi zW^JAy@PqjC=j?s-X6fxuzg%s)pq&8epo~xhuTdMjA2hI^fP})q(1UhMY+9T}Yhht= z06V>q6y!14+#`LlP1k`9g9R@YcNk0EUJjk>?7;P_bMusab6oZ z2(=@D;yiJ>6iBMdeQ=#LjBk8{f8_meI_it<^d9O{?K(&d{DWkX@6e-1G{Y6p8 zF25t9(Bn9>CZtSy+jFPq=WOd@+W?4n0r7tkjHEdo+q@WGdpDP=!|BSI z)xR{Sr?x5u&ud7jLc-`mT7EENNHRae;oJlfg`p|&6Yl*ihYhhk zz)-0X`H#(IiC+^Z#e%vrmX#pM!Uo?Zla@g5=xurf#6%1oW5XY!w;C4%%$+?ih1^xH zHxGq4Y;qS2K)(XDB#KLD| zgyMQA&@O?>=M14?_f)6s&@(upyl)O&8lC|pQTwLd1jcSq6$W~31P|6p=S#cE2{$0b zOyG(nxbIPA{IW2zyEiM-)jWd$8CL{2FnSL$&~5I*)O z48!45mV*FpWK=NpR}pe=av~^|T5o##d9U_A zd`V(*Om^KL%$-|@bz&^ItIg~N6_QV3>lyNVlnbzDZB+;=R1Zrq*2JV(jRBLV?0>?` zE{%ZV5!Y9}_p1oHY(Se)xQO|D(wRxs;rY-sddjOtw9ffFT`sH=K8V;as#7<7<6`u^ zcQ(Cy>D>(K;5+-@KCAA8R?-^xCw?9DEPS|B{Xl*pyc$UGsv?k5YF54aazr;t4#kre z(IaSHty5RhHKId;LQ<-3lATqr175QZTBm4q**9@0a{jl|~8Bl!Wlw?laEqvm)JODy zuy7^4+O>+gF{P63me;AT|7zXW|5r<`_^|xlo6(!z+Boy9dM`K)hskkJ)+_OU0A1&U z`T3fBOM{pgtA5ocZ!(Az%M&v|*3NM^s1>R5sY9Xub>N7u1?t~@fsK-k&~}iCtOt5l z9nguRI0g20<zz;jTD#d+sX*DYP9vG&--E?iuOgS(6AEpOC&8 zcHZdUpu+WO38aXy^vybLt3l75)hKy!eH`?ATJvD&PV!yzu2F~GFI#A5MI&%RSqVbs z9d4Q|w-V;6=f&wyo%X0um=tAVak3%iBG&6Xq96fsqQ*ugvxmHmJzxf=OlUd|hLZ2L zX}yZXh&m+}Kvgjf@?t0y?{j>SGV{rl5(Xgjf8G<)f{pmkKQGue)o5paer8%BDRJO1 z@F^3^QAKeT6xlU9k>F$?QYv6##jCy3p;0g=DA8^9 zCb|e{=(LrkiSGz+<&Q)^>sK~dnMa&TG&JgU%ijTrj& z=3o4VtmC&r zJNth4#UgzU?7&zM>BCetE-(+}sIA@CQJ5}G2kuf7N{6y-NNATuoDb8{CH{7r^cN4GsHVXUNa_Mv4D8_X}Q7Xw+#X>PR^sCcZ zp^2nJ(in0le7I0!eL~Z?>evS)1B^O#>APdhtp^`;yn1#8x65%8=LcgMOe?cLFqyqQ zvq@Gt{rY4ecyWD1ACt#M-SsIQmoLp14`1fnQSHjS1?N0Hjm}8j&vwivDGuz=?lf^P z@+fX2MY5y%DNQw}OaOEBhZeqdVVPKI6#Op8Vj`8UOK7^^8Ag;zlCCwnUxvS}^jX zO*8&b?h-c=$x}Dc8+^2r;96VsAMCRb4H;H;0#?k!-z+wd(K>9L#X{7zndC8=+@J~x z@_m(uJat+!t(Cpze%N{VixR9h9~P{S-7fPpAKU*c#>>29-$JcBT62Rk3x+LA$?O{E zEZKanUQrs3ecp@R&xN#s49_|5QueV7BUkz2O>U=wbn~Qaq4CDSmUTninm?NC37X=k8=vz>dw7i8v+uGzWGkoVL@@@ry*%~q$36J((S?{jiY zfUt(*k|>f0rTX-Z;3Fh06s8(8G#SE0q#(RU0M%;>%^|38+cJ8Cy?CYvoCjWWJ>UrT zxyZ|)j98%lvvMdA5(nPZSctr~h$=u<`UDf_jQSjpBC_1D+T^a!q9~m#9U7GH!Ik;K zZr3isnUGz>)&Rl_O=u?U1D+Q~J@cmjB=h1&J`ps>b+m|uvdCKSOmuQIF-RZUATME$ zsf?Ya4IpX{JWE)kga!Xt{trTnwp%`4`1YS&;0=#qf{6FkUw_BRxUdvF{LNoTtOGAB zT9bt(f#Q}^WC>JdPBRc$Xpq_>&g-%a_g*%13Z6~k7adqP$zL>z?O*4$qP)Sa&1x8D zgv1l>_+p}UU`PN-dcTxy5yioxv5AVon1{aTESW9wY;<|*R2*@XDd$#(qLaDSzZ0}= zfgYW{JZ%xE9AvrW1_5h&hom(8oOkgUXe!I4lY9$Dmx9*rQ>UWQ74%y9jVbo964n96 z%&9a#wLM;5(z7#AU^Pm0A^uf&PTT1U(Q>ZXrCoK3K)n!#WNs0C4btPNVi!$2w2cj4 zwgqOCA2RF!0SiO&GbJ-~r>6`hq;%MI8s zz%GHQ!Fy`m3z*BYUPTM^C1SM*Y>Rey&3D(@NoK_`pZs8DW@#)y!WNesKK#-DxEt4; zW8eCx%Vgzi6Ba2pS)n#i9MJu(qhgTQr$G+d_xCbQ?E!{pBD z4POdEx#@J13|Aoo+Z#rQA=hV*rKsO54jv@^2C>88!)*zZe2gGbEnf_Lc&}K$@*Wc? z6jB`QU~;J#Oy^_D02!V%>qO-~hsVVWdt5aY;KE>tRZmL~WRRMxo>*Vj;oid*d))Qe zq1eFXi?Ic`+y{!%PPpy!M5N$<_5jG>QPIzg*Z~!mSESnqV}B<#;w4kGx|zg}B%qDe z&tKG1TpC5zVh+AEJU(O#SsxCIKZdqXN#mf9t3j@BRaoKM=?M*XJ#_Z;D%CZnMbs#p z@0UFtF4^OY?)M!H{yXnCh#d|sf9C`EhmO05*Zs2gr{=5}7RV>kqcm8$o=Zb#KW6i; zdiJr; zve?r;nok9WlRCjk0c>)y{JnxMcKgg5Z?7DV8dYnd+i%5y!|r4FUJ4B9EV^lpEx681VtOcGG4 znA}zhn$`)zTK6tRS}5d>TVX*3CO7Qc*(~0`C9_)JBFVOKsP=#(v_RA)m(HOtNJ^L- zRSgX;5Z>1JMm??>(sCC(DP}(u4f}rYbO>@hOY&5P&01zybb3Q&K6S~EC~5j3`$hN*76cT0;F!&`;+0zhU$$R@iGB2 z$Rlgka&C*_k-+1lS7 z)XWU@;~V@W4EQ9!nQ>T}*+%zme>CSfcUY5RfwerHE>h?etNb-L-IIL(I#-OGT0k%W zsXNUo|4*UU4;LlO=&knz&SWGC#fX3Ms6~{m!D^0_L4((PARC|a0`D95->ARvuST%fveBC=T44ChR@$<848f$TmR*n7;E&3KJi4O;+o5pGop?vM(o04G^_%Z< z?tb{0pc5jk5=TYu^Z8sJGi!qjJW=C}2ZZo_1EuDPEqZhdU1Na`tDR* z8}XgXBDeYVxE@i!J@cF!xLI}TCGT#Wc_VlMb5;%S0Z$B!??Iz-AhSBSYXjnWV&2&I z%&V$jDHbLRmDcM-XCsl-Bv!0R2`U1I7ukyEiTA2H$-omd;J0I&i`LG)ZQjlLQC?z^ zlvmf9J2iaDQykd(Sn$p2m;ym3UFh<}@2r8u~?~s;pH>(F4=4yZP;1sx8m+CtxP11DU5A3Ss{p+$#$F$*QJ zwW9lzAq=T}Bmk%WDD-p0j?z>!-ST@-+SAq<=OP+h+e1NVI>jxOpjo@w8HJ2 zkKx`sTsIdh-sp^XJbw!B!=A*v2@QTnB8F&YYz(C5kQ!o$kQb1sRnf<$nQM!^CQFtD zC9+GTm8}UZcUeMk-KkdtFB6@?T%A&_XY)edp4Gy7b&e{Xegp~2DrRRuvF8w>@)zgl1!!qs9CWpuTKYM044lEm#5&PUKCK*?u8O?diNIt*awgVFy95Gq5c2gYm zyKkdna#To$qR$%afZiU%se$WD(?=PJf-migvPfjJt#M8eLav~LM(UiK?r3o~i!U=J zF6%t7Rmd1NHVgv+AFPO29Z^Z&msCShr)}WJ16Kc^y*GhtDn0kdJ>oeeFNSOck_@0g z1cRt7h7on*Oy_pmX*>6J@9kaM+jhA9kGFf9PJf--7WW-dK}XaOWRqRM4HX0zz#SM+ z6h)0hbQ~N6br@9mKMx5?0?`~oxMytJ&%nvqg6I97=Y8I1`F;n@LC@VB71UL(x&1%3 zt%Yo69`0u1?a9UO^~35p5PHBWDYG31hTj^}hu8Zu5VP6)dMlt zTZrqVBgz5?_NSO(N^r*O9>iDKwt29z92}l@>@b|)> zps&)PLJIm!nCshthJGwwQ3s!`^~~(r$J&2)Rxc-aXXRI!+%2}s>Mx1au%o%&Kkr;8 zm`AK+S{2D5AgBNp*JFWY8(2C9ZJUe6;A7pR-SS^p7BIKxteZ-1b2yB)+)5zob4vM; zB3-y7$qqIF(_Ts#Hfk&;)gau{A|2GMQ4odA8aRdMj)&&ahXWr(80otaCh1k)En0`l zPg%5CRwFY~mjULmPIge84@FGZ6$w*UO+^OC?Y`S%G9z&>ln{3*)&li;7O>3@kmaX( zmrCv^pwCFJS?|@VS{}JPvO%R+bxSw&ccsEA2P?XD$SK+mo-VGJ*2ln|QUi^US-j_Y zIIvBA&M!&PC0pxrTBN%s$)k&Xd*#Kxo1omPBpRt?uxeSiD8w|mRq@}Bp~>1#j>RW- zhO(^r>`!eA?>Ot>+9{=5$}f~-d5#X#IB~O3<2D0I+K~9MNeSJCP|N^}WQee&Yi`ff zZB#52qF9-(b-r#hvrw2L&UhV`KR^~~bEU3wUOb~)a5&T`FxT_6(O3Wq=bD>}7Swqs z@Uc28Kg!7KB(18%pcNhfKU`jCO~aK{Vo-;)(JyauCBGi{ksY&m{|B&g$i%UO)hFN2 z^meBCh{JI_JIzNM=d_65efOGgIlU>c7L?js$g+utNsZ(dk-L6SH+Z{4C$gq187aYjay8oSNbf1N0G$=1^U~smAp0#U{n9oL! zs!6%P5WO*GOqbbga4~lAnQ`M9SABKy-`DENEc7$Lf38Ch8P!}|K4%PlsU5?_(Cb{i z=km=j&9jv}wWBPGZlbAQofeZJDNv)hj@RFsn6AN9>KH#%#~vsy8C?Act?$pu)8lRw zQ`7bP6lcgeMH>a3G%7jc?;bB5vK!-Ot-H?fHm8Xv;GDs#>8dN=)5eBzXJ|0 zovpzZ-J#{CzniHNY^5o%}dic6-)^H%rshO`wxL_WyGT z6J!5pje94vI@7iypR?gKJJt6W7*N+0KAzki&@~r@nVPhj-@hvPP+B`C`6uf^b7-8$ z+ChRpk(G}`dDPyzG{q9&fA}A=Wn>4t3MCFl!$8D&P`6|;rG)7vkBTkibrX!v-vE8) z6|eWYo&Kkkko!{%p?79oW-{oHJxhIhgqMZD)ugxq>^y(Fp*R3i&Q02E?_y@4k;o)Z zp`U&FMBd>aCw%AvJq`yU_9Fko8n>hHsYUBwue5;1^U=dgWHX1eLVK)WkVPreDUwRX zVw0C%(@h2~vKFimp6Z7NN=sgm^sBQNY3%liXVSdx@%!8g=Wg<71@%d@9Hpai^ASC9 z=f-|SU+K_Atjx>!8`J*yAGS#OLY2naK~Ju**f%>kJNQ%jPJ~f_g|YXifbgHx1o=CZ zwLL*@dNgUXXoEPuzjC_RHz&9TO8D>i6@wH66x_SOS_~dOC(twaHlx06A3cxMc_(a% zSlBFUI6M*Ap>0W_u?V)5t`*#z-43k9J;Ekv@Yovp?yg{iq8=zI-$mAzHKa|h zS6vRhEF24mc0|}2b)*cv;o~WPo^GrD#RehVjRz4sT4Afbj%%8wmw3H$;9u!4R>2T_ z(o7(CNQNSY?Y<=-Hwip(_h;dwsHWq}#bBrBrDueI-xu4J2!>jaeRb)%DfMn^GtW|;gKKHmyALCa- zUuspr1!Xt(lVvg`|=tH5=!#Kt^#Buk;)w78Wmj316 zF8kRM|FM}|IP8Sjp$XhTI{BBQPkA?K7yVOzM70qjsu^?wv=-k9s}^JQ1lB72#lmC! zR;DCijOH5lR>o!^UG#BpyqDK_%Ota8dDwC2b7g0EILl8*wp&d}>nP)t()tkY|jbuu> znIef)>>_Dgz#@{tyD9+&p?|&x35-Tv_+pqt{Rwm&knNI8oC=~m4d+P4}?96xWeSaDMgfN~-*BAu_>TT)_ zVVW0);rB2_bi1g`qcq@Czgoe8m{a7q_(8-2K_zo1qAU7tM2SYP$rk1Tr%s7_ui8bg z0z240#s8oU*SreFSEg@&-(pNIeRt_Ba-Utom&0q@4OabMOU5e~jwiiHXN?P5$ml@} zEAsR$<)`vi#-Nz&%fFwqtzh4sJ!O||x`(sXD0UJ!XNXDKBHb@Z276N#)vQE;QUjBw zDHpUTK9aV|s{)dxCxppN2Yt@3MT#t5F0wT?mSO~^$L0>M`N(d@&HEnjM!3aLtxW3N zMe;cuY&m7+j+If$0~926Vz&l$LSxsI?#p;xh z?P2=^Gog2IDXDOlukGoej{rY|rmbN%pWfNI4R-kXPRoD2HOGP<&j&S0WFLp|(`bdC zYD#&GB9$PbEhq+d(5{Fc&vMbaph^Y_a1wckNg>YwDZWN!rY6%pgqr_I!;xlAeLfDJZc+((JbVCAQ6gcDm?V;K4U(uLTzI%ED7Y zw<42wn{?7eUvGhx*v{Zu*^}V0nF0o$4vPU8xMWOzvl{@;Ml=rZ{mGlQtSX%KV(iep zB>S|gPBEK!weIcokvR{5+7a{#Whr9dW&_GC0JCqt2WoqK6qUf=EZ(8{^PRAZW73br z0=uJR794{d+t1nRPA6Lea%GdC5j%#l!Fz86*n z^dG>=lQbSGE02}h%ZM=I^87K&;L`XR^T&DErUdN7(wn8{qx#%1c{lKoA9!xj=v6y{ z`rJ?~3CW}}nTs{5Ib%{ThA$24VQ}nCbNVb+uPS#V8-73h->(TRuIz71CjN=6;jk+U zDyM_W?zd6OEf_qFJw@`P?s%^Tq8()YGsCk^mMb}^PU~Z15&B@`<5)5fqFMMGB=U8E(r4i{X&$j3JVEaC4=r(GP6RM8JWN@5rhA` zcv_u%mhS?{L`iE!UA$hmG$0X7_GqPbc#C=(Zc!)0WA{9ruDxc3_GdQ6fi36W0sCk_ zah6$WOFuib_?o*ZRbq!^cE#7ON|wX=_<{Tw#MX-CH`Iopv9xpU;L*7w=4?lh{rUC{%W##q?;SZ8t+N=LQ!#;j(#g))aM-m;uqts_G+x<9k;hc*X6CA- zLUF?VipVHPfYOE>@neY4U1ZvTo4Z2VD@P9Rd-Ni5lh-B4@#v?a1uEP{q=#4Ko-`57 z2(}tFsOm-abT<&&nwd1<3}KG2TiO-f65J9D>?=$oSg7;iCMimDbji0&X$MB|+HeEE zbWW=xo&m1?(s+wUxB~lQ}XDw3h60mdFqw#cfZA;7A#(U0NQMk7%PdpT(>DJ z4u8|uOxdX;Fo##o1dPzyheMG<<3sC}H>gYybuF8mMMLF^M2C#)&=M-Ibv-jDyrd(t z2JByM)>p8z2Ao}IB`?fO`y6|uW$-R4yQRo*oh)dA+%$4-ETnEII{HaBwgwsAIwHu8X^Yx4cMIgH7gFz6{(R@Z zmc8DjZHu}U^^oKc^QDPA0~m?;@Z}Spe9iFI(VwoE@Ui+FgvIfKZzNH7->j9LQ#WZh zPv4^Ij=t+#KxYP}s4Jz*C*U8`E6<0fe0@;yH1o|?Oc?6bkLX=^q3^@`gdy(=t9OCw zoACO5+qPIX#iAUZ8SOO8J|s=bBBmMGYtN`|_+n<{K#1AD9y-QD9_}dba1cYI{H-s( z-6p|`S8US@Y*4}BmAsu~M22J&uZ+LiqgB!4nH^jiP!QFmE#y@I@!%tR7Y{Pu(wiQK zfDh;%zg#qCIB^gjPi@AF2X=JMF#P=Ar>iXUl=jq*50TWD#+^QFHLDd-N~n>@Ma776 zx`?^JWKC<|%k=Hbw_2&+hV0UU081R|jWPQg+xkaG8{$ z!P2_*F&UcA$T`uRarb64~sK-@9AvOZBT2e<90#vKWb!ay>=jG1^y7 z$Ai7d(BM+G4vfT69+J#`Dq5{@B?5*WI!9qK?8PqJ$qE+Z|LZ3Y^Z6E7OxyCU8nSG- zG!%z@58#;_R0y$&Qf{P30u_5nxzOh(ta7%5f))pJo)js{Ak&=#UdJb|>u!G|evFy~ z2FBS@u)u^daLJhbwhxxq-F{#3mIWBUYE1q%IXWCLIJ{K3ZUu~wC}k~0&X}78>S+T% z+q+wu#4DYX0o)s(@fHd7lXMuzYn_rYCBs+u8Sl`nTK6=GIg|$RA&}<4QZNV{H89A9 zjhF5Y*f(hvqbuZH5t%mxHcj9z8*s)KC5?nCLF zh&I}Mtq$*nn!PLx84Fa|cp@YD=r|zlqw1eN_?M(tET{_pW!byr!{JcH;XKq{D^%^I zlu)9+gNi*j{}#yvIRV{YZ;^f_s=wSKh_WO__CJHf0PHR2H9){PNzwXMNIG;u^c8*~ zjk+0IMsQzyyi9_NKKW5B0Kj5Mk|;vn`_*rBfB5k)fAahPku0N>izyPv3I~pVo8xEN zhsxiKS3agJPl6+t~;{EFMOh{2W!VsOxNH_&}^xn#9nyaBt6N;mE;ZjGcT9CUeEhiJMuj7 zJ$&9hTyf0)t2bP_kFx>(*iBVS|GuKF#l9wsQf7nd4)ip(Lna$@*;htMlY#8J9SQgVF?(#VRbnMBR#bHL^ zi@GQ(75`&Cv_R~;A)=3MDWN!PGVG8%EfQin;PFuaEMJ8>b3>eK0p|#=?5=5Dh_TK0JC(Nmr*_&rvFgk6ATcad z4k1a1YV$i|YQsxJdR1m*&C9*oY$?smY3zJ$ypT@`@?$G;;_8wL*EqC9!|DN#I zleQv2U#Nn8J5oP?xuF0N4)qpIs@N!aGTkKDLRN&v3HM7HqH1{!Q5DjKAi0ppyBKAX zrUWIBD&8^RWNC>>Bge>2`XTvn+KSLUl7u-uo|i-0Wp&<}bPeg0;aQ{anT!nKLk!Aj z1N=EGd0U0^TNIWk*yi=27s)0Lr|ouIO)8m`5_qr;R4nR8w$W>q<)SuH67Y%oV6d)1 zRU7^xQ>w0@TNML4VP?sV+%lyNs%X>cyR$0!X%bz0_&TpT@MT6H3eU#SDNsDS&8Tm& zGJE4bDXU4gCCOmpt8h5xZiiAsuW6@m^2+HO3SFYOOOh)&G%XzpM09ZzbzSnck=KbK z601Sb3pPT{QA$`!m`kcdh79}&0X5`4qrD?&PuG8YTWra9Dz{%C8#!!CimYr&8l^N) zz-$|PP0-@I%qxijmL_?+=Wj2lA=jqF;EpGfe0SDKL#R&tvs zm(d<-tNXDsB-D(3ccQy zZmdwm-AbKC((^a51x2ZXP1i|0y8>4ZJ1BWpFN2;^Zl%bFRP36$8^bqGtxjcN8 znL(vXO5YfiZOJb6NlojlGi0lO zgDlg%NdAfX7IF1dJ5rpuI+6Xo(>s3gs>P5rfAw}I+09`$w%*D#9HW$#6e*`-AAK1T zGSGna=LhldoXoL3;q5H@O; zGp&llOuN_0;38%VbJphy-3{k|BEK%mmF%5!R#F;l08YU?`kL=mNf*6qYIk(XT%(|F z{th}D)R&zF&WM5WqRH|F2I57B*-ey9r{{k9;zvaJW)5?s{b;#R^*MDP=rG zRu2%d>w3HzRM4x(v_|DmH3^*NM|sh+AMUtgI2rCr*Y2~AlQpToZTyA>6Srqoeo9Vo z*tE1+jsHbTd7dI4QL)z*J2WOmH7Ri~iB47>3heo^UUg8@6`7>C74zY;p)i53TN|`EEQ4M~ol28#r1Q4>URIQV1OJzW!ftX`1{dG)M)gdH`F+k{ zb31<1wOQAqbamhD`LZMKj$=4_k#S*r{KXiwkE7uVs4f=Ynq*5q$=UR>9bU|Maf@$O zK&zr1l<$nRlfwYTm&Q76CQQKoOBb}XSUOWW-XRybIgEw~a zG{2JpH`CX_9W2N1Nuj|HBxt0OoGsexw}cXYy`{ zR(tIaOM-0jx}eYfT#pm>5a~A;P6FV?m$QRNkg-!vTc)Rn-wrxO7IAobf_Uwq#+(h5 zaveq1P_Yd_Z~|)E^UnH1|7PNx^6+F?vh0e-uCXBAu^A5sL1A;1>p!vsMWG=nceVu- zDT*KEkUi|;D%_12fj2W4LdPlP5sDn9V%z>1SuzbwnMbE$x!-zbv)D-Dpw-3*r7P_r zwXzC+k-RXhR+czp_w@D=5A~>e#v~PS=l5sz3dT?fkf3703CVbVW=PW13m$ z^62cK0b@rYHuyppZ|i>W%Ui(~*W}lQ$JUXw381$$cup^&l)EUBPsN^;Kae-dF38Qi zuRRbkNcP#GS^M>T<-WN+P+wNaOQE6T*1eu?l+{7~*h6MZ=rP_o*`DAEsrgErr_<+k zMfX5Kz$Wqb$){ceGalvqd_D9TjmSI?ozI8f*vUPtUR>@@*KgN->5grW@Lns_JZ-zp zR%FoFc54n?VXSHuQ#LtA+|T>jEJYoJvzo_}JPJiS?(jE;mw7ys?Z~5Cub<)+RGfsv z^Dbuv74F8scQ zc_|3}AJ7~l3$>>w<28M5iE{kE?(}5c8ow*y_oB^vf$Ub5&2$rFQ5vY`A6X#ADUc7f zi^J`C9+vGMV+HbYk-xq-bA|;=rPKbjoaC}wM|1ZrAt$VmcaT!q06AR`z9%>VY5(vb_PIkyDuJ1Qrw?=Au2~$O%|XvMLUW_u}>n37`SM!pGj(n z)aCI`%q|aqIP(cq^du>csLP<vx^|d4?OYu-^;d3 zBQ|a@hpSucETpn%tktfj(RLKiUO91nbS-oNjm&~L{+x$9(;@pa+=VXQH~y8qKerW) z=j?LXjtqPTRO4edahuQD=||L<;xRXa0e49Wh@W9tUx$0mt`v~#n-#!yakE-gNxU*u zx#&*V)=0hTB!7kQI{7rpFwv2WAKQU(2m$urf6EcJ-}>V7SZ~xaH-gL0?7OgMl zQTuy5k#6Yln21FV`4>C#;D=l^20Pe2c+PTL9NhWqZ`oQ9J1;8MVs|MH&BF4Z{0@t@ zws3_hcLXp02k+@~8nA{-<#QG z*CK(y?SUD36&~nw%Ye`jE@N_pJ)R5sR{_LW+fo{(S7oaTL}m#CbW=JMD}=bG&+Whr zygG&6C(c#(xy6e+!TCit+Ggcir6Cgfo`H`Gc_y2bqtG%mDxe?st#Hng+BcFv3jVgn zHW$QM&&5u5r{S$f^LME)6QiI-@<47BoD&x-lBHd|bApSa{ejzkKiD7GplzPNm^wDU z*A0KS&TkwOfxQvI!MTxk=xY0L7{&%td#XR$X$z4rl&P@;n?=g={`V*ARZAj|M;jQF z3waXpAVP;)1{rj6ND7@EnC!Oz_q7Dy_0t(5O+X5{W(=wqMg$wDv%)V%j(ymyJDa0- zrsIRd5fwWfAEtgiXkD%v{PP(m(3440SU5wCG_r=SkA4IM#)uhQEgyhJCs#OGL&j_* z&>3iNM)u#*WBAvouAVA=weQD-ivOT}-m}lv0PO2E7UzBdgT$;E8ld z@5};jSn#{YWVlBXC3C^(4s%QUwqlqmX8bR{oB89{ED_oNxLv=M9AFobJNwiUb^-#OK8-1LB=}2tNu)giEJBPuPgS&RVKJi^!e}%KQ z#*S}t?)9q@U|UJk7@?s zovuN(%sw|vr9bvu?WMzvPL@YJ)bZvBU0OQjoG}?WZeDcde^c)^t1UCrw3NTCCz%}1 zJ0G)}YxYvg5{m4CKC#KiLz}c)fmrlV$l<^&x@4jrW}3}Yo3#0}Z_m`L_R13jo0v?x zNt>!^Wol(z3M1VX)x}#g_j-hR_V`4x`L&~TFDT}A_?{DHc~mkN$&+BLv&B1%vfXhG z3WFz;7Y#1E_<hwJIRBuq<#Cu5A~{AS=JatOl*eBQCQeq z=JJoMU_t%c;lKXIw$Q*%h90NqI@ulX3VI2YW8mzKELfv#mmPXM4n2B@?lih5v3m5} z4M(kd{q67Dk_56TKH=`ajM$+R1W8Fng?x$fpt^{cMVIrSe2xGM6_VHDjI_eN6!;{~ zy@)13o4ggIp35eeP2N4F$NP9_uEeEID?rpCybtzCo^6jVz-{jK{TZf~FM7ZY`TQ;u z6z~4X7yr|%9x(+XC_rmf!X8LX07|BawN9;sYZus7xmFjt53Z#=!M6f}%PuAa~yq+ra zD4p|&#EY**qHxT@fRj*5z8r*VkAyaiMGx4>EQrnP?t+hhGQ9I6Tla&rUg>lEB3+Fn zo*&O(FAKCssM4a?a?igQo4d5XOl_rgdZPzX;1nvhPKHvFX}qn%EFMNnpr}s0Ly7U#WaiN8adVTUCxj{VX;lKh z5w!=@ovstaPj!9Zo2IeqX?wvI~Yx>`-@elE9WxdoQKB)>AbIrKE={GTNxMhhLFfCFXj1443>3Wy4~rTge_SiNEi9!r z@oE^DuRIS;>f$wOF~f6nc9S-nsUrCv4>a*iWq@9Fj_i?-(u?8rdtm)mjG7HQwJPMB zwvl+wPD6HL>sS;P&)nfxdDqIWN`Ut>x>J%AxewZcvH~z>Qsz-Es$|UD3z8Z1QE)+{ zTke;oDT&UavlT8ge`vj&8hRI*zy80AO~_;Xq|Gxg6N{xOb~BtI%h_2P?%Gh2VP$DH zQ_4h&tfyj8r?@;kMcgfIr*XC{iN@K}+;yJDOH%Ch9wpo0;25*Ok%M;}ozvb;3Cn-e zf{gqr%so;u94|G8$G^pD{A(%Y8H$`VkAE9f3oRx)=mym#AXi&}&BXU>_%%iR@`p~2;dfzhkV=`=5hV?p~uHd9Tb z1UVKHtQey*juAX_#~+*%XWD<}!>mQP`_7*I`u{BOvDlpE85e#_5;GoC*&AAr-j48IHMJ)8Ym?Q zT%Dz2F>Z_8pH14O6P9U@NDc`Os+Ub@Rs5x#MqZ0F-m=Ip#ZB$v32mA^WCJvb7(zayjl77srgIzYrF@} zTNYUj)m(AHHT+uEYpIcfkoBbH5cHm%v?HVVs=ozRrg;%p$yN?qpHeHJpygCdAfI#&=O1&} z)Fg}_9DQ6gUfD;H$AF-(U7wG<)(zS#Uu&8FNApLc-~!XAML#$vD3dM_U7MKTuGbVm zb`d0(qdt{YLW@R+yoTO9wMu<$Vw3h-%vDK?;(=^YP!ccbi6Iwjx~fHFhV zCcsaqy@$Gd1@d}ty{a>!c225z^V9~*{r+g)fhUzk-IsPn=LFxQ3jmwRA?ff398_-& z>JeTR?wf9sn$OJSWe41nE#p5422wEj=BbU^DspxD0cMeVCa+ddL+44M_7wh*UPXg; zV1G75gEE9`c!CY4Nswlrqoi20gIuyVa_180_tE{9gCo?TPsf!hZ6a<6hxnxvH}Q8#HZIE{NT z7WwR``AqXY&1d$y;hDGxzxmlc^Lg!2DdG~f1FrwbAnB-)v3K4O1PXE;syFF*r98J9g}we!H-^bjm{26F`0S@Isn~| z28(LecpHQlNUPjH_qr{cbzfTM0e!IIS~+ME!S(q*N1+ONWps`-(QGZxcuQJi55H|53+i`I%Tx`G1@z3@x*lT|(Ocq#x z^!kO(N69J^AS9C+W>iy`4(be?%H&%LpGsWDJvgkh#K>PO!W{7+kgf6k_~>`?YMsM^7D zU}Yxs2^7!imNuxMTf}Sy(jf@_xfBJbUELKjcnsNLW9%v{SYcy)K<>?U+m87!RLR~R zD`S#HDWNncm5N32x-P|9Ws+hSv)bzdtbRIZs0B$@F`ee^sM9ch_wnPT2K{=Hq~&`aYGU$Vl` zX-Wxn+f`I-$`4M>|KkVdpH14&$e!S}%D6CahoKDgxO@@+(4+$)=lnis&|LM zYLNnN&O=f9lqU)!?*XqyZa(Fd@)8`|8iamqEhNA227AQWQ*cfHY%iX*lV0ilP?0^Q zIk+aKL0j?dt8X*{70eZ24LL>jNJaukr+{V+JNutT0K#)?-dz7h_gHczr=~uMG-}f6 zLiZDz?BERF9a3FlkWq72hRoW!4SbWdS^6wYfv5fbm z65h6LW8vA2y_!_b-q5Mkt10n%pk_*`y=W)S*^SImBF_A6`;45BQy#yi*buip90lBrpZmJ z)4)AtR-$2Nvz$fy0O}q1?GVZbPGW`jaer92;KDl=P`z=q>xbkdhe37A3RIUUB}f@H zP_aep?zVLYW=O_3E;EHWBz_eJx7WaqALKD6^s zeerX*bD@iOvwH5_t?$a#{qZd~3n217fA0elHvvkf2FFnoDCJs;tfFF3S)^Y-0%rzf z@7gB)ztHA(u+H1X!#~{ZSwVuk$(Knd;}UI)Zdg^YbJkxYc1m>2i>U%2!Cec6PVcpB z7V18rAYKaOe(p!d2|Iwlxmnt(KpB)CPmIR&@btF{Q<|^X zdLf)eiyilU}>Yy7N(3*XvBVzzE01K}d}7?jyg99VAwc7YS`SKR62s zJNzCP*I46G#$Tdrl7pX+q=*ytsXvNwjnd)s2gX5AuvmCL|N9(0$_fW~ieFZ#QVm!uJR657-*Hg#Kju zN7Hd@G1aKWZn-?^z?myZi#P*k&t61&d`O}3HZEIG(P<07*>R6hk9quO}v|1agGYjD*Ldn5Wzmh=R zhyF&v;b|Fkx&{PTN0jPi^?;wA%n|`luXFW=eT!p!`|JEKI#Ue#pZ%T)2INt57o>)2 zMHNDnD>Hu_o%h?pQDOrE2X8qlr`ZR=aB1VB?(x6#vP?Tay+Yn4>tC8_r_gHJNu`wF z-EO5~x6WMW2OfGz@Y;KctjN6xl$Yh7Ybe?OJ> zB9PL(B02!!E|jg)VItt;8DoeMhoCW*Z~g#e*+f0Z-M!el+*ZGv%?&slHnk%|zH~x5 ztTz^URfIl@D5Ww<7UXER&NR`>Cfw8-`0M<-7a#{J<; zXi$tw6d#P)a1CG_HX^62AXG*v0aN?1m+(`+-I7}QJ$?n<4ei5==hmo@j|}(M2~2d?++H`_ zP9&891KTKyIj|ik%;VR@n9nJkV-R1JU||a0%e*gtcBNlQR3%7Zc0)7%XaQ_AF!jvB zj2^)3W9qE#cCT%i;zD9-|kn?ET8`Xs+8`FpYUsC)xhLYE5JG!bM!2eS0flL zNH}03*rCRO8;;T`pP#<_ZCkIF4K}!2`2dsDus(3PV1=*;LZiAyk31;YT^F?11AD^~ z=ah#d0c^ZaiI0oYdv=D56*TPJ!xg)`exLW1FFMI(1~gn4z3O<)ic01bSqCbZSY3!r z+s)E!VBt6(njtI@VSxzl*Q>IkdIGZO10g*?(2wd4F18S?hTavEh8^74pG$^`s8Io1 zr!X>$e{%Bc-)JElIPA{uva$+0C}j!-MU$}`!;eeucrTs01Q;-o7CJ?IAgY8s2Kf(E zZP@8`8Db|{bd~#ss4E^u8qbLH9+O542P@>Pz=0p^jO-}xW(AUQ>2AMj|D^>aZ{#2M zCm*v5yK;DCWqe(!*fiM?iMqY*^u4e%0>~TH(*`~gAs7Ytg-b*Gg8NmF5`*-rB*jC5 zRlp{}#{%3+QViS>iG6Fs3w%1KnD2)7T7WqyoFcnP=^UIQ<7RE1RT=Qa&(grw4u1?L z$BP-xGWz4FIA3;5hZ+!m3g?7+w2wN;vxJxijL8z{1liY*I2JHsH@Pi{n1(;K~u zpmr$}%IEJX?$Vu-c%F-=wIN2v;dySzm9E`!*Ddv~ZM!mO>r(7AAb&1hI@b^hZ9GW| zB*R^+>?Rn2xvt2Dem0EGK=YuvC#ZBzt0G-|Re?e0gWT@;uHYP1+g>o^d6-mUlwX8B;~p#%!J2 zq)nC{7k83Yrk7tC1oTj%QgNE(U~sQnJd+z_oV(P|{7D~F#y`$m8d@Yjr|EUebib!8 zoV{#TuUpo%<&%#4b^7+Yoz#31ktW&2>vg*&>GshFHH24$A5`_a;dfqB-vgG$T6tvv zYU`wgI1msTBTx=O@QYCT_B;814Ya^)$#4DFkdz6e%xZiJDJ2j;XG6u2e`m;L#{3g0 z2%5CVfT7@=2KP2WebJ?PDIq&vsSMbo{%{&HQ(`fELDc>JqsPc~vPTW%Ui4N~aoA;8 zE8|i2@AixSIe2^&9~3FgiB2c6}EN#oSo!x5qQ*N#nP!&yNRp3b!AM9!rkTfB!G>%%28MYS4ZR zjL(f)*LY4(M`7gS{`3|u*KiiDz;q zB?bYT0C2gPuf^<7FMnOoJDI`buz*8&*9!6_2N`T6L6BoYd$*)vwu{Zn$k%A2 z3f5t6#ty&UUwv$_by+y;^V;F6N(4?F%nM*A&<#b7_^x1|C-C&3@-*fMnt)6>Kgtpj zGzxm?^m&z<4EnkvVd|=>2$=1@SQaxHZtO6_(Z9<{TW9aKPVZ-hm2p!S-y9!gnT+!K zez2DqIP9{NTg{TYC}loHa;Vr96O4k7zE%82+x$3TCB$f(v?pWQAxaz4@ybQ$ge`w%r?#G6A7ctIf<#Dd4qq-uwIs5ULw62W^QNti$w0A}C?)}BrEoS9|>;tz+&2aUc99|pWv@$hK zl(LZ`^;9fooExA37+1O16)9pvWVWVr7PhPw#}virFe`P#*G$iX9c<9dmPM~rY*n=~ zM?$MKWuRjJn;3@rXsuCl1L0^Jcxd6*)Hs91B}Gcl-FeqCV@Bm7KIl@7t-tbp}O zXlXGDFw>-0^~yK#(G;c7r&ad__1?N%;SOb;3`!q_i>F-#b&(SJcBnzlqp||Bg?Ut| zn_plP9_DUB4aimp7Z;aPN@$eY4_uiZH+=U<^1V0u zA!J~625Ey`yjDe#{Lq{wvc3qAH&!3>I}(cAnad;V1o2`N-9@i@C=~9?c63 z)>Pp+R~1>(OwFJ1P%pn+ROV60Gcd^hp1|K5(=FQQ=d2Afn*xEWK{ciT!m)x@bMoG=exv)tkAL}--~W$f8KqoIk+`9&pfNBKgTI0f`ub5~K}E5f z;S5>+(x4*43Kg3vWgTSPhVX(sn_l3q#l9Ezt^oy z^`T#-U%uy9zH)3H>R3Kib|#y%S;fOsf9_-3+ccD7jKg6QJN-@>^fHnOBKDc^J4cMl zz->{hy%yw34yw`F%n`QJd*wTnIHmNt9g92`Y0e_u^Kd=3St0t_wBj=JFmPl0h>p_# z;y-O=s$P*f0FNZbj9YrB8IpiLFNK`I3EZs!i|0ma@ zWngq1_dEI_^Yk*8Z=4M1GKejHN2*}cbrR3vBqCIe5ArYdlyWOYK15FfMBkwgqzPjC zdqHjk`b|nm2~c@pe@VI~!M#}ogdRzX$^g)QkyW`TO&k-d+!0v{mk&70<;wWDpY*wdf>*29h$bw-uPTcfOjM%!#=ouXZ_P?#(|A=Il{Bv&P69yRm>d5NgScMNR@JL541 z2paebzowo~zG*QYD}H_G0Xf5A<8jBzcwC{B(6M#_>+~*zr0Uk7{hsO5*7MJ4u1W4G z4yreXx6Xxf{*XRJna9SET8Mx&E2^QY1uOB+``_`~7_w4O=CO&tM7GVn1hR!qT1?O& zHT!|6B8Zu;Aj|#E(dM%@hNJ<{6#h(P@E-eT_q~eWgm+#kNTDw?wX)+7#$Ow>IBb1P zt770U!MkD&V4N^cxIe5;RV(XKIEuOv*`8;eJ^FW2>mbquyCiXvM*<& zXm(V5*Lau|)x;2K&MaeIxSodG^5;^760-H{x)kw@S-`JaUcj6eU7fz3Y4m%<({;&j zKsWVDS*h0o*G*W?BIN}GKH|fykTU*vDgSEtqQ_frF#BRs=ygH2WEFCAeSQs;9AL@Z zmMOQwcG4-6Z-qJf{*U;v55ABMGzMSi>TOogpq8v}{e^8weg z{Y+Y90(Jx|(;|=aI(+erF?oAO@@!rAM4#Q6-CN7qkXv@%Z)?CB$X=nxt zH%XBLrk}RG$k!-9=e1(&MPwX?>oX*c$rdZPj+b9ll3!U?Li>O0B_ns(rEfS~3b4|O zKj;yq?54;ADmI$|0v|y62Q*za%K$f`?iqW9Zf4grM zjbifVDoR{#B?jfHmxbdAy7l5EVR(lR@}!M#K!2>NKvV*X*g7_4fw*Ktt3 z&hkL2hYeoXQJV!*oNiHwX>zAbAG9_yZdRQP0!HuNVp0R4FcfO=GU>8cphmAw23%XS zQfSc8h6wb#@R_p&LMM05U1%#{#Ky(pF#7Ds9o0d#u1Siexkz^Qp{c!CxfTuK2@_pS-+g7hPp?}sP_MW!7GeF!pMG-e*~2>F=@nzM!S2~})}t-{ z&X<4kEsO7BpnASe&T=@4`-xQ)w}nzRQ=|zL#7TR|=YHuLqo8YEy(~*}L7E$r5}Yhu z5qd4KR@5v1gqO|h6Q_!A$(DGmnF#UO$ZX#VIv%(LtH@2LX~5L?q=u@N&@zwDm0cjzn;jeG@n~+7 zyiMEB=!#nr1||=>w+;OB(tTk|HNdB+si3#H*L&ZdzI1Atqbos&kw0{iJi|OF9hXJ8(KnXJ$raM|I9RsB!7EY@!TXyI_>eYzEVHc`rr6iJ|BlYoh?EHF2wXYO|2r6dEEC;NB{=WbVD4sD2P4nO8v zXbc@V$H6l6T37C|4;H#OZ=S7euN`&VBkIGT{gD*_YtO@?0#UORRdQYW?%8=H4n1W% z*Sd1I{ii%k5{|pYV=c=+on#9M&PL+wV0AsewV!}VuUa~FODOaO{sXJl$Y+Pu8LwZc z54Lz3zb-tsj-+we)2OtXDoQ9N6hr1yvFY35xGVe&?)Lp zMZCDfw?mHmQ#Hrdv?bvig4BrCQ;QaYuCxWAr?Jrueu(j?8IAA3T<0m@tz z1Z`EUgXnCZ+j^hf0ZEE<38>DU$Zg;#7z-q_!wf7^2&M3g+dAG0+ zov(kz?}lx1kWG}8!(OYM9H&u`?2`#>EPB;7dJ(xQ!LHAe={I~+#kJv&BHHPV6N~6F z4>UV^RVI)Ko+I}hUD-Zw2wr?_B;a}RU7bJUFW)SxvE@Ey0}&3}7&{csU5d3zv)5uS zaC*OiDe|q5*2x-FCTS{fk4ASb*fjYRNe#@PGr`ClpzlW?nzlde{S+F`pyTEO*J)~C zGIQ>YK6L_r|C@1h-_N46CXM=3H4?-OH*qk;hQ~6_HVpMIV!wd0kMCIs&Yp``}OPurc^b*KS{{c%--GBpvFPaClm= z!&AB?uvvMMu7U-6N_44sOJJ#Ig9>|mN&+v7DnREOizQR&iy|Z`xWMQ`H_2L62Q`Qq z*Mou`7M@xN*8&0VZr?4*_=V5T5X+jU`i%{xB%NL90f+H(#LBPTO(_9T1ypRQcP6jQ zBSQ=vC(;7ZO2zG&8-w>t)_CK7qu^H529hm32QJAuzpatIZfj$*LfRuLp}_hRb%MH@ zUOEBK+ZtIZT{Z#fEOyYFg4c<&m@MXu>Nqp%2pb`wo+4{tW3&-=ZK`LdEqmY>%DC8J z6kMfUA>6Dy_u9bk25q_MGsy8ErB0IK+-o_(c(fY8gL#}LZ4Ymcq?1ksF|}SdGigb$ zTaoy*U#hswBVUs0*(!hREM3mi8Px|KZPRvGLoKd$f=Am{TO3p#X-qpWMFAAAw0ZhXW zmSm7pc7Wk->=`fTwf^H2y1==Hg6^I1E6gdrT~u{1Da+5c@1FVEXw zC|&fP#DXRH-`;E>>o^?R1#;U#jjAb>avMdqV5911x%l3^hSI@kwf3sxv5%C8{#1|x)nsl+QM^ONIQ#~xSkgybmYIIl<+vwBbzCxJq zaRyoy&kIVTA#4mPeZX9a7orTGUiH7IznKNiN^MAPxpHDRN!K8^+(PZ?$+~onNqS$D zz(4DMQ*fBv^E@2*{8?~hV6<;+SR8pSe%lf==r4LbWvgN#wNc5+ZXh^YT=~|FOvlt+olAid7(a4HNE$p4R1aA8tPSjNb2UlzvWB6f}&dCz|l?rQvH?nUojso zef|B5^MB=13+W6@oPoiMuFcg5e_)$tT(flsKs@W-YAs(TpX4^H4oZY3+1)m^`hG|7lCAblQzM<%%hY};N?n_ zrLC&CupH=2U8XG#)2lWGR|PzXPM_C0tBZGvxUvP&NmDcuJUetv*_1g4oykCQIPhX8 z18J^xDV?7z)$I-_ozoq0J)++(HUVi^x3rB;q5D_aNs0_j5wmKxnPckGJiV%48ahdF zN@%{sMrdK&3zUFsUnF{@OW-JaPr=rv@)jG%}6@=y>2FXCBH{d zMQ};9k7p9(@N!2D$ zKL>xB?}T@c6QU4fuUjR*&KoA>kRH!+xUwCjuWD2VJ{Y;#J-pj9YveVmwkW;k1Kr23 zV~t;yN16opCHw5q{A~x2Mc^N95-b6t<|GCF;TfRmX#SHOSA58W+Rs`}N6zmg>{jg{JA zwE#$@l5Y>+c(^c@A!pi9PNQ|+_t zz&jRLym7SahvX!O(~-BV2Ko}EgzR+#6^j|jBn4`TwbAI&tW~ztSHibWIy<9q?o|Ph z5_*++RLr|Qvo?HlOrKi=gX;2n4G3C`PC|86k^<>BkV;6e>Xsf3{M_#tEWvT{jtOo& zmZZQWP+V9xgN!Y@dReDux3pP$KN<@o`rK-PN40|A1(Yn!Qk*b!>!Tmhn7NK;RtbUJ zL-7c{t%uhRiExKu=O|pPotlwf}yX@056P!MyZ& z-HMfp<r-D*KOz;2|yGVLp+U-#;Ivn@_q{mQ@EY-71 zfvf1U$)FRhy%AAOZx$a^m(9C1>mI)v65J?Jb~rH2t3jJa-n|#LC+J<&9FJ#q0&V6_ zcdQU@(qi&Dk+%?{S9n65ED!jbFz;!fSA8epd(jpM zeP>^;ft0ZGY`7b~B6U`PJ4z`lC{l)MoLw;YU_5hoK!2FoOb*}YhJE7YqP|z_0@p`( zMPm(Bg?x!JZWhjp{RLPl^qp7wGqZ`}(m9t&Vh}dY9fi5+#+y4RT}TB1qhp zicOPCp`fQcyofpQ`a0f%>yjK{oougP23_Ht^W+-LnNjHs^6ozHU25we+i@%_XjBQr zw|bL*`n+8ptAl#onn0B@&Ff?TQFtra{VN=Op6u@G{Mqc@Vb12g3;%o4ZlT4v{I+D` zpU4_^#)Z2xFUZypF7De#DYxJs8~f=SS0!0Yj`v9jpcn-kAf$6G>Jj+}7`#7|e=NwL z^LZbF#%q?Qlixmb%+T&A$_HjA=ke><9e(ZiXqv72V#iC^2m%aPP+K{rOr_h*z=!!a z6n|+_RDd>6V$fEq>8rQq-dH2vM}6@%};#eBjc+U3)B47+nHqd1dw?j{AwPfl)$l2PQ|VZTEaYlXhs5m z7Zmh=a6_TX_3sVMn$zuBI;T}p5e}7(P**K!;(;dsTrmrfqV&3A2V|SErgD=PNWTj@ zXq{PP%FHQR$28LChzXXn$uh`U-|$^Cw_TxEp^|&HA_Ecty>6xSLg8K6K}~{lnQ0fo z2$1$%q&)W^D;ROtSG+g(+n*~eFxtF6^di|bf$X+|Q6{BKqlf|GCGs^Md6Apuf_THy zxpm$~LA$Ig`mQW%+8U4Z{<~h$A>mxR3~Tqc2^#0L4qiaM#{Y|JrftUoAk<-w{3`jT45eLg%bUM!Kv}ZK9M$id>~)aYeIt z&eEwH!*NSjIi(=V1kqlsNzIn%ot)h$XmbYO-qXPU0a_XH`O zQ}4M?b4K-mzUqHeqdOG3JqCiuvJARco(2C^GSz+w{N(9Lyv~q5w+yHi$DW6S>LOk% zbH&3_W6}mZ69?62B`Ara2W9{r3b^4Pp~J55wupG9(|5J3&uyRd*yP&qB{F25-YmwD zWG0=jv>PIH?xk>GDt_T9!@oQBUy8pPzS9>oV34xp2D101@thm15LiVik5Z(9ioGt| zCp;0h0^GIu@OO`gu8KK43v!TjqO93xH>42vPRol^#XIr!JJkZOMV3Oamir%ts0>VBc-1BK{~W?QOFHnqmwb?Jy5 z63%AnIT~dPPWm9*1kOQNMv+fiNs9Wa6JdwF ze35X8=WQq^Ug6O%^-vOxS%Jnmr<7x4y9PaU2Zwpk6=So>&Wu%l^G{Q4=^{7_3Om#g zSZtn21L0Ktrvoo3krD##>nd=eQwyf-A>jj}wSgOGVT?>%Q@RI$VkGTq-UkmIOKob6-`_`Z?9v;CxC&=eByLfo- zQ{IOJX^@jTF1;2BVPm>NQ0!~ox0K)P+o9-;KwVISI9Ymrg!0UT>I!s-59Z}z<*%@Jtvs=o9BJr=lMS0=c&hjW&X@l zmB}vsQct@NP=G#eTCm7 z#$x^|)k)vsyIk!uIgAtQiFNWhBb-d0b40knT8(_Po|gNDH$gCXg@ycK zT$mnsIo-}Uh~-h}b~c_FVR`bb-Yo@@W}igzH_c8`G=-e8Ff7L@8A#e4regoq0Yu8v zWEBS`Fk5_k*-KNEwd`6bB7yWZ>z%39AVdpRLMbj>r%aXX(L-l-{W^e7N-%)fzJ`(WQMK`!LcpAoxZi~eKAYNJ#o&O zDC?OU@5C{hDEBow!XK~vr*(m^+lF&&2#w^Z_Otrt-jrgJuvG*J$+&q2?j^s1sB(G6 z)Im-H2PAe{Mf((*uBaNHe$kM}W?m(bg1{VsgPXuHYE?BUv2j zv2SMt<4L<8H5vZfY;PtH{>Kbrm_k-s>~J4avL1@uqhhfO?^)Fop~-0+@WglK zIYJ5rmLMX37xV}sn#2R%nF1^nS9Q$_qlugIu%I} zt=kh>6Ix4j^oO9Mh(jycIHDhUgL9>k0z~+-vKRvH8?U)fPFMB zd9LMIR--s!daVuAc+Q{AHM8k;Tgt*OnjU|kGUKix_U$dCWD04tSm$ae*>Q>-Ll#3W z)Ls{n>p_67GS7VWwwXFQjyL3iJqEOjWYsPIjWabxeofpgvYCTjps|sx1_QYFJ&!3+ zc}s&=Q)i?|*LocwWz%Zl2ssEEERZr$zfj^XiQs5d=S0PN?4CYe+u=eGJp+Uz2aG`E zzS%&-`>(FGPG^nYn7i-q56rWXTgJ_lY!gM|sn}v!f-IT6Dd=P9^wcQX67ue}cyPDd zCtyQ33QYDPc+_n!zhE1|Kl|ekeqjxY&q(sy23%Hi(nRm_V88W1iXB%*&VO( za%2wegUdLmsgC?oa1VFYJ^oDda^oRT9V9uQ8Riubtww~M4p1_XV=Ef&YTLsZJN?8(S~EtPpll zAXD7}t}zsXOIPr)s{4?)Ub5e-OQmTOH}kv2Dbf{?@C2pczPTCf-sn;Zl5S8AU4s`^ zgyHEPsN!&P_S`(3&Opm}kulc&6^EW|wC;w@1TAjtsgzq-nPP0$W9;5K1zhwy;2D8b-E??(sKBwW1Q2ZP1M~G(M<| zHObDo)>f?jck2WK6R5av{E67egLFV1B#!Ek{{10zTR+b|DL6JOBltXbH3wyrfXpl0 zCP3XGtiF9r4g{mV?7>%7Eirum>}zj2d%3ZmLx)jf{U9fweZnidiaHP!vt)iXalVT=(le4~qh+TQ2P zKL4Eo>#VemM6aofqcPgiDZK?$@Pm%F5EyGoxv5qz0F zyvqskSV301F`)$t??&ufQYhJGifqD0Dsc;S^f9mEm%#(o(T$>;UPbI`x*yyLjLNdm>FOv6-in1%((d3`|}G(F|*yT8!s^DEKpQU$$%hy6zh(1 zpd6?&{CezGA~xJol(#Nla2B*&hCJG!IJ^u>60wsprmrxHwhToIP>`T- zaYf)==<$V-HmxE}Jply=j$&baGyVL#ZCjr;8_piHj+eY>$>+3?Y&yq(?Ued2eEiGH zi*J9e@%42;rC7yDk-q)gy6Dw2*9SZSh3M*-OZ@tngX~8vgu^PRc14nR3bV+==%I`_?kuUl_#+!if1wlej?N94XS+J!B) zy6HUNP=c})=CQAO6-H@td95Pd%d5n$Lc-{2Vu+2=4{*^HM%bAAlhBsf7riH!p5F;k z^Bi70`>MQweiYN^{gK?2D)P1`&Mr*YzTIW7Fv0{yQ-A8V7W8midtpOqq>2lDszcYz z%K=Ud>Z+a$?PWiP!u3XR9bHck%CU%jKnz?aS{u{?1x8xMv6p+<86h~fE~G{1Vmo3N zP@b(KBM)4L5#xiI0Y%M9QoiU5%N@>^@H@W7-UF4&tNbC4VgA(}-wi&XON2bn|65oP z9E`9rx%Jmq1HNqD44?Y>l@CenSYj`39A5`*ml1)J7E0Dck+Ycay)H!kG-G@n_Ap=R z@IUy4KK8z-^YT;Ea@l3R3A1ZL_r-urynjm+cIMj3F7VyTZi2n=rl8^gg9Mvc)GWs8 zvNXTjvujB^H$l*7zVDN}um`rOM@ojWTBKLw{U_lrz50F5P49N7!hi3z4*$3%c*g)c zj<gj zQA&1*A_u70%t$?u@--N4)4@g=7;AlsTmEgcD!kei`;{dDsIRwfCU&{k(feao%grw* zO9s4evUmFD2^z&2^dL48)It9MjLa5;#lq<9d#ttyQ_QwgijA_dd!2JI$pCA1EH}m> zPO(n!{_yoVKEpBnt7IujVul{~&0+z2Z6tcKC>f+R)2Y~2(V1varPI-<+L;if(ea)v z-0!t5#t^kp)+*9QCa^SD_*a0pjr0?ZiM&)1m?1baP7$!@PnW|OdH$ovHRA2QHFJm= zBPAZ0XUJMN&IjgL%=$J;mQ0Z(6n3wK@A}c;YSt>&rl6V`SNVUamz?sb6c;2F}2tPFa(6sc(o9gpwJ@lI}oj3ARRcWLme1MJ%tCbIO zKDo`*%c4x!GT$|rV}!1;*Bhsx$@(E*|HIyJYqE*k?rv|CH1)Ga+tFD!5@`LYgTV zungX#ViU=3=&~0VR>{}VHOdN3x|gOeuoTM61~~)lC(!Il&(dJ!BNjuJMkhs^w8f#l zmZDREfAR#kAVDS^8_vMa+?$bde9)0bV#oUHPd%P;Mhwu;GZ-jni?g>RuDrgLJZKt@392X)J$hWy?u)!k)u_0LZ{Jg&@05aoF2OM4Mr5R zK&2x^icjB5-4&NcWwAA-v!Rc)Pa-E}>iG6gc7uqaw*tfM2FzFk4Xf^Dyy&czW1(rH zw2%zoVPibDbRlwOwF)3!PcZkIfvFhhtc^93pLUJeMyab~?^)}v4{p16w$aC-$rp6h z#r;Hswn3{nEh`R52*2-_CemaGkf?B2aev+hX%En2HCfY51llB@N+WstAiIno4*)W< z;Y@9pP=_NSJb|^*?+!zlyZ_PI-ZBOM!S*i)qMdKN@!n#9#C?B$Rb)E16u z0Y{C73(bYezw@@rCq!jF6o(7IeW1ltI6zfc_yyI>K=mWpvEb!~_o3dlm<0+OFGdn~-&-??%09 zj*=A2U_BsJQ^+-o@IoUc18LF*D)w$fm1OlSj4WssP2m-RrP0`2Qw9%(K&4zbvf=E7h}r}QQ}Zeh8+Jpxn-kB4sb>t%10Erb6U z!*c{8m>)90HuHy%=1eEs-*O{H>|EYE8lg|A*=l?AtoGK@hr&E1IO(;r4;t90}eZ?t}z zl_7`1P)w&JB#~%(p&3nzw1a<;y(PSJVShwh#6G$hszI1HpUhvL>-2oWEL$3(m2TdjJs_~cI&4vb{ z_*=vU^s+T6(p=sNFvkXXL+mN>iS52UH)SdsRt9MLG(+g@*O=7;Oj0%fE^a(-vKt#D z8*Z`zqW@f0yzn)GRT<(Kw6@<6Ga@|K}f#hn(xi z1SuQj+~RLs7cNcsCN`^R_C7m1&F?JPJZ&8-j#tS)FS|MaHfa|nE=W~WcolK*e9b(q zPZOAjYoWWLy;U7OJhq*?HQ*qc8OeEBJ$DZ%$)WClbNF&jh1YgbP6%2ury-09dh9VZ zjA3HqhW|`&pKndhb=#1y4Z3Z;A`>LlcPKlhcbEKDt2i&ug&qQ#o|}2=LbDew=hf3y zk{0!4*>b;>CD(-)`CXEFWn08~wbSO-#zza69h-IA&vVu*jJ$MeLEYVkAhX|-KlJ@_ zlIg~NPo>4iZ$Bj~phzAS+bi7&>M7^gor=q$UEBkb`(O~drI$oEqJZX~O-g4!CR$N1 zw5C8uW%bMk2Luv?tl-^;>c)M-qpAd!r3SmRCdN_7SeXpx{W{N{ z#0WBzvK#*2Thel}JlPm##>%a4+$kZs%*aq~?C926 zAnPzCE2l^)w7v67;-qn_=&rC%X<6_9I}OxFi`e^TwnYrh>xE74Ir4y0&AJwYuRWC{ z#heV)(T_ltp(||o%FX#tLBq6&jU)T!rHS=wZ0=g(pEdnPBnAN1OVh_I17){1{C)gRZ zj#Wyp8>VwZ-0qmFhjfES+~TCNLW}twed5TBkA8~tuQ0MLlh5Az-ZAScBe&fK*{B;r z55z!Yz-_>%lHY;V@${A1CG)!^=b$zoI4rVEU{Y2DCI_|C*i&57Ms%}Mpn=-Q^jU&` zV=7`W;$ehVu_H1`;3B%EBY3eg49*7&JdqJzCY{{%)3?5BwkHDlZzhpOH?}8tE$qn+ zN_LeZmvKvkq2Lth33iDrgT`jOSn(`J(&GtsH;q*x1K!t#I=I|Vy7)OfC|e!yJ{ejP z0?sT3V23=gegwBjIdml?)SgNXc@~gPvKJ&6a-dkGQ)(huo`8|V{jrRWF=7<^PvcGN>O3}WAV|9h|e z-VCvI>k}4`Zf5R|8*iaDSmY)KDcJx;9#gR!d``lSsGHv#eLOUQrDHXL{xd)lb*N0~ z*(PxpparYuv6B9Vuw0%dstS4Mr3%@QN0S&ieN`ca{HKzX$Rc@4z-@X>(0cAZVTWXM zxF(U81KHRVX-edlsaL)Fy$=fvf$6~Q%Am1d#F}tTZ}i~>CD2CV2xwR#i`4W4Ksx4y zQAM=?A_a(W(?gFc2Yl~yH9J`kgz&3+rTO6fTE%U;B*659Z2`+)M!sQHhz5JOK`uE6 zf4w|GHLKLONTO*G*M#C@fARkpTAzpMxHk&($CGhzV|Uv|#-W|NT7F+r6><RWGfgH>{m#Zbe=7x)s$a-EYr&HNF{R zfD)^({P^DYd3U$H`j+B<7fQg}F};q-YoG^3W~)pC zb0F@6yLRYU?E0;mqbPaaERXkf=}>%+b^hsvBKg_KNugSMysv@o@CUg};4W;81aiZG z&wG4qADbdY1{zZS?m@GmROOk#2SEBP_R6OwV1aQA3FDpT=ihJJJ1*H`pkMUhO~E&A zKvyBX`k1i!<&R0GqMKg{w40s2nO@y=ChIg=4>6rpb8_UX<~$Beo3QqojIr?iX>`g6 z+gKQbMM&lU_LaW~%{wK@zg}%2n?5r;B}iV5kSEKaWIHI5O2wwID-_MZgxX8vqN1a> zhx8DQfqQ&9zOatM&5CYTB|Cdj=bVbbIG}W`n+VI|Sn|>-fO*bCnx~Ru*2;!9lqIo# z49R3k(HZnsPOGSn)i2Uui-i@uT;Y(1i}Q4&{v8ayp`-3~R1jZ*4=zFA7I4H@LLsPBI7^c&~ZOPBUa2RRp6J#=|Yfp2$6s_L4clulOZ)q|Xa zY~(>zPpgNP<8|~!<$!lQXMo-%YEkd}>7B*r)lWULrt4UEe)R`;mfZR;ypqd0MXEWO z!d;?k!Y*-(y8D-j)b=lbyrk+&<7V72_%2xObe?xQzHf>)_rsOm=Jn_BYhP<78(uJ- z=3I+aCY_RPr{J@Wz05i#IZUOIRB{nycCSgVgl2M%%wHFp3Zn4C*FnKWv`;hvZmW5U z7-C?26U*Rk+xGBpuWVAw%;@m`_|avujhT1i#?g`k7P!$-GFVfyz&n|Hm5)RSeNJQIb@z~CP+D8bFMIU@Nrz@MraST}*KULuV#v~XFu(-U zp|?4@Z(Fw{8od{G<0Z#Nqvnd3CfS8K=|Ls_YdC|!OKW0k*-Q26)LA9ciqKV@qn;)H zX9!e2##DsgQ`EAX>5D8q$Q|}@YS|_JPocSLJz2*~SePVerVj{QWogER2}B6f$+mO6 zZM#yn&o9?{*E1R|?(2FH8+#j*fJ1MzCOxRf7(6qz^j;I*IiqX_ltLOy?&!i&sDumR*mJCxbgb3%VK>=resMJNuXl0 z$Yu`i$e@pVn5IRcqZ2uIrj^gaO!6)NiLO>-0t*)*V9dwbhDFY=dw*r^uG;VbGg+Ci zDmRL&IGenBX6WdBycPu}?lB3H!`?xTgp>n+)m2pJG5=jmIrU>d#06Jvf8b-ptlZZu z*F=9qWG&q?dRcK}C&Wg#-gThfqE}~1hxp~}WYr)CL`{8~*gHgLh5N&9g-yf)G7@uk zTaA%7O~A2j5T{oy{>_Wts0wR1pw9Jh(e*`NAp3y|kP)278(Nk~cH{8l(|7xHM5NAunj=UjR?nym&y37bbS@@wt(X%!t5Y222nq@4L-)Ohs0-FPvxq3^vuWHYIvy8{jd9t+$S zT@t--+FrVyl+t)Uz{m4W);YPU7yz$dA-G#Yisw*pJGXaUAwP%iib!A`qW91Z3l0T# zsfVA4&4c58PrlS7>mr5xT45I%lpRwJ-vjweE z*iEJu&@vih?-_Hu`7rdY&99qn+U3*{I3fKmAhCO^d(qk`=c`!4nJ!Z2P1$_Ikocj3Vo1X zjg9KlSS9{i)$Q5C1}Oo$r!WE1#2|?I0&IH?{UBEv@EbF>UMV;kKt31?Tkg9@AnPnv z#Q{q8m?C{ttmdB`()`7@NRN1nulDQveYJ{RGD9Gq4S?o*5j&IBB{{sHo!;X6SUluW zC3-4Jm-fsUQWiuZD)#$o;up3>q)6ZD0A6g?^o)>1&f6W*5_sTU@o5lA`d5c^Eyt9f z#qEEqD36D$?J*!${G&r^kd%Smr%}WN?@kLjytsmmG~Y}CP%Jpf5NO`AByQodE1}zF zw()OD@hzYDf^;X!uvLaP0P*vQFD7_v2`Eg2tRt0l2Z+(*(9B5e(uIQ!+=@V?Dc6!3 z8geCoFR|2kZknt@x(aW=-Z@BT~;V*_H~)8(YQgue7Umg4%?s@%d1b4 zyb|=H*XC-2Qo#)}Ht>UtR1g{s&^$UCd{0=-f{sa?EKvObalPfd%J3xiVS1;(7P__3 zU6Mpkkb1)+sl(tk;X_=;`xrM?T$FiWgq2Cf-~QKC%Dkct`V5{WPd&zflpC*Tz^fWT zv`V048z{1lij8Nl4uKRV7M>iRKe2Jnk>6-LKr(XpgdN)k$-{=;QtNOLlNtdxcC2+4 zP)Vg^uzGHRtas$)MQ2q=q$!rJ3yqU@O8!zFlO>2JaIkYu(wwaUT>=Alki*2a9p6OS zftryX#rWQg!QD&R7JOu#x%`Y^!$wYXE4yrVpXlIhgD8W3OuD6aq{l+nKxUl8v&x^ijnC-RP;Ch7t{1+1+xy-p0@iP-9^M$|xCBsultj z20C$K|L}IMj&AbQDl#LhB%0GuiwaGVcslQL->ssWP^|(JMiA$!R6dNr(N{=i_$?O3 zS#@+XAH>OSu`nZqqxC_MP}0$HzVUM#X(v~WFfbEiEG3qgoM|x_!DiBT&&_F^Ylh7q zD%8~^-;H6@U;&$Blnf}EY+b?P%1=UD;GbM zJeu1xL$BT@NEF1&jl=1Sq(GR@-Zs-9NfWIN(W}pp^nh;Ao5Abb!f{LjE0J@Nd?>j+ zx56typ5w1oRL3Ma&6RZ~e2idn5{4MThpMjCPJP)7pEvgX{8wa$8^h;_1$^=;87xwJ zs8}pITfwV{c`Sj1b0<_TVwko~MUf+?48CP{+(0a@NhW*9&j23g&AUcy}SkjH(3 z2>@(wT^b!XPY18MEUw0sCuV70@XSf@uzn09xJ=sdLEWh^GhBv6??%(H+&HjiL)s4I zQns@WD{n-X%t{k=1|u6HPk`+QG<&ADPu&PgRk(l+dBAp=-%bw*H%8{F>OwRI5Dvmt z1=-NX@hL|;`<5^hBo=zO1=BRStbUR)qmq9FL}Lt~7*i`jQ3j`Njhtt%xvl-WZ|NE8w3 z_ih5&j{fL6L1EN-XgQ$S@71MR%}Eoz%R>T93TLMpoFuKPnibE^6Bs+coeVe$3gMcB zpaiIAdGpP^KF#!QpJG;rf2-*7BBL@9=zuKQ`2~CtH()!UQGtWIFK6Rxv_W&sr2XR! zuNtKpMYA&gl|=S1Ly;RtX=*J{R8Gl2+jl<|i$#wQxE0|a19Ju$FFm9_=I*@x$^*g! z0=+st=$yFXrHlSgJ&OG=`kx}3Ne`!-EG< z?dUIuuQ|;o8Eu9Wf#awR8uPw0B@MyfGy~`NUBACc-kU-`u=vEUP_j!Dxj@A}`CPX& zg}stj1Yg8s+(P~_ZVvCJI-Xs~fAYCL(dDqG9{6)Zfa&Mv&zpXh(p~BtUOXFLJ0>?> zJI#OUkq><%J`}d7HwEpX`$?YQsH#T!Kzd!eDQNX9kcSu7k$lk;L8AoUPp@v299z&W zO^VU0O}|A#_-fyalBW{WQ|jqjPDjKB7M8Z+`)syfsMm?}GS+u_)U`1uVT_Bm7mP82 zV($Cx6~BJ*$JUaaHk2`!EyI>JD|i{e1VP86Q`#UZ2`lp36=YI!$`DlYHwCpQ450d6 zGXo?s&F3H(Hx(I5&}2rHIek}9FS~47-wdsb)OwtNmoa90{FfNv#eF@|Tiogc))~ps zp0xYoo!H1bo(%1WjFMK7P2VCZ`Vayenr$nKd@tHc)H>2Tk%lkJUC`<6w*haFRr?{nI?N8}qaG=NFJ-HxAgGv+%E~DH*gwISO6h z=ykkDq#A65>1T`j*So0dSKGeO{pI*q?kxGWUcL6^9=gCclTMFZJLUK5rkq#n)h)j1 zLD@imf{nJc8A#vTNj;qe6o!kCoGg>n%CE4JfC-!p<=V)?EMiwd(D^Z0J0+X`NL(ga zH>E}W`*BNtj6aGq%1ts~a0oe{+{6e&)UvlW9riL~=>Ww5op>&yDq}_$!t73VrFul-RANG+i>}? z9J9ejnOu!t4ooh0;;j2q^XoHkB8;M6Sv?Pr_p{0xO~}>iv0wxG+N?C}%Gi@Xm~Ml^wAkFSeEv*JKKq;ic*IwO`y1 zcVjOl&tf&%LCI2a+ZEeJc7jGbN~7kpfQ4Ndo*ZOq#GNu#(;z`DAy9^&zaw%556ej= z$gD7jhsC$vWhac_G3kdlejj7a&vsjLVuOo~#sSwEh|!R$P{Tw4z>U>X8Ze!TR#8P%0|z$0 zH`x!oHwLW?s|@d*wkjYC1eQ~zSl2(~fjL9WFy_FPrrF!nD>_Ar-2Oz47Rtkhb!FP5 zxx!&BGUWW5x|j^V8ZS-|?BhYnq9QN2Q+QNa4Bbr}wnfhelg&lPt`{7I9wRVKnzK9Z zj&+jSZLN!qg!GLVRDvk=KI;vYKUnv3P*Cv_6n;P-8u5U1IOY7$R#~rPzcSPF7SWti zCJ44KF!&64oK-y$Zt&6I)pvt0L5tAj1(ndiCV_=DhNV&9+zj|M8S8B0!ty{w)?v6< zOs5O5WAS`PPZ(ima(Vmuuou0#Y&-Y7_#&_lussY`I)60#-LNWNC3~lTyjLTHVDqMJ zV6}4(c&(Y53kfIez;`X?_UsS5&qK!;ohV6gGIXOXk<(2dl+?1X`DcbTgRHZ2Fk&@_ z&VY>7Bb{=}2r`toZ%44TGsC2Q#Es)DHe4TU8enc4YwA7h5_CLwD(!&JbD!ag0TJ21 z&bQXA`iua=hMJWDa+1jYHx{qg(dge`wJOTL=Yib&1jEA)TsSA{$`Cw1Kl!x}$3uW} z<7A)>0u<{0bVW3>Zc5X*xvaY(O{|lwdgXmtCkXv?Np^|!>UiH0)-~3ts1GAbqcgm+ zrq?JRv66#oLQjz%VRF#X;9Zadzsm0fr3M_`=)FhPO%K1)AJGcvGT4$;9 z0Vt`VCdl|9XFrPUa(#yfF~rR1;Xhn-tz?OLt2FP;_uEN@8*i0bEw)N$C>b!o>!{cU zAa>_3?f|8h6|xLLRnC;v$3$BgNqqBs2Gg_m{ut%rWEX z@T|YCCHp=zgt~Vv(07!Q9iqqqDz*U72TCsgd#)=w;&TxGzD6*a!>f z39xT*C7mHiSL|Rzk|4O7ugPTX=k4d^2{P#S6o*xviXo43`6{v2$5GS5bSfCbgXtSM zIQ|&<(QIpVA13f{W0%K<7ElfkYv~gNnXFY%(AEvqmv$e$`i2mz3o=<95knE((8stt z1S{-80G5>yUL`U>sZtBHGKv)ekmX@0o0DHrf~WN_4X@ zN{TJU)6}PB8ofdb5m$`HJ1d%H=daDA$CTZVbjm$Ivp;P^F(E5p0!Hj>2FLcN5GzgbqZ#^8%)b*F1jKO?v}u`MRT8W(OG3b1SZ|fdi5iSf3(vW2SIU$k0HZj;I0m> zgYMy|{qgZPF}m61ON4Ia`Ad2s>ePk<%xLmru| zeCXtZe9-N{GBH=#!^NH_gwO(Iy@6hT?l}mqb%Y?+H4kj037( zJ)j<6ERAmGCPi!Ra8CKwD>6AaAct;~#(}glzI=e}HV(n3xo@w5-Z!3Zv+m~Tw(gFN z{*3Kh48Xqc-=fZdJ^cfEWypDT6=Elney;`~EKmWpBr{rL@n*-~Y3x3-Z4W|qo*Lj<5?IIjJ2!#c0`{aYPv=ABVveejc4BdfBgV#hS7mwb z9IQQkIJx&luW&Axw*b2dTXYP0jHo-j17+@}9|vOVSGk&e!)e$T(V2n3Pd~DCEfAauUsWt!OO)M;a(pQ388Dax9H_+7r~7g zv@A`AD}CYf2P*Er=mDB-tmQl?`A|UfAKdn$rkqyM6kZW%kfbiubjm(v)y%jG4yr+P zlf5~_%$gGvg0_9NvCW-@Z0tV-{2eD)YIpJFY>$8rq@dson+mgkhRDm zOEF$h*m1vY+@!x9xbgXN2yVQr*pNfO*zSGL#}eRW-}h|iJ`i?^iy+*8lYJc)jb;0k zg;CIXDsJk;m=l~KrqHlsR-H1$HZ*i`H-d@Te3W=(o&h}(C=DiFe~SI(Z(@J@=b!v) z>7RsaDA`Ji#JQ&In!&%1*hnYdjc6xEC7vEm9PcqXB+O?e2&yEN&>&DwdM%Zj6j_i$!EFB?GAJqGCJzb6J_3 zJrGrgt*Lh-_c6IWTazBy7QR}3O}Ji~F71lyWh2+|pkSvz>{(|$@!k{J%RT_Zw2GpI z`@J@Th_WlO!vrd$d`{P(#Q0QZPzlKY!AY(eDoZYIJ3-!YW2o%4fJz!A+eVRODt4bT zU)15(skkQ8z6|^-XmhE6$ZET>%x8yazgM~!Ts=M?|F{wmBj=PcI7Z&=!eeIO_*Wp? z$J+T|LQ8n@+!BNm()9H52hjhpCNLwUhwcIrLN8?TI;D7Zg2_)zfaF5ljC!;nvoF_r zZ_5N4ZX9CWXJKV_QnECPY@=e2vvvo=QUvus8jOc%6^-Kio|@G&S3z90(Q~&@(<{kk z)q~!(R#6`INP5+4LTn8adHY<0g%~x*|FQRSg_nW8CEgKm3KGTG4Z_3+A09E} zf&F(Y`5p9~u=L0aBpp=14AMH%qG*9gyy^dO0Lc+}48!F4KOM&nvvo*6l#x!Gos>u4 z2|7(yFpIjnv9ZatFgBYh*(Qp_LkID*IsaJq=k}N0l$7#~CGIHFgIzsRqzU19z9`nCLEoj{`!GmoU;;?fYIJ6DgA4g+n~FoAtMhwc zxCZF#Mb$8Tjn7qn8^1#g^j@9ut*~67l6{(N3M%ox8`AF6#5Ok0W8OW_PAEO!@Ypee z399!0wXM{;P|!w&9e%E;KLTlv3dJ5s;CItq5lv#`o3(%hx{gkr(FGd%Pau7tGIbn= zo5`JKEbZ=surDJP{q;Bssz<(jHxFFJ@AH}ggl6N?BjLOctexoCVK~@YH!QoqE>Yj zR4OoTwk3Q^2pIeDLg*P)8Ju#Ab{&eEs!nWb+j0xHB^0wu_QwP-F*Sj4q~|J&(<+<#e)Ie3Jy41VInI zmQ%^ci#4IGqO1JmAWf_29)E*0-nY~b$5nUv%TlF+(0BM;5WK z@`3-xdPMJ6Hb!DYGqBDrtQ4t9KNox6q_Q4MI;A=`4lvrp993)ZiV&2MN|D~@wu*XU z1|!yUfLXx(! z|L1okh1n*|jqwK!&PI?Mawr)nx@S;58hrNPa z7wVv4OeYIEL+I%A8SOssV;~TIBnS12+PM{8I%O{G%(}u71REn=PmOUDP~!#6^X^|c z7Z(of{hm8--#k$N(%LnbBKfz!uaBu>_+5ibOT%*qvY~ zYC=u1USoJ6j<=R~OMFK<5wNfe2G;JPtLA>hb8W9RJu}|Qw9X;9EnCh;R;V_zHgaq5 zc9E&MChTk2X?~ULLgM<{=Ggmq$Dyw+VuLkr{2CPVh#}do;%)Ap2?lH)+WYn*Bu*1 z1DUv;QY|PzmQ6E9%dn4(j_#5qdTMZAoI|%mQsF(tkg_0Z$ODrcozjA+JVExN9QqO; z6AsYDtWDSpCtAfR(Dx~ifp|1V=%4@qvrW}>KVLJzz8lg*j!)MNb7;ni9p;#LuDvi1 z7#}K_EmCzSLw?0FFH*mI_fb7r>Bbb@EQ@7l3nfdW$VMu5SO60nUw+_SF1Z4t#hApJ zfDFW_sj?FlqwaI{1!k!D{C9r3)HUHiJV66k(gD!AIvLNmMhahHLD*~ zW;V%lBCqkVfb*O^6O&^qFmM}#x_s4AaYKLm}qrgsT6+!!`E_$O3vnD9pbSxB?mo}9KGdmsrH@;X& zCz4f>kV40JJETH;B@ZJ|T*=fgu{63LI8ZfrDmND!=&=^=KK;Fs=)Z8%b-H~cHCykbJ61oUK~H zOXnL~o1vHn+CQ^UmnqxVByNW`A1RS>^DF5rL)~6~|gZfb@Vwmqh_Zg>P>=U%vI${hw+Zg-z)Ju8O z%>K)u?DTq~bz}bpcF7|Gu!WQissZ*=v7OR#$$ihfncaTJ{ky@}MgQgMYrAMoy0m9T zjbI12BGA-oRddK&FIgokjLM)Z0=vcQK|B}Pm%ysYVs}G_yb?tnNf!4-yM`R-6mT$} z`sd$Z8ysUqWhT`P?pW}q88DasZS_3T$1E!2zDqILY$05@YO-wkWb%}XT^+tN9RIYa zyNv&zSHljQtck3n3st4^%hIPF)tplK!|10TI0}Cjs`TbpaD?e~yt*zF$7h7-)vLsK z|9x&rSl`^Ipy^ZWUlR7zzcO-VsiKjezDrem?9=96(|7PqN39@Y&Wv|fOk z64)Z6H7SioEtV87ohU3TYTs#dX`8u(J#lAKqx zuMRGL>apv`tG<>0s_E=B?*y+@+4#SYfz_x!u{PYi2(3@)*-r|XIn-{v*FS9mmkLS- z_&G?$=CZOEJ&h|2L z;SZuU-SkyxL#?A9K|u!Idv8$%2TJa++-~?X-ZSKJ$G3CZDM*K4YQx~12^ z*^$Gha>R|5s+x*mGk$(oRJ)1nW`-X(rrlIq;HQ+5L5W2H72BfNK@E5(fD(Kj4?OKI zx{|w5hAeGXhDOSJGIx-3~5&j!F;Wi0H;cnIO3A z(r)P`(J9XryvvJFV&2J%uJcecV#=LxIO~sTvtKo%reFr^0jZiopeAZWc%_k&LB_Fx zioHm5tas+Flp!@3k{%h1- zs#;+hXdL1@4zPRKyMpRy({JL?EIDs3o_&Ew9XK*IpAo;O;@kgxz%fH=R_eFw$(kvk zQ92S*TPfKViX>98$VY+j|7lVXWlR|~h~gF+XsoM8=E{UO<6{C0M<&?Odl&)3ef@y$ zqY*9EtZ}zxS=iu#XOYbu#AmB0R|vVQ6zKtibc-PmY?9L|dc=ofq6Ek|fU>h-E$=wN z?j}xOOvb_y6YPFst;TwibFVSN?xduri-@)DV6ta&V<*Oj3Cs|zoSmdL6%;`8B;(;XA(nt<^n2=0zoc==Xb>ISEgq1jNFPYkM9C2U(Ns?xly^lz^87~3a_9z)+nFJcE1?5EIBvI3F(_BH ziY_nGtPJa+wLz)!x=?6+6q+2AquLmgzEHC-GC62pB&u6D@WhxCkAHJ(WkwK zBd^)EFT2-<{NqJ$5rDEWts+b(B(RJHuyr7u+34A}a0PE?NWBtPAe4TXkOjs(ISg^& z$do#Ck8K=`mA(A_B7M;$vpG@iKKNHM=;5|;L{RG<={#?wWbaU9EfovGQ^44Jd6n2m z(gfnS)4%ifOoTmnu;)3>9>WL^lV57?{DHM=%Ea$*<7~DKM>a>bl4Xi}q)u-z(t>sL z!P%GD+x_lHHw9e-Vc%1vmt7|;Vn3MOJ435DK&rSC0x)|3Wxc968$4q@&u{6EUw_dV zz6oEtE=+>HLT?=s9*t>M?ooAdZ%cLyFSC2RCuD9%ey;Y;^T>N9;+T=)pklc<9*)Ow zxbcR?hT*{6!b8}QV{^+pgz;;GQ-=16i z@fIBpvTT|v`(}9p zrq${#7@3VpgG)W%wo^As8`Kg$AT0@N5Wz+ORC6-IKxqK`zU2toh^Y@;uW(R!6tqmV zpj(5HgECmJCt^=v-&|OH-|)ohlXLzlv$=3v$N$goZ~eyC%$t-_zb8K-b<8q)ZtRpA zEVeWkDOn3enyA>*Uqp3{3a`X3;LndB`KwnKNm>+`E;z9vD~#tP2d$p+j(Y3DF4dYTm@Y1p)XI?@u7}>hAA|POdC2+gVc+-P z8+kMs>*1NfZ@~>B0$fmKPX>N=k2%e2dQD#uG9TZ8WV%xbz3$6+D>SXUz zyi~<4*s>I{kF!t*e!p^!H<|u~-R-&q z+lKPD`!hWJ%m`_nbMX%(*^McYB^GGNresjysHH%V1v`Tu<5mLMu^S?+@A7oeA3PhV zkR8%Q&*P*Jtin5_Z1zp>Twz6EqGzMHj!q423>vH6GGl7zApvXFtTDmdYTRkl`w3zjd=>OtXBfr>;K6m%cb0h~h zB}SM5krkA^l@v1bS={$uM&7f)btNSO^6w!mD=^Sl9N#K3mdB$W=`I%5brT8!U3Ick5NtcXqc~1ITT6%Qhf!@73$V zzjsKFEq(}urwqD4rk7W+KUCf#b@UFZ;VX?lK%ubn3wHkJUn<{d;}1yNxrV5N3;T(V zK1cWX<YT`kwE-Xah23BBtU%Dz&*U+tpwJV z(4Ls_1H$em63;<^)yocn!15+WrjaTT&s18gow_Z6Z0MJoia9pPPfbez%Jem9MR)@1 zxTqrhG%KDnKyQ&Za_g1LReJTu1dj{-mV0*74N(=r6<&qxW`=td%g@@LGlxxwJr5tp z7-l&6*!O44er;Z<{;_G@9CB-{J&+r(M{6v~6#FUJLyGiJu^0$Ct9l~rf^d*#3zV7c zq1TXqRs?EQd!R%*k!Y}6bPilAT3iv>w0MKhNl=>yA*uHj=YVgZR~E-yA<04A^g6Pf zhs%$qMp?mGc2rp+E2TI1m;z29bGi5y!RI|9dfAq#poLk*t)q+BRiJrj;##Lj;}>4S zW+$*`_cs_-ceAD!L>Xn26FFJa+qu~wlV^Nq)Zg6$dH_dN22r+ZgU`pndC&$mipu3k z=mu5<*heUdS?j9{=>%QoWOgP8uRIh&-JMF~wxx--;rPlr@vEL@=+%9$>TR;hjW<`I zs61k;olMD+D3U zQtCqB=k1rV%Vs&y6d#6A2 zI*Tdr?UYn=GKDSb_Qm*$tLMkw>9BGYTpk%_G#&06{pyI@vUHi*e3VW*DJ5 zb6jn)rah))eH3{>#SS-4c}Hy!mZ~4kF$AVaQ%PNDmUpK22N4j_jmnJzf)px}bxK?P zp%~}_xhpmal$o!-&*|~k)Jl5jJ%Y87Mk-Z1_XN9>)g?K+082wIiK;mVML?w)@HVv% z?3A8hCvvXx3qdyxx@?(hi?EJ1Z`uavf89eL=NIzxS&g7>qgBL(wVNJ_IubZEM`aqA z>AiXu>cN^CQ{pXJg|SS*wgN2>@+!-qWD>0W3w&?H_u@Hyjm)8Q}BsBe_>E}%J- z|NL96);+5~BQ%+Ei)HRvhwFL?Dt?2wudEoNg;>&ZPJn(7p5L06r8*Y1L0U!ki$KzW zZkb=eHbm*wTHYo8Hr7MQ)0x|5XL%o);Zk!u9;i5Bg|qpH;<;b%wQfs1TA;{{P0=0; zKy0UEDHPcZttEW>0?#N)r=gH)(vlrZjDj<;C)Yfb^hy&z0=y=$AnMH#|G9Mb;p2B0 z05+%2IWO2=Z12DG6_&So{W|>G*P6)&X6u(5@5!JJdW5r_PRX`YFh8+tf-aC&QNhfs zd=1Ks=pgMFH!lZRca?M#K<3EHT18THBYlMKi*_|!o=q-O+&sI{#Rpfvyk@G}j1EKO z%W34O8*h#-S)#ilT<8jh70K|RW(aT_e1ZI1lDf0ZWaoU>D4DgYlZJf63NPi zRopJFfs4fblc8rwdVpRHn__?jj;o?CMHxih^kKR`8X}1Ep+H3Pgj+1g^gryK5S~Ka z`N`$)9SE+Y*G@^1?(}box(^%VHff`{3iiuaSno)(RJ9zP*SaZ=!-*+A7h z8DmdH=@K56Mw`*}-GlowNvRv7smTIOCn#AJMJhl|MA=982~2;Zz>^--LoTv<=qC%i z13r$*US!~&6Qr=aK-wjdToztpo#E@%6)#<0l*it`W?@@GQACw zbiXH|r^!ip1+Q06E2fWy9G{MV^y(hEWASxi9MvJsVpok{5!~_B>C@rlFnWe>WQL-E zAEy3-Z${Dj^c%k<8$UBE7D(rfSh03evNVcpqhdS!%Y=s)|wvDM8|cvRV<=!On{ z=V;h8XZh~+zboD-8;p1up+PNI&|iRnU^evm!{Tm?N0&m)ELN*Hz9QK>?{ONfcSLTRn6k0^x#Me;+CY+NZ}dl?ZPg8$cbYbF4(R-LO0i=_O4!W_X>Mbl z24<7(@Z2%Jk=V8M^=A~hXV(W`mu z$Tj1L3a_mJx$HjBUBTD-qCXUtfv(E{-5gjeJjO3t*yDf5Q?I@*ye9mR+vER1zz$_Q z_sHVE99ExOF*x`A{fUk4Hidj$ z)b7YCNgZ7VQsI@*DM_#1525wU$eZe=_o5Ch)T{4B6|(i}`~I1p&GdHHMqr;I&Q5bU z61>b>)F=G*935IkEM+!^CfOG$tSD*j)mZhQ7&AZ)Ie+Y;t8Ezaisj%3;7Ev;=8M#z!BUF%p94@fG<3NDyK3&7P zB|hxaDLFOmFxBl-EGT2&5O&jPA&+Q$eSmJ8ewa!C0o_vNc2T3ak-t{nqJIC|+rDU6 z{JVzNHCw3O#jZ)4+lzsbMq|{m3jkr;W{eeO`|3AC`>bVq+;(eYLx`tWx>1%HnGFsR zN`C|ENDr#sDN?)u8{7>(?XGPKZas%gfMI>FE6*7LV^V!$(a&FW4VGg6KG6#DfJTFo zs@S8t>75pmP45W8=JGofI%o-fgPxc&qTzXEf(_SKTMqwLOx@E*x$Zb;33?6R0t+{Cd4?QJ2PmyLS_MZ5>yf=Dv zz*hDjy5*NyD}n|&fBfWHiu8SYD?1m&QCG`u&ep3j1C$ck$SnmS)poBeL0<4mUZzoy zb@*mvMI8%4p7^`G?V=vK?8j)r4lgJR{@a^J7he-XS6Oy%v|hb6;DGPtMNfo%U+spL zBv(PQJArk9{7bLKACLG?gg5+$2k%<)<_apo-n@_l&>H32FuPzM9y9Lz;NSmZhSqmF z&Joh<#tv?h1to6fWLex~GDyW@p<=QBE>RAhDD4Ne-8K9xtgFhwh|7yoRmWZ#fX(mf zDfO&VlBbfLzH6pHPnszwzTdalbd)*0V*1wT;p-1x+x5*iPnlmTRPIoI^3pq`Qn7yF zC+9U`)tviYD|p2*!#8)!m&-DtOlkNW@4w;yP-2j5i!r^~J0wfcJ@?_W5vFH6jCjZ$ zzGwH#88b{{@9{3u1zv+}(-}UiDSU`u6KXogN^j(azkrPIv7Bc4p9AN;W=7%TFMau^ zq{fX=c*DYEHB+*46coh8<}dD$<{QhLHh*3>>#F>mppDw(i;Ou61 zakrD@e)Tkp#^AV0;3fC6v!>^&nw3cb6;T7yYFO^teVPEZw`F}mrhDqqyrAv#86kUA zC6awHItlK)Rh*NdyTPO-2)f1Dq(P+xX@H&nt>G2ZajZ_YRbUVeupQXT zPmHm$&p?eF#mM2N{u$$|{k$1eKf2wVPL6zLR!zMH1RE$>9Ytz^koUz#U@Q#IzVEf0 zTx9i=9e%gxVxaqR%v0$D4svQX$qn469#!-Jy_Y`7`T!bly{A~NGI0AN5<+pTfo=j` z&H(#?s@ykm!M2&%kz1z^u+Q=DLxG)Moy#g>?-OnE`AAe7xYs9PVdcx42|h87x7X*M zBA11{2|RZIa>LW?TIdbG3-h0`Kb@CjEpGKf#gA=BSYfYSlW@p=S*PrP0OUOYFnHX% z@POcf_fF+iA@odNoWM#5-_6dC?h%)Hj$?Z5|KJ?~g8ff-e57p~^!m!{pR?{8$s~g8 z#>@vBT_&5LLFB67hhCantUQ5^t_-haZ}IJ%(IkfIZRj*gHbT$QCNV^kjQu8hrOWO} z4I;d$UGagU5k~Ewn>|-kHmi`2CpYIeaC$!P0*PlwpfrZB?Koy^-1$xElXbLtfqL|v zpwnapvjxhHQx4FGZN#pBGbP(Zk$5T=A!PWS${w-Pje*v6$v#@Wj%Aha#rHKpxs%lo6Xmhw`pE4hm0gudS6{Ra6QBlXjbq3mw zo)tERLn~wdZ8KNEwhJGRBXc3^g*9%=63DHn6lsmJNqib~N3dr5n4)Z&Dg0POdZFa0 zj9(D7j#@`8TejDynKu1S4VG-t3c_7UdcImBY#VH50sD;uy<28zEHh5hhDa2YKM5U&l6!; zaH_1v=LBb2E(=E&_?1F}Vt~$GRIRG=&Y_P(3qD}{o|E?olY{osnXG&kw(!#25T1@a zu&9mSA?}K3661h0zq4d4>%nO1?SqvACWX;&F?z6VSbejuP$@M-Xj@X~C9>6xSFkb* zxa^^1yD5@M#qObT)rjLYvin6{bdD;W2kox2RVdH_8U%tNj~3QGAcXgmGd|_=k0cd= zU2ucuBS{zik>o5XbF~=lctA0FDxG)5HYmn$btYeY=Ueyw+YFS+ga0vu7^aYw7Qz3A zl&ps$_o&!`g~@EAuoyH0k!%SuH9X7}HgKSoVJ>u^$_UF4RPr|kb+Q^ItL4{(XA~Rw zsiELUXcZ`KigIN)ggSb_TT?v)oq%>|s@F|#S+Wa!*C7v-jnJSY+ek`P1Ktn3Q)VY@bY)%c$ z96uZ3e)nA?{^I?5lJ(~6GZMnFu~kcv>J_=rpe;q(196%C{1v=RC}c=pP#GQ07nw+=7B0+usj6EgJ-X8f5iW5$L?*8zaptmgH}mour6aEYFQ88qo4~ zL@e((C4*@2VJdc$;2^spCYO~%?~g=jGNg}wB=2Btn^_OE;V6`-V%I9Q6B?^JKlPAwyHm(Q(C-uz#C-ymDzCR?}NfDa#UF#^VB0=byNu% zR5f(pi5T#HLpq1Zn36^vyO#A@fwu6TLFE=K{PI+AWDO3c*nRE zf$K;j#~{+&;cN-dfQIxS!V0Br&id{<{o{7RM~#nQ_Vsq#!l(-=IRSPGrL}Xjro(ql zdRCP@f6crKugi-}M_L6c^W=c0&uRX7d9i;x7whBk7{_JH3YEKsM}wV(2V2wUFivbe z&`H-B;lzEd_v_B*>3OtvU^g+U;jw z0azt*I$2tft}`~EO<`9kn#86skOtZCJN+@6fur#y4gQrPeK_|tF4x;dcOo$9fgx$V ze3iH&aGB%5z;R*DmOC@z`dE0f=bCSh*jr_Gh#O+x-a<;IkVcD-zlM?>r^qqzB!h0w z?*Rvw3rI=zp2SCJ~o>i>_uZ-Hy-Ow;#>CnO(++z2G6peO<4 zVvArHK{wi&ZfAC8LN74a_E zs3_tEL9muq!3#x&|NADUgCfu z9{!6S;;dbm7~(!>A9_hJ5&QcNaiGPpJoveM8QJ2(Nd#c68nkvUp_pA1baHF=@%M;L zD>RI`V5}ue(&>940Fp<=Aqk;vN}L%IeTqYL`hW+~O)%Nu0RwFagCJc%Vq}I7{(XXM z3e0xbEe<~{y*;~_2NCQup?%6`2Xo%rh2}x?#W0%=1vn=(Ph3)}t+XYadA@wpg)8t_ zNNCLXK^?dbMU&s%=u9ZE!vO9hTGs@!8s^RMG=iMq;!r>g|1fsANUy9bnqsjh?ek_` zC7CX4Ps*+A$##m_Mv**9yW97$-@8|Zw`D71;zZf1KjOYoQ~%~QC-b%7C%-38gNc#;c8&H;fQ(zWVUSEZfgEL6WZ3oV?GF_aYts0vkRjIRG9|NK07D|nlTMH`>dMQ2P1xhtW?&9Hfiq38iYV?C%Hh7 z9lxB)7TuS>TmAhH^+aIc57_ma7Iw_desT9d*# zh#50poV29ohL`s?Z&}W_;KeQ2EuJL2G1`5mvKVX1qrjhR41^cFH&o`%q4%V&Z${g6{||kXUeG{YPXKH&fxl zaX0W%2APTsidj#QG)imAVO3%I6auM}qQ zxhKN{z;P#Q+J4n&yMS@o_&v*FWou+&OeYjYrYlp)O-M{)4OaSO=oG5-F9i;&JMQg@ zc6y=tzj3i5oCh1m-El+8{v=LzW5O2Ce@&QZu^UU5hu4t-HC?7kp5zj{vp4MtP+N3Ju{+Tlww4kf`^%5YnO8BGa|fSd3l7s}k4n(gH4$ zUQfITPeJkkZpnc5l27Elu%aHVDY64ggXWP@Y>pCn+_2=8`-{KWdL=JZ z=I3W7oFf@S^@O-?ehEaY2Ro*P6l0)BE~U-jCH&y3bc5^+2ubXX$&qYSmrFZ9uIv-_ zK6Nz|%3x`^QPso0ISL`O771YP=Ha*pHEC0_Qtk%Y!{0zj|HK$$jrH=(FmsQn8#9RU2RiheYO+Lht;s#g{8p; z#OiM6?Offx@t?kHnO;`>q2fL{_0mK{Zd*+}EfjN!A{QtvM$zI$(C~a2G$gVGh>bt@ zyZxna?nNaqeTS{~(Df2Lm=sYXfZATVlg^>@_$f2=swc{Gu>)>*y$tkY8C=#WFW_A! zhOo4NJi1zz$kfR`7k7b}be_5($RtvYP0tCT7*4R2b}ukq)GJ%1F?F=1l2vnbsXpeW_@}_y>z5V) zCE;{sM5_>O71*W**-eklG5L`r4Epe2J^R_QgN6OsV>vS^?eTAHBla)ULWm`bkHwAa zB3s1h()21@#3)QO`f|iEe8YAP{1~1;+FRK@nj61a`)k{x>X*cRu#{lqeDM*#9VmQs zyL=jjm;tKv$dZ`J74)!LN|>=WqDKkjPtZ|s>L|k!6lB@J%XMwaNAz`a+lh{iT_9qg zQbvoB5%2p>+>bW=%;L(3L~FlGF1$3Z%zZ0Yrh{VIC~}?Bq8i?gsLy~?C1<*>j3hGY z%C!qJWIaj~EHK!XC8^}^6azmCc4@bR7t$JpNYR-^&CN(MLbd?=vkOHPp_^lLXsr4s zL+}o7(HX+p0BS49L()X{`azMfTDOaT9Rh;)6{n@i0blr4(@6c&3V%a!cLxx=T#>GY zQexx*DU>|+G_^u*@=2W80rlD?zU|=2#FHN3syPL`ef$m}q{cGRZc?EhYN{I}(%3P` z<;)Pcd#mN`b7QGEaABv0rQ!g~?lWY^L@=lx{m?}J5Cn#SX3E1LOXBs95Qtg;$E^#v2P}>da&tv!0ozni~2Azsv07C_=&KXd3yO@z+QCGC@*KaZKeEbMDpA| zPgoTJ;RM-W`7GvM=W^VTFq8H-gp>-BmY^1({%CzW5bb<8A ztaRnZu3A8iW^`8`)#w;E{lr-E8CYzF&iit1?^MruZ6hq?cqbJafw&g ztS7T}^Gs}D$j-j-P-s5iOw7){0CnW>*o}@sX3LSE&-N@%=$QDs-<|%AErAo48O(L< zAz|U6GV^V3_U(5wmoZ=mDh4q2hc|+}LyxkRPKsC|Nc8Dzcj0xi-)rGQ^oJLVy2M$MUfIF}^u6dc-nVCsffeKV5giT22Eol{^ZDnG_Q?1A zf7Nd*{phkbg+SI!SWCPVshGlLkbt_^CnF zPHQiH?)&FD;Lkn4Ne4I?6{_E(@4bJtc(tWIkKZQsLseM1u;~Ea|3R%k7bwO^0S%8f ziPt4=A*=Zb!FtsJ#Y%aX5=3NL2v%M{J7B)1Onhzns#w!m+&lp6PDombbtpy^QZ2U% z_qZp9od?$30XICV=aq#(DnL{y=|Dx%M!C7B<3ZnmPLENYx) zfrNEl4Pe`$VF#IwZul1PvIQA(jB$hGa1bhjTImeA?yqs8Ty=`T09Wc-XLZqkjuT-D z5bA|?M5PG2=x%w23^Rhu0!(raZNOL2=V@9_;hBRFG%>1l$CaDBZ-(X3$U|C17mF=b zawStP^IB;fVr1GjCGS!MnEx{g#>R`VpntjOHZYuyG-%^LT9yGGVbIt|Yp95b%Xay! zR9$0i>mposPTOl`pq`PF5Fwd!qu^2Qln@Gzg36=ja}d-H3;uQeZ%KbO`5C92G1Hn5x+I2(xZLKc__)O z3~JIGCXdKI6}ZrlhdLw4W-2}ID3>x9WR*^&xGRx)|uS)0M5l44*A(^VyB6gE8&E?_3_UMZIpy z5z$6qcsD&R(6z!!k8ZMha;@-jXeqBvSre1bzeV0LYRbOVIsg1m z5*D<7b?aNXZ*HZs-e`X-|MmNCuU@cx!8uLa{NL+Uk3;*FMhzBm;=T1We$eI6HPL$% zz0!WS-Ap!!%vulQ0(Oc|Lk@iLB=aJr@oYm08&@1I6T0 zB!|*wM;3}OzTBoPr!z%|cy|TMpG9d1-n>fUYjI2vwXR zk_J7msnG>NIpCaP;2!E4kGbbV#dH-Pd&3TcwAsiJ7uz3>%l;U?U?;9cIMye)?t3>q z>x2J((fMOgv~_WKqrAZrt17U=&7sL7yU}VFRPBD6<1Xap-lYBMy?_3_WsceZwwIjT z<~GN;a0|goEBEFR#q?6-KBaw1$DgY%@a~o6sT-g>1jc#heo-nktHA**4^2Wn2q>6%G6h4=5WqcvPLh=Eu9Nmf-hVzj-~_g1-9K z*OZf7Zs>Di=epL)!`VYIyD74h((aI6iTYw@ADts9r#6xtNe?loFUb1H_E(=j)1-NK z1@%9z3$7q5Yr68DtS9RV6wwouR0slrpSgU;Z0z8b*%eimx;5_ z6Ly?^gD36S_|=#{SYY#CjlT{fpSti;eU+6NdQ352P~-u2tQhExfgS3P_<3q`Eg`h} z^HAd^h1Wp0dNuNOxRDZ80!lWGav1*T-7)*a(E&TCzT*2K*p~sf0?~ulx&sX~@>+o= z=}Kr#F+#NkbfJJ04%Q`t+NqZjE^L%<)I0%4c?F3R>6Uw=25E*2^or|4x5D`@C`QG5 zb-lBaLeGjD`^B1a3y7y#@XwK_)818>uG9_>V-}_%(#wEnV8vJaI-%SH8I0Jj2$i?EUbF# zMR%}~GX!e`b-kXAP|4M=$dfpBS{e4c{qm$eqyOdAl*G5lWUYr&53S9=$djojR07e>`SD^wLw49rPe zDJ@Pvn5{rSV4F-?$o}vhKBek-QJ-g%YPauoV(#?@*g&^!GYFuZ6>kNVuqrwsI584Q zLF+=dOBXrOUX2Y!_90?4SA|ah=}g*!l?UGsI6>kkk!-6kc@4!RQ)D%zHRpyfjsmY{ zA72lFzGfLp-Z}AWWBHZr{5DzccHUw3cXB8j@3Pfd2A}=E4X}WsXy7M%Nj5hF;==3e zDl2RkQ_K#E6jIuPw~TL8ef9JCMtN;0N@W#ArV3gjiXt;WTsj?c2UT=Ey?p_eU6cgh zg-)-dia4J>sH-?BP7A>6Tf{ljyJ!?3tO+cJHCDC25ayKE!cOErooa^s*BSV4*s-*2 zz6CPhs{2tfsd&Ns6=2sJRE2z!VvbSd2&BeETbMkcG)j^jR}ajp5+?)K*=Mt!NO6|Y zVXH1Kb#>U7aA)>`l_NkiB$5Un z90jV_&E@;Akqg8ErxLg9Q)D?e3*^G_3>1#j1JitF|a6(`Kt^FRk(qdwY~=xss94{A1FAZuMXhzG&jK}KaO#ei$InbKmc z3&Zr#aDb6J#P=!5TVCvI&T`ZhPyqLW8v*|;Uy3YZ)eY}KF=k=`@^yMp+)?N2ZXMY3|V6Ybe`vV zUyE#)HOmT^b<+mi@WR{S4T>hRMYh*(x1vFw?3Wpdd%7e+zuaN#=H~}@)X=iP?D;1; z=E(nk>n8;x3mCYtQ_CV?&>Xdc-$tANMmIK}j)MjTbY|OVc`Owa13fBxDQzFWOxjI9oYF1mce_Wg zAz8ugurfvhP1I7ar?HH@NrRM|A2U^6-SipB6TvRJG3E%V<7Wh5dHLn3aZ~lGb^dph zt7et}6JEdD&6yn}E3ij+H0%zTrUKp(*t;9z!2VymfHl~K8DZN|(EGaaA3m}riD2Pk zNRQkjHO|@QRUW%1=r4=T&%Yw=P+&cNljajB_U{L|BFMdwdb;f!N#kb5Ie)E=`rbx( z4*UB$MDWdZU%~f3t-oxW`QhUGy6$gASaL!~d4`aCGBhTw!lS$+ipq#00Qgm93GY&< zKEN1(eo7ORd~cvBFOc#Xu(DAfMOTT>LADG-J|~zaKVWyfPO{x^jw+V>Y~Y-OAv2hseaI*_&Z+wk zyyyWsbYxdR0UDAL#f$Xn&A~@QkUS(yVx}KKwrCy3`P-C@!qRE2bkD3BUZJSce;ig& zqaF8A3?-*qIJsXFu4?Z7?L{XwI``&PDW-YiG>xi5)3-{i#7ASQ75k-4nh)PFpQ)L4 zXnK{nTCv*40h5M(qrnv(_C0m%S8;*|<^NUPZ)_bIm*odxaZTb$pC>lx7fO2grr83C zLidGq(;cS7AduwbOmC;#gE5t8q5XXT{Z;8ec(YhE;D+tEIJcN|EU*shqIfs-NwF$k}H0mc^|`Xj2S?j@H6pm7ykG!3oMSz>_s29lFFCH zk!iMaWKK~Ggly_5?G4{9d7rY_yIq#6u8IKC#?%E{Krp><1`4OA2uvyJC1eYo1nx2j zsvP$(2d?R=h%5=D8@uEy<+`KLH3F$7Z*vPD`oDW)PfOuJxoY%jR(!K&zWIT=nt7%) zCBhkX-Y_?b%5S==h(_;2WYsH6r|7mvcF@}+d&w!UMdttPw0}-Q-!Ry;514ZU6TGAr zJSqP6H6uyph9?(ZSU~~Ipd4iu#cZO;MoNoT>9%)y)D|@^%)z-u(}TiJ_;2Yn(0Qw( zlX!Rz|HXc#F4~y|bWg`=bO?ERxAXqJ7ys!#l?5T16K_|LoS_O}TsSy$$O;F$DF#UF zw*wA%s|7uzO;Jug7M6Ri311?5{)iF6E*N_Gm@WZEymkCG#be>;QI*qEBUS=E&2flS zwnr6sm+)Hur9pTg29Mfl4mb_h=#XJ!8aRETeHXLOT+?SCc>hJ`4rv5}KUAtMqVSIq zIuX;AyM0?caL*Z|JnVOFF9e2th!ak-Ltu!|^u*)Y$9n}9QxdlEdq>GKZj+4*M`3`k zVbEl=j$+nQB!$xAIvYzWw?btcDv2UvBm`dSq8o(dbI~m6T`U`Vs4zyXn+yIOvR_pCV^7w?@o##!`^$8~7dG4@IlYsv`$~QLM^j(5 z?S*D(A~rL~=~Y;^3XPv4et!83Dh<5&c!}> z4IF)lTtajdc|P=WOTza}H7Qzb_1ew5;oF32HD-y#9KPc$zTL2~nZQ|gheHCVb1w;O z!mdWIooAV2j6YeLN6KBejPiojtZ|HDj!@(yN}C`!sW=N{(6xbuq84crAL_9IpXxuw ziB@S=0LgH=vIYVsa1;Rg~i%yBdVay5Y%aRd+hdTpKSi)IQuuk z2t7M(Cxo6|KVBQlLP@!8fS*Mi;JBhwjv2BGyl&c{&I;NieF7~uSVFXqk9CU|dEMev zIFcti_a*z9vwsgezt?B4oU?zPr_bl~YPxQ?;n??odGmG4{PUQf^K(+=!t>8%tNG_N z#X!XWD5cG(Q%En#2+pPBL@9!eY7`f!Wb`m^Ecd`Au()7r^e%1UvCO(e6epTWc9@6~is%4+L_fPLrrUf9=53p|32DX4x2Au*1 z$u8(vEl}n`9`X{?1C7q<%A<Q#&9Hb=A}$%-^0MfY zAYg$?bjL`F5;I7NF;J{f;eB6@nWlTd$1>_%_ISX_eh__mB{w+LHvT4bEG@9EyI>?N zEwCunWM*IJ0Qw!1hK5&Fgi)0bd<=S3jxSIxU`_y+g6YZ)!QG@$j*)+aVX=p<9;nW+ z^r~8YXigUZ(A@esDtq+hkPW@%7=R7^?6DmE#-?xoQe&CY+9RiCkb~U9cCI^FlZ#dp z;0cO3N|8EBo9(;9`-vc1gnvWTleE!0cb>Vy1WPubK#EG-z^HireUYTQGx1E=JWzvg-I@pDi=0DEP9ma@|42@JbQYf zB+&;up1VNB2?J93?)Mz6P%z?{*o})3pXAsJTEBPoFLH~OIsVPJu8=h@>^K)&*^Mm} z1JUYDls4HfO<5~FL)=flpjW~o4!O)6;)5yf-Nx6XXt9-E`HL? z6I1F!`=n<_7h~b}_1it{R&PAF<_soV*w6iv#b@477mc!O!K($d;F8Dyp`ER0+lG zqDT>?H9!scNio)YZ=*M=FF;Vc-uF7MiMNY95u6~&0mfI$$X-Rav@R5=zolE%Rl*iO zSddL^($s~P`1ULAD|(dOzSkm)$a=|1@snBgwE12q1aw9RF)O^m&S*e1{5_4D|0w*W z1vInc+Gmm$ZlH1D8RwA|Xu2ro6N+?D+O1Tjs@JnlhKzNk(I)b`li{V)kiZbDwK0c= zI^Cu+Z(Jnd_1*F|)t?u=uUDOu?i437sRVVqlR#?@*)m&&f5KB4KIciLhv_QZER9|s zsaNajcmuOfs~>|h!712u5AR_Wg|K)p>3?3JW7(obSOq)mZ7Mk zVK+N^ZOm4Wg;?u=E-r-k(E6kWIAOYZ9#7nH>x}7Uh@Xqg`ex$e`*YUV7Tj@qfXmXa z5tdRsGr=y*G7x|{6uMt(ZbJeFO})xYf{EEk3tdq9YbfM2;D%zhU@Ie9A~1T`HVrC) z@hUSrRhRfIeLcbnL7LGZ%+|rd{pT?j<}LFh2jCh=mq!q;%S81?d>7e%Ezp zOOjwk82^xB?os3}1=>hWq(z(<32G7ryn2&QTN_gu0bVV-x9Ncmss=@sc)bLQcxreB zQWRSra5M5yC(yhHU5`lBj!|D|20(AM(j!yV{dZ+V2LU*!h1f~w2Y}=RAM+GX`$$jdjaGks_0MMQPkTQ(G2nCEJ=2x z`Jp;2H*c0TPU)oKnb>Y%1WRVNf~~?vc{?3%{xD2$Aa@ZE0K2TW%Z^4-T<9;STadT( z-6^X{)=RVQs+oH)Zdb&cFC+fWmjGbXVK6E5C3_`aJ3 zF-2ee=1*ktOM@7_6=G5-CXphmC@mV2J%N2-QXt{kEHmda$HiH;nVsx{gv~wUvD;Ab z?_0-`d~)I74NLL~%la12Zj_@N1ca$`n}M$(3o5d71-xtluv@7yy$zF>dA>kz*g&^~ z5ivz`I%p(%0JeG1FcCm6Q=w$9-wu9P)OlGisR)IOszJb@8Fo&2)zcVCITgNR&Q9@a z?~~#Vn7-gvoR0J={0!uwZxN?Y9!2w|U1(ynEA}7f3_uU&{zY5*)fbBNjD=(sCmwU4 zNbl9>W%7Vh1b018d$)oggV}8-tc8aA^0|!vaL;k{Id*5|o&4FEp_Zwqq-%FPF}QI0 zsm^LL*+(&ZD6*T%$#GROCj z%%M{g7nF=-6dZ{Zc9aa!)kOWQ@V_qEw$Qrl+MA^*c4bT+tg#0|K913?dZlvSYVRVx zfj$7C|2{f*UX41BUKz82Um10Z?1<8xkHVGB_!yK?A;ghTV0jiNgiLrS`sX=57TePN z>V7G)KG^<1uBdE8R~hCI!1pz&C|ByiA&(ugAxO+iwAlv7{$R}mV43nJbM)PXCWU^gQ8T7(tM@S_vurrd(mkD zhY9|@f9?kLZr%}h1AS?pfj;8iF6(f-3hr27V?BF03NYs9{M}Dx&XZZJi}$8E{iXVJEIbr&M0%x73&NNCC3$UqC~Gt zq)oM0q$duY@>zcMQGK^8&vN8pPIlgPW2if#QZ|0g0+Ahoyt}0OrI~=PSRuWEVop)y zIL1}F=ncH`kRALp@JFvg5dst69kzNJRoOgDETIa;A`A&PLIS=^j_c!_dQ=L? zN6{5!>I{Vr2SrbGdF+QGY~DtSdC($$THVkpu7gtH)M(HI6|D)@rI4kOtuXOb`j}jWyZ5-fIeu01|HaD!sb9B{cSx!Wua}^dYEXB6 z4#hw;E|b!hfwkKzJs4CTQ!lA}r4y=L^s27cu)G(8AocXxz$Za%s!bvLL(8dTzv|F+ z{KcaE|7Q>1d;5>tU;XxN?f?GjcmMTUt(0cm2t^IGGV1<#%O8Jg!N|J5y!kb9-i0v& zS*1Z3xk)iMC~}R`UL?JqO&Vm%%%UHXq=>WXUhgE{O;2oJc?e{eYeYGopR1~YusU~M zBC|`BCxT*cX^x~TI4=@vx=q*XRb`}F2sKxlirMwBph{$tc)5e_ZPK(t3svsC%b{zc z&-iWQt!5e&)wB^9tKgGOS9eSCld6Q(z!#hyiN_igM@W_UhUsYyin{`%rj{60H%-?A z&xGYu&S^%Ny~oqMSlglDjSZ)Xm#Y0Z_qN0Wv>f5O>mp04q38ym;!@L&Gmv)lnP z_-^NIa|6i3lAV(k7Jy_;{p}w}`b*;gm0H0mk7Bk^B#Y8weJ3(5F7>=6G%=;12b9Ds zlq3nN__#RN-Q=I6xANQRW1@W_twLO84Y-~4+^0TEABb^wbcZ{to*574K?Q)*hD3X=nBra^5AjBz!MI4G1Xl^1)osZNT`WR=^X z6u4rxQB&r7S8z^alwX7h36xiw?phq48Ia>?)NB`J(NExfoZmUkG14t{ELzNC<`B#b zKm4OQ`;uTLLLPDXca~Y{^&NFn$R{rBwydx!!|J1$hZMPol7A=#3T09eK);IfN}&Xk z%L8txY_2n(&ETyNEeS7`Kwc8p#H+nKq3-`&>Nzj_xu1Mmc8;rML`9zBQYJslLtKGYb z{zRD{*(^Idts7>ujNrRZe;BjxUh~}_zAdsRx>Rv4_5!V2 z?v3kwtU)Y=ziUJAGipPj8nIFS0#;%Tu{fI$UiCMnZLQ41;8|HIEM)780|mnM*l+!j$T3=f#g2Cd0&QOr$>+@Q40vM*#v74ekzL5EY0 zfv)uU5Zd6tr56uK6PfHt-AOTqmC>_*z{7)hfnK%6{|LQLq@QWJt5J@u<1HYozOYe_ z!J`X26ml}lZlDgXE&+b^5CM#PfodTJm$8Q z%i2OQz`b{Y(pJvzB*@>Wy9gU80;tzSLWW%Rt=9(J&cLBB6uXq^F^|8p+xN88)G>gK zZ2A0KK=**UP1~ghLyGu)N?1W7p>P!+B;XWh{4!+89y+YTD)i2$^#L3Gs`zUo@_@z2 ztQ)u-`d7}wnG9K<^4OdwLF)wPy|K9f|35Yd!afGNA~-Ezz%5?1Tyz^Beu>}hc}Q?* zI_m7O`KE&aVjsQ+Pcn|#9`qv&xycP*|Mf5X^e?*6Dd6SP?ZSt!yt%JvP?V8}Akw^i z(uzs{XLOzcgcB&J`|pPcl@_D3J~i|rS;uWv?ZOT%bR`U0Rp(O-5Ts>O+NTt-g_1i$ zFa_=Yz2v(1rV`_@*$}L&4MC%z+Zwr8)CCN&RS~CD+3v{RsaJgj!f`j1wE}J>RL=&+ zzKt>28v8zCR9AKV*l(7tx;huH)`i0_EF$sP2?vZ2&U!CB*0+MO~VA zTKzOar`ze1E%+dx-xCd#Q(>#;W_xV#Ss~J)AcNWG9&lS6y@B5WTvsbYVfC$AEyxg> z=<>PakTae}>)3Ea-8V1qtDI_?mcF(9Uw=b34MiK_!YgHvQ6Dt<6j2NqnS4rH;@e6V zPkrKd>$L*kWt!gTeuaUDN%)-U8$nYeVbTrdBhLY^Ws~~duuCpYovMs`#k{vN_L1i) z=#t9{#OVHXZUR@{INS|{B3}U1Xn``g1hJy9#8#~v5wqU>jNEhf)jqJ&wZA>Yb(jYqI{7%Zymp!LYdEU zuN$-K>9av?ifpD8#QM(A>m}G5g%;~JAGx|O&1>+yIVTM|1+snb`JE!|vZHjWq+NJ{ zXD-!U!b9evOv$}yJdZWnO_~~BO-!l)_xHw{pACT%&{5w8Ehcy#XdZ>jC#8^NKJJft zRXy5qW49Z&XGS-W+{VpiYyI}@KJlJya+1qR%!R!smQ-do4?AM^Dk=rt^gUUFuyXqL zm{PwS$vXbW^Y7BzW1Qv*4UIGV@jdkY&N@0N;^%v8m*_4#A+apFk*!l#q3$Nzg5pFb z>7K3Lc|a(EdU6K3MQWfw^3xU3#U6#C_2V_}_K){dJLRkqwtg-rdp2>#$+;i@(c&X3 z^OPd;xeKQk60Mx~0g8D{kuNCipKmBz=ii4eq>t!+#a4cU(7=1}+G0@+FGE%Xk{Ek@ zS13C)xS#L7Td~b+NpO>z?^Wbv?33H45+EndcXXPCl@SLP;ywAxufVIjGNt)ZCDGu)T%%$)69;F-UU9nf; zK^>aE;%8oyWlNWfK91?o{0(<&{ML|D;YWp_LIm48Jd0x*T$gK_69TEXmfyG=VnJY7 z*56V|z6&Gpu+;*7FU3HoOEFMOK(B6XEUeEq26n)L%c4kD7P3D)k8YMBXP(YPC|(!SB{W-b#{C_~4rgFfc2SFP}X<>z+M6<$lk-RPn38{6E*-E=)Q zVI%%)ZHqYWaRSbS|BCF7{Z9*U{*=7pRnqCgz*%Nxx_T+*K1I4I?USH*uWmYvcSL=H z)Pc%ayRx1@b@@dxR``KHl&O$iR~2zERT122#Q;9>pHJ_|>q^KObA{ zG2mvP9rvtiStoC`_XQB30v1px4NT(oD|DUEsnsmI0+oaii!fi2#KZF!#Bg?l?DnjL z8Tl3pI9;i@C;HZFsWG1l+ms!%I(f~?c7@*SC8<;2_VbTU*cK(btWU*KW{hl~&9Ykm zOpnrNWP5~eFK91MR~GTN(W%T)AP~UZMr}xiFnPTCIwwrvYMD41m{DKO$udn$KXCX9 zZ;NI6LCxk1WbIIKYS)cn0dL#jIQ3SFfoh-4l-4|1HK{sgHOjGOV!#cXYx2|$be|_s zg<`A`8Dj_BI-$-NmS%w%tUTyl21mv=Vc9WmT#@s1CbQ$wW4eSLIse`Nm8mbf5kamV z)VKVcYz#dg^*Bsd?9l-d3)tkla>`+Us2@)dfR_DJ2KA>g=rMhxBi=k3D2(_}2VB4o zg&`u`6Tc=)`TSLj5h)M*-{oZcP)5Xs!#2mPjL1QX*-w!&N}D9P?%xrmS9e5R_aBJD zZe>*CHK8V{)&OY)&O;}~SoM_@(I>ByU^Q5e;&PBdeNNM)G4k@%O-wGS;XMr5 z6u8B7Uy~+ZT^IU*hwRl*eGE@}7?LG%oXr^xqMjOuZD+Ez+BmBewi|TdWWBxyWVjF zJ<`Z<1j0y5V^BU}*(AN{(a-L_UcDfPiu=y;1$d`(P{a{2 zLs$!k3$vI|#|bf%dd91Pein1`^D`69kqj3$Cosf4*`d%Q)1Cxngnlt=jVQ|_i|(7dY!dJd=gvF)+ToA`(mv1I;i*iM=AdM! zI4h`j-d$ynxL@oXpx~&C5iAZfC$d!SN1tc<L%9&E;eD?Ds&ciJu|cuYv$xzT-+#g=H=-_?=0p&VNm zULAw3?jZY;N-;?k`4IVqkun&?5X{sld#3d$^{R0L$Dq-*3nzoFAFoYzocR35-xdd1 zF!6_7hm*-xZUyHqybP+bLQ5&dfKKxcTn2U1Si|uEYQECuESsI9zC3g7s6&8#k>OEJGW;S^KJ{?(W~lYTVyGuZCZnPZA7cEioVS2Q?|`M z3pp%DW6mx-4duDm9&<3zK77yTb=%$*E;SD>Twl!6Ue`{e7?W9$$y{S@X1;gyh}$tx z_JP8V2RZnXw#+)-_bjF)o9g=}Iqkx(>Zex5;R?k-U3n9w#dJ@r5c^9KLi6Yj8l_7i zCzz++?2Gl_$pPJ4z@F7$B;v$V-`l=Dm@zI z&9ZLa59PprgsbT!3C8WR-Af?QZ~maAykww!s|rYr(RIgkKtgIqRJOoCrv-re0;5+g z_E;wp$H0`QB@UHqEHOH6)K+(&2gXwb1ck;0}i)y zAp^xf`Ui_;3Qd^#V{(L>ljFiJahsJLJ5MnmQ=|c1Vq}$p()MZX^jSKIXOasu8#b&^ z$Jk8|KWDmLjUio7f%j_Un=0?o&}Grp!8^t6^s32i(=ZR76rqR=?Nz|8q_ z(WXfIlM;>v*2s3vUU*&Gs5|Ow0haMsdl!<8EPXrmB z38BkooA=5jxcBa@Hw=Hj?cZ;Fw@FhHTo#o6b)&pCH2v$_zxPK3S3yvtfp)~^>mTFj+dR#2yPRe(YL}!w>i@So;a1_s&@HmhqbjH0@HIx^dw+qK1l>Fp z!an}^Xo(COZTn#~=%#VlOuqi3PhRw-0Q#q;u%K=S#?Q0>oO$+pHYt`x?1Uw98+LFN zndY3^f&+AmShq`&A~4Ss?ewRfrO{Opi9TKQLYykFp}S4lDK||SjyEaxV_?`x!r`dk zWJ)F|dA{GX4dl6O8i^%nx7)YHLzfoND(wDg!Z&N?=b=zGl7cvX%tsn5J25fRV;p{- z6DB5}dZ+!pD2pp1cAvkM6m#=&T{t9j+RAaMrWnXf9>CU7D1BNk*zB7RM8GHg_srI- znq?-rp2UE3C6d<?&jo$51Rs8%& zyt*9c*@c?ibS3%j;m@#T6;D)t8L!TTQ2KP(uFSf0o z*lWK4pf`ES9Ug?Op5DD}$qr-mSteNJEuI zBZBtRhr-4TgAoR63@91l$zwa6zrAn9D;7f&^t;M;$OdkP#)UHmd#w!3Hi`j(!< zbVyxjn=&on253T^k6avpp}`ZRJqp>5-Z@5^cpKfJ;UwfC*%ELy=-z8Lm5?JS4#}gx z@H-4$E(N^qz%|~tXE||V8v|6pj*JZs!`;1b&7SkN0XvuVQ&>W9*v@Yb`y*vYhO9>k z8ES}nTm#nmEJ?rHfmd}md~>EJ`(;5~_q=exZI_}(md9U0`aG-X26-k-D|=(BBC;df zlt_SPpg)h&?F={&wE`Lk%zqpTjIlw%dR7^3Xt|!_=FH4JTC&GBzw|sEy$ic6ELkLE z5674*%r|=QUaEo2^%3lDigYLj>>3+8fnwLQ9C;BtQ2M+KabnBFvoiDAzmr544z_K# zno2fPOa?{PQ(Dx)C<2XIq<=1oz5tDq^$<{v7o~d`WUZ>o>884EGxK-l_;EtpiG@Md z!I2}*{w#JZ43Qq1c(UVW;fo%OMdtX8p$)<$NDh`pwbM7~hh7HA=UfzH6E-%<;+Z4U zk_73d>11cfrjUNcW?r!P=D(e3O}Jxad|<-~D-%|p|3GZ3cjB_135&J~cKuXE zT$T2c&*X-%gX$E)3d)%y&-$L)0}0lfop^#1BqpBd)J6QnGQC{-S?MEUd}-D--By!P z8^v6w$Q9@++UX8q)V zn{&G9byLc`KhWhb=n96AG31__G+0}4a}KcQF!`XUk-&VSh@09j%mXgARw4duO`1cY zE#9TkZLmz=DlG`w^2!KMw+EDiERzE_heveNimz|AouyuqAPY-^v1VQpFVjQUr>vNL zf}DjGqyzN5XkcmF=!f*Kl>%Ki5WhCcS1r&X!L)%!hF5dLBgWbW{~2M(hyUV-IXh;o z4D;-<4s%qIu#V4;p zsZ|mWRaIK~rQ)NC(MRUS41|%)$SBdkX(pSXT^I4Lt>T5tIyWpD7AMF-_5Scmf8?7t zaeI6{A+!!+YN%to+yiTK@oZ@{rV_j8^vKL9RbI|A4_1RV8i-guz)3rtAVPgBBvxq~ z#C2JS?6tBi&%&|?)kxZ*SD96_2HaMAfO(LVi^@eAvRk-+*TZh;4%d$cMf3#7F-=%~tBdWv3t(@d_g^4RSl!poo4+lje-y%9vzH zk1|=(3&ho*D(*}vQC9jJ=z8cmN|qdjb@JWtQ^HD*ir_dkR$Mx4kca-JM?dsKKhIG| zIT=6dYwx^uYUE^)3tJMFWRQ8b1bPJ!w#bhBGRNp-Pz*JQ_5o$6JDq!s9VoXpm2Hf) zOgxJBb(_hq7p#}{l-2Zdh+--zQch{HFC$lNV#5DSZmy)fp#};28~~shR~?>JO+NLuLTvbJlq^(A$F^L(PAlx*(_&+M0Cn zFw1P?CGiYXm&&`w-neKVg4pI1_Tyv47dwKkW_)YA?fk3f(U?l|4PxGC}pg#|w-R>*j(=$`SYU$)K9vO80po;h1wB_c1|etT-9w z{(Fn^z7=XQH6>lUq4AHKkL$t(PjyyCXdlG@a&}YN>&ks0O_~btRNgxIZpkXmY7wra zdz6(Dr2EEl{v&*(+r6&5%UdVuk}nlx2CS3h@N4H+PS>j{yw8MQ5ES^`m2P~ca=O7| zKQBkJL{v`Yc;-mj6ggA(&2+RX<(QCTZF`2n0ryis_)g8ESPODI?;lMk`&>AFd(H|A z^%MgIQ8knnX%)@_<2Qy_Q~8Fei^H?s`%PV4l^&IX2hmT}&=;l!TvsLs92S%WBRf=? z;tu3}t}BcAcc!*_=~dSwG9z)nonI27S2ZX~#ky>dxX4|x*^;YLjN@Q!EvgsilGD=D z(xiwaUb2@Fn(v!bho+CkBXS@%tdW9y+#&Jo4)%#@sM>~8+Yg-yupp^u;3s=Y_E7GQ z3r|T^R_H3Gm>m=;q_jC?jUX3z3lGg{lo#49y$2P?mGaJO|N+`V* zFAlE~V~Bq*SfYWiG*0gYO?y2^?p)&U^tm(-<4vpiJH@C5XMbDav%wg7!yzyl`C~X= z!49K~>$B5-XMxP?JL;y8Plkew3on~iSTQ&EQOrY%+@rMpL2ar=?=E?tQdeV2`e0*z zng3Cht|I7;%&5u&TEkY=C(t2}CyHn2K6|AN7@cb+7o>0q8eg%8e!#66q;$$iwNQsA za;5FW7||RNFKXgJvl8B(uEc{0epjTaKJn14q(f5Pa{4mCli2VI{`cIH@B(R)AYQam znF*Y%1-xuQD{W3`A@A99q_>d3A_DXEaHGQUmegV0g z<&T33)fVs_*Zy@aDdFZbxv(W`u(D*e6my6o6_j?bU!PYBRApUa;(XRlt`+2vcDiS7 zR$vb4Q8vhOAWxxdQt4HNA~aFh$A8VYiY|rTS!ALq6<_A*ZVQfh9|>+y=q?cW>m4{^ z2s2FmJgQN?!KXPW*$*#1G#RUq+mu&=`W0;uQ8cPnh{o)fJH)mwIxXTjP&i>~yq^#4S?SB@ot@j&nZsVs97N>6l(;I8MRs58QL ze-ylg6GoUijjDDiA^k*!B9_H7#^N#!8iMCH#|oC`pXit)-+uFNXUAf3TzHknVsYA( z1%b^$7#M2O6o=dkzRc_Oycvc~)_weHb)`pX$l3^OWio13yt(>~%?qj*B)@t6^}4U# zd;Rj;abIoHJPi2=#N^)T{PsKNpvVgEUHw`n^g8HO#}y#>L*xC`^nko#Y8O=HTqj2P z0J!eW5hdYc2G>|Y#A0m5;*Xn9xaXVS=(5!ia!&!AFhMIqcX;=C zW=qRs_XK4G-!;(c;~ zcsIW|xam!D$%WVTy;lCpClu2`kv2+;4s^QmA+UB2xZ!rWYNxmy-hF6)q;u!hsDT_$ zvYOH1;eC8;=HJ2W3H;1&7r7)hoGZbnWa$D z0Ya(}iG(gqiTUUISk?nNO}CZ9ct4|s51V!2etc-4v9H;NSo%-+aK>l(UKX(Zx`n($ zQioc{xv=HgWd*<-ih+9eOiJ5Le=`57^iMaGwICt3$)^Y6iRsF?=&g}^!;^TWA=_hW zVGfE5HUcKM^4lSmdPa^?m`*o3qa_0#daNCJlpQyR>QaAWn<91DTsliS6zce(Y^Y4# z$;*OJvksR?rQ+M(j#Lw{cQEWU9QHoWVb`$(UnY1ZJ? zRwku{V!*}}QCf2j%N&uuE7%Fb;*b)p2u@et_sH|zHMbc;Dr>){D~&dUwF+T%9=#mY zCuxkS`k3)J&0M(f!Up){2bqg46a!Ren<#Dhx35Y;r5K8AvZkFN3F>%WMeJS=h|$ga zT#ikV5b~dTe#&9~_83SO#cU-X$K*h3W#`yyutpwzJ0WG`b&UMnv0kyk?^|-71tfns z_TVU4!VM%YoG!|-0?As6Nufw0XySxm+X~Jpx@=yv3_VQUZeBB!B{>k~lmzxLLu5ZH zhIxn+PWcqB_)l9tGM8P|vM`LTXR?_-X_4m!8K@2a6B>pd2)kji%VGWct_~`3CI3%w=xJ$7-{QAa++YM-9J9PaJI#E{QBcLJIL;# z;$tpsAUMpJx}HtF8@clpj>tEO=qQ^)AFqTT7D%jrt~fLl7h zQia+DrkE~%oC9h+M~N#dkPWqCjS}vA?6^v7+s=H|f}w=9-~9(@rZ)(oSSt9)w4UgV>V#Je{GIgCki&lciUwhW;JUJ&-Y2V4HEFI%FUW2} z&L>Bby!Sq|p-#;LJk|%DQelr&itz=pmtwf4hMUmx{wu7z`CNerR<5}w{EykE(G%~Em{1Kyi!1f6Z z-Nb^261VJAWVs7ZOTd*nXf?c^V$vv*3j80yaM&18>7O5&$h0dq&eh=otPaOIrMhYR zWDamRxO}-IkM)-m1()ZJ^;%9Z`oypO(s!zDwG&I3kh$weuiD4|97?It#e+4ck(VTC zlw-woB9jd=y^tnRH$#2lq7FKdSrWdUmkp85EPAtVEAaMLgl5R}YLobj`JQwoGOys8 zb#qKU@Sx&S7SyeJlt`9}vCy;hUC%aU?pQ+_T*m6lQL|@jY@@|#tX$V9yG>|#MPe~8 zIl^_Eb%ln=s^2|G=6s54O@YEl`((gJ#A zsBS(k5tmV+d2B$n>uz>5=Uv(PZ?;XgFG)EZOZ)9jC`s8h|Az7+)TA_NdZX`%jdL;s z&c-$fw|N~6iDy=A5RU&U>}u)xse274?#Qm;y1pcZL)MIc98!`;JFB z?!!IOA?Mg_%212tx4X}rwkchLsBDjYA>#m*FUO0E zsrm9Q=Wl5*pIUD#hV?=j56Bu10>Cp>`z#4Mt9f*lxGDmguvL%;aIjpfy<=nF$lLog z$6d#6F7hg3GHfxy#c_4vY!M4uj)H56y3(j(Urhs6js;y5kIQ6YJD#qJmRDl${pdxn zk;ZoU?IA@nr1=Mnfc@~zvI^CB6)+ASYWs$H9Gr&1KNM`U8-*b}&j}0Oq_)|TTe@tC z0SoJ9s}K`;=cnC>G|&zFept7zwBDjTCESlkdbe!x2V`(OTahGk*mCQN zl@>t=zDaY6pQ+k4FOAeF$HkmrL!Eo&XZzl^7^AnYy7!RFE^LhY ztc=m86myFrH_bXDw_a;d=+y^ zM;5)r%i_!8M(EP)lE=df|M)9~k~=<$K3Ke3Bzy2`yMKy%vlz;!K|wLW&jBp9L;rTq z9pyX>&pp982e=J`%a*}?*q?omAZ;q4q4vpsc@XIHYV-Ei4p4U0j;)Nq4hr4khL z0&9Tj;37o#L0g=k5A(xa;1t}c=yp^jge!Ib24ZX z@0XRj|IlJL{&n}i^^wL&p-3~O#Yr85sIYu7sps->pYs2-9WJ*`c_h zS9OW2d9C~`x;;vltHuxP;lny9Q#GLKqVv=Zbf4$;kUzs+{5{j|1?pPCCCQ>6(IsKa zB7hJ#Vr9&SK=x;T3acl}eJGy}sbl_S0|2L?blsY{-#sq+tG@+f>co|Y$YySHlk5Iw z&~r3+mA`{x3Mpd15^XG>``8y^eGnSbVVluv4?QT>?03ifLlf^@Y!ir@=hKIX?$SIU z|}@!?~*py$!qj)@l2vUGWP9T{*V zo2|a%WQti$k(HFTTCtRYi343moW;hQe&X3LY0p>r?Ba3U;56B|?s8%C`}rldbQ(aO zHYgUGLNSRHS%o#v>mtiSk|fok%OkO2^r*B>;0!xE;voJXBu4y-#&W?QKUELd(rLKt z8i<7;;{@3e)hzqaOSf!xWax-YZ`=j3~Taa*ZI$YREWr0)`Z5DvWTvS_NUZAF#glsSQinuIwfyLdJz3AgsQaK4$#)Bu9 zQxtQYBK4HkBy1o|BArnkARPvOqVapiptq@&EM1wwOJvX)SstlZ!)psKHP#IbxMAQ{ z4@tp1-+P|vY_632QKJCc*j zrY%-jj7`?m-~NH5zcgW!QY)L8M=@I{l7&Szde!2&<+13!#ES}~MN}IEeQW&zI=E3N zy5oLTc`zhys$QMor`zUrmPUmwEcAv|1WCReU+xHcX79dyh!eJcu=ZPTj)e%! zh0O~K5g7U)cihvJSj7uu=g3%#YMt1@>U7t{b^z@~h3$RL-v2{+ae^%m0vCVQbyxU= zg*D-%_z^)xv^q(5Af}AA^0CMXIR!5e0}oS2;|mUZkpWi1*~ zl`M7P$`zOx28C|YC?=I6NmxeL2GhYx`C&n!sL!hu(nFWYm1$R|VZqz@jxv0;_71f5 zC^&zQ-L)8^tbgK}53)*USzwV9kh_lT6Vs-><`D%qFi!U+2Eer4*P{Gp&7$b@VS9}ixSK$i5bw1va8SwNh<~K1F>L6g+3%6(9$A}w_f#T%9h5T@l z1wfa6m@u2%b78x(#!4u0=|m=OBI&2JWiH8z~jq=@4>vn`(l<8F+ z5K-%*t3tLg-SP}s7HtI4M4WA^LhjGn$_FJFX<{VGk9EviB04PU^DOf_uErn=!UyhF zobkg&KVDHYZ`B;|cST*|C1eYoBv|dyPTz;6f3jDj*9p-_q9Vu;Rz>8|)iR^^ZRNgL z^Uqf4aVRvOzuvUSU+b}hIV;}EN2Mu4*!IXiFRUhRr#pE)kc7`sSAq^+ZAd3?U38cD z5X_C;@&myum2sjTP%m!+)tXGzsqj^E+IU^!c+pbN7f>>B2s<|6QcBSqo)#FE|NTFn zB_B>A1y%+ilVUbd@D^#!71*eh+7k`2)GAQi>EYjgrP90At3;Vke=-YocXe@^lm6p3 zt8jk19rN;z_!Lf)iR&&Yw*PkO8@99wE}IKrp-89_KTvK3edt>=+jyu$e_XxM{X^wG zITQNO^G!V2OL^_WbY-Wsl)gMQTdGR|9gh^|pz4hJ|FQQia80FW`ksLkk`F^} z1d}5mNCbnZ2!@JGqjTx>vTb*}ZMWNQcf0KjZO3l6o$aQx+bg1YMJ@^|q5 zOn_w+13950DzaIMRVEtDQefo~_OTU)4MTfku@I9Hm}jU7?4DN!ISfcExam_7c%oH- zu*q~U;av6XARWP)CYMaeM|Sct9*4;XEcl;blVb~NOg-SSM}NFW*`em$ihwN2h?+<4 zk4};iCx@*m2VsV%*G+L0vz{VrsK{d3X%B>q2DLtrUt_Xj_@+Z-4IoVKn4CRUfOzev z%GZ6~Hv%H?SJS7HYYuEiR+_Au1}UbWB6q0BR7LX4c;Helf;d~esF7}R66w_(S z)x1`<;p`R9%b*w52rHUXUbUPDUUj@KkeE6-uZSGtUzHEi_hc)AqDe0I9;j=k^HXNx zRo(jYgY;D>={20kv#WX0&>r;oY&5wG=kL`;mCPbgR4JOo_ZmJQNngGwh|z9DsG+HQ$QGS2^vTc+AX_Spz&% zi({`nQ$Vu<9yN3ETVGNbfu}zHlS3rKnN*v6;Y%o{m?8yKDHwa1w}^%$JAG<-d7RqtM4ny~ZjiIv{f>JseHHGzsjBBRxO@)s zl<6~%x@i?nl53Fbe$VX9n*KfG*qgAS8OLWo?-o`=`kpD1cnH7b-?t>Kq?LNx`P-%7gV+m-@Rp`et5*J%|f zNmLbl0Ius<@Qk5&9wOMDvklp7Kh65g|sr&c0UZv&I<;vlkp(k{JF`#Aku6>LUNE z(jMMXk8JLCMcGu4hfrosujCDejX3oRb&^#JPrF1*K8m>HdQS{Tq+Og6XWU9r4$pfq zZ^WtM?e%Z0rFux2*V)%LLs0qnmpdXxoO&ebyli^K+!3d2db9VSU{EmPv?lQLFZV7O zaXS7FYrh5e`L+d*I2E`&@~U<_>^$`5sc+wYd&KF}pdNAZS1aD_S%QKTn`c(|-VLl+ zJmQo9P2jg8dW9oSm$*Bbd*ap3_}t&;QB8CK_517GHO?bWc}qWETKsnWH(rfsAocOv zKm6)fjUYVw?!{ZA&Y5(Y%yKOh(@c?*RAeXr8n?)Ghzz(DakbpNvJr^V#s#MY_5@uF ztm6)F3qo7fySa(s`#D7(o4NSKzaqX1+S(~_whQ{8wThh{8$DXpQG&rRpbj@)6~{^N zs0nXXkH|_nhGWQzdwFh(>ZqWCigjrTt>Jgl_&Igl)2bnlE8@6d!*!>;8pGG}?z&Zb zMzJqTw#{5v@2^;IC=6FxHWy>WavgUf`PJ9vUomgDb=V-0g$CNuH!h1i`FmxjNDbhx zn{4(TrYor0&$Q+B16*P^<>*iykz==h*=^=c;=yV>CBzpATLjTV9zYYPpHk z%%zwtiez9g;9_7c2iAPzXeJIDm4a5QMr1J_j~0C7Qpy1-VcE4s=Y&Jdb&sAPXk~L= z>zvE2pbqiZIiDh>L1pSS&h|x`)wp`r=1nk?Odl81Kk#GoDvX!Pm9|isvE93i!I~R= zG4sNdz_f_%%0GPg@sfB>N?<(KzGH3rhGEb8!GGpfyB@G?ROY^Q<TUDIDK?-NVB3ccq1tcYgIRh`sfEt zm+TRHck>14Fga~KR3_IXTmDkMR8VnwE8^?sF&c+;O)bP@3IwSUAAa~@TyX8{Q2U)L z$c4pK5|p{5hv@W`zOBp}3M+nVIbHG!;G9aMTNx|_zEARjcQaaGV3X*j*T9~grW{*P zY{0e2mtw+pSfO}I-mK$m%)8z!w2LLo>ER8z)dO1~@(o^>riWAor-?LJ*M#-@Rl$Z6 zTE(WRmt2dcm8niB^2F!dk9yj_$7B3sF=4NL%m=X>3x~}E{^!G%pYV)}nmH-oX(X%0 z(!6rqs5OC=%2V~9n<-`!MG~mUPv6+jJ_)#!61-YXRtSMvp#;ypl0 zgk6>+voCVWgvdGH37o9@a)=o#rIUdpcJO)PTGZKrA6@}F<*`?UN7`<$`-b$qGZ)P{~9C75@uIjoy%p${!% z<_NS|4@jE$pM_S=YzM*ki-9enU7VJQD)nc9AS>a@;)xS;DLc4+R((6~#jb_}_oi8J zHJYVKekdcGE5qh7S{x3aPPSG%R5~!Fj0akVXL~4&T%;f)AfkcnZ@ToAga=}OIGBuLXa9VNiz$J zcW#`Vt$b-VUS|27zf|+do8>=!_hm2JMn3KIkX5{M%KJa6U4pS)tiVn3ivq&-1{zt* zc27&-VY68_-Qv+VE8QoTUdLHO8K^((jbruzo*W#r0`QdY-Oh|R?~8U=zr#YOGcw5} zkq+S_d6TR~p$px>T=g7Ss8y{}ph5)7K-e1#rZ$HKGE8r@^By~7%s$q*NBKC+k@nX_ zlHfqBcnSo5z zreSUZLHeMwu-gkR1?wbjlC$1>rUOk&H^fx;x$feeR8=y)(n>Bi89sjGL_&)NaE!L( zIX@;4OzeQO@O$U4c^iSFdws!WlKRRd{VPmhRzNX8_>&D>nvy{RGBmI}-vVXdD|iEv zb(c)vW(QSv$IlV(EIEbdRF9Wi~wuxHEQKl88&JiyAZyypAQ zX(cavvk>~UxbB7K)HFIRXfh(9V~v*;fEepGyB}i(h$((Q+Ox;cIMWo4{$L+Te`RKx zDiceyhhm`lIiHF|>cf2B^}uy_=bg*q?2xR`4ADVljCYlAn?NUNq;o@al^LQ6YJ)mO zl_6?Xw|nBXpMf5h;o3^qcJVgBzVHlDv|k0)8-Q2Y#qwAIu;=(LhEw)HTjj`iV$E#` zn;eY;r}ZqfC!(1_L$j4rq1;HV{I*?&`^7``VnDq38TPp1UEWvz$$SCK1`v*0lx>L^ z%+SA+;Bhn*TAM+8tX175AAo`dpcW)GfhC-yE|Uw97bD||rLYG8EI!jWQgLmXadv5& zckTm{#BO$RU{e92txr_|WKj%cbhO9-k1_x)3ZTm9rRzY86C`Gm_saJkp(xu2zm3e9mnP>|1vd#sQ_ zt@w7ILT;S8wj>6gC!5Ebx*T{twAaMz&7m0R1Wl(RGiL6h8)uZMKjSAYTD7p!^}6^0 zv(XE;%hYM&6+sV}bn&O2eWXEu6&`OGkE@t+Rj*F0Uood#J=dU3?5bR_BYMz zU@a7oZsk^nV&q(tz(dBZPB86<0=G%P2gse1t-2jh3X%1p1sw~(?F5m`^gzhhL9t3a zlnvSWt&cxuZr`^KzZaY6gyXJ3b6WT*FMG>)5BRm!ivYctWiH{gG58a^Kd`oK^0dT_ zh_l{M*c+mz&u0bL*A5oN?DsX!GLM+!Yl+r@%MK2i%qYba1Cq_VfF^~vTb95p3f1iA zekRNUPDu=@KpY->v|1GS?;V%>GN2^QgLTW^S>V4Se;hWjR{Au#VBtY7{m^ZtRO_y( z5ckM)=ak5DR1@DfWIbTaCYEP*O$Xf8Jyw93BDl2W=QE6e`BuTZ@00Bg444BZfGMOH zpkmH}0s`m|&-bqquJsrM=9DT}4ets`jVM!IV3tqAAgzwSb#VcwIe4Xa?}A!MtNQS) zs|#+)w|qHAbz?rdxjQ^{{2cl`vpS^mb-PWBw!+0$SXjEB6Hi@) z=D={dZUUDJ6myOuEmY(P`P6NzBrc*3*p8zFIv>=1hu&p5D3HH-S-cvwar?P3&PRRM zyQhmTi&r`KN#jJFq*aYJ=s3ACzj#Tne28=k(_L2!3PQUIZ@9_N7$7<%*%w6S-SEL1?2o0WpB#(Ay z0YzKE^^8e+?g2|sdiFRwaQ*n|h&b2?uC>X1dq|NpIc_qBl@tRuU_TXEAlS)Vhiq=1 zVk8jD6)%Z5`)U%y(1WOYYv*DNQzg&Gf)Q{pj(C-ElIMRG3Y)pWO^my^gV$+_r>*1c zlkH?M0A}B2jbW?* z$<>ITKWNVWj;wKD-t#IuTSFgn58wewnX6E19Y&4FXyN zez&YkUg^^tP#Og0Fj=0dOa&efJ+=&MFEoQ64$|@R?b9~Q1>S@SuzO<-HA`HLF(Hlh zRNIfebK*j^dHChkh#5dh_tSanofMNs0V_A~wx5M^ZG#}pJ}<1e#I@6?(~j^QP?KFl zG|?_PAA}N~0_=dq(Y*!Br%gf-X}n1y+fnd>C(xT-+xZ*gO!M33pUok+#tP;-Zup3- zGs#0fq?iX38Kfe2b1w;zaid>;XknHrUywdk(+rYD>AZ_j&Y)GnLJKlM`gjcz9cUGY z0#}7JOZR!rBs8I+{tp3RO|_AlA&i-8jwIV`J!$bsZ};G2UQq=)Ix7bv$QHW z9U9(~<+yy;DmoV=kg~b=$!_j`fsP;R(JV#LiQU{9SfTaHPmok^Sm_VW)A`_q*>Gx? z`?^JqvQ7{t?-O@$(?xK)0}9ETrOEOP_bx?M@Ih4^>Ld3_wSF5y4l6r^*W{Sj#I`8x zr5^~>$+~IHOUO+ce0yfjv*iW#}QQH0DXS0l!Y4FF=Rb=N_mdSwuP-8Mll~N33vJ0umF4;jv zkw*p9#$1^>;#BFnl6PoE?9^VmCG-whmM8#4YQ%YYjPnYyE_9W1rE84yxxjN#T84^2EgEPKp3e6Q~>HiE|O zUSl%Z@4ySEGbR9Opcv?0t3|?PeH57Prg7%c!MbVKZHFg_f@iq(%d??bEelllsuztq z;VLRywZju%l|5w>cP>lZr z#(5_zP`vir-~T=1Yy^c`_ko5ie`QL6VFmS6DR3gi#8G5DE~$_}a6lMMmO~~}Qx>+* z3v*hN4UKm*^gNAsS&c$mYB~VPd&jto0e@!LHmFI>^-C1SW+Dk zLOR&l4F~peM@{U;U5dF)ksgTtF8Fmbj_03$9LwSlMH*8RA65f;U z1b+PiNd;9lH&M{49*`I=uX+8NTMJa4WA!;w^(K*8-zs4*aMpCaab?M6@o_Q;InNm9 zI$o{skZYB2m2->w!Pj%%-u!0KS694qVaYD4=SO?Kcj%v6)hjrwLb^DHE31TtYtN|D zproi(y=8WzQpe8>FH`l>*i@IJy34Dcd2d0hI+GkzofKb}mivQ(m;Mi^5{|_nF-|m^ zqtn*?XCTe_>(B4MVZ_`+ZpM#E)hmO!izb*mNimRHYlPZ!PAqSj?vicxjuJn;oh;u- zQkVfRtpd*&2%dZ97J0;b>G+_kCg_r-fLdBHXufnqtK?m`K6u7M;4e#OqGsfADxo2& zHZ*1W8qzN6t$WgM15x9R&=u9){sy&~a13msNnAQPL% z+0D)HISf(fVmimKIk;7gBoNh{OqglnIJLf;c{}}D)lJd{I!#o?!6J$pIzHkg!1xgq zz@~u~Ek1<}j9DNHF0~Q#>^_Xe)u%T5y#F6Y19W3?{A_ZA-HhhIKKN>rEXD(h8Kg)* z6?qN(oM`H*yo`fd!&*hDXC_st(9}pqWLcoXR}XDdwY&`XgJ9W!;EMhP$_O!Iffpsu z&xdHTCQXzrj+wm@`o6btE{p$)kHRG#UV(fx=+l?Wbm1BlOmC9y0hQw1Ion#G4u-KO`2wpYO>lRj?*Y!>Ar4yjth`R zqX{+;)sH&ir(IxxI|)*6Nx&`*kH7p~{DE{R3A|CKXu-!Wy)>q1ZN%PF)^uVwXMSF{ z`(fVk)k)2cp6 z>Up?@`i||MXBfOMO3)M346PX>PFMIRgJQgObYpm>E1u~IT1W8e7PbEQq0JOGJ$3~((MP!Kym_``X%np2ei)1E-~Q9E zz#J$qRmjl-9I&jSO@htUTE7%nu45S&%u-#R(E(M#8K7X^&rhDJsZ#D_@?AAi9`)X% zPPp*PrEkoK!i*%pH4voM*k(zdHwrev;s1nr;77i%{evy#6%K4sERH`buQYg);UKGbG6KZszc(lWF)M-GdRc!zz zwGr7zE;>3-(MjL+-!li&p?G~BbZYcKUOf+rk8y!(7vx!suIIV7_JG)%qo)^}b7Pg7 zyyz#SQn}Evk}rtkRL`tru9Lg+E_r+SAnE5GhlZqj2s5Pf&Pr~0Ww?#{x2ik&f4M5J zT5?Z@F^UJ$4!h@@$Nhfo@lv09CaZ6+<7NqOe)?{#x%lF<4y)t-{=|YHx%iJQVw2~!)+24h82cg zvmX)3xRGr%B~O8ILN_fVJ5L!GVJ>##u*@?Sgl%qwG;^>Z0(!fn7V7AGVhp$CLj89u zlPAUys$FIYivz?01QwrRj}w+bFjg9CP5&){xpU&Mz_8$M+@297z%2D%*^u~iaxt)v zH1T7(F##EFLt(X?;%UoWwkx}q?4%!G7td`~C;9c!17TYw7(F`*T?Of!9Bw{$Xu%Dc zy->sgFJpTg_5ce!lNDa5|9!imG0f=9es_OCIw@o4%sTLV)NJBo9ibQ~Hm{^2uh7>? zb9g7eIedNCS>d4grs^giD;TybR~YIP^yl+9eZ1|;ME#NG@J9YU6|(N+a<)pc1GI`y zCH>;_o_F}IY8}5~F&<0e6!FsNNBW~$1w0=-wpD`5d_0~-4|7k0U9%mDHe=3CNL>4u zjC+4uHJ>etY`U80dGmxCY>dLG>&ATBhnTfhunrTfA8Ydsi@v+`Pv*tp&-$|t+-hr~ zVBFAGfGr34f>Eau&%M5y7#kVUxab z;OWK!{|J^R6fe0d|HD0*F0@sRd=oLwcre-Zm`{s3hg-$DJ3rp=zq-T z5fArrTsjw=f%ECEtDIZau^yN~?3a)Fck=f`9fJ0qDmr)Za=}sZ$0{f+&ZZA4a2l%h zMbhDXZiZ@fAv!-TP@?&Ge>o@4>#kdiv`Bo-ZI}C~{~7giUIU%zzsD1Ql}Fac53GdU z*mKP|j#>-Eyw9#mZE5CqU|VFNnfvC_eyGk|BhWLT$kzz^!yYpPqXfAIxlRm?P@40? z4xgQj&gZbQk~cy8CeI(1@q*;}PqqE*ycI69%ym#K2#DMT*^t$|8oJA~gzgNj3~nTQ zeRGw&rxk>^s`1}md182#u$FVzBjL*aJ9flvfvnDW?;PLYS!v=n^`PbIq+7h_+Trn}S>PS9ij*kAuU!8c4HZ;j`8_ zPYgNiM0^#V{#xj_pG7w7BeKs9{)@nfp%2QY{ttymuIHw>KEBfO^bzF?|9s{PKMr9p% zN&96hrLF2j(aNAl0Ut%Iif2J$+P6Lr&*%@xf87c58Cw#!-J;vj@6Y94mAAt2{lLPLt~gB>3!?-u*C>jGm9BV1tJpfH z0F*^GdxItn9BoxM(4Wzuws*+wV*Sq-T!Z$t;bV=JmEYg7 zZnf(%R-kZP*QD|%A^Xi!KhK8w9k@u=BB_U=7Z$|(Vk3s#F}*WL>|0820dCd-*uu?r zJZu@jKb&{$mm=e0H-o>qgTy&7FHoV$DqKr3sTA1?q%Ix_yi!lFUpg9F`Zjr7_R|vt zRr*wh-3TBlOxk zRl!Mq`GRZQRes3aj4}7g0D{RE+genZ+%yS0>`?L79zmyz(QGt)spc z=7q7ZNY2k8yUFELvu~q0)sqE|BxImQ@ykoX zBnVYB^T7$k&Z*5Ir6P<~dkQ_SuZZz{KQ|?Ck81}x z3+;J$?YO?kEXHlD@2FMBo;N<0&aoOd$6fdylD6`kjNVG&gZKYPqMSh&{^^Wf0>x~g z$U4-D&lhybF9qKX>+#5hl&A*97REKEmfoqA$E|{dmCu^U%U%&kocrI@`j=f^E#s?~ z#r@oPAXmb&mp&TXa!yGL!EiM~*$Wa4?B?zZ8k8r@DTP>Y5;xDIl#>#eDgFpZXH54!E<&d70w4-~-UB(ZOw)*Q$;SE>U%d zj>s~_T^zev#8(3@)uQ%4n3p%cRFPH-MT{58uzPw4w0(gv$Oa%;U#Etpuwiq9D{`&p zd6on7sAiXYl|Exs1{9;H4RP4V?hFOv2F}w)!4BJ&?M&@8{~1$^UWqDm|NoFjPL?ko zD>Ub@JctD1vQM#ZZ=jfU6j@D0qU_8;WeHcaTXu3@&G;$#Z5t*lkYd{hjPq{d9{h>9 z$|9TKi37*fEQm$s3x?eWNeSm1R40~&UG@A#h=OR#P~;Jnulk{YU>UMsX!2C;i^|1i zVH<;TX`Q4#43>TTWr!x!cF?Opbn-E_t5sAiE*543VTwM%?6n zd+o5AE?$c(KX%{Wh>zdwu8Sv`4vddl6MU3W40t3(RAh-~fnbPKEY>O>Ex5Pfy0jx$ zlg`6k)QNrXp7cG?Kr04$vs^lfv(KwIbgfih)Bw7{UTqR!M9ESm%Xfv=ayA8wI<-mA zdVr2v9h76D+;p*QLIE`n)L1qS6POreg_CVFPYRUpwlA@HtJVFq6Z-H9k1m%k+Hl+;7CM`~}}=R+gUI za)hj7mkM;;A}X@W#0YJpm@O1Zq9U_Z=Qx?1Y(cWTENrv)X19ys1->ongW+j(nYt_D zj9RP6qO%3D+{u}o^drFz9A6Bcy|-$wy;=XV&p8Ex{>3m$v_j3~5owj5fyT2MG78C4 zN1b9|-T^g$xfkW-qNt#KvkrQc`>vr*hc|hz1$5NO@H)+PPMJ$9sF4AH^h9tnO z6_FZ*w@)-g*x_JOrW*r3Sxp)JBU@a}r5PNyhTei^MB{gKZ*Q56iKbJkMu1MTyejwv z$>pr!bdx>GK@Kifc7AQk>}X!s!WKv!WAJvp`$^Rv<@JRtKo7cW;r)fj!iU^YFQ$)9 zUW#SD?cyfsK`8h=qyE!Xd8%I-4RWe*8Gb~iYnu3xXdo<8Ttc4~Hc8|CupECSCy9Gw zK3;Q&L{aw_jt6cN2C3(uZ9Lu$D>sHp`PR>R&D|Ia^ZzC&nU3M1ESM%rup;PCV2c91 z!yFpCM=X^Eveq^UE>b|w03>sh{3iZuJeFA=*Eo#zpb0+4Y8)IlNBFf5rhMfCqZ^Zc zzrmT@cHq@ftckwD2*rF(ks&JbmbgQV(GKj#NS2@eCQLmgoHhm-s1JamLcXX#fOJcS zrBiwc)JxXDl~IDiuwiZ&^En3>QAk$Z&+VhzId|Pai{^70Ppp-v%9G_iL8&3Qr{_@7 zGf0j)Vad4RG5C-w<-G;lrzMJz#u&*Xqg|Fy)9cq5Ce1F+VMxU{(#^r!r(K1`)-Lz@ zz>eV4G_c$D%Twl@^teuULe5o3p9rrH%ZBe9@+Bk^Ou!#!poDM8t(JGrufuaE%%B(N zmM+Q{H31Q>!-_1h`esadctmmdH6s95t`4e)3TAc>aNKb44ii5*o?>DtvX+XB<0X1m zh88U<76L}}O6AOll%HMeOXB`&Y@fnfhlw=;{+xMYZGIU}+t!H5i0*Kkrwmfv&j2mbsG z;V=`lhbOq`yAE^c9iAA}s|r3Al4Mt0#uCWbWHu6@BP|nSKcV9_a--AGP|^p&nPk{U@iG%W)<;2%jR2mpfznih1OQwaKv&C=uK(Sj1r=CD1GmPC&9 zW@%Yil~gA=!UcgxgGN1GiVm*fa)=J0+ucal@lbdS1kLWaBJQ$PIvNBZQNq3|c+eFJ zBanj*9o*;uoew?<`;s(wTpMVVcf{WaRnMw|2l%Hwx@o+)iC#mY>|?SV_Yk>wOTIwU zqUcgwoz?~9?a4vcmCixEG!l6oD>bIXM7o^in#M(LyXg_nyv zT}!#CAvI98bX}Uu8Jw?GoLPL-rGY-E>hshp&U!bJQP)bCSnTdYjQ7huo-Zt$dn`{A zW0aOSuIIR?KfL!hMjQ3p=AX?Ww_cfS!#We|^^js7P-KvbtPi{jA-`7j=@}L7yFdbV z6|l(M7jEXJ2VGm#s%{so1F7kIZbyOcEDhRT50hPVs^9PPsJL(czWv|!Em>hax6*Jx z2M3PK!;4o^1=R1cfE}qlumV)8SnYg!$;Pi_e{aRt_bpiiM24OGQqMBq8rs^Y#fdg^ z`@_0q#nU!~cgYL6cR~M1tGEyd5y{}?Jp7z=PMio!4R69<6yqE(KM?*Iy=(Cq=_+T) zoxmq`MI4%ef0x7BuY2G7P zmJ$ncz$X-K!hZR>xmv{@*KSqv{CaQjhVnS+ynSAf?@-~yeS@cYb9Cb4MHpnFLS?Fo{u3NH|I!LyH>+$N2#ZRkj(u1~<#U4ht>AST% z&XG;*+$IMO1(cWslQJm=cy4x3k%RMeKDl#%iVBx(XmFC{Nq%?fUQmnH;F>>6)g`Zj z+EOIm!<>TV*3!?GXrhAZco`~tlf&D9601ggychZORoyV}lY2!1Mi$!rs;6o_YkjL{ zRtd4kGZ&0X_k_?}u^s_yAi?^f6L7;mx5QTSWjPy2IPOZ4l$schC&F$~U%3pcle6Bp z4$U(6>bPA3+s}n zhg<`GObkQiaoQy7-4o}5ZW=&C>sjtv>(L>tWUjfT(^l7!;F7bhp0tm#IeBT>s^zKa;i2AY%D+XFv+YK;_pa zNDk?12hv@zdM-1(Ood(lI$27@s1ugwbjjmgk54k^7ps?Q@@Lz3hgGD{aaRrPS-<-5 zWhc_r(LEkzsXzI!ROTNwQb5Um5Rbd%K&Jqs$sgRzBC(%QkKuxn<5&_G~M*14Fbg^?-9XKEo zZIVYCpqRT9xlKhjx%AR4z&(~5(8XLNUGnaK_%Kg(DOk&tg<&Rp)T!%@D2U;wMl>ok z@oz%t3$7b=>L5Mt8mtSAb*+TN)2I{1_@VmJ@PK_@$9S3tvXoifs(eu{B(Qs-E$XD| zp*VRe#Ocm2xelaA2}JX!>*Tgn57lOA3uF;6-kU7{!*zmcZ1_-^2y^Egh3Fd|&_q$o z1;*=J71(4o>V#>d47UTba_NU|`4F)gBvnGqeo38=VaX1eb^L7n6E#)AT8YEvkMiy% zIlCEQSMil^oF?lXxXL)s#1l`Yn5`6i(IZ=cXRFM&$0KI;O6XJ2Ap2^sv>%FRpe$bB zQMo)~*ezdl)TMZ7{KPM1#x@x&43N={8T&aCebll68YAmI<^I3k{go~8HwT`@EX3c? z8?IioEK9Y|t3`2I2-RA9VKp~tPRkd=Z6$!dxM7mE&S-zS+q^2=VOyar)OjNhALLiC zcxfe59NJ0e+doh?*6>&}#$(-Q?_;ct#A}DskNL`tQ^=OY!1H7?JI}>|jl^D)2`7hQ zpiDd+_`RhuJPe`141iPKTt!mWueZ(7bt8{$?a&CB~?eMY&E02u@i=P%TVC@OZlhW?*%w0Cq zI03Evz&nu5H{*mrrZ>)RUG2SGMDBwxnznD z`QMSCE<-Q-Kn-qJal3P!P-}A{dJZH zVT|R2?+Z6my0mr>V$%!5xTj7{VLJN%f+G zs-nZkd#=rJqNqmaef`x^!^-By zsD7WTT<=%rnFDg*P@d0;^~hGmc$l0x6sVJ35pN3ab1MuhKdytC1=Tdt?K%~$A{bGdpU``Ub=uz@=NC3XBqFc5-9AZjj^w`3LuX!V&!*H z&C7D-FT0Ql7SUAD4FQ^cUMrkLPf!yd><_0uqE($BiuFP&hhFh^4|};MEbuY`=D~WXvBC>=?_7Sn z#5g|*?+X8uB)l@~q+*lVES+K?gtv`~?BVtBHcZn-lm725#aweNj zoL36|tp=H4%EGiDQwxk->0z6~mlaa>!qJnF9WS7sJityH@fQy#x*H+#vrFVXlIXxL z%x;rG&!8B{0;W=t$KSpzHYk?#@z!vn1fW+UtK@B&-KvfOMU%7M>j)mJgvOsOv-4)P z`N#X=)pj3p7_o7Ux}8S)>Bq7&BL`ex|K`gsR&sM@pxvp91%1?XLG*qc1gQu ziyBBYLvU}9w0j~sVkf^>ct=)4;~k09Zp_~naENcX)37&aKje%HE>F!rPY>}#ZnJ~S z&TkKQnuky<@H6cK8p=4%SyCdeWNPT-kQk;%zYmI8@5R|q6vo24vHJB}xZhrf*}Z;; zrEU8qzvS=yj8jeF=nwXhbO)Yls!XPuJrq+!k$ft0jdL_p#UGZW%&D6>;?yjx5ckcv z3oGOMWQ*vUdbM*Wzn#-3Ua@GzDOK@Ml;VGvj^hlw-*G`Akz4$$3mRykhv#OAD+4>o zPP$)~Mz6NdO2}#~KV?45H|>FEc3A$-gMZ)tvP0%P8FEYKWT`s&G0w+%b-W!OtDN&x zeG6lpH@mkoJN=H#*y*nZ?DSnqw+#y2?@=2*56IyS!zKUn&-_-{Lduae~!1 z)^T%&ZiRw(%`0V}tyOT`-yhQ7PxrAssh%~^7eML{x_wYP81?t^mO}5bt5$&mwZmL3 zlRKx9XO}#;1-!9`o-*-=Ekk+7KCh9#tx8#J6H60iP#WCG6Clr=?*F1t*2*>V z#Dn2QPrSK9ciz0X=@rTNStxrt#J?&pAXh>R^5jqr z2^<*PIN7Rf&y}1@uG!EB($DP$d7EAc>kdg;73eg6K7pz@saCw=}9Vovboup#@8Os1H8ihSXPhw#t|$me`Cto1YS`navj(YFJq3X^&lGacDo(yj$Hl&Vho&=I;Ec=s~tFSl5Ju`woptG zMG~n@mj#D^NLX02w&_pI!n#K?@p>qEPGy))%y<(Um^g+?~|_N6QlwiqFUX z#a!Wm4JI78g~NjO!!bR_aWpT3i}|}4k5i&Dx>2{ax9tq?er&XOC^7KBEs5}y% zBSfYMEfijR`_GI1!{}_>SR6l_+;Ctw zbhQa-#RG~Nq)0y%IqY5)Y>1WS3r>OesoAgWqffh@cKyiLpxN9n*C5j;b~NPDC+8MY zg;Wi6QEdDB*f-+e!gLIDhT=>!1Y8ptGtqU`FJQ1i5Y{ne!i!LUk|sK&s-YoZvXO@i z8$F@yP48L|maq3-F}E}*NmQ{|gUn4@1-iv;AV8156Eb6`y;=K>tRGxmTJn>^Zz2~q zbfNB+)k=El9bb$8;e{pb|91WztpYPjX;4tx1z{6SkH=9LtpdeFA1|RW5)7;ryJcO9 za;3vESFrk0y!KY|)UUnl%;vK%0e)v6k0J0MVdI=E;08R)(IkW9|VvY`orrYhF&0xOrd+Y9UJ#B}h?1G4$#XqvY_~h!Gj`-Rs=TKrq~n zv&qxnI;+DdPv#WUyY08fYHoQg_uoGJSM&J1!-B&?#626>g`1_A+tVsKG3D!C3GbWthB_=PEOdJ!VQjJ-WiVS5NU66W1eP8cfSoiG z@nkHm1jUoZcUbM3ECzWGm;w>u2-@ zayRBDPsRW4xZZJXmJZJEo7F5$mS<9B;$(STaP8~8wBda5Qe27R|8c4{!tZsR9cCuCo|AEH&}y_9jfwETh}f&qRQ=FV=~$Q%5-!A(FO1Q|NO1y zzkl-nd%qJmP)r>~YN<#J%d8=n#bAz*=>lb+u5ovGY7iO75xNcnj(Qr3N}rEOpV&Y? z`XQcOJ9UWUi#qw+r)dm&n46|vf}%JSYr!X_hpd%vcgb<73O-KqNv}||$>ZXj4GjKU zO}h1B@mj^YMP*?R-S89GBhCN?yW@^E;tV%jSk*PrVLDkheP{-%_u}bJ{_^Qp!cqd8_=Ds=Ng(ItE1Xw-^on5B zf;w=&JXxMH^C*2ofWLngx0pAw5Z$z=pl5e9u`+vKK#!Fbqm*loZ8eY4SO{)(E+_)Q ztRb>3;4ZyfkS{W3`EdQ(!$a<;9uGYgB!)kpdQV^cW_H+OJqULA9maYjt3hzwIM>b3 zgJzj`Ejz3`ZJ`fYN29d7j!t*M2!8{}te)}*ilypBqfQqIa$qORF)X7$m@My=9^szy zpOgx1*4Wh>gt5>zvB%h9cl$Q(+kabBlxXH4^@M1;>N&75lUreb>FKdX*c$wdb)UVD zZCbZpxxnZcXZU4pCdIFeWBiGUV_Zcsz^w5R75S&D@;~2`;e5E8+*hs&*5t{v=_+Vm z&KIGKIj)}IlE10;x)*{%cWuBFqZzQq(N$8TmSwoKf@kpSiOJ8Eq~O8--aAess!`FV#INrlQ8K(d6nkAZ%sET!T;yOq zq@VF)?1AJb&$JBbs>O*v3VuAjB=_whMX!u=eB1;-l@xP;BKxUGL;qqNr%8sZa8!A~ z3W%f1IOrzGrSm*c^%FA=(18e^OQL)ds&6L0xby^8dp#_}wy0 zkqx_}d$b{>MimDHxqE$MJw}}-ZZ5L{SX|ykrXPj0f>oO)f>qn@#VRDRJqCvuh-Bc{<_dKy38%c>zvm~w>(ao za)Wz1uaQ0!P%ciGq&(?35MtZVTX}ct#$RqVcTgNQW@o{3TJK)3LajOMJcJg+UMa?% zHCU{hEZ;i|MPMe$XEnTYHbZUYC{4&^mZzDqs@ebY(^=nt+3TsUK#^P(-O0ZY*v+qW zJtyqu*8r_kS=dn*og^pZarw-o!h;PsCXrx+W&7eepNdakc6z58Nls8^P$z%0cVGA& zx*UiDbLc{^5m~GH1Vk|Qgm3rga@i|PAo1K0SqB;BHqfW|=k@z7(x{+RzfInCJUrW? zZjtVytCW4@jM^@Ch8<+=>Q9*1Ay)Q;`ub1*^p}^NixQ}?-)Z7^t45qEgS8->u|t^y z;y;zHt*+WdNsCr3+(qvdKjNKGNRVi`Psfx+T8JUA&n!ua0p1b#&=Jc`vf*KtEu2LipK%=I!IcBx}Q0dqHZh@fPh zR*~mC}kxP_0zMuMg~!uXe^`d7iz}HO?(+{CP&LRh0XeyV}bv z7?CqR1f$Dh0eTa3m=%I4amUV6!}CQ~ zuy+*-wCoIw@#ABH5b*rRu)@b{r%DHZB``+SJ}8^|KP1+HlYaRoakA|clY*{MZTV}bVb?#Fl?|D4;7c2?h8m|Bq05p?vd4j^ zn3E<`Obx|=2=DIIG%J8PKNpfFOSM`fx7WNus6B11{q0np_?KhR|6P2sKhnJw*;v zkq4Dp)v((=*#&N&TdNwXBCD3%lXZsFi3iC}RpAm%3GiiW6?NQ#@S~(VRI6G6O(NS} z(iK{6h3F16kX)j3g3@_sB|2G`EZ=`)P^W7w@08bOZWly(>jR?%2gB>QD|wyJTyjsg zl`Nlj#_O8fT2Lx8{5(yc;8!WvF*cydj!?2`(ZmW$)DO~bFfV&9sF2(0RU?a;(=P|P zfF{|dfND-k;IhHMG$>O+KAv4r+21BP>AjgZ8J@0TN=T*%F?7!T`lrA9m3S4!te{Ag zMdM-TMe-Bd#0ng*x$T@f+m;#?$L(h!7HU)sYo}wfiy#hyRVmerz`^O{_lcAIYUnHe z{ag$tu9J0fAg-9|*GFH60$tE!74G3)LoN3p(cpPq=msWfaaC}a{JOL-!LeS;gNx}~+`TTxq6ub_bHnKpfixsk1AFKI9c3L3Ez^h%RJ(=LbB;HCM3J1$$jF)^@Tlr<;DYa%DaF~;SUz{wQ(kxhRvuZdz4 z_I2Qj1q+&rSAhicFIVOL;uKN=V&9GQhzDqgR`D>fY~m|&Z2{QRtFbY#Ytx~J z-T}R~WC|P@G8QrgBM_$1`E<#(s=LrGI4rm*XjQjxust!}FNZD`W=~Qg^6{{-K*r-+ zCgrDJ@*B&|+k@B`69-;YTWB2WRI~$eGc0;bJfTAsCnY40s-kzO?|~kCv8;%}{Ysxw z;8fQ4$L0d<%Qe@usmqq-&KTw{vuD(X1xlVa7shQn0_OUeZ@zByVq1Q&GMiL5lXE7{ z;>Q$olp;r{$lctPOsnh!i3;ioYT{Q34+nPgZ_kJl#PWurFXoa^$FB(-74GmzcSQ-) zgD&+T$$%*yyb`L;cqxGyK;_aTjN??g=K6NYGTg3A!`Lg{g*ly8bpw=E--aGWC`|#W zc2tTfp0?bj7nptw!H^b@jl_mXs?BwcV~orfhin)lcHX4JYIGb{r_FXXV(AZ>v%e#2 z9C$^PV*>K66a$gejZ|d6eAGXmi{hDouA+f0mwOhLK#huY-Zn{@>`#vm*#bhAIALcv zSU%i_%h_Re>6N$NGnc}%pg0#5^f02=dl!9s#@g^IezhbIr~+HnUmOE^!H7FFDtV;Q zpD0oxUb>FNaqf8Lb1Pk&_(N{XIfY?&-KuBWag5lF$rmQVuw}zbf5ENn#>8Rm$l0}C zm&|3H*?2Av?9*nNSe;~w*+`LiDl$1_XhEMFQm*br%BUs7ZrPq2|7qLbSHE$``_$wX zz5v9mhW88LIMJtXzyIz3c-hm@{qjmtOXwZCk3I!$rhmOEFNSpViXhm~=cW1RFxB#6 zd7J%9M8kBt_|fzlqE+mHZ->+?FIJ1(8vG6>O3`k>=Gw#KTx*al7f6_%}9-&fA1yzHGo1zxv3w=!qu zKmOUGt`goS*Dwi8w<^!f&-cf(4I-$cO~_RO*RQlH=s_FX0szNte8W?iEhPaIhW<9G5X)-ozecHxv%*Y%GbZ`!`#Xd=9)? zv5@3za>>-IewQm(25FM~x)cUMnq~XEknU|mcxr^53ku9|$O_{xoVO~>6=Kk%}}k~nZ^)q*9#+TK?6KG$8G zI@Klb7_tNE8V}7lMVh3?L-)B(et4AyGHe=Ywm0v-)OYqRqaRWB*NA}lRQ41+dQ|^czm#K=Eofr@)tM>@@2+G28 zLUKaVM7yBZ2PC#nlD&!{UX=IbdLh=We~p9Ab{uWG-ZBWrSo6GgX0`KQ%}bLUwjS0( zu@PE@LqF7@(o>Xo6mZmI%om*uIFvhQH4m#^GPrBQwTf%p&bc8d~}aE`jL z{_?XMobj6qcAvxcYE~$j^2_KQ+2+--FBQMFg__rF6>?hZB%O4$V4a{4iqY0UwVEO^ ztU#$r3@a0AaLU;cj_i^do;~tHs*t)Q?Bru_A&PZ?Bu62gJ7YU6iS~INgscvPWNUeN zW};Cwei)g}J8W_Cg6l2AXw5Odli@}uW=pOto$O2EK&1zlS$HeZ%FzK5~(xL1H*;j?Q{(FLifGzNnTn>T@Z$MS3`J zn0qPM2HZ^;#LN(9*nI(xo^$%{$-@Wc8-eserMi|BzA`{MW&)%`6jMQweL%GUOXfc5 zL+LOuYz)f}EgS}x>ofd%NdvT$wW>SmE_t_XmD?G9TkuhWybFeHAZ$?&yOoLt!?0vd zCn;B^iH?#Tvp{%}li-p&Cr6dR#bfRMiO^r)E6nom7Z--v02tFTn`ro$K7IUqi_UcX z*gOx!rh>_V19lcNM76#tfhC-E5=}n%I+uhXK!l7>6E@J+4~2D*u>N@yb>+tIPU|w) zkAFoVVL@%aK&e&WinrMBj$|-k!|YygkTtux>A<>Dy=bzLVI3@B+L);Ew(>b|eJkUi zrN)Wo*x!A#oopN{?Bc+F$sUvGWGBU>QDi$d3uMwO1by_zH!h2}Nro2Wsmd1Zlr%&X zlDpE)Gf&f-XQDQ5o<}JMYImV?3#Am&=nH|9fD5akVB%9)Ii5CLX4$R1IZriW&X3`+ z>tzd!7pG?oFTClUEI&av30oEoNRqij;^l&k{xxK`Z2#PWSuN@H zuNsmzYrtz@)){pixT~e|B5pLvp)it-%Gu}H5onIqP z_#t~0r|AO81mS3WSTO;P*^UhWdFocnSQ#VULcN>1^Gkt7BU9R20xy!CH|xMzqIwha zv!7ys^|pkHJQTQ1lF3b$BNL@w#j=pbfX#8H(nm8W9272BFIQ_&Ocjfn@@Mr5alJk2 z)Gyz_?`n&Bb;d*7EEy!9qO6#8PA}71^k$m4M(nRHtbA!zvb!D(^3uAnU;H-CAQy*gbJ# zu4ZLRrWDD4kzk&2c34DM$RY1rd=jb)j=E?vmC>`2(g(Xlp*|iU(ldRP+xkH054}jz zL^l@}3v=gaXVmc^cH1TIAbrwpARe|h{HDJqW_CI71nmjI9I}0~0M_7T%y0kG2PXz1 zCQo69m&P-a-0pBAPB9v10=Q_o2csex=@X?RIE zvUi~Zvrcwbp6`()x+}jkHh%T{4t+*{(e5o5vk-YFd0$kO$3)*F4&REi#+lHm%ET5+@*b!)=FgweUz* zW`vsg$%pk*jV|oN6vu8A2*p%|D>H=!AUYWot;y}lY`+y_}Y>_YAn zCx;}HstFAzClj0!-fMdZ+cCzahYN+ zQsg`pxs8)UcP>aE33}0Y9latbT~r8Nqss*-pRVbZpHlSFiD6CjDH{2WQNta}7K{nR zHc4qfRk+?6#x+uEga%daQX`6HZ=04H(WpR=C^w*s0SbS-Gy$q)E(NEDz%EqT!Gm$f zKwG(As`;3dc%m<*lXTOXlv$$-Dt(aM8$~n0;m%SO(AnI4S72zy1=EWbO>7bNw5Mb9 zwgo$dhf@-!QReY{hjn@_MCgw|*J!i=hf{;;3q@_ftzPI8l&6~K5hN!SGe79QHZB=nEw zMb{2Hdsrx=M&X4x9!&8%s9H)BL4f3{XR|bxi*>NQ(zt0^sx;B0Iim2M*^KpM=JeQ2 zBARolUh|BeYsH-ZSxpMY%Kka)S7n4$%3W^%(G} z4#Yit&@Cl8Gowxy+zWZhQzr;^tPu6Yjy!L}f;nJ$*fOFLe(~4K=B@brI%L2N7bk6!VYd#I9%=>3LP=TJlBX&LX4HD61;GE}R`q51L1h(GJc4>q zK&!f4luZx2HOk`sYJGRnr?_X8?V=NkL7;WZrL`(7Y-^jl!{_ciWWw%}=RqfJ#>`HB z9_JL@8_+H?oa>Khplj$MNr`7CfBUpL-gcLmIS;(*c($0HaU$k<2Iq@&?1A#+4$JoD zlWhlG&6hk5>&IAH#DGm`ljMmt&HOT7)RjByZOb4&d24JptQmc_&zy+sEsy>fg#}Z7 z60iBmZ;bxSZ<~KMhum^t#>aK9J>6CRkYXNCWRQwHO{eo(6k1gcNN4Z$g_so*$6|y8 zCJvZb62tQR4+XA*V*K3^Z45XjP$%uH|O<$VV8o-$$GDHvd0yq za%FcQFZlSOe+I{e)>Hw%DOSW`Yd&^_^+O~VsDmV$AyTmz8%6r{A8cq7Yk(%QO;bzg zSl(VmC%2!wfib8IX=)@opH8}eL9+ZL>80^^!{1D+KrY)+r*7Kt2PN>J>kZatgAlz# zze&?cpY^>O0?h^lq*|?k-xtca%wRBfkPcPFw)ziyV^^{Cc zThmK*PsZ-*o-R;PR76m4Cx9%GMPyNNVYNUgSFk`)+&~bO!VN_#D*Vq$LKTTb^C00L z?Wy`HFYn!VgYSOt+;h+U&UfS|W~uKmSwW{nYW%Alk{|5(K3f0yEtv3O9-S>tS$_P( zcoB{p_jH^3!#uGOD{1@MF zU3!>eQ6cbN@HmCe_8yp%#y#nwk*-nT&h)wY$&epXF*lA^EE;gay`^?04tP@PRN+No zno&xuhT92wuRY8%!F4`tA?sWtg%y*Mm_|3OpEy$FjRZ^<%;)f@pM@F=*E2eo&UoHg z{%!nBRjp6Ge5p?*WDTbLIN|Lbbk_SjzO#417S0;yeg*`~XuyE7CEyVRYBot9dw0{9 zLNmPkfm*N|)?7=Rqb8z&aUGo}T`^uupCA_`QEnP#KggcNjBgBI^}~!;ZoPi)dqv+a zd$TEgqzEt{0GlJi{0m0-nbkB(6@L8Se}88*M*H4#m60w7UW~6aS&ToV6nzxAPem9c zR&>l3eq+GtfJ{|f*cQRI$%l!7Sh7T@69A8dUds0Z@@&A`)>^@leDQMo0ar=2laRz#yb=zU899Uu0LP2(=aZRZQV#VV`?+D(`ESNQFdeD3yu z7-9vul!s{&tDxb|Xs)oxT^~NEm|HB;(y3xKuH4oHol}4rvpM>Pw@R$fvO?cDop9;T z%r!mOY(_Y62-<=INEN3mu#!Ha(x{@xSJKhKFJ;$fpA>3csyM~IdwoxOEFa$-zKdH) zp9(lDsu%P@+h}2(Z)@P4F<^g=5K|+xQ`X{#m!L?rhr*I0ages-|a-FL> zQ-wvywKICSnW}o&{ZhtmGv`8YC|z zG5Q*cuuBdbONAWR_`2d=DXo-7`x%oFw)otkql7u)c;3$7Zh8^2%n$mWkR)+q!jPF` zHydV+oqC$5Ya_GtL^rDb{LAJXzZYccvcL#TBkOqi!M%EZjh=3kLZ|ra)7=uq`{$*? zhH1{2R$^s^j3>s=ekWQtZw!9N=<;5lQ+11c=D;p*r-{pZmQtLi$mbaKO9tjkhkO<3 zpzli>mGL~>@lfY`rNI8P8tZ_s@iK&jfxCyPAN_JICr7MvEp^KZQfH8BFd_BG^|>in zunMmxPwSBC*YiD4=CfYZC#@jRG;Lb0a3e=uGUb>@8`SbsYh2suMp=)%+3ytescz$U z%PQe3;Y(3bvBK}1|0%jDXtOGgv(`)9Cv9`Bh3)SY`kLFbYPqsP%hL=B>uXO(8Y_qS zrHJo%z5l!~r*^oW=NQ!8)$u&cajT^Z+)>*QJcwyUoQoVoXy;j}Lr;IKEU&Uf2g~k9 z+?*XcKKme}OJ%e?MVsqh zG&%je3LxsKJ5~5&zzJnB&oUT>!-!ow)2=*2>g6XU7n4d}52*I`^T0$TauR{v(*-*d zYrPui7%F4_hG|+s5`$F*KKo`ICx&WPCNW1QuJchBx;)EM zjf>e(ko_Gai2R~w{tG!g97WxMqj&8lz&cGSKBq`M6@hWPTzX|_0uNU%nkk#0$ZQoa zla8BMJTZkn3uz52-7;a1B1ep6Ib(T=UTc}MX_@qPs{4lnKdAb_O6)A1qZ|Cx`0SvU<3NWG|K;v>m}_hRdge%lxWkuoaw-`R@BhV9kl@m`N@;uyuN9 z0-|n8af2eAR79h)lKyONnLcB-l3z<76O~L^1#sG`Eb{LYGz3J>N%Fr6TU!aUiu_lF zrcNDj!qzKz4Q}zgZdhgG%KxscSafLm1yPH%UQ{$sogz+w&CfiK^`RQ&XYza88-iGu zj^H&4yxb=pa59jVbHu0IPWg7z`{@i=hT+N{|L>GPB%jS~VUig9Ws;}unYz^?0R_bbdv|$`s~{ZUvW3 z%MI<}HU@*jD?d)$B>RHd_)U%SihE;lC9Mlfc3;IgZqGt*qeIqkn6S;cBY)_qo@TU7 z%RZVMOHv%zHkFy!rd^Z*xINjJgO%;Q(ybSa!B($T@^;R~m+O4v#F1h9=f&|JkakWF zw+be|=I~xlquWE$OHydO1}svzibn<9b_Vk#K;oCP!l#3OgFmcs81(}m9R~C!V${FG za2>F+E>vIZosZ3}i-p$`;PtV~r>?C146%%PQR2YCup|j}a1Y^RdMBU)d z#&OW&MG-2z6E$}2GZ=;kF{&XM?pfOC9IGL5T=LyN?N}xpZ?rDq9UrMllmlBAkaHcv zjZL5waTHk%y7Ma3NyNMbU|NWy!jB1azzMM@C@o>gDzI9**^cK~fpwhx=knjaWyHkc z56*u< zN)1}L0%{DO@lAy_SdaXMXcxCJTt{+eHHDW*7iyG`0}DCr%n9eUVW;SIkOO#~WXw$U+2m;&X}!Eygk`>vRu=8{IB@Wr7ACk}a?9|p4}f}L^a^1&-46Qn z8f6x@(EYJ@Cx0O;EL#P_A@1yG4%uLK-ijZ8zwgVbNE`1FPMy=SZ5TEE(=)efjLxK&bn`In8s?b*gU}D+Oiq!;qJHBIaU^j z@~wI0+&>n%J%z@#ZzHVM;{uXgv`~0QohIl32RVg4P42p*uLbFUPz-JoNHxw~IiX{0 zvT4}BY0z%i3qnKt{p)wqC;nq?X_k96aFDnKo87(SHjR5-NcBP;fdjLUA2G%r4jW_) z9XWd+fA`V3PiGltl|k7OHlY*;=4n|FH0huXf_D{j!DuvcA}OTzrqCV1xFk#T+9zn1 z=lXW>@Qk|5p$X$_!;}>t(jQlAM4gZ7l-Uo6wH8bFkknG{5+>ddO1j) z9@OlyOjan*QYB5W5z5%0XV9)Z>!&S{CEr{A1E~=`qC2ABk@(?u6dZT?M)FPklvGLq zS_50Ch}Et&AbXQ5yeB^miBcfg*~MEQpjYvHAYJOy#^0~Z7oh+-P`!W=4mPEoDVus= zmi=9q5dy&$WBt@VEc+`r_iSi2kES@R(u9RD&$92=gZ|Np$vZ$%WWQ^RqFtWq636Qn zVL6L6ig;d^e?9*q^SQKLei+iH7N)wd3|(kww-e~&Ib>m_bcgFf-wONrwj;#FS2p6Q zgFA|gmT@smT%Y>4Bj?2CZi)rU5@wm9)Wgo;Vo8(i78-CVq;w#cMGJ>oP@b~?+c@OG zuo-|Mx7q!W)c`oIp%z4a_ou$biDpL1CkZ6Wffv8kCX>ouN>NOa0xIHBSoh>)|BK$Y zLK3}JagNZNeT)1Lll+&9{M#Y^R!1j@HDN{m7~80xeV5#&cX1PVSDC}4Rg?zQC``iJ zL1Q0}d3WW9$!^JU|LY$4FRz?=pFXbK<8oJGW4yCL^jCId*aO>woE4&}(ADSO9L>_$ zfrD`t^1km)ui|X*j^)MswFX83F+D|GIjNsl7y3YcgTIrzNqWKeE?FYIC%DQS;-L_h zjbUSHu-U`OQct!$Kl`KC{@KrHSw>P&$$7?Fj+X z;tKmxzHta+gM&xWv@HZ5E{f z1)EgdhsL!s5;`IB275$EO+f}RQla8FoiwJT=v+JLHQ<^7d)S0lj%f(HJx9sId-~{& z3MfzSdQ?|Gm{jmPb8V|96Q3N|kXTT#>X+wAI;ESSLU4`WNe?J1EyY^FMeen9wMTP! zqgYGd4ZSjUH7#I_SRr6kW6JJk%RbBB*WaD{yvy!I^N<3kMy)%2;s~nN4Vnhr+{3`h zO?{@VTEp4ni=+t@w8Kk>{6LJ4drXAK>YK8oByAcNz-VVm`N)(oY!0?aTS>W>I+c@W zJW)rN2P84Kz7Z!*oQnmlHOiBISnOJj#9>U*&k>?j6aF5ZkUfXO)=1B5qp^Ra6c2ti z{!Hf@1kcOqJ)YPS+v_{Rf*OX7fKWxQ1xNd3(rUZ|plR%KM5>mSDI_ErT2UyskCvUdn8Ot_kIADzb z^81G+M(0NU$+sHG+84&T*=ge3q)>`&6xob=B^7hQg!Inpn2Om3jY_2IMysy*!zMwT zcxzC-yj*lezjx3A;odq)4bO@{>!rowN# zw9)O>H((NyH_g*AYRrOcB_;FJyC6knLs+K@U0WM%;#l?US&p_!6=R@6l z=GNm$kJo|s4lURS-0JBJPWFXzZml2{+|Ie+g0l^ZpmheuarEl-oj|2ShG&a3VRBjU zs=5EjHo`J$-WyP+$c^($%9Q{5JxO9Wzc?^)u%L!4+-&EdIFnWqpo1=8GidMA3_BU%1u ze6ySzWi8Ajx6iqE-J8RcV0p65`B-SKq&(ob|K-=0`tSPITCO4U?KW8&wrWzDqAd6@ zdCZhSPDl?og-&+Q^2gSY`9@>QEAE}JZP`G6B@y?)&_R)|8EsYQy;60EltW`p{6Z^%PVJ_Lb^Ro!%| z0{duG-V1p^VtM^eAP`BHh~qd_p10?0k!_mQ7yxA(NSwGTxKC2&TkY{6^bo&ZriD5U zTg5L(ROZ3>?K6Z%1b$RJ;m;)2ff1NvGL3Dc6q|9*jlk$xJt%B&CQ3v2~xSAfS|QmSf@!Cn%cJu$Hc1pKolM|kQ2C?Z!Jg3-MSUKaNX zknuFi_!&hED8HgyY@RfUS?e(%sSN?zuMEm4V{Ng$9tAF)a0hk+#66TY!VG;_%WaZ& z1tNnFMbEJI1`mDA9Axy~CpzY|evFz5d?EJrK%&AWxd0#zh!T zVo-WV-a#jjNS8b+kIGaTWPWpf^MkV`CDJ&~G8vw?hu>rqY*;_!R=2c_jbXf`aqoU> zdxv>mH5=l%124=hWMN|s%se1IG>0Fi3&m;T{Ztiw*t3V*B-47PiL<$Pm^BKlWr!*G zJLptlEN_Q&Kh>f@4gABNcOg&vY;cD(#VcP_2h}kTbE~=bKo`qR(BVys)tF(^aEGk$ zf_aTghh2GDs9JfAQ^mne#2s|A@`5Oz&hpBZ3=)9K$Ewkfm0@#jFv1CAwu0F0RyXRpO~d_!098ihRmL)!8cim>RT_fkLPd8eEjE z!h|-B5?dHfsj)hpAweZpktjZm)!?Qthuvy6j)|pV9wB^&Hy>Y}we)S)pJxZ3Km7h= zy!pcE1(}pAEVJ^Y=Vvy~{b=C^K_&U4;RA#QnJhXAJYBtd8r?WI+4uVF-ST)45w_2a zV?LZC0F!wacKJNs^Y%ZRr%=Bj@UW0@eU%#}?1TEat&-g$i0%Z$aeD%jCngCls_up& z_hzT~{1{=vIG}8xVHXTo==96Izil4PK7#|3IW6onAiwI0`@K0QoVA{DEa@zPiq}?8ymG@OA$Q{>QSXb|YVoY(Iey5Beo!~(5ZF!kXoqQ>|D#(Xb zc!j$Sy&GFW#9(RIL&4G=mi^(fAjLbQVN^JcF9wG?R&DJ*9}J$QmjlN7zB5B!hOwLEjE(M=_4vvV|OEx43a&7xt`)WjRVIz^ARDBDOi76lu6^{1R|v zH$aiDNMR+g^fr?(70sdGrqZeIRm#n(df95y6^H?si=pYHO|(aX(jW&^JzTt=#f^8{ zC~o5yKp&r6VQL_O_9$%-mG7OslF65!a>M31UTu7A-W-lqO+m&!EJ3_te2!PDd(8Oa z(DO4#JXKkXJ$9?(uUij+Ijh5?KkzJ@VnoyTazFSD+2+7A&jAyB=XyD)5-M6`OX7HHl$CFh8kfn&GF^v3 z*ck~jb}>!10RXG9wW=|zDQVnE$x+F>E>{H%S*2}N}ccP$7u-^HKyN*tmK>8$*d@oDl z0xDIt^bXe!x;t>hw&R)OI7)0iv;UDcc>B(RSaVGwHp?9cUKU%BMncug3Q@kO%mx20 z5EclFLy&A~C^&)q82h1MxKUz(jNxv#+qrFXXLe6F+L7L0$d;4s4m|BtnplP+N&y6g zJSrmo`-^uih<&BZMSpOA!A9|47yoY(b@N-8z4p2LW1fy%OCMNJPTvc;Cuj~os^|ve zfq8AqsLSrXZV=3?2QEqiSd`wm;tV>Bo)FyH8sI>y#+!3R4aXm9;L3Uj0yc1`-QM zB|Y$3$AW}e1-|Wb20#)K*F=%fE{eJ@zcy#N0NX@PfpTxE3z7ogEH1XPsVNycTq6g7 z!L|(T%a+IN%*ubfb?}&ZnB@iWQY?h(aF-BM?u$hwswG6nJw3S?iczDV@_(VaK^%Ve zaqm8Jj;+HoCM~dLBZWvQILp*2QD&u$pFqKxpfZ;v{m(SPf$5o$CWJMMc7?S-rcsga z6;L{CXKHwNL9V4n+%0Q#gZv}U$KL6#WiDmBN4C~EG(POn+7YlOgTrRipX}_&2g&nV zMTldY6u69uCXktVpMY?D3)iqc*Bfqa8Nqbv)! z%WHvJF_-yA1(os)0bWT8dd#z%kMhJ=utdib_uK2ld)vnSjWRBYAG---L9_x8Lz|D;@%2v@zw(dem$oy= zxNm2YxMfQ9en}UTsY-U&&d{i0##G5rKf+d^LcfR^GdLK(=#7^@_^o+N)nWZo3sKUg zuV0e1D*ByP`t&;wFK6VwY~gIOBiBh|EZd{r*xaU z7f%X5J%3I_1E6Ka_?7dznV9kAzN-|eP;E7p6Blq$by|9nX$-1&#$1bTSuZnos9}K^ zv=WvE*yiUgBj)GXGbZtk?!_x-HhxN0IdH@sw%&#~3|lD0Mv5dtj`8#}ygNU>B-!JQ zrRy3wNz5@&Kra2+CCN=0)+FoYBzj$vTpLU4!43#EA8(5tS_Z-}Oe0Ft%ZoNo{bRE- zabVxmLZV=WUyR?jiEC!6v1UBR!WQF3FB+1HX*u3&z0^BNJL%w8Oe+r2(m-DXtQpq0 z4KmCuy~jTVkr7nW%>pNOF(z%GGG(II;@c$K`?mXSU^%T9o{_hc#VAg;GA*(MG{gHaz+*5T}pTt5;)Id~=e))*>jO!X#q%-Gu#q%&Z#|8su^^-I- zv~rwT_B#6Jtc=P_nY&K4I~!-5+HX|;oh%zqK#p&SVNawK2^5K=B3h*Q;1?}glnxc1 z+JXw8l2aT9<%*!p5UPJH8q2t}l~3RJW{v)-TH6D=rw=+n#N|&Bzy58+@BaI}-z@x- zWI3f+N|DI1^F=KDyJk)L)oaiD0w&w7h?6!YRkB->C)Ic~&W#kF7DYn;fBy12KXfzBEdN}Ue1@!d;I&Ml$&{B)DN-qtjP+cSm_xn|vK-HgAO(uEoEqt3 z=u%=KcRgpm9kkkXf+C-Ig1aXg&CL3P(UQchjGITg+3}ejc+YmVi8|?HN};3311bV*Mn(HQ4s4O0o3Dnw zH|8X_ng?Z+kzR)U)D-%Ne5+7h?z>%q!hJh9aU7)6FG>wU?Ex)aBEyyOF4rxtNeqf( zr*hzCtQ&)Zlbv!X&>XroOrK*8ImC1#lq9P0%ax!8Y9)Qid)Gvb@|x^`zAufkA)svP z1yLLShI@~^S=B!$-L;jp1{hQhPKi*(vJI+SbpscncXn6c{lFT(6K-wX`+-pt?+2Dm zRbvy>>nY+~L#}u+?ZV z@s>ZO6hOW&LvK0R9Sg&^NDujz@nEuw5@tv`K!pvnjB~_L0}bqip*&&%_dI7q&>mk5 z&#EJZ6_ZN+a8@hzQ?K9^gDlSJfD9FikPbK@X&U8q9?}ndQ^dO==yZl?IVZ)@9^1Y2 zqH(Ljw?h1BvdX_<6qutv{uJw;!JRn}Us zDo^0P&PZamPpa}<>$TSf3n*6e49ClOn-z~FOPm{mvxRw_n=hxr_e`T9uVu1$Ss?{Ggys@^LT?6iXNemoc`oxNovr@ zR&S#P-muol^yE|EW%<&X_v&^%?;*fsYV9wR|26j2vu_@IC6_LwcYES=6U$5V+VcGy z?`f3F-%1eoaPM>L(tPVsM$l?Nnp8$sE+6%yA6oND}xSwAQ_ z+akpv&wx{-a*H%Zc$K@<2g7o7N}~sCEV;%?dSgQA%*6i|{c)hxMTNswFEsHST-O`+sjs$wBW*S{DXc0MaZ~o=2(Awa`x0b$Pm^Q@(u+=DDv1jNJy- z*SFdI=x0AI_{7`~cG&Q+1<$q;M*71<>cfvd+(v!)$L8=4w^ARr@dxIlaXpS$#jNrR19%oOZ~Vzicn-M3lanw90CKAO3w$-_7c{o>@fCbET{ zhvUH8oBK@WtQ<-KrX&OFm{r1-A(}9%{+p zex^pbOr~)ia4K@|pplKCQR1A1GM08j7CTe&q-$b5R#>)-|6uv6=K0GO5`S0lia2$^ zf4ep(P1VbR70TpZQth#pdr8t5e*K+#(4=SzU!!P)y}b%vHTTe1Wp}cFX<0s^5jJP{ z(mE`2v+vz+yz{)%yc4}p5C5#_3`a|!<1RW%(9uLUbG55|nRmlL$7;9^bFV#5v4ZQk zUpDg6p7)ULI+vrOj~<0}PyVCf=;t?nz2}YN|I{4*>06g1;F0VBsXrZk&m%>AR2bz` zsVWs!b33GU5RJi`GKHm6wD!zUwHhc-&K7XSeo$cr6zX`)$CtiibVmH5XZ{O0%+48c z;AKp^i8FGVQhZL4dMe^D=_C8;OA;sy@6qjn#Fi58HG)I@Lig>koJr%x_!Wa!vV>^4 z#hk|Qbf|5&5!%G~bxC?c`kmI17^Wu#V=#Al^|G(8<0rg+mlyAr4-@TSQsS=Rl(>K0 zB7vjz#5*xZ?xJL)0Bbd@h3_v8Iq18SbVHfERPjmfWwydlql4kF*tBqx9c15#Em}Xx z2rTd4mMtWk#zQT*p@}9tCJ)n$Df;jgl;Fx$8V9=`lOSyqCS!V`ktrg)S%|@ zK9UdDxw+R)a1o+13RNArx?JvDZ%*}dux>%uJQ z(Oz;N49op$?;^6fDWdXY;~)E*B*)(<_OEVxPjDTACY4trbYu|drnxaK>X4#rtJUz zZ;e=ab!W|Fa>IeK5@W*l)lrHE6uF00nU5(A6a{@47fXaCyJtem>ZjbtK&V}$os!_1 z;ttAh%3bn!-V(*d$r@#(u$=?ji#pdYd~tz{?U1CqM2MNJk7PT!$hAu4AoI3PvdlHs zZ4arJFAwVBg3fl0U!qsqG%b*ZHOhrK-sPf&NX3nHGpx0DhbH@`yY@Tv%e7F!EK+#O zFWPM-ubh4)7;tKk<#8I7hTotAXvxU?O>sXSyf8kr&2^#LAZvLQewS1rxvy}!7n0(2 zdrqcMjUNVlPt*8is9EN)R9{x$8`rsPZYFI6U+<5+j*=zM8IcdNT!t90b(CT)MPk7v z0*&>9zJ**MeV!6!0FCduFP14=#?}{MH+v1++Da=m?7j9pX&ETWKYeG$A|oh@Jx|C< zCp!#q=2iOiz(Lcz0jDeOo$>}5c9H0% zTcFl0xIOu#OnoNgqGz!@+g**l=@x9AT<&_5tbwdf{DvHH73Z>d$JCSL>u<=B82Ud- zhV(z+WVkhd;@#;t1R7NzsV0VB41a@SIevepDxTLPNn$p**3;$u+jPuVweGFNqy@Lx z4yTPkSNr^1p}!G%*Z!%ih-43EjvRP}R%>F8_EQR2AQV#(QJ@)e3jF%Z{4?$!^^hCP zABUc44DWh<@%A6DdFAGUvv2QRQ1_Fp*Q?)Zf)c;TzKHS5m>(sqqbud7MUN%eL322& z+1&*7-B_NEKI5ws;LlVE_G`06gRK!_D@$hCYE7BAPj00S0J5BkIf#TmH%X`pR|?z*{Yf4Cp`=~*Ktk$k~Jmo{D;w`{H|q(2f5>C=QLqO{^~N9y&j2P5E0`UPCw?w@fIx+BEP5GK2j_; zg4R$}*yQ+6PKm30j3%n~)ph$xh69_ZLna>1Zb|{d1Nl_Mt_4?^kMBeEfW4uKLM`|t zFuSS}LUu}Yf{SE}RLkj|UB&4NYz{x7(g1mHtJh$=BF+xiI$tdN`zhTo-!1P0e;UuE z(C7S?IG^+=@hz}r(O><5*$4GrJ*bq~76ridL5@cY7wlq_lL# zS&>F*GOgI;MKyLZTKL%VcNyB3W%%s;^V{6fS=rNDph8{>BzI@3%IO+TDpX*!OAW3luWJOI`CL42pk8!A;jh>-nI(Em@hrtWY zl_e|)INYC|#7&*5rB|WYOVh;L&N0HsIfg)M(SmKSXq1L?hSDFnUQlNN>hSVB6%w1cex? zLg5O;HXwk3N-#PBUcf0MkHY^^Ft0oyox6Wtx#$X5{&YzR)S6dsl6D0a2Y2e@D94p~ z+zz_KN8Ra>4FwF&fDf!eyp{NS14 zUPjF14ZORTq_V?|1FyQvO)yhHDPRSagVHyK=rTwltJJ7cjk0XZ1lRfM%aS}Ai4|>p z)UZKUXWQg;qE&8d1JoHREuAZjbu&~6T{X8y7VXg!Qb7isN~G0Ox4R|qY_F+o6Uf+M zA};UfP(AM{Qn<`PQ=yk7!rE%7K^VOS-Qmdvlk}x>%DE_dgtB++<~ZwKyWv`~9`E5g zvBNo5v*Sx^UGB&y8VyRztTP{xM0N(naW~jVk%>XcpcEi|uc0DvOUY0#1J%{lh$h@0 zz`Z7P$C8*zy34MN-i2WE1lB%t?LVx{c}q| zku54#So3Cc7*Es@t?iTmG@qdBy9_td`0ekf;3Vn&QMjkwKL@$9k{L! zCLe=ZUc0!JG*V&C`L7Js`c%)jD89?X=x*}Vlj76l!DO9ar{uxp{quHvC%;*>;36b* zXK{N%u)sC;*9ven^wS1`zXzP2+@IT|(S{`x>w?rWTlW~6+Q zK(ZV-+6%$mAr#NOlme>Q7f=ybxlJ@KP-;SuiGYiEtcg<_a?_{G1&>aX0vC;P^~+I{ z)mx3>2I@w6N}}5oG=izTz&=B4yXqIo0t_$dip0&MiL^QJ!q`HjEzzqR_;4p_oOg;t3@Xv!qUabrVaTK^ zU$A$9`m%czu;#kv6ks;gM2%O`3~X@4z2<{e;bqFzuKIoF8oxZC-PMbV=GorVSP!tL z_^Nh5pY!`ON&<|gWM$Ib03?vy~Gdb*D8-75B#A=I}QDrFSk#KDomD?Y=xtyoy>ytt0L7oo-8o_55zpt>Ch0xuH2V>KOH2ITZTB)eN$|Q8pUH^vdaj zDqwW`ZBl?F5hv0=SAv;L*z&MvHWW)i9cU=H!mMPr!j9-;29t)+pkNpnQJFcsN#-!xAIPm%b+{X}H`tNX}tS0MDH2Vr6eiqOaL&`4Qj5o|SJRzb#SA8hrV*7VbM@6Yd= zlek}ywXy|zHx?jBVmd(_V3+*FEZ7+`6umB|yLjsZ)EIcl_011fH!J(3J@TzV0G#tg zJL7|gAf8IvQ9N2AXF|i8A zDFqZruN|}sJx~!xque9E4bsiXE@*?XM{Y{ak1t6shW3#jE>_8j9$!SKD&mDFg6;rm z`z*73d~>*=-QUaUJnmikn)?QBECllI0HUz7CW!3cn6(8`shgo&q-2%3#c}TNP{dWE z#Of=Dd~ed7s(VsY`@9jjd4yFhMx~xJEMlHGX*Tp$eq=SJP@>;B{mR_*IP6}d1!HoT zztfPX*dk34;~x#^M)xo&;xcZ&NWEX&F*SjlUC_f{H5ckI0jlmzUm{!;j$W#kKJ1Lf zrVuJL*>NTsoiVX;+(*!i6;#G)HmwerW;8v^KAId$QeGJMrOd=o?V=R0tDTMYD>1x+ zqBN;NpL+Cg*C=52LE+XmNH1uHl~yZ>4B5nm`Klh)?1qwpM^uSkIoxzteXdd$gBs&` z+-8p`=bgdpC!q6e2b_%#UxRjM1OR%*E#4ctUV7el7;`{iK2O@fFAj+dDBxi@-=NzI zqB5xNsijd^12+nLXLU?%1VN(){z-nvR6MhbdqK2$`h7W)sahD__sib~c_QwB6K22a zTrWdfip@6|Z86cUp%vMh#fc{Y++fAelJFD`kl~aX#>lWZRgrd z7RJvSTVml>yPn>5K$~m6MPMPe{_*0?Ku$~$FY_uw z6aRqIUl(HoJa>oN5v=Z-8ao@sy1S+}@&C4X)8rD6e}`48mOcUDGIg%xpf4^x{&#V% zq(^2m3vSJ`9<Teg#HKbnxYYp4k1SK{n~ zjNJjJ#;i{>eJ~#iKSiRQe@k}gN&;YFOyL7$095U&b$qMWiZLth+nAY^>m74Vo<43)2g>0+PNT%V^^Dh+?K2k<bX- zT>!;_XCA1WJ_J)mlwub}@-Tn`%Vn>|0FAOcuu59&c~hoy?SKSY2>9>y(7DF*GE^uH z)d_Y9n-xPCFF7nM1rJ=CE^zbQQ3(Z6B_)8Juhirg7t_ zUEh?O$55UW_;Fx!V}ZHd$U$woQvqF*b^wDrao$1SvfwWzk;3NibjfYHM}E>{Sy(J@ z+r%1ibNDeRv9!%ALG>ss$Gb5+8|q{jHk`uVPhViWovVKXG)1VOPkH=$APH{7C7DU&MmO|$q^pJV#No8!P|io z`pv|XemcgGriy%(6!F=KnKW9VUK)1=?5GkoXbeUQpTYa>^>cdFFBiRRoRXZxsuGg< z!c0k@noLQ1C%gBg`?oCN}Wf zfLVE$l&bW(i;@0F^!*Yx1yXTOxc!Sg?|ksdpm3xh$NdJVJ)SIrN~Mu0t17f?1Yl=rvp0lk&^#g77c-j<*8*D zZqiF%P7aHB-b-`!2^s=wq}!!sE+|fFBxN_vJ40`ndqaTH;nQyY^3UD-0?&$&{dR0> z90#=C^Ml)<$`Mu`w!K$6WH_vjs13gDCO=I&vhfF2Kp7X+eo1Mb7-S)%CReyMY_a}7 z3~cIiYwG3eeb6N1hp~KHSs7v*Ki{Wy$L`rXtc~7$@ZUC#H_j*F9UrMl)C=Q>K)&IS zrEvnKh@;4AoPDm(2D!3E4*I4AzDtBv)1Q#Re%ip>?wWPbw%zwx+tYjhdg1TphT;WT zz*w+%x98lRlR`(j8Dg~RT;G_Np+p~W?Z?v68HUF?P}q1eYXvg?$kNq|*DnQz4l0*k@nZw&??)wtWwF zyt8MWhnmVAVJFCiFrCO|II|KyCN9d;K3IVatD01cYMjg!cpP>=$AYE~(&kZHKSw;s zsh@HK%tqPN;t+$zz8c9RsX<8$R_#N5eRYmFN7$&W@ymsLADfNwun&9`13zqkwmQkq zPpaA2xx`$?i_O&Iz*Z*P#L8@?6q_g*vxtXpT#|sbkoL>Z`JeaOFu9r=>sBaVE=cv) zac}z6(h1_d((L)ws?;&`x>;h$R#16nKd)b!a8x)r2Nh5FGl_Lz4t|cwcyFT=n-QE5 zJq*^5!Elw1&X`}Nf}odQtLTahSOIaIZO#t^vpHE1moHY;iZ9r1T6o5zWjWw>$SNNG z7i4EV-uXLct2yn+0$&HE&LGFZ;~aO1uo?A^<6lmnvxi)tX>T(<%)isxhd<1%c0I-J z;XCXKH|GxnJ?89rhxI%x@ZqsA>qSWv(=iYYD zc1QgJq@gX`%)cR6*c`qhtVG(u@131Ctz8-EpCc)Oq*0A>t=C?cGP@{rTMOlRmpFc-yd?^7~ZTm z<(vKL`A8l#Foj(J^OhxJ9ao0zynN|8X3I4YvRrBj8hkBy<~M6W)A(o5|OYSw6!d1D@7 zw;Hd*+93O#WQEt4epK@AFU-TKFG!rsLh$vN2rF`Dd*2uAbN=j`DA2Wc-aU^pmz|O{ zLA(4B|61@dZgV(RbzkGP)IXk=MjlDZTuL}Q-8OLT?w?oVy3fupEyqC+qMQC@X;~YUIUeR>z3&l z1Jlq@xm}GsR4hPgESY_k8zsyUgUkffw1rskasO@_Ypt*4R**BkSx{i za_SJb-vgOC$CS0=`;tz&m@Wfx0)3x*J+?V_gxUaFqrlIwpt2RTSm9@!_Hf7pkJUFvM=fkfAB|7LqnIjMtE)U27l6$FJR!|C5O0|J|!EUgv z#Spt*N8fthAB%gQcZ~paPc_PZ`99~zq?>M#rO+$pjQIfLP*^j_qaQN%J+Lzrt2V8? zM~oh4kyGk%vVt8;)qxE~rir20NGTF2l0ZdV5P|F+ung7lJR}vNu1I-6Cynyo5Gipf zbQxnti#0eJT%xc?!^rQL@XPEX^MuRiDxKRxT4fL9Fl5nvfhlzFoElz%FILn3%=x~& z!1rqSy;t@wC;-0kd2*Lt%XCBhbg6$Yok=ez{Vy+b?UyHnZiTXI(TaHTKiSknSepIM z7HJQ6KfQ0-X=&|TXg6oT+qTAt4LSx>lh2uuz|raqh+^K zmSHeV=H)nH?$46U1+5%5x^6+D>iRnufe(b+DIM~oByAdGRLct{^n~aHh2o3iOsHmZ z(xaY_>dob%Zc)*<3x8Gi=8E@^{_N!I*t(m(2W+8}qE34J(!->o9(W7jIaMMH{&$(&sE81$a)8!W(rLPJe^X&)?G3c(MO6n zm2^Enj&s|6=iKYFGwBVJb=>Q-ck!;zzB;!#yqfF=-55OECy4XFEtx&yUA#Twt7AYQ zV?XYCQ)0UTw$KwhBO>X$>Sk_49CpUBU>17h_XCp{%qz!0SsGc#!gcUu15TMLT^M=|xpe;=jWXL^jU$HY04e?hPP?5U2I_`e_Ja+udG_d)=vWWblODma zXGS}4>;}qVcVyllf9X^69T|r;Fc!9owDd#Je?JiVb!Mh**Ra+?#R{le+m%h*v4ZNj zU+H?*DvdTJ{OEh7B<+QXs8yJl#zIPwPeBYMB4hp?rWgv@qo;aIiIWpV*Koet5OhJL z-aLIE@S3E-L(kOC6>j8U9i`opP9Wk|@?v>VF6m}qQqavnP+Cl4GUg|8*Fw?1<*pY* zSwV+vok3v*!iF_=PoA_Xz{Y*{-8GTs^!F!YZ4Ml3wZPRcntgueMbCEs-dV8|(9!Lk zRSPNM*Jq$YQ7w)CSMlzSY51NM0&F|TPwmaJjrrO8t`(zEM|51OHnE_NxNE@`rgXu5 z(1O@E?WkyF*fmnlO?JP^oRVUptTuiTTubrZBF*y8AsC3+uId!82udD{d~a6p`O0cM zb>wYw@xOjP^#k)@y~Da876S7}Ag=cJE6i>E$Zz&bCtD{z@;jo^sQz~?XDv5vS|3a? zT23Kne`u|Em%KRm-25Z-9!?=d+)&oJ%B`Kzc&(q%KDkp>A8-|#mVRB#-2pBuDY)FQQ$Wm)cHhkM$KEh#fV?hw&Q!c8;Kcpl5vtg~3 z1oGb(J@<<1<;5a(O-MIfKu(E935Ah7ZHj=WjT&2YJgqNVUSoxhaqlku?Yj$%uI%?} zeo{b6#}laMJVg8F7^V1(B8RDnW_i9yACkcGn3W*Tftr==Of*UMt)&Zu1&~GEAj31u zWCMZmyz`>-q7v@C*-6Y6Y2R#(@?J=JK)S2GJQ)_n!^@Yvw+dt6wif9v(P}0~T*ZlX zOXX;!pL?mxxmCPODEJV^Ipmwlx#W%It=j0a;4+^2nhbIqX_FDsoP-iPg+`eGIX#=-TKH#)7_=3v=vd4SC39E6X&|Q;o zk3gebv}h3)Eyg%u1sUus1{e*fFO3xfN0dG`Ca=W^5*ruIpSI84Xr3B>wRaKO?7&+9 zB_?(=n^M5mb2=5#Mv(Q0bfz5fnSf?M%Sl)vskPE9)o#xmaiIz}o5@)zoNAG7_RA3# zPDv40c@}Wgs07^(x>*ON7fvZs*4i(-XSAa~0()oGEv$^Aj{*R!kPwj)=KBJ zGu<>4o)&W=4J@Y#Z*L-ZalZioxvSHh$m4Vma2nC=O|7GGXbf z@jAoV!pEYa?F=R^W0rC!y+#2mXaNHyq=)8On&~VqA}F>HLXN&X3%7 z24gk1BcXfeefgcN*X!PF6XUDazz`U|tKU1eg#iYWzbOzos1 z&WkdFTbb>u!=7lavCcKq4LidONc9DBjZ+YeQNTs(U)ds68-Bf2b77GI zt3vpNFg#gLFKS^@L)M3Clu3-bL6)J?xwb10`670+f^uBoJL+lt4%k^SoOM{!7XC$r zMr?##8h=A8iF?jU-{zT^Bn_q5N|8-e#9jIZ?2<-NSC~G6I=1Khcg(GW`jk6BVxx}U z8(PI%?{So@^(u$(&n@~(Ti2GfWEt37A4@&i_Ppg8?x^$m4D$-NPdeTX%$=w-0mxT% zY}@!}I4E&^i8;(GKea1`kbN% zHf7QMj`qu>ahbQm?JBp{r(V9)r;Up{vi7KbY&zPbK#*xiHvS}^HHS9OI)XEnhp8ƙyn@9Rm?3p2yio6InWCs+T22i*Ps#%#g}ot_6C#BVzqxcK#EWNOiI)WKtzfM%+5;9hNz9in*XeEsIYqEW zL9N0fkl;@ZI_!(lphb8aC_3=9KplbwOApgQBX^=-o+<8>@1KY3u|uTN5C4odTZ;9& zFr5D~{5dmpv-{}Aovrzi5k$;tpSQ`$7Y0P#CLp>*Db7>mEER#dR`I+J`T?llA(x^J zkPsJ;qjpy|V_Rqsii2IVGs-r4( zyZaKKHb?`jr7wtd6V&)yVkAkUEa$FKbV4car`v5iB5ZiveSP++ko%@MHv4%`t%B3Lb@?L+ONhp zU2@R7Q~r=Ng|`Mq*`+Xy{_&4?_@6q4qqxX!vfU#o__?pq9L-4iB!Of(@YY4OiGkTm zDPUn#Kt-UKHWm-4=H7;E>@?MNvV)Tsa!pnX(XdY!jSd$y$A7J|{x&CRYfjOzhmRgsmn*r8CHCo_UX$IK; zUTo-|ROpWj#+#Q~c35({h2qJ$AC*Gqg+$8ZIDOLXQ#DE~4K!Bb4MPXWI!p|`V{DFD z8IzYDTzWUyJh00KRF1nU1jY8DyQW`-eD}^fA;JA{^eTV?1k7H)eG)8jbI2Ei2Pm~% z>DTO0#;xMC2JVCXbQ{|e3n;J-PG1`mII_@_ZBOs_J~<`KINNwHU7btzy)e%2X%m3d zQHmN0m}e1#Rd&_I@@{&yYYe#rwR1OeqJ$OXk((NsW1Vx(s1m)jupEGlfRpZx@&Lm+3$85@;>J7{ zfX`6Db~|j_qioH}dFyyM*Y(Ij{YbPvfR#MW2%uzwD|R2Ohg;|WYzUSZp(Ow0Ta9Eb zJ19ABm;rXThuWGHO0kV1n+R(#;!)0jwULS9I*XLI;guNc3hoWq{` zsZ`-%&x_#LB-mqLV3+|K0YZl9#rCIucI~I{2O1HQ=9RI5>Ks&Ic>l!22eOTdYTkh<(H;kY{-RFyN)I}q+8NwOw)orus%wt;psE{8N<2tb ztxp`JJxUs}^=H z=$f=$ngtoT2;gl?Sx(Tggc>*er{Ve*L2}s{AqS?|KyK*}cIiP%0iuk1DB#s!1wnu|zJWf7{>Ok* zIX7|+Iz9NQ7vgx)Zm@IWes2zlG%@wgnc&(ayVpCXOR~W^ismJH?FYMrRIVIQQ;V8% z$_w-IfO`sbtqrdAG+ME(E-^MNwq|3jJ=LnPxhU>kw*NIFdZO3=;9XL|4m}R+7}>XmyuACOvl z#dyP+dSwjM13vcV-UW|-zUj4V3!1_o_+@F0b*l<%p zXqU8%cW+Ly6d>wbHnlliJL8dHyIU9VpzGjE_D$U@Et20GS-mwIBWl?Bjt!Cq+wS`9 zHR-d(Y#i9?SP-*`4{cZW$eV%8Ug}rlmn(s+2Av>MsHeGOo-SIO$jMu6D>E_pP>21= z5&^^ZVULrR=dT{g=6&W28ix((S>VL<&gy^{`M$i`1Bq5Ec-_n?x=#?r$>VN(xjB3@ zX9tZ^M+NdcNYi}C$rYA(-*!LfyM987&u#jU{Wc4X1IZE-U%&PC0K$+yEn{LB?gr(w z&VSci##!m`2j{;a)eh_zx0%c{jg+E+A}27lyOQ25*)8vr6uO@R`QE#-ld8|@`|^G# zSo}<`BUPLlaV6acqDH&8l{8)%JQEB4NfhUZD&9GkSHY?0At~>y=nSxfE9o{8Ib|8B z-qeVj0t|Q8DN}jp<|oeGBQA8$QeB^26VkzN*X!b(m7WsS%lgP}NuOk+xPn(PZJqNW za?ox41lG>;h+*|Lz^pnNf*7j~*}5Nv!xD zwoO=Bl5r;+ep1diPEfC$+4w10#ctW;z>&WlCezs#O0kh5iByDvOVtcYT1G`1psg~s z1b3oPUQ2y~>{j;iA}5bXVutNt8QQaj(#XDr{ar6{U9OoEJRR1{wZJ^>VGL?&*|7PD z%N!VCOgzd|^%0{SL5?tGjKcr=5nwmi><4Y6cd)Y?Wxx5*bDGg^Ec_5Xrvfq8#rnuRzvXIP$56UcKoP$2+*79Ky%w z5F^J5uNH9I8N3&2P={?1bP_C?=_##F|QN-f~;dKl)Kh(H6Wk2A!v`U z9@fC5Jtc}0YkjtcXo#BoTaSr%(LpI7>UN2W$Q5o^T;pV@ z+W7r$?aK93+^ey#{B7}HQaqQ=0EdCo&TN3pk^v_y2iyoHwQvywS~I~&ri*cb{E-&v zN?Ec@jeCBday#^q+c-`eeZ?J_2o-*P0`(GMJF}d(OorVw$kGHoKzfRG%lWs@-`@E0 z5+L(-(4F-5aJ;&ObMUoz-o7^5&PdZ8%C66Xi< zgphh*HGp^{7NN3JSk82;M+hg=PVDfhsK$0Bs^1ht?ZZs`<_S7SgPFM52!9`c`)Q6kh}v9NC1O4au_OTqcgpBrqg!Y?zY=w zx0n2P$9CJ!?xxe-Zd*K1Jh)U)Q6b7D2q*+pRL=2sR8UY9FEA0*VdPL`@ZkUZk>HR> zG%qAfcE6O`MJ$fbPNyv+cb1@>x7on3%Fek6`+1Ctsh%1|M67r{(9T_h^9nnup* zshKopvZ48%qcYG&GO~VxWhT1F#Lx^$SpeuS$Dr5^sz-+ZvkR!kfSD2S8U4^6pqld0 ze>4-*R#WPtJxx}!bLl(rCX{1_nT-^)fg&kXOst^#4cvP&jk^z)$S=kM_{Tmu^alTS zekVPFK|okSR>-i&lG`|kv?2DO#DolCxA3G#Tj7Fmg$t;$P6BctLC@e1JRf++ZU8DV{x_$Z6Gv^foB23hd4?vub!L8(+RuXG{(~^TFb18 z-YwVQOA0VhRRK2o2-@9XUZ+BI5^2hI54|2Hsqz%nVl>b3jTg7^`$!jr8#922ew%M^ zB+`kY6;|n+uj-it*85{g9(@Tu0#ZU`aeL)2P2goM;6`j{<2rnabYvyqCf}8-N?!GR z%WBxph3lV zzIpg{G{Ld!`C7N!pp?i=dV}gnWRH6%xLy(%EC6Z@?Q%V-`ZO#AH@EJ6NfVRN0pgf3i$sVPl9p zv6s(E%1c>fk>^>_MNZF9R@Cn3v}p~Bv}xpcN!QkHgPRn;T z&N!}&`&P!cTS=`Xg@^eGnLwtD`3XDe4O}fGz-3W?Nc(2y`z_2433j1wRv!sY6!p*; zax9^9{db8ED(^)+c6W&67#9Xi+m9olz)nf|t#F`xJjG;AyaQP&Cc7%E1>3rTuk*{J zHw9@5!A?giIb8%AH$@S(er*f-cv+K8-*} z-2s7`omB4v__kHXx<1Hk?BJl26>9lRmtukc20K0txMPYeCcoYc8w|twkVbolPo_9l zVBl|x?kAegFoUF?-|w!MS3y7OdP$PJLS*pOh8qU}7dI5T*F)#NC4zI|ng_t}RL@U@ zT-Z!{dC0l&{0K<#&QV?D;M5zqN&ea7hCC&*3ZCc(`Y7XkYgLffuMJO?0AzWn=oQ9+x`1@VfY~H|2kM07`EUFP^i;qf=TCxXM#6 zuZ_y(-X>YxqY=Ahx`@?(bK!@M2doh!tYJG6TMj&C2V4Gkl@&bG)-rSRcaM`5?6wx? z-82aljf~h@(kNy;1 zuXnxaJwR`gJb?;EBextzsj-dGh`XHa+%|eu6gpj+eDn}#P88w!Drph7Goo2t!!PwV ze~&Mo%}`tk-Z-O=qXot~d^SUY@6sYaOS&L~lRXB)tWZVGZ7+Z2 zD~#%8>p`o|3;~95_Fl?0B3*L5y?7_r2=RPP{LgSKD-|B%5Z3qC0owGAWR3o>Q+Sbzy+Vk*%p-> z&VH}4ea&n}&}iF^?Qg?+WcK^{QLlQGs7{$9Kn^&_JnEFL0xku}+RdZy3D?N?%+wqN z3iYLe?qFm_z&|;HY9Z#r#fI$h?B!OIZPSxESl0X)G8Z)0=Ee(7kYstWf4T#D){hGV zU*LlE8o%rT>r!NtRPf0kv|f&U)e&P%;~r8)j9G0xQI&2_WhqgIf~u#V@yek!gUSog zD|3T?78DO)5?yuhRiQRqGdu>%;GybF+D2m}Bbk?>a!AUzeGv1q^=E&d6=EiRaB$vo zOFNp)UgW%vMPkM3-bpW7l&M07l?=t+nTxP%=b}8ikK9r>b1J9fX${cJJqb+XWreTx zN#qoAw#a(|jw?FYHgCgv?9YD{nE9&H04-%of}4f;o~4|6ey6Zr zn8hs@l|yF_%v2bdw}D$2eo}UMepyr^rwi8q_N-26o5x`vopP`8EIAb2tge?n30yg) z&10JNz^) z%k-adAm@wSYb-}E!4fNJM9D8sfAXfuU-PY9zxse|eQn&m2h6-Xg%krdF1g_Dg$QCI z>3}LGP>hZML74%zwl@B!kU5@5Bjr>x6fE^}%lz&KRJv`4K75}lZ_la=&QQepToW%9 z)cFqUPN+Y<3?wgEE33E14KAxQkMBljs5F^tf{RO#%Tf}@4lYi-d1aZwC6!{*d*jS7xn>dLTZwV8Fdo=&1F#wn{eR!rBB5x_&zN-+uS1v$Xfh zN=du`c@{FfFVeu|vLZY~feZ|TE)`J+A>*JTVwbW-0>!gnRf~3Wp$r7_VX~<`f2Un| z@13@{m%O7XaVv3TF|Ez3Z)IC?(Be;GwdG7c-tB7`PlVBlWr&psV;``Ip@O@|f{youMdYGHGOq8KP?Kn3KSC%&8{1 z;R#G0y*Z#mnhvc+u;U2SKSCTr7nv&GIW%1_v`ETOV7)h*`kN$o!_XOMv!-J^wyek~ z)|YL=q*$~2jqxZhPRs{mMRDolZ16WkBmhBkyDZsN(+uri8x?v5?r*q-O|a%-J1%Tn zPkW6zKN#>Y3r#X-&PV6lNTn0Ut6R*Z$Z3i>Ns(jN>lN)8%qPao;Ip)5!2P=Tnz+k* zxnI3>`P-1{kSaH}2-eH{C2by9&6*NbB&g**340Kx$tF0ULVQYX2m1CG(H@_uO7 zxy^gfr9KpDX2W{J2HXeS?~%>Q)_EC-b3?4Ia7SXex~b zT(%>GwTSrQfib{_m2xxb2cP}&(mN(R4R`LmYPM|8T3#BzJY2Fxm*+1+>oP;}sj@B@ z$c;Be8$8g=JVv8C-e|2pmKJ&C!V2dH4zq(bGlh}@$Oli0OkmO?tHI+BmB&(~0IP?J z5aWchu_h*S9S|WoHduJ`opQ;7!N<96Kh3Q8s1~StEB?;&tUi84gQ(HXjUMq_cchW zLaV(s+opF-J1d2fh@o1hJWz6yUB@DDY(vqAQDXu6h{qjvxS)6H-&=CsIV+lp752JR zd5Je9oEbIVYy z^LQ-AkZ7trUu*zfIiCY93HAu8$#HRmw80bK2vi?!cs{3}R6(L&9$giRJ-F~wz)m`} zV{@Q%3m>YT+2f5w;Yj0&WtzYzf`pvq^XsK4k6+tfELp*yPsMU#XHJ=!+LlW(U}$Nnm;(zieswoQ8?xv)=csH^2Z8wJYV{Jql`q5b zLFIP(CcksqBa%yZO}jn2a6aUFN|p#7x%Wm^@sSxOhfbU3!0n2i;QB%{HXgAPMr=k~ z7hK$*t?n96p{)}`$V#Cth9a}+{J`TNIX(+o0pZK#fd`c<+@HFXMi$F3}yrGjk+zk>xS-{Oe%HipAw-EI8}aBc_cD* zD(*Nn{A2WLb%7|4PM7pcy4_Z*k(RN@(_WZi2M9|!GRnD~Fk)SfjFO%;so|tB#UBsc(SNTy-9KwLR`ifKAW8!Y%d%Q zExSFzWG5UMKbO_StD^_2`zD)!vwUU95$GD^v>pN=XdNkck|<^^MOINU2bCq5MG=_3lC3Hb z=)iQ!o4LV%3q-C5U8+LC?;={F{=}5#GBJR$1tAt&iami~buLy2nH=|z1(~mUVRa#} z9sUWcsxie*E6_5Jd1+i79nXms)N(F!d)%)IZ_hp}$aAZmFuq^boKjYo)6O?op@Z7+ zc=a{QHfl~QKcGwW2u(JFVzy9Z6BTn(c7>x2M@17oDVQvtM?Z$Z&@N@6BtwCubot^= zU_QlqHeF%}Jk3xZRTY{^?{RD3@0FD)G_fI< zyfv4+ZD+I}2ZDx?#O~D`C|1v`BWN^RxodWNp=BwPRr_eP37pqJvO=iX04(#yhDoUU zVz3c?`q(uq9ViqrQM^TRRou)SbV=sL0+lxg3$W>RIxmk-<_)?W2`+?w*myNth1bvf zJ7&YP1HPjJJqqi78DKw_RAn6nIY0)sL}gH6y82|axKM=R}s0;7}}&ytVw zEITRl>#Bn93x*xKx~YF8YuP1vI-9!^U8zlsz+A!`evfo9z)9sy=!d zbQaJ~-|AZF+pNYbH{aHjEPP0odSp3R)%jxeScB<{M;&@}?H|$?zH5RtC9QJ zAOl(u-u8LdM9B5rtlr{Z89IEeL6IAgA8{yXi3+Ri`o;ZWkS83O4ZJhh#3z>06utXJ z7oVGeP3+*4|833cZ#1j%uuu3{e8YQqNaX_C)u16@um%=8&Btss8zgGu5B$bbyK&mR z%8I^%iGOHn-=A~fEzNeaj@Je}uv!r|JEc3uvuN(#Y(e!FF8P>y!`iV3nE@Vi~UDzcuQ`0)x$=m-u(oN=pV=uLoW%jjQ)o5 z9Q^?3ft%I&oQukHo-5S-kwuU@m#;dfJmK0oF=x?CYB6`s7G#BHY!V39#1 z2aA$A={EYPE0#Baa^iI?c!R&DN1EZE5Ppi_wlmaSw2@OL)VtR3@@D4Iw`XfQNOnMr zq$jFOn5(KrC*V`r8Nov`z)gql7FD!PSOy&gw|OJ4UKKajwM14s)D%M}G@ey)htIJw zDa1Ax>mger!GBo@zU_gxE;GIIG6-ehXKLnjF+~xY13`FVA<1l!F=h&?!%ocQ6f3+< zzO1~|ZE1nOQgH?=c6Mddq7R=)wP4Tu8yq&Oo76htd5>60v$}&MN1hcmt2g;v4#5#x z<`mG(SS+sjS8A`bu|URN@BN~AtV5<|>9i{@CUI2w`3HX@u}&>ItF_Yi z_*zu5YqMvD;$p;AH>h$fS2ijRsuDS{hruoZeO2`G=HOmg^AL~Tgvb$CTHCNRV#~1| z4&xg6L+|sQ|GW4cQo|ok>vr#<)41`$sP)8#7<{+7ZuNY^>kc{1-#5{OthNJVTnDi__b_kGhu%R(wc z+hr--yf^RD>!Xchdm|kII`#+~i8j;mxaH0B6P$hjx4iM#El#X}Sg~7LBxk|uJqvkE zS|0R~RA{zO2X>fQ1CW(F=mG>%pqbz<<)FCV^DBT?*rH?fS?quH=i%?{{Hm32%@NJ(M2QM^{8A^U{O5 z>7|hGn9O_N)(gZgopd}k{0B45?nmBt$MA0_U9%vWw`pNT^y0;91a16H3zK;TaCVMP ziPGqRIV_VtPa34zs;bcQ1aq43U7HoXvMoU9HRzHOxu0GMh3SQn1KbAbVPcTrKdYMH z)yjHiH#!JO7y8eyP8trK(AEG zpv&^W<$*aKIUaziWFAsB?ej{PG%HR6$ycg;xwKh!4!VRxUl~k`?SwkfctL?^b3nVQ zN~se$!Wd!uOCMzyz)N=jsk72Voi_ygM|t;Y+60-t@A;o1OPyFn)|z2?9mOP3WUbL` z%HjemA*QKRg|78ML|>kN#*vf=R`6saTCF_oxXY~2I(fR}$BJ1dP%LcCUrP#Jn}oUJ zW)M0+G5aV|M#Z$sR|4Bg6}>z>(Px8y0@LXEKz1&=hujeNlSI)28N~9SB07)Gdox3^ zk89w6D&P3!Uhbheb-_l-a8h`WeV)4Dx!&(d;C9tv9u(+H&-o<; z*V(fI+D52Zk%yL-?FT*I&%aapsxwCwKz#_RUZ~Mvyty~B0*sw4^YHwx=KwT1a)ila zeGOQwwe?9HaEFyxnRM$wz~=cTvC{a%jHNHx`{-X+k@(#r}0 zam-Vfqkb2a+eOCXzNq+My{yP{nPB)zKXV}h@8xrj620v18`rsdSsHg}_Mj&9co4|~ zUzUR7<>%wToppRoI@Z37W5Uw!PV^lo%h+vEPHcv6H$&Naib>Azi7>jSFI%-ge@)sHz@k;{Tkc;fkCpYh0)(N^3>+&EOGjP|*{ zdNW0oVA|E>305UuVo>t|6%RI%IWQVY&24}0E&XD=qKWf z6O*aL-cjcOOMSEjXBQ}-9vr{oXG!^Fg+(cjw_BFX1Bzrw6$)MNX^<3?JHkPiL&1&U z%*&-USP8n~SIs_6(FKAo*F30N2U^E%RMNsc`V{HmHbtjS9dt<#(!UowZ^f@#zm9w= zdim8KX}{a?wMPp-k12Wo`rG#WRod+riND!Vzv;1Ew7&C?l1${h?v+ky?#b~)3!L{~ zMyx~&8Wrb+Sg|_nu)gcL*Jrsj1O16hA+YqB4|p8=$!QOaVlK8Pl3#!`_8@=3WOlrD zBXhy;E$K|1wtmbCb!j69o7HzSM*jTc+xa#8Bf$pAL((w4+tC4Q_DeskeAscFN12M1e4~EpKIi+E$t`S>C7syXv62dd zJadpLNen?6Om!;Zo_Px+T@xiSMy!~P2paLk#2gM@%9t)O**d~*;crRu6jE$vBxorH z3arwp7+vI0)-v)dVSo|IC6I_h8;-g~V`!Zy7CSybn?hr`9eTMO5z2Y~d$IAm9sWrg zU!R}G?pIj-I-U2&mGjG^y_R)iY{Z5WYadp~2=(#;!7Xm=JYD2=f+-0d^maOjn>+*Z z=m75`3=9nswqwQfUZ0fEV&25KPn;F#AK&@Ag;88vOYUqV&I;`Ld*Hn{61*S`s`87)KhSbZ-{Nf$i zf$^F28WTwDYIQyPC}szm3;R7ScN(WvQ>|=)*{TCU=sS&fgC=)X!ka*@18Fom;T0gE z%oO#IL#iS2gA3F4yrKD2nc~$(9|-D!&P5KlP{z3E_O!Ll7JK-F-4^)Cy2RgEHlTmL z{?Cb}g_Sn*Xr++#Dg4#)u|| ztLDq+dGIWA+FlbK)=Fv~de4kJf!M4}(UiH~7x-Q?Ws)zN-RMy(1H?zIIe6t_09ah4+^dH4>@4WHWyiH5qv5W{l zuh2NJ=})Yb6ZN|{gkoptxEVQOO}%Rt(s`unvGxq}k*j{^jDXcT5}4aHC$QHI%QK?TF7h2p?4@LjR{{>cs|sY+CqPPy;YT zZu8s)q-`mY&@EWDF>04mlLM6{=tBM^@Kg75^k+dCid)l)g+uj0E9PuaWz%=&=;x$B z58=NgaSbNtGBy&%$`BnRhz>h~&H< z(i}JNdjb-ezS%tite2SWSO3L$cHsEpIY%F{0tb~Qd$(>pQ9&m*r>#T<8x@U;As(K+ zK9Ce6EQWsZdSL(hzo_6dU|z5E^|lQ8IIY0262#d!<2W$PVh;77i?w6Ci&bb4!-NJ8 z6C*mZ@Ju$aZQR#+05v;i#@~&L)JTbRx&R#}S5B^&X-in)17htUV$PC&xJqJ4Oo|Qh8 zX+D!;=H7Pd#N4MYxlkwnh~SZd7?1Swv)mhjn#&%AP(ZtQ@#4pxZT!t&?TuU+u4zMt zZK#sn!Q7NR_5>Qh$DZ|41Jv20KL>9mL}6h`_WT?=c20coD$iT;I^LDwjWaZOa#7eW zTjkj&ZxzDjfSX~Rbg4J~Gayg}k7=;ejiv{BNFyumS|5#;v}EpaTFGRE zi^o9MyP~}mHv_0yc6&AO@o;nSP>Ko?VUIyi9gkpOBap`PK5U<5Cy;I*f5*$sB#{0~ zbK&1fJUfBp#7>x8GaG9Y#iUYX9TgKR$cDJmXVP@Z2?2J|tP*M-c&wFbKIR%E#u5(= zCOb{cu7-=t21u~4vBkrf$Jc%0Df{0A4FhcmLeIV?i))Z+uq+?wCU1Hp@i~^=7YGL2 zMTe>zP<7Gi;R?a!<^C6t+s9Do=#Mo}1oB{dD4G zA=nEhLt#8i56Yz12QA^OksJ7TpyVSp9A6Vd8$*wQPp1Y-T3drVq`)+(z9B}1p>Ot~ zHxqrTU}WOgI>YLMH;VrAqjL*d7gW*4`S5+3)r}CP+^0mF3&)OKEV0CrjUJe1!(I+% zD*RMVxTb-B$;egqmQb;k7h@VTv421~Man~S4%P2nlVUz}YiDj$W z`2D0atcFhUsPc~w$06G|af0rt8D0=ROo{B^WUD^r=BO@mKKHrFxhfu*iWMkhUt}xa z@z7;E0N_Q_kNF-eg=JEZ;M4qhCb$GHTeFMoXQ!|@vFN#A2Cm~2bA%#wRLpUOmUlDk z7`>G{==Rj5VdlEfR5_X+z1;27(TpkO;87->A9$M7b9?9vYTLWm?5>a82s}%6s5(L5 zjKInFHKz-VJ#kXHa=QYK40aov*;Fhka`OAN;@W1y_|l zmkDBkZ3GEQw5p!a!qJWps_eWjsuo zRvA&x@A7^sONoL@jsVHa5n!ngjB@S(Ah7%SZ!7zAik}HDPnc7yNXBcUvK}%+S24vv z=F?6p<`~c+G^@3ojycJmcczt#SIuiyUyJG|ZTvd8vxK@SubjRs>a-LZwqeotM1hxPc0WuI^53jDQLt51f#57S zUlW**xlh6#grVrkP$V;`<@7v_~{Hi-kSTgSmDDRWc+=jl1F z2^%g!E=x(A6JsOG3>&EwvyLK3=z7CaQv+QcykCsUNrs|YSS&p1hX#+MJZaV!&SqR# zpTZ$`Sm9!_h97(Je;2UCSxu>~*W3zFAd4 zZJ{ctvRUT>i#!|EaqpjaXAAYw#;-qKwD#>L^)b4Mxiz<0-N)Gt1;@weVmjA9hsKnF zi=2GU-sl=xEhmw4%l*(C9j8W80~T$4@R7h;U`VJE;*)mTRcw>9=s;1h!rk`_~3F+y}Saza!POtXhsLPiM(U<3aIVK_wJMett?6BbkQAR zty>CQ%TR3l#qRf#{$b-+wF~V8oEI$a7%}yN8SQgxH$J^OF$27n0 z+m?(@W2Pi!>%sasV>1Q7D{jozWt;hTI4~R zwX#5o$@vLhFt|aA4G*u7yqOi^yYf$jZTto)vK(u9iM#}_b3pEGSNyPm$4e!e&1-MN zKbC52o$vdpFM#((A;gREk-R*s{q?>}2wUih1CENZ71So1O)&7jwluUp_0` zC`y}tdtTpLXUH;MR^VxVsSFdL&(ginbEPH<+CZspgg2}4-deYGPE&9e)W&a;+zo?F zN6y;Fyf@D%<2_IFPw;WJX7w7*hUsxLD(9BUdZZg8OQGw+QB@7C^E|4=_k8Tr3v9P* zeJb7hICa6sSz`rXpfW>MvY?go0E)-(^YlY6hZgw>K{tI}+(+M( zZ}m@6HL7#DN!&^>4F7Ud-n8I>3>!k=TNFm-(o2~qfjv-llIFfD%J?RSx!bu;TQO!G zaic|&=FFb@#FCMN&9>&e8#b}h_pVde3uwJO|H8Bk|HjZp&r3pNYXB|KuJXtR(O`2L zgH3bUb5Kd{t7)kMr&2(#3JLqnYnn2Vs2981{HIcmoCA4x;ySo(L>Bn*fMji+e42| z8VR4EM!Je#>(&WvrLd{GBnZGy7hAL z*}bwI47jHA=z7U5w@-xVU`*uPfmV5IdFgyj4GnK7+6_h0y<80r?~AH-vl&~&%aJ0_3Ru(POOf9W(L6<6a$Pym#G*e{Mjulf?j58JhD|SlFRdp zxSA^E4hA)oOuA74opVyTkQIEAtUHnccsIT<>Oo~iGN&Y?F zu;pQ^3CO_Ux62YkyS$7;Vg;z#Yi@Cy=o{Xs{hKCTjD5_ZfjJ~-wbPigRTx=47kcpI zsJ3|{2Ywa5i(boJycjJ6Ogqh^cR?HML}=pql9na^!{8egczBWfFTelJmrM|A`J3Ch zq--=<=EOxw=gfd~m}0>7SOskt6-QtXOY)DKktsS9lqx?74Hz?22Lk~J5tY8p>Nb8# zWN&0<h^5xdA*Oj5Bz3` zo@R9x`Sm{TshPiCl)xnUFXwd2>!nM;NftMwSU9=}vbqfBHyjBT^LJu*W?fSK*t6oi zWk!zEs!~>ZC}MV}=GNRxk^#A9eP}Lp!&?#BE=zWusD2kC*TQB5jC^PUj;$kL6sv3U zyS}&6Uv*S;$@TJ7w>(KDZ>gY5zKO{cRfB8r|H_p{f{hg_CVgxE&pr+|Z8@c#B}+-( zYopabLF|ZfrF|3weX&ZYnC4;?o;n-(VM^XCr0@?i#tdoQ zHIfGY4u1nEJB>iytl1Zhdbg2$^JW-QaH95I1t$GnIjYr|gx*OPQsdfjMy{sK2D?=? z*ZhY0+~YZ`FF&<8!h{^5>w+w@$BBK3XUworLot;UIY7mn^*PLK<5z|@M6GdMr_w4q zg$+?pghxq@#K2GFG^-Q*mxY}3%LNLU)4q@8TRhXeGJ|$=wVW3DG5P>~AAC9q!Ptv6 zR?r8>`0UPzTEE25G_M;G&1yW~B*_VCqaQ$U6vvkZbd7_QVZ>_Lyd{~h{tM4WOh4vB zR>Ft+j&{L{nI;JR*Prx<$WC^`$B8vz=r=w>`0S^ca*C8vF^l4+^t%@XmUv$fu9$L5 zz32>Wn0@ChjKW`w>hazDRP;)7TBylH1pQ1da2G2+|`Pw!KvKU z0L-iYy9C{DA@oYUGZnk%o%6j)-+2oh-#*Z<&-Z_U^!e$M>z%G9=p6e}&0olJcF=KR zhtLi)=xm^v6pAELF}O`(0p`%QR4>yBw*yOSiv)Wi**UId!>`|F`D4Re_8e0E;g>d9 z!oq1yA1e!uX@1+h_?gNCk8Jl!x7DOdv1%TkoeDTdr$8JIbLyUm zGwA|u*Z1ld>=xaEVxltFWclXtQJ&be5MxvXZJc{N!&oha^KRL4%?EYk@$WkCzl>P% z?{5AZtd#8G#X~M}m-pU~K6xSUIK;~n!V7umQ`-qO{&9k{(n4O9@OTtnt>gEQ8)CiR zlfYv%5~Er^+o;Al|9Vqi4DsxG>61W=Y3F+v@+#f1tMHZ(?HrtVInR7*98t%A>Vo&{ z_|=jd;$voST@iIml<&PI1V5fF0?=lm^+J7^i1s&p$S3BV13 z7a-vg_;i7FrY8!*2%jj@y)YKlNuQKm;h-IDKaIxjSg>K8Z5hH!YEI5S*m{zfHn36` z?P;>|6^jhynC-S3DP{vjQh-@RmIR$Vi#)aTLsB02>$N}mb+T&$6C2V)10N#93=DAH zmS4Or-?%MfSfRz13GFn-I$BC@&id!cCS)vM8FGXSx{z&VUv?73tfj~*DyB|Z9bD*# zImpnM4E4(R2`z(v_Ep&OB|jTHfd|&V2fIL3Y19I+<4ia{4Cx5RsjyHlR{bAVQYFE`Ga8^WQh+e|;b;5aH!hq)yq! z&!N+%0i6b?diwgQ-R{-XKT#y}Hgc8-AdO82+rBO`Tea7x!84hMk7$hJ>Xbvo=luKb zsqzYOoYxRp`qX}+4bM?!(s7O5 zm)1kCfxSM}yt6(H(n3y->JfbQMEG~ylqQOLXq+4y#$IN6!7+O;;=JchwPwJ*L5dkP zc{F6ZzaB|+&YMl((({KuS`rvJ>raJPp(M(8?I5)waSSBjxMSO-l~ZmDPkcFk%2St> zQy#g;lU1BXbvIPyHL3C14*w2x%yc0v80O{$tf@G8oHjpNm9?(L)vWu zlfr6a!3q z`>B{!QMvRkNwKhsGXS`5WtK8|Z(_ptUEWd8?NAez7I;gY#MP-AA-kp9r$6$Fuu1*Y z1+vE9z$*|GfW%TfKPADxHwf_W)dc@8^KMiZx#6pE#8sbIDv@I+h}#X>&xDhGAkGfi zA9O0GSXR-nNkeyHhmV!|nhowp$tIFGuamCjU7vL$VvVF(eI$6RvQLIr@wL{cGPE6X zk#}=jIRo51yz6`ee~W)Ur!-_K563jCi^Lasdf8)sY)GkhFOWTB_9!}p>iM1H%d>hO z0``O1^K*~q%FT5@o*Q72Jbgcvt{^*{Sn|}ENuE-Q0l#;lk-kC7c!}-c2XG!l?_mD$E4WPNlD=O>qZYH%p-rN*|PsCvraFw1ydrP%$sK zGU6o9evX@GD`YPiJe$x?2dt#c!`+`y7UZ-NvCUFHgB$R13@) z*W~vlMV?LS3`Ls6Ze$jttok^SHOh?kxpc+=W_B+i9i5-t^7gSY(RO^?xgqgNB8JFgI3;uwJ9;AGU^XoDn~hjZ`r zxTmp(4 z8_!|iAxUyAxPTHvJH#3OJ3Y!pHv$d5h1|8Uhn15WKJ?}BY*(#ugITqW~E{FrLj|UBqnoQi5=n2|Ej0^VgDs(XivZZcB393_AxX2Gp7|Dh74o z`UPW&vpEE_Ce3bZymfHeY2YoP)PZ#qm)j1DC>(qr@)9CMWwWGCt0|8=0uwAEz3`J3O7RCZ>M z6KgO!vyCZ-Vj#gdlZwHFkEgQzu6unN6&I%Mc>~QJ$hmR;h!b#x2RxTo&oLD6RGiQXs zyOUx-r0jqK{>X}GZFnb*Vd#5MEr2!j8om4xv>FAk@0&;c62Wug?f>|zQ9CJBLB;>h(M&*B3J9JcFC!bc*gUBn23m(po2>Bfy-ZT1 z<*kb!k`3&38z*+Mm70Mjn_{4nKZA-fR<9UYJ920vZ--`>@BzD|FO{zG(43jpH+z-z za5ylQUJ}-Fs<_*{Q@QaVdhB8UT3=B%02qC8$**3MP_#9wEPf5W zI=aX+L(v*m8h#{rrJ$N*LyBI#G|#Ql4eM#QdS(WIAD`|gm2S(XJbABzdi0~!?`-|M zm#*_W_72$HtHDUF!o5xTyaKdvtf5ywQ=Ca97cA)%E&1MP2C^Rg>8AW3 zB#@SP?~o+9ZSWrs`e;SRWqEXeSm*3Aai!Z5zy0%&69y`;hW<7C$tby7c3Us=)f>av z$kk#?ZRWJ8R|D}K1D4Hwip{nIEyX|*dO8)Om+w)o47xxLD$j@$y!I%elyMKxZ)Ste zc}6_wGC*g~uZ8vns~{Knmb{LKS-LuV=aHDNf$3~a5E*qQ`(CuJON`3+k|EWuwZ`_+t zU3A=&-kCu)yu6tiidX?wH8iPLxZ^w6Prlh1J!3-3Xb*9k2TpurSbP4}l~2jx*9JXp zX6R|8m!{a`;-5Z(8v!R}em zlKb8F~b^r|SV zFH7>T^(*s6HrfGsAM{$y4=jw_6}56oo7?A%mbo%3eoBU7SJW6f^4b|zRwdVH(0S3d zKYafWe&b23a^gMCN@A6fmIl*;H5)_~(T8SYj|Th6j~9K}=J#xZg!v;Aa`sObcg$Cq zB#rvi&km98P8>|AHj^DC6a%K+E-L2s>@wN$sD3EnZ{zpAYfPwGJ+E9GH)Wk?lXSpy zrJyjplEJ#n6;mqR;-=tP+`P>JE2e-{5T#2JLyz(5Wwk)big!+UJqpuvN*v-pwh0AR zf9|Fm4!{K~Rheq~X7+tcvWVBjCuoIQqKO|b(5a7!hF=@iRl8r~2%|pr?w5RXEQ1bCoBUuUj8LaE$d8eB+91*73$)?s zUOQBng=cK9TQ2K@gr4}|wyA?IySbkTE1)6wVR9(2OHv@H^mY8!GvH<(q?mjP!j>^H;V9v<1v-{z~I<1mF~%y4nsRy2e; zc5q3knB%|L1Q(s(85wDJVz|Va1wkHC40PG(hGbqkKcFQr39_g#Ax2XmYU8&^2HXvF z4iNk%^AQJ*#@^f=AVyZXQ%A6?>ia8FXpn?DR;LS5JQs)*{IWPvmG;%k-QJ zbR|^08Fixkg(Y;uG))_Q02uDkkB$C9J+RQ91%LM$w{_w1^Ae(fLU=&lsmu}gK$pg= zq41V*Gs+Z;G?)fz{3OtrL+`yb2-z<%;TXfI=rGR`>p2INmGezKA2ERg(Tm9(i}L7? zRp%v(kQQUWBbj$Cs#A*dK}Gv`kvF_K+543h@+SZ2&)=Exz6p7^=hWUKC!830?Ph}N z62)Ag$ayLz+oM4m8&V(!d<^Gd!Ahsf8>9ms+VCL`tq#t8nZb9hCvqNDx*b>4N)83D z2(Ooddk3i2G{@-7sa2tv(u0By0f5C*aAS*ND2X_a#tgw@Lko{tUDf2l?3*->*%@(A zxja142U=(Jy=Cylj{#CJqLefA2`+>hBzov&2F*KQQ`M1nm2Dg7{|L5h*eh9KYm(x< z|M>4;nXpylHfIL8F@-!eTf#1i>7>XVD(1eV$!8U$D8N?>`IyPPJik1;JF-)X?M!su zgDyF$UipB$9UL8*qEBVFXE!P`pmAKitPESV;!SfqS<)kDQX4h;d>%jHf7uf>JjAoBpy9yWPeq<)Jb}bebh!R6Li1HOFaIW)9c?(lP zFLAI-EP`UI>!Ywd=rP_NmdIHtxGUf4l`yN2!ZpWg-aKd=)a{n9I;biF(zgBL7FS*5 zlc|fe;g%cU|H#UYCSojBo7KuXZ|2PDcm1kf^?vq<*L+SDv@!+~NN8+a0uDn1{XjfK zE|#S{CPKb^X!Zg>hI;9(xv*+zUzVV$r7X%*s$r;A;7ljonJhN6|g6L^E1(Rr6!Se@|*8pBkM-1 zv~uFm2DBd^q08@}m`sXnrDE+xj1hQJ43ccdjfzDt@C*mCrfC zEur%Rcf5h&43xBubM1kQ7q8FQsx=&!p8D+5S_jV6zb;c+#^0aMNp)T!3F*rtLGp|R zy)-CO)GOCq69ZM9*DfgOK}Eim=s8=ab*m--3D3TIJg`}wm_6nA6YH=j-0P{cbVNT7 z3nz|1SaD9bLo>Jp#znYelA}ByH!_y>GzF)iTB3b|1|pnCARcc8VnVI~r)VV9HE_44<$@3A+creJSy8`6iSsytwgWvf7Eao}Fgvo=!0v zDYAizSt2Om7@Nwy_{h;~JmQ;Yx7;Hhbnqc75i;pt-&=Iw*8~n#!m4Vr&56MQe4itP zLJ`FjP$Z9v8B|^vO5f{us}dr)^}e?tbz;WyK+QQ}7BEGm%Hup&@iZyiyf-`P<$H+rCg(!+}tphs5+jiEL;NKmlD%dl)PdvyqM*x{vM@weZZV#15s@S%n* zabmoHOJoFIQYa>oB5SA^ETA*eHD*H08%np0j3E;!JBAn1HcSkUp14}Qd({(ZE)k?mFZI@Q_RTU#9>{uusQ%+O zcR=ml>4=*P8i7C>6L5>Y9+IQbR0hphyxXiOCU+cPhs72W!{Bf);jufK)y}))*uWo;s)@ zzBU(;uuHIm>EN^}mIpqDzSrP%1^21jgaL#tJRHauVumSp@ObajAAEb72_A;|Vb{qP zc8(Ay)=etRz_g2Ez(UTUVjg`tjk|kZd&D;7)&R}cfMi}(s5U$yyhRc#xZ|G4L#OOo zH)Q42U=yDLQKj1{X3M-q?Gal8R=67<1~RN`QOl);;rl}zQ-{eOTf1Un8rGC(QHA)$ng_({DUo$CtQ>K za0zl*O5#SdV4OG`Aj?cjq*BZ}iXeDQ*t@*3cSXjU7^8;dzDp{erw3qM--&SN`P zC)%`7wyeJKmv7T_b{k}Uui18f;hOR6D^4satn4dhevxAL^1yC?8t7&F`wvb zp&f6%j_D$}?3fw3lU^U%7N#>MRbtq`jlMm5qW2M-7vFqs9DZh9h|H|JyXJeQ%|*T@ z=sj}QiES6i(j8&DT&I{T6uCsj^hnbscX{hW`-ujV>am1*t81?7W@wbU($jc^LALGF zD-^{-WNO$Gb%Jw>DG5pp?eaQ%U3jqPWu5ZAjl8ZlqsdHujPt|>ELum?D87?PPfy0WSgJ2h>pQWcs?_@FA!2*f zbi!aCBC+!hLgagPjsrZTQ-s4Y3&)&z_i=16H`0TsL1l?>i>R^ZWMS-y)wl zvE;d9CV8$<%q5Ckpkj8*l6eNYO`!!^tJNF>eb+Mi>YmIE0@ni97Xx5dZ-(-+ha^5ONDoC8wVXEap;ph0cU$Zj-^7ToV$6#%Vh%?fvRVx1^(ynPm|Zg+hV2L zNu%PN(3r|tAUKP%D9OX%JK*s#x6&774C*sgN}UK;=>6I}unZfrX^#V4OfI)yQbS{f zB<3n$LgZp#PYc4f;AtMsjAoU5iD0Eqo=O*)DmOsUTDs(fz)ru8G5-cEB+Q6YyZj_X zEj|Te(+(#4>sJ~{(re?`Ei&8hGAU*&MK)70DWSzY2m@eQ<8Dy`WKiqnu>xfMgalf_ zpi5@xp1GB7nV}D%Vh#)3Fz4Bhj8hr9t&Dcq=o!1nap@`@04=MB0V0CF84U~zZy-^x0GsTV; zsdA(&OXPf%31!zBoFZG*tmxwa8+6rBI#eRZ5YY+8#@?dFN3fE0xQa4E@sQ|*8mx0P zRzX+#ZkwJV>6D)Lj-Qbs$qlay{-}xz6!smQEAXjsTH~X)xV`dzN%iz9A!8>Z3 z|7%u1dgr?MFE`~?5Rtf~UIh%-7@zpd0~vNF#Z#>ydw!c+W$21I>!-H!bD8^GJ(J6H zN_W!L-gUvv>V3-V;*S@cw`ae}aU;h}ERBwWv0VMZ@#4r=y+QU3{($=`pl678t%wGy z#OT6E)Mxr;w|M}qRG?mwF|RZvLveq$9w_KfsxqM9r{k?PK;&>$u+Brv>6^WMdM|e; zw5GkMI7g>|(SBx{j-cDm0aC<1Fc`~1zxWwedlMC*YHs?bNgMi8{(;-%xD#tbx6HJm zW{PQ|$T=zoeJWUZnC^9kXqF0)u?qHl#CI$o9D|y?Lc)nkZIo_#|m|4Z| zqOtN_ldo!(Z+GvNXDF&?7DiqUyvS+b>-=;>$kVidhj@d(HvG_>D;!M@+}Grg9l$37 zE!j0VJc-xguSxN0_09kt40`t*mVJyq3h+%|4lbdOMzk3u@V(j z%ePj{wiFdk8=tiD<6pGN2UwnNsTY-a_lRpGrP2AE#<}O{HNp00H+b1H*$9e>O4WV# zaoK@l^X3nCTiQ3T6pKcUnfNHBnB5d9H0}~uVU|f}h755sRe@{&jOt@>nhi|Vpzppq zIG0-z^bp7kp-F|2Rb!~R^@Fend}JE~B^F7WF#7>7p^L4JpF-eo5Bq8?HUsJ9>gF>L z*yF%DkPeCCTrA5EUlZS(m14zB%qC{3peG8Q!ix=bvjUqJSI@Om+=tC-*@zZ4Q`%+h zm#4iSVY$1o!Gsg1^;+4Bw*26_7`v?Rh^f-e zj^FrMg5vqz<<+x6j*E?y-a{!cn*-Wa7dby z$;p_%tS6;TY@O7bsTXw=Q%#XWR18K|kUI6atV#%Uw)mn~ltcHwp&L+RIOu{|*v6>d z8G@~5a7>1xcm62?#7jq^MD=4X4l(K#c(a(!_3w>bCB@;;OH`Z|IOvinY7ILO)CPpg zdiinj5y-$CH^OWL%*&MxixDjG@!W9hfEgv(a`K?xS0|09Cct?!b%5h*q(S_w)@(I7 z5}Ze01q>Pa5+Fr02!hIV1E&d?ySgLO!b?SiE;?C5z)&cwRy*T{TDU+Z&=m@IIZbxg)VZN_jPeMeN$d4%cb*G zX;2A&PkLom54}g&H!-;#jcBGfnH-P*!TJJELk>@-sB#5>U^W@4d= zVt|KdKNYiV<|W8|JuKEM^zzb>HpR8M3Czbp2nY>nhgkknEmLX*jJ&4X9VZX{-ExWEI4RGU-ZR=p8yjNzy@S%I(=HQ8myc zLa#^^88fed|2?uX6erpzIqZx1nz4dT;CQGC)kU-qy}=6r>j)UW5Tj-{tpoNv?aq^% zfVD9t_!3EbZGcr~2CQ6)fl?GL6|>cK$(ws+JD3*9kjoCc1(XJT0$!HGvPzGON=>df z$3v6lQNp?Dokw5B!Xu)WVAQ%oU6@}X@S*gshQvu*lvX*^{tYc$ZhNGyy#B7<09I<$7_reo*jbNUI! z{?dbV3$O?9CyI2hv!q^{9`xsBz@ya=0 zyCt8`E9L9A!rF89jY`ig^F9l^8k!bft^}L7dLr7iJilPJAmaI_9Dl_+BGT_3Oph?h z5TWaWEV9Rm-79CzZ$lLVm!KE8Vb5R1wh=&3&0(Zt(a7GN$ncmKavLU6eFQi^xInuoyi0C5L&R&4+AQz5Mft zCSjY${*WPf!F2p_Z$P%dWmKW@l8fU7ly$g_=H?Ii+RE17njp0QXaAqQFM(?+J=gb$ zCnO(+Yy@-80J21o#SwvUMQxbrOlP`Im$~h9yIpSOwzaqIo%W`4Z#%t36a@tY1sBu+ zDyxVh3Zk;QfIB#-D5A27j-m*P45RRW-y|xDL~|f;Gv0saS4qzD16{oe0gp2tZ} zZoM)NO01Et=jRmLPmu>yL=Lsx`w^qr_~t5pQE(Or7}m>oiCX7mi!|3<7K@5PK(bFJV~+__`EarIW{Vmd94Z5G>cJe3@amBmDhvy&JB*3CutAfF%#t#GgX-o z#lxkqpXf-Kr!SRj>g=~lv3EXEaDk^;6IkkzEo!Bs1<}|AxygRrms)8(1q6OhlAzfE zifbvev$K0-$I`vaijX06d14_uPF8|xFWgh z+9f_iH>w{%gUA#fX2EgWG>h1(CDGR?rI)SDRTo(BNqlp%vsQoq-n!W*MnIpQNx+WlG=t%bJqObE8 zi_X$OjEC9AK3=`Nm!Am!&Pw@uCk>9UZf317sv<)30RJ4QP`42DLc!DRqFdh>a4e_0 zNv?M{!L=1z9dCu~;=#tO$7Llje55d%Ud4ksS#(cG7rlwXt2cfHTNx_^XIx_h$HQ>& z$biGrVfSVNEn}W!KSYXjAFowrK&md*0Z$FcSmoJ_^! zuP3KYG1bE}BRdx z_O*(|({6-3avoq(g((p5Z2?A}c18=8EoBaK3NNw|Yj~oGv;gs6+S|9N!M%5m^dzL{=y7p&z@hVtS_C z=O6Jap10ZkjN~)d!()%JnKL5#{u_fT zA#Q`<&{M`WlU+aA`A+lO%is9RMQ&CbiFY<5nP zG^*=^+gz@B4uIx&j#I5~v+D&xp-ZbUeO4}it=m42I^q7XGp<9@)8raN3j4?)Xp}s; zV~?U=+DezIuFu)Y-xZwddNg<`5BQ^0Wvo_pz#GwsT=VVBVrq>~j#C=wAZb;9(-gk3 z1#(5J_zA9=F7a$+Kqb9`pk_Ue)~IgaZRIBf;%`w&cg)}BUHpBR8@j3Yc-?9ZcPUXIuYG3>?ed=qK!wJwyXmi;P1fCaViEfE*b@ttKvGWdR z2*tQu@Y3KPmuO!uZlm>oKqs(Yy&7?u$^JiTY5JneI@SjsaBNVwF)K*d>~rcJ((8^n zPD9=qp4+AOL)Xh{==#7mrj+Vo?@u5>3`gtqLW^R}liaLR$?Nz3}7WQRJ$$ zLB0R0jj$kMmGCP5ZKgIj+O5=aNF2w%Ajh^K$iUnr%JUtK3Zl6TGV9{kBQi(90AAYfdRKY<|SAC9sg!#I{0?Cegl0tidb^2KO@QO4(^rwLo8tV*Tl` zs<%R35wh5-4XSefBWav-(X@i_Z8HX-)~!@^1@=E>p4-DV@H*TR0&nqeDvm-$=q26; zvP86ARtX`WN?DoYDE~13jBA~6dswVx2%5?iQuQqutjg8Ttf}mBS`j9GnUZt=G^gM^8We&0ORwvy1HmV1la_BgI zfjCi8;nJw?lqZF?GpF23+%w$k-5b^I>@HrEpqIW;fi%q`R8!7E}<%%mDJ0aIEf>V{(1Qj{oH3Uw)+bQTD*4LDw(JE%ws z!v-jnGJ;+N*U0%8e!(WE1_`iAxKwzKZ%f?RU}Z%dnG99aaCZOyt#Xz+MG;UZ3^;1A zuBAx^buET|{A~I&Sto5xTVnF8M|jHt79-p}eusYK{k8Jf48E-Q@5>gF_1t#Ew(F%6 zaB7F`h;t|w+Gw}Hj#zZTw=WD!*HUI9?U}yQdap-;SSOK{sk+Y3bh+hPIxkyv%e4#I zDY2EH*Kg3TNw#C=B}j2MOUJiTodZ;!U!2AdDU&{*`{O9nrbwH0Q_M6#qOvdU+0$UX z05OI)_L>09Sb~rV{rW#aE+_BuPI|_7Ot-sc_KA8voF*BR z#o@QK#xL1xtJD9=*VD2`Z}vMhZYQ{XIc>HeuKn!qbH+l{VZ$zm8KMp-uLVZ~+)u*` zVKSR=(3v{Hytrk)R;fFpI|r6PjSeWX?8;0mOkW^Ci&>21|CqYsIbQ`K~gWr>@RYkfArCiFT4U`&lD!c zuXSn;eLwWR<3}IGG8-Aabn;SpoHOn?uJap#yD8g0J?JdZW#hQT&e!Lb^YP!0KKe*6 z!VIx>I?lO+&LBmk%pF^L^r>YyFUeIEiPI%FXAcEO0_6t&B0FRcW@p14c-=?l06PkV zz0l}zKRGfOUlD}ySMWUM84%|9pf-h+*l-mksAvw0q}5RDVG7#%BOWMF6LYbs4ha9d z#8;$eB&^*^-=2Vg}`_Dtp% zOzY)ms>~hE&uuS!2D#szN26#9}6;Z zZPwp820v4Jrm|5DdIW#_OKl@BSYGi$Q*?4dijS14eGnnC;^^N z57-{oqeu_v<=>|V?NS{Miw5mZORvosw9C{TiF8W}gX4yg@XBj5y5wW6t>gTKuzXb= z)7PSD<(BsahSkb{!*^@RG8pFO$w@Z?PR=p_lNODM93ib#Q)&urOB zMH2(W&A_gq+3eZ7U@!kPKUtC>IVayWvrE1Rk1txGShR#}p%Vn1P7WEWW>|89n&lQ{%vCQ}nX^9| zOEtC)Gd7#4#zv7W+#aSdQB&K*pQ_5lQBxO=!Gq=Y)gE@WMUWVF{Y1TV_}Hz^G^6;6 zxGZM!kFCNsXz|c0Zv|wE(v@o@NvgPM=L8Q)BEhD_bY97F|;{7wGIfCb@nQhp0HIp|;2SwgKUWY74mN+kVzTRPN71jx>cqd5T z>{hxXxQlKFnx~1Gece=YgTRE$zW(ADUa|H*$%fg-utA@^gUsu zsGGs}Yu#3pEc&!0Hf-m-Gtgt-53=AGPfqiVcS?sI`8-Lh5YJ@N>w&Tr?}!`+ZXzq8 zX0+fidmks%P-UJ!Dl}c3%o?ka%DHW}XoHwpwpKeoP@I$BrIBkPUsglbvFl}zy%+9e zQurP8Udh5iyM;;2<>|WD6SQ=3hmzUZx9W*s6n&~S6_Dlv4jbO?m{Ao+vV$ijw!KVF zP`_hVP?g6R)|S!Fhh?A`{rIE4^Y^m+>4pHxx3>T4w`7wIN7nWmg+%fxHkTqKv zbx3hI=-my%;AQT0vg;FDlg$Mg=I7D+XSgB5(=p@+rjZkyMTVJJNV#~GS1X-7ZzV4> zpqk{+sj~h#t-?%tt(yj!pft6-qlyOkwgvZPy`frN9e>V*w2^W_$WseqeLy@nA}54U z%y$}To?)r^WPCkk|gO1c)FjtOq( z!l|$j70=Ducpu<|iphWa!M}gcv@g(%YI_tg6Lq`j|J7?(u6Hd1{-DCL`05K(Q++5<^9_fVey= zZAXdv#1F&%S2ih#pqJx%*q2}eUw-yp%s+l$0L1rBQSXpXUYShqJtIK0QS3E}G^1P| zR!iw?E)CsZt@M7M6lRszVS$b#U#|?0k=CTpTyLGQ6xL(r&>5c1;L3dFieFE?`>)It!W>U`5eXytdy*b47YvSgX)f(ieEOPyvVao#+IEZ;YW{l$KwtJ_+LT z8We-i3aa*NXDU@koVPkbh9E2}2sHMhLJgbOb19_a9S#?3lUX6>BClOKnO!`Y z3{nwQf_mpjQR2ME-W{?M?{1fFx>|5U-k`p-@aPY7-&qrJPMyu`o892t%WqUan&0JC z_Szn43bWe2k5@@AFiG&?U^9c@~=h)O}Ea&=IV^`XaAfQYY_a>Kv;CmsEvN ze~n}4=#zz7RRu|-uS)w~?{L2xdJ1}?_IPh#wbO1YYH0ivB)h8&`_!#box@8Ml)YBS z`y?!bUvHl|y-|JJsWYI%y+zz6&+r@s8O_+KtwJ1GtE}=Ugx>+l#%iFZI}!MB8`VfF zhDFJcdV*!JPS_ItfZqwTCaw{R#Ho!bxz=Q&B}#M#0`EIUrQPS z2SGvj@tg;9vK=;ruAjHoEzM_~44TDu%dDScB|~L0jQ_8yUNeoxa9Pi6c!Omoc7x9F zfMcWxU7$kVQG3mCXrGG`p<^^ob8qsLKy$k*B_E>`lx*xT`(m^Mo% z8!f-S7BqptwL~#06uV@WTx%Ev^0g zOn1>-))pJCG%%A>Oq;K{6;P+x&{{CmL|f6ZM3pOa(3(a7Gj;UOADA3@03o z6J|yl_6HYk`i=z|HoRLhgN)^G-F@dqc!A=q^HyrvR|oBK;NLW#0moc#6zD~@(}~0d zr^RTb2^_a$?9^?EQw^TSjo&o?oy2n6z}T?O$TnJxHc%`qn`@|uZ7K}r?)HggRx%|Z zjjKtJoFM1swZi@-jlKvWJZ#h-zolk`ndD?OCeT&cdG2P6lK~t5+DsObByQNSVXLvz z2pbs`3!OU~sffF7x0nuDq9C1@z(W}yt+H)S3$%}3lU@lMR9xcsI9B`Kp)oE8z348d zf^%rJFcV=(xM9TLql|zxYaiM=5as?wXNSDyj8*U&ET2zd3SF9^9Bm?|^tq|9WR9P^ zX#x&#@0b5~@j}h)e02nGAf%iI++V=6P!7@{y&g;5+IiJ}J<^n*$dHnN zxG~W}ah+pJfb?`5!*%d6Tz}Rz>q}oYEGW@yzViX8;I^RHaPX_yXhAtcu_q~VjEd+4 zeyBtNw)eI%d6M{Pg}l$gHFd(<(sbw#2Ak7KmkZ(3~rs2gSA^)~>)o2ABeI9*~^tPvWJSwB{3!dVZcH zGNfLf6q+W9qVVtuk~&YTOr2Ne4h%wO8GuHdUrva63awGT81Xk4`QWA>zaQyrK-BwH zn=g%c;3b8rm3A;SM2DqfSLCqW!S4!gP{Q=q%d_ch zkK7qe-kGXSMJddDxi~&-To<;npk(Cvwa!t=WO0(*fRf|i_~sR|dL-o@8@4XHjIgkU zVu4I|6BW_#SSLK_w+u)NJAl`xBRHN#{i(a8Ot5GXvRtE~&?@u22f&HVg*4j@UZH)R zVJ*032v;hL+VXD&s-_z1$uLlm$b_ymYRfYkTqR@TNJGdyx zsT{`AJJHs)G^54S2uqN8aE#)NOZx|D|86iXH|MRINp9M3=4QE3Wx;0@+eeXJDq;<@ zSd;~hL?T(o!&l45-^$z}k}BD$Ks{Rxium9)1&S>^l<=vSV_Dk(y=_62szrRA(L@W7 zeK3p0!@2_n;oH4S`1rC%uJwu!s`jmyYswtz1ko9?!cgpe&(z@bc3 zQD_R&335UDdlCh@p)vuDfI6SQZ3f9_sLcz7E93(VD!U|w*3rnpm>z^jVjML1+xC-s zd3j)*YcjuV0g$bf^SAIaT?|@ODGb`dOw}tQiCftI;T?$qdl|yDEhJ&2WzB{obooZh znwDZWQe-_9fjwcsnys&ygLZaL6h^)(?GDNFxZ;uz^^SL>F{008J&qkT7P3s#ZhaIc zXcPv8`z_{|w1oc*NiYrWnu&_GGhM;RMpCDU_JKWYwJ(-4u5x%hOM@?(WI87EI(emF z%(O(7{(+nyx1|G*@jCO5dxXwz+x_zG6Zb7JH6C2t9ozlOh#9kjavqpuZjvRwrpbd) z-(Auni^7zlq)MmH8YN2bg5?fRVKMgc8P;`S%;1hQ&cUTIGn7>izxT!xH&?^z^3yYu z&XH6b_H6bUtt5FA3ubN`N}HxI=iOF7G-Z#+{TX`ZHk@AEFCSFy@z5OgND!nj$mbYE zmiT0l)q+cGFVg^33g_K4+kNu^AIR!9j!VpFrhm9oYt=a)HW()}H%WEM!`XC)&BdD7 zaH`16-usMjH8f-N(NV7V1Pg2Ed-PSm%TU3(5=2f*VG{+1*<2tXE%!}j8`X7z9i(1( z6bM)9h2w8jMhU55lRrK$cSofb|JYzrM4~m{As4s>>ueY#4~*eVDIw1{3Y_n}+E z+(El6x<{%%UPK?Fdjy+!FqWjtO@C~$>an17-X8Dj@Q>oYd_~%+SVFpd8r2PRvE+N# zf&^Z-OOIe#$lq>|7vXi3NE>dyM*sQ+ApIX|vcl9HJj2FC^EoDy@~`L)X3HIK6U9;!P0AGi zt5p?*w=oC9A1JDLnGl6r3?(#o=?8NfL3S(0caP$X=(OzK)Z3u^zFt*CALc(6RFmC7 z*XPFgr3Ne)4BFw5J&MDw`JuJ+ZeG!>Iz`EXG@lK;S~_2r=hZ4)O|qw3)68(k&QqiE z{0l2$XYztMc7HKo=X-5u2>%e=FJP3-b1sc$ZMXXiUHQ!cXoGOe=t zgS_u0zCZX*Rrq$gkgoEr7jA&!y9VA#Wuo9XjQMxebHk6hBR@g*N;kTd1#k8RVF#$w zJMpiz5x2rGyczZVs_;g2s_X>G=G`F6r>y5B#)Hv zulik|Wlc-;BICi?O1)?>E1pYFK9ll)i?hE;)*S*#sC3Uw_L_F4AUIa8!5qj$0pZ!) zSqO<|H%!p0kl(w_R2mj8^@*P!BpD-RA8pt$R2aE7g%n#rff{5)O3*E7k06JR6x|QI zC57H}h_e(y{3H#!)IV{_p?kg*^Ew{S^1|aebmEsbxIVdVHAz<0(ET0@ZvllvSs=bS zBgucQ*#DMv9DrjXGM;x|jE-0zF6Ot~fAX+%8_Sm4hS6kZ%bor%vUyapWr6*nb@Cee zQov=uJpNLi7WC|*Md$$UmF%;BF#C)s+W|(OP2ejdihO9wH-DU`(=?ely6F>5|LV#)eId8N$vx@>JO|5w;Tm zKT9PL_gq@AjadiUz!=&c*TrOndA1ltBiuWlhrA|zxXaOCO-jG=^|NFpx0sy`!vg5p zhw1iiq}cTo%tFK&e!CJ4N`<~%0IdkJEM++MGT8juNo#I7BWwEv26Q=bbm)xBctT>> z*h^0AsE`@ZA-*U614*>uwFdSr!{$4kV!`!D1=3`n2IvgH_91Q12}lCQ1cAJuM=h_E z8i0DrG!R#g35uRl5|AM&rHaFLI4==R5WSqmCSlleGHh+@*C&fivbIx zLDWPP(f#>HSEAlBn2{Ybc=t%94V#fGMrNdrVoy=zI2BPZFLOBJb`EOvlR`na9XdsH z-DY_bdbt?`OU2)*K{>g2g96lB*m9p4%GUr*IZJO>%kt1a(BJfDTY!_>UEwP@svQmM1!9 z&oS>fDv%5_L0n#Pu~D4N&ZGx_nsn@oz6%;I!>p5~%c0jyZFbl(tw>ei)efoV^*)*= z=Cu9iB*U>%h_uF49_8XAept>KH@^$iX8Ik<#f_@LkaF?8fKx)vk{swVeU^Z#X_6@R^X!(M!Q~6 zlzFzfGz6oj>RQJ=(g&*3Bug48LO^MiyA}6=D>YM<>y7Ud0{42i^A<*{7B;AteE(MX z5-Q%W+G(uP?k}3gF(T_lLyzg)?y0l>W?DUGrXUTCIPMwpseB6h(V-Q2)3o?mV7}9X zQkV^X8y5)-q&U6|@!&w*2>j-<)q)W4xEF+)nm!TVrgv zO76VT8dFQLM<`NFMJ)A;5v5K`R-JNhbnoSNLCz5SV;A*60Ir|liF;6*a?h_Q^aNR< zhW=quOtGmHS%)RK9UzYxJ#&Mft`iM6)L;-$ z+y$EHp$ncqE?YmhN50=DPm~8_?B(JKL6;?HvW^YlhQCFh&X~>^7N2!VeizAFZi|l% zZ*cY);UkM;K{+HH3RU=ZzEPwP@;eRchM!%Po>ModaTAQL_Xz@$io)FY8lYb@5Gb*8?vY({Wu}8dg-osVdFT7?=|QnN-Mstmb@s4T!N3nPNj1rIME|(+dE|^EXq7mI z23v9Kl?V*yXzY>U3$M~AbT#-g zHY_1kY30`R^i!e(tMpZKsg=lTIsWYdbF!1f!qj2O}o`jkEFV#a@> z`jIp6%_Z~kM}x8fx?;-%v27N|$dl|>jY)}Qe7_7_;BS2G+eNd#HtmVm^ZJDkNwN)l zJ-dx?v5jIuu}n)vq)E#Hug@)#EPVQ+L0v9>NUp2TscYzk3-i2nXKN%{@C1inCXHSqDOrItPa+) zWYdK{MIedRIj3v^5?12*LZ7o@WUMSwv7ucsTSpI%{NJUL z4tI3aQ4VccP@Q5t7pv!|i313pzj+*wmD2)`6iem6 z426;2$F4VL9~Nx!TE)+sdBkr!bQ?`5G&mq*LNmPFka5guUiD0a$*_~CN=PQRU6c)H zoemipn7tGW?(Ob8*(q!l?t1g8G>U3t@_`9vqgR_thNPDq^=M+d zT-y~#J+k<@9!o?Cyax5oAKVXL``-RdvU11y2i*p>cRGK18%-`EM~{pMGv$QfRLC3Q<0AzUb;) zFj2L4>Q#bnX#bp+IjzD0=3q!abkjph0#`4jde>i%cV~EB3j$6u6k2MVQv*rzD%gqL zkROG(GgljCJ%GFzK3r`WCo?p8+U<+JU-YPPy*!RzARcsXS3!maHh{aBF!hoxHO89&^+TMdyGL+vF7=E-S}h%$Tm z*H1TXH|6TES*N$y2o6sPFR}a%>vW@M9+xdb^MV~oi>Xf}up}h}A`K>ngoo^K-9h=+ z&jAk!f+CbA*#LCm%N+_`_ISjMW3BoaC*SPkjr9aKJ2P*|FU|!U?2Onxd@I@Y%GjCH zM!>41*fNUjry|P2QUf0OXVY==8rWCg6J$ZrM3xl1p*ZJc5(%Bv7$M$4+Q}h)hitoO zw@ZnrEBMZ9t03T(%DW5EnvFq?1h^aOLG>>Yvh974E^7#E3e2Jh?ey<()cU~t?kLRj zWH2k(oG}9KiIFk+Iaat$n)EG?-S5vfKmm@d*L&fE{V4$bS2=RBALo$QH%3lN1Jvcr1G4n>!I$nHpJy%1H=AgbwG z1~o4U5R!Hqz`&`%bEkzgRJn-;qi>ab07Od$KUJK zpu+fe!>zwgf3lz2{;S`TO;g}&IvhUv6brlvS)l(7U$IQ*LZ42>7TI+unM|Y4sWa(v z@m1*~>1Dr$;5zvkUksiO+TEicLb2okKS6-k-}cSm_o@f&?yHM@&(eP#wELe6yaYe% zpyRoz96MZ|8`3HVIKjnsgBvmBxqt99z-8%&ZgC{dhPU6a;tcC=+)1&ZZ<9+!pwEJq z3b)irp&-?_-M1hZNa_Nzp$(9+4>io}KdG!*(P5+5dgV$Epw&@V>T*rEJ%v3cgfcEuMee`H3X zqCwfNOopW3`M?g@6Z#ho9%z!4Okcr+-j!InzIO%sH&bANtW#W`hI^?*!BY+^WCTy6 zKXJh^w&~hI^CXyu09`=6zA;u9104~1t$3ivr;3)$H`h>VMf5#|A;%XZ z;?g|&ekq*89}Ahd4HxE^Arr5cWBq8fUkyQVzcTm7b2M7sUSEV&gEA!u*#dBDmJk^2 ztW{Rb-4p}_6Ix|8Y|!vcmUO^T(@k={Pq1*jNPI+wN0azh6vq{ym9b6L&h+`j&(tc> zK5DG9UXKH+MnnnMgQ7X{-|;f6YxzST6q9sr8!_8mv7tP9_$s)IVs}s^kBUfVi=Zs) zp7hC2eQi#i_@S&`xobdNNaN#XC29(63?Ba9wP@oK%2&MCJl*8;_3*A1>!u9?6P zupYAG^>1Ibb`C;6iPnG&S2$R5{R!nS{w6g*M)8lg8px_w1~Pz{Em<;!K;_)9N4th%6Dbl8Eon@lOS1#?2I%(Zg`n7*6$I~) zINOR5Q_KD;%O9s@<5}w}^DF-dr3bbff&cHGMgRD~fQm~$F8Z7_aI+gW?56Y>fxDez zTPSjcia1NB3-onyAd@RxO+~*^=I}rPsVZTaCum$KlO)}4UGg$MDlDBQ$q?$xRBd)a zPZ<+h55p?KB}Q%P3UFFcBCrmc_%uiS^6A@x+k&M|OP%6@Ddi|>bgooEiN;}R7c}`} zA}o_$7BXa~*Vq8JIh&3Wol>UTWALg$UhQ|=cQB;dZzV4>;99U&m8trizOCqX1AZ;v zWg(h0aFo&Q)~W_yT=`@9w}qqkYr*9&`eyYUY(8$<)@!p_2s4ekkpZc^)3T#b12A&hHuQ(Y_yf|CF3~W$e*SBYV_Dv5gct2f0qL4Y(sufy7l>fWGVx zvX%b3f)WL&oxT;MxsWk0n;vj%Vg`VT1U5{s#XCJYuIq!)mxVQp3qbuFFV@@u-aXuT zou#3Z2czpvG8|rKsIcwfu&%SH!exIjv`y5}`(N*n!7l3*e11`QuP{<{(-kU0z>e8> z$YAeU}6^?Mi{z4u?-aT z$wZ{QSNO{t5-$4)`Es!^@ifT|K?@S~^Pc_Z?>mEYL;PWoHR zrlk7x(Feti>RjHz>DdkkLk9R*_|?lt`O@Rzwe-R1U9{G#AuwMMJ*8W?b6yg|jgxjl z@$*vDJiEZWEz7&ce1EC|SvP*u{C5&-!=dPGBNT0*SXheJ=!=evUGu{ii*y_U#jYQ3 zb^#ZrTL>}0Ocm1KG=lBvRcy3UGVR%OHTQ}$4$KW?<>TgYFoN?xsiy9`p{7j>HtYPD zX;A1?6uK0NOH}y(6sAWI!^0yF!>Wb3yhddLZ;elh3gz081a)#)g2b13>t;qx$cCFU zMsWSGG9E@6juS?xk3TsP`bDoz*Y9(XzO}+n6D3*(RB2E?+HJqxH(!R?gS$207Mv-hAURi)oZ&*O7gkoVYRY*m2(Ac34o@fg5 z5E!U6*tw?fR72epsOW>XTHI%s)6in;1F)$DGHHELaSEfsvXDH{fMewRqRPViW_q7$m;&ZlmOOd*Bvv=v`BkU z?OP_$U}NN2ah|AGG2pBjU<$-rowEH)WXA=X49`|!obzVSN?=cdLgRa(SS6(Ed9>0~ zH#5xWL1Jb=D;?$plF5y20smnd7Uoh+WWxbKGtq7RGNs=!AN5EOXprj|`gy?uP2}pN zn{7)mWF=@=JWT7F$Q)DQ6fUT+;S8i1nUy~KwVvnZEel#$=U64!;au$pJLwdr*|kns z=1{?|bBvZGNZNRn3RD(Om>&zP(_Zhr6G5Y8DMX$vHu%conI83pp2`C^Ojj2(Mtrx+ zkas&%L@)JD5~WFpyibsJ<+_=*zB%+!)dAl-ujNl4@;<3NuDZl_N_P2bl~IEBkb3)) zc@yw;wAH7}$%xzTx2>g3`_q3mIIq7d{I>y8&n=v5yNNS$*T~gvp;!=JX`&+fNH;y; z)Ja1$6}t~QRjtAY4!}9%H{{)k)USDx64_yQTynZ-48%tJX;OKWz!94$$f6TTqF|?D zM=<)i_3|{jOMXjGA<$F`L8lf@>M0PHjfd<kiiPf+5-I{y&M8c_FM1`|Tf2DLF_H)le;K5mv4(6Tq$KfNtaBe&rK+%PDm_sw za2v@=gRlEH@bGMcBqgZIquBxOnZ8sWD?c2hSJc2)J!Hr)!L?3-oLbg`p(V&NV-rR` zJQg?&yLZOV{)hUF0Y8%;|8NHBu;GaB5+jRQ3j9>aHcPq*67k_hSSZu#RL|F6f<1jezz75pSi;axs{ODI zyv%*6T1eeb9SAgnm;&m&@O9g>(oJDK`Wj+-~(QEQYccH)(1i{`9Z&Uwo+HG zb!QGZ;6#CMj3GT*y-_UGHAzb0AjPzWhfaQmX}DN z*p(ECp(1dyfp&!M7elLQ9`yS@o6@JxZ~|Uf0G|o?EuQ~~Sx3~DOnn=honkY7OeUnO zPx2E2BOPKWJ;%-d*Z)^$*@8))fT-Q`lg#ZazM&xYhf@qTLf!G9hAg(>wJFW$Tc1R+ z2^3jLMIfCKHj&`^p{WoTfC<3T_WuP9I7=`AM*sOq8|NdOFfsW~=MO!`l0vd!f7eV3 z3CT!wy%E*Eh@ZpKc#uFSbjkMb3!QM=Vj+em97J?b*}JA>R$Nj@HjD~0f(ZzPR$+%U zj$aGjY=iSfh{}P5(Qkn!8W)zrVxn=eU-U+Tg=r&9`8UbVijOOFP4Q^9F)TjdgtZ}i zJbD=jC;1OJ=6Y`oY6(n`+;?vWzIME7{GYYOPuJLGA7``u%J6`EeSrFsX*Jv{Vmr+6 zy6SzB4jMJRa#i*63cr2QR{C?`6xXMUCVFea0pQ~f>fs*{@#U?d{s#Z^uYYSlNOHKv z)@(Sgf7HlZ-AA!|DY6>^v7kS?%&|HgLY&HKP(Pk_Gal<&wLrE^LTh!4W0Blo@?Ok)ZHBUM!g!`U!OLY&v_4h(F4AHKzDXnRpinvzNyfx@~ZME2tGlyUKpiaCpoNYb}M9#63u$qtvR_fn!NRQ zK~4bV;q!T@R*FZ{AjGGysIt=j$*S0I2x*79 z!AsR(kbX^ICS5^Lej-m&;WEI0&_Ei!8QQ+FK%o;V6OgxespCz!9xn%}3~Snp;j=U5 z$sazL6<>4O5Kr>h{oQMC7#69Jk3MN7W!x4i8{Yaf87)$$C>Bgl4Hc2h#}-DUhQ>x| z>}kfRI?7RM^qfGT$CT#Nu54$Tp*5s~Mnyv8EIJk3FZ;x)R{1#;Um`(F(zHDugLX(< zlS3&ajsBeN~6_e&uL+AeNqqHw4y?Z4* z&1W>vWfZ@%;T8^0bo3uGiaR-dpKaG??A$AHf7f7Wj{fTMZE~2Kp|Rb+jIn-ha^76;?b0;&qrSCdm-vp)HV>4}Kq)XJ*Q=5A(7VJne)+TWys?RPkGO~4Mt?%m zd~=yDaG&o9(q^0>8J<1#2ERJtJ@G@wGp>b_0XiM>-5u}de7X4DB43VQE)L%gQ!B#p zJ-4sO=BVP&zG;Bf&5)|QB>M(g$1^;1X9r!KvUtxknJkZbB?2At}IOL1WwrO@GIODC;;)2q|hfGz*Dek>r;JmPwNimA(Hz8)<|J^9E+4_rLsR zESX0e_J7P|9`{P}Lpy|z1-oXp%*l7H4BG%2(0}dn$>n8o*32<_#-n*-sihdr*kie8 z^T)#44-HF9`a__2xnslbOPo>t+#tn%PLY0;DNXi*+Gwa#KqV$*^3ey`U^@*lJunoQ zFc`8+0H-w#NXRlgo5hu|e@@`Vv*|&acIIv;(#cITPhN2_0PwGx^M58WuZ)kAW#r?mr`QyVtfnIV zdR2-{Q;q5@emSpX)+K&5sSn)AU*+|g>tjKgLk3wb$QxEKIhJp?W%pWYn{=+OWkXu! zsvm`QJuoe)GgG;iz)Rp2f%Iz@FvIV3f8^IecleB3;Ppj+oJG&_MFWh}IrDzONY$-2 zTb$}V>qn1GDj26dWENZ$cT1Fn1Or$R|WGBCc zht3T$Nr4H3?z9eYc2ML38q2P^bgF7*6@|7lM?#Ow>V-O%m_z(7dMVqe(BKYRt2*Fa z=1?gpR)98~5;I(FF84#8WE?ad^d9>ZCYuHY4b%*}^Je7yA-lfUF`_9auL5&?FMp3m^px{V9RHB(7Oy|9@#&hg3}VFkLt15V zKp$`xHmDQWG)b+jmcH$KPS+*54b<4L&#m#n9?O@9>_&X%FF!a|w7mQiMt8;4*@;zO zbRs*9&0@t1Do^MFy507;{6Jq2MqPmGG=AbVR?VuOD!yY8ffHXb`S0~Uw!7j6Xv+upW zE9|f<&~APXZPX~8jLg0KV<3!&=Zg5pUGTnbq)7&vcg^AzkcMcI<%0VAHEBMLglV;k z#nbMG01+Q{%&i7G*dqQzk2D`V3|QaAZ)aLyJ=_YJatsgT(1VH``z#fqDNcKoO zfm5ppMktbLu>Y<}hCZ1-tv+g?TDpFX<~o40Qa!+orQ+O%qk(2B&g;Y-!a+~$p+qUA z!w#FLmMfR?I)s=l7`N5QoZ+w(7Uu3BujApj7F;ovPd1}FTrbByWW58JC#m*bF&9>X zET+PxUXB5{I=W^S_Qb_BPFZClxUCvCyE$+F>F@vTb;EL2>i=KM$@W)f zIXhJ}Z4fkWcWaxv_ounD7` z(P_I4{ZqXsmaV4;VHKC*ASNsuoX@ywm9>f@kd4D7@(^@)_tCq2OZcNJ(|R7Cw)!Rb z)vvF1iVf(A-FWTSBz`2mc^lrFZ#TloW{ORv$T~=A^O`{9=);?*{b+UJr=n6o|Juo>S-0I~R1gf-z5FM*@v(~{(cExj!<($FMz~3#*wqwSMMb31J@QnLFj*IdmFvgsfdcQ; z)EE(N*RhP<+7thh`LY}rFB!s=yohXgWfqnaqlG1xVxjVBD;0sM;s=V;((`0v zP=4qDzk{xXX0|$E%G67Ko$&7yU!+5(Z->jGvHxwc`Xp3pFCkkXwpbq6Ij*r;OR+M{ z)I57RN*q{n+gGVyC!5AX%!C&j*#a?ULk1O@63vRKY4oS6V!7s!YqTG((IjgP+S1A^~p$dT`Bg`%B@+P<2jWGAparUi_b|2G<}Jo^^H ztFeN@%>A6OV!Kg-a<5~5H|2-1S)bJm;|qq;BSSjqhmgZPJxznj-4v!l*&)kPRS*rn zD3OmBJV!zYC&-R8z&IV_1lh?m6csDJZ16~;*L>##Qo+q5v0+EG*~lX~L$N0*a*T?I zVN0p}nGfV8P(i=>tGB{izw(!hZ~SEo_1V|s-$Zw{QJp8r3hJSuR~`ao_kLCKqcvYG z4sZ7DA?4z8Eb!)jt?4V};@YqeQ@`Etb4~R(^kba)&Y#L0NOE|`t0jxzH?1l zGOI&Yt0)(5rwd`T1=iDN-K?ZJsDX{@5xW&xL3lQqTz4!NU*OIa&Ix};wvyW$t}yLe z@U0y`Fd*uCr>J+xCpO$oa?c1OZ4`TrBF$6;D%fLnEtb?`bZn)QrVyloa~ULP)+!T; zrW)RYLOscopnS*>#Nx{WTWL@M*70|1kQ=0zIT(VKD7$%AcsE=wcxfIuqyjm7nj}r4 zm+Bj`LvCjD)vyHwI2*hV3oeS21b8zzYg$#W;)Z-bG>qu0OfVYLPY!^XU-b<9rFz0J z#5-h{1A^Ds#iAow1S0U4xOWqUn~_lVA}n#UTi>pHf2XN;%%uq2cKteH#>r2R)XP^n zoPgCYQl!Dg`Yt*$AWzzWGd`y`bd?jb6P1Ux0Yxhm=Dh?2FPaEW=SG1hPUDaE;-6IB zIXll_g&ZH$rjQbD>zeKQa^#$m&8neT=%cQpBBDVFDN)d&BZg^`?SCD3G}9!PN$ERS zr;QbETIF`%g5V}upAYN>+cB^wC=zc|WsnZAS&^_a-1V9UiDLFoKkEorAl(eI$fqy| zl^LWSNU<tKTc+zc1eu|;bh;qft3{q6(co%;HBI$60A#cUp zMEfT3kR8gs6#}EwK~;|wggW&l^kaCsBmDJNec&UEKiWgw76O~Cgu3&skAqBgh@bDj zw_*R;j5cwMU*61O|00(Rr({7keUW#>Z@H*~+z;)SHL8u3NgYv96KCFI1H}yDE z76WF)X&t|V32ig3NenJchHz~QN#N$v*s!U|H*#sT6uXfk>#2x+3(B180zaSgXnvR5 z3iU>1k;kASn>T28T(&!)PNsjC%Ucql^=b&jGuAa6W@pl3fIRg@%noJM6aOyQ{`YVL zK)!Y4`@6_N8{Tr0AyWeb+&|)uebH z{%j?Y)2~QSfv=LT6Gn;-glUyGq-BAFc5A)Y%LeUMx~+$+*1KK`7__SsKZGL>-Fl#X z_Mk9LI%s!BQUqgGx#E%asuSWtyIztGQJwqrYLe#DA#4P$v>v*OiFVEP9yOe}1LP@E zEP63o)R*`TqR?!^kf^=y{ttSsAME7H!9oh>uB1>-yfqXK>7D89ta5Uvomsq}AW_ zDgD1yGiB(2L@q?!)ZJD9izx7Sqf+kJiwrB%hecEWi^OqrU2HfSlV`M!ZKBu>=x0S>eU-lX12n9ATy{C^_BpA{aj0;C z<`As8IxO84*5Xn+Zx=r+X#7lz$$W8z$GD|cO@6P#!SR27%K(oZGkEt%?w*IN1mrnc(N(XM$wwUt^AK6HNBS<@-BT9ddog@JuE2X9eCrTzrh>50mtHp6b5^0 zGCW&_JKPF+@c65J>*YY(u8L>1%2G*(yFS&e8)tVo+zMY_AUWuSrec*z*YnQx z5TtLNlL#B#4h5zKpp~s3RDX*?k*0aT2^pt2%_H}@u*fwqbCJmdf5BUBR^Mi$FZ0Tr z_n1P8OMbwHEsmLQ#udzPLz+siui@2XLln0i zIC^hEQ-7Xl|Lgl-*Q>i}>Y(LJ*C1UlEb_p>B$j+=Fj|QPA+7vcUAIUZN2|IS!g%4`@? zW?0H_5lfcj0t-DRol58J@dkQ3=cYhSjD6|6cBXwsF60Cz%(;OX;RKE6XK>t(r2gma zA=8#=o82mz>E%X}soep`W!M*@BVb0mkqPOc)j%eMCsLSV*9yTI*Zd$1jtoHB}y$A&PfKwagKzhLE`smEDpg?Yv~K|6v?6@ zs^7RO{kRQ!M$*|>XcKreAGd0`E*F{QQ#LHT6#kdIA4Y#IX37yiETv2JIS^hhgzOO0 z9~w8c$Rkp8Dj50RI}}-};s9XR9WNQn(E(sM9?VbTSk7=VbX33Ru|+`!xG4TxvzhFC zWkN6z85q{nSV6G|DN+id0^Obmw>o*ww*(E!JjqT4Z2$E#6%EQ(Ar`x|DN<#agYHzF zCk0|nixf3eAx?}_kb;0LUDt&Fvb zH#~N{4aeS-*JDAN{@&GZe)yJY$(@;6HS99prULa3tXb3xhxh7OxvukX&Mp#bcJVL6 z77_TkbUV?832H|l#e*IDQ?pn;c!};kdBMVbR zvCvn$L%)3e4VSN;1@~pwgnhGnX0$U2ymb;xP8UIaaJ1jD5UgzhCRQ2N31drbf#R$) zG)c;H;iy)5Rgme@sn{YbRMgO^(@qnp4xazeIn%}Jj?E~*|HO0{{j>njoG$t0$<2Ni z?tZ$Mi?Z#~GhlPUKt~HQ={?f+;2krw*e>}*ID$rhmc@_L=&td9^NQFKtvKaku)&sy z?ZdZ{U9XHSIc;Q1Dk-*%BKv_xjO=FCLrnD~IZ0AUAIWvz&D+S72VM|g@xZHDuyTM`-;KFJbdTfUy^EWHpYe{EYQU?Y)5pKV$V?IBo%=T0$SycVD0?f0Xg(7 z!HtljP`ywe(!|G5t->m|%i>Jf4wb>XJCG&K@I?M+ydGV8sBVW!nKoT(6sCyqJ#LWD ztueoDW;5h$b1Pck1lh#3i8E*TK{fow6RXll9BAP94H3-z{EUbUF|8 z^nexhwxV}_P3QxuX6O7Swn+vFBAqOS_43QlL4gyno_8sp8_!w?KHkFrs?-1##XsI^ zAggRR7qr7@^3y1G6Gb*)&wV+)ET|VsO>(80jX{m1fGKc;Cf00vy-yEW<+?HG=IlB} z8~xBpbI@;D&?>L08CD6mzibXhH|H-Kb?oOk8IwtZ6K}6|HX!8vs?8V38XHDPp%Fs1 zQfxX!Hd7J3a|RqM{Itpy>TUj*0Za(|#JN>i3i2Uol1}Iw%!D2BPJSh{)ZP*tbuQ-( zIF_p7rlpFO%pZRjbaW^g|HX$BN~nGDGhQD{lF)`j_-2xXJy0Lq6|jYUpxEr3XrIel zKBbi;&)w|a99~De9Z6VH>y$HvztVUmWeO(673TtyIry%A9f?8C&W5f93|eKUz4648M{e>5xj1{ zPtgWD)Av2E6C}!p?F^WeVPTTh6uXKdaa07xC$Jbr(=CL2F&2D3a>g8S`MB~Bj`;X3 zo#qj59>;^6%*rGe&1(lt8=7r46lFq?g9eXX-HwO|FsXlYcLh*)WA2$>5hVBfM$Rd0{XKA;2;Sg3mJil-NgF1l*^ zWi>pwsfnrK4cQ?9Sc5WAu-hfis|5-dFkF``zaLtuI^v9<@=^?23_8;#XN>P(23qr= zTXyUxb7cmAiSLR3KoYq{WNjFF1xCXF+xnpD-si$#dsN37+T1vG#$}JKX+ZTF&;0__Stvk`e^C zG17}f!?p+s|1(vMB!1%B+Mms#rKosz$Hctap3J&VHW-ej%L9&(Av>FOt{~DnJSwt^ zV&f*9Jd*hGr6ztno~|WxJRYC>vN6ps8zUw6z7_C|(y=gN*l?uP3?l~e zDxFgHk-f}1VE@|+9BnM&Cc-b<>MJpBVaD`9y+a_I3+Smz%<<5co_(+b^c!# zK3eDb#9(~r7Dv)v8BYVsU55o-cTz0WKITGlhe3r39l0^aKkA+1*`&}q*c@Tp71otB z`jHf`sCTZIyVB_huyYLAH8DNH9Ukc3*2}l4Izf21l-?*>v}nhvH^vgBrNd}%PUd!f0m)Oq za2sqInVi6j9q;3>3T6yU+l(LB;jZ{MC*Nrp!zP9_Fxo=;Tt_%m=fmDH+ zT1KntQ|$J{{H2As3Y!M1ft6;zY9~Z}1zrm&^S`}x&h%<%CBP<`3tsCzA4Vj?IlS6@ zr4dSkE)YD2$GfIi2=KSO!ylhw(m8k{eXCWiRVIq^r6~f9Ic?e&(J0T3Ss9x#8lp-& z(^}jNIKXo|zWLNhFWTT_K=(~m=KOH-9osf^w{<%#{qc>^DZL2T*LuABCKLo_Fe?HV z@)AXQag}1JAJ(v4^k2ejkYAhHNp}hs4i~Mp{Z$1|fche|Iq@7P;8H)4pS}I6=i0Dq z3VSFt*mczct>X^pgE_+h_P8={9))usUstE&$v+F(*+CNtjVw<$>TS!5#*g~=Wh z6l5@K)im+S6n9`Tv1USx?E08$rr#qk5~V$|=#&0^inQ2rzGf-ESiCQ^Ui!c%JqQ;7 zX19ZagE4BEvOsjl6Ty6&%;p?8+XKKS-e_o_+#s0c`L`B7OL)^V<;j_&NH9R*na=)6B=b`cn6DIu30u#fx-+!O93|5ig#<~2@t)O#* zQnphN#+!J%fOvpeOb!VEEqDy$}4eLD8-Fb8|di(Gp%VNl$O~>GLQS^~*uk0r)0cG28dni;MWH!9#gsYT zKUU|3-Du76NJONkn|)8o6DJ{q9CC;CGS#fnm7@q8nj0E$Hw!q~5h(lhKf@Ikv}KL` z-5*KDYm?wBwZdU8rGzToENmF-^=_O{8VDSipenyi*-CdQFtNA57fTTy&CK(x9KT!I z;)(3s*^y~JdZ?x>4P6<&Kg_7;b$Mg5!w6!ryt59^`|&UTVmpIknW(f1uZounuJS%r zUy|y0&GE}*<^1fhi;@CKp%il@W*+GMQA?PO-p&^8V*lUe{9@T3<M}dbFiFumcDx5d`L}GT6n%e;-=U&3e zn~nHF(^ukyEWt-n{||PN?AIpv&|3u`+bJbmE-S6BnMmsQL#>Gde9=@Txl&? z7kMST`>j&-O<|p+3KZY*yauXc)!X2o6bcSJnf5zlY?J%hOQgFU5a&3d~GH(I&8N6= zh;rJ>ryQb`ptf5>B^1)l0iE=P@CN?o&_fEnxG8?4_udJ(bt*n(TBbM94g=HHveuB^p<~?B228O?l>hvJ|#C%SzKb~1Tp+8~^jgumvF;*N8z;D2I`c`2< zuwIN$)~ePqpH0Rm4+)@-0B?YgdEw?Dc(VJD)Jy$u|7yL4DX>(Egz8$^ZYYgw6dLI+ zfQ^y9Wzx{nU|u*E#Fs1hz5GnsGG!Z7I3QWiknC@lK^z2$UC$oMMN7Z;-(R<-QQ=a& z@5T!SETk*dfd-~sUhAnT3@wsj-@9qm33L4=A*tkGn2uf`>hY=^uc;Il1I-_n-wg{} zw_^wyitX*YH!}DB$F^nLZoBx*QizQ#kVbF_n}W2eBvJC4^CorByW;AlXxu`O4(pHu zqbyY7I#sIYWv`d3!C&?~XPxI1u#C*E=zsTBhX<-OAung;R1leF`2b&*L8tK*1f0 z@ibL3<2cPfTWP$>Wq!qBkUW{2IF5rC-pL6hBNx8=);C`D;#=f6Ntf)2Ymp_txi(CL zF^7SH(tMn(=BRJ@ZUz!asJd*6!pOw#^2-;#av3>+_e>d?<~1IPKKv;`_l-X#9$ z&-as!Zd{#GX_X1tK`EgXy^u;kwF1m$bSdsf9OONguL?dD*XL2CNRwO^?hC~wf6#M^ zyClu_Y?bwS9FV3=s|~E?HOa3>l)%}`(k}kaK-9Bsj>jRYjxO;(q&li}#R%jWW}bKx z%jeJ`^4t?RVTKy}%eg?d1RsC& z-&)lYAI|lQL$jTofEkVu*6M%cEswx#+dSDTP!@eJBtN)V zyao9Sf+|5fw0v4GGl%!c*A)TiKzo~iml+fxzxl)e(fr?^|Na;M^;_wDN;#Jzi8djp z_3m>Ialrw0e=%5n+>#Iae`KqS!NrHTaqt0h2Lm`TR#VD!ilk8qr{apEfc#zuwMp1; zyn}ZJmT<8b2o9Nh8;7ZUz+!t&&cg+GVSD5o{qOatE%VzW3IAA0O5AvU+hAqu8cGQh z-u+a zDYX`?LA7pdicgCHxxb+K(QA}Qet&$%1=1+Tm7TrnjM17o@t=-^ros%S zRp@#+*@vFz9E!_96kMR2JkEj<^|D2kWW#HNk$qMe*-9y)f}e>aTYVC9CSjF$qgnwZaSO+Hz@Rd~8;3^($3tt}3@Gky~lFa%j)2Y6mRLPT|=t z!?N`sipY|obWpglqhRTvD1xbNr@$zDENG=yP3wDegKxH1a=>GSZd|KRbNtn)VmhB3 zmTd6f9Jw^$3O`%2HMk-eWt0v@C68&2$F^gwsv;^qs0FC=b@U_eV#OxbcHSYG&~C^) z8%>_u8q!~YHueU9i?1D!Fx@}qr1h2fIwoYAU)_} zaHoHVqCHxxn*ET!cKqzEUI)YOc^?)Th3(M?1P0gB4p`nZXM3~Q$JvM3XD)*!(3XC1 z;oMNPsNLA1uxL>?ee05RrDp^5bMBAZ<+Ua*5lR+u|4`_GfSs`$T(6>IZE*WxWPP8j z_x$c<7;MSe+&1~n!WxHS!$=$UZjY*0r;pdFc0#vIz4VX*#ruk2LX*$yeyejr9epkSoN|s|F5RGLmzS#H?D(Yu zd`7lsn;Ne=qSmU9lI1}c#6>X`Ue~7-#gxWeV={S|!PBbFO*bfd`DN1Mq&sGrawo4@ zwLGXn&MoM*A1lv|EL;z8f)_WJlXrNU;$O?5WCNdabjOM(I~e%S03JraW<)f!PuAfM_~n z+aOu`0Fqt{sUrFzKONdCHpRnYUXn;RcAnhzQ(q2%upUI7n$9&F?0^^~jvevl_0v;q zS8BLq(A?Oou&m)!PAwHDiF&;?X3C-kUgp(bZV`o3OlKd0obpwi9-?Hl<=?i8(4 z|9x`Ijc2|$tY*F!DCIedoS_mxQZ{O*6#0&|>I>rPK+|fSSz@Q(qYGq0Gnh7dnR1`H zYXT4p;CkNHs2{e<46PGfdM zcOR72>*!ven$SnTY@5+E2A6T*U?X(-LDmG)_Wd5MP;HH*6d;))ZWShnl}7a`kf1^z zp3bkBj$gZpm+Nnr!?F)C*6wELdoKqMH=pD6#kG&$wBXA_qAn-7ugxmYCst@Iqm;l8 zxs6IlhWrca3~HKrr{Z>xE0CDfuX~wHVHbbncq9hMF-?Us{pyLPiD%~AP6;~!7^;Ms z)itI@u)^n&6i@4*Hd3=WvNWoVt_ed;z>WM4Cy#bcMcK>2WA_zXV-u&_j#Vs9zC~6R zI&abw>JZH>*)3V=Xl>;GbMjALHm9Gg`>u-o?OQ#qw!Y-~xsUt){RxZ5Fw!_!X|DR; zi{7v9B=aIu0=i@FMJI}ycqx)=Q`b*u22U_2@CYpYb`!lgFA|0vt;&2Ert3|#IF;is zgd>k|Fekkaq+{52iZc_8r8}3+EWl02e3xxbJA6_J+RlRp&6l zj4bL^uO;h5nnq!1U@pBU&cMJNdTSIO*Wmu@*{D^#cg9EQvz&sHThXVDLhAd2_3=H3!Na_ zNsb`nhNpK(I9O)U#fs0R(7T~J zz%Pt#j@PMsy*ugszH3Oov_{%63?cf(MTfk0g>jifp!e zkr|Y71x1!q2^swS>9ypb;>5W1!c%c=QHdg)N`9=Tm(SYgd3w0pU0(h|pG5XAe+kaK zg43_beMjin$49?6l$u62_7iMn)F78-mSAp#{$Zo6Edr@6>d8)3k05c(usaTS#rM@h z!A!>DFjl99oh7T>I2E$f%9G_$N^ou2I4jv6x=XQ1i9-Xdr z+l4$XISx1WAS}y%NF$F@1DfiJHc7*kRz+gtbNO3H-1@u`;AB>U0zk zxfX~FOoQTfj0QK#`K4n_8>pBACvp8B^%{4~==DyVhO1DxFbUPwW&zK2qWRDsra2f^ zMCYihA#n@(7OI+1eQ2qq%=g@MAQy%h7qYaD=vdeVmuC*f(Jju9VfQw-B$E5K&LJ2MdYG zidmPW|G3J4#&2Ynq%<^t!i|W0-hxpmegum-!Y!asT*P1f=E6~R^d7oRg4<`sXI#@N zY;8XKFkfkl1*=`l4Dr0#g1M3 zoFK(M`V5j3qq=_c(*r}9ExWPzVVNy&^~ws)l;zT^#`KWp_;o?|y|2I3O|p5ZWP9lO zu+?vJLjfn#40f^oFdFP`*Bw(mX#duqEHQ8PgSt`V)+n-c-AEkUqksb^qvuJZ5 znJOp=S;I3mM&O`-!?&95ftFuTZS_SFBV^#zAUzS*pq8oIc~CvtA}b{q#8`h+8q)>s zyn7=xopNlIzz6J+7{_5>Ls3Ban2v~E8FpzF&P;l{&tt)?)o&gO1 z+&7}UynJ=M?IQAP5-YPTIVXuwJrXiEy84|DDU;b@o*UZ6E}pr z&KKD}c|s?%WQ@f(8K*~FBJ12Zlefoe?p;VJw@@UPNZ;@5MclMhP zXFw?@?%wgtVCF}stJnJ(qSHvxj4t8gsVG^iRW&P%!C5|;Rb>Ftn+p-a&Hm3jFk8Dh zYoD8kONd$Z`A;oq`RTs(Unb|=7%e?kXt_=)+bCk760|`j6Opa&E`xz_3GX5g!{RIa zHE~+M&lHgA?PYohUbGSTmXb&>|FcjKL|FXBJP2|yA5%S{LY2WR3%faeS#a+(%{8x1 zDe`UXrSHXPRcqDIl}MkT)$4uTzc5(SDb0w?U^c4zqqXXN!BupxZygl8;S)ea?2+%* zsC~h=L2%)o_hUwL9tK2QY6VGN1523l|tc^CgFtemb4?#(fL9@_91>5uR zbL7bL5akB8O9y}b|Jl|rxNYeZOGN^%Iz!G*55bR)&X$B)#8;CdZY`>AJkx?Q-2r5^`za-idAq2D6iJWw zzR=_`=fe($7R5jxOGKq0k6#3RQ9Bg}0+5;KpL)6s$|-yKDS{Ng{UP@zRPzqejgbH8 z5chfPin|wJ-hu`g;RW$kdU;Tv$CZdVf-O-sf(kEubSE+J?$9-YET`4sIo%C834ZGh z)a^hEd>$*0E+SfP@NwTjAL!Kw!l#&0LXF)Ph>&>cytasTUXuz}JT>Q(U9ixgqr3Rn zF3{lF#Yb9OsN>erYbUg+ap3~WbA8$(@*-=(wkdY13cO%?n8D!iR5nf@mIa+uuJBq; z0gRJCGVn3;dsju+Dul6!Zk-qBL|KZBdc0Qz<%-NDN5j!wHt=tZ+jhQ9+uK~e&&}7n zZMWI~U9a6RlysjPFKx4=`^<$grb4=0S`W!6B*HF`7?>2^dcS3p9a(oA3fSz#H-@f5 z`5Mmm{`-v|UL9>2UfO?s;qPRT8&8sSRs+UbN(mA6N-CiuybXxFsual+YbIO}8<-YZ zuQxU{Hp+{5`#e{F0apgv6vIjK#lhOcYH)7e?8Be?Crq_pwnH<~!sTImJ8SdxTLKpdV z!=1*E6LgYbNRtz87UMd@imhXsVq=>B6a&bw}u-}bt23UM;2k_Iu2L)%-F&M#zEG6LI~DMd-{(>6T_*ik3Rl?yzM$~U?Z35P=HulHr;zr+ zs;}HQGMR6+Y}P|5?^C1`n6x9!{_X^wEdZ{ET=gFM3@=f%TY8>kFvUS@CuGrgX-!__ zlCU0foIVc3O}dE}r6qCw9;o_qNvZ?n;M!I%A3b9oRCgXFkL3-Jp3&gWK1G9YpW=?c zY3G7?oS^fAT;Z0U4s@{L@EAPsAuiZvcTxW_C!o%j zF4t|dzAV(ZpFmaCHK>Ouj#)2X5m*_vTbl1%!Yc?qM3)Rl>6Nv;?1g~!U9R6_hrnRP zHCxB;?X}gOa$8uis7c|{b8TRX02|$}d!LHa944I*z_iFvUVF384&GHzYd1}=aRjJ} z-NC!zy96l2`aQ64;;KJ-{{tk|vjvE1N@ffxHh+P6?ZwLr0`?a?-SJbL@G|n!mva+r z%Y)q(FDwPWAZ!_j6Yneu%5y(ePj>Z~F^Gxm!^9vry7&+`O#C|e$m%Fd2qN~Jv6*al zG=4AND8vL(2Ta3_kU4A#FPY>#KjmcX3=uKUZs&%b zZzk1mmsqf~QMl>~NpWNB6j@ISA;xjY%Yg+j`BFS_ zN%|pHSCz&f-O~eLlNy#Bu|oiPW=EH8aRSJQrIFv;V;i$@DbsZ0ybVi)W0sC+QsvT> z6XL(HcF!P;^1u9-Zw4zDGQdcfo5XF!gIyNCtETA{O9QMuh;a zs+TDi*GSGtbEJ86o1(}2ruP=QMv|}2^t@`38;)*zg!aloftuwsjBI3Kqkb-xW- zdCqn$VsZKRBTCgN(=xr=rM2Xmc&DPz13St3JbJw6@cKOFi_gQK*uYv#<_mVv&Z{1J z$^Q)ZH~W$YI_Wegm+!vm&hIq;^nF`$Lbshiv(OE$V~pb<*Fq8}mH4A9NIgjvZ1&mZ zrBz)NEuGjLujgaBMK>`u`gYVK`NA-*YSs8Q$gObreLD&o?&w8l;OMv6e;+sW{^Xkz zT5T(ISqfn-B2hU+?De%-4CL4HAtA>L@ z%^vhTAAGtVN2!MmJNzw^ou*&x`jD*UHra9Ca0)aB1{P-LQ%Y!D+)O3R3r-VdNvohr z0at~K_@y+$p(1<*DS~dID(I>{rmB*4!g5e0Ii|YtMuTuOzbIx4z0<#b3W_~v(V)lc z(ltfLuwp@v3nBA$*{rv0Dag1aOWYVMEQDp{Up@~V)y?sXn48QR+2i0o#U;P9!1mKF zH!ugJ2BydRj-q7J5nyq@LnC7YYV(#bRf-C)Iif>!m8<1v90!UEQ38*y_^+}b{MzDF zMn3-G1Y+doRNS`$kj%BJhPqEFJ1KG-nC+NlKz=)4&=B4t%l0fmYJdy|mwqaF6)LRf zZSdSIeJb7yk^quqiDvmMt!keN*#uTAVP~7)LQz^wS4^4jHa~NrcQsGT91v)Zk}m$e z(f3A!?w!|lZ_OP~efWy-Oc`qa0+-bt&lcHj@e-33Blg!qW+g{`7UooXdNow5Y7$4A zBmsKJG03B&koKrt^#)+7!SyDbi$O96I<^hW_SiLXo&LLonnLNm&{W?!kw~|P3XRZ| z0=xxUURh{WSlbj+D#mf#JwL=b@5qP|pRJyM{JR$1F-t<;BgeVn&W(fP4y(cJ5~Vy( zk+Vp$i;IPNc?xg!n@Nz5E22=uvxD9}J{OFZD_S0On>MJL;P^KF1@V*BZS09elJ8w{ z^-}z=BqSANbu`#(R2rk3c$6#x(Gp8Fa|W|ovOMVXfGSw_EDOD3a&@gfpYZeGdk%&b zNf3SUexT;3O2FU{wzRW1m*5`I5o znKIe;aRi>p=iR1{k(}w^^gw+Ni-3z2CxGXuiPtK`#>QMao0mgUBzSFYR1Z+2l<+{y zL~V!VUv2~Q70yM7z2Exs>)5AbUfch(sm0Ww(OEXt1Z;g92ZM#>CY~lFYz>LwF zdm!t2nr!!~4O|S7%-qQ}VftyNg+dH+w)r1c*9K$m`olfa8%!SXaA4G-d$UPi8>o8| zkQL>UvWKDJ#dULXZg^SU?$=;j=J0&j;>O`GOECi$!W2h!$C#vnHw9tsrh(~*a$=Za z{VQePvkxci_q*y0J5E|lKA-4qnJn#j_q`Kji5u5OnPDU#}>ys)#fxSvtRW#Btx1HDq{hOK$W^EW2xeOO5>_ftUQN>lJw|Oo*5(XP#io%Zk zeu4!Z-9M4fC!4rUrQLW6RAq&w5=sd~zeP}IC%of%*Sm6jn*z12suX&0deDvN?a-Q5 z_U+{f`@fv|wT53d#~%z!6cs`z{-M}*c@b_zgrr9z>{J@iRQ4rJ-CGEf(ZY1}52Y1zOmgM=iR7||jfV)C17!!$VU$%YPVOtn_?@Ol^G z&3X%Si+K`&wukVNgPGAR^xdprCdgl#Z*ELa7iNVMw6C0k?^ zE)|^r#$VcIobB574~LT=MFVMbXmNGYBF z3ZA(FB)?Cdx0YudhpW!`RJ`AgK-v-L*+P-~{>!1K`Kt1d;TC``%IMrqwz_d1<){^8 zDk$Y%3hJa1+CbbOThIgsy&sVc^vMS1uqVuVnp8V@i>KlOE)tPt(F-Cqt9f%o*uT0p zDmg3{)cY!w+hY&&awceU={<4Tk?Y5-5Fhre;%l(bKnHP3zB*T3F>Wz1@|qW&8$1tt zE(ARSI}HBPSML%;y!2aNegV5P8YI&);@zsB*4c(MZaXbviM3GeB1xoGmxL4t;m`Ui zy>6N_Ed26!ZF7Y^kg&OBSa!KV;_=2yi|s7(zX2rIl~Y?}$DkFfMON>J)jJ1B9(1H* z*C@6d53e2UF@VJj56g>x@+F)M;J(3bz|mi2+6KE^=E|>L7+e@y4D7!7QcYoKi)`-{ zP~hth(d-O_Mrcu@$UJX`V>;+QO!eFln-@7OaOZ&Xry<7^B)al5bWQg3SG|SoJnx_k zE3Pvbqhj5(`QDoSur6E@dKJnMahPe5r9mZ|DI0tz`oF+1Ad zl_tk$b8~L*3peZhECFlFlrw)K>24gbmRJR>Ih1lE1qq*orCt>RU5G)sh*<~0Dz+s< z)uBOEMj8}NDl>H}GOXhzJN#=Vv=ki5vvJ}t6xS-;21@Pkqm%W9>u=(C9=%8Amhuobd59 z65+Jr5Rf=|nvw7QE|6DOEZkso?~`%R0}@sOy?Nv<0^9gwIW zASzTwDIt*AMkQQ~?IDTLnYw&~XqH z?je;v4eE@y?kMuqC5q?RAjdbQDoX&k$o{ZEf%lKV2-G@>Zb71;pSGIY(c-Y&!dYu&s!v~Cz(P# z+8@1I_9zdxj!hb?_&(rhmnD8yyb zW8`}1W#L8t1AJ%yf(QBQ!yyOW?BEY_@=A(O`yn~-!(ekRY{wn?$u#{iqs?|f;Q1N{_f1=a zWbnX&YBi-yr%0N)Mh9vecJNx1iP3ij#i9#d_odjSs#WDfyL7L&qZ1>C-ZD4G=-^j4 zbQ`;4-1jN9?p4qDmiQZZTj=Cbz0mAFFSuml{84=#$)kEko}0c_J%7|?DGJi`i0=bC z$AVG!ygv`f0S1n<%)(KQ{l+;Nz%hrw=_?$%m>m#mV%n9IGL<6BsDvb?K5Tga&>%y$Wh*fd)M%-r@Ax+ehXDpQHnI~HHg~yxmm4ak zbzc96Z7(}Zdw7d%Pt+!LqdW^ zA9teVIGM?`lPrm*Tih+46F(;&3X2LubJTl7N(t0E!E>4Ls!1#3M$iu=!L~ATiUh-1IC4){}B(z1*3e z>^XE;4mbo}wud?KI4AIq>|Sx>A7d<`iE(dL z)IpmW(kFZb|F{4?4xz<*6ofq|!sH2xxhyG<>}vS&2Qw^adH>+|x0C(cGAr)uE6F)4 zY<)^84^yNT)gSVhO8<<}A0n?4GWy~mxz(#c0_6=*^?)Q@OTF^_`xHi!u09@C;awS) z!OTweO7)uE=VAUd3a@$fdo;?SCR+p#YLs_-;cdC<0tsB2J~fAK37_5Xfu$4~Os8-w z^h$4y>h~~HHMa_}k{7xTf{IN%@OUR@csd7s&!O?T16<+Rn-_ckU#rBS*ZW*ZDHvEE znhCZ1-AIo*emms!_m9mGRRFo1R)wO9oDIzN#Sgfi6E+6i=jzQ5KJ<7a!s2Q+2IZ_G z#jlO4ft>AtVN6dcK|F6amC!paoey!iQ8sttSuqY@SZG?d5F(dFP;P%f_t${T+fCRPq8Bbx z?h2@dd^=F_AbH;{`BAcQd>21CAUWW100ILjaU~kGf3MG8_135ySlTjIo&6~AsXNA@8NnOryzKQNWXrBZF-+eT;#^3b}WhiEXnR^XZ_Cb)&*tJca#m@ zb0-zgxC0|n#f&UTpW^)F9sjRw!34SBtI2<}jZmIn_;KT~mZjA4xI8-&HGx)pX?FV0 zh3sKbD=7~?An1wc^lplO$SV}#o;GM% zWS>Vlqnp?v?`8TtE=t!ZwLv8lvqDbA8v+e5iLW7tpk`wm|Ju|?ASb?Z<{3vQ-#LSd zH5_@-3GS$42i135b_Uok3UC>U+}IVdtZn2%du!U1qDbtb2gXVroxzk(Yxln}u?*%( zADS6|u!|Ku63CIEWLYBguYTABtjZ}+HN22lN+-jdyHIqI*UP^Zg8Y@4yb^zX_!WM( zB$s|h8RRrq=5$nX2hgB*>?@BRsjxVttg*lQBgx=4Wpd*%rqs$+-ix%ib7EvOmk2A zHVR!i|FJR(cY9mi?y?>B(Pofhg%SVf?{N=?LI>r>BM=K6)KcFPWfpygS4cMpbkf(D zmEpK>pjB0d?S{$qXOj)Q)1F%OGUXA!200R3b}_o+YJBDYo_*T0AL-ba0$aYG(l_CXD@c)jv&b(p<7MiC~ zAFZK2+C=Z6S|e*gJa+Z7oxijw&j`x-6;T@{_hfw@o2ZX2My1AWlo%AP!u`^YfSl>a zC79pqpiBJch|W7Q=Xs%x?Szb#_ZW2Nh2O&ce~e?*#k5>?<5P&nZM1`K#^kNo^^gxC}dPJkenxhv8)Iz;au z>#X%hC!o#R`8b5F=Wn`p@ZY{8wG2>-e}1WnEO+CjAsAr>tQKZb%JmdkOC_x0cLW#% zdPt{tE?pXX8G4wEvLgQBsd?&(@y+o|$U(t9S%a`a*dZ>A&F6jmm%kk&t0p?ju+P8J zr$-#K`P187w!z86P_utUebsgnEEt0DdBfxN+CA6EX|Z1 z7In~D=<-Poic^C8zHaYvX6!3y6S5q0z-h;X&43a^XO?U9RInI&-yD^Fa4 z%b>!I&xIno{a?2wkmE8Mxo?#*8?`UiLTS$NJH3(nN0Xzj=IQy5BZk4b*qYov0NLK< z`u!7ee^uMI%0Its=*EU;X`eUnYQhZ6Uit_u9v%qLi&0KW1M->Bo4h%;RE+|LT2%u` zH&l+lAJG_IF%AeiI_Y+4q9`x2ChTC?m2eZARI%T=>CdiB4!9`Qmtpx)Ho3rFMibL)xqDl)sS-Akc34qN=NVc0+?2{3523hQ_R4m#UBX%Ul=&-Yr%i9 zL0oc{-8T%9s~=hLw#fJEmWj;A`FlSjhunA~bIocZ(@ZH(QRD3isZ~^%!#aW*QWds?SL>I??>-6iHby)-hI$(@D00)0ii3q&27kog+08 zphC`rwc+lk{TyfXdN$F`^q=8O-U(#GpEo71IT5>9ACTEZYT=Rv{ifrm6tVT^MTAsGqwj zNb?D8@av#I&YX5a-sEe-AGVAXQAHF=e*gsgS(}b&SpN?b2*%Zh#^BL2QliQCKO@z*NWP@joMLWH%!6 zd5?S_g|#U*$(H)+r(YD|bveNX=B@xsu>bZb?8(6^yE)T1z`!O0GUa2S7_*Z0gm09}py&+mNKGT1OnLf#|CUz@?E!)ma(L@Ccx?7+DCsG88`*c0RSO1Dby2~S69RabfCvBfh^O0q)sM{B3v z@obJ?#N32`SIdq?)x-JY0Y`xQZQqQdn8ophp*=CB;zUvXlzpLqy~y2x$#0g0&WT$k zfOhHO6m_-#)N{|cBqY_TVCsX}z zFQ7L3_~?w)U#@s(Ewy}RCiVIIpU>)?fq!bQymR+k#+l1!Y6^b1_^Vsq-Tlt7A9c<6 zs42ctj%D3fe78-j_wE!Vigx-Rmt?p|HNXjQvCI5iFIw_iJIF?Iafm9u8+pU(QOgjP8FZMYEu5aG6h&_C$hPj!<9nnJ6xW4kc%Ag=G5cZ@PVe!Y< zayNGx76UC192V(7WHN(U7t|oW8PO`NnR!W?3uJMZOu`yiRCE;l^6ncR{>2)&QiXyQf|7YLerh zWu#f5RUK0m@s9_b2-k-%4Y)%Z8kCu2Dm|L^8U&{JU<=a*7U3_QG ztW4^oc@!Rc{QbKhw7>IoH{+X4vs%7&`KRUIt)2DJ1?j55W0NXjMzeUbcJe%6B!l@- zKn1^tTp?>!9TDBUbdUl#O^#07Lg%acI8MDC_zBoCjT!V3PJW2}C-D1kFFNUIK~>!s zs{TRdaT}7{xK;{UQ3uR^(kW#cMN&ZD2$VpNLPJ1-2wN1FzP0$x4qAiFLc`H|YHKyS z0b_fki}%0r{UdMLMk*{rJZ3yMjMpG-EpB{>oVLhyq!*_OkP8X*jjwnO<3gR?@6`Oe za`Bsd_IUOm;UGb*`~I%9zlrr1ZGRP^n`db9|?RQRs7;H&}U==G|3=Gz|49~>g=Dpivcbi*=WtTm^%6p@8`cV2(-54t@ zeW`_^MKY5*JW?YUim;Fgh8VMUPzHk;q%Qu>z_!KK)E+E3FH~+N1_yqNtmf1B^Sm&|dG_ zz}1oi{7Vd0&E-V(dLN!tM>qQJoUoqv2|pJijC7{$KOu{`g(?$48-#2Huemy&{4tZeO&RWz}eQUa~pPAXxIY+g_X z|K_wK>UwE<(7EYF2=FMq+9vS?*Vfa}yX->R(>xAguWIXa< zih17{bu1<`&b(b2wlxsX?F+p*{pPf+kPBY<&Lm;s0;p#$gl)@tkYfi_q0gBQrdYtU zXX0NMkU}?3pn@>vfUt8PrQA)CGAf}tRu_71dc8W2KRD)aRI_{|`^a%<*X(gEFAb@YCrA-YZd?T-=(Th(h zoo8=5f6Mm5!p_Gz?J{n#py%f4L>8y=r_#}XB}?7dsT5c_mGzWzEmo)}0EevNPuJ<@ z*gW+$??2;SQ#`6a^^M;sxy!rd-yE+SdoN&z|2kEtBE{)!6qW}$`7k$l zEcqx_G|mDK^|D2kWW!(qiyNm)LA7+iIJA{g0)}){0!FY&qxI9FuFUfey=_{Hvcjvu zGX)wtFL>RP!KyB1QrAUZ3GZPph;vAbtS&5tmnzxrWn%g}OzwHZT)KrpLP%3F+6HH? z;6AlDcMLs4)6=b|PjlV|JBEH3{&9(|J9$kK?<~#-E4r~J07XtM>g}6-c7V8Ti|pKV zO)0GRbcnlRcKCOa{pumqA#yVwci?b)zN1%DTcSUkX>ljfADwI^72NWy?%R+=TC5z+ zF-m!aB6UDuFvU2oRal~YNEgyO=vp#gFi)N`Ez`S-J`c@1SUjYkaLrpUJ_D`?rF?2Z znV|-h;*v-&9~xi7o8y~!twL-a*v3Z+IP<;Lz6K_pnD474c&0pbtr7zvd`|v^mSCM% zEu?6c`rVReFxbz5&vLv9k~^NBx*z*C90b(we>rKYZ56EB2DL2maC<|#K|VVNs#r7q zKAYU&*~QNS{w=NQu^=tBm%klm(MBMC?2qo^15L3sRdCr~Pybd+{T8~(J_`lrWooO1rzyqCe?%ydn_xk-&G z^NDt8V=T@^@dw;AwK*Q!dCg}^=@q01s)EW$on-c!fXA+(rXP}7EI&zmUNhHgxS7Rm zDHs1yw&@+2Wt2*lOg-rj$@Jok=Ai(`*KFoNs`HNtNQ>go9yi7TQ`^ z^V_CK74#4+jc<`H5Zz!vOu1rQ<@j~OLDuCMcDW*v`}Lf#J3?O6bNW>udP@9_bc;-@ z?o)uKZ4qo@m1?;h2sj>jf1*AqTNPOr+V8Ph`q^8_Vg11c5(CrYof5E-zbIC#TEp8D z>dJw~{?dX$A`I@v=Wqu%oP74dj>Az;T#4t5&15^bkk^e%3+k;LPZg!CprC6o0jr~d z#5$r;jtk8_-h1M-svClp;ROAEuD@sP?F3%JwKr6XFms z>6l|01b%}a;(#!lOOqudc(3|slpcgFf2X14&qVQKu1Cvd}`mpwyXcvo^2v*{)+oe|EpCwzzp2&E4NE845F?8)xuXm;tXao76BZhKgwm z7e9))Y~y3z4t+*5DHdyR!zA!xU|UP2l}I9GMFnc&A&prfWgE3eJ}G*j^S}okHhWYOMtxT7hC4|SjMqmULdnc+F+Sr zHy*9FS&dd3DW#So>!^eji4L05u?cJkuSKaTko0=*p^Ic0{0dbo|B%Aun36;&-fdC^ zLelkvVTICRNaZ<-gqIj{$fwyo(=QKvHapCMk&=!bNkr$y)24$~cqpfo(DSncrcDrs zWH8tGbs<@jGQ}MlYMU_B#nkQrD1pJwj4eONl4M7opbG+V4?>kDUIv5pQ5S=61sovQ zE`~xL#&H^afL3*s9F&w!egJv8a?t^TPt7PDr)SiT`vQOj~d* zW*B~%1OL~#y#Imsxq6%1`*+(+%-zA_f*fz;T zdw|!O{SAuYVXj6SR5B4N!wn1yrQ(;MGfLrgh_CUJNSAQ&R8xadj{2&9qa3GLkib** zDbixEP1UqTr0^C}=6_%4rZ}RB#o;)1_(fo3N7G;d?%OLqK5EM$=(Y%BAverm%E-<@ zP^iP4NsH`GG|ZMWm{Re1kokB>XGs7(JG^!{Q6lO^uOH8GgvQGr=d|;jpfPf){I2$^ zmZ526*9SMrK{uW?U$q*BPEpDe6gmEc=IyGu4w7cs;D{hk-zeW7)}-nom?AR^=xUCr zwdy&7UjDi1tESD1ObN(@YfsC1Ap4di0B@zI+Z4wGu$T%`alv}f)a>`biS#DrRYfap zCQLEW#pS9u`0FMv_0{n=glg63bBw^Hf~13e9!U0+6=D&f0)@?80Y+JNB)3T+M_f7P zTb|s-0d?GP^}%=E`}-IRu8h+oE|GO^jH^9XxGJQS(6^dPCG3c86EByT|1?)wRD{=0 zNt|>53M;nyUYfj9y2`tV50w~G^7uL423k8Mi%wR8JQs)dhB%tQ_I8);b25SZuEDe&T=YfT+`rofFD!ZfUjD5R(+W?ixLkdD zOeur2=35~~6Fu|iiia@1Xf8_o6s0i-qz08{XYlqn4XP#>fn1Jaj^KSVZn$a={q)Dc z4ZO;4U;N8>%V;z2Ph*yntidMO?u&5(#R~%_*xM*&5e1X(1k5w%MWzZC^42QvdmjQ> zIs@?Jm&Rc6Zhmk|cy{D91{W4jhikU@;BsF+@6Onx1ZPLMB-rb7JOGil!5=494|zS_ zE<%)}D0&Jf*m!DsCvLE#=qF$FU1D35$)#7ojgvJjwWU}63q_ci+Z@`U=we!wn*dFB zl;!*zf+b!Z^qrZ_rXxp{9T5eRYF4k5d9Fv9>FtFNg4_{yw_7H{#lj}80D7H)NyZj6drE8nu4QkGGqgi0t7 z0kb}oDs3EBFU9|G-^>iAlWfE05jsVJX%z!gF)mA@3q%sJZfG~c0w+*_ksypZ=qBEB zImR|~M2&LrOA|kVc9go1TOo}=us%1$sbh}a?q^_(!m@2;`yuWu_|nI_eqe#lxqmNx zNSfSupy{*%gMm_lveZQ?Avpkt7|gf#do1y)i%EX7-y>U48e1$zAt%(Q$(Q1SVgXFv zm&O8ptxp3a%Weed=nK+je!oXLEJ0#ME;%5d*F!FXzz6oAY4GObL^~F9ZaT>F%?~;a z>+2azGa#=;wor63NOO&Uihfd@F>l(YiI*onh6vb96tsE5(SYQD`9Wzh4RoStsV_bR zVOKyp#P~Pp_0g&_M!z&db|!FG@WgWX#mM3|ZvEk{Fa6n8<<4zq)hzORs06bNn$KkLwK$@QS^rEUd9T=<8Z=-AM8Uf#CG=?TN#F>GH9Qd7AU$^I~kr*DUZg z>yTYhYi@;H^+$Q3D@*}uJm~0jo^E2p)XH&qu~yX>-Y?cotPsGeq;6tc1Rl)fRnbpG zw$@JQk0@~BGL(a0VsRnPn`N_uY5(uqzGqwQ@S23gEOiiAewi4(dlD*vV*hNv2U2O9 z2rYBey*_$So5})dAtQ`iPK{w+aj3NKlfnlmDeG!_pWCK-H9 zK9ftg3bV&;7?(#kLCKP)RdpREI2nT0pgxbs0cQii>mWNnwp8B=neUDBZexfBMPE*M zCdpg{?;%1PQSTCXvqQV(?ZcV2VjgZg*07K#pNhkd!~&67$EViQ+${9JG2GZ?f0tPJ z#Sdk-|6rrdAO7bDej%38rf$}XJtT))YR-)perl`+&7G8T2Stjhg!%FLq>kRHLrq$2rfd&9tT^<6{FyE-!B)#xKg%OK-56U?aycL}KR_v~DWazmj5O*SR!+?rZR!%KS3i=Xq?9IGfZ6=k z2L#-0U8x(67!fKoyQXbzQd9wb9j1!aLisjw@cIbfa~b_lXd zH3;WHTEE`|yV7n1R8B19p_@Rh)qW3@c*iV3uMe(5&66*e7rBHa)oe(6%>RO+#bs#V z~Bw_cEDE?8W{F$0Lnim?}W& z2$&>Gm-11o8=C6ZlG4D`iK%2>qz2lBrsDbg^Z|imqldlb8wyJ7{SHU2Tq1YnI9vwAI8Tl2yp6t?*b%fPKI#5$&{2{!SV4jt z=jR|q84$Cqpp?rgvXn~LqN*MXrys< z2>B2o3;!gch9_Emc26=zuS@y){+oS@qo4%0K$5S{lANTEf3a$GazL)Sfkvu9tqQLP zT9=zJq5-e2mr-b19>CByFA6!As}#u-A?Pg>6-E|D;%yC3xT}L^0eEH?5Hc8}QUyEw z@k#?QFe6Vxcm~rd#PImObiL=ph%6vqD2+)22AM8?B{3CrF7-Mg&Z4iV3nVq6cYwxO zvn)1WR7c}V1N4mRg}SGm1onZBZ4`6CaSLpfe>h0lmQcEF;Q~uhBq+ARC}Wz`Bu4l8 zT#iI`TyrHplF<%a4RUsz!usRqYH!vDIr#*q0A}P^&B;H6TU^qjjLz+3D>s+q##5xD zRuOUqrG$xeIS`G-mrPpuM!u>!s6k#OG=NL$g`$NUf?D5e@kl>+g)vJEoe8QEHpidv zx$B+fRl_WfKczSs*ynLtcrvg_en5IMu$NyZ#V(T4@IuuIShH@9ua_dX`})YM5q0$J zCmkrWodZa&XtTqgI0bh-1@+_Jx8P^iCtW`xM_#dsbh{Nw&QZ!U6ltOoP=&FZP9oh< z=8{D2{{Q5C30za@xxPm{A^9+5Baj?{5y`-?I6^Q~#71Z9GF|Sy?Y+0#^me&Jdu!W& zJGV`3XS#rj;s&UI0vbRzSzK6LP*ykG1qOr>R|L^_-tS$WhrX(8)0TVQ3~u#oVd`W<0eyU!Z&?EW@{G&MTv4p}Mr4Cxgq zd@8>;y-jN83dk!dP)Fd&xkkkV z5xiOh2TP}PJomA~VT=tpb-Fl2JD&W218=}t@&mgSTNp#vJ$SGUGzN9M=|q99U3SD- zhr9DLfuAZfB)B<80+Wpi4$HoT+*vrJnlHhiOzcOoMxB6zUib`IqRl_ z*x?l?TQd1y&Nb=H%Xyz}2ykGJm!)X8Z{8WNc*#0<|{BfOiCv(D9(%xii3UYvC}>OXd7(dWrwuU|U-g zE1@)EL!SJ#F`U?;^k%v5TWX^VtoVy=GubhPoHFq$Ybd6YA{8Uic`Q^NRP2QU#dPHz z=Z%bRKPW+UDeED`fv52oZM^UZI!Q5}e>$KkVASa%(H173-VZ?usG)_%WURDpa88n7 zXhYW>wmAUXH_9-8FY5qxnDI}M%_il^^x)CBr-*%0#_$y!H+r^~e_SRpA}C9ka)rb@ zaN-Hr#71Qv^%Rp%k&RRoR(!)|W>|`GSoMk^l`xqPZX zCl(92hnyM|W#Zu?sd6Bc>h-)ePq*47Q;pHzcG6U zH>%mx(!<0Z?XunoWV_1`+91L44)VKX=q@8y43>;vm+wpF2b8))a~oU!FJbm9+sF>H%83EYNLNG`S}zH``6}4KT*stT$xtW$wB;XUFzids6&KZnleJs=NZFM*Zl&DyP??zV~tbn-q$K6t)GJO1r} zVdAaD59}MC5_}QB&CXFH+sHg_TJ=to(5kaRS zbH-uW8OUvy(G_!-0wu*sRhuk&!R7gR{JwC#s*dg#r+C|HuAetGoU!n{0c~`h9SdV@ z5UGXHe|zWoj1UgI2C*%J^XTUmCOFO%EB{2oU3a3aOTUvUenIxX!iNz#yHL6G#PGOg$O#*)(%%{u+C2^xeQ_j4bW5&>E}C)1B&|fzD09Sq$Q?URp74zE|K&dy zt+=6g8@Ad&bQN5w!RQRCj$Zd1=L|pjPAxyO6cv__Xq!9do((NEM}@=I*09jW54c|8 zEf;JKNDX)(#^h!n)H&45!7@?%uNmgEVBtKQKW6vWocxl>bHA+KKG(S7xZXdSO!jbF zaU3`d++wog)Kd&_`qqsI{+0TlgzEGI0kxq1i1OP=KtJSUXlcOW^Wo?n-muB}WJC!h zTRX&udBfb1rM!zgT&uDnsf9#(Lr#}HuPV#MNedGAx>Bh1>nDYhPUXi7wt#DlqQ>`K z`oe3-a`$6?ABitXcPPqLW%I`YR||2+zO+2^iuLdm2H)8#G$Lt1>xP45knP`G9g9v(0fnID&)Y z=+@u9>b~D^UN(Z`!8`J0WXl%@9N@=}0!JyuKwhAjis}o;;tW*w#O@b#K`T||uoc2H zI!p(l6m2!XK#C-2y0zhWM~Nrif%kPma$T3L#%d3QQMPB@ymarXumgdGqG+2pSaAdm zyZx{pA7eiG_o600xe+wSzwynBWSs-YTXvejU<<_nCFdq8Duy{DDfc`o8~NL;&6#rN z&68i<{>}U;yWVd5UeD`o+7^iB-v?Q(7~XNs{(xBTYVm=<2{0MfP5YQLWz&0~BmG8v?zE`%J9sPtbe(5nU_t3$G-oltz&OBbt; zEAVV}KpJSMl}a$&w$3?E6XSoI#^4*avZcw&6!^Ly8V8d2nSv&|(`O^uA8>x=fj}UY z0k)agz-6?ARNiT6va(#-Eo`EX(rYCL0-^PQU&CvFPl0Ev z1P{WIRu7lAF5za)X-&HKatQjJQXL~*&?nO&#LtGG9N&8vD5>jJCG>&7RLEJ^({=8h zN@&CL&2;V$y#dAW@gC_OZQ5pevJ%5Wsl2;N{NyOZ{sIof}Q*WFI%j$AJ;kX0qZPrx-{%9HOFzCAmPG3H6!us%H1CydkF!VXnr&tPf>W z(j?zRpt#C|QkpW)8YpeK5Q06O7{J5bJRak-_|FCLXFT0-j{_F$Abl1VRp6M=9yMnZ zh@%fVA;Z=hCR2^iLj)*`zarvOS+A=FVoa>K-~cmrXXsg1Sl{{C7g{BEyx%o~X?{(l znp_#HD$s$qIcFI}1}NzzYNPOD;; zf^aQp((+F)=qFGLn8aW0zLu}|JuhC%&wnY2uj>L1;%fdH<_sUVTRL1D&yz>41)8P+ zyaQ8MaRk9K3m0b1=nxu4FOu#Lz&`So zuj(;lYB)J^0wjD;_zO8z@J;dJLj0_H2t6h%k1Aq3c0%8}Ap-r^MQKpR)e-svE_SbE zOMAaz#M}o3`)-h<+zgllqw$&v8rvuaB1TPARI<{rbJrE{kbwB43JC|033Sc0-B8k} z+aM{FVD;ayh9z#>&?aH#0+d4ffyeEUdxJbzQ{dMfw(nKF3IJO!$q2!Nin(#5UEauR z0fiCdDfmSAF*sYgVkitVT&i*zxd{W>W#E53Ak8kiW26KM4>x;aU1V9@oCW#_zsVJmD(UplifCN7SvRArj zQ4bWvCo8+@hYH-_>kdqx(7Xh<`L}55Y;qI3*A%IXo@&mr#6nNBEo_)7d}qMD*-$19 zJT%Z;tGnr59|#6L1A4jbSiv&Zm+<0!*JJEv9M|dm-oO0EeDlT`mi`oj$Js8c)bxVT zz+Hd>GNWOve={@+RD^BOB=~;n^%=jHU^IWievsS%f|hkGnA!du`;YO(0!UJ4goklW zX@9l#J(A?W;h<8J^&^L3AnmEAq6}Fmy=FKMg@U)q%Bqk-5LX%CgKG<3i5lz4^(wqC z%e4;#pUT3qS#Xp05t1md^NyT5en-3m+hcb0h@9{{>1$ViIG{7aE;# zt3;DX;^N5+Y^@$qQ7FFBrp=>LBs&AJz-mBwRk26XAiv{WPY((be7AeVIHw6)fi4#7 zJku9^5?bxnrac^(MoQ%evWbckx#tB{ecocOAg%{&?p3@MD1ed#p`TK(}|zuV465Y~UG_U~jVx2=@}!*h#?WlN%%wG@d5hW%Ii z!po&ML+bzx>Ao@4LxQjJUCII)uvZ-0zv zXTTQfv-!(rr~9S}^{PHart@`ZuEgY%fp{gG->JlN5ZOarj((3a1*)}U{^_w-uJ4f3 z*|3H7pY@?#qy>UtUS0sHr z7u=;t<2{P(@lJ5r@0Te!E9(`+@vyuC+3{=XLVhhBD>@p`A}l=Z*6p&b}R<^{KY<&wSt2FtyqE5d+qg5*H zDOpVN9e6D}Y_gW^p_p=tlt2JE_@3g+O(L~ zJ|L-ft9M=v&h^S^L*iEPYL`TT;Ykc8XR7=7NazL9&7@4+rrqkc&wm4h?Z9V=t%Xu8 zLXKsgvQ5Fm2|1K@%eVgMX+%!*&vq{)sSb>sJtoL0pqM<0WK&U({hA`qiPP1~r({!` ze^L95-Sgw7paiPkEg9CFIxnc?lypmw#<RQUmplL>P;`iICcgaQv zHYdAH%*j@Yfw=G%DylivFp1%Wk_Ke7o-tqv%p0mA7K;u{@1a5c z?WKBu;EjSx+J)~QSHL$<8=8;o6TTSh3*$NgV~t^tbDSo?acdmsU-c6W8u$4ZzCGn_ z@|hEH*bPpqi4j>vF>w@$rJ}a|?Dm431=&Bp{+3-$!81|x9GE`yS$kgO2GfC`&UjU2 zTvD_reo{%YxJ6tX7w4qL1Un@Z18zu>p&lY#eOj|v1l%Bp1B>Z88I}khS44|G^wi5B zJt+jP*-nrb>sBN>565${s%9gvBB06zr8SrMW8UEbRFGBEhhFOo2Zjm5y|%{Ua}YMN z=sf?gjp26U@1HuGH(|2$I<_<0+*38%+(-Um;RiAmE|o`%RttJ~muAKV{^>(HO9DKq z>PBd|$1k;9+C0bB-{sR^ftBCx(*xP~45#0(;|4tWOFl_8=UIF@G3~gYA56vQChpI< z7l&YrQat~v@`L-}!nO$eBKsla-{qa8sdC8g6 z)?NjomhJG%zBB*60#_+~?t%D`bPw>U#`FF>GNcVxCab^mF_(O^FK!$!*?+ZTl?FKM z?%{XK7r$rTfXyYM=D@x!OKY~FAA7*HUG`Y1tES_5`Sd1_H8UHWtHoF8Bwrgs*AjGD z4sT0Fv)w)4-ueBFGmR!?!Igs5r0@%~=Q(O(8TU~P1og|Qs4hBI(C7{VaXFgZUU{HN zLF+Z0K$ccDy~+hQHu>~M|6XDU1SKmMPpzI;1> z`of?&KA*lI#c1qV?{2RusPo?g!dghSf&S}(0Azka`J*Ht-LOB#EC-dDn_)RVthzI% zuIIJMM&D!U^5Db3U&0Az4h)mcCZ0+H#jK{t3gk^dE?;Eu)@8eu@UD8En>WD*#xPNq zAi}CC9p^*;lNI>~^H99Qt|%-y0whRj(`Nf!n{zqh#>{i`Qd~2n9l~==y6UE|KQunD zECR{pGW-_pWA~8B3?tX#Z6E1*X0T=3-?sWalHpo}q|%~TzbE}ybH;?tM2pl(gCDvc;rw`q&{ zy>yCqr7BiXqWoBy!mk4Q-Ym@tkYvM=mruFsnI-8X1ztyd;-*+@gRH~T(+FcdW2HIa zhNsiIseD%>o+{pYyNRqCt5e>A7n=eTjHOe|Mha#o3Y*9_dv*xp$@Q>qx(xmg{XEDe zm;cgIx$XopbR;IvIuKB0*Lud$lgJq{qaU{8F*{-|rf%h#7tnt}d|8%q`>K#`us^km z*q}DZV57d{!eYu=6EU7Svj{BXc>^n-{lkxMdH-|ytyM0&NFl^RKpK|Om3UrovHzxJ z%(<}y5o11R=aUV8%0Fw~PU^6JD@%hY=2xzU*3(x#yTTfT{lIK`W9CB=>%T^gQqS#5 zYO9N_nsdi)cwgdsJ?gU=26B5`L zkTG(MlS^Lt?QL_Ny)Ve3!lKGIKgQUtTx5JuzWI845+#X3mt`S74fW;iw4 z8%Q^dTst%@EPg?1b6yjhebqMkf}ae+=|}Cj>2>|8DNW{adM?(+asSJpt9UfMb_f%gOlgDsG0+S*2=936`K!D)dL5((6;QbgomxBn zdtL9kc58t%p%iMUp4`%?dU8{V=N1_bVe|bF2h2RZ(2P1`W5`UJ6X2pYS0mwKOdJ?9 zEZQXlbZQ{>e_}}i78x869;NH0o1EhUE2txaZPF|d7%uzIt=~_1{qcWq_|F#YA%cWF z%cnrlcV3;K_!aCIZPOl)r~z_~_3%`u;()Y4c!+;Z*#xmz-UZA0a-f=} zJ4~;s6ylW*VT1T=*fw&6`@Nafwq`xfILFr=OTHz>gP{h1Uw}Y2;%;PW}e|N8*Q~GI!Kn!b+HG z@zDtAgp(CT#Dr%zvrie%X4*7IFA)`-|SgzKirJ4OMFi z$_C&d94>&vT@se+@AkrzCqSaFW$qp46U=!KE=mndB_|jZhd@DxYT+Ha5M&!|zGgjH zYsNSGPsXwJ5hss+azp&V^zjt+J8;f`rKle%=VD(;4b&wDRm@FR_J`NQVgeCAe31j? zPeV?~VuIep&&?DW7o+#erv90%=C&9) z@CFOW<3=q;no!M~F!7lZ!V>;YG7u1=3!=6coh-t5!Pxo-J%FWKmIE*L4i zZ-g~WSZ;}PdVBN~+VlACsmu5Oxl%}t@|vl7&!)Ngb9afBgFCs!cSochR~z=pv&3Ss zKW)D&AN}O`FBcjuMw!oP1?m36Sd3T`X1IqGbB`i-Ar{Hcr*DN~M~ALgj?#!}(3{+> zDwITvmW!^-Zz@3{wwf*qD3+nFnc@GKf5f#0=`@g2{mB(nusX@R61v&FC;(N2khikH z51)o0@fN0QI*1o33xL|GnxEiX6!7H1<)WL)C$IL%4ozR@elrv_u@*MUt3wjqW7K$z zZ2_ozg#Pqm5n|cUOmIs+tg4lx8F1?cH);*DLvi}mLKhu=oRJT`WmvP#VK+{Jyn83! zG9vHcSHAi)a=?MBXD*wpK1~#Jh9W1ij83nqm6dt+@k_9#q(lBccIR!8YD^2LM4~s zJ-FLAOAOSgXpXgqoGlqAN12-K(wMI`f~rgXQX1JmRtb#*@3}uR0on;2Lw;5B zR`?}FtcWO-Tmc1{B;Wi0asAD$keop^ z-5aan!SWy{)J)cV>FsyN)4Sxru^g7(r79N`p)eGN?2|SFB~%yG3q#ls*P;Sxm#i$j zPm!+1PNm06ybX6oSUr+H0cC(zgNg+VwmSWC(W&2?6B2SsgE}xwSa=nO0-_^zmGh$` z69rZ9Z$l;4eV4v)6gZEgx&SfuS&dkxj0-YYhOo=&KP>r9pw#HSD*o|CGf8mZ4fZw@ ze>8()Hc=#%ifSCLx4Y?;0}TW1S#r&6!VR3kZcp}Ss_B+)g?4+NrmX2RT^y!en}^&@wfr}fGh z5hnAug>8UnMWG~7unj7pbR$|+#_Gx>-%eErjh%_QzVM2?f!fnI2OFKAzd&h29CZ5?vTfg;V*0ZjpA zu$1iJXY!D^UsoZ$2@6LZuRF9+j)AyQ=^`s_Hbl&InAu&Zam$(q;sC%S(W0i1o742FV}UY6$2#ATr@-$5G? zJAyR3|7C~rCqMn+kIkJKhuvGVxF-!!mq4YpSI|4JRFML@F6(@9X+1w5DANX{$0Hs^ zWJ)L0JK=%?yU+aSH(*D>n0sh#&q-h7QdK z5TdIN!GK+rAc>DlC!VU6)xuCHE@6A{*=WTajSZU|zxrlkgLy81CA=OToT-M8w_2xn z0|8|3>m&UgoxB5q)ghT`bTfyX(6`A{7m2nom{A|?-bL${1ZTnq_8~!*unXc*ymF)55zD^` z8%@J=ycf$q&&Zvee&>$+GhNj5$#>0bCmeQZVri&tm+cQIfm~73^eREMxQp)cw+d+^ zGi&`{OR#GF=+-~U39FM!KiYiq72}%ZB+=|4xl_nN6Jx%MVj$MFgNi~;=6K0bMYXt7 z`5AqiR0(w5AcC5!Nf&j|2LdvA4X(}JRnws~7H$Ea1-d6}$O)7AP$&zJwyCa2Q~Z*r z>d>Ay10i4o-Kp#e)2r}y+s9BXnV#p^iIL0J&&3!|a+@B9f zF1kBxbHIV=3o#dvz+X8h-MuVq4Fh4_vT(ek57{7axMXFAaHo5wpfa!-*zGcCi1Ibc z3njz*bb4W%Dp8c{21J`XA`kO&=*8rcw9sWCmfjqX*bPOY*b&^IC=-`EALPS;hLer* zTfTZ1=;dsb<6GhNvhX~0K7CxH!#{iEx^muCUV$dT3wwsk!gqVY&rhRE%tDlaL! zX&_(IlzKXB5e=tb#H2sYi`?yI1m4dtkax*iZceo0h84iZjxG<)q8L~L)1g-Z>ZY(n zxW^4r+Sla;Jj`t;D@*+oAyy0>rJy$fsm?(~2K|w^I;72OseF@jxp*R0MJsLW7*^4? zr*{0iKw{1*;jn%V3%eQi7GNlNx7RjUMpwDzX>7mivE*B~V2&*r#^(2MvKo^c0^a=_ z&uBG%ckIDYvSh59D#!KQNtTIiSWhvD6j?(>6^Tkcvmy#Si>dC=3h5n}4!4WE0^WAl z3Gd2<`LzHk!yc8*q41mm*^x3s`n&EIeLs}vTsgappY68Y~fiX`p1aTP5xlo>zWsx!uvo#k*r9DRR@{8b{Q62S893*h+9ai#RO&PnPASSeEaz1@b6dlbA<<(r2b|IJIcy9E_OdEpiKxD(HljuG7u{yEV+uKC@-^2`OeIAsfRWUr6B@EQ zgh?-@3XcEV*0;9Ocj?7mn|zz6*MdY?79ZOus_A!oXX?;9HwE#e0|gxp|MI zu|MqbZ_{2_a7l{VaQEJLq7ir5bL8%_R=yxRC6=@_il-&{4s*w>cF9y5 z%0cZ9fT!omB1k;_nBC9(Z2Qg+%_~{BK*E9jV3vZMe0ry6#xQBTfv>fj?!q+T%Bh_` zxsV#%?74gHX{n)d7u(&t=!&_W%D&J6*Jk(P09_Y-9(GeXK@Wi+7*xcOpA{29&8h*{ z9{R9i8-r>FAcF&4`kQ2@rTbphA%#fI9AK_8l;I%#aI&&eRqkVGrnd@=7NKd>3S!Nn zHPE~C!>*hcop|cOt6kGGBv^U=@vCV}J$;lO-auoL7b~)mNv=bv>_n{Vj0|-r5T#NO4<4HfIF;4OXVnCkgUwrZ1%*e6TJ#~6o#C(x;4R* zD1Bqg#m73}JcT6Vi6;|3-Xm_nsj8d*syVj;7f07|ttJ~YKFScJ$#v`fdqVq^XWfs@ zIW0Zwu2)qGt7&}I{;litvAwOz1$)23k8*x&Uvu!8u!sJpn^ba5AV}R^g*V zQp8pE2>sdyU*kp7)6f6UX=?F5m*zMohzdo8q6X(eQJX3~@)7K5V5x$x5Pl2LkLA;y zK#tkSFQyOjo5agpYh5q;=~aV@z0L#tql!$4Zcwo{;zLi}so-eQYIl6mAvE+W+NPO;#y>l6!V0x~D}d-U;}==+Ld!D7LK zSx=v-r#pm(vx!Q*2Jh@so{<;ZzZKEiiMM7lxje&ZRig6huKI{o7-JlB0nC{9azjyZCjat_-rL$J>QmxBZQ3)E zs*q^W1A;06E0muok@zQ`SLZYS*&g$&W;w|L87w2%=QKBDB!2prYe7c)^8N>%m1OH! z1-1^HTs&f8bM{iqE{c>=QFrJI;#|o1bO@tG#}%Ih*3z4(72mr)AM@uu^nmMT&n8X3 z`@5Nfc-|e)cR}WD)*;{`L)%jJ_OLoio3`ZZb@K;YtLepp6XJHJ6_^Y0-guyvz(=cG z9w|Q-+YWLZ;53FkdFqtSu*D9i72heVmKdRvB}}zWcDU^g9uDs2L7QztG9W@In;5KQipi}V(S`N{eKk|wf9EF0SPURV|3fGHa@AX8~OM2;TQc!sb z*d6e{gH9bTIrBOdDE^FlQ9oEeTvBS&RJG;-l=sV|w zUfFI1s=>gGvYVk;I@t{r5A`%sJv0upt1pF3PcB4JI#sPO`8(!cRhIEU+eT?;#TvtG zSvT>v)+l-!Ihcf8L)muZ>TwwKm=RL!_KZHo+((@4nO zsYFLMLxOkQ6dZAn^~$9$d)${71P?l;%wF%kkr{Nt=hEGK6@yMOk-H_20tcP&u3heR zUV~2hmtx`HC&Idgwq@J37HLl{sd(aPz*v3e)~m~ZHy+I+2QGSH(L6d3kSnSu!;Gic zsvIxD{V-d9LP`KjbHODb54CSW&?CFwptqx7pb(9tTdW#(*Uc zZdSdK_s>;TS<1wS+O)ufDbhpYag*~V_uUaOE^XSabh%W|E0Z*OH+rWA^!se3PX&MG zli?rFJLu7>I8Ev;s-CsvJ7>SgN|dkC3dGsNA^8y+ak}val0KS{mIG1%;9Uw>)*1EXxzwQQKwRibUs& z5tXwWJq}5sQa4%zS!hYDAXZ?)o*$4^UdzVv(J4moZ`{9MxGz(KcF zCj32*DQ18o50O|oS(yh^>L*ppAgl)2oZzD*U0v?eME8?g&0+WD?wC?8_gTYW*loyZ zATo&$S!Sq!p}Q2z;6_~sRLEj=|1RG7FC(qkhF2B~YBfo|5I*whQ|dMblz8^gH9%h2 zr|b*I=+FvfIuQD9QZEBts%BmhT|y(Z;r-AC#YT^7p@Rxtrn=Mg(hd;-jH z@Am3c-W4wA?SOPtmvE(fLdX|`#%FTeH&x+tjreoDe>9ow`NBkKT1+A|^%MiuS#{8_ z=YN;JB)#w2rrjOUN>&T5kZtaHPzPS-`5EaIZ1Tv1W+N+&6#D{c&UimIfw;R;PWB}0<@`@LVfPizYcU72yR31xTkV+3(R)_D|q1ZFC zgGQ?{zRCXl+hsWz2N+mJw9RSu&`!niexZ4miHq~>!0r)CPO4C}+e^>e70|4T7S&9{ zuIpwMD%xNP=Rv3rZkMH~cLiW6)^S;MzVQ?Nqy@$mY4)NIJ4h9` z70H1uOuNa7bc$k*QxN-$!fuClS%DO)$eUG(0wg3HaE*@aq6;NyvYa3=J2{$8`uR-QZ&?aB!_DG85;s657Y@?roQNAu9 zajk+jrgB5Ucrm!pxUlRHjD}+S3Pp+8X$H^hQ6{APtXZL zfvFN+JA;OPyYIfRPH)|51D5Z9JVB!g>~To zIV@Q}^@`>QIj*Rt2L*M$H)fR}^8#cY*H1mc=m8<+AkwIxS>>{O?p@bzg??s&?^2LP zvOi)tpk%a1VK%3>f5{CcYxF6{%r`_Y)V`Tz+hgdXPxET`*2OtPOFcLy`{;Y&Lrx!w zduWunz@p7zA7ryduE5KtZ7 zz+{ni0z)80hfE0_fCgyw9&kOuR88OLf1M=xc0mJCqM(%rhDRpEHki+W zC5SM8W+Gm5Ld4{-MfIiT+MO@d0>h%`xl+-|gCa{Wy{4DG=lWPV93|=TeiW!zVX=!Y zUeYMPAZ~&7p=S5B66|%|BU>~9II;vaW_H6~ro#k`?5J5L_&D4=bnmb$3`>|Ex88ac zHpf?}%EB=W0&z1#jlz|nQ~^fNuDb644#i8pdo;S@;n;k}oNGZk^gQgn7q<=)!yqx4 zLR-IJhs0P*OTo;hzl}%o!+|5SERr8h;%q*KPBC5z?FeoIZXh&so+K+*yT?QKkbztE zwEJoI4t@u}kB>?;Nb>YgY-HZ67yq02o|j8{$APPKS$4?i$v%K$gcD?i@(6iA?+j>w zs(}W1j8DAe4k_?!ncGV5h-?RF9Om_VbSm?t(IU7G%)0i6iIsN8)@J6Z@x10;bp9f` zB#pu#V`6a%i-dMc_#`;oYje*c;_kGkW$)UP|Zm|i)x zJ2;Q3bMI7QaY>2i5ua>(@UgH>DmT=yjbi)rt8%8cj0YPITtdKtjf$|M0O0k{r}Y|R zdI_yZyR0|j4l3E-79({}F;z_2e@QV;CpT=^DKat!w2BLo2hF3Y4!f0MiJW%R!%Bz1 zno}qmaE1CuEU=Cz*TZt@;f9X=nr&f_3GP!2Cwm*^na+)26EN{yr`)i|9b=Ace2RS$ z8DkSOS)KLOQu8KeE(?nT8xNLV>@GvK0Gf$*S>iMu?q9H2AU>oXdJsVf7E5ta1{MSI zNRFH4kF2;CHHm^;O@UvBTO)ti@?f|Y-BuJW%Ctjz$2xpH6-lu%x+YC3csuOoG-CmY z-*2nlB^w>MfMmCcbFr0Taw)QfiaIY&o>3E&OE*mK3PXO%1fOQFHtiMZT_s|mL-jF~ zs;10XB8mqQ)?M@J!ZK&HX;-^n4Qmy3gH-&A$i>0C=GoQ6jA!y~KI5YXt=>yD6W@l|k!P;pt;scdk) zHV0mW4242!$c7z5T(I-hqBG_R+u-Wae zvwMo)L%J5)7>dKILY6~~-}SIeL5BoL+CUtvU)n3cbNd52Bn28&hB~R*8E}O3DKn*d zRf>1FWM{xl!71QuN%2mD_6XEi`9ODm!Ik+t-YWZ%t*yV!zHZxm%b$kG`zy8q&)<9(R^Y%LafX3q$+krE7jrt>!Y0u_hYq0Xf zXxj@ZP;T_a*ju({9r?zq#zm+i=+DbZkpu66j+uaXKgH~&$Sx|XI%Kt1nrtT@5;cCq z4V`JSg-vv(0NPA7UCMe$4eQQ^bb(8ejIM7ng-s23?zJE6;a)uzmt?lOr3TofE7l{U&C3QZl5L11E)9TDmfM+dXcEb_gGmGI6`|kN~Mme?9p1bFY2;X5ak&xAWfm zH4q>|Z^$W<&#Q88RaE#i`h5Coy`)+D>#JYsc)R2){crcZ9rN0+u@wISsex$BlZX4> zZh33>Yv<+<(1m_akAY{KwNC~y+}onH#$)4z8WR)4;f1A88Fe4K_4)Uo|M5M_xK2HG zdwiM)4lS7cky|J9 zFZ<+=5$>NPZ41mZuQYJjd=N{aqfr|so9yt&hQeqa2G-hT*XCf?ITX1Khgy!RG6h&_ zfRux4BGMx#BJgGbLdLQKc1H^*giKCpd1Z&dXflFQ|LrJQ_JT2H=a`t06pC3-kwnl- z1qs|XZPJ`>O>y`JNd?p;^^(iddOD6*>6J8R*=%bTgmFxrN(kvUaC5eK4QaFV?gP1-Xh^&Pwg=FQxTjMiA zmLp~?EAyOlRv=~M5q89kwE%s)Vb7{Sqj@RqDOpVN$1*PtTpD}W#JudGm~x7gU`8<~ ztWX3RK<5`M`nz6L!dp29lgj(ThL=t_w>;e7;YC zD2L?J`9X2+z(1i~@80>c9yl)gq)^c!=~kefhz>4<{6oKN!WJa1P#E6?TmCdV6vi+t z)ZRj`|1ei8d!e?@M@YeK8`2YK`@3@nUtTq_=7N`yCQa;pG|g7A*W4# zsI?RWg=YJxs4l2!$L0s5FG4nHATT-(VsLo62$IYC$kN~rVY;d`xXJ~2L7V8c5ogF6 z<_xKUQ2UTmnHY^f4$vht>~M#ikbVX$(F&nM0<%oWYJ?YYY@`^lGAE#fMf*XE_K|#>N7CHtkbB`> zicbP^z+-Fx!*Wy+Ey#F|IZP%eE91DCmt#LF{G-~4DaBuOo5>Cb&Pbm!0ecO_R8pjZipp~B zr7J)WW(~0Nof6gtY@VK=+~#r$0y7u=@`A6;?$+FqWk}ADyMS}7qiECC$_%^esu}wO zwvm(k7||feY(|SV1#~L^x{jYH_}e=E3KzZVF8$c^uAuehW`1=*qM%oh3fb#*B$YYB z=K^6iRM`dEY;VTlJU3|n<4R1nc_)Cw`oJu;{k6h+I$6EbEkT*>wmobsy;xKdk>;@l zD2!u74T=QcI-hEBk9dRRqbyxPDYzk{OPYgAkhwNbMz)QCy`>-P-G<)g?&V^fo-a# zey26M^~xAg9=(kG-9EGa%n7m#63-mME*CjLVp8o6ugU*h3U{5aHNcY9rcEV1UKLc5 zYtFnQf+W}SnU%ur^gc~1u%8#ej;+vV6>q7$-?c$rkEPqflo=BMG0Pxj?q}HrMJ9vl zUP@eM&fxWQsLp}+JvYJ&(zx1wThvht!ROH!x7E08o&_`RcumK(|Z&kI7C``m@vSn z>p*`X7tJjY_cc%e407J%*4`|zxaGm8ytM;4-4j*cZ6^W%Q-M`SQ=YS zsqW40r8n^ID{jmxS3DlpgTEVm3KZm*1|xs%113(?7e458dBzI=K4_R)M~Wvxhu3jA zNN2tLk@tTaZOb;7*{_hx4!pk^G_e+UD5i%ZA5&476>5}YU4Jr&KVfRHTHL}k3cKjO z$oqwwhi zD(&q#E=HuiuRHq(62}c`j+<~J(4;rYaoR{R$rM>fMH$3CfZ(If2N|S}I1f2By02jh zq1bBMjMJ~S%Z@mAO;OiL!ce{}lCjX^Q6~`mS(Y_bE0}sxKH_0vtySNoVH3->}Nk)@B^cl zlJd8&yg|->VZ4-kCSJ-lin&UW%T!d6I#Zpi`E>Sn@p&lw&DB7CXMD&BCOtCQC0{ly z8hwV}sanirc&0$LPPFJD!F{|@hINM`QP2;i=ZS(if~_lg>L#&nm0Nw}7EpRleR&9W z)t7<7;#%lVqQfn};jsePtv|V?EF4SMk)3G3HB&O=)D8U$h6=~siY=hzmQNRnY6y0A z#|94hHLIFcnER@H=~ievQ0yhR?9@EVoNUB}S~7c23W5!pNlh zAAA48+{fXPgK*%DHVc<;)9dNqU-H(b1-YOwpX)gQoXri>bxX(=z*BWdH@$x<#AM`I zGdhG=HjFs4Ka_BOt$-6JSXg`Wh5zx&_~+UzUG}W^N18*}3;8ZpmxXsKH+df+i2@)u zL>9ziYHx5Oa19|7G)8}Qi9y)Mi|)nkFU5nZk$v%E1x`l15Njb|(R085bJK6XW%Pw# zJJ|OFa(oK8ZsJ6qqnK8TG*eO7S9%Q!Wrx`!!E)agquMq|BSZ3M!RFVYEA^zf!8u>F z8h)FQD25;@hQ9FVtoKbXUADB7hkDDKHBc{?K@adBeW?i;D+?w0K@HAGKZ&I5t00|R z9fE1>F} z!&zilu9@>Jm(lR$jAK?tX2-g}&zT!ei-F_DHNN(>DR1}~!S$Fqv4ZHiZL}OXQx0|K zqcR2CDW-@bTd62i5??Cl3*Y8`Msk*Uo;ZcQw3*9gW|)d2QNe9>)K=gokDK zX4lgfrH_CbvNCLya{biP@JqNH_GAP*8Kv>#{y6})^71_${Z(_<_Y30IusF&ONE=f` zm%wHTBX_zKw+7fh74yIb04!3 zBhMXoyf?g@u;}aN>BkqU29zbqi2EWG5V-GBO_%%hDVM4B8iNknd1BD;tezmB);L34 z05IN3IKE*&UcNP`3J7SyrHfg8gk+9ys zI$$M(rFofx6n>9a3O`ko1^K1o@I1O!bOpMk;sWoA3wX7Gtw00bBd(lX#;b!?7CWsl zR$$1?li~si&i9s0`uRF@HZv{@jpO=UKwm!EGs&PBV8=@ZGCXjI@4A2+johocVt$(! z#%PwgUMGd3GqkRR*Ut3I4hf)y3oBE=pEeXd*nOelVr5Ja!ggJo3B7^CUknkR;7YI5Qr2X36b^mDODr195wT@-dgpaowF-vcL=Kk z(wJNt87b}w&%hStl{bYKC;I^ay+tx&ycz_nA^t*M!aJ~iVsas-FXCjWT@aRSHW55>&) zzib4@J2#rr$-c1&?i_esX)^)Laf$&+n?pe1Eyn`#)PT}p9g^7Ss$6>g@zqgsaZa2l z7bwiixMlDk^H2M?KT|IvJ*?JP-g4L88Z@6TjqJoJ=wQ z7VQ3-{c9e>$UE*2)YmyST4}_@hP8nokrZw&tpkT+fl6~!I5v-BKxHkHit44yyzaXW zL{^G3p@bpUYnQScj7SlXX0~bf%)A-g&J^&osopnRWfw_@@G2BneQ*^7EdEqL6;Xd` z4!b975Zca#c^anUGAmCHXR|Y}zbq{Lh7l+q6zsb}j!q%hOupeZifN%p6BT7BQ9ST+ zCzL*CN^sfNHS;ngHRA1J6!`c^oFB9Y%rItx47YZ;?G4^N_cRm?6=-&aL6)aiK5Vep zh8qr`I|AwjhMaPsM4(WTtlZ*TF<19!`WD|V*&XL(!KXg3F{zr4E+#M-`snbD+GV%J ztCUInEPg$G6e6i>JuWIRorG_*QFb%5EUX)7crBdBam~r|?E9G4Et@t@Hpp?)`){ZI zZk2iT)?tfVSR%DmAtezV&d*Mw#p?Bdtse&>D3VrBYFwGHXIE))yx=ht8yuLX}j2- zKLiUQ|Fr$UGnS#(s=JQw4&G!=7{diCj=S|EEc7zC<%j5sI8%_|@_1$!DD{Esjv|!{ zk=|@S=vH)xt#UI)Q|z7#GEABk=&)+GxjwXN-nR8RF)HQ2Bh|N! zhHCE*{N5oa95~B#!z8ZMPBD;0XrZE3!Vf=3+C>bEGHcyB6&ICyRhqw{ZA|yz)jmZu zDBb7=m9?_lqE=8#stl|Q9CB(_RZYu-GUZktwoU7nM)t@LecsGA7}8oPDwY|z^z+p9FiNRs7d@y*k|#k#%k=*i zc}rHIYn=AO5>RqN*QBrhV9g5iO_sx2AeNLznHVLzpva|pSj?+kRuy(!U~!1hlJC^g zdo06*rK8#IK5m#;QTNiH&2@<#785K=!%4oK!=dLK&5c=51%+H5=m_5n&YyV-GLSo6 zx^2wH@tn!B3>gOV@SLG+dYKb4CchW9X#aS~@*Fpk4eT$Yo#sr60a1%IDhf4p246Wm zPS#@lR zoiPX76aBv)FB#pHZx#LbugE3`c31Y9OnNcJz{WZcdZUzQ#f`#J^|jFbyj0Z^U>;j8 zDhlg&tpWmheErxD?t>v0~dZ+mS_X<(NMz&L_EngY)oKKfnnblgvb%x9FkqvH%pkqAXdJ)1j@gYbHo8ZO1D;|3WAPAly-flMLA;WY zyUL}H4=m#Y#Iv5h)*>wNFoLE1)zA4WbgSidF zhBD3qUM2+GS1quEzhKc^Ja_Ib8pZZEvCsV&B{!3P^G>SDJha9VlWGkZa9u4JpsQS( z1d9dt6{+4!Anm_G-r*J>vdgcYeyG?HnG4OrRW7Gg^@<|VU9qjH*;u};=gx$^U4ORE z#bwjzxV~KR1)9$@nvCBadvKI2;pU|}@OBv7rBN0C>nSFYB5SCq0?k3My)R*T_^GKI zy&?X(li%x_?N-9O>N(-o#!_&zV9;)-cl>d~EpxhUNA-ya(5#P|=WL3}q{wC}>Xhm# zBxCPOv8iA!5UxKU#>_C*EsYRT76oh%!zKU_RPU80tCN+Ls&XH_%Dzsr#cfOzZrZpl zFVWvgh$QyFXQzr^C1r>AKWofd0 z85FfahY1v(oaa@|E2gl26{8s)P)Cgfx!2}I!xA~Llf&%a2wRby!^~WHmb07denXaF#VfH&82%;_8=UuZ}v1?*To8ut`F7S23)T~Bh|`?0}@^I zR2}LmjbkCSbV(b-wpnn>`W25IS?|1D`0?|*N*vdPBP?Ad%{(M=*d^WOeos~_jF;e4 zBL%l!6&HC$e1g;k^?Du-j1RdG(jU1cQ5cC5LPW&8z`_50H9qrLhh(!?#8p!#isufg3v@$a>g}dXgPMzwi9XYcG0n zOEUBr^(o^em9x+Dnw%FyClWNrhvq@4OXb|%AR{{=1qh?RZ4g!>zxsc_*ZuJ)zx~-C zel1-_F-s^CJ(i6a^_2Y&*-UBhAtxUBhf}jg`dn&XOlb9l|AV zUXrH!rhj|Dbqn3#S|?Z%T==ssG?Kk!eH9CIQ#B|fdi;l5>|MBGy}7V2SU3TN`c351X;{&AMU^jM%a#x+MlhXm;{QfHdH|sif{wgD9@)mge|o0Vnm|z1O|9N zpMMJ>^!fc0b+GQ2|M+y85fpyEt$LSioI-Y+e4AS-28tE8P*LZ2&99~etrm>@ZPvCx z3DJE;pR&dIj`IQGdEUnF_s!q+;|6)YOPh9o!2f6OOW>Nyvb=r53(3WhjX>TjP?0FI zs4Rw7v0_!P)!oz6+jP%2(>>E)b(@)Kr@MNaOm%mC-G!D5E+8nlpeCp+f{3!Xpt8AP zX{kV=;(`l{XjxGNr6@|_J0}Sx2}JTBVV>1f^{evbEf;+EpL5SW%l~f-%@*Dx)pQD5 zC~E=f@MEI8%29ZI!@m(H-|#KTxOpFAFOkBaXTe0e$%IRXSsL=D(rQf3eEaWi zkagTv82(}(l*5BB_T!jtuNdmXpLCZXj2qbQ`gO=ZG^_Oy2*#~~V7>N!?VPaG0 z+Tg3x3Zz9dpC8q0%cY-?JFtXS)7z)Dk)oM-(=7K?%TW6lchb4j+z*qHp~U?#@%ZRB z3+-D5=SbMvCiPyzqXf0QABU z+;j2q()wn5z69P{3I_}bXJ8Z6YA#Q@1)P*nc!#Vo9dn^U6l+zm>+720aV#ciBJPYP za?WK2oSJW&y%d+xM}99SgE8)nzou+@+2f`g!_EfQ&uM_ob~e2j7)sGMXrb%pYzw_N zt3;IyBw+6M69&!~2V@L<%zY`yy{~M z9y1D@Ly+46Gw26&J&E2#S>20gzR%8+ULIFBf~3NLpl1@ruBJ$WMPmUpx}VeNFT#Ju ze3G4;Vz)9W4Vqp&!O2LBzxTk1Cm56r6(o1{HVmo&qx+;^+jR=>N-hp`y* zG}0TVFA?DwEB(5dGH*Ou3O#!U^)BE2(;BpUh3C}yLcCTDj7TPF{9o5t0eb%Y%o^Z zl)-*RvE3ATLh0J*a&aQNOoT;&MyL_gXHPLrIVfBko+mpxIbD%q*3Q1qk z;xNfP>GTTFQmO(QSidYnS>)X6x)Qs5mbH5lhu;Z2#oNW5rQZ!8U*)D7HJ{5E+ zs5$zY;sTUPBSByXo%3xo6J$K~MAal-qS!p~O3*bnLte<73cA7E2v`lJ(JlhPVJ; zqXE&-(89BWLBaJ*Me#dr_JKLx2J{_-+E$2)1G)r24SC-WxsCO8la~eF_gn4jf@$OU zI-Q(lj-!tLc1~XL_*rcS|0uE=mhjErKTVc%Gb}s~)`Euafc@15id|2UWJ~R$Sqp1B zRMR(s&J09t+kj!TC$3V{Lubyf`d=bI1mSVNSpeM?M$lVUqBq3R|%9 zWT?vizs9BCEWP#Shu-$|Y+sPvjsrnVtTln{&XLe%6C0%_UwtNIjN50MDMa_i;a(^W zd{h(}iFdDO+CKyEr;{tsb5~-(v+g`Je&hIom%X&;nC!CLpguM=hkoj_Z2JCC{Xt=g zGy|H8SN-$D1!HXooStZR&gu|jb8<1peQ(|@_1_km)_!^NycSiyPv11`H@*;9BQB=% zV=eUs&qV3KSdcnr`Rq)#Ahbc-!gN8^K@QzT+Ekx`M}k){?7LWy6zE30ISsAnR%F1@ z5SVb$sc)ZJ@lKG{AAQE2T}cew{81i9X^-0YqXiVZnI;2zou%25#`_^(dzC_y$uSUFO{iWL}-I-BHh`0U01fUn=QtMS=ffuL4!6|LE+J z3y?Ix6xkd)2N?%N`TtK`{E}(;AA5!FFV5$x_DZ66zcjg%B zhsp$^ZwEo#hf0WpP44rk@!BbGQ-R*<9`96eVlRpflk|9hs-HnqJ!2DF7*INIX`EqF z9bHMEkhef8x&$oEK42v5ziO$M1n)OFymX$0G{1!@n~XGtOQ#n|+hFwJD#DLBL=~J;3ORTUe_qgyvpuP#hF=-SO!LD5lU3I0ztMxeNAzAroJuaBHiQL4r$24 zGe=*-#|3XrniKPFqeb1N-7_ytwXSGgKUFR#JGl8qJVp?Z?F{gZiYOM;$nq)OV-*Hu z3P1sKTPTKP4#{gHKc`4 zO@pgFURhMD#39Uw9pXJf<^v;L9nv-H8jA{E#j!nd6Ky^{>IKA=?|a~vc;LGm$uXxz zfxq9#nb)TO;Ad9Y{M}8jPI8^c-gA!)oYRGiPmknHmP(srQyZn0DlF8l z%hEguX@!0K6=``S`pwm0z*Cp1saL0J48EWrbRh6kMS0`~(NblpW=-IE%~ScwuoT(S zh*oG@McpBgGgDw(^8Cc@!afgN;Phz>$lO5-8*=<4m>K1OTkL&O7=XX!;UT-`-Ss^f z_?>X#c+gjc)AwLJ37>uA_g1(uS&W1{T>U$xgRda2;zK%ofRsI56tz=3#K*lu|WX;$sQ$j>=-r2^`=!b;?*SHJLX zKx<$zZSd{!T?p&wZuyZ=3|QY0E+M$>JEBPv!0GbHHGUsROT3c9cT1iHx)wh!K+W@& zZM0As@VsNx7`-%a*;ipW;Kt6>K$`l#pT!%1tbm{tc0f(g;+{8ViRsV(O3uFM&mS10 z1II6V2*N1-n(n_#|Ka1;tEmazPp+VlDF-h;t!Sy>M zdOwhkp}B`(3KKtZ3pm4{LEdSXFNzLaiahUiI7W|SrZKh3Bw3bbi=>q{G547zlB*y( zmf$^tCBa!hJ^y-+8zaW4LU;ExKW2}IZdM(&V_Hhdo@i>Wg-sXEzNwH-VIZWv) z>EnT?)fuAGG_Xg}g}|ZrDB_&BL3>|x%Fj6Uc;GJZ!&@hJ1k}-+{I807J<=6tp=Y^b zata*x&2I79MHkIkIiW$DD>y~){zsy6>3-6ns-uyzd9hcAyg`ejc1PV)^$7RNcM6-x z2yj0bk?!~=4E`zH&+MpGzGGkf!zEhHUuz2#jRRwisT7++kt9lY4L0LI9)rQG6rmv) zO=D{CeZLD}8eQwQI5Qbe!@`-tTzF61>RET~9T+aK;IUwYrJ3FmhkrJxx5Ob`89uQgwk!t!%%M+twJO@&$W_P$$ztUwG}N{M}~e$^vXtM)ee=V8IU4r4v^~kLB9W>78FfKpTvF=ROX#6&!1H#I25<@)#O*#%oeH$kn^nbG?AUmwBSHhWhaOaKsC?O3QB6@j3L~Wck4-hVNf!!MDXM}thxK_J(d>!Y z2JXYyO!4P0p37W+{?Hg4IJOf*C6lH<{HGcA)MQ+oQ66tu91xjVs(B6S0XnUv^7{<7 zD&a;q4_cTiHI-x8ZO;Wst{~iS7KDG@b~)2t>zfNUcLg_C22*X|IO*CjLrj)Ud6;0xU!c}CmCPRKkxn1^HaefB=zXYm9uFo*i zCTVeCv34w=FcKuVfTU+wkT_m;u0M-DX-^Wx+o-w&+9y=>DO4SucT#XVG>hIlCtpz= z*A&-Gf8vX|))Lh&L6bB^)~m|!-aIv4QZBfoZ6fiLk_5K`ZptwO_mCv|8|gLd)%oW& z*}~$OXj^?o=;x}s=KSxl-qPq#_Y9X0T-i>^S_hyY{bpbqt6SUe1SgJZx9;ipiW z=f5KU+uGxt^6K>%>fb&ONB@q6`EfF@V+)yKvEYgu66ME#J=wYxX%bgfkZn9Jy*^^I z?Chafh@9q7x>_aDp|-(>7K@x)X-gUp87ZNN?kbB#_3d7T0VP3DDuw0#P}+)GL)ean zXItrb?*yt=*{RwRjr}QZQXO)IMRCQ9+mFG?qKw-jES~>{6))pE|LtRPl3Q^RkHhfI zHt4!Yv2_%=KSx_-0Z`#^mJ*M>z1Y%|VcAvvj^%cM}m>F3acK&s91zwXFQI$p4 ztCK}1BT)Mg!PreMMYbyN$tJpm!97#0-_`j&!bKQ->A#GH*GN268Usxv*j(L6%Kqur zzdwFG!8@PAf0M*XV*N%)7h%ImfvV2W@&ojEh(Fb7=>k`v(AG>1_sJe_Lw{Z}<3aqm zemk`D;2+5|9!m&pu|Y~A#jc{r3QBiIbsdO!&FPtIic?;rZ$%yYRSk32i5VVd1ovF# zWM{^|_M^Xa+6xG9*+ug>!s0+~AWKtDUkLl`Ris+zn%O$p+#+PI>=jzNh0N_dM_`RJ z(tG{0pg#z+#p&j3F=mSMWSA?4jxlJ~RO6*eaW&naLe7(|@otStWP4T3s$NxHs2&Xw zy4X+s_JHvi9aNn_+;B$1;q}pAvG01tWtr8cY!`31O_F%*+vMBW6a&SAk=aD)PKF_y zRIB(IkkXl?7hY}9mQ8?GdGZvvVeSSLMVy})Ke0h;R2KxdfSc3hnzGd4Upn_GAAU?% z+`-AnP#yp0;VFB5D&Fp~9WYLi5Yy zzZ1fv>T|?I?-#a%oz9lWOTdVWlLY(8a-SaNmOP)%_9|3;AZ^gzl0S}ZRUPpD1lTbC z^W%jF7W}6>X2um0@FLV)Z=+~m9PIrUu|-8E|^W1<%>2cuTR z7O`**%boOT3<}5}4Oa#1oUxD319X7+bY3N{fC?K zZjuz4O|}ay++x^@9xu75bcZ=~JWqJVNesPs5I058ZlC?61j?TQ`nISf!F#?dU0$X-lGknX?w@}c z^vXtaf|F6_Z%p(L4Oc%FTD=#K5RWBf8IQe}oi>&tjbhhPWGz?V}T0$pWzrKPbDiF?W>-68%l9d$8jacvA(ooPEStUx*i8~Bteu69K_G$O>mMKIeuOv!hvy5~=STLl=km{v;V?uo z!nh+_qDmvJi^*t{GT{gW~V-b;Pfiv4e@69SvC6mhnSIxO|Q^Wxo&JF0IySO34=WSDleK|I7 z7as=&)+b<o`eWdIs!{d`mq>bi+@}XIQ}I@G$;Je%}hSr^gxMc`#jbPyHq99_L4m@z^7cR?3&dv6`7)&_S3IMG$Y4PJwmpb?)K4wKyA_3Hya{{Cp!?anN}c zY2&fcS!P3M`w7K9rpO~o_m^g-7&vD8xkD|K$QveRc$t(vj6PY^$uu)7L}y9eWTcoY zARkHfJ-+?+s4=oeXr!}gph52OO%_=Ofp9FCskSK00#%R|etEM@0RRf)0hN7G5x73Pll%x*>hz5ghub3nrtuNs6^TE z%-F}UF*_EDtw{y`2FnE6`ZRVc`&!N-hG?qqec}DOc z{$8BQo^I*Cdeol$jkm6l13I^EpOe7Gwb@I5W-jy(fRN29ZBayK_$6^c%u3}FNv^!l zqX+1_i^=2Iu2_QxIEhi{@4S~$rr*MJOzjFuoUU*8S|8aYZue@2YAKM~O4XP`S2A0` z>&_3hch0p5mOgyd`y?f*zUhIYxi=N9mc3O1I9 zxq)Y{S7)PrCX%=I$3brK46rsq$xdXYrkdU5v(N9Mf3Ip6)azP^ zdN-5epdKbyo+QeFEfug(8Y3Q0E)GJCq`&rhoO}}qeP!?{Y9nRQ#XtpB>l*pm5g=+< zCuVrO41b4XME%R8KgBXu^lVHCyFxbbcyTJRK}t5oLW(kj(iN++HESl|jGL9Gp$8Vk zcED3^VNi2tQ5Mu5?7d;n==G(2naN%{55e=IqSGN#2 z4Vy7pS)PNC6Kuxig(ps&XkB+&=SSTnTRcgrjd|ZqvAZaeMd?nkN6Z!JPG1Y3 zEwl=#J|Rn^VbfCaN}_LLAf0h%H z9WFmFM&(GZazpZW_WtZ&?Ad~N>#;as23j6`U62JUVzKm~1{nqs2<9rxb}w^svr6>T z$Dm#+X!O>%c>_s9PWZkU?0QX`;BLz|VvLMr;yFCS4I`62`qgqzt9znt{hgjH_9R9d z_aud4lPI#9(v`;S4z*}}wfpN)Ob8v5v0Y)Tufe%j&*7c9=vGdzp1)4W59z-j|FZ9Z zkxZe_V|DaWVw+Vli()|#%831BSbl)}^=5O` z_Z{K3fGys8{riLNd9nk6?~@KlA!4D_u~6jA$b(?Y)%kmbk47Sb$7#+-gepWBr=P<8 z=ARbVE)iRC^4jH%$H^)lN8=$nGoYk9onkjqBo%b;fg#}zWc)UVVW#gADP+o6DA0{A zk+%Tz(K3L=Q!xcJG4J~($xw}Zm8wP1sVWlO5VU}1-2W~!zx>0`ey@A;S3lGJ@y%cV_E$PN z&3X`u8Y)NA`tH;}gjvl;k*RPA$>Fxl@YnYyCvA+)L5hW3U?HVjCQ4VlFUb{D(-^zQ z9qt~GenP53C{n5ypmx+}Q7b_q3e;H5iiCY07#QwVrOdqm>l22CFU~87DG1Jmm8fj8 z9%Wh+p(GgjTPzP4)UJ%AaI{1NothzzM`(72cMU9d)(>s{S8UHt=d zfm6S+=zrFoeVW&^Yl@x1)9NbVX2(132;<{+Lz5Ab=JvP6r^MFK`+_Wl z$vM#Ly34l?a+I*&HgBdY<`~o#R#Mkkr;*^KA1@C6bP0SMo1Op8`~AE2T6DZM8o-e} zVB52mVmBjOmaZ*ksn0#AvR||aOVDsr+d*4(Zl#~dJ7e1b7SJ7reG|YD2-XBCUGx>w0x6n; zF-{zC1&90KU^_A=aE#wR-nZL6wB{ftWt5>PvHqIk*kp_Fj{cL7%y2a1B zxWe;b#2Qh8Cwd>BkYdQ|Yzf9-Y=Qq`&=E>wGUu0uZkKI|-Wzv@siT)Fp*2I61)=(F z0l1gJr|^p9$tLkJP<1t^o53(Hv~-S+U5yvVR_~hX-@hZZZg*C0x%KZP zaj5MMk72Xh#(~{NvFQ}qNa?yr1H|bwH7TMiQcLFMzG$`gU3rFgA@sE$6dKg=5o?u5 zvT{MX)b&7cFoB%$;b0tBpLUFo{}fl>^049~|I=T-L*jXy!vUSK0h^g*icO@*DoTeu zZV=YEH1R&DY+ySw_71!JyYIxDI5quFjN|HC98doc8<_ExvQtmyTOm>Zk4v*iDUTs> z*#;8tQ|u{<9H(^9Ks>cglLb6oX-xa1vjjIS_anAM?-80}^cc)&i^5aoQt0FZBJMtq zGn(pm5bt)I`r^d-kz$M!md1VE_l?g) zRxE{Y{{CsQoX1$ow87E_id|2UWJ-55d|zm-A`*4+@JTcdKy|e3*V5 z%Z-S7r0@On=Qsi5^QYW@zzrDPKbd>LUY4D=K{O~58L;l8QfvxEk|^CR#in`qUxT(- z+U;FTH)uBnUX3^}&ybI0xqx+MorQ(_fnfU+IKhJY#w(Xzf7u&6voy!1Ho`sz`MmNZ zz~hBscXPFTlB~f0>im_WW))CYtII?!khMh)rPK10B7L@|WUfJ-CqWXF5w9Ay?|#Tn z&H`n~HDt&UU2~EXP{vR5s=H{f5XD>j;Xv~#Nmi@Gp0GmtwjfDx#|KmS=C~dv`7<@h zGMO4|`X=UO3WD1eEzAZ*9z-G?B-^J%k|OFRspPAbcv|pGj z*y?pdUIWrH$X?MNkTSPYd@`&#s+!(5IVU_plNvi}SU8%#(Za>ixT8JK2`*HpKVj@c zDO_T*FJF-5tn^~w2A}gjY0UeQ4tmLi93V`>a}{A5pkD{CG-!{!aZ{cn$n!fh*KP2z zn`X{vvU}cLAKAKNKex~A@ODXZkaXHPw;0MBpG4_5MDGth8MaecIu8odl4Kn;l9p_p zS{#H6R--gw_D(pja>~`zT!{ z@GuqJBu!zhufYWSU@V~Y>hTck$xO{EUa8u64a~839l06 z3JO()0gJBAH-=QImW6l1g~wAZgO$o!!B@fP8q}z`3@W2BD+4Ud=Z6EQb2P`?%u3`k z(J}~d$F54>bo)Zr7 z|Lo`-8X33r<;qrtnJWok7ggewKF{S9k1j*ZXy7r#v)mAK^;;^HY!ob(04ui|f2p>pTIFeV7aSXDYzEMy z@L>y+34Lj~<_3Cvw$Ed`UkmW!uJEoABZDcHZniLqbGO6+uZ&`yuRcR`38G5tC$x$2 z@35OLW&h1(c>UbQx#l4FCl&Fu6@gtp2tG@e@OZJyuvzTZQS4fZtf6$4!j*?aU#~_! zA1u&9(%Jus@)akic?iSd-cy{+()gS=eg5>ar=*ibdt$mtrz%ZYrq(}V3Pbj){wDz8 zB(Jr5XMf1iIYDCFIQqxU_K@Hbdf{<7oP(MMv?}|A*9A}z*C9^>&!%Edg4-L7hngv; zY|2m%y7}Cff4V>1&$@K{uxi_7vYy9FN1@Hqv6Et9E!jrt>c8IO6Cd};rv>sv*olz_ z8k$?cvoMo@CJSn4EYipI9ZD8*HSux!B@(2)M~MxSYiOq(Pm)vfKC~C$cu*Hy;bu5y zmd$LouRq{oIC$*hI;d7SHn|o2k4{E^|CJKK)%lewED+b1hwflOGXboKiFVn(*_qko z0wZU}bJaaN{-Y|$9vHmcEjyU%n@sQ2yr8nV#?UscvHk1H!QC08nA@h9$620Qn^ovI#a2?JjM6=hZDET;AFBJL zne$HwA$15^QAcN+fMvD~KGh2jc-IA$OvNikdTsCl?^Bu-Va}w>s)zK#g?D4tk<;=_ zI$5||aYtJb7B4w4_s+y!^Ev`{g+6$t*P|lr5&hu%9}0`+H3jts;pa^Hz}%LZ9m?C9 z#ggn%+afOJW$4Aq!CB{^?! z-J$vx=2Jy^WW09*b$r2Eq5fo8K8*o}YzTFg1U*ri82v7&dDq8NH{}^3gF3^j+*glb z(o9XU3Zr2u1VU!H(^~1)5GY3hah>wWT{N)Idu2n=0Vzkp`H&q3-eCNpn*^nyyCpXS z>56Sr_6eUrDLs_0G|}hOm9VSLnY2_1l;!PSX-vA{iFar0W1kd}z7^8V=ln8b3>t{M z-I$x~)g){Z8|l^<%O?0}7*=P0M#F=)PASfpbVTS1Ydlnx7~4g{hDoF_rj6>Mui5?Uy10ftEINNxq51aym09tstKOf>R; z<;s&q9dmO9J-)Z16NBH2z@Rbq$Kl<0Jx`JxZmyk?)`O32q%Tc;MDKfL3F#5mDDHS8 z3#khhO!pWhhjzy43*6BC+kgIM%VaBvG>I!K$Tl8>=!gx7_E0SJJm*llR{Fjl?v~JO zfd~9v_!X$hD-3A#?(--VTm#uC$mt@Kf?(xZim@tMzT&}E`n zQ0Z)=bh~^nMYckiy+gV1HreL8KWHHa?YBtMn3UMhKtu|S*G}JS5Du=97R}2fCrA!` ziM}}RGKBTZV4Y49R0*(FubtL62ke&I1p5XIxN!wiT#z>fZVo#;MZa8IOK*mc{ZBz9 zMjC^IVh!MNO$%5t$W!?5oP>FX{a#p)@mO5vr|I84|^C!M~`|H>M z(ImDEZ%`kdodwy88oCBF^z1P13UoSfusEK!Mf5MqE~VNttn+sF=79P=hrTxp|LgPE zEji0>iCcscLOs5ixJ1XS4eS*DYd0gd@4>a7xXrd}fA)(;<7A7*eK-1pU(U1IBA+Lx z)5t*{uU!{ytV=b;LPbRtrK^!vO49w4MaJN2vehdg%%Hvwg77E9t_nMRDoDD2hM+;4 zE7>TknNlBf0^FW0fu;}>h)lJD8~zy- zuO1cHyV4Fk);hYIWYU)*$(tT>azT**JD7L*o(}C~QrMMDDb)nqrOmI@Yd`+cP5DJ_ zuSa^w;n$wub(KC5SW4vvZV;AI`{kwbdhIIM(^>9I9&jI2V1|}bMtLDquiXcAR2Q|Q zntNk35$<;&U{y`<|_;JIGTk+!91A+Lb9>1#= zX@w?njv#YtrTs!Z?Asja4V=+2(jUa}IZo){?{0m(*NTbujtrOl1dpSQ4&0s$L4sed zpNX!UwjwN1`&e;i?wYvUk=;Ibl`TMb1IQO%4eMoGVAq@m6eox~J?iLhT>gj83hf#E zK|*jqfo&SauA|6WN{7Ws`$E&0dOtkcMXw0iFRP1&QqrIqY(4s2lGf|SL zntp!$Tts|uy*4G{!=Q5!7q#1c@?$gU9wF9iR?`L3H-{EQG-%t%Bf4Ci8}?z)y2u9Y zs#(B-Im7ljELC3)YV!@+WYPu!n{}gU#2pcs1_JAa$0LUoH!yA09Qm(TtYDh<>#~Jp z(@^|D{B^L%0ULnrqS!2o>_Ekb4AC9oVo8%2xelKbrM`{nxGXx6#b$+c!PB3&Ey$vq zqKad%E*KfRv9kB=JV~{nTj5rDs?ot`YmVeLPj)pKV2|)Z@ZhC6cyVC}z@_P6nf_SVi7S z=ef5?Z#1YFZe~Yw-7zYLU>F)#bZ*?cloct@e4m{sy&l{K;O%aoY_-8v62-2jNCKrp zvQ#rXAOVlZW9^b#=XQAbj0OZFe^&HX!9Ps10^&D&Ppl<7x#f@e>#RdGci@h#m}2)) zWDlj=Olnm7G|2+U2F-7lc0q)87kyfAN)<0Dm)3|kOevS1p1U~YA#lo6Dtc90Ll1>P zfKd2Yv0auFmm;&jb8T={P*qR`JhgGU>!CF~YF_B7^j{n~JRCeJn4G`Vej~&sc*5hX zql0aeMKr?P16L@xENG)M>9r6w>0t__c`_^M7=1D}DQpoGNhm6aE5$$@Oah!92|gT; z=DI5z_tgH2eQ=Gp8)XOKG7DvY8i?$gVtPH+2A>cCBlJ;C2U#ZCE-Z(>G^DZn-z*`b zkkeQHB-6Zkx^>0zkZBH*EFL?n$8FY-{S;e7k-Zr8Yfu?!V+7Xk0Wo_^z>*NGY{6ZM zIn}>DGEZWH&cSZ7hsDBzG{)TAD1ua#yp>$`2P~C`rn9I3P#kNe74PRStf6zE`p=bX z$w;h3|I#t&PuCd6I}iSmJp019u3KzSmPoOyD6)dmA))HB>E)67 zG-g-yzA@Cnvzg=J(PMM_n0}a3HQEhPG+6I zai|{{zws^m;E015#IwLh82l=3k@7z?(sRT%95ZP zPz-(G2F;PsJy7A{O3Q6G!6Qcs0>*AI7k>PLAm!I@N1hptEFzEh!VY8+TU9BMD@7P^ z>lALEf-NVBY<$Flx!1#v#{!wZ_dT|h%ADUBTdytqKP9gfy|LljRX@15;G(w7`w0ki z8q^nL2Zgy%{8$PC0w+ZkQ*gAT$Qqw2QL^Y0l1yA(Eit+;fQ=b-A^gXUi+%6VQ+vr( zPEIj@!)kkMd=n$ZLY!?gh@O$VP%Me!4=v2r=)FRuT}=<^qVr?HN~SRfL(4;}L-bhF zZ4t4;_D7Vra-(%L&*+E|;!->e9<1x}w~FnvISvveRcsc$0T^ocLyqiR1d1)(lH=zU zo=@ZU&xltfc=md%@NDp|q96Fw_}!Ezcp5{BrFdV5{Mob)`qIDLl)LpckN6$CbrIsY z59P11mhFlAi?4N~^YccskfeNJLb7{pjLCM21rCNSlrAe~t?=EO^4}f&`faGK{C$IV z?}BEg0c!A*qb~*6spb9OPHE0vUGUpcp!&iM&hIgs_RM2J7 z2N1|IPCXvj;??2PM4t&Ol9?d#oh~Q}KsBhGNe^QlD?p!JhW8lM4=1h}J%#bmZ(^ix z`uUt^|ZxDt!8g^S_0tNXMeki4wu1J7_gD2jLGDM{d z%m+!47J9>nB+=hQW0!9N1&_7TP;>xYL@mr&C?8J`F@i^v9`XqOZ(;fFg}z;s5wBIE zC?j+SfJ(UKdY<>9MJU)>t4spA2fSu!ZtI%WF?BPGWb@d_>p-$%^yq*7Js9!P5MLEe z=pV<(WBR6ALDuzC<#Mv)3$t=T!~THs&mxM2r7R!#HX%iLD_~nd3v)H>vf#$FKBydT zVSr&ymL&L$Y?yLz=FaFffmw7Bw2jmY@N*8mmn8c2NLS8n&|*nLm7r3b6qzKz`%Z>o zc(_J*Wt2@ILrmnzaz%roH002uIvFXw6ge8fMgB&ffM{qSXx38f8j7r>bQZbid$VAt zv_vw-Vy*+H${C|G@cywnK9vNZQ*6#|zf zikG-`FE_-TIs*%{VHo0Bcb>Zb;Mfd-b#?j8`(3BWd)!tR9v2;Kw=op!DK?oRiLft{ z-6jVDk#wXD>Sq&yM`{&YM_--4Wy~1=pUxd-^$frqsk^xxf9k)pop-d-xtX&CPc3cVkEH*w~H+iiOVRI!d>~^TXHneeb}6 zMG2m1%pC};ETNj2D+`{en#5&4$o=tI=zl+_&KFh*_C@!4+>$qmS9%(P@A~(AcgO1) zuT}o=qhH+p{?<2gzPaufZC_sHxk9upV3pt;ps-1d;=eUYLvVfM3DJ(=IyysyXL0mw zx<}OunvN?y*HXVpquOH9e_Z#Kb)M%uKV8)J$+#~!1mpR9A>4!v zShdP*_!)S)fAAB;GJIjcc{zgBxdi?g!^;9fCqSnB`kyi;TH)0?Kk6pg!VO+Lwp^t) z@Y+qWyC{-H>2AzzQ9ll8P@K>-PP{;j5VkbVxhdWWLKMvbU9(E(l}wA5#7i*A*r3e< zl9rRA4!TOT+P9n3g+3CMOW!9QP(XEF^VE%5kWr$=iFqH{EsNHkDhsrt<);_MT_ow; z(86OM`H&4-@+cPSHZv_jL1g^SlVRZR?BvpU_aWeI7V|T!@!e*yYYq8&G6cA{mJq|CieYW+TtW`Q(Cf;TKwfRSwhhTs~7ey=bC2%aqTi7iYjs-V;h zEcM>zQt%2v8Bjs3eLda(5-IaO|d6KaU z`f#>Q!`Kv8U6$odJ^8m)QG?knWTmSLY7RgmO@lg9Q*M@;g}5M^u59RyHK-4$b^wDS zE^Ux-a(?$T@_FHm5Vx&Ywtt6b(trAvr*-wxw*F2}7V~)ZGTJyVDHNMTk=2wgU)1Dl zq%Q{c_^kIk6`CY_>T^?g*ZoxuwlC8B$*EZ$;yHJp;uPi?U$ymi=F6^4o#6RCse>Af zWuA?p2O^ex-jc(Y9U52DqKkvD6s$>%!d8{g3*DxA=g10fdutv?Vt|=rK%j3U#imjuh0-0L*)Sz< zdLo-cCq-_Ua@X&O><*OW83KzZB6oL3z-srWA8cls1CDH-bN7eajVW{OTTH$nTNwwP zGM3UN6g9~3UL1W_P#$?GqE@XhfyylXb4r(5p#NT}?D56>GBr8$Zpj+IRdb3}d*&ed zt~*T-2c!&Mirfg7@Z<^qAhFtz-xW{zBU$4~fHZx;mv1Y@ZbnO?E0?X0-J!WDuTfUh zrin>{GVk_@x23n4oFEhgH>jV+nI;xNZxT{1x${-pybZ4J)#hP0e&A%u`D=*9tKSHm zV}*y2`PV$MpT~Qe4{dOJl46fhq=M45(Zzuobn>g^k!^JME0wPt53E;O&KwWi>Qw^u z#!JHzB>MNr4%o%wP4$6I;&{oWXuNU9%<37(11&9f`bu#%1mm8{t3*5Hk9;bopm6!X z+g5wGF!7SE*wuowq+l{AumqSwIwD5*UEm7v!Pdj!nNh)?6TrvGzVzV(do#mZ{sjm2 zuNiXI?=~~>)Dxi4&JY}&(F#iT+updl;O5^OziLq5ls}aNm4DE~`DLPX$>=8C`I+Gi z59fy-)t#K+L7l#T#!qQAGiSd2cQ?p7Zk{ZUvqZr0Ghi{@L9q~d*$S0Zz%I6N`Vt7R zwL(p}IWg1NW8g|vcA>LMWYOxB68AkV#Cih0U8L&|DkMuitA z9$r-T!||g@CGt2v?;w@<=d)~KNGY^6=Fn$UJG6R?$RfQwo-U4E?VIS2Ay^arP7AZo z`~R7Ij=+tFX?>O8@T6sddC*^gC`*PM51fv1k3||OSOzRgPE0-bTYMf#KU7#Piu956 zEt1U5qVU%xBLy}VC4*w2sym(1We70KctBMK)d%+@uCckohcRVSP6w|p z2eE|pivmy}=FZo9JzC(ug0GZ$-~aMa@wR|?Nu?$!vQkq`f2u5jC){gAoWs?C_2l`X z>%lZ&#QV#>@qxV%J#V)(4y5I8gW7mbc#T&(2nLl+9&?qc=Wo-QNq&C!*d1{)9^*Z$ zFUSI{PDlNl2NsddUl^z3ppDa!OR?D$*-7crqPP2W$F32j2H%UR5yj8mID1z}L2!c> z3MOOk%rP-|wn6(|1n!;x(hLMGH{>6SZ^|EulKoD_l#93drB2zQDPQoG`%H*4$a-!u z1|7K#YQMVw_y6T@#mcprmw!i6xnYIJyBVM)9I)lhq*&-CGEh3?$*-8Ctj z127bpB#0lApsJJ5wOuI20KVLy8~Q!3O8eD?w+L~tUSvVV^7+uq;5lT{&!)AiHqLD# z%RJ*HiP3nx+_ToJ+rJyw(r?Yp5#E$<_3soK)FrAXQP|DpK7YbG)gyulfX3ZO;RF+E z@|;2+#_EJ@ObNR}Ht;yTS7Nj5Wm7CDpk+|Hp15*x2dR%~r#t9_0G7WtGk<-ex~6^i zi7FeK)*G}Z!?pzsx}^Xb53gyD!vFCmH^OD4K(QDY_rZl5P(Jw4zieA*1vdB*6=xpq) zzyuC>t{9;2RY8M=F`#0CzC)R=u+Rd@HUa%0WOlOX(s__6jBNJb#6q*-HN{G13v7xO zi*Djlmjf~~I8+*g)zt4pbWIVRjPP6Qo$R4U)yluap6&DOSwxC+WfEB13dGGzxAP zn&`8BkA-6tjt7&%dB)vmhHOUWW|Pz@!58fVa1P>JSnE|WN3Yjo-SS;|ZD3{O5$Kb0 zquICldp-Z)7V{O(88hZebgR~}hG(iJEo`dQRU#_eCS0q>2|_iey)?mpx;1H6sd|HnxW&G)U) z$e?<@P0sT;)b_CrW^YjJHHtJ)x>gnPS#-zV@mU#{`ub*|I04p(a0t^wFL?B3k}lrs z-y}8!7lmihdgQ3+i%JsZfNU;gqLAn!37nKi-`Gr${sJ%QEf2-d#z!z7VdiXVKxUNK z^+1Qw=h5SPShFVv_%hO%`=a%c#j$NssBUvCv=ieXx8+d10F{mUdUYa;U8!ki_1)rF z`~gj3O9%vi<6*u=oRB}rg&EC>7k_uSA%EPLO_BDQLN4wLkM}+f5}t{3b3(Snp%Yvb zzCC2S@7>u&ij!fx-N>@!aK4;_ki*B_b|L(DLxa7%GH+vO4g`}8aYYf=6q`f!_x&D4 ztbruYn6OL?ni6NGc+f**aDtPS7$5rcaS!cDD!w3A!U3fPR+F^CT7v(g({)XeD?cLw z+It7g_wJiAXAopvP2BfO3^8)ksV~QU&FY$b^7U`LNse(#?(lfyd&|a=siW8n6#0PC zJ(Ayu=_Ki-ChYkUXhaD=G$`O<(X@Mhs z5=ge!tCK|-bJ?OUia_4J`=UHhXt^)SicE?`GFqfAcp%S$g=~3bffOkGl#Noe^k*h8 zj($dO*0w^Q*o|o&0ev2rcg8lcG{#7GGIi23s-41Cpg}gumPQu}^$n2u0#d)2ZJ`#a z2kdGFia+2@Zei-xNunemfO{5IDA3Ig z+2eQ^xP#qsu3%|gvRApZ8EWdAdC4jH z+TVp*jZ?Vs-4v3|%_HLRUiyTMM|6NiJE{tNz>sW}v} zH57@9#~iKB7BIO7lP?&R8(jBX$)5D4b(MMgpg=)xd|_6Zr#7q11B$&*k-JEajWv#K zn!C{UgQAC>v1`0DqAlEAP(JxSDGWgVY|F_srf@+4Z4A8)j&zOUj&~WP5AqjuGOggN z8o>`XG54W@^6Gp&E-y`>O?7@^pGS&sc`$h1f@5I@b&aZ6S}E$3o6`#Qf*xPYYGy;} zWDbq2gJ%Es1hG7^JQ|-c-&v_>om{It5SSfp&_E7cRT5;rvs0BW$fnCfpG5U}pfe7M zb0EV#4_@{l{5xw+hgp|Kec2wV>x+yn_R7*eRWcT2iRxwM1c31$Dk%c6Td zU|OUm=Jy|q7YlZ1P10S_r%2bV?Ngeg8bBcCrX01u3+c|Oy{e6}gEP*ObjgQ7r+`^y zy=c(P*}t)oeIEV%kL>3T_Y^1dME&T$AMCK_Y;r)2)XG#7Otir_xI>;1{dm?Mc8MgJ z9GkJAyCxD|4TP7YydYOPpqbzdnuUkrfxdVqLxVOcgvL zl{^mM+^|`XYAE&`Mb1ExzqpCSPco=6#@i}>2Cd3P0S`i;p-51vDuztc_9?fbky&&@ zT<`RH?Pr4Bk~`W*6Kd&7a*8~rGuTRUQQM$JJr4|`^o3rc^JIl|4^su|s$GIsMPAU2 z0A$rq7Iphnk`noysI|eL24YLfM%i9rfiz3dSSTj7A<8lp?;nv(;*NFhi_1v9 z0-Uf;g?)H1aFP{L-`V|(Uy`k!hu?70T2FK+TdxRo~prnRwVM?a<_<;&bk^r|@?OqRw ziM|q1G^-Zku&B$D7IBAx!tQMW7+}tmbqcpnLDiMqX-&ShLVc&Pwgc-k5-{65!wK2rwsl;2&0g_^ zxBjpL4VSFw8%$R8O-U2oB;7ARO;^A!sZHD&w^3FFDaY*C{;2`b!S4U)V0VKCMojb zRAAf)KBxhLjdSV@uk{m5z$#Y-d+i)LS(M?b&4%hJlhl&FA90m~%fU1VXCp(``QG2}uxIAy z?H1Vq*M6ya^JAo2Ap!z93M3!g?$Zetq5())AXC@`X+b!|5*K7jNtsv{1Kma4Bo64)clllyB(fRNW}=XKi1hDZTbO0j z*GD3eLZ0jzFy#U$Zphbzba$$(be=^**;Q-F3+82H5P89XF1o}GBC5tGK_S*fsrvPG zr6iNvqQqmT22_Ry?3YU@wvYmjeqE)aR|V^jV3Xp0)JEB3#RlQc87-}93zH$Qq;G?WkU@>RYFv?!;p2AHViAgCqUMZSn(s#jvO#N% z%R_my+t2;#2UgfHt3$p;E^vblk9`@F4Qy^w>~)G